Instructions to use mlabonne/AlphaMonarch-7B-2bit-HQQ with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use mlabonne/AlphaMonarch-7B-2bit-HQQ with Transformers:

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

pipe = pipeline("text-generation", model="mlabonne/AlphaMonarch-7B-2bit-HQQ")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("mlabonne/AlphaMonarch-7B-2bit-HQQ")
model = AutoModelForCausalLM.from_pretrained("mlabonne/AlphaMonarch-7B-2bit-HQQ")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use mlabonne/AlphaMonarch-7B-2bit-HQQ with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "mlabonne/AlphaMonarch-7B-2bit-HQQ"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "mlabonne/AlphaMonarch-7B-2bit-HQQ",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/mlabonne/AlphaMonarch-7B-2bit-HQQ

SGLang

How to use mlabonne/AlphaMonarch-7B-2bit-HQQ 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 "mlabonne/AlphaMonarch-7B-2bit-HQQ" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "mlabonne/AlphaMonarch-7B-2bit-HQQ",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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 "mlabonne/AlphaMonarch-7B-2bit-HQQ" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "mlabonne/AlphaMonarch-7B-2bit-HQQ",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use mlabonne/AlphaMonarch-7B-2bit-HQQ with Docker Model Runner:
```
docker model run hf.co/mlabonne/AlphaMonarch-7B-2bit-HQQ
```

Amazing model

by deleted - opened Mar 30, 2024

Discussion

deleted

Mar 30, 2024

Hello,
This is mind-blowing! It is only 2.43 GB in size!
Is the token embedding matrix and attention layers weights also in 2-bit?
Also Java and C is Turing-complete, so, in theory, if reimplement forward pass of this model, process of loading weights and tokenizer in one of these languages, then it is probably possible to run this model on a smartphone with very optimized Android system and no other apps open.

mlabonne

Owner Mar 31, 2024

Thanks @CatUkraine! No, as far as I know, only the linear layers are replaced using HQQ.

deleted

Apr 1, 2024

Thank you for response @mlabonne ! I am starting to understand rotary positional encoding and attention, and it is not as hard as i expected. I am going to port your model(and probably some other, smaller models) to some platforms and my devices.

Kalemnor

Apr 2, 2024

Thanks @CatUkraine! No, as far as I know, only the linear layers are replaced using HQQ.

Any chance to see a HQQ+ version of this?

mlabonne

Owner Apr 2, 2024

@Kalemnor No plans at the moment but it would be fun!

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