Add files using upload-large-folder tool

.gitattributes

@@ -1574,3 +1574,9 @@ finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=3/lora_ans_no_es/0.0-0.
 finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=0/lora_ans/0.0-1.0/0.7-1.0/checkpoint/lit_model.pth.lora filter=lfs diff=lfs merge=lfs -text
 finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=0/lora_ans/0.0-0.7/0.7-1.0/checkpoint/lit_model.pth.lora filter=lfs diff=lfs merge=lfs -text
 finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=0/lora_ans_no_es/0.0-1.0/0.7-1.0/checkpoint/lit_model.pth.lora filter=lfs diff=lfs merge=lfs -text
+finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=0/lora_ans_no_es/0.0-0.7/0.0-0.3/checkpoint/lit_model.pth.lora filter=lfs diff=lfs merge=lfs -text
+finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=1/lora_ans/0.0-1.0/0.7-1.0/checkpoint/lit_model.pth.lora filter=lfs diff=lfs merge=lfs -text
+finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=1/lora_ans/0.0-0.7/0.7-1.0/checkpoint/lit_model.pth.lora filter=lfs diff=lfs merge=lfs -text
+finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=1/lora_ans_no_es/0.0-1.0/0.7-1.0/checkpoint/lit_model.pth.lora filter=lfs diff=lfs merge=lfs -text
+finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=1/lora_ans_no_es/0.0-0.7/0.0-0.3/checkpoint/lit_model.pth.lora filter=lfs diff=lfs merge=lfs -text
+finetune_lora/qwen2.5-7b-instruct/banking77/size=16/seed=4/lora_ans/0.0-1.0/0.7-1.0/checkpoint/lit_model.pth.lora filter=lfs diff=lfs merge=lfs -text

finetune_lora/qwen2.5-7b-instruct/banking77/size=16/seed=4/lora_ans/0.0-0.7/0.7-1.0/best.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0db411d48812655d1c5f052a730aef52a6ba9991e8ed6edf40e66a586f1ca40d
+size 42951934

finetune_lora/qwen2.5-7b-instruct/banking77/size=16/seed=4/lora_ans/0.0-0.7/0.7-1.0/checkpoint/config.json

@@ -0,0 +1,27 @@
+{
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "model_type": "qwen2",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "rms_norm_eps": 1e-06,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.43.1",
+  "use_cache": true,
+  "use_sliding_window": false,
+  "vocab_size": 152064
+}

finetune_lora/qwen2.5-7b-instruct/banking77/size=16/seed=4/lora_ans/0.0-0.7/0.7-1.0/checkpoint/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "pad_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_p": 0.8,
+  "top_k": 20,
+  "transformers_version": "4.37.0"
+}

finetune_lora/qwen2.5-7b-instruct/banking77/size=16/seed=4/lora_ans/0.0-0.7/0.7-1.0/checkpoint/model_config.yaml

@@ -0,0 +1,37 @@
+attention_logit_softcapping: null
+attention_scores_scalar: null
+attn_bias: true
+bias: false
+block_size: 131072
+final_logit_softcapping: null
+gelu_approximate: none
+head_size: 128
+hf_config:
+  name: Qwen2.5-7B-Instruct
+  org: Qwen
+intermediate_size: 18944
+lm_head_bias: false
+mlp_class_name: LLaMAMLP
+n_embd: 3584
+n_expert: 0
+n_expert_per_token: 0
+n_head: 28
+n_layer: 28
+n_query_groups: 4
+name: Qwen2.5-7B-Instruct
+norm_class_name: RMSNorm
+norm_eps: 1.0e-06
+padded_vocab_size: 152064
+padding_multiple: 512
+parallel_residual: false
+post_attention_norm: false
+post_mlp_norm: false
+rope_adjustments: null
+rope_base: 1000000
+rope_condense_ratio: 1
+rotary_percentage: 1.0
+scale_embeddings: false
+shared_attention_norm: false
+sliding_window_layer_placing: null
+sliding_window_size: null
+vocab_size: 151643

finetune_lora/qwen2.5-7b-instruct/banking77/size=16/seed=4/lora_ans/0.0-0.7/0.7-1.0/last.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:14b9b2d3eb808423fd95e61dd8637650b493b981057e8e993a995fc96997ac57
+size 42951934

finetune_lora/qwen2.5-7b-instruct/banking77/size=16/seed=4/lora_ans/0.0-0.7/0.7-1.0/logs/events.out.tfevents.1734803539.gamma.3827549.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8ef1b07a95dce0083ee407aa859858c9c304ee81055ba4464a2ba52f2c194242
+size 54843

finetune_lora/qwen2.5-7b-instruct/banking77/size=16/seed=4/lora_ans/0.0-0.7/0.7-1.0/logs/hparams.yaml

@@ -0,0 +1 @@
+{}

finetune_lora/qwen2.5-7b-instruct/banking77/size=16/seed=4/lora_ans/0.0-0.7/0.7-1.0/logs/metrics.csv

@@ -0,0 +1,257 @@
+epoch,iter,iter_time,step,train/loss,val/loss
+0,1,0.41801123321056366,0,15.245295206705729,
+0,8,0.10661979764699936,1,13.468572124358147,
+0,16,0.08192970603704453,2,12.807103838239398,
+0,24,0.08319449797272682,3,11.434923648834229,
+0,32,0.08193360269069672,4,8.681313514709473,
+0,40,0.08204995095729828,5,6.688623336053664,
+0,48,0.07599947601556778,6,6.781939400566949,
+0,56,0.07942227274179459,7,4.500931331089565,
+0,64,0.07595022395253181,8,4.407195840563093,
+0,72,0.07605744525790215,9,3.5039620933861566,
+0,80,0.08175421133637428,10,3.1178311017843394,
+0,88,0.07897650450468063,11,3.023450704721304,
+0,96,0.07681397721171379,12,2.7905956506729126,
+0,104,0.0815720409154892,13,2.4336187413760593,
+1,112,0.0761033184826374,14,1.980389792578561,
+1,120,0.07973353192210197,15,1.9931339876992362,
+1,128,0.08694154769182205,16,2.008383246985349,
+,,,16,,2.6267958818855934
+1,136,0.08132371306419373,17,2.5170993294034685,
+1,144,0.07975330948829651,18,2.212132157950566,
+1,152,0.07661822438240051,19,2.207448954765613,
+1,160,0.07666566222906113,20,2.7828273262296404,
+1,168,0.08055181056261063,21,1.1784269785881043,
+1,176,0.081810861825943,22,1.8498126888275146,
+1,184,0.07554763928055763,23,1.846410722732544,
+1,192,0.08249428123235703,24,2.250374401316923,
+1,200,0.077521201223135,25,1.3879836375514667,
+1,208,0.0803658738732338,26,2.109172882572297,
+1,216,0.07579009607434273,27,2.1057448100160667,
+2,224,0.07578283175826073,28,1.10992347240448,
+2,232,0.07849422469735146,29,1.7119302831847092,
+2,240,0.08102121204137802,30,0.9734124172301519,
+2,248,0.07945103570818901,31,1.1411668062210083,
+2,256,0.0782124400138855,32,1.533616192638874,
+,,,32,,1.9445154222391419
+2,264,0.07599280774593353,33,0.5895678848028183,
+2,272,0.08003127947449684,34,1.5822145501772562,
+2,280,0.08132949844002724,35,1.5129140269756318,
+2,288,0.07663846015930176,36,1.5641627487014322,
+2,296,0.07849453017115593,37,1.2844155366931642,
+2,304,0.08224019035696983,38,1.338078805378505,
+2,312,0.07608083263039589,39,1.1991653948119192,
+2,320,0.07631854712963104,40,1.1972227019648398,
+3,328,0.08251668512821198,41,1.1251970356510532,
+3,336,0.07888041064143181,42,1.1884995436668395,
+3,344,0.07658852636814117,43,0.9476699064175288,
+3,352,0.07850195840001106,44,0.6889174205285532,
+3,360,0.07570690289139748,45,0.5112864623467127,
+3,368,0.07637650147080421,46,0.6372244626283645,
+3,376,0.07663038372993469,47,0.5155495659068778,
+3,384,0.08066761121153831,48,0.6537449024617672,
+,,,48,,1.456997456523658
+3,392,0.07540604844689369,49,0.8660666452300164,
+3,400,0.07997581735253334,50,0.3456831289564862,
+3,408,0.08019159734249115,51,0.59176118850708,
+3,416,0.07979369163513184,52,0.7916472737605755,
+3,424,0.08033589646220207,53,0.6650819067771618,
+3,432,0.0795188620686531,54,0.444424509401283,
+4,440,0.080206960439682,55,0.33103326729365756,
+4,448,0.07583343237638474,56,0.2951349520257541,
+4,456,0.07670791819691658,57,0.3400797193700617,
+4,464,0.08102433755993843,58,0.31007088974118235,
+4,472,0.07964625582098961,59,0.22732141523650198,
+4,480,0.08171048760414124,60,0.5153657476659175,
+4,488,0.07617775350809097,61,0.10149463026651315,
+4,496,0.07952700182795525,62,0.38331171444484163,
+4,504,0.07995658367872238,63,0.15525524652209774,
+4,512,0.07716748863458633,64,0.261320523917675,
+,,,64,,1.2392874679996468
+4,520,0.07842296734452248,65,0.13272464900676692,
+4,528,0.082512516528368,66,0.9336077170101581,
+4,536,0.07599904015660286,67,0.3285862146211522,
+5,544,0.0797199234366417,68,0.3883792461344489,
+5,552,0.07869511097669601,69,0.13799265958368778,
+5,560,0.08104588091373444,70,0.2141539692878723,
+5,568,0.07710668444633484,71,0.13665436466152853,
+5,576,0.07971390709280968,72,0.07650123164057732,
+5,584,0.07601956278085709,73,0.01841307829858528,
+5,592,0.07583951205015182,74,0.12733550760895013,
+5,600,0.07970942184329033,75,0.06060662712291654,
+5,608,0.07581551373004913,76,0.16055354276864692,
+5,616,0.07567897439002991,77,0.10011633023105819,
+5,624,0.07577453181147575,78,0.0373442519384508,
+5,632,0.07998546585440636,79,0.051642198355928544,
+5,640,0.07571514323353767,80,0.07324450952000916,
+,,,80,,1.220179757155941
+5,648,0.08041854947805405,81,0.03427613151049422,
+6,656,0.08035837858915329,82,0.018245021411185634,
+6,664,0.0757586807012558,83,0.026896495630757675,
+6,672,0.0766090452671051,84,0.031050517057467784,
+6,680,0.08087816834449768,85,0.08722809751634486,
+6,688,0.08124937489628792,86,0.10045788701972924,
+6,696,0.08173618838191032,87,0.04810646304901203,
+6,704,0.0767863355576992,88,0.015122977174961796,
+6,712,0.08045336231589317,89,0.008250255844202536,
+6,720,0.07572482526302338,90,0.005870503788658729,
+6,728,0.07988850027322769,91,0.0039164314833309115,
+6,736,0.07896571978926659,92,0.05440512783825398,
+6,744,0.07544690370559692,93,0.014060104723154967,
+6,752,0.07678967341780663,94,0.007574873343110085,
+7,760,0.08017173409461975,95,0.009290854320051554,
+7,768,0.08091559633612633,96,0.0037136566310297125,
+,,,96,,1.334627054505429
+7,776,0.07518190145492554,97,0.0026499477098695934,
+7,784,0.08751776441931725,98,0.015139012409445759,
+7,792,0.08122460171580315,99,0.003162139758037833,
+7,800,0.08049898967146873,100,0.0029983498638362755,
+7,808,0.07630405947566032,101,0.002129281434463337,
+7,816,0.08032820001244545,102,0.0025433075497858226,
+7,824,0.07648511230945587,103,0.002474746289502444,
+7,832,0.07677393779158592,104,0.09051269446499646,
+7,840,0.07649029046297073,105,0.0017086559632194362,
+7,848,0.07640264183282852,106,0.12156336870290996,
+7,856,0.0797690823674202,107,0.004290321783628315,
+7,864,0.08241377770900726,108,0.029583712056693103,
+8,872,0.08153984695672989,109,0.003109156208289466,
+8,880,0.07651563733816147,110,0.003980762682788607,
+8,888,0.07592421397566795,111,0.002226233565514641,
+8,896,0.07850825041532516,112,0.003909367095348576,
+,,,112,,1.3644537629380737
+8,904,0.07586736232042313,113,0.001750759679082505,
+8,912,0.07655619457364082,114,0.00316919437997664,
+8,920,0.07655161246657372,115,0.04217005980451806,
+8,928,0.08035317808389664,116,0.004546249871728597,
+8,936,0.08037881553173065,117,0.002945837671480452,
+8,944,0.07642089948058128,118,0.004391685235672272,
+8,952,0.07318853959441185,119,0.0021036160843712942,
+8,960,0.0757434032857418,120,0.0018103123606684117,
+8,968,0.08203768730163574,121,0.002336886105256091,
+9,976,0.08017420023679733,122,0.002482170578358429,
+9,984,0.08124836906790733,123,0.0017026159142504535,
+9,992,0.07860083505511284,124,0.0017314481374341995,
+9,1000,0.07872269302606583,125,0.002313698805385054,
+9,1008,0.08234187960624695,126,0.003353067442731117,
+9,1016,0.07541193068027496,127,0.0019318842887878418,
+9,1024,0.07562727108597755,128,0.0019356367298431934,
+,,,128,,1.4059044789459745
+9,1032,0.07581831887364388,129,0.001956676916840176,
+9,1040,0.08199923112988472,130,0.0017174848579560166,
+9,1048,0.0805906243622303,131,0.005637636017733409,
+9,1056,0.08028692379593849,132,0.0010265738403956806,
+9,1064,0.07822137326002121,133,0.0025562918573782288,
+9,1072,0.07715917751193047,134,0.003451792076230049,
+9,1080,0.08294691890478134,135,0.0012422075066881786,
+10,1088,0.0802244134247303,136,0.0025929521472225414,
+10,1096,0.07667588070034981,137,0.0022301371092908083,
+10,1104,0.08019435033202171,138,0.001588386707366086,
+10,1112,0.08342451229691505,139,0.0013748745805060025,
+10,1120,0.07861519977450371,140,0.0019766525505110623,
+10,1128,0.07851975038647652,141,0.001517053209245205,
+10,1136,0.07589851692318916,142,0.0014517174109421444,
+10,1144,0.07873744890093803,143,0.002652840534273278,
+10,1152,0.07638392969965935,144,0.0022880356191308238,
+,,,144,,1.428448607019112
+10,1160,0.0801188237965107,145,0.002163147094856518,
+10,1168,0.07911596074700356,146,0.00181398067252303,
+10,1176,0.07729433849453926,147,0.0010964423547395402,
+10,1184,0.07888704165816307,148,0.0014495904300323066,
+11,1192,0.07915829494595528,149,0.0031000313846844024,
+11,1200,0.07857371866703033,150,0.0023745037291923333,
+11,1208,0.07675604894757271,151,0.0024424809186408916,
+11,1216,0.08103244006633759,152,0.0018206863727148932,
+11,1224,0.07935509085655212,153,0.0008022674221799455,
+11,1232,0.08044067770242691,154,0.0014515669137271571,
+11,1240,0.08136998489499092,155,0.0016064035589806736,
+11,1248,0.07888821139931679,156,0.0022734016871682424,
+11,1256,0.08190657570958138,157,0.0007576886116741941,
+11,1264,0.08128445595502853,158,0.002296472684500496,
+11,1272,0.08002636581659317,159,0.001984355048517938,
+11,1280,0.07539411261677742,160,0.0012432260632825394,
+,,,160,,1.4431297172934323
+11,1288,0.0766761302947998,161,0.001251848966509489,
+11,1296,0.07675955817103386,162,0.0013161608949303627,
+12,1304,0.07578713074326515,163,0.0014664484406239353,
+12,1312,0.08188273757696152,164,0.001062643605862455,
+12,1320,0.08134699240326881,165,0.0008801866206340492,
+12,1328,0.0808706171810627,166,0.00235078040690672,
+12,1336,0.07948964089155197,167,0.002048665467494478,
+12,1344,0.08113948628306389,168,0.001369765927342491,
+12,1352,0.0757298655807972,169,0.0009977765439543873,
+12,1360,0.07644052430987358,170,0.0013666578258077304,
+12,1368,0.08014019578695297,171,0.0007469795714132488,
+12,1376,0.0853777565062046,172,0.0018483813347605369,
+12,1384,0.07632169127464294,173,0.0012686533842828464,
+12,1392,0.08055079728364944,174,0.0021629807953205374,
+12,1400,0.07989292219281197,175,0.001345583886307265,
+13,1408,0.07691758498549461,176,0.0019089055675712804,
+,,,176,,1.4550310555150954
+13,1416,0.07565921917557716,177,0.001476082220223957,
+13,1424,0.07553005591034889,178,0.0017441685447314133,
+13,1432,0.0762014240026474,179,0.0016184770020431485,
+13,1440,0.07696964219212532,180,0.0017657826732223232,
+13,1448,0.07974260672926903,181,0.0016460686238133349,
+13,1456,0.0758756548166275,182,0.0008050930815183424,
+13,1464,0.07945945113897324,183,0.0013644157582893967,
+13,1472,0.08039956167340279,184,0.0016187235451070592,
+13,1480,0.07823250815272331,185,0.0017591769552651655,
+13,1488,0.080325186252594,186,0.0011306559716575538,
+13,1496,0.07607422396540642,187,0.0010171088852941814,
+13,1504,0.0798703022301197,188,0.001641997423458604,
+13,1512,0.08069155737757683,189,0.0007013407267742988,
+14,1520,0.08059044554829597,190,0.001396438034134917,
+14,1528,0.07578818500041962,191,0.0013882955117151142,
+14,1536,0.07934626936912537,192,0.0008619158938729569,
+,,,192,,1.4619449249095162
+14,1544,0.08102718740701675,193,0.00198127578914864,
+14,1552,0.08570212870836258,194,0.001800365976974248,
+14,1560,0.08734514564275742,195,0.0012209131174521254,
+14,1568,0.08161161467432976,196,0.0018001355066964472,
+14,1576,0.07588807120919228,197,0.0010678037186153233,
+14,1584,0.07575773447751999,198,0.0007969227394667165,
+14,1592,0.07965559884905815,199,0.0009699558338616043,
+14,1600,0.08192871138453484,200,0.0014347147225635126,
+14,1608,0.08045091852545738,201,0.0009239678493031533,
+14,1616,0.08015654981136322,202,0.0013737400198200096,
+15,1624,0.08006387203931808,203,0.001007056453843312,
+15,1632,0.0758742168545723,204,0.0014523388253468456,
+15,1640,0.08171569555997849,205,0.0013339696411878385,
+15,1648,0.07627980411052704,206,0.001162861017898346,
+15,1656,0.07653063163161278,207,0.0019701303747503766,
+15,1664,0.07834725454449654,208,0.0010178024385822937,
+,,,208,,1.4749231715660311
+15,1672,0.07587892189621925,209,0.0013492611552854733,
+15,1680,0.07548464089632034,210,0.0022430306233997854,
+15,1688,0.08024438098073006,211,0.0014101047612105806,
+15,1696,0.08149268478155136,212,0.0008350604267824176,
+15,1704,0.07638537511229515,213,0.0010546634887130214,
+15,1712,0.07661915197968483,214,0.0009221237305609975,
+15,1720,0.08141110092401505,215,0.001081652287601693,
+15,1728,0.0815996453166008,216,0.0007339539627234141,
+16,1736,0.07660701870918274,217,0.000886707014918405,
+16,1744,0.07634572684764862,218,0.0016906476094542692,
+16,1752,0.07945265993475914,219,0.0008056441678807302,
+16,1760,0.08188224956393242,220,0.000925458868517092,
+16,1768,0.07933010905981064,221,0.001020455122880566,
+16,1776,0.07691051810979843,222,0.0012126970520642186,
+16,1784,0.07874795794487,223,0.0009672197649100174,
+16,1792,0.07989288866519928,224,0.0009207261609844863,
+,,,224,,1.4826744639940854
+16,1800,0.07566746324300766,225,0.001855694290222318,
+16,1808,0.07404480874538422,226,0.0012396624600835915,
+16,1816,0.07673775404691696,227,0.0013457523076795042,
+16,1824,0.08036114647984505,228,0.0009607873317095097,
+16,1832,0.0763234794139862,229,0.0007727841241285205,
+17,1840,0.07973895967006683,230,0.0018329469625873922,
+17,1848,0.0813264399766922,231,0.0015189680112102492,
+17,1856,0.08228092640638351,232,0.001438708872800427,
+17,1864,0.08009340614080429,233,0.0010031821568393046,
+17,1872,0.07870448380708694,234,0.0009355457171935726,
+17,1880,0.07899034395813942,235,0.0011530033266171813,
+17,1888,0.07948676124215126,236,0.0011397790456671387,
+17,1896,0.07565326616168022,237,0.0010101297294669268,
+17,1904,0.07669680193066597,238,0.0011359528510365635,
+17,1912,0.07655368000268936,239,0.0010693810663410816,
+17,1920,0.08091424778103828,240,0.0008892255383684779,
+,,,240,,1.4971246234441207

finetune_lora/qwen2.5-7b-instruct/banking77/size=16/seed=4/lora_ans/0.0-0.7/0.7-1.0/train_args.yaml

@@ -0,0 +1,10 @@
+K: null
+global_batch_size: 8
+learning_rate: 0.0001
+loss: ans
+max_steps: -1
+micro_batch_size: 1
+optimizer_name: adamw
+patience: 10
+val_check_interval: 16
+weight_decay: 0.0

finetune_lora/qwen2.5-7b-instruct/banking77/size=16/seed=4/lora_ans/0.0-1.0/0.7-1.0/best.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b559320c9c9b26a330736a85b2c00d4c39a3c711789f2eec32c4c244a77a3e33
+size 42951934

finetune_lora/qwen2.5-7b-instruct/banking77/size=16/seed=4/lora_ans/0.0-1.0/0.7-1.0/checkpoint/config.json

@@ -0,0 +1,27 @@
+{
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "model_type": "qwen2",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "rms_norm_eps": 1e-06,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.43.1",
+  "use_cache": true,
+  "use_sliding_window": false,
+  "vocab_size": 152064
+}

finetune_lora/qwen2.5-7b-instruct/banking77/size=16/seed=4/lora_ans/0.0-1.0/0.7-1.0/checkpoint/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "pad_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_p": 0.8,
+  "top_k": 20,
+  "transformers_version": "4.37.0"
+}

finetune_lora/qwen2.5-7b-instruct/banking77/size=16/seed=4/lora_ans/0.0-1.0/0.7-1.0/checkpoint/lit_model.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f2893700d468552b27b00bb16e3906c7a01f49c09f3c786c252a2b5d36d179f
+size 15231297418

finetune_lora/qwen2.5-7b-instruct/banking77/size=16/seed=4/lora_ans/0.0-1.0/0.7-1.0/checkpoint/lit_model.pth.lora

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9fe6068e4d9759fb1c20c28651ec1e8af9674895506f29da8300f2d9bcaa4a74
+size 42951614

finetune_lora/qwen2.5-7b-instruct/banking77/size=16/seed=4/lora_ans/0.0-1.0/0.7-1.0/checkpoint/model_config.yaml

@@ -0,0 +1,37 @@
+attention_logit_softcapping: null
+attention_scores_scalar: null
+attn_bias: true
+bias: false
+block_size: 131072
+final_logit_softcapping: null
+gelu_approximate: none
+head_size: 128
+hf_config:
+  name: Qwen2.5-7B-Instruct
+  org: Qwen
+intermediate_size: 18944
+lm_head_bias: false
+mlp_class_name: LLaMAMLP
+n_embd: 3584
+n_expert: 0
+n_expert_per_token: 0
+n_head: 28
+n_layer: 28
+n_query_groups: 4
+name: Qwen2.5-7B-Instruct
+norm_class_name: RMSNorm
+norm_eps: 1.0e-06
+padded_vocab_size: 152064
+padding_multiple: 512
+parallel_residual: false
+post_attention_norm: false
+post_mlp_norm: false
+rope_adjustments: null
+rope_base: 1000000
+rope_condense_ratio: 1
+rotary_percentage: 1.0
+scale_embeddings: false
+shared_attention_norm: false
+sliding_window_layer_placing: null
+sliding_window_size: null
+vocab_size: 151643

finetune_lora/qwen2.5-7b-instruct/banking77/size=16/seed=4/lora_ans/0.0-1.0/0.7-1.0/checkpoint/tokenizer_config.json

@@ -0,0 +1,207 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}\n    {%- endif %}\n    {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n    {%- else %}\n        {{- '<|im_start|>system\\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + message.content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

finetune_lora/qwen2.5-7b-instruct/banking77/size=16/seed=4/lora_ans/0.0-1.0/0.7-1.0/last.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d90321b7bb761b9130426ffb709e61f77b08d1da18f0db5b08f00e78a3455726
+size 42951934

finetune_lora/qwen2.5-7b-instruct/banking77/size=16/seed=4/lora_ans/0.0-1.0/0.7-1.0/logs/events.out.tfevents.1734805986.gamma.3831644.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a1a31d87f98c76c37145d1f1fde5a4f9a05de0b8f09345aa35e6c954af74d7c
+size 18290

finetune_lora/qwen2.5-7b-instruct/banking77/size=16/seed=4/lora_ans/0.0-1.0/0.7-1.0/logs/hparams.yaml

@@ -0,0 +1 @@
+{}

finetune_lora/qwen2.5-7b-instruct/banking77/size=16/seed=4/lora_ans/0.0-1.0/0.7-1.0/logs/metrics.csv

@@ -0,0 +1,87 @@
+epoch,iter,iter_time,step,train/loss,val/loss
+0,1,0.4123350605368614,0,12.11691665649414,
+0,8,0.10280553624033928,1,11.946681051543266,
+0,16,0.07627921178936958,2,12.145848044033709,
+0,24,0.08037809655070305,3,10.488243877887726,
+0,32,0.08190082013607025,4,10.462685223283438,
+0,40,0.08398445695638657,5,8.56721435054656,
+0,48,0.08043656125664711,6,5.304779767990112,
+0,56,0.07871370762586594,7,3.7038332121712823,
+0,64,0.0827261172235012,8,3.9963007313864574,
+0,72,0.07606184110045433,9,3.590405491420201,
+0,80,0.08197446912527084,10,2.288888775385343,
+0,88,0.08026864752173424,11,2.8127734565734865,
+0,96,0.08372917026281357,12,2.824228882789612,
+0,104,0.0800313800573349,13,2.5621902501141585,
+0,112,0.08824199438095093,14,2.6932331981330084,
+0,120,0.08051637560129166,15,2.207143089988015,
+0,128,0.07703807577490807,16,3.3674790700276693,
+,,,16,,2.4488709875419317
+0,136,0.07896118238568306,17,3.7004852845118594,
+0,144,0.07713618129491806,18,2.7902996081572313,
+0,152,0.08011943846940994,19,1.5338764108460525,
+1,160,0.08054821565747261,20,2.225591220855713,
+1,168,0.0815880298614502,21,2.170739907771349,
+1,176,0.07644334435462952,22,2.2126090356281827,
+1,184,0.0761294811964035,23,2.033785240990775,
+1,192,0.08202523365616798,24,1.9919723123311996,
+1,200,0.07628072425723076,25,1.7949061763697658,
+1,208,0.0784546248614788,26,1.5366853833198548,
+1,216,0.0869893953204155,27,1.2262009448475308,
+1,224,0.07972022145986557,28,1.497286565842167,
+1,232,0.08121772855520248,29,1.4650954405466716,
+1,240,0.07640029862523079,30,1.76709544217145,
+1,248,0.08160492032766342,31,1.195439001609539,
+1,256,0.08063583821058273,32,1.9514090542135567,
+,,,32,,1.5661012466344457
+1,264,0.0758480355143547,33,1.3703056919959284,
+1,272,0.07965521514415741,34,1.7205631125856329,
+1,280,0.07957543060183525,35,2.1078760952785096,
+1,288,0.0808374248445034,36,1.391715651466733,
+1,296,0.07660144194960594,37,1.5229190270105997,
+1,304,0.07613057643175125,38,1.1857397437095643,
+2,312,0.08105170354247093,39,1.4523041761583753,
+2,320,0.07964178919792175,40,0.8246737561727825,
+2,328,0.08194279670715332,41,1.1756470587945753,
+2,336,0.07796105369925499,42,1.2552748719851177,
+2,344,0.08080613985657692,43,0.9264966470223887,
+2,352,0.08059879392385483,44,0.9736070151512439,
+2,360,0.0762358009815216,45,1.401765607021473,
+2,368,0.08251360431313515,46,1.0940892631188035,
+2,376,0.08006580546498299,47,0.6867778870192441,
+2,384,0.07554114982485771,48,0.8855126520683025,
+,,,48,,0.6725706542279087
+2,392,0.08119768649339676,49,0.5306956503126357,
+2,400,0.07640928775072098,50,0.7529445330684001,
+2,408,0.07648879289627075,51,0.9772441033273935,
+2,416,0.08261218294501305,52,0.806821551322937,
+2,424,0.07634934410452843,53,0.6013531816693453,
+2,432,0.08255348727107048,54,0.4793394379890882,
+2,440,0.07707996293902397,55,1.045152180153748,
+2,448,0.08096887916326523,56,0.47096291308601695,
+2,456,0.08088582381606102,57,0.4629532111187776,
+3,464,0.08024600520730019,58,0.411201739105685,
+3,472,0.07615400850772858,59,0.39088416420694055,
+3,480,0.0852845124900341,60,0.4160480417551533,
+3,488,0.08170796930789948,61,0.2052328262863488,
+3,496,0.07765119895339012,62,0.5011219114065171,
+3,504,0.07743249088525772,63,0.2920545196581271,
+3,512,0.08184167742729187,64,0.06196220939119275,
+,,,64,,0.2965837898900953
+3,520,0.07639065012335777,65,0.1769609681710049,
+3,528,0.07876966893672943,66,0.4924048261406521,
+3,536,0.08122127503156662,67,0.17013415841963783,
+3,544,0.08725094050168991,68,0.18195277379293526,
+3,552,0.07747167348861694,69,0.3597257772150139,
+3,560,0.07957847788929939,70,0.4243425264440734,
+3,568,0.08156620338559151,71,0.08218356514615673,
+3,576,0.08031244203448296,72,0.9947862213929043,
+3,584,0.07941707968711853,73,0.18634830034253272,
+3,592,0.07876205816864967,74,0.5666589183466775,
+3,600,0.07881605997681618,75,0.41652951103945574,
+3,608,0.08015795797109604,76,0.23235163522454408,
+3,616,0.08064737170934677,77,0.19293122462025195,
+4,624,0.07412714138627052,78,0.055583399906754496,
+4,632,0.08050036430358887,79,0.26034141704440117,
+4,640,0.076687753200531,80,0.1070822188258171,
+,,,80,,0.1379297547421213

finetune_lora/qwen2.5-7b-instruct/banking77/size=16/seed=4/lora_ans/0.0-1.0/0.7-1.0/test=banking77/list=test_1000/labels.csv

@@ -0,0 +1,1000 @@
+6197,55
+6441,63
+5254,42
+4058,38
+5596,27
+114,11
+2806,66
+1087,12
+1139,4
+12996,2
+10447,41
+11288,73
+12620,37
+10691,8
+8838,20
+11829,25
+5055,26
+5484,52
+2936,54
+9239,0
+5848,25
+2035,47
+1266,14
+11887,55
+4073,38
+4823,44
+2662,15
+5923,25
+2923,54
+7478,35
+12409,39
+11263,38
+12272,50
+8760,5
+1955,47
+7756,71
+12943,9
+8885,20
+6776,53
+12715,5
+11892,55
+4231,73
+3603,16
+1096,12
+2786,66
+12791,31
+7613,65
+12357,65
+11662,42
+9066,57
+63,11
+5620,27
+6193,55
+871,46
+10854,66
+8147,43
+3305,6
+3248,61
+2603,15
+3482,16
+4639,28
+7136,7
+6570,70
+7884,39
+12307,35
+10342,4
+257,13
+10939,40
+597,34
+3076,40
+8498,59
+7988,58
+575,34
+10692,8
+416,17
+2046,47
+1802,23
+12925,9
+5787,51
+2827,66
+4537,22
+3049,40
+11130,30
+2725,66
+1417,33
+12180,7
+6157,55
+8864,20
+12995,2
+2324,60
+2251,60
+1989,47
+16,11
+8827,20
+8493,59
+4953,26
+10807,15
+7101,7
+8210,76
+2312,60
+4330,62
+12365,71
+11466,3
+1666,1
+4063,38
+4622,3
+3836,74
+11584,26
+9371,19
+3407,6
+2680,15
+3845,74
+10741,60
+9232,0
+5932,25
+6113,48
+5411,52
+12422,39
+12101,53
+10322,12
+963,36
+11640,45
+5790,51
+519,17
+9694,2
+96,11
+7911,39
+6069,48
+13002,2
+7853,39
+8660,5
+2817,66
+11,11
+1701,49
+5397,52
+2685,15
+425,17
+1958,47
+12202,7
+5036,26
+6054,48
+10515,1
+12164,7
+4865,44
+12189,7
+8355,76
+7023,7
+12813,57
+7527,35
+2435,75
+9940,24
+7402,50
+32,11
+1541,41
+3094,40
+12947,9
+5604,27
+1775,49
+8117,43
+9587,9
+2008,47
+2741,66
+582,34
+2562,15
+7763,71
+7034,7
+6376,63
+9092,57
+202,13
+11193,74
+6028,48
+1476,41
+10315,12
+9472,19
+6092,48
+1919,56
+2986,54
+6078,48
+6698,67
+11430,22
+8139,43
+7580,65
+9696,2
+11768,51
+7030,7
+3711,30
+8118,43
+4895,44
+8450,37
+2225,8
+6803,53
+3040,40
+2834,66
+10998,10
+3955,68
+11164,74
+3599,16
+4288,62
+2846,66
+4278,62
+1150,4
+8719,5
+9210,0
+10575,23
+317,32
+6922,21
+10171,34
+8842,20
+11121,16
+6293,18
+3204,61
+591,34
+9236,0
+10820,15
+996,36
+8481,59
+10583,23
+1555,41
+3754,30
+2257,60
+11351,62
+12515,43
+1792,23
+10659,47
+8290,76
+9713,2
+12555,72
+7311,64
+2599,15
+10911,54
+6332,63
+5152,45
+4602,3
+6340,63
+10385,14
+1807,23
+8458,37
+3319,6
+10078,13
+8380,37
+2739,66
+1115,12
+9783,69
+2158,8
+5652,27
+12173,7
+5744,51
+3782,74
+3456,6
+9414,19
+4810,28
+11760,27
+8925,20
+7440,35
+1883,56
+9930,24
+12920,19
+6413,63
+7596,65
+11819,25
+6624,67
+11903,55
+3858,74
+9103,57
+3555,16
+4457,29
+2409,75
+9174,57
+7133,7
+3159,61
+10129,17
+3150,10
+2201,8
+7666,65
+2865,66
+12815,57
+3282,61
+3886,74
+775,46
+6393,63
+10102,32
+7652,65
+6523,70
+7589,65
+866,46
+4929,26
+5588,27
+3759,30
+10532,49
+9122,57
+12466,58
+4214,73
+5337,42
+574,34
+9839,69
+12404,39
+296,32
+7868,39
+12839,57
+6542,70
+122,11
+750,46
+10136,17
+12891,19
+3235,61
+7444,35
+12924,9
+8768,5
+7057,7
+8917,20
+7602,65
+9870,69
+3316,6
+5694,51
+2078,8
+8022,58
+4839,44
+730,34
+12123,21
+4332,62
+7453,35
+4418,29
+6744,53
+9728,2
+1611,1
+10779,75
+6783,53
+4472,22
+367,32
+10636,56
+1696,49
+3281,61
+6690,67
+4800,28
+8328,76
+4827,44
+530,17
+3977,68
+12840,57
+10417,33
+3545,16
+2396,75
+5344,42
+1829,56
+10084,32
+6306,63
+3692,30
+11054,6
+12948,9
+1656,1
+2338,75
+12260,50
+1372,33
+6644,67
+4988,26
+6775,53
+8683,5
+2032,47
+6371,63
+6946,21
+3676,30
+2414,75
+636,34
+12581,76
+6547,70
+7395,50
+3136,10
+7163,64
+8426,37
+5921,25
+6807,53
+9788,69
+12732,20
+8341,76
+2233,8
+3528,16
+9238,0
+2532,15
+9960,24
+8977,31
+11800,51
+3432,6
+8708,5
+6902,21
+4076,38
+6444,63
+812,46
+7036,7
+1971,47
+9335,0
+3942,68
+5775,51
+9428,19
+8656,5
+9465,19
+5128,45
+10072,13
+9836,69
+9955,24
+1312,14
+3554,16
+10852,66
+5417,52
+180,13
+2776,66
+11345,62
+3655,30
+7277,64
+13060,24
+887,36
+1485,41
+8296,76
+12531,72
+11407,22
+9162,57
+4452,29
+12471,58
+3355,6
+4651,28
+3943,68
+7060,7
+3775,74
+3853,74
+1191,4
+3992,68
+7669,65
+5048,26
+9818,69
+4304,62
+10406,33
+9906,24
+11208,68
+1419,33
+4781,28
+3255,61
+455,17
+6948,21
+6594,67
+9470,19
+1844,56
+11622,45
+372,32
+109,11
+1525,41
+6630,67
+2468,75
+583,34
+11698,52
+5557,27
+13074,24
+6325,63
+11086,16
+1524,41
+6932,21
+3715,30
+6936,21
+914,36
+10052,13
+3295,61
+7266,64
+4371,29
+5666,27
+11409,22
+7414,50
+11770,51
+11278,38
+1869,56
+9650,2
+2715,66
+811,46
+11947,18
+547,17
+4377,29
+1147,4
+9112,57
+11843,48
+7886,39
+12316,35
+6926,21
+12835,57
+2451,75
+7372,50
+9207,0
+8537,59
+11889,55
+12083,53
+9037,31
+12988,2
+1114,12
+12685,5
+2915,54
+9370,19
+2337,75
+2649,15
+4955,26
+9692,2
+9643,9
+3059,40
+3724,30
+8145,43
+6717,67
+5525,52
+6551,70
+578,34
+2871,54
+9211,0
+11002,10
+2539,15
+6279,18
+10878,66
+804,46
+10910,54
+6890,21
+9849,69
+3021,40
+4061,38
+8237,76
+12476,58
+3211,61
+7895,39
+4703,28
+5236,45
+8303,76
+2867,66
+6756,53
+4719,28
+2131,8
+12739,20
+6798,53
+5817,51
+11425,22
+11553,44
+10949,40
+10671,47
+9781,69
+5571,27
+1493,41
+1399,33
+1126,4
+494,17
+527,17
+12562,72
+274,13
+1354,14
+7309,64
+10444,41
+11669,42
+7329,64
+4329,62
+2104,8
+12551,72
+11595,26
+2130,8
+12940,9
+4023,38
+1740,49
+686,34
+9012,31
+4606,3
+4133,73
+7875,39
+841,46
+5378,52
+9804,69
+11946,18
+7709,71
+3143,10
+834,46
+4445,29
+9734,2
+10973,10
+9572,9
+11203,68
+747,46
+12687,5
+2988,54
+8714,5
+6366,63
+3100,10
+11038,61
+12437,39
+4657,28
+3653,30
+9914,24
+3517,16
+9979,24
+7470,35
+8547,59
+533,17
+5114,45
+8769,5
+640,34
+3198,61
+10835,15
+9364,19
+10943,40
+9819,69
+3048,40
+1884,56
+9637,9
+11507,28
+363,32
+8159,43
+12295,35
+4648,28
+4052,38
+11134,30
+7619,65
+3556,16
+863,46
+4437,29
+2423,75
+9873,69
+11736,27
+5320,42
+4359,29
+8285,76
+12299,35
+9013,31
+9314,0
+5011,26
+3694,30
+3720,30
+8497,59
+7067,7
+5140,45
+5258,42
+520,17
+1946,47
+6438,63
+6916,21
+5745,51
+10890,54
+12326,65
+11163,74
+6403,63
+8797,20
+1556,41
+6049,48
+2499,75
+5195,45
+9966,24
+12532,72
+8402,37
+12369,71
+2493,75
+7910,39
+11521,28
+239,13
+6718,67
+11847,48
+576,34
+1566,1
+5690,51
+6944,21
+8723,5
+6622,67
+4659,28
+8658,5
+7814,39
+8586,59
+8160,43
+3709,30
+11620,45
+11840,25
+10249,36
+12833,57
+12484,43
+13072,24
+669,34
+6232,55
+1112,12
+8281,76
+5428,52
+8909,20
+12116,53
+11534,44
+1330,14
+11429,22
+4934,26
+112,11
+3690,30
+5058,26
+2723,66
+10599,23
+3824,74
+4482,22
+12187,7
+5339,42
+1794,23
+1813,23
+9626,9
+2502,75
+356,32
+3176,61
+6263,18
+4373,29
+5736,51
+10314,12
+9450,19
+12498,43
+5155,45
+3811,74
+2668,15
+906,36
+2883,54
+6849,53
+2237,60
+11850,48
+6414,63
+9383,19
+851,46
+7802,71
+7696,71
+3634,16
+5640,27
+8722,5
+4731,28
+9543,9
+1183,4
+9019,31
+5836,25
+7845,39
+11955,18
+2361,75
+13019,69
+1878,56
+5198,45
+917,36
+10725,60
+8619,5
+4928,26
+9789,69
+5300,42
+9756,2
+3586,16
+8692,5
+12781,31
+6175,55
+11744,27
+11863,48
+4357,29
+11624,45
+10726,60
+9652,2
+5939,25
+3114,10
+4292,62
+5810,51
+6885,21
+8479,59
+1629,1
+7561,35
+12056,67
+3330,6
+10125,17
+10358,4
+10849,66
+1857,56
+2073,47
+10603,56
+6613,67
+0,11
+726,34
+8246,76
+9434,19
+7770,71
+6508,70
+9700,2
+8075,43
+8105,43
+566,17
+7824,39
+7778,71
+11714,52
+3427,6
+7165,64
+8802,20
+8944,31
+12936,9
+12156,21
+12265,50
+3131,10
+1923,56
+4866,44
+11664,42
+2905,54
+6947,21
+2114,8
+7536,35
+10215,46
+5319,42
+2093,8
+12463,58
+4978,26
+6038,48
+1113,12
+11394,29
+1422,33
+6544,70
+8647,5
+1679,49
+12356,65
+12070,67
+12277,50
+774,46
+9699,2
+8452,37
+5649,27
+6680,67
+9679,2
+10266,36
+3356,6
+764,46
+7725,71
+250,13
+10111,32
+1219,4
+3056,40
+11428,22
+5368,42
+352,32
+12923,9
+12676,59
+24,11
+10540,49
+10208,46
+3068,40
+4997,26
+9194,0
+9071,57
+1020,12
+8119,43
+3856,74
+2772,66
+5789,51
+4760,28
+4568,3
+10079,13
+7062,7
+5173,45
+7731,71
+9629,9
+6458,63
+11468,3
+12564,76
+171,13
+10128,17
+11103,16
+12171,7
+285,13
+7717,71
+6847,53
+8517,59
+4927,26
+3145,10
+2188,8
+9437,19
+8682,5
+8784,20
+12106,53
+7129,7
+6716,67
+2291,60
+9988,24
+2379,75
+5882,25
+5051,26
+8308,76
+12004,70
+6426,63
+7593,65
+2106,8
+8528,59
+5135,45
+5635,27
+1675,49
+4101,73
+12311,35
+2377,75
+7452,35
+2192,8
+1103,12
+4739,28
+7524,35
+2036,47
+11540,44
+1744,49
+6503,70
+5827,51
+7515,35
+6215,55
+8696,5
+9185,0
+9842,69
+3405,6
+4479,22
+3431,6
+11262,38
+10399,14
+1179,4
+5724,51
+10804,15
+7540,35
+7606,65
+12830,57
+12883,19
+623,34
+10616,56
+7358,50
+10683,8
+9646,2
+3727,30
+7598,65
+1901,56
+2397,75
+10740,60
+11073,6
+11680,42
+1347,14
+10570,23
+3078,40
+10134,17
+2862,66
+3728,30
+3507,16
+8727,5
+6883,21
+4761,28
+11283,73
+8201,72
+9821,69
+9349,19
+7231,64
+5608,27
+6367,63
+13061,24
+12027,70
+1538,41
+11943,18
+12606,37
+6153,55
+11988,63
+5526,52
+6668,67
+10397,14
+10564,23
+3058,40
+11410,22
+7777,71
+5475,52
+3772,74
+9440,19
+722,34
+5249,45
+8912,20
+11317,73
+6921,21
+11175,74
+3919,68
+9682,2
+7156,7
+3571,16
+11953,18
+5269,42
+626,34
+761,46
+8172,72
+10331,4
+8158,43
+11477,3
+8219,76
+6909,21
+437,17
+4306,62
+10138,17
+6498,70
+9589,9
+11147,30
+12803,57
+185,13
+1823,56
+11964,63
+517,17
+7164,64
+12251,50
+12896,19
+19,11
+4888,44
+3508,16
+14,11
+10830,15
+5634,27
+4901,44
+3225,61
+11711,52
+956,36
+2176,8
+10410,33
+8943,20

finetune_lora/qwen2.5-7b-instruct/banking77/size=16/seed=4/lora_ans/0.0-1.0/0.7-1.0/train_args.yaml

@@ -0,0 +1,10 @@
+K: null
+global_batch_size: 8
+learning_rate: 0.0001
+loss: ans
+max_steps: 80
+micro_batch_size: 1
+optimizer_name: adamw
+patience: 10
+val_check_interval: 16
+weight_decay: 0.0

finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=0/lora_ans_no_es/0.0-0.7/0.0-0.3/best.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8ab4428e8cbfe27581a9f2d0b955a4bcfabaac91d8ab022885e89935261a895d
+size 42951934

finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=0/lora_ans_no_es/0.0-0.7/0.0-0.3/checkpoint/config.json

@@ -0,0 +1,27 @@
+{
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "model_type": "qwen2",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "rms_norm_eps": 1e-06,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.43.1",
+  "use_cache": true,
+  "use_sliding_window": false,
+  "vocab_size": 152064
+}

finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=0/lora_ans_no_es/0.0-0.7/0.0-0.3/checkpoint/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "pad_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_p": 0.8,
+  "top_k": 20,
+  "transformers_version": "4.37.0"
+}

finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=0/lora_ans_no_es/0.0-0.7/0.0-0.3/checkpoint/lit_model.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f2893700d468552b27b00bb16e3906c7a01f49c09f3c786c252a2b5d36d179f
+size 15231297418

finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=0/lora_ans_no_es/0.0-0.7/0.0-0.3/checkpoint/lit_model.pth.lora

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:30b223c02d1e1f841f0a1ebfc4adc640ea3f92b0e0e9dbebb11b1f2189372c8b
+size 42951614

finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=0/lora_ans_no_es/0.0-0.7/0.0-0.3/checkpoint/model_config.yaml

@@ -0,0 +1,37 @@
+attention_logit_softcapping: null
+attention_scores_scalar: null
+attn_bias: true
+bias: false
+block_size: 131072
+final_logit_softcapping: null
+gelu_approximate: none
+head_size: 128
+hf_config:
+  name: Qwen2.5-7B-Instruct
+  org: Qwen
+intermediate_size: 18944
+lm_head_bias: false
+mlp_class_name: LLaMAMLP
+n_embd: 3584
+n_expert: 0
+n_expert_per_token: 0
+n_head: 28
+n_layer: 28
+n_query_groups: 4
+name: Qwen2.5-7B-Instruct
+norm_class_name: RMSNorm
+norm_eps: 1.0e-06
+padded_vocab_size: 152064
+padding_multiple: 512
+parallel_residual: false
+post_attention_norm: false
+post_mlp_norm: false
+rope_adjustments: null
+rope_base: 1000000
+rope_condense_ratio: 1
+rotary_percentage: 1.0
+scale_embeddings: false
+shared_attention_norm: false
+sliding_window_layer_placing: null
+sliding_window_size: null
+vocab_size: 151643

finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=0/lora_ans_no_es/0.0-0.7/0.0-0.3/checkpoint/tokenizer_config.json

@@ -0,0 +1,207 @@
+{
+  "add_bos_token": false,
+  "add_prefix_space": false,
+  "added_tokens_decoder": {
+    "151643": {
+      "content": "<|endoftext|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151644": {
+      "content": "<|im_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151645": {
+      "content": "<|im_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151646": {
+      "content": "<|object_ref_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151647": {
+      "content": "<|object_ref_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151648": {
+      "content": "<|box_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151649": {
+      "content": "<|box_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151650": {
+      "content": "<|quad_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151651": {
+      "content": "<|quad_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151652": {
+      "content": "<|vision_start|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151653": {
+      "content": "<|vision_end|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151654": {
+      "content": "<|vision_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151655": {
+      "content": "<|image_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151656": {
+      "content": "<|video_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "151657": {
+      "content": "<tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151658": {
+      "content": "</tool_call>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151659": {
+      "content": "<|fim_prefix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151660": {
+      "content": "<|fim_middle|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151661": {
+      "content": "<|fim_suffix|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151662": {
+      "content": "<|fim_pad|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151663": {
+      "content": "<|repo_name|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    },
+    "151664": {
+      "content": "<|file_sep|>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": false
+    }
+  },
+  "additional_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "bos_token": null,
+  "chat_template": "{%- if tools %}\n    {{- '<|im_start|>system\\n' }}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- messages[0]['content'] }}\n    {%- else %}\n        {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}\n    {%- endif %}\n    {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n    {%- for tool in tools %}\n        {{- \"\\n\" }}\n        {{- tool | tojson }}\n    {%- endfor %}\n    {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n    {%- if messages[0]['role'] == 'system' %}\n        {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n    {%- else %}\n        {{- '<|im_start|>system\\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\\n' }}\n    {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n    {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n        {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n    {%- elif message.role == \"assistant\" %}\n        {{- '<|im_start|>' + message.role }}\n        {%- if message.content %}\n            {{- '\\n' + message.content }}\n        {%- endif %}\n        {%- for tool_call in message.tool_calls %}\n            {%- if tool_call.function is defined %}\n                {%- set tool_call = tool_call.function %}\n            {%- endif %}\n            {{- '\\n<tool_call>\\n{\"name\": \"' }}\n            {{- tool_call.name }}\n            {{- '\", \"arguments\": ' }}\n            {{- tool_call.arguments | tojson }}\n            {{- '}\\n</tool_call>' }}\n        {%- endfor %}\n        {{- '<|im_end|>\\n' }}\n    {%- elif message.role == \"tool\" %}\n        {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n            {{- '<|im_start|>user' }}\n        {%- endif %}\n        {{- '\\n<tool_response>\\n' }}\n        {{- message.content }}\n        {{- '\\n</tool_response>' }}\n        {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n            {{- '<|im_end|>\\n' }}\n        {%- endif %}\n    {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n    {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n",
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=0/lora_ans_no_es/0.0-0.7/0.0-0.3/last.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:84955cef7d1e787b0c026015aa914237537283fa4bd894aff9be50340811ffe2
+size 42951934

finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=0/lora_ans_no_es/0.0-0.7/0.0-0.3/logs/events.out.tfevents.1735554367.gamma.2619525.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7d1fbbe02db3129a6dbff5d9cab1e24a19cdde29d6ef2e25a22f9a6b1798217d
+size 926766

finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=0/lora_ans_no_es/0.0-0.7/0.0-0.3/logs/hparams.yaml

@@ -0,0 +1 @@
+{}

finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=0/lora_ans_no_es/0.0-0.7/0.0-0.3/test=banking77/list=0.7-1.0/labels.csv

@@ -0,0 +1,184 @@
+9777,69
+8433,37
+11008,61
+6369,63
+6636,67
+8438,37
+1882,56
+5097,45
+273,13
+3266,61
+7981,58
+7837,39
+3200,61
+8368,76
+8197,72
+3656,30
+6897,21
+8754,5
+1481,41
+10932,40
+4724,28
+11068,6
+1579,1
+136,11
+12898,19
+9254,0
+2148,8
+3300,61
+11393,29
+4538,22
+9002,31
+7758,71
+6516,70
+11448,3
+9127,57
+2343,75
+13070,24
+1782,23
+8089,43
+11050,6
+3312,6
+10687,8
+4653,28
+6935,21
+8834,20
+3650,30
+147,11
+269,13
+3392,6
+1130,4
+2755,66
+7194,64
+10176,34
+7513,35
+251,13
+11949,18
+9876,24
+3415,6
+4125,73
+8373,37
+3445,6
+7880,39
+8343,76
+4848,44
+10060,13
+8039,58
+2110,8
+823,46
+4368,29
+3990,68
+11342,62
+8812,20
+1080,12
+8477,59
+5003,26
+12015,70
+7863,39
+9775,69
+3047,40
+6635,67
+7985,58
+7879,39
+5990,48
+9153,57
+3165,61
+1366,33
+4038,38
+9618,9
+5586,27
+2428,75
+1511,41
+1177,4
+2375,75
+3834,74
+11590,26
+2873,54
+9502,19
+12922,19
+295,32
+2212,8
+7253,64
+4282,62
+3481,6
+9148,57
+7525,35
+4364,29
+571,34
+6694,67
+10210,46
+11039,61
+7935,58
+10621,56
+12678,59
+756,46
+5380,52
+6449,63
+4707,28
+12368,71
+6534,70
+1573,1
+11911,55
+360,32
+9179,0
+181,13
+8623,5
+2391,75
+1188,4
+11166,74
+4465,22
+148,11
+9875,24
+1204,4
+8360,76
+6969,21
+7860,39
+6461,63
+11385,29
+4908,44
+7804,39
+2191,8
+6205,55
+4576,3
+2064,47
+755,46
+5834,51
+5545,27
+1418,33
+12932,9
+2382,75
+3380,6
+12490,43
+7744,71
+9298,0
+4818,44
+1624,1
+10508,1
+12271,50
+9331,0
+6634,67
+5224,45
+8272,76
+7548,35
+8182,72
+10496,1
+11618,45
+2489,75
+9379,19
+11519,28
+4949,26
+6880,53
+1588,1
+12544,72
+8514,59
+10040,11
+1393,33
+4609,3
+1201,4
+9261,0
+5129,45
+9309,0
+3262,61
+11560,44
+2705,66
+12677,59

finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=0/lora_ans_no_es/0.0-0.7/0.0-0.3/test=banking77/list=test_1000/labels.csv

@@ -0,0 +1,1000 @@
+6197,55
+6441,63
+5254,42
+4058,38
+5596,27
+114,11
+2806,66
+1087,12
+1139,4
+12996,2
+10447,41
+11288,73
+12620,37
+10691,8
+8838,20
+11829,25
+5055,26
+5484,52
+2936,54
+9239,0
+5848,25
+2035,47
+1266,14
+11887,55
+4073,38
+4823,44
+2662,15
+5923,25
+2923,54
+7478,35
+12409,39
+11263,38
+12272,50
+8760,5
+1955,47
+7756,71
+12943,9
+8885,20
+6776,53
+12715,5
+11892,55
+4231,73
+3603,16
+1096,12
+2786,66
+12791,31
+7613,65
+12357,65
+11662,42
+9066,57
+63,11
+5620,27
+6193,55
+871,46
+10854,66
+8147,43
+3305,6
+3248,61
+2603,15
+3482,16
+4639,28
+7136,7
+6570,70
+7884,39
+12307,35
+10342,4
+257,13
+10939,40
+597,34
+3076,40
+8498,59
+7988,58
+575,34
+10692,8
+416,17
+2046,47
+1802,23
+12925,9
+5787,51
+2827,66
+4537,22
+3049,40
+11130,30
+2725,66
+1417,33
+12180,7
+6157,55
+8864,20
+12995,2
+2324,60
+2251,60
+1989,47
+16,11
+8827,20
+8493,59
+4953,26
+10807,15
+7101,7
+8210,76
+2312,60
+4330,62
+12365,71
+11466,3
+1666,1
+4063,38
+4622,3
+3836,74
+11584,26
+9371,19
+3407,6
+2680,15
+3845,74
+10741,60
+9232,0
+5932,25
+6113,48
+5411,52
+12422,39
+12101,53
+10322,12
+963,36
+11640,45
+5790,51
+519,17
+9694,2
+96,11
+7911,39
+6069,48
+13002,2
+7853,39
+8660,5
+2817,66
+11,11
+1701,49
+5397,52
+2685,15
+425,17
+1958,47
+12202,7
+5036,26
+6054,48
+10515,1
+12164,7
+4865,44
+12189,7
+8355,76
+7023,7
+12813,57
+7527,35
+2435,75
+9940,24
+7402,50
+32,11
+1541,41
+3094,40
+12947,9
+5604,27
+1775,49
+8117,43
+9587,9
+2008,47
+2741,66
+582,34
+2562,15
+7763,71
+7034,7
+6376,63
+9092,57
+202,13
+11193,74
+6028,48
+1476,41
+10315,12
+9472,19
+6092,48
+1919,56
+2986,54
+6078,48
+6698,67
+11430,22
+8139,43
+7580,65
+9696,2
+11768,51
+7030,7
+3711,30
+8118,43
+4895,44
+8450,37
+2225,8
+6803,53
+3040,40
+2834,66
+10998,10
+3955,68
+11164,74
+3599,16
+4288,62
+2846,66
+4278,62
+1150,4
+8719,5
+9210,0
+10575,23
+317,32
+6922,21
+10171,34
+8842,20
+11121,16
+6293,18
+3204,61
+591,34
+9236,0
+10820,15
+996,36
+8481,59
+10583,23
+1555,41
+3754,30
+2257,60
+11351,62
+12515,43
+1792,23
+10659,47
+8290,76
+9713,2
+12555,72
+7311,64
+2599,15
+10911,54
+6332,63
+5152,45
+4602,3
+6340,63
+10385,14
+1807,23
+8458,37
+3319,6
+10078,13
+8380,37
+2739,66
+1115,12
+9783,69
+2158,8
+5652,27
+12173,7
+5744,51
+3782,74
+3456,6
+9414,19
+4810,28
+11760,27
+8925,20
+7440,35
+1883,56
+9930,24
+12920,19
+6413,63
+7596,65
+11819,25
+6624,67
+11903,55
+3858,74
+9103,57
+3555,16
+4457,29
+2409,75
+9174,57
+7133,7
+3159,61
+10129,17
+3150,10
+2201,8
+7666,65
+2865,66
+12815,57
+3282,61
+3886,74
+775,46
+6393,63
+10102,32
+7652,65
+6523,70
+7589,65
+866,46
+4929,26
+5588,27
+3759,30
+10532,49
+9122,57
+12466,58
+4214,73
+5337,42
+574,34
+9839,69
+12404,39
+296,32
+7868,39
+12839,57
+6542,70
+122,11
+750,46
+10136,17
+12891,19
+3235,61
+7444,35
+12924,9
+8768,5
+7057,7
+8917,20
+7602,65
+9870,69
+3316,6
+5694,51
+2078,8
+8022,58
+4839,44
+730,34
+12123,21
+4332,62
+7453,35
+4418,29
+6744,53
+9728,2
+1611,1
+10779,75
+6783,53
+4472,22
+367,32
+10636,56
+1696,49
+3281,61
+6690,67
+4800,28
+8328,76
+4827,44
+530,17
+3977,68
+12840,57
+10417,33
+3545,16
+2396,75
+5344,42
+1829,56
+10084,32
+6306,63
+3692,30
+11054,6
+12948,9
+1656,1
+2338,75
+12260,50
+1372,33
+6644,67
+4988,26
+6775,53
+8683,5
+2032,47
+6371,63
+6946,21
+3676,30
+2414,75
+636,34
+12581,76
+6547,70
+7395,50
+3136,10
+7163,64
+8426,37
+5921,25
+6807,53
+9788,69
+12732,20
+8341,76
+2233,8
+3528,16
+9238,0
+2532,15
+9960,24
+8977,31
+11800,51
+3432,6
+8708,5
+6902,21
+4076,38
+6444,63
+812,46
+7036,7
+1971,47
+9335,0
+3942,68
+5775,51
+9428,19
+8656,5
+9465,19
+5128,45
+10072,13
+9836,69
+9955,24
+1312,14
+3554,16
+10852,66
+5417,52
+180,13
+2776,66
+11345,62
+3655,30
+7277,64
+13060,24
+887,36
+1485,41
+8296,76
+12531,72
+11407,22
+9162,57
+4452,29
+12471,58
+3355,6
+4651,28
+3943,68
+7060,7
+3775,74
+3853,74
+1191,4
+3992,68
+7669,65
+5048,26
+9818,69
+4304,62
+10406,33
+9906,24
+11208,68
+1419,33
+4781,28
+3255,61
+455,17
+6948,21
+6594,67
+9470,19
+1844,56
+11622,45
+372,32
+109,11
+1525,41
+6630,67
+2468,75
+583,34
+11698,52
+5557,27
+13074,24
+6325,63
+11086,16
+1524,41
+6932,21
+3715,30
+6936,21
+914,36
+10052,13
+3295,61
+7266,64
+4371,29
+5666,27
+11409,22
+7414,50
+11770,51
+11278,38
+1869,56
+9650,2
+2715,66
+811,46
+11947,18
+547,17
+4377,29
+1147,4
+9112,57
+11843,48
+7886,39
+12316,35
+6926,21
+12835,57
+2451,75
+7372,50
+9207,0
+8537,59
+11889,55
+12083,53
+9037,31
+12988,2
+1114,12
+12685,5
+2915,54
+9370,19
+2337,75
+2649,15
+4955,26
+9692,2
+9643,9
+3059,40
+3724,30
+8145,43
+6717,67
+5525,52
+6551,70
+578,34
+2871,54
+9211,0
+11002,10
+2539,15
+6279,18
+10878,66
+804,46
+10910,54
+6890,21
+9849,69
+3021,40
+4061,38
+8237,76
+12476,58
+3211,61
+7895,39
+4703,28
+5236,45
+8303,76
+2867,66
+6756,53
+4719,28
+2131,8
+12739,20
+6798,53
+5817,51
+11425,22
+11553,44
+10949,40
+10671,47
+9781,69
+5571,27
+1493,41
+1399,33
+1126,4
+494,17
+527,17
+12562,72
+274,13
+1354,14
+7309,64
+10444,41
+11669,42
+7329,64
+4329,62
+2104,8
+12551,72
+11595,26
+2130,8
+12940,9
+4023,38
+1740,49
+686,34
+9012,31
+4606,3
+4133,73
+7875,39
+841,46
+5378,52
+9804,69
+11946,18
+7709,71
+3143,10
+834,46
+4445,29
+9734,2
+10973,10
+9572,9
+11203,68
+747,46
+12687,5
+2988,54
+8714,5
+6366,63
+3100,10
+11038,61
+12437,39
+4657,28
+3653,30
+9914,24
+3517,16
+9979,24
+7470,35
+8547,59
+533,17
+5114,45
+8769,5
+640,34
+3198,61
+10835,15
+9364,19
+10943,40
+9819,69
+3048,40
+1884,56
+9637,9
+11507,28
+363,32
+8159,43
+12295,35
+4648,28
+4052,38
+11134,30
+7619,65
+3556,16
+863,46
+4437,29
+2423,75
+9873,69
+11736,27
+5320,42
+4359,29
+8285,76
+12299,35
+9013,31
+9314,0
+5011,26
+3694,30
+3720,30
+8497,59
+7067,7
+5140,45
+5258,42
+520,17
+1946,47
+6438,63
+6916,21
+5745,51
+10890,54
+12326,65
+11163,74
+6403,63
+8797,20
+1556,41
+6049,48
+2499,75
+5195,45
+9966,24
+12532,72
+8402,37
+12369,71
+2493,75
+7910,39
+11521,28
+239,13
+6718,67
+11847,48
+576,34
+1566,1
+5690,51
+6944,21
+8723,5
+6622,67
+4659,28
+8658,5
+7814,39
+8586,59
+8160,43
+3709,30
+11620,45
+11840,25
+10249,36
+12833,57
+12484,43
+13072,24
+669,34
+6232,55
+1112,12
+8281,76
+5428,52
+8909,20
+12116,53
+11534,44
+1330,14
+11429,22
+4934,26
+112,11
+3690,30
+5058,26
+2723,66
+10599,23
+3824,74
+4482,22
+12187,7
+5339,42
+1794,23
+1813,23
+9626,9
+2502,75
+356,32
+3176,61
+6263,18
+4373,29
+5736,51
+10314,12
+9450,19
+12498,43
+5155,45
+3811,74
+2668,15
+906,36
+2883,54
+6849,53
+2237,60
+11850,48
+6414,63
+9383,19
+851,46
+7802,71
+7696,71
+3634,16
+5640,27
+8722,5
+4731,28
+9543,9
+1183,4
+9019,31
+5836,25
+7845,39
+11955,18
+2361,75
+13019,69
+1878,56
+5198,45
+917,36
+10725,60
+8619,5
+4928,26
+9789,69
+5300,42
+9756,2
+3586,16
+8692,5
+12781,31
+6175,55
+11744,27
+11863,48
+4357,29
+11624,45
+10726,60
+9652,2
+5939,25
+3114,10
+4292,62
+5810,51
+6885,21
+8479,59
+1629,1
+7561,35
+12056,67
+3330,6
+10125,17
+10358,4
+10849,66
+1857,56
+2073,47
+10603,56
+6613,67
+0,11
+726,34
+8246,76
+9434,19
+7770,71
+6508,70
+9700,2
+8075,43
+8105,43
+566,17
+7824,39
+7778,71
+11714,52
+3427,6
+7165,64
+8802,20
+8944,31
+12936,9
+12156,21
+12265,50
+3131,10
+1923,56
+4866,44
+11664,42
+2905,54
+6947,21
+2114,8
+7536,35
+10215,46
+5319,42
+2093,8
+12463,58
+4978,26
+6038,48
+1113,12
+11394,29
+1422,33
+6544,70
+8647,5
+1679,49
+12356,65
+12070,67
+12277,50
+774,46
+9699,2
+8452,37
+5649,27
+6680,67
+9679,2
+10266,36
+3356,6
+764,46
+7725,71
+250,13
+10111,32
+1219,4
+3056,40
+11428,22
+5368,42
+352,32
+12923,9
+12676,59
+24,11
+10540,49
+10208,46
+3068,40
+4997,26
+9194,0
+9071,57
+1020,12
+8119,43
+3856,74
+2772,66
+5789,51
+4760,28
+4568,3
+10079,13
+7062,7
+5173,45
+7731,71
+9629,9
+6458,63
+11468,3
+12564,76
+171,13
+10128,17
+11103,16
+12171,7
+285,13
+7717,71
+6847,53
+8517,59
+4927,26
+3145,10
+2188,8
+9437,19
+8682,5
+8784,20
+12106,53
+7129,7
+6716,67
+2291,60
+9988,24
+2379,75
+5882,25
+5051,26
+8308,76
+12004,70
+6426,63
+7593,65
+2106,8
+8528,59
+5135,45
+5635,27
+1675,49
+4101,73
+12311,35
+2377,75
+7452,35
+2192,8
+1103,12
+4739,28
+7524,35
+2036,47
+11540,44
+1744,49
+6503,70
+5827,51
+7515,35
+6215,55
+8696,5
+9185,0
+9842,69
+3405,6
+4479,22
+3431,6
+11262,38
+10399,14
+1179,4
+5724,51
+10804,15
+7540,35
+7606,65
+12830,57
+12883,19
+623,34
+10616,56
+7358,50
+10683,8
+9646,2
+3727,30
+7598,65
+1901,56
+2397,75
+10740,60
+11073,6
+11680,42
+1347,14
+10570,23
+3078,40
+10134,17
+2862,66
+3728,30
+3507,16
+8727,5
+6883,21
+4761,28
+11283,73
+8201,72
+9821,69
+9349,19
+7231,64
+5608,27
+6367,63
+13061,24
+12027,70
+1538,41
+11943,18
+12606,37
+6153,55
+11988,63
+5526,52
+6668,67
+10397,14
+10564,23
+3058,40
+11410,22
+7777,71
+5475,52
+3772,74
+9440,19
+722,34
+5249,45
+8912,20
+11317,73
+6921,21
+11175,74
+3919,68
+9682,2
+7156,7
+3571,16
+11953,18
+5269,42
+626,34
+761,46
+8172,72
+10331,4
+8158,43
+11477,3
+8219,76
+6909,21
+437,17
+4306,62
+10138,17
+6498,70
+9589,9
+11147,30
+12803,57
+185,13
+1823,56
+11964,63
+517,17
+7164,64
+12251,50
+12896,19
+19,11
+4888,44
+3508,16
+14,11
+10830,15
+5634,27
+4901,44
+3225,61
+11711,52
+956,36
+2176,8
+10410,33
+8943,20

finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=0/lora_ans_no_es/0.0-0.7/0.0-0.3/train_args.yaml

@@ -0,0 +1,10 @@
+K: null
+global_batch_size: 8
+learning_rate: 0.0001
+loss: ans
+max_steps: -1
+micro_batch_size: 1
+optimizer_name: adamw
+patience: 10
+val_check_interval: 16
+weight_decay: 0.0

finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=1/lora_ans/0.0-0.7/0.7-1.0/best.ckpt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:628d3831db10567a4d5ff02481fd7661e359cf5801b4be05284de1cecc89fe21
+size 42951934

finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=1/lora_ans/0.0-0.7/0.7-1.0/checkpoint/config.json

@@ -0,0 +1,27 @@
+{
+  "architectures": [
+    "Qwen2ForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "model_type": "qwen2",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "rms_norm_eps": 1e-06,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "tie_word_embeddings": false,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.43.1",
+  "use_cache": true,
+  "use_sliding_window": false,
+  "vocab_size": 152064
+}

finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=1/lora_ans/0.0-0.7/0.7-1.0/checkpoint/generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "pad_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "repetition_penalty": 1.05,
+  "temperature": 0.7,
+  "top_p": 0.8,
+  "top_k": 20,
+  "transformers_version": "4.37.0"
+}

finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=1/lora_ans/0.0-0.7/0.7-1.0/checkpoint/lit_model.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8f2893700d468552b27b00bb16e3906c7a01f49c09f3c786c252a2b5d36d179f
+size 15231297418

finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=1/lora_ans/0.0-0.7/0.7-1.0/checkpoint/lit_model.pth.lora

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:96b4534d2d3b82d6223c2627abed89d24a35d207b8c900128d7a5465c2b77769
+size 42951614

finetune_lora/qwen2.5-7b-instruct/banking77/size=64/seed=1/lora_ans/0.0-0.7/0.7-1.0/checkpoint/model_config.yaml

@@ -0,0 +1,37 @@
+attention_logit_softcapping: null
+attention_scores_scalar: null
+attn_bias: true
+bias: false
+block_size: 131072
+final_logit_softcapping: null
+gelu_approximate: none
+head_size: 128
+hf_config:
+  name: Qwen2.5-7B-Instruct
+  org: Qwen
+intermediate_size: 18944
+lm_head_bias: false
+mlp_class_name: LLaMAMLP
+n_embd: 3584
+n_expert: 0
+n_expert_per_token: 0
+n_head: 28
+n_layer: 28
+n_query_groups: 4
+name: Qwen2.5-7B-Instruct
+norm_class_name: RMSNorm
+norm_eps: 1.0e-06
+padded_vocab_size: 152064
+padding_multiple: 512
+parallel_residual: false
+post_attention_norm: false
+post_mlp_norm: false
+rope_adjustments: null
+rope_base: 1000000
+rope_condense_ratio: 1
+rotary_percentage: 1.0
+scale_embeddings: false
+shared_attention_norm: false
+sliding_window_layer_placing: null
+sliding_window_size: null
+vocab_size: 151643