--- tags: - unsloth --- # qwen2.5-coder-7b-sft-v2-huyen-889 Model finetuned on the full prompt - exactly as used in the reward generation pipeline.