Instructions to use stabilityai/StableBeluga2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use stabilityai/StableBeluga2 with Transformers:

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

pipe = pipeline("text-generation", model="stabilityai/StableBeluga2")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("stabilityai/StableBeluga2")
model = AutoModelForCausalLM.from_pretrained("stabilityai/StableBeluga2")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use stabilityai/StableBeluga2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "stabilityai/StableBeluga2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "stabilityai/StableBeluga2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/stabilityai/StableBeluga2

SGLang

How to use stabilityai/StableBeluga2 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/StableBeluga2" \
    --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/StableBeluga2",
		"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/StableBeluga2" \
        --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/StableBeluga2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use stabilityai/StableBeluga2 with Docker Model Runner:
```
docker model run hf.co/stabilityai/StableBeluga2
```

Upload 16 bit precision weights

#29

by mallorbc - opened Aug 4, 2023

Discussion

mallorbc

Aug 4, 2023

These weights are the full fp32. To save bandwidth and disk space, upload 16 bits.

breadlicker45

Aug 4, 2023

These weights are the full fp32. To save bandwidth and disk space, upload 16 bits.

lol

mallorbc

Aug 4, 2023

These weights are the full fp32. To save bandwidth and disk space, upload 16 bits.

lol

What is funny about this?

breadlicker45

Aug 5, 2023

•

edited Aug 5, 2023

These weights are the full fp32. To save bandwidth and disk space, upload 16 bits.

lol

What is funny about this?

I really don't think they are going to lower the quality just for someone but someone will upload a 16bit version of it at some point in time.

mallorbc

Aug 7, 2023

These weights are the full fp32. To save bandwidth and disk space, upload 16 bits.

lol

What is funny about this?

I really don't think they are going to lower the quality just for someone but someone will upload a 16bit version of it at some point in time.

Upload weights at different precisions. Or upload bf16

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment