Instructions to use TheBloke/stable-vicuna-13B-GPTQ with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use TheBloke/stable-vicuna-13B-GPTQ with Transformers:

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

pipe = pipeline("text-generation", model="TheBloke/stable-vicuna-13B-GPTQ")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("TheBloke/stable-vicuna-13B-GPTQ")
model = AutoModelForCausalLM.from_pretrained("TheBloke/stable-vicuna-13B-GPTQ", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use TheBloke/stable-vicuna-13B-GPTQ with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "TheBloke/stable-vicuna-13B-GPTQ"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "TheBloke/stable-vicuna-13B-GPTQ",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/TheBloke/stable-vicuna-13B-GPTQ

SGLang

How to use TheBloke/stable-vicuna-13B-GPTQ 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 "TheBloke/stable-vicuna-13B-GPTQ" \
    --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": "TheBloke/stable-vicuna-13B-GPTQ",
		"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 "TheBloke/stable-vicuna-13B-GPTQ" \
        --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": "TheBloke/stable-vicuna-13B-GPTQ",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use TheBloke/stable-vicuna-13B-GPTQ with Docker Model Runner:
```
docker model run hf.co/TheBloke/stable-vicuna-13B-GPTQ
```

ask

#12

by ReD2401 - opened May 4, 2023

Discussion

ReD2401

May 4, 2023

•

edited May 5, 2023

Hello TheBloke, I hope you are well.

I thank you for all your effort O)/

TheBloke

Owner May 4, 2023

•

edited May 4, 2023

I'm happy to take a look. There are some complications with GPT-J models. llama.cpp can't load them, and the latest and best 4bit quantisation code for CPU doesn't work with them. It would be possible to use them in GPT4ALL-Chat though, or there's a CLI version that also supports GPT-J for CPU.

GPTQ 4bit for GPU should be possible, but not as well supported.

I will give it a go and let you know!

ReD2401

May 4, 2023

thank you so much i really appreciate it

TheBloke

Owner May 6, 2023

•

edited May 6, 2023

Why did you edit the model out of your comment? Do you not want me to look at it any more? Or did someone else do it?

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