Updated benchmarks

This commit is contained in:
Donato Capitella
2025-09-17 10:41:14 +01:00
parent f4506ca711
commit 006aaa64e1
408 changed files with 14136 additions and 9467 deletions
@@ -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 |
| ------------------------------ | ---------: | ---------: | ---------- | --: | ---: | --------------: | -------------------: |
| 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: 1fe00296 (6182)
HW Exception by GPU node-1 (Agent handle: 0x37f5d310) reason :GPU Hang
✖ ! [rocm6_4_3] gpt-oss-120b-F16 __hblt0 failed (exit 134)