File size: 31,359 Bytes
0fad588 250460a 0fad588 250460a 0fad588 250460a 0fad588 af7125e 0fad588 d3e3eab 0fad588 d3e3eab 0fad588 9c96887 0fad588 9c96887 0fad588 9c96887 0fad588 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 |
---
license: apache-2.0
---
<div align="center">
<img src="https://github.com/baichuan-inc/Baichuan-Omni-1.5/raw/main/assets/logo.png" width="300em" ></img>
<!-- <img src="https://raw.githubusercontent.com/baichuan-inc/Baichuan-Omni-1.5/refs/heads/main/assets/logo.png" width="300em" ></img>
<img src="https://github.com/baichuan-inc/Baichuan-Omni-1.5/raw/main/assets/train-pipeline.png" width="300em" ></img> -->
<!-- <img src="https://github.com/OpenBMB/MiniCPM-o/raw/main/assets/minicpm-o-26-framework-v2.png" width="300em" ></img> -->
**Open-source Omni-modal Foundation Model Supporting Text, Image, Video, and Audio Inputs as Well as Text and Audio Outputs**
<p align="center">
Baichuan-Omni-1.5 <a href="https://huggingface.co/baichuan-inc/Baichuan-Omni-1d5">🤗</a> | Baichuan-Omni-1.5-Base <a href="https://huggingface.co/baichuan-inc/Baichuan-Omni-1d5-Base">🤗</a> |Github <a href="https://github.com/baichuan-inc/Baichuan-Omni-1.5/">📖 </a> | Report <a href="https://github.com/baichuan-inc/Baichuan-Omni-1.5/raw/main/baichuan_omni_1_5.pdf">📖</a>
</p>
</p>
<p align="center">
OpenMM-Medical <a href="https://huggingface.co/datasets/baichuan-inc/OpenMM-Medical">🤗</a> | OpenAudioBench <a href="https://huggingface.co/datasets/baichuan-inc/OpenAudioBench">🤗</a>
</p>
</div>
<!-- ## 介绍
**Baichuan-Omni-1.5** 是从 Baichuan-omni 升级的最新的、端到端训练的、支持全模态输入/双模态输出的多模态大模型。该模型使用Qwen2.5-7B昨晚大语言模型基座，可以以端到端方式，接受图像、视频、文本、音频作为输入，并且以可控的方式生成高质量文本和语音。
- **Baichuan-Omni-1.5-Base**: 为促进全模态大模型发展，我们开源了使用高质量海量数据训练的全模态基座模型。该模型未经SFT指令微调，可塑性强，是**业内首个**开源的**全模态基座模型**。
- **Baichuan-Omni-1.5**: 基于性能强悍的Baichuan-Omni-1.5-base，使用高质量的全模态对齐数据，进行端到端的多模态指令数据训练。Baichuan-Omni-1.5的纯文本、图像、视频、音频理解能力达到了 GPT-4o-mini 级别。可控音频生成的能力十分强大，在xxx和xxx评测集上取得最高表现。 -->
## Baichuan-Omni-1.5
The Baichuan-Omni-1.5 is the latest, top-performing model in the Baichuan-omni series. This model is trained and inferred in an end-to-end manner. Compared with Baichuan-omni, this model has significant improvements in text/image/audio/video understanding and text/audio generation, and supports new features such as controllable real-time voice conversations and multi-modal real-time interactions. The main features of Baichuan-Omni-1.5 include:
- 🔥 **Possess Multimodal Understanding and Interaction Capabilities.**
Baichuan-Omni-1.5 not only supports images, videos, text, and audio as input, and generates high-quality text and voice output, but also **supports continuous video and audio streaming, and real-time voice interaction with users**. In OminiBench, a comprehensive evaluation benchmark for omnimodal understanding, Baichuan-Omni-1.5 has achieved the first-class level of the open source community and surpassed GPT-4o-mini.
- 💪 **Strong Visual Capability.**
Baichuan-Omni-1.5 has an average score of 73.3 on the OpenCompass list (comprehensive 10 mainstream multimodal evaluation benchmarks). **With the size of 7B, it surpasses mainstream commercial closed-source multimodal large models such as GPT-4o-mini, Gemini 1.5 Pro and Claude 3.5 Sonnet in single-image understanding**. In addition, its video understanding performance is also better than GPT-4V and Claude 3.5 Sonnet and open source omnimodal models.
- 🚀 **Leading Medical Image Understanding Capabilities.**
Baichuan-Omni-1.5 achieved the best performance on GMAI-MMBench and Openmm-Medical. Using only 7B LLM, the average score exceeded Qwen2-VL-72b by 3%, i.e. 80.7% v.s 83.8%.
- 🎙 **Excellent Voice Capabilities.**
Baichuan-Omni-1.5 **supports high-quality, controllable voice bilingual real-time conversations in Chinese and English**. It **outperforms GPT-4o-realtime** in speech understanding tasks (such as ASR and STT, etc.), and demonstrates **the highest speech generation performance among open source models** in semantic and acoustic evaluation of voice conversations.
- 🎬 **Powerful Real-world Understanding and Other Features.**
Baichuan-Omni-1.5 further optimizes the many visual understanding capabilities of Baichuan-omni. It can process images of any aspect ratio and up to 1.8 million pixels (such as 1344x1344). It scored 68.8 points on RealWorldQA, **surpassing commercial closed-source models such as GPT-4o-mini** and recently open-sourced omnimodal models. It scored 85.6/83.6 on the English/Chinese evaluation subsets of MMBench, respectively, which is also in the first echelon of models with the same size.
- 💫 **Provides [🤗 Base Model](https://huggingface.co/baichuan-inc/Baichuan-Omni-1d5-Base) and [🤗 Instruct Model](https://huggingface.co/baichuan-inc/Baichuan-Omni-1d5).**
Baichuan-Omni-1.5-Base is a high-performance foundational omni-modal model in the industry. Based on the powerful base, Baichuan-Omni-1.5 employs high-quality omnimodal alignment data to perform end-to-end multimodal instruction data training.
**Model Architecture**
<div align="center">
<img src="https://github.com/baichuan-inc/Baichuan-Omni-1.5/raw/main/assets/train-pipeline.png", width=80%></img>
</div>
<br>
- **End-to-end Omni-modal Architecture.** We carefully design **multi-stage and end-to-end** progressive training of different modal encoding/decoding modules to make full use of the rich knowledge in different modalities, we expect different modal knowledge to complement each other.
Notably, the model is fully trained end-to-end using NTP loss in the whole pre-training stage.
- **High-quality Controllable Audio Solution.** Multimodal system prompts have been redesigned to include traditional text system prompts and **speech system prompts** for specifying model sounds. It provides the flexibility to control voice style through text or speech samples at inference time, and supports advanced capabilities such as end-to-end voice cloning and timbre creation.
### Open-source Evaluation Datasets
**OpenMM-Medical**
To comprehensively evaluate the model's multi-modal medical capabilities, we have constructed OpenMM-Medical, which includes data from 42 publicly available medical image datasets such as ACRIMA (retinal images), BioMediTech (microscope images), and CoronaHack (X-rays), totaling 88,996 images.
**OpenAudioBench**
To efficiently assess the model's "IQ" issues, we developed OpenAudioBench, comprising five end-to-end audio understanding sub-datasets: four public benchmarks (Llama Question, WEB QA, TriviaQA, AlpacaEval), and an internally created speech logical reasoning dataset by the Baichuan team, totaling 2,701 entries. This suite reflects the model's comprehensive "IQ" level.
<!-- **High-quality Medical Image Evaluation Dataset--Openmm-Medical**
- We have built a more diverse medical evaluation dataset named **Openmm-Medical** to evaluate large models in medical scenarios.
- The images in Openmm-Medical come from **42 public medical image datasets**, such as ACRIMA (fundus images), BioMediTech (microscope images), and CoronaHack (X-rays).
- **Openmm-Medical contains a total of 88,996 images**, and each image is designed as a **multiple-choice question to facilitate the evaluation of different large models.**
- To promote the development of omnimodal large models in the medical field, we will soon **open** this evaluation dataset.
-->
### Evaluation
We sugguest readers to refer to our [**Github**](https://github.com/baichuan-inc/Baichuan-Omni-1.5/) for more details.
<div align="center">
<img src="https://github.com/baichuan-inc/Baichuan-Omni-1.5/raw/main/assets/performance.png" , width=80%>
</div>
<br>
<details>
<summary>click to view</summary>
#### Pure Text Understanding
<div align="center">
<table style="margin: 0 auto; text-align: center;">
<thead>
<tr>
<th class="tg-c3ow" colspan="7">Comprehensive Tasks</th>
</tr>
</thead>
<tbody>
<tr>
<td>Model</td>
<td>Size</td>
<td>MMLU (Acc.)</td>
<td>CMMLU (Acc.)</td>
<td>AGIEval (Acc.)</td>
<td>C-Eval (Acc.)</td>
<td>GAOKAO (Acc.)</td>
</tr>
<tr>
<td colspan="7">Proprietary Models</td>
</tr>
<tr>
<td>GPT 4o</td>
<td>-</td>
<td><b>88.0♢<br></td>
<td><b>78.3♢<br></td>
<td><b>62.3♢<br></td>
<td><b>86.0♢<br></td>
<td>-</td>
</tr>
<tr>
<td>GPT 4o mini</td>
<td>-</td>
<td>82.0</td>
<td>67.6</td>
<td>52.2</td>
<td>63.6</td>
<td>70.8</td>
</tr>
<tr>
<td colspan="7">Open-source Models (Pure text)</td>
</tr>
<tr>
<td>MAP-Neo</td>
<td>7B</td>
<td>58.2</td>
<td>55.1</td>
<td>33.9</td>
<td>57.5</td>
<td>-</td>
</tr>
<tr>
<td>Qwen1.5-Chat</td>
<td>7B</td>
<td>61.5</td>
<td>68.0</td>
<td>39.3</td>
<td>68.8</td>
<td>-</td>
</tr>
<tr>
<td>Llama3-Instruct</td>
<td>8B</td>
<td>67.1</td>
<td>51.7</td>
<td>38.4</td>
<td>50.7</td>
<td>-</td>
</tr>
<tr>
<td>OLMo</td>
<td>7B</td>
<td>28.4</td>
<td>25.6</td>
<td>19.9</td>
<td>27.3</td>
<td>-</td>
</tr>
<tr>
<td colspan="7">Open-source Models (Omni-modal)</td>
</tr>
<tr>
<td>VITA</td>
<td>8x7B</td>
<td>71.0*</td>
<td>46.6</td>
<td>46.2*</td>
<td>56.7*</td>
<td>-</td>
</tr>
<tr>
<td>VITA-1.5</td>
<td>7B</td>
<td>71.0</td>
<td>75.1</td>
<td>47.9</td>
<td>65.6</td>
<td>57.4</td>
</tr>
<tr>
<td>Baichuan-Omni</td>
<td>7B</td>
<td>65.3</td>
<td>72.2</td>
<td>47.7</td>
<td>68.9</td>
<td>-</td>
</tr>
<tr>
<td>MiniCPM-o 2.6</td>
<td>7B</td>
<td>65.3</td>
<td>63.3</td>
<td>50.9</td>
<td>61.5</td>
<td>56.3</td>
</tr>
<tr>
<td><b>Baichuan-Omni-1.5<br></td>
<td>7B</td>
<td>72.2</td>
<td>75.5</td>
<td>54.4</td>
<td>73.1</td>
<td><b>73.5<br></td>
</tr>
</tbody>
</table>
</div>
</details>
<details>
<summary>click to view</summary>
#### Image Understanding
<div align="center">
<table style="margin: 0 auto; text-align: center;">
<thead>
<tr>
<th class="tg-c3ow" colspan="9">Multi-choice & Yes-or-No Question</th>
</tr>
</thead>
<tbody>
<tr>
<td>Model</td>
<td>Size</td>
<td>MMBench-EN (Acc.)</td>
<td>MMbench-CN (Acc.)</td>
<td>SEED-IMG (Acc.)</td>
<td>MMMU-val (Acc.)</td>
<td>HallusionBench (Acc.)</td>
</tr>
<tr>
<td colspan="9">Proprietary Models</td>
</tr>
<tr>
<td>GPT-4o</td>
<td>-</td>
<td>83.4♢</td>
<td>82.1♢</td>
<td>-</td>
<td><b>69.1♢<br></td>
<td><b>55.0♢<br></td>
</tr>
<tr>
<td>GPT-4o-mini</td>
<td>-</td>
<td>77.7</td>
<td>76.9</td>
<td>72.3</td>
<td>60.0♢</td>
<td>46.1♢</td>
</tr>
<tr>
<td colspan="9">Open Source Models (Vision-Language)</td>
</tr>
<tr>
<td>Qwen2-VL-7B</td>
<td>7B</td>
<td><b>86.4<br></td>
<td>81.9</td>
<td><b>76.5<br></td>
<td>52.7</td>
<td>50.6∗</td>
</tr>
<tr>
<td>MiniCPM-Llama3-V 2.5</td>
<td>8B</td>
<td>76.7</td>
<td>73.3</td>
<td>72.4</td>
<td>45.8∗</td>
<td>42.5</td>
</tr>
<tr>
<td colspan="9">Open Source Models (Omni-modal)</td>
</tr>
<tr>
<td>VITA</td>
<td>8x7B</td>
<td>74.7</td>
<td>71.4</td>
<td>72.6</td>
<td>45.3</td>
<td>39.7∗</td>
</tr>
<tr>
<td>VITA-1.5</td>
<td>7B</td>
<td>80.8</td>
<td>80.2</td>
<td>74.2</td>
<td>53.1</td>
<td>44.1</td>
</tr>
<tr>
<td>Baichuan-Omni</td>
<td>7B</td>
<td>76.2</td>
<td>74.9</td>
<td>74.1</td>
<td>47.3</td>
<td>47.8</td>
</tr>
<tr>
<td>MiniCPM-o 2.6</td>
<td>7B</td>
<td>83.6</td>
<td>81.8</td>
<td>75.4</td>
<td>51.1</td>
<td>50.1</td>
</tr>
<tr>
<td><b>Baichuan-Omni-1.5<br></td>
<td>7B</td>
<td>85.6</td>
<td><b>83.6<br></td>
<td>75.7</td>
<td>53.9</td>
<td>49.7</td>
</tr>
</tbody>
</table>
</div>
<br>
<div align="center">
<table style="margin: 0 auto; text-align: center;">
<thead>
<tr>
<th class="tg-c3ow" colspan="9">Visual Question Answering</th>
</tr>
</thead>
<tbody>
<tr>
<td>Model</td>
<td>Size</td>
<td>RealWorldQA (Acc.)</td>
<td>MathVista-mini (Acc.)</td>
<td>TextVQA-val (Acc.)</td>
<td>ChartQA (Acc.)</td>
<td>OCRBench (Acc.)</td>
</tr>
<tr>
<td colspan="8">Proprietary Models</td>
</tr>
<tr>
<td>GPT-4o</td>
<td>-</td>
<td><b>75.4♢<br></td>
<td>63.8♢</td>
<td>-</td>
<td>85.7♢</td>
<td>73.6♢</td>
</tr>
<tr>
<td>GPT-4o-mini</td>
<td>-</td>
<td>66.3</td>
<td>53.4</td>
<td>66.8</td>
<td>-</td>
<td>77.4</td>
</tr>
<tr>
<td colspan="8">Open Source Models (Vision-Language)</td>
</tr>
<tr>
<td>Qwen2-VL-7B</td>
<td>7B</td>
<td>69.7</td>
<td>58.2∗</td>
<td><b>84.3∗<br></td>
<td>83.0∗</td>
<td>84.5∗</td>
</tr>
<tr>
<td>MiniCPM-Llama3-V 2.5</td>
<td>8B</td>
<td>63.5</td>
<td>54.3∗</td>
<td>76.6</td>
<td>72.0</td>
<td>72.5</td>
</tr>
<tr>
<td colspan="8">Open Source Models (Omni-modal)</td>
</tr>
<tr>
<td>VITA</td>
<td>8x7B</td>
<td>59.0</td>
<td>44.9∗</td>
<td>71.8</td>
<td>76.6</td>
<td>68.5∗</td>
</tr>
<tr>
<td>VITA-1.5</td>
<td>7B</td>
<td>66.8</td>
<td><b>66.5<br></td>
<td>74.9</td>
<td>79.6</td>
<td>73.3</td>
</tr>
<tr>
<td>Baichuan-Omni</td>
<td>7B</td>
<td>62.6</td>
<td>51.9</td>
<td>74.3</td>
<td>79.6</td>
<td>70.0</td>
</tr>
<tr>
<td>MiniCPM-o 2.6</td>
<td>7B</td>
<td>67.7</td>
<td>64.6</td>
<td>80.1</td>
<td><b>87.6<br></td>
<td><b>89.7∗<br></td>
</tr>
<tr>
<td>Baichuan-Omni-1.5 </td>
<td>7B</td>
<td>68.8</td>
<td>63.6</td>
<td>83.2</td>
<td>84.9</td>
<td>84.0</td>
</tr>
</tbody>
</table>
</div>
</details>
<details>
<summary>click to view</summary>
#### Video Understanding
<div align="center">
<table style="margin: 0 auto; text-align: center;">
<thead>
<tr>
<th colspan="7">General VQA </th>
</tr>
</thead>
<tbody>
<tr>
<td>Model</td>
<td>Size</td>
<td># Frames</td>
<td>MVBench (Acc.)</td>
<td>Egoschema (Acc.)</td>
<td>VideoMME (Acc.)</td>
<td>Perception-Test (Acc.)</td>
</tr>
<tr>
<td colspan="7">Proprietary Models</td>
</tr>
<tr>
<td>Gemini 1.5 Pro</td>
<td>-</td>
<td>-</td>
<td><b>81.3♢<br></td>
<td>63.2*</td>
<td><b>75.0♢<br></td>
<td>-</td>
</tr>
<tr>
<td>GPT 4o mini</td>
<td>-</td>
<td>-</td>
<td>55.2</td>
<td>58.5</td>
<td>63.6</td>
<td>48.2</td>
</tr>
<tr>
<td>GPT 4o</td>
<td>-</td>
<td>-</td>
<td>-</td>
<td><b>77.2*<br></td>
<td>71.9♢</td>
<td>-</td>
</tr>
<tr>
<td>GPT 4V</td>
<td>-</td>
<td>-</td>
<td>43.7♢</td>
<td>55.6*</td>
<td>59.9♢</td>
<td>-</td>
</tr>
<tr>
<td colspan="7">Open-source Models (Vision-language)</td>
</tr>
<tr>
<td>Qwen2-VL-7B</td>
<td>7B</td>
<td>2 fps (max 768)</td>
<td>67.0* | 64.4</td>
<td>66.7* | 66.6</td>
<td>63.3* | 59.0</td>
<td>62.3* | 60.3</td>
</tr>
<tr>
<td>AnyGPT</td>
<td>8B</td>
<td>48</td>
<td>33.2</td>
<td>32.1</td>
<td>29.8</td>
<td>29.1</td>
</tr>
<tr>
<td>VideoLLaMA 2</td>
<td>7B</td>
<td>16</td>
<td>54.6*</td>
<td>51.7*</td>
<td>46.6*</td>
<td>51.4*</td>
</tr>
<tr>
<td>VideoChat2</td>
<td>7B</td>
<td>16</td>
<td>51.1*</td>
<td>42.1♢</td>
<td>33.7♢</td>
<td>47.3♢</td>
</tr>
<tr>
<td>LLaVA-NeXT-Video</td>
<td>7B</td>
<td>32</td>
<td>46.5♢</td>
<td>43.9♢</td>
<td>33.7♢</td>
<td>48.8♢</td>
</tr>
<tr>
<td>Video-LLaVA</td>
<td>7B</td>
<td>8</td>
<td>41.0♢</td>
<td>38.4♢</td>
<td>39.9♢</td>
<td>44.3♢</td>
</tr>
<tr>
<td colspan="7">Open-source Models (Omni-modal)</td>
</tr>
<tr>
<td>VITA</td>
<td>8x7B</td>
<td>1 fps (max 32)</td>
<td>53.4</td>
<td>53.9</td>
<td>56.1</td>
<td>56.2</td>
</tr>
<tr>
<td>VITA-1.5</td>
<td>7B</td>
<td>1 fps (max 32)</td>
<td>55.5</td>
<td>54.7</td>
<td>57.3</td>
<td>57.6</td>
</tr>
<tr>
<td>Baichuan-Omni</td>
<td>7B</td>
<td>1 fps (max 32)</td>
<td>60.9</td>
<td>58.8</td>
<td>58.2</td>
<td>56.8</td>
</tr>
<tr>
<td>MiniCPM-o 2.6</td>
<td>7B</td>
<td>1 fps (max 64)</td>
<td>58.6</td>
<td>50.7</td>
<td>63.4</td>
<td>66.6</td>
</tr>
<tr>
<td>Baichuan-Omini-1.5</td>
<td>7B</td>
<td>1 fps (max 32)</td>
<td> 63.7 </td>
<td> 62.4 </td>
<td> 60.1 </td>
<td> <b>68.9 <br> </td>
</tr>
</tbody>
</table>
</div>
<br>
<div align="center">
<table style="margin: 0 auto; text-align: center;">
<thead>
<tr>
<th colspan="7">Open-ended VQA</th>
</tr>
</thead>
<tbody>
<tr>
<td rowspan="2">Model</td>
<td rowspan="2">Size</td>
<td rowspan="2"># Frames</td>
<td colspan="2">ActivityNet-QA</td>
<td colspan="2">MSVD-QA</td>
</tr>
<tr>
<td>(Acc.)</td>
<td>(Score)</td>
<td>(Acc.)</td>
<td>(Score)</td>
</tr>
<tr>
<td colspan="7">Proprietary Models</td>
</tr>
<tr>
<td>Gemini 1.5 Pro</td>
<td>-</td>
<td>-</td>
<td>56.7*</td>
<td>-</td>
<td>-</td>
<td>-</td>
</tr>
<tr>
<td>GPT 4o mini</td>
<td>-</td>
<td>1 fps (max 32)</td>
<td>62.1</td>
<td>3.1</td>
<td>67.5</td>
<td>3.3</td>
</tr>
<tr>
<td>GPT 4o</td>
<td>-</td>
<td>-</td>
<td>61.9*</td>
<td>-</td>
<td>-</td>
<td>-</td>
</tr>
<tr>
<td>GPT 4V</td>
<td>-</td>
<td>-</td>
<td>59.5*</td>
<td>-</td>
<td>-</td>
<td>-</td>
</tr>
<tr>
<td colspan="7">Open-source Models (Vision-language)</td>
</tr>
<tr>
<td>Qwen2 VL</td>
<td>7B</td>
<td>2 fps (max 768)</td>
<td>17.4</td>
<td>1.9</td>
<td>61.1</td>
<td>3.5</td>
</tr>
<tr>
<td>VideoLLaMA 2</td>
<td>7B</td>
<td>16</td>
<td>50.2*</td>
<td>3.3*</td>
<td>70.9*</td>
<td>3.8*</td>
</tr>
<tr>
<td>VideoChat2</td>
<td>7B</td>
<td>16</td>
<td>49.1*</td>
<td>3.3*</td>
<td>70.0*</td>
<td>3.9*</td>
</tr>
<tr>
<td>LLaVA-NeXT-Video</td>
<td>7B</td>
<td>32</td>
<td>53.5*</td>
<td>3.2*</td>
<td>67.4</td>
<td>3.4</td>
</tr>
<tr>
<td>Video-LLaVA</td>
<td>7B</td>
<td>8</td>
<td>45.3*</td>
<td>3.3*</td>
<td>70.7*</td>
<td>3.9*</td>
</tr>
<tr>
<td colspan="7">Open-source Models (Omni-modal)</td>
</tr>
<tr>
<td>VITA</td>
<td>8x7B</td>
<td>1 fps (max 32)</td>
<td>55.0</td>
<td>3.5</td>
<td>63.9</td>
<td>3.7</td>
</tr>
<tr>
<td>VITA-1.5</td>
<td>7B</td>
<td>1 fps (max 32)</td>
<td>59.6</td>
<td>3.0</td>
<td>67.6</td>
<td>3.3</td>
</tr>
<tr>
<td>Baichuan-Omni</td>
<td>7B</td>
<td>1 fps (max 48)</td>
<td>58.6</td>
<td><b>3.7<br></td>
<td>72.2</td>
<td> <b>4.0<br> </td>
</tr>
<tr>
<td>MiniCPM-o 2.6</td>
<td>7B</td>
<td>1 fps (max 64)</td>
<td><b>63.0<br></td>
<td>3.1</td>
<td>73.7</td>
<td>3.6</td>
</tr>
<tr>
<td>Baichuan-Omni-1.5</td>
<td>7B</td>
<td>1 fps (max 48)</td>
<td> 62.0</td>
<td> 3.1</td>
<td> <b> 74.2 <br></td>
<td> 3.6</td>
</tr>
</tbody>
</table>
</div>
</details>
<details>
<summary>click to view</summary>
#### Audio Comprehensive and Speech Generation
<div align="center">
<table style="margin: 0 auto; text-align: center;">
<thead>
<tr>
<th colspan="12">Audio Comprehensive Capacity</th>
</tr></thead>
<tbody>
<tr>
<td rowspan="2">Model</td>
<td rowspan="2">Size</td>
<td colspan="2">Reasoning QA</td>
<td colspan="2">Llama Questions</td>
<td colspan="2">Web Questions</td>
<td colspan="2">TriviaQA</td>
<td colspan="2">AlpacaEval</td>
</tr>
<tr>
<td>s→t</td>
<td>s→s</td>
<td>s→t</td>
<td>s→s</td>
<td>s→t</td>
<td>s→s</td>
<td>s→t</td>
<td>s→s</td>
<td>s→t</td>
<td>s→s</td>
</tr>
<tr>
<td colspan="12">Proprietary Models</td>
</tr>
<tr>
<td>GPT-4o-Audio</td>
<td>-</td>
<td><b>55.6</td>
<td>-</td>
<td><b>88.4</td>
<td>-</td>
<td><b>8.10</td>
<td>-</td>
<td><b>9.06</td>
<td>-</td>
<td><b>8.01</td>
<td>-</td>
</tr>
<tr>
<td colspan="12">Open-source Models (Pure Audio)</td>
</tr>
<tr>
<td>GLM-4-Voice</td>
<td>9B</td>
<td>-</td>
<td>26.5</td>
<td>-</td>
<td>71.0</td>
<td>-</td>
<td>5.15</td>
<td>-</td>
<td>4.66</td>
<td>-</td>
<td>4.89</td>
</tr>
<tr>
<td colspan="12">Open-source Models (Omni-modal)</td>
</tr>
<tr>
<td>VITA-1.5</td>
<td>7B</td>
<td>41.0</td>
<td>-</td>
<td>74.2</td>
<td>-</td>
<td>5.73</td>
<td>-</td>
<td>4.68</td>
<td>-</td>
<td>6.82</td>
<td>-</td>
</tr>
<tr>
<td>MiniCPM-o 2.6</td>
<td>7B</td>
<td>38.6</td>
<td>-</td>
<td>77.8</td>
<td>-</td>
<td>6.86</td>
<td>-</td>
<td>6.19</td>
<td>-</td>
<td>5.18</td>
<td>-</td>
</tr>
<tr>
<td><b>Baichuan-Omni-1.5</td>
<td>7B</td>
<td>50.0</td>
<td><b>40.9</td>
<td>78.5</td>
<td><b>75.3</td>
<td>5.91</td>
<td><b>5.52</td>
<td>5.72</td>
<td>5.31</td>
<td>7.79</td>
<td><b>6.94</td>
</tr>
</tbody>
</table>
</div>
</details>
<details>
<summary>click to view</summary>
#### Omni-modal Understanding
<div align="center">
<table style="margin: 0 auto; text-align: center;">
<thead>
<tr>
<th colspan="7">Omni-Undesratnding </th>
</tr>
<thead>
<tbody>
<tr>
<td>Model</td>
<td>Size</td>
<td>Image & Audio</td>
<td>Image Caption & Audio</td>
<td>Image & Audio Transcript</td>
<td>Image Caption & Audio Transcript</td>
</tr>
</thead>
<tr>
<td colspan="6">Proprietary Models</td>
</tr>
<tr>
<td>GPT4o-mini</td>
<td>-</td>
<td>-</td>
<td>-</td>
<td>37.0</td>
<td>37.7</td>
</tr>
<tr>
<td colspan="6">Open-source Models (Omni-modal)</td>
</tr>
<tr>
<td>VITA</td>
<td>8x7B</td>
<td>33.1</td>
<td>31.8</td>
<td>42.0</td>
<td>44.2</td>
</tr>
<tr>
<td>VITA-1.5</td>
<td>7B</td>
<td>33.4</td>
<td>29.6</td>
<td>48.5</td>
<td><b>47.2<br></td>
</tr>
<tr>
<td>Baichuan-Omni</td>
<td>7B</td>
<td>32.2</td>
<td>26.5</td>
<td>42.6</td>
<td>44.2</td>
</tr>
<tr>
<td>MiniCPM-o 2.6</td>
<td>7B</td>
<td>40.5</td>
<td>30.8</td>
<td><b>53.2<br></td>
<td>46.3</td>
</tr>
<tr>
<td><b>Baichuan-Omni-1.5<br></td>
<td>7B</td>
<td><b>42.9<br></td>
<td><b>37.7<br></td>
<td>47.9</td>
<td>46.9</td>
</tr>
</tbody>
</table>
</div>
</details>
<details>
<summary>click to view</summary>
#### Medical Image Understanding Capabilities
<div align="center">
<table style="margin: 0 auto; text-align: center;">
<thead>
<tr>
<th colspan="7">Medical Understanding </th>
</tr>
</thead>
<tbody>
<tr>
<td>Model</td>
<td>Size</td>
<td>GMAI-MMB-VAL (Acc.)</td>
<td>OpenMM-Medical (Acc.)</td>
</tr>
</thead>
<tr>
<td colspan="4">Proprietary Models</td>
</tr>
<tr>
<td>GPT4o-mini</td>
<td>-</td>
<td>46.4</td>
<td>74.3</td>
</tr>
<tr>
<td colspan="4">Open-source Models (Vision-Language)</td>
</tr>
<tr>
<td>Qwen2 VL</td>
<td>7B</td>
<td>46.3</td>
<td>76.9</td>
</tr>
<tr>
<td>Qwen2 VL</td>
<td>72B</td>
<td><b>50.7<br></td>
<td>80.7</td>
</tr>
<tr>
<td colspan="4">Open-source Models (Omni-modal)</td>
</tr>
<tr>
<td>VITA-1.5</td>
<td>7B</td>
<td>36.7</td>
<td>67.1</td>
</tr>
<tr>
<td>MiniCPM-o 2.6</td>
<td>7B</td>
<td>41.5</td>
<td>73.6</td>
</tr>
<tr>
<td><b>Baichuan-Omni-1.5<br></td>
<td>7B</td>
<td>49.9</td>
<td><b>83.8<br></td>
</tr>
</tbody>
</table>
</div>
</details>
## Examples
<br>
<div style="display: flex; flex-direction: column; align-items: center;">
<img src="https://github.com/baichuan-inc/Baichuan-Omni-1.5/raw/main/assets/pipeline.png" alt="pipeline" style="margin-bottom: 5px;">
<img src="https://github.com/baichuan-inc/Baichuan-Omni-1.5/raw/main/assets/math.png" alt="math" style="margin-bottom: 5px;">
<img src="https://github.com/baichuan-inc/Baichuan-Omni-1.5/raw/main/assets/fly_bill.png" alt="fly_bill" style="margin-bottom: 5px;">
</div>
## 🚀 Quick Start
We recommend interested scholars to visit our github repo for more details. [**Github**](https://github.com/baichuan-inc/Baichuan-Omni-1.5/)
### Statement
- We hereby declare that our team has not developed any applications based on Baichuan-Omni-1.5/Baichuan-Omni-1.5-base models, not on iOS, Android, the web, or any other platform. We strongly call on all users not to use Baichuan-Omni-1.5/Baichuan-Omni-1.5-base models for any activities that harm national / social security or violate the law. Also, we ask users not to use Baichuan-Omni-1.5/Baichuan-Omni-1.5-base models for Internet services that have not undergone appropriate security reviews and filings. We hope that all users can abide by this principle and ensure that the development of technology proceeds in a regulated and legal environment.
- We have done our best to ensure the compliance of the data used in the model training process. However, despite our considerable efforts, there may still be some unforeseeable issues due to the complexity of the model and data. Therefore, if any problems arise due to the use of Baichuan-Omni-1.5/Baichuan-Omni-1.5-base open-source models, including but not limited to data security issues, public opinion risks, or any risks and problems brought about by the model being misled, abused, spread or improperly exploited, we will not assume any responsibility.
### License
The community usage of Baichuan-Omni-1.5/Baichuan-Omni-1.5-base requires adherence to [Apache 2.0](https://github.com/baichuan-inc/Baichuan-Omni-1.5/blob/main/LICENSE) and [Community License for Baichuan-Omni-1.5 Models](https://github.com/baichuan-inc/Baichuan-Omni-1.5/blob/main/LICENSE). The Baichuan-Omni-1.5/Baichuan-Omni-1.5-base models supports commercial use. If you plan to use the Baichuan-Omni-1.5/Baichuan-Omni-1.5-base models or its derivatives for commercial purposes, please ensure that your entity meets the following conditions:
1. The Daily Active Users (DAU) of your or your affiliate's service or product is less than 1 million.
2. Neither you nor your affiliates are software service providers or cloud service providers.
3. There is no possibility for you or your affiliates to grant the commercial license given to you, to reauthorize it to other third parties without Baichuan's permission.
Upon meeting the above conditions, you need to submit the application materials required by the Baichuan-Omni-1.5 Model Community License Agreement via the following contact email: [email protected]. Once approved, Baichuan will hereby grant you a non-exclusive, global, non-transferable, non-sublicensable, revocable commercial copyright license.
<!-- ### Citation
If you find our work helpful, please consider citing our papers 📝 and liking this project ❤️！
```bib
@article{
} -->
```
|