--- library_name: transformers --- # refiner-gpt54-rubric_V3-2-gpt54-rl_5e-6-answer_only — step 50 GRPO checkpoint trained from `lihaoxin2020/qwen3-4b-refiner-gpt54-ep2`. ## Training run - W&B: https://wandb.ai/lihaoxin2020-yale-university/refiner-sft-grpo/runs/griuo1l8/overview