|
|
--- |
|
|
library_name: transformers |
|
|
tags: |
|
|
- falcon-h1 |
|
|
license: other |
|
|
license_name: falcon-llm-license |
|
|
license_link: https://falconllm.tii.ae/falcon-terms-and-conditions.html |
|
|
--- |
|
|
|
|
|
# Table of Contents |
|
|
|
|
|
0. [TL;DR](#TL;DR) |
|
|
1. [Model Details](#model-details) |
|
|
2. [Training Details](#training-details) |
|
|
3. [Usage](#usage) |
|
|
4. [Evaluation](#evaluation) |
|
|
5. [Citation](#citation) |
|
|
|
|
|
# TL;DR |
|
|
|
|
|
# Model Details |
|
|
|
|
|
## Model Description |
|
|
|
|
|
- **Developed by:** [https://www.tii.ae](https://www.tii.ae) |
|
|
- **Model type:** Causal decoder-only |
|
|
- **Architecture:** Hybrid Transformers + Mamba architecture |
|
|
- **Language(s) (NLP):** English, Multilingual |
|
|
- **License:** Falcon-LLM License |
|
|
|
|
|
# Training details |
|
|
|
|
|
For more details about the training protocol of this model, please refer to the [Falcon-H1 technical blogpost](https://falcon-lm.github.io/blog/falcon-h1/). |
|
|
|
|
|
# Usage |
|
|
|
|
|
Currently to use this model you can either rely on Hugging Face `transformers`, `vLLM` or our custom fork of `llama.cpp` library. |
|
|
|
|
|
## Inference |
|
|
|
|
|
Make sure to install the latest version of `transformers` or `vllm`, eventually install these packages from source: |
|
|
|
|
|
```bash |
|
|
pip install git+https://github.com/huggingface/transformers.git |
|
|
``` |
|
|
|
|
|
Refer to [the official vLLM documentation for more details on building vLLM from source](https://docs.vllm.ai/en/latest/getting_started/installation/gpu.html#build-wheel-from-source). |
|
|
|
|
|
### 🤗 transformers |
|
|
|
|
|
Refer to the snippet below to run H1 models using 🤗 transformers: |
|
|
|
|
|
```python |
|
|
import torch |
|
|
from transformers import AutoModelForCausalLM, AutoTokenizer |
|
|
|
|
|
model_id = "tiiuae/Falcon-H1-1B-Base" |
|
|
|
|
|
model = AutoModelForCausalLM.from_pretrained( |
|
|
model_id, |
|
|
torch_dtype=torch.bfloat16, |
|
|
device_map="auto" |
|
|
) |
|
|
|
|
|
# Perform text generation |
|
|
``` |
|
|
|
|
|
### vLLM |
|
|
|
|
|
For vLLM, simply start a server by executing the command below: |
|
|
|
|
|
``` |
|
|
# pip install vllm |
|
|
vllm serve tiiuae/Falcon-H1-1B-Instruct --tensor-parallel-size 2 --data-parallel-size 1 |
|
|
``` |
|
|
|
|
|
### `llama.cpp` |
|
|
|
|
|
While we are working on integrating our architecture directly into `llama.cpp` library, you can install our fork of the library and use it directly: https://github.com/tiiuae/llama.cpp-Falcon-H1 |
|
|
Use the same installing guidelines as `llama.cpp`. |
|
|
|
|
|
# Evaluation |
|
|
|
|
|
Falcon-H1 series perform very well on a variety of tasks, including reasoning tasks. |
|
|
|
|
|
| Tasks | Falcon-H1-34B | Qwen2.5-72B | Qwen2.5-32B | Gemma3-27B | Llama3.1-70B | Llama4-scout | |
|
|
| --- | --- | --- | --- | --- | --- | --- | |
|
|
| **General** | | | | | | |
|
|
| BBH | **69.36** | 67.77 | 67.45 | 61.6 | 62.78 | 61.71 | |
|
|
| MMLU | 83.46 | **85.96** | 83.18 | 78.32 | 78.49 | 77.98 | |
|
|
| ARC-C | 71.25 | **72.44** | 70.48 | 70.31 | 69.2 | 62.97 | |
|
|
| HellaSwag | 85.68 | 87.57 | 85.13 | 86.19 | **87.78** | 84.01 | |
|
|
| Winogrande | 82.72 | 83.74 | 82.32 | 82.4 | **85.32** | 78.93 | |
|
|
| **Math** | | | | | | |
|
|
| GSM8k | 76.5 | 89.76 | **90.14** | 81.35 | 80.52 | 83.24 | |
|
|
| MATH lvl5 | **40.71** | 38.14 | 36.4 | 25.38 | 18.81 | 27.19 | |
|
|
| **Science** | | | | | | |
|
|
| GPQA | **42.7** | 42.28 | 39.68 | 35.82 | 36.49 | 35.99 | |
|
|
| MMLU-Pro | 57.18 | **60.22** | 58.05 | 49.64 | 47.07 | 50.16 | |
|
|
| MMLU-stem | 83.82 | **84.81** | 82.81 | 76.59 | 70.35 | 72.57 | |
|
|
| **Code** | | | | | | |
|
|
| HumanEval | **70.12** | 59.15 | 59.76 | 48.78 | 57.32 | 57.32 | |
|
|
| HumanEval+ | **64.63** | 51.22 | 51.83 | 40.85 | 50.61 | 48.78 | |
|
|
| MBPP | 83.33 | **87.04** | 83.07 | 76.19 | 78.84 | 77.78 | |
|
|
| MBPP+ | 70.37 | **70.63** | 68.78 | 61.64 | 66.67 | 64.29 | |
|
|
|
|
|
You can check more in detail on our [our release blogpost](https://falcon-lm.github.io/blog/falcon-h1/), detailed benchmarks. |
|
|
|
|
|
# Useful links |
|
|
|
|
|
- View [our release blogpost](https://falcon-lm.github.io/blog/falcon-h1/). |
|
|
- Feel free to join [our discord server](https://discord.gg/fwXpMyGc) if you have any questions or to interact with our researchers and developers. |
|
|
|
|
|
# Citation |
|
|
|
|
|
If the Falcon-H1 family of models were helpful to your work, feel free to give us a cite. |
|
|
|
|
|
``` |
|
|
@misc{tiifalconh1, |
|
|
title = {Falcon-H1: A Family of Hybrid-Head Language Models Redefining Efficiency and Performance}, |
|
|
url = {https://falcon-lm.github.io/blog/falcon-h1}, |
|
|
author = {Falcon-LLM Team}, |
|
|
month = {May}, |
|
|
year = {2025} |
|
|
} |
|
|
``` |
