github/tinygrad
1
1
Fork 0
mirror of https://github.com/tinygrad/tinygrad.git synced 2026-01-21 04:47:56 -05:00
Code Issues Packages Projects Releases Wiki Activity

FUSE_ARANGE=1 in bert bench (#10263)

Browse Source 
still fails, something multi related maybe

Co-authored-by: qazal <77887910+Qazalin@users.noreply.github.com>
This commit is contained in:
chenyu
2025-05-13 09:12:19 -04:00
committed by GitHub
parent a2d6b0afe0
commit ad5cb2717d
1 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
.github/workflows/benchmark.yml vendored
View File

@@ -321,7 +321,7 @@ jobs:
run: NV=1 CAPTURE_PROCESS_REPLAY=0 DEFAULT_FLOAT=HALF BENCHMARK=10 BS=1536 GPUS=6 MODEL=resnet python3 examples/mlperf/model_train.py | tee train_resnet.txt
- name: Run 10 MLPerf Bert training steps (6 gpu)
# TODO: remove BERT_LAYERS once scheduler is fast
run: NV=1 CAPTURE_PROCESS_REPLAY=0 DEFAULT_FLOAT=HALF BENCHMARK=10 BS=66 GPUS=6 BERT_LAYERS=2 MODEL=bert python3 examples/mlperf/model_train.py | tee train_bert.txt
run: NV=1 CAPTURE_PROCESS_REPLAY=0 DEFAULT_FLOAT=HALF BENCHMARK=10 BS=66 GPUS=6 BERT_LAYERS=2 FUSE_ARANGE=1 FUSE_ARANGE_UINT=0 MODEL=bert python3 examples/mlperf/model_train.py | tee train_bert.txt
- uses: actions/upload-artifact@v4
with:
name: Speed (NVIDIA Training)
@@ -517,7 +517,7 @@ jobs:
run: AMD=1 CAPTURE_PROCESS_REPLAY=0 DEFAULT_FLOAT=HALF BENCHMARK=10 BS=1536 GPUS=6 MODEL=resnet python3 examples/mlperf/model_train.py | tee train_resnet.txt
- name: Run 10 MLPerf Bert training steps (6 gpu)
# TODO: remove BERT_LAYERS once scheduler is fast
run: AMD=1 CAPTURE_PROCESS_REPLAY=0 DEFAULT_FLOAT=HALF BENCHMARK=10 BS=66 GPUS=6 BERT_LAYERS=2 MODEL=bert python3 examples/mlperf/model_train.py | tee train_bert.txt
run: AMD=1 CAPTURE_PROCESS_REPLAY=0 DEFAULT_FLOAT=HALF BENCHMARK=10 BS=66 GPUS=6 BERT_LAYERS=2 FUSE_ARANGE=1 FUSE_ARANGE_UINT=0 MODEL=bert python3 examples/mlperf/model_train.py | tee train_bert.txt
- uses: actions/upload-artifact@v4
with:
name: Speed (AMD Training)