Text Generation
Transformers
Safetensors
Chinese
llama
code
chinese
easypl
minicpm
lora
sft
conversational
text-generation-inference
Instructions to use XuehangCang/EasyPL-1B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use XuehangCang/EasyPL-1B with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="XuehangCang/EasyPL-1B") messages = [ {"role": "user", "content": "Who are you?"}, ] pipe(messages)# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("XuehangCang/EasyPL-1B") model = AutoModelForCausalLM.from_pretrained("XuehangCang/EasyPL-1B", device_map="auto") messages = [ {"role": "user", "content": "Who are you?"}, ] inputs = tokenizer.apply_chat_template( messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt", ).to(model.device) outputs = model.generate(**inputs, max_new_tokens=40) print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:])) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use XuehangCang/EasyPL-1B with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "XuehangCang/EasyPL-1B" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "XuehangCang/EasyPL-1B", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/XuehangCang/EasyPL-1B
- SGLang
How to use XuehangCang/EasyPL-1B with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "XuehangCang/EasyPL-1B" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "XuehangCang/EasyPL-1B", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "XuehangCang/EasyPL-1B" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "XuehangCang/EasyPL-1B", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }' - Docker Model Runner
How to use XuehangCang/EasyPL-1B with Docker Model Runner:
docker model run hf.co/XuehangCang/EasyPL-1B
| license: mit | |
| datasets: | |
| - XuehangCang/e_style_code | |
| language: | |
| - zh | |
| base_model: | |
| - openbmb/MiniCPM5-1B | |
| base_model_relation: finetune | |
| pipeline_tag: text-generation | |
| library_name: transformers | |
| tags: | |
| - text-generation | |
| - code | |
| - chinese | |
| - easypl | |
| - minicpm | |
| - lora | |
| - sft | |
| # EasyPL-1B | |
| EasyPL-1B 是一个面向 EasyPL(易语言 Easy Programming Language)编程场景的中文代码助手模型 | |
| 模型目标是根据中文需求生成清晰、简洁、可运行的代码,同时提供必要的中文解释,适合教学示例、语法演示和轻量级编程辅助 | |
| ## Model Details | |
| | 项目 | 内容 | | |
| | --- | --- | | |
| | Model name | EasyPL-1B | | |
| | Base model | `openbmb/MiniCPM5-1B` | | |
| | Dataset | `XuehangCang/e_style_code` | | |
| | Language | Chinese | | |
| | Training method | SFT + LoRA | | |
| | Task | Text generation / Code generation | | |
| | License | MIT | | |
| ## Intended Use | |
| EasyPL-1B 主要适用于: | |
| - 根据中文需求生成示例代码 | |
| - 解释代码片段的执行逻辑 | |
| - 生成函数、字符串处理、循环、条件判断等基础编程示例 | |
| - 将结构化题目、自然语言描述整理为代码或中文说明 | |
| 该模型更偏向编程教学与示例生成,不建议作为通用问答模型或复杂工程自动化模型使用 | |
| ## Coding Style Skill | |
| EasyPL-1B 配套提供 [易语言风格 Python 编码规范 SKILL.md](./易语言风格%20Python%20编码规范%20SKILL.md),用于说明模型推荐遵循的中文化 Python 编码风格。 | |
| 该规范覆盖中文函数名和变量名、模块组织、函数设计、中文注释、异常处理、PEP8 格式约束、常用代码模板与质量检查清单。需要了解 EasyPL 风格代码的命名规则、生成边界和示例模板时,可以点击上方链接查看详情文件。 | |
| ## How to Use | |
| ```python | |
| from transformers import AutoModelForCausalLM, AutoTokenizer | |
| model_id = "XuehangCang/EasyPL-1B" | |
| tokenizer = AutoTokenizer.from_pretrained(model_id, trust_remote_code=True) | |
| model = AutoModelForCausalLM.from_pretrained( | |
| model_id, | |
| torch_dtype="auto", | |
| device_map="auto", | |
| trust_remote_code=True, | |
| ) | |
| messages = [ | |
| {"role": "user", "content": "生成随机UUID"}, | |
| ] | |
| inputs = tokenizer.apply_chat_template( | |
| messages, | |
| tokenize=True, | |
| add_generation_prompt=True, | |
| enable_thinking=False, | |
| return_tensors="pt", | |
| ) | |
| inputs = {key: value.to(model.device) for key, value in inputs.items()} | |
| outputs = model.generate(**inputs, max_new_tokens=2048) | |
| prompt_len = inputs["input_ids"].shape[-1] | |
| print(tokenizer.decode(outputs[0][prompt_len:], skip_special_tokens=True)) | |
| ``` | |
| 如果使用的 `transformers` 或 tokenizer 版本不支持 `enable_thinking` 参数,可以移除该参数后再运行 | |
| ## Example Outputs | |
| 以下示例均使用模型实际调用生成 | |
| ### 生成随机UUID | |
| ```python | |
| import uuid | |
| def 生成随机UUID(): | |
| """生成随机UUID""" | |
| return str(uuid.uuid4()) | |
| # 测试 | |
| print(生成随机UUID()) | |
| ``` | |
| ### 判断一个数字是否为偶数,并输出结果 | |
| ```python | |
| def 取模( 数字, 基数): | |
| """取模函数,返回 数字 % 基数 结果""" | |
| return 数字 % 基数 | |
| def 取余( 数字, 基数): | |
| """取余函数,返回 数字 % 基数 结果""" | |
| return 取模( 数字, 基数) | |
| def 判断偶数( 数字): | |
| """判断数字是否为偶数""" | |
| return 取余( 数字, 2) == 0 | |
| def 判断奇数( 数字): | |
| """判断数字是否为奇数""" | |
| return 取余( 数字, 2) != 0 | |
| # 测试 | |
| if __name__ == '__main__': | |
| 结果1 = 判断偶数(64) | |
| print(f"64 是偶数:{结果1}") | |
| 结果2 = 判断奇数(64) | |
| print(f"64 是奇数:{结果2}") | |
| 结果3 = 判断偶数(7) | |
| print(f"7 是偶数:{结果3}") | |
| 结果4 = 判断奇数(10) | |
| print(f"10 是奇数:{结果4}") | |
| 结果5 = 判断偶数(3) | |
| print(f"3 是偶数:{结果5}") | |
| ``` | |
| ### 计算字符串的MD5值 | |
| ```python | |
| import hashlib | |
| def 计算MD5(数据文本): | |
| """ | |
| 计算字符串的MD5值 | |
| 参数: | |
| 数据文本: 需要计算MD5的字符串 | |
| 返回: | |
| MD5值字符串 | |
| """ | |
| md5 = hashlib.md5() | |
| md5.update(数据文本.encode('utf-8')) | |
| return md5.hexdigest() | |
| ``` | |
| 这些输出体现了模型对易语言风格的适配:短中文需求进入,返回带中文函数名、中文注释、测试代码或简要说明的代码片段 | |
| ## Limitations | |
| - EasyPL-1B 是针对中文编程场景微调的模型,通用知识问答能力不是主要优化目标 | |
| - 生成代码仍可能出现语法或逻辑错误,建议在实际使用前进行人工检查和运行验证 | |
| - 模型能力受训练数据质量、训练样本规模、训练轮数和基座模型能力影响 | |
| ## Citation | |
| 如果你使用了本模型,请同时关注并引用相关基座模型与数据集: | |
| - Base model: `openbmb/MiniCPM5-1B` | |
| - Dataset: `XuehangCang/e_style_code` | |
| ## License | |
| This model card is released under the MIT License. | |