Qwen2.5 7B Chinese Instruct

Published by @zh_tuner · Community adapter

Sharpens Qwen2.5's Mandarin instruction-following on a 60 k curated dialogue dataset. Improved handling of formal registers, classical idioms, and structured output.

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-zh-instruct-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