{ "best_metric": 2.8585827350616455, "best_model_checkpoint": "bert_tiny_lda_5_v1_book/checkpoint-350000", "epoch": 25.0, "eval_steps": 10000, "global_step": 355850, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0351271603203597, "grad_norm": 2.4258768558502197, "learning_rate": 5e-06, "loss": 11.1937, "step": 500 }, { "epoch": 0.0702543206407194, "grad_norm": 0.9785349369049072, "learning_rate": 1e-05, "loss": 9.0871, "step": 1000 }, { "epoch": 0.1053814809610791, "grad_norm": 0.5848240852355957, "learning_rate": 1.5e-05, "loss": 7.8439, "step": 1500 }, { "epoch": 0.1405086412814388, "grad_norm": 0.5948261618614197, "learning_rate": 2e-05, "loss": 7.6427, "step": 2000 }, { "epoch": 0.1756358016017985, "grad_norm": 0.7661160230636597, "learning_rate": 2.5e-05, "loss": 7.5522, "step": 2500 }, { "epoch": 0.2107629619221582, "grad_norm": 0.7159568667411804, "learning_rate": 3e-05, "loss": 7.4907, "step": 3000 }, { "epoch": 0.2458901222425179, "grad_norm": 0.6961402297019958, "learning_rate": 3.5e-05, "loss": 7.4519, "step": 3500 }, { "epoch": 0.2810172825628776, "grad_norm": 0.6688615679740906, "learning_rate": 4e-05, "loss": 7.4202, "step": 4000 }, { "epoch": 0.31614444288323734, "grad_norm": 0.6611530184745789, "learning_rate": 4.5e-05, "loss": 7.3944, "step": 4500 }, { "epoch": 0.351271603203597, "grad_norm": 0.8425453305244446, "learning_rate": 5e-05, "loss": 7.373, "step": 5000 }, { "epoch": 0.38639876352395675, "grad_norm": 0.7047624588012695, "learning_rate": 5.500000000000001e-05, "loss": 7.3492, "step": 5500 }, { "epoch": 0.4215259238443164, "grad_norm": 0.514307975769043, "learning_rate": 6e-05, "loss": 7.3374, "step": 6000 }, { "epoch": 0.45665308416467615, "grad_norm": 0.5747005939483643, "learning_rate": 6.500000000000001e-05, "loss": 7.3227, "step": 6500 }, { "epoch": 0.4917802444850358, "grad_norm": 0.7502753734588623, "learning_rate": 7e-05, "loss": 7.3126, "step": 7000 }, { "epoch": 0.5269074048053956, "grad_norm": 0.6385812163352966, "learning_rate": 7.500000000000001e-05, "loss": 7.2987, "step": 7500 }, { "epoch": 0.5620345651257552, "grad_norm": 0.6978061199188232, "learning_rate": 8e-05, "loss": 7.2858, "step": 8000 }, { "epoch": 0.5971617254461149, "grad_norm": 0.6714794039726257, "learning_rate": 8.5e-05, "loss": 7.2755, "step": 8500 }, { "epoch": 0.6322888857664747, "grad_norm": 0.8273310661315918, "learning_rate": 9e-05, "loss": 7.2633, "step": 9000 }, { "epoch": 0.6674160460868344, "grad_norm": 0.6619166731834412, "learning_rate": 9.5e-05, "loss": 7.2587, "step": 9500 }, { "epoch": 0.702543206407194, "grad_norm": 0.6027777791023254, "learning_rate": 0.0001, "loss": 7.2508, "step": 10000 }, { "epoch": 0.702543206407194, "eval_accuracy": 0.16390076959612057, "eval_loss": 7.091331958770752, "eval_runtime": 189.2701, "eval_samples_per_second": 637.745, "eval_steps_per_second": 3.989, "step": 10000 }, { "epoch": 0.7376703667275537, "grad_norm": 0.5365111827850342, "learning_rate": 9.985542865404077e-05, "loss": 7.2431, "step": 10500 }, { "epoch": 0.7727975270479135, "grad_norm": 0.6523528695106506, "learning_rate": 9.971085730808155e-05, "loss": 7.2368, "step": 11000 }, { "epoch": 0.8079246873682732, "grad_norm": 0.5658899545669556, "learning_rate": 9.956628596212231e-05, "loss": 7.2337, "step": 11500 }, { "epoch": 0.8430518476886328, "grad_norm": 0.5877309441566467, "learning_rate": 9.942171461616309e-05, "loss": 7.2251, "step": 12000 }, { "epoch": 0.8781790080089925, "grad_norm": 0.6072720289230347, "learning_rate": 9.927714327020385e-05, "loss": 7.221, "step": 12500 }, { "epoch": 0.9133061683293523, "grad_norm": 1.3007500171661377, "learning_rate": 9.913257192424462e-05, "loss": 7.1999, "step": 13000 }, { "epoch": 0.948433328649712, "grad_norm": 0.8876389265060425, "learning_rate": 9.89880005782854e-05, "loss": 7.1393, "step": 13500 }, { "epoch": 0.9835604889700716, "grad_norm": 0.7199205756187439, "learning_rate": 9.884342923232616e-05, "loss": 7.0857, "step": 14000 }, { "epoch": 1.0186876492904313, "grad_norm": 0.8176659345626831, "learning_rate": 9.869885788636692e-05, "loss": 7.051, "step": 14500 }, { "epoch": 1.053814809610791, "grad_norm": 0.5849608182907104, "learning_rate": 9.855428654040769e-05, "loss": 7.0423, "step": 15000 }, { "epoch": 1.0889419699311507, "grad_norm": 0.7167512774467468, "learning_rate": 9.840971519444846e-05, "loss": 7.0334, "step": 15500 }, { "epoch": 1.1240691302515105, "grad_norm": 0.8312996029853821, "learning_rate": 9.826514384848924e-05, "loss": 7.0287, "step": 16000 }, { "epoch": 1.1591962905718702, "grad_norm": 0.6854020357131958, "learning_rate": 9.812057250253e-05, "loss": 7.0234, "step": 16500 }, { "epoch": 1.1943234508922298, "grad_norm": 0.8675721883773804, "learning_rate": 9.797600115657077e-05, "loss": 7.0202, "step": 17000 }, { "epoch": 1.2294506112125896, "grad_norm": 0.6203986406326294, "learning_rate": 9.783142981061155e-05, "loss": 7.0109, "step": 17500 }, { "epoch": 1.2645777715329491, "grad_norm": 1.117289423942566, "learning_rate": 9.768685846465231e-05, "loss": 6.9859, "step": 18000 }, { "epoch": 1.299704931853309, "grad_norm": 1.048002004623413, "learning_rate": 9.754228711869307e-05, "loss": 6.9095, "step": 18500 }, { "epoch": 1.3348320921736687, "grad_norm": 1.234838604927063, "learning_rate": 9.739771577273384e-05, "loss": 6.8278, "step": 19000 }, { "epoch": 1.3699592524940285, "grad_norm": 1.856527328491211, "learning_rate": 9.725314442677463e-05, "loss": 6.3418, "step": 19500 }, { "epoch": 1.405086412814388, "grad_norm": 1.4836328029632568, "learning_rate": 9.71085730808154e-05, "loss": 5.6868, "step": 20000 }, { "epoch": 1.405086412814388, "eval_accuracy": 0.40739290530376704, "eval_loss": 5.007092475891113, "eval_runtime": 226.6891, "eval_samples_per_second": 532.474, "eval_steps_per_second": 3.331, "step": 20000 }, { "epoch": 1.4402135731347478, "grad_norm": 1.3329079151153564, "learning_rate": 9.696400173485616e-05, "loss": 5.2708, "step": 20500 }, { "epoch": 1.4753407334551074, "grad_norm": 1.3452800512313843, "learning_rate": 9.681943038889692e-05, "loss": 5.0332, "step": 21000 }, { "epoch": 1.5104678937754672, "grad_norm": 1.0615885257720947, "learning_rate": 9.66748590429377e-05, "loss": 4.88, "step": 21500 }, { "epoch": 1.545595054095827, "grad_norm": 1.1207612752914429, "learning_rate": 9.653028769697846e-05, "loss": 4.7666, "step": 22000 }, { "epoch": 1.5807222144161865, "grad_norm": 1.0668436288833618, "learning_rate": 9.638571635101923e-05, "loss": 4.6837, "step": 22500 }, { "epoch": 1.6158493747365463, "grad_norm": 1.0427242517471313, "learning_rate": 9.624114500506e-05, "loss": 4.6082, "step": 23000 }, { "epoch": 1.650976535056906, "grad_norm": 1.1569322347640991, "learning_rate": 9.609657365910077e-05, "loss": 4.5442, "step": 23500 }, { "epoch": 1.6861036953772657, "grad_norm": 0.9855100512504578, "learning_rate": 9.595200231314155e-05, "loss": 4.4842, "step": 24000 }, { "epoch": 1.7212308556976255, "grad_norm": 1.155909538269043, "learning_rate": 9.580743096718231e-05, "loss": 4.4335, "step": 24500 }, { "epoch": 1.7563580160179852, "grad_norm": 1.0347774028778076, "learning_rate": 9.566285962122307e-05, "loss": 4.3805, "step": 25000 }, { "epoch": 1.7914851763383448, "grad_norm": 1.0738935470581055, "learning_rate": 9.551828827526385e-05, "loss": 4.3329, "step": 25500 }, { "epoch": 1.8266123366587044, "grad_norm": 1.0274145603179932, "learning_rate": 9.537371692930462e-05, "loss": 4.2901, "step": 26000 }, { "epoch": 1.8617394969790642, "grad_norm": 0.9598429203033447, "learning_rate": 9.522914558334538e-05, "loss": 4.2446, "step": 26500 }, { "epoch": 1.896866657299424, "grad_norm": 1.1881024837493896, "learning_rate": 9.508457423738616e-05, "loss": 4.2118, "step": 27000 }, { "epoch": 1.9319938176197837, "grad_norm": 1.1907711029052734, "learning_rate": 9.494000289142692e-05, "loss": 4.1784, "step": 27500 }, { "epoch": 1.9671209779401433, "grad_norm": 1.011011004447937, "learning_rate": 9.47954315454677e-05, "loss": 4.1448, "step": 28000 }, { "epoch": 2.002248138260503, "grad_norm": 0.97409987449646, "learning_rate": 9.465086019950846e-05, "loss": 4.1251, "step": 28500 }, { "epoch": 2.0373752985808626, "grad_norm": 1.1129947900772095, "learning_rate": 9.450628885354923e-05, "loss": 4.0923, "step": 29000 }, { "epoch": 2.0725024589012224, "grad_norm": 1.101016640663147, "learning_rate": 9.436171750758999e-05, "loss": 4.0649, "step": 29500 }, { "epoch": 2.107629619221582, "grad_norm": 1.0833418369293213, "learning_rate": 9.421714616163077e-05, "loss": 4.0487, "step": 30000 }, { "epoch": 2.107629619221582, "eval_accuracy": 0.5617256537619557, "eval_loss": 3.696725606918335, "eval_runtime": 189.3675, "eval_samples_per_second": 637.417, "eval_steps_per_second": 3.987, "step": 30000 }, { "epoch": 2.142756779541942, "grad_norm": 0.9855577349662781, "learning_rate": 9.407257481567155e-05, "loss": 4.0266, "step": 30500 }, { "epoch": 2.1778839398623013, "grad_norm": 0.9121841192245483, "learning_rate": 9.392800346971231e-05, "loss": 4.003, "step": 31000 }, { "epoch": 2.213011100182661, "grad_norm": 0.9467945098876953, "learning_rate": 9.378343212375307e-05, "loss": 3.9889, "step": 31500 }, { "epoch": 2.248138260503021, "grad_norm": 0.9951153993606567, "learning_rate": 9.363886077779385e-05, "loss": 3.971, "step": 32000 }, { "epoch": 2.2832654208233807, "grad_norm": 1.076171875, "learning_rate": 9.349428943183462e-05, "loss": 3.9541, "step": 32500 }, { "epoch": 2.3183925811437405, "grad_norm": 1.1674124002456665, "learning_rate": 9.334971808587538e-05, "loss": 3.9352, "step": 33000 }, { "epoch": 2.3535197414641003, "grad_norm": 0.958756685256958, "learning_rate": 9.320514673991614e-05, "loss": 3.9194, "step": 33500 }, { "epoch": 2.3886469017844596, "grad_norm": 0.9853763580322266, "learning_rate": 9.306057539395692e-05, "loss": 3.9051, "step": 34000 }, { "epoch": 2.4237740621048194, "grad_norm": 0.9711935520172119, "learning_rate": 9.29160040479977e-05, "loss": 3.8887, "step": 34500 }, { "epoch": 2.458901222425179, "grad_norm": 1.1217039823532104, "learning_rate": 9.277143270203846e-05, "loss": 3.8763, "step": 35000 }, { "epoch": 2.494028382745539, "grad_norm": 1.090285062789917, "learning_rate": 9.262686135607923e-05, "loss": 3.8645, "step": 35500 }, { "epoch": 2.5291555430658983, "grad_norm": 1.019071340560913, "learning_rate": 9.248229001012e-05, "loss": 3.8516, "step": 36000 }, { "epoch": 2.564282703386258, "grad_norm": 1.0942332744598389, "learning_rate": 9.233771866416077e-05, "loss": 3.8374, "step": 36500 }, { "epoch": 2.599409863706618, "grad_norm": 1.0246928930282593, "learning_rate": 9.219314731820153e-05, "loss": 3.8253, "step": 37000 }, { "epoch": 2.6345370240269776, "grad_norm": 0.9755086302757263, "learning_rate": 9.20485759722423e-05, "loss": 3.818, "step": 37500 }, { "epoch": 2.6696641843473374, "grad_norm": 1.0182682275772095, "learning_rate": 9.190400462628307e-05, "loss": 3.8054, "step": 38000 }, { "epoch": 2.704791344667697, "grad_norm": 0.9583987593650818, "learning_rate": 9.175943328032385e-05, "loss": 3.7967, "step": 38500 }, { "epoch": 2.739918504988057, "grad_norm": 1.1440130472183228, "learning_rate": 9.161486193436461e-05, "loss": 3.7845, "step": 39000 }, { "epoch": 2.7750456653084163, "grad_norm": 0.9983710050582886, "learning_rate": 9.147029058840538e-05, "loss": 3.7781, "step": 39500 }, { "epoch": 2.810172825628776, "grad_norm": 0.9105831980705261, "learning_rate": 9.132571924244616e-05, "loss": 3.7657, "step": 40000 }, { "epoch": 2.810172825628776, "eval_accuracy": 0.5989295466076175, "eval_loss": 3.442207098007202, "eval_runtime": 197.2369, "eval_samples_per_second": 611.985, "eval_steps_per_second": 3.828, "step": 40000 }, { "epoch": 2.845299985949136, "grad_norm": 0.9889848232269287, "learning_rate": 9.118114789648692e-05, "loss": 3.7549, "step": 40500 }, { "epoch": 2.8804271462694957, "grad_norm": 0.9526759386062622, "learning_rate": 9.103657655052768e-05, "loss": 3.7473, "step": 41000 }, { "epoch": 2.915554306589855, "grad_norm": 0.9214153289794922, "learning_rate": 9.089200520456845e-05, "loss": 3.7391, "step": 41500 }, { "epoch": 2.950681466910215, "grad_norm": 0.9279940724372864, "learning_rate": 9.074743385860923e-05, "loss": 3.7343, "step": 42000 }, { "epoch": 2.9858086272305746, "grad_norm": 0.9647195339202881, "learning_rate": 9.060286251265e-05, "loss": 3.7277, "step": 42500 }, { "epoch": 3.0209357875509344, "grad_norm": 0.9329895377159119, "learning_rate": 9.045829116669077e-05, "loss": 3.7132, "step": 43000 }, { "epoch": 3.056062947871294, "grad_norm": 1.0447282791137695, "learning_rate": 9.031371982073153e-05, "loss": 3.7097, "step": 43500 }, { "epoch": 3.091190108191654, "grad_norm": 0.9059212803840637, "learning_rate": 9.016914847477231e-05, "loss": 3.7061, "step": 44000 }, { "epoch": 3.1263172685120133, "grad_norm": 0.9500852823257446, "learning_rate": 9.002457712881307e-05, "loss": 3.6966, "step": 44500 }, { "epoch": 3.161444428832373, "grad_norm": 0.9555869102478027, "learning_rate": 8.988000578285384e-05, "loss": 3.692, "step": 45000 }, { "epoch": 3.196571589152733, "grad_norm": 1.055525779724121, "learning_rate": 8.973543443689461e-05, "loss": 3.6823, "step": 45500 }, { "epoch": 3.2316987494730927, "grad_norm": 0.9254573583602905, "learning_rate": 8.959086309093538e-05, "loss": 3.6789, "step": 46000 }, { "epoch": 3.2668259097934524, "grad_norm": 1.0057339668273926, "learning_rate": 8.944629174497616e-05, "loss": 3.6711, "step": 46500 }, { "epoch": 3.301953070113812, "grad_norm": 0.9492254257202148, "learning_rate": 8.930172039901692e-05, "loss": 3.6655, "step": 47000 }, { "epoch": 3.3370802304341716, "grad_norm": 0.8773466944694519, "learning_rate": 8.915714905305768e-05, "loss": 3.6598, "step": 47500 }, { "epoch": 3.3722073907545314, "grad_norm": 1.0529727935791016, "learning_rate": 8.901257770709845e-05, "loss": 3.6548, "step": 48000 }, { "epoch": 3.407334551074891, "grad_norm": 0.9654408693313599, "learning_rate": 8.886800636113923e-05, "loss": 3.6478, "step": 48500 }, { "epoch": 3.442461711395251, "grad_norm": 0.9600367546081543, "learning_rate": 8.872343501518e-05, "loss": 3.64, "step": 49000 }, { "epoch": 3.4775888717156107, "grad_norm": 1.0016140937805176, "learning_rate": 8.857886366922077e-05, "loss": 3.6361, "step": 49500 }, { "epoch": 3.51271603203597, "grad_norm": 0.9173470735549927, "learning_rate": 8.843429232326153e-05, "loss": 3.6336, "step": 50000 }, { "epoch": 3.51271603203597, "eval_accuracy": 0.6176236758029259, "eval_loss": 3.3141584396362305, "eval_runtime": 241.2795, "eval_samples_per_second": 500.274, "eval_steps_per_second": 3.129, "step": 50000 }, { "epoch": 3.54784319235633, "grad_norm": 0.9954687356948853, "learning_rate": 8.828972097730231e-05, "loss": 3.6285, "step": 50500 }, { "epoch": 3.5829703526766896, "grad_norm": 0.9681750535964966, "learning_rate": 8.814514963134307e-05, "loss": 3.6211, "step": 51000 }, { "epoch": 3.6180975129970494, "grad_norm": 0.9054061770439148, "learning_rate": 8.800057828538384e-05, "loss": 3.6202, "step": 51500 }, { "epoch": 3.6532246733174087, "grad_norm": 1.1536766290664673, "learning_rate": 8.78560069394246e-05, "loss": 3.6134, "step": 52000 }, { "epoch": 3.6883518336377685, "grad_norm": 1.057255506515503, "learning_rate": 8.771143559346538e-05, "loss": 3.6081, "step": 52500 }, { "epoch": 3.7234789939581283, "grad_norm": 0.9315850734710693, "learning_rate": 8.756686424750616e-05, "loss": 3.6054, "step": 53000 }, { "epoch": 3.758606154278488, "grad_norm": 0.9245489835739136, "learning_rate": 8.742229290154692e-05, "loss": 3.6014, "step": 53500 }, { "epoch": 3.793733314598848, "grad_norm": 0.9240121245384216, "learning_rate": 8.727772155558768e-05, "loss": 3.5927, "step": 54000 }, { "epoch": 3.8288604749192077, "grad_norm": 0.9152762293815613, "learning_rate": 8.713315020962846e-05, "loss": 3.591, "step": 54500 }, { "epoch": 3.8639876352395675, "grad_norm": 1.0582739114761353, "learning_rate": 8.698857886366923e-05, "loss": 3.5835, "step": 55000 }, { "epoch": 3.899114795559927, "grad_norm": 0.938005805015564, "learning_rate": 8.684400751770999e-05, "loss": 3.5798, "step": 55500 }, { "epoch": 3.9342419558802866, "grad_norm": 0.9770874977111816, "learning_rate": 8.669943617175075e-05, "loss": 3.5773, "step": 56000 }, { "epoch": 3.9693691162006464, "grad_norm": 1.0012571811676025, "learning_rate": 8.655486482579153e-05, "loss": 3.5736, "step": 56500 }, { "epoch": 4.004496276521006, "grad_norm": 1.0326473712921143, "learning_rate": 8.641029347983231e-05, "loss": 3.5672, "step": 57000 }, { "epoch": 4.0396234368413655, "grad_norm": 1.0577622652053833, "learning_rate": 8.626572213387307e-05, "loss": 3.5605, "step": 57500 }, { "epoch": 4.074750597161725, "grad_norm": 1.0534971952438354, "learning_rate": 8.612115078791384e-05, "loss": 3.5591, "step": 58000 }, { "epoch": 4.109877757482085, "grad_norm": 1.0737226009368896, "learning_rate": 8.597657944195461e-05, "loss": 3.5555, "step": 58500 }, { "epoch": 4.145004917802445, "grad_norm": 1.0102324485778809, "learning_rate": 8.583200809599538e-05, "loss": 3.5498, "step": 59000 }, { "epoch": 4.180132078122805, "grad_norm": 0.9273024201393127, "learning_rate": 8.568743675003614e-05, "loss": 3.5522, "step": 59500 }, { "epoch": 4.215259238443164, "grad_norm": 1.1412433385849, "learning_rate": 8.55428654040769e-05, "loss": 3.5449, "step": 60000 }, { "epoch": 4.215259238443164, "eval_accuracy": 0.6291410872988004, "eval_loss": 3.2371888160705566, "eval_runtime": 170.6019, "eval_samples_per_second": 707.53, "eval_steps_per_second": 4.426, "step": 60000 }, { "epoch": 4.250386398763524, "grad_norm": 0.9087731838226318, "learning_rate": 8.539829405811768e-05, "loss": 3.5457, "step": 60500 }, { "epoch": 4.285513559083884, "grad_norm": 0.9552867412567139, "learning_rate": 8.525372271215846e-05, "loss": 3.5376, "step": 61000 }, { "epoch": 4.320640719404244, "grad_norm": 0.883446991443634, "learning_rate": 8.510915136619923e-05, "loss": 3.5334, "step": 61500 }, { "epoch": 4.355767879724603, "grad_norm": 0.9625371098518372, "learning_rate": 8.496458002023999e-05, "loss": 3.5312, "step": 62000 }, { "epoch": 4.3908950400449624, "grad_norm": 0.9487352967262268, "learning_rate": 8.482000867428075e-05, "loss": 3.5327, "step": 62500 }, { "epoch": 4.426022200365322, "grad_norm": 0.9547644853591919, "learning_rate": 8.467543732832153e-05, "loss": 3.5283, "step": 63000 }, { "epoch": 4.461149360685682, "grad_norm": 0.9435824751853943, "learning_rate": 8.45308659823623e-05, "loss": 3.5223, "step": 63500 }, { "epoch": 4.496276521006042, "grad_norm": 0.988957405090332, "learning_rate": 8.438629463640307e-05, "loss": 3.5212, "step": 64000 }, { "epoch": 4.531403681326402, "grad_norm": 1.0021283626556396, "learning_rate": 8.424172329044384e-05, "loss": 3.5174, "step": 64500 }, { "epoch": 4.566530841646761, "grad_norm": 1.0738072395324707, "learning_rate": 8.409715194448461e-05, "loss": 3.5144, "step": 65000 }, { "epoch": 4.601658001967121, "grad_norm": 0.964025616645813, "learning_rate": 8.395258059852538e-05, "loss": 3.5121, "step": 65500 }, { "epoch": 4.636785162287481, "grad_norm": 1.0260933637619019, "learning_rate": 8.380800925256614e-05, "loss": 3.509, "step": 66000 }, { "epoch": 4.671912322607841, "grad_norm": 0.9634905457496643, "learning_rate": 8.36634379066069e-05, "loss": 3.5091, "step": 66500 }, { "epoch": 4.7070394829282005, "grad_norm": 0.9776327610015869, "learning_rate": 8.351886656064768e-05, "loss": 3.5014, "step": 67000 }, { "epoch": 4.742166643248559, "grad_norm": 0.966543972492218, "learning_rate": 8.337429521468846e-05, "loss": 3.5019, "step": 67500 }, { "epoch": 4.777293803568919, "grad_norm": 0.9388764500617981, "learning_rate": 8.322972386872923e-05, "loss": 3.4993, "step": 68000 }, { "epoch": 4.812420963889279, "grad_norm": 0.9204984903335571, "learning_rate": 8.308515252276999e-05, "loss": 3.4979, "step": 68500 }, { "epoch": 4.847548124209639, "grad_norm": 0.9963185787200928, "learning_rate": 8.294058117681077e-05, "loss": 3.4914, "step": 69000 }, { "epoch": 4.8826752845299985, "grad_norm": 0.9862936735153198, "learning_rate": 8.279600983085153e-05, "loss": 3.4912, "step": 69500 }, { "epoch": 4.917802444850358, "grad_norm": 1.016281247138977, "learning_rate": 8.26514384848923e-05, "loss": 3.4893, "step": 70000 }, { "epoch": 4.917802444850358, "eval_accuracy": 0.6376040431695125, "eval_loss": 3.1787805557250977, "eval_runtime": 172.1839, "eval_samples_per_second": 701.029, "eval_steps_per_second": 4.385, "step": 70000 }, { "epoch": 4.952929605170718, "grad_norm": 0.9065674543380737, "learning_rate": 8.250686713893306e-05, "loss": 3.4873, "step": 70500 }, { "epoch": 4.988056765491078, "grad_norm": 0.9568567276000977, "learning_rate": 8.236229579297384e-05, "loss": 3.4831, "step": 71000 }, { "epoch": 5.023183925811438, "grad_norm": 0.9480956792831421, "learning_rate": 8.221772444701461e-05, "loss": 3.4823, "step": 71500 }, { "epoch": 5.0583110861317975, "grad_norm": 0.937225878238678, "learning_rate": 8.207315310105538e-05, "loss": 3.4753, "step": 72000 }, { "epoch": 5.093438246452157, "grad_norm": 1.0734835863113403, "learning_rate": 8.192858175509614e-05, "loss": 3.4715, "step": 72500 }, { "epoch": 5.128565406772516, "grad_norm": 0.9612576365470886, "learning_rate": 8.178401040913692e-05, "loss": 3.4686, "step": 73000 }, { "epoch": 5.163692567092876, "grad_norm": 0.9185591340065002, "learning_rate": 8.163943906317768e-05, "loss": 3.4684, "step": 73500 }, { "epoch": 5.198819727413236, "grad_norm": 0.9828550815582275, "learning_rate": 8.149486771721845e-05, "loss": 3.4642, "step": 74000 }, { "epoch": 5.2339468877335955, "grad_norm": 0.9762160181999207, "learning_rate": 8.135029637125921e-05, "loss": 3.4663, "step": 74500 }, { "epoch": 5.269074048053955, "grad_norm": 0.9395500421524048, "learning_rate": 8.120572502529999e-05, "loss": 3.4625, "step": 75000 }, { "epoch": 5.304201208374315, "grad_norm": 0.9051805734634399, "learning_rate": 8.106115367934077e-05, "loss": 3.4603, "step": 75500 }, { "epoch": 5.339328368694675, "grad_norm": 0.9372847080230713, "learning_rate": 8.091658233338153e-05, "loss": 3.4575, "step": 76000 }, { "epoch": 5.374455529015035, "grad_norm": 1.1480001211166382, "learning_rate": 8.07720109874223e-05, "loss": 3.4553, "step": 76500 }, { "epoch": 5.409582689335394, "grad_norm": 1.0730514526367188, "learning_rate": 8.062743964146306e-05, "loss": 3.4511, "step": 77000 }, { "epoch": 5.444709849655754, "grad_norm": 1.0617109537124634, "learning_rate": 8.048286829550384e-05, "loss": 3.4471, "step": 77500 }, { "epoch": 5.479837009976113, "grad_norm": 1.0057247877120972, "learning_rate": 8.03382969495446e-05, "loss": 3.4482, "step": 78000 }, { "epoch": 5.514964170296473, "grad_norm": 0.9427891373634338, "learning_rate": 8.019372560358536e-05, "loss": 3.4459, "step": 78500 }, { "epoch": 5.550091330616833, "grad_norm": 1.1544243097305298, "learning_rate": 8.004915425762614e-05, "loss": 3.4443, "step": 79000 }, { "epoch": 5.5852184909371925, "grad_norm": 0.95564866065979, "learning_rate": 7.990458291166692e-05, "loss": 3.4425, "step": 79500 }, { "epoch": 5.620345651257552, "grad_norm": 0.9871929883956909, "learning_rate": 7.976001156570768e-05, "loss": 3.4397, "step": 80000 }, { "epoch": 5.620345651257552, "eval_accuracy": 0.6442254979948139, "eval_loss": 3.1367175579071045, "eval_runtime": 171.5804, "eval_samples_per_second": 703.495, "eval_steps_per_second": 4.4, "step": 80000 }, { "epoch": 5.655472811577912, "grad_norm": 0.9615615606307983, "learning_rate": 7.961544021974845e-05, "loss": 3.4361, "step": 80500 }, { "epoch": 5.690599971898272, "grad_norm": 0.9359455108642578, "learning_rate": 7.947086887378921e-05, "loss": 3.4423, "step": 81000 }, { "epoch": 5.725727132218632, "grad_norm": 1.0585249662399292, "learning_rate": 7.932629752782999e-05, "loss": 3.4363, "step": 81500 }, { "epoch": 5.760854292538991, "grad_norm": 0.9661962389945984, "learning_rate": 7.918172618187075e-05, "loss": 3.433, "step": 82000 }, { "epoch": 5.795981452859351, "grad_norm": 1.0288212299346924, "learning_rate": 7.903715483591153e-05, "loss": 3.4333, "step": 82500 }, { "epoch": 5.83110861317971, "grad_norm": 1.0270296335220337, "learning_rate": 7.88925834899523e-05, "loss": 3.4311, "step": 83000 }, { "epoch": 5.866235773500071, "grad_norm": 1.0638517141342163, "learning_rate": 7.874801214399307e-05, "loss": 3.4297, "step": 83500 }, { "epoch": 5.90136293382043, "grad_norm": 1.1506760120391846, "learning_rate": 7.860344079803384e-05, "loss": 3.4268, "step": 84000 }, { "epoch": 5.936490094140789, "grad_norm": 0.9845248460769653, "learning_rate": 7.84588694520746e-05, "loss": 3.4239, "step": 84500 }, { "epoch": 5.971617254461149, "grad_norm": 1.0493260622024536, "learning_rate": 7.831429810611536e-05, "loss": 3.4235, "step": 85000 }, { "epoch": 6.006744414781509, "grad_norm": 0.9381208419799805, "learning_rate": 7.816972676015614e-05, "loss": 3.4231, "step": 85500 }, { "epoch": 6.041871575101869, "grad_norm": 0.9385048747062683, "learning_rate": 7.802515541419692e-05, "loss": 3.4169, "step": 86000 }, { "epoch": 6.076998735422229, "grad_norm": 0.991669237613678, "learning_rate": 7.788058406823768e-05, "loss": 3.4188, "step": 86500 }, { "epoch": 6.112125895742588, "grad_norm": 0.9762731790542603, "learning_rate": 7.773601272227845e-05, "loss": 3.4127, "step": 87000 }, { "epoch": 6.147253056062948, "grad_norm": 0.9846952557563782, "learning_rate": 7.759144137631922e-05, "loss": 3.4114, "step": 87500 }, { "epoch": 6.182380216383308, "grad_norm": 0.8991774320602417, "learning_rate": 7.744687003035999e-05, "loss": 3.4108, "step": 88000 }, { "epoch": 6.217507376703667, "grad_norm": 0.9704461097717285, "learning_rate": 7.730229868440075e-05, "loss": 3.4108, "step": 88500 }, { "epoch": 6.252634537024027, "grad_norm": 0.9730648398399353, "learning_rate": 7.715772733844152e-05, "loss": 3.4061, "step": 89000 }, { "epoch": 6.287761697344386, "grad_norm": 1.0883005857467651, "learning_rate": 7.70131559924823e-05, "loss": 3.4057, "step": 89500 }, { "epoch": 6.322888857664746, "grad_norm": 1.0396915674209595, "learning_rate": 7.686858464652307e-05, "loss": 3.4066, "step": 90000 }, { "epoch": 6.322888857664746, "eval_accuracy": 0.6490976777167169, "eval_loss": 3.105360507965088, "eval_runtime": 176.7917, "eval_samples_per_second": 682.758, "eval_steps_per_second": 4.271, "step": 90000 }, { "epoch": 6.358016017985106, "grad_norm": 0.9814009666442871, "learning_rate": 7.672401330056384e-05, "loss": 3.4048, "step": 90500 }, { "epoch": 6.393143178305466, "grad_norm": 0.9409726858139038, "learning_rate": 7.65794419546046e-05, "loss": 3.4015, "step": 91000 }, { "epoch": 6.4282703386258255, "grad_norm": 1.011364221572876, "learning_rate": 7.643487060864536e-05, "loss": 3.3997, "step": 91500 }, { "epoch": 6.463397498946185, "grad_norm": 0.9707909822463989, "learning_rate": 7.629029926268614e-05, "loss": 3.3986, "step": 92000 }, { "epoch": 6.498524659266545, "grad_norm": 0.9443272948265076, "learning_rate": 7.61457279167269e-05, "loss": 3.3982, "step": 92500 }, { "epoch": 6.533651819586905, "grad_norm": 0.948948860168457, "learning_rate": 7.600115657076767e-05, "loss": 3.3985, "step": 93000 }, { "epoch": 6.568778979907265, "grad_norm": 1.1618703603744507, "learning_rate": 7.585658522480845e-05, "loss": 3.3942, "step": 93500 }, { "epoch": 6.603906140227624, "grad_norm": 0.9058728218078613, "learning_rate": 7.571201387884922e-05, "loss": 3.3918, "step": 94000 }, { "epoch": 6.639033300547983, "grad_norm": 0.9888197183609009, "learning_rate": 7.556744253288999e-05, "loss": 3.3901, "step": 94500 }, { "epoch": 6.674160460868343, "grad_norm": 0.9335415363311768, "learning_rate": 7.542287118693075e-05, "loss": 3.3897, "step": 95000 }, { "epoch": 6.709287621188703, "grad_norm": 0.9256088733673096, "learning_rate": 7.527829984097152e-05, "loss": 3.3879, "step": 95500 }, { "epoch": 6.744414781509063, "grad_norm": 0.9326155781745911, "learning_rate": 7.51337284950123e-05, "loss": 3.3922, "step": 96000 }, { "epoch": 6.7795419418294225, "grad_norm": 0.954921305179596, "learning_rate": 7.498915714905306e-05, "loss": 3.3845, "step": 96500 }, { "epoch": 6.814669102149782, "grad_norm": 1.016191840171814, "learning_rate": 7.484458580309382e-05, "loss": 3.3861, "step": 97000 }, { "epoch": 6.849796262470142, "grad_norm": 0.9487805366516113, "learning_rate": 7.47000144571346e-05, "loss": 3.3804, "step": 97500 }, { "epoch": 6.884923422790502, "grad_norm": 1.1292203664779663, "learning_rate": 7.455544311117538e-05, "loss": 3.3803, "step": 98000 }, { "epoch": 6.920050583110862, "grad_norm": 1.0362935066223145, "learning_rate": 7.441087176521614e-05, "loss": 3.3783, "step": 98500 }, { "epoch": 6.955177743431221, "grad_norm": 0.9618114829063416, "learning_rate": 7.42663004192569e-05, "loss": 3.3785, "step": 99000 }, { "epoch": 6.99030490375158, "grad_norm": 0.9607150554656982, "learning_rate": 7.412172907329767e-05, "loss": 3.374, "step": 99500 }, { "epoch": 7.02543206407194, "grad_norm": 1.0077358484268188, "learning_rate": 7.397715772733845e-05, "loss": 3.3758, "step": 100000 }, { "epoch": 7.02543206407194, "eval_accuracy": 0.6534222780368004, "eval_loss": 3.0733838081359863, "eval_runtime": 175.7741, "eval_samples_per_second": 686.711, "eval_steps_per_second": 4.295, "step": 100000 }, { "epoch": 7.0605592243923, "grad_norm": 0.9707036018371582, "learning_rate": 7.383258638137921e-05, "loss": 3.3731, "step": 100500 }, { "epoch": 7.09568638471266, "grad_norm": 0.9526292085647583, "learning_rate": 7.368801503541999e-05, "loss": 3.3727, "step": 101000 }, { "epoch": 7.1308135450330195, "grad_norm": 1.0115597248077393, "learning_rate": 7.354344368946075e-05, "loss": 3.3726, "step": 101500 }, { "epoch": 7.165940705353379, "grad_norm": 0.9818953275680542, "learning_rate": 7.339887234350153e-05, "loss": 3.3693, "step": 102000 }, { "epoch": 7.201067865673739, "grad_norm": 0.9782551527023315, "learning_rate": 7.32543009975423e-05, "loss": 3.368, "step": 102500 }, { "epoch": 7.236195025994099, "grad_norm": 0.9621304273605347, "learning_rate": 7.310972965158306e-05, "loss": 3.3668, "step": 103000 }, { "epoch": 7.271322186314459, "grad_norm": 0.9694998860359192, "learning_rate": 7.296515830562382e-05, "loss": 3.3674, "step": 103500 }, { "epoch": 7.306449346634818, "grad_norm": 1.0466331243515015, "learning_rate": 7.28205869596646e-05, "loss": 3.3619, "step": 104000 }, { "epoch": 7.341576506955178, "grad_norm": 0.9282737374305725, "learning_rate": 7.267601561370536e-05, "loss": 3.3582, "step": 104500 }, { "epoch": 7.376703667275537, "grad_norm": 1.0513064861297607, "learning_rate": 7.253144426774614e-05, "loss": 3.3614, "step": 105000 }, { "epoch": 7.411830827595897, "grad_norm": 1.0153560638427734, "learning_rate": 7.23868729217869e-05, "loss": 3.3655, "step": 105500 }, { "epoch": 7.446957987916257, "grad_norm": 0.9592369794845581, "learning_rate": 7.224230157582768e-05, "loss": 3.3629, "step": 106000 }, { "epoch": 7.482085148236616, "grad_norm": 0.9639871120452881, "learning_rate": 7.209773022986845e-05, "loss": 3.3592, "step": 106500 }, { "epoch": 7.517212308556976, "grad_norm": 1.0919930934906006, "learning_rate": 7.195315888390921e-05, "loss": 3.3555, "step": 107000 }, { "epoch": 7.552339468877336, "grad_norm": 0.9628298282623291, "learning_rate": 7.180858753794997e-05, "loss": 3.3543, "step": 107500 }, { "epoch": 7.587466629197696, "grad_norm": 0.9425886869430542, "learning_rate": 7.166401619199075e-05, "loss": 3.356, "step": 108000 }, { "epoch": 7.6225937895180556, "grad_norm": 1.054260015487671, "learning_rate": 7.151944484603153e-05, "loss": 3.3568, "step": 108500 }, { "epoch": 7.657720949838415, "grad_norm": 0.9053565263748169, "learning_rate": 7.13748735000723e-05, "loss": 3.3528, "step": 109000 }, { "epoch": 7.692848110158775, "grad_norm": 0.9711061120033264, "learning_rate": 7.123030215411306e-05, "loss": 3.3525, "step": 109500 }, { "epoch": 7.727975270479135, "grad_norm": 0.9495022296905518, "learning_rate": 7.108573080815382e-05, "loss": 3.3548, "step": 110000 }, { "epoch": 7.727975270479135, "eval_accuracy": 0.6570935898665028, "eval_loss": 3.0504095554351807, "eval_runtime": 181.6704, "eval_samples_per_second": 664.423, "eval_steps_per_second": 4.156, "step": 110000 }, { "epoch": 7.763102430799494, "grad_norm": 0.9641624093055725, "learning_rate": 7.09411594621946e-05, "loss": 3.351, "step": 110500 }, { "epoch": 7.798229591119854, "grad_norm": 0.9371187090873718, "learning_rate": 7.079658811623536e-05, "loss": 3.3476, "step": 111000 }, { "epoch": 7.833356751440213, "grad_norm": 1.0344008207321167, "learning_rate": 7.065201677027613e-05, "loss": 3.3472, "step": 111500 }, { "epoch": 7.868483911760573, "grad_norm": 0.9815768599510193, "learning_rate": 7.05074454243169e-05, "loss": 3.3487, "step": 112000 }, { "epoch": 7.903611072080933, "grad_norm": 0.9961532354354858, "learning_rate": 7.036287407835768e-05, "loss": 3.3473, "step": 112500 }, { "epoch": 7.938738232401293, "grad_norm": 1.0052876472473145, "learning_rate": 7.021830273239845e-05, "loss": 3.3401, "step": 113000 }, { "epoch": 7.9738653927216525, "grad_norm": 1.0125106573104858, "learning_rate": 7.007373138643921e-05, "loss": 3.3417, "step": 113500 }, { "epoch": 8.008992553042011, "grad_norm": 1.035673975944519, "learning_rate": 6.992916004047997e-05, "loss": 3.3438, "step": 114000 }, { "epoch": 8.044119713362372, "grad_norm": 0.9409092664718628, "learning_rate": 6.978458869452075e-05, "loss": 3.3387, "step": 114500 }, { "epoch": 8.079246873682731, "grad_norm": 0.9636377096176147, "learning_rate": 6.964001734856152e-05, "loss": 3.3381, "step": 115000 }, { "epoch": 8.114374034003092, "grad_norm": 0.9915282726287842, "learning_rate": 6.949544600260228e-05, "loss": 3.3376, "step": 115500 }, { "epoch": 8.14950119432345, "grad_norm": 0.9477564096450806, "learning_rate": 6.935087465664306e-05, "loss": 3.3383, "step": 116000 }, { "epoch": 8.184628354643811, "grad_norm": 0.9475392699241638, "learning_rate": 6.920630331068383e-05, "loss": 3.3378, "step": 116500 }, { "epoch": 8.21975551496417, "grad_norm": 0.9279679656028748, "learning_rate": 6.90617319647246e-05, "loss": 3.3333, "step": 117000 }, { "epoch": 8.25488267528453, "grad_norm": 1.008705735206604, "learning_rate": 6.891716061876536e-05, "loss": 3.3339, "step": 117500 }, { "epoch": 8.29000983560489, "grad_norm": 0.9312646985054016, "learning_rate": 6.877258927280613e-05, "loss": 3.3309, "step": 118000 }, { "epoch": 8.325136995925249, "grad_norm": 0.9859389662742615, "learning_rate": 6.86280179268469e-05, "loss": 3.3297, "step": 118500 }, { "epoch": 8.36026415624561, "grad_norm": 0.9474372863769531, "learning_rate": 6.848344658088767e-05, "loss": 3.3303, "step": 119000 }, { "epoch": 8.395391316565968, "grad_norm": 0.9352207183837891, "learning_rate": 6.833887523492845e-05, "loss": 3.3303, "step": 119500 }, { "epoch": 8.430518476886329, "grad_norm": 0.9617791771888733, "learning_rate": 6.819430388896921e-05, "loss": 3.3302, "step": 120000 }, { "epoch": 8.430518476886329, "eval_accuracy": 0.6599465012923535, "eval_loss": 3.030395984649658, "eval_runtime": 186.4499, "eval_samples_per_second": 647.391, "eval_steps_per_second": 4.049, "step": 120000 }, { "epoch": 8.465645637206688, "grad_norm": 0.9561007618904114, "learning_rate": 6.804973254300999e-05, "loss": 3.3295, "step": 120500 }, { "epoch": 8.500772797527048, "grad_norm": 0.9580948948860168, "learning_rate": 6.790516119705075e-05, "loss": 3.3297, "step": 121000 }, { "epoch": 8.535899957847407, "grad_norm": 1.0268235206604004, "learning_rate": 6.776058985109152e-05, "loss": 3.3297, "step": 121500 }, { "epoch": 8.571027118167768, "grad_norm": 1.0322978496551514, "learning_rate": 6.761601850513228e-05, "loss": 3.3262, "step": 122000 }, { "epoch": 8.606154278488127, "grad_norm": 0.9905152320861816, "learning_rate": 6.747144715917306e-05, "loss": 3.3243, "step": 122500 }, { "epoch": 8.641281438808488, "grad_norm": 0.9595542550086975, "learning_rate": 6.732687581321382e-05, "loss": 3.3235, "step": 123000 }, { "epoch": 8.676408599128846, "grad_norm": 1.0159380435943604, "learning_rate": 6.71823044672546e-05, "loss": 3.3196, "step": 123500 }, { "epoch": 8.711535759449205, "grad_norm": 1.0143177509307861, "learning_rate": 6.703773312129536e-05, "loss": 3.3176, "step": 124000 }, { "epoch": 8.746662919769566, "grad_norm": 0.9414626955986023, "learning_rate": 6.689316177533613e-05, "loss": 3.3149, "step": 124500 }, { "epoch": 8.781790080089925, "grad_norm": 0.9970999956130981, "learning_rate": 6.67485904293769e-05, "loss": 3.3187, "step": 125000 }, { "epoch": 8.816917240410286, "grad_norm": 1.0234370231628418, "learning_rate": 6.660401908341767e-05, "loss": 3.3181, "step": 125500 }, { "epoch": 8.852044400730644, "grad_norm": 0.940746009349823, "learning_rate": 6.645944773745843e-05, "loss": 3.3172, "step": 126000 }, { "epoch": 8.887171561051005, "grad_norm": 0.9791254997253418, "learning_rate": 6.631487639149921e-05, "loss": 3.3177, "step": 126500 }, { "epoch": 8.922298721371364, "grad_norm": 1.0222547054290771, "learning_rate": 6.617030504553999e-05, "loss": 3.3154, "step": 127000 }, { "epoch": 8.957425881691725, "grad_norm": 1.0297785997390747, "learning_rate": 6.602573369958075e-05, "loss": 3.3208, "step": 127500 }, { "epoch": 8.992553042012084, "grad_norm": 0.9728857278823853, "learning_rate": 6.588116235362152e-05, "loss": 3.3197, "step": 128000 }, { "epoch": 9.027680202332444, "grad_norm": 1.0127958059310913, "learning_rate": 6.573659100766228e-05, "loss": 3.311, "step": 128500 }, { "epoch": 9.062807362652803, "grad_norm": 0.9339101910591125, "learning_rate": 6.559201966170306e-05, "loss": 3.3104, "step": 129000 }, { "epoch": 9.097934522973162, "grad_norm": 0.995980978012085, "learning_rate": 6.544744831574382e-05, "loss": 3.3125, "step": 129500 }, { "epoch": 9.133061683293523, "grad_norm": 0.9456177353858948, "learning_rate": 6.530287696978458e-05, "loss": 3.3087, "step": 130000 }, { "epoch": 9.133061683293523, "eval_accuracy": 0.6619766581709358, "eval_loss": 3.015723466873169, "eval_runtime": 186.0907, "eval_samples_per_second": 648.641, "eval_steps_per_second": 4.057, "step": 130000 }, { "epoch": 9.168188843613882, "grad_norm": 1.1779860258102417, "learning_rate": 6.515830562382536e-05, "loss": 3.3094, "step": 130500 }, { "epoch": 9.203316003934242, "grad_norm": 0.950018584728241, "learning_rate": 6.501373427786614e-05, "loss": 3.3108, "step": 131000 }, { "epoch": 9.238443164254601, "grad_norm": 1.0023565292358398, "learning_rate": 6.48691629319069e-05, "loss": 3.309, "step": 131500 }, { "epoch": 9.273570324574962, "grad_norm": 0.9661219120025635, "learning_rate": 6.472459158594767e-05, "loss": 3.3113, "step": 132000 }, { "epoch": 9.30869748489532, "grad_norm": 0.9721055626869202, "learning_rate": 6.458002023998843e-05, "loss": 3.3082, "step": 132500 }, { "epoch": 9.343824645215681, "grad_norm": 0.9974498152732849, "learning_rate": 6.443544889402921e-05, "loss": 3.3056, "step": 133000 }, { "epoch": 9.37895180553604, "grad_norm": 0.9510700702667236, "learning_rate": 6.429087754806997e-05, "loss": 3.3064, "step": 133500 }, { "epoch": 9.414078965856401, "grad_norm": 0.9505058526992798, "learning_rate": 6.414630620211074e-05, "loss": 3.304, "step": 134000 }, { "epoch": 9.44920612617676, "grad_norm": 0.9862971305847168, "learning_rate": 6.400173485615151e-05, "loss": 3.3017, "step": 134500 }, { "epoch": 9.484333286497119, "grad_norm": 0.9923340678215027, "learning_rate": 6.385716351019229e-05, "loss": 3.3012, "step": 135000 }, { "epoch": 9.51946044681748, "grad_norm": 0.9438260197639465, "learning_rate": 6.371259216423306e-05, "loss": 3.3014, "step": 135500 }, { "epoch": 9.554587607137838, "grad_norm": 1.0171432495117188, "learning_rate": 6.356802081827382e-05, "loss": 3.3013, "step": 136000 }, { "epoch": 9.589714767458199, "grad_norm": 1.0318517684936523, "learning_rate": 6.342344947231458e-05, "loss": 3.3025, "step": 136500 }, { "epoch": 9.624841927778558, "grad_norm": 0.9766752123832703, "learning_rate": 6.327887812635536e-05, "loss": 3.2982, "step": 137000 }, { "epoch": 9.659969088098919, "grad_norm": 1.015612244606018, "learning_rate": 6.313430678039613e-05, "loss": 3.2985, "step": 137500 }, { "epoch": 9.695096248419278, "grad_norm": 1.0225369930267334, "learning_rate": 6.29897354344369e-05, "loss": 3.2943, "step": 138000 }, { "epoch": 9.730223408739638, "grad_norm": 1.0277986526489258, "learning_rate": 6.284516408847767e-05, "loss": 3.3013, "step": 138500 }, { "epoch": 9.765350569059997, "grad_norm": 1.0086792707443237, "learning_rate": 6.270059274251843e-05, "loss": 3.2946, "step": 139000 }, { "epoch": 9.800477729380358, "grad_norm": 0.9675197005271912, "learning_rate": 6.255602139655921e-05, "loss": 3.2955, "step": 139500 }, { "epoch": 9.835604889700717, "grad_norm": 1.0447670221328735, "learning_rate": 6.241145005059997e-05, "loss": 3.2942, "step": 140000 }, { "epoch": 9.835604889700717, "eval_accuracy": 0.6654035466550119, "eval_loss": 2.998171091079712, "eval_runtime": 188.8315, "eval_samples_per_second": 639.226, "eval_steps_per_second": 3.998, "step": 140000 }, { "epoch": 9.870732050021076, "grad_norm": 0.9500594735145569, "learning_rate": 6.226687870464074e-05, "loss": 3.2944, "step": 140500 }, { "epoch": 9.905859210341436, "grad_norm": 1.0142775774002075, "learning_rate": 6.212230735868151e-05, "loss": 3.2922, "step": 141000 }, { "epoch": 9.940986370661795, "grad_norm": 0.9775634407997131, "learning_rate": 6.197773601272228e-05, "loss": 3.2935, "step": 141500 }, { "epoch": 9.976113530982156, "grad_norm": 0.9756135940551758, "learning_rate": 6.183316466676306e-05, "loss": 3.2897, "step": 142000 }, { "epoch": 10.011240691302515, "grad_norm": 0.994945228099823, "learning_rate": 6.168859332080382e-05, "loss": 3.2918, "step": 142500 }, { "epoch": 10.046367851622875, "grad_norm": 1.0026745796203613, "learning_rate": 6.154402197484458e-05, "loss": 3.2911, "step": 143000 }, { "epoch": 10.081495011943234, "grad_norm": 0.9682114720344543, "learning_rate": 6.139945062888536e-05, "loss": 3.2866, "step": 143500 }, { "epoch": 10.116622172263595, "grad_norm": 1.0748752355575562, "learning_rate": 6.125487928292613e-05, "loss": 3.286, "step": 144000 }, { "epoch": 10.151749332583954, "grad_norm": 0.9581155776977539, "learning_rate": 6.111030793696689e-05, "loss": 3.2874, "step": 144500 }, { "epoch": 10.186876492904315, "grad_norm": 1.0997650623321533, "learning_rate": 6.096573659100766e-05, "loss": 3.2864, "step": 145000 }, { "epoch": 10.222003653224673, "grad_norm": 1.036073088645935, "learning_rate": 6.082116524504844e-05, "loss": 3.2884, "step": 145500 }, { "epoch": 10.257130813545032, "grad_norm": 0.9324133396148682, "learning_rate": 6.067659389908921e-05, "loss": 3.288, "step": 146000 }, { "epoch": 10.292257973865393, "grad_norm": 0.9933338165283203, "learning_rate": 6.053202255312997e-05, "loss": 3.2815, "step": 146500 }, { "epoch": 10.327385134185752, "grad_norm": 1.0416404008865356, "learning_rate": 6.0387451207170744e-05, "loss": 3.2831, "step": 147000 }, { "epoch": 10.362512294506113, "grad_norm": 0.9802959561347961, "learning_rate": 6.024287986121151e-05, "loss": 3.2831, "step": 147500 }, { "epoch": 10.397639454826471, "grad_norm": 0.975288450717926, "learning_rate": 6.009830851525228e-05, "loss": 3.281, "step": 148000 }, { "epoch": 10.432766615146832, "grad_norm": 0.9557508826255798, "learning_rate": 5.995373716929304e-05, "loss": 3.2848, "step": 148500 }, { "epoch": 10.467893775467191, "grad_norm": 0.9647365808486938, "learning_rate": 5.980916582333381e-05, "loss": 3.2795, "step": 149000 }, { "epoch": 10.503020935787552, "grad_norm": 1.0546215772628784, "learning_rate": 5.966459447737459e-05, "loss": 3.2782, "step": 149500 }, { "epoch": 10.53814809610791, "grad_norm": 0.9828442931175232, "learning_rate": 5.952002313141536e-05, "loss": 3.2799, "step": 150000 }, { "epoch": 10.53814809610791, "eval_accuracy": 0.6671639322564016, "eval_loss": 2.9831271171569824, "eval_runtime": 185.657, "eval_samples_per_second": 650.156, "eval_steps_per_second": 4.067, "step": 150000 }, { "epoch": 10.573275256428271, "grad_norm": 0.9615412950515747, "learning_rate": 5.9375451785456126e-05, "loss": 3.2834, "step": 150500 }, { "epoch": 10.60840241674863, "grad_norm": 0.9674803018569946, "learning_rate": 5.9230880439496896e-05, "loss": 3.277, "step": 151000 }, { "epoch": 10.643529577068989, "grad_norm": 0.9678474068641663, "learning_rate": 5.908630909353766e-05, "loss": 3.2777, "step": 151500 }, { "epoch": 10.67865673738935, "grad_norm": 0.9506127238273621, "learning_rate": 5.894173774757843e-05, "loss": 3.2777, "step": 152000 }, { "epoch": 10.713783897709709, "grad_norm": 0.9495745897293091, "learning_rate": 5.8797166401619195e-05, "loss": 3.276, "step": 152500 }, { "epoch": 10.74891105803007, "grad_norm": 0.954150915145874, "learning_rate": 5.865259505565997e-05, "loss": 3.276, "step": 153000 }, { "epoch": 10.784038218350428, "grad_norm": 0.9836591482162476, "learning_rate": 5.8508023709700743e-05, "loss": 3.2775, "step": 153500 }, { "epoch": 10.819165378670789, "grad_norm": 0.9521194100379944, "learning_rate": 5.8363452363741514e-05, "loss": 3.275, "step": 154000 }, { "epoch": 10.854292538991148, "grad_norm": 0.9946979880332947, "learning_rate": 5.821888101778228e-05, "loss": 3.2767, "step": 154500 }, { "epoch": 10.889419699311508, "grad_norm": 0.9655107855796814, "learning_rate": 5.807430967182305e-05, "loss": 3.2756, "step": 155000 }, { "epoch": 10.924546859631867, "grad_norm": 0.9714339375495911, "learning_rate": 5.792973832586381e-05, "loss": 3.2752, "step": 155500 }, { "epoch": 10.959674019952226, "grad_norm": 1.0766863822937012, "learning_rate": 5.7785166979904584e-05, "loss": 3.2735, "step": 156000 }, { "epoch": 10.994801180272587, "grad_norm": 1.0243186950683594, "learning_rate": 5.764059563394536e-05, "loss": 3.2726, "step": 156500 }, { "epoch": 11.029928340592946, "grad_norm": 1.0681743621826172, "learning_rate": 5.7496024287986125e-05, "loss": 3.2663, "step": 157000 }, { "epoch": 11.065055500913306, "grad_norm": 0.9860267043113708, "learning_rate": 5.7351452942026896e-05, "loss": 3.2706, "step": 157500 }, { "epoch": 11.100182661233665, "grad_norm": 0.9802544713020325, "learning_rate": 5.720688159606766e-05, "loss": 3.2691, "step": 158000 }, { "epoch": 11.135309821554026, "grad_norm": 1.0232839584350586, "learning_rate": 5.706231025010843e-05, "loss": 3.2668, "step": 158500 }, { "epoch": 11.170436981874385, "grad_norm": 1.0054973363876343, "learning_rate": 5.6917738904149195e-05, "loss": 3.2686, "step": 159000 }, { "epoch": 11.205564142194746, "grad_norm": 1.0075745582580566, "learning_rate": 5.6773167558189966e-05, "loss": 3.2685, "step": 159500 }, { "epoch": 11.240691302515105, "grad_norm": 0.9830793142318726, "learning_rate": 5.6628596212230736e-05, "loss": 3.271, "step": 160000 }, { "epoch": 11.240691302515105, "eval_accuracy": 0.6686676073076556, "eval_loss": 2.9750185012817383, "eval_runtime": 183.5917, "eval_samples_per_second": 657.47, "eval_steps_per_second": 4.112, "step": 160000 }, { "epoch": 11.275818462835465, "grad_norm": 0.9880683422088623, "learning_rate": 5.6484024866271514e-05, "loss": 3.2645, "step": 160500 }, { "epoch": 11.310945623155824, "grad_norm": 1.0423779487609863, "learning_rate": 5.633945352031228e-05, "loss": 3.2694, "step": 161000 }, { "epoch": 11.346072783476183, "grad_norm": 1.021085500717163, "learning_rate": 5.619488217435305e-05, "loss": 3.2648, "step": 161500 }, { "epoch": 11.381199943796544, "grad_norm": 1.0164035558700562, "learning_rate": 5.605031082839381e-05, "loss": 3.2659, "step": 162000 }, { "epoch": 11.416327104116903, "grad_norm": 0.9693009257316589, "learning_rate": 5.5905739482434584e-05, "loss": 3.2622, "step": 162500 }, { "epoch": 11.451454264437263, "grad_norm": 0.9684035778045654, "learning_rate": 5.576116813647535e-05, "loss": 3.2589, "step": 163000 }, { "epoch": 11.486581424757622, "grad_norm": 0.969419002532959, "learning_rate": 5.561659679051612e-05, "loss": 3.2628, "step": 163500 }, { "epoch": 11.521708585077983, "grad_norm": 1.0042866468429565, "learning_rate": 5.5472025444556896e-05, "loss": 3.263, "step": 164000 }, { "epoch": 11.556835745398342, "grad_norm": 1.0603992938995361, "learning_rate": 5.532745409859767e-05, "loss": 3.2609, "step": 164500 }, { "epoch": 11.591962905718702, "grad_norm": 1.127294898033142, "learning_rate": 5.518288275263843e-05, "loss": 3.2613, "step": 165000 }, { "epoch": 11.627090066039061, "grad_norm": 1.0210075378417969, "learning_rate": 5.50383114066792e-05, "loss": 3.2615, "step": 165500 }, { "epoch": 11.662217226359422, "grad_norm": 1.0078051090240479, "learning_rate": 5.4893740060719965e-05, "loss": 3.2587, "step": 166000 }, { "epoch": 11.69734438667978, "grad_norm": 0.9809885621070862, "learning_rate": 5.4749168714760736e-05, "loss": 3.259, "step": 166500 }, { "epoch": 11.73247154700014, "grad_norm": 1.0912126302719116, "learning_rate": 5.46045973688015e-05, "loss": 3.2547, "step": 167000 }, { "epoch": 11.7675987073205, "grad_norm": 1.0941123962402344, "learning_rate": 5.446002602284227e-05, "loss": 3.2557, "step": 167500 }, { "epoch": 11.80272586764086, "grad_norm": 0.997536838054657, "learning_rate": 5.431545467688305e-05, "loss": 3.2576, "step": 168000 }, { "epoch": 11.83785302796122, "grad_norm": 0.9559796452522278, "learning_rate": 5.417088333092382e-05, "loss": 3.2553, "step": 168500 }, { "epoch": 11.872980188281579, "grad_norm": 0.9806485176086426, "learning_rate": 5.402631198496458e-05, "loss": 3.2544, "step": 169000 }, { "epoch": 11.90810734860194, "grad_norm": 1.0185707807540894, "learning_rate": 5.3881740639005354e-05, "loss": 3.2542, "step": 169500 }, { "epoch": 11.943234508922298, "grad_norm": 0.9899264574050903, "learning_rate": 5.373716929304612e-05, "loss": 3.2545, "step": 170000 }, { "epoch": 11.943234508922298, "eval_accuracy": 0.6703399472442935, "eval_loss": 2.9623563289642334, "eval_runtime": 182.8722, "eval_samples_per_second": 660.057, "eval_steps_per_second": 4.129, "step": 170000 }, { "epoch": 11.978361669242659, "grad_norm": 0.9951636791229248, "learning_rate": 5.359259794708689e-05, "loss": 3.2541, "step": 170500 }, { "epoch": 12.013488829563018, "grad_norm": 1.0922133922576904, "learning_rate": 5.344802660112765e-05, "loss": 3.2548, "step": 171000 }, { "epoch": 12.048615989883379, "grad_norm": 1.0029044151306152, "learning_rate": 5.330345525516843e-05, "loss": 3.2531, "step": 171500 }, { "epoch": 12.083743150203738, "grad_norm": 0.9556426405906677, "learning_rate": 5.31588839092092e-05, "loss": 3.2494, "step": 172000 }, { "epoch": 12.118870310524096, "grad_norm": 0.9693633317947388, "learning_rate": 5.3014312563249965e-05, "loss": 3.2518, "step": 172500 }, { "epoch": 12.153997470844457, "grad_norm": 1.0057485103607178, "learning_rate": 5.2869741217290736e-05, "loss": 3.2548, "step": 173000 }, { "epoch": 12.189124631164816, "grad_norm": 0.9814544916152954, "learning_rate": 5.272516987133151e-05, "loss": 3.2509, "step": 173500 }, { "epoch": 12.224251791485177, "grad_norm": 1.0081578493118286, "learning_rate": 5.258059852537227e-05, "loss": 3.2461, "step": 174000 }, { "epoch": 12.259378951805536, "grad_norm": 0.9391810894012451, "learning_rate": 5.243602717941304e-05, "loss": 3.2518, "step": 174500 }, { "epoch": 12.294506112125896, "grad_norm": 0.9642152786254883, "learning_rate": 5.229145583345382e-05, "loss": 3.2459, "step": 175000 }, { "epoch": 12.329633272446255, "grad_norm": 0.9983788728713989, "learning_rate": 5.214688448749458e-05, "loss": 3.2501, "step": 175500 }, { "epoch": 12.364760432766616, "grad_norm": 1.0108500719070435, "learning_rate": 5.2002313141535354e-05, "loss": 3.2466, "step": 176000 }, { "epoch": 12.399887593086975, "grad_norm": 0.9797979593276978, "learning_rate": 5.185774179557612e-05, "loss": 3.2461, "step": 176500 }, { "epoch": 12.435014753407334, "grad_norm": 1.0821365118026733, "learning_rate": 5.171317044961689e-05, "loss": 3.2473, "step": 177000 }, { "epoch": 12.470141913727694, "grad_norm": 1.174897313117981, "learning_rate": 5.156859910365765e-05, "loss": 3.2452, "step": 177500 }, { "epoch": 12.505269074048053, "grad_norm": 1.1057952642440796, "learning_rate": 5.1424027757698423e-05, "loss": 3.2479, "step": 178000 }, { "epoch": 12.540396234368414, "grad_norm": 0.9830456972122192, "learning_rate": 5.127945641173919e-05, "loss": 3.2483, "step": 178500 }, { "epoch": 12.575523394688773, "grad_norm": 1.0125759840011597, "learning_rate": 5.113488506577997e-05, "loss": 3.2446, "step": 179000 }, { "epoch": 12.610650555009133, "grad_norm": 1.0111823081970215, "learning_rate": 5.0990313719820736e-05, "loss": 3.2446, "step": 179500 }, { "epoch": 12.645777715329492, "grad_norm": 0.9823210835456848, "learning_rate": 5.0845742373861507e-05, "loss": 3.2444, "step": 180000 }, { "epoch": 12.645777715329492, "eval_accuracy": 0.6723193510957107, "eval_loss": 2.9492604732513428, "eval_runtime": 177.8714, "eval_samples_per_second": 678.614, "eval_steps_per_second": 4.245, "step": 180000 }, { "epoch": 12.680904875649853, "grad_norm": 1.0771335363388062, "learning_rate": 5.070117102790227e-05, "loss": 3.2415, "step": 180500 }, { "epoch": 12.716032035970212, "grad_norm": 1.0094541311264038, "learning_rate": 5.055659968194304e-05, "loss": 3.2434, "step": 181000 }, { "epoch": 12.751159196290573, "grad_norm": 0.9946825504302979, "learning_rate": 5.0412028335983805e-05, "loss": 3.2444, "step": 181500 }, { "epoch": 12.786286356610931, "grad_norm": 0.9406523108482361, "learning_rate": 5.0267456990024576e-05, "loss": 3.2435, "step": 182000 }, { "epoch": 12.82141351693129, "grad_norm": 1.0183651447296143, "learning_rate": 5.0122885644065354e-05, "loss": 3.2461, "step": 182500 }, { "epoch": 12.856540677251651, "grad_norm": 1.0511256456375122, "learning_rate": 4.997831429810612e-05, "loss": 3.2416, "step": 183000 }, { "epoch": 12.89166783757201, "grad_norm": 1.1275522708892822, "learning_rate": 4.983374295214689e-05, "loss": 3.2413, "step": 183500 }, { "epoch": 12.92679499789237, "grad_norm": 0.9592730402946472, "learning_rate": 4.968917160618766e-05, "loss": 3.238, "step": 184000 }, { "epoch": 12.96192215821273, "grad_norm": 0.9847075343132019, "learning_rate": 4.954460026022842e-05, "loss": 3.243, "step": 184500 }, { "epoch": 12.99704931853309, "grad_norm": 1.0094618797302246, "learning_rate": 4.9400028914269194e-05, "loss": 3.2405, "step": 185000 }, { "epoch": 13.032176478853449, "grad_norm": 0.9838051199913025, "learning_rate": 4.9255457568309965e-05, "loss": 3.2361, "step": 185500 }, { "epoch": 13.06730363917381, "grad_norm": 1.119706630706787, "learning_rate": 4.9110886222350736e-05, "loss": 3.236, "step": 186000 }, { "epoch": 13.102430799494169, "grad_norm": 1.129960536956787, "learning_rate": 4.89663148763915e-05, "loss": 3.2374, "step": 186500 }, { "epoch": 13.13755795981453, "grad_norm": 1.002439022064209, "learning_rate": 4.882174353043227e-05, "loss": 3.2339, "step": 187000 }, { "epoch": 13.172685120134888, "grad_norm": 0.9733322262763977, "learning_rate": 4.867717218447304e-05, "loss": 3.2332, "step": 187500 }, { "epoch": 13.207812280455247, "grad_norm": 1.0108827352523804, "learning_rate": 4.853260083851381e-05, "loss": 3.2356, "step": 188000 }, { "epoch": 13.242939440775608, "grad_norm": 1.054499864578247, "learning_rate": 4.8388029492554576e-05, "loss": 3.2344, "step": 188500 }, { "epoch": 13.278066601095967, "grad_norm": 0.9512261152267456, "learning_rate": 4.824345814659535e-05, "loss": 3.2333, "step": 189000 }, { "epoch": 13.313193761416327, "grad_norm": 1.063110113143921, "learning_rate": 4.809888680063612e-05, "loss": 3.2367, "step": 189500 }, { "epoch": 13.348320921736686, "grad_norm": 0.9670858383178711, "learning_rate": 4.795431545467688e-05, "loss": 3.2336, "step": 190000 }, { "epoch": 13.348320921736686, "eval_accuracy": 0.6731444141876055, "eval_loss": 2.942819595336914, "eval_runtime": 177.6004, "eval_samples_per_second": 679.649, "eval_steps_per_second": 4.251, "step": 190000 }, { "epoch": 13.383448082057047, "grad_norm": 1.0234127044677734, "learning_rate": 4.780974410871765e-05, "loss": 3.2295, "step": 190500 }, { "epoch": 13.418575242377406, "grad_norm": 0.9976109266281128, "learning_rate": 4.766517276275842e-05, "loss": 3.2334, "step": 191000 }, { "epoch": 13.453702402697767, "grad_norm": 0.9884486198425293, "learning_rate": 4.7520601416799194e-05, "loss": 3.2296, "step": 191500 }, { "epoch": 13.488829563018125, "grad_norm": 1.0319344997406006, "learning_rate": 4.737603007083996e-05, "loss": 3.2322, "step": 192000 }, { "epoch": 13.523956723338486, "grad_norm": 0.9803874492645264, "learning_rate": 4.723145872488073e-05, "loss": 3.2337, "step": 192500 }, { "epoch": 13.559083883658845, "grad_norm": 0.989548921585083, "learning_rate": 4.70868873789215e-05, "loss": 3.2307, "step": 193000 }, { "epoch": 13.594211043979204, "grad_norm": 1.0503300428390503, "learning_rate": 4.694231603296227e-05, "loss": 3.2313, "step": 193500 }, { "epoch": 13.629338204299565, "grad_norm": 1.063338279724121, "learning_rate": 4.6797744687003034e-05, "loss": 3.2285, "step": 194000 }, { "epoch": 13.664465364619923, "grad_norm": 0.9926460981369019, "learning_rate": 4.665317334104381e-05, "loss": 3.2277, "step": 194500 }, { "epoch": 13.699592524940284, "grad_norm": 1.0616544485092163, "learning_rate": 4.6508601995084576e-05, "loss": 3.2295, "step": 195000 }, { "epoch": 13.734719685260643, "grad_norm": 1.0530409812927246, "learning_rate": 4.6364030649125346e-05, "loss": 3.2284, "step": 195500 }, { "epoch": 13.769846845581004, "grad_norm": 0.9996173977851868, "learning_rate": 4.621945930316611e-05, "loss": 3.228, "step": 196000 }, { "epoch": 13.804974005901363, "grad_norm": 1.060828685760498, "learning_rate": 4.607488795720689e-05, "loss": 3.226, "step": 196500 }, { "epoch": 13.840101166221723, "grad_norm": 0.968076765537262, "learning_rate": 4.593031661124765e-05, "loss": 3.2275, "step": 197000 }, { "epoch": 13.875228326542082, "grad_norm": 1.0310499668121338, "learning_rate": 4.578574526528842e-05, "loss": 3.2255, "step": 197500 }, { "epoch": 13.910355486862443, "grad_norm": 1.092532992362976, "learning_rate": 4.564117391932919e-05, "loss": 3.225, "step": 198000 }, { "epoch": 13.945482647182802, "grad_norm": 1.133365273475647, "learning_rate": 4.5496602573369964e-05, "loss": 3.2245, "step": 198500 }, { "epoch": 13.98060980750316, "grad_norm": 1.0174368619918823, "learning_rate": 4.535203122741073e-05, "loss": 3.2306, "step": 199000 }, { "epoch": 14.015736967823521, "grad_norm": 1.0136277675628662, "learning_rate": 4.52074598814515e-05, "loss": 3.2239, "step": 199500 }, { "epoch": 14.05086412814388, "grad_norm": 0.9779367446899414, "learning_rate": 4.506288853549227e-05, "loss": 3.2254, "step": 200000 }, { "epoch": 14.05086412814388, "eval_accuracy": 0.6745767911163286, "eval_loss": 2.9315974712371826, "eval_runtime": 179.7076, "eval_samples_per_second": 671.68, "eval_steps_per_second": 4.201, "step": 200000 }, { "epoch": 14.08599128846424, "grad_norm": 0.9721986651420593, "learning_rate": 4.491831718953304e-05, "loss": 3.2212, "step": 200500 }, { "epoch": 14.1211184487846, "grad_norm": 1.0022870302200317, "learning_rate": 4.4773745843573805e-05, "loss": 3.2231, "step": 201000 }, { "epoch": 14.15624560910496, "grad_norm": 0.9985385537147522, "learning_rate": 4.4629174497614575e-05, "loss": 3.2214, "step": 201500 }, { "epoch": 14.19137276942532, "grad_norm": 1.0058878660202026, "learning_rate": 4.4484603151655346e-05, "loss": 3.2197, "step": 202000 }, { "epoch": 14.22649992974568, "grad_norm": 1.0070774555206299, "learning_rate": 4.434003180569612e-05, "loss": 3.2215, "step": 202500 }, { "epoch": 14.261627090066039, "grad_norm": 1.0218665599822998, "learning_rate": 4.419546045973688e-05, "loss": 3.2223, "step": 203000 }, { "epoch": 14.2967542503864, "grad_norm": 1.0088073015213013, "learning_rate": 4.405088911377765e-05, "loss": 3.2201, "step": 203500 }, { "epoch": 14.331881410706758, "grad_norm": 1.0248451232910156, "learning_rate": 4.390631776781842e-05, "loss": 3.2225, "step": 204000 }, { "epoch": 14.367008571027117, "grad_norm": 1.0222102403640747, "learning_rate": 4.3761746421859187e-05, "loss": 3.2214, "step": 204500 }, { "epoch": 14.402135731347478, "grad_norm": 1.008377194404602, "learning_rate": 4.361717507589996e-05, "loss": 3.2192, "step": 205000 }, { "epoch": 14.437262891667837, "grad_norm": 1.0465419292449951, "learning_rate": 4.347260372994073e-05, "loss": 3.2205, "step": 205500 }, { "epoch": 14.472390051988198, "grad_norm": 1.0380550622940063, "learning_rate": 4.33280323839815e-05, "loss": 3.2191, "step": 206000 }, { "epoch": 14.507517212308557, "grad_norm": 1.0289161205291748, "learning_rate": 4.318346103802226e-05, "loss": 3.217, "step": 206500 }, { "epoch": 14.542644372628917, "grad_norm": 1.006243109703064, "learning_rate": 4.3038889692063034e-05, "loss": 3.2161, "step": 207000 }, { "epoch": 14.577771532949276, "grad_norm": 1.1275827884674072, "learning_rate": 4.2894318346103804e-05, "loss": 3.2159, "step": 207500 }, { "epoch": 14.612898693269637, "grad_norm": 1.0984758138656616, "learning_rate": 4.2749747000144575e-05, "loss": 3.218, "step": 208000 }, { "epoch": 14.648025853589996, "grad_norm": 0.9622736573219299, "learning_rate": 4.260517565418534e-05, "loss": 3.2157, "step": 208500 }, { "epoch": 14.683153013910356, "grad_norm": 1.04253351688385, "learning_rate": 4.246060430822611e-05, "loss": 3.2166, "step": 209000 }, { "epoch": 14.718280174230715, "grad_norm": 1.1333566904067993, "learning_rate": 4.231603296226688e-05, "loss": 3.215, "step": 209500 }, { "epoch": 14.753407334551074, "grad_norm": 1.0053073167800903, "learning_rate": 4.217146161630765e-05, "loss": 3.2143, "step": 210000 }, { "epoch": 14.753407334551074, "eval_accuracy": 0.6758909772748868, "eval_loss": 2.9231269359588623, "eval_runtime": 177.1572, "eval_samples_per_second": 681.35, "eval_steps_per_second": 4.262, "step": 210000 }, { "epoch": 14.788534494871435, "grad_norm": 1.093920111656189, "learning_rate": 4.2026890270348416e-05, "loss": 3.2142, "step": 210500 }, { "epoch": 14.823661655191794, "grad_norm": 1.1499086618423462, "learning_rate": 4.1882318924389186e-05, "loss": 3.2139, "step": 211000 }, { "epoch": 14.858788815512154, "grad_norm": 1.1073871850967407, "learning_rate": 4.173774757842996e-05, "loss": 3.2157, "step": 211500 }, { "epoch": 14.893915975832513, "grad_norm": 1.0481630563735962, "learning_rate": 4.159317623247073e-05, "loss": 3.2129, "step": 212000 }, { "epoch": 14.929043136152874, "grad_norm": 1.0672996044158936, "learning_rate": 4.144860488651149e-05, "loss": 3.2142, "step": 212500 }, { "epoch": 14.964170296473233, "grad_norm": 0.9917321801185608, "learning_rate": 4.130403354055227e-05, "loss": 3.2151, "step": 213000 }, { "epoch": 14.999297456793593, "grad_norm": 0.9837495684623718, "learning_rate": 4.1159462194593033e-05, "loss": 3.2147, "step": 213500 }, { "epoch": 15.034424617113952, "grad_norm": 1.0623294115066528, "learning_rate": 4.1014890848633804e-05, "loss": 3.2089, "step": 214000 }, { "epoch": 15.069551777434313, "grad_norm": 1.0003976821899414, "learning_rate": 4.087031950267457e-05, "loss": 3.2106, "step": 214500 }, { "epoch": 15.104678937754672, "grad_norm": 0.9986251592636108, "learning_rate": 4.0725748156715346e-05, "loss": 3.2113, "step": 215000 }, { "epoch": 15.13980609807503, "grad_norm": 1.0745850801467896, "learning_rate": 4.058117681075611e-05, "loss": 3.2078, "step": 215500 }, { "epoch": 15.174933258395392, "grad_norm": 1.0086036920547485, "learning_rate": 4.043660546479688e-05, "loss": 3.211, "step": 216000 }, { "epoch": 15.21006041871575, "grad_norm": 0.9925144910812378, "learning_rate": 4.0292034118837645e-05, "loss": 3.2102, "step": 216500 }, { "epoch": 15.245187579036111, "grad_norm": 1.026578426361084, "learning_rate": 4.014746277287842e-05, "loss": 3.2094, "step": 217000 }, { "epoch": 15.28031473935647, "grad_norm": 0.977218747138977, "learning_rate": 4.0002891426919186e-05, "loss": 3.21, "step": 217500 }, { "epoch": 15.31544189967683, "grad_norm": 0.9964417219161987, "learning_rate": 3.985832008095996e-05, "loss": 3.2088, "step": 218000 }, { "epoch": 15.35056905999719, "grad_norm": 1.049068808555603, "learning_rate": 3.971374873500072e-05, "loss": 3.2096, "step": 218500 }, { "epoch": 15.38569622031755, "grad_norm": 1.0806366205215454, "learning_rate": 3.95691773890415e-05, "loss": 3.2075, "step": 219000 }, { "epoch": 15.42082338063791, "grad_norm": 1.0102629661560059, "learning_rate": 3.942460604308226e-05, "loss": 3.2083, "step": 219500 }, { "epoch": 15.45595054095827, "grad_norm": 1.0897108316421509, "learning_rate": 3.928003469712303e-05, "loss": 3.2058, "step": 220000 }, { "epoch": 15.45595054095827, "eval_accuracy": 0.6771663097769833, "eval_loss": 2.9153566360473633, "eval_runtime": 173.1439, "eval_samples_per_second": 697.143, "eval_steps_per_second": 4.361, "step": 220000 }, { "epoch": 15.491077701278629, "grad_norm": 1.016409158706665, "learning_rate": 3.9135463351163804e-05, "loss": 3.2076, "step": 220500 }, { "epoch": 15.526204861598988, "grad_norm": 1.2338467836380005, "learning_rate": 3.899089200520457e-05, "loss": 3.2086, "step": 221000 }, { "epoch": 15.561332021919348, "grad_norm": 1.0713613033294678, "learning_rate": 3.884632065924534e-05, "loss": 3.2069, "step": 221500 }, { "epoch": 15.596459182239707, "grad_norm": 1.003710150718689, "learning_rate": 3.870174931328611e-05, "loss": 3.206, "step": 222000 }, { "epoch": 15.631586342560068, "grad_norm": 1.0036053657531738, "learning_rate": 3.855717796732688e-05, "loss": 3.2039, "step": 222500 }, { "epoch": 15.666713502880427, "grad_norm": 1.0012390613555908, "learning_rate": 3.8412606621367644e-05, "loss": 3.2072, "step": 223000 }, { "epoch": 15.701840663200787, "grad_norm": 1.0264482498168945, "learning_rate": 3.8268035275408415e-05, "loss": 3.2058, "step": 223500 }, { "epoch": 15.736967823521146, "grad_norm": 1.0620832443237305, "learning_rate": 3.812346392944918e-05, "loss": 3.2036, "step": 224000 }, { "epoch": 15.772094983841507, "grad_norm": 1.0413259267807007, "learning_rate": 3.797889258348996e-05, "loss": 3.2021, "step": 224500 }, { "epoch": 15.807222144161866, "grad_norm": 1.0166760683059692, "learning_rate": 3.783432123753072e-05, "loss": 3.2007, "step": 225000 }, { "epoch": 15.842349304482227, "grad_norm": 1.0761804580688477, "learning_rate": 3.768974989157149e-05, "loss": 3.2008, "step": 225500 }, { "epoch": 15.877476464802585, "grad_norm": 0.986355721950531, "learning_rate": 3.754517854561226e-05, "loss": 3.2052, "step": 226000 }, { "epoch": 15.912603625122944, "grad_norm": 1.0473321676254272, "learning_rate": 3.740060719965303e-05, "loss": 3.2043, "step": 226500 }, { "epoch": 15.947730785443305, "grad_norm": 0.9840390682220459, "learning_rate": 3.72560358536938e-05, "loss": 3.2019, "step": 227000 }, { "epoch": 15.982857945763664, "grad_norm": 1.1408594846725464, "learning_rate": 3.711146450773457e-05, "loss": 3.2033, "step": 227500 }, { "epoch": 16.017985106084023, "grad_norm": 1.0580511093139648, "learning_rate": 3.696689316177534e-05, "loss": 3.1989, "step": 228000 }, { "epoch": 16.053112266404383, "grad_norm": 1.237089991569519, "learning_rate": 3.682232181581611e-05, "loss": 3.1974, "step": 228500 }, { "epoch": 16.088239426724744, "grad_norm": 1.1202374696731567, "learning_rate": 3.667775046985687e-05, "loss": 3.1984, "step": 229000 }, { "epoch": 16.123366587045105, "grad_norm": 1.0699679851531982, "learning_rate": 3.6533179123897644e-05, "loss": 3.1988, "step": 229500 }, { "epoch": 16.158493747365462, "grad_norm": 1.0159801244735718, "learning_rate": 3.6388607777938415e-05, "loss": 3.2014, "step": 230000 }, { "epoch": 16.158493747365462, "eval_accuracy": 0.6780014612548544, "eval_loss": 2.909538745880127, "eval_runtime": 172.134, "eval_samples_per_second": 701.233, "eval_steps_per_second": 4.386, "step": 230000 }, { "epoch": 16.193620907685823, "grad_norm": 1.0414690971374512, "learning_rate": 3.6244036431979186e-05, "loss": 3.1964, "step": 230500 }, { "epoch": 16.228748068006183, "grad_norm": 1.0194602012634277, "learning_rate": 3.609946508601995e-05, "loss": 3.1986, "step": 231000 }, { "epoch": 16.26387522832654, "grad_norm": 1.0621025562286377, "learning_rate": 3.595489374006073e-05, "loss": 3.1951, "step": 231500 }, { "epoch": 16.2990023886469, "grad_norm": 1.0132440328598022, "learning_rate": 3.581032239410149e-05, "loss": 3.1996, "step": 232000 }, { "epoch": 16.33412954896726, "grad_norm": 1.1425501108169556, "learning_rate": 3.566575104814226e-05, "loss": 3.196, "step": 232500 }, { "epoch": 16.369256709287622, "grad_norm": 1.0304450988769531, "learning_rate": 3.5521179702183026e-05, "loss": 3.1971, "step": 233000 }, { "epoch": 16.40438386960798, "grad_norm": 1.0570875406265259, "learning_rate": 3.5376608356223804e-05, "loss": 3.1965, "step": 233500 }, { "epoch": 16.43951102992834, "grad_norm": 0.9665914177894592, "learning_rate": 3.523203701026457e-05, "loss": 3.1958, "step": 234000 }, { "epoch": 16.4746381902487, "grad_norm": 1.2075748443603516, "learning_rate": 3.508746566430534e-05, "loss": 3.1979, "step": 234500 }, { "epoch": 16.50976535056906, "grad_norm": 0.993384838104248, "learning_rate": 3.49428943183461e-05, "loss": 3.1957, "step": 235000 }, { "epoch": 16.54489251088942, "grad_norm": 1.0302695035934448, "learning_rate": 3.479832297238687e-05, "loss": 3.193, "step": 235500 }, { "epoch": 16.58001967120978, "grad_norm": 1.0362904071807861, "learning_rate": 3.4653751626427644e-05, "loss": 3.1965, "step": 236000 }, { "epoch": 16.61514683153014, "grad_norm": 1.0113791227340698, "learning_rate": 3.4509180280468415e-05, "loss": 3.1931, "step": 236500 }, { "epoch": 16.650273991850497, "grad_norm": 1.0897748470306396, "learning_rate": 3.436460893450918e-05, "loss": 3.2013, "step": 237000 }, { "epoch": 16.685401152170858, "grad_norm": 1.1453965902328491, "learning_rate": 3.422003758854995e-05, "loss": 3.1944, "step": 237500 }, { "epoch": 16.72052831249122, "grad_norm": 1.0244419574737549, "learning_rate": 3.407546624259072e-05, "loss": 3.1939, "step": 238000 }, { "epoch": 16.75565547281158, "grad_norm": 1.048883080482483, "learning_rate": 3.393089489663149e-05, "loss": 3.1956, "step": 238500 }, { "epoch": 16.790782633131936, "grad_norm": 1.0635311603546143, "learning_rate": 3.378632355067226e-05, "loss": 3.1919, "step": 239000 }, { "epoch": 16.825909793452297, "grad_norm": 1.0601235628128052, "learning_rate": 3.3641752204713026e-05, "loss": 3.1919, "step": 239500 }, { "epoch": 16.861036953772658, "grad_norm": 0.9861056804656982, "learning_rate": 3.3497180858753797e-05, "loss": 3.1923, "step": 240000 }, { "epoch": 16.861036953772658, "eval_accuracy": 0.6788273287342915, "eval_loss": 2.904675006866455, "eval_runtime": 172.5993, "eval_samples_per_second": 699.342, "eval_steps_per_second": 4.374, "step": 240000 }, { "epoch": 16.89616411409302, "grad_norm": 1.1252508163452148, "learning_rate": 3.335260951279456e-05, "loss": 3.1908, "step": 240500 }, { "epoch": 16.931291274413375, "grad_norm": 1.0311353206634521, "learning_rate": 3.320803816683534e-05, "loss": 3.1951, "step": 241000 }, { "epoch": 16.966418434733736, "grad_norm": 1.0423134565353394, "learning_rate": 3.30634668208761e-05, "loss": 3.1906, "step": 241500 }, { "epoch": 17.001545595054097, "grad_norm": 1.0416908264160156, "learning_rate": 3.291889547491687e-05, "loss": 3.1963, "step": 242000 }, { "epoch": 17.036672755374454, "grad_norm": 1.032296061515808, "learning_rate": 3.277432412895764e-05, "loss": 3.1929, "step": 242500 }, { "epoch": 17.071799915694815, "grad_norm": 1.0529314279556274, "learning_rate": 3.2629752782998414e-05, "loss": 3.1898, "step": 243000 }, { "epoch": 17.106927076015175, "grad_norm": 1.0260303020477295, "learning_rate": 3.248518143703918e-05, "loss": 3.1891, "step": 243500 }, { "epoch": 17.142054236335536, "grad_norm": 1.1377393007278442, "learning_rate": 3.234061009107995e-05, "loss": 3.188, "step": 244000 }, { "epoch": 17.177181396655893, "grad_norm": 1.0354127883911133, "learning_rate": 3.219603874512072e-05, "loss": 3.1872, "step": 244500 }, { "epoch": 17.212308556976254, "grad_norm": 1.0202113389968872, "learning_rate": 3.205146739916149e-05, "loss": 3.1914, "step": 245000 }, { "epoch": 17.247435717296614, "grad_norm": 1.0626760721206665, "learning_rate": 3.1906896053202255e-05, "loss": 3.1869, "step": 245500 }, { "epoch": 17.282562877616975, "grad_norm": 1.0529558658599854, "learning_rate": 3.1762324707243026e-05, "loss": 3.1909, "step": 246000 }, { "epoch": 17.317690037937332, "grad_norm": 1.0054688453674316, "learning_rate": 3.1617753361283796e-05, "loss": 3.1872, "step": 246500 }, { "epoch": 17.352817198257693, "grad_norm": 0.9858493208885193, "learning_rate": 3.147318201532457e-05, "loss": 3.1869, "step": 247000 }, { "epoch": 17.387944358578054, "grad_norm": 1.0260921716690063, "learning_rate": 3.132861066936533e-05, "loss": 3.1897, "step": 247500 }, { "epoch": 17.42307151889841, "grad_norm": 1.0405819416046143, "learning_rate": 3.11840393234061e-05, "loss": 3.1827, "step": 248000 }, { "epoch": 17.45819867921877, "grad_norm": 1.026649832725525, "learning_rate": 3.103946797744687e-05, "loss": 3.1859, "step": 248500 }, { "epoch": 17.493325839539132, "grad_norm": 1.2595336437225342, "learning_rate": 3.0894896631487643e-05, "loss": 3.1835, "step": 249000 }, { "epoch": 17.528452999859493, "grad_norm": 1.0458557605743408, "learning_rate": 3.075032528552841e-05, "loss": 3.1825, "step": 249500 }, { "epoch": 17.56358016017985, "grad_norm": 1.0119978189468384, "learning_rate": 3.060575393956918e-05, "loss": 3.1846, "step": 250000 }, { "epoch": 17.56358016017985, "eval_accuracy": 0.679679251707914, "eval_loss": 2.8981587886810303, "eval_runtime": 178.3449, "eval_samples_per_second": 676.812, "eval_steps_per_second": 4.233, "step": 250000 }, { "epoch": 17.59870732050021, "grad_norm": 1.0216857194900513, "learning_rate": 3.046118259360995e-05, "loss": 3.1895, "step": 250500 }, { "epoch": 17.63383448082057, "grad_norm": 1.0001978874206543, "learning_rate": 3.0316611247650716e-05, "loss": 3.1828, "step": 251000 }, { "epoch": 17.668961641140932, "grad_norm": 1.1086599826812744, "learning_rate": 3.0172039901691484e-05, "loss": 3.1848, "step": 251500 }, { "epoch": 17.70408880146129, "grad_norm": 1.064321756362915, "learning_rate": 3.0027468555732258e-05, "loss": 3.1868, "step": 252000 }, { "epoch": 17.73921596178165, "grad_norm": 1.0282937288284302, "learning_rate": 2.9882897209773025e-05, "loss": 3.1854, "step": 252500 }, { "epoch": 17.77434312210201, "grad_norm": 1.0527218580245972, "learning_rate": 2.9738325863813793e-05, "loss": 3.1843, "step": 253000 }, { "epoch": 17.809470282422367, "grad_norm": 1.025094747543335, "learning_rate": 2.959375451785456e-05, "loss": 3.186, "step": 253500 }, { "epoch": 17.844597442742728, "grad_norm": 1.0321406126022339, "learning_rate": 2.9449183171895334e-05, "loss": 3.1824, "step": 254000 }, { "epoch": 17.87972460306309, "grad_norm": 0.9949146509170532, "learning_rate": 2.93046118259361e-05, "loss": 3.1875, "step": 254500 }, { "epoch": 17.91485176338345, "grad_norm": 1.144135594367981, "learning_rate": 2.916004047997687e-05, "loss": 3.1856, "step": 255000 }, { "epoch": 17.949978923703807, "grad_norm": 1.0128605365753174, "learning_rate": 2.9015469134017636e-05, "loss": 3.1856, "step": 255500 }, { "epoch": 17.985106084024167, "grad_norm": 1.1451584100723267, "learning_rate": 2.887089778805841e-05, "loss": 3.1819, "step": 256000 }, { "epoch": 18.020233244344528, "grad_norm": 1.0377037525177002, "learning_rate": 2.8726326442099178e-05, "loss": 3.1848, "step": 256500 }, { "epoch": 18.05536040466489, "grad_norm": 1.0408658981323242, "learning_rate": 2.8581755096139945e-05, "loss": 3.1833, "step": 257000 }, { "epoch": 18.090487564985246, "grad_norm": 1.1872942447662354, "learning_rate": 2.843718375018072e-05, "loss": 3.1789, "step": 257500 }, { "epoch": 18.125614725305606, "grad_norm": 1.082833170890808, "learning_rate": 2.8292612404221487e-05, "loss": 3.183, "step": 258000 }, { "epoch": 18.160741885625967, "grad_norm": 1.0577306747436523, "learning_rate": 2.8148041058262254e-05, "loss": 3.1816, "step": 258500 }, { "epoch": 18.195869045946324, "grad_norm": 1.0363332033157349, "learning_rate": 2.800346971230302e-05, "loss": 3.1813, "step": 259000 }, { "epoch": 18.230996206266685, "grad_norm": 1.0489052534103394, "learning_rate": 2.7858898366343792e-05, "loss": 3.1818, "step": 259500 }, { "epoch": 18.266123366587045, "grad_norm": 1.089944839477539, "learning_rate": 2.7714327020384563e-05, "loss": 3.1797, "step": 260000 }, { "epoch": 18.266123366587045, "eval_accuracy": 0.6805455812064435, "eval_loss": 2.8921754360198975, "eval_runtime": 177.4594, "eval_samples_per_second": 680.189, "eval_steps_per_second": 4.254, "step": 260000 }, { "epoch": 18.301250526907406, "grad_norm": 1.0893577337265015, "learning_rate": 2.756975567442533e-05, "loss": 3.1797, "step": 260500 }, { "epoch": 18.336377687227763, "grad_norm": 1.036550760269165, "learning_rate": 2.7425184328466098e-05, "loss": 3.1781, "step": 261000 }, { "epoch": 18.371504847548124, "grad_norm": 1.091452717781067, "learning_rate": 2.728061298250687e-05, "loss": 3.1818, "step": 261500 }, { "epoch": 18.406632007868485, "grad_norm": 1.0211600065231323, "learning_rate": 2.7136041636547636e-05, "loss": 3.1791, "step": 262000 }, { "epoch": 18.441759168188845, "grad_norm": 1.03214693069458, "learning_rate": 2.6991470290588407e-05, "loss": 3.1753, "step": 262500 }, { "epoch": 18.476886328509202, "grad_norm": 1.0473188161849976, "learning_rate": 2.6846898944629178e-05, "loss": 3.1766, "step": 263000 }, { "epoch": 18.512013488829563, "grad_norm": 1.045701026916504, "learning_rate": 2.6702327598669945e-05, "loss": 3.1762, "step": 263500 }, { "epoch": 18.547140649149924, "grad_norm": 1.064491629600525, "learning_rate": 2.6557756252710713e-05, "loss": 3.1754, "step": 264000 }, { "epoch": 18.58226780947028, "grad_norm": 1.0177878141403198, "learning_rate": 2.641318490675148e-05, "loss": 3.1757, "step": 264500 }, { "epoch": 18.61739496979064, "grad_norm": 1.125398874282837, "learning_rate": 2.6268613560792254e-05, "loss": 3.1806, "step": 265000 }, { "epoch": 18.652522130111002, "grad_norm": 1.0414868593215942, "learning_rate": 2.612404221483302e-05, "loss": 3.1771, "step": 265500 }, { "epoch": 18.687649290431363, "grad_norm": 1.0502114295959473, "learning_rate": 2.597947086887379e-05, "loss": 3.1741, "step": 266000 }, { "epoch": 18.72277645075172, "grad_norm": 1.1341712474822998, "learning_rate": 2.5834899522914556e-05, "loss": 3.1787, "step": 266500 }, { "epoch": 18.75790361107208, "grad_norm": 1.1443614959716797, "learning_rate": 2.569032817695533e-05, "loss": 3.1772, "step": 267000 }, { "epoch": 18.79303077139244, "grad_norm": 1.0344430208206177, "learning_rate": 2.5545756830996098e-05, "loss": 3.1799, "step": 267500 }, { "epoch": 18.828157931712802, "grad_norm": 1.0747827291488647, "learning_rate": 2.5401185485036865e-05, "loss": 3.1753, "step": 268000 }, { "epoch": 18.86328509203316, "grad_norm": 1.0839353799819946, "learning_rate": 2.5256614139077633e-05, "loss": 3.176, "step": 268500 }, { "epoch": 18.89841225235352, "grad_norm": 1.0615566968917847, "learning_rate": 2.5112042793118407e-05, "loss": 3.1764, "step": 269000 }, { "epoch": 18.93353941267388, "grad_norm": 1.103611946105957, "learning_rate": 2.4967471447159174e-05, "loss": 3.1748, "step": 269500 }, { "epoch": 18.968666572994238, "grad_norm": 1.0903996229171753, "learning_rate": 2.4822900101199945e-05, "loss": 3.1768, "step": 270000 }, { "epoch": 18.968666572994238, "eval_accuracy": 0.6813137842220225, "eval_loss": 2.888620138168335, "eval_runtime": 175.6741, "eval_samples_per_second": 687.102, "eval_steps_per_second": 4.298, "step": 270000 }, { "epoch": 19.0037937333146, "grad_norm": 1.0990253686904907, "learning_rate": 2.4678328755240712e-05, "loss": 3.1749, "step": 270500 }, { "epoch": 19.03892089363496, "grad_norm": 1.0721038579940796, "learning_rate": 2.4533757409281483e-05, "loss": 3.1733, "step": 271000 }, { "epoch": 19.07404805395532, "grad_norm": 1.1202887296676636, "learning_rate": 2.438918606332225e-05, "loss": 3.1738, "step": 271500 }, { "epoch": 19.109175214275677, "grad_norm": 1.116477131843567, "learning_rate": 2.424461471736302e-05, "loss": 3.1725, "step": 272000 }, { "epoch": 19.144302374596037, "grad_norm": 1.1108098030090332, "learning_rate": 2.410004337140379e-05, "loss": 3.1738, "step": 272500 }, { "epoch": 19.179429534916398, "grad_norm": 1.0893597602844238, "learning_rate": 2.395547202544456e-05, "loss": 3.173, "step": 273000 }, { "epoch": 19.21455669523676, "grad_norm": 1.156237244606018, "learning_rate": 2.3810900679485327e-05, "loss": 3.1699, "step": 273500 }, { "epoch": 19.249683855557116, "grad_norm": 1.0929256677627563, "learning_rate": 2.3666329333526098e-05, "loss": 3.1712, "step": 274000 }, { "epoch": 19.284811015877477, "grad_norm": 1.0944771766662598, "learning_rate": 2.3521757987566865e-05, "loss": 3.1714, "step": 274500 }, { "epoch": 19.319938176197837, "grad_norm": 1.0623080730438232, "learning_rate": 2.3377186641607636e-05, "loss": 3.1722, "step": 275000 }, { "epoch": 19.355065336518194, "grad_norm": 1.1324524879455566, "learning_rate": 2.3232615295648403e-05, "loss": 3.1749, "step": 275500 }, { "epoch": 19.390192496838555, "grad_norm": 1.0073175430297852, "learning_rate": 2.3088043949689174e-05, "loss": 3.1687, "step": 276000 }, { "epoch": 19.425319657158916, "grad_norm": 1.0207949876785278, "learning_rate": 2.2943472603729945e-05, "loss": 3.1738, "step": 276500 }, { "epoch": 19.460446817479276, "grad_norm": 1.1397725343704224, "learning_rate": 2.2798901257770712e-05, "loss": 3.1697, "step": 277000 }, { "epoch": 19.495573977799634, "grad_norm": 1.0806381702423096, "learning_rate": 2.265432991181148e-05, "loss": 3.173, "step": 277500 }, { "epoch": 19.530701138119994, "grad_norm": 1.1235697269439697, "learning_rate": 2.250975856585225e-05, "loss": 3.169, "step": 278000 }, { "epoch": 19.565828298440355, "grad_norm": 1.112440824508667, "learning_rate": 2.2365187219893018e-05, "loss": 3.1715, "step": 278500 }, { "epoch": 19.600955458760716, "grad_norm": 1.094836711883545, "learning_rate": 2.2220615873933785e-05, "loss": 3.1681, "step": 279000 }, { "epoch": 19.636082619081073, "grad_norm": 1.0511667728424072, "learning_rate": 2.2076044527974556e-05, "loss": 3.1718, "step": 279500 }, { "epoch": 19.671209779401433, "grad_norm": 1.0339254140853882, "learning_rate": 2.1931473182015323e-05, "loss": 3.1696, "step": 280000 }, { "epoch": 19.671209779401433, "eval_accuracy": 0.6821722159572469, "eval_loss": 2.8827872276306152, "eval_runtime": 177.2578, "eval_samples_per_second": 680.963, "eval_steps_per_second": 4.259, "step": 280000 }, { "epoch": 19.706336939721794, "grad_norm": 0.9928497076034546, "learning_rate": 2.1786901836056094e-05, "loss": 3.1711, "step": 280500 }, { "epoch": 19.74146410004215, "grad_norm": 1.0955822467803955, "learning_rate": 2.164233049009686e-05, "loss": 3.171, "step": 281000 }, { "epoch": 19.776591260362512, "grad_norm": 1.0685282945632935, "learning_rate": 2.1497759144137632e-05, "loss": 3.1709, "step": 281500 }, { "epoch": 19.811718420682872, "grad_norm": 1.0688941478729248, "learning_rate": 2.1353187798178403e-05, "loss": 3.1686, "step": 282000 }, { "epoch": 19.846845581003233, "grad_norm": 1.0664271116256714, "learning_rate": 2.120861645221917e-05, "loss": 3.1682, "step": 282500 }, { "epoch": 19.88197274132359, "grad_norm": 1.1146113872528076, "learning_rate": 2.106404510625994e-05, "loss": 3.1682, "step": 283000 }, { "epoch": 19.91709990164395, "grad_norm": 1.1959316730499268, "learning_rate": 2.091947376030071e-05, "loss": 3.171, "step": 283500 }, { "epoch": 19.95222706196431, "grad_norm": 1.0600392818450928, "learning_rate": 2.077490241434148e-05, "loss": 3.1665, "step": 284000 }, { "epoch": 19.987354222284672, "grad_norm": 1.0286040306091309, "learning_rate": 2.0630331068382247e-05, "loss": 3.1675, "step": 284500 }, { "epoch": 20.02248138260503, "grad_norm": 1.0849840641021729, "learning_rate": 2.0485759722423017e-05, "loss": 3.1708, "step": 285000 }, { "epoch": 20.05760854292539, "grad_norm": 1.0760403871536255, "learning_rate": 2.0341188376463785e-05, "loss": 3.1662, "step": 285500 }, { "epoch": 20.09273570324575, "grad_norm": 1.1143406629562378, "learning_rate": 2.0196617030504556e-05, "loss": 3.1655, "step": 286000 }, { "epoch": 20.127862863566108, "grad_norm": 1.0620818138122559, "learning_rate": 2.0052045684545323e-05, "loss": 3.1655, "step": 286500 }, { "epoch": 20.16299002388647, "grad_norm": 1.1125637292861938, "learning_rate": 1.9907474338586094e-05, "loss": 3.1653, "step": 287000 }, { "epoch": 20.19811718420683, "grad_norm": 1.031412959098816, "learning_rate": 1.976290299262686e-05, "loss": 3.1647, "step": 287500 }, { "epoch": 20.23324434452719, "grad_norm": 1.078932523727417, "learning_rate": 1.9618331646667632e-05, "loss": 3.1651, "step": 288000 }, { "epoch": 20.268371504847547, "grad_norm": 1.0613499879837036, "learning_rate": 1.9473760300708403e-05, "loss": 3.162, "step": 288500 }, { "epoch": 20.303498665167908, "grad_norm": 1.098341703414917, "learning_rate": 1.932918895474917e-05, "loss": 3.1667, "step": 289000 }, { "epoch": 20.33862582548827, "grad_norm": 1.0405348539352417, "learning_rate": 1.918461760878994e-05, "loss": 3.1657, "step": 289500 }, { "epoch": 20.37375298580863, "grad_norm": 1.0748839378356934, "learning_rate": 1.9040046262830708e-05, "loss": 3.1656, "step": 290000 }, { "epoch": 20.37375298580863, "eval_accuracy": 0.6826159586267788, "eval_loss": 2.8787484169006348, "eval_runtime": 174.499, "eval_samples_per_second": 691.729, "eval_steps_per_second": 4.327, "step": 290000 }, { "epoch": 20.408880146128986, "grad_norm": 1.0512151718139648, "learning_rate": 1.889547491687148e-05, "loss": 3.164, "step": 290500 }, { "epoch": 20.444007306449347, "grad_norm": 1.125925898551941, "learning_rate": 1.8750903570912246e-05, "loss": 3.1664, "step": 291000 }, { "epoch": 20.479134466769708, "grad_norm": 1.0517058372497559, "learning_rate": 1.8606332224953017e-05, "loss": 3.1621, "step": 291500 }, { "epoch": 20.514261627090065, "grad_norm": 1.0715209245681763, "learning_rate": 1.8461760878993785e-05, "loss": 3.165, "step": 292000 }, { "epoch": 20.549388787410425, "grad_norm": 1.0268090963363647, "learning_rate": 1.8317189533034555e-05, "loss": 3.1618, "step": 292500 }, { "epoch": 20.584515947730786, "grad_norm": 1.0581594705581665, "learning_rate": 1.8172618187075323e-05, "loss": 3.165, "step": 293000 }, { "epoch": 20.619643108051147, "grad_norm": 1.0758394002914429, "learning_rate": 1.8028046841116093e-05, "loss": 3.1649, "step": 293500 }, { "epoch": 20.654770268371504, "grad_norm": 1.0578275918960571, "learning_rate": 1.788347549515686e-05, "loss": 3.1621, "step": 294000 }, { "epoch": 20.689897428691864, "grad_norm": 1.063185691833496, "learning_rate": 1.7738904149197628e-05, "loss": 3.1629, "step": 294500 }, { "epoch": 20.725024589012225, "grad_norm": 1.0353269577026367, "learning_rate": 1.75943328032384e-05, "loss": 3.162, "step": 295000 }, { "epoch": 20.760151749332586, "grad_norm": 1.0861767530441284, "learning_rate": 1.7449761457279166e-05, "loss": 3.165, "step": 295500 }, { "epoch": 20.795278909652943, "grad_norm": 1.0341486930847168, "learning_rate": 1.7305190111319937e-05, "loss": 3.1597, "step": 296000 }, { "epoch": 20.830406069973304, "grad_norm": 1.0556368827819824, "learning_rate": 1.7160618765360705e-05, "loss": 3.1625, "step": 296500 }, { "epoch": 20.865533230293664, "grad_norm": 1.0311942100524902, "learning_rate": 1.7016047419401475e-05, "loss": 3.1591, "step": 297000 }, { "epoch": 20.90066039061402, "grad_norm": 1.0713430643081665, "learning_rate": 1.6871476073442243e-05, "loss": 3.164, "step": 297500 }, { "epoch": 20.935787550934382, "grad_norm": 1.0596240758895874, "learning_rate": 1.6726904727483014e-05, "loss": 3.1618, "step": 298000 }, { "epoch": 20.970914711254743, "grad_norm": 1.0879930257797241, "learning_rate": 1.658233338152378e-05, "loss": 3.1635, "step": 298500 }, { "epoch": 21.006041871575103, "grad_norm": 1.0523048639297485, "learning_rate": 1.6437762035564552e-05, "loss": 3.1625, "step": 299000 }, { "epoch": 21.04116903189546, "grad_norm": 1.0494343042373657, "learning_rate": 1.629319068960532e-05, "loss": 3.162, "step": 299500 }, { "epoch": 21.07629619221582, "grad_norm": 1.1549487113952637, "learning_rate": 1.614861934364609e-05, "loss": 3.1581, "step": 300000 }, { "epoch": 21.07629619221582, "eval_accuracy": 0.6834358038843739, "eval_loss": 2.8755788803100586, "eval_runtime": 177.8527, "eval_samples_per_second": 678.685, "eval_steps_per_second": 4.245, "step": 300000 }, { "epoch": 21.111423352536182, "grad_norm": 1.068939447402954, "learning_rate": 1.6004047997686857e-05, "loss": 3.1589, "step": 300500 }, { "epoch": 21.14655051285654, "grad_norm": 1.052692174911499, "learning_rate": 1.5859476651727628e-05, "loss": 3.1615, "step": 301000 }, { "epoch": 21.1816776731769, "grad_norm": 1.0974150896072388, "learning_rate": 1.57149053057684e-05, "loss": 3.16, "step": 301500 }, { "epoch": 21.21680483349726, "grad_norm": 1.0353583097457886, "learning_rate": 1.5570333959809166e-05, "loss": 3.1629, "step": 302000 }, { "epoch": 21.25193199381762, "grad_norm": 1.0160739421844482, "learning_rate": 1.5425762613849937e-05, "loss": 3.1579, "step": 302500 }, { "epoch": 21.287059154137978, "grad_norm": 1.123225450515747, "learning_rate": 1.5281191267890704e-05, "loss": 3.1593, "step": 303000 }, { "epoch": 21.32218631445834, "grad_norm": 1.077954888343811, "learning_rate": 1.5136619921931475e-05, "loss": 3.159, "step": 303500 }, { "epoch": 21.3573134747787, "grad_norm": 1.0142537355422974, "learning_rate": 1.4992048575972243e-05, "loss": 3.1554, "step": 304000 }, { "epoch": 21.39244063509906, "grad_norm": 1.1051485538482666, "learning_rate": 1.4847477230013013e-05, "loss": 3.1597, "step": 304500 }, { "epoch": 21.427567795419417, "grad_norm": 1.1120260953903198, "learning_rate": 1.470290588405378e-05, "loss": 3.1577, "step": 305000 }, { "epoch": 21.462694955739778, "grad_norm": 1.113200306892395, "learning_rate": 1.4558334538094551e-05, "loss": 3.1578, "step": 305500 }, { "epoch": 21.49782211606014, "grad_norm": 1.092349648475647, "learning_rate": 1.4413763192135319e-05, "loss": 3.1586, "step": 306000 }, { "epoch": 21.532949276380496, "grad_norm": 1.1539112329483032, "learning_rate": 1.4269191846176088e-05, "loss": 3.1579, "step": 306500 }, { "epoch": 21.568076436700856, "grad_norm": 1.078343152999878, "learning_rate": 1.4124620500216857e-05, "loss": 3.1566, "step": 307000 }, { "epoch": 21.603203597021217, "grad_norm": 1.062975287437439, "learning_rate": 1.3980049154257626e-05, "loss": 3.1576, "step": 307500 }, { "epoch": 21.638330757341578, "grad_norm": 1.0585559606552124, "learning_rate": 1.3835477808298397e-05, "loss": 3.1634, "step": 308000 }, { "epoch": 21.673457917661935, "grad_norm": 1.0667996406555176, "learning_rate": 1.3690906462339164e-05, "loss": 3.1593, "step": 308500 }, { "epoch": 21.708585077982296, "grad_norm": 1.082306981086731, "learning_rate": 1.3546335116379935e-05, "loss": 3.1573, "step": 309000 }, { "epoch": 21.743712238302656, "grad_norm": 1.071520447731018, "learning_rate": 1.3401763770420702e-05, "loss": 3.1559, "step": 309500 }, { "epoch": 21.778839398623017, "grad_norm": 1.0488178730010986, "learning_rate": 1.3257192424461473e-05, "loss": 3.1566, "step": 310000 }, { "epoch": 21.778839398623017, "eval_accuracy": 0.6841967701215512, "eval_loss": 2.8690459728240967, "eval_runtime": 174.5272, "eval_samples_per_second": 691.617, "eval_steps_per_second": 4.326, "step": 310000 }, { "epoch": 21.813966558943374, "grad_norm": 1.0526132583618164, "learning_rate": 1.311262107850224e-05, "loss": 3.1539, "step": 310500 }, { "epoch": 21.849093719263735, "grad_norm": 1.068835973739624, "learning_rate": 1.2968049732543011e-05, "loss": 3.1534, "step": 311000 }, { "epoch": 21.884220879584095, "grad_norm": 1.0928212404251099, "learning_rate": 1.2823478386583779e-05, "loss": 3.1585, "step": 311500 }, { "epoch": 21.919348039904452, "grad_norm": 1.119399070739746, "learning_rate": 1.267890704062455e-05, "loss": 3.1538, "step": 312000 }, { "epoch": 21.954475200224813, "grad_norm": 1.100266456604004, "learning_rate": 1.2534335694665317e-05, "loss": 3.1565, "step": 312500 }, { "epoch": 21.989602360545174, "grad_norm": 1.062166452407837, "learning_rate": 1.2389764348706088e-05, "loss": 3.1576, "step": 313000 }, { "epoch": 22.024729520865534, "grad_norm": 1.1449230909347534, "learning_rate": 1.2245193002746857e-05, "loss": 3.1544, "step": 313500 }, { "epoch": 22.05985668118589, "grad_norm": 1.0918468236923218, "learning_rate": 1.2100621656787626e-05, "loss": 3.1511, "step": 314000 }, { "epoch": 22.094983841506252, "grad_norm": 1.1008789539337158, "learning_rate": 1.1956050310828395e-05, "loss": 3.1571, "step": 314500 }, { "epoch": 22.130111001826613, "grad_norm": 1.0781457424163818, "learning_rate": 1.1811478964869162e-05, "loss": 3.154, "step": 315000 }, { "epoch": 22.165238162146974, "grad_norm": 1.0556070804595947, "learning_rate": 1.1666907618909931e-05, "loss": 3.1502, "step": 315500 }, { "epoch": 22.20036532246733, "grad_norm": 1.069411039352417, "learning_rate": 1.15223362729507e-05, "loss": 3.1515, "step": 316000 }, { "epoch": 22.23549248278769, "grad_norm": 1.1082756519317627, "learning_rate": 1.137776492699147e-05, "loss": 3.1541, "step": 316500 }, { "epoch": 22.270619643108052, "grad_norm": 1.0453790426254272, "learning_rate": 1.123319358103224e-05, "loss": 3.1546, "step": 317000 }, { "epoch": 22.30574680342841, "grad_norm": 1.124457597732544, "learning_rate": 1.108862223507301e-05, "loss": 3.1539, "step": 317500 }, { "epoch": 22.34087396374877, "grad_norm": 1.0583232641220093, "learning_rate": 1.0944050889113779e-05, "loss": 3.1527, "step": 318000 }, { "epoch": 22.37600112406913, "grad_norm": 1.0516482591629028, "learning_rate": 1.0799479543154548e-05, "loss": 3.1551, "step": 318500 }, { "epoch": 22.41112828438949, "grad_norm": 1.0380626916885376, "learning_rate": 1.0654908197195317e-05, "loss": 3.1522, "step": 319000 }, { "epoch": 22.44625544470985, "grad_norm": 1.0822714567184448, "learning_rate": 1.0510336851236086e-05, "loss": 3.1508, "step": 319500 }, { "epoch": 22.48138260503021, "grad_norm": 1.1107425689697266, "learning_rate": 1.0365765505276855e-05, "loss": 3.1508, "step": 320000 }, { "epoch": 22.48138260503021, "eval_accuracy": 0.6844732547641824, "eval_loss": 2.8670718669891357, "eval_runtime": 181.4613, "eval_samples_per_second": 665.189, "eval_steps_per_second": 4.161, "step": 320000 }, { "epoch": 22.51650976535057, "grad_norm": 1.115052342414856, "learning_rate": 1.0221194159317624e-05, "loss": 3.1529, "step": 320500 }, { "epoch": 22.55163692567093, "grad_norm": 1.1568660736083984, "learning_rate": 1.0076622813358393e-05, "loss": 3.1558, "step": 321000 }, { "epoch": 22.586764085991287, "grad_norm": 1.0568718910217285, "learning_rate": 9.932051467399162e-06, "loss": 3.1548, "step": 321500 }, { "epoch": 22.621891246311648, "grad_norm": 1.1229430437088013, "learning_rate": 9.787480121439931e-06, "loss": 3.15, "step": 322000 }, { "epoch": 22.65701840663201, "grad_norm": 1.0613257884979248, "learning_rate": 9.6429087754807e-06, "loss": 3.1536, "step": 322500 }, { "epoch": 22.692145566952366, "grad_norm": 1.0614995956420898, "learning_rate": 9.49833742952147e-06, "loss": 3.152, "step": 323000 }, { "epoch": 22.727272727272727, "grad_norm": 1.0684412717819214, "learning_rate": 9.353766083562238e-06, "loss": 3.151, "step": 323500 }, { "epoch": 22.762399887593087, "grad_norm": 1.0731546878814697, "learning_rate": 9.209194737603008e-06, "loss": 3.149, "step": 324000 }, { "epoch": 22.797527047913448, "grad_norm": 1.1367193460464478, "learning_rate": 9.064623391643777e-06, "loss": 3.1533, "step": 324500 }, { "epoch": 22.832654208233805, "grad_norm": 1.05376136302948, "learning_rate": 8.920052045684546e-06, "loss": 3.1525, "step": 325000 }, { "epoch": 22.867781368554166, "grad_norm": 1.0761877298355103, "learning_rate": 8.775480699725315e-06, "loss": 3.1519, "step": 325500 }, { "epoch": 22.902908528874526, "grad_norm": 1.0843102931976318, "learning_rate": 8.630909353766084e-06, "loss": 3.1538, "step": 326000 }, { "epoch": 22.938035689194887, "grad_norm": 1.0343738794326782, "learning_rate": 8.486338007806853e-06, "loss": 3.1508, "step": 326500 }, { "epoch": 22.973162849515244, "grad_norm": 1.079969048500061, "learning_rate": 8.341766661847622e-06, "loss": 3.147, "step": 327000 }, { "epoch": 23.008290009835605, "grad_norm": 1.0848047733306885, "learning_rate": 8.197195315888391e-06, "loss": 3.1509, "step": 327500 }, { "epoch": 23.043417170155966, "grad_norm": 1.0909361839294434, "learning_rate": 8.05262396992916e-06, "loss": 3.1513, "step": 328000 }, { "epoch": 23.078544330476323, "grad_norm": 1.0889477729797363, "learning_rate": 7.90805262396993e-06, "loss": 3.146, "step": 328500 }, { "epoch": 23.113671490796683, "grad_norm": 1.0752896070480347, "learning_rate": 7.763481278010698e-06, "loss": 3.1496, "step": 329000 }, { "epoch": 23.148798651117044, "grad_norm": 1.1438133716583252, "learning_rate": 7.618909932051468e-06, "loss": 3.1513, "step": 329500 }, { "epoch": 23.183925811437405, "grad_norm": 1.04530668258667, "learning_rate": 7.474338586092237e-06, "loss": 3.1496, "step": 330000 }, { "epoch": 23.183925811437405, "eval_accuracy": 0.6848984702814148, "eval_loss": 2.864835262298584, "eval_runtime": 177.8517, "eval_samples_per_second": 678.689, "eval_steps_per_second": 4.245, "step": 330000 }, { "epoch": 23.219052971757762, "grad_norm": 1.1062647104263306, "learning_rate": 7.3297672401330065e-06, "loss": 3.1444, "step": 330500 }, { "epoch": 23.254180132078123, "grad_norm": 1.1085108518600464, "learning_rate": 7.1851958941737756e-06, "loss": 3.1489, "step": 331000 }, { "epoch": 23.289307292398483, "grad_norm": 1.0674808025360107, "learning_rate": 7.040624548214545e-06, "loss": 3.1484, "step": 331500 }, { "epoch": 23.324434452718844, "grad_norm": 1.046091914176941, "learning_rate": 6.896053202255314e-06, "loss": 3.152, "step": 332000 }, { "epoch": 23.3595616130392, "grad_norm": 1.0943750143051147, "learning_rate": 6.751481856296083e-06, "loss": 3.1494, "step": 332500 }, { "epoch": 23.39468877335956, "grad_norm": 1.0809797048568726, "learning_rate": 6.606910510336851e-06, "loss": 3.1459, "step": 333000 }, { "epoch": 23.429815933679922, "grad_norm": 1.0576915740966797, "learning_rate": 6.46233916437762e-06, "loss": 3.1483, "step": 333500 }, { "epoch": 23.46494309400028, "grad_norm": 1.0860077142715454, "learning_rate": 6.317767818418389e-06, "loss": 3.1484, "step": 334000 }, { "epoch": 23.50007025432064, "grad_norm": 1.1696947813034058, "learning_rate": 6.173196472459159e-06, "loss": 3.1503, "step": 334500 }, { "epoch": 23.535197414641, "grad_norm": 1.0658084154129028, "learning_rate": 6.028625126499928e-06, "loss": 3.1505, "step": 335000 }, { "epoch": 23.57032457496136, "grad_norm": 1.0921698808670044, "learning_rate": 5.884053780540697e-06, "loss": 3.1481, "step": 335500 }, { "epoch": 23.60545173528172, "grad_norm": 1.0856083631515503, "learning_rate": 5.739482434581466e-06, "loss": 3.1457, "step": 336000 }, { "epoch": 23.64057889560208, "grad_norm": 1.1023163795471191, "learning_rate": 5.5949110886222355e-06, "loss": 3.1434, "step": 336500 }, { "epoch": 23.67570605592244, "grad_norm": 1.210658073425293, "learning_rate": 5.4503397426630046e-06, "loss": 3.1482, "step": 337000 }, { "epoch": 23.7108332162428, "grad_norm": 1.0992660522460938, "learning_rate": 5.305768396703774e-06, "loss": 3.1428, "step": 337500 }, { "epoch": 23.745960376563158, "grad_norm": 1.0835355520248413, "learning_rate": 5.161197050744543e-06, "loss": 3.1449, "step": 338000 }, { "epoch": 23.78108753688352, "grad_norm": 1.0921368598937988, "learning_rate": 5.016625704785312e-06, "loss": 3.1453, "step": 338500 }, { "epoch": 23.81621469720388, "grad_norm": 1.0912210941314697, "learning_rate": 4.872054358826081e-06, "loss": 3.1468, "step": 339000 }, { "epoch": 23.851341857524236, "grad_norm": 1.0831457376480103, "learning_rate": 4.72748301286685e-06, "loss": 3.1471, "step": 339500 }, { "epoch": 23.886469017844597, "grad_norm": 1.0877655744552612, "learning_rate": 4.582911666907619e-06, "loss": 3.1475, "step": 340000 }, { "epoch": 23.886469017844597, "eval_accuracy": 0.685332350059987, "eval_loss": 2.8612310886383057, "eval_runtime": 177.4884, "eval_samples_per_second": 680.078, "eval_steps_per_second": 4.254, "step": 340000 }, { "epoch": 23.921596178164958, "grad_norm": 1.0923268795013428, "learning_rate": 4.438340320948388e-06, "loss": 3.1454, "step": 340500 }, { "epoch": 23.956723338485318, "grad_norm": 1.0587596893310547, "learning_rate": 4.293768974989157e-06, "loss": 3.1457, "step": 341000 }, { "epoch": 23.991850498805675, "grad_norm": 1.083365559577942, "learning_rate": 4.149197629029926e-06, "loss": 3.1468, "step": 341500 }, { "epoch": 24.026977659126036, "grad_norm": 1.0678715705871582, "learning_rate": 4.004626283070695e-06, "loss": 3.1477, "step": 342000 }, { "epoch": 24.062104819446397, "grad_norm": 1.1342980861663818, "learning_rate": 3.8600549371114645e-06, "loss": 3.145, "step": 342500 }, { "epoch": 24.097231979766757, "grad_norm": 1.0564827919006348, "learning_rate": 3.7154835911522336e-06, "loss": 3.1447, "step": 343000 }, { "epoch": 24.132359140087114, "grad_norm": 1.0578299760818481, "learning_rate": 3.5709122451930026e-06, "loss": 3.1468, "step": 343500 }, { "epoch": 24.167486300407475, "grad_norm": 1.077895998954773, "learning_rate": 3.426340899233772e-06, "loss": 3.1482, "step": 344000 }, { "epoch": 24.202613460727836, "grad_norm": 1.0786484479904175, "learning_rate": 3.2817695532745412e-06, "loss": 3.1467, "step": 344500 }, { "epoch": 24.237740621048193, "grad_norm": 1.079455852508545, "learning_rate": 3.1371982073153103e-06, "loss": 3.1487, "step": 345000 }, { "epoch": 24.272867781368554, "grad_norm": 1.084189534187317, "learning_rate": 2.9926268613560794e-06, "loss": 3.1449, "step": 345500 }, { "epoch": 24.307994941688914, "grad_norm": 1.1079171895980835, "learning_rate": 2.8480555153968485e-06, "loss": 3.1438, "step": 346000 }, { "epoch": 24.343122102009275, "grad_norm": 1.1567984819412231, "learning_rate": 2.7034841694376176e-06, "loss": 3.1428, "step": 346500 }, { "epoch": 24.378249262329632, "grad_norm": 1.1012502908706665, "learning_rate": 2.5589128234783866e-06, "loss": 3.1453, "step": 347000 }, { "epoch": 24.413376422649993, "grad_norm": 1.1289618015289307, "learning_rate": 2.414341477519156e-06, "loss": 3.1447, "step": 347500 }, { "epoch": 24.448503582970353, "grad_norm": 1.0563510656356812, "learning_rate": 2.269770131559925e-06, "loss": 3.1456, "step": 348000 }, { "epoch": 24.483630743290714, "grad_norm": 1.0858873128890991, "learning_rate": 2.125198785600694e-06, "loss": 3.1423, "step": 348500 }, { "epoch": 24.51875790361107, "grad_norm": 1.0456035137176514, "learning_rate": 1.980627439641463e-06, "loss": 3.1415, "step": 349000 }, { "epoch": 24.553885063931432, "grad_norm": 1.0833024978637695, "learning_rate": 1.8360560936822323e-06, "loss": 3.1426, "step": 349500 }, { "epoch": 24.589012224251793, "grad_norm": 1.0850093364715576, "learning_rate": 1.6914847477230013e-06, "loss": 3.1459, "step": 350000 }, { "epoch": 24.589012224251793, "eval_accuracy": 0.6859052336269196, "eval_loss": 2.8585827350616455, "eval_runtime": 175.3319, "eval_samples_per_second": 688.443, "eval_steps_per_second": 4.306, "step": 350000 }, { "epoch": 24.62413938457215, "grad_norm": 1.056212306022644, "learning_rate": 1.5469134017637704e-06, "loss": 3.1441, "step": 350500 }, { "epoch": 24.65926654489251, "grad_norm": 1.0897036790847778, "learning_rate": 1.4023420558045395e-06, "loss": 3.1426, "step": 351000 }, { "epoch": 24.69439370521287, "grad_norm": 1.096097469329834, "learning_rate": 1.2577707098453088e-06, "loss": 3.1428, "step": 351500 }, { "epoch": 24.72952086553323, "grad_norm": 1.1135632991790771, "learning_rate": 1.1131993638860779e-06, "loss": 3.1441, "step": 352000 }, { "epoch": 24.76464802585359, "grad_norm": 1.1100594997406006, "learning_rate": 9.68628017926847e-07, "loss": 3.1434, "step": 352500 }, { "epoch": 24.79977518617395, "grad_norm": 1.1389775276184082, "learning_rate": 8.240566719676161e-07, "loss": 3.1428, "step": 353000 }, { "epoch": 24.83490234649431, "grad_norm": 1.0750815868377686, "learning_rate": 6.794853260083851e-07, "loss": 3.1424, "step": 353500 }, { "epoch": 24.870029506814667, "grad_norm": 1.1358802318572998, "learning_rate": 5.349139800491543e-07, "loss": 3.1449, "step": 354000 }, { "epoch": 24.905156667135028, "grad_norm": 1.1005890369415283, "learning_rate": 3.903426340899234e-07, "loss": 3.1444, "step": 354500 }, { "epoch": 24.94028382745539, "grad_norm": 1.085300087928772, "learning_rate": 2.457712881306925e-07, "loss": 3.14, "step": 355000 }, { "epoch": 24.97541098777575, "grad_norm": 1.0869441032409668, "learning_rate": 1.0119994217146162e-07, "loss": 3.1434, "step": 355500 }, { "epoch": 25.0, "step": 355850, "total_flos": 3.037727852389325e+18, "train_loss": 3.5510173419085684, "train_runtime": 125655.4376, "train_samples_per_second": 453.093, "train_steps_per_second": 2.832 } ], "logging_steps": 500, "max_steps": 355850, "num_input_tokens_seen": 0, "num_train_epochs": 25, "save_steps": 10000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 3.037727852389325e+18, "train_batch_size": 160, "trial_name": null, "trial_params": null }