enable new model uts to xpu and fix some failures on xpu #41386
Conversation
Signed-off-by: Yao, Matrix <[email protected]>
Signed-off-by: Yao, Matrix <[email protected]>
Signed-off-by: Yao, Matrix <[email protected]>
ydshieh
left a comment
ydshieh
left a comment
There was a problem hiding this comment.
Thank you @yao-matrix ! 3 nit comments 🙏
|
Also FYI: I merged a PR #40727, and we can run tests like PATCH_TESTING_METHODS_TO_COLLECT_OUTPUTS=yes HF_HOME=/mnt/cache HF_HUB_READ_TOKEN=hf_XXX RUN_SLOW=1 python3 -m pytest -v .... It will produce an output file named The main advantage is that: if a test has several assert statements and fails at the first one, it won't stop but continue. Hope you find it useful in the future. |
|
Ping me once the nit comments are addressed , thanks. |
|
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
Signed-off-by: Yao, Matrix <[email protected]>
|
@ydshieh , done, pls help review again, thx very much. |
|
[For maintainers] Suggested jobs to run (before merge) run-slow: cohere2_vision, ernie4_5, internvl, llava, llava_next_video, phi3, qwen2_5_omni, qwen3, whisper, xglm |
…#41386) * enable new model uts to xpu and fix some failures on xpu Signed-off-by: Yao, Matrix <[email protected]> * add more Signed-off-by: Yao, Matrix <[email protected]> * fix style Signed-off-by: Yao, Matrix <[email protected]> * Update test_modeling_internvl.py * Update test_modeling_llava.py * Update test_modeling_qwen2_5_omni.py * Update test_modeling_llava_next_video.py * Update test_modeling_qwen3.py * Update test_modeling_whisper.py * Update test_modeling_whisper.py * Update test_modeling_llava.py * Update test_modeling_llava.py * Update test_modeling_qwen2_5_omni.py * fix style Signed-off-by: Yao, Matrix <[email protected]> --------- Signed-off-by: Yao, Matrix <[email protected]>
4 participants
covered modeling:
@SunMarc @ydshieh , pls help review, thx very much.