Instructions to use mente-ai/uyu-1-10M with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use mente-ai/uyu-1-10M with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="mente-ai/uyu-1-10M")# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("mente-ai/uyu-1-10M") model = AutoModelForCausalLM.from_pretrained("mente-ai/uyu-1-10M") - Notebooks
- Google Colab
- Kaggle
- Local Apps
- vLLM
How to use mente-ai/uyu-1-10M with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "mente-ai/uyu-1-10M" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "mente-ai/uyu-1-10M", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/mente-ai/uyu-1-10M
- SGLang
How to use mente-ai/uyu-1-10M 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 "mente-ai/uyu-1-10M" \ --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": "mente-ai/uyu-1-10M", "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 "mente-ai/uyu-1-10M" \ --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": "mente-ai/uyu-1-10M", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use mente-ai/uyu-1-10M with Docker Model Runner:
docker model run hf.co/mente-ai/uyu-1-10M
uyu
This is a nanoGPT checkpoint converted to a Hugging Face GPT-2-compatible
GPT2LMHeadModel.
The model weights load with:
from transformers import AutoTokenizer, GPT2LMHeadModel, pipeline
model = GPT2LMHeadModel.from_pretrained(".")
tokenizer = AutoTokenizer.from_pretrained(".", trust_remote_code=True)
pipe = pipeline(
"text-generation",
model="mente-ai/uyu-1-10M",
trust_remote_code=True,
)
Use eos_token_id=3 during generation to stop at <STORY_END>.
The tokenizer is a SentencePiece model stored as uyu.model. It is not a
standard GPT-2 byte-level BPE tokenizer.
Original checkpoint: ckpt.pt
- Downloads last month
- 107