diff --git a/.gitattributes b/.gitattributes index e21e34e887fa0b3e7f048109efb0d1b477b6ef26..976cb1dd80b17cece7a2066ad46c1407f36b7627 100644 --- a/.gitattributes +++ b/.gitattributes @@ -84,3 +84,31 @@ llava_modality_gap/llava_modality_gap_SEEDBench_IMG_openai_result.xlsx filter=lf llava_modality_gap/llava_modality_gap_ScienceQA_VAL.xlsx filter=lfs diff=lfs merge=lfs -text llava_modality_gap/llava_modality_gap_ScienceQA_VAL_openai_result.xlsx filter=lfs diff=lfs merge=lfs -text llava_modality_gap/llava_modality_gap_TextVQA_VAL.xlsx filter=lfs diff=lfs merge=lfs -text +llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_A-OKVQA_openai_result.xlsx filter=lfs diff=lfs merge=lfs -text +llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_AI2D_TEST.xlsx filter=lfs diff=lfs merge=lfs -text +llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_AI2D_TEST_openai_result.xlsx filter=lfs diff=lfs merge=lfs -text +llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_MME.xlsx filter=lfs diff=lfs merge=lfs -text +llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_MME_auxmatch.xlsx filter=lfs diff=lfs merge=lfs -text +llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_MMMU_DEV_VAL.xlsx filter=lfs diff=lfs merge=lfs -text +llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_MMMU_DEV_VAL_openai_result.xlsx filter=lfs diff=lfs merge=lfs -text +llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_POPE.xlsx filter=lfs diff=lfs merge=lfs -text +llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_POPE_auxmatch.xlsx filter=lfs diff=lfs merge=lfs -text +llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_SEEDBench_IMG.xlsx filter=lfs diff=lfs merge=lfs -text +llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_SEEDBench_IMG_openai_result.xlsx filter=lfs diff=lfs merge=lfs -text +llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_ScienceQA_VAL.xlsx filter=lfs diff=lfs merge=lfs -text +llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_ScienceQA_VAL_openai_result.xlsx filter=lfs diff=lfs merge=lfs -text +llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_TextVQA_VAL.xlsx filter=lfs diff=lfs merge=lfs -text +llava_v1.5_7b/llava_v1.5_7b_A-OKVQA_openai_result.xlsx filter=lfs diff=lfs merge=lfs -text +llava_v1.5_7b/llava_v1.5_7b_AI2D_TEST.xlsx filter=lfs diff=lfs merge=lfs -text +llava_v1.5_7b/llava_v1.5_7b_AI2D_TEST_openai_result.xlsx filter=lfs diff=lfs merge=lfs -text +llava_v1.5_7b/llava_v1.5_7b_MME.xlsx filter=lfs diff=lfs merge=lfs -text +llava_v1.5_7b/llava_v1.5_7b_MME_auxmatch.xlsx filter=lfs diff=lfs merge=lfs -text +llava_v1.5_7b/llava_v1.5_7b_MMMU_DEV_VAL.xlsx filter=lfs diff=lfs merge=lfs -text +llava_v1.5_7b/llava_v1.5_7b_MMMU_DEV_VAL_openai_result.xlsx filter=lfs diff=lfs merge=lfs -text +llava_v1.5_7b/llava_v1.5_7b_POPE.xlsx filter=lfs diff=lfs merge=lfs -text +llava_v1.5_7b/llava_v1.5_7b_POPE_auxmatch.xlsx filter=lfs diff=lfs merge=lfs -text +llava_v1.5_7b/llava_v1.5_7b_SEEDBench_IMG.xlsx filter=lfs diff=lfs merge=lfs -text +llava_v1.5_7b/llava_v1.5_7b_SEEDBench_IMG_openai_result.xlsx filter=lfs diff=lfs merge=lfs -text +llava_v1.5_7b/llava_v1.5_7b_ScienceQA_VAL.xlsx filter=lfs diff=lfs merge=lfs -text +llava_v1.5_7b/llava_v1.5_7b_ScienceQA_VAL_openai_result.xlsx filter=lfs diff=lfs merge=lfs -text +llava_v1.5_7b/llava_v1.5_7b_TextVQA_VAL.xlsx filter=lfs diff=lfs merge=lfs -text diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_A-OKVQA.xlsx b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_A-OKVQA.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..4a5d54ab900cb098ad651a1166884735b20735ac Binary files /dev/null and b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_A-OKVQA.xlsx differ diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_A-OKVQA_PREV.pkl b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_A-OKVQA_PREV.pkl new file mode 100644 index 0000000000000000000000000000000000000000..2a35974226c244c9fcac364b99d5a6bdc2bdfa1f --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_A-OKVQA_PREV.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8df89784362bb89af63a75b1129dbcda0faff1fa05bda5f09fc57c00dc5dd586 +size 5576 diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_A-OKVQA_acc.csv b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_A-OKVQA_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..23a8477ddf2b90ab11950558558ce68f6d381179 --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_A-OKVQA_acc.csv @@ -0,0 +1,2 @@ +"split","Overall","ALL" +"val","0.7921397379912664","0.7921397379912664" diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_A-OKVQA_openai_result.pkl b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_A-OKVQA_openai_result.pkl new file mode 100644 index 0000000000000000000000000000000000000000..e3c85bda8d34395509927de3ca10d152ec0dc454 --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_A-OKVQA_openai_result.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2ebf6d8c15ae23a1311658ece08ebb3fba343bb9f38dac94ba4c502cb4204b3d +size 65068 diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_A-OKVQA_openai_result.xlsx b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_A-OKVQA_openai_result.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..4620ab72202522a00b25a48c3044f4d660769fa6 --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_A-OKVQA_openai_result.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e73c8a8f9c248c2ef574ac62ca118282f80247e4c3f93a432214dddc633efbeb +size 104029 diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_AI2D_TEST.xlsx b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_AI2D_TEST.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..f6c0c27e95858f3401510f48486fe794edc104de --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_AI2D_TEST.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:73bb6a1588b82855b20c66b2ae84f82c755849640adc74703efd8f969ec7e25b +size 243350 diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_AI2D_TEST_acc.csv b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_AI2D_TEST_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..945e1f3b497624bcc11733d6bfdccd4a777d758b --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_AI2D_TEST_acc.csv @@ -0,0 +1,2 @@ +"split","Overall","atomStructure","eclipses","faultsEarthquakes","foodChainsWebs","lifeCycles","moonPhaseEquinox","partsOfA","partsOfTheEarth","photosynthesisRespiration","rockCycle","rockStrata","solarSystem","typesOf","volcano","waterCNPCycle" +"none","0.5272020725388601","0.5","0.4838709677419355","0.5714285714285714","0.5791738382099828","0.4676258992805755","0.4368231046931408","0.5811088295687885","0.5769230769230769","0.5189873417721519","0.3880597014925373","0.5609756097560976","0.5","0.45481049562682213","0.6875","0.36363636363636365" diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_AI2D_TEST_openai_result.pkl b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_AI2D_TEST_openai_result.pkl new file mode 100644 index 0000000000000000000000000000000000000000..7461509cd0ce7d30eaa0a1511e2260b0af400ad0 --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_AI2D_TEST_openai_result.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c05eba72af00e5af0f2acf3bec6a5793e0b304bed3ec2a63064a12d9a5263a4e +size 167991 diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_AI2D_TEST_openai_result.xlsx b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_AI2D_TEST_openai_result.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..a1857308d53fbb0697392f2bbd0253914c68f6f5 --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_AI2D_TEST_openai_result.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0d79b962d5bc7753cc90aa2c7ca7280c86319a33eccdbcb12a5d175e6abc2116 +size 262396 diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_HallusionBench.xlsx b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_HallusionBench.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..26312dc51338d4e874c3f086f51c5190e02d9811 Binary files /dev/null and b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_HallusionBench.xlsx differ diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_HallusionBench_auxmatch.xlsx b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_HallusionBench_auxmatch.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..892411d1205d9425ad989bc32bb1752e17ea49ed Binary files /dev/null and b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_HallusionBench_auxmatch.xlsx differ diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_HallusionBench_score.csv b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_HallusionBench_score.csv new file mode 100644 index 0000000000000000000000000000000000000000..e618e5d0e0c0a3259a070fc28f3d72450ebb05d2 --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_HallusionBench_score.csv @@ -0,0 +1,13 @@ +"split","aAcc","fAcc","qAcc" +"Overall","36.38275499474238","17.91907514450867","11.868131868131867" +"VS","21.38888888888889","6.896551724137931","3.932584269662921" +"VD","45.5160744500846","23.47826086956522","16.967509025270758" +"VD_video","29.411764705882355","4.166666666666666","2.898550724637681" +"VS_chart","13.076923076923078","0.0","0.0" +"VD_ocr","55.0561797752809","32.55813953488372","30.23255813953488" +"VD_illusion","53.47222222222222","29.03225806451613","19.444444444444446" +"VS_table","11.607142857142858","0.0","2.3255813953488373" +"VD_math","41.66666666666667","13.88888888888889","16.666666666666664" +"VS_map","37.5","4.545454545454546","6.25" +"VS_ocr","42.592592592592595","26.923076923076923","14.814814814814813" +"VD_figure","60.0","36.58536585365854","23.076923076923077" diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_MME.xlsx b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_MME.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..054b345e3cebdb948658e6d044558e4775bfc6b0 --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_MME.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cfb1f35fc991baddd60a6a1ab70b81bcff6815e0b14b7ffc77972a1866c4c600 +size 102129 diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_MME_auxmatch.xlsx b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_MME_auxmatch.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..33e5b74cb4f62a5910020ce06ef794218e544018 --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_MME_auxmatch.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:baa5a3b445dd1d53a197477cea2404bcbc6b78e8585c871e434185c9c7aa8c1b +size 116568 diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_MME_score.csv b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_MME_score.csv new file mode 100644 index 0000000000000000000000000000000000000000..ce7d13d186f5a04cd70f4f2469d4524dc4c3b726 --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_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" +"1398.8100240096037","317.5","115.0","115.25","133.52941176470588","50.0","165.0","105.0","126.66666666666666","190.0","147.5","62.5","101.66666666666666","143.19727891156464","161.0","100.0" diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_MMMU_DEV_VAL.xlsx b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_MMMU_DEV_VAL.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..321f3cc1ed7a84467fd7ec528e81239245151185 --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_MMMU_DEV_VAL.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b7a1c9641f40aba56f949a58ab184ed34f5b10dafd44413065ca0e434734e51c +size 262384 diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_MMMU_DEV_VAL_acc.csv b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_MMMU_DEV_VAL_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..60ae0414e3ac0302906a5c503effeec2b14b9677 --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_MMMU_DEV_VAL_acc.csv @@ -0,0 +1,3 @@ +"split","Overall","Accounting","Agriculture","Architecture_and_Engineering","Art","Art_Theory","Basic_Medical_Science","Biology","Chemistry","Clinical_Medicine","Computer_Science","Design","Diagnostics_and_Laboratory_Medicine","Economics","Electronics","Energy_and_Power","Finance","Geography","History","Literature","Manage","Marketing","Materials","Math","Mechanical_Engineering","Music","Pharmacy","Physics","Psychology","Public_Health","Sociology","Art & Design","Business","Health & Medicine","Humanities & Social Science","Science","Tech & Engineering" +"validation","0.31555555555555553","0.26666666666666666","0.4","0.26666666666666666","0.6","0.5333333333333333","0.3","0.23333333333333334","0.13333333333333333","0.23333333333333334","0.16666666666666666","0.5333333333333333","0.4","0.43333333333333335","0.23333333333333334","0.36666666666666664","0.06666666666666667","0.26666666666666666","0.5","0.6666666666666666","0.13333333333333333","0.23333333333333334","0.3333333333333333","0.3","0.23333333333333334","0.13333333333333333","0.2","0.26666666666666666","0.23333333333333334","0.36666666666666664","0.43333333333333335","0.45","0.22666666666666666","0.3","0.4583333333333333","0.24","0.2857142857142857" +"dev","0.3","0.0","0.2","0.0","0.4","0.8","0.4","0.4","0.4","0.2","0.4","0.6","0.2","0.0","0.0","0.4","0.2","0.0","0.6","0.6","0.4","0.0","0.0","0.4","0.2","0.2","0.2","0.2","0.8","0.0","0.8","0.5","0.12","0.2","0.7","0.28","0.17142857142857143" diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_MMMU_DEV_VAL_openai_result.pkl b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_MMMU_DEV_VAL_openai_result.pkl new file mode 100644 index 0000000000000000000000000000000000000000..c5ebf8335c5e47297d9a9ea0684621bb4e85d893 --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_MMMU_DEV_VAL_openai_result.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:65a5d18d3746a46a0ef381e30cc60ec5b2aedcd6fa44c530bea7dca7dc430313 +size 69604 diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_MMMU_DEV_VAL_openai_result.xlsx b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_MMMU_DEV_VAL_openai_result.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..4488573eb10c9391f0580ca98f85fbaf3eb1824b --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_MMMU_DEV_VAL_openai_result.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c6b039ca7e96b2cbdbc52ff99b4e22e96c7e250d911b94318d208b2e86626c8b +size 273245 diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_POPE.xlsx b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_POPE.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..f5bd4451b11d23b106564a07a701a1f81ff33542 --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_POPE.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:690eb505027e729f9eef4d6d79980580fcbad612fdf9750e46f58e202019c0a2 +size 161580 diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_POPE_auxmatch.xlsx b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_POPE_auxmatch.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..0b39bd1f47bf82bc2b5b0f6287bdf905b0239c2d --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_POPE_auxmatch.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7b235f4aa47c083fccd204a3ef2c809c35569b19ce2569919276490089d12ab1 +size 186570 diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_POPE_score.csv b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_POPE_score.csv new file mode 100644 index 0000000000000000000000000000000000000000..cbbd914176c869d61e803b23c66f4873697e10a3 --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_POPE_score.csv @@ -0,0 +1,5 @@ +"split","Overall","acc","precision","recall" +"Overall","85.62269975068266","86.54444444444445","91.91944940096865","80.13333333333334" +"adversarial","83.38536246964966","84.03333333333333","86.91250903832248","80.13333333333334" +"popular","85.97997138769671","86.93333333333332","92.74691358024691","80.13333333333334" +"random","87.60932944606414","88.66666666666667","96.62379421221866","80.13333333333334" diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_Q-Bench1_VAL.xlsx b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_Q-Bench1_VAL.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..fb811e64e71fbeaa03d69aded874293cff04a5c1 Binary files /dev/null and b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_Q-Bench1_VAL.xlsx differ diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_Q-Bench1_VAL_acc.csv b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_Q-Bench1_VAL_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..7ddd86f866da5c3ad21d4ce2374c9a2ba7c3cd54 --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_Q-Bench1_VAL_acc.csv @@ -0,0 +1,2 @@ +"split","Overall","type_0_concern_0","type_0_concern_1","type_0_concern_2","type_0_concern_3","type_1_concern_0","type_1_concern_1","type_1_concern_2","type_1_concern_3","type_2_concern_0","type_2_concern_1","type_2_concern_2","type_2_concern_3" +"val","0.5939799331103679","0.5425531914893617","0.7085714285714285","0.6410256410256411","0.7857142857142857","0.4","0.7321428571428571","0.56","0.7333333333333333","0.4715909090909091","0.6275862068965518","0.47126436781609193","0.6235294117647059" diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_Q-Bench1_VAL_openai_result.pkl b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_Q-Bench1_VAL_openai_result.pkl new file mode 100644 index 0000000000000000000000000000000000000000..073b06cebbbb1c5e0de2414a244da23f824d0bab --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_Q-Bench1_VAL_openai_result.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:78afa9af7afb2e64377b2c7c7cd73509946d052a6d1e0edde262ed523036ae4e +size 80117 diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_Q-Bench1_VAL_openai_result.xlsx b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_Q-Bench1_VAL_openai_result.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..a71ff44edc818edbbdc31ac84b5700f122030ace Binary files /dev/null and b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_Q-Bench1_VAL_openai_result.xlsx differ diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_RealWorldQA.xlsx b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_RealWorldQA.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..8656a948a169264b0106baf399bc663191abeb14 Binary files /dev/null and b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_RealWorldQA.xlsx differ diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_RealWorldQA_acc.csv b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_RealWorldQA_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..1c402dc980b8c2128c138b7768c1b0895505e8a7 --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_RealWorldQA_acc.csv @@ -0,0 +1,2 @@ +"split","Overall" +"none","0.48104575163398694" diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_RealWorldQA_openai_result.pkl b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_RealWorldQA_openai_result.pkl new file mode 100644 index 0000000000000000000000000000000000000000..692857c6c1ce312eec4a313368971fc36287a7c4 --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_RealWorldQA_openai_result.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:681e0c2c54efb830a79e15f80b596a1c82f96818f559a35d5cda9e7084f08c7e +size 48781 diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_RealWorldQA_openai_result.xlsx b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_RealWorldQA_openai_result.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..ebc876e0025901404e1e920994fa44563ad2ba94 Binary files /dev/null and b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_RealWorldQA_openai_result.xlsx differ diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_SEEDBench_IMG.xlsx b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_SEEDBench_IMG.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..8f261a2278e0ce9296ecf4985f05cc8174c488f8 --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_SEEDBench_IMG.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fc62c92523a8140834a13d47616a5ed21a7f2a43af08f0d808414ef6ed1697d7 +size 977442 diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_SEEDBench_IMG_acc.csv b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_SEEDBench_IMG_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..2d08486b9f6f00b77812b8894a37d18ffdd2ea2d --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_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","0.6058881394041596","0.6805764680576468","0.6663025669033316","0.7319587628865979","0.5899795501022495","0.2738046587658357","0.727042431918936","0.5007610350076104","0.5476190476190477","0.7552870090634441" diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_SEEDBench_IMG_openai_result.pkl b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_SEEDBench_IMG_openai_result.pkl new file mode 100644 index 0000000000000000000000000000000000000000..fea1d87a68ce7d6cd057e2612d0f7e8fa6bbe3cd --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_SEEDBench_IMG_openai_result.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ec546a53ea4d18bc85dcdceb91a59c860685a788e6a0d13749e167a686c967ba +size 881070 diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_SEEDBench_IMG_openai_result.xlsx b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_SEEDBench_IMG_openai_result.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..425206a40e7db7ad90209574359491423d4a0d44 --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_SEEDBench_IMG_openai_result.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a88e41ca1b7d55db1b4f0e68d39ba9b63aa4556ad14304501b366d1611c38ea9 +size 1063839 diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_ScienceQA_VAL.xlsx b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_ScienceQA_VAL.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..633c55da916f8d58de84705eb04fcaebcc614cbc --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_ScienceQA_VAL.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6d0d3d5fc9fdd09ee7627a1d2f4f577b595f8a33b6c6f833b2af24c1f608902e +size 343587 diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_ScienceQA_VAL_acc.csv b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_ScienceQA_VAL_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..1d597a2b457b75d389f9bd39056d7c9323994b49 --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_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","0.7083333333333334","1.0","1.0","0.0","1.0","1.0","0.45652173913043476","0.6428571428571429","0.421875","1.0","0.625","0.8095238095238095","0.9574468085106383","1.0","0.6428571428571429","0.6","1.0","0.967741935483871","0.6","0.0","1.0","0.75","0.52","0.8507462686567164","0.9245283018867925","0.4358974358974359","0.52","0.5294117647058824","0.14893617021276595","0.4084507042253521","1.0","1.0","0.8","0.38345864661654133","0.5","0.7413793103448276","0.5","0.5","0.2765957446808511","1.0","1.0","0.2647058823529412","0.6111111111111112","0.7441860465116279","0.5","1.0","1.0","1.0","0.8780487804878049","0.5","0.9047619047619048","0.5522388059701493","0.9716981132075472","0.6666666666666666","0.7894736842105263","0.6666666666666666","0.3333333333333333","0.6666666666666666","0.0","1.0","0.5555555555555556","0.29411764705882354","0.8888888888888888","1.0","0.3125","0.6666666666666666" diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_ScienceQA_VAL_openai_result.pkl b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_ScienceQA_VAL_openai_result.pkl new file mode 100644 index 0000000000000000000000000000000000000000..bb574a9602c868313d660f4f08c3710fe959a48d --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_ScienceQA_VAL_openai_result.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0c5b6fd4715f46bf4522dbcb27ffacb230d3a6fd5f943c01b61b8c47505f6869 +size 117729 diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_ScienceQA_VAL_openai_result.xlsx b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_ScienceQA_VAL_openai_result.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..2495812fed71a7f7b91836b5df46e10c4a8435f0 --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_ScienceQA_VAL_openai_result.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c3b9200a6cc9d6858b2cfef203caaecfb7dc4ebd4adb1c7c0e90b180aeb793ba +size 360662 diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_TextVQA_VAL.xlsx b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_TextVQA_VAL.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..6503fab63adec797e25bbd3ae8b2239ffcac42e0 --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_TextVQA_VAL.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:30b101b9d49571c470df6d39216d70c5af9307ff16660d6f1502d406885c7ee1 +size 328465 diff --git a/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_TextVQA_VAL_acc.csv b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_TextVQA_VAL_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..49a69003db896245b0a68e8c0f0ca1c9ce7f69f5 --- /dev/null +++ b/llava-v1.5-7b-finetune-moe/llava-v1.5-7b-finetune-moe_TextVQA_VAL_acc.csv @@ -0,0 +1,2 @@ +"Overall" +"14.228" diff --git a/llava_v1.5_7b/llava_v1.5_7b_A-OKVQA.xlsx b/llava_v1.5_7b/llava_v1.5_7b_A-OKVQA.xlsx index 26158b99e1058e2bc3d2b744b43179e1c3019580..1ecb29628431b37a9f7d6a9a67c37c5df44442e1 100644 Binary files a/llava_v1.5_7b/llava_v1.5_7b_A-OKVQA.xlsx and b/llava_v1.5_7b/llava_v1.5_7b_A-OKVQA.xlsx differ diff --git a/llava_v1.5_7b/llava_v1.5_7b_A-OKVQA_PREV.pkl b/llava_v1.5_7b/llava_v1.5_7b_A-OKVQA_PREV.pkl new file mode 100644 index 0000000000000000000000000000000000000000..a88d5f0b35d5d95c732562067e17578c55b1983e --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_A-OKVQA_PREV.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b06833b2c146f6022e81ddfc02299d14e91adcaaddce32d0e1eae28daa8c56b9 +size 5557 diff --git a/llava_v1.5_7b/llava_v1.5_7b_A-OKVQA_acc.csv b/llava_v1.5_7b/llava_v1.5_7b_A-OKVQA_acc.csv index 621f5a850c15ef5476ef3c0282c5b3541e5190b7..73f4cdd5255b6862e53ecfe877cea72c46923d61 100644 --- a/llava_v1.5_7b/llava_v1.5_7b_A-OKVQA_acc.csv +++ b/llava_v1.5_7b/llava_v1.5_7b_A-OKVQA_acc.csv @@ -1,2 +1,2 @@ "split","Overall","ALL" -"val","0.7877729257641921","0.7877729257641921" +"val","0.7851528384279476","0.7851528384279476" diff --git a/llava_v1.5_7b/llava_v1.5_7b_A-OKVQA_openai_result.pkl b/llava_v1.5_7b/llava_v1.5_7b_A-OKVQA_openai_result.pkl index 66afb1c478f35ecad06d2eb72fff664f05be464a..204f8e2444b78ade5d4934175a57eb40d706a41f 100644 --- a/llava_v1.5_7b/llava_v1.5_7b_A-OKVQA_openai_result.pkl +++ b/llava_v1.5_7b/llava_v1.5_7b_A-OKVQA_openai_result.pkl @@ -1,3 +1,3 @@ version https://git-lfs.github.com/spec/v1 -oid sha256:f370b42d90c6d613fd8cdc362de48d226c89fae66f9d85de1524617df6f3f0e1 -size 65151 +oid sha256:c7725a3e2910fbca54a7630bbc149f44de2e04ccb166072e95d07973b71e2e54 +size 65229 diff --git a/llava_v1.5_7b/llava_v1.5_7b_A-OKVQA_openai_result.xlsx b/llava_v1.5_7b/llava_v1.5_7b_A-OKVQA_openai_result.xlsx index af9b6ad4944b53a5eabfe87d0477a9a7ff0c90f8..9422d5b682e5890385fcb1a7c9757498a126f04a 100644 Binary files a/llava_v1.5_7b/llava_v1.5_7b_A-OKVQA_openai_result.xlsx and b/llava_v1.5_7b/llava_v1.5_7b_A-OKVQA_openai_result.xlsx differ diff --git a/llava_v1.5_7b/llava_v1.5_7b_AI2D_TEST.xlsx b/llava_v1.5_7b/llava_v1.5_7b_AI2D_TEST.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..13d8a47f4d6e737e52df3d3219814ccdeefe8a3f --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_AI2D_TEST.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bc00344bc1e1c1b9d97daa6eabdbff779326739279cc243bd0c20f8cff0e3d4b +size 243211 diff --git a/llava_v1.5_7b/llava_v1.5_7b_AI2D_TEST_acc.csv b/llava_v1.5_7b/llava_v1.5_7b_AI2D_TEST_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..7f08a93e1087c97a08918fae5e923c502b0e11a4 --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_AI2D_TEST_acc.csv @@ -0,0 +1,2 @@ +"split","Overall","atomStructure","eclipses","faultsEarthquakes","foodChainsWebs","lifeCycles","moonPhaseEquinox","partsOfA","partsOfTheEarth","photosynthesisRespiration","rockCycle","rockStrata","solarSystem","typesOf","volcano","waterCNPCycle" +"none","0.5197538860103627","0.5","0.5483870967741935","0.5714285714285714","0.5473321858864028","0.4628297362110312","0.41155234657039713","0.5811088295687885","0.5769230769230769","0.5189873417721519","0.417910447761194","0.6097560975609756","0.4444444444444444","0.5043731778425656","0.625","0.4318181818181818" diff --git a/llava_v1.5_7b/llava_v1.5_7b_AI2D_TEST_openai_result.pkl b/llava_v1.5_7b/llava_v1.5_7b_AI2D_TEST_openai_result.pkl new file mode 100644 index 0000000000000000000000000000000000000000..1140d2c8a2cfda73e24ddd5080c6ca52a6b057a4 --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_AI2D_TEST_openai_result.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7615c6e1b4657ea463348da12a980be3e0c78e94f24988cc035385ea76e1cb8a +size 168654 diff --git a/llava_v1.5_7b/llava_v1.5_7b_AI2D_TEST_openai_result.xlsx b/llava_v1.5_7b/llava_v1.5_7b_AI2D_TEST_openai_result.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..16fab17314074248acc6e01ffc88c1b76572c370 --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_AI2D_TEST_openai_result.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a46fe8afd221bcf60d18cbf5a413adcb2180c759e846d30aec2e1a90de682b9f +size 262115 diff --git a/llava_v1.5_7b/llava_v1.5_7b_HallusionBench.xlsx b/llava_v1.5_7b/llava_v1.5_7b_HallusionBench.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..03dfb9fe603c58aeb3f90322d793737df0dfd3dc Binary files /dev/null and b/llava_v1.5_7b/llava_v1.5_7b_HallusionBench.xlsx differ diff --git a/llava_v1.5_7b/llava_v1.5_7b_HallusionBench_auxmatch.xlsx b/llava_v1.5_7b/llava_v1.5_7b_HallusionBench_auxmatch.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..cb97225dc7ab4c3032843188a469ba8eee11b6ca Binary files /dev/null and b/llava_v1.5_7b/llava_v1.5_7b_HallusionBench_auxmatch.xlsx differ diff --git a/llava_v1.5_7b/llava_v1.5_7b_HallusionBench_score.csv b/llava_v1.5_7b/llava_v1.5_7b_HallusionBench_score.csv new file mode 100644 index 0000000000000000000000000000000000000000..81f8dfaf3c29808ce069652a1a6498bb807413b0 --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_HallusionBench_score.csv @@ -0,0 +1,13 @@ +"split","aAcc","fAcc","qAcc" +"Overall","35.646687697160885","18.786127167630056","11.20879120879121" +"VS","21.38888888888889","8.620689655172415","7.303370786516854" +"VD","44.33164128595601","23.91304347826087","13.718411552346572" +"VD_illusion","46.52777777777778","30.64516129032258","6.944444444444445" +"VD_video","29.411764705882355","4.166666666666666","2.898550724637681" +"VS_map","37.5","9.090909090909092","9.375" +"VD_ocr","62.92134831460674","34.883720930232556","32.55813953488372" +"VS_table","8.035714285714286","0.0","0.0" +"VD_math","37.96296296296296","11.11111111111111","11.11111111111111" +"VD_figure","60.0","36.58536585365854","28.205128205128204" +"VS_ocr","38.88888888888889","26.923076923076923","14.814814814814813" +"VS_chart","17.692307692307693","2.5","7.894736842105263" diff --git a/llava_v1.5_7b/llava_v1.5_7b_MME.xlsx b/llava_v1.5_7b/llava_v1.5_7b_MME.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..cda8fcb6aad2ac2aa1b49d1a0ae97745036911dc --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_MME.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ee35bca5934bec796b893bb590615d9301138ad2e7cb5b8285b38445d231dec4 +size 102187 diff --git a/llava_v1.5_7b/llava_v1.5_7b_MME_auxmatch.xlsx b/llava_v1.5_7b/llava_v1.5_7b_MME_auxmatch.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..3c4b62fdbf68942d1523160cb4b9f5bb42d1d0c6 --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_MME_auxmatch.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3150f6a0f063351994386f8165bbb190e74969c126c2c6a9ffd360e1b3272b15 +size 116732 diff --git a/llava_v1.5_7b/llava_v1.5_7b_MME_score.csv b/llava_v1.5_7b/llava_v1.5_7b_MME_score.csv new file mode 100644 index 0000000000000000000000000000000000000000..8def5751eaa7e62d309b2c6aea01fe3282c8cfd0 --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_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" +"1348.906462585034","302.1428571428571","130.0","112.75","125.0","62.5","156.66666666666669","107.14285714285714","93.33333333333333","185.0","137.0","40.0","115.0","141.156462585034","153.0","92.5" diff --git a/llava_v1.5_7b/llava_v1.5_7b_MMMU_DEV_VAL.xlsx b/llava_v1.5_7b/llava_v1.5_7b_MMMU_DEV_VAL.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..db5a52a6bfbba7a257339418e1b54f3d18e6216d --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_MMMU_DEV_VAL.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:256404155bd70da19e4d7ed7b70bb3e86fc254ba5e708935fea17e813ff9408c +size 261141 diff --git a/llava_v1.5_7b/llava_v1.5_7b_MMMU_DEV_VAL_acc.csv b/llava_v1.5_7b/llava_v1.5_7b_MMMU_DEV_VAL_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..a31a14b01976cf6bf429f92bd2c3bf69c8f1e74a --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_MMMU_DEV_VAL_acc.csv @@ -0,0 +1,3 @@ +"split","Overall","Accounting","Agriculture","Architecture_and_Engineering","Art","Art_Theory","Basic_Medical_Science","Biology","Chemistry","Clinical_Medicine","Computer_Science","Design","Diagnostics_and_Laboratory_Medicine","Economics","Electronics","Energy_and_Power","Finance","Geography","History","Literature","Manage","Marketing","Materials","Math","Mechanical_Engineering","Music","Pharmacy","Physics","Psychology","Public_Health","Sociology","Art & Design","Business","Health & Medicine","Humanities & Social Science","Science","Tech & Engineering" +"validation","0.3233333333333333","0.23333333333333334","0.4666666666666667","0.23333333333333334","0.6333333333333333","0.6666666666666666","0.4","0.23333333333333334","0.16666666666666666","0.26666666666666666","0.2","0.5666666666666667","0.3333333333333333","0.4666666666666667","0.16666666666666666","0.36666666666666664","0.1","0.2","0.4666666666666667","0.7333333333333333","0.16666666666666666","0.16666666666666666","0.3333333333333333","0.3333333333333333","0.23333333333333334","0.16666666666666666","0.2","0.26666666666666666","0.26666666666666666","0.23333333333333334","0.43333333333333335","0.5083333333333333","0.22666666666666666","0.2866666666666667","0.475","0.24","0.2857142857142857" +"dev","0.28","0.2","0.2","0.0","0.4","0.8","0.8","0.2","0.0","0.2","0.4","0.4","0.2","0.0","0.0","0.2","0.0","0.2","0.8","0.6","0.0","0.2","0.0","0.4","0.2","0.2","0.4","0.0","0.6","0.0","0.8","0.45","0.08","0.32","0.7","0.16","0.14285714285714285" diff --git a/llava_v1.5_7b/llava_v1.5_7b_MMMU_DEV_VAL_openai_result.pkl b/llava_v1.5_7b/llava_v1.5_7b_MMMU_DEV_VAL_openai_result.pkl new file mode 100644 index 0000000000000000000000000000000000000000..dcec7e70360df1286acbfd89a54bcbaa5020a70a --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_MMMU_DEV_VAL_openai_result.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fa9f9442f3561777a4d2f84b71df5b2a150cbf811032a46fe9ad6021c3b5c505 +size 70857 diff --git a/llava_v1.5_7b/llava_v1.5_7b_MMMU_DEV_VAL_openai_result.xlsx b/llava_v1.5_7b/llava_v1.5_7b_MMMU_DEV_VAL_openai_result.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..ba511e2ccd1e4f7b24e581ea17658bd84f733a98 --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_MMMU_DEV_VAL_openai_result.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:70aa149a30305ee56541e0f97986a16e7f259c0eccf210e1f292c9bbc0aaf886 +size 271502 diff --git a/llava_v1.5_7b/llava_v1.5_7b_POPE.xlsx b/llava_v1.5_7b/llava_v1.5_7b_POPE.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..402541dcaef5abe86536d3dde047f3ca0a1ab45d --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_POPE.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cebdb3d6b956dadde493ea877224a3d389a3eb8d3414a81226b67611d44fd747 +size 161986 diff --git a/llava_v1.5_7b/llava_v1.5_7b_POPE_auxmatch.xlsx b/llava_v1.5_7b/llava_v1.5_7b_POPE_auxmatch.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..8b2ce00a1ff4d209167e002bf97601b0ffeb003d --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_POPE_auxmatch.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1dd83871599d8b0fc5d60d4fc80729ec5f8fb14aefcdde08066cf9565df9a682 +size 188196 diff --git a/llava_v1.5_7b/llava_v1.5_7b_POPE_score.csv b/llava_v1.5_7b/llava_v1.5_7b_POPE_score.csv new file mode 100644 index 0000000000000000000000000000000000000000..1c227e8d748a4f8ef2723dd09efe65250af7171d --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_POPE_score.csv @@ -0,0 +1,5 @@ +"split","Overall","acc","precision","recall" +"Overall","79.8510242085661","81.05555555555556","90.46413502109705","71.46666666666667" +"popular","80.32971150243536","81.46666666666667","91.70230966638152","71.46666666666667" +"adversarial","77.51265365148228","78.10000000000001","84.67614533965245","71.46666666666667" +"random","81.83206106870229","83.6","95.71428571428572","71.46666666666667" diff --git a/llava_v1.5_7b/llava_v1.5_7b_Q-Bench1_VAL.xlsx b/llava_v1.5_7b/llava_v1.5_7b_Q-Bench1_VAL.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..eead4685dc095a0ae68df1483333edcbc1afeef3 Binary files /dev/null and b/llava_v1.5_7b/llava_v1.5_7b_Q-Bench1_VAL.xlsx differ diff --git a/llava_v1.5_7b/llava_v1.5_7b_Q-Bench1_VAL_acc.csv b/llava_v1.5_7b/llava_v1.5_7b_Q-Bench1_VAL_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..596ac81ed1db792469dca84764374bd46068cc80 --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_Q-Bench1_VAL_acc.csv @@ -0,0 +1,2 @@ +"split","Overall","type_0_concern_0","type_0_concern_1","type_0_concern_2","type_0_concern_3","type_1_concern_0","type_1_concern_1","type_1_concern_2","type_1_concern_3","type_2_concern_0","type_2_concern_1","type_2_concern_2","type_2_concern_3" +"val","0.5826086956521739","0.601063829787234","0.7028571428571428","0.6410256410256411","0.7285714285714285","0.4066666666666667","0.6964285714285714","0.57","0.6777777777777778","0.4659090909090909","0.5379310344827586","0.47126436781609193","0.6" diff --git a/llava_v1.5_7b/llava_v1.5_7b_Q-Bench1_VAL_openai_result.pkl b/llava_v1.5_7b/llava_v1.5_7b_Q-Bench1_VAL_openai_result.pkl new file mode 100644 index 0000000000000000000000000000000000000000..4be1932a8609357bc332f38b20511be323a93dfc --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_Q-Bench1_VAL_openai_result.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5b28c0093dab1efff3ff6a2fff0edfd5b74fbf593674d303431f0ed0257cc58c +size 79142 diff --git a/llava_v1.5_7b/llava_v1.5_7b_Q-Bench1_VAL_openai_result.xlsx b/llava_v1.5_7b/llava_v1.5_7b_Q-Bench1_VAL_openai_result.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..704415a76e2803ba62770954e4d8c9b02a47199a Binary files /dev/null and b/llava_v1.5_7b/llava_v1.5_7b_Q-Bench1_VAL_openai_result.xlsx differ diff --git a/llava_v1.5_7b/llava_v1.5_7b_RealWorldQA.xlsx b/llava_v1.5_7b/llava_v1.5_7b_RealWorldQA.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..ec77e8844f0398d559cb64d6189ef6e18337f91e Binary files /dev/null and b/llava_v1.5_7b/llava_v1.5_7b_RealWorldQA.xlsx differ diff --git a/llava_v1.5_7b/llava_v1.5_7b_RealWorldQA_acc.csv b/llava_v1.5_7b/llava_v1.5_7b_RealWorldQA_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..4277b5e14f5a5e5bdb5f9d4530c0395e1da65bca --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_RealWorldQA_acc.csv @@ -0,0 +1,2 @@ +"split","Overall" +"none","0.48627450980392156" diff --git a/llava_v1.5_7b/llava_v1.5_7b_RealWorldQA_openai_result.pkl b/llava_v1.5_7b/llava_v1.5_7b_RealWorldQA_openai_result.pkl new file mode 100644 index 0000000000000000000000000000000000000000..78fae08b2bb71bf1a9ef6b75dc2bf028670722af --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_RealWorldQA_openai_result.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:14267442dd8b89504849a0cd80672755cf21501dfc0415c870988211eb1544e5 +size 47481 diff --git a/llava_v1.5_7b/llava_v1.5_7b_RealWorldQA_openai_result.xlsx b/llava_v1.5_7b/llava_v1.5_7b_RealWorldQA_openai_result.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..e656bd4d696271a1306fd9c922173e370c9ea959 Binary files /dev/null and b/llava_v1.5_7b/llava_v1.5_7b_RealWorldQA_openai_result.xlsx differ diff --git a/llava_v1.5_7b/llava_v1.5_7b_SEEDBench_IMG.xlsx b/llava_v1.5_7b/llava_v1.5_7b_SEEDBench_IMG.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..e21112a87883b718eb0e1639ef45bf1bb7553150 --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_SEEDBench_IMG.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:de3426d90aced9f861fff0bfa7824a236dcf7c99666f807d7ca3168e32c4a79f +size 976856 diff --git a/llava_v1.5_7b/llava_v1.5_7b_SEEDBench_IMG_acc.csv b/llava_v1.5_7b/llava_v1.5_7b_SEEDBench_IMG_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..033ff161a6bb4277ea57e34fd6132dc3da7ad8c2 --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_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","0.600407532321529","0.6670251667025167","0.661387220098307","0.711340206185567","0.5797546012269938","0.2738046587658357","0.7298923369221026","0.5053272450532724","0.4523809523809524","0.7613293051359517" diff --git a/llava_v1.5_7b/llava_v1.5_7b_SEEDBench_IMG_openai_result.pkl b/llava_v1.5_7b/llava_v1.5_7b_SEEDBench_IMG_openai_result.pkl new file mode 100644 index 0000000000000000000000000000000000000000..670f0136a0bb12f4bed3a9366149af791c52ec9e --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_SEEDBench_IMG_openai_result.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c00006ac5031f4eea261454ea49120094f45291f72f0c4aa0ebda92881397ee4 +size 874548 diff --git a/llava_v1.5_7b/llava_v1.5_7b_SEEDBench_IMG_openai_result.xlsx b/llava_v1.5_7b/llava_v1.5_7b_SEEDBench_IMG_openai_result.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..e7bc43a7a58ec82df678ae6e867d006a390e69ef --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_SEEDBench_IMG_openai_result.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d15c042669e953fa74e5244684ae8fb622387dacbeafa78844eea1a22ba8432f +size 1057418 diff --git a/llava_v1.5_7b/llava_v1.5_7b_ScienceQA_VAL.xlsx b/llava_v1.5_7b/llava_v1.5_7b_ScienceQA_VAL.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..3c4cda2039532d87495f2a425b346ccba11e1697 --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_ScienceQA_VAL.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fff3af4b89468cd61c5dae69394bc7502a69dd2a66196a23b34eaa552fd0e3d4 +size 343136 diff --git a/llava_v1.5_7b/llava_v1.5_7b_ScienceQA_VAL_PREV.pkl b/llava_v1.5_7b/llava_v1.5_7b_ScienceQA_VAL_PREV.pkl new file mode 100644 index 0000000000000000000000000000000000000000..7bab0e3fa10c3c5ed6c23128ce2e8647d8c88f0f --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_ScienceQA_VAL_PREV.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9e6e3a0b44b5a9212aa9fab63b317ade9bc2fa31a68abc52be13a117d09c796a +size 10460 diff --git a/llava_v1.5_7b/llava_v1.5_7b_ScienceQA_VAL_acc.csv b/llava_v1.5_7b/llava_v1.5_7b_ScienceQA_VAL_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..ffb8bdb90dbc2fd02f4148cdb2abedfa7be43ba0 --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_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","0.6552217453505007","0.8333333333333334","1.0","1.0","0.0","1.0","1.0","0.4782608695652174","0.7619047619047619","0.421875","1.0","0.5","0.8452380952380952","0.8936170212765957","1.0","0.5714285714285714","0.8","1.0","0.9193548387096774","0.4","0.0","1.0","0.75","0.56","0.8208955223880597","0.9622641509433962","0.46153846153846156","0.64","0.47058823529411764","0.19148936170212766","0.4788732394366197","1.0","1.0","0.8","0.3308270676691729","0.4032258064516129","0.6896551724137931","0.0","0.5","0.3617021276595745","1.0","1.0","0.23529411764705882","0.3888888888888889","0.7906976744186046","0.6","1.0","1.0","1.0","0.926829268292683","0.5","0.9047619047619048","0.5970149253731343","0.9779874213836478","0.6666666666666666","0.7105263157894737","0.6666666666666666","0.3","1.0","0.0","1.0","0.6666666666666666","0.17647058823529413","0.8888888888888888","1.0","0.28125","0.6666666666666666" diff --git a/llava_v1.5_7b/llava_v1.5_7b_ScienceQA_VAL_openai_result.pkl b/llava_v1.5_7b/llava_v1.5_7b_ScienceQA_VAL_openai_result.pkl new file mode 100644 index 0000000000000000000000000000000000000000..2f4f252c24c0e127fbb7f36f919d3bab92f9bb65 --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_ScienceQA_VAL_openai_result.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:37e8f2bce39d1ec02a0a4e3935f605418795a01efdaff5e4f084fe666b830027 +size 115377 diff --git a/llava_v1.5_7b/llava_v1.5_7b_ScienceQA_VAL_openai_result.xlsx b/llava_v1.5_7b/llava_v1.5_7b_ScienceQA_VAL_openai_result.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..81dac2325dfcc717a93d37d8516727577d4420d7 --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_ScienceQA_VAL_openai_result.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8c4201e3521bb6be240940a3da41e2ec6afd54883d5c3173b92f85463d891176 +size 359863 diff --git a/llava_v1.5_7b/llava_v1.5_7b_TextVQA_VAL.xlsx b/llava_v1.5_7b/llava_v1.5_7b_TextVQA_VAL.xlsx new file mode 100644 index 0000000000000000000000000000000000000000..a96e0753feeafc247949ca786fcd9ed2f1eb3efd --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_TextVQA_VAL.xlsx @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3d83b8260122ca840f7d90906004df25c47e3120e3ef496c4e310353a04edb41 +size 328440 diff --git a/llava_v1.5_7b/llava_v1.5_7b_TextVQA_VAL_PREV.pkl b/llava_v1.5_7b/llava_v1.5_7b_TextVQA_VAL_PREV.pkl new file mode 100644 index 0000000000000000000000000000000000000000..f027558fe5e4ac135a577cf9acdaf9d930b651e0 --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_TextVQA_VAL_PREV.pkl @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ce11d823a10204e15a6fd0658779985b72b68d44151e365d4c892159603186fe +size 47895 diff --git a/llava_v1.5_7b/llava_v1.5_7b_TextVQA_VAL_acc.csv b/llava_v1.5_7b/llava_v1.5_7b_TextVQA_VAL_acc.csv new file mode 100644 index 0000000000000000000000000000000000000000..e07a4bcc04f0610883f16d5baf70f5fa155fd638 --- /dev/null +++ b/llava_v1.5_7b/llava_v1.5_7b_TextVQA_VAL_acc.csv @@ -0,0 +1,2 @@ +"Overall" +"21.758000000000003"