LilMoo-v0.2 / evals_for_reference.csv
shiza-fatimah's picture
Inital commit
7d82049
Model,arc_hi_acc,arc_hi_acc_norm,hellaswag_hi_acc,hellaswag_hi_acc_norm,mmlu_hi_acc,mmlu_hi_acc_norm,truthfulqa_hi_mc1,truthfulqa_hi_mc2,copa_hi_acc,copa_hi_acc_norm,iitp_mr_hi_acc,iitp_mr_hi_acc_norm,indicxnli_acc,indicxnli_acc_norm,milu_hi_acc,milu_hi_acc_norm,csqa_hi_acc,csqa_hi_acc_norm,global_piqa_acc,global_piqa_acc_norm
Llama-3.2-1B-Instruct,0.2260273972602739,0.2611301369863014,0.2911002548853016,0.3273152081563296,0.2823832113853823,0.2944439977486532,0.2755498059508409,0.4671780924298416,0.512249443207127,0.4988864142538975,0.535483870967742,0.535483870967742,0.3983935742971887,0.3907630522088353,0.2848088463353786,0.298428966354258,0.3638303927148548,0.3287990893568582,,
Qwen3-0.6B,0.2063356164383561,0.2602739726027397,0.2689039932030586,0.2890824129141886,0.2905041408699847,0.2965345340516201,0.2561448900388098,0.4567232616277077,0.5367483296213809,0.5033407572383074,0.4419354838709677,0.4419354838709677,0.3702811244979919,0.3642570281124498,0.2751668801833996,0.2886521475288247,0.315509391007399,0.2995731360273193,0.46,0.54
HindiLLM-medium,0.1900684931506849,0.2226027397260274,0.2672047578589634,0.2804800339847069,0.2378387070837018,0.2628447374768834,0.2509702457956015,0.4469643888228802,0.534521158129176,0.5367483296213809,0.3935483870967742,0.3935483870967742,0.3433734939759036,0.3389558232931727,0.2548715528285348,0.2704470366125008,0.3596186681844052,0.322965281730222,0.48,0.6
Qwen2.5-3B,0.2311643835616438,0.2662671232876712,0.2907816482582838,0.3276338147833475,0.3610195384739085,0.3572404920800836,0.2652005174644243,0.4380496715887901,0.5233853006681515,0.4988864142538975,0.5516129032258065,0.5516129032258065,0.4096385542168674,0.3995983935742971,0.3645742026835682,0.3658553030813836,0.3543540125213432,0.3223676721684689,,
gemma-3-1b-it,0.2106164383561644,0.2465753424657534,0.2892948173322005,0.3177570093457944,0.2830264533247568,0.2922730562032644,0.2897800776196636,0.4721026764339953,0.532293986636971,0.5278396436525612,0.5064516129032258,0.5064516129032258,0.3967871485943775,0.395582329317269,0.2934394174364507,0.3004517564560717,0.3598178713716562,0.3206886738759248,,
Qwen3-1.7B,0.2559931506849315,0.285958904109589,0.284197111299915,0.3181818181818182,0.3558736029589129,0.360778322746643,0.2690815006468305,0.4728831115100759,0.5679287305122495,0.5389755011135857,0.5161290322580645,0.5161290322580645,0.4012048192771084,0.3919678714859437,0.3679455195199245,0.369833456948284,0.3369664200341491,0.3178998292544109,,
Qwen2.5-1.5B,0.2226027397260274,0.2363013698630137,0.2803738317757009,0.3026762956669498,0.3132588244753558,0.3159926027176972,0.276843467011643,0.4529494179609675,0.5256124721603563,0.5077951002227171,0.3612903225806451,0.3612903225806451,0.3799196787148594,0.3791164658634538,0.3039579259658823,0.310296001618232,0.3336368810472396,0.3065167899829254,,
Llama-3.2-3B-Instruct,0.2731164383561644,0.3073630136986301,0.330607476635514,0.3948598130841121,0.3315912197475275,0.332636487899011,0.2664941785252264,0.446896635332512,0.5278396436525612,0.512249443207127,0.5806451612903226,0.5806451612903226,0.4124497991967871,0.4128514056224899,0.3605960488166678,0.3645742026835682,0.3844052361980649,0.3511098463289698,,
Qwen-2.5-1.5B-Instruct,0.2268835616438356,0.2422945205479452,0.280267629566695,0.3039507221750212,0.311087882929967,0.3208973225054273,0.2742561448900388,0.4536794735165977,0.5501113585746102,0.5278396436525612,0.3967741935483871,0.3967741935483871,0.385140562248996,0.3811244979919678,0.3040253523026094,0.3086103432000539,0.3342629482071713,0.3067444507683551,,
Airavata,0.2243150684931507,0.2645547945205479,0.2948173322005097,0.3322005097706032,0.272171745597813,0.2854386105974109,0.2626131953428202,0.4500189832322431,0.5456570155902004,0.5256124721603563,0.3516129032258064,0.3516129032258064,0.4281124497991967,0.3971887550200803,0.2811003978153867,0.2912817746611826,0.3560045532157086,0.3366818440523619,0.58,0.62
Qwen2.5-3B-Instruct,0.247431506849315,0.2799657534246575,0.2933305012744265,0.3292268479184367,0.3762965345340516,0.3688992522312455,0.2613195342820181,0.4348808885322328,0.5300668151447662,0.5011135857461024,0.5612903225806452,0.5612903225806452,0.3887550200803213,0.3847389558232931,0.3796777021104443,0.3800822601308071,0.357057484348321,0.3261525327262379,,
Llama-3.2-3B,0.2636986301369863,0.3056506849315068,0.3280586236193712,0.4034621920135939,0.3190480019297258,0.324354747929565,0.2677878395860285,0.4396524010886695,0.532293986636971,0.5011135857461024,0.5483870967741935,0.5483870967741935,0.3859437751004016,0.3771084337349397,0.3562807632661317,0.3617422965410289,0.3955036994877632,0.3668184405236198,0.54,0.64
muril-large-cased,0.2208904109589041,0.2525684931506849,0.2512744265080713,0.2582837723024639,0.2241698158719948,0.2434670740532282,0.2276843467011643,0.5001468455946757,0.4766146993318486,0.4810690423162583,0.3838709677419354,0.3838709677419354,0.3526104417670683,0.3317269076305221,0.2301260872496797,0.241588564493291,0.2903244166192373,0.2646272054638588,,
HindiLLM-small,0.1960616438356164,0.2320205479452055,0.2507434154630416,0.2584961767204758,0.2285921042051941,0.2593873120527458,0.2871927554980595,0.4846568348323389,0.512249443207127,0.5300668151447662,0.3161290322580645,0.3161290322580645,0.3253012048192771,0.3385542168674699,0.2519722203492684,0.2643786663070595,0.2999715424018213,0.2796528173022197,,
Qwen3-0.6B-Base,0.2277397260273972,0.2748287671232877,0.2751699235344095,0.2982158028887001,0.3098818042936399,0.3105250462330144,0.2794307891332471,0.4709799653282344,0.5389755011135857,0.534521158129176,0.3612903225806451,0.3612903225806451,0.3859437751004016,0.3582329317269076,0.3010585934866159,0.3093520329040523,0.3347467273762095,0.3096471257825839,0.49,0.53
Qwen3-1.7B-Base,0.2722602739726027,0.3013698630136986,0.2980033984706882,0.3411214953271028,0.3853019216852938,0.3827289539277961,0.278137128072445,0.4516788347373374,0.5523385300668151,0.5233853006681515,0.603225806451613,0.603225806451613,0.4228915662650602,0.4273092369477911,0.3922190007416897,0.3897916526195132,0.3634604439385316,0.3343767785998862,,
muril-base-cased,0.2208904109589041,0.2688356164383562,0.2527612574341546,0.241928632115548,0.2313258824475355,0.2408941062957305,0.2328589909443725,0.4986325168680121,0.5033407572383074,0.4766146993318486,0.332258064516129,0.332258064516129,0.3413654618473896,0.3313253012048193,0.2289124131885914,0.246375834400917,0.3449630051223676,0.3018497438816164,,
Qwen-2.5-0.5B-Instruct,0.2020547945205479,0.2337328767123287,0.2693288020390824,0.2875955819881053,0.2802122698399936,0.2872879311731124,0.2923673997412678,0.4801813355602639,0.5545657015590201,0.5256124721603563,0.3580645161290323,0.3580645161290323,0.3429718875502008,0.3365461847389558,0.2729418110714044,0.2831231879172005,0.3253272623790552,0.3011667615253273,,
gemma-3-1b-pt,0.2183219178082192,0.2491438356164383,0.2965165675446049,0.3259345794392523,0.2528744874165796,0.2633271689314143,0.2858990944372574,0.4718021928835578,0.4988864142538975,0.4832962138084632,0.3935483870967742,0.3935483870967742,0.3706827309236948,0.3618473895582329,0.2518373676758141,0.2625581552154271,0.3932555492316448,0.3592771770062606,0.52,0.64
Qwen-2.5-0.5B,0.2011986301369863,0.2371575342465753,0.2684791843670348,0.286321155480034,0.2720913403553912,0.2778805178097612,0.2871927554980595,0.482118582921091,0.5835189309576837,0.5367483296213809,0.4064516129032258,0.4064516129032258,0.329718875502008,0.3321285140562249,0.272132695030679,0.2848088463353786,0.3226237905520774,0.3001138303927148,0.5,0.57
Llama-3.2-1B,0.2337328767123287,0.2654109589041096,0.2920560747663551,0.3299702633814783,0.2615582535981346,0.2796494331430409,0.2690815006468305,0.4549373920832709,0.5367483296213809,0.5144766146993318,0.3806451612903225,0.3806451612903225,0.3939759036144578,0.3726907630522088,0.2731440900815858,0.2910794956510013,0.3702902675014228,0.3365964712578258,0.48,0.55
OpenHathi-7B-Hi-v0.1-Base,0.2328767123287671,0.273972602739726,0.3093670348343245,0.3557774001699235,0.2746643081128889,0.2878507678700651,0.2496765847347994,0.4451412752203544,0.5590200445434298,0.5300668151447662,0.3838709677419354,0.3838709677419354,0.4301204819277108,0.4096385542168674,0.2952599285280831,0.307598948149147,0.3616676152532726,0.3335799658508822,,
gemma-2-2b-it,0.2217465753424657,0.261986301369863,0.2918436703483432,0.3242353440951572,0.3085149151724692,0.3148669293237919,0.2949547218628719,0.4972516625270038,0.512249443207127,0.5011135857461024,0.6161290322580645,0.6161290322580645,0.4040160642570281,0.4128514056224899,0.3588429640617625,0.3663272874384735,0.3630620375640296,0.3214570290267501,,
gemma-2-2b,0.2046232876712329,0.2491438356164383,0.2615760407816482,0.2854715378079864,0.2445927474471335,0.2597089330224331,0.2949547218628719,0.5016770009850119,0.5167037861915368,0.512249443207127,0.3967741935483871,0.3967741935483871,0.4220883534136546,0.4080321285140562,0.3298496392690985,0.3380756523498078,0.3724530449630051,0.3406659077973819,,