Instructions to use China-NCTIEDA/ChipExpert-8B-Instruct with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use China-NCTIEDA/ChipExpert-8B-Instruct with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="China-NCTIEDA/ChipExpert-8B-Instruct")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("China-NCTIEDA/ChipExpert-8B-Instruct")
model = AutoModelForCausalLM.from_pretrained("China-NCTIEDA/ChipExpert-8B-Instruct")
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]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use China-NCTIEDA/ChipExpert-8B-Instruct with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "China-NCTIEDA/ChipExpert-8B-Instruct"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "China-NCTIEDA/ChipExpert-8B-Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/China-NCTIEDA/ChipExpert-8B-Instruct

SGLang

How to use China-NCTIEDA/ChipExpert-8B-Instruct 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 "China-NCTIEDA/ChipExpert-8B-Instruct" \
    --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": "China-NCTIEDA/ChipExpert-8B-Instruct",
		"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 "China-NCTIEDA/ChipExpert-8B-Instruct" \
        --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": "China-NCTIEDA/ChipExpert-8B-Instruct",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use China-NCTIEDA/ChipExpert-8B-Instruct with Docker Model Runner:
```
docker model run hf.co/China-NCTIEDA/ChipExpert-8B-Instruct
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

ChipExpert-8B-Instruct

The First Open-Source Integrated-Circuit-Design-Specific Large Language Model．

💻Github

Introduction:

ChipExpert is the first open-source, instructional LLM dedicated to the Integrated-Circuit-Design industry, covering knowledge across multiple sub-domains, including analog circuits, digital circuits, radio frequency (RF), semiconductor devices, electronic design automation (EDA), system-on-chip (SoC), computing-in-memory, and more. This model aims to provide teaching assistant services for students in the field of IC to learn fundamental knowledge, engineers to inquire about technical details, and researchers to investigate cutting-edge papers and research topics. The ultimate goal of this model is to help the integrated circuit industry reduce the learning barrier and lower the training costs.

Key Features:

The first Large Language Model (LLM) in the IC design field.
A professional corpus covering ten specialized areas of IC design.
Achieves superior performance in both foundational and cutting-edge knowledge compared to general LLMs.

Contributions

This project is the result of a collaborative effort：

Ning Xu^1,2    Zhaoyang Zhang^1,2    Lei Qi^1,2    Wensuo Wang¹    Chao Zhang¹    Zihao Ren²
Huaiyuan Zhang²    Yanqi Zhang²    Zhichao Liu²    Xing Wang²    Qingwen Wei²    Shiyang Wu²
Lanlan Yang²    Xin Geng²    Yuchen Ma²    Yutong Zhang²    Mengyao Kong²
Zhican Zhang²    Shiyang Wu²    Yao Wang²    Lanlan Yang¹    Chen Yang¹
Qianfeng Lu²    Yiqun Ma²    Zhengxuan Wang²    Yaoyao Xu²    Chengjie Liu¹
Mengyao Zhao²    Junbo Liu²    Yufan Song¹    Yuejian Shi²    Jun Yang^1,2

¹National Center of Technology Innovation for EDA, Nanjing, China
²Southeast University, Nanjing, China

Model Description

Developed by: NCTIEDA (National Center of Technology Innovation for EDA) and Southeast University
Model type: Instruction Model
Language(s): English
License: Apache License 2.0
Finetuned from model: Llama 3

Out-of-Scope Use

This is the first version of ChipExpert, more capable versions with enhanced abilities will be released soon.

Citation

If you find our work helpful, please consider citing the following paper:

@article{chipexpert2024,
  title={ChipExpert: The First Open-Source Integrated-Circuit-Design-Specific Large Language Model},
  author={Ning Xu, ZhaoyangZhang et al.},
  journal={arXiv preprint arXiv:2024.xxxxx},
  year={2024}
}

Model Card Contact

chipexpert@nctieda.com

Downloads last month: 6

Safetensors

Model size

8B params

Tensor type

BF16

Model tree for China-NCTIEDA/ChipExpert-8B-Instruct

Quantizations

1 model