Views
No views yet
transformers, vLLM, sglang, llama.cpp, ollama or mlx library.1import torch
2from transformers import AutoModelForCausalLM, AutoTokenizer
3
4model_id = "tiiuae/Falcon-H1-Tiny-90M-Instruct-pre-DPO"
5
6model = AutoModelForCausalLM.from_pretrained(
7 model_id,
8 torch_dtype=torch.bfloat16,
9 device_map="auto"
10)
11
12# Perform text generationtransformers serve tiiuae/Falcon-H1-Tiny-90M-Instruct-pre-DPOllama.cppllama.cpp under our official collection - an example setup could be:1brew install llama.cpp
2pip install huggingface_hub
3hf download tiiuae/Falcon-H1-Tiny-90M-Instruct-pre-DPO Falcon-H1-Tiny-90M-Instruct-pre-DPO-Q8_0.gguf --local-dir ./
4llama-cli ./ Falcon-H1-Tiny-90M-Instruct-pre-DPO-Q8_0.gguf -cnv ollamaollama run hf.co/tiiuae/Falcon-H1-Tiny-90M-Instruct-pre-DPO:Q8_0 mlxmlx_lm.chat --model tiiuae/Falcon-H1-Tiny-90M-Instruct-pre-DPO1# pip install vllm>=0.9.0
2vllm serve tiiuae/Falcon-H1-Tiny-90M-Instruct-pre-DPO --tensor-parallel-size 2 --data-parallel-size 11python -m sglang.launch_server \
2 --model ttiiuae/Falcon-H1-Tiny-90M-Instruct-pre-DPO \
3 --tensor-parallel-size 1 @misc{falcon_h1_tiny,
title={Falcon-H1-Tiny: A series of extremely small, yet powerful language models redefining capabilities at small scale},
author={Falcon-LLM Team},
year={2026},
}