{ "best_metric": 0.0909295603632927, "best_model_checkpoint": "vit_epochs5_batch32_lr5e-05_size224_tiles7_seed3_q3_dropout_v2/checkpoint-469", "epoch": 5.0, "eval_steps": 500, "global_step": 2345, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.010660980810234541, "grad_norm": 3.2466237545013428, "learning_rate": 4.989339019189766e-05, "loss": 0.1598, "step": 5 }, { "epoch": 0.021321961620469083, "grad_norm": 2.2757046222686768, "learning_rate": 4.978678038379531e-05, "loss": 0.0451, "step": 10 }, { "epoch": 0.031982942430703626, "grad_norm": 3.945024251937866, "learning_rate": 4.9680170575692967e-05, "loss": 0.1239, "step": 15 }, { "epoch": 0.042643923240938165, "grad_norm": 0.34108126163482666, "learning_rate": 4.957356076759062e-05, "loss": 0.0657, "step": 20 }, { "epoch": 0.053304904051172705, "grad_norm": 2.8693323135375977, "learning_rate": 4.9466950959488276e-05, "loss": 0.1324, "step": 25 }, { "epoch": 0.06396588486140725, "grad_norm": 1.6753228902816772, "learning_rate": 4.936034115138593e-05, "loss": 0.1054, "step": 30 }, { "epoch": 0.07462686567164178, "grad_norm": 3.1162660121917725, "learning_rate": 4.9253731343283586e-05, "loss": 0.0871, "step": 35 }, { "epoch": 0.08528784648187633, "grad_norm": 3.145806312561035, "learning_rate": 4.914712153518124e-05, "loss": 0.1381, "step": 40 }, { "epoch": 0.09594882729211088, "grad_norm": 3.556755304336548, "learning_rate": 4.904051172707889e-05, "loss": 0.1636, "step": 45 }, { "epoch": 0.10660980810234541, "grad_norm": 2.5963540077209473, "learning_rate": 4.893390191897655e-05, "loss": 0.0831, "step": 50 }, { "epoch": 0.11727078891257996, "grad_norm": 1.5497069358825684, "learning_rate": 4.88272921108742e-05, "loss": 0.1117, "step": 55 }, { "epoch": 0.1279317697228145, "grad_norm": 2.861238718032837, "learning_rate": 4.872068230277186e-05, "loss": 0.1124, "step": 60 }, { "epoch": 0.13859275053304904, "grad_norm": 0.7488825917243958, "learning_rate": 4.861407249466951e-05, "loss": 0.0953, "step": 65 }, { "epoch": 0.14925373134328357, "grad_norm": 1.254591464996338, "learning_rate": 4.850746268656717e-05, "loss": 0.074, "step": 70 }, { "epoch": 0.15991471215351813, "grad_norm": 2.7214035987854004, "learning_rate": 4.840085287846482e-05, "loss": 0.0931, "step": 75 }, { "epoch": 0.17057569296375266, "grad_norm": 0.1711389273405075, "learning_rate": 4.829424307036248e-05, "loss": 0.0891, "step": 80 }, { "epoch": 0.1812366737739872, "grad_norm": 5.988621711730957, "learning_rate": 4.8187633262260126e-05, "loss": 0.1111, "step": 85 }, { "epoch": 0.19189765458422176, "grad_norm": 1.670976161956787, "learning_rate": 4.808102345415779e-05, "loss": 0.0688, "step": 90 }, { "epoch": 0.2025586353944563, "grad_norm": 2.0074946880340576, "learning_rate": 4.7974413646055436e-05, "loss": 0.1074, "step": 95 }, { "epoch": 0.21321961620469082, "grad_norm": 6.26732873916626, "learning_rate": 4.78678038379531e-05, "loss": 0.0879, "step": 100 }, { "epoch": 0.22388059701492538, "grad_norm": 1.8915948867797852, "learning_rate": 4.7761194029850745e-05, "loss": 0.1087, "step": 105 }, { "epoch": 0.2345415778251599, "grad_norm": 4.518038272857666, "learning_rate": 4.765458422174841e-05, "loss": 0.0525, "step": 110 }, { "epoch": 0.24520255863539445, "grad_norm": 4.321690559387207, "learning_rate": 4.7547974413646055e-05, "loss": 0.1086, "step": 115 }, { "epoch": 0.255863539445629, "grad_norm": 1.0948755741119385, "learning_rate": 4.7441364605543716e-05, "loss": 0.1077, "step": 120 }, { "epoch": 0.26652452025586354, "grad_norm": 0.1039871796965599, "learning_rate": 4.7334754797441364e-05, "loss": 0.0252, "step": 125 }, { "epoch": 0.2771855010660981, "grad_norm": 1.5545969009399414, "learning_rate": 4.7228144989339026e-05, "loss": 0.0882, "step": 130 }, { "epoch": 0.2878464818763326, "grad_norm": 4.236177444458008, "learning_rate": 4.7121535181236674e-05, "loss": 0.0585, "step": 135 }, { "epoch": 0.29850746268656714, "grad_norm": 4.147879600524902, "learning_rate": 4.7014925373134335e-05, "loss": 0.0712, "step": 140 }, { "epoch": 0.3091684434968017, "grad_norm": 1.9625178575515747, "learning_rate": 4.690831556503198e-05, "loss": 0.0698, "step": 145 }, { "epoch": 0.31982942430703626, "grad_norm": 4.660637378692627, "learning_rate": 4.6801705756929645e-05, "loss": 0.0926, "step": 150 }, { "epoch": 0.3304904051172708, "grad_norm": 3.1571691036224365, "learning_rate": 4.669509594882729e-05, "loss": 0.0885, "step": 155 }, { "epoch": 0.3411513859275053, "grad_norm": 0.12366515398025513, "learning_rate": 4.658848614072495e-05, "loss": 0.0786, "step": 160 }, { "epoch": 0.35181236673773986, "grad_norm": 8.316474914550781, "learning_rate": 4.64818763326226e-05, "loss": 0.1273, "step": 165 }, { "epoch": 0.3624733475479744, "grad_norm": 0.920849621295929, "learning_rate": 4.637526652452026e-05, "loss": 0.0563, "step": 170 }, { "epoch": 0.373134328358209, "grad_norm": 2.456662893295288, "learning_rate": 4.626865671641791e-05, "loss": 0.1507, "step": 175 }, { "epoch": 0.3837953091684435, "grad_norm": 1.663049578666687, "learning_rate": 4.6162046908315566e-05, "loss": 0.0686, "step": 180 }, { "epoch": 0.39445628997867804, "grad_norm": 4.130738258361816, "learning_rate": 4.605543710021322e-05, "loss": 0.0758, "step": 185 }, { "epoch": 0.4051172707889126, "grad_norm": 0.3752882778644562, "learning_rate": 4.5948827292110876e-05, "loss": 0.156, "step": 190 }, { "epoch": 0.4157782515991471, "grad_norm": 0.5585218071937561, "learning_rate": 4.584221748400853e-05, "loss": 0.0933, "step": 195 }, { "epoch": 0.42643923240938164, "grad_norm": 2.0211336612701416, "learning_rate": 4.5735607675906185e-05, "loss": 0.0416, "step": 200 }, { "epoch": 0.43710021321961623, "grad_norm": 0.17304249107837677, "learning_rate": 4.562899786780384e-05, "loss": 0.0461, "step": 205 }, { "epoch": 0.44776119402985076, "grad_norm": 2.8942387104034424, "learning_rate": 4.5522388059701495e-05, "loss": 0.0452, "step": 210 }, { "epoch": 0.4584221748400853, "grad_norm": 0.815241813659668, "learning_rate": 4.541577825159915e-05, "loss": 0.1066, "step": 215 }, { "epoch": 0.4690831556503198, "grad_norm": 1.7086501121520996, "learning_rate": 4.5309168443496804e-05, "loss": 0.0973, "step": 220 }, { "epoch": 0.47974413646055436, "grad_norm": 3.537661075592041, "learning_rate": 4.520255863539446e-05, "loss": 0.067, "step": 225 }, { "epoch": 0.4904051172707889, "grad_norm": 3.7729461193084717, "learning_rate": 4.5095948827292114e-05, "loss": 0.0506, "step": 230 }, { "epoch": 0.5010660980810234, "grad_norm": 3.0081357955932617, "learning_rate": 4.498933901918977e-05, "loss": 0.0859, "step": 235 }, { "epoch": 0.511727078891258, "grad_norm": 1.8216296434402466, "learning_rate": 4.488272921108742e-05, "loss": 0.0263, "step": 240 }, { "epoch": 0.5223880597014925, "grad_norm": 2.64898419380188, "learning_rate": 4.477611940298508e-05, "loss": 0.0805, "step": 245 }, { "epoch": 0.5330490405117271, "grad_norm": 0.18348786234855652, "learning_rate": 4.466950959488273e-05, "loss": 0.0757, "step": 250 }, { "epoch": 0.5437100213219617, "grad_norm": 4.891566753387451, "learning_rate": 4.456289978678039e-05, "loss": 0.089, "step": 255 }, { "epoch": 0.5543710021321961, "grad_norm": 2.3679656982421875, "learning_rate": 4.445628997867804e-05, "loss": 0.059, "step": 260 }, { "epoch": 0.5650319829424307, "grad_norm": 2.391279935836792, "learning_rate": 4.43496801705757e-05, "loss": 0.0791, "step": 265 }, { "epoch": 0.5756929637526652, "grad_norm": 4.096751689910889, "learning_rate": 4.424307036247335e-05, "loss": 0.056, "step": 270 }, { "epoch": 0.5863539445628998, "grad_norm": 3.4114673137664795, "learning_rate": 4.4136460554371006e-05, "loss": 0.1029, "step": 275 }, { "epoch": 0.5970149253731343, "grad_norm": 2.888383388519287, "learning_rate": 4.402985074626866e-05, "loss": 0.0704, "step": 280 }, { "epoch": 0.6076759061833689, "grad_norm": 7.766504287719727, "learning_rate": 4.3923240938166316e-05, "loss": 0.0557, "step": 285 }, { "epoch": 0.6183368869936035, "grad_norm": 0.43232423067092896, "learning_rate": 4.381663113006397e-05, "loss": 0.0804, "step": 290 }, { "epoch": 0.6289978678038379, "grad_norm": 7.043906211853027, "learning_rate": 4.3710021321961625e-05, "loss": 0.0506, "step": 295 }, { "epoch": 0.6396588486140725, "grad_norm": 0.19176779687404633, "learning_rate": 4.360341151385928e-05, "loss": 0.0441, "step": 300 }, { "epoch": 0.650319829424307, "grad_norm": 10.689586639404297, "learning_rate": 4.3496801705756935e-05, "loss": 0.0592, "step": 305 }, { "epoch": 0.6609808102345416, "grad_norm": 1.1990717649459839, "learning_rate": 4.339019189765459e-05, "loss": 0.0276, "step": 310 }, { "epoch": 0.6716417910447762, "grad_norm": 4.2909064292907715, "learning_rate": 4.328358208955224e-05, "loss": 0.2166, "step": 315 }, { "epoch": 0.6823027718550106, "grad_norm": 0.43074071407318115, "learning_rate": 4.31769722814499e-05, "loss": 0.044, "step": 320 }, { "epoch": 0.6929637526652452, "grad_norm": 1.7041658163070679, "learning_rate": 4.307036247334755e-05, "loss": 0.128, "step": 325 }, { "epoch": 0.7036247334754797, "grad_norm": 0.7270706295967102, "learning_rate": 4.29637526652452e-05, "loss": 0.06, "step": 330 }, { "epoch": 0.7142857142857143, "grad_norm": 3.4168386459350586, "learning_rate": 4.2857142857142856e-05, "loss": 0.1149, "step": 335 }, { "epoch": 0.7249466950959488, "grad_norm": 1.9576493501663208, "learning_rate": 4.275053304904051e-05, "loss": 0.109, "step": 340 }, { "epoch": 0.7356076759061834, "grad_norm": 0.8163846135139465, "learning_rate": 4.2643923240938166e-05, "loss": 0.0863, "step": 345 }, { "epoch": 0.746268656716418, "grad_norm": 5.865867614746094, "learning_rate": 4.253731343283582e-05, "loss": 0.043, "step": 350 }, { "epoch": 0.7569296375266524, "grad_norm": 2.625739336013794, "learning_rate": 4.2430703624733475e-05, "loss": 0.0463, "step": 355 }, { "epoch": 0.767590618336887, "grad_norm": 0.488187313079834, "learning_rate": 4.232409381663113e-05, "loss": 0.0249, "step": 360 }, { "epoch": 0.7782515991471215, "grad_norm": 4.585363388061523, "learning_rate": 4.2217484008528785e-05, "loss": 0.1249, "step": 365 }, { "epoch": 0.7889125799573561, "grad_norm": 8.77475643157959, "learning_rate": 4.211087420042644e-05, "loss": 0.0768, "step": 370 }, { "epoch": 0.7995735607675906, "grad_norm": 1.645713448524475, "learning_rate": 4.2004264392324094e-05, "loss": 0.0664, "step": 375 }, { "epoch": 0.8102345415778252, "grad_norm": 1.4777451753616333, "learning_rate": 4.189765458422175e-05, "loss": 0.0676, "step": 380 }, { "epoch": 0.8208955223880597, "grad_norm": 6.782469749450684, "learning_rate": 4.1791044776119404e-05, "loss": 0.1109, "step": 385 }, { "epoch": 0.8315565031982942, "grad_norm": 6.149125576019287, "learning_rate": 4.168443496801706e-05, "loss": 0.1863, "step": 390 }, { "epoch": 0.8422174840085288, "grad_norm": 6.886513710021973, "learning_rate": 4.157782515991471e-05, "loss": 0.1473, "step": 395 }, { "epoch": 0.8528784648187633, "grad_norm": 0.4821918308734894, "learning_rate": 4.147121535181237e-05, "loss": 0.0431, "step": 400 }, { "epoch": 0.8635394456289979, "grad_norm": 5.310860633850098, "learning_rate": 4.136460554371002e-05, "loss": 0.0694, "step": 405 }, { "epoch": 0.8742004264392325, "grad_norm": 0.3728225529193878, "learning_rate": 4.125799573560768e-05, "loss": 0.0408, "step": 410 }, { "epoch": 0.8848614072494669, "grad_norm": 0.44299569725990295, "learning_rate": 4.115138592750533e-05, "loss": 0.1092, "step": 415 }, { "epoch": 0.8955223880597015, "grad_norm": 4.963627815246582, "learning_rate": 4.104477611940299e-05, "loss": 0.0794, "step": 420 }, { "epoch": 0.906183368869936, "grad_norm": 7.958082675933838, "learning_rate": 4.093816631130064e-05, "loss": 0.0868, "step": 425 }, { "epoch": 0.9168443496801706, "grad_norm": 2.6454625129699707, "learning_rate": 4.0831556503198296e-05, "loss": 0.0566, "step": 430 }, { "epoch": 0.9275053304904051, "grad_norm": 2.8813672065734863, "learning_rate": 4.072494669509595e-05, "loss": 0.0623, "step": 435 }, { "epoch": 0.9381663113006397, "grad_norm": 0.523626446723938, "learning_rate": 4.0618336886993606e-05, "loss": 0.0848, "step": 440 }, { "epoch": 0.9488272921108742, "grad_norm": 4.044142723083496, "learning_rate": 4.051172707889126e-05, "loss": 0.0681, "step": 445 }, { "epoch": 0.9594882729211087, "grad_norm": 1.8371537923812866, "learning_rate": 4.0405117270788915e-05, "loss": 0.046, "step": 450 }, { "epoch": 0.9701492537313433, "grad_norm": 3.6716814041137695, "learning_rate": 4.029850746268657e-05, "loss": 0.1216, "step": 455 }, { "epoch": 0.9808102345415778, "grad_norm": 1.372074842453003, "learning_rate": 4.0191897654584225e-05, "loss": 0.0333, "step": 460 }, { "epoch": 0.9914712153518124, "grad_norm": 7.915004730224609, "learning_rate": 4.008528784648188e-05, "loss": 0.0583, "step": 465 }, { "epoch": 1.0, "eval_accuracy": 0.9653333333333334, "eval_loss": 0.0909295603632927, "eval_runtime": 29.0701, "eval_samples_per_second": 128.999, "eval_steps_per_second": 4.059, "step": 469 }, { "epoch": 1.0021321961620469, "grad_norm": 0.4635755121707916, "learning_rate": 3.997867803837953e-05, "loss": 0.0578, "step": 470 }, { "epoch": 1.0127931769722816, "grad_norm": 1.7677839994430542, "learning_rate": 3.987206823027719e-05, "loss": 0.034, "step": 475 }, { "epoch": 1.023454157782516, "grad_norm": 0.08747611194849014, "learning_rate": 3.976545842217484e-05, "loss": 0.009, "step": 480 }, { "epoch": 1.0341151385927505, "grad_norm": 4.493967056274414, "learning_rate": 3.96588486140725e-05, "loss": 0.0093, "step": 485 }, { "epoch": 1.044776119402985, "grad_norm": 0.039758358150720596, "learning_rate": 3.9552238805970146e-05, "loss": 0.0121, "step": 490 }, { "epoch": 1.0554371002132197, "grad_norm": 0.21273668110370636, "learning_rate": 3.944562899786781e-05, "loss": 0.0242, "step": 495 }, { "epoch": 1.0660980810234542, "grad_norm": 2.2686619758605957, "learning_rate": 3.9339019189765456e-05, "loss": 0.0351, "step": 500 }, { "epoch": 1.0767590618336886, "grad_norm": 0.523878276348114, "learning_rate": 3.923240938166312e-05, "loss": 0.0127, "step": 505 }, { "epoch": 1.0874200426439233, "grad_norm": 0.06266263872385025, "learning_rate": 3.9125799573560765e-05, "loss": 0.0163, "step": 510 }, { "epoch": 1.0980810234541578, "grad_norm": 7.2656941413879395, "learning_rate": 3.901918976545843e-05, "loss": 0.0336, "step": 515 }, { "epoch": 1.1087420042643923, "grad_norm": 0.5467610955238342, "learning_rate": 3.8912579957356075e-05, "loss": 0.0207, "step": 520 }, { "epoch": 1.1194029850746268, "grad_norm": 7.460873603820801, "learning_rate": 3.8805970149253736e-05, "loss": 0.0526, "step": 525 }, { "epoch": 1.1300639658848615, "grad_norm": 0.05378120392560959, "learning_rate": 3.8699360341151384e-05, "loss": 0.0258, "step": 530 }, { "epoch": 1.140724946695096, "grad_norm": 0.09427474439144135, "learning_rate": 3.8592750533049046e-05, "loss": 0.0041, "step": 535 }, { "epoch": 1.1513859275053304, "grad_norm": 0.04945606738328934, "learning_rate": 3.8486140724946694e-05, "loss": 0.0335, "step": 540 }, { "epoch": 1.1620469083155651, "grad_norm": 0.4567449390888214, "learning_rate": 3.8379530916844355e-05, "loss": 0.0106, "step": 545 }, { "epoch": 1.1727078891257996, "grad_norm": 0.03384602069854736, "learning_rate": 3.8272921108742e-05, "loss": 0.0038, "step": 550 }, { "epoch": 1.183368869936034, "grad_norm": 1.0245600938796997, "learning_rate": 3.8166311300639665e-05, "loss": 0.0765, "step": 555 }, { "epoch": 1.1940298507462686, "grad_norm": 5.4920854568481445, "learning_rate": 3.805970149253731e-05, "loss": 0.0493, "step": 560 }, { "epoch": 1.2046908315565032, "grad_norm": 0.030089106410741806, "learning_rate": 3.7953091684434974e-05, "loss": 0.0413, "step": 565 }, { "epoch": 1.2153518123667377, "grad_norm": 0.20811010897159576, "learning_rate": 3.784648187633262e-05, "loss": 0.0051, "step": 570 }, { "epoch": 1.2260127931769722, "grad_norm": 0.027497880160808563, "learning_rate": 3.7739872068230284e-05, "loss": 0.0125, "step": 575 }, { "epoch": 1.236673773987207, "grad_norm": 1.7421993017196655, "learning_rate": 3.763326226012793e-05, "loss": 0.0057, "step": 580 }, { "epoch": 1.2473347547974414, "grad_norm": 0.03481829911470413, "learning_rate": 3.752665245202559e-05, "loss": 0.0034, "step": 585 }, { "epoch": 1.2579957356076759, "grad_norm": 2.799452066421509, "learning_rate": 3.742004264392324e-05, "loss": 0.0071, "step": 590 }, { "epoch": 1.2686567164179103, "grad_norm": 0.03254032880067825, "learning_rate": 3.73134328358209e-05, "loss": 0.0522, "step": 595 }, { "epoch": 1.279317697228145, "grad_norm": 10.43923282623291, "learning_rate": 3.720682302771855e-05, "loss": 0.0282, "step": 600 }, { "epoch": 1.2899786780383795, "grad_norm": 0.02785065956413746, "learning_rate": 3.710021321961621e-05, "loss": 0.005, "step": 605 }, { "epoch": 1.3006396588486142, "grad_norm": 0.02530420571565628, "learning_rate": 3.699360341151386e-05, "loss": 0.004, "step": 610 }, { "epoch": 1.3113006396588487, "grad_norm": 2.8372607231140137, "learning_rate": 3.6886993603411515e-05, "loss": 0.032, "step": 615 }, { "epoch": 1.3219616204690832, "grad_norm": 14.28488540649414, "learning_rate": 3.678038379530917e-05, "loss": 0.0393, "step": 620 }, { "epoch": 1.3326226012793176, "grad_norm": 0.02340145781636238, "learning_rate": 3.6673773987206824e-05, "loss": 0.0037, "step": 625 }, { "epoch": 1.3432835820895521, "grad_norm": 13.50255298614502, "learning_rate": 3.656716417910448e-05, "loss": 0.0632, "step": 630 }, { "epoch": 1.3539445628997868, "grad_norm": 8.260050773620605, "learning_rate": 3.6460554371002134e-05, "loss": 0.0112, "step": 635 }, { "epoch": 1.3646055437100213, "grad_norm": 0.027168547734618187, "learning_rate": 3.635394456289979e-05, "loss": 0.0232, "step": 640 }, { "epoch": 1.375266524520256, "grad_norm": 9.259380340576172, "learning_rate": 3.624733475479744e-05, "loss": 0.0722, "step": 645 }, { "epoch": 1.3859275053304905, "grad_norm": 0.033915404230356216, "learning_rate": 3.61407249466951e-05, "loss": 0.0351, "step": 650 }, { "epoch": 1.396588486140725, "grad_norm": 9.903706550598145, "learning_rate": 3.603411513859275e-05, "loss": 0.038, "step": 655 }, { "epoch": 1.4072494669509594, "grad_norm": 0.022436009719967842, "learning_rate": 3.592750533049041e-05, "loss": 0.0331, "step": 660 }, { "epoch": 1.417910447761194, "grad_norm": 0.05081154778599739, "learning_rate": 3.582089552238806e-05, "loss": 0.0061, "step": 665 }, { "epoch": 1.4285714285714286, "grad_norm": 0.02403384819626808, "learning_rate": 3.571428571428572e-05, "loss": 0.0796, "step": 670 }, { "epoch": 1.439232409381663, "grad_norm": 0.023996546864509583, "learning_rate": 3.560767590618337e-05, "loss": 0.0834, "step": 675 }, { "epoch": 1.4498933901918978, "grad_norm": 0.025445258244872093, "learning_rate": 3.5501066098081026e-05, "loss": 0.0266, "step": 680 }, { "epoch": 1.4605543710021323, "grad_norm": 14.832476615905762, "learning_rate": 3.539445628997868e-05, "loss": 0.0585, "step": 685 }, { "epoch": 1.4712153518123667, "grad_norm": 0.04287252202630043, "learning_rate": 3.5287846481876336e-05, "loss": 0.0276, "step": 690 }, { "epoch": 1.4818763326226012, "grad_norm": 0.3106297552585602, "learning_rate": 3.518123667377399e-05, "loss": 0.0425, "step": 695 }, { "epoch": 1.4925373134328357, "grad_norm": 0.824328601360321, "learning_rate": 3.5074626865671645e-05, "loss": 0.0308, "step": 700 }, { "epoch": 1.5031982942430704, "grad_norm": 0.028267953544855118, "learning_rate": 3.496801705756929e-05, "loss": 0.0078, "step": 705 }, { "epoch": 1.5138592750533049, "grad_norm": 3.530308485031128, "learning_rate": 3.4861407249466955e-05, "loss": 0.0333, "step": 710 }, { "epoch": 1.5245202558635396, "grad_norm": 0.7099125981330872, "learning_rate": 3.47547974413646e-05, "loss": 0.004, "step": 715 }, { "epoch": 1.535181236673774, "grad_norm": 0.027877703309059143, "learning_rate": 3.4648187633262264e-05, "loss": 0.0317, "step": 720 }, { "epoch": 1.5458422174840085, "grad_norm": 0.5809793472290039, "learning_rate": 3.454157782515991e-05, "loss": 0.019, "step": 725 }, { "epoch": 1.556503198294243, "grad_norm": 4.466644763946533, "learning_rate": 3.4434968017057574e-05, "loss": 0.0785, "step": 730 }, { "epoch": 1.5671641791044775, "grad_norm": 1.0263413190841675, "learning_rate": 3.432835820895522e-05, "loss": 0.0413, "step": 735 }, { "epoch": 1.5778251599147122, "grad_norm": 0.030618691816926003, "learning_rate": 3.422174840085288e-05, "loss": 0.0075, "step": 740 }, { "epoch": 1.5884861407249466, "grad_norm": 3.667579412460327, "learning_rate": 3.411513859275053e-05, "loss": 0.0061, "step": 745 }, { "epoch": 1.5991471215351813, "grad_norm": 0.135724738240242, "learning_rate": 3.400852878464819e-05, "loss": 0.0044, "step": 750 }, { "epoch": 1.6098081023454158, "grad_norm": 0.2659117579460144, "learning_rate": 3.390191897654584e-05, "loss": 0.004, "step": 755 }, { "epoch": 1.6204690831556503, "grad_norm": 0.3448447585105896, "learning_rate": 3.37953091684435e-05, "loss": 0.0555, "step": 760 }, { "epoch": 1.6311300639658848, "grad_norm": 0.2710099220275879, "learning_rate": 3.368869936034115e-05, "loss": 0.0041, "step": 765 }, { "epoch": 1.6417910447761193, "grad_norm": 3.1645729541778564, "learning_rate": 3.358208955223881e-05, "loss": 0.0322, "step": 770 }, { "epoch": 1.652452025586354, "grad_norm": 0.022952351719141006, "learning_rate": 3.347547974413646e-05, "loss": 0.0055, "step": 775 }, { "epoch": 1.6631130063965884, "grad_norm": 12.804929733276367, "learning_rate": 3.336886993603412e-05, "loss": 0.0201, "step": 780 }, { "epoch": 1.6737739872068231, "grad_norm": 0.025945674628019333, "learning_rate": 3.326226012793177e-05, "loss": 0.0026, "step": 785 }, { "epoch": 1.6844349680170576, "grad_norm": 3.4994208812713623, "learning_rate": 3.3155650319829424e-05, "loss": 0.0092, "step": 790 }, { "epoch": 1.695095948827292, "grad_norm": 0.022479137405753136, "learning_rate": 3.304904051172708e-05, "loss": 0.0339, "step": 795 }, { "epoch": 1.7057569296375266, "grad_norm": 1.1224740743637085, "learning_rate": 3.294243070362473e-05, "loss": 0.0479, "step": 800 }, { "epoch": 1.716417910447761, "grad_norm": 0.0518471822142601, "learning_rate": 3.283582089552239e-05, "loss": 0.0377, "step": 805 }, { "epoch": 1.7270788912579957, "grad_norm": 7.18663215637207, "learning_rate": 3.272921108742004e-05, "loss": 0.0203, "step": 810 }, { "epoch": 1.7377398720682304, "grad_norm": 3.9707489013671875, "learning_rate": 3.26226012793177e-05, "loss": 0.0416, "step": 815 }, { "epoch": 1.748400852878465, "grad_norm": 0.027207937091588974, "learning_rate": 3.251599147121535e-05, "loss": 0.0025, "step": 820 }, { "epoch": 1.7590618336886994, "grad_norm": 0.02161501534283161, "learning_rate": 3.240938166311301e-05, "loss": 0.0071, "step": 825 }, { "epoch": 1.7697228144989339, "grad_norm": 9.111710548400879, "learning_rate": 3.230277185501066e-05, "loss": 0.0172, "step": 830 }, { "epoch": 1.7803837953091683, "grad_norm": 0.10926713049411774, "learning_rate": 3.2196162046908317e-05, "loss": 0.0242, "step": 835 }, { "epoch": 1.7910447761194028, "grad_norm": 0.028723997995257378, "learning_rate": 3.208955223880597e-05, "loss": 0.0054, "step": 840 }, { "epoch": 1.8017057569296375, "grad_norm": 0.22235746681690216, "learning_rate": 3.1982942430703626e-05, "loss": 0.0041, "step": 845 }, { "epoch": 1.8123667377398722, "grad_norm": 0.024353526532649994, "learning_rate": 3.187633262260128e-05, "loss": 0.041, "step": 850 }, { "epoch": 1.8230277185501067, "grad_norm": 0.08913513273000717, "learning_rate": 3.1769722814498935e-05, "loss": 0.0065, "step": 855 }, { "epoch": 1.8336886993603412, "grad_norm": 0.043564602732658386, "learning_rate": 3.166311300639659e-05, "loss": 0.01, "step": 860 }, { "epoch": 1.8443496801705757, "grad_norm": 0.04333405941724777, "learning_rate": 3.1556503198294245e-05, "loss": 0.0086, "step": 865 }, { "epoch": 1.8550106609808101, "grad_norm": 0.02290233224630356, "learning_rate": 3.14498933901919e-05, "loss": 0.0062, "step": 870 }, { "epoch": 1.8656716417910446, "grad_norm": 0.32433441281318665, "learning_rate": 3.1343283582089554e-05, "loss": 0.0751, "step": 875 }, { "epoch": 1.8763326226012793, "grad_norm": 1.6093181371688843, "learning_rate": 3.123667377398721e-05, "loss": 0.0042, "step": 880 }, { "epoch": 1.886993603411514, "grad_norm": 11.231916427612305, "learning_rate": 3.1130063965884864e-05, "loss": 0.0354, "step": 885 }, { "epoch": 1.8976545842217485, "grad_norm": 0.02130195125937462, "learning_rate": 3.102345415778252e-05, "loss": 0.02, "step": 890 }, { "epoch": 1.908315565031983, "grad_norm": 0.020609725266695023, "learning_rate": 3.0916844349680173e-05, "loss": 0.006, "step": 895 }, { "epoch": 1.9189765458422174, "grad_norm": 4.776242733001709, "learning_rate": 3.081023454157783e-05, "loss": 0.008, "step": 900 }, { "epoch": 1.929637526652452, "grad_norm": 0.020240770652890205, "learning_rate": 3.070362473347548e-05, "loss": 0.0027, "step": 905 }, { "epoch": 1.9402985074626866, "grad_norm": 0.026981867849826813, "learning_rate": 3.059701492537314e-05, "loss": 0.0031, "step": 910 }, { "epoch": 1.950959488272921, "grad_norm": 3.7452125549316406, "learning_rate": 3.0490405117270792e-05, "loss": 0.0121, "step": 915 }, { "epoch": 1.9616204690831558, "grad_norm": 0.022721681743860245, "learning_rate": 3.0383795309168444e-05, "loss": 0.0225, "step": 920 }, { "epoch": 1.9722814498933903, "grad_norm": 0.03232937306165695, "learning_rate": 3.0277185501066102e-05, "loss": 0.012, "step": 925 }, { "epoch": 1.9829424307036247, "grad_norm": 0.016798535361886024, "learning_rate": 3.0170575692963753e-05, "loss": 0.0552, "step": 930 }, { "epoch": 1.9936034115138592, "grad_norm": 0.022736212238669395, "learning_rate": 3.006396588486141e-05, "loss": 0.0278, "step": 935 }, { "epoch": 2.0, "eval_accuracy": 0.972, "eval_loss": 0.10947595536708832, "eval_runtime": 30.9166, "eval_samples_per_second": 121.294, "eval_steps_per_second": 3.817, "step": 938 }, { "epoch": 2.0042643923240937, "grad_norm": 0.06007245555520058, "learning_rate": 2.9957356076759063e-05, "loss": 0.0338, "step": 940 }, { "epoch": 2.014925373134328, "grad_norm": 0.04219633340835571, "learning_rate": 2.9850746268656714e-05, "loss": 0.0021, "step": 945 }, { "epoch": 2.025586353944563, "grad_norm": 3.861257314682007, "learning_rate": 2.9744136460554372e-05, "loss": 0.01, "step": 950 }, { "epoch": 2.0362473347547976, "grad_norm": 0.016493400558829308, "learning_rate": 2.9637526652452023e-05, "loss": 0.0331, "step": 955 }, { "epoch": 2.046908315565032, "grad_norm": 0.052993178367614746, "learning_rate": 2.953091684434968e-05, "loss": 0.0035, "step": 960 }, { "epoch": 2.0575692963752665, "grad_norm": 0.029567304998636246, "learning_rate": 2.9424307036247333e-05, "loss": 0.0021, "step": 965 }, { "epoch": 2.068230277185501, "grad_norm": 0.03378676995635033, "learning_rate": 2.931769722814499e-05, "loss": 0.005, "step": 970 }, { "epoch": 2.0788912579957355, "grad_norm": 0.04222923889756203, "learning_rate": 2.9211087420042642e-05, "loss": 0.0022, "step": 975 }, { "epoch": 2.08955223880597, "grad_norm": 0.014676189981400967, "learning_rate": 2.91044776119403e-05, "loss": 0.0022, "step": 980 }, { "epoch": 2.100213219616205, "grad_norm": 0.015069925226271152, "learning_rate": 2.8997867803837952e-05, "loss": 0.0021, "step": 985 }, { "epoch": 2.1108742004264394, "grad_norm": 0.020251786336302757, "learning_rate": 2.889125799573561e-05, "loss": 0.0026, "step": 990 }, { "epoch": 2.121535181236674, "grad_norm": 0.05129149183630943, "learning_rate": 2.878464818763326e-05, "loss": 0.0018, "step": 995 }, { "epoch": 2.1321961620469083, "grad_norm": 0.017302637919783592, "learning_rate": 2.867803837953092e-05, "loss": 0.0022, "step": 1000 }, { "epoch": 2.142857142857143, "grad_norm": 0.013412648811936378, "learning_rate": 2.857142857142857e-05, "loss": 0.0048, "step": 1005 }, { "epoch": 2.1535181236673773, "grad_norm": 0.13972871005535126, "learning_rate": 2.846481876332623e-05, "loss": 0.0018, "step": 1010 }, { "epoch": 2.1641791044776117, "grad_norm": 0.0254370104521513, "learning_rate": 2.835820895522388e-05, "loss": 0.0072, "step": 1015 }, { "epoch": 2.1748400852878467, "grad_norm": 0.014770382083952427, "learning_rate": 2.825159914712154e-05, "loss": 0.0081, "step": 1020 }, { "epoch": 2.185501066098081, "grad_norm": 0.03373216465115547, "learning_rate": 2.814498933901919e-05, "loss": 0.0016, "step": 1025 }, { "epoch": 2.1961620469083156, "grad_norm": 0.01940246671438217, "learning_rate": 2.8038379530916848e-05, "loss": 0.0023, "step": 1030 }, { "epoch": 2.20682302771855, "grad_norm": 0.013438255526125431, "learning_rate": 2.79317697228145e-05, "loss": 0.0504, "step": 1035 }, { "epoch": 2.2174840085287846, "grad_norm": 0.015464778058230877, "learning_rate": 2.7825159914712157e-05, "loss": 0.0076, "step": 1040 }, { "epoch": 2.228144989339019, "grad_norm": 0.014380119740962982, "learning_rate": 2.771855010660981e-05, "loss": 0.002, "step": 1045 }, { "epoch": 2.2388059701492535, "grad_norm": 0.024420637637376785, "learning_rate": 2.7611940298507467e-05, "loss": 0.0017, "step": 1050 }, { "epoch": 2.2494669509594885, "grad_norm": 0.128773033618927, "learning_rate": 2.7505330490405118e-05, "loss": 0.002, "step": 1055 }, { "epoch": 2.260127931769723, "grad_norm": 0.014405138790607452, "learning_rate": 2.7398720682302776e-05, "loss": 0.0027, "step": 1060 }, { "epoch": 2.2707889125799574, "grad_norm": 0.013299096375703812, "learning_rate": 2.7292110874200428e-05, "loss": 0.0409, "step": 1065 }, { "epoch": 2.281449893390192, "grad_norm": 0.06791973114013672, "learning_rate": 2.7185501066098086e-05, "loss": 0.0021, "step": 1070 }, { "epoch": 2.2921108742004264, "grad_norm": 0.02083703875541687, "learning_rate": 2.7078891257995737e-05, "loss": 0.0054, "step": 1075 }, { "epoch": 2.302771855010661, "grad_norm": 0.08183146268129349, "learning_rate": 2.6972281449893395e-05, "loss": 0.0015, "step": 1080 }, { "epoch": 2.3134328358208958, "grad_norm": 0.01852445863187313, "learning_rate": 2.6865671641791047e-05, "loss": 0.0175, "step": 1085 }, { "epoch": 2.3240938166311302, "grad_norm": 0.012970099225640297, "learning_rate": 2.6759061833688705e-05, "loss": 0.0017, "step": 1090 }, { "epoch": 2.3347547974413647, "grad_norm": 0.02032940648496151, "learning_rate": 2.6652452025586356e-05, "loss": 0.0014, "step": 1095 }, { "epoch": 2.345415778251599, "grad_norm": 0.07824794203042984, "learning_rate": 2.6545842217484007e-05, "loss": 0.0066, "step": 1100 }, { "epoch": 2.3560767590618337, "grad_norm": 0.10589689016342163, "learning_rate": 2.6439232409381666e-05, "loss": 0.0015, "step": 1105 }, { "epoch": 2.366737739872068, "grad_norm": 0.014430535957217216, "learning_rate": 2.6332622601279317e-05, "loss": 0.0016, "step": 1110 }, { "epoch": 2.3773987206823026, "grad_norm": 0.011324513703584671, "learning_rate": 2.6226012793176975e-05, "loss": 0.0015, "step": 1115 }, { "epoch": 2.388059701492537, "grad_norm": 0.0126158706843853, "learning_rate": 2.6119402985074626e-05, "loss": 0.0015, "step": 1120 }, { "epoch": 2.398720682302772, "grad_norm": 4.089925289154053, "learning_rate": 2.6012793176972285e-05, "loss": 0.0081, "step": 1125 }, { "epoch": 2.4093816631130065, "grad_norm": 0.011434156447649002, "learning_rate": 2.5906183368869936e-05, "loss": 0.0014, "step": 1130 }, { "epoch": 2.420042643923241, "grad_norm": 0.011946013197302818, "learning_rate": 2.5799573560767594e-05, "loss": 0.0014, "step": 1135 }, { "epoch": 2.4307036247334755, "grad_norm": 0.01641424186527729, "learning_rate": 2.5692963752665245e-05, "loss": 0.0018, "step": 1140 }, { "epoch": 2.44136460554371, "grad_norm": 0.011022225022315979, "learning_rate": 2.5586353944562904e-05, "loss": 0.0012, "step": 1145 }, { "epoch": 2.4520255863539444, "grad_norm": 0.012972844764590263, "learning_rate": 2.5479744136460555e-05, "loss": 0.0017, "step": 1150 }, { "epoch": 2.4626865671641793, "grad_norm": 0.011477009393274784, "learning_rate": 2.537313432835821e-05, "loss": 0.0387, "step": 1155 }, { "epoch": 2.473347547974414, "grad_norm": 0.015750903636217117, "learning_rate": 2.5266524520255864e-05, "loss": 0.0015, "step": 1160 }, { "epoch": 2.4840085287846483, "grad_norm": 0.01056817825883627, "learning_rate": 2.515991471215352e-05, "loss": 0.0013, "step": 1165 }, { "epoch": 2.4946695095948828, "grad_norm": 0.010986260138452053, "learning_rate": 2.5053304904051174e-05, "loss": 0.0012, "step": 1170 }, { "epoch": 2.5053304904051172, "grad_norm": 0.010363342240452766, "learning_rate": 2.494669509594883e-05, "loss": 0.0324, "step": 1175 }, { "epoch": 2.5159914712153517, "grad_norm": 1.7450491189956665, "learning_rate": 2.4840085287846483e-05, "loss": 0.0026, "step": 1180 }, { "epoch": 2.526652452025586, "grad_norm": 0.01395554468035698, "learning_rate": 2.4733475479744138e-05, "loss": 0.0371, "step": 1185 }, { "epoch": 2.5373134328358207, "grad_norm": 0.015563605353236198, "learning_rate": 2.4626865671641793e-05, "loss": 0.0016, "step": 1190 }, { "epoch": 2.5479744136460556, "grad_norm": 0.011689956299960613, "learning_rate": 2.4520255863539444e-05, "loss": 0.0021, "step": 1195 }, { "epoch": 2.55863539445629, "grad_norm": 0.009155909530818462, "learning_rate": 2.44136460554371e-05, "loss": 0.0014, "step": 1200 }, { "epoch": 2.5692963752665245, "grad_norm": 0.009336567483842373, "learning_rate": 2.4307036247334754e-05, "loss": 0.0021, "step": 1205 }, { "epoch": 2.579957356076759, "grad_norm": 0.01069782767444849, "learning_rate": 2.420042643923241e-05, "loss": 0.0014, "step": 1210 }, { "epoch": 2.5906183368869935, "grad_norm": 0.40958771109580994, "learning_rate": 2.4093816631130063e-05, "loss": 0.0051, "step": 1215 }, { "epoch": 2.6012793176972284, "grad_norm": 0.0775865837931633, "learning_rate": 2.3987206823027718e-05, "loss": 0.0015, "step": 1220 }, { "epoch": 2.611940298507463, "grad_norm": 0.013319520279765129, "learning_rate": 2.3880597014925373e-05, "loss": 0.0013, "step": 1225 }, { "epoch": 2.6226012793176974, "grad_norm": 0.010163035243749619, "learning_rate": 2.3773987206823027e-05, "loss": 0.0016, "step": 1230 }, { "epoch": 2.633262260127932, "grad_norm": 0.020594105124473572, "learning_rate": 2.3667377398720682e-05, "loss": 0.0063, "step": 1235 }, { "epoch": 2.6439232409381663, "grad_norm": 0.010080802254378796, "learning_rate": 2.3560767590618337e-05, "loss": 0.0012, "step": 1240 }, { "epoch": 2.654584221748401, "grad_norm": 0.03885523974895477, "learning_rate": 2.345415778251599e-05, "loss": 0.0013, "step": 1245 }, { "epoch": 2.6652452025586353, "grad_norm": 0.012037910521030426, "learning_rate": 2.3347547974413646e-05, "loss": 0.0012, "step": 1250 }, { "epoch": 2.6759061833688698, "grad_norm": 0.015458466485142708, "learning_rate": 2.32409381663113e-05, "loss": 0.0011, "step": 1255 }, { "epoch": 2.6865671641791042, "grad_norm": 0.014528359286487103, "learning_rate": 2.3134328358208956e-05, "loss": 0.0021, "step": 1260 }, { "epoch": 2.697228144989339, "grad_norm": 0.014770982787013054, "learning_rate": 2.302771855010661e-05, "loss": 0.0067, "step": 1265 }, { "epoch": 2.7078891257995736, "grad_norm": 0.012316780164837837, "learning_rate": 2.2921108742004265e-05, "loss": 0.0011, "step": 1270 }, { "epoch": 2.718550106609808, "grad_norm": 0.02276039682328701, "learning_rate": 2.281449893390192e-05, "loss": 0.0036, "step": 1275 }, { "epoch": 2.7292110874200426, "grad_norm": 0.04132280498743057, "learning_rate": 2.2707889125799575e-05, "loss": 0.0032, "step": 1280 }, { "epoch": 2.739872068230277, "grad_norm": 0.010015731677412987, "learning_rate": 2.260127931769723e-05, "loss": 0.0011, "step": 1285 }, { "epoch": 2.750533049040512, "grad_norm": 0.008909180760383606, "learning_rate": 2.2494669509594884e-05, "loss": 0.0224, "step": 1290 }, { "epoch": 2.7611940298507465, "grad_norm": 0.020119719207286835, "learning_rate": 2.238805970149254e-05, "loss": 0.0032, "step": 1295 }, { "epoch": 2.771855010660981, "grad_norm": 8.630078315734863, "learning_rate": 2.2281449893390194e-05, "loss": 0.0335, "step": 1300 }, { "epoch": 2.7825159914712154, "grad_norm": 0.013341797515749931, "learning_rate": 2.217484008528785e-05, "loss": 0.0017, "step": 1305 }, { "epoch": 2.79317697228145, "grad_norm": 0.009243636392056942, "learning_rate": 2.2068230277185503e-05, "loss": 0.0012, "step": 1310 }, { "epoch": 2.8038379530916844, "grad_norm": 0.009263358078897, "learning_rate": 2.1961620469083158e-05, "loss": 0.0014, "step": 1315 }, { "epoch": 2.814498933901919, "grad_norm": 0.009164247661828995, "learning_rate": 2.1855010660980813e-05, "loss": 0.0143, "step": 1320 }, { "epoch": 2.8251599147121533, "grad_norm": 0.01843312196433544, "learning_rate": 2.1748400852878467e-05, "loss": 0.0024, "step": 1325 }, { "epoch": 2.835820895522388, "grad_norm": 0.009804385714232922, "learning_rate": 2.164179104477612e-05, "loss": 0.0015, "step": 1330 }, { "epoch": 2.8464818763326227, "grad_norm": 0.011213825084269047, "learning_rate": 2.1535181236673773e-05, "loss": 0.0011, "step": 1335 }, { "epoch": 2.857142857142857, "grad_norm": 0.118233822286129, "learning_rate": 2.1428571428571428e-05, "loss": 0.0031, "step": 1340 }, { "epoch": 2.8678038379530917, "grad_norm": 0.009663242846727371, "learning_rate": 2.1321961620469083e-05, "loss": 0.0038, "step": 1345 }, { "epoch": 2.878464818763326, "grad_norm": 0.009211266413331032, "learning_rate": 2.1215351812366738e-05, "loss": 0.0076, "step": 1350 }, { "epoch": 2.8891257995735606, "grad_norm": 0.008291327394545078, "learning_rate": 2.1108742004264392e-05, "loss": 0.024, "step": 1355 }, { "epoch": 2.8997867803837956, "grad_norm": 5.672647953033447, "learning_rate": 2.1002132196162047e-05, "loss": 0.0034, "step": 1360 }, { "epoch": 2.91044776119403, "grad_norm": 0.009524866007268429, "learning_rate": 2.0895522388059702e-05, "loss": 0.001, "step": 1365 }, { "epoch": 2.9211087420042645, "grad_norm": 0.009426102973520756, "learning_rate": 2.0788912579957357e-05, "loss": 0.0011, "step": 1370 }, { "epoch": 2.931769722814499, "grad_norm": 0.943455696105957, "learning_rate": 2.068230277185501e-05, "loss": 0.0014, "step": 1375 }, { "epoch": 2.9424307036247335, "grad_norm": 0.007855545729398727, "learning_rate": 2.0575692963752666e-05, "loss": 0.0011, "step": 1380 }, { "epoch": 2.953091684434968, "grad_norm": 0.007699600420892239, "learning_rate": 2.046908315565032e-05, "loss": 0.0009, "step": 1385 }, { "epoch": 2.9637526652452024, "grad_norm": 0.08365332335233688, "learning_rate": 2.0362473347547976e-05, "loss": 0.0455, "step": 1390 }, { "epoch": 2.974413646055437, "grad_norm": 0.00963526964187622, "learning_rate": 2.025586353944563e-05, "loss": 0.0009, "step": 1395 }, { "epoch": 2.9850746268656714, "grad_norm": 0.24307560920715332, "learning_rate": 2.0149253731343285e-05, "loss": 0.0011, "step": 1400 }, { "epoch": 2.9957356076759063, "grad_norm": 2.380711078643799, "learning_rate": 2.004264392324094e-05, "loss": 0.002, "step": 1405 }, { "epoch": 3.0, "eval_accuracy": 0.9746666666666667, "eval_loss": 0.11213460564613342, "eval_runtime": 28.8728, "eval_samples_per_second": 129.88, "eval_steps_per_second": 4.087, "step": 1407 }, { "epoch": 3.0063965884861408, "grad_norm": 0.008787929080426693, "learning_rate": 1.9936034115138594e-05, "loss": 0.001, "step": 1410 }, { "epoch": 3.0170575692963753, "grad_norm": 0.008867541328072548, "learning_rate": 1.982942430703625e-05, "loss": 0.0009, "step": 1415 }, { "epoch": 3.0277185501066097, "grad_norm": 0.008270635269582272, "learning_rate": 1.9722814498933904e-05, "loss": 0.001, "step": 1420 }, { "epoch": 3.038379530916844, "grad_norm": 0.012295828200876713, "learning_rate": 1.961620469083156e-05, "loss": 0.001, "step": 1425 }, { "epoch": 3.0490405117270787, "grad_norm": 4.370051383972168, "learning_rate": 1.9509594882729213e-05, "loss": 0.0028, "step": 1430 }, { "epoch": 3.0597014925373136, "grad_norm": 0.007836264558136463, "learning_rate": 1.9402985074626868e-05, "loss": 0.0009, "step": 1435 }, { "epoch": 3.070362473347548, "grad_norm": 0.00735406344756484, "learning_rate": 1.9296375266524523e-05, "loss": 0.0009, "step": 1440 }, { "epoch": 3.0810234541577826, "grad_norm": 0.029614465311169624, "learning_rate": 1.9189765458422178e-05, "loss": 0.001, "step": 1445 }, { "epoch": 3.091684434968017, "grad_norm": 0.008709880523383617, "learning_rate": 1.9083155650319832e-05, "loss": 0.0021, "step": 1450 }, { "epoch": 3.1023454157782515, "grad_norm": 0.008091866038739681, "learning_rate": 1.8976545842217487e-05, "loss": 0.0009, "step": 1455 }, { "epoch": 3.113006396588486, "grad_norm": 0.007227215450257063, "learning_rate": 1.8869936034115142e-05, "loss": 0.0009, "step": 1460 }, { "epoch": 3.1236673773987205, "grad_norm": 0.00812044832855463, "learning_rate": 1.8763326226012797e-05, "loss": 0.0009, "step": 1465 }, { "epoch": 3.1343283582089554, "grad_norm": 0.20463325083255768, "learning_rate": 1.865671641791045e-05, "loss": 0.0012, "step": 1470 }, { "epoch": 3.14498933901919, "grad_norm": 0.01566215418279171, "learning_rate": 1.8550106609808106e-05, "loss": 0.001, "step": 1475 }, { "epoch": 3.1556503198294243, "grad_norm": 0.009645018726587296, "learning_rate": 1.8443496801705757e-05, "loss": 0.0009, "step": 1480 }, { "epoch": 3.166311300639659, "grad_norm": 0.008649914525449276, "learning_rate": 1.8336886993603412e-05, "loss": 0.0009, "step": 1485 }, { "epoch": 3.1769722814498933, "grad_norm": 0.008244159631431103, "learning_rate": 1.8230277185501067e-05, "loss": 0.0009, "step": 1490 }, { "epoch": 3.1876332622601278, "grad_norm": 0.009652147069573402, "learning_rate": 1.812366737739872e-05, "loss": 0.0044, "step": 1495 }, { "epoch": 3.1982942430703627, "grad_norm": 0.00760348467156291, "learning_rate": 1.8017057569296376e-05, "loss": 0.0009, "step": 1500 }, { "epoch": 3.208955223880597, "grad_norm": 0.014630252495408058, "learning_rate": 1.791044776119403e-05, "loss": 0.0013, "step": 1505 }, { "epoch": 3.2196162046908317, "grad_norm": 0.00809138361364603, "learning_rate": 1.7803837953091686e-05, "loss": 0.0008, "step": 1510 }, { "epoch": 3.230277185501066, "grad_norm": 0.008098193444311619, "learning_rate": 1.769722814498934e-05, "loss": 0.0009, "step": 1515 }, { "epoch": 3.2409381663113006, "grad_norm": 0.007939455099403858, "learning_rate": 1.7590618336886995e-05, "loss": 0.0008, "step": 1520 }, { "epoch": 3.251599147121535, "grad_norm": 0.007237904239445925, "learning_rate": 1.7484008528784647e-05, "loss": 0.0008, "step": 1525 }, { "epoch": 3.2622601279317696, "grad_norm": 0.0077867875806987286, "learning_rate": 1.73773987206823e-05, "loss": 0.0023, "step": 1530 }, { "epoch": 3.272921108742004, "grad_norm": 0.07109486311674118, "learning_rate": 1.7270788912579956e-05, "loss": 0.001, "step": 1535 }, { "epoch": 3.283582089552239, "grad_norm": 0.007528245914727449, "learning_rate": 1.716417910447761e-05, "loss": 0.0009, "step": 1540 }, { "epoch": 3.2942430703624734, "grad_norm": 0.01986989937722683, "learning_rate": 1.7057569296375266e-05, "loss": 0.0009, "step": 1545 }, { "epoch": 3.304904051172708, "grad_norm": 0.0075040943920612335, "learning_rate": 1.695095948827292e-05, "loss": 0.0008, "step": 1550 }, { "epoch": 3.3155650319829424, "grad_norm": 0.006988045293837786, "learning_rate": 1.6844349680170575e-05, "loss": 0.0008, "step": 1555 }, { "epoch": 3.326226012793177, "grad_norm": 0.006291010417044163, "learning_rate": 1.673773987206823e-05, "loss": 0.0007, "step": 1560 }, { "epoch": 3.3368869936034113, "grad_norm": 0.007796446327120066, "learning_rate": 1.6631130063965885e-05, "loss": 0.0008, "step": 1565 }, { "epoch": 3.3475479744136463, "grad_norm": 0.006688463501632214, "learning_rate": 1.652452025586354e-05, "loss": 0.0008, "step": 1570 }, { "epoch": 3.3582089552238807, "grad_norm": 0.008288230746984482, "learning_rate": 1.6417910447761194e-05, "loss": 0.0009, "step": 1575 }, { "epoch": 3.368869936034115, "grad_norm": 0.007838346064090729, "learning_rate": 1.631130063965885e-05, "loss": 0.0009, "step": 1580 }, { "epoch": 3.3795309168443497, "grad_norm": 0.006377186626195908, "learning_rate": 1.6204690831556504e-05, "loss": 0.0008, "step": 1585 }, { "epoch": 3.390191897654584, "grad_norm": 0.007430005352944136, "learning_rate": 1.6098081023454158e-05, "loss": 0.0008, "step": 1590 }, { "epoch": 3.4008528784648187, "grad_norm": 0.0068603018298745155, "learning_rate": 1.5991471215351813e-05, "loss": 0.0008, "step": 1595 }, { "epoch": 3.411513859275053, "grad_norm": 0.007665066979825497, "learning_rate": 1.5884861407249468e-05, "loss": 0.0008, "step": 1600 }, { "epoch": 3.4221748400852876, "grad_norm": 0.007227443624287844, "learning_rate": 1.5778251599147122e-05, "loss": 0.0008, "step": 1605 }, { "epoch": 3.4328358208955225, "grad_norm": 0.00955430418252945, "learning_rate": 1.5671641791044777e-05, "loss": 0.0008, "step": 1610 }, { "epoch": 3.443496801705757, "grad_norm": 0.006856875494122505, "learning_rate": 1.5565031982942432e-05, "loss": 0.0008, "step": 1615 }, { "epoch": 3.4541577825159915, "grad_norm": 0.01420837827026844, "learning_rate": 1.5458422174840087e-05, "loss": 0.0008, "step": 1620 }, { "epoch": 3.464818763326226, "grad_norm": 0.06159917637705803, "learning_rate": 1.535181236673774e-05, "loss": 0.0008, "step": 1625 }, { "epoch": 3.4754797441364604, "grad_norm": 0.007254160474985838, "learning_rate": 1.5245202558635396e-05, "loss": 0.0007, "step": 1630 }, { "epoch": 3.486140724946695, "grad_norm": 0.012399489991366863, "learning_rate": 1.5138592750533051e-05, "loss": 0.0009, "step": 1635 }, { "epoch": 3.49680170575693, "grad_norm": 0.011903755366802216, "learning_rate": 1.5031982942430706e-05, "loss": 0.0013, "step": 1640 }, { "epoch": 3.5074626865671643, "grad_norm": 0.006795087829232216, "learning_rate": 1.4925373134328357e-05, "loss": 0.0008, "step": 1645 }, { "epoch": 3.518123667377399, "grad_norm": 0.007012196350842714, "learning_rate": 1.4818763326226012e-05, "loss": 0.0008, "step": 1650 }, { "epoch": 3.5287846481876333, "grad_norm": 0.006799119058996439, "learning_rate": 1.4712153518123666e-05, "loss": 0.0008, "step": 1655 }, { "epoch": 3.5394456289978677, "grad_norm": 0.0064195431768894196, "learning_rate": 1.4605543710021321e-05, "loss": 0.0007, "step": 1660 }, { "epoch": 3.550106609808102, "grad_norm": 0.008582387119531631, "learning_rate": 1.4498933901918976e-05, "loss": 0.0008, "step": 1665 }, { "epoch": 3.5607675906183367, "grad_norm": 0.006130632013082504, "learning_rate": 1.439232409381663e-05, "loss": 0.0007, "step": 1670 }, { "epoch": 3.571428571428571, "grad_norm": 0.0104199955239892, "learning_rate": 1.4285714285714285e-05, "loss": 0.0008, "step": 1675 }, { "epoch": 3.582089552238806, "grad_norm": 0.006977797485888004, "learning_rate": 1.417910447761194e-05, "loss": 0.0007, "step": 1680 }, { "epoch": 3.5927505330490406, "grad_norm": 0.07114434987306595, "learning_rate": 1.4072494669509595e-05, "loss": 0.0008, "step": 1685 }, { "epoch": 3.603411513859275, "grad_norm": 0.009811518713831902, "learning_rate": 1.396588486140725e-05, "loss": 0.0007, "step": 1690 }, { "epoch": 3.6140724946695095, "grad_norm": 0.0063069225288927555, "learning_rate": 1.3859275053304904e-05, "loss": 0.0008, "step": 1695 }, { "epoch": 3.624733475479744, "grad_norm": 0.006444647442549467, "learning_rate": 1.3752665245202559e-05, "loss": 0.0007, "step": 1700 }, { "epoch": 3.635394456289979, "grad_norm": 0.007482605520635843, "learning_rate": 1.3646055437100214e-05, "loss": 0.0007, "step": 1705 }, { "epoch": 3.6460554371002134, "grad_norm": 0.006347157526761293, "learning_rate": 1.3539445628997869e-05, "loss": 0.0007, "step": 1710 }, { "epoch": 3.656716417910448, "grad_norm": 0.006359875667840242, "learning_rate": 1.3432835820895523e-05, "loss": 0.0009, "step": 1715 }, { "epoch": 3.6673773987206824, "grad_norm": 0.01279941201210022, "learning_rate": 1.3326226012793178e-05, "loss": 0.0007, "step": 1720 }, { "epoch": 3.678038379530917, "grad_norm": 0.006938565522432327, "learning_rate": 1.3219616204690833e-05, "loss": 0.0474, "step": 1725 }, { "epoch": 3.6886993603411513, "grad_norm": 0.020562391728162766, "learning_rate": 1.3113006396588488e-05, "loss": 0.0012, "step": 1730 }, { "epoch": 3.699360341151386, "grad_norm": 0.006440839730203152, "learning_rate": 1.3006396588486142e-05, "loss": 0.0008, "step": 1735 }, { "epoch": 3.7100213219616203, "grad_norm": 0.006626409012824297, "learning_rate": 1.2899786780383797e-05, "loss": 0.0007, "step": 1740 }, { "epoch": 3.7206823027718547, "grad_norm": 0.007075931876897812, "learning_rate": 1.2793176972281452e-05, "loss": 0.0007, "step": 1745 }, { "epoch": 3.7313432835820897, "grad_norm": 0.0074788648635149, "learning_rate": 1.2686567164179105e-05, "loss": 0.0008, "step": 1750 }, { "epoch": 3.742004264392324, "grad_norm": 0.005989312659949064, "learning_rate": 1.257995735607676e-05, "loss": 0.0008, "step": 1755 }, { "epoch": 3.7526652452025586, "grad_norm": 7.463435173034668, "learning_rate": 1.2473347547974414e-05, "loss": 0.0027, "step": 1760 }, { "epoch": 3.763326226012793, "grad_norm": 0.0073138996958732605, "learning_rate": 1.2366737739872069e-05, "loss": 0.0008, "step": 1765 }, { "epoch": 3.7739872068230276, "grad_norm": 0.006524786818772554, "learning_rate": 1.2260127931769722e-05, "loss": 0.0008, "step": 1770 }, { "epoch": 3.7846481876332625, "grad_norm": 0.006607542745769024, "learning_rate": 1.2153518123667377e-05, "loss": 0.0007, "step": 1775 }, { "epoch": 3.795309168443497, "grad_norm": 0.006721988320350647, "learning_rate": 1.2046908315565032e-05, "loss": 0.0008, "step": 1780 }, { "epoch": 3.8059701492537314, "grad_norm": 0.006366103421896696, "learning_rate": 1.1940298507462686e-05, "loss": 0.0007, "step": 1785 }, { "epoch": 3.816631130063966, "grad_norm": 0.006934124510735273, "learning_rate": 1.1833688699360341e-05, "loss": 0.0007, "step": 1790 }, { "epoch": 3.8272921108742004, "grad_norm": 0.04416465014219284, "learning_rate": 1.1727078891257996e-05, "loss": 0.0007, "step": 1795 }, { "epoch": 3.837953091684435, "grad_norm": 0.00798969529569149, "learning_rate": 1.162046908315565e-05, "loss": 0.0007, "step": 1800 }, { "epoch": 3.8486140724946694, "grad_norm": 0.00854866299778223, "learning_rate": 1.1513859275053305e-05, "loss": 0.0007, "step": 1805 }, { "epoch": 3.859275053304904, "grad_norm": 0.05965569242835045, "learning_rate": 1.140724946695096e-05, "loss": 0.0008, "step": 1810 }, { "epoch": 3.8699360341151388, "grad_norm": 0.007432678714394569, "learning_rate": 1.1300639658848615e-05, "loss": 0.0008, "step": 1815 }, { "epoch": 3.8805970149253732, "grad_norm": 0.0058729397132992744, "learning_rate": 1.119402985074627e-05, "loss": 0.0007, "step": 1820 }, { "epoch": 3.8912579957356077, "grad_norm": 0.0063949706964194775, "learning_rate": 1.1087420042643924e-05, "loss": 0.0008, "step": 1825 }, { "epoch": 3.901918976545842, "grad_norm": 0.010465593077242374, "learning_rate": 1.0980810234541579e-05, "loss": 0.0007, "step": 1830 }, { "epoch": 3.9125799573560767, "grad_norm": 0.0066323308274149895, "learning_rate": 1.0874200426439234e-05, "loss": 0.0007, "step": 1835 }, { "epoch": 3.923240938166311, "grad_norm": 0.0066092172637581825, "learning_rate": 1.0767590618336887e-05, "loss": 0.0007, "step": 1840 }, { "epoch": 3.933901918976546, "grad_norm": 0.005982421338558197, "learning_rate": 1.0660980810234541e-05, "loss": 0.0007, "step": 1845 }, { "epoch": 3.9445628997867805, "grad_norm": 0.005778250750154257, "learning_rate": 1.0554371002132196e-05, "loss": 0.0007, "step": 1850 }, { "epoch": 3.955223880597015, "grad_norm": 0.006557360291481018, "learning_rate": 1.0447761194029851e-05, "loss": 0.0036, "step": 1855 }, { "epoch": 3.9658848614072495, "grad_norm": 0.006712673231959343, "learning_rate": 1.0341151385927506e-05, "loss": 0.0007, "step": 1860 }, { "epoch": 3.976545842217484, "grad_norm": 0.006519649177789688, "learning_rate": 1.023454157782516e-05, "loss": 0.0007, "step": 1865 }, { "epoch": 3.9872068230277184, "grad_norm": 0.005961988121271133, "learning_rate": 1.0127931769722815e-05, "loss": 0.0008, "step": 1870 }, { "epoch": 3.997867803837953, "grad_norm": 0.006635779049247503, "learning_rate": 1.002132196162047e-05, "loss": 0.0007, "step": 1875 }, { "epoch": 4.0, "eval_accuracy": 0.9773333333333334, "eval_loss": 0.10557926446199417, "eval_runtime": 28.9128, "eval_samples_per_second": 129.7, "eval_steps_per_second": 4.081, "step": 1876 }, { "epoch": 4.008528784648187, "grad_norm": 0.006346689537167549, "learning_rate": 9.914712153518125e-06, "loss": 0.0007, "step": 1880 }, { "epoch": 4.019189765458422, "grad_norm": 0.008006563410162926, "learning_rate": 9.80810234541578e-06, "loss": 0.0007, "step": 1885 }, { "epoch": 4.029850746268656, "grad_norm": 0.007809875067323446, "learning_rate": 9.701492537313434e-06, "loss": 0.0007, "step": 1890 }, { "epoch": 4.040511727078891, "grad_norm": 0.006803672760725021, "learning_rate": 9.594882729211089e-06, "loss": 0.0007, "step": 1895 }, { "epoch": 4.051172707889126, "grad_norm": 0.008930443786084652, "learning_rate": 9.488272921108744e-06, "loss": 0.0007, "step": 1900 }, { "epoch": 4.061833688699361, "grad_norm": 0.005703925155103207, "learning_rate": 9.381663113006398e-06, "loss": 0.0007, "step": 1905 }, { "epoch": 4.072494669509595, "grad_norm": 0.006208935286849737, "learning_rate": 9.275053304904053e-06, "loss": 0.0008, "step": 1910 }, { "epoch": 4.08315565031983, "grad_norm": 0.005720548331737518, "learning_rate": 9.168443496801706e-06, "loss": 0.0006, "step": 1915 }, { "epoch": 4.093816631130064, "grad_norm": 0.006536661647260189, "learning_rate": 9.06183368869936e-06, "loss": 0.0007, "step": 1920 }, { "epoch": 4.104477611940299, "grad_norm": 0.00596208218485117, "learning_rate": 8.955223880597016e-06, "loss": 0.0007, "step": 1925 }, { "epoch": 4.115138592750533, "grad_norm": 0.005358322523534298, "learning_rate": 8.84861407249467e-06, "loss": 0.0006, "step": 1930 }, { "epoch": 4.1257995735607675, "grad_norm": 0.006345430389046669, "learning_rate": 8.742004264392323e-06, "loss": 0.0006, "step": 1935 }, { "epoch": 4.136460554371002, "grad_norm": 0.005854356102645397, "learning_rate": 8.635394456289978e-06, "loss": 0.0007, "step": 1940 }, { "epoch": 4.1471215351812365, "grad_norm": 0.005065130069851875, "learning_rate": 8.528784648187633e-06, "loss": 0.0006, "step": 1945 }, { "epoch": 4.157782515991471, "grad_norm": 0.005898057483136654, "learning_rate": 8.422174840085288e-06, "loss": 0.0007, "step": 1950 }, { "epoch": 4.1684434968017055, "grad_norm": 0.00627669133245945, "learning_rate": 8.315565031982942e-06, "loss": 0.0007, "step": 1955 }, { "epoch": 4.17910447761194, "grad_norm": 0.005751720163971186, "learning_rate": 8.208955223880597e-06, "loss": 0.0007, "step": 1960 }, { "epoch": 4.189765458422174, "grad_norm": 0.00575893884524703, "learning_rate": 8.102345415778252e-06, "loss": 0.0006, "step": 1965 }, { "epoch": 4.20042643923241, "grad_norm": 0.006232505664229393, "learning_rate": 7.995735607675907e-06, "loss": 0.0007, "step": 1970 }, { "epoch": 4.211087420042644, "grad_norm": 0.006513633765280247, "learning_rate": 7.889125799573561e-06, "loss": 0.0007, "step": 1975 }, { "epoch": 4.221748400852879, "grad_norm": 0.005717046558856964, "learning_rate": 7.782515991471216e-06, "loss": 0.0007, "step": 1980 }, { "epoch": 4.232409381663113, "grad_norm": 0.014668038114905357, "learning_rate": 7.67590618336887e-06, "loss": 0.0006, "step": 1985 }, { "epoch": 4.243070362473348, "grad_norm": 0.006526192184537649, "learning_rate": 7.5692963752665255e-06, "loss": 0.0006, "step": 1990 }, { "epoch": 4.253731343283582, "grad_norm": 0.005813642870634794, "learning_rate": 7.4626865671641785e-06, "loss": 0.0007, "step": 1995 }, { "epoch": 4.264392324093817, "grad_norm": 0.006546695716679096, "learning_rate": 7.356076759061833e-06, "loss": 0.0007, "step": 2000 }, { "epoch": 4.275053304904051, "grad_norm": 0.00655017513781786, "learning_rate": 7.249466950959488e-06, "loss": 0.0007, "step": 2005 }, { "epoch": 4.285714285714286, "grad_norm": 0.006542718969285488, "learning_rate": 7.142857142857143e-06, "loss": 0.0006, "step": 2010 }, { "epoch": 4.29637526652452, "grad_norm": 0.0057178670540452, "learning_rate": 7.0362473347547975e-06, "loss": 0.0518, "step": 2015 }, { "epoch": 4.3070362473347545, "grad_norm": 0.006247959565371275, "learning_rate": 6.929637526652452e-06, "loss": 0.0006, "step": 2020 }, { "epoch": 4.317697228144989, "grad_norm": 0.007482743822038174, "learning_rate": 6.823027718550107e-06, "loss": 0.0007, "step": 2025 }, { "epoch": 4.3283582089552235, "grad_norm": 0.005513486452400684, "learning_rate": 6.716417910447762e-06, "loss": 0.0006, "step": 2030 }, { "epoch": 4.339019189765459, "grad_norm": 0.006352463271468878, "learning_rate": 6.609808102345416e-06, "loss": 0.0007, "step": 2035 }, { "epoch": 4.349680170575693, "grad_norm": 0.005849063862115145, "learning_rate": 6.503198294243071e-06, "loss": 0.0006, "step": 2040 }, { "epoch": 4.360341151385928, "grad_norm": 0.005921467673033476, "learning_rate": 6.396588486140726e-06, "loss": 0.0007, "step": 2045 }, { "epoch": 4.371002132196162, "grad_norm": 0.005727188196033239, "learning_rate": 6.28997867803838e-06, "loss": 0.0007, "step": 2050 }, { "epoch": 4.381663113006397, "grad_norm": 0.0059150028973817825, "learning_rate": 6.1833688699360345e-06, "loss": 0.0006, "step": 2055 }, { "epoch": 4.392324093816631, "grad_norm": 0.005900897551327944, "learning_rate": 6.076759061833688e-06, "loss": 0.0007, "step": 2060 }, { "epoch": 4.402985074626866, "grad_norm": 0.00620963703840971, "learning_rate": 5.970149253731343e-06, "loss": 0.0007, "step": 2065 }, { "epoch": 4.4136460554371, "grad_norm": 0.005051200278103352, "learning_rate": 5.863539445628998e-06, "loss": 0.0006, "step": 2070 }, { "epoch": 4.424307036247335, "grad_norm": 0.0057454826310276985, "learning_rate": 5.756929637526653e-06, "loss": 0.0006, "step": 2075 }, { "epoch": 4.434968017057569, "grad_norm": 0.005414177197962999, "learning_rate": 5.650319829424307e-06, "loss": 0.0006, "step": 2080 }, { "epoch": 4.445628997867804, "grad_norm": 0.005801606923341751, "learning_rate": 5.543710021321962e-06, "loss": 0.0007, "step": 2085 }, { "epoch": 4.456289978678038, "grad_norm": 0.0059316218830645084, "learning_rate": 5.437100213219617e-06, "loss": 0.0006, "step": 2090 }, { "epoch": 4.466950959488273, "grad_norm": 0.007300084922462702, "learning_rate": 5.330490405117271e-06, "loss": 0.0007, "step": 2095 }, { "epoch": 4.477611940298507, "grad_norm": 0.006561764515936375, "learning_rate": 5.2238805970149255e-06, "loss": 0.0007, "step": 2100 }, { "epoch": 4.4882729211087415, "grad_norm": 0.006763527635484934, "learning_rate": 5.11727078891258e-06, "loss": 0.0006, "step": 2105 }, { "epoch": 4.498933901918977, "grad_norm": 0.006273689679801464, "learning_rate": 5.010660980810235e-06, "loss": 0.0007, "step": 2110 }, { "epoch": 4.509594882729211, "grad_norm": 0.006041302345693111, "learning_rate": 4.90405117270789e-06, "loss": 0.0006, "step": 2115 }, { "epoch": 4.520255863539446, "grad_norm": 0.005818418692797422, "learning_rate": 4.797441364605544e-06, "loss": 0.0007, "step": 2120 }, { "epoch": 4.53091684434968, "grad_norm": 0.00663146423175931, "learning_rate": 4.690831556503199e-06, "loss": 0.0007, "step": 2125 }, { "epoch": 4.541577825159915, "grad_norm": 0.008564852178096771, "learning_rate": 4.584221748400853e-06, "loss": 0.0006, "step": 2130 }, { "epoch": 4.552238805970149, "grad_norm": 0.005502946674823761, "learning_rate": 4.477611940298508e-06, "loss": 0.0006, "step": 2135 }, { "epoch": 4.562899786780384, "grad_norm": 0.00622538710013032, "learning_rate": 4.371002132196162e-06, "loss": 0.0007, "step": 2140 }, { "epoch": 4.573560767590618, "grad_norm": 0.005693373270332813, "learning_rate": 4.264392324093816e-06, "loss": 0.0007, "step": 2145 }, { "epoch": 4.584221748400853, "grad_norm": 0.005666282959282398, "learning_rate": 4.157782515991471e-06, "loss": 0.0007, "step": 2150 }, { "epoch": 4.594882729211087, "grad_norm": 0.0053250049240887165, "learning_rate": 4.051172707889126e-06, "loss": 0.0007, "step": 2155 }, { "epoch": 4.605543710021322, "grad_norm": 0.005874867085367441, "learning_rate": 3.944562899786781e-06, "loss": 0.0006, "step": 2160 }, { "epoch": 4.616204690831556, "grad_norm": 0.005929911509156227, "learning_rate": 3.837953091684435e-06, "loss": 0.0006, "step": 2165 }, { "epoch": 4.6268656716417915, "grad_norm": 0.007802083622664213, "learning_rate": 3.7313432835820893e-06, "loss": 0.0006, "step": 2170 }, { "epoch": 4.637526652452026, "grad_norm": 0.005895972717553377, "learning_rate": 3.624733475479744e-06, "loss": 0.0006, "step": 2175 }, { "epoch": 4.6481876332622605, "grad_norm": 0.005571678746491671, "learning_rate": 3.5181236673773987e-06, "loss": 0.0006, "step": 2180 }, { "epoch": 4.658848614072495, "grad_norm": 0.005317580420523882, "learning_rate": 3.4115138592750535e-06, "loss": 0.0006, "step": 2185 }, { "epoch": 4.669509594882729, "grad_norm": 0.006017689593136311, "learning_rate": 3.304904051172708e-06, "loss": 0.0006, "step": 2190 }, { "epoch": 4.680170575692964, "grad_norm": 0.005719756707549095, "learning_rate": 3.198294243070363e-06, "loss": 0.0007, "step": 2195 }, { "epoch": 4.690831556503198, "grad_norm": 0.008723727427423, "learning_rate": 3.0916844349680173e-06, "loss": 0.0007, "step": 2200 }, { "epoch": 4.701492537313433, "grad_norm": 0.005117720924317837, "learning_rate": 2.9850746268656716e-06, "loss": 0.0006, "step": 2205 }, { "epoch": 4.712153518123667, "grad_norm": 0.0065277134999632835, "learning_rate": 2.8784648187633263e-06, "loss": 0.0006, "step": 2210 }, { "epoch": 4.722814498933902, "grad_norm": 0.00568021647632122, "learning_rate": 2.771855010660981e-06, "loss": 0.0006, "step": 2215 }, { "epoch": 4.733475479744136, "grad_norm": 0.005861042067408562, "learning_rate": 2.6652452025586354e-06, "loss": 0.0006, "step": 2220 }, { "epoch": 4.744136460554371, "grad_norm": 0.01781267486512661, "learning_rate": 2.55863539445629e-06, "loss": 0.0007, "step": 2225 }, { "epoch": 4.754797441364605, "grad_norm": 0.005976023152470589, "learning_rate": 2.452025586353945e-06, "loss": 0.0007, "step": 2230 }, { "epoch": 4.76545842217484, "grad_norm": 0.0059677972458302975, "learning_rate": 2.3454157782515996e-06, "loss": 0.0006, "step": 2235 }, { "epoch": 4.776119402985074, "grad_norm": 0.006011356599628925, "learning_rate": 2.238805970149254e-06, "loss": 0.0007, "step": 2240 }, { "epoch": 4.786780383795309, "grad_norm": 0.005780589301139116, "learning_rate": 2.132196162046908e-06, "loss": 0.0006, "step": 2245 }, { "epoch": 4.797441364605544, "grad_norm": 0.005922527052462101, "learning_rate": 2.025586353944563e-06, "loss": 0.0006, "step": 2250 }, { "epoch": 4.8081023454157785, "grad_norm": 0.0064440565183758736, "learning_rate": 1.9189765458422177e-06, "loss": 0.0007, "step": 2255 }, { "epoch": 4.818763326226013, "grad_norm": 0.00593932019546628, "learning_rate": 1.812366737739872e-06, "loss": 0.0007, "step": 2260 }, { "epoch": 4.8294243070362475, "grad_norm": 0.005869138054549694, "learning_rate": 1.7057569296375267e-06, "loss": 0.0006, "step": 2265 }, { "epoch": 4.840085287846482, "grad_norm": 0.02662317268550396, "learning_rate": 1.5991471215351815e-06, "loss": 0.0008, "step": 2270 }, { "epoch": 4.850746268656716, "grad_norm": 0.007721681147813797, "learning_rate": 1.4925373134328358e-06, "loss": 0.0006, "step": 2275 }, { "epoch": 4.861407249466951, "grad_norm": 0.005554463714361191, "learning_rate": 1.3859275053304905e-06, "loss": 0.0006, "step": 2280 }, { "epoch": 4.872068230277185, "grad_norm": 0.00487638870254159, "learning_rate": 1.279317697228145e-06, "loss": 0.0006, "step": 2285 }, { "epoch": 4.88272921108742, "grad_norm": 0.006217475514858961, "learning_rate": 1.1727078891257998e-06, "loss": 0.0007, "step": 2290 }, { "epoch": 4.893390191897654, "grad_norm": 0.005931223277002573, "learning_rate": 1.066098081023454e-06, "loss": 0.0006, "step": 2295 }, { "epoch": 4.904051172707889, "grad_norm": 0.006681455764919519, "learning_rate": 9.594882729211088e-07, "loss": 0.0007, "step": 2300 }, { "epoch": 4.914712153518123, "grad_norm": 0.005305519327521324, "learning_rate": 8.528784648187634e-07, "loss": 0.0007, "step": 2305 }, { "epoch": 4.925373134328359, "grad_norm": 0.005433112848550081, "learning_rate": 7.462686567164179e-07, "loss": 0.0006, "step": 2310 }, { "epoch": 4.936034115138593, "grad_norm": 0.005620833486318588, "learning_rate": 6.396588486140725e-07, "loss": 0.0006, "step": 2315 }, { "epoch": 4.946695095948828, "grad_norm": 0.006259509362280369, "learning_rate": 5.33049040511727e-07, "loss": 0.0006, "step": 2320 }, { "epoch": 4.957356076759062, "grad_norm": 0.05239493027329445, "learning_rate": 4.264392324093817e-07, "loss": 0.0006, "step": 2325 }, { "epoch": 4.968017057569297, "grad_norm": 0.005692152306437492, "learning_rate": 3.1982942430703626e-07, "loss": 0.0006, "step": 2330 }, { "epoch": 4.978678038379531, "grad_norm": 0.007721024565398693, "learning_rate": 2.1321961620469084e-07, "loss": 0.0007, "step": 2335 }, { "epoch": 4.9893390191897655, "grad_norm": 0.005226352252066135, "learning_rate": 1.0660980810234542e-07, "loss": 0.0006, "step": 2340 }, { "epoch": 5.0, "grad_norm": 0.006887008901685476, "learning_rate": 0.0, "loss": 0.0006, "step": 2345 }, { "epoch": 5.0, "eval_accuracy": 0.9784, "eval_loss": 0.10852671414613724, "eval_runtime": 28.9463, "eval_samples_per_second": 129.55, "eval_steps_per_second": 4.077, "step": 2345 }, { "epoch": 5.0, "step": 2345, "total_flos": 5.8118992210944e+18, "train_loss": 0.0236334878852023, "train_runtime": 1467.9028, "train_samples_per_second": 51.093, "train_steps_per_second": 1.598 } ], "logging_steps": 5, "max_steps": 2345, "num_input_tokens_seen": 0, "num_train_epochs": 5, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 5.8118992210944e+18, "train_batch_size": 32, "trial_name": null, "trial_params": null }