Skip to content

Commit

Permalink
[feat] remove registeration logic and adding language evaluation task…
Browse files Browse the repository at this point in the history 
…s. (#218)

* Load tasks only one time (#139)

* chore: Initialize tasks only once to avoid re-initialization

* chore: Initialize tasks only once to avoid re-initialization

* chore: Refactor task initialization to avoid re-initialization

* chore: Update task initialization to fix include_path issue

* chore: Update task initialization to fix include_path issue

* Upload live_bench results (#140)

* upload results

* add a readme

* chore: Update upload_results.py script to use shell syntax

* Update upload_results.py

* Update upload_results.py

* Add Muirbench (#143)

* handle gen kwargs in internvl2

* Add muirbench

* if no response directly return 0 (#142)

* merge ov evals (#144)

* chore: Update gpt_eval_model_name to "gpt-3.5-turbo" in mathvista.yaml

* Squashed commit of the following:

commit 994c9f97a2f8db3e9b7d7933d1e1680acde5b70b
Author: Yan Shu <570533048@qq.com>
Date:   Mon Jul 8 17:21:23 2024 +0800

    Add files via upload

* Squashed commit of the following:

commit e31cd7883d4555c7530795c7f102b8d78cbd372f
Author: Bo Li <drluodian@gmail.com>
Date:   Wed Jul 10 12:08:08 2024 +1000

    chore: Update lmms_eval/models/vila.py and lmms_eval/tasks/__init__.py

commit 1d8c980d1089f9d7702c3b92d5c85039f2809c6d
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Tue Jul 9 02:08:52 2024 +0000

    Rename xcomposer 4KHD

commit 6da76f36ecf5f9aa73057e767a4fcb60c99ff896
Author: Bo Li <drluodian@gmail.com>
Date:   Tue Jul 9 11:55:56 2024 +1000

    Upgrade lmms-eval to version 0.2.1

commit cd1858523fcd8630082cbefba8710e0de3ee8805
Author: Bo Li <drluodian@gmail.com>
Date:   Tue Jul 9 11:52:23 2024 +1000

    Upgrade lmms-eval to support more models and evaluation tasks

commit 672d7e5bb49dcb34e1b2fdeb09f3f4588dc583a6
Author: Bo Li <drluodian@gmail.com>
Date:   Tue Jul 9 11:43:41 2024 +1000

    feat: Add tie_weights parameter to Llava model initialization

commit 2037a86261b55fa42b8ba3a04eab192b3e69d6ea
Merge: e6844db1 a5c18692
Author: Bo Li <drluodian@gmail.com>
Date:   Tue Jul 9 11:37:12 2024 +1000

    Fix gen kwargs image aspect ratio in internvl2

commit a5c186925de989b616f58a35ece36065a32b4594
Merge: 2ebec77f 557083a1
Author: Li Bo <drluodian@gmail.com>
Date:   Tue Jul 9 09:15:56 2024 +0800

    Merge pull request #137 from shuyansy/main

    add MLVU task

commit 557083a156c3dd67ac79e22b4202e9b69b6b00f4
Author: Yan Shu <570533048@qq.com>
Date:   Mon Jul 8 16:56:50 2024 +0800

    Add files via upload

commit 2ebec77f5606d79e9a7b995970e32792050606a1
Merge: 211bfede b23d349e
Author: Li Bo <drluodian@gmail.com>
Date:   Mon Jul 8 11:53:06 2024 +0800

    Merge pull request #136 from Dousia/main

    Add detailcaps

commit b23d349e46d60dc149ffaa54d6e019f4996ed92d
Author: ByteDance <bytedance@MacBook-Pro.local>
Date:   Sun Jul 7 23:24:19 2024 +0800

    Add install capture_metric in env

commit c6e211d5f9dbb7572d3a141b6504cb1ca2007c33
Author: ByteDance <bytedance@MacBook-Pro.local>
Date:   Sun Jul 7 23:04:13 2024 +0800

    Add detailcaps

commit 211bfedebad243ef82a8b0be36c3b5a9b9cb2f72
Merge: 7c208b76 79514eee
Author: Li Bo <drluodian@gmail.com>
Date:   Tue Jul 2 23:05:12 2024 +0800

    Merge pull request #133 from EvolvingLMMs-Lab/dev/wild_vision

    Add wild vision bench

commit 79514eeebcfd6f655be2a10c776037d12a7b7214
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Mon Jul 1 15:10:02 2024 +0000

    Fixing handling None filtered score

commit 725fac2781446958b905e1e6c6eb3c0a8e582e49
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Mon Jul 1 08:25:42 2024 +0000

    Fixing dataset name

commit 8d963e132ac03fc0d835d480cfcfcabe72af143c
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Mon Jul 1 08:24:51 2024 +0000

    Fixing scoring logic

commit e2990d0a69e876721256fdf946c68ba7ae0cbdc1
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Mon Jul 1 06:06:57 2024 +0000

    Hardcode to keep image for wild vision

commit ed381736730d8fb785b4ee919fdb751734ecef25
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Mon Jul 1 06:06:38 2024 +0000

    Add wild vision 0617

commit 7c208b76640c986cfe94233dce735c3ca4ad4319
Author: Li Bo <drluodian@gmail.com>
Date:   Mon Jul 1 11:53:31 2024 +0800

    Update README.md

commit 39d40dea47bc59ff04e8b0cbc445345098debc9a
Merge: e19b43a3 ba7081c0
Author: Li Bo <drluodian@gmail.com>
Date:   Mon Jul 1 11:47:09 2024 +0800

    Merge pull request #129 from Dannoopsy/mmbench_ru

    add task MMBench-ru

commit e19b43a3a1e7212e623061b164b0419cc0dda689
Merge: 11fd7e3f a0de8970
Author: Li Bo <drluodian@gmail.com>
Date:   Mon Jul 1 11:46:58 2024 +0800

    Merge pull request #128 from Dannoopsy/gqa-ru

    add task gqa-ru

commit 11fd7e3fc05908aeb01e4a6161a7b55cd38b3122
Merge: 383e7fea a7522592
Author: Li Bo <drluodian@gmail.com>
Date:   Mon Jul 1 11:46:16 2024 +0800

    Merge pull request #130 from lscpku/vitatecs

    Add task VITATECS

commit a75225926e5954f85466d257f99acf0163fde596
Author: lscpku <lisc99@pku.edu.cn>
Date:   Fri Jun 28 20:37:06 2024 +0800

    create new task vitatecs

commit ba7081c0abac840002d320e30733e891298dfa11
Author: Dannoopsy <63581325+Dannoopsy@users.noreply.github.com>
Date:   Fri Jun 28 12:21:05 2024 +0300

    change prompt to ru

commit 27ea9c0055a8abf3a8198829b8617018479918e2
Author: Dannoopsy <belopolskikh.dd@phystech.edu>
Date:   Thu Jun 27 17:17:29 2024 +0000

    add mmbench_ru_dev

commit 383e7fead3138aedf62e9c0ec48303835ef26e2a
Merge: 06fa000f ed2e7f79
Author: Li Bo <drluodian@gmail.com>
Date:   Fri Jun 28 00:14:10 2024 +0800

    Merge pull request #126 from lorenzomammana/feature/external-package-integration

    External package integration using plugins

commit ed2e7f792151d21bce8f1c498270b9391e1d5c85
Merge: 03947e14 06fa000f
Author: Lorenzo Mammana <mammanalorenzo@outlook.it>
Date:   Thu Jun 27 15:38:10 2024 +0000

    Merge branch 'main' into feature/external-package-integration

commit a0de89708d5e6f259bb17f0eaace3c5b901b275c
Author: Dannoopsy <belopolskikh.dd@phystech.edu>
Date:   Tue Jun 25 11:11:37 2024 +0000

    new task gqa-ru

commit 06fa000f60d3e4d160fac8ceb9959ae92a98f752
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Tue Jun 25 06:41:13 2024 +0000

    Fix vid mme post prompt issue

commit b388d79e0df6f60068196cb7047453ebd22d6ef1
Author: Li Bo <drluodian@gmail.com>
Date:   Sun Jun 23 22:31:16 2024 +0800

    Update activitynetqa_generation.yaml

commit 8f9d620fcd9d0a0742ee6bcf51ea63bd6b088a36
Author: Li Bo <drluodian@gmail.com>
Date:   Sun Jun 23 14:02:25 2024 +0800

    Update pyproject.toml

commit 6341b7c15ce9fb28eb06b067ddb299d6cf2e16c3
Merge: fce85f1b 903b042b
Author: Li Bo <drluodian@gmail.com>
Date:   Sun Jun 23 14:02:02 2024 +0800

    Merge pull request #125 from EvolvingLMMs-Lab/dev/interleave

    [Model] aligned llava-interleave model results on video tasks

commit 903b042be016016d4ebeecb07701f3076a2d323c
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Sat Jun 22 12:07:13 2024 +0000

    Remove unnecessary lines for video llava

commit d78ec86407b729a964906a8c2e50704b4bc74d06
Merge: ebe7217a fce85f1b
Author: Li Bo <drluodian@gmail.com>
Date:   Sat Jun 22 13:57:31 2024 +0800

    Merge branch 'main' into dev/interleave

commit ebe7217a486c1e754e42c2cbdb834e09fbbcc9b0
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Sat Jun 22 02:57:08 2024 +0000

    Delete unnecessary lines

commit 120c474b056f9177c74e1fd9691d59e2f234b785
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Fri Jun 21 08:38:41 2024 +0000

    Revise model registry for llava_hf and longva

commit 7d6201f921088afd3f52a35076e3c6fcc9aa518c
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Fri Jun 21 08:38:24 2024 +0000

    Add longva

commit 12f480699c71a12a24d4349d9b0681933201a3a6
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Fri Jun 21 08:35:39 2024 +0000

    Remove unnecessary lines since use batched visuals now in llava

commit 12cea76f1f0f14b1fd1007c9d39a9b0557368637
Author: Bo Li <drluodian@gmail.com>
Date:   Thu Jun 20 18:15:32 2024 +0000

    chore: Add loguru for logging in lmms_eval package

commit 03947e14a46fd25b412931f7c9c25f4a2971d0b4
Author: Lorenzo Mammana <mammanalorenzo@outlook.it>
Date:   Wed Jun 5 13:40:41 2024 +0000

    feat: Allow including external tasks from plugins

commit b80a91f73e15ddd0b0ce1322d7d121fa14030eed
Author: Lorenzo Mammana <mammanalorenzo@outlook.it>
Date:   Wed Jun 5 13:04:55 2024 +0000

    feat: Allow loading model configurations from other packages

commit 8ef24740dd48a11c97eb627f2fff4aca107fef0d
Author: Bo Li <drluodian@gmail.com>
Date:   Thu Jun 20 12:11:03 2024 +0000

    chore: Remove unused models from lmms_eval package

commit af38885fc2e066f5ea44388f33e07176f836fe28
Author: Bo Li <drluodian@gmail.com>
Date:   Thu Jun 20 12:07:09 2024 +0000

    chore: Handle ImportError when importing models

    Handle the ImportError exception when importing models in the lmms_eval package. This change adds a try-except block to catch the ImportError and print an error message indicating the failed import. This will help with troubleshooting and identifying any issues with the model imports.

commit fce85f1b03ff7043b29dee787c5d17a08dd2687a
Merge: dbe63293 d94f83cb
Author: Li Bo <drluodian@gmail.com>
Date:   Thu Jun 20 20:02:12 2024 +0800

    Merge pull request #120 from EvolvingLMMs-Lab/pufanyi/hf_dataset_docs

    Add docs for datasets upload to HF

commit dbe63293245a5141fdfd80bda7657c304f6bd32f
Author: choiszt <ls2001927@sohu.com>
Date:   Thu Jun 20 15:14:21 2024 +0800

    update ablation for videomme datasets

commit d94f83cb3f08b61a2c75cc4326e58792100605b3
Author: Li Bo <drluodian@gmail.com>
Date:   Thu Jun 20 13:30:59 2024 +0800

    Update README.md

commit cab8159ff35db330536c0b6dfb4b0a3b24142209
Author: Li Bo <drluodian@gmail.com>
Date:   Thu Jun 20 13:30:29 2024 +0800

    Update README.md

commit 45876652a877a8006b828f32f5cc4660629f9190
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Thu Jun 20 03:55:30 2024 +0000

    Add llava_hf back to registry

commit 3463651b8c54d36cd94169e3d376f5ed225a195a
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Thu Jun 20 03:54:33 2024 +0000

    Remove handling non-visual loop in llava

commit cb0d3f49b72790b081f981e0e6147131542f7f68
Author: Fanyi Pu <FPU001@e.ntu.edu.sg>
Date:   Thu Jun 20 02:11:18 2024 +0800

    update readme

commit 813877bfe5ac590cdbe92dd74d18f83a2091f748
Author: Fanyi Pu <FPU001@e.ntu.edu.sg>
Date:   Wed Jun 19 15:37:52 2024 +0800

    to sh script

commit a14684b8557d5894976448a5c559ed7a66a6cf16
Author: Fanyi Pu <FPU001@e.ntu.edu.sg>
Date:   Wed Jun 19 15:37:04 2024 +0800

    lint

commit d0f8851d42ba31f5da2a7a65e91499db45174dbc
Author: Fanyi Pu <FPU001@e.ntu.edu.sg>
Date:   Wed Jun 19 15:36:48 2024 +0800

    small fix

commit 63748e9718f287ad433afc90e340b5e17a89c1ed
Author: Fanyi Pu <FPU001@e.ntu.edu.sg>
Date:   Wed Jun 19 15:36:43 2024 +0800

    small fix

commit 7f1159a1fe04cfb783dc31d4fbdef3bda0ce19e4
Author: Fanyi Pu <FPU001@e.ntu.edu.sg>
Date:   Wed Jun 19 15:35:05 2024 +0800

    update preparation

commit 19f9bd621c76a483ff98f8c7eb78f64753da683a
Author: Fanyi Pu <FPU001@e.ntu.edu.sg>
Date:   Wed Jun 19 15:23:24 2024 +0800

    docs

commit ce6f889ba02d819979c7922f6336cf4f1f718f65
Author: Fanyi Pu <FPU001@e.ntu.edu.sg>
Date:   Wed Jun 19 15:04:16 2024 +0800

    tutorial

commit f513c520c2a3dad26d2b2ca5c4ed4db05a493c73
Author: Bo Li <drluodian@gmail.com>
Date:   Wed Jun 19 06:51:19 2024 +0000

    chore: Update dependencies to fix potential risks and improve compatibility

commit efb529552c5e4ba039a4cba8e9aa5cb7ba65bf90
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Wed Jun 19 10:25:58 2024 +0800

    Release llava-wilder

commit 742651fc9daf97e2f57831ed6e6e7ee7ead7d555
Author: Fanyi Pu <FPU001@e.ntu.edu.sg>
Date:   Wed Jun 19 07:44:26 2024 +0800

    feat: Add support for auto downloading tar format videos

commit 511b6259828212fcba954cdeb8cf90d6e5daabf8
Merge: 22a4958e 050b2c37
Author: Bo Li <drluodian@gmail.com>
Date:   Tue Jun 18 17:01:03 2024 +0000

    Merge branch 'main' of https://github.com/EvolvingLMMs-Lab/lmms-eval

commit 050b2c370017e9b97475dd6cf01fd051b5ca5c86
Merge: 74facb41 ef306512
Author: Li Bo <drluodian@gmail.com>
Date:   Tue Jun 18 13:13:38 2024 +0800

    Merge pull request #114 from zjysteven/add-tinyllava

    add tinyllava

commit ef306512e5135f76dffa383f600b8733015836e8
Author: Jingyang Zhang <jingyang.zhang@duke.edu>
Date:   Mon Jun 17 17:57:02 2024 -0400

    fix typo

commit 9bab67732a4238097725deddf867fb1946ffee40
Merge: dbfb2387 74facb41
Author: Jingyang Zhang <jingyang.zhang@duke.edu>
Date:   Sun Jun 16 10:56:05 2024 -0400

    Merge branch 'EvolvingLMMs-Lab:main' into add-tinyllava

commit 74facb41a826691dfce4458cf1d8659b34fc5bf5
Merge: 8ba192f9 d5df72de
Author: Li Bo <drluodian@gmail.com>
Date:   Sun Jun 16 17:59:19 2024 +0800

    Merge pull request #118 from teowu/main

    Fix the potential risk by PR #117

commit d5df72de2d03108d6b365818ecc3551ac9aa6302
Merge: 5bf59ed2 8ba192f9
Author: Teo (Timothy) Wu Haoning <38696372+teowu@users.noreply.github.com>
Date:   Sun Jun 16 15:32:13 2024 +0800

    Merge branch 'EvolvingLMMs-Lab:main' into main

commit 5bf59ed250da98a408a94e214a73caa400cba842
Author: teowu <realtimothyhwu@gmail.com>
Date:   Sun Jun 16 07:27:28 2024 +0000

    fix #117, allow auto download with tar format videos

commit 98b3955cb808e36303c030aea78eb037d1ec59ce
Merge: a056f118 be9dada8
Author: teowu <realtimothyhwu@gmail.com>
Date:   Sun Jun 16 07:25:07 2024 +0000

    Merge branch 'main' of https://github.com/teowu/lmms-eval into main

commit a056f118704eccec86ce32ab86981ce4bc1e1deb
Author: teowu <realtimothyhwu@gmail.com>
Date:   Sun Jun 16 07:23:54 2024 +0000

    fix #117, allow auto download with tar format videos

commit 8ba192f94edf5d99598983445d5faa4f8807c49f
Merge: 7cc28907 be9dada8
Author: Li Bo <drluodian@gmail.com>
Date:   Sat Jun 15 17:30:59 2024 +0800

    Merge pull request #117 from teowu/main

    LongVideoBench for LMMs-Eval

commit be9dada8b4189c53c08e1674ab273242cf2f80a0
Merge: 62ea8ceb 7cc28907
Author: Teo (Timothy) Wu Haoning <38696372+teowu@users.noreply.github.com>
Date:   Sat Jun 15 16:39:20 2024 +0800

    Merge pull request #1 from EvolvingLMMs-Lab/main

    Merge pull request #113 from teowu/main

commit 62ea8ceb223ef2b51ebab2bcd50d5cf339c35cfe
Author: teowu <realtimothyhwu@gmail.com>
Date:   Sat Jun 15 08:30:11 2024 +0000

    LongVideoBench support: image LMMs (idefics2, phi3) and video LMMs (LLaVA-Next-Video-34B)

commit 7cc28907edbb4eb58ee1398772a48110ea35dd96
Merge: 4bc7224d ea14cd4b
Author: Li Bo <drluodian@gmail.com>
Date:   Sat Jun 15 14:10:22 2024 +0800

    Merge pull request #113 from teowu/main

    Q-Bench, Q-Bench2, A-Bench

commit dbfb23873979f789477f4797ee2d6071e0fd921e
Author: Jingyang <jingyang.zhang@duke.edu>
Date:   Fri Jun 14 16:20:42 2024 -0400

    add tinyllava

commit ea14cd4b361f4c95b3665cbdb95bc51754090eb5
Author: teowu <realtimothyhwu@gmail.com>
Date:   Fri Jun 14 15:01:52 2024 +0000

    Add qbench, qbench2, abench; fix phi3v as its current implementation does not support multi-image

commit 4bc7224dcd27fe8b288bfc3fed4d7a9da9635658
Merge: 2797987f bf14cb85
Author: Li Bo <drluodian@gmail.com>
Date:   Fri Jun 14 02:14:43 2024 +0800

    Merge pull request #111 from XinrunDu/main

    add II-Bench

commit bf14cb8527b2b7ac438a36567a875168bc02d294
Author: XinrunDu <duxinrun2000@gmail.com>
Date:   Thu Jun 13 09:37:02 2024 +0000

    fix dataset_path

commit 6248113f4e11a0ac396d31fa1b032a142fea8cb4
Author: XinrunDu <duxinrun2000@gmail.com>
Date:   Thu Jun 13 09:32:06 2024 +0000

    add II-Bench

commit 2797987f5b88b87bd172714b678a75a1d8051826
Merge: 63d82f1f 66d4bb2d
Author: Li Bo <drluodian@gmail.com>
Date:   Thu Jun 13 11:14:47 2024 +0800

    Merge pull request #109 from EvolvingLMMs-Lab/pufanyi/update_version

    [Small Update] Update the version of LMMs-Eval

commit 66d4bb2d9c9afbbdea40196d4ad80e214d0b14b6
Author: Fanyi Pu <FPU001@e.ntu.edu.sg>
Date:   Thu Jun 13 11:13:00 2024 +0800

    update version

commit 63d82f1ff11eb430d91a15d6788a1f0b4d596850
Author: Li Bo <drluodian@gmail.com>
Date:   Thu Jun 13 11:04:32 2024 +0800

    Update README.md

commit 44a33799671cb668f55366d5e5a4ddb051a3a1b4
Merge: 5ed00356 0ce46d08
Author: Li Bo <drluodian@gmail.com>
Date:   Thu Jun 13 04:00:12 2024 +0800

    Merge pull request #105 from tianyu-z/main

    Include VCR

commit 0ce46d088e473d12d63de44f17c67dceab25658c
Author: Suyuchen <suyuchen.wang@umontreal.ca>
Date:   Wed Jun 12 15:56:34 2024 -0400

    update README.md

commit 46a88d8b0199ed44d2ff459fb372f2e006960cea
Merge: 47b13b9b 5ed00356
Author: Suyuchen <suyuchen.wang@umontreal.ca>
Date:   Wed Jun 12 15:50:26 2024 -0400

    merged readme.md

commit 47b13b9b320d36ac53b3622557e31239f7c22621
Author: Suyuchen <suyuchen.wang@umontreal.ca>
Date:   Wed Jun 12 15:30:52 2024 -0400

    update aggregation function for vcr_wiki

commit 5ed00356676cf5d0ff056cf27d1b519b8e303ff7
Author: Li Bo <drluodian@gmail.com>
Date:   Thu Jun 13 03:21:42 2024 +0800

    Update README.md

commit ed8806839db5988ced672bd162b7b046edb4863a
Author: Li Bo <drluodian@gmail.com>
Date:   Thu Jun 13 03:13:59 2024 +0800

    Update README.md

commit fea3806026932a6e2bd6e538bcc413e33abdf245
Merge: d99a24ab 05dc8e85
Author: Li Bo <drluodian@gmail.com>
Date:   Thu Jun 13 03:11:49 2024 +0800

    Merge pull request #108 from EvolvingLMMs-Lab/internal_main_dev

    [Upgrade to v0.2] Embracing Video Evaluations with LMMs-Eval

commit 05dc8e853eab7c6bc782a1e2662d2efe7422f767
Author: Bo Li <drluodian@gmail.com>
Date:   Wed Jun 12 15:56:04 2024 +0000

    chore: Update lmms-eval to support video evaluations for LLaVA models

commit cbeee20bc4ffb510a2b23d96cdaf4077be7c2a9e
Author: Bo Li <drluodian@gmail.com>
Date:   Wed Jun 12 15:50:30 2024 +0000

    chore: Update lmms-eval to support video evaluations for LLaVA models

commit f00d5498b69dd4f7e54c907ac906abc7c128f000
Author: Bo Li <drluodian@gmail.com>
Date:   Wed Jun 12 15:46:33 2024 +0000

    Update image alignment in README.md

commit 34156335db74cef9e3f0915d7172fd6b22456c15
Author: Bo Li <drluodian@gmail.com>
Date:   Wed Jun 12 15:43:16 2024 +0000

    Update llava conv_template in lmms_eval/models/llava.py

commit 50575a950736bc8fc1e191310314cbb5fdff5720
Author: Bo Li <drluodian@gmail.com>
Date:   Wed Jun 12 15:39:03 2024 +0000

    chore: Update lmms-eval to support video evaluations for LLaVA models

commit c9b2252fb8a15dd04252af5e6b4613855afd6ada
Author: Bo Li <drluodian@gmail.com>
Date:   Wed Jun 12 15:33:48 2024 +0000

    Bump version to 0.2.0.dev0

commit 465bd4205e8097e9c037b24a3ed08dd6a7694efa
Merge: e43bd840 d99a24ab
Author: Bo Li <drluodian@gmail.com>
Date:   Wed Jun 12 15:04:25 2024 +0000

    Merge branch 'main' of https://github.com/EvolvingLMMs-Lab/lmms-eval into internal_main_dev

commit e43bd840b63eb499856e36d9d2ba45c924abcead
Author: Bo Li <drluodian@gmail.com>
Date:   Wed Jun 12 14:54:06 2024 +0000

    chore: Remove unnecessary files and code related to live_bench and sft_eval tasks

commit d99a24abd06df10d07e5a4d0ad5030613f92f2e7
Merge: 374590be a66003be
Author: Li Bo <drluodian@gmail.com>
Date:   Wed Jun 12 19:45:57 2024 +0800

    Merge pull request #107 from AtsuMiyai/new_task/upd_update

    update gpt-3.5-turbo version

commit a66003befe4175824a1be6ed59f5f5b88c15f792
Author: AtsuMiyai <miyai.atsuyuki.practice@gmail.com>
Date:   Wed Jun 12 17:05:17 2024 +0900

    update gpt-3.5-turbo version

commit ee91f272985f32eeb9cd6faa41afdd8eb49cac30
Author: AtsuMiyai <miyai.atsuyuki.practice@gmail.com>
Date:   Wed Jun 12 16:50:53 2024 +0900

    update gpt-3.5-turbo version

commit 326b9694fc77398592b8caf3ba0bc2e2bb903813
Author: tianyu-z <zhangtianyupro@gmail.com>
Date:   Mon Jun 10 20:07:40 2024 -0400

    include std and confidence interval

commit cd050d4a721d01a2ace0cd030cf7f8dc67eb8c4d
Author: Suyuchen <suyuchen.wang@umontreal.ca>
Date:   Mon Jun 10 18:49:47 2024 -0400

    update vcr_wiki tasks in README.md

commit 205721e0aad76dde30255e56149bbed121883356
Author: Suyuchen <suyuchen.wang@umontreal.ca>
Date:   Mon Jun 10 18:43:15 2024 -0400

    update vcr_wiki tasks

commit db8e718b502469e8536ee359c5559de87635ffc7
Author: tianyu-z <zhangtianyupro@gmail.com>
Date:   Mon Jun 10 16:13:58 2024 -0400

    include the try-except logic for spacy

commit 427dabb790118f538b64e4e5bf6a7aab9689b3d9
Author: Suyuchen <suyuchen.wang@umontreal.ca>
Date:   Mon Jun 10 15:51:05 2024 -0400

    add crossed_text to vcr_wiki output

commit 043b483eb55f7be4fea75c9bc0b9b03d251b109b
Author: tianyu-z <zhangtianyupro@gmail.com>
Date:   Mon Jun 10 15:47:00 2024 -0400

    switch logic

commit e1f04db8f58dd10591fde335ea13f74cda7c79bd
Author: tianyu-z <zhangtianyupro@gmail.com>
Date:   Mon Jun 10 02:38:21 2024 -0400

    modify the form of VCR

commit 96e8d9867c9549ab7490f4b12cfeb6a06238e0aa
Author: tianyu-z <zhangtianyupro@gmail.com>
Date:   Mon Jun 10 00:10:30 2024 -0400

    init include vcr

commit 374590be62f988a76cf6704cfe394cd8ae7d4cb6
Merge: 504685e2 cb3b9ce7
Author: Kaichen Zhang - NTU <kaichenzhang358@outlook.com>
Date:   Fri Jun 7 20:25:48 2024 +0800

    Merge pull request #101 from Gumpest/main

    Update conbench in README

commit 504685e20b17659b913cf46f3012c16bf429e09d
Author: Li Bo <drluodian@gmail.com>
Date:   Thu Jun 6 15:42:15 2024 +0800

    Update README.md

commit cb3b9ce71411da862ff01342a9122a3c656ffbd1
Merge: c9793b38 67b64ea4
Author: Yuan Zhang <56063339+Gumpest@users.noreply.github.com>
Date:   Thu Jun 6 11:22:24 2024 +0800

    Merge branch 'EvolvingLMMs-Lab:main' into main

commit c9793b3883714f254a700230b7bee781d6110e73
Author: Yuan Zhang <gump_well_done@163.com>
Date:   Thu Jun 6 11:21:05 2024 +0800

    update README

commit 67b64ea44a5a39d96c7a196a8a8345a7486bd912
Merge: 8ee7848a 5fd68451
Author: Li Bo <drluodian@gmail.com>
Date:   Wed Jun 5 23:12:58 2024 +0800

    Merge pull request #100 from Gumpest/main

    add Conbench

commit 5fd684515c55ef643726c1b6c720c7cbd2183ba1
Author: Yuan Zhang <gump_well_done@163.com>
Date:   Wed Jun 5 21:52:31 2024 +0800

    add conbench

commit 8ee7848aaa6383aa1f919c3f21199c81db3fff89
Merge: 747e1978 6fefaf7c
Author: Li Bo <drluodian@gmail.com>
Date:   Tue Jun 4 17:09:33 2024 +0800

    Merge pull request #95 from AtsuMiyai/new_task/upd

    add MM-UPD

commit 747e19782996065cdce7157ee8c5e15beb5b6c59
Merge: 4854a34d 05843072
Author: Li Bo <drluodian@gmail.com>
Date:   Tue Jun 4 17:09:04 2024 +0800

    Merge pull request #97 from CaraJ7/update

    Add MathVerse in README.md

commit 6fefaf7cea504e35583ee7217449da290295a7a4
Author: AtsuMiyai <miyai.atsuyuki.practice@gmail.com>
Date:   Tue Jun 4 17:36:39 2024 +0900

    update utils.py for leaderboard submission

commit 5f4fe360def1c48ea0cb1da6409d192784882308
Author: AtsuMiyai <miyai.atsuyuki.practice@gmail.com>
Date:   Sun Jun 2 23:28:27 2024 +0900

    slightly change query_prompt for the reproduction

commit 05843072d608b970bcada1cd0db65a3c80864060
Author: CaraJ7 <1350074492@qq.com>
Date:   Sun Jun 2 17:05:28 2024 +0800

    Add MathVerse in README.md

commit 0581ab3cfb362e2024988b46fbbb00324f1233c9
Author: AtsuMiyai <miyai.atsuyuki.practice@gmail.com>
Date:   Fri May 31 16:09:45 2024 +0900

    merge model_specific_prompt_kwargs and dataset_name into each task yaml

commit 4854a34d4d37efb5e201f2691ecdb054590cf20b
Author: Pu Fanyi <FPU001@e.ntu.edu.sg>
Date:   Sat May 4 19:23:39 2024 +0800

    Group MMMU images into one image (#83)

    * update

    * update font

    * Add matplotlib.font_manager import in utils.py

    * Refactor font handling in add_order_label function in utils.py

    * group mmmu

    ---------

    Co-authored-by: Li Bo <drluodian@gmail.com>

commit d224794c49520f4d28a31862cf977198cd6cbc5e
Author: AtsuMiyai <miyai.atsuyuki.practice@gmail.com>
Date:   Wed May 29 15:15:59 2024 +0900

    add upd

commit 453e7936424220f02b99517059ca71babfbe5f5a
Author: AtsuMiyai <miyai.atsuyuki.practice@gmail.com>
Date:   Wed May 29 15:03:30 2024 +0900

    add upd

commit 909edd6769ddcf8a546be4fdd129416687516878
Author: AtsuMiyai <miyai.atsuyuki.practice@gmail.com>
Date:   Wed May 29 12:52:21 2024 +0900

    add upd

commit 7c1ac9706cafc4801fa4da181d2f610b7838c7b8
Author: AtsuMiyai <miyai.atsuyuki.practice@gmail.com>
Date:   Wed May 29 12:50:32 2024 +0900

    add upd

commit 811301c5280ddd74986645086f026ab730c8848c
Author: AtsuMiyai <miyai.atsuyuki.practice@gmail.com>
Date:   Wed May 29 12:46:58 2024 +0900

    add upd

commit 71401bafd1d515f704f86ab4817a758542bc4672
Author: AtsuMiyai <miyai.atsuyuki.practice@gmail.com>
Date:   Wed May 29 12:41:21 2024 +0900

    add upd

commit 24dc435908d921e9f1a5706e3141b12e5d838d18
Author: Bo Li <drluodian@gmail.com>
Date:   Mon May 27 10:17:32 2024 +0000

    fix compatibility issue of older version llava

commit 616edf43731415b35f0f5e97748ed2e017a2891d
Author: Bo Li <drluodian@gmail.com>
Date:   Mon May 27 09:32:26 2024 +0000

    [Fix] import issues of multilingual llava and olympiadbench

commit 4c5a99e21a63fb0ee1c7d15546d18066e1d9894b
Merge: 45c05b2b b05c3e22
Author: Li Bo <drluodian@gmail.com>
Date:   Mon May 27 14:19:53 2024 +0800

    Merge pull request #87 from vfragoso/vifragos/phi3v

    Adding microsoft/Phi-3-vision-128k-instruct model.

commit b05c3e222fabd308dd7af4e04c1c6a0812962fe6
Author: Victor Fragoso <victor.fragoso@microsoft.com>
Date:   Fri May 24 16:36:37 2024 +0000

    Adding documentation of Phi3v class.

commit c2008971308ce8168d57c24d00b725832f099244
Author: Victor Fragoso <victor.fragoso@microsoft.com>
Date:   Fri May 24 16:25:02 2024 +0000

    Adding prompt arguments for Phi3v on MathVista-TestMini

commit 7f9fb6bcc6cd24a7b8011b8753d0ea98cc2451fd
Author: Victor Fragoso <victor.fragoso@microsoft.com>
Date:   Fri May 24 13:24:16 2024 +0000

    Adding Phi3v model.

commit 45c05b2b2bece76e06849a52a0d034f9c0ac2367
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Thu May 23 03:47:36 2024 +0000

    Set printing info for llava_hf to debug level

commit 53f013ed8278776551ca992562253387cc9968d2
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Thu May 23 03:41:39 2024 +0000

    Fix pope random name in pope full

commit 22520a95f13334b75eee0cf0387151067a6bf516
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Thu May 23 03:41:14 2024 +0000

    Add separated pope tasks by category

commit d1eefb1565014b47287ffa6b350229062f8f602f
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Thu May 9 08:36:02 2024 +0000

    Update gitignore

commit b2b4dbd2dc13432c79208db35abf7f55c97f1790
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Mon May 20 07:45:11 2024 +0000

    Comment out Spice in caption task so that don't need to download stanford nlp model

commit 662f05ce4c62a46a83f819d3a5925a9bd20059b5
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Mon May 20 03:13:13 2024 +0000

    Comment out parse result in xcomposer

commit 09329322916bfbb604d72ddaf50441a0947f8805
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Thu May 16 03:55:39 2024 +0000

    Fix instructblip qformer size mismatch and multi-images problem

commit 557a6a3b15e07e506bc05e2cc76ff6a2f8c93964
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Thu May 16 03:11:41 2024 +0000

    Remove redundant code in fuyu

commit 6aeb5504e74ed1980b53700d8e4d4dcf7d1b38fc
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Thu May 16 01:45:24 2024 +0000

    Fix idefics2 llava in the wild bugs

commit aea80e6a71f716951353e1e5d68380243396b4d6
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Wed May 15 11:07:35 2024 +0000

    Better task list_with_num

commit 3c12a080d66b9c38f615b961befca7c30f82fa39
Author: Li Bo <drluodian@gmail.com>
Date:   Sat May 18 02:35:52 2024 +0800

    Update LICENSE

commit 82317a635a4978b32e095a06cc295d0ae23661c2
Author: Li Bo <drluodian@gmail.com>
Date:   Sat May 18 02:29:09 2024 +0800

    Update LICENSE

commit a8bba1cdb51061a0d27bf9a98cca1505b5c58ea5
Author: Li Bo <drluodian@gmail.com>
Date:   Sat May 18 02:28:03 2024 +0800

    Create LICENSE

commit caa5893b5fd2c1d32c72b97f371ccd9a8d9ec3a0
Merge: c0944486 423b0060
Author: Li Bo <drluodian@gmail.com>
Date:   Mon May 13 11:45:26 2024 +0800

    Merge pull request #73 from EvolvingLMMs-Lab/kc/qwen_vl_api

    [Feat] Add qwen vl api

commit c09444860362a136f17641f8b2a1f91c2bbc3715
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Sat May 11 06:11:19 2024 +0000

    Fix llava_hf image tokens number issue

commit 64f07e497f53e5bcbe9e8fb5830cc7a1daaf7ff1
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Thu May 9 02:04:10 2024 +0000

    Fix endless warning for llava_hf generation

commit 8aaa828108da8514dd9cd23a9d6d83a8b67f2d65
Author: Bo Li <drluodian@gmail.com>
Date:   Thu May 2 06:13:56 2024 +0000

    Add model_name parameter to Llava constructor

commit 7847dc4d8efe60605102414bb071b1da9851228e
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Tue May 7 03:15:59 2024 +0000

    Parse result for llava_hf 1.6

commit 3e56b4f92db39a2ce92903b0c43a34f1d14d59ec
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Tue May 7 03:09:56 2024 +0000

    Fix llava_hf generation for 1.6

commit fa3ff92b07ea5aaa633a2039818c310744f84d07
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Mon May 6 08:32:57 2024 +0000

    Fix llava conv template for llama3

commit 423b00606aa77fd6b324c19e3d480b73ab852db6
Author: kcz358 <kaichenzhang358@outlook.com>
Date:   Sun May 5 07:54:52 2024 +0000

    Add qwen vl api

commit b7fd7a9f7aa3c0e1e50374047dfffc46a7462b90
Merge: 986139a9 c5a130b6
Author: Li Bo <drluodian@gmail.com>
Date:   Sun May 5 13:19:48 2024 +0800

    Merge pull request #59 from EvolvingLMMs-Lab/add_idefics2

    add idefics2

commit 986139a9a31154679bdea029b09639f84712db27
Merge: b46239ca 8d3526c0
Author: Li Bo <drluodian@gmail.com>
Date:   Fri May 3 01:18:18 2024 +0800

    Merge pull request #36 from cocoshe/main

    [Fix] repr llava doc

commit b46239cabab7b545ec99d9eae6c851e531b18374
Merge: bc69a744 373265f2
Author: Li Bo <drluodian@gmail.com>
Date:   Fri May 3 01:17:34 2024 +0800

    Merge pull request #56 from gagan3012/main

    Multilingual LLava bench

commit bc69a744d2cffeb06eba62e843bcc7869e27613a
Merge: eef3aeb6 626e8a91
Author: Li Bo <drluodian@gmail.com>
Date:   Fri May 3 01:12:14 2024 +0800

    Merge pull request #70 from hunterheiden/hsh/new_task/WebSRC

    Bugfix: WebSRC should be token-level F1 NOT character-level

commit 626e8a91a4af2dd5dd774fc130cc2f4d74b2bc37
Author: Hunter Heidenreich <hunter.heidenreich@rootsautomation.com>
Date:   Thu May 2 09:31:03 2024 -0400

    Bugfix: WebSRC should be token-level F1 NOT character-level

commit eef3aeb6ab589bb1d5045af5b5c1984a69402d19
Merge: c4e9dd9f 9bca4413
Author: Li Bo <drluodian@gmail.com>
Date:   Thu May 2 14:38:17 2024 +0800

    Merge pull request #69 from hunterheiden/hsh/new_task/WebSRC

    [New Task] WebSRC (multimodal Q&A on web screenshots)

commit 9bca441376325173128e5c50087f068e519c48da
Author: Hunter Heidenreich <hunter.heidenreich@rootsautomation.com>
Date:   Wed May 1 11:07:29 2024 -0400

    Add code to enable compilation of submission for WebSRC test split

commit 7687495b1ed552eeba088cb9ad5aaf1170e7fff9
Author: Hunter Heidenreich <hunter.heidenreich@rootsautomation.com>
Date:   Wed May 1 10:47:32 2024 -0400

    Draft and validate websrc eval on dev split

commit 4eebd3e5d7ab3b8c3116eea57318db72d2ce32bb
Author: Hunter Heidenreich <hunter.heidenreich@rootsautomation.com>
Date:   Wed May 1 10:46:54 2024 -0400

    Update main README with new task names

commit 35fe80b67656114a8824eb59574089663bdc4c9a
Author: Hunter Heidenreich <hunter.heidenreich@rootsautomation.com>
Date:   Wed May 1 10:46:20 2024 -0400

    Draft README for WebSRC

commit 955bd0635cc6c14a96ad869f1002e6dbefdc5071
Author: Hunter Heidenreich <hunter.heidenreich@rootsautomation.com>
Date:   Tue Apr 30 10:16:21 2024 -0400

    Init webSRC

commit c4e9dd9f6e40e8586587c4a75987aa109a37f14b
Merge: d8a3a99f 319afccb
Author: Li Bo <drluodian@gmail.com>
Date:   Fri Apr 26 14:37:22 2024 +0800

    Merge pull request #63 from hunterheiden/hsh/new_task/screenspot

    New Task: ScreenSpot - Grounding (REC) and instruction generation (REG) on screens

commit 319afccbe713ddf40a8a6fa28501e64c0ad34725
Author: Hunter Heidenreich <hunter.heidenreich@rootsautomation.com>
Date:   Thu Apr 25 11:44:34 2024 -0400

    slight update

commit 2f3811ca1bbad6a441016b05fde09a571900fca8
Author: Hunter Heidenreich <hunter.heidenreich@rootsautomation.com>
Date:   Thu Apr 25 11:41:04 2024 -0400

    Add README file specific to ScreenSpot

commit 28962cbe83631ec5d6481aaea4907a7c96fec848
Author: Hunter Heidenreich <hunter.heidenreich@rootsautomation.com>
Date:   Wed Apr 24 11:52:33 2024 -0400

    Update README to reflect new tasks

commit e457cfb4f2d6869e8367d6d5b03ad25ee4acc363
Author: Hunter Heidenreich <hunter.heidenreich@rootsautomation.com>
Date:   Tue Apr 23 18:33:16 2024 -0400

    Create ScreenSpot on clean branch

commit d8a3a99ff6142fe101fa3c188cc7f29593c44345
Merge: 3dcd0158 ed171293
Author: Li Bo <drluodian@gmail.com>
Date:   Tue Apr 23 10:34:03 2024 +0800

    Merge pull request #61 from tupini07/patch-1

    Fix typo in Qwen-VL that was causing "reference before assignment"

commit ed171293d1e82075c5c6a847fc91ecbfd45cf89f
Author: Andrea Tupini <tupini07@gmail.com>
Date:   Mon Apr 22 14:56:41 2024 -0600

    refactor query construction for clarity

commit cd874201c46f32a2903ddffae85f9db73e14adfd
Author: Andrea Tupini <tupini07@gmail.com>
Date:   Mon Apr 22 14:54:29 2024 -0600

    convert contexts to list if necessary and remove unnecessary construction of `questions`

commit 85573674e90c8d505312ba18c5102e0051255078
Author: Andrea Tupini <tupini07@gmail.com>
Date:   Mon Apr 22 14:47:33 2024 -0600

    Fix typo in qwen_vl that was causing "reference before assignment"

commit 3dcd01582b719555bcf8eb25d91cc5e42abd2c5f
Merge: 95df9fee 743673a1
Author: Li Bo <drluodian@gmail.com>
Date:   Sat Apr 20 22:03:16 2024 +0800

    Merge pull request #60 from CaraJ7/main

    Add MathVerse

commit 743673a1419b6e729e18c96f148745cc739d4c71
Merge: c1a54721 95df9fee
Author: CaraJ7 <1350074492@qq.com>
Date:   Sat Apr 20 21:49:02 2024 +0800

    Merge branch 'main' of https://github.com/EvolvingLMMs-Lab/lmms-eval

commit c1a5472135c3b84061b64d997ab50dda0412ba4f
Author: CaraJ7 <1350074492@qq.com>
Date:   Sat Apr 20 21:45:34 2024 +0800

    Add MathVerse

commit 373265f24e7a89cbd49ab724a2e388cc0930be78
Author: Gagan Bhatia <49101362+gagan3012@users.noreply.github.com>
Date:   Fri Apr 12 17:21:39 2024 -0700

    Add files via upload

commit d8530514a5ef9378d2adeaceb228b60ec25a6718
Author: Gagan Bhatia <49101362+gagan3012@users.noreply.github.com>
Date:   Fri Apr 12 17:19:49 2024 -0700

    Create README.md

commit 22a4958e993463edff352ac033014f9a485706cc
Author: Bo Li <bo.li01@bytedance.com>
Date:   Thu Apr 4 17:12:43 2024 +0000

    [WIP] adding mmbench dev evaluation (#75)

    * WIP

    * Update GPT evaluation model name and sys prompt

    * 🛠️ Scale accuracy to percentage

    The accuracy value is now multiplied by 100 in the aggregation function to represent it as a percentage. Regarding the evaluation process, `math` module importation and refactoring reduce progress log verbosity by logging every 100 evaluations instead of 10. It prevents potential logging overflow. Handling of NaN values is added to ensure 'default_value' is set in case of missing data, avoiding errors in split, category, and l2-category assignments. Finally, reporting of categorical and l2-categorical accuracies is streamlined through a new `calculate_hit_rates` function, improving code readability and maintenance.

    Issue refs: #1427, #1533

    * Update GPT evaluation model name and API configuration

    * Refactor MMBench_Evaluator class to handle missing columns

    * Add print statements for detailed results in MMBench-CN(CC), MMBench-CN(Dev), and MMBench-EN(Dev) evaluations

    * Refactor MMBench-CN and MMBench-EN evaluation functions

    * 🔄 Refactor result processing and logging logic

    - Simplified the result processing functions across different utility modules (`cc_utils.py`, `cn_utils.py`, `en_utils.py`) to unify the handling of multiple-choice options. Now, all options ("A" to "E") are dynamically added to the result data, and default to "nan" if not provided in the document.
    - Removed redundant keys directly from the process results dict creation to avoid clutter and align with the new dynamic addition of options.
    - In `mmbench_evals.py`, removed the unnecessary check for all splits being 'dev' and streamlined the evaluation loop by eliminating the progress bar (tqdm) for a cleaner log output.
    - Commented-out code and verbose logging during evaluation, which may have interfered with performance, has been removed for a more efficient and less intrusive logging experience.

    This cleanup reduces redundancy in the codebase and improves evaluation performance.

    Refs #2045

    ---------

    Co-authored-by: Bo Li <bo.li01@bytedance.com>
    (cherry picked from commit a19278c2ea6ddcbca64d3cc7f4efec7fe5775121)

commit 8d3526c0869f0ad7747ff6bb02441140792b461c
Author: cocoshe <1228759711@qq.com>
Date:   Thu Mar 28 13:38:36 2024 +0800

    fix doc

* feat: Add LlavaOneVision model to available models

chore: Update sqlitedict dependency to version 2.1.0

* Revert "Squashed commit of the following:"

This reverts commit 11b00999df3c43cb225482e030b791b2d454124c.

* Refactor available models in lmms_eval

Remove duplicate entries for "llava_hf", "llava_onevision", and "longva" in the AVAILABLE_MODELS dictionary in lmms_eval/models/__init__.py.

* fix: Handle import errors in lmms_eval models/__init__.py

The code changes in this commit fix the handling of import errors in the lmms_eval/models/__init__.py file. Previously, when an import error occurred, the code simply ignored it. This commit updates the code to log an error message using the logger module when an import error occurs.

This commit also removes duplicate entries for "llava_hf", "llava_onevision", and "longva" in the AVAILABLE_MODELS dictionary.

Recent user commits:
- Refactor available models in lmms_eval
- Revert "Squashed commit of the following:"
- feat: Add LlavaOneVision model to available models
- chore: Update sqlitedict dependency to version 2.1.0

* fix: Handle import errors in lmms_eval models/__init__.py

* chore: Remove unused imports in lmms_eval/models/__init__.py and lmms_eval/tasks/vcr_wiki/utils.py

* Remove unused imports in lmms_eval/tasks/vcr_wiki/utils.py

* chore: Update lmms_eval/tasks/vcr_wiki/utils.py

This commit updates the `lmms_eval/tasks/vcr_wiki/utils.py` file. It removes unused imports and fixes the condition for loading Spacy models based on the `load_package` value in the config file. Additionally, it adds a debug log message when the Spacy models are not loaded due to `load_package` being set to False.

Remove unused imports in `lmms_eval/tasks/vcr_wiki/utils.py`

* feat: Add new subtasks to overall score calculation

The code changes in this commit add new subtasks to the overall score calculation in the `overall_score` function. The subtasks "ScanQA", "BLINK", "MathVerse", "SciVerse", and "Mantis" are included in the `categories` dictionary. This ensures that the scores for these subtasks are calculated and included in the evaluation results.

Remove unused imports and update subtask categories in `utils.py`

* feat: Add new subtasks to overall score calculation

* chore: Update lmms_eval/tasks/llava_interleave_bench/_default_template_interleave_yaml

Update the image aspect ratio in the default template for the llava_interleave_bench task. Change the value of "image_aspect_ratio" from "original" to "pad". This ensures that the generated images have a padded aspect ratio.

* if no response directly return 0

* Squashed commit of the following:

commit b2a009b6bbf8353172f5a1dd9c29ea1f67610c02
Author: Pu Fanyi <FPU001@e.ntu.edu.sg>
Date:   Mon Jul 15 19:12:25 2024 -0700

    if no response directly return 0 (#142)

commit 5fc5f2f5acf454fc99448b0d62eb52b4bffba0d5
Author: Kaichen Zhang - NTU <kaichenzhang358@outlook.com>
Date:   Tue Jul 16 10:12:11 2024 +0800

    Add Muirbench (#143)

    * handle gen kwargs in internvl2

    * Add muirbench

* Add files via upload

(cherry picked from commit 557083a156c3dd67ac79e22b4202e9b69b6b00f4)

* update

---------

Co-authored-by: Fanyi Pu <FPU001@e.ntu.edu.sg>
Co-authored-by: Yan Shu <570533048@qq.com>

* Fix llava onevision loglikelihood video bug

* LiveBench July (#146)

* claude auto detect json mode

* extract information

* use claude to generate

* fix bugs

* fix

* generate data

* chore: Update dataset name and version for live_bench task

* gpt-4-turbo => gpt-4o

* chore: Update dataset capture settings in create_dataset.py

* everything use gpt-4o

* websites

* livebench_july

* Refactor code to simplify data assignment in example.ipynb

* chore: Update dataset name for live_bench task

* Add xcomposer2d5 from fanyi, revise something for better usage (#145)

* internvl2

* fix some bugs

* fix

* lint

* feat: Add XComposer2D5 model to AVAILABLE_MODELS

* xcomposer

* Fix llava vid error when using public

* Fix xcomposer2d5

* Add generation tokens

---------

Co-authored-by: Fanyi Pu <FPU001@e.ntu.edu.sg>

* Dev/ov evals (#147)

* fix doc

* [WIP] adding mmbench dev evaluation (#75)

* WIP

* Update GPT evaluation model name and sys prompt

* 🛠️ Scale accuracy to percentage

The accuracy value is now multiplied by 100 in the aggregation function to represent it as a percentage. Regarding the evaluation process, `math` module importation and refactoring reduce progress log verbosity by logging every 100 evaluations instead of 10. It prevents potential logging overflow. Handling of NaN values is added to ensure 'default_value' is set in case of missing data, avoiding errors in split, category, and l2-category assignments. Finally, reporting of categorical and l2-categorical accuracies is streamlined through a new `calculate_hit_rates` function, improving code readability and maintenance.

Issue refs: #1427, #1533

* Update GPT evaluation model name and API configuration

* Refactor MMBench_Evaluator class to handle missing columns

* Add print statements for detailed results in MMBench-CN(CC), MMBench-CN(Dev), and MMBench-EN(Dev) evaluations

* Refactor MMBench-CN and MMBench-EN evaluation functions

* 🔄 Refactor result processing and logging logic

- Simplified the result processing functions across different utility modules (`cc_utils.py`, `cn_utils.py`, `en_utils.py`) to unify the handling of multiple-choice options. Now, all options ("A" to "E") are dynamically added to the result data, and default to "nan" if not provided in the document.
- Removed redundant keys directly from the process results dict creation to avoid clutter and align with the new dynamic addition of options.
- In `mmbench_evals.py`, removed the unnecessary check for all splits being 'dev' and streamlined the evaluation loop by eliminating the progress bar (tqdm) for a cleaner log output.
- Commented-out code and verbose logging during evaluation, which may have interfered with performance, has been removed for a more efficient and less intrusive logging experience.

This cleanup reduces redundancy in the codebase and improves evaluation performance.

Refs #2045

---------

Co-authored-by: Bo Li <bo.li01@bytedance.com>
(cherry picked from commit a19278c2ea6ddcbca64d3cc7f4efec7fe5775121)

* Create README.md

* Add files via upload

* Add MathVerse

* Fix typo in qwen_vl that was causing "reference before assignment"

* convert contexts to list if necessary and remove unnecessary construction of `questions`

* refactor query construction for clarity

* Create ScreenSpot on clean branch

* Update README to reflect new tasks

* Add README file specific to ScreenSpot

* slight update

* Init webSRC

* Draft README for WebSRC

* Update main README with new task names

* Draft and validate websrc eval on dev split

* Add code to enable compilation of submission for WebSRC test split

* Bugfix: WebSRC should be token-level F1 NOT character-level

* Add qwen vl api

* Fix llava conv template for llama3

* Fix llava_hf generation for 1.6

* Parse result for llava_hf 1.6

* Add model_name parameter to Llava constructor

* Fix endless warning for llava_hf generation

* Fix llava_hf image tokens number issue

* Create LICENSE

* Update LICENSE

* Update LICENSE

* Better task list_with_num

* Fix idefics2 llava in the wild bugs

* Remove redundant code in fuyu

* Fix instructblip qformer size mismatch and multi-images problem

* Comment out parse result in xcomposer

* Comment out Spice in caption task so that don't need to download stanford nlp model

* Update gitignore

* Add separated pope tasks by category

* Fix pope random name in pope full

* Set printing info for llava_hf to debug level

* Adding Phi3v model.

* Adding prompt arguments for Phi3v on MathVista-TestMini

* Adding documentation of Phi3v class.

* [Fix] import issues of multilingual llava and olympiadbench

* fix compatibility issue of older version llava

* add upd

* add upd

* add upd

* add upd

* add upd

* add upd

* Group MMMU images into one image (#83)

* update

* update font

* Add matplotlib.font_manager import in utils.py

* Refactor font handling in add_order_label function in utils.py

* group mmmu

---------

Co-authored-by: Li Bo <drluodian@gmail.com>

* merge model_specific_prompt_kwargs and dataset_name into each task yaml

* Add MathVerse in README.md

* slightly change query_prompt for the reproduction

* update utils.py for leaderboard submission

* add conbench

* update README

* Update README.md

* init include vcr

* modify the form of VCR

* switch logic

* add crossed_text to vcr_wiki output

* include the try-except logic for spacy

* update vcr_wiki tasks

* update vcr_wiki tasks in README.md

* include std and confidence interval

* update gpt-3.5-turbo version

* update gpt-3.5-turbo version

* chore: Remove unnecessary files and code related to live_bench and sft_eval tasks

* Bump version to 0.2.0.dev0

* chore: Update lmms-eval to support video evaluations for LLaVA models

* Update llava conv_template in lmms_eval/models/llava.py

* Update image alignment in README.md

* chore: Update lmms-eval to support video evaluations for LLaVA models

* chore: Update lmms-eval to support video evaluations for LLaVA models

* Update README.md

* Update README.md

* update aggregation function for vcr_wiki

* update README.md

* Update README.md

* update version

* add II-Bench

* fix dataset_path

* Add qbench, qbench2, abench; fix phi3v as its current implementation does not support multi-image

* add tinyllava

* LongVideoBench support: image LMMs (idefics2, phi3) and video LMMs (LLaVA-Next-Video-34B)

* fix #117, allow auto download with tar format videos

* fix #117, allow auto download with tar format videos

* fix typo

* feat: Add support for auto downloading tar format videos

* Release llava-wilder

* chore: Update dependencies to fix potential risks and improve compatibility

* tutorial

* docs

* update preparation

* small fix

* small fix

* lint

* to sh script

* update readme

* Remove handling non-visual loop in llava

* Add llava_hf back to registry

* Update README.md

* Update README.md

* update ablation for videomme datasets

* chore: Handle ImportError when importing models

Handle the ImportError exception when importing models in the lmms_eval package. This change adds a try-except block to catch the ImportError and print an error message indicating the failed import. This will help with troubleshooting and identifying any issues with the model imports.

* chore: Remove unused models from lmms_eval package

* feat: Allow loading model configurations from other packages

* feat: Allow including external tasks from plugins

* chore: Add loguru for logging in lmms_eval package

* Remove unnecessary lines since use batched visuals now in llava

* Add longva

* Revise model registry for llava_hf and longva

* Delete unnecessary lines

* Remove unnecessary lines for video llava

* Update pyproject.toml

* Update activitynetqa_generation.yaml

* Fix vid mme post prompt issue

* new task gqa-ru

* add mmbench_ru_dev

* change prompt to ru

* create new task vitatecs

* Update README.md

* Add wild vision 0617

* Hardcode to keep image for wild vision

* Fixing scoring logic

* Fixing dataset name

* Fixing handling None filtered score

* Add detailcaps

* Add install capture_metric in env

* Add files via upload

* feat: Add tie_weights parameter to Llava model initialization

* Upgrade lmms-eval to support more models and evaluation tasks

* Upgrade lmms-eval to version 0.2.1

* Rename xcomposer 4KHD

* chore: Update lmms_eval/models/vila.py and lmms_eval/tasks/__init__.py

* Update utils.py

* Update _default_template_vcr_yaml

* add process sync via temp file in lmms_eval/evaluator.py

* Update utils.py

* Update _default_template_vcr_yaml

* Add muirbench

* Squashed commit of the following:

commit dfdba507b5fbe985b0030ffec575f9f2638bc1ed
Author: Li Bo <drluodian@gmail.com>
Date:   Tue Jul 16 11:13:52 2024 +0800

    merge ov evals (#144)

    * chore: Update gpt_eval_model_name to "gpt-3.5-turbo" in mathvista.yaml

    * Squashed commit of the following:

    commit 994c9f97a2f8db3e9b7d7933d1e1680acde5b70b
    Author: Yan Shu <570533048@qq.com>
    Date:   Mon Jul 8 17:21:23 2024 +0800

        Add files via upload

    * Squashed commit of the following:

    commit e31cd7883d4555c7530795c7f102b8d78cbd372f
    Author: Bo Li <drluodian@gmail.com>
    Date:   Wed Jul 10 12:08:08 2024 +1000

        chore: Update lmms_eval/models/vila.py and lmms_eval/tasks/__init__.py

    commit 1d8c980d1089f9d7702c3b92d5c85039f2809c6d
    Author: kcz358 <kaichenzhang358@outlook.com>
    Date:   Tue Jul 9 02:08:52 2024 +0000

        Rename xcomposer 4KHD

    commit 6da76f36ecf5f9aa73057e767a4fcb60c99ff896
    Author: Bo Li <drluodian@gmail.com>
    Date:   Tue Jul 9 11:55:56 2024 +1000

        Upgrade lmms-eval to version 0.2.1

    commit cd1858523fcd8630082cbefba8710e0de3ee8805
    Author: Bo Li <drluodian@gmail.com>
    Date:   Tue Jul 9 11:52:23 2024 +1000

        Upgrade lmms-eval to support more models and evaluation tasks

    commit 672d7e5bb49dcb34e1b2fdeb09f3f4588dc583a6
    Author: Bo Li <drluodian@gmail.com>
    Date:   Tue Jul 9 11:43:41 2024 +1000

        feat: Add tie_weights parameter to Llava model initialization

    commit 2037a86261b55fa42b8ba3a04eab192b3e69d6ea
    Merge: e6844db1 a5c18692
    Author: Bo Li <drluodian@gmail.com>
    Date:   Tue Jul 9 11:37:12 2024 +1000

        Fix gen kwargs image aspect ratio in internvl2

    commit a5c186925de989b616f58a35ece36065a32b4594
    Merge: 2ebec77f 557083a1
    Author: Li Bo <drluodian@gmail.com>
    Date:   Tue Jul 9 09:15:56 2024 +0800

        Merge pull request #137 from shuyansy/main

        add MLVU task

    commit 557083a156c3dd67ac79e22b4202e9b69b6b00f4
    Author: Yan Shu <570533048@qq.com>
    Date:   Mon Jul 8 16:56:50 2024 +0800

        Add files via upload

    commit 2ebec77f5606d79e9a7b995970e32792050606a1
    Merge: 211bfede b23d349e
    Author: Li Bo <drluodian@gmail.com>
    Date:   Mon Jul 8 11:53:06 2024 +0800

        Merge pull request #136 from Dousia/main

        Add detailcaps

    commit b23d349e46d60dc149ffaa54d6e019f4996ed92d
    Author: ByteDance <bytedance@MacBook-Pro.local>
    Date:   Sun Jul 7 23:24:19 2024 +0800

        Add install capture_metric in env

    commit c6e211d5f9dbb7572d3a141b6504cb1ca2007c33
    Author: ByteDance <bytedance@MacBook-Pro.local>
    Date:   Sun Jul 7 23:04:13 2024 +0800

        Add detailcaps

    commit 211bfedebad243ef82a8b0be36c3b5a9b9cb2f72
    Merge: 7c208b76 79514eee
    Author: Li Bo <drluodian@gmail.com>
    Date:   Tue Jul 2 23:05:12 2024 +0800

        Merge pull request #133 from EvolvingLMMs-Lab/dev/wild_vision

        Add wild vision bench

    commit 79514eeebcfd6f655be2a10c776037d12a7b7214
    Author: kcz358 <kaichenzhang358@outlook.com>
    Date:   Mon Jul 1 15:10:02 2024 +0000

        Fixing handling None filtered score

    commit 725fac2781446958b905e1e6c6eb3c0a8e582e49
    Author: kcz358 <kaichenzhang358@outlook.com>
    Date:   Mon Jul 1 08:25:42 2024 +0000

        Fixing dataset name

    commit 8d963e132ac03fc0d835d480cfcfcabe72af143c
    Author: kcz358 <kaichenzhang358@outlook.com>
    Date:   Mon Jul 1 08:24:51 2024 +0000

        Fixing scoring logic

    commit e2990d0a69e876721256fdf946c68ba7ae0cbdc1
    Author: kcz358 <kaichenzhang358@outlook.com>
    Date:   Mon Jul 1 06:06:57 2024 +0000

        Hardcode to keep image for wild vision

    commit ed381736730d8fb785b4ee919fdb751734ecef25
    Author: kcz358 <kaichenzhang358@outlook.com>
    Date:   Mon Jul 1 06:06:38 2024 +0000

        Add wild vision 0617

    commit 7c208b76640c986cfe94233dce735c3ca4ad4319
    Author: Li Bo <drluodian@gmail.com>
    Date:   Mon Jul 1 11:53:31 2024 +0800

        Update README.md

    commit 39d40dea47bc59ff04e8b0cbc445345098debc9a
    Merge: e19b43a3 ba7081c0
    Author: Li Bo <drluodian@gmail.com>
    Date:   Mon Jul 1 11:47:09 2024 +0800

        Merge pull request #129 from Dannoopsy/mmbench_ru

        add task MMBench-ru

    commit e19b43a3a1e7212e623061b164b0419cc0dda689
    Merge: 11fd7e3f a0de8970
    Author: Li Bo <drluodian@gmail.com>
    Date:   Mon Jul 1 11:46:58 2024 +0800

        Merge pull request #128 from Dannoopsy/gqa-ru

        add task gqa-ru

    commit 11fd7e3fc05908aeb01e4a6161a7b55cd38b3122
    Merge: 383e7fea a7522592
    Author: Li Bo <drluodian@gmail.com>
    Date:   Mon Jul 1 11:46:16 2024 +0800

        Merge pull request #130 from lscpku/vitatecs

        Add task VITATECS

    commit a75225926e5954f85466d257f99acf0163fde596
    Author: lscpku <lisc99@pku.edu.cn>
    Date:   Fri Jun 28 20:37:06 2024 +0800

        create new task vitatecs

    commit ba7081c0abac840002d320e30733e891298dfa11
    Author: Dannoopsy <63581325+Dannoopsy@users.noreply.github.com>
    Date:   Fri Jun 28 12:21:05 2024 +0300

        change prompt to ru

    commit 27ea9c0055a8abf3a8198829b8617018479918e2
    Author: Dannoopsy <belopolskikh.dd@phystech.edu>
    Date:   Thu Jun 27 17:17:29 2024 +0000

        add mmbench_ru_dev

    commit 383e7fead3138aedf62e9c0ec48303835ef26e2a
    Merge: 06fa000f ed2e7f79
    Author: Li Bo <drluodian@gmail.com>
    Date:   Fri Jun 28 00:14:10 2024 +0800

        Merge pull request #126 from lorenzomammana/feature/external-package-integration

        External package integration using plugins

    commit ed2e7f792151d21bce8f1c498270b9391e1d5c85
    Merge: 03947e14 06fa000f
    Author: Lorenzo Mammana <mammanalorenzo@outlook.it>
    Date:   Thu Jun 27 15:38:10 2024 +0000

        Merge branch 'main' into feature/external-package-integration

    commit a0de89708d5e6f259bb17f0eaace3c5b901b275c
    Author: Dannoopsy <belopolskikh.dd@phystech.edu>
    Date:   Tue Jun 25 11:11:37 2024 +0000

        new task gqa-ru

    commit 06fa000f60d3e4d160fac8ceb9959ae92a98f752
    Author: kcz358 <kaichenzhang358@outlook.com>
    Date:   Tue Jun 25 06:41:13 2024 +0000

        Fix vid mme post prompt issue

    commit b388d79e0df6f60068196cb7047453ebd22d6ef1
    Author: Li Bo <drluodian@gmail.com>
    Date:   Sun Jun 23 22:31:16 2024 +0800

        Update activitynetqa_generation.yaml

    commit 8f9d620fcd9d0a0742ee6bcf51ea63bd6b088a36
    Author: Li Bo <drluodian@gmail.com>
    Date:   Sun Jun 23 14:02:25 2024 +0800

        Update pyproject.toml

    commit 6341b7c15ce9fb28eb06b067ddb299d6cf2e16c3
    Merge: fce85f1b 903b042b
    Author: Li Bo <drluodian@gmail.com>
    Date:   Sun Jun 23 14:02:02 2024 +0800

        Merge pull request #125 from EvolvingLMMs-Lab/dev/interleave

        [Model] aligned llava-interleave model results on video tasks

    commit 903b042be016016d4ebeecb07701f3076a2d323c
    Author: kcz358 <kaichenzhang358@outlook.com>
    Date:   Sat Jun 22 12:07:13 2024 +0000

        Remove unnecessary lines for video llava

    commit d78ec86407b729a964906a8c2e50704b4bc74d06
    Merge: ebe7217a fce85f1b
    Author: Li Bo <drluodian@gmail.com>
    Date:   Sat Jun 22 13:57:31 2024 +0800

        Merge branch 'main' into dev/interleave

    commit ebe7217a486c1e754e42c2cbdb834e09fbbcc9b0
    Author: kcz358 <kaichenzhang358@outlook.com>
    Date:   Sat Jun 22 02:57:08 2024 +0000

        Delete unnecessary lines

    commit 120c474b056f9177c74e1fd9691d59e2f234b785
    Author: kcz358 <kaichenzhang358@outlook.com>
    Date:   Fri Jun 21 08:38:41 2024 +0000

        Revise model registry for llava_hf and longva

    commit 7d6201f921088afd3f52a35076e3c6fcc9aa518c
    Author: kcz358 <kaichenzhang358@outlook.com>
    Date:   Fri Jun 21 08:38:24 2024 +0000

        Add longva

    commit 12f480699c71a12a24d4349d9b0681933201a3a6
    Author: kcz358 <kaichenzhang358@outlook.com>
    Date:   Fri Jun 21 08:35:39 2024 +0000

        Remove unnecessary lines since use batched visuals now in llava

    commit 12cea76f1f0f14b1fd1007c9d39a9b0557368637
    Author: Bo Li <drluodian@gmail.com>
    Date:   Thu Jun 20 18:15:32 2024 +0000

        chore: Add loguru for logging in lmms_eval package

    commit 03947e14a46fd25b412931f7c9c25f4a2971d0b4
    Author: Lorenzo Mammana <mammanalorenzo@outlook.it>
    Date:   Wed Jun 5 13:40:41 2024 +0000

        feat: Allow including external tasks from plugins

    commit b80a91f73e15ddd0b0ce1322d7d121fa14030eed
    Author: Lorenzo Mammana <mammanalorenzo@outlook.it>
    Date:   Wed Jun 5 13:04:55 2024 +0000

        feat: Allow loading model configurations from other packages

    commit 8ef24740dd48a11c97eb627f2fff4aca107fef0d
    Author: Bo Li <drluodian@gmail.com>
    Date:   Thu Jun 20 12:11:03 2024 +0000

        chore: Remove unused models from lmms_eval package

    commit af38885fc2e066f5ea44388f33e07176f836fe28
    Author: Bo Li <drluodian@gmail.com>
    Date:   Thu Jun 20 12:07:09 2024 +0000

        chore: Handle ImportError when importing models

        Handle the ImportError exception when importing models in the lmms_eval package. This change adds a try-except block to catch the ImportError and print an error message indicating the failed import. This will help with troubleshooting and identifying any issues with the model imports.

    commit fce85f1b03ff7043b29dee787c5d17a08dd2687a
    Merge: dbe63293 d94f83cb
    Author: Li Bo <drluodian@gmail.com>
    Date:   Thu Jun 20 20:02:12 2024 +0800

        Merge pull request #120 from EvolvingLMMs-Lab/pufanyi/hf_dataset_docs

        Add docs for datasets upload to HF

    commit dbe63293245a5141fdfd80bda7657c304f6bd32f
    Author: choiszt <ls2001927@sohu.com>
    Date:   Thu Jun 20 15:14:21 2024 +0800

        update ablation for videomme datasets

    commit d94f83cb3f08b61a2c75cc4326e58792100605b3
    Author: Li Bo <drluodian@gmail.com>
    Date:   Thu Jun 20 13:30:59 2024 +0800

        Update README.md

    commit cab8159ff35db330536c0b6dfb4b0a3b24142209
    Author: Li Bo <drluodian@gmail.com>
    Date:   Thu Jun 20 13:30:29 2024 +0800

        Update README.md

    commit 45876652a877a8006b828f32f5cc4660629f9190
    Author: kcz358 <kaichenzhang358@outlook.com>
    Date:   Thu Jun 20 03:55:30 2024 +0000

        Add llava_hf back to registry

    commit 3463651b8c54d36cd94169e3d376f5ed225a195a
    Author: kcz358 <kaichenzhang358@outlook.com>
    Date:   Thu Jun 20 03:54:33 2024 +0000

        Remove handling non-visual loop in llava

    commit cb0d3f49b72790b081f981e0e6147131542f7f68
    Author: Fanyi Pu <FPU001@e.ntu.edu.sg>
    Date:   Thu Jun 20 02:11:18 2024 +0800

        update readme

    commit 813877bfe5ac590cdbe92dd74d18f83a2091f748
    Author: Fanyi Pu <FPU001@e.ntu.edu.sg>
    Date:   Wed Jun 19 15:37:52 2024 +0800

        to sh script

    commit a14684b8557d5894976448a5c559ed7a66a6cf16
    Author: Fanyi Pu <FPU001@e.ntu.edu.sg>
    Date:   Wed Jun 19 15:37:04 2024 +0800

        lint

    commit d0f8851d42ba31f5da2a7a65e91499db45174dbc
    Author: Fanyi Pu <FPU001@e.ntu.edu.sg>
    Date:   Wed Jun 19 15:36:48 2024 +0800

        small fix

    commit 63748e9718f287ad433afc90e340b5e17a89c1ed
    Author: Fanyi Pu <FPU001@e.ntu.edu.sg>
    Date:   Wed Jun 19 15:36:43 2024 +0800

        small fix

    commit 7f1159a1fe04cfb783dc31d4fbdef3bda0ce19e4
    Author: Fanyi Pu <FPU001@e.ntu.edu.sg>
    Date:   Wed Jun 19 15:35:05 2024 +0800

        update preparation

    commit 19f9bd621c76a483ff98f8c7eb78f64753da683a
    Author: Fanyi Pu <FPU001@e.ntu.edu.sg>
    Date:   Wed Jun 19 15:23:24 2024 +0800

        docs

    commit ce6f889ba02d819979c7922f6336cf4f1f718f65
    Author: Fanyi Pu <FPU001@e.ntu.edu.sg>
    Date:   Wed Jun 19 15:04:16 2024 +0800

        tutorial

    commit f513c520c2a3dad26d2b2ca5c4ed4db05a493c73
    Author: Bo Li <drluodian@gmail.com>
    Date:   Wed Jun 19 06:51:19 2024 +0000

        chore: Update dependencies to fix potential risks and improve compatibility

    commit efb529552c5e4ba039a4cba8e9aa5cb7ba65bf90
    Author: kcz358 <kaichenzhang358@outlook.com>
    Date:   Wed Jun 19 10:25:58 2024 +0800

        Release llava-wilder

    commit 742651fc9daf97e2f57831ed6e6e7ee7ead7d555
    Author: Fanyi Pu <FPU001@e.ntu.edu.sg>
    Date:   Wed Jun 19 07:44:26 2024 +0800

        feat: Add support for auto downloading tar format videos

    commit 511b6259828212fcba954cdeb8cf90d6e5daabf8
    Merge: 22a4958e 050b2c37
    Author: Bo Li <drluodian@gmail.com>
    Date:   Tue Jun 18 17:01:03 2024 +0000

        Merge branch 'main' of https://github.com/EvolvingLMMs-Lab/lmms-eval

    commit 050b2c370017e9b97475dd6cf01fd051b5ca5c86
    Merge: 74facb41 ef306512
    Author: Li Bo <drluodian@gmail.com>
    Date:   Tue Jun 18 13:13:38 2024 +0800

        Merge pull request #114 from zjysteven/add-tinyllava

        add tinyllava

    commit ef306512e5135f76dffa383f600b8733015836e8
    Author: Jingyang Zhang <jingyang.zhang@duke.edu>
    Date:   Mon Jun 17 17:57:02 2024 -0400

        fix typo

    commit 9bab67732a4238097725deddf867fb1946ffee40
    Merge: dbfb2387 74facb41
    Author: Jingyang Zhang <jingyang.zhang@duke.edu>
    Date:   Sun Jun 16 10:56:05 2024 -0400

        Merge branch 'EvolvingLMMs-Lab:main' into add-tinyllava

    commit 74facb41a826691dfce4458cf1d8659b34fc5bf5
    Merge: 8ba192f9 d5df72de
    Author: Li Bo <drluodian@gmail.com>
    Date:   Sun Jun 16 17:59:19 2024 +0800

        Merge pull request #118 from teowu/main

        Fix the potential risk by PR #117

    commit d5df72de2d03108d6b365818ecc3551ac9aa6302
    Merge: 5bf59ed2 8ba192f9
    Author: Teo (Timothy) Wu Haoning <38696372+teowu@users.noreply.github.com>
    Date:   Sun Jun 16 15:32:13 2024 +0800

        Merge branch 'EvolvingLMMs-Lab:main' into main

    commit 5bf59ed250da98a408a94e214a73caa400cba842
    Author: teowu <realtimothyhwu@gmail.com>
    Date:   Sun Jun 16 07:27:28 2024 +0000

        fix #117, allow auto download with tar format videos

    commit 98b3955cb808e…
  • Loading branch information
@Luodian @pufanyi
@kcz358 @shuyansy @cocoshe @gagan3012 @CaraJ7 @tupini07 @hunterheiden @vfragoso @AtsuMiyai @Gumpest @tianyu-z @sheryc @XinrunDu @teowu @zjysteven @choiszt @lorenzomammana @Dannoopsy @lscpku @Dousia @KairuiHu
29 people authored Sep 1, 2024
1 parent 7adbd69 commit 7548ebe
Show file tree
Hide file tree
Showing 680 changed files with 16,847 additions and 1,953 deletions.
10 changes: 6 additions & 4 deletions .github/workflows/black.yml → .github/workflows/lint.yml
View file
Original file line number Diff line number Diff line change
Expand Up @@ -11,7 +11,9 @@ jobs:
uses: actions/setup-python@v4
with:
python-version: '3.9'
- name: Install specific version of Black
run: pip install black==23.9.1
- name: Run Black
run: black --line-length=240 ./
- name: Install pre-commit hook
run: |
python -m pip install pre-commit
pre-commit install
- name: Linting
run: pre-commit run --all-files
1 change: 1 addition & 0 deletions .gitignore
View file
Original file line number Diff line number Diff line change
Expand Up @@ -30,6 +30,7 @@ ckpt
pretrained/
LLaVA/
*logs
*.isorted
temp/
InternVL/
logs/
Expand Down
9 changes: 8 additions & 1 deletion .pre-commit-config.yaml
View file
Original file line number Diff line number Diff line change
Expand Up @@ -3,4 +3,11 @@ repos:
rev: 23.12.1
hooks:
- id: black
language_version: python3
language_version: python3
args: ["--line-length=240"]
- repo: https://github.com/PyCQA/isort
rev: 5.13.2
hooks:
- id: isort
language_version: python3
args: ["--profile", "black"]
193 changes: 188 additions & 5 deletions docs/commands.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -8,19 +8,202 @@ Equivalently, running the library can be done via the `lmms_eval` entrypoint at

This mode supports a number of command-line arguments, the details of which can be also be seen via running with `-h` or `--help`:

* `--model` : Selects which model type or provider is evaluated. Must be a mdoels registered under lmms_eval/models. For example, `--model qwen_vl` or `--model llava`.
- `--model` : Selects which model type or provider is evaluated. Must be a string corresponding to the name of the model type/provider being used. See [the main README](https://github.com/EleutherAI/lm-evaluation-harness/tree/main#model-apis-and-inference-servers) for a full list of enabled model names and supported libraries or APIs.

* `--model_args` : Controls parameters passed to the model constructor. Accepts a string containing comma-separated keyword arguments to the model class of the format `"arg1=val1,arg2=val2,..."`, such as, for example `--model_args pretrained=liuhaotian/llava-v1.5-7b,batch_size=1`. For a full list of what keyword arguments, see the initialization of the corresponding model class in `lmms_eval/models/`.

* `--tasks` : Determines which tasks or task groups are evaluated. Accepts a comma-separated list of task names or task group names. Must be solely comprised of valid tasks/groups. You can use `--tasks list` to see all the available tasks. If you add your own tasks but not shown on the list, you can try to set `--verbosity=DEBUG` to view the error message. You can also use `--tasks list_with_num` to check every tasks and the number of question each task contains. However, `list_with_num` will download all the available datasets and may require lots of memory and time.

* `--batch_size` : Sets the batch size used for evaluation. Can be a positive integer or `"auto"` to automatically select the largest batch size that will fit in memory, speeding up evaluation. One can pass `--batch_size auto:N` to re-select the maximum batch size `N` times during evaluation. This can help accelerate evaluation further, since `lm-eval` sorts documents in descending order of context length.
- `--num_fewshot` : Sets the number of few-shot examples to place in context. Must be an integer.

* `--output_path` : A string of the form `dir/file.jsonl` or `dir/`. Provides a path where high-level results will be saved, either into the file named or into the directory named. If `--log_samples` is passed as well, then per-document outputs and metrics will be saved into the directory as well.
- `--gen_kwargs` : takes an arg string in same format as `--model_args` and creates a dictionary of keyword arguments. These will be passed to the models for all called `generate_until` (free-form or greedy generation task) tasks, to set options such as the sampling temperature or `top_p` / `top_k`. For a list of what args are supported for each model type, reference the respective library's documentation (for example, the documentation for `transformers.AutoModelForCausalLM.generate()`.) These kwargs will be applied to all `generate_until` tasks called--we do not currently support unique gen_kwargs or batch_size values per task in a single run of the library. To control these on a per-task level, set them in that task's YAML file.

* `--log_samples` : If this flag is passed, then the model's outputs, and the text fed into the model, will be saved at per-document granularity. Must be used with `--output_path`.
- `--batch_size` : Sets the batch size used for evaluation. Can be a positive integer or `"auto"` to automatically select the largest batch size that will fit in memory, speeding up evaluation. One can pass `--batch_size auto:N` to re-select the maximum batch size `N` times during evaluation. This can help accelerate evaluation further, since `lm-eval` sorts documents in descending order of context length.

- `--max_batch_size` : Sets the maximum batch size to try to fit in memory, if `--batch_size auto` is passed.

- `--device` : Sets which device to place the model onto. Must be a string, for example, `"cuda", "cuda:0", "cpu", "mps"`. Defaults to "cuda", and can be ignored if running multi-GPU or running a non-local model type.

- `--output_path` : A string of the form `dir/file.jsonl` or `dir/`. Provides a path where high-level results will be saved, either into the file named or into the directory named. If `--log_samples` is passed as well, then per-document outputs and metrics will be saved into the directory as well.

- `--log_samples` : If this flag is passed, then the model's outputs, and the text fed into the model, will be saved at per-document granularity. Must be used with `--output_path`.

- `--limit` : Accepts an integer, or a float between 0.0 and 1.0 . If passed, will limit the number of documents to evaluate to the first X documents (if an integer) per task or first X% of documents per task. Useful for debugging, especially on costly API models.

- `--use_cache` : Should be a path where a sqlite db file can be written to. Takes a string of format `/path/to/sqlite_cache_` in order to create a cache db at `/path/to/sqlite_cache_rank{i}.db` for each process (0-NUM_GPUS). This allows results of prior runs to be cached, so that there is no need to re-run results in order to re-score or re-run a given (model, task) pair again.

- `--cache_requests` : Can be "true", "refresh", or "delete". "true" means that the cache should be used. "refresh" means that you wish to regenerate the cache, which you should run if you change your dataset configuration for a given task. "delete" will delete the cache. Cached files are stored under lm_eval/cache/.cache unless you specify a different path via the environment variable: `LM_HARNESS_CACHE_PATH`. e.g. `LM_HARNESS_CACHE_PATH=~/Documents/cache_for_lm_harness`.

- `--check_integrity` : If this flag is used, the library tests for each task selected are run to confirm task integrity.

- `--write_out` : Used for diagnostic purposes to observe the format of task documents passed to a model. If this flag is used, then prints the prompt and gold target string for the first document of each task.

- `--show_config` : If used, prints the full `lm_eval.api.task.TaskConfig` contents (non-default settings the task YAML file) for each task which was run, at the completion of an evaluation. Useful for when one is modifying a task's configuration YAML locally to transmit the exact configurations used for debugging or for reproducibility purposes.

- `--include_path` : Accepts a path to a folder. If passed, then all YAML files containing `lm-eval` compatible task configurations will be added to the task registry as available tasks. Used for when one is writing config files for their own task in a folder other than `lm_eval/tasks/`.

- `--system_instruction`: Specifies a system instruction string to prepend to the prompt.

- `--apply_chat_template` : This flag specifies whether to apply a chat template to the prompt. It can be used in the following ways:
- `--apply_chat_template` : When used without an argument, applies the only available chat template to the prompt. For Hugging Face models, if no dedicated chat template exists, the default chat template will be applied.
- `--apply_chat_template template_name` : If the model has multiple chat templates, apply the specified template to the prompt.

For Hugging Face models, the default chat template can be found in the [`default_chat_template`](https://github.com/huggingface/transformers/blob/fc35907f95459d7a6c5281dfadd680b6f7b620e3/src/transformers/tokenization_utils_base.py#L1912) property of the Transformers Tokenizer.

- `--fewshot_as_multiturn` : If this flag is on, the Fewshot examples are treated as a multi-turn conversation. Questions are provided as user content and answers are provided as assistant responses. Requires `--num_fewshot` to be set to be greater than 0, and `--apply_chat_template` to be on.

- `--predict_only`: Generates the model outputs without computing metrics. Use with `--log_samples` to retrieve decoded results.

* `--seed`: Set seed for python's random, numpy and torch. Accepts a comma-separated list of 3 values for python's random, numpy, and torch seeds, respectively, or a single integer to set the same seed for all three. The values are either an integer or 'None' to not set the seed. Default is `0,1234,1234` (for backward compatibility). E.g. `--seed 0,None,8` sets `random.seed(0)` and `torch.manual_seed(8)`. Here numpy's seed is not set since the second value is `None`. E.g, `--seed 42` sets all three seeds to 42.

* `--wandb_args`: Tracks logging to Weights and Biases for evaluation runs and includes args passed to `wandb.init`, such as `project` and `job_type`. Full list [here](https://docs.wandb.ai/ref/python/init). e.g., ```--wandb_args project=test-project,name=test-run```

* `--hf_hub_log_args` : Logs evaluation results to Hugging Face Hub. Accepts a string with the arguments separated by commas. Available arguments:
* `hub_results_org` - organization name on Hugging Face Hub, e.g., `EleutherAI`. If not provided, the results will be pushed to the owner of the Hugging Face token,
* `hub_repo_name` - repository name on Hugging Face Hub (deprecated, `details_repo_name` and `results_repo_name` should be used instead), e.g., `lm-eval-results`,
* `details_repo_name` - repository name on Hugging Face Hub to store details, e.g., `lm-eval-results`,
* `results_repo_name` - repository name on Hugging Face Hub to store results, e.g., `lm-eval-results`,
* `push_results_to_hub` - whether to push results to Hugging Face Hub, can be `True` or `False`,
* `push_samples_to_hub` - whether to push samples results to Hugging Face Hub, can be `True` or `False`. Requires `--log_samples` to be set,
* `public_repo` - whether the repository is public, can be `True` or `False`,
* `leaderboard_url` - URL to the leaderboard, e.g., `https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard`.
* `point_of_contact` - Point of contact for the results dataset, e.g., `yourname@example.com`.
* `gated` - whether to gate the details dataset, can be `True` or `False`.

## External Library Usage

We also support using the library's external API for use within model training loops or other scripts.

`lmms_eval` supplies two functions for external import and use: `lmms_eval.evaluate()` and `lmms_eval.simple_evaluate()`.

`simple_evaluate()` can be used by simply creating an `lmms_eval.api.model.LM` subclass that implements the methods described in the [Model Guide](https://github.com/EleutherAI/lm-evaluation-harness/tree/main/docs/model_guide.md), and wrapping your custom model in that class as follows:

```python
import lmms_eval
...

my_model = initialize_my_model() # create your model (could be running finetuning with some custom modeling code)
...
# instantiate an LM subclass that takes your initialized model and can run
# - `Your_LMM.loglikelihood()`
# - `Your_LMM.generate_until()`
lmm_obj = Your_LMM(model=my_model, batch_size=16)

# indexes all tasks from the `lmms_eval/tasks` subdirectory.
# Alternatively, you can set `TaskManager(include_path="path/to/my/custom/task/configs")`
# to include a set of tasks in a separate directory.
task_manager = lmms_eval.tasks.TaskManager()

# Setting `task_manager` to the one above is optional and should generally be done
# if you want to include tasks from paths other than ones in `lmms_eval/tasks`.
# `simple_evaluate` will instantiate its own task_manager if it is set to None here.
results = lmms_eval.simple_evaluate( # call simple_evaluate
model=lmm_obj,
tasks=["taskname1", "taskname2"],
num_fewshot=0,
task_manager=task_manager,
...
)
```

See the `simple_evaluate()` and `evaluate()` functions in [lmms_eval/evaluator.py](../lmms_eval/evaluator.py#:~:text=simple_evaluate) for a full description of all arguments available. All keyword arguments to simple_evaluate share the same role as the command-line flags described previously.

Additionally, the `evaluate()` function offers the core evaluation functionality provided by the library, but without some of the special handling and simplification + abstraction provided by `simple_evaluate()`.

As a brief example usage of `evaluate()`:

```python
import lmms_eval

# suppose you've defined a custom lm_eval.api.Task subclass in your own external codebase
from my_tasks import MyTask1
...

# create your model (could be running finetuning with some custom modeling code)
my_model = initialize_my_model()
...

# instantiate an LM subclass that takes your initialized model and can run
# - `Your_LM.loglikelihood()`
# - `Your_LM.loglikelihood_rolling()`
# - `Your_LM.generate_until()`
lmm_obj = Your_LMM(model=my_model, batch_size=16)

# optional: the task_manager indexes tasks including ones
# specified by the user through `include_path`.
task_manager = lmms_eval.tasks.TaskManager(
include_path="/path/to/custom/yaml"
)

# To get a task dict for `evaluate`
task_dict = lmms_eval.tasks.get_task_dict(
[
"mmlu", # A stock task
"my_custom_task", # A custom task
{
"task": ..., # A dict that configures a task
"doc_to_text": ...,
},
MyTask1 # A task object from `lm_eval.task.Task`
],
task_manager # A task manager that allows lm_eval to
# load the task during evaluation.
# If none is provided, `get_task_dict`
# will instantiate one itself, but this
# only includes the stock tasks so users
# will need to set this if including
# custom paths is required.
)

results = evaluate(
lm=lmm_obj,
task_dict=task_dict,
...
)
```

## Usage with SRT API

> install sglang
```bash
git clone https://github.com/sgl-project/sglang.git
cd sglang;
pip install -e "python[srt]"
python3 -m pip install flashinfer -i https://flashinfer.ai/whl/cu121/torch2.3/
```

> run sglang backend service with the following command
```bash

CKPT_PATH=$1
TASK=$2
MODALITY=$3
TP_SIZE=$4
echo $TASK
TASK_SUFFIX="${TASK//,/_}"
echo $TASK_SUFFIX

python3 -m lmms_eval \
--model srt_api \
--model_args modality=$MODALITY,model_version=$CKPT_PATH,tp=$TP_SIZE,host=127.0.0.1,port=30000,timeout=600 \
--tasks $TASK \
--batch_size 1 \
--log_samples \
--log_samples_suffix $TASK_SUFFIX \
--output_path ./logs/

```

You may need to install some dependencies for the above command to work (if you encounter some errors).

```bash
pip install httpx==0.23.3
pip install protobuf==3.20
pip install flashinfer -i https://flashinfer.ai/whl/cu121/torch2.3/
```

* `--limit` : Accepts an integer, or a float between 0.0 and 1.0 . If passed, will limit the number of documents to evaluate to the first X documents (if an integer) per task or first X% of documents per task. Useful for debugging, especially on costly API models.

## Usage with SRT API

Expand Down
2 changes: 1 addition & 1 deletion docs/model_guide.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -63,7 +63,7 @@ Congrats on implementing your model! Now it's time to test it out.

To make your model usable via the command line interface to `lmms_eval`, you'll need to tell `lmms_eval` what your model's name is.

This is done via a *decorator*, `lmms_eval.api.registry.register_model`. Using `register_model()`, one can both tell the package what the model's name(s) to be used are when invoking it with `python -m lm_eval --model <name>` and alert `lmms_eval` to the model's existence.
This is done via a *decorator*, `lmms_eval.api.registry.register_model`. Using `register_model()`, one can both tell the package what the model's name(s) to be used are when invoking it with `python -m lmms-eval --model <name>` and alert `lmms_eval` to the model's existence.

```python
from lmms_eval.api.registry import register_model
Expand Down
Loading

0 comments on commit 7548ebe

Please sign in to comment.