Instructions to use FasterDecoding/medusa-1.0-vicuna-33b-v1.3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use FasterDecoding/medusa-1.0-vicuna-33b-v1.3 with Transformers:

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

pipe = pipeline("text-generation", model="FasterDecoding/medusa-1.0-vicuna-33b-v1.3")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("FasterDecoding/medusa-1.0-vicuna-33b-v1.3")
model = AutoModelForCausalLM.from_pretrained("FasterDecoding/medusa-1.0-vicuna-33b-v1.3", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use FasterDecoding/medusa-1.0-vicuna-33b-v1.3 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "FasterDecoding/medusa-1.0-vicuna-33b-v1.3"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "FasterDecoding/medusa-1.0-vicuna-33b-v1.3",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/FasterDecoding/medusa-1.0-vicuna-33b-v1.3

SGLang

How to use FasterDecoding/medusa-1.0-vicuna-33b-v1.3 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 "FasterDecoding/medusa-1.0-vicuna-33b-v1.3" \
    --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": "FasterDecoding/medusa-1.0-vicuna-33b-v1.3",
		"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 "FasterDecoding/medusa-1.0-vicuna-33b-v1.3" \
        --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": "FasterDecoding/medusa-1.0-vicuna-33b-v1.3",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use FasterDecoding/medusa-1.0-vicuna-33b-v1.3 with Docker Model Runner:
```
docker model run hf.co/FasterDecoding/medusa-1.0-vicuna-33b-v1.3
```

medusa-1.0-vicuna-33b-v1.3 / pytorch_model-00006-of-00007.bin

tianlecai

Upload folder using huggingface_hub

10b04df over 2 years ago

Download

History Blame Contribute Delete

9.96 GB

This file is stored with Xet . It is too big to display, but you can still download it.

Xet Pointer Details

( Raw pointer file )

Xet hash:: 25dd2c0498851429feeab6734acb0549a7004c26e3a57ddad2b0e305c4d5cec3
Size of remote file:: 9.96 GB
SHA256:: 7a01022826e6007b5ece34985dc401eae1733148c9a3839e2d4250c82773095d

Xet efficiently stores Large Files inside Git, intelligently splitting files into unique chunks and accelerating uploads and downloads. More info.