🏠 Homepage｜🛠 Extensions VS Code, Jetbrains｜🤗 HF Repo | 🪧 HF DEMO

English | 中文 | 日本語

We introduce CodeGeeX4-ALL-9B, the open-source version of the latest CodeGeeX4 model series. It is a multilingual code generation model continually trained on the GLM-4-9B, significantly enhancing its code generation capabilities. Using a single CodeGeeX4-ALL-9B model, it can support comprehensive functions such as code completion and generation, code interpreter, web search, function call, repository-level code Q&A, covering various scenarios of software development. CodeGeeX4-ALL-9B has achieved highly competitive performance on public benchmarks, such as BigCodeBench and NaturalCodeBench. It is currently the most powerful code generation model with less than 10B parameters, even surpassing much larger general-purpose models, achieving the best balance in terms of inference speed and model performance.

CodeGeeX4 is now available on Ollama! Please install Ollama 0.2 or later and run the following command:

ollama run codegeex4

To connect the local model to our VS Code / Jetbrains extensions, please check Local Mode Guideline.

Use 4.39.0<=transformers<=4.40.2 to quickly launch codegeex4-all-9b：

import torch
from transformers import AutoTokenizer, AutoModelForCausalLM

device = "cuda" if torch.cuda.is_available() else "cpu"
tokenizer = AutoTokenizer.from_pretrained("THUDM/codegeex4-all-9b", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(
    "THUDM/codegeex4-all-9b",
    torch_dtype=torch.bfloat16,
    low_cpu_mem_usage=True,
    trust_remote_code=True
).to(device).eval()
inputs = tokenizer.apply_chat_template([{"role": "user", "content": "write a quick sort"}], add_generation_prompt=True, tokenize=True, return_tensors="pt", return_dict=True ).to(device)
with torch.no_grad():
    outputs = model.generate(**inputs)
    outputs = outputs[:, inputs['input_ids'].shape[1]:]
    print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Use vllm==0.5.1 to quickly launch codegeex4-all-9b:

from transformers import AutoTokenizer
from vllm import LLM, SamplingParams

# CodeGeeX4-ALL-9B
# max_model_len, tp_size = 1048576, 4
# If OOM，please reduce max_model_len，or increase tp_size
max_model_len, tp_size = 131072, 1
model_name = "codegeex4-all-9b"
prompt = [{"role": "user", "content": "Hello"}]

tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
llm = LLM(
    model=model_name,
    tensor_parallel_size=tp_size,
    max_model_len=max_model_len,
    trust_remote_code=True,
    enforce_eager=True,
    # If OOM，try using follong parameters
    # enable_chunked_prefill=True,
    # max_num_batched_tokens=8192
)
stop_token_ids = [151329, 151336, 151338]
sampling_params = SamplingParams(temperature=0.95, max_tokens=1024, stop_token_ids=stop_token_ids)

inputs = tokenizer.apply_chat_template(prompt, tokenize=False, add_generation_prompt=True)
outputs = llm.generate(prompts=inputs, sampling_params=sampling_params)

print(outputs[0].outputs[0].text)

Set up OpenAI Compatible Server via vllm, detailed please check OpenAI Compatible Server

python -m vllm.entrypoints.openai.api_server \
     --model THUDM/codegeex4-all-9b \
     --trust_remote_code

Codegeex4 now suport Candle framwork Repo

Use Rust to launch codegeex4-all-9b:

	cd candle_demo
	cargo build -p codegeex4-cli --release --features cuda # for Cuda
	cargo build -p codegeex4-cli --release # for cpu
	./target/release/codegeex4-cli --sample-len 512

CodeGeeX4-ALL-9B provides three user guides to help users quickly understand and use the model:

1. System Prompt Guideline: This guide introduces how to use system prompts in CodeGeeX4-ALL-9B, including the VSCode extension official system prompt, customized system prompts, and some tips for maintaining multi-turn dialogue history.

2. Infilling Guideline: This guide explains the VSCode extension official infilling format, covering general infilling, cross-file infilling, and generating a new file in a repository.

3. Repository Tasks Guideline: This guide demonstrates how to use repository tasks in CodeGeeX4-ALL-9B, including QA tasks at the repository level and how to trigger the aicommiter capability of CodeGeeX4-ALL-9B to perform deletions, additions, and changes to files at the repository level.

4. Local Mode Guideline：This guide introduces how to deploy CodeGeeX4-ALL-9B locally and connect it to Visual Studio Code / Jetbrains extensions.

These guides aim to provide a comprehensive understanding and facilitate efficient use of the model.

CodeGeeX4-ALL-9B is ranked as the most powerful model under 10 billion parameters, even surpassing general models several times its size, achieving the best balance between inference performance and model effectiveness.

CodeGeeX4-ALL-9B scored 48.9 and 40.4 for the complete and instruct tasks of BigCodeBench, which are the highest scores among models with less than 20 billion parameters. In CRUXEval, a benchmark for testing code reasoning, understanding, and execution capabilities, CodeGeeX4-ALL-9B presented remarkable results with its COT (chain-of-thought) abilities. From easy code generation tasks in HumanEval and MBPP, to very challenging tasks in NaturalCodeBench, CodeGeeX4-ALL-9B also achieved outstanding performance at its scale. It is currently the only code model that supports Function Call capabilities and even achieves a better execution success rate than GPT-4. Furthermore, in the "Code Needle In A Haystack" (NIAH) evaluation, the CodeGeeX4-ALL-9B model demonstrated its ability to retrieve code within contexts up to 128K, achieving a 100% retrieval accuracy in all python scripts.

Details of the evaluation results can be found in the Evaluation.

The code in this repository is open source under the Apache-2.0 license. The model weights are licensed under the Model License. CodeGeeX4-9B weights are open for academic research. For users who wish to use the models for commercial purposes, please fill in the registration form.

If you find our work helpful, please feel free to cite the following paper:

@inproceedings{zheng2023codegeex,
  title={CodeGeeX: A Pre-Trained Model for Code Generation with Multilingual Benchmarking on HumanEval-X},
  author={Qinkai Zheng and Xiao Xia and Xu Zou and Yuxiao Dong and Shan Wang and Yufei Xue and Zihan Wang and Lei Shen and Andi Wang and Yang Li and Teng Su and Zhilin Yang and Jie Tang},
  booktitle={Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining},
  pages={5673--5684},
  year={2023}
}