File size: 8,069 Bytes
2b69b8f |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 |
---
language:
- en
tags:
- falcon3
---
# Table of Contents
0. [TL;DR](#TL;DR)
1. [Model Details](#model-details)
2. [Usage](#usage)
3. [Training Details](#training-details)
4. [Evaluation](#evaluation)
# TL;DR
Falcon 3 family of Open Foundation Models is a set of pretrained and instruct LLMs ranging from 1B to 10B.
This repository contains the Falcon3-7B-Instruct, the best Instruct LLM under 8B at the time of release.
# Model Details
## Model Description
- **Developed by:** [https://www.tii.ae](https://www.tii.ae)
- **Model type:** Causal decoder-only
- **Architecture:** Transformer-base
- **Language(s) (NLP):** Mainly English
- **License:** TII Falcon-LLM License 2.0
<br>
# Usage
Find below an example on how to use the model in `transformers` (Make sure to have the latest transformers, or the one built from source):
<details>
<summary> Click to expand </summary>
```python
from transformers import AutoTokenizer, AutoModelForCausalLM
from transformers import AutoModelForCausalLM, AutoTokenizer
model_name = "tiiuae/Falcon3-7B-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>
# Training Details
Based on `tiiuae/Falcon3-7B-Base`, post-training stage is comprised of supervised finetuning followed by human preference alignement (DPO).
## Supervised finetuning
### Training Data
1.2 million diverse, high-quality samples Tulu-3, Open-Hermes, Numina an Apigen.
| Data type | ratio |
|--------------------------------------|-------|
| Conversations | 32% |
| STEM | 32% |
| Code | 12% |
| Safety | 9.1% |
| Multi lingual | 8.3% |
| Function call | 3.3% |
| NLP (summarization, generation, QA) | 3.2% |
#### Training Hyperparameters
<style type="text/css">
.tg {border-collapse:collapse;border-spacing:0;}
.tg td{border-color:black;border-style:solid;border-width:1px;font-family:Arial, sans-serif;font-size:14px;
overflow:hidden;padding:10px 5px;word-break:normal;}
.tg th{border-color:black;border-style:solid;border-width:1px;font-family:Arial, sans-serif;font-size:14px;
font-weight:normal;overflow:hidden;padding:10px 5px;word-break:normal;}
.tg .tg-c3ow{border-color:inherit;text-align:center;vertical-align:top}
.tg .tg-7btt{border-color:inherit;font-weight:bold;text-align:center;vertical-align:top}
.tg .tg-0pky{border-color:inherit;text-align:left;vertical-align:top}
.tg .tg-ihkz{border-color:inherit;text-align:center;vertical-align:top}
.tg .tg-pcvp{border-color:inherit;text-align:left;vertical-align:top}
.tg .tg-j2vi{border-color:inherit;font-weight:bold;text-align:center;vertical-align:top}
.tg .tg-amwm{border-color:inherit;text-align:left;vertical-align:top}
.tg .tg-0lax{border-color:inherit;font-weight:bold;text-align:center;vertical-align:top}
</style>
<table class="tg"><thead>
<tr>
<th class="tg-7btt" rowspan="3">AdamW</th>
<th class="tg-c3ow">β1</th>
<th class="tg-0pky">0.9</th>
</tr>
<tr>
<th class="tg-ihkz">β2</th>
<th class="tg-pcvp">0.999</th>
</tr>
<tr>
<th class="tg-c3ow">weight decay</th>
<th class="tg-0pky">0.01</th>
</tr></thead>
<tbody>
<tr>
<td class="tg-j2vi" rowspan="4">Learning rate</td>
<td class="tg-ihkz">type</td>
<td class="tg-pcvp">linear decay</td>
</tr>
<tr>
<td class="tg-c3ow">init lr</td>
<td class="tg-0pky">5e-6</td>
</tr>
<tr>
<td class="tg-ihkz">final lr</td>
<td class="tg-pcvp">0</td>
</tr>
<tr>
<td class="tg-c3ow">warm rate</td>
<td class="tg-0pky">0.03</td>
</tr>
<tr>
<td class="tg-j2vi">Batch size</td>
<td class="tg-ihkz"></td>
<td class="tg-pcvp">64</td>
</tr>
<tr>
<td class="tg-amwm">Epochs</td>
<td class="tg-0lax"></td>
<td class="tg-0lax">2</td>
</tr>
</tbody>
</table>
## Human preference alignment - DPO
### Training Data
TO DO DO DO DO
#### Training Hyperparameters
TODODODODOD
# Evaluation
We report in the following table our internal pipeline benchmarks:
<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-3.1-8B-Instruct</th>
<th>Qwen2-7B-Instruct</th>
<th>Qwen2.5-7B-Instruct</th>
<th>Falcon3-7B-Instruct</th>
</tr>
</thead>
<tbody>
<tr>
<td rowspan="3">General</td>
<td>MMLU (5-shot)</td>
<td>-</td>
<td>-</td>
<td>-</td>
<td>-</td>
</tr>
<tr>
<td>MMLU-PRO (5-shot)</td>
<td>-</td>
<td>-</td>
<td>-</td>
<td>-</td>
</tr>
<tr>
<td>IFEval</td>
<td>-</td>
<td>-</td>
<td>-</td>
<td>-</td>
</tr>
<tr>
<td rowspan="2">Math</td>
<td>GSM8K (5-shot)</td>
<td>-</td>
<td>-</td>
<td>-</td>
<td>-</td>
</tr>
<tr>
<td>MATH(4-shot)</td>
<td>-</td>
<td>-</td>
<td>-</td>
<td>-</td>
</tr>
<tr>
<td rowspan="4">Reasoning</td>
<td>Arc Challenge (25-shot)</td>
<td>-</td>
<td>-</td>
<td>-</td>
<td>-</td>
</tr>
<tr>
<td>GPQA (0-shot)</td>
<td>-</td>
<td>-</td>
<td>-</td>
<td>-</td>
</tr>
<tr>
<td>MUSR (0-shot)</td>
<td>-</td>
<td>-</td>
<td>-</td>
<td>-</td>
</tr>
<tr>
<td>BBH (3-shot)</td>
<td>-</td>
<td>-</td>
<td>-</td>
<td>-</td>
</tr>
<tr>
<td rowspan="4">CommonSense Understanding</td>
<td>PIQA (0-shot)</td>
<td>-</td>
<td>-</td>
<td>-</td>
<td>-</td>
</tr>
<tr>
<td>SciQ (0-shot)</td>
<td>-</td>
<td>-</td>
<td>-</td>
<td>-</td>
</tr>
<tr>
<td>Winogrande (0-shot)</td>
<td>-</td>
<td>-</td>
<td>-</td>
<td>-</td>
</tr>
<tr>
<td>OpenbookQA (0-shot)</td>
<td>-</td>
<td>-</td>
<td>-</td>
<td>-</td>
</tr>
</tbody>
</table>
# Citation
If Falcon3 series were helpful to your work, feel free to give us a cite.
```
@misc{Falcon3,
title = {Falcon 3 family of Open Foundation Models},
author = {TII Team},
month = {December},
year = {2024}
}
```
|