Commit History

  1. Update README with comprehensive model card

    Verified signature
    df8a81e
  2. Add Google Colab training notebook for V2 GRPO training (free T4 path)

    Verified signature
    55ef8ec

  1. v2: 5K subset for A10G, fix escaping

    Verified signature
    3c818d7
  2. fix: escape syntax in quality_reward

    Verified signature
    75be256
  3. GRPO training complete — smart contract security auditor

    Verified signature
    93b6a9a
  4. add: GRPO v2 training script with 4 reward functions + dataset builder

    Verified signature
    9ab390c
  5. fix: total_mem -> total_memory for PyTorch compat

    Verified signature
    39535c8
  6. fix: disable all Hub calls during trainer init to prevent 401

    Verified signature
    3716618
  7. Upload tokenizer

    Verified signature
    1fb5e81
  8. Upload Qwen2ForCausalLM

    Verified signature
    044e65a
  9. Upload train_grpo_job.py with huggingface_hub

    Verified signature
    df05b8e
  10. Upload train_grpo_job.py with huggingface_hub

    Verified signature
    0ee8b77
  11. Upload train_grpo_job.py with huggingface_hub

    Verified signature
    eac5c9b
  12. Upload train_grpo_job.py with huggingface_hub

    Verified signature
    7168e35

  1. Upload train_grpo_job.py with huggingface_hub

    Verified signature
    74022f8
  2. initial commit

    Verified signature
    c4b5a68