Instructions to use bunsenfeng/PFA_switcher_2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use bunsenfeng/PFA_switcher_2 with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="bunsenfeng/PFA_switcher_2")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("bunsenfeng/PFA_switcher_2", dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use bunsenfeng/PFA_switcher_2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "bunsenfeng/PFA_switcher_2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "bunsenfeng/PFA_switcher_2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/bunsenfeng/PFA_switcher_2

SGLang

How to use bunsenfeng/PFA_switcher_2 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 "bunsenfeng/PFA_switcher_2" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "bunsenfeng/PFA_switcher_2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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 "bunsenfeng/PFA_switcher_2" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "bunsenfeng/PFA_switcher_2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use bunsenfeng/PFA_switcher_2 with Docker Model Runner:
```
docker model run hf.co/bunsenfeng/PFA_switcher_2
```

Improve model card: Add paper details, metadata, and usage for Switch Generation

by nielsr HF Staff - opened Oct 17, 2025

base: refs/heads/main

←

from: refs/pr/1

Discussion Files changed

+65

-150

nielsr

Oct 17, 2025

This PR significantly improves the model card for the Switch Generation model.

Key changes include:

Updating the metadata with pipeline_tag: text-generation and license: apache-2.0 for better discoverability and clarity. The library_name: transformers has been confirmed based on file evidence (tokenizer_config.json, generation_config.json, adapter_config.json).
Replacing placeholder content with a detailed model description derived from the paper's abstract.
Adding direct links to the research paper (Don't Throw Away Your Pretrained Model) and the GitHub repository (BunsenFeng/switch_generation).
Including the "Quick Start" code snippet directly from the GitHub README to provide immediate, actionable sample usage instructions.
Adding a proper BibTeX citation for the paper.

These updates will greatly enhance the model's visibility, provide essential context, and improve the user experience on the Hugging Face Hub.

Improve model card: Add paper details, metadata, and usage for Switch Generationdb6532b5

bunsenfeng changed pull request status to merged Oct 17, 2025

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment