diff --git "a/pretrained_adapter/trainer_state.json" "b/pretrained_adapter/trainer_state.json" new file mode 100644--- /dev/null +++ "b/pretrained_adapter/trainer_state.json" @@ -0,0 +1,30569 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 1.0, + "eval_steps": 500, + "global_step": 4361, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.00022930520522815867, + "grad_norm": 43.657429986391875, + "learning_rate": 7.633587786259541e-06, + "loss": 6.2095, + "step": 1 + }, + { + "epoch": 0.00045861041045631735, + "grad_norm": 49.45528156333248, + "learning_rate": 1.5267175572519083e-05, + "loss": 6.1589, + "step": 2 + }, + { + "epoch": 0.000687915615684476, + "grad_norm": 41.59624725004239, + "learning_rate": 2.2900763358778628e-05, + "loss": 6.4097, + "step": 3 + }, + { + "epoch": 0.0009172208209126347, + "grad_norm": 25.873539186042887, + "learning_rate": 3.0534351145038166e-05, + "loss": 5.7889, + "step": 4 + }, + { + "epoch": 0.0011465260261407934, + "grad_norm": 92.55934581005494, + "learning_rate": 3.816793893129771e-05, + "loss": 6.0161, + "step": 5 + }, + { + "epoch": 0.001375831231368952, + "grad_norm": 8.283168225665385, + "learning_rate": 4.5801526717557256e-05, + "loss": 5.3858, + "step": 6 + }, + { + "epoch": 0.0016051364365971107, + "grad_norm": 9.992769926528503, + "learning_rate": 5.3435114503816794e-05, + "loss": 5.4802, + "step": 7 + }, + { + "epoch": 0.0018344416418252694, + "grad_norm": 15.570698649538418, + "learning_rate": 6.106870229007633e-05, + "loss": 5.2895, + "step": 8 + }, + { + "epoch": 0.0020637468470534283, + "grad_norm": 11.026631209996928, + "learning_rate": 6.870229007633588e-05, + "loss": 5.188, + "step": 9 + }, + { + "epoch": 0.002293052052281587, + "grad_norm": 7.335926577246894, + "learning_rate": 7.633587786259542e-05, + "loss": 5.0984, + "step": 10 + }, + { + "epoch": 0.0025223572575097455, + "grad_norm": 16.915624231108723, + "learning_rate": 8.396946564885496e-05, + "loss": 5.1816, + "step": 11 + }, + { + "epoch": 0.002751662462737904, + "grad_norm": 87.58159343124493, + "learning_rate": 9.160305343511451e-05, + "loss": 5.0624, + "step": 12 + }, + { + "epoch": 0.002980967667966063, + "grad_norm": 10.415277737117835, + "learning_rate": 9.923664122137405e-05, + "loss": 4.935, + "step": 13 + }, + { + "epoch": 0.0032102728731942215, + "grad_norm": 10.287419269187977, + "learning_rate": 0.00010687022900763359, + "loss": 4.9442, + "step": 14 + }, + { + "epoch": 0.00343957807842238, + "grad_norm": 4.329805640101233, + "learning_rate": 0.00011450381679389313, + "loss": 4.6681, + "step": 15 + }, + { + "epoch": 0.0036688832836505388, + "grad_norm": 9.459906795722558, + "learning_rate": 0.00012213740458015266, + "loss": 4.861, + "step": 16 + }, + { + "epoch": 0.0038981884888786974, + "grad_norm": 74.17495468492244, + "learning_rate": 0.00012977099236641222, + "loss": 4.8099, + "step": 17 + }, + { + "epoch": 0.0041274936941068565, + "grad_norm": 5.4007233460759245, + "learning_rate": 0.00013740458015267177, + "loss": 4.861, + "step": 18 + }, + { + "epoch": 0.004356798899335015, + "grad_norm": 11.215337137567843, + "learning_rate": 0.0001450381679389313, + "loss": 4.6569, + "step": 19 + }, + { + "epoch": 0.004586104104563174, + "grad_norm": 15.070438505562352, + "learning_rate": 0.00015267175572519084, + "loss": 4.7286, + "step": 20 + }, + { + "epoch": 0.004815409309791332, + "grad_norm": 4.8868753278874175, + "learning_rate": 0.00016030534351145037, + "loss": 4.6244, + "step": 21 + }, + { + "epoch": 0.005044714515019491, + "grad_norm": 3.395333775871932, + "learning_rate": 0.00016793893129770992, + "loss": 4.5792, + "step": 22 + }, + { + "epoch": 0.005274019720247649, + "grad_norm": 5.405545372118402, + "learning_rate": 0.00017557251908396944, + "loss": 4.5674, + "step": 23 + }, + { + "epoch": 0.005503324925475808, + "grad_norm": 3.0804034971931307, + "learning_rate": 0.00018320610687022902, + "loss": 4.537, + "step": 24 + }, + { + "epoch": 0.005732630130703967, + "grad_norm": 2.4130774312305387, + "learning_rate": 0.00019083969465648857, + "loss": 4.597, + "step": 25 + }, + { + "epoch": 0.005961935335932126, + "grad_norm": 2.7195343832354597, + "learning_rate": 0.0001984732824427481, + "loss": 4.5617, + "step": 26 + }, + { + "epoch": 0.006191240541160285, + "grad_norm": 1.9722898992449724, + "learning_rate": 0.00020610687022900765, + "loss": 4.5245, + "step": 27 + }, + { + "epoch": 0.006420545746388443, + "grad_norm": 1.811546394048573, + "learning_rate": 0.00021374045801526718, + "loss": 4.4559, + "step": 28 + }, + { + "epoch": 0.006649850951616602, + "grad_norm": 4.26397897033763, + "learning_rate": 0.00022137404580152673, + "loss": 4.4735, + "step": 29 + }, + { + "epoch": 0.00687915615684476, + "grad_norm": 1.8778334921648363, + "learning_rate": 0.00022900763358778625, + "loss": 4.4258, + "step": 30 + }, + { + "epoch": 0.007108461362072919, + "grad_norm": 4.247766829745934, + "learning_rate": 0.0002366412213740458, + "loss": 4.3768, + "step": 31 + }, + { + "epoch": 0.0073377665673010775, + "grad_norm": 1.7872755674447072, + "learning_rate": 0.00024427480916030533, + "loss": 4.1895, + "step": 32 + }, + { + "epoch": 0.007567071772529237, + "grad_norm": 1.6887101624634633, + "learning_rate": 0.0002519083969465649, + "loss": 4.4767, + "step": 33 + }, + { + "epoch": 0.007796376977757395, + "grad_norm": 1.4449573378774971, + "learning_rate": 0.00025954198473282443, + "loss": 4.4314, + "step": 34 + }, + { + "epoch": 0.008025682182985553, + "grad_norm": 1.5695251559624077, + "learning_rate": 0.000267175572519084, + "loss": 4.314, + "step": 35 + }, + { + "epoch": 0.008254987388213713, + "grad_norm": 1.7070708687249543, + "learning_rate": 0.00027480916030534353, + "loss": 4.4272, + "step": 36 + }, + { + "epoch": 0.008484292593441871, + "grad_norm": 3.045205258221281, + "learning_rate": 0.0002824427480916031, + "loss": 4.2841, + "step": 37 + }, + { + "epoch": 0.00871359779867003, + "grad_norm": 2.976949173909087, + "learning_rate": 0.0002900763358778626, + "loss": 4.2029, + "step": 38 + }, + { + "epoch": 0.008942903003898188, + "grad_norm": 1.6023754618732007, + "learning_rate": 0.00029770992366412214, + "loss": 4.2374, + "step": 39 + }, + { + "epoch": 0.009172208209126348, + "grad_norm": 2.104925568876093, + "learning_rate": 0.0003053435114503817, + "loss": 4.099, + "step": 40 + }, + { + "epoch": 0.009401513414354506, + "grad_norm": 3.119824341496828, + "learning_rate": 0.00031297709923664124, + "loss": 4.1661, + "step": 41 + }, + { + "epoch": 0.009630818619582664, + "grad_norm": 1.5054626119062398, + "learning_rate": 0.00032061068702290074, + "loss": 4.2114, + "step": 42 + }, + { + "epoch": 0.009860123824810824, + "grad_norm": 1.8374755359884836, + "learning_rate": 0.0003282442748091603, + "loss": 4.0753, + "step": 43 + }, + { + "epoch": 0.010089429030038982, + "grad_norm": 1.5063143087650968, + "learning_rate": 0.00033587786259541984, + "loss": 3.9873, + "step": 44 + }, + { + "epoch": 0.01031873423526714, + "grad_norm": 1.840731812431801, + "learning_rate": 0.0003435114503816794, + "loss": 4.0491, + "step": 45 + }, + { + "epoch": 0.010548039440495299, + "grad_norm": 0.976719929071258, + "learning_rate": 0.0003511450381679389, + "loss": 3.9026, + "step": 46 + }, + { + "epoch": 0.010777344645723459, + "grad_norm": 1.3524236654705835, + "learning_rate": 0.0003587786259541985, + "loss": 4.1842, + "step": 47 + }, + { + "epoch": 0.011006649850951617, + "grad_norm": 1.7056387173319887, + "learning_rate": 0.00036641221374045805, + "loss": 4.1666, + "step": 48 + }, + { + "epoch": 0.011235955056179775, + "grad_norm": 1.0347299093114959, + "learning_rate": 0.0003740458015267176, + "loss": 4.0625, + "step": 49 + }, + { + "epoch": 0.011465260261407933, + "grad_norm": 1.1246606418193017, + "learning_rate": 0.00038167938931297715, + "loss": 4.0474, + "step": 50 + }, + { + "epoch": 0.011694565466636093, + "grad_norm": 1.0921510765626554, + "learning_rate": 0.00038931297709923665, + "loss": 3.9034, + "step": 51 + }, + { + "epoch": 0.011923870671864251, + "grad_norm": 1.3754325239699374, + "learning_rate": 0.0003969465648854962, + "loss": 3.9762, + "step": 52 + }, + { + "epoch": 0.01215317587709241, + "grad_norm": 1.0200475655129955, + "learning_rate": 0.00040458015267175575, + "loss": 4.085, + "step": 53 + }, + { + "epoch": 0.01238248108232057, + "grad_norm": 0.9186687434517378, + "learning_rate": 0.0004122137404580153, + "loss": 4.076, + "step": 54 + }, + { + "epoch": 0.012611786287548728, + "grad_norm": 0.8292711276430443, + "learning_rate": 0.0004198473282442748, + "loss": 4.095, + "step": 55 + }, + { + "epoch": 0.012841091492776886, + "grad_norm": 0.8386189966982689, + "learning_rate": 0.00042748091603053435, + "loss": 4.0092, + "step": 56 + }, + { + "epoch": 0.013070396698005044, + "grad_norm": 1.121138848221922, + "learning_rate": 0.0004351145038167939, + "loss": 3.904, + "step": 57 + }, + { + "epoch": 0.013299701903233204, + "grad_norm": 0.9918106436295577, + "learning_rate": 0.00044274809160305345, + "loss": 3.788, + "step": 58 + }, + { + "epoch": 0.013529007108461362, + "grad_norm": 0.7470972768453017, + "learning_rate": 0.00045038167938931295, + "loss": 4.1037, + "step": 59 + }, + { + "epoch": 0.01375831231368952, + "grad_norm": 1.0325915990066685, + "learning_rate": 0.0004580152671755725, + "loss": 4.0703, + "step": 60 + }, + { + "epoch": 0.013987617518917679, + "grad_norm": 1.3436434034049543, + "learning_rate": 0.00046564885496183206, + "loss": 4.0432, + "step": 61 + }, + { + "epoch": 0.014216922724145839, + "grad_norm": 0.9754554443836098, + "learning_rate": 0.0004732824427480916, + "loss": 3.8279, + "step": 62 + }, + { + "epoch": 0.014446227929373997, + "grad_norm": 0.7954646850546606, + "learning_rate": 0.00048091603053435116, + "loss": 3.9552, + "step": 63 + }, + { + "epoch": 0.014675533134602155, + "grad_norm": 0.6829437279967754, + "learning_rate": 0.0004885496183206107, + "loss": 3.9941, + "step": 64 + }, + { + "epoch": 0.014904838339830313, + "grad_norm": 0.7778555767964662, + "learning_rate": 0.0004961832061068703, + "loss": 3.9241, + "step": 65 + }, + { + "epoch": 0.015134143545058473, + "grad_norm": 0.7451842100981538, + "learning_rate": 0.0005038167938931298, + "loss": 3.8935, + "step": 66 + }, + { + "epoch": 0.015363448750286631, + "grad_norm": 0.78979675438245, + "learning_rate": 0.0005114503816793893, + "loss": 3.9091, + "step": 67 + }, + { + "epoch": 0.01559275395551479, + "grad_norm": 1.8909002751449382, + "learning_rate": 0.0005190839694656489, + "loss": 3.9927, + "step": 68 + }, + { + "epoch": 0.01582205916074295, + "grad_norm": 1.0355065516812128, + "learning_rate": 0.0005267175572519084, + "loss": 3.9133, + "step": 69 + }, + { + "epoch": 0.016051364365971106, + "grad_norm": 0.8265066987884603, + "learning_rate": 0.000534351145038168, + "loss": 3.8042, + "step": 70 + }, + { + "epoch": 0.016280669571199266, + "grad_norm": 1.0114397844327805, + "learning_rate": 0.0005419847328244275, + "loss": 3.7794, + "step": 71 + }, + { + "epoch": 0.016509974776427426, + "grad_norm": 1.1209193805030204, + "learning_rate": 0.0005496183206106871, + "loss": 3.7373, + "step": 72 + }, + { + "epoch": 0.016739279981655582, + "grad_norm": 10.28923756843068, + "learning_rate": 0.0005572519083969466, + "loss": 3.9898, + "step": 73 + }, + { + "epoch": 0.016968585186883742, + "grad_norm": 1.1356132215026602, + "learning_rate": 0.0005648854961832062, + "loss": 3.9183, + "step": 74 + }, + { + "epoch": 0.017197890392111902, + "grad_norm": 1.2088788079396195, + "learning_rate": 0.0005725190839694656, + "loss": 3.9502, + "step": 75 + }, + { + "epoch": 0.01742719559734006, + "grad_norm": 0.8909750677383539, + "learning_rate": 0.0005801526717557252, + "loss": 3.8122, + "step": 76 + }, + { + "epoch": 0.01765650080256822, + "grad_norm": 0.9539502365816872, + "learning_rate": 0.0005877862595419848, + "loss": 3.8765, + "step": 77 + }, + { + "epoch": 0.017885806007796375, + "grad_norm": 1.1754757020564015, + "learning_rate": 0.0005954198473282443, + "loss": 3.9915, + "step": 78 + }, + { + "epoch": 0.018115111213024535, + "grad_norm": 1.0463829145524763, + "learning_rate": 0.0006030534351145039, + "loss": 3.8822, + "step": 79 + }, + { + "epoch": 0.018344416418252695, + "grad_norm": 1.0538392879285288, + "learning_rate": 0.0006106870229007634, + "loss": 3.9204, + "step": 80 + }, + { + "epoch": 0.01857372162348085, + "grad_norm": 1.371409186983254, + "learning_rate": 0.000618320610687023, + "loss": 3.7461, + "step": 81 + }, + { + "epoch": 0.01880302682870901, + "grad_norm": 1.3673638514801891, + "learning_rate": 0.0006259541984732825, + "loss": 3.6478, + "step": 82 + }, + { + "epoch": 0.01903233203393717, + "grad_norm": 1.3848742828977436, + "learning_rate": 0.000633587786259542, + "loss": 3.8512, + "step": 83 + }, + { + "epoch": 0.019261637239165328, + "grad_norm": 1.2850019389221101, + "learning_rate": 0.0006412213740458015, + "loss": 3.7993, + "step": 84 + }, + { + "epoch": 0.019490942444393488, + "grad_norm": 1.4832580180173334, + "learning_rate": 0.0006488549618320611, + "loss": 3.6547, + "step": 85 + }, + { + "epoch": 0.019720247649621648, + "grad_norm": 1.6080172339415582, + "learning_rate": 0.0006564885496183206, + "loss": 3.8711, + "step": 86 + }, + { + "epoch": 0.019949552854849804, + "grad_norm": 1.1154400479139512, + "learning_rate": 0.0006641221374045802, + "loss": 3.7169, + "step": 87 + }, + { + "epoch": 0.020178858060077964, + "grad_norm": 1.645372262208749, + "learning_rate": 0.0006717557251908397, + "loss": 3.6277, + "step": 88 + }, + { + "epoch": 0.02040816326530612, + "grad_norm": 1.3893525245974192, + "learning_rate": 0.0006793893129770993, + "loss": 3.6639, + "step": 89 + }, + { + "epoch": 0.02063746847053428, + "grad_norm": 1.0053743979193044, + "learning_rate": 0.0006870229007633588, + "loss": 3.6649, + "step": 90 + }, + { + "epoch": 0.02086677367576244, + "grad_norm": 1.1614418252686407, + "learning_rate": 0.0006946564885496184, + "loss": 3.74, + "step": 91 + }, + { + "epoch": 0.021096078880990597, + "grad_norm": 1.3575380054537982, + "learning_rate": 0.0007022900763358778, + "loss": 3.7537, + "step": 92 + }, + { + "epoch": 0.021325384086218757, + "grad_norm": 0.8479568463773359, + "learning_rate": 0.0007099236641221374, + "loss": 3.7162, + "step": 93 + }, + { + "epoch": 0.021554689291446917, + "grad_norm": 0.9333785701617681, + "learning_rate": 0.000717557251908397, + "loss": 3.7085, + "step": 94 + }, + { + "epoch": 0.021783994496675074, + "grad_norm": 1.2254535687787216, + "learning_rate": 0.0007251908396946565, + "loss": 3.7966, + "step": 95 + }, + { + "epoch": 0.022013299701903234, + "grad_norm": 1.2026348322813127, + "learning_rate": 0.0007328244274809161, + "loss": 3.6644, + "step": 96 + }, + { + "epoch": 0.022242604907131393, + "grad_norm": 1.313103580799827, + "learning_rate": 0.0007404580152671756, + "loss": 3.6401, + "step": 97 + }, + { + "epoch": 0.02247191011235955, + "grad_norm": 1.4553173839481668, + "learning_rate": 0.0007480916030534352, + "loss": 3.7639, + "step": 98 + }, + { + "epoch": 0.02270121531758771, + "grad_norm": 1.4874611160600395, + "learning_rate": 0.0007557251908396947, + "loss": 3.7916, + "step": 99 + }, + { + "epoch": 0.022930520522815866, + "grad_norm": 0.961311524797651, + "learning_rate": 0.0007633587786259543, + "loss": 3.7239, + "step": 100 + }, + { + "epoch": 0.023159825728044026, + "grad_norm": 1.3238883807235111, + "learning_rate": 0.0007709923664122137, + "loss": 3.7521, + "step": 101 + }, + { + "epoch": 0.023389130933272186, + "grad_norm": 1.2368082221691818, + "learning_rate": 0.0007786259541984733, + "loss": 3.6317, + "step": 102 + }, + { + "epoch": 0.023618436138500343, + "grad_norm": 1.0295504364699906, + "learning_rate": 0.0007862595419847328, + "loss": 3.703, + "step": 103 + }, + { + "epoch": 0.023847741343728503, + "grad_norm": 1.6403304293669907, + "learning_rate": 0.0007938931297709924, + "loss": 3.632, + "step": 104 + }, + { + "epoch": 0.024077046548956663, + "grad_norm": 1.2592293742883371, + "learning_rate": 0.0008015267175572519, + "loss": 3.6537, + "step": 105 + }, + { + "epoch": 0.02430635175418482, + "grad_norm": 1.1722378580638597, + "learning_rate": 0.0008091603053435115, + "loss": 3.6283, + "step": 106 + }, + { + "epoch": 0.02453565695941298, + "grad_norm": 1.486360364790245, + "learning_rate": 0.000816793893129771, + "loss": 3.7221, + "step": 107 + }, + { + "epoch": 0.02476496216464114, + "grad_norm": 1.3740626606953747, + "learning_rate": 0.0008244274809160306, + "loss": 3.7322, + "step": 108 + }, + { + "epoch": 0.024994267369869295, + "grad_norm": 1.3500993442128448, + "learning_rate": 0.0008320610687022901, + "loss": 3.6416, + "step": 109 + }, + { + "epoch": 0.025223572575097455, + "grad_norm": 1.4204953440476773, + "learning_rate": 0.0008396946564885496, + "loss": 3.7296, + "step": 110 + }, + { + "epoch": 0.025452877780325612, + "grad_norm": 1.1692671777562835, + "learning_rate": 0.0008473282442748091, + "loss": 3.6707, + "step": 111 + }, + { + "epoch": 0.025682182985553772, + "grad_norm": 1.3082742014808966, + "learning_rate": 0.0008549618320610687, + "loss": 3.8596, + "step": 112 + }, + { + "epoch": 0.025911488190781932, + "grad_norm": 1.2525466738219675, + "learning_rate": 0.0008625954198473283, + "loss": 3.5688, + "step": 113 + }, + { + "epoch": 0.02614079339601009, + "grad_norm": 1.2789522216259475, + "learning_rate": 0.0008702290076335878, + "loss": 3.6605, + "step": 114 + }, + { + "epoch": 0.026370098601238248, + "grad_norm": 0.9353019984471383, + "learning_rate": 0.0008778625954198474, + "loss": 3.6194, + "step": 115 + }, + { + "epoch": 0.026599403806466408, + "grad_norm": 1.1905551040705071, + "learning_rate": 0.0008854961832061069, + "loss": 3.5916, + "step": 116 + }, + { + "epoch": 0.026828709011694565, + "grad_norm": 1.1117767772113902, + "learning_rate": 0.0008931297709923665, + "loss": 3.5889, + "step": 117 + }, + { + "epoch": 0.027058014216922725, + "grad_norm": 1.1615736582968985, + "learning_rate": 0.0009007633587786259, + "loss": 3.5911, + "step": 118 + }, + { + "epoch": 0.027287319422150885, + "grad_norm": 1.0545247904515538, + "learning_rate": 0.0009083969465648855, + "loss": 3.6321, + "step": 119 + }, + { + "epoch": 0.02751662462737904, + "grad_norm": 0.9323799983940683, + "learning_rate": 0.000916030534351145, + "loss": 3.5986, + "step": 120 + }, + { + "epoch": 0.0277459298326072, + "grad_norm": 1.1232069380257588, + "learning_rate": 0.0009236641221374046, + "loss": 3.7034, + "step": 121 + }, + { + "epoch": 0.027975235037835357, + "grad_norm": 1.5021141347218998, + "learning_rate": 0.0009312977099236641, + "loss": 3.5506, + "step": 122 + }, + { + "epoch": 0.028204540243063517, + "grad_norm": 1.2578817530450708, + "learning_rate": 0.0009389312977099237, + "loss": 3.5578, + "step": 123 + }, + { + "epoch": 0.028433845448291677, + "grad_norm": 1.2163769500940085, + "learning_rate": 0.0009465648854961832, + "loss": 3.68, + "step": 124 + }, + { + "epoch": 0.028663150653519834, + "grad_norm": 1.2195019880340285, + "learning_rate": 0.0009541984732824428, + "loss": 3.6427, + "step": 125 + }, + { + "epoch": 0.028892455858747994, + "grad_norm": 0.9789936794037456, + "learning_rate": 0.0009618320610687023, + "loss": 3.5575, + "step": 126 + }, + { + "epoch": 0.029121761063976154, + "grad_norm": 1.0562878148693131, + "learning_rate": 0.0009694656488549618, + "loss": 3.5596, + "step": 127 + }, + { + "epoch": 0.02935106626920431, + "grad_norm": 1.0884006811205837, + "learning_rate": 0.0009770992366412213, + "loss": 3.5169, + "step": 128 + }, + { + "epoch": 0.02958037147443247, + "grad_norm": 1.2436561855869455, + "learning_rate": 0.0009847328244274808, + "loss": 3.7166, + "step": 129 + }, + { + "epoch": 0.029809676679660627, + "grad_norm": 0.8714828373462264, + "learning_rate": 0.0009923664122137405, + "loss": 3.491, + "step": 130 + }, + { + "epoch": 0.030038981884888787, + "grad_norm": 1.137292451130883, + "learning_rate": 0.001, + "loss": 3.4241, + "step": 131 + }, + { + "epoch": 0.030268287090116947, + "grad_norm": 1.0364456612334985, + "learning_rate": 0.0009999998621016723, + "loss": 3.5921, + "step": 132 + }, + { + "epoch": 0.030497592295345103, + "grad_norm": 1.1714409677055964, + "learning_rate": 0.0009999994484067654, + "loss": 3.5809, + "step": 133 + }, + { + "epoch": 0.030726897500573263, + "grad_norm": 0.9500675071445576, + "learning_rate": 0.0009999987589155075, + "loss": 3.4387, + "step": 134 + }, + { + "epoch": 0.030956202705801423, + "grad_norm": 0.9950755587536414, + "learning_rate": 0.0009999977936282788, + "loss": 3.4851, + "step": 135 + }, + { + "epoch": 0.03118550791102958, + "grad_norm": 1.0122469094702187, + "learning_rate": 0.0009999965525456119, + "loss": 3.6214, + "step": 136 + }, + { + "epoch": 0.03141481311625774, + "grad_norm": 1.221925949579536, + "learning_rate": 0.0009999950356681913, + "loss": 3.3586, + "step": 137 + }, + { + "epoch": 0.0316441183214859, + "grad_norm": 0.9106232995773293, + "learning_rate": 0.0009999932429968536, + "loss": 3.563, + "step": 138 + }, + { + "epoch": 0.03187342352671406, + "grad_norm": 1.1296793049433504, + "learning_rate": 0.0009999911745325876, + "loss": 3.6507, + "step": 139 + }, + { + "epoch": 0.03210272873194221, + "grad_norm": 1.1132850880938479, + "learning_rate": 0.0009999888302765346, + "loss": 3.4183, + "step": 140 + }, + { + "epoch": 0.03233203393717037, + "grad_norm": 0.9119552771954645, + "learning_rate": 0.0009999862102299873, + "loss": 3.4932, + "step": 141 + }, + { + "epoch": 0.03256133914239853, + "grad_norm": 1.4882333920676796, + "learning_rate": 0.000999983314394391, + "loss": 3.4328, + "step": 142 + }, + { + "epoch": 0.03279064434762669, + "grad_norm": 1.1216191754543268, + "learning_rate": 0.0009999801427713433, + "loss": 3.5056, + "step": 143 + }, + { + "epoch": 0.03301994955285485, + "grad_norm": 1.1871993462544397, + "learning_rate": 0.000999976695362593, + "loss": 3.4926, + "step": 144 + }, + { + "epoch": 0.03324925475808301, + "grad_norm": 1.254553900932972, + "learning_rate": 0.0009999729721700424, + "loss": 3.3816, + "step": 145 + }, + { + "epoch": 0.033478559963311165, + "grad_norm": 1.074228933849284, + "learning_rate": 0.0009999689731957449, + "loss": 3.3642, + "step": 146 + }, + { + "epoch": 0.033707865168539325, + "grad_norm": 1.4261044002301715, + "learning_rate": 0.000999964698441906, + "loss": 3.5398, + "step": 147 + }, + { + "epoch": 0.033937170373767485, + "grad_norm": 1.4026959041470632, + "learning_rate": 0.000999960147910884, + "loss": 3.5494, + "step": 148 + }, + { + "epoch": 0.034166475578995645, + "grad_norm": 1.032053196964685, + "learning_rate": 0.0009999553216051892, + "loss": 3.402, + "step": 149 + }, + { + "epoch": 0.034395780784223805, + "grad_norm": 1.0203255640846303, + "learning_rate": 0.000999950219527483, + "loss": 3.4487, + "step": 150 + }, + { + "epoch": 0.03462508598945196, + "grad_norm": 2.655366289277392, + "learning_rate": 0.00099994484168058, + "loss": 3.5826, + "step": 151 + }, + { + "epoch": 0.03485439119468012, + "grad_norm": 1.0475157676092506, + "learning_rate": 0.0009999391880674469, + "loss": 3.4501, + "step": 152 + }, + { + "epoch": 0.03508369639990828, + "grad_norm": 0.9884649784558653, + "learning_rate": 0.0009999332586912019, + "loss": 3.6144, + "step": 153 + }, + { + "epoch": 0.03531300160513644, + "grad_norm": 1.2458384570520489, + "learning_rate": 0.0009999270535551157, + "loss": 3.61, + "step": 154 + }, + { + "epoch": 0.0355423068103646, + "grad_norm": 1.045652407817945, + "learning_rate": 0.0009999205726626108, + "loss": 3.2503, + "step": 155 + }, + { + "epoch": 0.03577161201559275, + "grad_norm": 1.0337607391404555, + "learning_rate": 0.0009999138160172623, + "loss": 3.3663, + "step": 156 + }, + { + "epoch": 0.03600091722082091, + "grad_norm": 1.0777086576493733, + "learning_rate": 0.000999906783622797, + "loss": 3.3049, + "step": 157 + }, + { + "epoch": 0.03623022242604907, + "grad_norm": 1.1290542813637987, + "learning_rate": 0.000999899475483094, + "loss": 3.4173, + "step": 158 + }, + { + "epoch": 0.03645952763127723, + "grad_norm": 1.062003019988298, + "learning_rate": 0.0009998918916021842, + "loss": 3.4903, + "step": 159 + }, + { + "epoch": 0.03668883283650539, + "grad_norm": 1.070674829499474, + "learning_rate": 0.000999884031984251, + "loss": 3.5304, + "step": 160 + }, + { + "epoch": 0.03691813804173355, + "grad_norm": 1.1537811508154514, + "learning_rate": 0.0009998758966336297, + "loss": 3.6249, + "step": 161 + }, + { + "epoch": 0.0371474432469617, + "grad_norm": 0.9458276012767253, + "learning_rate": 0.0009998674855548075, + "loss": 3.5036, + "step": 162 + }, + { + "epoch": 0.03737674845218986, + "grad_norm": 1.0383950512244278, + "learning_rate": 0.0009998587987524242, + "loss": 3.5277, + "step": 163 + }, + { + "epoch": 0.03760605365741802, + "grad_norm": 1.1172395375353856, + "learning_rate": 0.0009998498362312714, + "loss": 3.3946, + "step": 164 + }, + { + "epoch": 0.03783535886264618, + "grad_norm": 0.9609060450853866, + "learning_rate": 0.0009998405979962926, + "loss": 3.4285, + "step": 165 + }, + { + "epoch": 0.03806466406787434, + "grad_norm": 1.142265047385587, + "learning_rate": 0.0009998310840525833, + "loss": 3.3064, + "step": 166 + }, + { + "epoch": 0.038293969273102496, + "grad_norm": 1.138922678010045, + "learning_rate": 0.000999821294405392, + "loss": 3.3877, + "step": 167 + }, + { + "epoch": 0.038523274478330656, + "grad_norm": 1.2111400415539255, + "learning_rate": 0.0009998112290601179, + "loss": 3.3752, + "step": 168 + }, + { + "epoch": 0.038752579683558816, + "grad_norm": 1.1335128501537075, + "learning_rate": 0.0009998008880223134, + "loss": 3.4471, + "step": 169 + }, + { + "epoch": 0.038981884888786976, + "grad_norm": 1.0795860265679984, + "learning_rate": 0.0009997902712976824, + "loss": 3.4655, + "step": 170 + }, + { + "epoch": 0.039211190094015136, + "grad_norm": 0.9835498254971039, + "learning_rate": 0.000999779378892081, + "loss": 3.3123, + "step": 171 + }, + { + "epoch": 0.039440495299243296, + "grad_norm": 1.034562182523929, + "learning_rate": 0.0009997682108115176, + "loss": 3.2369, + "step": 172 + }, + { + "epoch": 0.03966980050447145, + "grad_norm": 1.0623380009015113, + "learning_rate": 0.0009997567670621522, + "loss": 3.2932, + "step": 173 + }, + { + "epoch": 0.03989910570969961, + "grad_norm": 1.3384801429781477, + "learning_rate": 0.000999745047650297, + "loss": 3.4682, + "step": 174 + }, + { + "epoch": 0.04012841091492777, + "grad_norm": 1.2694620662496918, + "learning_rate": 0.0009997330525824165, + "loss": 3.3586, + "step": 175 + }, + { + "epoch": 0.04035771612015593, + "grad_norm": 0.971496903836351, + "learning_rate": 0.0009997207818651273, + "loss": 3.2737, + "step": 176 + }, + { + "epoch": 0.04058702132538409, + "grad_norm": 1.0828837854987434, + "learning_rate": 0.0009997082355051976, + "loss": 3.4618, + "step": 177 + }, + { + "epoch": 0.04081632653061224, + "grad_norm": 1.021151045929391, + "learning_rate": 0.0009996954135095479, + "loss": 3.3249, + "step": 178 + }, + { + "epoch": 0.0410456317358404, + "grad_norm": 1.1266832037486092, + "learning_rate": 0.000999682315885251, + "loss": 3.4349, + "step": 179 + }, + { + "epoch": 0.04127493694106856, + "grad_norm": 1.1164390762099567, + "learning_rate": 0.0009996689426395307, + "loss": 3.3486, + "step": 180 + }, + { + "epoch": 0.04150424214629672, + "grad_norm": 1.057576897228051, + "learning_rate": 0.0009996552937797645, + "loss": 3.4308, + "step": 181 + }, + { + "epoch": 0.04173354735152488, + "grad_norm": 1.0906076885433837, + "learning_rate": 0.0009996413693134805, + "loss": 3.3097, + "step": 182 + }, + { + "epoch": 0.04196285255675304, + "grad_norm": 1.288777444726708, + "learning_rate": 0.0009996271692483596, + "loss": 3.223, + "step": 183 + }, + { + "epoch": 0.042192157761981194, + "grad_norm": 1.0853940568590343, + "learning_rate": 0.0009996126935922342, + "loss": 3.3798, + "step": 184 + }, + { + "epoch": 0.042421462967209354, + "grad_norm": 1.0019366279442203, + "learning_rate": 0.0009995979423530893, + "loss": 3.3368, + "step": 185 + }, + { + "epoch": 0.042650768172437514, + "grad_norm": 1.0898888444893517, + "learning_rate": 0.0009995829155390612, + "loss": 3.2859, + "step": 186 + }, + { + "epoch": 0.042880073377665674, + "grad_norm": 1.1295669536010502, + "learning_rate": 0.000999567613158439, + "loss": 3.2967, + "step": 187 + }, + { + "epoch": 0.043109378582893834, + "grad_norm": 0.9548335608499282, + "learning_rate": 0.000999552035219663, + "loss": 3.2426, + "step": 188 + }, + { + "epoch": 0.04333868378812199, + "grad_norm": 1.3243135168783589, + "learning_rate": 0.0009995361817313263, + "loss": 3.3668, + "step": 189 + }, + { + "epoch": 0.04356798899335015, + "grad_norm": 0.9458823626717223, + "learning_rate": 0.000999520052702173, + "loss": 3.2774, + "step": 190 + }, + { + "epoch": 0.04379729419857831, + "grad_norm": 1.1610321809645539, + "learning_rate": 0.0009995036481411004, + "loss": 3.308, + "step": 191 + }, + { + "epoch": 0.04402659940380647, + "grad_norm": 0.977824215850698, + "learning_rate": 0.0009994869680571567, + "loss": 3.3226, + "step": 192 + }, + { + "epoch": 0.04425590460903463, + "grad_norm": 1.0900848726826127, + "learning_rate": 0.0009994700124595429, + "loss": 3.4231, + "step": 193 + }, + { + "epoch": 0.04448520981426279, + "grad_norm": 1.1639120687083, + "learning_rate": 0.000999452781357611, + "loss": 3.3758, + "step": 194 + }, + { + "epoch": 0.04471451501949094, + "grad_norm": 1.0035618983609145, + "learning_rate": 0.0009994352747608663, + "loss": 3.2155, + "step": 195 + }, + { + "epoch": 0.0449438202247191, + "grad_norm": 1.1327229816023607, + "learning_rate": 0.0009994174926789647, + "loss": 3.3447, + "step": 196 + }, + { + "epoch": 0.04517312542994726, + "grad_norm": 0.9301783418701456, + "learning_rate": 0.0009993994351217151, + "loss": 3.344, + "step": 197 + }, + { + "epoch": 0.04540243063517542, + "grad_norm": 1.1187600644943954, + "learning_rate": 0.0009993811020990778, + "loss": 3.3586, + "step": 198 + }, + { + "epoch": 0.04563173584040358, + "grad_norm": 1.153374166908871, + "learning_rate": 0.000999362493621165, + "loss": 3.3738, + "step": 199 + }, + { + "epoch": 0.04586104104563173, + "grad_norm": 1.1873371912438468, + "learning_rate": 0.0009993436096982413, + "loss": 3.2883, + "step": 200 + }, + { + "epoch": 0.04609034625085989, + "grad_norm": 1.0728664436302189, + "learning_rate": 0.0009993244503407226, + "loss": 3.3925, + "step": 201 + }, + { + "epoch": 0.04631965145608805, + "grad_norm": 1.1597971597425816, + "learning_rate": 0.0009993050155591775, + "loss": 3.3938, + "step": 202 + }, + { + "epoch": 0.04654895666131621, + "grad_norm": 1.0814629641930933, + "learning_rate": 0.0009992853053643258, + "loss": 3.2363, + "step": 203 + }, + { + "epoch": 0.04677826186654437, + "grad_norm": 1.0643628224186295, + "learning_rate": 0.0009992653197670393, + "loss": 3.2865, + "step": 204 + }, + { + "epoch": 0.04700756707177253, + "grad_norm": 1.1278350576369938, + "learning_rate": 0.0009992450587783426, + "loss": 3.2514, + "step": 205 + }, + { + "epoch": 0.047236872277000685, + "grad_norm": 1.0427722665619799, + "learning_rate": 0.0009992245224094108, + "loss": 3.3365, + "step": 206 + }, + { + "epoch": 0.047466177482228845, + "grad_norm": 2.167880796155055, + "learning_rate": 0.000999203710671572, + "loss": 3.2142, + "step": 207 + }, + { + "epoch": 0.047695482687457005, + "grad_norm": 1.1496632004661065, + "learning_rate": 0.000999182623576306, + "loss": 3.181, + "step": 208 + }, + { + "epoch": 0.047924787892685165, + "grad_norm": 0.9003746203865626, + "learning_rate": 0.0009991612611352438, + "loss": 3.2984, + "step": 209 + }, + { + "epoch": 0.048154093097913325, + "grad_norm": 1.6112700708350443, + "learning_rate": 0.0009991396233601693, + "loss": 3.3348, + "step": 210 + }, + { + "epoch": 0.04838339830314148, + "grad_norm": 3.028136822203153, + "learning_rate": 0.0009991177102630173, + "loss": 3.2876, + "step": 211 + }, + { + "epoch": 0.04861270350836964, + "grad_norm": 1.0463689422450377, + "learning_rate": 0.0009990955218558751, + "loss": 3.1823, + "step": 212 + }, + { + "epoch": 0.0488420087135978, + "grad_norm": 0.9555840254060373, + "learning_rate": 0.0009990730581509817, + "loss": 3.2277, + "step": 213 + }, + { + "epoch": 0.04907131391882596, + "grad_norm": 1.0389375896589454, + "learning_rate": 0.000999050319160728, + "loss": 3.2467, + "step": 214 + }, + { + "epoch": 0.04930061912405412, + "grad_norm": 1.004388679381557, + "learning_rate": 0.0009990273048976566, + "loss": 3.1771, + "step": 215 + }, + { + "epoch": 0.04952992432928228, + "grad_norm": 0.9777513635352231, + "learning_rate": 0.0009990040153744619, + "loss": 3.2885, + "step": 216 + }, + { + "epoch": 0.04975922953451043, + "grad_norm": 1.304643595455851, + "learning_rate": 0.0009989804506039905, + "loss": 3.1976, + "step": 217 + }, + { + "epoch": 0.04998853473973859, + "grad_norm": 0.9571725192615003, + "learning_rate": 0.0009989566105992403, + "loss": 3.3532, + "step": 218 + }, + { + "epoch": 0.05021783994496675, + "grad_norm": 1.1457824516292576, + "learning_rate": 0.0009989324953733614, + "loss": 3.3044, + "step": 219 + }, + { + "epoch": 0.05044714515019491, + "grad_norm": 1.0406355481610479, + "learning_rate": 0.0009989081049396557, + "loss": 3.1551, + "step": 220 + }, + { + "epoch": 0.05067645035542307, + "grad_norm": 1.0599438564902408, + "learning_rate": 0.0009988834393115766, + "loss": 3.1607, + "step": 221 + }, + { + "epoch": 0.050905755560651224, + "grad_norm": 1.2010848538067076, + "learning_rate": 0.0009988584985027296, + "loss": 3.2686, + "step": 222 + }, + { + "epoch": 0.051135060765879384, + "grad_norm": 1.0274972925932762, + "learning_rate": 0.000998833282526872, + "loss": 3.2456, + "step": 223 + }, + { + "epoch": 0.051364365971107544, + "grad_norm": 1.0009277746643406, + "learning_rate": 0.0009988077913979125, + "loss": 3.3953, + "step": 224 + }, + { + "epoch": 0.051593671176335704, + "grad_norm": 1.4574029831215065, + "learning_rate": 0.0009987820251299122, + "loss": 3.3138, + "step": 225 + }, + { + "epoch": 0.051822976381563864, + "grad_norm": 0.8934497425747058, + "learning_rate": 0.0009987559837370832, + "loss": 3.1325, + "step": 226 + }, + { + "epoch": 0.052052281586792024, + "grad_norm": 0.9285886844438986, + "learning_rate": 0.00099872966723379, + "loss": 3.3287, + "step": 227 + }, + { + "epoch": 0.05228158679202018, + "grad_norm": 1.0079443482006392, + "learning_rate": 0.0009987030756345486, + "loss": 3.3361, + "step": 228 + }, + { + "epoch": 0.052510891997248337, + "grad_norm": 0.8719242499879856, + "learning_rate": 0.0009986762089540266, + "loss": 3.2021, + "step": 229 + }, + { + "epoch": 0.052740197202476496, + "grad_norm": 1.0425837059275858, + "learning_rate": 0.0009986490672070436, + "loss": 3.1574, + "step": 230 + }, + { + "epoch": 0.052969502407704656, + "grad_norm": 0.9250813240435634, + "learning_rate": 0.0009986216504085709, + "loss": 3.0972, + "step": 231 + }, + { + "epoch": 0.053198807612932816, + "grad_norm": 0.9394997575751759, + "learning_rate": 0.0009985939585737311, + "loss": 3.205, + "step": 232 + }, + { + "epoch": 0.05342811281816097, + "grad_norm": 1.1436388764777794, + "learning_rate": 0.0009985659917177991, + "loss": 3.1559, + "step": 233 + }, + { + "epoch": 0.05365741802338913, + "grad_norm": 0.9984568095528824, + "learning_rate": 0.0009985377498562013, + "loss": 3.225, + "step": 234 + }, + { + "epoch": 0.05388672322861729, + "grad_norm": 0.9522568621882477, + "learning_rate": 0.0009985092330045155, + "loss": 3.2499, + "step": 235 + }, + { + "epoch": 0.05411602843384545, + "grad_norm": 0.84315761400127, + "learning_rate": 0.0009984804411784715, + "loss": 3.318, + "step": 236 + }, + { + "epoch": 0.05434533363907361, + "grad_norm": 1.2205057168307794, + "learning_rate": 0.0009984513743939508, + "loss": 3.2882, + "step": 237 + }, + { + "epoch": 0.05457463884430177, + "grad_norm": 0.9698624226029762, + "learning_rate": 0.0009984220326669862, + "loss": 3.2309, + "step": 238 + }, + { + "epoch": 0.05480394404952992, + "grad_norm": 0.9452031708822771, + "learning_rate": 0.0009983924160137626, + "loss": 3.1691, + "step": 239 + }, + { + "epoch": 0.05503324925475808, + "grad_norm": 0.8736428043772054, + "learning_rate": 0.0009983625244506162, + "loss": 3.1989, + "step": 240 + }, + { + "epoch": 0.05526255445998624, + "grad_norm": 0.843316796499873, + "learning_rate": 0.000998332357994035, + "loss": 3.0978, + "step": 241 + }, + { + "epoch": 0.0554918596652144, + "grad_norm": 1.3921214089500296, + "learning_rate": 0.0009983019166606589, + "loss": 3.2572, + "step": 242 + }, + { + "epoch": 0.05572116487044256, + "grad_norm": 0.9751176192438809, + "learning_rate": 0.0009982712004672786, + "loss": 3.1578, + "step": 243 + }, + { + "epoch": 0.055950470075670715, + "grad_norm": 1.0239918636751226, + "learning_rate": 0.0009982402094308373, + "loss": 3.2572, + "step": 244 + }, + { + "epoch": 0.056179775280898875, + "grad_norm": 7.819389299742299, + "learning_rate": 0.0009982089435684295, + "loss": 3.2374, + "step": 245 + }, + { + "epoch": 0.056409080486127035, + "grad_norm": 1.0050795100212049, + "learning_rate": 0.0009981774028973013, + "loss": 3.25, + "step": 246 + }, + { + "epoch": 0.056638385691355195, + "grad_norm": 1.2705976862806314, + "learning_rate": 0.0009981455874348499, + "loss": 3.2297, + "step": 247 + }, + { + "epoch": 0.056867690896583355, + "grad_norm": 1.6563562763905781, + "learning_rate": 0.0009981134971986248, + "loss": 3.2024, + "step": 248 + }, + { + "epoch": 0.05709699610181151, + "grad_norm": 3.8102609430811745, + "learning_rate": 0.0009980811322063269, + "loss": 3.0728, + "step": 249 + }, + { + "epoch": 0.05732630130703967, + "grad_norm": 2.9878557402074737, + "learning_rate": 0.0009980484924758083, + "loss": 3.3346, + "step": 250 + }, + { + "epoch": 0.05755560651226783, + "grad_norm": 1.0762401398518555, + "learning_rate": 0.0009980155780250728, + "loss": 3.0971, + "step": 251 + }, + { + "epoch": 0.05778491171749599, + "grad_norm": 1.0873643994388487, + "learning_rate": 0.0009979823888722759, + "loss": 3.2877, + "step": 252 + }, + { + "epoch": 0.05801421692272415, + "grad_norm": 1.346699136085296, + "learning_rate": 0.0009979489250357243, + "loss": 3.26, + "step": 253 + }, + { + "epoch": 0.05824352212795231, + "grad_norm": 1.0735072635275993, + "learning_rate": 0.000997915186533877, + "loss": 3.0959, + "step": 254 + }, + { + "epoch": 0.05847282733318046, + "grad_norm": 1.1011660382793467, + "learning_rate": 0.0009978811733853431, + "loss": 3.2332, + "step": 255 + }, + { + "epoch": 0.05870213253840862, + "grad_norm": 1.1180156548382405, + "learning_rate": 0.000997846885608885, + "loss": 3.2701, + "step": 256 + }, + { + "epoch": 0.05893143774363678, + "grad_norm": 1.2780734576750397, + "learning_rate": 0.0009978123232234147, + "loss": 3.172, + "step": 257 + }, + { + "epoch": 0.05916074294886494, + "grad_norm": 2.1847801527787176, + "learning_rate": 0.0009977774862479972, + "loss": 3.2837, + "step": 258 + }, + { + "epoch": 0.0593900481540931, + "grad_norm": 1.4019510321227602, + "learning_rate": 0.000997742374701848, + "loss": 3.0309, + "step": 259 + }, + { + "epoch": 0.05961935335932125, + "grad_norm": 0.9934766339423722, + "learning_rate": 0.0009977069886043344, + "loss": 3.251, + "step": 260 + }, + { + "epoch": 0.05984865856454941, + "grad_norm": 1.1791812331589984, + "learning_rate": 0.0009976713279749754, + "loss": 3.2611, + "step": 261 + }, + { + "epoch": 0.06007796376977757, + "grad_norm": 1.2747624055030427, + "learning_rate": 0.0009976353928334409, + "loss": 3.2223, + "step": 262 + }, + { + "epoch": 0.06030726897500573, + "grad_norm": 1.1707972506611497, + "learning_rate": 0.0009975991831995528, + "loss": 3.1925, + "step": 263 + }, + { + "epoch": 0.06053657418023389, + "grad_norm": 2.670792838623812, + "learning_rate": 0.0009975626990932837, + "loss": 3.1786, + "step": 264 + }, + { + "epoch": 0.06076587938546205, + "grad_norm": 1.0997698649526895, + "learning_rate": 0.0009975259405347581, + "loss": 3.1672, + "step": 265 + }, + { + "epoch": 0.060995184590690206, + "grad_norm": 1.036433825727608, + "learning_rate": 0.0009974889075442521, + "loss": 3.2499, + "step": 266 + }, + { + "epoch": 0.061224489795918366, + "grad_norm": 1.2418420105704162, + "learning_rate": 0.0009974516001421926, + "loss": 3.1985, + "step": 267 + }, + { + "epoch": 0.061453795001146526, + "grad_norm": 0.959294350488388, + "learning_rate": 0.0009974140183491578, + "loss": 3.1524, + "step": 268 + }, + { + "epoch": 0.061683100206374686, + "grad_norm": 1.2518633504393066, + "learning_rate": 0.000997376162185878, + "loss": 3.1648, + "step": 269 + }, + { + "epoch": 0.061912405411602846, + "grad_norm": 1.0581775056768956, + "learning_rate": 0.0009973380316732343, + "loss": 3.1934, + "step": 270 + }, + { + "epoch": 0.062141710616831, + "grad_norm": 0.909088118732758, + "learning_rate": 0.0009972996268322594, + "loss": 3.1933, + "step": 271 + }, + { + "epoch": 0.06237101582205916, + "grad_norm": 0.9010582473989799, + "learning_rate": 0.0009972609476841367, + "loss": 3.223, + "step": 272 + }, + { + "epoch": 0.06260032102728733, + "grad_norm": 1.0639569159915934, + "learning_rate": 0.0009972219942502017, + "loss": 3.1512, + "step": 273 + }, + { + "epoch": 0.06282962623251548, + "grad_norm": 1.0574852345732595, + "learning_rate": 0.000997182766551941, + "loss": 3.159, + "step": 274 + }, + { + "epoch": 0.06305893143774363, + "grad_norm": 0.8548345761995526, + "learning_rate": 0.0009971432646109918, + "loss": 3.058, + "step": 275 + }, + { + "epoch": 0.0632882366429718, + "grad_norm": 0.8868324385755864, + "learning_rate": 0.0009971034884491436, + "loss": 3.0911, + "step": 276 + }, + { + "epoch": 0.06351754184819995, + "grad_norm": 0.9482194723272221, + "learning_rate": 0.0009970634380883365, + "loss": 3.0641, + "step": 277 + }, + { + "epoch": 0.06374684705342812, + "grad_norm": 1.052675566562753, + "learning_rate": 0.0009970231135506618, + "loss": 3.168, + "step": 278 + }, + { + "epoch": 0.06397615225865627, + "grad_norm": 1.6799634023160757, + "learning_rate": 0.0009969825148583627, + "loss": 3.107, + "step": 279 + }, + { + "epoch": 0.06420545746388442, + "grad_norm": 1.0285766179795208, + "learning_rate": 0.0009969416420338329, + "loss": 3.0272, + "step": 280 + }, + { + "epoch": 0.06443476266911259, + "grad_norm": 0.9637523631099039, + "learning_rate": 0.0009969004950996173, + "loss": 3.1189, + "step": 281 + }, + { + "epoch": 0.06466406787434074, + "grad_norm": 0.8979891527308242, + "learning_rate": 0.000996859074078413, + "loss": 3.1781, + "step": 282 + }, + { + "epoch": 0.06489337307956891, + "grad_norm": 0.9381681228832114, + "learning_rate": 0.0009968173789930668, + "loss": 3.2497, + "step": 283 + }, + { + "epoch": 0.06512267828479706, + "grad_norm": 0.9908594937876559, + "learning_rate": 0.0009967754098665778, + "loss": 3.0489, + "step": 284 + }, + { + "epoch": 0.06535198349002522, + "grad_norm": 1.1703556055299162, + "learning_rate": 0.0009967331667220958, + "loss": 3.0562, + "step": 285 + }, + { + "epoch": 0.06558128869525338, + "grad_norm": 0.9837003244162177, + "learning_rate": 0.000996690649582922, + "loss": 3.1327, + "step": 286 + }, + { + "epoch": 0.06581059390048154, + "grad_norm": 2.208418391236925, + "learning_rate": 0.0009966478584725086, + "loss": 3.1995, + "step": 287 + }, + { + "epoch": 0.0660398991057097, + "grad_norm": 0.9243395884473828, + "learning_rate": 0.0009966047934144584, + "loss": 3.1817, + "step": 288 + }, + { + "epoch": 0.06626920431093786, + "grad_norm": 0.9873502938409483, + "learning_rate": 0.0009965614544325263, + "loss": 3.2363, + "step": 289 + }, + { + "epoch": 0.06649850951616602, + "grad_norm": 0.947398968152614, + "learning_rate": 0.0009965178415506175, + "loss": 2.9103, + "step": 290 + }, + { + "epoch": 0.06672781472139418, + "grad_norm": 0.862002501156526, + "learning_rate": 0.000996473954792789, + "loss": 3.1955, + "step": 291 + }, + { + "epoch": 0.06695711992662233, + "grad_norm": 0.8404930885626188, + "learning_rate": 0.0009964297941832479, + "loss": 3.0126, + "step": 292 + }, + { + "epoch": 0.0671864251318505, + "grad_norm": 0.9117853136351894, + "learning_rate": 0.0009963853597463532, + "loss": 2.9866, + "step": 293 + }, + { + "epoch": 0.06741573033707865, + "grad_norm": 1.0518412821468357, + "learning_rate": 0.0009963406515066148, + "loss": 2.9505, + "step": 294 + }, + { + "epoch": 0.06764503554230682, + "grad_norm": 0.8630530426598456, + "learning_rate": 0.000996295669488693, + "loss": 3.1561, + "step": 295 + }, + { + "epoch": 0.06787434074753497, + "grad_norm": 0.8636358094022285, + "learning_rate": 0.0009962504137173997, + "loss": 3.0899, + "step": 296 + }, + { + "epoch": 0.06810364595276312, + "grad_norm": 0.8543047521315504, + "learning_rate": 0.0009962048842176979, + "loss": 3.019, + "step": 297 + }, + { + "epoch": 0.06833295115799129, + "grad_norm": 0.8067362550628355, + "learning_rate": 0.0009961590810147014, + "loss": 3.0485, + "step": 298 + }, + { + "epoch": 0.06856225636321944, + "grad_norm": 0.8599453574752765, + "learning_rate": 0.0009961130041336748, + "loss": 3.1779, + "step": 299 + }, + { + "epoch": 0.06879156156844761, + "grad_norm": 1.1172794058898665, + "learning_rate": 0.0009960666536000338, + "loss": 2.9346, + "step": 300 + }, + { + "epoch": 0.06902086677367576, + "grad_norm": 0.8584307444564421, + "learning_rate": 0.0009960200294393449, + "loss": 2.8795, + "step": 301 + }, + { + "epoch": 0.06925017197890392, + "grad_norm": 0.8688547498984888, + "learning_rate": 0.000995973131677326, + "loss": 2.9967, + "step": 302 + }, + { + "epoch": 0.06947947718413208, + "grad_norm": 0.8989380348953748, + "learning_rate": 0.0009959259603398453, + "loss": 3.0073, + "step": 303 + }, + { + "epoch": 0.06970878238936024, + "grad_norm": 0.9601397796803984, + "learning_rate": 0.0009958785154529225, + "loss": 3.0658, + "step": 304 + }, + { + "epoch": 0.0699380875945884, + "grad_norm": 0.8922700674332504, + "learning_rate": 0.0009958307970427275, + "loss": 3.0652, + "step": 305 + }, + { + "epoch": 0.07016739279981656, + "grad_norm": 2.7100088918497502, + "learning_rate": 0.0009957828051355816, + "loss": 3.0306, + "step": 306 + }, + { + "epoch": 0.07039669800504471, + "grad_norm": 0.9106570103062104, + "learning_rate": 0.0009957345397579572, + "loss": 2.9837, + "step": 307 + }, + { + "epoch": 0.07062600321027288, + "grad_norm": 1.025150541399293, + "learning_rate": 0.0009956860009364766, + "loss": 3.026, + "step": 308 + }, + { + "epoch": 0.07085530841550103, + "grad_norm": 0.9178635886980168, + "learning_rate": 0.0009956371886979138, + "loss": 3.0885, + "step": 309 + }, + { + "epoch": 0.0710846136207292, + "grad_norm": 0.8705055567978421, + "learning_rate": 0.000995588103069193, + "loss": 2.9579, + "step": 310 + }, + { + "epoch": 0.07131391882595735, + "grad_norm": 1.2627409068728102, + "learning_rate": 0.00099553874407739, + "loss": 3.0095, + "step": 311 + }, + { + "epoch": 0.0715432240311855, + "grad_norm": 1.09360093098442, + "learning_rate": 0.0009954891117497306, + "loss": 3.158, + "step": 312 + }, + { + "epoch": 0.07177252923641367, + "grad_norm": 0.9672812927927841, + "learning_rate": 0.0009954392061135916, + "loss": 3.0698, + "step": 313 + }, + { + "epoch": 0.07200183444164182, + "grad_norm": 1.170671285111154, + "learning_rate": 0.0009953890271965006, + "loss": 3.0059, + "step": 314 + }, + { + "epoch": 0.07223113964686999, + "grad_norm": 1.1836176046023192, + "learning_rate": 0.0009953385750261364, + "loss": 2.9707, + "step": 315 + }, + { + "epoch": 0.07246044485209814, + "grad_norm": 3.3797758697518097, + "learning_rate": 0.0009952878496303273, + "loss": 2.9177, + "step": 316 + }, + { + "epoch": 0.07268975005732631, + "grad_norm": 1.2214210320378394, + "learning_rate": 0.0009952368510370538, + "loss": 2.9885, + "step": 317 + }, + { + "epoch": 0.07291905526255446, + "grad_norm": 1.1095405623412478, + "learning_rate": 0.000995185579274446, + "loss": 2.8863, + "step": 318 + }, + { + "epoch": 0.07314836046778261, + "grad_norm": 1.0379026309321624, + "learning_rate": 0.0009951340343707852, + "loss": 3.0298, + "step": 319 + }, + { + "epoch": 0.07337766567301078, + "grad_norm": 5.154761075849086, + "learning_rate": 0.000995082216354503, + "loss": 3.0098, + "step": 320 + }, + { + "epoch": 0.07360697087823893, + "grad_norm": 1.2402248543643328, + "learning_rate": 0.0009950301252541823, + "loss": 2.9633, + "step": 321 + }, + { + "epoch": 0.0738362760834671, + "grad_norm": 0.9827538213041006, + "learning_rate": 0.000994977761098556, + "loss": 2.8997, + "step": 322 + }, + { + "epoch": 0.07406558128869525, + "grad_norm": 1.806223150421278, + "learning_rate": 0.0009949251239165075, + "loss": 2.9671, + "step": 323 + }, + { + "epoch": 0.0742948864939234, + "grad_norm": 2.8090959268140807, + "learning_rate": 0.0009948722137370716, + "loss": 2.9243, + "step": 324 + }, + { + "epoch": 0.07452419169915157, + "grad_norm": 1.16442668709804, + "learning_rate": 0.000994819030589433, + "loss": 2.9007, + "step": 325 + }, + { + "epoch": 0.07475349690437973, + "grad_norm": 0.950574711982085, + "learning_rate": 0.000994765574502927, + "loss": 2.8594, + "step": 326 + }, + { + "epoch": 0.0749828021096079, + "grad_norm": 0.9333147482105119, + "learning_rate": 0.00099471184550704, + "loss": 2.8726, + "step": 327 + }, + { + "epoch": 0.07521210731483605, + "grad_norm": 1.019676888265282, + "learning_rate": 0.0009946578436314081, + "loss": 3.039, + "step": 328 + }, + { + "epoch": 0.0754414125200642, + "grad_norm": 1.0900137773937537, + "learning_rate": 0.0009946035689058189, + "loss": 3.0159, + "step": 329 + }, + { + "epoch": 0.07567071772529237, + "grad_norm": 0.8104336271798247, + "learning_rate": 0.0009945490213602093, + "loss": 2.882, + "step": 330 + }, + { + "epoch": 0.07590002293052052, + "grad_norm": 1.0385395401206012, + "learning_rate": 0.0009944942010246681, + "loss": 2.8353, + "step": 331 + }, + { + "epoch": 0.07612932813574869, + "grad_norm": 1.3779976013980697, + "learning_rate": 0.0009944391079294334, + "loss": 3.1837, + "step": 332 + }, + { + "epoch": 0.07635863334097684, + "grad_norm": 1.0425197199750043, + "learning_rate": 0.0009943837421048942, + "loss": 3.0145, + "step": 333 + }, + { + "epoch": 0.07658793854620499, + "grad_norm": 2.0983332147130582, + "learning_rate": 0.00099432810358159, + "loss": 3.0756, + "step": 334 + }, + { + "epoch": 0.07681724375143316, + "grad_norm": 0.905610088395742, + "learning_rate": 0.0009942721923902106, + "loss": 2.9677, + "step": 335 + }, + { + "epoch": 0.07704654895666131, + "grad_norm": 1.0809217391615422, + "learning_rate": 0.0009942160085615961, + "loss": 2.9796, + "step": 336 + }, + { + "epoch": 0.07727585416188948, + "grad_norm": 0.9263874365247076, + "learning_rate": 0.0009941595521267377, + "loss": 3.043, + "step": 337 + }, + { + "epoch": 0.07750515936711763, + "grad_norm": 3.315849084021983, + "learning_rate": 0.0009941028231167757, + "loss": 3.0335, + "step": 338 + }, + { + "epoch": 0.0777344645723458, + "grad_norm": 6.8892127158012295, + "learning_rate": 0.0009940458215630017, + "loss": 2.9292, + "step": 339 + }, + { + "epoch": 0.07796376977757395, + "grad_norm": 1.155735912518752, + "learning_rate": 0.0009939885474968572, + "loss": 2.9437, + "step": 340 + }, + { + "epoch": 0.0781930749828021, + "grad_norm": 0.9481527161280847, + "learning_rate": 0.0009939310009499348, + "loss": 2.9579, + "step": 341 + }, + { + "epoch": 0.07842238018803027, + "grad_norm": 0.997952770519535, + "learning_rate": 0.0009938731819539761, + "loss": 2.8857, + "step": 342 + }, + { + "epoch": 0.07865168539325842, + "grad_norm": 2.671444412856791, + "learning_rate": 0.000993815090540874, + "loss": 3.0917, + "step": 343 + }, + { + "epoch": 0.07888099059848659, + "grad_norm": 1.006180158399254, + "learning_rate": 0.0009937567267426714, + "loss": 2.9792, + "step": 344 + }, + { + "epoch": 0.07911029580371474, + "grad_norm": 1.1974972782076565, + "learning_rate": 0.000993698090591561, + "loss": 3.0216, + "step": 345 + }, + { + "epoch": 0.0793396010089429, + "grad_norm": 0.9450082928959268, + "learning_rate": 0.0009936391821198866, + "loss": 2.9482, + "step": 346 + }, + { + "epoch": 0.07956890621417106, + "grad_norm": 1.212766194959965, + "learning_rate": 0.0009935800013601416, + "loss": 2.8871, + "step": 347 + }, + { + "epoch": 0.07979821141939922, + "grad_norm": 1.044671542840282, + "learning_rate": 0.0009935205483449694, + "loss": 2.8302, + "step": 348 + }, + { + "epoch": 0.08002751662462738, + "grad_norm": 0.9633956302998411, + "learning_rate": 0.000993460823107164, + "loss": 2.9648, + "step": 349 + }, + { + "epoch": 0.08025682182985554, + "grad_norm": 0.8128119783497872, + "learning_rate": 0.0009934008256796696, + "loss": 3.0528, + "step": 350 + }, + { + "epoch": 0.08048612703508369, + "grad_norm": 1.0563647322571343, + "learning_rate": 0.0009933405560955803, + "loss": 2.9304, + "step": 351 + }, + { + "epoch": 0.08071543224031186, + "grad_norm": 0.8231782979913692, + "learning_rate": 0.0009932800143881404, + "loss": 3.0125, + "step": 352 + }, + { + "epoch": 0.08094473744554001, + "grad_norm": 0.8913153343617132, + "learning_rate": 0.0009932192005907446, + "loss": 2.9225, + "step": 353 + }, + { + "epoch": 0.08117404265076818, + "grad_norm": 0.7669451689637047, + "learning_rate": 0.0009931581147369368, + "loss": 2.873, + "step": 354 + }, + { + "epoch": 0.08140334785599633, + "grad_norm": 0.9523078229719566, + "learning_rate": 0.0009930967568604118, + "loss": 2.9176, + "step": 355 + }, + { + "epoch": 0.08163265306122448, + "grad_norm": 0.9226895865435402, + "learning_rate": 0.0009930351269950143, + "loss": 2.978, + "step": 356 + }, + { + "epoch": 0.08186195826645265, + "grad_norm": 0.9396520606709691, + "learning_rate": 0.000992973225174739, + "loss": 3.0344, + "step": 357 + }, + { + "epoch": 0.0820912634716808, + "grad_norm": 1.2785613682144803, + "learning_rate": 0.00099291105143373, + "loss": 2.9049, + "step": 358 + }, + { + "epoch": 0.08232056867690897, + "grad_norm": 0.8997170979840782, + "learning_rate": 0.0009928486058062827, + "loss": 3.0059, + "step": 359 + }, + { + "epoch": 0.08254987388213712, + "grad_norm": 0.782548134567164, + "learning_rate": 0.000992785888326841, + "loss": 2.8835, + "step": 360 + }, + { + "epoch": 0.08277917908736529, + "grad_norm": 0.9492404589907806, + "learning_rate": 0.0009927228990299999, + "loss": 2.9134, + "step": 361 + }, + { + "epoch": 0.08300848429259344, + "grad_norm": 0.8374066212043736, + "learning_rate": 0.0009926596379505037, + "loss": 2.8325, + "step": 362 + }, + { + "epoch": 0.0832377894978216, + "grad_norm": 0.8405849685844833, + "learning_rate": 0.0009925961051232468, + "loss": 2.9656, + "step": 363 + }, + { + "epoch": 0.08346709470304976, + "grad_norm": 0.9267581625200724, + "learning_rate": 0.0009925323005832735, + "loss": 2.9874, + "step": 364 + }, + { + "epoch": 0.08369639990827792, + "grad_norm": 0.8042331319506728, + "learning_rate": 0.000992468224365778, + "loss": 3.063, + "step": 365 + }, + { + "epoch": 0.08392570511350608, + "grad_norm": 0.8582192568791892, + "learning_rate": 0.000992403876506104, + "loss": 2.9403, + "step": 366 + }, + { + "epoch": 0.08415501031873424, + "grad_norm": 0.8443647710188339, + "learning_rate": 0.000992339257039746, + "loss": 2.9177, + "step": 367 + }, + { + "epoch": 0.08438431552396239, + "grad_norm": 0.9930849818130272, + "learning_rate": 0.0009922743660023467, + "loss": 2.8824, + "step": 368 + }, + { + "epoch": 0.08461362072919056, + "grad_norm": 0.8255814220163399, + "learning_rate": 0.0009922092034297006, + "loss": 2.9033, + "step": 369 + }, + { + "epoch": 0.08484292593441871, + "grad_norm": 0.7997437977447099, + "learning_rate": 0.0009921437693577503, + "loss": 2.8907, + "step": 370 + }, + { + "epoch": 0.08507223113964688, + "grad_norm": 0.9020599899237935, + "learning_rate": 0.0009920780638225891, + "loss": 2.8297, + "step": 371 + }, + { + "epoch": 0.08530153634487503, + "grad_norm": 0.8699762594984202, + "learning_rate": 0.0009920120868604594, + "loss": 2.7341, + "step": 372 + }, + { + "epoch": 0.08553084155010318, + "grad_norm": 0.8273464119312244, + "learning_rate": 0.0009919458385077538, + "loss": 2.9607, + "step": 373 + }, + { + "epoch": 0.08576014675533135, + "grad_norm": 0.7839916253836987, + "learning_rate": 0.0009918793188010146, + "loss": 2.8113, + "step": 374 + }, + { + "epoch": 0.0859894519605595, + "grad_norm": 0.8716165812950134, + "learning_rate": 0.0009918125277769336, + "loss": 2.8665, + "step": 375 + }, + { + "epoch": 0.08621875716578767, + "grad_norm": 0.8263188521699705, + "learning_rate": 0.0009917454654723523, + "loss": 2.8533, + "step": 376 + }, + { + "epoch": 0.08644806237101582, + "grad_norm": 0.9119588060637683, + "learning_rate": 0.0009916781319242614, + "loss": 3.019, + "step": 377 + }, + { + "epoch": 0.08667736757624397, + "grad_norm": 0.9229517736194497, + "learning_rate": 0.0009916105271698021, + "loss": 3.0087, + "step": 378 + }, + { + "epoch": 0.08690667278147214, + "grad_norm": 0.8258634156965033, + "learning_rate": 0.0009915426512462646, + "loss": 2.8105, + "step": 379 + }, + { + "epoch": 0.0871359779867003, + "grad_norm": 0.7754371646048179, + "learning_rate": 0.0009914745041910888, + "loss": 2.8897, + "step": 380 + }, + { + "epoch": 0.08736528319192846, + "grad_norm": 0.8677124233734469, + "learning_rate": 0.0009914060860418644, + "loss": 2.7721, + "step": 381 + }, + { + "epoch": 0.08759458839715661, + "grad_norm": 0.752217493160148, + "learning_rate": 0.0009913373968363298, + "loss": 2.8279, + "step": 382 + }, + { + "epoch": 0.08782389360238478, + "grad_norm": 0.812901871321775, + "learning_rate": 0.000991268436612374, + "loss": 2.8011, + "step": 383 + }, + { + "epoch": 0.08805319880761293, + "grad_norm": 0.8149240270666838, + "learning_rate": 0.0009911992054080347, + "loss": 2.738, + "step": 384 + }, + { + "epoch": 0.08828250401284109, + "grad_norm": 0.8090818592648013, + "learning_rate": 0.0009911297032614997, + "loss": 2.8366, + "step": 385 + }, + { + "epoch": 0.08851180921806925, + "grad_norm": 0.8714808383919462, + "learning_rate": 0.0009910599302111057, + "loss": 2.8506, + "step": 386 + }, + { + "epoch": 0.08874111442329741, + "grad_norm": 0.8912332072376428, + "learning_rate": 0.000990989886295339, + "loss": 3.0616, + "step": 387 + }, + { + "epoch": 0.08897041962852557, + "grad_norm": 0.9591608403475229, + "learning_rate": 0.0009909195715528355, + "loss": 2.8223, + "step": 388 + }, + { + "epoch": 0.08919972483375373, + "grad_norm": 0.9448477150145642, + "learning_rate": 0.0009908489860223804, + "loss": 2.9351, + "step": 389 + }, + { + "epoch": 0.08942903003898188, + "grad_norm": 0.8827376195736059, + "learning_rate": 0.000990778129742908, + "loss": 2.7375, + "step": 390 + }, + { + "epoch": 0.08965833524421005, + "grad_norm": 0.8498179668712234, + "learning_rate": 0.000990707002753502, + "loss": 2.9516, + "step": 391 + }, + { + "epoch": 0.0898876404494382, + "grad_norm": 0.8933116935584072, + "learning_rate": 0.000990635605093396, + "loss": 2.7459, + "step": 392 + }, + { + "epoch": 0.09011694565466637, + "grad_norm": 0.8272555896673921, + "learning_rate": 0.0009905639368019724, + "loss": 2.7434, + "step": 393 + }, + { + "epoch": 0.09034625085989452, + "grad_norm": 0.7739504853964513, + "learning_rate": 0.0009904919979187628, + "loss": 2.8446, + "step": 394 + }, + { + "epoch": 0.09057555606512267, + "grad_norm": 1.9440957490406832, + "learning_rate": 0.0009904197884834482, + "loss": 2.7257, + "step": 395 + }, + { + "epoch": 0.09080486127035084, + "grad_norm": 0.8196587654975309, + "learning_rate": 0.0009903473085358587, + "loss": 2.867, + "step": 396 + }, + { + "epoch": 0.09103416647557899, + "grad_norm": 0.8825720992141164, + "learning_rate": 0.0009902745581159742, + "loss": 2.8372, + "step": 397 + }, + { + "epoch": 0.09126347168080716, + "grad_norm": 0.9812949595279852, + "learning_rate": 0.000990201537263923, + "loss": 2.6709, + "step": 398 + }, + { + "epoch": 0.09149277688603531, + "grad_norm": 0.9790109033785863, + "learning_rate": 0.0009901282460199829, + "loss": 2.7087, + "step": 399 + }, + { + "epoch": 0.09172208209126347, + "grad_norm": 0.7991673007702061, + "learning_rate": 0.000990054684424581, + "loss": 2.978, + "step": 400 + }, + { + "epoch": 0.09195138729649163, + "grad_norm": 0.7621643484543408, + "learning_rate": 0.0009899808525182935, + "loss": 2.9154, + "step": 401 + }, + { + "epoch": 0.09218069250171979, + "grad_norm": 0.9040498276986862, + "learning_rate": 0.0009899067503418453, + "loss": 2.8052, + "step": 402 + }, + { + "epoch": 0.09240999770694795, + "grad_norm": 0.7194540675110556, + "learning_rate": 0.0009898323779361107, + "loss": 2.784, + "step": 403 + }, + { + "epoch": 0.0926393029121761, + "grad_norm": 1.3739502958199954, + "learning_rate": 0.0009897577353421131, + "loss": 2.7725, + "step": 404 + }, + { + "epoch": 0.09286860811740426, + "grad_norm": 1.697519128025193, + "learning_rate": 0.000989682822601025, + "loss": 2.9074, + "step": 405 + }, + { + "epoch": 0.09309791332263243, + "grad_norm": 0.7253517897799665, + "learning_rate": 0.0009896076397541676, + "loss": 2.6869, + "step": 406 + }, + { + "epoch": 0.09332721852786058, + "grad_norm": 0.7733207223981025, + "learning_rate": 0.0009895321868430113, + "loss": 2.7048, + "step": 407 + }, + { + "epoch": 0.09355652373308875, + "grad_norm": 0.7761318659890364, + "learning_rate": 0.0009894564639091755, + "loss": 2.7588, + "step": 408 + }, + { + "epoch": 0.0937858289383169, + "grad_norm": 0.9497811180430108, + "learning_rate": 0.0009893804709944281, + "loss": 2.7994, + "step": 409 + }, + { + "epoch": 0.09401513414354506, + "grad_norm": 0.78220230286778, + "learning_rate": 0.0009893042081406866, + "loss": 2.931, + "step": 410 + }, + { + "epoch": 0.09424443934877322, + "grad_norm": 0.7388621942904008, + "learning_rate": 0.0009892276753900174, + "loss": 2.8168, + "step": 411 + }, + { + "epoch": 0.09447374455400137, + "grad_norm": 0.722602720171425, + "learning_rate": 0.0009891508727846348, + "loss": 2.789, + "step": 412 + }, + { + "epoch": 0.09470304975922954, + "grad_norm": 0.7685793595266897, + "learning_rate": 0.0009890738003669028, + "loss": 2.7602, + "step": 413 + }, + { + "epoch": 0.09493235496445769, + "grad_norm": 0.8004248191732187, + "learning_rate": 0.0009889964581793343, + "loss": 2.7717, + "step": 414 + }, + { + "epoch": 0.09516166016968586, + "grad_norm": 0.732439586058133, + "learning_rate": 0.0009889188462645904, + "loss": 2.7597, + "step": 415 + }, + { + "epoch": 0.09539096537491401, + "grad_norm": 0.7894537813446656, + "learning_rate": 0.0009888409646654818, + "loss": 2.8522, + "step": 416 + }, + { + "epoch": 0.09562027058014216, + "grad_norm": 0.7719965261163033, + "learning_rate": 0.0009887628134249667, + "loss": 2.7978, + "step": 417 + }, + { + "epoch": 0.09584957578537033, + "grad_norm": 0.7969794514355631, + "learning_rate": 0.0009886843925861536, + "loss": 2.6893, + "step": 418 + }, + { + "epoch": 0.09607888099059848, + "grad_norm": 0.7809179338841662, + "learning_rate": 0.0009886057021922983, + "loss": 2.8197, + "step": 419 + }, + { + "epoch": 0.09630818619582665, + "grad_norm": 0.7908241061193355, + "learning_rate": 0.0009885267422868062, + "loss": 2.8523, + "step": 420 + }, + { + "epoch": 0.0965374914010548, + "grad_norm": 0.755787386120347, + "learning_rate": 0.0009884475129132311, + "loss": 2.8054, + "step": 421 + }, + { + "epoch": 0.09676679660628296, + "grad_norm": 0.7111414750754166, + "learning_rate": 0.0009883680141152752, + "loss": 2.8172, + "step": 422 + }, + { + "epoch": 0.09699610181151112, + "grad_norm": 0.8728319488173084, + "learning_rate": 0.0009882882459367897, + "loss": 2.8401, + "step": 423 + }, + { + "epoch": 0.09722540701673928, + "grad_norm": 0.8458846201251554, + "learning_rate": 0.0009882082084217742, + "loss": 2.7181, + "step": 424 + }, + { + "epoch": 0.09745471222196744, + "grad_norm": 0.8085219915911656, + "learning_rate": 0.0009881279016143766, + "loss": 2.7504, + "step": 425 + }, + { + "epoch": 0.0976840174271956, + "grad_norm": 0.7382658207307717, + "learning_rate": 0.0009880473255588936, + "loss": 2.8059, + "step": 426 + }, + { + "epoch": 0.09791332263242375, + "grad_norm": 0.7959030202045049, + "learning_rate": 0.0009879664802997707, + "loss": 2.6899, + "step": 427 + }, + { + "epoch": 0.09814262783765192, + "grad_norm": 0.7834320151425577, + "learning_rate": 0.0009878853658816015, + "loss": 2.8299, + "step": 428 + }, + { + "epoch": 0.09837193304288007, + "grad_norm": 0.7719033036819419, + "learning_rate": 0.000987803982349128, + "loss": 2.8278, + "step": 429 + }, + { + "epoch": 0.09860123824810824, + "grad_norm": 0.7437018354138782, + "learning_rate": 0.0009877223297472411, + "loss": 2.7102, + "step": 430 + }, + { + "epoch": 0.09883054345333639, + "grad_norm": 0.7472088751586442, + "learning_rate": 0.0009876404081209796, + "loss": 2.7846, + "step": 431 + }, + { + "epoch": 0.09905984865856456, + "grad_norm": 0.8047920228301283, + "learning_rate": 0.000987558217515531, + "loss": 2.8273, + "step": 432 + }, + { + "epoch": 0.09928915386379271, + "grad_norm": 0.7421556074267647, + "learning_rate": 0.000987475757976231, + "loss": 2.8615, + "step": 433 + }, + { + "epoch": 0.09951845906902086, + "grad_norm": 0.7473129784194349, + "learning_rate": 0.000987393029548564, + "loss": 2.6819, + "step": 434 + }, + { + "epoch": 0.09974776427424903, + "grad_norm": 0.7213130573603368, + "learning_rate": 0.000987310032278162, + "loss": 2.8557, + "step": 435 + }, + { + "epoch": 0.09997706947947718, + "grad_norm": 0.7495479172832743, + "learning_rate": 0.0009872267662108063, + "loss": 2.8639, + "step": 436 + }, + { + "epoch": 0.10020637468470535, + "grad_norm": 0.7148391204537173, + "learning_rate": 0.0009871432313924254, + "loss": 2.8439, + "step": 437 + }, + { + "epoch": 0.1004356798899335, + "grad_norm": 0.7623188003646107, + "learning_rate": 0.000987059427869097, + "loss": 2.9049, + "step": 438 + }, + { + "epoch": 0.10066498509516165, + "grad_norm": 0.7303372151006089, + "learning_rate": 0.000986975355687046, + "loss": 2.8557, + "step": 439 + }, + { + "epoch": 0.10089429030038982, + "grad_norm": 0.7771839911919463, + "learning_rate": 0.0009868910148926465, + "loss": 2.7839, + "step": 440 + }, + { + "epoch": 0.10112359550561797, + "grad_norm": 0.7826991892954378, + "learning_rate": 0.0009868064055324204, + "loss": 2.7672, + "step": 441 + }, + { + "epoch": 0.10135290071084614, + "grad_norm": 0.8363566083642435, + "learning_rate": 0.0009867215276530372, + "loss": 2.7453, + "step": 442 + }, + { + "epoch": 0.1015822059160743, + "grad_norm": 0.7618378226605663, + "learning_rate": 0.0009866363813013153, + "loss": 2.7592, + "step": 443 + }, + { + "epoch": 0.10181151112130245, + "grad_norm": 0.8016269406523818, + "learning_rate": 0.0009865509665242208, + "loss": 2.82, + "step": 444 + }, + { + "epoch": 0.10204081632653061, + "grad_norm": 0.7618305455537179, + "learning_rate": 0.0009864652833688676, + "loss": 2.8814, + "step": 445 + }, + { + "epoch": 0.10227012153175877, + "grad_norm": 0.7301839006668606, + "learning_rate": 0.0009863793318825185, + "loss": 2.7119, + "step": 446 + }, + { + "epoch": 0.10249942673698693, + "grad_norm": 0.750330346098155, + "learning_rate": 0.0009862931121125836, + "loss": 2.8623, + "step": 447 + }, + { + "epoch": 0.10272873194221509, + "grad_norm": 0.7069661183929277, + "learning_rate": 0.000986206624106621, + "loss": 2.8073, + "step": 448 + }, + { + "epoch": 0.10295803714744324, + "grad_norm": 0.6869090668049882, + "learning_rate": 0.000986119867912337, + "loss": 2.7406, + "step": 449 + }, + { + "epoch": 0.10318734235267141, + "grad_norm": 0.7188620201465696, + "learning_rate": 0.0009860328435775856, + "loss": 2.6793, + "step": 450 + }, + { + "epoch": 0.10341664755789956, + "grad_norm": 0.7406178480243968, + "learning_rate": 0.000985945551150369, + "loss": 2.698, + "step": 451 + }, + { + "epoch": 0.10364595276312773, + "grad_norm": 0.587499177958355, + "learning_rate": 0.000985857990678837, + "loss": 2.6585, + "step": 452 + }, + { + "epoch": 0.10387525796835588, + "grad_norm": 0.7867316512350357, + "learning_rate": 0.0009857701622112876, + "loss": 2.9362, + "step": 453 + }, + { + "epoch": 0.10410456317358405, + "grad_norm": 0.7664148500037959, + "learning_rate": 0.0009856820657961661, + "loss": 2.768, + "step": 454 + }, + { + "epoch": 0.1043338683788122, + "grad_norm": 0.6544748164155397, + "learning_rate": 0.000985593701482066, + "loss": 2.7518, + "step": 455 + }, + { + "epoch": 0.10456317358404035, + "grad_norm": 0.7491645305353452, + "learning_rate": 0.0009855050693177284, + "loss": 2.7479, + "step": 456 + }, + { + "epoch": 0.10479247878926852, + "grad_norm": 0.673863276584435, + "learning_rate": 0.0009854161693520424, + "loss": 2.7725, + "step": 457 + }, + { + "epoch": 0.10502178399449667, + "grad_norm": 0.7634810891432753, + "learning_rate": 0.0009853270016340447, + "loss": 2.8976, + "step": 458 + }, + { + "epoch": 0.10525108919972484, + "grad_norm": 0.6598027601115576, + "learning_rate": 0.0009852375662129194, + "loss": 2.7388, + "step": 459 + }, + { + "epoch": 0.10548039440495299, + "grad_norm": 0.7632534006211183, + "learning_rate": 0.0009851478631379982, + "loss": 2.722, + "step": 460 + }, + { + "epoch": 0.10570969961018115, + "grad_norm": 0.7107735959233006, + "learning_rate": 0.0009850578924587613, + "loss": 2.928, + "step": 461 + }, + { + "epoch": 0.10593900481540931, + "grad_norm": 0.6368400984051914, + "learning_rate": 0.0009849676542248356, + "loss": 2.7491, + "step": 462 + }, + { + "epoch": 0.10616831002063747, + "grad_norm": 0.6964242575368829, + "learning_rate": 0.000984877148485996, + "loss": 2.7749, + "step": 463 + }, + { + "epoch": 0.10639761522586563, + "grad_norm": 0.7508930643660485, + "learning_rate": 0.0009847863752921648, + "loss": 2.7718, + "step": 464 + }, + { + "epoch": 0.10662692043109379, + "grad_norm": 0.6859263654286484, + "learning_rate": 0.000984695334693412, + "loss": 2.6189, + "step": 465 + }, + { + "epoch": 0.10685622563632194, + "grad_norm": 0.8033621269050825, + "learning_rate": 0.0009846040267399547, + "loss": 2.7966, + "step": 466 + }, + { + "epoch": 0.1070855308415501, + "grad_norm": 0.7807945353774183, + "learning_rate": 0.000984512451482158, + "loss": 2.8824, + "step": 467 + }, + { + "epoch": 0.10731483604677826, + "grad_norm": 0.7353635974657357, + "learning_rate": 0.0009844206089705343, + "loss": 2.7413, + "step": 468 + }, + { + "epoch": 0.10754414125200643, + "grad_norm": 0.7568497320076317, + "learning_rate": 0.0009843284992557431, + "loss": 2.7294, + "step": 469 + }, + { + "epoch": 0.10777344645723458, + "grad_norm": 0.6808167530969954, + "learning_rate": 0.0009842361223885915, + "loss": 2.7095, + "step": 470 + }, + { + "epoch": 0.10800275166246273, + "grad_norm": 0.6160979814934168, + "learning_rate": 0.000984143478420034, + "loss": 2.7514, + "step": 471 + }, + { + "epoch": 0.1082320568676909, + "grad_norm": 0.6465866649318431, + "learning_rate": 0.0009840505674011727, + "loss": 2.6936, + "step": 472 + }, + { + "epoch": 0.10846136207291905, + "grad_norm": 0.6081227781917411, + "learning_rate": 0.0009839573893832563, + "loss": 2.8015, + "step": 473 + }, + { + "epoch": 0.10869066727814722, + "grad_norm": 0.6680632230237991, + "learning_rate": 0.0009838639444176812, + "loss": 2.7059, + "step": 474 + }, + { + "epoch": 0.10891997248337537, + "grad_norm": 0.7094933944881141, + "learning_rate": 0.000983770232555991, + "loss": 2.6308, + "step": 475 + }, + { + "epoch": 0.10914927768860354, + "grad_norm": 0.7039831162610235, + "learning_rate": 0.0009836762538498768, + "loss": 2.9155, + "step": 476 + }, + { + "epoch": 0.10937858289383169, + "grad_norm": 0.7257170553939399, + "learning_rate": 0.0009835820083511765, + "loss": 2.7542, + "step": 477 + }, + { + "epoch": 0.10960788809905984, + "grad_norm": 0.7324791637793726, + "learning_rate": 0.0009834874961118754, + "loss": 2.7749, + "step": 478 + }, + { + "epoch": 0.10983719330428801, + "grad_norm": 0.6938562178647415, + "learning_rate": 0.0009833927171841055, + "loss": 2.7306, + "step": 479 + }, + { + "epoch": 0.11006649850951616, + "grad_norm": 0.6824148967185918, + "learning_rate": 0.0009832976716201466, + "loss": 2.6903, + "step": 480 + }, + { + "epoch": 0.11029580371474433, + "grad_norm": 0.6936566497463247, + "learning_rate": 0.0009832023594724246, + "loss": 2.7067, + "step": 481 + }, + { + "epoch": 0.11052510891997248, + "grad_norm": 0.6948901968845296, + "learning_rate": 0.0009831067807935137, + "loss": 2.6356, + "step": 482 + }, + { + "epoch": 0.11075441412520064, + "grad_norm": 0.7089095109711848, + "learning_rate": 0.0009830109356361344, + "loss": 2.6875, + "step": 483 + }, + { + "epoch": 0.1109837193304288, + "grad_norm": 0.7196810863623381, + "learning_rate": 0.0009829148240531537, + "loss": 2.7687, + "step": 484 + }, + { + "epoch": 0.11121302453565696, + "grad_norm": 0.7004433239078077, + "learning_rate": 0.0009828184460975867, + "loss": 2.8476, + "step": 485 + }, + { + "epoch": 0.11144232974088512, + "grad_norm": 0.6479432593261939, + "learning_rate": 0.0009827218018225944, + "loss": 2.6644, + "step": 486 + }, + { + "epoch": 0.11167163494611328, + "grad_norm": 0.7136649314527803, + "learning_rate": 0.0009826248912814855, + "loss": 2.7719, + "step": 487 + }, + { + "epoch": 0.11190094015134143, + "grad_norm": 0.6934638480541544, + "learning_rate": 0.0009825277145277147, + "loss": 2.6695, + "step": 488 + }, + { + "epoch": 0.1121302453565696, + "grad_norm": 0.669807491547042, + "learning_rate": 0.0009824302716148847, + "loss": 2.8512, + "step": 489 + }, + { + "epoch": 0.11235955056179775, + "grad_norm": 0.6771108909846232, + "learning_rate": 0.0009823325625967442, + "loss": 2.7739, + "step": 490 + }, + { + "epoch": 0.11258885576702592, + "grad_norm": 0.6593917196855329, + "learning_rate": 0.0009822345875271884, + "loss": 2.629, + "step": 491 + }, + { + "epoch": 0.11281816097225407, + "grad_norm": 0.6308851082254566, + "learning_rate": 0.00098213634646026, + "loss": 2.7679, + "step": 492 + }, + { + "epoch": 0.11304746617748222, + "grad_norm": 0.801445975062801, + "learning_rate": 0.0009820378394501481, + "loss": 2.8072, + "step": 493 + }, + { + "epoch": 0.11327677138271039, + "grad_norm": 0.7371724403660551, + "learning_rate": 0.0009819390665511884, + "loss": 2.7641, + "step": 494 + }, + { + "epoch": 0.11350607658793854, + "grad_norm": 0.767569840085392, + "learning_rate": 0.0009818400278178636, + "loss": 2.7971, + "step": 495 + }, + { + "epoch": 0.11373538179316671, + "grad_norm": 0.7499921213469136, + "learning_rate": 0.0009817407233048028, + "loss": 2.7758, + "step": 496 + }, + { + "epoch": 0.11396468699839486, + "grad_norm": 0.7384017422351868, + "learning_rate": 0.0009816411530667814, + "loss": 2.8142, + "step": 497 + }, + { + "epoch": 0.11419399220362302, + "grad_norm": 0.6919941821059186, + "learning_rate": 0.0009815413171587219, + "loss": 2.7599, + "step": 498 + }, + { + "epoch": 0.11442329740885118, + "grad_norm": 0.6611516509348537, + "learning_rate": 0.000981441215635693, + "loss": 2.6207, + "step": 499 + }, + { + "epoch": 0.11465260261407934, + "grad_norm": 0.6953949124600061, + "learning_rate": 0.0009813408485529103, + "loss": 2.6663, + "step": 500 + }, + { + "epoch": 0.1148819078193075, + "grad_norm": 0.6291675610794889, + "learning_rate": 0.0009812402159657353, + "loss": 2.6502, + "step": 501 + }, + { + "epoch": 0.11511121302453566, + "grad_norm": 0.6951266190345152, + "learning_rate": 0.000981139317929676, + "loss": 2.6274, + "step": 502 + }, + { + "epoch": 0.11534051822976382, + "grad_norm": 0.6197744768413845, + "learning_rate": 0.000981038154500388, + "loss": 2.7559, + "step": 503 + }, + { + "epoch": 0.11556982343499198, + "grad_norm": 0.652435050420378, + "learning_rate": 0.0009809367257336717, + "loss": 2.5876, + "step": 504 + }, + { + "epoch": 0.11579912864022013, + "grad_norm": 0.7324016557107166, + "learning_rate": 0.0009808350316854746, + "loss": 2.7548, + "step": 505 + }, + { + "epoch": 0.1160284338454483, + "grad_norm": 0.6602566678675983, + "learning_rate": 0.0009807330724118904, + "loss": 2.6352, + "step": 506 + }, + { + "epoch": 0.11625773905067645, + "grad_norm": 0.6933786025165083, + "learning_rate": 0.0009806308479691594, + "loss": 2.7354, + "step": 507 + }, + { + "epoch": 0.11648704425590461, + "grad_norm": 0.7131051843492375, + "learning_rate": 0.000980528358413668, + "loss": 2.6334, + "step": 508 + }, + { + "epoch": 0.11671634946113277, + "grad_norm": 0.6870518053284176, + "learning_rate": 0.0009804256038019482, + "loss": 2.7694, + "step": 509 + }, + { + "epoch": 0.11694565466636092, + "grad_norm": 0.7515222842640336, + "learning_rate": 0.0009803225841906792, + "loss": 2.7591, + "step": 510 + }, + { + "epoch": 0.11717495987158909, + "grad_norm": 0.7140231128425132, + "learning_rate": 0.0009802192996366857, + "loss": 2.7241, + "step": 511 + }, + { + "epoch": 0.11740426507681724, + "grad_norm": 0.6883991526105611, + "learning_rate": 0.000980115750196939, + "loss": 2.6909, + "step": 512 + }, + { + "epoch": 0.11763357028204541, + "grad_norm": 0.6772701423556428, + "learning_rate": 0.0009800119359285563, + "loss": 2.6568, + "step": 513 + }, + { + "epoch": 0.11786287548727356, + "grad_norm": 0.6269267349696597, + "learning_rate": 0.0009799078568888007, + "loss": 2.8011, + "step": 514 + }, + { + "epoch": 0.11809218069250171, + "grad_norm": 0.6655753734458618, + "learning_rate": 0.0009798035131350813, + "loss": 2.6531, + "step": 515 + }, + { + "epoch": 0.11832148589772988, + "grad_norm": 0.5944395419394383, + "learning_rate": 0.0009796989047249538, + "loss": 2.6979, + "step": 516 + }, + { + "epoch": 0.11855079110295803, + "grad_norm": 0.6506160689943615, + "learning_rate": 0.0009795940317161194, + "loss": 2.7876, + "step": 517 + }, + { + "epoch": 0.1187800963081862, + "grad_norm": 0.6463566562436602, + "learning_rate": 0.0009794888941664253, + "loss": 2.7341, + "step": 518 + }, + { + "epoch": 0.11900940151341435, + "grad_norm": 0.6876761531394692, + "learning_rate": 0.0009793834921338646, + "loss": 2.6636, + "step": 519 + }, + { + "epoch": 0.1192387067186425, + "grad_norm": 0.6868792766004788, + "learning_rate": 0.0009792778256765763, + "loss": 2.6194, + "step": 520 + }, + { + "epoch": 0.11946801192387067, + "grad_norm": 0.6833202183156338, + "learning_rate": 0.0009791718948528457, + "loss": 2.7384, + "step": 521 + }, + { + "epoch": 0.11969731712909883, + "grad_norm": 0.6052524257915569, + "learning_rate": 0.0009790656997211032, + "loss": 2.7991, + "step": 522 + }, + { + "epoch": 0.119926622334327, + "grad_norm": 0.7192917969863138, + "learning_rate": 0.0009789592403399252, + "loss": 2.6414, + "step": 523 + }, + { + "epoch": 0.12015592753955515, + "grad_norm": 0.7676649920487703, + "learning_rate": 0.0009788525167680345, + "loss": 2.7516, + "step": 524 + }, + { + "epoch": 0.12038523274478331, + "grad_norm": 0.6155914460785528, + "learning_rate": 0.0009787455290642985, + "loss": 2.738, + "step": 525 + }, + { + "epoch": 0.12061453795001147, + "grad_norm": 0.70792291345927, + "learning_rate": 0.0009786382772877312, + "loss": 2.6681, + "step": 526 + }, + { + "epoch": 0.12084384315523962, + "grad_norm": 0.7003559406480586, + "learning_rate": 0.000978530761497492, + "loss": 2.9823, + "step": 527 + }, + { + "epoch": 0.12107314836046779, + "grad_norm": 0.7247681130160939, + "learning_rate": 0.0009784229817528859, + "loss": 2.7487, + "step": 528 + }, + { + "epoch": 0.12130245356569594, + "grad_norm": 0.7549613986938164, + "learning_rate": 0.0009783149381133633, + "loss": 2.7271, + "step": 529 + }, + { + "epoch": 0.1215317587709241, + "grad_norm": 0.6484877269505425, + "learning_rate": 0.0009782066306385207, + "loss": 2.7607, + "step": 530 + }, + { + "epoch": 0.12176106397615226, + "grad_norm": 0.6701921173372634, + "learning_rate": 0.0009780980593880992, + "loss": 2.7142, + "step": 531 + }, + { + "epoch": 0.12199036918138041, + "grad_norm": 0.678565873052309, + "learning_rate": 0.0009779892244219865, + "loss": 2.7357, + "step": 532 + }, + { + "epoch": 0.12221967438660858, + "grad_norm": 0.6824800176741677, + "learning_rate": 0.0009778801258002153, + "loss": 2.5991, + "step": 533 + }, + { + "epoch": 0.12244897959183673, + "grad_norm": 0.7473270547309727, + "learning_rate": 0.0009777707635829631, + "loss": 2.82, + "step": 534 + }, + { + "epoch": 0.1226782847970649, + "grad_norm": 0.6334021900501278, + "learning_rate": 0.000977661137830554, + "loss": 2.6827, + "step": 535 + }, + { + "epoch": 0.12290759000229305, + "grad_norm": 0.6930374932754324, + "learning_rate": 0.0009775512486034563, + "loss": 2.766, + "step": 536 + }, + { + "epoch": 0.1231368952075212, + "grad_norm": 0.694379345935452, + "learning_rate": 0.0009774410959622845, + "loss": 2.6137, + "step": 537 + }, + { + "epoch": 0.12336620041274937, + "grad_norm": 0.7236256012949815, + "learning_rate": 0.0009773306799677978, + "loss": 2.7149, + "step": 538 + }, + { + "epoch": 0.12359550561797752, + "grad_norm": 0.6984376130796863, + "learning_rate": 0.000977220000680901, + "loss": 2.6056, + "step": 539 + }, + { + "epoch": 0.12382481082320569, + "grad_norm": 0.6693461537234031, + "learning_rate": 0.0009771090581626445, + "loss": 2.6417, + "step": 540 + }, + { + "epoch": 0.12405411602843384, + "grad_norm": 0.6716844853444026, + "learning_rate": 0.000976997852474223, + "loss": 2.5949, + "step": 541 + }, + { + "epoch": 0.124283421233662, + "grad_norm": 0.6130823330590883, + "learning_rate": 0.0009768863836769769, + "loss": 2.6423, + "step": 542 + }, + { + "epoch": 0.12451272643889016, + "grad_norm": 0.8043723029616219, + "learning_rate": 0.0009767746518323914, + "loss": 2.6557, + "step": 543 + }, + { + "epoch": 0.12474203164411832, + "grad_norm": 0.6729989448054221, + "learning_rate": 0.0009766626570020974, + "loss": 2.8096, + "step": 544 + }, + { + "epoch": 0.12497133684934648, + "grad_norm": 0.6394073581857131, + "learning_rate": 0.0009765503992478704, + "loss": 2.7676, + "step": 545 + }, + { + "epoch": 0.12520064205457465, + "grad_norm": 0.6782248526257117, + "learning_rate": 0.000976437878631631, + "loss": 2.6452, + "step": 546 + }, + { + "epoch": 0.1254299472598028, + "grad_norm": 0.6996135454530884, + "learning_rate": 0.0009763250952154449, + "loss": 2.7665, + "step": 547 + }, + { + "epoch": 0.12565925246503096, + "grad_norm": 0.6632957025926736, + "learning_rate": 0.0009762120490615225, + "loss": 2.7448, + "step": 548 + }, + { + "epoch": 0.1258885576702591, + "grad_norm": 0.6319746203399706, + "learning_rate": 0.0009760987402322195, + "loss": 2.5745, + "step": 549 + }, + { + "epoch": 0.12611786287548726, + "grad_norm": 0.5304779937529719, + "learning_rate": 0.000975985168790036, + "loss": 2.7035, + "step": 550 + }, + { + "epoch": 0.12634716808071544, + "grad_norm": 0.6053523352238981, + "learning_rate": 0.0009758713347976178, + "loss": 2.7953, + "step": 551 + }, + { + "epoch": 0.1265764732859436, + "grad_norm": 0.6552841140263906, + "learning_rate": 0.0009757572383177545, + "loss": 2.5658, + "step": 552 + }, + { + "epoch": 0.12680577849117175, + "grad_norm": 0.6681378204267486, + "learning_rate": 0.000975642879413381, + "loss": 2.7313, + "step": 553 + }, + { + "epoch": 0.1270350836963999, + "grad_norm": 0.6267904031745176, + "learning_rate": 0.0009755282581475768, + "loss": 2.7629, + "step": 554 + }, + { + "epoch": 0.12726438890162806, + "grad_norm": 0.6331980253128872, + "learning_rate": 0.0009754133745835665, + "loss": 2.6122, + "step": 555 + }, + { + "epoch": 0.12749369410685624, + "grad_norm": 0.700999622367925, + "learning_rate": 0.0009752982287847192, + "loss": 2.8405, + "step": 556 + }, + { + "epoch": 0.1277229993120844, + "grad_norm": 0.7160228974589439, + "learning_rate": 0.0009751828208145482, + "loss": 2.713, + "step": 557 + }, + { + "epoch": 0.12795230451731254, + "grad_norm": 0.6789430622143611, + "learning_rate": 0.0009750671507367119, + "loss": 2.674, + "step": 558 + }, + { + "epoch": 0.1281816097225407, + "grad_norm": 0.6125509810929403, + "learning_rate": 0.0009749512186150131, + "loss": 2.7448, + "step": 559 + }, + { + "epoch": 0.12841091492776885, + "grad_norm": 0.6900677386177575, + "learning_rate": 0.0009748350245133994, + "loss": 2.733, + "step": 560 + }, + { + "epoch": 0.12864022013299703, + "grad_norm": 0.6249071621698773, + "learning_rate": 0.0009747185684959625, + "loss": 2.7298, + "step": 561 + }, + { + "epoch": 0.12886952533822518, + "grad_norm": 0.6230230765925919, + "learning_rate": 0.0009746018506269387, + "loss": 2.6519, + "step": 562 + }, + { + "epoch": 0.12909883054345334, + "grad_norm": 0.6035855170514268, + "learning_rate": 0.000974484870970709, + "loss": 2.6397, + "step": 563 + }, + { + "epoch": 0.1293281357486815, + "grad_norm": 0.5907711827478068, + "learning_rate": 0.0009743676295917985, + "loss": 2.6599, + "step": 564 + }, + { + "epoch": 0.12955744095390964, + "grad_norm": 0.6605561257233306, + "learning_rate": 0.0009742501265548767, + "loss": 2.6592, + "step": 565 + }, + { + "epoch": 0.12978674615913782, + "grad_norm": 0.6349898715789875, + "learning_rate": 0.0009741323619247575, + "loss": 2.6606, + "step": 566 + }, + { + "epoch": 0.13001605136436598, + "grad_norm": 0.6779980138308443, + "learning_rate": 0.0009740143357663993, + "loss": 2.5652, + "step": 567 + }, + { + "epoch": 0.13024535656959413, + "grad_norm": 0.636032413183963, + "learning_rate": 0.0009738960481449042, + "loss": 2.7024, + "step": 568 + }, + { + "epoch": 0.13047466177482228, + "grad_norm": 0.5991674225823588, + "learning_rate": 0.000973777499125519, + "loss": 2.6747, + "step": 569 + }, + { + "epoch": 0.13070396698005043, + "grad_norm": 0.6546694561038477, + "learning_rate": 0.0009736586887736346, + "loss": 2.6718, + "step": 570 + }, + { + "epoch": 0.13093327218527862, + "grad_norm": 0.6018149358056906, + "learning_rate": 0.0009735396171547859, + "loss": 2.6957, + "step": 571 + }, + { + "epoch": 0.13116257739050677, + "grad_norm": 0.6069532483598061, + "learning_rate": 0.0009734202843346522, + "loss": 2.7095, + "step": 572 + }, + { + "epoch": 0.13139188259573492, + "grad_norm": 0.6126007373733499, + "learning_rate": 0.0009733006903790564, + "loss": 2.7771, + "step": 573 + }, + { + "epoch": 0.13162118780096307, + "grad_norm": 0.6082657008941991, + "learning_rate": 0.0009731808353539659, + "loss": 2.7277, + "step": 574 + }, + { + "epoch": 0.13185049300619123, + "grad_norm": 0.5989831836472285, + "learning_rate": 0.0009730607193254922, + "loss": 2.8076, + "step": 575 + }, + { + "epoch": 0.1320797982114194, + "grad_norm": 0.6489308878916809, + "learning_rate": 0.0009729403423598899, + "loss": 2.7944, + "step": 576 + }, + { + "epoch": 0.13230910341664756, + "grad_norm": 0.6629988740513078, + "learning_rate": 0.0009728197045235585, + "loss": 2.5402, + "step": 577 + }, + { + "epoch": 0.13253840862187571, + "grad_norm": 0.6257727503732321, + "learning_rate": 0.0009726988058830409, + "loss": 2.7351, + "step": 578 + }, + { + "epoch": 0.13276771382710387, + "grad_norm": 0.6401228810988389, + "learning_rate": 0.0009725776465050242, + "loss": 2.7293, + "step": 579 + }, + { + "epoch": 0.13299701903233205, + "grad_norm": 0.6564615097625293, + "learning_rate": 0.0009724562264563388, + "loss": 2.6339, + "step": 580 + }, + { + "epoch": 0.1332263242375602, + "grad_norm": 0.678440627987337, + "learning_rate": 0.0009723345458039594, + "loss": 2.5594, + "step": 581 + }, + { + "epoch": 0.13345562944278835, + "grad_norm": 0.70146621247164, + "learning_rate": 0.0009722126046150042, + "loss": 2.8054, + "step": 582 + }, + { + "epoch": 0.1336849346480165, + "grad_norm": 0.665161710982368, + "learning_rate": 0.000972090402956735, + "loss": 2.6068, + "step": 583 + }, + { + "epoch": 0.13391423985324466, + "grad_norm": 0.7273197422383503, + "learning_rate": 0.0009719679408965576, + "loss": 2.7203, + "step": 584 + }, + { + "epoch": 0.13414354505847284, + "grad_norm": 0.7017015043550033, + "learning_rate": 0.0009718452185020212, + "loss": 2.6685, + "step": 585 + }, + { + "epoch": 0.134372850263701, + "grad_norm": 0.5976923093397456, + "learning_rate": 0.0009717222358408186, + "loss": 2.608, + "step": 586 + }, + { + "epoch": 0.13460215546892915, + "grad_norm": 0.6046118345971886, + "learning_rate": 0.0009715989929807862, + "loss": 2.765, + "step": 587 + }, + { + "epoch": 0.1348314606741573, + "grad_norm": 0.6775771599615457, + "learning_rate": 0.0009714754899899041, + "loss": 2.547, + "step": 588 + }, + { + "epoch": 0.13506076587938545, + "grad_norm": 0.5935351117071331, + "learning_rate": 0.0009713517269362955, + "loss": 2.6254, + "step": 589 + }, + { + "epoch": 0.13529007108461363, + "grad_norm": 0.5878137333004145, + "learning_rate": 0.0009712277038882273, + "loss": 2.6681, + "step": 590 + }, + { + "epoch": 0.1355193762898418, + "grad_norm": 0.6343083734576159, + "learning_rate": 0.0009711034209141101, + "loss": 2.6178, + "step": 591 + }, + { + "epoch": 0.13574868149506994, + "grad_norm": 0.6554398027286941, + "learning_rate": 0.0009709788780824971, + "loss": 2.7209, + "step": 592 + }, + { + "epoch": 0.1359779867002981, + "grad_norm": 0.6891517620033676, + "learning_rate": 0.0009708540754620856, + "loss": 2.7463, + "step": 593 + }, + { + "epoch": 0.13620729190552625, + "grad_norm": 1.7034752649824192, + "learning_rate": 0.0009707290131217159, + "loss": 2.6862, + "step": 594 + }, + { + "epoch": 0.13643659711075443, + "grad_norm": 0.5901079112520189, + "learning_rate": 0.0009706036911303713, + "loss": 2.6749, + "step": 595 + }, + { + "epoch": 0.13666590231598258, + "grad_norm": 0.5837179951361651, + "learning_rate": 0.0009704781095571787, + "loss": 2.8239, + "step": 596 + }, + { + "epoch": 0.13689520752121073, + "grad_norm": 0.6317688273345596, + "learning_rate": 0.0009703522684714083, + "loss": 2.6813, + "step": 597 + }, + { + "epoch": 0.13712451272643889, + "grad_norm": 0.6001484853192836, + "learning_rate": 0.0009702261679424728, + "loss": 2.7215, + "step": 598 + }, + { + "epoch": 0.13735381793166704, + "grad_norm": 0.6518499199520018, + "learning_rate": 0.0009700998080399286, + "loss": 2.5446, + "step": 599 + }, + { + "epoch": 0.13758312313689522, + "grad_norm": 0.6029281372509081, + "learning_rate": 0.000969973188833475, + "loss": 2.7374, + "step": 600 + }, + { + "epoch": 0.13781242834212337, + "grad_norm": 0.5642590137506862, + "learning_rate": 0.0009698463103929542, + "loss": 2.5242, + "step": 601 + }, + { + "epoch": 0.13804173354735153, + "grad_norm": 0.7380067365673283, + "learning_rate": 0.0009697191727883516, + "loss": 2.7038, + "step": 602 + }, + { + "epoch": 0.13827103875257968, + "grad_norm": 0.6204010152898631, + "learning_rate": 0.0009695917760897954, + "loss": 2.5391, + "step": 603 + }, + { + "epoch": 0.13850034395780783, + "grad_norm": 0.6033978284249149, + "learning_rate": 0.0009694641203675569, + "loss": 2.8181, + "step": 604 + }, + { + "epoch": 0.138729649163036, + "grad_norm": 0.5805864543911133, + "learning_rate": 0.0009693362056920501, + "loss": 2.5163, + "step": 605 + }, + { + "epoch": 0.13895895436826416, + "grad_norm": 0.6281844600648565, + "learning_rate": 0.0009692080321338317, + "loss": 2.6995, + "step": 606 + }, + { + "epoch": 0.13918825957349232, + "grad_norm": 0.6412086853441177, + "learning_rate": 0.0009690795997636015, + "loss": 2.617, + "step": 607 + }, + { + "epoch": 0.13941756477872047, + "grad_norm": 0.6325907988053894, + "learning_rate": 0.0009689509086522018, + "loss": 2.7036, + "step": 608 + }, + { + "epoch": 0.13964686998394862, + "grad_norm": 0.6426067433650404, + "learning_rate": 0.0009688219588706179, + "loss": 2.6415, + "step": 609 + }, + { + "epoch": 0.1398761751891768, + "grad_norm": 0.6879340254446905, + "learning_rate": 0.0009686927504899778, + "loss": 2.8106, + "step": 610 + }, + { + "epoch": 0.14010548039440496, + "grad_norm": 0.6489938508016905, + "learning_rate": 0.0009685632835815518, + "loss": 2.6238, + "step": 611 + }, + { + "epoch": 0.1403347855996331, + "grad_norm": 0.5960706018998392, + "learning_rate": 0.0009684335582167529, + "loss": 2.6182, + "step": 612 + }, + { + "epoch": 0.14056409080486126, + "grad_norm": 0.7614060881473732, + "learning_rate": 0.0009683035744671367, + "loss": 2.6708, + "step": 613 + }, + { + "epoch": 0.14079339601008942, + "grad_norm": 0.5445056330911738, + "learning_rate": 0.0009681733324044017, + "loss": 2.7523, + "step": 614 + }, + { + "epoch": 0.1410227012153176, + "grad_norm": 0.6222406085233773, + "learning_rate": 0.0009680428321003883, + "loss": 2.7427, + "step": 615 + }, + { + "epoch": 0.14125200642054575, + "grad_norm": 0.5596539714846597, + "learning_rate": 0.0009679120736270794, + "loss": 2.6523, + "step": 616 + }, + { + "epoch": 0.1414813116257739, + "grad_norm": 0.5621370495117682, + "learning_rate": 0.000967781057056601, + "loss": 2.6313, + "step": 617 + }, + { + "epoch": 0.14171061683100206, + "grad_norm": 0.592223393893703, + "learning_rate": 0.0009676497824612206, + "loss": 2.6134, + "step": 618 + }, + { + "epoch": 0.1419399220362302, + "grad_norm": 0.5889756140393635, + "learning_rate": 0.0009675182499133485, + "loss": 2.7767, + "step": 619 + }, + { + "epoch": 0.1421692272414584, + "grad_norm": 0.5585003987213765, + "learning_rate": 0.0009673864594855372, + "loss": 2.5383, + "step": 620 + }, + { + "epoch": 0.14239853244668654, + "grad_norm": 0.6404907151548638, + "learning_rate": 0.0009672544112504813, + "loss": 2.5597, + "step": 621 + }, + { + "epoch": 0.1426278376519147, + "grad_norm": 0.6436544543956512, + "learning_rate": 0.0009671221052810178, + "loss": 2.6733, + "step": 622 + }, + { + "epoch": 0.14285714285714285, + "grad_norm": 0.6192079741512292, + "learning_rate": 0.0009669895416501257, + "loss": 2.5517, + "step": 623 + }, + { + "epoch": 0.143086448062371, + "grad_norm": 0.6059247303003883, + "learning_rate": 0.0009668567204309265, + "loss": 2.6347, + "step": 624 + }, + { + "epoch": 0.14331575326759918, + "grad_norm": 0.6448685877099977, + "learning_rate": 0.0009667236416966833, + "loss": 2.6846, + "step": 625 + }, + { + "epoch": 0.14354505847282734, + "grad_norm": 0.590526105583618, + "learning_rate": 0.0009665903055208013, + "loss": 2.6816, + "step": 626 + }, + { + "epoch": 0.1437743636780555, + "grad_norm": 0.6344726183883758, + "learning_rate": 0.0009664567119768281, + "loss": 2.6537, + "step": 627 + }, + { + "epoch": 0.14400366888328364, + "grad_norm": 0.6048841875322886, + "learning_rate": 0.0009663228611384527, + "loss": 2.7244, + "step": 628 + }, + { + "epoch": 0.14423297408851182, + "grad_norm": 0.6592368374122102, + "learning_rate": 0.0009661887530795067, + "loss": 2.6341, + "step": 629 + }, + { + "epoch": 0.14446227929373998, + "grad_norm": 0.6312179604413195, + "learning_rate": 0.0009660543878739629, + "loss": 2.5721, + "step": 630 + }, + { + "epoch": 0.14469158449896813, + "grad_norm": 0.5838432851593107, + "learning_rate": 0.0009659197655959365, + "loss": 2.579, + "step": 631 + }, + { + "epoch": 0.14492088970419628, + "grad_norm": 0.6626650982568616, + "learning_rate": 0.000965784886319684, + "loss": 2.515, + "step": 632 + }, + { + "epoch": 0.14515019490942443, + "grad_norm": 0.6394597946634709, + "learning_rate": 0.000965649750119604, + "loss": 2.5623, + "step": 633 + }, + { + "epoch": 0.14537950011465262, + "grad_norm": 0.6598368312916912, + "learning_rate": 0.0009655143570702369, + "loss": 2.7348, + "step": 634 + }, + { + "epoch": 0.14560880531988077, + "grad_norm": 0.6071594208714, + "learning_rate": 0.0009653787072462643, + "loss": 2.8042, + "step": 635 + }, + { + "epoch": 0.14583811052510892, + "grad_norm": 0.6203611508327429, + "learning_rate": 0.0009652428007225099, + "loss": 2.6284, + "step": 636 + }, + { + "epoch": 0.14606741573033707, + "grad_norm": 0.6934048008159673, + "learning_rate": 0.0009651066375739388, + "loss": 2.6956, + "step": 637 + }, + { + "epoch": 0.14629672093556523, + "grad_norm": 0.5759947310223896, + "learning_rate": 0.0009649702178756578, + "loss": 2.7537, + "step": 638 + }, + { + "epoch": 0.1465260261407934, + "grad_norm": 0.5948265514364672, + "learning_rate": 0.000964833541702915, + "loss": 2.6137, + "step": 639 + }, + { + "epoch": 0.14675533134602156, + "grad_norm": 0.5830973725624155, + "learning_rate": 0.0009646966091311, + "loss": 2.6428, + "step": 640 + }, + { + "epoch": 0.14698463655124971, + "grad_norm": 0.660534046147951, + "learning_rate": 0.0009645594202357438, + "loss": 2.4997, + "step": 641 + }, + { + "epoch": 0.14721394175647787, + "grad_norm": 0.588409647688694, + "learning_rate": 0.0009644219750925191, + "loss": 2.654, + "step": 642 + }, + { + "epoch": 0.14744324696170602, + "grad_norm": 0.6043405116215858, + "learning_rate": 0.0009642842737772397, + "loss": 2.5742, + "step": 643 + }, + { + "epoch": 0.1476725521669342, + "grad_norm": 0.6484862534590432, + "learning_rate": 0.0009641463163658606, + "loss": 2.6853, + "step": 644 + }, + { + "epoch": 0.14790185737216235, + "grad_norm": 0.7148634098770481, + "learning_rate": 0.0009640081029344782, + "loss": 2.6899, + "step": 645 + }, + { + "epoch": 0.1481311625773905, + "grad_norm": 0.5860005704349507, + "learning_rate": 0.0009638696335593303, + "loss": 2.6331, + "step": 646 + }, + { + "epoch": 0.14836046778261866, + "grad_norm": 0.609567351051576, + "learning_rate": 0.0009637309083167956, + "loss": 2.6405, + "step": 647 + }, + { + "epoch": 0.1485897729878468, + "grad_norm": 0.6314003942093319, + "learning_rate": 0.0009635919272833937, + "loss": 2.5536, + "step": 648 + }, + { + "epoch": 0.148819078193075, + "grad_norm": 0.6141163071191318, + "learning_rate": 0.0009634526905357859, + "loss": 2.7894, + "step": 649 + }, + { + "epoch": 0.14904838339830315, + "grad_norm": 0.6594325448258949, + "learning_rate": 0.0009633131981507744, + "loss": 2.7054, + "step": 650 + }, + { + "epoch": 0.1492776886035313, + "grad_norm": 0.557003064896759, + "learning_rate": 0.000963173450205302, + "loss": 2.5808, + "step": 651 + }, + { + "epoch": 0.14950699380875945, + "grad_norm": 0.6186758194271333, + "learning_rate": 0.0009630334467764528, + "loss": 2.7644, + "step": 652 + }, + { + "epoch": 0.1497362990139876, + "grad_norm": 0.5380611410141701, + "learning_rate": 0.0009628931879414517, + "loss": 2.6727, + "step": 653 + }, + { + "epoch": 0.1499656042192158, + "grad_norm": 0.5933085508088959, + "learning_rate": 0.0009627526737776646, + "loss": 2.6068, + "step": 654 + }, + { + "epoch": 0.15019490942444394, + "grad_norm": 0.5766532157134324, + "learning_rate": 0.0009626119043625983, + "loss": 2.6547, + "step": 655 + }, + { + "epoch": 0.1504242146296721, + "grad_norm": 0.5870428747226673, + "learning_rate": 0.0009624708797739001, + "loss": 2.7609, + "step": 656 + }, + { + "epoch": 0.15065351983490025, + "grad_norm": 0.6548073363122533, + "learning_rate": 0.0009623296000893582, + "loss": 2.7199, + "step": 657 + }, + { + "epoch": 0.1508828250401284, + "grad_norm": 0.6482518922589047, + "learning_rate": 0.0009621880653869016, + "loss": 2.5979, + "step": 658 + }, + { + "epoch": 0.15111213024535658, + "grad_norm": 0.6362148718062363, + "learning_rate": 0.0009620462757446, + "loss": 2.7116, + "step": 659 + }, + { + "epoch": 0.15134143545058473, + "grad_norm": 0.6213171346968954, + "learning_rate": 0.0009619042312406634, + "loss": 2.6524, + "step": 660 + }, + { + "epoch": 0.15157074065581289, + "grad_norm": 0.6112367783024394, + "learning_rate": 0.0009617619319534428, + "loss": 2.689, + "step": 661 + }, + { + "epoch": 0.15180004586104104, + "grad_norm": 0.6288114664960663, + "learning_rate": 0.0009616193779614293, + "loss": 2.8252, + "step": 662 + }, + { + "epoch": 0.1520293510662692, + "grad_norm": 0.5794315101771313, + "learning_rate": 0.000961476569343255, + "loss": 2.5951, + "step": 663 + }, + { + "epoch": 0.15225865627149737, + "grad_norm": 0.5753831747082726, + "learning_rate": 0.000961333506177692, + "loss": 2.7472, + "step": 664 + }, + { + "epoch": 0.15248796147672553, + "grad_norm": 0.5912617865598487, + "learning_rate": 0.0009611901885436529, + "loss": 2.6514, + "step": 665 + }, + { + "epoch": 0.15271726668195368, + "grad_norm": 0.6248466296102673, + "learning_rate": 0.0009610466165201911, + "loss": 2.8405, + "step": 666 + }, + { + "epoch": 0.15294657188718183, + "grad_norm": 0.6189250350207397, + "learning_rate": 0.0009609027901864996, + "loss": 2.6268, + "step": 667 + }, + { + "epoch": 0.15317587709240998, + "grad_norm": 0.6015562123334194, + "learning_rate": 0.0009607587096219124, + "loss": 2.6302, + "step": 668 + }, + { + "epoch": 0.15340518229763817, + "grad_norm": 0.5947161991292068, + "learning_rate": 0.0009606143749059029, + "loss": 2.6487, + "step": 669 + }, + { + "epoch": 0.15363448750286632, + "grad_norm": 0.5659480225953492, + "learning_rate": 0.0009604697861180856, + "loss": 2.6911, + "step": 670 + }, + { + "epoch": 0.15386379270809447, + "grad_norm": 0.6119350132430725, + "learning_rate": 0.0009603249433382144, + "loss": 2.5461, + "step": 671 + }, + { + "epoch": 0.15409309791332262, + "grad_norm": 0.6043189625307019, + "learning_rate": 0.0009601798466461838, + "loss": 2.646, + "step": 672 + }, + { + "epoch": 0.1543224031185508, + "grad_norm": 0.6552529333964608, + "learning_rate": 0.0009600344961220282, + "loss": 2.6908, + "step": 673 + }, + { + "epoch": 0.15455170832377896, + "grad_norm": 0.5857913418285996, + "learning_rate": 0.0009598888918459219, + "loss": 2.6616, + "step": 674 + }, + { + "epoch": 0.1547810135290071, + "grad_norm": 0.5631602296450658, + "learning_rate": 0.0009597430338981791, + "loss": 2.691, + "step": 675 + }, + { + "epoch": 0.15501031873423526, + "grad_norm": 0.5710511410203849, + "learning_rate": 0.0009595969223592542, + "loss": 2.5518, + "step": 676 + }, + { + "epoch": 0.15523962393946342, + "grad_norm": 0.6004490287071396, + "learning_rate": 0.0009594505573097414, + "loss": 2.6558, + "step": 677 + }, + { + "epoch": 0.1554689291446916, + "grad_norm": 0.6201680806095327, + "learning_rate": 0.0009593039388303747, + "loss": 2.4901, + "step": 678 + }, + { + "epoch": 0.15569823434991975, + "grad_norm": 0.6093908472182034, + "learning_rate": 0.0009591570670020277, + "loss": 2.7125, + "step": 679 + }, + { + "epoch": 0.1559275395551479, + "grad_norm": 0.6109352782477433, + "learning_rate": 0.0009590099419057142, + "loss": 2.6977, + "step": 680 + }, + { + "epoch": 0.15615684476037606, + "grad_norm": 0.6044355819553532, + "learning_rate": 0.0009588625636225871, + "loss": 2.5917, + "step": 681 + }, + { + "epoch": 0.1563861499656042, + "grad_norm": 0.5775329648987793, + "learning_rate": 0.0009587149322339394, + "loss": 2.604, + "step": 682 + }, + { + "epoch": 0.1566154551708324, + "grad_norm": 0.5699873950877192, + "learning_rate": 0.0009585670478212036, + "loss": 2.6572, + "step": 683 + }, + { + "epoch": 0.15684476037606054, + "grad_norm": 0.5420877965736458, + "learning_rate": 0.0009584189104659517, + "loss": 2.5951, + "step": 684 + }, + { + "epoch": 0.1570740655812887, + "grad_norm": 0.6020956049115337, + "learning_rate": 0.0009582705202498956, + "loss": 2.7016, + "step": 685 + }, + { + "epoch": 0.15730337078651685, + "grad_norm": 0.666618586750053, + "learning_rate": 0.0009581218772548859, + "loss": 2.5868, + "step": 686 + }, + { + "epoch": 0.157532675991745, + "grad_norm": 0.5690464555760235, + "learning_rate": 0.0009579729815629133, + "loss": 2.54, + "step": 687 + }, + { + "epoch": 0.15776198119697318, + "grad_norm": 0.5673873056584894, + "learning_rate": 0.0009578238332561075, + "loss": 2.5601, + "step": 688 + }, + { + "epoch": 0.15799128640220134, + "grad_norm": 0.6040269769467606, + "learning_rate": 0.0009576744324167379, + "loss": 2.5941, + "step": 689 + }, + { + "epoch": 0.1582205916074295, + "grad_norm": 0.6242194604124018, + "learning_rate": 0.0009575247791272129, + "loss": 2.845, + "step": 690 + }, + { + "epoch": 0.15844989681265764, + "grad_norm": 0.6078637106174865, + "learning_rate": 0.0009573748734700804, + "loss": 2.7203, + "step": 691 + }, + { + "epoch": 0.1586792020178858, + "grad_norm": 0.6964390868352294, + "learning_rate": 0.0009572247155280271, + "loss": 2.6848, + "step": 692 + }, + { + "epoch": 0.15890850722311398, + "grad_norm": 0.5531919953685518, + "learning_rate": 0.0009570743053838796, + "loss": 2.7356, + "step": 693 + }, + { + "epoch": 0.15913781242834213, + "grad_norm": 0.6007013750751182, + "learning_rate": 0.0009569236431206025, + "loss": 2.6968, + "step": 694 + }, + { + "epoch": 0.15936711763357028, + "grad_norm": 0.6030955628831624, + "learning_rate": 0.0009567727288213005, + "loss": 2.5857, + "step": 695 + }, + { + "epoch": 0.15959642283879844, + "grad_norm": 0.6018325185106664, + "learning_rate": 0.0009566215625692167, + "loss": 2.6349, + "step": 696 + }, + { + "epoch": 0.1598257280440266, + "grad_norm": 0.5886114673168791, + "learning_rate": 0.0009564701444477337, + "loss": 2.7652, + "step": 697 + }, + { + "epoch": 0.16005503324925477, + "grad_norm": 0.5838176631181529, + "learning_rate": 0.0009563184745403723, + "loss": 2.5431, + "step": 698 + }, + { + "epoch": 0.16028433845448292, + "grad_norm": 0.608391640127884, + "learning_rate": 0.000956166552930793, + "loss": 2.7264, + "step": 699 + }, + { + "epoch": 0.16051364365971107, + "grad_norm": 0.609305065761074, + "learning_rate": 0.0009560143797027946, + "loss": 2.5435, + "step": 700 + }, + { + "epoch": 0.16074294886493923, + "grad_norm": 0.5441199355070299, + "learning_rate": 0.0009558619549403147, + "loss": 2.6642, + "step": 701 + }, + { + "epoch": 0.16097225407016738, + "grad_norm": 0.61212430961782, + "learning_rate": 0.0009557092787274299, + "loss": 2.6435, + "step": 702 + }, + { + "epoch": 0.16120155927539556, + "grad_norm": 0.579142039050964, + "learning_rate": 0.0009555563511483555, + "loss": 2.6274, + "step": 703 + }, + { + "epoch": 0.16143086448062371, + "grad_norm": 0.5684231135018838, + "learning_rate": 0.0009554031722874451, + "loss": 2.6221, + "step": 704 + }, + { + "epoch": 0.16166016968585187, + "grad_norm": 0.5815475716598019, + "learning_rate": 0.0009552497422291912, + "loss": 2.581, + "step": 705 + }, + { + "epoch": 0.16188947489108002, + "grad_norm": 0.6391704090918481, + "learning_rate": 0.000955096061058225, + "loss": 2.7009, + "step": 706 + }, + { + "epoch": 0.16211878009630817, + "grad_norm": 0.5988123004247159, + "learning_rate": 0.0009549421288593157, + "loss": 2.6297, + "step": 707 + }, + { + "epoch": 0.16234808530153635, + "grad_norm": 0.6405099171873008, + "learning_rate": 0.0009547879457173714, + "loss": 2.6836, + "step": 708 + }, + { + "epoch": 0.1625773905067645, + "grad_norm": 0.6158100046515668, + "learning_rate": 0.0009546335117174385, + "loss": 2.6731, + "step": 709 + }, + { + "epoch": 0.16280669571199266, + "grad_norm": 0.6335830548640157, + "learning_rate": 0.0009544788269447018, + "loss": 2.6118, + "step": 710 + }, + { + "epoch": 0.1630360009172208, + "grad_norm": 0.6508892042892909, + "learning_rate": 0.0009543238914844843, + "loss": 2.6047, + "step": 711 + }, + { + "epoch": 0.16326530612244897, + "grad_norm": 0.5511317230203788, + "learning_rate": 0.0009541687054222474, + "loss": 2.6292, + "step": 712 + }, + { + "epoch": 0.16349461132767715, + "grad_norm": 0.6001105870210803, + "learning_rate": 0.0009540132688435907, + "loss": 2.5425, + "step": 713 + }, + { + "epoch": 0.1637239165329053, + "grad_norm": 0.6062404137287182, + "learning_rate": 0.0009538575818342521, + "loss": 2.6111, + "step": 714 + }, + { + "epoch": 0.16395322173813345, + "grad_norm": 0.5765795554557571, + "learning_rate": 0.0009537016444801074, + "loss": 2.6199, + "step": 715 + }, + { + "epoch": 0.1641825269433616, + "grad_norm": 0.5529368035558798, + "learning_rate": 0.0009535454568671704, + "loss": 2.7374, + "step": 716 + }, + { + "epoch": 0.16441183214858976, + "grad_norm": 0.562787531750508, + "learning_rate": 0.0009533890190815935, + "loss": 2.6371, + "step": 717 + }, + { + "epoch": 0.16464113735381794, + "grad_norm": 0.5690087338068902, + "learning_rate": 0.0009532323312096665, + "loss": 2.6103, + "step": 718 + }, + { + "epoch": 0.1648704425590461, + "grad_norm": 0.58197031183848, + "learning_rate": 0.0009530753933378173, + "loss": 2.6114, + "step": 719 + }, + { + "epoch": 0.16509974776427425, + "grad_norm": 0.5858991615860932, + "learning_rate": 0.0009529182055526121, + "loss": 2.651, + "step": 720 + }, + { + "epoch": 0.1653290529695024, + "grad_norm": 0.5997177334278254, + "learning_rate": 0.0009527607679407545, + "loss": 2.6044, + "step": 721 + }, + { + "epoch": 0.16555835817473058, + "grad_norm": 0.5151546572407538, + "learning_rate": 0.000952603080589086, + "loss": 2.6105, + "step": 722 + }, + { + "epoch": 0.16578766337995873, + "grad_norm": 0.5442087381762091, + "learning_rate": 0.0009524451435845857, + "loss": 2.6147, + "step": 723 + }, + { + "epoch": 0.16601696858518689, + "grad_norm": 0.5988729470498328, + "learning_rate": 0.000952286957014371, + "loss": 2.5924, + "step": 724 + }, + { + "epoch": 0.16624627379041504, + "grad_norm": 0.5392621094566636, + "learning_rate": 0.0009521285209656963, + "loss": 2.5473, + "step": 725 + }, + { + "epoch": 0.1664755789956432, + "grad_norm": 0.5533761284726162, + "learning_rate": 0.0009519698355259538, + "loss": 2.6615, + "step": 726 + }, + { + "epoch": 0.16670488420087137, + "grad_norm": 1.5017098866913354, + "learning_rate": 0.0009518109007826734, + "loss": 2.5838, + "step": 727 + }, + { + "epoch": 0.16693418940609953, + "grad_norm": 0.5725750108562958, + "learning_rate": 0.0009516517168235226, + "loss": 2.6337, + "step": 728 + }, + { + "epoch": 0.16716349461132768, + "grad_norm": 0.649333546401536, + "learning_rate": 0.0009514922837363059, + "loss": 2.5499, + "step": 729 + }, + { + "epoch": 0.16739279981655583, + "grad_norm": 0.5701059970151839, + "learning_rate": 0.0009513326016089658, + "loss": 2.7237, + "step": 730 + }, + { + "epoch": 0.16762210502178398, + "grad_norm": 0.6276476365425432, + "learning_rate": 0.0009511726705295817, + "loss": 2.6717, + "step": 731 + }, + { + "epoch": 0.16785141022701217, + "grad_norm": 0.5478487964951606, + "learning_rate": 0.0009510124905863709, + "loss": 2.5865, + "step": 732 + }, + { + "epoch": 0.16808071543224032, + "grad_norm": 0.6132210320656695, + "learning_rate": 0.000950852061867687, + "loss": 2.8059, + "step": 733 + }, + { + "epoch": 0.16831002063746847, + "grad_norm": 0.7725317071471209, + "learning_rate": 0.0009506913844620217, + "loss": 2.6077, + "step": 734 + }, + { + "epoch": 0.16853932584269662, + "grad_norm": 0.5770644311531319, + "learning_rate": 0.0009505304584580038, + "loss": 2.4925, + "step": 735 + }, + { + "epoch": 0.16876863104792478, + "grad_norm": 0.5877654315941752, + "learning_rate": 0.0009503692839443987, + "loss": 2.689, + "step": 736 + }, + { + "epoch": 0.16899793625315296, + "grad_norm": 0.6335517426107904, + "learning_rate": 0.0009502078610101092, + "loss": 2.6976, + "step": 737 + }, + { + "epoch": 0.1692272414583811, + "grad_norm": 0.6295678396301381, + "learning_rate": 0.0009500461897441752, + "loss": 2.6649, + "step": 738 + }, + { + "epoch": 0.16945654666360926, + "grad_norm": 0.7998091842577779, + "learning_rate": 0.0009498842702357736, + "loss": 2.6557, + "step": 739 + }, + { + "epoch": 0.16968585186883742, + "grad_norm": 0.6500856921823084, + "learning_rate": 0.0009497221025742177, + "loss": 2.5899, + "step": 740 + }, + { + "epoch": 0.16991515707406557, + "grad_norm": 0.6171625169949876, + "learning_rate": 0.0009495596868489587, + "loss": 2.6472, + "step": 741 + }, + { + "epoch": 0.17014446227929375, + "grad_norm": 0.5994391543812632, + "learning_rate": 0.0009493970231495835, + "loss": 2.5048, + "step": 742 + }, + { + "epoch": 0.1703737674845219, + "grad_norm": 0.6374449161346082, + "learning_rate": 0.0009492341115658167, + "loss": 2.6204, + "step": 743 + }, + { + "epoch": 0.17060307268975006, + "grad_norm": 0.6511337554371005, + "learning_rate": 0.0009490709521875189, + "loss": 2.5691, + "step": 744 + }, + { + "epoch": 0.1708323778949782, + "grad_norm": 0.5923434539384794, + "learning_rate": 0.0009489075451046879, + "loss": 2.5123, + "step": 745 + }, + { + "epoch": 0.17106168310020636, + "grad_norm": 0.7226956991186433, + "learning_rate": 0.000948743890407458, + "loss": 2.6432, + "step": 746 + }, + { + "epoch": 0.17129098830543454, + "grad_norm": 0.8178768014220795, + "learning_rate": 0.0009485799881861, + "loss": 2.7401, + "step": 747 + }, + { + "epoch": 0.1715202935106627, + "grad_norm": 0.8094745617695469, + "learning_rate": 0.0009484158385310212, + "loss": 2.5994, + "step": 748 + }, + { + "epoch": 0.17174959871589085, + "grad_norm": 0.9786412697671586, + "learning_rate": 0.0009482514415327654, + "loss": 2.8047, + "step": 749 + }, + { + "epoch": 0.171978903921119, + "grad_norm": 0.5333782945165835, + "learning_rate": 0.0009480867972820131, + "loss": 2.6698, + "step": 750 + }, + { + "epoch": 0.17220820912634716, + "grad_norm": 0.6199017511949049, + "learning_rate": 0.000947921905869581, + "loss": 2.4844, + "step": 751 + }, + { + "epoch": 0.17243751433157534, + "grad_norm": 0.740961655640854, + "learning_rate": 0.0009477567673864216, + "loss": 2.6353, + "step": 752 + }, + { + "epoch": 0.1726668195368035, + "grad_norm": 0.5707042574604405, + "learning_rate": 0.0009475913819236248, + "loss": 2.5913, + "step": 753 + }, + { + "epoch": 0.17289612474203164, + "grad_norm": 2.4249880486951634, + "learning_rate": 0.0009474257495724156, + "loss": 2.5931, + "step": 754 + }, + { + "epoch": 0.1731254299472598, + "grad_norm": 0.6543197075178601, + "learning_rate": 0.0009472598704241561, + "loss": 2.6907, + "step": 755 + }, + { + "epoch": 0.17335473515248795, + "grad_norm": 0.6322432807067385, + "learning_rate": 0.000947093744570344, + "loss": 2.6234, + "step": 756 + }, + { + "epoch": 0.17358404035771613, + "grad_norm": 0.6653107738685661, + "learning_rate": 0.0009469273721026131, + "loss": 2.7499, + "step": 757 + }, + { + "epoch": 0.17381334556294428, + "grad_norm": 0.8160252327410751, + "learning_rate": 0.0009467607531127335, + "loss": 2.5971, + "step": 758 + }, + { + "epoch": 0.17404265076817244, + "grad_norm": 0.6807836414044495, + "learning_rate": 0.0009465938876926111, + "loss": 2.7408, + "step": 759 + }, + { + "epoch": 0.1742719559734006, + "grad_norm": 0.8177524730208959, + "learning_rate": 0.0009464267759342875, + "loss": 2.5085, + "step": 760 + }, + { + "epoch": 0.17450126117862874, + "grad_norm": 0.5446335022022794, + "learning_rate": 0.0009462594179299406, + "loss": 2.6048, + "step": 761 + }, + { + "epoch": 0.17473056638385692, + "grad_norm": 0.6484490993229063, + "learning_rate": 0.0009460918137718841, + "loss": 2.4542, + "step": 762 + }, + { + "epoch": 0.17495987158908508, + "grad_norm": 0.8324411494600152, + "learning_rate": 0.0009459239635525672, + "loss": 2.7773, + "step": 763 + }, + { + "epoch": 0.17518917679431323, + "grad_norm": 0.6308451194674497, + "learning_rate": 0.000945755867364575, + "loss": 2.5085, + "step": 764 + }, + { + "epoch": 0.17541848199954138, + "grad_norm": 1.2166826046762598, + "learning_rate": 0.0009455875253006281, + "loss": 2.6686, + "step": 765 + }, + { + "epoch": 0.17564778720476956, + "grad_norm": 0.8743362953551981, + "learning_rate": 0.0009454189374535829, + "loss": 2.7066, + "step": 766 + }, + { + "epoch": 0.17587709240999772, + "grad_norm": 0.6940876280531686, + "learning_rate": 0.0009452501039164315, + "loss": 2.6395, + "step": 767 + }, + { + "epoch": 0.17610639761522587, + "grad_norm": 0.8096104413148612, + "learning_rate": 0.0009450810247823012, + "loss": 2.7467, + "step": 768 + }, + { + "epoch": 0.17633570282045402, + "grad_norm": 0.7861313246371997, + "learning_rate": 0.0009449117001444549, + "loss": 2.514, + "step": 769 + }, + { + "epoch": 0.17656500802568217, + "grad_norm": 0.7742671602143275, + "learning_rate": 0.0009447421300962911, + "loss": 2.8212, + "step": 770 + }, + { + "epoch": 0.17679431323091035, + "grad_norm": 0.6523645590122369, + "learning_rate": 0.0009445723147313433, + "loss": 2.5136, + "step": 771 + }, + { + "epoch": 0.1770236184361385, + "grad_norm": 0.5945209094967606, + "learning_rate": 0.0009444022541432806, + "loss": 2.6559, + "step": 772 + }, + { + "epoch": 0.17725292364136666, + "grad_norm": 0.6129112282154288, + "learning_rate": 0.0009442319484259074, + "loss": 2.539, + "step": 773 + }, + { + "epoch": 0.17748222884659481, + "grad_norm": 0.5497447677531998, + "learning_rate": 0.000944061397673163, + "loss": 2.4888, + "step": 774 + }, + { + "epoch": 0.17771153405182297, + "grad_norm": 0.5995892971928244, + "learning_rate": 0.0009438906019791222, + "loss": 2.6028, + "step": 775 + }, + { + "epoch": 0.17794083925705115, + "grad_norm": 0.5740342853122383, + "learning_rate": 0.0009437195614379946, + "loss": 2.6488, + "step": 776 + }, + { + "epoch": 0.1781701444622793, + "grad_norm": 0.5560732265699864, + "learning_rate": 0.0009435482761441251, + "loss": 2.6951, + "step": 777 + }, + { + "epoch": 0.17839944966750745, + "grad_norm": 0.5489530121007206, + "learning_rate": 0.0009433767461919937, + "loss": 2.6916, + "step": 778 + }, + { + "epoch": 0.1786287548727356, + "grad_norm": 0.5964204269343198, + "learning_rate": 0.000943204971676215, + "loss": 2.5014, + "step": 779 + }, + { + "epoch": 0.17885806007796376, + "grad_norm": 0.5695015854386745, + "learning_rate": 0.0009430329526915386, + "loss": 2.6551, + "step": 780 + }, + { + "epoch": 0.17908736528319194, + "grad_norm": 0.6652279072023947, + "learning_rate": 0.0009428606893328493, + "loss": 2.5821, + "step": 781 + }, + { + "epoch": 0.1793166704884201, + "grad_norm": 0.6561019167959505, + "learning_rate": 0.000942688181695166, + "loss": 2.6763, + "step": 782 + }, + { + "epoch": 0.17954597569364825, + "grad_norm": 0.5907684857195635, + "learning_rate": 0.0009425154298736432, + "loss": 2.6508, + "step": 783 + }, + { + "epoch": 0.1797752808988764, + "grad_norm": 1.3187286248590215, + "learning_rate": 0.0009423424339635694, + "loss": 2.7418, + "step": 784 + }, + { + "epoch": 0.18000458610410455, + "grad_norm": 0.5763848341079852, + "learning_rate": 0.0009421691940603678, + "loss": 2.7216, + "step": 785 + }, + { + "epoch": 0.18023389130933273, + "grad_norm": 0.9180851048852515, + "learning_rate": 0.0009419957102595968, + "loss": 2.7524, + "step": 786 + }, + { + "epoch": 0.1804631965145609, + "grad_norm": 0.5191743787719795, + "learning_rate": 0.0009418219826569488, + "loss": 2.6016, + "step": 787 + }, + { + "epoch": 0.18069250171978904, + "grad_norm": 0.5919716563827763, + "learning_rate": 0.0009416480113482504, + "loss": 2.5083, + "step": 788 + }, + { + "epoch": 0.1809218069250172, + "grad_norm": 0.7446835537286906, + "learning_rate": 0.0009414737964294635, + "loss": 2.6046, + "step": 789 + }, + { + "epoch": 0.18115111213024535, + "grad_norm": 0.6606175729707061, + "learning_rate": 0.0009412993379966836, + "loss": 2.7897, + "step": 790 + }, + { + "epoch": 0.18138041733547353, + "grad_norm": 0.7585292841545105, + "learning_rate": 0.000941124636146141, + "loss": 2.6743, + "step": 791 + }, + { + "epoch": 0.18160972254070168, + "grad_norm": 0.7767794638833198, + "learning_rate": 0.0009409496909741998, + "loss": 2.6716, + "step": 792 + }, + { + "epoch": 0.18183902774592983, + "grad_norm": 0.7079747371164424, + "learning_rate": 0.0009407745025773589, + "loss": 2.5226, + "step": 793 + }, + { + "epoch": 0.18206833295115799, + "grad_norm": 0.6690434769326156, + "learning_rate": 0.0009405990710522508, + "loss": 2.5469, + "step": 794 + }, + { + "epoch": 0.18229763815638614, + "grad_norm": 0.6034789612058635, + "learning_rate": 0.0009404233964956423, + "loss": 2.7595, + "step": 795 + }, + { + "epoch": 0.18252694336161432, + "grad_norm": 0.7213721673218417, + "learning_rate": 0.0009402474790044346, + "loss": 2.5675, + "step": 796 + }, + { + "epoch": 0.18275624856684247, + "grad_norm": 0.5543748983402408, + "learning_rate": 0.0009400713186756625, + "loss": 2.5448, + "step": 797 + }, + { + "epoch": 0.18298555377207062, + "grad_norm": 0.6256323714791832, + "learning_rate": 0.0009398949156064948, + "loss": 2.6014, + "step": 798 + }, + { + "epoch": 0.18321485897729878, + "grad_norm": 0.5937434850745722, + "learning_rate": 0.0009397182698942342, + "loss": 2.6156, + "step": 799 + }, + { + "epoch": 0.18344416418252693, + "grad_norm": 0.5615055681133202, + "learning_rate": 0.0009395413816363175, + "loss": 2.7526, + "step": 800 + }, + { + "epoch": 0.1836734693877551, + "grad_norm": 0.5993578396900704, + "learning_rate": 0.0009393642509303149, + "loss": 2.7641, + "step": 801 + }, + { + "epoch": 0.18390277459298326, + "grad_norm": 0.5490009901076589, + "learning_rate": 0.0009391868778739305, + "loss": 2.7441, + "step": 802 + }, + { + "epoch": 0.18413207979821142, + "grad_norm": 0.6457507062830766, + "learning_rate": 0.0009390092625650023, + "loss": 2.6624, + "step": 803 + }, + { + "epoch": 0.18436138500343957, + "grad_norm": 0.525735364967052, + "learning_rate": 0.0009388314051015015, + "loss": 2.5551, + "step": 804 + }, + { + "epoch": 0.18459069020866772, + "grad_norm": 0.5673561460176977, + "learning_rate": 0.0009386533055815332, + "loss": 2.5921, + "step": 805 + }, + { + "epoch": 0.1848199954138959, + "grad_norm": 0.49443941310856265, + "learning_rate": 0.0009384749641033358, + "loss": 2.5846, + "step": 806 + }, + { + "epoch": 0.18504930061912406, + "grad_norm": 0.6084843069441535, + "learning_rate": 0.0009382963807652813, + "loss": 2.5999, + "step": 807 + }, + { + "epoch": 0.1852786058243522, + "grad_norm": 0.6364948838449934, + "learning_rate": 0.0009381175556658752, + "loss": 2.6101, + "step": 808 + }, + { + "epoch": 0.18550791102958036, + "grad_norm": 0.5683972274554904, + "learning_rate": 0.000937938488903756, + "loss": 2.6452, + "step": 809 + }, + { + "epoch": 0.18573721623480852, + "grad_norm": 0.5594411938152388, + "learning_rate": 0.0009377591805776961, + "loss": 2.7243, + "step": 810 + }, + { + "epoch": 0.1859665214400367, + "grad_norm": 0.5929923047821192, + "learning_rate": 0.0009375796307866003, + "loss": 2.6293, + "step": 811 + }, + { + "epoch": 0.18619582664526485, + "grad_norm": 0.5500391126163324, + "learning_rate": 0.0009373998396295074, + "loss": 2.5653, + "step": 812 + }, + { + "epoch": 0.186425131850493, + "grad_norm": 0.5546683080358317, + "learning_rate": 0.0009372198072055888, + "loss": 2.6396, + "step": 813 + }, + { + "epoch": 0.18665443705572116, + "grad_norm": 0.5354065403063325, + "learning_rate": 0.0009370395336141495, + "loss": 2.4853, + "step": 814 + }, + { + "epoch": 0.18688374226094934, + "grad_norm": 0.6069158422855603, + "learning_rate": 0.0009368590189546268, + "loss": 2.7075, + "step": 815 + }, + { + "epoch": 0.1871130474661775, + "grad_norm": 0.5755902715034543, + "learning_rate": 0.0009366782633265916, + "loss": 2.6619, + "step": 816 + }, + { + "epoch": 0.18734235267140564, + "grad_norm": 0.5661075052400032, + "learning_rate": 0.0009364972668297474, + "loss": 2.7042, + "step": 817 + }, + { + "epoch": 0.1875716578766338, + "grad_norm": 0.5330995315932455, + "learning_rate": 0.0009363160295639308, + "loss": 2.5621, + "step": 818 + }, + { + "epoch": 0.18780096308186195, + "grad_norm": 0.5392806939467549, + "learning_rate": 0.0009361345516291111, + "loss": 2.5642, + "step": 819 + }, + { + "epoch": 0.18803026828709013, + "grad_norm": 0.5255055524236761, + "learning_rate": 0.0009359528331253901, + "loss": 2.6862, + "step": 820 + }, + { + "epoch": 0.18825957349231828, + "grad_norm": 0.7639667009350554, + "learning_rate": 0.0009357708741530024, + "loss": 2.7202, + "step": 821 + }, + { + "epoch": 0.18848887869754644, + "grad_norm": 0.5230276315731069, + "learning_rate": 0.0009355886748123158, + "loss": 2.6324, + "step": 822 + }, + { + "epoch": 0.1887181839027746, + "grad_norm": 0.5587985669787732, + "learning_rate": 0.00093540623520383, + "loss": 2.5469, + "step": 823 + }, + { + "epoch": 0.18894748910800274, + "grad_norm": 0.6119545972498001, + "learning_rate": 0.0009352235554281774, + "loss": 2.627, + "step": 824 + }, + { + "epoch": 0.18917679431323092, + "grad_norm": 0.7157899269572204, + "learning_rate": 0.000935040635586123, + "loss": 2.575, + "step": 825 + }, + { + "epoch": 0.18940609951845908, + "grad_norm": 0.5302161603296106, + "learning_rate": 0.0009348574757785642, + "loss": 2.6811, + "step": 826 + }, + { + "epoch": 0.18963540472368723, + "grad_norm": 0.6117908313216828, + "learning_rate": 0.0009346740761065305, + "loss": 2.657, + "step": 827 + }, + { + "epoch": 0.18986470992891538, + "grad_norm": 0.5807608209577789, + "learning_rate": 0.0009344904366711844, + "loss": 2.6594, + "step": 828 + }, + { + "epoch": 0.19009401513414353, + "grad_norm": 0.5910531676732003, + "learning_rate": 0.0009343065575738197, + "loss": 2.573, + "step": 829 + }, + { + "epoch": 0.19032332033937172, + "grad_norm": 0.5625976671356495, + "learning_rate": 0.0009341224389158631, + "loss": 2.5851, + "step": 830 + }, + { + "epoch": 0.19055262554459987, + "grad_norm": 0.5564019079878183, + "learning_rate": 0.0009339380807988733, + "loss": 2.606, + "step": 831 + }, + { + "epoch": 0.19078193074982802, + "grad_norm": 0.5429420169269378, + "learning_rate": 0.0009337534833245409, + "loss": 2.6639, + "step": 832 + }, + { + "epoch": 0.19101123595505617, + "grad_norm": 0.4941502180305949, + "learning_rate": 0.0009335686465946887, + "loss": 2.755, + "step": 833 + }, + { + "epoch": 0.19124054116028433, + "grad_norm": 0.5920802142251423, + "learning_rate": 0.0009333835707112713, + "loss": 2.7178, + "step": 834 + }, + { + "epoch": 0.1914698463655125, + "grad_norm": 0.5631837557305065, + "learning_rate": 0.0009331982557763754, + "loss": 2.4964, + "step": 835 + }, + { + "epoch": 0.19169915157074066, + "grad_norm": 0.5688623374177513, + "learning_rate": 0.0009330127018922195, + "loss": 2.511, + "step": 836 + }, + { + "epoch": 0.19192845677596881, + "grad_norm": 0.5827358330979518, + "learning_rate": 0.0009328269091611537, + "loss": 2.6407, + "step": 837 + }, + { + "epoch": 0.19215776198119697, + "grad_norm": 0.6027346442736435, + "learning_rate": 0.00093264087768566, + "loss": 2.7276, + "step": 838 + }, + { + "epoch": 0.19238706718642512, + "grad_norm": 0.5819861110000321, + "learning_rate": 0.0009324546075683524, + "loss": 2.6522, + "step": 839 + }, + { + "epoch": 0.1926163723916533, + "grad_norm": 0.6332172385137345, + "learning_rate": 0.0009322680989119763, + "loss": 2.6077, + "step": 840 + }, + { + "epoch": 0.19284567759688145, + "grad_norm": 0.5386713900220237, + "learning_rate": 0.0009320813518194083, + "loss": 2.5216, + "step": 841 + }, + { + "epoch": 0.1930749828021096, + "grad_norm": 0.6053643331860359, + "learning_rate": 0.000931894366393657, + "loss": 2.463, + "step": 842 + }, + { + "epoch": 0.19330428800733776, + "grad_norm": 0.6349710740242767, + "learning_rate": 0.0009317071427378624, + "loss": 2.4853, + "step": 843 + }, + { + "epoch": 0.1935335932125659, + "grad_norm": 0.502990198450512, + "learning_rate": 0.0009315196809552956, + "loss": 2.6211, + "step": 844 + }, + { + "epoch": 0.1937628984177941, + "grad_norm": 0.5544241246719038, + "learning_rate": 0.0009313319811493594, + "loss": 2.6627, + "step": 845 + }, + { + "epoch": 0.19399220362302225, + "grad_norm": 0.6567656185973659, + "learning_rate": 0.0009311440434235878, + "loss": 2.6886, + "step": 846 + }, + { + "epoch": 0.1942215088282504, + "grad_norm": 0.5531198879670239, + "learning_rate": 0.000930955867881646, + "loss": 2.5192, + "step": 847 + }, + { + "epoch": 0.19445081403347855, + "grad_norm": 0.5888553837740109, + "learning_rate": 0.0009307674546273301, + "loss": 2.492, + "step": 848 + }, + { + "epoch": 0.1946801192387067, + "grad_norm": 0.5209765468275656, + "learning_rate": 0.0009305788037645681, + "loss": 2.653, + "step": 849 + }, + { + "epoch": 0.1949094244439349, + "grad_norm": 0.6295068266150337, + "learning_rate": 0.000930389915397418, + "loss": 2.6716, + "step": 850 + }, + { + "epoch": 0.19513872964916304, + "grad_norm": 0.5087674639448507, + "learning_rate": 0.0009302007896300697, + "loss": 2.679, + "step": 851 + }, + { + "epoch": 0.1953680348543912, + "grad_norm": 0.5144599970454604, + "learning_rate": 0.0009300114265668436, + "loss": 2.6181, + "step": 852 + }, + { + "epoch": 0.19559734005961935, + "grad_norm": 0.5904983644852464, + "learning_rate": 0.0009298218263121911, + "loss": 2.5825, + "step": 853 + }, + { + "epoch": 0.1958266452648475, + "grad_norm": 0.5346568081310518, + "learning_rate": 0.0009296319889706946, + "loss": 2.628, + "step": 854 + }, + { + "epoch": 0.19605595047007568, + "grad_norm": 0.4966316935401516, + "learning_rate": 0.0009294419146470668, + "loss": 2.6229, + "step": 855 + }, + { + "epoch": 0.19628525567530383, + "grad_norm": 0.5150950006621237, + "learning_rate": 0.0009292516034461515, + "loss": 2.6791, + "step": 856 + }, + { + "epoch": 0.19651456088053199, + "grad_norm": 0.5474758535013369, + "learning_rate": 0.0009290610554729234, + "loss": 2.5434, + "step": 857 + }, + { + "epoch": 0.19674386608576014, + "grad_norm": 0.532797826158212, + "learning_rate": 0.0009288702708324872, + "loss": 2.6831, + "step": 858 + }, + { + "epoch": 0.19697317129098832, + "grad_norm": 0.5778632997902853, + "learning_rate": 0.0009286792496300784, + "loss": 2.7651, + "step": 859 + }, + { + "epoch": 0.19720247649621647, + "grad_norm": 0.5208076696456018, + "learning_rate": 0.0009284879919710631, + "loss": 2.6199, + "step": 860 + }, + { + "epoch": 0.19743178170144463, + "grad_norm": 0.5176748844084559, + "learning_rate": 0.0009282964979609379, + "loss": 2.6449, + "step": 861 + }, + { + "epoch": 0.19766108690667278, + "grad_norm": 0.5514669615379315, + "learning_rate": 0.0009281047677053293, + "loss": 2.726, + "step": 862 + }, + { + "epoch": 0.19789039211190093, + "grad_norm": 0.5556388938496857, + "learning_rate": 0.0009279128013099947, + "loss": 2.6839, + "step": 863 + }, + { + "epoch": 0.1981196973171291, + "grad_norm": 0.5254182591860521, + "learning_rate": 0.0009277205988808214, + "loss": 2.5386, + "step": 864 + }, + { + "epoch": 0.19834900252235727, + "grad_norm": 0.5345640915169386, + "learning_rate": 0.0009275281605238268, + "loss": 2.5766, + "step": 865 + }, + { + "epoch": 0.19857830772758542, + "grad_norm": 0.5098576068596419, + "learning_rate": 0.0009273354863451589, + "loss": 2.6978, + "step": 866 + }, + { + "epoch": 0.19880761293281357, + "grad_norm": 0.5352239138586762, + "learning_rate": 0.0009271425764510953, + "loss": 2.6082, + "step": 867 + }, + { + "epoch": 0.19903691813804172, + "grad_norm": 0.5610565090983125, + "learning_rate": 0.000926949430948044, + "loss": 2.67, + "step": 868 + }, + { + "epoch": 0.1992662233432699, + "grad_norm": 0.656920503629115, + "learning_rate": 0.0009267560499425423, + "loss": 2.6164, + "step": 869 + }, + { + "epoch": 0.19949552854849806, + "grad_norm": 0.5213250962371483, + "learning_rate": 0.0009265624335412584, + "loss": 2.5341, + "step": 870 + }, + { + "epoch": 0.1997248337537262, + "grad_norm": 0.4974392263182893, + "learning_rate": 0.0009263685818509895, + "loss": 2.5849, + "step": 871 + }, + { + "epoch": 0.19995413895895436, + "grad_norm": 0.5431010367085529, + "learning_rate": 0.000926174494978663, + "loss": 2.5553, + "step": 872 + }, + { + "epoch": 0.20018344416418252, + "grad_norm": 0.5328290809766733, + "learning_rate": 0.000925980173031336, + "loss": 2.6237, + "step": 873 + }, + { + "epoch": 0.2004127493694107, + "grad_norm": 0.5687733720991106, + "learning_rate": 0.0009257856161161952, + "loss": 2.5584, + "step": 874 + }, + { + "epoch": 0.20064205457463885, + "grad_norm": 0.5171564416988184, + "learning_rate": 0.0009255908243405567, + "loss": 2.6519, + "step": 875 + }, + { + "epoch": 0.200871359779867, + "grad_norm": 0.523661667774422, + "learning_rate": 0.0009253957978118663, + "loss": 2.4875, + "step": 876 + }, + { + "epoch": 0.20110066498509516, + "grad_norm": 0.5581359438441194, + "learning_rate": 0.0009252005366376996, + "loss": 2.7369, + "step": 877 + }, + { + "epoch": 0.2013299701903233, + "grad_norm": 0.6498100200900407, + "learning_rate": 0.0009250050409257611, + "loss": 2.5333, + "step": 878 + }, + { + "epoch": 0.2015592753955515, + "grad_norm": 0.7945156719929152, + "learning_rate": 0.0009248093107838852, + "loss": 2.5651, + "step": 879 + }, + { + "epoch": 0.20178858060077964, + "grad_norm": 0.5008801563156086, + "learning_rate": 0.0009246133463200352, + "loss": 2.5529, + "step": 880 + }, + { + "epoch": 0.2020178858060078, + "grad_norm": 0.5450016252619178, + "learning_rate": 0.0009244171476423036, + "loss": 2.7562, + "step": 881 + }, + { + "epoch": 0.20224719101123595, + "grad_norm": 0.5700906371909651, + "learning_rate": 0.0009242207148589125, + "loss": 2.6107, + "step": 882 + }, + { + "epoch": 0.2024764962164641, + "grad_norm": 0.5466015085352527, + "learning_rate": 0.0009240240480782129, + "loss": 2.6282, + "step": 883 + }, + { + "epoch": 0.20270580142169228, + "grad_norm": 0.5531639582988874, + "learning_rate": 0.000923827147408685, + "loss": 2.5381, + "step": 884 + }, + { + "epoch": 0.20293510662692044, + "grad_norm": 0.5364550504396982, + "learning_rate": 0.0009236300129589376, + "loss": 2.7647, + "step": 885 + }, + { + "epoch": 0.2031644118321486, + "grad_norm": 0.5344788567885463, + "learning_rate": 0.0009234326448377088, + "loss": 2.6058, + "step": 886 + }, + { + "epoch": 0.20339371703737674, + "grad_norm": 1.0295169536134363, + "learning_rate": 0.0009232350431538657, + "loss": 2.6056, + "step": 887 + }, + { + "epoch": 0.2036230222426049, + "grad_norm": 0.53998905284473, + "learning_rate": 0.000923037208016404, + "loss": 2.634, + "step": 888 + }, + { + "epoch": 0.20385232744783308, + "grad_norm": 1.1933123418069547, + "learning_rate": 0.0009228391395344482, + "loss": 2.6943, + "step": 889 + }, + { + "epoch": 0.20408163265306123, + "grad_norm": 0.5685751240770236, + "learning_rate": 0.0009226408378172515, + "loss": 2.5791, + "step": 890 + }, + { + "epoch": 0.20431093785828938, + "grad_norm": 0.5847302119200467, + "learning_rate": 0.000922442302974196, + "loss": 2.5628, + "step": 891 + }, + { + "epoch": 0.20454024306351754, + "grad_norm": 0.5753828572521712, + "learning_rate": 0.000922243535114792, + "loss": 2.6234, + "step": 892 + }, + { + "epoch": 0.2047695482687457, + "grad_norm": 0.5168751342100233, + "learning_rate": 0.0009220445343486785, + "loss": 2.5979, + "step": 893 + }, + { + "epoch": 0.20499885347397387, + "grad_norm": 0.5619775081728319, + "learning_rate": 0.0009218453007856231, + "loss": 2.6188, + "step": 894 + }, + { + "epoch": 0.20522815867920202, + "grad_norm": 0.4785793376231115, + "learning_rate": 0.0009216458345355217, + "loss": 2.676, + "step": 895 + }, + { + "epoch": 0.20545746388443017, + "grad_norm": 0.5320873318255053, + "learning_rate": 0.0009214461357083985, + "loss": 2.5624, + "step": 896 + }, + { + "epoch": 0.20568676908965833, + "grad_norm": 0.5132371793775127, + "learning_rate": 0.0009212462044144061, + "loss": 2.5334, + "step": 897 + }, + { + "epoch": 0.20591607429488648, + "grad_norm": 0.5336375736810005, + "learning_rate": 0.0009210460407638253, + "loss": 2.4232, + "step": 898 + }, + { + "epoch": 0.20614537950011466, + "grad_norm": 0.531079753771586, + "learning_rate": 0.0009208456448670648, + "loss": 2.4036, + "step": 899 + }, + { + "epoch": 0.20637468470534281, + "grad_norm": 0.6120219765899287, + "learning_rate": 0.0009206450168346618, + "loss": 2.5536, + "step": 900 + }, + { + "epoch": 0.20660398991057097, + "grad_norm": 0.5875126020105468, + "learning_rate": 0.0009204441567772816, + "loss": 2.715, + "step": 901 + }, + { + "epoch": 0.20683329511579912, + "grad_norm": 0.653455073582347, + "learning_rate": 0.0009202430648057168, + "loss": 2.6308, + "step": 902 + }, + { + "epoch": 0.20706260032102727, + "grad_norm": 0.6015342073429464, + "learning_rate": 0.0009200417410308888, + "loss": 2.5986, + "step": 903 + }, + { + "epoch": 0.20729190552625545, + "grad_norm": 0.6205148875664964, + "learning_rate": 0.0009198401855638461, + "loss": 2.7317, + "step": 904 + }, + { + "epoch": 0.2075212107314836, + "grad_norm": 0.61337286387987, + "learning_rate": 0.0009196383985157656, + "loss": 2.5348, + "step": 905 + }, + { + "epoch": 0.20775051593671176, + "grad_norm": 0.5501680995401871, + "learning_rate": 0.0009194363799979515, + "loss": 2.6267, + "step": 906 + }, + { + "epoch": 0.2079798211419399, + "grad_norm": 0.487552733132786, + "learning_rate": 0.000919234130121836, + "loss": 2.7257, + "step": 907 + }, + { + "epoch": 0.2082091263471681, + "grad_norm": 0.543431421745895, + "learning_rate": 0.000919031648998979, + "loss": 2.5156, + "step": 908 + }, + { + "epoch": 0.20843843155239625, + "grad_norm": 0.5322136580686172, + "learning_rate": 0.0009188289367410672, + "loss": 2.6086, + "step": 909 + }, + { + "epoch": 0.2086677367576244, + "grad_norm": 0.5150592538859086, + "learning_rate": 0.0009186259934599157, + "loss": 2.4665, + "step": 910 + }, + { + "epoch": 0.20889704196285255, + "grad_norm": 0.5012686455449378, + "learning_rate": 0.0009184228192674666, + "loss": 2.4397, + "step": 911 + }, + { + "epoch": 0.2091263471680807, + "grad_norm": 0.5730578195264024, + "learning_rate": 0.0009182194142757893, + "loss": 2.6007, + "step": 912 + }, + { + "epoch": 0.2093556523733089, + "grad_norm": 0.6009739424543851, + "learning_rate": 0.0009180157785970808, + "loss": 2.6129, + "step": 913 + }, + { + "epoch": 0.20958495757853704, + "grad_norm": 0.5780652263747186, + "learning_rate": 0.0009178119123436651, + "loss": 2.6672, + "step": 914 + }, + { + "epoch": 0.2098142627837652, + "grad_norm": 0.5840424782687995, + "learning_rate": 0.0009176078156279932, + "loss": 2.5964, + "step": 915 + }, + { + "epoch": 0.21004356798899335, + "grad_norm": 0.5977680364605673, + "learning_rate": 0.000917403488562644, + "loss": 2.5607, + "step": 916 + }, + { + "epoch": 0.2102728731942215, + "grad_norm": 0.5704494571409638, + "learning_rate": 0.0009171989312603226, + "loss": 2.5673, + "step": 917 + }, + { + "epoch": 0.21050217839944968, + "grad_norm": 0.548475922551634, + "learning_rate": 0.0009169941438338616, + "loss": 2.7532, + "step": 918 + }, + { + "epoch": 0.21073148360467783, + "grad_norm": 0.477943647854946, + "learning_rate": 0.0009167891263962202, + "loss": 2.522, + "step": 919 + }, + { + "epoch": 0.21096078880990599, + "grad_norm": 0.5544815332903971, + "learning_rate": 0.0009165838790604848, + "loss": 2.5535, + "step": 920 + }, + { + "epoch": 0.21119009401513414, + "grad_norm": 0.5060801690290005, + "learning_rate": 0.0009163784019398685, + "loss": 2.6259, + "step": 921 + }, + { + "epoch": 0.2114193992203623, + "grad_norm": 0.5395792819648405, + "learning_rate": 0.0009161726951477107, + "loss": 2.7474, + "step": 922 + }, + { + "epoch": 0.21164870442559047, + "grad_norm": 0.5498084076583595, + "learning_rate": 0.0009159667587974785, + "loss": 2.4593, + "step": 923 + }, + { + "epoch": 0.21187800963081863, + "grad_norm": 0.521792132507052, + "learning_rate": 0.0009157605930027646, + "loss": 2.5721, + "step": 924 + }, + { + "epoch": 0.21210731483604678, + "grad_norm": 0.5056597553272131, + "learning_rate": 0.0009155541978772887, + "loss": 2.6136, + "step": 925 + }, + { + "epoch": 0.21233662004127493, + "grad_norm": 1.010583709585516, + "learning_rate": 0.0009153475735348973, + "loss": 2.3837, + "step": 926 + }, + { + "epoch": 0.21256592524650308, + "grad_norm": 0.8388733590376442, + "learning_rate": 0.0009151407200895625, + "loss": 2.5716, + "step": 927 + }, + { + "epoch": 0.21279523045173127, + "grad_norm": 0.5691235938436544, + "learning_rate": 0.0009149336376553836, + "loss": 2.657, + "step": 928 + }, + { + "epoch": 0.21302453565695942, + "grad_norm": 0.5305276873848594, + "learning_rate": 0.000914726326346586, + "loss": 2.6007, + "step": 929 + }, + { + "epoch": 0.21325384086218757, + "grad_norm": 0.5310170243816872, + "learning_rate": 0.0009145187862775209, + "loss": 2.5741, + "step": 930 + }, + { + "epoch": 0.21348314606741572, + "grad_norm": 0.6269594379373113, + "learning_rate": 0.0009143110175626661, + "loss": 2.4888, + "step": 931 + }, + { + "epoch": 0.21371245127264388, + "grad_norm": 0.5331903873081018, + "learning_rate": 0.0009141030203166256, + "loss": 2.615, + "step": 932 + }, + { + "epoch": 0.21394175647787206, + "grad_norm": 0.5136636992797036, + "learning_rate": 0.0009138947946541291, + "loss": 2.4911, + "step": 933 + }, + { + "epoch": 0.2141710616831002, + "grad_norm": 0.5444544422713049, + "learning_rate": 0.0009136863406900328, + "loss": 2.6708, + "step": 934 + }, + { + "epoch": 0.21440036688832836, + "grad_norm": 0.5320310483291407, + "learning_rate": 0.0009134776585393181, + "loss": 2.5368, + "step": 935 + }, + { + "epoch": 0.21462967209355652, + "grad_norm": 0.5547536960605317, + "learning_rate": 0.0009132687483170929, + "loss": 2.4371, + "step": 936 + }, + { + "epoch": 0.21485897729878467, + "grad_norm": 0.5060142962291608, + "learning_rate": 0.0009130596101385906, + "loss": 2.6906, + "step": 937 + }, + { + "epoch": 0.21508828250401285, + "grad_norm": 0.5401074233411572, + "learning_rate": 0.0009128502441191706, + "loss": 2.6423, + "step": 938 + }, + { + "epoch": 0.215317587709241, + "grad_norm": 0.6335432161804118, + "learning_rate": 0.0009126406503743174, + "loss": 2.6224, + "step": 939 + }, + { + "epoch": 0.21554689291446916, + "grad_norm": 0.5222481800170111, + "learning_rate": 0.000912430829019642, + "loss": 2.6118, + "step": 940 + }, + { + "epoch": 0.2157761981196973, + "grad_norm": 0.7640229004235045, + "learning_rate": 0.0009122207801708802, + "loss": 2.5032, + "step": 941 + }, + { + "epoch": 0.21600550332492546, + "grad_norm": 0.49862227381525803, + "learning_rate": 0.0009120105039438936, + "loss": 2.5482, + "step": 942 + }, + { + "epoch": 0.21623480853015364, + "grad_norm": 0.5688020591874653, + "learning_rate": 0.0009118000004546689, + "loss": 2.6418, + "step": 943 + }, + { + "epoch": 0.2164641137353818, + "grad_norm": 0.5127149268557829, + "learning_rate": 0.0009115892698193189, + "loss": 2.4224, + "step": 944 + }, + { + "epoch": 0.21669341894060995, + "grad_norm": 0.5163176987972456, + "learning_rate": 0.0009113783121540807, + "loss": 2.6322, + "step": 945 + }, + { + "epoch": 0.2169227241458381, + "grad_norm": 0.5368786951257792, + "learning_rate": 0.0009111671275753176, + "loss": 2.5972, + "step": 946 + }, + { + "epoch": 0.21715202935106626, + "grad_norm": 0.5837626705930372, + "learning_rate": 0.0009109557161995172, + "loss": 2.6256, + "step": 947 + }, + { + "epoch": 0.21738133455629444, + "grad_norm": 0.5302500224455239, + "learning_rate": 0.000910744078143293, + "loss": 2.7256, + "step": 948 + }, + { + "epoch": 0.2176106397615226, + "grad_norm": 0.5294311089932505, + "learning_rate": 0.0009105322135233828, + "loss": 2.4725, + "step": 949 + }, + { + "epoch": 0.21783994496675074, + "grad_norm": 0.8470589838551166, + "learning_rate": 0.0009103201224566498, + "loss": 2.6599, + "step": 950 + }, + { + "epoch": 0.2180692501719789, + "grad_norm": 0.5209527048995176, + "learning_rate": 0.0009101078050600821, + "loss": 2.6536, + "step": 951 + }, + { + "epoch": 0.21829855537720708, + "grad_norm": 0.5796596805168186, + "learning_rate": 0.0009098952614507925, + "loss": 2.6234, + "step": 952 + }, + { + "epoch": 0.21852786058243523, + "grad_norm": 0.451958764991328, + "learning_rate": 0.0009096824917460186, + "loss": 2.5795, + "step": 953 + }, + { + "epoch": 0.21875716578766338, + "grad_norm": 0.4784277651375499, + "learning_rate": 0.0009094694960631228, + "loss": 2.5255, + "step": 954 + }, + { + "epoch": 0.21898647099289154, + "grad_norm": 0.5063292201253459, + "learning_rate": 0.0009092562745195921, + "loss": 2.7511, + "step": 955 + }, + { + "epoch": 0.2192157761981197, + "grad_norm": 0.5122986304273726, + "learning_rate": 0.000909042827233038, + "loss": 2.5746, + "step": 956 + }, + { + "epoch": 0.21944508140334787, + "grad_norm": 0.5457757463738323, + "learning_rate": 0.0009088291543211967, + "loss": 2.5409, + "step": 957 + }, + { + "epoch": 0.21967438660857602, + "grad_norm": 0.5039247218218906, + "learning_rate": 0.0009086152559019286, + "loss": 2.5413, + "step": 958 + }, + { + "epoch": 0.21990369181380418, + "grad_norm": 0.48828117188097825, + "learning_rate": 0.0009084011320932188, + "loss": 2.6686, + "step": 959 + }, + { + "epoch": 0.22013299701903233, + "grad_norm": 0.5351606792137646, + "learning_rate": 0.0009081867830131765, + "loss": 2.4472, + "step": 960 + }, + { + "epoch": 0.22036230222426048, + "grad_norm": 0.5147409069178347, + "learning_rate": 0.0009079722087800352, + "loss": 2.4826, + "step": 961 + }, + { + "epoch": 0.22059160742948866, + "grad_norm": 0.5045050827840369, + "learning_rate": 0.0009077574095121526, + "loss": 2.5122, + "step": 962 + }, + { + "epoch": 0.22082091263471681, + "grad_norm": 0.5415820913997286, + "learning_rate": 0.0009075423853280106, + "loss": 2.5975, + "step": 963 + }, + { + "epoch": 0.22105021783994497, + "grad_norm": 0.5544232974656174, + "learning_rate": 0.000907327136346215, + "loss": 2.5216, + "step": 964 + }, + { + "epoch": 0.22127952304517312, + "grad_norm": 0.541819781579045, + "learning_rate": 0.0009071116626854958, + "loss": 2.6886, + "step": 965 + }, + { + "epoch": 0.22150882825040127, + "grad_norm": 0.5167262959710406, + "learning_rate": 0.0009068959644647068, + "loss": 2.6615, + "step": 966 + }, + { + "epoch": 0.22173813345562945, + "grad_norm": 0.5150459211695976, + "learning_rate": 0.0009066800418028256, + "loss": 2.5313, + "step": 967 + }, + { + "epoch": 0.2219674386608576, + "grad_norm": 0.5336663199947982, + "learning_rate": 0.0009064638948189539, + "loss": 2.481, + "step": 968 + }, + { + "epoch": 0.22219674386608576, + "grad_norm": 0.5408920121521198, + "learning_rate": 0.0009062475236323168, + "loss": 2.5807, + "step": 969 + }, + { + "epoch": 0.2224260490713139, + "grad_norm": 0.5273482146916263, + "learning_rate": 0.0009060309283622632, + "loss": 2.4994, + "step": 970 + }, + { + "epoch": 0.22265535427654207, + "grad_norm": 0.47232805981379405, + "learning_rate": 0.0009058141091282656, + "loss": 2.488, + "step": 971 + }, + { + "epoch": 0.22288465948177025, + "grad_norm": 0.4855818823472147, + "learning_rate": 0.0009055970660499199, + "loss": 2.5077, + "step": 972 + }, + { + "epoch": 0.2231139646869984, + "grad_norm": 0.5417324685894963, + "learning_rate": 0.0009053797992469461, + "loss": 2.7115, + "step": 973 + }, + { + "epoch": 0.22334326989222655, + "grad_norm": 0.5349294729164095, + "learning_rate": 0.0009051623088391865, + "loss": 2.8105, + "step": 974 + }, + { + "epoch": 0.2235725750974547, + "grad_norm": 0.5247823767519383, + "learning_rate": 0.0009049445949466078, + "loss": 2.6013, + "step": 975 + }, + { + "epoch": 0.22380188030268286, + "grad_norm": 0.5008178574992427, + "learning_rate": 0.0009047266576892992, + "loss": 2.6316, + "step": 976 + }, + { + "epoch": 0.22403118550791104, + "grad_norm": 0.4964383840423843, + "learning_rate": 0.0009045084971874737, + "loss": 2.4409, + "step": 977 + }, + { + "epoch": 0.2242604907131392, + "grad_norm": 0.5379750601137271, + "learning_rate": 0.0009042901135614671, + "loss": 2.5027, + "step": 978 + }, + { + "epoch": 0.22448979591836735, + "grad_norm": 0.5577625532124341, + "learning_rate": 0.0009040715069317382, + "loss": 2.4434, + "step": 979 + }, + { + "epoch": 0.2247191011235955, + "grad_norm": 0.4781053000142138, + "learning_rate": 0.0009038526774188691, + "loss": 2.5334, + "step": 980 + }, + { + "epoch": 0.22494840632882365, + "grad_norm": 0.5039668307246159, + "learning_rate": 0.0009036336251435648, + "loss": 2.6482, + "step": 981 + }, + { + "epoch": 0.22517771153405183, + "grad_norm": 0.5017793042597786, + "learning_rate": 0.0009034143502266528, + "loss": 2.4849, + "step": 982 + }, + { + "epoch": 0.22540701673928, + "grad_norm": 0.5253246563702937, + "learning_rate": 0.0009031948527890839, + "loss": 2.5671, + "step": 983 + }, + { + "epoch": 0.22563632194450814, + "grad_norm": 0.5745613897116919, + "learning_rate": 0.0009029751329519311, + "loss": 2.7763, + "step": 984 + }, + { + "epoch": 0.2258656271497363, + "grad_norm": 0.5049994988434976, + "learning_rate": 0.000902755190836391, + "loss": 2.62, + "step": 985 + }, + { + "epoch": 0.22609493235496445, + "grad_norm": 0.4764317549785879, + "learning_rate": 0.0009025350265637815, + "loss": 2.4488, + "step": 986 + }, + { + "epoch": 0.22632423756019263, + "grad_norm": 0.6136668051799411, + "learning_rate": 0.0009023146402555442, + "loss": 2.7373, + "step": 987 + }, + { + "epoch": 0.22655354276542078, + "grad_norm": 0.573267137519285, + "learning_rate": 0.0009020940320332425, + "loss": 2.7075, + "step": 988 + }, + { + "epoch": 0.22678284797064893, + "grad_norm": 0.5200485892597959, + "learning_rate": 0.0009018732020185624, + "loss": 2.6355, + "step": 989 + }, + { + "epoch": 0.22701215317587708, + "grad_norm": 0.5517029014530853, + "learning_rate": 0.0009016521503333126, + "loss": 2.5468, + "step": 990 + }, + { + "epoch": 0.22724145838110524, + "grad_norm": 0.5187567700103933, + "learning_rate": 0.0009014308770994235, + "loss": 2.5284, + "step": 991 + }, + { + "epoch": 0.22747076358633342, + "grad_norm": 0.5338278943154022, + "learning_rate": 0.0009012093824389477, + "loss": 2.4849, + "step": 992 + }, + { + "epoch": 0.22770006879156157, + "grad_norm": 0.5213474484846345, + "learning_rate": 0.0009009876664740605, + "loss": 2.6313, + "step": 993 + }, + { + "epoch": 0.22792937399678972, + "grad_norm": 0.737677774653088, + "learning_rate": 0.0009007657293270588, + "loss": 2.5594, + "step": 994 + }, + { + "epoch": 0.22815867920201788, + "grad_norm": 0.49880705645881085, + "learning_rate": 0.0009005435711203618, + "loss": 2.5148, + "step": 995 + }, + { + "epoch": 0.22838798440724603, + "grad_norm": 0.5629653771431715, + "learning_rate": 0.0009003211919765101, + "loss": 2.6256, + "step": 996 + }, + { + "epoch": 0.2286172896124742, + "grad_norm": 0.49859375930866207, + "learning_rate": 0.000900098592018167, + "loss": 2.4286, + "step": 997 + }, + { + "epoch": 0.22884659481770236, + "grad_norm": 0.539892282995032, + "learning_rate": 0.0008998757713681169, + "loss": 2.6096, + "step": 998 + }, + { + "epoch": 0.22907590002293052, + "grad_norm": 0.5240294961587105, + "learning_rate": 0.0008996527301492663, + "loss": 2.7128, + "step": 999 + }, + { + "epoch": 0.22930520522815867, + "grad_norm": 0.4965783387747989, + "learning_rate": 0.0008994294684846431, + "loss": 2.5179, + "step": 1000 + }, + { + "epoch": 0.22953451043338685, + "grad_norm": 0.4971742842645663, + "learning_rate": 0.0008992059864973972, + "loss": 2.5032, + "step": 1001 + }, + { + "epoch": 0.229763815638615, + "grad_norm": 0.5261705297572478, + "learning_rate": 0.0008989822843107994, + "loss": 2.553, + "step": 1002 + }, + { + "epoch": 0.22999312084384316, + "grad_norm": 0.4942009964466662, + "learning_rate": 0.0008987583620482427, + "loss": 2.6587, + "step": 1003 + }, + { + "epoch": 0.2302224260490713, + "grad_norm": 0.5470500020729047, + "learning_rate": 0.0008985342198332407, + "loss": 2.5163, + "step": 1004 + }, + { + "epoch": 0.23045173125429946, + "grad_norm": 0.49731768039026913, + "learning_rate": 0.0008983098577894292, + "loss": 2.6093, + "step": 1005 + }, + { + "epoch": 0.23068103645952764, + "grad_norm": 0.5397997217999787, + "learning_rate": 0.0008980852760405646, + "loss": 2.6671, + "step": 1006 + }, + { + "epoch": 0.2309103416647558, + "grad_norm": 0.5042882522946076, + "learning_rate": 0.0008978604747105246, + "loss": 2.4498, + "step": 1007 + }, + { + "epoch": 0.23113964686998395, + "grad_norm": 0.5252371742467071, + "learning_rate": 0.0008976354539233082, + "loss": 2.5411, + "step": 1008 + }, + { + "epoch": 0.2313689520752121, + "grad_norm": 0.5053049408711577, + "learning_rate": 0.0008974102138030354, + "loss": 2.6499, + "step": 1009 + }, + { + "epoch": 0.23159825728044026, + "grad_norm": 0.5284940365914116, + "learning_rate": 0.0008971847544739468, + "loss": 2.5135, + "step": 1010 + }, + { + "epoch": 0.23182756248566844, + "grad_norm": 0.5943663504458296, + "learning_rate": 0.000896959076060405, + "loss": 2.6329, + "step": 1011 + }, + { + "epoch": 0.2320568676908966, + "grad_norm": 0.5508334099740166, + "learning_rate": 0.0008967331786868919, + "loss": 2.5565, + "step": 1012 + }, + { + "epoch": 0.23228617289612474, + "grad_norm": 0.48899355256932564, + "learning_rate": 0.0008965070624780116, + "loss": 2.5628, + "step": 1013 + }, + { + "epoch": 0.2325154781013529, + "grad_norm": 0.5456480639323239, + "learning_rate": 0.0008962807275584879, + "loss": 2.4011, + "step": 1014 + }, + { + "epoch": 0.23274478330658105, + "grad_norm": 0.5063274528006585, + "learning_rate": 0.0008960541740531658, + "loss": 2.6973, + "step": 1015 + }, + { + "epoch": 0.23297408851180923, + "grad_norm": 0.5160372321241986, + "learning_rate": 0.0008958274020870107, + "loss": 2.6304, + "step": 1016 + }, + { + "epoch": 0.23320339371703738, + "grad_norm": 0.5039441049761458, + "learning_rate": 0.0008956004117851083, + "loss": 2.6694, + "step": 1017 + }, + { + "epoch": 0.23343269892226554, + "grad_norm": 0.50325095179086, + "learning_rate": 0.0008953732032726652, + "loss": 2.4769, + "step": 1018 + }, + { + "epoch": 0.2336620041274937, + "grad_norm": 0.5579252727151873, + "learning_rate": 0.0008951457766750079, + "loss": 2.556, + "step": 1019 + }, + { + "epoch": 0.23389130933272184, + "grad_norm": 0.5259799046213871, + "learning_rate": 0.0008949181321175837, + "loss": 2.6571, + "step": 1020 + }, + { + "epoch": 0.23412061453795002, + "grad_norm": 0.5177437891476427, + "learning_rate": 0.0008946902697259593, + "loss": 2.5354, + "step": 1021 + }, + { + "epoch": 0.23434991974317818, + "grad_norm": 0.5158957646185591, + "learning_rate": 0.0008944621896258225, + "loss": 2.5086, + "step": 1022 + }, + { + "epoch": 0.23457922494840633, + "grad_norm": 0.5201106333192204, + "learning_rate": 0.0008942338919429805, + "loss": 2.6123, + "step": 1023 + }, + { + "epoch": 0.23480853015363448, + "grad_norm": 0.48642162646007747, + "learning_rate": 0.0008940053768033609, + "loss": 2.5356, + "step": 1024 + }, + { + "epoch": 0.23503783535886263, + "grad_norm": 0.4899997776690642, + "learning_rate": 0.0008937766443330113, + "loss": 2.5966, + "step": 1025 + }, + { + "epoch": 0.23526714056409082, + "grad_norm": 0.5615183999627873, + "learning_rate": 0.0008935476946580987, + "loss": 2.4837, + "step": 1026 + }, + { + "epoch": 0.23549644576931897, + "grad_norm": 0.5138603061366842, + "learning_rate": 0.0008933185279049103, + "loss": 2.683, + "step": 1027 + }, + { + "epoch": 0.23572575097454712, + "grad_norm": 0.5420677446149121, + "learning_rate": 0.000893089144199853, + "loss": 2.6164, + "step": 1028 + }, + { + "epoch": 0.23595505617977527, + "grad_norm": 0.555253942105103, + "learning_rate": 0.0008928595436694532, + "loss": 2.6691, + "step": 1029 + }, + { + "epoch": 0.23618436138500343, + "grad_norm": 0.5553336959782843, + "learning_rate": 0.0008926297264403571, + "loss": 2.6005, + "step": 1030 + }, + { + "epoch": 0.2364136665902316, + "grad_norm": 0.6090328975269361, + "learning_rate": 0.0008923996926393305, + "loss": 2.4319, + "step": 1031 + }, + { + "epoch": 0.23664297179545976, + "grad_norm": 0.49865684384909376, + "learning_rate": 0.0008921694423932583, + "loss": 2.4669, + "step": 1032 + }, + { + "epoch": 0.23687227700068791, + "grad_norm": 0.5166166331944503, + "learning_rate": 0.0008919389758291449, + "loss": 2.6791, + "step": 1033 + }, + { + "epoch": 0.23710158220591607, + "grad_norm": 0.5855508999648725, + "learning_rate": 0.0008917082930741145, + "loss": 2.5652, + "step": 1034 + }, + { + "epoch": 0.23733088741114422, + "grad_norm": 0.49363019873474334, + "learning_rate": 0.0008914773942554098, + "loss": 2.655, + "step": 1035 + }, + { + "epoch": 0.2375601926163724, + "grad_norm": 0.6432936139330713, + "learning_rate": 0.0008912462795003932, + "loss": 2.4738, + "step": 1036 + }, + { + "epoch": 0.23778949782160055, + "grad_norm": 0.4808859213891284, + "learning_rate": 0.000891014948936546, + "loss": 2.7105, + "step": 1037 + }, + { + "epoch": 0.2380188030268287, + "grad_norm": 0.43935370038009275, + "learning_rate": 0.0008907834026914686, + "loss": 2.6436, + "step": 1038 + }, + { + "epoch": 0.23824810823205686, + "grad_norm": 0.4963050107157757, + "learning_rate": 0.0008905516408928804, + "loss": 2.7099, + "step": 1039 + }, + { + "epoch": 0.238477413437285, + "grad_norm": 0.5251177996111115, + "learning_rate": 0.0008903196636686197, + "loss": 2.6491, + "step": 1040 + }, + { + "epoch": 0.2387067186425132, + "grad_norm": 0.5759197515923149, + "learning_rate": 0.0008900874711466434, + "loss": 2.5828, + "step": 1041 + }, + { + "epoch": 0.23893602384774135, + "grad_norm": 0.4861852262461765, + "learning_rate": 0.0008898550634550275, + "loss": 2.4383, + "step": 1042 + }, + { + "epoch": 0.2391653290529695, + "grad_norm": 0.5276377965934083, + "learning_rate": 0.0008896224407219666, + "loss": 2.7241, + "step": 1043 + }, + { + "epoch": 0.23939463425819765, + "grad_norm": 0.5300281267479063, + "learning_rate": 0.0008893896030757737, + "loss": 2.6695, + "step": 1044 + }, + { + "epoch": 0.23962393946342583, + "grad_norm": 0.5524665295435216, + "learning_rate": 0.0008891565506448804, + "loss": 2.6149, + "step": 1045 + }, + { + "epoch": 0.239853244668654, + "grad_norm": 0.5421014814915385, + "learning_rate": 0.0008889232835578372, + "loss": 2.5782, + "step": 1046 + }, + { + "epoch": 0.24008254987388214, + "grad_norm": 0.579159155019549, + "learning_rate": 0.0008886898019433122, + "loss": 2.4587, + "step": 1047 + }, + { + "epoch": 0.2403118550791103, + "grad_norm": 0.5187542029158786, + "learning_rate": 0.0008884561059300926, + "loss": 2.4827, + "step": 1048 + }, + { + "epoch": 0.24054116028433845, + "grad_norm": 0.4783819950900938, + "learning_rate": 0.0008882221956470836, + "loss": 2.5858, + "step": 1049 + }, + { + "epoch": 0.24077046548956663, + "grad_norm": 0.542018979850284, + "learning_rate": 0.0008879880712233084, + "loss": 2.545, + "step": 1050 + }, + { + "epoch": 0.24099977069479478, + "grad_norm": 0.4819173036695961, + "learning_rate": 0.0008877537327879086, + "loss": 2.4774, + "step": 1051 + }, + { + "epoch": 0.24122907590002293, + "grad_norm": 0.45930696269158283, + "learning_rate": 0.0008875191804701434, + "loss": 2.5968, + "step": 1052 + }, + { + "epoch": 0.24145838110525109, + "grad_norm": 0.4998085049356983, + "learning_rate": 0.0008872844143993908, + "loss": 2.543, + "step": 1053 + }, + { + "epoch": 0.24168768631047924, + "grad_norm": 0.5140253296165845, + "learning_rate": 0.0008870494347051458, + "loss": 2.6009, + "step": 1054 + }, + { + "epoch": 0.24191699151570742, + "grad_norm": 0.5037619181216958, + "learning_rate": 0.0008868142415170218, + "loss": 2.534, + "step": 1055 + }, + { + "epoch": 0.24214629672093557, + "grad_norm": 0.49669300757681084, + "learning_rate": 0.0008865788349647495, + "loss": 2.5822, + "step": 1056 + }, + { + "epoch": 0.24237560192616373, + "grad_norm": 0.490300024138838, + "learning_rate": 0.0008863432151781781, + "loss": 2.6952, + "step": 1057 + }, + { + "epoch": 0.24260490713139188, + "grad_norm": 0.4957298651622268, + "learning_rate": 0.0008861073822872734, + "loss": 2.5649, + "step": 1058 + }, + { + "epoch": 0.24283421233662003, + "grad_norm": 0.47465606846850383, + "learning_rate": 0.0008858713364221195, + "loss": 2.4574, + "step": 1059 + }, + { + "epoch": 0.2430635175418482, + "grad_norm": 0.4749530394654881, + "learning_rate": 0.0008856350777129178, + "loss": 2.6805, + "step": 1060 + }, + { + "epoch": 0.24329282274707636, + "grad_norm": 0.46454049686182275, + "learning_rate": 0.0008853986062899868, + "loss": 2.5025, + "step": 1061 + }, + { + "epoch": 0.24352212795230452, + "grad_norm": 0.4906128731001337, + "learning_rate": 0.0008851619222837626, + "loss": 2.5863, + "step": 1062 + }, + { + "epoch": 0.24375143315753267, + "grad_norm": 0.50613744770709, + "learning_rate": 0.0008849250258247986, + "loss": 2.5134, + "step": 1063 + }, + { + "epoch": 0.24398073836276082, + "grad_norm": 0.4813469102420131, + "learning_rate": 0.0008846879170437652, + "loss": 2.6177, + "step": 1064 + }, + { + "epoch": 0.244210043567989, + "grad_norm": 0.4686676650248179, + "learning_rate": 0.0008844505960714503, + "loss": 2.4974, + "step": 1065 + }, + { + "epoch": 0.24443934877321716, + "grad_norm": 0.4779800286403916, + "learning_rate": 0.0008842130630387581, + "loss": 2.5809, + "step": 1066 + }, + { + "epoch": 0.2446686539784453, + "grad_norm": 0.4684773708236581, + "learning_rate": 0.0008839753180767108, + "loss": 2.5176, + "step": 1067 + }, + { + "epoch": 0.24489795918367346, + "grad_norm": 0.4691611670023575, + "learning_rate": 0.0008837373613164464, + "loss": 2.4646, + "step": 1068 + }, + { + "epoch": 0.24512726438890162, + "grad_norm": 0.48195796246336153, + "learning_rate": 0.0008834991928892204, + "loss": 2.6203, + "step": 1069 + }, + { + "epoch": 0.2453565695941298, + "grad_norm": 0.4639361311924014, + "learning_rate": 0.000883260812926405, + "loss": 2.6019, + "step": 1070 + }, + { + "epoch": 0.24558587479935795, + "grad_norm": 0.48991299033488817, + "learning_rate": 0.000883022221559489, + "loss": 2.6021, + "step": 1071 + }, + { + "epoch": 0.2458151800045861, + "grad_norm": 0.5422763879292297, + "learning_rate": 0.0008827834189200778, + "loss": 2.5834, + "step": 1072 + }, + { + "epoch": 0.24604448520981426, + "grad_norm": 0.5354848393128536, + "learning_rate": 0.0008825444051398934, + "loss": 2.5458, + "step": 1073 + }, + { + "epoch": 0.2462737904150424, + "grad_norm": 0.5280306279209465, + "learning_rate": 0.000882305180350774, + "loss": 2.6138, + "step": 1074 + }, + { + "epoch": 0.2465030956202706, + "grad_norm": 0.5708039739003191, + "learning_rate": 0.0008820657446846745, + "loss": 2.5907, + "step": 1075 + }, + { + "epoch": 0.24673240082549874, + "grad_norm": 0.5131646093862618, + "learning_rate": 0.0008818260982736661, + "loss": 2.6099, + "step": 1076 + }, + { + "epoch": 0.2469617060307269, + "grad_norm": 0.48077250466470284, + "learning_rate": 0.000881586241249936, + "loss": 2.6029, + "step": 1077 + }, + { + "epoch": 0.24719101123595505, + "grad_norm": 0.4854943807651703, + "learning_rate": 0.0008813461737457877, + "loss": 2.5615, + "step": 1078 + }, + { + "epoch": 0.2474203164411832, + "grad_norm": 0.45968794270189606, + "learning_rate": 0.0008811058958936411, + "loss": 2.5084, + "step": 1079 + }, + { + "epoch": 0.24764962164641138, + "grad_norm": 0.5123374836376756, + "learning_rate": 0.0008808654078260317, + "loss": 2.7263, + "step": 1080 + }, + { + "epoch": 0.24787892685163954, + "grad_norm": 0.4914702841048258, + "learning_rate": 0.000880624709675611, + "loss": 2.5844, + "step": 1081 + }, + { + "epoch": 0.2481082320568677, + "grad_norm": 0.49932382810014997, + "learning_rate": 0.0008803838015751466, + "loss": 2.5884, + "step": 1082 + }, + { + "epoch": 0.24833753726209584, + "grad_norm": 0.49873649322776337, + "learning_rate": 0.000880142683657522, + "loss": 2.5246, + "step": 1083 + }, + { + "epoch": 0.248566842467324, + "grad_norm": 0.4907465211750858, + "learning_rate": 0.0008799013560557359, + "loss": 2.5571, + "step": 1084 + }, + { + "epoch": 0.24879614767255218, + "grad_norm": 0.563148687747315, + "learning_rate": 0.0008796598189029029, + "loss": 2.5124, + "step": 1085 + }, + { + "epoch": 0.24902545287778033, + "grad_norm": 0.5006720035844668, + "learning_rate": 0.0008794180723322536, + "loss": 2.4724, + "step": 1086 + }, + { + "epoch": 0.24925475808300848, + "grad_norm": 0.49641827238921576, + "learning_rate": 0.0008791761164771338, + "loss": 2.6358, + "step": 1087 + }, + { + "epoch": 0.24948406328823663, + "grad_norm": 0.4889936173606407, + "learning_rate": 0.0008789339514710046, + "loss": 2.4548, + "step": 1088 + }, + { + "epoch": 0.2497133684934648, + "grad_norm": 0.5199002003119678, + "learning_rate": 0.0008786915774474424, + "loss": 2.4451, + "step": 1089 + }, + { + "epoch": 0.24994267369869297, + "grad_norm": 0.5234817125911233, + "learning_rate": 0.0008784489945401395, + "loss": 2.5193, + "step": 1090 + }, + { + "epoch": 0.2501719789039211, + "grad_norm": 0.6088853919216173, + "learning_rate": 0.0008782062028829027, + "loss": 2.4918, + "step": 1091 + }, + { + "epoch": 0.2504012841091493, + "grad_norm": 0.5626126708582475, + "learning_rate": 0.0008779632026096544, + "loss": 2.6122, + "step": 1092 + }, + { + "epoch": 0.2506305893143774, + "grad_norm": 0.508402142718814, + "learning_rate": 0.0008777199938544318, + "loss": 2.642, + "step": 1093 + }, + { + "epoch": 0.2508598945196056, + "grad_norm": 0.4723880085860993, + "learning_rate": 0.0008774765767513875, + "loss": 2.465, + "step": 1094 + }, + { + "epoch": 0.25108919972483373, + "grad_norm": 0.5336725860568139, + "learning_rate": 0.0008772329514347883, + "loss": 2.4481, + "step": 1095 + }, + { + "epoch": 0.2513185049300619, + "grad_norm": 0.48420255112945715, + "learning_rate": 0.0008769891180390167, + "loss": 2.4641, + "step": 1096 + }, + { + "epoch": 0.2515478101352901, + "grad_norm": 0.5390196331840846, + "learning_rate": 0.0008767450766985694, + "loss": 2.4694, + "step": 1097 + }, + { + "epoch": 0.2517771153405182, + "grad_norm": 0.49615558332117154, + "learning_rate": 0.000876500827548058, + "loss": 2.638, + "step": 1098 + }, + { + "epoch": 0.2520064205457464, + "grad_norm": 0.4982617077768434, + "learning_rate": 0.0008762563707222086, + "loss": 2.6133, + "step": 1099 + }, + { + "epoch": 0.2522357257509745, + "grad_norm": 0.4697030763231646, + "learning_rate": 0.0008760117063558621, + "loss": 2.5581, + "step": 1100 + }, + { + "epoch": 0.2524650309562027, + "grad_norm": 0.48782918194309155, + "learning_rate": 0.0008757668345839738, + "loss": 2.5719, + "step": 1101 + }, + { + "epoch": 0.2526943361614309, + "grad_norm": 0.4520224024503562, + "learning_rate": 0.0008755217555416131, + "loss": 2.6474, + "step": 1102 + }, + { + "epoch": 0.252923641366659, + "grad_norm": 0.46900524429103424, + "learning_rate": 0.0008752764693639638, + "loss": 2.5279, + "step": 1103 + }, + { + "epoch": 0.2531529465718872, + "grad_norm": 0.47496852611029156, + "learning_rate": 0.0008750309761863246, + "loss": 2.7758, + "step": 1104 + }, + { + "epoch": 0.2533822517771153, + "grad_norm": 0.4617813265166214, + "learning_rate": 0.0008747852761441078, + "loss": 2.5684, + "step": 1105 + }, + { + "epoch": 0.2536115569823435, + "grad_norm": 0.4935818265619656, + "learning_rate": 0.0008745393693728395, + "loss": 2.4454, + "step": 1106 + }, + { + "epoch": 0.2538408621875717, + "grad_norm": 0.4534204409758379, + "learning_rate": 0.0008742932560081607, + "loss": 2.5445, + "step": 1107 + }, + { + "epoch": 0.2540701673927998, + "grad_norm": 0.48104075556177256, + "learning_rate": 0.0008740469361858255, + "loss": 2.527, + "step": 1108 + }, + { + "epoch": 0.254299472598028, + "grad_norm": 0.49017885558172, + "learning_rate": 0.0008738004100417025, + "loss": 2.4788, + "step": 1109 + }, + { + "epoch": 0.2545287778032561, + "grad_norm": 0.5086096680884407, + "learning_rate": 0.0008735536777117739, + "loss": 2.5219, + "step": 1110 + }, + { + "epoch": 0.2547580830084843, + "grad_norm": 0.4743716989843402, + "learning_rate": 0.0008733067393321355, + "loss": 2.6753, + "step": 1111 + }, + { + "epoch": 0.2549873882137125, + "grad_norm": 0.5324106224921763, + "learning_rate": 0.0008730595950389968, + "loss": 2.6529, + "step": 1112 + }, + { + "epoch": 0.2552166934189406, + "grad_norm": 0.49458383314759397, + "learning_rate": 0.000872812244968681, + "loss": 2.5619, + "step": 1113 + }, + { + "epoch": 0.2554459986241688, + "grad_norm": 0.498129857057272, + "learning_rate": 0.0008725646892576249, + "loss": 2.6742, + "step": 1114 + }, + { + "epoch": 0.2556753038293969, + "grad_norm": 0.4849206870682493, + "learning_rate": 0.0008723169280423783, + "loss": 2.6567, + "step": 1115 + }, + { + "epoch": 0.2559046090346251, + "grad_norm": 0.49036166851292057, + "learning_rate": 0.0008720689614596049, + "loss": 2.6391, + "step": 1116 + }, + { + "epoch": 0.25613391423985327, + "grad_norm": 0.4609416683876693, + "learning_rate": 0.0008718207896460811, + "loss": 2.6237, + "step": 1117 + }, + { + "epoch": 0.2563632194450814, + "grad_norm": 0.4961892837311079, + "learning_rate": 0.0008715724127386971, + "loss": 2.5858, + "step": 1118 + }, + { + "epoch": 0.2565925246503096, + "grad_norm": 0.4940233738481095, + "learning_rate": 0.0008713238308744557, + "loss": 2.6878, + "step": 1119 + }, + { + "epoch": 0.2568218298555377, + "grad_norm": 0.5202702237019, + "learning_rate": 0.0008710750441904733, + "loss": 2.6289, + "step": 1120 + }, + { + "epoch": 0.2570511350607659, + "grad_norm": 0.484533386809111, + "learning_rate": 0.0008708260528239789, + "loss": 2.4638, + "step": 1121 + }, + { + "epoch": 0.25728044026599406, + "grad_norm": 0.5049864110107426, + "learning_rate": 0.0008705768569123141, + "loss": 2.5153, + "step": 1122 + }, + { + "epoch": 0.2575097454712222, + "grad_norm": 0.617813980789451, + "learning_rate": 0.000870327456592934, + "loss": 2.5645, + "step": 1123 + }, + { + "epoch": 0.25773905067645037, + "grad_norm": 0.49965358624221207, + "learning_rate": 0.0008700778520034062, + "loss": 2.6124, + "step": 1124 + }, + { + "epoch": 0.2579683558816785, + "grad_norm": 0.47655373452127353, + "learning_rate": 0.0008698280432814107, + "loss": 2.579, + "step": 1125 + }, + { + "epoch": 0.25819766108690667, + "grad_norm": 0.5211856403995138, + "learning_rate": 0.0008695780305647405, + "loss": 2.633, + "step": 1126 + }, + { + "epoch": 0.25842696629213485, + "grad_norm": 0.5017082136430949, + "learning_rate": 0.000869327813991301, + "loss": 2.5808, + "step": 1127 + }, + { + "epoch": 0.258656271497363, + "grad_norm": 0.5120622624466028, + "learning_rate": 0.0008690773936991098, + "loss": 2.6286, + "step": 1128 + }, + { + "epoch": 0.25888557670259116, + "grad_norm": 0.4456092287132823, + "learning_rate": 0.0008688267698262971, + "loss": 2.6133, + "step": 1129 + }, + { + "epoch": 0.2591148819078193, + "grad_norm": 0.4717211157006013, + "learning_rate": 0.0008685759425111055, + "loss": 2.5746, + "step": 1130 + }, + { + "epoch": 0.25934418711304746, + "grad_norm": 0.49478591052321796, + "learning_rate": 0.0008683249118918894, + "loss": 2.4677, + "step": 1131 + }, + { + "epoch": 0.25957349231827564, + "grad_norm": 0.4949558462481698, + "learning_rate": 0.0008680736781071159, + "loss": 2.6614, + "step": 1132 + }, + { + "epoch": 0.25980279752350377, + "grad_norm": 0.5571247518357129, + "learning_rate": 0.0008678222412953637, + "loss": 2.4424, + "step": 1133 + }, + { + "epoch": 0.26003210272873195, + "grad_norm": 0.4950089162465125, + "learning_rate": 0.0008675706015953237, + "loss": 2.6619, + "step": 1134 + }, + { + "epoch": 0.2602614079339601, + "grad_norm": 0.5375659887176111, + "learning_rate": 0.0008673187591457987, + "loss": 2.7488, + "step": 1135 + }, + { + "epoch": 0.26049071313918826, + "grad_norm": 0.5373483645758782, + "learning_rate": 0.0008670667140857032, + "loss": 2.7848, + "step": 1136 + }, + { + "epoch": 0.26072001834441644, + "grad_norm": 0.4943424638199161, + "learning_rate": 0.0008668144665540639, + "loss": 2.5763, + "step": 1137 + }, + { + "epoch": 0.26094932354964456, + "grad_norm": 0.5109674707122965, + "learning_rate": 0.0008665620166900184, + "loss": 2.6209, + "step": 1138 + }, + { + "epoch": 0.26117862875487274, + "grad_norm": 0.502006320136588, + "learning_rate": 0.0008663093646328167, + "loss": 2.5498, + "step": 1139 + }, + { + "epoch": 0.26140793396010087, + "grad_norm": 0.4639371666050806, + "learning_rate": 0.0008660565105218196, + "loss": 2.5837, + "step": 1140 + }, + { + "epoch": 0.26163723916532905, + "grad_norm": 0.5151227024664373, + "learning_rate": 0.0008658034544965003, + "loss": 2.6324, + "step": 1141 + }, + { + "epoch": 0.26186654437055723, + "grad_norm": 0.4931657893920925, + "learning_rate": 0.0008655501966964423, + "loss": 2.5459, + "step": 1142 + }, + { + "epoch": 0.26209584957578536, + "grad_norm": 0.482863426191544, + "learning_rate": 0.0008652967372613412, + "loss": 2.5568, + "step": 1143 + }, + { + "epoch": 0.26232515478101354, + "grad_norm": 0.4919893611789088, + "learning_rate": 0.0008650430763310034, + "loss": 2.5722, + "step": 1144 + }, + { + "epoch": 0.26255445998624166, + "grad_norm": 0.5134170997734894, + "learning_rate": 0.0008647892140453466, + "loss": 2.6177, + "step": 1145 + }, + { + "epoch": 0.26278376519146984, + "grad_norm": 0.46508880603955516, + "learning_rate": 0.0008645351505443997, + "loss": 2.5842, + "step": 1146 + }, + { + "epoch": 0.263013070396698, + "grad_norm": 0.5197524699447577, + "learning_rate": 0.0008642808859683021, + "loss": 2.5677, + "step": 1147 + }, + { + "epoch": 0.26324237560192615, + "grad_norm": 0.530405228218678, + "learning_rate": 0.0008640264204573048, + "loss": 2.4295, + "step": 1148 + }, + { + "epoch": 0.26347168080715433, + "grad_norm": 0.5148344448409018, + "learning_rate": 0.0008637717541517689, + "loss": 2.5707, + "step": 1149 + }, + { + "epoch": 0.26370098601238245, + "grad_norm": 0.5193504344466714, + "learning_rate": 0.0008635168871921669, + "loss": 2.6809, + "step": 1150 + }, + { + "epoch": 0.26393029121761064, + "grad_norm": 0.48928385986932144, + "learning_rate": 0.0008632618197190816, + "loss": 2.6269, + "step": 1151 + }, + { + "epoch": 0.2641595964228388, + "grad_norm": 0.47431776469562065, + "learning_rate": 0.0008630065518732067, + "loss": 2.5606, + "step": 1152 + }, + { + "epoch": 0.26438890162806694, + "grad_norm": 0.45513158275387317, + "learning_rate": 0.0008627510837953458, + "loss": 2.6326, + "step": 1153 + }, + { + "epoch": 0.2646182068332951, + "grad_norm": 0.47621137600237873, + "learning_rate": 0.0008624954156264137, + "loss": 2.4948, + "step": 1154 + }, + { + "epoch": 0.26484751203852325, + "grad_norm": 0.4728461004283487, + "learning_rate": 0.0008622395475074355, + "loss": 2.5271, + "step": 1155 + }, + { + "epoch": 0.26507681724375143, + "grad_norm": 0.45242595030714783, + "learning_rate": 0.0008619834795795459, + "loss": 2.5382, + "step": 1156 + }, + { + "epoch": 0.2653061224489796, + "grad_norm": 0.48007710174222706, + "learning_rate": 0.0008617272119839903, + "loss": 2.4087, + "step": 1157 + }, + { + "epoch": 0.26553542765420773, + "grad_norm": 0.4746142676259672, + "learning_rate": 0.0008614707448621244, + "loss": 2.6802, + "step": 1158 + }, + { + "epoch": 0.2657647328594359, + "grad_norm": 0.500618338137613, + "learning_rate": 0.0008612140783554136, + "loss": 2.4403, + "step": 1159 + }, + { + "epoch": 0.2659940380646641, + "grad_norm": 0.4962357480839508, + "learning_rate": 0.0008609572126054336, + "loss": 2.531, + "step": 1160 + }, + { + "epoch": 0.2662233432698922, + "grad_norm": 0.5062076370637191, + "learning_rate": 0.0008607001477538696, + "loss": 2.5725, + "step": 1161 + }, + { + "epoch": 0.2664526484751204, + "grad_norm": 0.4970356205581144, + "learning_rate": 0.0008604428839425171, + "loss": 2.5252, + "step": 1162 + }, + { + "epoch": 0.2666819536803485, + "grad_norm": 0.47514844299343717, + "learning_rate": 0.0008601854213132807, + "loss": 2.67, + "step": 1163 + }, + { + "epoch": 0.2669112588855767, + "grad_norm": 0.5344136648728673, + "learning_rate": 0.0008599277600081757, + "loss": 2.6533, + "step": 1164 + }, + { + "epoch": 0.2671405640908049, + "grad_norm": 0.5293819850056689, + "learning_rate": 0.0008596699001693256, + "loss": 2.5111, + "step": 1165 + }, + { + "epoch": 0.267369869296033, + "grad_norm": 0.46962064726276515, + "learning_rate": 0.0008594118419389648, + "loss": 2.5876, + "step": 1166 + }, + { + "epoch": 0.2675991745012612, + "grad_norm": 0.41505991875829235, + "learning_rate": 0.000859153585459436, + "loss": 2.5199, + "step": 1167 + }, + { + "epoch": 0.2678284797064893, + "grad_norm": 0.46107295689654254, + "learning_rate": 0.0008588951308731921, + "loss": 2.4707, + "step": 1168 + }, + { + "epoch": 0.2680577849117175, + "grad_norm": 0.47681205307579866, + "learning_rate": 0.0008586364783227949, + "loss": 2.6184, + "step": 1169 + }, + { + "epoch": 0.2682870901169457, + "grad_norm": 0.4576989662157205, + "learning_rate": 0.0008583776279509152, + "loss": 2.3935, + "step": 1170 + }, + { + "epoch": 0.2685163953221738, + "grad_norm": 0.5233279462730294, + "learning_rate": 0.0008581185799003332, + "loss": 2.6635, + "step": 1171 + }, + { + "epoch": 0.268745700527402, + "grad_norm": 0.4872758195794469, + "learning_rate": 0.0008578593343139383, + "loss": 2.5685, + "step": 1172 + }, + { + "epoch": 0.2689750057326301, + "grad_norm": 0.513573095798337, + "learning_rate": 0.0008575998913347283, + "loss": 2.5576, + "step": 1173 + }, + { + "epoch": 0.2692043109378583, + "grad_norm": 0.48607547215003566, + "learning_rate": 0.0008573402511058104, + "loss": 2.5018, + "step": 1174 + }, + { + "epoch": 0.2694336161430865, + "grad_norm": 0.5051018070200041, + "learning_rate": 0.0008570804137704004, + "loss": 2.4948, + "step": 1175 + }, + { + "epoch": 0.2696629213483146, + "grad_norm": 0.4882405348874768, + "learning_rate": 0.0008568203794718228, + "loss": 2.4663, + "step": 1176 + }, + { + "epoch": 0.2698922265535428, + "grad_norm": 0.531810103347527, + "learning_rate": 0.0008565601483535108, + "loss": 2.5372, + "step": 1177 + }, + { + "epoch": 0.2701215317587709, + "grad_norm": 0.4647151659982516, + "learning_rate": 0.0008562997205590062, + "loss": 2.4085, + "step": 1178 + }, + { + "epoch": 0.2703508369639991, + "grad_norm": 0.5079274814930771, + "learning_rate": 0.0008560390962319591, + "loss": 2.3617, + "step": 1179 + }, + { + "epoch": 0.27058014216922727, + "grad_norm": 0.531648888531459, + "learning_rate": 0.0008557782755161281, + "loss": 2.6464, + "step": 1180 + }, + { + "epoch": 0.2708094473744554, + "grad_norm": 0.502930483920416, + "learning_rate": 0.0008555172585553804, + "loss": 2.6057, + "step": 1181 + }, + { + "epoch": 0.2710387525796836, + "grad_norm": 0.4992453990229745, + "learning_rate": 0.000855256045493691, + "loss": 2.5133, + "step": 1182 + }, + { + "epoch": 0.2712680577849117, + "grad_norm": 0.45893417084869076, + "learning_rate": 0.0008549946364751435, + "loss": 2.4142, + "step": 1183 + }, + { + "epoch": 0.2714973629901399, + "grad_norm": 0.5102066869472344, + "learning_rate": 0.000854733031643929, + "loss": 2.5742, + "step": 1184 + }, + { + "epoch": 0.27172666819536806, + "grad_norm": 0.44662980133574903, + "learning_rate": 0.0008544712311443475, + "loss": 2.5683, + "step": 1185 + }, + { + "epoch": 0.2719559734005962, + "grad_norm": 0.4488783075298677, + "learning_rate": 0.0008542092351208057, + "loss": 2.5661, + "step": 1186 + }, + { + "epoch": 0.27218527860582437, + "grad_norm": 0.4783189069530046, + "learning_rate": 0.0008539470437178196, + "loss": 2.5487, + "step": 1187 + }, + { + "epoch": 0.2724145838110525, + "grad_norm": 0.45360340967562923, + "learning_rate": 0.0008536846570800117, + "loss": 2.5158, + "step": 1188 + }, + { + "epoch": 0.27264388901628067, + "grad_norm": 0.47960514199011517, + "learning_rate": 0.000853422075352113, + "loss": 2.5444, + "step": 1189 + }, + { + "epoch": 0.27287319422150885, + "grad_norm": 0.5043239628366347, + "learning_rate": 0.0008531592986789616, + "loss": 2.5591, + "step": 1190 + }, + { + "epoch": 0.273102499426737, + "grad_norm": 0.4598895880335181, + "learning_rate": 0.0008528963272055035, + "loss": 2.4888, + "step": 1191 + }, + { + "epoch": 0.27333180463196516, + "grad_norm": 0.452808275199236, + "learning_rate": 0.0008526331610767919, + "loss": 2.5536, + "step": 1192 + }, + { + "epoch": 0.2735611098371933, + "grad_norm": 0.5293615237415722, + "learning_rate": 0.0008523698004379877, + "loss": 2.6786, + "step": 1193 + }, + { + "epoch": 0.27379041504242146, + "grad_norm": 0.48863718735892275, + "learning_rate": 0.0008521062454343586, + "loss": 2.4642, + "step": 1194 + }, + { + "epoch": 0.27401972024764965, + "grad_norm": 0.45640242686952726, + "learning_rate": 0.00085184249621128, + "loss": 2.5907, + "step": 1195 + }, + { + "epoch": 0.27424902545287777, + "grad_norm": 0.4883112585947825, + "learning_rate": 0.0008515785529142339, + "loss": 2.6752, + "step": 1196 + }, + { + "epoch": 0.27447833065810595, + "grad_norm": 0.5506246993003119, + "learning_rate": 0.0008513144156888101, + "loss": 2.6242, + "step": 1197 + }, + { + "epoch": 0.2747076358633341, + "grad_norm": 0.4728631927145634, + "learning_rate": 0.0008510500846807045, + "loss": 2.5216, + "step": 1198 + }, + { + "epoch": 0.27493694106856226, + "grad_norm": 0.44281701836616083, + "learning_rate": 0.0008507855600357207, + "loss": 2.6359, + "step": 1199 + }, + { + "epoch": 0.27516624627379044, + "grad_norm": 0.4912393375234709, + "learning_rate": 0.0008505208418997684, + "loss": 2.5605, + "step": 1200 + }, + { + "epoch": 0.27539555147901856, + "grad_norm": 0.4575599206652079, + "learning_rate": 0.0008502559304188644, + "loss": 2.6283, + "step": 1201 + }, + { + "epoch": 0.27562485668424674, + "grad_norm": 0.4946302335851822, + "learning_rate": 0.0008499908257391324, + "loss": 2.6109, + "step": 1202 + }, + { + "epoch": 0.27585416188947487, + "grad_norm": 0.4786671385596687, + "learning_rate": 0.0008497255280068019, + "loss": 2.5885, + "step": 1203 + }, + { + "epoch": 0.27608346709470305, + "grad_norm": 0.4750425215095902, + "learning_rate": 0.0008494600373682098, + "loss": 2.6358, + "step": 1204 + }, + { + "epoch": 0.27631277229993123, + "grad_norm": 0.46044728793722567, + "learning_rate": 0.0008491943539697986, + "loss": 2.4136, + "step": 1205 + }, + { + "epoch": 0.27654207750515936, + "grad_norm": 0.5233321329623363, + "learning_rate": 0.0008489284779581177, + "loss": 2.5106, + "step": 1206 + }, + { + "epoch": 0.27677138271038754, + "grad_norm": 0.5022125908616752, + "learning_rate": 0.0008486624094798226, + "loss": 2.4979, + "step": 1207 + }, + { + "epoch": 0.27700068791561566, + "grad_norm": 0.49704378246481323, + "learning_rate": 0.0008483961486816747, + "loss": 2.5673, + "step": 1208 + }, + { + "epoch": 0.27722999312084384, + "grad_norm": 0.5344257676941607, + "learning_rate": 0.0008481296957105417, + "loss": 2.7492, + "step": 1209 + }, + { + "epoch": 0.277459298326072, + "grad_norm": 0.47916956266474997, + "learning_rate": 0.0008478630507133973, + "loss": 2.5225, + "step": 1210 + }, + { + "epoch": 0.27768860353130015, + "grad_norm": 0.4833466538287842, + "learning_rate": 0.0008475962138373213, + "loss": 2.5942, + "step": 1211 + }, + { + "epoch": 0.27791790873652833, + "grad_norm": 0.5005413376692454, + "learning_rate": 0.0008473291852294987, + "loss": 2.4629, + "step": 1212 + }, + { + "epoch": 0.27814721394175645, + "grad_norm": 0.44936853127879994, + "learning_rate": 0.0008470619650372211, + "loss": 2.5837, + "step": 1213 + }, + { + "epoch": 0.27837651914698464, + "grad_norm": 0.4893701967471812, + "learning_rate": 0.0008467945534078852, + "loss": 2.6035, + "step": 1214 + }, + { + "epoch": 0.2786058243522128, + "grad_norm": 0.4832812322690823, + "learning_rate": 0.0008465269504889934, + "loss": 2.7212, + "step": 1215 + }, + { + "epoch": 0.27883512955744094, + "grad_norm": 0.4980594712024276, + "learning_rate": 0.0008462591564281539, + "loss": 2.5155, + "step": 1216 + }, + { + "epoch": 0.2790644347626691, + "grad_norm": 0.483655450186786, + "learning_rate": 0.0008459911713730799, + "loss": 2.5588, + "step": 1217 + }, + { + "epoch": 0.27929373996789725, + "grad_norm": 0.49258710510241066, + "learning_rate": 0.0008457229954715902, + "loss": 2.5235, + "step": 1218 + }, + { + "epoch": 0.27952304517312543, + "grad_norm": 0.509752550000836, + "learning_rate": 0.0008454546288716089, + "loss": 2.6781, + "step": 1219 + }, + { + "epoch": 0.2797523503783536, + "grad_norm": 0.4584315780887492, + "learning_rate": 0.0008451860717211652, + "loss": 2.7229, + "step": 1220 + }, + { + "epoch": 0.27998165558358173, + "grad_norm": 0.5375971312769982, + "learning_rate": 0.0008449173241683935, + "loss": 2.5535, + "step": 1221 + }, + { + "epoch": 0.2802109607888099, + "grad_norm": 0.48104900207522083, + "learning_rate": 0.0008446483863615329, + "loss": 2.6391, + "step": 1222 + }, + { + "epoch": 0.28044026599403804, + "grad_norm": 0.5289462829006798, + "learning_rate": 0.0008443792584489281, + "loss": 2.6218, + "step": 1223 + }, + { + "epoch": 0.2806695711992662, + "grad_norm": 0.4828469349757339, + "learning_rate": 0.0008441099405790278, + "loss": 2.5669, + "step": 1224 + }, + { + "epoch": 0.2808988764044944, + "grad_norm": 0.4451710020019788, + "learning_rate": 0.0008438404329003863, + "loss": 2.6329, + "step": 1225 + }, + { + "epoch": 0.2811281816097225, + "grad_norm": 0.4582297249184695, + "learning_rate": 0.000843570735561662, + "loss": 2.5133, + "step": 1226 + }, + { + "epoch": 0.2813574868149507, + "grad_norm": 0.48384748079623774, + "learning_rate": 0.0008433008487116183, + "loss": 2.624, + "step": 1227 + }, + { + "epoch": 0.28158679202017883, + "grad_norm": 0.46233608068105775, + "learning_rate": 0.000843030772499123, + "loss": 2.5637, + "step": 1228 + }, + { + "epoch": 0.281816097225407, + "grad_norm": 0.4755563149222165, + "learning_rate": 0.0008427605070731481, + "loss": 2.4905, + "step": 1229 + }, + { + "epoch": 0.2820454024306352, + "grad_norm": 0.4808264941922298, + "learning_rate": 0.0008424900525827706, + "loss": 2.4461, + "step": 1230 + }, + { + "epoch": 0.2822747076358633, + "grad_norm": 0.494963325122324, + "learning_rate": 0.0008422194091771708, + "loss": 2.7394, + "step": 1231 + }, + { + "epoch": 0.2825040128410915, + "grad_norm": 0.4547314298581302, + "learning_rate": 0.0008419485770056344, + "loss": 2.6251, + "step": 1232 + }, + { + "epoch": 0.2827333180463196, + "grad_norm": 0.4364214682946773, + "learning_rate": 0.0008416775562175503, + "loss": 2.4826, + "step": 1233 + }, + { + "epoch": 0.2829626232515478, + "grad_norm": 0.44773248835781765, + "learning_rate": 0.0008414063469624118, + "loss": 2.4605, + "step": 1234 + }, + { + "epoch": 0.283191928456776, + "grad_norm": 0.4498446032116315, + "learning_rate": 0.000841134949389816, + "loss": 2.6783, + "step": 1235 + }, + { + "epoch": 0.2834212336620041, + "grad_norm": 0.4480142856242567, + "learning_rate": 0.0008408633636494642, + "loss": 2.4421, + "step": 1236 + }, + { + "epoch": 0.2836505388672323, + "grad_norm": 0.48906912190328755, + "learning_rate": 0.0008405915898911611, + "loss": 2.6741, + "step": 1237 + }, + { + "epoch": 0.2838798440724604, + "grad_norm": 0.4908280881901036, + "learning_rate": 0.0008403196282648156, + "loss": 2.4174, + "step": 1238 + }, + { + "epoch": 0.2841091492776886, + "grad_norm": 0.43423159243689047, + "learning_rate": 0.0008400474789204396, + "loss": 2.5422, + "step": 1239 + }, + { + "epoch": 0.2843384544829168, + "grad_norm": 0.5055783504981841, + "learning_rate": 0.0008397751420081488, + "loss": 2.4972, + "step": 1240 + }, + { + "epoch": 0.2845677596881449, + "grad_norm": 0.4535038644999793, + "learning_rate": 0.0008395026176781626, + "loss": 2.6048, + "step": 1241 + }, + { + "epoch": 0.2847970648933731, + "grad_norm": 0.4671768999470344, + "learning_rate": 0.0008392299060808035, + "loss": 2.5307, + "step": 1242 + }, + { + "epoch": 0.2850263700986012, + "grad_norm": 0.4634391163103519, + "learning_rate": 0.0008389570073664976, + "loss": 2.5395, + "step": 1243 + }, + { + "epoch": 0.2852556753038294, + "grad_norm": 0.46737711167334406, + "learning_rate": 0.0008386839216857738, + "loss": 2.5853, + "step": 1244 + }, + { + "epoch": 0.2854849805090576, + "grad_norm": 0.5071586618530647, + "learning_rate": 0.0008384106491892642, + "loss": 2.4182, + "step": 1245 + }, + { + "epoch": 0.2857142857142857, + "grad_norm": 0.45642946868140283, + "learning_rate": 0.0008381371900277045, + "loss": 2.6378, + "step": 1246 + }, + { + "epoch": 0.2859435909195139, + "grad_norm": 0.4822591355904201, + "learning_rate": 0.0008378635443519327, + "loss": 2.5014, + "step": 1247 + }, + { + "epoch": 0.286172896124742, + "grad_norm": 0.5471517398844807, + "learning_rate": 0.0008375897123128897, + "loss": 2.6363, + "step": 1248 + }, + { + "epoch": 0.2864022013299702, + "grad_norm": 0.46028724779549485, + "learning_rate": 0.0008373156940616199, + "loss": 2.5448, + "step": 1249 + }, + { + "epoch": 0.28663150653519837, + "grad_norm": 0.4956532494955046, + "learning_rate": 0.0008370414897492696, + "loss": 2.5892, + "step": 1250 + }, + { + "epoch": 0.2868608117404265, + "grad_norm": 0.4958472869375327, + "learning_rate": 0.0008367670995270882, + "loss": 2.6121, + "step": 1251 + }, + { + "epoch": 0.28709011694565467, + "grad_norm": 0.4486767794758269, + "learning_rate": 0.0008364925235464274, + "loss": 2.5902, + "step": 1252 + }, + { + "epoch": 0.28731942215088285, + "grad_norm": 0.4384232877969696, + "learning_rate": 0.0008362177619587416, + "loss": 2.5059, + "step": 1253 + }, + { + "epoch": 0.287548727356111, + "grad_norm": 0.5082029370210748, + "learning_rate": 0.0008359428149155875, + "loss": 2.6585, + "step": 1254 + }, + { + "epoch": 0.28777803256133916, + "grad_norm": 0.4647030868361388, + "learning_rate": 0.0008356676825686238, + "loss": 2.4757, + "step": 1255 + }, + { + "epoch": 0.2880073377665673, + "grad_norm": 0.4473912018740268, + "learning_rate": 0.0008353923650696118, + "loss": 2.5804, + "step": 1256 + }, + { + "epoch": 0.28823664297179546, + "grad_norm": 0.445949600020368, + "learning_rate": 0.0008351168625704147, + "loss": 2.5356, + "step": 1257 + }, + { + "epoch": 0.28846594817702365, + "grad_norm": 0.45678754331695715, + "learning_rate": 0.0008348411752229982, + "loss": 2.5536, + "step": 1258 + }, + { + "epoch": 0.28869525338225177, + "grad_norm": 0.4495464884108481, + "learning_rate": 0.0008345653031794292, + "loss": 2.6279, + "step": 1259 + }, + { + "epoch": 0.28892455858747995, + "grad_norm": 0.45397108223100546, + "learning_rate": 0.000834289246591877, + "loss": 2.6447, + "step": 1260 + }, + { + "epoch": 0.2891538637927081, + "grad_norm": 0.482127469317215, + "learning_rate": 0.0008340130056126125, + "loss": 2.5702, + "step": 1261 + }, + { + "epoch": 0.28938316899793626, + "grad_norm": 0.45775977746693314, + "learning_rate": 0.0008337365803940084, + "loss": 2.6205, + "step": 1262 + }, + { + "epoch": 0.28961247420316444, + "grad_norm": 0.43339110042809237, + "learning_rate": 0.0008334599710885394, + "loss": 2.554, + "step": 1263 + }, + { + "epoch": 0.28984177940839256, + "grad_norm": 0.4992739456013745, + "learning_rate": 0.0008331831778487807, + "loss": 2.4613, + "step": 1264 + }, + { + "epoch": 0.29007108461362074, + "grad_norm": 0.4703768157877051, + "learning_rate": 0.0008329062008274098, + "loss": 2.5355, + "step": 1265 + }, + { + "epoch": 0.29030038981884887, + "grad_norm": 0.4771560096519856, + "learning_rate": 0.0008326290401772056, + "loss": 2.5527, + "step": 1266 + }, + { + "epoch": 0.29052969502407705, + "grad_norm": 0.5072234627194462, + "learning_rate": 0.000832351696051048, + "loss": 2.6118, + "step": 1267 + }, + { + "epoch": 0.29075900022930523, + "grad_norm": 0.4706104980811716, + "learning_rate": 0.0008320741686019179, + "loss": 2.5964, + "step": 1268 + }, + { + "epoch": 0.29098830543453336, + "grad_norm": 0.4935028344410878, + "learning_rate": 0.000831796457982898, + "loss": 2.6756, + "step": 1269 + }, + { + "epoch": 0.29121761063976154, + "grad_norm": 0.4688998604595926, + "learning_rate": 0.0008315185643471713, + "loss": 2.4847, + "step": 1270 + }, + { + "epoch": 0.29144691584498966, + "grad_norm": 0.49108347943374514, + "learning_rate": 0.0008312404878480222, + "loss": 2.5359, + "step": 1271 + }, + { + "epoch": 0.29167622105021784, + "grad_norm": 0.5220715365534428, + "learning_rate": 0.0008309622286388356, + "loss": 2.6677, + "step": 1272 + }, + { + "epoch": 0.291905526255446, + "grad_norm": 0.48257706881535034, + "learning_rate": 0.0008306837868730979, + "loss": 2.5522, + "step": 1273 + }, + { + "epoch": 0.29213483146067415, + "grad_norm": 0.48271436014368235, + "learning_rate": 0.0008304051627043953, + "loss": 2.6932, + "step": 1274 + }, + { + "epoch": 0.29236413666590233, + "grad_norm": 0.44329634246525484, + "learning_rate": 0.0008301263562864152, + "loss": 2.5706, + "step": 1275 + }, + { + "epoch": 0.29259344187113046, + "grad_norm": 0.4425880278072451, + "learning_rate": 0.0008298473677729452, + "loss": 2.4903, + "step": 1276 + }, + { + "epoch": 0.29282274707635864, + "grad_norm": 0.5494082673292668, + "learning_rate": 0.0008295681973178737, + "loss": 2.5368, + "step": 1277 + }, + { + "epoch": 0.2930520522815868, + "grad_norm": 0.4720440761489996, + "learning_rate": 0.0008292888450751892, + "loss": 2.5001, + "step": 1278 + }, + { + "epoch": 0.29328135748681494, + "grad_norm": 0.45452625879638303, + "learning_rate": 0.0008290093111989804, + "loss": 2.602, + "step": 1279 + }, + { + "epoch": 0.2935106626920431, + "grad_norm": 0.4647155316302802, + "learning_rate": 0.0008287295958434364, + "loss": 2.5923, + "step": 1280 + }, + { + "epoch": 0.29373996789727125, + "grad_norm": 0.4717134399276198, + "learning_rate": 0.0008284496991628465, + "loss": 2.4243, + "step": 1281 + }, + { + "epoch": 0.29396927310249943, + "grad_norm": 0.4410078224490797, + "learning_rate": 0.0008281696213115995, + "loss": 2.4968, + "step": 1282 + }, + { + "epoch": 0.2941985783077276, + "grad_norm": 0.4651688306691343, + "learning_rate": 0.0008278893624441847, + "loss": 2.4953, + "step": 1283 + }, + { + "epoch": 0.29442788351295573, + "grad_norm": 0.4255207907961841, + "learning_rate": 0.000827608922715191, + "loss": 2.4209, + "step": 1284 + }, + { + "epoch": 0.2946571887181839, + "grad_norm": 0.4436456689448577, + "learning_rate": 0.000827328302279307, + "loss": 2.4871, + "step": 1285 + }, + { + "epoch": 0.29488649392341204, + "grad_norm": 0.5075969454094265, + "learning_rate": 0.0008270475012913211, + "loss": 2.6085, + "step": 1286 + }, + { + "epoch": 0.2951157991286402, + "grad_norm": 0.4679770451479592, + "learning_rate": 0.0008267665199061211, + "loss": 2.3834, + "step": 1287 + }, + { + "epoch": 0.2953451043338684, + "grad_norm": 0.5070527285606218, + "learning_rate": 0.0008264853582786949, + "loss": 2.39, + "step": 1288 + }, + { + "epoch": 0.2955744095390965, + "grad_norm": 0.477235611918014, + "learning_rate": 0.0008262040165641288, + "loss": 2.6524, + "step": 1289 + }, + { + "epoch": 0.2958037147443247, + "grad_norm": 0.4996305508159406, + "learning_rate": 0.0008259224949176094, + "loss": 2.472, + "step": 1290 + }, + { + "epoch": 0.29603301994955283, + "grad_norm": 0.49067605065701503, + "learning_rate": 0.0008256407934944219, + "loss": 2.51, + "step": 1291 + }, + { + "epoch": 0.296262325154781, + "grad_norm": 0.458960906293898, + "learning_rate": 0.0008253589124499512, + "loss": 2.6964, + "step": 1292 + }, + { + "epoch": 0.2964916303600092, + "grad_norm": 0.4616340714349238, + "learning_rate": 0.0008250768519396807, + "loss": 2.576, + "step": 1293 + }, + { + "epoch": 0.2967209355652373, + "grad_norm": 0.6870554376413907, + "learning_rate": 0.0008247946121191933, + "loss": 2.5755, + "step": 1294 + }, + { + "epoch": 0.2969502407704655, + "grad_norm": 0.4347392728016095, + "learning_rate": 0.0008245121931441706, + "loss": 2.4389, + "step": 1295 + }, + { + "epoch": 0.2971795459756936, + "grad_norm": 0.5272861093189205, + "learning_rate": 0.0008242295951703929, + "loss": 2.5568, + "step": 1296 + }, + { + "epoch": 0.2974088511809218, + "grad_norm": 0.44124025067177547, + "learning_rate": 0.0008239468183537393, + "loss": 2.5208, + "step": 1297 + }, + { + "epoch": 0.29763815638615, + "grad_norm": 0.4463387603740826, + "learning_rate": 0.0008236638628501878, + "loss": 2.5838, + "step": 1298 + }, + { + "epoch": 0.2978674615913781, + "grad_norm": 0.4799103071998853, + "learning_rate": 0.0008233807288158146, + "loss": 2.5474, + "step": 1299 + }, + { + "epoch": 0.2980967667966063, + "grad_norm": 0.49081208694177686, + "learning_rate": 0.0008230974164067949, + "loss": 2.5523, + "step": 1300 + }, + { + "epoch": 0.2983260720018344, + "grad_norm": 0.4457061187360201, + "learning_rate": 0.0008228139257794012, + "loss": 2.4326, + "step": 1301 + }, + { + "epoch": 0.2985553772070626, + "grad_norm": 0.4807577479858049, + "learning_rate": 0.0008225302570900058, + "loss": 2.6558, + "step": 1302 + }, + { + "epoch": 0.2987846824122908, + "grad_norm": 0.5157913010888612, + "learning_rate": 0.0008222464104950778, + "loss": 2.4929, + "step": 1303 + }, + { + "epoch": 0.2990139876175189, + "grad_norm": 0.4765768845881543, + "learning_rate": 0.0008219623861511855, + "loss": 2.5185, + "step": 1304 + }, + { + "epoch": 0.2992432928227471, + "grad_norm": 0.4654560132402044, + "learning_rate": 0.000821678184214995, + "loss": 2.5111, + "step": 1305 + }, + { + "epoch": 0.2994725980279752, + "grad_norm": 0.472860150892364, + "learning_rate": 0.0008213938048432696, + "loss": 2.5452, + "step": 1306 + }, + { + "epoch": 0.2997019032332034, + "grad_norm": 0.47356690353200936, + "learning_rate": 0.0008211092481928716, + "loss": 2.5111, + "step": 1307 + }, + { + "epoch": 0.2999312084384316, + "grad_norm": 0.4624931130944497, + "learning_rate": 0.0008208245144207603, + "loss": 2.4125, + "step": 1308 + }, + { + "epoch": 0.3001605136436597, + "grad_norm": 0.45100984733734195, + "learning_rate": 0.0008205396036839927, + "loss": 2.6359, + "step": 1309 + }, + { + "epoch": 0.3003898188488879, + "grad_norm": 0.5021925967291027, + "learning_rate": 0.0008202545161397241, + "loss": 2.4525, + "step": 1310 + }, + { + "epoch": 0.300619124054116, + "grad_norm": 0.48083195072366375, + "learning_rate": 0.0008199692519452069, + "loss": 2.6186, + "step": 1311 + }, + { + "epoch": 0.3008484292593442, + "grad_norm": 0.4965882169409774, + "learning_rate": 0.0008196838112577904, + "loss": 2.6372, + "step": 1312 + }, + { + "epoch": 0.30107773446457237, + "grad_norm": 0.4767534494741337, + "learning_rate": 0.0008193981942349224, + "loss": 2.5358, + "step": 1313 + }, + { + "epoch": 0.3013070396698005, + "grad_norm": 0.5448815512208761, + "learning_rate": 0.0008191124010341468, + "loss": 2.5202, + "step": 1314 + }, + { + "epoch": 0.3015363448750287, + "grad_norm": 0.4988600773592009, + "learning_rate": 0.0008188264318131056, + "loss": 2.5207, + "step": 1315 + }, + { + "epoch": 0.3017656500802568, + "grad_norm": 0.46658756563470727, + "learning_rate": 0.0008185402867295373, + "loss": 2.5721, + "step": 1316 + }, + { + "epoch": 0.301994955285485, + "grad_norm": 0.42922520594522656, + "learning_rate": 0.0008182539659412776, + "loss": 2.522, + "step": 1317 + }, + { + "epoch": 0.30222426049071316, + "grad_norm": 0.5257366196460637, + "learning_rate": 0.0008179674696062593, + "loss": 2.4038, + "step": 1318 + }, + { + "epoch": 0.3024535656959413, + "grad_norm": 0.48697859788237674, + "learning_rate": 0.0008176807978825118, + "loss": 2.4257, + "step": 1319 + }, + { + "epoch": 0.30268287090116947, + "grad_norm": 0.47735221441151554, + "learning_rate": 0.0008173939509281613, + "loss": 2.4459, + "step": 1320 + }, + { + "epoch": 0.3029121761063976, + "grad_norm": 0.4493651565207939, + "learning_rate": 0.0008171069289014306, + "loss": 2.5185, + "step": 1321 + }, + { + "epoch": 0.30314148131162577, + "grad_norm": 0.4507402338408838, + "learning_rate": 0.0008168197319606392, + "loss": 2.6408, + "step": 1322 + }, + { + "epoch": 0.30337078651685395, + "grad_norm": 0.4905063565327578, + "learning_rate": 0.0008165323602642028, + "loss": 2.6167, + "step": 1323 + }, + { + "epoch": 0.3036000917220821, + "grad_norm": 0.5400771526205824, + "learning_rate": 0.0008162448139706341, + "loss": 2.6781, + "step": 1324 + }, + { + "epoch": 0.30382939692731026, + "grad_norm": 0.5143752363777043, + "learning_rate": 0.0008159570932385414, + "loss": 2.5871, + "step": 1325 + }, + { + "epoch": 0.3040587021325384, + "grad_norm": 0.5043163029342392, + "learning_rate": 0.0008156691982266298, + "loss": 2.4029, + "step": 1326 + }, + { + "epoch": 0.30428800733776656, + "grad_norm": 0.4566381870254535, + "learning_rate": 0.0008153811290936999, + "loss": 2.405, + "step": 1327 + }, + { + "epoch": 0.30451731254299474, + "grad_norm": 0.4748705665090741, + "learning_rate": 0.0008150928859986488, + "loss": 2.5591, + "step": 1328 + }, + { + "epoch": 0.30474661774822287, + "grad_norm": 0.4469888433871677, + "learning_rate": 0.0008148044691004698, + "loss": 2.5788, + "step": 1329 + }, + { + "epoch": 0.30497592295345105, + "grad_norm": 0.47675983422680657, + "learning_rate": 0.0008145158785582514, + "loss": 2.4196, + "step": 1330 + }, + { + "epoch": 0.3052052281586792, + "grad_norm": 0.45030667493017884, + "learning_rate": 0.0008142271145311783, + "loss": 2.5696, + "step": 1331 + }, + { + "epoch": 0.30543453336390736, + "grad_norm": 0.5893444904654627, + "learning_rate": 0.0008139381771785308, + "loss": 2.5735, + "step": 1332 + }, + { + "epoch": 0.30566383856913554, + "grad_norm": 0.47996037805049646, + "learning_rate": 0.000813649066659685, + "loss": 2.5532, + "step": 1333 + }, + { + "epoch": 0.30589314377436366, + "grad_norm": 0.4609698217063107, + "learning_rate": 0.000813359783134112, + "loss": 2.6293, + "step": 1334 + }, + { + "epoch": 0.30612244897959184, + "grad_norm": 0.48106520459434443, + "learning_rate": 0.0008130703267613787, + "loss": 2.6066, + "step": 1335 + }, + { + "epoch": 0.30635175418481997, + "grad_norm": 0.46926938796697404, + "learning_rate": 0.0008127806977011474, + "loss": 2.5737, + "step": 1336 + }, + { + "epoch": 0.30658105939004815, + "grad_norm": 0.49967187067676566, + "learning_rate": 0.0008124908961131759, + "loss": 2.6087, + "step": 1337 + }, + { + "epoch": 0.30681036459527633, + "grad_norm": 0.5443473520204855, + "learning_rate": 0.0008122009221573161, + "loss": 2.5524, + "step": 1338 + }, + { + "epoch": 0.30703966980050446, + "grad_norm": 0.46925320530386466, + "learning_rate": 0.0008119107759935163, + "loss": 2.377, + "step": 1339 + }, + { + "epoch": 0.30726897500573264, + "grad_norm": 0.5213044649489694, + "learning_rate": 0.000811620457781819, + "loss": 2.4201, + "step": 1340 + }, + { + "epoch": 0.30749828021096076, + "grad_norm": 0.44971875992129146, + "learning_rate": 0.0008113299676823615, + "loss": 2.5156, + "step": 1341 + }, + { + "epoch": 0.30772758541618894, + "grad_norm": 0.5293860216094946, + "learning_rate": 0.0008110393058553765, + "loss": 2.5186, + "step": 1342 + }, + { + "epoch": 0.3079568906214171, + "grad_norm": 0.4496365531353882, + "learning_rate": 0.0008107484724611911, + "loss": 2.6339, + "step": 1343 + }, + { + "epoch": 0.30818619582664525, + "grad_norm": 0.4834592683789248, + "learning_rate": 0.000810457467660227, + "loss": 2.6172, + "step": 1344 + }, + { + "epoch": 0.30841550103187343, + "grad_norm": 0.4878062022896494, + "learning_rate": 0.0008101662916130006, + "loss": 2.6117, + "step": 1345 + }, + { + "epoch": 0.3086448062371016, + "grad_norm": 0.634842676348823, + "learning_rate": 0.0008098749444801224, + "loss": 2.4618, + "step": 1346 + }, + { + "epoch": 0.30887411144232974, + "grad_norm": 0.4790973089056852, + "learning_rate": 0.0008095834264222979, + "loss": 2.4705, + "step": 1347 + }, + { + "epoch": 0.3091034166475579, + "grad_norm": 0.47998150399588535, + "learning_rate": 0.0008092917376003262, + "loss": 2.5379, + "step": 1348 + }, + { + "epoch": 0.30933272185278604, + "grad_norm": 0.45362639031882945, + "learning_rate": 0.0008089998781751009, + "loss": 2.6565, + "step": 1349 + }, + { + "epoch": 0.3095620270580142, + "grad_norm": 0.4586549437521482, + "learning_rate": 0.00080870784830761, + "loss": 2.5996, + "step": 1350 + }, + { + "epoch": 0.3097913322632424, + "grad_norm": 0.45753986424108206, + "learning_rate": 0.0008084156481589349, + "loss": 2.4815, + "step": 1351 + }, + { + "epoch": 0.31002063746847053, + "grad_norm": 0.4545079608868822, + "learning_rate": 0.0008081232778902516, + "loss": 2.6065, + "step": 1352 + }, + { + "epoch": 0.3102499426736987, + "grad_norm": 0.4309980226648794, + "learning_rate": 0.0008078307376628291, + "loss": 2.5075, + "step": 1353 + }, + { + "epoch": 0.31047924787892683, + "grad_norm": 0.46450131565271896, + "learning_rate": 0.0008075380276380311, + "loss": 2.6567, + "step": 1354 + }, + { + "epoch": 0.310708553084155, + "grad_norm": 0.4302767017948274, + "learning_rate": 0.0008072451479773143, + "loss": 2.5173, + "step": 1355 + }, + { + "epoch": 0.3109378582893832, + "grad_norm": 0.44110764031605, + "learning_rate": 0.0008069520988422291, + "loss": 2.4832, + "step": 1356 + }, + { + "epoch": 0.3111671634946113, + "grad_norm": 0.41558449154330984, + "learning_rate": 0.0008066588803944195, + "loss": 2.6051, + "step": 1357 + }, + { + "epoch": 0.3113964686998395, + "grad_norm": 0.42599619047435977, + "learning_rate": 0.0008063654927956227, + "loss": 2.6715, + "step": 1358 + }, + { + "epoch": 0.3116257739050676, + "grad_norm": 0.4466151575516139, + "learning_rate": 0.0008060719362076697, + "loss": 2.5252, + "step": 1359 + }, + { + "epoch": 0.3118550791102958, + "grad_norm": 0.4628843958076107, + "learning_rate": 0.000805778210792484, + "loss": 2.6472, + "step": 1360 + }, + { + "epoch": 0.312084384315524, + "grad_norm": 0.48027381028210464, + "learning_rate": 0.0008054843167120826, + "loss": 2.5837, + "step": 1361 + }, + { + "epoch": 0.3123136895207521, + "grad_norm": 0.4256467022264815, + "learning_rate": 0.0008051902541285758, + "loss": 2.6147, + "step": 1362 + }, + { + "epoch": 0.3125429947259803, + "grad_norm": 0.41368786073869857, + "learning_rate": 0.0008048960232041663, + "loss": 2.4602, + "step": 1363 + }, + { + "epoch": 0.3127722999312084, + "grad_norm": 0.447536472891922, + "learning_rate": 0.00080460162410115, + "loss": 2.5588, + "step": 1364 + }, + { + "epoch": 0.3130016051364366, + "grad_norm": 0.45105347933687595, + "learning_rate": 0.0008043070569819153, + "loss": 2.5002, + "step": 1365 + }, + { + "epoch": 0.3132309103416648, + "grad_norm": 0.4440444608040981, + "learning_rate": 0.0008040123220089437, + "loss": 2.4771, + "step": 1366 + }, + { + "epoch": 0.3134602155468929, + "grad_norm": 0.48398284518385515, + "learning_rate": 0.0008037174193448089, + "loss": 2.6474, + "step": 1367 + }, + { + "epoch": 0.3136895207521211, + "grad_norm": 0.462946400440165, + "learning_rate": 0.0008034223491521775, + "loss": 2.5765, + "step": 1368 + }, + { + "epoch": 0.3139188259573492, + "grad_norm": 0.45179153789558973, + "learning_rate": 0.0008031271115938077, + "loss": 2.5481, + "step": 1369 + }, + { + "epoch": 0.3141481311625774, + "grad_norm": 0.42742712387483756, + "learning_rate": 0.000802831706832551, + "loss": 2.4135, + "step": 1370 + }, + { + "epoch": 0.3143774363678056, + "grad_norm": 0.42116822706246704, + "learning_rate": 0.0008025361350313505, + "loss": 2.3618, + "step": 1371 + }, + { + "epoch": 0.3146067415730337, + "grad_norm": 0.4644889445323345, + "learning_rate": 0.0008022403963532419, + "loss": 2.5191, + "step": 1372 + }, + { + "epoch": 0.3148360467782619, + "grad_norm": 0.4622532857598247, + "learning_rate": 0.0008019444909613523, + "loss": 2.5396, + "step": 1373 + }, + { + "epoch": 0.31506535198349, + "grad_norm": 0.4970975592627013, + "learning_rate": 0.0008016484190189012, + "loss": 2.7058, + "step": 1374 + }, + { + "epoch": 0.3152946571887182, + "grad_norm": 0.4892662920013599, + "learning_rate": 0.0008013521806892003, + "loss": 2.5907, + "step": 1375 + }, + { + "epoch": 0.31552396239394637, + "grad_norm": 0.4590186938556544, + "learning_rate": 0.0008010557761356522, + "loss": 2.6187, + "step": 1376 + }, + { + "epoch": 0.3157532675991745, + "grad_norm": 0.45451691229801006, + "learning_rate": 0.000800759205521752, + "loss": 2.4002, + "step": 1377 + }, + { + "epoch": 0.3159825728044027, + "grad_norm": 0.46419975708523564, + "learning_rate": 0.0008004624690110859, + "loss": 2.5378, + "step": 1378 + }, + { + "epoch": 0.3162118780096308, + "grad_norm": 0.438626321286265, + "learning_rate": 0.0008001655667673318, + "loss": 2.6798, + "step": 1379 + }, + { + "epoch": 0.316441183214859, + "grad_norm": 0.48325262832412436, + "learning_rate": 0.0007998684989542588, + "loss": 2.7027, + "step": 1380 + }, + { + "epoch": 0.31667048842008716, + "grad_norm": 0.4577464705563416, + "learning_rate": 0.0007995712657357279, + "loss": 2.5916, + "step": 1381 + }, + { + "epoch": 0.3168997936253153, + "grad_norm": 0.48274188485573666, + "learning_rate": 0.0007992738672756908, + "loss": 2.6608, + "step": 1382 + }, + { + "epoch": 0.31712909883054347, + "grad_norm": 0.41413380186501525, + "learning_rate": 0.0007989763037381904, + "loss": 2.6107, + "step": 1383 + }, + { + "epoch": 0.3173584040357716, + "grad_norm": 0.420835106339783, + "learning_rate": 0.0007986785752873609, + "loss": 2.3988, + "step": 1384 + }, + { + "epoch": 0.31758770924099977, + "grad_norm": 0.4100566796120642, + "learning_rate": 0.0007983806820874271, + "loss": 2.4617, + "step": 1385 + }, + { + "epoch": 0.31781701444622795, + "grad_norm": 0.4529920672002737, + "learning_rate": 0.0007980826243027052, + "loss": 2.5744, + "step": 1386 + }, + { + "epoch": 0.3180463196514561, + "grad_norm": 0.4368871564815729, + "learning_rate": 0.0007977844020976016, + "loss": 2.7454, + "step": 1387 + }, + { + "epoch": 0.31827562485668426, + "grad_norm": 0.45282641434372195, + "learning_rate": 0.0007974860156366139, + "loss": 2.5049, + "step": 1388 + }, + { + "epoch": 0.3185049300619124, + "grad_norm": 0.43766849323810186, + "learning_rate": 0.00079718746508433, + "loss": 2.5842, + "step": 1389 + }, + { + "epoch": 0.31873423526714056, + "grad_norm": 0.4735515966275632, + "learning_rate": 0.0007968887506054282, + "loss": 2.524, + "step": 1390 + }, + { + "epoch": 0.31896354047236875, + "grad_norm": 0.4691545282898903, + "learning_rate": 0.0007965898723646776, + "loss": 2.6418, + "step": 1391 + }, + { + "epoch": 0.31919284567759687, + "grad_norm": 0.5037755674359052, + "learning_rate": 0.0007962908305269375, + "loss": 2.496, + "step": 1392 + }, + { + "epoch": 0.31942215088282505, + "grad_norm": 0.4430768440783589, + "learning_rate": 0.0007959916252571573, + "loss": 2.5475, + "step": 1393 + }, + { + "epoch": 0.3196514560880532, + "grad_norm": 0.47194478149564145, + "learning_rate": 0.0007956922567203766, + "loss": 2.3864, + "step": 1394 + }, + { + "epoch": 0.31988076129328136, + "grad_norm": 0.45295180378981814, + "learning_rate": 0.000795392725081725, + "loss": 2.5655, + "step": 1395 + }, + { + "epoch": 0.32011006649850954, + "grad_norm": 0.4651980956726357, + "learning_rate": 0.0007950930305064223, + "loss": 2.5018, + "step": 1396 + }, + { + "epoch": 0.32033937170373766, + "grad_norm": 0.42867710237288636, + "learning_rate": 0.000794793173159778, + "loss": 2.4605, + "step": 1397 + }, + { + "epoch": 0.32056867690896584, + "grad_norm": 0.48227774496913745, + "learning_rate": 0.0007944931532071912, + "loss": 2.6782, + "step": 1398 + }, + { + "epoch": 0.32079798211419397, + "grad_norm": 0.4640458886224156, + "learning_rate": 0.0007941929708141513, + "loss": 2.6967, + "step": 1399 + }, + { + "epoch": 0.32102728731942215, + "grad_norm": 0.439836423860641, + "learning_rate": 0.0007938926261462366, + "loss": 2.4886, + "step": 1400 + }, + { + "epoch": 0.32125659252465033, + "grad_norm": 0.4457817417643202, + "learning_rate": 0.0007935921193691153, + "loss": 2.4373, + "step": 1401 + }, + { + "epoch": 0.32148589772987846, + "grad_norm": 0.4519837695272278, + "learning_rate": 0.000793291450648545, + "loss": 2.659, + "step": 1402 + }, + { + "epoch": 0.32171520293510664, + "grad_norm": 0.4127653641151133, + "learning_rate": 0.0007929906201503722, + "loss": 2.4186, + "step": 1403 + }, + { + "epoch": 0.32194450814033476, + "grad_norm": 0.4300550352248573, + "learning_rate": 0.0007926896280405334, + "loss": 2.6568, + "step": 1404 + }, + { + "epoch": 0.32217381334556294, + "grad_norm": 0.4470031536205589, + "learning_rate": 0.0007923884744850536, + "loss": 2.3887, + "step": 1405 + }, + { + "epoch": 0.3224031185507911, + "grad_norm": 0.4533588734943564, + "learning_rate": 0.0007920871596500472, + "loss": 2.4954, + "step": 1406 + }, + { + "epoch": 0.32263242375601925, + "grad_norm": 0.4402943495833976, + "learning_rate": 0.0007917856837017176, + "loss": 2.5927, + "step": 1407 + }, + { + "epoch": 0.32286172896124743, + "grad_norm": 0.46351193494675563, + "learning_rate": 0.0007914840468063566, + "loss": 2.5915, + "step": 1408 + }, + { + "epoch": 0.32309103416647555, + "grad_norm": 0.4763694895405192, + "learning_rate": 0.0007911822491303452, + "loss": 2.3563, + "step": 1409 + }, + { + "epoch": 0.32332033937170374, + "grad_norm": 0.45112235449549964, + "learning_rate": 0.0007908802908401532, + "loss": 2.4679, + "step": 1410 + }, + { + "epoch": 0.3235496445769319, + "grad_norm": 0.4352580371143475, + "learning_rate": 0.0007905781721023382, + "loss": 2.2662, + "step": 1411 + }, + { + "epoch": 0.32377894978216004, + "grad_norm": 0.4777391418466409, + "learning_rate": 0.0007902758930835477, + "loss": 2.5716, + "step": 1412 + }, + { + "epoch": 0.3240082549873882, + "grad_norm": 0.44358085225373534, + "learning_rate": 0.000789973453950516, + "loss": 2.53, + "step": 1413 + }, + { + "epoch": 0.32423756019261635, + "grad_norm": 0.5088948904352971, + "learning_rate": 0.000789670854870067, + "loss": 2.5783, + "step": 1414 + }, + { + "epoch": 0.32446686539784453, + "grad_norm": 0.4891702871103524, + "learning_rate": 0.000789368096009112, + "loss": 2.5438, + "step": 1415 + }, + { + "epoch": 0.3246961706030727, + "grad_norm": 0.45657035318663786, + "learning_rate": 0.0007890651775346512, + "loss": 2.5482, + "step": 1416 + }, + { + "epoch": 0.32492547580830083, + "grad_norm": 0.48164723760490885, + "learning_rate": 0.0007887620996137721, + "loss": 2.5413, + "step": 1417 + }, + { + "epoch": 0.325154781013529, + "grad_norm": 0.5336299946957686, + "learning_rate": 0.0007884588624136504, + "loss": 2.5557, + "step": 1418 + }, + { + "epoch": 0.32538408621875714, + "grad_norm": 0.45357982305616823, + "learning_rate": 0.0007881554661015497, + "loss": 2.5647, + "step": 1419 + }, + { + "epoch": 0.3256133914239853, + "grad_norm": 0.461045767312388, + "learning_rate": 0.0007878519108448218, + "loss": 2.7376, + "step": 1420 + }, + { + "epoch": 0.3258426966292135, + "grad_norm": 0.4239906208188082, + "learning_rate": 0.0007875481968109051, + "loss": 2.5166, + "step": 1421 + }, + { + "epoch": 0.3260720018344416, + "grad_norm": 0.47252828039811373, + "learning_rate": 0.0007872443241673269, + "loss": 2.6449, + "step": 1422 + }, + { + "epoch": 0.3263013070396698, + "grad_norm": 0.4010967571029355, + "learning_rate": 0.0007869402930817007, + "loss": 2.4663, + "step": 1423 + }, + { + "epoch": 0.32653061224489793, + "grad_norm": 0.4233055649425182, + "learning_rate": 0.0007866361037217284, + "loss": 2.4252, + "step": 1424 + }, + { + "epoch": 0.3267599174501261, + "grad_norm": 0.4628825447708869, + "learning_rate": 0.0007863317562551987, + "loss": 2.5535, + "step": 1425 + }, + { + "epoch": 0.3269892226553543, + "grad_norm": 0.43902825852110017, + "learning_rate": 0.0007860272508499876, + "loss": 2.4404, + "step": 1426 + }, + { + "epoch": 0.3272185278605824, + "grad_norm": 0.4734716312726826, + "learning_rate": 0.0007857225876740584, + "loss": 2.4145, + "step": 1427 + }, + { + "epoch": 0.3274478330658106, + "grad_norm": 0.4420165106208353, + "learning_rate": 0.0007854177668954609, + "loss": 2.4756, + "step": 1428 + }, + { + "epoch": 0.3276771382710387, + "grad_norm": 0.45504556555851605, + "learning_rate": 0.0007851127886823327, + "loss": 2.5487, + "step": 1429 + }, + { + "epoch": 0.3279064434762669, + "grad_norm": 0.4966339861554591, + "learning_rate": 0.0007848076532028974, + "loss": 2.7324, + "step": 1430 + }, + { + "epoch": 0.3281357486814951, + "grad_norm": 0.47342524257636737, + "learning_rate": 0.0007845023606254658, + "loss": 2.6494, + "step": 1431 + }, + { + "epoch": 0.3283650538867232, + "grad_norm": 0.4791559207922651, + "learning_rate": 0.0007841969111184352, + "loss": 2.435, + "step": 1432 + }, + { + "epoch": 0.3285943590919514, + "grad_norm": 0.43133367264324024, + "learning_rate": 0.0007838913048502894, + "loss": 2.4489, + "step": 1433 + }, + { + "epoch": 0.3288236642971795, + "grad_norm": 0.5203301273514411, + "learning_rate": 0.0007835855419895992, + "loss": 2.4723, + "step": 1434 + }, + { + "epoch": 0.3290529695024077, + "grad_norm": 0.4978655454142866, + "learning_rate": 0.0007832796227050208, + "loss": 2.5384, + "step": 1435 + }, + { + "epoch": 0.3292822747076359, + "grad_norm": 0.4228075332791875, + "learning_rate": 0.0007829735471652976, + "loss": 2.5447, + "step": 1436 + }, + { + "epoch": 0.329511579912864, + "grad_norm": 0.4575240585457995, + "learning_rate": 0.0007826673155392587, + "loss": 2.5775, + "step": 1437 + }, + { + "epoch": 0.3297408851180922, + "grad_norm": 0.48735152579713525, + "learning_rate": 0.0007823609279958194, + "loss": 2.5541, + "step": 1438 + }, + { + "epoch": 0.32997019032332037, + "grad_norm": 0.46394789060529296, + "learning_rate": 0.000782054384703981, + "loss": 2.5449, + "step": 1439 + }, + { + "epoch": 0.3301994955285485, + "grad_norm": 0.4563055792590887, + "learning_rate": 0.0007817476858328308, + "loss": 2.4926, + "step": 1440 + }, + { + "epoch": 0.3304288007337767, + "grad_norm": 0.46024076058691116, + "learning_rate": 0.0007814408315515418, + "loss": 2.487, + "step": 1441 + }, + { + "epoch": 0.3306581059390048, + "grad_norm": 0.4524794601120097, + "learning_rate": 0.0007811338220293729, + "loss": 2.4024, + "step": 1442 + }, + { + "epoch": 0.330887411144233, + "grad_norm": 0.4772763114981455, + "learning_rate": 0.0007808266574356683, + "loss": 2.5534, + "step": 1443 + }, + { + "epoch": 0.33111671634946116, + "grad_norm": 0.44906036632567925, + "learning_rate": 0.000780519337939858, + "loss": 2.6134, + "step": 1444 + }, + { + "epoch": 0.3313460215546893, + "grad_norm": 0.5092228714450693, + "learning_rate": 0.0007802118637114573, + "loss": 2.527, + "step": 1445 + }, + { + "epoch": 0.33157532675991747, + "grad_norm": 0.5538991725927475, + "learning_rate": 0.0007799042349200671, + "loss": 2.5611, + "step": 1446 + }, + { + "epoch": 0.3318046319651456, + "grad_norm": 0.4818543637288862, + "learning_rate": 0.0007795964517353734, + "loss": 2.5247, + "step": 1447 + }, + { + "epoch": 0.33203393717037377, + "grad_norm": 0.48926931658555306, + "learning_rate": 0.0007792885143271472, + "loss": 2.4761, + "step": 1448 + }, + { + "epoch": 0.33226324237560195, + "grad_norm": 0.473539160451376, + "learning_rate": 0.0007789804228652449, + "loss": 2.4097, + "step": 1449 + }, + { + "epoch": 0.3324925475808301, + "grad_norm": 0.4594445790064673, + "learning_rate": 0.0007786721775196073, + "loss": 2.5455, + "step": 1450 + }, + { + "epoch": 0.33272185278605826, + "grad_norm": 0.4425068755963015, + "learning_rate": 0.0007783637784602609, + "loss": 2.5489, + "step": 1451 + }, + { + "epoch": 0.3329511579912864, + "grad_norm": 0.44288508223813816, + "learning_rate": 0.0007780552258573163, + "loss": 2.4912, + "step": 1452 + }, + { + "epoch": 0.33318046319651456, + "grad_norm": 0.4535272705374163, + "learning_rate": 0.0007777465198809692, + "loss": 2.5947, + "step": 1453 + }, + { + "epoch": 0.33340976840174275, + "grad_norm": 0.454291227478551, + "learning_rate": 0.0007774376607014994, + "loss": 2.4771, + "step": 1454 + }, + { + "epoch": 0.33363907360697087, + "grad_norm": 0.4180050191218943, + "learning_rate": 0.0007771286484892722, + "loss": 2.537, + "step": 1455 + }, + { + "epoch": 0.33386837881219905, + "grad_norm": 0.44401522382370046, + "learning_rate": 0.0007768194834147362, + "loss": 2.6954, + "step": 1456 + }, + { + "epoch": 0.3340976840174272, + "grad_norm": 0.4727939753772184, + "learning_rate": 0.000776510165648425, + "loss": 2.5163, + "step": 1457 + }, + { + "epoch": 0.33432698922265536, + "grad_norm": 0.47147720163498796, + "learning_rate": 0.000776200695360956, + "loss": 2.6601, + "step": 1458 + }, + { + "epoch": 0.33455629442788354, + "grad_norm": 0.4765088406699159, + "learning_rate": 0.0007758910727230311, + "loss": 2.4979, + "step": 1459 + }, + { + "epoch": 0.33478559963311166, + "grad_norm": 0.6016517162085594, + "learning_rate": 0.000775581297905436, + "loss": 2.5184, + "step": 1460 + }, + { + "epoch": 0.33501490483833984, + "grad_norm": 0.4994854874096803, + "learning_rate": 0.0007752713710790404, + "loss": 2.881, + "step": 1461 + }, + { + "epoch": 0.33524421004356797, + "grad_norm": 0.4398888716236374, + "learning_rate": 0.000774961292414798, + "loss": 2.3851, + "step": 1462 + }, + { + "epoch": 0.33547351524879615, + "grad_norm": 0.5040059316557297, + "learning_rate": 0.0007746510620837459, + "loss": 2.4507, + "step": 1463 + }, + { + "epoch": 0.33570282045402433, + "grad_norm": 0.4755178157969526, + "learning_rate": 0.0007743406802570053, + "loss": 2.4162, + "step": 1464 + }, + { + "epoch": 0.33593212565925246, + "grad_norm": 0.4318575081878421, + "learning_rate": 0.0007740301471057807, + "loss": 2.4879, + "step": 1465 + }, + { + "epoch": 0.33616143086448064, + "grad_norm": 0.4561328600739601, + "learning_rate": 0.0007737194628013599, + "loss": 2.5349, + "step": 1466 + }, + { + "epoch": 0.33639073606970876, + "grad_norm": 0.4631068249680571, + "learning_rate": 0.0007734086275151146, + "loss": 2.4814, + "step": 1467 + }, + { + "epoch": 0.33662004127493694, + "grad_norm": 0.4554174259704821, + "learning_rate": 0.0007730976414184992, + "loss": 2.4745, + "step": 1468 + }, + { + "epoch": 0.3368493464801651, + "grad_norm": 0.49374926381317313, + "learning_rate": 0.0007727865046830517, + "loss": 2.57, + "step": 1469 + }, + { + "epoch": 0.33707865168539325, + "grad_norm": 0.6386187177607843, + "learning_rate": 0.0007724752174803929, + "loss": 2.6771, + "step": 1470 + }, + { + "epoch": 0.33730795689062143, + "grad_norm": 0.4492259515035345, + "learning_rate": 0.0007721637799822269, + "loss": 2.606, + "step": 1471 + }, + { + "epoch": 0.33753726209584956, + "grad_norm": 0.47718533452675455, + "learning_rate": 0.0007718521923603404, + "loss": 2.3461, + "step": 1472 + }, + { + "epoch": 0.33776656730107774, + "grad_norm": 0.5077032367722294, + "learning_rate": 0.0007715404547866032, + "loss": 2.4709, + "step": 1473 + }, + { + "epoch": 0.3379958725063059, + "grad_norm": 0.4483840855716774, + "learning_rate": 0.0007712285674329676, + "loss": 2.5595, + "step": 1474 + }, + { + "epoch": 0.33822517771153404, + "grad_norm": 0.459282069616094, + "learning_rate": 0.0007709165304714685, + "loss": 2.6095, + "step": 1475 + }, + { + "epoch": 0.3384544829167622, + "grad_norm": 0.4716725858160699, + "learning_rate": 0.0007706043440742235, + "loss": 2.4759, + "step": 1476 + }, + { + "epoch": 0.33868378812199035, + "grad_norm": 0.45983599585231527, + "learning_rate": 0.0007702920084134324, + "loss": 2.4363, + "step": 1477 + }, + { + "epoch": 0.33891309332721853, + "grad_norm": 0.4924512804763021, + "learning_rate": 0.0007699795236613776, + "loss": 2.555, + "step": 1478 + }, + { + "epoch": 0.3391423985324467, + "grad_norm": 0.4724767894171939, + "learning_rate": 0.0007696668899904236, + "loss": 2.553, + "step": 1479 + }, + { + "epoch": 0.33937170373767483, + "grad_norm": 0.43347549496761484, + "learning_rate": 0.000769354107573017, + "loss": 2.5693, + "step": 1480 + }, + { + "epoch": 0.339601008942903, + "grad_norm": 0.49305996933569846, + "learning_rate": 0.0007690411765816864, + "loss": 2.7451, + "step": 1481 + }, + { + "epoch": 0.33983031414813114, + "grad_norm": 0.46304442412548324, + "learning_rate": 0.0007687280971890425, + "loss": 2.5762, + "step": 1482 + }, + { + "epoch": 0.3400596193533593, + "grad_norm": 0.5300414438729194, + "learning_rate": 0.0007684148695677778, + "loss": 2.6085, + "step": 1483 + }, + { + "epoch": 0.3402889245585875, + "grad_norm": 0.4906532952553116, + "learning_rate": 0.0007681014938906666, + "loss": 2.5286, + "step": 1484 + }, + { + "epoch": 0.3405182297638156, + "grad_norm": 0.43650016976548733, + "learning_rate": 0.000767787970330565, + "loss": 2.5298, + "step": 1485 + }, + { + "epoch": 0.3407475349690438, + "grad_norm": 0.4140200401041763, + "learning_rate": 0.0007674742990604101, + "loss": 2.5771, + "step": 1486 + }, + { + "epoch": 0.34097684017427193, + "grad_norm": 0.4684281173411318, + "learning_rate": 0.000767160480253221, + "loss": 2.5544, + "step": 1487 + }, + { + "epoch": 0.3412061453795001, + "grad_norm": 0.5111443479162884, + "learning_rate": 0.0007668465140820984, + "loss": 2.6454, + "step": 1488 + }, + { + "epoch": 0.3414354505847283, + "grad_norm": 0.44932033526043536, + "learning_rate": 0.0007665324007202235, + "loss": 2.614, + "step": 1489 + }, + { + "epoch": 0.3416647557899564, + "grad_norm": 0.4441712034886222, + "learning_rate": 0.0007662181403408592, + "loss": 2.5903, + "step": 1490 + }, + { + "epoch": 0.3418940609951846, + "grad_norm": 0.4341221211926292, + "learning_rate": 0.0007659037331173498, + "loss": 2.5791, + "step": 1491 + }, + { + "epoch": 0.3421233662004127, + "grad_norm": 0.4482313653111539, + "learning_rate": 0.0007655891792231197, + "loss": 2.4999, + "step": 1492 + }, + { + "epoch": 0.3423526714056409, + "grad_norm": 0.48325499915900383, + "learning_rate": 0.0007652744788316752, + "loss": 2.4012, + "step": 1493 + }, + { + "epoch": 0.3425819766108691, + "grad_norm": 0.45280205593639294, + "learning_rate": 0.0007649596321166025, + "loss": 2.469, + "step": 1494 + }, + { + "epoch": 0.3428112818160972, + "grad_norm": 0.4936464838474039, + "learning_rate": 0.0007646446392515692, + "loss": 2.5808, + "step": 1495 + }, + { + "epoch": 0.3430405870213254, + "grad_norm": 0.4902859738149358, + "learning_rate": 0.0007643295004103232, + "loss": 2.5282, + "step": 1496 + }, + { + "epoch": 0.3432698922265535, + "grad_norm": 0.4906335856935621, + "learning_rate": 0.000764014215766693, + "loss": 2.5303, + "step": 1497 + }, + { + "epoch": 0.3434991974317817, + "grad_norm": 0.4912756307291836, + "learning_rate": 0.0007636987854945874, + "loss": 2.575, + "step": 1498 + }, + { + "epoch": 0.3437285026370099, + "grad_norm": 0.4501538058266363, + "learning_rate": 0.0007633832097679958, + "loss": 2.4795, + "step": 1499 + }, + { + "epoch": 0.343957807842238, + "grad_norm": 0.6174835404987802, + "learning_rate": 0.0007630674887609876, + "loss": 2.5775, + "step": 1500 + }, + { + "epoch": 0.3441871130474662, + "grad_norm": 0.4508198654605851, + "learning_rate": 0.0007627516226477122, + "loss": 2.5739, + "step": 1501 + }, + { + "epoch": 0.3444164182526943, + "grad_norm": 0.42150939615725913, + "learning_rate": 0.0007624356116023996, + "loss": 2.3911, + "step": 1502 + }, + { + "epoch": 0.3446457234579225, + "grad_norm": 0.41466120895934055, + "learning_rate": 0.0007621194557993589, + "loss": 2.6728, + "step": 1503 + }, + { + "epoch": 0.3448750286631507, + "grad_norm": 0.4789782677829923, + "learning_rate": 0.00076180315541298, + "loss": 2.5951, + "step": 1504 + }, + { + "epoch": 0.3451043338683788, + "grad_norm": 0.44868013681214725, + "learning_rate": 0.0007614867106177319, + "loss": 2.4698, + "step": 1505 + }, + { + "epoch": 0.345333639073607, + "grad_norm": 0.45630788639134284, + "learning_rate": 0.0007611701215881635, + "loss": 2.4408, + "step": 1506 + }, + { + "epoch": 0.3455629442788351, + "grad_norm": 0.5858642338203837, + "learning_rate": 0.0007608533884989029, + "loss": 2.5908, + "step": 1507 + }, + { + "epoch": 0.3457922494840633, + "grad_norm": 0.42907831194124024, + "learning_rate": 0.000760536511524658, + "loss": 2.4987, + "step": 1508 + }, + { + "epoch": 0.34602155468929147, + "grad_norm": 0.45379335937104404, + "learning_rate": 0.0007602194908402166, + "loss": 2.5033, + "step": 1509 + }, + { + "epoch": 0.3462508598945196, + "grad_norm": 0.46041228741143897, + "learning_rate": 0.0007599023266204445, + "loss": 2.5773, + "step": 1510 + }, + { + "epoch": 0.34648016509974777, + "grad_norm": 0.47488563988277616, + "learning_rate": 0.0007595850190402877, + "loss": 2.5231, + "step": 1511 + }, + { + "epoch": 0.3467094703049759, + "grad_norm": 0.5143777000630925, + "learning_rate": 0.0007592675682747708, + "loss": 2.5226, + "step": 1512 + }, + { + "epoch": 0.3469387755102041, + "grad_norm": 0.424491029654998, + "learning_rate": 0.0007589499744989976, + "loss": 2.4157, + "step": 1513 + }, + { + "epoch": 0.34716808071543226, + "grad_norm": 0.420177309007689, + "learning_rate": 0.0007586322378881508, + "loss": 2.5642, + "step": 1514 + }, + { + "epoch": 0.3473973859206604, + "grad_norm": 0.4601538461625998, + "learning_rate": 0.0007583143586174916, + "loss": 2.675, + "step": 1515 + }, + { + "epoch": 0.34762669112588857, + "grad_norm": 0.46088930571970194, + "learning_rate": 0.0007579963368623601, + "loss": 2.5508, + "step": 1516 + }, + { + "epoch": 0.3478559963311167, + "grad_norm": 0.41087249343858323, + "learning_rate": 0.000757678172798175, + "loss": 2.6368, + "step": 1517 + }, + { + "epoch": 0.34808530153634487, + "grad_norm": 0.4451230790611183, + "learning_rate": 0.0007573598666004335, + "loss": 2.4509, + "step": 1518 + }, + { + "epoch": 0.34831460674157305, + "grad_norm": 0.47259232492821546, + "learning_rate": 0.0007570414184447112, + "loss": 2.5569, + "step": 1519 + }, + { + "epoch": 0.3485439119468012, + "grad_norm": 0.4213996862140523, + "learning_rate": 0.000756722828506662, + "loss": 2.5612, + "step": 1520 + }, + { + "epoch": 0.34877321715202936, + "grad_norm": 0.45501781208829645, + "learning_rate": 0.0007564040969620179, + "loss": 2.5509, + "step": 1521 + }, + { + "epoch": 0.3490025223572575, + "grad_norm": 0.4301351859938062, + "learning_rate": 0.0007560852239865889, + "loss": 2.4809, + "step": 1522 + }, + { + "epoch": 0.34923182756248566, + "grad_norm": 0.44837816803612623, + "learning_rate": 0.0007557662097562636, + "loss": 2.7425, + "step": 1523 + }, + { + "epoch": 0.34946113276771384, + "grad_norm": 0.483812632328673, + "learning_rate": 0.0007554470544470079, + "loss": 2.5025, + "step": 1524 + }, + { + "epoch": 0.34969043797294197, + "grad_norm": 0.43256063270842043, + "learning_rate": 0.0007551277582348658, + "loss": 2.5217, + "step": 1525 + }, + { + "epoch": 0.34991974317817015, + "grad_norm": 0.45838586316372154, + "learning_rate": 0.0007548083212959587, + "loss": 2.3634, + "step": 1526 + }, + { + "epoch": 0.3501490483833983, + "grad_norm": 0.41879786801154045, + "learning_rate": 0.0007544887438064862, + "loss": 2.6072, + "step": 1527 + }, + { + "epoch": 0.35037835358862646, + "grad_norm": 0.47146569843168024, + "learning_rate": 0.000754169025942725, + "loss": 2.5274, + "step": 1528 + }, + { + "epoch": 0.35060765879385464, + "grad_norm": 0.5061033374669957, + "learning_rate": 0.0007538491678810294, + "loss": 2.4791, + "step": 1529 + }, + { + "epoch": 0.35083696399908276, + "grad_norm": 0.42678399925122107, + "learning_rate": 0.0007535291697978309, + "loss": 2.6463, + "step": 1530 + }, + { + "epoch": 0.35106626920431094, + "grad_norm": 0.4203836815522087, + "learning_rate": 0.0007532090318696381, + "loss": 2.4568, + "step": 1531 + }, + { + "epoch": 0.3512955744095391, + "grad_norm": 0.45122942696118906, + "learning_rate": 0.0007528887542730374, + "loss": 2.4785, + "step": 1532 + }, + { + "epoch": 0.35152487961476725, + "grad_norm": 0.4253041357420022, + "learning_rate": 0.0007525683371846913, + "loss": 2.5003, + "step": 1533 + }, + { + "epoch": 0.35175418481999543, + "grad_norm": 0.4853202306237496, + "learning_rate": 0.00075224778078134, + "loss": 2.5132, + "step": 1534 + }, + { + "epoch": 0.35198349002522356, + "grad_norm": 0.49605755153863784, + "learning_rate": 0.0007519270852398001, + "loss": 2.4174, + "step": 1535 + }, + { + "epoch": 0.35221279523045174, + "grad_norm": 0.5112977322415262, + "learning_rate": 0.0007516062507369654, + "loss": 2.6789, + "step": 1536 + }, + { + "epoch": 0.3524421004356799, + "grad_norm": 0.3966874178093317, + "learning_rate": 0.000751285277449806, + "loss": 2.481, + "step": 1537 + }, + { + "epoch": 0.35267140564090804, + "grad_norm": 0.4755627637003679, + "learning_rate": 0.0007509641655553683, + "loss": 2.5351, + "step": 1538 + }, + { + "epoch": 0.3529007108461362, + "grad_norm": 0.4494825190386488, + "learning_rate": 0.0007506429152307756, + "loss": 2.4213, + "step": 1539 + }, + { + "epoch": 0.35313001605136435, + "grad_norm": 0.45359869861321644, + "learning_rate": 0.0007503215266532276, + "loss": 2.6156, + "step": 1540 + }, + { + "epoch": 0.35335932125659253, + "grad_norm": 0.4559465246624271, + "learning_rate": 0.00075, + "loss": 2.5933, + "step": 1541 + }, + { + "epoch": 0.3535886264618207, + "grad_norm": 0.4695989657807095, + "learning_rate": 0.0007496783354484449, + "loss": 2.561, + "step": 1542 + }, + { + "epoch": 0.35381793166704884, + "grad_norm": 0.41477783501836835, + "learning_rate": 0.00074935653317599, + "loss": 2.4859, + "step": 1543 + }, + { + "epoch": 0.354047236872277, + "grad_norm": 0.4236573113113202, + "learning_rate": 0.0007490345933601394, + "loss": 2.5304, + "step": 1544 + }, + { + "epoch": 0.35427654207750514, + "grad_norm": 0.4932144267610413, + "learning_rate": 0.000748712516178473, + "loss": 2.5177, + "step": 1545 + }, + { + "epoch": 0.3545058472827333, + "grad_norm": 0.44112843729278184, + "learning_rate": 0.0007483903018086466, + "loss": 2.4919, + "step": 1546 + }, + { + "epoch": 0.3547351524879615, + "grad_norm": 0.43972036558992045, + "learning_rate": 0.0007480679504283911, + "loss": 2.6384, + "step": 1547 + }, + { + "epoch": 0.35496445769318963, + "grad_norm": 0.4602594774517598, + "learning_rate": 0.0007477454622155135, + "loss": 2.6439, + "step": 1548 + }, + { + "epoch": 0.3551937628984178, + "grad_norm": 0.46762597187298294, + "learning_rate": 0.0007474228373478964, + "loss": 2.5872, + "step": 1549 + }, + { + "epoch": 0.35542306810364593, + "grad_norm": 0.45152376609426587, + "learning_rate": 0.0007471000760034972, + "loss": 2.4519, + "step": 1550 + }, + { + "epoch": 0.3556523733088741, + "grad_norm": 0.4279672587478386, + "learning_rate": 0.0007467771783603492, + "loss": 2.628, + "step": 1551 + }, + { + "epoch": 0.3558816785141023, + "grad_norm": 0.44051093865626095, + "learning_rate": 0.0007464541445965602, + "loss": 2.6212, + "step": 1552 + }, + { + "epoch": 0.3561109837193304, + "grad_norm": 0.4132637820407449, + "learning_rate": 0.0007461309748903138, + "loss": 2.5144, + "step": 1553 + }, + { + "epoch": 0.3563402889245586, + "grad_norm": 0.4382601095486323, + "learning_rate": 0.0007458076694198681, + "loss": 2.522, + "step": 1554 + }, + { + "epoch": 0.3565695941297867, + "grad_norm": 0.4520843297991037, + "learning_rate": 0.0007454842283635562, + "loss": 2.5574, + "step": 1555 + }, + { + "epoch": 0.3567988993350149, + "grad_norm": 0.4306977969895921, + "learning_rate": 0.0007451606518997862, + "loss": 2.4198, + "step": 1556 + }, + { + "epoch": 0.3570282045402431, + "grad_norm": 0.4347956299346095, + "learning_rate": 0.0007448369402070404, + "loss": 2.5165, + "step": 1557 + }, + { + "epoch": 0.3572575097454712, + "grad_norm": 0.43949985844034534, + "learning_rate": 0.0007445130934638764, + "loss": 2.4824, + "step": 1558 + }, + { + "epoch": 0.3574868149506994, + "grad_norm": 0.45142878921409, + "learning_rate": 0.0007441891118489254, + "loss": 2.5829, + "step": 1559 + }, + { + "epoch": 0.3577161201559275, + "grad_norm": 0.4596560327080393, + "learning_rate": 0.0007438649955408941, + "loss": 2.6087, + "step": 1560 + }, + { + "epoch": 0.3579454253611557, + "grad_norm": 0.468380296038909, + "learning_rate": 0.0007435407447185622, + "loss": 2.4966, + "step": 1561 + }, + { + "epoch": 0.3581747305663839, + "grad_norm": 0.7461400321878416, + "learning_rate": 0.000743216359560785, + "loss": 2.5929, + "step": 1562 + }, + { + "epoch": 0.358404035771612, + "grad_norm": 0.4187905948102027, + "learning_rate": 0.0007428918402464908, + "loss": 2.5827, + "step": 1563 + }, + { + "epoch": 0.3586333409768402, + "grad_norm": 0.5679216416678062, + "learning_rate": 0.0007425671869546822, + "loss": 2.6221, + "step": 1564 + }, + { + "epoch": 0.3588626461820683, + "grad_norm": 0.4296229899001695, + "learning_rate": 0.0007422423998644359, + "loss": 2.5272, + "step": 1565 + }, + { + "epoch": 0.3590919513872965, + "grad_norm": 0.4947674316523603, + "learning_rate": 0.0007419174791549023, + "loss": 2.5719, + "step": 1566 + }, + { + "epoch": 0.3593212565925247, + "grad_norm": 0.4405843114799296, + "learning_rate": 0.0007415924250053055, + "loss": 2.5735, + "step": 1567 + }, + { + "epoch": 0.3595505617977528, + "grad_norm": 0.4745365070429128, + "learning_rate": 0.0007412672375949431, + "loss": 2.5088, + "step": 1568 + }, + { + "epoch": 0.359779867002981, + "grad_norm": 0.4273545052400285, + "learning_rate": 0.0007409419171031865, + "loss": 2.5723, + "step": 1569 + }, + { + "epoch": 0.3600091722082091, + "grad_norm": 0.43980388008120397, + "learning_rate": 0.0007406164637094801, + "loss": 2.4507, + "step": 1570 + }, + { + "epoch": 0.3602384774134373, + "grad_norm": 0.43856996222845607, + "learning_rate": 0.0007402908775933419, + "loss": 2.4909, + "step": 1571 + }, + { + "epoch": 0.36046778261866547, + "grad_norm": 0.43402018745562054, + "learning_rate": 0.0007399651589343632, + "loss": 2.5719, + "step": 1572 + }, + { + "epoch": 0.3606970878238936, + "grad_norm": 0.4571477355779323, + "learning_rate": 0.0007396393079122077, + "loss": 2.5061, + "step": 1573 + }, + { + "epoch": 0.3609263930291218, + "grad_norm": 0.46735035475603165, + "learning_rate": 0.0007393133247066133, + "loss": 2.4561, + "step": 1574 + }, + { + "epoch": 0.3611556982343499, + "grad_norm": 0.4003758320792675, + "learning_rate": 0.0007389872094973896, + "loss": 2.5363, + "step": 1575 + }, + { + "epoch": 0.3613850034395781, + "grad_norm": 0.41225977125194024, + "learning_rate": 0.00073866096246442, + "loss": 2.568, + "step": 1576 + }, + { + "epoch": 0.36161430864480626, + "grad_norm": 0.4612906905312467, + "learning_rate": 0.00073833458378766, + "loss": 2.5674, + "step": 1577 + }, + { + "epoch": 0.3618436138500344, + "grad_norm": 0.4338209170038769, + "learning_rate": 0.0007380080736471378, + "loss": 2.5969, + "step": 1578 + }, + { + "epoch": 0.36207291905526257, + "grad_norm": 0.4349904726123216, + "learning_rate": 0.0007376814322229544, + "loss": 2.5602, + "step": 1579 + }, + { + "epoch": 0.3623022242604907, + "grad_norm": 0.41453883377822115, + "learning_rate": 0.0007373546596952828, + "loss": 2.4226, + "step": 1580 + }, + { + "epoch": 0.36253152946571887, + "grad_norm": 0.5066049556280079, + "learning_rate": 0.0007370277562443688, + "loss": 2.538, + "step": 1581 + }, + { + "epoch": 0.36276083467094705, + "grad_norm": 0.4353140486932172, + "learning_rate": 0.00073670072205053, + "loss": 2.5708, + "step": 1582 + }, + { + "epoch": 0.3629901398761752, + "grad_norm": 0.4545146892761879, + "learning_rate": 0.0007363735572941564, + "loss": 2.5234, + "step": 1583 + }, + { + "epoch": 0.36321944508140336, + "grad_norm": 0.45881450616755687, + "learning_rate": 0.0007360462621557095, + "loss": 2.5672, + "step": 1584 + }, + { + "epoch": 0.3634487502866315, + "grad_norm": 0.4394160329537354, + "learning_rate": 0.0007357188368157236, + "loss": 2.4832, + "step": 1585 + }, + { + "epoch": 0.36367805549185966, + "grad_norm": 0.4354008541203151, + "learning_rate": 0.0007353912814548042, + "loss": 2.5173, + "step": 1586 + }, + { + "epoch": 0.36390736069708784, + "grad_norm": 0.46602563211587283, + "learning_rate": 0.0007350635962536284, + "loss": 2.5723, + "step": 1587 + }, + { + "epoch": 0.36413666590231597, + "grad_norm": 0.4261726748583994, + "learning_rate": 0.0007347357813929454, + "loss": 2.4962, + "step": 1588 + }, + { + "epoch": 0.36436597110754415, + "grad_norm": 0.41926435681220803, + "learning_rate": 0.0007344078370535756, + "loss": 2.5184, + "step": 1589 + }, + { + "epoch": 0.3645952763127723, + "grad_norm": 0.4365965231651376, + "learning_rate": 0.000734079763416411, + "loss": 2.5325, + "step": 1590 + }, + { + "epoch": 0.36482458151800046, + "grad_norm": 0.4458088478786541, + "learning_rate": 0.0007337515606624148, + "loss": 2.6198, + "step": 1591 + }, + { + "epoch": 0.36505388672322864, + "grad_norm": 0.4394659036515887, + "learning_rate": 0.0007334232289726212, + "loss": 2.5231, + "step": 1592 + }, + { + "epoch": 0.36528319192845676, + "grad_norm": 0.4158965969095272, + "learning_rate": 0.0007330947685281362, + "loss": 2.4199, + "step": 1593 + }, + { + "epoch": 0.36551249713368494, + "grad_norm": 0.4115465431658202, + "learning_rate": 0.000732766179510136, + "loss": 2.6258, + "step": 1594 + }, + { + "epoch": 0.36574180233891307, + "grad_norm": 0.4500296154450618, + "learning_rate": 0.0007324374620998682, + "loss": 2.338, + "step": 1595 + }, + { + "epoch": 0.36597110754414125, + "grad_norm": 0.4145295232934788, + "learning_rate": 0.0007321086164786512, + "loss": 2.4627, + "step": 1596 + }, + { + "epoch": 0.36620041274936943, + "grad_norm": 0.4289288611462709, + "learning_rate": 0.000731779642827874, + "loss": 2.6636, + "step": 1597 + }, + { + "epoch": 0.36642971795459756, + "grad_norm": 0.4842213197157591, + "learning_rate": 0.0007314505413289963, + "loss": 2.4307, + "step": 1598 + }, + { + "epoch": 0.36665902315982574, + "grad_norm": 0.4639909712140274, + "learning_rate": 0.0007311213121635483, + "loss": 2.5485, + "step": 1599 + }, + { + "epoch": 0.36688832836505386, + "grad_norm": 0.445409622094119, + "learning_rate": 0.0007307919555131306, + "loss": 2.5191, + "step": 1600 + }, + { + "epoch": 0.36711763357028204, + "grad_norm": 0.4463872969583862, + "learning_rate": 0.0007304624715594139, + "loss": 2.4523, + "step": 1601 + }, + { + "epoch": 0.3673469387755102, + "grad_norm": 0.4048544598473617, + "learning_rate": 0.0007301328604841399, + "loss": 2.498, + "step": 1602 + }, + { + "epoch": 0.36757624398073835, + "grad_norm": 0.45080435271083497, + "learning_rate": 0.0007298031224691193, + "loss": 2.5663, + "step": 1603 + }, + { + "epoch": 0.36780554918596653, + "grad_norm": 0.43960097886649024, + "learning_rate": 0.0007294732576962335, + "loss": 2.7083, + "step": 1604 + }, + { + "epoch": 0.36803485439119465, + "grad_norm": 0.43976525332200117, + "learning_rate": 0.0007291432663474339, + "loss": 2.6496, + "step": 1605 + }, + { + "epoch": 0.36826415959642284, + "grad_norm": 0.44166155112178607, + "learning_rate": 0.0007288131486047414, + "loss": 2.4218, + "step": 1606 + }, + { + "epoch": 0.368493464801651, + "grad_norm": 0.4423272455571056, + "learning_rate": 0.0007284829046502467, + "loss": 2.6015, + "step": 1607 + }, + { + "epoch": 0.36872277000687914, + "grad_norm": 0.40188962269176204, + "learning_rate": 0.0007281525346661102, + "loss": 2.5019, + "step": 1608 + }, + { + "epoch": 0.3689520752121073, + "grad_norm": 0.42910358021069883, + "learning_rate": 0.0007278220388345619, + "loss": 2.4727, + "step": 1609 + }, + { + "epoch": 0.36918138041733545, + "grad_norm": 0.44270240426052976, + "learning_rate": 0.0007274914173379009, + "loss": 2.542, + "step": 1610 + }, + { + "epoch": 0.36941068562256363, + "grad_norm": 0.4598812174459891, + "learning_rate": 0.0007271606703584958, + "loss": 2.408, + "step": 1611 + }, + { + "epoch": 0.3696399908277918, + "grad_norm": 0.41114614916045544, + "learning_rate": 0.0007268297980787846, + "loss": 2.4222, + "step": 1612 + }, + { + "epoch": 0.36986929603301993, + "grad_norm": 0.4536226260467649, + "learning_rate": 0.000726498800681274, + "loss": 2.4005, + "step": 1613 + }, + { + "epoch": 0.3700986012382481, + "grad_norm": 0.4372113806992469, + "learning_rate": 0.0007261676783485402, + "loss": 2.4896, + "step": 1614 + }, + { + "epoch": 0.37032790644347624, + "grad_norm": 0.41779259246575, + "learning_rate": 0.0007258364312632279, + "loss": 2.5009, + "step": 1615 + }, + { + "epoch": 0.3705572116487044, + "grad_norm": 0.4353943542692925, + "learning_rate": 0.0007255050596080509, + "loss": 2.5263, + "step": 1616 + }, + { + "epoch": 0.3707865168539326, + "grad_norm": 0.46040696949221105, + "learning_rate": 0.0007251735635657915, + "loss": 2.6131, + "step": 1617 + }, + { + "epoch": 0.3710158220591607, + "grad_norm": 0.42307678534646237, + "learning_rate": 0.0007248419433193007, + "loss": 2.5097, + "step": 1618 + }, + { + "epoch": 0.3712451272643889, + "grad_norm": 0.4315978713465072, + "learning_rate": 0.000724510199051498, + "loss": 2.5211, + "step": 1619 + }, + { + "epoch": 0.37147443246961703, + "grad_norm": 0.4272718859490884, + "learning_rate": 0.0007241783309453714, + "loss": 2.5073, + "step": 1620 + }, + { + "epoch": 0.3717037376748452, + "grad_norm": 0.4354028164449237, + "learning_rate": 0.0007238463391839769, + "loss": 2.5873, + "step": 1621 + }, + { + "epoch": 0.3719330428800734, + "grad_norm": 0.4448220645445737, + "learning_rate": 0.0007235142239504391, + "loss": 2.6474, + "step": 1622 + }, + { + "epoch": 0.3721623480853015, + "grad_norm": 0.4099816682421195, + "learning_rate": 0.0007231819854279508, + "loss": 2.5597, + "step": 1623 + }, + { + "epoch": 0.3723916532905297, + "grad_norm": 0.4352587193987876, + "learning_rate": 0.0007228496237997722, + "loss": 2.4611, + "step": 1624 + }, + { + "epoch": 0.3726209584957579, + "grad_norm": 0.45609161849131213, + "learning_rate": 0.0007225171392492316, + "loss": 2.5315, + "step": 1625 + }, + { + "epoch": 0.372850263700986, + "grad_norm": 0.42523181929940795, + "learning_rate": 0.0007221845319597258, + "loss": 2.4484, + "step": 1626 + }, + { + "epoch": 0.3730795689062142, + "grad_norm": 0.4293347382748502, + "learning_rate": 0.0007218518021147182, + "loss": 2.4379, + "step": 1627 + }, + { + "epoch": 0.3733088741114423, + "grad_norm": 0.47227272394638203, + "learning_rate": 0.0007215189498977409, + "loss": 2.5416, + "step": 1628 + }, + { + "epoch": 0.3735381793166705, + "grad_norm": 0.46714289210314414, + "learning_rate": 0.0007211859754923923, + "loss": 2.4708, + "step": 1629 + }, + { + "epoch": 0.3737674845218987, + "grad_norm": 0.4452914670761041, + "learning_rate": 0.0007208528790823394, + "loss": 2.6833, + "step": 1630 + }, + { + "epoch": 0.3739967897271268, + "grad_norm": 0.45886602843024177, + "learning_rate": 0.0007205196608513158, + "loss": 2.3982, + "step": 1631 + }, + { + "epoch": 0.374226094932355, + "grad_norm": 0.4109357616869329, + "learning_rate": 0.0007201863209831223, + "loss": 2.4476, + "step": 1632 + }, + { + "epoch": 0.3744554001375831, + "grad_norm": 0.41434639128184025, + "learning_rate": 0.0007198528596616272, + "loss": 2.5324, + "step": 1633 + }, + { + "epoch": 0.3746847053428113, + "grad_norm": 0.4333797010582042, + "learning_rate": 0.0007195192770707654, + "loss": 2.5518, + "step": 1634 + }, + { + "epoch": 0.37491401054803947, + "grad_norm": 0.40475795169458634, + "learning_rate": 0.0007191855733945387, + "loss": 2.4924, + "step": 1635 + }, + { + "epoch": 0.3751433157532676, + "grad_norm": 0.40744258509507336, + "learning_rate": 0.0007188517488170159, + "loss": 2.3873, + "step": 1636 + }, + { + "epoch": 0.3753726209584958, + "grad_norm": 0.43988620884930274, + "learning_rate": 0.0007185178035223327, + "loss": 2.5308, + "step": 1637 + }, + { + "epoch": 0.3756019261637239, + "grad_norm": 0.4365397678380721, + "learning_rate": 0.0007181837376946905, + "loss": 2.3408, + "step": 1638 + }, + { + "epoch": 0.3758312313689521, + "grad_norm": 0.4356975940497111, + "learning_rate": 0.0007178495515183583, + "loss": 2.3936, + "step": 1639 + }, + { + "epoch": 0.37606053657418026, + "grad_norm": 0.4496128130866457, + "learning_rate": 0.0007175152451776708, + "loss": 2.3738, + "step": 1640 + }, + { + "epoch": 0.3762898417794084, + "grad_norm": 0.4375977675806726, + "learning_rate": 0.000717180818857029, + "loss": 2.4919, + "step": 1641 + }, + { + "epoch": 0.37651914698463657, + "grad_norm": 0.45003708593875336, + "learning_rate": 0.0007168462727409003, + "loss": 2.3783, + "step": 1642 + }, + { + "epoch": 0.3767484521898647, + "grad_norm": 0.46313812683925815, + "learning_rate": 0.0007165116070138182, + "loss": 2.5096, + "step": 1643 + }, + { + "epoch": 0.37697775739509287, + "grad_norm": 0.45765856790647236, + "learning_rate": 0.0007161768218603821, + "loss": 2.4696, + "step": 1644 + }, + { + "epoch": 0.37720706260032105, + "grad_norm": 0.459312157890754, + "learning_rate": 0.0007158419174652569, + "loss": 2.4643, + "step": 1645 + }, + { + "epoch": 0.3774363678055492, + "grad_norm": 0.465708990480305, + "learning_rate": 0.000715506894013174, + "loss": 2.4202, + "step": 1646 + }, + { + "epoch": 0.37766567301077736, + "grad_norm": 0.4331832828911476, + "learning_rate": 0.00071517175168893, + "loss": 2.5486, + "step": 1647 + }, + { + "epoch": 0.3778949782160055, + "grad_norm": 0.4523616274044757, + "learning_rate": 0.0007148364906773871, + "loss": 2.4667, + "step": 1648 + }, + { + "epoch": 0.37812428342123366, + "grad_norm": 0.4130993264126267, + "learning_rate": 0.0007145011111634732, + "loss": 2.4728, + "step": 1649 + }, + { + "epoch": 0.37835358862646185, + "grad_norm": 0.4350373645580089, + "learning_rate": 0.000714165613332181, + "loss": 2.4738, + "step": 1650 + }, + { + "epoch": 0.37858289383168997, + "grad_norm": 0.4139492610712652, + "learning_rate": 0.0007138299973685694, + "loss": 2.5886, + "step": 1651 + }, + { + "epoch": 0.37881219903691815, + "grad_norm": 0.428278136948053, + "learning_rate": 0.0007134942634577614, + "loss": 2.5677, + "step": 1652 + }, + { + "epoch": 0.3790415042421463, + "grad_norm": 0.43020689990965766, + "learning_rate": 0.0007131584117849459, + "loss": 2.657, + "step": 1653 + }, + { + "epoch": 0.37927080944737446, + "grad_norm": 0.43347614819662494, + "learning_rate": 0.0007128224425353764, + "loss": 2.4519, + "step": 1654 + }, + { + "epoch": 0.37950011465260264, + "grad_norm": 0.4037368959439688, + "learning_rate": 0.0007124863558943713, + "loss": 2.6926, + "step": 1655 + }, + { + "epoch": 0.37972941985783076, + "grad_norm": 0.42618013749633205, + "learning_rate": 0.0007121501520473137, + "loss": 2.4335, + "step": 1656 + }, + { + "epoch": 0.37995872506305894, + "grad_norm": 0.4061876818910433, + "learning_rate": 0.0007118138311796514, + "loss": 2.4657, + "step": 1657 + }, + { + "epoch": 0.38018803026828707, + "grad_norm": 0.42699022356210803, + "learning_rate": 0.0007114773934768965, + "loss": 2.5694, + "step": 1658 + }, + { + "epoch": 0.38041733547351525, + "grad_norm": 0.40644089057986926, + "learning_rate": 0.0007111408391246262, + "loss": 2.4078, + "step": 1659 + }, + { + "epoch": 0.38064664067874343, + "grad_norm": 0.43174240305503125, + "learning_rate": 0.0007108041683084814, + "loss": 2.6085, + "step": 1660 + }, + { + "epoch": 0.38087594588397156, + "grad_norm": 0.4394475596983873, + "learning_rate": 0.0007104673812141675, + "loss": 2.5297, + "step": 1661 + }, + { + "epoch": 0.38110525108919974, + "grad_norm": 0.38531196984056687, + "learning_rate": 0.000710130478027454, + "loss": 2.4336, + "step": 1662 + }, + { + "epoch": 0.38133455629442786, + "grad_norm": 0.414892721324069, + "learning_rate": 0.0007097934589341745, + "loss": 2.4895, + "step": 1663 + }, + { + "epoch": 0.38156386149965604, + "grad_norm": 0.40226933038627805, + "learning_rate": 0.0007094563241202266, + "loss": 2.4438, + "step": 1664 + }, + { + "epoch": 0.3817931667048842, + "grad_norm": 0.4369107954760128, + "learning_rate": 0.0007091190737715711, + "loss": 2.4491, + "step": 1665 + }, + { + "epoch": 0.38202247191011235, + "grad_norm": 0.4080482869369189, + "learning_rate": 0.0007087817080742337, + "loss": 2.6198, + "step": 1666 + }, + { + "epoch": 0.38225177711534053, + "grad_norm": 0.463628129771531, + "learning_rate": 0.0007084442272143026, + "loss": 2.435, + "step": 1667 + }, + { + "epoch": 0.38248108232056865, + "grad_norm": 0.47292916888828035, + "learning_rate": 0.0007081066313779301, + "loss": 2.5277, + "step": 1668 + }, + { + "epoch": 0.38271038752579684, + "grad_norm": 0.43370959846182905, + "learning_rate": 0.000707768920751332, + "loss": 2.477, + "step": 1669 + }, + { + "epoch": 0.382939692731025, + "grad_norm": 0.41370407953223964, + "learning_rate": 0.0007074310955207869, + "loss": 2.5895, + "step": 1670 + }, + { + "epoch": 0.38316899793625314, + "grad_norm": 0.40051156807385807, + "learning_rate": 0.0007070931558726373, + "loss": 2.5624, + "step": 1671 + }, + { + "epoch": 0.3833983031414813, + "grad_norm": 0.4100870795557393, + "learning_rate": 0.0007067551019932879, + "loss": 2.6426, + "step": 1672 + }, + { + "epoch": 0.38362760834670945, + "grad_norm": 0.44277987085186393, + "learning_rate": 0.0007064169340692076, + "loss": 2.5619, + "step": 1673 + }, + { + "epoch": 0.38385691355193763, + "grad_norm": 0.4455167139198393, + "learning_rate": 0.0007060786522869271, + "loss": 2.6223, + "step": 1674 + }, + { + "epoch": 0.3840862187571658, + "grad_norm": 0.41624796878976394, + "learning_rate": 0.0007057402568330407, + "loss": 2.6025, + "step": 1675 + }, + { + "epoch": 0.38431552396239393, + "grad_norm": 0.4881028870655976, + "learning_rate": 0.0007054017478942047, + "loss": 2.3482, + "step": 1676 + }, + { + "epoch": 0.3845448291676221, + "grad_norm": 0.4571683819931176, + "learning_rate": 0.0007050631256571389, + "loss": 2.5196, + "step": 1677 + }, + { + "epoch": 0.38477413437285024, + "grad_norm": 0.45211934698837536, + "learning_rate": 0.0007047243903086245, + "loss": 2.5104, + "step": 1678 + }, + { + "epoch": 0.3850034395780784, + "grad_norm": 0.43017155933884094, + "learning_rate": 0.000704385542035506, + "loss": 2.5025, + "step": 1679 + }, + { + "epoch": 0.3852327447833066, + "grad_norm": 0.44357778050991864, + "learning_rate": 0.0007040465810246897, + "loss": 2.5863, + "step": 1680 + }, + { + "epoch": 0.3854620499885347, + "grad_norm": 0.42309622826578624, + "learning_rate": 0.000703707507463144, + "loss": 2.5437, + "step": 1681 + }, + { + "epoch": 0.3856913551937629, + "grad_norm": 0.472316889031239, + "learning_rate": 0.0007033683215379002, + "loss": 2.7034, + "step": 1682 + }, + { + "epoch": 0.38592066039899103, + "grad_norm": 0.43845031974716175, + "learning_rate": 0.0007030290234360505, + "loss": 2.4196, + "step": 1683 + }, + { + "epoch": 0.3861499656042192, + "grad_norm": 0.45067633921611083, + "learning_rate": 0.0007026896133447497, + "loss": 2.5708, + "step": 1684 + }, + { + "epoch": 0.3863792708094474, + "grad_norm": 0.4812150109993902, + "learning_rate": 0.0007023500914512139, + "loss": 2.4687, + "step": 1685 + }, + { + "epoch": 0.3866085760146755, + "grad_norm": 0.42470829273159166, + "learning_rate": 0.0007020104579427214, + "loss": 2.5203, + "step": 1686 + }, + { + "epoch": 0.3868378812199037, + "grad_norm": 0.4427180582055635, + "learning_rate": 0.0007016707130066116, + "loss": 2.6594, + "step": 1687 + }, + { + "epoch": 0.3870671864251318, + "grad_norm": 0.43460325118413135, + "learning_rate": 0.0007013308568302854, + "loss": 2.5337, + "step": 1688 + }, + { + "epoch": 0.38729649163036, + "grad_norm": 0.47092156374644994, + "learning_rate": 0.0007009908896012055, + "loss": 2.5771, + "step": 1689 + }, + { + "epoch": 0.3875257968355882, + "grad_norm": 0.4396394200075872, + "learning_rate": 0.0007006508115068954, + "loss": 2.534, + "step": 1690 + }, + { + "epoch": 0.3877551020408163, + "grad_norm": 0.4630576700173389, + "learning_rate": 0.0007003106227349399, + "loss": 2.4684, + "step": 1691 + }, + { + "epoch": 0.3879844072460445, + "grad_norm": 0.426966582493572, + "learning_rate": 0.0006999703234729847, + "loss": 2.6329, + "step": 1692 + }, + { + "epoch": 0.3882137124512726, + "grad_norm": 0.47590243684214334, + "learning_rate": 0.000699629913908737, + "loss": 2.5385, + "step": 1693 + }, + { + "epoch": 0.3884430176565008, + "grad_norm": 0.4301915039344611, + "learning_rate": 0.0006992893942299641, + "loss": 2.3708, + "step": 1694 + }, + { + "epoch": 0.388672322861729, + "grad_norm": 0.43692902398543587, + "learning_rate": 0.0006989487646244943, + "loss": 2.4159, + "step": 1695 + }, + { + "epoch": 0.3889016280669571, + "grad_norm": 0.469330206356673, + "learning_rate": 0.000698608025280217, + "loss": 2.661, + "step": 1696 + }, + { + "epoch": 0.3891309332721853, + "grad_norm": 0.4533697142415042, + "learning_rate": 0.0006982671763850814, + "loss": 2.4781, + "step": 1697 + }, + { + "epoch": 0.3893602384774134, + "grad_norm": 0.4573470025941725, + "learning_rate": 0.0006979262181270976, + "loss": 2.4769, + "step": 1698 + }, + { + "epoch": 0.3895895436826416, + "grad_norm": 0.46699226148744843, + "learning_rate": 0.0006975851506943359, + "loss": 2.5779, + "step": 1699 + }, + { + "epoch": 0.3898188488878698, + "grad_norm": 0.5477273369361095, + "learning_rate": 0.000697243974274927, + "loss": 2.5348, + "step": 1700 + }, + { + "epoch": 0.3900481540930979, + "grad_norm": 0.4612856551450056, + "learning_rate": 0.0006969026890570611, + "loss": 2.5926, + "step": 1701 + }, + { + "epoch": 0.3902774592983261, + "grad_norm": 0.4461172405667249, + "learning_rate": 0.000696561295228989, + "loss": 2.5081, + "step": 1702 + }, + { + "epoch": 0.3905067645035542, + "grad_norm": 0.42302279596501313, + "learning_rate": 0.0006962197929790216, + "loss": 2.4191, + "step": 1703 + }, + { + "epoch": 0.3907360697087824, + "grad_norm": 0.42540614485446515, + "learning_rate": 0.0006958781824955288, + "loss": 2.5209, + "step": 1704 + }, + { + "epoch": 0.39096537491401057, + "grad_norm": 0.44792366877183193, + "learning_rate": 0.0006955364639669409, + "loss": 2.5957, + "step": 1705 + }, + { + "epoch": 0.3911946801192387, + "grad_norm": 0.45103571430268585, + "learning_rate": 0.0006951946375817474, + "loss": 2.4143, + "step": 1706 + }, + { + "epoch": 0.39142398532446687, + "grad_norm": 0.42728447198479874, + "learning_rate": 0.0006948527035284978, + "loss": 2.4848, + "step": 1707 + }, + { + "epoch": 0.391653290529695, + "grad_norm": 0.45689028309076274, + "learning_rate": 0.0006945106619958002, + "loss": 2.4507, + "step": 1708 + }, + { + "epoch": 0.3918825957349232, + "grad_norm": 0.4183385977878037, + "learning_rate": 0.0006941685131723225, + "loss": 2.4406, + "step": 1709 + }, + { + "epoch": 0.39211190094015136, + "grad_norm": 0.4911704181393975, + "learning_rate": 0.0006938262572467919, + "loss": 2.5472, + "step": 1710 + }, + { + "epoch": 0.3923412061453795, + "grad_norm": 0.43899483545564166, + "learning_rate": 0.0006934838944079943, + "loss": 2.5744, + "step": 1711 + }, + { + "epoch": 0.39257051135060766, + "grad_norm": 0.4685165283335075, + "learning_rate": 0.0006931414248447749, + "loss": 2.4391, + "step": 1712 + }, + { + "epoch": 0.3927998165558358, + "grad_norm": 0.4735787517519185, + "learning_rate": 0.0006927988487460378, + "loss": 2.4995, + "step": 1713 + }, + { + "epoch": 0.39302912176106397, + "grad_norm": 0.4694903105159986, + "learning_rate": 0.000692456166300745, + "loss": 2.5952, + "step": 1714 + }, + { + "epoch": 0.39325842696629215, + "grad_norm": 0.41458461659645335, + "learning_rate": 0.0006921133776979186, + "loss": 2.4588, + "step": 1715 + }, + { + "epoch": 0.3934877321715203, + "grad_norm": 0.4418011690502116, + "learning_rate": 0.0006917704831266381, + "loss": 2.5571, + "step": 1716 + }, + { + "epoch": 0.39371703737674846, + "grad_norm": 0.4289830187965507, + "learning_rate": 0.0006914274827760418, + "loss": 2.5643, + "step": 1717 + }, + { + "epoch": 0.39394634258197664, + "grad_norm": 0.4127425781753242, + "learning_rate": 0.0006910843768353266, + "loss": 2.3905, + "step": 1718 + }, + { + "epoch": 0.39417564778720476, + "grad_norm": 0.4116134617808664, + "learning_rate": 0.0006907411654937475, + "loss": 2.3761, + "step": 1719 + }, + { + "epoch": 0.39440495299243294, + "grad_norm": 0.4255339709686001, + "learning_rate": 0.0006903978489406174, + "loss": 2.4926, + "step": 1720 + }, + { + "epoch": 0.39463425819766107, + "grad_norm": 0.41316033032335175, + "learning_rate": 0.0006900544273653075, + "loss": 2.5847, + "step": 1721 + }, + { + "epoch": 0.39486356340288925, + "grad_norm": 0.4307029126632216, + "learning_rate": 0.0006897109009572468, + "loss": 2.5081, + "step": 1722 + }, + { + "epoch": 0.39509286860811743, + "grad_norm": 0.43297067037420306, + "learning_rate": 0.000689367269905922, + "loss": 2.3895, + "step": 1723 + }, + { + "epoch": 0.39532217381334556, + "grad_norm": 0.42732053723818125, + "learning_rate": 0.000689023534400878, + "loss": 2.3119, + "step": 1724 + }, + { + "epoch": 0.39555147901857374, + "grad_norm": 0.4277546933513115, + "learning_rate": 0.0006886796946317168, + "loss": 2.3906, + "step": 1725 + }, + { + "epoch": 0.39578078422380186, + "grad_norm": 0.4135394010464934, + "learning_rate": 0.0006883357507880985, + "loss": 2.4507, + "step": 1726 + }, + { + "epoch": 0.39601008942903004, + "grad_norm": 0.6329219605417933, + "learning_rate": 0.0006879917030597397, + "loss": 2.464, + "step": 1727 + }, + { + "epoch": 0.3962393946342582, + "grad_norm": 0.415594397823351, + "learning_rate": 0.0006876475516364149, + "loss": 2.4672, + "step": 1728 + }, + { + "epoch": 0.39646869983948635, + "grad_norm": 0.4103165760746096, + "learning_rate": 0.0006873032967079561, + "loss": 2.4725, + "step": 1729 + }, + { + "epoch": 0.39669800504471453, + "grad_norm": 0.4437017800728669, + "learning_rate": 0.0006869589384642516, + "loss": 2.6562, + "step": 1730 + }, + { + "epoch": 0.39692731024994266, + "grad_norm": 0.40617569936634396, + "learning_rate": 0.0006866144770952474, + "loss": 2.4586, + "step": 1731 + }, + { + "epoch": 0.39715661545517084, + "grad_norm": 0.42386841484979937, + "learning_rate": 0.000686269912790946, + "loss": 2.4508, + "step": 1732 + }, + { + "epoch": 0.397385920660399, + "grad_norm": 0.8369546659627066, + "learning_rate": 0.0006859252457414067, + "loss": 2.499, + "step": 1733 + }, + { + "epoch": 0.39761522586562714, + "grad_norm": 0.4272925146536701, + "learning_rate": 0.0006855804761367456, + "loss": 2.4908, + "step": 1734 + }, + { + "epoch": 0.3978445310708553, + "grad_norm": 0.44468361679580143, + "learning_rate": 0.0006852356041671351, + "loss": 2.3732, + "step": 1735 + }, + { + "epoch": 0.39807383627608345, + "grad_norm": 0.4609041749531052, + "learning_rate": 0.0006848906300228046, + "loss": 2.5255, + "step": 1736 + }, + { + "epoch": 0.39830314148131163, + "grad_norm": 0.4854508293267031, + "learning_rate": 0.0006845455538940394, + "loss": 2.5623, + "step": 1737 + }, + { + "epoch": 0.3985324466865398, + "grad_norm": 0.4548619056729797, + "learning_rate": 0.0006842003759711811, + "loss": 2.4991, + "step": 1738 + }, + { + "epoch": 0.39876175189176793, + "grad_norm": 0.4478801152646248, + "learning_rate": 0.0006838550964446276, + "loss": 2.5187, + "step": 1739 + }, + { + "epoch": 0.3989910570969961, + "grad_norm": 0.445478138600466, + "learning_rate": 0.0006835097155048329, + "loss": 2.5865, + "step": 1740 + }, + { + "epoch": 0.39922036230222424, + "grad_norm": 0.4249757965491911, + "learning_rate": 0.0006831642333423067, + "loss": 2.393, + "step": 1741 + }, + { + "epoch": 0.3994496675074524, + "grad_norm": 0.4502579581095641, + "learning_rate": 0.0006828186501476144, + "loss": 2.4672, + "step": 1742 + }, + { + "epoch": 0.3996789727126806, + "grad_norm": 0.4566260133080201, + "learning_rate": 0.000682472966111378, + "loss": 2.4447, + "step": 1743 + }, + { + "epoch": 0.3999082779179087, + "grad_norm": 0.42136114092914767, + "learning_rate": 0.0006821271814242738, + "loss": 2.4349, + "step": 1744 + }, + { + "epoch": 0.4001375831231369, + "grad_norm": 0.4416778786475387, + "learning_rate": 0.0006817812962770348, + "loss": 2.6127, + "step": 1745 + }, + { + "epoch": 0.40036688832836503, + "grad_norm": 0.43249632197838456, + "learning_rate": 0.0006814353108604487, + "loss": 2.3862, + "step": 1746 + }, + { + "epoch": 0.4005961935335932, + "grad_norm": 0.44943828109409306, + "learning_rate": 0.0006810892253653589, + "loss": 2.6304, + "step": 1747 + }, + { + "epoch": 0.4008254987388214, + "grad_norm": 0.43898374386957945, + "learning_rate": 0.0006807430399826637, + "loss": 2.552, + "step": 1748 + }, + { + "epoch": 0.4010548039440495, + "grad_norm": 0.519104511141394, + "learning_rate": 0.0006803967549033167, + "loss": 2.549, + "step": 1749 + }, + { + "epoch": 0.4012841091492777, + "grad_norm": 0.500870139181494, + "learning_rate": 0.0006800503703183265, + "loss": 2.6204, + "step": 1750 + }, + { + "epoch": 0.4015134143545058, + "grad_norm": 0.4548349093686276, + "learning_rate": 0.0006797038864187564, + "loss": 2.4573, + "step": 1751 + }, + { + "epoch": 0.401742719559734, + "grad_norm": 0.5489434661967886, + "learning_rate": 0.0006793573033957248, + "loss": 2.531, + "step": 1752 + }, + { + "epoch": 0.4019720247649622, + "grad_norm": 0.47774460085349835, + "learning_rate": 0.0006790106214404043, + "loss": 2.7072, + "step": 1753 + }, + { + "epoch": 0.4022013299701903, + "grad_norm": 0.4241975128153764, + "learning_rate": 0.0006786638407440224, + "loss": 2.4289, + "step": 1754 + }, + { + "epoch": 0.4024306351754185, + "grad_norm": 0.894487124852329, + "learning_rate": 0.0006783169614978614, + "loss": 2.496, + "step": 1755 + }, + { + "epoch": 0.4026599403806466, + "grad_norm": 0.5996831633604233, + "learning_rate": 0.000677969983893257, + "loss": 2.524, + "step": 1756 + }, + { + "epoch": 0.4028892455858748, + "grad_norm": 0.4723132171676838, + "learning_rate": 0.0006776229081216001, + "loss": 2.5824, + "step": 1757 + }, + { + "epoch": 0.403118550791103, + "grad_norm": 0.48296155533142654, + "learning_rate": 0.0006772757343743352, + "loss": 2.4637, + "step": 1758 + }, + { + "epoch": 0.4033478559963311, + "grad_norm": 0.5269316123561463, + "learning_rate": 0.0006769284628429611, + "loss": 2.479, + "step": 1759 + }, + { + "epoch": 0.4035771612015593, + "grad_norm": 0.5901566976139577, + "learning_rate": 0.0006765810937190307, + "loss": 2.5515, + "step": 1760 + }, + { + "epoch": 0.4038064664067874, + "grad_norm": 0.4908342646401223, + "learning_rate": 0.0006762336271941498, + "loss": 2.4328, + "step": 1761 + }, + { + "epoch": 0.4040357716120156, + "grad_norm": 0.47201084915998176, + "learning_rate": 0.0006758860634599792, + "loss": 2.4473, + "step": 1762 + }, + { + "epoch": 0.4042650768172438, + "grad_norm": 0.6146451010290656, + "learning_rate": 0.0006755384027082326, + "loss": 2.4185, + "step": 1763 + }, + { + "epoch": 0.4044943820224719, + "grad_norm": 0.5581858638744899, + "learning_rate": 0.0006751906451306774, + "loss": 2.4407, + "step": 1764 + }, + { + "epoch": 0.4047236872277001, + "grad_norm": 0.5023355985397836, + "learning_rate": 0.0006748427909191342, + "loss": 2.4614, + "step": 1765 + }, + { + "epoch": 0.4049529924329282, + "grad_norm": 1.0403942529035695, + "learning_rate": 0.0006744948402654771, + "loss": 2.4858, + "step": 1766 + }, + { + "epoch": 0.4051822976381564, + "grad_norm": 0.5248572817865174, + "learning_rate": 0.0006741467933616335, + "loss": 2.5791, + "step": 1767 + }, + { + "epoch": 0.40541160284338457, + "grad_norm": 0.5684777556700432, + "learning_rate": 0.0006737986503995832, + "loss": 2.431, + "step": 1768 + }, + { + "epoch": 0.4056409080486127, + "grad_norm": 0.4766230865415869, + "learning_rate": 0.0006734504115713604, + "loss": 2.4359, + "step": 1769 + }, + { + "epoch": 0.4058702132538409, + "grad_norm": 0.47302510952383714, + "learning_rate": 0.0006731020770690505, + "loss": 2.5272, + "step": 1770 + }, + { + "epoch": 0.406099518459069, + "grad_norm": 0.4720106737704816, + "learning_rate": 0.0006727536470847932, + "loss": 2.44, + "step": 1771 + }, + { + "epoch": 0.4063288236642972, + "grad_norm": 0.48903980690567606, + "learning_rate": 0.0006724051218107795, + "loss": 2.5806, + "step": 1772 + }, + { + "epoch": 0.40655812886952536, + "grad_norm": 0.44225775856182553, + "learning_rate": 0.000672056501439254, + "loss": 2.5479, + "step": 1773 + }, + { + "epoch": 0.4067874340747535, + "grad_norm": 0.4630554816994286, + "learning_rate": 0.0006717077861625134, + "loss": 2.4889, + "step": 1774 + }, + { + "epoch": 0.40701673927998167, + "grad_norm": 0.43636390000728753, + "learning_rate": 0.0006713589761729063, + "loss": 2.5195, + "step": 1775 + }, + { + "epoch": 0.4072460444852098, + "grad_norm": 0.43840431048689515, + "learning_rate": 0.0006710100716628344, + "loss": 2.5513, + "step": 1776 + }, + { + "epoch": 0.40747534969043797, + "grad_norm": 0.43260507044870433, + "learning_rate": 0.0006706610728247508, + "loss": 2.4661, + "step": 1777 + }, + { + "epoch": 0.40770465489566615, + "grad_norm": 0.4258916058955056, + "learning_rate": 0.0006703119798511612, + "loss": 2.3897, + "step": 1778 + }, + { + "epoch": 0.4079339601008943, + "grad_norm": 0.4886634000721582, + "learning_rate": 0.0006699627929346227, + "loss": 2.5824, + "step": 1779 + }, + { + "epoch": 0.40816326530612246, + "grad_norm": 1.11281501740796, + "learning_rate": 0.0006696135122677447, + "loss": 2.5886, + "step": 1780 + }, + { + "epoch": 0.4083925705113506, + "grad_norm": 0.3866930280352948, + "learning_rate": 0.0006692641380431879, + "loss": 2.4572, + "step": 1781 + }, + { + "epoch": 0.40862187571657876, + "grad_norm": 0.871739426665547, + "learning_rate": 0.0006689146704536646, + "loss": 2.615, + "step": 1782 + }, + { + "epoch": 0.40885118092180694, + "grad_norm": 0.4297647437904479, + "learning_rate": 0.0006685651096919393, + "loss": 2.507, + "step": 1783 + }, + { + "epoch": 0.40908048612703507, + "grad_norm": 0.4154358808933437, + "learning_rate": 0.0006682154559508269, + "loss": 2.5351, + "step": 1784 + }, + { + "epoch": 0.40930979133226325, + "grad_norm": 0.39218030497977924, + "learning_rate": 0.0006678657094231944, + "loss": 2.5271, + "step": 1785 + }, + { + "epoch": 0.4095390965374914, + "grad_norm": 0.5210818257844745, + "learning_rate": 0.0006675158703019593, + "loss": 2.5977, + "step": 1786 + }, + { + "epoch": 0.40976840174271956, + "grad_norm": 0.4545505299047796, + "learning_rate": 0.0006671659387800909, + "loss": 2.4874, + "step": 1787 + }, + { + "epoch": 0.40999770694794774, + "grad_norm": 0.4617554144062717, + "learning_rate": 0.0006668159150506089, + "loss": 2.5314, + "step": 1788 + }, + { + "epoch": 0.41022701215317586, + "grad_norm": 0.42577796014582125, + "learning_rate": 0.000666465799306584, + "loss": 2.5731, + "step": 1789 + }, + { + "epoch": 0.41045631735840404, + "grad_norm": 0.462524419420046, + "learning_rate": 0.0006661155917411378, + "loss": 2.4951, + "step": 1790 + }, + { + "epoch": 0.41068562256363217, + "grad_norm": 0.42765150168119836, + "learning_rate": 0.0006657652925474423, + "loss": 2.5843, + "step": 1791 + }, + { + "epoch": 0.41091492776886035, + "grad_norm": 0.43609021265223186, + "learning_rate": 0.0006654149019187205, + "loss": 2.5347, + "step": 1792 + }, + { + "epoch": 0.41114423297408853, + "grad_norm": 0.41399616120666194, + "learning_rate": 0.000665064420048245, + "loss": 2.4258, + "step": 1793 + }, + { + "epoch": 0.41137353817931666, + "grad_norm": 0.5155459745654638, + "learning_rate": 0.0006647138471293399, + "loss": 2.5682, + "step": 1794 + }, + { + "epoch": 0.41160284338454484, + "grad_norm": 0.4555520552954321, + "learning_rate": 0.0006643631833553785, + "loss": 2.508, + "step": 1795 + }, + { + "epoch": 0.41183214858977296, + "grad_norm": 0.4691744012618711, + "learning_rate": 0.0006640124289197845, + "loss": 2.5483, + "step": 1796 + }, + { + "epoch": 0.41206145379500114, + "grad_norm": 0.42455930757253346, + "learning_rate": 0.000663661584016032, + "loss": 2.5345, + "step": 1797 + }, + { + "epoch": 0.4122907590002293, + "grad_norm": 0.4098516488562056, + "learning_rate": 0.0006633106488376445, + "loss": 2.6348, + "step": 1798 + }, + { + "epoch": 0.41252006420545745, + "grad_norm": 0.4293035414777929, + "learning_rate": 0.0006629596235781957, + "loss": 2.5364, + "step": 1799 + }, + { + "epoch": 0.41274936941068563, + "grad_norm": 0.4320301110589936, + "learning_rate": 0.0006626085084313083, + "loss": 2.6932, + "step": 1800 + }, + { + "epoch": 0.41297867461591375, + "grad_norm": 0.4685182608190543, + "learning_rate": 0.0006622573035906556, + "loss": 2.2801, + "step": 1801 + }, + { + "epoch": 0.41320797982114194, + "grad_norm": 0.44010287118391456, + "learning_rate": 0.0006619060092499597, + "loss": 2.5765, + "step": 1802 + }, + { + "epoch": 0.4134372850263701, + "grad_norm": 0.4362671536921874, + "learning_rate": 0.0006615546256029921, + "loss": 2.3292, + "step": 1803 + }, + { + "epoch": 0.41366659023159824, + "grad_norm": 0.48823520704001044, + "learning_rate": 0.0006612031528435737, + "loss": 2.6275, + "step": 1804 + }, + { + "epoch": 0.4138958954368264, + "grad_norm": 0.4332895220565784, + "learning_rate": 0.0006608515911655743, + "loss": 2.4323, + "step": 1805 + }, + { + "epoch": 0.41412520064205455, + "grad_norm": 0.4892138236724828, + "learning_rate": 0.0006604999407629137, + "loss": 2.578, + "step": 1806 + }, + { + "epoch": 0.41435450584728273, + "grad_norm": 0.425831859353221, + "learning_rate": 0.0006601482018295591, + "loss": 2.5654, + "step": 1807 + }, + { + "epoch": 0.4145838110525109, + "grad_norm": 0.4581336258013433, + "learning_rate": 0.0006597963745595276, + "loss": 2.5274, + "step": 1808 + }, + { + "epoch": 0.41481311625773903, + "grad_norm": 0.44363929622100057, + "learning_rate": 0.0006594444591468851, + "loss": 2.5189, + "step": 1809 + }, + { + "epoch": 0.4150424214629672, + "grad_norm": 0.40804976774722296, + "learning_rate": 0.0006590924557857454, + "loss": 2.3367, + "step": 1810 + }, + { + "epoch": 0.4152717266681954, + "grad_norm": 0.43698627893635295, + "learning_rate": 0.0006587403646702713, + "loss": 2.4844, + "step": 1811 + }, + { + "epoch": 0.4155010318734235, + "grad_norm": 0.4178464656677865, + "learning_rate": 0.0006583881859946739, + "loss": 2.3568, + "step": 1812 + }, + { + "epoch": 0.4157303370786517, + "grad_norm": 0.4324029140074503, + "learning_rate": 0.0006580359199532126, + "loss": 2.591, + "step": 1813 + }, + { + "epoch": 0.4159596422838798, + "grad_norm": 0.4092805108787261, + "learning_rate": 0.0006576835667401953, + "loss": 2.4604, + "step": 1814 + }, + { + "epoch": 0.416188947489108, + "grad_norm": 0.41850205194100015, + "learning_rate": 0.000657331126549977, + "loss": 2.4917, + "step": 1815 + }, + { + "epoch": 0.4164182526943362, + "grad_norm": 0.4503415244594472, + "learning_rate": 0.000656978599576962, + "loss": 2.549, + "step": 1816 + }, + { + "epoch": 0.4166475578995643, + "grad_norm": 0.403692478611071, + "learning_rate": 0.0006566259860156014, + "loss": 2.4723, + "step": 1817 + }, + { + "epoch": 0.4168768631047925, + "grad_norm": 0.4329704494231919, + "learning_rate": 0.0006562732860603949, + "loss": 2.5584, + "step": 1818 + }, + { + "epoch": 0.4171061683100206, + "grad_norm": 0.49647467252799854, + "learning_rate": 0.0006559204999058888, + "loss": 2.4586, + "step": 1819 + }, + { + "epoch": 0.4173354735152488, + "grad_norm": 0.47475185732738534, + "learning_rate": 0.0006555676277466782, + "loss": 2.638, + "step": 1820 + }, + { + "epoch": 0.417564778720477, + "grad_norm": 0.5133986747134613, + "learning_rate": 0.0006552146697774049, + "loss": 2.612, + "step": 1821 + }, + { + "epoch": 0.4177940839257051, + "grad_norm": 0.4642463727428177, + "learning_rate": 0.0006548616261927578, + "loss": 2.4878, + "step": 1822 + }, + { + "epoch": 0.4180233891309333, + "grad_norm": 0.4395973130839919, + "learning_rate": 0.0006545084971874737, + "loss": 2.346, + "step": 1823 + }, + { + "epoch": 0.4182526943361614, + "grad_norm": 0.42962092225905674, + "learning_rate": 0.0006541552829563362, + "loss": 2.3417, + "step": 1824 + }, + { + "epoch": 0.4184819995413896, + "grad_norm": 0.5043655725079967, + "learning_rate": 0.0006538019836941758, + "loss": 2.3713, + "step": 1825 + }, + { + "epoch": 0.4187113047466178, + "grad_norm": 0.5059160363791787, + "learning_rate": 0.0006534485995958699, + "loss": 2.3757, + "step": 1826 + }, + { + "epoch": 0.4189406099518459, + "grad_norm": 0.43573373766503226, + "learning_rate": 0.0006530951308563431, + "loss": 2.514, + "step": 1827 + }, + { + "epoch": 0.4191699151570741, + "grad_norm": 0.46238049354091737, + "learning_rate": 0.0006527415776705662, + "loss": 2.5381, + "step": 1828 + }, + { + "epoch": 0.4193992203623022, + "grad_norm": 0.4066218445786038, + "learning_rate": 0.0006523879402335567, + "loss": 2.6294, + "step": 1829 + }, + { + "epoch": 0.4196285255675304, + "grad_norm": 0.4615357753487828, + "learning_rate": 0.0006520342187403789, + "loss": 2.4483, + "step": 1830 + }, + { + "epoch": 0.41985783077275857, + "grad_norm": 0.4170880323774979, + "learning_rate": 0.0006516804133861429, + "loss": 2.5587, + "step": 1831 + }, + { + "epoch": 0.4200871359779867, + "grad_norm": 0.46755122947608335, + "learning_rate": 0.0006513265243660057, + "loss": 2.4443, + "step": 1832 + }, + { + "epoch": 0.4203164411832149, + "grad_norm": 0.5308616851364631, + "learning_rate": 0.0006509725518751698, + "loss": 2.6544, + "step": 1833 + }, + { + "epoch": 0.420545746388443, + "grad_norm": 0.4794382401355046, + "learning_rate": 0.0006506184961088844, + "loss": 2.5343, + "step": 1834 + }, + { + "epoch": 0.4207750515936712, + "grad_norm": 0.4518163256388653, + "learning_rate": 0.0006502643572624438, + "loss": 2.6005, + "step": 1835 + }, + { + "epoch": 0.42100435679889936, + "grad_norm": 0.5021839706745717, + "learning_rate": 0.0006499101355311891, + "loss": 2.6965, + "step": 1836 + }, + { + "epoch": 0.4212336620041275, + "grad_norm": 0.42758226833723717, + "learning_rate": 0.0006495558311105064, + "loss": 2.3624, + "step": 1837 + }, + { + "epoch": 0.42146296720935567, + "grad_norm": 0.5606092194724246, + "learning_rate": 0.0006492014441958277, + "loss": 2.5779, + "step": 1838 + }, + { + "epoch": 0.4216922724145838, + "grad_norm": 0.487928524685822, + "learning_rate": 0.0006488469749826305, + "loss": 2.6181, + "step": 1839 + }, + { + "epoch": 0.42192157761981197, + "grad_norm": 0.49972484181909727, + "learning_rate": 0.0006484924236664376, + "loss": 2.501, + "step": 1840 + }, + { + "epoch": 0.42215088282504015, + "grad_norm": 0.5253044138802362, + "learning_rate": 0.000648137790442817, + "loss": 2.5728, + "step": 1841 + }, + { + "epoch": 0.4223801880302683, + "grad_norm": 0.44619554824779833, + "learning_rate": 0.0006477830755073822, + "loss": 2.5379, + "step": 1842 + }, + { + "epoch": 0.42260949323549646, + "grad_norm": 0.434634198187184, + "learning_rate": 0.0006474282790557916, + "loss": 2.5478, + "step": 1843 + }, + { + "epoch": 0.4228387984407246, + "grad_norm": 0.38525911937489127, + "learning_rate": 0.0006470734012837485, + "loss": 2.4406, + "step": 1844 + }, + { + "epoch": 0.42306810364595276, + "grad_norm": 0.4086822523918052, + "learning_rate": 0.000646718442387001, + "loss": 2.5527, + "step": 1845 + }, + { + "epoch": 0.42329740885118095, + "grad_norm": 0.43633439106606475, + "learning_rate": 0.0006463634025613422, + "loss": 2.481, + "step": 1846 + }, + { + "epoch": 0.42352671405640907, + "grad_norm": 0.40435375269466384, + "learning_rate": 0.0006460082820026094, + "loss": 2.4889, + "step": 1847 + }, + { + "epoch": 0.42375601926163725, + "grad_norm": 0.4457135611323216, + "learning_rate": 0.0006456530809066852, + "loss": 2.5846, + "step": 1848 + }, + { + "epoch": 0.4239853244668654, + "grad_norm": 0.4068160664218977, + "learning_rate": 0.0006452977994694959, + "loss": 2.4428, + "step": 1849 + }, + { + "epoch": 0.42421462967209356, + "grad_norm": 0.452706186623504, + "learning_rate": 0.0006449424378870123, + "loss": 2.4838, + "step": 1850 + }, + { + "epoch": 0.42444393487732174, + "grad_norm": 0.43256943266083925, + "learning_rate": 0.0006445869963552496, + "loss": 2.6111, + "step": 1851 + }, + { + "epoch": 0.42467324008254986, + "grad_norm": 0.4391362283452171, + "learning_rate": 0.0006442314750702666, + "loss": 2.5686, + "step": 1852 + }, + { + "epoch": 0.42490254528777804, + "grad_norm": 0.42337752734490475, + "learning_rate": 0.0006438758742281672, + "loss": 2.486, + "step": 1853 + }, + { + "epoch": 0.42513185049300617, + "grad_norm": 0.4293872105681496, + "learning_rate": 0.0006435201940250977, + "loss": 2.513, + "step": 1854 + }, + { + "epoch": 0.42536115569823435, + "grad_norm": 0.42773430951431846, + "learning_rate": 0.0006431644346572495, + "loss": 2.5119, + "step": 1855 + }, + { + "epoch": 0.42559046090346253, + "grad_norm": 0.4371607866758792, + "learning_rate": 0.0006428085963208567, + "loss": 2.529, + "step": 1856 + }, + { + "epoch": 0.42581976610869066, + "grad_norm": 0.43321832259656445, + "learning_rate": 0.0006424526792121974, + "loss": 2.6172, + "step": 1857 + }, + { + "epoch": 0.42604907131391884, + "grad_norm": 0.40729503361891695, + "learning_rate": 0.0006420966835275933, + "loss": 2.4231, + "step": 1858 + }, + { + "epoch": 0.42627837651914696, + "grad_norm": 0.46047413136978865, + "learning_rate": 0.0006417406094634089, + "loss": 2.4324, + "step": 1859 + }, + { + "epoch": 0.42650768172437514, + "grad_norm": 0.41057377456416494, + "learning_rate": 0.0006413844572160527, + "loss": 2.4619, + "step": 1860 + }, + { + "epoch": 0.4267369869296033, + "grad_norm": 0.433410034503615, + "learning_rate": 0.0006410282269819756, + "loss": 2.3197, + "step": 1861 + }, + { + "epoch": 0.42696629213483145, + "grad_norm": 0.4221008904444366, + "learning_rate": 0.000640671918957672, + "loss": 2.4424, + "step": 1862 + }, + { + "epoch": 0.42719559734005963, + "grad_norm": 0.4502190502270232, + "learning_rate": 0.0006403155333396787, + "loss": 2.5168, + "step": 1863 + }, + { + "epoch": 0.42742490254528775, + "grad_norm": 0.4384639862821971, + "learning_rate": 0.0006399590703245759, + "loss": 2.5689, + "step": 1864 + }, + { + "epoch": 0.42765420775051594, + "grad_norm": 0.4395706065806961, + "learning_rate": 0.0006396025301089863, + "loss": 2.5418, + "step": 1865 + }, + { + "epoch": 0.4278835129557441, + "grad_norm": 0.43017065628789686, + "learning_rate": 0.0006392459128895747, + "loss": 2.4857, + "step": 1866 + }, + { + "epoch": 0.42811281816097224, + "grad_norm": 0.4343694032021137, + "learning_rate": 0.0006388892188630493, + "loss": 2.5064, + "step": 1867 + }, + { + "epoch": 0.4283421233662004, + "grad_norm": 0.4211615255452742, + "learning_rate": 0.0006385324482261597, + "loss": 2.464, + "step": 1868 + }, + { + "epoch": 0.42857142857142855, + "grad_norm": 0.4100774992933267, + "learning_rate": 0.0006381756011756982, + "loss": 2.7968, + "step": 1869 + }, + { + "epoch": 0.42880073377665673, + "grad_norm": 0.3960285988364611, + "learning_rate": 0.0006378186779084996, + "loss": 2.5358, + "step": 1870 + }, + { + "epoch": 0.4290300389818849, + "grad_norm": 0.41016099248493104, + "learning_rate": 0.0006374616786214403, + "loss": 2.4992, + "step": 1871 + }, + { + "epoch": 0.42925934418711303, + "grad_norm": 0.4051891438714947, + "learning_rate": 0.0006371046035114384, + "loss": 2.6222, + "step": 1872 + }, + { + "epoch": 0.4294886493923412, + "grad_norm": 0.40271625786156223, + "learning_rate": 0.0006367474527754544, + "loss": 2.4197, + "step": 1873 + }, + { + "epoch": 0.42971795459756934, + "grad_norm": 0.47812920892667554, + "learning_rate": 0.0006363902266104902, + "loss": 2.4795, + "step": 1874 + }, + { + "epoch": 0.4299472598027975, + "grad_norm": 0.4159954275256249, + "learning_rate": 0.0006360329252135894, + "loss": 2.3122, + "step": 1875 + }, + { + "epoch": 0.4301765650080257, + "grad_norm": 0.41967995282238046, + "learning_rate": 0.0006356755487818371, + "loss": 2.4435, + "step": 1876 + }, + { + "epoch": 0.4304058702132538, + "grad_norm": 0.4349284678090585, + "learning_rate": 0.0006353180975123595, + "loss": 2.4825, + "step": 1877 + }, + { + "epoch": 0.430635175418482, + "grad_norm": 0.4437813232187703, + "learning_rate": 0.0006349605716023246, + "loss": 2.4181, + "step": 1878 + }, + { + "epoch": 0.43086448062371013, + "grad_norm": 0.43195971717495546, + "learning_rate": 0.0006346029712489413, + "loss": 2.6244, + "step": 1879 + }, + { + "epoch": 0.4310937858289383, + "grad_norm": 0.4235250140979753, + "learning_rate": 0.0006342452966494593, + "loss": 2.4665, + "step": 1880 + }, + { + "epoch": 0.4313230910341665, + "grad_norm": 0.4262817015178558, + "learning_rate": 0.0006338875480011698, + "loss": 2.5324, + "step": 1881 + }, + { + "epoch": 0.4315523962393946, + "grad_norm": 0.3991689333813372, + "learning_rate": 0.0006335297255014043, + "loss": 2.4211, + "step": 1882 + }, + { + "epoch": 0.4317817014446228, + "grad_norm": 0.3989927056514101, + "learning_rate": 0.0006331718293475357, + "loss": 2.5805, + "step": 1883 + }, + { + "epoch": 0.4320110066498509, + "grad_norm": 0.43296015219919043, + "learning_rate": 0.0006328138597369768, + "loss": 2.4852, + "step": 1884 + }, + { + "epoch": 0.4322403118550791, + "grad_norm": 0.40493985054916504, + "learning_rate": 0.0006324558168671811, + "loss": 2.371, + "step": 1885 + }, + { + "epoch": 0.4324696170603073, + "grad_norm": 0.4230108184036763, + "learning_rate": 0.0006320977009356432, + "loss": 2.518, + "step": 1886 + }, + { + "epoch": 0.4326989222655354, + "grad_norm": 0.4441314772571616, + "learning_rate": 0.0006317395121398968, + "loss": 2.4759, + "step": 1887 + }, + { + "epoch": 0.4329282274707636, + "grad_norm": 0.4274954010199445, + "learning_rate": 0.0006313812506775169, + "loss": 2.6627, + "step": 1888 + }, + { + "epoch": 0.4331575326759917, + "grad_norm": 0.41050846912464684, + "learning_rate": 0.0006310229167461179, + "loss": 2.5058, + "step": 1889 + }, + { + "epoch": 0.4333868378812199, + "grad_norm": 0.49491647090813506, + "learning_rate": 0.0006306645105433546, + "loss": 2.4976, + "step": 1890 + }, + { + "epoch": 0.4336161430864481, + "grad_norm": 0.43683801565004, + "learning_rate": 0.0006303060322669214, + "loss": 2.4716, + "step": 1891 + }, + { + "epoch": 0.4338454482916762, + "grad_norm": 0.37430370404736424, + "learning_rate": 0.0006299474821145523, + "loss": 2.4301, + "step": 1892 + }, + { + "epoch": 0.4340747534969044, + "grad_norm": 0.43039669207187226, + "learning_rate": 0.0006295888602840214, + "loss": 2.4401, + "step": 1893 + }, + { + "epoch": 0.4343040587021325, + "grad_norm": 0.43597862550523403, + "learning_rate": 0.0006292301669731421, + "loss": 2.4809, + "step": 1894 + }, + { + "epoch": 0.4345333639073607, + "grad_norm": 0.42180944102611806, + "learning_rate": 0.0006288714023797671, + "loss": 2.4898, + "step": 1895 + }, + { + "epoch": 0.4347626691125889, + "grad_norm": 0.41351345028788994, + "learning_rate": 0.0006285125667017886, + "loss": 2.3136, + "step": 1896 + }, + { + "epoch": 0.434991974317817, + "grad_norm": 0.4656742291867835, + "learning_rate": 0.000628153660137138, + "loss": 2.5174, + "step": 1897 + }, + { + "epoch": 0.4352212795230452, + "grad_norm": 0.46575433329003957, + "learning_rate": 0.000627794682883786, + "loss": 2.4492, + "step": 1898 + }, + { + "epoch": 0.4354505847282733, + "grad_norm": 0.5295417251032559, + "learning_rate": 0.0006274356351397413, + "loss": 2.5893, + "step": 1899 + }, + { + "epoch": 0.4356798899335015, + "grad_norm": 0.45154392205536753, + "learning_rate": 0.000627076517103053, + "loss": 2.4951, + "step": 1900 + }, + { + "epoch": 0.43590919513872967, + "grad_norm": 0.4022159011529797, + "learning_rate": 0.0006267173289718079, + "loss": 2.401, + "step": 1901 + }, + { + "epoch": 0.4361385003439578, + "grad_norm": 0.4305686440962281, + "learning_rate": 0.0006263580709441319, + "loss": 2.6092, + "step": 1902 + }, + { + "epoch": 0.43636780554918597, + "grad_norm": 0.402283555584964, + "learning_rate": 0.000625998743218189, + "loss": 2.5446, + "step": 1903 + }, + { + "epoch": 0.43659711075441415, + "grad_norm": 0.39590084017401544, + "learning_rate": 0.0006256393459921824, + "loss": 2.6249, + "step": 1904 + }, + { + "epoch": 0.4368264159596423, + "grad_norm": 0.4112255858038538, + "learning_rate": 0.000625279879464353, + "loss": 2.4844, + "step": 1905 + }, + { + "epoch": 0.43705572116487046, + "grad_norm": 0.37488230810248235, + "learning_rate": 0.0006249203438329799, + "loss": 2.5432, + "step": 1906 + }, + { + "epoch": 0.4372850263700986, + "grad_norm": 0.43627984500941347, + "learning_rate": 0.000624560739296381, + "loss": 2.4216, + "step": 1907 + }, + { + "epoch": 0.43751433157532676, + "grad_norm": 0.42568023290011553, + "learning_rate": 0.0006242010660529115, + "loss": 2.5909, + "step": 1908 + }, + { + "epoch": 0.43774363678055495, + "grad_norm": 0.4086783659853399, + "learning_rate": 0.0006238413243009648, + "loss": 2.4481, + "step": 1909 + }, + { + "epoch": 0.43797294198578307, + "grad_norm": 0.4236790985869651, + "learning_rate": 0.0006234815142389719, + "loss": 2.4179, + "step": 1910 + }, + { + "epoch": 0.43820224719101125, + "grad_norm": 0.40659184350553934, + "learning_rate": 0.000623121636065402, + "loss": 2.55, + "step": 1911 + }, + { + "epoch": 0.4384315523962394, + "grad_norm": 0.39674378676715655, + "learning_rate": 0.0006227616899787612, + "loss": 2.5664, + "step": 1912 + }, + { + "epoch": 0.43866085760146756, + "grad_norm": 0.4252331686737038, + "learning_rate": 0.0006224016761775933, + "loss": 2.4145, + "step": 1913 + }, + { + "epoch": 0.43889016280669574, + "grad_norm": 0.40673513045227405, + "learning_rate": 0.0006220415948604796, + "loss": 2.5976, + "step": 1914 + }, + { + "epoch": 0.43911946801192386, + "grad_norm": 0.4123676313609636, + "learning_rate": 0.0006216814462260386, + "loss": 2.4399, + "step": 1915 + }, + { + "epoch": 0.43934877321715204, + "grad_norm": 0.40601601553564093, + "learning_rate": 0.0006213212304729259, + "loss": 2.6446, + "step": 1916 + }, + { + "epoch": 0.43957807842238017, + "grad_norm": 0.4609384316085945, + "learning_rate": 0.0006209609477998338, + "loss": 2.5988, + "step": 1917 + }, + { + "epoch": 0.43980738362760835, + "grad_norm": 0.4056172397774213, + "learning_rate": 0.0006206005984054923, + "loss": 2.4166, + "step": 1918 + }, + { + "epoch": 0.44003668883283653, + "grad_norm": 0.4399183444579953, + "learning_rate": 0.0006202401824886674, + "loss": 2.5318, + "step": 1919 + }, + { + "epoch": 0.44026599403806466, + "grad_norm": 0.44464884343384226, + "learning_rate": 0.0006198797002481622, + "loss": 2.6567, + "step": 1920 + }, + { + "epoch": 0.44049529924329284, + "grad_norm": 0.43867737061144807, + "learning_rate": 0.0006195191518828162, + "loss": 2.5576, + "step": 1921 + }, + { + "epoch": 0.44072460444852096, + "grad_norm": 0.4357959662190321, + "learning_rate": 0.0006191585375915056, + "loss": 2.4423, + "step": 1922 + }, + { + "epoch": 0.44095390965374914, + "grad_norm": 0.4463751804890018, + "learning_rate": 0.0006187978575731427, + "loss": 2.5188, + "step": 1923 + }, + { + "epoch": 0.4411832148589773, + "grad_norm": 0.42954629693342894, + "learning_rate": 0.0006184371120266761, + "loss": 2.6098, + "step": 1924 + }, + { + "epoch": 0.44141252006420545, + "grad_norm": 0.41763450963748705, + "learning_rate": 0.0006180763011510911, + "loss": 2.5698, + "step": 1925 + }, + { + "epoch": 0.44164182526943363, + "grad_norm": 0.43762791592003814, + "learning_rate": 0.0006177154251454082, + "loss": 2.5955, + "step": 1926 + }, + { + "epoch": 0.44187113047466176, + "grad_norm": 0.38994081502629996, + "learning_rate": 0.000617354484208684, + "loss": 2.2868, + "step": 1927 + }, + { + "epoch": 0.44210043567988994, + "grad_norm": 0.4269930197828653, + "learning_rate": 0.0006169934785400116, + "loss": 2.3899, + "step": 1928 + }, + { + "epoch": 0.4423297408851181, + "grad_norm": 0.45319865853616126, + "learning_rate": 0.0006166324083385189, + "loss": 2.5476, + "step": 1929 + }, + { + "epoch": 0.44255904609034624, + "grad_norm": 0.43118906238298316, + "learning_rate": 0.0006162712738033699, + "loss": 2.6547, + "step": 1930 + }, + { + "epoch": 0.4427883512955744, + "grad_norm": 0.46371037655038955, + "learning_rate": 0.0006159100751337642, + "loss": 2.5492, + "step": 1931 + }, + { + "epoch": 0.44301765650080255, + "grad_norm": 0.405193936764988, + "learning_rate": 0.0006155488125289363, + "loss": 2.4209, + "step": 1932 + }, + { + "epoch": 0.44324696170603073, + "grad_norm": 0.40850470460477634, + "learning_rate": 0.0006151874861881565, + "loss": 2.4745, + "step": 1933 + }, + { + "epoch": 0.4434762669112589, + "grad_norm": 0.4317117130979621, + "learning_rate": 0.0006148260963107297, + "loss": 2.6118, + "step": 1934 + }, + { + "epoch": 0.44370557211648703, + "grad_norm": 0.40276576879946396, + "learning_rate": 0.0006144646430959964, + "loss": 2.6579, + "step": 1935 + }, + { + "epoch": 0.4439348773217152, + "grad_norm": 0.3912647463597528, + "learning_rate": 0.0006141031267433316, + "loss": 2.5016, + "step": 1936 + }, + { + "epoch": 0.44416418252694334, + "grad_norm": 0.3973641567700097, + "learning_rate": 0.0006137415474521454, + "loss": 2.4996, + "step": 1937 + }, + { + "epoch": 0.4443934877321715, + "grad_norm": 0.44973500180782694, + "learning_rate": 0.0006133799054218825, + "loss": 2.5848, + "step": 1938 + }, + { + "epoch": 0.4446227929373997, + "grad_norm": 0.4079369389821792, + "learning_rate": 0.0006130182008520222, + "loss": 2.5143, + "step": 1939 + }, + { + "epoch": 0.4448520981426278, + "grad_norm": 0.41185376036854404, + "learning_rate": 0.0006126564339420784, + "loss": 2.6882, + "step": 1940 + }, + { + "epoch": 0.445081403347856, + "grad_norm": 0.3895838706852536, + "learning_rate": 0.000612294604891599, + "loss": 2.4585, + "step": 1941 + }, + { + "epoch": 0.44531070855308413, + "grad_norm": 0.4020348555486623, + "learning_rate": 0.000611932713900167, + "loss": 2.4621, + "step": 1942 + }, + { + "epoch": 0.4455400137583123, + "grad_norm": 0.3966018871524096, + "learning_rate": 0.0006115707611673986, + "loss": 2.4182, + "step": 1943 + }, + { + "epoch": 0.4457693189635405, + "grad_norm": 0.43986734080265244, + "learning_rate": 0.0006112087468929449, + "loss": 2.5167, + "step": 1944 + }, + { + "epoch": 0.4459986241687686, + "grad_norm": 0.42678204308840306, + "learning_rate": 0.0006108466712764902, + "loss": 2.5016, + "step": 1945 + }, + { + "epoch": 0.4462279293739968, + "grad_norm": 0.4374016520801921, + "learning_rate": 0.000610484534517753, + "loss": 2.6483, + "step": 1946 + }, + { + "epoch": 0.4464572345792249, + "grad_norm": 0.43430201703681504, + "learning_rate": 0.0006101223368164858, + "loss": 2.5739, + "step": 1947 + }, + { + "epoch": 0.4466865397844531, + "grad_norm": 0.41339758288057776, + "learning_rate": 0.0006097600783724742, + "loss": 2.5301, + "step": 1948 + }, + { + "epoch": 0.4469158449896813, + "grad_norm": 0.4359910055722659, + "learning_rate": 0.0006093977593855375, + "loss": 2.5213, + "step": 1949 + }, + { + "epoch": 0.4471451501949094, + "grad_norm": 0.40971139342479734, + "learning_rate": 0.0006090353800555286, + "loss": 2.5199, + "step": 1950 + }, + { + "epoch": 0.4473744554001376, + "grad_norm": 0.4178390586831926, + "learning_rate": 0.0006086729405823335, + "loss": 2.446, + "step": 1951 + }, + { + "epoch": 0.4476037606053657, + "grad_norm": 0.434793408575898, + "learning_rate": 0.0006083104411658713, + "loss": 2.4579, + "step": 1952 + }, + { + "epoch": 0.4478330658105939, + "grad_norm": 0.41665467101055964, + "learning_rate": 0.0006079478820060943, + "loss": 2.5386, + "step": 1953 + }, + { + "epoch": 0.4480623710158221, + "grad_norm": 0.4178433559707151, + "learning_rate": 0.0006075852633029877, + "loss": 2.4518, + "step": 1954 + }, + { + "epoch": 0.4482916762210502, + "grad_norm": 0.3969986298101345, + "learning_rate": 0.0006072225852565695, + "loss": 2.4471, + "step": 1955 + }, + { + "epoch": 0.4485209814262784, + "grad_norm": 0.692853808781834, + "learning_rate": 0.0006068598480668905, + "loss": 2.502, + "step": 1956 + }, + { + "epoch": 0.4487502866315065, + "grad_norm": 0.42887071793351145, + "learning_rate": 0.0006064970519340341, + "loss": 2.4411, + "step": 1957 + }, + { + "epoch": 0.4489795918367347, + "grad_norm": 0.41763962468050075, + "learning_rate": 0.0006061341970581165, + "loss": 2.4762, + "step": 1958 + }, + { + "epoch": 0.4492088970419629, + "grad_norm": 0.42240037450986817, + "learning_rate": 0.0006057712836392856, + "loss": 2.4041, + "step": 1959 + }, + { + "epoch": 0.449438202247191, + "grad_norm": 0.4090412962001092, + "learning_rate": 0.0006054083118777221, + "loss": 2.4633, + "step": 1960 + }, + { + "epoch": 0.4496675074524192, + "grad_norm": 0.41730419194992247, + "learning_rate": 0.0006050452819736389, + "loss": 2.5467, + "step": 1961 + }, + { + "epoch": 0.4498968126576473, + "grad_norm": 0.4236491537748718, + "learning_rate": 0.000604682194127281, + "loss": 2.4896, + "step": 1962 + }, + { + "epoch": 0.4501261178628755, + "grad_norm": 0.4228898206088045, + "learning_rate": 0.000604319048538925, + "loss": 2.558, + "step": 1963 + }, + { + "epoch": 0.45035542306810367, + "grad_norm": 0.41968741376038227, + "learning_rate": 0.0006039558454088796, + "loss": 2.4719, + "step": 1964 + }, + { + "epoch": 0.4505847282733318, + "grad_norm": 0.42642898233125004, + "learning_rate": 0.0006035925849374855, + "loss": 2.6825, + "step": 1965 + }, + { + "epoch": 0.45081403347856, + "grad_norm": 0.41139797305604925, + "learning_rate": 0.0006032292673251142, + "loss": 2.4531, + "step": 1966 + }, + { + "epoch": 0.4510433386837881, + "grad_norm": 0.40843532151635853, + "learning_rate": 0.0006028658927721697, + "loss": 2.5359, + "step": 1967 + }, + { + "epoch": 0.4512726438890163, + "grad_norm": 0.4251099296848785, + "learning_rate": 0.000602502461479087, + "loss": 2.5559, + "step": 1968 + }, + { + "epoch": 0.45150194909424446, + "grad_norm": 0.481368532376938, + "learning_rate": 0.0006021389736463321, + "loss": 2.3814, + "step": 1969 + }, + { + "epoch": 0.4517312542994726, + "grad_norm": 0.40930654316705456, + "learning_rate": 0.0006017754294744027, + "loss": 2.4401, + "step": 1970 + }, + { + "epoch": 0.45196055950470077, + "grad_norm": 0.545008361656244, + "learning_rate": 0.0006014118291638271, + "loss": 2.2956, + "step": 1971 + }, + { + "epoch": 0.4521898647099289, + "grad_norm": 0.42942642152267746, + "learning_rate": 0.0006010481729151649, + "loss": 2.548, + "step": 1972 + }, + { + "epoch": 0.45241916991515707, + "grad_norm": 0.40348909087892604, + "learning_rate": 0.0006006844609290065, + "loss": 2.4999, + "step": 1973 + }, + { + "epoch": 0.45264847512038525, + "grad_norm": 0.6843519838534303, + "learning_rate": 0.000600320693405973, + "loss": 2.5882, + "step": 1974 + }, + { + "epoch": 0.4528777803256134, + "grad_norm": 0.4620650111290832, + "learning_rate": 0.0005999568705467161, + "loss": 2.5189, + "step": 1975 + }, + { + "epoch": 0.45310708553084156, + "grad_norm": 0.4123795902341295, + "learning_rate": 0.000599592992551918, + "loss": 2.4686, + "step": 1976 + }, + { + "epoch": 0.4533363907360697, + "grad_norm": 0.43570273353902916, + "learning_rate": 0.0005992290596222915, + "loss": 2.5397, + "step": 1977 + }, + { + "epoch": 0.45356569594129786, + "grad_norm": 0.4708190608981544, + "learning_rate": 0.0005988650719585795, + "loss": 2.6789, + "step": 1978 + }, + { + "epoch": 0.45379500114652604, + "grad_norm": 0.40778261667209703, + "learning_rate": 0.0005985010297615551, + "loss": 2.4812, + "step": 1979 + }, + { + "epoch": 0.45402430635175417, + "grad_norm": 0.4130981680705708, + "learning_rate": 0.0005981369332320217, + "loss": 2.6788, + "step": 1980 + }, + { + "epoch": 0.45425361155698235, + "grad_norm": 0.3948063798067752, + "learning_rate": 0.0005977727825708123, + "loss": 2.4417, + "step": 1981 + }, + { + "epoch": 0.4544829167622105, + "grad_norm": 0.4246683114573599, + "learning_rate": 0.0005974085779787901, + "loss": 2.5658, + "step": 1982 + }, + { + "epoch": 0.45471222196743866, + "grad_norm": 0.3750760900767624, + "learning_rate": 0.0005970443196568478, + "loss": 2.4905, + "step": 1983 + }, + { + "epoch": 0.45494152717266684, + "grad_norm": 0.4111385887737466, + "learning_rate": 0.000596680007805908, + "loss": 2.5272, + "step": 1984 + }, + { + "epoch": 0.45517083237789496, + "grad_norm": 0.38863281390457494, + "learning_rate": 0.0005963156426269227, + "loss": 2.4569, + "step": 1985 + }, + { + "epoch": 0.45540013758312314, + "grad_norm": 0.4749701254671495, + "learning_rate": 0.0005959512243208731, + "loss": 2.5089, + "step": 1986 + }, + { + "epoch": 0.45562944278835127, + "grad_norm": 0.38755001844653425, + "learning_rate": 0.0005955867530887702, + "loss": 2.4411, + "step": 1987 + }, + { + "epoch": 0.45585874799357945, + "grad_norm": 0.4730725271087748, + "learning_rate": 0.0005952222291316535, + "loss": 2.5407, + "step": 1988 + }, + { + "epoch": 0.45608805319880763, + "grad_norm": 0.3883613449181855, + "learning_rate": 0.0005948576526505923, + "loss": 2.4357, + "step": 1989 + }, + { + "epoch": 0.45631735840403576, + "grad_norm": 0.4163887638061282, + "learning_rate": 0.0005944930238466845, + "loss": 2.5114, + "step": 1990 + }, + { + "epoch": 0.45654666360926394, + "grad_norm": 0.4444078621488862, + "learning_rate": 0.0005941283429210568, + "loss": 2.5607, + "step": 1991 + }, + { + "epoch": 0.45677596881449206, + "grad_norm": 0.4049211576592178, + "learning_rate": 0.0005937636100748648, + "loss": 2.3071, + "step": 1992 + }, + { + "epoch": 0.45700527401972024, + "grad_norm": 0.411759164417734, + "learning_rate": 0.0005933988255092926, + "loss": 2.4583, + "step": 1993 + }, + { + "epoch": 0.4572345792249484, + "grad_norm": 0.4312319324360527, + "learning_rate": 0.0005930339894255532, + "loss": 2.4823, + "step": 1994 + }, + { + "epoch": 0.45746388443017655, + "grad_norm": 0.43882274429451745, + "learning_rate": 0.0005926691020248874, + "loss": 2.4968, + "step": 1995 + }, + { + "epoch": 0.45769318963540473, + "grad_norm": 0.49182388107902025, + "learning_rate": 0.000592304163508565, + "loss": 2.2532, + "step": 1996 + }, + { + "epoch": 0.4579224948406329, + "grad_norm": 0.43191706816565084, + "learning_rate": 0.0005919391740778833, + "loss": 2.49, + "step": 1997 + }, + { + "epoch": 0.45815180004586104, + "grad_norm": 0.5162431791786872, + "learning_rate": 0.0005915741339341682, + "loss": 2.5633, + "step": 1998 + }, + { + "epoch": 0.4583811052510892, + "grad_norm": 0.4438186793798423, + "learning_rate": 0.0005912090432787736, + "loss": 2.6474, + "step": 1999 + }, + { + "epoch": 0.45861041045631734, + "grad_norm": 0.40612204551779696, + "learning_rate": 0.0005908439023130807, + "loss": 2.2976, + "step": 2000 + }, + { + "epoch": 0.4588397156615455, + "grad_norm": 0.5154024081925999, + "learning_rate": 0.000590478711238499, + "loss": 2.4562, + "step": 2001 + }, + { + "epoch": 0.4590690208667737, + "grad_norm": 0.47948434264467954, + "learning_rate": 0.0005901134702564656, + "loss": 2.4922, + "step": 2002 + }, + { + "epoch": 0.45929832607200183, + "grad_norm": 0.444903977938652, + "learning_rate": 0.0005897481795684446, + "loss": 2.5206, + "step": 2003 + }, + { + "epoch": 0.45952763127723, + "grad_norm": 0.4070529343809612, + "learning_rate": 0.0005893828393759281, + "loss": 2.4752, + "step": 2004 + }, + { + "epoch": 0.45975693648245813, + "grad_norm": 0.45494451543217523, + "learning_rate": 0.0005890174498804355, + "loss": 2.6417, + "step": 2005 + }, + { + "epoch": 0.4599862416876863, + "grad_norm": 0.4584443980106787, + "learning_rate": 0.0005886520112835128, + "loss": 2.4625, + "step": 2006 + }, + { + "epoch": 0.4602155468929145, + "grad_norm": 0.4275333481671138, + "learning_rate": 0.0005882865237867339, + "loss": 2.545, + "step": 2007 + }, + { + "epoch": 0.4604448520981426, + "grad_norm": 0.44945428199728865, + "learning_rate": 0.0005879209875916989, + "loss": 2.3915, + "step": 2008 + }, + { + "epoch": 0.4606741573033708, + "grad_norm": 0.42800562996723573, + "learning_rate": 0.0005875554029000353, + "loss": 2.5201, + "step": 2009 + }, + { + "epoch": 0.4609034625085989, + "grad_norm": 0.450374363436087, + "learning_rate": 0.0005871897699133973, + "loss": 2.503, + "step": 2010 + }, + { + "epoch": 0.4611327677138271, + "grad_norm": 0.42712103348260033, + "learning_rate": 0.0005868240888334653, + "loss": 2.6142, + "step": 2011 + }, + { + "epoch": 0.4613620729190553, + "grad_norm": 0.4096235105743487, + "learning_rate": 0.0005864583598619468, + "loss": 2.4463, + "step": 2012 + }, + { + "epoch": 0.4615913781242834, + "grad_norm": 0.40730540049499514, + "learning_rate": 0.0005860925832005753, + "loss": 2.4814, + "step": 2013 + }, + { + "epoch": 0.4618206833295116, + "grad_norm": 0.49986345079125905, + "learning_rate": 0.0005857267590511108, + "loss": 2.4509, + "step": 2014 + }, + { + "epoch": 0.4620499885347397, + "grad_norm": 0.42290195052983176, + "learning_rate": 0.0005853608876153395, + "loss": 2.5344, + "step": 2015 + }, + { + "epoch": 0.4622792937399679, + "grad_norm": 0.4134584133249253, + "learning_rate": 0.0005849949690950735, + "loss": 2.3932, + "step": 2016 + }, + { + "epoch": 0.4625085989451961, + "grad_norm": 0.40959055196593114, + "learning_rate": 0.0005846290036921512, + "loss": 2.4953, + "step": 2017 + }, + { + "epoch": 0.4627379041504242, + "grad_norm": 0.42892830895847744, + "learning_rate": 0.0005842629916084365, + "loss": 2.4295, + "step": 2018 + }, + { + "epoch": 0.4629672093556524, + "grad_norm": 0.4329839084591894, + "learning_rate": 0.0005838969330458195, + "loss": 2.5656, + "step": 2019 + }, + { + "epoch": 0.4631965145608805, + "grad_norm": 0.421728936961171, + "learning_rate": 0.0005835308282062153, + "loss": 2.5655, + "step": 2020 + }, + { + "epoch": 0.4634258197661087, + "grad_norm": 0.4243343803781863, + "learning_rate": 0.0005831646772915651, + "loss": 2.4682, + "step": 2021 + }, + { + "epoch": 0.4636551249713369, + "grad_norm": 0.4077765557656573, + "learning_rate": 0.0005827984805038352, + "loss": 2.5637, + "step": 2022 + }, + { + "epoch": 0.463884430176565, + "grad_norm": 0.4230905964994004, + "learning_rate": 0.0005824322380450173, + "loss": 2.5877, + "step": 2023 + }, + { + "epoch": 0.4641137353817932, + "grad_norm": 0.4121770885184156, + "learning_rate": 0.0005820659501171284, + "loss": 2.5331, + "step": 2024 + }, + { + "epoch": 0.4643430405870213, + "grad_norm": 0.42910687168836686, + "learning_rate": 0.0005816996169222102, + "loss": 2.3962, + "step": 2025 + }, + { + "epoch": 0.4645723457922495, + "grad_norm": 0.3879030205165921, + "learning_rate": 0.00058133323866233, + "loss": 2.6819, + "step": 2026 + }, + { + "epoch": 0.46480165099747767, + "grad_norm": 0.43385501885180194, + "learning_rate": 0.0005809668155395793, + "loss": 2.4038, + "step": 2027 + }, + { + "epoch": 0.4650309562027058, + "grad_norm": 0.419876196144891, + "learning_rate": 0.0005806003477560748, + "loss": 2.5586, + "step": 2028 + }, + { + "epoch": 0.465260261407934, + "grad_norm": 0.4113945938145781, + "learning_rate": 0.0005802338355139578, + "loss": 2.4495, + "step": 2029 + }, + { + "epoch": 0.4654895666131621, + "grad_norm": 0.49256512764143734, + "learning_rate": 0.0005798672790153937, + "loss": 2.5572, + "step": 2030 + }, + { + "epoch": 0.4657188718183903, + "grad_norm": 0.4006945850198939, + "learning_rate": 0.0005795006784625728, + "loss": 2.5237, + "step": 2031 + }, + { + "epoch": 0.46594817702361846, + "grad_norm": 0.4009880771135874, + "learning_rate": 0.0005791340340577094, + "loss": 2.4889, + "step": 2032 + }, + { + "epoch": 0.4661774822288466, + "grad_norm": 0.4155818624688708, + "learning_rate": 0.0005787673460030423, + "loss": 2.5356, + "step": 2033 + }, + { + "epoch": 0.46640678743407477, + "grad_norm": 0.5064532987263431, + "learning_rate": 0.000578400614500834, + "loss": 2.4543, + "step": 2034 + }, + { + "epoch": 0.4666360926393029, + "grad_norm": 0.43322290160536175, + "learning_rate": 0.000578033839753371, + "loss": 2.3387, + "step": 2035 + }, + { + "epoch": 0.46686539784453107, + "grad_norm": 0.419644667642737, + "learning_rate": 0.0005776670219629642, + "loss": 2.6276, + "step": 2036 + }, + { + "epoch": 0.46709470304975925, + "grad_norm": 0.45071106953185436, + "learning_rate": 0.0005773001613319476, + "loss": 2.568, + "step": 2037 + }, + { + "epoch": 0.4673240082549874, + "grad_norm": 0.4426388152222326, + "learning_rate": 0.0005769332580626792, + "loss": 2.5049, + "step": 2038 + }, + { + "epoch": 0.46755331346021556, + "grad_norm": 0.4422388871586187, + "learning_rate": 0.00057656631235754, + "loss": 2.5279, + "step": 2039 + }, + { + "epoch": 0.4677826186654437, + "grad_norm": 0.5013417997051348, + "learning_rate": 0.0005761993244189354, + "loss": 2.4497, + "step": 2040 + }, + { + "epoch": 0.46801192387067186, + "grad_norm": 0.4248235421982216, + "learning_rate": 0.0005758322944492929, + "loss": 2.4747, + "step": 2041 + }, + { + "epoch": 0.46824122907590005, + "grad_norm": 0.41141993847843333, + "learning_rate": 0.0005754652226510644, + "loss": 2.618, + "step": 2042 + }, + { + "epoch": 0.46847053428112817, + "grad_norm": 0.44513810261256054, + "learning_rate": 0.0005750981092267237, + "loss": 2.5351, + "step": 2043 + }, + { + "epoch": 0.46869983948635635, + "grad_norm": 0.40401286826103033, + "learning_rate": 0.0005747309543787683, + "loss": 2.5569, + "step": 2044 + }, + { + "epoch": 0.4689291446915845, + "grad_norm": 0.3967611424672895, + "learning_rate": 0.0005743637583097183, + "loss": 2.4905, + "step": 2045 + }, + { + "epoch": 0.46915844989681266, + "grad_norm": 0.41870906694409116, + "learning_rate": 0.0005739965212221168, + "loss": 2.4972, + "step": 2046 + }, + { + "epoch": 0.46938775510204084, + "grad_norm": 0.44739464283465624, + "learning_rate": 0.0005736292433185291, + "loss": 2.4583, + "step": 2047 + }, + { + "epoch": 0.46961706030726896, + "grad_norm": 0.39611578124715496, + "learning_rate": 0.0005732619248015433, + "loss": 2.421, + "step": 2048 + }, + { + "epoch": 0.46984636551249714, + "grad_norm": 0.4002060559573953, + "learning_rate": 0.0005728945658737699, + "loss": 2.5447, + "step": 2049 + }, + { + "epoch": 0.47007567071772527, + "grad_norm": 0.3934428246151609, + "learning_rate": 0.0005725271667378418, + "loss": 2.4703, + "step": 2050 + }, + { + "epoch": 0.47030497592295345, + "grad_norm": 0.4342415657408605, + "learning_rate": 0.0005721597275964133, + "loss": 2.4925, + "step": 2051 + }, + { + "epoch": 0.47053428112818163, + "grad_norm": 0.42081603377456767, + "learning_rate": 0.000571792248652162, + "loss": 2.5625, + "step": 2052 + }, + { + "epoch": 0.47076358633340976, + "grad_norm": 0.42540348816104706, + "learning_rate": 0.0005714247301077865, + "loss": 2.4657, + "step": 2053 + }, + { + "epoch": 0.47099289153863794, + "grad_norm": 0.40856371745875253, + "learning_rate": 0.0005710571721660077, + "loss": 2.4891, + "step": 2054 + }, + { + "epoch": 0.47122219674386606, + "grad_norm": 0.43178647214201094, + "learning_rate": 0.0005706895750295682, + "loss": 2.5871, + "step": 2055 + }, + { + "epoch": 0.47145150194909424, + "grad_norm": 0.418567749500074, + "learning_rate": 0.0005703219389012318, + "loss": 2.471, + "step": 2056 + }, + { + "epoch": 0.4716808071543224, + "grad_norm": 0.5189769314406872, + "learning_rate": 0.0005699542639837844, + "loss": 2.37, + "step": 2057 + }, + { + "epoch": 0.47191011235955055, + "grad_norm": 0.417786871327731, + "learning_rate": 0.0005695865504800327, + "loss": 2.4453, + "step": 2058 + }, + { + "epoch": 0.47213941756477873, + "grad_norm": 0.42719985022691753, + "learning_rate": 0.0005692187985928055, + "loss": 2.514, + "step": 2059 + }, + { + "epoch": 0.47236872277000685, + "grad_norm": 0.4160457013747338, + "learning_rate": 0.000568851008524952, + "loss": 2.6173, + "step": 2060 + }, + { + "epoch": 0.47259802797523504, + "grad_norm": 0.49123550151141215, + "learning_rate": 0.0005684831804793427, + "loss": 2.4072, + "step": 2061 + }, + { + "epoch": 0.4728273331804632, + "grad_norm": 0.4101438163050544, + "learning_rate": 0.0005681153146588693, + "loss": 2.5214, + "step": 2062 + }, + { + "epoch": 0.47305663838569134, + "grad_norm": 0.38080956758000445, + "learning_rate": 0.0005677474112664438, + "loss": 2.4375, + "step": 2063 + }, + { + "epoch": 0.4732859435909195, + "grad_norm": 0.41824188603256335, + "learning_rate": 0.0005673794705049995, + "loss": 2.4942, + "step": 2064 + }, + { + "epoch": 0.47351524879614765, + "grad_norm": 0.4342755915211105, + "learning_rate": 0.0005670114925774899, + "loss": 2.401, + "step": 2065 + }, + { + "epoch": 0.47374455400137583, + "grad_norm": 0.4486001046944782, + "learning_rate": 0.0005666434776868895, + "loss": 2.4546, + "step": 2066 + }, + { + "epoch": 0.473973859206604, + "grad_norm": 0.5029506146980202, + "learning_rate": 0.0005662754260361924, + "loss": 2.6153, + "step": 2067 + }, + { + "epoch": 0.47420316441183213, + "grad_norm": 0.4417740539808205, + "learning_rate": 0.0005659073378284137, + "loss": 2.4224, + "step": 2068 + }, + { + "epoch": 0.4744324696170603, + "grad_norm": 0.4323886334880185, + "learning_rate": 0.0005655392132665884, + "loss": 2.5957, + "step": 2069 + }, + { + "epoch": 0.47466177482228844, + "grad_norm": 0.3869778787049649, + "learning_rate": 0.0005651710525537714, + "loss": 2.4837, + "step": 2070 + }, + { + "epoch": 0.4748910800275166, + "grad_norm": 0.42112262741076695, + "learning_rate": 0.000564802855893038, + "loss": 2.4087, + "step": 2071 + }, + { + "epoch": 0.4751203852327448, + "grad_norm": 0.44844834875830747, + "learning_rate": 0.0005644346234874824, + "loss": 2.4874, + "step": 2072 + }, + { + "epoch": 0.4753496904379729, + "grad_norm": 0.37787578431172864, + "learning_rate": 0.0005640663555402198, + "loss": 2.4875, + "step": 2073 + }, + { + "epoch": 0.4755789956432011, + "grad_norm": 0.4100645198565528, + "learning_rate": 0.0005636980522543838, + "loss": 2.4955, + "step": 2074 + }, + { + "epoch": 0.47580830084842923, + "grad_norm": 0.42160985102623844, + "learning_rate": 0.0005633297138331285, + "loss": 2.4555, + "step": 2075 + }, + { + "epoch": 0.4760376060536574, + "grad_norm": 0.6683046734423032, + "learning_rate": 0.0005629613404796266, + "loss": 2.4071, + "step": 2076 + }, + { + "epoch": 0.4762669112588856, + "grad_norm": 0.5148914434535207, + "learning_rate": 0.0005625929323970705, + "loss": 2.5123, + "step": 2077 + }, + { + "epoch": 0.4764962164641137, + "grad_norm": 0.4344349832648187, + "learning_rate": 0.0005622244897886716, + "loss": 2.5688, + "step": 2078 + }, + { + "epoch": 0.4767255216693419, + "grad_norm": 0.39102776776013787, + "learning_rate": 0.0005618560128576603, + "loss": 2.5623, + "step": 2079 + }, + { + "epoch": 0.47695482687457, + "grad_norm": 0.45585153517369814, + "learning_rate": 0.0005614875018072861, + "loss": 2.326, + "step": 2080 + }, + { + "epoch": 0.4771841320797982, + "grad_norm": 0.42879083913254457, + "learning_rate": 0.0005611189568408173, + "loss": 2.501, + "step": 2081 + }, + { + "epoch": 0.4774134372850264, + "grad_norm": 0.781987335091487, + "learning_rate": 0.0005607503781615407, + "loss": 2.5925, + "step": 2082 + }, + { + "epoch": 0.4776427424902545, + "grad_norm": 0.4295872116186448, + "learning_rate": 0.0005603817659727619, + "loss": 2.589, + "step": 2083 + }, + { + "epoch": 0.4778720476954827, + "grad_norm": 0.44766151733937176, + "learning_rate": 0.0005600131204778049, + "loss": 2.499, + "step": 2084 + }, + { + "epoch": 0.4781013529007108, + "grad_norm": 0.4410378749791594, + "learning_rate": 0.0005596444418800121, + "loss": 2.6293, + "step": 2085 + }, + { + "epoch": 0.478330658105939, + "grad_norm": 0.4346059788891013, + "learning_rate": 0.000559275730382744, + "loss": 2.5272, + "step": 2086 + }, + { + "epoch": 0.4785599633111672, + "grad_norm": 0.42419387239593925, + "learning_rate": 0.0005589069861893798, + "loss": 2.43, + "step": 2087 + }, + { + "epoch": 0.4787892685163953, + "grad_norm": 0.46704528816734536, + "learning_rate": 0.0005585382095033159, + "loss": 2.4496, + "step": 2088 + }, + { + "epoch": 0.4790185737216235, + "grad_norm": 0.4171691887788205, + "learning_rate": 0.0005581694005279673, + "loss": 2.5494, + "step": 2089 + }, + { + "epoch": 0.47924787892685167, + "grad_norm": 0.8563998944283588, + "learning_rate": 0.0005578005594667665, + "loss": 2.5639, + "step": 2090 + }, + { + "epoch": 0.4794771841320798, + "grad_norm": 0.5946852778116933, + "learning_rate": 0.0005574316865231637, + "loss": 2.338, + "step": 2091 + }, + { + "epoch": 0.479706489337308, + "grad_norm": 0.5103083772633821, + "learning_rate": 0.0005570627819006268, + "loss": 2.5149, + "step": 2092 + }, + { + "epoch": 0.4799357945425361, + "grad_norm": 0.43329498643797065, + "learning_rate": 0.0005566938458026411, + "loss": 2.4757, + "step": 2093 + }, + { + "epoch": 0.4801650997477643, + "grad_norm": 0.42407891240051376, + "learning_rate": 0.0005563248784327092, + "loss": 2.4251, + "step": 2094 + }, + { + "epoch": 0.48039440495299246, + "grad_norm": 0.47339543094566083, + "learning_rate": 0.0005559558799943514, + "loss": 2.5897, + "step": 2095 + }, + { + "epoch": 0.4806237101582206, + "grad_norm": 0.43627543120293705, + "learning_rate": 0.0005555868506911044, + "loss": 2.5276, + "step": 2096 + }, + { + "epoch": 0.48085301536344877, + "grad_norm": 0.45197767900402985, + "learning_rate": 0.0005552177907265223, + "loss": 2.5608, + "step": 2097 + }, + { + "epoch": 0.4810823205686769, + "grad_norm": 0.4371637745794847, + "learning_rate": 0.0005548487003041762, + "loss": 2.4712, + "step": 2098 + }, + { + "epoch": 0.48131162577390507, + "grad_norm": 0.6685445801570307, + "learning_rate": 0.000554479579627654, + "loss": 2.557, + "step": 2099 + }, + { + "epoch": 0.48154093097913325, + "grad_norm": 0.44929311857523613, + "learning_rate": 0.0005541104289005602, + "loss": 2.528, + "step": 2100 + }, + { + "epoch": 0.4817702361843614, + "grad_norm": 0.4481141074063933, + "learning_rate": 0.0005537412483265157, + "loss": 2.4323, + "step": 2101 + }, + { + "epoch": 0.48199954138958956, + "grad_norm": 0.4512739411820372, + "learning_rate": 0.0005533720381091582, + "loss": 2.4737, + "step": 2102 + }, + { + "epoch": 0.4822288465948177, + "grad_norm": 0.4375414230178747, + "learning_rate": 0.0005530027984521413, + "loss": 2.5197, + "step": 2103 + }, + { + "epoch": 0.48245815180004586, + "grad_norm": 0.4806836674053585, + "learning_rate": 0.0005526335295591354, + "loss": 2.4679, + "step": 2104 + }, + { + "epoch": 0.48268745700527405, + "grad_norm": 0.4154785794038622, + "learning_rate": 0.0005522642316338268, + "loss": 2.4048, + "step": 2105 + }, + { + "epoch": 0.48291676221050217, + "grad_norm": 0.4606747793746695, + "learning_rate": 0.0005518949048799176, + "loss": 2.4895, + "step": 2106 + }, + { + "epoch": 0.48314606741573035, + "grad_norm": 0.470828686552855, + "learning_rate": 0.0005515255495011259, + "loss": 2.5873, + "step": 2107 + }, + { + "epoch": 0.4833753726209585, + "grad_norm": 0.5384535455922705, + "learning_rate": 0.0005511561657011859, + "loss": 2.4955, + "step": 2108 + }, + { + "epoch": 0.48360467782618666, + "grad_norm": 0.6602255419150592, + "learning_rate": 0.0005507867536838472, + "loss": 2.5182, + "step": 2109 + }, + { + "epoch": 0.48383398303141484, + "grad_norm": 0.4976405037883742, + "learning_rate": 0.0005504173136528746, + "loss": 2.5746, + "step": 2110 + }, + { + "epoch": 0.48406328823664296, + "grad_norm": 0.4463071630388425, + "learning_rate": 0.0005500478458120492, + "loss": 2.509, + "step": 2111 + }, + { + "epoch": 0.48429259344187114, + "grad_norm": 0.45577970611811147, + "learning_rate": 0.0005496783503651667, + "loss": 2.4794, + "step": 2112 + }, + { + "epoch": 0.48452189864709927, + "grad_norm": 0.4024922689882915, + "learning_rate": 0.0005493088275160387, + "loss": 2.448, + "step": 2113 + }, + { + "epoch": 0.48475120385232745, + "grad_norm": 0.42915534114426196, + "learning_rate": 0.0005489392774684911, + "loss": 2.5245, + "step": 2114 + }, + { + "epoch": 0.48498050905755563, + "grad_norm": 0.4260531408793905, + "learning_rate": 0.0005485697004263657, + "loss": 2.5264, + "step": 2115 + }, + { + "epoch": 0.48520981426278376, + "grad_norm": 0.4993458354015788, + "learning_rate": 0.0005482000965935182, + "loss": 2.5138, + "step": 2116 + }, + { + "epoch": 0.48543911946801194, + "grad_norm": 0.40017227769905256, + "learning_rate": 0.0005478304661738199, + "loss": 2.5617, + "step": 2117 + }, + { + "epoch": 0.48566842467324006, + "grad_norm": 0.40211006500657265, + "learning_rate": 0.0005474608093711565, + "loss": 2.5316, + "step": 2118 + }, + { + "epoch": 0.48589772987846824, + "grad_norm": 0.6522156325889068, + "learning_rate": 0.0005470911263894279, + "loss": 2.4558, + "step": 2119 + }, + { + "epoch": 0.4861270350836964, + "grad_norm": 0.4154569181501013, + "learning_rate": 0.0005467214174325494, + "loss": 2.4282, + "step": 2120 + }, + { + "epoch": 0.48635634028892455, + "grad_norm": 0.4770746147903165, + "learning_rate": 0.0005463516827044491, + "loss": 2.4815, + "step": 2121 + }, + { + "epoch": 0.48658564549415273, + "grad_norm": 0.43029828248743607, + "learning_rate": 0.000545981922409071, + "loss": 2.5447, + "step": 2122 + }, + { + "epoch": 0.48681495069938086, + "grad_norm": 0.5059668647528426, + "learning_rate": 0.000545612136750372, + "loss": 2.5586, + "step": 2123 + }, + { + "epoch": 0.48704425590460904, + "grad_norm": 0.45755950459270484, + "learning_rate": 0.0005452423259323235, + "loss": 2.5252, + "step": 2124 + }, + { + "epoch": 0.4872735611098372, + "grad_norm": 0.4488563525701135, + "learning_rate": 0.0005448724901589107, + "loss": 2.3606, + "step": 2125 + }, + { + "epoch": 0.48750286631506534, + "grad_norm": 0.5545392299170803, + "learning_rate": 0.0005445026296341325, + "loss": 2.4236, + "step": 2126 + }, + { + "epoch": 0.4877321715202935, + "grad_norm": 0.46672817517528675, + "learning_rate": 0.0005441327445620014, + "loss": 2.465, + "step": 2127 + }, + { + "epoch": 0.48796147672552165, + "grad_norm": 0.5270492526614021, + "learning_rate": 0.0005437628351465435, + "loss": 2.4977, + "step": 2128 + }, + { + "epoch": 0.48819078193074983, + "grad_norm": 0.4474838325778462, + "learning_rate": 0.0005433929015917988, + "loss": 2.5577, + "step": 2129 + }, + { + "epoch": 0.488420087135978, + "grad_norm": 0.43857156675711184, + "learning_rate": 0.0005430229441018197, + "loss": 2.5373, + "step": 2130 + }, + { + "epoch": 0.48864939234120613, + "grad_norm": 0.4985831181307391, + "learning_rate": 0.0005426529628806724, + "loss": 2.535, + "step": 2131 + }, + { + "epoch": 0.4888786975464343, + "grad_norm": 0.44986877795163793, + "learning_rate": 0.0005422829581324361, + "loss": 2.4766, + "step": 2132 + }, + { + "epoch": 0.48910800275166244, + "grad_norm": 0.5154163309322665, + "learning_rate": 0.0005419129300612029, + "loss": 2.4346, + "step": 2133 + }, + { + "epoch": 0.4893373079568906, + "grad_norm": 0.5031060664768799, + "learning_rate": 0.000541542878871078, + "loss": 2.4732, + "step": 2134 + }, + { + "epoch": 0.4895666131621188, + "grad_norm": 1.5244772358437393, + "learning_rate": 0.000541172804766179, + "loss": 2.4978, + "step": 2135 + }, + { + "epoch": 0.4897959183673469, + "grad_norm": 0.42252371618073253, + "learning_rate": 0.0005408027079506361, + "loss": 2.5121, + "step": 2136 + }, + { + "epoch": 0.4900252235725751, + "grad_norm": 0.5158450384819795, + "learning_rate": 0.0005404325886285927, + "loss": 2.5623, + "step": 2137 + }, + { + "epoch": 0.49025452877780323, + "grad_norm": 0.8462381132972169, + "learning_rate": 0.0005400624470042037, + "loss": 2.5182, + "step": 2138 + }, + { + "epoch": 0.4904838339830314, + "grad_norm": 0.6428200802216733, + "learning_rate": 0.000539692283281637, + "loss": 2.3407, + "step": 2139 + }, + { + "epoch": 0.4907131391882596, + "grad_norm": 0.5441967451393024, + "learning_rate": 0.0005393220976650724, + "loss": 2.5149, + "step": 2140 + }, + { + "epoch": 0.4909424443934877, + "grad_norm": 0.8918750299410828, + "learning_rate": 0.0005389518903587017, + "loss": 2.6005, + "step": 2141 + }, + { + "epoch": 0.4911717495987159, + "grad_norm": 0.5248647118582926, + "learning_rate": 0.0005385816615667287, + "loss": 2.5975, + "step": 2142 + }, + { + "epoch": 0.491401054803944, + "grad_norm": 0.5580441224051625, + "learning_rate": 0.0005382114114933695, + "loss": 2.465, + "step": 2143 + }, + { + "epoch": 0.4916303600091722, + "grad_norm": 0.4385147200097133, + "learning_rate": 0.0005378411403428513, + "loss": 2.5096, + "step": 2144 + }, + { + "epoch": 0.4918596652144004, + "grad_norm": 0.41051987589825806, + "learning_rate": 0.0005374708483194132, + "loss": 2.3313, + "step": 2145 + }, + { + "epoch": 0.4920889704196285, + "grad_norm": 0.4209846153179242, + "learning_rate": 0.0005371005356273057, + "loss": 2.4798, + "step": 2146 + }, + { + "epoch": 0.4923182756248567, + "grad_norm": 0.42227842999738774, + "learning_rate": 0.000536730202470791, + "loss": 2.4339, + "step": 2147 + }, + { + "epoch": 0.4925475808300848, + "grad_norm": 1.509315015341058, + "learning_rate": 0.0005363598490541425, + "loss": 2.526, + "step": 2148 + }, + { + "epoch": 0.492776886035313, + "grad_norm": 0.4626218987667178, + "learning_rate": 0.0005359894755816443, + "loss": 2.5376, + "step": 2149 + }, + { + "epoch": 0.4930061912405412, + "grad_norm": 0.4037603371470717, + "learning_rate": 0.0005356190822575921, + "loss": 2.4859, + "step": 2150 + }, + { + "epoch": 0.4932354964457693, + "grad_norm": 0.4271101406323239, + "learning_rate": 0.0005352486692862926, + "loss": 2.6607, + "step": 2151 + }, + { + "epoch": 0.4934648016509975, + "grad_norm": 0.4203370671926582, + "learning_rate": 0.0005348782368720626, + "loss": 2.4682, + "step": 2152 + }, + { + "epoch": 0.4936941068562256, + "grad_norm": 0.5203099990107406, + "learning_rate": 0.0005345077852192307, + "loss": 2.5027, + "step": 2153 + }, + { + "epoch": 0.4939234120614538, + "grad_norm": 0.409297013890328, + "learning_rate": 0.0005341373145321351, + "loss": 2.5298, + "step": 2154 + }, + { + "epoch": 0.494152717266682, + "grad_norm": 0.4324630843548042, + "learning_rate": 0.0005337668250151254, + "loss": 2.4333, + "step": 2155 + }, + { + "epoch": 0.4943820224719101, + "grad_norm": 0.6110657819353865, + "learning_rate": 0.0005333963168725609, + "loss": 2.4436, + "step": 2156 + }, + { + "epoch": 0.4946113276771383, + "grad_norm": 0.4260775835443568, + "learning_rate": 0.0005330257903088111, + "loss": 2.4882, + "step": 2157 + }, + { + "epoch": 0.4948406328823664, + "grad_norm": 0.41480311694783356, + "learning_rate": 0.0005326552455282566, + "loss": 2.5934, + "step": 2158 + }, + { + "epoch": 0.4950699380875946, + "grad_norm": 0.44624790469722253, + "learning_rate": 0.000532284682735287, + "loss": 2.37, + "step": 2159 + }, + { + "epoch": 0.49529924329282277, + "grad_norm": 0.42506185881546904, + "learning_rate": 0.0005319141021343024, + "loss": 2.5581, + "step": 2160 + }, + { + "epoch": 0.4955285484980509, + "grad_norm": 0.4256469132808583, + "learning_rate": 0.0005315435039297124, + "loss": 2.5181, + "step": 2161 + }, + { + "epoch": 0.49575785370327907, + "grad_norm": 1.763725438533562, + "learning_rate": 0.0005311728883259368, + "loss": 2.392, + "step": 2162 + }, + { + "epoch": 0.4959871589085072, + "grad_norm": 0.4260930507813381, + "learning_rate": 0.0005308022555274046, + "loss": 2.5098, + "step": 2163 + }, + { + "epoch": 0.4962164641137354, + "grad_norm": 0.44501000490734033, + "learning_rate": 0.0005304316057385541, + "loss": 2.3076, + "step": 2164 + }, + { + "epoch": 0.49644576931896356, + "grad_norm": 0.46479911411451447, + "learning_rate": 0.0005300609391638336, + "loss": 2.5724, + "step": 2165 + }, + { + "epoch": 0.4966750745241917, + "grad_norm": 0.43207876325826944, + "learning_rate": 0.0005296902560077, + "loss": 2.3981, + "step": 2166 + }, + { + "epoch": 0.49690437972941986, + "grad_norm": 0.5022990749257071, + "learning_rate": 0.0005293195564746201, + "loss": 2.5079, + "step": 2167 + }, + { + "epoch": 0.497133684934648, + "grad_norm": 0.4384066048435348, + "learning_rate": 0.0005289488407690685, + "loss": 2.5359, + "step": 2168 + }, + { + "epoch": 0.49736299013987617, + "grad_norm": 0.39859481098359306, + "learning_rate": 0.0005285781090955304, + "loss": 2.4284, + "step": 2169 + }, + { + "epoch": 0.49759229534510435, + "grad_norm": 0.4421562723872016, + "learning_rate": 0.0005282073616584982, + "loss": 2.4957, + "step": 2170 + }, + { + "epoch": 0.4978216005503325, + "grad_norm": 0.4441158824748812, + "learning_rate": 0.0005278365986624743, + "loss": 2.4998, + "step": 2171 + }, + { + "epoch": 0.49805090575556066, + "grad_norm": 0.4430012434540808, + "learning_rate": 0.0005274658203119684, + "loss": 2.4917, + "step": 2172 + }, + { + "epoch": 0.4982802109607888, + "grad_norm": 0.45591431248662445, + "learning_rate": 0.0005270950268115001, + "loss": 2.5681, + "step": 2173 + }, + { + "epoch": 0.49850951616601696, + "grad_norm": 0.43365455603592995, + "learning_rate": 0.0005267242183655961, + "loss": 2.4486, + "step": 2174 + }, + { + "epoch": 0.49873882137124514, + "grad_norm": 0.6605529943641315, + "learning_rate": 0.0005263533951787919, + "loss": 2.4286, + "step": 2175 + }, + { + "epoch": 0.49896812657647327, + "grad_norm": 0.4948711821908847, + "learning_rate": 0.0005259825574556315, + "loss": 2.4924, + "step": 2176 + }, + { + "epoch": 0.49919743178170145, + "grad_norm": 0.5305774952671369, + "learning_rate": 0.000525611705400666, + "loss": 2.4674, + "step": 2177 + }, + { + "epoch": 0.4994267369869296, + "grad_norm": 0.5224502885620547, + "learning_rate": 0.0005252408392184552, + "loss": 2.4425, + "step": 2178 + }, + { + "epoch": 0.49965604219215776, + "grad_norm": 0.4546056432511788, + "learning_rate": 0.0005248699591135664, + "loss": 2.505, + "step": 2179 + }, + { + "epoch": 0.49988534739738594, + "grad_norm": 0.4886270817473413, + "learning_rate": 0.0005244990652905744, + "loss": 2.5918, + "step": 2180 + }, + { + "epoch": 0.5001146526026141, + "grad_norm": 0.4393260766847396, + "learning_rate": 0.0005241281579540618, + "loss": 2.5573, + "step": 2181 + }, + { + "epoch": 0.5003439578078422, + "grad_norm": 0.43503091024150886, + "learning_rate": 0.0005237572373086188, + "loss": 2.533, + "step": 2182 + }, + { + "epoch": 0.5005732630130704, + "grad_norm": 0.43389637040401036, + "learning_rate": 0.0005233863035588427, + "loss": 2.4652, + "step": 2183 + }, + { + "epoch": 0.5008025682182986, + "grad_norm": 0.5490173897807289, + "learning_rate": 0.0005230153569093377, + "loss": 2.4465, + "step": 2184 + }, + { + "epoch": 0.5010318734235267, + "grad_norm": 0.418568999624073, + "learning_rate": 0.0005226443975647161, + "loss": 2.3663, + "step": 2185 + }, + { + "epoch": 0.5012611786287549, + "grad_norm": 0.4461511435427459, + "learning_rate": 0.0005222734257295963, + "loss": 2.5313, + "step": 2186 + }, + { + "epoch": 0.501490483833983, + "grad_norm": 0.46287785040478174, + "learning_rate": 0.0005219024416086036, + "loss": 2.605, + "step": 2187 + }, + { + "epoch": 0.5017197890392112, + "grad_norm": 0.4375733558499431, + "learning_rate": 0.0005215314454063707, + "loss": 2.3554, + "step": 2188 + }, + { + "epoch": 0.5019490942444393, + "grad_norm": 0.43390714833516664, + "learning_rate": 0.0005211604373275366, + "loss": 2.3405, + "step": 2189 + }, + { + "epoch": 0.5021783994496675, + "grad_norm": 0.4618318562410262, + "learning_rate": 0.000520789417576747, + "loss": 2.4946, + "step": 2190 + }, + { + "epoch": 0.5024077046548957, + "grad_norm": 0.43785889225711017, + "learning_rate": 0.0005204183863586533, + "loss": 2.5342, + "step": 2191 + }, + { + "epoch": 0.5026370098601238, + "grad_norm": 0.43630377309306195, + "learning_rate": 0.0005200473438779147, + "loss": 2.5714, + "step": 2192 + }, + { + "epoch": 0.502866315065352, + "grad_norm": 0.47032933235401325, + "learning_rate": 0.0005196762903391951, + "loss": 2.6601, + "step": 2193 + }, + { + "epoch": 0.5030956202705802, + "grad_norm": 0.4260944467359662, + "learning_rate": 0.0005193052259471655, + "loss": 2.5556, + "step": 2194 + }, + { + "epoch": 0.5033249254758083, + "grad_norm": 0.4342620292528818, + "learning_rate": 0.0005189341509065023, + "loss": 2.5026, + "step": 2195 + }, + { + "epoch": 0.5035542306810364, + "grad_norm": 0.4259345922043714, + "learning_rate": 0.0005185630654218881, + "loss": 2.4334, + "step": 2196 + }, + { + "epoch": 0.5037835358862646, + "grad_norm": 0.5673466317160978, + "learning_rate": 0.0005181919696980112, + "loss": 2.4035, + "step": 2197 + }, + { + "epoch": 0.5040128410914928, + "grad_norm": 0.4356715768551664, + "learning_rate": 0.0005178208639395655, + "loss": 2.6512, + "step": 2198 + }, + { + "epoch": 0.5042421462967209, + "grad_norm": 0.44466700568858963, + "learning_rate": 0.0005174497483512506, + "loss": 2.3855, + "step": 2199 + }, + { + "epoch": 0.504471451501949, + "grad_norm": 0.4134948803632293, + "learning_rate": 0.0005170786231377711, + "loss": 2.3367, + "step": 2200 + }, + { + "epoch": 0.5047007567071773, + "grad_norm": 0.49047759882382014, + "learning_rate": 0.0005167074885038374, + "loss": 2.545, + "step": 2201 + }, + { + "epoch": 0.5049300619124054, + "grad_norm": 0.41341644686224704, + "learning_rate": 0.0005163363446541647, + "loss": 2.4819, + "step": 2202 + }, + { + "epoch": 0.5051593671176335, + "grad_norm": 1.3960664570854024, + "learning_rate": 0.0005159651917934735, + "loss": 2.5136, + "step": 2203 + }, + { + "epoch": 0.5053886723228618, + "grad_norm": 0.413989407233478, + "learning_rate": 0.0005155940301264896, + "loss": 2.377, + "step": 2204 + }, + { + "epoch": 0.5056179775280899, + "grad_norm": 0.4233767273448, + "learning_rate": 0.0005152228598579428, + "loss": 2.4685, + "step": 2205 + }, + { + "epoch": 0.505847282733318, + "grad_norm": 0.4256030471162949, + "learning_rate": 0.0005148516811925684, + "loss": 2.4679, + "step": 2206 + }, + { + "epoch": 0.5060765879385462, + "grad_norm": 0.3984957607652719, + "learning_rate": 0.000514480494335106, + "loss": 2.4306, + "step": 2207 + }, + { + "epoch": 0.5063058931437744, + "grad_norm": 0.4402713907211792, + "learning_rate": 0.0005141092994902998, + "loss": 2.5565, + "step": 2208 + }, + { + "epoch": 0.5065351983490025, + "grad_norm": 0.39770696905612646, + "learning_rate": 0.0005137380968628983, + "loss": 2.4963, + "step": 2209 + }, + { + "epoch": 0.5067645035542306, + "grad_norm": 0.41486746032029564, + "learning_rate": 0.0005133668866576544, + "loss": 2.3575, + "step": 2210 + }, + { + "epoch": 0.5069938087594589, + "grad_norm": 0.40681886781262455, + "learning_rate": 0.0005129956690793255, + "loss": 2.5218, + "step": 2211 + }, + { + "epoch": 0.507223113964687, + "grad_norm": 0.39538492742104725, + "learning_rate": 0.0005126244443326722, + "loss": 2.3489, + "step": 2212 + }, + { + "epoch": 0.5074524191699151, + "grad_norm": 0.4425131097717793, + "learning_rate": 0.0005122532126224601, + "loss": 2.5174, + "step": 2213 + }, + { + "epoch": 0.5076817243751434, + "grad_norm": 0.41247114825687553, + "learning_rate": 0.0005118819741534577, + "loss": 2.4004, + "step": 2214 + }, + { + "epoch": 0.5079110295803715, + "grad_norm": 0.41028273936860343, + "learning_rate": 0.0005115107291304378, + "loss": 2.4288, + "step": 2215 + }, + { + "epoch": 0.5081403347855996, + "grad_norm": 0.41886707600605394, + "learning_rate": 0.000511139477758177, + "loss": 2.5178, + "step": 2216 + }, + { + "epoch": 0.5083696399908277, + "grad_norm": 0.40458510104485346, + "learning_rate": 0.0005107682202414544, + "loss": 2.4499, + "step": 2217 + }, + { + "epoch": 0.508598945196056, + "grad_norm": 0.40299237806008353, + "learning_rate": 0.0005103969567850538, + "loss": 2.457, + "step": 2218 + }, + { + "epoch": 0.5088282504012841, + "grad_norm": 0.4082496032852673, + "learning_rate": 0.0005100256875937613, + "loss": 2.4473, + "step": 2219 + }, + { + "epoch": 0.5090575556065122, + "grad_norm": 0.40505262789849006, + "learning_rate": 0.0005096544128723667, + "loss": 2.448, + "step": 2220 + }, + { + "epoch": 0.5092868608117405, + "grad_norm": 0.4228224648456252, + "learning_rate": 0.0005092831328256625, + "loss": 2.4135, + "step": 2221 + }, + { + "epoch": 0.5095161660169686, + "grad_norm": 0.42635403919342163, + "learning_rate": 0.0005089118476584442, + "loss": 2.4729, + "step": 2222 + }, + { + "epoch": 0.5097454712221967, + "grad_norm": 0.43412215092799156, + "learning_rate": 0.0005085405575755105, + "loss": 2.5093, + "step": 2223 + }, + { + "epoch": 0.509974776427425, + "grad_norm": 0.4116578632258648, + "learning_rate": 0.0005081692627816624, + "loss": 2.3996, + "step": 2224 + }, + { + "epoch": 0.5102040816326531, + "grad_norm": 0.4186429702006985, + "learning_rate": 0.0005077979634817034, + "loss": 2.5861, + "step": 2225 + }, + { + "epoch": 0.5104333868378812, + "grad_norm": 0.3768852998676308, + "learning_rate": 0.0005074266598804402, + "loss": 2.4582, + "step": 2226 + }, + { + "epoch": 0.5106626920431094, + "grad_norm": 0.4009082274862905, + "learning_rate": 0.0005070553521826808, + "loss": 2.4025, + "step": 2227 + }, + { + "epoch": 0.5108919972483376, + "grad_norm": 0.5675821011982236, + "learning_rate": 0.0005066840405932363, + "loss": 2.6226, + "step": 2228 + }, + { + "epoch": 0.5111213024535657, + "grad_norm": 0.4196601732905991, + "learning_rate": 0.00050631272531692, + "loss": 2.4869, + "step": 2229 + }, + { + "epoch": 0.5113506076587938, + "grad_norm": 0.4019177557479886, + "learning_rate": 0.0005059414065585465, + "loss": 2.443, + "step": 2230 + }, + { + "epoch": 0.511579912864022, + "grad_norm": 0.3827107025950233, + "learning_rate": 0.0005055700845229327, + "loss": 2.3424, + "step": 2231 + }, + { + "epoch": 0.5118092180692502, + "grad_norm": 0.427636638435452, + "learning_rate": 0.0005051987594148979, + "loss": 2.5062, + "step": 2232 + }, + { + "epoch": 0.5120385232744783, + "grad_norm": 0.5175129784950776, + "learning_rate": 0.000504827431439262, + "loss": 2.4487, + "step": 2233 + }, + { + "epoch": 0.5122678284797065, + "grad_norm": 0.41754377213977983, + "learning_rate": 0.0005044561008008472, + "loss": 2.5052, + "step": 2234 + }, + { + "epoch": 0.5124971336849347, + "grad_norm": 0.4110117925899411, + "learning_rate": 0.000504084767704477, + "loss": 2.5831, + "step": 2235 + }, + { + "epoch": 0.5127264388901628, + "grad_norm": 0.42649388784949765, + "learning_rate": 0.0005037134323549763, + "loss": 2.3226, + "step": 2236 + }, + { + "epoch": 0.512955744095391, + "grad_norm": 0.44534959373772104, + "learning_rate": 0.0005033420949571712, + "loss": 2.511, + "step": 2237 + }, + { + "epoch": 0.5131850493006191, + "grad_norm": 0.436927194055244, + "learning_rate": 0.0005029707557158887, + "loss": 2.4255, + "step": 2238 + }, + { + "epoch": 0.5134143545058473, + "grad_norm": 0.44018552237438097, + "learning_rate": 0.0005025994148359574, + "loss": 2.4731, + "step": 2239 + }, + { + "epoch": 0.5136436597110754, + "grad_norm": 0.5374482836896353, + "learning_rate": 0.0005022280725222063, + "loss": 2.4679, + "step": 2240 + }, + { + "epoch": 0.5138729649163036, + "grad_norm": 0.41197253081735485, + "learning_rate": 0.0005018567289794651, + "loss": 2.4599, + "step": 2241 + }, + { + "epoch": 0.5141022701215318, + "grad_norm": 0.4640454505688113, + "learning_rate": 0.0005014853844125648, + "loss": 2.4078, + "step": 2242 + }, + { + "epoch": 0.5143315753267599, + "grad_norm": 0.3966224835973591, + "learning_rate": 0.0005011140390263362, + "loss": 2.4671, + "step": 2243 + }, + { + "epoch": 0.5145608805319881, + "grad_norm": 0.42620958579602736, + "learning_rate": 0.0005007426930256113, + "loss": 2.3244, + "step": 2244 + }, + { + "epoch": 0.5147901857372162, + "grad_norm": 0.4586731746673869, + "learning_rate": 0.0005003713466152218, + "loss": 2.4717, + "step": 2245 + }, + { + "epoch": 0.5150194909424444, + "grad_norm": 0.40694691805442845, + "learning_rate": 0.0005, + "loss": 2.4364, + "step": 2246 + }, + { + "epoch": 0.5152487961476726, + "grad_norm": 0.43917359238149567, + "learning_rate": 0.0004996286533847783, + "loss": 2.6707, + "step": 2247 + }, + { + "epoch": 0.5154781013529007, + "grad_norm": 0.44731210114618136, + "learning_rate": 0.0004992573069743887, + "loss": 2.4208, + "step": 2248 + }, + { + "epoch": 0.5157074065581289, + "grad_norm": 0.4075568770086156, + "learning_rate": 0.000498885960973664, + "loss": 2.3954, + "step": 2249 + }, + { + "epoch": 0.515936711763357, + "grad_norm": 0.44869351367798743, + "learning_rate": 0.0004985146155874353, + "loss": 2.4879, + "step": 2250 + }, + { + "epoch": 0.5161660169685852, + "grad_norm": 0.3886994679594167, + "learning_rate": 0.000498143271020535, + "loss": 2.5888, + "step": 2251 + }, + { + "epoch": 0.5163953221738133, + "grad_norm": 0.44420108294478017, + "learning_rate": 0.0004977719274777938, + "loss": 2.4716, + "step": 2252 + }, + { + "epoch": 0.5166246273790415, + "grad_norm": 0.40392450855246526, + "learning_rate": 0.0004974005851640428, + "loss": 2.4404, + "step": 2253 + }, + { + "epoch": 0.5168539325842697, + "grad_norm": 0.3996611289065144, + "learning_rate": 0.0004970292442841113, + "loss": 2.4852, + "step": 2254 + }, + { + "epoch": 0.5170832377894978, + "grad_norm": 0.4181344782351684, + "learning_rate": 0.000496657905042829, + "loss": 2.4496, + "step": 2255 + }, + { + "epoch": 0.517312542994726, + "grad_norm": 0.4097762258789988, + "learning_rate": 0.0004962865676450238, + "loss": 2.3739, + "step": 2256 + }, + { + "epoch": 0.5175418481999542, + "grad_norm": 0.40108167996072147, + "learning_rate": 0.0004959152322955232, + "loss": 2.4516, + "step": 2257 + }, + { + "epoch": 0.5177711534051823, + "grad_norm": 0.404222519180616, + "learning_rate": 0.0004955438991991528, + "loss": 2.5132, + "step": 2258 + }, + { + "epoch": 0.5180004586104104, + "grad_norm": 0.3924642333613604, + "learning_rate": 0.0004951725685607382, + "loss": 2.4512, + "step": 2259 + }, + { + "epoch": 0.5182297638156386, + "grad_norm": 0.39672483989989077, + "learning_rate": 0.0004948012405851022, + "loss": 2.4381, + "step": 2260 + }, + { + "epoch": 0.5184590690208668, + "grad_norm": 0.41208938960988123, + "learning_rate": 0.0004944299154770673, + "loss": 2.5629, + "step": 2261 + }, + { + "epoch": 0.5186883742260949, + "grad_norm": 0.4427552200040141, + "learning_rate": 0.0004940585934414535, + "loss": 2.6199, + "step": 2262 + }, + { + "epoch": 0.518917679431323, + "grad_norm": 0.40192760503267755, + "learning_rate": 0.0004936872746830802, + "loss": 2.5309, + "step": 2263 + }, + { + "epoch": 0.5191469846365513, + "grad_norm": 0.42588213841649036, + "learning_rate": 0.0004933159594067637, + "loss": 2.6225, + "step": 2264 + }, + { + "epoch": 0.5193762898417794, + "grad_norm": 0.42163153621725463, + "learning_rate": 0.0004929446478173195, + "loss": 2.563, + "step": 2265 + }, + { + "epoch": 0.5196055950470075, + "grad_norm": 0.372622274058016, + "learning_rate": 0.00049257334011956, + "loss": 2.3714, + "step": 2266 + }, + { + "epoch": 0.5198349002522358, + "grad_norm": 0.40337769060055517, + "learning_rate": 0.0004922020365182968, + "loss": 2.4133, + "step": 2267 + }, + { + "epoch": 0.5200642054574639, + "grad_norm": 0.3842767003310132, + "learning_rate": 0.0004918307372183378, + "loss": 2.369, + "step": 2268 + }, + { + "epoch": 0.520293510662692, + "grad_norm": 0.38113373788595406, + "learning_rate": 0.0004914594424244897, + "loss": 2.4619, + "step": 2269 + }, + { + "epoch": 0.5205228158679202, + "grad_norm": 0.45593155562405774, + "learning_rate": 0.0004910881523415558, + "loss": 2.5382, + "step": 2270 + }, + { + "epoch": 0.5207521210731484, + "grad_norm": 0.429516733912204, + "learning_rate": 0.0004907168671743376, + "loss": 2.502, + "step": 2271 + }, + { + "epoch": 0.5209814262783765, + "grad_norm": 0.42683974790069246, + "learning_rate": 0.0004903455871276334, + "loss": 2.5519, + "step": 2272 + }, + { + "epoch": 0.5212107314836046, + "grad_norm": 0.5249532296294631, + "learning_rate": 0.0004899743124062387, + "loss": 2.5586, + "step": 2273 + }, + { + "epoch": 0.5214400366888329, + "grad_norm": 0.4569992901104077, + "learning_rate": 0.0004896030432149462, + "loss": 2.5273, + "step": 2274 + }, + { + "epoch": 0.521669341894061, + "grad_norm": 0.45662529555076264, + "learning_rate": 0.0004892317797585456, + "loss": 2.6642, + "step": 2275 + }, + { + "epoch": 0.5218986470992891, + "grad_norm": 0.4394161329821294, + "learning_rate": 0.0004888605222418232, + "loss": 2.5043, + "step": 2276 + }, + { + "epoch": 0.5221279523045174, + "grad_norm": 0.3980508662197439, + "learning_rate": 0.0004884892708695623, + "loss": 2.5582, + "step": 2277 + }, + { + "epoch": 0.5223572575097455, + "grad_norm": 0.4545543744306781, + "learning_rate": 0.00048811802584654236, + "loss": 2.4175, + "step": 2278 + }, + { + "epoch": 0.5225865627149736, + "grad_norm": 0.4222421258234301, + "learning_rate": 0.0004877467873775402, + "loss": 2.5744, + "step": 2279 + }, + { + "epoch": 0.5228158679202017, + "grad_norm": 0.40084517770540296, + "learning_rate": 0.00048737555566732787, + "loss": 2.4818, + "step": 2280 + }, + { + "epoch": 0.52304517312543, + "grad_norm": 0.412582058879235, + "learning_rate": 0.00048700433092067473, + "loss": 2.5119, + "step": 2281 + }, + { + "epoch": 0.5232744783306581, + "grad_norm": 0.42824844666723133, + "learning_rate": 0.0004866331133423456, + "loss": 2.2419, + "step": 2282 + }, + { + "epoch": 0.5235037835358862, + "grad_norm": 0.42265314163284706, + "learning_rate": 0.0004862619031371019, + "loss": 2.5694, + "step": 2283 + }, + { + "epoch": 0.5237330887411145, + "grad_norm": 0.4256267280420792, + "learning_rate": 0.0004858907005097004, + "loss": 2.3704, + "step": 2284 + }, + { + "epoch": 0.5239623939463426, + "grad_norm": 0.46397604351525135, + "learning_rate": 0.0004855195056648942, + "loss": 2.5646, + "step": 2285 + }, + { + "epoch": 0.5241916991515707, + "grad_norm": 0.44422731936223614, + "learning_rate": 0.0004851483188074317, + "loss": 2.5033, + "step": 2286 + }, + { + "epoch": 0.524421004356799, + "grad_norm": 0.4244062132014539, + "learning_rate": 0.00048477714014205734, + "loss": 2.4613, + "step": 2287 + }, + { + "epoch": 0.5246503095620271, + "grad_norm": 0.4567358919051459, + "learning_rate": 0.0004844059698735105, + "loss": 2.5206, + "step": 2288 + }, + { + "epoch": 0.5248796147672552, + "grad_norm": 0.4167893607953319, + "learning_rate": 0.00048403480820652644, + "loss": 2.3985, + "step": 2289 + }, + { + "epoch": 0.5251089199724833, + "grad_norm": 0.42007423147734024, + "learning_rate": 0.00048366365534583533, + "loss": 2.5313, + "step": 2290 + }, + { + "epoch": 0.5253382251777116, + "grad_norm": 0.4234092416192092, + "learning_rate": 0.0004832925114961629, + "loss": 2.6587, + "step": 2291 + }, + { + "epoch": 0.5255675303829397, + "grad_norm": 0.3965469855209245, + "learning_rate": 0.0004829213768622289, + "loss": 2.388, + "step": 2292 + }, + { + "epoch": 0.5257968355881678, + "grad_norm": 0.4021460222386118, + "learning_rate": 0.0004825502516487497, + "loss": 2.5572, + "step": 2293 + }, + { + "epoch": 0.526026140793396, + "grad_norm": 0.3771439851903524, + "learning_rate": 0.00048217913606043457, + "loss": 2.3489, + "step": 2294 + }, + { + "epoch": 0.5262554459986242, + "grad_norm": 0.4185225493136941, + "learning_rate": 0.00048180803030198896, + "loss": 2.4919, + "step": 2295 + }, + { + "epoch": 0.5264847512038523, + "grad_norm": 0.4144097647269471, + "learning_rate": 0.0004814369345781121, + "loss": 2.515, + "step": 2296 + }, + { + "epoch": 0.5267140564090805, + "grad_norm": 0.39798057145566457, + "learning_rate": 0.0004810658490934979, + "loss": 2.4122, + "step": 2297 + }, + { + "epoch": 0.5269433616143087, + "grad_norm": 0.43720425800065926, + "learning_rate": 0.0004806947740528347, + "loss": 2.5203, + "step": 2298 + }, + { + "epoch": 0.5271726668195368, + "grad_norm": 0.4426142282027905, + "learning_rate": 0.000480323709660805, + "loss": 2.5776, + "step": 2299 + }, + { + "epoch": 0.5274019720247649, + "grad_norm": 0.4411436188779373, + "learning_rate": 0.00047995265612208544, + "loss": 2.4511, + "step": 2300 + }, + { + "epoch": 0.5276312772299931, + "grad_norm": 0.41867362937665287, + "learning_rate": 0.0004795816136413467, + "loss": 2.4305, + "step": 2301 + }, + { + "epoch": 0.5278605824352213, + "grad_norm": 0.39518952899712434, + "learning_rate": 0.00047921058242325317, + "loss": 2.4421, + "step": 2302 + }, + { + "epoch": 0.5280898876404494, + "grad_norm": 0.44998605010401777, + "learning_rate": 0.00047883956267246353, + "loss": 2.5008, + "step": 2303 + }, + { + "epoch": 0.5283191928456776, + "grad_norm": 0.40344905167108797, + "learning_rate": 0.00047846855459362934, + "loss": 2.5562, + "step": 2304 + }, + { + "epoch": 0.5285484980509058, + "grad_norm": 0.41417328847122376, + "learning_rate": 0.00047809755839139657, + "loss": 2.5319, + "step": 2305 + }, + { + "epoch": 0.5287778032561339, + "grad_norm": 0.4035995357363959, + "learning_rate": 0.00047772657427040384, + "loss": 2.603, + "step": 2306 + }, + { + "epoch": 0.5290071084613621, + "grad_norm": 0.4097517594785107, + "learning_rate": 0.0004773556024352841, + "loss": 2.6278, + "step": 2307 + }, + { + "epoch": 0.5292364136665902, + "grad_norm": 0.435229878453973, + "learning_rate": 0.00047698464309066233, + "loss": 2.5505, + "step": 2308 + }, + { + "epoch": 0.5294657188718184, + "grad_norm": 0.37662577395392266, + "learning_rate": 0.00047661369644115754, + "loss": 2.5349, + "step": 2309 + }, + { + "epoch": 0.5296950240770465, + "grad_norm": 0.41832631156628336, + "learning_rate": 0.0004762427626913813, + "loss": 2.5425, + "step": 2310 + }, + { + "epoch": 0.5299243292822747, + "grad_norm": 0.4369701380454565, + "learning_rate": 0.0004758718420459383, + "loss": 2.5378, + "step": 2311 + }, + { + "epoch": 0.5301536344875029, + "grad_norm": 0.41093038656815495, + "learning_rate": 0.00047550093470942576, + "loss": 2.5034, + "step": 2312 + }, + { + "epoch": 0.530382939692731, + "grad_norm": 0.4117571814688317, + "learning_rate": 0.0004751300408864339, + "loss": 2.4367, + "step": 2313 + }, + { + "epoch": 0.5306122448979592, + "grad_norm": 0.3883977389205128, + "learning_rate": 0.0004747591607815449, + "loss": 2.3751, + "step": 2314 + }, + { + "epoch": 0.5308415501031873, + "grad_norm": 0.4385905870831135, + "learning_rate": 0.00047438829459933414, + "loss": 2.4238, + "step": 2315 + }, + { + "epoch": 0.5310708553084155, + "grad_norm": 0.4116399012323293, + "learning_rate": 0.00047401744254436864, + "loss": 2.4382, + "step": 2316 + }, + { + "epoch": 0.5313001605136437, + "grad_norm": 0.43111238214205055, + "learning_rate": 0.0004736466048212082, + "loss": 2.3633, + "step": 2317 + }, + { + "epoch": 0.5315294657188718, + "grad_norm": 0.46806759906604634, + "learning_rate": 0.000473275781634404, + "loss": 2.5336, + "step": 2318 + }, + { + "epoch": 0.5317587709241, + "grad_norm": 0.4407318433137697, + "learning_rate": 0.0004729049731885002, + "loss": 2.4047, + "step": 2319 + }, + { + "epoch": 0.5319880761293282, + "grad_norm": 0.38444077039345204, + "learning_rate": 0.00047253417968803164, + "loss": 2.5162, + "step": 2320 + }, + { + "epoch": 0.5322173813345563, + "grad_norm": 0.416029490371269, + "learning_rate": 0.000472163401337526, + "loss": 2.4129, + "step": 2321 + }, + { + "epoch": 0.5324466865397844, + "grad_norm": 0.36621545473542527, + "learning_rate": 0.0004717926383415019, + "loss": 2.3239, + "step": 2322 + }, + { + "epoch": 0.5326759917450126, + "grad_norm": 0.44368027970672214, + "learning_rate": 0.00047142189090446985, + "loss": 2.3668, + "step": 2323 + }, + { + "epoch": 0.5329052969502408, + "grad_norm": 0.4333951656949447, + "learning_rate": 0.0004710511592309315, + "loss": 2.4195, + "step": 2324 + }, + { + "epoch": 0.5331346021554689, + "grad_norm": 0.4404711039368185, + "learning_rate": 0.0004706804435253802, + "loss": 2.385, + "step": 2325 + }, + { + "epoch": 0.533363907360697, + "grad_norm": 0.39419427636700693, + "learning_rate": 0.0004703097439923, + "loss": 2.471, + "step": 2326 + }, + { + "epoch": 0.5335932125659253, + "grad_norm": 0.3999085224145573, + "learning_rate": 0.0004699390608361665, + "loss": 2.3765, + "step": 2327 + }, + { + "epoch": 0.5338225177711534, + "grad_norm": 0.48949107300057804, + "learning_rate": 0.0004695683942614459, + "loss": 2.5119, + "step": 2328 + }, + { + "epoch": 0.5340518229763815, + "grad_norm": 0.44293819044774335, + "learning_rate": 0.0004691977444725955, + "loss": 2.5105, + "step": 2329 + }, + { + "epoch": 0.5342811281816098, + "grad_norm": 0.44937909393452363, + "learning_rate": 0.0004688271116740632, + "loss": 2.4825, + "step": 2330 + }, + { + "epoch": 0.5345104333868379, + "grad_norm": 0.4517007777054763, + "learning_rate": 0.0004684564960702877, + "loss": 2.4944, + "step": 2331 + }, + { + "epoch": 0.534739738592066, + "grad_norm": 0.3987809517879934, + "learning_rate": 0.0004680858978656976, + "loss": 2.3866, + "step": 2332 + }, + { + "epoch": 0.5349690437972942, + "grad_norm": 0.3920900045648854, + "learning_rate": 0.0004677153172647131, + "loss": 2.36, + "step": 2333 + }, + { + "epoch": 0.5351983490025224, + "grad_norm": 0.41940923733080754, + "learning_rate": 0.00046734475447174345, + "loss": 2.4975, + "step": 2334 + }, + { + "epoch": 0.5354276542077505, + "grad_norm": 0.38379136510360573, + "learning_rate": 0.00046697420969118894, + "loss": 2.4781, + "step": 2335 + }, + { + "epoch": 0.5356569594129786, + "grad_norm": 0.39822941580126975, + "learning_rate": 0.0004666036831274392, + "loss": 2.5004, + "step": 2336 + }, + { + "epoch": 0.5358862646182069, + "grad_norm": 0.3930593425722889, + "learning_rate": 0.00046623317498487466, + "loss": 2.4225, + "step": 2337 + }, + { + "epoch": 0.536115569823435, + "grad_norm": 0.523705005518738, + "learning_rate": 0.0004658626854678648, + "loss": 2.4389, + "step": 2338 + }, + { + "epoch": 0.5363448750286631, + "grad_norm": 0.4149250237085271, + "learning_rate": 0.0004654922147807694, + "loss": 2.4248, + "step": 2339 + }, + { + "epoch": 0.5365741802338914, + "grad_norm": 0.4216228804733719, + "learning_rate": 0.00046512176312793734, + "loss": 2.5507, + "step": 2340 + }, + { + "epoch": 0.5368034854391195, + "grad_norm": 0.4082164290689722, + "learning_rate": 0.00046475133071370757, + "loss": 2.6263, + "step": 2341 + }, + { + "epoch": 0.5370327906443476, + "grad_norm": 0.47369641369530396, + "learning_rate": 0.0004643809177424078, + "loss": 2.5224, + "step": 2342 + }, + { + "epoch": 0.5372620958495757, + "grad_norm": 0.39776667504714125, + "learning_rate": 0.00046401052441835574, + "loss": 2.3246, + "step": 2343 + }, + { + "epoch": 0.537491401054804, + "grad_norm": 0.4421694316959335, + "learning_rate": 0.0004636401509458575, + "loss": 2.4237, + "step": 2344 + }, + { + "epoch": 0.5377207062600321, + "grad_norm": 0.4046353157541414, + "learning_rate": 0.000463269797529209, + "loss": 2.334, + "step": 2345 + }, + { + "epoch": 0.5379500114652602, + "grad_norm": 0.4158625569290717, + "learning_rate": 0.0004628994643726941, + "loss": 2.3543, + "step": 2346 + }, + { + "epoch": 0.5381793166704885, + "grad_norm": 0.550915400222488, + "learning_rate": 0.00046252915168058697, + "loss": 2.5267, + "step": 2347 + }, + { + "epoch": 0.5384086218757166, + "grad_norm": 0.4100655512770807, + "learning_rate": 0.00046215885965714875, + "loss": 2.4296, + "step": 2348 + }, + { + "epoch": 0.5386379270809447, + "grad_norm": 0.40825239145972014, + "learning_rate": 0.0004617885885066305, + "loss": 2.3921, + "step": 2349 + }, + { + "epoch": 0.538867232286173, + "grad_norm": 0.4295998908010766, + "learning_rate": 0.0004614183384332712, + "loss": 2.4379, + "step": 2350 + }, + { + "epoch": 0.5390965374914011, + "grad_norm": 0.40037709213306655, + "learning_rate": 0.0004610481096412984, + "loss": 2.4854, + "step": 2351 + }, + { + "epoch": 0.5393258426966292, + "grad_norm": 0.41183095179910073, + "learning_rate": 0.0004606779023349277, + "loss": 2.3872, + "step": 2352 + }, + { + "epoch": 0.5395551479018573, + "grad_norm": 0.40672366642484986, + "learning_rate": 0.000460307716718363, + "loss": 2.5664, + "step": 2353 + }, + { + "epoch": 0.5397844531070856, + "grad_norm": 0.411827756402086, + "learning_rate": 0.00045993755299579623, + "loss": 2.411, + "step": 2354 + }, + { + "epoch": 0.5400137583123137, + "grad_norm": 0.460481512659573, + "learning_rate": 0.0004595674113714074, + "loss": 2.5419, + "step": 2355 + }, + { + "epoch": 0.5402430635175418, + "grad_norm": 0.3910684104396047, + "learning_rate": 0.0004591972920493638, + "loss": 2.4677, + "step": 2356 + }, + { + "epoch": 0.54047236872277, + "grad_norm": 0.4318139819047241, + "learning_rate": 0.0004588271952338212, + "loss": 2.4878, + "step": 2357 + }, + { + "epoch": 0.5407016739279982, + "grad_norm": 0.38480854820778365, + "learning_rate": 0.00045845712112892195, + "loss": 2.4856, + "step": 2358 + }, + { + "epoch": 0.5409309791332263, + "grad_norm": 0.4062470669582389, + "learning_rate": 0.00045808706993879714, + "loss": 2.553, + "step": 2359 + }, + { + "epoch": 0.5411602843384545, + "grad_norm": 0.40694046981463655, + "learning_rate": 0.0004577170418675638, + "loss": 2.447, + "step": 2360 + }, + { + "epoch": 0.5413895895436827, + "grad_norm": 0.3920165904134509, + "learning_rate": 0.00045734703711932767, + "loss": 2.3509, + "step": 2361 + }, + { + "epoch": 0.5416188947489108, + "grad_norm": 0.387326247123156, + "learning_rate": 0.00045697705589818036, + "loss": 2.5268, + "step": 2362 + }, + { + "epoch": 0.5418481999541389, + "grad_norm": 0.43454856623788957, + "learning_rate": 0.0004566070984082013, + "loss": 2.5358, + "step": 2363 + }, + { + "epoch": 0.5420775051593671, + "grad_norm": 0.5367945822628232, + "learning_rate": 0.0004562371648534564, + "loss": 2.4135, + "step": 2364 + }, + { + "epoch": 0.5423068103645953, + "grad_norm": 0.4358215922971502, + "learning_rate": 0.00045586725543799865, + "loss": 2.4151, + "step": 2365 + }, + { + "epoch": 0.5425361155698234, + "grad_norm": 0.40785277870164555, + "learning_rate": 0.0004554973703658676, + "loss": 2.3746, + "step": 2366 + }, + { + "epoch": 0.5427654207750516, + "grad_norm": 0.4018013104357245, + "learning_rate": 0.00045512750984108937, + "loss": 2.3874, + "step": 2367 + }, + { + "epoch": 0.5429947259802798, + "grad_norm": 0.41665356278933297, + "learning_rate": 0.0004547576740676765, + "loss": 2.4338, + "step": 2368 + }, + { + "epoch": 0.5432240311855079, + "grad_norm": 0.4209818770922377, + "learning_rate": 0.000454387863249628, + "loss": 2.571, + "step": 2369 + }, + { + "epoch": 0.5434533363907361, + "grad_norm": 0.408631412307893, + "learning_rate": 0.00045401807759092895, + "loss": 2.4917, + "step": 2370 + }, + { + "epoch": 0.5436826415959642, + "grad_norm": 0.42880608895236677, + "learning_rate": 0.00045364831729555096, + "loss": 2.4506, + "step": 2371 + }, + { + "epoch": 0.5439119468011924, + "grad_norm": 0.3969894777121708, + "learning_rate": 0.0004532785825674507, + "loss": 2.4905, + "step": 2372 + }, + { + "epoch": 0.5441412520064205, + "grad_norm": 0.4006800146992838, + "learning_rate": 0.0004529088736105721, + "loss": 2.3787, + "step": 2373 + }, + { + "epoch": 0.5443705572116487, + "grad_norm": 0.4155088450293192, + "learning_rate": 0.00045253919062884356, + "loss": 2.5269, + "step": 2374 + }, + { + "epoch": 0.5445998624168769, + "grad_norm": 0.4174232571193688, + "learning_rate": 0.0004521695338261802, + "loss": 2.6293, + "step": 2375 + }, + { + "epoch": 0.544829167622105, + "grad_norm": 0.4319016728387888, + "learning_rate": 0.0004517999034064818, + "loss": 2.4991, + "step": 2376 + }, + { + "epoch": 0.5450584728273332, + "grad_norm": 0.4125240512504525, + "learning_rate": 0.0004514302995736344, + "loss": 2.5271, + "step": 2377 + }, + { + "epoch": 0.5452877780325613, + "grad_norm": 0.410185207405131, + "learning_rate": 0.0004510607225315088, + "loss": 2.4223, + "step": 2378 + }, + { + "epoch": 0.5455170832377895, + "grad_norm": 0.39044141394279014, + "learning_rate": 0.0004506911724839613, + "loss": 2.4571, + "step": 2379 + }, + { + "epoch": 0.5457463884430177, + "grad_norm": 0.40532619609568193, + "learning_rate": 0.0004503216496348331, + "loss": 2.513, + "step": 2380 + }, + { + "epoch": 0.5459756936482458, + "grad_norm": 0.4416730677003992, + "learning_rate": 0.0004499521541879508, + "loss": 2.5747, + "step": 2381 + }, + { + "epoch": 0.546204998853474, + "grad_norm": 0.4289793496999646, + "learning_rate": 0.0004495826863471254, + "loss": 2.3668, + "step": 2382 + }, + { + "epoch": 0.5464343040587021, + "grad_norm": 0.46735404960788773, + "learning_rate": 0.00044921324631615303, + "loss": 2.4383, + "step": 2383 + }, + { + "epoch": 0.5466636092639303, + "grad_norm": 0.430157697544011, + "learning_rate": 0.0004488438342988141, + "loss": 2.4008, + "step": 2384 + }, + { + "epoch": 0.5468929144691584, + "grad_norm": 0.3980692003178823, + "learning_rate": 0.0004484744504988742, + "loss": 2.6264, + "step": 2385 + }, + { + "epoch": 0.5471222196743866, + "grad_norm": 0.3636110017249839, + "learning_rate": 0.0004481050951200824, + "loss": 2.472, + "step": 2386 + }, + { + "epoch": 0.5473515248796148, + "grad_norm": 0.3997247323704613, + "learning_rate": 0.00044773576836617336, + "loss": 2.4791, + "step": 2387 + }, + { + "epoch": 0.5475808300848429, + "grad_norm": 0.42334499610679704, + "learning_rate": 0.0004473664704408646, + "loss": 2.3946, + "step": 2388 + }, + { + "epoch": 0.547810135290071, + "grad_norm": 0.39335451917560327, + "learning_rate": 0.0004469972015478588, + "loss": 2.4821, + "step": 2389 + }, + { + "epoch": 0.5480394404952993, + "grad_norm": 0.42558476122940847, + "learning_rate": 0.00044662796189084194, + "loss": 2.4606, + "step": 2390 + }, + { + "epoch": 0.5482687457005274, + "grad_norm": 0.45964454717666187, + "learning_rate": 0.0004462587516734844, + "loss": 2.5318, + "step": 2391 + }, + { + "epoch": 0.5484980509057555, + "grad_norm": 0.413656249754307, + "learning_rate": 0.00044588957109943986, + "loss": 2.6802, + "step": 2392 + }, + { + "epoch": 0.5487273561109837, + "grad_norm": 0.3990623397177412, + "learning_rate": 0.00044552042037234596, + "loss": 2.3997, + "step": 2393 + }, + { + "epoch": 0.5489566613162119, + "grad_norm": 0.3959487014169831, + "learning_rate": 0.0004451512996958237, + "loss": 2.4661, + "step": 2394 + }, + { + "epoch": 0.54918596652144, + "grad_norm": 0.42930961286630936, + "learning_rate": 0.00044478220927347774, + "loss": 2.4428, + "step": 2395 + }, + { + "epoch": 0.5494152717266682, + "grad_norm": 0.425025279634262, + "learning_rate": 0.00044441314930889565, + "loss": 2.5077, + "step": 2396 + }, + { + "epoch": 0.5496445769318964, + "grad_norm": 0.44599553645463125, + "learning_rate": 0.00044404412000564875, + "loss": 2.3909, + "step": 2397 + }, + { + "epoch": 0.5498738821371245, + "grad_norm": 0.4116843795370194, + "learning_rate": 0.0004436751215672906, + "loss": 2.5691, + "step": 2398 + }, + { + "epoch": 0.5501031873423526, + "grad_norm": 0.40923994317152734, + "learning_rate": 0.000443306154197359, + "loss": 2.5365, + "step": 2399 + }, + { + "epoch": 0.5503324925475809, + "grad_norm": 0.43957089519507003, + "learning_rate": 0.00044293721809937316, + "loss": 2.4941, + "step": 2400 + }, + { + "epoch": 0.550561797752809, + "grad_norm": 0.4066845449588877, + "learning_rate": 0.00044256831347683646, + "loss": 2.6311, + "step": 2401 + }, + { + "epoch": 0.5507911029580371, + "grad_norm": 0.4036423105158156, + "learning_rate": 0.0004421994405332336, + "loss": 2.3499, + "step": 2402 + }, + { + "epoch": 0.5510204081632653, + "grad_norm": 0.4035396769866045, + "learning_rate": 0.0004418305994720328, + "loss": 2.3051, + "step": 2403 + }, + { + "epoch": 0.5512497133684935, + "grad_norm": 0.37860124098923237, + "learning_rate": 0.0004414617904966841, + "loss": 2.5374, + "step": 2404 + }, + { + "epoch": 0.5514790185737216, + "grad_norm": 0.44040192742015516, + "learning_rate": 0.0004410930138106203, + "loss": 2.6276, + "step": 2405 + }, + { + "epoch": 0.5517083237789497, + "grad_norm": 0.4342513844088, + "learning_rate": 0.0004407242696172559, + "loss": 2.3006, + "step": 2406 + }, + { + "epoch": 0.551937628984178, + "grad_norm": 0.40656619818532114, + "learning_rate": 0.000440355558119988, + "loss": 2.6102, + "step": 2407 + }, + { + "epoch": 0.5521669341894061, + "grad_norm": 0.40142413921517905, + "learning_rate": 0.0004399868795221951, + "loss": 2.3826, + "step": 2408 + }, + { + "epoch": 0.5523962393946342, + "grad_norm": 0.4167446712954756, + "learning_rate": 0.00043961823402723814, + "loss": 2.4933, + "step": 2409 + }, + { + "epoch": 0.5526255445998625, + "grad_norm": 0.36550050641701143, + "learning_rate": 0.00043924962183845953, + "loss": 2.4782, + "step": 2410 + }, + { + "epoch": 0.5528548498050906, + "grad_norm": 0.3982446247308319, + "learning_rate": 0.0004388810431591829, + "loss": 2.5043, + "step": 2411 + }, + { + "epoch": 0.5530841550103187, + "grad_norm": 0.36804553261212286, + "learning_rate": 0.00043851249819271403, + "loss": 2.246, + "step": 2412 + }, + { + "epoch": 0.553313460215547, + "grad_norm": 0.4046077710770028, + "learning_rate": 0.0004381439871423398, + "loss": 2.4585, + "step": 2413 + }, + { + "epoch": 0.5535427654207751, + "grad_norm": 0.46547342115420914, + "learning_rate": 0.0004377755102113286, + "loss": 2.4027, + "step": 2414 + }, + { + "epoch": 0.5537720706260032, + "grad_norm": 0.38248194386287465, + "learning_rate": 0.00043740706760292966, + "loss": 2.5485, + "step": 2415 + }, + { + "epoch": 0.5540013758312313, + "grad_norm": 0.4399849184898659, + "learning_rate": 0.00043703865952037353, + "loss": 2.5423, + "step": 2416 + }, + { + "epoch": 0.5542306810364596, + "grad_norm": 0.43769856514989053, + "learning_rate": 0.0004366702861668716, + "loss": 2.4635, + "step": 2417 + }, + { + "epoch": 0.5544599862416877, + "grad_norm": 0.4086658042648592, + "learning_rate": 0.00043630194774561625, + "loss": 2.5131, + "step": 2418 + }, + { + "epoch": 0.5546892914469158, + "grad_norm": 0.41810704734409304, + "learning_rate": 0.00043593364445978036, + "loss": 2.509, + "step": 2419 + }, + { + "epoch": 0.554918596652144, + "grad_norm": 0.38829063380618056, + "learning_rate": 0.0004355653765125178, + "loss": 2.3279, + "step": 2420 + }, + { + "epoch": 0.5551479018573722, + "grad_norm": 0.43024146778079997, + "learning_rate": 0.0004351971441069622, + "loss": 2.5731, + "step": 2421 + }, + { + "epoch": 0.5553772070626003, + "grad_norm": 0.38346952997427053, + "learning_rate": 0.00043482894744622874, + "loss": 2.4031, + "step": 2422 + }, + { + "epoch": 0.5556065122678285, + "grad_norm": 0.4130312145686704, + "learning_rate": 0.0004344607867334116, + "loss": 2.5308, + "step": 2423 + }, + { + "epoch": 0.5558358174730567, + "grad_norm": 0.4204079675333204, + "learning_rate": 0.0004340926621715864, + "loss": 2.4809, + "step": 2424 + }, + { + "epoch": 0.5560651226782848, + "grad_norm": 0.4060375286455885, + "learning_rate": 0.00043372457396380766, + "loss": 2.392, + "step": 2425 + }, + { + "epoch": 0.5562944278835129, + "grad_norm": 0.42050991842737495, + "learning_rate": 0.00043335652231311076, + "loss": 2.3646, + "step": 2426 + }, + { + "epoch": 0.5565237330887411, + "grad_norm": 0.42851035204196175, + "learning_rate": 0.00043298850742251013, + "loss": 2.3005, + "step": 2427 + }, + { + "epoch": 0.5567530382939693, + "grad_norm": 0.4236768548642689, + "learning_rate": 0.0004326205294950006, + "loss": 2.4846, + "step": 2428 + }, + { + "epoch": 0.5569823434991974, + "grad_norm": 0.42065507714274175, + "learning_rate": 0.0004322525887335563, + "loss": 2.4852, + "step": 2429 + }, + { + "epoch": 0.5572116487044256, + "grad_norm": 0.41403857200160105, + "learning_rate": 0.0004318846853411309, + "loss": 2.505, + "step": 2430 + }, + { + "epoch": 0.5574409539096538, + "grad_norm": 0.3939564570719575, + "learning_rate": 0.00043151681952065734, + "loss": 2.4159, + "step": 2431 + }, + { + "epoch": 0.5576702591148819, + "grad_norm": 0.4036211754245581, + "learning_rate": 0.0004311489914750482, + "loss": 2.5294, + "step": 2432 + }, + { + "epoch": 0.5578995643201101, + "grad_norm": 0.42116576760935787, + "learning_rate": 0.00043078120140719456, + "loss": 2.4952, + "step": 2433 + }, + { + "epoch": 0.5581288695253382, + "grad_norm": 0.4126156315651797, + "learning_rate": 0.0004304134495199674, + "loss": 2.6461, + "step": 2434 + }, + { + "epoch": 0.5583581747305664, + "grad_norm": 0.3770587262052428, + "learning_rate": 0.0004300457360162158, + "loss": 2.5135, + "step": 2435 + }, + { + "epoch": 0.5585874799357945, + "grad_norm": 0.39464334805197526, + "learning_rate": 0.00042967806109876853, + "loss": 2.4407, + "step": 2436 + }, + { + "epoch": 0.5588167851410227, + "grad_norm": 0.39110462919693423, + "learning_rate": 0.0004293104249704319, + "loss": 2.4902, + "step": 2437 + }, + { + "epoch": 0.5590460903462509, + "grad_norm": 0.40229874979150926, + "learning_rate": 0.00042894282783399246, + "loss": 2.5331, + "step": 2438 + }, + { + "epoch": 0.559275395551479, + "grad_norm": 0.3879603473280568, + "learning_rate": 0.00042857526989221355, + "loss": 2.466, + "step": 2439 + }, + { + "epoch": 0.5595047007567072, + "grad_norm": 0.36570570009248515, + "learning_rate": 0.00042820775134783817, + "loss": 2.4074, + "step": 2440 + }, + { + "epoch": 0.5597340059619353, + "grad_norm": 0.3936927713805096, + "learning_rate": 0.00042784027240358674, + "loss": 2.3369, + "step": 2441 + }, + { + "epoch": 0.5599633111671635, + "grad_norm": 0.4105152437948424, + "learning_rate": 0.00042747283326215847, + "loss": 2.5623, + "step": 2442 + }, + { + "epoch": 0.5601926163723917, + "grad_norm": 0.40136686467117944, + "learning_rate": 0.0004271054341262301, + "loss": 2.488, + "step": 2443 + }, + { + "epoch": 0.5604219215776198, + "grad_norm": 0.35558725639165345, + "learning_rate": 0.00042673807519845675, + "loss": 2.2854, + "step": 2444 + }, + { + "epoch": 0.560651226782848, + "grad_norm": 0.39798738675335593, + "learning_rate": 0.000426370756681471, + "loss": 2.5898, + "step": 2445 + }, + { + "epoch": 0.5608805319880761, + "grad_norm": 0.46494696267557284, + "learning_rate": 0.0004260034787778833, + "loss": 2.5141, + "step": 2446 + }, + { + "epoch": 0.5611098371933043, + "grad_norm": 0.41170248034403517, + "learning_rate": 0.0004256362416902817, + "loss": 2.4174, + "step": 2447 + }, + { + "epoch": 0.5613391423985324, + "grad_norm": 0.3924872581634847, + "learning_rate": 0.00042526904562123193, + "loss": 2.4121, + "step": 2448 + }, + { + "epoch": 0.5615684476037606, + "grad_norm": 0.4193718467050702, + "learning_rate": 0.00042490189077327637, + "loss": 2.3709, + "step": 2449 + }, + { + "epoch": 0.5617977528089888, + "grad_norm": 0.4137912260458225, + "learning_rate": 0.0004245347773489359, + "loss": 2.4508, + "step": 2450 + }, + { + "epoch": 0.5620270580142169, + "grad_norm": 0.43308425710767373, + "learning_rate": 0.00042416770555070703, + "loss": 2.4841, + "step": 2451 + }, + { + "epoch": 0.562256363219445, + "grad_norm": 0.397750759621236, + "learning_rate": 0.0004238006755810648, + "loss": 2.3965, + "step": 2452 + }, + { + "epoch": 0.5624856684246733, + "grad_norm": 0.3954979158118363, + "learning_rate": 0.00042343368764246, + "loss": 2.4778, + "step": 2453 + }, + { + "epoch": 0.5627149736299014, + "grad_norm": 0.4053634658723481, + "learning_rate": 0.000423066741937321, + "loss": 2.5763, + "step": 2454 + }, + { + "epoch": 0.5629442788351295, + "grad_norm": 0.39602895975488145, + "learning_rate": 0.0004226998386680524, + "loss": 2.4949, + "step": 2455 + }, + { + "epoch": 0.5631735840403577, + "grad_norm": 0.4338685325491798, + "learning_rate": 0.00042233297803703583, + "loss": 2.4818, + "step": 2456 + }, + { + "epoch": 0.5634028892455859, + "grad_norm": 0.854105723228055, + "learning_rate": 0.000421966160246629, + "loss": 2.5024, + "step": 2457 + }, + { + "epoch": 0.563632194450814, + "grad_norm": 0.385841755200706, + "learning_rate": 0.0004215993854991662, + "loss": 2.5091, + "step": 2458 + }, + { + "epoch": 0.5638614996560422, + "grad_norm": 0.40809903262890873, + "learning_rate": 0.00042123265399695783, + "loss": 2.4034, + "step": 2459 + }, + { + "epoch": 0.5640908048612704, + "grad_norm": 0.38485372271196905, + "learning_rate": 0.00042086596594229065, + "loss": 2.4518, + "step": 2460 + }, + { + "epoch": 0.5643201100664985, + "grad_norm": 0.42291813022500846, + "learning_rate": 0.0004204993215374273, + "loss": 2.3462, + "step": 2461 + }, + { + "epoch": 0.5645494152717266, + "grad_norm": 0.46117495777497675, + "learning_rate": 0.0004201327209846065, + "loss": 2.4626, + "step": 2462 + }, + { + "epoch": 0.5647787204769549, + "grad_norm": 0.4535580028505749, + "learning_rate": 0.00041976616448604226, + "loss": 2.4623, + "step": 2463 + }, + { + "epoch": 0.565008025682183, + "grad_norm": 0.38098062796691873, + "learning_rate": 0.0004193996522439253, + "loss": 2.626, + "step": 2464 + }, + { + "epoch": 0.5652373308874111, + "grad_norm": 0.4083650379551673, + "learning_rate": 0.00041903318446042076, + "loss": 2.6054, + "step": 2465 + }, + { + "epoch": 0.5654666360926393, + "grad_norm": 0.3985598823078232, + "learning_rate": 0.0004186667613376702, + "loss": 2.4488, + "step": 2466 + }, + { + "epoch": 0.5656959412978675, + "grad_norm": 0.4484682075878348, + "learning_rate": 0.00041830038307778984, + "loss": 2.5723, + "step": 2467 + }, + { + "epoch": 0.5659252465030956, + "grad_norm": 0.38302777401563765, + "learning_rate": 0.0004179340498828718, + "loss": 2.491, + "step": 2468 + }, + { + "epoch": 0.5661545517083237, + "grad_norm": 0.5862669368355644, + "learning_rate": 0.0004175677619549828, + "loss": 2.4304, + "step": 2469 + }, + { + "epoch": 0.566383856913552, + "grad_norm": 0.40785501472799846, + "learning_rate": 0.00041720151949616496, + "loss": 2.4768, + "step": 2470 + }, + { + "epoch": 0.5666131621187801, + "grad_norm": 0.41862331726050045, + "learning_rate": 0.000416835322708435, + "loss": 2.4495, + "step": 2471 + }, + { + "epoch": 0.5668424673240082, + "grad_norm": 0.4158138036183523, + "learning_rate": 0.0004164691717937848, + "loss": 2.5665, + "step": 2472 + }, + { + "epoch": 0.5670717725292365, + "grad_norm": 0.42581814379272165, + "learning_rate": 0.00041610306695418056, + "loss": 2.5735, + "step": 2473 + }, + { + "epoch": 0.5673010777344646, + "grad_norm": 0.41201672818678803, + "learning_rate": 0.0004157370083915636, + "loss": 2.4348, + "step": 2474 + }, + { + "epoch": 0.5675303829396927, + "grad_norm": 0.4006040519082231, + "learning_rate": 0.0004153709963078488, + "loss": 2.4764, + "step": 2475 + }, + { + "epoch": 0.5677596881449208, + "grad_norm": 0.41484932913121975, + "learning_rate": 0.0004150050309049267, + "loss": 2.4952, + "step": 2476 + }, + { + "epoch": 0.5679889933501491, + "grad_norm": 0.4021980441935514, + "learning_rate": 0.0004146391123846606, + "loss": 2.306, + "step": 2477 + }, + { + "epoch": 0.5682182985553772, + "grad_norm": 0.3900954251236672, + "learning_rate": 0.00041427324094888947, + "loss": 2.4607, + "step": 2478 + }, + { + "epoch": 0.5684476037606053, + "grad_norm": 0.3887635693020848, + "learning_rate": 0.0004139074167994249, + "loss": 2.702, + "step": 2479 + }, + { + "epoch": 0.5686769089658336, + "grad_norm": 0.39937563742361903, + "learning_rate": 0.0004135416401380535, + "loss": 2.4576, + "step": 2480 + }, + { + "epoch": 0.5689062141710617, + "grad_norm": 0.3732711218121983, + "learning_rate": 0.00041317591116653486, + "loss": 2.506, + "step": 2481 + }, + { + "epoch": 0.5691355193762898, + "grad_norm": 0.3941353453049478, + "learning_rate": 0.000412810230086603, + "loss": 2.4496, + "step": 2482 + }, + { + "epoch": 0.569364824581518, + "grad_norm": 1.936027337410314, + "learning_rate": 0.0004124445970999648, + "loss": 2.3946, + "step": 2483 + }, + { + "epoch": 0.5695941297867462, + "grad_norm": 0.4538718932569087, + "learning_rate": 0.00041207901240830126, + "loss": 2.4731, + "step": 2484 + }, + { + "epoch": 0.5698234349919743, + "grad_norm": 0.40860449676474125, + "learning_rate": 0.00041171347621326627, + "loss": 2.5569, + "step": 2485 + }, + { + "epoch": 0.5700527401972024, + "grad_norm": 0.42036349898550407, + "learning_rate": 0.0004113479887164873, + "loss": 2.3486, + "step": 2486 + }, + { + "epoch": 0.5702820454024307, + "grad_norm": 0.4381801331540965, + "learning_rate": 0.00041098255011956465, + "loss": 2.7221, + "step": 2487 + }, + { + "epoch": 0.5705113506076588, + "grad_norm": 0.4071846334360373, + "learning_rate": 0.00041061716062407204, + "loss": 2.5355, + "step": 2488 + }, + { + "epoch": 0.5707406558128869, + "grad_norm": 0.44352953224426006, + "learning_rate": 0.00041025182043155547, + "loss": 2.4868, + "step": 2489 + }, + { + "epoch": 0.5709699610181151, + "grad_norm": 0.41929648840943984, + "learning_rate": 0.00040988652974353467, + "loss": 2.391, + "step": 2490 + }, + { + "epoch": 0.5711992662233433, + "grad_norm": 0.46390651060425564, + "learning_rate": 0.000409521288761501, + "loss": 2.2916, + "step": 2491 + }, + { + "epoch": 0.5714285714285714, + "grad_norm": 0.4271518737404197, + "learning_rate": 0.0004091560976869194, + "loss": 2.4749, + "step": 2492 + }, + { + "epoch": 0.5716578766337996, + "grad_norm": 0.5812939891110038, + "learning_rate": 0.00040879095672122646, + "loss": 2.3859, + "step": 2493 + }, + { + "epoch": 0.5718871818390278, + "grad_norm": 0.5039646766183624, + "learning_rate": 0.00040842586606583176, + "loss": 2.6786, + "step": 2494 + }, + { + "epoch": 0.5721164870442559, + "grad_norm": 0.5487742362853093, + "learning_rate": 0.0004080608259221167, + "loss": 2.441, + "step": 2495 + }, + { + "epoch": 0.572345792249484, + "grad_norm": 0.7064573814952789, + "learning_rate": 0.00040769583649143517, + "loss": 2.614, + "step": 2496 + }, + { + "epoch": 0.5725750974547122, + "grad_norm": 0.4402996245633843, + "learning_rate": 0.0004073308979751126, + "loss": 2.5324, + "step": 2497 + }, + { + "epoch": 0.5728044026599404, + "grad_norm": 0.5510063363211549, + "learning_rate": 0.00040696601057444695, + "loss": 2.4021, + "step": 2498 + }, + { + "epoch": 0.5730337078651685, + "grad_norm": 0.44447447892286124, + "learning_rate": 0.0004066011744907074, + "loss": 2.5452, + "step": 2499 + }, + { + "epoch": 0.5732630130703967, + "grad_norm": 0.4201934817072395, + "learning_rate": 0.0004062363899251353, + "loss": 2.4471, + "step": 2500 + }, + { + "epoch": 0.5734923182756249, + "grad_norm": 0.4699607593097993, + "learning_rate": 0.00040587165707894326, + "loss": 2.4335, + "step": 2501 + }, + { + "epoch": 0.573721623480853, + "grad_norm": 0.4201189192489125, + "learning_rate": 0.00040550697615331565, + "loss": 2.5844, + "step": 2502 + }, + { + "epoch": 0.5739509286860812, + "grad_norm": 0.4140418689280386, + "learning_rate": 0.0004051423473494076, + "loss": 2.3897, + "step": 2503 + }, + { + "epoch": 0.5741802338913093, + "grad_norm": 0.4146514457729819, + "learning_rate": 0.0004047777708683466, + "loss": 2.4141, + "step": 2504 + }, + { + "epoch": 0.5744095390965375, + "grad_norm": 0.48129718790534975, + "learning_rate": 0.0004044132469112299, + "loss": 2.4482, + "step": 2505 + }, + { + "epoch": 0.5746388443017657, + "grad_norm": 0.42929092904581684, + "learning_rate": 0.00040404877567912695, + "loss": 2.3399, + "step": 2506 + }, + { + "epoch": 0.5748681495069938, + "grad_norm": 0.4223438330424067, + "learning_rate": 0.00040368435737307733, + "loss": 2.448, + "step": 2507 + }, + { + "epoch": 0.575097454712222, + "grad_norm": 0.48419444580084425, + "learning_rate": 0.00040331999219409204, + "loss": 2.3442, + "step": 2508 + }, + { + "epoch": 0.5753267599174501, + "grad_norm": 0.4419151923932312, + "learning_rate": 0.00040295568034315224, + "loss": 2.4667, + "step": 2509 + }, + { + "epoch": 0.5755560651226783, + "grad_norm": 0.4354045216676486, + "learning_rate": 0.00040259142202121, + "loss": 2.5867, + "step": 2510 + }, + { + "epoch": 0.5757853703279064, + "grad_norm": 0.4496883980132637, + "learning_rate": 0.0004022272174291878, + "loss": 2.4213, + "step": 2511 + }, + { + "epoch": 0.5760146755331346, + "grad_norm": 1.3648610136709516, + "learning_rate": 0.00040186306676797843, + "loss": 2.4964, + "step": 2512 + }, + { + "epoch": 0.5762439807383628, + "grad_norm": 0.4537313753617338, + "learning_rate": 0.0004014989702384449, + "loss": 2.5098, + "step": 2513 + }, + { + "epoch": 0.5764732859435909, + "grad_norm": 0.44472314545427605, + "learning_rate": 0.00040113492804142066, + "loss": 2.4733, + "step": 2514 + }, + { + "epoch": 0.576702591148819, + "grad_norm": 0.4716417015417961, + "learning_rate": 0.00040077094037770843, + "loss": 2.3469, + "step": 2515 + }, + { + "epoch": 0.5769318963540473, + "grad_norm": 0.5465571504480445, + "learning_rate": 0.00040040700744808206, + "loss": 2.3901, + "step": 2516 + }, + { + "epoch": 0.5771612015592754, + "grad_norm": 0.4362579178648901, + "learning_rate": 0.0004000431294532838, + "loss": 2.5332, + "step": 2517 + }, + { + "epoch": 0.5773905067645035, + "grad_norm": 0.44315469085723913, + "learning_rate": 0.0003996793065940271, + "loss": 2.507, + "step": 2518 + }, + { + "epoch": 0.5776198119697317, + "grad_norm": 0.709717523852723, + "learning_rate": 0.0003993155390709935, + "loss": 2.4726, + "step": 2519 + }, + { + "epoch": 0.5778491171749599, + "grad_norm": 0.4295985642212992, + "learning_rate": 0.0003989518270848351, + "loss": 2.3081, + "step": 2520 + }, + { + "epoch": 0.578078422380188, + "grad_norm": 0.5505988741025957, + "learning_rate": 0.0003985881708361729, + "loss": 2.5474, + "step": 2521 + }, + { + "epoch": 0.5783077275854162, + "grad_norm": 0.4176371572597361, + "learning_rate": 0.0003982245705255974, + "loss": 2.4873, + "step": 2522 + }, + { + "epoch": 0.5785370327906444, + "grad_norm": 0.41640051027501934, + "learning_rate": 0.00039786102635366784, + "loss": 2.3325, + "step": 2523 + }, + { + "epoch": 0.5787663379958725, + "grad_norm": 0.42039631458469, + "learning_rate": 0.00039749753852091305, + "loss": 2.4592, + "step": 2524 + }, + { + "epoch": 0.5789956432011006, + "grad_norm": 0.4521090737274071, + "learning_rate": 0.0003971341072278302, + "loss": 2.3261, + "step": 2525 + }, + { + "epoch": 0.5792249484063289, + "grad_norm": 0.43392948229793915, + "learning_rate": 0.0003967707326748857, + "loss": 2.4287, + "step": 2526 + }, + { + "epoch": 0.579454253611557, + "grad_norm": 0.41478412334046216, + "learning_rate": 0.00039640741506251457, + "loss": 2.5412, + "step": 2527 + }, + { + "epoch": 0.5796835588167851, + "grad_norm": 0.40574943514285056, + "learning_rate": 0.0003960441545911204, + "loss": 2.3461, + "step": 2528 + }, + { + "epoch": 0.5799128640220133, + "grad_norm": 0.38427147773923975, + "learning_rate": 0.00039568095146107495, + "loss": 2.5065, + "step": 2529 + }, + { + "epoch": 0.5801421692272415, + "grad_norm": 0.41311171155440324, + "learning_rate": 0.0003953178058727191, + "loss": 2.4712, + "step": 2530 + }, + { + "epoch": 0.5803714744324696, + "grad_norm": 0.42624558760442915, + "learning_rate": 0.00039495471802636096, + "loss": 2.4832, + "step": 2531 + }, + { + "epoch": 0.5806007796376977, + "grad_norm": 0.3883874368065423, + "learning_rate": 0.000394591688122278, + "loss": 2.4257, + "step": 2532 + }, + { + "epoch": 0.580830084842926, + "grad_norm": 0.43774473513800816, + "learning_rate": 0.0003942287163607145, + "loss": 2.4254, + "step": 2533 + }, + { + "epoch": 0.5810593900481541, + "grad_norm": 0.433181395383144, + "learning_rate": 0.0003938658029418837, + "loss": 2.6078, + "step": 2534 + }, + { + "epoch": 0.5812886952533822, + "grad_norm": 0.3875788155561691, + "learning_rate": 0.0003935029480659658, + "loss": 2.3513, + "step": 2535 + }, + { + "epoch": 0.5815180004586105, + "grad_norm": 0.7324694193020477, + "learning_rate": 0.00039314015193310953, + "loss": 2.5839, + "step": 2536 + }, + { + "epoch": 0.5817473056638386, + "grad_norm": 0.49690597446003215, + "learning_rate": 0.00039277741474343054, + "loss": 2.4439, + "step": 2537 + }, + { + "epoch": 0.5819766108690667, + "grad_norm": 0.39444301821867306, + "learning_rate": 0.00039241473669701237, + "loss": 2.4673, + "step": 2538 + }, + { + "epoch": 0.5822059160742948, + "grad_norm": 0.402858520548484, + "learning_rate": 0.0003920521179939057, + "loss": 2.4983, + "step": 2539 + }, + { + "epoch": 0.5824352212795231, + "grad_norm": 0.42446051599220036, + "learning_rate": 0.00039168955883412874, + "loss": 2.5285, + "step": 2540 + }, + { + "epoch": 0.5826645264847512, + "grad_norm": 0.4317934630882923, + "learning_rate": 0.00039132705941766644, + "loss": 2.5634, + "step": 2541 + }, + { + "epoch": 0.5828938316899793, + "grad_norm": 0.4274389052553112, + "learning_rate": 0.00039096461994447146, + "loss": 2.4587, + "step": 2542 + }, + { + "epoch": 0.5831231368952076, + "grad_norm": 0.39611739862832024, + "learning_rate": 0.0003906022406144624, + "loss": 2.4906, + "step": 2543 + }, + { + "epoch": 0.5833524421004357, + "grad_norm": 0.44403154626684677, + "learning_rate": 0.000390239921627526, + "loss": 2.4988, + "step": 2544 + }, + { + "epoch": 0.5835817473056638, + "grad_norm": 0.4032488126144617, + "learning_rate": 0.0003898776631835143, + "loss": 2.4744, + "step": 2545 + }, + { + "epoch": 0.583811052510892, + "grad_norm": 0.3926047218304753, + "learning_rate": 0.0003895154654822471, + "loss": 2.4703, + "step": 2546 + }, + { + "epoch": 0.5840403577161202, + "grad_norm": 0.3956293936774464, + "learning_rate": 0.00038915332872350994, + "loss": 2.4386, + "step": 2547 + }, + { + "epoch": 0.5842696629213483, + "grad_norm": 0.4222022098655776, + "learning_rate": 0.00038879125310705526, + "loss": 2.4902, + "step": 2548 + }, + { + "epoch": 0.5844989681265764, + "grad_norm": 0.4087958377292356, + "learning_rate": 0.00038842923883260135, + "loss": 2.3835, + "step": 2549 + }, + { + "epoch": 0.5847282733318047, + "grad_norm": 0.47343188428170224, + "learning_rate": 0.0003880672860998331, + "loss": 2.5372, + "step": 2550 + }, + { + "epoch": 0.5849575785370328, + "grad_norm": 0.4855466591742376, + "learning_rate": 0.00038770539510840093, + "loss": 2.4434, + "step": 2551 + }, + { + "epoch": 0.5851868837422609, + "grad_norm": 0.47799309527327954, + "learning_rate": 0.0003873435660579217, + "loss": 2.5536, + "step": 2552 + }, + { + "epoch": 0.5854161889474891, + "grad_norm": 0.38588784817036104, + "learning_rate": 0.00038698179914797783, + "loss": 2.4012, + "step": 2553 + }, + { + "epoch": 0.5856454941527173, + "grad_norm": 0.4114268190680933, + "learning_rate": 0.00038662009457811754, + "loss": 2.3883, + "step": 2554 + }, + { + "epoch": 0.5858747993579454, + "grad_norm": 0.41635697612911704, + "learning_rate": 0.0003862584525478545, + "loss": 2.5829, + "step": 2555 + }, + { + "epoch": 0.5861041045631736, + "grad_norm": 0.5052929717905738, + "learning_rate": 0.0003858968732566685, + "loss": 2.6735, + "step": 2556 + }, + { + "epoch": 0.5863334097684018, + "grad_norm": 0.4076121071686265, + "learning_rate": 0.00038553535690400353, + "loss": 2.5387, + "step": 2557 + }, + { + "epoch": 0.5865627149736299, + "grad_norm": 0.4152692032450925, + "learning_rate": 0.00038517390368927044, + "loss": 2.4698, + "step": 2558 + }, + { + "epoch": 0.586792020178858, + "grad_norm": 0.390714262344046, + "learning_rate": 0.00038481251381184355, + "loss": 2.5626, + "step": 2559 + }, + { + "epoch": 0.5870213253840862, + "grad_norm": 0.4209023167341848, + "learning_rate": 0.0003844511874710638, + "loss": 2.5533, + "step": 2560 + }, + { + "epoch": 0.5872506305893144, + "grad_norm": 0.4239132719128045, + "learning_rate": 0.00038408992486623584, + "loss": 2.3425, + "step": 2561 + }, + { + "epoch": 0.5874799357945425, + "grad_norm": 0.40743770182894595, + "learning_rate": 0.00038372872619663013, + "loss": 2.4397, + "step": 2562 + }, + { + "epoch": 0.5877092409997707, + "grad_norm": 0.4127903262367354, + "learning_rate": 0.00038336759166148117, + "loss": 2.343, + "step": 2563 + }, + { + "epoch": 0.5879385462049989, + "grad_norm": 0.4287291579871143, + "learning_rate": 0.00038300652145998855, + "loss": 2.45, + "step": 2564 + }, + { + "epoch": 0.588167851410227, + "grad_norm": 0.39288053024744457, + "learning_rate": 0.0003826455157913159, + "loss": 2.4015, + "step": 2565 + }, + { + "epoch": 0.5883971566154552, + "grad_norm": 0.45066847142426314, + "learning_rate": 0.00038228457485459187, + "loss": 2.6173, + "step": 2566 + }, + { + "epoch": 0.5886264618206833, + "grad_norm": 0.45149147885368457, + "learning_rate": 0.00038192369884890886, + "loss": 2.5423, + "step": 2567 + }, + { + "epoch": 0.5888557670259115, + "grad_norm": 0.3999259326736763, + "learning_rate": 0.00038156288797332384, + "loss": 2.3187, + "step": 2568 + }, + { + "epoch": 0.5890850722311396, + "grad_norm": 0.4593607917945761, + "learning_rate": 0.00038120214242685723, + "loss": 2.4121, + "step": 2569 + }, + { + "epoch": 0.5893143774363678, + "grad_norm": 0.40013703792556876, + "learning_rate": 0.00038084146240849456, + "loss": 2.3961, + "step": 2570 + }, + { + "epoch": 0.589543682641596, + "grad_norm": 0.4722661332479663, + "learning_rate": 0.00038048084811718373, + "loss": 2.5672, + "step": 2571 + }, + { + "epoch": 0.5897729878468241, + "grad_norm": 0.4680658626034723, + "learning_rate": 0.00038012029975183795, + "loss": 2.4453, + "step": 2572 + }, + { + "epoch": 0.5900022930520523, + "grad_norm": 0.44360653122441956, + "learning_rate": 0.0003797598175113327, + "loss": 2.4437, + "step": 2573 + }, + { + "epoch": 0.5902315982572804, + "grad_norm": 0.44367459723280495, + "learning_rate": 0.00037939940159450775, + "loss": 2.4723, + "step": 2574 + }, + { + "epoch": 0.5904609034625086, + "grad_norm": 0.418467805117952, + "learning_rate": 0.0003790390522001662, + "loss": 2.4206, + "step": 2575 + }, + { + "epoch": 0.5906902086677368, + "grad_norm": 0.4137324785833235, + "learning_rate": 0.00037867876952707425, + "loss": 2.7063, + "step": 2576 + }, + { + "epoch": 0.5909195138729649, + "grad_norm": 0.4162920870086698, + "learning_rate": 0.0003783185537739615, + "loss": 2.4492, + "step": 2577 + }, + { + "epoch": 0.591148819078193, + "grad_norm": 0.3825500763486464, + "learning_rate": 0.0003779584051395205, + "loss": 2.3327, + "step": 2578 + }, + { + "epoch": 0.5913781242834212, + "grad_norm": 0.41295234596404473, + "learning_rate": 0.00037759832382240697, + "loss": 2.2931, + "step": 2579 + }, + { + "epoch": 0.5916074294886494, + "grad_norm": 0.42456140107452783, + "learning_rate": 0.000377238310021239, + "loss": 2.4275, + "step": 2580 + }, + { + "epoch": 0.5918367346938775, + "grad_norm": 0.3786528263990366, + "learning_rate": 0.00037687836393459826, + "loss": 2.44, + "step": 2581 + }, + { + "epoch": 0.5920660398991057, + "grad_norm": 0.41277026608402806, + "learning_rate": 0.00037651848576102814, + "loss": 2.5962, + "step": 2582 + }, + { + "epoch": 0.5922953451043339, + "grad_norm": 0.3862419808818825, + "learning_rate": 0.0003761586756990354, + "loss": 2.3593, + "step": 2583 + }, + { + "epoch": 0.592524650309562, + "grad_norm": 0.4227416379686459, + "learning_rate": 0.00037579893394708866, + "loss": 2.5016, + "step": 2584 + }, + { + "epoch": 0.5927539555147902, + "grad_norm": 0.4183189626300753, + "learning_rate": 0.0003754392607036191, + "loss": 2.4233, + "step": 2585 + }, + { + "epoch": 0.5929832607200184, + "grad_norm": 0.4958203184226247, + "learning_rate": 0.00037507965616702014, + "loss": 2.629, + "step": 2586 + }, + { + "epoch": 0.5932125659252465, + "grad_norm": 0.41454523979696123, + "learning_rate": 0.0003747201205356472, + "loss": 2.254, + "step": 2587 + }, + { + "epoch": 0.5934418711304746, + "grad_norm": 0.3967426654385581, + "learning_rate": 0.0003743606540078177, + "loss": 2.4082, + "step": 2588 + }, + { + "epoch": 0.5936711763357028, + "grad_norm": 0.4342871085842698, + "learning_rate": 0.0003740012567818111, + "loss": 2.5679, + "step": 2589 + }, + { + "epoch": 0.593900481540931, + "grad_norm": 0.46584777431172414, + "learning_rate": 0.00037364192905586823, + "loss": 2.3645, + "step": 2590 + }, + { + "epoch": 0.5941297867461591, + "grad_norm": 0.6036775692933483, + "learning_rate": 0.0003732826710281922, + "loss": 2.3913, + "step": 2591 + }, + { + "epoch": 0.5943590919513873, + "grad_norm": 0.4208275578752697, + "learning_rate": 0.000372923482896947, + "loss": 2.4231, + "step": 2592 + }, + { + "epoch": 0.5945883971566155, + "grad_norm": 0.41120416257735104, + "learning_rate": 0.0003725643648602588, + "loss": 2.4077, + "step": 2593 + }, + { + "epoch": 0.5948177023618436, + "grad_norm": 0.45331753796983654, + "learning_rate": 0.00037220531711621417, + "loss": 2.4127, + "step": 2594 + }, + { + "epoch": 0.5950470075670717, + "grad_norm": 0.4219250437403737, + "learning_rate": 0.0003718463398628621, + "loss": 2.5472, + "step": 2595 + }, + { + "epoch": 0.5952763127723, + "grad_norm": 0.4070181399289423, + "learning_rate": 0.00037148743329821143, + "loss": 2.4976, + "step": 2596 + }, + { + "epoch": 0.5955056179775281, + "grad_norm": 0.4088261584032599, + "learning_rate": 0.0003711285976202331, + "loss": 2.417, + "step": 2597 + }, + { + "epoch": 0.5957349231827562, + "grad_norm": 0.5914871579049782, + "learning_rate": 0.000370769833026858, + "loss": 2.4262, + "step": 2598 + }, + { + "epoch": 0.5959642283879845, + "grad_norm": 0.45727042772722115, + "learning_rate": 0.0003704111397159787, + "loss": 2.3884, + "step": 2599 + }, + { + "epoch": 0.5961935335932126, + "grad_norm": 0.41750593346449094, + "learning_rate": 0.0003700525178854478, + "loss": 2.4648, + "step": 2600 + }, + { + "epoch": 0.5964228387984407, + "grad_norm": 0.4091172737131766, + "learning_rate": 0.0003696939677330788, + "loss": 2.5066, + "step": 2601 + }, + { + "epoch": 0.5966521440036688, + "grad_norm": 0.45017496402306717, + "learning_rate": 0.00036933548945664545, + "loss": 2.5551, + "step": 2602 + }, + { + "epoch": 0.5968814492088971, + "grad_norm": 0.4061201432162814, + "learning_rate": 0.00036897708325388213, + "loss": 2.5156, + "step": 2603 + }, + { + "epoch": 0.5971107544141252, + "grad_norm": 0.4892369832838835, + "learning_rate": 0.00036861874932248316, + "loss": 2.4132, + "step": 2604 + }, + { + "epoch": 0.5973400596193533, + "grad_norm": 0.4230239875055197, + "learning_rate": 0.0003682604878601034, + "loss": 2.4947, + "step": 2605 + }, + { + "epoch": 0.5975693648245816, + "grad_norm": 0.38651272475693343, + "learning_rate": 0.000367902299064357, + "loss": 2.3546, + "step": 2606 + }, + { + "epoch": 0.5977986700298097, + "grad_norm": 0.5877198635849098, + "learning_rate": 0.000367544183132819, + "loss": 2.5365, + "step": 2607 + }, + { + "epoch": 0.5980279752350378, + "grad_norm": 0.39830864893488394, + "learning_rate": 0.00036718614026302334, + "loss": 2.5248, + "step": 2608 + }, + { + "epoch": 0.598257280440266, + "grad_norm": 0.4581886233792683, + "learning_rate": 0.0003668281706524645, + "loss": 2.4581, + "step": 2609 + }, + { + "epoch": 0.5984865856454942, + "grad_norm": 0.7439170947468554, + "learning_rate": 0.0003664702744985957, + "loss": 2.4058, + "step": 2610 + }, + { + "epoch": 0.5987158908507223, + "grad_norm": 0.3898784755376088, + "learning_rate": 0.0003661124519988304, + "loss": 2.332, + "step": 2611 + }, + { + "epoch": 0.5989451960559504, + "grad_norm": 0.3933296447716806, + "learning_rate": 0.00036575470335054077, + "loss": 2.5688, + "step": 2612 + }, + { + "epoch": 0.5991745012611787, + "grad_norm": 0.38481804922357904, + "learning_rate": 0.00036539702875105893, + "loss": 2.5966, + "step": 2613 + }, + { + "epoch": 0.5994038064664068, + "grad_norm": 0.4215528443758271, + "learning_rate": 0.00036503942839767546, + "loss": 2.5001, + "step": 2614 + }, + { + "epoch": 0.5996331116716349, + "grad_norm": 0.42262219428243125, + "learning_rate": 0.0003646819024876406, + "loss": 2.4412, + "step": 2615 + }, + { + "epoch": 0.5998624168768631, + "grad_norm": 0.3878713255712672, + "learning_rate": 0.00036432445121816306, + "loss": 2.4513, + "step": 2616 + }, + { + "epoch": 0.6000917220820913, + "grad_norm": 0.38518035164580705, + "learning_rate": 0.0003639670747864107, + "loss": 2.4103, + "step": 2617 + }, + { + "epoch": 0.6003210272873194, + "grad_norm": 0.4357090809875327, + "learning_rate": 0.00036360977338950986, + "loss": 2.5384, + "step": 2618 + }, + { + "epoch": 0.6005503324925476, + "grad_norm": 0.41123916962764206, + "learning_rate": 0.00036325254722454584, + "loss": 2.3532, + "step": 2619 + }, + { + "epoch": 0.6007796376977758, + "grad_norm": 0.4406475656344627, + "learning_rate": 0.0003628953964885616, + "loss": 2.5159, + "step": 2620 + }, + { + "epoch": 0.6010089429030039, + "grad_norm": 0.3945673050346948, + "learning_rate": 0.00036253832137855997, + "loss": 2.5137, + "step": 2621 + }, + { + "epoch": 0.601238248108232, + "grad_norm": 0.38846271281528677, + "learning_rate": 0.00036218132209150044, + "loss": 2.4687, + "step": 2622 + }, + { + "epoch": 0.6014675533134602, + "grad_norm": 0.4197119507879678, + "learning_rate": 0.00036182439882430183, + "loss": 2.4992, + "step": 2623 + }, + { + "epoch": 0.6016968585186884, + "grad_norm": 0.4019895351057925, + "learning_rate": 0.0003614675517738405, + "loss": 2.5576, + "step": 2624 + }, + { + "epoch": 0.6019261637239165, + "grad_norm": 0.41402671200949187, + "learning_rate": 0.00036111078113695096, + "loss": 2.439, + "step": 2625 + }, + { + "epoch": 0.6021554689291447, + "grad_norm": 0.44740264085739456, + "learning_rate": 0.00036075408711042534, + "loss": 2.5222, + "step": 2626 + }, + { + "epoch": 0.6023847741343729, + "grad_norm": 0.4047945118246008, + "learning_rate": 0.0003603974698910139, + "loss": 2.4356, + "step": 2627 + }, + { + "epoch": 0.602614079339601, + "grad_norm": 0.4518404571708384, + "learning_rate": 0.0003600409296754241, + "loss": 2.4955, + "step": 2628 + }, + { + "epoch": 0.6028433845448292, + "grad_norm": 0.41117267070991503, + "learning_rate": 0.0003596844666603214, + "loss": 2.4235, + "step": 2629 + }, + { + "epoch": 0.6030726897500573, + "grad_norm": 0.39755742408831835, + "learning_rate": 0.0003593280810423282, + "loss": 2.5042, + "step": 2630 + }, + { + "epoch": 0.6033019949552855, + "grad_norm": 0.41978244366053435, + "learning_rate": 0.0003589717730180245, + "loss": 2.3886, + "step": 2631 + }, + { + "epoch": 0.6035313001605136, + "grad_norm": 0.4444080461395485, + "learning_rate": 0.0003586155427839473, + "loss": 2.506, + "step": 2632 + }, + { + "epoch": 0.6037606053657418, + "grad_norm": 0.4079728566195103, + "learning_rate": 0.00035825939053659117, + "loss": 2.6012, + "step": 2633 + }, + { + "epoch": 0.60398991057097, + "grad_norm": 0.4329865265844062, + "learning_rate": 0.00035790331647240684, + "loss": 2.5183, + "step": 2634 + }, + { + "epoch": 0.6042192157761981, + "grad_norm": 0.4303042555154398, + "learning_rate": 0.00035754732078780273, + "loss": 2.5159, + "step": 2635 + }, + { + "epoch": 0.6044485209814263, + "grad_norm": 0.4010987483049038, + "learning_rate": 0.0003571914036791435, + "loss": 2.4839, + "step": 2636 + }, + { + "epoch": 0.6046778261866544, + "grad_norm": 0.4082902550747299, + "learning_rate": 0.00035683556534275076, + "loss": 2.265, + "step": 2637 + }, + { + "epoch": 0.6049071313918826, + "grad_norm": 0.41664640283978427, + "learning_rate": 0.0003564798059749023, + "loss": 2.5195, + "step": 2638 + }, + { + "epoch": 0.6051364365971108, + "grad_norm": 0.4007553065817275, + "learning_rate": 0.00035612412577183303, + "loss": 2.3858, + "step": 2639 + }, + { + "epoch": 0.6053657418023389, + "grad_norm": 0.39907293361214985, + "learning_rate": 0.0003557685249297333, + "loss": 2.4104, + "step": 2640 + }, + { + "epoch": 0.6055950470075671, + "grad_norm": 0.3912899996968911, + "learning_rate": 0.00035541300364475063, + "loss": 2.415, + "step": 2641 + }, + { + "epoch": 0.6058243522127952, + "grad_norm": 0.43243503302436237, + "learning_rate": 0.00035505756211298776, + "loss": 2.5077, + "step": 2642 + }, + { + "epoch": 0.6060536574180234, + "grad_norm": 0.40340463930962206, + "learning_rate": 0.0003547022005305043, + "loss": 2.3819, + "step": 2643 + }, + { + "epoch": 0.6062829626232515, + "grad_norm": 0.3707964556633149, + "learning_rate": 0.00035434691909331487, + "loss": 2.3423, + "step": 2644 + }, + { + "epoch": 0.6065122678284797, + "grad_norm": 0.41791219438854316, + "learning_rate": 0.0003539917179973907, + "loss": 2.5452, + "step": 2645 + }, + { + "epoch": 0.6067415730337079, + "grad_norm": 0.4276644904684411, + "learning_rate": 0.00035363659743865795, + "loss": 2.4413, + "step": 2646 + }, + { + "epoch": 0.606970878238936, + "grad_norm": 0.41572109759043613, + "learning_rate": 0.00035328155761299917, + "loss": 2.4476, + "step": 2647 + }, + { + "epoch": 0.6072001834441642, + "grad_norm": 0.404145169601064, + "learning_rate": 0.00035292659871625153, + "loss": 2.3139, + "step": 2648 + }, + { + "epoch": 0.6074294886493924, + "grad_norm": 0.39401501391657245, + "learning_rate": 0.0003525717209442085, + "loss": 2.4229, + "step": 2649 + }, + { + "epoch": 0.6076587938546205, + "grad_norm": 0.3903518283501721, + "learning_rate": 0.00035221692449261785, + "loss": 2.5557, + "step": 2650 + }, + { + "epoch": 0.6078880990598486, + "grad_norm": 0.38654857146378496, + "learning_rate": 0.00035186220955718306, + "loss": 2.4404, + "step": 2651 + }, + { + "epoch": 0.6081174042650768, + "grad_norm": 0.40267295425398, + "learning_rate": 0.0003515075763335625, + "loss": 2.4911, + "step": 2652 + }, + { + "epoch": 0.608346709470305, + "grad_norm": 0.3959440578890467, + "learning_rate": 0.0003511530250173696, + "loss": 2.5399, + "step": 2653 + }, + { + "epoch": 0.6085760146755331, + "grad_norm": 0.4132409291687141, + "learning_rate": 0.0003507985558041723, + "loss": 2.6272, + "step": 2654 + }, + { + "epoch": 0.6088053198807613, + "grad_norm": 0.4233766625390764, + "learning_rate": 0.00035044416888949364, + "loss": 2.4351, + "step": 2655 + }, + { + "epoch": 0.6090346250859895, + "grad_norm": 0.4023588845970447, + "learning_rate": 0.00035008986446881087, + "loss": 2.4761, + "step": 2656 + }, + { + "epoch": 0.6092639302912176, + "grad_norm": 0.3839933511487952, + "learning_rate": 0.0003497356427375562, + "loss": 2.4793, + "step": 2657 + }, + { + "epoch": 0.6094932354964457, + "grad_norm": 0.3808405158806429, + "learning_rate": 0.0003493815038911157, + "loss": 2.5144, + "step": 2658 + }, + { + "epoch": 0.609722540701674, + "grad_norm": 0.39250804382995425, + "learning_rate": 0.00034902744812483034, + "loss": 2.4525, + "step": 2659 + }, + { + "epoch": 0.6099518459069021, + "grad_norm": 0.38944984258017445, + "learning_rate": 0.0003486734756339943, + "loss": 2.5223, + "step": 2660 + }, + { + "epoch": 0.6101811511121302, + "grad_norm": 0.4235396656913754, + "learning_rate": 0.00034831958661385714, + "loss": 2.4821, + "step": 2661 + }, + { + "epoch": 0.6104104563173584, + "grad_norm": 0.4564659904594352, + "learning_rate": 0.00034796578125962117, + "loss": 2.6034, + "step": 2662 + }, + { + "epoch": 0.6106397615225866, + "grad_norm": 0.3913073768390961, + "learning_rate": 0.0003476120597664434, + "loss": 2.4177, + "step": 2663 + }, + { + "epoch": 0.6108690667278147, + "grad_norm": 0.3952011736862222, + "learning_rate": 0.0003472584223294339, + "loss": 2.438, + "step": 2664 + }, + { + "epoch": 0.6110983719330428, + "grad_norm": 0.41908100116694935, + "learning_rate": 0.00034690486914365704, + "loss": 2.6428, + "step": 2665 + }, + { + "epoch": 0.6113276771382711, + "grad_norm": 0.40602577431888065, + "learning_rate": 0.00034655140040413005, + "loss": 2.3634, + "step": 2666 + }, + { + "epoch": 0.6115569823434992, + "grad_norm": 0.43608734259265003, + "learning_rate": 0.00034619801630582435, + "loss": 2.6102, + "step": 2667 + }, + { + "epoch": 0.6117862875487273, + "grad_norm": 0.3910950196527191, + "learning_rate": 0.0003458447170436638, + "loss": 2.3409, + "step": 2668 + }, + { + "epoch": 0.6120155927539556, + "grad_norm": 0.3725104871167605, + "learning_rate": 0.00034549150281252633, + "loss": 2.519, + "step": 2669 + }, + { + "epoch": 0.6122448979591837, + "grad_norm": 0.42964183618554225, + "learning_rate": 0.0003451383738072422, + "loss": 2.2679, + "step": 2670 + }, + { + "epoch": 0.6124742031644118, + "grad_norm": 0.408122282797986, + "learning_rate": 0.0003447853302225952, + "loss": 2.3963, + "step": 2671 + }, + { + "epoch": 0.6127035083696399, + "grad_norm": 0.9539396037032627, + "learning_rate": 0.00034443237225332174, + "loss": 2.6068, + "step": 2672 + }, + { + "epoch": 0.6129328135748682, + "grad_norm": 0.43219053216182834, + "learning_rate": 0.00034407950009411126, + "loss": 2.3759, + "step": 2673 + }, + { + "epoch": 0.6131621187800963, + "grad_norm": 0.4372371184686368, + "learning_rate": 0.00034372671393960517, + "loss": 2.3655, + "step": 2674 + }, + { + "epoch": 0.6133914239853244, + "grad_norm": 0.4095337288277439, + "learning_rate": 0.00034337401398439873, + "loss": 2.3922, + "step": 2675 + }, + { + "epoch": 0.6136207291905527, + "grad_norm": 0.4337953958230529, + "learning_rate": 0.0003430214004230381, + "loss": 2.4988, + "step": 2676 + }, + { + "epoch": 0.6138500343957808, + "grad_norm": 0.4141592809531555, + "learning_rate": 0.00034266887345002305, + "loss": 2.3436, + "step": 2677 + }, + { + "epoch": 0.6140793396010089, + "grad_norm": 0.45374910178276434, + "learning_rate": 0.0003423164332598049, + "loss": 2.5063, + "step": 2678 + }, + { + "epoch": 0.6143086448062371, + "grad_norm": 0.4640475960876632, + "learning_rate": 0.0003419640800467874, + "loss": 2.5707, + "step": 2679 + }, + { + "epoch": 0.6145379500114653, + "grad_norm": 0.4022130457774907, + "learning_rate": 0.0003416118140053261, + "loss": 2.3428, + "step": 2680 + }, + { + "epoch": 0.6147672552166934, + "grad_norm": 0.41650125864509147, + "learning_rate": 0.0003412596353297288, + "loss": 2.4737, + "step": 2681 + }, + { + "epoch": 0.6149965604219215, + "grad_norm": 0.3929763405289858, + "learning_rate": 0.0003409075442142546, + "loss": 2.3643, + "step": 2682 + }, + { + "epoch": 0.6152258656271498, + "grad_norm": 0.3839478849350019, + "learning_rate": 0.00034055554085311493, + "loss": 2.4128, + "step": 2683 + }, + { + "epoch": 0.6154551708323779, + "grad_norm": 0.5139210457057161, + "learning_rate": 0.0003402036254404722, + "loss": 2.5594, + "step": 2684 + }, + { + "epoch": 0.615684476037606, + "grad_norm": 0.39647959948426964, + "learning_rate": 0.00033985179817044105, + "loss": 2.4347, + "step": 2685 + }, + { + "epoch": 0.6159137812428342, + "grad_norm": 0.40146864330088033, + "learning_rate": 0.0003395000592370864, + "loss": 2.4435, + "step": 2686 + }, + { + "epoch": 0.6161430864480624, + "grad_norm": 0.42985586729376524, + "learning_rate": 0.0003391484088344257, + "loss": 2.5959, + "step": 2687 + }, + { + "epoch": 0.6163723916532905, + "grad_norm": 0.45189441430394695, + "learning_rate": 0.0003387968471564263, + "loss": 2.4951, + "step": 2688 + }, + { + "epoch": 0.6166016968585187, + "grad_norm": 0.39425358232554775, + "learning_rate": 0.00033844537439700807, + "loss": 2.4161, + "step": 2689 + }, + { + "epoch": 0.6168310020637469, + "grad_norm": 0.38270760849742574, + "learning_rate": 0.00033809399075004036, + "loss": 2.4663, + "step": 2690 + }, + { + "epoch": 0.617060307268975, + "grad_norm": 0.4021811753963198, + "learning_rate": 0.00033774269640934445, + "loss": 2.4645, + "step": 2691 + }, + { + "epoch": 0.6172896124742032, + "grad_norm": 0.42779749654340876, + "learning_rate": 0.00033739149156869166, + "loss": 2.3893, + "step": 2692 + }, + { + "epoch": 0.6175189176794313, + "grad_norm": 0.44679440168224527, + "learning_rate": 0.0003370403764218045, + "loss": 2.4164, + "step": 2693 + }, + { + "epoch": 0.6177482228846595, + "grad_norm": 0.41837197986512886, + "learning_rate": 0.0003366893511623555, + "loss": 2.4163, + "step": 2694 + }, + { + "epoch": 0.6179775280898876, + "grad_norm": 0.3897204046688793, + "learning_rate": 0.000336338415983968, + "loss": 2.5066, + "step": 2695 + }, + { + "epoch": 0.6182068332951158, + "grad_norm": 0.4172760344717078, + "learning_rate": 0.00033598757108021545, + "loss": 2.4654, + "step": 2696 + }, + { + "epoch": 0.618436138500344, + "grad_norm": 0.42389631369192227, + "learning_rate": 0.00033563681664462155, + "loss": 2.4645, + "step": 2697 + }, + { + "epoch": 0.6186654437055721, + "grad_norm": 0.47722266213546105, + "learning_rate": 0.00033528615287066005, + "loss": 2.302, + "step": 2698 + }, + { + "epoch": 0.6188947489108003, + "grad_norm": 0.37275831608000815, + "learning_rate": 0.000334935579951755, + "loss": 2.4207, + "step": 2699 + }, + { + "epoch": 0.6191240541160284, + "grad_norm": 0.3965555130644273, + "learning_rate": 0.0003345850980812795, + "loss": 2.4598, + "step": 2700 + }, + { + "epoch": 0.6193533593212566, + "grad_norm": 0.4422383481362954, + "learning_rate": 0.0003342347074525578, + "loss": 2.4894, + "step": 2701 + }, + { + "epoch": 0.6195826645264848, + "grad_norm": 0.3804542119507746, + "learning_rate": 0.00033388440825886216, + "loss": 2.5459, + "step": 2702 + }, + { + "epoch": 0.6198119697317129, + "grad_norm": 0.3980149491589244, + "learning_rate": 0.0003335342006934161, + "loss": 2.5094, + "step": 2703 + }, + { + "epoch": 0.6200412749369411, + "grad_norm": 0.3789722125789966, + "learning_rate": 0.0003331840849493911, + "loss": 2.4572, + "step": 2704 + }, + { + "epoch": 0.6202705801421692, + "grad_norm": 0.4013224447393776, + "learning_rate": 0.00033283406121990914, + "loss": 2.3254, + "step": 2705 + }, + { + "epoch": 0.6204998853473974, + "grad_norm": 0.44983211181013366, + "learning_rate": 0.00033248412969804063, + "loss": 2.4408, + "step": 2706 + }, + { + "epoch": 0.6207291905526255, + "grad_norm": 0.40538317913071303, + "learning_rate": 0.0003321342905768057, + "loss": 2.4934, + "step": 2707 + }, + { + "epoch": 0.6209584957578537, + "grad_norm": 0.4129779991647362, + "learning_rate": 0.0003317845440491731, + "loss": 2.4108, + "step": 2708 + }, + { + "epoch": 0.6211878009630819, + "grad_norm": 0.41223386800044465, + "learning_rate": 0.00033143489030806086, + "loss": 2.4299, + "step": 2709 + }, + { + "epoch": 0.62141710616831, + "grad_norm": 0.44595966387912145, + "learning_rate": 0.0003310853295463353, + "loss": 2.5733, + "step": 2710 + }, + { + "epoch": 0.6216464113735382, + "grad_norm": 0.39293561919259906, + "learning_rate": 0.00033073586195681227, + "loss": 2.5831, + "step": 2711 + }, + { + "epoch": 0.6218757165787664, + "grad_norm": 0.4289314025615685, + "learning_rate": 0.0003303864877322553, + "loss": 2.4861, + "step": 2712 + }, + { + "epoch": 0.6221050217839945, + "grad_norm": 0.40397057823507687, + "learning_rate": 0.00033003720706537736, + "loss": 2.374, + "step": 2713 + }, + { + "epoch": 0.6223343269892226, + "grad_norm": 0.4372207141479431, + "learning_rate": 0.0003296880201488387, + "loss": 2.3822, + "step": 2714 + }, + { + "epoch": 0.6225636321944508, + "grad_norm": 0.44358662859411996, + "learning_rate": 0.0003293389271752492, + "loss": 2.4084, + "step": 2715 + }, + { + "epoch": 0.622792937399679, + "grad_norm": 0.3974373669738123, + "learning_rate": 0.0003289899283371657, + "loss": 2.3552, + "step": 2716 + }, + { + "epoch": 0.6230222426049071, + "grad_norm": 0.41253814885520246, + "learning_rate": 0.00032864102382709374, + "loss": 2.4407, + "step": 2717 + }, + { + "epoch": 0.6232515478101353, + "grad_norm": 0.4170650157786748, + "learning_rate": 0.0003282922138374868, + "loss": 2.3914, + "step": 2718 + }, + { + "epoch": 0.6234808530153635, + "grad_norm": 0.4181556035116046, + "learning_rate": 0.000327943498560746, + "loss": 2.4707, + "step": 2719 + }, + { + "epoch": 0.6237101582205916, + "grad_norm": 0.4053271986733598, + "learning_rate": 0.0003275948781892205, + "loss": 2.3566, + "step": 2720 + }, + { + "epoch": 0.6239394634258197, + "grad_norm": 0.40743128818438173, + "learning_rate": 0.00032724635291520694, + "loss": 2.3514, + "step": 2721 + }, + { + "epoch": 0.624168768631048, + "grad_norm": 0.4283149363752754, + "learning_rate": 0.00032689792293094937, + "loss": 2.4878, + "step": 2722 + }, + { + "epoch": 0.6243980738362761, + "grad_norm": 0.43032251540384403, + "learning_rate": 0.00032654958842863967, + "loss": 2.3872, + "step": 2723 + }, + { + "epoch": 0.6246273790415042, + "grad_norm": 0.4484193068856885, + "learning_rate": 0.0003262013496004167, + "loss": 2.6189, + "step": 2724 + }, + { + "epoch": 0.6248566842467324, + "grad_norm": 0.3838728276736482, + "learning_rate": 0.0003258532066383667, + "loss": 2.4729, + "step": 2725 + }, + { + "epoch": 0.6250859894519606, + "grad_norm": 0.4763259757293702, + "learning_rate": 0.00032550515973452294, + "loss": 2.5378, + "step": 2726 + }, + { + "epoch": 0.6253152946571887, + "grad_norm": 0.41746469994648444, + "learning_rate": 0.000325157209080866, + "loss": 2.4921, + "step": 2727 + }, + { + "epoch": 0.6255445998624168, + "grad_norm": 0.45712654996979524, + "learning_rate": 0.00032480935486932263, + "loss": 2.438, + "step": 2728 + }, + { + "epoch": 0.6257739050676451, + "grad_norm": 0.41235461049267513, + "learning_rate": 0.00032446159729176743, + "loss": 2.4125, + "step": 2729 + }, + { + "epoch": 0.6260032102728732, + "grad_norm": 0.41426971868196466, + "learning_rate": 0.0003241139365400207, + "loss": 2.5095, + "step": 2730 + }, + { + "epoch": 0.6262325154781013, + "grad_norm": 0.4152786842023299, + "learning_rate": 0.0003237663728058502, + "loss": 2.6621, + "step": 2731 + }, + { + "epoch": 0.6264618206833296, + "grad_norm": 0.41826576978479724, + "learning_rate": 0.0003234189062809695, + "loss": 2.5762, + "step": 2732 + }, + { + "epoch": 0.6266911258885577, + "grad_norm": 0.3932224219368948, + "learning_rate": 0.0003230715371570389, + "loss": 2.3902, + "step": 2733 + }, + { + "epoch": 0.6269204310937858, + "grad_norm": 0.40969696800459493, + "learning_rate": 0.00032272426562566484, + "loss": 2.5518, + "step": 2734 + }, + { + "epoch": 0.6271497362990139, + "grad_norm": 0.4184898843452659, + "learning_rate": 0.00032237709187839996, + "loss": 2.4942, + "step": 2735 + }, + { + "epoch": 0.6273790415042422, + "grad_norm": 0.4244259815431634, + "learning_rate": 0.0003220300161067432, + "loss": 2.4619, + "step": 2736 + }, + { + "epoch": 0.6276083467094703, + "grad_norm": 0.5644826927561171, + "learning_rate": 0.0003216830385021388, + "loss": 2.3927, + "step": 2737 + }, + { + "epoch": 0.6278376519146984, + "grad_norm": 0.4073321525559593, + "learning_rate": 0.0003213361592559777, + "loss": 2.4324, + "step": 2738 + }, + { + "epoch": 0.6280669571199267, + "grad_norm": 0.5436373477132271, + "learning_rate": 0.0003209893785595959, + "loss": 2.4731, + "step": 2739 + }, + { + "epoch": 0.6282962623251548, + "grad_norm": 0.3770086807595444, + "learning_rate": 0.0003206426966042754, + "loss": 2.2435, + "step": 2740 + }, + { + "epoch": 0.6285255675303829, + "grad_norm": 0.42220610053159885, + "learning_rate": 0.00032029611358124366, + "loss": 2.3883, + "step": 2741 + }, + { + "epoch": 0.6287548727356111, + "grad_norm": 0.3978283469446332, + "learning_rate": 0.00031994962968167365, + "loss": 2.2787, + "step": 2742 + }, + { + "epoch": 0.6289841779408393, + "grad_norm": 0.39797410153468804, + "learning_rate": 0.00031960324509668336, + "loss": 2.4712, + "step": 2743 + }, + { + "epoch": 0.6292134831460674, + "grad_norm": 0.3714433231360344, + "learning_rate": 0.00031925696001733637, + "loss": 2.4918, + "step": 2744 + }, + { + "epoch": 0.6294427883512955, + "grad_norm": 0.4020307102562083, + "learning_rate": 0.0003189107746346412, + "loss": 2.4474, + "step": 2745 + }, + { + "epoch": 0.6296720935565238, + "grad_norm": 0.401963060997167, + "learning_rate": 0.0003185646891395514, + "loss": 2.5483, + "step": 2746 + }, + { + "epoch": 0.6299013987617519, + "grad_norm": 0.4070020663855686, + "learning_rate": 0.0003182187037229653, + "loss": 2.3313, + "step": 2747 + }, + { + "epoch": 0.63013070396698, + "grad_norm": 0.3908856209112336, + "learning_rate": 0.00031787281857572627, + "loss": 2.461, + "step": 2748 + }, + { + "epoch": 0.6303600091722082, + "grad_norm": 0.39702793525384944, + "learning_rate": 0.0003175270338886221, + "loss": 2.3438, + "step": 2749 + }, + { + "epoch": 0.6305893143774364, + "grad_norm": 0.40379224459686597, + "learning_rate": 0.00031718134985238565, + "loss": 2.5094, + "step": 2750 + }, + { + "epoch": 0.6308186195826645, + "grad_norm": 0.413193869957779, + "learning_rate": 0.00031683576665769345, + "loss": 2.5173, + "step": 2751 + }, + { + "epoch": 0.6310479247878927, + "grad_norm": 0.4216856433118937, + "learning_rate": 0.0003164902844951673, + "loss": 2.4664, + "step": 2752 + }, + { + "epoch": 0.6312772299931209, + "grad_norm": 0.4014016293294973, + "learning_rate": 0.0003161449035553724, + "loss": 2.5471, + "step": 2753 + }, + { + "epoch": 0.631506535198349, + "grad_norm": 0.3723756497996698, + "learning_rate": 0.00031579962402881903, + "loss": 2.471, + "step": 2754 + }, + { + "epoch": 0.6317358404035771, + "grad_norm": 0.40305215821058304, + "learning_rate": 0.00031545444610596077, + "loss": 2.4359, + "step": 2755 + }, + { + "epoch": 0.6319651456088053, + "grad_norm": 0.417950909775346, + "learning_rate": 0.0003151093699771955, + "loss": 2.3196, + "step": 2756 + }, + { + "epoch": 0.6321944508140335, + "grad_norm": 0.4200859975716103, + "learning_rate": 0.000314764395832865, + "loss": 2.447, + "step": 2757 + }, + { + "epoch": 0.6324237560192616, + "grad_norm": 0.38276450732664985, + "learning_rate": 0.00031441952386325467, + "loss": 2.4684, + "step": 2758 + }, + { + "epoch": 0.6326530612244898, + "grad_norm": 0.41156507721559343, + "learning_rate": 0.0003140747542585934, + "loss": 2.4362, + "step": 2759 + }, + { + "epoch": 0.632882366429718, + "grad_norm": 0.3942924169063572, + "learning_rate": 0.0003137300872090542, + "loss": 2.5788, + "step": 2760 + }, + { + "epoch": 0.6331116716349461, + "grad_norm": 0.379338390275833, + "learning_rate": 0.00031338552290475266, + "loss": 2.6278, + "step": 2761 + }, + { + "epoch": 0.6333409768401743, + "grad_norm": 0.41014055455438936, + "learning_rate": 0.0003130410615357485, + "loss": 2.2815, + "step": 2762 + }, + { + "epoch": 0.6335702820454024, + "grad_norm": 0.44355322234797273, + "learning_rate": 0.00031269670329204396, + "loss": 2.5236, + "step": 2763 + }, + { + "epoch": 0.6337995872506306, + "grad_norm": 0.4410116659470103, + "learning_rate": 0.00031235244836358525, + "loss": 2.4871, + "step": 2764 + }, + { + "epoch": 0.6340288924558587, + "grad_norm": 0.37597407328450955, + "learning_rate": 0.0003120082969402604, + "loss": 2.3672, + "step": 2765 + }, + { + "epoch": 0.6342581976610869, + "grad_norm": 0.4067643972901847, + "learning_rate": 0.00031166424921190173, + "loss": 2.4425, + "step": 2766 + }, + { + "epoch": 0.6344875028663151, + "grad_norm": 0.395735821688463, + "learning_rate": 0.00031132030536828314, + "loss": 2.4986, + "step": 2767 + }, + { + "epoch": 0.6347168080715432, + "grad_norm": 0.4051049202827995, + "learning_rate": 0.00031097646559912205, + "loss": 2.396, + "step": 2768 + }, + { + "epoch": 0.6349461132767714, + "grad_norm": 0.42524628511005685, + "learning_rate": 0.00031063273009407805, + "loss": 2.3872, + "step": 2769 + }, + { + "epoch": 0.6351754184819995, + "grad_norm": 1.9268731226961358, + "learning_rate": 0.0003102890990427535, + "loss": 2.2699, + "step": 2770 + }, + { + "epoch": 0.6354047236872277, + "grad_norm": 0.4239536199374805, + "learning_rate": 0.00030994557263469265, + "loss": 2.3999, + "step": 2771 + }, + { + "epoch": 0.6356340288924559, + "grad_norm": 0.4297377792543152, + "learning_rate": 0.00030960215105938274, + "loss": 2.5193, + "step": 2772 + }, + { + "epoch": 0.635863334097684, + "grad_norm": 0.4444994661457377, + "learning_rate": 0.0003092588345062526, + "loss": 2.538, + "step": 2773 + }, + { + "epoch": 0.6360926393029122, + "grad_norm": 0.41350288743826547, + "learning_rate": 0.00030891562316467345, + "loss": 2.5371, + "step": 2774 + }, + { + "epoch": 0.6363219445081403, + "grad_norm": 0.4693394740510387, + "learning_rate": 0.0003085725172239582, + "loss": 2.4513, + "step": 2775 + }, + { + "epoch": 0.6365512497133685, + "grad_norm": 0.43070600294213546, + "learning_rate": 0.0003082295168733621, + "loss": 2.4541, + "step": 2776 + }, + { + "epoch": 0.6367805549185966, + "grad_norm": 0.8724730717145117, + "learning_rate": 0.0003078866223020815, + "loss": 2.4582, + "step": 2777 + }, + { + "epoch": 0.6370098601238248, + "grad_norm": 0.40971718312788774, + "learning_rate": 0.0003075438336992551, + "loss": 2.3769, + "step": 2778 + }, + { + "epoch": 0.637239165329053, + "grad_norm": 0.4537390633916842, + "learning_rate": 0.0003072011512539624, + "loss": 2.4675, + "step": 2779 + }, + { + "epoch": 0.6374684705342811, + "grad_norm": 0.4407730244327278, + "learning_rate": 0.00030685857515522517, + "loss": 2.5982, + "step": 2780 + }, + { + "epoch": 0.6376977757395093, + "grad_norm": 0.4145475437405502, + "learning_rate": 0.00030651610559200574, + "loss": 2.5332, + "step": 2781 + }, + { + "epoch": 0.6379270809447375, + "grad_norm": 0.5112126436958023, + "learning_rate": 0.0003061737427532083, + "loss": 2.506, + "step": 2782 + }, + { + "epoch": 0.6381563861499656, + "grad_norm": 0.4126973161010612, + "learning_rate": 0.00030583148682767757, + "loss": 2.4676, + "step": 2783 + }, + { + "epoch": 0.6383856913551937, + "grad_norm": 0.4210590748617979, + "learning_rate": 0.0003054893380042001, + "loss": 2.4438, + "step": 2784 + }, + { + "epoch": 0.638614996560422, + "grad_norm": 0.46543971226327074, + "learning_rate": 0.00030514729647150243, + "loss": 2.4725, + "step": 2785 + }, + { + "epoch": 0.6388443017656501, + "grad_norm": 0.43362505991728045, + "learning_rate": 0.00030480536241825264, + "loss": 2.4857, + "step": 2786 + }, + { + "epoch": 0.6390736069708782, + "grad_norm": 0.46247107666088016, + "learning_rate": 0.0003044635360330592, + "loss": 2.5322, + "step": 2787 + }, + { + "epoch": 0.6393029121761064, + "grad_norm": 0.525557225812654, + "learning_rate": 0.0003041218175044713, + "loss": 2.3418, + "step": 2788 + }, + { + "epoch": 0.6395322173813346, + "grad_norm": 0.4179636790128022, + "learning_rate": 0.00030378020702097845, + "loss": 2.532, + "step": 2789 + }, + { + "epoch": 0.6397615225865627, + "grad_norm": 0.43017181843913066, + "learning_rate": 0.0003034387047710111, + "loss": 2.5742, + "step": 2790 + }, + { + "epoch": 0.6399908277917908, + "grad_norm": 0.5363228659090693, + "learning_rate": 0.000303097310942939, + "loss": 2.5146, + "step": 2791 + }, + { + "epoch": 0.6402201329970191, + "grad_norm": 0.839912329127438, + "learning_rate": 0.0003027560257250733, + "loss": 2.3085, + "step": 2792 + }, + { + "epoch": 0.6404494382022472, + "grad_norm": 0.3877968399693849, + "learning_rate": 0.0003024148493056641, + "loss": 2.4213, + "step": 2793 + }, + { + "epoch": 0.6406787434074753, + "grad_norm": 0.40747119718747105, + "learning_rate": 0.00030207378187290245, + "loss": 2.536, + "step": 2794 + }, + { + "epoch": 0.6409080486127036, + "grad_norm": 0.43332915365388675, + "learning_rate": 0.00030173282361491865, + "loss": 2.4826, + "step": 2795 + }, + { + "epoch": 0.6411373538179317, + "grad_norm": 0.4638627507583677, + "learning_rate": 0.0003013919747197832, + "loss": 2.4576, + "step": 2796 + }, + { + "epoch": 0.6413666590231598, + "grad_norm": 0.41647008058690077, + "learning_rate": 0.0003010512353755057, + "loss": 2.4221, + "step": 2797 + }, + { + "epoch": 0.6415959642283879, + "grad_norm": 0.4136677633570984, + "learning_rate": 0.00030071060577003606, + "loss": 2.514, + "step": 2798 + }, + { + "epoch": 0.6418252694336162, + "grad_norm": 0.4155683730053769, + "learning_rate": 0.00030037008609126313, + "loss": 2.4918, + "step": 2799 + }, + { + "epoch": 0.6420545746388443, + "grad_norm": 0.3924106828596435, + "learning_rate": 0.00030002967652701534, + "loss": 2.4296, + "step": 2800 + }, + { + "epoch": 0.6422838798440724, + "grad_norm": 0.42003311142811744, + "learning_rate": 0.0002996893772650602, + "loss": 2.4156, + "step": 2801 + }, + { + "epoch": 0.6425131850493007, + "grad_norm": 0.47167962581509254, + "learning_rate": 0.0002993491884931048, + "loss": 2.4214, + "step": 2802 + }, + { + "epoch": 0.6427424902545288, + "grad_norm": 0.5938710813206443, + "learning_rate": 0.0002990091103987945, + "loss": 2.4945, + "step": 2803 + }, + { + "epoch": 0.6429717954597569, + "grad_norm": 0.4331963246668154, + "learning_rate": 0.0002986691431697148, + "loss": 2.4885, + "step": 2804 + }, + { + "epoch": 0.6432011006649851, + "grad_norm": 0.41768232633107194, + "learning_rate": 0.0002983292869933886, + "loss": 2.3891, + "step": 2805 + }, + { + "epoch": 0.6434304058702133, + "grad_norm": 0.43547433293852134, + "learning_rate": 0.00029798954205727886, + "loss": 2.4734, + "step": 2806 + }, + { + "epoch": 0.6436597110754414, + "grad_norm": 0.4332881092547548, + "learning_rate": 0.0002976499085487862, + "loss": 2.5093, + "step": 2807 + }, + { + "epoch": 0.6438890162806695, + "grad_norm": 0.42826450247645537, + "learning_rate": 0.0002973103866552506, + "loss": 2.4055, + "step": 2808 + }, + { + "epoch": 0.6441183214858978, + "grad_norm": 0.46519224263283554, + "learning_rate": 0.00029697097656394963, + "loss": 2.518, + "step": 2809 + }, + { + "epoch": 0.6443476266911259, + "grad_norm": 0.4245106771636753, + "learning_rate": 0.0002966316784621, + "loss": 2.2215, + "step": 2810 + }, + { + "epoch": 0.644576931896354, + "grad_norm": 0.4520301034879141, + "learning_rate": 0.00029629249253685595, + "loss": 2.3929, + "step": 2811 + }, + { + "epoch": 0.6448062371015822, + "grad_norm": 0.4403635120066087, + "learning_rate": 0.0002959534189753105, + "loss": 2.6074, + "step": 2812 + }, + { + "epoch": 0.6450355423068104, + "grad_norm": 0.4253712785236607, + "learning_rate": 0.00029561445796449416, + "loss": 2.4162, + "step": 2813 + }, + { + "epoch": 0.6452648475120385, + "grad_norm": 0.4488213689938859, + "learning_rate": 0.0002952756096913756, + "loss": 2.5209, + "step": 2814 + }, + { + "epoch": 0.6454941527172667, + "grad_norm": 0.6569411036417919, + "learning_rate": 0.0002949368743428612, + "loss": 2.4694, + "step": 2815 + }, + { + "epoch": 0.6457234579224949, + "grad_norm": 0.6832409636216549, + "learning_rate": 0.0002945982521057953, + "loss": 2.4355, + "step": 2816 + }, + { + "epoch": 0.645952763127723, + "grad_norm": 0.43927742164486455, + "learning_rate": 0.0002942597431669593, + "loss": 2.3366, + "step": 2817 + }, + { + "epoch": 0.6461820683329511, + "grad_norm": 0.44918261962053563, + "learning_rate": 0.000293921347713073, + "loss": 2.4779, + "step": 2818 + }, + { + "epoch": 0.6464113735381793, + "grad_norm": 0.3855656574457366, + "learning_rate": 0.0002935830659307924, + "loss": 2.448, + "step": 2819 + }, + { + "epoch": 0.6466406787434075, + "grad_norm": 0.49794642605870437, + "learning_rate": 0.0002932448980067122, + "loss": 2.3063, + "step": 2820 + }, + { + "epoch": 0.6468699839486356, + "grad_norm": 0.400805590083751, + "learning_rate": 0.0002929068441273629, + "loss": 2.445, + "step": 2821 + }, + { + "epoch": 0.6470992891538638, + "grad_norm": 0.4131756173291388, + "learning_rate": 0.0002925689044792132, + "loss": 2.483, + "step": 2822 + }, + { + "epoch": 0.647328594359092, + "grad_norm": 0.46281583845598057, + "learning_rate": 0.0002922310792486681, + "loss": 2.5115, + "step": 2823 + }, + { + "epoch": 0.6475578995643201, + "grad_norm": 0.415823037734872, + "learning_rate": 0.00029189336862207003, + "loss": 2.45, + "step": 2824 + }, + { + "epoch": 0.6477872047695483, + "grad_norm": 0.42270740896639253, + "learning_rate": 0.00029155577278569745, + "loss": 2.427, + "step": 2825 + }, + { + "epoch": 0.6480165099747764, + "grad_norm": 0.4314392561028178, + "learning_rate": 0.0002912182919257664, + "loss": 2.3527, + "step": 2826 + }, + { + "epoch": 0.6482458151800046, + "grad_norm": 0.39238740150019213, + "learning_rate": 0.00029088092622842895, + "loss": 2.5694, + "step": 2827 + }, + { + "epoch": 0.6484751203852327, + "grad_norm": 0.46541249635476184, + "learning_rate": 0.0002905436758797737, + "loss": 2.492, + "step": 2828 + }, + { + "epoch": 0.6487044255904609, + "grad_norm": 0.4091097365340356, + "learning_rate": 0.00029020654106582544, + "loss": 2.4349, + "step": 2829 + }, + { + "epoch": 0.6489337307956891, + "grad_norm": 0.3973462313297649, + "learning_rate": 0.000289869521972546, + "loss": 2.5564, + "step": 2830 + }, + { + "epoch": 0.6491630360009172, + "grad_norm": 0.4402711941905484, + "learning_rate": 0.0002895326187858326, + "loss": 2.3869, + "step": 2831 + }, + { + "epoch": 0.6493923412061454, + "grad_norm": 0.3949670953125137, + "learning_rate": 0.0002891958316915187, + "loss": 2.3706, + "step": 2832 + }, + { + "epoch": 0.6496216464113735, + "grad_norm": 0.5479583653317684, + "learning_rate": 0.00028885916087537377, + "loss": 2.4149, + "step": 2833 + }, + { + "epoch": 0.6498509516166017, + "grad_norm": 0.506513769745912, + "learning_rate": 0.0002885226065231035, + "loss": 2.4815, + "step": 2834 + }, + { + "epoch": 0.6500802568218299, + "grad_norm": 0.44820896928587384, + "learning_rate": 0.00028818616882034877, + "loss": 2.3701, + "step": 2835 + }, + { + "epoch": 0.650309562027058, + "grad_norm": 0.443477452793199, + "learning_rate": 0.00028784984795268643, + "loss": 2.5054, + "step": 2836 + }, + { + "epoch": 0.6505388672322862, + "grad_norm": 0.4127145880305316, + "learning_rate": 0.0002875136441056286, + "loss": 2.3366, + "step": 2837 + }, + { + "epoch": 0.6507681724375143, + "grad_norm": 0.42592259875685373, + "learning_rate": 0.00028717755746462373, + "loss": 2.4288, + "step": 2838 + }, + { + "epoch": 0.6509974776427425, + "grad_norm": 0.4446836040684875, + "learning_rate": 0.000286841588215054, + "loss": 2.4577, + "step": 2839 + }, + { + "epoch": 0.6512267828479706, + "grad_norm": 0.450662317152369, + "learning_rate": 0.0002865057365422386, + "loss": 2.4222, + "step": 2840 + }, + { + "epoch": 0.6514560880531988, + "grad_norm": 0.4424497406343933, + "learning_rate": 0.0002861700026314308, + "loss": 2.4728, + "step": 2841 + }, + { + "epoch": 0.651685393258427, + "grad_norm": 0.39561599588030233, + "learning_rate": 0.0002858343866678191, + "loss": 2.4633, + "step": 2842 + }, + { + "epoch": 0.6519146984636551, + "grad_norm": 0.6871973223954196, + "learning_rate": 0.00028549888883652686, + "loss": 2.5785, + "step": 2843 + }, + { + "epoch": 0.6521440036688833, + "grad_norm": 0.3983091832033729, + "learning_rate": 0.0002851635093226129, + "loss": 2.5763, + "step": 2844 + }, + { + "epoch": 0.6523733088741115, + "grad_norm": 0.42684971227221075, + "learning_rate": 0.00028482824831107, + "loss": 2.5383, + "step": 2845 + }, + { + "epoch": 0.6526026140793396, + "grad_norm": 0.385382324761664, + "learning_rate": 0.0002844931059868261, + "loss": 2.4981, + "step": 2846 + }, + { + "epoch": 0.6528319192845677, + "grad_norm": 0.46444195264946664, + "learning_rate": 0.000284158082534743, + "loss": 2.3713, + "step": 2847 + }, + { + "epoch": 0.6530612244897959, + "grad_norm": 0.435482454357957, + "learning_rate": 0.000283823178139618, + "loss": 2.3983, + "step": 2848 + }, + { + "epoch": 0.6532905296950241, + "grad_norm": 0.4313604180272572, + "learning_rate": 0.00028348839298618177, + "loss": 2.5171, + "step": 2849 + }, + { + "epoch": 0.6535198349002522, + "grad_norm": 0.43149726682054307, + "learning_rate": 0.0002831537272590997, + "loss": 2.4547, + "step": 2850 + }, + { + "epoch": 0.6537491401054804, + "grad_norm": 0.4362986665444281, + "learning_rate": 0.0002828191811429709, + "loss": 2.479, + "step": 2851 + }, + { + "epoch": 0.6539784453107086, + "grad_norm": 0.4305716334649344, + "learning_rate": 0.00028248475482232937, + "loss": 2.5289, + "step": 2852 + }, + { + "epoch": 0.6542077505159367, + "grad_norm": 0.41686674706203863, + "learning_rate": 0.00028215044848164164, + "loss": 2.3153, + "step": 2853 + }, + { + "epoch": 0.6544370557211648, + "grad_norm": 0.4477042458906509, + "learning_rate": 0.0002818162623053094, + "loss": 2.406, + "step": 2854 + }, + { + "epoch": 0.6546663609263931, + "grad_norm": 0.4091583560936637, + "learning_rate": 0.00028148219647766747, + "loss": 2.3886, + "step": 2855 + }, + { + "epoch": 0.6548956661316212, + "grad_norm": 0.38726231127198607, + "learning_rate": 0.00028114825118298413, + "loss": 2.4614, + "step": 2856 + }, + { + "epoch": 0.6551249713368493, + "grad_norm": 0.48319154309749185, + "learning_rate": 0.00028081442660546124, + "loss": 2.3714, + "step": 2857 + }, + { + "epoch": 0.6553542765420775, + "grad_norm": 0.42846219631524607, + "learning_rate": 0.00028048072292923465, + "loss": 2.5275, + "step": 2858 + }, + { + "epoch": 0.6555835817473057, + "grad_norm": 0.4052599172208735, + "learning_rate": 0.0002801471403383728, + "loss": 2.3563, + "step": 2859 + }, + { + "epoch": 0.6558128869525338, + "grad_norm": 0.3915313056857332, + "learning_rate": 0.00027981367901687773, + "loss": 2.3774, + "step": 2860 + }, + { + "epoch": 0.6560421921577619, + "grad_norm": 0.4500532864744177, + "learning_rate": 0.00027948033914868415, + "loss": 2.3992, + "step": 2861 + }, + { + "epoch": 0.6562714973629902, + "grad_norm": 0.40709766080319676, + "learning_rate": 0.0002791471209176606, + "loss": 2.4278, + "step": 2862 + }, + { + "epoch": 0.6565008025682183, + "grad_norm": 0.3924701989194399, + "learning_rate": 0.00027881402450760775, + "loss": 2.2987, + "step": 2863 + }, + { + "epoch": 0.6567301077734464, + "grad_norm": 0.4163194144336759, + "learning_rate": 0.00027848105010225933, + "loss": 2.3854, + "step": 2864 + }, + { + "epoch": 0.6569594129786747, + "grad_norm": 0.3921893411431849, + "learning_rate": 0.00027814819788528165, + "loss": 2.3656, + "step": 2865 + }, + { + "epoch": 0.6571887181839028, + "grad_norm": 0.4236525284195649, + "learning_rate": 0.00027781546804027445, + "loss": 2.4823, + "step": 2866 + }, + { + "epoch": 0.6574180233891309, + "grad_norm": 0.43391109405678385, + "learning_rate": 0.00027748286075076836, + "loss": 2.4653, + "step": 2867 + }, + { + "epoch": 0.657647328594359, + "grad_norm": 0.3969913865116351, + "learning_rate": 0.00027715037620022797, + "loss": 2.3939, + "step": 2868 + }, + { + "epoch": 0.6578766337995873, + "grad_norm": 0.4107107380684914, + "learning_rate": 0.00027681801457204937, + "loss": 2.3703, + "step": 2869 + }, + { + "epoch": 0.6581059390048154, + "grad_norm": 0.3893903174889239, + "learning_rate": 0.0002764857760495609, + "loss": 2.4334, + "step": 2870 + }, + { + "epoch": 0.6583352442100435, + "grad_norm": 0.4005843667029635, + "learning_rate": 0.00027615366081602306, + "loss": 2.4529, + "step": 2871 + }, + { + "epoch": 0.6585645494152718, + "grad_norm": 0.4247732028898823, + "learning_rate": 0.0002758216690546287, + "loss": 2.4512, + "step": 2872 + }, + { + "epoch": 0.6587938546204999, + "grad_norm": 0.6131213358190342, + "learning_rate": 0.0002754898009485021, + "loss": 2.5931, + "step": 2873 + }, + { + "epoch": 0.659023159825728, + "grad_norm": 0.5852018635713097, + "learning_rate": 0.00027515805668069937, + "loss": 2.4644, + "step": 2874 + }, + { + "epoch": 0.6592524650309562, + "grad_norm": 0.4169021022441875, + "learning_rate": 0.0002748264364342085, + "loss": 2.4189, + "step": 2875 + }, + { + "epoch": 0.6594817702361844, + "grad_norm": 0.44686161167754784, + "learning_rate": 0.000274494940391949, + "loss": 2.34, + "step": 2876 + }, + { + "epoch": 0.6597110754414125, + "grad_norm": 0.5006220607632837, + "learning_rate": 0.00027416356873677204, + "loss": 2.636, + "step": 2877 + }, + { + "epoch": 0.6599403806466407, + "grad_norm": 0.41868835967921714, + "learning_rate": 0.00027383232165145976, + "loss": 2.3193, + "step": 2878 + }, + { + "epoch": 0.6601696858518689, + "grad_norm": 0.396074362851744, + "learning_rate": 0.0002735011993187258, + "loss": 2.3246, + "step": 2879 + }, + { + "epoch": 0.660398991057097, + "grad_norm": 0.4017740352121225, + "learning_rate": 0.0002731702019212156, + "loss": 2.5448, + "step": 2880 + }, + { + "epoch": 0.6606282962623251, + "grad_norm": 0.3955197012258001, + "learning_rate": 0.0002728393296415042, + "loss": 2.512, + "step": 2881 + }, + { + "epoch": 0.6608576014675533, + "grad_norm": 0.4035551258733985, + "learning_rate": 0.00027250858266209915, + "loss": 2.5139, + "step": 2882 + }, + { + "epoch": 0.6610869066727815, + "grad_norm": 0.3985409516126535, + "learning_rate": 0.00027217796116543817, + "loss": 2.3981, + "step": 2883 + }, + { + "epoch": 0.6613162118780096, + "grad_norm": 0.4284184679680021, + "learning_rate": 0.0002718474653338899, + "loss": 2.3817, + "step": 2884 + }, + { + "epoch": 0.6615455170832378, + "grad_norm": 0.47578507524368296, + "learning_rate": 0.0002715170953497532, + "loss": 2.4503, + "step": 2885 + }, + { + "epoch": 0.661774822288466, + "grad_norm": 0.43032491036745, + "learning_rate": 0.0002711868513952587, + "loss": 2.4408, + "step": 2886 + }, + { + "epoch": 0.6620041274936941, + "grad_norm": 0.4295097599483036, + "learning_rate": 0.00027085673365256614, + "loss": 2.4454, + "step": 2887 + }, + { + "epoch": 0.6622334326989223, + "grad_norm": 0.40657176490929336, + "learning_rate": 0.0002705267423037666, + "loss": 2.5045, + "step": 2888 + }, + { + "epoch": 0.6624627379041504, + "grad_norm": 0.43206408042447775, + "learning_rate": 0.00027019687753088075, + "loss": 2.5458, + "step": 2889 + }, + { + "epoch": 0.6626920431093786, + "grad_norm": 0.40557942903307964, + "learning_rate": 0.0002698671395158603, + "loss": 2.2752, + "step": 2890 + }, + { + "epoch": 0.6629213483146067, + "grad_norm": 0.41888366190671644, + "learning_rate": 0.00026953752844058597, + "loss": 2.5597, + "step": 2891 + }, + { + "epoch": 0.6631506535198349, + "grad_norm": 0.39057667880816693, + "learning_rate": 0.0002692080444868695, + "loss": 2.5052, + "step": 2892 + }, + { + "epoch": 0.6633799587250631, + "grad_norm": 0.3979132147082578, + "learning_rate": 0.0002688786878364516, + "loss": 2.5253, + "step": 2893 + }, + { + "epoch": 0.6636092639302912, + "grad_norm": 0.35251795436298694, + "learning_rate": 0.00026854945867100377, + "loss": 2.3727, + "step": 2894 + }, + { + "epoch": 0.6638385691355194, + "grad_norm": 0.4063156445630605, + "learning_rate": 0.00026822035717212597, + "loss": 2.5366, + "step": 2895 + }, + { + "epoch": 0.6640678743407475, + "grad_norm": 0.4349172098842921, + "learning_rate": 0.0002678913835213488, + "loss": 2.3446, + "step": 2896 + }, + { + "epoch": 0.6642971795459757, + "grad_norm": 0.4198099004466601, + "learning_rate": 0.00026756253790013193, + "loss": 2.4687, + "step": 2897 + }, + { + "epoch": 0.6645264847512039, + "grad_norm": 0.4292180772199803, + "learning_rate": 0.0002672338204898642, + "loss": 2.4686, + "step": 2898 + }, + { + "epoch": 0.664755789956432, + "grad_norm": 0.40438577013770866, + "learning_rate": 0.0002669052314718641, + "loss": 2.3602, + "step": 2899 + }, + { + "epoch": 0.6649850951616602, + "grad_norm": 0.3770347014926654, + "learning_rate": 0.00026657677102737875, + "loss": 2.6425, + "step": 2900 + }, + { + "epoch": 0.6652144003668883, + "grad_norm": 0.39604062662616496, + "learning_rate": 0.0002662484393375855, + "loss": 2.4559, + "step": 2901 + }, + { + "epoch": 0.6654437055721165, + "grad_norm": 0.4187121585861029, + "learning_rate": 0.00026592023658358904, + "loss": 2.5109, + "step": 2902 + }, + { + "epoch": 0.6656730107773446, + "grad_norm": 0.41022242139517967, + "learning_rate": 0.00026559216294642446, + "loss": 2.4808, + "step": 2903 + }, + { + "epoch": 0.6659023159825728, + "grad_norm": 0.43816437456581714, + "learning_rate": 0.00026526421860705474, + "loss": 2.4614, + "step": 2904 + }, + { + "epoch": 0.666131621187801, + "grad_norm": 0.39062880562884156, + "learning_rate": 0.0002649364037463718, + "loss": 2.5494, + "step": 2905 + }, + { + "epoch": 0.6663609263930291, + "grad_norm": 0.4890345453849536, + "learning_rate": 0.0002646087185451959, + "loss": 2.4707, + "step": 2906 + }, + { + "epoch": 0.6665902315982573, + "grad_norm": 0.4093148689830726, + "learning_rate": 0.0002642811631842764, + "loss": 2.4386, + "step": 2907 + }, + { + "epoch": 0.6668195368034855, + "grad_norm": 0.4026197780670048, + "learning_rate": 0.00026395373784429053, + "loss": 2.4449, + "step": 2908 + }, + { + "epoch": 0.6670488420087136, + "grad_norm": 0.42732392554582116, + "learning_rate": 0.0002636264427058439, + "loss": 2.3546, + "step": 2909 + }, + { + "epoch": 0.6672781472139417, + "grad_norm": 0.7015163028255288, + "learning_rate": 0.00026329927794947, + "loss": 2.3967, + "step": 2910 + }, + { + "epoch": 0.6675074524191699, + "grad_norm": 0.4056486118248027, + "learning_rate": 0.00026297224375563123, + "loss": 2.4324, + "step": 2911 + }, + { + "epoch": 0.6677367576243981, + "grad_norm": 0.4614216056738353, + "learning_rate": 0.00026264534030471723, + "loss": 2.5398, + "step": 2912 + }, + { + "epoch": 0.6679660628296262, + "grad_norm": 0.5353354967237349, + "learning_rate": 0.00026231856777704575, + "loss": 2.3394, + "step": 2913 + }, + { + "epoch": 0.6681953680348544, + "grad_norm": 0.4131799790389513, + "learning_rate": 0.0002619919263528622, + "loss": 2.4678, + "step": 2914 + }, + { + "epoch": 0.6684246732400826, + "grad_norm": 0.41701819378201127, + "learning_rate": 0.00026166541621234026, + "loss": 2.4215, + "step": 2915 + }, + { + "epoch": 0.6686539784453107, + "grad_norm": 0.4033720184717444, + "learning_rate": 0.0002613390375355801, + "loss": 2.4875, + "step": 2916 + }, + { + "epoch": 0.6688832836505388, + "grad_norm": 0.41848611889301435, + "learning_rate": 0.00026101279050261045, + "loss": 2.3959, + "step": 2917 + }, + { + "epoch": 0.6691125888557671, + "grad_norm": 0.4571186660166063, + "learning_rate": 0.0002606866752933869, + "loss": 2.4827, + "step": 2918 + }, + { + "epoch": 0.6693418940609952, + "grad_norm": 0.4436533718661509, + "learning_rate": 0.00026036069208779247, + "loss": 2.5694, + "step": 2919 + }, + { + "epoch": 0.6695711992662233, + "grad_norm": 0.410556355907968, + "learning_rate": 0.00026003484106563703, + "loss": 2.4131, + "step": 2920 + }, + { + "epoch": 0.6698005044714515, + "grad_norm": 0.4136041429968752, + "learning_rate": 0.0002597091224066581, + "loss": 2.3895, + "step": 2921 + }, + { + "epoch": 0.6700298096766797, + "grad_norm": 0.4077016154341346, + "learning_rate": 0.00025938353629052004, + "loss": 2.517, + "step": 2922 + }, + { + "epoch": 0.6702591148819078, + "grad_norm": 0.39102173607874297, + "learning_rate": 0.00025905808289681365, + "loss": 2.4104, + "step": 2923 + }, + { + "epoch": 0.6704884200871359, + "grad_norm": 0.5164429763260153, + "learning_rate": 0.00025873276240505685, + "loss": 2.4053, + "step": 2924 + }, + { + "epoch": 0.6707177252923642, + "grad_norm": 0.40124882876778895, + "learning_rate": 0.0002584075749946946, + "loss": 2.3389, + "step": 2925 + }, + { + "epoch": 0.6709470304975923, + "grad_norm": 0.507240691114077, + "learning_rate": 0.0002580825208450978, + "loss": 2.4824, + "step": 2926 + }, + { + "epoch": 0.6711763357028204, + "grad_norm": 0.4634755971661481, + "learning_rate": 0.00025775760013556424, + "loss": 2.5207, + "step": 2927 + }, + { + "epoch": 0.6714056409080487, + "grad_norm": 0.3857920551384695, + "learning_rate": 0.00025743281304531785, + "loss": 2.3338, + "step": 2928 + }, + { + "epoch": 0.6716349461132768, + "grad_norm": 0.4318714389720274, + "learning_rate": 0.0002571081597535095, + "loss": 2.4893, + "step": 2929 + }, + { + "epoch": 0.6718642513185049, + "grad_norm": 0.37981193378438455, + "learning_rate": 0.000256783640439215, + "loss": 2.3726, + "step": 2930 + }, + { + "epoch": 0.672093556523733, + "grad_norm": 0.42333685615840555, + "learning_rate": 0.00025645925528143776, + "loss": 2.47, + "step": 2931 + }, + { + "epoch": 0.6723228617289613, + "grad_norm": 0.4095811672251681, + "learning_rate": 0.0002561350044591061, + "loss": 2.4293, + "step": 2932 + }, + { + "epoch": 0.6725521669341894, + "grad_norm": 0.3893626303967803, + "learning_rate": 0.0002558108881510747, + "loss": 2.3852, + "step": 2933 + }, + { + "epoch": 0.6727814721394175, + "grad_norm": 0.4382833864637244, + "learning_rate": 0.0002554869065361237, + "loss": 2.4423, + "step": 2934 + }, + { + "epoch": 0.6730107773446458, + "grad_norm": 0.39442228476385965, + "learning_rate": 0.00025516305979295963, + "loss": 2.5124, + "step": 2935 + }, + { + "epoch": 0.6732400825498739, + "grad_norm": 0.4158310658905497, + "learning_rate": 0.000254839348100214, + "loss": 2.5109, + "step": 2936 + }, + { + "epoch": 0.673469387755102, + "grad_norm": 0.44970716500145697, + "learning_rate": 0.0002545157716364439, + "loss": 2.4328, + "step": 2937 + }, + { + "epoch": 0.6736986929603302, + "grad_norm": 0.4649518921748269, + "learning_rate": 0.00025419233058013193, + "loss": 2.4624, + "step": 2938 + }, + { + "epoch": 0.6739279981655584, + "grad_norm": 0.41628737949090994, + "learning_rate": 0.00025386902510968624, + "loss": 2.4298, + "step": 2939 + }, + { + "epoch": 0.6741573033707865, + "grad_norm": 0.38977182837832647, + "learning_rate": 0.00025354585540343985, + "loss": 2.3109, + "step": 2940 + }, + { + "epoch": 0.6743866085760146, + "grad_norm": 0.38635733220896695, + "learning_rate": 0.00025322282163965095, + "loss": 2.568, + "step": 2941 + }, + { + "epoch": 0.6746159137812429, + "grad_norm": 0.42889475549966216, + "learning_rate": 0.0002528999239965027, + "loss": 2.4006, + "step": 2942 + }, + { + "epoch": 0.674845218986471, + "grad_norm": 0.4144904947008498, + "learning_rate": 0.00025257716265210384, + "loss": 2.4843, + "step": 2943 + }, + { + "epoch": 0.6750745241916991, + "grad_norm": 0.4324497281564362, + "learning_rate": 0.00025225453778448647, + "loss": 2.4484, + "step": 2944 + }, + { + "epoch": 0.6753038293969273, + "grad_norm": 0.4361220641899095, + "learning_rate": 0.0002519320495716091, + "loss": 2.4866, + "step": 2945 + }, + { + "epoch": 0.6755331346021555, + "grad_norm": 0.4253219391538684, + "learning_rate": 0.00025160969819135366, + "loss": 2.5803, + "step": 2946 + }, + { + "epoch": 0.6757624398073836, + "grad_norm": 0.3793578978272472, + "learning_rate": 0.00025128748382152716, + "loss": 2.4093, + "step": 2947 + }, + { + "epoch": 0.6759917450126118, + "grad_norm": 0.44070824962376176, + "learning_rate": 0.0002509654066398607, + "loss": 2.3885, + "step": 2948 + }, + { + "epoch": 0.67622105021784, + "grad_norm": 0.45442871867254797, + "learning_rate": 0.00025064346682401016, + "loss": 2.556, + "step": 2949 + }, + { + "epoch": 0.6764503554230681, + "grad_norm": 0.39272604261891764, + "learning_rate": 0.0002503216645515553, + "loss": 2.4922, + "step": 2950 + }, + { + "epoch": 0.6766796606282962, + "grad_norm": 0.404474261076567, + "learning_rate": 0.0002500000000000001, + "loss": 2.4399, + "step": 2951 + }, + { + "epoch": 0.6769089658335244, + "grad_norm": 0.42068450710407645, + "learning_rate": 0.00024967847334677235, + "loss": 2.3546, + "step": 2952 + }, + { + "epoch": 0.6771382710387526, + "grad_norm": 0.40647435485164407, + "learning_rate": 0.0002493570847692246, + "loss": 2.5164, + "step": 2953 + }, + { + "epoch": 0.6773675762439807, + "grad_norm": 0.4257784960015678, + "learning_rate": 0.00024903583444463187, + "loss": 2.4724, + "step": 2954 + }, + { + "epoch": 0.6775968814492089, + "grad_norm": 0.40296513665986367, + "learning_rate": 0.00024871472255019424, + "loss": 2.3776, + "step": 2955 + }, + { + "epoch": 0.6778261866544371, + "grad_norm": 0.37911318926106663, + "learning_rate": 0.0002483937492630345, + "loss": 2.4724, + "step": 2956 + }, + { + "epoch": 0.6780554918596652, + "grad_norm": 0.3885333003921011, + "learning_rate": 0.00024807291476019994, + "loss": 2.4599, + "step": 2957 + }, + { + "epoch": 0.6782847970648934, + "grad_norm": 0.4121780641875266, + "learning_rate": 0.00024775221921866005, + "loss": 2.457, + "step": 2958 + }, + { + "epoch": 0.6785141022701215, + "grad_norm": 0.4019697414906326, + "learning_rate": 0.00024743166281530877, + "loss": 2.3373, + "step": 2959 + }, + { + "epoch": 0.6787434074753497, + "grad_norm": 0.39255890146691613, + "learning_rate": 0.0002471112457269628, + "loss": 2.2655, + "step": 2960 + }, + { + "epoch": 0.6789727126805778, + "grad_norm": 0.4172313419286418, + "learning_rate": 0.000246790968130362, + "loss": 2.4102, + "step": 2961 + }, + { + "epoch": 0.679202017885806, + "grad_norm": 0.6007899649845201, + "learning_rate": 0.00024647083020216915, + "loss": 2.3416, + "step": 2962 + }, + { + "epoch": 0.6794313230910342, + "grad_norm": 0.40271654561877424, + "learning_rate": 0.0002461508321189706, + "loss": 2.2374, + "step": 2963 + }, + { + "epoch": 0.6796606282962623, + "grad_norm": 0.4289559658554842, + "learning_rate": 0.000245830974057275, + "loss": 2.3436, + "step": 2964 + }, + { + "epoch": 0.6798899335014905, + "grad_norm": 0.4170483585982145, + "learning_rate": 0.00024551125619351385, + "loss": 2.5108, + "step": 2965 + }, + { + "epoch": 0.6801192387067186, + "grad_norm": 0.43659506877033366, + "learning_rate": 0.00024519167870404123, + "loss": 2.4212, + "step": 2966 + }, + { + "epoch": 0.6803485439119468, + "grad_norm": 0.4143811832805458, + "learning_rate": 0.00024487224176513453, + "loss": 2.479, + "step": 2967 + }, + { + "epoch": 0.680577849117175, + "grad_norm": 0.4138974366731961, + "learning_rate": 0.00024455294555299217, + "loss": 2.3807, + "step": 2968 + }, + { + "epoch": 0.6808071543224031, + "grad_norm": 0.4275690304049134, + "learning_rate": 0.0002442337902437365, + "loss": 2.2668, + "step": 2969 + }, + { + "epoch": 0.6810364595276313, + "grad_norm": 0.3860706228249406, + "learning_rate": 0.000243914776013411, + "loss": 2.4256, + "step": 2970 + }, + { + "epoch": 0.6812657647328595, + "grad_norm": 0.42161785492245996, + "learning_rate": 0.0002435959030379824, + "loss": 2.5306, + "step": 2971 + }, + { + "epoch": 0.6814950699380876, + "grad_norm": 0.4067212246351068, + "learning_rate": 0.0002432771714933381, + "loss": 2.5706, + "step": 2972 + }, + { + "epoch": 0.6817243751433157, + "grad_norm": 0.40249019548752596, + "learning_rate": 0.00024295858155528888, + "loss": 2.3974, + "step": 2973 + }, + { + "epoch": 0.6819536803485439, + "grad_norm": 0.41192350429635105, + "learning_rate": 0.0002426401333995666, + "loss": 2.4899, + "step": 2974 + }, + { + "epoch": 0.6821829855537721, + "grad_norm": 0.37069895113896706, + "learning_rate": 0.00024232182720182523, + "loss": 2.4433, + "step": 2975 + }, + { + "epoch": 0.6824122907590002, + "grad_norm": 0.41989140986193685, + "learning_rate": 0.00024200366313764, + "loss": 2.526, + "step": 2976 + }, + { + "epoch": 0.6826415959642284, + "grad_norm": 0.3766652442584587, + "learning_rate": 0.00024168564138250855, + "loss": 2.5292, + "step": 2977 + }, + { + "epoch": 0.6828709011694566, + "grad_norm": 0.3937291725327558, + "learning_rate": 0.00024136776211184936, + "loss": 2.494, + "step": 2978 + }, + { + "epoch": 0.6831002063746847, + "grad_norm": 0.41708496416761603, + "learning_rate": 0.00024105002550100246, + "loss": 2.4554, + "step": 2979 + }, + { + "epoch": 0.6833295115799128, + "grad_norm": 0.5483558809633806, + "learning_rate": 0.00024073243172522912, + "loss": 2.3904, + "step": 2980 + }, + { + "epoch": 0.6835588167851411, + "grad_norm": 0.4254933645712717, + "learning_rate": 0.00024041498095971254, + "loss": 2.4468, + "step": 2981 + }, + { + "epoch": 0.6837881219903692, + "grad_norm": 0.40195105646549534, + "learning_rate": 0.00024009767337955553, + "loss": 2.4415, + "step": 2982 + }, + { + "epoch": 0.6840174271955973, + "grad_norm": 0.3889128512222097, + "learning_rate": 0.0002397805091597835, + "loss": 2.5034, + "step": 2983 + }, + { + "epoch": 0.6842467324008255, + "grad_norm": 0.39187727663872174, + "learning_rate": 0.00023946348847534194, + "loss": 2.3085, + "step": 2984 + }, + { + "epoch": 0.6844760376060537, + "grad_norm": 0.3691445536148744, + "learning_rate": 0.0002391466115010973, + "loss": 2.5045, + "step": 2985 + }, + { + "epoch": 0.6847053428112818, + "grad_norm": 0.5635047050308312, + "learning_rate": 0.00023882987841183657, + "loss": 2.5163, + "step": 2986 + }, + { + "epoch": 0.6849346480165099, + "grad_norm": 0.3947403266975219, + "learning_rate": 0.00023851328938226808, + "loss": 2.3298, + "step": 2987 + }, + { + "epoch": 0.6851639532217382, + "grad_norm": 0.40755118695928516, + "learning_rate": 0.00023819684458701996, + "loss": 2.4324, + "step": 2988 + }, + { + "epoch": 0.6853932584269663, + "grad_norm": 0.39788328822937075, + "learning_rate": 0.00023788054420064109, + "loss": 2.4648, + "step": 2989 + }, + { + "epoch": 0.6856225636321944, + "grad_norm": 0.40402196478889496, + "learning_rate": 0.0002375643883976004, + "loss": 2.4373, + "step": 2990 + }, + { + "epoch": 0.6858518688374227, + "grad_norm": 0.4332109555969562, + "learning_rate": 0.00023724837735228773, + "loss": 2.5326, + "step": 2991 + }, + { + "epoch": 0.6860811740426508, + "grad_norm": 0.8936903847881699, + "learning_rate": 0.00023693251123901244, + "loss": 2.3998, + "step": 2992 + }, + { + "epoch": 0.6863104792478789, + "grad_norm": 0.4058830729901206, + "learning_rate": 0.00023661679023200422, + "loss": 2.4002, + "step": 2993 + }, + { + "epoch": 0.686539784453107, + "grad_norm": 0.4052595670183673, + "learning_rate": 0.00023630121450541247, + "loss": 2.3196, + "step": 2994 + }, + { + "epoch": 0.6867690896583353, + "grad_norm": 0.43189059892040993, + "learning_rate": 0.00023598578423330714, + "loss": 2.4771, + "step": 2995 + }, + { + "epoch": 0.6869983948635634, + "grad_norm": 0.6157040216338957, + "learning_rate": 0.00023567049958967678, + "loss": 2.3363, + "step": 2996 + }, + { + "epoch": 0.6872277000687915, + "grad_norm": 0.43931680677148693, + "learning_rate": 0.00023535536074843083, + "loss": 2.3183, + "step": 2997 + }, + { + "epoch": 0.6874570052740198, + "grad_norm": 0.38912329908976356, + "learning_rate": 0.0002350403678833976, + "loss": 2.4612, + "step": 2998 + }, + { + "epoch": 0.6876863104792479, + "grad_norm": 0.3954547406649831, + "learning_rate": 0.00023472552116832502, + "loss": 2.4337, + "step": 2999 + }, + { + "epoch": 0.687915615684476, + "grad_norm": 0.4480955418188767, + "learning_rate": 0.00023441082077688025, + "loss": 2.4523, + "step": 3000 + }, + { + "epoch": 0.6881449208897042, + "grad_norm": 0.41601263749223505, + "learning_rate": 0.0002340962668826503, + "loss": 2.3874, + "step": 3001 + }, + { + "epoch": 0.6883742260949324, + "grad_norm": 0.4323565730594705, + "learning_rate": 0.0002337818596591408, + "loss": 2.4151, + "step": 3002 + }, + { + "epoch": 0.6886035313001605, + "grad_norm": 0.4601647078647517, + "learning_rate": 0.00023346759927977663, + "loss": 2.4937, + "step": 3003 + }, + { + "epoch": 0.6888328365053886, + "grad_norm": 0.45875902108433025, + "learning_rate": 0.00023315348591790164, + "loss": 2.4269, + "step": 3004 + }, + { + "epoch": 0.6890621417106169, + "grad_norm": 0.44517508131767575, + "learning_rate": 0.0002328395197467789, + "loss": 2.5071, + "step": 3005 + }, + { + "epoch": 0.689291446915845, + "grad_norm": 0.39584806992212357, + "learning_rate": 0.00023252570093958998, + "loss": 2.4223, + "step": 3006 + }, + { + "epoch": 0.6895207521210731, + "grad_norm": 0.37700058312419626, + "learning_rate": 0.00023221202966943515, + "loss": 2.381, + "step": 3007 + }, + { + "epoch": 0.6897500573263013, + "grad_norm": 0.40211315013230375, + "learning_rate": 0.00023189850610933328, + "loss": 2.38, + "step": 3008 + }, + { + "epoch": 0.6899793625315295, + "grad_norm": 0.4126233960899161, + "learning_rate": 0.0002315851304322223, + "loss": 2.498, + "step": 3009 + }, + { + "epoch": 0.6902086677367576, + "grad_norm": 0.39685035368410304, + "learning_rate": 0.0002312719028109575, + "loss": 2.5137, + "step": 3010 + }, + { + "epoch": 0.6904379729419858, + "grad_norm": 0.415527166252937, + "learning_rate": 0.0002309588234183137, + "loss": 2.5074, + "step": 3011 + }, + { + "epoch": 0.690667278147214, + "grad_norm": 0.40524822847019354, + "learning_rate": 0.00023064589242698314, + "loss": 2.3914, + "step": 3012 + }, + { + "epoch": 0.6908965833524421, + "grad_norm": 0.42886441306675, + "learning_rate": 0.00023033311000957653, + "loss": 2.4794, + "step": 3013 + }, + { + "epoch": 0.6911258885576702, + "grad_norm": 0.3691091135582987, + "learning_rate": 0.00023002047633862232, + "loss": 2.2923, + "step": 3014 + }, + { + "epoch": 0.6913551937628984, + "grad_norm": 0.4197564465583856, + "learning_rate": 0.00022970799158656758, + "loss": 2.4547, + "step": 3015 + }, + { + "epoch": 0.6915844989681266, + "grad_norm": 0.382562692177307, + "learning_rate": 0.0002293956559257766, + "loss": 2.3729, + "step": 3016 + }, + { + "epoch": 0.6918138041733547, + "grad_norm": 0.4348636455187477, + "learning_rate": 0.0002290834695285316, + "loss": 2.5053, + "step": 3017 + }, + { + "epoch": 0.6920431093785829, + "grad_norm": 0.40778900506488547, + "learning_rate": 0.00022877143256703236, + "loss": 2.5688, + "step": 3018 + }, + { + "epoch": 0.6922724145838111, + "grad_norm": 0.4036419665656172, + "learning_rate": 0.00022845954521339678, + "loss": 2.4589, + "step": 3019 + }, + { + "epoch": 0.6925017197890392, + "grad_norm": 0.3952048249835875, + "learning_rate": 0.0002281478076396596, + "loss": 2.4333, + "step": 3020 + }, + { + "epoch": 0.6927310249942674, + "grad_norm": 0.39512425206444923, + "learning_rate": 0.0002278362200177732, + "loss": 2.4547, + "step": 3021 + }, + { + "epoch": 0.6929603301994955, + "grad_norm": 0.3967962261395788, + "learning_rate": 0.000227524782519607, + "loss": 2.4206, + "step": 3022 + }, + { + "epoch": 0.6931896354047237, + "grad_norm": 0.40261959665156505, + "learning_rate": 0.00022721349531694852, + "loss": 2.4612, + "step": 3023 + }, + { + "epoch": 0.6934189406099518, + "grad_norm": 0.3749434594973393, + "learning_rate": 0.00022690235858150083, + "loss": 2.4578, + "step": 3024 + }, + { + "epoch": 0.69364824581518, + "grad_norm": 0.41484315887123935, + "learning_rate": 0.0002265913724848855, + "loss": 2.4431, + "step": 3025 + }, + { + "epoch": 0.6938775510204082, + "grad_norm": 0.4116267996679295, + "learning_rate": 0.00022628053719864018, + "loss": 2.509, + "step": 3026 + }, + { + "epoch": 0.6941068562256363, + "grad_norm": 0.40229891532722695, + "learning_rate": 0.00022596985289421946, + "loss": 2.2846, + "step": 3027 + }, + { + "epoch": 0.6943361614308645, + "grad_norm": 0.38981476249789865, + "learning_rate": 0.00022565931974299463, + "loss": 2.334, + "step": 3028 + }, + { + "epoch": 0.6945654666360926, + "grad_norm": 0.39565240719731315, + "learning_rate": 0.00022534893791625405, + "loss": 2.485, + "step": 3029 + }, + { + "epoch": 0.6947947718413208, + "grad_norm": 0.38893190922016263, + "learning_rate": 0.00022503870758520212, + "loss": 2.2379, + "step": 3030 + }, + { + "epoch": 0.695024077046549, + "grad_norm": 0.4424049294782965, + "learning_rate": 0.00022472862892095968, + "loss": 2.5077, + "step": 3031 + }, + { + "epoch": 0.6952533822517771, + "grad_norm": 0.42594929736436976, + "learning_rate": 0.000224418702094564, + "loss": 2.5624, + "step": 3032 + }, + { + "epoch": 0.6954826874570053, + "grad_norm": 0.41618165296379, + "learning_rate": 0.00022410892727696896, + "loss": 2.5045, + "step": 3033 + }, + { + "epoch": 0.6957119926622334, + "grad_norm": 0.4614704165154477, + "learning_rate": 0.00022379930463904408, + "loss": 2.4358, + "step": 3034 + }, + { + "epoch": 0.6959412978674616, + "grad_norm": 0.405935920819258, + "learning_rate": 0.0002234898343515751, + "loss": 2.3621, + "step": 3035 + }, + { + "epoch": 0.6961706030726897, + "grad_norm": 0.41145292316780147, + "learning_rate": 0.00022318051658526368, + "loss": 2.4847, + "step": 3036 + }, + { + "epoch": 0.6963999082779179, + "grad_norm": 0.48640966335705815, + "learning_rate": 0.00022287135151072792, + "loss": 2.4205, + "step": 3037 + }, + { + "epoch": 0.6966292134831461, + "grad_norm": 0.4149909184917952, + "learning_rate": 0.00022256233929850044, + "loss": 2.3366, + "step": 3038 + }, + { + "epoch": 0.6968585186883742, + "grad_norm": 0.4212946387796949, + "learning_rate": 0.00022225348011903096, + "loss": 2.5476, + "step": 3039 + }, + { + "epoch": 0.6970878238936024, + "grad_norm": 0.41395704587239873, + "learning_rate": 0.00022194477414268384, + "loss": 2.4408, + "step": 3040 + }, + { + "epoch": 0.6973171290988306, + "grad_norm": 0.46998896719612354, + "learning_rate": 0.0002216362215397393, + "loss": 2.4866, + "step": 3041 + }, + { + "epoch": 0.6975464343040587, + "grad_norm": 0.40698865818001595, + "learning_rate": 0.0002213278224803927, + "loss": 2.348, + "step": 3042 + }, + { + "epoch": 0.6977757395092868, + "grad_norm": 0.3888170463300552, + "learning_rate": 0.00022101957713475522, + "loss": 2.3139, + "step": 3043 + }, + { + "epoch": 0.698005044714515, + "grad_norm": 0.4002376026150465, + "learning_rate": 0.00022071148567285283, + "loss": 2.5085, + "step": 3044 + }, + { + "epoch": 0.6982343499197432, + "grad_norm": 0.4322864957493867, + "learning_rate": 0.00022040354826462666, + "loss": 2.4764, + "step": 3045 + }, + { + "epoch": 0.6984636551249713, + "grad_norm": 0.4026342554470211, + "learning_rate": 0.00022009576507993274, + "loss": 2.4183, + "step": 3046 + }, + { + "epoch": 0.6986929603301995, + "grad_norm": 0.4029881383497249, + "learning_rate": 0.0002197881362885426, + "loss": 2.4204, + "step": 3047 + }, + { + "epoch": 0.6989222655354277, + "grad_norm": 0.4197351125536147, + "learning_rate": 0.00021948066206014207, + "loss": 2.6134, + "step": 3048 + }, + { + "epoch": 0.6991515707406558, + "grad_norm": 0.3652880077606984, + "learning_rate": 0.0002191733425643318, + "loss": 2.287, + "step": 3049 + }, + { + "epoch": 0.6993808759458839, + "grad_norm": 0.4225420282624902, + "learning_rate": 0.00021886617797062702, + "loss": 2.4354, + "step": 3050 + }, + { + "epoch": 0.6996101811511122, + "grad_norm": 0.412990896785503, + "learning_rate": 0.00021855916844845826, + "loss": 2.4251, + "step": 3051 + }, + { + "epoch": 0.6998394863563403, + "grad_norm": 0.7013233020886593, + "learning_rate": 0.0002182523141671691, + "loss": 2.3807, + "step": 3052 + }, + { + "epoch": 0.7000687915615684, + "grad_norm": 0.4049168940924462, + "learning_rate": 0.00021794561529601898, + "loss": 2.2913, + "step": 3053 + }, + { + "epoch": 0.7002980967667966, + "grad_norm": 0.41237067483353906, + "learning_rate": 0.00021763907200418066, + "loss": 2.3522, + "step": 3054 + }, + { + "epoch": 0.7005274019720248, + "grad_norm": 0.42247810641827177, + "learning_rate": 0.00021733268446074138, + "loss": 2.4442, + "step": 3055 + }, + { + "epoch": 0.7007567071772529, + "grad_norm": 0.398137956802751, + "learning_rate": 0.00021702645283470234, + "loss": 2.298, + "step": 3056 + }, + { + "epoch": 0.700986012382481, + "grad_norm": 0.373861685583105, + "learning_rate": 0.00021672037729497917, + "loss": 2.5182, + "step": 3057 + }, + { + "epoch": 0.7012153175877093, + "grad_norm": 0.4042908759398654, + "learning_rate": 0.00021641445801040088, + "loss": 2.3704, + "step": 3058 + }, + { + "epoch": 0.7014446227929374, + "grad_norm": 0.40068206876277496, + "learning_rate": 0.0002161086951497106, + "loss": 2.3817, + "step": 3059 + }, + { + "epoch": 0.7016739279981655, + "grad_norm": 0.4388874208992351, + "learning_rate": 0.00021580308888156496, + "loss": 2.5365, + "step": 3060 + }, + { + "epoch": 0.7019032332033938, + "grad_norm": 0.39458415623154275, + "learning_rate": 0.00021549763937453442, + "loss": 2.3648, + "step": 3061 + }, + { + "epoch": 0.7021325384086219, + "grad_norm": 0.4160829901095841, + "learning_rate": 0.00021519234679710282, + "loss": 2.4218, + "step": 3062 + }, + { + "epoch": 0.70236184361385, + "grad_norm": 0.4844083514768985, + "learning_rate": 0.00021488721131766736, + "loss": 2.5565, + "step": 3063 + }, + { + "epoch": 0.7025911488190782, + "grad_norm": 0.4126383562752624, + "learning_rate": 0.00021458223310453912, + "loss": 2.5014, + "step": 3064 + }, + { + "epoch": 0.7028204540243064, + "grad_norm": 0.3919910614968368, + "learning_rate": 0.00021427741232594183, + "loss": 2.5692, + "step": 3065 + }, + { + "epoch": 0.7030497592295345, + "grad_norm": 0.4095651241712313, + "learning_rate": 0.00021397274915001252, + "loss": 2.4022, + "step": 3066 + }, + { + "epoch": 0.7032790644347626, + "grad_norm": 0.4405646848291948, + "learning_rate": 0.0002136682437448013, + "loss": 2.4818, + "step": 3067 + }, + { + "epoch": 0.7035083696399909, + "grad_norm": 0.410707147554731, + "learning_rate": 0.00021336389627827168, + "loss": 2.6146, + "step": 3068 + }, + { + "epoch": 0.703737674845219, + "grad_norm": 0.4142805809653468, + "learning_rate": 0.0002130597069182994, + "loss": 2.5441, + "step": 3069 + }, + { + "epoch": 0.7039669800504471, + "grad_norm": 0.3756293404407109, + "learning_rate": 0.0002127556758326733, + "loss": 2.4687, + "step": 3070 + }, + { + "epoch": 0.7041962852556753, + "grad_norm": 0.4473054068166196, + "learning_rate": 0.0002124518031890948, + "loss": 2.4416, + "step": 3071 + }, + { + "epoch": 0.7044255904609035, + "grad_norm": 0.4230665832915528, + "learning_rate": 0.00021214808915517847, + "loss": 2.4252, + "step": 3072 + }, + { + "epoch": 0.7046548956661316, + "grad_norm": 0.3983709553945093, + "learning_rate": 0.0002118445338984502, + "loss": 2.2984, + "step": 3073 + }, + { + "epoch": 0.7048842008713598, + "grad_norm": 0.4001352118091537, + "learning_rate": 0.0002115411375863497, + "loss": 2.2554, + "step": 3074 + }, + { + "epoch": 0.705113506076588, + "grad_norm": 0.394550725241349, + "learning_rate": 0.00021123790038622808, + "loss": 2.4532, + "step": 3075 + }, + { + "epoch": 0.7053428112818161, + "grad_norm": 0.41081266701130426, + "learning_rate": 0.00021093482246534894, + "loss": 2.3469, + "step": 3076 + }, + { + "epoch": 0.7055721164870442, + "grad_norm": 0.39001066477643676, + "learning_rate": 0.0002106319039908879, + "loss": 2.4359, + "step": 3077 + }, + { + "epoch": 0.7058014216922724, + "grad_norm": 0.443711575910713, + "learning_rate": 0.00021032914512993312, + "loss": 2.5146, + "step": 3078 + }, + { + "epoch": 0.7060307268975006, + "grad_norm": 0.409998107587763, + "learning_rate": 0.00021002654604948412, + "loss": 2.4532, + "step": 3079 + }, + { + "epoch": 0.7062600321027287, + "grad_norm": 0.44234145332866154, + "learning_rate": 0.0002097241069164526, + "loss": 2.4009, + "step": 3080 + }, + { + "epoch": 0.7064893373079569, + "grad_norm": 0.40313636361057664, + "learning_rate": 0.00020942182789766172, + "loss": 2.5072, + "step": 3081 + }, + { + "epoch": 0.7067186425131851, + "grad_norm": 0.4158593704950678, + "learning_rate": 0.00020911970915984702, + "loss": 2.4415, + "step": 3082 + }, + { + "epoch": 0.7069479477184132, + "grad_norm": 0.410611924627721, + "learning_rate": 0.00020881775086965492, + "loss": 2.3401, + "step": 3083 + }, + { + "epoch": 0.7071772529236414, + "grad_norm": 0.47574789264065087, + "learning_rate": 0.00020851595319364358, + "loss": 2.5136, + "step": 3084 + }, + { + "epoch": 0.7074065581288695, + "grad_norm": 0.4112093734034053, + "learning_rate": 0.00020821431629828246, + "loss": 2.4566, + "step": 3085 + }, + { + "epoch": 0.7076358633340977, + "grad_norm": 0.4406733899282893, + "learning_rate": 0.00020791284034995294, + "loss": 2.4818, + "step": 3086 + }, + { + "epoch": 0.7078651685393258, + "grad_norm": 0.43008843476047404, + "learning_rate": 0.00020761152551494643, + "loss": 2.4677, + "step": 3087 + }, + { + "epoch": 0.708094473744554, + "grad_norm": 0.4415755375627261, + "learning_rate": 0.00020731037195946677, + "loss": 2.4628, + "step": 3088 + }, + { + "epoch": 0.7083237789497822, + "grad_norm": 0.4400306061625708, + "learning_rate": 0.00020700937984962798, + "loss": 2.5226, + "step": 3089 + }, + { + "epoch": 0.7085530841550103, + "grad_norm": 0.4134885527901022, + "learning_rate": 0.00020670854935145533, + "loss": 2.5083, + "step": 3090 + }, + { + "epoch": 0.7087823893602385, + "grad_norm": 0.41151954643196975, + "learning_rate": 0.0002064078806308848, + "loss": 2.4181, + "step": 3091 + }, + { + "epoch": 0.7090116945654666, + "grad_norm": 0.38309638568950116, + "learning_rate": 0.00020610737385376348, + "loss": 2.3398, + "step": 3092 + }, + { + "epoch": 0.7092409997706948, + "grad_norm": 0.3926103377967803, + "learning_rate": 0.00020580702918584882, + "loss": 2.386, + "step": 3093 + }, + { + "epoch": 0.709470304975923, + "grad_norm": 0.3906686369560157, + "learning_rate": 0.00020550684679280885, + "loss": 2.4577, + "step": 3094 + }, + { + "epoch": 0.7096996101811511, + "grad_norm": 0.3853183906569975, + "learning_rate": 0.000205206826840222, + "loss": 2.4531, + "step": 3095 + }, + { + "epoch": 0.7099289153863793, + "grad_norm": 0.41138802903367727, + "learning_rate": 0.00020490696949357774, + "loss": 2.4967, + "step": 3096 + }, + { + "epoch": 0.7101582205916074, + "grad_norm": 0.43747421817136595, + "learning_rate": 0.0002046072749182751, + "loss": 2.4536, + "step": 3097 + }, + { + "epoch": 0.7103875257968356, + "grad_norm": 0.4302065909654247, + "learning_rate": 0.0002043077432796236, + "loss": 2.425, + "step": 3098 + }, + { + "epoch": 0.7106168310020637, + "grad_norm": 0.42476943681547424, + "learning_rate": 0.00020400837474284273, + "loss": 2.584, + "step": 3099 + }, + { + "epoch": 0.7108461362072919, + "grad_norm": 0.3992233076819081, + "learning_rate": 0.00020370916947306267, + "loss": 2.376, + "step": 3100 + }, + { + "epoch": 0.7110754414125201, + "grad_norm": 0.4036219382005645, + "learning_rate": 0.0002034101276353224, + "loss": 2.4798, + "step": 3101 + }, + { + "epoch": 0.7113047466177482, + "grad_norm": 0.46572177136856996, + "learning_rate": 0.00020311124939457193, + "loss": 2.4945, + "step": 3102 + }, + { + "epoch": 0.7115340518229764, + "grad_norm": 0.6792710147274534, + "learning_rate": 0.00020281253491567027, + "loss": 2.4127, + "step": 3103 + }, + { + "epoch": 0.7117633570282046, + "grad_norm": 0.42320335229230405, + "learning_rate": 0.0002025139843633863, + "loss": 2.5307, + "step": 3104 + }, + { + "epoch": 0.7119926622334327, + "grad_norm": 0.394741280930373, + "learning_rate": 0.0002022155979023984, + "loss": 2.4042, + "step": 3105 + }, + { + "epoch": 0.7122219674386608, + "grad_norm": 0.4104606315340626, + "learning_rate": 0.00020191737569729491, + "loss": 2.4752, + "step": 3106 + }, + { + "epoch": 0.712451272643889, + "grad_norm": 0.4505703308716284, + "learning_rate": 0.000201619317912573, + "loss": 2.5973, + "step": 3107 + }, + { + "epoch": 0.7126805778491172, + "grad_norm": 0.40916553540365724, + "learning_rate": 0.00020132142471263932, + "loss": 2.3703, + "step": 3108 + }, + { + "epoch": 0.7129098830543453, + "grad_norm": 0.3976211989619038, + "learning_rate": 0.00020102369626180962, + "loss": 2.4988, + "step": 3109 + }, + { + "epoch": 0.7131391882595735, + "grad_norm": 0.4350735481272384, + "learning_rate": 0.00020072613272430922, + "loss": 2.4264, + "step": 3110 + }, + { + "epoch": 0.7133684934648017, + "grad_norm": 0.4217369110174488, + "learning_rate": 0.0002004287342642721, + "loss": 2.4442, + "step": 3111 + }, + { + "epoch": 0.7135977986700298, + "grad_norm": 0.3557263621039467, + "learning_rate": 0.00020013150104574123, + "loss": 2.3857, + "step": 3112 + }, + { + "epoch": 0.7138271038752579, + "grad_norm": 0.3971127094592139, + "learning_rate": 0.00019983443323266824, + "loss": 2.3681, + "step": 3113 + }, + { + "epoch": 0.7140564090804862, + "grad_norm": 0.3968575330796799, + "learning_rate": 0.0001995375309889143, + "loss": 2.4422, + "step": 3114 + }, + { + "epoch": 0.7142857142857143, + "grad_norm": 0.41177156045493685, + "learning_rate": 0.00019924079447824805, + "loss": 2.5024, + "step": 3115 + }, + { + "epoch": 0.7145150194909424, + "grad_norm": 0.4135055774142241, + "learning_rate": 0.0001989442238643478, + "loss": 2.4491, + "step": 3116 + }, + { + "epoch": 0.7147443246961706, + "grad_norm": 0.4297016015471969, + "learning_rate": 0.00019864781931079977, + "loss": 2.4305, + "step": 3117 + }, + { + "epoch": 0.7149736299013988, + "grad_norm": 0.4117646512060056, + "learning_rate": 0.0001983515809810988, + "loss": 2.3891, + "step": 3118 + }, + { + "epoch": 0.7152029351066269, + "grad_norm": 0.38972945988546187, + "learning_rate": 0.00019805550903864773, + "loss": 2.5946, + "step": 3119 + }, + { + "epoch": 0.715432240311855, + "grad_norm": 0.3983780454550496, + "learning_rate": 0.00019775960364675822, + "loss": 2.4494, + "step": 3120 + }, + { + "epoch": 0.7156615455170833, + "grad_norm": 0.38116342001442377, + "learning_rate": 0.00019746386496864948, + "loss": 2.2753, + "step": 3121 + }, + { + "epoch": 0.7158908507223114, + "grad_norm": 0.3799134031867374, + "learning_rate": 0.00019716829316744912, + "loss": 2.5266, + "step": 3122 + }, + { + "epoch": 0.7161201559275395, + "grad_norm": 0.3970701562475162, + "learning_rate": 0.00019687288840619226, + "loss": 2.34, + "step": 3123 + }, + { + "epoch": 0.7163494611327678, + "grad_norm": 0.40202991563786905, + "learning_rate": 0.00019657765084782282, + "loss": 2.4709, + "step": 3124 + }, + { + "epoch": 0.7165787663379959, + "grad_norm": 0.3897852121442939, + "learning_rate": 0.0001962825806551911, + "loss": 2.5113, + "step": 3125 + }, + { + "epoch": 0.716808071543224, + "grad_norm": 0.3870559157263271, + "learning_rate": 0.00019598767799105637, + "loss": 2.3824, + "step": 3126 + }, + { + "epoch": 0.7170373767484521, + "grad_norm": 0.4146693691492978, + "learning_rate": 0.0001956929430180846, + "loss": 2.4609, + "step": 3127 + }, + { + "epoch": 0.7172666819536804, + "grad_norm": 0.4402103209880072, + "learning_rate": 0.00019539837589885024, + "loss": 2.4832, + "step": 3128 + }, + { + "epoch": 0.7174959871589085, + "grad_norm": 0.39929597819907864, + "learning_rate": 0.00019510397679583374, + "loss": 2.4297, + "step": 3129 + }, + { + "epoch": 0.7177252923641366, + "grad_norm": 0.38673588917206064, + "learning_rate": 0.00019480974587142425, + "loss": 2.4937, + "step": 3130 + }, + { + "epoch": 0.7179545975693649, + "grad_norm": 0.4412370876880804, + "learning_rate": 0.0001945156832879174, + "loss": 2.4184, + "step": 3131 + }, + { + "epoch": 0.718183902774593, + "grad_norm": 0.4199448631961088, + "learning_rate": 0.00019422178920751616, + "loss": 2.5397, + "step": 3132 + }, + { + "epoch": 0.7184132079798211, + "grad_norm": 0.4077330000141336, + "learning_rate": 0.00019392806379233036, + "loss": 2.2653, + "step": 3133 + }, + { + "epoch": 0.7186425131850493, + "grad_norm": 0.4076731046171923, + "learning_rate": 0.0001936345072043773, + "loss": 2.3169, + "step": 3134 + }, + { + "epoch": 0.7188718183902775, + "grad_norm": 0.44841908276654885, + "learning_rate": 0.00019334111960558065, + "loss": 2.5434, + "step": 3135 + }, + { + "epoch": 0.7191011235955056, + "grad_norm": 0.39928162489159014, + "learning_rate": 0.0001930479011577711, + "loss": 2.4221, + "step": 3136 + }, + { + "epoch": 0.7193304288007337, + "grad_norm": 0.40488666189985445, + "learning_rate": 0.00019275485202268573, + "loss": 2.5835, + "step": 3137 + }, + { + "epoch": 0.719559734005962, + "grad_norm": 0.41411173929445344, + "learning_rate": 0.0001924619723619691, + "loss": 2.4283, + "step": 3138 + }, + { + "epoch": 0.7197890392111901, + "grad_norm": 0.42332296256012975, + "learning_rate": 0.00019216926233717085, + "loss": 2.3799, + "step": 3139 + }, + { + "epoch": 0.7200183444164182, + "grad_norm": 0.42743567906207963, + "learning_rate": 0.00019187672210974855, + "loss": 2.4719, + "step": 3140 + }, + { + "epoch": 0.7202476496216464, + "grad_norm": 0.42419499364761015, + "learning_rate": 0.00019158435184106498, + "loss": 2.4701, + "step": 3141 + }, + { + "epoch": 0.7204769548268746, + "grad_norm": 0.42221978832871015, + "learning_rate": 0.00019129215169239017, + "loss": 2.4473, + "step": 3142 + }, + { + "epoch": 0.7207062600321027, + "grad_norm": 0.4750421289496919, + "learning_rate": 0.00019100012182489905, + "loss": 2.5379, + "step": 3143 + }, + { + "epoch": 0.7209355652373309, + "grad_norm": 0.41908001249936727, + "learning_rate": 0.00019070826239967387, + "loss": 2.312, + "step": 3144 + }, + { + "epoch": 0.7211648704425591, + "grad_norm": 0.4235332068514487, + "learning_rate": 0.00019041657357770226, + "loss": 2.4584, + "step": 3145 + }, + { + "epoch": 0.7213941756477872, + "grad_norm": 0.3706895615496749, + "learning_rate": 0.00019012505551987762, + "loss": 2.3502, + "step": 3146 + }, + { + "epoch": 0.7216234808530153, + "grad_norm": 0.386150029116593, + "learning_rate": 0.00018983370838699943, + "loss": 2.4089, + "step": 3147 + }, + { + "epoch": 0.7218527860582435, + "grad_norm": 0.4132693192162246, + "learning_rate": 0.00018954253233977297, + "loss": 2.4127, + "step": 3148 + }, + { + "epoch": 0.7220820912634717, + "grad_norm": 0.4351030481279718, + "learning_rate": 0.00018925152753880892, + "loss": 2.5332, + "step": 3149 + }, + { + "epoch": 0.7223113964686998, + "grad_norm": 0.41594707656027535, + "learning_rate": 0.00018896069414462357, + "loss": 2.3687, + "step": 3150 + }, + { + "epoch": 0.722540701673928, + "grad_norm": 0.40776627977581126, + "learning_rate": 0.00018867003231763847, + "loss": 2.5124, + "step": 3151 + }, + { + "epoch": 0.7227700068791562, + "grad_norm": 0.3867723641850582, + "learning_rate": 0.00018837954221818132, + "loss": 2.5184, + "step": 3152 + }, + { + "epoch": 0.7229993120843843, + "grad_norm": 0.42344928672317855, + "learning_rate": 0.00018808922400648375, + "loss": 2.4558, + "step": 3153 + }, + { + "epoch": 0.7232286172896125, + "grad_norm": 0.4099948067173013, + "learning_rate": 0.0001877990778426839, + "loss": 2.6339, + "step": 3154 + }, + { + "epoch": 0.7234579224948406, + "grad_norm": 0.41593905621607147, + "learning_rate": 0.00018750910388682428, + "loss": 2.5058, + "step": 3155 + }, + { + "epoch": 0.7236872277000688, + "grad_norm": 0.3973721400267247, + "learning_rate": 0.0001872193022988526, + "loss": 2.4778, + "step": 3156 + }, + { + "epoch": 0.723916532905297, + "grad_norm": 0.3958860292525971, + "learning_rate": 0.00018692967323862125, + "loss": 2.3728, + "step": 3157 + }, + { + "epoch": 0.7241458381105251, + "grad_norm": 0.41691497337885, + "learning_rate": 0.0001866402168658881, + "loss": 2.4452, + "step": 3158 + }, + { + "epoch": 0.7243751433157533, + "grad_norm": 0.383058571217824, + "learning_rate": 0.00018635093334031517, + "loss": 2.3597, + "step": 3159 + }, + { + "epoch": 0.7246044485209814, + "grad_norm": 0.41419285912186954, + "learning_rate": 0.00018606182282146922, + "loss": 2.5216, + "step": 3160 + }, + { + "epoch": 0.7248337537262096, + "grad_norm": 0.4328825353400251, + "learning_rate": 0.00018577288546882165, + "loss": 2.4258, + "step": 3161 + }, + { + "epoch": 0.7250630589314377, + "grad_norm": 0.4215760045024753, + "learning_rate": 0.00018548412144174859, + "loss": 2.3758, + "step": 3162 + }, + { + "epoch": 0.7252923641366659, + "grad_norm": 0.39973667131428214, + "learning_rate": 0.00018519553089953023, + "loss": 2.3484, + "step": 3163 + }, + { + "epoch": 0.7255216693418941, + "grad_norm": 0.435796611353287, + "learning_rate": 0.0001849071140013512, + "loss": 2.468, + "step": 3164 + }, + { + "epoch": 0.7257509745471222, + "grad_norm": 0.4009757581964912, + "learning_rate": 0.0001846188709063001, + "loss": 2.3301, + "step": 3165 + }, + { + "epoch": 0.7259802797523504, + "grad_norm": 0.40282098366268076, + "learning_rate": 0.00018433080177337043, + "loss": 2.366, + "step": 3166 + }, + { + "epoch": 0.7262095849575786, + "grad_norm": 0.3946365790949136, + "learning_rate": 0.00018404290676145857, + "loss": 2.4309, + "step": 3167 + }, + { + "epoch": 0.7264388901628067, + "grad_norm": 0.4064977323712654, + "learning_rate": 0.00018375518602936592, + "loss": 2.4243, + "step": 3168 + }, + { + "epoch": 0.7266681953680348, + "grad_norm": 0.5776392537920116, + "learning_rate": 0.00018346763973579722, + "loss": 2.5011, + "step": 3169 + }, + { + "epoch": 0.726897500573263, + "grad_norm": 0.4274853104083807, + "learning_rate": 0.000183180268039361, + "loss": 2.3077, + "step": 3170 + }, + { + "epoch": 0.7271268057784912, + "grad_norm": 0.4186084387214672, + "learning_rate": 0.00018289307109856939, + "loss": 2.5581, + "step": 3171 + }, + { + "epoch": 0.7273561109837193, + "grad_norm": 0.4231215973230345, + "learning_rate": 0.00018260604907183875, + "loss": 2.535, + "step": 3172 + }, + { + "epoch": 0.7275854161889475, + "grad_norm": 0.4334382627116434, + "learning_rate": 0.0001823192021174882, + "loss": 2.3629, + "step": 3173 + }, + { + "epoch": 0.7278147213941757, + "grad_norm": 0.41501053324522086, + "learning_rate": 0.00018203253039374075, + "loss": 2.4561, + "step": 3174 + }, + { + "epoch": 0.7280440265994038, + "grad_norm": 0.3870574813825533, + "learning_rate": 0.0001817460340587223, + "loss": 2.5093, + "step": 3175 + }, + { + "epoch": 0.7282733318046319, + "grad_norm": 0.3980975988502954, + "learning_rate": 0.00018145971327046274, + "loss": 2.4056, + "step": 3176 + }, + { + "epoch": 0.7285026370098602, + "grad_norm": 0.4641955023947692, + "learning_rate": 0.00018117356818689445, + "loss": 2.4728, + "step": 3177 + }, + { + "epoch": 0.7287319422150883, + "grad_norm": 0.4257068654913183, + "learning_rate": 0.00018088759896585323, + "loss": 2.5204, + "step": 3178 + }, + { + "epoch": 0.7289612474203164, + "grad_norm": 0.37737671306480725, + "learning_rate": 0.00018060180576507756, + "loss": 2.4007, + "step": 3179 + }, + { + "epoch": 0.7291905526255446, + "grad_norm": 0.5087372727548889, + "learning_rate": 0.00018031618874220963, + "loss": 2.4617, + "step": 3180 + }, + { + "epoch": 0.7294198578307728, + "grad_norm": 0.42081472040208706, + "learning_rate": 0.00018003074805479313, + "loss": 2.3928, + "step": 3181 + }, + { + "epoch": 0.7296491630360009, + "grad_norm": 0.4655489680770466, + "learning_rate": 0.00017974548386027583, + "loss": 2.2924, + "step": 3182 + }, + { + "epoch": 0.729878468241229, + "grad_norm": 0.42001637504389266, + "learning_rate": 0.00017946039631600724, + "loss": 2.4942, + "step": 3183 + }, + { + "epoch": 0.7301077734464573, + "grad_norm": 0.39985987544172874, + "learning_rate": 0.0001791754855792399, + "loss": 2.3769, + "step": 3184 + }, + { + "epoch": 0.7303370786516854, + "grad_norm": 0.4130238609257555, + "learning_rate": 0.00017889075180712837, + "loss": 2.4949, + "step": 3185 + }, + { + "epoch": 0.7305663838569135, + "grad_norm": 0.42039703913959503, + "learning_rate": 0.0001786061951567303, + "loss": 2.4361, + "step": 3186 + }, + { + "epoch": 0.7307956890621418, + "grad_norm": 0.4402163178360052, + "learning_rate": 0.00017832181578500512, + "loss": 2.4577, + "step": 3187 + }, + { + "epoch": 0.7310249942673699, + "grad_norm": 0.4003605124061562, + "learning_rate": 0.00017803761384881445, + "loss": 2.4354, + "step": 3188 + }, + { + "epoch": 0.731254299472598, + "grad_norm": 0.44308068588143695, + "learning_rate": 0.0001777535895049221, + "loss": 2.4872, + "step": 3189 + }, + { + "epoch": 0.7314836046778261, + "grad_norm": 0.405425009978453, + "learning_rate": 0.0001774697429099943, + "loss": 2.2565, + "step": 3190 + }, + { + "epoch": 0.7317129098830544, + "grad_norm": 0.5235346625503996, + "learning_rate": 0.0001771860742205988, + "loss": 2.4661, + "step": 3191 + }, + { + "epoch": 0.7319422150882825, + "grad_norm": 0.4168933253517395, + "learning_rate": 0.00017690258359320533, + "loss": 2.5104, + "step": 3192 + }, + { + "epoch": 0.7321715202935106, + "grad_norm": 0.4068219953881811, + "learning_rate": 0.00017661927118418525, + "loss": 2.3974, + "step": 3193 + }, + { + "epoch": 0.7324008254987389, + "grad_norm": 0.3917869693491349, + "learning_rate": 0.00017633613714981234, + "loss": 2.5775, + "step": 3194 + }, + { + "epoch": 0.732630130703967, + "grad_norm": 0.4727401600997679, + "learning_rate": 0.00017605318164626066, + "loss": 2.5291, + "step": 3195 + }, + { + "epoch": 0.7328594359091951, + "grad_norm": 0.42419106897637704, + "learning_rate": 0.0001757704048296072, + "loss": 2.5228, + "step": 3196 + }, + { + "epoch": 0.7330887411144233, + "grad_norm": 0.408684478311777, + "learning_rate": 0.00017548780685582949, + "loss": 2.393, + "step": 3197 + }, + { + "epoch": 0.7333180463196515, + "grad_norm": 0.4164274223023187, + "learning_rate": 0.00017520538788080677, + "loss": 2.4895, + "step": 3198 + }, + { + "epoch": 0.7335473515248796, + "grad_norm": 0.40054511590512476, + "learning_rate": 0.00017492314806031922, + "loss": 2.5502, + "step": 3199 + }, + { + "epoch": 0.7337766567301077, + "grad_norm": 0.3995885682013317, + "learning_rate": 0.00017464108755004882, + "loss": 2.5149, + "step": 3200 + }, + { + "epoch": 0.734005961935336, + "grad_norm": 0.3941007821355203, + "learning_rate": 0.00017435920650557806, + "loss": 2.4007, + "step": 3201 + }, + { + "epoch": 0.7342352671405641, + "grad_norm": 0.40547605201980885, + "learning_rate": 0.00017407750508239067, + "loss": 2.4201, + "step": 3202 + }, + { + "epoch": 0.7344645723457922, + "grad_norm": 0.3968623540520974, + "learning_rate": 0.00017379598343587112, + "loss": 2.4134, + "step": 3203 + }, + { + "epoch": 0.7346938775510204, + "grad_norm": 0.43713911951062495, + "learning_rate": 0.0001735146417213051, + "loss": 2.3793, + "step": 3204 + }, + { + "epoch": 0.7349231827562486, + "grad_norm": 0.40986133260673696, + "learning_rate": 0.00017323348009387878, + "loss": 2.2943, + "step": 3205 + }, + { + "epoch": 0.7351524879614767, + "grad_norm": 0.40277988227581835, + "learning_rate": 0.00017295249870867896, + "loss": 2.449, + "step": 3206 + }, + { + "epoch": 0.7353817931667049, + "grad_norm": 0.4164028331546371, + "learning_rate": 0.0001726716977206929, + "loss": 2.3378, + "step": 3207 + }, + { + "epoch": 0.7356110983719331, + "grad_norm": 0.488843379028829, + "learning_rate": 0.00017239107728480912, + "loss": 2.4595, + "step": 3208 + }, + { + "epoch": 0.7358404035771612, + "grad_norm": 0.4510040041305662, + "learning_rate": 0.00017211063755581525, + "loss": 2.3302, + "step": 3209 + }, + { + "epoch": 0.7360697087823893, + "grad_norm": 0.42382572175031286, + "learning_rate": 0.0001718303786884005, + "loss": 2.443, + "step": 3210 + }, + { + "epoch": 0.7362990139876175, + "grad_norm": 0.484697921722537, + "learning_rate": 0.0001715503008371536, + "loss": 2.5092, + "step": 3211 + }, + { + "epoch": 0.7365283191928457, + "grad_norm": 0.4482761502911404, + "learning_rate": 0.00017127040415656364, + "loss": 2.3662, + "step": 3212 + }, + { + "epoch": 0.7367576243980738, + "grad_norm": 0.4140569213215031, + "learning_rate": 0.0001709906888010196, + "loss": 2.3541, + "step": 3213 + }, + { + "epoch": 0.736986929603302, + "grad_norm": 0.4481016026646461, + "learning_rate": 0.00017071115492481087, + "loss": 2.3541, + "step": 3214 + }, + { + "epoch": 0.7372162348085302, + "grad_norm": 0.44440720872605577, + "learning_rate": 0.00017043180268212638, + "loss": 2.5956, + "step": 3215 + }, + { + "epoch": 0.7374455400137583, + "grad_norm": 0.3817860499111489, + "learning_rate": 0.0001701526322270549, + "loss": 2.4472, + "step": 3216 + }, + { + "epoch": 0.7376748452189865, + "grad_norm": 0.4100344222985941, + "learning_rate": 0.00016987364371358481, + "loss": 2.4118, + "step": 3217 + }, + { + "epoch": 0.7379041504242146, + "grad_norm": 0.39153712995977746, + "learning_rate": 0.00016959483729560472, + "loss": 2.4282, + "step": 3218 + }, + { + "epoch": 0.7381334556294428, + "grad_norm": 0.3871743790212973, + "learning_rate": 0.00016931621312690214, + "loss": 2.4094, + "step": 3219 + }, + { + "epoch": 0.7383627608346709, + "grad_norm": 0.374436833098019, + "learning_rate": 0.00016903777136116437, + "loss": 2.3109, + "step": 3220 + }, + { + "epoch": 0.7385920660398991, + "grad_norm": 0.41641039646354183, + "learning_rate": 0.00016875951215197777, + "loss": 2.3648, + "step": 3221 + }, + { + "epoch": 0.7388213712451273, + "grad_norm": 0.4433129564678002, + "learning_rate": 0.0001684814356528288, + "loss": 2.5062, + "step": 3222 + }, + { + "epoch": 0.7390506764503554, + "grad_norm": 0.42520198384542784, + "learning_rate": 0.00016820354201710214, + "loss": 2.517, + "step": 3223 + }, + { + "epoch": 0.7392799816555836, + "grad_norm": 0.4085724543690626, + "learning_rate": 0.00016792583139808197, + "loss": 2.4519, + "step": 3224 + }, + { + "epoch": 0.7395092868608117, + "grad_norm": 0.4022069048974689, + "learning_rate": 0.00016764830394895203, + "loss": 2.1958, + "step": 3225 + }, + { + "epoch": 0.7397385920660399, + "grad_norm": 0.4033396331864117, + "learning_rate": 0.00016737095982279442, + "loss": 2.3647, + "step": 3226 + }, + { + "epoch": 0.7399678972712681, + "grad_norm": 0.403865504210829, + "learning_rate": 0.00016709379917259027, + "loss": 2.3795, + "step": 3227 + }, + { + "epoch": 0.7401972024764962, + "grad_norm": 0.4065955776660432, + "learning_rate": 0.00016681682215121936, + "loss": 2.3763, + "step": 3228 + }, + { + "epoch": 0.7404265076817244, + "grad_norm": 0.40366851773247486, + "learning_rate": 0.00016654002891146091, + "loss": 2.3199, + "step": 3229 + }, + { + "epoch": 0.7406558128869525, + "grad_norm": 0.5345515606201237, + "learning_rate": 0.00016626341960599157, + "loss": 2.5939, + "step": 3230 + }, + { + "epoch": 0.7408851180921807, + "grad_norm": 0.4272107256545189, + "learning_rate": 0.00016598699438738764, + "loss": 2.4493, + "step": 3231 + }, + { + "epoch": 0.7411144232974088, + "grad_norm": 0.43452477718957644, + "learning_rate": 0.0001657107534081232, + "loss": 2.4708, + "step": 3232 + }, + { + "epoch": 0.741343728502637, + "grad_norm": 0.4080480265497662, + "learning_rate": 0.00016543469682057105, + "loss": 2.5619, + "step": 3233 + }, + { + "epoch": 0.7415730337078652, + "grad_norm": 0.4132361584912087, + "learning_rate": 0.0001651588247770019, + "loss": 2.4337, + "step": 3234 + }, + { + "epoch": 0.7418023389130933, + "grad_norm": 0.3920715761441055, + "learning_rate": 0.00016488313742958526, + "loss": 2.5475, + "step": 3235 + }, + { + "epoch": 0.7420316441183215, + "grad_norm": 0.3773565841831819, + "learning_rate": 0.0001646076349303884, + "loss": 2.4572, + "step": 3236 + }, + { + "epoch": 0.7422609493235497, + "grad_norm": 0.40738576041857877, + "learning_rate": 0.00016433231743137646, + "loss": 2.3828, + "step": 3237 + }, + { + "epoch": 0.7424902545287778, + "grad_norm": 0.38290632144063314, + "learning_rate": 0.00016405718508441264, + "loss": 2.3093, + "step": 3238 + }, + { + "epoch": 0.7427195597340059, + "grad_norm": 0.4392925075938457, + "learning_rate": 0.0001637822380412584, + "loss": 2.4167, + "step": 3239 + }, + { + "epoch": 0.7429488649392341, + "grad_norm": 0.39914153879757713, + "learning_rate": 0.00016350747645357266, + "loss": 2.4028, + "step": 3240 + }, + { + "epoch": 0.7431781701444623, + "grad_norm": 0.43080776235657275, + "learning_rate": 0.00016323290047291195, + "loss": 2.4863, + "step": 3241 + }, + { + "epoch": 0.7434074753496904, + "grad_norm": 0.40498531821811723, + "learning_rate": 0.0001629585102507304, + "loss": 2.3583, + "step": 3242 + }, + { + "epoch": 0.7436367805549186, + "grad_norm": 0.46757113140911905, + "learning_rate": 0.0001626843059383803, + "loss": 2.5096, + "step": 3243 + }, + { + "epoch": 0.7438660857601468, + "grad_norm": 0.39436403236885703, + "learning_rate": 0.0001624102876871103, + "loss": 2.4865, + "step": 3244 + }, + { + "epoch": 0.7440953909653749, + "grad_norm": 0.4032040219896239, + "learning_rate": 0.00016213645564806752, + "loss": 2.4854, + "step": 3245 + }, + { + "epoch": 0.744324696170603, + "grad_norm": 0.4203042963447584, + "learning_rate": 0.0001618628099722957, + "loss": 2.4106, + "step": 3246 + }, + { + "epoch": 0.7445540013758313, + "grad_norm": 0.41243102592758574, + "learning_rate": 0.0001615893508107359, + "loss": 2.4107, + "step": 3247 + }, + { + "epoch": 0.7447833065810594, + "grad_norm": 0.39905189623014053, + "learning_rate": 0.00016131607831422633, + "loss": 2.3912, + "step": 3248 + }, + { + "epoch": 0.7450126117862875, + "grad_norm": 0.40060882709536716, + "learning_rate": 0.00016104299263350252, + "loss": 2.4659, + "step": 3249 + }, + { + "epoch": 0.7452419169915158, + "grad_norm": 0.40585600382333015, + "learning_rate": 0.0001607700939191966, + "loss": 2.4951, + "step": 3250 + }, + { + "epoch": 0.7454712221967439, + "grad_norm": 0.44940342294395774, + "learning_rate": 0.00016049738232183758, + "loss": 2.5238, + "step": 3251 + }, + { + "epoch": 0.745700527401972, + "grad_norm": 0.44347328161486094, + "learning_rate": 0.00016022485799185128, + "loss": 2.3862, + "step": 3252 + }, + { + "epoch": 0.7459298326072001, + "grad_norm": 0.40036358081544526, + "learning_rate": 0.0001599525210795606, + "loss": 2.5191, + "step": 3253 + }, + { + "epoch": 0.7461591378124284, + "grad_norm": 0.40534142901900266, + "learning_rate": 0.0001596803717351845, + "loss": 2.3869, + "step": 3254 + }, + { + "epoch": 0.7463884430176565, + "grad_norm": 0.39477958286789755, + "learning_rate": 0.00015940841010883889, + "loss": 2.344, + "step": 3255 + }, + { + "epoch": 0.7466177482228846, + "grad_norm": 0.419321883239225, + "learning_rate": 0.00015913663635053576, + "loss": 2.3859, + "step": 3256 + }, + { + "epoch": 0.7468470534281129, + "grad_norm": 0.4161461433895471, + "learning_rate": 0.00015886505061018413, + "loss": 2.3031, + "step": 3257 + }, + { + "epoch": 0.747076358633341, + "grad_norm": 0.4182221315442342, + "learning_rate": 0.00015859365303758834, + "loss": 2.4128, + "step": 3258 + }, + { + "epoch": 0.7473056638385691, + "grad_norm": 0.4378706955852613, + "learning_rate": 0.0001583224437824498, + "loss": 2.3906, + "step": 3259 + }, + { + "epoch": 0.7475349690437973, + "grad_norm": 0.38774490025832914, + "learning_rate": 0.0001580514229943657, + "loss": 2.3751, + "step": 3260 + }, + { + "epoch": 0.7477642742490255, + "grad_norm": 0.45645030817906557, + "learning_rate": 0.0001577805908228293, + "loss": 2.4213, + "step": 3261 + }, + { + "epoch": 0.7479935794542536, + "grad_norm": 0.3908782637702008, + "learning_rate": 0.0001575099474172295, + "loss": 2.3595, + "step": 3262 + }, + { + "epoch": 0.7482228846594817, + "grad_norm": 0.431082119509026, + "learning_rate": 0.00015723949292685191, + "loss": 2.4375, + "step": 3263 + }, + { + "epoch": 0.74845218986471, + "grad_norm": 0.39570581670042426, + "learning_rate": 0.00015696922750087715, + "loss": 2.2749, + "step": 3264 + }, + { + "epoch": 0.7486814950699381, + "grad_norm": 0.3957128672039852, + "learning_rate": 0.0001566991512883818, + "loss": 2.2893, + "step": 3265 + }, + { + "epoch": 0.7489108002751662, + "grad_norm": 0.4332885947376619, + "learning_rate": 0.00015642926443833798, + "loss": 2.4623, + "step": 3266 + }, + { + "epoch": 0.7491401054803944, + "grad_norm": 0.4281752162108278, + "learning_rate": 0.00015615956709961378, + "loss": 2.4724, + "step": 3267 + }, + { + "epoch": 0.7493694106856226, + "grad_norm": 0.6549198556375135, + "learning_rate": 0.0001558900594209723, + "loss": 2.2354, + "step": 3268 + }, + { + "epoch": 0.7495987158908507, + "grad_norm": 0.4598595225529424, + "learning_rate": 0.00015562074155107215, + "loss": 2.4471, + "step": 3269 + }, + { + "epoch": 0.7498280210960789, + "grad_norm": 0.41110559842148814, + "learning_rate": 0.00015535161363846705, + "loss": 2.3425, + "step": 3270 + }, + { + "epoch": 0.7500573263013071, + "grad_norm": 0.4275108275753036, + "learning_rate": 0.0001550826758316068, + "loss": 2.3811, + "step": 3271 + }, + { + "epoch": 0.7502866315065352, + "grad_norm": 0.4375882501284239, + "learning_rate": 0.00015481392827883489, + "loss": 2.4739, + "step": 3272 + }, + { + "epoch": 0.7505159367117633, + "grad_norm": 0.3738321936828146, + "learning_rate": 0.00015454537112839122, + "loss": 2.3368, + "step": 3273 + }, + { + "epoch": 0.7507452419169915, + "grad_norm": 0.4062795595660475, + "learning_rate": 0.00015427700452840998, + "loss": 2.4139, + "step": 3274 + }, + { + "epoch": 0.7509745471222197, + "grad_norm": 0.47191382875180093, + "learning_rate": 0.00015400882862692033, + "loss": 2.4956, + "step": 3275 + }, + { + "epoch": 0.7512038523274478, + "grad_norm": 0.4083346830796121, + "learning_rate": 0.0001537408435718462, + "loss": 2.2921, + "step": 3276 + }, + { + "epoch": 0.751433157532676, + "grad_norm": 0.41017927599946175, + "learning_rate": 0.00015347304951100665, + "loss": 2.3086, + "step": 3277 + }, + { + "epoch": 0.7516624627379042, + "grad_norm": 0.3938677255426243, + "learning_rate": 0.000153205446592115, + "loss": 2.5384, + "step": 3278 + }, + { + "epoch": 0.7518917679431323, + "grad_norm": 0.40150927346947196, + "learning_rate": 0.00015293803496277907, + "loss": 2.3541, + "step": 3279 + }, + { + "epoch": 0.7521210731483605, + "grad_norm": 0.3912715053335062, + "learning_rate": 0.00015267081477050133, + "loss": 2.5847, + "step": 3280 + }, + { + "epoch": 0.7523503783535886, + "grad_norm": 0.4157005361465217, + "learning_rate": 0.00015240378616267886, + "loss": 2.4685, + "step": 3281 + }, + { + "epoch": 0.7525796835588168, + "grad_norm": 0.38501927690400295, + "learning_rate": 0.00015213694928660276, + "loss": 2.3528, + "step": 3282 + }, + { + "epoch": 0.7528089887640449, + "grad_norm": 0.4079708048430353, + "learning_rate": 0.00015187030428945843, + "loss": 2.2546, + "step": 3283 + }, + { + "epoch": 0.7530382939692731, + "grad_norm": 0.40006514527707315, + "learning_rate": 0.00015160385131832532, + "loss": 2.357, + "step": 3284 + }, + { + "epoch": 0.7532675991745013, + "grad_norm": 0.4020539111274743, + "learning_rate": 0.0001513375905201776, + "loss": 2.5239, + "step": 3285 + }, + { + "epoch": 0.7534969043797294, + "grad_norm": 0.44403438874251894, + "learning_rate": 0.0001510715220418823, + "loss": 2.473, + "step": 3286 + }, + { + "epoch": 0.7537262095849576, + "grad_norm": 0.4283545109566083, + "learning_rate": 0.00015080564603020142, + "loss": 2.5369, + "step": 3287 + }, + { + "epoch": 0.7539555147901857, + "grad_norm": 0.42661160449306457, + "learning_rate": 0.00015053996263179032, + "loss": 2.47, + "step": 3288 + }, + { + "epoch": 0.7541848199954139, + "grad_norm": 0.3941492242886424, + "learning_rate": 0.0001502744719931982, + "loss": 2.5173, + "step": 3289 + }, + { + "epoch": 0.7544141252006421, + "grad_norm": 0.4190322472222547, + "learning_rate": 0.0001500091742608677, + "loss": 2.4744, + "step": 3290 + }, + { + "epoch": 0.7546434304058702, + "grad_norm": 0.39916779403016855, + "learning_rate": 0.00014974406958113558, + "loss": 2.3701, + "step": 3291 + }, + { + "epoch": 0.7548727356110984, + "grad_norm": 0.40646463600204197, + "learning_rate": 0.00014947915810023171, + "loss": 2.3731, + "step": 3292 + }, + { + "epoch": 0.7551020408163265, + "grad_norm": 0.38713477083236275, + "learning_rate": 0.00014921443996427947, + "loss": 2.3907, + "step": 3293 + }, + { + "epoch": 0.7553313460215547, + "grad_norm": 0.4386187497062649, + "learning_rate": 0.0001489499153192954, + "loss": 2.3962, + "step": 3294 + }, + { + "epoch": 0.7555606512267828, + "grad_norm": 0.43401713850941775, + "learning_rate": 0.0001486855843111901, + "loss": 2.3975, + "step": 3295 + }, + { + "epoch": 0.755789956432011, + "grad_norm": 0.4089592162259285, + "learning_rate": 0.00014842144708576604, + "loss": 2.5018, + "step": 3296 + }, + { + "epoch": 0.7560192616372392, + "grad_norm": 0.3868477886158715, + "learning_rate": 0.0001481575037887201, + "loss": 2.4152, + "step": 3297 + }, + { + "epoch": 0.7562485668424673, + "grad_norm": 0.43479219515489625, + "learning_rate": 0.00014789375456564135, + "loss": 2.5202, + "step": 3298 + }, + { + "epoch": 0.7564778720476955, + "grad_norm": 0.39113452399018767, + "learning_rate": 0.00014763019956201253, + "loss": 2.4802, + "step": 3299 + }, + { + "epoch": 0.7567071772529237, + "grad_norm": 0.43217570596552746, + "learning_rate": 0.00014736683892320812, + "loss": 2.332, + "step": 3300 + }, + { + "epoch": 0.7569364824581518, + "grad_norm": 0.4112887998869391, + "learning_rate": 0.0001471036727944966, + "loss": 2.4885, + "step": 3301 + }, + { + "epoch": 0.7571657876633799, + "grad_norm": 0.41040657799752767, + "learning_rate": 0.00014684070132103854, + "loss": 2.3424, + "step": 3302 + }, + { + "epoch": 0.7573950928686081, + "grad_norm": 0.41550522073432894, + "learning_rate": 0.0001465779246478872, + "loss": 2.3512, + "step": 3303 + }, + { + "epoch": 0.7576243980738363, + "grad_norm": 0.41848925856497493, + "learning_rate": 0.00014631534291998833, + "loss": 2.3809, + "step": 3304 + }, + { + "epoch": 0.7578537032790644, + "grad_norm": 0.43353356082743516, + "learning_rate": 0.00014605295628218045, + "loss": 2.432, + "step": 3305 + }, + { + "epoch": 0.7580830084842926, + "grad_norm": 0.3977096833871388, + "learning_rate": 0.0001457907648791943, + "loss": 2.3941, + "step": 3306 + }, + { + "epoch": 0.7583123136895208, + "grad_norm": 0.4631481884231966, + "learning_rate": 0.0001455287688556527, + "loss": 2.3533, + "step": 3307 + }, + { + "epoch": 0.7585416188947489, + "grad_norm": 0.41090633041742197, + "learning_rate": 0.00014526696835607088, + "loss": 2.4754, + "step": 3308 + }, + { + "epoch": 0.758770924099977, + "grad_norm": 0.40822705423182937, + "learning_rate": 0.00014500536352485673, + "loss": 2.4674, + "step": 3309 + }, + { + "epoch": 0.7590002293052053, + "grad_norm": 0.41961214273484887, + "learning_rate": 0.00014474395450630895, + "loss": 2.3677, + "step": 3310 + }, + { + "epoch": 0.7592295345104334, + "grad_norm": 0.39817126157901495, + "learning_rate": 0.00014448274144461965, + "loss": 2.421, + "step": 3311 + }, + { + "epoch": 0.7594588397156615, + "grad_norm": 0.44085925523332453, + "learning_rate": 0.00014422172448387178, + "loss": 2.3588, + "step": 3312 + }, + { + "epoch": 0.7596881449208897, + "grad_norm": 0.5229076291224334, + "learning_rate": 0.00014396090376804112, + "loss": 2.4217, + "step": 3313 + }, + { + "epoch": 0.7599174501261179, + "grad_norm": 0.4249392078359076, + "learning_rate": 0.0001437002794409939, + "loss": 2.625, + "step": 3314 + }, + { + "epoch": 0.760146755331346, + "grad_norm": 0.3905704186344273, + "learning_rate": 0.00014343985164648926, + "loss": 2.3705, + "step": 3315 + }, + { + "epoch": 0.7603760605365741, + "grad_norm": 0.4322149924073928, + "learning_rate": 0.0001431796205281773, + "loss": 2.371, + "step": 3316 + }, + { + "epoch": 0.7606053657418024, + "grad_norm": 0.3942234956774646, + "learning_rate": 0.00014291958622959973, + "loss": 2.4631, + "step": 3317 + }, + { + "epoch": 0.7608346709470305, + "grad_norm": 0.4281122818313435, + "learning_rate": 0.00014265974889418959, + "loss": 2.2538, + "step": 3318 + }, + { + "epoch": 0.7610639761522586, + "grad_norm": 0.40723307476395115, + "learning_rate": 0.00014240010866527176, + "loss": 2.5096, + "step": 3319 + }, + { + "epoch": 0.7612932813574869, + "grad_norm": 0.44134818827963407, + "learning_rate": 0.00014214066568606182, + "loss": 2.4318, + "step": 3320 + }, + { + "epoch": 0.761522586562715, + "grad_norm": 0.4071605516274004, + "learning_rate": 0.00014188142009966686, + "loss": 2.2334, + "step": 3321 + }, + { + "epoch": 0.7617518917679431, + "grad_norm": 0.399917925181231, + "learning_rate": 0.00014162237204908478, + "loss": 2.4266, + "step": 3322 + }, + { + "epoch": 0.7619811969731712, + "grad_norm": 0.3889185999832242, + "learning_rate": 0.0001413635216772053, + "loss": 2.3897, + "step": 3323 + }, + { + "epoch": 0.7622105021783995, + "grad_norm": 0.3938532876772715, + "learning_rate": 0.00014110486912680793, + "loss": 2.3434, + "step": 3324 + }, + { + "epoch": 0.7624398073836276, + "grad_norm": 0.48784536343773655, + "learning_rate": 0.000140846414540564, + "loss": 2.3416, + "step": 3325 + }, + { + "epoch": 0.7626691125888557, + "grad_norm": 0.3879826590358214, + "learning_rate": 0.00014058815806103543, + "loss": 2.4091, + "step": 3326 + }, + { + "epoch": 0.762898417794084, + "grad_norm": 0.4042193819664549, + "learning_rate": 0.00014033009983067452, + "loss": 2.3357, + "step": 3327 + }, + { + "epoch": 0.7631277229993121, + "grad_norm": 0.43763705594185387, + "learning_rate": 0.00014007223999182444, + "loss": 2.3636, + "step": 3328 + }, + { + "epoch": 0.7633570282045402, + "grad_norm": 0.5400638692809757, + "learning_rate": 0.00013981457868671927, + "loss": 2.4238, + "step": 3329 + }, + { + "epoch": 0.7635863334097684, + "grad_norm": 0.40943088915281306, + "learning_rate": 0.00013955711605748305, + "loss": 2.3384, + "step": 3330 + }, + { + "epoch": 0.7638156386149966, + "grad_norm": 0.6419166544140655, + "learning_rate": 0.0001392998522461305, + "loss": 2.4733, + "step": 3331 + }, + { + "epoch": 0.7640449438202247, + "grad_norm": 0.45927518161445374, + "learning_rate": 0.00013904278739456643, + "loss": 2.6413, + "step": 3332 + }, + { + "epoch": 0.7642742490254528, + "grad_norm": 0.41532827827010005, + "learning_rate": 0.00013878592164458635, + "loss": 2.2961, + "step": 3333 + }, + { + "epoch": 0.7645035542306811, + "grad_norm": 0.41655555801140237, + "learning_rate": 0.00013852925513787567, + "loss": 2.417, + "step": 3334 + }, + { + "epoch": 0.7647328594359092, + "grad_norm": 0.4521547520218105, + "learning_rate": 0.00013827278801600978, + "loss": 2.4207, + "step": 3335 + }, + { + "epoch": 0.7649621646411373, + "grad_norm": 0.40744860488767104, + "learning_rate": 0.00013801652042045416, + "loss": 2.5293, + "step": 3336 + }, + { + "epoch": 0.7651914698463655, + "grad_norm": 0.4236727315405918, + "learning_rate": 0.0001377604524925647, + "loss": 2.6073, + "step": 3337 + }, + { + "epoch": 0.7654207750515937, + "grad_norm": 0.42770235475109647, + "learning_rate": 0.0001375045843735862, + "loss": 2.4745, + "step": 3338 + }, + { + "epoch": 0.7656500802568218, + "grad_norm": 0.4007407792765845, + "learning_rate": 0.00013724891620465424, + "loss": 2.4268, + "step": 3339 + }, + { + "epoch": 0.76587938546205, + "grad_norm": 0.4155421103863451, + "learning_rate": 0.00013699344812679353, + "loss": 2.4165, + "step": 3340 + }, + { + "epoch": 0.7661086906672782, + "grad_norm": 0.40340655331150227, + "learning_rate": 0.0001367381802809185, + "loss": 2.3862, + "step": 3341 + }, + { + "epoch": 0.7663379958725063, + "grad_norm": 0.406924123030159, + "learning_rate": 0.0001364831128078331, + "loss": 2.3281, + "step": 3342 + }, + { + "epoch": 0.7665673010777345, + "grad_norm": 0.3931338814198388, + "learning_rate": 0.00013622824584823113, + "loss": 2.3682, + "step": 3343 + }, + { + "epoch": 0.7667966062829626, + "grad_norm": 0.4286903861085394, + "learning_rate": 0.00013597357954269534, + "loss": 2.4424, + "step": 3344 + }, + { + "epoch": 0.7670259114881908, + "grad_norm": 0.41514234408414996, + "learning_rate": 0.00013571911403169795, + "loss": 2.48, + "step": 3345 + }, + { + "epoch": 0.7672552166934189, + "grad_norm": 0.4395554692982556, + "learning_rate": 0.00013546484945560028, + "loss": 2.3499, + "step": 3346 + }, + { + "epoch": 0.7674845218986471, + "grad_norm": 0.4170782601088326, + "learning_rate": 0.0001352107859546533, + "loss": 2.331, + "step": 3347 + }, + { + "epoch": 0.7677138271038753, + "grad_norm": 0.4399278664661611, + "learning_rate": 0.00013495692366899658, + "loss": 2.413, + "step": 3348 + }, + { + "epoch": 0.7679431323091034, + "grad_norm": 0.42567084320541787, + "learning_rate": 0.00013470326273865886, + "loss": 2.4531, + "step": 3349 + }, + { + "epoch": 0.7681724375143316, + "grad_norm": 0.42525387204269643, + "learning_rate": 0.0001344498033035576, + "loss": 2.4248, + "step": 3350 + }, + { + "epoch": 0.7684017427195597, + "grad_norm": 0.43284341069620935, + "learning_rate": 0.00013419654550349985, + "loss": 2.4831, + "step": 3351 + }, + { + "epoch": 0.7686310479247879, + "grad_norm": 0.4161151170363166, + "learning_rate": 0.0001339434894781803, + "loss": 2.4953, + "step": 3352 + }, + { + "epoch": 0.7688603531300161, + "grad_norm": 0.3984324974996877, + "learning_rate": 0.00013369063536718346, + "loss": 2.4907, + "step": 3353 + }, + { + "epoch": 0.7690896583352442, + "grad_norm": 0.4103827735029681, + "learning_rate": 0.0001334379833099817, + "loss": 2.2573, + "step": 3354 + }, + { + "epoch": 0.7693189635404724, + "grad_norm": 0.40995514666786487, + "learning_rate": 0.00013318553344593632, + "loss": 2.3177, + "step": 3355 + }, + { + "epoch": 0.7695482687457005, + "grad_norm": 0.44434982582844385, + "learning_rate": 0.0001329332859142967, + "loss": 2.486, + "step": 3356 + }, + { + "epoch": 0.7697775739509287, + "grad_norm": 0.41251660503267845, + "learning_rate": 0.00013268124085420136, + "loss": 2.3979, + "step": 3357 + }, + { + "epoch": 0.7700068791561568, + "grad_norm": 0.38848190615809747, + "learning_rate": 0.00013242939840467637, + "loss": 2.4696, + "step": 3358 + }, + { + "epoch": 0.770236184361385, + "grad_norm": 0.41531066254556404, + "learning_rate": 0.0001321777587046364, + "loss": 2.4818, + "step": 3359 + }, + { + "epoch": 0.7704654895666132, + "grad_norm": 0.48779230895458875, + "learning_rate": 0.00013192632189288405, + "loss": 2.4998, + "step": 3360 + }, + { + "epoch": 0.7706947947718413, + "grad_norm": 0.4181081923616364, + "learning_rate": 0.00013167508810811059, + "loss": 2.4677, + "step": 3361 + }, + { + "epoch": 0.7709240999770695, + "grad_norm": 0.425896683340428, + "learning_rate": 0.00013142405748889458, + "loss": 2.4553, + "step": 3362 + }, + { + "epoch": 0.7711534051822977, + "grad_norm": 0.4175228721164271, + "learning_rate": 0.0001311732301737029, + "loss": 2.462, + "step": 3363 + }, + { + "epoch": 0.7713827103875258, + "grad_norm": 0.4096287167008857, + "learning_rate": 0.00013092260630089014, + "loss": 2.5293, + "step": 3364 + }, + { + "epoch": 0.7716120155927539, + "grad_norm": 0.428742846685001, + "learning_rate": 0.0001306721860086991, + "loss": 2.5001, + "step": 3365 + }, + { + "epoch": 0.7718413207979821, + "grad_norm": 0.3830599398916876, + "learning_rate": 0.00013042196943525942, + "loss": 2.4302, + "step": 3366 + }, + { + "epoch": 0.7720706260032103, + "grad_norm": 0.37803688153267273, + "learning_rate": 0.00013017195671858928, + "loss": 2.3963, + "step": 3367 + }, + { + "epoch": 0.7722999312084384, + "grad_norm": 0.386815797590407, + "learning_rate": 0.0001299221479965939, + "loss": 2.3796, + "step": 3368 + }, + { + "epoch": 0.7725292364136666, + "grad_norm": 0.3949502836149769, + "learning_rate": 0.0001296725434070661, + "loss": 2.5094, + "step": 3369 + }, + { + "epoch": 0.7727585416188948, + "grad_norm": 0.39935356542340833, + "learning_rate": 0.00012942314308768594, + "loss": 2.54, + "step": 3370 + }, + { + "epoch": 0.7729878468241229, + "grad_norm": 0.3997780341498846, + "learning_rate": 0.00012917394717602121, + "loss": 2.4728, + "step": 3371 + }, + { + "epoch": 0.773217152029351, + "grad_norm": 0.4573235691676702, + "learning_rate": 0.00012892495580952667, + "loss": 2.532, + "step": 3372 + }, + { + "epoch": 0.7734464572345793, + "grad_norm": 0.38632673893721314, + "learning_rate": 0.00012867616912554426, + "loss": 2.3919, + "step": 3373 + }, + { + "epoch": 0.7736757624398074, + "grad_norm": 0.4301815111176681, + "learning_rate": 0.00012842758726130281, + "loss": 2.295, + "step": 3374 + }, + { + "epoch": 0.7739050676450355, + "grad_norm": 0.3764873398403183, + "learning_rate": 0.00012817921035391882, + "loss": 2.4766, + "step": 3375 + }, + { + "epoch": 0.7741343728502637, + "grad_norm": 0.5685444042316482, + "learning_rate": 0.00012793103854039518, + "loss": 2.349, + "step": 3376 + }, + { + "epoch": 0.7743636780554919, + "grad_norm": 0.39902368971907487, + "learning_rate": 0.00012768307195762168, + "loss": 2.3872, + "step": 3377 + }, + { + "epoch": 0.77459298326072, + "grad_norm": 0.4011503656731011, + "learning_rate": 0.00012743531074237507, + "loss": 2.4635, + "step": 3378 + }, + { + "epoch": 0.7748222884659481, + "grad_norm": 0.41325993381531, + "learning_rate": 0.00012718775503131908, + "loss": 2.3975, + "step": 3379 + }, + { + "epoch": 0.7750515936711764, + "grad_norm": 0.39472935519842034, + "learning_rate": 0.00012694040496100318, + "loss": 2.3764, + "step": 3380 + }, + { + "epoch": 0.7752808988764045, + "grad_norm": 0.43059162489314273, + "learning_rate": 0.0001266932606678646, + "loss": 2.4904, + "step": 3381 + }, + { + "epoch": 0.7755102040816326, + "grad_norm": 0.4218436699741544, + "learning_rate": 0.00012644632228822617, + "loss": 2.4087, + "step": 3382 + }, + { + "epoch": 0.7757395092868609, + "grad_norm": 0.44805252213520363, + "learning_rate": 0.00012619958995829756, + "loss": 2.4141, + "step": 3383 + }, + { + "epoch": 0.775968814492089, + "grad_norm": 0.40270946605148206, + "learning_rate": 0.00012595306381417448, + "loss": 2.5316, + "step": 3384 + }, + { + "epoch": 0.7761981196973171, + "grad_norm": 0.3828257213363296, + "learning_rate": 0.0001257067439918394, + "loss": 2.4511, + "step": 3385 + }, + { + "epoch": 0.7764274249025452, + "grad_norm": 0.39551786667063854, + "learning_rate": 0.0001254606306271607, + "loss": 2.3727, + "step": 3386 + }, + { + "epoch": 0.7766567301077735, + "grad_norm": 0.3990754729834297, + "learning_rate": 0.00012521472385589234, + "loss": 2.4334, + "step": 3387 + }, + { + "epoch": 0.7768860353130016, + "grad_norm": 0.42196024646711133, + "learning_rate": 0.00012496902381367543, + "loss": 2.3745, + "step": 3388 + }, + { + "epoch": 0.7771153405182297, + "grad_norm": 0.4523988891025732, + "learning_rate": 0.00012472353063603626, + "loss": 2.3579, + "step": 3389 + }, + { + "epoch": 0.777344645723458, + "grad_norm": 1.872472331357063, + "learning_rate": 0.00012447824445838718, + "loss": 2.4302, + "step": 3390 + }, + { + "epoch": 0.7775739509286861, + "grad_norm": 0.4181020952121823, + "learning_rate": 0.0001242331654160263, + "loss": 2.3776, + "step": 3391 + }, + { + "epoch": 0.7778032561339142, + "grad_norm": 0.43726262455818127, + "learning_rate": 0.00012398829364413793, + "loss": 2.4583, + "step": 3392 + }, + { + "epoch": 0.7780325613391424, + "grad_norm": 0.3924474043889944, + "learning_rate": 0.0001237436292777914, + "loss": 2.4738, + "step": 3393 + }, + { + "epoch": 0.7782618665443706, + "grad_norm": 0.41595924301438614, + "learning_rate": 0.00012349917245194214, + "loss": 2.3573, + "step": 3394 + }, + { + "epoch": 0.7784911717495987, + "grad_norm": 0.39361023509118864, + "learning_rate": 0.00012325492330143061, + "loss": 2.4205, + "step": 3395 + }, + { + "epoch": 0.7787204769548268, + "grad_norm": 0.3942915450544963, + "learning_rate": 0.0001230108819609833, + "loss": 2.4582, + "step": 3396 + }, + { + "epoch": 0.7789497821600551, + "grad_norm": 0.4150747909084578, + "learning_rate": 0.00012276704856521175, + "loss": 2.3936, + "step": 3397 + }, + { + "epoch": 0.7791790873652832, + "grad_norm": 0.514599917298864, + "learning_rate": 0.0001225234232486127, + "loss": 2.3077, + "step": 3398 + }, + { + "epoch": 0.7794083925705113, + "grad_norm": 0.4114066317446174, + "learning_rate": 0.00012228000614556816, + "loss": 2.4906, + "step": 3399 + }, + { + "epoch": 0.7796376977757395, + "grad_norm": 0.4187982497632212, + "learning_rate": 0.00012203679739034579, + "loss": 2.4143, + "step": 3400 + }, + { + "epoch": 0.7798670029809677, + "grad_norm": 0.42860322307419635, + "learning_rate": 0.00012179379711709738, + "loss": 2.3992, + "step": 3401 + }, + { + "epoch": 0.7800963081861958, + "grad_norm": 0.41336839851378654, + "learning_rate": 0.00012155100545986065, + "loss": 2.4859, + "step": 3402 + }, + { + "epoch": 0.780325613391424, + "grad_norm": 0.405663298263079, + "learning_rate": 0.0001213084225525577, + "loss": 2.4299, + "step": 3403 + }, + { + "epoch": 0.7805549185966522, + "grad_norm": 0.3990812030634922, + "learning_rate": 0.00012106604852899566, + "loss": 2.5131, + "step": 3404 + }, + { + "epoch": 0.7807842238018803, + "grad_norm": 0.44753359725588426, + "learning_rate": 0.00012082388352286627, + "loss": 2.5905, + "step": 3405 + }, + { + "epoch": 0.7810135290071084, + "grad_norm": 0.39863677281162774, + "learning_rate": 0.00012058192766774639, + "loss": 2.611, + "step": 3406 + }, + { + "epoch": 0.7812428342123366, + "grad_norm": 0.4096629230972119, + "learning_rate": 0.00012034018109709716, + "loss": 2.5366, + "step": 3407 + }, + { + "epoch": 0.7814721394175648, + "grad_norm": 0.41514175806941983, + "learning_rate": 0.00012009864394426434, + "loss": 2.4183, + "step": 3408 + }, + { + "epoch": 0.7817014446227929, + "grad_norm": 0.3734200002808564, + "learning_rate": 0.00011985731634247809, + "loss": 2.4081, + "step": 3409 + }, + { + "epoch": 0.7819307498280211, + "grad_norm": 0.38664367154254103, + "learning_rate": 0.00011961619842485333, + "loss": 2.381, + "step": 3410 + }, + { + "epoch": 0.7821600550332493, + "grad_norm": 0.43388321004940095, + "learning_rate": 0.00011937529032438904, + "loss": 2.4447, + "step": 3411 + }, + { + "epoch": 0.7823893602384774, + "grad_norm": 0.4017238020461609, + "learning_rate": 0.00011913459217396843, + "loss": 2.3692, + "step": 3412 + }, + { + "epoch": 0.7826186654437056, + "grad_norm": 0.3770955604812929, + "learning_rate": 0.00011889410410635887, + "loss": 2.3294, + "step": 3413 + }, + { + "epoch": 0.7828479706489337, + "grad_norm": 0.42461577358056446, + "learning_rate": 0.00011865382625421239, + "loss": 2.37, + "step": 3414 + }, + { + "epoch": 0.7830772758541619, + "grad_norm": 0.39909867854514086, + "learning_rate": 0.0001184137587500641, + "loss": 2.3692, + "step": 3415 + }, + { + "epoch": 0.78330658105939, + "grad_norm": 0.40678449767479824, + "learning_rate": 0.00011817390172633402, + "loss": 2.256, + "step": 3416 + }, + { + "epoch": 0.7835358862646182, + "grad_norm": 0.42193245884363234, + "learning_rate": 0.00011793425531532564, + "loss": 2.2586, + "step": 3417 + }, + { + "epoch": 0.7837651914698464, + "grad_norm": 0.39794930274711476, + "learning_rate": 0.00011769481964922618, + "loss": 2.3806, + "step": 3418 + }, + { + "epoch": 0.7839944966750745, + "grad_norm": 0.39322429044692014, + "learning_rate": 0.00011745559486010671, + "loss": 2.3962, + "step": 3419 + }, + { + "epoch": 0.7842238018803027, + "grad_norm": 0.4021587011061479, + "learning_rate": 0.00011721658107992223, + "loss": 2.4761, + "step": 3420 + }, + { + "epoch": 0.7844531070855308, + "grad_norm": 0.4224296662265774, + "learning_rate": 0.00011697777844051105, + "loss": 2.4647, + "step": 3421 + }, + { + "epoch": 0.784682412290759, + "grad_norm": 0.3994387762447641, + "learning_rate": 0.0001167391870735951, + "loss": 2.4951, + "step": 3422 + }, + { + "epoch": 0.7849117174959872, + "grad_norm": 0.40298756784242024, + "learning_rate": 0.00011650080711077964, + "loss": 2.4711, + "step": 3423 + }, + { + "epoch": 0.7851410227012153, + "grad_norm": 0.583797913861257, + "learning_rate": 0.00011626263868355375, + "loss": 2.3571, + "step": 3424 + }, + { + "epoch": 0.7853703279064435, + "grad_norm": 0.4051562678375087, + "learning_rate": 0.00011602468192328936, + "loss": 2.583, + "step": 3425 + }, + { + "epoch": 0.7855996331116716, + "grad_norm": 0.4069949789151716, + "learning_rate": 0.00011578693696124193, + "loss": 2.4344, + "step": 3426 + }, + { + "epoch": 0.7858289383168998, + "grad_norm": 0.39685827293527387, + "learning_rate": 0.00011554940392854973, + "loss": 2.511, + "step": 3427 + }, + { + "epoch": 0.7860582435221279, + "grad_norm": 0.4042427049225674, + "learning_rate": 0.00011531208295623485, + "loss": 2.4093, + "step": 3428 + }, + { + "epoch": 0.7862875487273561, + "grad_norm": 0.39565579675922574, + "learning_rate": 0.00011507497417520146, + "loss": 2.3893, + "step": 3429 + }, + { + "epoch": 0.7865168539325843, + "grad_norm": 0.4003761687711523, + "learning_rate": 0.0001148380777162375, + "loss": 2.5601, + "step": 3430 + }, + { + "epoch": 0.7867461591378124, + "grad_norm": 0.38632534771229365, + "learning_rate": 0.00011460139371001339, + "loss": 2.4802, + "step": 3431 + }, + { + "epoch": 0.7869754643430406, + "grad_norm": 0.39545304810334164, + "learning_rate": 0.00011436492228708239, + "loss": 2.3702, + "step": 3432 + }, + { + "epoch": 0.7872047695482688, + "grad_norm": 0.40939784570739907, + "learning_rate": 0.00011412866357788049, + "loss": 2.402, + "step": 3433 + }, + { + "epoch": 0.7874340747534969, + "grad_norm": 0.41227137170390704, + "learning_rate": 0.00011389261771272663, + "loss": 2.4195, + "step": 3434 + }, + { + "epoch": 0.787663379958725, + "grad_norm": 0.42891755305821244, + "learning_rate": 0.00011365678482182207, + "loss": 2.2644, + "step": 3435 + }, + { + "epoch": 0.7878926851639533, + "grad_norm": 0.39168096220692045, + "learning_rate": 0.00011342116503525057, + "loss": 2.3707, + "step": 3436 + }, + { + "epoch": 0.7881219903691814, + "grad_norm": 0.3964948705395691, + "learning_rate": 0.0001131857584829783, + "loss": 2.2261, + "step": 3437 + }, + { + "epoch": 0.7883512955744095, + "grad_norm": 0.3844633895889368, + "learning_rate": 0.00011295056529485426, + "loss": 2.5705, + "step": 3438 + }, + { + "epoch": 0.7885806007796377, + "grad_norm": 0.4578444704204237, + "learning_rate": 0.0001127155856006093, + "loss": 2.3818, + "step": 3439 + }, + { + "epoch": 0.7888099059848659, + "grad_norm": 0.4318764417672927, + "learning_rate": 0.00011248081952985656, + "loss": 2.559, + "step": 3440 + }, + { + "epoch": 0.789039211190094, + "grad_norm": 0.4157239247507073, + "learning_rate": 0.00011224626721209141, + "loss": 2.336, + "step": 3441 + }, + { + "epoch": 0.7892685163953221, + "grad_norm": 0.4616872925163364, + "learning_rate": 0.0001120119287766917, + "loss": 2.259, + "step": 3442 + }, + { + "epoch": 0.7894978216005504, + "grad_norm": 0.41870237090561196, + "learning_rate": 0.0001117778043529164, + "loss": 2.4651, + "step": 3443 + }, + { + "epoch": 0.7897271268057785, + "grad_norm": 0.42127214186746736, + "learning_rate": 0.0001115438940699074, + "loss": 2.5309, + "step": 3444 + }, + { + "epoch": 0.7899564320110066, + "grad_norm": 0.4223412279437213, + "learning_rate": 0.0001113101980566879, + "loss": 2.4136, + "step": 3445 + }, + { + "epoch": 0.7901857372162349, + "grad_norm": 0.41487056816981566, + "learning_rate": 0.00011107671644216305, + "loss": 2.3215, + "step": 3446 + }, + { + "epoch": 0.790415042421463, + "grad_norm": 0.40917381394048313, + "learning_rate": 0.00011084344935511958, + "loss": 2.3144, + "step": 3447 + }, + { + "epoch": 0.7906443476266911, + "grad_norm": 0.42452638266632475, + "learning_rate": 0.00011061039692422637, + "loss": 2.4051, + "step": 3448 + }, + { + "epoch": 0.7908736528319192, + "grad_norm": 0.40224316675693095, + "learning_rate": 0.00011037755927803345, + "loss": 2.4295, + "step": 3449 + }, + { + "epoch": 0.7911029580371475, + "grad_norm": 0.3988269255925655, + "learning_rate": 0.00011014493654497254, + "loss": 2.4424, + "step": 3450 + }, + { + "epoch": 0.7913322632423756, + "grad_norm": 0.38198553112144346, + "learning_rate": 0.00010991252885335651, + "loss": 2.4138, + "step": 3451 + }, + { + "epoch": 0.7915615684476037, + "grad_norm": 0.3923911078921115, + "learning_rate": 0.0001096803363313803, + "loss": 2.4038, + "step": 3452 + }, + { + "epoch": 0.791790873652832, + "grad_norm": 0.40153548102675835, + "learning_rate": 0.00010944835910711958, + "loss": 2.4968, + "step": 3453 + }, + { + "epoch": 0.7920201788580601, + "grad_norm": 0.4100325139524148, + "learning_rate": 0.00010921659730853145, + "loss": 2.3782, + "step": 3454 + }, + { + "epoch": 0.7922494840632882, + "grad_norm": 0.4003781427380612, + "learning_rate": 0.00010898505106345396, + "loss": 2.4569, + "step": 3455 + }, + { + "epoch": 0.7924787892685164, + "grad_norm": 0.4212010986549088, + "learning_rate": 0.00010875372049960696, + "loss": 2.4704, + "step": 3456 + }, + { + "epoch": 0.7927080944737446, + "grad_norm": 0.40850455436862093, + "learning_rate": 0.00010852260574459022, + "loss": 2.4657, + "step": 3457 + }, + { + "epoch": 0.7929373996789727, + "grad_norm": 0.39922760641890903, + "learning_rate": 0.00010829170692588558, + "loss": 2.3208, + "step": 3458 + }, + { + "epoch": 0.7931667048842008, + "grad_norm": 0.4362254387641223, + "learning_rate": 0.00010806102417085512, + "loss": 2.409, + "step": 3459 + }, + { + "epoch": 0.7933960100894291, + "grad_norm": 0.39897684196237254, + "learning_rate": 0.00010783055760674182, + "loss": 2.3914, + "step": 3460 + }, + { + "epoch": 0.7936253152946572, + "grad_norm": 0.38801578260155634, + "learning_rate": 0.00010760030736066951, + "loss": 2.3986, + "step": 3461 + }, + { + "epoch": 0.7938546204998853, + "grad_norm": 0.42034284397077337, + "learning_rate": 0.00010737027355964284, + "loss": 2.4093, + "step": 3462 + }, + { + "epoch": 0.7940839257051135, + "grad_norm": 0.7181673951428934, + "learning_rate": 0.00010714045633054687, + "loss": 2.3459, + "step": 3463 + }, + { + "epoch": 0.7943132309103417, + "grad_norm": 0.4069720202274433, + "learning_rate": 0.00010691085580014714, + "loss": 2.4978, + "step": 3464 + }, + { + "epoch": 0.7945425361155698, + "grad_norm": 0.4103208109918724, + "learning_rate": 0.00010668147209508971, + "loss": 2.466, + "step": 3465 + }, + { + "epoch": 0.794771841320798, + "grad_norm": 0.4148999112113308, + "learning_rate": 0.00010645230534190148, + "loss": 2.2579, + "step": 3466 + }, + { + "epoch": 0.7950011465260262, + "grad_norm": 0.4142335550235184, + "learning_rate": 0.00010622335566698877, + "loss": 2.4412, + "step": 3467 + }, + { + "epoch": 0.7952304517312543, + "grad_norm": 0.3844164875089216, + "learning_rate": 0.00010599462319663906, + "loss": 2.4921, + "step": 3468 + }, + { + "epoch": 0.7954597569364824, + "grad_norm": 0.39951630170438185, + "learning_rate": 0.00010576610805701942, + "loss": 2.4266, + "step": 3469 + }, + { + "epoch": 0.7956890621417106, + "grad_norm": 0.43983411995865906, + "learning_rate": 0.0001055378103741777, + "loss": 2.4847, + "step": 3470 + }, + { + "epoch": 0.7959183673469388, + "grad_norm": 0.4188385198397135, + "learning_rate": 0.00010530973027404073, + "loss": 2.4753, + "step": 3471 + }, + { + "epoch": 0.7961476725521669, + "grad_norm": 0.42126689233211556, + "learning_rate": 0.00010508186788241647, + "loss": 2.3974, + "step": 3472 + }, + { + "epoch": 0.7963769777573951, + "grad_norm": 0.3864371464313023, + "learning_rate": 0.00010485422332499212, + "loss": 2.5004, + "step": 3473 + }, + { + "epoch": 0.7966062829626233, + "grad_norm": 0.39508244406623994, + "learning_rate": 0.00010462679672733489, + "loss": 2.2343, + "step": 3474 + }, + { + "epoch": 0.7968355881678514, + "grad_norm": 0.3770479027837959, + "learning_rate": 0.00010439958821489165, + "loss": 2.2722, + "step": 3475 + }, + { + "epoch": 0.7970648933730796, + "grad_norm": 0.39381144256024486, + "learning_rate": 0.00010417259791298939, + "loss": 2.449, + "step": 3476 + }, + { + "epoch": 0.7972941985783077, + "grad_norm": 0.4045416492215021, + "learning_rate": 0.00010394582594683428, + "loss": 2.4355, + "step": 3477 + }, + { + "epoch": 0.7975235037835359, + "grad_norm": 0.43206530438884705, + "learning_rate": 0.0001037192724415122, + "loss": 2.2417, + "step": 3478 + }, + { + "epoch": 0.797752808988764, + "grad_norm": 0.37855382021631695, + "learning_rate": 0.0001034929375219884, + "loss": 2.3832, + "step": 3479 + }, + { + "epoch": 0.7979821141939922, + "grad_norm": 0.442914900928133, + "learning_rate": 0.00010326682131310816, + "loss": 2.3434, + "step": 3480 + }, + { + "epoch": 0.7982114193992204, + "grad_norm": 0.4307007463441936, + "learning_rate": 0.00010304092393959514, + "loss": 2.436, + "step": 3481 + }, + { + "epoch": 0.7984407246044485, + "grad_norm": 0.3865650019740518, + "learning_rate": 0.00010281524552605315, + "loss": 2.3795, + "step": 3482 + }, + { + "epoch": 0.7986700298096767, + "grad_norm": 0.4134636344685394, + "learning_rate": 0.00010258978619696468, + "loss": 2.3688, + "step": 3483 + }, + { + "epoch": 0.7988993350149048, + "grad_norm": 0.4169112893578526, + "learning_rate": 0.00010236454607669199, + "loss": 2.337, + "step": 3484 + }, + { + "epoch": 0.799128640220133, + "grad_norm": 0.41865195950237244, + "learning_rate": 0.00010213952528947551, + "loss": 2.4385, + "step": 3485 + }, + { + "epoch": 0.7993579454253612, + "grad_norm": 0.4022567597585622, + "learning_rate": 0.00010191472395943552, + "loss": 2.4683, + "step": 3486 + }, + { + "epoch": 0.7995872506305893, + "grad_norm": 0.4154365657485822, + "learning_rate": 0.00010169014221057089, + "loss": 2.2738, + "step": 3487 + }, + { + "epoch": 0.7998165558358175, + "grad_norm": 0.3791900369437855, + "learning_rate": 0.00010146578016675934, + "loss": 2.3884, + "step": 3488 + }, + { + "epoch": 0.8000458610410456, + "grad_norm": 0.4035851107931036, + "learning_rate": 0.00010124163795175734, + "loss": 2.3573, + "step": 3489 + }, + { + "epoch": 0.8002751662462738, + "grad_norm": 0.40688074185158224, + "learning_rate": 0.0001010177156892006, + "loss": 2.4874, + "step": 3490 + }, + { + "epoch": 0.8005044714515019, + "grad_norm": 0.3960797087400139, + "learning_rate": 0.00010079401350260287, + "loss": 2.2502, + "step": 3491 + }, + { + "epoch": 0.8007337766567301, + "grad_norm": 1.1890110644341723, + "learning_rate": 0.0001005705315153569, + "loss": 2.3812, + "step": 3492 + }, + { + "epoch": 0.8009630818619583, + "grad_norm": 0.3916352236909438, + "learning_rate": 0.00010034726985073362, + "loss": 2.4433, + "step": 3493 + }, + { + "epoch": 0.8011923870671864, + "grad_norm": 0.7384677956782891, + "learning_rate": 0.00010012422863188314, + "loss": 2.542, + "step": 3494 + }, + { + "epoch": 0.8014216922724146, + "grad_norm": 0.38898906465530225, + "learning_rate": 9.9901407981833e-05, + "loss": 2.4061, + "step": 3495 + }, + { + "epoch": 0.8016509974776428, + "grad_norm": 0.42817665665821, + "learning_rate": 9.967880802348989e-05, + "loss": 2.5256, + "step": 3496 + }, + { + "epoch": 0.8018803026828709, + "grad_norm": 0.4202354509617054, + "learning_rate": 9.94564288796384e-05, + "loss": 2.4785, + "step": 3497 + }, + { + "epoch": 0.802109607888099, + "grad_norm": 0.40134636691396314, + "learning_rate": 9.923427067294133e-05, + "loss": 2.4151, + "step": 3498 + }, + { + "epoch": 0.8023389130933272, + "grad_norm": 0.41531505774854577, + "learning_rate": 9.901233352593953e-05, + "loss": 2.331, + "step": 3499 + }, + { + "epoch": 0.8025682182985554, + "grad_norm": 0.40519597729436524, + "learning_rate": 9.879061756105234e-05, + "loss": 2.5124, + "step": 3500 + }, + { + "epoch": 0.8027975235037835, + "grad_norm": 0.41295871675494894, + "learning_rate": 9.856912290057668e-05, + "loss": 2.2938, + "step": 3501 + }, + { + "epoch": 0.8030268287090117, + "grad_norm": 0.4412871691205104, + "learning_rate": 9.834784966668747e-05, + "loss": 2.4293, + "step": 3502 + }, + { + "epoch": 0.8032561339142399, + "grad_norm": 0.4052727102190155, + "learning_rate": 9.812679798143748e-05, + "loss": 2.3318, + "step": 3503 + }, + { + "epoch": 0.803485439119468, + "grad_norm": 0.4160792701992066, + "learning_rate": 9.79059679667575e-05, + "loss": 2.4617, + "step": 3504 + }, + { + "epoch": 0.8037147443246961, + "grad_norm": 0.4101199209278493, + "learning_rate": 9.768535974445586e-05, + "loss": 2.6707, + "step": 3505 + }, + { + "epoch": 0.8039440495299244, + "grad_norm": 0.3939259639045875, + "learning_rate": 9.746497343621857e-05, + "loss": 2.3275, + "step": 3506 + }, + { + "epoch": 0.8041733547351525, + "grad_norm": 0.38697461259738725, + "learning_rate": 9.724480916360906e-05, + "loss": 2.3708, + "step": 3507 + }, + { + "epoch": 0.8044026599403806, + "grad_norm": 0.39826104283766306, + "learning_rate": 9.702486704806896e-05, + "loss": 2.3062, + "step": 3508 + }, + { + "epoch": 0.8046319651456088, + "grad_norm": 0.44098710925793133, + "learning_rate": 9.68051472109162e-05, + "loss": 2.4958, + "step": 3509 + }, + { + "epoch": 0.804861270350837, + "grad_norm": 0.41523725118859817, + "learning_rate": 9.658564977334727e-05, + "loss": 2.3619, + "step": 3510 + }, + { + "epoch": 0.8050905755560651, + "grad_norm": 0.40267592328910756, + "learning_rate": 9.636637485643529e-05, + "loss": 2.4159, + "step": 3511 + }, + { + "epoch": 0.8053198807612932, + "grad_norm": 0.3885781246350191, + "learning_rate": 9.614732258113096e-05, + "loss": 2.5193, + "step": 3512 + }, + { + "epoch": 0.8055491859665215, + "grad_norm": 0.43209463188163083, + "learning_rate": 9.592849306826174e-05, + "loss": 2.4694, + "step": 3513 + }, + { + "epoch": 0.8057784911717496, + "grad_norm": 0.4432943446300484, + "learning_rate": 9.570988643853296e-05, + "loss": 2.5608, + "step": 3514 + }, + { + "epoch": 0.8060077963769777, + "grad_norm": 0.4402406025064212, + "learning_rate": 9.549150281252633e-05, + "loss": 2.6389, + "step": 3515 + }, + { + "epoch": 0.806237101582206, + "grad_norm": 0.4114711062708968, + "learning_rate": 9.527334231070084e-05, + "loss": 2.3228, + "step": 3516 + }, + { + "epoch": 0.8064664067874341, + "grad_norm": 0.45719826422369986, + "learning_rate": 9.505540505339223e-05, + "loss": 2.2851, + "step": 3517 + }, + { + "epoch": 0.8066957119926622, + "grad_norm": 0.4347674031998547, + "learning_rate": 9.483769116081348e-05, + "loss": 2.4252, + "step": 3518 + }, + { + "epoch": 0.8069250171978903, + "grad_norm": 0.42327771060302494, + "learning_rate": 9.4620200753054e-05, + "loss": 2.4005, + "step": 3519 + }, + { + "epoch": 0.8071543224031186, + "grad_norm": 0.42829251474185653, + "learning_rate": 9.440293395008004e-05, + "loss": 2.4513, + "step": 3520 + }, + { + "epoch": 0.8073836276083467, + "grad_norm": 0.41540301671254104, + "learning_rate": 9.418589087173441e-05, + "loss": 2.3061, + "step": 3521 + }, + { + "epoch": 0.8076129328135748, + "grad_norm": 0.39640455091790294, + "learning_rate": 9.396907163773694e-05, + "loss": 2.3638, + "step": 3522 + }, + { + "epoch": 0.8078422380188031, + "grad_norm": 0.3849541433470688, + "learning_rate": 9.375247636768325e-05, + "loss": 2.3949, + "step": 3523 + }, + { + "epoch": 0.8080715432240312, + "grad_norm": 0.3768313231841476, + "learning_rate": 9.35361051810461e-05, + "loss": 2.4375, + "step": 3524 + }, + { + "epoch": 0.8083008484292593, + "grad_norm": 0.40378088963890707, + "learning_rate": 9.331995819717443e-05, + "loss": 2.52, + "step": 3525 + }, + { + "epoch": 0.8085301536344875, + "grad_norm": 0.4232648040437427, + "learning_rate": 9.310403553529334e-05, + "loss": 2.4849, + "step": 3526 + }, + { + "epoch": 0.8087594588397157, + "grad_norm": 0.41982495855379237, + "learning_rate": 9.288833731450419e-05, + "loss": 2.4592, + "step": 3527 + }, + { + "epoch": 0.8089887640449438, + "grad_norm": 0.41383790090133843, + "learning_rate": 9.267286365378503e-05, + "loss": 2.3919, + "step": 3528 + }, + { + "epoch": 0.809218069250172, + "grad_norm": 0.395770351025285, + "learning_rate": 9.245761467198948e-05, + "loss": 2.3311, + "step": 3529 + }, + { + "epoch": 0.8094473744554002, + "grad_norm": 0.4626602154746187, + "learning_rate": 9.22425904878475e-05, + "loss": 2.48, + "step": 3530 + }, + { + "epoch": 0.8096766796606283, + "grad_norm": 0.4068872824065589, + "learning_rate": 9.20277912199648e-05, + "loss": 2.3138, + "step": 3531 + }, + { + "epoch": 0.8099059848658564, + "grad_norm": 0.4458957530527226, + "learning_rate": 9.181321698682354e-05, + "loss": 2.499, + "step": 3532 + }, + { + "epoch": 0.8101352900710846, + "grad_norm": 0.41829572599051784, + "learning_rate": 9.159886790678123e-05, + "loss": 2.5768, + "step": 3533 + }, + { + "epoch": 0.8103645952763128, + "grad_norm": 0.4231935435643373, + "learning_rate": 9.138474409807145e-05, + "loss": 2.4402, + "step": 3534 + }, + { + "epoch": 0.8105939004815409, + "grad_norm": 0.44584139392127997, + "learning_rate": 9.11708456788033e-05, + "loss": 2.4282, + "step": 3535 + }, + { + "epoch": 0.8108232056867691, + "grad_norm": 0.37817240142678216, + "learning_rate": 9.095717276696214e-05, + "loss": 2.5451, + "step": 3536 + }, + { + "epoch": 0.8110525108919973, + "grad_norm": 0.4115096079972369, + "learning_rate": 9.074372548040793e-05, + "loss": 2.4718, + "step": 3537 + }, + { + "epoch": 0.8112818160972254, + "grad_norm": 0.37998110364609927, + "learning_rate": 9.053050393687723e-05, + "loss": 2.3295, + "step": 3538 + }, + { + "epoch": 0.8115111213024536, + "grad_norm": 0.39184993788150985, + "learning_rate": 9.031750825398145e-05, + "loss": 2.3079, + "step": 3539 + }, + { + "epoch": 0.8117404265076817, + "grad_norm": 0.40251340681159, + "learning_rate": 9.010473854920759e-05, + "loss": 2.4015, + "step": 3540 + }, + { + "epoch": 0.8119697317129099, + "grad_norm": 0.43169632130937524, + "learning_rate": 8.98921949399179e-05, + "loss": 2.3616, + "step": 3541 + }, + { + "epoch": 0.812199036918138, + "grad_norm": 0.4090569896905003, + "learning_rate": 8.967987754335022e-05, + "loss": 2.3309, + "step": 3542 + }, + { + "epoch": 0.8124283421233662, + "grad_norm": 0.40843528653456296, + "learning_rate": 8.94677864766173e-05, + "loss": 2.3419, + "step": 3543 + }, + { + "epoch": 0.8126576473285944, + "grad_norm": 0.4073445890366409, + "learning_rate": 8.925592185670711e-05, + "loss": 2.4737, + "step": 3544 + }, + { + "epoch": 0.8128869525338225, + "grad_norm": 0.40051948522935654, + "learning_rate": 8.904428380048269e-05, + "loss": 2.4231, + "step": 3545 + }, + { + "epoch": 0.8131162577390507, + "grad_norm": 0.39416429623939964, + "learning_rate": 8.883287242468242e-05, + "loss": 2.4116, + "step": 3546 + }, + { + "epoch": 0.8133455629442788, + "grad_norm": 0.41726063169918043, + "learning_rate": 8.862168784591929e-05, + "loss": 2.4182, + "step": 3547 + }, + { + "epoch": 0.813574868149507, + "grad_norm": 0.38398675759763246, + "learning_rate": 8.841073018068125e-05, + "loss": 2.4663, + "step": 3548 + }, + { + "epoch": 0.8138041733547352, + "grad_norm": 0.4049866566960841, + "learning_rate": 8.819999954533115e-05, + "loss": 2.4996, + "step": 3549 + }, + { + "epoch": 0.8140334785599633, + "grad_norm": 0.3902428093791268, + "learning_rate": 8.798949605610662e-05, + "loss": 2.5462, + "step": 3550 + }, + { + "epoch": 0.8142627837651915, + "grad_norm": 0.41302738544378087, + "learning_rate": 8.777921982911996e-05, + "loss": 2.333, + "step": 3551 + }, + { + "epoch": 0.8144920889704196, + "grad_norm": 0.41743122214109984, + "learning_rate": 8.756917098035799e-05, + "loss": 2.4903, + "step": 3552 + }, + { + "epoch": 0.8147213941756478, + "grad_norm": 0.4444502004522551, + "learning_rate": 8.735934962568253e-05, + "loss": 2.4159, + "step": 3553 + }, + { + "epoch": 0.8149506993808759, + "grad_norm": 0.4122920986742243, + "learning_rate": 8.714975588082951e-05, + "loss": 2.3207, + "step": 3554 + }, + { + "epoch": 0.8151800045861041, + "grad_norm": 0.42340966282474385, + "learning_rate": 8.694038986140945e-05, + "loss": 2.3203, + "step": 3555 + }, + { + "epoch": 0.8154093097913323, + "grad_norm": 0.41485984559666855, + "learning_rate": 8.673125168290713e-05, + "loss": 2.3831, + "step": 3556 + }, + { + "epoch": 0.8156386149965604, + "grad_norm": 0.4346299045986973, + "learning_rate": 8.652234146068206e-05, + "loss": 2.4194, + "step": 3557 + }, + { + "epoch": 0.8158679202017886, + "grad_norm": 0.3764827331102309, + "learning_rate": 8.631365930996732e-05, + "loss": 2.3902, + "step": 3558 + }, + { + "epoch": 0.8160972254070168, + "grad_norm": 0.3953960627714742, + "learning_rate": 8.610520534587086e-05, + "loss": 2.3785, + "step": 3559 + }, + { + "epoch": 0.8163265306122449, + "grad_norm": 0.41648292727584746, + "learning_rate": 8.589697968337446e-05, + "loss": 2.4101, + "step": 3560 + }, + { + "epoch": 0.816555835817473, + "grad_norm": 0.4438777549168453, + "learning_rate": 8.568898243733397e-05, + "loss": 2.4751, + "step": 3561 + }, + { + "epoch": 0.8167851410227012, + "grad_norm": 0.42545203274544435, + "learning_rate": 8.548121372247918e-05, + "loss": 2.5096, + "step": 3562 + }, + { + "epoch": 0.8170144462279294, + "grad_norm": 0.4008490376149219, + "learning_rate": 8.527367365341409e-05, + "loss": 2.298, + "step": 3563 + }, + { + "epoch": 0.8172437514331575, + "grad_norm": 0.4080171434066841, + "learning_rate": 8.50663623446164e-05, + "loss": 2.5604, + "step": 3564 + }, + { + "epoch": 0.8174730566383857, + "grad_norm": 0.4269369704845385, + "learning_rate": 8.485927991043757e-05, + "loss": 2.141, + "step": 3565 + }, + { + "epoch": 0.8177023618436139, + "grad_norm": 0.40788824377220145, + "learning_rate": 8.465242646510279e-05, + "loss": 2.3235, + "step": 3566 + }, + { + "epoch": 0.817931667048842, + "grad_norm": 0.3965804528196065, + "learning_rate": 8.444580212271125e-05, + "loss": 2.3626, + "step": 3567 + }, + { + "epoch": 0.8181609722540701, + "grad_norm": 0.41031659015042343, + "learning_rate": 8.423940699723548e-05, + "loss": 2.4775, + "step": 3568 + }, + { + "epoch": 0.8183902774592984, + "grad_norm": 0.3953307449147809, + "learning_rate": 8.403324120252159e-05, + "loss": 2.4466, + "step": 3569 + }, + { + "epoch": 0.8186195826645265, + "grad_norm": 0.40201813918483575, + "learning_rate": 8.382730485228923e-05, + "loss": 2.4143, + "step": 3570 + }, + { + "epoch": 0.8188488878697546, + "grad_norm": 0.4014437880196209, + "learning_rate": 8.362159806013175e-05, + "loss": 2.3919, + "step": 3571 + }, + { + "epoch": 0.8190781930749828, + "grad_norm": 0.41596964420427135, + "learning_rate": 8.341612093951523e-05, + "loss": 2.596, + "step": 3572 + }, + { + "epoch": 0.819307498280211, + "grad_norm": 0.4045933554300166, + "learning_rate": 8.321087360377988e-05, + "loss": 2.3227, + "step": 3573 + }, + { + "epoch": 0.8195368034854391, + "grad_norm": 0.45401862921557623, + "learning_rate": 8.300585616613854e-05, + "loss": 2.5503, + "step": 3574 + }, + { + "epoch": 0.8197661086906672, + "grad_norm": 0.38951329629460174, + "learning_rate": 8.280106873967752e-05, + "loss": 2.4166, + "step": 3575 + }, + { + "epoch": 0.8199954138958955, + "grad_norm": 0.4451926808271857, + "learning_rate": 8.259651143735603e-05, + "loss": 2.3637, + "step": 3576 + }, + { + "epoch": 0.8202247191011236, + "grad_norm": 0.4185179005828439, + "learning_rate": 8.239218437200679e-05, + "loss": 2.4189, + "step": 3577 + }, + { + "epoch": 0.8204540243063517, + "grad_norm": 0.4127695076924484, + "learning_rate": 8.218808765633512e-05, + "loss": 2.3131, + "step": 3578 + }, + { + "epoch": 0.82068332951158, + "grad_norm": 0.44422227992527924, + "learning_rate": 8.198422140291939e-05, + "loss": 2.4348, + "step": 3579 + }, + { + "epoch": 0.8209126347168081, + "grad_norm": 0.43223971843735, + "learning_rate": 8.178058572421077e-05, + "loss": 2.5096, + "step": 3580 + }, + { + "epoch": 0.8211419399220362, + "grad_norm": 0.39649851690020943, + "learning_rate": 8.157718073253351e-05, + "loss": 2.4363, + "step": 3581 + }, + { + "epoch": 0.8213712451272643, + "grad_norm": 0.4553377409810759, + "learning_rate": 8.137400654008436e-05, + "loss": 2.4559, + "step": 3582 + }, + { + "epoch": 0.8216005503324926, + "grad_norm": 0.41288350725272516, + "learning_rate": 8.117106325893287e-05, + "loss": 2.3212, + "step": 3583 + }, + { + "epoch": 0.8218298555377207, + "grad_norm": 0.4246928922825332, + "learning_rate": 8.096835100102107e-05, + "loss": 2.5037, + "step": 3584 + }, + { + "epoch": 0.8220591607429488, + "grad_norm": 0.40957635261828473, + "learning_rate": 8.076586987816404e-05, + "loss": 2.5128, + "step": 3585 + }, + { + "epoch": 0.8222884659481771, + "grad_norm": 0.45990793111157, + "learning_rate": 8.056362000204847e-05, + "loss": 2.3218, + "step": 3586 + }, + { + "epoch": 0.8225177711534052, + "grad_norm": 0.4079797137884439, + "learning_rate": 8.036160148423449e-05, + "loss": 2.6416, + "step": 3587 + }, + { + "epoch": 0.8227470763586333, + "grad_norm": 0.3942149774458509, + "learning_rate": 8.0159814436154e-05, + "loss": 2.4269, + "step": 3588 + }, + { + "epoch": 0.8229763815638615, + "grad_norm": 0.43743271248180654, + "learning_rate": 7.995825896911141e-05, + "loss": 2.3048, + "step": 3589 + }, + { + "epoch": 0.8232056867690897, + "grad_norm": 0.9368730067886182, + "learning_rate": 7.975693519428318e-05, + "loss": 2.3521, + "step": 3590 + }, + { + "epoch": 0.8234349919743178, + "grad_norm": 0.4001131239046266, + "learning_rate": 7.955584322271853e-05, + "loss": 2.5256, + "step": 3591 + }, + { + "epoch": 0.8236642971795459, + "grad_norm": 0.4156119360438316, + "learning_rate": 7.935498316533818e-05, + "loss": 2.2361, + "step": 3592 + }, + { + "epoch": 0.8238936023847742, + "grad_norm": 0.4006838773480312, + "learning_rate": 7.915435513293523e-05, + "loss": 2.4902, + "step": 3593 + }, + { + "epoch": 0.8241229075900023, + "grad_norm": 0.4282694133282178, + "learning_rate": 7.89539592361747e-05, + "loss": 2.5546, + "step": 3594 + }, + { + "epoch": 0.8243522127952304, + "grad_norm": 0.4632021146714906, + "learning_rate": 7.875379558559387e-05, + "loss": 2.5246, + "step": 3595 + }, + { + "epoch": 0.8245815180004586, + "grad_norm": 0.4393012387949964, + "learning_rate": 7.85538642916015e-05, + "loss": 2.5298, + "step": 3596 + }, + { + "epoch": 0.8248108232056868, + "grad_norm": 0.4142815345784967, + "learning_rate": 7.835416546447838e-05, + "loss": 2.4754, + "step": 3597 + }, + { + "epoch": 0.8250401284109149, + "grad_norm": 0.3908698998299526, + "learning_rate": 7.81546992143769e-05, + "loss": 2.3995, + "step": 3598 + }, + { + "epoch": 0.8252694336161431, + "grad_norm": 0.42675960308173966, + "learning_rate": 7.795546565132167e-05, + "loss": 2.3663, + "step": 3599 + }, + { + "epoch": 0.8254987388213713, + "grad_norm": 0.41126178711291517, + "learning_rate": 7.775646488520816e-05, + "loss": 2.4015, + "step": 3600 + }, + { + "epoch": 0.8257280440265994, + "grad_norm": 0.39765014541017013, + "learning_rate": 7.755769702580412e-05, + "loss": 2.4018, + "step": 3601 + }, + { + "epoch": 0.8259573492318275, + "grad_norm": 0.4039045777606054, + "learning_rate": 7.735916218274857e-05, + "loss": 2.3519, + "step": 3602 + }, + { + "epoch": 0.8261866544370557, + "grad_norm": 0.3997444467330455, + "learning_rate": 7.716086046555193e-05, + "loss": 2.3906, + "step": 3603 + }, + { + "epoch": 0.8264159596422839, + "grad_norm": 0.44588523293865157, + "learning_rate": 7.696279198359607e-05, + "loss": 2.3627, + "step": 3604 + }, + { + "epoch": 0.826645264847512, + "grad_norm": 0.4312449598878249, + "learning_rate": 7.676495684613432e-05, + "loss": 2.2729, + "step": 3605 + }, + { + "epoch": 0.8268745700527402, + "grad_norm": 0.4311606008201497, + "learning_rate": 7.656735516229125e-05, + "loss": 2.2767, + "step": 3606 + }, + { + "epoch": 0.8271038752579684, + "grad_norm": 0.4298996386882859, + "learning_rate": 7.636998704106252e-05, + "loss": 2.5339, + "step": 3607 + }, + { + "epoch": 0.8273331804631965, + "grad_norm": 0.4153284612009962, + "learning_rate": 7.617285259131501e-05, + "loss": 2.5001, + "step": 3608 + }, + { + "epoch": 0.8275624856684247, + "grad_norm": 0.39997124357996067, + "learning_rate": 7.597595192178702e-05, + "loss": 2.4284, + "step": 3609 + }, + { + "epoch": 0.8277917908736528, + "grad_norm": 0.40709983427374646, + "learning_rate": 7.577928514108745e-05, + "loss": 2.3117, + "step": 3610 + }, + { + "epoch": 0.828021096078881, + "grad_norm": 0.40445763380293487, + "learning_rate": 7.558285235769646e-05, + "loss": 2.4476, + "step": 3611 + }, + { + "epoch": 0.8282504012841091, + "grad_norm": 0.40245102631349816, + "learning_rate": 7.538665367996484e-05, + "loss": 2.3842, + "step": 3612 + }, + { + "epoch": 0.8284797064893373, + "grad_norm": 0.4345813411339542, + "learning_rate": 7.519068921611494e-05, + "loss": 2.4437, + "step": 3613 + }, + { + "epoch": 0.8287090116945655, + "grad_norm": 0.414392805610035, + "learning_rate": 7.499495907423887e-05, + "loss": 2.397, + "step": 3614 + }, + { + "epoch": 0.8289383168997936, + "grad_norm": 0.4329893963552301, + "learning_rate": 7.479946336230047e-05, + "loss": 2.4389, + "step": 3615 + }, + { + "epoch": 0.8291676221050218, + "grad_norm": 0.3854615850554792, + "learning_rate": 7.46042021881338e-05, + "loss": 2.3904, + "step": 3616 + }, + { + "epoch": 0.8293969273102499, + "grad_norm": 0.4373203505969336, + "learning_rate": 7.440917565944349e-05, + "loss": 2.3972, + "step": 3617 + }, + { + "epoch": 0.8296262325154781, + "grad_norm": 0.42900361665514036, + "learning_rate": 7.421438388380486e-05, + "loss": 2.4022, + "step": 3618 + }, + { + "epoch": 0.8298555377207063, + "grad_norm": 0.4329294126685716, + "learning_rate": 7.4019826968664e-05, + "loss": 2.4529, + "step": 3619 + }, + { + "epoch": 0.8300848429259344, + "grad_norm": 0.40924540292058853, + "learning_rate": 7.382550502133694e-05, + "loss": 2.4388, + "step": 3620 + }, + { + "epoch": 0.8303141481311626, + "grad_norm": 0.4371530138115317, + "learning_rate": 7.363141814901053e-05, + "loss": 2.3835, + "step": 3621 + }, + { + "epoch": 0.8305434533363908, + "grad_norm": 0.4244398492337022, + "learning_rate": 7.343756645874162e-05, + "loss": 2.425, + "step": 3622 + }, + { + "epoch": 0.8307727585416189, + "grad_norm": 0.38980683481396144, + "learning_rate": 7.32439500574577e-05, + "loss": 2.3357, + "step": 3623 + }, + { + "epoch": 0.831002063746847, + "grad_norm": 0.42936441584572094, + "learning_rate": 7.305056905195618e-05, + "loss": 2.5282, + "step": 3624 + }, + { + "epoch": 0.8312313689520752, + "grad_norm": 0.4016215969455211, + "learning_rate": 7.285742354890473e-05, + "loss": 2.405, + "step": 3625 + }, + { + "epoch": 0.8314606741573034, + "grad_norm": 0.44877835648738956, + "learning_rate": 7.266451365484105e-05, + "loss": 2.4681, + "step": 3626 + }, + { + "epoch": 0.8316899793625315, + "grad_norm": 0.4046986568809434, + "learning_rate": 7.247183947617325e-05, + "loss": 2.4565, + "step": 3627 + }, + { + "epoch": 0.8319192845677597, + "grad_norm": 0.40232642232750426, + "learning_rate": 7.227940111917863e-05, + "loss": 2.3521, + "step": 3628 + }, + { + "epoch": 0.8321485897729879, + "grad_norm": 0.40269243604507543, + "learning_rate": 7.20871986900053e-05, + "loss": 2.5575, + "step": 3629 + }, + { + "epoch": 0.832377894978216, + "grad_norm": 0.41939601542090904, + "learning_rate": 7.189523229467071e-05, + "loss": 2.468, + "step": 3630 + }, + { + "epoch": 0.8326072001834441, + "grad_norm": 0.39537561023640794, + "learning_rate": 7.170350203906218e-05, + "loss": 2.411, + "step": 3631 + }, + { + "epoch": 0.8328365053886724, + "grad_norm": 0.4408669431717679, + "learning_rate": 7.151200802893681e-05, + "loss": 2.4096, + "step": 3632 + }, + { + "epoch": 0.8330658105939005, + "grad_norm": 0.4179422732039931, + "learning_rate": 7.132075036992158e-05, + "loss": 2.4327, + "step": 3633 + }, + { + "epoch": 0.8332951157991286, + "grad_norm": 0.3937953804014827, + "learning_rate": 7.112972916751287e-05, + "loss": 2.3313, + "step": 3634 + }, + { + "epoch": 0.8335244210043568, + "grad_norm": 0.39689178344567977, + "learning_rate": 7.093894452707666e-05, + "loss": 2.3937, + "step": 3635 + }, + { + "epoch": 0.833753726209585, + "grad_norm": 0.4276784862497425, + "learning_rate": 7.074839655384835e-05, + "loss": 2.4945, + "step": 3636 + }, + { + "epoch": 0.8339830314148131, + "grad_norm": 0.4275367799881659, + "learning_rate": 7.055808535293334e-05, + "loss": 2.3886, + "step": 3637 + }, + { + "epoch": 0.8342123366200412, + "grad_norm": 0.4064162428295308, + "learning_rate": 7.036801102930551e-05, + "loss": 2.4071, + "step": 3638 + }, + { + "epoch": 0.8344416418252695, + "grad_norm": 0.39795789891740735, + "learning_rate": 7.017817368780888e-05, + "loss": 2.3844, + "step": 3639 + }, + { + "epoch": 0.8346709470304976, + "grad_norm": 0.41376509252058985, + "learning_rate": 6.998857343315634e-05, + "loss": 2.4819, + "step": 3640 + }, + { + "epoch": 0.8349002522357257, + "grad_norm": 0.4390437736381063, + "learning_rate": 6.979921036993042e-05, + "loss": 2.4053, + "step": 3641 + }, + { + "epoch": 0.835129557440954, + "grad_norm": 0.43239748497178465, + "learning_rate": 6.9610084602582e-05, + "loss": 2.4307, + "step": 3642 + }, + { + "epoch": 0.8353588626461821, + "grad_norm": 0.4453729466749027, + "learning_rate": 6.942119623543202e-05, + "loss": 2.4509, + "step": 3643 + }, + { + "epoch": 0.8355881678514102, + "grad_norm": 0.4325181690435561, + "learning_rate": 6.923254537266988e-05, + "loss": 2.4394, + "step": 3644 + }, + { + "epoch": 0.8358174730566383, + "grad_norm": 0.38985025863550976, + "learning_rate": 6.904413211835414e-05, + "loss": 2.4895, + "step": 3645 + }, + { + "epoch": 0.8360467782618666, + "grad_norm": 0.40496848247963113, + "learning_rate": 6.885595657641213e-05, + "loss": 2.3068, + "step": 3646 + }, + { + "epoch": 0.8362760834670947, + "grad_norm": 0.40723914609996736, + "learning_rate": 6.866801885064056e-05, + "loss": 2.5038, + "step": 3647 + }, + { + "epoch": 0.8365053886723228, + "grad_norm": 0.4151520226071435, + "learning_rate": 6.848031904470441e-05, + "loss": 2.5018, + "step": 3648 + }, + { + "epoch": 0.8367346938775511, + "grad_norm": 0.4530529381525359, + "learning_rate": 6.829285726213769e-05, + "loss": 2.4151, + "step": 3649 + }, + { + "epoch": 0.8369639990827792, + "grad_norm": 0.3875785962827794, + "learning_rate": 6.810563360634297e-05, + "loss": 2.3976, + "step": 3650 + }, + { + "epoch": 0.8371933042880073, + "grad_norm": 0.4088893370730233, + "learning_rate": 6.79186481805918e-05, + "loss": 2.3538, + "step": 3651 + }, + { + "epoch": 0.8374226094932355, + "grad_norm": 0.4177853321427478, + "learning_rate": 6.773190108802374e-05, + "loss": 2.3766, + "step": 3652 + }, + { + "epoch": 0.8376519146984637, + "grad_norm": 0.44014152495723796, + "learning_rate": 6.754539243164754e-05, + "loss": 2.4568, + "step": 3653 + }, + { + "epoch": 0.8378812199036918, + "grad_norm": 0.4588513997110853, + "learning_rate": 6.735912231433993e-05, + "loss": 2.5498, + "step": 3654 + }, + { + "epoch": 0.8381105251089199, + "grad_norm": 0.7025805746666476, + "learning_rate": 6.717309083884654e-05, + "loss": 2.3546, + "step": 3655 + }, + { + "epoch": 0.8383398303141482, + "grad_norm": 0.40694106640313077, + "learning_rate": 6.698729810778065e-05, + "loss": 2.4124, + "step": 3656 + }, + { + "epoch": 0.8385691355193763, + "grad_norm": 0.4578649285641327, + "learning_rate": 6.680174422362468e-05, + "loss": 2.4636, + "step": 3657 + }, + { + "epoch": 0.8387984407246044, + "grad_norm": 0.420659801815045, + "learning_rate": 6.661642928872874e-05, + "loss": 2.4134, + "step": 3658 + }, + { + "epoch": 0.8390277459298326, + "grad_norm": 0.42029535666292805, + "learning_rate": 6.643135340531136e-05, + "loss": 2.4331, + "step": 3659 + }, + { + "epoch": 0.8392570511350608, + "grad_norm": 0.3947303995873364, + "learning_rate": 6.6246516675459e-05, + "loss": 2.3532, + "step": 3660 + }, + { + "epoch": 0.8394863563402889, + "grad_norm": 0.4247491338121949, + "learning_rate": 6.606191920112664e-05, + "loss": 2.3545, + "step": 3661 + }, + { + "epoch": 0.8397156615455171, + "grad_norm": 0.412420290512582, + "learning_rate": 6.587756108413684e-05, + "loss": 2.4008, + "step": 3662 + }, + { + "epoch": 0.8399449667507453, + "grad_norm": 0.4173376655824018, + "learning_rate": 6.569344242618036e-05, + "loss": 2.4722, + "step": 3663 + }, + { + "epoch": 0.8401742719559734, + "grad_norm": 0.4065372051366088, + "learning_rate": 6.550956332881564e-05, + "loss": 2.3865, + "step": 3664 + }, + { + "epoch": 0.8404035771612015, + "grad_norm": 0.41671361914897354, + "learning_rate": 6.532592389346958e-05, + "loss": 2.4418, + "step": 3665 + }, + { + "epoch": 0.8406328823664297, + "grad_norm": 0.43439746470351387, + "learning_rate": 6.514252422143591e-05, + "loss": 2.5645, + "step": 3666 + }, + { + "epoch": 0.8408621875716579, + "grad_norm": 0.44477493454389117, + "learning_rate": 6.495936441387713e-05, + "loss": 2.4773, + "step": 3667 + }, + { + "epoch": 0.841091492776886, + "grad_norm": 0.4313397259352959, + "learning_rate": 6.477644457182275e-05, + "loss": 2.4963, + "step": 3668 + }, + { + "epoch": 0.8413207979821142, + "grad_norm": 0.41196423586182296, + "learning_rate": 6.459376479617013e-05, + "loss": 2.3811, + "step": 3669 + }, + { + "epoch": 0.8415501031873424, + "grad_norm": 0.4827253939786443, + "learning_rate": 6.441132518768417e-05, + "loss": 2.4368, + "step": 3670 + }, + { + "epoch": 0.8417794083925705, + "grad_norm": 0.4545399490532656, + "learning_rate": 6.422912584699752e-05, + "loss": 2.3148, + "step": 3671 + }, + { + "epoch": 0.8420087135977987, + "grad_norm": 0.37311210966889313, + "learning_rate": 6.404716687461004e-05, + "loss": 2.3827, + "step": 3672 + }, + { + "epoch": 0.8422380188030268, + "grad_norm": 0.44476949001144134, + "learning_rate": 6.386544837088904e-05, + "loss": 2.4454, + "step": 3673 + }, + { + "epoch": 0.842467324008255, + "grad_norm": 0.392363236373028, + "learning_rate": 6.36839704360691e-05, + "loss": 2.5104, + "step": 3674 + }, + { + "epoch": 0.8426966292134831, + "grad_norm": 0.38410944379713935, + "learning_rate": 6.350273317025251e-05, + "loss": 2.5182, + "step": 3675 + }, + { + "epoch": 0.8429259344187113, + "grad_norm": 0.4152235470661938, + "learning_rate": 6.33217366734084e-05, + "loss": 2.3636, + "step": 3676 + }, + { + "epoch": 0.8431552396239395, + "grad_norm": 0.4233296521983836, + "learning_rate": 6.314098104537324e-05, + "loss": 2.4151, + "step": 3677 + }, + { + "epoch": 0.8433845448291676, + "grad_norm": 0.42181164526328574, + "learning_rate": 6.29604663858505e-05, + "loss": 2.5057, + "step": 3678 + }, + { + "epoch": 0.8436138500343958, + "grad_norm": 0.4007251040147993, + "learning_rate": 6.278019279441122e-05, + "loss": 2.2689, + "step": 3679 + }, + { + "epoch": 0.8438431552396239, + "grad_norm": 0.40560028806530163, + "learning_rate": 6.260016037049259e-05, + "loss": 2.2521, + "step": 3680 + }, + { + "epoch": 0.8440724604448521, + "grad_norm": 0.4113068262912769, + "learning_rate": 6.242036921339972e-05, + "loss": 2.4089, + "step": 3681 + }, + { + "epoch": 0.8443017656500803, + "grad_norm": 0.3989308728168822, + "learning_rate": 6.224081942230403e-05, + "loss": 2.5117, + "step": 3682 + }, + { + "epoch": 0.8445310708553084, + "grad_norm": 0.4054095281322586, + "learning_rate": 6.206151109624402e-05, + "loss": 2.4311, + "step": 3683 + }, + { + "epoch": 0.8447603760605366, + "grad_norm": 0.4458649679191281, + "learning_rate": 6.188244433412482e-05, + "loss": 2.5429, + "step": 3684 + }, + { + "epoch": 0.8449896812657647, + "grad_norm": 0.4109255327030491, + "learning_rate": 6.170361923471868e-05, + "loss": 2.5311, + "step": 3685 + }, + { + "epoch": 0.8452189864709929, + "grad_norm": 0.40151794242209854, + "learning_rate": 6.152503589666425e-05, + "loss": 2.357, + "step": 3686 + }, + { + "epoch": 0.845448291676221, + "grad_norm": 0.3814116663904621, + "learning_rate": 6.134669441846691e-05, + "loss": 2.4413, + "step": 3687 + }, + { + "epoch": 0.8456775968814492, + "grad_norm": 0.4828075249321938, + "learning_rate": 6.116859489849846e-05, + "loss": 2.5472, + "step": 3688 + }, + { + "epoch": 0.8459069020866774, + "grad_norm": 0.4152837228993579, + "learning_rate": 6.099073743499772e-05, + "loss": 2.5016, + "step": 3689 + }, + { + "epoch": 0.8461362072919055, + "grad_norm": 0.4277460508275668, + "learning_rate": 6.081312212606943e-05, + "loss": 2.5023, + "step": 3690 + }, + { + "epoch": 0.8463655124971337, + "grad_norm": 0.49008156242683737, + "learning_rate": 6.063574906968511e-05, + "loss": 2.4236, + "step": 3691 + }, + { + "epoch": 0.8465948177023619, + "grad_norm": 0.4051654050492244, + "learning_rate": 6.045861836368244e-05, + "loss": 2.4215, + "step": 3692 + }, + { + "epoch": 0.84682412290759, + "grad_norm": 0.41597514552972603, + "learning_rate": 6.028173010576582e-05, + "loss": 2.3586, + "step": 3693 + }, + { + "epoch": 0.8470534281128181, + "grad_norm": 0.4384238081384734, + "learning_rate": 6.0105084393505206e-05, + "loss": 2.4169, + "step": 3694 + }, + { + "epoch": 0.8472827333180463, + "grad_norm": 0.4036758068247421, + "learning_rate": 5.9928681324337544e-05, + "loss": 2.3663, + "step": 3695 + }, + { + "epoch": 0.8475120385232745, + "grad_norm": 0.4433267541860026, + "learning_rate": 5.975252099556544e-05, + "loss": 2.408, + "step": 3696 + }, + { + "epoch": 0.8477413437285026, + "grad_norm": 0.4083914255341007, + "learning_rate": 5.957660350435773e-05, + "loss": 2.3394, + "step": 3697 + }, + { + "epoch": 0.8479706489337308, + "grad_norm": 0.40861072286095695, + "learning_rate": 5.940092894774929e-05, + "loss": 2.3341, + "step": 3698 + }, + { + "epoch": 0.848199954138959, + "grad_norm": 0.3997603048610276, + "learning_rate": 5.922549742264122e-05, + "loss": 2.4143, + "step": 3699 + }, + { + "epoch": 0.8484292593441871, + "grad_norm": 0.40375394938784726, + "learning_rate": 5.9050309025800275e-05, + "loss": 2.6126, + "step": 3700 + }, + { + "epoch": 0.8486585645494152, + "grad_norm": 0.4396269545804237, + "learning_rate": 5.8875363853859166e-05, + "loss": 2.4868, + "step": 3701 + }, + { + "epoch": 0.8488878697546435, + "grad_norm": 0.4240624106243453, + "learning_rate": 5.8700662003316366e-05, + "loss": 2.3175, + "step": 3702 + }, + { + "epoch": 0.8491171749598716, + "grad_norm": 0.38991855721334046, + "learning_rate": 5.852620357053651e-05, + "loss": 2.3898, + "step": 3703 + }, + { + "epoch": 0.8493464801650997, + "grad_norm": 0.4295103466351228, + "learning_rate": 5.835198865174956e-05, + "loss": 2.6028, + "step": 3704 + }, + { + "epoch": 0.8495757853703279, + "grad_norm": 0.4035978533771102, + "learning_rate": 5.8178017343051336e-05, + "loss": 2.4329, + "step": 3705 + }, + { + "epoch": 0.8498050905755561, + "grad_norm": 0.39546280723859195, + "learning_rate": 5.800428974040311e-05, + "loss": 2.3216, + "step": 3706 + }, + { + "epoch": 0.8500343957807842, + "grad_norm": 0.4216090561350578, + "learning_rate": 5.783080593963219e-05, + "loss": 2.4578, + "step": 3707 + }, + { + "epoch": 0.8502637009860123, + "grad_norm": 0.4495316627074711, + "learning_rate": 5.7657566036430686e-05, + "loss": 2.3251, + "step": 3708 + }, + { + "epoch": 0.8504930061912406, + "grad_norm": 0.40771076691275837, + "learning_rate": 5.748457012635683e-05, + "loss": 2.317, + "step": 3709 + }, + { + "epoch": 0.8507223113964687, + "grad_norm": 0.4127560527126349, + "learning_rate": 5.731181830483395e-05, + "loss": 2.267, + "step": 3710 + }, + { + "epoch": 0.8509516166016968, + "grad_norm": 0.39554781550232626, + "learning_rate": 5.713931066715078e-05, + "loss": 2.4244, + "step": 3711 + }, + { + "epoch": 0.8511809218069251, + "grad_norm": 0.41048679991373155, + "learning_rate": 5.696704730846136e-05, + "loss": 2.4885, + "step": 3712 + }, + { + "epoch": 0.8514102270121532, + "grad_norm": 0.42329633077066403, + "learning_rate": 5.679502832378497e-05, + "loss": 2.4839, + "step": 3713 + }, + { + "epoch": 0.8516395322173813, + "grad_norm": 0.39568470089776786, + "learning_rate": 5.662325380800631e-05, + "loss": 2.466, + "step": 3714 + }, + { + "epoch": 0.8518688374226095, + "grad_norm": 0.40659443187287136, + "learning_rate": 5.645172385587482e-05, + "loss": 2.3723, + "step": 3715 + }, + { + "epoch": 0.8520981426278377, + "grad_norm": 0.39763201428417116, + "learning_rate": 5.6280438562005434e-05, + "loss": 2.4031, + "step": 3716 + }, + { + "epoch": 0.8523274478330658, + "grad_norm": 0.39702930434451333, + "learning_rate": 5.6109398020877834e-05, + "loss": 2.5075, + "step": 3717 + }, + { + "epoch": 0.8525567530382939, + "grad_norm": 0.3927818069615744, + "learning_rate": 5.593860232683712e-05, + "loss": 2.404, + "step": 3718 + }, + { + "epoch": 0.8527860582435222, + "grad_norm": 0.40832398927160446, + "learning_rate": 5.576805157409265e-05, + "loss": 2.3856, + "step": 3719 + }, + { + "epoch": 0.8530153634487503, + "grad_norm": 0.5785180259086619, + "learning_rate": 5.5597745856719394e-05, + "loss": 2.2192, + "step": 3720 + }, + { + "epoch": 0.8532446686539784, + "grad_norm": 0.4453794023462607, + "learning_rate": 5.542768526865677e-05, + "loss": 2.3459, + "step": 3721 + }, + { + "epoch": 0.8534739738592066, + "grad_norm": 0.3919606999866897, + "learning_rate": 5.525786990370901e-05, + "loss": 2.3644, + "step": 3722 + }, + { + "epoch": 0.8537032790644348, + "grad_norm": 0.4236747160788929, + "learning_rate": 5.508829985554509e-05, + "loss": 2.3261, + "step": 3723 + }, + { + "epoch": 0.8539325842696629, + "grad_norm": 0.4113536130114633, + "learning_rate": 5.4918975217698844e-05, + "loss": 2.3842, + "step": 3724 + }, + { + "epoch": 0.8541618894748911, + "grad_norm": 0.41268652677841944, + "learning_rate": 5.474989608356856e-05, + "loss": 2.473, + "step": 3725 + }, + { + "epoch": 0.8543911946801193, + "grad_norm": 0.4130265297030074, + "learning_rate": 5.458106254641715e-05, + "loss": 2.4206, + "step": 3726 + }, + { + "epoch": 0.8546204998853474, + "grad_norm": 0.3939304547253736, + "learning_rate": 5.441247469937194e-05, + "loss": 2.3565, + "step": 3727 + }, + { + "epoch": 0.8548498050905755, + "grad_norm": 0.47586154941121384, + "learning_rate": 5.424413263542522e-05, + "loss": 2.3426, + "step": 3728 + }, + { + "epoch": 0.8550791102958037, + "grad_norm": 0.42567642119567783, + "learning_rate": 5.407603644743286e-05, + "loss": 2.433, + "step": 3729 + }, + { + "epoch": 0.8553084155010319, + "grad_norm": 0.40452181172855795, + "learning_rate": 5.390818622811594e-05, + "loss": 2.398, + "step": 3730 + }, + { + "epoch": 0.85553772070626, + "grad_norm": 0.4112927616202549, + "learning_rate": 5.374058207005944e-05, + "loss": 2.3592, + "step": 3731 + }, + { + "epoch": 0.8557670259114882, + "grad_norm": 0.4065363652984248, + "learning_rate": 5.357322406571269e-05, + "loss": 2.4406, + "step": 3732 + }, + { + "epoch": 0.8559963311167164, + "grad_norm": 0.41326600291811943, + "learning_rate": 5.3406112307389066e-05, + "loss": 2.3222, + "step": 3733 + }, + { + "epoch": 0.8562256363219445, + "grad_norm": 0.4392726940582819, + "learning_rate": 5.3239246887266554e-05, + "loss": 2.5164, + "step": 3734 + }, + { + "epoch": 0.8564549415271727, + "grad_norm": 0.4250902458811768, + "learning_rate": 5.3072627897386926e-05, + "loss": 2.2847, + "step": 3735 + }, + { + "epoch": 0.8566842467324008, + "grad_norm": 0.4433869345237931, + "learning_rate": 5.29062554296561e-05, + "loss": 2.3006, + "step": 3736 + }, + { + "epoch": 0.856913551937629, + "grad_norm": 0.3943436674156202, + "learning_rate": 5.27401295758439e-05, + "loss": 2.4061, + "step": 3737 + }, + { + "epoch": 0.8571428571428571, + "grad_norm": 0.39007549259820407, + "learning_rate": 5.25742504275844e-05, + "loss": 2.3997, + "step": 3738 + }, + { + "epoch": 0.8573721623480853, + "grad_norm": 0.4148871162772051, + "learning_rate": 5.2408618076375315e-05, + "loss": 2.5388, + "step": 3739 + }, + { + "epoch": 0.8576014675533135, + "grad_norm": 0.4206664875881806, + "learning_rate": 5.224323261357844e-05, + "loss": 2.3144, + "step": 3740 + }, + { + "epoch": 0.8578307727585416, + "grad_norm": 0.42639164551592224, + "learning_rate": 5.207809413041914e-05, + "loss": 2.3266, + "step": 3741 + }, + { + "epoch": 0.8580600779637698, + "grad_norm": 0.3888357658767153, + "learning_rate": 5.1913202717987e-05, + "loss": 2.2526, + "step": 3742 + }, + { + "epoch": 0.8582893831689979, + "grad_norm": 0.4084073154777171, + "learning_rate": 5.174855846723459e-05, + "loss": 2.3049, + "step": 3743 + }, + { + "epoch": 0.8585186883742261, + "grad_norm": 0.41828078021601744, + "learning_rate": 5.158416146897887e-05, + "loss": 2.5995, + "step": 3744 + }, + { + "epoch": 0.8587479935794543, + "grad_norm": 0.4110528304522503, + "learning_rate": 5.1420011813900104e-05, + "loss": 2.2832, + "step": 3745 + }, + { + "epoch": 0.8589772987846824, + "grad_norm": 0.41087020338842617, + "learning_rate": 5.125610959254212e-05, + "loss": 2.4692, + "step": 3746 + }, + { + "epoch": 0.8592066039899106, + "grad_norm": 0.4078399713884646, + "learning_rate": 5.109245489531211e-05, + "loss": 2.3671, + "step": 3747 + }, + { + "epoch": 0.8594359091951387, + "grad_norm": 0.4550220873848347, + "learning_rate": 5.092904781248115e-05, + "loss": 2.4239, + "step": 3748 + }, + { + "epoch": 0.8596652144003669, + "grad_norm": 0.4344315113504398, + "learning_rate": 5.0765888434183446e-05, + "loss": 2.438, + "step": 3749 + }, + { + "epoch": 0.859894519605595, + "grad_norm": 0.4196686750041355, + "learning_rate": 5.060297685041659e-05, + "loss": 2.359, + "step": 3750 + }, + { + "epoch": 0.8601238248108232, + "grad_norm": 0.41422843019625255, + "learning_rate": 5.0440313151041364e-05, + "loss": 2.3927, + "step": 3751 + }, + { + "epoch": 0.8603531300160514, + "grad_norm": 0.42454817312414883, + "learning_rate": 5.027789742578226e-05, + "loss": 2.4033, + "step": 3752 + }, + { + "epoch": 0.8605824352212795, + "grad_norm": 0.43039488684911287, + "learning_rate": 5.011572976422657e-05, + "loss": 2.4035, + "step": 3753 + }, + { + "epoch": 0.8608117404265077, + "grad_norm": 0.4444296862941777, + "learning_rate": 4.995381025582485e-05, + "loss": 2.3181, + "step": 3754 + }, + { + "epoch": 0.8610410456317359, + "grad_norm": 0.3994153468825797, + "learning_rate": 4.9792138989890825e-05, + "loss": 2.312, + "step": 3755 + }, + { + "epoch": 0.861270350836964, + "grad_norm": 0.38294941232279206, + "learning_rate": 4.9630716055601434e-05, + "loss": 2.3935, + "step": 3756 + }, + { + "epoch": 0.8614996560421921, + "grad_norm": 0.3874432640609366, + "learning_rate": 4.9469541541996234e-05, + "loss": 2.4238, + "step": 3757 + }, + { + "epoch": 0.8617289612474203, + "grad_norm": 0.37916816697723404, + "learning_rate": 4.9308615537978216e-05, + "loss": 2.2529, + "step": 3758 + }, + { + "epoch": 0.8619582664526485, + "grad_norm": 0.4180372290717237, + "learning_rate": 4.914793813231305e-05, + "loss": 2.3681, + "step": 3759 + }, + { + "epoch": 0.8621875716578766, + "grad_norm": 0.41865381899248616, + "learning_rate": 4.8987509413629284e-05, + "loss": 2.55, + "step": 3760 + }, + { + "epoch": 0.8624168768631048, + "grad_norm": 0.4224516016377883, + "learning_rate": 4.882732947041818e-05, + "loss": 2.433, + "step": 3761 + }, + { + "epoch": 0.862646182068333, + "grad_norm": 0.4305753694722786, + "learning_rate": 4.866739839103418e-05, + "loss": 2.4633, + "step": 3762 + }, + { + "epoch": 0.8628754872735611, + "grad_norm": 0.4210861498837944, + "learning_rate": 4.850771626369416e-05, + "loss": 2.3178, + "step": 3763 + }, + { + "epoch": 0.8631047924787892, + "grad_norm": 0.41363391534793026, + "learning_rate": 4.8348283176477545e-05, + "loss": 2.2175, + "step": 3764 + }, + { + "epoch": 0.8633340976840175, + "grad_norm": 0.43603734648642584, + "learning_rate": 4.818909921732662e-05, + "loss": 2.5733, + "step": 3765 + }, + { + "epoch": 0.8635634028892456, + "grad_norm": 0.4227910717322414, + "learning_rate": 4.8030164474046286e-05, + "loss": 2.3725, + "step": 3766 + }, + { + "epoch": 0.8637927080944737, + "grad_norm": 0.40988835668890367, + "learning_rate": 4.787147903430383e-05, + "loss": 2.546, + "step": 3767 + }, + { + "epoch": 0.8640220132997019, + "grad_norm": 0.39825586676293645, + "learning_rate": 4.771304298562912e-05, + "loss": 2.2932, + "step": 3768 + }, + { + "epoch": 0.8642513185049301, + "grad_norm": 0.40071883415387455, + "learning_rate": 4.755485641541424e-05, + "loss": 2.4917, + "step": 3769 + }, + { + "epoch": 0.8644806237101582, + "grad_norm": 0.4233696371823567, + "learning_rate": 4.7396919410914174e-05, + "loss": 2.6093, + "step": 3770 + }, + { + "epoch": 0.8647099289153863, + "grad_norm": 0.41367962536337494, + "learning_rate": 4.723923205924557e-05, + "loss": 2.5393, + "step": 3771 + }, + { + "epoch": 0.8649392341206146, + "grad_norm": 0.420016187002705, + "learning_rate": 4.708179444738791e-05, + "loss": 2.4381, + "step": 3772 + }, + { + "epoch": 0.8651685393258427, + "grad_norm": 0.4311466223900657, + "learning_rate": 4.6924606662182736e-05, + "loss": 2.4485, + "step": 3773 + }, + { + "epoch": 0.8653978445310708, + "grad_norm": 0.42121546398896925, + "learning_rate": 4.6767668790333675e-05, + "loss": 2.3876, + "step": 3774 + }, + { + "epoch": 0.8656271497362991, + "grad_norm": 0.4269679956003662, + "learning_rate": 4.6610980918406596e-05, + "loss": 2.4272, + "step": 3775 + }, + { + "epoch": 0.8658564549415272, + "grad_norm": 0.4265185215767201, + "learning_rate": 4.645454313282965e-05, + "loss": 2.3007, + "step": 3776 + }, + { + "epoch": 0.8660857601467553, + "grad_norm": 0.3991748822758438, + "learning_rate": 4.629835551989276e-05, + "loss": 2.3964, + "step": 3777 + }, + { + "epoch": 0.8663150653519834, + "grad_norm": 0.4185779633753731, + "learning_rate": 4.614241816574799e-05, + "loss": 2.4725, + "step": 3778 + }, + { + "epoch": 0.8665443705572117, + "grad_norm": 0.40453463379923477, + "learning_rate": 4.5986731156409224e-05, + "loss": 2.2953, + "step": 3779 + }, + { + "epoch": 0.8667736757624398, + "grad_norm": 0.4086469619326245, + "learning_rate": 4.583129457775253e-05, + "loss": 2.3521, + "step": 3780 + }, + { + "epoch": 0.8670029809676679, + "grad_norm": 0.41776776214834993, + "learning_rate": 4.567610851551568e-05, + "loss": 2.5034, + "step": 3781 + }, + { + "epoch": 0.8672322861728962, + "grad_norm": 0.4210153009772555, + "learning_rate": 4.5521173055298234e-05, + "loss": 2.3564, + "step": 3782 + }, + { + "epoch": 0.8674615913781243, + "grad_norm": 0.4060882229440686, + "learning_rate": 4.536648828256146e-05, + "loss": 2.5194, + "step": 3783 + }, + { + "epoch": 0.8676908965833524, + "grad_norm": 0.43125618934897797, + "learning_rate": 4.52120542826287e-05, + "loss": 2.4073, + "step": 3784 + }, + { + "epoch": 0.8679202017885806, + "grad_norm": 0.4456009632262567, + "learning_rate": 4.505787114068433e-05, + "loss": 2.5701, + "step": 3785 + }, + { + "epoch": 0.8681495069938088, + "grad_norm": 0.42319224021774876, + "learning_rate": 4.490393894177508e-05, + "loss": 2.4515, + "step": 3786 + }, + { + "epoch": 0.8683788121990369, + "grad_norm": 0.4028511972226411, + "learning_rate": 4.4750257770808764e-05, + "loss": 2.3847, + "step": 3787 + }, + { + "epoch": 0.868608117404265, + "grad_norm": 0.4280035466554648, + "learning_rate": 4.459682771255497e-05, + "loss": 2.4489, + "step": 3788 + }, + { + "epoch": 0.8688374226094933, + "grad_norm": 0.4574825785589019, + "learning_rate": 4.444364885164448e-05, + "loss": 2.4664, + "step": 3789 + }, + { + "epoch": 0.8690667278147214, + "grad_norm": 0.44661794114510345, + "learning_rate": 4.4290721272570056e-05, + "loss": 2.314, + "step": 3790 + }, + { + "epoch": 0.8692960330199495, + "grad_norm": 1.1904167721133525, + "learning_rate": 4.413804505968533e-05, + "loss": 2.4046, + "step": 3791 + }, + { + "epoch": 0.8695253382251777, + "grad_norm": 0.4092118806360508, + "learning_rate": 4.3985620297205487e-05, + "loss": 2.2786, + "step": 3792 + }, + { + "epoch": 0.8697546434304059, + "grad_norm": 0.3937551320768139, + "learning_rate": 4.3833447069206944e-05, + "loss": 2.4412, + "step": 3793 + }, + { + "epoch": 0.869983948635634, + "grad_norm": 0.4092712826873307, + "learning_rate": 4.3681525459627616e-05, + "loss": 2.441, + "step": 3794 + }, + { + "epoch": 0.8702132538408622, + "grad_norm": 0.41608704115077805, + "learning_rate": 4.352985555226635e-05, + "loss": 2.4138, + "step": 3795 + }, + { + "epoch": 0.8704425590460904, + "grad_norm": 0.4387990213929117, + "learning_rate": 4.337843743078329e-05, + "loss": 2.5469, + "step": 3796 + }, + { + "epoch": 0.8706718642513185, + "grad_norm": 0.4059199347580236, + "learning_rate": 4.322727117869951e-05, + "loss": 2.3914, + "step": 3797 + }, + { + "epoch": 0.8709011694565466, + "grad_norm": 0.4127778006147335, + "learning_rate": 4.3076356879397594e-05, + "loss": 2.524, + "step": 3798 + }, + { + "epoch": 0.8711304746617748, + "grad_norm": 0.4020780627590755, + "learning_rate": 4.29256946161205e-05, + "loss": 2.3437, + "step": 3799 + }, + { + "epoch": 0.871359779867003, + "grad_norm": 0.3735647449315572, + "learning_rate": 4.277528447197282e-05, + "loss": 2.4026, + "step": 3800 + }, + { + "epoch": 0.8715890850722311, + "grad_norm": 0.46399866099140746, + "learning_rate": 4.262512652991968e-05, + "loss": 2.3767, + "step": 3801 + }, + { + "epoch": 0.8718183902774593, + "grad_norm": 0.4438800055521492, + "learning_rate": 4.247522087278716e-05, + "loss": 2.3389, + "step": 3802 + }, + { + "epoch": 0.8720476954826875, + "grad_norm": 0.4120996673130382, + "learning_rate": 4.2325567583262113e-05, + "loss": 2.5221, + "step": 3803 + }, + { + "epoch": 0.8722770006879156, + "grad_norm": 0.4194409851554206, + "learning_rate": 4.2176166743892576e-05, + "loss": 2.5389, + "step": 3804 + }, + { + "epoch": 0.8725063058931438, + "grad_norm": 0.4282324775789959, + "learning_rate": 4.2027018437086895e-05, + "loss": 2.5094, + "step": 3805 + }, + { + "epoch": 0.8727356110983719, + "grad_norm": 0.41984733868363894, + "learning_rate": 4.1878122745114265e-05, + "loss": 2.5257, + "step": 3806 + }, + { + "epoch": 0.8729649163036001, + "grad_norm": 0.4197098342812893, + "learning_rate": 4.172947975010449e-05, + "loss": 2.416, + "step": 3807 + }, + { + "epoch": 0.8731942215088283, + "grad_norm": 0.4258918175726162, + "learning_rate": 4.158108953404832e-05, + "loss": 2.3533, + "step": 3808 + }, + { + "epoch": 0.8734235267140564, + "grad_norm": 0.3744300396970033, + "learning_rate": 4.143295217879645e-05, + "loss": 2.416, + "step": 3809 + }, + { + "epoch": 0.8736528319192846, + "grad_norm": 0.39804139107963954, + "learning_rate": 4.128506776606067e-05, + "loss": 2.5608, + "step": 3810 + }, + { + "epoch": 0.8738821371245127, + "grad_norm": 0.4193259466739966, + "learning_rate": 4.113743637741296e-05, + "loss": 2.5307, + "step": 3811 + }, + { + "epoch": 0.8741114423297409, + "grad_norm": 0.4047404501333747, + "learning_rate": 4.0990058094285966e-05, + "loss": 2.5475, + "step": 3812 + }, + { + "epoch": 0.874340747534969, + "grad_norm": 0.39505423111767024, + "learning_rate": 4.084293299797226e-05, + "loss": 2.5197, + "step": 3813 + }, + { + "epoch": 0.8745700527401972, + "grad_norm": 0.41189289841174453, + "learning_rate": 4.069606116962532e-05, + "loss": 2.3278, + "step": 3814 + }, + { + "epoch": 0.8747993579454254, + "grad_norm": 0.3935629549048426, + "learning_rate": 4.054944269025862e-05, + "loss": 2.5128, + "step": 3815 + }, + { + "epoch": 0.8750286631506535, + "grad_norm": 0.4140191904963766, + "learning_rate": 4.0403077640745855e-05, + "loss": 2.5059, + "step": 3816 + }, + { + "epoch": 0.8752579683558817, + "grad_norm": 0.38251204454462734, + "learning_rate": 4.025696610182095e-05, + "loss": 2.4765, + "step": 3817 + }, + { + "epoch": 0.8754872735611099, + "grad_norm": 0.38465423024611495, + "learning_rate": 4.0111108154078205e-05, + "loss": 2.4754, + "step": 3818 + }, + { + "epoch": 0.875716578766338, + "grad_norm": 0.4064575332865436, + "learning_rate": 3.996550387797187e-05, + "loss": 2.3843, + "step": 3819 + }, + { + "epoch": 0.8759458839715661, + "grad_norm": 0.40705142339202244, + "learning_rate": 3.9820153353816226e-05, + "loss": 2.3528, + "step": 3820 + }, + { + "epoch": 0.8761751891767943, + "grad_norm": 0.43308971327126994, + "learning_rate": 3.9675056661785556e-05, + "loss": 2.2176, + "step": 3821 + }, + { + "epoch": 0.8764044943820225, + "grad_norm": 0.42066047745000423, + "learning_rate": 3.9530213881914576e-05, + "loss": 2.3912, + "step": 3822 + }, + { + "epoch": 0.8766337995872506, + "grad_norm": 0.44568986403078215, + "learning_rate": 3.9385625094097154e-05, + "loss": 2.4737, + "step": 3823 + }, + { + "epoch": 0.8768631047924788, + "grad_norm": 0.41105205523671956, + "learning_rate": 3.924129037808777e-05, + "loss": 2.4441, + "step": 3824 + }, + { + "epoch": 0.877092409997707, + "grad_norm": 0.4368976784745604, + "learning_rate": 3.909720981350034e-05, + "loss": 2.403, + "step": 3825 + }, + { + "epoch": 0.8773217152029351, + "grad_norm": 0.4242519544329515, + "learning_rate": 3.895338347980898e-05, + "loss": 2.7203, + "step": 3826 + }, + { + "epoch": 0.8775510204081632, + "grad_norm": 0.4426637574963083, + "learning_rate": 3.880981145634704e-05, + "loss": 2.517, + "step": 3827 + }, + { + "epoch": 0.8777803256133915, + "grad_norm": 0.41402117386542225, + "learning_rate": 3.8666493822308077e-05, + "loss": 2.4226, + "step": 3828 + }, + { + "epoch": 0.8780096308186196, + "grad_norm": 0.42640724639216576, + "learning_rate": 3.852343065674507e-05, + "loss": 2.4396, + "step": 3829 + }, + { + "epoch": 0.8782389360238477, + "grad_norm": 0.41017168925222663, + "learning_rate": 3.838062203857073e-05, + "loss": 2.3902, + "step": 3830 + }, + { + "epoch": 0.8784682412290759, + "grad_norm": 0.4158620307036415, + "learning_rate": 3.8238068046557276e-05, + "loss": 2.3318, + "step": 3831 + }, + { + "epoch": 0.8786975464343041, + "grad_norm": 0.4017217423939412, + "learning_rate": 3.8095768759336626e-05, + "loss": 2.6127, + "step": 3832 + }, + { + "epoch": 0.8789268516395322, + "grad_norm": 0.41616075377818823, + "learning_rate": 3.795372425540006e-05, + "loss": 2.6039, + "step": 3833 + }, + { + "epoch": 0.8791561568447603, + "grad_norm": 0.4036751799232379, + "learning_rate": 3.7811934613098394e-05, + "loss": 2.4925, + "step": 3834 + }, + { + "epoch": 0.8793854620499886, + "grad_norm": 0.41695953369211175, + "learning_rate": 3.76703999106418e-05, + "loss": 2.2581, + "step": 3835 + }, + { + "epoch": 0.8796147672552167, + "grad_norm": 0.4402461400693039, + "learning_rate": 3.752912022610006e-05, + "loss": 2.4717, + "step": 3836 + }, + { + "epoch": 0.8798440724604448, + "grad_norm": 0.4472902125839286, + "learning_rate": 3.7388095637401754e-05, + "loss": 2.4601, + "step": 3837 + }, + { + "epoch": 0.8800733776656731, + "grad_norm": 0.4129757823103584, + "learning_rate": 3.72473262223354e-05, + "loss": 2.4356, + "step": 3838 + }, + { + "epoch": 0.8803026828709012, + "grad_norm": 0.41600137307216595, + "learning_rate": 3.7106812058548376e-05, + "loss": 2.335, + "step": 3839 + }, + { + "epoch": 0.8805319880761293, + "grad_norm": 0.41198020247808254, + "learning_rate": 3.696655322354736e-05, + "loss": 2.3937, + "step": 3840 + }, + { + "epoch": 0.8807612932813574, + "grad_norm": 0.4209923885987185, + "learning_rate": 3.682654979469807e-05, + "loss": 2.501, + "step": 3841 + }, + { + "epoch": 0.8809905984865857, + "grad_norm": 0.4390890837443518, + "learning_rate": 3.668680184922563e-05, + "loss": 2.3501, + "step": 3842 + }, + { + "epoch": 0.8812199036918138, + "grad_norm": 0.3893623714771188, + "learning_rate": 3.654730946421403e-05, + "loss": 2.2904, + "step": 3843 + }, + { + "epoch": 0.8814492088970419, + "grad_norm": 0.40086004537038183, + "learning_rate": 3.6408072716606344e-05, + "loss": 2.5144, + "step": 3844 + }, + { + "epoch": 0.8816785141022702, + "grad_norm": 0.41319937591639627, + "learning_rate": 3.6269091683204466e-05, + "loss": 2.4954, + "step": 3845 + }, + { + "epoch": 0.8819078193074983, + "grad_norm": 0.42465698691199927, + "learning_rate": 3.613036644066969e-05, + "loss": 2.4396, + "step": 3846 + }, + { + "epoch": 0.8821371245127264, + "grad_norm": 0.4063600458942431, + "learning_rate": 3.5991897065521693e-05, + "loss": 2.3439, + "step": 3847 + }, + { + "epoch": 0.8823664297179546, + "grad_norm": 0.41022972116218126, + "learning_rate": 3.5853683634139434e-05, + "loss": 2.5119, + "step": 3848 + }, + { + "epoch": 0.8825957349231828, + "grad_norm": 0.43713948400863184, + "learning_rate": 3.571572622276026e-05, + "loss": 2.5094, + "step": 3849 + }, + { + "epoch": 0.8828250401284109, + "grad_norm": 0.42416202142971365, + "learning_rate": 3.557802490748097e-05, + "loss": 2.5595, + "step": 3850 + }, + { + "epoch": 0.883054345333639, + "grad_norm": 0.45631928863137305, + "learning_rate": 3.544057976425619e-05, + "loss": 2.2706, + "step": 3851 + }, + { + "epoch": 0.8832836505388673, + "grad_norm": 0.42573386461771856, + "learning_rate": 3.5303390868900164e-05, + "loss": 2.4453, + "step": 3852 + }, + { + "epoch": 0.8835129557440954, + "grad_norm": 0.3902306012949812, + "learning_rate": 3.5166458297085146e-05, + "loss": 2.2672, + "step": 3853 + }, + { + "epoch": 0.8837422609493235, + "grad_norm": 0.38764133893731517, + "learning_rate": 3.502978212434227e-05, + "loss": 2.3888, + "step": 3854 + }, + { + "epoch": 0.8839715661545517, + "grad_norm": 0.4201090787325488, + "learning_rate": 3.489336242606111e-05, + "loss": 2.4969, + "step": 3855 + }, + { + "epoch": 0.8842008713597799, + "grad_norm": 0.40994277237362675, + "learning_rate": 3.4757199277490105e-05, + "loss": 2.2531, + "step": 3856 + }, + { + "epoch": 0.884430176565008, + "grad_norm": 0.44085431311347306, + "learning_rate": 3.462129275373577e-05, + "loss": 2.5139, + "step": 3857 + }, + { + "epoch": 0.8846594817702362, + "grad_norm": 0.4033991231804382, + "learning_rate": 3.448564292976319e-05, + "loss": 2.4176, + "step": 3858 + }, + { + "epoch": 0.8848887869754644, + "grad_norm": 0.4055925194670613, + "learning_rate": 3.4350249880395924e-05, + "loss": 2.5165, + "step": 3859 + }, + { + "epoch": 0.8851180921806925, + "grad_norm": 0.3970172163843095, + "learning_rate": 3.4215113680316046e-05, + "loss": 2.5696, + "step": 3860 + }, + { + "epoch": 0.8853473973859206, + "grad_norm": 0.39876742012895594, + "learning_rate": 3.408023440406355e-05, + "loss": 2.3667, + "step": 3861 + }, + { + "epoch": 0.8855767025911488, + "grad_norm": 0.4166766654609906, + "learning_rate": 3.394561212603708e-05, + "loss": 2.3985, + "step": 3862 + }, + { + "epoch": 0.885806007796377, + "grad_norm": 0.3963084069406601, + "learning_rate": 3.381124692049331e-05, + "loss": 2.4352, + "step": 3863 + }, + { + "epoch": 0.8860353130016051, + "grad_norm": 0.39879312807911416, + "learning_rate": 3.367713886154733e-05, + "loss": 2.3636, + "step": 3864 + }, + { + "epoch": 0.8862646182068333, + "grad_norm": 0.4236568607705268, + "learning_rate": 3.354328802317197e-05, + "loss": 2.5135, + "step": 3865 + }, + { + "epoch": 0.8864939234120615, + "grad_norm": 0.4504605802511009, + "learning_rate": 3.340969447919873e-05, + "loss": 2.3729, + "step": 3866 + }, + { + "epoch": 0.8867232286172896, + "grad_norm": 0.4317718977331243, + "learning_rate": 3.327635830331677e-05, + "loss": 2.4122, + "step": 3867 + }, + { + "epoch": 0.8869525338225178, + "grad_norm": 0.4146208209148113, + "learning_rate": 3.314327956907354e-05, + "loss": 2.4134, + "step": 3868 + }, + { + "epoch": 0.8871818390277459, + "grad_norm": 0.39596937681325634, + "learning_rate": 3.3010458349874206e-05, + "loss": 2.3447, + "step": 3869 + }, + { + "epoch": 0.8874111442329741, + "grad_norm": 0.4344616894559595, + "learning_rate": 3.287789471898223e-05, + "loss": 2.3561, + "step": 3870 + }, + { + "epoch": 0.8876404494382022, + "grad_norm": 0.4132896753711769, + "learning_rate": 3.2745588749518775e-05, + "loss": 2.3743, + "step": 3871 + }, + { + "epoch": 0.8878697546434304, + "grad_norm": 0.39912156311800534, + "learning_rate": 3.2613540514462914e-05, + "loss": 2.3763, + "step": 3872 + }, + { + "epoch": 0.8880990598486586, + "grad_norm": 0.44243554657962914, + "learning_rate": 3.248175008665161e-05, + "loss": 2.5494, + "step": 3873 + }, + { + "epoch": 0.8883283650538867, + "grad_norm": 0.3836351324600238, + "learning_rate": 3.235021753877942e-05, + "loss": 2.4526, + "step": 3874 + }, + { + "epoch": 0.8885576702591149, + "grad_norm": 0.42246849046096197, + "learning_rate": 3.221894294339911e-05, + "loss": 2.5179, + "step": 3875 + }, + { + "epoch": 0.888786975464343, + "grad_norm": 0.4370117846401229, + "learning_rate": 3.208792637292057e-05, + "loss": 2.4276, + "step": 3876 + }, + { + "epoch": 0.8890162806695712, + "grad_norm": 0.40539662532244075, + "learning_rate": 3.1957167899611836e-05, + "loss": 2.4267, + "step": 3877 + }, + { + "epoch": 0.8892455858747994, + "grad_norm": 0.43463780878591085, + "learning_rate": 3.1826667595598435e-05, + "loss": 2.4377, + "step": 3878 + }, + { + "epoch": 0.8894748910800275, + "grad_norm": 0.4186036879146607, + "learning_rate": 3.169642553286334e-05, + "loss": 2.299, + "step": 3879 + }, + { + "epoch": 0.8897041962852557, + "grad_norm": 0.43079884439372307, + "learning_rate": 3.156644178324719e-05, + "loss": 2.4154, + "step": 3880 + }, + { + "epoch": 0.8899335014904838, + "grad_norm": 0.39685120061092743, + "learning_rate": 3.143671641844831e-05, + "loss": 2.3212, + "step": 3881 + }, + { + "epoch": 0.890162806695712, + "grad_norm": 0.45029382191413214, + "learning_rate": 3.1307249510022225e-05, + "loss": 2.3062, + "step": 3882 + }, + { + "epoch": 0.8903921119009401, + "grad_norm": 0.4162068350053989, + "learning_rate": 3.117804112938205e-05, + "loss": 2.5359, + "step": 3883 + }, + { + "epoch": 0.8906214171061683, + "grad_norm": 0.39685277900203364, + "learning_rate": 3.104909134779821e-05, + "loss": 2.3295, + "step": 3884 + }, + { + "epoch": 0.8908507223113965, + "grad_norm": 0.4289830126593849, + "learning_rate": 3.092040023639869e-05, + "loss": 2.3678, + "step": 3885 + }, + { + "epoch": 0.8910800275166246, + "grad_norm": 0.4137408544815711, + "learning_rate": 3.079196786616839e-05, + "loss": 2.497, + "step": 3886 + }, + { + "epoch": 0.8913093327218528, + "grad_norm": 0.3883914661153665, + "learning_rate": 3.066379430795002e-05, + "loss": 2.3104, + "step": 3887 + }, + { + "epoch": 0.891538637927081, + "grad_norm": 0.44840952093963343, + "learning_rate": 3.053587963244303e-05, + "loss": 2.4158, + "step": 3888 + }, + { + "epoch": 0.8917679431323091, + "grad_norm": 0.4011952074213191, + "learning_rate": 3.040822391020459e-05, + "loss": 2.4443, + "step": 3889 + }, + { + "epoch": 0.8919972483375372, + "grad_norm": 0.42505146710170455, + "learning_rate": 3.0280827211648377e-05, + "loss": 2.3474, + "step": 3890 + }, + { + "epoch": 0.8922265535427654, + "grad_norm": 0.3981849401059799, + "learning_rate": 3.0153689607045842e-05, + "loss": 2.5641, + "step": 3891 + }, + { + "epoch": 0.8924558587479936, + "grad_norm": 0.3995286406727298, + "learning_rate": 3.0026811166525114e-05, + "loss": 2.3586, + "step": 3892 + }, + { + "epoch": 0.8926851639532217, + "grad_norm": 0.40779116942729504, + "learning_rate": 2.9900191960071545e-05, + "loss": 2.2652, + "step": 3893 + }, + { + "epoch": 0.8929144691584499, + "grad_norm": 0.458547922583732, + "learning_rate": 2.9773832057527263e-05, + "loss": 2.3332, + "step": 3894 + }, + { + "epoch": 0.8931437743636781, + "grad_norm": 0.4300720079123421, + "learning_rate": 2.9647731528591848e-05, + "loss": 2.4031, + "step": 3895 + }, + { + "epoch": 0.8933730795689062, + "grad_norm": 0.4431071229492851, + "learning_rate": 2.9521890442821275e-05, + "loss": 2.3548, + "step": 3896 + }, + { + "epoch": 0.8936023847741343, + "grad_norm": 0.4208120511453034, + "learning_rate": 2.9396308869628795e-05, + "loss": 2.3409, + "step": 3897 + }, + { + "epoch": 0.8938316899793626, + "grad_norm": 0.40884088706157856, + "learning_rate": 2.9270986878284167e-05, + "loss": 2.3813, + "step": 3898 + }, + { + "epoch": 0.8940609951845907, + "grad_norm": 0.40570311085020766, + "learning_rate": 2.914592453791448e-05, + "loss": 2.4509, + "step": 3899 + }, + { + "epoch": 0.8942903003898188, + "grad_norm": 0.4430647670656077, + "learning_rate": 2.902112191750289e-05, + "loss": 2.3955, + "step": 3900 + }, + { + "epoch": 0.8945196055950471, + "grad_norm": 0.4104077619597906, + "learning_rate": 2.8896579085889994e-05, + "loss": 2.3878, + "step": 3901 + }, + { + "epoch": 0.8947489108002752, + "grad_norm": 0.41205041224926103, + "learning_rate": 2.8772296111772677e-05, + "loss": 2.5899, + "step": 3902 + }, + { + "epoch": 0.8949782160055033, + "grad_norm": 0.4147461637797838, + "learning_rate": 2.86482730637046e-05, + "loss": 2.3549, + "step": 3903 + }, + { + "epoch": 0.8952075212107314, + "grad_norm": 0.42063573804315446, + "learning_rate": 2.8524510010095984e-05, + "loss": 2.4423, + "step": 3904 + }, + { + "epoch": 0.8954368264159597, + "grad_norm": 0.44319344093545293, + "learning_rate": 2.840100701921383e-05, + "loss": 2.2982, + "step": 3905 + }, + { + "epoch": 0.8956661316211878, + "grad_norm": 0.4113429577772726, + "learning_rate": 2.8277764159181486e-05, + "loss": 2.2123, + "step": 3906 + }, + { + "epoch": 0.8958954368264159, + "grad_norm": 0.40327859361814944, + "learning_rate": 2.8154781497978898e-05, + "loss": 2.6557, + "step": 3907 + }, + { + "epoch": 0.8961247420316442, + "grad_norm": 0.41073856847944373, + "learning_rate": 2.8032059103442364e-05, + "loss": 2.2906, + "step": 3908 + }, + { + "epoch": 0.8963540472368723, + "grad_norm": 0.46635458712437056, + "learning_rate": 2.7909597043265013e-05, + "loss": 2.337, + "step": 3909 + }, + { + "epoch": 0.8965833524421004, + "grad_norm": 0.4148363601177566, + "learning_rate": 2.7787395384995862e-05, + "loss": 2.434, + "step": 3910 + }, + { + "epoch": 0.8968126576473286, + "grad_norm": 0.4237618184733734, + "learning_rate": 2.7665454196040662e-05, + "loss": 2.4825, + "step": 3911 + }, + { + "epoch": 0.8970419628525568, + "grad_norm": 0.40677782402236656, + "learning_rate": 2.7543773543661167e-05, + "loss": 2.338, + "step": 3912 + }, + { + "epoch": 0.8972712680577849, + "grad_norm": 0.4177312741290501, + "learning_rate": 2.7422353494975905e-05, + "loss": 2.3159, + "step": 3913 + }, + { + "epoch": 0.897500573263013, + "grad_norm": 0.41065167392693314, + "learning_rate": 2.7301194116959084e-05, + "loss": 2.4477, + "step": 3914 + }, + { + "epoch": 0.8977298784682413, + "grad_norm": 0.4319691461187236, + "learning_rate": 2.7180295476441573e-05, + "loss": 2.4437, + "step": 3915 + }, + { + "epoch": 0.8979591836734694, + "grad_norm": 0.41482979608711934, + "learning_rate": 2.70596576401102e-05, + "loss": 2.5857, + "step": 3916 + }, + { + "epoch": 0.8981884888786975, + "grad_norm": 0.39754903599850006, + "learning_rate": 2.6939280674508016e-05, + "loss": 2.5108, + "step": 3917 + }, + { + "epoch": 0.8984177940839257, + "grad_norm": 0.4146657935953936, + "learning_rate": 2.681916464603401e-05, + "loss": 2.3641, + "step": 3918 + }, + { + "epoch": 0.8986470992891539, + "grad_norm": 0.4088838979675896, + "learning_rate": 2.669930962094358e-05, + "loss": 2.463, + "step": 3919 + }, + { + "epoch": 0.898876404494382, + "grad_norm": 0.4414382300520191, + "learning_rate": 2.6579715665347894e-05, + "loss": 2.4385, + "step": 3920 + }, + { + "epoch": 0.8991057096996102, + "grad_norm": 0.4028675065029393, + "learning_rate": 2.6460382845214126e-05, + "loss": 2.4155, + "step": 3921 + }, + { + "epoch": 0.8993350149048384, + "grad_norm": 0.4062255564237616, + "learning_rate": 2.6341311226365406e-05, + "loss": 2.4189, + "step": 3922 + }, + { + "epoch": 0.8995643201100665, + "grad_norm": 0.43364004709321763, + "learning_rate": 2.6222500874481025e-05, + "loss": 2.4698, + "step": 3923 + }, + { + "epoch": 0.8997936253152946, + "grad_norm": 0.42274254424608193, + "learning_rate": 2.6103951855095897e-05, + "loss": 2.4161, + "step": 3924 + }, + { + "epoch": 0.9000229305205228, + "grad_norm": 0.4290982310182714, + "learning_rate": 2.5985664233600827e-05, + "loss": 2.3731, + "step": 3925 + }, + { + "epoch": 0.900252235725751, + "grad_norm": 0.4481104974272491, + "learning_rate": 2.586763807524245e-05, + "loss": 2.4939, + "step": 3926 + }, + { + "epoch": 0.9004815409309791, + "grad_norm": 0.42096222483173157, + "learning_rate": 2.574987344512336e-05, + "loss": 2.4465, + "step": 3927 + }, + { + "epoch": 0.9007108461362073, + "grad_norm": 0.4040890699096655, + "learning_rate": 2.5632370408201488e-05, + "loss": 2.464, + "step": 3928 + }, + { + "epoch": 0.9009401513414355, + "grad_norm": 0.4524109057235387, + "learning_rate": 2.5515129029290984e-05, + "loss": 2.4735, + "step": 3929 + }, + { + "epoch": 0.9011694565466636, + "grad_norm": 0.40654134256465785, + "learning_rate": 2.539814937306123e-05, + "loss": 2.4167, + "step": 3930 + }, + { + "epoch": 0.9013987617518918, + "grad_norm": 0.40129284449775754, + "learning_rate": 2.5281431504037556e-05, + "loss": 2.4925, + "step": 3931 + }, + { + "epoch": 0.90162806695712, + "grad_norm": 0.3792768741474, + "learning_rate": 2.5164975486600572e-05, + "loss": 2.34, + "step": 3932 + }, + { + "epoch": 0.9018573721623481, + "grad_norm": 0.40704707073696866, + "learning_rate": 2.504878138498684e-05, + "loss": 2.3232, + "step": 3933 + }, + { + "epoch": 0.9020866773675762, + "grad_norm": 0.39922454757467524, + "learning_rate": 2.4932849263288147e-05, + "loss": 2.2539, + "step": 3934 + }, + { + "epoch": 0.9023159825728044, + "grad_norm": 0.39080871426481273, + "learning_rate": 2.48171791854519e-05, + "loss": 2.503, + "step": 3935 + }, + { + "epoch": 0.9025452877780326, + "grad_norm": 0.40807366694300906, + "learning_rate": 2.470177121528089e-05, + "loss": 2.4231, + "step": 3936 + }, + { + "epoch": 0.9027745929832607, + "grad_norm": 0.41332855657376194, + "learning_rate": 2.4586625416433473e-05, + "loss": 2.3919, + "step": 3937 + }, + { + "epoch": 0.9030038981884889, + "grad_norm": 0.4326379531309123, + "learning_rate": 2.4471741852423235e-05, + "loss": 2.2895, + "step": 3938 + }, + { + "epoch": 0.903233203393717, + "grad_norm": 0.38755786787601676, + "learning_rate": 2.435712058661921e-05, + "loss": 2.5175, + "step": 3939 + }, + { + "epoch": 0.9034625085989452, + "grad_norm": 0.4381521920073401, + "learning_rate": 2.424276168224565e-05, + "loss": 2.2984, + "step": 3940 + }, + { + "epoch": 0.9036918138041734, + "grad_norm": 0.4039084891635872, + "learning_rate": 2.4128665202382327e-05, + "loss": 2.417, + "step": 3941 + }, + { + "epoch": 0.9039211190094015, + "grad_norm": 0.40588123082363775, + "learning_rate": 2.4014831209963893e-05, + "loss": 2.3644, + "step": 3942 + }, + { + "epoch": 0.9041504242146297, + "grad_norm": 0.3974404050073574, + "learning_rate": 2.3901259767780515e-05, + "loss": 2.3551, + "step": 3943 + }, + { + "epoch": 0.9043797294198578, + "grad_norm": 0.4083461527727942, + "learning_rate": 2.378795093847752e-05, + "loss": 2.2433, + "step": 3944 + }, + { + "epoch": 0.904609034625086, + "grad_norm": 0.4024728329150473, + "learning_rate": 2.367490478455514e-05, + "loss": 2.3073, + "step": 3945 + }, + { + "epoch": 0.9048383398303141, + "grad_norm": 0.3964063172910701, + "learning_rate": 2.3562121368368937e-05, + "loss": 2.4671, + "step": 3946 + }, + { + "epoch": 0.9050676450355423, + "grad_norm": 0.43283526851540954, + "learning_rate": 2.3449600752129597e-05, + "loss": 2.4583, + "step": 3947 + }, + { + "epoch": 0.9052969502407705, + "grad_norm": 0.38938966615060494, + "learning_rate": 2.3337342997902633e-05, + "loss": 2.4932, + "step": 3948 + }, + { + "epoch": 0.9055262554459986, + "grad_norm": 0.44060919613148036, + "learning_rate": 2.3225348167608685e-05, + "loss": 2.5904, + "step": 3949 + }, + { + "epoch": 0.9057555606512268, + "grad_norm": 0.4213510537969173, + "learning_rate": 2.3113616323023277e-05, + "loss": 2.4579, + "step": 3950 + }, + { + "epoch": 0.905984865856455, + "grad_norm": 0.43106664834025843, + "learning_rate": 2.3002147525777118e-05, + "loss": 2.5355, + "step": 3951 + }, + { + "epoch": 0.9062141710616831, + "grad_norm": 0.41368894088762154, + "learning_rate": 2.2890941837355518e-05, + "loss": 2.3344, + "step": 3952 + }, + { + "epoch": 0.9064434762669112, + "grad_norm": 0.4123739782175195, + "learning_rate": 2.2779999319098856e-05, + "loss": 2.498, + "step": 3953 + }, + { + "epoch": 0.9066727814721394, + "grad_norm": 0.40857607239010313, + "learning_rate": 2.2669320032202178e-05, + "loss": 2.4977, + "step": 3954 + }, + { + "epoch": 0.9069020866773676, + "grad_norm": 0.42204430582850244, + "learning_rate": 2.255890403771571e-05, + "loss": 2.481, + "step": 3955 + }, + { + "epoch": 0.9071313918825957, + "grad_norm": 0.4029412742913259, + "learning_rate": 2.2448751396543787e-05, + "loss": 2.318, + "step": 3956 + }, + { + "epoch": 0.9073606970878239, + "grad_norm": 0.4358741014171541, + "learning_rate": 2.233886216944614e-05, + "loss": 2.3121, + "step": 3957 + }, + { + "epoch": 0.9075900022930521, + "grad_norm": 0.4086411545722723, + "learning_rate": 2.222923641703689e-05, + "loss": 2.2466, + "step": 3958 + }, + { + "epoch": 0.9078193074982802, + "grad_norm": 0.4157162237026957, + "learning_rate": 2.211987419978484e-05, + "loss": 2.3027, + "step": 3959 + }, + { + "epoch": 0.9080486127035083, + "grad_norm": 0.43576300094287007, + "learning_rate": 2.2010775578013343e-05, + "loss": 2.3461, + "step": 3960 + }, + { + "epoch": 0.9082779179087366, + "grad_norm": 0.37386537409785464, + "learning_rate": 2.1901940611900705e-05, + "loss": 2.3279, + "step": 3961 + }, + { + "epoch": 0.9085072231139647, + "grad_norm": 0.4369878785780353, + "learning_rate": 2.179336936147941e-05, + "loss": 2.3612, + "step": 3962 + }, + { + "epoch": 0.9087365283191928, + "grad_norm": 0.42169277002714683, + "learning_rate": 2.168506188663666e-05, + "loss": 2.5067, + "step": 3963 + }, + { + "epoch": 0.908965833524421, + "grad_norm": 0.42265870634689845, + "learning_rate": 2.157701824711411e-05, + "loss": 2.2754, + "step": 3964 + }, + { + "epoch": 0.9091951387296492, + "grad_norm": 0.38116451095518766, + "learning_rate": 2.1469238502507925e-05, + "loss": 2.424, + "step": 3965 + }, + { + "epoch": 0.9094244439348773, + "grad_norm": 0.4112250782279852, + "learning_rate": 2.136172271226877e-05, + "loss": 2.556, + "step": 3966 + }, + { + "epoch": 0.9096537491401054, + "grad_norm": 0.45992568909133924, + "learning_rate": 2.125447093570154e-05, + "loss": 2.2389, + "step": 3967 + }, + { + "epoch": 0.9098830543453337, + "grad_norm": 0.40059150879108907, + "learning_rate": 2.114748323196558e-05, + "loss": 2.4371, + "step": 3968 + }, + { + "epoch": 0.9101123595505618, + "grad_norm": 0.4250831953777384, + "learning_rate": 2.1040759660074793e-05, + "loss": 2.4141, + "step": 3969 + }, + { + "epoch": 0.9103416647557899, + "grad_norm": 0.40322799151407024, + "learning_rate": 2.0934300278896867e-05, + "loss": 2.4271, + "step": 3970 + }, + { + "epoch": 0.9105709699610182, + "grad_norm": 0.3914699025929932, + "learning_rate": 2.0828105147154273e-05, + "loss": 2.2875, + "step": 3971 + }, + { + "epoch": 0.9108002751662463, + "grad_norm": 0.42745731214791044, + "learning_rate": 2.0722174323423605e-05, + "loss": 2.5284, + "step": 3972 + }, + { + "epoch": 0.9110295803714744, + "grad_norm": 0.39562355245971126, + "learning_rate": 2.061650786613545e-05, + "loss": 2.4053, + "step": 3973 + }, + { + "epoch": 0.9112588855767025, + "grad_norm": 0.4145075752859725, + "learning_rate": 2.0511105833574683e-05, + "loss": 2.4003, + "step": 3974 + }, + { + "epoch": 0.9114881907819308, + "grad_norm": 0.40838277539399626, + "learning_rate": 2.040596828388058e-05, + "loss": 2.5963, + "step": 3975 + }, + { + "epoch": 0.9117174959871589, + "grad_norm": 0.4403439582364569, + "learning_rate": 2.0301095275046144e-05, + "loss": 2.3714, + "step": 3976 + }, + { + "epoch": 0.911946801192387, + "grad_norm": 0.42658130831070845, + "learning_rate": 2.019648686491865e-05, + "loss": 2.5177, + "step": 3977 + }, + { + "epoch": 0.9121761063976153, + "grad_norm": 0.4451097724677879, + "learning_rate": 2.009214311119939e-05, + "loss": 2.466, + "step": 3978 + }, + { + "epoch": 0.9124054116028434, + "grad_norm": 0.4076707400612246, + "learning_rate": 1.9988064071443767e-05, + "loss": 2.3049, + "step": 3979 + }, + { + "epoch": 0.9126347168080715, + "grad_norm": 0.42776456903784954, + "learning_rate": 1.9884249803060915e-05, + "loss": 2.3915, + "step": 3980 + }, + { + "epoch": 0.9128640220132997, + "grad_norm": 0.40852629968973, + "learning_rate": 1.9780700363314253e-05, + "loss": 2.5798, + "step": 3981 + }, + { + "epoch": 0.9130933272185279, + "grad_norm": 0.4368434085708607, + "learning_rate": 1.9677415809320865e-05, + "loss": 2.406, + "step": 3982 + }, + { + "epoch": 0.913322632423756, + "grad_norm": 0.4210190089194239, + "learning_rate": 1.957439619805196e-05, + "loss": 2.3595, + "step": 3983 + }, + { + "epoch": 0.9135519376289841, + "grad_norm": 0.40688568312913076, + "learning_rate": 1.947164158633219e-05, + "loss": 2.3407, + "step": 3984 + }, + { + "epoch": 0.9137812428342124, + "grad_norm": 0.4177096482361305, + "learning_rate": 1.9369152030840554e-05, + "loss": 2.4217, + "step": 3985 + }, + { + "epoch": 0.9140105480394405, + "grad_norm": 0.40622316663286506, + "learning_rate": 1.926692758810955e-05, + "loss": 2.5126, + "step": 3986 + }, + { + "epoch": 0.9142398532446686, + "grad_norm": 0.44474868510077814, + "learning_rate": 1.916496831452552e-05, + "loss": 2.5085, + "step": 3987 + }, + { + "epoch": 0.9144691584498968, + "grad_norm": 0.4163684117605436, + "learning_rate": 1.9063274266328368e-05, + "loss": 2.37, + "step": 3988 + }, + { + "epoch": 0.914698463655125, + "grad_norm": 0.40346384435503196, + "learning_rate": 1.8961845499611998e-05, + "loss": 2.5901, + "step": 3989 + }, + { + "epoch": 0.9149277688603531, + "grad_norm": 0.4595232999303116, + "learning_rate": 1.886068207032382e-05, + "loss": 2.3405, + "step": 3990 + }, + { + "epoch": 0.9151570740655813, + "grad_norm": 0.4183420125743051, + "learning_rate": 1.8759784034264925e-05, + "loss": 2.3523, + "step": 3991 + }, + { + "epoch": 0.9153863792708095, + "grad_norm": 0.4016045842180325, + "learning_rate": 1.865915144708985e-05, + "loss": 2.5229, + "step": 3992 + }, + { + "epoch": 0.9156156844760376, + "grad_norm": 0.42202035189965514, + "learning_rate": 1.855878436430708e-05, + "loss": 2.3285, + "step": 3993 + }, + { + "epoch": 0.9158449896812658, + "grad_norm": 0.3962833324848743, + "learning_rate": 1.8458682841278162e-05, + "loss": 2.3253, + "step": 3994 + }, + { + "epoch": 0.916074294886494, + "grad_norm": 0.409023153525658, + "learning_rate": 1.835884693321871e-05, + "loss": 2.4584, + "step": 3995 + }, + { + "epoch": 0.9163036000917221, + "grad_norm": 0.44152383108123444, + "learning_rate": 1.825927669519728e-05, + "loss": 2.3973, + "step": 3996 + }, + { + "epoch": 0.9165329052969502, + "grad_norm": 0.40196005009207086, + "learning_rate": 1.8159972182136386e-05, + "loss": 2.481, + "step": 3997 + }, + { + "epoch": 0.9167622105021784, + "grad_norm": 0.40419128327920206, + "learning_rate": 1.8060933448811546e-05, + "loss": 2.275, + "step": 3998 + }, + { + "epoch": 0.9169915157074066, + "grad_norm": 0.4215065671053842, + "learning_rate": 1.7962160549851945e-05, + "loss": 2.3471, + "step": 3999 + }, + { + "epoch": 0.9172208209126347, + "grad_norm": 0.44564667592643126, + "learning_rate": 1.7863653539740122e-05, + "loss": 2.4523, + "step": 4000 + }, + { + "epoch": 0.9174501261178629, + "grad_norm": 0.40318000102022195, + "learning_rate": 1.776541247281177e-05, + "loss": 2.4097, + "step": 4001 + }, + { + "epoch": 0.917679431323091, + "grad_norm": 0.40266298708314696, + "learning_rate": 1.7667437403255938e-05, + "loss": 2.3605, + "step": 4002 + }, + { + "epoch": 0.9179087365283192, + "grad_norm": 0.41137713655667946, + "learning_rate": 1.7569728385115224e-05, + "loss": 2.3235, + "step": 4003 + }, + { + "epoch": 0.9181380417335474, + "grad_norm": 0.410131471068046, + "learning_rate": 1.7472285472285186e-05, + "loss": 2.2718, + "step": 4004 + }, + { + "epoch": 0.9183673469387755, + "grad_norm": 0.41588352129816103, + "learning_rate": 1.7375108718514665e-05, + "loss": 2.4708, + "step": 4005 + }, + { + "epoch": 0.9185966521440037, + "grad_norm": 0.40057990886302725, + "learning_rate": 1.7278198177405614e-05, + "loss": 2.3054, + "step": 4006 + }, + { + "epoch": 0.9188259573492318, + "grad_norm": 0.404856322872949, + "learning_rate": 1.7181553902413438e-05, + "loss": 2.3689, + "step": 4007 + }, + { + "epoch": 0.91905526255446, + "grad_norm": 0.4268488547267201, + "learning_rate": 1.7085175946846333e-05, + "loss": 2.5102, + "step": 4008 + }, + { + "epoch": 0.9192845677596881, + "grad_norm": 0.4369596385165686, + "learning_rate": 1.698906436386577e-05, + "loss": 2.2453, + "step": 4009 + }, + { + "epoch": 0.9195138729649163, + "grad_norm": 0.6800877297893606, + "learning_rate": 1.689321920648623e-05, + "loss": 2.5084, + "step": 4010 + }, + { + "epoch": 0.9197431781701445, + "grad_norm": 0.4094920662353081, + "learning_rate": 1.679764052757532e-05, + "loss": 2.3984, + "step": 4011 + }, + { + "epoch": 0.9199724833753726, + "grad_norm": 0.39439259025289847, + "learning_rate": 1.6702328379853525e-05, + "loss": 2.307, + "step": 4012 + }, + { + "epoch": 0.9202017885806008, + "grad_norm": 0.39410514882384, + "learning_rate": 1.6607282815894464e-05, + "loss": 2.3144, + "step": 4013 + }, + { + "epoch": 0.920431093785829, + "grad_norm": 0.3926747183243991, + "learning_rate": 1.6512503888124643e-05, + "loss": 2.4075, + "step": 4014 + }, + { + "epoch": 0.9206603989910571, + "grad_norm": 0.4225767554150859, + "learning_rate": 1.6417991648823405e-05, + "loss": 2.3989, + "step": 4015 + }, + { + "epoch": 0.9208897041962852, + "grad_norm": 0.39944442811582104, + "learning_rate": 1.6323746150122997e-05, + "loss": 2.4672, + "step": 4016 + }, + { + "epoch": 0.9211190094015134, + "grad_norm": 0.43039598790007977, + "learning_rate": 1.6229767444008835e-05, + "loss": 2.3054, + "step": 4017 + }, + { + "epoch": 0.9213483146067416, + "grad_norm": 1.6906182205100768, + "learning_rate": 1.6136055582318844e-05, + "loss": 2.3947, + "step": 4018 + }, + { + "epoch": 0.9215776198119697, + "grad_norm": 0.3944302699513334, + "learning_rate": 1.604261061674378e-05, + "loss": 2.4184, + "step": 4019 + }, + { + "epoch": 0.9218069250171979, + "grad_norm": 0.4047197065280542, + "learning_rate": 1.5949432598827307e-05, + "loss": 2.4088, + "step": 4020 + }, + { + "epoch": 0.9220362302224261, + "grad_norm": 0.4389590871477028, + "learning_rate": 1.5856521579965865e-05, + "loss": 2.4815, + "step": 4021 + }, + { + "epoch": 0.9222655354276542, + "grad_norm": 0.43081593203697727, + "learning_rate": 1.576387761140846e-05, + "loss": 2.3945, + "step": 4022 + }, + { + "epoch": 0.9224948406328823, + "grad_norm": 0.41127481739062033, + "learning_rate": 1.5671500744256938e-05, + "loss": 2.2959, + "step": 4023 + }, + { + "epoch": 0.9227241458381106, + "grad_norm": 0.3897922152299257, + "learning_rate": 1.5579391029465705e-05, + "loss": 2.1859, + "step": 4024 + }, + { + "epoch": 0.9229534510433387, + "grad_norm": 0.4008217640900677, + "learning_rate": 1.5487548517841953e-05, + "loss": 2.4069, + "step": 4025 + }, + { + "epoch": 0.9231827562485668, + "grad_norm": 0.416670586742061, + "learning_rate": 1.539597326004527e-05, + "loss": 2.5481, + "step": 4026 + }, + { + "epoch": 0.923412061453795, + "grad_norm": 0.4104096846722136, + "learning_rate": 1.530466530658814e-05, + "loss": 2.4536, + "step": 4027 + }, + { + "epoch": 0.9236413666590232, + "grad_norm": 0.387393942572769, + "learning_rate": 1.5213624707835271e-05, + "loss": 2.2543, + "step": 4028 + }, + { + "epoch": 0.9238706718642513, + "grad_norm": 0.4076068878534301, + "learning_rate": 1.5122851514004054e-05, + "loss": 2.3958, + "step": 4029 + }, + { + "epoch": 0.9240999770694794, + "grad_norm": 0.4049657500255033, + "learning_rate": 1.503234577516438e-05, + "loss": 2.2842, + "step": 4030 + }, + { + "epoch": 0.9243292822747077, + "grad_norm": 0.40884238744903906, + "learning_rate": 1.4942107541238703e-05, + "loss": 2.4354, + "step": 4031 + }, + { + "epoch": 0.9245585874799358, + "grad_norm": 0.3948322071573605, + "learning_rate": 1.4852136862001764e-05, + "loss": 2.3221, + "step": 4032 + }, + { + "epoch": 0.9247878926851639, + "grad_norm": 0.4180102517863457, + "learning_rate": 1.4762433787080809e-05, + "loss": 2.4978, + "step": 4033 + }, + { + "epoch": 0.9250171978903922, + "grad_norm": 0.3938449442261802, + "learning_rate": 1.4672998365955314e-05, + "loss": 2.3127, + "step": 4034 + }, + { + "epoch": 0.9252465030956203, + "grad_norm": 0.43306971671721534, + "learning_rate": 1.4583830647957541e-05, + "loss": 2.4751, + "step": 4035 + }, + { + "epoch": 0.9254758083008484, + "grad_norm": 0.43287612178912854, + "learning_rate": 1.449493068227159e-05, + "loss": 2.5235, + "step": 4036 + }, + { + "epoch": 0.9257051135060765, + "grad_norm": 0.42406213008436017, + "learning_rate": 1.4406298517934068e-05, + "loss": 2.4343, + "step": 4037 + }, + { + "epoch": 0.9259344187113048, + "grad_norm": 0.38750525565215155, + "learning_rate": 1.4317934203834037e-05, + "loss": 2.5018, + "step": 4038 + }, + { + "epoch": 0.9261637239165329, + "grad_norm": 0.4211682316164755, + "learning_rate": 1.4229837788712562e-05, + "loss": 2.3827, + "step": 4039 + }, + { + "epoch": 0.926393029121761, + "grad_norm": 0.416831520328127, + "learning_rate": 1.4142009321163051e-05, + "loss": 2.5414, + "step": 4040 + }, + { + "epoch": 0.9266223343269893, + "grad_norm": 0.3958566791514544, + "learning_rate": 1.4054448849631085e-05, + "loss": 2.4079, + "step": 4041 + }, + { + "epoch": 0.9268516395322174, + "grad_norm": 0.4328111546366986, + "learning_rate": 1.3967156422414474e-05, + "loss": 2.463, + "step": 4042 + }, + { + "epoch": 0.9270809447374455, + "grad_norm": 0.42134658982150625, + "learning_rate": 1.3880132087663145e-05, + "loss": 2.429, + "step": 4043 + }, + { + "epoch": 0.9273102499426737, + "grad_norm": 0.4390810636907507, + "learning_rate": 1.379337589337909e-05, + "loss": 2.3886, + "step": 4044 + }, + { + "epoch": 0.9275395551479019, + "grad_norm": 0.43623341322809184, + "learning_rate": 1.3706887887416419e-05, + "loss": 2.3039, + "step": 4045 + }, + { + "epoch": 0.92776886035313, + "grad_norm": 0.43915933718302397, + "learning_rate": 1.3620668117481471e-05, + "loss": 2.3041, + "step": 4046 + }, + { + "epoch": 0.9279981655583581, + "grad_norm": 0.4261505274652647, + "learning_rate": 1.3534716631132316e-05, + "loss": 2.367, + "step": 4047 + }, + { + "epoch": 0.9282274707635864, + "grad_norm": 0.39721348275884044, + "learning_rate": 1.344903347577936e-05, + "loss": 2.3422, + "step": 4048 + }, + { + "epoch": 0.9284567759688145, + "grad_norm": 0.3970146645291773, + "learning_rate": 1.3363618698684853e-05, + "loss": 2.3823, + "step": 4049 + }, + { + "epoch": 0.9286860811740426, + "grad_norm": 0.4167516711289773, + "learning_rate": 1.3278472346962944e-05, + "loss": 2.362, + "step": 4050 + }, + { + "epoch": 0.9289153863792708, + "grad_norm": 0.3817768556341581, + "learning_rate": 1.3193594467579728e-05, + "loss": 2.3748, + "step": 4051 + }, + { + "epoch": 0.929144691584499, + "grad_norm": 0.36544238782620864, + "learning_rate": 1.3108985107353477e-05, + "loss": 2.2579, + "step": 4052 + }, + { + "epoch": 0.9293739967897271, + "grad_norm": 0.40903101398044917, + "learning_rate": 1.3024644312954026e-05, + "loss": 2.3405, + "step": 4053 + }, + { + "epoch": 0.9296033019949553, + "grad_norm": 0.4127563583547811, + "learning_rate": 1.2940572130903161e-05, + "loss": 2.3834, + "step": 4054 + }, + { + "epoch": 0.9298326072001835, + "grad_norm": 0.4167435630937459, + "learning_rate": 1.2856768607574564e-05, + "loss": 2.4569, + "step": 4055 + }, + { + "epoch": 0.9300619124054116, + "grad_norm": 0.4005151771451197, + "learning_rate": 1.2773233789193816e-05, + "loss": 2.2552, + "step": 4056 + }, + { + "epoch": 0.9302912176106397, + "grad_norm": 0.41836044191796407, + "learning_rate": 1.2689967721837947e-05, + "loss": 2.463, + "step": 4057 + }, + { + "epoch": 0.930520522815868, + "grad_norm": 0.4191427488786073, + "learning_rate": 1.2606970451436105e-05, + "loss": 2.3102, + "step": 4058 + }, + { + "epoch": 0.9307498280210961, + "grad_norm": 0.42915107610024783, + "learning_rate": 1.2524242023769006e-05, + "loss": 2.334, + "step": 4059 + }, + { + "epoch": 0.9309791332263242, + "grad_norm": 0.43128136343921897, + "learning_rate": 1.2441782484469145e-05, + "loss": 2.3472, + "step": 4060 + }, + { + "epoch": 0.9312084384315524, + "grad_norm": 0.3907339612546228, + "learning_rate": 1.2359591879020526e-05, + "loss": 2.3674, + "step": 4061 + }, + { + "epoch": 0.9314377436367806, + "grad_norm": 0.45546746179806064, + "learning_rate": 1.2277670252759e-05, + "loss": 2.4602, + "step": 4062 + }, + { + "epoch": 0.9316670488420087, + "grad_norm": 0.412184521310007, + "learning_rate": 1.2196017650872081e-05, + "loss": 2.3757, + "step": 4063 + }, + { + "epoch": 0.9318963540472369, + "grad_norm": 0.4145701646220118, + "learning_rate": 1.2114634118398637e-05, + "loss": 2.3727, + "step": 4064 + }, + { + "epoch": 0.932125659252465, + "grad_norm": 0.41032478540403283, + "learning_rate": 1.2033519700229367e-05, + "loss": 2.5632, + "step": 4065 + }, + { + "epoch": 0.9323549644576932, + "grad_norm": 0.4184180386520196, + "learning_rate": 1.1952674441106482e-05, + "loss": 2.3814, + "step": 4066 + }, + { + "epoch": 0.9325842696629213, + "grad_norm": 0.44986960072535875, + "learning_rate": 1.1872098385623586e-05, + "loss": 2.3422, + "step": 4067 + }, + { + "epoch": 0.9328135748681495, + "grad_norm": 0.4023316040374927, + "learning_rate": 1.1791791578226018e-05, + "loss": 2.4126, + "step": 4068 + }, + { + "epoch": 0.9330428800733777, + "grad_norm": 0.39654036627291384, + "learning_rate": 1.1711754063210289e-05, + "loss": 2.4757, + "step": 4069 + }, + { + "epoch": 0.9332721852786058, + "grad_norm": 0.40495399180289793, + "learning_rate": 1.1631985884724804e-05, + "loss": 2.3247, + "step": 4070 + }, + { + "epoch": 0.933501490483834, + "grad_norm": 0.43565251253673115, + "learning_rate": 1.155248708676887e-05, + "loss": 2.4116, + "step": 4071 + }, + { + "epoch": 0.9337307956890621, + "grad_norm": 0.44107379231685223, + "learning_rate": 1.147325771319374e-05, + "loss": 2.3339, + "step": 4072 + }, + { + "epoch": 0.9339601008942903, + "grad_norm": 0.4352238406535254, + "learning_rate": 1.1394297807701737e-05, + "loss": 2.4583, + "step": 4073 + }, + { + "epoch": 0.9341894060995185, + "grad_norm": 0.39763212952061544, + "learning_rate": 1.1315607413846518e-05, + "loss": 2.4565, + "step": 4074 + }, + { + "epoch": 0.9344187113047466, + "grad_norm": 0.4071798634885264, + "learning_rate": 1.1237186575033254e-05, + "loss": 2.3645, + "step": 4075 + }, + { + "epoch": 0.9346480165099748, + "grad_norm": 0.42155043358658095, + "learning_rate": 1.115903533451834e-05, + "loss": 2.3357, + "step": 4076 + }, + { + "epoch": 0.9348773217152029, + "grad_norm": 0.41850207165384484, + "learning_rate": 1.1081153735409522e-05, + "loss": 2.4068, + "step": 4077 + }, + { + "epoch": 0.9351066269204311, + "grad_norm": 0.396970091534828, + "learning_rate": 1.1003541820665763e-05, + "loss": 2.4565, + "step": 4078 + }, + { + "epoch": 0.9353359321256592, + "grad_norm": 0.4200256242206688, + "learning_rate": 1.0926199633097156e-05, + "loss": 2.4507, + "step": 4079 + }, + { + "epoch": 0.9355652373308874, + "grad_norm": 0.40197164948474584, + "learning_rate": 1.0849127215365296e-05, + "loss": 2.3549, + "step": 4080 + }, + { + "epoch": 0.9357945425361156, + "grad_norm": 0.4252943112562058, + "learning_rate": 1.0772324609982787e-05, + "loss": 2.4171, + "step": 4081 + }, + { + "epoch": 0.9360238477413437, + "grad_norm": 0.390174197309122, + "learning_rate": 1.0695791859313298e-05, + "loss": 2.3852, + "step": 4082 + }, + { + "epoch": 0.9362531529465719, + "grad_norm": 0.4178534088702647, + "learning_rate": 1.0619529005571893e-05, + "loss": 2.4502, + "step": 4083 + }, + { + "epoch": 0.9364824581518001, + "grad_norm": 0.40181789675534857, + "learning_rate": 1.0543536090824702e-05, + "loss": 2.3542, + "step": 4084 + }, + { + "epoch": 0.9367117633570282, + "grad_norm": 0.4014256940197799, + "learning_rate": 1.0467813156988748e-05, + "loss": 2.3876, + "step": 4085 + }, + { + "epoch": 0.9369410685622563, + "grad_norm": 0.392214938848277, + "learning_rate": 1.03923602458324e-05, + "loss": 2.276, + "step": 4086 + }, + { + "epoch": 0.9371703737674846, + "grad_norm": 0.39885879698852855, + "learning_rate": 1.0317177398975031e-05, + "loss": 2.5046, + "step": 4087 + }, + { + "epoch": 0.9373996789727127, + "grad_norm": 0.42146928793600924, + "learning_rate": 1.0242264657886858e-05, + "loss": 2.4031, + "step": 4088 + }, + { + "epoch": 0.9376289841779408, + "grad_norm": 0.4218406149933754, + "learning_rate": 1.0167622063889326e-05, + "loss": 2.4129, + "step": 4089 + }, + { + "epoch": 0.937858289383169, + "grad_norm": 0.4103233387827228, + "learning_rate": 1.0093249658154835e-05, + "loss": 2.4149, + "step": 4090 + }, + { + "epoch": 0.9380875945883972, + "grad_norm": 0.4491813580971542, + "learning_rate": 1.0019147481706625e-05, + "loss": 2.4742, + "step": 4091 + }, + { + "epoch": 0.9383168997936253, + "grad_norm": 0.38432149124617304, + "learning_rate": 9.945315575419e-06, + "loss": 2.2985, + "step": 4092 + }, + { + "epoch": 0.9385462049988534, + "grad_norm": 0.4370671148754573, + "learning_rate": 9.871753980017051e-06, + "loss": 2.4575, + "step": 4093 + }, + { + "epoch": 0.9387755102040817, + "grad_norm": 0.3829199328187466, + "learning_rate": 9.79846273607704e-06, + "loss": 2.3888, + "step": 4094 + }, + { + "epoch": 0.9390048154093098, + "grad_norm": 0.43017895381527693, + "learning_rate": 9.725441884025855e-06, + "loss": 2.385, + "step": 4095 + }, + { + "epoch": 0.9392341206145379, + "grad_norm": 0.4053800245947016, + "learning_rate": 9.652691464141272e-06, + "loss": 2.2938, + "step": 4096 + }, + { + "epoch": 0.9394634258197662, + "grad_norm": 0.4141069863367104, + "learning_rate": 9.580211516551862e-06, + "loss": 2.4349, + "step": 4097 + }, + { + "epoch": 0.9396927310249943, + "grad_norm": 0.4421142365100417, + "learning_rate": 9.50800208123731e-06, + "loss": 2.5688, + "step": 4098 + }, + { + "epoch": 0.9399220362302224, + "grad_norm": 0.4187328342506791, + "learning_rate": 9.436063198027589e-06, + "loss": 2.2925, + "step": 4099 + }, + { + "epoch": 0.9401513414354505, + "grad_norm": 0.42470763118915594, + "learning_rate": 9.364394906603902e-06, + "loss": 2.4271, + "step": 4100 + }, + { + "epoch": 0.9403806466406788, + "grad_norm": 0.4015228747968419, + "learning_rate": 9.292997246497959e-06, + "loss": 2.2648, + "step": 4101 + }, + { + "epoch": 0.9406099518459069, + "grad_norm": 0.45386301630886194, + "learning_rate": 9.221870257092203e-06, + "loss": 2.405, + "step": 4102 + }, + { + "epoch": 0.940839257051135, + "grad_norm": 0.4123372591357491, + "learning_rate": 9.151013977619693e-06, + "loss": 2.5091, + "step": 4103 + }, + { + "epoch": 0.9410685622563633, + "grad_norm": 0.41291014101365353, + "learning_rate": 9.080428447164502e-06, + "loss": 2.4579, + "step": 4104 + }, + { + "epoch": 0.9412978674615914, + "grad_norm": 0.40753617991381264, + "learning_rate": 9.010113704661038e-06, + "loss": 2.4076, + "step": 4105 + }, + { + "epoch": 0.9415271726668195, + "grad_norm": 0.42578466501525086, + "learning_rate": 8.940069788894389e-06, + "loss": 2.4529, + "step": 4106 + }, + { + "epoch": 0.9417564778720477, + "grad_norm": 0.410163773473269, + "learning_rate": 8.870296738500316e-06, + "loss": 2.4115, + "step": 4107 + }, + { + "epoch": 0.9419857830772759, + "grad_norm": 0.410626173199923, + "learning_rate": 8.800794591965256e-06, + "loss": 2.5131, + "step": 4108 + }, + { + "epoch": 0.942215088282504, + "grad_norm": 0.4229300612369237, + "learning_rate": 8.731563387626095e-06, + "loss": 2.39, + "step": 4109 + }, + { + "epoch": 0.9424443934877321, + "grad_norm": 0.4228804484360852, + "learning_rate": 8.662603163670236e-06, + "loss": 2.4267, + "step": 4110 + }, + { + "epoch": 0.9426736986929604, + "grad_norm": 0.41679647378427376, + "learning_rate": 8.59391395813569e-06, + "loss": 2.4165, + "step": 4111 + }, + { + "epoch": 0.9429030038981885, + "grad_norm": 0.44856587860292263, + "learning_rate": 8.525495808911155e-06, + "loss": 2.2693, + "step": 4112 + }, + { + "epoch": 0.9431323091034166, + "grad_norm": 0.41735532637810735, + "learning_rate": 8.457348753735328e-06, + "loss": 2.4735, + "step": 4113 + }, + { + "epoch": 0.9433616143086448, + "grad_norm": 0.41844217307909726, + "learning_rate": 8.389472830197808e-06, + "loss": 2.4687, + "step": 4114 + }, + { + "epoch": 0.943590919513873, + "grad_norm": 0.38415220287562185, + "learning_rate": 8.321868075738593e-06, + "loss": 2.3234, + "step": 4115 + }, + { + "epoch": 0.9438202247191011, + "grad_norm": 0.3984950040568315, + "learning_rate": 8.25453452764785e-06, + "loss": 2.4183, + "step": 4116 + }, + { + "epoch": 0.9440495299243293, + "grad_norm": 0.42744786272942986, + "learning_rate": 8.187472223066371e-06, + "loss": 2.3899, + "step": 4117 + }, + { + "epoch": 0.9442788351295575, + "grad_norm": 0.3903557937166001, + "learning_rate": 8.120681198985291e-06, + "loss": 2.5352, + "step": 4118 + }, + { + "epoch": 0.9445081403347856, + "grad_norm": 0.39724678034919586, + "learning_rate": 8.054161492246136e-06, + "loss": 2.3581, + "step": 4119 + }, + { + "epoch": 0.9447374455400137, + "grad_norm": 0.3886093464474875, + "learning_rate": 7.98791313954067e-06, + "loss": 2.4519, + "step": 4120 + }, + { + "epoch": 0.944966750745242, + "grad_norm": 0.3882273709986534, + "learning_rate": 7.921936177411049e-06, + "loss": 2.4822, + "step": 4121 + }, + { + "epoch": 0.9451960559504701, + "grad_norm": 0.40426433566448666, + "learning_rate": 7.856230642249718e-06, + "loss": 2.3754, + "step": 4122 + }, + { + "epoch": 0.9454253611556982, + "grad_norm": 0.4634779691925973, + "learning_rate": 7.790796570299463e-06, + "loss": 2.2566, + "step": 4123 + }, + { + "epoch": 0.9456546663609264, + "grad_norm": 0.4182880163070719, + "learning_rate": 7.725633997653193e-06, + "loss": 2.4072, + "step": 4124 + }, + { + "epoch": 0.9458839715661546, + "grad_norm": 0.4145547533269478, + "learning_rate": 7.660742960254207e-06, + "loss": 2.3545, + "step": 4125 + }, + { + "epoch": 0.9461132767713827, + "grad_norm": 0.42392327124074064, + "learning_rate": 7.59612349389599e-06, + "loss": 2.4825, + "step": 4126 + }, + { + "epoch": 0.9463425819766109, + "grad_norm": 0.39868458529394213, + "learning_rate": 7.531775634222138e-06, + "loss": 2.2953, + "step": 4127 + }, + { + "epoch": 0.946571887181839, + "grad_norm": 0.42673190994935006, + "learning_rate": 7.467699416726537e-06, + "loss": 2.3564, + "step": 4128 + }, + { + "epoch": 0.9468011923870672, + "grad_norm": 0.40705129009924607, + "learning_rate": 7.403894876753192e-06, + "loss": 2.4027, + "step": 4129 + }, + { + "epoch": 0.9470304975922953, + "grad_norm": 0.3980443700773334, + "learning_rate": 7.340362049496285e-06, + "loss": 2.3038, + "step": 4130 + }, + { + "epoch": 0.9472598027975235, + "grad_norm": 0.43631628843857734, + "learning_rate": 7.277100970000061e-06, + "loss": 2.3899, + "step": 4131 + }, + { + "epoch": 0.9474891080027517, + "grad_norm": 0.4346105401142604, + "learning_rate": 7.214111673158941e-06, + "loss": 2.4267, + "step": 4132 + }, + { + "epoch": 0.9477184132079798, + "grad_norm": 0.4322859261430703, + "learning_rate": 7.151394193717408e-06, + "loss": 2.4523, + "step": 4133 + }, + { + "epoch": 0.947947718413208, + "grad_norm": 0.4129198260461464, + "learning_rate": 7.088948566269959e-06, + "loss": 2.2743, + "step": 4134 + }, + { + "epoch": 0.9481770236184361, + "grad_norm": 0.3959602485005294, + "learning_rate": 7.026774825261151e-06, + "loss": 2.4241, + "step": 4135 + }, + { + "epoch": 0.9484063288236643, + "grad_norm": 0.4084085499722853, + "learning_rate": 6.964873004985717e-06, + "loss": 2.343, + "step": 4136 + }, + { + "epoch": 0.9486356340288925, + "grad_norm": 0.40960784748153434, + "learning_rate": 6.903243139588233e-06, + "loss": 2.2128, + "step": 4137 + }, + { + "epoch": 0.9488649392341206, + "grad_norm": 0.40552141670951697, + "learning_rate": 6.841885263063341e-06, + "loss": 2.4043, + "step": 4138 + }, + { + "epoch": 0.9490942444393488, + "grad_norm": 0.4043788193510212, + "learning_rate": 6.780799409255522e-06, + "loss": 2.3457, + "step": 4139 + }, + { + "epoch": 0.9493235496445769, + "grad_norm": 0.42582654202743864, + "learning_rate": 6.7199856118594916e-06, + "loss": 2.389, + "step": 4140 + }, + { + "epoch": 0.9495528548498051, + "grad_norm": 0.42157673699546255, + "learning_rate": 6.659443904419637e-06, + "loss": 2.5503, + "step": 4141 + }, + { + "epoch": 0.9497821600550332, + "grad_norm": 0.43247395469658567, + "learning_rate": 6.599174320330359e-06, + "loss": 2.5842, + "step": 4142 + }, + { + "epoch": 0.9500114652602614, + "grad_norm": 0.3693732526046562, + "learning_rate": 6.539176892836008e-06, + "loss": 2.4069, + "step": 4143 + }, + { + "epoch": 0.9502407704654896, + "grad_norm": 0.36418924076633313, + "learning_rate": 6.479451655030721e-06, + "loss": 2.4539, + "step": 4144 + }, + { + "epoch": 0.9504700756707177, + "grad_norm": 0.39744856544217877, + "learning_rate": 6.4199986398585375e-06, + "loss": 2.5202, + "step": 4145 + }, + { + "epoch": 0.9506993808759459, + "grad_norm": 0.40017342409548545, + "learning_rate": 6.3608178801133345e-06, + "loss": 2.3561, + "step": 4146 + }, + { + "epoch": 0.9509286860811741, + "grad_norm": 0.4197638968642321, + "learning_rate": 6.3019094084388884e-06, + "loss": 2.3865, + "step": 4147 + }, + { + "epoch": 0.9511579912864022, + "grad_norm": 0.4117559348838071, + "learning_rate": 6.243273257328652e-06, + "loss": 2.4912, + "step": 4148 + }, + { + "epoch": 0.9513872964916303, + "grad_norm": 0.39664767938730205, + "learning_rate": 6.18490945912592e-06, + "loss": 2.3655, + "step": 4149 + }, + { + "epoch": 0.9516166016968585, + "grad_norm": 0.3838152660837382, + "learning_rate": 6.1268180460238854e-06, + "loss": 2.4259, + "step": 4150 + }, + { + "epoch": 0.9518459069020867, + "grad_norm": 0.45605335900928456, + "learning_rate": 6.068999050065249e-06, + "loss": 2.4252, + "step": 4151 + }, + { + "epoch": 0.9520752121073148, + "grad_norm": 0.4135521451823031, + "learning_rate": 6.011452503142667e-06, + "loss": 2.3716, + "step": 4152 + }, + { + "epoch": 0.952304517312543, + "grad_norm": 0.4277162819219882, + "learning_rate": 5.9541784369983586e-06, + "loss": 2.3805, + "step": 4153 + }, + { + "epoch": 0.9525338225177712, + "grad_norm": 0.4113149112522117, + "learning_rate": 5.897176883224442e-06, + "loss": 2.419, + "step": 4154 + }, + { + "epoch": 0.9527631277229993, + "grad_norm": 0.4085869936708412, + "learning_rate": 5.840447873262433e-06, + "loss": 2.4936, + "step": 4155 + }, + { + "epoch": 0.9529924329282274, + "grad_norm": 0.3978603357864399, + "learning_rate": 5.7839914384038014e-06, + "loss": 2.3769, + "step": 4156 + }, + { + "epoch": 0.9532217381334557, + "grad_norm": 0.45245940491682085, + "learning_rate": 5.727807609789471e-06, + "loss": 2.278, + "step": 4157 + }, + { + "epoch": 0.9534510433386838, + "grad_norm": 0.4397613746749079, + "learning_rate": 5.671896418410094e-06, + "loss": 2.394, + "step": 4158 + }, + { + "epoch": 0.9536803485439119, + "grad_norm": 0.4935658695801412, + "learning_rate": 5.616257895105892e-06, + "loss": 2.3821, + "step": 4159 + }, + { + "epoch": 0.95390965374914, + "grad_norm": 0.38725792292457933, + "learning_rate": 5.560892070566703e-06, + "loss": 2.3279, + "step": 4160 + }, + { + "epoch": 0.9541389589543683, + "grad_norm": 0.4248061632860381, + "learning_rate": 5.505798975331933e-06, + "loss": 2.415, + "step": 4161 + }, + { + "epoch": 0.9543682641595964, + "grad_norm": 0.41407357933652017, + "learning_rate": 5.4509786397906046e-06, + "loss": 2.546, + "step": 4162 + }, + { + "epoch": 0.9545975693648245, + "grad_norm": 0.4475567103772526, + "learning_rate": 5.396431094181198e-06, + "loss": 2.448, + "step": 4163 + }, + { + "epoch": 0.9548268745700528, + "grad_norm": 0.3913579814636102, + "learning_rate": 5.3421563685919214e-06, + "loss": 2.5032, + "step": 4164 + }, + { + "epoch": 0.9550561797752809, + "grad_norm": 0.40882156157601585, + "learning_rate": 5.288154492960107e-06, + "loss": 2.3985, + "step": 4165 + }, + { + "epoch": 0.955285484980509, + "grad_norm": 0.4095976186749967, + "learning_rate": 5.234425497072981e-06, + "loss": 2.4831, + "step": 4166 + }, + { + "epoch": 0.9555147901857373, + "grad_norm": 0.43409086140166525, + "learning_rate": 5.1809694105671155e-06, + "loss": 2.4779, + "step": 4167 + }, + { + "epoch": 0.9557440953909654, + "grad_norm": 0.40481285254916494, + "learning_rate": 5.127786262928424e-06, + "loss": 2.4617, + "step": 4168 + }, + { + "epoch": 0.9559734005961935, + "grad_norm": 0.3907575313906487, + "learning_rate": 5.074876083492441e-06, + "loss": 2.2379, + "step": 4169 + }, + { + "epoch": 0.9562027058014216, + "grad_norm": 0.40430135387224214, + "learning_rate": 5.022238901444098e-06, + "loss": 2.4362, + "step": 4170 + }, + { + "epoch": 0.9564320110066499, + "grad_norm": 0.4015339431098572, + "learning_rate": 4.96987474581767e-06, + "loss": 2.4515, + "step": 4171 + }, + { + "epoch": 0.956661316211878, + "grad_norm": 0.42470946977683255, + "learning_rate": 4.917783645496887e-06, + "loss": 2.4461, + "step": 4172 + }, + { + "epoch": 0.9568906214171061, + "grad_norm": 0.4141443917434427, + "learning_rate": 4.865965629214819e-06, + "loss": 2.1921, + "step": 4173 + }, + { + "epoch": 0.9571199266223344, + "grad_norm": 0.4230063537596685, + "learning_rate": 4.8144207255539914e-06, + "loss": 2.3534, + "step": 4174 + }, + { + "epoch": 0.9573492318275625, + "grad_norm": 0.40302636209880327, + "learning_rate": 4.763148962946218e-06, + "loss": 2.4831, + "step": 4175 + }, + { + "epoch": 0.9575785370327906, + "grad_norm": 0.3745534143311019, + "learning_rate": 4.712150369672652e-06, + "loss": 2.2529, + "step": 4176 + }, + { + "epoch": 0.9578078422380188, + "grad_norm": 0.3952062527087716, + "learning_rate": 4.661424973863681e-06, + "loss": 2.2992, + "step": 4177 + }, + { + "epoch": 0.958037147443247, + "grad_norm": 0.3799045449541699, + "learning_rate": 4.610972803499313e-06, + "loss": 2.3339, + "step": 4178 + }, + { + "epoch": 0.9582664526484751, + "grad_norm": 0.40697288726163694, + "learning_rate": 4.560793886408398e-06, + "loss": 2.3452, + "step": 4179 + }, + { + "epoch": 0.9584957578537033, + "grad_norm": 0.42103958748441694, + "learning_rate": 4.510888250269407e-06, + "loss": 2.4397, + "step": 4180 + }, + { + "epoch": 0.9587250630589315, + "grad_norm": 0.4131878639244808, + "learning_rate": 4.461255922609986e-06, + "loss": 2.3865, + "step": 4181 + }, + { + "epoch": 0.9589543682641596, + "grad_norm": 0.4337387940527842, + "learning_rate": 4.411896930806902e-06, + "loss": 2.598, + "step": 4182 + }, + { + "epoch": 0.9591836734693877, + "grad_norm": 0.40339590477137666, + "learning_rate": 4.362811302086267e-06, + "loss": 2.4142, + "step": 4183 + }, + { + "epoch": 0.959412978674616, + "grad_norm": 0.41675789649716644, + "learning_rate": 4.313999063523422e-06, + "loss": 2.4578, + "step": 4184 + }, + { + "epoch": 0.9596422838798441, + "grad_norm": 0.37937019928844873, + "learning_rate": 4.265460242042885e-06, + "loss": 2.2274, + "step": 4185 + }, + { + "epoch": 0.9598715890850722, + "grad_norm": 0.4147407343125415, + "learning_rate": 4.2171948644182946e-06, + "loss": 2.4331, + "step": 4186 + }, + { + "epoch": 0.9601008942903004, + "grad_norm": 0.412137861368512, + "learning_rate": 4.169202957272522e-06, + "loss": 2.3731, + "step": 4187 + }, + { + "epoch": 0.9603301994955286, + "grad_norm": 0.3997070368641328, + "learning_rate": 4.121484547077614e-06, + "loss": 2.398, + "step": 4188 + }, + { + "epoch": 0.9605595047007567, + "grad_norm": 0.401836740748224, + "learning_rate": 4.074039660154738e-06, + "loss": 2.4413, + "step": 4189 + }, + { + "epoch": 0.9607888099059849, + "grad_norm": 0.3971575071270662, + "learning_rate": 4.026868322674127e-06, + "loss": 2.4151, + "step": 4190 + }, + { + "epoch": 0.961018115111213, + "grad_norm": 0.42384477419919214, + "learning_rate": 3.9799705606551325e-06, + "loss": 2.4074, + "step": 4191 + }, + { + "epoch": 0.9612474203164412, + "grad_norm": 0.4178606043252162, + "learning_rate": 3.933346399966398e-06, + "loss": 2.6086, + "step": 4192 + }, + { + "epoch": 0.9614767255216693, + "grad_norm": 0.4190747602428646, + "learning_rate": 3.886995866325294e-06, + "loss": 2.3267, + "step": 4193 + }, + { + "epoch": 0.9617060307268975, + "grad_norm": 0.4366439924826478, + "learning_rate": 3.840918985298647e-06, + "loss": 2.3334, + "step": 4194 + }, + { + "epoch": 0.9619353359321257, + "grad_norm": 0.42110350310177264, + "learning_rate": 3.795115782302072e-06, + "loss": 2.5129, + "step": 4195 + }, + { + "epoch": 0.9621646411373538, + "grad_norm": 0.4950071505011167, + "learning_rate": 3.7495862826003592e-06, + "loss": 2.431, + "step": 4196 + }, + { + "epoch": 0.962393946342582, + "grad_norm": 0.40246130708465694, + "learning_rate": 3.704330511307197e-06, + "loss": 2.1951, + "step": 4197 + }, + { + "epoch": 0.9626232515478101, + "grad_norm": 0.4445058802053536, + "learning_rate": 3.659348493385395e-06, + "loss": 2.4795, + "step": 4198 + }, + { + "epoch": 0.9628525567530383, + "grad_norm": 0.4176072511754996, + "learning_rate": 3.614640253646828e-06, + "loss": 2.3814, + "step": 4199 + }, + { + "epoch": 0.9630818619582665, + "grad_norm": 0.40223692819690704, + "learning_rate": 3.5702058167521035e-06, + "loss": 2.3899, + "step": 4200 + }, + { + "epoch": 0.9633111671634946, + "grad_norm": 0.411765815150142, + "learning_rate": 3.5260452072110594e-06, + "loss": 2.5357, + "step": 4201 + }, + { + "epoch": 0.9635404723687228, + "grad_norm": 0.4624064426526554, + "learning_rate": 3.4821584493823776e-06, + "loss": 2.4878, + "step": 4202 + }, + { + "epoch": 0.9637697775739509, + "grad_norm": 0.39620546562415665, + "learning_rate": 3.4385455674737498e-06, + "loss": 2.5086, + "step": 4203 + }, + { + "epoch": 0.9639990827791791, + "grad_norm": 0.41786550278880574, + "learning_rate": 3.3952065855415994e-06, + "loss": 2.2167, + "step": 4204 + }, + { + "epoch": 0.9642283879844072, + "grad_norm": 0.430035740852092, + "learning_rate": 3.3521415274915256e-06, + "loss": 2.5018, + "step": 4205 + }, + { + "epoch": 0.9644576931896354, + "grad_norm": 0.4125601273837197, + "learning_rate": 3.309350417077972e-06, + "loss": 2.441, + "step": 4206 + }, + { + "epoch": 0.9646869983948636, + "grad_norm": 0.3941858257182658, + "learning_rate": 3.2668332779041133e-06, + "loss": 2.4744, + "step": 4207 + }, + { + "epoch": 0.9649163036000917, + "grad_norm": 0.4060597700713832, + "learning_rate": 3.2245901334221893e-06, + "loss": 2.2854, + "step": 4208 + }, + { + "epoch": 0.9651456088053199, + "grad_norm": 0.42527769201840415, + "learning_rate": 3.1826210069332838e-06, + "loss": 2.3916, + "step": 4209 + }, + { + "epoch": 0.9653749140105481, + "grad_norm": 0.39364100457621887, + "learning_rate": 3.1409259215871567e-06, + "loss": 2.3291, + "step": 4210 + }, + { + "epoch": 0.9656042192157762, + "grad_norm": 0.42139765275475344, + "learning_rate": 3.0995049003826324e-06, + "loss": 2.3989, + "step": 4211 + }, + { + "epoch": 0.9658335244210043, + "grad_norm": 0.43444629588892103, + "learning_rate": 3.058357966167213e-06, + "loss": 2.2903, + "step": 4212 + }, + { + "epoch": 0.9660628296262325, + "grad_norm": 0.4159655035370089, + "learning_rate": 3.017485141637355e-06, + "loss": 2.3862, + "step": 4213 + }, + { + "epoch": 0.9662921348314607, + "grad_norm": 0.4408938029850508, + "learning_rate": 2.9768864493381344e-06, + "loss": 2.3103, + "step": 4214 + }, + { + "epoch": 0.9665214400366888, + "grad_norm": 0.41423756610211654, + "learning_rate": 2.9365619116636376e-06, + "loss": 2.3067, + "step": 4215 + }, + { + "epoch": 0.966750745241917, + "grad_norm": 0.41427746376914454, + "learning_rate": 2.896511550856462e-06, + "loss": 2.5025, + "step": 4216 + }, + { + "epoch": 0.9669800504471452, + "grad_norm": 0.42847210267667807, + "learning_rate": 2.856735389008269e-06, + "loss": 2.3382, + "step": 4217 + }, + { + "epoch": 0.9672093556523733, + "grad_norm": 0.4243127449011545, + "learning_rate": 2.8172334480591755e-06, + "loss": 2.4655, + "step": 4218 + }, + { + "epoch": 0.9674386608576014, + "grad_norm": 0.4355628674854465, + "learning_rate": 2.778005749798307e-06, + "loss": 2.3798, + "step": 4219 + }, + { + "epoch": 0.9676679660628297, + "grad_norm": 0.4328016610178888, + "learning_rate": 2.739052315863355e-06, + "loss": 2.3618, + "step": 4220 + }, + { + "epoch": 0.9678972712680578, + "grad_norm": 0.40412069902841297, + "learning_rate": 2.700373167740744e-06, + "loss": 2.3076, + "step": 4221 + }, + { + "epoch": 0.9681265764732859, + "grad_norm": 0.4267938300254307, + "learning_rate": 2.6619683267656845e-06, + "loss": 2.4176, + "step": 4222 + }, + { + "epoch": 0.968355881678514, + "grad_norm": 0.40883803906820104, + "learning_rate": 2.62383781412201e-06, + "loss": 2.4838, + "step": 4223 + }, + { + "epoch": 0.9685851868837423, + "grad_norm": 0.4059715276045125, + "learning_rate": 2.5859816508422286e-06, + "loss": 2.2756, + "step": 4224 + }, + { + "epoch": 0.9688144920889704, + "grad_norm": 0.3868090117179569, + "learning_rate": 2.5483998578076373e-06, + "loss": 2.3221, + "step": 4225 + }, + { + "epoch": 0.9690437972941985, + "grad_norm": 0.41674574467588155, + "learning_rate": 2.511092455747932e-06, + "loss": 2.3329, + "step": 4226 + }, + { + "epoch": 0.9692731024994268, + "grad_norm": 0.37864925838829105, + "learning_rate": 2.4740594652418736e-06, + "loss": 2.2894, + "step": 4227 + }, + { + "epoch": 0.9695024077046549, + "grad_norm": 0.41737559489917386, + "learning_rate": 2.4373009067163443e-06, + "loss": 2.2823, + "step": 4228 + }, + { + "epoch": 0.969731712909883, + "grad_norm": 0.42136131361364704, + "learning_rate": 2.4008168004472917e-06, + "loss": 2.3717, + "step": 4229 + }, + { + "epoch": 0.9699610181151113, + "grad_norm": 0.4210719418044602, + "learning_rate": 2.364607166559063e-06, + "loss": 2.4069, + "step": 4230 + }, + { + "epoch": 0.9701903233203394, + "grad_norm": 0.40001978083157574, + "learning_rate": 2.3286720250246253e-06, + "loss": 2.345, + "step": 4231 + }, + { + "epoch": 0.9704196285255675, + "grad_norm": 0.4184644423451954, + "learning_rate": 2.2930113956655674e-06, + "loss": 2.3449, + "step": 4232 + }, + { + "epoch": 0.9706489337307956, + "grad_norm": 0.424251652655412, + "learning_rate": 2.2576252981520994e-06, + "loss": 2.2756, + "step": 4233 + }, + { + "epoch": 0.9708782389360239, + "grad_norm": 0.4220588291413228, + "learning_rate": 2.2225137520029416e-06, + "loss": 2.3004, + "step": 4234 + }, + { + "epoch": 0.971107544141252, + "grad_norm": 0.40780442935764677, + "learning_rate": 2.1876767765853233e-06, + "loss": 2.3429, + "step": 4235 + }, + { + "epoch": 0.9713368493464801, + "grad_norm": 0.43874849834238755, + "learning_rate": 2.153114391115152e-06, + "loss": 2.3402, + "step": 4236 + }, + { + "epoch": 0.9715661545517084, + "grad_norm": 0.443674754990103, + "learning_rate": 2.118826614656788e-06, + "loss": 2.5385, + "step": 4237 + }, + { + "epoch": 0.9717954597569365, + "grad_norm": 0.4303924938838481, + "learning_rate": 2.084813466123159e-06, + "loss": 2.2114, + "step": 4238 + }, + { + "epoch": 0.9720247649621646, + "grad_norm": 0.3947256785453821, + "learning_rate": 2.051074964275701e-06, + "loss": 2.494, + "step": 4239 + }, + { + "epoch": 0.9722540701673928, + "grad_norm": 0.422906841263788, + "learning_rate": 2.0176111277242502e-06, + "loss": 2.3626, + "step": 4240 + }, + { + "epoch": 0.972483375372621, + "grad_norm": 0.41167809596161886, + "learning_rate": 1.984421974927375e-06, + "loss": 2.3176, + "step": 4241 + }, + { + "epoch": 0.9727126805778491, + "grad_norm": 0.4547287383129543, + "learning_rate": 1.95150752419182e-06, + "loss": 2.3452, + "step": 4242 + }, + { + "epoch": 0.9729419857830772, + "grad_norm": 0.4226594648635396, + "learning_rate": 1.9188677936731734e-06, + "loss": 2.3773, + "step": 4243 + }, + { + "epoch": 0.9731712909883055, + "grad_norm": 0.4056493532143581, + "learning_rate": 1.8865028013751451e-06, + "loss": 2.3689, + "step": 4244 + }, + { + "epoch": 0.9734005961935336, + "grad_norm": 0.41601675344700995, + "learning_rate": 1.8544125651501208e-06, + "loss": 2.4156, + "step": 4245 + }, + { + "epoch": 0.9736299013987617, + "grad_norm": 0.4001198568478675, + "learning_rate": 1.8225971026987753e-06, + "loss": 2.4545, + "step": 4246 + }, + { + "epoch": 0.97385920660399, + "grad_norm": 0.4361331232878844, + "learning_rate": 1.7910564315704035e-06, + "loss": 2.4387, + "step": 4247 + }, + { + "epoch": 0.9740885118092181, + "grad_norm": 0.39504861431423616, + "learning_rate": 1.7597905691625893e-06, + "loss": 2.3982, + "step": 4248 + }, + { + "epoch": 0.9743178170144462, + "grad_norm": 0.43138135622145396, + "learning_rate": 1.7287995327214257e-06, + "loss": 2.3482, + "step": 4249 + }, + { + "epoch": 0.9745471222196744, + "grad_norm": 0.3978025167754971, + "learning_rate": 1.6980833393412388e-06, + "loss": 2.3596, + "step": 4250 + }, + { + "epoch": 0.9747764274249026, + "grad_norm": 0.4213869860433641, + "learning_rate": 1.6676420059649754e-06, + "loss": 2.39, + "step": 4251 + }, + { + "epoch": 0.9750057326301307, + "grad_norm": 0.40409873303062077, + "learning_rate": 1.6374755493838155e-06, + "loss": 2.396, + "step": 4252 + }, + { + "epoch": 0.9752350378353588, + "grad_norm": 0.4387177870662379, + "learning_rate": 1.6075839862374485e-06, + "loss": 2.3572, + "step": 4253 + }, + { + "epoch": 0.975464343040587, + "grad_norm": 0.4175386280434851, + "learning_rate": 1.5779673330137966e-06, + "loss": 2.4407, + "step": 4254 + }, + { + "epoch": 0.9756936482458152, + "grad_norm": 0.3972037791687638, + "learning_rate": 1.5486256060492366e-06, + "loss": 2.4583, + "step": 4255 + }, + { + "epoch": 0.9759229534510433, + "grad_norm": 0.4234526887072704, + "learning_rate": 1.5195588215283773e-06, + "loss": 2.4427, + "step": 4256 + }, + { + "epoch": 0.9761522586562715, + "grad_norm": 0.3952556725425923, + "learning_rate": 1.4907669954844495e-06, + "loss": 2.3656, + "step": 4257 + }, + { + "epoch": 0.9763815638614997, + "grad_norm": 0.4456496513412184, + "learning_rate": 1.462250143798638e-06, + "loss": 2.4214, + "step": 4258 + }, + { + "epoch": 0.9766108690667278, + "grad_norm": 0.4065978709061907, + "learning_rate": 1.434008282200805e-06, + "loss": 2.3566, + "step": 4259 + }, + { + "epoch": 0.976840174271956, + "grad_norm": 0.3963630652437679, + "learning_rate": 1.406041426268878e-06, + "loss": 2.2354, + "step": 4260 + }, + { + "epoch": 0.9770694794771841, + "grad_norm": 0.41553588326671015, + "learning_rate": 1.3783495914291844e-06, + "loss": 2.3584, + "step": 4261 + }, + { + "epoch": 0.9772987846824123, + "grad_norm": 0.45036559110398505, + "learning_rate": 1.3509327929563942e-06, + "loss": 2.3313, + "step": 4262 + }, + { + "epoch": 0.9775280898876404, + "grad_norm": 0.44075897937742975, + "learning_rate": 1.3237910459734104e-06, + "loss": 2.4276, + "step": 4263 + }, + { + "epoch": 0.9777573950928686, + "grad_norm": 0.4138599303822533, + "learning_rate": 1.2969243654514796e-06, + "loss": 2.5161, + "step": 4264 + }, + { + "epoch": 0.9779867002980968, + "grad_norm": 0.43775147343172194, + "learning_rate": 1.270332766210025e-06, + "loss": 2.2763, + "step": 4265 + }, + { + "epoch": 0.9782160055033249, + "grad_norm": 0.4121562891016972, + "learning_rate": 1.2440162629168138e-06, + "loss": 2.4829, + "step": 4266 + }, + { + "epoch": 0.9784453107085531, + "grad_norm": 0.4433359043943498, + "learning_rate": 1.2179748700879012e-06, + "loss": 2.4164, + "step": 4267 + }, + { + "epoch": 0.9786746159137812, + "grad_norm": 0.3956100057818747, + "learning_rate": 1.1922086020874634e-06, + "loss": 2.4152, + "step": 4268 + }, + { + "epoch": 0.9789039211190094, + "grad_norm": 0.44036134612578914, + "learning_rate": 1.1667174731280205e-06, + "loss": 2.5406, + "step": 4269 + }, + { + "epoch": 0.9791332263242376, + "grad_norm": 0.40869519265689375, + "learning_rate": 1.1415014972703808e-06, + "loss": 2.39, + "step": 4270 + }, + { + "epoch": 0.9793625315294657, + "grad_norm": 0.4369477100875798, + "learning_rate": 1.1165606884234182e-06, + "loss": 2.4194, + "step": 4271 + }, + { + "epoch": 0.9795918367346939, + "grad_norm": 0.4018345309272542, + "learning_rate": 1.0918950603444055e-06, + "loss": 2.4824, + "step": 4272 + }, + { + "epoch": 0.9798211419399221, + "grad_norm": 0.415071194861111, + "learning_rate": 1.0675046266386268e-06, + "loss": 2.4888, + "step": 4273 + }, + { + "epoch": 0.9800504471451502, + "grad_norm": 0.41727399326599096, + "learning_rate": 1.0433894007597643e-06, + "loss": 2.3927, + "step": 4274 + }, + { + "epoch": 0.9802797523503783, + "grad_norm": 0.3954003655410629, + "learning_rate": 1.019549396009567e-06, + "loss": 2.3531, + "step": 4275 + }, + { + "epoch": 0.9805090575556065, + "grad_norm": 0.6222911351978521, + "learning_rate": 9.959846255381267e-07, + "loss": 2.3819, + "step": 4276 + }, + { + "epoch": 0.9807383627608347, + "grad_norm": 0.41927772390321794, + "learning_rate": 9.726951023434348e-07, + "loss": 2.3733, + "step": 4277 + }, + { + "epoch": 0.9809676679660628, + "grad_norm": 0.4307334670714879, + "learning_rate": 9.496808392719935e-07, + "loss": 2.429, + "step": 4278 + }, + { + "epoch": 0.981196973171291, + "grad_norm": 0.4151977173931741, + "learning_rate": 9.269418490182591e-07, + "loss": 2.4504, + "step": 4279 + }, + { + "epoch": 0.9814262783765192, + "grad_norm": 0.441009852668005, + "learning_rate": 9.044781441249205e-07, + "loss": 2.3618, + "step": 4280 + }, + { + "epoch": 0.9816555835817473, + "grad_norm": 0.39709350176038943, + "learning_rate": 8.822897369827332e-07, + "loss": 2.4645, + "step": 4281 + }, + { + "epoch": 0.9818848887869754, + "grad_norm": 0.4504985139108724, + "learning_rate": 8.603766398307956e-07, + "loss": 2.4633, + "step": 4282 + }, + { + "epoch": 0.9821141939922037, + "grad_norm": 0.4365096829722913, + "learning_rate": 8.387388647561611e-07, + "loss": 2.3877, + "step": 4283 + }, + { + "epoch": 0.9823434991974318, + "grad_norm": 0.41396561851326635, + "learning_rate": 8.173764236940606e-07, + "loss": 2.4027, + "step": 4284 + }, + { + "epoch": 0.9825728044026599, + "grad_norm": 0.42779110774392276, + "learning_rate": 7.962893284279016e-07, + "loss": 2.3732, + "step": 4285 + }, + { + "epoch": 0.982802109607888, + "grad_norm": 0.39912794272671337, + "learning_rate": 7.754775905891575e-07, + "loss": 2.3994, + "step": 4286 + }, + { + "epoch": 0.9830314148131163, + "grad_norm": 0.4315213051613043, + "learning_rate": 7.549412216574791e-07, + "loss": 2.4502, + "step": 4287 + }, + { + "epoch": 0.9832607200183444, + "grad_norm": 0.390302652425847, + "learning_rate": 7.346802329605828e-07, + "loss": 2.4843, + "step": 4288 + }, + { + "epoch": 0.9834900252235725, + "grad_norm": 0.4332373845248671, + "learning_rate": 7.146946356743067e-07, + "loss": 2.443, + "step": 4289 + }, + { + "epoch": 0.9837193304288008, + "grad_norm": 0.41271366583449254, + "learning_rate": 6.949844408225548e-07, + "loss": 2.2594, + "step": 4290 + }, + { + "epoch": 0.9839486356340289, + "grad_norm": 0.41774030951128654, + "learning_rate": 6.755496592773524e-07, + "loss": 2.4732, + "step": 4291 + }, + { + "epoch": 0.984177940839257, + "grad_norm": 0.3900682422286095, + "learning_rate": 6.563903017587358e-07, + "loss": 2.3969, + "step": 4292 + }, + { + "epoch": 0.9844072460444853, + "grad_norm": 0.4036073066071854, + "learning_rate": 6.375063788349733e-07, + "loss": 2.4604, + "step": 4293 + }, + { + "epoch": 0.9846365512497134, + "grad_norm": 0.40076515653814376, + "learning_rate": 6.188979009222329e-07, + "loss": 2.5151, + "step": 4294 + }, + { + "epoch": 0.9848658564549415, + "grad_norm": 0.41681490123208553, + "learning_rate": 6.005648782848594e-07, + "loss": 2.5804, + "step": 4295 + }, + { + "epoch": 0.9850951616601696, + "grad_norm": 0.40838077123318617, + "learning_rate": 5.825073210352083e-07, + "loss": 2.521, + "step": 4296 + }, + { + "epoch": 0.9853244668653979, + "grad_norm": 0.4175172018055849, + "learning_rate": 5.647252391337565e-07, + "loss": 2.5333, + "step": 4297 + }, + { + "epoch": 0.985553772070626, + "grad_norm": 0.42162470223277254, + "learning_rate": 5.472186423889358e-07, + "loss": 2.2854, + "step": 4298 + }, + { + "epoch": 0.9857830772758541, + "grad_norm": 0.4038845337513078, + "learning_rate": 5.299875404572441e-07, + "loss": 2.2921, + "step": 4299 + }, + { + "epoch": 0.9860123824810824, + "grad_norm": 0.4100262565010712, + "learning_rate": 5.130319428433005e-07, + "loss": 2.3815, + "step": 4300 + }, + { + "epoch": 0.9862416876863105, + "grad_norm": 0.41604160400067763, + "learning_rate": 4.963518588996796e-07, + "loss": 2.3184, + "step": 4301 + }, + { + "epoch": 0.9864709928915386, + "grad_norm": 0.4196452082455565, + "learning_rate": 4.799472978269659e-07, + "loss": 2.3269, + "step": 4302 + }, + { + "epoch": 0.9867002980967668, + "grad_norm": 0.402082109073192, + "learning_rate": 4.638182686738657e-07, + "loss": 2.5792, + "step": 4303 + }, + { + "epoch": 0.986929603301995, + "grad_norm": 0.42869927255312845, + "learning_rate": 4.4796478033704015e-07, + "loss": 2.238, + "step": 4304 + }, + { + "epoch": 0.9871589085072231, + "grad_norm": 0.42428450763448344, + "learning_rate": 4.3238684156110543e-07, + "loss": 2.2957, + "step": 4305 + }, + { + "epoch": 0.9873882137124512, + "grad_norm": 0.4901876637154223, + "learning_rate": 4.170844609387991e-07, + "loss": 2.5098, + "step": 4306 + }, + { + "epoch": 0.9876175189176795, + "grad_norm": 0.4174584975571827, + "learning_rate": 4.020576469108139e-07, + "loss": 2.3921, + "step": 4307 + }, + { + "epoch": 0.9878468241229076, + "grad_norm": 0.40709059062552727, + "learning_rate": 3.873064077657973e-07, + "loss": 2.2436, + "step": 4308 + }, + { + "epoch": 0.9880761293281357, + "grad_norm": 0.40143116475678137, + "learning_rate": 3.7283075164046274e-07, + "loss": 2.4361, + "step": 4309 + }, + { + "epoch": 0.988305434533364, + "grad_norm": 0.4194802000884478, + "learning_rate": 3.5863068651953436e-07, + "loss": 2.3671, + "step": 4310 + }, + { + "epoch": 0.9885347397385921, + "grad_norm": 0.46908938418991264, + "learning_rate": 3.4470622023557995e-07, + "loss": 2.3219, + "step": 4311 + }, + { + "epoch": 0.9887640449438202, + "grad_norm": 0.4164366956792953, + "learning_rate": 3.3105736046928903e-07, + "loss": 2.5135, + "step": 4312 + }, + { + "epoch": 0.9889933501490484, + "grad_norm": 0.42061378601809807, + "learning_rate": 3.176841147492504e-07, + "loss": 2.2572, + "step": 4313 + }, + { + "epoch": 0.9892226553542766, + "grad_norm": 0.4122356009348431, + "learning_rate": 3.0458649045211895e-07, + "loss": 2.2465, + "step": 4314 + }, + { + "epoch": 0.9894519605595047, + "grad_norm": 0.4328460843671504, + "learning_rate": 2.9176449480244895e-07, + "loss": 2.3361, + "step": 4315 + }, + { + "epoch": 0.9896812657647328, + "grad_norm": 0.40570057847212226, + "learning_rate": 2.792181348726941e-07, + "loss": 2.3266, + "step": 4316 + }, + { + "epoch": 0.989910570969961, + "grad_norm": 0.40119328352712497, + "learning_rate": 2.6694741758342967e-07, + "loss": 2.3097, + "step": 4317 + }, + { + "epoch": 0.9901398761751892, + "grad_norm": 0.42219240587859763, + "learning_rate": 2.549523497030193e-07, + "loss": 2.6148, + "step": 4318 + }, + { + "epoch": 0.9903691813804173, + "grad_norm": 0.39021051295171066, + "learning_rate": 2.432329378478926e-07, + "loss": 2.404, + "step": 4319 + }, + { + "epoch": 0.9905984865856455, + "grad_norm": 0.399118722878862, + "learning_rate": 2.3178918848243413e-07, + "loss": 2.479, + "step": 4320 + }, + { + "epoch": 0.9908277917908737, + "grad_norm": 0.4555223476481628, + "learning_rate": 2.2062110791892798e-07, + "loss": 2.5814, + "step": 4321 + }, + { + "epoch": 0.9910570969961018, + "grad_norm": 0.4265290221457086, + "learning_rate": 2.097287023176131e-07, + "loss": 2.4912, + "step": 4322 + }, + { + "epoch": 0.99128640220133, + "grad_norm": 0.42724699123402116, + "learning_rate": 1.9911197768662792e-07, + "loss": 2.3737, + "step": 4323 + }, + { + "epoch": 0.9915157074065581, + "grad_norm": 0.4667527956214658, + "learning_rate": 1.8877093988212135e-07, + "loss": 2.4017, + "step": 4324 + }, + { + "epoch": 0.9917450126117863, + "grad_norm": 0.4222090265453155, + "learning_rate": 1.7870559460814173e-07, + "loss": 2.2441, + "step": 4325 + }, + { + "epoch": 0.9919743178170144, + "grad_norm": 0.41342257256674936, + "learning_rate": 1.6891594741663684e-07, + "loss": 2.5071, + "step": 4326 + }, + { + "epoch": 0.9922036230222426, + "grad_norm": 0.4133031749430915, + "learning_rate": 1.5940200370750947e-07, + "loss": 2.5045, + "step": 4327 + }, + { + "epoch": 0.9924329282274708, + "grad_norm": 0.39939735706812257, + "learning_rate": 1.5016376872867277e-07, + "loss": 2.3626, + "step": 4328 + }, + { + "epoch": 0.9926622334326989, + "grad_norm": 0.4356305929916336, + "learning_rate": 1.4120124757577291e-07, + "loss": 2.512, + "step": 4329 + }, + { + "epoch": 0.9928915386379271, + "grad_norm": 0.4504626069090871, + "learning_rate": 1.3251444519246647e-07, + "loss": 2.4874, + "step": 4330 + }, + { + "epoch": 0.9931208438431552, + "grad_norm": 0.4145266290230297, + "learning_rate": 1.2410336637047603e-07, + "loss": 2.4538, + "step": 4331 + }, + { + "epoch": 0.9933501490483834, + "grad_norm": 0.4324598882532313, + "learning_rate": 1.1596801574914606e-07, + "loss": 2.2556, + "step": 4332 + }, + { + "epoch": 0.9935794542536116, + "grad_norm": 0.39699063699426346, + "learning_rate": 1.081083978159425e-07, + "loss": 2.4312, + "step": 4333 + }, + { + "epoch": 0.9938087594588397, + "grad_norm": 0.4382295483359266, + "learning_rate": 1.0052451690617526e-07, + "loss": 2.4931, + "step": 4334 + }, + { + "epoch": 0.9940380646640679, + "grad_norm": 0.40988875809777964, + "learning_rate": 9.321637720310915e-08, + "loss": 2.46, + "step": 4335 + }, + { + "epoch": 0.994267369869296, + "grad_norm": 0.4187657434435745, + "learning_rate": 8.618398273779749e-08, + "loss": 2.445, + "step": 4336 + }, + { + "epoch": 0.9944966750745242, + "grad_norm": 0.4111901260294163, + "learning_rate": 7.942733738924845e-08, + "loss": 2.3134, + "step": 4337 + }, + { + "epoch": 0.9947259802797523, + "grad_norm": 0.4262476185726667, + "learning_rate": 7.294644488442526e-08, + "loss": 2.3271, + "step": 4338 + }, + { + "epoch": 0.9949552854849805, + "grad_norm": 0.41827747223204526, + "learning_rate": 6.6741308798135e-08, + "loss": 2.2649, + "step": 4339 + }, + { + "epoch": 0.9951845906902087, + "grad_norm": 0.4267508926620777, + "learning_rate": 6.081193255313977e-08, + "loss": 2.3091, + "step": 4340 + }, + { + "epoch": 0.9954138958954368, + "grad_norm": 0.44989668014063816, + "learning_rate": 5.5158319419934546e-08, + "loss": 2.4361, + "step": 4341 + }, + { + "epoch": 0.995643201100665, + "grad_norm": 0.3886287875400758, + "learning_rate": 4.978047251708029e-08, + "loss": 2.4266, + "step": 4342 + }, + { + "epoch": 0.9958725063058932, + "grad_norm": 0.39139735035164147, + "learning_rate": 4.4678394810981906e-08, + "loss": 2.3129, + "step": 4343 + }, + { + "epoch": 0.9961018115111213, + "grad_norm": 0.40770550229183855, + "learning_rate": 3.9852089115888226e-08, + "loss": 2.3505, + "step": 4344 + }, + { + "epoch": 0.9963311167163494, + "grad_norm": 0.4055509382780643, + "learning_rate": 3.5301558093947527e-08, + "loss": 2.33, + "step": 4345 + }, + { + "epoch": 0.9965604219215776, + "grad_norm": 0.396937196431943, + "learning_rate": 3.102680425520754e-08, + "loss": 2.3525, + "step": 4346 + }, + { + "epoch": 0.9967897271268058, + "grad_norm": 0.4487719922823648, + "learning_rate": 2.7027829957559925e-08, + "loss": 2.3993, + "step": 4347 + }, + { + "epoch": 0.9970190323320339, + "grad_norm": 0.4258524375924717, + "learning_rate": 2.3304637406906802e-08, + "loss": 2.355, + "step": 4348 + }, + { + "epoch": 0.997248337537262, + "grad_norm": 0.4041747176720719, + "learning_rate": 1.985722865682771e-08, + "loss": 2.3417, + "step": 4349 + }, + { + "epoch": 0.9974776427424903, + "grad_norm": 0.4201530980470467, + "learning_rate": 1.6685605608968145e-08, + "loss": 2.6461, + "step": 4350 + }, + { + "epoch": 0.9977069479477184, + "grad_norm": 0.3836635588343828, + "learning_rate": 1.3789770012762048e-08, + "loss": 2.3564, + "step": 4351 + }, + { + "epoch": 0.9979362531529465, + "grad_norm": 0.4262332637329797, + "learning_rate": 1.116972346548728e-08, + "loss": 2.404, + "step": 4352 + }, + { + "epoch": 0.9981655583581748, + "grad_norm": 0.42417723201798285, + "learning_rate": 8.825467412376665e-09, + "loss": 2.4422, + "step": 4353 + }, + { + "epoch": 0.9983948635634029, + "grad_norm": 0.4239703940461041, + "learning_rate": 6.757003146506957e-09, + "loss": 2.5329, + "step": 4354 + }, + { + "epoch": 0.998624168768631, + "grad_norm": 0.4182237934425861, + "learning_rate": 4.96433180879885e-09, + "loss": 2.2959, + "step": 4355 + }, + { + "epoch": 0.9988534739738592, + "grad_norm": 0.4369746331886112, + "learning_rate": 3.4474543881279906e-09, + "loss": 2.3, + "step": 4356 + }, + { + "epoch": 0.9990827791790874, + "grad_norm": 0.41910233866190966, + "learning_rate": 2.206371721158451e-09, + "loss": 2.2559, + "step": 4357 + }, + { + "epoch": 0.9993120843843155, + "grad_norm": 0.425580901507808, + "learning_rate": 1.241084492509259e-09, + "loss": 2.3766, + "step": 4358 + }, + { + "epoch": 0.9995413895895436, + "grad_norm": 0.4406177288767297, + "learning_rate": 5.515932345323549e-10, + "loss": 2.6794, + "step": 4359 + }, + { + "epoch": 0.9997706947947719, + "grad_norm": 0.4283902412991119, + "learning_rate": 1.37898327645658e-10, + "loss": 2.4155, + "step": 4360 + }, + { + "epoch": 1.0, + "grad_norm": 0.4129659068816933, + "learning_rate": 0.0, + "loss": 2.3808, + "step": 4361 + }, + { + "epoch": 1.0, + "step": 4361, + "total_flos": 54895682469888.0, + "train_loss": 2.5851684883903183, + "train_runtime": 11505.4297, + "train_samples_per_second": 48.51, + "train_steps_per_second": 0.379 + } + ], + "logging_steps": 1.0, + "max_steps": 4361, + "num_input_tokens_seen": 0, + "num_train_epochs": 1, + "save_steps": 50000.0, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": false, + "should_training_stop": false + }, + "attributes": {} + } + }, + "total_flos": 54895682469888.0, + "train_batch_size": 16, + "trial_name": null, + "trial_params": null +}