File size: 7,673 Bytes
f7e1c38 17f0bdd f7e1c38 17f0bdd f7e1c38 796f921 f7e1c38 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 |
---
language:
- en
- fr
- es
- pt
tags:
- falcon3
base_model: tiiuae/Falcon3-1B-Base
license: other
license_name: falcon-llm-license
license_link: https://falconllm.tii.ae/falcon-terms-and-conditions.html
library_name: transformers
---
<div align="center">
<img src="https://huggingface.co/datasets/tiiuae/documentation-images/resolve/main/general/falco3-logo.png" alt="drawing" width="500"/>
</div>
# Falcon3-1B-Instruct
**Falcon3** family of Open Foundation Models is a set of pretrained and instruct LLMs ranging from 1B to 10B parameters.
This repository contains the **Falcon3-1B-Instruct**. It achieves strong results on reasoning, language understanding, instruction following, code and mathematics tasks.
Falcon3-1B-Instruct supports 4 languages (English, French, Spanish, Portuguese) and a context length of up to 8K.
## Model Details
- Architecture
- Transformer-based causal decoder-only architecture
- 18 decoder blocks
- Grouped Query Attention (GQA) for faster inference: 8 query heads and 4 key-value heads
- Wider head dimension: 256
- High RoPE value to support long context understanding: 1000042
- Uses SwiGLU and RMSNorm
- 8K context length
- 131K vocab size
- Pruned and healed using larger Falcon models (3B and 7B respectively) on only 80 Gigatokens of datasets comprising of web, code, STEM, high quality and multilingual data using 256 H100 GPU chips
- Posttrained on 1.2 million samples of STEM, conversational, code, safety and function call data
- Supports EN, FR, ES, PT
- Developed by [Technology Innovation Institute](https://www.tii.ae)
- License: TII Falcon-LLM License 2.0
- Model Release Date: December 2024
## Getting started
<details>
<summary> Click to expand </summary>
```python
from transformers import AutoTokenizer, AutoModelForCausalLM
model_name = "tiiuae/Falcon3-1B-Instruct"
model = AutoModelForCausalLM.from_pretrained(
model_name,
torch_dtype="auto",
device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained(model_name)
prompt = "How many hours in one day?"
messages = [
{"role": "system", "content": "You are a helpful friendly assistant Falcon3 from TII, try to follow instructions as much as possible."},
{"role": "user", "content": prompt}
]
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True
)
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
generated_ids = model.generate(
**model_inputs,
max_new_tokens=1024
)
generated_ids = [
output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
]
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
print(response)
```
</details>
<br>
## Benchmarks
We report in the following table our internal pipeline benchmarks.
- We use [lm-evaluation harness](https://github.com/EleutherAI/lm-evaluation-harness).
- We report **raw scores** obtained by applying chat template **without fewshot_as_multiturn** (unlike Llama3.1).
- We use same batch-size across all models.
<table border="1" style="width: 100%; text-align: center; border-collapse: collapse;">
<colgroup>
<col style="width: 10%;">
<col style="width: 10%;">
<col style="width: 7%;">
<col style="width: 7%;">
<col style="width: 7%;">
<col style="background-color: rgba(80, 15, 213, 0.5); width: 7%;">
</colgroup>
<thead>
<tr>
<th>Category</th>
<th>Benchmark</th>
<th>Llama-3.2-1B</th>
<th>Qwen2.5-1.5B</th>
<th>SmolLM2-1.7B</th>
<th>Falcon3-1B-Instruct</th>
</tr>
</thead>
<tbody>
<tr>
<td rowspan="3">General</td>
<td>MMLU (5-shot)</td>
<td>23.4</td>
<td><b>58.4</b></td>
<td>48.4</td>
<td>43.9</td>
</tr>
<tr>
<td>MMLU-PRO (5-shot)</td>
<td>11.3</td>
<td><b>21.3</b></td>
<td>17.2</td>
<td>18.6</td>
</tr>
<tr>
<td>IFEval</td>
<td><b>55.8</b></td>
<td>44.4</td>
<td>53.0</td>
<td>54.4</td>
</tr>
<tr>
<td rowspan="3">Math</td>
<td>GSM8K (5-shot)</td>
<td>37.4</td>
<td><b>57.2</b></td>
<td>43.4</td>
<td>38.6</td>
</tr>
<tr>
<td>GSM8K (8-shot, COT)</td>
<td>35.6</td>
<td><b>62.2</b></td>
<td>47.2</td>
<td>41.8</td>
</tr>
<tr>
<td>MATH Lvl-5 (4-shot)</td>
<td><b>3.9</b></td>
<td>0.2</td>
<td>0.1</td>
<td>1.0</td>
</tr>
<tr>
<td rowspan="6">Reasoning</td>
<td>Arc Challenge (25-shot)</td>
<td>34.1</td>
<td>47.0</td>
<td><b>47.6</b></td>
<td>45.9</td>
</tr>
<tr>
<td>GPQA (0-shot)</td>
<td>25.3</td>
<td><b>29.6</b></td>
<td>28.7</td>
<td>26.5</td>
</tr>
<tr>
<td>GPQA (0-shot, COT)</td>
<td>13.2</td>
<td>9.2</td>
<td>16.0</td>
<td><b>21.3</b></td>
</tr>
<tr>
<td>MUSR (0-shot)</td>
<td>32.4</td>
<td>36.8</td>
<td>33.0</td>
<td><b>40.7</b></td>
</tr>
<tr>
<td>BBH (3-shot)</td>
<td>30.3</td>
<td><b>38.5</b></td>
<td>33.1</td>
<td>35.1</td>
</tr>
<tr>
<td>BBH (3-shot, COT)</td>
<td>0.0</td>
<td>20.3</td>
<td>0.8</td>
<td><b>30.5</b></td>
</tr>
<tr>
<td rowspan="5">CommonSense Understanding</td>
<td>PIQA (0-shot)</td>
<td>72.1</td>
<td>73.2</td>
<td><b>74.4</b></td>
<td>72.0</td>
</tr>
<tr>
<td>SciQ (0-shot)</td>
<td>61.8</td>
<td>69.5</td>
<td>71.4</td>
<td><b>86.8</b></td>
</tr>
<tr>
<td>Winogrande (0-shot)</td>
<td>-</td>
<td>-</td>
<td>-</td>
<td><b>60.2</b></td>
</tr>
<tr>
<td>OpenbookQA (0-shot)</td>
<td>40.2</td>
<td>40.4</td>
<td><b>42.8</b></td>
<td>40.0</td>
</tr>
<tr>
<td>MT-Bench (avg)</td>
<td>5.4</td>
<td><b>7.1</b></td>
<td>6.1</td>
<td>5.5</td>
</tr>
<tr>
<td rowspan="1">Instructions following</td>
<td>Alpaca (WC)</td>
<td><b>8.6</b></td>
<td><b>8.6</b></td>
<td>5.4</td>
<td>6.1</td>
</tr>
</tbody>
</table>
## Useful links
- View our [release blogpost](https://huggingface.co/blog/falcon3).
- Feel free to join [our discord server](https://discord.gg/fwXpMyGc) if you have any questions or to interact with our researchers and developers.
## Technical Report
Coming soon....
## Citation
If the Falcon3 family of models were helpful to your work, feel free to give us a cite.
```
@misc{Falcon3,
title = {The Falcon 3 Family of Open Models},
url = {https://huggingface.co/blog/falcon3},
author = {Falcon-LLM Team},
month = {December},
year = {2024}
}
``` |