medical-expert/v3_sftmerged_grounding_grpo_step1600

v3sftmergedgroundinggrpostep1600

GRPO LoRA adapter for chest X-ray phrase grounding, on top of the v3 SFT-merged Lingshu base (lingshu_v3_sft_merged, Qwen2.5-VL-7B). Global step 1600.

• —Base model: local merged base lingshu_v3_sft_merged (see adapter_config.json)
• —Method: GRPO (EasyR1/veRL), dual-expert (MDETR/TransVG) tool-calling rollout
• —Reward: overall = 0.1format + 0.9IoU
• —Hyperparams: lr 5e-6, klcoef 1e-1, klpenalty lowvarkl, enforce_eager, LoRA rank 32 / alpha 64, all-linear, vision frozen
• —rollout: n=2, temperature 1.0, tp=4, gpu_util 0.6

Notes

• —Best val IoU of this run so far: 0.1816 @ step 1400 (val evals every 100 steps oscillate 0.174-0.182).
• —Near step 1600: train format ~1.0, accuracy(IoU) ~0.34, anyexpertrate ~0.58 (MDETR-dominated), reward overall ~0.41.
• —basemodel in adapterconfig.json points to the local merged base path; load against that base.