Instructions to use distilbert/distilgpt2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use distilbert/distilgpt2 with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="distilbert/distilgpt2")# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("distilbert/distilgpt2") model = AutoModelForCausalLM.from_pretrained("distilbert/distilgpt2") - Notebooks
- Google Colab
- Kaggle
- Local Apps
- vLLM
How to use distilbert/distilgpt2 with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "distilbert/distilgpt2" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "distilbert/distilgpt2", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/distilbert/distilgpt2
- SGLang
How to use distilbert/distilgpt2 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 "distilbert/distilgpt2" \ --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": "distilbert/distilgpt2", "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 "distilbert/distilgpt2" \ --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": "distilbert/distilgpt2", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use distilbert/distilgpt2 with Docker Model Runner:
docker model run hf.co/distilbert/distilgpt2
New model card for distilgpt2
Expanded model card for distilgpt2, including information about training, use, limitations, emissions, and evaluation. Also added metadata for evaluation results
That looks great to me!
Thanks a lot @Marissa
What does the author think @VictorSanh - good for merge?
LGTM too!
Alright, let's merge it - hope that's fine for you @VictorSanh ! Otherwise we can still revert afterwards (#feature request haha @julien-c ) :-)
lgtm!
only thing i would note: **Hours used:** 8 -> it's actually in the order of magnitude of 1 week on a node with 8 16GB v100
would you like to change that @Marissa ? :)
can we wait for a little bit more feedback from the team before merging, next time?
In particular, not a huge fan of the collapsible sections UX-wise. What do others think?
I don't like collapsable sections either: they can be used for certain very long sections but not everywhere like this and the main content should not be collapsed.
Sounds good, so general review policy might help here @julien-c
thanks for the feedback -- will create a new PR with those changes!
Sounds good!