Instructions to use helloollel/vicuna-13b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use helloollel/vicuna-13b with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="helloollel/vicuna-13b")# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("helloollel/vicuna-13b") model = AutoModelForCausalLM.from_pretrained("helloollel/vicuna-13b") - Notebooks
- Google Colab
- Kaggle
- Local Apps
- vLLM
How to use helloollel/vicuna-13b with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "helloollel/vicuna-13b" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "helloollel/vicuna-13b", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/helloollel/vicuna-13b
- SGLang
How to use helloollel/vicuna-13b 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 "helloollel/vicuna-13b" \ --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": "helloollel/vicuna-13b", "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 "helloollel/vicuna-13b" \ --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": "helloollel/vicuna-13b", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use helloollel/vicuna-13b with Docker Model Runner:
docker model run hf.co/helloollel/vicuna-13b
notebook script fails on LambdaLabs cloud
running python 3.8.x
Complains about versionitis deep down in tensorboard.protobuf.
Do I just need a later python, then reinstall everything?
Well, try a newer version please. The more recent it is, the better it will be.
Ok, but turns out lambdacloud is a pain to use unless you want to keep the instance running 24/7
Do you think I can run 13B on a pair of K80s? each one has 24GB and 2 gpus. They are pretty cheap (~$100 on ebay), appreciate your thoughts.
Yes, you can. Additionally, ensure that there is at least 24G of free memory. To lower the memory cost, use the following command:
python3 -m fastchat.serve.cli --model-name /content/drive/MyDrive/AI/fastchat/vicuna-13b --load-8bit