Instructions to use Faradaylab/ARIA-70B-V3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Faradaylab/ARIA-70B-V3 with Transformers:

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

pipe = pipeline("text-generation", model="Faradaylab/ARIA-70B-V3")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Faradaylab/ARIA-70B-V3")
model = AutoModelForCausalLM.from_pretrained("Faradaylab/ARIA-70B-V3")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Faradaylab/ARIA-70B-V3 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Faradaylab/ARIA-70B-V3"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Faradaylab/ARIA-70B-V3",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/Faradaylab/ARIA-70B-V3

SGLang

How to use Faradaylab/ARIA-70B-V3 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 "Faradaylab/ARIA-70B-V3" \
    --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": "Faradaylab/ARIA-70B-V3",
		"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 "Faradaylab/ARIA-70B-V3" \
        --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": "Faradaylab/ARIA-70B-V3",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use Faradaylab/ARIA-70B-V3 with Docker Model Runner:
```
docker model run hf.co/Faradaylab/ARIA-70B-V3
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

ARIA V3 has been trained over 100.000 high quality french language with a focus on data bias, grammar and overall language/writing capacities of the model.

The training has been done on Nvidia GPU in the cloud with Amazon Sagemaker.

Base Model : Llama2-70B-Chat-HF

Dataset : private dataset.

Added value : French Language / Writing / Content Creation / Data bias reduction

Feel free to reach out to us ! [email protected]

Downloads last month: 291

Safetensors

Model size

69B params

Tensor type

F16

F32

Model tree for Faradaylab/ARIA-70B-V3

Quantizations

1 model