File size: 71,566 Bytes
6e70ed0 5869b18 6e70ed0 5869b18 6e70ed0 5869b18 6e70ed0 5869b18 6e70ed0 5869b18 cb9dd07 6e70ed0 bbead0c 12279a7 6e70ed0 12279a7 6e70ed0 12279a7 6e70ed0 12279a7 6e70ed0 12279a7 6e70ed0 12279a7 6e70ed0 bbead0c 6e70ed0 bbead0c 6e70ed0 bbead0c 6e70ed0 426509c 6e70ed0 bbead0c 6e70ed0 5869b18 6e70ed0 bbead0c 6e70ed0 bbead0c 426509c bbead0c 6e70ed0 bbead0c 6e70ed0 bbead0c 6e70ed0 5869b18 6e70ed0 5869b18 6e70ed0 5869b18 6e70ed0 5869b18 6e70ed0 5869b18 6e70ed0 5869b18 6e70ed0 5869b18 6e70ed0 5869b18 6e70ed0 5869b18 6e70ed0 5869b18 6e70ed0 5869b18 6e70ed0 5869b18 6e70ed0 5869b18 6e70ed0 5869b18 6e70ed0 5869b18 6e70ed0 5869b18 6e70ed0 5869b18 6e70ed0 5869b18 6e70ed0 5869b18 6e70ed0 bbead0c 6e70ed0 bbead0c 6e70ed0 bbead0c 6e70ed0 5869b18 6e70ed0 bbead0c 6e70ed0 bbead0c 6e70ed0 bbead0c 6e70ed0 bbead0c 6e70ed0 5869b18 6e70ed0 5869b18 6e70ed0 5869b18 |
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 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 |
---
base_model:
- OpenGVLab/InternVL2_5-78B
datasets:
- OpenGVLab/MMPR-v1.1
language:
- multilingual
library_name: transformers
license: mit
license_name: qwen
license_link: https://huggingface.co/Qwen/Qwen2.5-72B-Instruct/blob/main/LICENSE
pipeline_tag: image-text-to-text
tags:
- internvl
- custom_code
base_model_relation: finetune
---
# InternVL2_5-78B-MPO
## Paper
The model was presented in the paper [Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling](https://huggingface.co/papers/2412.05271).
### Abstract
We introduce InternVL 2.5, an advanced multimodal large language model (MLLM) series that builds upon InternVL 2.0, maintaining its core model architecture while introducing significant enhancements in training and testing strategies as well as data quality. In this work, we delve into the relationship between model scaling and performance, systematically exploring the performance trends in vision encoders, language models, dataset sizes, and test-time configurations. Through extensive evaluations on a wide range of benchmarks, including multi-discipline reasoning, document understanding, multi-image / video understanding, real-world comprehension, multimodal hallucination detection, visual grounding, multilingual capabilities, and pure language processing, InternVL 2.5 exhibits competitive performance, rivaling leading commercial models such as GPT-4o and Claude-3.5-Sonnet. Notably, our model is the first open-source MLLMs to surpass 70% on the MMMU benchmark, achieving a 3.7-point improvement through Chain-of-Thought (CoT) reasoning and showcasing strong potential for test-time scaling. We hope this model contributes to the open-source community by setting new standards for developing and applying multimodal AI systems. HuggingFace demo see this https URL
---
[\[📂 GitHub\]](https://github.com/OpenGVLab/InternVL) [\[📜 InternVL 1.0\]](https://huggingface.co/papers/2312.14238) [\[📜 InternVL 1.5\]](https://huggingface.co/papers/2404.16821) [\[📜 InternVL 2.5\]](https://huggingface.co/papers/2412.05271) [\[📜 InternVL2.5-MPO\]](https://huggingface.co/papers/2411.10442)
[\[🆕 Blog\]](https://internvl.github.io/blog/) [\[🗨️ Chat Demo\]](https://internvl.opengvlab.com/) [\[🤗 HF Demo\]](https://huggingface.co/spaces/OpenGVLab/InternVL) [\[🚀 Quick Start\]](#quick-start) [\[📖 Documents\]](https://internvl.readthedocs.io/en/latest/)
<div align="center">
<img width="500" alt="image" src="https://cdn-uploads.huggingface.co/production/uploads/64006c09330a45b03605bba3/zJsd2hqd3EevgXo6fNgC-.png">
</div>
## Introduction
We introduce InternVL2.5-MPO, an advanced multimodal large language model (MLLM) series that demonstrates superior overall performance. This series builds upon InternVL2.5 and Mixed Preference Optimization.
## InternVL 2.5 Family
In the following table, we provide an overview of the InternVL2.5-MPO series.
| Model Name | Vision Part | Language Part | HF Link |
| :-----------------: | :-------------------------------------------------------------------------------------: | :----------------------------------------------------------------------------: | :------------------------------------------------------------: |
| InternVL2_5-1B-MPO | [InternViT-300M-448px-V2_5](https://huggingface.co/OpenGVLab/InternViT-300M-448px-V2_5) | [Qwen2.5-0.5B-Instruct](https://huggingface.co/Qwen/Qwen2.5-0.5B-Instruct) | [🤗 link](https://huggingface.co/OpenGVLab/InternVL2_5-1B-MPO) |
| InternVL2_5-2B-MPO | [InternViT-300M-448px-V2_5](https://huggingface.co/OpenGVLab/InternViT-300M-448px-V2_5) | [internlm2_5-1_8b-chat](https://huggingface.co/internlm/internlm2_5-1_8b-chat) | [🤗 link](https://huggingface.co/OpenGVLab/InternVL2_5-2B-MPO) |
| InternVL2_5-4B-MPO | [InternViT-300M-448px-V2_5](https://huggingface.co/OpenGVLab/InternViT-300M-448px-V2_5) | [Qwen2.5-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-3B-Instruct) | [🤗 link](https://huggingface.co/OpenGVLab/InternVL2_5-4B-MPO) |
| InternVL2_5-8B-MPO | [InternViT-300M-448px-V2_5](https://huggingface.co/OpenGVLab/InternViT-300M-448px-V2_5) | [internlm2_5-7b-chat](https://huggingface.co/internlm/internlm2_5-7b-chat) | [🤗 link](https://huggingface.co/OpenGVLab/InternVL2_5-8B-MPO) |
| InternVL2_5-26B-MPO | [InternViT-6B-448px-V2_5](https://huggingface.co/OpenGVLab/InternViT-6B-448px-V2_5) | [internlm2_5-20b-chat](https://huggingface.co/internlm/internlm2_5-20b-chat) | [🤗 link](https://huggingface.co/OpenGVLab/InternVL2_5-26B-MPO) |
| InternVL2_5-38B-MPO | [InternViT-6B-448px-V2_5](https://huggingface.co/OpenGVLab/InternViT-6B-448px-V2_5) | [Qwen2.5-32B-Instruct](https://huggingface.co/Qwen/Qwen2.5-32B-Instruct) | [🤗 link](https://huggingface.co/OpenGVLab/InternVL2_5-38B-MPO) |
| InternVL2_5-78B-MPO | [InternViT-6B-448px-V2_5](https://huggingface.co/OpenGVLab/InternViT-6B-448px-V2_5) | [Qwen2.5-72B-Instruct](https://huggingface.co/Qwen/Qwen2.5-72B-Instruct) | [🤗 link](https://huggingface.co/OpenGVLab/InternVL2_5-78B-MPO) |
## Model Architecture
As shown in the following figure, [InternVL2.5-MPO](https://internvl.github.io/blog/2024-12-20-InternVL-2.5-MPO/) retains the same model architecture as [InternVL 2.5](https://internvl.github.io/blog/2024-12-05-InternVL-2.5/) and its predecessors, InternVL 1.5 and 2.0, following the "ViT-MLP-LLM" paradigm. In this new version, we integrate a newly incrementally pre-trained InternViT with various pre-trained LLMs, including InternLM 2.5 and Qwen 2.5, using a randomly initialized MLP projector.
As in the previous version, we applied a pixel unshuffle operation, reducing the number of visual tokens to one-quarter of the original. Besides, we adopted a similar dynamic resolution strategy as InternVL 1.5, dividing images into tiles of 448×448 pixels. The key difference, starting from InternVL 2.0, is that we additionally introduced support for multi-image and video data.
## Key Designs
### Multi-Modal Preference Dataset
MMPR is a large-scale and high-quality multimodal reasoning preference dataset. This dataset includes about 3 million samples.
To construct this dataset, we propose an efficient data construction pipeline. Specifically, we categorize the multimodal data into **samples with clear ground truths** and **samples without clear ground truths**.
- **For samples with clear ground truths:**
the model is prompted to first provide the reasoning process and then give the final answer in the format like `Final Answer: ***`.
Responses matching the ground truth answer constitute the positive set \\(\mathcal{Y}_p\\), while those that do not match make up the negative set \\(\mathcal{Y}_n\\). Additionally, responses that fail to provide a clear final answer are also merged into \\(\mathcal{Y}_n\\).
Given these responses labeled as positive or negative, we build the preference pairs by selecting a chosen response \\(y_c\\) from \\(\mathcal{Y}_p\\) and a negative response \\(y_r\\) from \\(\mathcal{Y}_n\\).
- **For samples without clear ground truths:**
we propose a simple yet effective method: Dropout Next-Token Prediction (Dropout NTP).
Specifically, we use the responses generated by InternVL2-8B as chosen answers.
Given the chosen answer, we truncate it by half and then prompt InternVL2-8B to complete the remaining
portion of the truncated answer without access to the image input.
This generated completion serves as the rejected answer for the paired sample.
It is worth noting that while the responses generated by InternVL2-8B may not be perfect,
the completions generated without the image input will introduce more hallucinations than those
generated with the image input.
Therefore, the partial order relationship between the chosen and rejected responses holds true.
The data construction pipeline is open-sourced, see more details in our [document](https://internvl.readthedocs.io/en/latest/internvl2.0/preference_optimization.html#generate-additional-preference-data).
### Mixed Preference Optimization
The key insight behind MPO is that *an effective PO process should enable the model to learn the relative preference between pairs of responses, the absolute quality of individual responses, and the process for generating preferred responses.* We define the training objective as a combination of
preference loss \\(\mathcal{L}_{\text{p}}\\),
quality loss \\(\mathcal{L}_{\text{q}}\\),
and generation loss \\(\mathcal{L}_{\text{g}}\\),
referred to as Mixed Preference Optimization:
$$
\mathcal{L}=w_{p}\cdot\mathcal{L}_{\text{p}} + w_{q}\cdot\mathcal{L}_{\text{q}} + w_{g}\cdot\mathcal{L}_{\text{g}},
$$
where \\(w_{*}\\) represents the weight assigned to each loss component.
In this work, we empirically compare different variants of preference loss.
Based on the experimental results, we use DPO as our preference loss and BCO as our quality loss.
Specifically, the DPO serves as the preference loss to enable the model to learn the
relative preference between chosen and rejected responses.
This algorithm optimizes the following loss function:
$$
\mathcal{L}_{\text{p}}=-\log \sigma\left(\beta \log \frac{\pi_\theta\left(y_c \mid x\right)}{\pi_0\left(y_c \mid x\right)}-\beta \log \frac{\pi_\theta\left(y_r \mid x\right)}{\pi_0\left(y_r \mid x\right)}\right),
$$
where \\(\beta\\) is the KL penalty coefficient, and \\(x\\), \\(y_c\\), and \\(y_r\\) are user query, chosen response, and rejected response, respectively.
The policy model \\(\pi_\theta\\) is initialized from model \\(\pi_0\\).
Additionally, the BCO loss is employed as the quality loss, which helps the model to understand the absolute quality of individual responses.
The loss function is defined as:
$$
\mathcal{L}_{\text{q}}=\mathcal{L}_{\text{q}}^+ + \mathcal{L}_{\text{q}}^-,
$$
where \\(\mathcal{L}_{\text{q}}^{+}\\) and \\(\mathcal{L}_{\text{q}}^{+}\\) represent the loss for chosen and rejected responses, respectively.
Each response type's loss is calculated independently, requiring the model to differentiate the absolute quality of individual responses. The loss terms are given by:
$$
\mathcal{L}_{\text{q}}^+=-\log \sigma\left(\beta \log \frac{\pi_\theta\left(y_c \mid x\right)}{\pi_0\left(y_c \mid x\right)} - \delta\right),
$$
$$
\mathcal{L}_{\text{q}}^-=-\log \sigma\left(-\left(\beta \log \frac{\pi_\theta\left(y_r \mid x\right)}{\pi_0\left(y_r \mid x\right)} - \delta\right) \right),
$$
where \\(\delta\\) represents the reward shift, calculated as the moving average of previous rewards to stabilize training.
Finally, the SFT loss is used as the generation loss to help the model learn the generation process of preferred responses.
The loss function is defined as:
$$
\mathcal{L}_{\text{gen}}=-\frac{\log\pi_\theta\left(y_c \mid x\right)}{\left| y_c \right|}.
$$
## Evaluation on Multimodal Capability
To comprehensively compare InternVL's performance before and after MPO, we employ the benchmarks from OpenCompass Learderboard, including both well-established classic datasets and newly introduced ones. These benchmarks span a wide range of categories, aiming to provide a thorough and balanced assessment of InternVL’s capabilities across various multimodal tasks. We provide the evaluation results in the tables behind.
| Model | Avg. | MMBench v1.1 | MMStar | MMMU | MathVista | HallusionBench | AI2D | OCRBench | MMVet |
| ------------------- | ---- | ------------ | ------ | ---- | --------- | -------------- | ---- | -------- | ----- |
| InternVL2-5-1B | 54.9 | 66.5 | 51.3 | 41.2 | 47.1 | 39.4 | 69.0 | 77.4 | 47.2 |
| InternVL2-5-1B-MPO | 56.4 | 67.2 | 49.7 | 40.8 | 53.0 | 40.0 | 69.4 | 83.6 | 47.2 |
| InternVL2-5-2B | 59.9 | 70.9 | 54.3 | 43.2 | 51.1 | 42.3 | 74.9 | 80.2 | 62.6 |
| InternVL2-5-2B-MPO | 62.0 | 71.6 | 55.0 | 45.0 | 56.4 | 43.0 | 75.3 | 84.2 | 65.4 |
| InternVL2-5-4B | 65.1 | 78.2 | 58.7 | 51.8 | 60.8 | 46.6 | 81.4 | 82.0 | 61.5 |
| InternVL2-5-4B-MPO | 67.6 | 78.6 | 60.2 | 51.6 | 65.3 | 47.8 | 82.0 | 88.0 | 67.1 |
| InternVL2-5-8B | 68.9 | 82.5 | 63.2 | 56.2 | 64.5 | 49.0 | 84.6 | 82.1 | 62.8 |
| InternVL2-5-8B-MPO | 70.4 | 82.4 | 65.7 | 54.9 | 68.9 | 51.4 | 84.5 | 88.3 | 66.9 |
| InternVL2-5-26B | 71.6 | 84.6 | 66.5 | 60.7 | 68.0 | 55.8 | 86.2 | 85.4 | 65.4 |
| InternVL2-5-26B-MPO | 72.7 | 84.2 | 67.2 | 57.7 | 72.8 | 55.3 | 86.2 | 91.2 | 67.1 |
| InternVL2-5-38B | 73.5 | 85.4 | 68.5 | 64.6 | 72.4 | 57.9 | 87.6 | 84.1 | 67.2 |
| InternVL2-5-38B-MPO | 75.5 | 85.6 | 69.8 | 64.1 | 73.8 | 61.5 | 88.1 | 88.5 | 72.5 |
| InternVL2-5-78B | 75.2 | 87.5 | 69.5 | 70.0 | 70.6 | 57.4 | 89.1 | 85.3 | 71.8 |
| InternVL2-5-78B-MPO | 76.6 | 87.3 | 73.1 | 68.3 | 73.8 | 58.7 | 89.3 | 91.2 | 71.4 |
## Quick Start
We provide an example code to run `InternVL2_5-78B-MPO` using `transformers`.
> Please use transformers>=4.37.2 to ensure the model works normally.
### Model Loading
#### 16-bit (bf16 / fp16)
```python
import torch
from transformers import AutoTokenizer, AutoModel
path = "OpenGVLab/InternVL2_5-78B-MPO"
model = AutoModel.from_pretrained(
path,
torch_dtype=torch.bfloat16,
low_cpu_mem_usage=True,
use_flash_attn=True,
trust_remote_code=True).eval().cuda()
```
#### BNB 8-bit Quantization
```python
import torch
from transformers import AutoTokenizer, AutoModel
path = "OpenGVLab/InternVL2_5-78B-MPO"
model = AutoModel.from_pretrained(
path,
torch_dtype=torch.bfloat16,
load_in_8bit=True,
low_cpu_mem_usage=True,
use_flash_attn=True,
trust_remote_code=True).eval()
```
#### Multiple GPUs
The reason for writing the code this way is to avoid errors that occur during multi-GPU inference due to tensors not being on the same device. By ensuring that the first and last layers of the large language model (LLM) are on the same device, we prevent such errors.
```python
import math
import torch
from transformers import AutoTokenizer, AutoModel
def split_model(model_name):
device_map = {}
world_size = torch.cuda.device_count()
num_layers = {
'InternVL2_5-1B': 24, 'InternVL2_5-2B': 24, 'InternVL2_5-4B': 36, 'InternVL2_5-8B': 32,
'InternVL2_5-26B': 48, 'InternVL2_5-38B': 64, 'InternVL2_5-78B': 80}[model_name]
# Since the first GPU will be used for ViT, treat it as half a GPU.
num_layers_per_gpu = math.ceil(num_layers / (world_size - 0.5))
num_layers_per_gpu = [num_layers_per_gpu] * world_size
num_layers_per_gpu[0] = math.ceil(num_layers_per_gpu[0] * 0.5)
layer_cnt = 0
for i, num_layer in enumerate(num_layers_per_gpu):
for j in range(num_layer):
device_map[f'language_model.model.layers.{layer_cnt}'] = i
layer_cnt += 1
device_map['vision_model'] = 0
device_map['mlp1'] = 0
device_map['language_model.model.tok_embeddings'] = 0
device_map['language_model.model.embed_tokens'] = 0
device_map['language_model.output'] = 0
device_map['language_model.model.norm'] = 0
device_map['language_model.model.rotary_emb'] = 0
device_map['language_model.lm_head'] = 0
device_map[f'language_model.model.layers.{num_layers - 1}'] = 0
return device_map
path = "OpenGVLab/InternVL2_5-78B-MPO"
device_map = split_model('InternVL2_5-78B')
model = AutoModel.from_pretrained(
path,
torch_dtype=torch.bfloat16,
load_in_8bit=False,
low_cpu_mem_usage=True,
use_flash_attn=True,
trust_remote_code=True,
device_map=device_map).eval()
```
### Inference with Transformers
```python
import math
import numpy as np
import torch
import torchvision.transforms as T
from decord import VideoReader, cpu
from PIL import Image
from torchvision.transforms.functional import InterpolationMode
from transformers import AutoModel, AutoTokenizer
IMAGENET_MEAN = (0.485, 0.456, 0.406)
IMAGENET_STD = (0.229, 0.224, 0.225)
def build_transform(input_size):
MEAN, STD = IMAGENET_MEAN, IMAGENET_STD
transform = T.Compose([
T.Lambda(lambda img: img.convert('RGB') if img.mode != 'RGB' else img),
T.Resize((input_size, input_size), interpolation=InterpolationMode.BICUBIC),
T.ToTensor(),
T.Normalize(mean=MEAN, std=STD)
])
return transform
def find_closest_aspect_ratio(aspect_ratio, target_ratios, width, height, image_size):
best_ratio_diff = float('inf')
best_ratio = (1, 1)
area = width * height
for ratio in target_ratios:
target_aspect_ratio = ratio[0] / ratio[1]
ratio_diff = abs(aspect_ratio - target_aspect_ratio)
if ratio_diff < best_ratio_diff:
best_ratio_diff = ratio_diff
best_ratio = ratio
elif ratio_diff == best_ratio_diff:
if area > 0.5 * image_size * image_size * ratio[0] * ratio[1]:
best_ratio = ratio
return best_ratio
def dynamic_preprocess(image, min_num=1, max_num=12, image_size=448, use_thumbnail=False):
orig_width, orig_height = image.size
aspect_ratio = orig_width / orig_height
# calculate the existing image aspect ratio
target_ratios = set(
(i, j) for n in range(min_num, max_num + 1) for i in range(1, n + 1) for j in range(1, n + 1) if
i * j <= max_num and i * j >= min_num)
target_ratios = sorted(target_ratios, key=lambda x: x[0] * x[1])
# find the closest aspect ratio to the target
target_aspect_ratio = find_closest_aspect_ratio(
aspect_ratio, target_ratios, orig_width, orig_height, image_size)
# calculate the target width and height
target_width = image_size * target_aspect_ratio[0]
target_height = image_size * target_aspect_ratio[1]
blocks = target_aspect_ratio[0] * target_aspect_ratio[1]
# resize the image
resized_img = image.resize((target_width, target_height))
processed_images = []
for i in range(blocks):
box = (
(i % (target_width // image_size)) * image_size,
(i // (target_width // image_size)) * image_size,
((i % (target_width // image_size)) + 1) * image_size,
((i // (target_width // image_size)) + 1) * image_size
)
# split the image
split_img = resized_img.crop(box)
processed_images.append(split_img)
assert len(processed_images) == blocks
if use_thumbnail and len(processed_images) != 1:
thumbnail_img = image.resize((image_size, image_size))
processed_images.append(thumbnail_img)
return processed_images
def load_image(image_file, input_size=448, max_num=12):
image = Image.open(image_file).convert('RGB')
transform = build_transform(input_size=input_size)
images = dynamic_preprocess(image, image_size=input_size, use_thumbnail=True, max_num=max_num)
pixel_values = [transform(image) for image in images]
pixel_values = torch.stack(pixel_values)
return pixel_values
def split_model(model_name):
device_map = {}
world_size = torch.cuda.device_count()
num_layers = {
'InternVL2_5-1B': 24, 'InternVL2_5-2B': 24, 'InternVL2_5-4B': 36, 'InternVL2_5-8B': 32,
'InternVL2_5-26B': 48, 'InternVL2_5-38B': 64, 'InternVL2_5-78B': 80}[model_name]
# Since the first GPU will be used for ViT, treat it as half a GPU.
num_layers_per_gpu = math.ceil(num_layers / (world_size - 0.5))
num_layers_per_gpu = [num_layers_per_gpu] * world_size
num_layers_per_gpu[0] = math.ceil(num_layers_per_gpu[0] * 0.5)
layer_cnt = 0
for i, num_layer in enumerate(num_layers_per_gpu):
for j in range(num_layer):
device_map[f'language_model.model.layers.{layer_cnt}'] = i
layer_cnt += 1
device_map['vision_model'] = 0
device_map['mlp1'] = 0
device_map['language_model.model.tok_embeddings'] = 0
device_map['language_model.model.embed_tokens'] = 0
device_map['language_model.output'] = 0
device_map['language_model.model.norm'] = 0
device_map['language_model.model.rotary_emb'] = 0
device_map['language_model.lm_head'] = 0
device_map[f'language_model.model.layers.{num_layers - 1}'] = 0
return device_map
# If you set `load_in_8bit=True`, you will need two 80GB GPUs.
# If you set `load_in_8bit=False`, you will need at least three 80GB GPUs.
path = 'OpenGVLab/InternVL2_5-78B-MPO'
device_map = split_model('InternVL2_5-78B')
model = AutoModel.from_pretrained(
path,
torch_dtype=torch.bfloat16,
load_in_8bit=False,
low_cpu_mem_usage=True,
use_flash_attn=True,
trust_remote_code=True,
device_map=device_map).eval()
tokenizer = AutoTokenizer.from_pretrained(path, trust_remote_code=True, use_fast=False)
# set the max number of tiles in `max_num`
pixel_values = load_image('./examples/image1.jpg', max_num=12).to(torch.bfloat16).cuda()
generation_config = dict(max_new_tokens=1024, do_sample=True)
# pure-text conversation (纯文本对话)
question = 'Hello, who are you?'
response, history = model.chat(tokenizer, None, question, generation_config, history=None, return_history=True)
print(f'User: {question}
Assistant: {response}')
question = 'Can you tell me a story?'
response, history = model.chat(tokenizer, None, question, generation_config, history=history, return_history=True)
print(f'User: {question}
Assistant: {response}')
# single-image single-round conversation (单图单轮对话)
question = '<image>
Please describe the image shortly.'
response = model.chat(tokenizer, pixel_values, question, generation_config)
print(f'User: {question}
Assistant: {response}')
# single-image multi-round conversation (单图多轮对话)
question = '<image>
Please describe the image in detail.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config, history=None, return_history=True)
print(f'User: {question}
Assistant: {response}')
question = 'Please write a poem according to the image.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config, history=history, return_history=True)
print(f'User: {question}
Assistant: {response}')
# multi-image multi-round conversation, combined images (多图多轮对话，拼接图像)
pixel_values1 = load_image('./examples/image1.jpg', max_num=12).to(torch.bfloat16).cuda()
pixel_values2 = load_image('./examples/image2.jpg', max_num=12).to(torch.bfloat16).cuda()
pixel_values = torch.cat((pixel_values1, pixel_values2), dim=0)
question = '<image>
Describe the two images in detail.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config,
history=None, return_history=True)
print(f'User: {question}
Assistant: {response}')
question = 'What are the similarities and differences between these two images.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config,
history=history, return_history=True)
print(f'User: {question}
Assistant: {response}')
# multi-image multi-round conversation, separate images (多图多轮对话，独立图像)
pixel_values1 = load_image('./examples/image1.jpg', max_num=12).to(torch.bfloat16).cuda()
pixel_values2 = load_image('./examples/image2.jpg', max_num=12).to(torch.bfloat16).cuda()
pixel_values = torch.cat((pixel_values1, pixel_values2), dim=0)
num_patches_list = [pixel_values1.size(0), pixel_values2.size(0)]
question = 'Image-1: <image>
Image-2: <image>
Describe the two images in detail.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config,
num_patches_list=num_patches_list,
history=None, return_history=True)
print(f'User: {question}
Assistant: {response}')
question = 'What are the similarities and differences between these two images.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config,
num_patches_list=num_patches_list,
history=history, return_history=True)
print(f'User: {question}
Assistant: {response}')
# batch inference, single image per sample (单图批处理)
pixel_values1 = load_image('./examples/image1.jpg', max_num=12).to(torch.bfloat16).cuda()
pixel_values2 = load_image('./examples/image2.jpg', max_num=12).to(torch.bfloat16).cuda()
num_patches_list = [pixel_values1.size(0), pixel_values2.size(0)]
pixel_values = torch.cat((pixel_values1, pixel_values2), dim=0)
questions = ['<image>
Describe the image in detail.'] * len(num_patches_list)
responses = model.batch_chat(tokenizer, pixel_values,
num_patches_list=num_patches_list,
questions=questions,
generation_config=generation_config)
for question, response in zip(questions, responses):
print(f'User: {question}
Assistant: {response}')
# video multi-round conversation (视频多轮对话)
def get_index(bound, fps, max_frame, first_idx=0, num_segments=32):
if bound:
start, end = bound[0], bound[1]
else:
start, end = -100000, 100000
start_idx = max(first_idx, round(start * fps))
end_idx = min(round(end * fps), max_frame)
seg_size = float(end_idx - start_idx) / num_segments
frame_indices = np.array([
int(start_idx + (seg_size / 2) + np.round(seg_size * idx))
for idx in range(num_segments)
])
return frame_indices
def load_video(video_path, bound=None, input_size=448, max_num=1, num_segments=32):
vr = VideoReader(video_path, ctx=cpu(0), num_threads=1)
max_frame = len(vr) - 1
fps = float(vr.get_avg_fps())
pixel_values_list, num_patches_list = [], []
transform = build_transform(input_size=input_size)
frame_indices = get_index(bound, fps, max_frame, first_idx=0, num_segments=num_segments)
for frame_index in frame_indices:
img = Image.fromarray(vr[frame_index].asnumpy()).convert('RGB')
img = dynamic_preprocess(img, image_size=input_size, use_thumbnail=True, max_num=max_num)
pixel_values = [transform(tile) for tile in img]
pixel_values = torch.stack(pixel_values)
num_patches_list.append(pixel_values.shape[0])
pixel_values_list.append(pixel_values)
pixel_values = torch.cat(pixel_values_list)
return pixel_values, num_patches_list
video_path = './examples/red-panda.mp4'
pixel_values, num_patches_list = load_video(video_path, num_segments=8, max_num=1)
pixel_values = pixel_values.to(torch.bfloat16).cuda()
video_prefix = ''.join([f'Frame{i+1}: <image>
' for i in range(len(num_patches_list))])
question = video_prefix + 'What is the red panda doing?'
# Frame1: <image>
Frame2: <image>
...
Frame8: <image>
{question}
response, history = model.chat(tokenizer, pixel_values, question, generation_config,
num_patches_list=num_patches_list, history=None, return_history=True)
print(f'User: {question}
Assistant: {response}')
question = 'Describe this video in detail.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config,
num_patches_list=num_patches_list, history=history, return_history=True)
print(f'User: {question}
Assistant: {response}')
```
#### Streaming Output
Besides this method, you can also use the following code to get streamed output.
```python
from transformers import TextIteratorStreamer
from threading import Thread
# Initialize the streamer
streamer = TextIteratorStreamer(tokenizer, skip_prompt=True, skip_special_tokens=True, timeout=10)
# Define the generation configuration
generation_config = dict(max_new_tokens=1024, do_sample=False, streamer=streamer)
# Start the model chat in a separate thread
thread = Thread(target=model.chat, kwargs=dict(
tokenizer=tokenizer, pixel_values=pixel_values, question=question,
history=None, return_history=False, generation_config=generation_config,
))
thread.start()
# Initialize an empty string to store the generated text
generated_text = ''
# Loop through the streamer to get the new text as it is generated
for new_text in streamer:
if new_text == model.conv_template.sep:
break
generated_text += new_text
print(new_text, end='', flush=True) # Print each new chunk of generated text on the same line
```
## Finetune
Many repositories now support fine-tuning of the InternVL series models, including [InternVL](https://github.com/OpenGVLab/InternVL), [SWIFT](https://github.com/modelscope/ms-swift), [XTurner](https://github.com/InternLM/xtuner), and others. Please refer to their documentation for more details on fine-tuning.
## Deployment
### LMDeploy
LMDeploy is a toolkit for compressing, deploying, and serving LLMs & VLMs.
```sh
pip install lmdeploy>=0.6.4
```
LMDeploy abstracts the complex inference process of multi-modal Vision-Language Models (VLM) into an easy-to-use pipeline, similar to the Large Language Model (LLM) inference pipeline.
#### A 'Hello, world' Example
```python
from lmdeploy import pipeline, TurbomindEngineConfig
from lmdeploy.vl import load_image
model = 'OpenGVLab/InternVL2_5-78B-MPO'
image = load_image('https://raw.githubusercontent.com/open-mmlab/mmdeploy/main/tests/data/tiger.jpeg')
pipe = pipeline(model, backend_config=TurbomindEngineConfig(session_len=8192, tp=4))
response = pipe(('describe this image', image))
print(response.text)
```
If `ImportError` occurs while executing this case, please install the required dependency packages as prompted.
#### Multi-images Inference
When dealing with multiple images, you can put them all in one list. Keep in mind that multiple images will lead to a higher number of input tokens, and as a result, the size of the context window typically needs to be increased.
```python
from lmdeploy import pipeline, TurbomindEngineConfig
from lmdeploy.vl import load_image
from lmdeploy.vl.constants import IMAGE_TOKEN
model = 'OpenGVLab/InternVL2_5-78B-MPO'
pipe = pipeline(model, backend_config=TurbomindEngineConfig(session_len=8192, tp=4))
image_urls=[
'https://raw.githubusercontent.com/open-mmlab/mmdeploy/main/demo/resources/human-pose.jpg',
'https://raw.githubusercontent.com/open-mmlab/mmdeploy/main/demo/resources/det.jpg'
]
images = [load_image(img_url) for img_url in image_urls]
# Numbering images improves multi-image conversations
response = pipe((f'Image-1: {IMAGE_TOKEN}
Image-2: {IMAGE_TOKEN}
describe these two images', images))
print(response.text)
```
#### Batch Prompts Inference
Conducting inference with batch prompts is quite straightforward; just place them within a list structure:
```python
from lmdeploy import pipeline, TurbomindEngineConfig
from lmdeploy.vl import load_image
model = 'OpenGVLab/InternVL2_5-78B-MPO'
pipe = pipeline(model, backend_config=TurbomindEngineConfig(session_len=8192, tp=4))
image_urls=[
"https://raw.githubusercontent.com/open-mmlab/mmdeploy/main/demo/resources/human-pose.jpg",
"https://raw.githubusercontent.com/open-mmlab/mmdeploy/main/demo/resources/det.jpg"
]
prompts = [('describe this image', load_image(img_url)) for img_url in image_urls]
response = pipe(prompts)
print(response)
```
#### Multi-turn Conversation
There are two ways to do the multi-turn conversations with the pipeline. One is to construct messages according to the format of OpenAI and use above introduced method, the other is to use the `pipeline.chat` interface.
```python
from lmdeploy import pipeline, TurbomindEngineConfig, GenerationConfig
from lmdeploy.vl import load_image
model = 'OpenGVLab/InternVL2_5-78B-MPO'
pipe = pipeline(model, backend_config=TurbomindEngineConfig(session_len=8192, tp=4))
image = load_image('https://raw.githubusercontent.com/open-mmlab/mmdeploy/main/demo/resources/human-pose.jpg')
gen_config = GenerationConfig(top_k=40, top_p=0.8, temperature=0.8)
sess = pipe.chat(('describe this image', image), gen_config=gen_config)
print(sess.response.text)
sess = pipe.chat('What is the woman doing?', session=sess, gen_config=gen_config)
print(sess.response.text)
```
#### Service
LMDeploy's `api_server` enables models to be easily packed into services with a single command. The provided RESTful APIs are compatible with OpenAI's interfaces. Below are an example of service startup:
```shell
lmdeploy serve api_server OpenGVLab/InternVL2_5-78B-MPO --server-port 23333 --tp 4
```
To use the OpenAI-style interface, you need to install OpenAI:
```shell
pip install openai
```
Then, use the code below to make the API call:
```python
from openai import OpenAI
client = OpenAI(api_key='YOUR_API_KEY', base_url='http://0.0.0.0:23333/v1')
model_name = client.models.list().data[0].id
response = client.chat.completions.create(
model=model_name,
messages=[{
'role':
'user',
'content': [{
'type': 'text',
'text': 'describe this image',
}, {
'type': 'image_url',
'image_url': {
'url':
'https://modelscope.oss-cn-beijing.aliyuncs.com/resource/tiger.jpeg',
},
}],
}],
temperature=0.8,
top_p=0.8)
print(response)
```
## License
This project is released under the MIT License. This project uses the pre-trained Qwen2.5-72B-Instruct as a component, which is licensed under the Qwen License.
## Citation
If you find this project useful in your research, please consider citing:
```BibTeX
@article{wang2024mpo,
title={Enhancing the Reasoning Ability of Multimodal Large Language Models via Mixed Preference Optimization},
author={Wang, Weiyun and Chen, Zhe and Wang, Wenhai and Cao, Yue and Liu, Yangzhou and Gao, Zhangwei and Zhu, Jinguo and Zhu, Xizhou and Lu, Lewei and Qiao, Yu and Dai, Jifeng},
journal={arXiv preprint arXiv:2411.10442},
year={2024}
}
@article{chen2024expanding,
title={Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling},
author={Chen, Zhe and Wang, Weiyun and Cao, Yue and Liu, Yangzhou and Gao, Zhangwei and Cui, Erfei and Zhu, Jinguo and Ye, Shenglong and Tian, Hao and Liu, Zhaoyang and others},
journal={arXiv preprint arXiv:2412.05271},
year={2024}
}
@article{chen2024far,
title={How Far Are We to GPT-4V? Closing the Gap to Commercial Multimodal Models with Open-Source Suites},
author={Chen, Zhe and Wang, Weiyun and Tian, Hao and Ye, Shenglong and Gao, Zhangwei and Cui, Erfei and Tong, Wenwen and Hu, Kongzhi and Luo, Jiapeng and Ma, Zheng and others},
journal={Science China Information Sciences},
volume={67},
number={12},
pages={220101},
year={2024},
publisher={Springer}
}
@inproceedings{chen2024internvl,
title={Internvl: Scaling up vision foundation models and aligning for generic visual-linguistic tasks},
author={Chen, Zhe and Wu, Jiannan and Wang, Wenhai and Su, Weijie and Chen, Guo and Xing, Sen and Zhong, Muyan and Zhang, Qinglong and Zhu, Xizhou and Lu, Lewei and others},
booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
pages={24185--24198},
year={2024}
}
```
## What can InternVL do?
<details>
<summary>Visual Perception (click to expand)</summary>
- Linear-Probe Image Classification [[see details]](./classification#-evaluation)
ViT-22B uses the private JFT-3B dataset.
| method | #param | IN-1K | IN-ReaL | IN-V2 | IN-A | IN-R | IN-Sketch |
| ------------------- | :----: | :---: | :-----: | :---: | :---: | :---: | :-------: |
| OpenCLIP-G | 1.8B | 86.2 | 89.4 | 77.2 | 63.8 | 87.8 | 66.4 |
| DINOv2-g | 1.1B | 86.5 | 89.6 | 78.4 | 75.9 | 78.8 | 62.5 |
| EVA-01-CLIP-g | 1.1B | 86.5 | 89.3 | 77.4 | 70.5 | 87.7 | 63.1 |
| MAWS-ViT-6.5B | 6.5B | 87.8 | - | - | - | - | - |
| ViT-22B\* | 21.7B | 89.5 | 90.9 | 83.2 | 83.8 | 87.4 | - |
| InternViT-6B (ours) | 5.9B | 88.2 | 90.4 | 79.9 | 77.5 | 89.8 | 69.1 |
- Semantic Segmentation [[see details]](./segmentation#-evaluation)
| method | decoder | #param (train/total) | crop size | mIoU |
| --------------------- | :-----: | :------------------: | :-------: | ------------ |
| OpenCLIP-G (frozen) | Linear | 0.3M / 1.8B | 512 | 39.3 |
| ViT-22B (frozen) | Linear | 0.9M / 21.7B | 504 | 34.6 |
| InternViT-6B (frozen) | Linear | 0.5M / 5.9B | 504 | 47.2 (+12.6) |
| ViT-22B (frozen) | UperNet | 0.8B / 22.5B | 504 | 52.7 |
| InternViT-6B (frozen) | UperNet | 0.4B / 6.3B | 504 | 54.9 (+2.2) |
| ViT-22B | UperNet | 22.5B / 22.5B | 504 | 55.3 |
| InternViT-6B | UperNet | 6.3B / 6.3B | 504 | 58.9 (+3.6) |
- Zero-Shot Image Classification [[see details]](./clip_benchmark#imagenet-variants-and-objectnet)
| method | IN-1K | IN-A | IN-R | IN-V2 | IN-Sketch | ObjectNet |
| ----------------- | :---: | :---: | :---: | :---: | :-------: | :-------: |
| OpenCLIP-G | 80.1 | 69.3 | 92.1 | 73.6 | 68.9 | 73.0 |
| EVA-02-CLIP-E+ | 82.0 | 82.1 | 94.5 | 75.7 | 71.6 | 79.6 |
| ViT-22B\* | 85.9 | 90.1 | 96.0 | 80.9 | - | 87.6 |
| InternVL-C (ours) | 83.2 | 83.8 | 95.5 | 77.3 | 73.9 | 80.6 |
- Multilingual Zero-Shot Image Classification [[see details]](./clip_benchmark#multilingual-imagenet-1k)
EN: English, ZH: Chinese, JP: Japanese, Ar: Arabic, IT: Italian
| method | IN-1K (EN) | IN-1K (ZH) | IN-1K (JP) | IN-1K (AR) | IN-1K (IT) |
| ----------------- | :--------: | :--------: | :--------: | :--------: | :--------: |
| Taiyi-CLIP-ViT-H | - | 54.4 | - | - | - |
| WuKong-ViT-L-G | - | 57.5 | - | - | - |
| CN-CLIP-ViT-H | - | 59.6 | - | - | - |
| AltCLIP-ViT-L | 74.5 | 59.6 | - | - | - |
| EVA-02-CLIP-E+ | 82.0 | - | - | - | 41.2 |
| OpenCLIP-XLM-R-H | 77.0 | 55.7 | 53.1 | 37.0 | 56.8 |
| InternVL-C (ours) | 83.2 | 64.5 | 61.5 | 44.9 | 65.7 |
- Zero-Shot Video Classification
| method | #frame | K400 | K600 | K700 |
| ----------------- | :----: | :---: | :---: | :---: |
| OpenCLIP-G | 1 | 65.9 | 66.1 | 59.2 |
| EVA-02-CLIP-E+ | 1 | 69.8 | 69.3 | 63.4 |
| InternVL-C (ours) | 1 | 71.0 | 71.3 | 65.7 |
| ViCLIP | 8 | 75.7 | 73.5 | 66.4 |
| InternVL-C (ours) | 8 | 79.4 | 78.8 | 71.5 |
</details>
<details>
<summary>Cross-Modal Retrieval (click to expand)</summary>
- English Zero-Shot Image-Text Retrieval [[see details]](./clip_benchmark#flickr30k--coco)
<table>
<tr align=center>
<td rowspan="3" align=left><b>model</b></td>
<td colspan="6" align=center><b>Flickr30K</b></td>
<td colspan="6" align=center><b>COCO</b></td>
<td rowspan="3" align=center><b>avg</b></td>
</tr>
<tr align=center>
<td colspan="3" align=center><b>image-to-text</b></td>
<td colspan="3" align=center><b>text-to-image</b></td>
<td colspan="3" align=center><b>image-to-text</b></td>
<td colspan="3" align=center><b>text-to-image</b></td>
</tr>
<tr>
<td>R@1</td>
<td>R@5</td>
<td>R@10</td>
<td>R@1</td>
<td>R@5</td>
<td>R@10</td>
<td>R@1</td>
<td>R@5</td>
<td>R@10</td>
<td>R@1</td>
<td>R@5</td>
<td>R@10</td>
</tr>
<tr align=center>
<td align=left>OpenCLIP-G</td>
<td>92.9</td>
<td>99.3</td>
<td>99.8</td>
<td>79.5</td>
<td>95.0</td>
<td>97.1</td>
<td>67.3</td>
<td>86.9</td>
<td>92.6</td>
<td>51.4</td>
<td>74.9</td>
<td>83.0</td>
<td>85.0</td>
</tr>
<tr align=center>
<td align=left>EVA-02-CLIP-E+</td>
<td>93.9</td>
<td>99.4</td>
<td>99.8</td>
<td>78.8</td>
<td>94.2</td>
<td>96.8</td>
<td>68.8</td>
<td>87.8</td>
<td>92.8</td>
<td>51.1</td>
<td>75.0</td>
<td>82.7</td>
<td>85.1</td>
</tr>
<tr align=center>
<td align=left>EVA-CLIP-8B</td>
<td>95.6</td>
<td>99.6</td>
<td>99.9</td>
<td>80.8</td>
<td>95.5</td>
<td>97.6</td>
<td>70.3</td>
<td>89.3</td>
<td>93.9</td>
<td>53.0</td>
<td>76.0</td>
<td>83.4</td>
<td>86.2</td>
</tr>
<tr align=center>
<td align=left>InternVL-C (ours)</td>
<td>94.7</td>
<td>99.6</td>
<td>99.9</td>
<td>81.7</td>
<td>96.0</td>
<td>98.2</td>
<td>70.6</td>
<td>89.0</td>
<td>93.5</td>
<td>54.1</td>
<td>77.3</td>
<td>84.6</td>
<td>86.6</td>
</tr>
<tr align=center>
<td align=left>InternVL-G (ours)</td>
<td>95.7</td>
<td>99.7</td>
<td>99.9</td>
<td>85.0</td>
<td>97.0</td>
<td>98.6</td>
<td>74.9</td>
<td>91.3</td>
<td>95.2</td>
<td>58.6</td>
<td>81.3</td>
<td>88.0</td>
<td>88.8</td>
</tr>
</table>
- Chinese Zero-Shot Image-Text Retrieval [[see details]](./clip_benchmark#flickr30k-cn--coco-cn)
<table>
<tr align=center>
<td rowspan="3" align=left><b>model</b></td>
<td colspan="6" align=center><b>Flickr30K-CN</b></td>
<td colspan="6" align=center><b>COCO-CN</b></td>
<td rowspan="3" align=center><b>avg</b></td>
</tr>
<tr align=center>
<td colspan="3" align=center><b>image-to-text</b></td>
<td colspan="3" align=center><b>text-to-image</b></td>
<td colspan="3" align=center><b>image-to-text</b></td>
<td colspan="3" align=center><b>text-to-image</b></td>
</tr>
<tr>
<td>R@1</td>
<td>R@5</td>
<td>R@10</td>
<td>R@1</td>
<td>R@5</td>
<td>R@10</td>
<td>R@1</td>
<td>R@5</td>
<td>R@10</td>
<td>R@1</td>
<td>R@5</td>
<td>R@10</td>
</tr>
<tr align=center>
<td align=left>CN-CLIP-ViT-H</td>
<td>81.6</td>
<td>97.5</td>
<td>98.8</td>
<td>71.2</td>
<td>91.4</td>
<td>95.5</td>
<td>63.0</td>
<td>86.6</td>
<td>92.9</td>
<td>69.2</td>
<td>89.9</td>
<td>96.1</td>
<td>86.1</td>
</tr>
<tr align=center>
<td align=left>OpenCLIP-XLM-R-H</td>
<td>86.1</td>
<td>97.5</td>
<td>99.2</td>
<td>71.0</td>
<td>90.5</td>
<td>94.9</td>
<td>70.0</td>
<td>91.5</td>
<td>97.0</td>
<td>66.1</td>
<td>90.8</td>
<td>96.0</td>
<td>87.6</td>
</tr>
<tr align=center>
<td align=left>InternVL-C (ours)</td>
<td>90.3</td>
<td>98.8</td>
<td>99.7</td>
<td>75.1</td>
<td>92.9</td>
<td>96.4</td>
<td>68.8</td>
<td>92.0</td>
<td>96.7</td>
<td>68.9</td>
<td>91.9</td>
<td>96.5</td>
<td>89.0</td>
</tr>
<tr align=center>
<td align=left>InternVL-G (ours)</td>
<td>92.9</td>
<td>99.4</td>
<td>99.8</td>
<td>77.7</td>
<td>94.8</td>
<td>97.3</td>
<td>71.4</td>
<td>93.9</td>
<td>97.7</td>
<td>73.8</td>
<td>94.4</td>
<td>98.1</td>
<td>90.9</td>
</tr>
</table>
- Multilingual Zero-Shot Image-Text Retrieval on XTD [[see details]](./clip_benchmark#xtd)
| method | EN | ES | FR | ZH | IT | KO | RU | JP | average |
| ----------------- | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :-----: |
| AltCLIP | 95.4 | 94.1 | 92.9 | 95.1 | 94.2 | 94.4 | 91.8 | 91.7 | 93.7 |
| OpenCLIP-XLM-R-H | 97.3 | 96.1 | 94.5 | 94.7 | 96.0 | 90.2 | 93.9 | 94.0 | 94.6 |
| InternVL-C (ours) | 97.3 | 95.7 | 95.1 | 95.6 | 96.0 | 92.2 | 93.3 | 95.5 | 95.1 |
| InternVL-G (ours) | 98.6 | 97.7 | 96.5 | 96.7 | 96.9 | 95.1 | 94.8 | 96.1 | 96.6 |
</details>
<details>
<summary>Multimodal Dialogue</summary>
</details>
## Quick Start with HuggingFace
<details>
<summary>using InternViT-6B for visual feature extraction (click to expand)</summary>
```python
import torch
from PIL import Image
from transformers import AutoModel, CLIPImageProcessor
model = AutoModel.from_pretrained(
'OpenGVLab/InternViT-6B-448px-V2_5',
torch_dtype=torch.bfloat16,
low_cpu_mem_usage=True,
trust_remote_code=True).cuda().eval()
image = Image.open('./examples/image1.jpg').convert('RGB')
image_processor = CLIPImageProcessor.from_pretrained('OpenGVLab/InternViT-6B-448px-V1-5')
pixel_values = image_processor(images=image, return_tensors='pt').pixel_values
pixel_values = pixel_values.to(torch.bfloat16).cuda()
outputs = model(pixel_values)
```
</details>
<details>
<summary>using InternVL-C(ontrastive) and InternVL-G(enerative) for cross-modal retrieval (click to expand)</summary>
```python
import torch
from PIL import Image
from transformers import AutoModel, CLIPImageProcessor
from transformers import AutoTokenizer
model = AutoModel.from_pretrained(
'OpenGVLab/InternVL-14B-224px',
torch_dtype=torch.bfloat16,
low_cpu_mem_usage=True,
trust_remote_code=True).cuda().eval()
image_processor = CLIPImageProcessor.from_pretrained('OpenGVLab/InternVL-14B-224px')
tokenizer = AutoTokenizer.from_pretrained(
'OpenGVLab/InternVL-14B-224px', use_fast=False, add_eos_token=True)
tokenizer.pad_token_id = 0 # set pad_token_id to 0
images = [
Image.open('./examples/image1.jpg').convert('RGB'),
Image.open('./examples/image2.jpg').convert('RGB'),
Image.open('./examples/image3.jpg').convert('RGB')
]
prefix = 'summarize:'
texts = [
prefix + 'a photo of a red panda', # English
prefix + '一张熊猫的照片', # Chinese
prefix + '二匹の猫の写真' # Japanese
]
pixel_values = image_processor(images=images, return_tensors='pt').pixel_values
pixel_values = pixel_values.to(torch.bfloat16).cuda()
input_ids = tokenizer(texts, return_tensors='pt', max_length=80,
truncation=True, padding='max_length').input_ids.cuda()
# InternVL-C
logits_per_image, logits_per_text = model(
image=pixel_values, text=input_ids, mode='InternVL-C')
probs = logits_per_image.softmax(dim=-1)
# tensor([[9.9609e-01, 5.2185e-03, 6.0070e-08],
# [2.2949e-02, 9.7656e-01, 5.9903e-06],
# [3.2932e-06, 7.4863e-05, 1.0000e+00]], device='cuda:0',
# dtype=torch.bfloat16, grad_fn=<SoftmaxBackward0>)
# InternVL-G
logits_per_image, logits_per_text = model(
image=pixel_values, text=input_ids, mode='InternVL-G')
probs = logits_per_image.softmax(dim=-1)
# tensor([[9.9609e-01, 3.1738e-03, 3.6322e-08],
# [8.6060e-03, 9.9219e-01, 2.8759e-06],
# [1.7583e-06, 3.1233e-05, 1.0000e+00]], device='cuda:0',
# dtype=torch.bfloat16, grad_fn=<SoftmaxBackward0>)
# please set add_eos_token to False for generation
tokenizer.add_eos_token = False
image = Image.open('./examples/image1.jpg').convert('RGB')
pixel_values = image_processor(images=image, return_tensors='pt').pixel_values
pixel_values = pixel_values.to(torch.bfloat16).cuda()
tokenized = tokenizer("English caption:", return_tensors='pt')
pred = model.generate(
pixel_values=pixel_values,
input_ids=tokenized.input_ids.cuda(),
attention_mask=tokenized.attention_mask.cuda(),
num_beams=5,
min_new_tokens=8,
)
caption = tokenizer.decode(pred[0].cpu(), skip_special_tokens=True).strip()
# English caption: a red panda sitting on top of a wooden platform
```
</details>
<details>
<summary>using InternVL 2.5 for multimodal chat (click to expand)</summary>
Here, we take the smaller `OpenGVLab/InternVL2_5-8B` as an example:
```python
import numpy as np
import torch
import torchvision.transforms as T
from decord import VideoReader, cpu
from PIL import Image
from torchvision.transforms.functional import InterpolationMode
from transformers import AutoModel, AutoTokenizer
IMAGENET_MEAN = (0.485, 0.456, 0.406)
IMAGENET_STD = (0.229, 0.224, 0.225)
def build_transform(input_size):
MEAN, STD = IMAGENET_MEAN, IMAGENET_STD
transform = T.Compose([
T.Lambda(lambda img: img.convert('RGB') if img.mode != 'RGB' else img),
T.Resize((input_size, input_size), interpolation=InterpolationMode.BICUBIC),
T.ToTensor(),
T.Normalize(mean=MEAN, std=STD)
])
return transform
def find_closest_aspect_ratio(aspect_ratio, target_ratios, width, height, image_size):
best_ratio_diff = float('inf')
best_ratio = (1, 1)
area = width * height
for ratio in target_ratios:
target_aspect_ratio = ratio[0] / ratio[1]
ratio_diff = abs(aspect_ratio - target_aspect_ratio)
if ratio_diff < best_ratio_diff:
best_ratio_diff = ratio_diff
best_ratio = ratio
elif ratio_diff == best_ratio_diff:
if area > 0.5 * image_size * image_size * ratio[0] * ratio[1]:
best_ratio = ratio
return best_ratio
def dynamic_preprocess(image, min_num=1, max_num=12, image_size=448, use_thumbnail=False):
orig_width, orig_height = image.size
aspect_ratio = orig_width / orig_height
# calculate the existing image aspect ratio
target_ratios = set(
(i, j) for n in range(min_num, max_num + 1) for i in range(1, n + 1) for j in range(1, n + 1) if
i * j <= max_num and i * j >= min_num)
target_ratios = sorted(target_ratios, key=lambda x: x[0] * x[1])
# find the closest aspect ratio to the target
target_aspect_ratio = find_closest_aspect_ratio(
aspect_ratio, target_ratios, orig_width, orig_height, image_size)
# calculate the target width and height
target_width = image_size * target_aspect_ratio[0]
target_height = image_size * target_aspect_ratio[1]
blocks = target_aspect_ratio[0] * target_aspect_ratio[1]
# resize the image
resized_img = image.resize((target_width, target_height))
processed_images = []
for i in range(blocks):
box = (
(i % (target_width // image_size)) * image_size,
(i // (target_width // image_size)) * image_size,
((i % (target_width // image_size)) + 1) * image_size,
((i // (target_width // image_size)) + 1) * image_size
)
# split the image
split_img = resized_img.crop(box)
processed_images.append(split_img)
assert len(processed_images) == blocks
if use_thumbnail and len(processed_images) != 1:
thumbnail_img = image.resize((image_size, image_size))
processed_images.append(thumbnail_img)
return processed_images
def load_image(image_file, input_size=448, max_num=12):
image = Image.open(image_file).convert('RGB')
transform = build_transform(input_size=input_size)
images = dynamic_preprocess(image, image_size=input_size, use_thumbnail=True, max_num=max_num)
pixel_values = [transform(image) for image in images]
pixel_values = torch.stack(pixel_values)
return pixel_values
# If you have an 80G A100 GPU, you can put the entire model on a single GPU.
# Otherwise, you need to load a model using multiple GPUs, please refer to the `Multiple GPUs` section.
path = 'OpenGVLab/InternVL2_5-8B'
model = AutoModel.from_pretrained(
path,
torch_dtype=torch.bfloat16,
low_cpu_mem_usage=True,
trust_remote_code=True).eval().cuda()
tokenizer = AutoTokenizer.from_pretrained(path, trust_remote_code=True, use_fast=False)
# set the max number of tiles in `max_num`
pixel_values = load_image('./examples/image1.jpg', max_num=12).to(torch.bfloat16).cuda()
generation_config = dict(max_new_tokens=1024, do_sample=False)
# pure-text conversation (纯文本对话)
question = 'Hello, who are you?'
response, history = model.chat(tokenizer, None, question, generation_config, history=None, return_history=True)
print(f'User: {question}
Assistant: {response}')
question = 'Can you tell me a story?'
response, history = model.chat(tokenizer, None, question, generation_config, history=history, return_history=True)
print(f'User: {question}
Assistant: {response}')
# single-image single-round conversation (单图单轮对话)
question = '<image>
Please describe the image shortly.'
response = model.chat(tokenizer, pixel_values, question, generation_config)
print(f'User: {question}
Assistant: {response}')
# single-image multi-round conversation (单图多轮对话)
question = '<image>
Please describe the image in detail.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config, history=None, return_history=True)
print(f'User: {question}
Assistant: {response}')
question = 'Please write a poem according to the image.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config, history=history, return_history=True)
print(f'User: {question}
Assistant: {response}')
# multi-image multi-round conversation, combined images (多图多轮对话，拼接图像)
pixel_values1 = load_image('./examples/image1.jpg', max_num=12).to(torch.bfloat16).cuda()
pixel_values2 = load_image('./examples/image2.jpg', max_num=12).to(torch.bfloat16).cuda()
pixel_values = torch.cat((pixel_values1, pixel_values2), dim=0)
question = '<image>
Describe the two images in detail.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config,
history=None, return_history=True)
print(f'User: {question}
Assistant: {response}')
question = 'What are the similarities and differences between these two images.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config,
history=history, return_history=True)
print(f'User: {question}
Assistant: {response}')
# multi-image multi-round conversation, separate images (多图多轮对话，独立图像)
pixel_values1 = load_image('./examples/image1.jpg', max_num=12).to(torch.bfloat16).cuda()
pixel_values2 = load_image('./examples/image2.jpg', max_num=12).to(torch.bfloat16).cuda()
pixel_values = torch.cat((pixel_values1, pixel_values2), dim=0)
num_patches_list = [pixel_values1.size(0), pixel_values2.size(0)]
question = 'Image-1: <image>
Image-2: <image>
Describe the two images in detail.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config,
num_patches_list=num_patches_list,
history=None, return_history=True)
print(f'User: {question}
Assistant: {response}')
question = 'What are the similarities and differences between these two images.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config,
num_patches_list=num_patches_list, history=history, return_history=True)
print(f'User: {question}
Assistant: {response}')
# batch inference, single image per sample (单图批处理)
pixel_values1 = load_image('./examples/image1.jpg', max_num=12).to(torch.bfloat16).cuda()
pixel_values2 = load_image('./examples/image2.jpg', max_num=12).to(torch.bfloat16).cuda()
num_patches_list = [pixel_values1.size(0), pixel_values2.size(0)]
pixel_values = torch.cat((pixel_values1, pixel_values2), dim=0)
questions = ['<image>
Describe the image in detail.'] * len(num_patches_list)
responses = model.batch_chat(tokenizer, pixel_values,
num_patches_list=num_patches_list,
questions=questions,
generation_config=generation_config)
for question, response in zip(questions, responses):
print(f'User: {question}
Assistant: {response}')
# video multi-round conversation (视频多轮对话)
def get_index(bound, fps, max_frame, first_idx=0, num_segments=32):
if bound:
start, end = bound[0], bound[1]
else:
start, end = -100000, 100000
start_idx = max(first_idx, round(start * fps))
end_idx = min(round(end * fps), max_frame)
seg_size = float(end_idx - start_idx) / num_segments
frame_indices = np.array([
int(start_idx + (seg_size / 2) + np.round(seg_size * idx))
for idx in range(num_segments)
])
return frame_indices
def load_video(video_path, bound=None, input_size=448, max_num=1, num_segments=32):
vr = VideoReader(video_path, ctx=cpu(0), num_threads=1)
max_frame = len(vr) - 1
fps = float(vr.get_avg_fps())
pixel_values_list, num_patches_list = [], []
transform = build_transform(input_size=input_size)
frame_indices = get_index(bound, fps, max_frame, first_idx=0, num_segments=num_segments)
for frame_index in frame_indices:
img = Image.fromarray(vr[frame_index].asnumpy()).convert('RGB')
img = dynamic_preprocess(img, image_size=input_size, use_thumbnail=True, max_num=max_num)
pixel_values = [transform(tile) for tile in img]
pixel_values = torch.stack(pixel_values)
num_patches_list.append(pixel_values.shape[0])
pixel_values_list.append(pixel_values)
pixel_values = torch.cat(pixel_values_list)
return pixel_values, num_patches_list
video_path = './examples/red-panda.mp4'
pixel_values, num_patches_list = load_video(video_path, num_segments=8, max_num=1)
pixel_values = pixel_values.to(torch.bfloat16).cuda()
video_prefix = ''.join([f'Frame-{i+1}: <image>
' for i in range(len(num_patches_list))])
question = video_prefix + 'What is the red panda doing?'
# Frame1: <image>
Frame2: <image>
...
Frame8: <image>
{question}
response, history = model.chat(tokenizer, pixel_values, question, generation_config,
num_patches_list=num_patches_list, history=None, return_history=True)
print(f'User: {question}
Assistant: {response}')
question = 'Describe this video in detail.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config,
num_patches_list=num_patches_list, history=history, return_history=True)
print(f'User: {question}
Assistant: {response}')
```
</details>
## License
This project is released under the MIT License. This project uses the pre-trained Qwen2.5-72B-Instruct as a component, which is licensed under the Qwen License.
## Citation
If you find this project useful in your research, please consider citing:
```BibTeX
@article{wang2024mpo,
title={Enhancing the Reasoning Ability of Multimodal Large Language Models via Mixed Preference Optimization},
author={Wang, Weiyun and Chen, Zhe and Wang, Wenhai and Cao, Yue and Liu, Yangzhou and Gao, Zhangwei and Zhu, Jinguo and Zhu, Xizhou and Lu, Lewei and Qiao, Yu and Dai, Jifeng},
journal={arXiv preprint arXiv:2411.10442},
year={2024}
}
@article{chen2024expanding,
title={Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling},
author={Chen, Zhe and Wang, Weiyun and Cao, Yue and Liu, Yangzhou and Gao, Zhangwei and Cui, Erfei and Zhu, Jinguo and Ye, Shenglong and Tian, Hao and Liu, Zhaoyang and others},
journal={arXiv preprint arXiv:2412.05271},
year={2024}
}
@article{chen2024far,
title={How Far Are We to GPT-4V? Closing the Gap to Commercial Multimodal Models with Open-Source Suites},
author={Chen, Zhe and Wang, Weiyun and Tian, Hao and Ye, Shenglong and Gao, Zhangwei and Cui, Erfei and Tong, Wenwen and Hu, Kongzhi and Luo, Jiapeng and Ma, Zheng and others},
journal={Science China Information Sciences},
volume={67},
number={12},
pages={220101},
year={2024},
publisher={Springer}
}
@inproceedings{chen2024internvl,
title={Internvl: Scaling up vision foundation models and aligning for generic visual-linguistic tasks},
author={Chen, Zhe and Wu, Jiannan and Wang, Wenhai and Su, Weijie and Chen, Guo and Xing, Sen and Zhong, Muyan and Zhang, Qinglong and Zhu, Xizhou and Lu, Lewei and others},
booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
pages={24185--24198},
year={2024}
}
```
## What can InternVL do?
<details>
<summary>Visual Perception (click to expand)</summary>
- Linear-Probe Image Classification [[see details]](./classification#-evaluation)
ViT-22B uses the private JFT-3B dataset.
| method | #param | IN-1K | IN-ReaL | IN-V2 | IN-A | IN-R | IN-Sketch |
| ------------------- | :----: | :---: | :-----: | :---: | :---: | :---: | :-------: |
| OpenCLIP-G | 1.8B | 86.2 | 89.4 | 77.2 | 63.8 | 87.8 | 66.4 |
| DINOv2-g | 1.1B | 86.5 | 89.6 | 78.4 | 75.9 | 78.8 | 62.5 |
| EVA-01-CLIP-g | 1.1B | 86.5 | 89.3 | 77.4 | 70.5 | 87.7 | 63.1 |
| MAWS-ViT-6.5B | 6.5B | 87.8 | - | - | - | - | - |
| ViT-22B\* | 21.7B | 89.5 | 90.9 | 83.2 | 83.8 | 87.4 | - |
| InternViT-6B (ours) | 5.9B | 88.2 | 90.4 | 79.9 | 77.5 | 89.8 | 69.1 |
- Semantic Segmentation [[see details]](./segmentation#-evaluation)
| method | decoder | #param (train/total) | crop size | mIoU |
| --------------------- | :-----: | :------------------: | :-------: | ------------ |
| OpenCLIP-G (frozen) | Linear | 0.3M / 1.8B | 512 | 39.3 |
| ViT-22B (frozen) | Linear | 0.9M / 21.7B | 504 | 34.6 |
| InternViT-6B (frozen) | Linear | 0.5M / 5.9B | 504 | 47.2 (+12.6) |
| ViT-22B (frozen) | UperNet | 0.8B / 22.5B | 504 | 52.7 |
| InternViT-6B (frozen) | UperNet | 0.4B / 6.3B | 504 | 54.9 (+2.2) |
| ViT-22B | UperNet | 22.5B / 22.5B | 504 | 55.3 |
| InternViT-6B | UperNet | 6.3B / 6.3B | 504 | 58.9 (+3.6) |
- Zero-Shot Image Classification [[see details]](./clip_benchmark#imagenet-variants-and-objectnet)
| method | IN-1K | IN-A | IN-R | IN-V2 | IN-Sketch | ObjectNet |
| ----------------- | :---: | :---: | :---: | :---: | :-------: | :-------: |
| OpenCLIP-G | 80.1 | 69.3 | 92.1 | 73.6 | 68.9 | 73.0 |
| EVA-02-CLIP-E+ | 82.0 | 82.1 | 94.5 | 75.7 | 71.6 | 79.6 |
| ViT-22B\* | 85.9 | 90.1 | 96.0 | 80.9 | - | 87.6 |
| InternVL-C (ours) | 83.2 | 83.8 | 95.5 | 77.3 | 73.9 | 80.6 |
- Multilingual Zero-Shot Image Classification [[see details]](./clip_benchmark#multilingual-imagenet-1k)
EN: English, ZH: Chinese, JP: Japanese, Ar: Arabic, IT: Italian
| method | IN-1K (EN) | IN-1K (ZH) | IN-1K (JP) | IN-1K (AR) | IN-1K (IT) |
| ----------------- | :--------: | :--------: | :--------: | :--------: | :--------: |
| Taiyi-CLIP-ViT-H | - | 54.4 | - | - | - |
| WuKong-ViT-L-G | - | 57.5 | - | - | - |
| CN-CLIP-ViT-H | - | 59.6 | - | - | - |
| AltCLIP-ViT-L | 74.5 | 59.6 | - | - | - |
| EVA-02-CLIP-E+ | 82.0 | - | - | - | 41.2 |
| OpenCLIP-XLM-R-H | 77.0 | 55.7 | 53.1 | 37.0 | 56.8 |
| InternVL-C (ours) | 83.2 | 64.5 | 61.5 | 44.9 | 65.7 |
- Zero-Shot Video Classification
| method | #frame | K400 | K600 | K700 |
| ----------------- | :----: | :---: | :---: | :---: |
| OpenCLIP-G | 1 | 65.9 | 66.1 | 59.2 |
| EVA-02-CLIP-E+ | 1 | 69.8 | 69.3 | 63.4 |
| InternVL-C (ours) | 1 | 71.0 | 71.3 | 65.7 |
| ViCLIP | 8 | 75.7 | 73.5 | 66.4 |
| InternVL-C (ours) | 8 | 79.4 | 78.8 | 71.5 |
</details>
<details>
<summary>Cross-Modal Retrieval (click to expand)</summary>
- English Zero-Shot Image-Text Retrieval [[see details]](./clip_benchmark#flickr30k--coco)
<table>
<tr align=center>
<td rowspan="3" align=left><b>model</b></td>
<td colspan="6" align=center><b>Flickr30K</b></td>
<td colspan="6" align=center><b>COCO</b></td>
<td rowspan="3" align=center><b>avg</b></td>
</tr>
<tr align=center>
<td colspan="3" align=center><b>image-to-text</b></td>
<td colspan="3" align=center><b>text-to-image</b></td>
<td colspan="3" align=center><b>image-to-text</b></td>
<td colspan="3" align=center><b>text-to-image</b></td>
</tr>
<tr>
<td>R@1</td>
<td>R@5</td>
<td>R@10</td>
<td>R@1</td>
<td>R@5</td>
<td>R@10</td>
<td>R@1</td>
<td>R@5</td>
<td>R@10</td>
<td>R@1</td>
<td>R@5</td>
<td>R@10</td>
</tr>
<tr align=center>
<td align=left>OpenCLIP-G</td>
<td>92.9</td>
<td>99.3</td>
<td>99.8</td>
<td>79.5</td>
<td>95.0</td>
<td>97.1</td>
<td>67.3</td>
<td>86.9</td>
<td>92.6</td>
<td>51.4</td>
<td>74.9</td>
<td>83.0</td>
<td>85.0</td>
</tr>
<tr align=center>
<td align=left>EVA-02-CLIP-E+</td>
<td>93.9</td>
<td>99.4</td>
<td>99.8</td>
<td>78.8</td>
<td>94.2</td>
<td>96.8</td>
<td>68.8</td>
<td>87.8</td>
<td>92.8</td>
<td>51.1</td>
<td>75.0</td>
<td>82.7</td>
<td>85.1</td>
</tr>
<tr align=center>
<td align=left>EVA-CLIP-8B</td>
<td>95.6</td>
<td>99.6</td>
<td>99.9</td>
<td>80.8</td>
<td>95.5</td>
<td>97.6</td>
<td>70.3</td>
<td>89.3</td>
<td>93.9</td>
<td>53.0</td>
<td>76.0</td>
<td>83.4</td>
<td>86.2</td>
</tr>
<tr align=center>
<td align=left>InternVL-C (ours)</td>
<td>94.7</td>
<td>99.6</td>
<td>99.9</td>
<td>81.7</td>
<td>96.0</td>
<td>98.2</td>
<td>70.6</td>
<td>89.0</td>
<td>93.5</td>
<td>54.1</td>
<td>77.3</td>
<td>84.6</td>
<td>86.6</td>
</tr>
<tr align=center>
<td align=left>InternVL-G (ours)</td>
<td>95.7</td>
<td>99.7</td>
<td>99.9</td>
<td>85.0</td>
<td>97.0</td>
<td>98.6</td>
<td>74.9</td>
<td>91.3</td>
<td>95.2</td>
<td>58.6</td>
<td>81.3</td>
<td>88.0</td>
<td>88.8</td>
</tr>
</table>
- Chinese Zero-Shot Image-Text Retrieval [[see details]](./clip_benchmark#flickr30k-cn--coco-cn)
<table>
<tr align=center>
<td rowspan="3" align=left><b>model</b></td>
<td colspan="6" align=center><b>Flickr30K-CN</b></td>
<td colspan="6" align=center><b>COCO-CN</b></td>
<td rowspan="3" align=center><b>avg</b></td>
</tr>
<tr align=center>
<td colspan="3" align=center><b>image-to-text</b></td>
<td colspan="3" align=center><b>text-to-image</b></td>
<td colspan="3" align=center><b>image-to-text</b></td>
<td colspan="3" align=center><b>text-to-image</b></td>
</tr>
<tr>
<td>R@1</td>
<td>R@5</td>
<td>R@10</td>
<td>R@1</td>
<td>R@5</td>
<td>R@10</td>
<td>R@1</td>
<td>R@5</td>
<td>R@10</td>
<td>R@1</td>
<td>R@5</td>
<td>R@10</td>
</tr>
<tr align=center>
<td align=left>CN-CLIP-ViT-H</td>
<td>81.6</td>
<td>97.5</td>
<td>98.8</td>
<td>71.2</td>
<td>91.4</td>
<td>95.5</td>
<td>63.0</td>
<td>86.6</td>
<td>92.9</td>
<td>69.2</td>
<td>89.9</td>
<td>96.1</td>
<td>86.1</td>
</tr>
<tr align=center>
<td align=left>OpenCLIP-XLM-R-H</td>
<td>86.1</td>
<td>97.5</td>
<td>99.2</td>
<td>71.0</td>
<td>90.5</td>
<td>94.9</td>
<td>70.0</td>
<td>91.5</td>
<td>97.0</td>
<td>66.1</td>
<td>90.8</td>
<td>96.0</td>
<td>87.6</td>
</tr>
<tr align=center>
<td align=left>InternVL-C (ours)</td>
<td>90.3</td>
<td>98.8</td>
<td>99.7</td>
<td>75.1</td>
<td>92.9</td>
<td>96.4</td>
<td>68.8</td>
<td>92.0</td>
<td>96.7</td>
<td>68.9</td>
<td>91.9</td>
<td>96.5</td>
<td>89.0</td>
</tr>
<tr align=center>
<td align=left>InternVL-G (ours)</td>
<td>92.9</td>
<td>99.4</td>
<td>99.8</td>
<td>77.7</td>
<td>94.8</td>
<td>97.3</td>
<td>71.4</td>
<td>93.9</td>
<td>97.7</td>
<td>73.8</td>
<td>94.4</td>
<td>98.1</td>
<td>90.9</td>
</tr>
</table>
- Multilingual Zero-Shot Image-Text Retrieval on XTD [[see details]](./clip_benchmark#xtd)
| method | EN | ES | FR | ZH | IT | KO | RU | JP | average |
| ----------------- | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :-----: |
| AltCLIP | 95.4 | 94.1 | 92.9 | 95.1 | 94.2 | 94.4 | 91.8 | 91.7 | 93.7 |
| OpenCLIP-XLM-R-H | 97.3 | 96.1 | 94.5 | 94.7 | 96.0 | 90.2 | 93.9 | 94.0 | 94.6 |
| InternVL-C (ours) | 97.3 | 95.7 | 95.1 | 95.6 | 96.0 | 92.2 | 93.3 | 95.5 | 95.1 |
| InternVL-G (ours) | 98.6 | 97.7 | 96.5 | 96.7 | 96.9 | 95.1 | 94.8 | 96.1 | 96.6 |
</details>
<details>
<summary>Multimodal Dialogue</summary>
</details> |