Torch-free layer-wise test experiment

[hengtaoguo]

Description

1. Fix the rand seed, generate weights and inputs;
2. Get the output array, calculate stats such as max/min/std, and hard-code into the codes.
3. When on CI, only run MT forward pass and ignore HF forward pass due to lack of torch in CI.

Problem: Even with fixed rand seed, the weights/inputs could be numerically offset among difference devices.

Tests

Please describe how you tested this change, and include any instructions and/or
commands to reproduce.

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

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!