Qwen2.5 7B Chain-of-Thought

Published by @reasoning_lab · Community adapter

Chain-of-thought reasoning adapter trained on mathematical problem solutions and logical deduction traces. Shows intermediate steps, catches errors in reasoning chains, and self-corrects.

Usage

from peft import PeftModel
from transformers import AutoModelForCausalLM, AutoTokenizer

tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen2.5-7B-Instruct")
base = AutoModelForCausalLM.from_pretrained("Qwen/Qwen2.5-7B-Instruct")
model = PeftModel.from_pretrained(base, "modelforgelab/qwen25-7b-cot-lora")

inputs = tokenizer("Your prompt here", return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=256)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Compatibility

• transformers>=4.40
• vLLM