Instructions to use SinclairSchneider/dbrx-instruct-quantization-fixed with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use SinclairSchneider/dbrx-instruct-quantization-fixed with Transformers:

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

pipe = pipeline("text-generation", model="SinclairSchneider/dbrx-instruct-quantization-fixed", trust_remote_code=True)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("SinclairSchneider/dbrx-instruct-quantization-fixed", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("SinclairSchneider/dbrx-instruct-quantization-fixed", trust_remote_code=True)

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use SinclairSchneider/dbrx-instruct-quantization-fixed with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/SinclairSchneider/dbrx-instruct-quantization-fixed

SGLang

How to use SinclairSchneider/dbrx-instruct-quantization-fixed 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 "SinclairSchneider/dbrx-instruct-quantization-fixed" \
    --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": "SinclairSchneider/dbrx-instruct-quantization-fixed",
		"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 "SinclairSchneider/dbrx-instruct-quantization-fixed" \
        --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": "SinclairSchneider/dbrx-instruct-quantization-fixed",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use SinclairSchneider/dbrx-instruct-quantization-fixed with Docker Model Runner:
```
docker model run hf.co/SinclairSchneider/dbrx-instruct-quantization-fixed
```

Quantisation question?

by pjw000 - opened Apr 1, 2024

Discussion

pjw000

Apr 1, 2024

•

edited Apr 1, 2024

Using this model should I be able to create a GPTQ 4 bit version? My naive attempts using the official intruct model produce the following error which I have no idea how to deal with:

ValueError: Block pattern could not be match. Pass block_name_to_quantize argument in quantize_model

Specifically: I have no idea what the block name might be!

SinclairSchneider

Owner Apr 1, 2024

There are already people trying to get a GPTQ version. I'd suggest to have a look at this github discussion: https://github.com/AutoGPTQ/AutoGPTQ/issues/621

pjw000

Apr 2, 2024

Thank you. That's a VERY helpful link! Much appreciated.

pjw000 changed discussion status to closed Apr 2, 2024

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