Move src/MaxText/distillation to src/maxtext/trainers/post_train/distillation

[SurbhiJainUSC]

Description

• Move src/MaxText/distillation to src/maxtext/trainers/post_train/distillation as part of repo restructuring work as planned here: https://github.com/AI-Hypercomputer/maxtext/blob/main/RESTRUCTURE.md.
• Created shim file in src/MaxText/distillation with a deprecation warning to ensure backward compatibility.

Tests

Distillation old command that works:

python3 -m MaxText.distillation.train_distill \
src/MaxText/configs/distillation.yml \
run_name=$RUN_NAME \
base_output_directory=$BASE_OUTPUT_DIRECTORY \
checkpoint_period=5 \
hf_access_token=$HF_TOKEN \
steps=10 \
save_checkpoint_on_completion=True \
teacher_overrides.load_parameters_path=gs://maxtext-model-checkpoints/llama3.1-8b/2025-01-17-04-13/scanned/0/items

This old command will work but will also print a warning: 'src.MaxText.distillation.train_distill' is deprecated; use 'src.maxtext.trainers.post_train.distillation.train_distill' instead..

Distillation new command that works:

python3 -m maxtext.trainers.post_train.distillation.train_distill \
src/MaxText/configs/distillation.yml \
run_name=$RUN_NAME \
base_output_directory=$BASE_OUTPUT_DIRECTORY \
checkpoint_period=5 \
hf_access_token=$HF_TOKEN \
steps=10 \
save_checkpoint_on_completion=True \
teacher_overrides.load_parameters_path=gs://maxtext-model-checkpoints/llama3.1-8b/2025-01-17-04-13/scanned/0/items

Checklist

Before submitting this PR, please make sure (put X in square brackets):

• I have performed a self-review of my code. For an optional AI review, add the gemini-review label.
• I have necessary comments in my code, particularly in hard-to-understand areas.
• I have run end-to-end tests tests and provided workload links above if applicable.
• I have made or will make corresponding changes to the doc if needed, including adding new documentation pages to the relevant Table of Contents (toctree directive) as explained in our documentation.

[codecov]

Codecov Report

❌ Patch coverage is 44.06780% with 99 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
.../trainers/post_train/distillation/train_distill.py	44.06%	94 Missing and 5 partials ⚠️

📢 Thoughts on this report? Let us know!

[shralex]

@codecov-ai-reviewer review

[github-actions]

🤖 Hi @SurbhiJainUSC, I've received your request, and I'm working on it now! You can track my progress in the logs for more details.

[github-actions]

📋 Review Summary

This Pull Request successfully moves the distillation module to a new location (src/maxtext/trainers/post_train/distillation) as part of the repo restructuring. The changes are well-implemented with backward compatibility in mind through a shim, and relevant configuration files have been updated.

🔍 General Feedback

• The approach of using a shim for backward compatibility is a good practice to minimize disruption during code migration.
• The updates to .coveragerc, codecov.yml, and GitHub Actions workflows (run_pathways_tests.yml, run_tests_against_package.yml) are thorough and correctly reflect the new module structure.

[gagika]

Thanks