Instructions to use Duke-CEI-SVD/Llama-3.1-8B-Dense-Basis with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Duke-CEI-SVD/Llama-3.1-8B-Dense-Basis with Transformers:

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

pipe = pipeline("text-generation", model="Duke-CEI-SVD/Llama-3.1-8B-Dense-Basis")

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Duke-CEI-SVD/Llama-3.1-8B-Dense-Basis")
model = AutoModelForCausalLM.from_pretrained("Duke-CEI-SVD/Llama-3.1-8B-Dense-Basis", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Duke-CEI-SVD/Llama-3.1-8B-Dense-Basis with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Duke-CEI-SVD/Llama-3.1-8B-Dense-Basis"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Duke-CEI-SVD/Llama-3.1-8B-Dense-Basis",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/Duke-CEI-SVD/Llama-3.1-8B-Dense-Basis

SGLang

How to use Duke-CEI-SVD/Llama-3.1-8B-Dense-Basis 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 "Duke-CEI-SVD/Llama-3.1-8B-Dense-Basis" \
    --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": "Duke-CEI-SVD/Llama-3.1-8B-Dense-Basis",
		"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 "Duke-CEI-SVD/Llama-3.1-8B-Dense-Basis" \
        --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": "Duke-CEI-SVD/Llama-3.1-8B-Dense-Basis",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use Duke-CEI-SVD/Llama-3.1-8B-Dense-Basis with Docker Model Runner:
```
docker model run hf.co/Duke-CEI-SVD/Llama-3.1-8B-Dense-Basis
```

You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Llama 3.1 8B Dense Reference (FP16)

Built with Llama.

This is the dense reference model used with the BASIS research project. It is an FP16 reserialization of the Meta Llama 3.1 8B pretrained base model at revision d04e592bb4f6aa9cfee91e2e20afa771667e1d4b. The source weights were BF16. This export converts them to FP16; it does not apply BASIS compression, recovery, or additional training. Numerical results can differ from the BF16 source.

The repository includes a complete Transformers configuration, tokenizer, and sharded safetensors weights. Load it with:

from transformers import AutoModelForCausalLM, AutoTokenizer

repo = "Duke-CEI-SVD/Llama-3.1-8B-Dense-Basis"
tokenizer = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(repo, dtype="float16", device_map="auto")

Use is subject to the included Llama 3.1 Community License and Meta's Acceptable Use Policy. The required attribution is in NOTICE.

Downloads last month: 8

Safetensors

Model size

8B params

Tensor type

F16

Model tree for Duke-CEI-SVD/Llama-3.1-8B-Dense-Basis

Base model

meta-llama/Llama-3.1-8B

Finetuned

(1498)

this model