dei-model / train_configs

Commit History

  1. Add HuggingFace Hub checkpoint persistence - upload and download checkpoints between jobs

    renpas22
    3b46388
  2. Restore SFT to 10000 steps - will resume from checkpoint at 7000

    renpas22
    e9e0301
  3. Reduce training steps to fit within job timeout (SFT: 5k, PRM: 2k, RL: 3k)

    renpas22
    5cfd8d6

  1. Quote mixed_precision value

    renpas22
    37e8f2f
  2. Add inference config parameters

    renpas22
    2d1ba1a
  3. Add missing RL/PPO config parameters

    renpas22
    5af9eca

  1. Fix model name to use valid Qwen2-VL model

    renpas22
    61dbc34
  2. Enable 8-bit quantization and reduce batch size for memory

    renpas22
    d4b8544
  3. Add required top-level config keys for trainer

    renpas22
    9770aa1
  4. Add training scripts and configs

    renpas22
    2b8876a