Text Generation
Transformers
PyTorch
Safetensors
English
falcon
custom_code
text-generation-inference
Instructions to use tiiuae/falcon-7b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use tiiuae/falcon-7b with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="tiiuae/falcon-7b", trust_remote_code=True)# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("tiiuae/falcon-7b", trust_remote_code=True) model = AutoModelForCausalLM.from_pretrained("tiiuae/falcon-7b", trust_remote_code=True) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use tiiuae/falcon-7b with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "tiiuae/falcon-7b" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "tiiuae/falcon-7b", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/tiiuae/falcon-7b
- SGLang
How to use tiiuae/falcon-7b 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 "tiiuae/falcon-7b" \ --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": "tiiuae/falcon-7b", "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 "tiiuae/falcon-7b" \ --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": "tiiuae/falcon-7b", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Docker Model Runner
How to use tiiuae/falcon-7b with Docker Model Runner:
docker model run hf.co/tiiuae/falcon-7b
Anyone discovered "Mini" yet in prompting?
#83
by YoYo1234Qwerty - opened
It seems like it's still a mystery what dataset TII used, and how to prompt this model. I've noticed from experiments that it's eager to output a format like:
User How are you
Mini Fine thanks
For instance:
User Nice to meet you. How's the weather?
Mini The weather is actually terrible, with rain and thunderstorms predicted. Are you and your friends still planning on seeing a movie together tonight?
User Oh wow, we might have umbrellas handy then.... So have you seen any good movies recently?
Mini As an AI I'm not capable off `watching` movies per-said, but are some favorites include Titanic 1922, Interstellar, Avatar, and Avengers movies.
I've seen in the tokenizer things like >>QUESTION<< and >>ANSWER<<, but it's not clear if they just ripped off this tokenizer from somewhere, and never actually trained it like this.
Other users suggest this format, but I've never seen it verified:
User: blah
Assistant: blah
Anyone have any insights into the Instruct dataset, or how this was intended to be prompted?