Instructions to use TheBloke/CodeLlama-7B-Instruct-AWQ with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use TheBloke/CodeLlama-7B-Instruct-AWQ with Transformers:

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

pipe = pipeline("text-generation", model="TheBloke/CodeLlama-7B-Instruct-AWQ")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("TheBloke/CodeLlama-7B-Instruct-AWQ")
model = AutoModelForCausalLM.from_pretrained("TheBloke/CodeLlama-7B-Instruct-AWQ")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use TheBloke/CodeLlama-7B-Instruct-AWQ with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "TheBloke/CodeLlama-7B-Instruct-AWQ"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "TheBloke/CodeLlama-7B-Instruct-AWQ",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/TheBloke/CodeLlama-7B-Instruct-AWQ

SGLang

How to use TheBloke/CodeLlama-7B-Instruct-AWQ 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/CodeLlama-7B-Instruct-AWQ" \
    --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/CodeLlama-7B-Instruct-AWQ",
		"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/CodeLlama-7B-Instruct-AWQ" \
        --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/CodeLlama-7B-Instruct-AWQ",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use TheBloke/CodeLlama-7B-Instruct-AWQ with Docker Model Runner:
```
docker model run hf.co/TheBloke/CodeLlama-7B-Instruct-AWQ
```

What GPU recommendations for this?

by hazrpg - opened Nov 30, 2023

Discussion

hazrpg

Nov 30, 2023

Hey, thanks so much for releasing these.

I currently have a 3060 12GB and it fails to load - it fills up the available 11GB (presumably ~1GB is used for the OS) and then pytorch errors out with out of memory.

I might try the GGUF versions for now - but I'm curious what your recommendations are for this model, how much vRAM, etc.

YaTharThShaRma999

Nov 30, 2023

It should only take like 6gb vram? Infact you could probably load a 13b model as well. What are you using to load the model?

kmclester

Dec 4, 2023

If you're using vLLM, it pre-computes and reserves the necessary memory for cache and prompt conversions. So, a 6gb model may actually reserve >12gb of vram.

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