Instructions to use ISTA-DASLab/Meta-Llama-3.1-70B-AQLM-PV-2Bit-1x16 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ISTA-DASLab/Meta-Llama-3.1-70B-AQLM-PV-2Bit-1x16 with Transformers:

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

pipe = pipeline("text-generation", model="ISTA-DASLab/Meta-Llama-3.1-70B-AQLM-PV-2Bit-1x16")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("ISTA-DASLab/Meta-Llama-3.1-70B-AQLM-PV-2Bit-1x16")
model = AutoModelForCausalLM.from_pretrained("ISTA-DASLab/Meta-Llama-3.1-70B-AQLM-PV-2Bit-1x16")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use ISTA-DASLab/Meta-Llama-3.1-70B-AQLM-PV-2Bit-1x16 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ISTA-DASLab/Meta-Llama-3.1-70B-AQLM-PV-2Bit-1x16"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ISTA-DASLab/Meta-Llama-3.1-70B-AQLM-PV-2Bit-1x16",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/ISTA-DASLab/Meta-Llama-3.1-70B-AQLM-PV-2Bit-1x16

SGLang

How to use ISTA-DASLab/Meta-Llama-3.1-70B-AQLM-PV-2Bit-1x16 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 "ISTA-DASLab/Meta-Llama-3.1-70B-AQLM-PV-2Bit-1x16" \
    --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": "ISTA-DASLab/Meta-Llama-3.1-70B-AQLM-PV-2Bit-1x16",
		"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 "ISTA-DASLab/Meta-Llama-3.1-70B-AQLM-PV-2Bit-1x16" \
        --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": "ISTA-DASLab/Meta-Llama-3.1-70B-AQLM-PV-2Bit-1x16",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use ISTA-DASLab/Meta-Llama-3.1-70B-AQLM-PV-2Bit-1x16 with Docker Model Runner:
```
docker model run hf.co/ISTA-DASLab/Meta-Llama-3.1-70B-AQLM-PV-2Bit-1x16
```

Official AQLM quantization of meta-llama/Meta-Llama-3.1-70B finetuned with PV-Tuning.

For this quantization, we used 1 codebook of 16 bits and groupsize of 8.

Results:

Model	Quantization	MMLU (5-shot)	ArcC	ArcE	Hellaswag	PiQA	Winogrande	Model size, Gb
	fp16	0.7839	0.6058	0.8729	0.6650	0.8292	0.7964	141
	1x16g8	0.7353	0.4556	0.7731	0.6335	0.7590	0.7703	21.9

Downloads last month: 11

Safetensors

Model size

11B params

Tensor type

F16

I16

Model tree for ISTA-DASLab/Meta-Llama-3.1-70B-AQLM-PV-2Bit-1x16

Base model

meta-llama/Llama-3.1-70B

Quantized

(79)

this model

Space using ISTA-DASLab/Meta-Llama-3.1-70B-AQLM-PV-2Bit-1x16 1

Papers for ISTA-DASLab/Meta-Llama-3.1-70B-AQLM-PV-2Bit-1x16

PV-Tuning: Beyond Straight-Through Estimation for Extreme LLM Compression

Paper • 2405.14852 • Published May 23, 2024 • 2

Extreme Compression of Large Language Models via Additive Quantization

Paper • 2401.06118 • Published Jan 11, 2024 • 14