julian/amd-strix-halo-toolboxes
1
0
Fork 0
Code Issues Pull Requests 2 Actions Packages Projects Releases Wiki Activity

update benchmarks

Browse Source
This commit is contained in:
Donato Capitella
2026-04-11 11:18:45 +01:00
parent a821bcb91d
commit 7aa6e6dea9
514 changed files with 7863 additions and 13522 deletions
Download Patch File Download Diff File Expand all files Collapse all files
benchmark/results/gpt-oss-20b-mxfp4__rocm7-nightlies__fa1.log
+3 -3
View File
@@ -2,7 +2,7 @@ ggml_cuda_init: found 1 ROCm devices (Total VRAM: 126976 MiB):
Device 0: Radeon 8060S Graphics, gfx1151 (0x1151), VMM: no, Wave Size: 32, VRAM: 126976 MiB
| model | size | params | backend | ngl | fa | mmap | test | t/s |
| ------------------------------ | ---------: | ---------: | ---------- | --: | -: | ---: | --------------: | -------------------: |
| gpt-oss 20B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 1668.50 ± 13.61 |
| gpt-oss 20B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 72.68 ± 0.10 |
| gpt-oss 20B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | pp512 | 1635.73 ± 10.16 |
| gpt-oss 20B MXFP4 MoE | 11.27 GiB | 20.91 B | ROCm | 99 | 1 | 0 | tg128 | 72.47 ± 0.03 |
build: 2405d59cb (8577)
build: 3f8752b55 (8743)