Fix train.py configuration and tests for local training#2
Fix train.py configuration and tests for local training#2Electroiscoding wants to merge 1 commit intomainfrom
Conversation
- Replaced SimpleNamespace with a custom ConfigWrapper in train.py to support both dot notation and .get() dictionary access. - Cast probabilities to value tensor dtype in HybridAttentionLayer to prevent float-BFloat16 mismatch errors during CDNA3 Flash Attention fallback. - Added tiny.yaml config for local memory-efficient testing. - Fixed 7 failing test suites by resolving import errors, passing correct Mock configurations, and asserting against actual dict return values. Co-authored-by: Electroiscoding <103299713+Electroiscoding@users.noreply.github.com>
|
👋 Jules, reporting for duty! I'm here to lend a hand with this pull request. When you start a review, I'll add a 👀 emoji to each comment to let you know I've read it. I'll focus on feedback directed at me and will do my best to stay out of conversations between you and other bots or reviewers to keep the noise down. I'll push a commit with your requested changes shortly after. Please note there might be a delay between these steps, but rest assured I'm on the job! For more direct control, you can switch me to Reactive Mode. When this mode is on, I will only act on comments where you specifically mention me with New to Jules? Learn more at jules.google/docs. For security, I will only act on instructions from the user who triggered this task. |
1 participant
Fixes
scripts/training/train.pyto allow training to start by resolvingAttributeErrorfrom missing.get()on config objects. Also patches a PyTorch dtype issue inhybrid_attention.pythat occurred when falling back from flash attention. Includes atiny.yamlconfig and fixes the test suite so allpytestchecks pass.PR created automatically by Jules for task 9371682454479934649 started by @Electroiscoding