Text Generation
Transformers
Safetensors
English
gemma
jepa
world-models
omnimodal
arc-challenge
mmlu
gsm8k
image-generation
video-generation
audio-generation
Mixture of Experts
sparse-moe
punica
dag-reasoning
compiler-safety
os-computer-use
casp15
structural-biology
custom_code
text-generation-inference
Instructions to use clevrpwn/gmma-jepa with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use clevrpwn/gmma-jepa with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="clevrpwn/gmma-jepa", trust_remote_code=True)# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("clevrpwn/gmma-jepa", trust_remote_code=True) model = AutoModelForCausalLM.from_pretrained("clevrpwn/gmma-jepa", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use clevrpwn/gmma-jepa with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "clevrpwn/gmma-jepa" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "clevrpwn/gmma-jepa", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/clevrpwn/gmma-jepa
- SGLang
How to use clevrpwn/gmma-jepa 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 "clevrpwn/gmma-jepa" \ --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": "clevrpwn/gmma-jepa", "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 "clevrpwn/gmma-jepa" \ --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": "clevrpwn/gmma-jepa", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use clevrpwn/gmma-jepa with Docker Model Runner:
docker model run hf.co/clevrpwn/gmma-jepa
| { | |
| "backend": "tokenizers", | |
| "bos_token": "<bos>", | |
| "clean_up_tokenization_spaces": false, | |
| "eos_token": "<eos>", | |
| "extra_special_tokens": [ | |
| "<start_of_turn>", | |
| "<end_of_turn>" | |
| ], | |
| "is_local": false, | |
| "local_files_only": false, | |
| "mask_token": "<mask>", | |
| "model_max_length": 1000000000000000019884624838656, | |
| "pad_token": "<pad>", | |
| "sp_model_kwargs": {}, | |
| "spaces_between_special_tokens": false, | |
| "tokenizer_class": "GemmaTokenizer", | |
| "unk_token": "<unk>", | |
| "use_default_system_prompt": false | |
| } | |