Instructions to use Dawn-AI/Dawn-superlite with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Dawn-AI/Dawn-superlite with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="Dawn-AI/Dawn-superlite", trust_remote_code=True)# pip install -U transformers accelerate # Load model directly from transformers import AutoModelForCausalLM model = AutoModelForCausalLM.from_pretrained("Dawn-AI/Dawn-superlite", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use Dawn-AI/Dawn-superlite with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "Dawn-AI/Dawn-superlite" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Dawn-AI/Dawn-superlite", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/Dawn-AI/Dawn-superlite
- SGLang
How to use Dawn-AI/Dawn-superlite 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 "Dawn-AI/Dawn-superlite" \ --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": "Dawn-AI/Dawn-superlite", "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 "Dawn-AI/Dawn-superlite" \ --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": "Dawn-AI/Dawn-superlite", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use Dawn-AI/Dawn-superlite with Docker Model Runner:
docker model run hf.co/Dawn-AI/Dawn-superlite
Download setup.sh from Dawn-AI/Dawn-superlite: direct link, hf CLI and curl.
- Browser
- Download file 1.27 kB
-
https://huggingface.co/Dawn-AI/Dawn-superlite/resolve/main/setup.sh
- Command line
-
hf download hf://Dawn-AI/Dawn-superlite/setup.sh
-
curl -L -o setup.sh https://huggingface.co/Dawn-AI/Dawn-superlite/resolve/main/setup.sh
1.27 kB
| set -eo pipefail | |
| CONDA_ROOT="/lustre/fsw/portfolios/nvr/projects/nvr_lpr_llm/users/yongganf/anaconda" | |
| ENV_NAME="nemotron_flash" | |
| PYTHON_VERSION="3.10.12" | |
| source "${CONDA_ROOT}/etc/profile.d/conda.sh" | |
| if conda env list | awk '{print $1}' | grep -qx "${ENV_NAME}"; then | |
| echo "Conda env '${ENV_NAME}' already exists, skipping creation." | |
| else | |
| echo "Creating conda env '${ENV_NAME}' with Python ${PYTHON_VERSION}..." | |
| conda create -n "${ENV_NAME}" "python=${PYTHON_VERSION}" -y | |
| fi | |
| conda activate "${ENV_NAME}" | |
| echo "Installing CUDA toolkit 12.8..." | |
| conda install -c nvidia cuda-toolkit=12.8 -y | |
| export CUDA_HOME | |
| CUDA_HOME="$(dirname "$(dirname "$(which nvcc)")")" | |
| echo "CUDA_HOME set to ${CUDA_HOME}" | |
| echo "Installing PyTorch and core Python packages..." | |
| pip install torch==2.8.0 transformers==4.56.2 psutil | |
| echo "Installing causal-conv1d from source..." | |
| pip install --no-build-isolation causal-conv1d | |
| echo "Installing flash-attention from source..." | |
| pip install --no-build-isolation flash-attn==2.7.3 | |
| echo "Installing mamba-ssm from source..." | |
| pip install --no-build-isolation mamba-ssm | |
| echo "Installing flash-linear-attention..." | |
| pip install --no-build-isolation flash-linear-attention | |
| echo "Done! Environment '${ENV_NAME}' is ready." | |