Text Generation
Transformers
PyTorch
Safetensors
English
qed
causal-lm
decoder-only
rope
rmsnorm
swiglu
custom-architecture
custom_code
Instructions to use levossadtchi/QED-75M with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use levossadtchi/QED-75M with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="levossadtchi/QED-75M", trust_remote_code=True)# Load model directly from transformers import AutoModelForCausalLM model = AutoModelForCausalLM.from_pretrained("levossadtchi/QED-75M", trust_remote_code=True, dtype="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps
- vLLM
How to use levossadtchi/QED-75M with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "levossadtchi/QED-75M" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "levossadtchi/QED-75M", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/levossadtchi/QED-75M
- SGLang
How to use levossadtchi/QED-75M 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 "levossadtchi/QED-75M" \ --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": "levossadtchi/QED-75M", "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 "levossadtchi/QED-75M" \ --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": "levossadtchi/QED-75M", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use levossadtchi/QED-75M with Docker Model Runner:
docker model run hf.co/levossadtchi/QED-75M
| #!/usr/bin/env python3 | |
| """ | |
| Example: generate text from QED-75M on Hugging Face. | |
| Run: | |
| python generate_gravity_example.py | |
| """ | |
| from __future__ import annotations | |
| import torch | |
| from transformers import AutoModelForCausalLM, AutoTokenizer | |
| def main() -> None: | |
| repo_id = "levossadtchi/QED-75M" | |
| prompt = "Explain gravity in one sentence. \n<|assistant|>" | |
| # trust_remote_code=True is required because QED is a custom architecture. | |
| tokenizer = AutoTokenizer.from_pretrained(repo_id, trust_remote_code=True) | |
| model = AutoModelForCausalLM.from_pretrained( | |
| repo_id, | |
| trust_remote_code=True, | |
| torch_dtype=torch.float32, | |
| ) | |
| model.eval() | |
| device = "cuda" if torch.cuda.is_available() else "cpu" | |
| model.to(device) | |
| inputs = tokenizer(prompt, return_tensors="pt", add_special_tokens=False).to(device) | |
| with torch.no_grad(): | |
| out_ids = model.generate( | |
| **inputs, | |
| max_new_tokens=64, | |
| do_sample=True, | |
| temperature=0.8, | |
| top_k=50, | |
| eos_token_id=tokenizer.eos_token_id, | |
| pad_token_id=tokenizer.pad_token_id, | |
| ) | |
| text = tokenizer.decode(out_ids[0], skip_special_tokens=True) | |
| print(text) | |
| if __name__ == "__main__": | |
| main() | |