The PeterJinGo/SearchR1-nq_hotpotqa_train-qwen2.5-7b-em-grpo-v0.3 galaxy

1 models are built on PeterJinGo/SearchR1-nq_hotpotqa_train-qwen2.5-7b-em-grpo-v0.3: 0 quantized, 0 fine-tuned, 1 adapters and 0 merges. Together they were downloaded 95,682 times in the last 30 days.

Most downloaded direct derivatives

• reasonrag/decision-boundary-searchr1-7b-dpo-v3-overled-lora64 (adapter), 6 downloads in 30 days

Sister project: Paper Pulse, the upvote history of every Hugging Face Daily Paper.