{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.200009225517782, "eval_steps": 500, "global_step": 3252, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.000369020711287421, "grad_norm": 32.204185485839844, "learning_rate": 0.0, "loss": 6.701247215270996, "step": 1 }, { "epoch": 0.000738041422574842, "grad_norm": 32.250003814697266, "learning_rate": 2.0000000000000003e-06, "loss": 6.400952339172363, "step": 2 }, { "epoch": 0.001107062133862263, "grad_norm": 24.697145462036133, "learning_rate": 4.000000000000001e-06, "loss": 6.332036972045898, "step": 3 }, { "epoch": 0.001476082845149684, "grad_norm": 18.79606819152832, "learning_rate": 6e-06, "loss": 6.286121845245361, "step": 4 }, { "epoch": 0.001845103556437105, "grad_norm": 18.390621185302734, "learning_rate": 8.000000000000001e-06, "loss": 6.1857757568359375, "step": 5 }, { "epoch": 0.002214124267724526, "grad_norm": 15.014426231384277, "learning_rate": 1e-05, "loss": 6.110501766204834, "step": 6 }, { "epoch": 0.002583144979011947, "grad_norm": 15.987100601196289, "learning_rate": 1.2e-05, "loss": 5.976349830627441, "step": 7 }, { "epoch": 0.002952165690299368, "grad_norm": 18.343597412109375, "learning_rate": 1.4000000000000001e-05, "loss": 6.035942077636719, "step": 8 }, { "epoch": 0.0033211864015867893, "grad_norm": 13.626900672912598, "learning_rate": 1.6000000000000003e-05, "loss": 6.0868072509765625, "step": 9 }, { "epoch": 0.00369020711287421, "grad_norm": 11.279500007629395, "learning_rate": 1.8e-05, "loss": 6.021449089050293, "step": 10 }, { "epoch": 0.004059227824161631, "grad_norm": 12.686036109924316, "learning_rate": 2e-05, "loss": 6.056863784790039, "step": 11 }, { "epoch": 0.004428248535449052, "grad_norm": 10.581238746643066, "learning_rate": 2.2000000000000003e-05, "loss": 5.860020637512207, "step": 12 }, { "epoch": 0.004797269246736473, "grad_norm": 8.909687042236328, "learning_rate": 2.4e-05, "loss": 5.828363418579102, "step": 13 }, { "epoch": 0.005166289958023894, "grad_norm": 6.024884223937988, "learning_rate": 2.6000000000000002e-05, "loss": 5.932955741882324, "step": 14 }, { "epoch": 0.0055353106693113155, "grad_norm": 7.050922393798828, "learning_rate": 2.8000000000000003e-05, "loss": 5.747481822967529, "step": 15 }, { "epoch": 0.005904331380598736, "grad_norm": 5.779654502868652, "learning_rate": 3e-05, "loss": 5.645102024078369, "step": 16 }, { "epoch": 0.006273352091886157, "grad_norm": 6.848283290863037, "learning_rate": 3.2000000000000005e-05, "loss": 6.164463996887207, "step": 17 }, { "epoch": 0.0066423728031735785, "grad_norm": 5.797471046447754, "learning_rate": 3.4000000000000007e-05, "loss": 5.728422164916992, "step": 18 }, { "epoch": 0.007011393514460999, "grad_norm": 6.957800388336182, "learning_rate": 3.6e-05, "loss": 5.617619514465332, "step": 19 }, { "epoch": 0.00738041422574842, "grad_norm": 6.550650596618652, "learning_rate": 3.8e-05, "loss": 5.684273719787598, "step": 20 }, { "epoch": 0.007749434937035841, "grad_norm": 5.633530616760254, "learning_rate": 4e-05, "loss": 5.696666240692139, "step": 21 }, { "epoch": 0.008118455648323262, "grad_norm": 5.457925319671631, "learning_rate": 4.2e-05, "loss": 5.586109638214111, "step": 22 }, { "epoch": 0.008487476359610684, "grad_norm": 7.086246967315674, "learning_rate": 4.4000000000000006e-05, "loss": 5.715998649597168, "step": 23 }, { "epoch": 0.008856497070898105, "grad_norm": 6.937062740325928, "learning_rate": 4.600000000000001e-05, "loss": 5.574023246765137, "step": 24 }, { "epoch": 0.009225517782185525, "grad_norm": 6.063978672027588, "learning_rate": 4.8e-05, "loss": 5.798689365386963, "step": 25 }, { "epoch": 0.009594538493472946, "grad_norm": 5.2796807289123535, "learning_rate": 5e-05, "loss": 5.642449855804443, "step": 26 }, { "epoch": 0.009963559204760367, "grad_norm": 6.248167991638184, "learning_rate": 5.2000000000000004e-05, "loss": 5.889494895935059, "step": 27 }, { "epoch": 0.010332579916047788, "grad_norm": 5.35820198059082, "learning_rate": 5.4000000000000005e-05, "loss": 5.48094367980957, "step": 28 }, { "epoch": 0.010701600627335208, "grad_norm": 5.648565292358398, "learning_rate": 5.6000000000000006e-05, "loss": 5.6747331619262695, "step": 29 }, { "epoch": 0.011070621338622631, "grad_norm": 4.546300888061523, "learning_rate": 5.8e-05, "loss": 5.303746223449707, "step": 30 }, { "epoch": 0.011439642049910052, "grad_norm": 4.514847278594971, "learning_rate": 6e-05, "loss": 5.3704681396484375, "step": 31 }, { "epoch": 0.011808662761197472, "grad_norm": 3.8123302459716797, "learning_rate": 6.2e-05, "loss": 5.293277263641357, "step": 32 }, { "epoch": 0.012177683472484893, "grad_norm": 6.512507438659668, "learning_rate": 6.400000000000001e-05, "loss": 5.347484588623047, "step": 33 }, { "epoch": 0.012546704183772314, "grad_norm": 6.077487468719482, "learning_rate": 6.6e-05, "loss": 5.459488868713379, "step": 34 }, { "epoch": 0.012915724895059735, "grad_norm": 9.170805931091309, "learning_rate": 6.800000000000001e-05, "loss": 5.849956512451172, "step": 35 }, { "epoch": 0.013284745606347157, "grad_norm": 6.619089126586914, "learning_rate": 7e-05, "loss": 5.7309160232543945, "step": 36 }, { "epoch": 0.013653766317634578, "grad_norm": 8.339101791381836, "learning_rate": 7.2e-05, "loss": 5.443577766418457, "step": 37 }, { "epoch": 0.014022787028921999, "grad_norm": 6.749815464019775, "learning_rate": 7.4e-05, "loss": 5.468483924865723, "step": 38 }, { "epoch": 0.01439180774020942, "grad_norm": 6.140146732330322, "learning_rate": 7.6e-05, "loss": 5.542171001434326, "step": 39 }, { "epoch": 0.01476082845149684, "grad_norm": 5.653714656829834, "learning_rate": 7.800000000000001e-05, "loss": 5.644989490509033, "step": 40 }, { "epoch": 0.01512984916278426, "grad_norm": 5.6543474197387695, "learning_rate": 8e-05, "loss": 5.580202102661133, "step": 41 }, { "epoch": 0.015498869874071682, "grad_norm": 6.0401506423950195, "learning_rate": 8.2e-05, "loss": 6.081020832061768, "step": 42 }, { "epoch": 0.015867890585359104, "grad_norm": 4.91713809967041, "learning_rate": 8.4e-05, "loss": 5.453029155731201, "step": 43 }, { "epoch": 0.016236911296646523, "grad_norm": 4.726991653442383, "learning_rate": 8.6e-05, "loss": 5.531949520111084, "step": 44 }, { "epoch": 0.016605932007933945, "grad_norm": 7.298604488372803, "learning_rate": 8.800000000000001e-05, "loss": 5.451730728149414, "step": 45 }, { "epoch": 0.016974952719221368, "grad_norm": 5.089509963989258, "learning_rate": 9e-05, "loss": 5.463062763214111, "step": 46 }, { "epoch": 0.017343973430508787, "grad_norm": 5.357382774353027, "learning_rate": 9.200000000000001e-05, "loss": 5.308757781982422, "step": 47 }, { "epoch": 0.01771299414179621, "grad_norm": 5.978388786315918, "learning_rate": 9.4e-05, "loss": 5.5749430656433105, "step": 48 }, { "epoch": 0.01808201485308363, "grad_norm": 5.822226047515869, "learning_rate": 9.6e-05, "loss": 5.463890075683594, "step": 49 }, { "epoch": 0.01845103556437105, "grad_norm": 6.441373348236084, "learning_rate": 9.8e-05, "loss": 5.169106483459473, "step": 50 }, { "epoch": 0.01882005627565847, "grad_norm": 4.273548603057861, "learning_rate": 0.0001, "loss": 5.433943748474121, "step": 51 }, { "epoch": 0.019189076986945892, "grad_norm": 6.303419589996338, "learning_rate": 0.0001, "loss": 5.323245048522949, "step": 52 }, { "epoch": 0.019558097698233315, "grad_norm": 6.22584342956543, "learning_rate": 0.0001, "loss": 5.212267875671387, "step": 53 }, { "epoch": 0.019927118409520734, "grad_norm": 6.891298770904541, "learning_rate": 0.0001, "loss": 5.431396484375, "step": 54 }, { "epoch": 0.020296139120808156, "grad_norm": 7.352054595947266, "learning_rate": 0.0001, "loss": 5.629201889038086, "step": 55 }, { "epoch": 0.020665159832095575, "grad_norm": 4.706995487213135, "learning_rate": 0.0001, "loss": 5.119016647338867, "step": 56 }, { "epoch": 0.021034180543382998, "grad_norm": 6.846929550170898, "learning_rate": 0.0001, "loss": 5.710336685180664, "step": 57 }, { "epoch": 0.021403201254670417, "grad_norm": 5.81605339050293, "learning_rate": 0.0001, "loss": 5.506718158721924, "step": 58 }, { "epoch": 0.02177222196595784, "grad_norm": 5.613030910491943, "learning_rate": 0.0001, "loss": 5.485965728759766, "step": 59 }, { "epoch": 0.022141242677245262, "grad_norm": 4.727166652679443, "learning_rate": 0.0001, "loss": 5.300134181976318, "step": 60 }, { "epoch": 0.02251026338853268, "grad_norm": 6.181882381439209, "learning_rate": 0.0001, "loss": 5.327762126922607, "step": 61 }, { "epoch": 0.022879284099820103, "grad_norm": 4.523162364959717, "learning_rate": 0.0001, "loss": 5.624878406524658, "step": 62 }, { "epoch": 0.023248304811107522, "grad_norm": 5.7826828956604, "learning_rate": 0.0001, "loss": 5.246135711669922, "step": 63 }, { "epoch": 0.023617325522394945, "grad_norm": 3.8491437435150146, "learning_rate": 0.0001, "loss": 5.223575592041016, "step": 64 }, { "epoch": 0.023986346233682364, "grad_norm": 6.430501937866211, "learning_rate": 0.0001, "loss": 5.349424839019775, "step": 65 }, { "epoch": 0.024355366944969786, "grad_norm": 4.793713092803955, "learning_rate": 0.0001, "loss": 5.306467056274414, "step": 66 }, { "epoch": 0.02472438765625721, "grad_norm": 5.871583461761475, "learning_rate": 0.0001, "loss": 5.41292142868042, "step": 67 }, { "epoch": 0.025093408367544628, "grad_norm": 3.803302764892578, "learning_rate": 0.0001, "loss": 5.451375484466553, "step": 68 }, { "epoch": 0.02546242907883205, "grad_norm": 4.517539978027344, "learning_rate": 0.0001, "loss": 5.264144420623779, "step": 69 }, { "epoch": 0.02583144979011947, "grad_norm": 4.541337966918945, "learning_rate": 0.0001, "loss": 5.2370219230651855, "step": 70 }, { "epoch": 0.02620047050140689, "grad_norm": 7.2052812576293945, "learning_rate": 0.0001, "loss": 5.298012733459473, "step": 71 }, { "epoch": 0.026569491212694314, "grad_norm": 4.4917707443237305, "learning_rate": 0.0001, "loss": 5.396561145782471, "step": 72 }, { "epoch": 0.026938511923981733, "grad_norm": 6.187958240509033, "learning_rate": 0.0001, "loss": 5.577371597290039, "step": 73 }, { "epoch": 0.027307532635269156, "grad_norm": 5.212502956390381, "learning_rate": 0.0001, "loss": 5.232723712921143, "step": 74 }, { "epoch": 0.027676553346556575, "grad_norm": 4.552979946136475, "learning_rate": 0.0001, "loss": 5.3533616065979, "step": 75 }, { "epoch": 0.028045574057843997, "grad_norm": 5.144165515899658, "learning_rate": 0.0001, "loss": 5.281284809112549, "step": 76 }, { "epoch": 0.028414594769131416, "grad_norm": 4.348478317260742, "learning_rate": 0.0001, "loss": 5.15976619720459, "step": 77 }, { "epoch": 0.02878361548041884, "grad_norm": 3.5937416553497314, "learning_rate": 0.0001, "loss": 5.613420009613037, "step": 78 }, { "epoch": 0.02915263619170626, "grad_norm": 5.10581111907959, "learning_rate": 0.0001, "loss": 5.1962571144104, "step": 79 }, { "epoch": 0.02952165690299368, "grad_norm": 5.265993595123291, "learning_rate": 0.0001, "loss": 5.335290431976318, "step": 80 }, { "epoch": 0.029890677614281103, "grad_norm": 3.9692020416259766, "learning_rate": 0.0001, "loss": 5.45660924911499, "step": 81 }, { "epoch": 0.03025969832556852, "grad_norm": 3.537494421005249, "learning_rate": 0.0001, "loss": 5.203220844268799, "step": 82 }, { "epoch": 0.030628719036855944, "grad_norm": 3.4767425060272217, "learning_rate": 0.0001, "loss": 5.189380645751953, "step": 83 }, { "epoch": 0.030997739748143363, "grad_norm": 3.707651138305664, "learning_rate": 0.0001, "loss": 5.333460330963135, "step": 84 }, { "epoch": 0.031366760459430786, "grad_norm": 3.983680486679077, "learning_rate": 0.0001, "loss": 5.286036491394043, "step": 85 }, { "epoch": 0.03173578117071821, "grad_norm": 5.241593837738037, "learning_rate": 0.0001, "loss": 5.400127410888672, "step": 86 }, { "epoch": 0.03210480188200563, "grad_norm": 3.111454963684082, "learning_rate": 0.0001, "loss": 5.3072919845581055, "step": 87 }, { "epoch": 0.032473822593293046, "grad_norm": 3.6329469680786133, "learning_rate": 0.0001, "loss": 5.42648458480835, "step": 88 }, { "epoch": 0.03284284330458047, "grad_norm": 6.178943634033203, "learning_rate": 0.0001, "loss": 5.166748046875, "step": 89 }, { "epoch": 0.03321186401586789, "grad_norm": 4.884125232696533, "learning_rate": 0.0001, "loss": 5.268251895904541, "step": 90 }, { "epoch": 0.033580884727155313, "grad_norm": 4.925684452056885, "learning_rate": 0.0001, "loss": 5.187631130218506, "step": 91 }, { "epoch": 0.033949905438442736, "grad_norm": 5.778391361236572, "learning_rate": 0.0001, "loss": 4.894795894622803, "step": 92 }, { "epoch": 0.03431892614973015, "grad_norm": 4.655902862548828, "learning_rate": 0.0001, "loss": 5.2550225257873535, "step": 93 }, { "epoch": 0.034687946861017574, "grad_norm": 5.219376564025879, "learning_rate": 0.0001, "loss": 5.362205505371094, "step": 94 }, { "epoch": 0.035056967572304996, "grad_norm": 3.6282198429107666, "learning_rate": 0.0001, "loss": 5.384284019470215, "step": 95 }, { "epoch": 0.03542598828359242, "grad_norm": 4.7581787109375, "learning_rate": 0.0001, "loss": 5.144690036773682, "step": 96 }, { "epoch": 0.035795008994879834, "grad_norm": 3.4947946071624756, "learning_rate": 0.0001, "loss": 5.272378444671631, "step": 97 }, { "epoch": 0.03616402970616726, "grad_norm": 4.599036693572998, "learning_rate": 0.0001, "loss": 4.959323406219482, "step": 98 }, { "epoch": 0.03653305041745468, "grad_norm": 4.1109418869018555, "learning_rate": 0.0001, "loss": 5.293513774871826, "step": 99 }, { "epoch": 0.0369020711287421, "grad_norm": 4.760090351104736, "learning_rate": 0.0001, "loss": 5.218464374542236, "step": 100 }, { "epoch": 0.037271091840029524, "grad_norm": 3.34501314163208, "learning_rate": 0.0001, "loss": 5.282923698425293, "step": 101 }, { "epoch": 0.03764011255131694, "grad_norm": 3.4620864391326904, "learning_rate": 0.0001, "loss": 5.078286170959473, "step": 102 }, { "epoch": 0.03800913326260436, "grad_norm": 4.38137674331665, "learning_rate": 0.0001, "loss": 5.183228969573975, "step": 103 }, { "epoch": 0.038378153973891785, "grad_norm": 4.770188331604004, "learning_rate": 0.0001, "loss": 5.21339750289917, "step": 104 }, { "epoch": 0.03874717468517921, "grad_norm": 4.081193923950195, "learning_rate": 0.0001, "loss": 5.423886299133301, "step": 105 }, { "epoch": 0.03911619539646663, "grad_norm": 4.0991010665893555, "learning_rate": 0.0001, "loss": 5.067556381225586, "step": 106 }, { "epoch": 0.039485216107754045, "grad_norm": 4.759697437286377, "learning_rate": 0.0001, "loss": 5.147449016571045, "step": 107 }, { "epoch": 0.03985423681904147, "grad_norm": 4.935500621795654, "learning_rate": 0.0001, "loss": 5.575446128845215, "step": 108 }, { "epoch": 0.04022325753032889, "grad_norm": 4.266407012939453, "learning_rate": 0.0001, "loss": 5.134807586669922, "step": 109 }, { "epoch": 0.04059227824161631, "grad_norm": 3.929640769958496, "learning_rate": 0.0001, "loss": 5.2398786544799805, "step": 110 }, { "epoch": 0.04096129895290373, "grad_norm": 4.449552059173584, "learning_rate": 0.0001, "loss": 5.156042575836182, "step": 111 }, { "epoch": 0.04133031966419115, "grad_norm": 3.860306739807129, "learning_rate": 0.0001, "loss": 4.952601432800293, "step": 112 }, { "epoch": 0.04169934037547857, "grad_norm": 4.714514255523682, "learning_rate": 0.0001, "loss": 5.524988174438477, "step": 113 }, { "epoch": 0.042068361086765996, "grad_norm": 4.664876461029053, "learning_rate": 0.0001, "loss": 4.995589256286621, "step": 114 }, { "epoch": 0.04243738179805342, "grad_norm": 4.218993663787842, "learning_rate": 0.0001, "loss": 5.228067874908447, "step": 115 }, { "epoch": 0.042806402509340834, "grad_norm": 4.040130138397217, "learning_rate": 0.0001, "loss": 5.018726348876953, "step": 116 }, { "epoch": 0.043175423220628256, "grad_norm": 4.016865253448486, "learning_rate": 0.0001, "loss": 5.552411079406738, "step": 117 }, { "epoch": 0.04354444393191568, "grad_norm": 4.608650207519531, "learning_rate": 0.0001, "loss": 5.315113067626953, "step": 118 }, { "epoch": 0.0439134646432031, "grad_norm": 4.381293296813965, "learning_rate": 0.0001, "loss": 5.192525386810303, "step": 119 }, { "epoch": 0.044282485354490524, "grad_norm": 3.6763710975646973, "learning_rate": 0.0001, "loss": 5.22324275970459, "step": 120 }, { "epoch": 0.04465150606577794, "grad_norm": 4.947065830230713, "learning_rate": 0.0001, "loss": 5.236631870269775, "step": 121 }, { "epoch": 0.04502052677706536, "grad_norm": 3.920776844024658, "learning_rate": 0.0001, "loss": 5.249703407287598, "step": 122 }, { "epoch": 0.045389547488352784, "grad_norm": 2.573857545852661, "learning_rate": 0.0001, "loss": 4.9618330001831055, "step": 123 }, { "epoch": 0.04575856819964021, "grad_norm": 3.178628921508789, "learning_rate": 0.0001, "loss": 5.127320766448975, "step": 124 }, { "epoch": 0.04612758891092763, "grad_norm": 3.242992877960205, "learning_rate": 0.0001, "loss": 5.084701061248779, "step": 125 }, { "epoch": 0.046496609622215045, "grad_norm": 3.685932159423828, "learning_rate": 0.0001, "loss": 5.189405918121338, "step": 126 }, { "epoch": 0.04686563033350247, "grad_norm": 2.7726705074310303, "learning_rate": 0.0001, "loss": 5.2822041511535645, "step": 127 }, { "epoch": 0.04723465104478989, "grad_norm": 3.1517961025238037, "learning_rate": 0.0001, "loss": 5.11806583404541, "step": 128 }, { "epoch": 0.04760367175607731, "grad_norm": 3.016364097595215, "learning_rate": 0.0001, "loss": 5.419538974761963, "step": 129 }, { "epoch": 0.04797269246736473, "grad_norm": 3.950747489929199, "learning_rate": 0.0001, "loss": 5.245120525360107, "step": 130 }, { "epoch": 0.04834171317865215, "grad_norm": 3.8114898204803467, "learning_rate": 0.0001, "loss": 4.765509128570557, "step": 131 }, { "epoch": 0.04871073388993957, "grad_norm": 3.9141814708709717, "learning_rate": 0.0001, "loss": 5.1021881103515625, "step": 132 }, { "epoch": 0.049079754601226995, "grad_norm": 4.04232931137085, "learning_rate": 0.0001, "loss": 4.959789276123047, "step": 133 }, { "epoch": 0.04944877531251442, "grad_norm": 2.6271684169769287, "learning_rate": 0.0001, "loss": 5.169219493865967, "step": 134 }, { "epoch": 0.04981779602380183, "grad_norm": 3.3536503314971924, "learning_rate": 0.0001, "loss": 5.00529146194458, "step": 135 }, { "epoch": 0.050186816735089256, "grad_norm": 4.068567276000977, "learning_rate": 0.0001, "loss": 4.860372543334961, "step": 136 }, { "epoch": 0.05055583744637668, "grad_norm": 4.450196743011475, "learning_rate": 0.0001, "loss": 5.2418718338012695, "step": 137 }, { "epoch": 0.0509248581576641, "grad_norm": 3.7781050205230713, "learning_rate": 0.0001, "loss": 5.009589672088623, "step": 138 }, { "epoch": 0.05129387886895152, "grad_norm": 4.003530979156494, "learning_rate": 0.0001, "loss": 5.156592845916748, "step": 139 }, { "epoch": 0.05166289958023894, "grad_norm": 4.052554130554199, "learning_rate": 0.0001, "loss": 5.234455585479736, "step": 140 }, { "epoch": 0.05203192029152636, "grad_norm": 3.1616437435150146, "learning_rate": 0.0001, "loss": 4.669750213623047, "step": 141 }, { "epoch": 0.05240094100281378, "grad_norm": 3.804659366607666, "learning_rate": 0.0001, "loss": 5.179162979125977, "step": 142 }, { "epoch": 0.052769961714101206, "grad_norm": 3.394329071044922, "learning_rate": 0.0001, "loss": 5.047037601470947, "step": 143 }, { "epoch": 0.05313898242538863, "grad_norm": 3.5721750259399414, "learning_rate": 0.0001, "loss": 5.083056926727295, "step": 144 }, { "epoch": 0.053508003136676044, "grad_norm": 3.9708645343780518, "learning_rate": 0.0001, "loss": 5.124134063720703, "step": 145 }, { "epoch": 0.053877023847963466, "grad_norm": 3.5934338569641113, "learning_rate": 0.0001, "loss": 5.215319633483887, "step": 146 }, { "epoch": 0.05424604455925089, "grad_norm": 4.272321701049805, "learning_rate": 0.0001, "loss": 5.307074546813965, "step": 147 }, { "epoch": 0.05461506527053831, "grad_norm": 3.845202922821045, "learning_rate": 0.0001, "loss": 4.817156791687012, "step": 148 }, { "epoch": 0.05498408598182573, "grad_norm": 5.178013801574707, "learning_rate": 0.0001, "loss": 5.578018665313721, "step": 149 }, { "epoch": 0.05535310669311315, "grad_norm": 4.21903133392334, "learning_rate": 0.0001, "loss": 5.05971622467041, "step": 150 }, { "epoch": 0.05572212740440057, "grad_norm": 3.9957923889160156, "learning_rate": 0.0001, "loss": 5.0507731437683105, "step": 151 }, { "epoch": 0.056091148115687994, "grad_norm": 4.366008281707764, "learning_rate": 0.0001, "loss": 4.8964433670043945, "step": 152 }, { "epoch": 0.05646016882697542, "grad_norm": 5.036316871643066, "learning_rate": 0.0001, "loss": 4.990871429443359, "step": 153 }, { "epoch": 0.05682918953826283, "grad_norm": 3.921596050262451, "learning_rate": 0.0001, "loss": 5.315304756164551, "step": 154 }, { "epoch": 0.057198210249550255, "grad_norm": 3.0889859199523926, "learning_rate": 0.0001, "loss": 5.343647003173828, "step": 155 }, { "epoch": 0.05756723096083768, "grad_norm": 3.3512563705444336, "learning_rate": 0.0001, "loss": 5.151203155517578, "step": 156 }, { "epoch": 0.0579362516721251, "grad_norm": 3.630702018737793, "learning_rate": 0.0001, "loss": 4.823637008666992, "step": 157 }, { "epoch": 0.05830527238341252, "grad_norm": 3.516183853149414, "learning_rate": 0.0001, "loss": 4.9256272315979, "step": 158 }, { "epoch": 0.05867429309469994, "grad_norm": 3.8045616149902344, "learning_rate": 0.0001, "loss": 4.9611992835998535, "step": 159 }, { "epoch": 0.05904331380598736, "grad_norm": 4.987512111663818, "learning_rate": 0.0001, "loss": 5.136240005493164, "step": 160 }, { "epoch": 0.05941233451727478, "grad_norm": 3.295520305633545, "learning_rate": 0.0001, "loss": 4.939165115356445, "step": 161 }, { "epoch": 0.059781355228562205, "grad_norm": 3.4286487102508545, "learning_rate": 0.0001, "loss": 5.1253533363342285, "step": 162 }, { "epoch": 0.06015037593984962, "grad_norm": 3.3474066257476807, "learning_rate": 0.0001, "loss": 4.942944526672363, "step": 163 }, { "epoch": 0.06051939665113704, "grad_norm": 3.4838764667510986, "learning_rate": 0.0001, "loss": 4.937647819519043, "step": 164 }, { "epoch": 0.060888417362424466, "grad_norm": 3.3820011615753174, "learning_rate": 0.0001, "loss": 4.728400230407715, "step": 165 }, { "epoch": 0.06125743807371189, "grad_norm": 3.039099931716919, "learning_rate": 0.0001, "loss": 5.29592227935791, "step": 166 }, { "epoch": 0.06162645878499931, "grad_norm": 3.884432554244995, "learning_rate": 0.0001, "loss": 5.097470283508301, "step": 167 }, { "epoch": 0.061995479496286726, "grad_norm": 3.927171468734741, "learning_rate": 0.0001, "loss": 5.287883758544922, "step": 168 }, { "epoch": 0.06236450020757415, "grad_norm": 3.2847440242767334, "learning_rate": 0.0001, "loss": 4.842876434326172, "step": 169 }, { "epoch": 0.06273352091886157, "grad_norm": 3.7847704887390137, "learning_rate": 0.0001, "loss": 5.119341850280762, "step": 170 }, { "epoch": 0.06310254163014899, "grad_norm": 3.4630985260009766, "learning_rate": 0.0001, "loss": 5.680130958557129, "step": 171 }, { "epoch": 0.06347156234143642, "grad_norm": 2.9577441215515137, "learning_rate": 0.0001, "loss": 5.309061050415039, "step": 172 }, { "epoch": 0.06384058305272383, "grad_norm": 3.408936023712158, "learning_rate": 0.0001, "loss": 4.966146945953369, "step": 173 }, { "epoch": 0.06420960376401126, "grad_norm": 2.8536202907562256, "learning_rate": 0.0001, "loss": 5.047423362731934, "step": 174 }, { "epoch": 0.06457862447529868, "grad_norm": 2.683049201965332, "learning_rate": 0.0001, "loss": 5.021755218505859, "step": 175 }, { "epoch": 0.06494764518658609, "grad_norm": 8.568255424499512, "learning_rate": 0.0001, "loss": 5.02105712890625, "step": 176 }, { "epoch": 0.06531666589787352, "grad_norm": 3.26704478263855, "learning_rate": 0.0001, "loss": 4.734854221343994, "step": 177 }, { "epoch": 0.06568568660916094, "grad_norm": 3.1402342319488525, "learning_rate": 0.0001, "loss": 5.746152400970459, "step": 178 }, { "epoch": 0.06605470732044837, "grad_norm": 3.1304996013641357, "learning_rate": 0.0001, "loss": 5.069252014160156, "step": 179 }, { "epoch": 0.06642372803173578, "grad_norm": 3.1906208992004395, "learning_rate": 0.0001, "loss": 5.060096263885498, "step": 180 }, { "epoch": 0.0667927487430232, "grad_norm": 3.0051543712615967, "learning_rate": 0.0001, "loss": 5.333315849304199, "step": 181 }, { "epoch": 0.06716176945431063, "grad_norm": 2.9292314052581787, "learning_rate": 0.0001, "loss": 4.98300838470459, "step": 182 }, { "epoch": 0.06753079016559804, "grad_norm": 3.8607776165008545, "learning_rate": 0.0001, "loss": 5.22333288192749, "step": 183 }, { "epoch": 0.06789981087688547, "grad_norm": 5.044845104217529, "learning_rate": 0.0001, "loss": 4.9512128829956055, "step": 184 }, { "epoch": 0.06826883158817289, "grad_norm": 3.930791139602661, "learning_rate": 0.0001, "loss": 4.994528770446777, "step": 185 }, { "epoch": 0.0686378522994603, "grad_norm": 4.407391548156738, "learning_rate": 0.0001, "loss": 5.1336493492126465, "step": 186 }, { "epoch": 0.06900687301074773, "grad_norm": 3.2872698307037354, "learning_rate": 0.0001, "loss": 4.903454303741455, "step": 187 }, { "epoch": 0.06937589372203515, "grad_norm": 2.9913930892944336, "learning_rate": 0.0001, "loss": 4.907477378845215, "step": 188 }, { "epoch": 0.06974491443332256, "grad_norm": 3.690415382385254, "learning_rate": 0.0001, "loss": 5.368701457977295, "step": 189 }, { "epoch": 0.07011393514460999, "grad_norm": 3.2525527477264404, "learning_rate": 0.0001, "loss": 4.97088623046875, "step": 190 }, { "epoch": 0.07048295585589741, "grad_norm": 3.62608003616333, "learning_rate": 0.0001, "loss": 5.350979328155518, "step": 191 }, { "epoch": 0.07085197656718484, "grad_norm": 2.343998908996582, "learning_rate": 0.0001, "loss": 4.678924083709717, "step": 192 }, { "epoch": 0.07122099727847225, "grad_norm": 2.5523736476898193, "learning_rate": 0.0001, "loss": 5.1422343254089355, "step": 193 }, { "epoch": 0.07159001798975967, "grad_norm": 3.007969856262207, "learning_rate": 0.0001, "loss": 5.103295803070068, "step": 194 }, { "epoch": 0.0719590387010471, "grad_norm": 2.446394205093384, "learning_rate": 0.0001, "loss": 4.8293867111206055, "step": 195 }, { "epoch": 0.07232805941233451, "grad_norm": 2.6612651348114014, "learning_rate": 0.0001, "loss": 5.177353858947754, "step": 196 }, { "epoch": 0.07269708012362194, "grad_norm": 2.940788507461548, "learning_rate": 0.0001, "loss": 4.852959156036377, "step": 197 }, { "epoch": 0.07306610083490936, "grad_norm": 3.0821218490600586, "learning_rate": 0.0001, "loss": 4.744450569152832, "step": 198 }, { "epoch": 0.07343512154619677, "grad_norm": 3.23681640625, "learning_rate": 0.0001, "loss": 5.011826515197754, "step": 199 }, { "epoch": 0.0738041422574842, "grad_norm": 2.6175684928894043, "learning_rate": 0.0001, "loss": 5.107472896575928, "step": 200 }, { "epoch": 0.07417316296877162, "grad_norm": 3.0995991230010986, "learning_rate": 0.0001, "loss": 4.885207653045654, "step": 201 }, { "epoch": 0.07454218368005905, "grad_norm": 2.8786182403564453, "learning_rate": 0.0001, "loss": 4.845186233520508, "step": 202 }, { "epoch": 0.07491120439134646, "grad_norm": 3.120236396789551, "learning_rate": 0.0001, "loss": 4.899354457855225, "step": 203 }, { "epoch": 0.07528022510263388, "grad_norm": 2.4766738414764404, "learning_rate": 0.0001, "loss": 5.172772407531738, "step": 204 }, { "epoch": 0.07564924581392131, "grad_norm": 2.8088064193725586, "learning_rate": 0.0001, "loss": 4.881832599639893, "step": 205 }, { "epoch": 0.07601826652520872, "grad_norm": 3.229881763458252, "learning_rate": 0.0001, "loss": 5.073505401611328, "step": 206 }, { "epoch": 0.07638728723649615, "grad_norm": 3.5004501342773438, "learning_rate": 0.0001, "loss": 4.852757930755615, "step": 207 }, { "epoch": 0.07675630794778357, "grad_norm": 3.0608553886413574, "learning_rate": 0.0001, "loss": 4.712113380432129, "step": 208 }, { "epoch": 0.07712532865907099, "grad_norm": 3.229390859603882, "learning_rate": 0.0001, "loss": 4.8641276359558105, "step": 209 }, { "epoch": 0.07749434937035841, "grad_norm": 2.9614827632904053, "learning_rate": 0.0001, "loss": 5.257576942443848, "step": 210 }, { "epoch": 0.07786337008164583, "grad_norm": 2.8287434577941895, "learning_rate": 0.0001, "loss": 4.992201805114746, "step": 211 }, { "epoch": 0.07823239079293326, "grad_norm": 2.8483452796936035, "learning_rate": 0.0001, "loss": 4.637844562530518, "step": 212 }, { "epoch": 0.07860141150422068, "grad_norm": 2.9546141624450684, "learning_rate": 0.0001, "loss": 4.944921016693115, "step": 213 }, { "epoch": 0.07897043221550809, "grad_norm": 4.089045524597168, "learning_rate": 0.0001, "loss": 5.075927734375, "step": 214 }, { "epoch": 0.07933945292679552, "grad_norm": 2.8543574810028076, "learning_rate": 0.0001, "loss": 4.585363388061523, "step": 215 }, { "epoch": 0.07970847363808294, "grad_norm": 2.697917938232422, "learning_rate": 0.0001, "loss": 5.132508277893066, "step": 216 }, { "epoch": 0.08007749434937037, "grad_norm": 2.9316647052764893, "learning_rate": 0.0001, "loss": 5.069299221038818, "step": 217 }, { "epoch": 0.08044651506065778, "grad_norm": 2.8044543266296387, "learning_rate": 0.0001, "loss": 4.850956439971924, "step": 218 }, { "epoch": 0.0808155357719452, "grad_norm": 2.7793681621551514, "learning_rate": 0.0001, "loss": 5.32306432723999, "step": 219 }, { "epoch": 0.08118455648323263, "grad_norm": 2.586214542388916, "learning_rate": 0.0001, "loss": 4.600836277008057, "step": 220 }, { "epoch": 0.08155357719452004, "grad_norm": 3.0345284938812256, "learning_rate": 0.0001, "loss": 4.788100719451904, "step": 221 }, { "epoch": 0.08192259790580746, "grad_norm": 2.685389518737793, "learning_rate": 0.0001, "loss": 4.780126571655273, "step": 222 }, { "epoch": 0.08229161861709489, "grad_norm": 2.843975305557251, "learning_rate": 0.0001, "loss": 4.858603000640869, "step": 223 }, { "epoch": 0.0826606393283823, "grad_norm": 3.5730929374694824, "learning_rate": 0.0001, "loss": 4.825096130371094, "step": 224 }, { "epoch": 0.08302966003966973, "grad_norm": 3.2879269123077393, "learning_rate": 0.0001, "loss": 4.757035255432129, "step": 225 }, { "epoch": 0.08339868075095715, "grad_norm": 3.235781669616699, "learning_rate": 0.0001, "loss": 5.257933139801025, "step": 226 }, { "epoch": 0.08376770146224456, "grad_norm": 2.9078941345214844, "learning_rate": 0.0001, "loss": 4.914604663848877, "step": 227 }, { "epoch": 0.08413672217353199, "grad_norm": 2.557096004486084, "learning_rate": 0.0001, "loss": 4.683157920837402, "step": 228 }, { "epoch": 0.08450574288481941, "grad_norm": 3.4715662002563477, "learning_rate": 0.0001, "loss": 4.910725116729736, "step": 229 }, { "epoch": 0.08487476359610684, "grad_norm": 3.880197048187256, "learning_rate": 0.0001, "loss": 4.976219177246094, "step": 230 }, { "epoch": 0.08524378430739425, "grad_norm": 3.1803884506225586, "learning_rate": 0.0001, "loss": 4.871642112731934, "step": 231 }, { "epoch": 0.08561280501868167, "grad_norm": 5.288760662078857, "learning_rate": 0.0001, "loss": 4.9952287673950195, "step": 232 }, { "epoch": 0.0859818257299691, "grad_norm": 3.4574294090270996, "learning_rate": 0.0001, "loss": 5.026848793029785, "step": 233 }, { "epoch": 0.08635084644125651, "grad_norm": 2.983919382095337, "learning_rate": 0.0001, "loss": 4.823044776916504, "step": 234 }, { "epoch": 0.08671986715254394, "grad_norm": 2.493537664413452, "learning_rate": 0.0001, "loss": 4.764354705810547, "step": 235 }, { "epoch": 0.08708888786383136, "grad_norm": 3.7901830673217773, "learning_rate": 0.0001, "loss": 5.213089942932129, "step": 236 }, { "epoch": 0.08745790857511877, "grad_norm": 2.74006986618042, "learning_rate": 0.0001, "loss": 4.810093879699707, "step": 237 }, { "epoch": 0.0878269292864062, "grad_norm": 2.547126531600952, "learning_rate": 0.0001, "loss": 4.611993312835693, "step": 238 }, { "epoch": 0.08819594999769362, "grad_norm": 2.8028151988983154, "learning_rate": 0.0001, "loss": 4.800611972808838, "step": 239 }, { "epoch": 0.08856497070898105, "grad_norm": 2.3688666820526123, "learning_rate": 0.0001, "loss": 4.928980827331543, "step": 240 }, { "epoch": 0.08893399142026846, "grad_norm": 2.43617844581604, "learning_rate": 0.0001, "loss": 5.174886703491211, "step": 241 }, { "epoch": 0.08930301213155588, "grad_norm": 4.133114814758301, "learning_rate": 0.0001, "loss": 4.833251476287842, "step": 242 }, { "epoch": 0.08967203284284331, "grad_norm": 2.575796365737915, "learning_rate": 0.0001, "loss": 5.019968032836914, "step": 243 }, { "epoch": 0.09004105355413072, "grad_norm": 3.121090888977051, "learning_rate": 0.0001, "loss": 4.538406848907471, "step": 244 }, { "epoch": 0.09041007426541815, "grad_norm": 3.4044008255004883, "learning_rate": 0.0001, "loss": 5.054136753082275, "step": 245 }, { "epoch": 0.09077909497670557, "grad_norm": 2.7993178367614746, "learning_rate": 0.0001, "loss": 4.802661418914795, "step": 246 }, { "epoch": 0.09114811568799298, "grad_norm": 2.870140552520752, "learning_rate": 0.0001, "loss": 5.112711429595947, "step": 247 }, { "epoch": 0.09151713639928041, "grad_norm": 3.2912027835845947, "learning_rate": 0.0001, "loss": 4.869266510009766, "step": 248 }, { "epoch": 0.09188615711056783, "grad_norm": 2.9518210887908936, "learning_rate": 0.0001, "loss": 4.794067859649658, "step": 249 }, { "epoch": 0.09225517782185526, "grad_norm": 2.7578539848327637, "learning_rate": 0.0001, "loss": 4.8072967529296875, "step": 250 }, { "epoch": 0.09262419853314267, "grad_norm": 2.4775185585021973, "learning_rate": 0.0001, "loss": 4.741370677947998, "step": 251 }, { "epoch": 0.09299321924443009, "grad_norm": 2.850309133529663, "learning_rate": 0.0001, "loss": 4.818919658660889, "step": 252 }, { "epoch": 0.09336223995571752, "grad_norm": 3.1108920574188232, "learning_rate": 0.0001, "loss": 5.032450199127197, "step": 253 }, { "epoch": 0.09373126066700493, "grad_norm": 2.813636541366577, "learning_rate": 0.0001, "loss": 4.874199867248535, "step": 254 }, { "epoch": 0.09410028137829235, "grad_norm": 2.8555026054382324, "learning_rate": 0.0001, "loss": 4.985494136810303, "step": 255 }, { "epoch": 0.09446930208957978, "grad_norm": 2.5190069675445557, "learning_rate": 0.0001, "loss": 4.894311904907227, "step": 256 }, { "epoch": 0.0948383228008672, "grad_norm": 2.6791582107543945, "learning_rate": 0.0001, "loss": 4.945681095123291, "step": 257 }, { "epoch": 0.09520734351215462, "grad_norm": 2.9527032375335693, "learning_rate": 0.0001, "loss": 4.8298468589782715, "step": 258 }, { "epoch": 0.09557636422344204, "grad_norm": 2.7807869911193848, "learning_rate": 0.0001, "loss": 4.615987777709961, "step": 259 }, { "epoch": 0.09594538493472946, "grad_norm": 2.682555675506592, "learning_rate": 0.0001, "loss": 4.74656867980957, "step": 260 }, { "epoch": 0.09631440564601688, "grad_norm": 3.3544092178344727, "learning_rate": 0.0001, "loss": 4.863458633422852, "step": 261 }, { "epoch": 0.0966834263573043, "grad_norm": 2.737848997116089, "learning_rate": 0.0001, "loss": 4.7740397453308105, "step": 262 }, { "epoch": 0.09705244706859173, "grad_norm": 2.7295894622802734, "learning_rate": 0.0001, "loss": 4.383846759796143, "step": 263 }, { "epoch": 0.09742146777987915, "grad_norm": 2.332279682159424, "learning_rate": 0.0001, "loss": 4.982947826385498, "step": 264 }, { "epoch": 0.09779048849116656, "grad_norm": 2.687039613723755, "learning_rate": 0.0001, "loss": 4.721158981323242, "step": 265 }, { "epoch": 0.09815950920245399, "grad_norm": 2.505675792694092, "learning_rate": 0.0001, "loss": 4.891917705535889, "step": 266 }, { "epoch": 0.0985285299137414, "grad_norm": 2.8479607105255127, "learning_rate": 0.0001, "loss": 4.618453502655029, "step": 267 }, { "epoch": 0.09889755062502883, "grad_norm": 2.6204135417938232, "learning_rate": 0.0001, "loss": 4.758787631988525, "step": 268 }, { "epoch": 0.09926657133631625, "grad_norm": 3.021299362182617, "learning_rate": 0.0001, "loss": 4.7189507484436035, "step": 269 }, { "epoch": 0.09963559204760367, "grad_norm": 2.4360415935516357, "learning_rate": 0.0001, "loss": 4.794710636138916, "step": 270 }, { "epoch": 0.1000046127588911, "grad_norm": 2.366312026977539, "learning_rate": 0.0001, "loss": 4.708616256713867, "step": 271 }, { "epoch": 0.10037363347017851, "grad_norm": 2.5458579063415527, "learning_rate": 0.0001, "loss": 4.5073394775390625, "step": 272 }, { "epoch": 0.10074265418146594, "grad_norm": 2.9471662044525146, "learning_rate": 0.0001, "loss": 4.495577335357666, "step": 273 }, { "epoch": 0.10111167489275336, "grad_norm": 3.0150680541992188, "learning_rate": 0.0001, "loss": 4.821500301361084, "step": 274 }, { "epoch": 0.10148069560404077, "grad_norm": 2.578234910964966, "learning_rate": 0.0001, "loss": 4.77611780166626, "step": 275 }, { "epoch": 0.1018497163153282, "grad_norm": 2.579213857650757, "learning_rate": 0.0001, "loss": 4.857175350189209, "step": 276 }, { "epoch": 0.10221873702661562, "grad_norm": 2.65431547164917, "learning_rate": 0.0001, "loss": 4.517264366149902, "step": 277 }, { "epoch": 0.10258775773790305, "grad_norm": 3.1191048622131348, "learning_rate": 0.0001, "loss": 5.008939266204834, "step": 278 }, { "epoch": 0.10295677844919046, "grad_norm": 2.8840596675872803, "learning_rate": 0.0001, "loss": 4.874368190765381, "step": 279 }, { "epoch": 0.10332579916047788, "grad_norm": 2.867748975753784, "learning_rate": 0.0001, "loss": 4.8059797286987305, "step": 280 }, { "epoch": 0.1036948198717653, "grad_norm": 3.3783462047576904, "learning_rate": 0.0001, "loss": 4.704605579376221, "step": 281 }, { "epoch": 0.10406384058305272, "grad_norm": 2.759956121444702, "learning_rate": 0.0001, "loss": 4.987379550933838, "step": 282 }, { "epoch": 0.10443286129434015, "grad_norm": 2.6871654987335205, "learning_rate": 0.0001, "loss": 4.723685264587402, "step": 283 }, { "epoch": 0.10480188200562757, "grad_norm": 2.8111624717712402, "learning_rate": 0.0001, "loss": 4.588558673858643, "step": 284 }, { "epoch": 0.10517090271691498, "grad_norm": 2.244652509689331, "learning_rate": 0.0001, "loss": 4.661073207855225, "step": 285 }, { "epoch": 0.10553992342820241, "grad_norm": 2.8253273963928223, "learning_rate": 0.0001, "loss": 4.67581033706665, "step": 286 }, { "epoch": 0.10590894413948983, "grad_norm": 2.362382650375366, "learning_rate": 0.0001, "loss": 4.8821916580200195, "step": 287 }, { "epoch": 0.10627796485077726, "grad_norm": 2.725433111190796, "learning_rate": 0.0001, "loss": 4.793010234832764, "step": 288 }, { "epoch": 0.10664698556206467, "grad_norm": 2.443126678466797, "learning_rate": 0.0001, "loss": 4.6530046463012695, "step": 289 }, { "epoch": 0.10701600627335209, "grad_norm": 2.5819406509399414, "learning_rate": 0.0001, "loss": 5.03138542175293, "step": 290 }, { "epoch": 0.10738502698463952, "grad_norm": 2.345172166824341, "learning_rate": 0.0001, "loss": 4.908970832824707, "step": 291 }, { "epoch": 0.10775404769592693, "grad_norm": 3.7422969341278076, "learning_rate": 0.0001, "loss": 4.663451194763184, "step": 292 }, { "epoch": 0.10812306840721435, "grad_norm": 2.690535068511963, "learning_rate": 0.0001, "loss": 4.976020812988281, "step": 293 }, { "epoch": 0.10849208911850178, "grad_norm": 3.6050312519073486, "learning_rate": 0.0001, "loss": 4.173515319824219, "step": 294 }, { "epoch": 0.1088611098297892, "grad_norm": 2.6177196502685547, "learning_rate": 0.0001, "loss": 4.570525169372559, "step": 295 }, { "epoch": 0.10923013054107662, "grad_norm": 3.2142410278320312, "learning_rate": 0.0001, "loss": 5.103811740875244, "step": 296 }, { "epoch": 0.10959915125236404, "grad_norm": 4.097013473510742, "learning_rate": 0.0001, "loss": 4.417541027069092, "step": 297 }, { "epoch": 0.10996817196365145, "grad_norm": 2.6971688270568848, "learning_rate": 0.0001, "loss": 4.503973007202148, "step": 298 }, { "epoch": 0.11033719267493888, "grad_norm": 2.4972405433654785, "learning_rate": 0.0001, "loss": 4.668544769287109, "step": 299 }, { "epoch": 0.1107062133862263, "grad_norm": 3.145050287246704, "learning_rate": 0.0001, "loss": 4.453399658203125, "step": 300 }, { "epoch": 0.11107523409751373, "grad_norm": 2.5468809604644775, "learning_rate": 0.0001, "loss": 4.50095272064209, "step": 301 }, { "epoch": 0.11144425480880114, "grad_norm": 2.6690492630004883, "learning_rate": 0.0001, "loss": 4.6230950355529785, "step": 302 }, { "epoch": 0.11181327552008856, "grad_norm": 2.255617141723633, "learning_rate": 0.0001, "loss": 4.648458003997803, "step": 303 }, { "epoch": 0.11218229623137599, "grad_norm": 2.83013916015625, "learning_rate": 0.0001, "loss": 4.58681583404541, "step": 304 }, { "epoch": 0.1125513169426634, "grad_norm": 2.867711067199707, "learning_rate": 0.0001, "loss": 4.499338150024414, "step": 305 }, { "epoch": 0.11292033765395083, "grad_norm": 2.580369234085083, "learning_rate": 0.0001, "loss": 4.6251220703125, "step": 306 }, { "epoch": 0.11328935836523825, "grad_norm": 2.3685848712921143, "learning_rate": 0.0001, "loss": 4.643671989440918, "step": 307 }, { "epoch": 0.11365837907652566, "grad_norm": 3.4852209091186523, "learning_rate": 0.0001, "loss": 4.932046890258789, "step": 308 }, { "epoch": 0.1140273997878131, "grad_norm": 2.830315351486206, "learning_rate": 0.0001, "loss": 4.5891923904418945, "step": 309 }, { "epoch": 0.11439642049910051, "grad_norm": 3.2897260189056396, "learning_rate": 0.0001, "loss": 4.485490798950195, "step": 310 }, { "epoch": 0.11476544121038794, "grad_norm": 2.5602848529815674, "learning_rate": 0.0001, "loss": 4.726596832275391, "step": 311 }, { "epoch": 0.11513446192167535, "grad_norm": 2.8062198162078857, "learning_rate": 0.0001, "loss": 4.881383895874023, "step": 312 }, { "epoch": 0.11550348263296277, "grad_norm": 2.9454972743988037, "learning_rate": 0.0001, "loss": 4.755986213684082, "step": 313 }, { "epoch": 0.1158725033442502, "grad_norm": 2.8830761909484863, "learning_rate": 0.0001, "loss": 5.087071418762207, "step": 314 }, { "epoch": 0.11624152405553762, "grad_norm": 2.050590991973877, "learning_rate": 0.0001, "loss": 4.456117630004883, "step": 315 }, { "epoch": 0.11661054476682504, "grad_norm": 2.8796334266662598, "learning_rate": 0.0001, "loss": 4.842302322387695, "step": 316 }, { "epoch": 0.11697956547811246, "grad_norm": 2.1899056434631348, "learning_rate": 0.0001, "loss": 4.674966335296631, "step": 317 }, { "epoch": 0.11734858618939988, "grad_norm": 2.62785267829895, "learning_rate": 0.0001, "loss": 4.86789608001709, "step": 318 }, { "epoch": 0.1177176069006873, "grad_norm": 2.5556700229644775, "learning_rate": 0.0001, "loss": 4.4173264503479, "step": 319 }, { "epoch": 0.11808662761197472, "grad_norm": 3.35113525390625, "learning_rate": 0.0001, "loss": 4.507681369781494, "step": 320 }, { "epoch": 0.11845564832326215, "grad_norm": 3.0669710636138916, "learning_rate": 0.0001, "loss": 4.245882987976074, "step": 321 }, { "epoch": 0.11882466903454957, "grad_norm": 2.4834563732147217, "learning_rate": 0.0001, "loss": 4.907164573669434, "step": 322 }, { "epoch": 0.11919368974583698, "grad_norm": 2.5204923152923584, "learning_rate": 0.0001, "loss": 4.891724586486816, "step": 323 }, { "epoch": 0.11956271045712441, "grad_norm": 2.367497205734253, "learning_rate": 0.0001, "loss": 4.811779022216797, "step": 324 }, { "epoch": 0.11993173116841183, "grad_norm": 2.360835552215576, "learning_rate": 0.0001, "loss": 4.670647144317627, "step": 325 }, { "epoch": 0.12030075187969924, "grad_norm": 2.2293601036071777, "learning_rate": 0.0001, "loss": 4.589348793029785, "step": 326 }, { "epoch": 0.12066977259098667, "grad_norm": 2.154679775238037, "learning_rate": 0.0001, "loss": 4.731776237487793, "step": 327 }, { "epoch": 0.12103879330227409, "grad_norm": 2.914649486541748, "learning_rate": 0.0001, "loss": 4.931624889373779, "step": 328 }, { "epoch": 0.12140781401356152, "grad_norm": 2.734717845916748, "learning_rate": 0.0001, "loss": 4.975724697113037, "step": 329 }, { "epoch": 0.12177683472484893, "grad_norm": 2.389617443084717, "learning_rate": 0.0001, "loss": 4.796389102935791, "step": 330 }, { "epoch": 0.12214585543613635, "grad_norm": 1.979083776473999, "learning_rate": 0.0001, "loss": 4.383009433746338, "step": 331 }, { "epoch": 0.12251487614742378, "grad_norm": 2.6853559017181396, "learning_rate": 0.0001, "loss": 4.8341898918151855, "step": 332 }, { "epoch": 0.12288389685871119, "grad_norm": 2.5915274620056152, "learning_rate": 0.0001, "loss": 4.717322826385498, "step": 333 }, { "epoch": 0.12325291756999862, "grad_norm": 2.812741994857788, "learning_rate": 0.0001, "loss": 4.969723701477051, "step": 334 }, { "epoch": 0.12362193828128604, "grad_norm": 2.57147216796875, "learning_rate": 0.0001, "loss": 4.582840919494629, "step": 335 }, { "epoch": 0.12399095899257345, "grad_norm": 2.5397863388061523, "learning_rate": 0.0001, "loss": 4.691579341888428, "step": 336 }, { "epoch": 0.12435997970386088, "grad_norm": 2.942898988723755, "learning_rate": 0.0001, "loss": 4.7678728103637695, "step": 337 }, { "epoch": 0.1247290004151483, "grad_norm": 2.3598756790161133, "learning_rate": 0.0001, "loss": 4.659327030181885, "step": 338 }, { "epoch": 0.12509802112643573, "grad_norm": 2.3905725479125977, "learning_rate": 0.0001, "loss": 4.716471195220947, "step": 339 }, { "epoch": 0.12546704183772314, "grad_norm": 2.3143739700317383, "learning_rate": 0.0001, "loss": 4.569877624511719, "step": 340 }, { "epoch": 0.12583606254901056, "grad_norm": 2.9056665897369385, "learning_rate": 0.0001, "loss": 4.622424602508545, "step": 341 }, { "epoch": 0.12620508326029797, "grad_norm": 2.454422950744629, "learning_rate": 0.0001, "loss": 4.7714080810546875, "step": 342 }, { "epoch": 0.12657410397158542, "grad_norm": 2.410001516342163, "learning_rate": 0.0001, "loss": 4.591516494750977, "step": 343 }, { "epoch": 0.12694312468287283, "grad_norm": 4.945067405700684, "learning_rate": 0.0001, "loss": 4.671370983123779, "step": 344 }, { "epoch": 0.12731214539416025, "grad_norm": 2.9925453662872314, "learning_rate": 0.0001, "loss": 4.712247848510742, "step": 345 }, { "epoch": 0.12768116610544766, "grad_norm": 2.5714759826660156, "learning_rate": 0.0001, "loss": 4.530534744262695, "step": 346 }, { "epoch": 0.12805018681673508, "grad_norm": 2.7953736782073975, "learning_rate": 0.0001, "loss": 4.63610315322876, "step": 347 }, { "epoch": 0.12841920752802252, "grad_norm": 2.2027106285095215, "learning_rate": 0.0001, "loss": 4.753771781921387, "step": 348 }, { "epoch": 0.12878822823930994, "grad_norm": 2.802049398422241, "learning_rate": 0.0001, "loss": 4.764090061187744, "step": 349 }, { "epoch": 0.12915724895059735, "grad_norm": 2.5809788703918457, "learning_rate": 0.0001, "loss": 4.627032279968262, "step": 350 }, { "epoch": 0.12952626966188477, "grad_norm": 2.522502899169922, "learning_rate": 0.0001, "loss": 4.610208988189697, "step": 351 }, { "epoch": 0.12989529037317218, "grad_norm": 2.322376012802124, "learning_rate": 0.0001, "loss": 4.736202239990234, "step": 352 }, { "epoch": 0.13026431108445963, "grad_norm": 2.6052095890045166, "learning_rate": 0.0001, "loss": 4.679316997528076, "step": 353 }, { "epoch": 0.13063333179574704, "grad_norm": 2.5171854496002197, "learning_rate": 0.0001, "loss": 4.587606906890869, "step": 354 }, { "epoch": 0.13100235250703446, "grad_norm": 2.7908267974853516, "learning_rate": 0.0001, "loss": 4.818106651306152, "step": 355 }, { "epoch": 0.13137137321832187, "grad_norm": 2.3033812046051025, "learning_rate": 0.0001, "loss": 4.679779529571533, "step": 356 }, { "epoch": 0.1317403939296093, "grad_norm": 2.204928159713745, "learning_rate": 0.0001, "loss": 4.498159408569336, "step": 357 }, { "epoch": 0.13210941464089673, "grad_norm": 2.2332544326782227, "learning_rate": 0.0001, "loss": 4.56221866607666, "step": 358 }, { "epoch": 0.13247843535218415, "grad_norm": 2.151496648788452, "learning_rate": 0.0001, "loss": 4.715498924255371, "step": 359 }, { "epoch": 0.13284745606347156, "grad_norm": 2.1087806224823, "learning_rate": 0.0001, "loss": 4.749574184417725, "step": 360 }, { "epoch": 0.13321647677475898, "grad_norm": 2.1617531776428223, "learning_rate": 0.0001, "loss": 4.659748077392578, "step": 361 }, { "epoch": 0.1335854974860464, "grad_norm": 2.515841245651245, "learning_rate": 0.0001, "loss": 4.4495697021484375, "step": 362 }, { "epoch": 0.13395451819733384, "grad_norm": 2.170407772064209, "learning_rate": 0.0001, "loss": 4.438388347625732, "step": 363 }, { "epoch": 0.13432353890862125, "grad_norm": 2.7017438411712646, "learning_rate": 0.0001, "loss": 4.622103214263916, "step": 364 }, { "epoch": 0.13469255961990867, "grad_norm": 3.747965097427368, "learning_rate": 0.0001, "loss": 4.560365676879883, "step": 365 }, { "epoch": 0.13506158033119608, "grad_norm": 2.727411985397339, "learning_rate": 0.0001, "loss": 4.737141132354736, "step": 366 }, { "epoch": 0.1354306010424835, "grad_norm": 2.811324119567871, "learning_rate": 0.0001, "loss": 4.603142261505127, "step": 367 }, { "epoch": 0.13579962175377094, "grad_norm": 2.113163948059082, "learning_rate": 0.0001, "loss": 4.581015110015869, "step": 368 }, { "epoch": 0.13616864246505836, "grad_norm": 2.3354477882385254, "learning_rate": 0.0001, "loss": 4.416581630706787, "step": 369 }, { "epoch": 0.13653766317634577, "grad_norm": 1.9050804376602173, "learning_rate": 0.0001, "loss": 4.300870418548584, "step": 370 }, { "epoch": 0.1369066838876332, "grad_norm": 2.382469654083252, "learning_rate": 0.0001, "loss": 4.5361223220825195, "step": 371 }, { "epoch": 0.1372757045989206, "grad_norm": 2.4976327419281006, "learning_rate": 0.0001, "loss": 4.676791191101074, "step": 372 }, { "epoch": 0.13764472531020805, "grad_norm": 2.061070680618286, "learning_rate": 0.0001, "loss": 4.431769371032715, "step": 373 }, { "epoch": 0.13801374602149546, "grad_norm": 2.2746779918670654, "learning_rate": 0.0001, "loss": 4.311607837677002, "step": 374 }, { "epoch": 0.13838276673278288, "grad_norm": 2.6687610149383545, "learning_rate": 0.0001, "loss": 5.022364139556885, "step": 375 }, { "epoch": 0.1387517874440703, "grad_norm": 2.4456582069396973, "learning_rate": 0.0001, "loss": 4.58386754989624, "step": 376 }, { "epoch": 0.1391208081553577, "grad_norm": 3.6035640239715576, "learning_rate": 0.0001, "loss": 4.901766777038574, "step": 377 }, { "epoch": 0.13948982886664513, "grad_norm": 2.442969799041748, "learning_rate": 0.0001, "loss": 4.6150665283203125, "step": 378 }, { "epoch": 0.13985884957793257, "grad_norm": 2.26698899269104, "learning_rate": 0.0001, "loss": 4.721152305603027, "step": 379 }, { "epoch": 0.14022787028921999, "grad_norm": 2.289684534072876, "learning_rate": 0.0001, "loss": 4.594067573547363, "step": 380 }, { "epoch": 0.1405968910005074, "grad_norm": 2.3382251262664795, "learning_rate": 0.0001, "loss": 4.569343566894531, "step": 381 }, { "epoch": 0.14096591171179482, "grad_norm": 2.277287006378174, "learning_rate": 0.0001, "loss": 4.555330276489258, "step": 382 }, { "epoch": 0.14133493242308223, "grad_norm": 2.5999021530151367, "learning_rate": 0.0001, "loss": 4.632719993591309, "step": 383 }, { "epoch": 0.14170395313436968, "grad_norm": 2.2470548152923584, "learning_rate": 0.0001, "loss": 4.36204719543457, "step": 384 }, { "epoch": 0.1420729738456571, "grad_norm": 2.7499003410339355, "learning_rate": 0.0001, "loss": 4.743986129760742, "step": 385 }, { "epoch": 0.1424419945569445, "grad_norm": 2.742394208908081, "learning_rate": 0.0001, "loss": 4.636870384216309, "step": 386 }, { "epoch": 0.14281101526823192, "grad_norm": 2.180415391921997, "learning_rate": 0.0001, "loss": 4.016098976135254, "step": 387 }, { "epoch": 0.14318003597951934, "grad_norm": 2.76760196685791, "learning_rate": 0.0001, "loss": 4.450103759765625, "step": 388 }, { "epoch": 0.14354905669080678, "grad_norm": 2.0674028396606445, "learning_rate": 0.0001, "loss": 4.38997745513916, "step": 389 }, { "epoch": 0.1439180774020942, "grad_norm": 2.3727712631225586, "learning_rate": 0.0001, "loss": 4.650741100311279, "step": 390 }, { "epoch": 0.1442870981133816, "grad_norm": 3.3699772357940674, "learning_rate": 0.0001, "loss": 4.770174026489258, "step": 391 }, { "epoch": 0.14465611882466903, "grad_norm": 2.5660290718078613, "learning_rate": 0.0001, "loss": 4.350974082946777, "step": 392 }, { "epoch": 0.14502513953595644, "grad_norm": 2.490234613418579, "learning_rate": 0.0001, "loss": 4.715450286865234, "step": 393 }, { "epoch": 0.1453941602472439, "grad_norm": 2.437030076980591, "learning_rate": 0.0001, "loss": 4.426624298095703, "step": 394 }, { "epoch": 0.1457631809585313, "grad_norm": 2.3875768184661865, "learning_rate": 0.0001, "loss": 4.674262046813965, "step": 395 }, { "epoch": 0.14613220166981872, "grad_norm": 3.1936697959899902, "learning_rate": 0.0001, "loss": 4.431836128234863, "step": 396 }, { "epoch": 0.14650122238110613, "grad_norm": 2.755526542663574, "learning_rate": 0.0001, "loss": 4.50398588180542, "step": 397 }, { "epoch": 0.14687024309239355, "grad_norm": 2.3031511306762695, "learning_rate": 0.0001, "loss": 4.393855094909668, "step": 398 }, { "epoch": 0.147239263803681, "grad_norm": 2.156735420227051, "learning_rate": 0.0001, "loss": 4.521570205688477, "step": 399 }, { "epoch": 0.1476082845149684, "grad_norm": 2.1634793281555176, "learning_rate": 0.0001, "loss": 4.347168922424316, "step": 400 }, { "epoch": 0.14797730522625582, "grad_norm": 2.3540008068084717, "learning_rate": 0.0001, "loss": 4.335336685180664, "step": 401 }, { "epoch": 0.14834632593754324, "grad_norm": 2.099828004837036, "learning_rate": 0.0001, "loss": 4.490433216094971, "step": 402 }, { "epoch": 0.14871534664883065, "grad_norm": 2.301783800125122, "learning_rate": 0.0001, "loss": 4.512866973876953, "step": 403 }, { "epoch": 0.1490843673601181, "grad_norm": 4.869900226593018, "learning_rate": 0.0001, "loss": 4.349788665771484, "step": 404 }, { "epoch": 0.1494533880714055, "grad_norm": 2.6026854515075684, "learning_rate": 0.0001, "loss": 4.820313930511475, "step": 405 }, { "epoch": 0.14982240878269293, "grad_norm": 2.8290839195251465, "learning_rate": 0.0001, "loss": 4.873462200164795, "step": 406 }, { "epoch": 0.15019142949398034, "grad_norm": 2.684792995452881, "learning_rate": 0.0001, "loss": 4.517108917236328, "step": 407 }, { "epoch": 0.15056045020526776, "grad_norm": 2.500267505645752, "learning_rate": 0.0001, "loss": 4.553900241851807, "step": 408 }, { "epoch": 0.1509294709165552, "grad_norm": 2.6919009685516357, "learning_rate": 0.0001, "loss": 4.5190510749816895, "step": 409 }, { "epoch": 0.15129849162784262, "grad_norm": 2.3876686096191406, "learning_rate": 0.0001, "loss": 4.545931339263916, "step": 410 }, { "epoch": 0.15166751233913003, "grad_norm": 2.572681427001953, "learning_rate": 0.0001, "loss": 4.756655693054199, "step": 411 }, { "epoch": 0.15203653305041745, "grad_norm": 2.4582159519195557, "learning_rate": 0.0001, "loss": 4.263705730438232, "step": 412 }, { "epoch": 0.15240555376170487, "grad_norm": 2.2141013145446777, "learning_rate": 0.0001, "loss": 4.258913040161133, "step": 413 }, { "epoch": 0.1527745744729923, "grad_norm": 2.1214869022369385, "learning_rate": 0.0001, "loss": 4.433398723602295, "step": 414 }, { "epoch": 0.15314359518427972, "grad_norm": 3.5143837928771973, "learning_rate": 0.0001, "loss": 4.403622627258301, "step": 415 }, { "epoch": 0.15351261589556714, "grad_norm": 2.399301767349243, "learning_rate": 0.0001, "loss": 4.545055389404297, "step": 416 }, { "epoch": 0.15388163660685455, "grad_norm": 2.3257532119750977, "learning_rate": 0.0001, "loss": 4.454653263092041, "step": 417 }, { "epoch": 0.15425065731814197, "grad_norm": 2.069754123687744, "learning_rate": 0.0001, "loss": 4.391440391540527, "step": 418 }, { "epoch": 0.1546196780294294, "grad_norm": 2.3238942623138428, "learning_rate": 0.0001, "loss": 4.560251235961914, "step": 419 }, { "epoch": 0.15498869874071683, "grad_norm": 2.3678717613220215, "learning_rate": 0.0001, "loss": 4.415225028991699, "step": 420 }, { "epoch": 0.15535771945200424, "grad_norm": 2.5803234577178955, "learning_rate": 0.0001, "loss": 4.502828598022461, "step": 421 }, { "epoch": 0.15572674016329166, "grad_norm": 2.0687503814697266, "learning_rate": 0.0001, "loss": 4.573302268981934, "step": 422 }, { "epoch": 0.15609576087457908, "grad_norm": 2.0890164375305176, "learning_rate": 0.0001, "loss": 4.3898725509643555, "step": 423 }, { "epoch": 0.15646478158586652, "grad_norm": 6.507479667663574, "learning_rate": 0.0001, "loss": 4.818697452545166, "step": 424 }, { "epoch": 0.15683380229715393, "grad_norm": 2.371757745742798, "learning_rate": 0.0001, "loss": 4.283136367797852, "step": 425 }, { "epoch": 0.15720282300844135, "grad_norm": 2.1233596801757812, "learning_rate": 0.0001, "loss": 4.306228160858154, "step": 426 }, { "epoch": 0.15757184371972877, "grad_norm": 1.91828191280365, "learning_rate": 0.0001, "loss": 4.135748386383057, "step": 427 }, { "epoch": 0.15794086443101618, "grad_norm": 2.200847864151001, "learning_rate": 0.0001, "loss": 4.567739486694336, "step": 428 }, { "epoch": 0.15830988514230362, "grad_norm": 2.3396151065826416, "learning_rate": 0.0001, "loss": 4.297729015350342, "step": 429 }, { "epoch": 0.15867890585359104, "grad_norm": 2.224308967590332, "learning_rate": 0.0001, "loss": 4.248779773712158, "step": 430 }, { "epoch": 0.15904792656487846, "grad_norm": 2.5491580963134766, "learning_rate": 0.0001, "loss": 4.218733787536621, "step": 431 }, { "epoch": 0.15941694727616587, "grad_norm": 2.4283812046051025, "learning_rate": 0.0001, "loss": 4.502690315246582, "step": 432 }, { "epoch": 0.1597859679874533, "grad_norm": 2.155597448348999, "learning_rate": 0.0001, "loss": 4.561605930328369, "step": 433 }, { "epoch": 0.16015498869874073, "grad_norm": 2.9684088230133057, "learning_rate": 0.0001, "loss": 4.341606140136719, "step": 434 }, { "epoch": 0.16052400941002815, "grad_norm": 2.1386497020721436, "learning_rate": 0.0001, "loss": 4.319955825805664, "step": 435 }, { "epoch": 0.16089303012131556, "grad_norm": 2.0225462913513184, "learning_rate": 0.0001, "loss": 4.451445579528809, "step": 436 }, { "epoch": 0.16126205083260298, "grad_norm": 2.4696478843688965, "learning_rate": 0.0001, "loss": 4.350612163543701, "step": 437 }, { "epoch": 0.1616310715438904, "grad_norm": 3.6279401779174805, "learning_rate": 0.0001, "loss": 4.606565952301025, "step": 438 }, { "epoch": 0.16200009225517784, "grad_norm": 2.111154556274414, "learning_rate": 0.0001, "loss": 4.30889368057251, "step": 439 }, { "epoch": 0.16236911296646525, "grad_norm": 3.213124990463257, "learning_rate": 0.0001, "loss": 4.794126987457275, "step": 440 }, { "epoch": 0.16273813367775267, "grad_norm": 2.129725933074951, "learning_rate": 0.0001, "loss": 4.6410322189331055, "step": 441 }, { "epoch": 0.16310715438904008, "grad_norm": 1.9005858898162842, "learning_rate": 0.0001, "loss": 4.66891622543335, "step": 442 }, { "epoch": 0.1634761751003275, "grad_norm": 2.47507381439209, "learning_rate": 0.0001, "loss": 3.983210802078247, "step": 443 }, { "epoch": 0.1638451958116149, "grad_norm": 2.970912456512451, "learning_rate": 0.0001, "loss": 4.768887996673584, "step": 444 }, { "epoch": 0.16421421652290236, "grad_norm": 2.7908215522766113, "learning_rate": 0.0001, "loss": 4.4359517097473145, "step": 445 }, { "epoch": 0.16458323723418977, "grad_norm": 2.090733289718628, "learning_rate": 0.0001, "loss": 4.485992908477783, "step": 446 }, { "epoch": 0.1649522579454772, "grad_norm": 2.2121379375457764, "learning_rate": 0.0001, "loss": 4.506903648376465, "step": 447 }, { "epoch": 0.1653212786567646, "grad_norm": 2.533843517303467, "learning_rate": 0.0001, "loss": 4.350931644439697, "step": 448 }, { "epoch": 0.16569029936805202, "grad_norm": 2.2086493968963623, "learning_rate": 0.0001, "loss": 3.9907069206237793, "step": 449 }, { "epoch": 0.16605932007933946, "grad_norm": 2.3507115840911865, "learning_rate": 0.0001, "loss": 4.416835308074951, "step": 450 }, { "epoch": 0.16642834079062688, "grad_norm": 2.2729458808898926, "learning_rate": 0.0001, "loss": 4.586258411407471, "step": 451 }, { "epoch": 0.1667973615019143, "grad_norm": 2.1393730640411377, "learning_rate": 0.0001, "loss": 4.258632183074951, "step": 452 }, { "epoch": 0.1671663822132017, "grad_norm": 2.0558552742004395, "learning_rate": 0.0001, "loss": 4.597503662109375, "step": 453 }, { "epoch": 0.16753540292448912, "grad_norm": 2.473813533782959, "learning_rate": 0.0001, "loss": 4.395957946777344, "step": 454 }, { "epoch": 0.16790442363577657, "grad_norm": 2.095021963119507, "learning_rate": 0.0001, "loss": 4.062214374542236, "step": 455 }, { "epoch": 0.16827344434706398, "grad_norm": 2.6224169731140137, "learning_rate": 0.0001, "loss": 4.3939208984375, "step": 456 }, { "epoch": 0.1686424650583514, "grad_norm": 2.2001235485076904, "learning_rate": 0.0001, "loss": 4.333123683929443, "step": 457 }, { "epoch": 0.16901148576963881, "grad_norm": 2.21736741065979, "learning_rate": 0.0001, "loss": 4.0645246505737305, "step": 458 }, { "epoch": 0.16938050648092623, "grad_norm": 2.2809412479400635, "learning_rate": 0.0001, "loss": 4.523746967315674, "step": 459 }, { "epoch": 0.16974952719221367, "grad_norm": 2.6346044540405273, "learning_rate": 0.0001, "loss": 4.581881999969482, "step": 460 }, { "epoch": 0.1701185479035011, "grad_norm": 2.1072700023651123, "learning_rate": 0.0001, "loss": 4.362997531890869, "step": 461 }, { "epoch": 0.1704875686147885, "grad_norm": 2.147351026535034, "learning_rate": 0.0001, "loss": 4.131965637207031, "step": 462 }, { "epoch": 0.17085658932607592, "grad_norm": 2.3557586669921875, "learning_rate": 0.0001, "loss": 4.431070327758789, "step": 463 }, { "epoch": 0.17122561003736333, "grad_norm": 2.1692962646484375, "learning_rate": 0.0001, "loss": 4.711268901824951, "step": 464 }, { "epoch": 0.17159463074865078, "grad_norm": 2.254636764526367, "learning_rate": 0.0001, "loss": 4.4809041023254395, "step": 465 }, { "epoch": 0.1719636514599382, "grad_norm": 2.2716152667999268, "learning_rate": 0.0001, "loss": 4.050970077514648, "step": 466 }, { "epoch": 0.1723326721712256, "grad_norm": 2.9789211750030518, "learning_rate": 0.0001, "loss": 4.269569396972656, "step": 467 }, { "epoch": 0.17270169288251302, "grad_norm": 2.0386388301849365, "learning_rate": 0.0001, "loss": 4.284242630004883, "step": 468 }, { "epoch": 0.17307071359380044, "grad_norm": 1.8008029460906982, "learning_rate": 0.0001, "loss": 4.464146137237549, "step": 469 }, { "epoch": 0.17343973430508788, "grad_norm": 1.944584608078003, "learning_rate": 0.0001, "loss": 4.559577465057373, "step": 470 }, { "epoch": 0.1738087550163753, "grad_norm": 4.818556785583496, "learning_rate": 0.0001, "loss": 4.378731727600098, "step": 471 }, { "epoch": 0.17417777572766271, "grad_norm": 2.3885557651519775, "learning_rate": 0.0001, "loss": 4.05840539932251, "step": 472 }, { "epoch": 0.17454679643895013, "grad_norm": 2.1333820819854736, "learning_rate": 0.0001, "loss": 4.016430377960205, "step": 473 }, { "epoch": 0.17491581715023755, "grad_norm": 1.8299129009246826, "learning_rate": 0.0001, "loss": 4.615049362182617, "step": 474 }, { "epoch": 0.175284837861525, "grad_norm": 2.0134215354919434, "learning_rate": 0.0001, "loss": 4.2402191162109375, "step": 475 }, { "epoch": 0.1756538585728124, "grad_norm": 2.5740880966186523, "learning_rate": 0.0001, "loss": 4.674981594085693, "step": 476 }, { "epoch": 0.17602287928409982, "grad_norm": 3.1095097064971924, "learning_rate": 0.0001, "loss": 4.765364646911621, "step": 477 }, { "epoch": 0.17639189999538724, "grad_norm": 2.1833994388580322, "learning_rate": 0.0001, "loss": 4.4260969161987305, "step": 478 }, { "epoch": 0.17676092070667465, "grad_norm": 2.396684408187866, "learning_rate": 0.0001, "loss": 4.116750717163086, "step": 479 }, { "epoch": 0.1771299414179621, "grad_norm": 2.2355973720550537, "learning_rate": 0.0001, "loss": 4.551278114318848, "step": 480 }, { "epoch": 0.1774989621292495, "grad_norm": 2.0249733924865723, "learning_rate": 0.0001, "loss": 4.321774959564209, "step": 481 }, { "epoch": 0.17786798284053693, "grad_norm": 2.2655394077301025, "learning_rate": 0.0001, "loss": 4.832188129425049, "step": 482 }, { "epoch": 0.17823700355182434, "grad_norm": 2.384474754333496, "learning_rate": 0.0001, "loss": 4.327328681945801, "step": 483 }, { "epoch": 0.17860602426311176, "grad_norm": 2.0592844486236572, "learning_rate": 0.0001, "loss": 4.406461238861084, "step": 484 }, { "epoch": 0.1789750449743992, "grad_norm": 2.1442856788635254, "learning_rate": 0.0001, "loss": 4.520481109619141, "step": 485 }, { "epoch": 0.17934406568568662, "grad_norm": 2.215745449066162, "learning_rate": 0.0001, "loss": 4.652613162994385, "step": 486 }, { "epoch": 0.17971308639697403, "grad_norm": 2.274385452270508, "learning_rate": 0.0001, "loss": 4.071110248565674, "step": 487 }, { "epoch": 0.18008210710826145, "grad_norm": 2.5120763778686523, "learning_rate": 0.0001, "loss": 4.499899387359619, "step": 488 }, { "epoch": 0.18045112781954886, "grad_norm": 2.1488053798675537, "learning_rate": 0.0001, "loss": 4.293880939483643, "step": 489 }, { "epoch": 0.1808201485308363, "grad_norm": 2.8950445652008057, "learning_rate": 0.0001, "loss": 4.485057830810547, "step": 490 }, { "epoch": 0.18118916924212372, "grad_norm": 2.100581407546997, "learning_rate": 0.0001, "loss": 4.113532066345215, "step": 491 }, { "epoch": 0.18155818995341114, "grad_norm": 2.3395795822143555, "learning_rate": 0.0001, "loss": 4.322099208831787, "step": 492 }, { "epoch": 0.18192721066469855, "grad_norm": 2.030890703201294, "learning_rate": 0.0001, "loss": 4.442522048950195, "step": 493 }, { "epoch": 0.18229623137598597, "grad_norm": 1.977658987045288, "learning_rate": 0.0001, "loss": 4.220512866973877, "step": 494 }, { "epoch": 0.1826652520872734, "grad_norm": 3.0135514736175537, "learning_rate": 0.0001, "loss": 4.858924388885498, "step": 495 }, { "epoch": 0.18303427279856083, "grad_norm": 2.4730112552642822, "learning_rate": 0.0001, "loss": 4.4013543128967285, "step": 496 }, { "epoch": 0.18340329350984824, "grad_norm": 2.456479549407959, "learning_rate": 0.0001, "loss": 4.416372776031494, "step": 497 }, { "epoch": 0.18377231422113566, "grad_norm": 2.1360464096069336, "learning_rate": 0.0001, "loss": 4.0488104820251465, "step": 498 }, { "epoch": 0.18414133493242307, "grad_norm": 2.367122173309326, "learning_rate": 0.0001, "loss": 4.326754570007324, "step": 499 }, { "epoch": 0.18451035564371052, "grad_norm": 2.3929057121276855, "learning_rate": 0.0001, "loss": 4.3895063400268555, "step": 500 }, { "epoch": 0.18487937635499793, "grad_norm": 1.9481403827667236, "learning_rate": 0.0001, "loss": 4.406152248382568, "step": 501 }, { "epoch": 0.18524839706628535, "grad_norm": 2.086376190185547, "learning_rate": 0.0001, "loss": 4.4221906661987305, "step": 502 }, { "epoch": 0.18561741777757276, "grad_norm": 2.2188632488250732, "learning_rate": 0.0001, "loss": 4.660634994506836, "step": 503 }, { "epoch": 0.18598643848886018, "grad_norm": 2.2751147747039795, "learning_rate": 0.0001, "loss": 4.23121452331543, "step": 504 }, { "epoch": 0.18635545920014762, "grad_norm": 2.1705162525177, "learning_rate": 0.0001, "loss": 4.265094757080078, "step": 505 }, { "epoch": 0.18672447991143504, "grad_norm": 2.121332883834839, "learning_rate": 0.0001, "loss": 4.205435752868652, "step": 506 }, { "epoch": 0.18709350062272245, "grad_norm": 1.964434266090393, "learning_rate": 0.0001, "loss": 4.287982940673828, "step": 507 }, { "epoch": 0.18746252133400987, "grad_norm": 2.236156463623047, "learning_rate": 0.0001, "loss": 4.284115791320801, "step": 508 }, { "epoch": 0.18783154204529728, "grad_norm": 2.2012174129486084, "learning_rate": 0.0001, "loss": 4.316831111907959, "step": 509 }, { "epoch": 0.1882005627565847, "grad_norm": 1.9142396450042725, "learning_rate": 0.0001, "loss": 4.5977678298950195, "step": 510 }, { "epoch": 0.18856958346787214, "grad_norm": 1.975645899772644, "learning_rate": 0.0001, "loss": 4.377478122711182, "step": 511 }, { "epoch": 0.18893860417915956, "grad_norm": 2.210484027862549, "learning_rate": 0.0001, "loss": 4.167498588562012, "step": 512 }, { "epoch": 0.18930762489044697, "grad_norm": 1.8951791524887085, "learning_rate": 0.0001, "loss": 4.046764373779297, "step": 513 }, { "epoch": 0.1896766456017344, "grad_norm": 2.129709005355835, "learning_rate": 0.0001, "loss": 4.1118035316467285, "step": 514 }, { "epoch": 0.1900456663130218, "grad_norm": 1.926705241203308, "learning_rate": 0.0001, "loss": 3.9770379066467285, "step": 515 }, { "epoch": 0.19041468702430925, "grad_norm": 2.785149335861206, "learning_rate": 0.0001, "loss": 4.119317054748535, "step": 516 }, { "epoch": 0.19078370773559666, "grad_norm": 2.28780198097229, "learning_rate": 0.0001, "loss": 3.9503774642944336, "step": 517 }, { "epoch": 0.19115272844688408, "grad_norm": 2.1383399963378906, "learning_rate": 0.0001, "loss": 4.462101459503174, "step": 518 }, { "epoch": 0.1915217491581715, "grad_norm": 1.8943592309951782, "learning_rate": 0.0001, "loss": 4.395929336547852, "step": 519 }, { "epoch": 0.1918907698694589, "grad_norm": 2.1494009494781494, "learning_rate": 0.0001, "loss": 4.108682155609131, "step": 520 }, { "epoch": 0.19225979058074635, "grad_norm": 2.092240810394287, "learning_rate": 0.0001, "loss": 4.348626613616943, "step": 521 }, { "epoch": 0.19262881129203377, "grad_norm": 2.151693820953369, "learning_rate": 0.0001, "loss": 4.288055896759033, "step": 522 }, { "epoch": 0.19299783200332118, "grad_norm": 1.9059604406356812, "learning_rate": 0.0001, "loss": 4.300743103027344, "step": 523 }, { "epoch": 0.1933668527146086, "grad_norm": 2.5519795417785645, "learning_rate": 0.0001, "loss": 4.22593355178833, "step": 524 }, { "epoch": 0.19373587342589602, "grad_norm": 1.7208030223846436, "learning_rate": 0.0001, "loss": 4.30415153503418, "step": 525 }, { "epoch": 0.19410489413718346, "grad_norm": 1.7663142681121826, "learning_rate": 0.0001, "loss": 4.290582180023193, "step": 526 }, { "epoch": 0.19447391484847087, "grad_norm": 2.063027858734131, "learning_rate": 0.0001, "loss": 4.227534770965576, "step": 527 }, { "epoch": 0.1948429355597583, "grad_norm": 2.6470508575439453, "learning_rate": 0.0001, "loss": 4.313108921051025, "step": 528 }, { "epoch": 0.1952119562710457, "grad_norm": 1.9663456678390503, "learning_rate": 0.0001, "loss": 4.123837947845459, "step": 529 }, { "epoch": 0.19558097698233312, "grad_norm": 2.032116413116455, "learning_rate": 0.0001, "loss": 3.8263444900512695, "step": 530 }, { "epoch": 0.19594999769362056, "grad_norm": 3.4389142990112305, "learning_rate": 0.0001, "loss": 4.467240333557129, "step": 531 }, { "epoch": 0.19631901840490798, "grad_norm": 1.9850727319717407, "learning_rate": 0.0001, "loss": 4.335421085357666, "step": 532 }, { "epoch": 0.1966880391161954, "grad_norm": 2.0280203819274902, "learning_rate": 0.0001, "loss": 4.235400676727295, "step": 533 }, { "epoch": 0.1970570598274828, "grad_norm": 2.8351972103118896, "learning_rate": 0.0001, "loss": 4.130883693695068, "step": 534 }, { "epoch": 0.19742608053877023, "grad_norm": 2.1816749572753906, "learning_rate": 0.0001, "loss": 4.412903785705566, "step": 535 }, { "epoch": 0.19779510125005767, "grad_norm": 2.455271005630493, "learning_rate": 0.0001, "loss": 4.324191093444824, "step": 536 }, { "epoch": 0.19816412196134509, "grad_norm": 2.081894874572754, "learning_rate": 0.0001, "loss": 4.476474761962891, "step": 537 }, { "epoch": 0.1985331426726325, "grad_norm": 2.1476821899414062, "learning_rate": 0.0001, "loss": 4.266482830047607, "step": 538 }, { "epoch": 0.19890216338391992, "grad_norm": 1.989898681640625, "learning_rate": 0.0001, "loss": 4.313356876373291, "step": 539 }, { "epoch": 0.19927118409520733, "grad_norm": 2.1811554431915283, "learning_rate": 0.0001, "loss": 4.685174465179443, "step": 540 }, { "epoch": 0.19964020480649478, "grad_norm": 2.0105929374694824, "learning_rate": 0.0001, "loss": 4.369940280914307, "step": 541 }, { "epoch": 0.2000092255177822, "grad_norm": 1.7295829057693481, "learning_rate": 0.0001, "loss": 3.995760679244995, "step": 542 }, { "epoch": 0.2003782462290696, "grad_norm": 2.0666468143463135, "learning_rate": 0.0001, "loss": 4.386075973510742, "step": 543 }, { "epoch": 0.20074726694035702, "grad_norm": 2.109861135482788, "learning_rate": 0.0001, "loss": 4.321488380432129, "step": 544 }, { "epoch": 0.20111628765164444, "grad_norm": 2.504140853881836, "learning_rate": 0.0001, "loss": 3.944672107696533, "step": 545 }, { "epoch": 0.20148530836293188, "grad_norm": 2.5921802520751953, "learning_rate": 0.0001, "loss": 4.286891460418701, "step": 546 }, { "epoch": 0.2018543290742193, "grad_norm": 1.767426609992981, "learning_rate": 0.0001, "loss": 4.364938259124756, "step": 547 }, { "epoch": 0.2022233497855067, "grad_norm": 2.15252685546875, "learning_rate": 0.0001, "loss": 4.374980449676514, "step": 548 }, { "epoch": 0.20259237049679413, "grad_norm": 1.9988367557525635, "learning_rate": 0.0001, "loss": 4.222935676574707, "step": 549 }, { "epoch": 0.20296139120808154, "grad_norm": 1.9735081195831299, "learning_rate": 0.0001, "loss": 4.353124618530273, "step": 550 }, { "epoch": 0.203330411919369, "grad_norm": 2.0647451877593994, "learning_rate": 0.0001, "loss": 4.040650367736816, "step": 551 }, { "epoch": 0.2036994326306564, "grad_norm": 1.9258832931518555, "learning_rate": 0.0001, "loss": 4.263591289520264, "step": 552 }, { "epoch": 0.20406845334194382, "grad_norm": 2.213529348373413, "learning_rate": 0.0001, "loss": 4.239419937133789, "step": 553 }, { "epoch": 0.20443747405323123, "grad_norm": 2.0212087631225586, "learning_rate": 0.0001, "loss": 3.937243700027466, "step": 554 }, { "epoch": 0.20480649476451865, "grad_norm": 2.1524879932403564, "learning_rate": 0.0001, "loss": 4.281352996826172, "step": 555 }, { "epoch": 0.2051755154758061, "grad_norm": 2.441908359527588, "learning_rate": 0.0001, "loss": 4.405776500701904, "step": 556 }, { "epoch": 0.2055445361870935, "grad_norm": 1.824597716331482, "learning_rate": 0.0001, "loss": 3.8622190952301025, "step": 557 }, { "epoch": 0.20591355689838092, "grad_norm": 2.2437732219696045, "learning_rate": 0.0001, "loss": 4.441629409790039, "step": 558 }, { "epoch": 0.20628257760966834, "grad_norm": 2.655653715133667, "learning_rate": 0.0001, "loss": 4.283421039581299, "step": 559 }, { "epoch": 0.20665159832095575, "grad_norm": 2.883206367492676, "learning_rate": 0.0001, "loss": 4.253350257873535, "step": 560 }, { "epoch": 0.2070206190322432, "grad_norm": 2.066970109939575, "learning_rate": 0.0001, "loss": 4.132658958435059, "step": 561 }, { "epoch": 0.2073896397435306, "grad_norm": 1.9143986701965332, "learning_rate": 0.0001, "loss": 4.017167091369629, "step": 562 }, { "epoch": 0.20775866045481803, "grad_norm": 1.868736982345581, "learning_rate": 0.0001, "loss": 4.328319549560547, "step": 563 }, { "epoch": 0.20812768116610544, "grad_norm": 2.3013720512390137, "learning_rate": 0.0001, "loss": 4.320085525512695, "step": 564 }, { "epoch": 0.20849670187739286, "grad_norm": 1.9836920499801636, "learning_rate": 0.0001, "loss": 4.308595657348633, "step": 565 }, { "epoch": 0.2088657225886803, "grad_norm": 2.0807578563690186, "learning_rate": 0.0001, "loss": 4.460840225219727, "step": 566 }, { "epoch": 0.20923474329996772, "grad_norm": 2.0584962368011475, "learning_rate": 0.0001, "loss": 4.269477367401123, "step": 567 }, { "epoch": 0.20960376401125513, "grad_norm": 1.9931753873825073, "learning_rate": 0.0001, "loss": 4.077987194061279, "step": 568 }, { "epoch": 0.20997278472254255, "grad_norm": 2.14865779876709, "learning_rate": 0.0001, "loss": 3.9619553089141846, "step": 569 }, { "epoch": 0.21034180543382996, "grad_norm": 2.4904727935791016, "learning_rate": 0.0001, "loss": 4.423192024230957, "step": 570 }, { "epoch": 0.2107108261451174, "grad_norm": 2.386594295501709, "learning_rate": 0.0001, "loss": 4.001859188079834, "step": 571 }, { "epoch": 0.21107984685640482, "grad_norm": 2.5853636264801025, "learning_rate": 0.0001, "loss": 4.339788436889648, "step": 572 }, { "epoch": 0.21144886756769224, "grad_norm": 2.092083692550659, "learning_rate": 0.0001, "loss": 4.241220474243164, "step": 573 }, { "epoch": 0.21181788827897965, "grad_norm": 1.840279221534729, "learning_rate": 0.0001, "loss": 3.8160362243652344, "step": 574 }, { "epoch": 0.21218690899026707, "grad_norm": 2.173718214035034, "learning_rate": 0.0001, "loss": 4.18009090423584, "step": 575 }, { "epoch": 0.2125559297015545, "grad_norm": 2.0022976398468018, "learning_rate": 0.0001, "loss": 4.075108051300049, "step": 576 }, { "epoch": 0.21292495041284193, "grad_norm": 2.2780895233154297, "learning_rate": 0.0001, "loss": 4.3333234786987305, "step": 577 }, { "epoch": 0.21329397112412934, "grad_norm": 1.7476773262023926, "learning_rate": 0.0001, "loss": 4.126917362213135, "step": 578 }, { "epoch": 0.21366299183541676, "grad_norm": 1.9226078987121582, "learning_rate": 0.0001, "loss": 3.9116811752319336, "step": 579 }, { "epoch": 0.21403201254670418, "grad_norm": 1.8429948091506958, "learning_rate": 0.0001, "loss": 4.143172264099121, "step": 580 }, { "epoch": 0.2144010332579916, "grad_norm": 1.9218641519546509, "learning_rate": 0.0001, "loss": 3.8314731121063232, "step": 581 }, { "epoch": 0.21477005396927903, "grad_norm": 1.9137080907821655, "learning_rate": 0.0001, "loss": 4.0515828132629395, "step": 582 }, { "epoch": 0.21513907468056645, "grad_norm": 3.729459524154663, "learning_rate": 0.0001, "loss": 4.178072929382324, "step": 583 }, { "epoch": 0.21550809539185387, "grad_norm": 1.6776292324066162, "learning_rate": 0.0001, "loss": 4.100445747375488, "step": 584 }, { "epoch": 0.21587711610314128, "grad_norm": 2.338021993637085, "learning_rate": 0.0001, "loss": 4.202198505401611, "step": 585 }, { "epoch": 0.2162461368144287, "grad_norm": 1.988771915435791, "learning_rate": 0.0001, "loss": 4.170504570007324, "step": 586 }, { "epoch": 0.21661515752571614, "grad_norm": 1.8706896305084229, "learning_rate": 0.0001, "loss": 3.869433641433716, "step": 587 }, { "epoch": 0.21698417823700356, "grad_norm": 2.2407386302948, "learning_rate": 0.0001, "loss": 3.876044273376465, "step": 588 }, { "epoch": 0.21735319894829097, "grad_norm": 2.3377597332000732, "learning_rate": 0.0001, "loss": 3.978440046310425, "step": 589 }, { "epoch": 0.2177222196595784, "grad_norm": 2.2066588401794434, "learning_rate": 0.0001, "loss": 4.019556045532227, "step": 590 }, { "epoch": 0.2180912403708658, "grad_norm": 2.0315310955047607, "learning_rate": 0.0001, "loss": 4.088963508605957, "step": 591 }, { "epoch": 0.21846026108215325, "grad_norm": 1.8932534456253052, "learning_rate": 0.0001, "loss": 4.147191524505615, "step": 592 }, { "epoch": 0.21882928179344066, "grad_norm": 2.4088425636291504, "learning_rate": 0.0001, "loss": 4.037857532501221, "step": 593 }, { "epoch": 0.21919830250472808, "grad_norm": 1.8900609016418457, "learning_rate": 0.0001, "loss": 4.142814636230469, "step": 594 }, { "epoch": 0.2195673232160155, "grad_norm": 2.220301866531372, "learning_rate": 0.0001, "loss": 4.271300315856934, "step": 595 }, { "epoch": 0.2199363439273029, "grad_norm": 2.1047005653381348, "learning_rate": 0.0001, "loss": 4.119107723236084, "step": 596 }, { "epoch": 0.22030536463859035, "grad_norm": 2.2996177673339844, "learning_rate": 0.0001, "loss": 4.6851725578308105, "step": 597 }, { "epoch": 0.22067438534987777, "grad_norm": 2.628307342529297, "learning_rate": 0.0001, "loss": 4.41997766494751, "step": 598 }, { "epoch": 0.22104340606116518, "grad_norm": 1.8094673156738281, "learning_rate": 0.0001, "loss": 4.277165412902832, "step": 599 }, { "epoch": 0.2214124267724526, "grad_norm": 1.794776201248169, "learning_rate": 0.0001, "loss": 4.271518230438232, "step": 600 }, { "epoch": 0.22178144748374, "grad_norm": 1.8499183654785156, "learning_rate": 0.0001, "loss": 4.174226760864258, "step": 601 }, { "epoch": 0.22215046819502746, "grad_norm": 1.818971037864685, "learning_rate": 0.0001, "loss": 4.11065673828125, "step": 602 }, { "epoch": 0.22251948890631487, "grad_norm": 2.016084909439087, "learning_rate": 0.0001, "loss": 4.37334680557251, "step": 603 }, { "epoch": 0.2228885096176023, "grad_norm": 2.3137686252593994, "learning_rate": 0.0001, "loss": 4.305994987487793, "step": 604 }, { "epoch": 0.2232575303288897, "grad_norm": 2.139861583709717, "learning_rate": 0.0001, "loss": 4.259668350219727, "step": 605 }, { "epoch": 0.22362655104017712, "grad_norm": 1.8052783012390137, "learning_rate": 0.0001, "loss": 4.200760364532471, "step": 606 }, { "epoch": 0.22399557175146456, "grad_norm": 2.228914737701416, "learning_rate": 0.0001, "loss": 4.199825286865234, "step": 607 }, { "epoch": 0.22436459246275198, "grad_norm": 2.5564470291137695, "learning_rate": 0.0001, "loss": 3.8638081550598145, "step": 608 }, { "epoch": 0.2247336131740394, "grad_norm": 2.1222283840179443, "learning_rate": 0.0001, "loss": 3.7322850227355957, "step": 609 }, { "epoch": 0.2251026338853268, "grad_norm": 2.641051769256592, "learning_rate": 0.0001, "loss": 4.397209167480469, "step": 610 }, { "epoch": 0.22547165459661422, "grad_norm": 2.2051429748535156, "learning_rate": 0.0001, "loss": 4.315612316131592, "step": 611 }, { "epoch": 0.22584067530790167, "grad_norm": 2.2115986347198486, "learning_rate": 0.0001, "loss": 4.010114669799805, "step": 612 }, { "epoch": 0.22620969601918908, "grad_norm": 1.726802945137024, "learning_rate": 0.0001, "loss": 4.505970001220703, "step": 613 }, { "epoch": 0.2265787167304765, "grad_norm": 1.6395057439804077, "learning_rate": 0.0001, "loss": 4.453322410583496, "step": 614 }, { "epoch": 0.2269477374417639, "grad_norm": 2.0230486392974854, "learning_rate": 0.0001, "loss": 3.912433385848999, "step": 615 }, { "epoch": 0.22731675815305133, "grad_norm": 2.265876054763794, "learning_rate": 0.0001, "loss": 4.255008697509766, "step": 616 }, { "epoch": 0.22768577886433877, "grad_norm": 2.2303671836853027, "learning_rate": 0.0001, "loss": 4.022263526916504, "step": 617 }, { "epoch": 0.2280547995756262, "grad_norm": 2.465766668319702, "learning_rate": 0.0001, "loss": 4.011572360992432, "step": 618 }, { "epoch": 0.2284238202869136, "grad_norm": 1.9247350692749023, "learning_rate": 0.0001, "loss": 4.222007751464844, "step": 619 }, { "epoch": 0.22879284099820102, "grad_norm": 2.3126766681671143, "learning_rate": 0.0001, "loss": 4.250060558319092, "step": 620 }, { "epoch": 0.22916186170948843, "grad_norm": 1.9733937978744507, "learning_rate": 0.0001, "loss": 3.859759569168091, "step": 621 }, { "epoch": 0.22953088242077588, "grad_norm": 2.1442854404449463, "learning_rate": 0.0001, "loss": 3.7772088050842285, "step": 622 }, { "epoch": 0.2298999031320633, "grad_norm": 1.9578642845153809, "learning_rate": 0.0001, "loss": 4.250098705291748, "step": 623 }, { "epoch": 0.2302689238433507, "grad_norm": 2.1606431007385254, "learning_rate": 0.0001, "loss": 4.332361221313477, "step": 624 }, { "epoch": 0.23063794455463812, "grad_norm": 1.9267970323562622, "learning_rate": 0.0001, "loss": 4.073374271392822, "step": 625 }, { "epoch": 0.23100696526592554, "grad_norm": 1.9631792306900024, "learning_rate": 0.0001, "loss": 3.9883084297180176, "step": 626 }, { "epoch": 0.23137598597721298, "grad_norm": 1.8955382108688354, "learning_rate": 0.0001, "loss": 3.8627865314483643, "step": 627 }, { "epoch": 0.2317450066885004, "grad_norm": 1.6978048086166382, "learning_rate": 0.0001, "loss": 4.236217975616455, "step": 628 }, { "epoch": 0.23211402739978781, "grad_norm": 2.5762033462524414, "learning_rate": 0.0001, "loss": 3.6205215454101562, "step": 629 }, { "epoch": 0.23248304811107523, "grad_norm": 1.9607030153274536, "learning_rate": 0.0001, "loss": 3.698427438735962, "step": 630 }, { "epoch": 0.23285206882236265, "grad_norm": 2.8456151485443115, "learning_rate": 0.0001, "loss": 4.37019681930542, "step": 631 }, { "epoch": 0.2332210895336501, "grad_norm": 1.8558168411254883, "learning_rate": 0.0001, "loss": 4.02039909362793, "step": 632 }, { "epoch": 0.2335901102449375, "grad_norm": 1.9798914194107056, "learning_rate": 0.0001, "loss": 3.885150194168091, "step": 633 }, { "epoch": 0.23395913095622492, "grad_norm": 2.588681697845459, "learning_rate": 0.0001, "loss": 4.107516288757324, "step": 634 }, { "epoch": 0.23432815166751234, "grad_norm": 1.9660398960113525, "learning_rate": 0.0001, "loss": 4.198762893676758, "step": 635 }, { "epoch": 0.23469717237879975, "grad_norm": 1.718747615814209, "learning_rate": 0.0001, "loss": 3.67313551902771, "step": 636 }, { "epoch": 0.2350661930900872, "grad_norm": 1.7743703126907349, "learning_rate": 0.0001, "loss": 3.94923996925354, "step": 637 }, { "epoch": 0.2354352138013746, "grad_norm": 2.002715587615967, "learning_rate": 0.0001, "loss": 4.1457109451293945, "step": 638 }, { "epoch": 0.23580423451266203, "grad_norm": 2.4004321098327637, "learning_rate": 0.0001, "loss": 4.041502952575684, "step": 639 }, { "epoch": 0.23617325522394944, "grad_norm": 2.460703134536743, "learning_rate": 0.0001, "loss": 4.037372589111328, "step": 640 }, { "epoch": 0.23654227593523686, "grad_norm": 1.6822046041488647, "learning_rate": 0.0001, "loss": 3.8982839584350586, "step": 641 }, { "epoch": 0.2369112966465243, "grad_norm": 1.6533682346343994, "learning_rate": 0.0001, "loss": 3.9182395935058594, "step": 642 }, { "epoch": 0.23728031735781172, "grad_norm": 1.8066723346710205, "learning_rate": 0.0001, "loss": 3.917219638824463, "step": 643 }, { "epoch": 0.23764933806909913, "grad_norm": 1.741705298423767, "learning_rate": 0.0001, "loss": 3.824510335922241, "step": 644 }, { "epoch": 0.23801835878038655, "grad_norm": 2.118285894393921, "learning_rate": 0.0001, "loss": 3.926053285598755, "step": 645 }, { "epoch": 0.23838737949167396, "grad_norm": 2.231577157974243, "learning_rate": 0.0001, "loss": 4.288327217102051, "step": 646 }, { "epoch": 0.23875640020296138, "grad_norm": 2.0882058143615723, "learning_rate": 0.0001, "loss": 4.010497093200684, "step": 647 }, { "epoch": 0.23912542091424882, "grad_norm": 1.7932448387145996, "learning_rate": 0.0001, "loss": 3.9600930213928223, "step": 648 }, { "epoch": 0.23949444162553624, "grad_norm": 2.145545721054077, "learning_rate": 0.0001, "loss": 3.8396453857421875, "step": 649 }, { "epoch": 0.23986346233682365, "grad_norm": 1.8074281215667725, "learning_rate": 0.0001, "loss": 4.326630592346191, "step": 650 }, { "epoch": 0.24023248304811107, "grad_norm": 1.7010893821716309, "learning_rate": 0.0001, "loss": 3.7501673698425293, "step": 651 }, { "epoch": 0.24060150375939848, "grad_norm": 1.8481875658035278, "learning_rate": 0.0001, "loss": 4.053319454193115, "step": 652 }, { "epoch": 0.24097052447068593, "grad_norm": 2.8165059089660645, "learning_rate": 0.0001, "loss": 3.818509578704834, "step": 653 }, { "epoch": 0.24133954518197334, "grad_norm": 2.072035551071167, "learning_rate": 0.0001, "loss": 4.141574859619141, "step": 654 }, { "epoch": 0.24170856589326076, "grad_norm": 1.8036015033721924, "learning_rate": 0.0001, "loss": 3.6866941452026367, "step": 655 }, { "epoch": 0.24207758660454817, "grad_norm": 1.6921944618225098, "learning_rate": 0.0001, "loss": 4.3356547355651855, "step": 656 }, { "epoch": 0.2424466073158356, "grad_norm": 1.7246938943862915, "learning_rate": 0.0001, "loss": 4.07541036605835, "step": 657 }, { "epoch": 0.24281562802712303, "grad_norm": 1.855697512626648, "learning_rate": 0.0001, "loss": 4.163908004760742, "step": 658 }, { "epoch": 0.24318464873841045, "grad_norm": 1.7613106966018677, "learning_rate": 0.0001, "loss": 3.863572359085083, "step": 659 }, { "epoch": 0.24355366944969786, "grad_norm": 1.9448808431625366, "learning_rate": 0.0001, "loss": 4.276730537414551, "step": 660 }, { "epoch": 0.24392269016098528, "grad_norm": 2.0929248332977295, "learning_rate": 0.0001, "loss": 3.9885942935943604, "step": 661 }, { "epoch": 0.2442917108722727, "grad_norm": 2.1055312156677246, "learning_rate": 0.0001, "loss": 3.8733935356140137, "step": 662 }, { "epoch": 0.24466073158356014, "grad_norm": 1.7790578603744507, "learning_rate": 0.0001, "loss": 3.9471259117126465, "step": 663 }, { "epoch": 0.24502975229484755, "grad_norm": 1.8341556787490845, "learning_rate": 0.0001, "loss": 4.12840461730957, "step": 664 }, { "epoch": 0.24539877300613497, "grad_norm": 1.8000987768173218, "learning_rate": 0.0001, "loss": 3.5752415657043457, "step": 665 }, { "epoch": 0.24576779371742238, "grad_norm": 1.6630133390426636, "learning_rate": 0.0001, "loss": 4.189026355743408, "step": 666 }, { "epoch": 0.2461368144287098, "grad_norm": 2.0056142807006836, "learning_rate": 0.0001, "loss": 4.161567687988281, "step": 667 }, { "epoch": 0.24650583513999724, "grad_norm": 2.0732192993164062, "learning_rate": 0.0001, "loss": 4.169988632202148, "step": 668 }, { "epoch": 0.24687485585128466, "grad_norm": 3.8518526554107666, "learning_rate": 0.0001, "loss": 4.432511806488037, "step": 669 }, { "epoch": 0.24724387656257207, "grad_norm": 2.418034076690674, "learning_rate": 0.0001, "loss": 3.945303440093994, "step": 670 }, { "epoch": 0.2476128972738595, "grad_norm": 2.0764167308807373, "learning_rate": 0.0001, "loss": 3.717202663421631, "step": 671 }, { "epoch": 0.2479819179851469, "grad_norm": 1.8870404958724976, "learning_rate": 0.0001, "loss": 3.550865650177002, "step": 672 }, { "epoch": 0.24835093869643435, "grad_norm": 1.7513939142227173, "learning_rate": 0.0001, "loss": 3.9279532432556152, "step": 673 }, { "epoch": 0.24871995940772176, "grad_norm": 2.062190294265747, "learning_rate": 0.0001, "loss": 4.258135795593262, "step": 674 }, { "epoch": 0.24908898011900918, "grad_norm": 2.407712697982788, "learning_rate": 0.0001, "loss": 3.9930434226989746, "step": 675 }, { "epoch": 0.2494580008302966, "grad_norm": 4.875779151916504, "learning_rate": 0.0001, "loss": 4.053095817565918, "step": 676 }, { "epoch": 0.249827021541584, "grad_norm": 2.107910633087158, "learning_rate": 0.0001, "loss": 3.8471646308898926, "step": 677 }, { "epoch": 0.25019604225287145, "grad_norm": 1.7939298152923584, "learning_rate": 0.0001, "loss": 4.007920742034912, "step": 678 }, { "epoch": 0.25056506296415887, "grad_norm": 1.8970980644226074, "learning_rate": 0.0001, "loss": 3.9944920539855957, "step": 679 }, { "epoch": 0.2509340836754463, "grad_norm": 2.2298221588134766, "learning_rate": 0.0001, "loss": 3.799461603164673, "step": 680 }, { "epoch": 0.2513031043867337, "grad_norm": 1.9469724893569946, "learning_rate": 0.0001, "loss": 3.9451560974121094, "step": 681 }, { "epoch": 0.2516721250980211, "grad_norm": 1.773236632347107, "learning_rate": 0.0001, "loss": 4.007401943206787, "step": 682 }, { "epoch": 0.25204114580930853, "grad_norm": 1.9999133348464966, "learning_rate": 0.0001, "loss": 3.8146023750305176, "step": 683 }, { "epoch": 0.25241016652059595, "grad_norm": 1.7420272827148438, "learning_rate": 0.0001, "loss": 3.947847366333008, "step": 684 }, { "epoch": 0.25277918723188336, "grad_norm": 1.816912055015564, "learning_rate": 0.0001, "loss": 4.101921081542969, "step": 685 }, { "epoch": 0.25314820794317083, "grad_norm": 1.8957693576812744, "learning_rate": 0.0001, "loss": 3.8831582069396973, "step": 686 }, { "epoch": 0.25351722865445825, "grad_norm": 2.2412970066070557, "learning_rate": 0.0001, "loss": 4.020118236541748, "step": 687 }, { "epoch": 0.25388624936574566, "grad_norm": 2.245518684387207, "learning_rate": 0.0001, "loss": 4.086484909057617, "step": 688 }, { "epoch": 0.2542552700770331, "grad_norm": 1.9446709156036377, "learning_rate": 0.0001, "loss": 3.790587902069092, "step": 689 }, { "epoch": 0.2546242907883205, "grad_norm": 1.8118345737457275, "learning_rate": 0.0001, "loss": 3.856565237045288, "step": 690 }, { "epoch": 0.2549933114996079, "grad_norm": 2.1692206859588623, "learning_rate": 0.0001, "loss": 4.035465717315674, "step": 691 }, { "epoch": 0.2553623322108953, "grad_norm": 1.748687505722046, "learning_rate": 0.0001, "loss": 4.366494178771973, "step": 692 }, { "epoch": 0.25573135292218274, "grad_norm": 1.870940923690796, "learning_rate": 0.0001, "loss": 3.9635026454925537, "step": 693 }, { "epoch": 0.25610037363347016, "grad_norm": 1.7080210447311401, "learning_rate": 0.0001, "loss": 3.9478514194488525, "step": 694 }, { "epoch": 0.2564693943447576, "grad_norm": 2.287992477416992, "learning_rate": 0.0001, "loss": 3.604180097579956, "step": 695 }, { "epoch": 0.25683841505604504, "grad_norm": 1.6620538234710693, "learning_rate": 0.0001, "loss": 3.694093942642212, "step": 696 }, { "epoch": 0.25720743576733246, "grad_norm": 2.257546901702881, "learning_rate": 0.0001, "loss": 3.9968767166137695, "step": 697 }, { "epoch": 0.2575764564786199, "grad_norm": 3.5476861000061035, "learning_rate": 0.0001, "loss": 3.4779582023620605, "step": 698 }, { "epoch": 0.2579454771899073, "grad_norm": 2.1635072231292725, "learning_rate": 0.0001, "loss": 4.081256866455078, "step": 699 }, { "epoch": 0.2583144979011947, "grad_norm": 1.9570531845092773, "learning_rate": 0.0001, "loss": 4.1506123542785645, "step": 700 }, { "epoch": 0.2586835186124821, "grad_norm": 2.0834312438964844, "learning_rate": 0.0001, "loss": 3.844961404800415, "step": 701 }, { "epoch": 0.25905253932376954, "grad_norm": 1.7518153190612793, "learning_rate": 0.0001, "loss": 4.191310405731201, "step": 702 }, { "epoch": 0.25942156003505695, "grad_norm": 1.8884351253509521, "learning_rate": 0.0001, "loss": 3.9044816493988037, "step": 703 }, { "epoch": 0.25979058074634437, "grad_norm": 2.1218340396881104, "learning_rate": 0.0001, "loss": 3.7818799018859863, "step": 704 }, { "epoch": 0.2601596014576318, "grad_norm": 2.3949546813964844, "learning_rate": 0.0001, "loss": 3.93882155418396, "step": 705 }, { "epoch": 0.26052862216891925, "grad_norm": 1.8599239587783813, "learning_rate": 0.0001, "loss": 4.148632526397705, "step": 706 }, { "epoch": 0.26089764288020667, "grad_norm": 2.2492785453796387, "learning_rate": 0.0001, "loss": 3.665086507797241, "step": 707 }, { "epoch": 0.2612666635914941, "grad_norm": 2.4720396995544434, "learning_rate": 0.0001, "loss": 4.0655364990234375, "step": 708 }, { "epoch": 0.2616356843027815, "grad_norm": 1.889915943145752, "learning_rate": 0.0001, "loss": 4.048415660858154, "step": 709 }, { "epoch": 0.2620047050140689, "grad_norm": 2.4719953536987305, "learning_rate": 0.0001, "loss": 3.8619840145111084, "step": 710 }, { "epoch": 0.26237372572535633, "grad_norm": 1.9330310821533203, "learning_rate": 0.0001, "loss": 4.062207221984863, "step": 711 }, { "epoch": 0.26274274643664375, "grad_norm": 2.2420597076416016, "learning_rate": 0.0001, "loss": 3.8035707473754883, "step": 712 }, { "epoch": 0.26311176714793116, "grad_norm": 1.9284963607788086, "learning_rate": 0.0001, "loss": 3.9348812103271484, "step": 713 }, { "epoch": 0.2634807878592186, "grad_norm": 2.1408097743988037, "learning_rate": 0.0001, "loss": 3.5326733589172363, "step": 714 }, { "epoch": 0.263849808570506, "grad_norm": 2.412616729736328, "learning_rate": 0.0001, "loss": 4.056027412414551, "step": 715 }, { "epoch": 0.26421882928179347, "grad_norm": 2.032498836517334, "learning_rate": 0.0001, "loss": 3.840789318084717, "step": 716 }, { "epoch": 0.2645878499930809, "grad_norm": 1.7631964683532715, "learning_rate": 0.0001, "loss": 4.013442039489746, "step": 717 }, { "epoch": 0.2649568707043683, "grad_norm": 1.8257538080215454, "learning_rate": 0.0001, "loss": 3.9977736473083496, "step": 718 }, { "epoch": 0.2653258914156557, "grad_norm": 1.7667533159255981, "learning_rate": 0.0001, "loss": 3.76548433303833, "step": 719 }, { "epoch": 0.26569491212694313, "grad_norm": 2.3263771533966064, "learning_rate": 0.0001, "loss": 4.297848224639893, "step": 720 }, { "epoch": 0.26606393283823054, "grad_norm": 2.0309314727783203, "learning_rate": 0.0001, "loss": 4.072471618652344, "step": 721 }, { "epoch": 0.26643295354951796, "grad_norm": 1.7921390533447266, "learning_rate": 0.0001, "loss": 3.7482833862304688, "step": 722 }, { "epoch": 0.2668019742608054, "grad_norm": 2.0791127681732178, "learning_rate": 0.0001, "loss": 3.7967019081115723, "step": 723 }, { "epoch": 0.2671709949720928, "grad_norm": 2.238356828689575, "learning_rate": 0.0001, "loss": 4.120265007019043, "step": 724 }, { "epoch": 0.2675400156833802, "grad_norm": 2.021831750869751, "learning_rate": 0.0001, "loss": 3.8354411125183105, "step": 725 }, { "epoch": 0.2679090363946677, "grad_norm": 2.2298364639282227, "learning_rate": 0.0001, "loss": 4.065906524658203, "step": 726 }, { "epoch": 0.2682780571059551, "grad_norm": 1.9646517038345337, "learning_rate": 0.0001, "loss": 3.930744171142578, "step": 727 }, { "epoch": 0.2686470778172425, "grad_norm": 1.642250895500183, "learning_rate": 0.0001, "loss": 3.8111205101013184, "step": 728 }, { "epoch": 0.2690160985285299, "grad_norm": 1.8207128047943115, "learning_rate": 0.0001, "loss": 3.982748031616211, "step": 729 }, { "epoch": 0.26938511923981734, "grad_norm": 1.8667646646499634, "learning_rate": 0.0001, "loss": 3.8368825912475586, "step": 730 }, { "epoch": 0.26975413995110475, "grad_norm": 2.267573833465576, "learning_rate": 0.0001, "loss": 3.8946995735168457, "step": 731 }, { "epoch": 0.27012316066239217, "grad_norm": 2.065119743347168, "learning_rate": 0.0001, "loss": 3.72292423248291, "step": 732 }, { "epoch": 0.2704921813736796, "grad_norm": 2.073775053024292, "learning_rate": 0.0001, "loss": 3.938802719116211, "step": 733 }, { "epoch": 0.270861202084967, "grad_norm": 1.9956557750701904, "learning_rate": 0.0001, "loss": 3.7231907844543457, "step": 734 }, { "epoch": 0.2712302227962544, "grad_norm": 1.6602935791015625, "learning_rate": 0.0001, "loss": 3.8295085430145264, "step": 735 }, { "epoch": 0.2715992435075419, "grad_norm": 1.9623558521270752, "learning_rate": 0.0001, "loss": 3.7618637084960938, "step": 736 }, { "epoch": 0.2719682642188293, "grad_norm": 1.961132287979126, "learning_rate": 0.0001, "loss": 3.813999652862549, "step": 737 }, { "epoch": 0.2723372849301167, "grad_norm": 1.8608838319778442, "learning_rate": 0.0001, "loss": 3.8973870277404785, "step": 738 }, { "epoch": 0.27270630564140413, "grad_norm": 1.9059594869613647, "learning_rate": 0.0001, "loss": 3.350216865539551, "step": 739 }, { "epoch": 0.27307532635269155, "grad_norm": 1.9583384990692139, "learning_rate": 0.0001, "loss": 4.086672782897949, "step": 740 }, { "epoch": 0.27344434706397897, "grad_norm": 1.7702605724334717, "learning_rate": 0.0001, "loss": 3.8397998809814453, "step": 741 }, { "epoch": 0.2738133677752664, "grad_norm": 1.857429027557373, "learning_rate": 0.0001, "loss": 3.7530627250671387, "step": 742 }, { "epoch": 0.2741823884865538, "grad_norm": 1.9428619146347046, "learning_rate": 0.0001, "loss": 3.851217031478882, "step": 743 }, { "epoch": 0.2745514091978412, "grad_norm": 3.188161849975586, "learning_rate": 0.0001, "loss": 3.9760873317718506, "step": 744 }, { "epoch": 0.2749204299091286, "grad_norm": 1.7686970233917236, "learning_rate": 0.0001, "loss": 3.7251675128936768, "step": 745 }, { "epoch": 0.2752894506204161, "grad_norm": 1.8680338859558105, "learning_rate": 0.0001, "loss": 3.719837188720703, "step": 746 }, { "epoch": 0.2756584713317035, "grad_norm": 4.727423667907715, "learning_rate": 0.0001, "loss": 4.332064151763916, "step": 747 }, { "epoch": 0.27602749204299093, "grad_norm": 1.9841293096542358, "learning_rate": 0.0001, "loss": 4.058561325073242, "step": 748 }, { "epoch": 0.27639651275427835, "grad_norm": 1.894529104232788, "learning_rate": 0.0001, "loss": 4.011638641357422, "step": 749 }, { "epoch": 0.27676553346556576, "grad_norm": 1.7220805883407593, "learning_rate": 0.0001, "loss": 4.114858150482178, "step": 750 }, { "epoch": 0.2771345541768532, "grad_norm": 1.9157108068466187, "learning_rate": 0.0001, "loss": 3.9608776569366455, "step": 751 }, { "epoch": 0.2775035748881406, "grad_norm": 3.2951736450195312, "learning_rate": 0.0001, "loss": 3.9270167350769043, "step": 752 }, { "epoch": 0.277872595599428, "grad_norm": 2.265681505203247, "learning_rate": 0.0001, "loss": 3.562108039855957, "step": 753 }, { "epoch": 0.2782416163107154, "grad_norm": 2.5440428256988525, "learning_rate": 0.0001, "loss": 3.4305877685546875, "step": 754 }, { "epoch": 0.27861063702200284, "grad_norm": 2.11076283454895, "learning_rate": 0.0001, "loss": 4.145153999328613, "step": 755 }, { "epoch": 0.27897965773329025, "grad_norm": 2.8526527881622314, "learning_rate": 0.0001, "loss": 3.981480598449707, "step": 756 }, { "epoch": 0.2793486784445777, "grad_norm": 2.2282848358154297, "learning_rate": 0.0001, "loss": 3.3889548778533936, "step": 757 }, { "epoch": 0.27971769915586514, "grad_norm": 2.0190341472625732, "learning_rate": 0.0001, "loss": 3.722273349761963, "step": 758 }, { "epoch": 0.28008671986715256, "grad_norm": 1.8118723630905151, "learning_rate": 0.0001, "loss": 3.598466634750366, "step": 759 }, { "epoch": 0.28045574057843997, "grad_norm": 2.110595703125, "learning_rate": 0.0001, "loss": 3.7021384239196777, "step": 760 }, { "epoch": 0.2808247612897274, "grad_norm": 2.296208381652832, "learning_rate": 0.0001, "loss": 3.875143051147461, "step": 761 }, { "epoch": 0.2811937820010148, "grad_norm": 1.8150382041931152, "learning_rate": 0.0001, "loss": 3.756319761276245, "step": 762 }, { "epoch": 0.2815628027123022, "grad_norm": 2.0273423194885254, "learning_rate": 0.0001, "loss": 3.5769379138946533, "step": 763 }, { "epoch": 0.28193182342358963, "grad_norm": 1.7625868320465088, "learning_rate": 0.0001, "loss": 3.7091686725616455, "step": 764 }, { "epoch": 0.28230084413487705, "grad_norm": 1.7284501791000366, "learning_rate": 0.0001, "loss": 3.5987820625305176, "step": 765 }, { "epoch": 0.28266986484616446, "grad_norm": 1.8203482627868652, "learning_rate": 0.0001, "loss": 3.814981698989868, "step": 766 }, { "epoch": 0.28303888555745194, "grad_norm": 2.4606571197509766, "learning_rate": 0.0001, "loss": 3.69838809967041, "step": 767 }, { "epoch": 0.28340790626873935, "grad_norm": 1.9143778085708618, "learning_rate": 0.0001, "loss": 3.6593539714813232, "step": 768 }, { "epoch": 0.28377692698002677, "grad_norm": 1.8607871532440186, "learning_rate": 0.0001, "loss": 3.706435441970825, "step": 769 }, { "epoch": 0.2841459476913142, "grad_norm": 1.801198959350586, "learning_rate": 0.0001, "loss": 3.9967586994171143, "step": 770 }, { "epoch": 0.2845149684026016, "grad_norm": 1.9508602619171143, "learning_rate": 0.0001, "loss": 4.045716762542725, "step": 771 }, { "epoch": 0.284883989113889, "grad_norm": 1.8430317640304565, "learning_rate": 0.0001, "loss": 3.9594106674194336, "step": 772 }, { "epoch": 0.28525300982517643, "grad_norm": 2.1014292240142822, "learning_rate": 0.0001, "loss": 3.7480573654174805, "step": 773 }, { "epoch": 0.28562203053646384, "grad_norm": 2.1231205463409424, "learning_rate": 0.0001, "loss": 3.96742582321167, "step": 774 }, { "epoch": 0.28599105124775126, "grad_norm": 1.7807960510253906, "learning_rate": 0.0001, "loss": 3.686154842376709, "step": 775 }, { "epoch": 0.2863600719590387, "grad_norm": 1.891921877861023, "learning_rate": 0.0001, "loss": 3.6867451667785645, "step": 776 }, { "epoch": 0.28672909267032615, "grad_norm": 2.125894784927368, "learning_rate": 0.0001, "loss": 3.893233299255371, "step": 777 }, { "epoch": 0.28709811338161356, "grad_norm": 1.9936643838882446, "learning_rate": 0.0001, "loss": 3.717493772506714, "step": 778 }, { "epoch": 0.287467134092901, "grad_norm": 1.9470746517181396, "learning_rate": 0.0001, "loss": 3.8600313663482666, "step": 779 }, { "epoch": 0.2878361548041884, "grad_norm": 1.8602983951568604, "learning_rate": 0.0001, "loss": 3.70097017288208, "step": 780 }, { "epoch": 0.2882051755154758, "grad_norm": 2.0043935775756836, "learning_rate": 0.0001, "loss": 3.5539088249206543, "step": 781 }, { "epoch": 0.2885741962267632, "grad_norm": 1.8311423063278198, "learning_rate": 0.0001, "loss": 3.9076876640319824, "step": 782 }, { "epoch": 0.28894321693805064, "grad_norm": 1.6583877801895142, "learning_rate": 0.0001, "loss": 3.9241816997528076, "step": 783 }, { "epoch": 0.28931223764933806, "grad_norm": 1.9217376708984375, "learning_rate": 0.0001, "loss": 3.4714879989624023, "step": 784 }, { "epoch": 0.28968125836062547, "grad_norm": 2.757667064666748, "learning_rate": 0.0001, "loss": 4.184150218963623, "step": 785 }, { "epoch": 0.2900502790719129, "grad_norm": 1.8339331150054932, "learning_rate": 0.0001, "loss": 3.9348061084747314, "step": 786 }, { "epoch": 0.29041929978320036, "grad_norm": 2.1208674907684326, "learning_rate": 0.0001, "loss": 3.872300624847412, "step": 787 }, { "epoch": 0.2907883204944878, "grad_norm": 1.9651638269424438, "learning_rate": 0.0001, "loss": 3.7166500091552734, "step": 788 }, { "epoch": 0.2911573412057752, "grad_norm": 1.7782893180847168, "learning_rate": 0.0001, "loss": 3.5736327171325684, "step": 789 }, { "epoch": 0.2915263619170626, "grad_norm": 1.8201782703399658, "learning_rate": 0.0001, "loss": 3.7780208587646484, "step": 790 }, { "epoch": 0.29189538262835, "grad_norm": 1.97230064868927, "learning_rate": 0.0001, "loss": 4.034017562866211, "step": 791 }, { "epoch": 0.29226440333963744, "grad_norm": 1.8389756679534912, "learning_rate": 0.0001, "loss": 3.8383772373199463, "step": 792 }, { "epoch": 0.29263342405092485, "grad_norm": 1.6989628076553345, "learning_rate": 0.0001, "loss": 3.8918330669403076, "step": 793 }, { "epoch": 0.29300244476221227, "grad_norm": 1.7311633825302124, "learning_rate": 0.0001, "loss": 3.9464480876922607, "step": 794 }, { "epoch": 0.2933714654734997, "grad_norm": 1.7128499746322632, "learning_rate": 0.0001, "loss": 3.7909626960754395, "step": 795 }, { "epoch": 0.2937404861847871, "grad_norm": 1.6355345249176025, "learning_rate": 0.0001, "loss": 3.874768018722534, "step": 796 }, { "epoch": 0.29410950689607457, "grad_norm": 2.4454221725463867, "learning_rate": 0.0001, "loss": 4.041829586029053, "step": 797 }, { "epoch": 0.294478527607362, "grad_norm": 1.5545457601547241, "learning_rate": 0.0001, "loss": 3.8092539310455322, "step": 798 }, { "epoch": 0.2948475483186494, "grad_norm": 1.7109586000442505, "learning_rate": 0.0001, "loss": 3.7526321411132812, "step": 799 }, { "epoch": 0.2952165690299368, "grad_norm": 2.030221462249756, "learning_rate": 0.0001, "loss": 4.1812052726745605, "step": 800 }, { "epoch": 0.29558558974122423, "grad_norm": 2.7777583599090576, "learning_rate": 0.0001, "loss": 3.9486491680145264, "step": 801 }, { "epoch": 0.29595461045251165, "grad_norm": 1.742997646331787, "learning_rate": 0.0001, "loss": 3.8623580932617188, "step": 802 }, { "epoch": 0.29632363116379906, "grad_norm": 1.7032355070114136, "learning_rate": 0.0001, "loss": 3.8116557598114014, "step": 803 }, { "epoch": 0.2966926518750865, "grad_norm": 2.0460309982299805, "learning_rate": 0.0001, "loss": 3.848754405975342, "step": 804 }, { "epoch": 0.2970616725863739, "grad_norm": 2.204787492752075, "learning_rate": 0.0001, "loss": 4.090125560760498, "step": 805 }, { "epoch": 0.2974306932976613, "grad_norm": 1.7479274272918701, "learning_rate": 0.0001, "loss": 3.6636033058166504, "step": 806 }, { "epoch": 0.2977997140089488, "grad_norm": 1.8795173168182373, "learning_rate": 0.0001, "loss": 3.6704483032226562, "step": 807 }, { "epoch": 0.2981687347202362, "grad_norm": 1.8553147315979004, "learning_rate": 0.0001, "loss": 3.73581600189209, "step": 808 }, { "epoch": 0.2985377554315236, "grad_norm": 1.4317312240600586, "learning_rate": 0.0001, "loss": 3.5331602096557617, "step": 809 }, { "epoch": 0.298906776142811, "grad_norm": 1.722379446029663, "learning_rate": 0.0001, "loss": 3.584402561187744, "step": 810 }, { "epoch": 0.29927579685409844, "grad_norm": 2.114041328430176, "learning_rate": 0.0001, "loss": 3.937901258468628, "step": 811 }, { "epoch": 0.29964481756538586, "grad_norm": 2.0994956493377686, "learning_rate": 0.0001, "loss": 3.636784076690674, "step": 812 }, { "epoch": 0.3000138382766733, "grad_norm": 2.4144608974456787, "learning_rate": 0.0001, "loss": 3.838789701461792, "step": 813 }, { "epoch": 0.3003828589879607, "grad_norm": 1.8704757690429688, "learning_rate": 0.0001, "loss": 3.7378387451171875, "step": 814 }, { "epoch": 0.3007518796992481, "grad_norm": 1.9785248041152954, "learning_rate": 0.0001, "loss": 3.7420191764831543, "step": 815 }, { "epoch": 0.3011209004105355, "grad_norm": 1.805633783340454, "learning_rate": 0.0001, "loss": 3.6539623737335205, "step": 816 }, { "epoch": 0.30148992112182293, "grad_norm": 1.926710844039917, "learning_rate": 0.0001, "loss": 4.070534706115723, "step": 817 }, { "epoch": 0.3018589418331104, "grad_norm": 1.697633147239685, "learning_rate": 0.0001, "loss": 3.381309986114502, "step": 818 }, { "epoch": 0.3022279625443978, "grad_norm": 2.511188268661499, "learning_rate": 0.0001, "loss": 3.8093338012695312, "step": 819 }, { "epoch": 0.30259698325568524, "grad_norm": 1.627577304840088, "learning_rate": 0.0001, "loss": 3.5759048461914062, "step": 820 }, { "epoch": 0.30296600396697265, "grad_norm": 2.0722548961639404, "learning_rate": 0.0001, "loss": 3.7673254013061523, "step": 821 }, { "epoch": 0.30333502467826007, "grad_norm": 2.092536687850952, "learning_rate": 0.0001, "loss": 3.8161582946777344, "step": 822 }, { "epoch": 0.3037040453895475, "grad_norm": 1.985236406326294, "learning_rate": 0.0001, "loss": 3.6728384494781494, "step": 823 }, { "epoch": 0.3040730661008349, "grad_norm": 2.135077714920044, "learning_rate": 0.0001, "loss": 3.6908416748046875, "step": 824 }, { "epoch": 0.3044420868121223, "grad_norm": 2.1780848503112793, "learning_rate": 0.0001, "loss": 3.6705384254455566, "step": 825 }, { "epoch": 0.30481110752340973, "grad_norm": 2.005251169204712, "learning_rate": 0.0001, "loss": 3.790651798248291, "step": 826 }, { "epoch": 0.30518012823469715, "grad_norm": 1.9229953289031982, "learning_rate": 0.0001, "loss": 3.68874454498291, "step": 827 }, { "epoch": 0.3055491489459846, "grad_norm": 2.0374631881713867, "learning_rate": 0.0001, "loss": 4.054494857788086, "step": 828 }, { "epoch": 0.30591816965727203, "grad_norm": 1.770068645477295, "learning_rate": 0.0001, "loss": 3.5029234886169434, "step": 829 }, { "epoch": 0.30628719036855945, "grad_norm": 1.9545096158981323, "learning_rate": 0.0001, "loss": 3.672738552093506, "step": 830 }, { "epoch": 0.30665621107984686, "grad_norm": 1.8246434926986694, "learning_rate": 0.0001, "loss": 3.822554349899292, "step": 831 }, { "epoch": 0.3070252317911343, "grad_norm": 2.029484510421753, "learning_rate": 0.0001, "loss": 3.7952938079833984, "step": 832 }, { "epoch": 0.3073942525024217, "grad_norm": 1.7776564359664917, "learning_rate": 0.0001, "loss": 3.626302480697632, "step": 833 }, { "epoch": 0.3077632732137091, "grad_norm": 1.9499139785766602, "learning_rate": 0.0001, "loss": 3.8773744106292725, "step": 834 }, { "epoch": 0.3081322939249965, "grad_norm": 1.7793817520141602, "learning_rate": 0.0001, "loss": 3.5810489654541016, "step": 835 }, { "epoch": 0.30850131463628394, "grad_norm": 2.1008458137512207, "learning_rate": 0.0001, "loss": 3.7431108951568604, "step": 836 }, { "epoch": 0.30887033534757136, "grad_norm": 1.6641654968261719, "learning_rate": 0.0001, "loss": 3.714667320251465, "step": 837 }, { "epoch": 0.3092393560588588, "grad_norm": 2.273031234741211, "learning_rate": 0.0001, "loss": 3.804274559020996, "step": 838 }, { "epoch": 0.30960837677014624, "grad_norm": 1.7359979152679443, "learning_rate": 0.0001, "loss": 3.8806774616241455, "step": 839 }, { "epoch": 0.30997739748143366, "grad_norm": 1.8702484369277954, "learning_rate": 0.0001, "loss": 3.7494711875915527, "step": 840 }, { "epoch": 0.3103464181927211, "grad_norm": 2.0235681533813477, "learning_rate": 0.0001, "loss": 3.859908103942871, "step": 841 }, { "epoch": 0.3107154389040085, "grad_norm": 1.7331302165985107, "learning_rate": 0.0001, "loss": 3.3850371837615967, "step": 842 }, { "epoch": 0.3110844596152959, "grad_norm": 1.866202712059021, "learning_rate": 0.0001, "loss": 4.035645484924316, "step": 843 }, { "epoch": 0.3114534803265833, "grad_norm": 1.813595175743103, "learning_rate": 0.0001, "loss": 3.852121591567993, "step": 844 }, { "epoch": 0.31182250103787074, "grad_norm": 1.8837628364562988, "learning_rate": 0.0001, "loss": 3.7275757789611816, "step": 845 }, { "epoch": 0.31219152174915815, "grad_norm": 1.8290719985961914, "learning_rate": 0.0001, "loss": 3.3215506076812744, "step": 846 }, { "epoch": 0.31256054246044557, "grad_norm": 1.6821497678756714, "learning_rate": 0.0001, "loss": 3.9793262481689453, "step": 847 }, { "epoch": 0.31292956317173304, "grad_norm": 2.5647542476654053, "learning_rate": 0.0001, "loss": 3.988278865814209, "step": 848 }, { "epoch": 0.31329858388302045, "grad_norm": 1.5824579000473022, "learning_rate": 0.0001, "loss": 3.4897401332855225, "step": 849 }, { "epoch": 0.31366760459430787, "grad_norm": 1.706549882888794, "learning_rate": 0.0001, "loss": 3.7646942138671875, "step": 850 }, { "epoch": 0.3140366253055953, "grad_norm": 1.542718768119812, "learning_rate": 0.0001, "loss": 3.734578847885132, "step": 851 }, { "epoch": 0.3144056460168827, "grad_norm": 1.674017071723938, "learning_rate": 0.0001, "loss": 3.6028411388397217, "step": 852 }, { "epoch": 0.3147746667281701, "grad_norm": 2.1733975410461426, "learning_rate": 0.0001, "loss": 3.9048941135406494, "step": 853 }, { "epoch": 0.31514368743945753, "grad_norm": 1.974860429763794, "learning_rate": 0.0001, "loss": 3.480330228805542, "step": 854 }, { "epoch": 0.31551270815074495, "grad_norm": 2.0926570892333984, "learning_rate": 0.0001, "loss": 3.58974552154541, "step": 855 }, { "epoch": 0.31588172886203236, "grad_norm": 1.7155181169509888, "learning_rate": 0.0001, "loss": 3.799851179122925, "step": 856 }, { "epoch": 0.3162507495733198, "grad_norm": 1.8020222187042236, "learning_rate": 0.0001, "loss": 3.9360482692718506, "step": 857 }, { "epoch": 0.31661977028460725, "grad_norm": 2.011735677719116, "learning_rate": 0.0001, "loss": 3.950687885284424, "step": 858 }, { "epoch": 0.31698879099589466, "grad_norm": 1.77201247215271, "learning_rate": 0.0001, "loss": 3.536175012588501, "step": 859 }, { "epoch": 0.3173578117071821, "grad_norm": 1.8835783004760742, "learning_rate": 0.0001, "loss": 3.696162700653076, "step": 860 }, { "epoch": 0.3177268324184695, "grad_norm": 1.772087574005127, "learning_rate": 0.0001, "loss": 3.6627166271209717, "step": 861 }, { "epoch": 0.3180958531297569, "grad_norm": 1.7648016214370728, "learning_rate": 0.0001, "loss": 3.8221614360809326, "step": 862 }, { "epoch": 0.3184648738410443, "grad_norm": 2.0680150985717773, "learning_rate": 0.0001, "loss": 4.0188493728637695, "step": 863 }, { "epoch": 0.31883389455233174, "grad_norm": 1.7627791166305542, "learning_rate": 0.0001, "loss": 3.635226011276245, "step": 864 }, { "epoch": 0.31920291526361916, "grad_norm": 1.9791538715362549, "learning_rate": 0.0001, "loss": 3.557415246963501, "step": 865 }, { "epoch": 0.3195719359749066, "grad_norm": 1.7755398750305176, "learning_rate": 0.0001, "loss": 3.798553705215454, "step": 866 }, { "epoch": 0.319940956686194, "grad_norm": 2.006571054458618, "learning_rate": 0.0001, "loss": 3.771108627319336, "step": 867 }, { "epoch": 0.32030997739748146, "grad_norm": 1.7845039367675781, "learning_rate": 0.0001, "loss": 3.7619147300720215, "step": 868 }, { "epoch": 0.3206789981087689, "grad_norm": 1.94892418384552, "learning_rate": 0.0001, "loss": 3.7109427452087402, "step": 869 }, { "epoch": 0.3210480188200563, "grad_norm": 1.978912353515625, "learning_rate": 0.0001, "loss": 3.640345573425293, "step": 870 }, { "epoch": 0.3214170395313437, "grad_norm": 1.7518224716186523, "learning_rate": 0.0001, "loss": 4.0331807136535645, "step": 871 }, { "epoch": 0.3217860602426311, "grad_norm": 1.7543238401412964, "learning_rate": 0.0001, "loss": 3.4129204750061035, "step": 872 }, { "epoch": 0.32215508095391854, "grad_norm": 1.6404772996902466, "learning_rate": 0.0001, "loss": 3.62764835357666, "step": 873 }, { "epoch": 0.32252410166520595, "grad_norm": 1.7205040454864502, "learning_rate": 0.0001, "loss": 3.2731800079345703, "step": 874 }, { "epoch": 0.32289312237649337, "grad_norm": 2.005343198776245, "learning_rate": 0.0001, "loss": 3.9949886798858643, "step": 875 }, { "epoch": 0.3232621430877808, "grad_norm": 1.806545376777649, "learning_rate": 0.0001, "loss": 3.6255760192871094, "step": 876 }, { "epoch": 0.3236311637990682, "grad_norm": 2.3675615787506104, "learning_rate": 0.0001, "loss": 3.56473708152771, "step": 877 }, { "epoch": 0.32400018451035567, "grad_norm": 2.2366559505462646, "learning_rate": 0.0001, "loss": 3.7560157775878906, "step": 878 }, { "epoch": 0.3243692052216431, "grad_norm": 1.8035284280776978, "learning_rate": 0.0001, "loss": 3.6027302742004395, "step": 879 }, { "epoch": 0.3247382259329305, "grad_norm": 1.9980396032333374, "learning_rate": 0.0001, "loss": 3.5838475227355957, "step": 880 }, { "epoch": 0.3251072466442179, "grad_norm": 1.575914978981018, "learning_rate": 0.0001, "loss": 3.6749353408813477, "step": 881 }, { "epoch": 0.32547626735550533, "grad_norm": 2.236276149749756, "learning_rate": 0.0001, "loss": 3.5917420387268066, "step": 882 }, { "epoch": 0.32584528806679275, "grad_norm": 1.8865281343460083, "learning_rate": 0.0001, "loss": 3.7568204402923584, "step": 883 }, { "epoch": 0.32621430877808016, "grad_norm": 2.1691057682037354, "learning_rate": 0.0001, "loss": 3.7836689949035645, "step": 884 }, { "epoch": 0.3265833294893676, "grad_norm": 1.5681935548782349, "learning_rate": 0.0001, "loss": 3.668919324874878, "step": 885 }, { "epoch": 0.326952350200655, "grad_norm": 1.8908230066299438, "learning_rate": 0.0001, "loss": 3.6291942596435547, "step": 886 }, { "epoch": 0.3273213709119424, "grad_norm": 1.7574917078018188, "learning_rate": 0.0001, "loss": 3.777817487716675, "step": 887 }, { "epoch": 0.3276903916232298, "grad_norm": 1.753098726272583, "learning_rate": 0.0001, "loss": 3.6071789264678955, "step": 888 }, { "epoch": 0.3280594123345173, "grad_norm": 1.7223573923110962, "learning_rate": 0.0001, "loss": 3.3281450271606445, "step": 889 }, { "epoch": 0.3284284330458047, "grad_norm": 1.7684091329574585, "learning_rate": 0.0001, "loss": 3.5951995849609375, "step": 890 }, { "epoch": 0.32879745375709213, "grad_norm": 1.7212754487991333, "learning_rate": 0.0001, "loss": 3.5466182231903076, "step": 891 }, { "epoch": 0.32916647446837954, "grad_norm": 1.7892823219299316, "learning_rate": 0.0001, "loss": 3.7357468605041504, "step": 892 }, { "epoch": 0.32953549517966696, "grad_norm": 1.5858311653137207, "learning_rate": 0.0001, "loss": 4.225461483001709, "step": 893 }, { "epoch": 0.3299045158909544, "grad_norm": 1.7122700214385986, "learning_rate": 0.0001, "loss": 3.4520514011383057, "step": 894 }, { "epoch": 0.3302735366022418, "grad_norm": 1.884587049484253, "learning_rate": 0.0001, "loss": 3.634110927581787, "step": 895 }, { "epoch": 0.3306425573135292, "grad_norm": 1.9494415521621704, "learning_rate": 0.0001, "loss": 3.80303955078125, "step": 896 }, { "epoch": 0.3310115780248166, "grad_norm": 1.9596128463745117, "learning_rate": 0.0001, "loss": 4.176259517669678, "step": 897 }, { "epoch": 0.33138059873610404, "grad_norm": 2.0338494777679443, "learning_rate": 0.0001, "loss": 3.864532470703125, "step": 898 }, { "epoch": 0.3317496194473915, "grad_norm": 2.142646551132202, "learning_rate": 0.0001, "loss": 3.832646369934082, "step": 899 }, { "epoch": 0.3321186401586789, "grad_norm": 1.5813355445861816, "learning_rate": 0.0001, "loss": 3.697361469268799, "step": 900 }, { "epoch": 0.33248766086996634, "grad_norm": 1.7064385414123535, "learning_rate": 0.0001, "loss": 3.5419235229492188, "step": 901 }, { "epoch": 0.33285668158125375, "grad_norm": 2.084634304046631, "learning_rate": 0.0001, "loss": 3.5576624870300293, "step": 902 }, { "epoch": 0.33322570229254117, "grad_norm": 1.9081051349639893, "learning_rate": 0.0001, "loss": 3.612410545349121, "step": 903 }, { "epoch": 0.3335947230038286, "grad_norm": 2.011467456817627, "learning_rate": 0.0001, "loss": 3.783489227294922, "step": 904 }, { "epoch": 0.333963743715116, "grad_norm": 1.6679588556289673, "learning_rate": 0.0001, "loss": 3.5102577209472656, "step": 905 }, { "epoch": 0.3343327644264034, "grad_norm": 2.222921848297119, "learning_rate": 0.0001, "loss": 3.4336118698120117, "step": 906 }, { "epoch": 0.33470178513769083, "grad_norm": 1.8418184518814087, "learning_rate": 0.0001, "loss": 3.6948063373565674, "step": 907 }, { "epoch": 0.33507080584897825, "grad_norm": 1.7070863246917725, "learning_rate": 0.0001, "loss": 3.321864128112793, "step": 908 }, { "epoch": 0.3354398265602657, "grad_norm": 1.7238003015518188, "learning_rate": 0.0001, "loss": 3.6659138202667236, "step": 909 }, { "epoch": 0.33580884727155313, "grad_norm": 1.581603765487671, "learning_rate": 0.0001, "loss": 3.8663277626037598, "step": 910 }, { "epoch": 0.33617786798284055, "grad_norm": 1.7234193086624146, "learning_rate": 0.0001, "loss": 3.7296018600463867, "step": 911 }, { "epoch": 0.33654688869412797, "grad_norm": 2.2194244861602783, "learning_rate": 0.0001, "loss": 3.3964855670928955, "step": 912 }, { "epoch": 0.3369159094054154, "grad_norm": 1.9598308801651, "learning_rate": 0.0001, "loss": 3.765475273132324, "step": 913 }, { "epoch": 0.3372849301167028, "grad_norm": 1.6702812910079956, "learning_rate": 0.0001, "loss": 3.781707763671875, "step": 914 }, { "epoch": 0.3376539508279902, "grad_norm": 1.733283519744873, "learning_rate": 0.0001, "loss": 3.7726798057556152, "step": 915 }, { "epoch": 0.33802297153927763, "grad_norm": 1.8327610492706299, "learning_rate": 0.0001, "loss": 3.653780937194824, "step": 916 }, { "epoch": 0.33839199225056504, "grad_norm": 1.8619647026062012, "learning_rate": 0.0001, "loss": 3.7552051544189453, "step": 917 }, { "epoch": 0.33876101296185246, "grad_norm": 2.694669008255005, "learning_rate": 0.0001, "loss": 3.7501096725463867, "step": 918 }, { "epoch": 0.33913003367313993, "grad_norm": 2.195939302444458, "learning_rate": 0.0001, "loss": 3.667315721511841, "step": 919 }, { "epoch": 0.33949905438442735, "grad_norm": 1.9822533130645752, "learning_rate": 0.0001, "loss": 3.8069558143615723, "step": 920 }, { "epoch": 0.33986807509571476, "grad_norm": 1.9777230024337769, "learning_rate": 0.0001, "loss": 3.6377172470092773, "step": 921 }, { "epoch": 0.3402370958070022, "grad_norm": 2.054548978805542, "learning_rate": 0.0001, "loss": 3.6718833446502686, "step": 922 }, { "epoch": 0.3406061165182896, "grad_norm": 1.8605886697769165, "learning_rate": 0.0001, "loss": 3.5356335639953613, "step": 923 }, { "epoch": 0.340975137229577, "grad_norm": 1.7587581872940063, "learning_rate": 0.0001, "loss": 3.4171085357666016, "step": 924 }, { "epoch": 0.3413441579408644, "grad_norm": 2.9351863861083984, "learning_rate": 0.0001, "loss": 3.959594964981079, "step": 925 }, { "epoch": 0.34171317865215184, "grad_norm": 1.9823545217514038, "learning_rate": 0.0001, "loss": 4.213414669036865, "step": 926 }, { "epoch": 0.34208219936343925, "grad_norm": 1.8164136409759521, "learning_rate": 0.0001, "loss": 3.5082969665527344, "step": 927 }, { "epoch": 0.34245122007472667, "grad_norm": 1.7285102605819702, "learning_rate": 0.0001, "loss": 3.7684812545776367, "step": 928 }, { "epoch": 0.34282024078601414, "grad_norm": 1.8018592596054077, "learning_rate": 0.0001, "loss": 3.9759693145751953, "step": 929 }, { "epoch": 0.34318926149730156, "grad_norm": 1.7711527347564697, "learning_rate": 0.0001, "loss": 3.807732582092285, "step": 930 }, { "epoch": 0.34355828220858897, "grad_norm": 2.3091368675231934, "learning_rate": 0.0001, "loss": 3.699556589126587, "step": 931 }, { "epoch": 0.3439273029198764, "grad_norm": 1.5658090114593506, "learning_rate": 0.0001, "loss": 3.676739454269409, "step": 932 }, { "epoch": 0.3442963236311638, "grad_norm": 1.5197257995605469, "learning_rate": 0.0001, "loss": 3.849608898162842, "step": 933 }, { "epoch": 0.3446653443424512, "grad_norm": 2.346301794052124, "learning_rate": 0.0001, "loss": 3.628758430480957, "step": 934 }, { "epoch": 0.34503436505373863, "grad_norm": 1.594332218170166, "learning_rate": 0.0001, "loss": 3.8458902835845947, "step": 935 }, { "epoch": 0.34540338576502605, "grad_norm": 1.9343355894088745, "learning_rate": 0.0001, "loss": 3.525843620300293, "step": 936 }, { "epoch": 0.34577240647631347, "grad_norm": 1.5463449954986572, "learning_rate": 0.0001, "loss": 3.4548423290252686, "step": 937 }, { "epoch": 0.3461414271876009, "grad_norm": 1.697885513305664, "learning_rate": 0.0001, "loss": 3.6326701641082764, "step": 938 }, { "epoch": 0.34651044789888835, "grad_norm": 1.5783311128616333, "learning_rate": 0.0001, "loss": 3.6567530632019043, "step": 939 }, { "epoch": 0.34687946861017577, "grad_norm": 1.5128939151763916, "learning_rate": 0.0001, "loss": 3.761467933654785, "step": 940 }, { "epoch": 0.3472484893214632, "grad_norm": 1.8232477903366089, "learning_rate": 0.0001, "loss": 3.546701431274414, "step": 941 }, { "epoch": 0.3476175100327506, "grad_norm": 1.5385538339614868, "learning_rate": 0.0001, "loss": 3.8010096549987793, "step": 942 }, { "epoch": 0.347986530744038, "grad_norm": 2.058393955230713, "learning_rate": 0.0001, "loss": 4.060273170471191, "step": 943 }, { "epoch": 0.34835555145532543, "grad_norm": 1.7603133916854858, "learning_rate": 0.0001, "loss": 3.323798179626465, "step": 944 }, { "epoch": 0.34872457216661285, "grad_norm": 1.3818413019180298, "learning_rate": 0.0001, "loss": 3.4097721576690674, "step": 945 }, { "epoch": 0.34909359287790026, "grad_norm": 1.720240831375122, "learning_rate": 0.0001, "loss": 3.3414576053619385, "step": 946 }, { "epoch": 0.3494626135891877, "grad_norm": 1.4738744497299194, "learning_rate": 0.0001, "loss": 3.571500301361084, "step": 947 }, { "epoch": 0.3498316343004751, "grad_norm": 1.9796428680419922, "learning_rate": 0.0001, "loss": 3.8501107692718506, "step": 948 }, { "epoch": 0.35020065501176256, "grad_norm": 2.292299270629883, "learning_rate": 0.0001, "loss": 3.959177017211914, "step": 949 }, { "epoch": 0.35056967572305, "grad_norm": 2.1303646564483643, "learning_rate": 0.0001, "loss": 3.801318645477295, "step": 950 }, { "epoch": 0.3509386964343374, "grad_norm": 2.0312278270721436, "learning_rate": 0.0001, "loss": 3.9173123836517334, "step": 951 }, { "epoch": 0.3513077171456248, "grad_norm": 1.567550778388977, "learning_rate": 0.0001, "loss": 3.4361815452575684, "step": 952 }, { "epoch": 0.3516767378569122, "grad_norm": 1.5868910551071167, "learning_rate": 0.0001, "loss": 3.2824997901916504, "step": 953 }, { "epoch": 0.35204575856819964, "grad_norm": 1.831191897392273, "learning_rate": 0.0001, "loss": 3.7019565105438232, "step": 954 }, { "epoch": 0.35241477927948706, "grad_norm": 2.257370948791504, "learning_rate": 0.0001, "loss": 3.5339605808258057, "step": 955 }, { "epoch": 0.35278379999077447, "grad_norm": 1.5606753826141357, "learning_rate": 0.0001, "loss": 3.533111333847046, "step": 956 }, { "epoch": 0.3531528207020619, "grad_norm": 1.5263773202896118, "learning_rate": 0.0001, "loss": 4.072247505187988, "step": 957 }, { "epoch": 0.3535218414133493, "grad_norm": 1.782421350479126, "learning_rate": 0.0001, "loss": 3.332890748977661, "step": 958 }, { "epoch": 0.3538908621246367, "grad_norm": 2.099050521850586, "learning_rate": 0.0001, "loss": 3.7435617446899414, "step": 959 }, { "epoch": 0.3542598828359242, "grad_norm": 2.5168917179107666, "learning_rate": 0.0001, "loss": 3.849274158477783, "step": 960 }, { "epoch": 0.3546289035472116, "grad_norm": 1.784055233001709, "learning_rate": 0.0001, "loss": 3.565821409225464, "step": 961 }, { "epoch": 0.354997924258499, "grad_norm": 2.2097904682159424, "learning_rate": 0.0001, "loss": 3.2546889781951904, "step": 962 }, { "epoch": 0.35536694496978644, "grad_norm": 1.6301053762435913, "learning_rate": 0.0001, "loss": 3.742492914199829, "step": 963 }, { "epoch": 0.35573596568107385, "grad_norm": 1.8436990976333618, "learning_rate": 0.0001, "loss": 3.4069957733154297, "step": 964 }, { "epoch": 0.35610498639236127, "grad_norm": 1.644563913345337, "learning_rate": 0.0001, "loss": 3.6898012161254883, "step": 965 }, { "epoch": 0.3564740071036487, "grad_norm": 1.6858733892440796, "learning_rate": 0.0001, "loss": 3.5746512413024902, "step": 966 }, { "epoch": 0.3568430278149361, "grad_norm": 1.733216643333435, "learning_rate": 0.0001, "loss": 3.218445062637329, "step": 967 }, { "epoch": 0.3572120485262235, "grad_norm": 2.5151119232177734, "learning_rate": 0.0001, "loss": 3.34220027923584, "step": 968 }, { "epoch": 0.35758106923751093, "grad_norm": 1.9844415187835693, "learning_rate": 0.0001, "loss": 3.0783324241638184, "step": 969 }, { "epoch": 0.3579500899487984, "grad_norm": 1.8496888875961304, "learning_rate": 0.0001, "loss": 3.6318705081939697, "step": 970 }, { "epoch": 0.3583191106600858, "grad_norm": 1.589284896850586, "learning_rate": 0.0001, "loss": 3.626168727874756, "step": 971 }, { "epoch": 0.35868813137137323, "grad_norm": 2.818904399871826, "learning_rate": 0.0001, "loss": 3.3815014362335205, "step": 972 }, { "epoch": 0.35905715208266065, "grad_norm": 2.175112009048462, "learning_rate": 0.0001, "loss": 3.623976230621338, "step": 973 }, { "epoch": 0.35942617279394806, "grad_norm": 1.9865856170654297, "learning_rate": 0.0001, "loss": 3.607475519180298, "step": 974 }, { "epoch": 0.3597951935052355, "grad_norm": 1.7502002716064453, "learning_rate": 0.0001, "loss": 3.6265413761138916, "step": 975 }, { "epoch": 0.3601642142165229, "grad_norm": 2.3163676261901855, "learning_rate": 0.0001, "loss": 4.394412040710449, "step": 976 }, { "epoch": 0.3605332349278103, "grad_norm": 1.9160200357437134, "learning_rate": 0.0001, "loss": 3.6283273696899414, "step": 977 }, { "epoch": 0.3609022556390977, "grad_norm": 2.278754234313965, "learning_rate": 0.0001, "loss": 3.5777151584625244, "step": 978 }, { "epoch": 0.36127127635038514, "grad_norm": 2.136258363723755, "learning_rate": 0.0001, "loss": 3.6353437900543213, "step": 979 }, { "epoch": 0.3616402970616726, "grad_norm": 1.4851433038711548, "learning_rate": 0.0001, "loss": 3.5095326900482178, "step": 980 }, { "epoch": 0.36200931777296, "grad_norm": 1.5341440439224243, "learning_rate": 0.0001, "loss": 3.5404481887817383, "step": 981 }, { "epoch": 0.36237833848424744, "grad_norm": 1.7951269149780273, "learning_rate": 0.0001, "loss": 3.618901014328003, "step": 982 }, { "epoch": 0.36274735919553486, "grad_norm": 1.8107537031173706, "learning_rate": 0.0001, "loss": 3.598160743713379, "step": 983 }, { "epoch": 0.3631163799068223, "grad_norm": 2.1145143508911133, "learning_rate": 0.0001, "loss": 3.4968204498291016, "step": 984 }, { "epoch": 0.3634854006181097, "grad_norm": 1.5084679126739502, "learning_rate": 0.0001, "loss": 3.5441575050354004, "step": 985 }, { "epoch": 0.3638544213293971, "grad_norm": 1.665592074394226, "learning_rate": 0.0001, "loss": 3.7481281757354736, "step": 986 }, { "epoch": 0.3642234420406845, "grad_norm": 1.7114098072052002, "learning_rate": 0.0001, "loss": 3.649366617202759, "step": 987 }, { "epoch": 0.36459246275197194, "grad_norm": 2.06028413772583, "learning_rate": 0.0001, "loss": 4.045251846313477, "step": 988 }, { "epoch": 0.36496148346325935, "grad_norm": 1.473903775215149, "learning_rate": 0.0001, "loss": 3.692885160446167, "step": 989 }, { "epoch": 0.3653305041745468, "grad_norm": 1.9220426082611084, "learning_rate": 0.0001, "loss": 3.5416295528411865, "step": 990 }, { "epoch": 0.36569952488583424, "grad_norm": 1.8098729848861694, "learning_rate": 0.0001, "loss": 3.6792027950286865, "step": 991 }, { "epoch": 0.36606854559712165, "grad_norm": 1.777948021888733, "learning_rate": 0.0001, "loss": 3.8671905994415283, "step": 992 }, { "epoch": 0.36643756630840907, "grad_norm": 2.385817050933838, "learning_rate": 0.0001, "loss": 3.4315450191497803, "step": 993 }, { "epoch": 0.3668065870196965, "grad_norm": 1.5102519989013672, "learning_rate": 0.0001, "loss": 3.7663180828094482, "step": 994 }, { "epoch": 0.3671756077309839, "grad_norm": 2.345006227493286, "learning_rate": 0.0001, "loss": 3.660745859146118, "step": 995 }, { "epoch": 0.3675446284422713, "grad_norm": 1.4960929155349731, "learning_rate": 0.0001, "loss": 3.458857536315918, "step": 996 }, { "epoch": 0.36791364915355873, "grad_norm": 1.7957972288131714, "learning_rate": 0.0001, "loss": 4.0572896003723145, "step": 997 }, { "epoch": 0.36828266986484615, "grad_norm": 1.6085453033447266, "learning_rate": 0.0001, "loss": 3.4816460609436035, "step": 998 }, { "epoch": 0.36865169057613356, "grad_norm": 1.5932081937789917, "learning_rate": 0.0001, "loss": 3.147345781326294, "step": 999 }, { "epoch": 0.36902071128742103, "grad_norm": 1.9356966018676758, "learning_rate": 0.0001, "loss": 3.3720457553863525, "step": 1000 }, { "epoch": 0.36938973199870845, "grad_norm": 2.6087594032287598, "learning_rate": 0.0001, "loss": 3.5508546829223633, "step": 1001 }, { "epoch": 0.36975875270999586, "grad_norm": 2.2257230281829834, "learning_rate": 0.0001, "loss": 3.638443946838379, "step": 1002 }, { "epoch": 0.3701277734212833, "grad_norm": 1.6002798080444336, "learning_rate": 0.0001, "loss": 3.2786178588867188, "step": 1003 }, { "epoch": 0.3704967941325707, "grad_norm": 2.1445670127868652, "learning_rate": 0.0001, "loss": 3.6076927185058594, "step": 1004 }, { "epoch": 0.3708658148438581, "grad_norm": 1.8145312070846558, "learning_rate": 0.0001, "loss": 3.7419614791870117, "step": 1005 }, { "epoch": 0.3712348355551455, "grad_norm": 1.9384288787841797, "learning_rate": 0.0001, "loss": 3.6455776691436768, "step": 1006 }, { "epoch": 0.37160385626643294, "grad_norm": 3.0492701530456543, "learning_rate": 0.0001, "loss": 3.756404399871826, "step": 1007 }, { "epoch": 0.37197287697772036, "grad_norm": 2.072761058807373, "learning_rate": 0.0001, "loss": 3.382974147796631, "step": 1008 }, { "epoch": 0.3723418976890078, "grad_norm": 1.9356118440628052, "learning_rate": 0.0001, "loss": 3.70412278175354, "step": 1009 }, { "epoch": 0.37271091840029524, "grad_norm": 3.073763370513916, "learning_rate": 0.0001, "loss": 4.081543445587158, "step": 1010 }, { "epoch": 0.37307993911158266, "grad_norm": 1.7098064422607422, "learning_rate": 0.0001, "loss": 3.724994421005249, "step": 1011 }, { "epoch": 0.3734489598228701, "grad_norm": 2.1610236167907715, "learning_rate": 0.0001, "loss": 3.382457971572876, "step": 1012 }, { "epoch": 0.3738179805341575, "grad_norm": 1.7166200876235962, "learning_rate": 0.0001, "loss": 3.514925241470337, "step": 1013 }, { "epoch": 0.3741870012454449, "grad_norm": 1.7985845804214478, "learning_rate": 0.0001, "loss": 3.369422435760498, "step": 1014 }, { "epoch": 0.3745560219567323, "grad_norm": 2.017040729522705, "learning_rate": 0.0001, "loss": 3.5744919776916504, "step": 1015 }, { "epoch": 0.37492504266801974, "grad_norm": 1.9383666515350342, "learning_rate": 0.0001, "loss": 3.6265766620635986, "step": 1016 }, { "epoch": 0.37529406337930715, "grad_norm": 2.7773749828338623, "learning_rate": 0.0001, "loss": 3.7150726318359375, "step": 1017 }, { "epoch": 0.37566308409059457, "grad_norm": 1.5019221305847168, "learning_rate": 0.0001, "loss": 3.675811767578125, "step": 1018 }, { "epoch": 0.376032104801882, "grad_norm": 1.6941287517547607, "learning_rate": 0.0001, "loss": 3.615675449371338, "step": 1019 }, { "epoch": 0.3764011255131694, "grad_norm": 1.8472751379013062, "learning_rate": 0.0001, "loss": 3.2442116737365723, "step": 1020 }, { "epoch": 0.37677014622445687, "grad_norm": 1.6253347396850586, "learning_rate": 0.0001, "loss": 3.6195101737976074, "step": 1021 }, { "epoch": 0.3771391669357443, "grad_norm": 2.329472780227661, "learning_rate": 0.0001, "loss": 3.5222408771514893, "step": 1022 }, { "epoch": 0.3775081876470317, "grad_norm": 1.6834906339645386, "learning_rate": 0.0001, "loss": 3.5763134956359863, "step": 1023 }, { "epoch": 0.3778772083583191, "grad_norm": 1.5031334161758423, "learning_rate": 0.0001, "loss": 3.567384958267212, "step": 1024 }, { "epoch": 0.37824622906960653, "grad_norm": 1.6308482885360718, "learning_rate": 0.0001, "loss": 3.451554536819458, "step": 1025 }, { "epoch": 0.37861524978089395, "grad_norm": 1.5196822881698608, "learning_rate": 0.0001, "loss": 3.275092601776123, "step": 1026 }, { "epoch": 0.37898427049218136, "grad_norm": 1.779683232307434, "learning_rate": 0.0001, "loss": 3.691413402557373, "step": 1027 }, { "epoch": 0.3793532912034688, "grad_norm": 1.9622735977172852, "learning_rate": 0.0001, "loss": 3.323134183883667, "step": 1028 }, { "epoch": 0.3797223119147562, "grad_norm": 1.7832732200622559, "learning_rate": 0.0001, "loss": 3.30794095993042, "step": 1029 }, { "epoch": 0.3800913326260436, "grad_norm": 1.7913005352020264, "learning_rate": 0.0001, "loss": 3.4632575511932373, "step": 1030 }, { "epoch": 0.3804603533373311, "grad_norm": 1.5579787492752075, "learning_rate": 0.0001, "loss": 3.09414005279541, "step": 1031 }, { "epoch": 0.3808293740486185, "grad_norm": 1.8105117082595825, "learning_rate": 0.0001, "loss": 3.809582233428955, "step": 1032 }, { "epoch": 0.3811983947599059, "grad_norm": 1.6644221544265747, "learning_rate": 0.0001, "loss": 3.6848762035369873, "step": 1033 }, { "epoch": 0.3815674154711933, "grad_norm": 2.0220589637756348, "learning_rate": 0.0001, "loss": 3.842792272567749, "step": 1034 }, { "epoch": 0.38193643618248074, "grad_norm": 1.869362711906433, "learning_rate": 0.0001, "loss": 3.38818621635437, "step": 1035 }, { "epoch": 0.38230545689376816, "grad_norm": 1.9640357494354248, "learning_rate": 0.0001, "loss": 3.6359920501708984, "step": 1036 }, { "epoch": 0.3826744776050556, "grad_norm": 1.9055365324020386, "learning_rate": 0.0001, "loss": 3.3396472930908203, "step": 1037 }, { "epoch": 0.383043498316343, "grad_norm": 1.4364265203475952, "learning_rate": 0.0001, "loss": 3.5074639320373535, "step": 1038 }, { "epoch": 0.3834125190276304, "grad_norm": 1.8021663427352905, "learning_rate": 0.0001, "loss": 3.385700225830078, "step": 1039 }, { "epoch": 0.3837815397389178, "grad_norm": 1.8658980131149292, "learning_rate": 0.0001, "loss": 3.4735209941864014, "step": 1040 }, { "epoch": 0.3841505604502053, "grad_norm": 1.6618491411209106, "learning_rate": 0.0001, "loss": 3.756561279296875, "step": 1041 }, { "epoch": 0.3845195811614927, "grad_norm": 1.5899429321289062, "learning_rate": 0.0001, "loss": 3.4793660640716553, "step": 1042 }, { "epoch": 0.3848886018727801, "grad_norm": 2.0102319717407227, "learning_rate": 0.0001, "loss": 3.4993884563446045, "step": 1043 }, { "epoch": 0.38525762258406754, "grad_norm": 1.47008216381073, "learning_rate": 0.0001, "loss": 3.5453593730926514, "step": 1044 }, { "epoch": 0.38562664329535495, "grad_norm": 1.6052583456039429, "learning_rate": 0.0001, "loss": 3.8380298614501953, "step": 1045 }, { "epoch": 0.38599566400664237, "grad_norm": 1.642446756362915, "learning_rate": 0.0001, "loss": 3.4695844650268555, "step": 1046 }, { "epoch": 0.3863646847179298, "grad_norm": 1.3856698274612427, "learning_rate": 0.0001, "loss": 3.4078400135040283, "step": 1047 }, { "epoch": 0.3867337054292172, "grad_norm": 1.4176830053329468, "learning_rate": 0.0001, "loss": 3.6066999435424805, "step": 1048 }, { "epoch": 0.3871027261405046, "grad_norm": 1.534299373626709, "learning_rate": 0.0001, "loss": 3.4636473655700684, "step": 1049 }, { "epoch": 0.38747174685179203, "grad_norm": 1.7126305103302002, "learning_rate": 0.0001, "loss": 3.399542808532715, "step": 1050 }, { "epoch": 0.3878407675630795, "grad_norm": 1.6469898223876953, "learning_rate": 0.0001, "loss": 3.617379665374756, "step": 1051 }, { "epoch": 0.3882097882743669, "grad_norm": 1.8253321647644043, "learning_rate": 0.0001, "loss": 3.5330235958099365, "step": 1052 }, { "epoch": 0.38857880898565433, "grad_norm": 2.0213844776153564, "learning_rate": 0.0001, "loss": 3.6431636810302734, "step": 1053 }, { "epoch": 0.38894782969694175, "grad_norm": 1.8043290376663208, "learning_rate": 0.0001, "loss": 3.80029034614563, "step": 1054 }, { "epoch": 0.38931685040822916, "grad_norm": 1.6341948509216309, "learning_rate": 0.0001, "loss": 3.9497902393341064, "step": 1055 }, { "epoch": 0.3896858711195166, "grad_norm": 1.924888014793396, "learning_rate": 0.0001, "loss": 3.4991908073425293, "step": 1056 }, { "epoch": 0.390054891830804, "grad_norm": 1.556576132774353, "learning_rate": 0.0001, "loss": 3.2436673641204834, "step": 1057 }, { "epoch": 0.3904239125420914, "grad_norm": 2.139651298522949, "learning_rate": 0.0001, "loss": 3.718282699584961, "step": 1058 }, { "epoch": 0.3907929332533788, "grad_norm": 1.961877703666687, "learning_rate": 0.0001, "loss": 3.5330257415771484, "step": 1059 }, { "epoch": 0.39116195396466624, "grad_norm": 1.4173837900161743, "learning_rate": 0.0001, "loss": 3.248403310775757, "step": 1060 }, { "epoch": 0.3915309746759537, "grad_norm": 1.7147531509399414, "learning_rate": 0.0001, "loss": 3.2063255310058594, "step": 1061 }, { "epoch": 0.39189999538724113, "grad_norm": 1.799842357635498, "learning_rate": 0.0001, "loss": 3.73169207572937, "step": 1062 }, { "epoch": 0.39226901609852854, "grad_norm": 1.7949094772338867, "learning_rate": 0.0001, "loss": 3.747079372406006, "step": 1063 }, { "epoch": 0.39263803680981596, "grad_norm": 1.8971632719039917, "learning_rate": 0.0001, "loss": 3.8412744998931885, "step": 1064 }, { "epoch": 0.3930070575211034, "grad_norm": 2.200519323348999, "learning_rate": 0.0001, "loss": 4.009820938110352, "step": 1065 }, { "epoch": 0.3933760782323908, "grad_norm": 1.7085025310516357, "learning_rate": 0.0001, "loss": 3.2844953536987305, "step": 1066 }, { "epoch": 0.3937450989436782, "grad_norm": 1.6495426893234253, "learning_rate": 0.0001, "loss": 3.320258140563965, "step": 1067 }, { "epoch": 0.3941141196549656, "grad_norm": 1.775577187538147, "learning_rate": 0.0001, "loss": 3.362485408782959, "step": 1068 }, { "epoch": 0.39448314036625304, "grad_norm": 1.8102372884750366, "learning_rate": 0.0001, "loss": 3.323610544204712, "step": 1069 }, { "epoch": 0.39485216107754045, "grad_norm": 1.716867446899414, "learning_rate": 0.0001, "loss": 3.5217385292053223, "step": 1070 }, { "epoch": 0.3952211817888279, "grad_norm": 2.428114414215088, "learning_rate": 0.0001, "loss": 3.3475725650787354, "step": 1071 }, { "epoch": 0.39559020250011534, "grad_norm": 1.6314780712127686, "learning_rate": 0.0001, "loss": 3.590587615966797, "step": 1072 }, { "epoch": 0.39595922321140276, "grad_norm": 1.8220868110656738, "learning_rate": 0.0001, "loss": 3.61910080909729, "step": 1073 }, { "epoch": 0.39632824392269017, "grad_norm": 2.0068199634552, "learning_rate": 0.0001, "loss": 3.285186290740967, "step": 1074 }, { "epoch": 0.3966972646339776, "grad_norm": 1.655479907989502, "learning_rate": 0.0001, "loss": 3.629521131515503, "step": 1075 }, { "epoch": 0.397066285345265, "grad_norm": 2.5227596759796143, "learning_rate": 0.0001, "loss": 3.4802355766296387, "step": 1076 }, { "epoch": 0.3974353060565524, "grad_norm": 1.5765255689620972, "learning_rate": 0.0001, "loss": 3.698794364929199, "step": 1077 }, { "epoch": 0.39780432676783983, "grad_norm": 1.5991921424865723, "learning_rate": 0.0001, "loss": 3.4378538131713867, "step": 1078 }, { "epoch": 0.39817334747912725, "grad_norm": 1.4769134521484375, "learning_rate": 0.0001, "loss": 3.7685511112213135, "step": 1079 }, { "epoch": 0.39854236819041466, "grad_norm": 2.4793806076049805, "learning_rate": 0.0001, "loss": 3.6917762756347656, "step": 1080 }, { "epoch": 0.39891138890170214, "grad_norm": 1.866546630859375, "learning_rate": 0.0001, "loss": 3.2043046951293945, "step": 1081 }, { "epoch": 0.39928040961298955, "grad_norm": 2.238892078399658, "learning_rate": 0.0001, "loss": 3.360795021057129, "step": 1082 }, { "epoch": 0.39964943032427697, "grad_norm": 1.4917564392089844, "learning_rate": 0.0001, "loss": 3.5305404663085938, "step": 1083 }, { "epoch": 0.4000184510355644, "grad_norm": 2.206327438354492, "learning_rate": 0.0001, "loss": 3.4937171936035156, "step": 1084 }, { "epoch": 0.4003874717468518, "grad_norm": 1.6631877422332764, "learning_rate": 0.0001, "loss": 3.1713926792144775, "step": 1085 }, { "epoch": 0.4007564924581392, "grad_norm": 1.7848588228225708, "learning_rate": 0.0001, "loss": 3.6137077808380127, "step": 1086 }, { "epoch": 0.40112551316942663, "grad_norm": 1.5225774049758911, "learning_rate": 0.0001, "loss": 3.8191094398498535, "step": 1087 }, { "epoch": 0.40149453388071404, "grad_norm": 2.3590917587280273, "learning_rate": 0.0001, "loss": 3.8867969512939453, "step": 1088 }, { "epoch": 0.40186355459200146, "grad_norm": 2.1857473850250244, "learning_rate": 0.0001, "loss": 3.3984169960021973, "step": 1089 }, { "epoch": 0.4022325753032889, "grad_norm": 1.5819369554519653, "learning_rate": 0.0001, "loss": 3.4486732482910156, "step": 1090 }, { "epoch": 0.4026015960145763, "grad_norm": 1.650406002998352, "learning_rate": 0.0001, "loss": 3.231422185897827, "step": 1091 }, { "epoch": 0.40297061672586376, "grad_norm": 1.7265455722808838, "learning_rate": 0.0001, "loss": 3.5539236068725586, "step": 1092 }, { "epoch": 0.4033396374371512, "grad_norm": 1.70913565158844, "learning_rate": 0.0001, "loss": 3.4585118293762207, "step": 1093 }, { "epoch": 0.4037086581484386, "grad_norm": 1.81973397731781, "learning_rate": 0.0001, "loss": 3.756134510040283, "step": 1094 }, { "epoch": 0.404077678859726, "grad_norm": 1.4688886404037476, "learning_rate": 0.0001, "loss": 3.1194019317626953, "step": 1095 }, { "epoch": 0.4044466995710134, "grad_norm": 1.6219803094863892, "learning_rate": 0.0001, "loss": 3.4928295612335205, "step": 1096 }, { "epoch": 0.40481572028230084, "grad_norm": 1.6186461448669434, "learning_rate": 0.0001, "loss": 3.453388214111328, "step": 1097 }, { "epoch": 0.40518474099358825, "grad_norm": 1.5223987102508545, "learning_rate": 0.0001, "loss": 3.459501266479492, "step": 1098 }, { "epoch": 0.40555376170487567, "grad_norm": 2.174037218093872, "learning_rate": 0.0001, "loss": 3.531449794769287, "step": 1099 }, { "epoch": 0.4059227824161631, "grad_norm": 1.6174062490463257, "learning_rate": 0.0001, "loss": 3.1327972412109375, "step": 1100 }, { "epoch": 0.4062918031274505, "grad_norm": 1.8244595527648926, "learning_rate": 0.0001, "loss": 3.7290871143341064, "step": 1101 }, { "epoch": 0.406660823838738, "grad_norm": 1.616381049156189, "learning_rate": 0.0001, "loss": 3.564664840698242, "step": 1102 }, { "epoch": 0.4070298445500254, "grad_norm": 2.438241958618164, "learning_rate": 0.0001, "loss": 3.4331419467926025, "step": 1103 }, { "epoch": 0.4073988652613128, "grad_norm": 1.5675773620605469, "learning_rate": 0.0001, "loss": 3.7543487548828125, "step": 1104 }, { "epoch": 0.4077678859726002, "grad_norm": 1.5436387062072754, "learning_rate": 0.0001, "loss": 3.64493465423584, "step": 1105 }, { "epoch": 0.40813690668388763, "grad_norm": 1.970644474029541, "learning_rate": 0.0001, "loss": 3.8694844245910645, "step": 1106 }, { "epoch": 0.40850592739517505, "grad_norm": 1.7549302577972412, "learning_rate": 0.0001, "loss": 3.5135347843170166, "step": 1107 }, { "epoch": 0.40887494810646247, "grad_norm": 1.5151584148406982, "learning_rate": 0.0001, "loss": 3.5717740058898926, "step": 1108 }, { "epoch": 0.4092439688177499, "grad_norm": 1.6516414880752563, "learning_rate": 0.0001, "loss": 3.412562131881714, "step": 1109 }, { "epoch": 0.4096129895290373, "grad_norm": 1.5741270780563354, "learning_rate": 0.0001, "loss": 3.829526901245117, "step": 1110 }, { "epoch": 0.4099820102403247, "grad_norm": 1.6586413383483887, "learning_rate": 0.0001, "loss": 3.552276134490967, "step": 1111 }, { "epoch": 0.4103510309516122, "grad_norm": 1.7744735479354858, "learning_rate": 0.0001, "loss": 3.5275754928588867, "step": 1112 }, { "epoch": 0.4107200516628996, "grad_norm": 1.7126959562301636, "learning_rate": 0.0001, "loss": 3.479656457901001, "step": 1113 }, { "epoch": 0.411089072374187, "grad_norm": 1.8186641931533813, "learning_rate": 0.0001, "loss": 3.313764810562134, "step": 1114 }, { "epoch": 0.41145809308547443, "grad_norm": 1.5473167896270752, "learning_rate": 0.0001, "loss": 3.2337112426757812, "step": 1115 }, { "epoch": 0.41182711379676185, "grad_norm": 1.4319747686386108, "learning_rate": 0.0001, "loss": 3.5199079513549805, "step": 1116 }, { "epoch": 0.41219613450804926, "grad_norm": 1.8948994874954224, "learning_rate": 0.0001, "loss": 3.231522798538208, "step": 1117 }, { "epoch": 0.4125651552193367, "grad_norm": 2.1671953201293945, "learning_rate": 0.0001, "loss": 3.368067502975464, "step": 1118 }, { "epoch": 0.4129341759306241, "grad_norm": 1.659521460533142, "learning_rate": 0.0001, "loss": 3.133955955505371, "step": 1119 }, { "epoch": 0.4133031966419115, "grad_norm": 1.9581573009490967, "learning_rate": 0.0001, "loss": 3.538813352584839, "step": 1120 }, { "epoch": 0.4136722173531989, "grad_norm": 1.950330376625061, "learning_rate": 0.0001, "loss": 3.265617609024048, "step": 1121 }, { "epoch": 0.4140412380644864, "grad_norm": 1.5767706632614136, "learning_rate": 0.0001, "loss": 3.3756749629974365, "step": 1122 }, { "epoch": 0.4144102587757738, "grad_norm": 1.7552682161331177, "learning_rate": 0.0001, "loss": 3.725167751312256, "step": 1123 }, { "epoch": 0.4147792794870612, "grad_norm": 1.6119760274887085, "learning_rate": 0.0001, "loss": 3.555713653564453, "step": 1124 }, { "epoch": 0.41514830019834864, "grad_norm": 1.9205046892166138, "learning_rate": 0.0001, "loss": 3.273820400238037, "step": 1125 }, { "epoch": 0.41551732090963606, "grad_norm": 1.6919970512390137, "learning_rate": 0.0001, "loss": 3.439239740371704, "step": 1126 }, { "epoch": 0.41588634162092347, "grad_norm": 1.6779921054840088, "learning_rate": 0.0001, "loss": 3.354769229888916, "step": 1127 }, { "epoch": 0.4162553623322109, "grad_norm": 2.210174083709717, "learning_rate": 0.0001, "loss": 3.5124754905700684, "step": 1128 }, { "epoch": 0.4166243830434983, "grad_norm": 1.7472038269042969, "learning_rate": 0.0001, "loss": 3.098078489303589, "step": 1129 }, { "epoch": 0.4169934037547857, "grad_norm": 3.019456386566162, "learning_rate": 0.0001, "loss": 3.047180652618408, "step": 1130 }, { "epoch": 0.41736242446607313, "grad_norm": 1.608566164970398, "learning_rate": 0.0001, "loss": 3.190749168395996, "step": 1131 }, { "epoch": 0.4177314451773606, "grad_norm": 1.9335167407989502, "learning_rate": 0.0001, "loss": 3.465040445327759, "step": 1132 }, { "epoch": 0.418100465888648, "grad_norm": 1.6094701290130615, "learning_rate": 0.0001, "loss": 3.4975686073303223, "step": 1133 }, { "epoch": 0.41846948659993544, "grad_norm": 1.7231701612472534, "learning_rate": 0.0001, "loss": 3.4584193229675293, "step": 1134 }, { "epoch": 0.41883850731122285, "grad_norm": 1.986169457435608, "learning_rate": 0.0001, "loss": 3.727247476577759, "step": 1135 }, { "epoch": 0.41920752802251027, "grad_norm": 2.0806264877319336, "learning_rate": 0.0001, "loss": 3.389432668685913, "step": 1136 }, { "epoch": 0.4195765487337977, "grad_norm": 1.7676399946212769, "learning_rate": 0.0001, "loss": 3.5187435150146484, "step": 1137 }, { "epoch": 0.4199455694450851, "grad_norm": 1.6268689632415771, "learning_rate": 0.0001, "loss": 3.0579276084899902, "step": 1138 }, { "epoch": 0.4203145901563725, "grad_norm": 2.9556727409362793, "learning_rate": 0.0001, "loss": 3.270014762878418, "step": 1139 }, { "epoch": 0.42068361086765993, "grad_norm": 1.4848302602767944, "learning_rate": 0.0001, "loss": 3.332440137863159, "step": 1140 }, { "epoch": 0.42105263157894735, "grad_norm": 1.721583604812622, "learning_rate": 0.0001, "loss": 3.0012364387512207, "step": 1141 }, { "epoch": 0.4214216522902348, "grad_norm": 1.8044254779815674, "learning_rate": 0.0001, "loss": 3.3614799976348877, "step": 1142 }, { "epoch": 0.42179067300152223, "grad_norm": 1.6166692972183228, "learning_rate": 0.0001, "loss": 3.351100444793701, "step": 1143 }, { "epoch": 0.42215969371280965, "grad_norm": 1.7357293367385864, "learning_rate": 0.0001, "loss": 3.5229084491729736, "step": 1144 }, { "epoch": 0.42252871442409706, "grad_norm": 1.5106472969055176, "learning_rate": 0.0001, "loss": 2.988489866256714, "step": 1145 }, { "epoch": 0.4228977351353845, "grad_norm": 1.5887922048568726, "learning_rate": 0.0001, "loss": 3.171905040740967, "step": 1146 }, { "epoch": 0.4232667558466719, "grad_norm": 1.8840998411178589, "learning_rate": 0.0001, "loss": 3.4596126079559326, "step": 1147 }, { "epoch": 0.4236357765579593, "grad_norm": 2.1984126567840576, "learning_rate": 0.0001, "loss": 3.387892246246338, "step": 1148 }, { "epoch": 0.4240047972692467, "grad_norm": 1.4740560054779053, "learning_rate": 0.0001, "loss": 3.4837212562561035, "step": 1149 }, { "epoch": 0.42437381798053414, "grad_norm": 1.4869908094406128, "learning_rate": 0.0001, "loss": 3.2742204666137695, "step": 1150 }, { "epoch": 0.42474283869182156, "grad_norm": 2.2522780895233154, "learning_rate": 0.0001, "loss": 3.855175018310547, "step": 1151 }, { "epoch": 0.425111859403109, "grad_norm": 2.3794596195220947, "learning_rate": 0.0001, "loss": 4.1811442375183105, "step": 1152 }, { "epoch": 0.42548088011439644, "grad_norm": 1.78865385055542, "learning_rate": 0.0001, "loss": 3.505410671234131, "step": 1153 }, { "epoch": 0.42584990082568386, "grad_norm": 1.6205072402954102, "learning_rate": 0.0001, "loss": 3.2110962867736816, "step": 1154 }, { "epoch": 0.4262189215369713, "grad_norm": 1.492835521697998, "learning_rate": 0.0001, "loss": 3.532453775405884, "step": 1155 }, { "epoch": 0.4265879422482587, "grad_norm": 1.6148616075515747, "learning_rate": 0.0001, "loss": 3.283496379852295, "step": 1156 }, { "epoch": 0.4269569629595461, "grad_norm": 1.7241994142532349, "learning_rate": 0.0001, "loss": 3.215756416320801, "step": 1157 }, { "epoch": 0.4273259836708335, "grad_norm": 1.485260009765625, "learning_rate": 0.0001, "loss": 3.5015108585357666, "step": 1158 }, { "epoch": 0.42769500438212094, "grad_norm": 1.8061119318008423, "learning_rate": 0.0001, "loss": 3.4037883281707764, "step": 1159 }, { "epoch": 0.42806402509340835, "grad_norm": 1.6641099452972412, "learning_rate": 0.0001, "loss": 3.097249746322632, "step": 1160 }, { "epoch": 0.42843304580469577, "grad_norm": 1.7615569829940796, "learning_rate": 0.0001, "loss": 3.1619954109191895, "step": 1161 }, { "epoch": 0.4288020665159832, "grad_norm": 1.934561014175415, "learning_rate": 0.0001, "loss": 3.3449654579162598, "step": 1162 }, { "epoch": 0.42917108722727065, "grad_norm": 1.7241837978363037, "learning_rate": 0.0001, "loss": 3.425506114959717, "step": 1163 }, { "epoch": 0.42954010793855807, "grad_norm": 1.5075688362121582, "learning_rate": 0.0001, "loss": 3.3891100883483887, "step": 1164 }, { "epoch": 0.4299091286498455, "grad_norm": 1.4058998823165894, "learning_rate": 0.0001, "loss": 3.4876890182495117, "step": 1165 }, { "epoch": 0.4302781493611329, "grad_norm": 1.6247506141662598, "learning_rate": 0.0001, "loss": 3.892271041870117, "step": 1166 }, { "epoch": 0.4306471700724203, "grad_norm": 1.5740920305252075, "learning_rate": 0.0001, "loss": 3.5151073932647705, "step": 1167 }, { "epoch": 0.43101619078370773, "grad_norm": 1.934751272201538, "learning_rate": 0.0001, "loss": 3.383263111114502, "step": 1168 }, { "epoch": 0.43138521149499515, "grad_norm": 1.4747998714447021, "learning_rate": 0.0001, "loss": 3.4744138717651367, "step": 1169 }, { "epoch": 0.43175423220628256, "grad_norm": 1.749432921409607, "learning_rate": 0.0001, "loss": 3.2744152545928955, "step": 1170 }, { "epoch": 0.43212325291757, "grad_norm": 1.667901873588562, "learning_rate": 0.0001, "loss": 3.490841865539551, "step": 1171 }, { "epoch": 0.4324922736288574, "grad_norm": 2.125776767730713, "learning_rate": 0.0001, "loss": 3.526581048965454, "step": 1172 }, { "epoch": 0.43286129434014486, "grad_norm": 1.6938812732696533, "learning_rate": 0.0001, "loss": 3.3852858543395996, "step": 1173 }, { "epoch": 0.4332303150514323, "grad_norm": 1.5756855010986328, "learning_rate": 0.0001, "loss": 3.314359664916992, "step": 1174 }, { "epoch": 0.4335993357627197, "grad_norm": 2.076364040374756, "learning_rate": 0.0001, "loss": 3.3884148597717285, "step": 1175 }, { "epoch": 0.4339683564740071, "grad_norm": 1.7696532011032104, "learning_rate": 0.0001, "loss": 3.628148078918457, "step": 1176 }, { "epoch": 0.4343373771852945, "grad_norm": 1.7007116079330444, "learning_rate": 0.0001, "loss": 3.4393184185028076, "step": 1177 }, { "epoch": 0.43470639789658194, "grad_norm": 1.9358893632888794, "learning_rate": 0.0001, "loss": 3.5703320503234863, "step": 1178 }, { "epoch": 0.43507541860786936, "grad_norm": 1.645951509475708, "learning_rate": 0.0001, "loss": 3.423347234725952, "step": 1179 }, { "epoch": 0.4354444393191568, "grad_norm": 3.4309563636779785, "learning_rate": 0.0001, "loss": 3.407176971435547, "step": 1180 }, { "epoch": 0.4358134600304442, "grad_norm": 2.664422035217285, "learning_rate": 0.0001, "loss": 3.5816760063171387, "step": 1181 }, { "epoch": 0.4361824807417316, "grad_norm": 1.5681055784225464, "learning_rate": 0.0001, "loss": 3.339106559753418, "step": 1182 }, { "epoch": 0.4365515014530191, "grad_norm": 1.5695600509643555, "learning_rate": 0.0001, "loss": 3.5136218070983887, "step": 1183 }, { "epoch": 0.4369205221643065, "grad_norm": 1.900416374206543, "learning_rate": 0.0001, "loss": 3.4979610443115234, "step": 1184 }, { "epoch": 0.4372895428755939, "grad_norm": 1.512001395225525, "learning_rate": 0.0001, "loss": 3.3145952224731445, "step": 1185 }, { "epoch": 0.4376585635868813, "grad_norm": 1.6252752542495728, "learning_rate": 0.0001, "loss": 3.6436784267425537, "step": 1186 }, { "epoch": 0.43802758429816874, "grad_norm": 1.743227481842041, "learning_rate": 0.0001, "loss": 3.059513568878174, "step": 1187 }, { "epoch": 0.43839660500945615, "grad_norm": 1.8757061958312988, "learning_rate": 0.0001, "loss": 3.7995994091033936, "step": 1188 }, { "epoch": 0.43876562572074357, "grad_norm": 1.7850326299667358, "learning_rate": 0.0001, "loss": 3.5283584594726562, "step": 1189 }, { "epoch": 0.439134646432031, "grad_norm": 1.9045826196670532, "learning_rate": 0.0001, "loss": 3.1428403854370117, "step": 1190 }, { "epoch": 0.4395036671433184, "grad_norm": 1.5550034046173096, "learning_rate": 0.0001, "loss": 3.2889909744262695, "step": 1191 }, { "epoch": 0.4398726878546058, "grad_norm": 1.4155323505401611, "learning_rate": 0.0001, "loss": 3.172661304473877, "step": 1192 }, { "epoch": 0.4402417085658933, "grad_norm": 1.8927156925201416, "learning_rate": 0.0001, "loss": 3.3489363193511963, "step": 1193 }, { "epoch": 0.4406107292771807, "grad_norm": 1.8200788497924805, "learning_rate": 0.0001, "loss": 3.3229305744171143, "step": 1194 }, { "epoch": 0.4409797499884681, "grad_norm": 1.9508037567138672, "learning_rate": 0.0001, "loss": 3.621987819671631, "step": 1195 }, { "epoch": 0.44134877069975553, "grad_norm": 1.429848551750183, "learning_rate": 0.0001, "loss": 3.2655892372131348, "step": 1196 }, { "epoch": 0.44171779141104295, "grad_norm": 1.5047922134399414, "learning_rate": 0.0001, "loss": 3.7596511840820312, "step": 1197 }, { "epoch": 0.44208681212233036, "grad_norm": 2.0604217052459717, "learning_rate": 0.0001, "loss": 3.6311607360839844, "step": 1198 }, { "epoch": 0.4424558328336178, "grad_norm": 1.5865529775619507, "learning_rate": 0.0001, "loss": 3.627190589904785, "step": 1199 }, { "epoch": 0.4428248535449052, "grad_norm": 1.4167044162750244, "learning_rate": 0.0001, "loss": 3.3320870399475098, "step": 1200 }, { "epoch": 0.4431938742561926, "grad_norm": 1.652715802192688, "learning_rate": 0.0001, "loss": 3.2706832885742188, "step": 1201 }, { "epoch": 0.44356289496748, "grad_norm": 1.5822203159332275, "learning_rate": 0.0001, "loss": 3.262948989868164, "step": 1202 }, { "epoch": 0.4439319156787675, "grad_norm": 1.417222023010254, "learning_rate": 0.0001, "loss": 3.130798578262329, "step": 1203 }, { "epoch": 0.4443009363900549, "grad_norm": 1.6606483459472656, "learning_rate": 0.0001, "loss": 3.337520122528076, "step": 1204 }, { "epoch": 0.44466995710134233, "grad_norm": 1.747694969177246, "learning_rate": 0.0001, "loss": 3.334962844848633, "step": 1205 }, { "epoch": 0.44503897781262974, "grad_norm": 1.5451515913009644, "learning_rate": 0.0001, "loss": 3.524030923843384, "step": 1206 }, { "epoch": 0.44540799852391716, "grad_norm": 1.8818703889846802, "learning_rate": 0.0001, "loss": 3.349241256713867, "step": 1207 }, { "epoch": 0.4457770192352046, "grad_norm": 1.8163851499557495, "learning_rate": 0.0001, "loss": 3.28201961517334, "step": 1208 }, { "epoch": 0.446146039946492, "grad_norm": 1.6565134525299072, "learning_rate": 0.0001, "loss": 3.61373233795166, "step": 1209 }, { "epoch": 0.4465150606577794, "grad_norm": 1.7229743003845215, "learning_rate": 0.0001, "loss": 3.5443825721740723, "step": 1210 }, { "epoch": 0.4468840813690668, "grad_norm": 1.6731629371643066, "learning_rate": 0.0001, "loss": 3.6706151962280273, "step": 1211 }, { "epoch": 0.44725310208035424, "grad_norm": 1.769970417022705, "learning_rate": 0.0001, "loss": 3.3441367149353027, "step": 1212 }, { "epoch": 0.4476221227916417, "grad_norm": 1.983640193939209, "learning_rate": 0.0001, "loss": 3.4111194610595703, "step": 1213 }, { "epoch": 0.4479911435029291, "grad_norm": 1.7668060064315796, "learning_rate": 0.0001, "loss": 3.356395721435547, "step": 1214 }, { "epoch": 0.44836016421421654, "grad_norm": 1.6508654356002808, "learning_rate": 0.0001, "loss": 3.1866378784179688, "step": 1215 }, { "epoch": 0.44872918492550395, "grad_norm": 1.5666056871414185, "learning_rate": 0.0001, "loss": 3.268158197402954, "step": 1216 }, { "epoch": 0.44909820563679137, "grad_norm": 1.6124550104141235, "learning_rate": 0.0001, "loss": 3.4840240478515625, "step": 1217 }, { "epoch": 0.4494672263480788, "grad_norm": 2.06943941116333, "learning_rate": 0.0001, "loss": 3.1811251640319824, "step": 1218 }, { "epoch": 0.4498362470593662, "grad_norm": 1.7592380046844482, "learning_rate": 0.0001, "loss": 3.2974326610565186, "step": 1219 }, { "epoch": 0.4502052677706536, "grad_norm": 1.4490355253219604, "learning_rate": 0.0001, "loss": 3.0902347564697266, "step": 1220 }, { "epoch": 0.45057428848194103, "grad_norm": 1.733947515487671, "learning_rate": 0.0001, "loss": 3.6533727645874023, "step": 1221 }, { "epoch": 0.45094330919322845, "grad_norm": 1.6569914817810059, "learning_rate": 0.0001, "loss": 3.244729995727539, "step": 1222 }, { "epoch": 0.4513123299045159, "grad_norm": 1.9674009084701538, "learning_rate": 0.0001, "loss": 3.2443857192993164, "step": 1223 }, { "epoch": 0.45168135061580333, "grad_norm": 1.5391271114349365, "learning_rate": 0.0001, "loss": 3.55330228805542, "step": 1224 }, { "epoch": 0.45205037132709075, "grad_norm": 1.7937438488006592, "learning_rate": 0.0001, "loss": 3.5578176975250244, "step": 1225 }, { "epoch": 0.45241939203837817, "grad_norm": 1.5524929761886597, "learning_rate": 0.0001, "loss": 3.3967649936676025, "step": 1226 }, { "epoch": 0.4527884127496656, "grad_norm": 1.8354557752609253, "learning_rate": 0.0001, "loss": 3.3173229694366455, "step": 1227 }, { "epoch": 0.453157433460953, "grad_norm": 1.737549066543579, "learning_rate": 0.0001, "loss": 3.4682083129882812, "step": 1228 }, { "epoch": 0.4535264541722404, "grad_norm": 1.562452793121338, "learning_rate": 0.0001, "loss": 3.33489990234375, "step": 1229 }, { "epoch": 0.4538954748835278, "grad_norm": 1.4506992101669312, "learning_rate": 0.0001, "loss": 3.272857666015625, "step": 1230 }, { "epoch": 0.45426449559481524, "grad_norm": 1.9846397638320923, "learning_rate": 0.0001, "loss": 3.21801495552063, "step": 1231 }, { "epoch": 0.45463351630610266, "grad_norm": 1.5797758102416992, "learning_rate": 0.0001, "loss": 3.4130430221557617, "step": 1232 }, { "epoch": 0.4550025370173901, "grad_norm": 1.5246052742004395, "learning_rate": 0.0001, "loss": 2.9863009452819824, "step": 1233 }, { "epoch": 0.45537155772867755, "grad_norm": 1.662542462348938, "learning_rate": 0.0001, "loss": 3.292523145675659, "step": 1234 }, { "epoch": 0.45574057843996496, "grad_norm": 2.635794162750244, "learning_rate": 0.0001, "loss": 3.5330426692962646, "step": 1235 }, { "epoch": 0.4561095991512524, "grad_norm": 1.6771011352539062, "learning_rate": 0.0001, "loss": 3.298444986343384, "step": 1236 }, { "epoch": 0.4564786198625398, "grad_norm": 2.0715532302856445, "learning_rate": 0.0001, "loss": 3.4189984798431396, "step": 1237 }, { "epoch": 0.4568476405738272, "grad_norm": 1.4649382829666138, "learning_rate": 0.0001, "loss": 3.052089214324951, "step": 1238 }, { "epoch": 0.4572166612851146, "grad_norm": 1.495627522468567, "learning_rate": 0.0001, "loss": 3.410895347595215, "step": 1239 }, { "epoch": 0.45758568199640204, "grad_norm": 1.3375694751739502, "learning_rate": 0.0001, "loss": 3.2738590240478516, "step": 1240 }, { "epoch": 0.45795470270768945, "grad_norm": 1.6306648254394531, "learning_rate": 0.0001, "loss": 3.3279998302459717, "step": 1241 }, { "epoch": 0.45832372341897687, "grad_norm": 1.4996439218521118, "learning_rate": 0.0001, "loss": 3.3534674644470215, "step": 1242 }, { "epoch": 0.4586927441302643, "grad_norm": 1.701925277709961, "learning_rate": 0.0001, "loss": 3.2298202514648438, "step": 1243 }, { "epoch": 0.45906176484155176, "grad_norm": 2.032834529876709, "learning_rate": 0.0001, "loss": 3.2606616020202637, "step": 1244 }, { "epoch": 0.45943078555283917, "grad_norm": 2.1584107875823975, "learning_rate": 0.0001, "loss": 3.666701316833496, "step": 1245 }, { "epoch": 0.4597998062641266, "grad_norm": 2.1172866821289062, "learning_rate": 0.0001, "loss": 3.753302574157715, "step": 1246 }, { "epoch": 0.460168826975414, "grad_norm": 2.0513622760772705, "learning_rate": 0.0001, "loss": 3.48362398147583, "step": 1247 }, { "epoch": 0.4605378476867014, "grad_norm": 1.8232686519622803, "learning_rate": 0.0001, "loss": 3.746429204940796, "step": 1248 }, { "epoch": 0.46090686839798883, "grad_norm": 1.693013310432434, "learning_rate": 0.0001, "loss": 3.8272290229797363, "step": 1249 }, { "epoch": 0.46127588910927625, "grad_norm": 1.5011879205703735, "learning_rate": 0.0001, "loss": 3.2806854248046875, "step": 1250 }, { "epoch": 0.46164490982056366, "grad_norm": 1.671715259552002, "learning_rate": 0.0001, "loss": 3.3165080547332764, "step": 1251 }, { "epoch": 0.4620139305318511, "grad_norm": 1.3232616186141968, "learning_rate": 0.0001, "loss": 3.110487699508667, "step": 1252 }, { "epoch": 0.4623829512431385, "grad_norm": 1.864763855934143, "learning_rate": 0.0001, "loss": 3.236311435699463, "step": 1253 }, { "epoch": 0.46275197195442597, "grad_norm": 2.454894542694092, "learning_rate": 0.0001, "loss": 3.6582586765289307, "step": 1254 }, { "epoch": 0.4631209926657134, "grad_norm": 1.8720130920410156, "learning_rate": 0.0001, "loss": 3.46455979347229, "step": 1255 }, { "epoch": 0.4634900133770008, "grad_norm": 1.5965237617492676, "learning_rate": 0.0001, "loss": 3.1986207962036133, "step": 1256 }, { "epoch": 0.4638590340882882, "grad_norm": 1.697632074356079, "learning_rate": 0.0001, "loss": 3.2565183639526367, "step": 1257 }, { "epoch": 0.46422805479957563, "grad_norm": 2.2041590213775635, "learning_rate": 0.0001, "loss": 3.4480080604553223, "step": 1258 }, { "epoch": 0.46459707551086304, "grad_norm": 1.6730577945709229, "learning_rate": 0.0001, "loss": 3.1107442378997803, "step": 1259 }, { "epoch": 0.46496609622215046, "grad_norm": 1.7489527463912964, "learning_rate": 0.0001, "loss": 3.313342571258545, "step": 1260 }, { "epoch": 0.4653351169334379, "grad_norm": 1.596701741218567, "learning_rate": 0.0001, "loss": 3.29213285446167, "step": 1261 }, { "epoch": 0.4657041376447253, "grad_norm": 1.2522304058074951, "learning_rate": 0.0001, "loss": 3.4877445697784424, "step": 1262 }, { "epoch": 0.4660731583560127, "grad_norm": 2.0764338970184326, "learning_rate": 0.0001, "loss": 3.0069572925567627, "step": 1263 }, { "epoch": 0.4664421790673002, "grad_norm": 2.137681007385254, "learning_rate": 0.0001, "loss": 3.4323811531066895, "step": 1264 }, { "epoch": 0.4668111997785876, "grad_norm": 1.4234166145324707, "learning_rate": 0.0001, "loss": 3.188433885574341, "step": 1265 }, { "epoch": 0.467180220489875, "grad_norm": 1.4589178562164307, "learning_rate": 0.0001, "loss": 3.2995429039001465, "step": 1266 }, { "epoch": 0.4675492412011624, "grad_norm": 1.3958278894424438, "learning_rate": 0.0001, "loss": 3.188019275665283, "step": 1267 }, { "epoch": 0.46791826191244984, "grad_norm": 2.1328070163726807, "learning_rate": 0.0001, "loss": 3.7158985137939453, "step": 1268 }, { "epoch": 0.46828728262373726, "grad_norm": 1.652588129043579, "learning_rate": 0.0001, "loss": 3.204348087310791, "step": 1269 }, { "epoch": 0.46865630333502467, "grad_norm": 1.8721615076065063, "learning_rate": 0.0001, "loss": 3.7498161792755127, "step": 1270 }, { "epoch": 0.4690253240463121, "grad_norm": 1.5559072494506836, "learning_rate": 0.0001, "loss": 3.393519401550293, "step": 1271 }, { "epoch": 0.4693943447575995, "grad_norm": 1.4716813564300537, "learning_rate": 0.0001, "loss": 2.8601036071777344, "step": 1272 }, { "epoch": 0.4697633654688869, "grad_norm": 1.797658920288086, "learning_rate": 0.0001, "loss": 3.5473480224609375, "step": 1273 }, { "epoch": 0.4701323861801744, "grad_norm": 1.7798032760620117, "learning_rate": 0.0001, "loss": 2.774407148361206, "step": 1274 }, { "epoch": 0.4705014068914618, "grad_norm": 1.511938452720642, "learning_rate": 0.0001, "loss": 3.343441963195801, "step": 1275 }, { "epoch": 0.4708704276027492, "grad_norm": 1.933148741722107, "learning_rate": 0.0001, "loss": 3.6382083892822266, "step": 1276 }, { "epoch": 0.47123944831403664, "grad_norm": 1.4988672733306885, "learning_rate": 0.0001, "loss": 3.4293012619018555, "step": 1277 }, { "epoch": 0.47160846902532405, "grad_norm": 1.7968255281448364, "learning_rate": 0.0001, "loss": 3.314704656600952, "step": 1278 }, { "epoch": 0.47197748973661147, "grad_norm": 1.4645870923995972, "learning_rate": 0.0001, "loss": 3.075185775756836, "step": 1279 }, { "epoch": 0.4723465104478989, "grad_norm": 1.8099424839019775, "learning_rate": 0.0001, "loss": 3.2947311401367188, "step": 1280 }, { "epoch": 0.4727155311591863, "grad_norm": 1.6789758205413818, "learning_rate": 0.0001, "loss": 3.0761632919311523, "step": 1281 }, { "epoch": 0.4730845518704737, "grad_norm": 1.6258026361465454, "learning_rate": 0.0001, "loss": 2.948410749435425, "step": 1282 }, { "epoch": 0.47345357258176113, "grad_norm": 1.705978274345398, "learning_rate": 0.0001, "loss": 3.57371187210083, "step": 1283 }, { "epoch": 0.4738225932930486, "grad_norm": 1.559913992881775, "learning_rate": 0.0001, "loss": 3.2678418159484863, "step": 1284 }, { "epoch": 0.474191614004336, "grad_norm": 1.6342334747314453, "learning_rate": 0.0001, "loss": 3.1184093952178955, "step": 1285 }, { "epoch": 0.47456063471562343, "grad_norm": 2.2108042240142822, "learning_rate": 0.0001, "loss": 3.415156602859497, "step": 1286 }, { "epoch": 0.47492965542691085, "grad_norm": 1.5237102508544922, "learning_rate": 0.0001, "loss": 3.342151165008545, "step": 1287 }, { "epoch": 0.47529867613819826, "grad_norm": 1.5422735214233398, "learning_rate": 0.0001, "loss": 3.4816832542419434, "step": 1288 }, { "epoch": 0.4756676968494857, "grad_norm": 1.3541908264160156, "learning_rate": 0.0001, "loss": 3.1162476539611816, "step": 1289 }, { "epoch": 0.4760367175607731, "grad_norm": 1.6186414957046509, "learning_rate": 0.0001, "loss": 3.2205629348754883, "step": 1290 }, { "epoch": 0.4764057382720605, "grad_norm": 1.4707603454589844, "learning_rate": 0.0001, "loss": 3.4080612659454346, "step": 1291 }, { "epoch": 0.4767747589833479, "grad_norm": 2.5382943153381348, "learning_rate": 0.0001, "loss": 3.398341417312622, "step": 1292 }, { "epoch": 0.47714377969463534, "grad_norm": 1.613637924194336, "learning_rate": 0.0001, "loss": 3.058744192123413, "step": 1293 }, { "epoch": 0.47751280040592275, "grad_norm": 1.4283666610717773, "learning_rate": 0.0001, "loss": 3.147404909133911, "step": 1294 }, { "epoch": 0.4778818211172102, "grad_norm": 1.7146626710891724, "learning_rate": 0.0001, "loss": 3.1472527980804443, "step": 1295 }, { "epoch": 0.47825084182849764, "grad_norm": 1.811388373374939, "learning_rate": 0.0001, "loss": 3.1171460151672363, "step": 1296 }, { "epoch": 0.47861986253978506, "grad_norm": 1.7085700035095215, "learning_rate": 0.0001, "loss": 3.4158060550689697, "step": 1297 }, { "epoch": 0.4789888832510725, "grad_norm": 1.8744767904281616, "learning_rate": 0.0001, "loss": 3.0952372550964355, "step": 1298 }, { "epoch": 0.4793579039623599, "grad_norm": 1.456732153892517, "learning_rate": 0.0001, "loss": 3.1159238815307617, "step": 1299 }, { "epoch": 0.4797269246736473, "grad_norm": 2.131995916366577, "learning_rate": 0.0001, "loss": 3.553405523300171, "step": 1300 }, { "epoch": 0.4800959453849347, "grad_norm": 1.3812084197998047, "learning_rate": 0.0001, "loss": 3.423631191253662, "step": 1301 }, { "epoch": 0.48046496609622213, "grad_norm": 1.6513025760650635, "learning_rate": 0.0001, "loss": 3.4310293197631836, "step": 1302 }, { "epoch": 0.48083398680750955, "grad_norm": 1.5494147539138794, "learning_rate": 0.0001, "loss": 3.401501178741455, "step": 1303 }, { "epoch": 0.48120300751879697, "grad_norm": 1.7790298461914062, "learning_rate": 0.0001, "loss": 3.169189214706421, "step": 1304 }, { "epoch": 0.48157202823008444, "grad_norm": 1.7407560348510742, "learning_rate": 0.0001, "loss": 3.7418973445892334, "step": 1305 }, { "epoch": 0.48194104894137185, "grad_norm": 1.738014817237854, "learning_rate": 0.0001, "loss": 3.295907735824585, "step": 1306 }, { "epoch": 0.48231006965265927, "grad_norm": 1.874145269393921, "learning_rate": 0.0001, "loss": 3.5947012901306152, "step": 1307 }, { "epoch": 0.4826790903639467, "grad_norm": 1.4572949409484863, "learning_rate": 0.0001, "loss": 3.1663944721221924, "step": 1308 }, { "epoch": 0.4830481110752341, "grad_norm": 1.5112180709838867, "learning_rate": 0.0001, "loss": 3.084444999694824, "step": 1309 }, { "epoch": 0.4834171317865215, "grad_norm": 1.6459461450576782, "learning_rate": 0.0001, "loss": 3.146885395050049, "step": 1310 }, { "epoch": 0.48378615249780893, "grad_norm": 2.100034475326538, "learning_rate": 0.0001, "loss": 3.2717232704162598, "step": 1311 }, { "epoch": 0.48415517320909635, "grad_norm": 1.7134156227111816, "learning_rate": 0.0001, "loss": 3.4034125804901123, "step": 1312 }, { "epoch": 0.48452419392038376, "grad_norm": 1.7136698961257935, "learning_rate": 0.0001, "loss": 3.1172027587890625, "step": 1313 }, { "epoch": 0.4848932146316712, "grad_norm": 1.878926157951355, "learning_rate": 0.0001, "loss": 3.0406594276428223, "step": 1314 }, { "epoch": 0.48526223534295865, "grad_norm": 1.5215249061584473, "learning_rate": 0.0001, "loss": 3.215932607650757, "step": 1315 }, { "epoch": 0.48563125605424606, "grad_norm": 1.617172122001648, "learning_rate": 0.0001, "loss": 2.9702274799346924, "step": 1316 }, { "epoch": 0.4860002767655335, "grad_norm": 1.7017861604690552, "learning_rate": 0.0001, "loss": 3.3277804851531982, "step": 1317 }, { "epoch": 0.4863692974768209, "grad_norm": 1.494105577468872, "learning_rate": 0.0001, "loss": 2.9950363636016846, "step": 1318 }, { "epoch": 0.4867383181881083, "grad_norm": 1.619275689125061, "learning_rate": 0.0001, "loss": 3.1625025272369385, "step": 1319 }, { "epoch": 0.4871073388993957, "grad_norm": 1.662200689315796, "learning_rate": 0.0001, "loss": 3.1082229614257812, "step": 1320 }, { "epoch": 0.48747635961068314, "grad_norm": 1.9424527883529663, "learning_rate": 0.0001, "loss": 3.2635717391967773, "step": 1321 }, { "epoch": 0.48784538032197056, "grad_norm": 1.6934415102005005, "learning_rate": 0.0001, "loss": 3.1008872985839844, "step": 1322 }, { "epoch": 0.48821440103325797, "grad_norm": 1.7076951265335083, "learning_rate": 0.0001, "loss": 3.237067699432373, "step": 1323 }, { "epoch": 0.4885834217445454, "grad_norm": 1.3889529705047607, "learning_rate": 0.0001, "loss": 3.492698907852173, "step": 1324 }, { "epoch": 0.48895244245583286, "grad_norm": 1.8343408107757568, "learning_rate": 0.0001, "loss": 3.2396481037139893, "step": 1325 }, { "epoch": 0.4893214631671203, "grad_norm": 1.6344548463821411, "learning_rate": 0.0001, "loss": 3.3105108737945557, "step": 1326 }, { "epoch": 0.4896904838784077, "grad_norm": 1.5573179721832275, "learning_rate": 0.0001, "loss": 3.3734488487243652, "step": 1327 }, { "epoch": 0.4900595045896951, "grad_norm": 1.9833474159240723, "learning_rate": 0.0001, "loss": 3.2984347343444824, "step": 1328 }, { "epoch": 0.4904285253009825, "grad_norm": 1.6462719440460205, "learning_rate": 0.0001, "loss": 3.137624740600586, "step": 1329 }, { "epoch": 0.49079754601226994, "grad_norm": 1.767473816871643, "learning_rate": 0.0001, "loss": 3.3016762733459473, "step": 1330 }, { "epoch": 0.49116656672355735, "grad_norm": 1.824905514717102, "learning_rate": 0.0001, "loss": 3.1734776496887207, "step": 1331 }, { "epoch": 0.49153558743484477, "grad_norm": 1.9755927324295044, "learning_rate": 0.0001, "loss": 3.2039339542388916, "step": 1332 }, { "epoch": 0.4919046081461322, "grad_norm": 2.6448943614959717, "learning_rate": 0.0001, "loss": 3.19694185256958, "step": 1333 }, { "epoch": 0.4922736288574196, "grad_norm": 1.4271053075790405, "learning_rate": 0.0001, "loss": 2.8365516662597656, "step": 1334 }, { "epoch": 0.49264264956870707, "grad_norm": 1.7595093250274658, "learning_rate": 0.0001, "loss": 3.4585306644439697, "step": 1335 }, { "epoch": 0.4930116702799945, "grad_norm": 2.378689765930176, "learning_rate": 0.0001, "loss": 3.1091604232788086, "step": 1336 }, { "epoch": 0.4933806909912819, "grad_norm": 1.5055756568908691, "learning_rate": 0.0001, "loss": 3.509739637374878, "step": 1337 }, { "epoch": 0.4937497117025693, "grad_norm": 1.7014816999435425, "learning_rate": 0.0001, "loss": 3.200786590576172, "step": 1338 }, { "epoch": 0.49411873241385673, "grad_norm": 1.4998637437820435, "learning_rate": 0.0001, "loss": 3.1514382362365723, "step": 1339 }, { "epoch": 0.49448775312514415, "grad_norm": 1.6783392429351807, "learning_rate": 0.0001, "loss": 3.6204285621643066, "step": 1340 }, { "epoch": 0.49485677383643156, "grad_norm": 1.9601293802261353, "learning_rate": 0.0001, "loss": 3.3752455711364746, "step": 1341 }, { "epoch": 0.495225794547719, "grad_norm": 1.6793802976608276, "learning_rate": 0.0001, "loss": 3.260241746902466, "step": 1342 }, { "epoch": 0.4955948152590064, "grad_norm": 1.6954846382141113, "learning_rate": 0.0001, "loss": 3.2194817066192627, "step": 1343 }, { "epoch": 0.4959638359702938, "grad_norm": 1.7891285419464111, "learning_rate": 0.0001, "loss": 3.294940233230591, "step": 1344 }, { "epoch": 0.4963328566815813, "grad_norm": 1.7117921113967896, "learning_rate": 0.0001, "loss": 3.281766414642334, "step": 1345 }, { "epoch": 0.4967018773928687, "grad_norm": 1.4939122200012207, "learning_rate": 0.0001, "loss": 3.192007064819336, "step": 1346 }, { "epoch": 0.4970708981041561, "grad_norm": 1.9519555568695068, "learning_rate": 0.0001, "loss": 3.4541361331939697, "step": 1347 }, { "epoch": 0.4974399188154435, "grad_norm": 1.585076928138733, "learning_rate": 0.0001, "loss": 3.1069955825805664, "step": 1348 }, { "epoch": 0.49780893952673094, "grad_norm": 1.6752030849456787, "learning_rate": 0.0001, "loss": 3.2131295204162598, "step": 1349 }, { "epoch": 0.49817796023801836, "grad_norm": 1.3977504968643188, "learning_rate": 0.0001, "loss": 3.1420865058898926, "step": 1350 }, { "epoch": 0.4985469809493058, "grad_norm": 1.845415472984314, "learning_rate": 0.0001, "loss": 3.4622037410736084, "step": 1351 }, { "epoch": 0.4989160016605932, "grad_norm": 1.5891289710998535, "learning_rate": 0.0001, "loss": 3.0782864093780518, "step": 1352 }, { "epoch": 0.4992850223718806, "grad_norm": 1.5672197341918945, "learning_rate": 0.0001, "loss": 3.485478401184082, "step": 1353 }, { "epoch": 0.499654043083168, "grad_norm": 1.501745343208313, "learning_rate": 0.0001, "loss": 3.184842586517334, "step": 1354 }, { "epoch": 0.5000230637944555, "grad_norm": 1.6000661849975586, "learning_rate": 0.0001, "loss": 3.1798150539398193, "step": 1355 }, { "epoch": 0.5003920845057429, "grad_norm": 1.5391520261764526, "learning_rate": 0.0001, "loss": 3.2730629444122314, "step": 1356 }, { "epoch": 0.5007611052170303, "grad_norm": 1.7756389379501343, "learning_rate": 0.0001, "loss": 3.3823647499084473, "step": 1357 }, { "epoch": 0.5011301259283177, "grad_norm": 1.656735897064209, "learning_rate": 0.0001, "loss": 3.2115426063537598, "step": 1358 }, { "epoch": 0.5014991466396052, "grad_norm": 1.5626742839813232, "learning_rate": 0.0001, "loss": 3.0972936153411865, "step": 1359 }, { "epoch": 0.5018681673508926, "grad_norm": 1.5979243516921997, "learning_rate": 0.0001, "loss": 3.2601230144500732, "step": 1360 }, { "epoch": 0.50223718806218, "grad_norm": 1.7928996086120605, "learning_rate": 0.0001, "loss": 3.3743858337402344, "step": 1361 }, { "epoch": 0.5026062087734674, "grad_norm": 1.6128671169281006, "learning_rate": 0.0001, "loss": 3.0439889430999756, "step": 1362 }, { "epoch": 0.5029752294847548, "grad_norm": 2.1466379165649414, "learning_rate": 0.0001, "loss": 3.456559658050537, "step": 1363 }, { "epoch": 0.5033442501960422, "grad_norm": 1.4360712766647339, "learning_rate": 0.0001, "loss": 3.4028847217559814, "step": 1364 }, { "epoch": 0.5037132709073296, "grad_norm": 1.6191432476043701, "learning_rate": 0.0001, "loss": 3.1056792736053467, "step": 1365 }, { "epoch": 0.5040822916186171, "grad_norm": 1.416353464126587, "learning_rate": 0.0001, "loss": 2.933607578277588, "step": 1366 }, { "epoch": 0.5044513123299045, "grad_norm": 1.6281343698501587, "learning_rate": 0.0001, "loss": 3.2463326454162598, "step": 1367 }, { "epoch": 0.5048203330411919, "grad_norm": 1.6688257455825806, "learning_rate": 0.0001, "loss": 3.1200292110443115, "step": 1368 }, { "epoch": 0.5051893537524793, "grad_norm": 1.4505757093429565, "learning_rate": 0.0001, "loss": 3.417482852935791, "step": 1369 }, { "epoch": 0.5055583744637667, "grad_norm": 1.6983534097671509, "learning_rate": 0.0001, "loss": 3.159592390060425, "step": 1370 }, { "epoch": 0.5059273951750543, "grad_norm": 1.8584877252578735, "learning_rate": 0.0001, "loss": 3.0950493812561035, "step": 1371 }, { "epoch": 0.5062964158863417, "grad_norm": 1.5846456289291382, "learning_rate": 0.0001, "loss": 3.389085531234741, "step": 1372 }, { "epoch": 0.5066654365976291, "grad_norm": 1.7630212306976318, "learning_rate": 0.0001, "loss": 3.5320043563842773, "step": 1373 }, { "epoch": 0.5070344573089165, "grad_norm": 1.4512522220611572, "learning_rate": 0.0001, "loss": 3.7379817962646484, "step": 1374 }, { "epoch": 0.5074034780202039, "grad_norm": 1.6806732416152954, "learning_rate": 0.0001, "loss": 3.149183750152588, "step": 1375 }, { "epoch": 0.5077724987314913, "grad_norm": 1.5226467847824097, "learning_rate": 0.0001, "loss": 3.300114154815674, "step": 1376 }, { "epoch": 0.5081415194427787, "grad_norm": 1.5075722932815552, "learning_rate": 0.0001, "loss": 3.723101854324341, "step": 1377 }, { "epoch": 0.5085105401540662, "grad_norm": 1.9363282918930054, "learning_rate": 0.0001, "loss": 3.1966476440429688, "step": 1378 }, { "epoch": 0.5088795608653536, "grad_norm": 1.7270959615707397, "learning_rate": 0.0001, "loss": 3.3071601390838623, "step": 1379 }, { "epoch": 0.509248581576641, "grad_norm": 1.9508857727050781, "learning_rate": 0.0001, "loss": 3.369410991668701, "step": 1380 }, { "epoch": 0.5096176022879284, "grad_norm": 1.787837028503418, "learning_rate": 0.0001, "loss": 3.188244104385376, "step": 1381 }, { "epoch": 0.5099866229992158, "grad_norm": 1.6709396839141846, "learning_rate": 0.0001, "loss": 3.5337071418762207, "step": 1382 }, { "epoch": 0.5103556437105032, "grad_norm": 1.329795241355896, "learning_rate": 0.0001, "loss": 2.846888542175293, "step": 1383 }, { "epoch": 0.5107246644217907, "grad_norm": 1.56487238407135, "learning_rate": 0.0001, "loss": 3.261615753173828, "step": 1384 }, { "epoch": 0.5110936851330781, "grad_norm": 1.4400935173034668, "learning_rate": 0.0001, "loss": 3.0846471786499023, "step": 1385 }, { "epoch": 0.5114627058443655, "grad_norm": 1.5131444931030273, "learning_rate": 0.0001, "loss": 2.98460054397583, "step": 1386 }, { "epoch": 0.5118317265556529, "grad_norm": 2.2951855659484863, "learning_rate": 0.0001, "loss": 3.0500059127807617, "step": 1387 }, { "epoch": 0.5122007472669403, "grad_norm": 1.9929873943328857, "learning_rate": 0.0001, "loss": 3.276970148086548, "step": 1388 }, { "epoch": 0.5125697679782277, "grad_norm": 1.3621002435684204, "learning_rate": 0.0001, "loss": 3.1971092224121094, "step": 1389 }, { "epoch": 0.5129387886895151, "grad_norm": 1.5580744743347168, "learning_rate": 0.0001, "loss": 3.28003191947937, "step": 1390 }, { "epoch": 0.5133078094008027, "grad_norm": 1.6657811403274536, "learning_rate": 0.0001, "loss": 2.7580995559692383, "step": 1391 }, { "epoch": 0.5136768301120901, "grad_norm": 1.5088671445846558, "learning_rate": 0.0001, "loss": 3.0004076957702637, "step": 1392 }, { "epoch": 0.5140458508233775, "grad_norm": 1.492072343826294, "learning_rate": 0.0001, "loss": 2.8936853408813477, "step": 1393 }, { "epoch": 0.5144148715346649, "grad_norm": 1.3926286697387695, "learning_rate": 0.0001, "loss": 3.16294002532959, "step": 1394 }, { "epoch": 0.5147838922459523, "grad_norm": 1.8348246812820435, "learning_rate": 0.0001, "loss": 3.0954291820526123, "step": 1395 }, { "epoch": 0.5151529129572398, "grad_norm": 1.6425052881240845, "learning_rate": 0.0001, "loss": 3.18729305267334, "step": 1396 }, { "epoch": 0.5155219336685272, "grad_norm": 1.539875864982605, "learning_rate": 0.0001, "loss": 3.116133451461792, "step": 1397 }, { "epoch": 0.5158909543798146, "grad_norm": 1.637787938117981, "learning_rate": 0.0001, "loss": 3.2384722232818604, "step": 1398 }, { "epoch": 0.516259975091102, "grad_norm": 1.4627083539962769, "learning_rate": 0.0001, "loss": 3.5018155574798584, "step": 1399 }, { "epoch": 0.5166289958023894, "grad_norm": 1.4987012147903442, "learning_rate": 0.0001, "loss": 3.1494905948638916, "step": 1400 }, { "epoch": 0.5169980165136768, "grad_norm": 1.7869046926498413, "learning_rate": 0.0001, "loss": 3.004969596862793, "step": 1401 }, { "epoch": 0.5173670372249642, "grad_norm": 1.725035309791565, "learning_rate": 0.0001, "loss": 3.2955503463745117, "step": 1402 }, { "epoch": 0.5177360579362517, "grad_norm": 1.5997917652130127, "learning_rate": 0.0001, "loss": 3.1477222442626953, "step": 1403 }, { "epoch": 0.5181050786475391, "grad_norm": 1.5094619989395142, "learning_rate": 0.0001, "loss": 3.082007646560669, "step": 1404 }, { "epoch": 0.5184740993588265, "grad_norm": 1.5752825736999512, "learning_rate": 0.0001, "loss": 3.4254229068756104, "step": 1405 }, { "epoch": 0.5188431200701139, "grad_norm": 2.708827495574951, "learning_rate": 0.0001, "loss": 3.7292110919952393, "step": 1406 }, { "epoch": 0.5192121407814013, "grad_norm": 1.5606101751327515, "learning_rate": 0.0001, "loss": 3.2026309967041016, "step": 1407 }, { "epoch": 0.5195811614926887, "grad_norm": 1.5307283401489258, "learning_rate": 0.0001, "loss": 3.1685783863067627, "step": 1408 }, { "epoch": 0.5199501822039762, "grad_norm": 1.5207116603851318, "learning_rate": 0.0001, "loss": 3.3560218811035156, "step": 1409 }, { "epoch": 0.5203192029152636, "grad_norm": 1.7686904668807983, "learning_rate": 0.0001, "loss": 3.2387309074401855, "step": 1410 }, { "epoch": 0.5206882236265511, "grad_norm": 1.7233319282531738, "learning_rate": 0.0001, "loss": 3.2740843296051025, "step": 1411 }, { "epoch": 0.5210572443378385, "grad_norm": 1.8312103748321533, "learning_rate": 0.0001, "loss": 3.4070029258728027, "step": 1412 }, { "epoch": 0.5214262650491259, "grad_norm": 1.2560760974884033, "learning_rate": 0.0001, "loss": 3.366732120513916, "step": 1413 }, { "epoch": 0.5217952857604133, "grad_norm": 1.5106955766677856, "learning_rate": 0.0001, "loss": 3.118375301361084, "step": 1414 }, { "epoch": 0.5221643064717008, "grad_norm": 1.7871521711349487, "learning_rate": 0.0001, "loss": 2.8388583660125732, "step": 1415 }, { "epoch": 0.5225333271829882, "grad_norm": 1.5334633588790894, "learning_rate": 0.0001, "loss": 3.2323946952819824, "step": 1416 }, { "epoch": 0.5229023478942756, "grad_norm": 1.470723271369934, "learning_rate": 0.0001, "loss": 3.4212357997894287, "step": 1417 }, { "epoch": 0.523271368605563, "grad_norm": 1.4914551973342896, "learning_rate": 0.0001, "loss": 3.307647466659546, "step": 1418 }, { "epoch": 0.5236403893168504, "grad_norm": 1.6958730220794678, "learning_rate": 0.0001, "loss": 3.6064822673797607, "step": 1419 }, { "epoch": 0.5240094100281378, "grad_norm": 1.5129966735839844, "learning_rate": 0.0001, "loss": 2.746400833129883, "step": 1420 }, { "epoch": 0.5243784307394252, "grad_norm": 1.5086040496826172, "learning_rate": 0.0001, "loss": 2.969712257385254, "step": 1421 }, { "epoch": 0.5247474514507127, "grad_norm": 1.791477918624878, "learning_rate": 0.0001, "loss": 3.373394250869751, "step": 1422 }, { "epoch": 0.5251164721620001, "grad_norm": 1.7548803091049194, "learning_rate": 0.0001, "loss": 2.970478057861328, "step": 1423 }, { "epoch": 0.5254854928732875, "grad_norm": 1.8303626775741577, "learning_rate": 0.0001, "loss": 3.006618022918701, "step": 1424 }, { "epoch": 0.5258545135845749, "grad_norm": 1.701167106628418, "learning_rate": 0.0001, "loss": 3.3926918506622314, "step": 1425 }, { "epoch": 0.5262235342958623, "grad_norm": 1.4916001558303833, "learning_rate": 0.0001, "loss": 2.933903932571411, "step": 1426 }, { "epoch": 0.5265925550071497, "grad_norm": 1.3432062864303589, "learning_rate": 0.0001, "loss": 3.294658660888672, "step": 1427 }, { "epoch": 0.5269615757184372, "grad_norm": 1.585404872894287, "learning_rate": 0.0001, "loss": 3.1204111576080322, "step": 1428 }, { "epoch": 0.5273305964297246, "grad_norm": 1.9491335153579712, "learning_rate": 0.0001, "loss": 3.3216922283172607, "step": 1429 }, { "epoch": 0.527699617141012, "grad_norm": 1.597486138343811, "learning_rate": 0.0001, "loss": 3.216310977935791, "step": 1430 }, { "epoch": 0.5280686378522994, "grad_norm": 2.0212347507476807, "learning_rate": 0.0001, "loss": 3.586266040802002, "step": 1431 }, { "epoch": 0.5284376585635869, "grad_norm": 1.457519769668579, "learning_rate": 0.0001, "loss": 3.044626235961914, "step": 1432 }, { "epoch": 0.5288066792748743, "grad_norm": 1.591465711593628, "learning_rate": 0.0001, "loss": 3.196074962615967, "step": 1433 }, { "epoch": 0.5291756999861618, "grad_norm": 1.8396682739257812, "learning_rate": 0.0001, "loss": 3.0668563842773438, "step": 1434 }, { "epoch": 0.5295447206974492, "grad_norm": 1.6969633102416992, "learning_rate": 0.0001, "loss": 3.1155049800872803, "step": 1435 }, { "epoch": 0.5299137414087366, "grad_norm": 1.395924687385559, "learning_rate": 0.0001, "loss": 2.8550267219543457, "step": 1436 }, { "epoch": 0.530282762120024, "grad_norm": 1.5051054954528809, "learning_rate": 0.0001, "loss": 3.4201347827911377, "step": 1437 }, { "epoch": 0.5306517828313114, "grad_norm": 1.5168818235397339, "learning_rate": 0.0001, "loss": 3.2115895748138428, "step": 1438 }, { "epoch": 0.5310208035425988, "grad_norm": 1.6491508483886719, "learning_rate": 0.0001, "loss": 3.357948064804077, "step": 1439 }, { "epoch": 0.5313898242538863, "grad_norm": 1.6566519737243652, "learning_rate": 0.0001, "loss": 3.1956992149353027, "step": 1440 }, { "epoch": 0.5317588449651737, "grad_norm": 1.888230323791504, "learning_rate": 0.0001, "loss": 3.300771713256836, "step": 1441 }, { "epoch": 0.5321278656764611, "grad_norm": 1.5399962663650513, "learning_rate": 0.0001, "loss": 3.1413931846618652, "step": 1442 }, { "epoch": 0.5324968863877485, "grad_norm": 1.7434382438659668, "learning_rate": 0.0001, "loss": 3.124591827392578, "step": 1443 }, { "epoch": 0.5328659070990359, "grad_norm": 1.5932893753051758, "learning_rate": 0.0001, "loss": 3.281287431716919, "step": 1444 }, { "epoch": 0.5332349278103233, "grad_norm": 1.825126051902771, "learning_rate": 0.0001, "loss": 3.75779390335083, "step": 1445 }, { "epoch": 0.5336039485216107, "grad_norm": 1.5416094064712524, "learning_rate": 0.0001, "loss": 3.127620220184326, "step": 1446 }, { "epoch": 0.5339729692328982, "grad_norm": 1.6902350187301636, "learning_rate": 0.0001, "loss": 3.11095929145813, "step": 1447 }, { "epoch": 0.5343419899441856, "grad_norm": 1.4231213331222534, "learning_rate": 0.0001, "loss": 3.3785438537597656, "step": 1448 }, { "epoch": 0.534711010655473, "grad_norm": 1.4861929416656494, "learning_rate": 0.0001, "loss": 3.5403685569763184, "step": 1449 }, { "epoch": 0.5350800313667604, "grad_norm": 1.7890992164611816, "learning_rate": 0.0001, "loss": 3.3114781379699707, "step": 1450 }, { "epoch": 0.5354490520780478, "grad_norm": 1.5725735425949097, "learning_rate": 0.0001, "loss": 3.291184425354004, "step": 1451 }, { "epoch": 0.5358180727893354, "grad_norm": 1.422446608543396, "learning_rate": 0.0001, "loss": 2.9178993701934814, "step": 1452 }, { "epoch": 0.5361870935006228, "grad_norm": 1.6171269416809082, "learning_rate": 0.0001, "loss": 3.1636667251586914, "step": 1453 }, { "epoch": 0.5365561142119102, "grad_norm": 1.600523829460144, "learning_rate": 0.0001, "loss": 3.0755040645599365, "step": 1454 }, { "epoch": 0.5369251349231976, "grad_norm": 1.3465650081634521, "learning_rate": 0.0001, "loss": 2.929584264755249, "step": 1455 }, { "epoch": 0.537294155634485, "grad_norm": 1.429313063621521, "learning_rate": 0.0001, "loss": 3.4685211181640625, "step": 1456 }, { "epoch": 0.5376631763457724, "grad_norm": 1.4593101739883423, "learning_rate": 0.0001, "loss": 3.0087366104125977, "step": 1457 }, { "epoch": 0.5380321970570598, "grad_norm": 1.438327670097351, "learning_rate": 0.0001, "loss": 3.515387535095215, "step": 1458 }, { "epoch": 0.5384012177683473, "grad_norm": 1.507686734199524, "learning_rate": 0.0001, "loss": 3.138439655303955, "step": 1459 }, { "epoch": 0.5387702384796347, "grad_norm": 1.5899978876113892, "learning_rate": 0.0001, "loss": 3.0492749214172363, "step": 1460 }, { "epoch": 0.5391392591909221, "grad_norm": 1.3913685083389282, "learning_rate": 0.0001, "loss": 2.9755537509918213, "step": 1461 }, { "epoch": 0.5395082799022095, "grad_norm": 1.577911376953125, "learning_rate": 0.0001, "loss": 3.5656211376190186, "step": 1462 }, { "epoch": 0.5398773006134969, "grad_norm": 1.581742525100708, "learning_rate": 0.0001, "loss": 3.3091001510620117, "step": 1463 }, { "epoch": 0.5402463213247843, "grad_norm": 2.854262590408325, "learning_rate": 0.0001, "loss": 3.3481838703155518, "step": 1464 }, { "epoch": 0.5406153420360718, "grad_norm": 1.5929034948349, "learning_rate": 0.0001, "loss": 3.2897744178771973, "step": 1465 }, { "epoch": 0.5409843627473592, "grad_norm": 1.9537807703018188, "learning_rate": 0.0001, "loss": 3.1020584106445312, "step": 1466 }, { "epoch": 0.5413533834586466, "grad_norm": 1.3209524154663086, "learning_rate": 0.0001, "loss": 3.0436618328094482, "step": 1467 }, { "epoch": 0.541722404169934, "grad_norm": 1.7131558656692505, "learning_rate": 0.0001, "loss": 2.984874963760376, "step": 1468 }, { "epoch": 0.5420914248812214, "grad_norm": 1.6406959295272827, "learning_rate": 0.0001, "loss": 3.3179259300231934, "step": 1469 }, { "epoch": 0.5424604455925088, "grad_norm": 1.538968801498413, "learning_rate": 0.0001, "loss": 3.1853573322296143, "step": 1470 }, { "epoch": 0.5428294663037962, "grad_norm": 2.0462467670440674, "learning_rate": 0.0001, "loss": 2.9979822635650635, "step": 1471 }, { "epoch": 0.5431984870150838, "grad_norm": 1.4344154596328735, "learning_rate": 0.0001, "loss": 3.1466450691223145, "step": 1472 }, { "epoch": 0.5435675077263712, "grad_norm": 1.4658862352371216, "learning_rate": 0.0001, "loss": 3.171497106552124, "step": 1473 }, { "epoch": 0.5439365284376586, "grad_norm": 1.4925707578659058, "learning_rate": 0.0001, "loss": 2.8434243202209473, "step": 1474 }, { "epoch": 0.544305549148946, "grad_norm": 1.762581467628479, "learning_rate": 0.0001, "loss": 2.983776330947876, "step": 1475 }, { "epoch": 0.5446745698602334, "grad_norm": 1.7249791622161865, "learning_rate": 0.0001, "loss": 3.3636794090270996, "step": 1476 }, { "epoch": 0.5450435905715209, "grad_norm": 1.3278249502182007, "learning_rate": 0.0001, "loss": 3.0298595428466797, "step": 1477 }, { "epoch": 0.5454126112828083, "grad_norm": 1.7934025526046753, "learning_rate": 0.0001, "loss": 2.9716615676879883, "step": 1478 }, { "epoch": 0.5457816319940957, "grad_norm": 1.443535566329956, "learning_rate": 0.0001, "loss": 3.3245465755462646, "step": 1479 }, { "epoch": 0.5461506527053831, "grad_norm": 1.5035440921783447, "learning_rate": 0.0001, "loss": 3.420717239379883, "step": 1480 }, { "epoch": 0.5465196734166705, "grad_norm": 1.6783944368362427, "learning_rate": 0.0001, "loss": 3.2124505043029785, "step": 1481 }, { "epoch": 0.5468886941279579, "grad_norm": 2.403007984161377, "learning_rate": 0.0001, "loss": 3.3220090866088867, "step": 1482 }, { "epoch": 0.5472577148392453, "grad_norm": 1.3949731588363647, "learning_rate": 0.0001, "loss": 3.121206521987915, "step": 1483 }, { "epoch": 0.5476267355505328, "grad_norm": 2.056257724761963, "learning_rate": 0.0001, "loss": 3.088494300842285, "step": 1484 }, { "epoch": 0.5479957562618202, "grad_norm": 1.5680383443832397, "learning_rate": 0.0001, "loss": 3.3057167530059814, "step": 1485 }, { "epoch": 0.5483647769731076, "grad_norm": 1.5357822179794312, "learning_rate": 0.0001, "loss": 3.3089511394500732, "step": 1486 }, { "epoch": 0.548733797684395, "grad_norm": 1.4477072954177856, "learning_rate": 0.0001, "loss": 3.4524927139282227, "step": 1487 }, { "epoch": 0.5491028183956824, "grad_norm": 1.9374648332595825, "learning_rate": 0.0001, "loss": 3.079282760620117, "step": 1488 }, { "epoch": 0.5494718391069698, "grad_norm": 1.4118106365203857, "learning_rate": 0.0001, "loss": 3.2847325801849365, "step": 1489 }, { "epoch": 0.5498408598182573, "grad_norm": 1.7036107778549194, "learning_rate": 0.0001, "loss": 3.3690595626831055, "step": 1490 }, { "epoch": 0.5502098805295447, "grad_norm": 1.479461908340454, "learning_rate": 0.0001, "loss": 3.3903355598449707, "step": 1491 }, { "epoch": 0.5505789012408322, "grad_norm": 2.033634901046753, "learning_rate": 0.0001, "loss": 3.4432337284088135, "step": 1492 }, { "epoch": 0.5509479219521196, "grad_norm": 1.5934860706329346, "learning_rate": 0.0001, "loss": 3.5054030418395996, "step": 1493 }, { "epoch": 0.551316942663407, "grad_norm": 1.4428513050079346, "learning_rate": 0.0001, "loss": 3.186854362487793, "step": 1494 }, { "epoch": 0.5516859633746944, "grad_norm": 1.551553726196289, "learning_rate": 0.0001, "loss": 3.50541090965271, "step": 1495 }, { "epoch": 0.5520549840859819, "grad_norm": 1.4098210334777832, "learning_rate": 0.0001, "loss": 3.071147918701172, "step": 1496 }, { "epoch": 0.5524240047972693, "grad_norm": 1.4409326314926147, "learning_rate": 0.0001, "loss": 3.0468363761901855, "step": 1497 }, { "epoch": 0.5527930255085567, "grad_norm": 1.6490522623062134, "learning_rate": 0.0001, "loss": 2.8445770740509033, "step": 1498 }, { "epoch": 0.5531620462198441, "grad_norm": 1.6029343605041504, "learning_rate": 0.0001, "loss": 2.9250729084014893, "step": 1499 }, { "epoch": 0.5535310669311315, "grad_norm": 1.4397848844528198, "learning_rate": 0.0001, "loss": 2.9054367542266846, "step": 1500 }, { "epoch": 0.5539000876424189, "grad_norm": 2.1724696159362793, "learning_rate": 0.0001, "loss": 3.502490997314453, "step": 1501 }, { "epoch": 0.5542691083537064, "grad_norm": 1.6301262378692627, "learning_rate": 0.0001, "loss": 3.2971608638763428, "step": 1502 }, { "epoch": 0.5546381290649938, "grad_norm": 1.5430877208709717, "learning_rate": 0.0001, "loss": 3.3104028701782227, "step": 1503 }, { "epoch": 0.5550071497762812, "grad_norm": 2.0268802642822266, "learning_rate": 0.0001, "loss": 2.9657490253448486, "step": 1504 }, { "epoch": 0.5553761704875686, "grad_norm": 1.6418383121490479, "learning_rate": 0.0001, "loss": 2.919212818145752, "step": 1505 }, { "epoch": 0.555745191198856, "grad_norm": 1.56089186668396, "learning_rate": 0.0001, "loss": 3.3910744190216064, "step": 1506 }, { "epoch": 0.5561142119101434, "grad_norm": 1.918686032295227, "learning_rate": 0.0001, "loss": 3.430589437484741, "step": 1507 }, { "epoch": 0.5564832326214308, "grad_norm": 1.5203773975372314, "learning_rate": 0.0001, "loss": 2.7875938415527344, "step": 1508 }, { "epoch": 0.5568522533327183, "grad_norm": 1.5869956016540527, "learning_rate": 0.0001, "loss": 3.224703311920166, "step": 1509 }, { "epoch": 0.5572212740440057, "grad_norm": 1.5257827043533325, "learning_rate": 0.0001, "loss": 3.476515293121338, "step": 1510 }, { "epoch": 0.5575902947552931, "grad_norm": 1.8196038007736206, "learning_rate": 0.0001, "loss": 3.351560354232788, "step": 1511 }, { "epoch": 0.5579593154665805, "grad_norm": 1.5189015865325928, "learning_rate": 0.0001, "loss": 3.3092334270477295, "step": 1512 }, { "epoch": 0.558328336177868, "grad_norm": 1.7151753902435303, "learning_rate": 0.0001, "loss": 3.152954339981079, "step": 1513 }, { "epoch": 0.5586973568891554, "grad_norm": 1.8838587999343872, "learning_rate": 0.0001, "loss": 3.2369325160980225, "step": 1514 }, { "epoch": 0.5590663776004429, "grad_norm": 2.727656841278076, "learning_rate": 0.0001, "loss": 2.9606804847717285, "step": 1515 }, { "epoch": 0.5594353983117303, "grad_norm": 1.6646603345870972, "learning_rate": 0.0001, "loss": 3.1642611026763916, "step": 1516 }, { "epoch": 0.5598044190230177, "grad_norm": 1.8451595306396484, "learning_rate": 0.0001, "loss": 3.3373947143554688, "step": 1517 }, { "epoch": 0.5601734397343051, "grad_norm": 1.5362062454223633, "learning_rate": 0.0001, "loss": 3.4962422847747803, "step": 1518 }, { "epoch": 0.5605424604455925, "grad_norm": 1.6662501096725464, "learning_rate": 0.0001, "loss": 3.258713960647583, "step": 1519 }, { "epoch": 0.5609114811568799, "grad_norm": 1.7864031791687012, "learning_rate": 0.0001, "loss": 3.5283095836639404, "step": 1520 }, { "epoch": 0.5612805018681674, "grad_norm": 1.65171217918396, "learning_rate": 0.0001, "loss": 3.401099681854248, "step": 1521 }, { "epoch": 0.5616495225794548, "grad_norm": 1.6977293491363525, "learning_rate": 0.0001, "loss": 3.3018059730529785, "step": 1522 }, { "epoch": 0.5620185432907422, "grad_norm": 1.6299015283584595, "learning_rate": 0.0001, "loss": 3.206380605697632, "step": 1523 }, { "epoch": 0.5623875640020296, "grad_norm": 1.4896444082260132, "learning_rate": 0.0001, "loss": 3.0151915550231934, "step": 1524 }, { "epoch": 0.562756584713317, "grad_norm": 1.5741418600082397, "learning_rate": 0.0001, "loss": 3.190915822982788, "step": 1525 }, { "epoch": 0.5631256054246044, "grad_norm": 1.5395468473434448, "learning_rate": 0.0001, "loss": 3.4873619079589844, "step": 1526 }, { "epoch": 0.5634946261358919, "grad_norm": 1.3901787996292114, "learning_rate": 0.0001, "loss": 3.2513246536254883, "step": 1527 }, { "epoch": 0.5638636468471793, "grad_norm": 1.6285467147827148, "learning_rate": 0.0001, "loss": 3.060163736343384, "step": 1528 }, { "epoch": 0.5642326675584667, "grad_norm": 1.6209940910339355, "learning_rate": 0.0001, "loss": 3.3560142517089844, "step": 1529 }, { "epoch": 0.5646016882697541, "grad_norm": 1.4346884489059448, "learning_rate": 0.0001, "loss": 3.169527053833008, "step": 1530 }, { "epoch": 0.5649707089810415, "grad_norm": 1.3554576635360718, "learning_rate": 0.0001, "loss": 3.060452461242676, "step": 1531 }, { "epoch": 0.5653397296923289, "grad_norm": 1.5099626779556274, "learning_rate": 0.0001, "loss": 2.8378283977508545, "step": 1532 }, { "epoch": 0.5657087504036165, "grad_norm": 1.6148101091384888, "learning_rate": 0.0001, "loss": 3.1122426986694336, "step": 1533 }, { "epoch": 0.5660777711149039, "grad_norm": 1.4931827783584595, "learning_rate": 0.0001, "loss": 2.9703516960144043, "step": 1534 }, { "epoch": 0.5664467918261913, "grad_norm": 1.783033013343811, "learning_rate": 0.0001, "loss": 3.2727267742156982, "step": 1535 }, { "epoch": 0.5668158125374787, "grad_norm": 2.223917245864868, "learning_rate": 0.0001, "loss": 3.3129425048828125, "step": 1536 }, { "epoch": 0.5671848332487661, "grad_norm": 1.668046236038208, "learning_rate": 0.0001, "loss": 3.5251851081848145, "step": 1537 }, { "epoch": 0.5675538539600535, "grad_norm": 1.862939715385437, "learning_rate": 0.0001, "loss": 2.987452507019043, "step": 1538 }, { "epoch": 0.567922874671341, "grad_norm": 1.4585286378860474, "learning_rate": 0.0001, "loss": 3.062586784362793, "step": 1539 }, { "epoch": 0.5682918953826284, "grad_norm": 1.3981331586837769, "learning_rate": 0.0001, "loss": 2.743325710296631, "step": 1540 }, { "epoch": 0.5686609160939158, "grad_norm": 1.5181610584259033, "learning_rate": 0.0001, "loss": 2.822594165802002, "step": 1541 }, { "epoch": 0.5690299368052032, "grad_norm": 1.640688180923462, "learning_rate": 0.0001, "loss": 3.172366142272949, "step": 1542 }, { "epoch": 0.5693989575164906, "grad_norm": 1.3805067539215088, "learning_rate": 0.0001, "loss": 2.910876750946045, "step": 1543 }, { "epoch": 0.569767978227778, "grad_norm": 1.7557059526443481, "learning_rate": 0.0001, "loss": 3.28838849067688, "step": 1544 }, { "epoch": 0.5701369989390654, "grad_norm": 1.6887164115905762, "learning_rate": 0.0001, "loss": 3.1530814170837402, "step": 1545 }, { "epoch": 0.5705060196503529, "grad_norm": 1.6673730611801147, "learning_rate": 0.0001, "loss": 3.093174934387207, "step": 1546 }, { "epoch": 0.5708750403616403, "grad_norm": 1.336640477180481, "learning_rate": 0.0001, "loss": 3.144160747528076, "step": 1547 }, { "epoch": 0.5712440610729277, "grad_norm": 1.649889588356018, "learning_rate": 0.0001, "loss": 2.7244925498962402, "step": 1548 }, { "epoch": 0.5716130817842151, "grad_norm": 1.4036601781845093, "learning_rate": 0.0001, "loss": 3.098292112350464, "step": 1549 }, { "epoch": 0.5719821024955025, "grad_norm": 2.5909788608551025, "learning_rate": 0.0001, "loss": 3.8458776473999023, "step": 1550 }, { "epoch": 0.5723511232067899, "grad_norm": 1.773421287536621, "learning_rate": 0.0001, "loss": 3.2020580768585205, "step": 1551 }, { "epoch": 0.5727201439180774, "grad_norm": 1.646235704421997, "learning_rate": 0.0001, "loss": 3.2537198066711426, "step": 1552 }, { "epoch": 0.5730891646293649, "grad_norm": 1.675258994102478, "learning_rate": 0.0001, "loss": 3.177224636077881, "step": 1553 }, { "epoch": 0.5734581853406523, "grad_norm": 1.3590600490570068, "learning_rate": 0.0001, "loss": 3.351827383041382, "step": 1554 }, { "epoch": 0.5738272060519397, "grad_norm": 1.581157922744751, "learning_rate": 0.0001, "loss": 3.174504518508911, "step": 1555 }, { "epoch": 0.5741962267632271, "grad_norm": 1.4486459493637085, "learning_rate": 0.0001, "loss": 3.0384104251861572, "step": 1556 }, { "epoch": 0.5745652474745145, "grad_norm": 1.7779186964035034, "learning_rate": 0.0001, "loss": 3.4975197315216064, "step": 1557 }, { "epoch": 0.574934268185802, "grad_norm": 1.3707894086837769, "learning_rate": 0.0001, "loss": 3.2995235919952393, "step": 1558 }, { "epoch": 0.5753032888970894, "grad_norm": 1.482908010482788, "learning_rate": 0.0001, "loss": 3.2311654090881348, "step": 1559 }, { "epoch": 0.5756723096083768, "grad_norm": 1.84312105178833, "learning_rate": 0.0001, "loss": 3.221818208694458, "step": 1560 }, { "epoch": 0.5760413303196642, "grad_norm": 1.4545091390609741, "learning_rate": 0.0001, "loss": 2.983891487121582, "step": 1561 }, { "epoch": 0.5764103510309516, "grad_norm": 1.3163419961929321, "learning_rate": 0.0001, "loss": 2.9315831661224365, "step": 1562 }, { "epoch": 0.576779371742239, "grad_norm": 1.2499723434448242, "learning_rate": 0.0001, "loss": 3.234877586364746, "step": 1563 }, { "epoch": 0.5771483924535264, "grad_norm": 1.5027834177017212, "learning_rate": 0.0001, "loss": 2.8867790699005127, "step": 1564 }, { "epoch": 0.5775174131648139, "grad_norm": 1.5805798768997192, "learning_rate": 0.0001, "loss": 3.25894832611084, "step": 1565 }, { "epoch": 0.5778864338761013, "grad_norm": 1.6571667194366455, "learning_rate": 0.0001, "loss": 3.343813896179199, "step": 1566 }, { "epoch": 0.5782554545873887, "grad_norm": 1.502855658531189, "learning_rate": 0.0001, "loss": 2.9952352046966553, "step": 1567 }, { "epoch": 0.5786244752986761, "grad_norm": 1.7565134763717651, "learning_rate": 0.0001, "loss": 3.260211706161499, "step": 1568 }, { "epoch": 0.5789934960099635, "grad_norm": 1.5395530462265015, "learning_rate": 0.0001, "loss": 2.9537479877471924, "step": 1569 }, { "epoch": 0.5793625167212509, "grad_norm": 1.6251996755599976, "learning_rate": 0.0001, "loss": 3.01223087310791, "step": 1570 }, { "epoch": 0.5797315374325384, "grad_norm": 1.9269543886184692, "learning_rate": 0.0001, "loss": 2.8629322052001953, "step": 1571 }, { "epoch": 0.5801005581438258, "grad_norm": 1.5990525484085083, "learning_rate": 0.0001, "loss": 2.809880495071411, "step": 1572 }, { "epoch": 0.5804695788551132, "grad_norm": 1.7015825510025024, "learning_rate": 0.0001, "loss": 3.1098175048828125, "step": 1573 }, { "epoch": 0.5808385995664007, "grad_norm": 1.7375601530075073, "learning_rate": 0.0001, "loss": 3.3175439834594727, "step": 1574 }, { "epoch": 0.5812076202776881, "grad_norm": 2.362367868423462, "learning_rate": 0.0001, "loss": 3.3696765899658203, "step": 1575 }, { "epoch": 0.5815766409889755, "grad_norm": 1.6514980792999268, "learning_rate": 0.0001, "loss": 2.8957343101501465, "step": 1576 }, { "epoch": 0.581945661700263, "grad_norm": 1.5710346698760986, "learning_rate": 0.0001, "loss": 3.0906119346618652, "step": 1577 }, { "epoch": 0.5823146824115504, "grad_norm": 1.6373987197875977, "learning_rate": 0.0001, "loss": 2.9068775177001953, "step": 1578 }, { "epoch": 0.5826837031228378, "grad_norm": 1.839416265487671, "learning_rate": 0.0001, "loss": 3.4652509689331055, "step": 1579 }, { "epoch": 0.5830527238341252, "grad_norm": 1.5160094499588013, "learning_rate": 0.0001, "loss": 3.1616334915161133, "step": 1580 }, { "epoch": 0.5834217445454126, "grad_norm": 1.5923758745193481, "learning_rate": 0.0001, "loss": 3.136388063430786, "step": 1581 }, { "epoch": 0.5837907652567, "grad_norm": 1.3367705345153809, "learning_rate": 0.0001, "loss": 3.005518913269043, "step": 1582 }, { "epoch": 0.5841597859679875, "grad_norm": 1.609127163887024, "learning_rate": 0.0001, "loss": 3.0415806770324707, "step": 1583 }, { "epoch": 0.5845288066792749, "grad_norm": 1.5081133842468262, "learning_rate": 0.0001, "loss": 3.313835859298706, "step": 1584 }, { "epoch": 0.5848978273905623, "grad_norm": 1.6494060754776, "learning_rate": 0.0001, "loss": 3.002715826034546, "step": 1585 }, { "epoch": 0.5852668481018497, "grad_norm": 2.044844388961792, "learning_rate": 0.0001, "loss": 3.10986328125, "step": 1586 }, { "epoch": 0.5856358688131371, "grad_norm": 1.6699491739273071, "learning_rate": 0.0001, "loss": 3.198453187942505, "step": 1587 }, { "epoch": 0.5860048895244245, "grad_norm": 1.6649705171585083, "learning_rate": 0.0001, "loss": 2.894638776779175, "step": 1588 }, { "epoch": 0.586373910235712, "grad_norm": 1.3813105821609497, "learning_rate": 0.0001, "loss": 2.938124895095825, "step": 1589 }, { "epoch": 0.5867429309469994, "grad_norm": 1.322628140449524, "learning_rate": 0.0001, "loss": 2.6089675426483154, "step": 1590 }, { "epoch": 0.5871119516582868, "grad_norm": 2.5670278072357178, "learning_rate": 0.0001, "loss": 3.186948776245117, "step": 1591 }, { "epoch": 0.5874809723695742, "grad_norm": 2.190114974975586, "learning_rate": 0.0001, "loss": 3.2568819522857666, "step": 1592 }, { "epoch": 0.5878499930808616, "grad_norm": 2.129382610321045, "learning_rate": 0.0001, "loss": 3.095207691192627, "step": 1593 }, { "epoch": 0.5882190137921491, "grad_norm": 1.6611095666885376, "learning_rate": 0.0001, "loss": 3.344822406768799, "step": 1594 }, { "epoch": 0.5885880345034366, "grad_norm": 1.5086028575897217, "learning_rate": 0.0001, "loss": 3.1396355628967285, "step": 1595 }, { "epoch": 0.588957055214724, "grad_norm": 1.3295549154281616, "learning_rate": 0.0001, "loss": 3.216621160507202, "step": 1596 }, { "epoch": 0.5893260759260114, "grad_norm": 1.695566177368164, "learning_rate": 0.0001, "loss": 2.7655792236328125, "step": 1597 }, { "epoch": 0.5896950966372988, "grad_norm": 1.84371018409729, "learning_rate": 0.0001, "loss": 3.026766300201416, "step": 1598 }, { "epoch": 0.5900641173485862, "grad_norm": 1.5515944957733154, "learning_rate": 0.0001, "loss": 2.8108599185943604, "step": 1599 }, { "epoch": 0.5904331380598736, "grad_norm": 2.0425801277160645, "learning_rate": 0.0001, "loss": 3.213899612426758, "step": 1600 }, { "epoch": 0.590802158771161, "grad_norm": 1.398847222328186, "learning_rate": 0.0001, "loss": 3.180758476257324, "step": 1601 }, { "epoch": 0.5911711794824485, "grad_norm": 1.437021017074585, "learning_rate": 0.0001, "loss": 2.807854175567627, "step": 1602 }, { "epoch": 0.5915402001937359, "grad_norm": 1.4648010730743408, "learning_rate": 0.0001, "loss": 2.975890636444092, "step": 1603 }, { "epoch": 0.5919092209050233, "grad_norm": 1.6542235612869263, "learning_rate": 0.0001, "loss": 3.090987205505371, "step": 1604 }, { "epoch": 0.5922782416163107, "grad_norm": 1.6483943462371826, "learning_rate": 0.0001, "loss": 3.0860087871551514, "step": 1605 }, { "epoch": 0.5926472623275981, "grad_norm": 1.5324894189834595, "learning_rate": 0.0001, "loss": 3.231109142303467, "step": 1606 }, { "epoch": 0.5930162830388855, "grad_norm": 1.59505033493042, "learning_rate": 0.0001, "loss": 2.975290298461914, "step": 1607 }, { "epoch": 0.593385303750173, "grad_norm": 1.5120759010314941, "learning_rate": 0.0001, "loss": 3.177901268005371, "step": 1608 }, { "epoch": 0.5937543244614604, "grad_norm": 1.679570198059082, "learning_rate": 0.0001, "loss": 2.8872320652008057, "step": 1609 }, { "epoch": 0.5941233451727478, "grad_norm": 1.4617884159088135, "learning_rate": 0.0001, "loss": 3.1202046871185303, "step": 1610 }, { "epoch": 0.5944923658840352, "grad_norm": 1.72306489944458, "learning_rate": 0.0001, "loss": 2.711467981338501, "step": 1611 }, { "epoch": 0.5948613865953226, "grad_norm": 1.4061599969863892, "learning_rate": 0.0001, "loss": 2.987332820892334, "step": 1612 }, { "epoch": 0.59523040730661, "grad_norm": 1.4733103513717651, "learning_rate": 0.0001, "loss": 3.1502363681793213, "step": 1613 }, { "epoch": 0.5955994280178976, "grad_norm": 1.5665700435638428, "learning_rate": 0.0001, "loss": 3.1351773738861084, "step": 1614 }, { "epoch": 0.595968448729185, "grad_norm": 1.4482078552246094, "learning_rate": 0.0001, "loss": 3.1212244033813477, "step": 1615 }, { "epoch": 0.5963374694404724, "grad_norm": 1.7150236368179321, "learning_rate": 0.0001, "loss": 3.253479242324829, "step": 1616 }, { "epoch": 0.5967064901517598, "grad_norm": 1.4574846029281616, "learning_rate": 0.0001, "loss": 3.266017436981201, "step": 1617 }, { "epoch": 0.5970755108630472, "grad_norm": 1.3656078577041626, "learning_rate": 0.0001, "loss": 3.157707452774048, "step": 1618 }, { "epoch": 0.5974445315743346, "grad_norm": 1.5180892944335938, "learning_rate": 0.0001, "loss": 3.25515079498291, "step": 1619 }, { "epoch": 0.597813552285622, "grad_norm": 1.510731816291809, "learning_rate": 0.0001, "loss": 3.1406009197235107, "step": 1620 }, { "epoch": 0.5981825729969095, "grad_norm": 1.4035090208053589, "learning_rate": 0.0001, "loss": 3.189007043838501, "step": 1621 }, { "epoch": 0.5985515937081969, "grad_norm": 1.4495316743850708, "learning_rate": 0.0001, "loss": 2.9883956909179688, "step": 1622 }, { "epoch": 0.5989206144194843, "grad_norm": 1.3949147462844849, "learning_rate": 0.0001, "loss": 2.9849445819854736, "step": 1623 }, { "epoch": 0.5992896351307717, "grad_norm": 1.3221460580825806, "learning_rate": 0.0001, "loss": 2.8309829235076904, "step": 1624 }, { "epoch": 0.5996586558420591, "grad_norm": 1.4396876096725464, "learning_rate": 0.0001, "loss": 2.965526580810547, "step": 1625 }, { "epoch": 0.6000276765533465, "grad_norm": 1.2181516885757446, "learning_rate": 0.0001, "loss": 2.977940559387207, "step": 1626 }, { "epoch": 0.600396697264634, "grad_norm": 1.4671798944473267, "learning_rate": 0.0001, "loss": 3.126061201095581, "step": 1627 }, { "epoch": 0.6007657179759214, "grad_norm": 1.3950921297073364, "learning_rate": 0.0001, "loss": 3.0236101150512695, "step": 1628 }, { "epoch": 0.6011347386872088, "grad_norm": 1.3503427505493164, "learning_rate": 0.0001, "loss": 3.053161859512329, "step": 1629 }, { "epoch": 0.6015037593984962, "grad_norm": 1.5075093507766724, "learning_rate": 0.0001, "loss": 2.8057613372802734, "step": 1630 }, { "epoch": 0.6018727801097836, "grad_norm": 1.5590583086013794, "learning_rate": 0.0001, "loss": 3.131702423095703, "step": 1631 }, { "epoch": 0.602241800821071, "grad_norm": 1.441806435585022, "learning_rate": 0.0001, "loss": 2.813474178314209, "step": 1632 }, { "epoch": 0.6026108215323585, "grad_norm": 1.6584491729736328, "learning_rate": 0.0001, "loss": 2.9506592750549316, "step": 1633 }, { "epoch": 0.6029798422436459, "grad_norm": 2.532684326171875, "learning_rate": 0.0001, "loss": 3.2481188774108887, "step": 1634 }, { "epoch": 0.6033488629549334, "grad_norm": 1.484161138534546, "learning_rate": 0.0001, "loss": 2.924128532409668, "step": 1635 }, { "epoch": 0.6037178836662208, "grad_norm": 1.7719470262527466, "learning_rate": 0.0001, "loss": 3.166921377182007, "step": 1636 }, { "epoch": 0.6040869043775082, "grad_norm": 1.5779889822006226, "learning_rate": 0.0001, "loss": 2.8806204795837402, "step": 1637 }, { "epoch": 0.6044559250887956, "grad_norm": 1.4906623363494873, "learning_rate": 0.0001, "loss": 2.676790714263916, "step": 1638 }, { "epoch": 0.6048249458000831, "grad_norm": 1.5765575170516968, "learning_rate": 0.0001, "loss": 3.1346771717071533, "step": 1639 }, { "epoch": 0.6051939665113705, "grad_norm": 1.5246917009353638, "learning_rate": 0.0001, "loss": 3.1085193157196045, "step": 1640 }, { "epoch": 0.6055629872226579, "grad_norm": 1.2811768054962158, "learning_rate": 0.0001, "loss": 2.9110164642333984, "step": 1641 }, { "epoch": 0.6059320079339453, "grad_norm": 1.5545175075531006, "learning_rate": 0.0001, "loss": 3.0606868267059326, "step": 1642 }, { "epoch": 0.6063010286452327, "grad_norm": 1.6818307638168335, "learning_rate": 0.0001, "loss": 2.910564422607422, "step": 1643 }, { "epoch": 0.6066700493565201, "grad_norm": 1.724615216255188, "learning_rate": 0.0001, "loss": 3.127122402191162, "step": 1644 }, { "epoch": 0.6070390700678076, "grad_norm": 1.7975995540618896, "learning_rate": 0.0001, "loss": 2.991077423095703, "step": 1645 }, { "epoch": 0.607408090779095, "grad_norm": 1.7335525751113892, "learning_rate": 0.0001, "loss": 2.7936394214630127, "step": 1646 }, { "epoch": 0.6077771114903824, "grad_norm": 1.5975278615951538, "learning_rate": 0.0001, "loss": 3.1368608474731445, "step": 1647 }, { "epoch": 0.6081461322016698, "grad_norm": 1.5442079305648804, "learning_rate": 0.0001, "loss": 3.59978985786438, "step": 1648 }, { "epoch": 0.6085151529129572, "grad_norm": 1.6264346837997437, "learning_rate": 0.0001, "loss": 2.999553680419922, "step": 1649 }, { "epoch": 0.6088841736242446, "grad_norm": 1.5625030994415283, "learning_rate": 0.0001, "loss": 3.002931594848633, "step": 1650 }, { "epoch": 0.609253194335532, "grad_norm": 1.5753135681152344, "learning_rate": 0.0001, "loss": 3.299349784851074, "step": 1651 }, { "epoch": 0.6096222150468195, "grad_norm": 1.3783560991287231, "learning_rate": 0.0001, "loss": 3.0232598781585693, "step": 1652 }, { "epoch": 0.6099912357581069, "grad_norm": 1.5048720836639404, "learning_rate": 0.0001, "loss": 3.138497829437256, "step": 1653 }, { "epoch": 0.6103602564693943, "grad_norm": 1.2929656505584717, "learning_rate": 0.0001, "loss": 3.0127711296081543, "step": 1654 }, { "epoch": 0.6107292771806818, "grad_norm": 1.6666373014450073, "learning_rate": 0.0001, "loss": 3.2772326469421387, "step": 1655 }, { "epoch": 0.6110982978919692, "grad_norm": 1.7976620197296143, "learning_rate": 0.0001, "loss": 3.1206836700439453, "step": 1656 }, { "epoch": 0.6114673186032566, "grad_norm": 1.4898130893707275, "learning_rate": 0.0001, "loss": 3.3054213523864746, "step": 1657 }, { "epoch": 0.6118363393145441, "grad_norm": 1.4448245763778687, "learning_rate": 0.0001, "loss": 3.1637535095214844, "step": 1658 }, { "epoch": 0.6122053600258315, "grad_norm": 1.5893950462341309, "learning_rate": 0.0001, "loss": 3.0473036766052246, "step": 1659 }, { "epoch": 0.6125743807371189, "grad_norm": 1.4983068704605103, "learning_rate": 0.0001, "loss": 3.289008617401123, "step": 1660 }, { "epoch": 0.6129434014484063, "grad_norm": 1.328446865081787, "learning_rate": 0.0001, "loss": 3.0265350341796875, "step": 1661 }, { "epoch": 0.6133124221596937, "grad_norm": 1.5527483224868774, "learning_rate": 0.0001, "loss": 3.1307101249694824, "step": 1662 }, { "epoch": 0.6136814428709811, "grad_norm": 1.402368426322937, "learning_rate": 0.0001, "loss": 2.912205219268799, "step": 1663 }, { "epoch": 0.6140504635822686, "grad_norm": 1.56800377368927, "learning_rate": 0.0001, "loss": 3.0869646072387695, "step": 1664 }, { "epoch": 0.614419484293556, "grad_norm": 1.5935865640640259, "learning_rate": 0.0001, "loss": 3.3098647594451904, "step": 1665 }, { "epoch": 0.6147885050048434, "grad_norm": 1.4028699398040771, "learning_rate": 0.0001, "loss": 3.077375888824463, "step": 1666 }, { "epoch": 0.6151575257161308, "grad_norm": 1.551713466644287, "learning_rate": 0.0001, "loss": 2.9939842224121094, "step": 1667 }, { "epoch": 0.6155265464274182, "grad_norm": 1.5303837060928345, "learning_rate": 0.0001, "loss": 2.9591264724731445, "step": 1668 }, { "epoch": 0.6158955671387056, "grad_norm": 1.6299644708633423, "learning_rate": 0.0001, "loss": 2.6475961208343506, "step": 1669 }, { "epoch": 0.616264587849993, "grad_norm": 1.5066560506820679, "learning_rate": 0.0001, "loss": 3.0980443954467773, "step": 1670 }, { "epoch": 0.6166336085612805, "grad_norm": 1.3605924844741821, "learning_rate": 0.0001, "loss": 3.140065908432007, "step": 1671 }, { "epoch": 0.6170026292725679, "grad_norm": 2.658493757247925, "learning_rate": 0.0001, "loss": 3.1797752380371094, "step": 1672 }, { "epoch": 0.6173716499838553, "grad_norm": 1.305162787437439, "learning_rate": 0.0001, "loss": 3.21160888671875, "step": 1673 }, { "epoch": 0.6177406706951427, "grad_norm": 1.433411955833435, "learning_rate": 0.0001, "loss": 3.3091678619384766, "step": 1674 }, { "epoch": 0.6181096914064302, "grad_norm": 1.3497594594955444, "learning_rate": 0.0001, "loss": 2.7126662731170654, "step": 1675 }, { "epoch": 0.6184787121177177, "grad_norm": 1.6366459131240845, "learning_rate": 0.0001, "loss": 2.8928656578063965, "step": 1676 }, { "epoch": 0.6188477328290051, "grad_norm": 1.3585864305496216, "learning_rate": 0.0001, "loss": 3.161496162414551, "step": 1677 }, { "epoch": 0.6192167535402925, "grad_norm": 1.654638648033142, "learning_rate": 0.0001, "loss": 2.8951926231384277, "step": 1678 }, { "epoch": 0.6195857742515799, "grad_norm": 1.5824251174926758, "learning_rate": 0.0001, "loss": 2.993499994277954, "step": 1679 }, { "epoch": 0.6199547949628673, "grad_norm": 1.8946468830108643, "learning_rate": 0.0001, "loss": 3.1146035194396973, "step": 1680 }, { "epoch": 0.6203238156741547, "grad_norm": 1.5007972717285156, "learning_rate": 0.0001, "loss": 3.0264508724212646, "step": 1681 }, { "epoch": 0.6206928363854421, "grad_norm": 1.3127504587173462, "learning_rate": 0.0001, "loss": 2.6989588737487793, "step": 1682 }, { "epoch": 0.6210618570967296, "grad_norm": 1.5104249715805054, "learning_rate": 0.0001, "loss": 2.9619827270507812, "step": 1683 }, { "epoch": 0.621430877808017, "grad_norm": 1.5483195781707764, "learning_rate": 0.0001, "loss": 2.853079080581665, "step": 1684 }, { "epoch": 0.6217998985193044, "grad_norm": 1.54996657371521, "learning_rate": 0.0001, "loss": 3.076469898223877, "step": 1685 }, { "epoch": 0.6221689192305918, "grad_norm": 1.577514886856079, "learning_rate": 0.0001, "loss": 2.824328660964966, "step": 1686 }, { "epoch": 0.6225379399418792, "grad_norm": 1.4631121158599854, "learning_rate": 0.0001, "loss": 2.8961539268493652, "step": 1687 }, { "epoch": 0.6229069606531666, "grad_norm": 1.8148789405822754, "learning_rate": 0.0001, "loss": 3.038398504257202, "step": 1688 }, { "epoch": 0.6232759813644541, "grad_norm": 1.419563889503479, "learning_rate": 0.0001, "loss": 3.0199737548828125, "step": 1689 }, { "epoch": 0.6236450020757415, "grad_norm": 2.0867295265197754, "learning_rate": 0.0001, "loss": 2.9508156776428223, "step": 1690 }, { "epoch": 0.6240140227870289, "grad_norm": 1.5577143430709839, "learning_rate": 0.0001, "loss": 3.0952770709991455, "step": 1691 }, { "epoch": 0.6243830434983163, "grad_norm": 1.7039523124694824, "learning_rate": 0.0001, "loss": 3.300767421722412, "step": 1692 }, { "epoch": 0.6247520642096037, "grad_norm": 1.5497665405273438, "learning_rate": 0.0001, "loss": 2.8786702156066895, "step": 1693 }, { "epoch": 0.6251210849208911, "grad_norm": 1.6092883348464966, "learning_rate": 0.0001, "loss": 2.570465087890625, "step": 1694 }, { "epoch": 0.6254901056321787, "grad_norm": 1.2568544149398804, "learning_rate": 0.0001, "loss": 2.7884633541107178, "step": 1695 }, { "epoch": 0.6258591263434661, "grad_norm": 1.5724081993103027, "learning_rate": 0.0001, "loss": 3.0927348136901855, "step": 1696 }, { "epoch": 0.6262281470547535, "grad_norm": 1.7444639205932617, "learning_rate": 0.0001, "loss": 2.9007298946380615, "step": 1697 }, { "epoch": 0.6265971677660409, "grad_norm": 1.578318476676941, "learning_rate": 0.0001, "loss": 3.1113481521606445, "step": 1698 }, { "epoch": 0.6269661884773283, "grad_norm": 1.488328456878662, "learning_rate": 0.0001, "loss": 3.2207224369049072, "step": 1699 }, { "epoch": 0.6273352091886157, "grad_norm": 1.7932227849960327, "learning_rate": 0.0001, "loss": 3.1064136028289795, "step": 1700 }, { "epoch": 0.6277042298999032, "grad_norm": 1.7112846374511719, "learning_rate": 0.0001, "loss": 3.062856674194336, "step": 1701 }, { "epoch": 0.6280732506111906, "grad_norm": 1.4303313493728638, "learning_rate": 0.0001, "loss": 2.7863383293151855, "step": 1702 }, { "epoch": 0.628442271322478, "grad_norm": 1.6389356851577759, "learning_rate": 0.0001, "loss": 3.166963577270508, "step": 1703 }, { "epoch": 0.6288112920337654, "grad_norm": 1.5803802013397217, "learning_rate": 0.0001, "loss": 3.153716564178467, "step": 1704 }, { "epoch": 0.6291803127450528, "grad_norm": 1.7736324071884155, "learning_rate": 0.0001, "loss": 3.2840754985809326, "step": 1705 }, { "epoch": 0.6295493334563402, "grad_norm": 1.4402971267700195, "learning_rate": 0.0001, "loss": 3.23294734954834, "step": 1706 }, { "epoch": 0.6299183541676276, "grad_norm": 1.6728215217590332, "learning_rate": 0.0001, "loss": 3.1475470066070557, "step": 1707 }, { "epoch": 0.6302873748789151, "grad_norm": 1.790436029434204, "learning_rate": 0.0001, "loss": 2.8499362468719482, "step": 1708 }, { "epoch": 0.6306563955902025, "grad_norm": 1.4944933652877808, "learning_rate": 0.0001, "loss": 2.9134278297424316, "step": 1709 }, { "epoch": 0.6310254163014899, "grad_norm": 1.4435248374938965, "learning_rate": 0.0001, "loss": 2.9818530082702637, "step": 1710 }, { "epoch": 0.6313944370127773, "grad_norm": 1.5078636407852173, "learning_rate": 0.0001, "loss": 3.0898208618164062, "step": 1711 }, { "epoch": 0.6317634577240647, "grad_norm": 1.378479242324829, "learning_rate": 0.0001, "loss": 3.294893980026245, "step": 1712 }, { "epoch": 0.6321324784353521, "grad_norm": 1.5981260538101196, "learning_rate": 0.0001, "loss": 3.2986881732940674, "step": 1713 }, { "epoch": 0.6325014991466396, "grad_norm": 1.5564826726913452, "learning_rate": 0.0001, "loss": 3.2673473358154297, "step": 1714 }, { "epoch": 0.632870519857927, "grad_norm": 1.6040616035461426, "learning_rate": 0.0001, "loss": 3.2563705444335938, "step": 1715 }, { "epoch": 0.6332395405692145, "grad_norm": 1.3881851434707642, "learning_rate": 0.0001, "loss": 3.0239639282226562, "step": 1716 }, { "epoch": 0.6336085612805019, "grad_norm": 1.8124662637710571, "learning_rate": 0.0001, "loss": 2.956881523132324, "step": 1717 }, { "epoch": 0.6339775819917893, "grad_norm": 1.4811569452285767, "learning_rate": 0.0001, "loss": 3.152526378631592, "step": 1718 }, { "epoch": 0.6343466027030767, "grad_norm": 1.3669099807739258, "learning_rate": 0.0001, "loss": 3.225633382797241, "step": 1719 }, { "epoch": 0.6347156234143642, "grad_norm": 1.4169130325317383, "learning_rate": 0.0001, "loss": 2.7808620929718018, "step": 1720 }, { "epoch": 0.6350846441256516, "grad_norm": 1.6478363275527954, "learning_rate": 0.0001, "loss": 3.184664726257324, "step": 1721 }, { "epoch": 0.635453664836939, "grad_norm": 1.2910592555999756, "learning_rate": 0.0001, "loss": 3.2132749557495117, "step": 1722 }, { "epoch": 0.6358226855482264, "grad_norm": 1.4628801345825195, "learning_rate": 0.0001, "loss": 3.0728073120117188, "step": 1723 }, { "epoch": 0.6361917062595138, "grad_norm": 1.8667229413986206, "learning_rate": 0.0001, "loss": 2.954141139984131, "step": 1724 }, { "epoch": 0.6365607269708012, "grad_norm": 2.564122200012207, "learning_rate": 0.0001, "loss": 3.096381902694702, "step": 1725 }, { "epoch": 0.6369297476820887, "grad_norm": 1.765616774559021, "learning_rate": 0.0001, "loss": 3.1382851600646973, "step": 1726 }, { "epoch": 0.6372987683933761, "grad_norm": 1.2878708839416504, "learning_rate": 0.0001, "loss": 3.0469586849212646, "step": 1727 }, { "epoch": 0.6376677891046635, "grad_norm": 1.4632006883621216, "learning_rate": 0.0001, "loss": 3.0743842124938965, "step": 1728 }, { "epoch": 0.6380368098159509, "grad_norm": 1.4671846628189087, "learning_rate": 0.0001, "loss": 2.769979476928711, "step": 1729 }, { "epoch": 0.6384058305272383, "grad_norm": 1.4776455163955688, "learning_rate": 0.0001, "loss": 2.88378643989563, "step": 1730 }, { "epoch": 0.6387748512385257, "grad_norm": 1.4656074047088623, "learning_rate": 0.0001, "loss": 2.7346935272216797, "step": 1731 }, { "epoch": 0.6391438719498131, "grad_norm": 1.5464832782745361, "learning_rate": 0.0001, "loss": 3.2409212589263916, "step": 1732 }, { "epoch": 0.6395128926611006, "grad_norm": 2.7446348667144775, "learning_rate": 0.0001, "loss": 2.8477261066436768, "step": 1733 }, { "epoch": 0.639881913372388, "grad_norm": 1.7659188508987427, "learning_rate": 0.0001, "loss": 2.8387794494628906, "step": 1734 }, { "epoch": 0.6402509340836754, "grad_norm": 1.5841516256332397, "learning_rate": 0.0001, "loss": 2.917804718017578, "step": 1735 }, { "epoch": 0.6406199547949629, "grad_norm": 1.4154927730560303, "learning_rate": 0.0001, "loss": 3.160935401916504, "step": 1736 }, { "epoch": 0.6409889755062503, "grad_norm": 1.7006475925445557, "learning_rate": 0.0001, "loss": 3.291802167892456, "step": 1737 }, { "epoch": 0.6413579962175378, "grad_norm": 1.3947173357009888, "learning_rate": 0.0001, "loss": 2.796884536743164, "step": 1738 }, { "epoch": 0.6417270169288252, "grad_norm": 1.3439325094223022, "learning_rate": 0.0001, "loss": 2.884575605392456, "step": 1739 }, { "epoch": 0.6420960376401126, "grad_norm": 1.6570194959640503, "learning_rate": 0.0001, "loss": 3.2111990451812744, "step": 1740 }, { "epoch": 0.6424650583514, "grad_norm": 1.582147240638733, "learning_rate": 0.0001, "loss": 2.898564577102661, "step": 1741 }, { "epoch": 0.6428340790626874, "grad_norm": 1.413657307624817, "learning_rate": 0.0001, "loss": 3.106001377105713, "step": 1742 }, { "epoch": 0.6432030997739748, "grad_norm": 1.5542272329330444, "learning_rate": 0.0001, "loss": 3.310490608215332, "step": 1743 }, { "epoch": 0.6435721204852622, "grad_norm": 1.5526739358901978, "learning_rate": 0.0001, "loss": 2.9970850944519043, "step": 1744 }, { "epoch": 0.6439411411965497, "grad_norm": 1.4884364604949951, "learning_rate": 0.0001, "loss": 2.9700937271118164, "step": 1745 }, { "epoch": 0.6443101619078371, "grad_norm": 1.3288850784301758, "learning_rate": 0.0001, "loss": 2.9514870643615723, "step": 1746 }, { "epoch": 0.6446791826191245, "grad_norm": 1.937334656715393, "learning_rate": 0.0001, "loss": 3.1158037185668945, "step": 1747 }, { "epoch": 0.6450482033304119, "grad_norm": 1.3582133054733276, "learning_rate": 0.0001, "loss": 2.8284592628479004, "step": 1748 }, { "epoch": 0.6454172240416993, "grad_norm": 1.419512152671814, "learning_rate": 0.0001, "loss": 2.8240017890930176, "step": 1749 }, { "epoch": 0.6457862447529867, "grad_norm": 1.7152806520462036, "learning_rate": 0.0001, "loss": 3.222334861755371, "step": 1750 }, { "epoch": 0.6461552654642742, "grad_norm": 1.4931284189224243, "learning_rate": 0.0001, "loss": 2.7273707389831543, "step": 1751 }, { "epoch": 0.6465242861755616, "grad_norm": 1.7655099630355835, "learning_rate": 0.0001, "loss": 2.9631106853485107, "step": 1752 }, { "epoch": 0.646893306886849, "grad_norm": 1.3547983169555664, "learning_rate": 0.0001, "loss": 2.9832592010498047, "step": 1753 }, { "epoch": 0.6472623275981364, "grad_norm": 2.1741528511047363, "learning_rate": 0.0001, "loss": 3.3106908798217773, "step": 1754 }, { "epoch": 0.6476313483094238, "grad_norm": 1.6267430782318115, "learning_rate": 0.0001, "loss": 3.373060464859009, "step": 1755 }, { "epoch": 0.6480003690207113, "grad_norm": 1.4111424684524536, "learning_rate": 0.0001, "loss": 2.8699936866760254, "step": 1756 }, { "epoch": 0.6483693897319988, "grad_norm": 1.6091253757476807, "learning_rate": 0.0001, "loss": 3.0611021518707275, "step": 1757 }, { "epoch": 0.6487384104432862, "grad_norm": 1.4287705421447754, "learning_rate": 0.0001, "loss": 3.1343512535095215, "step": 1758 }, { "epoch": 0.6491074311545736, "grad_norm": 1.9212841987609863, "learning_rate": 0.0001, "loss": 3.3366332054138184, "step": 1759 }, { "epoch": 0.649476451865861, "grad_norm": 1.9431073665618896, "learning_rate": 0.0001, "loss": 3.342465877532959, "step": 1760 }, { "epoch": 0.6498454725771484, "grad_norm": 1.4766361713409424, "learning_rate": 0.0001, "loss": 2.8368186950683594, "step": 1761 }, { "epoch": 0.6502144932884358, "grad_norm": 1.4235342741012573, "learning_rate": 0.0001, "loss": 3.209545135498047, "step": 1762 }, { "epoch": 0.6505835139997233, "grad_norm": 1.345210075378418, "learning_rate": 0.0001, "loss": 2.9051876068115234, "step": 1763 }, { "epoch": 0.6509525347110107, "grad_norm": 1.565621018409729, "learning_rate": 0.0001, "loss": 2.688246250152588, "step": 1764 }, { "epoch": 0.6513215554222981, "grad_norm": 1.5794657468795776, "learning_rate": 0.0001, "loss": 3.0677590370178223, "step": 1765 }, { "epoch": 0.6516905761335855, "grad_norm": 1.4312400817871094, "learning_rate": 0.0001, "loss": 2.8350164890289307, "step": 1766 }, { "epoch": 0.6520595968448729, "grad_norm": 1.4874008893966675, "learning_rate": 0.0001, "loss": 3.09765625, "step": 1767 }, { "epoch": 0.6524286175561603, "grad_norm": 1.5105725526809692, "learning_rate": 0.0001, "loss": 3.0668158531188965, "step": 1768 }, { "epoch": 0.6527976382674477, "grad_norm": 1.5277299880981445, "learning_rate": 0.0001, "loss": 2.898284673690796, "step": 1769 }, { "epoch": 0.6531666589787352, "grad_norm": 1.5099315643310547, "learning_rate": 0.0001, "loss": 3.0987911224365234, "step": 1770 }, { "epoch": 0.6535356796900226, "grad_norm": 1.4647843837738037, "learning_rate": 0.0001, "loss": 2.8843438625335693, "step": 1771 }, { "epoch": 0.65390470040131, "grad_norm": 1.4667749404907227, "learning_rate": 0.0001, "loss": 3.021373987197876, "step": 1772 }, { "epoch": 0.6542737211125974, "grad_norm": 1.37954843044281, "learning_rate": 0.0001, "loss": 3.0905230045318604, "step": 1773 }, { "epoch": 0.6546427418238848, "grad_norm": 1.6110352277755737, "learning_rate": 0.0001, "loss": 3.0917410850524902, "step": 1774 }, { "epoch": 0.6550117625351722, "grad_norm": 1.8625768423080444, "learning_rate": 0.0001, "loss": 2.915768623352051, "step": 1775 }, { "epoch": 0.6553807832464597, "grad_norm": 1.428928017616272, "learning_rate": 0.0001, "loss": 3.0600011348724365, "step": 1776 }, { "epoch": 0.6557498039577472, "grad_norm": 1.5121434926986694, "learning_rate": 0.0001, "loss": 2.9294376373291016, "step": 1777 }, { "epoch": 0.6561188246690346, "grad_norm": 1.46657395362854, "learning_rate": 0.0001, "loss": 2.7204599380493164, "step": 1778 }, { "epoch": 0.656487845380322, "grad_norm": 1.5031368732452393, "learning_rate": 0.0001, "loss": 2.7291100025177, "step": 1779 }, { "epoch": 0.6568568660916094, "grad_norm": 1.644879698753357, "learning_rate": 0.0001, "loss": 2.9993906021118164, "step": 1780 }, { "epoch": 0.6572258868028968, "grad_norm": 1.4957876205444336, "learning_rate": 0.0001, "loss": 3.0316343307495117, "step": 1781 }, { "epoch": 0.6575949075141843, "grad_norm": 1.3566728830337524, "learning_rate": 0.0001, "loss": 3.002734661102295, "step": 1782 }, { "epoch": 0.6579639282254717, "grad_norm": 1.2837598323822021, "learning_rate": 0.0001, "loss": 2.868499755859375, "step": 1783 }, { "epoch": 0.6583329489367591, "grad_norm": 1.3079769611358643, "learning_rate": 0.0001, "loss": 2.844012498855591, "step": 1784 }, { "epoch": 0.6587019696480465, "grad_norm": 1.5618178844451904, "learning_rate": 0.0001, "loss": 3.1040420532226562, "step": 1785 }, { "epoch": 0.6590709903593339, "grad_norm": 1.4766494035720825, "learning_rate": 0.0001, "loss": 2.9681851863861084, "step": 1786 }, { "epoch": 0.6594400110706213, "grad_norm": 1.5607540607452393, "learning_rate": 0.0001, "loss": 2.9938366413116455, "step": 1787 }, { "epoch": 0.6598090317819088, "grad_norm": 1.4804152250289917, "learning_rate": 0.0001, "loss": 2.7320704460144043, "step": 1788 }, { "epoch": 0.6601780524931962, "grad_norm": 1.5870630741119385, "learning_rate": 0.0001, "loss": 2.966982364654541, "step": 1789 }, { "epoch": 0.6605470732044836, "grad_norm": 3.4533731937408447, "learning_rate": 0.0001, "loss": 2.769727945327759, "step": 1790 }, { "epoch": 0.660916093915771, "grad_norm": 1.4948451519012451, "learning_rate": 0.0001, "loss": 3.295403003692627, "step": 1791 }, { "epoch": 0.6612851146270584, "grad_norm": 1.7767696380615234, "learning_rate": 0.0001, "loss": 3.1707565784454346, "step": 1792 }, { "epoch": 0.6616541353383458, "grad_norm": 1.899431586265564, "learning_rate": 0.0001, "loss": 2.736776351928711, "step": 1793 }, { "epoch": 0.6620231560496332, "grad_norm": 1.6180992126464844, "learning_rate": 0.0001, "loss": 3.4115140438079834, "step": 1794 }, { "epoch": 0.6623921767609207, "grad_norm": 1.6390759944915771, "learning_rate": 0.0001, "loss": 3.025819778442383, "step": 1795 }, { "epoch": 0.6627611974722081, "grad_norm": 1.3596994876861572, "learning_rate": 0.0001, "loss": 3.1112544536590576, "step": 1796 }, { "epoch": 0.6631302181834956, "grad_norm": 1.3268007040023804, "learning_rate": 0.0001, "loss": 2.923668146133423, "step": 1797 }, { "epoch": 0.663499238894783, "grad_norm": 1.4096367359161377, "learning_rate": 0.0001, "loss": 3.249220848083496, "step": 1798 }, { "epoch": 0.6638682596060704, "grad_norm": 1.5752679109573364, "learning_rate": 0.0001, "loss": 3.067765712738037, "step": 1799 }, { "epoch": 0.6642372803173578, "grad_norm": 1.4958841800689697, "learning_rate": 0.0001, "loss": 3.088843822479248, "step": 1800 }, { "epoch": 0.6646063010286453, "grad_norm": 1.3285409212112427, "learning_rate": 0.0001, "loss": 2.7293970584869385, "step": 1801 }, { "epoch": 0.6649753217399327, "grad_norm": 1.2899397611618042, "learning_rate": 0.0001, "loss": 2.649564027786255, "step": 1802 }, { "epoch": 0.6653443424512201, "grad_norm": 1.544846534729004, "learning_rate": 0.0001, "loss": 2.744021415710449, "step": 1803 }, { "epoch": 0.6657133631625075, "grad_norm": 1.423902988433838, "learning_rate": 0.0001, "loss": 2.7857630252838135, "step": 1804 }, { "epoch": 0.6660823838737949, "grad_norm": 1.3778213262557983, "learning_rate": 0.0001, "loss": 2.8091185092926025, "step": 1805 }, { "epoch": 0.6664514045850823, "grad_norm": 1.905591607093811, "learning_rate": 0.0001, "loss": 3.120008945465088, "step": 1806 }, { "epoch": 0.6668204252963698, "grad_norm": 1.7237114906311035, "learning_rate": 0.0001, "loss": 3.3375370502471924, "step": 1807 }, { "epoch": 0.6671894460076572, "grad_norm": 1.9337058067321777, "learning_rate": 0.0001, "loss": 2.9047162532806396, "step": 1808 }, { "epoch": 0.6675584667189446, "grad_norm": 1.570589542388916, "learning_rate": 0.0001, "loss": 2.9551525115966797, "step": 1809 }, { "epoch": 0.667927487430232, "grad_norm": 1.6417497396469116, "learning_rate": 0.0001, "loss": 3.305920362472534, "step": 1810 }, { "epoch": 0.6682965081415194, "grad_norm": 1.4718272686004639, "learning_rate": 0.0001, "loss": 2.9420154094696045, "step": 1811 }, { "epoch": 0.6686655288528068, "grad_norm": 1.3973904848098755, "learning_rate": 0.0001, "loss": 2.9395124912261963, "step": 1812 }, { "epoch": 0.6690345495640942, "grad_norm": 1.5156550407409668, "learning_rate": 0.0001, "loss": 3.0749330520629883, "step": 1813 }, { "epoch": 0.6694035702753817, "grad_norm": 1.4216914176940918, "learning_rate": 0.0001, "loss": 2.8323912620544434, "step": 1814 }, { "epoch": 0.6697725909866691, "grad_norm": 1.2921677827835083, "learning_rate": 0.0001, "loss": 2.778679370880127, "step": 1815 }, { "epoch": 0.6701416116979565, "grad_norm": 1.6747710704803467, "learning_rate": 0.0001, "loss": 3.270148992538452, "step": 1816 }, { "epoch": 0.670510632409244, "grad_norm": 1.3823071718215942, "learning_rate": 0.0001, "loss": 2.7984721660614014, "step": 1817 }, { "epoch": 0.6708796531205314, "grad_norm": 1.5142064094543457, "learning_rate": 0.0001, "loss": 2.7632479667663574, "step": 1818 }, { "epoch": 0.6712486738318189, "grad_norm": 1.4969267845153809, "learning_rate": 0.0001, "loss": 2.7298598289489746, "step": 1819 }, { "epoch": 0.6716176945431063, "grad_norm": 1.319321870803833, "learning_rate": 0.0001, "loss": 2.979416847229004, "step": 1820 }, { "epoch": 0.6719867152543937, "grad_norm": 1.7111543416976929, "learning_rate": 0.0001, "loss": 3.2181267738342285, "step": 1821 }, { "epoch": 0.6723557359656811, "grad_norm": 1.5696139335632324, "learning_rate": 0.0001, "loss": 3.252894401550293, "step": 1822 }, { "epoch": 0.6727247566769685, "grad_norm": 1.4902926683425903, "learning_rate": 0.0001, "loss": 3.4297800064086914, "step": 1823 }, { "epoch": 0.6730937773882559, "grad_norm": 1.4322476387023926, "learning_rate": 0.0001, "loss": 3.060048818588257, "step": 1824 }, { "epoch": 0.6734627980995433, "grad_norm": 1.9995813369750977, "learning_rate": 0.0001, "loss": 3.169847249984741, "step": 1825 }, { "epoch": 0.6738318188108308, "grad_norm": 1.564002513885498, "learning_rate": 0.0001, "loss": 2.913902759552002, "step": 1826 }, { "epoch": 0.6742008395221182, "grad_norm": 1.3918930292129517, "learning_rate": 0.0001, "loss": 2.819301128387451, "step": 1827 }, { "epoch": 0.6745698602334056, "grad_norm": 1.303747296333313, "learning_rate": 0.0001, "loss": 3.0108795166015625, "step": 1828 }, { "epoch": 0.674938880944693, "grad_norm": 1.416218638420105, "learning_rate": 0.0001, "loss": 2.9957869052886963, "step": 1829 }, { "epoch": 0.6753079016559804, "grad_norm": 1.5372474193572998, "learning_rate": 0.0001, "loss": 2.951519012451172, "step": 1830 }, { "epoch": 0.6756769223672678, "grad_norm": 1.55687415599823, "learning_rate": 0.0001, "loss": 3.3054733276367188, "step": 1831 }, { "epoch": 0.6760459430785553, "grad_norm": 1.3472673892974854, "learning_rate": 0.0001, "loss": 2.8613457679748535, "step": 1832 }, { "epoch": 0.6764149637898427, "grad_norm": 1.8348579406738281, "learning_rate": 0.0001, "loss": 2.94702410697937, "step": 1833 }, { "epoch": 0.6767839845011301, "grad_norm": 1.5582042932510376, "learning_rate": 0.0001, "loss": 3.1498701572418213, "step": 1834 }, { "epoch": 0.6771530052124175, "grad_norm": 1.2993719577789307, "learning_rate": 0.0001, "loss": 2.7699830532073975, "step": 1835 }, { "epoch": 0.6775220259237049, "grad_norm": 1.6542822122573853, "learning_rate": 0.0001, "loss": 3.0390725135803223, "step": 1836 }, { "epoch": 0.6778910466349923, "grad_norm": 1.3510726690292358, "learning_rate": 0.0001, "loss": 2.9722113609313965, "step": 1837 }, { "epoch": 0.6782600673462799, "grad_norm": 1.6047877073287964, "learning_rate": 0.0001, "loss": 3.0230305194854736, "step": 1838 }, { "epoch": 0.6786290880575673, "grad_norm": 1.6720974445343018, "learning_rate": 0.0001, "loss": 3.082628011703491, "step": 1839 }, { "epoch": 0.6789981087688547, "grad_norm": 1.4950906038284302, "learning_rate": 0.0001, "loss": 2.8173470497131348, "step": 1840 }, { "epoch": 0.6793671294801421, "grad_norm": 1.4866859912872314, "learning_rate": 0.0001, "loss": 3.35282826423645, "step": 1841 }, { "epoch": 0.6797361501914295, "grad_norm": 1.4296315908432007, "learning_rate": 0.0001, "loss": 2.9932985305786133, "step": 1842 }, { "epoch": 0.6801051709027169, "grad_norm": 1.50027596950531, "learning_rate": 0.0001, "loss": 2.8272571563720703, "step": 1843 }, { "epoch": 0.6804741916140044, "grad_norm": 1.6475785970687866, "learning_rate": 0.0001, "loss": 2.878445625305176, "step": 1844 }, { "epoch": 0.6808432123252918, "grad_norm": 1.2895855903625488, "learning_rate": 0.0001, "loss": 2.878225326538086, "step": 1845 }, { "epoch": 0.6812122330365792, "grad_norm": 1.3523904085159302, "learning_rate": 0.0001, "loss": 2.7548604011535645, "step": 1846 }, { "epoch": 0.6815812537478666, "grad_norm": 1.324211597442627, "learning_rate": 0.0001, "loss": 3.4955055713653564, "step": 1847 }, { "epoch": 0.681950274459154, "grad_norm": 1.528855562210083, "learning_rate": 0.0001, "loss": 2.970367908477783, "step": 1848 }, { "epoch": 0.6823192951704414, "grad_norm": 1.2832826375961304, "learning_rate": 0.0001, "loss": 2.6480512619018555, "step": 1849 }, { "epoch": 0.6826883158817288, "grad_norm": 1.600754976272583, "learning_rate": 0.0001, "loss": 2.7719860076904297, "step": 1850 }, { "epoch": 0.6830573365930163, "grad_norm": 1.3437533378601074, "learning_rate": 0.0001, "loss": 2.7696330547332764, "step": 1851 }, { "epoch": 0.6834263573043037, "grad_norm": 1.9593110084533691, "learning_rate": 0.0001, "loss": 3.0834431648254395, "step": 1852 }, { "epoch": 0.6837953780155911, "grad_norm": 1.5265679359436035, "learning_rate": 0.0001, "loss": 3.0833916664123535, "step": 1853 }, { "epoch": 0.6841643987268785, "grad_norm": 1.3539191484451294, "learning_rate": 0.0001, "loss": 2.8511109352111816, "step": 1854 }, { "epoch": 0.6845334194381659, "grad_norm": 1.2647582292556763, "learning_rate": 0.0001, "loss": 2.616400718688965, "step": 1855 }, { "epoch": 0.6849024401494533, "grad_norm": 1.359884262084961, "learning_rate": 0.0001, "loss": 2.8696694374084473, "step": 1856 }, { "epoch": 0.6852714608607408, "grad_norm": 1.615756630897522, "learning_rate": 0.0001, "loss": 2.9391608238220215, "step": 1857 }, { "epoch": 0.6856404815720283, "grad_norm": 1.4345788955688477, "learning_rate": 0.0001, "loss": 2.836974620819092, "step": 1858 }, { "epoch": 0.6860095022833157, "grad_norm": 1.4609493017196655, "learning_rate": 0.0001, "loss": 2.8592610359191895, "step": 1859 }, { "epoch": 0.6863785229946031, "grad_norm": 1.5064507722854614, "learning_rate": 0.0001, "loss": 3.0935704708099365, "step": 1860 }, { "epoch": 0.6867475437058905, "grad_norm": 1.2052319049835205, "learning_rate": 0.0001, "loss": 3.089925527572632, "step": 1861 }, { "epoch": 0.6871165644171779, "grad_norm": 1.4756081104278564, "learning_rate": 0.0001, "loss": 3.0707755088806152, "step": 1862 }, { "epoch": 0.6874855851284654, "grad_norm": 1.7576805353164673, "learning_rate": 0.0001, "loss": 2.8919661045074463, "step": 1863 }, { "epoch": 0.6878546058397528, "grad_norm": 2.2750089168548584, "learning_rate": 0.0001, "loss": 3.5638415813446045, "step": 1864 }, { "epoch": 0.6882236265510402, "grad_norm": 1.5081532001495361, "learning_rate": 0.0001, "loss": 2.7697250843048096, "step": 1865 }, { "epoch": 0.6885926472623276, "grad_norm": 1.324602723121643, "learning_rate": 0.0001, "loss": 3.077737808227539, "step": 1866 }, { "epoch": 0.688961667973615, "grad_norm": 1.7070655822753906, "learning_rate": 0.0001, "loss": 3.1621508598327637, "step": 1867 }, { "epoch": 0.6893306886849024, "grad_norm": 1.575618028640747, "learning_rate": 0.0001, "loss": 3.030515432357788, "step": 1868 }, { "epoch": 0.6896997093961899, "grad_norm": 1.24668550491333, "learning_rate": 0.0001, "loss": 2.677110195159912, "step": 1869 }, { "epoch": 0.6900687301074773, "grad_norm": 1.8279112577438354, "learning_rate": 0.0001, "loss": 2.918236255645752, "step": 1870 }, { "epoch": 0.6904377508187647, "grad_norm": 1.4794249534606934, "learning_rate": 0.0001, "loss": 3.092900037765503, "step": 1871 }, { "epoch": 0.6908067715300521, "grad_norm": 1.3672035932540894, "learning_rate": 0.0001, "loss": 3.224522590637207, "step": 1872 }, { "epoch": 0.6911757922413395, "grad_norm": 1.5928970575332642, "learning_rate": 0.0001, "loss": 3.1361751556396484, "step": 1873 }, { "epoch": 0.6915448129526269, "grad_norm": 1.6979690790176392, "learning_rate": 0.0001, "loss": 3.0488436222076416, "step": 1874 }, { "epoch": 0.6919138336639143, "grad_norm": 1.4633123874664307, "learning_rate": 0.0001, "loss": 3.1957483291625977, "step": 1875 }, { "epoch": 0.6922828543752018, "grad_norm": 1.536675214767456, "learning_rate": 0.0001, "loss": 3.0182981491088867, "step": 1876 }, { "epoch": 0.6926518750864892, "grad_norm": 1.6602140665054321, "learning_rate": 0.0001, "loss": 2.6960439682006836, "step": 1877 }, { "epoch": 0.6930208957977767, "grad_norm": 1.3703001737594604, "learning_rate": 0.0001, "loss": 3.015852928161621, "step": 1878 }, { "epoch": 0.6933899165090641, "grad_norm": 1.5959553718566895, "learning_rate": 0.0001, "loss": 3.180753231048584, "step": 1879 }, { "epoch": 0.6937589372203515, "grad_norm": 1.8855257034301758, "learning_rate": 0.0001, "loss": 2.9833192825317383, "step": 1880 }, { "epoch": 0.694127957931639, "grad_norm": 1.820612907409668, "learning_rate": 0.0001, "loss": 2.8149728775024414, "step": 1881 }, { "epoch": 0.6944969786429264, "grad_norm": 1.669589877128601, "learning_rate": 0.0001, "loss": 2.7853381633758545, "step": 1882 }, { "epoch": 0.6948659993542138, "grad_norm": 1.3921109437942505, "learning_rate": 0.0001, "loss": 2.9256534576416016, "step": 1883 }, { "epoch": 0.6952350200655012, "grad_norm": 1.373618721961975, "learning_rate": 0.0001, "loss": 2.686201572418213, "step": 1884 }, { "epoch": 0.6956040407767886, "grad_norm": 1.6283828020095825, "learning_rate": 0.0001, "loss": 2.891437530517578, "step": 1885 }, { "epoch": 0.695973061488076, "grad_norm": 1.5905407667160034, "learning_rate": 0.0001, "loss": 3.1153993606567383, "step": 1886 }, { "epoch": 0.6963420821993634, "grad_norm": 1.8783527612686157, "learning_rate": 0.0001, "loss": 2.97926664352417, "step": 1887 }, { "epoch": 0.6967111029106509, "grad_norm": 1.3532277345657349, "learning_rate": 0.0001, "loss": 2.6803927421569824, "step": 1888 }, { "epoch": 0.6970801236219383, "grad_norm": 1.5074236392974854, "learning_rate": 0.0001, "loss": 3.142784833908081, "step": 1889 }, { "epoch": 0.6974491443332257, "grad_norm": 1.893810749053955, "learning_rate": 0.0001, "loss": 2.6706981658935547, "step": 1890 }, { "epoch": 0.6978181650445131, "grad_norm": 1.4426836967468262, "learning_rate": 0.0001, "loss": 3.0196287631988525, "step": 1891 }, { "epoch": 0.6981871857558005, "grad_norm": 1.3750643730163574, "learning_rate": 0.0001, "loss": 2.9210805892944336, "step": 1892 }, { "epoch": 0.6985562064670879, "grad_norm": 1.8446600437164307, "learning_rate": 0.0001, "loss": 3.0644965171813965, "step": 1893 }, { "epoch": 0.6989252271783754, "grad_norm": 1.2527445554733276, "learning_rate": 0.0001, "loss": 3.040070056915283, "step": 1894 }, { "epoch": 0.6992942478896628, "grad_norm": 1.5171281099319458, "learning_rate": 0.0001, "loss": 2.7106070518493652, "step": 1895 }, { "epoch": 0.6996632686009502, "grad_norm": 1.9494513273239136, "learning_rate": 0.0001, "loss": 2.832226276397705, "step": 1896 }, { "epoch": 0.7000322893122376, "grad_norm": 1.60041081905365, "learning_rate": 0.0001, "loss": 3.051914930343628, "step": 1897 }, { "epoch": 0.7004013100235251, "grad_norm": 1.4158354997634888, "learning_rate": 0.0001, "loss": 2.8772706985473633, "step": 1898 }, { "epoch": 0.7007703307348125, "grad_norm": 1.3500134944915771, "learning_rate": 0.0001, "loss": 3.1860299110412598, "step": 1899 }, { "epoch": 0.7011393514461, "grad_norm": 1.5659834146499634, "learning_rate": 0.0001, "loss": 3.222621440887451, "step": 1900 }, { "epoch": 0.7015083721573874, "grad_norm": 1.7545522451400757, "learning_rate": 0.0001, "loss": 2.9130895137786865, "step": 1901 }, { "epoch": 0.7018773928686748, "grad_norm": 1.5776033401489258, "learning_rate": 0.0001, "loss": 2.6682968139648438, "step": 1902 }, { "epoch": 0.7022464135799622, "grad_norm": 1.5629349946975708, "learning_rate": 0.0001, "loss": 2.8258938789367676, "step": 1903 }, { "epoch": 0.7026154342912496, "grad_norm": 1.9601589441299438, "learning_rate": 0.0001, "loss": 3.34940767288208, "step": 1904 }, { "epoch": 0.702984455002537, "grad_norm": 1.2208720445632935, "learning_rate": 0.0001, "loss": 2.8936820030212402, "step": 1905 }, { "epoch": 0.7033534757138244, "grad_norm": 1.3387062549591064, "learning_rate": 0.0001, "loss": 2.675377607345581, "step": 1906 }, { "epoch": 0.7037224964251119, "grad_norm": 1.2717069387435913, "learning_rate": 0.0001, "loss": 2.920372247695923, "step": 1907 }, { "epoch": 0.7040915171363993, "grad_norm": 1.5965442657470703, "learning_rate": 0.0001, "loss": 2.995965003967285, "step": 1908 }, { "epoch": 0.7044605378476867, "grad_norm": 1.6870509386062622, "learning_rate": 0.0001, "loss": 2.7902636528015137, "step": 1909 }, { "epoch": 0.7048295585589741, "grad_norm": 1.4802327156066895, "learning_rate": 0.0001, "loss": 3.0240533351898193, "step": 1910 }, { "epoch": 0.7051985792702615, "grad_norm": 2.381451368331909, "learning_rate": 0.0001, "loss": 2.9454152584075928, "step": 1911 }, { "epoch": 0.7055675999815489, "grad_norm": 1.4864143133163452, "learning_rate": 0.0001, "loss": 2.8675129413604736, "step": 1912 }, { "epoch": 0.7059366206928364, "grad_norm": 1.3196614980697632, "learning_rate": 0.0001, "loss": 2.89752197265625, "step": 1913 }, { "epoch": 0.7063056414041238, "grad_norm": 1.751633644104004, "learning_rate": 0.0001, "loss": 2.87994647026062, "step": 1914 }, { "epoch": 0.7066746621154112, "grad_norm": 1.6252284049987793, "learning_rate": 0.0001, "loss": 3.0864391326904297, "step": 1915 }, { "epoch": 0.7070436828266986, "grad_norm": 2.2658016681671143, "learning_rate": 0.0001, "loss": 3.103297472000122, "step": 1916 }, { "epoch": 0.707412703537986, "grad_norm": 1.3630235195159912, "learning_rate": 0.0001, "loss": 2.8233394622802734, "step": 1917 }, { "epoch": 0.7077817242492734, "grad_norm": 1.6664007902145386, "learning_rate": 0.0001, "loss": 3.0205798149108887, "step": 1918 }, { "epoch": 0.708150744960561, "grad_norm": 1.5218522548675537, "learning_rate": 0.0001, "loss": 3.172025203704834, "step": 1919 }, { "epoch": 0.7085197656718484, "grad_norm": 1.5720617771148682, "learning_rate": 0.0001, "loss": 2.6702070236206055, "step": 1920 }, { "epoch": 0.7088887863831358, "grad_norm": 1.3254010677337646, "learning_rate": 0.0001, "loss": 2.821645736694336, "step": 1921 }, { "epoch": 0.7092578070944232, "grad_norm": 1.5990270376205444, "learning_rate": 0.0001, "loss": 3.2246832847595215, "step": 1922 }, { "epoch": 0.7096268278057106, "grad_norm": 1.5148303508758545, "learning_rate": 0.0001, "loss": 2.89823579788208, "step": 1923 }, { "epoch": 0.709995848516998, "grad_norm": 1.5021545886993408, "learning_rate": 0.0001, "loss": 3.2726035118103027, "step": 1924 }, { "epoch": 0.7103648692282855, "grad_norm": 1.4552438259124756, "learning_rate": 0.0001, "loss": 2.926816940307617, "step": 1925 }, { "epoch": 0.7107338899395729, "grad_norm": 1.5406982898712158, "learning_rate": 0.0001, "loss": 3.070967674255371, "step": 1926 }, { "epoch": 0.7111029106508603, "grad_norm": 1.4796096086502075, "learning_rate": 0.0001, "loss": 2.9255361557006836, "step": 1927 }, { "epoch": 0.7114719313621477, "grad_norm": 1.1283552646636963, "learning_rate": 0.0001, "loss": 2.9943838119506836, "step": 1928 }, { "epoch": 0.7118409520734351, "grad_norm": 1.5459461212158203, "learning_rate": 0.0001, "loss": 2.85526442527771, "step": 1929 }, { "epoch": 0.7122099727847225, "grad_norm": 1.5871220827102661, "learning_rate": 0.0001, "loss": 2.9002575874328613, "step": 1930 }, { "epoch": 0.71257899349601, "grad_norm": 1.3315916061401367, "learning_rate": 0.0001, "loss": 2.855234146118164, "step": 1931 }, { "epoch": 0.7129480142072974, "grad_norm": 1.361181616783142, "learning_rate": 0.0001, "loss": 3.015943765640259, "step": 1932 }, { "epoch": 0.7133170349185848, "grad_norm": 1.4395321607589722, "learning_rate": 0.0001, "loss": 2.9297475814819336, "step": 1933 }, { "epoch": 0.7136860556298722, "grad_norm": 1.298350214958191, "learning_rate": 0.0001, "loss": 3.027562141418457, "step": 1934 }, { "epoch": 0.7140550763411596, "grad_norm": 1.6706504821777344, "learning_rate": 0.0001, "loss": 2.9711947441101074, "step": 1935 }, { "epoch": 0.714424097052447, "grad_norm": 1.601120948791504, "learning_rate": 0.0001, "loss": 2.8355906009674072, "step": 1936 }, { "epoch": 0.7147931177637344, "grad_norm": 1.51227867603302, "learning_rate": 0.0001, "loss": 2.8757174015045166, "step": 1937 }, { "epoch": 0.7151621384750219, "grad_norm": 1.3440500497817993, "learning_rate": 0.0001, "loss": 2.8020594120025635, "step": 1938 }, { "epoch": 0.7155311591863094, "grad_norm": 1.471107006072998, "learning_rate": 0.0001, "loss": 2.82119083404541, "step": 1939 }, { "epoch": 0.7159001798975968, "grad_norm": 1.4539724588394165, "learning_rate": 0.0001, "loss": 2.9457240104675293, "step": 1940 }, { "epoch": 0.7162692006088842, "grad_norm": 1.3925291299819946, "learning_rate": 0.0001, "loss": 3.008293390274048, "step": 1941 }, { "epoch": 0.7166382213201716, "grad_norm": 1.518634557723999, "learning_rate": 0.0001, "loss": 2.6514289379119873, "step": 1942 }, { "epoch": 0.717007242031459, "grad_norm": 1.6765588521957397, "learning_rate": 0.0001, "loss": 2.8882694244384766, "step": 1943 }, { "epoch": 0.7173762627427465, "grad_norm": 1.4811643362045288, "learning_rate": 0.0001, "loss": 2.9772448539733887, "step": 1944 }, { "epoch": 0.7177452834540339, "grad_norm": 1.6854859590530396, "learning_rate": 0.0001, "loss": 3.078782796859741, "step": 1945 }, { "epoch": 0.7181143041653213, "grad_norm": 1.355280876159668, "learning_rate": 0.0001, "loss": 2.943895101547241, "step": 1946 }, { "epoch": 0.7184833248766087, "grad_norm": 1.670020341873169, "learning_rate": 0.0001, "loss": 3.156834602355957, "step": 1947 }, { "epoch": 0.7188523455878961, "grad_norm": 1.3620896339416504, "learning_rate": 0.0001, "loss": 2.8302195072174072, "step": 1948 }, { "epoch": 0.7192213662991835, "grad_norm": 1.3761087656021118, "learning_rate": 0.0001, "loss": 2.898786783218384, "step": 1949 }, { "epoch": 0.719590387010471, "grad_norm": 1.6319552659988403, "learning_rate": 0.0001, "loss": 2.79593825340271, "step": 1950 }, { "epoch": 0.7199594077217584, "grad_norm": 2.0058841705322266, "learning_rate": 0.0001, "loss": 2.978482961654663, "step": 1951 }, { "epoch": 0.7203284284330458, "grad_norm": 1.3885670900344849, "learning_rate": 0.0001, "loss": 3.054521083831787, "step": 1952 }, { "epoch": 0.7206974491443332, "grad_norm": 1.4734485149383545, "learning_rate": 0.0001, "loss": 2.9495396614074707, "step": 1953 }, { "epoch": 0.7210664698556206, "grad_norm": 1.3750038146972656, "learning_rate": 0.0001, "loss": 3.0440385341644287, "step": 1954 }, { "epoch": 0.721435490566908, "grad_norm": 1.4042837619781494, "learning_rate": 0.0001, "loss": 3.1852283477783203, "step": 1955 }, { "epoch": 0.7218045112781954, "grad_norm": 1.9339864253997803, "learning_rate": 0.0001, "loss": 2.9673991203308105, "step": 1956 }, { "epoch": 0.7221735319894829, "grad_norm": 1.4691051244735718, "learning_rate": 0.0001, "loss": 2.923004627227783, "step": 1957 }, { "epoch": 0.7225425527007703, "grad_norm": 1.4311037063598633, "learning_rate": 0.0001, "loss": 2.6666676998138428, "step": 1958 }, { "epoch": 0.7229115734120578, "grad_norm": 1.401097297668457, "learning_rate": 0.0001, "loss": 2.829680919647217, "step": 1959 }, { "epoch": 0.7232805941233452, "grad_norm": 1.5826247930526733, "learning_rate": 0.0001, "loss": 3.131427764892578, "step": 1960 }, { "epoch": 0.7236496148346326, "grad_norm": 1.3568289279937744, "learning_rate": 0.0001, "loss": 2.8599374294281006, "step": 1961 }, { "epoch": 0.72401863554592, "grad_norm": 1.329159140586853, "learning_rate": 0.0001, "loss": 2.6987149715423584, "step": 1962 }, { "epoch": 0.7243876562572075, "grad_norm": 1.5744143724441528, "learning_rate": 0.0001, "loss": 3.2294890880584717, "step": 1963 }, { "epoch": 0.7247566769684949, "grad_norm": 1.373091697692871, "learning_rate": 0.0001, "loss": 2.8129632472991943, "step": 1964 }, { "epoch": 0.7251256976797823, "grad_norm": 1.4948590993881226, "learning_rate": 0.0001, "loss": 2.887969970703125, "step": 1965 }, { "epoch": 0.7254947183910697, "grad_norm": 1.5744056701660156, "learning_rate": 0.0001, "loss": 2.9935696125030518, "step": 1966 }, { "epoch": 0.7258637391023571, "grad_norm": 1.5044643878936768, "learning_rate": 0.0001, "loss": 2.9768738746643066, "step": 1967 }, { "epoch": 0.7262327598136445, "grad_norm": 1.245267629623413, "learning_rate": 0.0001, "loss": 2.6165289878845215, "step": 1968 }, { "epoch": 0.726601780524932, "grad_norm": 1.9215524196624756, "learning_rate": 0.0001, "loss": 3.0411410331726074, "step": 1969 }, { "epoch": 0.7269708012362194, "grad_norm": 1.5462615489959717, "learning_rate": 0.0001, "loss": 2.519500494003296, "step": 1970 }, { "epoch": 0.7273398219475068, "grad_norm": 1.3163611888885498, "learning_rate": 0.0001, "loss": 2.7186384201049805, "step": 1971 }, { "epoch": 0.7277088426587942, "grad_norm": 1.416595220565796, "learning_rate": 0.0001, "loss": 3.1218972206115723, "step": 1972 }, { "epoch": 0.7280778633700816, "grad_norm": 1.4077136516571045, "learning_rate": 0.0001, "loss": 2.9508440494537354, "step": 1973 }, { "epoch": 0.728446884081369, "grad_norm": 1.5078394412994385, "learning_rate": 0.0001, "loss": 2.7629551887512207, "step": 1974 }, { "epoch": 0.7288159047926565, "grad_norm": 1.395969033241272, "learning_rate": 0.0001, "loss": 2.9125046730041504, "step": 1975 }, { "epoch": 0.7291849255039439, "grad_norm": 1.5188804864883423, "learning_rate": 0.0001, "loss": 2.8739659786224365, "step": 1976 }, { "epoch": 0.7295539462152313, "grad_norm": 1.355478048324585, "learning_rate": 0.0001, "loss": 2.929112672805786, "step": 1977 }, { "epoch": 0.7299229669265187, "grad_norm": 1.3358606100082397, "learning_rate": 0.0001, "loss": 2.790670871734619, "step": 1978 }, { "epoch": 0.7302919876378061, "grad_norm": 1.538364291191101, "learning_rate": 0.0001, "loss": 3.112060308456421, "step": 1979 }, { "epoch": 0.7306610083490936, "grad_norm": 1.472885012626648, "learning_rate": 0.0001, "loss": 2.865246534347534, "step": 1980 }, { "epoch": 0.7310300290603811, "grad_norm": 1.381571650505066, "learning_rate": 0.0001, "loss": 2.641606569290161, "step": 1981 }, { "epoch": 0.7313990497716685, "grad_norm": 1.3697189092636108, "learning_rate": 0.0001, "loss": 2.6383585929870605, "step": 1982 }, { "epoch": 0.7317680704829559, "grad_norm": 1.557041883468628, "learning_rate": 0.0001, "loss": 2.738731622695923, "step": 1983 }, { "epoch": 0.7321370911942433, "grad_norm": 1.3942391872406006, "learning_rate": 0.0001, "loss": 2.8994410037994385, "step": 1984 }, { "epoch": 0.7325061119055307, "grad_norm": 1.5076404809951782, "learning_rate": 0.0001, "loss": 2.6503405570983887, "step": 1985 }, { "epoch": 0.7328751326168181, "grad_norm": 1.468088984489441, "learning_rate": 0.0001, "loss": 2.5757386684417725, "step": 1986 }, { "epoch": 0.7332441533281056, "grad_norm": 1.445816159248352, "learning_rate": 0.0001, "loss": 2.9912917613983154, "step": 1987 }, { "epoch": 0.733613174039393, "grad_norm": 1.4270931482315063, "learning_rate": 0.0001, "loss": 2.8337669372558594, "step": 1988 }, { "epoch": 0.7339821947506804, "grad_norm": 1.6112340688705444, "learning_rate": 0.0001, "loss": 2.926030158996582, "step": 1989 }, { "epoch": 0.7343512154619678, "grad_norm": 1.798109769821167, "learning_rate": 0.0001, "loss": 2.8248329162597656, "step": 1990 }, { "epoch": 0.7347202361732552, "grad_norm": 1.4782241582870483, "learning_rate": 0.0001, "loss": 2.7862355709075928, "step": 1991 }, { "epoch": 0.7350892568845426, "grad_norm": 1.3989222049713135, "learning_rate": 0.0001, "loss": 2.6875412464141846, "step": 1992 }, { "epoch": 0.73545827759583, "grad_norm": 4.146024703979492, "learning_rate": 0.0001, "loss": 3.0937845706939697, "step": 1993 }, { "epoch": 0.7358272983071175, "grad_norm": 1.6432809829711914, "learning_rate": 0.0001, "loss": 2.841672420501709, "step": 1994 }, { "epoch": 0.7361963190184049, "grad_norm": 1.6712143421173096, "learning_rate": 0.0001, "loss": 2.8171515464782715, "step": 1995 }, { "epoch": 0.7365653397296923, "grad_norm": 1.3555632829666138, "learning_rate": 0.0001, "loss": 2.6634483337402344, "step": 1996 }, { "epoch": 0.7369343604409797, "grad_norm": 1.6375086307525635, "learning_rate": 0.0001, "loss": 3.071199417114258, "step": 1997 }, { "epoch": 0.7373033811522671, "grad_norm": 1.8251913785934448, "learning_rate": 0.0001, "loss": 3.201692819595337, "step": 1998 }, { "epoch": 0.7376724018635545, "grad_norm": 1.4557583332061768, "learning_rate": 0.0001, "loss": 2.930026054382324, "step": 1999 }, { "epoch": 0.7380414225748421, "grad_norm": 1.4042524099349976, "learning_rate": 0.0001, "loss": 3.025254011154175, "step": 2000 }, { "epoch": 0.7384104432861295, "grad_norm": 1.4500890970230103, "learning_rate": 0.0001, "loss": 3.1806561946868896, "step": 2001 }, { "epoch": 0.7387794639974169, "grad_norm": 1.2467339038848877, "learning_rate": 0.0001, "loss": 2.861098527908325, "step": 2002 }, { "epoch": 0.7391484847087043, "grad_norm": 1.600602149963379, "learning_rate": 0.0001, "loss": 2.7730369567871094, "step": 2003 }, { "epoch": 0.7395175054199917, "grad_norm": 1.4987252950668335, "learning_rate": 0.0001, "loss": 2.7257254123687744, "step": 2004 }, { "epoch": 0.7398865261312791, "grad_norm": 1.5078377723693848, "learning_rate": 0.0001, "loss": 2.6971306800842285, "step": 2005 }, { "epoch": 0.7402555468425666, "grad_norm": 1.33574378490448, "learning_rate": 0.0001, "loss": 2.877000331878662, "step": 2006 }, { "epoch": 0.740624567553854, "grad_norm": 1.274319052696228, "learning_rate": 0.0001, "loss": 2.800767421722412, "step": 2007 }, { "epoch": 0.7409935882651414, "grad_norm": 1.511022925376892, "learning_rate": 0.0001, "loss": 2.901646137237549, "step": 2008 }, { "epoch": 0.7413626089764288, "grad_norm": 1.5567617416381836, "learning_rate": 0.0001, "loss": 3.1679234504699707, "step": 2009 }, { "epoch": 0.7417316296877162, "grad_norm": 1.3998233079910278, "learning_rate": 0.0001, "loss": 2.9820773601531982, "step": 2010 }, { "epoch": 0.7421006503990036, "grad_norm": 1.3501412868499756, "learning_rate": 0.0001, "loss": 2.7470524311065674, "step": 2011 }, { "epoch": 0.742469671110291, "grad_norm": 1.5474804639816284, "learning_rate": 0.0001, "loss": 2.9939990043640137, "step": 2012 }, { "epoch": 0.7428386918215785, "grad_norm": 1.4111260175704956, "learning_rate": 0.0001, "loss": 2.9991772174835205, "step": 2013 }, { "epoch": 0.7432077125328659, "grad_norm": 1.6329808235168457, "learning_rate": 0.0001, "loss": 2.7530086040496826, "step": 2014 }, { "epoch": 0.7435767332441533, "grad_norm": 1.5864663124084473, "learning_rate": 0.0001, "loss": 2.7948145866394043, "step": 2015 }, { "epoch": 0.7439457539554407, "grad_norm": 1.3670873641967773, "learning_rate": 0.0001, "loss": 3.0830461978912354, "step": 2016 }, { "epoch": 0.7443147746667281, "grad_norm": 1.9852230548858643, "learning_rate": 0.0001, "loss": 2.976033926010132, "step": 2017 }, { "epoch": 0.7446837953780155, "grad_norm": 1.5994677543640137, "learning_rate": 0.0001, "loss": 3.044945240020752, "step": 2018 }, { "epoch": 0.745052816089303, "grad_norm": 1.7174354791641235, "learning_rate": 0.0001, "loss": 2.914764404296875, "step": 2019 }, { "epoch": 0.7454218368005905, "grad_norm": 1.4436919689178467, "learning_rate": 0.0001, "loss": 2.652708053588867, "step": 2020 }, { "epoch": 0.7457908575118779, "grad_norm": 1.4210989475250244, "learning_rate": 0.0001, "loss": 3.1993050575256348, "step": 2021 }, { "epoch": 0.7461598782231653, "grad_norm": 1.2840009927749634, "learning_rate": 0.0001, "loss": 2.909489154815674, "step": 2022 }, { "epoch": 0.7465288989344527, "grad_norm": 1.3904774188995361, "learning_rate": 0.0001, "loss": 2.797043800354004, "step": 2023 }, { "epoch": 0.7468979196457401, "grad_norm": 1.5861992835998535, "learning_rate": 0.0001, "loss": 3.2670280933380127, "step": 2024 }, { "epoch": 0.7472669403570276, "grad_norm": 1.6044795513153076, "learning_rate": 0.0001, "loss": 3.4284918308258057, "step": 2025 }, { "epoch": 0.747635961068315, "grad_norm": 1.3921892642974854, "learning_rate": 0.0001, "loss": 3.0070245265960693, "step": 2026 }, { "epoch": 0.7480049817796024, "grad_norm": 1.488046646118164, "learning_rate": 0.0001, "loss": 2.925715446472168, "step": 2027 }, { "epoch": 0.7483740024908898, "grad_norm": 1.5045620203018188, "learning_rate": 0.0001, "loss": 3.008768320083618, "step": 2028 }, { "epoch": 0.7487430232021772, "grad_norm": 1.3111097812652588, "learning_rate": 0.0001, "loss": 3.075441837310791, "step": 2029 }, { "epoch": 0.7491120439134646, "grad_norm": 1.3056654930114746, "learning_rate": 0.0001, "loss": 3.0047190189361572, "step": 2030 }, { "epoch": 0.7494810646247521, "grad_norm": 1.8801302909851074, "learning_rate": 0.0001, "loss": 3.139188766479492, "step": 2031 }, { "epoch": 0.7498500853360395, "grad_norm": 1.4859505891799927, "learning_rate": 0.0001, "loss": 2.8372485637664795, "step": 2032 }, { "epoch": 0.7502191060473269, "grad_norm": 1.3154526948928833, "learning_rate": 0.0001, "loss": 2.7965710163116455, "step": 2033 }, { "epoch": 0.7505881267586143, "grad_norm": 1.9756193161010742, "learning_rate": 0.0001, "loss": 2.9413933753967285, "step": 2034 }, { "epoch": 0.7509571474699017, "grad_norm": 1.9105099439620972, "learning_rate": 0.0001, "loss": 2.780482530593872, "step": 2035 }, { "epoch": 0.7513261681811891, "grad_norm": 1.4869881868362427, "learning_rate": 0.0001, "loss": 2.896604061126709, "step": 2036 }, { "epoch": 0.7516951888924766, "grad_norm": 1.5386766195297241, "learning_rate": 0.0001, "loss": 3.078188180923462, "step": 2037 }, { "epoch": 0.752064209603764, "grad_norm": 1.4198232889175415, "learning_rate": 0.0001, "loss": 2.8711838722229004, "step": 2038 }, { "epoch": 0.7524332303150514, "grad_norm": 1.4420826435089111, "learning_rate": 0.0001, "loss": 2.929636240005493, "step": 2039 }, { "epoch": 0.7528022510263388, "grad_norm": 1.36516535282135, "learning_rate": 0.0001, "loss": 2.804741859436035, "step": 2040 }, { "epoch": 0.7531712717376263, "grad_norm": 1.4359532594680786, "learning_rate": 0.0001, "loss": 3.135324716567993, "step": 2041 }, { "epoch": 0.7535402924489137, "grad_norm": 1.8480819463729858, "learning_rate": 0.0001, "loss": 3.6172842979431152, "step": 2042 }, { "epoch": 0.7539093131602012, "grad_norm": 1.681736946105957, "learning_rate": 0.0001, "loss": 3.114297389984131, "step": 2043 }, { "epoch": 0.7542783338714886, "grad_norm": 1.5505201816558838, "learning_rate": 0.0001, "loss": 2.80375599861145, "step": 2044 }, { "epoch": 0.754647354582776, "grad_norm": 1.2358726263046265, "learning_rate": 0.0001, "loss": 2.958815097808838, "step": 2045 }, { "epoch": 0.7550163752940634, "grad_norm": 1.4011297225952148, "learning_rate": 0.0001, "loss": 2.9087471961975098, "step": 2046 }, { "epoch": 0.7553853960053508, "grad_norm": 1.393249273300171, "learning_rate": 0.0001, "loss": 2.8047842979431152, "step": 2047 }, { "epoch": 0.7557544167166382, "grad_norm": 1.3402717113494873, "learning_rate": 0.0001, "loss": 2.883533239364624, "step": 2048 }, { "epoch": 0.7561234374279256, "grad_norm": 1.2944520711898804, "learning_rate": 0.0001, "loss": 2.856254816055298, "step": 2049 }, { "epoch": 0.7564924581392131, "grad_norm": 1.3056282997131348, "learning_rate": 0.0001, "loss": 2.8576488494873047, "step": 2050 }, { "epoch": 0.7568614788505005, "grad_norm": 1.5572243928909302, "learning_rate": 0.0001, "loss": 3.171898603439331, "step": 2051 }, { "epoch": 0.7572304995617879, "grad_norm": 1.3311290740966797, "learning_rate": 0.0001, "loss": 2.906850576400757, "step": 2052 }, { "epoch": 0.7575995202730753, "grad_norm": 1.2939544916152954, "learning_rate": 0.0001, "loss": 2.899794340133667, "step": 2053 }, { "epoch": 0.7579685409843627, "grad_norm": 1.3422843217849731, "learning_rate": 0.0001, "loss": 2.7020466327667236, "step": 2054 }, { "epoch": 0.7583375616956501, "grad_norm": 1.447121262550354, "learning_rate": 0.0001, "loss": 2.9802350997924805, "step": 2055 }, { "epoch": 0.7587065824069376, "grad_norm": 1.2349321842193604, "learning_rate": 0.0001, "loss": 2.825005531311035, "step": 2056 }, { "epoch": 0.759075603118225, "grad_norm": 1.5183897018432617, "learning_rate": 0.0001, "loss": 2.8741931915283203, "step": 2057 }, { "epoch": 0.7594446238295124, "grad_norm": 1.5560967922210693, "learning_rate": 0.0001, "loss": 2.605239152908325, "step": 2058 }, { "epoch": 0.7598136445407998, "grad_norm": 1.4999581575393677, "learning_rate": 0.0001, "loss": 2.7078697681427, "step": 2059 }, { "epoch": 0.7601826652520872, "grad_norm": 1.3336308002471924, "learning_rate": 0.0001, "loss": 2.785601854324341, "step": 2060 }, { "epoch": 0.7605516859633747, "grad_norm": 1.2708957195281982, "learning_rate": 0.0001, "loss": 2.5409998893737793, "step": 2061 }, { "epoch": 0.7609207066746622, "grad_norm": 1.3956071138381958, "learning_rate": 0.0001, "loss": 2.8070106506347656, "step": 2062 }, { "epoch": 0.7612897273859496, "grad_norm": 1.8451275825500488, "learning_rate": 0.0001, "loss": 3.173597574234009, "step": 2063 }, { "epoch": 0.761658748097237, "grad_norm": 1.7274500131607056, "learning_rate": 0.0001, "loss": 2.8510959148406982, "step": 2064 }, { "epoch": 0.7620277688085244, "grad_norm": 2.209097146987915, "learning_rate": 0.0001, "loss": 2.785370349884033, "step": 2065 }, { "epoch": 0.7623967895198118, "grad_norm": 1.4258315563201904, "learning_rate": 0.0001, "loss": 3.002084255218506, "step": 2066 }, { "epoch": 0.7627658102310992, "grad_norm": 1.4715402126312256, "learning_rate": 0.0001, "loss": 2.7496345043182373, "step": 2067 }, { "epoch": 0.7631348309423867, "grad_norm": 1.4964547157287598, "learning_rate": 0.0001, "loss": 2.913487195968628, "step": 2068 }, { "epoch": 0.7635038516536741, "grad_norm": 1.4587249755859375, "learning_rate": 0.0001, "loss": 2.671722888946533, "step": 2069 }, { "epoch": 0.7638728723649615, "grad_norm": 1.5689855813980103, "learning_rate": 0.0001, "loss": 2.7324419021606445, "step": 2070 }, { "epoch": 0.7642418930762489, "grad_norm": 1.3486673831939697, "learning_rate": 0.0001, "loss": 2.8513009548187256, "step": 2071 }, { "epoch": 0.7646109137875363, "grad_norm": 1.4199737310409546, "learning_rate": 0.0001, "loss": 2.5912599563598633, "step": 2072 }, { "epoch": 0.7649799344988237, "grad_norm": 1.4620234966278076, "learning_rate": 0.0001, "loss": 3.0144550800323486, "step": 2073 }, { "epoch": 0.7653489552101111, "grad_norm": 1.8056226968765259, "learning_rate": 0.0001, "loss": 2.740670680999756, "step": 2074 }, { "epoch": 0.7657179759213986, "grad_norm": 1.7832602262496948, "learning_rate": 0.0001, "loss": 2.845080852508545, "step": 2075 }, { "epoch": 0.766086996632686, "grad_norm": 1.45360267162323, "learning_rate": 0.0001, "loss": 3.0183839797973633, "step": 2076 }, { "epoch": 0.7664560173439734, "grad_norm": 1.7484989166259766, "learning_rate": 0.0001, "loss": 3.261448860168457, "step": 2077 }, { "epoch": 0.7668250380552608, "grad_norm": 1.515965223312378, "learning_rate": 0.0001, "loss": 3.1434056758880615, "step": 2078 }, { "epoch": 0.7671940587665482, "grad_norm": 1.4964405298233032, "learning_rate": 0.0001, "loss": 2.9421215057373047, "step": 2079 }, { "epoch": 0.7675630794778356, "grad_norm": 2.0405731201171875, "learning_rate": 0.0001, "loss": 3.0586788654327393, "step": 2080 }, { "epoch": 0.7679321001891232, "grad_norm": 1.150970697402954, "learning_rate": 0.0001, "loss": 2.6557629108428955, "step": 2081 }, { "epoch": 0.7683011209004106, "grad_norm": 1.497132658958435, "learning_rate": 0.0001, "loss": 2.8477468490600586, "step": 2082 }, { "epoch": 0.768670141611698, "grad_norm": 1.4393168687820435, "learning_rate": 0.0001, "loss": 2.7623608112335205, "step": 2083 }, { "epoch": 0.7690391623229854, "grad_norm": 1.3218929767608643, "learning_rate": 0.0001, "loss": 2.8696107864379883, "step": 2084 }, { "epoch": 0.7694081830342728, "grad_norm": 1.5987564325332642, "learning_rate": 0.0001, "loss": 2.9628794193267822, "step": 2085 }, { "epoch": 0.7697772037455602, "grad_norm": 1.3627783060073853, "learning_rate": 0.0001, "loss": 2.6830244064331055, "step": 2086 }, { "epoch": 0.7701462244568477, "grad_norm": 1.7660342454910278, "learning_rate": 0.0001, "loss": 3.2251641750335693, "step": 2087 }, { "epoch": 0.7705152451681351, "grad_norm": 1.3810138702392578, "learning_rate": 0.0001, "loss": 2.7580552101135254, "step": 2088 }, { "epoch": 0.7708842658794225, "grad_norm": 1.5499948263168335, "learning_rate": 0.0001, "loss": 2.6102685928344727, "step": 2089 }, { "epoch": 0.7712532865907099, "grad_norm": 1.4926974773406982, "learning_rate": 0.0001, "loss": 2.5623362064361572, "step": 2090 }, { "epoch": 0.7716223073019973, "grad_norm": 1.6133084297180176, "learning_rate": 0.0001, "loss": 2.6930906772613525, "step": 2091 }, { "epoch": 0.7719913280132847, "grad_norm": 1.5596905946731567, "learning_rate": 0.0001, "loss": 2.5821666717529297, "step": 2092 }, { "epoch": 0.7723603487245722, "grad_norm": 1.5830354690551758, "learning_rate": 0.0001, "loss": 3.011178731918335, "step": 2093 }, { "epoch": 0.7727293694358596, "grad_norm": 1.3379530906677246, "learning_rate": 0.0001, "loss": 3.2109463214874268, "step": 2094 }, { "epoch": 0.773098390147147, "grad_norm": 1.3382011651992798, "learning_rate": 0.0001, "loss": 2.9571571350097656, "step": 2095 }, { "epoch": 0.7734674108584344, "grad_norm": 4.172677040100098, "learning_rate": 0.0001, "loss": 3.2020719051361084, "step": 2096 }, { "epoch": 0.7738364315697218, "grad_norm": 1.396220326423645, "learning_rate": 0.0001, "loss": 2.995844841003418, "step": 2097 }, { "epoch": 0.7742054522810092, "grad_norm": 1.4525972604751587, "learning_rate": 0.0001, "loss": 2.5245261192321777, "step": 2098 }, { "epoch": 0.7745744729922966, "grad_norm": 1.7951250076293945, "learning_rate": 0.0001, "loss": 3.3059446811676025, "step": 2099 }, { "epoch": 0.7749434937035841, "grad_norm": 1.2973318099975586, "learning_rate": 0.0001, "loss": 2.734253406524658, "step": 2100 }, { "epoch": 0.7753125144148716, "grad_norm": 1.450643539428711, "learning_rate": 0.0001, "loss": 3.121037721633911, "step": 2101 }, { "epoch": 0.775681535126159, "grad_norm": 1.4518557786941528, "learning_rate": 0.0001, "loss": 2.4914710521698, "step": 2102 }, { "epoch": 0.7760505558374464, "grad_norm": 1.504050374031067, "learning_rate": 0.0001, "loss": 2.6494956016540527, "step": 2103 }, { "epoch": 0.7764195765487338, "grad_norm": 1.2799384593963623, "learning_rate": 0.0001, "loss": 2.440486431121826, "step": 2104 }, { "epoch": 0.7767885972600213, "grad_norm": 2.1463959217071533, "learning_rate": 0.0001, "loss": 2.7819137573242188, "step": 2105 }, { "epoch": 0.7771576179713087, "grad_norm": 1.534495234489441, "learning_rate": 0.0001, "loss": 2.9140965938568115, "step": 2106 }, { "epoch": 0.7775266386825961, "grad_norm": 2.2302684783935547, "learning_rate": 0.0001, "loss": 3.4044222831726074, "step": 2107 }, { "epoch": 0.7778956593938835, "grad_norm": 1.3098030090332031, "learning_rate": 0.0001, "loss": 2.99943470954895, "step": 2108 }, { "epoch": 0.7782646801051709, "grad_norm": 2.020843982696533, "learning_rate": 0.0001, "loss": 2.754470109939575, "step": 2109 }, { "epoch": 0.7786337008164583, "grad_norm": 1.2741507291793823, "learning_rate": 0.0001, "loss": 2.6821963787078857, "step": 2110 }, { "epoch": 0.7790027215277457, "grad_norm": 1.2599400281906128, "learning_rate": 0.0001, "loss": 2.64989972114563, "step": 2111 }, { "epoch": 0.7793717422390332, "grad_norm": 1.2309006452560425, "learning_rate": 0.0001, "loss": 3.040513753890991, "step": 2112 }, { "epoch": 0.7797407629503206, "grad_norm": 1.3600349426269531, "learning_rate": 0.0001, "loss": 2.787925958633423, "step": 2113 }, { "epoch": 0.780109783661608, "grad_norm": 1.387922763824463, "learning_rate": 0.0001, "loss": 2.947859048843384, "step": 2114 }, { "epoch": 0.7804788043728954, "grad_norm": 1.372092843055725, "learning_rate": 0.0001, "loss": 2.805586338043213, "step": 2115 }, { "epoch": 0.7808478250841828, "grad_norm": 1.3766084909439087, "learning_rate": 0.0001, "loss": 2.800447940826416, "step": 2116 }, { "epoch": 0.7812168457954702, "grad_norm": 1.3142141103744507, "learning_rate": 0.0001, "loss": 2.565390110015869, "step": 2117 }, { "epoch": 0.7815858665067577, "grad_norm": 1.264072299003601, "learning_rate": 0.0001, "loss": 2.6764588356018066, "step": 2118 }, { "epoch": 0.7819548872180451, "grad_norm": 1.576309323310852, "learning_rate": 0.0001, "loss": 2.548931121826172, "step": 2119 }, { "epoch": 0.7823239079293325, "grad_norm": 1.6410679817199707, "learning_rate": 0.0001, "loss": 2.7313623428344727, "step": 2120 }, { "epoch": 0.7826929286406199, "grad_norm": 1.7777352333068848, "learning_rate": 0.0001, "loss": 3.319873094558716, "step": 2121 }, { "epoch": 0.7830619493519074, "grad_norm": 1.8877464532852173, "learning_rate": 0.0001, "loss": 3.1459150314331055, "step": 2122 }, { "epoch": 0.7834309700631948, "grad_norm": 1.7062116861343384, "learning_rate": 0.0001, "loss": 2.8078203201293945, "step": 2123 }, { "epoch": 0.7837999907744823, "grad_norm": 1.2992392778396606, "learning_rate": 0.0001, "loss": 2.895521402359009, "step": 2124 }, { "epoch": 0.7841690114857697, "grad_norm": 1.6809476613998413, "learning_rate": 0.0001, "loss": 2.7038402557373047, "step": 2125 }, { "epoch": 0.7845380321970571, "grad_norm": 1.6137171983718872, "learning_rate": 0.0001, "loss": 2.717979907989502, "step": 2126 }, { "epoch": 0.7849070529083445, "grad_norm": 1.6056859493255615, "learning_rate": 0.0001, "loss": 2.887068033218384, "step": 2127 }, { "epoch": 0.7852760736196319, "grad_norm": 3.266723871231079, "learning_rate": 0.0001, "loss": 3.0682601928710938, "step": 2128 }, { "epoch": 0.7856450943309193, "grad_norm": 1.287630319595337, "learning_rate": 0.0001, "loss": 2.707812786102295, "step": 2129 }, { "epoch": 0.7860141150422068, "grad_norm": 1.775531530380249, "learning_rate": 0.0001, "loss": 2.6979894638061523, "step": 2130 }, { "epoch": 0.7863831357534942, "grad_norm": 1.709935188293457, "learning_rate": 0.0001, "loss": 2.8292548656463623, "step": 2131 }, { "epoch": 0.7867521564647816, "grad_norm": 2.4040794372558594, "learning_rate": 0.0001, "loss": 3.0629560947418213, "step": 2132 }, { "epoch": 0.787121177176069, "grad_norm": 2.395374298095703, "learning_rate": 0.0001, "loss": 3.3432130813598633, "step": 2133 }, { "epoch": 0.7874901978873564, "grad_norm": 1.4066134691238403, "learning_rate": 0.0001, "loss": 2.6744813919067383, "step": 2134 }, { "epoch": 0.7878592185986438, "grad_norm": 1.2354308366775513, "learning_rate": 0.0001, "loss": 2.7047038078308105, "step": 2135 }, { "epoch": 0.7882282393099312, "grad_norm": 1.3733752965927124, "learning_rate": 0.0001, "loss": 2.6595046520233154, "step": 2136 }, { "epoch": 0.7885972600212187, "grad_norm": 1.3449159860610962, "learning_rate": 0.0001, "loss": 2.943467140197754, "step": 2137 }, { "epoch": 0.7889662807325061, "grad_norm": 1.3382563591003418, "learning_rate": 0.0001, "loss": 3.1758599281311035, "step": 2138 }, { "epoch": 0.7893353014437935, "grad_norm": 1.2781132459640503, "learning_rate": 0.0001, "loss": 2.7101354598999023, "step": 2139 }, { "epoch": 0.7897043221550809, "grad_norm": 1.2511003017425537, "learning_rate": 0.0001, "loss": 2.619015693664551, "step": 2140 }, { "epoch": 0.7900733428663683, "grad_norm": 2.177109479904175, "learning_rate": 0.0001, "loss": 3.161228895187378, "step": 2141 }, { "epoch": 0.7904423635776558, "grad_norm": 1.2281290292739868, "learning_rate": 0.0001, "loss": 2.483548879623413, "step": 2142 }, { "epoch": 0.7908113842889433, "grad_norm": 1.5228513479232788, "learning_rate": 0.0001, "loss": 3.1720468997955322, "step": 2143 }, { "epoch": 0.7911804050002307, "grad_norm": 1.0985585451126099, "learning_rate": 0.0001, "loss": 2.552572250366211, "step": 2144 }, { "epoch": 0.7915494257115181, "grad_norm": 1.404768943786621, "learning_rate": 0.0001, "loss": 2.9454822540283203, "step": 2145 }, { "epoch": 0.7919184464228055, "grad_norm": 1.537357211112976, "learning_rate": 0.0001, "loss": 3.032482624053955, "step": 2146 }, { "epoch": 0.7922874671340929, "grad_norm": 1.6196309328079224, "learning_rate": 0.0001, "loss": 3.014634132385254, "step": 2147 }, { "epoch": 0.7926564878453803, "grad_norm": 1.4771015644073486, "learning_rate": 0.0001, "loss": 2.777270793914795, "step": 2148 }, { "epoch": 0.7930255085566678, "grad_norm": 1.604552984237671, "learning_rate": 0.0001, "loss": 2.6396617889404297, "step": 2149 }, { "epoch": 0.7933945292679552, "grad_norm": 1.349014401435852, "learning_rate": 0.0001, "loss": 2.728027105331421, "step": 2150 }, { "epoch": 0.7937635499792426, "grad_norm": 1.7365269660949707, "learning_rate": 0.0001, "loss": 3.0298194885253906, "step": 2151 }, { "epoch": 0.79413257069053, "grad_norm": 1.4154328107833862, "learning_rate": 0.0001, "loss": 2.9011080265045166, "step": 2152 }, { "epoch": 0.7945015914018174, "grad_norm": 1.322675108909607, "learning_rate": 0.0001, "loss": 2.9603075981140137, "step": 2153 }, { "epoch": 0.7948706121131048, "grad_norm": 1.478181004524231, "learning_rate": 0.0001, "loss": 2.936394214630127, "step": 2154 }, { "epoch": 0.7952396328243923, "grad_norm": 1.6514900922775269, "learning_rate": 0.0001, "loss": 2.9290857315063477, "step": 2155 }, { "epoch": 0.7956086535356797, "grad_norm": 1.4190984964370728, "learning_rate": 0.0001, "loss": 2.7810635566711426, "step": 2156 }, { "epoch": 0.7959776742469671, "grad_norm": 1.3994871377944946, "learning_rate": 0.0001, "loss": 2.6723806858062744, "step": 2157 }, { "epoch": 0.7963466949582545, "grad_norm": 1.5812629461288452, "learning_rate": 0.0001, "loss": 2.641390800476074, "step": 2158 }, { "epoch": 0.7967157156695419, "grad_norm": 1.3211990594863892, "learning_rate": 0.0001, "loss": 2.9209840297698975, "step": 2159 }, { "epoch": 0.7970847363808293, "grad_norm": 1.5397306680679321, "learning_rate": 0.0001, "loss": 2.8861312866210938, "step": 2160 }, { "epoch": 0.7974537570921167, "grad_norm": 1.635932445526123, "learning_rate": 0.0001, "loss": 3.055690050125122, "step": 2161 }, { "epoch": 0.7978227778034043, "grad_norm": 1.4576150178909302, "learning_rate": 0.0001, "loss": 2.966693878173828, "step": 2162 }, { "epoch": 0.7981917985146917, "grad_norm": 1.3281359672546387, "learning_rate": 0.0001, "loss": 2.501849889755249, "step": 2163 }, { "epoch": 0.7985608192259791, "grad_norm": 1.3197970390319824, "learning_rate": 0.0001, "loss": 2.7323904037475586, "step": 2164 }, { "epoch": 0.7989298399372665, "grad_norm": 1.5671765804290771, "learning_rate": 0.0001, "loss": 2.709599256515503, "step": 2165 }, { "epoch": 0.7992988606485539, "grad_norm": 1.4521089792251587, "learning_rate": 0.0001, "loss": 2.6312050819396973, "step": 2166 }, { "epoch": 0.7996678813598413, "grad_norm": 1.4199570417404175, "learning_rate": 0.0001, "loss": 2.578972339630127, "step": 2167 }, { "epoch": 0.8000369020711288, "grad_norm": 1.8000080585479736, "learning_rate": 0.0001, "loss": 2.8682992458343506, "step": 2168 }, { "epoch": 0.8004059227824162, "grad_norm": 1.4531214237213135, "learning_rate": 0.0001, "loss": 2.613320827484131, "step": 2169 }, { "epoch": 0.8007749434937036, "grad_norm": 1.4780194759368896, "learning_rate": 0.0001, "loss": 2.6423733234405518, "step": 2170 }, { "epoch": 0.801143964204991, "grad_norm": 1.5813509225845337, "learning_rate": 0.0001, "loss": 2.674984931945801, "step": 2171 }, { "epoch": 0.8015129849162784, "grad_norm": 1.6838464736938477, "learning_rate": 0.0001, "loss": 2.937983989715576, "step": 2172 }, { "epoch": 0.8018820056275658, "grad_norm": 1.8748642206192017, "learning_rate": 0.0001, "loss": 2.9707999229431152, "step": 2173 }, { "epoch": 0.8022510263388533, "grad_norm": 1.283101201057434, "learning_rate": 0.0001, "loss": 3.186964988708496, "step": 2174 }, { "epoch": 0.8026200470501407, "grad_norm": 1.5796922445297241, "learning_rate": 0.0001, "loss": 2.8713607788085938, "step": 2175 }, { "epoch": 0.8029890677614281, "grad_norm": 1.5021642446517944, "learning_rate": 0.0001, "loss": 2.656240224838257, "step": 2176 }, { "epoch": 0.8033580884727155, "grad_norm": 1.5513958930969238, "learning_rate": 0.0001, "loss": 3.0191383361816406, "step": 2177 }, { "epoch": 0.8037271091840029, "grad_norm": 1.796540379524231, "learning_rate": 0.0001, "loss": 3.0245022773742676, "step": 2178 }, { "epoch": 0.8040961298952903, "grad_norm": 1.4250386953353882, "learning_rate": 0.0001, "loss": 2.752445697784424, "step": 2179 }, { "epoch": 0.8044651506065778, "grad_norm": 1.8393974304199219, "learning_rate": 0.0001, "loss": 2.5349652767181396, "step": 2180 }, { "epoch": 0.8048341713178652, "grad_norm": 1.4794212579727173, "learning_rate": 0.0001, "loss": 2.6462838649749756, "step": 2181 }, { "epoch": 0.8052031920291526, "grad_norm": 1.5751160383224487, "learning_rate": 0.0001, "loss": 3.009155035018921, "step": 2182 }, { "epoch": 0.8055722127404401, "grad_norm": 1.7004485130310059, "learning_rate": 0.0001, "loss": 2.749986410140991, "step": 2183 }, { "epoch": 0.8059412334517275, "grad_norm": 1.598658561706543, "learning_rate": 0.0001, "loss": 2.9665043354034424, "step": 2184 }, { "epoch": 0.8063102541630149, "grad_norm": 1.4222527742385864, "learning_rate": 0.0001, "loss": 2.762340784072876, "step": 2185 }, { "epoch": 0.8066792748743024, "grad_norm": 1.829885482788086, "learning_rate": 0.0001, "loss": 3.0331404209136963, "step": 2186 }, { "epoch": 0.8070482955855898, "grad_norm": 1.8107000589370728, "learning_rate": 0.0001, "loss": 3.0382332801818848, "step": 2187 }, { "epoch": 0.8074173162968772, "grad_norm": 1.328037977218628, "learning_rate": 0.0001, "loss": 2.692668914794922, "step": 2188 }, { "epoch": 0.8077863370081646, "grad_norm": 1.3931996822357178, "learning_rate": 0.0001, "loss": 2.938567876815796, "step": 2189 }, { "epoch": 0.808155357719452, "grad_norm": 1.4583961963653564, "learning_rate": 0.0001, "loss": 2.5460586547851562, "step": 2190 }, { "epoch": 0.8085243784307394, "grad_norm": 1.579506278038025, "learning_rate": 0.0001, "loss": 2.9148147106170654, "step": 2191 }, { "epoch": 0.8088933991420268, "grad_norm": 1.2903050184249878, "learning_rate": 0.0001, "loss": 2.7893199920654297, "step": 2192 }, { "epoch": 0.8092624198533143, "grad_norm": 1.3901500701904297, "learning_rate": 0.0001, "loss": 2.569577932357788, "step": 2193 }, { "epoch": 0.8096314405646017, "grad_norm": 1.337532877922058, "learning_rate": 0.0001, "loss": 2.5224831104278564, "step": 2194 }, { "epoch": 0.8100004612758891, "grad_norm": 1.2765562534332275, "learning_rate": 0.0001, "loss": 2.8100814819335938, "step": 2195 }, { "epoch": 0.8103694819871765, "grad_norm": 1.8989681005477905, "learning_rate": 0.0001, "loss": 3.1207480430603027, "step": 2196 }, { "epoch": 0.8107385026984639, "grad_norm": 1.238049030303955, "learning_rate": 0.0001, "loss": 2.7239108085632324, "step": 2197 }, { "epoch": 0.8111075234097513, "grad_norm": 1.2485313415527344, "learning_rate": 0.0001, "loss": 2.784759759902954, "step": 2198 }, { "epoch": 0.8114765441210388, "grad_norm": 1.6310373544692993, "learning_rate": 0.0001, "loss": 2.528494358062744, "step": 2199 }, { "epoch": 0.8118455648323262, "grad_norm": 1.6298922300338745, "learning_rate": 0.0001, "loss": 2.5784249305725098, "step": 2200 }, { "epoch": 0.8122145855436136, "grad_norm": 1.3539034128189087, "learning_rate": 0.0001, "loss": 2.6481680870056152, "step": 2201 }, { "epoch": 0.812583606254901, "grad_norm": 1.2915791273117065, "learning_rate": 0.0001, "loss": 2.907034158706665, "step": 2202 }, { "epoch": 0.8129526269661885, "grad_norm": 1.3400006294250488, "learning_rate": 0.0001, "loss": 2.8638620376586914, "step": 2203 }, { "epoch": 0.813321647677476, "grad_norm": 1.3882219791412354, "learning_rate": 0.0001, "loss": 2.748488426208496, "step": 2204 }, { "epoch": 0.8136906683887634, "grad_norm": 1.3742769956588745, "learning_rate": 0.0001, "loss": 2.9029321670532227, "step": 2205 }, { "epoch": 0.8140596891000508, "grad_norm": 1.3791981935501099, "learning_rate": 0.0001, "loss": 2.8556437492370605, "step": 2206 }, { "epoch": 0.8144287098113382, "grad_norm": 1.8558706045150757, "learning_rate": 0.0001, "loss": 2.9949567317962646, "step": 2207 }, { "epoch": 0.8147977305226256, "grad_norm": 1.7949599027633667, "learning_rate": 0.0001, "loss": 2.913783550262451, "step": 2208 }, { "epoch": 0.815166751233913, "grad_norm": 1.3094298839569092, "learning_rate": 0.0001, "loss": 2.5644593238830566, "step": 2209 }, { "epoch": 0.8155357719452004, "grad_norm": 1.4265516996383667, "learning_rate": 0.0001, "loss": 2.583315134048462, "step": 2210 }, { "epoch": 0.8159047926564879, "grad_norm": 1.3417677879333496, "learning_rate": 0.0001, "loss": 2.842949390411377, "step": 2211 }, { "epoch": 0.8162738133677753, "grad_norm": 1.340125322341919, "learning_rate": 0.0001, "loss": 2.4887807369232178, "step": 2212 }, { "epoch": 0.8166428340790627, "grad_norm": 1.408888578414917, "learning_rate": 0.0001, "loss": 3.1525044441223145, "step": 2213 }, { "epoch": 0.8170118547903501, "grad_norm": 1.8607455492019653, "learning_rate": 0.0001, "loss": 3.201085090637207, "step": 2214 }, { "epoch": 0.8173808755016375, "grad_norm": 1.358332633972168, "learning_rate": 0.0001, "loss": 2.678880214691162, "step": 2215 }, { "epoch": 0.8177498962129249, "grad_norm": 1.268891453742981, "learning_rate": 0.0001, "loss": 2.776399850845337, "step": 2216 }, { "epoch": 0.8181189169242123, "grad_norm": 1.5241986513137817, "learning_rate": 0.0001, "loss": 2.90661883354187, "step": 2217 }, { "epoch": 0.8184879376354998, "grad_norm": 1.3744136095046997, "learning_rate": 0.0001, "loss": 2.7732930183410645, "step": 2218 }, { "epoch": 0.8188569583467872, "grad_norm": 1.511786937713623, "learning_rate": 0.0001, "loss": 2.9151206016540527, "step": 2219 }, { "epoch": 0.8192259790580746, "grad_norm": 1.1658916473388672, "learning_rate": 0.0001, "loss": 2.5359065532684326, "step": 2220 }, { "epoch": 0.819594999769362, "grad_norm": 1.5300307273864746, "learning_rate": 0.0001, "loss": 3.0853326320648193, "step": 2221 }, { "epoch": 0.8199640204806494, "grad_norm": 1.2514275312423706, "learning_rate": 0.0001, "loss": 2.7133419513702393, "step": 2222 }, { "epoch": 0.820333041191937, "grad_norm": 2.0529110431671143, "learning_rate": 0.0001, "loss": 2.9847869873046875, "step": 2223 }, { "epoch": 0.8207020619032244, "grad_norm": 1.3616807460784912, "learning_rate": 0.0001, "loss": 2.781620979309082, "step": 2224 }, { "epoch": 0.8210710826145118, "grad_norm": 1.4259276390075684, "learning_rate": 0.0001, "loss": 2.881051540374756, "step": 2225 }, { "epoch": 0.8214401033257992, "grad_norm": 1.66672945022583, "learning_rate": 0.0001, "loss": 2.9165854454040527, "step": 2226 }, { "epoch": 0.8218091240370866, "grad_norm": 1.46174955368042, "learning_rate": 0.0001, "loss": 2.743654727935791, "step": 2227 }, { "epoch": 0.822178144748374, "grad_norm": 1.674697756767273, "learning_rate": 0.0001, "loss": 2.7341601848602295, "step": 2228 }, { "epoch": 0.8225471654596614, "grad_norm": 1.8845542669296265, "learning_rate": 0.0001, "loss": 2.842817783355713, "step": 2229 }, { "epoch": 0.8229161861709489, "grad_norm": 1.63115394115448, "learning_rate": 0.0001, "loss": 3.0655999183654785, "step": 2230 }, { "epoch": 0.8232852068822363, "grad_norm": 1.4227349758148193, "learning_rate": 0.0001, "loss": 3.2023367881774902, "step": 2231 }, { "epoch": 0.8236542275935237, "grad_norm": 1.7390799522399902, "learning_rate": 0.0001, "loss": 2.977618932723999, "step": 2232 }, { "epoch": 0.8240232483048111, "grad_norm": 1.3881332874298096, "learning_rate": 0.0001, "loss": 2.9100148677825928, "step": 2233 }, { "epoch": 0.8243922690160985, "grad_norm": 1.2452030181884766, "learning_rate": 0.0001, "loss": 2.826486587524414, "step": 2234 }, { "epoch": 0.8247612897273859, "grad_norm": 1.3138855695724487, "learning_rate": 0.0001, "loss": 2.701992988586426, "step": 2235 }, { "epoch": 0.8251303104386734, "grad_norm": 1.3830678462982178, "learning_rate": 0.0001, "loss": 2.7444138526916504, "step": 2236 }, { "epoch": 0.8254993311499608, "grad_norm": 1.3492058515548706, "learning_rate": 0.0001, "loss": 2.808466672897339, "step": 2237 }, { "epoch": 0.8258683518612482, "grad_norm": 1.4770601987838745, "learning_rate": 0.0001, "loss": 2.7810497283935547, "step": 2238 }, { "epoch": 0.8262373725725356, "grad_norm": 1.511266827583313, "learning_rate": 0.0001, "loss": 2.5660064220428467, "step": 2239 }, { "epoch": 0.826606393283823, "grad_norm": 1.1650384664535522, "learning_rate": 0.0001, "loss": 2.8325390815734863, "step": 2240 }, { "epoch": 0.8269754139951104, "grad_norm": 1.231383204460144, "learning_rate": 0.0001, "loss": 2.784801483154297, "step": 2241 }, { "epoch": 0.8273444347063978, "grad_norm": 1.4804036617279053, "learning_rate": 0.0001, "loss": 2.9244391918182373, "step": 2242 }, { "epoch": 0.8277134554176853, "grad_norm": 1.1573939323425293, "learning_rate": 0.0001, "loss": 2.5701591968536377, "step": 2243 }, { "epoch": 0.8280824761289728, "grad_norm": 1.5606032609939575, "learning_rate": 0.0001, "loss": 2.8142495155334473, "step": 2244 }, { "epoch": 0.8284514968402602, "grad_norm": 1.3684754371643066, "learning_rate": 0.0001, "loss": 2.504342555999756, "step": 2245 }, { "epoch": 0.8288205175515476, "grad_norm": 1.4098914861679077, "learning_rate": 0.0001, "loss": 2.729292392730713, "step": 2246 }, { "epoch": 0.829189538262835, "grad_norm": 1.3850326538085938, "learning_rate": 0.0001, "loss": 2.7550501823425293, "step": 2247 }, { "epoch": 0.8295585589741225, "grad_norm": 1.6483633518218994, "learning_rate": 0.0001, "loss": 2.9975099563598633, "step": 2248 }, { "epoch": 0.8299275796854099, "grad_norm": 1.2655036449432373, "learning_rate": 0.0001, "loss": 2.6745851039886475, "step": 2249 }, { "epoch": 0.8302966003966973, "grad_norm": 3.4011499881744385, "learning_rate": 0.0001, "loss": 2.7984373569488525, "step": 2250 }, { "epoch": 0.8306656211079847, "grad_norm": 1.402586817741394, "learning_rate": 0.0001, "loss": 2.730355978012085, "step": 2251 }, { "epoch": 0.8310346418192721, "grad_norm": 1.4423519372940063, "learning_rate": 0.0001, "loss": 3.0755062103271484, "step": 2252 }, { "epoch": 0.8314036625305595, "grad_norm": 1.2773081064224243, "learning_rate": 0.0001, "loss": 2.6395621299743652, "step": 2253 }, { "epoch": 0.8317726832418469, "grad_norm": 1.1613749265670776, "learning_rate": 0.0001, "loss": 2.293677806854248, "step": 2254 }, { "epoch": 0.8321417039531344, "grad_norm": 1.2810988426208496, "learning_rate": 0.0001, "loss": 2.626866102218628, "step": 2255 }, { "epoch": 0.8325107246644218, "grad_norm": 1.659043550491333, "learning_rate": 0.0001, "loss": 2.8559231758117676, "step": 2256 }, { "epoch": 0.8328797453757092, "grad_norm": 1.3194257020950317, "learning_rate": 0.0001, "loss": 2.678562641143799, "step": 2257 }, { "epoch": 0.8332487660869966, "grad_norm": 1.2367075681686401, "learning_rate": 0.0001, "loss": 2.7288808822631836, "step": 2258 }, { "epoch": 0.833617786798284, "grad_norm": 1.4158155918121338, "learning_rate": 0.0001, "loss": 2.7426178455352783, "step": 2259 }, { "epoch": 0.8339868075095714, "grad_norm": 1.3224740028381348, "learning_rate": 0.0001, "loss": 2.6329636573791504, "step": 2260 }, { "epoch": 0.8343558282208589, "grad_norm": 1.3971599340438843, "learning_rate": 0.0001, "loss": 2.648425579071045, "step": 2261 }, { "epoch": 0.8347248489321463, "grad_norm": 1.1990060806274414, "learning_rate": 0.0001, "loss": 2.919184684753418, "step": 2262 }, { "epoch": 0.8350938696434337, "grad_norm": 1.3220272064208984, "learning_rate": 0.0001, "loss": 2.634798526763916, "step": 2263 }, { "epoch": 0.8354628903547212, "grad_norm": 1.5963809490203857, "learning_rate": 0.0001, "loss": 2.790233850479126, "step": 2264 }, { "epoch": 0.8358319110660086, "grad_norm": 1.471367359161377, "learning_rate": 0.0001, "loss": 2.8853023052215576, "step": 2265 }, { "epoch": 0.836200931777296, "grad_norm": 1.3974539041519165, "learning_rate": 0.0001, "loss": 2.7188940048217773, "step": 2266 }, { "epoch": 0.8365699524885835, "grad_norm": 1.6670260429382324, "learning_rate": 0.0001, "loss": 2.9337775707244873, "step": 2267 }, { "epoch": 0.8369389731998709, "grad_norm": 1.4588242769241333, "learning_rate": 0.0001, "loss": 2.508350372314453, "step": 2268 }, { "epoch": 0.8373079939111583, "grad_norm": 1.7735230922698975, "learning_rate": 0.0001, "loss": 3.147216558456421, "step": 2269 }, { "epoch": 0.8376770146224457, "grad_norm": 1.6487252712249756, "learning_rate": 0.0001, "loss": 2.345076084136963, "step": 2270 }, { "epoch": 0.8380460353337331, "grad_norm": 1.6008305549621582, "learning_rate": 0.0001, "loss": 2.619406223297119, "step": 2271 }, { "epoch": 0.8384150560450205, "grad_norm": 1.2398303747177124, "learning_rate": 0.0001, "loss": 3.0412392616271973, "step": 2272 }, { "epoch": 0.838784076756308, "grad_norm": 1.3599016666412354, "learning_rate": 0.0001, "loss": 2.7022361755371094, "step": 2273 }, { "epoch": 0.8391530974675954, "grad_norm": 1.428033471107483, "learning_rate": 0.0001, "loss": 2.538379669189453, "step": 2274 }, { "epoch": 0.8395221181788828, "grad_norm": 1.4936981201171875, "learning_rate": 0.0001, "loss": 2.7943711280822754, "step": 2275 }, { "epoch": 0.8398911388901702, "grad_norm": 1.3967269659042358, "learning_rate": 0.0001, "loss": 2.8803136348724365, "step": 2276 }, { "epoch": 0.8402601596014576, "grad_norm": 1.4555760622024536, "learning_rate": 0.0001, "loss": 2.561947822570801, "step": 2277 }, { "epoch": 0.840629180312745, "grad_norm": 1.9160515069961548, "learning_rate": 0.0001, "loss": 3.152907371520996, "step": 2278 }, { "epoch": 0.8409982010240324, "grad_norm": 1.4534456729888916, "learning_rate": 0.0001, "loss": 2.631615161895752, "step": 2279 }, { "epoch": 0.8413672217353199, "grad_norm": 1.1531647443771362, "learning_rate": 0.0001, "loss": 2.8475680351257324, "step": 2280 }, { "epoch": 0.8417362424466073, "grad_norm": 1.54391348361969, "learning_rate": 0.0001, "loss": 2.812819480895996, "step": 2281 }, { "epoch": 0.8421052631578947, "grad_norm": 1.6039685010910034, "learning_rate": 0.0001, "loss": 2.6106860637664795, "step": 2282 }, { "epoch": 0.8424742838691821, "grad_norm": 1.3731623888015747, "learning_rate": 0.0001, "loss": 2.7029008865356445, "step": 2283 }, { "epoch": 0.8428433045804696, "grad_norm": 1.2391259670257568, "learning_rate": 0.0001, "loss": 3.077003002166748, "step": 2284 }, { "epoch": 0.843212325291757, "grad_norm": 1.4879372119903564, "learning_rate": 0.0001, "loss": 2.9233322143554688, "step": 2285 }, { "epoch": 0.8435813460030445, "grad_norm": 1.4107441902160645, "learning_rate": 0.0001, "loss": 2.7926247119903564, "step": 2286 }, { "epoch": 0.8439503667143319, "grad_norm": 1.1961581707000732, "learning_rate": 0.0001, "loss": 2.502603054046631, "step": 2287 }, { "epoch": 0.8443193874256193, "grad_norm": 1.2598352432250977, "learning_rate": 0.0001, "loss": 2.5493571758270264, "step": 2288 }, { "epoch": 0.8446884081369067, "grad_norm": 1.538006067276001, "learning_rate": 0.0001, "loss": 2.6645119190216064, "step": 2289 }, { "epoch": 0.8450574288481941, "grad_norm": 1.68296217918396, "learning_rate": 0.0001, "loss": 2.634772539138794, "step": 2290 }, { "epoch": 0.8454264495594815, "grad_norm": 1.5997977256774902, "learning_rate": 0.0001, "loss": 2.8038954734802246, "step": 2291 }, { "epoch": 0.845795470270769, "grad_norm": 1.3056029081344604, "learning_rate": 0.0001, "loss": 2.6885428428649902, "step": 2292 }, { "epoch": 0.8461644909820564, "grad_norm": 1.4732667207717896, "learning_rate": 0.0001, "loss": 2.725029230117798, "step": 2293 }, { "epoch": 0.8465335116933438, "grad_norm": 1.1583454608917236, "learning_rate": 0.0001, "loss": 2.604356288909912, "step": 2294 }, { "epoch": 0.8469025324046312, "grad_norm": 1.3216207027435303, "learning_rate": 0.0001, "loss": 2.825326681137085, "step": 2295 }, { "epoch": 0.8472715531159186, "grad_norm": 1.3971141576766968, "learning_rate": 0.0001, "loss": 2.789093017578125, "step": 2296 }, { "epoch": 0.847640573827206, "grad_norm": 1.5605524778366089, "learning_rate": 0.0001, "loss": 2.5045242309570312, "step": 2297 }, { "epoch": 0.8480095945384934, "grad_norm": 1.6102603673934937, "learning_rate": 0.0001, "loss": 2.7107303142547607, "step": 2298 }, { "epoch": 0.8483786152497809, "grad_norm": 1.4063318967819214, "learning_rate": 0.0001, "loss": 2.5428497791290283, "step": 2299 }, { "epoch": 0.8487476359610683, "grad_norm": 1.3178786039352417, "learning_rate": 0.0001, "loss": 2.8552355766296387, "step": 2300 }, { "epoch": 0.8491166566723557, "grad_norm": 1.558769702911377, "learning_rate": 0.0001, "loss": 2.8863184452056885, "step": 2301 }, { "epoch": 0.8494856773836431, "grad_norm": 1.2626335620880127, "learning_rate": 0.0001, "loss": 2.959852457046509, "step": 2302 }, { "epoch": 0.8498546980949305, "grad_norm": 1.2213740348815918, "learning_rate": 0.0001, "loss": 2.9309041500091553, "step": 2303 }, { "epoch": 0.850223718806218, "grad_norm": 1.7077686786651611, "learning_rate": 0.0001, "loss": 2.557135820388794, "step": 2304 }, { "epoch": 0.8505927395175055, "grad_norm": 1.4547852277755737, "learning_rate": 0.0001, "loss": 2.7635810375213623, "step": 2305 }, { "epoch": 0.8509617602287929, "grad_norm": 1.6434897184371948, "learning_rate": 0.0001, "loss": 3.1130521297454834, "step": 2306 }, { "epoch": 0.8513307809400803, "grad_norm": 1.5651819705963135, "learning_rate": 0.0001, "loss": 2.70835018157959, "step": 2307 }, { "epoch": 0.8516998016513677, "grad_norm": 1.4460511207580566, "learning_rate": 0.0001, "loss": 2.854578971862793, "step": 2308 }, { "epoch": 0.8520688223626551, "grad_norm": 1.2943333387374878, "learning_rate": 0.0001, "loss": 2.697915554046631, "step": 2309 }, { "epoch": 0.8524378430739425, "grad_norm": 1.4202172756195068, "learning_rate": 0.0001, "loss": 3.2355120182037354, "step": 2310 }, { "epoch": 0.85280686378523, "grad_norm": 1.503556251525879, "learning_rate": 0.0001, "loss": 2.659749746322632, "step": 2311 }, { "epoch": 0.8531758844965174, "grad_norm": 1.4972689151763916, "learning_rate": 0.0001, "loss": 2.697551965713501, "step": 2312 }, { "epoch": 0.8535449052078048, "grad_norm": 1.409760594367981, "learning_rate": 0.0001, "loss": 2.9576363563537598, "step": 2313 }, { "epoch": 0.8539139259190922, "grad_norm": 1.9595894813537598, "learning_rate": 0.0001, "loss": 2.727663516998291, "step": 2314 }, { "epoch": 0.8542829466303796, "grad_norm": 1.4367308616638184, "learning_rate": 0.0001, "loss": 2.678309440612793, "step": 2315 }, { "epoch": 0.854651967341667, "grad_norm": 1.4826314449310303, "learning_rate": 0.0001, "loss": 2.901698350906372, "step": 2316 }, { "epoch": 0.8550209880529545, "grad_norm": 1.5045896768569946, "learning_rate": 0.0001, "loss": 3.123166561126709, "step": 2317 }, { "epoch": 0.8553900087642419, "grad_norm": 1.2063995599746704, "learning_rate": 0.0001, "loss": 2.5910580158233643, "step": 2318 }, { "epoch": 0.8557590294755293, "grad_norm": 1.7509044408798218, "learning_rate": 0.0001, "loss": 2.7925024032592773, "step": 2319 }, { "epoch": 0.8561280501868167, "grad_norm": 1.4055712223052979, "learning_rate": 0.0001, "loss": 3.0245532989501953, "step": 2320 }, { "epoch": 0.8564970708981041, "grad_norm": 1.3509148359298706, "learning_rate": 0.0001, "loss": 2.706692695617676, "step": 2321 }, { "epoch": 0.8568660916093915, "grad_norm": 1.427757740020752, "learning_rate": 0.0001, "loss": 3.01830792427063, "step": 2322 }, { "epoch": 0.857235112320679, "grad_norm": 1.4167133569717407, "learning_rate": 0.0001, "loss": 2.7517495155334473, "step": 2323 }, { "epoch": 0.8576041330319664, "grad_norm": 1.3209108114242554, "learning_rate": 0.0001, "loss": 2.324253559112549, "step": 2324 }, { "epoch": 0.8579731537432539, "grad_norm": 1.6592308282852173, "learning_rate": 0.0001, "loss": 2.8380730152130127, "step": 2325 }, { "epoch": 0.8583421744545413, "grad_norm": 1.358554720878601, "learning_rate": 0.0001, "loss": 2.6164937019348145, "step": 2326 }, { "epoch": 0.8587111951658287, "grad_norm": 1.5859087705612183, "learning_rate": 0.0001, "loss": 2.9774129390716553, "step": 2327 }, { "epoch": 0.8590802158771161, "grad_norm": 1.3010238409042358, "learning_rate": 0.0001, "loss": 2.5573437213897705, "step": 2328 }, { "epoch": 0.8594492365884036, "grad_norm": 1.4239884614944458, "learning_rate": 0.0001, "loss": 2.4256982803344727, "step": 2329 }, { "epoch": 0.859818257299691, "grad_norm": 1.359010100364685, "learning_rate": 0.0001, "loss": 2.6130564212799072, "step": 2330 }, { "epoch": 0.8601872780109784, "grad_norm": 1.371524453163147, "learning_rate": 0.0001, "loss": 2.9921374320983887, "step": 2331 }, { "epoch": 0.8605562987222658, "grad_norm": 1.3385703563690186, "learning_rate": 0.0001, "loss": 2.928964614868164, "step": 2332 }, { "epoch": 0.8609253194335532, "grad_norm": 1.2442500591278076, "learning_rate": 0.0001, "loss": 2.918653964996338, "step": 2333 }, { "epoch": 0.8612943401448406, "grad_norm": 1.6378728151321411, "learning_rate": 0.0001, "loss": 2.4350223541259766, "step": 2334 }, { "epoch": 0.861663360856128, "grad_norm": 1.2775866985321045, "learning_rate": 0.0001, "loss": 3.1638338565826416, "step": 2335 }, { "epoch": 0.8620323815674155, "grad_norm": 1.6095962524414062, "learning_rate": 0.0001, "loss": 2.9129040241241455, "step": 2336 }, { "epoch": 0.8624014022787029, "grad_norm": 1.3469468355178833, "learning_rate": 0.0001, "loss": 2.6665072441101074, "step": 2337 }, { "epoch": 0.8627704229899903, "grad_norm": 1.8031889200210571, "learning_rate": 0.0001, "loss": 2.673823356628418, "step": 2338 }, { "epoch": 0.8631394437012777, "grad_norm": 1.320042610168457, "learning_rate": 0.0001, "loss": 2.991584062576294, "step": 2339 }, { "epoch": 0.8635084644125651, "grad_norm": 1.289182186126709, "learning_rate": 0.0001, "loss": 2.9379591941833496, "step": 2340 }, { "epoch": 0.8638774851238525, "grad_norm": 1.3116093873977661, "learning_rate": 0.0001, "loss": 2.916102647781372, "step": 2341 }, { "epoch": 0.86424650583514, "grad_norm": 1.3502079248428345, "learning_rate": 0.0001, "loss": 2.7747695446014404, "step": 2342 }, { "epoch": 0.8646155265464274, "grad_norm": 1.2825478315353394, "learning_rate": 0.0001, "loss": 2.467130661010742, "step": 2343 }, { "epoch": 0.8649845472577148, "grad_norm": 1.2602667808532715, "learning_rate": 0.0001, "loss": 2.563481330871582, "step": 2344 }, { "epoch": 0.8653535679690023, "grad_norm": 1.451612949371338, "learning_rate": 0.0001, "loss": 2.6885299682617188, "step": 2345 }, { "epoch": 0.8657225886802897, "grad_norm": 1.2390000820159912, "learning_rate": 0.0001, "loss": 2.6427836418151855, "step": 2346 }, { "epoch": 0.8660916093915771, "grad_norm": 1.306930422782898, "learning_rate": 0.0001, "loss": 2.6665868759155273, "step": 2347 }, { "epoch": 0.8664606301028646, "grad_norm": 1.5203124284744263, "learning_rate": 0.0001, "loss": 2.701965570449829, "step": 2348 }, { "epoch": 0.866829650814152, "grad_norm": 1.5337237119674683, "learning_rate": 0.0001, "loss": 2.4476237297058105, "step": 2349 }, { "epoch": 0.8671986715254394, "grad_norm": 1.4455441236495972, "learning_rate": 0.0001, "loss": 2.7568483352661133, "step": 2350 }, { "epoch": 0.8675676922367268, "grad_norm": 1.5278493165969849, "learning_rate": 0.0001, "loss": 2.8219821453094482, "step": 2351 }, { "epoch": 0.8679367129480142, "grad_norm": 1.35123610496521, "learning_rate": 0.0001, "loss": 2.8713083267211914, "step": 2352 }, { "epoch": 0.8683057336593016, "grad_norm": 1.582343578338623, "learning_rate": 0.0001, "loss": 2.8528897762298584, "step": 2353 }, { "epoch": 0.868674754370589, "grad_norm": 1.3060811758041382, "learning_rate": 0.0001, "loss": 2.704202175140381, "step": 2354 }, { "epoch": 0.8690437750818765, "grad_norm": 1.5036711692810059, "learning_rate": 0.0001, "loss": 2.6231002807617188, "step": 2355 }, { "epoch": 0.8694127957931639, "grad_norm": 1.1934243440628052, "learning_rate": 0.0001, "loss": 2.4409382343292236, "step": 2356 }, { "epoch": 0.8697818165044513, "grad_norm": 1.5227559804916382, "learning_rate": 0.0001, "loss": 2.6822261810302734, "step": 2357 }, { "epoch": 0.8701508372157387, "grad_norm": 1.4023381471633911, "learning_rate": 0.0001, "loss": 2.7352123260498047, "step": 2358 }, { "epoch": 0.8705198579270261, "grad_norm": 1.215009331703186, "learning_rate": 0.0001, "loss": 2.4500010013580322, "step": 2359 }, { "epoch": 0.8708888786383135, "grad_norm": 1.4916762113571167, "learning_rate": 0.0001, "loss": 2.879578113555908, "step": 2360 }, { "epoch": 0.871257899349601, "grad_norm": 1.4884209632873535, "learning_rate": 0.0001, "loss": 2.529651165008545, "step": 2361 }, { "epoch": 0.8716269200608884, "grad_norm": 1.378793478012085, "learning_rate": 0.0001, "loss": 2.608342170715332, "step": 2362 }, { "epoch": 0.8719959407721758, "grad_norm": 1.3659100532531738, "learning_rate": 0.0001, "loss": 2.5141825675964355, "step": 2363 }, { "epoch": 0.8723649614834632, "grad_norm": 1.2427237033843994, "learning_rate": 0.0001, "loss": 2.6538009643554688, "step": 2364 }, { "epoch": 0.8727339821947507, "grad_norm": 1.3455301523208618, "learning_rate": 0.0001, "loss": 2.8389506340026855, "step": 2365 }, { "epoch": 0.8731030029060382, "grad_norm": 1.2972627878189087, "learning_rate": 0.0001, "loss": 2.7801151275634766, "step": 2366 }, { "epoch": 0.8734720236173256, "grad_norm": 1.2031877040863037, "learning_rate": 0.0001, "loss": 2.6963491439819336, "step": 2367 }, { "epoch": 0.873841044328613, "grad_norm": 1.1726205348968506, "learning_rate": 0.0001, "loss": 2.5077221393585205, "step": 2368 }, { "epoch": 0.8742100650399004, "grad_norm": 1.5344678163528442, "learning_rate": 0.0001, "loss": 2.817190647125244, "step": 2369 }, { "epoch": 0.8745790857511878, "grad_norm": 1.2676469087600708, "learning_rate": 0.0001, "loss": 2.676877975463867, "step": 2370 }, { "epoch": 0.8749481064624752, "grad_norm": 1.4374525547027588, "learning_rate": 0.0001, "loss": 2.5830302238464355, "step": 2371 }, { "epoch": 0.8753171271737626, "grad_norm": 1.8908612728118896, "learning_rate": 0.0001, "loss": 2.6553750038146973, "step": 2372 }, { "epoch": 0.8756861478850501, "grad_norm": 1.4351065158843994, "learning_rate": 0.0001, "loss": 2.56832218170166, "step": 2373 }, { "epoch": 0.8760551685963375, "grad_norm": 1.803560733795166, "learning_rate": 0.0001, "loss": 2.863875389099121, "step": 2374 }, { "epoch": 0.8764241893076249, "grad_norm": 1.5371648073196411, "learning_rate": 0.0001, "loss": 2.670315742492676, "step": 2375 }, { "epoch": 0.8767932100189123, "grad_norm": 1.704148292541504, "learning_rate": 0.0001, "loss": 2.7254858016967773, "step": 2376 }, { "epoch": 0.8771622307301997, "grad_norm": 1.5427805185317993, "learning_rate": 0.0001, "loss": 3.049015998840332, "step": 2377 }, { "epoch": 0.8775312514414871, "grad_norm": 1.625888705253601, "learning_rate": 0.0001, "loss": 2.918184757232666, "step": 2378 }, { "epoch": 0.8779002721527746, "grad_norm": 1.4127647876739502, "learning_rate": 0.0001, "loss": 3.011479377746582, "step": 2379 }, { "epoch": 0.878269292864062, "grad_norm": 1.5603810548782349, "learning_rate": 0.0001, "loss": 2.59173846244812, "step": 2380 }, { "epoch": 0.8786383135753494, "grad_norm": 1.61967933177948, "learning_rate": 0.0001, "loss": 2.6535842418670654, "step": 2381 }, { "epoch": 0.8790073342866368, "grad_norm": 1.4951385259628296, "learning_rate": 0.0001, "loss": 2.717947244644165, "step": 2382 }, { "epoch": 0.8793763549979242, "grad_norm": 1.412795066833496, "learning_rate": 0.0001, "loss": 2.7402877807617188, "step": 2383 }, { "epoch": 0.8797453757092116, "grad_norm": 1.607517123222351, "learning_rate": 0.0001, "loss": 2.8839831352233887, "step": 2384 }, { "epoch": 0.880114396420499, "grad_norm": 1.294025182723999, "learning_rate": 0.0001, "loss": 2.569945812225342, "step": 2385 }, { "epoch": 0.8804834171317866, "grad_norm": 1.5431630611419678, "learning_rate": 0.0001, "loss": 2.5957021713256836, "step": 2386 }, { "epoch": 0.880852437843074, "grad_norm": 1.2629653215408325, "learning_rate": 0.0001, "loss": 2.652252435684204, "step": 2387 }, { "epoch": 0.8812214585543614, "grad_norm": 1.3824316263198853, "learning_rate": 0.0001, "loss": 2.5284061431884766, "step": 2388 }, { "epoch": 0.8815904792656488, "grad_norm": 1.2346006631851196, "learning_rate": 0.0001, "loss": 2.512086868286133, "step": 2389 }, { "epoch": 0.8819594999769362, "grad_norm": 1.3740330934524536, "learning_rate": 0.0001, "loss": 2.4854819774627686, "step": 2390 }, { "epoch": 0.8823285206882236, "grad_norm": 1.3192955255508423, "learning_rate": 0.0001, "loss": 2.580254554748535, "step": 2391 }, { "epoch": 0.8826975413995111, "grad_norm": 1.4891587495803833, "learning_rate": 0.0001, "loss": 2.8962719440460205, "step": 2392 }, { "epoch": 0.8830665621107985, "grad_norm": 1.330406904220581, "learning_rate": 0.0001, "loss": 2.510892152786255, "step": 2393 }, { "epoch": 0.8834355828220859, "grad_norm": 1.3440054655075073, "learning_rate": 0.0001, "loss": 2.643061876296997, "step": 2394 }, { "epoch": 0.8838046035333733, "grad_norm": 1.7388200759887695, "learning_rate": 0.0001, "loss": 2.7714335918426514, "step": 2395 }, { "epoch": 0.8841736242446607, "grad_norm": 1.5938962697982788, "learning_rate": 0.0001, "loss": 2.6612141132354736, "step": 2396 }, { "epoch": 0.8845426449559481, "grad_norm": 1.940347671508789, "learning_rate": 0.0001, "loss": 2.5986690521240234, "step": 2397 }, { "epoch": 0.8849116656672356, "grad_norm": 1.3218764066696167, "learning_rate": 0.0001, "loss": 2.9979896545410156, "step": 2398 }, { "epoch": 0.885280686378523, "grad_norm": 1.2469706535339355, "learning_rate": 0.0001, "loss": 2.462606191635132, "step": 2399 }, { "epoch": 0.8856497070898104, "grad_norm": 1.1837810277938843, "learning_rate": 0.0001, "loss": 2.652838945388794, "step": 2400 }, { "epoch": 0.8860187278010978, "grad_norm": 2.0751419067382812, "learning_rate": 0.0001, "loss": 2.4346606731414795, "step": 2401 }, { "epoch": 0.8863877485123852, "grad_norm": 1.4119987487792969, "learning_rate": 0.0001, "loss": 2.690277338027954, "step": 2402 }, { "epoch": 0.8867567692236726, "grad_norm": 1.3241808414459229, "learning_rate": 0.0001, "loss": 2.4303760528564453, "step": 2403 }, { "epoch": 0.88712578993496, "grad_norm": 1.3303861618041992, "learning_rate": 0.0001, "loss": 2.5080244541168213, "step": 2404 }, { "epoch": 0.8874948106462475, "grad_norm": 1.3392122983932495, "learning_rate": 0.0001, "loss": 2.504693031311035, "step": 2405 }, { "epoch": 0.887863831357535, "grad_norm": 1.5703575611114502, "learning_rate": 0.0001, "loss": 2.8170089721679688, "step": 2406 }, { "epoch": 0.8882328520688224, "grad_norm": 1.246101975440979, "learning_rate": 0.0001, "loss": 2.8237814903259277, "step": 2407 }, { "epoch": 0.8886018727801098, "grad_norm": 1.2967649698257446, "learning_rate": 0.0001, "loss": 2.668259620666504, "step": 2408 }, { "epoch": 0.8889708934913972, "grad_norm": 1.4009569883346558, "learning_rate": 0.0001, "loss": 2.4728474617004395, "step": 2409 }, { "epoch": 0.8893399142026847, "grad_norm": 1.2578067779541016, "learning_rate": 0.0001, "loss": 2.669222116470337, "step": 2410 }, { "epoch": 0.8897089349139721, "grad_norm": 1.3995760679244995, "learning_rate": 0.0001, "loss": 2.4130091667175293, "step": 2411 }, { "epoch": 0.8900779556252595, "grad_norm": 1.7663567066192627, "learning_rate": 0.0001, "loss": 2.4469172954559326, "step": 2412 }, { "epoch": 0.8904469763365469, "grad_norm": 1.365085482597351, "learning_rate": 0.0001, "loss": 2.8002352714538574, "step": 2413 }, { "epoch": 0.8908159970478343, "grad_norm": 1.3992156982421875, "learning_rate": 0.0001, "loss": 2.665807008743286, "step": 2414 }, { "epoch": 0.8911850177591217, "grad_norm": 1.3104809522628784, "learning_rate": 0.0001, "loss": 2.4837777614593506, "step": 2415 }, { "epoch": 0.8915540384704091, "grad_norm": 1.2562874555587769, "learning_rate": 0.0001, "loss": 2.433007001876831, "step": 2416 }, { "epoch": 0.8919230591816966, "grad_norm": 2.0020346641540527, "learning_rate": 0.0001, "loss": 2.7686095237731934, "step": 2417 }, { "epoch": 0.892292079892984, "grad_norm": 1.1786705255508423, "learning_rate": 0.0001, "loss": 2.5072364807128906, "step": 2418 }, { "epoch": 0.8926611006042714, "grad_norm": 1.3780120611190796, "learning_rate": 0.0001, "loss": 2.5846095085144043, "step": 2419 }, { "epoch": 0.8930301213155588, "grad_norm": 1.3076527118682861, "learning_rate": 0.0001, "loss": 2.4574780464172363, "step": 2420 }, { "epoch": 0.8933991420268462, "grad_norm": 1.9450056552886963, "learning_rate": 0.0001, "loss": 2.93211030960083, "step": 2421 }, { "epoch": 0.8937681627381336, "grad_norm": 1.4180184602737427, "learning_rate": 0.0001, "loss": 2.6626830101013184, "step": 2422 }, { "epoch": 0.8941371834494211, "grad_norm": 1.6719318628311157, "learning_rate": 0.0001, "loss": 2.640284299850464, "step": 2423 }, { "epoch": 0.8945062041607085, "grad_norm": 1.8954707384109497, "learning_rate": 0.0001, "loss": 2.7420101165771484, "step": 2424 }, { "epoch": 0.8948752248719959, "grad_norm": 1.4159164428710938, "learning_rate": 0.0001, "loss": 2.5943408012390137, "step": 2425 }, { "epoch": 0.8952442455832834, "grad_norm": 1.2097530364990234, "learning_rate": 0.0001, "loss": 2.740488052368164, "step": 2426 }, { "epoch": 0.8956132662945708, "grad_norm": 1.2513972520828247, "learning_rate": 0.0001, "loss": 2.898742914199829, "step": 2427 }, { "epoch": 0.8959822870058582, "grad_norm": 1.4786025285720825, "learning_rate": 0.0001, "loss": 2.6297860145568848, "step": 2428 }, { "epoch": 0.8963513077171457, "grad_norm": 1.7812647819519043, "learning_rate": 0.0001, "loss": 2.6983418464660645, "step": 2429 }, { "epoch": 0.8967203284284331, "grad_norm": 1.3015216588974, "learning_rate": 0.0001, "loss": 2.794309377670288, "step": 2430 }, { "epoch": 0.8970893491397205, "grad_norm": 1.5914461612701416, "learning_rate": 0.0001, "loss": 2.6336257457733154, "step": 2431 }, { "epoch": 0.8974583698510079, "grad_norm": 1.1897107362747192, "learning_rate": 0.0001, "loss": 2.676433801651001, "step": 2432 }, { "epoch": 0.8978273905622953, "grad_norm": 1.5074164867401123, "learning_rate": 0.0001, "loss": 2.7378621101379395, "step": 2433 }, { "epoch": 0.8981964112735827, "grad_norm": 1.3481481075286865, "learning_rate": 0.0001, "loss": 2.5321526527404785, "step": 2434 }, { "epoch": 0.8985654319848702, "grad_norm": 1.394085168838501, "learning_rate": 0.0001, "loss": 2.6363251209259033, "step": 2435 }, { "epoch": 0.8989344526961576, "grad_norm": 1.3486318588256836, "learning_rate": 0.0001, "loss": 2.61332631111145, "step": 2436 }, { "epoch": 0.899303473407445, "grad_norm": 1.7329806089401245, "learning_rate": 0.0001, "loss": 2.9514331817626953, "step": 2437 }, { "epoch": 0.8996724941187324, "grad_norm": 1.361904263496399, "learning_rate": 0.0001, "loss": 2.684504985809326, "step": 2438 }, { "epoch": 0.9000415148300198, "grad_norm": 1.201570749282837, "learning_rate": 0.0001, "loss": 2.5842809677124023, "step": 2439 }, { "epoch": 0.9004105355413072, "grad_norm": 1.2957254648208618, "learning_rate": 0.0001, "loss": 2.6172399520874023, "step": 2440 }, { "epoch": 0.9007795562525946, "grad_norm": 1.312195062637329, "learning_rate": 0.0001, "loss": 2.624749183654785, "step": 2441 }, { "epoch": 0.9011485769638821, "grad_norm": 1.576440453529358, "learning_rate": 0.0001, "loss": 2.7436366081237793, "step": 2442 }, { "epoch": 0.9015175976751695, "grad_norm": 1.2228375673294067, "learning_rate": 0.0001, "loss": 2.6140825748443604, "step": 2443 }, { "epoch": 0.9018866183864569, "grad_norm": 1.2423906326293945, "learning_rate": 0.0001, "loss": 2.7037720680236816, "step": 2444 }, { "epoch": 0.9022556390977443, "grad_norm": 1.451062798500061, "learning_rate": 0.0001, "loss": 2.6453189849853516, "step": 2445 }, { "epoch": 0.9026246598090318, "grad_norm": 1.1727813482284546, "learning_rate": 0.0001, "loss": 2.7270729541778564, "step": 2446 }, { "epoch": 0.9029936805203193, "grad_norm": 2.0279793739318848, "learning_rate": 0.0001, "loss": 2.9855594635009766, "step": 2447 }, { "epoch": 0.9033627012316067, "grad_norm": 1.7079455852508545, "learning_rate": 0.0001, "loss": 2.4721243381500244, "step": 2448 }, { "epoch": 0.9037317219428941, "grad_norm": 1.2624309062957764, "learning_rate": 0.0001, "loss": 2.411201238632202, "step": 2449 }, { "epoch": 0.9041007426541815, "grad_norm": 1.8567713499069214, "learning_rate": 0.0001, "loss": 2.574373245239258, "step": 2450 }, { "epoch": 0.9044697633654689, "grad_norm": 1.325981616973877, "learning_rate": 0.0001, "loss": 2.842276096343994, "step": 2451 }, { "epoch": 0.9048387840767563, "grad_norm": 1.3725684881210327, "learning_rate": 0.0001, "loss": 2.5976645946502686, "step": 2452 }, { "epoch": 0.9052078047880437, "grad_norm": 1.1314144134521484, "learning_rate": 0.0001, "loss": 2.565751552581787, "step": 2453 }, { "epoch": 0.9055768254993312, "grad_norm": 1.339468002319336, "learning_rate": 0.0001, "loss": 2.547506809234619, "step": 2454 }, { "epoch": 0.9059458462106186, "grad_norm": 1.6339679956436157, "learning_rate": 0.0001, "loss": 2.9075393676757812, "step": 2455 }, { "epoch": 0.906314866921906, "grad_norm": 1.1771258115768433, "learning_rate": 0.0001, "loss": 2.824025869369507, "step": 2456 }, { "epoch": 0.9066838876331934, "grad_norm": 1.4108072519302368, "learning_rate": 0.0001, "loss": 2.3580262660980225, "step": 2457 }, { "epoch": 0.9070529083444808, "grad_norm": 1.169510006904602, "learning_rate": 0.0001, "loss": 2.7285428047180176, "step": 2458 }, { "epoch": 0.9074219290557682, "grad_norm": 1.4547724723815918, "learning_rate": 0.0001, "loss": 2.6937992572784424, "step": 2459 }, { "epoch": 0.9077909497670557, "grad_norm": 1.4349584579467773, "learning_rate": 0.0001, "loss": 3.0898327827453613, "step": 2460 }, { "epoch": 0.9081599704783431, "grad_norm": 1.5360443592071533, "learning_rate": 0.0001, "loss": 2.5042014122009277, "step": 2461 }, { "epoch": 0.9085289911896305, "grad_norm": 1.4699320793151855, "learning_rate": 0.0001, "loss": 2.813141107559204, "step": 2462 }, { "epoch": 0.9088980119009179, "grad_norm": 1.471962809562683, "learning_rate": 0.0001, "loss": 2.6840426921844482, "step": 2463 }, { "epoch": 0.9092670326122053, "grad_norm": 1.4424272775650024, "learning_rate": 0.0001, "loss": 3.027632474899292, "step": 2464 }, { "epoch": 0.9096360533234927, "grad_norm": 1.5802130699157715, "learning_rate": 0.0001, "loss": 2.761178493499756, "step": 2465 }, { "epoch": 0.9100050740347801, "grad_norm": 1.3871852159500122, "learning_rate": 0.0001, "loss": 2.7468292713165283, "step": 2466 }, { "epoch": 0.9103740947460677, "grad_norm": 1.2081438302993774, "learning_rate": 0.0001, "loss": 2.4132020473480225, "step": 2467 }, { "epoch": 0.9107431154573551, "grad_norm": 1.3163689374923706, "learning_rate": 0.0001, "loss": 2.6410627365112305, "step": 2468 }, { "epoch": 0.9111121361686425, "grad_norm": 2.4428391456604004, "learning_rate": 0.0001, "loss": 2.978120803833008, "step": 2469 }, { "epoch": 0.9114811568799299, "grad_norm": 1.2692396640777588, "learning_rate": 0.0001, "loss": 2.5074622631073, "step": 2470 }, { "epoch": 0.9118501775912173, "grad_norm": 1.710754632949829, "learning_rate": 0.0001, "loss": 2.925382137298584, "step": 2471 }, { "epoch": 0.9122191983025048, "grad_norm": 1.1821292638778687, "learning_rate": 0.0001, "loss": 2.4906058311462402, "step": 2472 }, { "epoch": 0.9125882190137922, "grad_norm": 2.642704963684082, "learning_rate": 0.0001, "loss": 2.764724016189575, "step": 2473 }, { "epoch": 0.9129572397250796, "grad_norm": 1.2302227020263672, "learning_rate": 0.0001, "loss": 2.8812432289123535, "step": 2474 }, { "epoch": 0.913326260436367, "grad_norm": 1.6255466938018799, "learning_rate": 0.0001, "loss": 2.826483964920044, "step": 2475 }, { "epoch": 0.9136952811476544, "grad_norm": 1.4263309240341187, "learning_rate": 0.0001, "loss": 2.375188112258911, "step": 2476 }, { "epoch": 0.9140643018589418, "grad_norm": 1.5001577138900757, "learning_rate": 0.0001, "loss": 2.5585086345672607, "step": 2477 }, { "epoch": 0.9144333225702292, "grad_norm": 1.506643295288086, "learning_rate": 0.0001, "loss": 2.633411169052124, "step": 2478 }, { "epoch": 0.9148023432815167, "grad_norm": 1.576504111289978, "learning_rate": 0.0001, "loss": 2.6867170333862305, "step": 2479 }, { "epoch": 0.9151713639928041, "grad_norm": 1.5607439279556274, "learning_rate": 0.0001, "loss": 2.623110771179199, "step": 2480 }, { "epoch": 0.9155403847040915, "grad_norm": 1.8771089315414429, "learning_rate": 0.0001, "loss": 2.8508970737457275, "step": 2481 }, { "epoch": 0.9159094054153789, "grad_norm": 1.387973666191101, "learning_rate": 0.0001, "loss": 2.6447036266326904, "step": 2482 }, { "epoch": 0.9162784261266663, "grad_norm": 1.3798539638519287, "learning_rate": 0.0001, "loss": 2.5550031661987305, "step": 2483 }, { "epoch": 0.9166474468379537, "grad_norm": 1.2975748777389526, "learning_rate": 0.0001, "loss": 2.6639719009399414, "step": 2484 }, { "epoch": 0.9170164675492412, "grad_norm": 1.4371088743209839, "learning_rate": 0.0001, "loss": 2.6786041259765625, "step": 2485 }, { "epoch": 0.9173854882605286, "grad_norm": 1.4656912088394165, "learning_rate": 0.0001, "loss": 2.5280263423919678, "step": 2486 }, { "epoch": 0.9177545089718161, "grad_norm": 1.2581617832183838, "learning_rate": 0.0001, "loss": 2.4089736938476562, "step": 2487 }, { "epoch": 0.9181235296831035, "grad_norm": 1.3985966444015503, "learning_rate": 0.0001, "loss": 2.8379788398742676, "step": 2488 }, { "epoch": 0.9184925503943909, "grad_norm": 1.2501823902130127, "learning_rate": 0.0001, "loss": 2.704857587814331, "step": 2489 }, { "epoch": 0.9188615711056783, "grad_norm": 2.003007411956787, "learning_rate": 0.0001, "loss": 2.85351824760437, "step": 2490 }, { "epoch": 0.9192305918169658, "grad_norm": 1.3592469692230225, "learning_rate": 0.0001, "loss": 2.423032283782959, "step": 2491 }, { "epoch": 0.9195996125282532, "grad_norm": 1.232752799987793, "learning_rate": 0.0001, "loss": 2.4551897048950195, "step": 2492 }, { "epoch": 0.9199686332395406, "grad_norm": 1.5268809795379639, "learning_rate": 0.0001, "loss": 2.688857078552246, "step": 2493 }, { "epoch": 0.920337653950828, "grad_norm": 1.874027967453003, "learning_rate": 0.0001, "loss": 2.74985671043396, "step": 2494 }, { "epoch": 0.9207066746621154, "grad_norm": 1.3269768953323364, "learning_rate": 0.0001, "loss": 2.5460526943206787, "step": 2495 }, { "epoch": 0.9210756953734028, "grad_norm": 2.811598062515259, "learning_rate": 0.0001, "loss": 2.660672664642334, "step": 2496 }, { "epoch": 0.9214447160846903, "grad_norm": 1.6315850019454956, "learning_rate": 0.0001, "loss": 3.2187280654907227, "step": 2497 }, { "epoch": 0.9218137367959777, "grad_norm": 1.0618568658828735, "learning_rate": 0.0001, "loss": 2.573913097381592, "step": 2498 }, { "epoch": 0.9221827575072651, "grad_norm": 1.2734028100967407, "learning_rate": 0.0001, "loss": 2.476067304611206, "step": 2499 }, { "epoch": 0.9225517782185525, "grad_norm": 1.3475615978240967, "learning_rate": 0.0001, "loss": 2.84903621673584, "step": 2500 }, { "epoch": 0.9229207989298399, "grad_norm": 1.7092598676681519, "learning_rate": 0.0001, "loss": 2.8901116847991943, "step": 2501 }, { "epoch": 0.9232898196411273, "grad_norm": 1.6815673112869263, "learning_rate": 0.0001, "loss": 2.8481507301330566, "step": 2502 }, { "epoch": 0.9236588403524147, "grad_norm": 1.3260475397109985, "learning_rate": 0.0001, "loss": 2.707507610321045, "step": 2503 }, { "epoch": 0.9240278610637022, "grad_norm": 1.438685417175293, "learning_rate": 0.0001, "loss": 2.385225534439087, "step": 2504 }, { "epoch": 0.9243968817749896, "grad_norm": 1.2652417421340942, "learning_rate": 0.0001, "loss": 2.465432643890381, "step": 2505 }, { "epoch": 0.924765902486277, "grad_norm": 1.7441188097000122, "learning_rate": 0.0001, "loss": 2.506162166595459, "step": 2506 }, { "epoch": 0.9251349231975645, "grad_norm": 1.3495367765426636, "learning_rate": 0.0001, "loss": 2.3380143642425537, "step": 2507 }, { "epoch": 0.9255039439088519, "grad_norm": 1.3084865808486938, "learning_rate": 0.0001, "loss": 2.6852970123291016, "step": 2508 }, { "epoch": 0.9258729646201393, "grad_norm": 1.3796690702438354, "learning_rate": 0.0001, "loss": 2.669874668121338, "step": 2509 }, { "epoch": 0.9262419853314268, "grad_norm": 1.4552221298217773, "learning_rate": 0.0001, "loss": 2.6849544048309326, "step": 2510 }, { "epoch": 0.9266110060427142, "grad_norm": 1.5176674127578735, "learning_rate": 0.0001, "loss": 2.795078754425049, "step": 2511 }, { "epoch": 0.9269800267540016, "grad_norm": 1.457411289215088, "learning_rate": 0.0001, "loss": 2.675358295440674, "step": 2512 }, { "epoch": 0.927349047465289, "grad_norm": 1.3915237188339233, "learning_rate": 0.0001, "loss": 2.6649587154388428, "step": 2513 }, { "epoch": 0.9277180681765764, "grad_norm": 1.2893636226654053, "learning_rate": 0.0001, "loss": 2.7491536140441895, "step": 2514 }, { "epoch": 0.9280870888878638, "grad_norm": 2.254154920578003, "learning_rate": 0.0001, "loss": 3.2393219470977783, "step": 2515 }, { "epoch": 0.9284561095991513, "grad_norm": 1.4001573324203491, "learning_rate": 0.0001, "loss": 2.658931255340576, "step": 2516 }, { "epoch": 0.9288251303104387, "grad_norm": 1.4120103120803833, "learning_rate": 0.0001, "loss": 2.5367226600646973, "step": 2517 }, { "epoch": 0.9291941510217261, "grad_norm": 1.3351088762283325, "learning_rate": 0.0001, "loss": 2.855226516723633, "step": 2518 }, { "epoch": 0.9295631717330135, "grad_norm": 1.2291979789733887, "learning_rate": 0.0001, "loss": 2.7481460571289062, "step": 2519 }, { "epoch": 0.9299321924443009, "grad_norm": 1.8275935649871826, "learning_rate": 0.0001, "loss": 3.0908970832824707, "step": 2520 }, { "epoch": 0.9303012131555883, "grad_norm": 1.2458549737930298, "learning_rate": 0.0001, "loss": 2.364227294921875, "step": 2521 }, { "epoch": 0.9306702338668758, "grad_norm": 1.4719657897949219, "learning_rate": 0.0001, "loss": 2.6455600261688232, "step": 2522 }, { "epoch": 0.9310392545781632, "grad_norm": 1.6506133079528809, "learning_rate": 0.0001, "loss": 2.7923343181610107, "step": 2523 }, { "epoch": 0.9314082752894506, "grad_norm": 1.5294411182403564, "learning_rate": 0.0001, "loss": 2.8154470920562744, "step": 2524 }, { "epoch": 0.931777296000738, "grad_norm": 1.7359119653701782, "learning_rate": 0.0001, "loss": 2.835257053375244, "step": 2525 }, { "epoch": 0.9321463167120254, "grad_norm": 1.31260347366333, "learning_rate": 0.0001, "loss": 2.96112060546875, "step": 2526 }, { "epoch": 0.9325153374233128, "grad_norm": 1.438057541847229, "learning_rate": 0.0001, "loss": 2.755451202392578, "step": 2527 }, { "epoch": 0.9328843581346004, "grad_norm": 1.564408779144287, "learning_rate": 0.0001, "loss": 2.803619384765625, "step": 2528 }, { "epoch": 0.9332533788458878, "grad_norm": 1.6691327095031738, "learning_rate": 0.0001, "loss": 2.5846962928771973, "step": 2529 }, { "epoch": 0.9336223995571752, "grad_norm": 1.3189988136291504, "learning_rate": 0.0001, "loss": 2.7232754230499268, "step": 2530 }, { "epoch": 0.9339914202684626, "grad_norm": 1.3274637460708618, "learning_rate": 0.0001, "loss": 2.453721761703491, "step": 2531 }, { "epoch": 0.93436044097975, "grad_norm": 1.8202317953109741, "learning_rate": 0.0001, "loss": 2.768386125564575, "step": 2532 }, { "epoch": 0.9347294616910374, "grad_norm": 1.4747159481048584, "learning_rate": 0.0001, "loss": 2.630366802215576, "step": 2533 }, { "epoch": 0.9350984824023248, "grad_norm": 1.5519587993621826, "learning_rate": 0.0001, "loss": 2.4446189403533936, "step": 2534 }, { "epoch": 0.9354675031136123, "grad_norm": 1.6022660732269287, "learning_rate": 0.0001, "loss": 2.5388131141662598, "step": 2535 }, { "epoch": 0.9358365238248997, "grad_norm": 2.0541179180145264, "learning_rate": 0.0001, "loss": 2.8741719722747803, "step": 2536 }, { "epoch": 0.9362055445361871, "grad_norm": 1.3456599712371826, "learning_rate": 0.0001, "loss": 2.571108341217041, "step": 2537 }, { "epoch": 0.9365745652474745, "grad_norm": 1.5520727634429932, "learning_rate": 0.0001, "loss": 2.653391122817993, "step": 2538 }, { "epoch": 0.9369435859587619, "grad_norm": 1.222963809967041, "learning_rate": 0.0001, "loss": 2.737966775894165, "step": 2539 }, { "epoch": 0.9373126066700493, "grad_norm": 1.1138067245483398, "learning_rate": 0.0001, "loss": 2.6676456928253174, "step": 2540 }, { "epoch": 0.9376816273813368, "grad_norm": 1.4528073072433472, "learning_rate": 0.0001, "loss": 2.6003029346466064, "step": 2541 }, { "epoch": 0.9380506480926242, "grad_norm": 1.5891335010528564, "learning_rate": 0.0001, "loss": 2.7663474082946777, "step": 2542 }, { "epoch": 0.9384196688039116, "grad_norm": 1.271596074104309, "learning_rate": 0.0001, "loss": 2.6594643592834473, "step": 2543 }, { "epoch": 0.938788689515199, "grad_norm": 1.2937078475952148, "learning_rate": 0.0001, "loss": 2.408064126968384, "step": 2544 }, { "epoch": 0.9391577102264864, "grad_norm": 1.585732102394104, "learning_rate": 0.0001, "loss": 2.8852663040161133, "step": 2545 }, { "epoch": 0.9395267309377738, "grad_norm": 1.2282239198684692, "learning_rate": 0.0001, "loss": 2.50693678855896, "step": 2546 }, { "epoch": 0.9398957516490613, "grad_norm": 1.5908827781677246, "learning_rate": 0.0001, "loss": 2.341029644012451, "step": 2547 }, { "epoch": 0.9402647723603488, "grad_norm": 1.3359893560409546, "learning_rate": 0.0001, "loss": 2.3609538078308105, "step": 2548 }, { "epoch": 0.9406337930716362, "grad_norm": 1.2563999891281128, "learning_rate": 0.0001, "loss": 2.46805739402771, "step": 2549 }, { "epoch": 0.9410028137829236, "grad_norm": 1.5582810640335083, "learning_rate": 0.0001, "loss": 2.864037275314331, "step": 2550 }, { "epoch": 0.941371834494211, "grad_norm": 1.2058285474777222, "learning_rate": 0.0001, "loss": 2.7398297786712646, "step": 2551 }, { "epoch": 0.9417408552054984, "grad_norm": 1.2350754737854004, "learning_rate": 0.0001, "loss": 2.5658228397369385, "step": 2552 }, { "epoch": 0.9421098759167859, "grad_norm": 1.2284866571426392, "learning_rate": 0.0001, "loss": 2.8050780296325684, "step": 2553 }, { "epoch": 0.9424788966280733, "grad_norm": 1.7097162008285522, "learning_rate": 0.0001, "loss": 2.8399977684020996, "step": 2554 }, { "epoch": 0.9428479173393607, "grad_norm": 1.1962225437164307, "learning_rate": 0.0001, "loss": 2.638280153274536, "step": 2555 }, { "epoch": 0.9432169380506481, "grad_norm": 1.5719633102416992, "learning_rate": 0.0001, "loss": 2.7287724018096924, "step": 2556 }, { "epoch": 0.9435859587619355, "grad_norm": 1.3931865692138672, "learning_rate": 0.0001, "loss": 2.6252529621124268, "step": 2557 }, { "epoch": 0.9439549794732229, "grad_norm": 1.4610967636108398, "learning_rate": 0.0001, "loss": 2.790597915649414, "step": 2558 }, { "epoch": 0.9443240001845103, "grad_norm": 1.2248364686965942, "learning_rate": 0.0001, "loss": 2.585798740386963, "step": 2559 }, { "epoch": 0.9446930208957978, "grad_norm": 1.2876465320587158, "learning_rate": 0.0001, "loss": 2.557786703109741, "step": 2560 }, { "epoch": 0.9450620416070852, "grad_norm": 1.2509464025497437, "learning_rate": 0.0001, "loss": 2.6418027877807617, "step": 2561 }, { "epoch": 0.9454310623183726, "grad_norm": 1.3864166736602783, "learning_rate": 0.0001, "loss": 2.7952234745025635, "step": 2562 }, { "epoch": 0.94580008302966, "grad_norm": 1.102947473526001, "learning_rate": 0.0001, "loss": 2.5732226371765137, "step": 2563 }, { "epoch": 0.9461691037409474, "grad_norm": 1.4126957654953003, "learning_rate": 0.0001, "loss": 2.6581919193267822, "step": 2564 }, { "epoch": 0.9465381244522348, "grad_norm": 1.613835334777832, "learning_rate": 0.0001, "loss": 2.3221960067749023, "step": 2565 }, { "epoch": 0.9469071451635223, "grad_norm": 1.4049272537231445, "learning_rate": 0.0001, "loss": 2.433567523956299, "step": 2566 }, { "epoch": 0.9472761658748097, "grad_norm": 1.3000819683074951, "learning_rate": 0.0001, "loss": 2.6969032287597656, "step": 2567 }, { "epoch": 0.9476451865860972, "grad_norm": 1.4271131753921509, "learning_rate": 0.0001, "loss": 2.478956460952759, "step": 2568 }, { "epoch": 0.9480142072973846, "grad_norm": 1.2680401802062988, "learning_rate": 0.0001, "loss": 2.377948760986328, "step": 2569 }, { "epoch": 0.948383228008672, "grad_norm": 1.3832354545593262, "learning_rate": 0.0001, "loss": 2.710078001022339, "step": 2570 }, { "epoch": 0.9487522487199594, "grad_norm": 1.4731502532958984, "learning_rate": 0.0001, "loss": 2.6471197605133057, "step": 2571 }, { "epoch": 0.9491212694312469, "grad_norm": 1.4623678922653198, "learning_rate": 0.0001, "loss": 2.8333146572113037, "step": 2572 }, { "epoch": 0.9494902901425343, "grad_norm": 1.515902042388916, "learning_rate": 0.0001, "loss": 2.558091402053833, "step": 2573 }, { "epoch": 0.9498593108538217, "grad_norm": 1.4958455562591553, "learning_rate": 0.0001, "loss": 2.620718240737915, "step": 2574 }, { "epoch": 0.9502283315651091, "grad_norm": 1.5462480783462524, "learning_rate": 0.0001, "loss": 2.482809543609619, "step": 2575 }, { "epoch": 0.9505973522763965, "grad_norm": 1.3152166604995728, "learning_rate": 0.0001, "loss": 2.5600030422210693, "step": 2576 }, { "epoch": 0.9509663729876839, "grad_norm": 1.5688284635543823, "learning_rate": 0.0001, "loss": 2.6229233741760254, "step": 2577 }, { "epoch": 0.9513353936989714, "grad_norm": 1.4280813932418823, "learning_rate": 0.0001, "loss": 2.6091513633728027, "step": 2578 }, { "epoch": 0.9517044144102588, "grad_norm": 1.2981430292129517, "learning_rate": 0.0001, "loss": 2.730076551437378, "step": 2579 }, { "epoch": 0.9520734351215462, "grad_norm": 1.1287400722503662, "learning_rate": 0.0001, "loss": 2.611839532852173, "step": 2580 }, { "epoch": 0.9524424558328336, "grad_norm": 1.2560960054397583, "learning_rate": 0.0001, "loss": 2.582897663116455, "step": 2581 }, { "epoch": 0.952811476544121, "grad_norm": 1.5811384916305542, "learning_rate": 0.0001, "loss": 2.9143199920654297, "step": 2582 }, { "epoch": 0.9531804972554084, "grad_norm": 1.3284416198730469, "learning_rate": 0.0001, "loss": 2.4791650772094727, "step": 2583 }, { "epoch": 0.9535495179666958, "grad_norm": 1.2874212265014648, "learning_rate": 0.0001, "loss": 2.4607291221618652, "step": 2584 }, { "epoch": 0.9539185386779833, "grad_norm": 1.2926467657089233, "learning_rate": 0.0001, "loss": 2.5478081703186035, "step": 2585 }, { "epoch": 0.9542875593892707, "grad_norm": 1.333652377128601, "learning_rate": 0.0001, "loss": 2.624563217163086, "step": 2586 }, { "epoch": 0.9546565801005581, "grad_norm": 1.337207555770874, "learning_rate": 0.0001, "loss": 2.483367443084717, "step": 2587 }, { "epoch": 0.9550256008118455, "grad_norm": 1.3198896646499634, "learning_rate": 0.0001, "loss": 2.604379177093506, "step": 2588 }, { "epoch": 0.955394621523133, "grad_norm": 1.3050955533981323, "learning_rate": 0.0001, "loss": 2.7053961753845215, "step": 2589 }, { "epoch": 0.9557636422344205, "grad_norm": 1.5816980600357056, "learning_rate": 0.0001, "loss": 2.663323163986206, "step": 2590 }, { "epoch": 0.9561326629457079, "grad_norm": 1.7195419073104858, "learning_rate": 0.0001, "loss": 2.6840455532073975, "step": 2591 }, { "epoch": 0.9565016836569953, "grad_norm": 1.6847760677337646, "learning_rate": 0.0001, "loss": 2.784318447113037, "step": 2592 }, { "epoch": 0.9568707043682827, "grad_norm": 1.4349342584609985, "learning_rate": 0.0001, "loss": 2.925220489501953, "step": 2593 }, { "epoch": 0.9572397250795701, "grad_norm": 3.007449150085449, "learning_rate": 0.0001, "loss": 2.9182851314544678, "step": 2594 }, { "epoch": 0.9576087457908575, "grad_norm": 1.5853283405303955, "learning_rate": 0.0001, "loss": 2.341871976852417, "step": 2595 }, { "epoch": 0.957977766502145, "grad_norm": 1.3766106367111206, "learning_rate": 0.0001, "loss": 2.799224615097046, "step": 2596 }, { "epoch": 0.9583467872134324, "grad_norm": 1.4085817337036133, "learning_rate": 0.0001, "loss": 2.438661575317383, "step": 2597 }, { "epoch": 0.9587158079247198, "grad_norm": 1.5021469593048096, "learning_rate": 0.0001, "loss": 2.7329893112182617, "step": 2598 }, { "epoch": 0.9590848286360072, "grad_norm": 1.5291149616241455, "learning_rate": 0.0001, "loss": 2.8366732597351074, "step": 2599 }, { "epoch": 0.9594538493472946, "grad_norm": 1.4384781122207642, "learning_rate": 0.0001, "loss": 2.5647504329681396, "step": 2600 }, { "epoch": 0.959822870058582, "grad_norm": 1.4433215856552124, "learning_rate": 0.0001, "loss": 2.5448992252349854, "step": 2601 }, { "epoch": 0.9601918907698694, "grad_norm": 1.9469406604766846, "learning_rate": 0.0001, "loss": 2.6081042289733887, "step": 2602 }, { "epoch": 0.9605609114811569, "grad_norm": 1.2604812383651733, "learning_rate": 0.0001, "loss": 2.7244648933410645, "step": 2603 }, { "epoch": 0.9609299321924443, "grad_norm": 1.2127017974853516, "learning_rate": 0.0001, "loss": 2.8785345554351807, "step": 2604 }, { "epoch": 0.9612989529037317, "grad_norm": 1.5300687551498413, "learning_rate": 0.0001, "loss": 2.952627658843994, "step": 2605 }, { "epoch": 0.9616679736150191, "grad_norm": 1.327938437461853, "learning_rate": 0.0001, "loss": 2.5551371574401855, "step": 2606 }, { "epoch": 0.9620369943263065, "grad_norm": 1.3031641244888306, "learning_rate": 0.0001, "loss": 2.3471217155456543, "step": 2607 }, { "epoch": 0.9624060150375939, "grad_norm": 1.578455924987793, "learning_rate": 0.0001, "loss": 3.085646629333496, "step": 2608 }, { "epoch": 0.9627750357488815, "grad_norm": 1.5054237842559814, "learning_rate": 0.0001, "loss": 2.9120266437530518, "step": 2609 }, { "epoch": 0.9631440564601689, "grad_norm": 1.2640669345855713, "learning_rate": 0.0001, "loss": 2.7620890140533447, "step": 2610 }, { "epoch": 0.9635130771714563, "grad_norm": 1.8280121088027954, "learning_rate": 0.0001, "loss": 2.441636085510254, "step": 2611 }, { "epoch": 0.9638820978827437, "grad_norm": 1.364216923713684, "learning_rate": 0.0001, "loss": 2.4871199131011963, "step": 2612 }, { "epoch": 0.9642511185940311, "grad_norm": 1.5485022068023682, "learning_rate": 0.0001, "loss": 2.5649526119232178, "step": 2613 }, { "epoch": 0.9646201393053185, "grad_norm": 1.6655040979385376, "learning_rate": 0.0001, "loss": 2.7898330688476562, "step": 2614 }, { "epoch": 0.964989160016606, "grad_norm": 1.1936808824539185, "learning_rate": 0.0001, "loss": 2.6647937297821045, "step": 2615 }, { "epoch": 0.9653581807278934, "grad_norm": 1.2943378686904907, "learning_rate": 0.0001, "loss": 2.669569253921509, "step": 2616 }, { "epoch": 0.9657272014391808, "grad_norm": 1.1280689239501953, "learning_rate": 0.0001, "loss": 2.4346325397491455, "step": 2617 }, { "epoch": 0.9660962221504682, "grad_norm": 1.4002063274383545, "learning_rate": 0.0001, "loss": 2.4830501079559326, "step": 2618 }, { "epoch": 0.9664652428617556, "grad_norm": 1.401254415512085, "learning_rate": 0.0001, "loss": 2.466660499572754, "step": 2619 }, { "epoch": 0.966834263573043, "grad_norm": 1.3727422952651978, "learning_rate": 0.0001, "loss": 2.7136168479919434, "step": 2620 }, { "epoch": 0.9672032842843304, "grad_norm": 1.2508078813552856, "learning_rate": 0.0001, "loss": 2.5349178314208984, "step": 2621 }, { "epoch": 0.9675723049956179, "grad_norm": 1.376368761062622, "learning_rate": 0.0001, "loss": 2.45225191116333, "step": 2622 }, { "epoch": 0.9679413257069053, "grad_norm": 1.3625285625457764, "learning_rate": 0.0001, "loss": 2.5948729515075684, "step": 2623 }, { "epoch": 0.9683103464181927, "grad_norm": 1.3083049058914185, "learning_rate": 0.0001, "loss": 2.5300045013427734, "step": 2624 }, { "epoch": 0.9686793671294801, "grad_norm": 1.5800106525421143, "learning_rate": 0.0001, "loss": 2.580919027328491, "step": 2625 }, { "epoch": 0.9690483878407675, "grad_norm": 1.3281010389328003, "learning_rate": 0.0001, "loss": 2.724635362625122, "step": 2626 }, { "epoch": 0.9694174085520549, "grad_norm": 1.809131145477295, "learning_rate": 0.0001, "loss": 2.8543689250946045, "step": 2627 }, { "epoch": 0.9697864292633424, "grad_norm": 1.273144006729126, "learning_rate": 0.0001, "loss": 2.1498217582702637, "step": 2628 }, { "epoch": 0.9701554499746299, "grad_norm": 1.2582924365997314, "learning_rate": 0.0001, "loss": 2.69679594039917, "step": 2629 }, { "epoch": 0.9705244706859173, "grad_norm": 1.3601304292678833, "learning_rate": 0.0001, "loss": 2.6128766536712646, "step": 2630 }, { "epoch": 0.9708934913972047, "grad_norm": 1.1981526613235474, "learning_rate": 0.0001, "loss": 2.3533787727355957, "step": 2631 }, { "epoch": 0.9712625121084921, "grad_norm": 1.209030270576477, "learning_rate": 0.0001, "loss": 2.3725507259368896, "step": 2632 }, { "epoch": 0.9716315328197795, "grad_norm": 1.2798311710357666, "learning_rate": 0.0001, "loss": 2.598292827606201, "step": 2633 }, { "epoch": 0.972000553531067, "grad_norm": 1.2821106910705566, "learning_rate": 0.0001, "loss": 2.6698920726776123, "step": 2634 }, { "epoch": 0.9723695742423544, "grad_norm": 1.321468472480774, "learning_rate": 0.0001, "loss": 2.639136791229248, "step": 2635 }, { "epoch": 0.9727385949536418, "grad_norm": 1.4109705686569214, "learning_rate": 0.0001, "loss": 2.7901482582092285, "step": 2636 }, { "epoch": 0.9731076156649292, "grad_norm": 1.6184871196746826, "learning_rate": 0.0001, "loss": 2.704902410507202, "step": 2637 }, { "epoch": 0.9734766363762166, "grad_norm": 1.3064790964126587, "learning_rate": 0.0001, "loss": 2.5556061267852783, "step": 2638 }, { "epoch": 0.973845657087504, "grad_norm": 1.2250337600708008, "learning_rate": 0.0001, "loss": 2.527682304382324, "step": 2639 }, { "epoch": 0.9742146777987915, "grad_norm": 1.5742082595825195, "learning_rate": 0.0001, "loss": 2.5938823223114014, "step": 2640 }, { "epoch": 0.9745836985100789, "grad_norm": 1.4700207710266113, "learning_rate": 0.0001, "loss": 2.9328396320343018, "step": 2641 }, { "epoch": 0.9749527192213663, "grad_norm": 1.206385850906372, "learning_rate": 0.0001, "loss": 2.457498788833618, "step": 2642 }, { "epoch": 0.9753217399326537, "grad_norm": 1.218274712562561, "learning_rate": 0.0001, "loss": 2.618913173675537, "step": 2643 }, { "epoch": 0.9756907606439411, "grad_norm": 1.4273556470870972, "learning_rate": 0.0001, "loss": 2.5836915969848633, "step": 2644 }, { "epoch": 0.9760597813552285, "grad_norm": 1.1255955696105957, "learning_rate": 0.0001, "loss": 2.3979835510253906, "step": 2645 }, { "epoch": 0.9764288020665159, "grad_norm": 1.243171215057373, "learning_rate": 0.0001, "loss": 2.590761184692383, "step": 2646 }, { "epoch": 0.9767978227778034, "grad_norm": 1.669752836227417, "learning_rate": 0.0001, "loss": 2.622347116470337, "step": 2647 }, { "epoch": 0.9771668434890908, "grad_norm": 1.7740674018859863, "learning_rate": 0.0001, "loss": 2.534968614578247, "step": 2648 }, { "epoch": 0.9775358642003783, "grad_norm": 1.2590816020965576, "learning_rate": 0.0001, "loss": 2.522012710571289, "step": 2649 }, { "epoch": 0.9779048849116657, "grad_norm": 1.4653773307800293, "learning_rate": 0.0001, "loss": 2.4802932739257812, "step": 2650 }, { "epoch": 0.9782739056229531, "grad_norm": 1.4919612407684326, "learning_rate": 0.0001, "loss": 2.7345774173736572, "step": 2651 }, { "epoch": 0.9786429263342405, "grad_norm": 1.5237891674041748, "learning_rate": 0.0001, "loss": 2.6211817264556885, "step": 2652 }, { "epoch": 0.979011947045528, "grad_norm": 1.7728335857391357, "learning_rate": 0.0001, "loss": 2.547435760498047, "step": 2653 }, { "epoch": 0.9793809677568154, "grad_norm": 1.393284797668457, "learning_rate": 0.0001, "loss": 2.2930915355682373, "step": 2654 }, { "epoch": 0.9797499884681028, "grad_norm": 1.3578563928604126, "learning_rate": 0.0001, "loss": 2.4874157905578613, "step": 2655 }, { "epoch": 0.9801190091793902, "grad_norm": 1.5419793128967285, "learning_rate": 0.0001, "loss": 2.704944372177124, "step": 2656 }, { "epoch": 0.9804880298906776, "grad_norm": 1.528920292854309, "learning_rate": 0.0001, "loss": 2.270540475845337, "step": 2657 }, { "epoch": 0.980857050601965, "grad_norm": 1.391210675239563, "learning_rate": 0.0001, "loss": 2.76560115814209, "step": 2658 }, { "epoch": 0.9812260713132525, "grad_norm": 1.6337140798568726, "learning_rate": 0.0001, "loss": 2.6468019485473633, "step": 2659 }, { "epoch": 0.9815950920245399, "grad_norm": 1.4911179542541504, "learning_rate": 0.0001, "loss": 2.741353988647461, "step": 2660 }, { "epoch": 0.9819641127358273, "grad_norm": 1.2552834749221802, "learning_rate": 0.0001, "loss": 2.634678602218628, "step": 2661 }, { "epoch": 0.9823331334471147, "grad_norm": 1.142063021659851, "learning_rate": 0.0001, "loss": 2.690155029296875, "step": 2662 }, { "epoch": 0.9827021541584021, "grad_norm": 1.451967477798462, "learning_rate": 0.0001, "loss": 2.7540202140808105, "step": 2663 }, { "epoch": 0.9830711748696895, "grad_norm": 1.3445749282836914, "learning_rate": 0.0001, "loss": 2.6504037380218506, "step": 2664 }, { "epoch": 0.983440195580977, "grad_norm": 1.221899390220642, "learning_rate": 0.0001, "loss": 2.515686273574829, "step": 2665 }, { "epoch": 0.9838092162922644, "grad_norm": 1.217468023300171, "learning_rate": 0.0001, "loss": 2.4935100078582764, "step": 2666 }, { "epoch": 0.9841782370035518, "grad_norm": 1.4214236736297607, "learning_rate": 0.0001, "loss": 2.497438907623291, "step": 2667 }, { "epoch": 0.9845472577148392, "grad_norm": 1.4258464574813843, "learning_rate": 0.0001, "loss": 2.7217655181884766, "step": 2668 }, { "epoch": 0.9849162784261266, "grad_norm": 1.1641238927841187, "learning_rate": 0.0001, "loss": 2.451544761657715, "step": 2669 }, { "epoch": 0.9852852991374141, "grad_norm": 1.2644168138504028, "learning_rate": 0.0001, "loss": 2.411421537399292, "step": 2670 }, { "epoch": 0.9856543198487016, "grad_norm": 1.2025909423828125, "learning_rate": 0.0001, "loss": 2.3580515384674072, "step": 2671 }, { "epoch": 0.986023340559989, "grad_norm": 1.7007675170898438, "learning_rate": 0.0001, "loss": 2.7507693767547607, "step": 2672 }, { "epoch": 0.9863923612712764, "grad_norm": 1.2869600057601929, "learning_rate": 0.0001, "loss": 2.5914969444274902, "step": 2673 }, { "epoch": 0.9867613819825638, "grad_norm": 1.5587964057922363, "learning_rate": 0.0001, "loss": 2.791196584701538, "step": 2674 }, { "epoch": 0.9871304026938512, "grad_norm": 1.203315258026123, "learning_rate": 0.0001, "loss": 2.2909860610961914, "step": 2675 }, { "epoch": 0.9874994234051386, "grad_norm": 1.2265089750289917, "learning_rate": 0.0001, "loss": 2.390126943588257, "step": 2676 }, { "epoch": 0.987868444116426, "grad_norm": 1.2214710712432861, "learning_rate": 0.0001, "loss": 2.715160846710205, "step": 2677 }, { "epoch": 0.9882374648277135, "grad_norm": 1.3411110639572144, "learning_rate": 0.0001, "loss": 2.419788122177124, "step": 2678 }, { "epoch": 0.9886064855390009, "grad_norm": 1.2270934581756592, "learning_rate": 0.0001, "loss": 2.706422805786133, "step": 2679 }, { "epoch": 0.9889755062502883, "grad_norm": 1.7891356945037842, "learning_rate": 0.0001, "loss": 2.3402068614959717, "step": 2680 }, { "epoch": 0.9893445269615757, "grad_norm": 1.5727932453155518, "learning_rate": 0.0001, "loss": 2.7061033248901367, "step": 2681 }, { "epoch": 0.9897135476728631, "grad_norm": 1.317122220993042, "learning_rate": 0.0001, "loss": 2.503577709197998, "step": 2682 }, { "epoch": 0.9900825683841505, "grad_norm": 1.8555914163589478, "learning_rate": 0.0001, "loss": 2.5686724185943604, "step": 2683 }, { "epoch": 0.990451589095438, "grad_norm": 1.3369981050491333, "learning_rate": 0.0001, "loss": 2.5570907592773438, "step": 2684 }, { "epoch": 0.9908206098067254, "grad_norm": 1.1378583908081055, "learning_rate": 0.0001, "loss": 2.360833168029785, "step": 2685 }, { "epoch": 0.9911896305180128, "grad_norm": 1.2291454076766968, "learning_rate": 0.0001, "loss": 2.573361873626709, "step": 2686 }, { "epoch": 0.9915586512293002, "grad_norm": 1.0740392208099365, "learning_rate": 0.0001, "loss": 2.5810983180999756, "step": 2687 }, { "epoch": 0.9919276719405876, "grad_norm": 1.3201264142990112, "learning_rate": 0.0001, "loss": 2.5796360969543457, "step": 2688 }, { "epoch": 0.992296692651875, "grad_norm": 1.3760223388671875, "learning_rate": 0.0001, "loss": 2.4639878273010254, "step": 2689 }, { "epoch": 0.9926657133631626, "grad_norm": 1.4512310028076172, "learning_rate": 0.0001, "loss": 2.798619031906128, "step": 2690 }, { "epoch": 0.99303473407445, "grad_norm": 1.17384934425354, "learning_rate": 0.0001, "loss": 2.059756278991699, "step": 2691 }, { "epoch": 0.9934037547857374, "grad_norm": 1.3536787033081055, "learning_rate": 0.0001, "loss": 2.3727731704711914, "step": 2692 }, { "epoch": 0.9937727754970248, "grad_norm": 1.2965606451034546, "learning_rate": 0.0001, "loss": 2.451420307159424, "step": 2693 }, { "epoch": 0.9941417962083122, "grad_norm": 1.4947526454925537, "learning_rate": 0.0001, "loss": 2.9389445781707764, "step": 2694 }, { "epoch": 0.9945108169195996, "grad_norm": 1.7509695291519165, "learning_rate": 0.0001, "loss": 3.2320077419281006, "step": 2695 }, { "epoch": 0.994879837630887, "grad_norm": 1.6314882040023804, "learning_rate": 0.0001, "loss": 2.8897805213928223, "step": 2696 }, { "epoch": 0.9952488583421745, "grad_norm": 2.0576696395874023, "learning_rate": 0.0001, "loss": 2.5224790573120117, "step": 2697 }, { "epoch": 0.9956178790534619, "grad_norm": 1.3766765594482422, "learning_rate": 0.0001, "loss": 2.9668025970458984, "step": 2698 }, { "epoch": 0.9959868997647493, "grad_norm": 1.392897605895996, "learning_rate": 0.0001, "loss": 2.6544203758239746, "step": 2699 }, { "epoch": 0.9963559204760367, "grad_norm": 1.9287497997283936, "learning_rate": 0.0001, "loss": 2.8124513626098633, "step": 2700 }, { "epoch": 0.9967249411873241, "grad_norm": 1.3017405271530151, "learning_rate": 0.0001, "loss": 2.1588127613067627, "step": 2701 }, { "epoch": 0.9970939618986115, "grad_norm": 1.5595961809158325, "learning_rate": 0.0001, "loss": 2.5245110988616943, "step": 2702 }, { "epoch": 0.997462982609899, "grad_norm": 1.3115499019622803, "learning_rate": 0.0001, "loss": 2.5114569664001465, "step": 2703 }, { "epoch": 0.9978320033211864, "grad_norm": 1.2463160753250122, "learning_rate": 0.0001, "loss": 2.6230061054229736, "step": 2704 }, { "epoch": 0.9982010240324738, "grad_norm": 2.6628470420837402, "learning_rate": 0.0001, "loss": 2.502009153366089, "step": 2705 }, { "epoch": 0.9985700447437612, "grad_norm": 1.2704349756240845, "learning_rate": 0.0001, "loss": 2.5808095932006836, "step": 2706 }, { "epoch": 0.9989390654550486, "grad_norm": 1.265252947807312, "learning_rate": 0.0001, "loss": 2.4881865978240967, "step": 2707 }, { "epoch": 0.999308086166336, "grad_norm": 1.3928241729736328, "learning_rate": 0.0001, "loss": 2.7869443893432617, "step": 2708 }, { "epoch": 0.9996771068776235, "grad_norm": 1.2805821895599365, "learning_rate": 0.0001, "loss": 2.7011942863464355, "step": 2709 }, { "epoch": 1.0, "grad_norm": 1.4393284320831299, "learning_rate": 0.0001, "loss": 2.9947524070739746, "step": 2710 }, { "epoch": 1.0003690207112874, "grad_norm": 1.260804533958435, "learning_rate": 0.0001, "loss": 2.1382710933685303, "step": 2711 }, { "epoch": 1.0007380414225748, "grad_norm": 1.1976549625396729, "learning_rate": 0.0001, "loss": 2.351989269256592, "step": 2712 }, { "epoch": 1.0011070621338622, "grad_norm": 1.1808687448501587, "learning_rate": 9.99999663781431e-05, "loss": 2.3851823806762695, "step": 2713 }, { "epoch": 1.0014760828451497, "grad_norm": 1.1574681997299194, "learning_rate": 9.999986551261765e-05, "loss": 2.3442330360412598, "step": 2714 }, { "epoch": 1.001845103556437, "grad_norm": 1.2709691524505615, "learning_rate": 9.999969740355926e-05, "loss": 2.159605026245117, "step": 2715 }, { "epoch": 1.0022141242677245, "grad_norm": 1.1577616930007935, "learning_rate": 9.999946205119403e-05, "loss": 2.3617849349975586, "step": 2716 }, { "epoch": 1.002583144979012, "grad_norm": 1.1567856073379517, "learning_rate": 9.999915945583851e-05, "loss": 2.5188002586364746, "step": 2717 }, { "epoch": 1.0029521656902993, "grad_norm": 1.0486716032028198, "learning_rate": 9.999878961789962e-05, "loss": 2.3041412830352783, "step": 2718 }, { "epoch": 1.0033211864015867, "grad_norm": 1.1860156059265137, "learning_rate": 9.999835253787473e-05, "loss": 2.452430486679077, "step": 2719 }, { "epoch": 1.0036902071128742, "grad_norm": 1.359322428703308, "learning_rate": 9.99978482163517e-05, "loss": 2.103452682495117, "step": 2720 }, { "epoch": 1.0040592278241616, "grad_norm": 1.160927653312683, "learning_rate": 9.999727665400875e-05, "loss": 2.43700909614563, "step": 2721 }, { "epoch": 1.004428248535449, "grad_norm": 1.2659995555877686, "learning_rate": 9.99966378516146e-05, "loss": 2.4356014728546143, "step": 2722 }, { "epoch": 1.0047972692467364, "grad_norm": 1.1249982118606567, "learning_rate": 9.999593181002829e-05, "loss": 2.1062188148498535, "step": 2723 }, { "epoch": 1.0051662899580238, "grad_norm": 1.4323338270187378, "learning_rate": 9.99951585301994e-05, "loss": 2.6172595024108887, "step": 2724 }, { "epoch": 1.0055353106693112, "grad_norm": 1.3080637454986572, "learning_rate": 9.999431801316791e-05, "loss": 2.2531561851501465, "step": 2725 }, { "epoch": 1.0059043313805986, "grad_norm": 1.06574547290802, "learning_rate": 9.999341026006419e-05, "loss": 2.3127198219299316, "step": 2726 }, { "epoch": 1.006273352091886, "grad_norm": 1.275451898574829, "learning_rate": 9.999243527210905e-05, "loss": 2.303448438644409, "step": 2727 }, { "epoch": 1.0066423728031735, "grad_norm": 1.1730908155441284, "learning_rate": 9.999139305061372e-05, "loss": 2.3254473209381104, "step": 2728 }, { "epoch": 1.007011393514461, "grad_norm": 1.165077805519104, "learning_rate": 9.999028359697987e-05, "loss": 2.1759743690490723, "step": 2729 }, { "epoch": 1.0073804142257483, "grad_norm": 1.1547991037368774, "learning_rate": 9.998910691269955e-05, "loss": 2.311336040496826, "step": 2730 }, { "epoch": 1.007749434937036, "grad_norm": 1.2483445405960083, "learning_rate": 9.99878629993553e-05, "loss": 2.354177951812744, "step": 2731 }, { "epoch": 1.0081184556483234, "grad_norm": 1.2146161794662476, "learning_rate": 9.998655185862e-05, "loss": 2.3205342292785645, "step": 2732 }, { "epoch": 1.0084874763596108, "grad_norm": 1.110552430152893, "learning_rate": 9.998517349225698e-05, "loss": 2.3592474460601807, "step": 2733 }, { "epoch": 1.0088564970708982, "grad_norm": 1.1161013841629028, "learning_rate": 9.998372790211993e-05, "loss": 2.2934389114379883, "step": 2734 }, { "epoch": 1.0092255177821856, "grad_norm": 1.308376431465149, "learning_rate": 9.998221509015304e-05, "loss": 2.2291321754455566, "step": 2735 }, { "epoch": 1.009594538493473, "grad_norm": 0.9436854720115662, "learning_rate": 9.998063505839083e-05, "loss": 2.458786964416504, "step": 2736 }, { "epoch": 1.0099635592047604, "grad_norm": 1.0987927913665771, "learning_rate": 9.997898780895824e-05, "loss": 2.3737964630126953, "step": 2737 }, { "epoch": 1.0103325799160479, "grad_norm": 1.0661914348602295, "learning_rate": 9.997727334407063e-05, "loss": 2.5442564487457275, "step": 2738 }, { "epoch": 1.0107016006273353, "grad_norm": 1.381758213043213, "learning_rate": 9.997549166603371e-05, "loss": 2.262239456176758, "step": 2739 }, { "epoch": 1.0110706213386227, "grad_norm": 1.10279381275177, "learning_rate": 9.997364277724361e-05, "loss": 2.113161325454712, "step": 2740 }, { "epoch": 1.01143964204991, "grad_norm": 1.2122479677200317, "learning_rate": 9.99717266801869e-05, "loss": 2.3161535263061523, "step": 2741 }, { "epoch": 1.0118086627611975, "grad_norm": 1.1132020950317383, "learning_rate": 9.996974337744046e-05, "loss": 2.4076192378997803, "step": 2742 }, { "epoch": 1.012177683472485, "grad_norm": 1.3932456970214844, "learning_rate": 9.996769287167158e-05, "loss": 2.2987520694732666, "step": 2743 }, { "epoch": 1.0125467041837724, "grad_norm": 1.9831929206848145, "learning_rate": 9.996557516563793e-05, "loss": 2.1648147106170654, "step": 2744 }, { "epoch": 1.0129157248950598, "grad_norm": 1.159448504447937, "learning_rate": 9.996339026218759e-05, "loss": 2.0185065269470215, "step": 2745 }, { "epoch": 1.0132847456063472, "grad_norm": 1.2310459613800049, "learning_rate": 9.996113816425894e-05, "loss": 2.1995739936828613, "step": 2746 }, { "epoch": 1.0136537663176346, "grad_norm": 1.185721516609192, "learning_rate": 9.99588188748808e-05, "loss": 2.483020544052124, "step": 2747 }, { "epoch": 1.014022787028922, "grad_norm": 1.0846699476242065, "learning_rate": 9.995643239717227e-05, "loss": 2.3773117065429688, "step": 2748 }, { "epoch": 1.0143918077402094, "grad_norm": 1.4346034526824951, "learning_rate": 9.995397873434292e-05, "loss": 2.2447404861450195, "step": 2749 }, { "epoch": 1.0147608284514968, "grad_norm": 1.3547903299331665, "learning_rate": 9.99514578896926e-05, "loss": 2.2144598960876465, "step": 2750 }, { "epoch": 1.0151298491627843, "grad_norm": 1.2307344675064087, "learning_rate": 9.994886986661153e-05, "loss": 2.2322051525115967, "step": 2751 }, { "epoch": 1.0154988698740717, "grad_norm": 1.23635733127594, "learning_rate": 9.994621466858027e-05, "loss": 2.04384446144104, "step": 2752 }, { "epoch": 1.015867890585359, "grad_norm": 1.0761336088180542, "learning_rate": 9.994349229916973e-05, "loss": 2.088040828704834, "step": 2753 }, { "epoch": 1.0162369112966465, "grad_norm": 1.2707788944244385, "learning_rate": 9.994070276204116e-05, "loss": 2.3106892108917236, "step": 2754 }, { "epoch": 1.016605932007934, "grad_norm": 1.373403549194336, "learning_rate": 9.993784606094612e-05, "loss": 2.4884133338928223, "step": 2755 }, { "epoch": 1.0169749527192213, "grad_norm": 1.1540249586105347, "learning_rate": 9.993492219972653e-05, "loss": 2.306023359298706, "step": 2756 }, { "epoch": 1.0173439734305088, "grad_norm": 1.1255766153335571, "learning_rate": 9.993193118231462e-05, "loss": 2.034877300262451, "step": 2757 }, { "epoch": 1.0177129941417962, "grad_norm": 1.1846975088119507, "learning_rate": 9.992887301273292e-05, "loss": 2.222283363342285, "step": 2758 }, { "epoch": 1.0180820148530836, "grad_norm": 1.1951572895050049, "learning_rate": 9.992574769509428e-05, "loss": 2.6116926670074463, "step": 2759 }, { "epoch": 1.018451035564371, "grad_norm": 1.275842308998108, "learning_rate": 9.992255523360186e-05, "loss": 2.384469509124756, "step": 2760 }, { "epoch": 1.0188200562756584, "grad_norm": 1.2275277376174927, "learning_rate": 9.991929563254914e-05, "loss": 2.201638698577881, "step": 2761 }, { "epoch": 1.0191890769869458, "grad_norm": 1.2368593215942383, "learning_rate": 9.991596889631986e-05, "loss": 2.221550464630127, "step": 2762 }, { "epoch": 1.0195580976982332, "grad_norm": 2.319199562072754, "learning_rate": 9.991257502938804e-05, "loss": 2.2228827476501465, "step": 2763 }, { "epoch": 1.0199271184095207, "grad_norm": 1.3081389665603638, "learning_rate": 9.990911403631805e-05, "loss": 2.6692709922790527, "step": 2764 }, { "epoch": 1.020296139120808, "grad_norm": 1.1656970977783203, "learning_rate": 9.990558592176446e-05, "loss": 2.104414701461792, "step": 2765 }, { "epoch": 1.0206651598320955, "grad_norm": 1.1566987037658691, "learning_rate": 9.990199069047214e-05, "loss": 2.202219009399414, "step": 2766 }, { "epoch": 1.021034180543383, "grad_norm": 1.108751654624939, "learning_rate": 9.989832834727623e-05, "loss": 2.3108620643615723, "step": 2767 }, { "epoch": 1.0214032012546703, "grad_norm": 1.1531026363372803, "learning_rate": 9.989459889710213e-05, "loss": 2.2549588680267334, "step": 2768 }, { "epoch": 1.0217722219659577, "grad_norm": 1.1012475490570068, "learning_rate": 9.989080234496547e-05, "loss": 2.1675665378570557, "step": 2769 }, { "epoch": 1.0221412426772452, "grad_norm": 1.1325420141220093, "learning_rate": 9.988693869597215e-05, "loss": 2.0958943367004395, "step": 2770 }, { "epoch": 1.0225102633885328, "grad_norm": 1.428666353225708, "learning_rate": 9.988300795531827e-05, "loss": 2.277865171432495, "step": 2771 }, { "epoch": 1.0228792840998202, "grad_norm": 1.2388602495193481, "learning_rate": 9.987901012829018e-05, "loss": 2.5396575927734375, "step": 2772 }, { "epoch": 1.0232483048111076, "grad_norm": 1.3019640445709229, "learning_rate": 9.987494522026449e-05, "loss": 2.1728811264038086, "step": 2773 }, { "epoch": 1.023617325522395, "grad_norm": 1.2613600492477417, "learning_rate": 9.987081323670795e-05, "loss": 2.332451343536377, "step": 2774 }, { "epoch": 1.0239863462336825, "grad_norm": 1.2500969171524048, "learning_rate": 9.986661418317759e-05, "loss": 2.2738547325134277, "step": 2775 }, { "epoch": 1.0243553669449699, "grad_norm": 1.1589046716690063, "learning_rate": 9.986234806532058e-05, "loss": 1.9583302736282349, "step": 2776 }, { "epoch": 1.0247243876562573, "grad_norm": 1.2765783071517944, "learning_rate": 9.985801488887435e-05, "loss": 2.297217607498169, "step": 2777 }, { "epoch": 1.0250934083675447, "grad_norm": 1.1659661531448364, "learning_rate": 9.985361465966643e-05, "loss": 2.1539950370788574, "step": 2778 }, { "epoch": 1.0254624290788321, "grad_norm": 1.0953938961029053, "learning_rate": 9.984914738361462e-05, "loss": 2.2637248039245605, "step": 2779 }, { "epoch": 1.0258314497901195, "grad_norm": 1.1690185070037842, "learning_rate": 9.984461306672681e-05, "loss": 2.4182748794555664, "step": 2780 }, { "epoch": 1.026200470501407, "grad_norm": 1.1313570737838745, "learning_rate": 9.984001171510112e-05, "loss": 2.432079792022705, "step": 2781 }, { "epoch": 1.0265694912126944, "grad_norm": 1.3420296907424927, "learning_rate": 9.983534333492575e-05, "loss": 2.226463556289673, "step": 2782 }, { "epoch": 1.0269385119239818, "grad_norm": 1.1634316444396973, "learning_rate": 9.983060793247911e-05, "loss": 2.352780818939209, "step": 2783 }, { "epoch": 1.0273075326352692, "grad_norm": 1.2736443281173706, "learning_rate": 9.982580551412972e-05, "loss": 2.5233426094055176, "step": 2784 }, { "epoch": 1.0276765533465566, "grad_norm": 1.5948350429534912, "learning_rate": 9.982093608633623e-05, "loss": 2.335979700088501, "step": 2785 }, { "epoch": 1.028045574057844, "grad_norm": 1.599293828010559, "learning_rate": 9.98159996556474e-05, "loss": 2.429558277130127, "step": 2786 }, { "epoch": 1.0284145947691314, "grad_norm": 1.1438393592834473, "learning_rate": 9.981099622870211e-05, "loss": 2.436469793319702, "step": 2787 }, { "epoch": 1.0287836154804189, "grad_norm": 1.148962140083313, "learning_rate": 9.980592581222934e-05, "loss": 2.2374677658081055, "step": 2788 }, { "epoch": 1.0291526361917063, "grad_norm": 1.2823269367218018, "learning_rate": 9.980078841304816e-05, "loss": 2.093261957168579, "step": 2789 }, { "epoch": 1.0295216569029937, "grad_norm": 1.343474268913269, "learning_rate": 9.979558403806772e-05, "loss": 2.5532610416412354, "step": 2790 }, { "epoch": 1.029890677614281, "grad_norm": 1.2379419803619385, "learning_rate": 9.97903126942873e-05, "loss": 2.52445912361145, "step": 2791 }, { "epoch": 1.0302596983255685, "grad_norm": 2.0419459342956543, "learning_rate": 9.978497438879612e-05, "loss": 2.385340452194214, "step": 2792 }, { "epoch": 1.030628719036856, "grad_norm": 1.2631980180740356, "learning_rate": 9.977956912877356e-05, "loss": 2.3819124698638916, "step": 2793 }, { "epoch": 1.0309977397481433, "grad_norm": 1.7171778678894043, "learning_rate": 9.977409692148905e-05, "loss": 2.2664334774017334, "step": 2794 }, { "epoch": 1.0313667604594308, "grad_norm": 1.2272237539291382, "learning_rate": 9.976855777430196e-05, "loss": 2.279372215270996, "step": 2795 }, { "epoch": 1.0317357811707182, "grad_norm": 1.187845230102539, "learning_rate": 9.976295169466178e-05, "loss": 2.2345821857452393, "step": 2796 }, { "epoch": 1.0321048018820056, "grad_norm": 1.1646171808242798, "learning_rate": 9.9757278690108e-05, "loss": 2.3630452156066895, "step": 2797 }, { "epoch": 1.032473822593293, "grad_norm": 1.3567818403244019, "learning_rate": 9.975153876827008e-05, "loss": 2.7453482151031494, "step": 2798 }, { "epoch": 1.0328428433045804, "grad_norm": 1.1872470378875732, "learning_rate": 9.974573193686747e-05, "loss": 2.405996799468994, "step": 2799 }, { "epoch": 1.0332118640158678, "grad_norm": 2.28593373298645, "learning_rate": 9.973985820370966e-05, "loss": 2.3356740474700928, "step": 2800 }, { "epoch": 1.0335808847271553, "grad_norm": 1.1443976163864136, "learning_rate": 9.973391757669608e-05, "loss": 2.3502440452575684, "step": 2801 }, { "epoch": 1.0339499054384427, "grad_norm": 1.0933480262756348, "learning_rate": 9.97279100638161e-05, "loss": 2.270904064178467, "step": 2802 }, { "epoch": 1.03431892614973, "grad_norm": 1.426459550857544, "learning_rate": 9.97218356731491e-05, "loss": 2.542168378829956, "step": 2803 }, { "epoch": 1.0346879468610175, "grad_norm": 1.102585792541504, "learning_rate": 9.971569441286434e-05, "loss": 2.3277781009674072, "step": 2804 }, { "epoch": 1.035056967572305, "grad_norm": 1.3864723443984985, "learning_rate": 9.970948629122108e-05, "loss": 2.7291107177734375, "step": 2805 }, { "epoch": 1.0354259882835923, "grad_norm": 1.8358428478240967, "learning_rate": 9.970321131656844e-05, "loss": 2.3083062171936035, "step": 2806 }, { "epoch": 1.0357950089948798, "grad_norm": 1.1997332572937012, "learning_rate": 9.969686949734548e-05, "loss": 2.21836256980896, "step": 2807 }, { "epoch": 1.0361640297061672, "grad_norm": 1.217188835144043, "learning_rate": 9.969046084208115e-05, "loss": 2.3839447498321533, "step": 2808 }, { "epoch": 1.0365330504174546, "grad_norm": 1.7005822658538818, "learning_rate": 9.968398535939427e-05, "loss": 2.502570867538452, "step": 2809 }, { "epoch": 1.036902071128742, "grad_norm": 1.3456088304519653, "learning_rate": 9.967744305799357e-05, "loss": 2.232293128967285, "step": 2810 }, { "epoch": 1.0372710918400294, "grad_norm": 1.4624797105789185, "learning_rate": 9.967083394667762e-05, "loss": 2.626893997192383, "step": 2811 }, { "epoch": 1.0376401125513168, "grad_norm": 1.3881340026855469, "learning_rate": 9.966415803433485e-05, "loss": 2.6140124797821045, "step": 2812 }, { "epoch": 1.0380091332626045, "grad_norm": 1.285071611404419, "learning_rate": 9.965741532994348e-05, "loss": 2.1296956539154053, "step": 2813 }, { "epoch": 1.0383781539738919, "grad_norm": 1.3330273628234863, "learning_rate": 9.965060584257164e-05, "loss": 2.496220588684082, "step": 2814 }, { "epoch": 1.0387471746851793, "grad_norm": 1.4019851684570312, "learning_rate": 9.964372958137724e-05, "loss": 2.222646951675415, "step": 2815 }, { "epoch": 1.0391161953964667, "grad_norm": 1.3267492055892944, "learning_rate": 9.963678655560798e-05, "loss": 2.6748855113983154, "step": 2816 }, { "epoch": 1.0394852161077541, "grad_norm": 1.2434436082839966, "learning_rate": 9.962977677460132e-05, "loss": 2.614943504333496, "step": 2817 }, { "epoch": 1.0398542368190415, "grad_norm": 1.163419485092163, "learning_rate": 9.962270024778458e-05, "loss": 2.1124489307403564, "step": 2818 }, { "epoch": 1.040223257530329, "grad_norm": 1.0768537521362305, "learning_rate": 9.961555698467479e-05, "loss": 2.238521099090576, "step": 2819 }, { "epoch": 1.0405922782416164, "grad_norm": 1.1708650588989258, "learning_rate": 9.960834699487873e-05, "loss": 2.269479274749756, "step": 2820 }, { "epoch": 1.0409612989529038, "grad_norm": 1.1022158861160278, "learning_rate": 9.960107028809293e-05, "loss": 2.4097166061401367, "step": 2821 }, { "epoch": 1.0413303196641912, "grad_norm": 1.529503583908081, "learning_rate": 9.959372687410364e-05, "loss": 2.0979671478271484, "step": 2822 }, { "epoch": 1.0416993403754786, "grad_norm": 1.1701993942260742, "learning_rate": 9.958631676278685e-05, "loss": 2.4261465072631836, "step": 2823 }, { "epoch": 1.042068361086766, "grad_norm": 1.3347166776657104, "learning_rate": 9.957883996410821e-05, "loss": 2.4081056118011475, "step": 2824 }, { "epoch": 1.0424373817980535, "grad_norm": 1.2888588905334473, "learning_rate": 9.957129648812307e-05, "loss": 2.5801122188568115, "step": 2825 }, { "epoch": 1.0428064025093409, "grad_norm": 1.2463241815567017, "learning_rate": 9.956368634497648e-05, "loss": 2.223165512084961, "step": 2826 }, { "epoch": 1.0431754232206283, "grad_norm": 1.1191731691360474, "learning_rate": 9.955600954490312e-05, "loss": 2.290614128112793, "step": 2827 }, { "epoch": 1.0435444439319157, "grad_norm": 1.1626617908477783, "learning_rate": 9.954826609822729e-05, "loss": 2.49528431892395, "step": 2828 }, { "epoch": 1.0439134646432031, "grad_norm": 1.117463231086731, "learning_rate": 9.954045601536299e-05, "loss": 2.4465551376342773, "step": 2829 }, { "epoch": 1.0442824853544905, "grad_norm": 1.1913187503814697, "learning_rate": 9.953257930681379e-05, "loss": 2.3271257877349854, "step": 2830 }, { "epoch": 1.044651506065778, "grad_norm": 1.2009834051132202, "learning_rate": 9.952463598317285e-05, "loss": 2.3904149532318115, "step": 2831 }, { "epoch": 1.0450205267770654, "grad_norm": 2.0127573013305664, "learning_rate": 9.951662605512297e-05, "loss": 2.4734532833099365, "step": 2832 }, { "epoch": 1.0453895474883528, "grad_norm": 1.3031402826309204, "learning_rate": 9.950854953343651e-05, "loss": 2.1786136627197266, "step": 2833 }, { "epoch": 1.0457585681996402, "grad_norm": 1.211361289024353, "learning_rate": 9.95004064289753e-05, "loss": 2.181720018386841, "step": 2834 }, { "epoch": 1.0461275889109276, "grad_norm": 1.1582212448120117, "learning_rate": 9.949219675269088e-05, "loss": 1.9957844018936157, "step": 2835 }, { "epoch": 1.046496609622215, "grad_norm": 1.7559356689453125, "learning_rate": 9.94839205156242e-05, "loss": 2.4555418491363525, "step": 2836 }, { "epoch": 1.0468656303335024, "grad_norm": 1.2164499759674072, "learning_rate": 9.947557772890575e-05, "loss": 2.2830417156219482, "step": 2837 }, { "epoch": 1.0472346510447899, "grad_norm": 1.1951940059661865, "learning_rate": 9.946716840375551e-05, "loss": 2.481210947036743, "step": 2838 }, { "epoch": 1.0476036717560773, "grad_norm": 1.3408992290496826, "learning_rate": 9.945869255148301e-05, "loss": 2.5232207775115967, "step": 2839 }, { "epoch": 1.0479726924673647, "grad_norm": 1.120118498802185, "learning_rate": 9.945015018348718e-05, "loss": 2.391134023666382, "step": 2840 }, { "epoch": 1.048341713178652, "grad_norm": 1.1508270502090454, "learning_rate": 9.944154131125642e-05, "loss": 2.217132806777954, "step": 2841 }, { "epoch": 1.0487107338899395, "grad_norm": 1.5447633266448975, "learning_rate": 9.943286594636863e-05, "loss": 2.2050719261169434, "step": 2842 }, { "epoch": 1.049079754601227, "grad_norm": 1.3981425762176514, "learning_rate": 9.942412410049102e-05, "loss": 2.2850115299224854, "step": 2843 }, { "epoch": 1.0494487753125143, "grad_norm": 1.0978604555130005, "learning_rate": 9.941531578538033e-05, "loss": 2.465747117996216, "step": 2844 }, { "epoch": 1.0498177960238018, "grad_norm": 1.263938546180725, "learning_rate": 9.940644101288259e-05, "loss": 2.1393165588378906, "step": 2845 }, { "epoch": 1.0501868167350892, "grad_norm": 1.0699031352996826, "learning_rate": 9.939749979493327e-05, "loss": 2.207099676132202, "step": 2846 }, { "epoch": 1.0505558374463766, "grad_norm": 1.380600094795227, "learning_rate": 9.938849214355721e-05, "loss": 2.2272696495056152, "step": 2847 }, { "epoch": 1.050924858157664, "grad_norm": 1.1381988525390625, "learning_rate": 9.937941807086853e-05, "loss": 2.2315499782562256, "step": 2848 }, { "epoch": 1.0512938788689514, "grad_norm": 1.313951015472412, "learning_rate": 9.937027758907075e-05, "loss": 2.5199499130249023, "step": 2849 }, { "epoch": 1.0516628995802388, "grad_norm": 1.2002795934677124, "learning_rate": 9.936107071045664e-05, "loss": 2.1761269569396973, "step": 2850 }, { "epoch": 1.0520319202915263, "grad_norm": 1.4535422325134277, "learning_rate": 9.935179744740832e-05, "loss": 2.173325300216675, "step": 2851 }, { "epoch": 1.052400941002814, "grad_norm": 1.218552589416504, "learning_rate": 9.934245781239714e-05, "loss": 2.192239761352539, "step": 2852 }, { "epoch": 1.0527699617141013, "grad_norm": 1.410374402999878, "learning_rate": 9.933305181798373e-05, "loss": 2.6459121704101562, "step": 2853 }, { "epoch": 1.0531389824253887, "grad_norm": 1.231536626815796, "learning_rate": 9.932357947681801e-05, "loss": 2.3720955848693848, "step": 2854 }, { "epoch": 1.0535080031366761, "grad_norm": 0.9815578460693359, "learning_rate": 9.931404080163906e-05, "loss": 2.1013834476470947, "step": 2855 }, { "epoch": 1.0538770238479636, "grad_norm": 1.221773624420166, "learning_rate": 9.930443580527519e-05, "loss": 2.356651782989502, "step": 2856 }, { "epoch": 1.054246044559251, "grad_norm": 1.2600326538085938, "learning_rate": 9.929476450064393e-05, "loss": 2.3809597492218018, "step": 2857 }, { "epoch": 1.0546150652705384, "grad_norm": 1.523032307624817, "learning_rate": 9.928502690075197e-05, "loss": 2.123838424682617, "step": 2858 }, { "epoch": 1.0549840859818258, "grad_norm": 1.09584641456604, "learning_rate": 9.927522301869515e-05, "loss": 2.174895763397217, "step": 2859 }, { "epoch": 1.0553531066931132, "grad_norm": 1.1009162664413452, "learning_rate": 9.926535286765844e-05, "loss": 1.9663336277008057, "step": 2860 }, { "epoch": 1.0557221274044006, "grad_norm": 1.68791663646698, "learning_rate": 9.925541646091599e-05, "loss": 2.624060869216919, "step": 2861 }, { "epoch": 1.056091148115688, "grad_norm": 1.3314361572265625, "learning_rate": 9.9245413811831e-05, "loss": 2.4428536891937256, "step": 2862 }, { "epoch": 1.0564601688269755, "grad_norm": 1.1084136962890625, "learning_rate": 9.923534493385575e-05, "loss": 2.229485273361206, "step": 2863 }, { "epoch": 1.0568291895382629, "grad_norm": 1.2766650915145874, "learning_rate": 9.922520984053167e-05, "loss": 2.3315391540527344, "step": 2864 }, { "epoch": 1.0571982102495503, "grad_norm": 1.1910954713821411, "learning_rate": 9.921500854548915e-05, "loss": 2.3932929039001465, "step": 2865 }, { "epoch": 1.0575672309608377, "grad_norm": 1.1520706415176392, "learning_rate": 9.920474106244763e-05, "loss": 2.3195106983184814, "step": 2866 }, { "epoch": 1.0579362516721251, "grad_norm": 1.4115525484085083, "learning_rate": 9.919440740521564e-05, "loss": 2.265500068664551, "step": 2867 }, { "epoch": 1.0583052723834125, "grad_norm": 1.242479681968689, "learning_rate": 9.918400758769063e-05, "loss": 2.0685365200042725, "step": 2868 }, { "epoch": 1.0586742930947, "grad_norm": 0.9896295666694641, "learning_rate": 9.917354162385898e-05, "loss": 2.0572893619537354, "step": 2869 }, { "epoch": 1.0590433138059874, "grad_norm": 1.2983309030532837, "learning_rate": 9.916300952779621e-05, "loss": 2.3481147289276123, "step": 2870 }, { "epoch": 1.0594123345172748, "grad_norm": 1.3420898914337158, "learning_rate": 9.915241131366657e-05, "loss": 2.510519027709961, "step": 2871 }, { "epoch": 1.0597813552285622, "grad_norm": 2.945176839828491, "learning_rate": 9.914174699572338e-05, "loss": 2.2072038650512695, "step": 2872 }, { "epoch": 1.0601503759398496, "grad_norm": 1.4629300832748413, "learning_rate": 9.91310165883088e-05, "loss": 2.0524699687957764, "step": 2873 }, { "epoch": 1.060519396651137, "grad_norm": 1.3528602123260498, "learning_rate": 9.912022010585384e-05, "loss": 2.2761552333831787, "step": 2874 }, { "epoch": 1.0608884173624245, "grad_norm": 1.2798396348953247, "learning_rate": 9.910935756287845e-05, "loss": 2.399897336959839, "step": 2875 }, { "epoch": 1.0612574380737119, "grad_norm": 1.1731154918670654, "learning_rate": 9.909842897399137e-05, "loss": 2.4510602951049805, "step": 2876 }, { "epoch": 1.0616264587849993, "grad_norm": 1.3091624975204468, "learning_rate": 9.90874343538902e-05, "loss": 2.7745285034179688, "step": 2877 }, { "epoch": 1.0619954794962867, "grad_norm": 1.1553939580917358, "learning_rate": 9.90763737173613e-05, "loss": 2.2136003971099854, "step": 2878 }, { "epoch": 1.0623645002075741, "grad_norm": 1.2175849676132202, "learning_rate": 9.906524707927982e-05, "loss": 2.3473761081695557, "step": 2879 }, { "epoch": 1.0627335209188615, "grad_norm": 1.234817385673523, "learning_rate": 9.905405445460972e-05, "loss": 2.3559083938598633, "step": 2880 }, { "epoch": 1.063102541630149, "grad_norm": 1.1417852640151978, "learning_rate": 9.904279585840365e-05, "loss": 2.4436240196228027, "step": 2881 }, { "epoch": 1.0634715623414364, "grad_norm": 1.1473355293273926, "learning_rate": 9.903147130580304e-05, "loss": 2.3275036811828613, "step": 2882 }, { "epoch": 1.0638405830527238, "grad_norm": 1.169655442237854, "learning_rate": 9.902008081203795e-05, "loss": 2.2820918560028076, "step": 2883 }, { "epoch": 1.0642096037640112, "grad_norm": 1.2267875671386719, "learning_rate": 9.900862439242719e-05, "loss": 2.205874443054199, "step": 2884 }, { "epoch": 1.0645786244752986, "grad_norm": 1.2290546894073486, "learning_rate": 9.89971020623782e-05, "loss": 2.4423248767852783, "step": 2885 }, { "epoch": 1.064947645186586, "grad_norm": 1.2750203609466553, "learning_rate": 9.898551383738706e-05, "loss": 2.3432581424713135, "step": 2886 }, { "epoch": 1.0653166658978734, "grad_norm": 1.2556260824203491, "learning_rate": 9.897385973303845e-05, "loss": 2.3541762828826904, "step": 2887 }, { "epoch": 1.0656856866091609, "grad_norm": 0.9851983189582825, "learning_rate": 9.896213976500573e-05, "loss": 2.4584689140319824, "step": 2888 }, { "epoch": 1.0660547073204483, "grad_norm": 1.1892993450164795, "learning_rate": 9.895035394905073e-05, "loss": 2.091233491897583, "step": 2889 }, { "epoch": 1.0664237280317357, "grad_norm": 1.3771134614944458, "learning_rate": 9.893850230102393e-05, "loss": 2.261171340942383, "step": 2890 }, { "epoch": 1.066792748743023, "grad_norm": 1.3230851888656616, "learning_rate": 9.892658483686427e-05, "loss": 2.3624632358551025, "step": 2891 }, { "epoch": 1.0671617694543105, "grad_norm": 1.1912976503372192, "learning_rate": 9.891460157259928e-05, "loss": 2.0878639221191406, "step": 2892 }, { "epoch": 1.067530790165598, "grad_norm": 1.1366764307022095, "learning_rate": 9.890255252434493e-05, "loss": 2.204204559326172, "step": 2893 }, { "epoch": 1.0678998108768856, "grad_norm": 1.0940247774124146, "learning_rate": 9.889043770830566e-05, "loss": 2.3325445652008057, "step": 2894 }, { "epoch": 1.068268831588173, "grad_norm": 1.2353836297988892, "learning_rate": 9.887825714077438e-05, "loss": 2.3457698822021484, "step": 2895 }, { "epoch": 1.0686378522994604, "grad_norm": 1.2287880182266235, "learning_rate": 9.886601083813243e-05, "loss": 2.3093037605285645, "step": 2896 }, { "epoch": 1.0690068730107478, "grad_norm": 1.3586639165878296, "learning_rate": 9.885369881684956e-05, "loss": 2.4313740730285645, "step": 2897 }, { "epoch": 1.0693758937220352, "grad_norm": 1.2711541652679443, "learning_rate": 9.884132109348386e-05, "loss": 2.281364917755127, "step": 2898 }, { "epoch": 1.0697449144333226, "grad_norm": 1.2252073287963867, "learning_rate": 9.882887768468183e-05, "loss": 2.327674388885498, "step": 2899 }, { "epoch": 1.07011393514461, "grad_norm": 1.1747812032699585, "learning_rate": 9.881636860717829e-05, "loss": 2.427927017211914, "step": 2900 }, { "epoch": 1.0704829558558975, "grad_norm": 1.0897465944290161, "learning_rate": 9.880379387779637e-05, "loss": 2.240711212158203, "step": 2901 }, { "epoch": 1.070851976567185, "grad_norm": 1.2367205619812012, "learning_rate": 9.87911535134475e-05, "loss": 2.317873001098633, "step": 2902 }, { "epoch": 1.0712209972784723, "grad_norm": 1.3059885501861572, "learning_rate": 9.877844753113138e-05, "loss": 2.230712413787842, "step": 2903 }, { "epoch": 1.0715900179897597, "grad_norm": 1.1270699501037598, "learning_rate": 9.876567594793598e-05, "loss": 2.2716009616851807, "step": 2904 }, { "epoch": 1.0719590387010471, "grad_norm": 1.9199767112731934, "learning_rate": 9.875283878103744e-05, "loss": 2.3663241863250732, "step": 2905 }, { "epoch": 1.0723280594123346, "grad_norm": 1.1901161670684814, "learning_rate": 9.873993604770017e-05, "loss": 2.2326900959014893, "step": 2906 }, { "epoch": 1.072697080123622, "grad_norm": 1.286529541015625, "learning_rate": 9.87269677652767e-05, "loss": 2.445389747619629, "step": 2907 }, { "epoch": 1.0730661008349094, "grad_norm": 1.196722149848938, "learning_rate": 9.871393395120774e-05, "loss": 2.5209763050079346, "step": 2908 }, { "epoch": 1.0734351215461968, "grad_norm": 1.2475699186325073, "learning_rate": 9.870083462302216e-05, "loss": 2.2954390048980713, "step": 2909 }, { "epoch": 1.0738041422574842, "grad_norm": 1.06890869140625, "learning_rate": 9.868766979833686e-05, "loss": 2.519423007965088, "step": 2910 }, { "epoch": 1.0741731629687716, "grad_norm": 1.1917078495025635, "learning_rate": 9.867443949485691e-05, "loss": 2.073939561843872, "step": 2911 }, { "epoch": 1.074542183680059, "grad_norm": 1.111032485961914, "learning_rate": 9.86611437303754e-05, "loss": 2.2018513679504395, "step": 2912 }, { "epoch": 1.0749112043913465, "grad_norm": 1.158103585243225, "learning_rate": 9.864778252277345e-05, "loss": 2.1112146377563477, "step": 2913 }, { "epoch": 1.0752802251026339, "grad_norm": 1.3009631633758545, "learning_rate": 9.863435589002022e-05, "loss": 2.303400993347168, "step": 2914 }, { "epoch": 1.0756492458139213, "grad_norm": 1.1233041286468506, "learning_rate": 9.862086385017283e-05, "loss": 2.3757100105285645, "step": 2915 }, { "epoch": 1.0760182665252087, "grad_norm": 1.2996214628219604, "learning_rate": 9.860730642137639e-05, "loss": 2.262042999267578, "step": 2916 }, { "epoch": 1.0763872872364961, "grad_norm": 1.3923972845077515, "learning_rate": 9.859368362186392e-05, "loss": 2.3264832496643066, "step": 2917 }, { "epoch": 1.0767563079477835, "grad_norm": 1.3784667253494263, "learning_rate": 9.857999546995638e-05, "loss": 2.382582426071167, "step": 2918 }, { "epoch": 1.077125328659071, "grad_norm": 1.0946407318115234, "learning_rate": 9.856624198406262e-05, "loss": 2.3233706951141357, "step": 2919 }, { "epoch": 1.0774943493703584, "grad_norm": 1.4874176979064941, "learning_rate": 9.855242318267934e-05, "loss": 2.2569785118103027, "step": 2920 }, { "epoch": 1.0778633700816458, "grad_norm": 1.4701104164123535, "learning_rate": 9.85385390843911e-05, "loss": 3.0135810375213623, "step": 2921 }, { "epoch": 1.0782323907929332, "grad_norm": 1.195450782775879, "learning_rate": 9.852458970787026e-05, "loss": 2.3072147369384766, "step": 2922 }, { "epoch": 1.0786014115042206, "grad_norm": 1.1978505849838257, "learning_rate": 9.8510575071877e-05, "loss": 2.3628156185150146, "step": 2923 }, { "epoch": 1.078970432215508, "grad_norm": 1.2323846817016602, "learning_rate": 9.84964951952592e-05, "loss": 2.3640780448913574, "step": 2924 }, { "epoch": 1.0793394529267955, "grad_norm": 1.086475133895874, "learning_rate": 9.848235009695255e-05, "loss": 2.3201472759246826, "step": 2925 }, { "epoch": 1.0797084736380829, "grad_norm": 1.1404470205307007, "learning_rate": 9.846813979598044e-05, "loss": 2.32676100730896, "step": 2926 }, { "epoch": 1.0800774943493703, "grad_norm": 1.2091211080551147, "learning_rate": 9.84538643114539e-05, "loss": 2.418729782104492, "step": 2927 }, { "epoch": 1.0804465150606577, "grad_norm": 1.0717134475708008, "learning_rate": 9.84395236625717e-05, "loss": 2.0758650302886963, "step": 2928 }, { "epoch": 1.0808155357719451, "grad_norm": 1.1892590522766113, "learning_rate": 9.842511786862019e-05, "loss": 2.1057724952697754, "step": 2929 }, { "epoch": 1.0811845564832325, "grad_norm": 1.0923703908920288, "learning_rate": 9.841064694897337e-05, "loss": 2.3451976776123047, "step": 2930 }, { "epoch": 1.08155357719452, "grad_norm": 1.0141979455947876, "learning_rate": 9.839611092309277e-05, "loss": 2.018157958984375, "step": 2931 }, { "epoch": 1.0819225979058074, "grad_norm": 1.583072543144226, "learning_rate": 9.838150981052757e-05, "loss": 2.501969575881958, "step": 2932 }, { "epoch": 1.082291618617095, "grad_norm": 1.200447916984558, "learning_rate": 9.83668436309144e-05, "loss": 2.025731325149536, "step": 2933 }, { "epoch": 1.0826606393283824, "grad_norm": 1.2629356384277344, "learning_rate": 9.835211240397741e-05, "loss": 2.201706886291504, "step": 2934 }, { "epoch": 1.0830296600396698, "grad_norm": 1.1073768138885498, "learning_rate": 9.833731614952829e-05, "loss": 2.3040077686309814, "step": 2935 }, { "epoch": 1.0833986807509572, "grad_norm": 1.1030486822128296, "learning_rate": 9.832245488746611e-05, "loss": 2.0193707942962646, "step": 2936 }, { "epoch": 1.0837677014622447, "grad_norm": 1.2273621559143066, "learning_rate": 9.830752863777741e-05, "loss": 2.1955220699310303, "step": 2937 }, { "epoch": 1.084136722173532, "grad_norm": 1.1562901735305786, "learning_rate": 9.829253742053612e-05, "loss": 2.0824837684631348, "step": 2938 }, { "epoch": 1.0845057428848195, "grad_norm": 1.285893440246582, "learning_rate": 9.827748125590354e-05, "loss": 2.3172109127044678, "step": 2939 }, { "epoch": 1.084874763596107, "grad_norm": 1.180080533027649, "learning_rate": 9.826236016412833e-05, "loss": 2.398594617843628, "step": 2940 }, { "epoch": 1.0852437843073943, "grad_norm": 1.3820878267288208, "learning_rate": 9.824717416554643e-05, "loss": 2.375526189804077, "step": 2941 }, { "epoch": 1.0856128050186817, "grad_norm": 1.4749921560287476, "learning_rate": 9.823192328058113e-05, "loss": 2.3514204025268555, "step": 2942 }, { "epoch": 1.0859818257299692, "grad_norm": 1.0625767707824707, "learning_rate": 9.821660752974293e-05, "loss": 2.2124929428100586, "step": 2943 }, { "epoch": 1.0863508464412566, "grad_norm": 1.3520872592926025, "learning_rate": 9.82012269336296e-05, "loss": 2.206320285797119, "step": 2944 }, { "epoch": 1.086719867152544, "grad_norm": 1.2864856719970703, "learning_rate": 9.818578151292612e-05, "loss": 2.209355354309082, "step": 2945 }, { "epoch": 1.0870888878638314, "grad_norm": 1.070597529411316, "learning_rate": 9.817027128840461e-05, "loss": 2.341501474380493, "step": 2946 }, { "epoch": 1.0874579085751188, "grad_norm": 1.8962551355361938, "learning_rate": 9.815469628092439e-05, "loss": 2.1929850578308105, "step": 2947 }, { "epoch": 1.0878269292864062, "grad_norm": 1.3590067625045776, "learning_rate": 9.813905651143189e-05, "loss": 2.2397091388702393, "step": 2948 }, { "epoch": 1.0881959499976936, "grad_norm": 1.1543856859207153, "learning_rate": 9.812335200096063e-05, "loss": 2.3019895553588867, "step": 2949 }, { "epoch": 1.088564970708981, "grad_norm": 1.0792970657348633, "learning_rate": 9.810758277063119e-05, "loss": 2.0411226749420166, "step": 2950 }, { "epoch": 1.0889339914202685, "grad_norm": 1.2972252368927002, "learning_rate": 9.809174884165122e-05, "loss": 2.098130464553833, "step": 2951 }, { "epoch": 1.089303012131556, "grad_norm": 1.0945857763290405, "learning_rate": 9.807585023531535e-05, "loss": 2.6416871547698975, "step": 2952 }, { "epoch": 1.0896720328428433, "grad_norm": 1.2332446575164795, "learning_rate": 9.805988697300522e-05, "loss": 2.2263262271881104, "step": 2953 }, { "epoch": 1.0900410535541307, "grad_norm": 1.2154067754745483, "learning_rate": 9.80438590761894e-05, "loss": 2.2590208053588867, "step": 2954 }, { "epoch": 1.0904100742654181, "grad_norm": 1.2669202089309692, "learning_rate": 9.80277665664234e-05, "loss": 2.066563606262207, "step": 2955 }, { "epoch": 1.0907790949767056, "grad_norm": 1.2902977466583252, "learning_rate": 9.801160946534961e-05, "loss": 2.2378411293029785, "step": 2956 }, { "epoch": 1.091148115687993, "grad_norm": 1.0957138538360596, "learning_rate": 9.799538779469734e-05, "loss": 2.019322395324707, "step": 2957 }, { "epoch": 1.0915171363992804, "grad_norm": 1.2171298265457153, "learning_rate": 9.797910157628265e-05, "loss": 2.128880023956299, "step": 2958 }, { "epoch": 1.0918861571105678, "grad_norm": 1.2721233367919922, "learning_rate": 9.796275083200847e-05, "loss": 2.8284707069396973, "step": 2959 }, { "epoch": 1.0922551778218552, "grad_norm": 1.5377930402755737, "learning_rate": 9.794633558386453e-05, "loss": 2.1659035682678223, "step": 2960 }, { "epoch": 1.0926241985331426, "grad_norm": 1.6050779819488525, "learning_rate": 9.792985585392722e-05, "loss": 2.2500600814819336, "step": 2961 }, { "epoch": 1.09299321924443, "grad_norm": 1.4523015022277832, "learning_rate": 9.791331166435973e-05, "loss": 2.5038907527923584, "step": 2962 }, { "epoch": 1.0933622399557175, "grad_norm": 1.079393982887268, "learning_rate": 9.789670303741192e-05, "loss": 2.142216205596924, "step": 2963 }, { "epoch": 1.0937312606670049, "grad_norm": 1.147013545036316, "learning_rate": 9.78800299954203e-05, "loss": 2.4091713428497314, "step": 2964 }, { "epoch": 1.0941002813782923, "grad_norm": 1.2202883958816528, "learning_rate": 9.786329256080801e-05, "loss": 2.437920093536377, "step": 2965 }, { "epoch": 1.0944693020895797, "grad_norm": 1.1605546474456787, "learning_rate": 9.784649075608482e-05, "loss": 2.19370436668396, "step": 2966 }, { "epoch": 1.0948383228008671, "grad_norm": 1.1556185483932495, "learning_rate": 9.782962460384701e-05, "loss": 2.3034353256225586, "step": 2967 }, { "epoch": 1.0952073435121545, "grad_norm": 1.2566462755203247, "learning_rate": 9.781269412677748e-05, "loss": 2.313375473022461, "step": 2968 }, { "epoch": 1.095576364223442, "grad_norm": 1.1213223934173584, "learning_rate": 9.779569934764552e-05, "loss": 2.5078439712524414, "step": 2969 }, { "epoch": 1.0959453849347294, "grad_norm": 1.5735787153244019, "learning_rate": 9.777864028930705e-05, "loss": 2.2396323680877686, "step": 2970 }, { "epoch": 1.0963144056460168, "grad_norm": 1.2157330513000488, "learning_rate": 9.77615169747043e-05, "loss": 2.141937255859375, "step": 2971 }, { "epoch": 1.0966834263573042, "grad_norm": 1.3616167306900024, "learning_rate": 9.774432942686601e-05, "loss": 2.358694076538086, "step": 2972 }, { "epoch": 1.0970524470685916, "grad_norm": 1.0903654098510742, "learning_rate": 9.772707766890726e-05, "loss": 2.412677526473999, "step": 2973 }, { "epoch": 1.097421467779879, "grad_norm": 1.178112268447876, "learning_rate": 9.77097617240295e-05, "loss": 2.037992000579834, "step": 2974 }, { "epoch": 1.0977904884911664, "grad_norm": 1.5634654760360718, "learning_rate": 9.769238161552047e-05, "loss": 2.2730789184570312, "step": 2975 }, { "epoch": 1.098159509202454, "grad_norm": 1.1706408262252808, "learning_rate": 9.767493736675429e-05, "loss": 2.154209852218628, "step": 2976 }, { "epoch": 1.0985285299137415, "grad_norm": 1.1626763343811035, "learning_rate": 9.765742900119121e-05, "loss": 2.6350579261779785, "step": 2977 }, { "epoch": 1.098897550625029, "grad_norm": 1.2628931999206543, "learning_rate": 9.763985654237786e-05, "loss": 2.39243221282959, "step": 2978 }, { "epoch": 1.0992665713363163, "grad_norm": 1.4186733961105347, "learning_rate": 9.762222001394692e-05, "loss": 2.4204227924346924, "step": 2979 }, { "epoch": 1.0996355920476037, "grad_norm": 1.2519240379333496, "learning_rate": 9.760451943961733e-05, "loss": 2.5161092281341553, "step": 2980 }, { "epoch": 1.1000046127588912, "grad_norm": 1.2000412940979004, "learning_rate": 9.758675484319413e-05, "loss": 2.2973248958587646, "step": 2981 }, { "epoch": 1.1003736334701786, "grad_norm": 1.4868764877319336, "learning_rate": 9.756892624856848e-05, "loss": 2.2293031215667725, "step": 2982 }, { "epoch": 1.100742654181466, "grad_norm": 1.2230395078659058, "learning_rate": 9.755103367971759e-05, "loss": 2.122448682785034, "step": 2983 }, { "epoch": 1.1011116748927534, "grad_norm": 1.5221542119979858, "learning_rate": 9.753307716070472e-05, "loss": 2.4165773391723633, "step": 2984 }, { "epoch": 1.1014806956040408, "grad_norm": 1.0263882875442505, "learning_rate": 9.751505671567913e-05, "loss": 2.006333351135254, "step": 2985 }, { "epoch": 1.1018497163153282, "grad_norm": 1.019101858139038, "learning_rate": 9.749697236887605e-05, "loss": 2.1152894496917725, "step": 2986 }, { "epoch": 1.1022187370266157, "grad_norm": 1.1862998008728027, "learning_rate": 9.747882414461664e-05, "loss": 2.2457492351531982, "step": 2987 }, { "epoch": 1.102587757737903, "grad_norm": 1.1038936376571655, "learning_rate": 9.746061206730802e-05, "loss": 2.197366237640381, "step": 2988 }, { "epoch": 1.1029567784491905, "grad_norm": 1.2376500368118286, "learning_rate": 9.74423361614431e-05, "loss": 2.2954468727111816, "step": 2989 }, { "epoch": 1.103325799160478, "grad_norm": 1.2686394453048706, "learning_rate": 9.74239964516007e-05, "loss": 2.5462355613708496, "step": 2990 }, { "epoch": 1.1036948198717653, "grad_norm": 1.1976101398468018, "learning_rate": 9.740559296244543e-05, "loss": 2.2806031703948975, "step": 2991 }, { "epoch": 1.1040638405830527, "grad_norm": 1.2570340633392334, "learning_rate": 9.738712571872763e-05, "loss": 2.4594249725341797, "step": 2992 }, { "epoch": 1.1044328612943402, "grad_norm": 1.2559014558792114, "learning_rate": 9.736859474528347e-05, "loss": 2.1627745628356934, "step": 2993 }, { "epoch": 1.1048018820056276, "grad_norm": 1.1326056718826294, "learning_rate": 9.735000006703475e-05, "loss": 2.374478340148926, "step": 2994 }, { "epoch": 1.105170902716915, "grad_norm": 1.1942623853683472, "learning_rate": 9.733134170898899e-05, "loss": 2.7001852989196777, "step": 2995 }, { "epoch": 1.1055399234282024, "grad_norm": 1.2085057497024536, "learning_rate": 9.731261969623932e-05, "loss": 2.068526268005371, "step": 2996 }, { "epoch": 1.1059089441394898, "grad_norm": 1.1206722259521484, "learning_rate": 9.72938340539645e-05, "loss": 2.610382556915283, "step": 2997 }, { "epoch": 1.1062779648507772, "grad_norm": 1.3056039810180664, "learning_rate": 9.727498480742886e-05, "loss": 2.352971315383911, "step": 2998 }, { "epoch": 1.1066469855620646, "grad_norm": 1.2749866247177124, "learning_rate": 9.725607198198227e-05, "loss": 2.281360626220703, "step": 2999 }, { "epoch": 1.107016006273352, "grad_norm": 1.2293897867202759, "learning_rate": 9.723709560306009e-05, "loss": 2.273200750350952, "step": 3000 }, { "epoch": 1.1073850269846395, "grad_norm": 1.2128939628601074, "learning_rate": 9.721805569618317e-05, "loss": 2.145307779312134, "step": 3001 }, { "epoch": 1.107754047695927, "grad_norm": 1.2133738994598389, "learning_rate": 9.719895228695781e-05, "loss": 2.287839651107788, "step": 3002 }, { "epoch": 1.1081230684072143, "grad_norm": 1.032503366470337, "learning_rate": 9.717978540107566e-05, "loss": 2.045119524002075, "step": 3003 }, { "epoch": 1.1084920891185017, "grad_norm": 1.4166245460510254, "learning_rate": 9.716055506431379e-05, "loss": 2.5032119750976562, "step": 3004 }, { "epoch": 1.1088611098297891, "grad_norm": 1.258292317390442, "learning_rate": 9.714126130253458e-05, "loss": 2.2132644653320312, "step": 3005 }, { "epoch": 1.1092301305410766, "grad_norm": 1.1345235109329224, "learning_rate": 9.712190414168572e-05, "loss": 2.202848434448242, "step": 3006 }, { "epoch": 1.109599151252364, "grad_norm": 1.1383850574493408, "learning_rate": 9.710248360780016e-05, "loss": 2.0112719535827637, "step": 3007 }, { "epoch": 1.1099681719636514, "grad_norm": 1.2097489833831787, "learning_rate": 9.708299972699607e-05, "loss": 2.2613463401794434, "step": 3008 }, { "epoch": 1.1103371926749388, "grad_norm": 1.5535197257995605, "learning_rate": 9.706345252547682e-05, "loss": 2.3805952072143555, "step": 3009 }, { "epoch": 1.1107062133862262, "grad_norm": 1.1583658456802368, "learning_rate": 9.704384202953094e-05, "loss": 2.0178611278533936, "step": 3010 }, { "epoch": 1.1110752340975136, "grad_norm": 1.3356847763061523, "learning_rate": 9.702416826553207e-05, "loss": 2.230762243270874, "step": 3011 }, { "epoch": 1.111444254808801, "grad_norm": 1.239138126373291, "learning_rate": 9.700443125993897e-05, "loss": 2.0663180351257324, "step": 3012 }, { "epoch": 1.1118132755200885, "grad_norm": 1.2251429557800293, "learning_rate": 9.698463103929542e-05, "loss": 2.3458542823791504, "step": 3013 }, { "epoch": 1.112182296231376, "grad_norm": 1.1106443405151367, "learning_rate": 9.696476763023023e-05, "loss": 2.317232370376587, "step": 3014 }, { "epoch": 1.1125513169426635, "grad_norm": 1.0846139192581177, "learning_rate": 9.694484105945719e-05, "loss": 2.2734031677246094, "step": 3015 }, { "epoch": 1.112920337653951, "grad_norm": 1.2062019109725952, "learning_rate": 9.692485135377503e-05, "loss": 2.359604597091675, "step": 3016 }, { "epoch": 1.1132893583652383, "grad_norm": 1.3950459957122803, "learning_rate": 9.690479854006738e-05, "loss": 2.1343746185302734, "step": 3017 }, { "epoch": 1.1136583790765258, "grad_norm": 1.2091808319091797, "learning_rate": 9.688468264530277e-05, "loss": 2.437082052230835, "step": 3018 }, { "epoch": 1.1140273997878132, "grad_norm": 1.0521266460418701, "learning_rate": 9.686450369653454e-05, "loss": 2.4125428199768066, "step": 3019 }, { "epoch": 1.1143964204991006, "grad_norm": 1.5490540266036987, "learning_rate": 9.684426172090085e-05, "loss": 2.4108211994171143, "step": 3020 }, { "epoch": 1.114765441210388, "grad_norm": 1.2043360471725464, "learning_rate": 9.682395674562458e-05, "loss": 2.2222847938537598, "step": 3021 }, { "epoch": 1.1151344619216754, "grad_norm": 1.1399905681610107, "learning_rate": 9.680358879801343e-05, "loss": 2.1693062782287598, "step": 3022 }, { "epoch": 1.1155034826329628, "grad_norm": 1.3601492643356323, "learning_rate": 9.678315790545966e-05, "loss": 2.3097658157348633, "step": 3023 }, { "epoch": 1.1158725033442503, "grad_norm": 1.2397693395614624, "learning_rate": 9.676266409544031e-05, "loss": 2.260146141052246, "step": 3024 }, { "epoch": 1.1162415240555377, "grad_norm": 1.2011525630950928, "learning_rate": 9.674210739551692e-05, "loss": 2.255570411682129, "step": 3025 }, { "epoch": 1.116610544766825, "grad_norm": 1.5030021667480469, "learning_rate": 9.672148783333572e-05, "loss": 2.339157819747925, "step": 3026 }, { "epoch": 1.1169795654781125, "grad_norm": 1.2525361776351929, "learning_rate": 9.67008054366274e-05, "loss": 2.2896413803100586, "step": 3027 }, { "epoch": 1.1173485861894, "grad_norm": 1.2128233909606934, "learning_rate": 9.668006023320719e-05, "loss": 2.164114475250244, "step": 3028 }, { "epoch": 1.1177176069006873, "grad_norm": 1.0033804178237915, "learning_rate": 9.665925225097478e-05, "loss": 2.087575912475586, "step": 3029 }, { "epoch": 1.1180866276119747, "grad_norm": 1.0399549007415771, "learning_rate": 9.663838151791431e-05, "loss": 2.3758840560913086, "step": 3030 }, { "epoch": 1.1184556483232622, "grad_norm": 1.0844656229019165, "learning_rate": 9.661744806209425e-05, "loss": 2.2420692443847656, "step": 3031 }, { "epoch": 1.1188246690345496, "grad_norm": 1.3967492580413818, "learning_rate": 9.65964519116675e-05, "loss": 2.1690375804901123, "step": 3032 }, { "epoch": 1.119193689745837, "grad_norm": 1.2111903429031372, "learning_rate": 9.657539309487123e-05, "loss": 2.4257168769836426, "step": 3033 }, { "epoch": 1.1195627104571244, "grad_norm": 1.1692126989364624, "learning_rate": 9.65542716400269e-05, "loss": 2.1197853088378906, "step": 3034 }, { "epoch": 1.1199317311684118, "grad_norm": 1.5416686534881592, "learning_rate": 9.653308757554022e-05, "loss": 2.513756275177002, "step": 3035 }, { "epoch": 1.1203007518796992, "grad_norm": 1.1827932596206665, "learning_rate": 9.651184092990108e-05, "loss": 2.422506093978882, "step": 3036 }, { "epoch": 1.1206697725909867, "grad_norm": 1.0790388584136963, "learning_rate": 9.649053173168357e-05, "loss": 2.278364419937134, "step": 3037 }, { "epoch": 1.121038793302274, "grad_norm": 1.2038443088531494, "learning_rate": 9.646916000954586e-05, "loss": 2.0369935035705566, "step": 3038 }, { "epoch": 1.1214078140135615, "grad_norm": 1.425893783569336, "learning_rate": 9.644772579223022e-05, "loss": 1.9707417488098145, "step": 3039 }, { "epoch": 1.121776834724849, "grad_norm": 1.2457895278930664, "learning_rate": 9.642622910856303e-05, "loss": 2.2378838062286377, "step": 3040 }, { "epoch": 1.1221458554361363, "grad_norm": 1.2137131690979004, "learning_rate": 9.640466998745456e-05, "loss": 2.2986092567443848, "step": 3041 }, { "epoch": 1.1225148761474237, "grad_norm": 1.285129427909851, "learning_rate": 9.638304845789917e-05, "loss": 2.140071153640747, "step": 3042 }, { "epoch": 1.1228838968587112, "grad_norm": 1.4199870824813843, "learning_rate": 9.636136454897504e-05, "loss": 2.355952739715576, "step": 3043 }, { "epoch": 1.1232529175699986, "grad_norm": 1.1842681169509888, "learning_rate": 9.633961828984437e-05, "loss": 2.065748929977417, "step": 3044 }, { "epoch": 1.123621938281286, "grad_norm": 1.4278918504714966, "learning_rate": 9.631780970975311e-05, "loss": 2.344302177429199, "step": 3045 }, { "epoch": 1.1239909589925734, "grad_norm": 1.3654305934906006, "learning_rate": 9.629593883803105e-05, "loss": 2.28863525390625, "step": 3046 }, { "epoch": 1.1243599797038608, "grad_norm": 1.0921337604522705, "learning_rate": 9.627400570409179e-05, "loss": 2.138638496398926, "step": 3047 }, { "epoch": 1.1247290004151482, "grad_norm": 1.304513692855835, "learning_rate": 9.625201033743261e-05, "loss": 2.2170708179473877, "step": 3048 }, { "epoch": 1.1250980211264356, "grad_norm": 1.0935540199279785, "learning_rate": 9.622995276763453e-05, "loss": 2.404672622680664, "step": 3049 }, { "epoch": 1.125467041837723, "grad_norm": 1.1630058288574219, "learning_rate": 9.620783302436219e-05, "loss": 2.2314352989196777, "step": 3050 }, { "epoch": 1.1258360625490105, "grad_norm": 1.2664902210235596, "learning_rate": 9.618565113736389e-05, "loss": 2.2113163471221924, "step": 3051 }, { "epoch": 1.126205083260298, "grad_norm": 1.4400297403335571, "learning_rate": 9.616340713647146e-05, "loss": 2.465211868286133, "step": 3052 }, { "epoch": 1.1265741039715853, "grad_norm": 1.3603750467300415, "learning_rate": 9.61411010516003e-05, "loss": 2.6725120544433594, "step": 3053 }, { "epoch": 1.1269431246828727, "grad_norm": 1.0848586559295654, "learning_rate": 9.611873291274927e-05, "loss": 2.448622941970825, "step": 3054 }, { "epoch": 1.1273121453941601, "grad_norm": 1.1034950017929077, "learning_rate": 9.609630275000072e-05, "loss": 2.2683541774749756, "step": 3055 }, { "epoch": 1.1276811661054476, "grad_norm": 1.1177915334701538, "learning_rate": 9.607381059352038e-05, "loss": 2.5854673385620117, "step": 3056 }, { "epoch": 1.128050186816735, "grad_norm": 1.1264052391052246, "learning_rate": 9.60512564735574e-05, "loss": 2.2541301250457764, "step": 3057 }, { "epoch": 1.1284192075280226, "grad_norm": 1.1862013339996338, "learning_rate": 9.602864042044422e-05, "loss": 2.2935705184936523, "step": 3058 }, { "epoch": 1.12878822823931, "grad_norm": 1.3225853443145752, "learning_rate": 9.600596246459659e-05, "loss": 2.217360496520996, "step": 3059 }, { "epoch": 1.1291572489505974, "grad_norm": 1.2997393608093262, "learning_rate": 9.598322263651352e-05, "loss": 2.1971118450164795, "step": 3060 }, { "epoch": 1.1295262696618849, "grad_norm": 1.1961925029754639, "learning_rate": 9.59604209667772e-05, "loss": 2.330679416656494, "step": 3061 }, { "epoch": 1.1298952903731723, "grad_norm": 1.1777751445770264, "learning_rate": 9.5937557486053e-05, "loss": 2.338290214538574, "step": 3062 }, { "epoch": 1.1302643110844597, "grad_norm": 1.2657216787338257, "learning_rate": 9.591463222508946e-05, "loss": 2.207080125808716, "step": 3063 }, { "epoch": 1.130633331795747, "grad_norm": 1.1900746822357178, "learning_rate": 9.589164521471817e-05, "loss": 2.1008081436157227, "step": 3064 }, { "epoch": 1.1310023525070345, "grad_norm": 1.5295573472976685, "learning_rate": 9.586859648585374e-05, "loss": 2.4141674041748047, "step": 3065 }, { "epoch": 1.131371373218322, "grad_norm": 2.153681993484497, "learning_rate": 9.584548606949383e-05, "loss": 2.4572765827178955, "step": 3066 }, { "epoch": 1.1317403939296093, "grad_norm": 1.1723103523254395, "learning_rate": 9.582231399671907e-05, "loss": 2.2941157817840576, "step": 3067 }, { "epoch": 1.1321094146408968, "grad_norm": 1.1888633966445923, "learning_rate": 9.579908029869293e-05, "loss": 1.9893760681152344, "step": 3068 }, { "epoch": 1.1324784353521842, "grad_norm": 1.3587697744369507, "learning_rate": 9.577578500666187e-05, "loss": 2.2886362075805664, "step": 3069 }, { "epoch": 1.1328474560634716, "grad_norm": 1.111478328704834, "learning_rate": 9.57524281519551e-05, "loss": 2.267953634262085, "step": 3070 }, { "epoch": 1.133216476774759, "grad_norm": 1.1176170110702515, "learning_rate": 9.572900976598463e-05, "loss": 1.994602084159851, "step": 3071 }, { "epoch": 1.1335854974860464, "grad_norm": 1.1192519664764404, "learning_rate": 9.570552988024526e-05, "loss": 2.22080135345459, "step": 3072 }, { "epoch": 1.1339545181973338, "grad_norm": 1.1567951440811157, "learning_rate": 9.568198852631452e-05, "loss": 2.233887195587158, "step": 3073 }, { "epoch": 1.1343235389086213, "grad_norm": 1.26155686378479, "learning_rate": 9.565838573585255e-05, "loss": 2.5856428146362305, "step": 3074 }, { "epoch": 1.1346925596199087, "grad_norm": 1.1886826753616333, "learning_rate": 9.563472154060211e-05, "loss": 2.264721393585205, "step": 3075 }, { "epoch": 1.135061580331196, "grad_norm": 1.6561039686203003, "learning_rate": 9.56109959723886e-05, "loss": 2.587336540222168, "step": 3076 }, { "epoch": 1.1354306010424835, "grad_norm": 1.2891184091567993, "learning_rate": 9.55872090631199e-05, "loss": 2.1480891704559326, "step": 3077 }, { "epoch": 1.135799621753771, "grad_norm": 1.1816635131835938, "learning_rate": 9.556336084478646e-05, "loss": 2.449404239654541, "step": 3078 }, { "epoch": 1.1361686424650583, "grad_norm": 1.0392423868179321, "learning_rate": 9.553945134946107e-05, "loss": 2.3077402114868164, "step": 3079 }, { "epoch": 1.1365376631763457, "grad_norm": 1.2148523330688477, "learning_rate": 9.551548060929904e-05, "loss": 2.1815848350524902, "step": 3080 }, { "epoch": 1.1369066838876332, "grad_norm": 1.126033067703247, "learning_rate": 9.5491448656538e-05, "loss": 2.38411808013916, "step": 3081 }, { "epoch": 1.1372757045989206, "grad_norm": 1.0256751775741577, "learning_rate": 9.546735552349787e-05, "loss": 2.1656641960144043, "step": 3082 }, { "epoch": 1.137644725310208, "grad_norm": 1.1608467102050781, "learning_rate": 9.544320124258092e-05, "loss": 2.3123373985290527, "step": 3083 }, { "epoch": 1.1380137460214954, "grad_norm": 1.2395098209381104, "learning_rate": 9.541898584627163e-05, "loss": 2.207996368408203, "step": 3084 }, { "epoch": 1.1383827667327828, "grad_norm": 1.7574105262756348, "learning_rate": 9.539470936713662e-05, "loss": 2.445319890975952, "step": 3085 }, { "epoch": 1.1387517874440702, "grad_norm": 1.3393627405166626, "learning_rate": 9.537037183782474e-05, "loss": 2.484121084213257, "step": 3086 }, { "epoch": 1.1391208081553577, "grad_norm": 1.192415714263916, "learning_rate": 9.534597329106688e-05, "loss": 1.801460862159729, "step": 3087 }, { "epoch": 1.139489828866645, "grad_norm": 1.0215504169464111, "learning_rate": 9.532151375967604e-05, "loss": 2.1395344734191895, "step": 3088 }, { "epoch": 1.1398588495779325, "grad_norm": 1.1312575340270996, "learning_rate": 9.52969932765472e-05, "loss": 2.1810903549194336, "step": 3089 }, { "epoch": 1.14022787028922, "grad_norm": 1.4220877885818481, "learning_rate": 9.527241187465734e-05, "loss": 2.4093544483184814, "step": 3090 }, { "epoch": 1.1405968910005073, "grad_norm": 1.2816299200057983, "learning_rate": 9.524776958706534e-05, "loss": 2.4554696083068848, "step": 3091 }, { "epoch": 1.1409659117117947, "grad_norm": 1.0947548151016235, "learning_rate": 9.522306644691198e-05, "loss": 2.112539768218994, "step": 3092 }, { "epoch": 1.1413349324230821, "grad_norm": 1.141553282737732, "learning_rate": 9.51983024874199e-05, "loss": 2.49861741065979, "step": 3093 }, { "epoch": 1.1417039531343698, "grad_norm": 1.1734540462493896, "learning_rate": 9.51734777418935e-05, "loss": 2.5351295471191406, "step": 3094 }, { "epoch": 1.1420729738456572, "grad_norm": 1.2637379169464111, "learning_rate": 9.514859224371893e-05, "loss": 1.952693223953247, "step": 3095 }, { "epoch": 1.1424419945569446, "grad_norm": 1.2385724782943726, "learning_rate": 9.512364602636405e-05, "loss": 2.3959014415740967, "step": 3096 }, { "epoch": 1.142811015268232, "grad_norm": 1.2813506126403809, "learning_rate": 9.509863912337842e-05, "loss": 2.479231357574463, "step": 3097 }, { "epoch": 1.1431800359795194, "grad_norm": 1.1192047595977783, "learning_rate": 9.507357156839315e-05, "loss": 2.1717641353607178, "step": 3098 }, { "epoch": 1.1435490566908069, "grad_norm": 1.2042735815048218, "learning_rate": 9.504844339512095e-05, "loss": 2.170778274536133, "step": 3099 }, { "epoch": 1.1439180774020943, "grad_norm": 1.2281179428100586, "learning_rate": 9.502325463735608e-05, "loss": 2.3095180988311768, "step": 3100 }, { "epoch": 1.1442870981133817, "grad_norm": 1.3607943058013916, "learning_rate": 9.499800532897426e-05, "loss": 2.156208038330078, "step": 3101 }, { "epoch": 1.144656118824669, "grad_norm": 1.1007684469223022, "learning_rate": 9.497269550393257e-05, "loss": 2.0497732162475586, "step": 3102 }, { "epoch": 1.1450251395359565, "grad_norm": 1.2566109895706177, "learning_rate": 9.494732519626961e-05, "loss": 1.9819095134735107, "step": 3103 }, { "epoch": 1.145394160247244, "grad_norm": 1.374106526374817, "learning_rate": 9.492189444010521e-05, "loss": 2.2789652347564697, "step": 3104 }, { "epoch": 1.1457631809585314, "grad_norm": 1.0990487337112427, "learning_rate": 9.489640326964058e-05, "loss": 2.018958806991577, "step": 3105 }, { "epoch": 1.1461322016698188, "grad_norm": 1.0797016620635986, "learning_rate": 9.487085171915811e-05, "loss": 2.0048415660858154, "step": 3106 }, { "epoch": 1.1465012223811062, "grad_norm": 1.0526022911071777, "learning_rate": 9.484523982302143e-05, "loss": 2.060316801071167, "step": 3107 }, { "epoch": 1.1468702430923936, "grad_norm": 1.593126893043518, "learning_rate": 9.48195676156753e-05, "loss": 2.384113073348999, "step": 3108 }, { "epoch": 1.147239263803681, "grad_norm": 1.3236359357833862, "learning_rate": 9.479383513164565e-05, "loss": 2.1066067218780518, "step": 3109 }, { "epoch": 1.1476082845149684, "grad_norm": 1.2667080163955688, "learning_rate": 9.476804240553943e-05, "loss": 2.666595458984375, "step": 3110 }, { "epoch": 1.1479773052262559, "grad_norm": 1.1191331148147583, "learning_rate": 9.474218947204459e-05, "loss": 2.2440185546875, "step": 3111 }, { "epoch": 1.1483463259375433, "grad_norm": 1.2326017618179321, "learning_rate": 9.47162763659301e-05, "loss": 2.2407429218292236, "step": 3112 }, { "epoch": 1.1487153466488307, "grad_norm": 1.1727313995361328, "learning_rate": 9.46903031220458e-05, "loss": 2.2380075454711914, "step": 3113 }, { "epoch": 1.149084367360118, "grad_norm": 1.16294264793396, "learning_rate": 9.466426977532247e-05, "loss": 2.0435376167297363, "step": 3114 }, { "epoch": 1.1494533880714055, "grad_norm": 1.2243717908859253, "learning_rate": 9.463817636077166e-05, "loss": 2.3480544090270996, "step": 3115 }, { "epoch": 1.149822408782693, "grad_norm": 1.8940809965133667, "learning_rate": 9.461202291348576e-05, "loss": 2.44993257522583, "step": 3116 }, { "epoch": 1.1501914294939803, "grad_norm": 1.2730379104614258, "learning_rate": 9.458580946863785e-05, "loss": 2.2375283241271973, "step": 3117 }, { "epoch": 1.1505604502052678, "grad_norm": 1.0598032474517822, "learning_rate": 9.455953606148172e-05, "loss": 2.21024751663208, "step": 3118 }, { "epoch": 1.1509294709165552, "grad_norm": 1.0919389724731445, "learning_rate": 9.453320272735178e-05, "loss": 2.005979299545288, "step": 3119 }, { "epoch": 1.1512984916278426, "grad_norm": 1.1598715782165527, "learning_rate": 9.45068095016631e-05, "loss": 2.110341787338257, "step": 3120 }, { "epoch": 1.15166751233913, "grad_norm": 1.576160192489624, "learning_rate": 9.448035641991121e-05, "loss": 2.5181822776794434, "step": 3121 }, { "epoch": 1.1520365330504174, "grad_norm": 1.1291695833206177, "learning_rate": 9.44538435176722e-05, "loss": 2.519340753555298, "step": 3122 }, { "epoch": 1.1524055537617048, "grad_norm": 1.3454173803329468, "learning_rate": 9.442727083060258e-05, "loss": 2.3823344707489014, "step": 3123 }, { "epoch": 1.1527745744729923, "grad_norm": 1.2988308668136597, "learning_rate": 9.440063839443929e-05, "loss": 2.4178874492645264, "step": 3124 }, { "epoch": 1.1531435951842797, "grad_norm": 1.0983692407608032, "learning_rate": 9.437394624499958e-05, "loss": 2.034365177154541, "step": 3125 }, { "epoch": 1.153512615895567, "grad_norm": 1.681356430053711, "learning_rate": 9.434719441818107e-05, "loss": 2.3911638259887695, "step": 3126 }, { "epoch": 1.1538816366068545, "grad_norm": 1.4751518964767456, "learning_rate": 9.432038294996155e-05, "loss": 2.630910634994507, "step": 3127 }, { "epoch": 1.154250657318142, "grad_norm": 1.301346778869629, "learning_rate": 9.429351187639914e-05, "loss": 2.380542039871216, "step": 3128 }, { "epoch": 1.1546196780294293, "grad_norm": 1.0771335363388062, "learning_rate": 9.4266581233632e-05, "loss": 2.1586148738861084, "step": 3129 }, { "epoch": 1.1549886987407167, "grad_norm": 1.4914519786834717, "learning_rate": 9.423959105787852e-05, "loss": 2.2884039878845215, "step": 3130 }, { "epoch": 1.1553577194520042, "grad_norm": 0.9678374528884888, "learning_rate": 9.421254138543702e-05, "loss": 2.3113045692443848, "step": 3131 }, { "epoch": 1.1557267401632916, "grad_norm": 1.2314658164978027, "learning_rate": 9.418543225268596e-05, "loss": 2.1425371170043945, "step": 3132 }, { "epoch": 1.156095760874579, "grad_norm": 1.0081827640533447, "learning_rate": 9.415826369608369e-05, "loss": 2.1523289680480957, "step": 3133 }, { "epoch": 1.1564647815858664, "grad_norm": 1.3371747732162476, "learning_rate": 9.413103575216853e-05, "loss": 2.363255500793457, "step": 3134 }, { "epoch": 1.1568338022971538, "grad_norm": 1.0438907146453857, "learning_rate": 9.41037484575586e-05, "loss": 2.0203635692596436, "step": 3135 }, { "epoch": 1.1572028230084412, "grad_norm": 1.5198173522949219, "learning_rate": 9.407640184895193e-05, "loss": 2.505650758743286, "step": 3136 }, { "epoch": 1.1575718437197287, "grad_norm": 1.3573668003082275, "learning_rate": 9.404899596312623e-05, "loss": 2.4251837730407715, "step": 3137 }, { "epoch": 1.157940864431016, "grad_norm": 1.0484641790390015, "learning_rate": 9.402153083693898e-05, "loss": 2.0261991024017334, "step": 3138 }, { "epoch": 1.1583098851423037, "grad_norm": 1.283238172531128, "learning_rate": 9.399400650732735e-05, "loss": 2.6753363609313965, "step": 3139 }, { "epoch": 1.1586789058535911, "grad_norm": 1.2240735292434692, "learning_rate": 9.396642301130807e-05, "loss": 2.188692331314087, "step": 3140 }, { "epoch": 1.1590479265648785, "grad_norm": 1.265328049659729, "learning_rate": 9.393878038597747e-05, "loss": 2.2494449615478516, "step": 3141 }, { "epoch": 1.159416947276166, "grad_norm": 1.0441162586212158, "learning_rate": 9.391107866851143e-05, "loss": 2.0810468196868896, "step": 3142 }, { "epoch": 1.1597859679874534, "grad_norm": 1.0633798837661743, "learning_rate": 9.388331789616527e-05, "loss": 2.2021102905273438, "step": 3143 }, { "epoch": 1.1601549886987408, "grad_norm": 1.0694890022277832, "learning_rate": 9.385549810627373e-05, "loss": 2.2702879905700684, "step": 3144 }, { "epoch": 1.1605240094100282, "grad_norm": 1.248457908630371, "learning_rate": 9.382761933625095e-05, "loss": 2.306567430496216, "step": 3145 }, { "epoch": 1.1608930301213156, "grad_norm": 1.5207546949386597, "learning_rate": 9.379968162359034e-05, "loss": 2.5696380138397217, "step": 3146 }, { "epoch": 1.161262050832603, "grad_norm": 1.2321131229400635, "learning_rate": 9.377168500586463e-05, "loss": 2.2960734367370605, "step": 3147 }, { "epoch": 1.1616310715438904, "grad_norm": 1.2443983554840088, "learning_rate": 9.374362952072575e-05, "loss": 2.1299962997436523, "step": 3148 }, { "epoch": 1.1620000922551779, "grad_norm": 1.1541646718978882, "learning_rate": 9.37155152059048e-05, "loss": 2.2903056144714355, "step": 3149 }, { "epoch": 1.1623691129664653, "grad_norm": 1.3018437623977661, "learning_rate": 9.368734209921201e-05, "loss": 2.155406951904297, "step": 3150 }, { "epoch": 1.1627381336777527, "grad_norm": 1.3526185750961304, "learning_rate": 9.365911023853663e-05, "loss": 2.4644646644592285, "step": 3151 }, { "epoch": 1.16310715438904, "grad_norm": 1.141622543334961, "learning_rate": 9.363081966184702e-05, "loss": 2.135586738586426, "step": 3152 }, { "epoch": 1.1634761751003275, "grad_norm": 1.147992730140686, "learning_rate": 9.360247040719039e-05, "loss": 2.359842300415039, "step": 3153 }, { "epoch": 1.163845195811615, "grad_norm": 1.3022924661636353, "learning_rate": 9.357406251269295e-05, "loss": 2.433149576187134, "step": 3154 }, { "epoch": 1.1642142165229024, "grad_norm": 1.0744410753250122, "learning_rate": 9.354559601655974e-05, "loss": 2.0761795043945312, "step": 3155 }, { "epoch": 1.1645832372341898, "grad_norm": 1.0704901218414307, "learning_rate": 9.351707095707465e-05, "loss": 1.95772385597229, "step": 3156 }, { "epoch": 1.1649522579454772, "grad_norm": 1.1841299533843994, "learning_rate": 9.348848737260025e-05, "loss": 2.342355966567993, "step": 3157 }, { "epoch": 1.1653212786567646, "grad_norm": 1.0198588371276855, "learning_rate": 9.34598453015779e-05, "loss": 2.3824853897094727, "step": 3158 }, { "epoch": 1.165690299368052, "grad_norm": 1.4849443435668945, "learning_rate": 9.343114478252758e-05, "loss": 2.2296078205108643, "step": 3159 }, { "epoch": 1.1660593200793394, "grad_norm": 1.4181482791900635, "learning_rate": 9.340238585404788e-05, "loss": 2.2069191932678223, "step": 3160 }, { "epoch": 1.1664283407906268, "grad_norm": 1.2179274559020996, "learning_rate": 9.337356855481591e-05, "loss": 2.1339528560638428, "step": 3161 }, { "epoch": 1.1667973615019143, "grad_norm": 1.176345944404602, "learning_rate": 9.334469292358736e-05, "loss": 2.321166515350342, "step": 3162 }, { "epoch": 1.1671663822132017, "grad_norm": 1.2189557552337646, "learning_rate": 9.331575899919631e-05, "loss": 2.3035945892333984, "step": 3163 }, { "epoch": 1.167535402924489, "grad_norm": 1.3463096618652344, "learning_rate": 9.328676682055524e-05, "loss": 2.1776459217071533, "step": 3164 }, { "epoch": 1.1679044236357765, "grad_norm": 1.3557798862457275, "learning_rate": 9.3257716426655e-05, "loss": 2.4240529537200928, "step": 3165 }, { "epoch": 1.168273444347064, "grad_norm": 1.1145424842834473, "learning_rate": 9.32286078565647e-05, "loss": 2.3938331604003906, "step": 3166 }, { "epoch": 1.1686424650583513, "grad_norm": 1.2242192029953003, "learning_rate": 9.319944114943171e-05, "loss": 2.2659354209899902, "step": 3167 }, { "epoch": 1.1690114857696388, "grad_norm": 1.42521071434021, "learning_rate": 9.317021634448161e-05, "loss": 2.2623727321624756, "step": 3168 }, { "epoch": 1.1693805064809262, "grad_norm": 1.1609444618225098, "learning_rate": 9.314093348101806e-05, "loss": 2.190453290939331, "step": 3169 }, { "epoch": 1.1697495271922136, "grad_norm": 1.224779725074768, "learning_rate": 9.311159259842283e-05, "loss": 2.4176273345947266, "step": 3170 }, { "epoch": 1.170118547903501, "grad_norm": 1.2663888931274414, "learning_rate": 9.308219373615574e-05, "loss": 2.3314459323883057, "step": 3171 }, { "epoch": 1.1704875686147884, "grad_norm": 1.2124848365783691, "learning_rate": 9.305273693375454e-05, "loss": 1.9994237422943115, "step": 3172 }, { "epoch": 1.1708565893260758, "grad_norm": 1.1816301345825195, "learning_rate": 9.302322223083495e-05, "loss": 1.9004156589508057, "step": 3173 }, { "epoch": 1.1712256100373633, "grad_norm": 1.3940671682357788, "learning_rate": 9.29936496670905e-05, "loss": 2.399667501449585, "step": 3174 }, { "epoch": 1.171594630748651, "grad_norm": 1.31154465675354, "learning_rate": 9.296401928229261e-05, "loss": 2.197573661804199, "step": 3175 }, { "epoch": 1.1719636514599383, "grad_norm": 1.2977360486984253, "learning_rate": 9.29343311162904e-05, "loss": 2.190610885620117, "step": 3176 }, { "epoch": 1.1723326721712257, "grad_norm": 1.2015266418457031, "learning_rate": 9.290458520901072e-05, "loss": 2.2624244689941406, "step": 3177 }, { "epoch": 1.1727016928825131, "grad_norm": 1.3382707834243774, "learning_rate": 9.287478160045808e-05, "loss": 2.512464761734009, "step": 3178 }, { "epoch": 1.1730707135938006, "grad_norm": 1.6516168117523193, "learning_rate": 9.284492033071459e-05, "loss": 2.483370304107666, "step": 3179 }, { "epoch": 1.173439734305088, "grad_norm": 1.0699466466903687, "learning_rate": 9.28150014399399e-05, "loss": 2.2230746746063232, "step": 3180 }, { "epoch": 1.1738087550163754, "grad_norm": 0.9074277877807617, "learning_rate": 9.278502496837116e-05, "loss": 2.058927297592163, "step": 3181 }, { "epoch": 1.1741777757276628, "grad_norm": 1.3944756984710693, "learning_rate": 9.275499095632294e-05, "loss": 2.3634066581726074, "step": 3182 }, { "epoch": 1.1745467964389502, "grad_norm": 1.065396785736084, "learning_rate": 9.272489944418724e-05, "loss": 2.318218231201172, "step": 3183 }, { "epoch": 1.1749158171502376, "grad_norm": 1.1903772354125977, "learning_rate": 9.269475047243332e-05, "loss": 2.243152618408203, "step": 3184 }, { "epoch": 1.175284837861525, "grad_norm": 1.2588956356048584, "learning_rate": 9.266454408160779e-05, "loss": 2.245896577835083, "step": 3185 }, { "epoch": 1.1756538585728125, "grad_norm": 1.0458413362503052, "learning_rate": 9.263428031233443e-05, "loss": 2.1319634914398193, "step": 3186 }, { "epoch": 1.1760228792840999, "grad_norm": 1.1485602855682373, "learning_rate": 9.260395920531421e-05, "loss": 2.269712448120117, "step": 3187 }, { "epoch": 1.1763918999953873, "grad_norm": 1.3907285928726196, "learning_rate": 9.257358080132523e-05, "loss": 1.8588699102401733, "step": 3188 }, { "epoch": 1.1767609207066747, "grad_norm": 1.1646180152893066, "learning_rate": 9.254314514122259e-05, "loss": 2.095424175262451, "step": 3189 }, { "epoch": 1.1771299414179621, "grad_norm": 0.977502703666687, "learning_rate": 9.251265226593844e-05, "loss": 2.0936083793640137, "step": 3190 }, { "epoch": 1.1774989621292495, "grad_norm": 1.2701300382614136, "learning_rate": 9.248210221648187e-05, "loss": 2.0806236267089844, "step": 3191 }, { "epoch": 1.177867982840537, "grad_norm": 1.2624444961547852, "learning_rate": 9.245149503393884e-05, "loss": 2.0586395263671875, "step": 3192 }, { "epoch": 1.1782370035518244, "grad_norm": 1.0514472723007202, "learning_rate": 9.242083075947216e-05, "loss": 1.9903342723846436, "step": 3193 }, { "epoch": 1.1786060242631118, "grad_norm": 1.1973508596420288, "learning_rate": 9.239010943432146e-05, "loss": 2.268153667449951, "step": 3194 }, { "epoch": 1.1789750449743992, "grad_norm": 1.3169218301773071, "learning_rate": 9.235933109980301e-05, "loss": 2.1548590660095215, "step": 3195 }, { "epoch": 1.1793440656856866, "grad_norm": 1.0195046663284302, "learning_rate": 9.232849579730984e-05, "loss": 2.2722513675689697, "step": 3196 }, { "epoch": 1.179713086396974, "grad_norm": 1.3852885961532593, "learning_rate": 9.229760356831154e-05, "loss": 2.508429527282715, "step": 3197 }, { "epoch": 1.1800821071082614, "grad_norm": 1.4471842050552368, "learning_rate": 9.226665445435428e-05, "loss": 2.449007749557495, "step": 3198 }, { "epoch": 1.1804511278195489, "grad_norm": 1.22493314743042, "learning_rate": 9.223564849706071e-05, "loss": 2.4319207668304443, "step": 3199 }, { "epoch": 1.1808201485308363, "grad_norm": 1.2292424440383911, "learning_rate": 9.220458573812996e-05, "loss": 2.480396270751953, "step": 3200 }, { "epoch": 1.1811891692421237, "grad_norm": 1.2916386127471924, "learning_rate": 9.217346621933754e-05, "loss": 2.125638246536255, "step": 3201 }, { "epoch": 1.181558189953411, "grad_norm": 1.4476752281188965, "learning_rate": 9.214228998253527e-05, "loss": 2.5323374271392822, "step": 3202 }, { "epoch": 1.1819272106646985, "grad_norm": 1.3327665328979492, "learning_rate": 9.211105706965128e-05, "loss": 2.379152774810791, "step": 3203 }, { "epoch": 1.182296231375986, "grad_norm": 1.208133578300476, "learning_rate": 9.20797675226899e-05, "loss": 2.452718496322632, "step": 3204 }, { "epoch": 1.1826652520872734, "grad_norm": 1.1764219999313354, "learning_rate": 9.204842138373167e-05, "loss": 2.117252826690674, "step": 3205 }, { "epoch": 1.1830342727985608, "grad_norm": 1.0734453201293945, "learning_rate": 9.201701869493317e-05, "loss": 2.110720157623291, "step": 3206 }, { "epoch": 1.1834032935098482, "grad_norm": 1.131759762763977, "learning_rate": 9.198555949852709e-05, "loss": 2.442760944366455, "step": 3207 }, { "epoch": 1.1837723142211356, "grad_norm": 1.1054357290267944, "learning_rate": 9.195404383682209e-05, "loss": 2.4476490020751953, "step": 3208 }, { "epoch": 1.184141334932423, "grad_norm": 1.4739761352539062, "learning_rate": 9.192247175220276e-05, "loss": 2.433220863342285, "step": 3209 }, { "epoch": 1.1845103556437104, "grad_norm": 0.9825840592384338, "learning_rate": 9.18908432871296e-05, "loss": 2.1265764236450195, "step": 3210 }, { "epoch": 1.1848793763549978, "grad_norm": 1.1555835008621216, "learning_rate": 9.185915848413893e-05, "loss": 2.196941614151001, "step": 3211 }, { "epoch": 1.1852483970662853, "grad_norm": 1.0551469326019287, "learning_rate": 9.182741738584278e-05, "loss": 2.299485445022583, "step": 3212 }, { "epoch": 1.1856174177775727, "grad_norm": 1.3282438516616821, "learning_rate": 9.179562003492898e-05, "loss": 2.1749956607818604, "step": 3213 }, { "epoch": 1.18598643848886, "grad_norm": 1.3330447673797607, "learning_rate": 9.176376647416096e-05, "loss": 2.441835880279541, "step": 3214 }, { "epoch": 1.1863554592001475, "grad_norm": 1.2996368408203125, "learning_rate": 9.173185674637775e-05, "loss": 2.157562255859375, "step": 3215 }, { "epoch": 1.186724479911435, "grad_norm": 1.1794039011001587, "learning_rate": 9.16998908944939e-05, "loss": 2.4779748916625977, "step": 3216 }, { "epoch": 1.1870935006227223, "grad_norm": 1.1913965940475464, "learning_rate": 9.16678689614995e-05, "loss": 2.1619136333465576, "step": 3217 }, { "epoch": 1.1874625213340098, "grad_norm": 1.222075343132019, "learning_rate": 9.163579099046e-05, "loss": 2.19252872467041, "step": 3218 }, { "epoch": 1.1878315420452972, "grad_norm": 1.2544031143188477, "learning_rate": 9.160365702451625e-05, "loss": 2.2551116943359375, "step": 3219 }, { "epoch": 1.1882005627565846, "grad_norm": 1.0641447305679321, "learning_rate": 9.15714671068844e-05, "loss": 2.100008010864258, "step": 3220 }, { "epoch": 1.1885695834678722, "grad_norm": 1.0371670722961426, "learning_rate": 9.153922128085581e-05, "loss": 2.046896457672119, "step": 3221 }, { "epoch": 1.1889386041791596, "grad_norm": 1.258983850479126, "learning_rate": 9.15069195897971e-05, "loss": 2.2679295539855957, "step": 3222 }, { "epoch": 1.189307624890447, "grad_norm": 1.1570239067077637, "learning_rate": 9.147456207714997e-05, "loss": 2.209266185760498, "step": 3223 }, { "epoch": 1.1896766456017345, "grad_norm": 1.197684407234192, "learning_rate": 9.14421487864312e-05, "loss": 1.89603853225708, "step": 3224 }, { "epoch": 1.1900456663130219, "grad_norm": 1.351925253868103, "learning_rate": 9.140967976123259e-05, "loss": 2.2843711376190186, "step": 3225 }, { "epoch": 1.1904146870243093, "grad_norm": 1.2932499647140503, "learning_rate": 9.137715504522091e-05, "loss": 2.371371030807495, "step": 3226 }, { "epoch": 1.1907837077355967, "grad_norm": 1.4813637733459473, "learning_rate": 9.13445746821378e-05, "loss": 2.300351142883301, "step": 3227 }, { "epoch": 1.1911527284468841, "grad_norm": 1.3303779363632202, "learning_rate": 9.131193871579975e-05, "loss": 2.278735876083374, "step": 3228 }, { "epoch": 1.1915217491581716, "grad_norm": 1.034745454788208, "learning_rate": 9.127924719009804e-05, "loss": 2.292043924331665, "step": 3229 }, { "epoch": 1.191890769869459, "grad_norm": 1.0290249586105347, "learning_rate": 9.124650014899867e-05, "loss": 2.113194465637207, "step": 3230 }, { "epoch": 1.1922597905807464, "grad_norm": 1.1803334951400757, "learning_rate": 9.121369763654228e-05, "loss": 2.5234076976776123, "step": 3231 }, { "epoch": 1.1926288112920338, "grad_norm": 1.1526269912719727, "learning_rate": 9.118083969684413e-05, "loss": 2.5803380012512207, "step": 3232 }, { "epoch": 1.1929978320033212, "grad_norm": 1.1297390460968018, "learning_rate": 9.114792637409401e-05, "loss": 2.0821688175201416, "step": 3233 }, { "epoch": 1.1933668527146086, "grad_norm": 1.060042381286621, "learning_rate": 9.111495771255623e-05, "loss": 2.453917980194092, "step": 3234 }, { "epoch": 1.193735873425896, "grad_norm": 1.3383007049560547, "learning_rate": 9.108193375656945e-05, "loss": 2.261434555053711, "step": 3235 }, { "epoch": 1.1941048941371835, "grad_norm": 1.131361961364746, "learning_rate": 9.104885455054677e-05, "loss": 2.0166025161743164, "step": 3236 }, { "epoch": 1.1944739148484709, "grad_norm": 1.0878463983535767, "learning_rate": 9.101572013897555e-05, "loss": 1.9969205856323242, "step": 3237 }, { "epoch": 1.1948429355597583, "grad_norm": 1.122007131576538, "learning_rate": 9.098253056641742e-05, "loss": 2.2814574241638184, "step": 3238 }, { "epoch": 1.1952119562710457, "grad_norm": 1.0480220317840576, "learning_rate": 9.094928587750818e-05, "loss": 2.103980541229248, "step": 3239 }, { "epoch": 1.1955809769823331, "grad_norm": 1.2849427461624146, "learning_rate": 9.091598611695772e-05, "loss": 2.2593774795532227, "step": 3240 }, { "epoch": 1.1959499976936205, "grad_norm": 1.179304838180542, "learning_rate": 9.088263132955011e-05, "loss": 2.3362348079681396, "step": 3241 }, { "epoch": 1.196319018404908, "grad_norm": 1.0887510776519775, "learning_rate": 9.084922156014328e-05, "loss": 2.2480661869049072, "step": 3242 }, { "epoch": 1.1966880391161954, "grad_norm": 1.3540750741958618, "learning_rate": 9.081575685366919e-05, "loss": 2.154998302459717, "step": 3243 }, { "epoch": 1.1970570598274828, "grad_norm": 1.2530829906463623, "learning_rate": 9.078223725513366e-05, "loss": 2.003631353378296, "step": 3244 }, { "epoch": 1.1974260805387702, "grad_norm": 1.7363594770431519, "learning_rate": 9.074866280961632e-05, "loss": 2.619837522506714, "step": 3245 }, { "epoch": 1.1977951012500576, "grad_norm": 1.851965069770813, "learning_rate": 9.071503356227063e-05, "loss": 2.3820676803588867, "step": 3246 }, { "epoch": 1.198164121961345, "grad_norm": 1.2144649028778076, "learning_rate": 9.068134955832363e-05, "loss": 2.2238597869873047, "step": 3247 }, { "epoch": 1.1985331426726324, "grad_norm": 1.0974757671356201, "learning_rate": 9.064761084307613e-05, "loss": 2.244767904281616, "step": 3248 }, { "epoch": 1.1989021633839199, "grad_norm": 1.1476398706436157, "learning_rate": 9.061381746190243e-05, "loss": 2.339564800262451, "step": 3249 }, { "epoch": 1.1992711840952073, "grad_norm": 1.1393887996673584, "learning_rate": 9.057996946025037e-05, "loss": 2.6328868865966797, "step": 3250 }, { "epoch": 1.1996402048064947, "grad_norm": 1.4006558656692505, "learning_rate": 9.05460668836413e-05, "loss": 2.344054698944092, "step": 3251 }, { "epoch": 1.200009225517782, "grad_norm": 1.1956920623779297, "learning_rate": 9.051210977766987e-05, "loss": 2.008216381072998, "step": 3252 } ], "logging_steps": 1, "max_steps": 5420, "num_input_tokens_seen": 0, "num_train_epochs": 2, "save_steps": 542, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 2.769950219146445e+18, "train_batch_size": 4, "trial_name": null, "trial_params": null }