Skip to content

[xpu] extend runs_on test markers to include xpu#10445

Open
singhalshubham03 wants to merge 2 commits into
hiyouga:mainfrom
singhalshubham03:xpu-accelerate-configs-and-test-support
Open

[xpu] extend runs_on test markers to include xpu#10445
singhalshubham03 wants to merge 2 commits into
hiyouga:mainfrom
singhalshubham03:xpu-accelerate-configs-and-test-support

Conversation

@singhalshubham03

@singhalshubham03 singhalshubham03 commented Apr 29, 2026

Copy link
Copy Markdown

Summary

Add "xpu" to @pytest.mark.runs_on across all data-processor, data, e2e, and eval test suites so tests are picked up on Intel XPU CI targets.

Files changed (14):

  • tests/data/processor/test_feedback.py
  • tests/data/processor/test_pairwise.py
  • tests/data/processor/test_processor_utils.py
  • tests/data/processor/test_supervised.py
  • tests/data/processor/test_unsupervised.py
  • tests/data/test_collator.py
  • tests/data/test_converter.py
  • tests/data/test_formatter.py
  • tests/data/test_loader.py
  • tests/data/test_mm_plugin.py
  • tests/data/test_template.py
  • tests/e2e/test_chat.py
  • tests/e2e/test_train.py
  • tests/eval/test_eval_template.py

Test plan

  • Confirm XPU-marked tests are picked up: pytest tests/ -m "runs_on" --device xpu
  • Verify existing CPU/MPS tests are unaffected

🤖 Generated with Claude Code

Add "xpu" to @pytest.mark.runs_on across data-processor, data, e2e, and
eval test suites so tests are picked up on Intel XPU CI targets.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
@singhalshubham03 singhalshubham03 force-pushed the xpu-accelerate-configs-and-test-support branch from 8f37b58 to c926777 Compare April 29, 2026 04:53
@singhalshubham03 singhalshubham03 changed the title [xpu] Add Intel XPU accelerate configs, Llama-3.3-70B training example, and XPU test markers [xpu] extend runs_on test markers to include xpu Apr 29, 2026

@gemini-code-assist gemini-code-assist Bot left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Code Review

This pull request introduces support for Intel XPU devices, specifically targeting fine-tuning of Llama-3.3-70B on 4x Intel Arc B60 GPUs. It adds various Accelerate and training configurations, a comprehensive README, and a launch script with optimized environment variables for XPU. Additionally, the test suite has been updated to include XPU in the supported execution environments. Feedback suggests improving the portability of the launch script by avoiding hardcoded virtual environment paths and making the WANDB integration configurable.

I am having trouble creating individual review comments. Click here to see my feedback.

examples/train_llama33_70b_4xpu/train.sh (19)

medium

Hardcoding the virtual environment path 'venv_main_2.11' limits the portability of this script. Consider using an environment variable or a relative path that is more flexible.

examples/train_llama33_70b_4xpu/train.sh (44)

medium

Disabling WANDB via an environment variable in the script is a good practice for reproducibility, but consider making this configurable via an environment variable to allow users to enable it if needed.

@singhalshubham03

Copy link
Copy Markdown
Author

@hiyouga can you please review this PR

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

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants