search

intel / intel-xpu-backend-for-triton

OpenAI Triton backend for Intel® GPUs
MIT License
126 stars 36 forks source link

Subset of E2E models for accuracy tests #2246

Open pbchekin opened 5 days ago

pbchekin commented 5 days ago

We currently have a subset of models to test E2E performance. We need similar subset for E2E accuracy tests. If the E2E accuracy workflow started with the subset, it should check that all models in the subset passed the accuracy check.

Periodic E2E jobs (E2E accuracy, Try latest PyTorch) need to use the subset mode to catch regressions early.

vlad-penkin commented 5 days ago

@pbchekin this ticket partially duplicates

Let's align scope between two tickets.