Instructions to use stabilityai/stablecode-completion-alpha-3b-4k with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use stabilityai/stablecode-completion-alpha-3b-4k with Transformers:

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

pipe = pipeline("text-generation", model="stabilityai/stablecode-completion-alpha-3b-4k")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("stabilityai/stablecode-completion-alpha-3b-4k")
model = AutoModelForCausalLM.from_pretrained("stabilityai/stablecode-completion-alpha-3b-4k")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use stabilityai/stablecode-completion-alpha-3b-4k with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "stabilityai/stablecode-completion-alpha-3b-4k"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "stabilityai/stablecode-completion-alpha-3b-4k",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/stabilityai/stablecode-completion-alpha-3b-4k

SGLang

How to use stabilityai/stablecode-completion-alpha-3b-4k 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 "stabilityai/stablecode-completion-alpha-3b-4k" \
    --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": "stabilityai/stablecode-completion-alpha-3b-4k",
		"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 "stabilityai/stablecode-completion-alpha-3b-4k" \
        --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": "stabilityai/stablecode-completion-alpha-3b-4k",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use stabilityai/stablecode-completion-alpha-3b-4k with Docker Model Runner:
```
docker model run hf.co/stabilityai/stablecode-completion-alpha-3b-4k
```

Activating the model on huggingface-vscode

by yurqua - opened Aug 9, 2023

Discussion

yurqua

Aug 9, 2023

Hi. Newbie here. Could you please help me to get around the ways of using the model in VS Code?

I'm trying to activate StableCode-Completion-Alpha-3B-4K on https://github.com/huggingface/huggingface-vscode
I have generated a HF token with a write permission and added it in VS Code extension settings. The bigcode/starcoder model works fine.
On the model page I pressed the '🚀 Deploy' button, selected 'Inference API' item, copied the https://api-inference.huggingface.co/models/stabilityai/stablecode-completion-alpha-3b-4k API endpoint URL and pasted it into the 'Hugging Face Code: Model ID Or Endpoint' settings field.
Unlike with the default bigcode/starcoder model, the extension doesn't seem to work. I tried to restart extensions host and also tried stabilityai/stablecode-completion-alpha-3b-4k as value with no luck.

When I type in my editor, I see 🔄 Status bar icon next to Hugging Face Code spinning for a split of a second, and then nothing happens.

The OUTPUT panel shows something like:
INPUT to API: (with parameters {"max_new_tokens":60,"temperature":null,"do_sample":false,"top_p":0.95,"stop":["<|endoftext|>"]})
def main

What am I doing wrong?