diff --git a/.gitattributes b/.gitattributes index a6344aac8c09253b3b630fb776ae94478aa0275b..12ae3ac19ddf750be8eb15c631164db1098cc1f0 100644 --- a/.gitattributes +++ b/.gitattributes @@ -33,3 +33,7 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text *.zip filter=lfs diff=lfs merge=lfs -text *.zst filter=lfs diff=lfs merge=lfs -text *tfevents* filter=lfs diff=lfs merge=lfs -text +llava_base/llava_base_SEEDBench_IMG_openai_result.xlsx filter=lfs diff=lfs merge=lfs -text +llava_base_v2/llava_base_v2_SEEDBench_IMG_openai_result.xlsx filter=lfs diff=lfs merge=lfs -text +llava_moe_e4t2/llava_moe_e4t2_SEEDBench_IMG_openai_result.xlsx filter=lfs diff=lfs merge=lfs -text +llava_moe_e5t3/llava_moe_e5t3_SEEDBench_IMG_openai_result.xlsx filter=lfs diff=lfs merge=lfs -text diff --git a/llava_base/01_MMMU_DEV_VAL.pkl b/llava_base/01_MMMU_DEV_VAL.pkl new file mode 100644 index 0000000000000000000000000000000000000000..f0a67f51aee7b2d213809e426c33090b8407f958 --- /dev/null +++ b/llava_base/01_MMMU_DEV_VAL.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:da276d484dc732bf2299ec90aed1b96016502c1a06adb344cf4bf5b23ace0127 +size 2689 diff --git a/llava_base/llava_base_MMBench_DEV_EN.xlsx b/llava_base/llava_base_MMBench_DEV_EN.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..d41254287bc69e0e7947f890100d6b8e056145dc Binary files /dev/null and b/llava_base/llava_base_MMBench_DEV_EN.xlsx differ diff --git a/llava_base/llava_base_MMBench_DEV_EN_acc.csv b/llava_base/llava_base_MMBench_DEV_EN_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..8f3beb93cd66a7e23b4ffa1079b1e43d84063c98 --- /dev/null +++ b/llava_base/llava_base_MMBench_DEV_EN_acc.csv @@ -0,0 +1,2 @@ +"split","Overall","AR","CP","FP-C","FP-S","LR","RR","action_recognition","attribute_comparison","attribute_recognition","celebrity_recognition","function_reasoning","future_prediction","identity_reasoning","image_emotion","image_quality","image_scene","image_style","image_topic","nature_relation","object_localization","ocr","physical_property_reasoning","physical_relation","social_relation","spatial_relationship","structuralized_imagetext_understanding" +"dev","0.6030927835051546","0.6884422110552764","0.7331081081081081","0.5804195804195804","0.5972696245733788","0.3305084745762712","0.4434782608695652","0.8703703703703703","0.5454545454545454","0.6891891891891891","0.7676767676767676","0.7468354430379747","0.4","0.9555555555555556","0.8","0.20754716981132076","0.9519230769230769","0.7358490566037735","0.7777777777777778","0.375","0.30864197530864196","0.5897435897435898","0.4666666666666667","0.25","0.627906976744186","0.26666666666666666","0.2948717948717949" diff --git a/llava_base/llava_base_MMBench_DEV_EN_openai_result.pkl b/llava_base/llava_base_MMBench_DEV_EN_openai_result.pkl new file mode 100644 index 0000000000000000000000000000000000000000..272bb4c07b57c82461f17b1cab765e57c5bd530d --- /dev/null +++ b/llava_base/llava_base_MMBench_DEV_EN_openai_result.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:06c63ec6ed5f9cdca6b4ccf69e141852ad475ee14284447c3b1d9452256faf47 +size 55132 diff --git a/llava_base/llava_base_MMBench_DEV_EN_openai_result.xlsx b/llava_base/llava_base_MMBench_DEV_EN_openai_result.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..637805eb088297992fea5deccee3e6ecede2fe4c Binary files /dev/null and b/llava_base/llava_base_MMBench_DEV_EN_openai_result.xlsx differ diff --git a/llava_base/llava_base_MME.xlsx b/llava_base/llava_base_MME.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..b124fde839debc3752cf518b5e9dff41b5dfcf41 Binary files /dev/null and b/llava_base/llava_base_MME.xlsx differ diff --git a/llava_base/llava_base_MME_PREV.pkl b/llava_base/llava_base_MME_PREV.pkl new file mode 100644 index 0000000000000000000000000000000000000000..d1b9170503ec017fa82585d3bc9c8698acec9808 --- /dev/null +++ b/llava_base/llava_base_MME_PREV.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:45cd4e356c0a3339e3c1181bc0c03bdb9f217b8f44bb754031a65f3aca410962 +size 44296 diff --git a/llava_base/llava_base_MME_auxmatch.xlsx b/llava_base/llava_base_MME_auxmatch.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..9054aa032eccdac4367fb209d88a36d287d05f68 Binary files /dev/null and b/llava_base/llava_base_MME_auxmatch.xlsx differ diff --git a/llava_base/llava_base_MME_score.csv b/llava_base/llava_base_MME_score.csv new file mode 100644 index 0000000000000000000000000000000000000000..72cd7fb8ad6cc80cc1e1c6c7bca03dd18794307b --- /dev/null +++ b/llava_base/llava_base_MME_score.csv @@ -0,0 +1,2 @@ +"perception","reasoning","OCR","artwork","celebrity","code_reasoning","color","commonsense_reasoning","count","existence","landmark","numerical_calculation","position","posters","scene","text_translation" +"1259.7040816326532","214.28571428571428","130.0","108.0","105.0","22.5","173.33333333333334","104.28571428571428","125.0","143.33333333333331","75.0","12.5","133.33333333333334","110.20408163265304","156.5","75.0" diff --git a/llava_base/llava_base_MMVet.xlsx b/llava_base/llava_base_MMVet.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..f5a5209358f96b97039b27f3aa61e27e1379c4d9 Binary files /dev/null and b/llava_base/llava_base_MMVet.xlsx differ diff --git a/llava_base/llava_base_MathVista_MINI.xlsx b/llava_base/llava_base_MathVista_MINI.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..76c73a835bf3a4bd2d917367f57cf23ebbd0bd2b Binary files /dev/null and b/llava_base/llava_base_MathVista_MINI.xlsx differ diff --git a/llava_base/llava_base_POPE.xlsx b/llava_base/llava_base_POPE.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..05dd8560f5c963d50c0c07d95e29ed76a91e1c7a Binary files /dev/null and b/llava_base/llava_base_POPE.xlsx differ diff --git a/llava_base/llava_base_POPE_auxmatch.xlsx b/llava_base/llava_base_POPE_auxmatch.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..ca92fce653f6ad9f78cfd0aa5dd2545d9e70131d Binary files /dev/null and b/llava_base/llava_base_POPE_auxmatch.xlsx differ diff --git a/llava_base/llava_base_POPE_score.csv b/llava_base/llava_base_POPE_score.csv new file mode 100644 index 0000000000000000000000000000000000000000..dfa7aaeddb79377758dcdd1af7981184f28dd9f6 --- /dev/null +++ b/llava_base/llava_base_POPE_score.csv @@ -0,0 +1,5 @@ +"split","Overall","acc","precision","recall" +"Overall","81.02643856920683","75.3","97.20149253731343","69.46666666666667" +"adversarial","80.37022753567297","77.3","95.33394327538883","69.46666666666667" +"popular","81.12105877773453","74.0","97.47427502338635","69.46666666666667" +"random","81.59749412685983","74.6","98.86148007590133","69.46666666666667" diff --git a/llava_base/llava_base_SEEDBench_IMG.xlsx b/llava_base/llava_base_SEEDBench_IMG.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..d2f7457730d68db265ae283c9f12d56291056d66 Binary files /dev/null and b/llava_base/llava_base_SEEDBench_IMG.xlsx differ diff --git a/llava_base/llava_base_SEEDBench_IMG_PREV.pkl b/llava_base/llava_base_SEEDBench_IMG_PREV.pkl new file mode 100644 index 0000000000000000000000000000000000000000..bfbb60a3b93f9f0cb9469dc1906a6999fc03e2ea --- /dev/null +++ b/llava_base/llava_base_SEEDBench_IMG_PREV.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ebaed15a17f9354992351ba742fbe8f96a156f164028239caa1513dc0b4128bc +size 82582 diff --git a/llava_base/llava_base_SEEDBench_IMG_acc.csv b/llava_base/llava_base_SEEDBench_IMG_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..90418ad3a5eceeb66a54623132a3388c3c73e2e8 --- /dev/null +++ b/llava_base/llava_base_SEEDBench_IMG_acc.csv @@ -0,0 +1,2 @@ +"split","Overall","Instance Attributes","Instance Identity","Instance Interaction","Instance Location","Instances Counting","Scene Understanding","Spatial Relation","Text Understanding","Visual Reasoning" +"none","0.5265598650927488","0.5091417509141751","0.590387766247952","0.6597938144329897","0.5040899795501023","0.27257866775643647","0.7178594046865104","0.4337899543378995","0.2857142857142857","0.743202416918429" diff --git a/llava_base/llava_base_SEEDBench_IMG_openai_result.pkl b/llava_base/llava_base_SEEDBench_IMG_openai_result.pkl new file mode 100644 index 0000000000000000000000000000000000000000..cdeb36f31132e3cd00a099d3d90a0e16ed799301 --- /dev/null +++ b/llava_base/llava_base_SEEDBench_IMG_openai_result.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:00aefe1f2727f4d934914a2bed5d384533c793f46ce6b634f6bf1132a32a765a +size 1052099 diff --git a/llava_base/llava_base_SEEDBench_IMG_openai_result.xlsx b/llava_base/llava_base_SEEDBench_IMG_openai_result.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..81b5935e81d352042edc7f4639d3b9e617d41dc3 --- /dev/null +++ b/llava_base/llava_base_SEEDBench_IMG_openai_result.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7b848ae407b25ea999e8e77dac01c39f68d147660cfdc70f0a9bf556535bb133 +size 1061587 diff --git a/llava_base/llava_base_ScienceQA_VAL.xlsx b/llava_base/llava_base_ScienceQA_VAL.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..2e6599e3babb6768c7c301cb7f8446ba6dca01ee Binary files /dev/null and b/llava_base/llava_base_ScienceQA_VAL.xlsx differ diff --git a/llava_base/llava_base_ScienceQA_VAL_acc.csv b/llava_base/llava_base_ScienceQA_VAL_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..265ee3e55d49b25a5ed9f65cb8e3f2bf3b9b887e --- /dev/null +++ b/llava_base/llava_base_ScienceQA_VAL_acc.csv @@ -0,0 +1,2 @@ +"split","Overall","Adaptations","Adaptations and natural selection","Age of Exploration","Ancient Egypt and Kush","Ancient Mesopotamia","Animals","Astronomy","Atoms and molecules","Basic economic principles","Chemical reactions","Cities","Classification","Classification and scientific names","Climate change","Colonial America","Context clues","Descriptive details","Designing experiments","Domain-specific vocabulary","Early 19th century American history","Early Americas","Earth events","Ecological interactions","Ecosystems","Engineering practices","English colonies in North America","Force and motion","Fossils","Genes to traits","Geography","Government","Independent reading comprehension","Informational texts: level 1","Magnets","Maps","Materials","Medieval Asia","Natural resources and human impacts","Oceania: geography","Oceans and continents","Oceans and continents ","Particle motion and energy","Persuasive strategies","Physical Geography","Plant reproduction","Plants","Plate tectonics","Read-alone texts","Rocks and minerals","Rome and the Byzantine Empire","Scientific names","Solutions","State capitals","States","States of matter","The American Revolution","The Americas: geography","The Antebellum period","The Civil War and Reconstruction","The Silk Road","Thermal energy","Velocity, acceleration, and forces","Visual elements","Water cycle","Weather and climate","World religions" +"val","0.5970433953266572","0.625","1.0","1.0","1.0","1.0","1.0","0.5869565217391305","0.6190476190476191","0.53125","1.0","0.125","0.5833333333333334","0.8723404255319149","1.0","0.21428571428571427","0.8","1.0","0.9032258064516129","0.6","0.0","0.5","1.0","0.52","0.7611940298507462","0.9811320754716981","0.1794871794871795","0.8","0.29411764705882354","0.1276595744680851","0.4225352112676056","1.0","1.0","0.8","0.23308270676691728","0.3387096774193548","0.646551724137931","0.5","0.5","0.19148936170212766","0.9","0.5","0.5147058823529411","0.5555555555555556","0.6046511627906976","0.6","0.8","0.75","1.0","0.926829268292683","0.5","0.8571428571428571","0.373134328358209","0.949685534591195","0.3333333333333333","0.8947368421052632","0.6666666666666666","0.13333333333333333","1.0","0.0","1.0","0.4444444444444444","0.19607843137254902","0.5555555555555556","1.0","0.34375","0.6666666666666666" diff --git a/llava_base/llava_base_ScienceQA_VAL_openai_result.pkl b/llava_base/llava_base_ScienceQA_VAL_openai_result.pkl new file mode 100644 index 0000000000000000000000000000000000000000..262b75b939ead23ec2bc9121ed50352f9ccf5bbf --- /dev/null +++ b/llava_base/llava_base_ScienceQA_VAL_openai_result.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:76de04b9a9defce7533e24f24b51d989b49d580b600c9c0c68cb882c3c2a865e +size 142786 diff --git a/llava_base/llava_base_ScienceQA_VAL_openai_result.xlsx b/llava_base/llava_base_ScienceQA_VAL_openai_result.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..8891047dc745bec089739e031815d0d211b3023b Binary files /dev/null and b/llava_base/llava_base_ScienceQA_VAL_openai_result.xlsx differ diff --git a/llava_base/llava_base_TextVQA_VAL.xlsx b/llava_base/llava_base_TextVQA_VAL.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..82f7be7ab41e406788c83b4d82411e36d854fbc0 Binary files /dev/null and b/llava_base/llava_base_TextVQA_VAL.xlsx differ diff --git a/llava_base/llava_base_TextVQA_VAL_acc.csv b/llava_base/llava_base_TextVQA_VAL_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..869088f6e6e914f7ee4b920a607db9593d1c7402 --- /dev/null +++ b/llava_base/llava_base_TextVQA_VAL_acc.csv @@ -0,0 +1,2 @@ +"Overall" +"15.498000000000001" diff --git a/llava_base_v2/01_MMMU_DEV_VAL.pkl b/llava_base_v2/01_MMMU_DEV_VAL.pkl new file mode 100644 index 0000000000000000000000000000000000000000..753b17a23a7646a92746dd46f3c0da4ad77a861c --- /dev/null +++ b/llava_base_v2/01_MMMU_DEV_VAL.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cd91a125b37689d1f7f7177eacba081afacbbd955c0a4688d74182f854fcc93c +size 1432 diff --git a/llava_base_v2/llava_base_v2_MME.xlsx b/llava_base_v2/llava_base_v2_MME.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..ef8228e8f41b91108e65ff2b79f63d12ac129b74 Binary files /dev/null and b/llava_base_v2/llava_base_v2_MME.xlsx differ diff --git a/llava_base_v2/llava_base_v2_MME_auxmatch.xlsx b/llava_base_v2/llava_base_v2_MME_auxmatch.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..d6b378d85645082e20656c86d5235fa12a907f02 Binary files /dev/null and b/llava_base_v2/llava_base_v2_MME_auxmatch.xlsx differ diff --git a/llava_base_v2/llava_base_v2_MME_score.csv b/llava_base_v2/llava_base_v2_MME_score.csv new file mode 100644 index 0000000000000000000000000000000000000000..ee4503300ae3238ddbfdf151a2662b3fa0a7917c --- /dev/null +++ b/llava_base_v2/llava_base_v2_MME_score.csv @@ -0,0 +1,2 @@ +"perception","reasoning","OCR","artwork","celebrity","code_reasoning","color","commonsense_reasoning","count","existence","landmark","numerical_calculation","position","posters","scene","text_translation" +"1268.2366946778711","265.0","95.0","110.5","69.11764705882354","57.5","185.0","115.0","153.33333333333334","190.0","56.0","47.5","128.33333333333334","130.95238095238096","150.0","45.0" diff --git a/llava_base_v2/llava_base_v2_MMVet.xlsx b/llava_base_v2/llava_base_v2_MMVet.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..1949b65aad6e258abf72b500bc2987f42e0bba16 Binary files /dev/null and b/llava_base_v2/llava_base_v2_MMVet.xlsx differ diff --git a/llava_base_v2/llava_base_v2_MathVista_MINI.xlsx b/llava_base_v2/llava_base_v2_MathVista_MINI.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..0ec4530b548f45f7ded369f86d370b65f882d6d7 Binary files /dev/null and b/llava_base_v2/llava_base_v2_MathVista_MINI.xlsx differ diff --git a/llava_base_v2/llava_base_v2_POPE.xlsx b/llava_base_v2/llava_base_v2_POPE.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..40f842ff525fd4c5b0f72964f4926bcb80dc17f5 Binary files /dev/null and b/llava_base_v2/llava_base_v2_POPE.xlsx differ diff --git a/llava_base_v2/llava_base_v2_POPE_auxmatch.xlsx b/llava_base_v2/llava_base_v2_POPE_auxmatch.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..65b848be637bfc2e036d39cf228010e28db792ef Binary files /dev/null and b/llava_base_v2/llava_base_v2_POPE_auxmatch.xlsx differ diff --git a/llava_base_v2/llava_base_v2_POPE_score.csv b/llava_base_v2/llava_base_v2_POPE_score.csv new file mode 100644 index 0000000000000000000000000000000000000000..57d1d8ca760e83d1330c79c0181e097c0f744b94 --- /dev/null +++ b/llava_base_v2/llava_base_v2_POPE_score.csv @@ -0,0 +1,5 @@ +"split","Overall","acc","precision","recall" +"Overall","82.78221208665907","84.83333333333334","96.28647214854111","72.6" +"random","83.6405529953917","85.66666666666667","98.6413043478261","72.6" +"adversarial","81.8489289740699","83.89999999999999","93.7984496124031","72.6" +"popular","82.87671232876713","84.93333333333334","96.54255319148936","72.6" diff --git a/llava_base_v2/llava_base_v2_SEEDBench_IMG.xlsx b/llava_base_v2/llava_base_v2_SEEDBench_IMG.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..9649c2c9697c07b56b28a98927fc69a7dbc83e5f Binary files /dev/null and b/llava_base_v2/llava_base_v2_SEEDBench_IMG.xlsx differ diff --git a/llava_base_v2/llava_base_v2_SEEDBench_IMG_acc.csv b/llava_base_v2/llava_base_v2_SEEDBench_IMG_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..2466311c90fd57f64e92126b555e4b5ea2aa655e --- /dev/null +++ b/llava_base_v2/llava_base_v2_SEEDBench_IMG_acc.csv @@ -0,0 +1,2 @@ +"split","Overall","Instance Attributes","Instance Identity","Instance Interaction","Instance Location","Instances Counting","Scene Understanding","Spatial Relation","Text Understanding","Visual Reasoning" +"none","0.604061270376616","0.6530436653043665","0.6810486073184052","0.7216494845360825","0.6278118609406953","0.2709440130772374","0.7412919569347688","0.5190258751902588","0.2976190476190476","0.7854984894259819" diff --git a/llava_base_v2/llava_base_v2_SEEDBench_IMG_openai_result.pkl b/llava_base_v2/llava_base_v2_SEEDBench_IMG_openai_result.pkl new file mode 100644 index 0000000000000000000000000000000000000000..f7821e10a859bd5cfe646a6ab09174b0f0afe815 --- /dev/null +++ b/llava_base_v2/llava_base_v2_SEEDBench_IMG_openai_result.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a2c0daae5b0338b94adaa9002dd43990b7911427a4e3789d34dbb48c9e567210 +size 895520 diff --git a/llava_base_v2/llava_base_v2_SEEDBench_IMG_openai_result.xlsx b/llava_base_v2/llava_base_v2_SEEDBench_IMG_openai_result.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..23c504324c334b5b5f61f860cd6d1d9fcf59a457 --- /dev/null +++ b/llava_base_v2/llava_base_v2_SEEDBench_IMG_openai_result.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8a1fc04a9084a0a9286ff0db0b50f97609316cb52d240b64c18401463689e9cc +size 1057192 diff --git a/llava_base_v2/llava_base_v2_ScienceQA_VAL.xlsx b/llava_base_v2/llava_base_v2_ScienceQA_VAL.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..d6ead0c5c8b92ef45fc26321c7dd47da46dd4185 Binary files /dev/null and b/llava_base_v2/llava_base_v2_ScienceQA_VAL.xlsx differ diff --git a/llava_base_v2/llava_base_v2_ScienceQA_VAL_acc.csv b/llava_base_v2/llava_base_v2_ScienceQA_VAL_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..dc4d46dbc159a80025eb6c1ae60bf4543957d944 --- /dev/null +++ b/llava_base_v2/llava_base_v2_ScienceQA_VAL_acc.csv @@ -0,0 +1,2 @@ +"split","Overall","Adaptations","Adaptations and natural selection","Age of Exploration","Ancient Egypt and Kush","Ancient Mesopotamia","Animals","Astronomy","Atoms and molecules","Basic economic principles","Chemical reactions","Cities","Classification","Classification and scientific names","Climate change","Colonial America","Context clues","Descriptive details","Designing experiments","Domain-specific vocabulary","Early 19th century American history","Early Americas","Earth events","Ecological interactions","Ecosystems","Engineering practices","English colonies in North America","Force and motion","Fossils","Genes to traits","Geography","Government","Independent reading comprehension","Informational texts: level 1","Magnets","Maps","Materials","Medieval Asia","Natural resources and human impacts","Oceania: geography","Oceans and continents","Oceans and continents ","Particle motion and energy","Persuasive strategies","Physical Geography","Plant reproduction","Plants","Plate tectonics","Read-alone texts","Rocks and minerals","Rome and the Byzantine Empire","Scientific names","Solutions","State capitals","States","States of matter","The American Revolution","The Americas: geography","The Antebellum period","The Civil War and Reconstruction","The Silk Road","Thermal energy","Velocity, acceleration, and forces","Visual elements","Water cycle","Weather and climate","World religions" +"val","0.6547448736289938","0.84375","1.0","1.0","1.0","0.5","1.0","0.5","0.6904761904761905","0.59375","1.0","0.375","0.7619047619047619","0.9361702127659575","1.0","0.5238095238095238","0.8","1.0","1.0","0.6","0.0","0.5","1.0","0.6","0.7164179104477612","1.0","0.4358974358974359","0.68","0.47058823529411764","0.0851063829787234","0.4647887323943662","1.0","1.0","1.0","0.2781954887218045","0.5483870967741935","0.8017241379310345","0.5","0.5","0.2765957446808511","0.8","1.0","0.25","0.5555555555555556","0.5348837209302325","0.7","0.8","1.0","1.0","1.0","0.5","0.8571428571428571","0.417910447761194","0.9905660377358491","0.3333333333333333","0.7894736842105263","0.6666666666666666","0.3","1.0","0.0","1.0","0.6666666666666666","0.13725490196078433","0.7777777777777778","1.0","0.359375","1.0" diff --git a/llava_base_v2/llava_base_v2_ScienceQA_VAL_openai_result.pkl b/llava_base_v2/llava_base_v2_ScienceQA_VAL_openai_result.pkl new file mode 100644 index 0000000000000000000000000000000000000000..09e96f9fa8c554ac386323d5f09ad36a06710940 --- /dev/null +++ b/llava_base_v2/llava_base_v2_ScienceQA_VAL_openai_result.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b210adcd797510b1b99b09948c943176b28ecb065130331bae80c133e5da35f4 +size 120112 diff --git a/llava_base_v2/llava_base_v2_ScienceQA_VAL_openai_result.xlsx b/llava_base_v2/llava_base_v2_ScienceQA_VAL_openai_result.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..cd4c0bcaaa01bc7cd957ccb11c9422e3d99bef52 Binary files /dev/null and b/llava_base_v2/llava_base_v2_ScienceQA_VAL_openai_result.xlsx differ diff --git a/llava_base_v2/llava_base_v2_TextVQA_VAL.xlsx b/llava_base_v2/llava_base_v2_TextVQA_VAL.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..7dec6284e500f502ffae6f753073d0f2273b8388 Binary files /dev/null and b/llava_base_v2/llava_base_v2_TextVQA_VAL.xlsx differ diff --git a/llava_base_v2/llava_base_v2_TextVQA_VAL_acc.csv b/llava_base_v2/llava_base_v2_TextVQA_VAL_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..0024460e81d94014e356b6d45c8eecbbf31bac78 --- /dev/null +++ b/llava_base_v2/llava_base_v2_TextVQA_VAL_acc.csv @@ -0,0 +1,2 @@ +"Overall" +"16.566" diff --git a/llava_moe_e4t2/01_MMMU_DEV_VAL.pkl b/llava_moe_e4t2/01_MMMU_DEV_VAL.pkl new file mode 100644 index 0000000000000000000000000000000000000000..e345f82fd5f75256770d92638e718a6797987a46 --- /dev/null +++ b/llava_moe_e4t2/01_MMMU_DEV_VAL.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8e93beea652c9611dccccc06ea9595d3f36ce5869400c2a0678269c360dcbe11 +size 1511 diff --git a/llava_moe_e4t2/llava_moe_e4t2_MME.xlsx b/llava_moe_e4t2/llava_moe_e4t2_MME.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..745523f718cb28144af5c451f6bc92690640e91c Binary files /dev/null and b/llava_moe_e4t2/llava_moe_e4t2_MME.xlsx differ diff --git a/llava_moe_e4t2/llava_moe_e4t2_MME_auxmatch.xlsx b/llava_moe_e4t2/llava_moe_e4t2_MME_auxmatch.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..2e558ccbe97de2552c40a68982e91a6b8931b9d2 Binary files /dev/null and b/llava_moe_e4t2/llava_moe_e4t2_MME_auxmatch.xlsx differ diff --git a/llava_moe_e4t2/llava_moe_e4t2_MME_score.csv b/llava_moe_e4t2/llava_moe_e4t2_MME_score.csv new file mode 100644 index 0000000000000000000000000000000000000000..6adf3b5b27079fc56019185d18cb265c7c52d0f5 --- /dev/null +++ b/llava_moe_e4t2/llava_moe_e4t2_MME_score.csv @@ -0,0 +1,2 @@ +"perception","reasoning","OCR","artwork","celebrity","code_reasoning","color","commonsense_reasoning","count","existence","landmark","numerical_calculation","position","posters","scene","text_translation" +"1280.2602040816328","259.2857142857143","115.0","106.0","80.0","57.5","180.0","104.28571428571428","140.0","185.0","66.75","47.5","123.33333333333333","142.1768707482993","142.0","50.0" diff --git a/llava_moe_e4t2/llava_moe_e4t2_MMVet.xlsx b/llava_moe_e4t2/llava_moe_e4t2_MMVet.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..f2ffc7e4436c2db11541593772c4c63caaaf41d5 Binary files /dev/null and b/llava_moe_e4t2/llava_moe_e4t2_MMVet.xlsx differ diff --git a/llava_moe_e4t2/llava_moe_e4t2_MathVista_MINI.xlsx b/llava_moe_e4t2/llava_moe_e4t2_MathVista_MINI.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..178981d374010e4250faf63b05d679beeca0da19 Binary files /dev/null and b/llava_moe_e4t2/llava_moe_e4t2_MathVista_MINI.xlsx differ diff --git a/llava_moe_e4t2/llava_moe_e4t2_POPE.xlsx b/llava_moe_e4t2/llava_moe_e4t2_POPE.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..fa9e383b21080cd2605bb3c6c70edbc3336e8847 Binary files /dev/null and b/llava_moe_e4t2/llava_moe_e4t2_POPE.xlsx differ diff --git a/llava_moe_e4t2/llava_moe_e4t2_POPE_auxmatch.xlsx b/llava_moe_e4t2/llava_moe_e4t2_POPE_auxmatch.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..da1e48419e2afa41008e98619f4f92b905ea0ae9 Binary files /dev/null and b/llava_moe_e4t2/llava_moe_e4t2_POPE_auxmatch.xlsx differ diff --git a/llava_moe_e4t2/llava_moe_e4t2_POPE_score.csv b/llava_moe_e4t2/llava_moe_e4t2_POPE_score.csv new file mode 100644 index 0000000000000000000000000000000000000000..ad60b028ac9733f22994cf77bc05c35115002391 --- /dev/null +++ b/llava_moe_e4t2/llava_moe_e4t2_POPE_score.csv @@ -0,0 +1,5 @@ +"split","Overall","acc","precision","recall" +"Overall","77.46876259572754","81.36666666666666","97.96126401630988","64.06666666666668" +"adversarial","77.03406813627255","80.9","96.58291457286433","64.06666666666668" +"popular","77.50000000000001","81.39999999999999","98.06122448979592","64.06666666666668" +"random","77.87682333873582","81.8","99.27685950413223","64.06666666666668" diff --git a/llava_moe_e4t2/llava_moe_e4t2_SEEDBench_IMG.xlsx b/llava_moe_e4t2/llava_moe_e4t2_SEEDBench_IMG.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..75458ea2fb02aa554cb210b04882ee78a2ec86e7 Binary files /dev/null and b/llava_moe_e4t2/llava_moe_e4t2_SEEDBench_IMG.xlsx differ diff --git a/llava_moe_e4t2/llava_moe_e4t2_SEEDBench_IMG_acc.csv b/llava_moe_e4t2/llava_moe_e4t2_SEEDBench_IMG_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..84fa616c4f4bac18a32f3c73ddaa9abac3571aad --- /dev/null +++ b/llava_moe_e4t2/llava_moe_e4t2_SEEDBench_IMG_acc.csv @@ -0,0 +1,2 @@ +"split","Overall","Instance Attributes","Instance Identity","Instance Interaction","Instance Location","Instances Counting","Scene Understanding","Spatial Relation","Text Understanding","Visual Reasoning" +"none","0.6124929735806633","0.6713271671327167","0.7001638448935008","0.7731958762886598","0.6278118609406953","0.2697180220678382","0.7384420519316023","0.5220700152207002","0.44047619047619047","0.7643504531722054" diff --git a/llava_moe_e4t2/llava_moe_e4t2_SEEDBench_IMG_openai_result.pkl b/llava_moe_e4t2/llava_moe_e4t2_SEEDBench_IMG_openai_result.pkl new file mode 100644 index 0000000000000000000000000000000000000000..9cc16c2e4ed1d70c9648a7c92052b2aace268568 --- /dev/null +++ b/llava_moe_e4t2/llava_moe_e4t2_SEEDBench_IMG_openai_result.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8ef88f1b512dec61c830d292f5dcd1dd2fd02c2a1835ca8efa551626baa609d4 +size 877642 diff --git a/llava_moe_e4t2/llava_moe_e4t2_SEEDBench_IMG_openai_result.xlsx b/llava_moe_e4t2/llava_moe_e4t2_SEEDBench_IMG_openai_result.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..77036ae393a9ebbcf4f73b84398d233b0fc2a4c1 --- /dev/null +++ b/llava_moe_e4t2/llava_moe_e4t2_SEEDBench_IMG_openai_result.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9927917613e699b95f8173dad936f07c264a222b38a4e61585cd5f73236e0070 +size 1057032 diff --git a/llava_moe_e4t2/llava_moe_e4t2_ScienceQA_VAL.xlsx b/llava_moe_e4t2/llava_moe_e4t2_ScienceQA_VAL.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..f7b73f21f4d7c1c7c1fac6bd6278455e302b12e4 Binary files /dev/null and b/llava_moe_e4t2/llava_moe_e4t2_ScienceQA_VAL.xlsx differ diff --git a/llava_moe_e4t2/llava_moe_e4t2_ScienceQA_VAL_acc.csv b/llava_moe_e4t2/llava_moe_e4t2_ScienceQA_VAL_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..e44c4fe7b72482ebaf84681a682c624d768aa333 --- /dev/null +++ b/llava_moe_e4t2/llava_moe_e4t2_ScienceQA_VAL_acc.csv @@ -0,0 +1,2 @@ +"split","Overall","Adaptations","Adaptations and natural selection","Age of Exploration","Ancient Egypt and Kush","Ancient Mesopotamia","Animals","Astronomy","Atoms and molecules","Basic economic principles","Chemical reactions","Cities","Classification","Classification and scientific names","Climate change","Colonial America","Context clues","Descriptive details","Designing experiments","Domain-specific vocabulary","Early 19th century American history","Early Americas","Earth events","Ecological interactions","Ecosystems","Engineering practices","English colonies in North America","Force and motion","Fossils","Genes to traits","Geography","Government","Independent reading comprehension","Informational texts: level 1","Magnets","Maps","Materials","Medieval Asia","Natural resources and human impacts","Oceania: geography","Oceans and continents","Oceans and continents ","Particle motion and energy","Persuasive strategies","Physical Geography","Plant reproduction","Plants","Plate tectonics","Read-alone texts","Rocks and minerals","Rome and the Byzantine Empire","Scientific names","Solutions","State capitals","States","States of matter","The American Revolution","The Americas: geography","The Antebellum period","The Civil War and Reconstruction","The Silk Road","Thermal energy","Velocity, acceleration, and forces","Visual elements","Water cycle","Weather and climate","World religions" +"val","0.6728659990462565","0.84375","1.0","1.0","1.0","0.5","1.0","0.6086956521739131","0.7380952380952381","0.609375","1.0","0.5","0.7857142857142857","0.9148936170212766","1.0","0.5238095238095238","0.8","0.6666666666666666","1.0","0.4","0.0","0.5","1.0","0.64","0.7164179104477612","1.0","0.38461538461538464","0.68","0.5294117647058824","0.10638297872340426","0.49295774647887325","1.0","1.0","1.0","0.2706766917293233","0.5967741935483871","0.8362068965517241","0.5","0.5","0.3191489361702128","0.9","1.0","0.29411764705882354","0.6111111111111112","0.7441860465116279","0.6","0.8","1.0","1.0","1.0","0.5","0.8809523809523809","0.40298507462686567","0.9937106918238994","0.3333333333333333","0.8421052631578947","0.6666666666666666","0.36666666666666664","1.0","0.0","1.0","0.6666666666666666","0.11764705882352941","0.7777777777777778","1.0","0.390625","1.0" diff --git a/llava_moe_e4t2/llava_moe_e4t2_ScienceQA_VAL_openai_result.pkl b/llava_moe_e4t2/llava_moe_e4t2_ScienceQA_VAL_openai_result.pkl new file mode 100644 index 0000000000000000000000000000000000000000..6109e2e870aa0a789bff6d0daa9a80338b0ab8b8 --- /dev/null +++ b/llava_moe_e4t2/llava_moe_e4t2_ScienceQA_VAL_openai_result.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:49939a6b74685510e73f86f828193d9e2055f54199f694ea5021d7e7d307ff7c +size 117302 diff --git a/llava_moe_e4t2/llava_moe_e4t2_ScienceQA_VAL_openai_result.xlsx b/llava_moe_e4t2/llava_moe_e4t2_ScienceQA_VAL_openai_result.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..8572c3c204ad3f116ad0515bfef2e2aaf63fb793 Binary files /dev/null and b/llava_moe_e4t2/llava_moe_e4t2_ScienceQA_VAL_openai_result.xlsx differ diff --git a/llava_moe_e4t2/llava_moe_e4t2_TextVQA_VAL.xlsx b/llava_moe_e4t2/llava_moe_e4t2_TextVQA_VAL.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..a7690968b84dea20342adc78f5e6d8e04c9d1fd2 Binary files /dev/null and b/llava_moe_e4t2/llava_moe_e4t2_TextVQA_VAL.xlsx differ diff --git a/llava_moe_e4t2/llava_moe_e4t2_TextVQA_VAL_acc.csv b/llava_moe_e4t2/llava_moe_e4t2_TextVQA_VAL_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..08f8fefc4b58d5edc235426bbbf9cd31c3da4b27 --- /dev/null +++ b/llava_moe_e4t2/llava_moe_e4t2_TextVQA_VAL_acc.csv @@ -0,0 +1,2 @@ +"Overall" +"19.916" diff --git a/llava_moe_e5t3/01_MMMU_DEV_VAL.pkl b/llava_moe_e5t3/01_MMMU_DEV_VAL.pkl new file mode 100644 index 0000000000000000000000000000000000000000..5d112bb086c2752ba7b9f32a878a0191e4a04ccc --- /dev/null +++ b/llava_moe_e5t3/01_MMMU_DEV_VAL.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:73f15453d25d286d97a74068fec83ba47c4f0774ce22f1ba78d41a24830fe713 +size 1387 diff --git a/llava_moe_e5t3/llava_moe_e5t3_MME.xlsx b/llava_moe_e5t3/llava_moe_e5t3_MME.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..5ac360345806e930d20530c11f479453241a64a6 Binary files /dev/null and b/llava_moe_e5t3/llava_moe_e5t3_MME.xlsx differ diff --git a/llava_moe_e5t3/llava_moe_e5t3_MME_auxmatch.xlsx b/llava_moe_e5t3/llava_moe_e5t3_MME_auxmatch.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..ce4b982e0d7f7f60f72cda7136d69249dcca9971 Binary files /dev/null and b/llava_moe_e5t3/llava_moe_e5t3_MME_auxmatch.xlsx differ diff --git a/llava_moe_e5t3/llava_moe_e5t3_MME_score.csv b/llava_moe_e5t3/llava_moe_e5t3_MME_score.csv new file mode 100644 index 0000000000000000000000000000000000000000..035525e9361b31f3d6bac42bd1d797119959230e --- /dev/null +++ b/llava_moe_e5t3/llava_moe_e5t3_MME_score.csv @@ -0,0 +1,2 @@ +"perception","reasoning","OCR","artwork","celebrity","code_reasoning","color","commonsense_reasoning","count","existence","landmark","numerical_calculation","position","posters","scene","text_translation" +"1251.123949579832","266.42857142857144","110.0","107.25","75.58823529411765","62.5","160.0","106.42857142857143","130.0","190.0","59.75","47.5","138.33333333333331","130.95238095238096","149.25","50.0" diff --git a/llava_moe_e5t3/llava_moe_e5t3_MMVet.xlsx b/llava_moe_e5t3/llava_moe_e5t3_MMVet.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..fb00c27da3a92108fe9116e4374ff30d9e48f8dd Binary files /dev/null and b/llava_moe_e5t3/llava_moe_e5t3_MMVet.xlsx differ diff --git a/llava_moe_e5t3/llava_moe_e5t3_MathVista_MINI.xlsx b/llava_moe_e5t3/llava_moe_e5t3_MathVista_MINI.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..11d2115cfcf3f91f27deec80b629372b75515e8f Binary files /dev/null and b/llava_moe_e5t3/llava_moe_e5t3_MathVista_MINI.xlsx differ diff --git a/llava_moe_e5t3/llava_moe_e5t3_POPE.xlsx b/llava_moe_e5t3/llava_moe_e5t3_POPE.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..0eacb72282218b3abb740fa6678735409d94a861 Binary files /dev/null and b/llava_moe_e5t3/llava_moe_e5t3_POPE.xlsx differ diff --git a/llava_moe_e5t3/llava_moe_e5t3_POPE_auxmatch.xlsx b/llava_moe_e5t3/llava_moe_e5t3_POPE_auxmatch.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..8382424890fcbbfebcd85ce5b00f97a8a49867b8 Binary files /dev/null and b/llava_moe_e5t3/llava_moe_e5t3_POPE_auxmatch.xlsx differ diff --git a/llava_moe_e5t3/llava_moe_e5t3_POPE_score.csv b/llava_moe_e5t3/llava_moe_e5t3_POPE_score.csv new file mode 100644 index 0000000000000000000000000000000000000000..fc13d26526aca84fad53bc6102e0bdc2e4b7a7a0 --- /dev/null +++ b/llava_moe_e5t3/llava_moe_e5t3_POPE_score.csv @@ -0,0 +1,5 @@ +"split","Overall","acc","precision","recall" +"Overall","78.32000000000001","81.93333333333334","97.89999999999999","65.26666666666667" +"random","78.76106194690264","82.39999999999999","99.29006085192698","65.26666666666667" +"adversarial","77.76012708498808","81.33333333333333","96.16895874263261","65.26666666666667" +"popular","78.44551282051283","82.06666666666666","98.29317269076306","65.26666666666667" diff --git a/llava_moe_e5t3/llava_moe_e5t3_SEEDBench_IMG.xlsx b/llava_moe_e5t3/llava_moe_e5t3_SEEDBench_IMG.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..341b083ab001097d488f13ba23edfaf12da4553c Binary files /dev/null and b/llava_moe_e5t3/llava_moe_e5t3_SEEDBench_IMG.xlsx differ diff --git a/llava_moe_e5t3/llava_moe_e5t3_SEEDBench_IMG_acc.csv b/llava_moe_e5t3/llava_moe_e5t3_SEEDBench_IMG_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..68c4448c620bc3cdbfc6bafca54fc9b162be8cdd --- /dev/null +++ b/llava_moe_e5t3/llava_moe_e5t3_SEEDBench_IMG_acc.csv @@ -0,0 +1,2 @@ +"split","Overall","Instance Attributes","Instance Identity","Instance Interaction","Instance Location","Instances Counting","Scene Understanding","Spatial Relation","Text Understanding","Visual Reasoning" +"none","0.6068015739179314","0.6665949666594967","0.685417804478427","0.7938144329896907","0.6063394683026585","0.2676747037188394","0.7381253958201394","0.5190258751902588","0.35714285714285715","0.770392749244713" diff --git a/llava_moe_e5t3/llava_moe_e5t3_SEEDBench_IMG_openai_result.pkl b/llava_moe_e5t3/llava_moe_e5t3_SEEDBench_IMG_openai_result.pkl new file mode 100644 index 0000000000000000000000000000000000000000..03a41dc8c2c1a94b77f1291da0e90d6e6c1ef519 --- /dev/null +++ b/llava_moe_e5t3/llava_moe_e5t3_SEEDBench_IMG_openai_result.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c1bdd817049777e5883cf9046d13fd04e3e945447a693d09fa48d1bbb52ffedd +size 885684 diff --git a/llava_moe_e5t3/llava_moe_e5t3_SEEDBench_IMG_openai_result.xlsx b/llava_moe_e5t3/llava_moe_e5t3_SEEDBench_IMG_openai_result.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..17b7af501102c92c01b310ef253a913b4c8b84d4 --- /dev/null +++ b/llava_moe_e5t3/llava_moe_e5t3_SEEDBench_IMG_openai_result.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:568a01029fe9f72c3802ae9a36e01520fdc87f0089e93bc5a9e444ddc7495834 +size 1057146 diff --git a/llava_moe_e5t3/llava_moe_e5t3_ScienceQA_VAL.xlsx b/llava_moe_e5t3/llava_moe_e5t3_ScienceQA_VAL.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..e880d4b57f8b2aef531d4a092b046987c3ca6026 Binary files /dev/null and b/llava_moe_e5t3/llava_moe_e5t3_ScienceQA_VAL.xlsx differ diff --git a/llava_moe_e5t3/llava_moe_e5t3_ScienceQA_VAL_acc.csv b/llava_moe_e5t3/llava_moe_e5t3_ScienceQA_VAL_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..01eaccb6f681cd8cfa33e1810ce3ee3bbc35543b --- /dev/null +++ b/llava_moe_e5t3/llava_moe_e5t3_ScienceQA_VAL_acc.csv @@ -0,0 +1,2 @@ +"split","Overall","Adaptations","Adaptations and natural selection","Age of Exploration","Ancient Egypt and Kush","Ancient Mesopotamia","Animals","Astronomy","Atoms and molecules","Basic economic principles","Chemical reactions","Cities","Classification","Classification and scientific names","Climate change","Colonial America","Context clues","Descriptive details","Designing experiments","Domain-specific vocabulary","Early 19th century American history","Early Americas","Earth events","Ecological interactions","Ecosystems","Engineering practices","English colonies in North America","Force and motion","Fossils","Genes to traits","Geography","Government","Independent reading comprehension","Informational texts: level 1","Magnets","Maps","Materials","Medieval Asia","Natural resources and human impacts","Oceania: geography","Oceans and continents","Oceans and continents ","Particle motion and energy","Persuasive strategies","Physical Geography","Plant reproduction","Plants","Plate tectonics","Read-alone texts","Rocks and minerals","Rome and the Byzantine Empire","Scientific names","Solutions","State capitals","States","States of matter","The American Revolution","The Americas: geography","The Antebellum period","The Civil War and Reconstruction","The Silk Road","Thermal energy","Velocity, acceleration, and forces","Visual elements","Water cycle","Weather and climate","World religions" +"val","0.6633285646161182","0.8333333333333334","1.0","1.0","1.0","1.0","1.0","0.6086956521739131","0.6666666666666666","0.6875","1.0","0.375","0.7976190476190477","0.9148936170212766","1.0","0.5238095238095238","0.8","1.0","1.0","0.6","0.0","0.5","1.0","0.6","0.7164179104477612","1.0","0.41025641025641024","0.76","0.47058823529411764","0.0851063829787234","0.43661971830985913","1.0","1.0","1.0","0.24812030075187969","0.6290322580645161","0.8275862068965517","0.5","0.5","0.3404255319148936","0.8","0.5","0.25","0.5555555555555556","0.3953488372093023","0.8","0.8","1.0","1.0","1.0","1.0","0.8809523809523809","0.417910447761194","0.9937106918238994","0.3333333333333333","0.8157894736842105","0.6666666666666666","0.3333333333333333","1.0","0.0","1.0","0.6666666666666666","0.13725490196078433","0.7777777777777778","1.0","0.375","1.0" diff --git a/llava_moe_e5t3/llava_moe_e5t3_ScienceQA_VAL_openai_result.pkl b/llava_moe_e5t3/llava_moe_e5t3_ScienceQA_VAL_openai_result.pkl new file mode 100644 index 0000000000000000000000000000000000000000..b179b4b32da8da9ce132190f68e82d48da849be9 --- /dev/null +++ b/llava_moe_e5t3/llava_moe_e5t3_ScienceQA_VAL_openai_result.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c73f54cd9b4564f8ce406aa983f6026d97860a011fb544f35521ce066eec45ab +size 119728 diff --git a/llava_moe_e5t3/llava_moe_e5t3_ScienceQA_VAL_openai_result.xlsx b/llava_moe_e5t3/llava_moe_e5t3_ScienceQA_VAL_openai_result.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..041d6b97dd3890d7b7c18c273511a54ac2bf9066 Binary files /dev/null and b/llava_moe_e5t3/llava_moe_e5t3_ScienceQA_VAL_openai_result.xlsx differ diff --git a/llava_moe_e5t3/llava_moe_e5t3_TextVQA_VAL.xlsx b/llava_moe_e5t3/llava_moe_e5t3_TextVQA_VAL.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..ffaf8db61e270a16ba5d1c50f174466fc615c275 Binary files /dev/null and b/llava_moe_e5t3/llava_moe_e5t3_TextVQA_VAL.xlsx differ diff --git a/llava_moe_e5t3/llava_moe_e5t3_TextVQA_VAL_acc.csv b/llava_moe_e5t3/llava_moe_e5t3_TextVQA_VAL_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..064c513e2024b32a2a9fe6e0c7f8d9f492dfc899 --- /dev/null +++ b/llava_moe_e5t3/llava_moe_e5t3_TextVQA_VAL_acc.csv @@ -0,0 +1,2 @@ +"Overall" +"17.488" diff --git a/llava_moe_e8t2/llava_moe_e8t2_MME.xlsx b/llava_moe_e8t2/llava_moe_e8t2_MME.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..65d594cbdf0fc333c75267c997de70c2864fd1d3 Binary files /dev/null and b/llava_moe_e8t2/llava_moe_e8t2_MME.xlsx differ diff --git a/llava_moe_e8t2/llava_moe_e8t2_MME_auxmatch.xlsx b/llava_moe_e8t2/llava_moe_e8t2_MME_auxmatch.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..f627775aacbfb90d9d09cb1fd8c503d1a1b6b293 Binary files /dev/null and b/llava_moe_e8t2/llava_moe_e8t2_MME_auxmatch.xlsx differ diff --git a/llava_moe_e8t2/llava_moe_e8t2_MME_score.csv b/llava_moe_e8t2/llava_moe_e8t2_MME_score.csv new file mode 100644 index 0000000000000000000000000000000000000000..0b649a63751f9b0010518d2e03ab18fcbc52a986 --- /dev/null +++ b/llava_moe_e8t2/llava_moe_e8t2_MME_score.csv @@ -0,0 +1,2 @@ +"perception","reasoning","OCR","artwork","celebrity","code_reasoning","color","commonsense_reasoning","count","existence","landmark","numerical_calculation","position","posters","scene","text_translation" +"1287.8229291716686","243.21428571428572","132.5","113.5","90.58823529411764","55.0","175.0","95.71428571428571","125.0","185.0","62.75","45.0","125.0","136.73469387755102","141.75","47.5" diff --git a/llava_moe_e8t2/llava_moe_e8t2_MMVet.xlsx b/llava_moe_e8t2/llava_moe_e8t2_MMVet.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..d74b880f91740fa7b4cdbc47899a76f93edd5a8b Binary files /dev/null and b/llava_moe_e8t2/llava_moe_e8t2_MMVet.xlsx differ diff --git a/llava_moe_e8t2/llava_moe_e8t2_POPE.xlsx b/llava_moe_e8t2/llava_moe_e8t2_POPE.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..70ad0f4367f06a8d37e6d5c5a7dca4373a9e0c03 Binary files /dev/null and b/llava_moe_e8t2/llava_moe_e8t2_POPE.xlsx differ diff --git a/llava_moe_e8t2/llava_moe_e8t2_POPE_auxmatch.xlsx b/llava_moe_e8t2/llava_moe_e8t2_POPE_auxmatch.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..0d095c9e25dd07ca96725bb8ce0a1b7395d44a91 Binary files /dev/null and b/llava_moe_e8t2/llava_moe_e8t2_POPE_auxmatch.xlsx differ diff --git a/llava_moe_e8t2/llava_moe_e8t2_POPE_score.csv b/llava_moe_e8t2/llava_moe_e8t2_POPE_score.csv new file mode 100644 index 0000000000000000000000000000000000000000..514db4422e7674210e07083e1fb9066ad0218bdb --- /dev/null +++ b/llava_moe_e8t2/llava_moe_e8t2_POPE_score.csv @@ -0,0 +1,5 @@ +"split","Overall","acc","precision","recall" +"Overall","78.00963081861958","81.73333333333333","97.98387096774194","64.8" +"popular","78.04094741067844","81.76666666666667","98.0827447023209","64.8" +"random","78.48203471941866","82.23333333333333","99.48822927328557","64.8" +"adversarial","77.51196172248804","81.2","96.42857142857143","64.8" diff --git a/llava_moe_e8t2/llava_moe_e8t2_SEEDBench_IMG.xlsx b/llava_moe_e8t2/llava_moe_e8t2_SEEDBench_IMG.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..afbc0ac2fd9ed525b73675812b333db33f2943e1 Binary files /dev/null and b/llava_moe_e8t2/llava_moe_e8t2_SEEDBench_IMG.xlsx differ diff --git a/llava_moe_e8t2/llava_moe_e8t2_SEEDBench_IMG_openai_result.pkl b/llava_moe_e8t2/llava_moe_e8t2_SEEDBench_IMG_openai_result.pkl new file mode 100644 index 0000000000000000000000000000000000000000..c2a743af9a8ae78ce50b2e8f6adfbd2dc265a981 Binary files /dev/null and b/llava_moe_e8t2/llava_moe_e8t2_SEEDBench_IMG_openai_result.pkl differ diff --git a/llava_moe_e8t2/llava_moe_e8t2_ScienceQA_VAL.xlsx b/llava_moe_e8t2/llava_moe_e8t2_ScienceQA_VAL.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..d3336d3f579854ba1aa8bf3ffb7b79ad792152fc Binary files /dev/null and b/llava_moe_e8t2/llava_moe_e8t2_ScienceQA_VAL.xlsx differ diff --git a/llava_moe_e8t2/llava_moe_e8t2_ScienceQA_VAL_acc.csv b/llava_moe_e8t2/llava_moe_e8t2_ScienceQA_VAL_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..fa67a3dfcbe68759b13ebf5c0baeb198be056fe9 --- /dev/null +++ b/llava_moe_e8t2/llava_moe_e8t2_ScienceQA_VAL_acc.csv @@ -0,0 +1,2 @@ +"split","Overall","Adaptations","Adaptations and natural selection","Age of Exploration","Ancient Egypt and Kush","Ancient Mesopotamia","Animals","Astronomy","Atoms and molecules","Basic economic principles","Chemical reactions","Cities","Classification","Classification and scientific names","Climate change","Colonial America","Context clues","Descriptive details","Designing experiments","Domain-specific vocabulary","Early 19th century American history","Early Americas","Earth events","Ecological interactions","Ecosystems","Engineering practices","English colonies in North America","Force and motion","Fossils","Genes to traits","Geography","Government","Independent reading comprehension","Informational texts: level 1","Magnets","Maps","Materials","Medieval Asia","Natural resources and human impacts","Oceania: geography","Oceans and continents","Oceans and continents ","Particle motion and energy","Persuasive strategies","Physical Geography","Plant reproduction","Plants","Plate tectonics","Read-alone texts","Rocks and minerals","Rome and the Byzantine Empire","Scientific names","Solutions","State capitals","States","States of matter","The American Revolution","The Americas: geography","The Antebellum period","The Civil War and Reconstruction","The Silk Road","Thermal energy","Velocity, acceleration, and forces","Visual elements","Water cycle","Weather and climate","World religions" +"val","0.6671435383881735","0.8854166666666666","1.0","1.0","1.0","0.5","1.0","0.5434782608695652","0.6666666666666666","0.59375","1.0","0.375","0.7738095238095238","0.9148936170212766","1.0","0.40476190476190477","0.8","1.0","1.0","0.6","0.0","0.5","1.0","0.68","0.7611940298507462","1.0","0.358974358974359","0.76","0.4117647058823529","0.0851063829787234","0.49295774647887325","1.0","1.0","1.0","0.2706766917293233","0.6612903225806451","0.8017241379310345","0.5","0.5","0.3404255319148936","0.9","1.0","0.20588235294117646","0.5555555555555556","0.6511627906976745","0.6","0.8","1.0","1.0","0.9512195121951219","1.0","0.8809523809523809","0.5223880597014925","0.9842767295597484","0.3333333333333333","0.8157894736842105","0.6666666666666666","0.43333333333333335","1.0","0.0","1.0","0.6666666666666666","0.11764705882352941","0.7777777777777778","1.0","0.375","1.0" diff --git a/llava_moe_e8t2/llava_moe_e8t2_ScienceQA_VAL_openai_result.pkl b/llava_moe_e8t2/llava_moe_e8t2_ScienceQA_VAL_openai_result.pkl new file mode 100644 index 0000000000000000000000000000000000000000..150af0d55b9151f469bf5cf182ebff24a52feb18 --- /dev/null +++ b/llava_moe_e8t2/llava_moe_e8t2_ScienceQA_VAL_openai_result.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7d31521e88212d326f5be802055313cfb8d89bbe4c75f0a96007de124887d5fd +size 117437 diff --git a/llava_moe_e8t2/llava_moe_e8t2_ScienceQA_VAL_openai_result.xlsx b/llava_moe_e8t2/llava_moe_e8t2_ScienceQA_VAL_openai_result.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..f84f9d30c43fb6f42cc844335301a1ded7068c2c Binary files /dev/null and b/llava_moe_e8t2/llava_moe_e8t2_ScienceQA_VAL_openai_result.xlsx differ diff --git a/llava_moe_e8t2/llava_moe_e8t2_TextVQA_VAL.xlsx b/llava_moe_e8t2/llava_moe_e8t2_TextVQA_VAL.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..fc125fb56fe65fe53f9d08be85f4ebdf9b071000 Binary files /dev/null and b/llava_moe_e8t2/llava_moe_e8t2_TextVQA_VAL.xlsx differ diff --git a/llava_moe_e8t2/llava_moe_e8t2_TextVQA_VAL_acc.csv b/llava_moe_e8t2/llava_moe_e8t2_TextVQA_VAL_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..f9adafcc41e7b6e13a079835c72861af292ceca3 --- /dev/null +++ b/llava_moe_e8t2/llava_moe_e8t2_TextVQA_VAL_acc.csv @@ -0,0 +1,2 @@ +"Overall" +"19.071999999999996"