Instructions to use Wannita/baseline_codecompletion with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Wannita/baseline_codecompletion with Transformers:

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

pipe = pipeline("text-generation", model="Wannita/baseline_codecompletion")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Wannita/baseline_codecompletion")
model = AutoModelForCausalLM.from_pretrained("Wannita/baseline_codecompletion")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Wannita/baseline_codecompletion with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Wannita/baseline_codecompletion"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Wannita/baseline_codecompletion",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/Wannita/baseline_codecompletion

SGLang

How to use Wannita/baseline_codecompletion 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 "Wannita/baseline_codecompletion" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Wannita/baseline_codecompletion",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "Wannita/baseline_codecompletion" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Wannita/baseline_codecompletion",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use Wannita/baseline_codecompletion with Docker Model Runner:
```
docker model run hf.co/Wannita/baseline_codecompletion
```

PyCoder

This repository contains the baseline model for the paper Syntax-Aware On-the-Fly Code Completion

The sample code to run the model can be found in the directory: "assets/notebooks/inference.ipynb" in our GitHub: https://github.com/awsm-research/pycoder.

PyCoder is an auto code completion model which leverages a Multi-Task Training technique (MTT) to cooperatively learn the code prediction task and the type prediction task. For the type prediction task, we propose to leverage the standard Python token type information (e.g., String, Number, Name, Keyword), which is readily available and lightweight, instead of using the AST information which requires source code to be parsable for an extraction, limiting its ability to perform on-the-fly code completion (see Section 2.3 in our paper).

More information can be found in our paper.

If you use our code or PyCoder, please cite our paper.

@article{takerngsaksiri2022syntax,
  title={Syntax-Aware On-the-Fly Code Completion},
  author={Takerngsaksiri, Wannita and Tantithamthavorn, Chakkrit and Li, Yuan-Fang},
  journal={arXiv preprint arXiv:2211.04673},
  year={2022}
}

Downloads last month: 8

Dataset used to train Wannita/baseline_codecompletion

Space using Wannita/baseline_codecompletion 1

Paper for Wannita/baseline_codecompletion

Syntax-Aware On-the-Fly Code Completion

Paper • 2211.04673 • Published Nov 9, 2022