diff --git a/README.md b/README.md index f670a3f..115e0a0 100644 --- a/README.md +++ b/README.md @@ -51,7 +51,6 @@ You can check the containers on DockerHub: https://hub.docker.com/r/kyuz0/amd-st | `rocm-6.4.2-rocwmma` | ROCm 6.4.2 (HIP) + ROCWMMA | ROCm with ROCWMMA enabled for improved flash attention on RDNA3+/CDNA. | | `rocm-6.4.3` | ROCm 6.4.3 (HIP) + hipBLASLt* | Latest stable ROCm. Great for BF16 models. Occasional crashes possible. | | `rocm-6.4.3-rocwmma` | ROCm 6.4.3 (HIP) + ROCWMMA + hipBLASLt* | ROCm with ROCWMMA enabled for improved flash attention on RDNA3+/CDNA. | -| `rocm-7beta` | ROCm 7.0 Beta (HIP) + hipBLASLt* | Latest ROCm beta. No real gain for Llama.cpp. Same model limits as 6.4.2. | | `rocm-7rc` | ROCm 7.0 RC (HIP) + hipBLASLt* | Release candidate for ROCm 7.0. Same behavior as beta. | | `rocm-7rc-rocwmma` | ROCm 7.0 RC (HIP) + ROCWMMA + hipBLASLt* | Release candidate for ROCm 7.0, with hipBLASLt and ROCWMMA for improved flash attention on RDNA3+/CDNA | diff --git a/benchmark/generate_results.json.py b/benchmark/generate_results.json.py index 4a0937a..c01279c 100644 --- a/benchmark/generate_results.json.py +++ b/benchmark/generate_results.json.py @@ -235,7 +235,7 @@ for path in sorted(glob.glob(os.path.join(RESULTS_DIR, "*.log"))): # Meta meta = { "generated_at": time.strftime("%Y-%m-%dT%H:%M:%SZ", time.gmtime()), - "os_kernel": "Fedora 42 — Linux fedora 6.16.0-264.vanilla.fc42.x86_64 (2025-07-28)", + "os_kernel": "Fedora 42 — Linux 6.15.9-201.fc42.x86_64 (Sat Aug 2 11:37:34 UTC 2025)", "llamacpp_builds": [{"hash": h, "number": n} for (h, n) in sorted(builds)], "environments": sorted(envs), "notes": "pp512 = prompt processing; tg128 = text generation; t/s = tokens/second", diff --git a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log index f7d6678..3e888bb 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log @@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x19cb8050) reason :GPU Hang +Memory access fault by GPU node-1 (Agent handle: 0x275a2540) on address 0x7f3fb2c08000. Reason: Page not present or supervisor privilege. ✖ ! [rocm6_4_2-rocwmma] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 failed (exit 134) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log index b800555..694fdbe 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log @@ -2,9 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 1 | 0 | pp512 | 139.31 ± 0.13 | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 1 | 0 | tg128 | 19.97 ± 0.00 | - -build: 79c1160b (6123) +HW Exception by GPU node-1 (Agent handle: 0x25d19540) reason :GPU Hang +✖ ! [rocm6_4_2-rocwmma] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2.log index 379088a..d19e880 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 999 | 0 | pp512 | 130.07 ± 0.32 | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 999 | 0 | tg128 | 19.48 ± 0.01 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 0 | pp512 | 131.14 ± 0.28 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 0 | tg128 | 20.15 ± 0.01 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2__fa1.log index 8df6842..e6a5d48 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2__fa1.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2__fa1.log @@ -2,5 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x50e2050) reason :GPU Hang -✖ ! [rocm6_4_2] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __fa1 failed (exit 134) +| model | size | params | backend | ngl | fa | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 1 | 0 | pp512 | 104.12 ± 0.05 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 1 | 0 | tg128 | 20.35 ± 0.00 | + +build: de219279 (6181) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_3-rocwmma.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_3-rocwmma.log new file mode 100644 index 0000000..7465f25 --- /dev/null +++ b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_3-rocwmma.log @@ -0,0 +1,15 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +rocBLAS error: No hipBLASLt solution found +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 0 | pp512 | 126.62 ± 0.10 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 0 | tg128 | 19.95 ± 0.02 | + +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_3-rocwmma__fa1.log similarity index 68% rename from benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc__fa1.log rename to benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_3-rocwmma__fa1.log index 94079a7..f8e8b6b 100644 --- a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc__fa1.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_3-rocwmma__fa1.log @@ -1,10 +1,10 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 999 | 1 | 0 | pp512 | 103.63 ± 0.10 | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 999 | 1 | 0 | tg128 | 20.09 ± 0.00 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 1 | 0 | pp512 | 135.10 ± 0.35 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 1 | 0 | tg128 | 20.14 ± 0.01 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_3-rocwmma__hblt0.log similarity index 79% rename from benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2.log rename to benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_3-rocwmma__hblt0.log index e1a550e..fad8a13 100644 --- a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_3-rocwmma__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 999 | 0 | pp512 | 129.88 ± 0.57 | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 999 | 0 | tg128 | 19.43 ± 0.00 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 0 | pp512 | 130.99 ± 0.36 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 0 | tg128 | 20.14 ± 0.01 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_beta__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_3-rocwmma__hblt0__fa1.log similarity index 68% rename from benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_beta__fa1.log rename to benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_3-rocwmma__hblt0__fa1.log index d5b577e..d4132eb 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_beta__fa1.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_3-rocwmma__hblt0__fa1.log @@ -1,10 +1,10 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 999 | 1 | 0 | pp512 | 100.80 ± 0.14 | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 999 | 1 | 0 | tg128 | 20.13 ± 0.00 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 1 | 0 | pp512 | 140.15 ± 0.41 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 1 | 0 | tg128 | 20.15 ± 0.01 | -build: 79c1160b (6123) +build: 1fe00296 (6182) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_3.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_3.log new file mode 100644 index 0000000..b84b584 --- /dev/null +++ b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_3.log @@ -0,0 +1,15 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +rocBLAS error: No hipBLASLt solution found +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 0 | pp512 | 126.66 ± 0.22 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 0 | tg128 | 20.14 ± 0.00 | + +build: 1fe00296 (6182) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_3__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_3__fa1.log new file mode 100644 index 0000000..350aa44 --- /dev/null +++ b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_3__fa1.log @@ -0,0 +1,10 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +| model | size | params | backend | ngl | fa | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 1 | 0 | pp512 | 100.20 ± 0.13 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 1 | 0 | tg128 | 20.30 ± 0.01 | + +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_3__hblt0.log similarity index 63% rename from benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log rename to benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_3__hblt0.log index d97d416..ec3889b 100644 --- a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_3__hblt0.log @@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x2edd2a90) reason :GPU Hang -✖ ! [rocm6_4_2-rocwmma] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 failed (exit 134) +HW Exception by GPU node-1 (Agent handle: 0x2624d340) reason :GPU Hang +✖ ! [rocm6_4_3] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __hblt0 failed (exit 134) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_3__hblt0__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_3__hblt0__fa1.log new file mode 100644 index 0000000..2d21418 --- /dev/null +++ b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_3__hblt0__fa1.log @@ -0,0 +1,6 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +Memory access fault by GPU node-1 (Agent handle: 0x37c5d340) on address 0x7f2e3516f000. Reason: Page not present or supervisor privilege. +✖ ! [rocm6_4_3] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_beta.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_beta.log deleted file mode 100644 index 00fe3c5..0000000 --- a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 999 | 0 | pp512 | 124.50 ± 0.25 | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 999 | 0 | tg128 | 20.02 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_beta__hblt0.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_beta__hblt0.log deleted file mode 100644 index 305470f..0000000 --- a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_beta__hblt0.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 0 | pp512 | 130.22 ± 0.35 | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 0 | tg128 | 20.00 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_beta__hblt0__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_beta__hblt0__fa1.log deleted file mode 100644 index 7cca3a8..0000000 --- a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_beta__hblt0__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x1f3f20c0) reason :GPU Hang -✖ ! [rocm7_beta] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma.log index a945543..d99dfb8 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 999 | 0 | pp512 | 120.16 ± 0.21 | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 999 | 0 | tg128 | 19.96 ± 0.01 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 0 | pp512 | 117.48 ± 0.53 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 0 | tg128 | 20.11 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__fa1.log index d65e6e5..b74a931 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__fa1.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 999 | 1 | 0 | pp512 | 133.91 ± 0.57 | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 999 | 1 | 0 | tg128 | 19.94 ± 0.00 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 1 | 0 | pp512 | 126.27 ± 0.47 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 1 | 0 | tg128 | 19.86 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0.log index 10c8b0a..906c8ff 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 0 | pp512 | 129.49 ± 0.48 | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 0 | tg128 | 19.95 ± 0.00 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 0 | pp512 | 158.54 ± 0.42 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 0 | tg128 | 20.11 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log index 1d0ab1d..d51baac 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 1 | 0 | pp512 | 138.34 ± 0.27 | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 1 | 0 | tg128 | 19.90 ± 0.00 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 1 | 0 | pp512 | 166.11 ± 0.32 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 1 | 0 | tg128 | 19.83 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc.log index 9927f0b..3c47854 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 999 | 0 | pp512 | 124.65 ± 0.23 | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 999 | 0 | tg128 | 19.91 ± 0.00 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 0 | pp512 | 89.60 ± 0.20 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 0 | tg128 | 20.22 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc__fa1.log index 86f99ad..bf63f0d 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc__fa1.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 999 | 1 | 0 | pp512 | 100.90 ± 0.22 | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 999 | 1 | 0 | tg128 | 20.15 ± 0.00 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 1 | 0 | pp512 | 64.66 ± 0.16 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 1 | 0 | tg128 | 20.35 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc__hblt0.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc__hblt0.log index 441e956..a3b2c08 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc__hblt0.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc__hblt0.log @@ -2,9 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 0 | pp512 | 129.49 ± 0.14 | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 0 | tg128 | 19.88 ± 0.00 | - -build: 79c1160b (6123) +HW Exception by GPU node-1 (Agent handle: 0x1d380ea0) reason :GPU Hang +✖ ! [rocm7_rc] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __hblt0 failed (exit 134) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc__hblt0__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc__hblt0__fa1.log index 28b4354..9cc7929 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc__hblt0__fa1.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc__hblt0__fa1.log @@ -2,9 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 1 | 0 | pp512 | 103.73 ± 0.14 | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 99 | 1 | 0 | tg128 | 20.07 ± 0.00 | - -build: 79c1160b (6123) +Memory access fault by GPU node-1 (Agent handle: 0x4a0fea0) on address 0x7f3bf796f000. Reason: Page not present or supervisor privilege. +✖ ! [rocm7_rc] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk.log index 15c9127..d8fa4d5 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | Vulkan | 999 | 0 | pp512 | 201.03 ± 0.31 | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | Vulkan | 999 | 0 | tg128 | 22.82 ± 0.00 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | Vulkan | 99 | 0 | pp512 | 197.95 ± 0.29 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | Vulkan | 99 | 0 | tg128 | 23.24 ± 0.00 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log index c0e6775..eece528 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | Vulkan | 999 | 1 | 0 | pp512 | 201.89 ± 0.37 | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | Vulkan | 999 | 1 | 0 | tg128 | 22.85 ± 0.01 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | Vulkan | 99 | 1 | 0 | pp512 | 199.40 ± 0.35 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | Vulkan | 99 | 1 | 0 | tg128 | 23.26 ± 0.00 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_radv.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_radv.log index c38f2a1..5b8bc47 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_radv.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_radv.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | Vulkan | 999 | 0 | pp512 | 128.01 ± 0.31 | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | Vulkan | 999 | 0 | tg128 | 22.92 ± 0.01 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | Vulkan | 99 | 0 | pp512 | 126.28 ± 0.17 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | Vulkan | 99 | 0 | tg128 | 23.33 ± 0.01 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_radv__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_radv__fa1.log index 12bf239..41d8077 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_radv__fa1.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_radv__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | Vulkan | 999 | 1 | 0 | pp512 | 132.56 ± 0.31 | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | Vulkan | 999 | 1 | 0 | tg128 | 23.31 ± 0.01 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | Vulkan | 99 | 1 | 0 | pp512 | 131.64 ± 0.32 | +| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | Vulkan | 99 | 1 | 0 | tg128 | 23.88 ± 0.01 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2-rocwmma.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2-rocwmma.log index deada9f..3ee3c3e 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2-rocwmma.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2-rocwmma.log @@ -2,9 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 99 | 0 | pp512 | 124.75 ± 0.42 | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 99 | 0 | tg128 | 15.43 ± 0.00 | - -build: 79c1160b (6123) +HW Exception by GPU node-1 (Agent handle: 0x3e28b540) reason :GPU Hang +✖ ! [rocm6_4_2-rocwmma] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 failed (exit 134) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2-rocwmma__fa1.log index 6e7bcaa..2b15919 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2-rocwmma__fa1.log @@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x2d9b050) reason :GPU Hang +Memory access fault by GPU node-1 (Agent handle: 0x2bdf8540) on address 0x7f5f95e35000. Reason: Page not present or supervisor privilege. ✖ ! [rocm6_4_2-rocwmma] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 failed (exit 134) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2.log index 685d734..63bd38e 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2.log @@ -2,9 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 999 | 0 | pp512 | 124.94 ± 0.42 | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 999 | 0 | tg128 | 15.35 ± 0.00 | - -build: 79c1160b (6123) +HW Exception by GPU node-1 (Agent handle: 0x3ff2d540) reason :GPU Hang +✖ ! [rocm6_4_2] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 failed (exit 134) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2__fa1.log index b9a03bd..18f04dd 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2__fa1.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2__fa1.log @@ -2,9 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 999 | 1 | 0 | pp512 | 100.41 ± 0.16 | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 999 | 1 | 0 | tg128 | 15.53 ± 0.01 | - -build: 79c1160b (6123) +HW Exception by GPU node-1 (Agent handle: 0x3bb3540) reason :GPU Hang +✖ ! [rocm6_4_2] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2-rocwmma.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_3-rocwmma.log similarity index 54% rename from benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2-rocwmma.log rename to benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_3-rocwmma.log index 5dc10c6..f5f0ad3 100644 --- a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2-rocwmma.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_3-rocwmma.log @@ -2,9 +2,14 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +rocBLAS error: No hipBLASLt solution found +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 999 | 0 | pp512 | 113.62 ± 0.21 | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 999 | 0 | tg128 | 15.47 ± 0.04 | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 99 | 0 | pp512 | 121.82 ± 0.35 | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 99 | 0 | tg128 | 15.59 ± 0.00 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_beta__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_3-rocwmma__fa1.log similarity index 69% rename from benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_beta__fa1.log rename to benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_3-rocwmma__fa1.log index 1208365..2bfbda7 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_beta__fa1.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_3-rocwmma__fa1.log @@ -1,10 +1,10 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 999 | 1 | 0 | pp512 | 90.24 ± 0.13 | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 999 | 1 | 0 | tg128 | 15.55 ± 0.04 | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 99 | 1 | 0 | pp512 | 126.60 ± 0.30 | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 99 | 1 | 0 | tg128 | 15.62 ± 0.04 | -build: 79c1160b (6123) +build: 1fe00296 (6182) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_3-rocwmma__hblt0.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_3-rocwmma__hblt0.log new file mode 100644 index 0000000..159477b --- /dev/null +++ b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_3-rocwmma__hblt0.log @@ -0,0 +1,6 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +Memory access fault by GPU node-1 (Agent handle: 0x26e36340) on address 0x7fcef3635000. Reason: Page not present or supervisor privilege. +✖ ! [rocm6_4_3-rocwmma] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __hblt0 failed (exit 134) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_3-rocwmma__hblt0__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_3-rocwmma__hblt0__fa1.log new file mode 100644 index 0000000..f625092 --- /dev/null +++ b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_3-rocwmma__hblt0__fa1.log @@ -0,0 +1,6 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +HW Exception by GPU node-1 (Agent handle: 0x35263340) reason :GPU Hang +✖ ! [rocm6_4_3-rocwmma] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_3.log similarity index 54% rename from benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2.log rename to benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_3.log index d1de7a1..8ac4440 100644 --- a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_3.log @@ -2,9 +2,14 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +rocBLAS error: No hipBLASLt solution found +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 999 | 0 | pp512 | 124.82 ± 0.18 | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 999 | 0 | tg128 | 15.35 ± 0.00 | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 99 | 0 | pp512 | 117.95 ± 0.30 | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 99 | 0 | tg128 | 15.65 ± 0.01 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_3__fa1.log similarity index 54% rename from benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2__fa1.log rename to benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_3__fa1.log index 5ed10e0..581bc16 100644 --- a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2__fa1.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_3__fa1.log @@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -Memory access fault by GPU node-1 (Agent handle: 0x1527fa90) on address 0x7f55d5f6f000. Reason: Page not present or supervisor privilege. -✖ ! [rocm6_4_2] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 failed (exit 134) +Memory access fault by GPU node-1 (Agent handle: 0x28aa3340) on address 0x7fb93761b000. Reason: Page not present or supervisor privilege. +✖ ! [rocm6_4_3] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 failed (exit 134) diff --git a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_3__hblt0.log similarity index 58% rename from benchmark/results/llama3.3-70.6B-Q4_K_M__rocm6_4_2-rocwmma__fa1.log rename to benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_3__hblt0.log index d800da3..486bf8f 100644 --- a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_3__hblt0.log @@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x3e596050) reason :GPU Hang -✖ ! [rocm6_4_2-rocwmma] llama3.3-70.6B-Q4_K_M __fa1 failed (exit 134) +HW Exception by GPU node-1 (Agent handle: 0x14d05340) reason :GPU Hang +✖ ! [rocm6_4_3] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __hblt0 failed (exit 134) diff --git a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_3__hblt0__fa1.log similarity index 57% rename from benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2.log rename to benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_3__hblt0__fa1.log index c119ae0..fb08717 100644 --- a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_3__hblt0__fa1.log @@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x2dab2050) reason :GPU Hang -✖ ! [rocm6_4_2] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 failed (exit 134) +HW Exception by GPU node-1 (Agent handle: 0x265e8340) reason :GPU Hang +✖ ! [rocm6_4_3] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_beta.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_beta.log deleted file mode 100644 index d839f3e..0000000 --- a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 999 | 0 | pp512 | 118.61 ± 0.54 | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 999 | 0 | tg128 | 15.51 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_beta__hblt0.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_beta__hblt0.log deleted file mode 100644 index 8a0ff5a..0000000 --- a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_beta__hblt0.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 99 | 0 | pp512 | 123.75 ± 0.39 | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 99 | 0 | tg128 | 15.48 ± 0.01 | - -build: 79c1160b (6123) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_beta__hblt0__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_beta__hblt0__fa1.log deleted file mode 100644 index 0aab0d1..0000000 --- a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_beta__hblt0__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -Memory access fault by GPU node-1 (Agent handle: 0x36bce0c0) on address 0x7f6ee1f6f000. Reason: Page not present or supervisor privilege. -✖ ! [rocm7_beta] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc-rocwmma.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc-rocwmma.log index 55995b2..8ca5c1e 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc-rocwmma.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc-rocwmma.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 999 | 0 | pp512 | 118.92 ± 0.39 | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 999 | 0 | tg128 | 15.47 ± 0.00 | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 99 | 0 | pp512 | 69.19 ± 0.20 | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 99 | 0 | tg128 | 15.64 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc-rocwmma__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc-rocwmma__fa1.log index 161145c..6eaa1df 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc-rocwmma__fa1.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 999 | 1 | 0 | pp512 | 127.14 ± 0.27 | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 999 | 1 | 0 | tg128 | 15.47 ± 0.00 | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 99 | 1 | 0 | pp512 | 114.61 ± 0.20 | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 99 | 1 | 0 | tg128 | 15.51 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc-rocwmma__hblt0.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc-rocwmma__hblt0.log index 5e8d103..5fb6167 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc-rocwmma__hblt0.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc-rocwmma__hblt0.log @@ -2,4 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc-rocwmma] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __hblt0 failed (exit 134) +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 99 | 0 | pp512 | 120.88 ± 0.92 | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 99 | 0 | tg128 | 15.61 ± 0.09 | + +build: de219279 (6181) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc-rocwmma__hblt0__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc-rocwmma__hblt0__fa1.log index d77ab42..ec552ad 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc-rocwmma__hblt0__fa1.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc-rocwmma__hblt0__fa1.log @@ -2,4 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc-rocwmma] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __hblt0__fa1 failed (exit 134) +| model | size | params | backend | ngl | fa | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 99 | 1 | 0 | pp512 | 150.07 ± 0.56 | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 99 | 1 | 0 | tg128 | 15.52 ± 0.00 | + +build: de219279 (6181) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc.log index 97c3a25..dc5e6ec 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 999 | 0 | pp512 | 118.52 ± 0.35 | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 999 | 0 | tg128 | 15.52 ± 0.00 | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 99 | 0 | pp512 | 69.52 ± 0.17 | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 99 | 0 | tg128 | 15.63 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc__fa1.log index c6194e3..c1980f8 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc__fa1.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 999 | 1 | 0 | pp512 | 97.36 ± 0.07 | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 999 | 1 | 0 | tg128 | 15.57 ± 0.02 | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 99 | 1 | 0 | pp512 | 74.02 ± 0.13 | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 99 | 1 | 0 | tg128 | 15.73 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc__hblt0.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc__hblt0.log index 7f9bb58..350f64d 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc__hblt0.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc__hblt0.log @@ -2,4 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __hblt0 failed (exit 134) +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 99 | 0 | pp512 | 142.67 ± 0.75 | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 99 | 0 | tg128 | 15.68 ± 0.00 | + +build: de219279 (6181) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc__hblt0__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc__hblt0__fa1.log index 8e70bdf..6a1fdfd 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc__hblt0__fa1.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc__hblt0__fa1.log @@ -2,4 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +Memory access fault by GPU node-1 (Agent handle: 0x1c536ea0) on address 0x7f623b57e000. Reason: Page not present or supervisor privilege. ✖ ! [rocm7_rc] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_amdvlk.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_amdvlk.log index f5209b4..1a18e9c 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_amdvlk.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_amdvlk.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | Vulkan | 999 | 0 | pp512 | 223.59 ± 0.50 | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | Vulkan | 999 | 0 | tg128 | 16.51 ± 0.01 | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | Vulkan | 99 | 0 | pp512 | 219.81 ± 0.70 | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | Vulkan | 99 | 0 | tg128 | 16.80 ± 0.00 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_amdvlk__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_amdvlk__fa1.log index 3ba2fc2..bc34d0a 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_amdvlk__fa1.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_amdvlk__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | Vulkan | 999 | 1 | 0 | pp512 | 225.75 ± 0.69 | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | Vulkan | 999 | 1 | 0 | tg128 | 16.53 ± 0.00 | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | Vulkan | 99 | 1 | 0 | pp512 | 222.20 ± 0.63 | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | Vulkan | 99 | 1 | 0 | tg128 | 16.82 ± 0.01 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_radv.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_radv.log index c01b8a2..1c621cf 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_radv.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_radv.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | Vulkan | 999 | 0 | pp512 | 127.35 ± 0.43 | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | Vulkan | 999 | 0 | tg128 | 16.80 ± 0.00 | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | Vulkan | 99 | 0 | pp512 | 126.55 ± 0.40 | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | Vulkan | 99 | 0 | tg128 | 17.07 ± 0.01 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_radv__fa1.log b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_radv__fa1.log index 3647c19..3dbeebf 100644 --- a/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_radv__fa1.log +++ b/benchmark/results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_radv__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | Vulkan | 999 | 1 | 0 | pp512 | 131.91 ± 0.42 | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | Vulkan | 999 | 1 | 0 | tg128 | 17.02 ± 0.00 | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | Vulkan | 99 | 1 | 0 | pp512 | 131.25 ± 0.50 | +| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | Vulkan | 99 | 1 | 0 | tg128 | 17.31 ± 0.00 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2__fa1.log b/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2__fa1.log deleted file mode 100644 index 49b6a40..0000000 --- a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -Memory access fault by GPU node-1 (Agent handle: 0xae0b050) on address 0x7f17943a9000. Reason: Page not present or supervisor privilege. -✖ ! [rocm6_4_2] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_beta.log b/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_beta.log deleted file mode 100644 index 820c8ea..0000000 --- a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen2 70B Q8_0 | 78.21 GiB | 72.71 B | ROCm | 999 | 0 | pp512 | 108.88 ± 0.21 | -| qwen2 70B Q8_0 | 78.21 GiB | 72.71 B | ROCm | 999 | 0 | tg128 | 2.65 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_beta__fa1.log b/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_beta__fa1.log deleted file mode 100644 index ecdf26e..0000000 --- a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_beta__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -Memory access fault by GPU node-1 (Agent handle: 0x1f7690e0) on address 0x7f6093d6f000. Reason: Page not present or supervisor privilege. -✖ ! [rocm7_beta] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_beta__hblt0.log b/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_beta__hblt0.log deleted file mode 100644 index 9bbdd27..0000000 --- a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_beta__hblt0.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x2ae290c0) reason :GPU Hang -✖ ! [rocm7_beta] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __hblt0 failed (exit 134) diff --git a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_beta__hblt0__fa1.log b/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_beta__hblt0__fa1.log deleted file mode 100644 index 3a354c6..0000000 --- a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_beta__hblt0__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x19f880e0) reason :GPU Hang -✖ ! [rocm7_beta] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma.log b/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma.log deleted file mode 100644 index 01916bd..0000000 --- a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen2 70B Q8_0 | 78.21 GiB | 72.71 B | ROCm | 999 | 0 | pp512 | 109.02 ± 0.07 | -| qwen2 70B Q8_0 | 78.21 GiB | 72.71 B | ROCm | 999 | 0 | tg128 | 2.65 ± 0.00 | - -build: 34c9d765 (6122) diff --git a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma__fa1.log b/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma__fa1.log deleted file mode 100644 index 4d3f05b..0000000 --- a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen2 70B Q8_0 | 78.21 GiB | 72.71 B | ROCm | 999 | 1 | 0 | pp512 | 117.34 ± 0.09 | -| qwen2 70B Q8_0 | 78.21 GiB | 72.71 B | ROCm | 999 | 1 | 0 | tg128 | 2.65 ± 0.00 | - -build: 34c9d765 (6122) diff --git a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0.log b/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0.log deleted file mode 100644 index 08bd375..0000000 --- a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0.log +++ /dev/null @@ -1,5 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc-rocwmma] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __hblt0 failed (exit 134) diff --git a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log b/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log deleted file mode 100644 index 3d74d91..0000000 --- a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log +++ /dev/null @@ -1,5 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc-rocwmma] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc.log b/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc.log deleted file mode 100644 index 343727e..0000000 --- a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen2 70B Q8_0 | 78.21 GiB | 72.71 B | ROCm | 999 | 0 | pp512 | 109.17 ± 0.12 | -| qwen2 70B Q8_0 | 78.21 GiB | 72.71 B | ROCm | 999 | 0 | tg128 | 2.65 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc__fa1.log b/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc__fa1.log deleted file mode 100644 index d9b5fe6..0000000 --- a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc__fa1.log +++ /dev/null @@ -1,5 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc__hblt0.log b/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc__hblt0.log deleted file mode 100644 index fac1830..0000000 --- a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc__hblt0.log +++ /dev/null @@ -1,5 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __hblt0 failed (exit 134) diff --git a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc__hblt0__fa1.log b/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc__hblt0__fa1.log deleted file mode 100644 index f08d646..0000000 --- a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc__hblt0__fa1.log +++ /dev/null @@ -1,5 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk.log b/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk.log deleted file mode 100644 index 4581b23..0000000 --- a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -ggml_vulkan: Device memory allocation of size 2491416576 failed. -ggml_vulkan: Requested buffer size exceeds device memory allocation limit: ErrorOutOfDeviceMemory -main: error: failed to load model '/mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf' -✖ ! [vulkan_amdvlk] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 failed (exit 1) diff --git a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log b/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log deleted file mode 100644 index 8835330..0000000 --- a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -ggml_vulkan: Device memory allocation of size 2491416576 failed. -ggml_vulkan: Requested buffer size exceeds device memory allocation limit: ErrorOutOfDeviceMemory -main: error: failed to load model '/mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf' -✖ ! [vulkan_amdvlk] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 failed (exit 1) diff --git a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_radv.log b/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_radv.log deleted file mode 100644 index f0955ad..0000000 --- a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_radv.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen2 70B Q8_0 | 78.21 GiB | 72.71 B | Vulkan | 999 | 0 | pp512 | 78.54 ± 0.14 | -| qwen2 70B Q8_0 | 78.21 GiB | 72.71 B | Vulkan | 999 | 0 | tg128 | 2.67 ± 0.00 | - -build: 34c9d765 (6122) diff --git a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_radv__fa1.log b/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_radv__fa1.log deleted file mode 100644 index 0c2bb42..0000000 --- a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_radv__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen2 70B Q8_0 | 78.21 GiB | 72.71 B | Vulkan | 999 | 1 | 0 | pp512 | 81.12 ± 0.08 | -| qwen2 70B Q8_0 | 78.21 GiB | 72.71 B | Vulkan | 999 | 1 | 0 | tg128 | 2.67 ± 0.00 | - -build: 34c9d765 (6122) diff --git a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log index 5c9071b..4fb737a 100644 --- a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log +++ b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log @@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0xd004050) reason :GPU Hang +HW Exception by GPU node-1 (Agent handle: 0x33b8a540) reason :GPU Hang ✖ ! [rocm6_4_2-rocwmma] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 failed (exit 134) diff --git a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log index 182cfd1..8ed4e21 100644 --- a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log @@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x1fdc2050) reason :GPU Hang +HW Exception by GPU node-1 (Agent handle: 0x20e35540) reason :GPU Hang ✖ ! [rocm6_4_2-rocwmma] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2.log index 1e5d45d..8ad5ab6 100644 --- a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2.log +++ b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2.log @@ -2,9 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 999 | 0 | pp512 | 33.28 ± 0.05 | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 999 | 0 | tg128 | 2.73 ± 0.00 | - -build: 79c1160b (6123) +HW Exception by GPU node-1 (Agent handle: 0x1b1ea540) reason :GPU Hang +✖ ! [rocm6_4_2] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 failed (exit 134) diff --git a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2__fa1.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2__fa1.log index db68588..7860063 100644 --- a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2__fa1.log +++ b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 999 | 1 | 0 | pp512 | 30.88 ± 0.02 | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 999 | 1 | 0 | tg128 | 2.73 ± 0.00 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 99 | 1 | 0 | pp512 | 16.16 ± 0.02 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 99 | 1 | 0 | tg128 | 2.78 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_3-rocwmma.log similarity index 54% rename from benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2.log rename to benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_3-rocwmma.log index 409a36b..769f2f3 100644 --- a/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2.log +++ b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_3-rocwmma.log @@ -2,9 +2,14 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +hipBLASLt error: Heuristic Fetch Failed! +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 999 | 0 | pp512 | 33.32 ± 0.04 | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 999 | 0 | tg128 | 2.73 ± 0.00 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 99 | 0 | pp512 | 98.02 ± 0.18 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 99 | 0 | tg128 | 2.77 ± 0.00 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2__fa1.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_3-rocwmma__fa1.log similarity index 79% rename from benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2__fa1.log rename to benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_3-rocwmma__fa1.log index 952dddb..4451e68 100644 --- a/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2__fa1.log +++ b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_3-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 999 | 1 | 0 | pp512 | 31.28 ± 0.02 | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 999 | 1 | 0 | tg128 | 2.74 ± 0.00 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 99 | 1 | 0 | pp512 | 101.83 ± 0.11 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 99 | 1 | 0 | tg128 | 2.77 ± 0.00 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_3-rocwmma__hblt0.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_3-rocwmma__hblt0.log new file mode 100644 index 0000000..2553f31 --- /dev/null +++ b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_3-rocwmma__hblt0.log @@ -0,0 +1,6 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +HW Exception by GPU node-1 (Agent handle: 0x21da1340) reason :GPU Hang +✖ ! [rocm6_4_3-rocwmma] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __hblt0 failed (exit 134) diff --git a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_3-rocwmma__hblt0__fa1.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_3-rocwmma__hblt0__fa1.log new file mode 100644 index 0000000..1a88dcd --- /dev/null +++ b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_3-rocwmma__hblt0__fa1.log @@ -0,0 +1,6 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +HW Exception by GPU node-1 (Agent handle: 0x15ac2340) reason :GPU Hang +✖ ! [rocm6_4_3-rocwmma] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_3.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_3.log new file mode 100644 index 0000000..56497ee --- /dev/null +++ b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_3.log @@ -0,0 +1,15 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +hipBLASLt error: Heuristic Fetch Failed! +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 99 | 0 | pp512 | 97.13 ± 0.17 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 99 | 0 | tg128 | 2.78 ± 0.00 | + +build: 1fe00296 (6182) diff --git a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm6_4_2__fa1.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_3__fa1.log similarity index 62% rename from benchmark/results/llama3.3-70.6B-Q4_K_M__rocm6_4_2__fa1.log rename to benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_3__fa1.log index cc4c84c..08a5922 100644 --- a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm6_4_2__fa1.log +++ b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_3__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 1 | 0 | pp512 | 31.69 ± 0.04 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 1 | 0 | tg128 | 4.62 ± 0.00 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 99 | 1 | 0 | pp512 | 80.42 ± 0.08 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 99 | 1 | 0 | tg128 | 2.78 ± 0.00 | -build: 79c1160b (6123) +build: 1fe00296 (6182) diff --git a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_3__hblt0.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_3__hblt0.log new file mode 100644 index 0000000..f5b7147 --- /dev/null +++ b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_3__hblt0.log @@ -0,0 +1,6 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +HW Exception by GPU node-1 (Agent handle: 0x2c1e5340) reason :GPU Hang +✖ ! [rocm6_4_3] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __hblt0 failed (exit 134) diff --git a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_3__hblt0__fa1.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_3__hblt0__fa1.log new file mode 100644 index 0000000..013fc7c --- /dev/null +++ b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_3__hblt0__fa1.log @@ -0,0 +1,6 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +Memory access fault by GPU node-1 (Agent handle: 0x3e536340) on address 0x7f9182f6f000. Reason: Page not present or supervisor privilege. +✖ ! [rocm6_4_3] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_beta.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_beta.log deleted file mode 100644 index 1275625..0000000 --- a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 999 | 0 | pp512 | 95.65 ± 0.23 | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 999 | 0 | tg128 | 2.74 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_beta__fa1.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_beta__fa1.log deleted file mode 100644 index d4fb01f..0000000 --- a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_beta__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -Memory access fault by GPU node-1 (Agent handle: 0x2e9460f0) on address 0x7f23cf58a000. Reason: Page not present or supervisor privilege. -✖ ! [rocm7_beta] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_beta__hblt0.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_beta__hblt0.log deleted file mode 100644 index 256df75..0000000 --- a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_beta__hblt0.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x2c3170e0) reason :GPU Hang -✖ ! [rocm7_beta] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __hblt0 failed (exit 134) diff --git a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_beta__hblt0__fa1.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_beta__hblt0__fa1.log deleted file mode 100644 index 51e9900..0000000 --- a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_beta__hblt0__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -Memory access fault by GPU node-1 (Agent handle: 0xe3f70e0) on address 0x7f4e23b6f000. Reason: Page not present or supervisor privilege. -✖ ! [rocm7_beta] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma.log index 2d8ff98..62ab0f7 100644 --- a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma.log +++ b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 999 | 0 | pp512 | 95.63 ± 0.19 | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 999 | 0 | tg128 | 2.73 ± 0.00 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 99 | 0 | pp512 | 97.31 ± 0.20 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 99 | 0 | tg128 | 2.78 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma__fa1.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma__fa1.log index 78ef763..1e31a60 100644 --- a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma__fa1.log +++ b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 999 | 1 | 0 | pp512 | 103.15 ± 0.13 | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 999 | 1 | 0 | tg128 | 2.73 ± 0.00 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 99 | 1 | 0 | pp512 | 100.85 ± 0.13 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 99 | 1 | 0 | tg128 | 2.77 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0.log index 8abfc94..c5612f6 100644 --- a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0.log +++ b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0.log @@ -2,4 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc-rocwmma] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __hblt0 failed (exit 134) +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 99 | 0 | pp512 | 93.00 ± 0.22 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 99 | 0 | tg128 | 2.78 ± 0.00 | + +build: de219279 (6181) diff --git a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log index b85ea71..e557123 100644 --- a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log +++ b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log @@ -2,4 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc-rocwmma] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __hblt0__fa1 failed (exit 134) +| model | size | params | backend | ngl | fa | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 99 | 1 | 0 | pp512 | 97.88 ± 0.09 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 99 | 1 | 0 | tg128 | 2.77 ± 0.00 | + +build: de219279 (6181) diff --git a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc.log index 7715818..7c6d5fd 100644 --- a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc.log +++ b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 999 | 0 | pp512 | 95.15 ± 0.14 | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 999 | 0 | tg128 | 2.74 ± 0.00 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 99 | 0 | pp512 | 99.41 ± 0.36 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 99 | 0 | tg128 | 2.77 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc__fa1.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc__fa1.log index a10eecc..f65845d 100644 --- a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc__fa1.log +++ b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc__fa1.log @@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -:0:rocdevice.cpp :3594: 448132897452 us: Callback: Queue 0x7f7ecc400000 aborting with error : HSA_STATUS_ERROR_EXCEPTION: An HSAIL operation resulted in a hardware exception. code: 0x1016 +Memory access fault by GPU node-1 (Agent handle: 0x1f66bec0) on address 0x7f3e84b6f000. Reason: Page not present or supervisor privilege. ✖ ! [rocm7_rc] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc__hblt0.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc__hblt0.log index e557f4b..44b15c4 100644 --- a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc__hblt0.log +++ b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc__hblt0.log @@ -2,4 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __hblt0 failed (exit 134) +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 99 | 0 | pp512 | 94.06 ± 0.09 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 99 | 0 | tg128 | 2.78 ± 0.00 | + +build: de219279 (6181) diff --git a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc__hblt0__fa1.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc__hblt0__fa1.log index f3c72e9..970522b 100644 --- a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc__hblt0__fa1.log +++ b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc__hblt0__fa1.log @@ -2,9 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 99 | 1 | 0 | pp512 | 30.04 ± 0.04 | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | ROCm | 99 | 1 | 0 | tg128 | 2.74 ± 0.00 | - -build: 79c1160b (6123) +Memory access fault by GPU node-1 (Agent handle: 0xac09ec0) on address 0x7f283f56f000. Reason: Page not present or supervisor privilege. +✖ ! [rocm7_rc] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk.log index ac2d0df..13252f9 100644 --- a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk.log +++ b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | Vulkan | 999 | 0 | pp512 | 98.20 ± 0.18 | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | Vulkan | 999 | 0 | tg128 | 2.75 ± 0.00 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | Vulkan | 99 | 0 | pp512 | 98.03 ± 0.24 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | Vulkan | 99 | 0 | tg128 | 2.78 ± 0.00 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log index 9e22472..5781898 100644 --- a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log +++ b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | Vulkan | 999 | 1 | 0 | pp512 | 99.14 ± 0.35 | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | Vulkan | 999 | 1 | 0 | tg128 | 2.74 ± 0.00 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | Vulkan | 99 | 1 | 0 | pp512 | 99.12 ± 0.25 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | Vulkan | 99 | 1 | 0 | tg128 | 2.77 ± 0.00 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_radv.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_radv.log index b4da67c..9500e62 100644 --- a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_radv.log +++ b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_radv.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | Vulkan | 999 | 0 | pp512 | 79.91 ± 0.16 | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | Vulkan | 999 | 0 | tg128 | 2.75 ± 0.00 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | Vulkan | 99 | 0 | pp512 | 75.59 ± 0.28 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | Vulkan | 99 | 0 | tg128 | 2.78 ± 0.00 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_radv__fa1.log b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_radv__fa1.log index 5e3f60f..6c47ac0 100644 --- a/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_radv__fa1.log +++ b/benchmark/results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_radv__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | Vulkan | 999 | 1 | 0 | pp512 | 82.40 ± 0.16 | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | Vulkan | 999 | 1 | 0 | tg128 | 2.75 ± 0.00 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | Vulkan | 99 | 1 | 0 | pp512 | 80.09 ± 0.38 | +| llama 70B Q8_0 | 75.65 GiB | 70.55 B | Vulkan | 99 | 1 | 0 | tg128 | 2.78 ± 0.00 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2-rocwmma.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2-rocwmma.log index b5a7155..6babfa8 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2-rocwmma.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2-rocwmma.log @@ -2,9 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 99 | 0 | pp512 | 134.21 ± 0.58 | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 99 | 0 | tg128 | 14.43 ± 0.00 | - -build: 79c1160b (6123) +HW Exception by GPU node-1 (Agent handle: 0x344ea540) reason :GPU Hang +✖ ! [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2-rocwmma__fa1.log index 3a30aaf..2f3524f 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2-rocwmma__fa1.log @@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x10997050) reason :GPU Hang +HW Exception by GPU node-1 (Agent handle: 0xe316540) reason :GPU Hang ✖ ! [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2.log index f81dd87..1009e19 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2.log @@ -2,9 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 999 | 0 | pp512 | 133.77 ± 0.46 | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 999 | 0 | tg128 | 14.30 ± 0.00 | - -build: 79c1160b (6123) +HW Exception by GPU node-1 (Agent handle: 0x17ade540) reason :GPU Hang +✖ ! [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2__fa1.log index b6d7516..c7625db 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2__fa1.log @@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -Memory access fault by GPU node-1 (Agent handle: 0x1732e050) on address 0x7fcb1a36f000. Reason: Page not present or supervisor privilege. +HW Exception by GPU node-1 (Agent handle: 0xe91f540) reason :GPU Hang ✖ ! [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_3-rocwmma.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_3-rocwmma.log new file mode 100644 index 0000000..ebcf552 --- /dev/null +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_3-rocwmma.log @@ -0,0 +1,11 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +rocBLAS error: No hipBLASLt solution found +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. +Memory access fault by GPU node-1 (Agent handle: 0x1a840340) on address 0x7f3babb56000. Reason: Page not present or supervisor privilege. +✖ ! [rocm6_4_3-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_beta__hblt0__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_3-rocwmma__fa1.log similarity index 68% rename from benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_beta__hblt0__fa1.log rename to benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_3-rocwmma__fa1.log index 6375788..ab25429 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_beta__hblt0__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_3-rocwmma__fa1.log @@ -1,10 +1,10 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 99 | 1 | 0 | pp512 | 103.96 ± 0.18 | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 99 | 1 | 0 | tg128 | 14.47 ± 0.02 | +| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 99 | 1 | 0 | pp512 | 291.08 ± 1.26 | +| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 99 | 1 | 0 | tg128 | 14.53 ± 0.00 | -build: 79c1160b (6123) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_3-rocwmma__hblt0.log similarity index 79% rename from benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2.log rename to benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_3-rocwmma__hblt0.log index f7132fb..65957d1 100644 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_3-rocwmma__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 999 | 0 | pp512 | 134.39 ± 0.32 | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 999 | 0 | tg128 | 14.33 ± 0.00 | +| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 99 | 0 | pp512 | 134.19 ± 1.49 | +| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 99 | 0 | tg128 | 14.56 ± 0.01 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_3-rocwmma__hblt0__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_3-rocwmma__hblt0__fa1.log new file mode 100644 index 0000000..dcff3bd --- /dev/null +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_3-rocwmma__hblt0__fa1.log @@ -0,0 +1,6 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +HW Exception by GPU node-1 (Agent handle: 0x1de78340) reason :GPU Hang +✖ ! [rocm6_4_3-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_3.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_3.log new file mode 100644 index 0000000..f75714d --- /dev/null +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_3.log @@ -0,0 +1,15 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +rocBLAS error: No hipBLASLt solution found +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 99 | 0 | pp512 | 270.28 ± 1.29 | +| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 99 | 0 | tg128 | 14.58 ± 0.03 | + +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_3__fa1.log similarity index 53% rename from benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2__fa1.log rename to benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_3__fa1.log index 53feea1..7c8dc76 100644 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_3__fa1.log @@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x3b11ea90) reason :GPU Hang -✖ ! [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 failed (exit 134) +Memory access fault by GPU node-1 (Agent handle: 0x2162b340) on address 0x7f500556f000. Reason: Page not present or supervisor privilege. +✖ ! [rocm6_4_3] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_3__hblt0.log similarity index 56% rename from benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log rename to benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_3__hblt0.log index d044208..96e7a94 100644 --- a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_3__hblt0.log @@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x432ea90) reason :GPU Hang -✖ ! [rocm6_4_2-rocwmma] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __fa1 failed (exit 134) +HW Exception by GPU node-1 (Agent handle: 0xdacf340) reason :GPU Hang +✖ ! [rocm6_4_3] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __hblt0 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_3__hblt0__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_3__hblt0__fa1.log new file mode 100644 index 0000000..65a2e33 --- /dev/null +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_3__hblt0__fa1.log @@ -0,0 +1,6 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +HW Exception by GPU node-1 (Agent handle: 0x3dc00340) reason :GPU Hang +✖ ! [rocm6_4_3] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_beta.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_beta.log deleted file mode 100644 index e2f4dbe..0000000 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_beta.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x225860e0) reason :GPU Hang -✖ ! [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_beta__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_beta__fa1.log deleted file mode 100644 index 311e082..0000000 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_beta__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -:0:rocdevice.cpp :3675: 454572762136 us: Callback: Queue 0x7fb3f1400000 aborting with error : HSA_STATUS_ERROR_EXCEPTION: An HSAIL operation resulted in a hardware exception. code: 0x1016 -✖ ! [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_beta__hblt0.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_beta__hblt0.log deleted file mode 100644 index 7188df3..0000000 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_beta__hblt0.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x11dec0e0) reason :GPU Hang -✖ ! [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __hblt0 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc-rocwmma.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc-rocwmma.log index be95461..64b52d5 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc-rocwmma.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc-rocwmma.log @@ -2,9 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 999 | 0 | pp512 | 273.64 ± 0.59 | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 999 | 0 | tg128 | 14.43 ± 0.00 | - -build: 34c9d765 (6122) +HW Exception by GPU node-1 (Agent handle: 0x3882bf60) reason :GPU Hang +✖ ! [rocm7_rc-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc-rocwmma__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc-rocwmma__fa1.log index 70811e5..11cbda1 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc-rocwmma__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 999 | 1 | 0 | pp512 | 293.87 ± 1.35 | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 999 | 1 | 0 | tg128 | 14.31 ± 0.00 | +| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 99 | 1 | 0 | pp512 | 285.84 ± 9.41 | +| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 99 | 1 | 0 | tg128 | 14.37 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc-rocwmma__hblt0.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc-rocwmma__hblt0.log index fa50f09..7575a7e 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc-rocwmma__hblt0.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc-rocwmma__hblt0.log @@ -2,4 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __hblt0 failed (exit 134) +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 99 | 0 | pp512 | 273.97 ± 1.67 | +| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 99 | 0 | tg128 | 14.57 ± 0.05 | + +build: de219279 (6181) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log index e9eb9a3..647f737 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log @@ -2,4 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __hblt0__fa1 failed (exit 134) +| model | size | params | backend | ngl | fa | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | +| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 99 | 1 | 0 | pp512 | 285.26 ± 1.79 | +| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 99 | 1 | 0 | tg128 | 14.33 ± 0.03 | + +build: de219279 (6181) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc.log index 07c550f..6ac1b55 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 999 | 0 | pp512 | 269.30 ± 1.99 | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 999 | 0 | tg128 | 14.43 ± 0.00 | +| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 99 | 0 | pp512 | 276.37 ± 1.65 | +| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 99 | 0 | tg128 | 14.57 ± 0.04 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc__fa1.log index 0a2d01d..bc46574 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc__fa1.log @@ -2,9 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 999 | 1 | 0 | pp512 | 225.70 ± 1.00 | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 999 | 1 | 0 | tg128 | 14.46 ± 0.00 | - -build: 79c1160b (6123) +Memory access fault by GPU node-1 (Agent handle: 0xa893ec0) on address 0x7f070a3a9000. Reason: Page not present or supervisor privilege. +✖ ! [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc__hblt0.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc__hblt0.log index c9103d5..ca30067 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc__hblt0.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 99 | 0 | pp512 | 135.16 ± 0.44 | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 99 | 0 | tg128 | 14.41 ± 0.00 | +| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 99 | 0 | pp512 | 269.17 ± 0.99 | +| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 99 | 0 | tg128 | 14.63 ± 0.01 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc__hblt0__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc__hblt0__fa1.log index 560fe07..56c35f6 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc__hblt0__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc__hblt0__fa1.log @@ -2,4 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +Memory access fault by GPU node-1 (Agent handle: 0x1db86ec0) on address 0x7f2273f6f000. Reason: Page not present or supervisor privilege. ✖ ! [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_amdvlk.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_amdvlk.log index 3cc2007..8c8e292 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_amdvlk.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_amdvlk.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | Vulkan | 999 | 0 | pp512 | 243.54 ± 1.24 | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | Vulkan | 999 | 0 | tg128 | 15.34 ± 0.00 | +| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | Vulkan | 99 | 0 | pp512 | 242.07 ± 1.05 | +| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | Vulkan | 99 | 0 | tg128 | 15.56 ± 0.01 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_amdvlk__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_amdvlk__fa1.log index 7dafa9d..71556fe 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_amdvlk__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_amdvlk__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | Vulkan | 999 | 1 | 0 | pp512 | 246.48 ± 1.35 | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | Vulkan | 999 | 1 | 0 | tg128 | 15.09 ± 0.01 | +| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | Vulkan | 99 | 1 | 0 | pp512 | 244.49 ± 1.13 | +| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | Vulkan | 99 | 1 | 0 | tg128 | 15.33 ± 0.00 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_radv.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_radv.log index 80c940d..33a8b80 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_radv.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_radv.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | Vulkan | 999 | 0 | pp512 | 147.36 ± 0.80 | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | Vulkan | 999 | 0 | tg128 | 15.30 ± 0.01 | +| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | Vulkan | 99 | 0 | pp512 | 147.08 ± 0.98 | +| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | Vulkan | 99 | 0 | tg128 | 15.50 ± 0.01 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_radv__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_radv__fa1.log index e72dffe..caf7973 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_radv__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_radv__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | Vulkan | 999 | 1 | 0 | pp512 | 150.06 ± 1.13 | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | Vulkan | 999 | 1 | 0 | tg128 | 15.27 ± 0.00 | +| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | Vulkan | 99 | 1 | 0 | pp512 | 149.97 ± 1.10 | +| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | Vulkan | 99 | 1 | 0 | tg128 | 15.49 ± 0.00 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2-rocwmma.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2-rocwmma.log index 0910303..a8331aa 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2-rocwmma.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2-rocwmma.log @@ -2,9 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 99 | 0 | pp512 | 135.23 ± 0.81 | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 99 | 0 | tg128 | 11.62 ± 0.00 | - -build: 79c1160b (6123) +HW Exception by GPU node-1 (Agent handle: 0x1019d540) reason :GPU Hang +✖ ! [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2-rocwmma__fa1.log index 07a24ef..b68e909 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2-rocwmma__fa1.log @@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0xf461050) reason :GPU Hang +HW Exception by GPU node-1 (Agent handle: 0x2ff5c540) reason :GPU Hang ✖ ! [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2.log index db50520..88fb55c 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2.log @@ -2,9 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 999 | 0 | pp512 | 135.29 ± 0.58 | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 999 | 0 | tg128 | 11.60 ± 0.00 | - -build: 79c1160b (6123) +HW Exception by GPU node-1 (Agent handle: 0x3db80540) reason :GPU Hang +✖ ! [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2__fa1.log index 4d84455..befc174 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2__fa1.log @@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -Memory access fault by GPU node-1 (Agent handle: 0x13dd2050) on address 0x7f6913b6f000. Reason: Page not present or supervisor privilege. +HW Exception by GPU node-1 (Agent handle: 0x24a4c540) reason :GPU Hang ✖ ! [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_3-rocwmma.log similarity index 54% rename from benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2.log rename to benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_3-rocwmma.log index 8678b7b..40cf34f 100644 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_3-rocwmma.log @@ -2,9 +2,14 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +rocBLAS error: No hipBLASLt solution found +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 999 | 0 | pp512 | 135.44 ± 0.76 | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 999 | 0 | tg128 | 11.61 ± 0.00 | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 99 | 0 | pp512 | 270.35 ± 3.39 | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 99 | 0 | tg128 | 11.78 ± 0.03 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_3-rocwmma__fa1.log similarity index 61% rename from benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log rename to benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_3-rocwmma__fa1.log index 911ffe7..4a19b96 100644 --- a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_3-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 153.97 ± 1.90 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 23.98 ± 0.01 | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 99 | 1 | 0 | pp512 | 292.23 ± 3.13 | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 99 | 1 | 0 | tg128 | 11.73 ± 0.03 | -build: 79c1160b (6123) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_3-rocwmma__hblt0.log similarity index 55% rename from benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2__fa1.log rename to benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_3-rocwmma__hblt0.log index 1849a77..65afec2 100644 --- a/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_3-rocwmma__hblt0.log @@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x64dea90) reason :GPU Hang -✖ ! [rocm6_4_2] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 failed (exit 134) +HW Exception by GPU node-1 (Agent handle: 0x5f69340) reason :GPU Hang +✖ ! [rocm6_4_3-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __hblt0 failed (exit 134) diff --git a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_3-rocwmma__hblt0__fa1.log similarity index 62% rename from benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log rename to benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_3-rocwmma__hblt0__fa1.log index bb5ac9d..98dc8c5 100644 --- a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_3-rocwmma__hblt0__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen2 70B Q8_0 | 78.21 GiB | 72.71 B | ROCm | 99 | 1 | 0 | pp512 | 33.87 ± 0.05 | -| qwen2 70B Q8_0 | 78.21 GiB | 72.71 B | ROCm | 99 | 1 | 0 | tg128 | 2.64 ± 0.00 | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 99 | 1 | 0 | pp512 | 140.27 ± 0.97 | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 99 | 1 | 0 | tg128 | 11.74 ± 0.00 | -build: 79c1160b (6123) +build: 1fe00296 (6182) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_3.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_3.log new file mode 100644 index 0000000..8ac0514 --- /dev/null +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_3.log @@ -0,0 +1,11 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +rocBLAS error: No hipBLASLt solution found +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. +HW Exception by GPU node-1 (Agent handle: 0x2079b340) reason :GPU Hang +✖ ! [rocm6_4_3] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_3__fa1.log similarity index 53% rename from benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2__fa1.log rename to benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_3__fa1.log index 099d9b2..8d6068b 100644 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_3__fa1.log @@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x2fba3a90) reason :GPU Hang -✖ ! [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 failed (exit 134) +Memory access fault by GPU node-1 (Agent handle: 0x37ff7340) on address 0x7fa76bba9000. Reason: Page not present or supervisor privilege. +✖ ! [rocm6_4_3] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_3__hblt0.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_3__hblt0.log new file mode 100644 index 0000000..2add86a --- /dev/null +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_3__hblt0.log @@ -0,0 +1,6 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +HW Exception by GPU node-1 (Agent handle: 0x2a344340) reason :GPU Hang +✖ ! [rocm6_4_3] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __hblt0 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_3__hblt0__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_3__hblt0__fa1.log new file mode 100644 index 0000000..977948c --- /dev/null +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_3__hblt0__fa1.log @@ -0,0 +1,6 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +Memory access fault by GPU node-1 (Agent handle: 0x27934340) on address 0x7f656656f000. Reason: Page not present or supervisor privilege. +✖ ! [rocm6_4_3] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_beta.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_beta.log deleted file mode 100644 index 1441b69..0000000 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 999 | 0 | pp512 | 262.13 ± 9.71 | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 999 | 0 | tg128 | 11.65 ± 0.01 | - -build: 79c1160b (6123) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_beta__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_beta__fa1.log deleted file mode 100644 index 4dc7de6..0000000 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_beta__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -Memory access fault by GPU node-1 (Agent handle: 0x2b4130e0) on address 0x7f8a7ed6f000. Reason: Page not present or supervisor privilege. -✖ ! [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_beta__hblt0.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_beta__hblt0.log deleted file mode 100644 index 236d063..0000000 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_beta__hblt0.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x12790e0) reason :GPU Hang -✖ ! [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __hblt0 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_beta__hblt0__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_beta__hblt0__fa1.log deleted file mode 100644 index 3db64ad..0000000 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_beta__hblt0__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -Memory access fault by GPU node-1 (Agent handle: 0x14e4a0e0) on address 0x7f859916f000. Reason: Page not present or supervisor privilege. -✖ ! [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc-rocwmma.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc-rocwmma.log index 930340d..03112eb 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc-rocwmma.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc-rocwmma.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 999 | 0 | pp512 | 267.45 ± 1.90 | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 999 | 0 | tg128 | 11.60 ± 0.05 | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 99 | 0 | pp512 | 279.13 ± 2.90 | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 99 | 0 | tg128 | 11.79 ± 0.07 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc-rocwmma__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc-rocwmma__fa1.log index 1ee598c..182bf9d 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc-rocwmma__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 999 | 1 | 0 | pp512 | 293.37 ± 7.08 | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 999 | 1 | 0 | tg128 | 11.54 ± 0.03 | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 99 | 1 | 0 | pp512 | 293.60 ± 3.84 | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 99 | 1 | 0 | tg128 | 11.62 ± 0.02 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc-rocwmma__hblt0.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc-rocwmma__hblt0.log index b62035e..66c4a41 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc-rocwmma__hblt0.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc-rocwmma__hblt0.log @@ -2,4 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __hblt0 failed (exit 134) +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 99 | 0 | pp512 | 264.02 ± 2.74 | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 99 | 0 | tg128 | 11.79 ± 0.06 | + +build: de219279 (6181) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc-rocwmma__hblt0__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc-rocwmma__hblt0__fa1.log index f49c465..5439b6b 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc-rocwmma__hblt0__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc-rocwmma__hblt0__fa1.log @@ -2,4 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __hblt0__fa1 failed (exit 134) +| model | size | params | backend | ngl | fa | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 99 | 1 | 0 | pp512 | 279.69 ± 2.30 | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 99 | 1 | 0 | tg128 | 11.60 ± 0.04 | + +build: de219279 (6181) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc.log index 8f33074..f3af1c9 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 999 | 0 | pp512 | 272.38 ± 1.28 | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 999 | 0 | tg128 | 11.64 ± 0.01 | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 99 | 0 | pp512 | 252.38 ± 7.70 | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 99 | 0 | tg128 | 11.35 ± 0.60 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc__fa1.log index 3ccfa82..9932c99 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc__fa1.log @@ -2,4 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +Memory access fault by GPU node-1 (Agent handle: 0x2e56aec0) on address 0x7f4102f6f000. Reason: Page not present or supervisor privilege. ✖ ! [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc__hblt0.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc__hblt0.log index 2758045..d31f283 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc__hblt0.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc__hblt0.log @@ -2,4 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __hblt0 failed (exit 134) +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 99 | 0 | pp512 | 271.54 ± 4.10 | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | ROCm | 99 | 0 | tg128 | 11.57 ± 0.58 | + +build: de219279 (6181) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc__hblt0__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc__hblt0__fa1.log index 0ab337b..986706b 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc__hblt0__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc__hblt0__fa1.log @@ -2,4 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +Memory access fault by GPU node-1 (Agent handle: 0x1aa83ec0) on address 0x7f9f1e96f000. Reason: Page not present or supervisor privilege. ✖ ! [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_amdvlk.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_amdvlk.log index 40cd552..089be05 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_amdvlk.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_amdvlk.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | Vulkan | 999 | 0 | pp512 | 255.55 ± 1.38 | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | Vulkan | 999 | 0 | tg128 | 12.27 ± 0.01 | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | Vulkan | 99 | 0 | pp512 | 258.54 ± 1.39 | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | Vulkan | 99 | 0 | tg128 | 12.45 ± 0.01 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_amdvlk__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_amdvlk__fa1.log index e8041dc..42f4672 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_amdvlk__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_amdvlk__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | Vulkan | 999 | 1 | 0 | pp512 | 259.07 ± 1.30 | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | Vulkan | 999 | 1 | 0 | tg128 | 12.11 ± 0.01 | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | Vulkan | 99 | 1 | 0 | pp512 | 262.84 ± 1.39 | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | Vulkan | 99 | 1 | 0 | tg128 | 12.30 ± 0.01 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_radv.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_radv.log index e52091e..e7df5fa 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_radv.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_radv.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | Vulkan | 999 | 0 | pp512 | 168.01 ± 0.85 | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | Vulkan | 999 | 0 | tg128 | 12.30 ± 0.00 | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | Vulkan | 99 | 0 | pp512 | 169.23 ± 0.84 | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | Vulkan | 99 | 0 | tg128 | 12.45 ± 0.00 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_radv__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_radv__fa1.log index 154221d..2776458 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_radv__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_radv__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | Vulkan | 999 | 1 | 0 | pp512 | 172.71 ± 0.91 | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | Vulkan | 999 | 1 | 0 | tg128 | 12.28 ± 0.00 | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | Vulkan | 99 | 1 | 0 | pp512 | 173.79 ± 0.85 | +| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | Vulkan | 99 | 1 | 0 | tg128 | 12.44 ± 0.01 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log index f13eb82..90103a9 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log @@ -2,9 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 0 | pp512 | 137.82 ± 0.73 | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 0 | tg128 | 17.41 ± 0.00 | - -build: 79c1160b (6123) +Memory access fault by GPU node-1 (Agent handle: 0x3e5ce540) on address 0x7f64d3b76000. Reason: Page not present or supervisor privilege. +✖ ! [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log index 5cd6f40..2e11ead 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log @@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x1624d050) reason :GPU Hang +HW Exception by GPU node-1 (Agent handle: 0x1239e540) reason :GPU Hang ✖ ! [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2.log index d0be2b8..b311bf5 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2.log @@ -2,9 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 999 | 0 | pp512 | 137.63 ± 0.80 | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 999 | 0 | tg128 | 17.29 ± 0.00 | - -build: 79c1160b (6123) +HW Exception by GPU node-1 (Agent handle: 0x101f4540) reason :GPU Hang +✖ ! [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2__fa1.log index e7a2f72..8ac1834 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2__fa1.log @@ -2,9 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 999 | 1 | 0 | pp512 | 122.98 ± 0.59 | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 999 | 1 | 0 | tg128 | 17.53 ± 0.00 | - -build: 79c1160b (6123) +Memory access fault by GPU node-1 (Agent handle: 0x15f12540) on address 0x7ef17d976000. Reason: Page not present or supervisor privilege. +✖ ! [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_3-rocwmma.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_3-rocwmma.log new file mode 100644 index 0000000..f500cd5 --- /dev/null +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_3-rocwmma.log @@ -0,0 +1,15 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +hipBLASLt error: Heuristic Fetch Failed! +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 0 | pp512 | 285.51 ± 1.64 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 0 | tg128 | 17.70 ± 0.01 | + +build: 1fe00296 (6182) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_beta__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_3-rocwmma__fa1.log similarity index 67% rename from benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_beta__fa1.log rename to benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_3-rocwmma__fa1.log index c46ae93..85f9bfb 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_beta__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_3-rocwmma__fa1.log @@ -1,10 +1,10 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 999 | 1 | 0 | pp512 | 233.14 ± 0.90 | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 999 | 1 | 0 | tg128 | 17.59 ± 0.00 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 1 | 0 | pp512 | 308.62 ± 2.62 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 1 | 0 | tg128 | 17.54 ± 0.00 | -build: 79c1160b (6123) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_3-rocwmma__hblt0.log similarity index 78% rename from benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2.log rename to benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_3-rocwmma__hblt0.log index 3ec496d..78e4255 100644 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_3-rocwmma__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 999 | 0 | pp512 | 138.27 ± 0.66 | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 999 | 0 | tg128 | 17.40 ± 0.00 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 0 | pp512 | 137.71 ± 0.62 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 0 | tg128 | 17.71 ± 0.01 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_3-rocwmma__hblt0__fa1.log similarity index 68% rename from benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc__fa1.log rename to benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_3-rocwmma__hblt0__fa1.log index bde171a..8945a72 100644 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_3-rocwmma__hblt0__fa1.log @@ -1,10 +1,10 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 999 | 1 | 0 | pp512 | 123.58 ± 0.18 | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 999 | 1 | 0 | tg128 | 17.55 ± 0.00 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 1 | 0 | pp512 | 142.62 ± 0.82 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 1 | 0 | tg128 | 17.55 ± 0.00 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_3.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_3.log new file mode 100644 index 0000000..5aa96bf --- /dev/null +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_3.log @@ -0,0 +1,15 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +rocBLAS error: No hipBLASLt solution found +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 0 | pp512 | 286.37 ± 1.44 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 0 | tg128 | 17.70 ± 0.00 | + +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_beta__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_3__fa1.log similarity index 67% rename from benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_beta__fa1.log rename to benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_3__fa1.log index ee0d484..98b05eb 100644 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_beta__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_3__fa1.log @@ -1,10 +1,10 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 999 | 1 | 0 | pp512 | 123.61 ± 0.50 | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 999 | 1 | 0 | tg128 | 17.60 ± 0.00 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 1 | 0 | pp512 | 234.68 ± 1.31 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 1 | 0 | tg128 | 17.71 ± 0.00 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_3__hblt0.log similarity index 55% rename from benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log rename to benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_3__hblt0.log index 2da2c5e..186cdf8 100644 --- a/benchmark/results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_3__hblt0.log @@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0xd98d050) reason :GPU Hang -✖ ! [rocm6_4_2-rocwmma] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 failed (exit 134) +HW Exception by GPU node-1 (Agent handle: 0xa3c7340) reason :GPU Hang +✖ ! [rocm6_4_3] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __hblt0 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_3__hblt0__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_3__hblt0__fa1.log new file mode 100644 index 0000000..19b75a4 --- /dev/null +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_3__hblt0__fa1.log @@ -0,0 +1,6 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +HW Exception by GPU node-1 (Agent handle: 0x118c6340) reason :GPU Hang +✖ ! [rocm6_4_3] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_beta.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_beta.log deleted file mode 100644 index cc14e7d..0000000 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 999 | 0 | pp512 | 281.87 ± 1.98 | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 999 | 0 | tg128 | 17.59 ± 0.01 | - -build: 79c1160b (6123) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_beta__hblt0.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_beta__hblt0.log deleted file mode 100644 index 05da15f..0000000 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_beta__hblt0.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x2334b0e0) reason :GPU Hang -✖ ! [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __hblt0 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_beta__hblt0__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_beta__hblt0__fa1.log deleted file mode 100644 index 20fd2ca..0000000 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_beta__hblt0__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x1b1f20f0) reason :GPU Hang -✖ ! [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma.log index 60b7302..54d6795 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma.log @@ -2,4 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc-rocwmma] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 failed (exit 134) +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 0 | pp512 | 290.54 ± 1.59 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 0 | tg128 | 17.67 ± 0.01 | + +build: de219279 (6181) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__fa1.log index 4935293..6f81260 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 999 | 1 | 0 | pp512 | 307.08 ± 2.67 | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 999 | 1 | 0 | tg128 | 17.34 ± 0.01 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 1 | 0 | pp512 | 304.99 ± 0.37 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 1 | 0 | tg128 | 17.28 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0.log index ac13496..721ffc5 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 0 | pp512 | 138.22 ± 0.46 | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 0 | tg128 | 17.45 ± 0.09 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 0 | pp512 | 283.93 ± 1.57 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 0 | tg128 | 17.65 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log index 12bb02b..241fb21 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log @@ -2,4 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc-rocwmma] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __hblt0__fa1 failed (exit 134) +| model | size | params | backend | ngl | fa | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 1 | 0 | pp512 | 300.13 ± 1.26 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 1 | 0 | tg128 | 17.27 ± 0.00 | + +build: de219279 (6181) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc.log index 1e4897a..19895f7 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 999 | 0 | pp512 | 281.24 ± 1.95 | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 999 | 0 | tg128 | 17.56 ± 0.00 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 0 | pp512 | 291.60 ± 1.95 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 0 | tg128 | 17.73 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc__fa1.log index 9eb1c08..d607e30 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc__fa1.log @@ -2,4 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +Memory access fault by GPU node-1 (Agent handle: 0x134adec0) on address 0x7f0318984000. Reason: Page not present or supervisor privilege. ✖ ! [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc__hblt0.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc__hblt0.log index ca4dda3..a0c7756 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc__hblt0.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc__hblt0.log @@ -2,4 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __hblt0 failed (exit 134) +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 0 | pp512 | 285.56 ± 1.41 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 0 | tg128 | 17.72 ± 0.00 | + +build: de219279 (6181) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc__hblt0__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc__hblt0__fa1.log index 9086eec..ae6829f 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc__hblt0__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc__hblt0__fa1.log @@ -2,4 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __hblt0__fa1 failed (exit 134) +| model | size | params | backend | ngl | fa | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 1 | 0 | pp512 | 227.75 ± 1.52 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 99 | 1 | 0 | tg128 | 17.73 ± 0.00 | + +build: de219279 (6181) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk.log index f03b1f2..f89d42c 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | Vulkan | 999 | 0 | pp512 | 218.27 ± 0.80 | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | Vulkan | 999 | 0 | tg128 | 20.09 ± 0.01 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | Vulkan | 99 | 0 | pp512 | 216.64 ± 2.76 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | Vulkan | 99 | 0 | tg128 | 20.39 ± 0.02 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log index 2706deb..7ee60d9 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | Vulkan | 999 | 1 | 0 | pp512 | 220.73 ± 0.69 | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | Vulkan | 999 | 1 | 0 | tg128 | 19.64 ± 0.00 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | Vulkan | 99 | 1 | 0 | pp512 | 217.68 ± 4.15 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | Vulkan | 99 | 1 | 0 | tg128 | 19.97 ± 0.01 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_radv.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_radv.log index 8058f91..002154b 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_radv.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_radv.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | Vulkan | 999 | 0 | pp512 | 152.77 ± 0.73 | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | Vulkan | 999 | 0 | tg128 | 20.02 ± 0.01 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | Vulkan | 99 | 0 | pp512 | 151.98 ± 0.60 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | Vulkan | 99 | 0 | tg128 | 20.26 ± 0.02 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_radv__fa1.log b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_radv__fa1.log index 953d42a..64d4625 100644 --- a/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_radv__fa1.log +++ b/benchmark/results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_radv__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | Vulkan | 999 | 1 | 0 | pp512 | 155.24 ± 1.01 | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | Vulkan | 999 | 1 | 0 | tg128 | 19.99 ± 0.00 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | Vulkan | 99 | 1 | 0 | pp512 | 154.96 ± 0.82 | +| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | Vulkan | 99 | 1 | 0 | tg128 | 20.28 ± 0.01 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2-rocwmma.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2-rocwmma.log index 24b7806..1adbae9 100644 --- a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2-rocwmma.log +++ b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2-rocwmma.log @@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x3eeda050) reason :GPU Hang +HW Exception by GPU node-1 (Agent handle: 0x2f5d1540) reason :GPU Hang ✖ ! [rocm6_4_2-rocwmma] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 failed (exit 134) diff --git a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2-rocwmma__fa1.log index c0f73f3..9a061f5 100644 --- a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2-rocwmma__fa1.log @@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x2d723050) reason :GPU Hang +HW Exception by GPU node-1 (Agent handle: 0xdc93540) reason :GPU Hang ✖ ! [rocm6_4_2-rocwmma] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 failed (exit 134) diff --git a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2.log index 95b0795..a1be58c 100644 --- a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2.log +++ b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2.log @@ -2,9 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 999 | 0 | pp512 | 73.83 ± 0.16 | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 999 | 0 | tg128 | 13.68 ± 0.01 | - -build: 79c1160b (6123) +HW Exception by GPU node-1 (Agent handle: 0xff7540) reason :GPU Hang +✖ ! [rocm6_4_2] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 failed (exit 134) diff --git a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2__fa1.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2__fa1.log index 22c3d0b..281a126 100644 --- a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2__fa1.log +++ b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2__fa1.log @@ -2,9 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 999 | 1 | 0 | pp512 | 61.47 ± 0.09 | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 999 | 1 | 0 | tg128 | 13.83 ± 0.00 | - -build: 79c1160b (6123) +HW Exception by GPU node-1 (Agent handle: 0x2607e540) reason :GPU Hang +✖ ! [rocm6_4_2] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_3-rocwmma.log similarity index 54% rename from benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2.log rename to benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_3-rocwmma.log index 1bc098e..d927ed6 100644 --- a/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2.log +++ b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_3-rocwmma.log @@ -2,9 +2,14 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +rocBLAS error: No hipBLASLt solution found +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 999 | 0 | pp512 | 74.15 ± 0.18 | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 999 | 0 | tg128 | 13.73 ± 0.00 | +| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 99 | 0 | pp512 | 130.11 ± 0.68 | +| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 99 | 0 | tg128 | 13.95 ± 0.04 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_3-rocwmma__fa1.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_3-rocwmma__fa1.log new file mode 100644 index 0000000..4638f26 --- /dev/null +++ b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_3-rocwmma__fa1.log @@ -0,0 +1,10 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +| model | size | params | backend | ngl | fa | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | +| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 99 | 1 | 0 | pp512 | 144.31 ± 0.80 | +| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 99 | 1 | 0 | tg128 | 13.71 ± 0.00 | + +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_3-rocwmma__hblt0.log similarity index 54% rename from benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log rename to benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_3-rocwmma__hblt0.log index a418f5b..d8218bd 100644 --- a/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log +++ b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_3-rocwmma__hblt0.log @@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0xcd80a90) reason :GPU Hang -✖ ! [rocm6_4_2-rocwmma] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 failed (exit 134) +HW Exception by GPU node-1 (Agent handle: 0x8063340) reason :GPU Hang +✖ ! [rocm6_4_3-rocwmma] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __hblt0 failed (exit 134) diff --git a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_3-rocwmma__hblt0__fa1.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_3-rocwmma__hblt0__fa1.log new file mode 100644 index 0000000..3a247e3 --- /dev/null +++ b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_3-rocwmma__hblt0__fa1.log @@ -0,0 +1,6 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +HW Exception by GPU node-1 (Agent handle: 0x18398340) reason :GPU Hang +✖ ! [rocm6_4_3-rocwmma] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_3.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_3.log new file mode 100644 index 0000000..82caf25 --- /dev/null +++ b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_3.log @@ -0,0 +1,15 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +rocBLAS error: No hipBLASLt solution found +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 99 | 0 | pp512 | 131.78 ± 1.03 | +| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 99 | 0 | tg128 | 13.68 ± 0.43 | + +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2__fa1.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_3__fa1.log similarity index 53% rename from benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2__fa1.log rename to benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_3__fa1.log index 40b3223..a0ed178 100644 --- a/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2__fa1.log +++ b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_3__fa1.log @@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -Memory access fault by GPU node-1 (Agent handle: 0x25011a90) on address 0x7fdcc1b6f000. Reason: Page not present or supervisor privilege. -✖ ! [rocm6_4_2] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 failed (exit 134) +Memory access fault by GPU node-1 (Agent handle: 0x50aa340) on address 0x7f7365ba9000. Reason: Page not present or supervisor privilege. +✖ ! [rocm6_4_3] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 failed (exit 134) diff --git a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_3__hblt0.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_3__hblt0.log new file mode 100644 index 0000000..3767019 --- /dev/null +++ b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_3__hblt0.log @@ -0,0 +1,6 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +HW Exception by GPU node-1 (Agent handle: 0x1990d340) reason :GPU Hang +✖ ! [rocm6_4_3] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __hblt0 failed (exit 134) diff --git a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_3__hblt0__fa1.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_3__hblt0__fa1.log new file mode 100644 index 0000000..f1dc1f5 --- /dev/null +++ b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_3__hblt0__fa1.log @@ -0,0 +1,6 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +Memory access fault by GPU node-1 (Agent handle: 0x180d4340) on address 0x7f11c8f6f000. Reason: Page not present or supervisor privilege. +✖ ! [rocm6_4_3] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_beta.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_beta.log deleted file mode 100644 index c4bfe32..0000000 --- a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_beta.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x359cb0e0) reason :GPU Hang -✖ ! [rocm7_beta] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 failed (exit 134) diff --git a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_beta__fa1.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_beta__fa1.log deleted file mode 100644 index 4707f93..0000000 --- a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_beta__fa1.log +++ /dev/null @@ -1,7 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -:0:rocdevice.cpp :3675: 456558403486 us: Callback: Queue 0x7f04ef600000 aborting with error : HSA_STATUS_ERROR_EXCEPTION: An HSAIL operation resulted in a hardware exception. code: 0x1016 -Memory access fault by GPU node-1 (Agent handle: 0x2e8f0f0) on address 0x7eeca7f6f000. Reason: Page not present or supervisor privilege. -✖ ! [rocm7_beta] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 failed (exit 134) diff --git a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_beta__hblt0.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_beta__hblt0.log deleted file mode 100644 index 4554363..0000000 --- a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_beta__hblt0.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x1c2260e0) reason :GPU Hang -✖ ! [rocm7_beta] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __hblt0 failed (exit 134) diff --git a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_beta__hblt0__fa1.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_beta__hblt0__fa1.log deleted file mode 100644 index d31397b..0000000 --- a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_beta__hblt0__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -Memory access fault by GPU node-1 (Agent handle: 0x11f900f0) on address 0x7f6f91d6f000. Reason: Page not present or supervisor privilege. -✖ ! [rocm7_beta] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc-rocwmma.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc-rocwmma.log index 440a82e..38826b0 100644 --- a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc-rocwmma.log +++ b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc-rocwmma.log @@ -2,9 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 999 | 0 | pp512 | 129.70 ± 0.81 | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 999 | 0 | tg128 | 13.66 ± 0.00 | - -build: 34c9d765 (6122) +Memory access fault by GPU node-1 (Agent handle: 0x3c89ef80) on address 0x7f777640a000. Reason: Page not present or supervisor privilege. +✖ ! [rocm7_rc-rocwmma] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 failed (exit 134) diff --git a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc-rocwmma__fa1.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc-rocwmma__fa1.log index 15a84cb..a74258c 100644 --- a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc-rocwmma__fa1.log +++ b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 999 | 1 | 0 | pp512 | 145.18 ± 0.48 | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 999 | 1 | 0 | tg128 | 13.43 ± 0.00 | +| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 99 | 1 | 0 | pp512 | 141.61 ± 0.92 | +| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 99 | 1 | 0 | tg128 | 13.34 ± 0.02 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc-rocwmma__hblt0.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc-rocwmma__hblt0.log index 7597916..966ba3e 100644 --- a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc-rocwmma__hblt0.log +++ b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc-rocwmma__hblt0.log @@ -2,4 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc-rocwmma] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __hblt0 failed (exit 134) +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 99 | 0 | pp512 | 133.33 ± 0.68 | +| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 99 | 0 | tg128 | 13.78 ± 0.04 | + +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc-rocwmma__hblt0__fa1.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc-rocwmma__hblt0__fa1.log index af8634b..22a4e47 100644 --- a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc-rocwmma__hblt0__fa1.log +++ b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc-rocwmma__hblt0__fa1.log @@ -2,4 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc-rocwmma] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __hblt0__fa1 failed (exit 134) +| model | size | params | backend | ngl | fa | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | +| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 99 | 1 | 0 | pp512 | 139.60 ± 0.47 | +| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 99 | 1 | 0 | tg128 | 13.03 ± 0.57 | + +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc.log index d8318fb..daaa296 100644 --- a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc.log +++ b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 999 | 0 | pp512 | 130.56 ± 0.46 | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 999 | 0 | tg128 | 13.87 ± 0.02 | +| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 99 | 0 | pp512 | 134.95 ± 0.76 | +| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 99 | 0 | tg128 | 13.99 ± 0.01 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc__fa1.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc__fa1.log index 3d86ec7..0d3acaf 100644 --- a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc__fa1.log +++ b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc__fa1.log @@ -2,9 +2,6 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 999 | 1 | 0 | pp512 | 97.08 ± 0.34 | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 999 | 1 | 0 | tg128 | 13.90 ± 0.03 | - -build: 79c1160b (6123) +:0:rocdevice.cpp :3675: 29915649820 us: Callback: Queue 0x7f500c700000 aborting with error : HSA_STATUS_ERROR_EXCEPTION: An HSAIL operation resulted in a hardware exception. code: 0x1016 +Memory access fault by GPU node-1 (Agent handle: 0x2aad9ec0) on address 0x7f37c576f000. Reason: Page not present or supervisor privilege. +✖ ! [rocm7_rc] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 failed (exit 134) diff --git a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc__hblt0.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc__hblt0.log index 3cb748b..8135bc1 100644 --- a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc__hblt0.log +++ b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc__hblt0.log @@ -2,4 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __hblt0 failed (exit 134) +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 99 | 0 | pp512 | 135.29 ± 0.51 | +| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | ROCm | 99 | 0 | tg128 | 13.97 ± 0.04 | + +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc__hblt0__fa1.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc__hblt0__fa1.log index 443bad3..fc4a37f 100644 --- a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc__hblt0__fa1.log +++ b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc__hblt0__fa1.log @@ -2,4 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +Memory access fault by GPU node-1 (Agent handle: 0x122f2ec0) on address 0x7f10537a9000. Reason: Page not present or supervisor privilege. ✖ ! [rocm7_rc] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_amdvlk.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_amdvlk.log index b14c784..ec978bb 100644 --- a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_amdvlk.log +++ b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_amdvlk.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | Vulkan | 999 | 0 | pp512 | 114.76 ± 0.62 | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | Vulkan | 999 | 0 | tg128 | 16.06 ± 0.00 | +| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | Vulkan | 99 | 0 | pp512 | 112.93 ± 0.63 | +| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | Vulkan | 99 | 0 | tg128 | 16.43 ± 0.01 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_amdvlk__fa1.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_amdvlk__fa1.log index c01f816..e560c0b 100644 --- a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_amdvlk__fa1.log +++ b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_amdvlk__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | Vulkan | 999 | 1 | 0 | pp512 | 116.18 ± 0.67 | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | Vulkan | 999 | 1 | 0 | tg128 | 15.90 ± 0.01 | +| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | Vulkan | 99 | 1 | 0 | pp512 | 114.35 ± 1.12 | +| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | Vulkan | 99 | 1 | 0 | tg128 | 16.27 ± 0.01 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_radv.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_radv.log index 077cf15..5fbf704 100644 --- a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_radv.log +++ b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_radv.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | Vulkan | 999 | 0 | pp512 | 64.79 ± 0.39 | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | Vulkan | 999 | 0 | tg128 | 16.61 ± 0.00 | +| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | Vulkan | 99 | 0 | pp512 | 64.60 ± 0.38 | +| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | Vulkan | 99 | 0 | tg128 | 17.03 ± 0.01 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_radv__fa1.log b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_radv__fa1.log index d9c6cb3..e878778 100644 --- a/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_radv__fa1.log +++ b/benchmark/results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_radv__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | Vulkan | 999 | 1 | 0 | pp512 | 66.84 ± 0.42 | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | Vulkan | 999 | 1 | 0 | tg128 | 16.86 ± 0.01 | +| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | Vulkan | 99 | 1 | 0 | pp512 | 66.60 ± 0.42 | +| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | Vulkan | 99 | 1 | 0 | tg128 | 17.28 ± 0.01 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2-rocwmma.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2-rocwmma.log index 562318d..4f8da20 100644 --- a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2-rocwmma.log +++ b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2-rocwmma.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 157.78 ± 2.71 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 24.56 ± 0.00 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 157.75 ± 2.58 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 24.62 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log index 77ad01b..598f905 100644 --- a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 161.64 ± 2.99 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 23.94 ± 0.00 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 161.90 ± 3.05 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 24.09 ± 0.02 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2.log index 9a6adc0..ae13cea 100644 --- a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2.log +++ b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | pp512 | 157.64 ± 2.49 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | tg128 | 23.93 ± 0.01 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 157.81 ± 2.51 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 24.61 ± 0.01 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2__fa1.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2__fa1.log index 000d477..2791323 100644 --- a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2__fa1.log +++ b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 1 | 0 | pp512 | 140.32 ± 1.99 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 1 | 0 | tg128 | 24.32 ± 0.00 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 140.24 ± 1.86 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 24.46 ± 0.02 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2-rocwmma.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_3-rocwmma.log similarity index 54% rename from benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2-rocwmma.log rename to benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_3-rocwmma.log index 3e8cb4e..cd9f858 100644 --- a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2-rocwmma.log +++ b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_3-rocwmma.log @@ -2,9 +2,14 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +hipBLASLt error: Heuristic Fetch Failed! +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 150.37 ± 1.75 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 24.49 ± 0.01 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 438.42 ± 4.14 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 24.57 ± 0.01 | -build: 79c1160b (6123) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_3-rocwmma__fa1.log similarity index 79% rename from benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log rename to benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_3-rocwmma__fa1.log index af7dc3f..3adfafe 100644 --- a/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_3-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 1 | 0 | pp512 | 162.19 ± 3.06 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 1 | 0 | tg128 | 24.03 ± 0.00 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 475.43 ± 7.40 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 24.08 ± 0.00 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2-rocwmma.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_3-rocwmma__hblt0.log similarity index 79% rename from benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2-rocwmma.log rename to benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_3-rocwmma__hblt0.log index 3c0cef6..96ee7c7 100644 --- a/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2-rocwmma.log +++ b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_3-rocwmma__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | pp512 | 157.95 ± 2.63 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | tg128 | 24.53 ± 0.01 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 158.13 ± 2.40 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 24.58 ± 0.00 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2__fa1.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_3-rocwmma__hblt0__fa1.log similarity index 79% rename from benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2__fa1.log rename to benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_3-rocwmma__hblt0__fa1.log index 86ac559..e330db0 100644 --- a/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2__fa1.log +++ b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_3-rocwmma__hblt0__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 1 | 0 | pp512 | 140.32 ± 2.10 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 1 | 0 | tg128 | 24.33 ± 0.00 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 163.40 ± 3.21 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 24.14 ± 0.00 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_3.log similarity index 54% rename from benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2.log rename to benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_3.log index dc8823d..f983fbc 100644 --- a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2.log +++ b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_3.log @@ -2,9 +2,14 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +hipBLASLt error: Heuristic Fetch Failed! +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | pp512 | 150.06 ± 1.71 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | tg128 | 23.13 ± 0.00 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 441.36 ± 3.35 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 24.60 ± 0.01 | -build: 79c1160b (6123) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_3__fa1.log similarity index 78% rename from benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log rename to benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_3__fa1.log index ded0220..04c9b9c 100644 --- a/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_3__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 1 | 0 | pp512 | 154.09 ± 1.98 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 1 | 0 | tg128 | 24.02 ± 0.01 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 337.36 ± 3.48 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 24.45 ± 0.00 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_3__hblt0.log similarity index 79% rename from benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2.log rename to benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_3__hblt0.log index 03365ca..c1ff6cb 100644 --- a/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2.log +++ b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_3__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | pp512 | 157.69 ± 2.52 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | tg128 | 23.89 ± 0.01 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 161.73 ± 1.23 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 24.58 ± 0.01 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2__fa1.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_3__hblt0__fa1.log similarity index 79% rename from benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2__fa1.log rename to benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_3__hblt0__fa1.log index 207a2a1..aaa3b75 100644 --- a/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2__fa1.log +++ b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_3__hblt0__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 1 | 0 | pp512 | 134.40 ± 1.47 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 1 | 0 | tg128 | 24.32 ± 0.01 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 143.05 ± 2.10 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 24.42 ± 0.01 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_beta.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_beta.log deleted file mode 100644 index 6b685b0..0000000 --- a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | pp512 | 424.74 ± 7.06 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | tg128 | 24.48 ± 0.01 | - -build: 79c1160b (6123) diff --git a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_beta__fa1.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_beta__fa1.log deleted file mode 100644 index 2940f7a..0000000 --- a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_beta__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -Memory access fault by GPU node-1 (Agent handle: 0x16acc0c0) on address 0x7f24fed6f000. Reason: Page not present or supervisor privilege. -✖ ! [rocm7_beta] Qwen3-30B-A3B-BF16-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_beta__hblt0.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_beta__hblt0.log deleted file mode 100644 index b98ad69..0000000 --- a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_beta__hblt0.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 154.45 ± 1.39 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 24.52 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_beta__hblt0__fa1.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_beta__hblt0__fa1.log deleted file mode 100644 index 8773673..0000000 --- a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_beta__hblt0__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 138.46 ± 1.64 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 24.29 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc-rocwmma.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc-rocwmma.log index 5088e35..7f744f7 100644 --- a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc-rocwmma.log +++ b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc-rocwmma.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | pp512 | 425.56 ± 3.28 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | tg128 | 24.80 ± 0.00 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 448.63 ± 5.90 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 24.96 ± 0.02 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc-rocwmma__fa1.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc-rocwmma__fa1.log index 20a8ebc..61f996e 100644 --- a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc-rocwmma__fa1.log +++ b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 1 | 0 | pp512 | 472.05 ± 4.59 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 1 | 0 | tg128 | 24.12 ± 0.00 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 473.34 ± 8.60 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 23.99 ± 0.01 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc-rocwmma__hblt0.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc-rocwmma__hblt0.log index 6fb55b7..8b322b7 100644 --- a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc-rocwmma__hblt0.log +++ b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc-rocwmma__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 153.54 ± 2.25 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 24.74 ± 0.01 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 338.07 ± 3.03 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 24.93 ± 0.03 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log index e0dd74e..a3675a1 100644 --- a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log +++ b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 158.20 ± 2.47 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 24.12 ± 0.00 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 342.57 ± 3.12 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 23.97 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc.log index 1c911fe..c7f3224 100644 --- a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc.log +++ b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | pp512 | 426.72 ± 7.55 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | tg128 | 24.57 ± 0.00 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 444.30 ± 6.78 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 24.66 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc__fa1.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc__fa1.log index d2a18a1..8384bda 100644 --- a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc__fa1.log +++ b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc__fa1.log @@ -2,4 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +Memory access fault by GPU node-1 (Agent handle: 0x38fecea0) on address 0x7f31ea76f000. Reason: Page not present or supervisor privilege. ✖ ! [rocm7_rc] Qwen3-30B-A3B-BF16-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc__hblt0.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc__hblt0.log index 25f3bf6..aab676c 100644 --- a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc__hblt0.log +++ b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 153.89 ± 1.73 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 24.57 ± 0.00 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 333.42 ± 6.83 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 24.69 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc__hblt0__fa1.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc__hblt0__fa1.log index 53b2312..dba441c 100644 --- a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc__hblt0__fa1.log +++ b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc__hblt0__fa1.log @@ -2,9 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 137.06 ± 2.00 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 24.32 ± 0.01 | - -build: 79c1160b (6123) +Memory access fault by GPU node-1 (Agent handle: 0x1f121ea0) on address 0x7fd78e16f000. Reason: Page not present or supervisor privilege. +✖ ! [rocm7_rc] Qwen3-30B-A3B-BF16-00001-of-00002 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_amdvlk.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_amdvlk.log index fc1a60f..67aa5d3 100644 --- a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_amdvlk.log +++ b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_amdvlk.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 0 | pp512 | 107.55 ± 0.11 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 0 | tg128 | 8.09 ± 0.00 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 99 | 0 | pp512 | 106.47 ± 0.10 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 99 | 0 | tg128 | 8.18 ± 0.00 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_amdvlk__fa1.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_amdvlk__fa1.log index f84135e..de3ec24 100644 --- a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_amdvlk__fa1.log +++ b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_amdvlk__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 1 | 0 | pp512 | 107.68 ± 0.13 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 1 | 0 | tg128 | 8.03 ± 0.00 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 99 | 1 | 0 | pp512 | 106.77 ± 0.12 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 99 | 1 | 0 | tg128 | 8.11 ± 0.00 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_radv.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_radv.log index 1458372..7ea35f3 100644 --- a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_radv.log +++ b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_radv.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 0 | pp512 | 86.02 ± 0.11 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 0 | tg128 | 7.46 ± 0.00 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 99 | 0 | pp512 | 84.71 ± 0.11 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 99 | 0 | tg128 | 7.52 ± 0.00 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_radv__fa1.log b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_radv__fa1.log index ec2a50a..2aa8bc0 100644 --- a/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_radv__fa1.log +++ b/benchmark/results/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_radv__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 1 | 0 | pp512 | 86.93 ± 0.15 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 1 | 0 | tg128 | 7.44 ± 0.00 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 99 | 1 | 0 | pp512 | 85.70 ± 0.10 | +| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 99 | 1 | 0 | tg128 | 7.52 ± 0.00 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2-rocwmma.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2-rocwmma.log index 015e9b4..14182c9 100644 --- a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2-rocwmma.log +++ b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2-rocwmma.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 387.45 ± 1.17 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 50.42 ± 0.01 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 387.23 ± 0.82 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 50.64 ± 0.01 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2-rocwmma__fa1.log index a5d6d7e..63e83fb 100644 --- a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 411.60 ± 0.78 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 48.14 ± 0.01 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 411.72 ± 1.04 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 48.78 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2.log index 4565d4e..f33f7c4 100644 --- a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2.log +++ b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 0 | pp512 | 385.52 ± 0.67 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 0 | tg128 | 50.06 ± 0.01 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 387.86 ± 1.41 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 50.65 ± 0.01 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2__fa1.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2__fa1.log index a0064ea..928cc4b 100644 --- a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2__fa1.log +++ b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 1 | 0 | pp512 | 300.86 ± 0.38 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 1 | 0 | tg128 | 49.71 ± 0.01 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 301.23 ± 0.49 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 50.07 ± 0.02 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_3-rocwmma.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_3-rocwmma.log new file mode 100644 index 0000000..7499112 --- /dev/null +++ b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_3-rocwmma.log @@ -0,0 +1,15 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +rocBLAS error: No hipBLASLt solution found +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 564.83 ± 6.58 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 50.68 ± 0.01 | + +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_3-rocwmma__fa1.log similarity index 79% rename from benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2-rocwmma__fa1.log rename to benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_3-rocwmma__fa1.log index 60992bd..8947515 100644 --- a/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_3-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 1 | 0 | pp512 | 412.35 ± 1.06 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 1 | 0 | tg128 | 48.26 ± 0.01 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 624.99 ± 3.81 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 48.64 ± 0.01 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2-rocwmma.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_3-rocwmma__hblt0.log similarity index 79% rename from benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2-rocwmma.log rename to benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_3-rocwmma__hblt0.log index dc6f1a9..1488828 100644 --- a/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2-rocwmma.log +++ b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_3-rocwmma__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 0 | pp512 | 388.77 ± 0.97 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 0 | tg128 | 50.31 ± 0.01 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 389.25 ± 2.01 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 50.66 ± 0.01 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2__fa1.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_3-rocwmma__hblt0__fa1.log similarity index 78% rename from benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2__fa1.log rename to benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_3-rocwmma__hblt0__fa1.log index 2a04531..d3d972e 100644 --- a/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2__fa1.log +++ b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_3-rocwmma__hblt0__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 1 | 0 | pp512 | 301.29 ± 0.54 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 1 | 0 | tg128 | 49.58 ± 0.00 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 412.18 ± 1.15 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 48.80 ± 0.01 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_3.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_3.log new file mode 100644 index 0000000..bb39f34 --- /dev/null +++ b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_3.log @@ -0,0 +1,15 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +rocBLAS error: No hipBLASLt solution found +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 562.86 ± 10.14 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 50.74 ± 0.01 | + +build: 1fe00296 (6182) diff --git a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_beta__fa1.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_3__fa1.log similarity index 68% rename from benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_beta__fa1.log rename to benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_3__fa1.log index bbdc595..e501cb6 100644 --- a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_beta__fa1.log +++ b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_3__fa1.log @@ -1,10 +1,10 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 1 | 0 | pp512 | 411.72 ± 2.56 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 1 | 0 | tg128 | 49.76 ± 0.00 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 418.07 ± 1.65 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 50.11 ± 0.01 | -build: 79c1160b (6123) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_3__hblt0.log similarity index 79% rename from benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2.log rename to benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_3__hblt0.log index bd9bc1c..87b8aec 100644 --- a/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2.log +++ b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_3__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 0 | pp512 | 388.72 ± 2.63 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 0 | tg128 | 50.19 ± 0.01 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 387.74 ± 1.70 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 50.65 ± 0.01 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_beta__hblt0__fa1.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_3__hblt0__fa1.log similarity index 68% rename from benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_beta__hblt0__fa1.log rename to benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_3__hblt0__fa1.log index 6a605f7..17416ee 100644 --- a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_beta__hblt0__fa1.log +++ b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_3__hblt0__fa1.log @@ -1,10 +1,10 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 300.58 ± 1.17 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 49.78 ± 0.00 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 301.31 ± 0.65 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 50.37 ± 0.02 | -build: 79c1160b (6123) +build: 1fe00296 (6182) diff --git a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_beta.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_beta.log deleted file mode 100644 index 7b4fa67..0000000 --- a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 0 | pp512 | 534.84 ± 2.48 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 0 | tg128 | 50.21 ± 0.01 | - -build: 79c1160b (6123) diff --git a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_beta__hblt0.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_beta__hblt0.log deleted file mode 100644 index f59b880..0000000 --- a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_beta__hblt0.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 387.34 ± 1.49 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 50.23 ± 0.01 | - -build: 79c1160b (6123) diff --git a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc-rocwmma.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc-rocwmma.log index bd849a1..444ca09 100644 --- a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc-rocwmma.log +++ b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc-rocwmma.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 0 | pp512 | 535.44 ± 6.90 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 0 | tg128 | 50.07 ± 0.01 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 570.31 ± 5.05 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 50.52 ± 0.02 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc-rocwmma__fa1.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc-rocwmma__fa1.log index 57f3363..c11bc62 100644 --- a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc-rocwmma__fa1.log +++ b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 1 | 0 | pp512 | 619.02 ± 7.73 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 1 | 0 | tg128 | 47.63 ± 0.00 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 612.79 ± 4.77 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 46.73 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc-rocwmma__hblt0.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc-rocwmma__hblt0.log index 922286f..e17b52a 100644 --- a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc-rocwmma__hblt0.log +++ b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc-rocwmma__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 387.98 ± 0.76 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 50.09 ± 0.01 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 572.09 ± 8.22 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 50.45 ± 0.01 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc-rocwmma__hblt0__fa1.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc-rocwmma__hblt0__fa1.log index a66f360..19e8321 100644 --- a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc-rocwmma__hblt0__fa1.log +++ b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc-rocwmma__hblt0__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 413.28 ± 2.05 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 47.63 ± 0.01 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 605.49 ± 1.47 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 46.73 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc.log index 625d68e..f2e3ece 100644 --- a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc.log +++ b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 0 | pp512 | 540.14 ± 5.22 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 0 | tg128 | 50.65 ± 0.01 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 573.05 ± 6.77 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 50.80 ± 0.01 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc__fa1.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc__fa1.log index 43c8e66..3fa7435 100644 --- a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc__fa1.log +++ b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 1 | 0 | pp512 | 418.60 ± 2.58 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 1 | 0 | tg128 | 49.63 ± 0.00 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 416.05 ± 3.44 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 50.33 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc__hblt0.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc__hblt0.log index 74fbd99..db8de56 100644 --- a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc__hblt0.log +++ b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 386.87 ± 1.67 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 50.50 ± 0.01 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 576.38 ± 3.91 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 50.85 ± 0.01 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc__hblt0__fa1.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc__hblt0__fa1.log index 763fdaf..f19e470 100644 --- a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc__hblt0__fa1.log +++ b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc__hblt0__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 300.40 ± 1.44 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 49.69 ± 0.00 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 414.62 ± 3.23 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 50.22 ± 0.01 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_amdvlk.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_amdvlk.log index 4f467f9..10609cc 100644 --- a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_amdvlk.log +++ b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_amdvlk.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | Vulkan | 999 | 0 | pp512 | 741.97 ± 2.92 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | Vulkan | 999 | 0 | tg128 | 57.22 ± 0.02 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | Vulkan | 99 | 0 | pp512 | 733.40 ± 2.59 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | Vulkan | 99 | 0 | tg128 | 59.36 ± 0.05 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_amdvlk__fa1.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_amdvlk__fa1.log index 8cbd25b..486113e 100644 --- a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_amdvlk__fa1.log +++ b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_amdvlk__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | Vulkan | 999 | 1 | 0 | pp512 | 731.64 ± 2.80 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | Vulkan | 999 | 1 | 0 | tg128 | 53.53 ± 0.02 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | Vulkan | 99 | 1 | 0 | pp512 | 725.54 ± 2.84 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | Vulkan | 99 | 1 | 0 | tg128 | 55.57 ± 0.02 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_radv.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_radv.log index 993ae07..00cd713 100644 --- a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_radv.log +++ b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_radv.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | Vulkan | 999 | 0 | pp512 | 396.38 ± 1.53 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | Vulkan | 999 | 0 | tg128 | 59.54 ± 0.02 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | Vulkan | 99 | 0 | pp512 | 392.54 ± 1.80 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | Vulkan | 99 | 0 | tg128 | 61.56 ± 0.02 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_radv__fa1.log b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_radv__fa1.log index 296a137..3123ba1 100644 --- a/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_radv__fa1.log +++ b/benchmark/results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_radv__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | Vulkan | 999 | 1 | 0 | pp512 | 406.84 ± 1.62 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | Vulkan | 999 | 1 | 0 | tg128 | 58.50 ± 0.10 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | Vulkan | 99 | 1 | 0 | pp512 | 403.74 ± 1.69 | +| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | Vulkan | 99 | 1 | 0 | tg128 | 60.57 ± 0.08 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2__fa1.log b/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2__fa1.log deleted file mode 100644 index d63b9d0..0000000 --- a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -Memory access fault by GPU node-1 (Agent handle: 0x168bc050) on address 0x7ef358d6f000. Reason: Page not present or supervisor privilege. -✖ ! [rocm6_4_2] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_beta.log b/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_beta.log deleted file mode 100644 index 4d0291a..0000000 --- a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | pp512 | 408.29 ± 1.82 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | tg128 | 24.53 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_beta__fa1.log b/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_beta__fa1.log deleted file mode 100644 index 0830d17..0000000 --- a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_beta__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -Memory access fault by GPU node-1 (Agent handle: 0xf2660e0) on address 0x7fb2199a9000. Reason: Page not present or supervisor privilege. -✖ ! [rocm7_beta] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_beta__hblt0.log b/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_beta__hblt0.log deleted file mode 100644 index eb240a1..0000000 --- a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_beta__hblt0.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 145.29 ± 1.91 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 24.53 ± 0.01 | - -build: 79c1160b (6123) diff --git a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_beta__hblt0__fa1.log b/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_beta__hblt0__fa1.log deleted file mode 100644 index d7843f5..0000000 --- a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_beta__hblt0__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 130.39 ± 1.57 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 24.31 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc-rocwmma.log b/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc-rocwmma.log deleted file mode 100644 index 0a1bfa8..0000000 --- a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc-rocwmma.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | pp512 | 414.47 ± 3.10 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | tg128 | 24.61 ± 0.01 | - -build: 34c9d765 (6122) diff --git a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc-rocwmma__fa1.log b/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc-rocwmma__fa1.log deleted file mode 100644 index 8c77605..0000000 --- a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc-rocwmma__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 1 | 0 | pp512 | 460.12 ± 5.58 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 1 | 0 | tg128 | 24.02 ± 0.01 | - -build: 34c9d765 (6122) diff --git a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc-rocwmma__hblt0.log b/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc-rocwmma__hblt0.log deleted file mode 100644 index 6c331f6..0000000 --- a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc-rocwmma__hblt0.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 145.43 ± 1.04 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 24.80 ± 0.00 | - -build: 34c9d765 (6122) diff --git a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log b/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log deleted file mode 100644 index b9050d5..0000000 --- a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 150.58 ± 1.93 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 24.13 ± 0.00 | - -build: 34c9d765 (6122) diff --git a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc.log b/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc.log deleted file mode 100644 index 04d0c86..0000000 --- a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | pp512 | 413.05 ± 2.36 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | tg128 | 24.15 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc__fa1.log b/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc__fa1.log deleted file mode 100644 index c7a5573..0000000 --- a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 1 | 0 | pp512 | 325.48 ± 1.77 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 1 | 0 | tg128 | 24.31 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc__hblt0.log b/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc__hblt0.log deleted file mode 100644 index c900ebe..0000000 --- a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc__hblt0.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | pp512 | 145.83 ± 2.39 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 0 | tg128 | 24.12 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc__hblt0__fa1.log b/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc__hblt0__fa1.log deleted file mode 100644 index a409195..0000000 --- a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc__hblt0__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | pp512 | 130.20 ± 1.39 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 99 | 1 | 0 | tg128 | 24.35 ± 0.01 | - -build: 79c1160b (6123) diff --git a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_amdvlk.log b/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_amdvlk.log deleted file mode 100644 index 6c09327..0000000 --- a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_amdvlk.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 0 | pp512 | 107.16 ± 0.06 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 0 | tg128 | 8.08 ± 0.00 | - -build: 34c9d765 (6122) diff --git a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_amdvlk__fa1.log b/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_amdvlk__fa1.log deleted file mode 100644 index 5f7c40c..0000000 --- a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_amdvlk__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 1 | 0 | pp512 | 107.26 ± 0.11 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 1 | 0 | tg128 | 8.04 ± 0.00 | - -build: 34c9d765 (6122) diff --git a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_radv.log b/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_radv.log deleted file mode 100644 index a65273a..0000000 --- a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_radv.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 0 | pp512 | 85.88 ± 0.10 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 0 | tg128 | 7.48 ± 0.00 | - -build: 34c9d765 (6122) diff --git a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_radv__fa1.log b/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_radv__fa1.log deleted file mode 100644 index a14f281..0000000 --- a/benchmark/results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_radv__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 1 | 0 | pp512 | 86.57 ± 0.11 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 1 | 0 | tg128 | 7.49 ± 0.00 | - -build: 34c9d765 (6122) diff --git a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2-rocwmma.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2-rocwmma.log index 4638587..a6af248 100644 --- a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2-rocwmma.log +++ b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2-rocwmma.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | pp512 | 192.14 ± 0.71 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | tg128 | 10.75 ± 3.44 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | pp512 | 222.91 ± 0.21 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | tg128 | 14.03 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2-rocwmma__fa1.log index 2b45f78..333ac47 100644 --- a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | pp512 | 229.77 ± 0.18 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | tg128 | 13.58 ± 0.00 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | pp512 | 229.15 ± 0.24 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | tg128 | 13.76 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2.log index fe52481..f26f454 100644 --- a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2.log +++ b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 0 | pp512 | 222.24 ± 0.39 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 0 | tg128 | 13.86 ± 0.00 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | pp512 | 222.59 ± 0.24 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | tg128 | 14.03 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2__fa1.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2__fa1.log index 62f3c9e..df5dd02 100644 --- a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2__fa1.log +++ b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 1 | 0 | pp512 | 201.58 ± 0.09 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 1 | 0 | tg128 | 13.57 ± 0.00 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | pp512 | 197.89 ± 3.40 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | tg128 | 13.76 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_3-rocwmma.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_3-rocwmma.log new file mode 100644 index 0000000..ba675a0 --- /dev/null +++ b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_3-rocwmma.log @@ -0,0 +1,15 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +rocBLAS error: No hipBLASLt solution found +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | pp512 | 734.26 ± 0.94 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | tg128 | 14.05 ± 0.00 | + +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_3-rocwmma__fa1.log similarity index 79% rename from benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2-rocwmma__fa1.log rename to benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_3-rocwmma__fa1.log index 348f5ed..a834fd9 100644 --- a/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_3-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 1 | 0 | pp512 | 229.77 ± 0.32 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 1 | 0 | tg128 | 13.59 ± 0.00 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | pp512 | 820.41 ± 1.59 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | tg128 | 13.77 ± 0.00 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2-rocwmma.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_3-rocwmma__hblt0.log similarity index 79% rename from benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2-rocwmma.log rename to benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_3-rocwmma__hblt0.log index 9f1e992..3b668bb 100644 --- a/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2-rocwmma.log +++ b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_3-rocwmma__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 0 | pp512 | 223.38 ± 0.29 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 0 | tg128 | 13.86 ± 0.00 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | pp512 | 213.40 ± 3.62 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | tg128 | 14.04 ± 0.00 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2__fa1.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_3-rocwmma__hblt0__fa1.log similarity index 79% rename from benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2__fa1.log rename to benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_3-rocwmma__hblt0__fa1.log index de6f8de..ed14086 100644 --- a/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2__fa1.log +++ b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_3-rocwmma__hblt0__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 1 | 0 | pp512 | 202.13 ± 0.24 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 1 | 0 | tg128 | 13.58 ± 0.00 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | pp512 | 224.20 ± 4.73 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | tg128 | 13.78 ± 0.00 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_3.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_3.log new file mode 100644 index 0000000..b6d34c8 --- /dev/null +++ b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_3.log @@ -0,0 +1,15 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +rocBLAS error: No hipBLASLt solution found +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | pp512 | 734.70 ± 1.48 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | tg128 | 14.03 ± 0.00 | + +build: 1fe00296 (6182) diff --git a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_beta__fa1.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_3__fa1.log similarity index 69% rename from benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_beta__fa1.log rename to benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_3__fa1.log index 214947a..a9a4641 100644 --- a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_beta__fa1.log +++ b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_3__fa1.log @@ -1,10 +1,10 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 1 | 0 | pp512 | 567.65 ± 0.94 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 1 | 0 | tg128 | 13.60 ± 0.00 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | pp512 | 554.49 ± 0.62 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | tg128 | 13.78 ± 0.00 | -build: 79c1160b (6123) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_3__hblt0.log similarity index 79% rename from benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2.log rename to benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_3__hblt0.log index 5872035..21731f6 100644 --- a/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2.log +++ b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_3__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 0 | pp512 | 222.86 ± 0.11 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 0 | tg128 | 13.85 ± 0.00 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | pp512 | 220.22 ± 1.60 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | tg128 | 14.04 ± 0.00 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_beta__hblt0__fa1.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_3__hblt0__fa1.log similarity index 69% rename from benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_beta__hblt0__fa1.log rename to benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_3__hblt0__fa1.log index 71ec3f9..ecb65ad 100644 --- a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_beta__hblt0__fa1.log +++ b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_3__hblt0__fa1.log @@ -1,10 +1,10 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | pp512 | 203.03 ± 0.17 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | tg128 | 13.58 ± 0.00 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | pp512 | 193.90 ± 1.19 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | tg128 | 13.77 ± 0.00 | -build: 79c1160b (6123) +build: 1fe00296 (6182) diff --git a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_beta.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_beta.log deleted file mode 100644 index c8ae56d..0000000 --- a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 0 | pp512 | 706.58 ± 0.96 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 0 | tg128 | 13.87 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_beta__hblt0.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_beta__hblt0.log deleted file mode 100644 index 305aaa3..0000000 --- a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_beta__hblt0.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | pp512 | 222.31 ± 0.28 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | tg128 | 13.88 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc-rocwmma.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc-rocwmma.log index 86106ee..f5c1c56 100644 --- a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc-rocwmma.log +++ b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc-rocwmma.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 0 | pp512 | 703.10 ± 0.68 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 0 | tg128 | 13.83 ± 0.00 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | pp512 | 751.04 ± 1.24 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | tg128 | 14.01 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc-rocwmma__fa1.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc-rocwmma__fa1.log index 3898840..c86174c 100644 --- a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc-rocwmma__fa1.log +++ b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 1 | 0 | pp512 | 818.63 ± 0.82 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 1 | 0 | tg128 | 13.47 ± 0.00 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | pp512 | 811.04 ± 1.22 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | tg128 | 13.45 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc-rocwmma__hblt0.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc-rocwmma__hblt0.log index 2e84f94..3f70b79 100644 --- a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc-rocwmma__hblt0.log +++ b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc-rocwmma__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | pp512 | 222.39 ± 0.17 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | tg128 | 13.81 ± 0.00 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | pp512 | 752.99 ± 1.44 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | tg128 | 14.00 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc-rocwmma__hblt0__fa1.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc-rocwmma__hblt0__fa1.log index a31080a..4cc26aa 100644 --- a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc-rocwmma__hblt0__fa1.log +++ b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc-rocwmma__hblt0__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | pp512 | 228.56 ± 0.31 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | tg128 | 13.51 ± 0.00 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | pp512 | 794.90 ± 1.42 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | tg128 | 13.45 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc.log index b52b25c..301a2b5 100644 --- a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc.log +++ b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 0 | pp512 | 706.92 ± 0.89 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 0 | tg128 | 13.87 ± 0.00 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | pp512 | 752.36 ± 0.48 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | tg128 | 14.05 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc__fa1.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc__fa1.log index 228b25b..1ddb96e 100644 --- a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc__fa1.log +++ b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 1 | 0 | pp512 | 554.98 ± 0.46 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 1 | 0 | tg128 | 13.61 ± 0.00 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | pp512 | 569.66 ± 0.60 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | tg128 | 13.78 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc__hblt0.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc__hblt0.log index 419821e..e22f092 100644 --- a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc__hblt0.log +++ b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | pp512 | 222.26 ± 0.30 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | tg128 | 13.86 ± 0.00 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | pp512 | 750.36 ± 1.88 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 0 | tg128 | 14.05 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc__hblt0__fa1.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc__hblt0__fa1.log index 8c0ebea..01fed10 100644 --- a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc__hblt0__fa1.log +++ b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc__hblt0__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | pp512 | 201.53 ± 0.07 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | tg128 | 13.59 ± 0.00 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | pp512 | 559.73 ± 0.51 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 99 | 1 | 0 | tg128 | 13.79 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__vulkan_amdvlk.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__vulkan_amdvlk.log index 670f9fa..6d1ed7b 100644 --- a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__vulkan_amdvlk.log +++ b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__vulkan_amdvlk.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | Vulkan | 999 | 0 | pp512 | 675.90 ± 1.28 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | Vulkan | 999 | 0 | tg128 | 14.26 ± 0.00 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | Vulkan | 99 | 0 | pp512 | 680.44 ± 0.55 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | Vulkan | 99 | 0 | tg128 | 14.39 ± 0.03 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__vulkan_amdvlk__fa1.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__vulkan_amdvlk__fa1.log index f680dfa..3c0a8e7 100644 --- a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__vulkan_amdvlk__fa1.log +++ b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__vulkan_amdvlk__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | Vulkan | 999 | 1 | 0 | pp512 | 371.03 ± 0.33 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | Vulkan | 999 | 1 | 0 | tg128 | 12.49 ± 0.00 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | Vulkan | 99 | 1 | 0 | pp512 | 371.66 ± 0.51 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | Vulkan | 99 | 1 | 0 | tg128 | 12.62 ± 0.00 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__vulkan_radv.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__vulkan_radv.log index 36cc6ea..5a3ee90 100644 --- a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__vulkan_radv.log +++ b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__vulkan_radv.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | Vulkan | 999 | 0 | pp512 | 504.61 ± 2.97 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | Vulkan | 999 | 0 | tg128 | 14.05 ± 0.00 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | Vulkan | 99 | 0 | pp512 | 502.88 ± 1.45 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | Vulkan | 99 | 0 | tg128 | 14.21 ± 0.00 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__vulkan_radv__fa1.log b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__vulkan_radv__fa1.log index df5009a..8f4867f 100644 --- a/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__vulkan_radv__fa1.log +++ b/benchmark/results/gemma-3-12b-it-UD-Q8_K_XL__vulkan_radv__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | Vulkan | 999 | 1 | 0 | pp512 | 495.37 ± 0.71 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | Vulkan | 999 | 1 | 0 | tg128 | 13.87 ± 0.00 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | Vulkan | 99 | 1 | 0 | pp512 | 496.33 ± 1.83 | +| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | Vulkan | 99 | 1 | 0 | tg128 | 14.02 ± 0.00 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2-rocwmma.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2-rocwmma.log index aab4706..0219357 100644 --- a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2-rocwmma.log +++ b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2-rocwmma.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | pp512 | 92.82 ± 0.46 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | tg128 | 4.05 ± 0.00 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | pp512 | 87.20 ± 3.70 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | tg128 | 4.09 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log index 43f28f1..8dcf6c9 100644 --- a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | pp512 | 94.62 ± 0.56 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | tg128 | 4.03 ± 0.00 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | pp512 | 68.87 ± 14.37 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | tg128 | 4.08 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2.log index b7e4cd2..627bb9e 100644 --- a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2.log +++ b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 0 | pp512 | 91.25 ± 0.44 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 0 | tg128 | 4.04 ± 0.00 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | pp512 | 82.57 ± 10.36 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | tg128 | 4.09 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2__fa1.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2__fa1.log index da3d8bb..b35b468 100644 --- a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2__fa1.log +++ b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 1 | 0 | pp512 | 84.81 ± 0.48 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 1 | 0 | tg128 | 4.04 ± 0.00 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | pp512 | 74.78 ± 10.12 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | tg128 | 4.09 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_3-rocwmma.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_3-rocwmma.log new file mode 100644 index 0000000..523552b --- /dev/null +++ b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_3-rocwmma.log @@ -0,0 +1,15 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +hipBLASLt error: Heuristic Fetch Failed! +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | pp512 | 395.28 ± 0.22 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | tg128 | 3.96 ± 0.00 | + +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2__fa1.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_3-rocwmma__fa1.log similarity index 79% rename from benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2__fa1.log rename to benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_3-rocwmma__fa1.log index d3b262b..f8b41b5 100644 --- a/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2__fa1.log +++ b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_3-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 1 | 0 | pp512 | 83.75 ± 0.35 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 1 | 0 | tg128 | 4.04 ± 0.00 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | pp512 | 468.37 ± 1.54 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | tg128 | 4.08 ± 0.00 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2-rocwmma.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_3-rocwmma__hblt0.log similarity index 79% rename from benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2-rocwmma.log rename to benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_3-rocwmma__hblt0.log index 96d541d..1ce39b8 100644 --- a/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2-rocwmma.log +++ b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_3-rocwmma__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 0 | pp512 | 92.52 ± 0.44 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 0 | tg128 | 4.05 ± 0.00 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | pp512 | 79.42 ± 0.41 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | tg128 | 3.97 ± 0.00 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_3-rocwmma__hblt0__fa1.log similarity index 79% rename from benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log rename to benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_3-rocwmma__hblt0__fa1.log index a5e826d..29c9209 100644 --- a/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_3-rocwmma__hblt0__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 1 | 0 | pp512 | 94.54 ± 0.52 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 1 | 0 | tg128 | 4.03 ± 0.00 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | pp512 | 89.19 ± 0.53 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | tg128 | 4.06 ± 0.00 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_3.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_3.log new file mode 100644 index 0000000..fcaf5b3 --- /dev/null +++ b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_3.log @@ -0,0 +1,15 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +hipBLASLt error: Heuristic Fetch Failed! +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | pp512 | 398.35 ± 1.07 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | tg128 | 4.09 ± 0.00 | + +build: 1fe00296 (6182) diff --git a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_beta__fa1.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_3__fa1.log similarity index 69% rename from benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_beta__fa1.log rename to benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_3__fa1.log index 6b1dd72..bb05e2f 100644 --- a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_beta__fa1.log +++ b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_3__fa1.log @@ -1,10 +1,10 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 1 | 0 | pp512 | 310.92 ± 0.73 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 1 | 0 | tg128 | 4.05 ± 0.00 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | pp512 | 302.82 ± 2.53 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | tg128 | 4.09 ± 0.00 | -build: 79c1160b (6123) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__rocm7_beta.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_3__hblt0.log similarity index 69% rename from benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__rocm7_beta.log rename to benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_3__hblt0.log index aaaffba..f41ad73 100644 --- a/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__rocm7_beta.log +++ b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_3__hblt0.log @@ -1,10 +1,10 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 0 | pp512 | 91.54 ± 0.50 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 0 | tg128 | 4.04 ± 0.00 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | pp512 | 59.13 ± 7.79 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | tg128 | 4.09 ± 0.00 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_beta__hblt0__fa1.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_3__hblt0__fa1.log similarity index 69% rename from benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_beta__hblt0__fa1.log rename to benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_3__hblt0__fa1.log index 846a5fd..c07bd16 100644 --- a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_beta__hblt0__fa1.log +++ b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_3__hblt0__fa1.log @@ -1,10 +1,10 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | pp512 | 82.85 ± 0.49 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | tg128 | 4.03 ± 0.00 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | pp512 | 61.26 ± 10.54 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | tg128 | 4.09 ± 0.00 | -build: 79c1160b (6123) +build: 1fe00296 (6182) diff --git a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_beta.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_beta.log deleted file mode 100644 index 6b535e0..0000000 --- a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 0 | pp512 | 405.35 ± 0.62 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 0 | tg128 | 4.04 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_beta__hblt0.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_beta__hblt0.log deleted file mode 100644 index 26890da..0000000 --- a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_beta__hblt0.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | pp512 | 86.80 ± 0.36 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | tg128 | 4.02 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc-rocwmma.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc-rocwmma.log index 8b094b5..d48d219 100644 --- a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc-rocwmma.log +++ b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc-rocwmma.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 0 | pp512 | 404.79 ± 0.61 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 0 | tg128 | 4.04 ± 0.00 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | pp512 | 418.46 ± 0.10 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | tg128 | 4.09 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc-rocwmma__fa1.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc-rocwmma__fa1.log index a690d20..ba0b9fa 100644 --- a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc-rocwmma__fa1.log +++ b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 1 | 0 | pp512 | 472.91 ± 1.05 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 1 | 0 | tg128 | 4.03 ± 0.00 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | pp512 | 466.83 ± 1.65 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | tg128 | 4.07 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc-rocwmma__hblt0.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc-rocwmma__hblt0.log index 6f151a6..f152fdb 100644 --- a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc-rocwmma__hblt0.log +++ b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc-rocwmma__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | pp512 | 91.08 ± 0.67 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | tg128 | 4.03 ± 0.01 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | pp512 | 454.10 ± 1.09 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | tg128 | 4.09 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log index 13775d4..a66ffac 100644 --- a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log +++ b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc-rocwmma__hblt0__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | pp512 | 93.26 ± 0.55 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | tg128 | 4.03 ± 0.00 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | pp512 | 499.43 ± 1.24 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | tg128 | 4.06 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc.log index 1615077..adb03dd 100644 --- a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc.log +++ b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 0 | pp512 | 368.33 ± 0.38 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 0 | tg128 | 3.71 ± 0.01 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | pp512 | 392.50 ± 0.50 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | tg128 | 3.97 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc__fa1.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc__fa1.log index 3803787..5ec86d3 100644 --- a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc__fa1.log +++ b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 1 | 0 | pp512 | 311.83 ± 0.31 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 1 | 0 | tg128 | 4.04 ± 0.00 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | pp512 | 311.25 ± 0.72 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | tg128 | 4.09 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc__hblt0.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc__hblt0.log index eba50a2..bec2363 100644 --- a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc__hblt0.log +++ b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | pp512 | 80.07 ± 0.21 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | tg128 | 4.00 ± 0.00 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | pp512 | 451.69 ± 0.62 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 0 | tg128 | 4.09 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc__hblt0__fa1.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc__hblt0__fa1.log index 4575c7f..3a00d5c 100644 --- a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc__hblt0__fa1.log +++ b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc__hblt0__fa1.log @@ -2,4 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc] gemma-3-27b-it-BF16-00001-of-00002 __hblt0__fa1 failed (exit 134) +| model | size | params | backend | ngl | fa | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | pp512 | 324.43 ± 0.22 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 99 | 1 | 0 | tg128 | 4.09 ± 0.00 | + +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__vulkan_amdvlk.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__vulkan_amdvlk.log index d74242d..f20e9bc 100644 --- a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__vulkan_amdvlk.log +++ b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__vulkan_amdvlk.log @@ -4,5 +4,5 @@ ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | ggml_vulkan: Device memory allocation of size 2819260416 failed. ggml_vulkan: Requested buffer size exceeds device memory allocation limit: ErrorOutOfDeviceMemory -main: error: failed to load model '/mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf' +main: error: failed to load model '/home/kyuz0/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf' ✖ ! [vulkan_amdvlk] gemma-3-27b-it-BF16-00001-of-00002 failed (exit 1) diff --git a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__vulkan_amdvlk__fa1.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__vulkan_amdvlk__fa1.log index a667917..2578dff 100644 --- a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__vulkan_amdvlk__fa1.log +++ b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__vulkan_amdvlk__fa1.log @@ -4,5 +4,5 @@ ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | ggml_vulkan: Device memory allocation of size 2819260416 failed. ggml_vulkan: Requested buffer size exceeds device memory allocation limit: ErrorOutOfDeviceMemory -main: error: failed to load model '/mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf' +main: error: failed to load model '/home/kyuz0/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf' ✖ ! [vulkan_amdvlk] gemma-3-27b-it-BF16-00001-of-00002 __fa1 failed (exit 1) diff --git a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__vulkan_radv.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__vulkan_radv.log index 73c8358..9dfdc23 100644 --- a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__vulkan_radv.log +++ b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__vulkan_radv.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | Vulkan | 999 | 0 | pp512 | 135.01 ± 0.28 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | Vulkan | 999 | 0 | tg128 | 4.03 ± 0.00 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | Vulkan | 99 | 0 | pp512 | 129.49 ± 0.34 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | Vulkan | 99 | 0 | tg128 | 4.06 ± 0.00 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__vulkan_radv__fa1.log b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__vulkan_radv__fa1.log index 8c6f730..96c0e93 100644 --- a/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__vulkan_radv__fa1.log +++ b/benchmark/results/gemma-3-27b-it-BF16-00001-of-00002__vulkan_radv__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | Vulkan | 999 | 1 | 0 | pp512 | 137.76 ± 0.25 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | Vulkan | 999 | 1 | 0 | tg128 | 4.03 ± 0.00 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | Vulkan | 99 | 1 | 0 | pp512 | 137.67 ± 1.25 | +| gemma3 27B BF16 | 50.31 GiB | 27.01 B | Vulkan | 99 | 1 | 0 | tg128 | 4.06 ± 0.00 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_2-rocwmma.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_2-rocwmma.log index 43d8ffa..84fba20 100644 --- a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_2-rocwmma.log +++ b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_2-rocwmma.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | pp512 | 727.59 ± 1.45 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | tg128 | 76.22 ± 0.03 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | pp512 | 728.70 ± 1.28 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | tg128 | 76.63 ± 0.03 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_2-rocwmma__fa1.log index 5f4bc59..073d72d 100644 --- a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_2-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | pp512 | 750.30 ± 1.03 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | tg128 | 69.96 ± 0.02 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | pp512 | 752.52 ± 0.83 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | tg128 | 70.93 ± 0.02 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_2.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_2.log index 8397b72..4ab8e49 100644 --- a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_2.log +++ b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_2.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 0 | pp512 | 728.24 ± 0.55 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 0 | tg128 | 75.89 ± 0.03 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | pp512 | 729.33 ± 1.93 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | tg128 | 76.79 ± 0.03 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_2__fa1.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_2__fa1.log index ac89e51..bdf1afb 100644 --- a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_2__fa1.log +++ b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_2__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 1 | 0 | pp512 | 643.29 ± 0.97 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 1 | 0 | tg128 | 69.53 ± 0.01 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | pp512 | 645.25 ± 0.89 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | tg128 | 70.31 ± 0.01 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_3-rocwmma.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_3-rocwmma.log new file mode 100644 index 0000000..4a3544a --- /dev/null +++ b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_3-rocwmma.log @@ -0,0 +1,15 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +rocBLAS error: No hipBLASLt solution found +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | pp512 | 2033.46 ± 5.16 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | tg128 | 76.47 ± 0.26 | + +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_3-rocwmma__fa1.log similarity index 78% rename from benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__rocm6_4_2-rocwmma__fa1.log rename to benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_3-rocwmma__fa1.log index 16dd036..78f4e0c 100644 --- a/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_3-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 1 | 0 | pp512 | 752.25 ± 0.73 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 1 | 0 | tg128 | 69.93 ± 0.01 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | pp512 | 2276.86 ± 9.60 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | tg128 | 70.76 ± 0.26 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__rocm6_4_2-rocwmma.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_3-rocwmma__hblt0.log similarity index 79% rename from benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__rocm6_4_2-rocwmma.log rename to benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_3-rocwmma__hblt0.log index 34e7e86..54a23d2 100644 --- a/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__rocm6_4_2-rocwmma.log +++ b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_3-rocwmma__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 0 | pp512 | 729.91 ± 1.22 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 0 | tg128 | 76.14 ± 0.03 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | pp512 | 727.18 ± 2.22 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | tg128 | 75.65 ± 0.74 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__rocm6_4_2__fa1.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_3-rocwmma__hblt0__fa1.log similarity index 78% rename from benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__rocm6_4_2__fa1.log rename to benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_3-rocwmma__hblt0__fa1.log index 08d39fe..24c6a23 100644 --- a/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__rocm6_4_2__fa1.log +++ b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_3-rocwmma__hblt0__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 1 | 0 | pp512 | 645.88 ± 0.61 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 1 | 0 | tg128 | 69.63 ± 0.01 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | pp512 | 740.27 ± 10.38 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | tg128 | 70.76 ± 0.11 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_3.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_3.log new file mode 100644 index 0000000..e669ba8 --- /dev/null +++ b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_3.log @@ -0,0 +1,15 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +rocBLAS error: No hipBLASLt solution found +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | pp512 | 2035.38 ± 4.03 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | tg128 | 75.40 ± 0.80 | + +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__rocm7_rc__fa1.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_3__fa1.log similarity index 69% rename from benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__rocm7_rc__fa1.log rename to benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_3__fa1.log index 4c71363..1562460 100644 --- a/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__rocm7_rc__fa1.log +++ b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_3__fa1.log @@ -1,10 +1,10 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 1 | 0 | pp512 | 646.16 ± 0.39 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 1 | 0 | tg128 | 69.53 ± 0.02 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | pp512 | 1515.55 ± 8.10 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | tg128 | 70.20 ± 0.39 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__rocm6_4_2.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_3__hblt0.log similarity index 79% rename from benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__rocm6_4_2.log rename to benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_3__hblt0.log index f07fba3..b49eeff 100644 --- a/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__rocm6_4_2.log +++ b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_3__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 0 | pp512 | 730.51 ± 1.49 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 0 | tg128 | 76.35 ± 0.02 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | pp512 | 714.75 ± 27.98 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | tg128 | 66.10 ± 5.25 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_beta__fa1.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_3__hblt0__fa1.log similarity index 68% rename from benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_beta__fa1.log rename to benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_3__hblt0__fa1.log index f67fdc0..f75c3de 100644 --- a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_beta__fa1.log +++ b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm6_4_3__hblt0__fa1.log @@ -1,10 +1,10 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 1 | 0 | pp512 | 1548.20 ± 4.48 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 1 | 0 | tg128 | 69.64 ± 0.01 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | pp512 | 596.86 ± 37.66 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | tg128 | 58.75 ± 3.09 | -build: 79c1160b (6123) +build: 1fe00296 (6182) diff --git a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_beta.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_beta.log deleted file mode 100644 index e24e049..0000000 --- a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 0 | pp512 | 1812.73 ± 7.38 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 0 | tg128 | 76.55 ± 0.02 | - -build: 79c1160b (6123) diff --git a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_beta__hblt0.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_beta__hblt0.log deleted file mode 100644 index 2791d8d..0000000 --- a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_beta__hblt0.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | pp512 | 729.03 ± 0.75 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | tg128 | 76.59 ± 0.03 | - -build: 79c1160b (6123) diff --git a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_beta__hblt0__fa1.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_beta__hblt0__fa1.log deleted file mode 100644 index e88558b..0000000 --- a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_beta__hblt0__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | pp512 | 651.26 ± 1.22 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | tg128 | 69.44 ± 0.01 | - -build: 79c1160b (6123) diff --git a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc-rocwmma.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc-rocwmma.log index 502bbc1..c7f6ba3 100644 --- a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc-rocwmma.log +++ b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc-rocwmma.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 0 | pp512 | 1799.45 ± 7.32 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 0 | tg128 | 75.43 ± 0.03 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | pp512 | 2014.60 ± 24.35 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | tg128 | 59.16 ± 3.76 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc-rocwmma__fa1.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc-rocwmma__fa1.log index c34e4a4..0862fc6 100644 --- a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc-rocwmma__fa1.log +++ b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 1 | 0 | pp512 | 2267.56 ± 6.61 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 1 | 0 | tg128 | 68.27 ± 0.01 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | pp512 | 2191.77 ± 78.21 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | tg128 | 54.32 ± 2.65 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc-rocwmma__hblt0.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc-rocwmma__hblt0.log index d86865b..4293b33 100644 --- a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc-rocwmma__hblt0.log +++ b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc-rocwmma__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | pp512 | 729.58 ± 0.87 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | tg128 | 75.48 ± 0.02 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | pp512 | 1991.71 ± 2.91 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | tg128 | 56.37 ± 3.40 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc-rocwmma__hblt0__fa1.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc-rocwmma__hblt0__fa1.log index 9ddd54a..1af2b23 100644 --- a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc-rocwmma__hblt0__fa1.log +++ b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc-rocwmma__hblt0__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | pp512 | 750.44 ± 0.80 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | tg128 | 68.27 ± 0.01 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | pp512 | 2096.22 ± 4.59 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | tg128 | 64.88 ± 0.05 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc.log index b4f3d2a..7e86d5a 100644 --- a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc.log +++ b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 0 | pp512 | 1812.27 ± 4.63 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 0 | tg128 | 76.22 ± 0.01 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | pp512 | 2027.41 ± 4.62 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | tg128 | 77.12 ± 0.03 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc__fa1.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc__fa1.log index a78e906..a3497bf 100644 --- a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc__fa1.log +++ b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 1 | 0 | pp512 | 1510.06 ± 4.96 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 1 | 0 | tg128 | 69.58 ± 0.02 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | pp512 | 1550.55 ± 4.52 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | tg128 | 70.54 ± 0.06 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc__hblt0.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc__hblt0.log index e4cd337..51f19f4 100644 --- a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc__hblt0.log +++ b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | pp512 | 729.81 ± 1.15 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | tg128 | 76.03 ± 0.04 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | pp512 | 1992.48 ± 7.34 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 0 | tg128 | 77.05 ± 0.03 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc__hblt0__fa1.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc__hblt0__fa1.log index a600775..a65b575 100644 --- a/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc__hblt0__fa1.log +++ b/benchmark/results/gemma-3-4b-it-Q3_K_S__rocm7_rc__hblt0__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | pp512 | 645.48 ± 1.40 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | tg128 | 69.67 ± 0.02 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | pp512 | 1474.15 ± 1.44 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 99 | 1 | 0 | tg128 | 70.44 ± 0.01 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gemma-3-4b-it-Q3_K_S__vulkan_amdvlk.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__vulkan_amdvlk.log index 8dafbf8..2624621 100644 --- a/benchmark/results/gemma-3-4b-it-Q3_K_S__vulkan_amdvlk.log +++ b/benchmark/results/gemma-3-4b-it-Q3_K_S__vulkan_amdvlk.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | Vulkan | 999 | 0 | pp512 | 1628.18 ± 1.73 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | Vulkan | 999 | 0 | tg128 | 84.23 ± 0.15 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | Vulkan | 99 | 0 | pp512 | 1593.62 ± 2.90 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | Vulkan | 99 | 0 | tg128 | 85.26 ± 0.26 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/gemma-3-4b-it-Q3_K_S__vulkan_amdvlk__fa1.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__vulkan_amdvlk__fa1.log index fc50285..20e82ce 100644 --- a/benchmark/results/gemma-3-4b-it-Q3_K_S__vulkan_amdvlk__fa1.log +++ b/benchmark/results/gemma-3-4b-it-Q3_K_S__vulkan_amdvlk__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | Vulkan | 999 | 1 | 0 | pp512 | 947.36 ± 1.47 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | Vulkan | 999 | 1 | 0 | tg128 | 60.35 ± 0.15 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | Vulkan | 99 | 1 | 0 | pp512 | 936.52 ± 2.35 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | Vulkan | 99 | 1 | 0 | tg128 | 60.89 ± 0.10 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/gemma-3-4b-it-Q3_K_S__vulkan_radv.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__vulkan_radv.log index 2ecedbd..6ba35a8 100644 --- a/benchmark/results/gemma-3-4b-it-Q3_K_S__vulkan_radv.log +++ b/benchmark/results/gemma-3-4b-it-Q3_K_S__vulkan_radv.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | Vulkan | 999 | 0 | pp512 | 1529.98 ± 0.80 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | Vulkan | 999 | 0 | tg128 | 86.95 ± 0.31 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | Vulkan | 99 | 0 | pp512 | 1515.05 ± 2.98 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | Vulkan | 99 | 0 | tg128 | 87.54 ± 0.18 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/gemma-3-4b-it-Q3_K_S__vulkan_radv__fa1.log b/benchmark/results/gemma-3-4b-it-Q3_K_S__vulkan_radv__fa1.log index 309d21b..a1a86a0 100644 --- a/benchmark/results/gemma-3-4b-it-Q3_K_S__vulkan_radv__fa1.log +++ b/benchmark/results/gemma-3-4b-it-Q3_K_S__vulkan_radv__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | Vulkan | 999 | 1 | 0 | pp512 | 1498.81 ± 1.70 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | Vulkan | 999 | 1 | 0 | tg128 | 81.29 ± 0.12 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | Vulkan | 99 | 1 | 0 | pp512 | 1476.16 ± 5.12 | +| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | Vulkan | 99 | 1 | 0 | tg128 | 82.48 ± 0.36 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-120b-F16__rocm6_4_2-rocwmma.log b/benchmark/results/gpt-oss-120b-F16__rocm6_4_2-rocwmma.log index aa28166..d8257a4 100644 --- a/benchmark/results/gpt-oss-120b-F16__rocm6_4_2-rocwmma.log +++ b/benchmark/results/gpt-oss-120b-F16__rocm6_4_2-rocwmma.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 353.66 ± 0.64 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 33.65 ± 0.00 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 355.59 ± 0.86 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 33.97 ± 0.01 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-120b-F16__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/gpt-oss-120b-F16__rocm6_4_2-rocwmma__fa1.log index 11185b5..c765b22 100644 --- a/benchmark/results/gpt-oss-120b-F16__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/gpt-oss-120b-F16__rocm6_4_2-rocwmma__fa1.log @@ -2,5 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x2ad71050) reason :GPU Hang -✖ ! [rocm6_4_2-rocwmma] gpt-oss-120b-F16 __fa1 failed (exit 134) +| model | size | params | backend | ngl | fa | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | pp512 | 390.43 ± 0.70 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | tg128 | 33.81 ± 0.01 | + +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-120b-F16__rocm6_4_2.log b/benchmark/results/gpt-oss-120b-F16__rocm6_4_2.log index e00d36f..306797a 100644 --- a/benchmark/results/gpt-oss-120b-F16__rocm6_4_2.log +++ b/benchmark/results/gpt-oss-120b-F16__rocm6_4_2.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 0 | pp512 | 352.40 ± 1.12 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 0 | tg128 | 31.99 ± 0.00 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 355.94 ± 1.35 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 33.97 ± 0.01 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-120b-F16__rocm6_4_2__fa1.log b/benchmark/results/gpt-oss-120b-F16__rocm6_4_2__fa1.log index 5f9c27c..a49785b 100644 --- a/benchmark/results/gpt-oss-120b-F16__rocm6_4_2__fa1.log +++ b/benchmark/results/gpt-oss-120b-F16__rocm6_4_2__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 1 | 0 | pp512 | 321.54 ± 0.46 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 1 | 0 | tg128 | 33.03 ± 0.00 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | pp512 | 322.57 ± 0.31 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | tg128 | 33.30 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-120b-F16__rocm6_4_3-rocwmma.log b/benchmark/results/gpt-oss-120b-F16__rocm6_4_3-rocwmma.log new file mode 100644 index 0000000..2b2057d --- /dev/null +++ b/benchmark/results/gpt-oss-120b-F16__rocm6_4_3-rocwmma.log @@ -0,0 +1,15 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +rocBLAS error: No hipBLASLt solution found +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 622.16 ± 6.71 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 33.91 ± 0.01 | + +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gpt-oss-120b-F16__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/gpt-oss-120b-F16__rocm6_4_3-rocwmma__fa1.log similarity index 79% rename from benchmark/results_08-08-2025/gpt-oss-120b-F16__rocm6_4_2-rocwmma__fa1.log rename to benchmark/results/gpt-oss-120b-F16__rocm6_4_3-rocwmma__fa1.log index dbb739d..a8e0637 100644 --- a/benchmark/results_08-08-2025/gpt-oss-120b-F16__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/gpt-oss-120b-F16__rocm6_4_3-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 1 | 0 | pp512 | 411.33 ± 1.01 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 1 | 0 | tg128 | 33.50 ± 0.00 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | pp512 | 743.09 ± 4.89 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | tg128 | 33.76 ± 0.00 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gpt-oss-120b-F16__rocm6_4_2-rocwmma.log b/benchmark/results/gpt-oss-120b-F16__rocm6_4_3-rocwmma__hblt0.log similarity index 79% rename from benchmark/results_08-08-2025/gpt-oss-120b-F16__rocm6_4_2-rocwmma.log rename to benchmark/results/gpt-oss-120b-F16__rocm6_4_3-rocwmma__hblt0.log index 2301d16..d94bfc6 100644 --- a/benchmark/results_08-08-2025/gpt-oss-120b-F16__rocm6_4_2-rocwmma.log +++ b/benchmark/results/gpt-oss-120b-F16__rocm6_4_3-rocwmma__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 0 | pp512 | 355.01 ± 0.57 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 0 | tg128 | 33.66 ± 0.00 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 354.98 ± 0.72 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 33.86 ± 0.00 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gpt-oss-120b-F16__rocm7_beta__fa1.log b/benchmark/results/gpt-oss-120b-F16__rocm6_4_3-rocwmma__hblt0__fa1.log similarity index 69% rename from benchmark/results_08-08-2025/gpt-oss-120b-F16__rocm7_beta__fa1.log rename to benchmark/results/gpt-oss-120b-F16__rocm6_4_3-rocwmma__hblt0__fa1.log index 707b558..95ce008 100644 --- a/benchmark/results_08-08-2025/gpt-oss-120b-F16__rocm7_beta__fa1.log +++ b/benchmark/results/gpt-oss-120b-F16__rocm6_4_3-rocwmma__hblt0__fa1.log @@ -1,10 +1,10 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 1 | 0 | pp512 | 249.65 ± 0.33 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 1 | 0 | tg128 | 33.04 ± 0.01 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | pp512 | 390.67 ± 0.97 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | tg128 | 33.79 ± 0.01 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-120b-F16__rocm6_4_3.log b/benchmark/results/gpt-oss-120b-F16__rocm6_4_3.log new file mode 100644 index 0000000..320bdde --- /dev/null +++ b/benchmark/results/gpt-oss-120b-F16__rocm6_4_3.log @@ -0,0 +1,15 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +hipBLASLt error: Heuristic Fetch Failed! +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 617.00 ± 4.97 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 33.90 ± 0.00 | + +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gpt-oss-120b-F16__rocm6_4_2__fa1.log b/benchmark/results/gpt-oss-120b-F16__rocm6_4_3__fa1.log similarity index 79% rename from benchmark/results_08-08-2025/gpt-oss-120b-F16__rocm6_4_2__fa1.log rename to benchmark/results/gpt-oss-120b-F16__rocm6_4_3__fa1.log index a62923c..aaf9547 100644 --- a/benchmark/results_08-08-2025/gpt-oss-120b-F16__rocm6_4_2__fa1.log +++ b/benchmark/results/gpt-oss-120b-F16__rocm6_4_3__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 1 | 0 | pp512 | 247.95 ± 0.40 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 1 | 0 | tg128 | 33.04 ± 0.00 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | pp512 | 543.39 ± 5.51 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | tg128 | 33.28 ± 0.00 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gpt-oss-120b-F16__rocm6_4_2.log b/benchmark/results/gpt-oss-120b-F16__rocm6_4_3__hblt0.log similarity index 79% rename from benchmark/results_08-08-2025/gpt-oss-120b-F16__rocm6_4_2.log rename to benchmark/results/gpt-oss-120b-F16__rocm6_4_3__hblt0.log index fc1ded3..717bdb1 100644 --- a/benchmark/results_08-08-2025/gpt-oss-120b-F16__rocm6_4_2.log +++ b/benchmark/results/gpt-oss-120b-F16__rocm6_4_3__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 0 | pp512 | 353.36 ± 0.53 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 0 | tg128 | 31.90 ± 0.01 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 354.18 ± 0.29 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 33.88 ± 0.00 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-120b-F16__rocm7_beta__fa1.log b/benchmark/results/gpt-oss-120b-F16__rocm6_4_3__hblt0__fa1.log similarity index 69% rename from benchmark/results/gpt-oss-120b-F16__rocm7_beta__fa1.log rename to benchmark/results/gpt-oss-120b-F16__rocm6_4_3__hblt0__fa1.log index 5a91196..a328319 100644 --- a/benchmark/results/gpt-oss-120b-F16__rocm7_beta__fa1.log +++ b/benchmark/results/gpt-oss-120b-F16__rocm6_4_3__hblt0__fa1.log @@ -1,10 +1,10 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 1 | 0 | pp512 | 548.27 ± 2.65 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 1 | 0 | tg128 | 33.07 ± 0.00 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | pp512 | 322.46 ± 0.46 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | tg128 | 33.33 ± 0.00 | -build: 79c1160b (6123) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-120b-F16__rocm7_beta.log b/benchmark/results/gpt-oss-120b-F16__rocm7_beta.log deleted file mode 100644 index 218c087..0000000 --- a/benchmark/results/gpt-oss-120b-F16__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 0 | pp512 | 604.24 ± 4.34 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 0 | tg128 | 33.69 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/gpt-oss-120b-F16__rocm7_beta__hblt0.log b/benchmark/results/gpt-oss-120b-F16__rocm7_beta__hblt0.log deleted file mode 100644 index 957c7c2..0000000 --- a/benchmark/results/gpt-oss-120b-F16__rocm7_beta__hblt0.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 355.23 ± 1.71 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 33.66 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/gpt-oss-120b-F16__rocm7_beta__hblt0__fa1.log b/benchmark/results/gpt-oss-120b-F16__rocm7_beta__hblt0__fa1.log deleted file mode 100644 index b665daa..0000000 --- a/benchmark/results/gpt-oss-120b-F16__rocm7_beta__hblt0__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | pp512 | 323.79 ± 0.87 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | tg128 | 33.04 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/gpt-oss-120b-F16__rocm7_rc-rocwmma.log b/benchmark/results/gpt-oss-120b-F16__rocm7_rc-rocwmma.log index f5db248..bfcbd06 100644 --- a/benchmark/results/gpt-oss-120b-F16__rocm7_rc-rocwmma.log +++ b/benchmark/results/gpt-oss-120b-F16__rocm7_rc-rocwmma.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 0 | pp512 | 592.27 ± 5.61 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 0 | tg128 | 33.68 ± 0.02 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 643.61 ± 7.14 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 33.91 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-120b-F16__rocm7_rc-rocwmma__fa1.log b/benchmark/results/gpt-oss-120b-F16__rocm7_rc-rocwmma__fa1.log index 60e0975..9e50477 100644 --- a/benchmark/results/gpt-oss-120b-F16__rocm7_rc-rocwmma__fa1.log +++ b/benchmark/results/gpt-oss-120b-F16__rocm7_rc-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 1 | 0 | pp512 | 735.02 ± 5.32 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 1 | 0 | tg128 | 33.34 ± 0.01 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | pp512 | 736.33 ± 3.33 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | tg128 | 33.74 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-120b-F16__rocm7_rc-rocwmma__hblt0.log b/benchmark/results/gpt-oss-120b-F16__rocm7_rc-rocwmma__hblt0.log index 7e97f90..a3eb53e 100644 --- a/benchmark/results/gpt-oss-120b-F16__rocm7_rc-rocwmma__hblt0.log +++ b/benchmark/results/gpt-oss-120b-F16__rocm7_rc-rocwmma__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 353.49 ± 1.71 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 33.63 ± 0.00 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 651.63 ± 3.08 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 33.88 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-120b-F16__rocm7_rc-rocwmma__hblt0__fa1.log b/benchmark/results/gpt-oss-120b-F16__rocm7_rc-rocwmma__hblt0__fa1.log index a07f108..422fcf2 100644 --- a/benchmark/results/gpt-oss-120b-F16__rocm7_rc-rocwmma__hblt0__fa1.log +++ b/benchmark/results/gpt-oss-120b-F16__rocm7_rc-rocwmma__hblt0__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | pp512 | 388.50 ± 1.06 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | tg128 | 33.28 ± 0.00 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | pp512 | 738.84 ± 9.12 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | tg128 | 33.79 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-120b-F16__rocm7_rc.log b/benchmark/results/gpt-oss-120b-F16__rocm7_rc.log index c05f1a3..8fc2f66 100644 --- a/benchmark/results/gpt-oss-120b-F16__rocm7_rc.log +++ b/benchmark/results/gpt-oss-120b-F16__rocm7_rc.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 0 | pp512 | 598.68 ± 9.32 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 0 | tg128 | 33.75 ± 0.00 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 649.28 ± 0.87 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 33.99 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-120b-F16__rocm7_rc__fa1.log b/benchmark/results/gpt-oss-120b-F16__rocm7_rc__fa1.log index f9a46a3..35770c3 100644 --- a/benchmark/results/gpt-oss-120b-F16__rocm7_rc__fa1.log +++ b/benchmark/results/gpt-oss-120b-F16__rocm7_rc__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 1 | 0 | pp512 | 546.30 ± 3.37 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 1 | 0 | tg128 | 33.04 ± 0.00 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | pp512 | 550.01 ± 3.85 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | tg128 | 33.38 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-120b-F16__rocm7_rc__hblt0.log b/benchmark/results/gpt-oss-120b-F16__rocm7_rc__hblt0.log index 19aa96b..7ed9087 100644 --- a/benchmark/results/gpt-oss-120b-F16__rocm7_rc__hblt0.log +++ b/benchmark/results/gpt-oss-120b-F16__rocm7_rc__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 354.34 ± 0.67 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 33.76 ± 0.00 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 659.79 ± 3.13 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 34.01 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-120b-F16__rocm7_rc__hblt0__fa1.log b/benchmark/results/gpt-oss-120b-F16__rocm7_rc__hblt0__fa1.log index 2733c29..af54d9f 100644 --- a/benchmark/results/gpt-oss-120b-F16__rocm7_rc__hblt0__fa1.log +++ b/benchmark/results/gpt-oss-120b-F16__rocm7_rc__hblt0__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | pp512 | 324.26 ± 0.80 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | tg128 | 33.05 ± 0.00 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | pp512 | 553.65 ± 2.40 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 99 | 1 | 0 | tg128 | 33.31 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-120b-F16__vulkan_amdvlk.log b/benchmark/results/gpt-oss-120b-F16__vulkan_amdvlk.log index 461cbc2..9f2e80d 100644 --- a/benchmark/results/gpt-oss-120b-F16__vulkan_amdvlk.log +++ b/benchmark/results/gpt-oss-120b-F16__vulkan_amdvlk.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | Vulkan | 999 | 0 | pp512 | 450.26 ± 1.46 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | Vulkan | 999 | 0 | tg128 | 33.56 ± 0.03 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | Vulkan | 99 | 0 | pp512 | 449.86 ± 1.68 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | Vulkan | 99 | 0 | tg128 | 34.19 ± 0.02 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-120b-F16__vulkan_amdvlk__fa1.log b/benchmark/results/gpt-oss-120b-F16__vulkan_amdvlk__fa1.log index 2219116..9fa4616 100644 --- a/benchmark/results/gpt-oss-120b-F16__vulkan_amdvlk__fa1.log +++ b/benchmark/results/gpt-oss-120b-F16__vulkan_amdvlk__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | Vulkan | 999 | 1 | 0 | pp512 | 499.80 ± 1.95 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | Vulkan | 999 | 1 | 0 | tg128 | 33.18 ± 0.01 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | Vulkan | 99 | 1 | 0 | pp512 | 496.21 ± 1.71 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | Vulkan | 99 | 1 | 0 | tg128 | 33.64 ± 0.01 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-120b-F16__vulkan_radv.log b/benchmark/results/gpt-oss-120b-F16__vulkan_radv.log index c45dee8..71ff53d 100644 --- a/benchmark/results/gpt-oss-120b-F16__vulkan_radv.log +++ b/benchmark/results/gpt-oss-120b-F16__vulkan_radv.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | Vulkan | 999 | 0 | pp512 | 230.22 ± 0.76 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | Vulkan | 999 | 0 | tg128 | 33.16 ± 0.01 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | Vulkan | 99 | 0 | pp512 | 230.09 ± 0.83 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | Vulkan | 99 | 0 | tg128 | 33.57 ± 0.02 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-120b-F16__vulkan_radv__fa1.log b/benchmark/results/gpt-oss-120b-F16__vulkan_radv__fa1.log index 718febd..547d915 100644 --- a/benchmark/results/gpt-oss-120b-F16__vulkan_radv__fa1.log +++ b/benchmark/results/gpt-oss-120b-F16__vulkan_radv__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | Vulkan | 999 | 1 | 0 | pp512 | 243.20 ± 1.11 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | Vulkan | 999 | 1 | 0 | tg128 | 33.15 ± 0.02 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | Vulkan | 99 | 1 | 0 | pp512 | 243.96 ± 0.96 | +| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | Vulkan | 99 | 1 | 0 | tg128 | 33.79 ± 0.01 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2-rocwmma.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2-rocwmma.log index c520d12..360ed4e 100644 --- a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2-rocwmma.log +++ b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2-rocwmma.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 352.37 ± 0.72 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 45.11 ± 0.02 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 353.20 ± 0.30 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 45.42 ± 0.01 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2-rocwmma__fa1.log index 9dc5fe3..6969d6b 100644 --- a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2-rocwmma__fa1.log @@ -2,5 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x3c5a6050) reason :GPU Hang -✖ ! [rocm6_4_2-rocwmma] gpt-oss-120b-mxfp4-00001-of-00003 __fa1 failed (exit 134) +| model | size | params | backend | ngl | fa | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 1 | 0 | pp512 | 387.10 ± 0.42 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 1 | 0 | tg128 | 45.16 ± 0.01 | + +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2.log index 03da684..32eae28 100644 --- a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2.log +++ b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2.log @@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x8bc5050) reason :GPU Hang +HW Exception by GPU node-1 (Agent handle: 0x2bea6540) reason :GPU Hang ✖ ! [rocm6_4_2] gpt-oss-120b-mxfp4-00001-of-00003 failed (exit 134) diff --git a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2__fa1.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2__fa1.log index 97f0889..1d395f6 100644 --- a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2__fa1.log +++ b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 1 | 0 | pp512 | 319.23 ± 0.62 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 1 | 0 | tg128 | 43.79 ± 0.02 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 1 | 0 | pp512 | 319.84 ± 0.73 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 1 | 0 | tg128 | 44.43 ± 0.02 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_3-rocwmma.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_3-rocwmma.log new file mode 100644 index 0000000..f3ebd8d --- /dev/null +++ b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_3-rocwmma.log @@ -0,0 +1,15 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +rocBLAS error: No hipBLASLt solution found +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 606.86 ± 5.18 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 45.26 ± 0.02 | + +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2__fa1.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_3-rocwmma__fa1.log similarity index 79% rename from benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2__fa1.log rename to benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_3-rocwmma__fa1.log index 769bedc..78fca14 100644 --- a/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2__fa1.log +++ b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_3-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 1 | 0 | pp512 | 246.76 ± 0.35 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 1 | 0 | tg128 | 43.67 ± 0.01 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 1 | 0 | pp512 | 732.72 ± 4.06 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 1 | 0 | tg128 | 45.14 ± 0.01 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2-rocwmma.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_3-rocwmma__hblt0.log similarity index 79% rename from benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2-rocwmma.log rename to benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_3-rocwmma__hblt0.log index 3f432b9..0436056 100644 --- a/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2-rocwmma.log +++ b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_3-rocwmma__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 0 | pp512 | 353.53 ± 0.62 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 0 | tg128 | 45.05 ± 0.08 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 351.42 ± 1.56 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 45.39 ± 0.01 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm6_4_2-rocwmma.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_3-rocwmma__hblt0__fa1.log similarity index 57% rename from benchmark/results/llama3.3-70.6B-Q4_K_M__rocm6_4_2-rocwmma.log rename to benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_3-rocwmma__hblt0__fa1.log index 1848af6..4fe67c3 100644 --- a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm6_4_2-rocwmma.log +++ b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_3-rocwmma__hblt0__fa1.log @@ -2,5 +2,5 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x1fec7050) reason :GPU Hang -✖ ! [rocm6_4_2-rocwmma] llama3.3-70.6B-Q4_K_M failed (exit 134) +HW Exception by GPU node-1 (Agent handle: 0x3273c340) reason :GPU Hang +✖ ! [rocm6_4_3-rocwmma] gpt-oss-120b-mxfp4-00001-of-00003 __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_3.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_3.log new file mode 100644 index 0000000..f5b3307 --- /dev/null +++ b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_3.log @@ -0,0 +1,15 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +rocBLAS error: No hipBLASLt solution found +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 608.20 ± 7.04 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 45.40 ± 0.01 | + +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_3__fa1.log similarity index 78% rename from benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2-rocwmma__fa1.log rename to benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_3__fa1.log index fa4767b..7a9d128 100644 --- a/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_3__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 1 | 0 | pp512 | 408.50 ± 1.91 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 1 | 0 | tg128 | 44.69 ± 0.18 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 1 | 0 | pp512 | 533.95 ± 3.58 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 1 | 0 | tg128 | 44.41 ± 0.03 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_3__hblt0.log similarity index 79% rename from benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2.log rename to benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_3__hblt0.log index c1f2f78..3f1f7ba 100644 --- a/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2.log +++ b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_3__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 0 | pp512 | 353.45 ± 1.22 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 0 | tg128 | 44.12 ± 0.01 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 352.53 ± 0.81 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 45.41 ± 0.01 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_beta__fa1.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_3__hblt0__fa1.log similarity index 68% rename from benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_beta__fa1.log rename to benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_3__hblt0__fa1.log index bbf6d17..fa1e33d 100644 --- a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_beta__fa1.log +++ b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_3__hblt0__fa1.log @@ -1,10 +1,10 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 1 | 0 | pp512 | 539.93 ± 1.23 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 1 | 0 | tg128 | 44.01 ± 0.00 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 1 | 0 | pp512 | 320.78 ± 0.96 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 1 | 0 | tg128 | 44.49 ± 0.03 | -build: 79c1160b (6123) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_beta.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_beta.log deleted file mode 100644 index 88c729a..0000000 --- a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 0 | pp512 | 589.45 ± 4.75 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 0 | tg128 | 45.00 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_beta__hblt0.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_beta__hblt0.log deleted file mode 100644 index 1b8a39f..0000000 --- a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_beta__hblt0.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x261760b0) reason :GPU Hang -✖ ! [rocm7_beta] gpt-oss-120b-mxfp4-00001-of-00003 __hblt0 failed (exit 134) diff --git a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_beta__hblt0__fa1.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_beta__hblt0__fa1.log deleted file mode 100644 index 2a30ca3..0000000 --- a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_beta__hblt0__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 1 | 0 | pp512 | 323.04 ± 0.94 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 1 | 0 | tg128 | 44.01 ± 0.01 | - -build: 79c1160b (6123) diff --git a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc-rocwmma.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc-rocwmma.log index 77cb354..17a14c7 100644 --- a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc-rocwmma.log +++ b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc-rocwmma.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 0 | pp512 | 586.82 ± 5.23 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 0 | tg128 | 44.72 ± 0.30 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 635.84 ± 5.72 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 45.26 ± 0.01 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc-rocwmma__fa1.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc-rocwmma__fa1.log index ae5c27f..8695248 100644 --- a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc-rocwmma__fa1.log +++ b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 1 | 0 | pp512 | 684.17 ± 67.05 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 1 | 0 | tg128 | 44.14 ± 0.27 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 1 | 0 | pp512 | 708.36 ± 12.96 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 1 | 0 | tg128 | 44.85 ± 0.01 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc-rocwmma__hblt0.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc-rocwmma__hblt0.log index c0da5bb..117d484 100644 --- a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc-rocwmma__hblt0.log +++ b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc-rocwmma__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 350.89 ± 1.88 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 44.93 ± 0.01 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 650.68 ± 9.08 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 45.26 ± 0.01 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc-rocwmma__hblt0__fa1.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc-rocwmma__hblt0__fa1.log index 4973b6a..173cfbe 100644 --- a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc-rocwmma__hblt0__fa1.log +++ b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc-rocwmma__hblt0__fa1.log @@ -2,4 +2,9 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc-rocwmma] gpt-oss-120b-mxfp4-00001-of-00003 __hblt0__fa1 failed (exit 134) +| model | size | params | backend | ngl | fa | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 1 | 0 | pp512 | 734.35 ± 10.26 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 1 | 0 | tg128 | 44.85 ± 0.00 | + +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc.log index 933469d..a5dcda7 100644 --- a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc.log +++ b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 0 | pp512 | 589.82 ± 5.12 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 0 | tg128 | 45.12 ± 0.01 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 646.07 ± 6.86 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 45.50 ± 0.01 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc__fa1.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc__fa1.log index 456548d..c8991e9 100644 --- a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc__fa1.log +++ b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 1 | 0 | pp512 | 540.27 ± 2.82 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 1 | 0 | tg128 | 43.89 ± 0.00 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 1 | 0 | pp512 | 541.57 ± 3.26 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 1 | 0 | tg128 | 44.31 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc__hblt0.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc__hblt0.log index b3222aa..f2a91fb 100644 --- a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc__hblt0.log +++ b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 354.60 ± 1.20 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 45.04 ± 0.01 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 0 | pp512 | 657.58 ± 3.78 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 0 | tg128 | 45.56 ± 0.01 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc__hblt0__fa1.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc__hblt0__fa1.log index b82d07b..698f338 100644 --- a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc__hblt0__fa1.log +++ b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc__hblt0__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 1 | 0 | pp512 | 319.46 ± 0.48 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 1 | 0 | tg128 | 43.90 ± 0.00 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 1 | 0 | pp512 | 550.79 ± 2.99 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 99 | 1 | 0 | tg128 | 44.41 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_amdvlk.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_amdvlk.log index 7d5b354..9d4e9e1 100644 --- a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_amdvlk.log +++ b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_amdvlk.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | Vulkan | 999 | 0 | pp512 | 488.47 ± 2.30 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | Vulkan | 999 | 0 | tg128 | 48.21 ± 0.02 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | Vulkan | 99 | 0 | pp512 | 485.54 ± 2.45 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | Vulkan | 99 | 0 | tg128 | 49.29 ± 0.03 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_amdvlk__fa1.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_amdvlk__fa1.log index 3441dc1..c4cd434 100644 --- a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_amdvlk__fa1.log +++ b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_amdvlk__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | Vulkan | 999 | 1 | 0 | pp512 | 547.53 ± 3.03 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | Vulkan | 999 | 1 | 0 | tg128 | 47.49 ± 0.08 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | Vulkan | 99 | 1 | 0 | pp512 | 540.81 ± 2.56 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | Vulkan | 99 | 1 | 0 | tg128 | 48.25 ± 0.03 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_radv.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_radv.log index f6cbe94..acd5b3a 100644 --- a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_radv.log +++ b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_radv.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | Vulkan | 999 | 0 | pp512 | 239.44 ± 1.23 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | Vulkan | 999 | 0 | tg128 | 49.15 ± 0.02 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | Vulkan | 99 | 0 | pp512 | 239.24 ± 1.27 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | Vulkan | 99 | 0 | tg128 | 50.39 ± 0.05 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_radv__fa1.log b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_radv__fa1.log index 5538ed2..f04d91c 100644 --- a/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_radv__fa1.log +++ b/benchmark/results/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_radv__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | Vulkan | 999 | 1 | 0 | pp512 | 255.37 ± 1.68 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | Vulkan | 999 | 1 | 0 | tg128 | 49.31 ± 0.08 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | Vulkan | 99 | 1 | 0 | pp512 | 255.50 ± 1.49 | +| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | Vulkan | 99 | 1 | 0 | tg128 | 50.41 ± 0.04 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-20b-F32__rocm6_4_2-rocwmma.log b/benchmark/results/gpt-oss-20b-F32__rocm6_4_2-rocwmma.log index f7c6172..6698000 100644 --- a/benchmark/results/gpt-oss-20b-F32__rocm6_4_2-rocwmma.log +++ b/benchmark/results/gpt-oss-20b-F32__rocm6_4_2-rocwmma.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 324.31 ± 4.50 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 26.87 ± 0.01 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 324.30 ± 4.23 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 27.10 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-20b-F32__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/gpt-oss-20b-F32__rocm6_4_2-rocwmma__fa1.log index 020f7b9..5f3dda7 100644 --- a/benchmark/results/gpt-oss-20b-F32__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/gpt-oss-20b-F32__rocm6_4_2-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 343.30 ± 5.27 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 26.76 ± 0.00 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 342.14 ± 4.83 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 27.05 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-20b-F32__rocm6_4_2.log b/benchmark/results/gpt-oss-20b-F32__rocm6_4_2.log index 88a7e15..5e1169a 100644 --- a/benchmark/results/gpt-oss-20b-F32__rocm6_4_2.log +++ b/benchmark/results/gpt-oss-20b-F32__rocm6_4_2.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 0 | pp512 | 322.55 ± 4.18 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 0 | tg128 | 24.90 ± 0.02 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 324.36 ± 4.35 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 27.12 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-20b-F32__rocm6_4_2__fa1.log b/benchmark/results/gpt-oss-20b-F32__rocm6_4_2__fa1.log index 41e5d8b..a1a5a05 100644 --- a/benchmark/results/gpt-oss-20b-F32__rocm6_4_2__fa1.log +++ b/benchmark/results/gpt-oss-20b-F32__rocm6_4_2__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 1 | 0 | pp512 | 304.86 ± 3.77 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 1 | 0 | tg128 | 26.58 ± 0.00 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 304.23 ± 3.73 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 26.85 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-20b-F32__rocm6_4_3-rocwmma.log b/benchmark/results/gpt-oss-20b-F32__rocm6_4_3-rocwmma.log new file mode 100644 index 0000000..45d2dea --- /dev/null +++ b/benchmark/results/gpt-oss-20b-F32__rocm6_4_3-rocwmma.log @@ -0,0 +1,15 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +hipBLASLt error: Heuristic Fetch Failed! +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 1198.51 ± 10.39 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 27.14 ± 0.00 | + +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gpt-oss-20b-F32__rocm6_4_2__fa1.log b/benchmark/results/gpt-oss-20b-F32__rocm6_4_3-rocwmma__fa1.log similarity index 78% rename from benchmark/results_08-08-2025/gpt-oss-20b-F32__rocm6_4_2__fa1.log rename to benchmark/results/gpt-oss-20b-F32__rocm6_4_3-rocwmma__fa1.log index 266806c..219d081 100644 --- a/benchmark/results_08-08-2025/gpt-oss-20b-F32__rocm6_4_2__fa1.log +++ b/benchmark/results/gpt-oss-20b-F32__rocm6_4_3-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 1 | 0 | pp512 | 257.11 ± 2.63 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 1 | 0 | tg128 | 26.47 ± 0.08 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 1506.44 ± 7.03 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 27.10 ± 0.01 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gpt-oss-20b-F32__rocm6_4_2-rocwmma.log b/benchmark/results/gpt-oss-20b-F32__rocm6_4_3-rocwmma__hblt0.log similarity index 79% rename from benchmark/results_08-08-2025/gpt-oss-20b-F32__rocm6_4_2-rocwmma.log rename to benchmark/results/gpt-oss-20b-F32__rocm6_4_3-rocwmma__hblt0.log index e1b0205..2c0a1a3 100644 --- a/benchmark/results_08-08-2025/gpt-oss-20b-F32__rocm6_4_2-rocwmma.log +++ b/benchmark/results/gpt-oss-20b-F32__rocm6_4_3-rocwmma__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 0 | pp512 | 324.54 ± 4.39 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 0 | tg128 | 26.87 ± 0.00 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 326.80 ± 4.56 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 27.13 ± 0.01 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gpt-oss-20b-F32__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/gpt-oss-20b-F32__rocm6_4_3-rocwmma__hblt0__fa1.log similarity index 79% rename from benchmark/results_08-08-2025/gpt-oss-20b-F32__rocm6_4_2-rocwmma__fa1.log rename to benchmark/results/gpt-oss-20b-F32__rocm6_4_3-rocwmma__hblt0__fa1.log index 8e851e8..b04117d 100644 --- a/benchmark/results_08-08-2025/gpt-oss-20b-F32__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/gpt-oss-20b-F32__rocm6_4_3-rocwmma__hblt0__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 1 | 0 | pp512 | 380.87 ± 8.21 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 1 | 0 | tg128 | 26.79 ± 0.00 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 350.18 ± 5.10 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 27.09 ± 0.00 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-20b-F32__rocm6_4_3.log b/benchmark/results/gpt-oss-20b-F32__rocm6_4_3.log new file mode 100644 index 0000000..8d5aeeb --- /dev/null +++ b/benchmark/results/gpt-oss-20b-F32__rocm6_4_3.log @@ -0,0 +1,15 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +hipBLASLt error: Heuristic Fetch Failed! +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 1185.57 ± 6.55 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 27.12 ± 0.01 | + +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-20b-F32__rocm7_beta__fa1.log b/benchmark/results/gpt-oss-20b-F32__rocm6_4_3__fa1.log similarity index 69% rename from benchmark/results/gpt-oss-20b-F32__rocm7_beta__fa1.log rename to benchmark/results/gpt-oss-20b-F32__rocm6_4_3__fa1.log index 3123235..4138942 100644 --- a/benchmark/results/gpt-oss-20b-F32__rocm7_beta__fa1.log +++ b/benchmark/results/gpt-oss-20b-F32__rocm6_4_3__fa1.log @@ -1,10 +1,10 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 1 | 0 | pp512 | 1011.32 ± 4.33 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 1 | 0 | tg128 | 26.65 ± 0.00 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 1000.77 ± 2.37 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 26.83 ± 0.00 | -build: 79c1160b (6123) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gpt-oss-20b-F32__rocm6_4_2.log b/benchmark/results/gpt-oss-20b-F32__rocm6_4_3__hblt0.log similarity index 79% rename from benchmark/results_08-08-2025/gpt-oss-20b-F32__rocm6_4_2.log rename to benchmark/results/gpt-oss-20b-F32__rocm6_4_3__hblt0.log index d9bd7eb..76ec711 100644 --- a/benchmark/results_08-08-2025/gpt-oss-20b-F32__rocm6_4_2.log +++ b/benchmark/results/gpt-oss-20b-F32__rocm6_4_3__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 0 | pp512 | 323.86 ± 4.33 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 0 | tg128 | 26.27 ± 0.00 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 322.00 ± 4.37 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 27.14 ± 0.01 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-20b-F32__rocm7_beta__hblt0__fa1.log b/benchmark/results/gpt-oss-20b-F32__rocm6_4_3__hblt0__fa1.log similarity index 69% rename from benchmark/results/gpt-oss-20b-F32__rocm7_beta__hblt0__fa1.log rename to benchmark/results/gpt-oss-20b-F32__rocm6_4_3__hblt0__fa1.log index 41d7fd7..9ce8ee4 100644 --- a/benchmark/results/gpt-oss-20b-F32__rocm7_beta__hblt0__fa1.log +++ b/benchmark/results/gpt-oss-20b-F32__rocm6_4_3__hblt0__fa1.log @@ -1,10 +1,10 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 301.30 ± 4.81 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 26.65 ± 0.00 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 303.26 ± 4.84 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 26.90 ± 0.00 | -build: 79c1160b (6123) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-20b-F32__rocm7_beta.log b/benchmark/results/gpt-oss-20b-F32__rocm7_beta.log deleted file mode 100644 index c97bff7..0000000 --- a/benchmark/results/gpt-oss-20b-F32__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 0 | pp512 | 1135.90 ± 9.10 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 0 | tg128 | 26.88 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/gpt-oss-20b-F32__rocm7_beta__hblt0.log b/benchmark/results/gpt-oss-20b-F32__rocm7_beta__hblt0.log deleted file mode 100644 index c01b9a9..0000000 --- a/benchmark/results/gpt-oss-20b-F32__rocm7_beta__hblt0.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 313.05 ± 6.96 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 26.86 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/gpt-oss-20b-F32__rocm7_rc-rocwmma.log b/benchmark/results/gpt-oss-20b-F32__rocm7_rc-rocwmma.log index 1912f8c..d80b5e6 100644 --- a/benchmark/results/gpt-oss-20b-F32__rocm7_rc-rocwmma.log +++ b/benchmark/results/gpt-oss-20b-F32__rocm7_rc-rocwmma.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 0 | pp512 | 1130.14 ± 7.45 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 0 | tg128 | 26.84 ± 0.00 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 1256.75 ± 10.54 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 27.11 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-20b-F32__rocm7_rc-rocwmma__fa1.log b/benchmark/results/gpt-oss-20b-F32__rocm7_rc-rocwmma__fa1.log index 5046a32..2648480 100644 --- a/benchmark/results/gpt-oss-20b-F32__rocm7_rc-rocwmma__fa1.log +++ b/benchmark/results/gpt-oss-20b-F32__rocm7_rc-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 1 | 0 | pp512 | 1502.62 ± 12.84 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 1 | 0 | tg128 | 26.67 ± 0.00 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 1481.17 ± 9.67 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 27.03 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-20b-F32__rocm7_rc-rocwmma__hblt0.log b/benchmark/results/gpt-oss-20b-F32__rocm7_rc-rocwmma__hblt0.log index c83d4a2..9c6df9d 100644 --- a/benchmark/results/gpt-oss-20b-F32__rocm7_rc-rocwmma__hblt0.log +++ b/benchmark/results/gpt-oss-20b-F32__rocm7_rc-rocwmma__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 319.92 ± 6.39 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 26.83 ± 0.00 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 1202.19 ± 5.53 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 27.10 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-20b-F32__rocm7_rc-rocwmma__hblt0__fa1.log b/benchmark/results/gpt-oss-20b-F32__rocm7_rc-rocwmma__hblt0__fa1.log index fe7a810..8b68b31 100644 --- a/benchmark/results/gpt-oss-20b-F32__rocm7_rc-rocwmma__hblt0__fa1.log +++ b/benchmark/results/gpt-oss-20b-F32__rocm7_rc-rocwmma__hblt0__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 338.36 ± 5.02 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 26.71 ± 0.00 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 1422.90 ± 11.48 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 27.04 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-20b-F32__rocm7_rc.log b/benchmark/results/gpt-oss-20b-F32__rocm7_rc.log index 1c550a3..8a71c43 100644 --- a/benchmark/results/gpt-oss-20b-F32__rocm7_rc.log +++ b/benchmark/results/gpt-oss-20b-F32__rocm7_rc.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 0 | pp512 | 1130.86 ± 14.88 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 0 | tg128 | 26.89 ± 0.00 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 1253.01 ± 23.20 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 27.11 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-20b-F32__rocm7_rc__fa1.log b/benchmark/results/gpt-oss-20b-F32__rocm7_rc__fa1.log index a59b260..eaf7b98 100644 --- a/benchmark/results/gpt-oss-20b-F32__rocm7_rc__fa1.log +++ b/benchmark/results/gpt-oss-20b-F32__rocm7_rc__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 1 | 0 | pp512 | 1007.82 ± 22.14 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 1 | 0 | tg128 | 26.66 ± 0.00 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 1005.24 ± 32.45 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 26.89 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-20b-F32__rocm7_rc__hblt0.log b/benchmark/results/gpt-oss-20b-F32__rocm7_rc__hblt0.log index 84a8d46..685527d 100644 --- a/benchmark/results/gpt-oss-20b-F32__rocm7_rc__hblt0.log +++ b/benchmark/results/gpt-oss-20b-F32__rocm7_rc__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 321.80 ± 6.18 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 26.83 ± 0.01 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 1220.02 ± 12.30 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 27.17 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-20b-F32__rocm7_rc__hblt0__fa1.log b/benchmark/results/gpt-oss-20b-F32__rocm7_rc__hblt0__fa1.log index ec0c58d..18915d7 100644 --- a/benchmark/results/gpt-oss-20b-F32__rocm7_rc__hblt0__fa1.log +++ b/benchmark/results/gpt-oss-20b-F32__rocm7_rc__hblt0__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 302.84 ± 5.01 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 26.61 ± 0.00 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 985.58 ± 10.64 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 26.88 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-20b-F32__vulkan_amdvlk.log b/benchmark/results/gpt-oss-20b-F32__vulkan_amdvlk.log index 6875f68..381cc89 100644 --- a/benchmark/results/gpt-oss-20b-F32__vulkan_amdvlk.log +++ b/benchmark/results/gpt-oss-20b-F32__vulkan_amdvlk.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | Vulkan | 999 | 0 | pp512 | 369.60 ± 1.30 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | Vulkan | 999 | 0 | tg128 | 8.72 ± 0.01 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | Vulkan | 99 | 0 | pp512 | 367.61 ± 1.90 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | Vulkan | 99 | 0 | tg128 | 8.69 ± 0.01 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-20b-F32__vulkan_amdvlk__fa1.log b/benchmark/results/gpt-oss-20b-F32__vulkan_amdvlk__fa1.log index b4f9322..4212c20 100644 --- a/benchmark/results/gpt-oss-20b-F32__vulkan_amdvlk__fa1.log +++ b/benchmark/results/gpt-oss-20b-F32__vulkan_amdvlk__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | Vulkan | 999 | 1 | 0 | pp512 | 389.96 ± 1.87 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | Vulkan | 999 | 1 | 0 | tg128 | 8.70 ± 0.01 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | Vulkan | 99 | 1 | 0 | pp512 | 386.12 ± 1.98 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | Vulkan | 99 | 1 | 0 | tg128 | 8.66 ± 0.01 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-20b-F32__vulkan_radv.log b/benchmark/results/gpt-oss-20b-F32__vulkan_radv.log index ad6bfa4..a3f7dda 100644 --- a/benchmark/results/gpt-oss-20b-F32__vulkan_radv.log +++ b/benchmark/results/gpt-oss-20b-F32__vulkan_radv.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | Vulkan | 999 | 0 | pp512 | 318.04 ± 1.50 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | Vulkan | 999 | 0 | tg128 | 7.89 ± 0.01 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | Vulkan | 99 | 0 | pp512 | 315.56 ± 1.40 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | Vulkan | 99 | 0 | tg128 | 7.86 ± 0.01 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-20b-F32__vulkan_radv__fa1.log b/benchmark/results/gpt-oss-20b-F32__vulkan_radv__fa1.log index 072c052..257f941 100644 --- a/benchmark/results/gpt-oss-20b-F32__vulkan_radv__fa1.log +++ b/benchmark/results/gpt-oss-20b-F32__vulkan_radv__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | Vulkan | 999 | 1 | 0 | pp512 | 334.64 ± 1.46 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | Vulkan | 999 | 1 | 0 | tg128 | 7.90 ± 0.01 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | Vulkan | 99 | 1 | 0 | pp512 | 333.31 ± 1.47 | +| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | Vulkan | 99 | 1 | 0 | tg128 | 7.92 ± 0.01 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_2-rocwmma.log b/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_2-rocwmma.log index 59744c9..28cbd07 100644 --- a/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_2-rocwmma.log +++ b/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_2-rocwmma.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 581.92 ± 2.00 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 64.34 ± 0.02 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 582.60 ± 4.90 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 64.91 ± 0.01 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_2-rocwmma__fa1.log index 97f911b..985eaf7 100644 --- a/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_2-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 642.40 ± 3.59 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 63.74 ± 0.00 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 644.05 ± 3.87 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 64.63 ± 0.01 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_2.log b/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_2.log index 2bd619e..c8bf125 100644 --- a/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_2.log +++ b/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_2.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 0 | pp512 | 582.94 ± 2.35 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 0 | tg128 | 64.35 ± 0.01 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 581.11 ± 2.96 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 65.00 ± 0.02 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_2__fa1.log b/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_2__fa1.log index 8a01f71..320f480 100644 --- a/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_2__fa1.log +++ b/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_2__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 1 | 0 | pp512 | 522.14 ± 1.92 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 1 | 0 | tg128 | 62.97 ± 0.01 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 522.29 ± 2.36 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 63.63 ± 0.00 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_3-rocwmma.log b/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_3-rocwmma.log new file mode 100644 index 0000000..2c92204 --- /dev/null +++ b/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_3-rocwmma.log @@ -0,0 +1,15 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +rocBLAS error: No hipBLASLt solution found +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 1184.03 ± 8.37 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 65.07 ± 0.00 | + +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__rocm6_4_2-rocwmma__fa1.log b/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_3-rocwmma__fa1.log similarity index 79% rename from benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__rocm6_4_2-rocwmma__fa1.log rename to benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_3-rocwmma__fa1.log index cb2c45b..cd2e38f 100644 --- a/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__rocm6_4_2-rocwmma__fa1.log +++ b/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_3-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 1 | 0 | pp512 | 649.48 ± 3.21 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 1 | 0 | tg128 | 64.18 ± 0.02 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 1480.28 ± 9.38 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 64.45 ± 0.02 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__rocm6_4_2-rocwmma.log b/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_3-rocwmma__hblt0.log similarity index 79% rename from benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__rocm6_4_2-rocwmma.log rename to benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_3-rocwmma__hblt0.log index c377132..f7e992e 100644 --- a/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__rocm6_4_2-rocwmma.log +++ b/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_3-rocwmma__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 0 | pp512 | 580.83 ± 2.46 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 0 | tg128 | 64.47 ± 0.02 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 584.04 ± 2.52 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 64.87 ± 0.02 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__rocm6_4_2__fa1.log b/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_3-rocwmma__hblt0__fa1.log similarity index 78% rename from benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__rocm6_4_2__fa1.log rename to benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_3-rocwmma__hblt0__fa1.log index 34d817d..2016c11 100644 --- a/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__rocm6_4_2__fa1.log +++ b/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_3-rocwmma__hblt0__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 1 | 0 | pp512 | 394.67 ± 1.08 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 1 | 0 | tg128 | 62.97 ± 0.00 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 643.25 ± 3.86 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 64.67 ± 0.01 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_3.log b/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_3.log new file mode 100644 index 0000000..d5473e4 --- /dev/null +++ b/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_3.log @@ -0,0 +1,15 @@ +ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no +ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no +ggml_cuda_init: found 1 ROCm devices: + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 +rocBLAS error: No hipBLASLt solution found +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_HIPBLASLT_ERROR environment variable is set. + +rocBLAS warning: hipBlasLT failed, falling back to tensile. +This message will be only be displayed once, unless the ROCBLAS_VERBOSE_TENSILE_ERROR environment variable is set. +| model | size | params | backend | ngl | mmap | test | t/s | +| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 1171.02 ± 7.04 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 64.94 ± 0.04 | + +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-20b-mxfp4__rocm7_beta__fa1.log b/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_3__fa1.log similarity index 69% rename from benchmark/results/gpt-oss-20b-mxfp4__rocm7_beta__fa1.log rename to benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_3__fa1.log index 9c6c567..abd5fd2 100644 --- a/benchmark/results/gpt-oss-20b-mxfp4__rocm7_beta__fa1.log +++ b/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_3__fa1.log @@ -1,10 +1,10 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 1 | 0 | pp512 | 1005.66 ± 1.52 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 1 | 0 | tg128 | 63.07 ± 0.01 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 996.31 ± 6.53 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 63.68 ± 0.01 | -build: 79c1160b (6123) +build: 1fe00296 (6182) diff --git a/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__rocm6_4_2.log b/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_3__hblt0.log similarity index 79% rename from benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__rocm6_4_2.log rename to benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_3__hblt0.log index 343b2b0..be6782a 100644 --- a/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__rocm6_4_2.log +++ b/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_3__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 0 | pp512 | 582.89 ± 2.32 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 0 | tg128 | 64.45 ± 0.02 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 582.51 ± 2.41 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 64.89 ± 0.01 | -build: cd6983d5 (6119) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-20b-mxfp4__rocm7_beta__hblt0__fa1.log b/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_3__hblt0__fa1.log similarity index 68% rename from benchmark/results/gpt-oss-20b-mxfp4__rocm7_beta__hblt0__fa1.log rename to benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_3__hblt0__fa1.log index 661d58f..90bc0d7 100644 --- a/benchmark/results/gpt-oss-20b-mxfp4__rocm7_beta__hblt0__fa1.log +++ b/benchmark/results/gpt-oss-20b-mxfp4__rocm6_4_3__hblt0__fa1.log @@ -1,10 +1,10 @@ ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 + Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 528.92 ± 2.02 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 63.00 ± 0.01 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 522.63 ± 1.74 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 63.66 ± 0.03 | -build: 79c1160b (6123) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-20b-mxfp4__rocm7_beta.log b/benchmark/results/gpt-oss-20b-mxfp4__rocm7_beta.log deleted file mode 100644 index 7ccf6e0..0000000 --- a/benchmark/results/gpt-oss-20b-mxfp4__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 0 | pp512 | 1128.54 ± 2.40 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 0 | tg128 | 64.39 ± 0.01 | - -build: 79c1160b (6123) diff --git a/benchmark/results/gpt-oss-20b-mxfp4__rocm7_beta__hblt0.log b/benchmark/results/gpt-oss-20b-mxfp4__rocm7_beta__hblt0.log deleted file mode 100644 index 57a687a..0000000 --- a/benchmark/results/gpt-oss-20b-mxfp4__rocm7_beta__hblt0.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 585.03 ± 1.84 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 64.36 ± 0.01 | - -build: 79c1160b (6123) diff --git a/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc-rocwmma.log b/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc-rocwmma.log index f49d7ff..8e4bdde 100644 --- a/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc-rocwmma.log +++ b/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc-rocwmma.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 0 | pp512 | 1124.54 ± 9.14 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 0 | tg128 | 64.19 ± 0.01 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 1236.64 ± 11.20 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 64.78 ± 0.01 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc-rocwmma__fa1.log b/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc-rocwmma__fa1.log index 82390bf..c87a02c 100644 --- a/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc-rocwmma__fa1.log +++ b/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc-rocwmma__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 1 | 0 | pp512 | 1474.70 ± 11.50 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 1 | 0 | tg128 | 63.31 ± 0.01 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 1460.58 ± 11.92 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 64.26 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc-rocwmma__hblt0.log b/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc-rocwmma__hblt0.log index b1e54db..2dd9efb 100644 --- a/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc-rocwmma__hblt0.log +++ b/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc-rocwmma__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 583.69 ± 2.09 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 64.26 ± 0.01 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 1299.34 ± 7.77 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 64.85 ± 0.00 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc-rocwmma__hblt0__fa1.log b/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc-rocwmma__hblt0__fa1.log index 3068d5b..fc2acae 100644 --- a/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc-rocwmma__hblt0__fa1.log +++ b/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc-rocwmma__hblt0__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 642.92 ± 1.97 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 63.28 ± 0.01 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 1516.33 ± 21.51 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 64.40 ± 0.01 | -build: 34c9d765 (6122) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc.log b/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc.log index d848311..34dc4e8 100644 --- a/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc.log +++ b/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 0 | pp512 | 1125.60 ± 1.90 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 0 | tg128 | 64.35 ± 0.01 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 1246.14 ± 8.32 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 65.15 ± 0.01 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc__fa1.log b/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc__fa1.log index bd4d588..f6feec3 100644 --- a/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc__fa1.log +++ b/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 1 | 0 | pp512 | 997.74 ± 8.16 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 1 | 0 | tg128 | 63.00 ± 0.00 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 1010.38 ± 6.35 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 63.49 ± 0.01 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc__hblt0.log b/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc__hblt0.log index 8b8a81a..425b7fd 100644 --- a/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc__hblt0.log +++ b/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc__hblt0.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 584.02 ± 1.44 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 64.50 ± 0.01 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | pp512 | 1303.74 ± 6.94 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 0 | tg128 | 65.10 ± 0.01 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc__hblt0__fa1.log b/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc__hblt0__fa1.log index d0567df..0bd1151 100644 --- a/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc__hblt0__fa1.log +++ b/benchmark/results/gpt-oss-20b-mxfp4__rocm7_rc__hblt0__fa1.log @@ -4,7 +4,7 @@ ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 525.48 ± 1.39 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 63.04 ± 0.01 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 1037.92 ± 11.67 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 63.63 ± 0.01 | -build: 79c1160b (6123) +build: de219279 (6181) diff --git a/benchmark/results/gpt-oss-20b-mxfp4__vulkan_amdvlk.log b/benchmark/results/gpt-oss-20b-mxfp4__vulkan_amdvlk.log index ec3e361..10fffaa 100644 --- a/benchmark/results/gpt-oss-20b-mxfp4__vulkan_amdvlk.log +++ b/benchmark/results/gpt-oss-20b-mxfp4__vulkan_amdvlk.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan | 999 | 0 | pp512 | 1218.18 ± 8.08 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan | 999 | 0 | tg128 | 69.76 ± 0.07 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan | 99 | 0 | pp512 | 1220.69 ± 8.95 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan | 99 | 0 | tg128 | 71.42 ± 0.20 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-20b-mxfp4__vulkan_amdvlk__fa1.log b/benchmark/results/gpt-oss-20b-mxfp4__vulkan_amdvlk__fa1.log index fbda2c7..185b49c 100644 --- a/benchmark/results/gpt-oss-20b-mxfp4__vulkan_amdvlk__fa1.log +++ b/benchmark/results/gpt-oss-20b-mxfp4__vulkan_amdvlk__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan | 999 | 1 | 0 | pp512 | 1482.59 ± 12.76 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan | 999 | 1 | 0 | tg128 | 68.63 ± 0.11 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan | 99 | 1 | 0 | pp512 | 1467.61 ± 12.70 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan | 99 | 1 | 0 | tg128 | 69.47 ± 0.09 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-20b-mxfp4__vulkan_radv.log b/benchmark/results/gpt-oss-20b-mxfp4__vulkan_radv.log index a2342cc..cf40790 100644 --- a/benchmark/results/gpt-oss-20b-mxfp4__vulkan_radv.log +++ b/benchmark/results/gpt-oss-20b-mxfp4__vulkan_radv.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan | 999 | 0 | pp512 | 649.86 ± 5.16 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan | 999 | 0 | tg128 | 70.72 ± 0.04 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan | 99 | 0 | pp512 | 651.21 ± 5.24 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan | 99 | 0 | tg128 | 72.35 ± 0.08 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/gpt-oss-20b-mxfp4__vulkan_radv__fa1.log b/benchmark/results/gpt-oss-20b-mxfp4__vulkan_radv__fa1.log index d1051a1..ed9a28f 100644 --- a/benchmark/results/gpt-oss-20b-mxfp4__vulkan_radv__fa1.log +++ b/benchmark/results/gpt-oss-20b-mxfp4__vulkan_radv__fa1.log @@ -2,7 +2,7 @@ ggml_vulkan: Found 1 Vulkan devices: ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat | model | size | params | backend | ngl | fa | mmap | test | t/s | | ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan | 999 | 1 | 0 | pp512 | 728.71 ± 8.40 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan | 999 | 1 | 0 | tg128 | 70.49 ± 0.04 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan | 99 | 1 | 0 | pp512 | 732.35 ± 7.51 | +| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan | 99 | 1 | 0 | tg128 | 72.05 ± 0.07 | -build: 34c9d765 (6122) +build: 1fe00296 (6182) diff --git a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm6_4_2.log b/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm6_4_2.log deleted file mode 100644 index 15e8d8f..0000000 --- a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm6_4_2.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 0 | pp512 | 33.76 ± 0.04 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 0 | tg128 | 4.48 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_beta.log b/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_beta.log deleted file mode 100644 index ab19f84..0000000 --- a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 0 | pp512 | 99.09 ± 0.10 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 0 | tg128 | 4.61 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_beta__fa1.log b/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_beta__fa1.log deleted file mode 100644 index 9bd37ad..0000000 --- a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_beta__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 1 | 0 | pp512 | 81.54 ± 0.11 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 1 | 0 | tg128 | 4.63 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_beta__hblt0.log b/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_beta__hblt0.log deleted file mode 100644 index b92eb71..0000000 --- a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_beta__hblt0.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x2595b0b0) reason :GPU Hang -✖ ! [rocm7_beta] llama3.3-70.6B-Q4_K_M __hblt0 failed (exit 134) diff --git a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_beta__hblt0__fa1.log b/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_beta__hblt0__fa1.log deleted file mode 100644 index 3fdb6c0..0000000 --- a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_beta__hblt0__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 99 | 1 | 0 | pp512 | 31.63 ± 0.02 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 99 | 1 | 0 | tg128 | 4.62 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_rc-rocwmma.log b/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_rc-rocwmma.log deleted file mode 100644 index 80974fd..0000000 --- a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_rc-rocwmma.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 0 | pp512 | 99.41 ± 0.11 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 0 | tg128 | 4.62 ± 0.00 | - -build: 34c9d765 (6122) diff --git a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_rc-rocwmma__fa1.log b/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_rc-rocwmma__fa1.log deleted file mode 100644 index 928b750..0000000 --- a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_rc-rocwmma__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 1 | 0 | pp512 | 106.70 ± 0.12 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 1 | 0 | tg128 | 4.60 ± 0.00 | - -build: 34c9d765 (6122) diff --git a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_rc-rocwmma__hblt0.log b/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_rc-rocwmma__hblt0.log deleted file mode 100644 index d78ff89..0000000 --- a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_rc-rocwmma__hblt0.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 99 | 0 | pp512 | 33.87 ± 0.08 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 99 | 0 | tg128 | 4.61 ± 0.00 | - -build: 34c9d765 (6122) diff --git a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_rc-rocwmma__hblt0__fa1.log b/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_rc-rocwmma__hblt0__fa1.log deleted file mode 100644 index 1757ab6..0000000 --- a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_rc-rocwmma__hblt0__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 99 | 1 | 0 | pp512 | 34.48 ± 0.05 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 99 | 1 | 0 | tg128 | 4.61 ± 0.00 | - -build: 34c9d765 (6122) diff --git a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_rc.log b/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_rc.log deleted file mode 100644 index 5075f85..0000000 --- a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_rc.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 0 | pp512 | 99.16 ± 0.09 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 0 | tg128 | 4.62 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_rc__fa1.log b/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_rc__fa1.log deleted file mode 100644 index afc44e9..0000000 --- a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_rc__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 1 | 0 | pp512 | 81.56 ± 0.09 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 1 | 0 | tg128 | 4.62 ± 0.00 | - -build: 79c1160b (6123) diff --git a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_rc__hblt0.log b/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_rc__hblt0.log deleted file mode 100644 index 9718332..0000000 --- a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_rc__hblt0.log +++ /dev/null @@ -1,5 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc] llama3.3-70.6B-Q4_K_M __hblt0 failed (exit 134) diff --git a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_rc__hblt0__fa1.log b/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_rc__hblt0__fa1.log deleted file mode 100644 index 9d5bb3c..0000000 --- a/benchmark/results/llama3.3-70.6B-Q4_K_M__rocm7_rc__hblt0__fa1.log +++ /dev/null @@ -1,5 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc] llama3.3-70.6B-Q4_K_M __hblt0__fa1 failed (exit 134) diff --git a/benchmark/results/llama3.3-70.6B-Q4_K_M__vulkan_amdvlk.log b/benchmark/results/llama3.3-70.6B-Q4_K_M__vulkan_amdvlk.log deleted file mode 100644 index f70a707..0000000 --- a/benchmark/results/llama3.3-70.6B-Q4_K_M__vulkan_amdvlk.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | Vulkan | 999 | 0 | pp512 | 72.73 ± 0.05 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | Vulkan | 999 | 0 | tg128 | 5.08 ± 0.00 | - -build: 34c9d765 (6122) diff --git a/benchmark/results/llama3.3-70.6B-Q4_K_M__vulkan_amdvlk__fa1.log b/benchmark/results/llama3.3-70.6B-Q4_K_M__vulkan_amdvlk__fa1.log deleted file mode 100644 index d13dc99..0000000 --- a/benchmark/results/llama3.3-70.6B-Q4_K_M__vulkan_amdvlk__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | Vulkan | 999 | 1 | 0 | pp512 | 73.47 ± 0.03 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | Vulkan | 999 | 1 | 0 | tg128 | 5.04 ± 0.00 | - -build: 34c9d765 (6122) diff --git a/benchmark/results/llama3.3-70.6B-Q4_K_M__vulkan_radv.log b/benchmark/results/llama3.3-70.6B-Q4_K_M__vulkan_radv.log deleted file mode 100644 index 5ccc1cf..0000000 --- a/benchmark/results/llama3.3-70.6B-Q4_K_M__vulkan_radv.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | Vulkan | 999 | 0 | pp512 | 78.79 ± 0.21 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | Vulkan | 999 | 0 | tg128 | 5.04 ± 0.00 | - -build: 34c9d765 (6122) diff --git a/benchmark/results/llama3.3-70.6B-Q4_K_M__vulkan_radv__fa1.log b/benchmark/results/llama3.3-70.6B-Q4_K_M__vulkan_radv__fa1.log deleted file mode 100644 index 375bd60..0000000 --- a/benchmark/results/llama3.3-70.6B-Q4_K_M__vulkan_radv__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | Vulkan | 999 | 1 | 0 | pp512 | 80.58 ± 0.13 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | Vulkan | 999 | 1 | 0 | tg128 | 5.03 ± 0.00 | - -build: 34c9d765 (6122) diff --git a/benchmark/results/run_benchmarks.log b/benchmark/results/run_benchmarks.log deleted file mode 100644 index b9e965c..0000000 --- a/benchmark/results/run_benchmarks.log +++ /dev/null @@ -1,1392 +0,0 @@ -Found 19 model(s) to bench: - • /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf - • /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf - • /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf - • /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf - • /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf - • /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf - • /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf - • /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf - • /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf - • /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf - • /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf - • /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf - • /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf - • /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf - • /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf - • /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf - • /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf - • /mnt/models/qwen-3-30B-A3B/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL.gguf - • /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf - - -▶ [rocm7_rc] gemma-3-27b-it-BF16-00001-of-00002 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf - - -▶ [rocm7_rc] gemma-3-27b-it-BF16-00001-of-00002 __fa1 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf -fa 1 - - -▶ [rocm7_beta] gemma-3-27b-it-BF16-00001-of-00002 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf - - -▶ [rocm7_beta] gemma-3-27b-it-BF16-00001-of-00002 __fa1 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf -fa 1 - - -▶ [rocm6_4_2-rocwmma] gemma-3-27b-it-BF16-00001-of-00002 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf - - -▶ [rocm6_4_2-rocwmma] gemma-3-27b-it-BF16-00001-of-00002 __fa1 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf -fa 1 - - -▶ [vulkan_radv] gemma-3-27b-it-BF16-00001-of-00002 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf - - -▶ [vulkan_radv] gemma-3-27b-it-BF16-00001-of-00002 __fa1 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf -fa 1 - - -▶ [vulkan_amdvlk] gemma-3-27b-it-BF16-00001-of-00002 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf - - * [vulkan_amdvlk] gemma-3-27b-it-BF16-00001-of-00002 : FAILED - -▶ [vulkan_amdvlk] gemma-3-27b-it-BF16-00001-of-00002 __fa1 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf -fa 1 - - * [vulkan_amdvlk] gemma-3-27b-it-BF16-00001-of-00002 __fa1 : FAILED - -▶ [rocm6_4_2] gemma-3-27b-it-BF16-00001-of-00002 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf - - -▶ [rocm6_4_2] gemma-3-27b-it-BF16-00001-of-00002 __fa1 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf -fa 1 - - -▶ [rocm7_rc-rocwmma] gemma-3-27b-it-BF16-00001-of-00002 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc-rocwmma.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf - - -▶ [rocm7_rc-rocwmma] gemma-3-27b-it-BF16-00001-of-00002 __fa1 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf -fa 1 - - -▶ [rocm7_rc] gemma-3-12b-it-UD-Q8_K_XL - → log: results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf - - -▶ [rocm7_rc] gemma-3-12b-it-UD-Q8_K_XL __fa1 - → log: results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf -fa 1 - - -▶ [rocm7_beta] gemma-3-12b-it-UD-Q8_K_XL - → log: results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf - - -▶ [rocm7_beta] gemma-3-12b-it-UD-Q8_K_XL __fa1 - → log: results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf -fa 1 - - -▶ [rocm6_4_2-rocwmma] gemma-3-12b-it-UD-Q8_K_XL - → log: results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf - - -▶ [rocm6_4_2-rocwmma] gemma-3-12b-it-UD-Q8_K_XL __fa1 - → log: results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf -fa 1 - - -▶ [vulkan_radv] gemma-3-12b-it-UD-Q8_K_XL - → log: results/gemma-3-12b-it-UD-Q8_K_XL__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf - - -▶ [vulkan_radv] gemma-3-12b-it-UD-Q8_K_XL __fa1 - → log: results/gemma-3-12b-it-UD-Q8_K_XL__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf -fa 1 - - -▶ [vulkan_amdvlk] gemma-3-12b-it-UD-Q8_K_XL - → log: results/gemma-3-12b-it-UD-Q8_K_XL__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf - - -▶ [vulkan_amdvlk] gemma-3-12b-it-UD-Q8_K_XL __fa1 - → log: results/gemma-3-12b-it-UD-Q8_K_XL__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf -fa 1 - - -▶ [rocm6_4_2] gemma-3-12b-it-UD-Q8_K_XL - → log: results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf - - -▶ [rocm6_4_2] gemma-3-12b-it-UD-Q8_K_XL __fa1 - → log: results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf -fa 1 - - -▶ [rocm7_rc-rocwmma] gemma-3-12b-it-UD-Q8_K_XL - → log: results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc-rocwmma.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf - - -▶ [rocm7_rc-rocwmma] gemma-3-12b-it-UD-Q8_K_XL __fa1 - → log: results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf -fa 1 - - -▶ [rocm7_rc] gemma-3-4b-it-Q3_K_S - → log: results/gemma-3-4b-it-Q3_K_S__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf - - -▶ [rocm7_rc] gemma-3-4b-it-Q3_K_S __fa1 - → log: results/gemma-3-4b-it-Q3_K_S__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf -fa 1 - - -▶ [rocm7_beta] gemma-3-4b-it-Q3_K_S - → log: results/gemma-3-4b-it-Q3_K_S__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf - - -▶ [rocm7_beta] gemma-3-4b-it-Q3_K_S __fa1 - → log: results/gemma-3-4b-it-Q3_K_S__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf -fa 1 - - -▶ [rocm6_4_2-rocwmma] gemma-3-4b-it-Q3_K_S - → log: results/gemma-3-4b-it-Q3_K_S__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf - - -▶ [rocm6_4_2-rocwmma] gemma-3-4b-it-Q3_K_S __fa1 - → log: results/gemma-3-4b-it-Q3_K_S__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf -fa 1 - - -▶ [vulkan_radv] gemma-3-4b-it-Q3_K_S - → log: results/gemma-3-4b-it-Q3_K_S__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf - - -▶ [vulkan_radv] gemma-3-4b-it-Q3_K_S __fa1 - → log: results/gemma-3-4b-it-Q3_K_S__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf -fa 1 - - -▶ [vulkan_amdvlk] gemma-3-4b-it-Q3_K_S - → log: results/gemma-3-4b-it-Q3_K_S__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf - - -▶ [vulkan_amdvlk] gemma-3-4b-it-Q3_K_S __fa1 - → log: results/gemma-3-4b-it-Q3_K_S__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf -fa 1 - - -▶ [rocm6_4_2] gemma-3-4b-it-Q3_K_S - → log: results/gemma-3-4b-it-Q3_K_S__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf - - -▶ [rocm6_4_2] gemma-3-4b-it-Q3_K_S __fa1 - → log: results/gemma-3-4b-it-Q3_K_S__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf -fa 1 - - -▶ [rocm7_rc-rocwmma] gemma-3-4b-it-Q3_K_S - → log: results/gemma-3-4b-it-Q3_K_S__rocm7_rc-rocwmma.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf - - -▶ [rocm7_rc-rocwmma] gemma-3-4b-it-Q3_K_S __fa1 - → log: results/gemma-3-4b-it-Q3_K_S__rocm7_rc-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf -fa 1 - - -▶ [rocm7_rc] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf - - -▶ [rocm7_rc] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [rocm7_beta] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf - - -▶ [rocm7_beta] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [rocm6_4_2-rocwmma] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf - - * [rocm6_4_2-rocwmma] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 : FAILED - -▶ [rocm6_4_2-rocwmma] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [vulkan_radv] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf - - -▶ [vulkan_radv] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [vulkan_amdvlk] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf - - -▶ [vulkan_amdvlk] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [rocm6_4_2] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf - - -▶ [rocm6_4_2] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - * [rocm6_4_2] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __fa1 : FAILED - -▶ [rocm7_rc-rocwmma] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf - - -▶ [rocm7_rc-rocwmma] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [rocm7_rc] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf - - -▶ [rocm7_rc] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf -fa 1 - - -▶ [rocm7_beta] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf - - -▶ [rocm7_beta] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf -fa 1 - - -▶ [rocm6_4_2-rocwmma] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf - - * [rocm6_4_2-rocwmma] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 : FAILED - -▶ [rocm6_4_2-rocwmma] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf -fa 1 - - * [rocm6_4_2-rocwmma] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 : FAILED - -▶ [vulkan_radv] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf - - -▶ [vulkan_radv] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf -fa 1 - - -▶ [vulkan_amdvlk] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf - - -▶ [vulkan_amdvlk] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf -fa 1 - - -▶ [rocm6_4_2] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf - - -▶ [rocm6_4_2] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf -fa 1 - - -▶ [rocm7_rc-rocwmma] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc-rocwmma.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf - - -▶ [rocm7_rc-rocwmma] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf -fa 1 - - -▶ [rocm7_rc] gpt-oss-120b-F16 - → log: results/gpt-oss-120b-F16__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf - - -▶ [rocm7_rc] gpt-oss-120b-F16 __fa1 - → log: results/gpt-oss-120b-F16__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf -fa 1 - - -▶ [rocm7_beta] gpt-oss-120b-F16 - → log: results/gpt-oss-120b-F16__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf - - -▶ [rocm7_beta] gpt-oss-120b-F16 __fa1 - → log: results/gpt-oss-120b-F16__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf -fa 1 - - -▶ [rocm6_4_2-rocwmma] gpt-oss-120b-F16 - → log: results/gpt-oss-120b-F16__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf - - * [rocm6_4_2-rocwmma] gpt-oss-120b-F16 : FAILED - -▶ [rocm6_4_2-rocwmma] gpt-oss-120b-F16 __fa1 - → log: results/gpt-oss-120b-F16__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf -fa 1 - - -▶ [vulkan_radv] gpt-oss-120b-F16 - → log: results/gpt-oss-120b-F16__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf - - -▶ [vulkan_radv] gpt-oss-120b-F16 __fa1 - → log: results/gpt-oss-120b-F16__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf -fa 1 - - -▶ [vulkan_amdvlk] gpt-oss-120b-F16 - → log: results/gpt-oss-120b-F16__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf - - -▶ [vulkan_amdvlk] gpt-oss-120b-F16 __fa1 - → log: results/gpt-oss-120b-F16__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf -fa 1 - - -▶ [rocm6_4_2] gpt-oss-120b-F16 - → log: results/gpt-oss-120b-F16__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf - - -▶ [rocm6_4_2] gpt-oss-120b-F16 __fa1 - → log: results/gpt-oss-120b-F16__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf -fa 1 - - -▶ [rocm7_rc-rocwmma] gpt-oss-120b-F16 - → log: results/gpt-oss-120b-F16__rocm7_rc-rocwmma.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf - - -▶ [rocm7_rc-rocwmma] gpt-oss-120b-F16 __fa1 - → log: results/gpt-oss-120b-F16__rocm7_rc-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf -fa 1 - - -▶ [rocm7_rc] gpt-oss-120b-mxfp4-00001-of-00003 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf - - -▶ [rocm7_rc] gpt-oss-120b-mxfp4-00001-of-00003 __fa1 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf -fa 1 - - -▶ [rocm7_beta] gpt-oss-120b-mxfp4-00001-of-00003 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf - - -▶ [rocm7_beta] gpt-oss-120b-mxfp4-00001-of-00003 __fa1 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf -fa 1 - - -▶ [rocm6_4_2-rocwmma] gpt-oss-120b-mxfp4-00001-of-00003 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf - - -▶ [rocm6_4_2-rocwmma] gpt-oss-120b-mxfp4-00001-of-00003 __fa1 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf -fa 1 - - -▶ [vulkan_radv] gpt-oss-120b-mxfp4-00001-of-00003 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf - - -▶ [vulkan_radv] gpt-oss-120b-mxfp4-00001-of-00003 __fa1 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf -fa 1 - - -▶ [vulkan_amdvlk] gpt-oss-120b-mxfp4-00001-of-00003 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf - - -▶ [vulkan_amdvlk] gpt-oss-120b-mxfp4-00001-of-00003 __fa1 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf -fa 1 - - -▶ [rocm6_4_2] gpt-oss-120b-mxfp4-00001-of-00003 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf - - * [rocm6_4_2] gpt-oss-120b-mxfp4-00001-of-00003 : FAILED - -▶ [rocm6_4_2] gpt-oss-120b-mxfp4-00001-of-00003 __fa1 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf -fa 1 - - -▶ [rocm7_rc-rocwmma] gpt-oss-120b-mxfp4-00001-of-00003 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc-rocwmma.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf - - -▶ [rocm7_rc-rocwmma] gpt-oss-120b-mxfp4-00001-of-00003 __fa1 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf -fa 1 - - -▶ [rocm7_rc] gpt-oss-20b-F32 - → log: results/gpt-oss-20b-F32__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf - - -▶ [rocm7_rc] gpt-oss-20b-F32 __fa1 - → log: results/gpt-oss-20b-F32__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf -fa 1 - - -▶ [rocm7_beta] gpt-oss-20b-F32 - → log: results/gpt-oss-20b-F32__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf - - -▶ [rocm7_beta] gpt-oss-20b-F32 __fa1 - → log: results/gpt-oss-20b-F32__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf -fa 1 - - -▶ [rocm6_4_2-rocwmma] gpt-oss-20b-F32 - → log: results/gpt-oss-20b-F32__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf - - -▶ [rocm6_4_2-rocwmma] gpt-oss-20b-F32 __fa1 - → log: results/gpt-oss-20b-F32__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf -fa 1 - - -▶ [vulkan_radv] gpt-oss-20b-F32 - → log: results/gpt-oss-20b-F32__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf - - -▶ [vulkan_radv] gpt-oss-20b-F32 __fa1 - → log: results/gpt-oss-20b-F32__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf -fa 1 - - -▶ [vulkan_amdvlk] gpt-oss-20b-F32 - → log: results/gpt-oss-20b-F32__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf - - -▶ [vulkan_amdvlk] gpt-oss-20b-F32 __fa1 - → log: results/gpt-oss-20b-F32__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf -fa 1 - - -▶ [rocm6_4_2] gpt-oss-20b-F32 - → log: results/gpt-oss-20b-F32__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf - - -▶ [rocm6_4_2] gpt-oss-20b-F32 __fa1 - → log: results/gpt-oss-20b-F32__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf -fa 1 - - -▶ [rocm7_rc-rocwmma] gpt-oss-20b-F32 - → log: results/gpt-oss-20b-F32__rocm7_rc-rocwmma.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf - - -▶ [rocm7_rc-rocwmma] gpt-oss-20b-F32 __fa1 - → log: results/gpt-oss-20b-F32__rocm7_rc-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf -fa 1 - - -▶ [rocm7_rc] gpt-oss-20b-mxfp4 - → log: results/gpt-oss-20b-mxfp4__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf - - -▶ [rocm7_rc] gpt-oss-20b-mxfp4 __fa1 - → log: results/gpt-oss-20b-mxfp4__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf -fa 1 - - -▶ [rocm7_beta] gpt-oss-20b-mxfp4 - → log: results/gpt-oss-20b-mxfp4__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf - - -▶ [rocm7_beta] gpt-oss-20b-mxfp4 __fa1 - → log: results/gpt-oss-20b-mxfp4__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf -fa 1 - - -▶ [rocm6_4_2-rocwmma] gpt-oss-20b-mxfp4 - → log: results/gpt-oss-20b-mxfp4__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf - - -▶ [rocm6_4_2-rocwmma] gpt-oss-20b-mxfp4 __fa1 - → log: results/gpt-oss-20b-mxfp4__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf -fa 1 - - -▶ [vulkan_radv] gpt-oss-20b-mxfp4 - → log: results/gpt-oss-20b-mxfp4__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf - - -▶ [vulkan_radv] gpt-oss-20b-mxfp4 __fa1 - → log: results/gpt-oss-20b-mxfp4__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf -fa 1 - - -▶ [vulkan_amdvlk] gpt-oss-20b-mxfp4 - → log: results/gpt-oss-20b-mxfp4__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf - - -▶ [vulkan_amdvlk] gpt-oss-20b-mxfp4 __fa1 - → log: results/gpt-oss-20b-mxfp4__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf -fa 1 - - -▶ [rocm6_4_2] gpt-oss-20b-mxfp4 - → log: results/gpt-oss-20b-mxfp4__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf - - -▶ [rocm6_4_2] gpt-oss-20b-mxfp4 __fa1 - → log: results/gpt-oss-20b-mxfp4__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf -fa 1 - - -▶ [rocm7_rc-rocwmma] gpt-oss-20b-mxfp4 - → log: results/gpt-oss-20b-mxfp4__rocm7_rc-rocwmma.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf - - -▶ [rocm7_rc-rocwmma] gpt-oss-20b-mxfp4 __fa1 - → log: results/gpt-oss-20b-mxfp4__rocm7_rc-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf -fa 1 - - -▶ [rocm7_rc] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf - - -▶ [rocm7_rc] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - * [rocm7_rc] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 : FAILED - -▶ [rocm7_beta] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf - - -▶ [rocm7_beta] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - * [rocm7_beta] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 : FAILED - -▶ [rocm6_4_2-rocwmma] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf - - -▶ [rocm6_4_2-rocwmma] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - * [rocm6_4_2-rocwmma] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 : FAILED - -▶ [vulkan_radv] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf - - -▶ [vulkan_radv] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [vulkan_amdvlk] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf - - * [vulkan_amdvlk] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 : FAILED - -▶ [vulkan_amdvlk] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - * [vulkan_amdvlk] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 : FAILED - -▶ [rocm6_4_2] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf - - * [rocm6_4_2] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 : FAILED - -▶ [rocm6_4_2] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - * [rocm6_4_2] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 : FAILED - -▶ [rocm7_rc-rocwmma] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf - - -▶ [rocm7_rc-rocwmma] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [rocm7_rc] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf - - -▶ [rocm7_rc] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - * [rocm7_rc] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 : FAILED - -▶ [rocm7_beta] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf - - -▶ [rocm7_beta] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - * [rocm7_beta] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 : FAILED - -▶ [rocm6_4_2-rocwmma] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf - - * [rocm6_4_2-rocwmma] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 : FAILED - -▶ [rocm6_4_2-rocwmma] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - * [rocm6_4_2-rocwmma] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 : FAILED - -▶ [vulkan_radv] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf - - -▶ [vulkan_radv] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [vulkan_amdvlk] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf - - -▶ [vulkan_amdvlk] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [rocm6_4_2] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf - - -▶ [rocm6_4_2] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [rocm7_rc-rocwmma] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf - - -▶ [rocm7_rc-rocwmma] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [rocm7_rc] llama3.3-70.6B-Q4_K_M - → log: results/llama3.3-70.6B-Q4_K_M__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf - - -▶ [rocm7_rc] llama3.3-70.6B-Q4_K_M __fa1 - → log: results/llama3.3-70.6B-Q4_K_M__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf -fa 1 - - -▶ [rocm7_beta] llama3.3-70.6B-Q4_K_M - → log: results/llama3.3-70.6B-Q4_K_M__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf - - -▶ [rocm7_beta] llama3.3-70.6B-Q4_K_M __fa1 - → log: results/llama3.3-70.6B-Q4_K_M__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf -fa 1 - - -▶ [rocm6_4_2-rocwmma] llama3.3-70.6B-Q4_K_M - → log: results/llama3.3-70.6B-Q4_K_M__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf - - -▶ [rocm6_4_2-rocwmma] llama3.3-70.6B-Q4_K_M __fa1 - → log: results/llama3.3-70.6B-Q4_K_M__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf -fa 1 - - -▶ [vulkan_radv] llama3.3-70.6B-Q4_K_M - → log: results/llama3.3-70.6B-Q4_K_M__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf - - -▶ [vulkan_radv] llama3.3-70.6B-Q4_K_M __fa1 - → log: results/llama3.3-70.6B-Q4_K_M__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf -fa 1 - - -▶ [vulkan_amdvlk] llama3.3-70.6B-Q4_K_M - → log: results/llama3.3-70.6B-Q4_K_M__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf - - -▶ [vulkan_amdvlk] llama3.3-70.6B-Q4_K_M __fa1 - → log: results/llama3.3-70.6B-Q4_K_M__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf -fa 1 - - -▶ [rocm6_4_2] llama3.3-70.6B-Q4_K_M - → log: results/llama3.3-70.6B-Q4_K_M__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf - - -▶ [rocm6_4_2] llama3.3-70.6B-Q4_K_M __fa1 - → log: results/llama3.3-70.6B-Q4_K_M__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf -fa 1 - - -▶ [rocm7_rc-rocwmma] llama3.3-70.6B-Q4_K_M - → log: results/llama3.3-70.6B-Q4_K_M__rocm7_rc-rocwmma.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf - - -▶ [rocm7_rc-rocwmma] llama3.3-70.6B-Q4_K_M __fa1 - → log: results/llama3.3-70.6B-Q4_K_M__rocm7_rc-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf -fa 1 - - -▶ [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf - - -▶ [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - * [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __fa1 : FAILED - -▶ [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf - - -▶ [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf - - * [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 : FAILED - -▶ [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - * [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __fa1 : FAILED - -▶ [vulkan_radv] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf - - -▶ [vulkan_radv] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [vulkan_amdvlk] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf - - -▶ [vulkan_amdvlk] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf - - -▶ [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [rocm7_rc-rocwmma] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf - - * [rocm7_rc-rocwmma] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 : FAILED - -▶ [rocm7_rc-rocwmma] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf - - -▶ [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf -fa 1 - - -▶ [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf - - * [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 : FAILED - -▶ [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf -fa 1 - - * [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 : FAILED - -▶ [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf - - * [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 : FAILED - -▶ [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf -fa 1 - - * [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 : FAILED - -▶ [vulkan_radv] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf - - -▶ [vulkan_radv] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf -fa 1 - - -▶ [vulkan_amdvlk] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf - - -▶ [vulkan_amdvlk] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf -fa 1 - - -▶ [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf - - -▶ [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf -fa 1 - - * [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 : FAILED - -▶ [rocm7_rc-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc-rocwmma.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf - - -▶ [rocm7_rc-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf -fa 1 - - -▶ [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf - - -▶ [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf -fa 1 - - * [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 : FAILED - -▶ [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf - - -▶ [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf -fa 1 - - * [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 : FAILED - -▶ [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf - - * [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 : FAILED - -▶ [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf -fa 1 - - * [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 : FAILED - -▶ [vulkan_radv] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf - - -▶ [vulkan_radv] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf -fa 1 - - -▶ [vulkan_amdvlk] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf - - -▶ [vulkan_amdvlk] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf -fa 1 - - -▶ [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf - - -▶ [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf -fa 1 - - * [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 : FAILED - -▶ [rocm7_rc-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc-rocwmma.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf - - -▶ [rocm7_rc-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf -fa 1 - - -▶ [rocm7_rc] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf - - -▶ [rocm7_rc] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf -fa 1 - - -▶ [rocm7_beta] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf - - * [rocm7_beta] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 : FAILED - -▶ [rocm7_beta] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf -fa 1 - - * [rocm7_beta] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 : FAILED - -▶ [rocm6_4_2-rocwmma] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf - - * [rocm6_4_2-rocwmma] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 : FAILED - -▶ [rocm6_4_2-rocwmma] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf -fa 1 - - * [rocm6_4_2-rocwmma] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 : FAILED - -▶ [vulkan_radv] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf - - -▶ [vulkan_radv] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf -fa 1 - - -▶ [vulkan_amdvlk] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf - - -▶ [vulkan_amdvlk] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf -fa 1 - - -▶ [rocm6_4_2] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf - - -▶ [rocm6_4_2] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf -fa 1 - - -▶ [rocm7_rc-rocwmma] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc-rocwmma.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf - - -▶ [rocm7_rc-rocwmma] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf -fa 1 - - -▶ [rocm7_rc] Qwen3-30B-A3B-BF16-00001-of-00002 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf - - -▶ [rocm7_rc] Qwen3-30B-A3B-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf -fa 1 - - * [rocm7_rc] Qwen3-30B-A3B-BF16-00001-of-00002 __fa1 : FAILED - -▶ [rocm7_beta] Qwen3-30B-A3B-BF16-00001-of-00002 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf - - -▶ [rocm7_beta] Qwen3-30B-A3B-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf -fa 1 - - * [rocm7_beta] Qwen3-30B-A3B-BF16-00001-of-00002 __fa1 : FAILED - -▶ [rocm6_4_2-rocwmma] Qwen3-30B-A3B-BF16-00001-of-00002 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf - - -▶ [rocm6_4_2-rocwmma] Qwen3-30B-A3B-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf -fa 1 - - -▶ [vulkan_radv] Qwen3-30B-A3B-BF16-00001-of-00002 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf - - -▶ [vulkan_radv] Qwen3-30B-A3B-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf -fa 1 - - -▶ [vulkan_amdvlk] Qwen3-30B-A3B-BF16-00001-of-00002 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf - - -▶ [vulkan_amdvlk] Qwen3-30B-A3B-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf -fa 1 - - -▶ [rocm6_4_2] Qwen3-30B-A3B-BF16-00001-of-00002 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf - - -▶ [rocm6_4_2] Qwen3-30B-A3B-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf -fa 1 - - -▶ [rocm7_rc-rocwmma] Qwen3-30B-A3B-BF16-00001-of-00002 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc-rocwmma.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf - - -▶ [rocm7_rc-rocwmma] Qwen3-30B-A3B-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf -fa 1 - - -▶ [rocm7_rc] Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL - → log: results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL.gguf - - -▶ [rocm7_rc] Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL __fa1 - → log: results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL.gguf -fa 1 - - -▶ [rocm7_beta] Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL - → log: results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL.gguf - - -▶ [rocm7_beta] Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL __fa1 - → log: results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL.gguf -fa 1 - - -▶ [rocm6_4_2-rocwmma] Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL - → log: results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL.gguf - - -▶ [rocm6_4_2-rocwmma] Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL __fa1 - → log: results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL.gguf -fa 1 - - -▶ [vulkan_radv] Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL - → log: results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL.gguf - - -▶ [vulkan_radv] Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL __fa1 - → log: results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL.gguf -fa 1 - - -▶ [vulkan_amdvlk] Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL - → log: results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL.gguf - - -▶ [vulkan_amdvlk] Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL __fa1 - → log: results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL.gguf -fa 1 - - -▶ [rocm6_4_2] Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL - → log: results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL.gguf - - -▶ [rocm6_4_2] Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL __fa1 - → log: results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL.gguf -fa 1 - - -▶ [rocm7_rc-rocwmma] Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL - → log: results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc-rocwmma.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL.gguf - - -▶ [rocm7_rc-rocwmma] Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL __fa1 - → log: results/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL.gguf -fa 1 - - -▶ [rocm7_rc] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf - - -▶ [rocm7_rc] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf -fa 1 - - -▶ [rocm7_beta] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf - - -▶ [rocm7_beta] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf -fa 1 - - * [rocm7_beta] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 __fa1 : FAILED - -▶ [rocm6_4_2-rocwmma] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf - - -▶ [rocm6_4_2-rocwmma] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf -fa 1 - - -▶ [vulkan_radv] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf - - -▶ [vulkan_radv] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf -fa 1 - - -▶ [vulkan_amdvlk] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf - - -▶ [vulkan_amdvlk] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf -fa 1 - - -▶ [rocm6_4_2] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf - - -▶ [rocm6_4_2] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf -fa 1 - - * [rocm6_4_2] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 __fa1 : FAILED - -▶ [rocm7_rc-rocwmma] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc-rocwmma.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf - - -▶ [rocm7_rc-rocwmma] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf -fa 1 - diff --git a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2__fa1.log b/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2__fa1.log deleted file mode 100644 index 268535b..0000000 --- a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -Memory access fault by GPU node-1 (Agent handle: 0x834aa90) on address 0x7f10fb96f000. Reason: Page not present or supervisor privilege. -✖ ! [rocm6_4_2] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_beta.log b/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_beta.log deleted file mode 100644 index 52deb8e..0000000 --- a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_beta.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x100d3790) reason :GPU Hang -✖ ! [rocm7_beta] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 failed (exit 134) diff --git a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_beta__fa1.log b/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_beta__fa1.log deleted file mode 100644 index 8039123..0000000 --- a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_beta__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -Memory access fault by GPU node-1 (Agent handle: 0x13829790) on address 0x7fa8ef9a9000. Reason: Page not present or supervisor privilege. -✖ ! [rocm7_beta] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc.log b/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc.log deleted file mode 100644 index fcf0f01..0000000 --- a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 999 | 0 | pp512 | 130.17 ± 0.38 | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | ROCm | 999 | 0 | tg128 | 19.83 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk.log b/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk.log deleted file mode 100644 index 4ef718e..0000000 --- a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | Vulkan | 999 | 0 | pp512 | 200.76 ± 0.32 | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | Vulkan | 999 | 0 | tg128 | 22.78 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log b/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log deleted file mode 100644 index 4bbf6de..0000000 --- a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | Vulkan | 999 | 1 | 0 | pp512 | 201.86 ± 0.27 | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | Vulkan | 999 | 1 | 0 | tg128 | 22.83 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_radv.log b/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_radv.log deleted file mode 100644 index 90347e7..0000000 --- a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_radv.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | Vulkan | 999 | 0 | pp512 | 127.73 ± 0.23 | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | Vulkan | 999 | 0 | tg128 | 22.88 ± 0.02 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_radv__fa1.log b/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_radv__fa1.log deleted file mode 100644 index cf98168..0000000 --- a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_radv__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | Vulkan | 999 | 1 | 0 | pp512 | 132.54 ± 0.34 | -| glm4moe 106B.A12B Q4_K - Medium | 68.01 GiB | 110.47 B | Vulkan | 999 | 1 | 0 | tg128 | 23.31 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2-rocwmma__fa1.log b/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2-rocwmma__fa1.log deleted file mode 100644 index a0c808c..0000000 --- a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2-rocwmma__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x2f508a90) reason :GPU Hang -✖ ! [rocm6_4_2-rocwmma] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_beta.log b/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_beta.log deleted file mode 100644 index 273166e..0000000 --- a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 999 | 0 | pp512 | 120.54 ± 0.30 | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 999 | 0 | tg128 | 15.49 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_beta__fa1.log b/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_beta__fa1.log deleted file mode 100644 index c23fe13..0000000 --- a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_beta__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x2a849790) reason :GPU Hang -✖ ! [rocm7_beta] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc.log b/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc.log deleted file mode 100644 index 5fbf5b3..0000000 --- a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 999 | 0 | pp512 | 124.18 ± 0.48 | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | ROCm | 999 | 0 | tg128 | 15.49 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc__fa1.log b/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc__fa1.log deleted file mode 100644 index 28ae734..0000000 --- a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc__fa1.log +++ /dev/null @@ -1,5 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_amdvlk.log b/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_amdvlk.log deleted file mode 100644 index 4247170..0000000 --- a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_amdvlk.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | Vulkan | 999 | 0 | pp512 | 223.02 ± 0.69 | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | Vulkan | 999 | 0 | tg128 | 16.47 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_amdvlk__fa1.log b/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_amdvlk__fa1.log deleted file mode 100644 index e3bc753..0000000 --- a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_amdvlk__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | Vulkan | 999 | 1 | 0 | pp512 | 224.54 ± 0.65 | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | Vulkan | 999 | 1 | 0 | tg128 | 16.49 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_radv.log b/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_radv.log deleted file mode 100644 index 5f0ace5..0000000 --- a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_radv.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | Vulkan | 999 | 0 | pp512 | 127.36 ± 0.46 | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | Vulkan | 999 | 0 | tg128 | 16.78 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_radv__fa1.log b/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_radv__fa1.log deleted file mode 100644 index 1973a52..0000000 --- a/benchmark/results_08-08-2025/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_radv__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | Vulkan | 999 | 1 | 0 | pp512 | 131.78 ± 0.46 | -| glm4moe 106B.A12B Q6_K | 94.57 GiB | 110.47 B | Vulkan | 999 | 1 | 0 | tg128 | 16.99 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log b/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log deleted file mode 100644 index 135d108..0000000 --- a/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x121f0a90) reason :GPU Hang -✖ ! [rocm6_4_2-rocwmma] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log b/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log deleted file mode 100644 index 29b2095..0000000 --- a/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x17018a90) reason :GPU Hang -✖ ! [rocm6_4_2-rocwmma] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2.log b/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2.log deleted file mode 100644 index 08dae7b..0000000 --- a/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x11442a90) reason :GPU Hang -✖ ! [rocm6_4_2] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_beta.log b/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_beta.log deleted file mode 100644 index e01b520..0000000 --- a/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_beta.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0xa636790) reason :GPU Hang -✖ ! [rocm7_beta] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_beta__fa1.log b/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_beta__fa1.log deleted file mode 100644 index 2f2342b..0000000 --- a/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_beta__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x1417b7b0) reason :GPU Hang -✖ ! [rocm7_beta] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc.log b/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc.log deleted file mode 100644 index c479337..0000000 --- a/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen2 70B Q8_0 | 78.21 GiB | 72.71 B | ROCm | 999 | 0 | pp512 | 33.30 ± 0.04 | -| qwen2 70B Q8_0 | 78.21 GiB | 72.71 B | ROCm | 999 | 0 | tg128 | 2.64 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc__fa1.log b/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc__fa1.log deleted file mode 100644 index 7b0ea20..0000000 --- a/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen2 70B Q8_0 | 78.21 GiB | 72.71 B | ROCm | 999 | 1 | 0 | pp512 | 31.09 ± 0.02 | -| qwen2 70B Q8_0 | 78.21 GiB | 72.71 B | ROCm | 999 | 1 | 0 | tg128 | 2.65 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk.log b/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk.log deleted file mode 100644 index 4581b23..0000000 --- a/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -ggml_vulkan: Device memory allocation of size 2491416576 failed. -ggml_vulkan: Requested buffer size exceeds device memory allocation limit: ErrorOutOfDeviceMemory -main: error: failed to load model '/mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf' -✖ ! [vulkan_amdvlk] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 failed (exit 1) diff --git a/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log b/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log deleted file mode 100644 index 8835330..0000000 --- a/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -ggml_vulkan: Device memory allocation of size 2491416576 failed. -ggml_vulkan: Requested buffer size exceeds device memory allocation limit: ErrorOutOfDeviceMemory -main: error: failed to load model '/mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf' -✖ ! [vulkan_amdvlk] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 failed (exit 1) diff --git a/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_radv.log b/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_radv.log deleted file mode 100644 index c6c72c5..0000000 --- a/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_radv.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen2 70B Q8_0 | 78.21 GiB | 72.71 B | Vulkan | 999 | 0 | pp512 | 78.70 ± 0.20 | -| qwen2 70B Q8_0 | 78.21 GiB | 72.71 B | Vulkan | 999 | 0 | tg128 | 2.66 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_radv__fa1.log b/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_radv__fa1.log deleted file mode 100644 index ea12120..0000000 --- a/benchmark/results_08-08-2025/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_radv__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen2 70B Q8_0 | 78.21 GiB | 72.71 B | Vulkan | 999 | 1 | 0 | pp512 | 81.29 ± 0.14 | -| qwen2 70B Q8_0 | 78.21 GiB | 72.71 B | Vulkan | 999 | 1 | 0 | tg128 | 2.66 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log b/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log deleted file mode 100644 index 3de552f..0000000 --- a/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x1496da90) reason :GPU Hang -✖ ! [rocm6_4_2-rocwmma] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_beta.log b/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_beta.log deleted file mode 100644 index 2acc073..0000000 --- a/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_beta.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0xfeef7b0) reason :GPU Hang -✖ ! [rocm7_beta] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_beta__fa1.log b/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_beta__fa1.log deleted file mode 100644 index 7a57ad3..0000000 --- a/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_beta__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -Memory access fault by GPU node-1 (Agent handle: 0x6d017c0) on address 0x7f967f1a9000. Reason: Page not present or supervisor privilege. -✖ ! [rocm7_beta] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc.log b/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc.log deleted file mode 100644 index b9ba150..0000000 --- a/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc.log +++ /dev/null @@ -1,5 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc__fa1.log b/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc__fa1.log deleted file mode 100644 index c55bab8..0000000 --- a/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc__fa1.log +++ /dev/null @@ -1,5 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk.log b/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk.log deleted file mode 100644 index eb3efec..0000000 --- a/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | Vulkan | 999 | 0 | pp512 | 98.14 ± 0.14 | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | Vulkan | 999 | 0 | tg128 | 2.73 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log b/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log deleted file mode 100644 index 966e109..0000000 --- a/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | Vulkan | 999 | 1 | 0 | pp512 | 99.24 ± 0.16 | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | Vulkan | 999 | 1 | 0 | tg128 | 2.72 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_radv.log b/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_radv.log deleted file mode 100644 index 80c3a0e..0000000 --- a/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_radv.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | Vulkan | 999 | 0 | pp512 | 80.11 ± 0.09 | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | Vulkan | 999 | 0 | tg128 | 2.73 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_radv__fa1.log b/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_radv__fa1.log deleted file mode 100644 index 5826f3e..0000000 --- a/benchmark/results_08-08-2025/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_radv__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | Vulkan | 999 | 1 | 0 | pp512 | 82.90 ± 0.14 | -| llama 70B Q8_0 | 75.65 GiB | 70.55 B | Vulkan | 999 | 1 | 0 | tg128 | 2.73 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2-rocwmma.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2-rocwmma.log deleted file mode 100644 index 40f418b..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2-rocwmma.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x28bb9a90) reason :GPU Hang -✖ ! [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2-rocwmma__fa1.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2-rocwmma__fa1.log deleted file mode 100644 index a94cdd6..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2-rocwmma__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x194fea90) reason :GPU Hang -✖ ! [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_beta.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_beta.log deleted file mode 100644 index 6d3b4ea..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_beta.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x17ad57b0) reason :GPU Hang -✖ ! [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_beta__fa1.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_beta__fa1.log deleted file mode 100644 index 107b01e..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_beta__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -Memory access fault by GPU node-1 (Agent handle: 0x2314b7b0) on address 0x7f38249a9000. Reason: Page not present or supervisor privilege. -✖ ! [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc.log deleted file mode 100644 index ccf7ac1..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 999 | 0 | pp512 | 135.25 ± 0.50 | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | ROCm | 999 | 0 | tg128 | 14.43 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc__fa1.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc__fa1.log deleted file mode 100644 index 8df0b3e..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc__fa1.log +++ /dev/null @@ -1,5 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_amdvlk.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_amdvlk.log deleted file mode 100644 index dc80b9d..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_amdvlk.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | Vulkan | 999 | 0 | pp512 | 243.45 ± 1.29 | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | Vulkan | 999 | 0 | tg128 | 15.29 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_amdvlk__fa1.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_amdvlk__fa1.log deleted file mode 100644 index 08242f2..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_amdvlk__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | Vulkan | 999 | 1 | 0 | pp512 | 247.48 ± 1.28 | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | Vulkan | 999 | 1 | 0 | tg128 | 15.03 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_radv.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_radv.log deleted file mode 100644 index ba7a655..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_radv.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | Vulkan | 999 | 0 | pp512 | 148.25 ± 0.91 | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | Vulkan | 999 | 0 | tg128 | 15.21 ± 0.06 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_radv__fa1.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_radv__fa1.log deleted file mode 100644 index 14f12dd..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_radv__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | Vulkan | 999 | 1 | 0 | pp512 | 149.82 ± 0.83 | -| llama4 17Bx16E (Scout) Q6_K | 82.35 GiB | 107.77 B | Vulkan | 999 | 1 | 0 | tg128 | 15.21 ± 0.04 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2-rocwmma.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2-rocwmma.log deleted file mode 100644 index 2faeaa3..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2-rocwmma.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x9ae6a90) reason :GPU Hang -✖ ! [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2-rocwmma__fa1.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2-rocwmma__fa1.log deleted file mode 100644 index 6ff4745..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2-rocwmma__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x6e9ba90) reason :GPU Hang -✖ ! [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_beta.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_beta.log deleted file mode 100644 index c768b8e..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_beta.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x4081f7b0) reason :GPU Hang -✖ ! [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_beta__fa1.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_beta__fa1.log deleted file mode 100644 index 98c472e..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_beta__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x3c0f27b0) reason :GPU Hang -✖ ! [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc.log deleted file mode 100644 index 9c06e2b..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc.log +++ /dev/null @@ -1,5 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc__fa1.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc__fa1.log deleted file mode 100644 index 3ccfa82..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc__fa1.log +++ /dev/null @@ -1,5 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_amdvlk.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_amdvlk.log deleted file mode 100644 index 3bdeae7..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_amdvlk.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | Vulkan | 999 | 0 | pp512 | 258.18 ± 1.38 | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | Vulkan | 999 | 0 | tg128 | 12.23 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_amdvlk__fa1.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_amdvlk__fa1.log deleted file mode 100644 index 2060565..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_amdvlk__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | Vulkan | 999 | 1 | 0 | pp512 | 260.16 ± 1.44 | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | Vulkan | 999 | 1 | 0 | tg128 | 12.09 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_radv.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_radv.log deleted file mode 100644 index d9b6ebc..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_radv.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | Vulkan | 999 | 0 | pp512 | 168.63 ± 0.81 | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | Vulkan | 999 | 0 | tg128 | 12.26 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_radv__fa1.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_radv__fa1.log deleted file mode 100644 index 579e532..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_radv__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | Vulkan | 999 | 1 | 0 | pp512 | 172.37 ± 0.92 | -| llama4 17Bx16E (Scout) Q8_0 | 106.65 GiB | 107.77 B | Vulkan | 999 | 1 | 0 | tg128 | 12.25 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log deleted file mode 100644 index 070646e..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x1a40fa90) reason :GPU Hang -✖ ! [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log deleted file mode 100644 index 3fa46c3..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x2e0ffa90) reason :GPU Hang -✖ ! [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2__fa1.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2__fa1.log deleted file mode 100644 index 9d0c061..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x3a741a90) reason :GPU Hang -✖ ! [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_beta.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_beta.log deleted file mode 100644 index fb93137..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 999 | 0 | pp512 | 138.90 ± 0.66 | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | ROCm | 999 | 0 | tg128 | 17.62 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc.log deleted file mode 100644 index 2e1a6fc..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc.log +++ /dev/null @@ -1,5 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk.log deleted file mode 100644 index 75ac351..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | Vulkan | 999 | 0 | pp512 | 218.18 ± 0.83 | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | Vulkan | 999 | 0 | tg128 | 20.04 ± 0.02 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log deleted file mode 100644 index a745a31..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | Vulkan | 999 | 1 | 0 | pp512 | 221.15 ± 0.74 | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | Vulkan | 999 | 1 | 0 | tg128 | 19.58 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_radv.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_radv.log deleted file mode 100644 index 4b78701..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_radv.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | Vulkan | 999 | 0 | pp512 | 152.21 ± 0.66 | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | Vulkan | 999 | 0 | tg128 | 19.98 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_radv__fa1.log b/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_radv__fa1.log deleted file mode 100644 index ee535dc..0000000 --- a/benchmark/results_08-08-2025/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_radv__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | Vulkan | 999 | 1 | 0 | pp512 | 155.22 ± 1.09 | -| llama4 17Bx16E (Scout) Q4_K - Medium | 57.73 GiB | 107.77 B | Vulkan | 999 | 1 | 0 | tg128 | 19.93 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2-rocwmma.log b/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2-rocwmma.log deleted file mode 100644 index aa6dfe3..0000000 --- a/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2-rocwmma.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x153dfa90) reason :GPU Hang -✖ ! [rocm6_4_2-rocwmma] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2-rocwmma__fa1.log b/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2-rocwmma__fa1.log deleted file mode 100644 index e2df164..0000000 --- a/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2-rocwmma__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x2bd2ba90) reason :GPU Hang -✖ ! [rocm6_4_2-rocwmma] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_beta.log b/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_beta.log deleted file mode 100644 index b5a6749..0000000 --- a/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_beta.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x513c7b0) reason :GPU Hang -✖ ! [rocm7_beta] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_beta__fa1.log b/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_beta__fa1.log deleted file mode 100644 index 7826050..0000000 --- a/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_beta__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -Memory access fault by GPU node-1 (Agent handle: 0x2567c7c0) on address 0x7ee66236f000. Reason: Page not present or supervisor privilege. -✖ ! [rocm7_beta] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc.log b/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc.log deleted file mode 100644 index dbd9c47..0000000 --- a/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc.log +++ /dev/null @@ -1,5 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc__fa1.log b/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc__fa1.log deleted file mode 100644 index 57b950a..0000000 --- a/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc__fa1.log +++ /dev/null @@ -1,5 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_amdvlk.log b/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_amdvlk.log deleted file mode 100644 index af5c138..0000000 --- a/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_amdvlk.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | Vulkan | 999 | 0 | pp512 | 114.49 ± 0.60 | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | Vulkan | 999 | 0 | tg128 | 15.98 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_amdvlk__fa1.log b/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_amdvlk__fa1.log deleted file mode 100644 index 19e5e37..0000000 --- a/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_amdvlk__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | Vulkan | 999 | 1 | 0 | pp512 | 116.07 ± 0.64 | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | Vulkan | 999 | 1 | 0 | tg128 | 15.84 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_radv.log b/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_radv.log deleted file mode 100644 index 2aefda4..0000000 --- a/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_radv.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | Vulkan | 999 | 0 | pp512 | 64.85 ± 0.38 | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | Vulkan | 999 | 0 | tg128 | 16.58 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_radv__fa1.log b/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_radv__fa1.log deleted file mode 100644 index c0359f0..0000000 --- a/benchmark/results_08-08-2025/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_radv__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | Vulkan | 999 | 1 | 0 | pp512 | 66.76 ± 0.43 | -| qwen3moe 235B.A22B Q3_K - Medium | 96.99 GiB | 235.09 B | Vulkan | 999 | 1 | 0 | tg128 | 16.83 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_beta.log b/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_beta.log deleted file mode 100644 index ea26bd0..0000000 --- a/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | pp512 | 153.49 ± 1.19 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | tg128 | 24.52 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_beta__fa1.log b/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_beta__fa1.log deleted file mode 100644 index bb2103f..0000000 --- a/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_beta__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 1 | 0 | pp512 | 138.49 ± 2.52 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 1 | 0 | tg128 | 24.35 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc.log b/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc.log deleted file mode 100644 index e446a9b..0000000 --- a/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | pp512 | 152.26 ± 2.41 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | tg128 | 24.55 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc__fa1.log b/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc__fa1.log deleted file mode 100644 index d73c640..0000000 --- a/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 1 | 0 | pp512 | 137.52 ± 1.75 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 1 | 0 | tg128 | 24.33 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_amdvlk.log b/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_amdvlk.log deleted file mode 100644 index 1687c7e..0000000 --- a/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_amdvlk.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 0 | pp512 | 107.48 ± 0.16 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 0 | tg128 | 8.04 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_amdvlk__fa1.log b/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_amdvlk__fa1.log deleted file mode 100644 index a9a752b..0000000 --- a/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_amdvlk__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 1 | 0 | pp512 | 107.64 ± 0.13 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 1 | 0 | tg128 | 7.96 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_radv.log b/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_radv.log deleted file mode 100644 index ccca043..0000000 --- a/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_radv.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 0 | pp512 | 85.97 ± 0.12 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 0 | tg128 | 7.38 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_radv__fa1.log b/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_radv__fa1.log deleted file mode 100644 index 48148ef..0000000 --- a/benchmark/results_08-08-2025/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_radv__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 1 | 0 | pp512 | 87.05 ± 0.10 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 1 | 0 | tg128 | 7.40 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_beta.log b/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_beta.log deleted file mode 100644 index a3987ef..0000000 --- a/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 0 | pp512 | 390.07 ± 0.40 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 0 | tg128 | 50.19 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_beta__fa1.log b/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_beta__fa1.log deleted file mode 100644 index a9ca9ef..0000000 --- a/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_beta__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 1 | 0 | pp512 | 300.60 ± 2.31 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 1 | 0 | tg128 | 49.78 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc.log b/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc.log deleted file mode 100644 index 8ff09a7..0000000 --- a/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 0 | pp512 | 388.99 ± 1.86 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 0 | tg128 | 50.31 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc__fa1.log b/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc__fa1.log deleted file mode 100644 index db6f9b0..0000000 --- a/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__rocm7_rc__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 1 | 0 | pp512 | 302.87 ± 0.88 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | ROCm | 999 | 1 | 0 | tg128 | 49.90 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_amdvlk.log b/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_amdvlk.log deleted file mode 100644 index 51b45f0..0000000 --- a/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_amdvlk.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | Vulkan | 999 | 0 | pp512 | 736.95 ± 3.72 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | Vulkan | 999 | 0 | tg128 | 56.89 ± 0.26 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_amdvlk__fa1.log b/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_amdvlk__fa1.log deleted file mode 100644 index 3f2a08e..0000000 --- a/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_amdvlk__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | Vulkan | 999 | 1 | 0 | pp512 | 727.71 ± 2.81 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | Vulkan | 999 | 1 | 0 | tg128 | 53.34 ± 0.31 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_radv.log b/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_radv.log deleted file mode 100644 index 5140ff3..0000000 --- a/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_radv.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | Vulkan | 999 | 0 | pp512 | 395.16 ± 1.55 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | Vulkan | 999 | 0 | tg128 | 58.95 ± 0.45 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_radv__fa1.log b/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_radv__fa1.log deleted file mode 100644 index 6bbc4f2..0000000 --- a/benchmark/results_08-08-2025/Qwen3-30B-A3B-Instruct-2507-UD-Q6_K_XL__vulkan_radv__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | Vulkan | 999 | 1 | 0 | pp512 | 405.61 ± 1.85 | -| qwen3moe 30B.A3B Q6_K | 24.53 GiB | 30.53 B | Vulkan | 999 | 1 | 0 | tg128 | 58.06 ± 0.28 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2-rocwmma.log b/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2-rocwmma.log deleted file mode 100644 index 6625574..0000000 --- a/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2-rocwmma.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | pp512 | 150.50 ± 1.69 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | tg128 | 24.55 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2.log b/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2.log deleted file mode 100644 index 222959d..0000000 --- a/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | pp512 | 150.34 ± 1.74 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | tg128 | 24.14 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_beta.log b/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_beta.log deleted file mode 100644 index cc48f94..0000000 --- a/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | pp512 | 146.55 ± 1.77 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | tg128 | 24.54 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_beta__fa1.log b/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_beta__fa1.log deleted file mode 100644 index 285bed2..0000000 --- a/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_beta__fa1.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -Memory access fault by GPU node-1 (Agent handle: 0x2bd8a7b0) on address 0x7fe0b0d6f000. Reason: Page not present or supervisor privilege. -✖ ! [rocm7_beta] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc.log b/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc.log deleted file mode 100644 index 29fa537..0000000 --- a/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | pp512 | 145.91 ± 1.76 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | ROCm | 999 | 0 | tg128 | 24.57 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc__fa1.log b/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc__fa1.log deleted file mode 100644 index 1416318..0000000 --- a/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc__fa1.log +++ /dev/null @@ -1,5 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_amdvlk.log b/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_amdvlk.log deleted file mode 100644 index 65ecb3e..0000000 --- a/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_amdvlk.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 0 | pp512 | 106.99 ± 0.10 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 0 | tg128 | 8.03 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_amdvlk__fa1.log b/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_amdvlk__fa1.log deleted file mode 100644 index 2b69233..0000000 --- a/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_amdvlk__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 1 | 0 | pp512 | 107.10 ± 0.08 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 1 | 0 | tg128 | 7.98 ± 0.02 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_radv.log b/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_radv.log deleted file mode 100644 index 3a2d167..0000000 --- a/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_radv.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 0 | pp512 | 85.50 ± 0.06 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 0 | tg128 | 7.42 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_radv__fa1.log b/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_radv__fa1.log deleted file mode 100644 index 9132fa2..0000000 --- a/benchmark/results_08-08-2025/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_radv__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 1 | 0 | pp512 | 86.52 ± 0.06 | -| qwen3moe 30B.A3B BF16 | 56.89 GiB | 30.53 B | Vulkan | 999 | 1 | 0 | tg128 | 7.40 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__rocm7_beta.log b/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__rocm7_beta.log deleted file mode 100644 index 6493650..0000000 --- a/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 0 | pp512 | 222.67 ± 0.37 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 0 | tg128 | 13.88 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__rocm7_beta__fa1.log b/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__rocm7_beta__fa1.log deleted file mode 100644 index a535d64..0000000 --- a/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__rocm7_beta__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 1 | 0 | pp512 | 203.12 ± 0.35 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 1 | 0 | tg128 | 13.60 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc.log b/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc.log deleted file mode 100644 index f1ec100..0000000 --- a/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 0 | pp512 | 222.49 ± 0.29 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 0 | tg128 | 13.86 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc__fa1.log b/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc__fa1.log deleted file mode 100644 index f4493e0..0000000 --- a/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 1 | 0 | pp512 | 201.47 ± 0.21 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | ROCm | 999 | 1 | 0 | tg128 | 13.61 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__vulkan_amdvlk.log b/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__vulkan_amdvlk.log deleted file mode 100644 index 5ac352f..0000000 --- a/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__vulkan_amdvlk.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | Vulkan | 999 | 0 | pp512 | 676.94 ± 0.85 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | Vulkan | 999 | 0 | tg128 | 13.99 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__vulkan_amdvlk__fa1.log b/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__vulkan_amdvlk__fa1.log deleted file mode 100644 index b3193bd..0000000 --- a/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__vulkan_amdvlk__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | Vulkan | 999 | 1 | 0 | pp512 | 371.17 ± 0.24 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | Vulkan | 999 | 1 | 0 | tg128 | 12.30 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__vulkan_radv.log b/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__vulkan_radv.log deleted file mode 100644 index b620676..0000000 --- a/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__vulkan_radv.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | Vulkan | 999 | 0 | pp512 | 503.27 ± 1.09 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | Vulkan | 999 | 0 | tg128 | 13.76 ± 0.02 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__vulkan_radv__fa1.log b/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__vulkan_radv__fa1.log deleted file mode 100644 index 5e9431a..0000000 --- a/benchmark/results_08-08-2025/gemma-3-12b-it-UD-Q8_K_XL__vulkan_radv__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | Vulkan | 999 | 1 | 0 | pp512 | 495.99 ± 2.36 | -| gemma3 12B Q8_0 | 13.40 GiB | 11.77 B | Vulkan | 999 | 1 | 0 | tg128 | 13.61 ± 0.03 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2.log b/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2.log deleted file mode 100644 index c646996..0000000 --- a/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2.log +++ /dev/null @@ -1,6 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -HW Exception by GPU node-1 (Agent handle: 0x10c4a90) reason :GPU Hang -✖ ! [rocm6_4_2] gemma-3-27b-it-BF16-00001-of-00002 failed (exit 134) diff --git a/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__rocm7_beta__fa1.log b/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__rocm7_beta__fa1.log deleted file mode 100644 index 18449f7..0000000 --- a/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__rocm7_beta__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 1 | 0 | pp512 | 83.61 ± 0.31 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 1 | 0 | tg128 | 4.04 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc.log b/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc.log deleted file mode 100644 index 9e9f25c..0000000 --- a/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 0 | pp512 | 55.68 ± 0.47 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 0 | tg128 | 3.11 ± 0.98 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc__fa1.log b/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc__fa1.log deleted file mode 100644 index f7ce012..0000000 --- a/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 1 | 0 | pp512 | 83.08 ± 0.42 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | ROCm | 999 | 1 | 0 | tg128 | 4.04 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__vulkan_amdvlk.log b/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__vulkan_amdvlk.log deleted file mode 100644 index d74242d..0000000 --- a/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__vulkan_amdvlk.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -ggml_vulkan: Device memory allocation of size 2819260416 failed. -ggml_vulkan: Requested buffer size exceeds device memory allocation limit: ErrorOutOfDeviceMemory -main: error: failed to load model '/mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf' -✖ ! [vulkan_amdvlk] gemma-3-27b-it-BF16-00001-of-00002 failed (exit 1) diff --git a/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__vulkan_amdvlk__fa1.log b/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__vulkan_amdvlk__fa1.log deleted file mode 100644 index a667917..0000000 --- a/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__vulkan_amdvlk__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -ggml_vulkan: Device memory allocation of size 2819260416 failed. -ggml_vulkan: Requested buffer size exceeds device memory allocation limit: ErrorOutOfDeviceMemory -main: error: failed to load model '/mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf' -✖ ! [vulkan_amdvlk] gemma-3-27b-it-BF16-00001-of-00002 __fa1 failed (exit 1) diff --git a/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__vulkan_radv.log b/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__vulkan_radv.log deleted file mode 100644 index 0c3a407..0000000 --- a/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__vulkan_radv.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | Vulkan | 999 | 0 | pp512 | 135.58 ± 0.45 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | Vulkan | 999 | 0 | tg128 | 4.00 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__vulkan_radv__fa1.log b/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__vulkan_radv__fa1.log deleted file mode 100644 index f2077af..0000000 --- a/benchmark/results_08-08-2025/gemma-3-27b-it-BF16-00001-of-00002__vulkan_radv__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | Vulkan | 999 | 1 | 0 | pp512 | 138.61 ± 0.55 | -| gemma3 27B BF16 | 50.31 GiB | 27.01 B | Vulkan | 999 | 1 | 0 | tg128 | 4.00 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__rocm7_beta.log b/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__rocm7_beta.log deleted file mode 100644 index ea76e52..0000000 --- a/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 0 | pp512 | 732.13 ± 1.42 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 0 | tg128 | 76.23 ± 0.03 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__rocm7_beta__fa1.log b/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__rocm7_beta__fa1.log deleted file mode 100644 index 76e9619..0000000 --- a/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__rocm7_beta__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 1 | 0 | pp512 | 652.29 ± 0.45 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 1 | 0 | tg128 | 69.62 ± 0.02 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__rocm7_rc.log b/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__rocm7_rc.log deleted file mode 100644 index ce94640..0000000 --- a/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__rocm7_rc.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 0 | pp512 | 730.59 ± 1.69 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | ROCm | 999 | 0 | tg128 | 76.01 ± 0.03 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__vulkan_amdvlk.log b/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__vulkan_amdvlk.log deleted file mode 100644 index b707702..0000000 --- a/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__vulkan_amdvlk.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | Vulkan | 999 | 0 | pp512 | 1614.72 ± 4.91 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | Vulkan | 999 | 0 | tg128 | 84.00 ± 0.23 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__vulkan_amdvlk__fa1.log b/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__vulkan_amdvlk__fa1.log deleted file mode 100644 index 6055d96..0000000 --- a/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__vulkan_amdvlk__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | Vulkan | 999 | 1 | 0 | pp512 | 942.34 ± 1.76 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | Vulkan | 999 | 1 | 0 | tg128 | 57.70 ± 0.22 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__vulkan_radv.log b/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__vulkan_radv.log deleted file mode 100644 index 5a56858..0000000 --- a/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__vulkan_radv.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | Vulkan | 999 | 0 | pp512 | 1527.75 ± 3.86 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | Vulkan | 999 | 0 | tg128 | 85.54 ± 0.99 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__vulkan_radv__fa1.log b/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__vulkan_radv__fa1.log deleted file mode 100644 index ab5608b..0000000 --- a/benchmark/results_08-08-2025/gemma-3-4b-it-Q3_K_S__vulkan_radv__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | Vulkan | 999 | 1 | 0 | pp512 | 1489.57 ± 4.71 | -| gemma3 4B Q3_K - Small | 1.80 GiB | 3.88 B | Vulkan | 999 | 1 | 0 | tg128 | 80.63 ± 0.22 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-120b-F16__rocm7_beta.log b/benchmark/results_08-08-2025/gpt-oss-120b-F16__rocm7_beta.log deleted file mode 100644 index 4e2c281..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-120b-F16__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 0 | pp512 | 357.38 ± 0.76 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 0 | tg128 | 33.62 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-120b-F16__rocm7_rc.log b/benchmark/results_08-08-2025/gpt-oss-120b-F16__rocm7_rc.log deleted file mode 100644 index 63dd9d9..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-120b-F16__rocm7_rc.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 0 | pp512 | 356.67 ± 0.74 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 0 | tg128 | 33.68 ± 0.02 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-120b-F16__rocm7_rc__fa1.log b/benchmark/results_08-08-2025/gpt-oss-120b-F16__rocm7_rc__fa1.log deleted file mode 100644 index 8096c36..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-120b-F16__rocm7_rc__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 1 | 0 | pp512 | 247.49 ± 0.65 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | ROCm | 999 | 1 | 0 | tg128 | 33.07 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-120b-F16__vulkan_amdvlk.log b/benchmark/results_08-08-2025/gpt-oss-120b-F16__vulkan_amdvlk.log deleted file mode 100644 index 755a9cf..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-120b-F16__vulkan_amdvlk.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | Vulkan | 999 | 0 | pp512 | 448.17 ± 1.37 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | Vulkan | 999 | 0 | tg128 | 33.39 ± 0.03 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-120b-F16__vulkan_amdvlk__fa1.log b/benchmark/results_08-08-2025/gpt-oss-120b-F16__vulkan_amdvlk__fa1.log deleted file mode 100644 index 152170f..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-120b-F16__vulkan_amdvlk__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | Vulkan | 999 | 1 | 0 | pp512 | 498.69 ± 2.19 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | Vulkan | 999 | 1 | 0 | tg128 | 33.06 ± 0.03 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-120b-F16__vulkan_radv.log b/benchmark/results_08-08-2025/gpt-oss-120b-F16__vulkan_radv.log deleted file mode 100644 index 5ab95e4..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-120b-F16__vulkan_radv.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | Vulkan | 999 | 0 | pp512 | 229.59 ± 0.74 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | Vulkan | 999 | 0 | tg128 | 33.08 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-120b-F16__vulkan_radv__fa1.log b/benchmark/results_08-08-2025/gpt-oss-120b-F16__vulkan_radv__fa1.log deleted file mode 100644 index 9d830ae..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-120b-F16__vulkan_radv__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | Vulkan | 999 | 1 | 0 | pp512 | 243.40 ± 0.99 | -| gpt-oss ?B F16 | 60.87 GiB | 116.83 B | Vulkan | 999 | 1 | 0 | tg128 | 33.07 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_beta.log b/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_beta.log deleted file mode 100644 index 3892e39..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 0 | pp512 | 354.82 ± 1.02 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 0 | tg128 | 45.00 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_beta__fa1.log b/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_beta__fa1.log deleted file mode 100644 index 69476e2..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_beta__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 1 | 0 | pp512 | 248.22 ± 0.50 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 1 | 0 | tg128 | 44.05 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc.log b/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc.log deleted file mode 100644 index 3a57ced..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 0 | pp512 | 353.20 ± 0.59 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | ROCm | 999 | 0 | tg128 | 45.15 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc__fa1.log b/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc__fa1.log deleted file mode 100644 index 93e7fca..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc__fa1.log +++ /dev/null @@ -1,5 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -✖ ! [rocm7_rc] gpt-oss-120b-mxfp4-00001-of-00003 __fa1 failed (exit 134) diff --git a/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_amdvlk.log b/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_amdvlk.log deleted file mode 100644 index d229658..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_amdvlk.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | Vulkan | 999 | 0 | pp512 | 486.90 ± 2.23 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | Vulkan | 999 | 0 | tg128 | 48.08 ± 0.03 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_amdvlk__fa1.log b/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_amdvlk__fa1.log deleted file mode 100644 index b556c96..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_amdvlk__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | Vulkan | 999 | 1 | 0 | pp512 | 546.41 ± 2.88 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | Vulkan | 999 | 1 | 0 | tg128 | 47.25 ± 0.02 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_radv.log b/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_radv.log deleted file mode 100644 index 802c652..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_radv.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | Vulkan | 999 | 0 | pp512 | 239.72 ± 1.23 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | Vulkan | 999 | 0 | tg128 | 49.01 ± 0.06 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_radv__fa1.log b/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_radv__fa1.log deleted file mode 100644 index 6b8a8c4..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_radv__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | Vulkan | 999 | 1 | 0 | pp512 | 255.17 ± 1.65 | -| gpt-oss ?B MXFP4 MoE | 59.02 GiB | 116.83 B | Vulkan | 999 | 1 | 0 | tg128 | 48.93 ± 0.02 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-20b-F32__rocm7_beta.log b/benchmark/results_08-08-2025/gpt-oss-20b-F32__rocm7_beta.log deleted file mode 100644 index d76138e..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-20b-F32__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 0 | pp512 | 322.43 ± 2.59 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 0 | tg128 | 26.89 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-20b-F32__rocm7_beta__fa1.log b/benchmark/results_08-08-2025/gpt-oss-20b-F32__rocm7_beta__fa1.log deleted file mode 100644 index 6dd4954..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-20b-F32__rocm7_beta__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 1 | 0 | pp512 | 254.08 ± 3.99 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 1 | 0 | tg128 | 26.62 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-20b-F32__rocm7_rc.log b/benchmark/results_08-08-2025/gpt-oss-20b-F32__rocm7_rc.log deleted file mode 100644 index 67b820b..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-20b-F32__rocm7_rc.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 0 | pp512 | 319.36 ± 3.07 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 0 | tg128 | 26.88 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-20b-F32__rocm7_rc__fa1.log b/benchmark/results_08-08-2025/gpt-oss-20b-F32__rocm7_rc__fa1.log deleted file mode 100644 index e07a069..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-20b-F32__rocm7_rc__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 1 | 0 | pp512 | 254.87 ± 2.27 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | ROCm | 999 | 1 | 0 | tg128 | 26.62 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-20b-F32__vulkan_amdvlk.log b/benchmark/results_08-08-2025/gpt-oss-20b-F32__vulkan_amdvlk.log deleted file mode 100644 index 52536d1..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-20b-F32__vulkan_amdvlk.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | Vulkan | 999 | 0 | pp512 | 369.69 ± 1.79 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | Vulkan | 999 | 0 | tg128 | 8.59 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-20b-F32__vulkan_amdvlk__fa1.log b/benchmark/results_08-08-2025/gpt-oss-20b-F32__vulkan_amdvlk__fa1.log deleted file mode 100644 index 974e845..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-20b-F32__vulkan_amdvlk__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | Vulkan | 999 | 1 | 0 | pp512 | 389.86 ± 2.13 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | Vulkan | 999 | 1 | 0 | tg128 | 8.58 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-20b-F32__vulkan_radv.log b/benchmark/results_08-08-2025/gpt-oss-20b-F32__vulkan_radv.log deleted file mode 100644 index 7decf08..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-20b-F32__vulkan_radv.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | Vulkan | 999 | 0 | pp512 | 319.09 ± 1.46 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | Vulkan | 999 | 0 | tg128 | 7.79 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-20b-F32__vulkan_radv__fa1.log b/benchmark/results_08-08-2025/gpt-oss-20b-F32__vulkan_radv__fa1.log deleted file mode 100644 index a9ce691..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-20b-F32__vulkan_radv__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | Vulkan | 999 | 1 | 0 | pp512 | 335.15 ± 1.80 | -| gpt-oss ?B BF16 | 38.97 GiB | 20.91 B | Vulkan | 999 | 1 | 0 | tg128 | 7.79 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__rocm7_beta.log b/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__rocm7_beta.log deleted file mode 100644 index 441cec1..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 0 | pp512 | 583.52 ± 2.76 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 0 | tg128 | 64.39 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__rocm7_beta__fa1.log b/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__rocm7_beta__fa1.log deleted file mode 100644 index e5f1e99..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__rocm7_beta__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 1 | 0 | pp512 | 396.75 ± 0.60 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 1 | 0 | tg128 | 62.98 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__rocm7_rc.log b/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__rocm7_rc.log deleted file mode 100644 index 97fab79..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__rocm7_rc.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 0 | pp512 | 581.83 ± 1.10 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 0 | tg128 | 64.50 ± 0.02 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__rocm7_rc__fa1.log b/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__rocm7_rc__fa1.log deleted file mode 100644 index 3e34f41..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__rocm7_rc__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 1 | 0 | pp512 | 394.87 ± 0.73 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 999 | 1 | 0 | tg128 | 63.06 ± 0.01 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__vulkan_amdvlk.log b/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__vulkan_amdvlk.log deleted file mode 100644 index 2d4b788..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__vulkan_amdvlk.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan | 999 | 0 | pp512 | 1205.02 ± 7.18 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan | 999 | 0 | tg128 | 68.84 ± 0.04 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__vulkan_amdvlk__fa1.log b/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__vulkan_amdvlk__fa1.log deleted file mode 100644 index 9a5c4c5..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__vulkan_amdvlk__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan | 999 | 1 | 0 | pp512 | 1472.56 ± 14.39 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan | 999 | 1 | 0 | tg128 | 67.78 ± 0.18 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__vulkan_radv.log b/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__vulkan_radv.log deleted file mode 100644 index f400d0f..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__vulkan_radv.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan | 999 | 0 | pp512 | 648.85 ± 6.28 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan | 999 | 0 | tg128 | 69.88 ± 0.04 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__vulkan_radv__fa1.log b/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__vulkan_radv__fa1.log deleted file mode 100644 index 1959c7e..0000000 --- a/benchmark/results_08-08-2025/gpt-oss-20b-mxfp4__vulkan_radv__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan | 999 | 1 | 0 | pp512 | 728.38 ± 8.17 | -| gpt-oss ?B MXFP4 MoE | 11.27 GiB | 20.91 B | Vulkan | 999 | 1 | 0 | tg128 | 69.80 ± 0.05 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__rocm6_4_2-rocwmma.log b/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__rocm6_4_2-rocwmma.log deleted file mode 100644 index e9da9da..0000000 --- a/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__rocm6_4_2-rocwmma.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 0 | pp512 | 33.47 ± 0.04 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 0 | tg128 | 4.62 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__rocm6_4_2-rocwmma__fa1.log b/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__rocm6_4_2-rocwmma__fa1.log deleted file mode 100644 index 0388774..0000000 --- a/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__rocm6_4_2-rocwmma__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 1 | 0 | pp512 | 34.51 ± 0.02 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 1 | 0 | tg128 | 4.61 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__rocm6_4_2.log b/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__rocm6_4_2.log deleted file mode 100644 index 01f32df..0000000 --- a/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__rocm6_4_2.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 0 | pp512 | 33.79 ± 0.03 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 0 | tg128 | 4.52 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__rocm6_4_2__fa1.log b/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__rocm6_4_2__fa1.log deleted file mode 100644 index f9ae86b..0000000 --- a/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__rocm6_4_2__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 1 | 0 | pp512 | 31.67 ± 0.04 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 1 | 0 | tg128 | 4.63 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__rocm7_beta.log b/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__rocm7_beta.log deleted file mode 100644 index f6959d1..0000000 --- a/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__rocm7_beta.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 0 | pp512 | 33.88 ± 0.02 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 0 | tg128 | 4.61 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__rocm7_beta__fa1.log b/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__rocm7_beta__fa1.log deleted file mode 100644 index 2869c45..0000000 --- a/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__rocm7_beta__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 1 | 0 | pp512 | 31.67 ± 0.02 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 1 | 0 | tg128 | 4.63 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__rocm7_rc.log b/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__rocm7_rc.log deleted file mode 100644 index 6bd1b01..0000000 --- a/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__rocm7_rc.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 0 | pp512 | 33.91 ± 0.03 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 0 | tg128 | 4.61 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__rocm7_rc__fa1.log b/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__rocm7_rc__fa1.log deleted file mode 100644 index 77dd920..0000000 --- a/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__rocm7_rc__fa1.log +++ /dev/null @@ -1,10 +0,0 @@ -ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no -ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no -ggml_cuda_init: found 1 ROCm devices: - Device 0: AMD Radeon Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32 -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 1 | 0 | pp512 | 31.66 ± 0.04 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | ROCm | 999 | 1 | 0 | tg128 | 4.63 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__vulkan_amdvlk.log b/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__vulkan_amdvlk.log deleted file mode 100644 index bc604f8..0000000 --- a/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__vulkan_amdvlk.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | Vulkan | 999 | 0 | pp512 | 72.75 ± 0.02 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | Vulkan | 999 | 0 | tg128 | 5.03 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__vulkan_amdvlk__fa1.log b/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__vulkan_amdvlk__fa1.log deleted file mode 100644 index 7ac44cb..0000000 --- a/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__vulkan_amdvlk__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (AMD open-source driver) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 32768 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | Vulkan | 999 | 1 | 0 | pp512 | 73.57 ± 0.02 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | Vulkan | 999 | 1 | 0 | tg128 | 5.00 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__vulkan_radv.log b/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__vulkan_radv.log deleted file mode 100644 index 4cc5212..0000000 --- a/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__vulkan_radv.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | Vulkan | 999 | 0 | pp512 | 78.99 ± 0.18 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | Vulkan | 999 | 0 | tg128 | 5.00 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__vulkan_radv__fa1.log b/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__vulkan_radv__fa1.log deleted file mode 100644 index 869327e..0000000 --- a/benchmark/results_08-08-2025/llama3.3-70.6B-Q4_K_M__vulkan_radv__fa1.log +++ /dev/null @@ -1,8 +0,0 @@ -ggml_vulkan: Found 1 Vulkan devices: -ggml_vulkan: 0 = Radeon 8060S Graphics (RADV GFX1151) (radv) | uma: 1 | fp16: 1 | bf16: 0 | warp size: 64 | shared memory: 65536 | int dot: 1 | matrix cores: KHR_coopmat -| model | size | params | backend | ngl | fa | mmap | test | t/s | -| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | Vulkan | 999 | 1 | 0 | pp512 | 80.92 ± 0.05 | -| llama 70B Q4_K - Medium | 39.59 GiB | 70.55 B | Vulkan | 999 | 1 | 0 | tg128 | 4.99 ± 0.00 | - -build: cd6983d5 (6119) diff --git a/benchmark/results_08-08-2025/run_benchmarks.log b/benchmark/results_08-08-2025/run_benchmarks.log deleted file mode 100644 index 073dde1..0000000 --- a/benchmark/results_08-08-2025/run_benchmarks.log +++ /dev/null @@ -1,1153 +0,0 @@ -Found 18 model(s) to bench: - • /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf - • /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf - • /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf - • /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf - • /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf - • /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf - • /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf - • /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf - • /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf - • /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf - • /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf - • /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf - • /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf - • /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf - • /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf - • /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf - • /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf - • /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf - - -▶ [rocm7_rc] gemma-3-27b-it-BF16-00001-of-00002 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf - - -▶ [rocm7_rc] gemma-3-27b-it-BF16-00001-of-00002 __fa1 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf -fa 1 - - -▶ [rocm7_beta] gemma-3-27b-it-BF16-00001-of-00002 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf - - -▶ [rocm7_beta] gemma-3-27b-it-BF16-00001-of-00002 __fa1 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf -fa 1 - - -▶ [rocm6_4_2-rocwmma] gemma-3-27b-it-BF16-00001-of-00002 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf - - -▶ [rocm6_4_2-rocwmma] gemma-3-27b-it-BF16-00001-of-00002 __fa1 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf -fa 1 - - -▶ [vulkan_radv] gemma-3-27b-it-BF16-00001-of-00002 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf - - -▶ [vulkan_radv] gemma-3-27b-it-BF16-00001-of-00002 __fa1 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf -fa 1 - - -▶ [vulkan_amdvlk] gemma-3-27b-it-BF16-00001-of-00002 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf - - * [vulkan_amdvlk] gemma-3-27b-it-BF16-00001-of-00002 : FAILED - -▶ [vulkan_amdvlk] gemma-3-27b-it-BF16-00001-of-00002 __fa1 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf -fa 1 - - * [vulkan_amdvlk] gemma-3-27b-it-BF16-00001-of-00002 __fa1 : FAILED - -▶ [rocm6_4_2] gemma-3-27b-it-BF16-00001-of-00002 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf - - * [rocm6_4_2] gemma-3-27b-it-BF16-00001-of-00002 : FAILED - -▶ [rocm6_4_2] gemma-3-27b-it-BF16-00001-of-00002 __fa1 - → log: results/gemma-3-27b-it-BF16-00001-of-00002__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/BF16/gemma-3-27b-it-BF16-00001-of-00002.gguf -fa 1 - - -▶ [rocm7_rc] gemma-3-12b-it-UD-Q8_K_XL - → log: results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf - - -▶ [rocm7_rc] gemma-3-12b-it-UD-Q8_K_XL __fa1 - → log: results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf -fa 1 - - -▶ [rocm7_beta] gemma-3-12b-it-UD-Q8_K_XL - → log: results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf - - -▶ [rocm7_beta] gemma-3-12b-it-UD-Q8_K_XL __fa1 - → log: results/gemma-3-12b-it-UD-Q8_K_XL__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf -fa 1 - - -▶ [rocm6_4_2-rocwmma] gemma-3-12b-it-UD-Q8_K_XL - → log: results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf - - -▶ [rocm6_4_2-rocwmma] gemma-3-12b-it-UD-Q8_K_XL __fa1 - → log: results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf -fa 1 - - -▶ [vulkan_radv] gemma-3-12b-it-UD-Q8_K_XL - → log: results/gemma-3-12b-it-UD-Q8_K_XL__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf - - -▶ [vulkan_radv] gemma-3-12b-it-UD-Q8_K_XL __fa1 - → log: results/gemma-3-12b-it-UD-Q8_K_XL__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf -fa 1 - - -▶ [vulkan_amdvlk] gemma-3-12b-it-UD-Q8_K_XL - → log: results/gemma-3-12b-it-UD-Q8_K_XL__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf - - -▶ [vulkan_amdvlk] gemma-3-12b-it-UD-Q8_K_XL __fa1 - → log: results/gemma-3-12b-it-UD-Q8_K_XL__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf -fa 1 - - -▶ [rocm6_4_2] gemma-3-12b-it-UD-Q8_K_XL - → log: results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf - - -▶ [rocm6_4_2] gemma-3-12b-it-UD-Q8_K_XL __fa1 - → log: results/gemma-3-12b-it-UD-Q8_K_XL__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-12b-it-UD-Q8_K_XL.gguf -fa 1 - - -▶ [rocm7_rc] gemma-3-4b-it-Q3_K_S - → log: results/gemma-3-4b-it-Q3_K_S__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf - - -▶ [rocm7_rc] gemma-3-4b-it-Q3_K_S __fa1 - → log: results/gemma-3-4b-it-Q3_K_S__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf -fa 1 - - -▶ [rocm7_beta] gemma-3-4b-it-Q3_K_S - → log: results/gemma-3-4b-it-Q3_K_S__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf - - -▶ [rocm7_beta] gemma-3-4b-it-Q3_K_S __fa1 - → log: results/gemma-3-4b-it-Q3_K_S__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf -fa 1 - - -▶ [rocm6_4_2-rocwmma] gemma-3-4b-it-Q3_K_S - → log: results/gemma-3-4b-it-Q3_K_S__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf - - -▶ [rocm6_4_2-rocwmma] gemma-3-4b-it-Q3_K_S __fa1 - → log: results/gemma-3-4b-it-Q3_K_S__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf -fa 1 - - -▶ [vulkan_radv] gemma-3-4b-it-Q3_K_S - → log: results/gemma-3-4b-it-Q3_K_S__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf - - -▶ [vulkan_radv] gemma-3-4b-it-Q3_K_S __fa1 - → log: results/gemma-3-4b-it-Q3_K_S__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf -fa 1 - - -▶ [vulkan_amdvlk] gemma-3-4b-it-Q3_K_S - → log: results/gemma-3-4b-it-Q3_K_S__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf - - -▶ [vulkan_amdvlk] gemma-3-4b-it-Q3_K_S __fa1 - → log: results/gemma-3-4b-it-Q3_K_S__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf -fa 1 - - -▶ [rocm6_4_2] gemma-3-4b-it-Q3_K_S - → log: results/gemma-3-4b-it-Q3_K_S__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf - - -▶ [rocm6_4_2] gemma-3-4b-it-Q3_K_S __fa1 - → log: results/gemma-3-4b-it-Q3_K_S__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gemma-3/gemma-3-4b-it-Q3_K_S.gguf -fa 1 - - -▶ [rocm7_rc] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf - - -▶ [rocm7_rc] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [rocm7_beta] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf - - * [rocm7_beta] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 : FAILED - -▶ [rocm7_beta] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - * [rocm7_beta] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __fa1 : FAILED - -▶ [rocm6_4_2-rocwmma] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf - - * [rocm6_4_2-rocwmma] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 : FAILED - -▶ [rocm6_4_2-rocwmma] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - * [rocm6_4_2-rocwmma] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __fa1 : FAILED - -▶ [vulkan_radv] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf - - -▶ [vulkan_radv] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [vulkan_amdvlk] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf - - -▶ [vulkan_amdvlk] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [rocm6_4_2] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf - - -▶ [rocm6_4_2] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q4_K_XL/GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - * [rocm6_4_2] GLM-4.5-Air-UD-Q4_K_XL-00001-of-00002 __fa1 : FAILED - -▶ [rocm7_rc] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf - - -▶ [rocm7_rc] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf -fa 1 - - * [rocm7_rc] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 : FAILED - -▶ [rocm7_beta] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf - - -▶ [rocm7_beta] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf -fa 1 - - * [rocm7_beta] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 : FAILED - -▶ [rocm6_4_2-rocwmma] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf - - -▶ [rocm6_4_2-rocwmma] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf -fa 1 - - * [rocm6_4_2-rocwmma] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 : FAILED - -▶ [vulkan_radv] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf - - -▶ [vulkan_radv] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf -fa 1 - - -▶ [vulkan_amdvlk] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf - - -▶ [vulkan_amdvlk] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf -fa 1 - - -▶ [rocm6_4_2] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf - - -▶ [rocm6_4_2] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 - → log: results/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/GLM-4.5-Air/UD-Q6_K_XL/GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003.gguf -fa 1 - - * [rocm6_4_2] GLM-4.5-Air-UD-Q6_K_XL-00001-of-00003 __fa1 : FAILED - -▶ [rocm7_rc] gpt-oss-120b-F16 - → log: results/gpt-oss-120b-F16__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf - - -▶ [rocm7_rc] gpt-oss-120b-F16 __fa1 - → log: results/gpt-oss-120b-F16__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf -fa 1 - - -▶ [rocm7_beta] gpt-oss-120b-F16 - → log: results/gpt-oss-120b-F16__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf - - -▶ [rocm7_beta] gpt-oss-120b-F16 __fa1 - → log: results/gpt-oss-120b-F16__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf -fa 1 - - -▶ [rocm6_4_2-rocwmma] gpt-oss-120b-F16 - → log: results/gpt-oss-120b-F16__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf - - -▶ [rocm6_4_2-rocwmma] gpt-oss-120b-F16 __fa1 - → log: results/gpt-oss-120b-F16__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf -fa 1 - - -▶ [vulkan_radv] gpt-oss-120b-F16 - → log: results/gpt-oss-120b-F16__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf - - -▶ [vulkan_radv] gpt-oss-120b-F16 __fa1 - → log: results/gpt-oss-120b-F16__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf -fa 1 - - -▶ [vulkan_amdvlk] gpt-oss-120b-F16 - → log: results/gpt-oss-120b-F16__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf - - -▶ [vulkan_amdvlk] gpt-oss-120b-F16 __fa1 - → log: results/gpt-oss-120b-F16__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf -fa 1 - - -▶ [rocm6_4_2] gpt-oss-120b-F16 - → log: results/gpt-oss-120b-F16__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf - - -▶ [rocm6_4_2] gpt-oss-120b-F16 __fa1 - → log: results/gpt-oss-120b-F16__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-F16.gguf -fa 1 - - -▶ [rocm7_rc] gpt-oss-120b-mxfp4-00001-of-00003 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf - - -▶ [rocm7_rc] gpt-oss-120b-mxfp4-00001-of-00003 __fa1 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf -fa 1 - - * [rocm7_rc] gpt-oss-120b-mxfp4-00001-of-00003 __fa1 : FAILED - -▶ [rocm7_beta] gpt-oss-120b-mxfp4-00001-of-00003 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf - - -▶ [rocm7_beta] gpt-oss-120b-mxfp4-00001-of-00003 __fa1 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf -fa 1 - - -▶ [rocm6_4_2-rocwmma] gpt-oss-120b-mxfp4-00001-of-00003 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf - - -▶ [rocm6_4_2-rocwmma] gpt-oss-120b-mxfp4-00001-of-00003 __fa1 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf -fa 1 - - -▶ [vulkan_radv] gpt-oss-120b-mxfp4-00001-of-00003 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf - - -▶ [vulkan_radv] gpt-oss-120b-mxfp4-00001-of-00003 __fa1 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf -fa 1 - - -▶ [vulkan_amdvlk] gpt-oss-120b-mxfp4-00001-of-00003 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf - - -▶ [vulkan_amdvlk] gpt-oss-120b-mxfp4-00001-of-00003 __fa1 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf -fa 1 - - -▶ [rocm6_4_2] gpt-oss-120b-mxfp4-00001-of-00003 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf - - -▶ [rocm6_4_2] gpt-oss-120b-mxfp4-00001-of-00003 __fa1 - → log: results/gpt-oss-120b-mxfp4-00001-of-00003__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-120b/gpt-oss-120b-mxfp4-00001-of-00003.gguf -fa 1 - - -▶ [rocm7_rc] gpt-oss-20b-F32 - → log: results/gpt-oss-20b-F32__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf - - -▶ [rocm7_rc] gpt-oss-20b-F32 __fa1 - → log: results/gpt-oss-20b-F32__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf -fa 1 - - -▶ [rocm7_beta] gpt-oss-20b-F32 - → log: results/gpt-oss-20b-F32__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf - - -▶ [rocm7_beta] gpt-oss-20b-F32 __fa1 - → log: results/gpt-oss-20b-F32__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf -fa 1 - - -▶ [rocm6_4_2-rocwmma] gpt-oss-20b-F32 - → log: results/gpt-oss-20b-F32__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf - - -▶ [rocm6_4_2-rocwmma] gpt-oss-20b-F32 __fa1 - → log: results/gpt-oss-20b-F32__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf -fa 1 - - -▶ [vulkan_radv] gpt-oss-20b-F32 - → log: results/gpt-oss-20b-F32__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf - - -▶ [vulkan_radv] gpt-oss-20b-F32 __fa1 - → log: results/gpt-oss-20b-F32__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf -fa 1 - - -▶ [vulkan_amdvlk] gpt-oss-20b-F32 - → log: results/gpt-oss-20b-F32__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf - - -▶ [vulkan_amdvlk] gpt-oss-20b-F32 __fa1 - → log: results/gpt-oss-20b-F32__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf -fa 1 - - -▶ [rocm6_4_2] gpt-oss-20b-F32 - → log: results/gpt-oss-20b-F32__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf - - -▶ [rocm6_4_2] gpt-oss-20b-F32 __fa1 - → log: results/gpt-oss-20b-F32__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-F32.gguf -fa 1 - - -▶ [rocm7_rc] gpt-oss-20b-mxfp4 - → log: results/gpt-oss-20b-mxfp4__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf - - -▶ [rocm7_rc] gpt-oss-20b-mxfp4 __fa1 - → log: results/gpt-oss-20b-mxfp4__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf -fa 1 - - -▶ [rocm7_beta] gpt-oss-20b-mxfp4 - → log: results/gpt-oss-20b-mxfp4__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf - - -▶ [rocm7_beta] gpt-oss-20b-mxfp4 __fa1 - → log: results/gpt-oss-20b-mxfp4__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf -fa 1 - - -▶ [rocm6_4_2-rocwmma] gpt-oss-20b-mxfp4 - → log: results/gpt-oss-20b-mxfp4__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf - - -▶ [rocm6_4_2-rocwmma] gpt-oss-20b-mxfp4 __fa1 - → log: results/gpt-oss-20b-mxfp4__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf -fa 1 - - -▶ [vulkan_radv] gpt-oss-20b-mxfp4 - → log: results/gpt-oss-20b-mxfp4__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf - - -▶ [vulkan_radv] gpt-oss-20b-mxfp4 __fa1 - → log: results/gpt-oss-20b-mxfp4__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf -fa 1 - - -▶ [vulkan_amdvlk] gpt-oss-20b-mxfp4 - → log: results/gpt-oss-20b-mxfp4__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf - - -▶ [vulkan_amdvlk] gpt-oss-20b-mxfp4 __fa1 - → log: results/gpt-oss-20b-mxfp4__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf -fa 1 - - -▶ [rocm6_4_2] gpt-oss-20b-mxfp4 - → log: results/gpt-oss-20b-mxfp4__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf - - -▶ [rocm6_4_2] gpt-oss-20b-mxfp4 __fa1 - → log: results/gpt-oss-20b-mxfp4__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/gpt-oss-20b/gpt-oss-20b-mxfp4.gguf -fa 1 - - -▶ [rocm7_rc] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf - - -▶ [rocm7_rc] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [rocm7_beta] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf - - * [rocm7_beta] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 : FAILED - -▶ [rocm7_beta] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - * [rocm7_beta] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 : FAILED - -▶ [rocm6_4_2-rocwmma] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf - - * [rocm6_4_2-rocwmma] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 : FAILED - -▶ [rocm6_4_2-rocwmma] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - * [rocm6_4_2-rocwmma] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 : FAILED - -▶ [vulkan_radv] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf - - -▶ [vulkan_radv] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [vulkan_amdvlk] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf - - * [vulkan_amdvlk] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 : FAILED - -▶ [vulkan_amdvlk] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - * [vulkan_amdvlk] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 : FAILED - -▶ [rocm6_4_2] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf - - * [rocm6_4_2] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 : FAILED - -▶ [rocm6_4_2] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/kimi-dev-72B-Q8_K_XL/UD-Q8_K_XL/Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - * [rocm6_4_2] Kimi-Dev-72B-UD-Q8_K_XL-00001-of-00002 __fa1 : FAILED - -▶ [rocm7_rc] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf - - * [rocm7_rc] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 : FAILED - -▶ [rocm7_rc] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - * [rocm7_rc] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 : FAILED - -▶ [rocm7_beta] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf - - * [rocm7_beta] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 : FAILED - -▶ [rocm7_beta] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - * [rocm7_beta] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 : FAILED - -▶ [rocm6_4_2-rocwmma] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf - - * [rocm6_4_2-rocwmma] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 : FAILED - -▶ [rocm6_4_2-rocwmma] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - * [rocm6_4_2-rocwmma] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 : FAILED - -▶ [vulkan_radv] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf - - -▶ [vulkan_radv] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [vulkan_amdvlk] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf - - -▶ [vulkan_amdvlk] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [rocm6_4_2] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf - - -▶ [rocm6_4_2] Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002 __fa1 - → log: results/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-70B-Instruct/UD-Q8_K_XL/Llama-3.3-70B-Instruct-UD-Q8_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [rocm7_rc] llama3.3-70.6B-Q4_K_M - → log: results/llama3.3-70.6B-Q4_K_M__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf - - -▶ [rocm7_rc] llama3.3-70.6B-Q4_K_M __fa1 - → log: results/llama3.3-70.6B-Q4_K_M__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf -fa 1 - - -▶ [rocm7_beta] llama3.3-70.6B-Q4_K_M - → log: results/llama3.3-70.6B-Q4_K_M__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf - - -▶ [rocm7_beta] llama3.3-70.6B-Q4_K_M __fa1 - → log: results/llama3.3-70.6B-Q4_K_M__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf -fa 1 - - -▶ [rocm6_4_2-rocwmma] llama3.3-70.6B-Q4_K_M - → log: results/llama3.3-70.6B-Q4_K_M__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf - - -▶ [rocm6_4_2-rocwmma] llama3.3-70.6B-Q4_K_M __fa1 - → log: results/llama3.3-70.6B-Q4_K_M__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf -fa 1 - - -▶ [vulkan_radv] llama3.3-70.6B-Q4_K_M - → log: results/llama3.3-70.6B-Q4_K_M__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf - - -▶ [vulkan_radv] llama3.3-70.6B-Q4_K_M __fa1 - → log: results/llama3.3-70.6B-Q4_K_M__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf -fa 1 - - -▶ [vulkan_amdvlk] llama3.3-70.6B-Q4_K_M - → log: results/llama3.3-70.6B-Q4_K_M__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf - - -▶ [vulkan_amdvlk] llama3.3-70.6B-Q4_K_M __fa1 - → log: results/llama3.3-70.6B-Q4_K_M__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf -fa 1 - - -▶ [rocm6_4_2] llama3.3-70.6B-Q4_K_M - → log: results/llama3.3-70.6B-Q4_K_M__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf - - -▶ [rocm6_4_2] llama3.3-70.6B-Q4_K_M __fa1 - → log: results/llama3.3-70.6B-Q4_K_M__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-3.3-Q4_K_M/llama3.3-70.6B-Q4_K_M.gguf -fa 1 - - -▶ [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf - - * [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 : FAILED - -▶ [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf - - -▶ [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf - - * [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 : FAILED - -▶ [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - * [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __fa1 : FAILED - -▶ [vulkan_radv] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf - - -▶ [vulkan_radv] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [vulkan_amdvlk] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf - - -▶ [vulkan_amdvlk] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - -▶ [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf - - -▶ [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q4_K_XL/Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002.gguf -fa 1 - - * [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-UD-Q4_K_XL-00001-of-00002 __fa1 : FAILED - -▶ [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf - - -▶ [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf -fa 1 - - * [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 : FAILED - -▶ [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf - - * [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 : FAILED - -▶ [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf -fa 1 - - * [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 : FAILED - -▶ [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf - - * [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 : FAILED - -▶ [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf -fa 1 - - * [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 : FAILED - -▶ [vulkan_radv] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf - - -▶ [vulkan_radv] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf -fa 1 - - -▶ [vulkan_amdvlk] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf - - -▶ [vulkan_amdvlk] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf -fa 1 - - -▶ [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf - - -▶ [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q6_K/Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002.gguf -fa 1 - - * [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-Q6_K-00001-of-00002 __fa1 : FAILED - -▶ [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf - - * [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 : FAILED - -▶ [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf -fa 1 - - * [rocm7_rc] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 : FAILED - -▶ [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf - - * [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 : FAILED - -▶ [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf -fa 1 - - * [rocm7_beta] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 : FAILED - -▶ [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf - - * [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 : FAILED - -▶ [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf -fa 1 - - * [rocm6_4_2-rocwmma] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 : FAILED - -▶ [vulkan_radv] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf - - -▶ [vulkan_radv] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf -fa 1 - - -▶ [vulkan_amdvlk] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf - - -▶ [vulkan_amdvlk] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf -fa 1 - - -▶ [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf - - -▶ [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 - → log: results/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/llama-4-scout-17b-16e/Q8_0/Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003.gguf -fa 1 - - * [rocm6_4_2] Llama-4-Scout-17B-16E-Instruct-Q8_0-00001-of-00003 __fa1 : FAILED - -▶ [rocm7_rc] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf - - * [rocm7_rc] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 : FAILED - -▶ [rocm7_rc] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf -fa 1 - - * [rocm7_rc] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 : FAILED - -▶ [rocm7_beta] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf - - * [rocm7_beta] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 : FAILED - -▶ [rocm7_beta] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf -fa 1 - - * [rocm7_beta] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 : FAILED - -▶ [rocm6_4_2-rocwmma] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf - - * [rocm6_4_2-rocwmma] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 : FAILED - -▶ [rocm6_4_2-rocwmma] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf -fa 1 - - * [rocm6_4_2-rocwmma] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 : FAILED - -▶ [vulkan_radv] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf - - -▶ [vulkan_radv] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf -fa 1 - - -▶ [vulkan_amdvlk] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf - - -▶ [vulkan_amdvlk] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf -fa 1 - - -▶ [rocm6_4_2] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf - - -▶ [rocm6_4_2] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 - → log: results/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-235B-Q3_K-XL/UD-Q3_K_XL/Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003.gguf -fa 1 - - * [rocm6_4_2] Qwen3-235B-A22B-Instruct-2507-UD-Q3_K_XL-00001-of-00003 __fa1 : FAILED - -▶ [rocm7_rc] Qwen3-30B-A3B-BF16-00001-of-00002 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf - - -▶ [rocm7_rc] Qwen3-30B-A3B-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf -fa 1 - - -▶ [rocm7_beta] Qwen3-30B-A3B-BF16-00001-of-00002 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf - - -▶ [rocm7_beta] Qwen3-30B-A3B-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf -fa 1 - - -▶ [rocm6_4_2-rocwmma] Qwen3-30B-A3B-BF16-00001-of-00002 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf - - -▶ [rocm6_4_2-rocwmma] Qwen3-30B-A3B-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf -fa 1 - - -▶ [vulkan_radv] Qwen3-30B-A3B-BF16-00001-of-00002 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf - - -▶ [vulkan_radv] Qwen3-30B-A3B-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf -fa 1 - - -▶ [vulkan_amdvlk] Qwen3-30B-A3B-BF16-00001-of-00002 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf - - -▶ [vulkan_amdvlk] Qwen3-30B-A3B-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf -fa 1 - - -▶ [rocm6_4_2] Qwen3-30B-A3B-BF16-00001-of-00002 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf - - -▶ [rocm6_4_2] Qwen3-30B-A3B-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-30B-A3B-BF16-00001-of-00002__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen-3-30B-A3B/BF16/Qwen3-30B-A3B-BF16-00001-of-00002.gguf -fa 1 - - -▶ [rocm7_rc] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf - - -▶ [rocm7_rc] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_rc__fa1.log - → cmd: toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf -fa 1 - - * [rocm7_rc] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 __fa1 : FAILED - -▶ [rocm7_beta] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_beta.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf - - -▶ [rocm7_beta] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm7_beta__fa1.log - → cmd: toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf -fa 1 - - * [rocm7_beta] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 __fa1 : FAILED - -▶ [rocm6_4_2-rocwmma] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2-rocwmma.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf - - -▶ [rocm6_4_2-rocwmma] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2-rocwmma__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf -fa 1 - - -▶ [vulkan_radv] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_radv.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf - - -▶ [vulkan_radv] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_radv__fa1.log - → cmd: toolbox run -c llama-vulkan-radv -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf -fa 1 - - -▶ [vulkan_amdvlk] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_amdvlk.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf - - -▶ [vulkan_amdvlk] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__vulkan_amdvlk__fa1.log - → cmd: toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf -fa 1 - - -▶ [rocm6_4_2] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf - - -▶ [rocm6_4_2] Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002 __fa1 - → log: results/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002__rocm6_4_2__fa1.log - → cmd: toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench -ngl 999 -mmp 0 -m /mnt/models/qwen3-coder-30B-A3B/BF16/Qwen3-Coder-30B-A3B-Instruct-BF16-00001-of-00002.gguf -fa 1 - diff --git a/benchmark/run_benchmarks.sh b/benchmark/run_benchmarks.sh index b4d15e7..151e1d1 100755 --- a/benchmark/run_benchmarks.sh +++ b/benchmark/run_benchmarks.sh @@ -28,7 +28,8 @@ echo declare -A CMDS=( [rocm6_4_2]="toolbox run -c llama-rocm-6.4.2 -- /usr/local/bin/llama-bench" [rocm6_4_2-rocwmma]="toolbox run -c llama-rocm-6.4.2-rocwmma -- /usr/local/bin/llama-bench" - [rocm7_beta]="toolbox run -c llama-rocm-7beta -- /usr/local/bin/llama-bench" + [rocm6_4_3]="toolbox run -c llama-rocm-6.4.3 -- /usr/local/bin/llama-bench" + [rocm6_4_3-rocwmma]="toolbox run -c llama-rocm-6.4.3-rocwmma -- /usr/local/bin/llama-bench" [rocm7_rc]="toolbox run -c llama-rocm-7rc -- /usr/local/bin/llama-bench" [rocm7_rc-rocwmma]="toolbox run -c llama-rocm-7rc-rocwmma -- /usr/local/bin/llama-bench" [vulkan_amdvlk]="toolbox run -c llama-vulkan-amdvlk -- /usr/sbin/llama-bench" @@ -41,8 +42,8 @@ for MODEL_PATH in "${MODEL_PATHS[@]}"; do for ENV in "${!CMDS[@]}"; do CMD="${CMDS[$ENV]}" - # For ROCm 7 envs, run default + HIPBLASLT=0 variants; others: default only - if [[ "$ENV" == rocm7_* ]]; then + # For ROCm 6.4.3 and 7 envs, run default + HIPBLASLT=0 variants; others: default only + if [[ "$ENV" == rocm7_* || "$ENV" == rocm6_4_3* ]]; then HBLT_MODES=( default off ) else HBLT_MODES=( default ) diff --git a/docs/index.html b/docs/index.html index b81fa7e..d8963c8 100644 --- a/docs/index.html +++ b/docs/index.html @@ -363,6 +363,7 @@ Repo: kyuz0/amd-strix-halo-toolboxes
+Platform: Framework Desktop, 128GB Unified RAM (accelerator-performance tuned profile)
Loading meta…
@@ -491,7 +492,7 @@