invoiceguard-code / training

Commit History

  1. Upload training/train_grpo.py with huggingface_hub

    Verified signature
    06b7563
  2. v6: GRPO with --resume-adapter to start from SFT checkpoint

    Verified signature
    59ed7f3
  3. v5d: save best-epoch checkpoint during SFT

    Verified signature
    a674764
  4. v5c: add --min-investigation-steps filter for submit-only with longer contexts

    Verified signature
    2857a20
  5. v5b: add --submit-only SFT mode

    Verified signature
    2a667e7
  6. v5: remove 4-bit quant, variable-length traces, submit oversampling, GRPO exploration fixes

    Verified signature
    1d75102
  7. Upload folder using huggingface_hub

    Verified signature
    4cd2128
  8. Fix: decode with skip_special_tokens=False to preserve <think> tags for stripping

    Verified signature
    231d76f
  9. Fix Qwen3 thinking mode + increase max_new_tokens: training/launch_hf_job.py

    Verified signature
    befff2d
  10. Fix Qwen3 thinking mode + increase max_new_tokens: training/eval_adapter.py

    Verified signature
    664fa8e
  11. Fix Qwen3 thinking mode + increase max_new_tokens: training/train_sft.py

    Verified signature
    7c48dcf
  12. Fix Qwen3 thinking mode + increase max_new_tokens: training/train_grpo.py

    Verified signature
    9d40fed
  13. Fix Qwen3 thinking mode + increase max_new_tokens: training/rollout.py

    Verified signature
    f788adf
  14. Add InvoiceGuard SFT and merge tooling

    Verified signature
    25ee43d
  15. Add InvoiceGuard adapter eval job script

    Verified signature
    90c05cf
  16. Sync InvoiceGuard code for GRPO training job

    Verified signature
    02767f3
  17. Sync InvoiceGuard code for GRPO training job

    Verified signature
    7db1128
  18. Sync InvoiceGuard code for GRPO training job

    Verified signature
    7c278c9
  19. Sync InvoiceGuard code for GRPO training job

    Verified signature
    c5cfe6d
  20. Sync InvoiceGuard code for GRPO training job

    Verified signature
    d1e9e92
  21. Sync InvoiceGuard code for GRPO training job

    Verified signature
    c01c661
  22. Sync InvoiceGuard code for GRPO training job

    Verified signature
    fdd2256
  23. Sync InvoiceGuard code for GRPO training job

    Verified signature
    25e6791
  24. Sync InvoiceGuard code for GRPO training job

    Verified signature
    9a88af0