New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[Tracking Bug] List of disabled workloads. #125

Open

mariecwhite opened this issue Aug 23, 2023 · 1 comment

Collaborator

mariecwhite commented Aug 23, 2023

This issue lists the workloads that we have disabled due to regressions in compilers or frameworks that we do not maintain.

Collaborator Author

mariecwhite commented Aug 23, 2023 •

edited

Loading

CUDA

XLA Compiler-Level:

BERT_LARGE_FP16_JAX_384XI32_BATCH{1, 16, 24, 32, 48, 64, 512, 1024, 1280}: [XLA-HLO:GPU] Type errors on BERT_LARGE_FP16_JAX_* models #117
T5_LARGE_{FP32, BF16, FP16}_JAX_512XI32_BATCH512: null.
T5_LARGE_FP32_TF_512XI32_BATCH512: null.

XLA Framework-Level:

T5_LARGE_FP32_TF_512XI32_BATCH512: null.
BERT_LARGE_FP32_JAX_384XI32_BATCH{1024, 1280}: 1 hour timeout.
T5_LARGE_FP32_JAX_512XI32_BATCH{64, 512}: resource exhausted.

Inductor Framework-Level:

RESNET50_FP32_PT_3X224X224XF32_BATCH2048: CUDA out of memory.
BERT_LARGE_FP32_PT_384XI32_BATCH{64, 512}: CUDA out of memory.
BERT_LARGE_FP32_PT_384XI32_BATCH{1024, 1280}: CUDA error: an illegal memory access.
BERT_LARGE_FP16_PT_384XI32_BATCH{16, 24, 32, 48, 64}: [PT-Inductor:GPU] Models failed with quantile() input tensor must be either float or double dtype #62

CPU

XLA Framework-Level:

EFFICIENTNETB7_FP32_TF_600X600X3XF32_BATCH128: 1 hour timeout.

Inductor Framework-Level:

BERT_LARGE_FP32_PT_384XI32_BATCH{32, 64}: null

pzread mentioned this issue

Automate artifact generations #131

Merged

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment