File size: 57,855 Bytes
dbf0c7e 9d00b78 dbf0c7e 90c58fe dbf0c7e e6b3665 fc26828 b2399b2 348203c 9806546 348203c b2399b2 fc26828 b2399b2 639d17a 9806546 639d17a b2399b2 639d17a b2399b2 639d17a b2399b2 fc26828 e6b3665 fc26828 9806546 fc26828 e6b3665 9806546 e6b3665 348203c 9806546 e6b3665 9806546 e6b3665 9d00b78 dbf0c7e 56dc8bb dbf0c7e e6b3665 639d17a d3783fe 639d17a 56dc8bb e6b3665 56dc8bb dbf0c7e 9c9bfe3 dbf0c7e 9c9bfe3 dbf0c7e 9d00b78 2ab9b5a c1a79df 2ab9b5a c1a79df 2ab9b5a c1a79df 2ab9b5a 46bbe74 9d00b78 ec21158 0faa7ad 94c5433 0b00e32 8485bb9 0b00e32 8485bb9 c1a79df 94c5433 c1a79df 3f2a2c7 0b00e32 c15b2e9 c1a79df 1c6dba3 94c5433 0faa7ad c7d8ebe 8485bb9 c7d8ebe 8485bb9 c1a79df c7d8ebe c1a79df 20a8116 c1a79df c7d8ebe c1a79df 1c6dba3 c1a79df 1c6dba3 c1a79df 1c6dba3 c1a79df 1c6dba3 c7d8ebe 9d00b78 dbf0c7e 3f2a2c7 dbf0c7e 3f2a2c7 dbf0c7e 7c0ce81 c7d8ebe dbf0c7e |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 |
---
license: cc-by-4.0
datasets:
- Salesforce/xlam-function-calling-60k
base_model: Qwen/Qwen2-7B-Instruct
---
# Hammer-7b Function Calling Model
## Introduction
Function calling enables LLMs to invoke specific functions, integrating external features, accessing real-world data, and extending beyond text generation. We present Hammer, a finetuned model based on Qwen2-7B-Instruct. Unlike previous works emphasizing on data refinement (cite xlam, IBM…), our focus is on applying novel training techniques to address recognized issues in existing function-calling models. Such issues are listed below:
1. Hallucination
- a) Function name hallucination: The model, rather than selecting from the provided function pool, has a tendency to generate a new function based on its own world knowledge.
- b) Parameter name hallucination: When the user fails to provide sufficient information to fulfill their request (lacking necessary parameters), the model is inclined to fill in the parameters relying on its own knowledge.
2. Overfitting
- a) Function name and parameter name: The model pays excessive attention to the function name and parameter name while neglecting other information such as description, input, and output. This leads to a lack of generalization and reduces the model's ability to handle diverse scenarios.
- b) Parameter filling: The model does not extract parameters based on the provided function definition. Instead, it fills in the parameters based on the learned knowledge from training. For instance, when expecting "San Francisco", "San Francisco, CA" might be filled in because in the training data, all "San Francisco"s are followed by "CA"s.
- c) Default value filling: The model fills in parameter default values according to patterns in the training data rather than the provided function definition. For example, when "default = inch" is most common in the training data, the model is likely to fill in "inch" instead of "cm", even though the latter is the provided default value in the function definition.
- d) Ordering of provided function list and parameter list: When the provided function list or parameter list have consistent orderings during training, it is possible that the model learns patterns that are not intended, such as remembering the orderings.
3. Instructions missing key information
Occasionally, user instructions may lack essential details vital for effective function execution. For instance, the command "Set an alarm to wake me up”, lacks a time specification. Ideally, in such instances, the model should either request additional information or merely output the function name, excluding the unspecified parameter. Existing methods either disregard such situations or output an “irrelevant” signal, indicating the query is unfulfillable with the given tools.
4. Prompt design
Inconsistency in instruction formatting between training and testing can result in a significant performance gap. For example, during the training phase, the default value is provided in the parameter description, while during testing, the default value is provided as a separate parameter in JSON format.
In this work, we focus on introducing function calling abilities with an inherent emphasis on addressing the aforementioned limitations. We summarize our techniques as follows:
1.Masking: We propose function/parameter mask technique, a dynamic data augmentation method. This approach enhances the model's focus on tool descriptions rather than the tool names within tool definitions. The masking operations include:
- a) Function Name Masking: Replacing the function name with a randomly generated string to ensure the model pays more attention to the function description rather than function names.
- b) Parameter Name Masking: Replacing the parameter name with a randomly generated string to ensure the model pays more attention to the parameter description rather than parameter names.
- c) Default Value Masking: Default values are replaced with random strings to prevent overfitting to specific values.
2.Function Shuffling: Random reordering of functions and parameters during training deters the model from memorizing their sequence.
3.Prompt Optimization: As our model concentrates on function/parameter descriptions, we incorporate default value information into those descriptions to boost performance at inference.
Addressing these multifaceted issues necessitates a refined and sophisticated approach to model training and optimization. To this end, we have meticulously developed an advanced function calling model through the fine-tuning of the *Qwen2-7B-instruct*. The ensuing sections provide a comprehensive overview of the methods and processes implemented during the training phase to mitigate these issues effectively:
1. **Data Extraction and Preparation**:
We extracted 7.5k sample data from *Salesforce/xlam-function-calling-60k* and removed the target tools from the candidate toolset to generate irrelevant data samples. This data was mixed with 60k XLAM data samples for training.
2. **Fine Tuning**:
Our fine-tuning process primarily leveraged the Low-Rank Adaptation (LoRA) technique, incorporating specific hyperparameters and strategies to ensure optimal model performance.
Masking And Function Shuffling Technique was used during the training process;The Training Setup is as follows:
- **LoRA Rank**: 32
- **Learning Rate**: 5e-5
- **Warmup Steps**: 100
- **LR Scheduler Type**: Cosine
- **Batch Size**: 4
- **Gradient Accumulation Steps**: 2
- **Hardware**: 4x A100 (80G) GPUs
3. **Inference**:
During inference, since our model focuses more on function/parameter descriptions, we added default value information in parameter descriptions to obtain better performance.
## Supported Function Calling Types
The model is capable of handling various function calling scenarios. Here, the supported types are classified based on the nature of inputs and outputs:
### Input Types
- 1. **Single Function Input**
- 2. **Multiple Functions Input**
### Output Types
- 1. **Simple Function Calling**
- 2. **Parallel Function Calling**
- 3. **Irrelevance**
- 4. **Relevance**
By categorizing function calling types based on inputs and outputs, our model provides robust support for a wide range of function calling scenarios, ensuring both flexibility and precision in handling diverse tasks.
## Performance
1. First, we evaluate our model on the Berkeley Function-Calling Leaderboard (BFCL), and the performance is as follows：
<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-9id2{color:#007BFF;text-align:center;vertical-align:middle}
.tg .tg-pchv{color:#212529;font-weight:bold;text-align:center;vertical-align:middle}
.tg .tg-qai4{color:#212529;text-align:center;vertical-align:middle}
.tg .tg-p59o{color:#00E;text-align:center;text-decoration:underline;vertical-align:top}
</style>
<table class="tg"><thead>
<tr>
<th class="tg-pchv"><span style="font-weight:700;font-style:normal;text-decoration:none;color:#212529">Rank</span></th>
<th class="tg-pchv"><span style="font-weight:700;font-style:normal;text-decoration:none;color:#212529">Overall</span> <span style="font-weight:700;font-style:normal;text-decoration:none;color:#212529">Acc</span></th>
<th class="tg-pchv"><span style="font-weight:700;font-style:normal;text-decoration:none;color:#212529">Model</span></th>
<th class="tg-pchv"><span style="font-weight:700;font-style:normal;text-decoration:none;color:#212529">AST</span> <span style="font-weight:700;font-style:normal;text-decoration:none;color:#212529">Summary</span></th>
<th class="tg-pchv"><span style="font-weight:700;font-style:normal;text-decoration:none;color:#212529">Exec</span> <span style="font-weight:700;font-style:normal;text-decoration:none;color:#212529">Summary</span></th>
<th class="tg-pchv"><span style="font-weight:700;font-style:normal;text-decoration:none;color:#212529">Irrelevance</span></th>
<th class="tg-pchv"><span style="font-weight:700;font-style:normal;text-decoration:none;color:#212529">Relevance</span></th>
<th class="tg-pchv"><span style="font-weight:700;font-style:normal;text-decoration:none;color:#212529">Organization</span></th>
<th class="tg-pchv"><span style="font-weight:700;font-style:normal;text-decoration:none;color:#212529">License</span></th>
</tr>
</thead>
<tbody>
<tr>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">1</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">85.79</span></td>
<td class="tg-p59o"><a href="https://platform.openai.com/docs/models/gpt-4-and-gpt-4-turbo">GPT-4-0125-Preview (Prompt)</a></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">85.5</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">89.25</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">61.35</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">97.56</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">OpenAI</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">Proprietary</span></td>
</tr>
<tr>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">2</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">85</span></td>
<td class="tg-p59o"><a href="https://platform.openai.com/docs/models/gpt-4-and-gpt-4-turbo">GPT-4-1106-Preview (Prompt)</a></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">86.31</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">87.38</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">64.98</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">90.24</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">OpenAI</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">Proprietary</span></td>
</tr>
<tr>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">3</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">84.74</span></td>
<td class="tg-p59o"><a href="https://platform.openai.com/docs/models/gpt-4-and-gpt-4-turbo">GPT-4-0613 (Prompt)</a></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">84.66</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">87.57</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">75.57</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">82.93</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">OpenAI</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">Proprietary</span></td>
</tr>
<tr>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">4</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">83.92</span></td>
<td class="tg-9id2"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#007BFF">Hammer-7b</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">78.7</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">89.71</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">72.87</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">92.68</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">MadeAgents</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">cc-by-nc-4.0</span></td>
</tr>
<tr>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">5</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">83.89</span></td>
<td class="tg-p59o"><a href="https://platform.openai.com/docs/models/gpt-4-and-gpt-4-turbo">GPT-4-turbo-2024-04-09 (Prompt)</a></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">85.41</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">88.12</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">61.82</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">82.93</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">OpenAI</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">Proprietary</span></td>
</tr>
<tr>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">6</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">83.35</span></td>
<td class="tg-p59o"><a href="https://openai.com/index/gpt-4o-mini-advancing-cost-efficient-intelligence/">GPT-4o-mini-2024-07-18 (Prompt)</a></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">80.51</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">87.95</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">79.2</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">80.49</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">OpenAI</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">Proprietary</span></td>
</tr>
<tr>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">7</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">83.13</span></td>
<td class="tg-p59o"><a href="https://openai.com/index/hello-gpt-4o/">GPT-4o-2024-05-13 (Prompt)</a></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">83.83</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">85.12</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">77.44</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">78.05</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">OpenAI</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">Proprietary</span></td>
</tr>
<tr>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">8</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">82.55</span></td>
<td class="tg-p59o"><a href="https://huggingface.co/meetkai/functionary-medium-v3.1">Functionary-Medium-v3.1 (FC)</a></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">81.06</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">89.32</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">73.23</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">70.73</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">MeetKai</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">MIT</span></td>
</tr>
<tr>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">9</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">81.78</span></td>
<td class="tg-p59o"><a href="https://platform.openai.com/docs/models/gpt-4-and-gpt-4-turbo">GPT-4-1106-Preview (FC)</a></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">77.95</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">87.61</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">72.7</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">82.93</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">OpenAI</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">Proprietary</span></td>
</tr>
<tr>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">10</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">81.59</span></td>
<td class="tg-p59o"><a href="https://llama.meta.com/llama3">Meta-Llama-3-70B-Instruct (Prompt)</a></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">80.15</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">88.04</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">50.47</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">92.68</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">Meta</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">Meta</span> <span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">Llama</span> <span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">3</span> <span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">Community</span></td>
</tr>
<tr>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">11</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">80.88</span></td>
<td class="tg-p59o"><a href="https://www.anthropic.com/news/claude-3-family">Claude-3-Opus-20240229 (Prompt)</a></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">79.42</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">87.39</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">56.15</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">85.37</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">Anthropic</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">Proprietary</span></td>
</tr>
<tr>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">12</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">80.87</span></td>
<td class="tg-p59o"><a href="https://platform.openai.com/docs/models/gpt-4-and-gpt-4-turbo">GPT-4-0125-Preview (FC)</a></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">77.02</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">85.3</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">74.03</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">85.37</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">OpenAI</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">Proprietary</span></td>
</tr>
<tr>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">13</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">80.23</span></td>
<td class="tg-p59o"><a href="https://huggingface.co/nvidia/nemotron-4-340b-instruct">Nemotron-4-340b-instruct (Prompt)</a></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">76.67</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">83.38</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">84.1</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">78.05</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">NVIDIA</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">nvidia-open-model-license</span></td>
</tr>
<tr>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">14</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">80.21</span></td>
<td class="tg-p59o"><a href="https://huggingface.co/meetkai/functionary-small-v3.1">Functionary-Small-v3.1 (FC)</a></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">78.64</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">83.45</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">68.36</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">85.37</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">MeetKai</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">MIT</span></td>
</tr>
<tr>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">15</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">79.66</span></td>
<td class="tg-p59o"><a href="https://mistral.ai/news/mistral-large-2407/">mistral-large-2407 (FC Any)</a></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">85.61</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">88.45</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">0.34</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">100</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">Mistral</span> <span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">AI</span></td>
<td class="tg-qai4"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#212529">Proprietary</span></td>
</tr>
</tbody></table>
*Note: The rankings are based on the performance metrics provided.*
2.In our evaluation, we assessed the function calling capabilities of various models, including our own fine-tuned models using both masked and non-masked approaches. Below are the results across several benchmarks, derived from evaluations performed in a zero-shot manner. Our model, **hammer-7b**, demonstrated superior performance compared to other models.
The table below replicates and extends the format found in ["Granite-Function Calling Model"](https://arxiv.org/abs/2407.00121), particularly Table 6: Function Calling Academic Benchmarks.
<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:12px 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:12px 5px;word-break:normal;}
.tg .tg-baqh{text-align:center;vertical-align:top}
.tg .tg-7geq{background-color:#ffffc7;text-align:center;vertical-align:top}
.tg .tg-k5c1{background-color:#ffffc7;font-weight:bold;text-align:center;vertical-align:top}
.tg .tg-nrix{text-align:center;vertical-align:middle}
.tg .tg-amwm{font-weight:bold;text-align:center;vertical-align:top}
</style>
<table class="tg"><thead>
<tr>
<th class="tg-nrix" rowspan="2"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Model</span></th>
<th class="tg-nrix" rowspan="2"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Size</span></th>
<th class="tg-baqh" colspan="2"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">API-Bank</span> <span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">L-1</span></th>
<th class="tg-baqh" colspan="2"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">API-Bank</span> <span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">L-2</span></th>
<th class="tg-baqh" colspan="2"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Tool-Alpaca</span></th>
<th class="tg-baqh" colspan="2"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Nexus</span> <span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Raven</span></th>
<th class="tg-baqh" colspan="2"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">F1</span> <span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Average</span></th>
</tr>
<tr>
<th class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">F1</span> <span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Func-Name</span></th>
<th class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">F1</span> <span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Args</span></th>
<th class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">F1</span> <span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Func-Name</span></th>
<th class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">F1</span> <span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Args</span></th>
<th class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">F1</span> <span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Func-Name</span></th>
<th class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">F1</span> <span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Args</span></th>
<th class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">F1</span> <span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Func-Name</span></th>
<th class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">F1</span> <span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Args</span></th>
<th class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">F1</span> <span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Func-Name</span></th>
<th class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">F1</span> <span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Args</span></th>
</tr>
</thead>
<tbody>
<tr>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Functionary-small-v2.4</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">7B</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">78.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">70.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">54.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">45.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">88.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">47.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">82.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">64.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">75.50%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">56.50%</span></td>
</tr>
<tr>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Gorilla-openfunctions-v2</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">7B</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">43.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">41.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">12.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">12.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">69.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">39.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">81.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">65.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">51.20%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">39.30%</span></td>
</tr>
<tr>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Hermes-2-Pro-Mistral</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">7B</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">93.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">77.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">54.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">25.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">80.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">26.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">90.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">63.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">79.30%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">47.80%</span></td>
</tr>
<tr>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Mistral-Instruct-v0.3</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">7B</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">79.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">69.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">69.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">46.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">33.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">33.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">71.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">54.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">63.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">50.50%</span></td>
</tr>
<tr>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">CodeGemma-Instruct</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">7B</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">77.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">57.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">59.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">38.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">59.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">31.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">84.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">68.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">69.80%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">48.50%</span></td>
</tr>
<tr>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Nexusflow-Raven-v2</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">13B</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">51.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">42.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">28.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">22.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">85.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">37.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">92.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">75.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">64.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">44.00%</span></td>
</tr>
<tr>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">C4AI-Command-R-v01</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">35B</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">93.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">76.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">77.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">54.00%</span></td>
<td class="tg-amwm"><span style="font-weight:700;font-style:normal;text-decoration:none;color:#000">90.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">42.00%</span></td>
<td class="tg-amwm"><span style="font-weight:700;font-style:normal;text-decoration:none;color:#000">93.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">71.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">88.30%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">60.80%</span></td>
</tr>
<tr>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Meta-Llama-3-70B-Instruct</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">70B</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">85.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">67.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">69.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">52.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">78.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">43.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">70.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">52.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">75.50%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">53.50%</span></td>
</tr>
<tr>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">GRANITE-20B-FUNCTIONCALLING</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">20B</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">91.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">71.00%</span></td>
<td class="tg-amwm"><span style="font-weight:700;font-style:normal;text-decoration:none;color:#000">83.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">60.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">89.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">44.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">92.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">72.00%</span></td>
<td class="tg-amwm"><span style="font-weight:700;font-style:normal;text-decoration:none;color:#000">88.80%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">61.80%</span></td>
</tr>
<tr>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">xlam-7b-fc-r</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">7B</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">90.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">80.70%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">68.90%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">60.70%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">67.30%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">59.00%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">54.10%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">57.50%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">70.10%</span></td>
<td class="tg-baqh"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">64.50%</span></td>
</tr>
<tr>
<td class="tg-7geq"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Hammer-7b</span></td>
<td class="tg-7geq"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">7B</span></td>
<td class="tg-k5c1"><span style="font-weight:700;font-style:normal;text-decoration:none;color:#000">93.80%</span></td>
<td class="tg-k5c1"><span style="font-weight:700;font-style:normal;text-decoration:none;color:#000">85.90%</span></td>
<td class="tg-7geq"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">79.20%</span></td>
<td class="tg-7geq"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">64.40%</span></td>
<td class="tg-7geq"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">82.30%</span></td>
<td class="tg-k5c1"><span style="font-weight:700;font-style:normal;text-decoration:none;color:#000">59.90%</span></td>
<td class="tg-7geq"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">92.50%</span></td>
<td class="tg-k5c1"><span style="font-weight:700;font-style:normal;text-decoration:none;color:#000">77.40%</span></td>
<td class="tg-7geq"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">86.90%</span></td>
<td class="tg-k5c1"><span style="font-weight:700;font-style:normal;text-decoration:none;color:#000">71.90%</span></td>
</tr>
</tbody></table>
3.Finally, we evaluate the performance of our model on the [Seal-Tools](https://arxiv.org/abs/2405.08355) dataset, which also achieves better performance.
<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:12px 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:12px 5px;word-break:normal;}
.tg .tg-9wq8{border-color:inherit;text-align:center;vertical-align:middle}
.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-mfhl{background-color:#ffffc7;border-color:inherit;text-align:center;vertical-align:top}
.tg .tg-py60{background-color:#ffffc7;border-color:inherit;font-weight:bold;text-align:center;vertical-align:top}
</style>
<table class="tg"><thead>
<tr>
<th class="tg-9wq8" rowspan="2"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Model</span></th>
<th class="tg-9wq8" rowspan="2"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Size</span></th>
<th class="tg-c3ow" colspan="2"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">SealTool(Single-Tool)</span></th>
</tr>
<tr>
<th class="tg-c3ow"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">F1</span> <span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Func-Name</span></th>
<th class="tg-c3ow"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">F1</span> <span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Args</span></th>
</tr></thead>
<tbody>
<tr>
<td class="tg-c3ow"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Gorilla-openfunctions-v2</span></td>
<td class="tg-c3ow"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">7B</span></td>
<td class="tg-c3ow"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">93.20%</span></td>
<td class="tg-c3ow"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">91.10%</span></td>
</tr>
<tr>
<td class="tg-c3ow"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">GRANITE-20B-FUNCTIONCALLING</span></td>
<td class="tg-c3ow"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">20B</span></td>
<td class="tg-c3ow"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">94.90%</span></td>
<td class="tg-7btt"><span style="font-weight:700;font-style:normal;text-decoration:none;color:#000">92.70%</span></td>
</tr>
<tr>
<td class="tg-c3ow"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">xlam-7b-fc-r</span></td>
<td class="tg-c3ow"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">7B</span></td>
<td class="tg-c3ow"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">79.00%</span></td>
<td class="tg-c3ow"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">76.90%</span></td>
</tr>
<tr>
<td class="tg-mfhl"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">Hammer-7b</span></td>
<td class="tg-mfhl"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">7B</span></td>
<td class="tg-py60"><span style="font-weight:700;font-style:normal;text-decoration:none;color:#000">97.40%</span></td>
<td class="tg-mfhl"><span style="font-weight:400;font-style:normal;text-decoration:none;color:#000">91.70%</span></td>
</tr>
</tbody></table>
## Upcoming Developments
We are actively working on preparing smaller models derived from this architecture, which will be open-sourced soon.
## Example Usage
This is a simple example of how to use our model.
~~~python
import json
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
model_name = "MadeAgents/Hammer-7b"
model = AutoModelForCausalLM.from_pretrained(model_name, device_map="auto", torch_dtype="auto", trust_remote_code=True)
tokenizer = AutoTokenizer.from_pretrained(model_name)
# Please use our provided instruction prompt for best performance
TASK_INSTRUCTION = """You are a tool calling assistant. In order to complete the user's request, you need to select one or more appropriate tools from the following tools and fill in the correct values for the tool parameters. Your specific tasks are:
1. Make one or more function/tool calls to meet the request based on the question.
2. If none of the function can be used, point it out and refuse to answer.
3. If the given question lacks the parameters required by the function, also point it out.
"""
FORMAT_INSTRUCTION = """
The output MUST strictly adhere to the following JSON format, and NO other text MUST be included.
The example format is as follows. Please make sure the parameter type is correct. If no function call is needed, please directly output an empty list '[]'
```
[
{"name": "func_name1", "arguments": {"argument1": "value1", "argument2": "value2"}},
... (more tool calls as required)
]
```
"""
# Define the input query and available tools
query = "Where can I find live giveaways for beta access and games? And what's the weather like in New York, US?"
live_giveaways_by_type = {
"name": "live_giveaways_by_type",
"description": "Retrieve live giveaways from the GamerPower API based on the specified type.",
"parameters": {
"type": "object",
"properties": {
"type": {
"type": "string",
"description": "The type of giveaways to retrieve (e.g., game, loot, beta).",
"default": "game"
}
},
"required": ["type"]
}
}
get_current_weather={
"name": "get_current_weather",
"description": "Get the current weather",
"parameters": {
"type": "object",
"properties": {
"location": {
"type": "string",
"description": "The city and state, e.g. San Francisco, CA"
}
},
"required": ["location"]
}
}
get_stock_price={
"name": "get_stock_price",
"description": "Retrieves the current stock price for a given ticker symbol. The ticker symbol must be a valid symbol for a publicly traded company on a major US stock exchange like NYSE or NASDAQ. The tool will return the latest trade price in USD. It should be used when the user asks about the current or most recent price of a specific stock. It will not provide any other information about the stock or company.",
"parameters": {
"type": "object",
"properties": {
"ticker": {
"type": "string",
"description": "The stock ticker symbol, e.g. AAPL for Apple Inc."
}
},
"required": ["ticker"]
}
}
def convert_to_format_tool(tools):
''''''
if isinstance(tools, dict):
format_tools = {
"name": tools["name"],
"description": tools["description"],
"parameters": tools["parameters"].get("properties", {}),
}
required = tools["parameters"].get("required", [])
for param in required:
format_tools["parameters"][param]["required"] = True
for param in format_tools["parameters"].keys():
if "default" in format_tools["parameters"][param]:
default = format_tools["parameters"][param]["default"]
format_tools["parameters"][param]["description"]+=f"default is \'{default}\'"
return format_tools
elif isinstance(tools, list):
return [convert_to_format_tool(tool) for tool in tools]
else:
return tools
# Helper function to build the input prompt for our model
def build_prompt(task_instruction: str, format_instruction: str, tools: list, query: str):
prompt = f"[BEGIN OF TASK INSTRUCTION]\n{task_instruction}\n[END OF TASK INSTRUCTION]\n\n"
prompt += f"[BEGIN OF AVAILABLE TOOLS]\n{json.dumps(tools)}\n[END OF AVAILABLE TOOLS]\n\n"
prompt += f"[BEGIN OF FORMAT INSTRUCTION]\n{format_instruction}\n[END OF FORMAT INSTRUCTION]\n\n"
prompt += f"[BEGIN OF QUERY]\n{query}\n[END OF QUERY]\n\n"
return prompt
# Build the input and start the inference
openai_format_tools = [live_giveaways_by_type, get_current_weather,get_stock_price]
format_tools = convert_to_format_tool(openai_format_tools)
content = build_prompt(TASK_INSTRUCTION, FORMAT_INSTRUCTION, format_tools, query)
messages=[
{ 'role': 'user', 'content': content}
]
inputs = tokenizer.apply_chat_template(messages, add_generation_prompt=True, return_tensors="pt").to(model.device)
# tokenizer.eos_token_id is the id of <|EOT|> token
outputs = model.generate(inputs, max_new_tokens=512, do_sample=False, num_return_sequences=1, eos_token_id=tokenizer.eos_token_id)
print(tokenizer.decode(outputs[0][len(inputs[0]):], skip_special_tokens=True))
~~~
---
## References
- 1.Yan F, Mao H, Ji C C-J, et al. Berkeley Function Calling Leaderboard.
- 2. Abdelaziz I, Basu K, Agarwal M, et al. Granite-Function Calling Model: Introducing Function Calling Abilities via Multi-task Learning of Granular Tasks[J]. arXiv preprint arXiv:2407.00121, 2024.
- 3. Wu M, Zhu T, Han H, et al. Seal-Tools: Self-Instruct Tool Learning Dataset for Agent Tuning and Detailed Benchmark[J]. arXiv preprint arXiv:2405.08355, 2024.
Feel free to reach out for further clarifications or contributions! |