How to use ivaning0919/pagent-hintflow-dpo-v2-adapter with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("/home/ivaning/prompt-r1r/Prompt-R1/Qwen/Qwen3-4B") model = PeftModel.from_pretrained(base_model, "ivaning0919/pagent-hintflow-dpo-v2-adapter")
LoRA adapter for the HintFlow orchestrator after offline DPO on v2 tree pairs.
See merged weights: ivaning0919/pagent-hintflow-dpo-v2-merged.
ivaning0919/pagent-hintflow-dpo-v2-merged
Training meta is in dpo_meta.json.
dpo_meta.json
Base model
