---
title: J
canonical_url: "https://www.modelscope.cn/models/jacksong123/J"
md_url: "https://www.modelscope.cn/models/jacksong123/J.md"
repository: jacksong123/J
chinese_name: J
last_updated: 2026-05-01
pipeline_tag: nli
tasks:
  - nli
parameters: 70.4M
tensor_type:
  - F32
library_name:
  - safetensors
  - pytorch
frameworks:
  - Pytorch
downloads: 5
stars: 0
---

# J

> J - jacksong123 在 ModelScope 开源的模型。微调Qwen-2.5-0.5b系列模型，通过独特的技术得到结果

- **Repository**: jacksong123/J
- **Tasks**: nli
- **Parameters**: 70.4M
- **Downloads**: 5
- **Stars**: 0
- **Last updated**: 2026-05-01

Source: https://www.modelscope.cn/models/jacksong123/J

---

# J - 对话助手模型

## 模型介绍

J 是一个基于 Qwen2-0.5B-Instruct 微调的对话助手模型，专注于提供友好、专业的对话体验。

## 特点

- 支持多轮对话
- 响应简洁明了
- 专业且友好的语气

## 技术规格

- **基础模型**: Qwen2-0.5B-Instruct
- **微调方法**: LoRA
- **参数规模**: 约1M可训练参数

## 使用方法

```python
from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel

# 加载基座模型
base_model = "Qwen/Qwen2-0.5B-Instruct"
model = AutoModelForCausalLM.from_pretrained(base_model)
tokenizer = AutoTokenizer.from_pretrained(base_model)

# 加载LoRA适配器
model = PeftModel.from_pretrained(model, "./Lora")

# 生成回复
messages = [{"role": "user", "content": "你好"}]
text = tokenizer.apply_chat_template(messages, tokenize=False)
inputs = tokenizer(text, return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=500)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
```

## 许可证

Apache-2.0

## 作者

jacksong123