train_sft: drop fp16, prefer bf16 (MPS-compatible without grad scaler) 35fa944 bpHigh Claude Opus 4.7 (1M context) commited on Apr 26
Phase 10: SFT training script (Qwen2.5-Coder-3B + LoRA via TRL) c803cd5 bpHigh Claude Opus 4.7 (1M context) commited on Apr 26