{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 14552, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 6.871907641561298e-05, "grad_norm": 22.647811889648438, "learning_rate": 6.864988558352403e-08, "loss": 2.7231, "step": 1 }, { "epoch": 0.00013743815283122595, "grad_norm": 25.14017677307129, "learning_rate": 1.3729977116704806e-07, "loss": 2.7655, "step": 2 }, { "epoch": 0.00020615722924683893, "grad_norm": 25.43830108642578, "learning_rate": 2.0594965675057208e-07, "loss": 2.733, "step": 3 }, { "epoch": 0.0002748763056624519, "grad_norm": 22.213796615600586, "learning_rate": 2.745995423340961e-07, "loss": 2.8417, "step": 4 }, { "epoch": 0.00034359538207806486, "grad_norm": 25.579633712768555, "learning_rate": 3.432494279176201e-07, "loss": 2.9468, "step": 5 }, { "epoch": 0.00041231445849367786, "grad_norm": 28.175825119018555, "learning_rate": 4.1189931350114416e-07, "loss": 2.8635, "step": 6 }, { "epoch": 0.0004810335349092908, "grad_norm": 23.89807891845703, "learning_rate": 4.805491990846682e-07, "loss": 2.7883, "step": 7 }, { "epoch": 0.0005497526113249038, "grad_norm": 22.523115158081055, "learning_rate": 5.491990846681922e-07, "loss": 2.8533, "step": 8 }, { "epoch": 0.0006184716877405168, "grad_norm": 24.960817337036133, "learning_rate": 6.178489702517163e-07, "loss": 2.8229, "step": 9 }, { "epoch": 0.0006871907641561297, "grad_norm": 25.564735412597656, "learning_rate": 6.864988558352402e-07, "loss": 2.7614, "step": 10 }, { "epoch": 0.0007559098405717427, "grad_norm": 21.750089645385742, "learning_rate": 7.551487414187643e-07, "loss": 2.6666, "step": 11 }, { "epoch": 0.0008246289169873557, "grad_norm": 22.495403289794922, "learning_rate": 8.237986270022883e-07, "loss": 2.6191, "step": 12 }, { "epoch": 0.0008933479934029687, "grad_norm": 19.478395462036133, "learning_rate": 8.924485125858125e-07, "loss": 2.5874, "step": 13 }, { "epoch": 0.0009620670698185816, "grad_norm": 22.656402587890625, "learning_rate": 9.610983981693364e-07, "loss": 2.6229, "step": 14 }, { "epoch": 0.0010307861462341946, "grad_norm": 19.629934310913086, "learning_rate": 1.0297482837528604e-06, "loss": 2.3967, "step": 15 }, { "epoch": 0.0010995052226498076, "grad_norm": 17.960237503051758, "learning_rate": 1.0983981693363845e-06, "loss": 2.3373, "step": 16 }, { "epoch": 0.0011682242990654205, "grad_norm": 17.880456924438477, "learning_rate": 1.1670480549199085e-06, "loss": 2.1947, "step": 17 }, { "epoch": 0.0012369433754810335, "grad_norm": 13.249114036560059, "learning_rate": 1.2356979405034326e-06, "loss": 2.1597, "step": 18 }, { "epoch": 0.0013056624518966466, "grad_norm": 14.289917945861816, "learning_rate": 1.3043478260869566e-06, "loss": 2.0542, "step": 19 }, { "epoch": 0.0013743815283122594, "grad_norm": 12.31395149230957, "learning_rate": 1.3729977116704805e-06, "loss": 1.9566, "step": 20 }, { "epoch": 0.0014431006047278725, "grad_norm": 13.818697929382324, "learning_rate": 1.4416475972540045e-06, "loss": 1.9345, "step": 21 }, { "epoch": 0.0015118196811434853, "grad_norm": 12.134113311767578, "learning_rate": 1.5102974828375285e-06, "loss": 1.8592, "step": 22 }, { "epoch": 0.0015805387575590984, "grad_norm": 10.896883964538574, "learning_rate": 1.5789473684210526e-06, "loss": 1.8806, "step": 23 }, { "epoch": 0.0016492578339747114, "grad_norm": 10.099254608154297, "learning_rate": 1.6475972540045766e-06, "loss": 1.7029, "step": 24 }, { "epoch": 0.0017179769103903243, "grad_norm": 11.785741806030273, "learning_rate": 1.7162471395881009e-06, "loss": 1.858, "step": 25 }, { "epoch": 0.0017866959868059373, "grad_norm": 10.147101402282715, "learning_rate": 1.784897025171625e-06, "loss": 1.7108, "step": 26 }, { "epoch": 0.0018554150632215504, "grad_norm": 9.865859985351562, "learning_rate": 1.8535469107551488e-06, "loss": 1.5348, "step": 27 }, { "epoch": 0.0019241341396371632, "grad_norm": 9.5696382522583, "learning_rate": 1.922196796338673e-06, "loss": 1.5748, "step": 28 }, { "epoch": 0.001992853216052776, "grad_norm": 9.457874298095703, "learning_rate": 1.990846681922197e-06, "loss": 1.5935, "step": 29 }, { "epoch": 0.002061572292468389, "grad_norm": 7.789409160614014, "learning_rate": 2.059496567505721e-06, "loss": 1.4632, "step": 30 }, { "epoch": 0.002130291368884002, "grad_norm": 7.926306247711182, "learning_rate": 2.1281464530892445e-06, "loss": 1.4375, "step": 31 }, { "epoch": 0.0021990104452996153, "grad_norm": 8.512276649475098, "learning_rate": 2.196796338672769e-06, "loss": 1.2944, "step": 32 }, { "epoch": 0.0022677295217152283, "grad_norm": 7.453540802001953, "learning_rate": 2.265446224256293e-06, "loss": 1.2664, "step": 33 }, { "epoch": 0.002336448598130841, "grad_norm": 7.3873748779296875, "learning_rate": 2.334096109839817e-06, "loss": 1.2464, "step": 34 }, { "epoch": 0.002405167674546454, "grad_norm": 7.915987014770508, "learning_rate": 2.402745995423341e-06, "loss": 1.2787, "step": 35 }, { "epoch": 0.002473886750962067, "grad_norm": 8.124571800231934, "learning_rate": 2.471395881006865e-06, "loss": 1.2902, "step": 36 }, { "epoch": 0.00254260582737768, "grad_norm": 8.436516761779785, "learning_rate": 2.540045766590389e-06, "loss": 1.1329, "step": 37 }, { "epoch": 0.002611324903793293, "grad_norm": 7.790831089019775, "learning_rate": 2.6086956521739132e-06, "loss": 1.1475, "step": 38 }, { "epoch": 0.002680043980208906, "grad_norm": 7.722904682159424, "learning_rate": 2.6773455377574373e-06, "loss": 1.1188, "step": 39 }, { "epoch": 0.002748763056624519, "grad_norm": 7.2353315353393555, "learning_rate": 2.745995423340961e-06, "loss": 1.0654, "step": 40 }, { "epoch": 0.002817482133040132, "grad_norm": 6.5284833908081055, "learning_rate": 2.8146453089244854e-06, "loss": 1.112, "step": 41 }, { "epoch": 0.002886201209455745, "grad_norm": 6.610606670379639, "learning_rate": 2.883295194508009e-06, "loss": 1.0334, "step": 42 }, { "epoch": 0.002954920285871358, "grad_norm": 7.103153705596924, "learning_rate": 2.9519450800915335e-06, "loss": 1.1221, "step": 43 }, { "epoch": 0.0030236393622869707, "grad_norm": 6.832851409912109, "learning_rate": 3.020594965675057e-06, "loss": 1.0077, "step": 44 }, { "epoch": 0.0030923584387025837, "grad_norm": 8.963631629943848, "learning_rate": 3.0892448512585815e-06, "loss": 0.982, "step": 45 }, { "epoch": 0.0031610775151181968, "grad_norm": 6.164712905883789, "learning_rate": 3.157894736842105e-06, "loss": 0.9607, "step": 46 }, { "epoch": 0.00322979659153381, "grad_norm": 5.981264591217041, "learning_rate": 3.226544622425629e-06, "loss": 0.8971, "step": 47 }, { "epoch": 0.003298515667949423, "grad_norm": 7.265660285949707, "learning_rate": 3.2951945080091533e-06, "loss": 0.8548, "step": 48 }, { "epoch": 0.0033672347443650355, "grad_norm": 5.986792087554932, "learning_rate": 3.3638443935926773e-06, "loss": 0.8755, "step": 49 }, { "epoch": 0.0034359538207806486, "grad_norm": 5.985032081604004, "learning_rate": 3.4324942791762018e-06, "loss": 0.8909, "step": 50 }, { "epoch": 0.0035046728971962616, "grad_norm": 6.479555130004883, "learning_rate": 3.5011441647597254e-06, "loss": 0.911, "step": 51 }, { "epoch": 0.0035733919736118747, "grad_norm": 5.790592193603516, "learning_rate": 3.56979405034325e-06, "loss": 0.8362, "step": 52 }, { "epoch": 0.0036421110500274878, "grad_norm": 5.732059001922607, "learning_rate": 3.6384439359267735e-06, "loss": 0.8157, "step": 53 }, { "epoch": 0.003710830126443101, "grad_norm": 6.173818111419678, "learning_rate": 3.7070938215102975e-06, "loss": 0.81, "step": 54 }, { "epoch": 0.0037795492028587134, "grad_norm": 5.504522800445557, "learning_rate": 3.775743707093822e-06, "loss": 0.7861, "step": 55 }, { "epoch": 0.0038482682792743265, "grad_norm": 5.516193389892578, "learning_rate": 3.844393592677346e-06, "loss": 0.6926, "step": 56 }, { "epoch": 0.0039169873556899396, "grad_norm": 6.244246959686279, "learning_rate": 3.913043478260869e-06, "loss": 0.7928, "step": 57 }, { "epoch": 0.003985706432105552, "grad_norm": 6.11889123916626, "learning_rate": 3.981693363844394e-06, "loss": 0.7514, "step": 58 }, { "epoch": 0.004054425508521166, "grad_norm": 5.907434940338135, "learning_rate": 4.050343249427918e-06, "loss": 0.7334, "step": 59 }, { "epoch": 0.004123144584936778, "grad_norm": 5.9996418952941895, "learning_rate": 4.118993135011442e-06, "loss": 0.7585, "step": 60 }, { "epoch": 0.004191863661352392, "grad_norm": 6.548773288726807, "learning_rate": 4.187643020594965e-06, "loss": 0.7114, "step": 61 }, { "epoch": 0.004260582737768004, "grad_norm": 6.755104064941406, "learning_rate": 4.256292906178489e-06, "loss": 0.6996, "step": 62 }, { "epoch": 0.004329301814183617, "grad_norm": 5.160214900970459, "learning_rate": 4.324942791762014e-06, "loss": 0.6795, "step": 63 }, { "epoch": 0.0043980208905992305, "grad_norm": 6.803913593292236, "learning_rate": 4.393592677345538e-06, "loss": 0.6732, "step": 64 }, { "epoch": 0.004466739967014843, "grad_norm": 5.484057426452637, "learning_rate": 4.4622425629290616e-06, "loss": 0.7242, "step": 65 }, { "epoch": 0.004535459043430457, "grad_norm": 5.926074504852295, "learning_rate": 4.530892448512586e-06, "loss": 0.6437, "step": 66 }, { "epoch": 0.004604178119846069, "grad_norm": 7.8767991065979, "learning_rate": 4.5995423340961105e-06, "loss": 0.6507, "step": 67 }, { "epoch": 0.004672897196261682, "grad_norm": 5.525781154632568, "learning_rate": 4.668192219679634e-06, "loss": 0.6167, "step": 68 }, { "epoch": 0.004741616272677295, "grad_norm": 5.893625259399414, "learning_rate": 4.736842105263158e-06, "loss": 0.665, "step": 69 }, { "epoch": 0.004810335349092908, "grad_norm": 5.334264755249023, "learning_rate": 4.805491990846682e-06, "loss": 0.5526, "step": 70 }, { "epoch": 0.0048790544255085215, "grad_norm": 5.3079423904418945, "learning_rate": 4.874141876430206e-06, "loss": 0.5745, "step": 71 }, { "epoch": 0.004947773501924134, "grad_norm": 5.758218765258789, "learning_rate": 4.94279176201373e-06, "loss": 0.6281, "step": 72 }, { "epoch": 0.005016492578339747, "grad_norm": 6.461379051208496, "learning_rate": 5.011441647597254e-06, "loss": 0.6162, "step": 73 }, { "epoch": 0.00508521165475536, "grad_norm": 5.237447738647461, "learning_rate": 5.080091533180778e-06, "loss": 0.5323, "step": 74 }, { "epoch": 0.005153930731170973, "grad_norm": 5.8103766441345215, "learning_rate": 5.148741418764302e-06, "loss": 0.5391, "step": 75 }, { "epoch": 0.005222649807586586, "grad_norm": 5.291259765625, "learning_rate": 5.2173913043478265e-06, "loss": 0.5929, "step": 76 }, { "epoch": 0.005291368884002199, "grad_norm": 5.118488311767578, "learning_rate": 5.28604118993135e-06, "loss": 0.5586, "step": 77 }, { "epoch": 0.005360087960417812, "grad_norm": 5.744769096374512, "learning_rate": 5.3546910755148746e-06, "loss": 0.5636, "step": 78 }, { "epoch": 0.005428807036833425, "grad_norm": 4.928773403167725, "learning_rate": 5.423340961098398e-06, "loss": 0.5782, "step": 79 }, { "epoch": 0.005497526113249038, "grad_norm": 5.171982765197754, "learning_rate": 5.491990846681922e-06, "loss": 0.5098, "step": 80 }, { "epoch": 0.005566245189664651, "grad_norm": 5.877892971038818, "learning_rate": 5.560640732265447e-06, "loss": 0.5038, "step": 81 }, { "epoch": 0.005634964266080264, "grad_norm": 5.041689872741699, "learning_rate": 5.629290617848971e-06, "loss": 0.4773, "step": 82 }, { "epoch": 0.0057036833424958765, "grad_norm": 4.9962029457092285, "learning_rate": 5.697940503432494e-06, "loss": 0.5484, "step": 83 }, { "epoch": 0.00577240241891149, "grad_norm": 4.987279891967773, "learning_rate": 5.766590389016018e-06, "loss": 0.5216, "step": 84 }, { "epoch": 0.005841121495327103, "grad_norm": 5.275233268737793, "learning_rate": 5.8352402745995424e-06, "loss": 0.534, "step": 85 }, { "epoch": 0.005909840571742716, "grad_norm": 5.686601638793945, "learning_rate": 5.903890160183067e-06, "loss": 0.5596, "step": 86 }, { "epoch": 0.005978559648158329, "grad_norm": 4.549646377563477, "learning_rate": 5.9725400457665905e-06, "loss": 0.5101, "step": 87 }, { "epoch": 0.006047278724573941, "grad_norm": 5.142058849334717, "learning_rate": 6.041189931350114e-06, "loss": 0.4555, "step": 88 }, { "epoch": 0.006115997800989555, "grad_norm": 4.022197246551514, "learning_rate": 6.109839816933639e-06, "loss": 0.4815, "step": 89 }, { "epoch": 0.0061847168774051674, "grad_norm": 4.709782600402832, "learning_rate": 6.178489702517163e-06, "loss": 0.4665, "step": 90 }, { "epoch": 0.006253435953820781, "grad_norm": 5.474863052368164, "learning_rate": 6.247139588100687e-06, "loss": 0.4959, "step": 91 }, { "epoch": 0.0063221550302363936, "grad_norm": 4.960180282592773, "learning_rate": 6.31578947368421e-06, "loss": 0.4998, "step": 92 }, { "epoch": 0.006390874106652006, "grad_norm": 5.2681169509887695, "learning_rate": 6.384439359267735e-06, "loss": 0.5105, "step": 93 }, { "epoch": 0.00645959318306762, "grad_norm": 5.513281345367432, "learning_rate": 6.453089244851258e-06, "loss": 0.5027, "step": 94 }, { "epoch": 0.006528312259483232, "grad_norm": 5.83230447769165, "learning_rate": 6.521739130434783e-06, "loss": 0.4701, "step": 95 }, { "epoch": 0.006597031335898846, "grad_norm": 4.777600288391113, "learning_rate": 6.5903890160183065e-06, "loss": 0.4822, "step": 96 }, { "epoch": 0.006665750412314458, "grad_norm": 4.752645492553711, "learning_rate": 6.659038901601831e-06, "loss": 0.4517, "step": 97 }, { "epoch": 0.006734469488730071, "grad_norm": 6.225427627563477, "learning_rate": 6.727688787185355e-06, "loss": 0.4506, "step": 98 }, { "epoch": 0.0068031885651456845, "grad_norm": 4.948246955871582, "learning_rate": 6.796338672768879e-06, "loss": 0.4655, "step": 99 }, { "epoch": 0.006871907641561297, "grad_norm": 6.330719470977783, "learning_rate": 6.8649885583524035e-06, "loss": 0.3929, "step": 100 }, { "epoch": 0.006940626717976911, "grad_norm": 5.323206424713135, "learning_rate": 6.933638443935927e-06, "loss": 0.4544, "step": 101 }, { "epoch": 0.007009345794392523, "grad_norm": 5.411411762237549, "learning_rate": 7.002288329519451e-06, "loss": 0.4171, "step": 102 }, { "epoch": 0.007078064870808137, "grad_norm": 5.343921661376953, "learning_rate": 7.070938215102974e-06, "loss": 0.4294, "step": 103 }, { "epoch": 0.007146783947223749, "grad_norm": 5.716175556182861, "learning_rate": 7.1395881006865e-06, "loss": 0.4259, "step": 104 }, { "epoch": 0.007215503023639362, "grad_norm": 4.502131938934326, "learning_rate": 7.208237986270023e-06, "loss": 0.4547, "step": 105 }, { "epoch": 0.0072842221000549755, "grad_norm": 5.371185302734375, "learning_rate": 7.276887871853547e-06, "loss": 0.446, "step": 106 }, { "epoch": 0.007352941176470588, "grad_norm": 4.269709587097168, "learning_rate": 7.3455377574370706e-06, "loss": 0.4686, "step": 107 }, { "epoch": 0.007421660252886202, "grad_norm": 5.291435241699219, "learning_rate": 7.414187643020595e-06, "loss": 0.4558, "step": 108 }, { "epoch": 0.007490379329301814, "grad_norm": 6.569545745849609, "learning_rate": 7.4828375286041195e-06, "loss": 0.4787, "step": 109 }, { "epoch": 0.007559098405717427, "grad_norm": 4.770885467529297, "learning_rate": 7.551487414187644e-06, "loss": 0.4294, "step": 110 }, { "epoch": 0.00762781748213304, "grad_norm": 5.073994159698486, "learning_rate": 7.620137299771168e-06, "loss": 0.4091, "step": 111 }, { "epoch": 0.007696536558548653, "grad_norm": 5.031019687652588, "learning_rate": 7.688787185354691e-06, "loss": 0.3782, "step": 112 }, { "epoch": 0.0077652556349642665, "grad_norm": 4.142268657684326, "learning_rate": 7.757437070938216e-06, "loss": 0.3768, "step": 113 }, { "epoch": 0.007833974711379879, "grad_norm": 6.32844352722168, "learning_rate": 7.826086956521738e-06, "loss": 0.4364, "step": 114 }, { "epoch": 0.007902693787795493, "grad_norm": 4.701157093048096, "learning_rate": 7.894736842105263e-06, "loss": 0.5013, "step": 115 }, { "epoch": 0.007971412864211104, "grad_norm": 4.476530075073242, "learning_rate": 7.963386727688787e-06, "loss": 0.4181, "step": 116 }, { "epoch": 0.008040131940626718, "grad_norm": 4.925069332122803, "learning_rate": 8.032036613272312e-06, "loss": 0.4094, "step": 117 }, { "epoch": 0.008108851017042331, "grad_norm": 4.64376974105835, "learning_rate": 8.100686498855836e-06, "loss": 0.3424, "step": 118 }, { "epoch": 0.008177570093457943, "grad_norm": 4.97096586227417, "learning_rate": 8.169336384439359e-06, "loss": 0.3361, "step": 119 }, { "epoch": 0.008246289169873557, "grad_norm": 4.443885326385498, "learning_rate": 8.237986270022884e-06, "loss": 0.3827, "step": 120 }, { "epoch": 0.00831500824628917, "grad_norm": 7.205092906951904, "learning_rate": 8.306636155606408e-06, "loss": 0.3989, "step": 121 }, { "epoch": 0.008383727322704784, "grad_norm": 5.541567802429199, "learning_rate": 8.37528604118993e-06, "loss": 0.3877, "step": 122 }, { "epoch": 0.008452446399120395, "grad_norm": 6.409497261047363, "learning_rate": 8.443935926773455e-06, "loss": 0.4158, "step": 123 }, { "epoch": 0.008521165475536009, "grad_norm": 5.462257385253906, "learning_rate": 8.512585812356978e-06, "loss": 0.412, "step": 124 }, { "epoch": 0.008589884551951622, "grad_norm": 4.933533668518066, "learning_rate": 8.581235697940504e-06, "loss": 0.3363, "step": 125 }, { "epoch": 0.008658603628367234, "grad_norm": 4.670529365539551, "learning_rate": 8.649885583524029e-06, "loss": 0.3722, "step": 126 }, { "epoch": 0.008727322704782848, "grad_norm": 4.549845218658447, "learning_rate": 8.718535469107551e-06, "loss": 0.3924, "step": 127 }, { "epoch": 0.008796041781198461, "grad_norm": 5.2527241706848145, "learning_rate": 8.787185354691076e-06, "loss": 0.316, "step": 128 }, { "epoch": 0.008864760857614073, "grad_norm": 4.70778751373291, "learning_rate": 8.8558352402746e-06, "loss": 0.3291, "step": 129 }, { "epoch": 0.008933479934029686, "grad_norm": 4.60440731048584, "learning_rate": 8.924485125858123e-06, "loss": 0.3537, "step": 130 }, { "epoch": 0.0090021990104453, "grad_norm": 5.95958137512207, "learning_rate": 8.993135011441648e-06, "loss": 0.3744, "step": 131 }, { "epoch": 0.009070918086860913, "grad_norm": 5.67908239364624, "learning_rate": 9.061784897025172e-06, "loss": 0.3673, "step": 132 }, { "epoch": 0.009139637163276525, "grad_norm": 8.624402046203613, "learning_rate": 9.130434782608697e-06, "loss": 0.3027, "step": 133 }, { "epoch": 0.009208356239692139, "grad_norm": 7.802802562713623, "learning_rate": 9.199084668192221e-06, "loss": 0.377, "step": 134 }, { "epoch": 0.009277075316107752, "grad_norm": 4.969829559326172, "learning_rate": 9.267734553775744e-06, "loss": 0.3425, "step": 135 }, { "epoch": 0.009345794392523364, "grad_norm": 4.135549545288086, "learning_rate": 9.336384439359268e-06, "loss": 0.3612, "step": 136 }, { "epoch": 0.009414513468938977, "grad_norm": 4.647878646850586, "learning_rate": 9.405034324942791e-06, "loss": 0.3559, "step": 137 }, { "epoch": 0.00948323254535459, "grad_norm": 4.619668960571289, "learning_rate": 9.473684210526315e-06, "loss": 0.3513, "step": 138 }, { "epoch": 0.009551951621770204, "grad_norm": 4.204782962799072, "learning_rate": 9.54233409610984e-06, "loss": 0.3663, "step": 139 }, { "epoch": 0.009620670698185816, "grad_norm": 4.2179131507873535, "learning_rate": 9.610983981693364e-06, "loss": 0.3583, "step": 140 }, { "epoch": 0.00968938977460143, "grad_norm": 5.504446029663086, "learning_rate": 9.679633867276889e-06, "loss": 0.3341, "step": 141 }, { "epoch": 0.009758108851017043, "grad_norm": 4.647984504699707, "learning_rate": 9.748283752860412e-06, "loss": 0.3294, "step": 142 }, { "epoch": 0.009826827927432655, "grad_norm": 6.047488689422607, "learning_rate": 9.816933638443936e-06, "loss": 0.3303, "step": 143 }, { "epoch": 0.009895547003848268, "grad_norm": 3.937889337539673, "learning_rate": 9.88558352402746e-06, "loss": 0.3831, "step": 144 }, { "epoch": 0.009964266080263882, "grad_norm": 4.736089706420898, "learning_rate": 9.954233409610983e-06, "loss": 0.3175, "step": 145 }, { "epoch": 0.010032985156679494, "grad_norm": 6.074359893798828, "learning_rate": 1.0022883295194508e-05, "loss": 0.2883, "step": 146 }, { "epoch": 0.010101704233095107, "grad_norm": 5.576590538024902, "learning_rate": 1.0091533180778032e-05, "loss": 0.3316, "step": 147 }, { "epoch": 0.01017042330951072, "grad_norm": 5.424705982208252, "learning_rate": 1.0160183066361557e-05, "loss": 0.3098, "step": 148 }, { "epoch": 0.010239142385926334, "grad_norm": 5.922277450561523, "learning_rate": 1.0228832951945081e-05, "loss": 0.366, "step": 149 }, { "epoch": 0.010307861462341946, "grad_norm": 4.489944934844971, "learning_rate": 1.0297482837528604e-05, "loss": 0.3094, "step": 150 }, { "epoch": 0.01037658053875756, "grad_norm": 4.293323516845703, "learning_rate": 1.0366132723112128e-05, "loss": 0.2626, "step": 151 }, { "epoch": 0.010445299615173173, "grad_norm": 3.744903564453125, "learning_rate": 1.0434782608695653e-05, "loss": 0.2884, "step": 152 }, { "epoch": 0.010514018691588784, "grad_norm": 4.6085124015808105, "learning_rate": 1.0503432494279176e-05, "loss": 0.2647, "step": 153 }, { "epoch": 0.010582737768004398, "grad_norm": 7.769456386566162, "learning_rate": 1.05720823798627e-05, "loss": 0.2705, "step": 154 }, { "epoch": 0.010651456844420011, "grad_norm": 4.854430198669434, "learning_rate": 1.0640732265446225e-05, "loss": 0.2838, "step": 155 }, { "epoch": 0.010720175920835623, "grad_norm": 6.2141265869140625, "learning_rate": 1.0709382151029749e-05, "loss": 0.2664, "step": 156 }, { "epoch": 0.010788894997251237, "grad_norm": 5.014756202697754, "learning_rate": 1.0778032036613274e-05, "loss": 0.3266, "step": 157 }, { "epoch": 0.01085761407366685, "grad_norm": 4.540536403656006, "learning_rate": 1.0846681922196796e-05, "loss": 0.2731, "step": 158 }, { "epoch": 0.010926333150082464, "grad_norm": 4.751138687133789, "learning_rate": 1.091533180778032e-05, "loss": 0.2872, "step": 159 }, { "epoch": 0.010995052226498075, "grad_norm": 5.525317668914795, "learning_rate": 1.0983981693363844e-05, "loss": 0.3289, "step": 160 }, { "epoch": 0.011063771302913689, "grad_norm": 3.9240145683288574, "learning_rate": 1.1052631578947368e-05, "loss": 0.3179, "step": 161 }, { "epoch": 0.011132490379329302, "grad_norm": 3.0794317722320557, "learning_rate": 1.1121281464530894e-05, "loss": 0.2557, "step": 162 }, { "epoch": 0.011201209455744914, "grad_norm": 5.090035438537598, "learning_rate": 1.1189931350114417e-05, "loss": 0.2594, "step": 163 }, { "epoch": 0.011269928532160528, "grad_norm": 5.80757999420166, "learning_rate": 1.1258581235697941e-05, "loss": 0.3579, "step": 164 }, { "epoch": 0.011338647608576141, "grad_norm": 9.971646308898926, "learning_rate": 1.1327231121281464e-05, "loss": 0.3074, "step": 165 }, { "epoch": 0.011407366684991753, "grad_norm": 3.7188661098480225, "learning_rate": 1.1395881006864989e-05, "loss": 0.2241, "step": 166 }, { "epoch": 0.011476085761407366, "grad_norm": 4.959718704223633, "learning_rate": 1.1464530892448513e-05, "loss": 0.2888, "step": 167 }, { "epoch": 0.01154480483782298, "grad_norm": 4.163051605224609, "learning_rate": 1.1533180778032036e-05, "loss": 0.2563, "step": 168 }, { "epoch": 0.011613523914238593, "grad_norm": 6.607916831970215, "learning_rate": 1.160183066361556e-05, "loss": 0.3473, "step": 169 }, { "epoch": 0.011682242990654205, "grad_norm": 4.601958274841309, "learning_rate": 1.1670480549199085e-05, "loss": 0.2558, "step": 170 }, { "epoch": 0.011750962067069819, "grad_norm": 4.330945014953613, "learning_rate": 1.173913043478261e-05, "loss": 0.3264, "step": 171 }, { "epoch": 0.011819681143485432, "grad_norm": 4.170353412628174, "learning_rate": 1.1807780320366134e-05, "loss": 0.2868, "step": 172 }, { "epoch": 0.011888400219901044, "grad_norm": 4.433043479919434, "learning_rate": 1.1876430205949657e-05, "loss": 0.2643, "step": 173 }, { "epoch": 0.011957119296316657, "grad_norm": 4.7110819816589355, "learning_rate": 1.1945080091533181e-05, "loss": 0.261, "step": 174 }, { "epoch": 0.012025838372732271, "grad_norm": 4.481231689453125, "learning_rate": 1.2013729977116706e-05, "loss": 0.2898, "step": 175 }, { "epoch": 0.012094557449147883, "grad_norm": 3.560685634613037, "learning_rate": 1.2082379862700228e-05, "loss": 0.2934, "step": 176 }, { "epoch": 0.012163276525563496, "grad_norm": 5.381150722503662, "learning_rate": 1.2151029748283753e-05, "loss": 0.3157, "step": 177 }, { "epoch": 0.01223199560197911, "grad_norm": 3.897331714630127, "learning_rate": 1.2219679633867277e-05, "loss": 0.2674, "step": 178 }, { "epoch": 0.012300714678394723, "grad_norm": 3.892125368118286, "learning_rate": 1.2288329519450802e-05, "loss": 0.2254, "step": 179 }, { "epoch": 0.012369433754810335, "grad_norm": 3.4671366214752197, "learning_rate": 1.2356979405034326e-05, "loss": 0.2784, "step": 180 }, { "epoch": 0.012438152831225948, "grad_norm": 4.165935516357422, "learning_rate": 1.2425629290617849e-05, "loss": 0.2414, "step": 181 }, { "epoch": 0.012506871907641562, "grad_norm": 4.749786376953125, "learning_rate": 1.2494279176201373e-05, "loss": 0.2711, "step": 182 }, { "epoch": 0.012575590984057174, "grad_norm": 5.419155597686768, "learning_rate": 1.2562929061784896e-05, "loss": 0.2956, "step": 183 }, { "epoch": 0.012644310060472787, "grad_norm": 5.571883678436279, "learning_rate": 1.263157894736842e-05, "loss": 0.28, "step": 184 }, { "epoch": 0.0127130291368884, "grad_norm": 5.024911403656006, "learning_rate": 1.2700228832951947e-05, "loss": 0.2745, "step": 185 }, { "epoch": 0.012781748213304012, "grad_norm": 3.4903249740600586, "learning_rate": 1.276887871853547e-05, "loss": 0.2241, "step": 186 }, { "epoch": 0.012850467289719626, "grad_norm": 3.2027268409729004, "learning_rate": 1.2837528604118994e-05, "loss": 0.2551, "step": 187 }, { "epoch": 0.01291918636613524, "grad_norm": 4.806751728057861, "learning_rate": 1.2906178489702517e-05, "loss": 0.2453, "step": 188 }, { "epoch": 0.012987905442550853, "grad_norm": 3.5513253211975098, "learning_rate": 1.2974828375286041e-05, "loss": 0.2204, "step": 189 }, { "epoch": 0.013056624518966465, "grad_norm": 4.020075798034668, "learning_rate": 1.3043478260869566e-05, "loss": 0.2228, "step": 190 }, { "epoch": 0.013125343595382078, "grad_norm": 4.0980730056762695, "learning_rate": 1.3112128146453089e-05, "loss": 0.2908, "step": 191 }, { "epoch": 0.013194062671797692, "grad_norm": 4.120733261108398, "learning_rate": 1.3180778032036613e-05, "loss": 0.253, "step": 192 }, { "epoch": 0.013262781748213303, "grad_norm": 5.574907302856445, "learning_rate": 1.3249427917620137e-05, "loss": 0.2519, "step": 193 }, { "epoch": 0.013331500824628917, "grad_norm": 4.821225166320801, "learning_rate": 1.3318077803203662e-05, "loss": 0.2398, "step": 194 }, { "epoch": 0.01340021990104453, "grad_norm": 4.308213233947754, "learning_rate": 1.3386727688787186e-05, "loss": 0.2139, "step": 195 }, { "epoch": 0.013468938977460142, "grad_norm": 3.7964706420898438, "learning_rate": 1.345537757437071e-05, "loss": 0.2293, "step": 196 }, { "epoch": 0.013537658053875756, "grad_norm": 4.772131443023682, "learning_rate": 1.3524027459954234e-05, "loss": 0.2392, "step": 197 }, { "epoch": 0.013606377130291369, "grad_norm": 3.972473382949829, "learning_rate": 1.3592677345537758e-05, "loss": 0.242, "step": 198 }, { "epoch": 0.013675096206706983, "grad_norm": 3.724640130996704, "learning_rate": 1.3661327231121281e-05, "loss": 0.2524, "step": 199 }, { "epoch": 0.013743815283122594, "grad_norm": 4.239178657531738, "learning_rate": 1.3729977116704807e-05, "loss": 0.2312, "step": 200 }, { "epoch": 0.013812534359538208, "grad_norm": 4.77308988571167, "learning_rate": 1.379862700228833e-05, "loss": 0.2427, "step": 201 }, { "epoch": 0.013881253435953821, "grad_norm": 4.048278331756592, "learning_rate": 1.3867276887871854e-05, "loss": 0.2174, "step": 202 }, { "epoch": 0.013949972512369433, "grad_norm": 3.618288040161133, "learning_rate": 1.3935926773455379e-05, "loss": 0.2144, "step": 203 }, { "epoch": 0.014018691588785047, "grad_norm": 5.2844624519348145, "learning_rate": 1.4004576659038902e-05, "loss": 0.2228, "step": 204 }, { "epoch": 0.01408741066520066, "grad_norm": 3.976106882095337, "learning_rate": 1.4073226544622426e-05, "loss": 0.2778, "step": 205 }, { "epoch": 0.014156129741616274, "grad_norm": 4.2854323387146, "learning_rate": 1.4141876430205949e-05, "loss": 0.2344, "step": 206 }, { "epoch": 0.014224848818031885, "grad_norm": 3.363224983215332, "learning_rate": 1.4210526315789473e-05, "loss": 0.1828, "step": 207 }, { "epoch": 0.014293567894447499, "grad_norm": 3.754612684249878, "learning_rate": 1.4279176201373e-05, "loss": 0.2235, "step": 208 }, { "epoch": 0.014362286970863112, "grad_norm": 8.145918846130371, "learning_rate": 1.4347826086956522e-05, "loss": 0.2175, "step": 209 }, { "epoch": 0.014431006047278724, "grad_norm": 5.262914657592773, "learning_rate": 1.4416475972540047e-05, "loss": 0.2488, "step": 210 }, { "epoch": 0.014499725123694338, "grad_norm": 4.774180889129639, "learning_rate": 1.448512585812357e-05, "loss": 0.229, "step": 211 }, { "epoch": 0.014568444200109951, "grad_norm": 5.528488636016846, "learning_rate": 1.4553775743707094e-05, "loss": 0.2142, "step": 212 }, { "epoch": 0.014637163276525563, "grad_norm": 4.704835891723633, "learning_rate": 1.4622425629290618e-05, "loss": 0.2557, "step": 213 }, { "epoch": 0.014705882352941176, "grad_norm": 3.362499237060547, "learning_rate": 1.4691075514874141e-05, "loss": 0.2373, "step": 214 }, { "epoch": 0.01477460142935679, "grad_norm": 5.358570575714111, "learning_rate": 1.4759725400457667e-05, "loss": 0.2501, "step": 215 }, { "epoch": 0.014843320505772403, "grad_norm": 3.8891983032226562, "learning_rate": 1.482837528604119e-05, "loss": 0.2208, "step": 216 }, { "epoch": 0.014912039582188015, "grad_norm": 4.194068908691406, "learning_rate": 1.4897025171624715e-05, "loss": 0.2216, "step": 217 }, { "epoch": 0.014980758658603628, "grad_norm": 4.894361972808838, "learning_rate": 1.4965675057208239e-05, "loss": 0.2, "step": 218 }, { "epoch": 0.015049477735019242, "grad_norm": 4.057684421539307, "learning_rate": 1.5034324942791763e-05, "loss": 0.2441, "step": 219 }, { "epoch": 0.015118196811434854, "grad_norm": 3.7686522006988525, "learning_rate": 1.5102974828375288e-05, "loss": 0.2534, "step": 220 }, { "epoch": 0.015186915887850467, "grad_norm": 4.611441135406494, "learning_rate": 1.5171624713958812e-05, "loss": 0.2083, "step": 221 }, { "epoch": 0.01525563496426608, "grad_norm": 3.7034616470336914, "learning_rate": 1.5240274599542335e-05, "loss": 0.2203, "step": 222 }, { "epoch": 0.015324354040681692, "grad_norm": 4.239235877990723, "learning_rate": 1.530892448512586e-05, "loss": 0.2452, "step": 223 }, { "epoch": 0.015393073117097306, "grad_norm": 3.68691349029541, "learning_rate": 1.5377574370709382e-05, "loss": 0.1927, "step": 224 }, { "epoch": 0.01546179219351292, "grad_norm": 3.591973066329956, "learning_rate": 1.5446224256292905e-05, "loss": 0.1876, "step": 225 }, { "epoch": 0.015530511269928533, "grad_norm": 3.505387544631958, "learning_rate": 1.551487414187643e-05, "loss": 0.211, "step": 226 }, { "epoch": 0.015599230346344145, "grad_norm": 4.52597713470459, "learning_rate": 1.5583524027459954e-05, "loss": 0.2467, "step": 227 }, { "epoch": 0.015667949422759758, "grad_norm": 3.919370174407959, "learning_rate": 1.5652173913043477e-05, "loss": 0.2123, "step": 228 }, { "epoch": 0.01573666849917537, "grad_norm": 3.979773759841919, "learning_rate": 1.5720823798627003e-05, "loss": 0.1804, "step": 229 }, { "epoch": 0.015805387575590985, "grad_norm": 3.536041736602783, "learning_rate": 1.5789473684210526e-05, "loss": 0.1859, "step": 230 }, { "epoch": 0.015874106652006597, "grad_norm": 3.296677827835083, "learning_rate": 1.585812356979405e-05, "loss": 0.2115, "step": 231 }, { "epoch": 0.01594282572842221, "grad_norm": 4.509555816650391, "learning_rate": 1.5926773455377575e-05, "loss": 0.1827, "step": 232 }, { "epoch": 0.016011544804837824, "grad_norm": 4.74887752532959, "learning_rate": 1.5995423340961098e-05, "loss": 0.2169, "step": 233 }, { "epoch": 0.016080263881253436, "grad_norm": 3.0517148971557617, "learning_rate": 1.6064073226544624e-05, "loss": 0.1873, "step": 234 }, { "epoch": 0.016148982957669047, "grad_norm": 4.007606029510498, "learning_rate": 1.613272311212815e-05, "loss": 0.2114, "step": 235 }, { "epoch": 0.016217702034084663, "grad_norm": 5.375392913818359, "learning_rate": 1.6201372997711673e-05, "loss": 0.2162, "step": 236 }, { "epoch": 0.016286421110500274, "grad_norm": 4.902636528015137, "learning_rate": 1.6270022883295195e-05, "loss": 0.2397, "step": 237 }, { "epoch": 0.016355140186915886, "grad_norm": 4.440280914306641, "learning_rate": 1.6338672768878718e-05, "loss": 0.2376, "step": 238 }, { "epoch": 0.0164238592633315, "grad_norm": 4.285326957702637, "learning_rate": 1.6407322654462244e-05, "loss": 0.2217, "step": 239 }, { "epoch": 0.016492578339747113, "grad_norm": 4.510012149810791, "learning_rate": 1.6475972540045767e-05, "loss": 0.2038, "step": 240 }, { "epoch": 0.01656129741616273, "grad_norm": 4.176109313964844, "learning_rate": 1.654462242562929e-05, "loss": 0.2809, "step": 241 }, { "epoch": 0.01663001649257834, "grad_norm": 3.514003038406372, "learning_rate": 1.6613272311212816e-05, "loss": 0.2028, "step": 242 }, { "epoch": 0.016698735568993952, "grad_norm": 5.285548686981201, "learning_rate": 1.668192219679634e-05, "loss": 0.2208, "step": 243 }, { "epoch": 0.016767454645409567, "grad_norm": 3.4300405979156494, "learning_rate": 1.675057208237986e-05, "loss": 0.1965, "step": 244 }, { "epoch": 0.01683617372182518, "grad_norm": 2.6903316974639893, "learning_rate": 1.6819221967963388e-05, "loss": 0.1911, "step": 245 }, { "epoch": 0.01690489279824079, "grad_norm": 4.645479679107666, "learning_rate": 1.688787185354691e-05, "loss": 0.2025, "step": 246 }, { "epoch": 0.016973611874656406, "grad_norm": 3.831782102584839, "learning_rate": 1.6956521739130433e-05, "loss": 0.1844, "step": 247 }, { "epoch": 0.017042330951072018, "grad_norm": 3.269639253616333, "learning_rate": 1.7025171624713956e-05, "loss": 0.1887, "step": 248 }, { "epoch": 0.01711105002748763, "grad_norm": 3.828476667404175, "learning_rate": 1.7093821510297486e-05, "loss": 0.1855, "step": 249 }, { "epoch": 0.017179769103903245, "grad_norm": 3.771472930908203, "learning_rate": 1.716247139588101e-05, "loss": 0.1957, "step": 250 }, { "epoch": 0.017248488180318856, "grad_norm": 4.260629177093506, "learning_rate": 1.723112128146453e-05, "loss": 0.2103, "step": 251 }, { "epoch": 0.017317207256734468, "grad_norm": 4.9855146408081055, "learning_rate": 1.7299771167048057e-05, "loss": 0.2207, "step": 252 }, { "epoch": 0.017385926333150083, "grad_norm": 4.666804313659668, "learning_rate": 1.736842105263158e-05, "loss": 0.2034, "step": 253 }, { "epoch": 0.017454645409565695, "grad_norm": 6.7025628089904785, "learning_rate": 1.7437070938215103e-05, "loss": 0.2148, "step": 254 }, { "epoch": 0.017523364485981307, "grad_norm": 3.453289270401001, "learning_rate": 1.750572082379863e-05, "loss": 0.2098, "step": 255 }, { "epoch": 0.017592083562396922, "grad_norm": 3.9793319702148438, "learning_rate": 1.7574370709382152e-05, "loss": 0.193, "step": 256 }, { "epoch": 0.017660802638812534, "grad_norm": 4.860001087188721, "learning_rate": 1.7643020594965675e-05, "loss": 0.2347, "step": 257 }, { "epoch": 0.017729521715228146, "grad_norm": 3.2719931602478027, "learning_rate": 1.77116704805492e-05, "loss": 0.2227, "step": 258 }, { "epoch": 0.01779824079164376, "grad_norm": 4.2679266929626465, "learning_rate": 1.7780320366132724e-05, "loss": 0.2117, "step": 259 }, { "epoch": 0.017866959868059373, "grad_norm": 3.8715946674346924, "learning_rate": 1.7848970251716246e-05, "loss": 0.1807, "step": 260 }, { "epoch": 0.017935678944474988, "grad_norm": 3.707181215286255, "learning_rate": 1.791762013729977e-05, "loss": 0.2158, "step": 261 }, { "epoch": 0.0180043980208906, "grad_norm": 3.629993438720703, "learning_rate": 1.7986270022883295e-05, "loss": 0.1974, "step": 262 }, { "epoch": 0.01807311709730621, "grad_norm": 2.8252458572387695, "learning_rate": 1.8054919908466818e-05, "loss": 0.1921, "step": 263 }, { "epoch": 0.018141836173721827, "grad_norm": 4.734744071960449, "learning_rate": 1.8123569794050344e-05, "loss": 0.1973, "step": 264 }, { "epoch": 0.01821055525013744, "grad_norm": 4.283486843109131, "learning_rate": 1.819221967963387e-05, "loss": 0.1935, "step": 265 }, { "epoch": 0.01827927432655305, "grad_norm": 5.084074020385742, "learning_rate": 1.8260869565217393e-05, "loss": 0.2007, "step": 266 }, { "epoch": 0.018347993402968665, "grad_norm": 5.0472092628479, "learning_rate": 1.8329519450800916e-05, "loss": 0.2076, "step": 267 }, { "epoch": 0.018416712479384277, "grad_norm": 4.071624755859375, "learning_rate": 1.8398169336384442e-05, "loss": 0.1794, "step": 268 }, { "epoch": 0.01848543155579989, "grad_norm": 5.0371527671813965, "learning_rate": 1.8466819221967965e-05, "loss": 0.2222, "step": 269 }, { "epoch": 0.018554150632215504, "grad_norm": 3.946063756942749, "learning_rate": 1.8535469107551488e-05, "loss": 0.2074, "step": 270 }, { "epoch": 0.018622869708631116, "grad_norm": 7.734065055847168, "learning_rate": 1.860411899313501e-05, "loss": 0.1813, "step": 271 }, { "epoch": 0.018691588785046728, "grad_norm": 5.364829063415527, "learning_rate": 1.8672768878718537e-05, "loss": 0.2193, "step": 272 }, { "epoch": 0.018760307861462343, "grad_norm": 3.095437526702881, "learning_rate": 1.874141876430206e-05, "loss": 0.1993, "step": 273 }, { "epoch": 0.018829026937877955, "grad_norm": 3.7303223609924316, "learning_rate": 1.8810068649885582e-05, "loss": 0.1645, "step": 274 }, { "epoch": 0.018897746014293566, "grad_norm": 6.917871475219727, "learning_rate": 1.8878718535469108e-05, "loss": 0.2, "step": 275 }, { "epoch": 0.01896646509070918, "grad_norm": 3.1666834354400635, "learning_rate": 1.894736842105263e-05, "loss": 0.177, "step": 276 }, { "epoch": 0.019035184167124793, "grad_norm": 5.193768501281738, "learning_rate": 1.9016018306636154e-05, "loss": 0.1939, "step": 277 }, { "epoch": 0.01910390324354041, "grad_norm": 3.894078016281128, "learning_rate": 1.908466819221968e-05, "loss": 0.1647, "step": 278 }, { "epoch": 0.01917262231995602, "grad_norm": 5.489270210266113, "learning_rate": 1.9153318077803206e-05, "loss": 0.2004, "step": 279 }, { "epoch": 0.019241341396371632, "grad_norm": 4.1796674728393555, "learning_rate": 1.922196796338673e-05, "loss": 0.229, "step": 280 }, { "epoch": 0.019310060472787247, "grad_norm": 4.52797269821167, "learning_rate": 1.9290617848970255e-05, "loss": 0.2074, "step": 281 }, { "epoch": 0.01937877954920286, "grad_norm": 4.283642768859863, "learning_rate": 1.9359267734553778e-05, "loss": 0.2684, "step": 282 }, { "epoch": 0.01944749862561847, "grad_norm": 3.2766826152801514, "learning_rate": 1.94279176201373e-05, "loss": 0.1921, "step": 283 }, { "epoch": 0.019516217702034086, "grad_norm": 5.192159652709961, "learning_rate": 1.9496567505720823e-05, "loss": 0.248, "step": 284 }, { "epoch": 0.019584936778449698, "grad_norm": 3.6420633792877197, "learning_rate": 1.956521739130435e-05, "loss": 0.2097, "step": 285 }, { "epoch": 0.01965365585486531, "grad_norm": 4.0534563064575195, "learning_rate": 1.9633867276887872e-05, "loss": 0.2182, "step": 286 }, { "epoch": 0.019722374931280925, "grad_norm": 3.570767879486084, "learning_rate": 1.9702517162471395e-05, "loss": 0.1619, "step": 287 }, { "epoch": 0.019791094007696537, "grad_norm": 3.0883734226226807, "learning_rate": 1.977116704805492e-05, "loss": 0.1723, "step": 288 }, { "epoch": 0.01985981308411215, "grad_norm": 3.477835178375244, "learning_rate": 1.9839816933638444e-05, "loss": 0.1795, "step": 289 }, { "epoch": 0.019928532160527763, "grad_norm": 3.796935558319092, "learning_rate": 1.9908466819221967e-05, "loss": 0.1964, "step": 290 }, { "epoch": 0.019997251236943375, "grad_norm": 3.094614267349243, "learning_rate": 1.9977116704805493e-05, "loss": 0.1879, "step": 291 }, { "epoch": 0.020065970313358987, "grad_norm": 3.3940539360046387, "learning_rate": 2.0045766590389016e-05, "loss": 0.1803, "step": 292 }, { "epoch": 0.020134689389774602, "grad_norm": 4.116355895996094, "learning_rate": 2.011441647597254e-05, "loss": 0.1738, "step": 293 }, { "epoch": 0.020203408466190214, "grad_norm": 3.9578816890716553, "learning_rate": 2.0183066361556065e-05, "loss": 0.1849, "step": 294 }, { "epoch": 0.020272127542605826, "grad_norm": 3.618885040283203, "learning_rate": 2.025171624713959e-05, "loss": 0.2138, "step": 295 }, { "epoch": 0.02034084661902144, "grad_norm": 4.058716297149658, "learning_rate": 2.0320366132723114e-05, "loss": 0.1805, "step": 296 }, { "epoch": 0.020409565695437053, "grad_norm": 4.8649983406066895, "learning_rate": 2.0389016018306636e-05, "loss": 0.1971, "step": 297 }, { "epoch": 0.020478284771852668, "grad_norm": 3.9856503009796143, "learning_rate": 2.0457665903890162e-05, "loss": 0.1375, "step": 298 }, { "epoch": 0.02054700384826828, "grad_norm": 4.298996448516846, "learning_rate": 2.0526315789473685e-05, "loss": 0.1942, "step": 299 }, { "epoch": 0.02061572292468389, "grad_norm": 4.277236461639404, "learning_rate": 2.0594965675057208e-05, "loss": 0.2241, "step": 300 }, { "epoch": 0.020684442001099507, "grad_norm": 4.343676567077637, "learning_rate": 2.0663615560640734e-05, "loss": 0.1757, "step": 301 }, { "epoch": 0.02075316107751512, "grad_norm": 3.2907192707061768, "learning_rate": 2.0732265446224257e-05, "loss": 0.1822, "step": 302 }, { "epoch": 0.02082188015393073, "grad_norm": 4.269043922424316, "learning_rate": 2.080091533180778e-05, "loss": 0.1797, "step": 303 }, { "epoch": 0.020890599230346345, "grad_norm": 4.160331726074219, "learning_rate": 2.0869565217391306e-05, "loss": 0.1617, "step": 304 }, { "epoch": 0.020959318306761957, "grad_norm": 4.29764461517334, "learning_rate": 2.093821510297483e-05, "loss": 0.1718, "step": 305 }, { "epoch": 0.02102803738317757, "grad_norm": 3.4751100540161133, "learning_rate": 2.100686498855835e-05, "loss": 0.1608, "step": 306 }, { "epoch": 0.021096756459593184, "grad_norm": 3.1855204105377197, "learning_rate": 2.1075514874141874e-05, "loss": 0.1817, "step": 307 }, { "epoch": 0.021165475536008796, "grad_norm": 3.661719799041748, "learning_rate": 2.11441647597254e-05, "loss": 0.1582, "step": 308 }, { "epoch": 0.021234194612424408, "grad_norm": 3.0187528133392334, "learning_rate": 2.1212814645308927e-05, "loss": 0.1438, "step": 309 }, { "epoch": 0.021302913688840023, "grad_norm": 3.4782466888427734, "learning_rate": 2.128146453089245e-05, "loss": 0.1751, "step": 310 }, { "epoch": 0.021371632765255635, "grad_norm": 4.296860218048096, "learning_rate": 2.1350114416475975e-05, "loss": 0.1827, "step": 311 }, { "epoch": 0.021440351841671246, "grad_norm": 3.085280656814575, "learning_rate": 2.1418764302059498e-05, "loss": 0.1555, "step": 312 }, { "epoch": 0.02150907091808686, "grad_norm": 6.193214416503906, "learning_rate": 2.148741418764302e-05, "loss": 0.2269, "step": 313 }, { "epoch": 0.021577789994502473, "grad_norm": 3.9016873836517334, "learning_rate": 2.1556064073226547e-05, "loss": 0.2047, "step": 314 }, { "epoch": 0.021646509070918085, "grad_norm": 4.168028831481934, "learning_rate": 2.162471395881007e-05, "loss": 0.1694, "step": 315 }, { "epoch": 0.0217152281473337, "grad_norm": 2.770247220993042, "learning_rate": 2.1693363844393593e-05, "loss": 0.1699, "step": 316 }, { "epoch": 0.021783947223749312, "grad_norm": 4.669060230255127, "learning_rate": 2.1762013729977116e-05, "loss": 0.184, "step": 317 }, { "epoch": 0.021852666300164927, "grad_norm": 4.542727947235107, "learning_rate": 2.183066361556064e-05, "loss": 0.1891, "step": 318 }, { "epoch": 0.02192138537658054, "grad_norm": 3.136165142059326, "learning_rate": 2.1899313501144164e-05, "loss": 0.1835, "step": 319 }, { "epoch": 0.02199010445299615, "grad_norm": 3.159848928451538, "learning_rate": 2.1967963386727687e-05, "loss": 0.1667, "step": 320 }, { "epoch": 0.022058823529411766, "grad_norm": 3.597182035446167, "learning_rate": 2.2036613272311213e-05, "loss": 0.1803, "step": 321 }, { "epoch": 0.022127542605827378, "grad_norm": 2.8625168800354004, "learning_rate": 2.2105263157894736e-05, "loss": 0.1571, "step": 322 }, { "epoch": 0.02219626168224299, "grad_norm": 4.466572284698486, "learning_rate": 2.217391304347826e-05, "loss": 0.1708, "step": 323 }, { "epoch": 0.022264980758658605, "grad_norm": 6.211597442626953, "learning_rate": 2.224256292906179e-05, "loss": 0.1989, "step": 324 }, { "epoch": 0.022333699835074217, "grad_norm": 3.3698079586029053, "learning_rate": 2.231121281464531e-05, "loss": 0.1756, "step": 325 }, { "epoch": 0.02240241891148983, "grad_norm": 2.9026718139648438, "learning_rate": 2.2379862700228834e-05, "loss": 0.1708, "step": 326 }, { "epoch": 0.022471137987905444, "grad_norm": 3.5659496784210205, "learning_rate": 2.244851258581236e-05, "loss": 0.1807, "step": 327 }, { "epoch": 0.022539857064321055, "grad_norm": 3.602508306503296, "learning_rate": 2.2517162471395883e-05, "loss": 0.1842, "step": 328 }, { "epoch": 0.022608576140736667, "grad_norm": 4.07927942276001, "learning_rate": 2.2585812356979406e-05, "loss": 0.2296, "step": 329 }, { "epoch": 0.022677295217152282, "grad_norm": 4.013204574584961, "learning_rate": 2.265446224256293e-05, "loss": 0.1613, "step": 330 }, { "epoch": 0.022746014293567894, "grad_norm": 4.45927619934082, "learning_rate": 2.2723112128146455e-05, "loss": 0.1605, "step": 331 }, { "epoch": 0.022814733369983506, "grad_norm": 3.6388678550720215, "learning_rate": 2.2791762013729977e-05, "loss": 0.2146, "step": 332 }, { "epoch": 0.02288345244639912, "grad_norm": 3.897397518157959, "learning_rate": 2.28604118993135e-05, "loss": 0.1966, "step": 333 }, { "epoch": 0.022952171522814733, "grad_norm": 3.903005361557007, "learning_rate": 2.2929061784897026e-05, "loss": 0.1735, "step": 334 }, { "epoch": 0.023020890599230345, "grad_norm": 3.3002736568450928, "learning_rate": 2.299771167048055e-05, "loss": 0.1666, "step": 335 }, { "epoch": 0.02308960967564596, "grad_norm": 4.10645055770874, "learning_rate": 2.3066361556064072e-05, "loss": 0.1784, "step": 336 }, { "epoch": 0.02315832875206157, "grad_norm": 3.841790199279785, "learning_rate": 2.3135011441647598e-05, "loss": 0.1414, "step": 337 }, { "epoch": 0.023227047828477187, "grad_norm": 3.226128101348877, "learning_rate": 2.320366132723112e-05, "loss": 0.1556, "step": 338 }, { "epoch": 0.0232957669048928, "grad_norm": 6.0019683837890625, "learning_rate": 2.3272311212814644e-05, "loss": 0.1728, "step": 339 }, { "epoch": 0.02336448598130841, "grad_norm": 6.635164260864258, "learning_rate": 2.334096109839817e-05, "loss": 0.1704, "step": 340 }, { "epoch": 0.023433205057724026, "grad_norm": 5.553908348083496, "learning_rate": 2.3409610983981696e-05, "loss": 0.1671, "step": 341 }, { "epoch": 0.023501924134139637, "grad_norm": 3.3888118267059326, "learning_rate": 2.347826086956522e-05, "loss": 0.1709, "step": 342 }, { "epoch": 0.02357064321055525, "grad_norm": 3.693002462387085, "learning_rate": 2.354691075514874e-05, "loss": 0.1924, "step": 343 }, { "epoch": 0.023639362286970864, "grad_norm": 3.499500036239624, "learning_rate": 2.3615560640732268e-05, "loss": 0.1961, "step": 344 }, { "epoch": 0.023708081363386476, "grad_norm": 3.2344932556152344, "learning_rate": 2.368421052631579e-05, "loss": 0.1257, "step": 345 }, { "epoch": 0.023776800439802088, "grad_norm": 5.107495307922363, "learning_rate": 2.3752860411899313e-05, "loss": 0.2397, "step": 346 }, { "epoch": 0.023845519516217703, "grad_norm": 4.3138298988342285, "learning_rate": 2.382151029748284e-05, "loss": 0.1799, "step": 347 }, { "epoch": 0.023914238592633315, "grad_norm": 3.840930223464966, "learning_rate": 2.3890160183066362e-05, "loss": 0.2368, "step": 348 }, { "epoch": 0.023982957669048927, "grad_norm": 3.8338029384613037, "learning_rate": 2.3958810068649885e-05, "loss": 0.173, "step": 349 }, { "epoch": 0.024051676745464542, "grad_norm": 3.5507044792175293, "learning_rate": 2.402745995423341e-05, "loss": 0.1945, "step": 350 }, { "epoch": 0.024120395821880154, "grad_norm": 3.469724178314209, "learning_rate": 2.4096109839816934e-05, "loss": 0.167, "step": 351 }, { "epoch": 0.024189114898295765, "grad_norm": 3.8239405155181885, "learning_rate": 2.4164759725400457e-05, "loss": 0.1802, "step": 352 }, { "epoch": 0.02425783397471138, "grad_norm": 4.300062656402588, "learning_rate": 2.423340961098398e-05, "loss": 0.1661, "step": 353 }, { "epoch": 0.024326553051126992, "grad_norm": 3.623924970626831, "learning_rate": 2.4302059496567506e-05, "loss": 0.1743, "step": 354 }, { "epoch": 0.024395272127542608, "grad_norm": 4.119725704193115, "learning_rate": 2.4370709382151032e-05, "loss": 0.1893, "step": 355 }, { "epoch": 0.02446399120395822, "grad_norm": 3.5345511436462402, "learning_rate": 2.4439359267734554e-05, "loss": 0.1624, "step": 356 }, { "epoch": 0.02453271028037383, "grad_norm": 4.097344398498535, "learning_rate": 2.450800915331808e-05, "loss": 0.1554, "step": 357 }, { "epoch": 0.024601429356789446, "grad_norm": 3.5665903091430664, "learning_rate": 2.4576659038901603e-05, "loss": 0.1686, "step": 358 }, { "epoch": 0.024670148433205058, "grad_norm": 5.310074806213379, "learning_rate": 2.4645308924485126e-05, "loss": 0.1686, "step": 359 }, { "epoch": 0.02473886750962067, "grad_norm": 4.585694313049316, "learning_rate": 2.4713958810068652e-05, "loss": 0.2107, "step": 360 }, { "epoch": 0.024807586586036285, "grad_norm": 3.1829283237457275, "learning_rate": 2.4782608695652175e-05, "loss": 0.1504, "step": 361 }, { "epoch": 0.024876305662451897, "grad_norm": 3.389416456222534, "learning_rate": 2.4851258581235698e-05, "loss": 0.1532, "step": 362 }, { "epoch": 0.02494502473886751, "grad_norm": 3.5007097721099854, "learning_rate": 2.491990846681922e-05, "loss": 0.1809, "step": 363 }, { "epoch": 0.025013743815283124, "grad_norm": 3.2445242404937744, "learning_rate": 2.4988558352402747e-05, "loss": 0.1479, "step": 364 }, { "epoch": 0.025082462891698735, "grad_norm": 4.312865257263184, "learning_rate": 2.505720823798627e-05, "loss": 0.1817, "step": 365 }, { "epoch": 0.025151181968114347, "grad_norm": 2.998418092727661, "learning_rate": 2.5125858123569792e-05, "loss": 0.1959, "step": 366 }, { "epoch": 0.025219901044529962, "grad_norm": 3.524000883102417, "learning_rate": 2.519450800915332e-05, "loss": 0.1845, "step": 367 }, { "epoch": 0.025288620120945574, "grad_norm": 2.843614101409912, "learning_rate": 2.526315789473684e-05, "loss": 0.1543, "step": 368 }, { "epoch": 0.025357339197361186, "grad_norm": 3.5980215072631836, "learning_rate": 2.5331807780320364e-05, "loss": 0.1639, "step": 369 }, { "epoch": 0.0254260582737768, "grad_norm": 4.278166770935059, "learning_rate": 2.5400457665903894e-05, "loss": 0.1853, "step": 370 }, { "epoch": 0.025494777350192413, "grad_norm": 4.016077041625977, "learning_rate": 2.5469107551487416e-05, "loss": 0.1626, "step": 371 }, { "epoch": 0.025563496426608025, "grad_norm": 3.7850985527038574, "learning_rate": 2.553775743707094e-05, "loss": 0.161, "step": 372 }, { "epoch": 0.02563221550302364, "grad_norm": 4.203721523284912, "learning_rate": 2.5606407322654465e-05, "loss": 0.1831, "step": 373 }, { "epoch": 0.02570093457943925, "grad_norm": 4.406505584716797, "learning_rate": 2.5675057208237988e-05, "loss": 0.1485, "step": 374 }, { "epoch": 0.025769653655854867, "grad_norm": 4.406871318817139, "learning_rate": 2.574370709382151e-05, "loss": 0.1583, "step": 375 }, { "epoch": 0.02583837273227048, "grad_norm": 3.555941343307495, "learning_rate": 2.5812356979405034e-05, "loss": 0.1525, "step": 376 }, { "epoch": 0.02590709180868609, "grad_norm": 3.7753219604492188, "learning_rate": 2.588100686498856e-05, "loss": 0.1535, "step": 377 }, { "epoch": 0.025975810885101706, "grad_norm": 4.202167987823486, "learning_rate": 2.5949656750572083e-05, "loss": 0.1894, "step": 378 }, { "epoch": 0.026044529961517317, "grad_norm": 6.254127502441406, "learning_rate": 2.6018306636155605e-05, "loss": 0.1399, "step": 379 }, { "epoch": 0.02611324903793293, "grad_norm": 4.170585632324219, "learning_rate": 2.608695652173913e-05, "loss": 0.1523, "step": 380 }, { "epoch": 0.026181968114348544, "grad_norm": 4.045107841491699, "learning_rate": 2.6155606407322654e-05, "loss": 0.1691, "step": 381 }, { "epoch": 0.026250687190764156, "grad_norm": 3.3169000148773193, "learning_rate": 2.6224256292906177e-05, "loss": 0.2195, "step": 382 }, { "epoch": 0.026319406267179768, "grad_norm": 3.5787882804870605, "learning_rate": 2.6292906178489703e-05, "loss": 0.1971, "step": 383 }, { "epoch": 0.026388125343595383, "grad_norm": 4.870965003967285, "learning_rate": 2.6361556064073226e-05, "loss": 0.1767, "step": 384 }, { "epoch": 0.026456844420010995, "grad_norm": 4.425243854522705, "learning_rate": 2.6430205949656752e-05, "loss": 0.1496, "step": 385 }, { "epoch": 0.026525563496426607, "grad_norm": 4.6036882400512695, "learning_rate": 2.6498855835240275e-05, "loss": 0.1793, "step": 386 }, { "epoch": 0.026594282572842222, "grad_norm": 2.980076551437378, "learning_rate": 2.65675057208238e-05, "loss": 0.1147, "step": 387 }, { "epoch": 0.026663001649257834, "grad_norm": 2.4807305335998535, "learning_rate": 2.6636155606407324e-05, "loss": 0.1366, "step": 388 }, { "epoch": 0.026731720725673445, "grad_norm": 3.080758571624756, "learning_rate": 2.6704805491990847e-05, "loss": 0.1573, "step": 389 }, { "epoch": 0.02680043980208906, "grad_norm": 3.748814344406128, "learning_rate": 2.6773455377574373e-05, "loss": 0.1329, "step": 390 }, { "epoch": 0.026869158878504672, "grad_norm": 3.0189850330352783, "learning_rate": 2.6842105263157896e-05, "loss": 0.1648, "step": 391 }, { "epoch": 0.026937877954920284, "grad_norm": 3.810770273208618, "learning_rate": 2.691075514874142e-05, "loss": 0.1493, "step": 392 }, { "epoch": 0.0270065970313359, "grad_norm": 3.513643503189087, "learning_rate": 2.6979405034324945e-05, "loss": 0.1383, "step": 393 }, { "epoch": 0.02707531610775151, "grad_norm": 2.9639782905578613, "learning_rate": 2.7048054919908467e-05, "loss": 0.1538, "step": 394 }, { "epoch": 0.027144035184167126, "grad_norm": 3.8626151084899902, "learning_rate": 2.711670480549199e-05, "loss": 0.1511, "step": 395 }, { "epoch": 0.027212754260582738, "grad_norm": 4.0711350440979, "learning_rate": 2.7185354691075516e-05, "loss": 0.1591, "step": 396 }, { "epoch": 0.02728147333699835, "grad_norm": 3.1105058193206787, "learning_rate": 2.725400457665904e-05, "loss": 0.1365, "step": 397 }, { "epoch": 0.027350192413413965, "grad_norm": 5.154309272766113, "learning_rate": 2.7322654462242562e-05, "loss": 0.2201, "step": 398 }, { "epoch": 0.027418911489829577, "grad_norm": 3.363769292831421, "learning_rate": 2.7391304347826085e-05, "loss": 0.157, "step": 399 }, { "epoch": 0.02748763056624519, "grad_norm": 3.6057794094085693, "learning_rate": 2.7459954233409614e-05, "loss": 0.1505, "step": 400 }, { "epoch": 0.027556349642660804, "grad_norm": 4.570315361022949, "learning_rate": 2.7528604118993137e-05, "loss": 0.185, "step": 401 }, { "epoch": 0.027625068719076416, "grad_norm": 4.205750942230225, "learning_rate": 2.759725400457666e-05, "loss": 0.1679, "step": 402 }, { "epoch": 0.027693787795492027, "grad_norm": 3.3512911796569824, "learning_rate": 2.7665903890160186e-05, "loss": 0.1689, "step": 403 }, { "epoch": 0.027762506871907643, "grad_norm": 3.0865864753723145, "learning_rate": 2.773455377574371e-05, "loss": 0.1717, "step": 404 }, { "epoch": 0.027831225948323254, "grad_norm": 3.4332940578460693, "learning_rate": 2.780320366132723e-05, "loss": 0.191, "step": 405 }, { "epoch": 0.027899945024738866, "grad_norm": 3.444995164871216, "learning_rate": 2.7871853546910758e-05, "loss": 0.1559, "step": 406 }, { "epoch": 0.02796866410115448, "grad_norm": 3.5106561183929443, "learning_rate": 2.794050343249428e-05, "loss": 0.1722, "step": 407 }, { "epoch": 0.028037383177570093, "grad_norm": 4.515082836151123, "learning_rate": 2.8009153318077803e-05, "loss": 0.1647, "step": 408 }, { "epoch": 0.028106102253985705, "grad_norm": 4.128320693969727, "learning_rate": 2.8077803203661326e-05, "loss": 0.167, "step": 409 }, { "epoch": 0.02817482133040132, "grad_norm": 3.992288589477539, "learning_rate": 2.8146453089244852e-05, "loss": 0.1396, "step": 410 }, { "epoch": 0.028243540406816932, "grad_norm": 3.5627903938293457, "learning_rate": 2.8215102974828375e-05, "loss": 0.1773, "step": 411 }, { "epoch": 0.028312259483232547, "grad_norm": 3.379153251647949, "learning_rate": 2.8283752860411898e-05, "loss": 0.1589, "step": 412 }, { "epoch": 0.02838097855964816, "grad_norm": 3.5480029582977295, "learning_rate": 2.8352402745995424e-05, "loss": 0.1659, "step": 413 }, { "epoch": 0.02844969763606377, "grad_norm": 3.425229072570801, "learning_rate": 2.8421052631578946e-05, "loss": 0.1834, "step": 414 }, { "epoch": 0.028518416712479386, "grad_norm": 3.586118221282959, "learning_rate": 2.8489702517162473e-05, "loss": 0.1323, "step": 415 }, { "epoch": 0.028587135788894998, "grad_norm": 5.99957275390625, "learning_rate": 2.8558352402746e-05, "loss": 0.1919, "step": 416 }, { "epoch": 0.02865585486531061, "grad_norm": 2.7336347103118896, "learning_rate": 2.862700228832952e-05, "loss": 0.1596, "step": 417 }, { "epoch": 0.028724573941726225, "grad_norm": 3.275486946105957, "learning_rate": 2.8695652173913044e-05, "loss": 0.1412, "step": 418 }, { "epoch": 0.028793293018141836, "grad_norm": 3.154313802719116, "learning_rate": 2.876430205949657e-05, "loss": 0.149, "step": 419 }, { "epoch": 0.028862012094557448, "grad_norm": 4.556890964508057, "learning_rate": 2.8832951945080093e-05, "loss": 0.1484, "step": 420 }, { "epoch": 0.028930731170973063, "grad_norm": 3.6872806549072266, "learning_rate": 2.8901601830663616e-05, "loss": 0.1737, "step": 421 }, { "epoch": 0.028999450247388675, "grad_norm": 3.661506414413452, "learning_rate": 2.897025171624714e-05, "loss": 0.173, "step": 422 }, { "epoch": 0.029068169323804287, "grad_norm": 2.6766583919525146, "learning_rate": 2.9038901601830665e-05, "loss": 0.1838, "step": 423 }, { "epoch": 0.029136888400219902, "grad_norm": 1.8896390199661255, "learning_rate": 2.9107551487414188e-05, "loss": 0.1354, "step": 424 }, { "epoch": 0.029205607476635514, "grad_norm": 4.8907647132873535, "learning_rate": 2.917620137299771e-05, "loss": 0.1799, "step": 425 }, { "epoch": 0.029274326553051126, "grad_norm": 2.403414011001587, "learning_rate": 2.9244851258581237e-05, "loss": 0.1342, "step": 426 }, { "epoch": 0.02934304562946674, "grad_norm": 3.636671304702759, "learning_rate": 2.931350114416476e-05, "loss": 0.1635, "step": 427 }, { "epoch": 0.029411764705882353, "grad_norm": 2.557176113128662, "learning_rate": 2.9382151029748282e-05, "loss": 0.1471, "step": 428 }, { "epoch": 0.029480483782297964, "grad_norm": 3.296290874481201, "learning_rate": 2.945080091533181e-05, "loss": 0.1381, "step": 429 }, { "epoch": 0.02954920285871358, "grad_norm": 4.164522647857666, "learning_rate": 2.9519450800915335e-05, "loss": 0.1817, "step": 430 }, { "epoch": 0.02961792193512919, "grad_norm": 4.208075523376465, "learning_rate": 2.9588100686498857e-05, "loss": 0.1798, "step": 431 }, { "epoch": 0.029686641011544807, "grad_norm": 4.203215599060059, "learning_rate": 2.965675057208238e-05, "loss": 0.1772, "step": 432 }, { "epoch": 0.029755360087960418, "grad_norm": 4.78367805480957, "learning_rate": 2.9725400457665906e-05, "loss": 0.1832, "step": 433 }, { "epoch": 0.02982407916437603, "grad_norm": 4.724888324737549, "learning_rate": 2.979405034324943e-05, "loss": 0.172, "step": 434 }, { "epoch": 0.029892798240791645, "grad_norm": 2.966081142425537, "learning_rate": 2.9862700228832952e-05, "loss": 0.182, "step": 435 }, { "epoch": 0.029961517317207257, "grad_norm": 4.538941860198975, "learning_rate": 2.9931350114416478e-05, "loss": 0.1683, "step": 436 }, { "epoch": 0.03003023639362287, "grad_norm": 4.349801063537598, "learning_rate": 3e-05, "loss": 0.176, "step": 437 }, { "epoch": 0.030098955470038484, "grad_norm": 3.3799338340759277, "learning_rate": 2.9999999628465417e-05, "loss": 0.1289, "step": 438 }, { "epoch": 0.030167674546454096, "grad_norm": 3.2664923667907715, "learning_rate": 2.9999998513861694e-05, "loss": 0.1663, "step": 439 }, { "epoch": 0.030236393622869707, "grad_norm": 3.458483934402466, "learning_rate": 2.9999996656188885e-05, "loss": 0.1547, "step": 440 }, { "epoch": 0.030305112699285323, "grad_norm": 3.7097301483154297, "learning_rate": 2.9999994055447076e-05, "loss": 0.1286, "step": 441 }, { "epoch": 0.030373831775700934, "grad_norm": 3.5156524181365967, "learning_rate": 2.99999907116364e-05, "loss": 0.1656, "step": 442 }, { "epoch": 0.030442550852116546, "grad_norm": 3.436678409576416, "learning_rate": 2.9999986624757027e-05, "loss": 0.1781, "step": 443 }, { "epoch": 0.03051126992853216, "grad_norm": 3.3391268253326416, "learning_rate": 2.9999981794809157e-05, "loss": 0.1338, "step": 444 }, { "epoch": 0.030579989004947773, "grad_norm": 3.0599429607391357, "learning_rate": 2.9999976221793024e-05, "loss": 0.1244, "step": 445 }, { "epoch": 0.030648708081363385, "grad_norm": 3.3995201587677, "learning_rate": 2.9999969905708908e-05, "loss": 0.1344, "step": 446 }, { "epoch": 0.030717427157779, "grad_norm": 2.9897408485412598, "learning_rate": 2.9999962846557118e-05, "loss": 0.1422, "step": 447 }, { "epoch": 0.030786146234194612, "grad_norm": 3.6955528259277344, "learning_rate": 2.9999955044338012e-05, "loss": 0.15, "step": 448 }, { "epoch": 0.030854865310610224, "grad_norm": 3.6499955654144287, "learning_rate": 2.9999946499051968e-05, "loss": 0.1723, "step": 449 }, { "epoch": 0.03092358438702584, "grad_norm": 3.2712342739105225, "learning_rate": 2.9999937210699414e-05, "loss": 0.1583, "step": 450 }, { "epoch": 0.03099230346344145, "grad_norm": 5.250085830688477, "learning_rate": 2.9999927179280813e-05, "loss": 0.1733, "step": 451 }, { "epoch": 0.031061022539857066, "grad_norm": 2.9345405101776123, "learning_rate": 2.9999916404796653e-05, "loss": 0.1622, "step": 452 }, { "epoch": 0.031129741616272678, "grad_norm": 3.3845314979553223, "learning_rate": 2.9999904887247478e-05, "loss": 0.168, "step": 453 }, { "epoch": 0.03119846069268829, "grad_norm": 3.031559467315674, "learning_rate": 2.9999892626633848e-05, "loss": 0.1798, "step": 454 }, { "epoch": 0.031267179769103905, "grad_norm": 4.808742523193359, "learning_rate": 2.999987962295638e-05, "loss": 0.1963, "step": 455 }, { "epoch": 0.031335898845519516, "grad_norm": 3.2196812629699707, "learning_rate": 2.999986587621571e-05, "loss": 0.1546, "step": 456 }, { "epoch": 0.03140461792193513, "grad_norm": 3.2685554027557373, "learning_rate": 2.9999851386412524e-05, "loss": 0.1311, "step": 457 }, { "epoch": 0.03147333699835074, "grad_norm": 3.325152635574341, "learning_rate": 2.9999836153547543e-05, "loss": 0.1507, "step": 458 }, { "epoch": 0.03154205607476635, "grad_norm": 4.301544189453125, "learning_rate": 2.9999820177621513e-05, "loss": 0.1636, "step": 459 }, { "epoch": 0.03161077515118197, "grad_norm": 3.457829713821411, "learning_rate": 2.999980345863523e-05, "loss": 0.1267, "step": 460 }, { "epoch": 0.03167949422759758, "grad_norm": 3.0617613792419434, "learning_rate": 2.9999785996589522e-05, "loss": 0.1401, "step": 461 }, { "epoch": 0.031748213304013194, "grad_norm": 4.792970180511475, "learning_rate": 2.999976779148526e-05, "loss": 0.1746, "step": 462 }, { "epoch": 0.031816932380428806, "grad_norm": 3.2135884761810303, "learning_rate": 2.999974884332333e-05, "loss": 0.2128, "step": 463 }, { "epoch": 0.03188565145684442, "grad_norm": 3.8818271160125732, "learning_rate": 2.9999729152104687e-05, "loss": 0.1798, "step": 464 }, { "epoch": 0.031954370533260036, "grad_norm": 5.075356483459473, "learning_rate": 2.9999708717830298e-05, "loss": 0.1371, "step": 465 }, { "epoch": 0.03202308960967565, "grad_norm": 3.945539951324463, "learning_rate": 2.9999687540501178e-05, "loss": 0.1699, "step": 466 }, { "epoch": 0.03209180868609126, "grad_norm": 4.012805938720703, "learning_rate": 2.999966562011837e-05, "loss": 0.1754, "step": 467 }, { "epoch": 0.03216052776250687, "grad_norm": 3.1253342628479004, "learning_rate": 2.9999642956682973e-05, "loss": 0.1551, "step": 468 }, { "epoch": 0.03222924683892248, "grad_norm": 3.4029693603515625, "learning_rate": 2.9999619550196097e-05, "loss": 0.178, "step": 469 }, { "epoch": 0.032297965915338095, "grad_norm": 3.6534807682037354, "learning_rate": 2.999959540065891e-05, "loss": 0.1686, "step": 470 }, { "epoch": 0.032366684991753714, "grad_norm": 4.110112190246582, "learning_rate": 2.9999570508072603e-05, "loss": 0.1752, "step": 471 }, { "epoch": 0.032435404068169325, "grad_norm": 3.834857225418091, "learning_rate": 2.999954487243841e-05, "loss": 0.1594, "step": 472 }, { "epoch": 0.03250412314458494, "grad_norm": 3.230417490005493, "learning_rate": 2.9999518493757598e-05, "loss": 0.1467, "step": 473 }, { "epoch": 0.03257284222100055, "grad_norm": 2.728452682495117, "learning_rate": 2.9999491372031485e-05, "loss": 0.1524, "step": 474 }, { "epoch": 0.03264156129741616, "grad_norm": 2.714285373687744, "learning_rate": 2.9999463507261403e-05, "loss": 0.1373, "step": 475 }, { "epoch": 0.03271028037383177, "grad_norm": 3.566642999649048, "learning_rate": 2.9999434899448734e-05, "loss": 0.1595, "step": 476 }, { "epoch": 0.03277899945024739, "grad_norm": 3.526045799255371, "learning_rate": 2.9999405548594905e-05, "loss": 0.1133, "step": 477 }, { "epoch": 0.032847718526663, "grad_norm": 3.375056743621826, "learning_rate": 2.9999375454701358e-05, "loss": 0.1304, "step": 478 }, { "epoch": 0.032916437603078615, "grad_norm": 3.6749989986419678, "learning_rate": 2.999934461776959e-05, "loss": 0.1272, "step": 479 }, { "epoch": 0.032985156679494226, "grad_norm": 3.8975980281829834, "learning_rate": 2.999931303780112e-05, "loss": 0.1542, "step": 480 }, { "epoch": 0.03305387575590984, "grad_norm": 3.292116165161133, "learning_rate": 2.999928071479753e-05, "loss": 0.1337, "step": 481 }, { "epoch": 0.03312259483232546, "grad_norm": 2.738342523574829, "learning_rate": 2.9999247648760408e-05, "loss": 0.1504, "step": 482 }, { "epoch": 0.03319131390874107, "grad_norm": 3.591148853302002, "learning_rate": 2.999921383969139e-05, "loss": 0.143, "step": 483 }, { "epoch": 0.03326003298515668, "grad_norm": 2.9360363483428955, "learning_rate": 2.9999179287592156e-05, "loss": 0.1403, "step": 484 }, { "epoch": 0.03332875206157229, "grad_norm": 4.114712715148926, "learning_rate": 2.9999143992464418e-05, "loss": 0.1885, "step": 485 }, { "epoch": 0.033397471137987904, "grad_norm": 4.003205299377441, "learning_rate": 2.9999107954309925e-05, "loss": 0.1658, "step": 486 }, { "epoch": 0.033466190214403516, "grad_norm": 4.701906681060791, "learning_rate": 2.9999071173130464e-05, "loss": 0.1524, "step": 487 }, { "epoch": 0.033534909290819134, "grad_norm": 3.629706382751465, "learning_rate": 2.9999033648927847e-05, "loss": 0.1309, "step": 488 }, { "epoch": 0.033603628367234746, "grad_norm": 3.2381696701049805, "learning_rate": 2.999899538170394e-05, "loss": 0.1282, "step": 489 }, { "epoch": 0.03367234744365036, "grad_norm": 6.146234512329102, "learning_rate": 2.9998956371460646e-05, "loss": 0.1406, "step": 490 }, { "epoch": 0.03374106652006597, "grad_norm": 3.3701884746551514, "learning_rate": 2.9998916618199884e-05, "loss": 0.1532, "step": 491 }, { "epoch": 0.03380978559648158, "grad_norm": 5.681600570678711, "learning_rate": 2.999887612192363e-05, "loss": 0.1849, "step": 492 }, { "epoch": 0.03387850467289719, "grad_norm": 2.327239990234375, "learning_rate": 2.9998834882633892e-05, "loss": 0.1461, "step": 493 }, { "epoch": 0.03394722374931281, "grad_norm": 3.53572940826416, "learning_rate": 2.9998792900332705e-05, "loss": 0.1484, "step": 494 }, { "epoch": 0.034015942825728424, "grad_norm": 2.4683291912078857, "learning_rate": 2.999875017502216e-05, "loss": 0.1528, "step": 495 }, { "epoch": 0.034084661902144035, "grad_norm": 3.890333890914917, "learning_rate": 2.9998706706704363e-05, "loss": 0.2098, "step": 496 }, { "epoch": 0.03415338097855965, "grad_norm": 3.9771604537963867, "learning_rate": 2.9998662495381474e-05, "loss": 0.1721, "step": 497 }, { "epoch": 0.03422210005497526, "grad_norm": 6.197415828704834, "learning_rate": 2.999861754105568e-05, "loss": 0.1226, "step": 498 }, { "epoch": 0.03429081913139088, "grad_norm": 4.372481822967529, "learning_rate": 2.999857184372921e-05, "loss": 0.2118, "step": 499 }, { "epoch": 0.03435953820780649, "grad_norm": 2.8908064365386963, "learning_rate": 2.9998525403404326e-05, "loss": 0.1294, "step": 500 }, { "epoch": 0.0344282572842221, "grad_norm": 3.636202812194824, "learning_rate": 2.9998478220083328e-05, "loss": 0.1275, "step": 501 }, { "epoch": 0.03449697636063771, "grad_norm": 3.258974075317383, "learning_rate": 2.9998430293768552e-05, "loss": 0.1356, "step": 502 }, { "epoch": 0.034565695437053325, "grad_norm": 3.6179091930389404, "learning_rate": 2.9998381624462378e-05, "loss": 0.1604, "step": 503 }, { "epoch": 0.034634414513468936, "grad_norm": 4.810796737670898, "learning_rate": 2.9998332212167216e-05, "loss": 0.1727, "step": 504 }, { "epoch": 0.034703133589884555, "grad_norm": 3.184286594390869, "learning_rate": 2.9998282056885507e-05, "loss": 0.1614, "step": 505 }, { "epoch": 0.03477185266630017, "grad_norm": 3.5860300064086914, "learning_rate": 2.9998231158619742e-05, "loss": 0.143, "step": 506 }, { "epoch": 0.03484057174271578, "grad_norm": 3.040266990661621, "learning_rate": 2.9998179517372443e-05, "loss": 0.1199, "step": 507 }, { "epoch": 0.03490929081913139, "grad_norm": 3.574176073074341, "learning_rate": 2.9998127133146163e-05, "loss": 0.1721, "step": 508 }, { "epoch": 0.034978009895547, "grad_norm": 2.7079575061798096, "learning_rate": 2.99980740059435e-05, "loss": 0.1137, "step": 509 }, { "epoch": 0.035046728971962614, "grad_norm": 3.8945140838623047, "learning_rate": 2.9998020135767084e-05, "loss": 0.1385, "step": 510 }, { "epoch": 0.03511544804837823, "grad_norm": 2.184969425201416, "learning_rate": 2.9997965522619587e-05, "loss": 0.1363, "step": 511 }, { "epoch": 0.035184167124793844, "grad_norm": 3.6004467010498047, "learning_rate": 2.9997910166503713e-05, "loss": 0.1672, "step": 512 }, { "epoch": 0.035252886201209456, "grad_norm": 3.739633798599243, "learning_rate": 2.9997854067422203e-05, "loss": 0.1698, "step": 513 }, { "epoch": 0.03532160527762507, "grad_norm": 5.088380813598633, "learning_rate": 2.9997797225377833e-05, "loss": 0.1276, "step": 514 }, { "epoch": 0.03539032435404068, "grad_norm": 3.460893154144287, "learning_rate": 2.9997739640373428e-05, "loss": 0.1516, "step": 515 }, { "epoch": 0.03545904343045629, "grad_norm": 2.5017166137695312, "learning_rate": 2.9997681312411837e-05, "loss": 0.1624, "step": 516 }, { "epoch": 0.03552776250687191, "grad_norm": 4.876821517944336, "learning_rate": 2.9997622241495943e-05, "loss": 0.13, "step": 517 }, { "epoch": 0.03559648158328752, "grad_norm": 7.058427333831787, "learning_rate": 2.9997562427628678e-05, "loss": 0.1319, "step": 518 }, { "epoch": 0.03566520065970313, "grad_norm": 3.169121265411377, "learning_rate": 2.9997501870813002e-05, "loss": 0.1322, "step": 519 }, { "epoch": 0.035733919736118745, "grad_norm": 3.3026745319366455, "learning_rate": 2.999744057105192e-05, "loss": 0.1341, "step": 520 }, { "epoch": 0.03580263881253436, "grad_norm": 4.316922664642334, "learning_rate": 2.9997378528348468e-05, "loss": 0.1154, "step": 521 }, { "epoch": 0.035871357888949976, "grad_norm": 6.8003315925598145, "learning_rate": 2.9997315742705708e-05, "loss": 0.1518, "step": 522 }, { "epoch": 0.03594007696536559, "grad_norm": 2.807241678237915, "learning_rate": 2.9997252214126768e-05, "loss": 0.1332, "step": 523 }, { "epoch": 0.0360087960417812, "grad_norm": 4.321352958679199, "learning_rate": 2.9997187942614784e-05, "loss": 0.1294, "step": 524 }, { "epoch": 0.03607751511819681, "grad_norm": 2.995412588119507, "learning_rate": 2.9997122928172942e-05, "loss": 0.1471, "step": 525 }, { "epoch": 0.03614623419461242, "grad_norm": 4.705594062805176, "learning_rate": 2.9997057170804458e-05, "loss": 0.1346, "step": 526 }, { "epoch": 0.036214953271028034, "grad_norm": 3.387735366821289, "learning_rate": 2.9996990670512603e-05, "loss": 0.1151, "step": 527 }, { "epoch": 0.03628367234744365, "grad_norm": 4.146975517272949, "learning_rate": 2.9996923427300658e-05, "loss": 0.1772, "step": 528 }, { "epoch": 0.036352391423859265, "grad_norm": 4.105936050415039, "learning_rate": 2.9996855441171958e-05, "loss": 0.1296, "step": 529 }, { "epoch": 0.03642111050027488, "grad_norm": 2.9803268909454346, "learning_rate": 2.9996786712129874e-05, "loss": 0.1755, "step": 530 }, { "epoch": 0.03648982957669049, "grad_norm": 3.7451791763305664, "learning_rate": 2.9996717240177805e-05, "loss": 0.1669, "step": 531 }, { "epoch": 0.0365585486531061, "grad_norm": 4.144136428833008, "learning_rate": 2.99966470253192e-05, "loss": 0.153, "step": 532 }, { "epoch": 0.03662726772952171, "grad_norm": 2.4652984142303467, "learning_rate": 2.999657606755753e-05, "loss": 0.1055, "step": 533 }, { "epoch": 0.03669598680593733, "grad_norm": 2.9748947620391846, "learning_rate": 2.9996504366896317e-05, "loss": 0.1072, "step": 534 }, { "epoch": 0.03676470588235294, "grad_norm": 3.6570518016815186, "learning_rate": 2.9996431923339107e-05, "loss": 0.1428, "step": 535 }, { "epoch": 0.036833424958768554, "grad_norm": 3.7767555713653564, "learning_rate": 2.999635873688949e-05, "loss": 0.159, "step": 536 }, { "epoch": 0.036902144035184166, "grad_norm": 3.2482025623321533, "learning_rate": 2.999628480755109e-05, "loss": 0.1434, "step": 537 }, { "epoch": 0.03697086311159978, "grad_norm": 4.568009376525879, "learning_rate": 2.9996210135327577e-05, "loss": 0.1349, "step": 538 }, { "epoch": 0.037039582188015396, "grad_norm": 2.918185234069824, "learning_rate": 2.999613472022264e-05, "loss": 0.1429, "step": 539 }, { "epoch": 0.03710830126443101, "grad_norm": 3.0404958724975586, "learning_rate": 2.9996058562240024e-05, "loss": 0.1493, "step": 540 }, { "epoch": 0.03717702034084662, "grad_norm": 2.848618745803833, "learning_rate": 2.9995981661383495e-05, "loss": 0.1251, "step": 541 }, { "epoch": 0.03724573941726223, "grad_norm": 3.1252992153167725, "learning_rate": 2.999590401765687e-05, "loss": 0.1234, "step": 542 }, { "epoch": 0.03731445849367784, "grad_norm": 4.563553333282471, "learning_rate": 2.9995825631063984e-05, "loss": 0.1516, "step": 543 }, { "epoch": 0.037383177570093455, "grad_norm": 3.0766305923461914, "learning_rate": 2.9995746501608728e-05, "loss": 0.1231, "step": 544 }, { "epoch": 0.037451896646509074, "grad_norm": 3.1019694805145264, "learning_rate": 2.999566662929502e-05, "loss": 0.1452, "step": 545 }, { "epoch": 0.037520615722924686, "grad_norm": 2.7688002586364746, "learning_rate": 2.999558601412682e-05, "loss": 0.1404, "step": 546 }, { "epoch": 0.0375893347993403, "grad_norm": 2.757249116897583, "learning_rate": 2.999550465610811e-05, "loss": 0.1502, "step": 547 }, { "epoch": 0.03765805387575591, "grad_norm": 3.2034246921539307, "learning_rate": 2.999542255524294e-05, "loss": 0.1254, "step": 548 }, { "epoch": 0.03772677295217152, "grad_norm": 2.688452959060669, "learning_rate": 2.9995339711535357e-05, "loss": 0.1251, "step": 549 }, { "epoch": 0.03779549202858713, "grad_norm": 1.8371212482452393, "learning_rate": 2.999525612498948e-05, "loss": 0.1187, "step": 550 }, { "epoch": 0.03786421110500275, "grad_norm": 2.7232730388641357, "learning_rate": 2.999517179560944e-05, "loss": 0.1695, "step": 551 }, { "epoch": 0.03793293018141836, "grad_norm": 4.185912132263184, "learning_rate": 2.9995086723399417e-05, "loss": 0.1573, "step": 552 }, { "epoch": 0.038001649257833975, "grad_norm": 2.566218137741089, "learning_rate": 2.9995000908363628e-05, "loss": 0.1262, "step": 553 }, { "epoch": 0.03807036833424959, "grad_norm": 3.6056532859802246, "learning_rate": 2.9994914350506324e-05, "loss": 0.1256, "step": 554 }, { "epoch": 0.0381390874106652, "grad_norm": 3.2860758304595947, "learning_rate": 2.999482704983179e-05, "loss": 0.1636, "step": 555 }, { "epoch": 0.03820780648708082, "grad_norm": 2.8866405487060547, "learning_rate": 2.999473900634435e-05, "loss": 0.1261, "step": 556 }, { "epoch": 0.03827652556349643, "grad_norm": 2.452105760574341, "learning_rate": 2.999465022004837e-05, "loss": 0.1616, "step": 557 }, { "epoch": 0.03834524463991204, "grad_norm": 3.3643765449523926, "learning_rate": 2.9994560690948244e-05, "loss": 0.1322, "step": 558 }, { "epoch": 0.03841396371632765, "grad_norm": 2.4583890438079834, "learning_rate": 2.999447041904841e-05, "loss": 0.112, "step": 559 }, { "epoch": 0.038482682792743264, "grad_norm": 2.928053140640259, "learning_rate": 2.9994379404353344e-05, "loss": 0.1963, "step": 560 }, { "epoch": 0.038551401869158876, "grad_norm": 3.8541100025177, "learning_rate": 2.999428764686754e-05, "loss": 0.1205, "step": 561 }, { "epoch": 0.038620120945574495, "grad_norm": 2.487278699874878, "learning_rate": 2.9994195146595556e-05, "loss": 0.1344, "step": 562 }, { "epoch": 0.038688840021990106, "grad_norm": 2.407606363296509, "learning_rate": 2.9994101903541972e-05, "loss": 0.1346, "step": 563 }, { "epoch": 0.03875755909840572, "grad_norm": 2.742018938064575, "learning_rate": 2.999400791771141e-05, "loss": 0.2213, "step": 564 }, { "epoch": 0.03882627817482133, "grad_norm": 3.801421642303467, "learning_rate": 2.9993913189108518e-05, "loss": 0.1466, "step": 565 }, { "epoch": 0.03889499725123694, "grad_norm": 2.6886000633239746, "learning_rate": 2.9993817717737993e-05, "loss": 0.1341, "step": 566 }, { "epoch": 0.03896371632765255, "grad_norm": 4.607528209686279, "learning_rate": 2.9993721503604565e-05, "loss": 0.1137, "step": 567 }, { "epoch": 0.03903243540406817, "grad_norm": 3.726102590560913, "learning_rate": 2.9993624546713e-05, "loss": 0.1297, "step": 568 }, { "epoch": 0.039101154480483784, "grad_norm": 3.86340594291687, "learning_rate": 2.99935268470681e-05, "loss": 0.1043, "step": 569 }, { "epoch": 0.039169873556899396, "grad_norm": 6.193768501281738, "learning_rate": 2.9993428404674703e-05, "loss": 0.1091, "step": 570 }, { "epoch": 0.03923859263331501, "grad_norm": 2.0518887042999268, "learning_rate": 2.999332921953769e-05, "loss": 0.0984, "step": 571 }, { "epoch": 0.03930731170973062, "grad_norm": 2.984464168548584, "learning_rate": 2.9993229291661974e-05, "loss": 0.1534, "step": 572 }, { "epoch": 0.03937603078614623, "grad_norm": 3.030172824859619, "learning_rate": 2.9993128621052505e-05, "loss": 0.1866, "step": 573 }, { "epoch": 0.03944474986256185, "grad_norm": 2.658539056777954, "learning_rate": 2.9993027207714264e-05, "loss": 0.1547, "step": 574 }, { "epoch": 0.03951346893897746, "grad_norm": 2.630455493927002, "learning_rate": 2.999292505165228e-05, "loss": 0.0963, "step": 575 }, { "epoch": 0.03958218801539307, "grad_norm": 2.5162649154663086, "learning_rate": 2.9992822152871617e-05, "loss": 0.1134, "step": 576 }, { "epoch": 0.039650907091808685, "grad_norm": 3.489565372467041, "learning_rate": 2.9992718511377366e-05, "loss": 0.1104, "step": 577 }, { "epoch": 0.0397196261682243, "grad_norm": 3.070950984954834, "learning_rate": 2.999261412717466e-05, "loss": 0.1384, "step": 578 }, { "epoch": 0.039788345244639915, "grad_norm": 2.274646759033203, "learning_rate": 2.999250900026868e-05, "loss": 0.1011, "step": 579 }, { "epoch": 0.03985706432105553, "grad_norm": 2.539151906967163, "learning_rate": 2.9992403130664626e-05, "loss": 0.0951, "step": 580 }, { "epoch": 0.03992578339747114, "grad_norm": 3.3916351795196533, "learning_rate": 2.999229651836774e-05, "loss": 0.1317, "step": 581 }, { "epoch": 0.03999450247388675, "grad_norm": 3.368335723876953, "learning_rate": 2.999218916338331e-05, "loss": 0.1451, "step": 582 }, { "epoch": 0.04006322155030236, "grad_norm": 3.0597846508026123, "learning_rate": 2.9992081065716653e-05, "loss": 0.1038, "step": 583 }, { "epoch": 0.040131940626717974, "grad_norm": 3.31697416305542, "learning_rate": 2.9991972225373123e-05, "loss": 0.1232, "step": 584 }, { "epoch": 0.04020065970313359, "grad_norm": 3.271181583404541, "learning_rate": 2.9991862642358107e-05, "loss": 0.1429, "step": 585 }, { "epoch": 0.040269378779549204, "grad_norm": 2.367365598678589, "learning_rate": 2.9991752316677042e-05, "loss": 0.1085, "step": 586 }, { "epoch": 0.040338097855964816, "grad_norm": 1.6097815036773682, "learning_rate": 2.9991641248335388e-05, "loss": 0.1218, "step": 587 }, { "epoch": 0.04040681693238043, "grad_norm": 2.925261974334717, "learning_rate": 2.9991529437338644e-05, "loss": 0.1276, "step": 588 }, { "epoch": 0.04047553600879604, "grad_norm": 2.1514954566955566, "learning_rate": 2.999141688369236e-05, "loss": 0.0921, "step": 589 }, { "epoch": 0.04054425508521165, "grad_norm": 2.6934545040130615, "learning_rate": 2.99913035874021e-05, "loss": 0.1102, "step": 590 }, { "epoch": 0.04061297416162727, "grad_norm": 2.906372547149658, "learning_rate": 2.999118954847348e-05, "loss": 0.1058, "step": 591 }, { "epoch": 0.04068169323804288, "grad_norm": 2.6554019451141357, "learning_rate": 2.999107476691215e-05, "loss": 0.1214, "step": 592 }, { "epoch": 0.040750412314458494, "grad_norm": 3.003293514251709, "learning_rate": 2.9990959242723802e-05, "loss": 0.1041, "step": 593 }, { "epoch": 0.040819131390874105, "grad_norm": 2.6214451789855957, "learning_rate": 2.999084297591415e-05, "loss": 0.1646, "step": 594 }, { "epoch": 0.04088785046728972, "grad_norm": 2.7035210132598877, "learning_rate": 2.9990725966488956e-05, "loss": 0.1304, "step": 595 }, { "epoch": 0.040956569543705336, "grad_norm": 2.3543355464935303, "learning_rate": 2.9990608214454016e-05, "loss": 0.1205, "step": 596 }, { "epoch": 0.04102528862012095, "grad_norm": 2.4838473796844482, "learning_rate": 2.999048971981517e-05, "loss": 0.1055, "step": 597 }, { "epoch": 0.04109400769653656, "grad_norm": 2.2398340702056885, "learning_rate": 2.9990370482578273e-05, "loss": 0.1062, "step": 598 }, { "epoch": 0.04116272677295217, "grad_norm": 2.289276123046875, "learning_rate": 2.999025050274925e-05, "loss": 0.1306, "step": 599 }, { "epoch": 0.04123144584936778, "grad_norm": 2.8925516605377197, "learning_rate": 2.999012978033403e-05, "loss": 0.1205, "step": 600 }, { "epoch": 0.041300164925783395, "grad_norm": 2.446601152420044, "learning_rate": 2.9990008315338604e-05, "loss": 0.1312, "step": 601 }, { "epoch": 0.04136888400219901, "grad_norm": 2.379640579223633, "learning_rate": 2.9989886107768982e-05, "loss": 0.117, "step": 602 }, { "epoch": 0.041437603078614625, "grad_norm": 2.9531803131103516, "learning_rate": 2.998976315763122e-05, "loss": 0.1102, "step": 603 }, { "epoch": 0.04150632215503024, "grad_norm": 2.5359978675842285, "learning_rate": 2.998963946493141e-05, "loss": 0.1197, "step": 604 }, { "epoch": 0.04157504123144585, "grad_norm": 3.331226110458374, "learning_rate": 2.9989515029675676e-05, "loss": 0.1574, "step": 605 }, { "epoch": 0.04164376030786146, "grad_norm": 3.0241568088531494, "learning_rate": 2.998938985187018e-05, "loss": 0.0871, "step": 606 }, { "epoch": 0.04171247938427707, "grad_norm": 2.4164884090423584, "learning_rate": 2.9989263931521136e-05, "loss": 0.1161, "step": 607 }, { "epoch": 0.04178119846069269, "grad_norm": 3.0708956718444824, "learning_rate": 2.998913726863477e-05, "loss": 0.1058, "step": 608 }, { "epoch": 0.0418499175371083, "grad_norm": 2.93009090423584, "learning_rate": 2.998900986321736e-05, "loss": 0.1106, "step": 609 }, { "epoch": 0.041918636613523914, "grad_norm": 3.4144489765167236, "learning_rate": 2.998888171527522e-05, "loss": 0.1063, "step": 610 }, { "epoch": 0.041987355689939526, "grad_norm": 5.995541095733643, "learning_rate": 2.9988752824814696e-05, "loss": 0.1498, "step": 611 }, { "epoch": 0.04205607476635514, "grad_norm": 3.148744821548462, "learning_rate": 2.9988623191842167e-05, "loss": 0.118, "step": 612 }, { "epoch": 0.04212479384277076, "grad_norm": 2.5194859504699707, "learning_rate": 2.9988492816364067e-05, "loss": 0.1068, "step": 613 }, { "epoch": 0.04219351291918637, "grad_norm": 4.616106033325195, "learning_rate": 2.9988361698386843e-05, "loss": 0.1383, "step": 614 }, { "epoch": 0.04226223199560198, "grad_norm": 3.258246898651123, "learning_rate": 2.9988229837916997e-05, "loss": 0.1247, "step": 615 }, { "epoch": 0.04233095107201759, "grad_norm": 5.036736011505127, "learning_rate": 2.9988097234961058e-05, "loss": 0.1471, "step": 616 }, { "epoch": 0.042399670148433204, "grad_norm": 10.568196296691895, "learning_rate": 2.99879638895256e-05, "loss": 0.1522, "step": 617 }, { "epoch": 0.042468389224848815, "grad_norm": 3.90811824798584, "learning_rate": 2.9987829801617223e-05, "loss": 0.1644, "step": 618 }, { "epoch": 0.042537108301264434, "grad_norm": 3.459172248840332, "learning_rate": 2.998769497124257e-05, "loss": 0.1832, "step": 619 }, { "epoch": 0.042605827377680046, "grad_norm": 4.366562843322754, "learning_rate": 2.998755939840832e-05, "loss": 0.1905, "step": 620 }, { "epoch": 0.04267454645409566, "grad_norm": 2.950479507446289, "learning_rate": 2.9987423083121193e-05, "loss": 0.1251, "step": 621 }, { "epoch": 0.04274326553051127, "grad_norm": 2.8339571952819824, "learning_rate": 2.998728602538794e-05, "loss": 0.1035, "step": 622 }, { "epoch": 0.04281198460692688, "grad_norm": 3.4357693195343018, "learning_rate": 2.9987148225215352e-05, "loss": 0.1537, "step": 623 }, { "epoch": 0.04288070368334249, "grad_norm": 2.3703691959381104, "learning_rate": 2.998700968261025e-05, "loss": 0.1133, "step": 624 }, { "epoch": 0.04294942275975811, "grad_norm": 2.3712260723114014, "learning_rate": 2.9986870397579495e-05, "loss": 0.1123, "step": 625 }, { "epoch": 0.04301814183617372, "grad_norm": 4.96372127532959, "learning_rate": 2.9986730370129997e-05, "loss": 0.131, "step": 626 }, { "epoch": 0.043086860912589335, "grad_norm": 3.8273496627807617, "learning_rate": 2.998658960026869e-05, "loss": 0.0951, "step": 627 }, { "epoch": 0.04315557998900495, "grad_norm": 2.66546368598938, "learning_rate": 2.9986448088002545e-05, "loss": 0.1391, "step": 628 }, { "epoch": 0.04322429906542056, "grad_norm": 2.6296207904815674, "learning_rate": 2.998630583333857e-05, "loss": 0.1059, "step": 629 }, { "epoch": 0.04329301814183617, "grad_norm": 5.72160005569458, "learning_rate": 2.998616283628382e-05, "loss": 0.125, "step": 630 }, { "epoch": 0.04336173721825179, "grad_norm": 2.8455193042755127, "learning_rate": 2.998601909684537e-05, "loss": 0.1038, "step": 631 }, { "epoch": 0.0434304562946674, "grad_norm": 5.943075656890869, "learning_rate": 2.9985874615030338e-05, "loss": 0.1211, "step": 632 }, { "epoch": 0.04349917537108301, "grad_norm": 2.614259719848633, "learning_rate": 2.9985729390845894e-05, "loss": 0.1129, "step": 633 }, { "epoch": 0.043567894447498624, "grad_norm": 4.510036945343018, "learning_rate": 2.998558342429922e-05, "loss": 0.1424, "step": 634 }, { "epoch": 0.043636613523914236, "grad_norm": 5.03054141998291, "learning_rate": 2.998543671539756e-05, "loss": 0.1602, "step": 635 }, { "epoch": 0.043705332600329855, "grad_norm": 2.752483367919922, "learning_rate": 2.998528926414817e-05, "loss": 0.1462, "step": 636 }, { "epoch": 0.04377405167674547, "grad_norm": 2.5503287315368652, "learning_rate": 2.9985141070558357e-05, "loss": 0.12, "step": 637 }, { "epoch": 0.04384277075316108, "grad_norm": 3.0037100315093994, "learning_rate": 2.9984992134635463e-05, "loss": 0.1088, "step": 638 }, { "epoch": 0.04391148982957669, "grad_norm": 3.537247896194458, "learning_rate": 2.9984842456386867e-05, "loss": 0.1326, "step": 639 }, { "epoch": 0.0439802089059923, "grad_norm": 5.240083694458008, "learning_rate": 2.9984692035819986e-05, "loss": 0.1177, "step": 640 }, { "epoch": 0.044048927982407914, "grad_norm": 2.9474263191223145, "learning_rate": 2.998454087294227e-05, "loss": 0.1313, "step": 641 }, { "epoch": 0.04411764705882353, "grad_norm": 2.6787116527557373, "learning_rate": 2.99843889677612e-05, "loss": 0.1201, "step": 642 }, { "epoch": 0.044186366135239144, "grad_norm": 2.3164477348327637, "learning_rate": 2.998423632028431e-05, "loss": 0.0894, "step": 643 }, { "epoch": 0.044255085211654756, "grad_norm": 3.823225259780884, "learning_rate": 2.998408293051916e-05, "loss": 0.1211, "step": 644 }, { "epoch": 0.04432380428807037, "grad_norm": 2.7460572719573975, "learning_rate": 2.9983928798473347e-05, "loss": 0.1309, "step": 645 }, { "epoch": 0.04439252336448598, "grad_norm": 2.9180748462677, "learning_rate": 2.9983773924154506e-05, "loss": 0.1322, "step": 646 }, { "epoch": 0.04446124244090159, "grad_norm": 2.247143268585205, "learning_rate": 2.9983618307570312e-05, "loss": 0.12, "step": 647 }, { "epoch": 0.04452996151731721, "grad_norm": 2.4070017337799072, "learning_rate": 2.9983461948728474e-05, "loss": 0.1217, "step": 648 }, { "epoch": 0.04459868059373282, "grad_norm": 3.569058656692505, "learning_rate": 2.998330484763673e-05, "loss": 0.1445, "step": 649 }, { "epoch": 0.04466739967014843, "grad_norm": 2.7817156314849854, "learning_rate": 2.9983147004302873e-05, "loss": 0.1389, "step": 650 }, { "epoch": 0.044736118746564045, "grad_norm": 2.101501941680908, "learning_rate": 2.9982988418734716e-05, "loss": 0.0955, "step": 651 }, { "epoch": 0.04480483782297966, "grad_norm": 4.099470615386963, "learning_rate": 2.9982829090940113e-05, "loss": 0.1496, "step": 652 }, { "epoch": 0.044873556899395275, "grad_norm": 2.500619888305664, "learning_rate": 2.9982669020926964e-05, "loss": 0.1223, "step": 653 }, { "epoch": 0.04494227597581089, "grad_norm": 2.730642557144165, "learning_rate": 2.9982508208703194e-05, "loss": 0.111, "step": 654 }, { "epoch": 0.0450109950522265, "grad_norm": 2.7717368602752686, "learning_rate": 2.998234665427677e-05, "loss": 0.1441, "step": 655 }, { "epoch": 0.04507971412864211, "grad_norm": 3.718820810317993, "learning_rate": 2.9982184357655693e-05, "loss": 0.1424, "step": 656 }, { "epoch": 0.04514843320505772, "grad_norm": 2.8288533687591553, "learning_rate": 2.998202131884801e-05, "loss": 0.1622, "step": 657 }, { "epoch": 0.045217152281473334, "grad_norm": 2.5851266384124756, "learning_rate": 2.998185753786179e-05, "loss": 0.1008, "step": 658 }, { "epoch": 0.04528587135788895, "grad_norm": 2.4692018032073975, "learning_rate": 2.9981693014705145e-05, "loss": 0.1191, "step": 659 }, { "epoch": 0.045354590434304565, "grad_norm": 2.3693277835845947, "learning_rate": 2.9981527749386235e-05, "loss": 0.1157, "step": 660 }, { "epoch": 0.045423309510720176, "grad_norm": 4.339116096496582, "learning_rate": 2.998136174191324e-05, "loss": 0.1103, "step": 661 }, { "epoch": 0.04549202858713579, "grad_norm": 2.1931216716766357, "learning_rate": 2.9981194992294382e-05, "loss": 0.1297, "step": 662 }, { "epoch": 0.0455607476635514, "grad_norm": 2.8306853771209717, "learning_rate": 2.9981027500537922e-05, "loss": 0.1057, "step": 663 }, { "epoch": 0.04562946673996701, "grad_norm": 2.440006732940674, "learning_rate": 2.9980859266652166e-05, "loss": 0.0947, "step": 664 }, { "epoch": 0.04569818581638263, "grad_norm": 3.1594276428222656, "learning_rate": 2.9980690290645436e-05, "loss": 0.1186, "step": 665 }, { "epoch": 0.04576690489279824, "grad_norm": 4.067253589630127, "learning_rate": 2.9980520572526113e-05, "loss": 0.0961, "step": 666 }, { "epoch": 0.045835623969213854, "grad_norm": 2.3573176860809326, "learning_rate": 2.9980350112302594e-05, "loss": 0.095, "step": 667 }, { "epoch": 0.045904343045629466, "grad_norm": 2.6601059436798096, "learning_rate": 2.9980178909983333e-05, "loss": 0.1094, "step": 668 }, { "epoch": 0.04597306212204508, "grad_norm": 2.582716941833496, "learning_rate": 2.9980006965576805e-05, "loss": 0.1179, "step": 669 }, { "epoch": 0.04604178119846069, "grad_norm": 2.7867209911346436, "learning_rate": 2.997983427909153e-05, "loss": 0.1273, "step": 670 }, { "epoch": 0.04611050027487631, "grad_norm": 3.098991632461548, "learning_rate": 2.9979660850536063e-05, "loss": 0.1395, "step": 671 }, { "epoch": 0.04617921935129192, "grad_norm": 2.2143633365631104, "learning_rate": 2.9979486679918993e-05, "loss": 0.0978, "step": 672 }, { "epoch": 0.04624793842770753, "grad_norm": 2.802797555923462, "learning_rate": 2.997931176724895e-05, "loss": 0.1055, "step": 673 }, { "epoch": 0.04631665750412314, "grad_norm": 3.079634189605713, "learning_rate": 2.9979136112534593e-05, "loss": 0.1319, "step": 674 }, { "epoch": 0.046385376580538755, "grad_norm": 2.827336549758911, "learning_rate": 2.9978959715784636e-05, "loss": 0.1208, "step": 675 }, { "epoch": 0.046454095656954374, "grad_norm": 3.3023929595947266, "learning_rate": 2.9978782577007808e-05, "loss": 0.1166, "step": 676 }, { "epoch": 0.046522814733369985, "grad_norm": 3.3801794052124023, "learning_rate": 2.9978604696212885e-05, "loss": 0.0782, "step": 677 }, { "epoch": 0.0465915338097856, "grad_norm": 2.350255012512207, "learning_rate": 2.997842607340868e-05, "loss": 0.0926, "step": 678 }, { "epoch": 0.04666025288620121, "grad_norm": 1.918501615524292, "learning_rate": 2.9978246708604043e-05, "loss": 0.1213, "step": 679 }, { "epoch": 0.04672897196261682, "grad_norm": 2.3829169273376465, "learning_rate": 2.997806660180786e-05, "loss": 0.1338, "step": 680 }, { "epoch": 0.04679769103903243, "grad_norm": 2.73325252532959, "learning_rate": 2.9977885753029046e-05, "loss": 0.1393, "step": 681 }, { "epoch": 0.04686641011544805, "grad_norm": 3.782999277114868, "learning_rate": 2.9977704162276568e-05, "loss": 0.1326, "step": 682 }, { "epoch": 0.04693512919186366, "grad_norm": 2.8819026947021484, "learning_rate": 2.997752182955942e-05, "loss": 0.1455, "step": 683 }, { "epoch": 0.047003848268279275, "grad_norm": 2.8145511150360107, "learning_rate": 2.9977338754886627e-05, "loss": 0.1179, "step": 684 }, { "epoch": 0.047072567344694886, "grad_norm": 3.5207886695861816, "learning_rate": 2.997715493826727e-05, "loss": 0.1006, "step": 685 }, { "epoch": 0.0471412864211105, "grad_norm": 4.236792087554932, "learning_rate": 2.997697037971045e-05, "loss": 0.1338, "step": 686 }, { "epoch": 0.04721000549752611, "grad_norm": 4.386802673339844, "learning_rate": 2.9976785079225305e-05, "loss": 0.1088, "step": 687 }, { "epoch": 0.04727872457394173, "grad_norm": 2.6514058113098145, "learning_rate": 2.9976599036821018e-05, "loss": 0.122, "step": 688 }, { "epoch": 0.04734744365035734, "grad_norm": 2.6371917724609375, "learning_rate": 2.997641225250681e-05, "loss": 0.1095, "step": 689 }, { "epoch": 0.04741616272677295, "grad_norm": 2.9872469902038574, "learning_rate": 2.9976224726291926e-05, "loss": 0.1348, "step": 690 }, { "epoch": 0.047484881803188564, "grad_norm": 1.8449310064315796, "learning_rate": 2.997603645818566e-05, "loss": 0.0677, "step": 691 }, { "epoch": 0.047553600879604176, "grad_norm": 4.735980033874512, "learning_rate": 2.9975847448197334e-05, "loss": 0.1128, "step": 692 }, { "epoch": 0.047622319956019794, "grad_norm": 4.935771465301514, "learning_rate": 2.9975657696336317e-05, "loss": 0.1017, "step": 693 }, { "epoch": 0.047691039032435406, "grad_norm": 3.087742805480957, "learning_rate": 2.9975467202612003e-05, "loss": 0.126, "step": 694 }, { "epoch": 0.04775975810885102, "grad_norm": 5.177886009216309, "learning_rate": 2.9975275967033835e-05, "loss": 0.0981, "step": 695 }, { "epoch": 0.04782847718526663, "grad_norm": 7.773315906524658, "learning_rate": 2.9975083989611283e-05, "loss": 0.1076, "step": 696 }, { "epoch": 0.04789719626168224, "grad_norm": 2.4344875812530518, "learning_rate": 2.997489127035386e-05, "loss": 0.0965, "step": 697 }, { "epoch": 0.04796591533809785, "grad_norm": 3.653546094894409, "learning_rate": 2.997469780927111e-05, "loss": 0.1218, "step": 698 }, { "epoch": 0.04803463441451347, "grad_norm": 3.6786844730377197, "learning_rate": 2.9974503606372615e-05, "loss": 0.1042, "step": 699 }, { "epoch": 0.048103353490929084, "grad_norm": 2.525902509689331, "learning_rate": 2.9974308661668e-05, "loss": 0.1349, "step": 700 }, { "epoch": 0.048172072567344695, "grad_norm": 2.2148547172546387, "learning_rate": 2.9974112975166916e-05, "loss": 0.1098, "step": 701 }, { "epoch": 0.04824079164376031, "grad_norm": 3.153125286102295, "learning_rate": 2.9973916546879062e-05, "loss": 0.0985, "step": 702 }, { "epoch": 0.04830951072017592, "grad_norm": 1.7190907001495361, "learning_rate": 2.997371937681417e-05, "loss": 0.1018, "step": 703 }, { "epoch": 0.04837822979659153, "grad_norm": 3.499164342880249, "learning_rate": 2.9973521464982004e-05, "loss": 0.1025, "step": 704 }, { "epoch": 0.04844694887300715, "grad_norm": 2.903007745742798, "learning_rate": 2.997332281139237e-05, "loss": 0.0853, "step": 705 }, { "epoch": 0.04851566794942276, "grad_norm": 3.5876097679138184, "learning_rate": 2.9973123416055098e-05, "loss": 0.1387, "step": 706 }, { "epoch": 0.04858438702583837, "grad_norm": 2.303130626678467, "learning_rate": 2.9972923278980088e-05, "loss": 0.0906, "step": 707 }, { "epoch": 0.048653106102253985, "grad_norm": 3.5312469005584717, "learning_rate": 2.9972722400177235e-05, "loss": 0.1046, "step": 708 }, { "epoch": 0.048721825178669596, "grad_norm": 2.9825758934020996, "learning_rate": 2.99725207796565e-05, "loss": 0.1126, "step": 709 }, { "epoch": 0.048790544255085215, "grad_norm": 5.0606818199157715, "learning_rate": 2.9972318417427866e-05, "loss": 0.1323, "step": 710 }, { "epoch": 0.04885926333150083, "grad_norm": 5.067862510681152, "learning_rate": 2.997211531350136e-05, "loss": 0.1376, "step": 711 }, { "epoch": 0.04892798240791644, "grad_norm": 3.2815144062042236, "learning_rate": 2.9971911467887045e-05, "loss": 0.1253, "step": 712 }, { "epoch": 0.04899670148433205, "grad_norm": 2.9604554176330566, "learning_rate": 2.9971706880595008e-05, "loss": 0.105, "step": 713 }, { "epoch": 0.04906542056074766, "grad_norm": 2.223402976989746, "learning_rate": 2.99715015516354e-05, "loss": 0.1498, "step": 714 }, { "epoch": 0.049134139637163274, "grad_norm": 3.3231945037841797, "learning_rate": 2.9971295481018388e-05, "loss": 0.1285, "step": 715 }, { "epoch": 0.04920285871357889, "grad_norm": 2.721881151199341, "learning_rate": 2.9971088668754174e-05, "loss": 0.1014, "step": 716 }, { "epoch": 0.049271577789994504, "grad_norm": 2.9034531116485596, "learning_rate": 2.9970881114853005e-05, "loss": 0.1238, "step": 717 }, { "epoch": 0.049340296866410116, "grad_norm": 3.2058234214782715, "learning_rate": 2.9970672819325166e-05, "loss": 0.1091, "step": 718 }, { "epoch": 0.04940901594282573, "grad_norm": 1.8299592733383179, "learning_rate": 2.9970463782180973e-05, "loss": 0.1036, "step": 719 }, { "epoch": 0.04947773501924134, "grad_norm": 1.889883041381836, "learning_rate": 2.9970254003430786e-05, "loss": 0.0937, "step": 720 }, { "epoch": 0.04954645409565695, "grad_norm": 2.45524263381958, "learning_rate": 2.997004348308499e-05, "loss": 0.1316, "step": 721 }, { "epoch": 0.04961517317207257, "grad_norm": 1.9404172897338867, "learning_rate": 2.9969832221154014e-05, "loss": 0.086, "step": 722 }, { "epoch": 0.04968389224848818, "grad_norm": 2.76347017288208, "learning_rate": 2.996962021764833e-05, "loss": 0.1274, "step": 723 }, { "epoch": 0.049752611324903794, "grad_norm": 2.4402658939361572, "learning_rate": 2.9969407472578435e-05, "loss": 0.0896, "step": 724 }, { "epoch": 0.049821330401319405, "grad_norm": 2.914762020111084, "learning_rate": 2.996919398595487e-05, "loss": 0.1164, "step": 725 }, { "epoch": 0.04989004947773502, "grad_norm": 2.7688772678375244, "learning_rate": 2.9968979757788214e-05, "loss": 0.1119, "step": 726 }, { "epoch": 0.04995876855415063, "grad_norm": 1.976208209991455, "learning_rate": 2.996876478808907e-05, "loss": 0.106, "step": 727 }, { "epoch": 0.05002748763056625, "grad_norm": 2.824897050857544, "learning_rate": 2.9968549076868098e-05, "loss": 0.1138, "step": 728 }, { "epoch": 0.05009620670698186, "grad_norm": 2.3160319328308105, "learning_rate": 2.996833262413598e-05, "loss": 0.1419, "step": 729 }, { "epoch": 0.05016492578339747, "grad_norm": 2.0143308639526367, "learning_rate": 2.996811542990343e-05, "loss": 0.1087, "step": 730 }, { "epoch": 0.05023364485981308, "grad_norm": 1.329952597618103, "learning_rate": 2.9967897494181217e-05, "loss": 0.0942, "step": 731 }, { "epoch": 0.050302363936228695, "grad_norm": 2.048495292663574, "learning_rate": 2.9967678816980138e-05, "loss": 0.125, "step": 732 }, { "epoch": 0.05037108301264431, "grad_norm": 2.400285482406616, "learning_rate": 2.9967459398311025e-05, "loss": 0.0919, "step": 733 }, { "epoch": 0.050439802089059925, "grad_norm": 2.614089012145996, "learning_rate": 2.996723923818474e-05, "loss": 0.1196, "step": 734 }, { "epoch": 0.05050852116547554, "grad_norm": 2.6183559894561768, "learning_rate": 2.9967018336612193e-05, "loss": 0.1199, "step": 735 }, { "epoch": 0.05057724024189115, "grad_norm": 3.233065605163574, "learning_rate": 2.996679669360433e-05, "loss": 0.109, "step": 736 }, { "epoch": 0.05064595931830676, "grad_norm": 3.2809059619903564, "learning_rate": 2.9966574309172133e-05, "loss": 0.1059, "step": 737 }, { "epoch": 0.05071467839472237, "grad_norm": 3.3712966442108154, "learning_rate": 2.996635118332661e-05, "loss": 0.1048, "step": 738 }, { "epoch": 0.05078339747113799, "grad_norm": 2.6072731018066406, "learning_rate": 2.9966127316078823e-05, "loss": 0.1211, "step": 739 }, { "epoch": 0.0508521165475536, "grad_norm": 3.061037540435791, "learning_rate": 2.9965902707439857e-05, "loss": 0.1049, "step": 740 }, { "epoch": 0.050920835623969214, "grad_norm": 3.8745768070220947, "learning_rate": 2.9965677357420836e-05, "loss": 0.1387, "step": 741 }, { "epoch": 0.050989554700384826, "grad_norm": 3.1730830669403076, "learning_rate": 2.9965451266032932e-05, "loss": 0.091, "step": 742 }, { "epoch": 0.05105827377680044, "grad_norm": 2.403613567352295, "learning_rate": 2.9965224433287335e-05, "loss": 0.0866, "step": 743 }, { "epoch": 0.05112699285321605, "grad_norm": 2.7800371646881104, "learning_rate": 2.9964996859195288e-05, "loss": 0.0853, "step": 744 }, { "epoch": 0.05119571192963167, "grad_norm": 2.928532361984253, "learning_rate": 2.9964768543768065e-05, "loss": 0.0949, "step": 745 }, { "epoch": 0.05126443100604728, "grad_norm": 3.4921681880950928, "learning_rate": 2.9964539487016973e-05, "loss": 0.1156, "step": 746 }, { "epoch": 0.05133315008246289, "grad_norm": 2.655487060546875, "learning_rate": 2.9964309688953362e-05, "loss": 0.0924, "step": 747 }, { "epoch": 0.0514018691588785, "grad_norm": 3.14526104927063, "learning_rate": 2.9964079149588613e-05, "loss": 0.1249, "step": 748 }, { "epoch": 0.051470588235294115, "grad_norm": 3.53985333442688, "learning_rate": 2.9963847868934154e-05, "loss": 0.1212, "step": 749 }, { "epoch": 0.051539307311709734, "grad_norm": 2.5671274662017822, "learning_rate": 2.9963615847001427e-05, "loss": 0.111, "step": 750 }, { "epoch": 0.051608026388125346, "grad_norm": 2.717520236968994, "learning_rate": 2.996338308380194e-05, "loss": 0.0994, "step": 751 }, { "epoch": 0.05167674546454096, "grad_norm": 2.99822735786438, "learning_rate": 2.9963149579347214e-05, "loss": 0.1196, "step": 752 }, { "epoch": 0.05174546454095657, "grad_norm": 3.3753750324249268, "learning_rate": 2.9962915333648824e-05, "loss": 0.117, "step": 753 }, { "epoch": 0.05181418361737218, "grad_norm": 2.9704091548919678, "learning_rate": 2.996268034671837e-05, "loss": 0.0988, "step": 754 }, { "epoch": 0.05188290269378779, "grad_norm": 2.3858346939086914, "learning_rate": 2.9962444618567494e-05, "loss": 0.097, "step": 755 }, { "epoch": 0.05195162177020341, "grad_norm": 2.160170555114746, "learning_rate": 2.9962208149207874e-05, "loss": 0.1127, "step": 756 }, { "epoch": 0.05202034084661902, "grad_norm": 3.330066442489624, "learning_rate": 2.9961970938651222e-05, "loss": 0.1193, "step": 757 }, { "epoch": 0.052089059923034635, "grad_norm": 4.444832801818848, "learning_rate": 2.9961732986909294e-05, "loss": 0.123, "step": 758 }, { "epoch": 0.05215777899945025, "grad_norm": 5.363932132720947, "learning_rate": 2.996149429399387e-05, "loss": 0.0946, "step": 759 }, { "epoch": 0.05222649807586586, "grad_norm": 2.8068532943725586, "learning_rate": 2.9961254859916777e-05, "loss": 0.1272, "step": 760 }, { "epoch": 0.05229521715228147, "grad_norm": 4.405323505401611, "learning_rate": 2.9961014684689876e-05, "loss": 0.1198, "step": 761 }, { "epoch": 0.05236393622869709, "grad_norm": 4.299930095672607, "learning_rate": 2.9960773768325073e-05, "loss": 0.1114, "step": 762 }, { "epoch": 0.0524326553051127, "grad_norm": 2.2484498023986816, "learning_rate": 2.996053211083429e-05, "loss": 0.118, "step": 763 }, { "epoch": 0.05250137438152831, "grad_norm": 3.1540040969848633, "learning_rate": 2.9960289712229505e-05, "loss": 0.1109, "step": 764 }, { "epoch": 0.052570093457943924, "grad_norm": 2.22041392326355, "learning_rate": 2.9960046572522722e-05, "loss": 0.1066, "step": 765 }, { "epoch": 0.052638812534359536, "grad_norm": 2.872384548187256, "learning_rate": 2.9959802691725993e-05, "loss": 0.1257, "step": 766 }, { "epoch": 0.052707531610775155, "grad_norm": 2.447678327560425, "learning_rate": 2.9959558069851397e-05, "loss": 0.1084, "step": 767 }, { "epoch": 0.052776250687190766, "grad_norm": 4.414535045623779, "learning_rate": 2.9959312706911044e-05, "loss": 0.1147, "step": 768 }, { "epoch": 0.05284496976360638, "grad_norm": 4.872952461242676, "learning_rate": 2.9959066602917093e-05, "loss": 0.147, "step": 769 }, { "epoch": 0.05291368884002199, "grad_norm": 2.447445869445801, "learning_rate": 2.995881975788174e-05, "loss": 0.0968, "step": 770 }, { "epoch": 0.0529824079164376, "grad_norm": 2.4139339923858643, "learning_rate": 2.9958572171817213e-05, "loss": 0.1117, "step": 771 }, { "epoch": 0.05305112699285321, "grad_norm": 2.484236478805542, "learning_rate": 2.9958323844735774e-05, "loss": 0.1154, "step": 772 }, { "epoch": 0.05311984606926883, "grad_norm": 2.4876461029052734, "learning_rate": 2.995807477664972e-05, "loss": 0.1101, "step": 773 }, { "epoch": 0.053188565145684444, "grad_norm": 2.699968099594116, "learning_rate": 2.99578249675714e-05, "loss": 0.0957, "step": 774 }, { "epoch": 0.053257284222100056, "grad_norm": 2.8864803314208984, "learning_rate": 2.9957574417513177e-05, "loss": 0.0906, "step": 775 }, { "epoch": 0.05332600329851567, "grad_norm": 2.657778739929199, "learning_rate": 2.9957323126487472e-05, "loss": 0.1005, "step": 776 }, { "epoch": 0.05339472237493128, "grad_norm": 1.9425623416900635, "learning_rate": 2.9957071094506728e-05, "loss": 0.0779, "step": 777 }, { "epoch": 0.05346344145134689, "grad_norm": 1.904491901397705, "learning_rate": 2.9956818321583435e-05, "loss": 0.1306, "step": 778 }, { "epoch": 0.05353216052776251, "grad_norm": 2.4733619689941406, "learning_rate": 2.9956564807730112e-05, "loss": 0.0911, "step": 779 }, { "epoch": 0.05360087960417812, "grad_norm": 3.499681234359741, "learning_rate": 2.9956310552959317e-05, "loss": 0.1029, "step": 780 }, { "epoch": 0.05366959868059373, "grad_norm": 2.2670822143554688, "learning_rate": 2.9956055557283647e-05, "loss": 0.1678, "step": 781 }, { "epoch": 0.053738317757009345, "grad_norm": 2.951045036315918, "learning_rate": 2.995579982071573e-05, "loss": 0.121, "step": 782 }, { "epoch": 0.05380703683342496, "grad_norm": 1.7880043983459473, "learning_rate": 2.9955543343268243e-05, "loss": 0.0916, "step": 783 }, { "epoch": 0.05387575590984057, "grad_norm": 2.5576300621032715, "learning_rate": 2.9955286124953883e-05, "loss": 0.1282, "step": 784 }, { "epoch": 0.05394447498625619, "grad_norm": 3.088212013244629, "learning_rate": 2.9955028165785392e-05, "loss": 0.1055, "step": 785 }, { "epoch": 0.0540131940626718, "grad_norm": 4.366748332977295, "learning_rate": 2.9954769465775556e-05, "loss": 0.0897, "step": 786 }, { "epoch": 0.05408191313908741, "grad_norm": 2.283332586288452, "learning_rate": 2.9954510024937185e-05, "loss": 0.095, "step": 787 }, { "epoch": 0.05415063221550302, "grad_norm": 3.0427558422088623, "learning_rate": 2.9954249843283135e-05, "loss": 0.0932, "step": 788 }, { "epoch": 0.054219351291918634, "grad_norm": 2.3616762161254883, "learning_rate": 2.995398892082629e-05, "loss": 0.0814, "step": 789 }, { "epoch": 0.05428807036833425, "grad_norm": 3.4076359272003174, "learning_rate": 2.9953727257579576e-05, "loss": 0.1072, "step": 790 }, { "epoch": 0.054356789444749865, "grad_norm": 2.8606648445129395, "learning_rate": 2.995346485355596e-05, "loss": 0.0811, "step": 791 }, { "epoch": 0.054425508521165476, "grad_norm": 2.9442999362945557, "learning_rate": 2.995320170876844e-05, "loss": 0.0958, "step": 792 }, { "epoch": 0.05449422759758109, "grad_norm": 2.1834683418273926, "learning_rate": 2.9952937823230048e-05, "loss": 0.1137, "step": 793 }, { "epoch": 0.0545629466739967, "grad_norm": 1.9369546175003052, "learning_rate": 2.9952673196953854e-05, "loss": 0.1034, "step": 794 }, { "epoch": 0.05463166575041231, "grad_norm": 2.900301456451416, "learning_rate": 2.995240782995298e-05, "loss": 0.0891, "step": 795 }, { "epoch": 0.05470038482682793, "grad_norm": 2.425760269165039, "learning_rate": 2.9952141722240555e-05, "loss": 0.0849, "step": 796 }, { "epoch": 0.05476910390324354, "grad_norm": 3.4682328701019287, "learning_rate": 2.9951874873829776e-05, "loss": 0.0766, "step": 797 }, { "epoch": 0.054837822979659154, "grad_norm": 2.173577308654785, "learning_rate": 2.995160728473385e-05, "loss": 0.1028, "step": 798 }, { "epoch": 0.054906542056074766, "grad_norm": 2.4643616676330566, "learning_rate": 2.9951338954966042e-05, "loss": 0.0896, "step": 799 }, { "epoch": 0.05497526113249038, "grad_norm": 2.3184618949890137, "learning_rate": 2.9951069884539634e-05, "loss": 0.1212, "step": 800 }, { "epoch": 0.05504398020890599, "grad_norm": 2.5010857582092285, "learning_rate": 2.995080007346797e-05, "loss": 0.1011, "step": 801 }, { "epoch": 0.05511269928532161, "grad_norm": 3.3774917125701904, "learning_rate": 2.99505295217644e-05, "loss": 0.0991, "step": 802 }, { "epoch": 0.05518141836173722, "grad_norm": 3.000581741333008, "learning_rate": 2.9950258229442342e-05, "loss": 0.1338, "step": 803 }, { "epoch": 0.05525013743815283, "grad_norm": 2.778444528579712, "learning_rate": 2.9949986196515226e-05, "loss": 0.0845, "step": 804 }, { "epoch": 0.05531885651456844, "grad_norm": 2.2092716693878174, "learning_rate": 2.9949713422996532e-05, "loss": 0.1138, "step": 805 }, { "epoch": 0.055387575590984055, "grad_norm": 2.9582698345184326, "learning_rate": 2.9949439908899768e-05, "loss": 0.1442, "step": 806 }, { "epoch": 0.05545629466739967, "grad_norm": 1.8930532932281494, "learning_rate": 2.994916565423848e-05, "loss": 0.1001, "step": 807 }, { "epoch": 0.055525013743815285, "grad_norm": 1.9916300773620605, "learning_rate": 2.9948890659026263e-05, "loss": 0.0788, "step": 808 }, { "epoch": 0.0555937328202309, "grad_norm": 1.928839921951294, "learning_rate": 2.9948614923276742e-05, "loss": 0.0729, "step": 809 }, { "epoch": 0.05566245189664651, "grad_norm": 2.7619400024414062, "learning_rate": 2.9948338447003565e-05, "loss": 0.0824, "step": 810 }, { "epoch": 0.05573117097306212, "grad_norm": 2.6236870288848877, "learning_rate": 2.9948061230220437e-05, "loss": 0.0751, "step": 811 }, { "epoch": 0.05579989004947773, "grad_norm": 1.751259207725525, "learning_rate": 2.994778327294109e-05, "loss": 0.0842, "step": 812 }, { "epoch": 0.05586860912589335, "grad_norm": 5.789602756500244, "learning_rate": 2.9947504575179283e-05, "loss": 0.0669, "step": 813 }, { "epoch": 0.05593732820230896, "grad_norm": 2.6662206649780273, "learning_rate": 2.9947225136948834e-05, "loss": 0.1133, "step": 814 }, { "epoch": 0.056006047278724574, "grad_norm": 2.528543472290039, "learning_rate": 2.9946944958263582e-05, "loss": 0.0977, "step": 815 }, { "epoch": 0.056074766355140186, "grad_norm": 3.9017438888549805, "learning_rate": 2.9946664039137402e-05, "loss": 0.0921, "step": 816 }, { "epoch": 0.0561434854315558, "grad_norm": 2.536738395690918, "learning_rate": 2.994638237958422e-05, "loss": 0.0903, "step": 817 }, { "epoch": 0.05621220450797141, "grad_norm": 2.4789912700653076, "learning_rate": 2.9946099979617986e-05, "loss": 0.1196, "step": 818 }, { "epoch": 0.05628092358438703, "grad_norm": 2.4855494499206543, "learning_rate": 2.9945816839252683e-05, "loss": 0.0906, "step": 819 }, { "epoch": 0.05634964266080264, "grad_norm": 2.2565524578094482, "learning_rate": 2.994553295850234e-05, "loss": 0.0954, "step": 820 }, { "epoch": 0.05641836173721825, "grad_norm": 3.1620230674743652, "learning_rate": 2.9945248337381023e-05, "loss": 0.0985, "step": 821 }, { "epoch": 0.056487080813633864, "grad_norm": 2.2642276287078857, "learning_rate": 2.994496297590283e-05, "loss": 0.0956, "step": 822 }, { "epoch": 0.056555799890049475, "grad_norm": 3.1903021335601807, "learning_rate": 2.9944676874081893e-05, "loss": 0.1165, "step": 823 }, { "epoch": 0.056624518966465094, "grad_norm": 2.7800774574279785, "learning_rate": 2.994439003193239e-05, "loss": 0.1216, "step": 824 }, { "epoch": 0.056693238042880706, "grad_norm": 2.0836093425750732, "learning_rate": 2.9944102449468532e-05, "loss": 0.0983, "step": 825 }, { "epoch": 0.05676195711929632, "grad_norm": 2.912212371826172, "learning_rate": 2.9943814126704563e-05, "loss": 0.127, "step": 826 }, { "epoch": 0.05683067619571193, "grad_norm": 1.9600672721862793, "learning_rate": 2.994352506365476e-05, "loss": 0.0857, "step": 827 }, { "epoch": 0.05689939527212754, "grad_norm": 1.7663404941558838, "learning_rate": 2.9943235260333453e-05, "loss": 0.0742, "step": 828 }, { "epoch": 0.05696811434854315, "grad_norm": 2.769035577774048, "learning_rate": 2.9942944716754992e-05, "loss": 0.0928, "step": 829 }, { "epoch": 0.05703683342495877, "grad_norm": 1.9186112880706787, "learning_rate": 2.9942653432933768e-05, "loss": 0.0956, "step": 830 }, { "epoch": 0.05710555250137438, "grad_norm": 2.8786630630493164, "learning_rate": 2.9942361408884216e-05, "loss": 0.0997, "step": 831 }, { "epoch": 0.057174271577789995, "grad_norm": 2.633418321609497, "learning_rate": 2.9942068644620802e-05, "loss": 0.1238, "step": 832 }, { "epoch": 0.05724299065420561, "grad_norm": 2.5109026432037354, "learning_rate": 2.994177514015802e-05, "loss": 0.0901, "step": 833 }, { "epoch": 0.05731170973062122, "grad_norm": 2.275055408477783, "learning_rate": 2.9941480895510428e-05, "loss": 0.1005, "step": 834 }, { "epoch": 0.05738042880703683, "grad_norm": 2.835205078125, "learning_rate": 2.9941185910692582e-05, "loss": 0.1153, "step": 835 }, { "epoch": 0.05744914788345245, "grad_norm": 2.894425392150879, "learning_rate": 2.9940890185719105e-05, "loss": 0.1039, "step": 836 }, { "epoch": 0.05751786695986806, "grad_norm": 1.9282283782958984, "learning_rate": 2.9940593720604647e-05, "loss": 0.0971, "step": 837 }, { "epoch": 0.05758658603628367, "grad_norm": 1.908587098121643, "learning_rate": 2.994029651536389e-05, "loss": 0.1062, "step": 838 }, { "epoch": 0.057655305112699284, "grad_norm": 1.941319465637207, "learning_rate": 2.9939998570011563e-05, "loss": 0.0975, "step": 839 }, { "epoch": 0.057724024189114896, "grad_norm": 1.977638602256775, "learning_rate": 2.993969988456242e-05, "loss": 0.0919, "step": 840 }, { "epoch": 0.05779274326553051, "grad_norm": 2.735022783279419, "learning_rate": 2.9939400459031264e-05, "loss": 0.0883, "step": 841 }, { "epoch": 0.05786146234194613, "grad_norm": 2.3624091148376465, "learning_rate": 2.9939100293432917e-05, "loss": 0.0973, "step": 842 }, { "epoch": 0.05793018141836174, "grad_norm": 2.69104266166687, "learning_rate": 2.9938799387782257e-05, "loss": 0.1091, "step": 843 }, { "epoch": 0.05799890049477735, "grad_norm": 2.7459614276885986, "learning_rate": 2.9938497742094188e-05, "loss": 0.0912, "step": 844 }, { "epoch": 0.05806761957119296, "grad_norm": 2.1227450370788574, "learning_rate": 2.9938195356383658e-05, "loss": 0.0872, "step": 845 }, { "epoch": 0.058136338647608574, "grad_norm": 2.0348424911499023, "learning_rate": 2.9937892230665633e-05, "loss": 0.0853, "step": 846 }, { "epoch": 0.05820505772402419, "grad_norm": 2.762383222579956, "learning_rate": 2.9937588364955143e-05, "loss": 0.0867, "step": 847 }, { "epoch": 0.058273776800439804, "grad_norm": 2.8377444744110107, "learning_rate": 2.9937283759267238e-05, "loss": 0.1228, "step": 848 }, { "epoch": 0.058342495876855416, "grad_norm": 3.2066519260406494, "learning_rate": 2.9936978413617e-05, "loss": 0.0949, "step": 849 }, { "epoch": 0.05841121495327103, "grad_norm": 2.32436203956604, "learning_rate": 2.9936672328019564e-05, "loss": 0.0968, "step": 850 }, { "epoch": 0.05847993402968664, "grad_norm": 2.1929244995117188, "learning_rate": 2.993636550249009e-05, "loss": 0.077, "step": 851 }, { "epoch": 0.05854865310610225, "grad_norm": 1.6907763481140137, "learning_rate": 2.9936057937043773e-05, "loss": 0.075, "step": 852 }, { "epoch": 0.05861737218251787, "grad_norm": 2.20955753326416, "learning_rate": 2.9935749631695857e-05, "loss": 0.0878, "step": 853 }, { "epoch": 0.05868609125893348, "grad_norm": 2.3409368991851807, "learning_rate": 2.993544058646161e-05, "loss": 0.086, "step": 854 }, { "epoch": 0.05875481033534909, "grad_norm": 2.4924752712249756, "learning_rate": 2.993513080135634e-05, "loss": 0.0555, "step": 855 }, { "epoch": 0.058823529411764705, "grad_norm": 2.5242133140563965, "learning_rate": 2.9934820276395397e-05, "loss": 0.0889, "step": 856 }, { "epoch": 0.05889224848818032, "grad_norm": 2.316401481628418, "learning_rate": 2.993450901159416e-05, "loss": 0.0836, "step": 857 }, { "epoch": 0.05896096756459593, "grad_norm": 2.49519419670105, "learning_rate": 2.9934197006968053e-05, "loss": 0.0915, "step": 858 }, { "epoch": 0.05902968664101155, "grad_norm": 1.9880847930908203, "learning_rate": 2.993388426253253e-05, "loss": 0.0917, "step": 859 }, { "epoch": 0.05909840571742716, "grad_norm": 2.3903043270111084, "learning_rate": 2.9933570778303083e-05, "loss": 0.0779, "step": 860 }, { "epoch": 0.05916712479384277, "grad_norm": 1.7417019605636597, "learning_rate": 2.9933256554295243e-05, "loss": 0.0724, "step": 861 }, { "epoch": 0.05923584387025838, "grad_norm": 1.8773279190063477, "learning_rate": 2.9932941590524572e-05, "loss": 0.0837, "step": 862 }, { "epoch": 0.059304562946673994, "grad_norm": 1.9991687536239624, "learning_rate": 2.993262588700668e-05, "loss": 0.0896, "step": 863 }, { "epoch": 0.05937328202308961, "grad_norm": 3.3438665866851807, "learning_rate": 2.9932309443757196e-05, "loss": 0.1014, "step": 864 }, { "epoch": 0.059442001099505225, "grad_norm": 3.6718642711639404, "learning_rate": 2.9931992260791804e-05, "loss": 0.1136, "step": 865 }, { "epoch": 0.059510720175920837, "grad_norm": 2.666943073272705, "learning_rate": 2.9931674338126214e-05, "loss": 0.1106, "step": 866 }, { "epoch": 0.05957943925233645, "grad_norm": 1.8237333297729492, "learning_rate": 2.9931355675776177e-05, "loss": 0.0693, "step": 867 }, { "epoch": 0.05964815832875206, "grad_norm": 2.826631546020508, "learning_rate": 2.9931036273757474e-05, "loss": 0.085, "step": 868 }, { "epoch": 0.05971687740516767, "grad_norm": 2.8149147033691406, "learning_rate": 2.9930716132085933e-05, "loss": 0.0952, "step": 869 }, { "epoch": 0.05978559648158329, "grad_norm": 2.8275415897369385, "learning_rate": 2.9930395250777416e-05, "loss": 0.1037, "step": 870 }, { "epoch": 0.0598543155579989, "grad_norm": 2.5551505088806152, "learning_rate": 2.9930073629847807e-05, "loss": 0.0902, "step": 871 }, { "epoch": 0.059923034634414514, "grad_norm": 3.1029069423675537, "learning_rate": 2.9929751269313047e-05, "loss": 0.1129, "step": 872 }, { "epoch": 0.059991753710830126, "grad_norm": 1.9971274137496948, "learning_rate": 2.9929428169189107e-05, "loss": 0.0803, "step": 873 }, { "epoch": 0.06006047278724574, "grad_norm": 1.8479725122451782, "learning_rate": 2.992910432949198e-05, "loss": 0.102, "step": 874 }, { "epoch": 0.06012919186366135, "grad_norm": 4.355798721313477, "learning_rate": 2.992877975023773e-05, "loss": 0.0932, "step": 875 }, { "epoch": 0.06019791094007697, "grad_norm": 3.3849079608917236, "learning_rate": 2.9928454431442412e-05, "loss": 0.1147, "step": 876 }, { "epoch": 0.06026663001649258, "grad_norm": 2.7940125465393066, "learning_rate": 2.9928128373122163e-05, "loss": 0.11, "step": 877 }, { "epoch": 0.06033534909290819, "grad_norm": 2.5787670612335205, "learning_rate": 2.9927801575293117e-05, "loss": 0.1143, "step": 878 }, { "epoch": 0.0604040681693238, "grad_norm": 2.6621789932250977, "learning_rate": 2.9927474037971474e-05, "loss": 0.0746, "step": 879 }, { "epoch": 0.060472787245739415, "grad_norm": 3.043715715408325, "learning_rate": 2.992714576117346e-05, "loss": 0.1249, "step": 880 }, { "epoch": 0.060541506322155034, "grad_norm": 3.3311352729797363, "learning_rate": 2.992681674491533e-05, "loss": 0.0956, "step": 881 }, { "epoch": 0.060610225398570645, "grad_norm": 4.271401405334473, "learning_rate": 2.9926486989213388e-05, "loss": 0.1044, "step": 882 }, { "epoch": 0.06067894447498626, "grad_norm": 1.8932572603225708, "learning_rate": 2.992615649408397e-05, "loss": 0.1242, "step": 883 }, { "epoch": 0.06074766355140187, "grad_norm": 3.2829864025115967, "learning_rate": 2.992582525954344e-05, "loss": 0.1114, "step": 884 }, { "epoch": 0.06081638262781748, "grad_norm": 2.2374134063720703, "learning_rate": 2.9925493285608218e-05, "loss": 0.0954, "step": 885 }, { "epoch": 0.06088510170423309, "grad_norm": 3.1802449226379395, "learning_rate": 2.9925160572294745e-05, "loss": 0.1271, "step": 886 }, { "epoch": 0.06095382078064871, "grad_norm": 3.146217107772827, "learning_rate": 2.9924827119619496e-05, "loss": 0.1111, "step": 887 }, { "epoch": 0.06102253985706432, "grad_norm": 4.443093299865723, "learning_rate": 2.9924492927599e-05, "loss": 0.0906, "step": 888 }, { "epoch": 0.061091258933479935, "grad_norm": 2.4132583141326904, "learning_rate": 2.9924157996249808e-05, "loss": 0.0833, "step": 889 }, { "epoch": 0.061159978009895546, "grad_norm": 2.226506233215332, "learning_rate": 2.992382232558851e-05, "loss": 0.0934, "step": 890 }, { "epoch": 0.06122869708631116, "grad_norm": 1.8288519382476807, "learning_rate": 2.992348591563174e-05, "loss": 0.0891, "step": 891 }, { "epoch": 0.06129741616272677, "grad_norm": 2.3609538078308105, "learning_rate": 2.992314876639616e-05, "loss": 0.0984, "step": 892 }, { "epoch": 0.06136613523914239, "grad_norm": 2.1655144691467285, "learning_rate": 2.9922810877898464e-05, "loss": 0.067, "step": 893 }, { "epoch": 0.061434854315558, "grad_norm": 2.273404121398926, "learning_rate": 2.99224722501554e-05, "loss": 0.1065, "step": 894 }, { "epoch": 0.06150357339197361, "grad_norm": 1.8811229467391968, "learning_rate": 2.9922132883183738e-05, "loss": 0.0931, "step": 895 }, { "epoch": 0.061572292468389224, "grad_norm": 2.031275749206543, "learning_rate": 2.9921792777000297e-05, "loss": 0.0848, "step": 896 }, { "epoch": 0.061641011544804836, "grad_norm": 3.0837557315826416, "learning_rate": 2.9921451931621914e-05, "loss": 0.1032, "step": 897 }, { "epoch": 0.06170973062122045, "grad_norm": 1.855838418006897, "learning_rate": 2.9921110347065483e-05, "loss": 0.0882, "step": 898 }, { "epoch": 0.061778449697636066, "grad_norm": 2.8757383823394775, "learning_rate": 2.9920768023347923e-05, "loss": 0.0922, "step": 899 }, { "epoch": 0.06184716877405168, "grad_norm": 2.046656847000122, "learning_rate": 2.992042496048619e-05, "loss": 0.0769, "step": 900 }, { "epoch": 0.06191588785046729, "grad_norm": 2.849346399307251, "learning_rate": 2.992008115849728e-05, "loss": 0.0992, "step": 901 }, { "epoch": 0.0619846069268829, "grad_norm": 4.271240234375, "learning_rate": 2.9919736617398222e-05, "loss": 0.0962, "step": 902 }, { "epoch": 0.06205332600329851, "grad_norm": 2.661719560623169, "learning_rate": 2.9919391337206082e-05, "loss": 0.0996, "step": 903 }, { "epoch": 0.06212204507971413, "grad_norm": 1.6221225261688232, "learning_rate": 2.9919045317937976e-05, "loss": 0.0758, "step": 904 }, { "epoch": 0.062190764156129744, "grad_norm": 2.0700974464416504, "learning_rate": 2.9918698559611033e-05, "loss": 0.0879, "step": 905 }, { "epoch": 0.062259483232545355, "grad_norm": 2.6182665824890137, "learning_rate": 2.991835106224243e-05, "loss": 0.105, "step": 906 }, { "epoch": 0.06232820230896097, "grad_norm": 2.7954633235931396, "learning_rate": 2.9918002825849398e-05, "loss": 0.0838, "step": 907 }, { "epoch": 0.06239692138537658, "grad_norm": 2.6861979961395264, "learning_rate": 2.9917653850449168e-05, "loss": 0.1015, "step": 908 }, { "epoch": 0.06246564046179219, "grad_norm": 2.226592779159546, "learning_rate": 2.9917304136059038e-05, "loss": 0.083, "step": 909 }, { "epoch": 0.06253435953820781, "grad_norm": 3.870797634124756, "learning_rate": 2.9916953682696328e-05, "loss": 0.0885, "step": 910 }, { "epoch": 0.06260307861462341, "grad_norm": 2.8626291751861572, "learning_rate": 2.99166024903784e-05, "loss": 0.1036, "step": 911 }, { "epoch": 0.06267179769103903, "grad_norm": 3.182389974594116, "learning_rate": 2.9916250559122654e-05, "loss": 0.0929, "step": 912 }, { "epoch": 0.06274051676745465, "grad_norm": 3.3900387287139893, "learning_rate": 2.9915897888946524e-05, "loss": 0.0814, "step": 913 }, { "epoch": 0.06280923584387026, "grad_norm": 3.955085277557373, "learning_rate": 2.9915544479867477e-05, "loss": 0.1147, "step": 914 }, { "epoch": 0.06287795492028588, "grad_norm": 3.0846030712127686, "learning_rate": 2.9915190331903013e-05, "loss": 0.0998, "step": 915 }, { "epoch": 0.06294667399670148, "grad_norm": 3.053560495376587, "learning_rate": 2.9914835445070693e-05, "loss": 0.1006, "step": 916 }, { "epoch": 0.0630153930731171, "grad_norm": 2.5688905715942383, "learning_rate": 2.991447981938809e-05, "loss": 0.1062, "step": 917 }, { "epoch": 0.0630841121495327, "grad_norm": 5.129382610321045, "learning_rate": 2.9914123454872815e-05, "loss": 0.1032, "step": 918 }, { "epoch": 0.06315283122594832, "grad_norm": 2.452625274658203, "learning_rate": 2.9913766351542527e-05, "loss": 0.1205, "step": 919 }, { "epoch": 0.06322155030236394, "grad_norm": 3.0296363830566406, "learning_rate": 2.9913408509414917e-05, "loss": 0.1457, "step": 920 }, { "epoch": 0.06329026937877955, "grad_norm": 2.8085432052612305, "learning_rate": 2.9913049928507712e-05, "loss": 0.0834, "step": 921 }, { "epoch": 0.06335898845519516, "grad_norm": 2.877746343612671, "learning_rate": 2.9912690608838668e-05, "loss": 0.0893, "step": 922 }, { "epoch": 0.06342770753161077, "grad_norm": 2.22729754447937, "learning_rate": 2.9912330550425594e-05, "loss": 0.0966, "step": 923 }, { "epoch": 0.06349642660802639, "grad_norm": 2.222036361694336, "learning_rate": 2.9911969753286325e-05, "loss": 0.1091, "step": 924 }, { "epoch": 0.063565145684442, "grad_norm": 2.2894675731658936, "learning_rate": 2.9911608217438726e-05, "loss": 0.0991, "step": 925 }, { "epoch": 0.06363386476085761, "grad_norm": 2.668795108795166, "learning_rate": 2.9911245942900715e-05, "loss": 0.0941, "step": 926 }, { "epoch": 0.06370258383727323, "grad_norm": 2.5117592811584473, "learning_rate": 2.991088292969024e-05, "loss": 0.098, "step": 927 }, { "epoch": 0.06377130291368883, "grad_norm": 3.6717300415039062, "learning_rate": 2.991051917782528e-05, "loss": 0.0985, "step": 928 }, { "epoch": 0.06384002199010445, "grad_norm": 2.3392105102539062, "learning_rate": 2.991015468732385e-05, "loss": 0.0839, "step": 929 }, { "epoch": 0.06390874106652007, "grad_norm": 1.86212158203125, "learning_rate": 2.9909789458204014e-05, "loss": 0.0922, "step": 930 }, { "epoch": 0.06397746014293568, "grad_norm": 1.7458009719848633, "learning_rate": 2.990942349048386e-05, "loss": 0.092, "step": 931 }, { "epoch": 0.0640461792193513, "grad_norm": 1.3424614667892456, "learning_rate": 2.990905678418152e-05, "loss": 0.0817, "step": 932 }, { "epoch": 0.0641148982957669, "grad_norm": 1.7877709865570068, "learning_rate": 2.9908689339315158e-05, "loss": 0.1121, "step": 933 }, { "epoch": 0.06418361737218252, "grad_norm": 1.7479221820831299, "learning_rate": 2.9908321155902977e-05, "loss": 0.0965, "step": 934 }, { "epoch": 0.06425233644859812, "grad_norm": 2.5267128944396973, "learning_rate": 2.9907952233963213e-05, "loss": 0.0908, "step": 935 }, { "epoch": 0.06432105552501374, "grad_norm": 2.502634286880493, "learning_rate": 2.990758257351415e-05, "loss": 0.0969, "step": 936 }, { "epoch": 0.06438977460142936, "grad_norm": 2.146109104156494, "learning_rate": 2.9907212174574094e-05, "loss": 0.086, "step": 937 }, { "epoch": 0.06445849367784497, "grad_norm": 1.9707906246185303, "learning_rate": 2.990684103716139e-05, "loss": 0.0909, "step": 938 }, { "epoch": 0.06452721275426059, "grad_norm": 2.639761447906494, "learning_rate": 2.9906469161294434e-05, "loss": 0.0918, "step": 939 }, { "epoch": 0.06459593183067619, "grad_norm": 1.9932564496994019, "learning_rate": 2.9906096546991638e-05, "loss": 0.0773, "step": 940 }, { "epoch": 0.06466465090709181, "grad_norm": 3.150486946105957, "learning_rate": 2.990572319427147e-05, "loss": 0.0997, "step": 941 }, { "epoch": 0.06473336998350743, "grad_norm": 3.674842596054077, "learning_rate": 2.9905349103152418e-05, "loss": 0.0956, "step": 942 }, { "epoch": 0.06480208905992303, "grad_norm": 2.677377462387085, "learning_rate": 2.9904974273653013e-05, "loss": 0.1055, "step": 943 }, { "epoch": 0.06487080813633865, "grad_norm": 2.1833815574645996, "learning_rate": 2.990459870579183e-05, "loss": 0.084, "step": 944 }, { "epoch": 0.06493952721275426, "grad_norm": 1.9125075340270996, "learning_rate": 2.9904222399587465e-05, "loss": 0.0816, "step": 945 }, { "epoch": 0.06500824628916987, "grad_norm": 2.1589889526367188, "learning_rate": 2.9903845355058566e-05, "loss": 0.0994, "step": 946 }, { "epoch": 0.06507696536558549, "grad_norm": 2.263845920562744, "learning_rate": 2.9903467572223803e-05, "loss": 0.0915, "step": 947 }, { "epoch": 0.0651456844420011, "grad_norm": 2.3695549964904785, "learning_rate": 2.9903089051101903e-05, "loss": 0.0868, "step": 948 }, { "epoch": 0.06521440351841672, "grad_norm": 2.5600149631500244, "learning_rate": 2.990270979171161e-05, "loss": 0.1079, "step": 949 }, { "epoch": 0.06528312259483232, "grad_norm": 2.0216591358184814, "learning_rate": 2.9902329794071717e-05, "loss": 0.0807, "step": 950 }, { "epoch": 0.06535184167124794, "grad_norm": 3.8542702198028564, "learning_rate": 2.990194905820104e-05, "loss": 0.1297, "step": 951 }, { "epoch": 0.06542056074766354, "grad_norm": 1.7542247772216797, "learning_rate": 2.9901567584118442e-05, "loss": 0.082, "step": 952 }, { "epoch": 0.06548927982407916, "grad_norm": 3.2224981784820557, "learning_rate": 2.9901185371842825e-05, "loss": 0.1132, "step": 953 }, { "epoch": 0.06555799890049478, "grad_norm": 1.8830537796020508, "learning_rate": 2.9900802421393117e-05, "loss": 0.093, "step": 954 }, { "epoch": 0.06562671797691039, "grad_norm": 2.1232681274414062, "learning_rate": 2.9900418732788296e-05, "loss": 0.0968, "step": 955 }, { "epoch": 0.065695437053326, "grad_norm": 1.4787240028381348, "learning_rate": 2.990003430604736e-05, "loss": 0.0693, "step": 956 }, { "epoch": 0.06576415612974161, "grad_norm": 2.6404523849487305, "learning_rate": 2.989964914118936e-05, "loss": 0.0929, "step": 957 }, { "epoch": 0.06583287520615723, "grad_norm": 2.38511323928833, "learning_rate": 2.9899263238233375e-05, "loss": 0.1032, "step": 958 }, { "epoch": 0.06590159428257285, "grad_norm": 2.3167715072631836, "learning_rate": 2.989887659719852e-05, "loss": 0.0981, "step": 959 }, { "epoch": 0.06597031335898845, "grad_norm": 1.9656964540481567, "learning_rate": 2.9898489218103952e-05, "loss": 0.0562, "step": 960 }, { "epoch": 0.06603903243540407, "grad_norm": 2.3484737873077393, "learning_rate": 2.9898101100968855e-05, "loss": 0.0784, "step": 961 }, { "epoch": 0.06610775151181968, "grad_norm": 4.376389026641846, "learning_rate": 2.989771224581246e-05, "loss": 0.091, "step": 962 }, { "epoch": 0.0661764705882353, "grad_norm": 2.501147508621216, "learning_rate": 2.9897322652654027e-05, "loss": 0.0762, "step": 963 }, { "epoch": 0.06624518966465091, "grad_norm": 2.1648435592651367, "learning_rate": 2.989693232151286e-05, "loss": 0.0947, "step": 964 }, { "epoch": 0.06631390874106652, "grad_norm": 2.717283248901367, "learning_rate": 2.989654125240829e-05, "loss": 0.1076, "step": 965 }, { "epoch": 0.06638262781748214, "grad_norm": 5.854992389678955, "learning_rate": 2.9896149445359692e-05, "loss": 0.1144, "step": 966 }, { "epoch": 0.06645134689389774, "grad_norm": 3.8556478023529053, "learning_rate": 2.9895756900386477e-05, "loss": 0.097, "step": 967 }, { "epoch": 0.06652006597031336, "grad_norm": 2.195192813873291, "learning_rate": 2.989536361750809e-05, "loss": 0.0814, "step": 968 }, { "epoch": 0.06658878504672897, "grad_norm": 2.8356270790100098, "learning_rate": 2.989496959674401e-05, "loss": 0.1128, "step": 969 }, { "epoch": 0.06665750412314458, "grad_norm": 3.3606839179992676, "learning_rate": 2.989457483811376e-05, "loss": 0.1033, "step": 970 }, { "epoch": 0.0667262231995602, "grad_norm": 2.7634987831115723, "learning_rate": 2.98941793416369e-05, "loss": 0.0905, "step": 971 }, { "epoch": 0.06679494227597581, "grad_norm": 4.129426002502441, "learning_rate": 2.989378310733301e-05, "loss": 0.1019, "step": 972 }, { "epoch": 0.06686366135239143, "grad_norm": 2.756744623184204, "learning_rate": 2.989338613522173e-05, "loss": 0.0897, "step": 973 }, { "epoch": 0.06693238042880703, "grad_norm": 2.8741846084594727, "learning_rate": 2.9892988425322714e-05, "loss": 0.0862, "step": 974 }, { "epoch": 0.06700109950522265, "grad_norm": 3.156949758529663, "learning_rate": 2.9892589977655676e-05, "loss": 0.1034, "step": 975 }, { "epoch": 0.06706981858163827, "grad_norm": 3.4729855060577393, "learning_rate": 2.9892190792240345e-05, "loss": 0.087, "step": 976 }, { "epoch": 0.06713853765805387, "grad_norm": 1.8695160150527954, "learning_rate": 2.9891790869096502e-05, "loss": 0.0894, "step": 977 }, { "epoch": 0.06720725673446949, "grad_norm": 2.0524816513061523, "learning_rate": 2.9891390208243957e-05, "loss": 0.0699, "step": 978 }, { "epoch": 0.0672759758108851, "grad_norm": 2.406322717666626, "learning_rate": 2.989098880970255e-05, "loss": 0.098, "step": 979 }, { "epoch": 0.06734469488730072, "grad_norm": 2.531477451324463, "learning_rate": 2.9890586673492176e-05, "loss": 0.1152, "step": 980 }, { "epoch": 0.06741341396371633, "grad_norm": 1.9710360765457153, "learning_rate": 2.9890183799632754e-05, "loss": 0.0828, "step": 981 }, { "epoch": 0.06748213304013194, "grad_norm": 1.6492676734924316, "learning_rate": 2.9889780188144237e-05, "loss": 0.0864, "step": 982 }, { "epoch": 0.06755085211654756, "grad_norm": 2.949826717376709, "learning_rate": 2.9889375839046616e-05, "loss": 0.0945, "step": 983 }, { "epoch": 0.06761957119296316, "grad_norm": 2.119957447052002, "learning_rate": 2.9888970752359935e-05, "loss": 0.1007, "step": 984 }, { "epoch": 0.06768829026937878, "grad_norm": 2.388761520385742, "learning_rate": 2.9888564928104248e-05, "loss": 0.0915, "step": 985 }, { "epoch": 0.06775700934579439, "grad_norm": 2.340548038482666, "learning_rate": 2.988815836629967e-05, "loss": 0.0911, "step": 986 }, { "epoch": 0.06782572842221, "grad_norm": 2.0265815258026123, "learning_rate": 2.9887751066966335e-05, "loss": 0.0723, "step": 987 }, { "epoch": 0.06789444749862562, "grad_norm": 2.9185738563537598, "learning_rate": 2.9887343030124417e-05, "loss": 0.0977, "step": 988 }, { "epoch": 0.06796316657504123, "grad_norm": 3.267960548400879, "learning_rate": 2.988693425579413e-05, "loss": 0.1074, "step": 989 }, { "epoch": 0.06803188565145685, "grad_norm": 1.9246410131454468, "learning_rate": 2.9886524743995728e-05, "loss": 0.0812, "step": 990 }, { "epoch": 0.06810060472787245, "grad_norm": 2.3009955883026123, "learning_rate": 2.9886114494749498e-05, "loss": 0.0766, "step": 991 }, { "epoch": 0.06816932380428807, "grad_norm": 1.7678526639938354, "learning_rate": 2.988570350807576e-05, "loss": 0.0823, "step": 992 }, { "epoch": 0.06823804288070369, "grad_norm": 2.3331689834594727, "learning_rate": 2.9885291783994874e-05, "loss": 0.0711, "step": 993 }, { "epoch": 0.0683067619571193, "grad_norm": 2.4606003761291504, "learning_rate": 2.988487932252723e-05, "loss": 0.1005, "step": 994 }, { "epoch": 0.06837548103353491, "grad_norm": 1.8149819374084473, "learning_rate": 2.9884466123693277e-05, "loss": 0.1093, "step": 995 }, { "epoch": 0.06844420010995052, "grad_norm": 2.4028570652008057, "learning_rate": 2.9884052187513463e-05, "loss": 0.0855, "step": 996 }, { "epoch": 0.06851291918636614, "grad_norm": 2.0700137615203857, "learning_rate": 2.988363751400831e-05, "loss": 0.0885, "step": 997 }, { "epoch": 0.06858163826278176, "grad_norm": 2.475687026977539, "learning_rate": 2.9883222103198353e-05, "loss": 0.0916, "step": 998 }, { "epoch": 0.06865035733919736, "grad_norm": 2.322943687438965, "learning_rate": 2.988280595510417e-05, "loss": 0.0837, "step": 999 }, { "epoch": 0.06871907641561298, "grad_norm": 2.7790563106536865, "learning_rate": 2.988238906974638e-05, "loss": 0.1112, "step": 1000 }, { "epoch": 0.06878779549202858, "grad_norm": 4.459034442901611, "learning_rate": 2.9881971447145634e-05, "loss": 0.0937, "step": 1001 }, { "epoch": 0.0688565145684442, "grad_norm": 1.9947988986968994, "learning_rate": 2.9881553087322615e-05, "loss": 0.0865, "step": 1002 }, { "epoch": 0.0689252336448598, "grad_norm": 1.6826523542404175, "learning_rate": 2.988113399029805e-05, "loss": 0.0925, "step": 1003 }, { "epoch": 0.06899395272127543, "grad_norm": 2.127340316772461, "learning_rate": 2.9880714156092704e-05, "loss": 0.0859, "step": 1004 }, { "epoch": 0.06906267179769104, "grad_norm": 1.4555394649505615, "learning_rate": 2.9880293584727372e-05, "loss": 0.0745, "step": 1005 }, { "epoch": 0.06913139087410665, "grad_norm": 3.83770751953125, "learning_rate": 2.987987227622289e-05, "loss": 0.0934, "step": 1006 }, { "epoch": 0.06920010995052227, "grad_norm": 2.536170244216919, "learning_rate": 2.9879450230600123e-05, "loss": 0.0997, "step": 1007 }, { "epoch": 0.06926882902693787, "grad_norm": 2.0735390186309814, "learning_rate": 2.9879027447879987e-05, "loss": 0.0923, "step": 1008 }, { "epoch": 0.06933754810335349, "grad_norm": 2.7743725776672363, "learning_rate": 2.9878603928083416e-05, "loss": 0.0834, "step": 1009 }, { "epoch": 0.06940626717976911, "grad_norm": 1.727285623550415, "learning_rate": 2.9878179671231402e-05, "loss": 0.0737, "step": 1010 }, { "epoch": 0.06947498625618471, "grad_norm": 3.2353897094726562, "learning_rate": 2.987775467734495e-05, "loss": 0.1101, "step": 1011 }, { "epoch": 0.06954370533260033, "grad_norm": 2.528282880783081, "learning_rate": 2.987732894644512e-05, "loss": 0.1175, "step": 1012 }, { "epoch": 0.06961242440901594, "grad_norm": 2.2640457153320312, "learning_rate": 2.9876902478553002e-05, "loss": 0.0952, "step": 1013 }, { "epoch": 0.06968114348543156, "grad_norm": 2.0164246559143066, "learning_rate": 2.9876475273689717e-05, "loss": 0.099, "step": 1014 }, { "epoch": 0.06974986256184718, "grad_norm": 1.4245723485946655, "learning_rate": 2.9876047331876435e-05, "loss": 0.1041, "step": 1015 }, { "epoch": 0.06981858163826278, "grad_norm": 2.1958653926849365, "learning_rate": 2.987561865313435e-05, "loss": 0.0808, "step": 1016 }, { "epoch": 0.0698873007146784, "grad_norm": 2.3182663917541504, "learning_rate": 2.9875189237484702e-05, "loss": 0.0716, "step": 1017 }, { "epoch": 0.069956019791094, "grad_norm": 2.251699447631836, "learning_rate": 2.9874759084948763e-05, "loss": 0.1034, "step": 1018 }, { "epoch": 0.07002473886750962, "grad_norm": 2.5201988220214844, "learning_rate": 2.987432819554784e-05, "loss": 0.0828, "step": 1019 }, { "epoch": 0.07009345794392523, "grad_norm": 2.9382989406585693, "learning_rate": 2.9873896569303273e-05, "loss": 0.0997, "step": 1020 }, { "epoch": 0.07016217702034085, "grad_norm": 1.9990320205688477, "learning_rate": 2.9873464206236452e-05, "loss": 0.0994, "step": 1021 }, { "epoch": 0.07023089609675646, "grad_norm": 2.266711950302124, "learning_rate": 2.9873031106368797e-05, "loss": 0.0871, "step": 1022 }, { "epoch": 0.07029961517317207, "grad_norm": 2.4669106006622314, "learning_rate": 2.9872597269721756e-05, "loss": 0.1074, "step": 1023 }, { "epoch": 0.07036833424958769, "grad_norm": 2.786977767944336, "learning_rate": 2.9872162696316822e-05, "loss": 0.0917, "step": 1024 }, { "epoch": 0.0704370533260033, "grad_norm": 4.403657913208008, "learning_rate": 2.987172738617552e-05, "loss": 0.0935, "step": 1025 }, { "epoch": 0.07050577240241891, "grad_norm": 3.5517377853393555, "learning_rate": 2.9871291339319425e-05, "loss": 0.0932, "step": 1026 }, { "epoch": 0.07057449147883453, "grad_norm": 2.124847173690796, "learning_rate": 2.9870854555770128e-05, "loss": 0.0938, "step": 1027 }, { "epoch": 0.07064321055525014, "grad_norm": 3.2337236404418945, "learning_rate": 2.987041703554927e-05, "loss": 0.0755, "step": 1028 }, { "epoch": 0.07071192963166575, "grad_norm": 3.002148389816284, "learning_rate": 2.986997877867852e-05, "loss": 0.1101, "step": 1029 }, { "epoch": 0.07078064870808136, "grad_norm": 2.698212146759033, "learning_rate": 2.98695397851796e-05, "loss": 0.0926, "step": 1030 }, { "epoch": 0.07084936778449698, "grad_norm": 2.4995617866516113, "learning_rate": 2.9869100055074247e-05, "loss": 0.084, "step": 1031 }, { "epoch": 0.07091808686091258, "grad_norm": 2.656590223312378, "learning_rate": 2.9868659588384246e-05, "loss": 0.0766, "step": 1032 }, { "epoch": 0.0709868059373282, "grad_norm": 2.7876620292663574, "learning_rate": 2.9868218385131416e-05, "loss": 0.0991, "step": 1033 }, { "epoch": 0.07105552501374382, "grad_norm": 2.016996383666992, "learning_rate": 2.986777644533762e-05, "loss": 0.0786, "step": 1034 }, { "epoch": 0.07112424409015942, "grad_norm": 2.6953682899475098, "learning_rate": 2.986733376902474e-05, "loss": 0.0813, "step": 1035 }, { "epoch": 0.07119296316657504, "grad_norm": 2.369501829147339, "learning_rate": 2.9866890356214715e-05, "loss": 0.0748, "step": 1036 }, { "epoch": 0.07126168224299065, "grad_norm": 3.005483627319336, "learning_rate": 2.9866446206929507e-05, "loss": 0.0793, "step": 1037 }, { "epoch": 0.07133040131940627, "grad_norm": 2.1576156616210938, "learning_rate": 2.9866001321191117e-05, "loss": 0.0759, "step": 1038 }, { "epoch": 0.07139912039582189, "grad_norm": 1.8459467887878418, "learning_rate": 2.986555569902158e-05, "loss": 0.1055, "step": 1039 }, { "epoch": 0.07146783947223749, "grad_norm": 2.615872859954834, "learning_rate": 2.9865109340442983e-05, "loss": 0.0892, "step": 1040 }, { "epoch": 0.07153655854865311, "grad_norm": 2.710530996322632, "learning_rate": 2.9864662245477424e-05, "loss": 0.0674, "step": 1041 }, { "epoch": 0.07160527762506871, "grad_norm": 2.6019930839538574, "learning_rate": 2.9864214414147066e-05, "loss": 0.1121, "step": 1042 }, { "epoch": 0.07167399670148433, "grad_norm": 2.3888490200042725, "learning_rate": 2.986376584647408e-05, "loss": 0.1069, "step": 1043 }, { "epoch": 0.07174271577789995, "grad_norm": 2.3097288608551025, "learning_rate": 2.9863316542480692e-05, "loss": 0.1029, "step": 1044 }, { "epoch": 0.07181143485431556, "grad_norm": 2.470060110092163, "learning_rate": 2.9862866502189163e-05, "loss": 0.0678, "step": 1045 }, { "epoch": 0.07188015393073117, "grad_norm": 2.3913755416870117, "learning_rate": 2.9862415725621784e-05, "loss": 0.0937, "step": 1046 }, { "epoch": 0.07194887300714678, "grad_norm": 2.658200740814209, "learning_rate": 2.9861964212800885e-05, "loss": 0.0883, "step": 1047 }, { "epoch": 0.0720175920835624, "grad_norm": 2.0163536071777344, "learning_rate": 2.986151196374883e-05, "loss": 0.0835, "step": 1048 }, { "epoch": 0.072086311159978, "grad_norm": 2.330658435821533, "learning_rate": 2.9861058978488032e-05, "loss": 0.1202, "step": 1049 }, { "epoch": 0.07215503023639362, "grad_norm": 3.811811685562134, "learning_rate": 2.9860605257040923e-05, "loss": 0.102, "step": 1050 }, { "epoch": 0.07222374931280924, "grad_norm": 1.7617384195327759, "learning_rate": 2.9860150799429983e-05, "loss": 0.087, "step": 1051 }, { "epoch": 0.07229246838922485, "grad_norm": 2.8592610359191895, "learning_rate": 2.985969560567772e-05, "loss": 0.1045, "step": 1052 }, { "epoch": 0.07236118746564046, "grad_norm": 1.89022696018219, "learning_rate": 2.985923967580669e-05, "loss": 0.0879, "step": 1053 }, { "epoch": 0.07242990654205607, "grad_norm": 2.6184990406036377, "learning_rate": 2.9858783009839477e-05, "loss": 0.0703, "step": 1054 }, { "epoch": 0.07249862561847169, "grad_norm": 2.600682020187378, "learning_rate": 2.98583256077987e-05, "loss": 0.1168, "step": 1055 }, { "epoch": 0.0725673446948873, "grad_norm": 2.048412799835205, "learning_rate": 2.985786746970702e-05, "loss": 0.09, "step": 1056 }, { "epoch": 0.07263606377130291, "grad_norm": 2.24877667427063, "learning_rate": 2.9857408595587136e-05, "loss": 0.0852, "step": 1057 }, { "epoch": 0.07270478284771853, "grad_norm": 2.217700719833374, "learning_rate": 2.985694898546177e-05, "loss": 0.0973, "step": 1058 }, { "epoch": 0.07277350192413413, "grad_norm": 2.221616506576538, "learning_rate": 2.9856488639353698e-05, "loss": 0.0824, "step": 1059 }, { "epoch": 0.07284222100054975, "grad_norm": 3.211618423461914, "learning_rate": 2.985602755728572e-05, "loss": 0.0989, "step": 1060 }, { "epoch": 0.07291094007696537, "grad_norm": 11.66296672821045, "learning_rate": 2.985556573928068e-05, "loss": 0.1019, "step": 1061 }, { "epoch": 0.07297965915338098, "grad_norm": 2.441450357437134, "learning_rate": 2.9855103185361458e-05, "loss": 0.0742, "step": 1062 }, { "epoch": 0.0730483782297966, "grad_norm": 4.204200744628906, "learning_rate": 2.9854639895550964e-05, "loss": 0.0931, "step": 1063 }, { "epoch": 0.0731170973062122, "grad_norm": 3.934990644454956, "learning_rate": 2.985417586987215e-05, "loss": 0.1201, "step": 1064 }, { "epoch": 0.07318581638262782, "grad_norm": 3.814692974090576, "learning_rate": 2.9853711108348e-05, "loss": 0.1163, "step": 1065 }, { "epoch": 0.07325453545904342, "grad_norm": 2.8711349964141846, "learning_rate": 2.9853245611001542e-05, "loss": 0.1137, "step": 1066 }, { "epoch": 0.07332325453545904, "grad_norm": 2.416491985321045, "learning_rate": 2.985277937785583e-05, "loss": 0.0988, "step": 1067 }, { "epoch": 0.07339197361187466, "grad_norm": 1.925836443901062, "learning_rate": 2.9852312408933966e-05, "loss": 0.0916, "step": 1068 }, { "epoch": 0.07346069268829027, "grad_norm": 2.7952942848205566, "learning_rate": 2.9851844704259084e-05, "loss": 0.1085, "step": 1069 }, { "epoch": 0.07352941176470588, "grad_norm": 2.6849236488342285, "learning_rate": 2.9851376263854346e-05, "loss": 0.1368, "step": 1070 }, { "epoch": 0.07359813084112149, "grad_norm": 2.540733814239502, "learning_rate": 2.9850907087742955e-05, "loss": 0.1007, "step": 1071 }, { "epoch": 0.07366684991753711, "grad_norm": 1.962433099746704, "learning_rate": 2.9850437175948167e-05, "loss": 0.1001, "step": 1072 }, { "epoch": 0.07373556899395273, "grad_norm": 4.170895576477051, "learning_rate": 2.9849966528493252e-05, "loss": 0.0919, "step": 1073 }, { "epoch": 0.07380428807036833, "grad_norm": 1.943647027015686, "learning_rate": 2.9849495145401526e-05, "loss": 0.0908, "step": 1074 }, { "epoch": 0.07387300714678395, "grad_norm": 2.2408156394958496, "learning_rate": 2.9849023026696335e-05, "loss": 0.0841, "step": 1075 }, { "epoch": 0.07394172622319956, "grad_norm": 2.581113815307617, "learning_rate": 2.9848550172401075e-05, "loss": 0.1149, "step": 1076 }, { "epoch": 0.07401044529961517, "grad_norm": 2.169886827468872, "learning_rate": 2.9848076582539165e-05, "loss": 0.1046, "step": 1077 }, { "epoch": 0.07407916437603079, "grad_norm": 1.6522423028945923, "learning_rate": 2.9847602257134067e-05, "loss": 0.0778, "step": 1078 }, { "epoch": 0.0741478834524464, "grad_norm": 2.784364700317383, "learning_rate": 2.984712719620928e-05, "loss": 0.093, "step": 1079 }, { "epoch": 0.07421660252886202, "grad_norm": 1.7367428541183472, "learning_rate": 2.9846651399788336e-05, "loss": 0.0771, "step": 1080 }, { "epoch": 0.07428532160527762, "grad_norm": 2.957118511199951, "learning_rate": 2.9846174867894804e-05, "loss": 0.0886, "step": 1081 }, { "epoch": 0.07435404068169324, "grad_norm": 1.9943588972091675, "learning_rate": 2.9845697600552295e-05, "loss": 0.0784, "step": 1082 }, { "epoch": 0.07442275975810884, "grad_norm": 3.118462324142456, "learning_rate": 2.9845219597784448e-05, "loss": 0.1062, "step": 1083 }, { "epoch": 0.07449147883452446, "grad_norm": 2.5219948291778564, "learning_rate": 2.9844740859614936e-05, "loss": 0.0876, "step": 1084 }, { "epoch": 0.07456019791094008, "grad_norm": 2.548877716064453, "learning_rate": 2.984426138606749e-05, "loss": 0.0903, "step": 1085 }, { "epoch": 0.07462891698735569, "grad_norm": 1.6517417430877686, "learning_rate": 2.9843781177165847e-05, "loss": 0.0801, "step": 1086 }, { "epoch": 0.0746976360637713, "grad_norm": 2.949676036834717, "learning_rate": 2.9843300232933806e-05, "loss": 0.0918, "step": 1087 }, { "epoch": 0.07476635514018691, "grad_norm": 1.4935328960418701, "learning_rate": 2.9842818553395184e-05, "loss": 0.0778, "step": 1088 }, { "epoch": 0.07483507421660253, "grad_norm": 2.32901668548584, "learning_rate": 2.9842336138573847e-05, "loss": 0.0739, "step": 1089 }, { "epoch": 0.07490379329301815, "grad_norm": 2.406278610229492, "learning_rate": 2.9841852988493696e-05, "loss": 0.0862, "step": 1090 }, { "epoch": 0.07497251236943375, "grad_norm": 2.128861427307129, "learning_rate": 2.984136910317866e-05, "loss": 0.0717, "step": 1091 }, { "epoch": 0.07504123144584937, "grad_norm": 2.012594223022461, "learning_rate": 2.984088448265271e-05, "loss": 0.0839, "step": 1092 }, { "epoch": 0.07510995052226498, "grad_norm": 2.6187520027160645, "learning_rate": 2.9840399126939858e-05, "loss": 0.1217, "step": 1093 }, { "epoch": 0.0751786695986806, "grad_norm": 2.000971555709839, "learning_rate": 2.983991303606414e-05, "loss": 0.0922, "step": 1094 }, { "epoch": 0.07524738867509621, "grad_norm": 1.957614541053772, "learning_rate": 2.9839426210049638e-05, "loss": 0.0882, "step": 1095 }, { "epoch": 0.07531610775151182, "grad_norm": 1.8712743520736694, "learning_rate": 2.9838938648920473e-05, "loss": 0.0824, "step": 1096 }, { "epoch": 0.07538482682792744, "grad_norm": 1.8827826976776123, "learning_rate": 2.9838450352700795e-05, "loss": 0.1065, "step": 1097 }, { "epoch": 0.07545354590434304, "grad_norm": 2.0738508701324463, "learning_rate": 2.983796132141479e-05, "loss": 0.0723, "step": 1098 }, { "epoch": 0.07552226498075866, "grad_norm": 2.193424940109253, "learning_rate": 2.9837471555086692e-05, "loss": 0.0851, "step": 1099 }, { "epoch": 0.07559098405717427, "grad_norm": 1.5578442811965942, "learning_rate": 2.9836981053740755e-05, "loss": 0.0599, "step": 1100 }, { "epoch": 0.07565970313358988, "grad_norm": 2.0676093101501465, "learning_rate": 2.9836489817401276e-05, "loss": 0.0952, "step": 1101 }, { "epoch": 0.0757284222100055, "grad_norm": 1.3085553646087646, "learning_rate": 2.98359978460926e-05, "loss": 0.0447, "step": 1102 }, { "epoch": 0.07579714128642111, "grad_norm": 1.3422486782073975, "learning_rate": 2.983550513983909e-05, "loss": 0.0932, "step": 1103 }, { "epoch": 0.07586586036283673, "grad_norm": 1.5411756038665771, "learning_rate": 2.9835011698665157e-05, "loss": 0.0963, "step": 1104 }, { "epoch": 0.07593457943925233, "grad_norm": 2.449307918548584, "learning_rate": 2.9834517522595242e-05, "loss": 0.0855, "step": 1105 }, { "epoch": 0.07600329851566795, "grad_norm": 2.011491537094116, "learning_rate": 2.9834022611653827e-05, "loss": 0.0997, "step": 1106 }, { "epoch": 0.07607201759208357, "grad_norm": 1.852790117263794, "learning_rate": 2.983352696586543e-05, "loss": 0.0704, "step": 1107 }, { "epoch": 0.07614073666849917, "grad_norm": 1.7666829824447632, "learning_rate": 2.983303058525461e-05, "loss": 0.078, "step": 1108 }, { "epoch": 0.07620945574491479, "grad_norm": 2.0170750617980957, "learning_rate": 2.9832533469845943e-05, "loss": 0.0917, "step": 1109 }, { "epoch": 0.0762781748213304, "grad_norm": 1.7106095552444458, "learning_rate": 2.9832035619664063e-05, "loss": 0.1049, "step": 1110 }, { "epoch": 0.07634689389774602, "grad_norm": 1.6974483728408813, "learning_rate": 2.9831537034733635e-05, "loss": 0.0782, "step": 1111 }, { "epoch": 0.07641561297416163, "grad_norm": 1.8042407035827637, "learning_rate": 2.9831037715079353e-05, "loss": 0.0905, "step": 1112 }, { "epoch": 0.07648433205057724, "grad_norm": 2.0530102252960205, "learning_rate": 2.983053766072595e-05, "loss": 0.0974, "step": 1113 }, { "epoch": 0.07655305112699286, "grad_norm": 4.038312911987305, "learning_rate": 2.9830036871698207e-05, "loss": 0.1098, "step": 1114 }, { "epoch": 0.07662177020340846, "grad_norm": 2.6972861289978027, "learning_rate": 2.9829535348020924e-05, "loss": 0.1045, "step": 1115 }, { "epoch": 0.07669048927982408, "grad_norm": 2.3428165912628174, "learning_rate": 2.9829033089718948e-05, "loss": 0.0737, "step": 1116 }, { "epoch": 0.07675920835623969, "grad_norm": 1.7748593091964722, "learning_rate": 2.9828530096817163e-05, "loss": 0.0834, "step": 1117 }, { "epoch": 0.0768279274326553, "grad_norm": 2.254084825515747, "learning_rate": 2.9828026369340478e-05, "loss": 0.0927, "step": 1118 }, { "epoch": 0.07689664650907092, "grad_norm": 2.123746633529663, "learning_rate": 2.9827521907313852e-05, "loss": 0.0881, "step": 1119 }, { "epoch": 0.07696536558548653, "grad_norm": 3.4641613960266113, "learning_rate": 2.982701671076228e-05, "loss": 0.0832, "step": 1120 }, { "epoch": 0.07703408466190215, "grad_norm": 2.335360050201416, "learning_rate": 2.982651077971078e-05, "loss": 0.0818, "step": 1121 }, { "epoch": 0.07710280373831775, "grad_norm": 2.4231419563293457, "learning_rate": 2.982600411418442e-05, "loss": 0.0863, "step": 1122 }, { "epoch": 0.07717152281473337, "grad_norm": 2.5061473846435547, "learning_rate": 2.9825496714208295e-05, "loss": 0.0807, "step": 1123 }, { "epoch": 0.07724024189114899, "grad_norm": 2.9821338653564453, "learning_rate": 2.982498857980754e-05, "loss": 0.0994, "step": 1124 }, { "epoch": 0.0773089609675646, "grad_norm": 1.961846947669983, "learning_rate": 2.9824479711007338e-05, "loss": 0.0669, "step": 1125 }, { "epoch": 0.07737768004398021, "grad_norm": 1.9513460397720337, "learning_rate": 2.982397010783288e-05, "loss": 0.0643, "step": 1126 }, { "epoch": 0.07744639912039582, "grad_norm": 2.3711469173431396, "learning_rate": 2.9823459770309426e-05, "loss": 0.0771, "step": 1127 }, { "epoch": 0.07751511819681144, "grad_norm": 1.492423415184021, "learning_rate": 2.982294869846225e-05, "loss": 0.086, "step": 1128 }, { "epoch": 0.07758383727322704, "grad_norm": 2.584911346435547, "learning_rate": 2.9822436892316663e-05, "loss": 0.1288, "step": 1129 }, { "epoch": 0.07765255634964266, "grad_norm": 2.2371652126312256, "learning_rate": 2.9821924351898034e-05, "loss": 0.0899, "step": 1130 }, { "epoch": 0.07772127542605828, "grad_norm": 2.1378250122070312, "learning_rate": 2.982141107723174e-05, "loss": 0.0942, "step": 1131 }, { "epoch": 0.07778999450247388, "grad_norm": 2.703505039215088, "learning_rate": 2.9820897068343215e-05, "loss": 0.1133, "step": 1132 }, { "epoch": 0.0778587135788895, "grad_norm": 2.2172904014587402, "learning_rate": 2.982038232525792e-05, "loss": 0.0814, "step": 1133 }, { "epoch": 0.0779274326553051, "grad_norm": 1.484122633934021, "learning_rate": 2.9819866848001356e-05, "loss": 0.0646, "step": 1134 }, { "epoch": 0.07799615173172073, "grad_norm": 1.6950709819793701, "learning_rate": 2.9819350636599058e-05, "loss": 0.0743, "step": 1135 }, { "epoch": 0.07806487080813634, "grad_norm": 3.0180907249450684, "learning_rate": 2.981883369107659e-05, "loss": 0.0713, "step": 1136 }, { "epoch": 0.07813358988455195, "grad_norm": 2.9549171924591064, "learning_rate": 2.9818316011459576e-05, "loss": 0.0898, "step": 1137 }, { "epoch": 0.07820230896096757, "grad_norm": 1.5188266038894653, "learning_rate": 2.981779759777364e-05, "loss": 0.0962, "step": 1138 }, { "epoch": 0.07827102803738317, "grad_norm": 1.5383788347244263, "learning_rate": 2.981727845004448e-05, "loss": 0.0917, "step": 1139 }, { "epoch": 0.07833974711379879, "grad_norm": 2.4145121574401855, "learning_rate": 2.9816758568297817e-05, "loss": 0.0774, "step": 1140 }, { "epoch": 0.07840846619021441, "grad_norm": 3.6882846355438232, "learning_rate": 2.9816237952559387e-05, "loss": 0.1052, "step": 1141 }, { "epoch": 0.07847718526663001, "grad_norm": 1.9797078371047974, "learning_rate": 2.9815716602854992e-05, "loss": 0.0966, "step": 1142 }, { "epoch": 0.07854590434304563, "grad_norm": 2.173556327819824, "learning_rate": 2.981519451921046e-05, "loss": 0.1051, "step": 1143 }, { "epoch": 0.07861462341946124, "grad_norm": 3.408672571182251, "learning_rate": 2.981467170165165e-05, "loss": 0.0975, "step": 1144 }, { "epoch": 0.07868334249587686, "grad_norm": 1.6493562459945679, "learning_rate": 2.9814148150204452e-05, "loss": 0.1034, "step": 1145 }, { "epoch": 0.07875206157229246, "grad_norm": 2.9825758934020996, "learning_rate": 2.981362386489482e-05, "loss": 0.1029, "step": 1146 }, { "epoch": 0.07882078064870808, "grad_norm": 1.8594664335250854, "learning_rate": 2.9813098845748715e-05, "loss": 0.0979, "step": 1147 }, { "epoch": 0.0788894997251237, "grad_norm": 3.9870707988739014, "learning_rate": 2.981257309279215e-05, "loss": 0.0924, "step": 1148 }, { "epoch": 0.0789582188015393, "grad_norm": 1.7028801441192627, "learning_rate": 2.9812046606051167e-05, "loss": 0.0733, "step": 1149 }, { "epoch": 0.07902693787795492, "grad_norm": 2.3017547130584717, "learning_rate": 2.9811519385551843e-05, "loss": 0.0852, "step": 1150 }, { "epoch": 0.07909565695437053, "grad_norm": 7.549728870391846, "learning_rate": 2.9810991431320302e-05, "loss": 0.0905, "step": 1151 }, { "epoch": 0.07916437603078615, "grad_norm": 2.789898157119751, "learning_rate": 2.9810462743382696e-05, "loss": 0.0829, "step": 1152 }, { "epoch": 0.07923309510720176, "grad_norm": 3.194683790206909, "learning_rate": 2.9809933321765215e-05, "loss": 0.0981, "step": 1153 }, { "epoch": 0.07930181418361737, "grad_norm": 6.441528797149658, "learning_rate": 2.9809403166494085e-05, "loss": 0.103, "step": 1154 }, { "epoch": 0.07937053326003299, "grad_norm": 5.304037570953369, "learning_rate": 2.980887227759557e-05, "loss": 0.1058, "step": 1155 }, { "epoch": 0.0794392523364486, "grad_norm": 2.337353229522705, "learning_rate": 2.9808340655095966e-05, "loss": 0.0908, "step": 1156 }, { "epoch": 0.07950797141286421, "grad_norm": 7.199710845947266, "learning_rate": 2.9807808299021612e-05, "loss": 0.1049, "step": 1157 }, { "epoch": 0.07957669048927983, "grad_norm": 3.4893546104431152, "learning_rate": 2.9807275209398877e-05, "loss": 0.0955, "step": 1158 }, { "epoch": 0.07964540956569544, "grad_norm": 2.393490791320801, "learning_rate": 2.9806741386254172e-05, "loss": 0.0827, "step": 1159 }, { "epoch": 0.07971412864211105, "grad_norm": 2.309373140335083, "learning_rate": 2.980620682961394e-05, "loss": 0.095, "step": 1160 }, { "epoch": 0.07978284771852666, "grad_norm": 3.450730562210083, "learning_rate": 2.980567153950466e-05, "loss": 0.1065, "step": 1161 }, { "epoch": 0.07985156679494228, "grad_norm": 10.743037223815918, "learning_rate": 2.9805135515952853e-05, "loss": 0.1009, "step": 1162 }, { "epoch": 0.07992028587135788, "grad_norm": 3.9979846477508545, "learning_rate": 2.9804598758985072e-05, "loss": 0.0759, "step": 1163 }, { "epoch": 0.0799890049477735, "grad_norm": 4.185831069946289, "learning_rate": 2.980406126862791e-05, "loss": 0.1255, "step": 1164 }, { "epoch": 0.08005772402418912, "grad_norm": 4.069482803344727, "learning_rate": 2.980352304490798e-05, "loss": 0.0754, "step": 1165 }, { "epoch": 0.08012644310060472, "grad_norm": 2.4200220108032227, "learning_rate": 2.9802984087851952e-05, "loss": 0.105, "step": 1166 }, { "epoch": 0.08019516217702034, "grad_norm": 2.9434292316436768, "learning_rate": 2.9802444397486533e-05, "loss": 0.0806, "step": 1167 }, { "epoch": 0.08026388125343595, "grad_norm": 2.0247490406036377, "learning_rate": 2.9801903973838447e-05, "loss": 0.0891, "step": 1168 }, { "epoch": 0.08033260032985157, "grad_norm": 1.7780437469482422, "learning_rate": 2.980136281693447e-05, "loss": 0.0784, "step": 1169 }, { "epoch": 0.08040131940626719, "grad_norm": 2.390028953552246, "learning_rate": 2.9800820926801407e-05, "loss": 0.0875, "step": 1170 }, { "epoch": 0.08047003848268279, "grad_norm": 3.590127944946289, "learning_rate": 2.9800278303466108e-05, "loss": 0.1083, "step": 1171 }, { "epoch": 0.08053875755909841, "grad_norm": 2.9428255558013916, "learning_rate": 2.979973494695545e-05, "loss": 0.1113, "step": 1172 }, { "epoch": 0.08060747663551401, "grad_norm": 2.534681558609009, "learning_rate": 2.9799190857296347e-05, "loss": 0.1047, "step": 1173 }, { "epoch": 0.08067619571192963, "grad_norm": 2.544119119644165, "learning_rate": 2.9798646034515758e-05, "loss": 0.0911, "step": 1174 }, { "epoch": 0.08074491478834525, "grad_norm": 1.3426134586334229, "learning_rate": 2.9798100478640665e-05, "loss": 0.0821, "step": 1175 }, { "epoch": 0.08081363386476086, "grad_norm": 4.373135089874268, "learning_rate": 2.9797554189698102e-05, "loss": 0.0866, "step": 1176 }, { "epoch": 0.08088235294117647, "grad_norm": 1.6937862634658813, "learning_rate": 2.9797007167715124e-05, "loss": 0.0683, "step": 1177 }, { "epoch": 0.08095107201759208, "grad_norm": 2.4853336811065674, "learning_rate": 2.9796459412718836e-05, "loss": 0.0927, "step": 1178 }, { "epoch": 0.0810197910940077, "grad_norm": 2.48781681060791, "learning_rate": 2.9795910924736366e-05, "loss": 0.1213, "step": 1179 }, { "epoch": 0.0810885101704233, "grad_norm": 1.373757243156433, "learning_rate": 2.9795361703794888e-05, "loss": 0.0652, "step": 1180 }, { "epoch": 0.08115722924683892, "grad_norm": 1.9913097620010376, "learning_rate": 2.979481174992161e-05, "loss": 0.0665, "step": 1181 }, { "epoch": 0.08122594832325454, "grad_norm": 3.041706085205078, "learning_rate": 2.9794261063143772e-05, "loss": 0.0716, "step": 1182 }, { "epoch": 0.08129466739967015, "grad_norm": 1.923303246498108, "learning_rate": 2.9793709643488664e-05, "loss": 0.0715, "step": 1183 }, { "epoch": 0.08136338647608576, "grad_norm": 2.249974012374878, "learning_rate": 2.9793157490983587e-05, "loss": 0.0712, "step": 1184 }, { "epoch": 0.08143210555250137, "grad_norm": 2.0102691650390625, "learning_rate": 2.9792604605655904e-05, "loss": 0.0776, "step": 1185 }, { "epoch": 0.08150082462891699, "grad_norm": 2.1144163608551025, "learning_rate": 2.9792050987533e-05, "loss": 0.086, "step": 1186 }, { "epoch": 0.0815695437053326, "grad_norm": 2.0061798095703125, "learning_rate": 2.9791496636642306e-05, "loss": 0.0681, "step": 1187 }, { "epoch": 0.08163826278174821, "grad_norm": 1.9888218641281128, "learning_rate": 2.9790941553011276e-05, "loss": 0.0769, "step": 1188 }, { "epoch": 0.08170698185816383, "grad_norm": 2.203568935394287, "learning_rate": 2.9790385736667408e-05, "loss": 0.0705, "step": 1189 }, { "epoch": 0.08177570093457943, "grad_norm": 1.83597731590271, "learning_rate": 2.9789829187638242e-05, "loss": 0.0787, "step": 1190 }, { "epoch": 0.08184442001099505, "grad_norm": 1.801870584487915, "learning_rate": 2.978927190595134e-05, "loss": 0.0732, "step": 1191 }, { "epoch": 0.08191313908741067, "grad_norm": 2.1216554641723633, "learning_rate": 2.9788713891634318e-05, "loss": 0.0768, "step": 1192 }, { "epoch": 0.08198185816382628, "grad_norm": 1.32426118850708, "learning_rate": 2.9788155144714814e-05, "loss": 0.069, "step": 1193 }, { "epoch": 0.0820505772402419, "grad_norm": 1.5764080286026, "learning_rate": 2.97875956652205e-05, "loss": 0.0865, "step": 1194 }, { "epoch": 0.0821192963166575, "grad_norm": 1.9841413497924805, "learning_rate": 2.9787035453179104e-05, "loss": 0.0958, "step": 1195 }, { "epoch": 0.08218801539307312, "grad_norm": 2.5688424110412598, "learning_rate": 2.9786474508618374e-05, "loss": 0.0857, "step": 1196 }, { "epoch": 0.08225673446948872, "grad_norm": 2.4420199394226074, "learning_rate": 2.9785912831566093e-05, "loss": 0.1256, "step": 1197 }, { "epoch": 0.08232545354590434, "grad_norm": 2.2861194610595703, "learning_rate": 2.9785350422050095e-05, "loss": 0.0864, "step": 1198 }, { "epoch": 0.08239417262231996, "grad_norm": 2.715195655822754, "learning_rate": 2.978478728009823e-05, "loss": 0.0814, "step": 1199 }, { "epoch": 0.08246289169873557, "grad_norm": 2.802676200866699, "learning_rate": 2.97842234057384e-05, "loss": 0.0659, "step": 1200 }, { "epoch": 0.08253161077515118, "grad_norm": 1.8152422904968262, "learning_rate": 2.9783658798998535e-05, "loss": 0.0678, "step": 1201 }, { "epoch": 0.08260032985156679, "grad_norm": 2.523510456085205, "learning_rate": 2.9783093459906612e-05, "loss": 0.0912, "step": 1202 }, { "epoch": 0.08266904892798241, "grad_norm": 2.301304817199707, "learning_rate": 2.9782527388490627e-05, "loss": 0.0759, "step": 1203 }, { "epoch": 0.08273776800439803, "grad_norm": 3.093546152114868, "learning_rate": 2.9781960584778632e-05, "loss": 0.088, "step": 1204 }, { "epoch": 0.08280648708081363, "grad_norm": 1.6740068197250366, "learning_rate": 2.9781393048798693e-05, "loss": 0.082, "step": 1205 }, { "epoch": 0.08287520615722925, "grad_norm": 3.3555638790130615, "learning_rate": 2.978082478057894e-05, "loss": 0.0881, "step": 1206 }, { "epoch": 0.08294392523364486, "grad_norm": 2.5668232440948486, "learning_rate": 2.9780255780147506e-05, "loss": 0.1274, "step": 1207 }, { "epoch": 0.08301264431006047, "grad_norm": 1.8678886890411377, "learning_rate": 2.9779686047532596e-05, "loss": 0.0698, "step": 1208 }, { "epoch": 0.08308136338647609, "grad_norm": 2.090348720550537, "learning_rate": 2.977911558276242e-05, "loss": 0.0978, "step": 1209 }, { "epoch": 0.0831500824628917, "grad_norm": 1.5796812772750854, "learning_rate": 2.9778544385865246e-05, "loss": 0.0794, "step": 1210 }, { "epoch": 0.08321880153930732, "grad_norm": 2.0259976387023926, "learning_rate": 2.9777972456869363e-05, "loss": 0.1167, "step": 1211 }, { "epoch": 0.08328752061572292, "grad_norm": 1.7238471508026123, "learning_rate": 2.9777399795803108e-05, "loss": 0.0778, "step": 1212 }, { "epoch": 0.08335623969213854, "grad_norm": 2.1139345169067383, "learning_rate": 2.977682640269485e-05, "loss": 0.0716, "step": 1213 }, { "epoch": 0.08342495876855414, "grad_norm": 1.9410244226455688, "learning_rate": 2.9776252277572992e-05, "loss": 0.1003, "step": 1214 }, { "epoch": 0.08349367784496976, "grad_norm": 2.3633432388305664, "learning_rate": 2.977567742046597e-05, "loss": 0.0686, "step": 1215 }, { "epoch": 0.08356239692138538, "grad_norm": 1.5476112365722656, "learning_rate": 2.977510183140227e-05, "loss": 0.1003, "step": 1216 }, { "epoch": 0.08363111599780099, "grad_norm": 1.7034157514572144, "learning_rate": 2.9774525510410398e-05, "loss": 0.0736, "step": 1217 }, { "epoch": 0.0836998350742166, "grad_norm": 2.0374088287353516, "learning_rate": 2.9773948457518915e-05, "loss": 0.0762, "step": 1218 }, { "epoch": 0.08376855415063221, "grad_norm": 1.887430191040039, "learning_rate": 2.9773370672756396e-05, "loss": 0.0878, "step": 1219 }, { "epoch": 0.08383727322704783, "grad_norm": 1.852992057800293, "learning_rate": 2.977279215615146e-05, "loss": 0.0757, "step": 1220 }, { "epoch": 0.08390599230346345, "grad_norm": 1.368425965309143, "learning_rate": 2.977221290773278e-05, "loss": 0.0709, "step": 1221 }, { "epoch": 0.08397471137987905, "grad_norm": 2.4216597080230713, "learning_rate": 2.977163292752904e-05, "loss": 0.068, "step": 1222 }, { "epoch": 0.08404343045629467, "grad_norm": 3.1920337677001953, "learning_rate": 2.9771052215568976e-05, "loss": 0.0752, "step": 1223 }, { "epoch": 0.08411214953271028, "grad_norm": 2.3636834621429443, "learning_rate": 2.9770470771881352e-05, "loss": 0.0825, "step": 1224 }, { "epoch": 0.0841808686091259, "grad_norm": 2.0045087337493896, "learning_rate": 2.9769888596494972e-05, "loss": 0.0696, "step": 1225 }, { "epoch": 0.08424958768554151, "grad_norm": 3.1520938873291016, "learning_rate": 2.9769305689438676e-05, "loss": 0.0907, "step": 1226 }, { "epoch": 0.08431830676195712, "grad_norm": 2.312044858932495, "learning_rate": 2.9768722050741345e-05, "loss": 0.0804, "step": 1227 }, { "epoch": 0.08438702583837274, "grad_norm": 1.710190773010254, "learning_rate": 2.9768137680431882e-05, "loss": 0.0803, "step": 1228 }, { "epoch": 0.08445574491478834, "grad_norm": 1.6789674758911133, "learning_rate": 2.976755257853924e-05, "loss": 0.0719, "step": 1229 }, { "epoch": 0.08452446399120396, "grad_norm": 1.6063143014907837, "learning_rate": 2.9766966745092405e-05, "loss": 0.0629, "step": 1230 }, { "epoch": 0.08459318306761957, "grad_norm": 1.3480403423309326, "learning_rate": 2.97663801801204e-05, "loss": 0.0578, "step": 1231 }, { "epoch": 0.08466190214403518, "grad_norm": 2.9865927696228027, "learning_rate": 2.9765792883652276e-05, "loss": 0.0858, "step": 1232 }, { "epoch": 0.0847306212204508, "grad_norm": 2.191270589828491, "learning_rate": 2.976520485571713e-05, "loss": 0.1014, "step": 1233 }, { "epoch": 0.08479934029686641, "grad_norm": 2.5614137649536133, "learning_rate": 2.9764616096344092e-05, "loss": 0.0857, "step": 1234 }, { "epoch": 0.08486805937328203, "grad_norm": 2.260117292404175, "learning_rate": 2.9764026605562326e-05, "loss": 0.0791, "step": 1235 }, { "epoch": 0.08493677844969763, "grad_norm": 1.7690922021865845, "learning_rate": 2.976343638340104e-05, "loss": 0.0793, "step": 1236 }, { "epoch": 0.08500549752611325, "grad_norm": 1.6885885000228882, "learning_rate": 2.9762845429889464e-05, "loss": 0.0602, "step": 1237 }, { "epoch": 0.08507421660252887, "grad_norm": 1.676914095878601, "learning_rate": 2.9762253745056878e-05, "loss": 0.0677, "step": 1238 }, { "epoch": 0.08514293567894447, "grad_norm": 4.08657693862915, "learning_rate": 2.976166132893259e-05, "loss": 0.0936, "step": 1239 }, { "epoch": 0.08521165475536009, "grad_norm": 2.4931800365448, "learning_rate": 2.976106818154595e-05, "loss": 0.0773, "step": 1240 }, { "epoch": 0.0852803738317757, "grad_norm": 1.4221181869506836, "learning_rate": 2.9760474302926337e-05, "loss": 0.0712, "step": 1241 }, { "epoch": 0.08534909290819132, "grad_norm": 1.9812510013580322, "learning_rate": 2.9759879693103176e-05, "loss": 0.1204, "step": 1242 }, { "epoch": 0.08541781198460692, "grad_norm": 1.6938742399215698, "learning_rate": 2.975928435210592e-05, "loss": 0.0747, "step": 1243 }, { "epoch": 0.08548653106102254, "grad_norm": 1.9092551469802856, "learning_rate": 2.975868827996406e-05, "loss": 0.0817, "step": 1244 }, { "epoch": 0.08555525013743816, "grad_norm": 1.8509624004364014, "learning_rate": 2.9758091476707125e-05, "loss": 0.0957, "step": 1245 }, { "epoch": 0.08562396921385376, "grad_norm": 1.6225253343582153, "learning_rate": 2.975749394236468e-05, "loss": 0.0579, "step": 1246 }, { "epoch": 0.08569268829026938, "grad_norm": 2.698106527328491, "learning_rate": 2.9756895676966325e-05, "loss": 0.0996, "step": 1247 }, { "epoch": 0.08576140736668499, "grad_norm": 1.3724385499954224, "learning_rate": 2.9756296680541697e-05, "loss": 0.0579, "step": 1248 }, { "epoch": 0.0858301264431006, "grad_norm": 2.156702995300293, "learning_rate": 2.975569695312047e-05, "loss": 0.0951, "step": 1249 }, { "epoch": 0.08589884551951622, "grad_norm": 2.0579493045806885, "learning_rate": 2.9755096494732352e-05, "loss": 0.0907, "step": 1250 }, { "epoch": 0.08596756459593183, "grad_norm": 2.0398061275482178, "learning_rate": 2.975449530540709e-05, "loss": 0.0932, "step": 1251 }, { "epoch": 0.08603628367234745, "grad_norm": 2.7284977436065674, "learning_rate": 2.9753893385174464e-05, "loss": 0.0894, "step": 1252 }, { "epoch": 0.08610500274876305, "grad_norm": 2.3995234966278076, "learning_rate": 2.975329073406429e-05, "loss": 0.0781, "step": 1253 }, { "epoch": 0.08617372182517867, "grad_norm": 1.982766032218933, "learning_rate": 2.9752687352106428e-05, "loss": 0.1034, "step": 1254 }, { "epoch": 0.08624244090159429, "grad_norm": 1.5644745826721191, "learning_rate": 2.975208323933076e-05, "loss": 0.0735, "step": 1255 }, { "epoch": 0.0863111599780099, "grad_norm": 1.9065078496932983, "learning_rate": 2.975147839576722e-05, "loss": 0.088, "step": 1256 }, { "epoch": 0.08637987905442551, "grad_norm": 1.5067585706710815, "learning_rate": 2.975087282144577e-05, "loss": 0.0666, "step": 1257 }, { "epoch": 0.08644859813084112, "grad_norm": 2.0938868522644043, "learning_rate": 2.9750266516396407e-05, "loss": 0.079, "step": 1258 }, { "epoch": 0.08651731720725674, "grad_norm": 1.6508110761642456, "learning_rate": 2.9749659480649163e-05, "loss": 0.0762, "step": 1259 }, { "epoch": 0.08658603628367234, "grad_norm": 2.1399917602539062, "learning_rate": 2.9749051714234114e-05, "loss": 0.0747, "step": 1260 }, { "epoch": 0.08665475536008796, "grad_norm": 1.965132713317871, "learning_rate": 2.9748443217181364e-05, "loss": 0.0955, "step": 1261 }, { "epoch": 0.08672347443650358, "grad_norm": 2.5416276454925537, "learning_rate": 2.9747833989521057e-05, "loss": 0.1056, "step": 1262 }, { "epoch": 0.08679219351291918, "grad_norm": 3.1682865619659424, "learning_rate": 2.974722403128338e-05, "loss": 0.1199, "step": 1263 }, { "epoch": 0.0868609125893348, "grad_norm": 2.7064907550811768, "learning_rate": 2.974661334249854e-05, "loss": 0.0782, "step": 1264 }, { "epoch": 0.0869296316657504, "grad_norm": 3.2442712783813477, "learning_rate": 2.9746001923196797e-05, "loss": 0.0868, "step": 1265 }, { "epoch": 0.08699835074216603, "grad_norm": 2.1181726455688477, "learning_rate": 2.974538977340843e-05, "loss": 0.0712, "step": 1266 }, { "epoch": 0.08706706981858164, "grad_norm": 1.6558963060379028, "learning_rate": 2.974477689316377e-05, "loss": 0.069, "step": 1267 }, { "epoch": 0.08713578889499725, "grad_norm": 2.0204784870147705, "learning_rate": 2.974416328249318e-05, "loss": 0.0897, "step": 1268 }, { "epoch": 0.08720450797141287, "grad_norm": 2.885711669921875, "learning_rate": 2.974354894142705e-05, "loss": 0.0868, "step": 1269 }, { "epoch": 0.08727322704782847, "grad_norm": 2.1242825984954834, "learning_rate": 2.9742933869995817e-05, "loss": 0.0678, "step": 1270 }, { "epoch": 0.08734194612424409, "grad_norm": 2.865819215774536, "learning_rate": 2.9742318068229955e-05, "loss": 0.069, "step": 1271 }, { "epoch": 0.08741066520065971, "grad_norm": 1.914265513420105, "learning_rate": 2.974170153615996e-05, "loss": 0.1034, "step": 1272 }, { "epoch": 0.08747938427707531, "grad_norm": 3.1928598880767822, "learning_rate": 2.974108427381638e-05, "loss": 0.0971, "step": 1273 }, { "epoch": 0.08754810335349093, "grad_norm": 1.6434850692749023, "learning_rate": 2.974046628122979e-05, "loss": 0.0833, "step": 1274 }, { "epoch": 0.08761682242990654, "grad_norm": 2.0538713932037354, "learning_rate": 2.973984755843081e-05, "loss": 0.098, "step": 1275 }, { "epoch": 0.08768554150632216, "grad_norm": 2.597914695739746, "learning_rate": 2.9739228105450077e-05, "loss": 0.0878, "step": 1276 }, { "epoch": 0.08775426058273776, "grad_norm": 2.567558526992798, "learning_rate": 2.9738607922318295e-05, "loss": 0.0993, "step": 1277 }, { "epoch": 0.08782297965915338, "grad_norm": 3.529686212539673, "learning_rate": 2.9737987009066173e-05, "loss": 0.0795, "step": 1278 }, { "epoch": 0.087891698735569, "grad_norm": 2.3445262908935547, "learning_rate": 2.9737365365724478e-05, "loss": 0.0804, "step": 1279 }, { "epoch": 0.0879604178119846, "grad_norm": 2.5317981243133545, "learning_rate": 2.9736742992323998e-05, "loss": 0.0789, "step": 1280 }, { "epoch": 0.08802913688840022, "grad_norm": 2.8752527236938477, "learning_rate": 2.973611988889557e-05, "loss": 0.1013, "step": 1281 }, { "epoch": 0.08809785596481583, "grad_norm": 1.9354466199874878, "learning_rate": 2.9735496055470056e-05, "loss": 0.0696, "step": 1282 }, { "epoch": 0.08816657504123145, "grad_norm": 2.0337984561920166, "learning_rate": 2.9734871492078366e-05, "loss": 0.0841, "step": 1283 }, { "epoch": 0.08823529411764706, "grad_norm": 1.5722805261611938, "learning_rate": 2.9734246198751436e-05, "loss": 0.0824, "step": 1284 }, { "epoch": 0.08830401319406267, "grad_norm": 1.7447541952133179, "learning_rate": 2.973362017552024e-05, "loss": 0.0793, "step": 1285 }, { "epoch": 0.08837273227047829, "grad_norm": 1.7142452001571655, "learning_rate": 2.973299342241579e-05, "loss": 0.103, "step": 1286 }, { "epoch": 0.08844145134689389, "grad_norm": 1.3775516748428345, "learning_rate": 2.9732365939469135e-05, "loss": 0.0864, "step": 1287 }, { "epoch": 0.08851017042330951, "grad_norm": 2.0331411361694336, "learning_rate": 2.9731737726711365e-05, "loss": 0.0847, "step": 1288 }, { "epoch": 0.08857888949972513, "grad_norm": 1.9378544092178345, "learning_rate": 2.973110878417359e-05, "loss": 0.0907, "step": 1289 }, { "epoch": 0.08864760857614074, "grad_norm": 2.3187007904052734, "learning_rate": 2.9730479111886976e-05, "loss": 0.0907, "step": 1290 }, { "epoch": 0.08871632765255635, "grad_norm": 2.7229387760162354, "learning_rate": 2.972984870988271e-05, "loss": 0.1045, "step": 1291 }, { "epoch": 0.08878504672897196, "grad_norm": 2.103992462158203, "learning_rate": 2.972921757819202e-05, "loss": 0.0977, "step": 1292 }, { "epoch": 0.08885376580538758, "grad_norm": 1.7725074291229248, "learning_rate": 2.9728585716846176e-05, "loss": 0.0656, "step": 1293 }, { "epoch": 0.08892248488180318, "grad_norm": 2.2938709259033203, "learning_rate": 2.9727953125876474e-05, "loss": 0.1006, "step": 1294 }, { "epoch": 0.0889912039582188, "grad_norm": 1.782270073890686, "learning_rate": 2.9727319805314255e-05, "loss": 0.057, "step": 1295 }, { "epoch": 0.08905992303463442, "grad_norm": 1.974712610244751, "learning_rate": 2.972668575519089e-05, "loss": 0.0768, "step": 1296 }, { "epoch": 0.08912864211105002, "grad_norm": 1.618503212928772, "learning_rate": 2.972605097553779e-05, "loss": 0.065, "step": 1297 }, { "epoch": 0.08919736118746564, "grad_norm": 2.214799404144287, "learning_rate": 2.97254154663864e-05, "loss": 0.0873, "step": 1298 }, { "epoch": 0.08926608026388125, "grad_norm": 1.5650640726089478, "learning_rate": 2.9724779227768203e-05, "loss": 0.0578, "step": 1299 }, { "epoch": 0.08933479934029687, "grad_norm": 1.8590044975280762, "learning_rate": 2.9724142259714712e-05, "loss": 0.0647, "step": 1300 }, { "epoch": 0.08940351841671249, "grad_norm": 1.7496365308761597, "learning_rate": 2.972350456225749e-05, "loss": 0.1081, "step": 1301 }, { "epoch": 0.08947223749312809, "grad_norm": 2.0565643310546875, "learning_rate": 2.9722866135428117e-05, "loss": 0.0719, "step": 1302 }, { "epoch": 0.08954095656954371, "grad_norm": 2.3572299480438232, "learning_rate": 2.972222697925823e-05, "loss": 0.0774, "step": 1303 }, { "epoch": 0.08960967564595931, "grad_norm": 2.129441738128662, "learning_rate": 2.9721587093779484e-05, "loss": 0.082, "step": 1304 }, { "epoch": 0.08967839472237493, "grad_norm": 1.5034427642822266, "learning_rate": 2.972094647902358e-05, "loss": 0.0603, "step": 1305 }, { "epoch": 0.08974711379879055, "grad_norm": 1.5013072490692139, "learning_rate": 2.972030513502225e-05, "loss": 0.0744, "step": 1306 }, { "epoch": 0.08981583287520616, "grad_norm": 2.0646843910217285, "learning_rate": 2.971966306180727e-05, "loss": 0.0702, "step": 1307 }, { "epoch": 0.08988455195162177, "grad_norm": 3.4211606979370117, "learning_rate": 2.9719020259410443e-05, "loss": 0.0758, "step": 1308 }, { "epoch": 0.08995327102803738, "grad_norm": 2.7438771724700928, "learning_rate": 2.9718376727863615e-05, "loss": 0.101, "step": 1309 }, { "epoch": 0.090021990104453, "grad_norm": 1.891187310218811, "learning_rate": 2.971773246719866e-05, "loss": 0.0722, "step": 1310 }, { "epoch": 0.0900907091808686, "grad_norm": 1.7729980945587158, "learning_rate": 2.97170874774475e-05, "loss": 0.0837, "step": 1311 }, { "epoch": 0.09015942825728422, "grad_norm": 2.1195054054260254, "learning_rate": 2.9716441758642086e-05, "loss": 0.0761, "step": 1312 }, { "epoch": 0.09022814733369984, "grad_norm": 2.167738914489746, "learning_rate": 2.9715795310814397e-05, "loss": 0.0803, "step": 1313 }, { "epoch": 0.09029686641011545, "grad_norm": 2.022583246231079, "learning_rate": 2.971514813399647e-05, "loss": 0.09, "step": 1314 }, { "epoch": 0.09036558548653106, "grad_norm": 2.3881916999816895, "learning_rate": 2.9714500228220347e-05, "loss": 0.0685, "step": 1315 }, { "epoch": 0.09043430456294667, "grad_norm": 2.5993688106536865, "learning_rate": 2.971385159351814e-05, "loss": 0.0789, "step": 1316 }, { "epoch": 0.09050302363936229, "grad_norm": 2.418602466583252, "learning_rate": 2.971320222992198e-05, "loss": 0.0778, "step": 1317 }, { "epoch": 0.0905717427157779, "grad_norm": 2.6494202613830566, "learning_rate": 2.971255213746402e-05, "loss": 0.0867, "step": 1318 }, { "epoch": 0.09064046179219351, "grad_norm": 2.240060567855835, "learning_rate": 2.9711901316176486e-05, "loss": 0.0707, "step": 1319 }, { "epoch": 0.09070918086860913, "grad_norm": 2.090278148651123, "learning_rate": 2.9711249766091602e-05, "loss": 0.0871, "step": 1320 }, { "epoch": 0.09077789994502473, "grad_norm": 2.7694408893585205, "learning_rate": 2.9710597487241647e-05, "loss": 0.0903, "step": 1321 }, { "epoch": 0.09084661902144035, "grad_norm": 2.53361177444458, "learning_rate": 2.970994447965894e-05, "loss": 0.0774, "step": 1322 }, { "epoch": 0.09091533809785597, "grad_norm": 2.5613789558410645, "learning_rate": 2.9709290743375826e-05, "loss": 0.0951, "step": 1323 }, { "epoch": 0.09098405717427158, "grad_norm": 2.0462446212768555, "learning_rate": 2.9708636278424687e-05, "loss": 0.0897, "step": 1324 }, { "epoch": 0.0910527762506872, "grad_norm": 1.5246104001998901, "learning_rate": 2.9707981084837948e-05, "loss": 0.0805, "step": 1325 }, { "epoch": 0.0911214953271028, "grad_norm": 1.497563362121582, "learning_rate": 2.9707325162648067e-05, "loss": 0.0805, "step": 1326 }, { "epoch": 0.09119021440351842, "grad_norm": 2.3225719928741455, "learning_rate": 2.9706668511887533e-05, "loss": 0.0759, "step": 1327 }, { "epoch": 0.09125893347993402, "grad_norm": 1.5931288003921509, "learning_rate": 2.9706011132588876e-05, "loss": 0.0812, "step": 1328 }, { "epoch": 0.09132765255634964, "grad_norm": 1.8735102415084839, "learning_rate": 2.970535302478466e-05, "loss": 0.1088, "step": 1329 }, { "epoch": 0.09139637163276526, "grad_norm": 2.0934667587280273, "learning_rate": 2.970469418850749e-05, "loss": 0.0704, "step": 1330 }, { "epoch": 0.09146509070918087, "grad_norm": 1.3412803411483765, "learning_rate": 2.9704034623790003e-05, "loss": 0.0673, "step": 1331 }, { "epoch": 0.09153380978559648, "grad_norm": 2.1552939414978027, "learning_rate": 2.9703374330664864e-05, "loss": 0.0598, "step": 1332 }, { "epoch": 0.09160252886201209, "grad_norm": 2.657620668411255, "learning_rate": 2.9702713309164795e-05, "loss": 0.0747, "step": 1333 }, { "epoch": 0.09167124793842771, "grad_norm": 2.0465712547302246, "learning_rate": 2.9702051559322537e-05, "loss": 0.0793, "step": 1334 }, { "epoch": 0.09173996701484333, "grad_norm": 3.552945613861084, "learning_rate": 2.970138908117087e-05, "loss": 0.0467, "step": 1335 }, { "epoch": 0.09180868609125893, "grad_norm": 1.046999216079712, "learning_rate": 2.9700725874742614e-05, "loss": 0.092, "step": 1336 }, { "epoch": 0.09187740516767455, "grad_norm": 2.2158989906311035, "learning_rate": 2.9700061940070616e-05, "loss": 0.0762, "step": 1337 }, { "epoch": 0.09194612424409015, "grad_norm": 2.3261749744415283, "learning_rate": 2.9699397277187775e-05, "loss": 0.0751, "step": 1338 }, { "epoch": 0.09201484332050577, "grad_norm": 1.5173531770706177, "learning_rate": 2.9698731886127013e-05, "loss": 0.06, "step": 1339 }, { "epoch": 0.09208356239692138, "grad_norm": 2.1954851150512695, "learning_rate": 2.9698065766921297e-05, "loss": 0.0767, "step": 1340 }, { "epoch": 0.092152281473337, "grad_norm": 2.578026056289673, "learning_rate": 2.969739891960362e-05, "loss": 0.0861, "step": 1341 }, { "epoch": 0.09222100054975262, "grad_norm": 1.6243007183074951, "learning_rate": 2.9696731344207008e-05, "loss": 0.0718, "step": 1342 }, { "epoch": 0.09228971962616822, "grad_norm": 1.6618294715881348, "learning_rate": 2.969606304076455e-05, "loss": 0.0695, "step": 1343 }, { "epoch": 0.09235843870258384, "grad_norm": 1.915857195854187, "learning_rate": 2.969539400930934e-05, "loss": 0.0904, "step": 1344 }, { "epoch": 0.09242715777899944, "grad_norm": 2.470335006713867, "learning_rate": 2.969472424987452e-05, "loss": 0.072, "step": 1345 }, { "epoch": 0.09249587685541506, "grad_norm": 7.586323261260986, "learning_rate": 2.9694053762493275e-05, "loss": 0.0997, "step": 1346 }, { "epoch": 0.09256459593183068, "grad_norm": 1.981920599937439, "learning_rate": 2.9693382547198815e-05, "loss": 0.0861, "step": 1347 }, { "epoch": 0.09263331500824629, "grad_norm": 1.840465784072876, "learning_rate": 2.969271060402439e-05, "loss": 0.0635, "step": 1348 }, { "epoch": 0.0927020340846619, "grad_norm": 1.3735918998718262, "learning_rate": 2.9692037933003292e-05, "loss": 0.0675, "step": 1349 }, { "epoch": 0.09277075316107751, "grad_norm": 2.5153002738952637, "learning_rate": 2.9691364534168843e-05, "loss": 0.0736, "step": 1350 }, { "epoch": 0.09283947223749313, "grad_norm": 2.001613140106201, "learning_rate": 2.9690690407554395e-05, "loss": 0.0776, "step": 1351 }, { "epoch": 0.09290819131390875, "grad_norm": 2.513209581375122, "learning_rate": 2.9690015553193348e-05, "loss": 0.1131, "step": 1352 }, { "epoch": 0.09297691039032435, "grad_norm": 1.8052184581756592, "learning_rate": 2.968933997111913e-05, "loss": 0.0811, "step": 1353 }, { "epoch": 0.09304562946673997, "grad_norm": 4.525395393371582, "learning_rate": 2.968866366136521e-05, "loss": 0.1202, "step": 1354 }, { "epoch": 0.09311434854315558, "grad_norm": 3.300668478012085, "learning_rate": 2.9687986623965095e-05, "loss": 0.0895, "step": 1355 }, { "epoch": 0.0931830676195712, "grad_norm": 2.2463982105255127, "learning_rate": 2.9687308858952315e-05, "loss": 0.0853, "step": 1356 }, { "epoch": 0.0932517866959868, "grad_norm": 2.2162773609161377, "learning_rate": 2.9686630366360456e-05, "loss": 0.0898, "step": 1357 }, { "epoch": 0.09332050577240242, "grad_norm": 1.9140098094940186, "learning_rate": 2.9685951146223117e-05, "loss": 0.0792, "step": 1358 }, { "epoch": 0.09338922484881804, "grad_norm": 2.141054153442383, "learning_rate": 2.9685271198573955e-05, "loss": 0.0893, "step": 1359 }, { "epoch": 0.09345794392523364, "grad_norm": 2.684979200363159, "learning_rate": 2.968459052344665e-05, "loss": 0.1006, "step": 1360 }, { "epoch": 0.09352666300164926, "grad_norm": 2.3214075565338135, "learning_rate": 2.9683909120874915e-05, "loss": 0.095, "step": 1361 }, { "epoch": 0.09359538207806486, "grad_norm": 2.4079251289367676, "learning_rate": 2.9683226990892517e-05, "loss": 0.0757, "step": 1362 }, { "epoch": 0.09366410115448048, "grad_norm": 2.3753597736358643, "learning_rate": 2.9682544133533236e-05, "loss": 0.0815, "step": 1363 }, { "epoch": 0.0937328202308961, "grad_norm": 1.668704867362976, "learning_rate": 2.968186054883091e-05, "loss": 0.0767, "step": 1364 }, { "epoch": 0.09380153930731171, "grad_norm": 2.334871292114258, "learning_rate": 2.9681176236819395e-05, "loss": 0.1425, "step": 1365 }, { "epoch": 0.09387025838372733, "grad_norm": 1.724042296409607, "learning_rate": 2.968049119753259e-05, "loss": 0.0868, "step": 1366 }, { "epoch": 0.09393897746014293, "grad_norm": 1.5472546815872192, "learning_rate": 2.9679805431004435e-05, "loss": 0.0762, "step": 1367 }, { "epoch": 0.09400769653655855, "grad_norm": 1.4619327783584595, "learning_rate": 2.96791189372689e-05, "loss": 0.0728, "step": 1368 }, { "epoch": 0.09407641561297417, "grad_norm": 2.5135767459869385, "learning_rate": 2.967843171635999e-05, "loss": 0.0663, "step": 1369 }, { "epoch": 0.09414513468938977, "grad_norm": 1.506057620048523, "learning_rate": 2.967774376831175e-05, "loss": 0.0884, "step": 1370 }, { "epoch": 0.09421385376580539, "grad_norm": 1.6248592138290405, "learning_rate": 2.9677055093158264e-05, "loss": 0.0679, "step": 1371 }, { "epoch": 0.094282572842221, "grad_norm": 1.746554970741272, "learning_rate": 2.9676365690933642e-05, "loss": 0.0758, "step": 1372 }, { "epoch": 0.09435129191863662, "grad_norm": 1.4647066593170166, "learning_rate": 2.9675675561672033e-05, "loss": 0.059, "step": 1373 }, { "epoch": 0.09442001099505222, "grad_norm": 1.906348705291748, "learning_rate": 2.9674984705407635e-05, "loss": 0.0743, "step": 1374 }, { "epoch": 0.09448873007146784, "grad_norm": 1.7488526105880737, "learning_rate": 2.9674293122174662e-05, "loss": 0.0764, "step": 1375 }, { "epoch": 0.09455744914788346, "grad_norm": 1.6583598852157593, "learning_rate": 2.9673600812007376e-05, "loss": 0.0821, "step": 1376 }, { "epoch": 0.09462616822429906, "grad_norm": 1.9935729503631592, "learning_rate": 2.9672907774940073e-05, "loss": 0.0655, "step": 1377 }, { "epoch": 0.09469488730071468, "grad_norm": 1.9460431337356567, "learning_rate": 2.9672214011007087e-05, "loss": 0.0709, "step": 1378 }, { "epoch": 0.09476360637713029, "grad_norm": 1.915809988975525, "learning_rate": 2.9671519520242782e-05, "loss": 0.0633, "step": 1379 }, { "epoch": 0.0948323254535459, "grad_norm": 2.2090201377868652, "learning_rate": 2.967082430268156e-05, "loss": 0.0636, "step": 1380 }, { "epoch": 0.09490104452996152, "grad_norm": 2.2879278659820557, "learning_rate": 2.9670128358357874e-05, "loss": 0.0731, "step": 1381 }, { "epoch": 0.09496976360637713, "grad_norm": 3.7926247119903564, "learning_rate": 2.9669431687306178e-05, "loss": 0.0862, "step": 1382 }, { "epoch": 0.09503848268279275, "grad_norm": 2.6045713424682617, "learning_rate": 2.9668734289561004e-05, "loss": 0.0903, "step": 1383 }, { "epoch": 0.09510720175920835, "grad_norm": 1.5090317726135254, "learning_rate": 2.9668036165156886e-05, "loss": 0.0727, "step": 1384 }, { "epoch": 0.09517592083562397, "grad_norm": 1.382483959197998, "learning_rate": 2.9667337314128414e-05, "loss": 0.0842, "step": 1385 }, { "epoch": 0.09524463991203959, "grad_norm": 1.783531665802002, "learning_rate": 2.9666637736510207e-05, "loss": 0.074, "step": 1386 }, { "epoch": 0.0953133589884552, "grad_norm": 1.9255015850067139, "learning_rate": 2.966593743233692e-05, "loss": 0.0805, "step": 1387 }, { "epoch": 0.09538207806487081, "grad_norm": 2.2544355392456055, "learning_rate": 2.9665236401643244e-05, "loss": 0.0765, "step": 1388 }, { "epoch": 0.09545079714128642, "grad_norm": 1.779158353805542, "learning_rate": 2.9664534644463907e-05, "loss": 0.0642, "step": 1389 }, { "epoch": 0.09551951621770204, "grad_norm": 1.587732195854187, "learning_rate": 2.9663832160833674e-05, "loss": 0.0653, "step": 1390 }, { "epoch": 0.09558823529411764, "grad_norm": 1.6918549537658691, "learning_rate": 2.9663128950787338e-05, "loss": 0.0878, "step": 1391 }, { "epoch": 0.09565695437053326, "grad_norm": 1.4387474060058594, "learning_rate": 2.9662425014359744e-05, "loss": 0.0641, "step": 1392 }, { "epoch": 0.09572567344694888, "grad_norm": 1.692023515701294, "learning_rate": 2.966172035158576e-05, "loss": 0.0791, "step": 1393 }, { "epoch": 0.09579439252336448, "grad_norm": 1.4089027643203735, "learning_rate": 2.9661014962500293e-05, "loss": 0.0711, "step": 1394 }, { "epoch": 0.0958631115997801, "grad_norm": 1.7097963094711304, "learning_rate": 2.9660308847138287e-05, "loss": 0.0698, "step": 1395 }, { "epoch": 0.0959318306761957, "grad_norm": 1.5652590990066528, "learning_rate": 2.9659602005534722e-05, "loss": 0.0814, "step": 1396 }, { "epoch": 0.09600054975261132, "grad_norm": 1.5115909576416016, "learning_rate": 2.965889443772461e-05, "loss": 0.0989, "step": 1397 }, { "epoch": 0.09606926882902694, "grad_norm": 2.008352518081665, "learning_rate": 2.9658186143743004e-05, "loss": 0.0589, "step": 1398 }, { "epoch": 0.09613798790544255, "grad_norm": 2.0150022506713867, "learning_rate": 2.9657477123624993e-05, "loss": 0.0719, "step": 1399 }, { "epoch": 0.09620670698185817, "grad_norm": 1.414870023727417, "learning_rate": 2.96567673774057e-05, "loss": 0.0569, "step": 1400 }, { "epoch": 0.09627542605827377, "grad_norm": 1.8415627479553223, "learning_rate": 2.9656056905120287e-05, "loss": 0.0589, "step": 1401 }, { "epoch": 0.09634414513468939, "grad_norm": 1.5775221586227417, "learning_rate": 2.9655345706803943e-05, "loss": 0.0566, "step": 1402 }, { "epoch": 0.09641286421110501, "grad_norm": 3.8095908164978027, "learning_rate": 2.9654633782491906e-05, "loss": 0.0684, "step": 1403 }, { "epoch": 0.09648158328752061, "grad_norm": 2.939591884613037, "learning_rate": 2.9653921132219436e-05, "loss": 0.0923, "step": 1404 }, { "epoch": 0.09655030236393623, "grad_norm": 2.0292840003967285, "learning_rate": 2.9653207756021844e-05, "loss": 0.0741, "step": 1405 }, { "epoch": 0.09661902144035184, "grad_norm": 1.3667855262756348, "learning_rate": 2.9652493653934462e-05, "loss": 0.075, "step": 1406 }, { "epoch": 0.09668774051676746, "grad_norm": 1.7072230577468872, "learning_rate": 2.9651778825992672e-05, "loss": 0.0885, "step": 1407 }, { "epoch": 0.09675645959318306, "grad_norm": 1.7183990478515625, "learning_rate": 2.965106327223188e-05, "loss": 0.0692, "step": 1408 }, { "epoch": 0.09682517866959868, "grad_norm": 2.0003278255462646, "learning_rate": 2.9650346992687536e-05, "loss": 0.0791, "step": 1409 }, { "epoch": 0.0968938977460143, "grad_norm": 1.4130903482437134, "learning_rate": 2.9649629987395117e-05, "loss": 0.0703, "step": 1410 }, { "epoch": 0.0969626168224299, "grad_norm": 1.393068552017212, "learning_rate": 2.9648912256390154e-05, "loss": 0.0612, "step": 1411 }, { "epoch": 0.09703133589884552, "grad_norm": 1.6097402572631836, "learning_rate": 2.964819379970819e-05, "loss": 0.0671, "step": 1412 }, { "epoch": 0.09710005497526113, "grad_norm": 1.6926010847091675, "learning_rate": 2.9647474617384825e-05, "loss": 0.0655, "step": 1413 }, { "epoch": 0.09716877405167675, "grad_norm": 1.9633493423461914, "learning_rate": 2.9646754709455677e-05, "loss": 0.0567, "step": 1414 }, { "epoch": 0.09723749312809236, "grad_norm": 1.4950464963912964, "learning_rate": 2.9646034075956412e-05, "loss": 0.0745, "step": 1415 }, { "epoch": 0.09730621220450797, "grad_norm": 3.3563055992126465, "learning_rate": 2.9645312716922732e-05, "loss": 0.1063, "step": 1416 }, { "epoch": 0.09737493128092359, "grad_norm": 2.5593671798706055, "learning_rate": 2.964459063239037e-05, "loss": 0.0802, "step": 1417 }, { "epoch": 0.09744365035733919, "grad_norm": 1.626767635345459, "learning_rate": 2.96438678223951e-05, "loss": 0.0691, "step": 1418 }, { "epoch": 0.09751236943375481, "grad_norm": 1.8433057069778442, "learning_rate": 2.9643144286972717e-05, "loss": 0.078, "step": 1419 }, { "epoch": 0.09758108851017043, "grad_norm": 2.219797134399414, "learning_rate": 2.9642420026159075e-05, "loss": 0.0806, "step": 1420 }, { "epoch": 0.09764980758658603, "grad_norm": 2.2013614177703857, "learning_rate": 2.964169503999005e-05, "loss": 0.0972, "step": 1421 }, { "epoch": 0.09771852666300165, "grad_norm": 2.0257086753845215, "learning_rate": 2.9640969328501554e-05, "loss": 0.0721, "step": 1422 }, { "epoch": 0.09778724573941726, "grad_norm": 2.2087326049804688, "learning_rate": 2.9640242891729537e-05, "loss": 0.0826, "step": 1423 }, { "epoch": 0.09785596481583288, "grad_norm": 5.045064449310303, "learning_rate": 2.9639515729709984e-05, "loss": 0.0706, "step": 1424 }, { "epoch": 0.09792468389224848, "grad_norm": 2.393740177154541, "learning_rate": 2.963878784247892e-05, "loss": 0.0718, "step": 1425 }, { "epoch": 0.0979934029686641, "grad_norm": 3.399815320968628, "learning_rate": 2.9638059230072405e-05, "loss": 0.0919, "step": 1426 }, { "epoch": 0.09806212204507972, "grad_norm": 1.6214542388916016, "learning_rate": 2.963732989252653e-05, "loss": 0.0752, "step": 1427 }, { "epoch": 0.09813084112149532, "grad_norm": 1.3035391569137573, "learning_rate": 2.963659982987742e-05, "loss": 0.0714, "step": 1428 }, { "epoch": 0.09819956019791094, "grad_norm": 3.295516014099121, "learning_rate": 2.9635869042161255e-05, "loss": 0.0796, "step": 1429 }, { "epoch": 0.09826827927432655, "grad_norm": 7.240596294403076, "learning_rate": 2.9635137529414218e-05, "loss": 0.0954, "step": 1430 }, { "epoch": 0.09833699835074217, "grad_norm": 1.687196135520935, "learning_rate": 2.9634405291672562e-05, "loss": 0.0682, "step": 1431 }, { "epoch": 0.09840571742715779, "grad_norm": 2.8977513313293457, "learning_rate": 2.9633672328972557e-05, "loss": 0.0872, "step": 1432 }, { "epoch": 0.09847443650357339, "grad_norm": 1.548365592956543, "learning_rate": 2.9632938641350508e-05, "loss": 0.1135, "step": 1433 }, { "epoch": 0.09854315557998901, "grad_norm": 2.4135594367980957, "learning_rate": 2.9632204228842757e-05, "loss": 0.0983, "step": 1434 }, { "epoch": 0.09861187465640461, "grad_norm": 3.2451415061950684, "learning_rate": 2.96314690914857e-05, "loss": 0.0733, "step": 1435 }, { "epoch": 0.09868059373282023, "grad_norm": 1.4462157487869263, "learning_rate": 2.9630733229315743e-05, "loss": 0.0913, "step": 1436 }, { "epoch": 0.09874931280923585, "grad_norm": 2.1520800590515137, "learning_rate": 2.9629996642369338e-05, "loss": 0.0683, "step": 1437 }, { "epoch": 0.09881803188565146, "grad_norm": 3.504732131958008, "learning_rate": 2.962925933068298e-05, "loss": 0.0885, "step": 1438 }, { "epoch": 0.09888675096206707, "grad_norm": 1.3221538066864014, "learning_rate": 2.962852129429319e-05, "loss": 0.075, "step": 1439 }, { "epoch": 0.09895547003848268, "grad_norm": 3.596738815307617, "learning_rate": 2.9627782533236533e-05, "loss": 0.0945, "step": 1440 }, { "epoch": 0.0990241891148983, "grad_norm": 1.7792508602142334, "learning_rate": 2.96270430475496e-05, "loss": 0.0757, "step": 1441 }, { "epoch": 0.0990929081913139, "grad_norm": 2.2215182781219482, "learning_rate": 2.9626302837269022e-05, "loss": 0.0757, "step": 1442 }, { "epoch": 0.09916162726772952, "grad_norm": 1.4904611110687256, "learning_rate": 2.962556190243148e-05, "loss": 0.077, "step": 1443 }, { "epoch": 0.09923034634414514, "grad_norm": 1.5227704048156738, "learning_rate": 2.9624820243073664e-05, "loss": 0.0589, "step": 1444 }, { "epoch": 0.09929906542056074, "grad_norm": 1.2434884309768677, "learning_rate": 2.9624077859232323e-05, "loss": 0.0824, "step": 1445 }, { "epoch": 0.09936778449697636, "grad_norm": 2.3442153930664062, "learning_rate": 2.9623334750944233e-05, "loss": 0.0947, "step": 1446 }, { "epoch": 0.09943650357339197, "grad_norm": 1.804431438446045, "learning_rate": 2.9622590918246205e-05, "loss": 0.0938, "step": 1447 }, { "epoch": 0.09950522264980759, "grad_norm": 2.1535143852233887, "learning_rate": 2.962184636117508e-05, "loss": 0.0671, "step": 1448 }, { "epoch": 0.0995739417262232, "grad_norm": 1.7621386051177979, "learning_rate": 2.962110107976775e-05, "loss": 0.0723, "step": 1449 }, { "epoch": 0.09964266080263881, "grad_norm": 1.7917230129241943, "learning_rate": 2.962035507406113e-05, "loss": 0.076, "step": 1450 }, { "epoch": 0.09971137987905443, "grad_norm": 1.6762405633926392, "learning_rate": 2.961960834409218e-05, "loss": 0.0801, "step": 1451 }, { "epoch": 0.09978009895547003, "grad_norm": 1.5727847814559937, "learning_rate": 2.9618860889897892e-05, "loss": 0.0614, "step": 1452 }, { "epoch": 0.09984881803188565, "grad_norm": 1.5463197231292725, "learning_rate": 2.9618112711515285e-05, "loss": 0.0639, "step": 1453 }, { "epoch": 0.09991753710830126, "grad_norm": 1.6005431413650513, "learning_rate": 2.961736380898143e-05, "loss": 0.0814, "step": 1454 }, { "epoch": 0.09998625618471688, "grad_norm": 1.8354583978652954, "learning_rate": 2.961661418233343e-05, "loss": 0.051, "step": 1455 }, { "epoch": 0.1000549752611325, "grad_norm": 1.746750831604004, "learning_rate": 2.9615863831608404e-05, "loss": 0.0718, "step": 1456 }, { "epoch": 0.1001236943375481, "grad_norm": 2.1697964668273926, "learning_rate": 2.961511275684354e-05, "loss": 0.0947, "step": 1457 }, { "epoch": 0.10019241341396372, "grad_norm": 1.3536142110824585, "learning_rate": 2.9614360958076032e-05, "loss": 0.0642, "step": 1458 }, { "epoch": 0.10026113249037932, "grad_norm": 1.582986831665039, "learning_rate": 2.9613608435343133e-05, "loss": 0.0593, "step": 1459 }, { "epoch": 0.10032985156679494, "grad_norm": 1.4089781045913696, "learning_rate": 2.9612855188682114e-05, "loss": 0.0648, "step": 1460 }, { "epoch": 0.10039857064321056, "grad_norm": 1.8583204746246338, "learning_rate": 2.9612101218130286e-05, "loss": 0.0753, "step": 1461 }, { "epoch": 0.10046728971962617, "grad_norm": 1.9465703964233398, "learning_rate": 2.9611346523725016e-05, "loss": 0.0651, "step": 1462 }, { "epoch": 0.10053600879604178, "grad_norm": 1.677047610282898, "learning_rate": 2.9610591105503672e-05, "loss": 0.0564, "step": 1463 }, { "epoch": 0.10060472787245739, "grad_norm": 2.3748950958251953, "learning_rate": 2.9609834963503686e-05, "loss": 0.0758, "step": 1464 }, { "epoch": 0.10067344694887301, "grad_norm": 1.790589451789856, "learning_rate": 2.9609078097762507e-05, "loss": 0.083, "step": 1465 }, { "epoch": 0.10074216602528863, "grad_norm": 1.6384632587432861, "learning_rate": 2.960832050831764e-05, "loss": 0.0564, "step": 1466 }, { "epoch": 0.10081088510170423, "grad_norm": 1.683727741241455, "learning_rate": 2.9607562195206607e-05, "loss": 0.077, "step": 1467 }, { "epoch": 0.10087960417811985, "grad_norm": 2.3921959400177, "learning_rate": 2.960680315846697e-05, "loss": 0.0983, "step": 1468 }, { "epoch": 0.10094832325453545, "grad_norm": 1.403456687927246, "learning_rate": 2.9606043398136343e-05, "loss": 0.0689, "step": 1469 }, { "epoch": 0.10101704233095107, "grad_norm": 1.6772300004959106, "learning_rate": 2.960528291425235e-05, "loss": 0.0684, "step": 1470 }, { "epoch": 0.10108576140736668, "grad_norm": 1.3615995645523071, "learning_rate": 2.960452170685267e-05, "loss": 0.063, "step": 1471 }, { "epoch": 0.1011544804837823, "grad_norm": 1.8742234706878662, "learning_rate": 2.9603759775975005e-05, "loss": 0.0707, "step": 1472 }, { "epoch": 0.10122319956019792, "grad_norm": 2.115016222000122, "learning_rate": 2.960299712165711e-05, "loss": 0.067, "step": 1473 }, { "epoch": 0.10129191863661352, "grad_norm": 1.8254194259643555, "learning_rate": 2.9602233743936762e-05, "loss": 0.0683, "step": 1474 }, { "epoch": 0.10136063771302914, "grad_norm": 1.8323243856430054, "learning_rate": 2.9601469642851772e-05, "loss": 0.0616, "step": 1475 }, { "epoch": 0.10142935678944474, "grad_norm": 1.8498414754867554, "learning_rate": 2.9600704818439995e-05, "loss": 0.0845, "step": 1476 }, { "epoch": 0.10149807586586036, "grad_norm": 2.0073752403259277, "learning_rate": 2.9599939270739317e-05, "loss": 0.0617, "step": 1477 }, { "epoch": 0.10156679494227598, "grad_norm": 2.0030572414398193, "learning_rate": 2.9599172999787668e-05, "loss": 0.072, "step": 1478 }, { "epoch": 0.10163551401869159, "grad_norm": 1.9848474264144897, "learning_rate": 2.9598406005623e-05, "loss": 0.0596, "step": 1479 }, { "epoch": 0.1017042330951072, "grad_norm": 1.3483909368515015, "learning_rate": 2.9597638288283317e-05, "loss": 0.0477, "step": 1480 }, { "epoch": 0.10177295217152281, "grad_norm": 1.1583404541015625, "learning_rate": 2.959686984780664e-05, "loss": 0.0882, "step": 1481 }, { "epoch": 0.10184167124793843, "grad_norm": 2.7666804790496826, "learning_rate": 2.9596100684231042e-05, "loss": 0.0569, "step": 1482 }, { "epoch": 0.10191039032435405, "grad_norm": 1.396380066871643, "learning_rate": 2.9595330797594622e-05, "loss": 0.0889, "step": 1483 }, { "epoch": 0.10197910940076965, "grad_norm": 4.5075788497924805, "learning_rate": 2.9594560187935523e-05, "loss": 0.0896, "step": 1484 }, { "epoch": 0.10204782847718527, "grad_norm": 2.3984169960021973, "learning_rate": 2.9593788855291917e-05, "loss": 0.0799, "step": 1485 }, { "epoch": 0.10211654755360088, "grad_norm": 2.1784307956695557, "learning_rate": 2.9593016799702017e-05, "loss": 0.0726, "step": 1486 }, { "epoch": 0.1021852666300165, "grad_norm": 2.19946026802063, "learning_rate": 2.9592244021204062e-05, "loss": 0.0787, "step": 1487 }, { "epoch": 0.1022539857064321, "grad_norm": 3.3720860481262207, "learning_rate": 2.9591470519836345e-05, "loss": 0.089, "step": 1488 }, { "epoch": 0.10232270478284772, "grad_norm": 2.454132080078125, "learning_rate": 2.9590696295637176e-05, "loss": 0.0838, "step": 1489 }, { "epoch": 0.10239142385926334, "grad_norm": 1.492220401763916, "learning_rate": 2.9589921348644904e-05, "loss": 0.0574, "step": 1490 }, { "epoch": 0.10246014293567894, "grad_norm": 1.8158925771713257, "learning_rate": 2.9589145678897928e-05, "loss": 0.0855, "step": 1491 }, { "epoch": 0.10252886201209456, "grad_norm": 2.9150476455688477, "learning_rate": 2.958836928643467e-05, "loss": 0.0604, "step": 1492 }, { "epoch": 0.10259758108851016, "grad_norm": 6.0129218101501465, "learning_rate": 2.958759217129359e-05, "loss": 0.0788, "step": 1493 }, { "epoch": 0.10266630016492578, "grad_norm": 1.7928640842437744, "learning_rate": 2.9586814333513183e-05, "loss": 0.084, "step": 1494 }, { "epoch": 0.1027350192413414, "grad_norm": 1.4486219882965088, "learning_rate": 2.958603577313198e-05, "loss": 0.0741, "step": 1495 }, { "epoch": 0.102803738317757, "grad_norm": 2.2505075931549072, "learning_rate": 2.9585256490188554e-05, "loss": 0.0914, "step": 1496 }, { "epoch": 0.10287245739417263, "grad_norm": 3.933818817138672, "learning_rate": 2.958447648472151e-05, "loss": 0.0823, "step": 1497 }, { "epoch": 0.10294117647058823, "grad_norm": 1.7995061874389648, "learning_rate": 2.9583695756769483e-05, "loss": 0.0632, "step": 1498 }, { "epoch": 0.10300989554700385, "grad_norm": 1.8620697259902954, "learning_rate": 2.9582914306371152e-05, "loss": 0.0915, "step": 1499 }, { "epoch": 0.10307861462341947, "grad_norm": 2.2234575748443604, "learning_rate": 2.958213213356523e-05, "loss": 0.0699, "step": 1500 }, { "epoch": 0.10314733369983507, "grad_norm": 1.576513409614563, "learning_rate": 2.9581349238390462e-05, "loss": 0.0653, "step": 1501 }, { "epoch": 0.10321605277625069, "grad_norm": 3.232984781265259, "learning_rate": 2.9580565620885627e-05, "loss": 0.0824, "step": 1502 }, { "epoch": 0.1032847718526663, "grad_norm": 1.3758571147918701, "learning_rate": 2.9579781281089546e-05, "loss": 0.0689, "step": 1503 }, { "epoch": 0.10335349092908191, "grad_norm": 2.033302068710327, "learning_rate": 2.9578996219041082e-05, "loss": 0.0768, "step": 1504 }, { "epoch": 0.10342221000549752, "grad_norm": 2.277113676071167, "learning_rate": 2.957821043477911e-05, "loss": 0.0949, "step": 1505 }, { "epoch": 0.10349092908191314, "grad_norm": 1.7144912481307983, "learning_rate": 2.957742392834257e-05, "loss": 0.0666, "step": 1506 }, { "epoch": 0.10355964815832876, "grad_norm": 1.2929816246032715, "learning_rate": 2.9576636699770417e-05, "loss": 0.0529, "step": 1507 }, { "epoch": 0.10362836723474436, "grad_norm": 2.7554380893707275, "learning_rate": 2.957584874910165e-05, "loss": 0.0762, "step": 1508 }, { "epoch": 0.10369708631115998, "grad_norm": 2.6773393154144287, "learning_rate": 2.9575060076375305e-05, "loss": 0.0778, "step": 1509 }, { "epoch": 0.10376580538757559, "grad_norm": 3.5014712810516357, "learning_rate": 2.957427068163045e-05, "loss": 0.0704, "step": 1510 }, { "epoch": 0.1038345244639912, "grad_norm": 1.788841724395752, "learning_rate": 2.9573480564906184e-05, "loss": 0.0655, "step": 1511 }, { "epoch": 0.10390324354040682, "grad_norm": 2.398635149002075, "learning_rate": 2.9572689726241657e-05, "loss": 0.0761, "step": 1512 }, { "epoch": 0.10397196261682243, "grad_norm": 3.012305498123169, "learning_rate": 2.9571898165676035e-05, "loss": 0.0718, "step": 1513 }, { "epoch": 0.10404068169323805, "grad_norm": 2.126781702041626, "learning_rate": 2.957110588324854e-05, "loss": 0.0804, "step": 1514 }, { "epoch": 0.10410940076965365, "grad_norm": 2.9326372146606445, "learning_rate": 2.957031287899842e-05, "loss": 0.0782, "step": 1515 }, { "epoch": 0.10417811984606927, "grad_norm": 1.6198583841323853, "learning_rate": 2.9569519152964952e-05, "loss": 0.0603, "step": 1516 }, { "epoch": 0.10424683892248489, "grad_norm": 2.3698389530181885, "learning_rate": 2.9568724705187454e-05, "loss": 0.0556, "step": 1517 }, { "epoch": 0.1043155579989005, "grad_norm": 1.8474323749542236, "learning_rate": 2.9567929535705294e-05, "loss": 0.0707, "step": 1518 }, { "epoch": 0.10438427707531611, "grad_norm": 1.6110050678253174, "learning_rate": 2.9567133644557853e-05, "loss": 0.0663, "step": 1519 }, { "epoch": 0.10445299615173172, "grad_norm": 1.4980560541152954, "learning_rate": 2.9566337031784555e-05, "loss": 0.0581, "step": 1520 }, { "epoch": 0.10452171522814734, "grad_norm": 1.4506834745407104, "learning_rate": 2.9565539697424872e-05, "loss": 0.0701, "step": 1521 }, { "epoch": 0.10459043430456294, "grad_norm": 1.729413390159607, "learning_rate": 2.95647416415183e-05, "loss": 0.0506, "step": 1522 }, { "epoch": 0.10465915338097856, "grad_norm": 2.2066783905029297, "learning_rate": 2.9563942864104366e-05, "loss": 0.0645, "step": 1523 }, { "epoch": 0.10472787245739418, "grad_norm": 1.6716660261154175, "learning_rate": 2.956314336522265e-05, "loss": 0.0805, "step": 1524 }, { "epoch": 0.10479659153380978, "grad_norm": 1.433236837387085, "learning_rate": 2.9562343144912745e-05, "loss": 0.1026, "step": 1525 }, { "epoch": 0.1048653106102254, "grad_norm": 2.4172513484954834, "learning_rate": 2.956154220321431e-05, "loss": 0.0869, "step": 1526 }, { "epoch": 0.104934029686641, "grad_norm": 1.5993080139160156, "learning_rate": 2.9560740540167004e-05, "loss": 0.0741, "step": 1527 }, { "epoch": 0.10500274876305662, "grad_norm": 1.8210828304290771, "learning_rate": 2.9559938155810547e-05, "loss": 0.0678, "step": 1528 }, { "epoch": 0.10507146783947224, "grad_norm": 2.2341744899749756, "learning_rate": 2.955913505018469e-05, "loss": 0.0764, "step": 1529 }, { "epoch": 0.10514018691588785, "grad_norm": 1.6175185441970825, "learning_rate": 2.9558331223329214e-05, "loss": 0.0733, "step": 1530 }, { "epoch": 0.10520890599230347, "grad_norm": 1.3811012506484985, "learning_rate": 2.955752667528394e-05, "loss": 0.0804, "step": 1531 }, { "epoch": 0.10527762506871907, "grad_norm": 1.459374189376831, "learning_rate": 2.9556721406088725e-05, "loss": 0.0593, "step": 1532 }, { "epoch": 0.10534634414513469, "grad_norm": 1.5395705699920654, "learning_rate": 2.955591541578346e-05, "loss": 0.0721, "step": 1533 }, { "epoch": 0.10541506322155031, "grad_norm": 1.2135851383209229, "learning_rate": 2.9555108704408065e-05, "loss": 0.0694, "step": 1534 }, { "epoch": 0.10548378229796591, "grad_norm": 2.040616750717163, "learning_rate": 2.9554301272002515e-05, "loss": 0.0724, "step": 1535 }, { "epoch": 0.10555250137438153, "grad_norm": 1.9813612699508667, "learning_rate": 2.9553493118606796e-05, "loss": 0.0577, "step": 1536 }, { "epoch": 0.10562122045079714, "grad_norm": 1.5523591041564941, "learning_rate": 2.9552684244260953e-05, "loss": 0.0641, "step": 1537 }, { "epoch": 0.10568993952721276, "grad_norm": 1.9437546730041504, "learning_rate": 2.9551874649005053e-05, "loss": 0.054, "step": 1538 }, { "epoch": 0.10575865860362836, "grad_norm": 2.597038984298706, "learning_rate": 2.9551064332879202e-05, "loss": 0.0759, "step": 1539 }, { "epoch": 0.10582737768004398, "grad_norm": 2.9643661975860596, "learning_rate": 2.9550253295923532e-05, "loss": 0.0968, "step": 1540 }, { "epoch": 0.1058960967564596, "grad_norm": 1.3954306840896606, "learning_rate": 2.954944153817823e-05, "loss": 0.0423, "step": 1541 }, { "epoch": 0.1059648158328752, "grad_norm": 2.001610040664673, "learning_rate": 2.954862905968351e-05, "loss": 0.079, "step": 1542 }, { "epoch": 0.10603353490929082, "grad_norm": 1.6642003059387207, "learning_rate": 2.9547815860479615e-05, "loss": 0.0494, "step": 1543 }, { "epoch": 0.10610225398570643, "grad_norm": 1.2532739639282227, "learning_rate": 2.954700194060683e-05, "loss": 0.0498, "step": 1544 }, { "epoch": 0.10617097306212205, "grad_norm": 1.6739997863769531, "learning_rate": 2.9546187300105484e-05, "loss": 0.0506, "step": 1545 }, { "epoch": 0.10623969213853766, "grad_norm": 1.8612629175186157, "learning_rate": 2.954537193901592e-05, "loss": 0.0619, "step": 1546 }, { "epoch": 0.10630841121495327, "grad_norm": 1.8455770015716553, "learning_rate": 2.954455585737853e-05, "loss": 0.0695, "step": 1547 }, { "epoch": 0.10637713029136889, "grad_norm": 1.8875877857208252, "learning_rate": 2.9543739055233748e-05, "loss": 0.0654, "step": 1548 }, { "epoch": 0.10644584936778449, "grad_norm": 1.80319082736969, "learning_rate": 2.9542921532622034e-05, "loss": 0.0648, "step": 1549 }, { "epoch": 0.10651456844420011, "grad_norm": 2.369126319885254, "learning_rate": 2.9542103289583883e-05, "loss": 0.0864, "step": 1550 }, { "epoch": 0.10658328752061573, "grad_norm": 2.1777961254119873, "learning_rate": 2.954128432615984e-05, "loss": 0.0771, "step": 1551 }, { "epoch": 0.10665200659703133, "grad_norm": 1.7156890630722046, "learning_rate": 2.9540464642390462e-05, "loss": 0.0672, "step": 1552 }, { "epoch": 0.10672072567344695, "grad_norm": 2.026038885116577, "learning_rate": 2.953964423831636e-05, "loss": 0.0647, "step": 1553 }, { "epoch": 0.10678944474986256, "grad_norm": 2.2420272827148438, "learning_rate": 2.9538823113978175e-05, "loss": 0.073, "step": 1554 }, { "epoch": 0.10685816382627818, "grad_norm": 1.3878508806228638, "learning_rate": 2.9538001269416583e-05, "loss": 0.0572, "step": 1555 }, { "epoch": 0.10692688290269378, "grad_norm": 1.2066726684570312, "learning_rate": 2.9537178704672295e-05, "loss": 0.0786, "step": 1556 }, { "epoch": 0.1069956019791094, "grad_norm": 1.1977702379226685, "learning_rate": 2.9536355419786063e-05, "loss": 0.066, "step": 1557 }, { "epoch": 0.10706432105552502, "grad_norm": 2.7613141536712646, "learning_rate": 2.9535531414798672e-05, "loss": 0.0751, "step": 1558 }, { "epoch": 0.10713304013194062, "grad_norm": 2.9528355598449707, "learning_rate": 2.9534706689750937e-05, "loss": 0.0703, "step": 1559 }, { "epoch": 0.10720175920835624, "grad_norm": 1.3206104040145874, "learning_rate": 2.9533881244683712e-05, "loss": 0.0628, "step": 1560 }, { "epoch": 0.10727047828477185, "grad_norm": 1.134541630744934, "learning_rate": 2.953305507963789e-05, "loss": 0.0901, "step": 1561 }, { "epoch": 0.10733919736118747, "grad_norm": 1.5158123970031738, "learning_rate": 2.95322281946544e-05, "loss": 0.0598, "step": 1562 }, { "epoch": 0.10740791643760308, "grad_norm": 1.809552550315857, "learning_rate": 2.9531400589774204e-05, "loss": 0.0729, "step": 1563 }, { "epoch": 0.10747663551401869, "grad_norm": 1.1769983768463135, "learning_rate": 2.9530572265038295e-05, "loss": 0.0549, "step": 1564 }, { "epoch": 0.10754535459043431, "grad_norm": 2.3794422149658203, "learning_rate": 2.952974322048771e-05, "loss": 0.0729, "step": 1565 }, { "epoch": 0.10761407366684991, "grad_norm": 2.658609390258789, "learning_rate": 2.9528913456163522e-05, "loss": 0.0703, "step": 1566 }, { "epoch": 0.10768279274326553, "grad_norm": 1.941475510597229, "learning_rate": 2.952808297210683e-05, "loss": 0.0792, "step": 1567 }, { "epoch": 0.10775151181968114, "grad_norm": 1.406612753868103, "learning_rate": 2.9527251768358772e-05, "loss": 0.0925, "step": 1568 }, { "epoch": 0.10782023089609676, "grad_norm": 2.3279995918273926, "learning_rate": 2.952641984496053e-05, "loss": 0.0484, "step": 1569 }, { "epoch": 0.10788894997251237, "grad_norm": 1.805437684059143, "learning_rate": 2.9525587201953314e-05, "loss": 0.0761, "step": 1570 }, { "epoch": 0.10795766904892798, "grad_norm": 2.657757043838501, "learning_rate": 2.9524753839378372e-05, "loss": 0.0859, "step": 1571 }, { "epoch": 0.1080263881253436, "grad_norm": 2.0609898567199707, "learning_rate": 2.9523919757276988e-05, "loss": 0.0736, "step": 1572 }, { "epoch": 0.1080951072017592, "grad_norm": 2.9728105068206787, "learning_rate": 2.9523084955690474e-05, "loss": 0.0826, "step": 1573 }, { "epoch": 0.10816382627817482, "grad_norm": 1.0895066261291504, "learning_rate": 2.9522249434660196e-05, "loss": 0.0656, "step": 1574 }, { "epoch": 0.10823254535459044, "grad_norm": 1.2356173992156982, "learning_rate": 2.9521413194227532e-05, "loss": 0.0691, "step": 1575 }, { "epoch": 0.10830126443100604, "grad_norm": 3.0264384746551514, "learning_rate": 2.952057623443392e-05, "loss": 0.0705, "step": 1576 }, { "epoch": 0.10836998350742166, "grad_norm": 2.4038102626800537, "learning_rate": 2.951973855532081e-05, "loss": 0.0847, "step": 1577 }, { "epoch": 0.10843870258383727, "grad_norm": 1.5876153707504272, "learning_rate": 2.9518900156929702e-05, "loss": 0.0678, "step": 1578 }, { "epoch": 0.10850742166025289, "grad_norm": 1.9912636280059814, "learning_rate": 2.9518061039302133e-05, "loss": 0.0579, "step": 1579 }, { "epoch": 0.1085761407366685, "grad_norm": 2.591609239578247, "learning_rate": 2.9517221202479665e-05, "loss": 0.0621, "step": 1580 }, { "epoch": 0.10864485981308411, "grad_norm": 2.2122087478637695, "learning_rate": 2.9516380646503904e-05, "loss": 0.0792, "step": 1581 }, { "epoch": 0.10871357888949973, "grad_norm": 1.4950019121170044, "learning_rate": 2.9515539371416492e-05, "loss": 0.0723, "step": 1582 }, { "epoch": 0.10878229796591533, "grad_norm": 2.366748809814453, "learning_rate": 2.9514697377259102e-05, "loss": 0.081, "step": 1583 }, { "epoch": 0.10885101704233095, "grad_norm": 2.273118495941162, "learning_rate": 2.9513854664073447e-05, "loss": 0.0647, "step": 1584 }, { "epoch": 0.10891973611874656, "grad_norm": 1.3420623540878296, "learning_rate": 2.951301123190127e-05, "loss": 0.0634, "step": 1585 }, { "epoch": 0.10898845519516218, "grad_norm": 2.5415327548980713, "learning_rate": 2.9512167080784353e-05, "loss": 0.0681, "step": 1586 }, { "epoch": 0.1090571742715778, "grad_norm": 2.6263768672943115, "learning_rate": 2.9511322210764515e-05, "loss": 0.089, "step": 1587 }, { "epoch": 0.1091258933479934, "grad_norm": 1.3175415992736816, "learning_rate": 2.951047662188361e-05, "loss": 0.0793, "step": 1588 }, { "epoch": 0.10919461242440902, "grad_norm": 1.7377058267593384, "learning_rate": 2.9509630314183522e-05, "loss": 0.0856, "step": 1589 }, { "epoch": 0.10926333150082462, "grad_norm": 1.6402006149291992, "learning_rate": 2.950878328770618e-05, "loss": 0.0724, "step": 1590 }, { "epoch": 0.10933205057724024, "grad_norm": 1.7518261671066284, "learning_rate": 2.9507935542493542e-05, "loss": 0.0807, "step": 1591 }, { "epoch": 0.10940076965365586, "grad_norm": 2.004511594772339, "learning_rate": 2.950708707858761e-05, "loss": 0.0523, "step": 1592 }, { "epoch": 0.10946948873007147, "grad_norm": 1.6677169799804688, "learning_rate": 2.9506237896030405e-05, "loss": 0.0716, "step": 1593 }, { "epoch": 0.10953820780648708, "grad_norm": 1.8242932558059692, "learning_rate": 2.9505387994864e-05, "loss": 0.0749, "step": 1594 }, { "epoch": 0.10960692688290269, "grad_norm": 2.6345531940460205, "learning_rate": 2.950453737513049e-05, "loss": 0.0877, "step": 1595 }, { "epoch": 0.10967564595931831, "grad_norm": 1.898119330406189, "learning_rate": 2.9503686036872024e-05, "loss": 0.082, "step": 1596 }, { "epoch": 0.10974436503573393, "grad_norm": 1.502055048942566, "learning_rate": 2.950283398013077e-05, "loss": 0.0692, "step": 1597 }, { "epoch": 0.10981308411214953, "grad_norm": 1.2578760385513306, "learning_rate": 2.9501981204948933e-05, "loss": 0.0619, "step": 1598 }, { "epoch": 0.10988180318856515, "grad_norm": 1.2933151721954346, "learning_rate": 2.9501127711368767e-05, "loss": 0.0559, "step": 1599 }, { "epoch": 0.10995052226498075, "grad_norm": 1.3840333223342896, "learning_rate": 2.9500273499432542e-05, "loss": 0.062, "step": 1600 }, { "epoch": 0.11001924134139637, "grad_norm": 2.0354602336883545, "learning_rate": 2.9499418569182583e-05, "loss": 0.0599, "step": 1601 }, { "epoch": 0.11008796041781198, "grad_norm": 1.7541197538375854, "learning_rate": 2.9498562920661237e-05, "loss": 0.0499, "step": 1602 }, { "epoch": 0.1101566794942276, "grad_norm": 1.5952192544937134, "learning_rate": 2.949770655391089e-05, "loss": 0.0655, "step": 1603 }, { "epoch": 0.11022539857064322, "grad_norm": 1.5572755336761475, "learning_rate": 2.949684946897397e-05, "loss": 0.0813, "step": 1604 }, { "epoch": 0.11029411764705882, "grad_norm": 2.3321285247802734, "learning_rate": 2.9495991665892928e-05, "loss": 0.0572, "step": 1605 }, { "epoch": 0.11036283672347444, "grad_norm": 1.2751384973526, "learning_rate": 2.9495133144710265e-05, "loss": 0.0564, "step": 1606 }, { "epoch": 0.11043155579989004, "grad_norm": 2.4116342067718506, "learning_rate": 2.9494273905468503e-05, "loss": 0.0802, "step": 1607 }, { "epoch": 0.11050027487630566, "grad_norm": 2.1407079696655273, "learning_rate": 2.9493413948210213e-05, "loss": 0.0895, "step": 1608 }, { "epoch": 0.11056899395272128, "grad_norm": 2.131193161010742, "learning_rate": 2.9492553272977988e-05, "loss": 0.0773, "step": 1609 }, { "epoch": 0.11063771302913689, "grad_norm": 1.797736406326294, "learning_rate": 2.9491691879814477e-05, "loss": 0.0767, "step": 1610 }, { "epoch": 0.1107064321055525, "grad_norm": 1.2523574829101562, "learning_rate": 2.9490829768762337e-05, "loss": 0.0492, "step": 1611 }, { "epoch": 0.11077515118196811, "grad_norm": 1.6414501667022705, "learning_rate": 2.9489966939864286e-05, "loss": 0.0685, "step": 1612 }, { "epoch": 0.11084387025838373, "grad_norm": 2.01462459564209, "learning_rate": 2.9489103393163065e-05, "loss": 0.0624, "step": 1613 }, { "epoch": 0.11091258933479935, "grad_norm": 1.3716456890106201, "learning_rate": 2.948823912870145e-05, "loss": 0.0652, "step": 1614 }, { "epoch": 0.11098130841121495, "grad_norm": 2.5252413749694824, "learning_rate": 2.9487374146522247e-05, "loss": 0.0606, "step": 1615 }, { "epoch": 0.11105002748763057, "grad_norm": 1.9859522581100464, "learning_rate": 2.9486508446668322e-05, "loss": 0.0679, "step": 1616 }, { "epoch": 0.11111874656404618, "grad_norm": 1.6926873922348022, "learning_rate": 2.9485642029182548e-05, "loss": 0.059, "step": 1617 }, { "epoch": 0.1111874656404618, "grad_norm": 1.05936861038208, "learning_rate": 2.9484774894107852e-05, "loss": 0.0808, "step": 1618 }, { "epoch": 0.1112561847168774, "grad_norm": 1.5406837463378906, "learning_rate": 2.9483907041487182e-05, "loss": 0.0579, "step": 1619 }, { "epoch": 0.11132490379329302, "grad_norm": 1.7852040529251099, "learning_rate": 2.9483038471363538e-05, "loss": 0.069, "step": 1620 }, { "epoch": 0.11139362286970864, "grad_norm": 1.9050500392913818, "learning_rate": 2.9482169183779948e-05, "loss": 0.0733, "step": 1621 }, { "epoch": 0.11146234194612424, "grad_norm": 1.2260042428970337, "learning_rate": 2.9481299178779465e-05, "loss": 0.0614, "step": 1622 }, { "epoch": 0.11153106102253986, "grad_norm": 2.046043872833252, "learning_rate": 2.9480428456405194e-05, "loss": 0.0538, "step": 1623 }, { "epoch": 0.11159978009895546, "grad_norm": 1.2984486818313599, "learning_rate": 2.9479557016700268e-05, "loss": 0.0646, "step": 1624 }, { "epoch": 0.11166849917537108, "grad_norm": 4.286612033843994, "learning_rate": 2.9478684859707856e-05, "loss": 0.0942, "step": 1625 }, { "epoch": 0.1117372182517867, "grad_norm": 1.6667174100875854, "learning_rate": 2.9477811985471166e-05, "loss": 0.0626, "step": 1626 }, { "epoch": 0.1118059373282023, "grad_norm": 1.6227195262908936, "learning_rate": 2.947693839403343e-05, "loss": 0.0559, "step": 1627 }, { "epoch": 0.11187465640461793, "grad_norm": 1.458451747894287, "learning_rate": 2.9476064085437935e-05, "loss": 0.0672, "step": 1628 }, { "epoch": 0.11194337548103353, "grad_norm": 2.230273723602295, "learning_rate": 2.9475189059727985e-05, "loss": 0.052, "step": 1629 }, { "epoch": 0.11201209455744915, "grad_norm": 2.927718162536621, "learning_rate": 2.9474313316946927e-05, "loss": 0.0801, "step": 1630 }, { "epoch": 0.11208081363386477, "grad_norm": 1.4344000816345215, "learning_rate": 2.9473436857138145e-05, "loss": 0.0648, "step": 1631 }, { "epoch": 0.11214953271028037, "grad_norm": 1.2153674364089966, "learning_rate": 2.9472559680345057e-05, "loss": 0.052, "step": 1632 }, { "epoch": 0.11221825178669599, "grad_norm": 1.4894062280654907, "learning_rate": 2.9471681786611118e-05, "loss": 0.0691, "step": 1633 }, { "epoch": 0.1122869708631116, "grad_norm": 1.3971396684646606, "learning_rate": 2.9470803175979817e-05, "loss": 0.059, "step": 1634 }, { "epoch": 0.11235568993952721, "grad_norm": 1.6378917694091797, "learning_rate": 2.946992384849468e-05, "loss": 0.0606, "step": 1635 }, { "epoch": 0.11242440901594282, "grad_norm": 2.2177417278289795, "learning_rate": 2.9469043804199256e-05, "loss": 0.0773, "step": 1636 }, { "epoch": 0.11249312809235844, "grad_norm": 1.2723857164382935, "learning_rate": 2.9468163043137154e-05, "loss": 0.0599, "step": 1637 }, { "epoch": 0.11256184716877406, "grad_norm": 1.679605484008789, "learning_rate": 2.9467281565352e-05, "loss": 0.0853, "step": 1638 }, { "epoch": 0.11263056624518966, "grad_norm": 1.4806387424468994, "learning_rate": 2.946639937088746e-05, "loss": 0.0699, "step": 1639 }, { "epoch": 0.11269928532160528, "grad_norm": 1.8962076902389526, "learning_rate": 2.9465516459787236e-05, "loss": 0.0921, "step": 1640 }, { "epoch": 0.11276800439802089, "grad_norm": 2.1140005588531494, "learning_rate": 2.946463283209507e-05, "loss": 0.0874, "step": 1641 }, { "epoch": 0.1128367234744365, "grad_norm": 2.7933828830718994, "learning_rate": 2.946374848785473e-05, "loss": 0.1008, "step": 1642 }, { "epoch": 0.11290544255085212, "grad_norm": 2.291383743286133, "learning_rate": 2.9462863427110022e-05, "loss": 0.0819, "step": 1643 }, { "epoch": 0.11297416162726773, "grad_norm": 1.2377433776855469, "learning_rate": 2.94619776499048e-05, "loss": 0.0748, "step": 1644 }, { "epoch": 0.11304288070368335, "grad_norm": 2.0254476070404053, "learning_rate": 2.9461091156282936e-05, "loss": 0.0903, "step": 1645 }, { "epoch": 0.11311159978009895, "grad_norm": 2.19692325592041, "learning_rate": 2.9460203946288344e-05, "loss": 0.0988, "step": 1646 }, { "epoch": 0.11318031885651457, "grad_norm": 1.6785231828689575, "learning_rate": 2.945931601996498e-05, "loss": 0.0775, "step": 1647 }, { "epoch": 0.11324903793293019, "grad_norm": 1.6552029848098755, "learning_rate": 2.9458427377356823e-05, "loss": 0.076, "step": 1648 }, { "epoch": 0.1133177570093458, "grad_norm": 1.7310131788253784, "learning_rate": 2.9457538018507905e-05, "loss": 0.072, "step": 1649 }, { "epoch": 0.11338647608576141, "grad_norm": 0.9958548545837402, "learning_rate": 2.945664794346228e-05, "loss": 0.0723, "step": 1650 }, { "epoch": 0.11345519516217702, "grad_norm": 2.5274147987365723, "learning_rate": 2.9455757152264027e-05, "loss": 0.0686, "step": 1651 }, { "epoch": 0.11352391423859264, "grad_norm": 1.78339421749115, "learning_rate": 2.945486564495729e-05, "loss": 0.0673, "step": 1652 }, { "epoch": 0.11359263331500824, "grad_norm": 1.2492752075195312, "learning_rate": 2.945397342158623e-05, "loss": 0.0584, "step": 1653 }, { "epoch": 0.11366135239142386, "grad_norm": 2.087890863418579, "learning_rate": 2.9453080482195036e-05, "loss": 0.0745, "step": 1654 }, { "epoch": 0.11373007146783948, "grad_norm": 1.1353931427001953, "learning_rate": 2.9452186826827956e-05, "loss": 0.0485, "step": 1655 }, { "epoch": 0.11379879054425508, "grad_norm": 1.2016814947128296, "learning_rate": 2.945129245552925e-05, "loss": 0.0522, "step": 1656 }, { "epoch": 0.1138675096206707, "grad_norm": 2.375480890274048, "learning_rate": 2.9450397368343224e-05, "loss": 0.085, "step": 1657 }, { "epoch": 0.1139362286970863, "grad_norm": 1.7769386768341064, "learning_rate": 2.9449501565314223e-05, "loss": 0.0618, "step": 1658 }, { "epoch": 0.11400494777350192, "grad_norm": 1.428658366203308, "learning_rate": 2.9448605046486623e-05, "loss": 0.0503, "step": 1659 }, { "epoch": 0.11407366684991754, "grad_norm": 1.0009559392929077, "learning_rate": 2.9447707811904833e-05, "loss": 0.0666, "step": 1660 }, { "epoch": 0.11414238592633315, "grad_norm": 1.7366750240325928, "learning_rate": 2.94468098616133e-05, "loss": 0.0753, "step": 1661 }, { "epoch": 0.11421110500274877, "grad_norm": 1.7553215026855469, "learning_rate": 2.944591119565651e-05, "loss": 0.068, "step": 1662 }, { "epoch": 0.11427982407916437, "grad_norm": 1.8277359008789062, "learning_rate": 2.9445011814078978e-05, "loss": 0.0573, "step": 1663 }, { "epoch": 0.11434854315557999, "grad_norm": 2.1221776008605957, "learning_rate": 2.944411171692526e-05, "loss": 0.0844, "step": 1664 }, { "epoch": 0.1144172622319956, "grad_norm": 2.6693193912506104, "learning_rate": 2.944321090423994e-05, "loss": 0.0833, "step": 1665 }, { "epoch": 0.11448598130841121, "grad_norm": 1.975422978401184, "learning_rate": 2.9442309376067645e-05, "loss": 0.081, "step": 1666 }, { "epoch": 0.11455470038482683, "grad_norm": 1.7058950662612915, "learning_rate": 2.944140713245304e-05, "loss": 0.0789, "step": 1667 }, { "epoch": 0.11462341946124244, "grad_norm": 2.069336175918579, "learning_rate": 2.9440504173440812e-05, "loss": 0.0647, "step": 1668 }, { "epoch": 0.11469213853765806, "grad_norm": 1.8804925680160522, "learning_rate": 2.94396004990757e-05, "loss": 0.0688, "step": 1669 }, { "epoch": 0.11476085761407366, "grad_norm": 1.4600805044174194, "learning_rate": 2.9438696109402463e-05, "loss": 0.0664, "step": 1670 }, { "epoch": 0.11482957669048928, "grad_norm": 1.4966977834701538, "learning_rate": 2.9437791004465903e-05, "loss": 0.0527, "step": 1671 }, { "epoch": 0.1148982957669049, "grad_norm": 1.811240315437317, "learning_rate": 2.9436885184310868e-05, "loss": 0.061, "step": 1672 }, { "epoch": 0.1149670148433205, "grad_norm": 1.178070068359375, "learning_rate": 2.943597864898221e-05, "loss": 0.0423, "step": 1673 }, { "epoch": 0.11503573391973612, "grad_norm": 1.2558339834213257, "learning_rate": 2.9435071398524857e-05, "loss": 0.0535, "step": 1674 }, { "epoch": 0.11510445299615173, "grad_norm": 1.799453616142273, "learning_rate": 2.9434163432983743e-05, "loss": 0.0793, "step": 1675 }, { "epoch": 0.11517317207256735, "grad_norm": 1.2941354513168335, "learning_rate": 2.943325475240385e-05, "loss": 0.0631, "step": 1676 }, { "epoch": 0.11524189114898296, "grad_norm": 1.5428601503372192, "learning_rate": 2.9432345356830182e-05, "loss": 0.0598, "step": 1677 }, { "epoch": 0.11531061022539857, "grad_norm": 1.848998785018921, "learning_rate": 2.9431435246307806e-05, "loss": 0.0691, "step": 1678 }, { "epoch": 0.11537932930181419, "grad_norm": 3.0086305141448975, "learning_rate": 2.943052442088179e-05, "loss": 0.0746, "step": 1679 }, { "epoch": 0.11544804837822979, "grad_norm": 3.534637928009033, "learning_rate": 2.942961288059727e-05, "loss": 0.0678, "step": 1680 }, { "epoch": 0.11551676745464541, "grad_norm": 1.3684864044189453, "learning_rate": 2.9428700625499386e-05, "loss": 0.0555, "step": 1681 }, { "epoch": 0.11558548653106102, "grad_norm": 2.2051291465759277, "learning_rate": 2.942778765563334e-05, "loss": 0.0751, "step": 1682 }, { "epoch": 0.11565420560747663, "grad_norm": 2.1432440280914307, "learning_rate": 2.9426873971044358e-05, "loss": 0.0711, "step": 1683 }, { "epoch": 0.11572292468389225, "grad_norm": 2.601938486099243, "learning_rate": 2.9425959571777702e-05, "loss": 0.0698, "step": 1684 }, { "epoch": 0.11579164376030786, "grad_norm": 1.6735713481903076, "learning_rate": 2.9425044457878663e-05, "loss": 0.0446, "step": 1685 }, { "epoch": 0.11586036283672348, "grad_norm": 1.685272455215454, "learning_rate": 2.9424128629392577e-05, "loss": 0.0904, "step": 1686 }, { "epoch": 0.11592908191313908, "grad_norm": 1.781658411026001, "learning_rate": 2.9423212086364813e-05, "loss": 0.094, "step": 1687 }, { "epoch": 0.1159978009895547, "grad_norm": 1.8060293197631836, "learning_rate": 2.9422294828840778e-05, "loss": 0.0606, "step": 1688 }, { "epoch": 0.11606652006597032, "grad_norm": 1.7114484310150146, "learning_rate": 2.9421376856865912e-05, "loss": 0.0617, "step": 1689 }, { "epoch": 0.11613523914238592, "grad_norm": 1.018959641456604, "learning_rate": 2.9420458170485678e-05, "loss": 0.0671, "step": 1690 }, { "epoch": 0.11620395821880154, "grad_norm": 1.1987180709838867, "learning_rate": 2.9419538769745598e-05, "loss": 0.0657, "step": 1691 }, { "epoch": 0.11627267729521715, "grad_norm": 1.7896560430526733, "learning_rate": 2.9418618654691215e-05, "loss": 0.0755, "step": 1692 }, { "epoch": 0.11634139637163277, "grad_norm": 1.1974798440933228, "learning_rate": 2.94176978253681e-05, "loss": 0.066, "step": 1693 }, { "epoch": 0.11641011544804838, "grad_norm": 1.3202587366104126, "learning_rate": 2.9416776281821882e-05, "loss": 0.0734, "step": 1694 }, { "epoch": 0.11647883452446399, "grad_norm": 1.5974280834197998, "learning_rate": 2.94158540240982e-05, "loss": 0.0641, "step": 1695 }, { "epoch": 0.11654755360087961, "grad_norm": 1.3792340755462646, "learning_rate": 2.9414931052242756e-05, "loss": 0.0771, "step": 1696 }, { "epoch": 0.11661627267729521, "grad_norm": 2.399723529815674, "learning_rate": 2.9414007366301257e-05, "loss": 0.0836, "step": 1697 }, { "epoch": 0.11668499175371083, "grad_norm": 1.0839849710464478, "learning_rate": 2.941308296631947e-05, "loss": 0.0607, "step": 1698 }, { "epoch": 0.11675371083012644, "grad_norm": 2.26743745803833, "learning_rate": 2.9412157852343185e-05, "loss": 0.0741, "step": 1699 }, { "epoch": 0.11682242990654206, "grad_norm": 2.3218321800231934, "learning_rate": 2.941123202441823e-05, "loss": 0.0649, "step": 1700 }, { "epoch": 0.11689114898295767, "grad_norm": 2.2380728721618652, "learning_rate": 2.9410305482590468e-05, "loss": 0.0594, "step": 1701 }, { "epoch": 0.11695986805937328, "grad_norm": 1.530924677848816, "learning_rate": 2.94093782269058e-05, "loss": 0.061, "step": 1702 }, { "epoch": 0.1170285871357889, "grad_norm": 0.9826022982597351, "learning_rate": 2.9408450257410156e-05, "loss": 0.0577, "step": 1703 }, { "epoch": 0.1170973062122045, "grad_norm": 1.4709402322769165, "learning_rate": 2.940752157414951e-05, "loss": 0.0705, "step": 1704 }, { "epoch": 0.11716602528862012, "grad_norm": 1.6973172426223755, "learning_rate": 2.9406592177169865e-05, "loss": 0.0696, "step": 1705 }, { "epoch": 0.11723474436503574, "grad_norm": 1.7132145166397095, "learning_rate": 2.940566206651727e-05, "loss": 0.0522, "step": 1706 }, { "epoch": 0.11730346344145134, "grad_norm": 1.714625358581543, "learning_rate": 2.9404731242237784e-05, "loss": 0.0863, "step": 1707 }, { "epoch": 0.11737218251786696, "grad_norm": 1.2132048606872559, "learning_rate": 2.940379970437753e-05, "loss": 0.0619, "step": 1708 }, { "epoch": 0.11744090159428257, "grad_norm": 1.4322912693023682, "learning_rate": 2.940286745298265e-05, "loss": 0.0591, "step": 1709 }, { "epoch": 0.11750962067069819, "grad_norm": 1.0954653024673462, "learning_rate": 2.9401934488099328e-05, "loss": 0.0672, "step": 1710 }, { "epoch": 0.1175783397471138, "grad_norm": 2.119910717010498, "learning_rate": 2.9401000809773786e-05, "loss": 0.1013, "step": 1711 }, { "epoch": 0.11764705882352941, "grad_norm": 1.4679090976715088, "learning_rate": 2.9400066418052264e-05, "loss": 0.0682, "step": 1712 }, { "epoch": 0.11771577789994503, "grad_norm": 1.78147292137146, "learning_rate": 2.9399131312981062e-05, "loss": 0.0453, "step": 1713 }, { "epoch": 0.11778449697636063, "grad_norm": 1.6451534032821655, "learning_rate": 2.9398195494606498e-05, "loss": 0.0669, "step": 1714 }, { "epoch": 0.11785321605277625, "grad_norm": 1.416390299797058, "learning_rate": 2.9397258962974924e-05, "loss": 0.0602, "step": 1715 }, { "epoch": 0.11792193512919186, "grad_norm": 1.5532162189483643, "learning_rate": 2.939632171813275e-05, "loss": 0.0706, "step": 1716 }, { "epoch": 0.11799065420560748, "grad_norm": 2.215268850326538, "learning_rate": 2.9395383760126388e-05, "loss": 0.0737, "step": 1717 }, { "epoch": 0.1180593732820231, "grad_norm": 1.2040982246398926, "learning_rate": 2.9394445089002313e-05, "loss": 0.0637, "step": 1718 }, { "epoch": 0.1181280923584387, "grad_norm": 1.2982048988342285, "learning_rate": 2.939350570480702e-05, "loss": 0.0532, "step": 1719 }, { "epoch": 0.11819681143485432, "grad_norm": 1.7737013101577759, "learning_rate": 2.9392565607587048e-05, "loss": 0.0672, "step": 1720 }, { "epoch": 0.11826553051126992, "grad_norm": 1.0142216682434082, "learning_rate": 2.9391624797388967e-05, "loss": 0.0498, "step": 1721 }, { "epoch": 0.11833424958768554, "grad_norm": 1.4414198398590088, "learning_rate": 2.939068327425938e-05, "loss": 0.0544, "step": 1722 }, { "epoch": 0.11840296866410116, "grad_norm": 1.6958986520767212, "learning_rate": 2.938974103824493e-05, "loss": 0.0608, "step": 1723 }, { "epoch": 0.11847168774051677, "grad_norm": 1.675709843635559, "learning_rate": 2.9388798089392293e-05, "loss": 0.0679, "step": 1724 }, { "epoch": 0.11854040681693238, "grad_norm": 2.9962596893310547, "learning_rate": 2.938785442774818e-05, "loss": 0.0725, "step": 1725 }, { "epoch": 0.11860912589334799, "grad_norm": 1.2834795713424683, "learning_rate": 2.938691005335934e-05, "loss": 0.0672, "step": 1726 }, { "epoch": 0.11867784496976361, "grad_norm": 1.4288638830184937, "learning_rate": 2.9385964966272548e-05, "loss": 0.0701, "step": 1727 }, { "epoch": 0.11874656404617923, "grad_norm": 1.4239888191223145, "learning_rate": 2.9385019166534635e-05, "loss": 0.0687, "step": 1728 }, { "epoch": 0.11881528312259483, "grad_norm": 1.725947380065918, "learning_rate": 2.9384072654192448e-05, "loss": 0.0838, "step": 1729 }, { "epoch": 0.11888400219901045, "grad_norm": 2.2466790676116943, "learning_rate": 2.938312542929287e-05, "loss": 0.0732, "step": 1730 }, { "epoch": 0.11895272127542605, "grad_norm": 1.5146512985229492, "learning_rate": 2.938217749188283e-05, "loss": 0.0593, "step": 1731 }, { "epoch": 0.11902144035184167, "grad_norm": 2.2990167140960693, "learning_rate": 2.9381228842009285e-05, "loss": 0.0724, "step": 1732 }, { "epoch": 0.11909015942825728, "grad_norm": 1.8050678968429565, "learning_rate": 2.938027947971923e-05, "loss": 0.0682, "step": 1733 }, { "epoch": 0.1191588785046729, "grad_norm": 2.185605764389038, "learning_rate": 2.9379329405059695e-05, "loss": 0.0689, "step": 1734 }, { "epoch": 0.11922759758108852, "grad_norm": 2.2622342109680176, "learning_rate": 2.9378378618077746e-05, "loss": 0.0555, "step": 1735 }, { "epoch": 0.11929631665750412, "grad_norm": 1.9296631813049316, "learning_rate": 2.9377427118820476e-05, "loss": 0.0729, "step": 1736 }, { "epoch": 0.11936503573391974, "grad_norm": 1.84392511844635, "learning_rate": 2.9376474907335027e-05, "loss": 0.0741, "step": 1737 }, { "epoch": 0.11943375481033534, "grad_norm": 1.6701384782791138, "learning_rate": 2.9375521983668568e-05, "loss": 0.0589, "step": 1738 }, { "epoch": 0.11950247388675096, "grad_norm": 1.4475433826446533, "learning_rate": 2.9374568347868304e-05, "loss": 0.0585, "step": 1739 }, { "epoch": 0.11957119296316658, "grad_norm": 1.3410913944244385, "learning_rate": 2.937361399998148e-05, "loss": 0.0683, "step": 1740 }, { "epoch": 0.11963991203958219, "grad_norm": 1.7816424369812012, "learning_rate": 2.937265894005537e-05, "loss": 0.0605, "step": 1741 }, { "epoch": 0.1197086311159978, "grad_norm": 1.586962342262268, "learning_rate": 2.937170316813728e-05, "loss": 0.0711, "step": 1742 }, { "epoch": 0.11977735019241341, "grad_norm": 1.6247246265411377, "learning_rate": 2.9370746684274563e-05, "loss": 0.0739, "step": 1743 }, { "epoch": 0.11984606926882903, "grad_norm": 1.338719367980957, "learning_rate": 2.9369789488514605e-05, "loss": 0.0653, "step": 1744 }, { "epoch": 0.11991478834524465, "grad_norm": 1.567333698272705, "learning_rate": 2.9368831580904815e-05, "loss": 0.0849, "step": 1745 }, { "epoch": 0.11998350742166025, "grad_norm": 1.2374533414840698, "learning_rate": 2.936787296149265e-05, "loss": 0.05, "step": 1746 }, { "epoch": 0.12005222649807587, "grad_norm": 1.2753788232803345, "learning_rate": 2.93669136303256e-05, "loss": 0.0653, "step": 1747 }, { "epoch": 0.12012094557449148, "grad_norm": 2.795503854751587, "learning_rate": 2.936595358745118e-05, "loss": 0.0676, "step": 1748 }, { "epoch": 0.1201896646509071, "grad_norm": 1.3185913562774658, "learning_rate": 2.9364992832916956e-05, "loss": 0.0725, "step": 1749 }, { "epoch": 0.1202583837273227, "grad_norm": 2.1619365215301514, "learning_rate": 2.9364031366770525e-05, "loss": 0.0655, "step": 1750 }, { "epoch": 0.12032710280373832, "grad_norm": 1.7006916999816895, "learning_rate": 2.9363069189059508e-05, "loss": 0.0529, "step": 1751 }, { "epoch": 0.12039582188015394, "grad_norm": 1.7013065814971924, "learning_rate": 2.9362106299831575e-05, "loss": 0.0615, "step": 1752 }, { "epoch": 0.12046454095656954, "grad_norm": 1.7892615795135498, "learning_rate": 2.936114269913442e-05, "loss": 0.0595, "step": 1753 }, { "epoch": 0.12053326003298516, "grad_norm": 1.3081766366958618, "learning_rate": 2.936017838701578e-05, "loss": 0.0409, "step": 1754 }, { "epoch": 0.12060197910940076, "grad_norm": 1.7016916275024414, "learning_rate": 2.9359213363523428e-05, "loss": 0.0825, "step": 1755 }, { "epoch": 0.12067069818581638, "grad_norm": 1.5074946880340576, "learning_rate": 2.935824762870517e-05, "loss": 0.0576, "step": 1756 }, { "epoch": 0.120739417262232, "grad_norm": 1.1601860523223877, "learning_rate": 2.935728118260884e-05, "loss": 0.0644, "step": 1757 }, { "epoch": 0.1208081363386476, "grad_norm": 1.559400200843811, "learning_rate": 2.9356314025282318e-05, "loss": 0.0598, "step": 1758 }, { "epoch": 0.12087685541506323, "grad_norm": 1.9627565145492554, "learning_rate": 2.9355346156773516e-05, "loss": 0.074, "step": 1759 }, { "epoch": 0.12094557449147883, "grad_norm": 2.4371306896209717, "learning_rate": 2.935437757713038e-05, "loss": 0.0768, "step": 1760 }, { "epoch": 0.12101429356789445, "grad_norm": 3.3034346103668213, "learning_rate": 2.9353408286400885e-05, "loss": 0.0641, "step": 1761 }, { "epoch": 0.12108301264431007, "grad_norm": 1.351914644241333, "learning_rate": 2.935243828463306e-05, "loss": 0.054, "step": 1762 }, { "epoch": 0.12115173172072567, "grad_norm": 3.019726514816284, "learning_rate": 2.9351467571874943e-05, "loss": 0.0847, "step": 1763 }, { "epoch": 0.12122045079714129, "grad_norm": 2.5074236392974854, "learning_rate": 2.935049614817463e-05, "loss": 0.0762, "step": 1764 }, { "epoch": 0.1212891698735569, "grad_norm": 2.4886088371276855, "learning_rate": 2.934952401358024e-05, "loss": 0.097, "step": 1765 }, { "epoch": 0.12135788894997251, "grad_norm": 1.2346774339675903, "learning_rate": 2.9348551168139934e-05, "loss": 0.0562, "step": 1766 }, { "epoch": 0.12142660802638812, "grad_norm": 1.8733620643615723, "learning_rate": 2.93475776119019e-05, "loss": 0.0603, "step": 1767 }, { "epoch": 0.12149532710280374, "grad_norm": 2.6768453121185303, "learning_rate": 2.9346603344914377e-05, "loss": 0.0652, "step": 1768 }, { "epoch": 0.12156404617921936, "grad_norm": 1.3838768005371094, "learning_rate": 2.934562836722561e-05, "loss": 0.061, "step": 1769 }, { "epoch": 0.12163276525563496, "grad_norm": 2.4499728679656982, "learning_rate": 2.9344652678883913e-05, "loss": 0.0754, "step": 1770 }, { "epoch": 0.12170148433205058, "grad_norm": 2.2928719520568848, "learning_rate": 2.934367627993761e-05, "loss": 0.0499, "step": 1771 }, { "epoch": 0.12177020340846618, "grad_norm": 2.0447704792022705, "learning_rate": 2.9342699170435074e-05, "loss": 0.0654, "step": 1772 }, { "epoch": 0.1218389224848818, "grad_norm": 1.7016568183898926, "learning_rate": 2.9341721350424708e-05, "loss": 0.0626, "step": 1773 }, { "epoch": 0.12190764156129742, "grad_norm": 1.9134193658828735, "learning_rate": 2.9340742819954954e-05, "loss": 0.0587, "step": 1774 }, { "epoch": 0.12197636063771303, "grad_norm": 1.395135521888733, "learning_rate": 2.933976357907428e-05, "loss": 0.0819, "step": 1775 }, { "epoch": 0.12204507971412865, "grad_norm": 2.148826837539673, "learning_rate": 2.9338783627831203e-05, "loss": 0.0604, "step": 1776 }, { "epoch": 0.12211379879054425, "grad_norm": 1.7414973974227905, "learning_rate": 2.9337802966274262e-05, "loss": 0.0723, "step": 1777 }, { "epoch": 0.12218251786695987, "grad_norm": 1.7491158246994019, "learning_rate": 2.933682159445204e-05, "loss": 0.0663, "step": 1778 }, { "epoch": 0.12225123694337547, "grad_norm": 1.3213285207748413, "learning_rate": 2.9335839512413155e-05, "loss": 0.0778, "step": 1779 }, { "epoch": 0.12231995601979109, "grad_norm": 2.450066089630127, "learning_rate": 2.9334856720206252e-05, "loss": 0.0743, "step": 1780 }, { "epoch": 0.12238867509620671, "grad_norm": 1.9005168676376343, "learning_rate": 2.933387321788001e-05, "loss": 0.0527, "step": 1781 }, { "epoch": 0.12245739417262232, "grad_norm": 1.851071834564209, "learning_rate": 2.933288900548317e-05, "loss": 0.0547, "step": 1782 }, { "epoch": 0.12252611324903794, "grad_norm": 2.1580617427825928, "learning_rate": 2.933190408306447e-05, "loss": 0.0598, "step": 1783 }, { "epoch": 0.12259483232545354, "grad_norm": 2.597165822982788, "learning_rate": 2.933091845067271e-05, "loss": 0.0606, "step": 1784 }, { "epoch": 0.12266355140186916, "grad_norm": 1.2908412218093872, "learning_rate": 2.9329932108356706e-05, "loss": 0.0443, "step": 1785 }, { "epoch": 0.12273227047828478, "grad_norm": 1.401051640510559, "learning_rate": 2.932894505616533e-05, "loss": 0.1017, "step": 1786 }, { "epoch": 0.12280098955470038, "grad_norm": 2.655468702316284, "learning_rate": 2.932795729414748e-05, "loss": 0.0672, "step": 1787 }, { "epoch": 0.122869708631116, "grad_norm": 1.8700593709945679, "learning_rate": 2.932696882235208e-05, "loss": 0.0807, "step": 1788 }, { "epoch": 0.1229384277075316, "grad_norm": 1.8961081504821777, "learning_rate": 2.9325979640828102e-05, "loss": 0.0812, "step": 1789 }, { "epoch": 0.12300714678394722, "grad_norm": 1.6558117866516113, "learning_rate": 2.932498974962454e-05, "loss": 0.0676, "step": 1790 }, { "epoch": 0.12307586586036284, "grad_norm": 1.6926639080047607, "learning_rate": 2.9323999148790437e-05, "loss": 0.071, "step": 1791 }, { "epoch": 0.12314458493677845, "grad_norm": 1.3137173652648926, "learning_rate": 2.9323007838374867e-05, "loss": 0.0706, "step": 1792 }, { "epoch": 0.12321330401319407, "grad_norm": 2.1826252937316895, "learning_rate": 2.9322015818426936e-05, "loss": 0.0662, "step": 1793 }, { "epoch": 0.12328202308960967, "grad_norm": 2.585253953933716, "learning_rate": 2.9321023088995786e-05, "loss": 0.0593, "step": 1794 }, { "epoch": 0.12335074216602529, "grad_norm": 2.489375352859497, "learning_rate": 2.9320029650130594e-05, "loss": 0.0744, "step": 1795 }, { "epoch": 0.1234194612424409, "grad_norm": 0.9218869209289551, "learning_rate": 2.9319035501880578e-05, "loss": 0.0551, "step": 1796 }, { "epoch": 0.12348818031885651, "grad_norm": 1.7640881538391113, "learning_rate": 2.9318040644294977e-05, "loss": 0.0796, "step": 1797 }, { "epoch": 0.12355689939527213, "grad_norm": 2.0926904678344727, "learning_rate": 2.931704507742308e-05, "loss": 0.078, "step": 1798 }, { "epoch": 0.12362561847168774, "grad_norm": 1.2110527753829956, "learning_rate": 2.9316048801314205e-05, "loss": 0.0562, "step": 1799 }, { "epoch": 0.12369433754810336, "grad_norm": 1.490302324295044, "learning_rate": 2.9315051816017707e-05, "loss": 0.0414, "step": 1800 }, { "epoch": 0.12376305662451896, "grad_norm": 1.3150806427001953, "learning_rate": 2.931405412158297e-05, "loss": 0.0516, "step": 1801 }, { "epoch": 0.12383177570093458, "grad_norm": 1.6475039720535278, "learning_rate": 2.931305571805942e-05, "loss": 0.0641, "step": 1802 }, { "epoch": 0.1239004947773502, "grad_norm": 1.3929314613342285, "learning_rate": 2.9312056605496515e-05, "loss": 0.0591, "step": 1803 }, { "epoch": 0.1239692138537658, "grad_norm": 3.1103432178497314, "learning_rate": 2.931105678394375e-05, "loss": 0.0663, "step": 1804 }, { "epoch": 0.12403793293018142, "grad_norm": 2.133100986480713, "learning_rate": 2.931005625345066e-05, "loss": 0.0642, "step": 1805 }, { "epoch": 0.12410665200659703, "grad_norm": 1.4177407026290894, "learning_rate": 2.93090550140668e-05, "loss": 0.0582, "step": 1806 }, { "epoch": 0.12417537108301265, "grad_norm": 1.5272419452667236, "learning_rate": 2.930805306584177e-05, "loss": 0.0727, "step": 1807 }, { "epoch": 0.12424409015942826, "grad_norm": 1.264459252357483, "learning_rate": 2.930705040882521e-05, "loss": 0.0809, "step": 1808 }, { "epoch": 0.12431280923584387, "grad_norm": 1.455634355545044, "learning_rate": 2.9306047043066786e-05, "loss": 0.0531, "step": 1809 }, { "epoch": 0.12438152831225949, "grad_norm": 1.4994449615478516, "learning_rate": 2.9305042968616202e-05, "loss": 0.0999, "step": 1810 }, { "epoch": 0.12445024738867509, "grad_norm": 1.376542091369629, "learning_rate": 2.9304038185523204e-05, "loss": 0.0692, "step": 1811 }, { "epoch": 0.12451896646509071, "grad_norm": 1.4521217346191406, "learning_rate": 2.930303269383756e-05, "loss": 0.074, "step": 1812 }, { "epoch": 0.12458768554150632, "grad_norm": 2.4114160537719727, "learning_rate": 2.930202649360908e-05, "loss": 0.0845, "step": 1813 }, { "epoch": 0.12465640461792193, "grad_norm": 1.2667818069458008, "learning_rate": 2.9301019584887612e-05, "loss": 0.0533, "step": 1814 }, { "epoch": 0.12472512369433755, "grad_norm": 1.7668564319610596, "learning_rate": 2.9300011967723037e-05, "loss": 0.0662, "step": 1815 }, { "epoch": 0.12479384277075316, "grad_norm": 2.6110544204711914, "learning_rate": 2.9299003642165266e-05, "loss": 0.0949, "step": 1816 }, { "epoch": 0.12486256184716878, "grad_norm": 2.2134759426116943, "learning_rate": 2.9297994608264256e-05, "loss": 0.063, "step": 1817 }, { "epoch": 0.12493128092358438, "grad_norm": 1.5806611776351929, "learning_rate": 2.9296984866069986e-05, "loss": 0.0683, "step": 1818 }, { "epoch": 0.125, "grad_norm": 1.6969441175460815, "learning_rate": 2.929597441563248e-05, "loss": 0.0778, "step": 1819 }, { "epoch": 0.12506871907641562, "grad_norm": 2.0560033321380615, "learning_rate": 2.9294963257001796e-05, "loss": 0.0638, "step": 1820 }, { "epoch": 0.12513743815283124, "grad_norm": 2.0285096168518066, "learning_rate": 2.9293951390228015e-05, "loss": 0.0733, "step": 1821 }, { "epoch": 0.12520615722924683, "grad_norm": 2.411750555038452, "learning_rate": 2.929293881536127e-05, "loss": 0.053, "step": 1822 }, { "epoch": 0.12527487630566245, "grad_norm": 1.4552677869796753, "learning_rate": 2.9291925532451726e-05, "loss": 0.061, "step": 1823 }, { "epoch": 0.12534359538207807, "grad_norm": 1.8295090198516846, "learning_rate": 2.9290911541549567e-05, "loss": 0.0631, "step": 1824 }, { "epoch": 0.12541231445849368, "grad_norm": 1.7499110698699951, "learning_rate": 2.928989684270504e-05, "loss": 0.0816, "step": 1825 }, { "epoch": 0.1254810335349093, "grad_norm": 2.072606325149536, "learning_rate": 2.9288881435968395e-05, "loss": 0.0493, "step": 1826 }, { "epoch": 0.1255497526113249, "grad_norm": 4.930244445800781, "learning_rate": 2.928786532138994e-05, "loss": 0.0657, "step": 1827 }, { "epoch": 0.1256184716877405, "grad_norm": 1.884161353111267, "learning_rate": 2.928684849902002e-05, "loss": 0.0637, "step": 1828 }, { "epoch": 0.12568719076415613, "grad_norm": 2.7641067504882812, "learning_rate": 2.928583096890899e-05, "loss": 0.079, "step": 1829 }, { "epoch": 0.12575590984057175, "grad_norm": 1.248298168182373, "learning_rate": 2.9284812731107267e-05, "loss": 0.0735, "step": 1830 }, { "epoch": 0.12582462891698737, "grad_norm": 2.5831971168518066, "learning_rate": 2.9283793785665288e-05, "loss": 0.06, "step": 1831 }, { "epoch": 0.12589334799340296, "grad_norm": 1.5468297004699707, "learning_rate": 2.928277413263353e-05, "loss": 0.0536, "step": 1832 }, { "epoch": 0.12596206706981858, "grad_norm": 2.181718111038208, "learning_rate": 2.9281753772062508e-05, "loss": 0.0655, "step": 1833 }, { "epoch": 0.1260307861462342, "grad_norm": 2.3720765113830566, "learning_rate": 2.9280732704002767e-05, "loss": 0.077, "step": 1834 }, { "epoch": 0.12609950522264982, "grad_norm": 2.131354570388794, "learning_rate": 2.9279710928504887e-05, "loss": 0.0599, "step": 1835 }, { "epoch": 0.1261682242990654, "grad_norm": 4.70466947555542, "learning_rate": 2.9278688445619485e-05, "loss": 0.0916, "step": 1836 }, { "epoch": 0.12623694337548103, "grad_norm": 1.4720964431762695, "learning_rate": 2.9277665255397215e-05, "loss": 0.0643, "step": 1837 }, { "epoch": 0.12630566245189664, "grad_norm": 1.9489505290985107, "learning_rate": 2.927664135788876e-05, "loss": 0.0758, "step": 1838 }, { "epoch": 0.12637438152831226, "grad_norm": 2.0674028396606445, "learning_rate": 2.9275616753144842e-05, "loss": 0.0742, "step": 1839 }, { "epoch": 0.12644310060472788, "grad_norm": 4.011932373046875, "learning_rate": 2.9274591441216224e-05, "loss": 0.062, "step": 1840 }, { "epoch": 0.12651181968114347, "grad_norm": 1.5030924081802368, "learning_rate": 2.9273565422153687e-05, "loss": 0.0687, "step": 1841 }, { "epoch": 0.1265805387575591, "grad_norm": 1.1542128324508667, "learning_rate": 2.927253869600807e-05, "loss": 0.0799, "step": 1842 }, { "epoch": 0.1266492578339747, "grad_norm": 2.0767509937286377, "learning_rate": 2.9271511262830227e-05, "loss": 0.0877, "step": 1843 }, { "epoch": 0.12671797691039033, "grad_norm": 1.6130280494689941, "learning_rate": 2.9270483122671056e-05, "loss": 0.0631, "step": 1844 }, { "epoch": 0.12678669598680595, "grad_norm": 6.487364768981934, "learning_rate": 2.9269454275581494e-05, "loss": 0.0951, "step": 1845 }, { "epoch": 0.12685541506322154, "grad_norm": 1.825284481048584, "learning_rate": 2.92684247216125e-05, "loss": 0.0549, "step": 1846 }, { "epoch": 0.12692413413963716, "grad_norm": 1.726995825767517, "learning_rate": 2.926739446081508e-05, "loss": 0.0859, "step": 1847 }, { "epoch": 0.12699285321605278, "grad_norm": 0.8405787944793701, "learning_rate": 2.926636349324027e-05, "loss": 0.0443, "step": 1848 }, { "epoch": 0.1270615722924684, "grad_norm": 1.1738733053207397, "learning_rate": 2.926533181893915e-05, "loss": 0.0594, "step": 1849 }, { "epoch": 0.127130291368884, "grad_norm": 1.2963387966156006, "learning_rate": 2.926429943796281e-05, "loss": 0.0584, "step": 1850 }, { "epoch": 0.1271990104452996, "grad_norm": 2.520582437515259, "learning_rate": 2.9263266350362407e-05, "loss": 0.0709, "step": 1851 }, { "epoch": 0.12726772952171522, "grad_norm": 1.436911702156067, "learning_rate": 2.9262232556189114e-05, "loss": 0.056, "step": 1852 }, { "epoch": 0.12733644859813084, "grad_norm": 3.1352286338806152, "learning_rate": 2.926119805549414e-05, "loss": 0.0645, "step": 1853 }, { "epoch": 0.12740516767454646, "grad_norm": 2.2824668884277344, "learning_rate": 2.9260162848328736e-05, "loss": 0.0867, "step": 1854 }, { "epoch": 0.12747388675096208, "grad_norm": 1.9401973485946655, "learning_rate": 2.9259126934744178e-05, "loss": 0.0739, "step": 1855 }, { "epoch": 0.12754260582737767, "grad_norm": 1.998288631439209, "learning_rate": 2.925809031479179e-05, "loss": 0.0739, "step": 1856 }, { "epoch": 0.1276113249037933, "grad_norm": 3.691746950149536, "learning_rate": 2.9257052988522923e-05, "loss": 0.0846, "step": 1857 }, { "epoch": 0.1276800439802089, "grad_norm": 2.0635437965393066, "learning_rate": 2.925601495598896e-05, "loss": 0.0658, "step": 1858 }, { "epoch": 0.12774876305662453, "grad_norm": 1.9099699258804321, "learning_rate": 2.9254976217241326e-05, "loss": 0.0782, "step": 1859 }, { "epoch": 0.12781748213304014, "grad_norm": 1.5423519611358643, "learning_rate": 2.9253936772331475e-05, "loss": 0.0566, "step": 1860 }, { "epoch": 0.12788620120945574, "grad_norm": 5.2345051765441895, "learning_rate": 2.9252896621310906e-05, "loss": 0.0727, "step": 1861 }, { "epoch": 0.12795492028587135, "grad_norm": 1.8715797662734985, "learning_rate": 2.9251855764231137e-05, "loss": 0.0636, "step": 1862 }, { "epoch": 0.12802363936228697, "grad_norm": 1.6430952548980713, "learning_rate": 2.9250814201143738e-05, "loss": 0.0922, "step": 1863 }, { "epoch": 0.1280923584387026, "grad_norm": 1.7056396007537842, "learning_rate": 2.9249771932100296e-05, "loss": 0.0629, "step": 1864 }, { "epoch": 0.1281610775151182, "grad_norm": 1.8125470876693726, "learning_rate": 2.924872895715245e-05, "loss": 0.0587, "step": 1865 }, { "epoch": 0.1282297965915338, "grad_norm": 2.2147350311279297, "learning_rate": 2.9247685276351872e-05, "loss": 0.0572, "step": 1866 }, { "epoch": 0.12829851566794942, "grad_norm": 1.9107158184051514, "learning_rate": 2.9246640889750254e-05, "loss": 0.0671, "step": 1867 }, { "epoch": 0.12836723474436504, "grad_norm": 3.44120454788208, "learning_rate": 2.9245595797399334e-05, "loss": 0.0872, "step": 1868 }, { "epoch": 0.12843595382078066, "grad_norm": 1.9303786754608154, "learning_rate": 2.9244549999350886e-05, "loss": 0.0689, "step": 1869 }, { "epoch": 0.12850467289719625, "grad_norm": 2.0480005741119385, "learning_rate": 2.924350349565672e-05, "loss": 0.0777, "step": 1870 }, { "epoch": 0.12857339197361187, "grad_norm": 1.5733941793441772, "learning_rate": 2.924245628636867e-05, "loss": 0.0828, "step": 1871 }, { "epoch": 0.12864211105002749, "grad_norm": 2.1431896686553955, "learning_rate": 2.9241408371538615e-05, "loss": 0.0731, "step": 1872 }, { "epoch": 0.1287108301264431, "grad_norm": 1.4477369785308838, "learning_rate": 2.9240359751218474e-05, "loss": 0.0737, "step": 1873 }, { "epoch": 0.12877954920285872, "grad_norm": 1.2288566827774048, "learning_rate": 2.9239310425460188e-05, "loss": 0.0565, "step": 1874 }, { "epoch": 0.12884826827927431, "grad_norm": 1.7812261581420898, "learning_rate": 2.9238260394315736e-05, "loss": 0.052, "step": 1875 }, { "epoch": 0.12891698735568993, "grad_norm": 3.093015432357788, "learning_rate": 2.9237209657837137e-05, "loss": 0.0698, "step": 1876 }, { "epoch": 0.12898570643210555, "grad_norm": 1.6457021236419678, "learning_rate": 2.9236158216076443e-05, "loss": 0.0657, "step": 1877 }, { "epoch": 0.12905442550852117, "grad_norm": 1.9258135557174683, "learning_rate": 2.9235106069085736e-05, "loss": 0.0639, "step": 1878 }, { "epoch": 0.1291231445849368, "grad_norm": 3.624006986618042, "learning_rate": 2.9234053216917145e-05, "loss": 0.0774, "step": 1879 }, { "epoch": 0.12919186366135238, "grad_norm": 1.2338483333587646, "learning_rate": 2.9232999659622818e-05, "loss": 0.0621, "step": 1880 }, { "epoch": 0.129260582737768, "grad_norm": 1.6681333780288696, "learning_rate": 2.923194539725495e-05, "loss": 0.0886, "step": 1881 }, { "epoch": 0.12932930181418362, "grad_norm": 1.0574829578399658, "learning_rate": 2.9230890429865766e-05, "loss": 0.0678, "step": 1882 }, { "epoch": 0.12939802089059924, "grad_norm": 1.9942249059677124, "learning_rate": 2.922983475750753e-05, "loss": 0.0624, "step": 1883 }, { "epoch": 0.12946673996701485, "grad_norm": 2.2837767601013184, "learning_rate": 2.9228778380232535e-05, "loss": 0.0757, "step": 1884 }, { "epoch": 0.12953545904343045, "grad_norm": 2.240910053253174, "learning_rate": 2.9227721298093112e-05, "loss": 0.0627, "step": 1885 }, { "epoch": 0.12960417811984606, "grad_norm": 1.6884351968765259, "learning_rate": 2.922666351114163e-05, "loss": 0.0605, "step": 1886 }, { "epoch": 0.12967289719626168, "grad_norm": 2.2084550857543945, "learning_rate": 2.9225605019430482e-05, "loss": 0.0564, "step": 1887 }, { "epoch": 0.1297416162726773, "grad_norm": 1.3352032899856567, "learning_rate": 2.9224545823012107e-05, "loss": 0.0411, "step": 1888 }, { "epoch": 0.12981033534909292, "grad_norm": 2.4381425380706787, "learning_rate": 2.922348592193898e-05, "loss": 0.0621, "step": 1889 }, { "epoch": 0.1298790544255085, "grad_norm": 0.8603202104568481, "learning_rate": 2.92224253162636e-05, "loss": 0.0597, "step": 1890 }, { "epoch": 0.12994777350192413, "grad_norm": 1.3126002550125122, "learning_rate": 2.922136400603851e-05, "loss": 0.1052, "step": 1891 }, { "epoch": 0.13001649257833975, "grad_norm": 1.3522933721542358, "learning_rate": 2.9220301991316287e-05, "loss": 0.0577, "step": 1892 }, { "epoch": 0.13008521165475537, "grad_norm": 1.5929310321807861, "learning_rate": 2.9219239272149538e-05, "loss": 0.0527, "step": 1893 }, { "epoch": 0.13015393073117099, "grad_norm": 2.066589117050171, "learning_rate": 2.9218175848590908e-05, "loss": 0.055, "step": 1894 }, { "epoch": 0.13022264980758658, "grad_norm": 2.0600733757019043, "learning_rate": 2.9217111720693076e-05, "loss": 0.0481, "step": 1895 }, { "epoch": 0.1302913688840022, "grad_norm": 1.6331031322479248, "learning_rate": 2.921604688850876e-05, "loss": 0.0763, "step": 1896 }, { "epoch": 0.13036008796041781, "grad_norm": 2.5165722370147705, "learning_rate": 2.9214981352090714e-05, "loss": 0.0602, "step": 1897 }, { "epoch": 0.13042880703683343, "grad_norm": 1.2769652605056763, "learning_rate": 2.921391511149171e-05, "loss": 0.0575, "step": 1898 }, { "epoch": 0.13049752611324905, "grad_norm": 2.6129539012908936, "learning_rate": 2.9212848166764576e-05, "loss": 0.0657, "step": 1899 }, { "epoch": 0.13056624518966464, "grad_norm": 1.3621997833251953, "learning_rate": 2.9211780517962164e-05, "loss": 0.0654, "step": 1900 }, { "epoch": 0.13063496426608026, "grad_norm": 1.2258408069610596, "learning_rate": 2.9210712165137365e-05, "loss": 0.0562, "step": 1901 }, { "epoch": 0.13070368334249588, "grad_norm": 1.5380048751831055, "learning_rate": 2.92096431083431e-05, "loss": 0.0669, "step": 1902 }, { "epoch": 0.1307724024189115, "grad_norm": 1.6045165061950684, "learning_rate": 2.9208573347632326e-05, "loss": 0.0725, "step": 1903 }, { "epoch": 0.1308411214953271, "grad_norm": 3.41879940032959, "learning_rate": 2.9207502883058044e-05, "loss": 0.0654, "step": 1904 }, { "epoch": 0.1309098405717427, "grad_norm": 2.013481378555298, "learning_rate": 2.920643171467328e-05, "loss": 0.0746, "step": 1905 }, { "epoch": 0.13097855964815833, "grad_norm": 2.245814561843872, "learning_rate": 2.9205359842531094e-05, "loss": 0.057, "step": 1906 }, { "epoch": 0.13104727872457395, "grad_norm": 1.3856902122497559, "learning_rate": 2.920428726668458e-05, "loss": 0.0646, "step": 1907 }, { "epoch": 0.13111599780098956, "grad_norm": 1.652858018875122, "learning_rate": 2.9203213987186885e-05, "loss": 0.0625, "step": 1908 }, { "epoch": 0.13118471687740516, "grad_norm": 1.7600111961364746, "learning_rate": 2.9202140004091168e-05, "loss": 0.0508, "step": 1909 }, { "epoch": 0.13125343595382077, "grad_norm": 1.8747128248214722, "learning_rate": 2.9201065317450637e-05, "loss": 0.0728, "step": 1910 }, { "epoch": 0.1313221550302364, "grad_norm": 1.688688039779663, "learning_rate": 2.919998992731852e-05, "loss": 0.0719, "step": 1911 }, { "epoch": 0.131390874106652, "grad_norm": 2.7806665897369385, "learning_rate": 2.91989138337481e-05, "loss": 0.0723, "step": 1912 }, { "epoch": 0.13145959318306763, "grad_norm": 1.5200575590133667, "learning_rate": 2.919783703679268e-05, "loss": 0.0749, "step": 1913 }, { "epoch": 0.13152831225948322, "grad_norm": 8.793168067932129, "learning_rate": 2.9196759536505603e-05, "loss": 0.0718, "step": 1914 }, { "epoch": 0.13159703133589884, "grad_norm": 1.862979769706726, "learning_rate": 2.9195681332940242e-05, "loss": 0.0648, "step": 1915 }, { "epoch": 0.13166575041231446, "grad_norm": 3.3506104946136475, "learning_rate": 2.9194602426150016e-05, "loss": 0.0646, "step": 1916 }, { "epoch": 0.13173446948873008, "grad_norm": 1.4785971641540527, "learning_rate": 2.9193522816188372e-05, "loss": 0.0554, "step": 1917 }, { "epoch": 0.1318031885651457, "grad_norm": 1.065229892730713, "learning_rate": 2.9192442503108785e-05, "loss": 0.0648, "step": 1918 }, { "epoch": 0.1318719076415613, "grad_norm": 1.8109219074249268, "learning_rate": 2.919136148696477e-05, "loss": 0.0544, "step": 1919 }, { "epoch": 0.1319406267179769, "grad_norm": 1.1958613395690918, "learning_rate": 2.9190279767809888e-05, "loss": 0.0821, "step": 1920 }, { "epoch": 0.13200934579439252, "grad_norm": 1.5297963619232178, "learning_rate": 2.918919734569772e-05, "loss": 0.0673, "step": 1921 }, { "epoch": 0.13207806487080814, "grad_norm": 1.3275593519210815, "learning_rate": 2.918811422068189e-05, "loss": 0.0494, "step": 1922 }, { "epoch": 0.13214678394722376, "grad_norm": 1.6192604303359985, "learning_rate": 2.9187030392816048e-05, "loss": 0.0692, "step": 1923 }, { "epoch": 0.13221550302363935, "grad_norm": 1.517513394355774, "learning_rate": 2.918594586215389e-05, "loss": 0.0669, "step": 1924 }, { "epoch": 0.13228422210005497, "grad_norm": 1.6822795867919922, "learning_rate": 2.9184860628749138e-05, "loss": 0.0623, "step": 1925 }, { "epoch": 0.1323529411764706, "grad_norm": 1.9141814708709717, "learning_rate": 2.918377469265555e-05, "loss": 0.0773, "step": 1926 }, { "epoch": 0.1324216602528862, "grad_norm": 2.091470956802368, "learning_rate": 2.9182688053926923e-05, "loss": 0.0936, "step": 1927 }, { "epoch": 0.13249037932930183, "grad_norm": 3.8118631839752197, "learning_rate": 2.9181600712617094e-05, "loss": 0.0686, "step": 1928 }, { "epoch": 0.13255909840571742, "grad_norm": 1.0499950647354126, "learning_rate": 2.918051266877992e-05, "loss": 0.0892, "step": 1929 }, { "epoch": 0.13262781748213304, "grad_norm": 2.4570202827453613, "learning_rate": 2.9179423922469302e-05, "loss": 0.0582, "step": 1930 }, { "epoch": 0.13269653655854866, "grad_norm": 2.2109055519104004, "learning_rate": 2.9178334473739176e-05, "loss": 0.0951, "step": 1931 }, { "epoch": 0.13276525563496427, "grad_norm": 3.6060969829559326, "learning_rate": 2.917724432264351e-05, "loss": 0.0662, "step": 1932 }, { "epoch": 0.13283397471137987, "grad_norm": 1.4982795715332031, "learning_rate": 2.9176153469236305e-05, "loss": 0.0622, "step": 1933 }, { "epoch": 0.13290269378779548, "grad_norm": 1.9031168222427368, "learning_rate": 2.91750619135716e-05, "loss": 0.067, "step": 1934 }, { "epoch": 0.1329714128642111, "grad_norm": 1.9407339096069336, "learning_rate": 2.9173969655703472e-05, "loss": 0.0583, "step": 1935 }, { "epoch": 0.13304013194062672, "grad_norm": 2.0777220726013184, "learning_rate": 2.917287669568603e-05, "loss": 0.0754, "step": 1936 }, { "epoch": 0.13310885101704234, "grad_norm": 1.8301743268966675, "learning_rate": 2.9171783033573412e-05, "loss": 0.0555, "step": 1937 }, { "epoch": 0.13317757009345793, "grad_norm": 0.9613759517669678, "learning_rate": 2.9170688669419803e-05, "loss": 0.0549, "step": 1938 }, { "epoch": 0.13324628916987355, "grad_norm": 2.0941953659057617, "learning_rate": 2.9169593603279407e-05, "loss": 0.0731, "step": 1939 }, { "epoch": 0.13331500824628917, "grad_norm": 1.7937880754470825, "learning_rate": 2.9168497835206476e-05, "loss": 0.0655, "step": 1940 }, { "epoch": 0.1333837273227048, "grad_norm": 1.3598791360855103, "learning_rate": 2.9167401365255292e-05, "loss": 0.0699, "step": 1941 }, { "epoch": 0.1334524463991204, "grad_norm": 2.4869332313537598, "learning_rate": 2.916630419348017e-05, "loss": 0.069, "step": 1942 }, { "epoch": 0.133521165475536, "grad_norm": 1.3812683820724487, "learning_rate": 2.9165206319935464e-05, "loss": 0.0551, "step": 1943 }, { "epoch": 0.13358988455195162, "grad_norm": 2.1234052181243896, "learning_rate": 2.916410774467556e-05, "loss": 0.0789, "step": 1944 }, { "epoch": 0.13365860362836723, "grad_norm": 1.3301620483398438, "learning_rate": 2.9163008467754877e-05, "loss": 0.0888, "step": 1945 }, { "epoch": 0.13372732270478285, "grad_norm": 1.525384783744812, "learning_rate": 2.9161908489227875e-05, "loss": 0.0618, "step": 1946 }, { "epoch": 0.13379604178119847, "grad_norm": 1.6584681272506714, "learning_rate": 2.916080780914904e-05, "loss": 0.0569, "step": 1947 }, { "epoch": 0.13386476085761406, "grad_norm": 1.4555351734161377, "learning_rate": 2.91597064275729e-05, "loss": 0.0471, "step": 1948 }, { "epoch": 0.13393347993402968, "grad_norm": 1.748567819595337, "learning_rate": 2.9158604344554012e-05, "loss": 0.0618, "step": 1949 }, { "epoch": 0.1340021990104453, "grad_norm": 1.899604320526123, "learning_rate": 2.9157501560146977e-05, "loss": 0.05, "step": 1950 }, { "epoch": 0.13407091808686092, "grad_norm": 3.283689260482788, "learning_rate": 2.915639807440642e-05, "loss": 0.0586, "step": 1951 }, { "epoch": 0.13413963716327654, "grad_norm": 1.6246956586837769, "learning_rate": 2.9155293887387006e-05, "loss": 0.0677, "step": 1952 }, { "epoch": 0.13420835623969213, "grad_norm": 2.0736193656921387, "learning_rate": 2.9154188999143436e-05, "loss": 0.072, "step": 1953 }, { "epoch": 0.13427707531610775, "grad_norm": 1.520501732826233, "learning_rate": 2.9153083409730444e-05, "loss": 0.0717, "step": 1954 }, { "epoch": 0.13434579439252337, "grad_norm": 1.5765622854232788, "learning_rate": 2.9151977119202796e-05, "loss": 0.0641, "step": 1955 }, { "epoch": 0.13441451346893898, "grad_norm": 2.635927200317383, "learning_rate": 2.91508701276153e-05, "loss": 0.0605, "step": 1956 }, { "epoch": 0.1344832325453546, "grad_norm": 2.9672820568084717, "learning_rate": 2.9149762435022785e-05, "loss": 0.0602, "step": 1957 }, { "epoch": 0.1345519516217702, "grad_norm": 1.6146481037139893, "learning_rate": 2.9148654041480134e-05, "loss": 0.0595, "step": 1958 }, { "epoch": 0.1346206706981858, "grad_norm": 1.5583661794662476, "learning_rate": 2.9147544947042245e-05, "loss": 0.0672, "step": 1959 }, { "epoch": 0.13468938977460143, "grad_norm": 1.853355884552002, "learning_rate": 2.914643515176407e-05, "loss": 0.0523, "step": 1960 }, { "epoch": 0.13475810885101705, "grad_norm": 1.7025078535079956, "learning_rate": 2.914532465570058e-05, "loss": 0.0583, "step": 1961 }, { "epoch": 0.13482682792743267, "grad_norm": 1.0929632186889648, "learning_rate": 2.914421345890679e-05, "loss": 0.0565, "step": 1962 }, { "epoch": 0.13489554700384826, "grad_norm": 1.2185420989990234, "learning_rate": 2.9143101561437742e-05, "loss": 0.0644, "step": 1963 }, { "epoch": 0.13496426608026388, "grad_norm": 1.6118162870407104, "learning_rate": 2.914198896334852e-05, "loss": 0.0765, "step": 1964 }, { "epoch": 0.1350329851566795, "grad_norm": 2.2040610313415527, "learning_rate": 2.9140875664694244e-05, "loss": 0.0573, "step": 1965 }, { "epoch": 0.13510170423309512, "grad_norm": 1.2751544713974, "learning_rate": 2.9139761665530055e-05, "loss": 0.042, "step": 1966 }, { "epoch": 0.1351704233095107, "grad_norm": 1.4677236080169678, "learning_rate": 2.9138646965911147e-05, "loss": 0.0546, "step": 1967 }, { "epoch": 0.13523914238592633, "grad_norm": 1.79599130153656, "learning_rate": 2.9137531565892732e-05, "loss": 0.0512, "step": 1968 }, { "epoch": 0.13530786146234194, "grad_norm": 1.527633547782898, "learning_rate": 2.913641546553007e-05, "loss": 0.0651, "step": 1969 }, { "epoch": 0.13537658053875756, "grad_norm": 2.1089305877685547, "learning_rate": 2.9135298664878447e-05, "loss": 0.0564, "step": 1970 }, { "epoch": 0.13544529961517318, "grad_norm": 1.1485321521759033, "learning_rate": 2.9134181163993193e-05, "loss": 0.0519, "step": 1971 }, { "epoch": 0.13551401869158877, "grad_norm": 1.2512222528457642, "learning_rate": 2.913306296292966e-05, "loss": 0.0537, "step": 1972 }, { "epoch": 0.1355827377680044, "grad_norm": 2.6210525035858154, "learning_rate": 2.9131944061743245e-05, "loss": 0.0674, "step": 1973 }, { "epoch": 0.13565145684442, "grad_norm": 2.0012729167938232, "learning_rate": 2.9130824460489375e-05, "loss": 0.0924, "step": 1974 }, { "epoch": 0.13572017592083563, "grad_norm": 2.2531514167785645, "learning_rate": 2.9129704159223513e-05, "loss": 0.0582, "step": 1975 }, { "epoch": 0.13578889499725125, "grad_norm": 1.843565583229065, "learning_rate": 2.912858315800116e-05, "loss": 0.0681, "step": 1976 }, { "epoch": 0.13585761407366684, "grad_norm": 1.1830995082855225, "learning_rate": 2.912746145687784e-05, "loss": 0.0579, "step": 1977 }, { "epoch": 0.13592633315008246, "grad_norm": 1.5701900720596313, "learning_rate": 2.9126339055909127e-05, "loss": 0.0489, "step": 1978 }, { "epoch": 0.13599505222649808, "grad_norm": 1.5623584985733032, "learning_rate": 2.912521595515062e-05, "loss": 0.0529, "step": 1979 }, { "epoch": 0.1360637713029137, "grad_norm": 1.5689940452575684, "learning_rate": 2.9124092154657953e-05, "loss": 0.0487, "step": 1980 }, { "epoch": 0.1361324903793293, "grad_norm": 1.7665215730667114, "learning_rate": 2.9122967654486795e-05, "loss": 0.0611, "step": 1981 }, { "epoch": 0.1362012094557449, "grad_norm": 2.0930120944976807, "learning_rate": 2.912184245469286e-05, "loss": 0.0698, "step": 1982 }, { "epoch": 0.13626992853216052, "grad_norm": 1.4963101148605347, "learning_rate": 2.912071655533188e-05, "loss": 0.0572, "step": 1983 }, { "epoch": 0.13633864760857614, "grad_norm": 1.2827008962631226, "learning_rate": 2.9119589956459634e-05, "loss": 0.0574, "step": 1984 }, { "epoch": 0.13640736668499176, "grad_norm": 1.7179368734359741, "learning_rate": 2.9118462658131934e-05, "loss": 0.0609, "step": 1985 }, { "epoch": 0.13647608576140738, "grad_norm": 1.423271894454956, "learning_rate": 2.911733466040462e-05, "loss": 0.0618, "step": 1986 }, { "epoch": 0.13654480483782297, "grad_norm": 1.7968275547027588, "learning_rate": 2.9116205963333567e-05, "loss": 0.0678, "step": 1987 }, { "epoch": 0.1366135239142386, "grad_norm": 1.2006639242172241, "learning_rate": 2.9115076566974693e-05, "loss": 0.0598, "step": 1988 }, { "epoch": 0.1366822429906542, "grad_norm": 1.3321417570114136, "learning_rate": 2.9113946471383947e-05, "loss": 0.0587, "step": 1989 }, { "epoch": 0.13675096206706983, "grad_norm": 1.193174958229065, "learning_rate": 2.911281567661731e-05, "loss": 0.0667, "step": 1990 }, { "epoch": 0.13681968114348544, "grad_norm": 1.0324628353118896, "learning_rate": 2.911168418273079e-05, "loss": 0.0605, "step": 1991 }, { "epoch": 0.13688840021990104, "grad_norm": 1.3145960569381714, "learning_rate": 2.9110551989780457e-05, "loss": 0.0633, "step": 1992 }, { "epoch": 0.13695711929631665, "grad_norm": 1.3727177381515503, "learning_rate": 2.910941909782239e-05, "loss": 0.0579, "step": 1993 }, { "epoch": 0.13702583837273227, "grad_norm": 1.2482707500457764, "learning_rate": 2.9108285506912704e-05, "loss": 0.0413, "step": 1994 }, { "epoch": 0.1370945574491479, "grad_norm": 1.494188666343689, "learning_rate": 2.9107151217107562e-05, "loss": 0.0511, "step": 1995 }, { "epoch": 0.1371632765255635, "grad_norm": 1.328032374382019, "learning_rate": 2.910601622846315e-05, "loss": 0.0755, "step": 1996 }, { "epoch": 0.1372319956019791, "grad_norm": 2.3247134685516357, "learning_rate": 2.9104880541035697e-05, "loss": 0.0707, "step": 1997 }, { "epoch": 0.13730071467839472, "grad_norm": 1.472182035446167, "learning_rate": 2.9103744154881456e-05, "loss": 0.077, "step": 1998 }, { "epoch": 0.13736943375481034, "grad_norm": 0.8720581531524658, "learning_rate": 2.910260707005673e-05, "loss": 0.036, "step": 1999 }, { "epoch": 0.13743815283122596, "grad_norm": 1.9385714530944824, "learning_rate": 2.910146928661784e-05, "loss": 0.0564, "step": 2000 }, { "epoch": 0.13750687190764155, "grad_norm": 1.243958830833435, "learning_rate": 2.9100330804621156e-05, "loss": 0.0517, "step": 2001 }, { "epoch": 0.13757559098405717, "grad_norm": 1.420423150062561, "learning_rate": 2.9099191624123074e-05, "loss": 0.0704, "step": 2002 }, { "epoch": 0.13764431006047279, "grad_norm": 1.1290013790130615, "learning_rate": 2.909805174518002e-05, "loss": 0.0566, "step": 2003 }, { "epoch": 0.1377130291368884, "grad_norm": 1.2906148433685303, "learning_rate": 2.909691116784847e-05, "loss": 0.0461, "step": 2004 }, { "epoch": 0.13778174821330402, "grad_norm": 0.9237934947013855, "learning_rate": 2.9095769892184925e-05, "loss": 0.053, "step": 2005 }, { "epoch": 0.1378504672897196, "grad_norm": 1.7221033573150635, "learning_rate": 2.909462791824592e-05, "loss": 0.0561, "step": 2006 }, { "epoch": 0.13791918636613523, "grad_norm": 1.7181503772735596, "learning_rate": 2.9093485246088022e-05, "loss": 0.0714, "step": 2007 }, { "epoch": 0.13798790544255085, "grad_norm": 1.614446997642517, "learning_rate": 2.9092341875767843e-05, "loss": 0.0584, "step": 2008 }, { "epoch": 0.13805662451896647, "grad_norm": 1.182671070098877, "learning_rate": 2.9091197807342015e-05, "loss": 0.0493, "step": 2009 }, { "epoch": 0.1381253435953821, "grad_norm": 1.7440829277038574, "learning_rate": 2.9090053040867218e-05, "loss": 0.057, "step": 2010 }, { "epoch": 0.13819406267179768, "grad_norm": 3.429442882537842, "learning_rate": 2.908890757640017e-05, "loss": 0.0844, "step": 2011 }, { "epoch": 0.1382627817482133, "grad_norm": 1.337247610092163, "learning_rate": 2.9087761413997598e-05, "loss": 0.0599, "step": 2012 }, { "epoch": 0.13833150082462892, "grad_norm": 1.463154673576355, "learning_rate": 2.9086614553716292e-05, "loss": 0.0538, "step": 2013 }, { "epoch": 0.13840021990104454, "grad_norm": 1.8224457502365112, "learning_rate": 2.9085466995613058e-05, "loss": 0.0608, "step": 2014 }, { "epoch": 0.13846893897746015, "grad_norm": 2.7929604053497314, "learning_rate": 2.9084318739744754e-05, "loss": 0.0814, "step": 2015 }, { "epoch": 0.13853765805387575, "grad_norm": 2.3854784965515137, "learning_rate": 2.9083169786168248e-05, "loss": 0.0665, "step": 2016 }, { "epoch": 0.13860637713029136, "grad_norm": 2.616438865661621, "learning_rate": 2.908202013494047e-05, "loss": 0.1021, "step": 2017 }, { "epoch": 0.13867509620670698, "grad_norm": 2.384176015853882, "learning_rate": 2.9080869786118366e-05, "loss": 0.0712, "step": 2018 }, { "epoch": 0.1387438152831226, "grad_norm": 1.692712426185608, "learning_rate": 2.907971873975892e-05, "loss": 0.0736, "step": 2019 }, { "epoch": 0.13881253435953822, "grad_norm": 1.1922433376312256, "learning_rate": 2.907856699591915e-05, "loss": 0.0846, "step": 2020 }, { "epoch": 0.1388812534359538, "grad_norm": 1.4192076921463013, "learning_rate": 2.9077414554656126e-05, "loss": 0.0542, "step": 2021 }, { "epoch": 0.13894997251236943, "grad_norm": 2.2813000679016113, "learning_rate": 2.9076261416026918e-05, "loss": 0.0704, "step": 2022 }, { "epoch": 0.13901869158878505, "grad_norm": 1.4107557535171509, "learning_rate": 2.907510758008866e-05, "loss": 0.0606, "step": 2023 }, { "epoch": 0.13908741066520067, "grad_norm": 2.575249671936035, "learning_rate": 2.9073953046898515e-05, "loss": 0.0651, "step": 2024 }, { "epoch": 0.13915612974161629, "grad_norm": 1.6100711822509766, "learning_rate": 2.9072797816513668e-05, "loss": 0.0499, "step": 2025 }, { "epoch": 0.13922484881803188, "grad_norm": 1.017929196357727, "learning_rate": 2.9071641888991348e-05, "loss": 0.0562, "step": 2026 }, { "epoch": 0.1392935678944475, "grad_norm": 1.7976194620132446, "learning_rate": 2.9070485264388823e-05, "loss": 0.0579, "step": 2027 }, { "epoch": 0.13936228697086311, "grad_norm": 2.142461061477661, "learning_rate": 2.906932794276338e-05, "loss": 0.0644, "step": 2028 }, { "epoch": 0.13943100604727873, "grad_norm": 1.730468511581421, "learning_rate": 2.9068169924172355e-05, "loss": 0.0645, "step": 2029 }, { "epoch": 0.13949972512369435, "grad_norm": 1.5758007764816284, "learning_rate": 2.906701120867312e-05, "loss": 0.0677, "step": 2030 }, { "epoch": 0.13956844420010994, "grad_norm": 1.468148946762085, "learning_rate": 2.9065851796323072e-05, "loss": 0.0379, "step": 2031 }, { "epoch": 0.13963716327652556, "grad_norm": 2.816110134124756, "learning_rate": 2.906469168717964e-05, "loss": 0.0855, "step": 2032 }, { "epoch": 0.13970588235294118, "grad_norm": 1.4914448261260986, "learning_rate": 2.9063530881300298e-05, "loss": 0.0509, "step": 2033 }, { "epoch": 0.1397746014293568, "grad_norm": 1.9615358114242554, "learning_rate": 2.906236937874255e-05, "loss": 0.0772, "step": 2034 }, { "epoch": 0.1398433205057724, "grad_norm": 0.9785048365592957, "learning_rate": 2.9061207179563935e-05, "loss": 0.0489, "step": 2035 }, { "epoch": 0.139912039582188, "grad_norm": 1.5445071458816528, "learning_rate": 2.9060044283822025e-05, "loss": 0.0547, "step": 2036 }, { "epoch": 0.13998075865860363, "grad_norm": 2.7915220260620117, "learning_rate": 2.9058880691574425e-05, "loss": 0.057, "step": 2037 }, { "epoch": 0.14004947773501925, "grad_norm": 0.8376253843307495, "learning_rate": 2.9057716402878782e-05, "loss": 0.043, "step": 2038 }, { "epoch": 0.14011819681143486, "grad_norm": 1.8317687511444092, "learning_rate": 2.905655141779277e-05, "loss": 0.0529, "step": 2039 }, { "epoch": 0.14018691588785046, "grad_norm": 1.943080186843872, "learning_rate": 2.90553857363741e-05, "loss": 0.064, "step": 2040 }, { "epoch": 0.14025563496426607, "grad_norm": 1.273436188697815, "learning_rate": 2.9054219358680512e-05, "loss": 0.0559, "step": 2041 }, { "epoch": 0.1403243540406817, "grad_norm": 2.6847498416900635, "learning_rate": 2.9053052284769797e-05, "loss": 0.0556, "step": 2042 }, { "epoch": 0.1403930731170973, "grad_norm": 2.5667436122894287, "learning_rate": 2.9051884514699762e-05, "loss": 0.0678, "step": 2043 }, { "epoch": 0.14046179219351293, "grad_norm": 1.0334084033966064, "learning_rate": 2.9050716048528256e-05, "loss": 0.0581, "step": 2044 }, { "epoch": 0.14053051126992852, "grad_norm": 1.2263928651809692, "learning_rate": 2.9049546886313168e-05, "loss": 0.0571, "step": 2045 }, { "epoch": 0.14059923034634414, "grad_norm": 1.6292179822921753, "learning_rate": 2.9048377028112405e-05, "loss": 0.054, "step": 2046 }, { "epoch": 0.14066794942275976, "grad_norm": 1.599839448928833, "learning_rate": 2.9047206473983928e-05, "loss": 0.0634, "step": 2047 }, { "epoch": 0.14073666849917538, "grad_norm": 1.7156742811203003, "learning_rate": 2.904603522398573e-05, "loss": 0.0876, "step": 2048 }, { "epoch": 0.140805387575591, "grad_norm": 1.0818771123886108, "learning_rate": 2.9044863278175817e-05, "loss": 0.0559, "step": 2049 }, { "epoch": 0.1408741066520066, "grad_norm": 1.2709518671035767, "learning_rate": 2.9043690636612254e-05, "loss": 0.0536, "step": 2050 }, { "epoch": 0.1409428257284222, "grad_norm": 1.4207316637039185, "learning_rate": 2.9042517299353134e-05, "loss": 0.074, "step": 2051 }, { "epoch": 0.14101154480483782, "grad_norm": 2.6015985012054443, "learning_rate": 2.904134326645657e-05, "loss": 0.0444, "step": 2052 }, { "epoch": 0.14108026388125344, "grad_norm": 1.583954095840454, "learning_rate": 2.904016853798073e-05, "loss": 0.0615, "step": 2053 }, { "epoch": 0.14114898295766906, "grad_norm": 1.4257380962371826, "learning_rate": 2.903899311398381e-05, "loss": 0.0541, "step": 2054 }, { "epoch": 0.14121770203408465, "grad_norm": 1.1719517707824707, "learning_rate": 2.903781699452403e-05, "loss": 0.067, "step": 2055 }, { "epoch": 0.14128642111050027, "grad_norm": 1.056066632270813, "learning_rate": 2.9036640179659658e-05, "loss": 0.06, "step": 2056 }, { "epoch": 0.1413551401869159, "grad_norm": 1.7605750560760498, "learning_rate": 2.9035462669448986e-05, "loss": 0.0528, "step": 2057 }, { "epoch": 0.1414238592633315, "grad_norm": 1.5145035982131958, "learning_rate": 2.9034284463950354e-05, "loss": 0.0552, "step": 2058 }, { "epoch": 0.14149257833974713, "grad_norm": 2.534074068069458, "learning_rate": 2.9033105563222126e-05, "loss": 0.0541, "step": 2059 }, { "epoch": 0.14156129741616272, "grad_norm": 1.27578604221344, "learning_rate": 2.9031925967322692e-05, "loss": 0.0497, "step": 2060 }, { "epoch": 0.14163001649257834, "grad_norm": 2.2386417388916016, "learning_rate": 2.9030745676310495e-05, "loss": 0.0507, "step": 2061 }, { "epoch": 0.14169873556899396, "grad_norm": 1.072872519493103, "learning_rate": 2.9029564690244007e-05, "loss": 0.0463, "step": 2062 }, { "epoch": 0.14176745464540957, "grad_norm": 1.536699652671814, "learning_rate": 2.9028383009181724e-05, "loss": 0.0606, "step": 2063 }, { "epoch": 0.14183617372182517, "grad_norm": 4.594063758850098, "learning_rate": 2.902720063318219e-05, "loss": 0.062, "step": 2064 }, { "epoch": 0.14190489279824078, "grad_norm": 2.3644540309906006, "learning_rate": 2.9026017562303977e-05, "loss": 0.087, "step": 2065 }, { "epoch": 0.1419736118746564, "grad_norm": 2.7649734020233154, "learning_rate": 2.902483379660569e-05, "loss": 0.0738, "step": 2066 }, { "epoch": 0.14204233095107202, "grad_norm": 1.4287198781967163, "learning_rate": 2.902364933614597e-05, "loss": 0.0551, "step": 2067 }, { "epoch": 0.14211105002748764, "grad_norm": 2.3251771926879883, "learning_rate": 2.9022464180983487e-05, "loss": 0.0611, "step": 2068 }, { "epoch": 0.14217976910390323, "grad_norm": 4.377283096313477, "learning_rate": 2.9021278331176963e-05, "loss": 0.076, "step": 2069 }, { "epoch": 0.14224848818031885, "grad_norm": 1.6297581195831299, "learning_rate": 2.9020091786785136e-05, "loss": 0.054, "step": 2070 }, { "epoch": 0.14231720725673447, "grad_norm": 2.405329942703247, "learning_rate": 2.9018904547866784e-05, "loss": 0.0648, "step": 2071 }, { "epoch": 0.1423859263331501, "grad_norm": 2.136078357696533, "learning_rate": 2.9017716614480728e-05, "loss": 0.0649, "step": 2072 }, { "epoch": 0.1424546454095657, "grad_norm": 1.4455718994140625, "learning_rate": 2.9016527986685807e-05, "loss": 0.0627, "step": 2073 }, { "epoch": 0.1425233644859813, "grad_norm": 3.561373710632324, "learning_rate": 2.9015338664540905e-05, "loss": 0.0521, "step": 2074 }, { "epoch": 0.14259208356239692, "grad_norm": 1.852992296218872, "learning_rate": 2.901414864810494e-05, "loss": 0.0644, "step": 2075 }, { "epoch": 0.14266080263881253, "grad_norm": 2.3063690662384033, "learning_rate": 2.9012957937436862e-05, "loss": 0.0653, "step": 2076 }, { "epoch": 0.14272952171522815, "grad_norm": 1.1840442419052124, "learning_rate": 2.901176653259566e-05, "loss": 0.0622, "step": 2077 }, { "epoch": 0.14279824079164377, "grad_norm": 1.8156594038009644, "learning_rate": 2.9010574433640352e-05, "loss": 0.0645, "step": 2078 }, { "epoch": 0.14286695986805936, "grad_norm": 1.4321231842041016, "learning_rate": 2.900938164062999e-05, "loss": 0.0519, "step": 2079 }, { "epoch": 0.14293567894447498, "grad_norm": 1.4491933584213257, "learning_rate": 2.9008188153623664e-05, "loss": 0.0578, "step": 2080 }, { "epoch": 0.1430043980208906, "grad_norm": 1.7442904710769653, "learning_rate": 2.9006993972680493e-05, "loss": 0.0715, "step": 2081 }, { "epoch": 0.14307311709730622, "grad_norm": 1.3390010595321655, "learning_rate": 2.900579909785964e-05, "loss": 0.055, "step": 2082 }, { "epoch": 0.14314183617372184, "grad_norm": 2.0120913982391357, "learning_rate": 2.90046035292203e-05, "loss": 0.0532, "step": 2083 }, { "epoch": 0.14321055525013743, "grad_norm": 2.825474262237549, "learning_rate": 2.9003407266821684e-05, "loss": 0.0684, "step": 2084 }, { "epoch": 0.14327927432655305, "grad_norm": 2.3631112575531006, "learning_rate": 2.9002210310723065e-05, "loss": 0.0855, "step": 2085 }, { "epoch": 0.14334799340296867, "grad_norm": 1.62999427318573, "learning_rate": 2.900101266098373e-05, "loss": 0.0562, "step": 2086 }, { "epoch": 0.14341671247938428, "grad_norm": 2.8821768760681152, "learning_rate": 2.8999814317663025e-05, "loss": 0.0778, "step": 2087 }, { "epoch": 0.1434854315557999, "grad_norm": 1.6027945280075073, "learning_rate": 2.899861528082029e-05, "loss": 0.0598, "step": 2088 }, { "epoch": 0.1435541506322155, "grad_norm": 2.0875909328460693, "learning_rate": 2.8997415550514938e-05, "loss": 0.0536, "step": 2089 }, { "epoch": 0.1436228697086311, "grad_norm": 2.6874656677246094, "learning_rate": 2.89962151268064e-05, "loss": 0.0615, "step": 2090 }, { "epoch": 0.14369158878504673, "grad_norm": 1.57175612449646, "learning_rate": 2.8995014009754134e-05, "loss": 0.0735, "step": 2091 }, { "epoch": 0.14376030786146235, "grad_norm": 2.1523666381835938, "learning_rate": 2.899381219941765e-05, "loss": 0.0596, "step": 2092 }, { "epoch": 0.14382902693787797, "grad_norm": 1.5903922319412231, "learning_rate": 2.899260969585648e-05, "loss": 0.0691, "step": 2093 }, { "epoch": 0.14389774601429356, "grad_norm": 2.554849863052368, "learning_rate": 2.8991406499130188e-05, "loss": 0.0618, "step": 2094 }, { "epoch": 0.14396646509070918, "grad_norm": 2.1526710987091064, "learning_rate": 2.899020260929839e-05, "loss": 0.0707, "step": 2095 }, { "epoch": 0.1440351841671248, "grad_norm": 7.342877388000488, "learning_rate": 2.8988998026420714e-05, "loss": 0.0779, "step": 2096 }, { "epoch": 0.14410390324354042, "grad_norm": 2.101358413696289, "learning_rate": 2.8987792750556835e-05, "loss": 0.0652, "step": 2097 }, { "epoch": 0.144172622319956, "grad_norm": 2.2984206676483154, "learning_rate": 2.8986586781766463e-05, "loss": 0.0614, "step": 2098 }, { "epoch": 0.14424134139637163, "grad_norm": 3.0480711460113525, "learning_rate": 2.8985380120109334e-05, "loss": 0.0652, "step": 2099 }, { "epoch": 0.14431006047278724, "grad_norm": 1.585041880607605, "learning_rate": 2.898417276564523e-05, "loss": 0.0572, "step": 2100 }, { "epoch": 0.14437877954920286, "grad_norm": 1.7166008949279785, "learning_rate": 2.8982964718433957e-05, "loss": 0.0618, "step": 2101 }, { "epoch": 0.14444749862561848, "grad_norm": 1.2385637760162354, "learning_rate": 2.898175597853536e-05, "loss": 0.0595, "step": 2102 }, { "epoch": 0.14451621770203407, "grad_norm": 2.617108106613159, "learning_rate": 2.8980546546009317e-05, "loss": 0.0827, "step": 2103 }, { "epoch": 0.1445849367784497, "grad_norm": 1.8720558881759644, "learning_rate": 2.8979336420915738e-05, "loss": 0.0716, "step": 2104 }, { "epoch": 0.1446536558548653, "grad_norm": 1.514114260673523, "learning_rate": 2.8978125603314576e-05, "loss": 0.0712, "step": 2105 }, { "epoch": 0.14472237493128093, "grad_norm": 2.232802152633667, "learning_rate": 2.8976914093265807e-05, "loss": 0.0756, "step": 2106 }, { "epoch": 0.14479109400769655, "grad_norm": 1.380678653717041, "learning_rate": 2.897570189082945e-05, "loss": 0.0428, "step": 2107 }, { "epoch": 0.14485981308411214, "grad_norm": 1.4192532300949097, "learning_rate": 2.8974488996065554e-05, "loss": 0.0707, "step": 2108 }, { "epoch": 0.14492853216052776, "grad_norm": 1.3564231395721436, "learning_rate": 2.8973275409034207e-05, "loss": 0.0658, "step": 2109 }, { "epoch": 0.14499725123694338, "grad_norm": 1.7984602451324463, "learning_rate": 2.897206112979552e-05, "loss": 0.0624, "step": 2110 }, { "epoch": 0.145065970313359, "grad_norm": 1.4086323976516724, "learning_rate": 2.897084615840965e-05, "loss": 0.0471, "step": 2111 }, { "epoch": 0.1451346893897746, "grad_norm": 2.078768253326416, "learning_rate": 2.8969630494936784e-05, "loss": 0.0668, "step": 2112 }, { "epoch": 0.1452034084661902, "grad_norm": 1.1067445278167725, "learning_rate": 2.8968414139437144e-05, "loss": 0.0589, "step": 2113 }, { "epoch": 0.14527212754260582, "grad_norm": 1.944395661354065, "learning_rate": 2.8967197091970988e-05, "loss": 0.0623, "step": 2114 }, { "epoch": 0.14534084661902144, "grad_norm": 1.4792667627334595, "learning_rate": 2.89659793525986e-05, "loss": 0.0613, "step": 2115 }, { "epoch": 0.14540956569543706, "grad_norm": 1.5606272220611572, "learning_rate": 2.896476092138031e-05, "loss": 0.0677, "step": 2116 }, { "epoch": 0.14547828477185268, "grad_norm": 1.7462468147277832, "learning_rate": 2.896354179837647e-05, "loss": 0.0708, "step": 2117 }, { "epoch": 0.14554700384826827, "grad_norm": 2.023568630218506, "learning_rate": 2.8962321983647486e-05, "loss": 0.0705, "step": 2118 }, { "epoch": 0.1456157229246839, "grad_norm": 1.620667576789856, "learning_rate": 2.8961101477253767e-05, "loss": 0.0342, "step": 2119 }, { "epoch": 0.1456844420010995, "grad_norm": 2.549578905105591, "learning_rate": 2.895988027925579e-05, "loss": 0.0819, "step": 2120 }, { "epoch": 0.14575316107751513, "grad_norm": 1.1738543510437012, "learning_rate": 2.8958658389714045e-05, "loss": 0.0527, "step": 2121 }, { "epoch": 0.14582188015393074, "grad_norm": 0.8657337427139282, "learning_rate": 2.8957435808689054e-05, "loss": 0.0967, "step": 2122 }, { "epoch": 0.14589059923034634, "grad_norm": 1.55435049533844, "learning_rate": 2.8956212536241395e-05, "loss": 0.0486, "step": 2123 }, { "epoch": 0.14595931830676195, "grad_norm": 1.2074878215789795, "learning_rate": 2.895498857243166e-05, "loss": 0.0571, "step": 2124 }, { "epoch": 0.14602803738317757, "grad_norm": 1.2397905588150024, "learning_rate": 2.8953763917320472e-05, "loss": 0.0519, "step": 2125 }, { "epoch": 0.1460967564595932, "grad_norm": 1.3727575540542603, "learning_rate": 2.8952538570968516e-05, "loss": 0.0608, "step": 2126 }, { "epoch": 0.1461654755360088, "grad_norm": 1.6707115173339844, "learning_rate": 2.895131253343648e-05, "loss": 0.0533, "step": 2127 }, { "epoch": 0.1462341946124244, "grad_norm": 1.7974352836608887, "learning_rate": 2.8950085804785107e-05, "loss": 0.0706, "step": 2128 }, { "epoch": 0.14630291368884002, "grad_norm": 1.0360219478607178, "learning_rate": 2.8948858385075162e-05, "loss": 0.0518, "step": 2129 }, { "epoch": 0.14637163276525564, "grad_norm": 1.7559075355529785, "learning_rate": 2.894763027436745e-05, "loss": 0.0424, "step": 2130 }, { "epoch": 0.14644035184167126, "grad_norm": 1.527764081954956, "learning_rate": 2.894640147272281e-05, "loss": 0.0539, "step": 2131 }, { "epoch": 0.14650907091808685, "grad_norm": 1.5230244398117065, "learning_rate": 2.8945171980202118e-05, "loss": 0.0598, "step": 2132 }, { "epoch": 0.14657778999450247, "grad_norm": 1.5934207439422607, "learning_rate": 2.8943941796866267e-05, "loss": 0.0702, "step": 2133 }, { "epoch": 0.14664650907091809, "grad_norm": 1.7954308986663818, "learning_rate": 2.8942710922776213e-05, "loss": 0.0575, "step": 2134 }, { "epoch": 0.1467152281473337, "grad_norm": 1.411750316619873, "learning_rate": 2.8941479357992925e-05, "loss": 0.0494, "step": 2135 }, { "epoch": 0.14678394722374932, "grad_norm": 1.3312351703643799, "learning_rate": 2.8940247102577407e-05, "loss": 0.0556, "step": 2136 }, { "epoch": 0.1468526663001649, "grad_norm": 1.7023886442184448, "learning_rate": 2.893901415659071e-05, "loss": 0.0644, "step": 2137 }, { "epoch": 0.14692138537658053, "grad_norm": 1.8062877655029297, "learning_rate": 2.893778052009391e-05, "loss": 0.0632, "step": 2138 }, { "epoch": 0.14699010445299615, "grad_norm": 1.6433078050613403, "learning_rate": 2.8936546193148122e-05, "loss": 0.069, "step": 2139 }, { "epoch": 0.14705882352941177, "grad_norm": 1.4098433256149292, "learning_rate": 2.893531117581448e-05, "loss": 0.0715, "step": 2140 }, { "epoch": 0.1471275426058274, "grad_norm": 0.9145337343215942, "learning_rate": 2.8934075468154178e-05, "loss": 0.0373, "step": 2141 }, { "epoch": 0.14719626168224298, "grad_norm": 1.524613618850708, "learning_rate": 2.8932839070228422e-05, "loss": 0.0641, "step": 2142 }, { "epoch": 0.1472649807586586, "grad_norm": 1.295087456703186, "learning_rate": 2.8931601982098465e-05, "loss": 0.0467, "step": 2143 }, { "epoch": 0.14733369983507422, "grad_norm": 1.6883684396743774, "learning_rate": 2.8930364203825582e-05, "loss": 0.061, "step": 2144 }, { "epoch": 0.14740241891148984, "grad_norm": 1.686668038368225, "learning_rate": 2.8929125735471104e-05, "loss": 0.0674, "step": 2145 }, { "epoch": 0.14747113798790545, "grad_norm": 1.145614743232727, "learning_rate": 2.892788657709637e-05, "loss": 0.0638, "step": 2146 }, { "epoch": 0.14753985706432104, "grad_norm": 1.9101309776306152, "learning_rate": 2.8926646728762768e-05, "loss": 0.0656, "step": 2147 }, { "epoch": 0.14760857614073666, "grad_norm": 1.3578566312789917, "learning_rate": 2.8925406190531722e-05, "loss": 0.0621, "step": 2148 }, { "epoch": 0.14767729521715228, "grad_norm": 2.274226427078247, "learning_rate": 2.8924164962464685e-05, "loss": 0.0708, "step": 2149 }, { "epoch": 0.1477460142935679, "grad_norm": 1.5110414028167725, "learning_rate": 2.892292304462314e-05, "loss": 0.0404, "step": 2150 }, { "epoch": 0.14781473336998352, "grad_norm": 1.826673150062561, "learning_rate": 2.892168043706861e-05, "loss": 0.0538, "step": 2151 }, { "epoch": 0.1478834524463991, "grad_norm": 1.3485395908355713, "learning_rate": 2.8920437139862656e-05, "loss": 0.0366, "step": 2152 }, { "epoch": 0.14795217152281473, "grad_norm": 1.1999167203903198, "learning_rate": 2.891919315306686e-05, "loss": 0.0462, "step": 2153 }, { "epoch": 0.14802089059923035, "grad_norm": 2.2561514377593994, "learning_rate": 2.891794847674286e-05, "loss": 0.0553, "step": 2154 }, { "epoch": 0.14808960967564597, "grad_norm": 1.4792383909225464, "learning_rate": 2.89167031109523e-05, "loss": 0.0634, "step": 2155 }, { "epoch": 0.14815832875206159, "grad_norm": 1.5227837562561035, "learning_rate": 2.8915457055756883e-05, "loss": 0.0521, "step": 2156 }, { "epoch": 0.14822704782847718, "grad_norm": 1.3312023878097534, "learning_rate": 2.891421031121833e-05, "loss": 0.0632, "step": 2157 }, { "epoch": 0.1482957669048928, "grad_norm": 1.557999610900879, "learning_rate": 2.8912962877398406e-05, "loss": 0.0705, "step": 2158 }, { "epoch": 0.1483644859813084, "grad_norm": 1.6181190013885498, "learning_rate": 2.89117147543589e-05, "loss": 0.0497, "step": 2159 }, { "epoch": 0.14843320505772403, "grad_norm": 1.2789968252182007, "learning_rate": 2.8910465942161654e-05, "loss": 0.0609, "step": 2160 }, { "epoch": 0.14850192413413962, "grad_norm": 1.3832417726516724, "learning_rate": 2.8909216440868518e-05, "loss": 0.0564, "step": 2161 }, { "epoch": 0.14857064321055524, "grad_norm": 1.9996662139892578, "learning_rate": 2.89079662505414e-05, "loss": 0.0638, "step": 2162 }, { "epoch": 0.14863936228697086, "grad_norm": 1.6049253940582275, "learning_rate": 2.8906715371242225e-05, "loss": 0.0644, "step": 2163 }, { "epoch": 0.14870808136338648, "grad_norm": 1.4661363363265991, "learning_rate": 2.8905463803032962e-05, "loss": 0.0566, "step": 2164 }, { "epoch": 0.1487768004398021, "grad_norm": 1.0771355628967285, "learning_rate": 2.8904211545975612e-05, "loss": 0.0448, "step": 2165 }, { "epoch": 0.1488455195162177, "grad_norm": 1.4349390268325806, "learning_rate": 2.8902958600132207e-05, "loss": 0.0522, "step": 2166 }, { "epoch": 0.1489142385926333, "grad_norm": 1.6414523124694824, "learning_rate": 2.8901704965564812e-05, "loss": 0.0613, "step": 2167 }, { "epoch": 0.14898295766904893, "grad_norm": 1.9568958282470703, "learning_rate": 2.890045064233554e-05, "loss": 0.0747, "step": 2168 }, { "epoch": 0.14905167674546455, "grad_norm": 1.321463942527771, "learning_rate": 2.8899195630506517e-05, "loss": 0.0735, "step": 2169 }, { "epoch": 0.14912039582188016, "grad_norm": 1.832391381263733, "learning_rate": 2.8897939930139924e-05, "loss": 0.0786, "step": 2170 }, { "epoch": 0.14918911489829575, "grad_norm": 1.3163652420043945, "learning_rate": 2.889668354129795e-05, "loss": 0.052, "step": 2171 }, { "epoch": 0.14925783397471137, "grad_norm": 1.8896701335906982, "learning_rate": 2.8895426464042852e-05, "loss": 0.0574, "step": 2172 }, { "epoch": 0.149326553051127, "grad_norm": 1.2923696041107178, "learning_rate": 2.889416869843689e-05, "loss": 0.0611, "step": 2173 }, { "epoch": 0.1493952721275426, "grad_norm": 1.6471734046936035, "learning_rate": 2.8892910244542383e-05, "loss": 0.0727, "step": 2174 }, { "epoch": 0.14946399120395823, "grad_norm": 1.5219652652740479, "learning_rate": 2.8891651102421658e-05, "loss": 0.063, "step": 2175 }, { "epoch": 0.14953271028037382, "grad_norm": 2.504763126373291, "learning_rate": 2.88903912721371e-05, "loss": 0.058, "step": 2176 }, { "epoch": 0.14960142935678944, "grad_norm": 1.5286451578140259, "learning_rate": 2.8889130753751116e-05, "loss": 0.043, "step": 2177 }, { "epoch": 0.14967014843320506, "grad_norm": 1.698801040649414, "learning_rate": 2.8887869547326148e-05, "loss": 0.0571, "step": 2178 }, { "epoch": 0.14973886750962068, "grad_norm": 1.1583162546157837, "learning_rate": 2.8886607652924676e-05, "loss": 0.063, "step": 2179 }, { "epoch": 0.1498075865860363, "grad_norm": 1.402256965637207, "learning_rate": 2.888534507060921e-05, "loss": 0.0451, "step": 2180 }, { "epoch": 0.1498763056624519, "grad_norm": 1.6113715171813965, "learning_rate": 2.8884081800442298e-05, "loss": 0.0564, "step": 2181 }, { "epoch": 0.1499450247388675, "grad_norm": 2.9103779792785645, "learning_rate": 2.8882817842486515e-05, "loss": 0.0625, "step": 2182 }, { "epoch": 0.15001374381528312, "grad_norm": 1.5986660718917847, "learning_rate": 2.888155319680448e-05, "loss": 0.0615, "step": 2183 }, { "epoch": 0.15008246289169874, "grad_norm": 1.8351677656173706, "learning_rate": 2.8880287863458837e-05, "loss": 0.0465, "step": 2184 }, { "epoch": 0.15015118196811436, "grad_norm": 1.1490105390548706, "learning_rate": 2.887902184251227e-05, "loss": 0.0639, "step": 2185 }, { "epoch": 0.15021990104452995, "grad_norm": 1.2980859279632568, "learning_rate": 2.88777551340275e-05, "loss": 0.0685, "step": 2186 }, { "epoch": 0.15028862012094557, "grad_norm": 1.7431894540786743, "learning_rate": 2.8876487738067266e-05, "loss": 0.0736, "step": 2187 }, { "epoch": 0.1503573391973612, "grad_norm": 6.726208686828613, "learning_rate": 2.8875219654694357e-05, "loss": 0.0715, "step": 2188 }, { "epoch": 0.1504260582737768, "grad_norm": 1.6633565425872803, "learning_rate": 2.88739508839716e-05, "loss": 0.0678, "step": 2189 }, { "epoch": 0.15049477735019243, "grad_norm": 2.301586627960205, "learning_rate": 2.8872681425961834e-05, "loss": 0.0597, "step": 2190 }, { "epoch": 0.15056349642660802, "grad_norm": 1.5147281885147095, "learning_rate": 2.887141128072795e-05, "loss": 0.0535, "step": 2191 }, { "epoch": 0.15063221550302364, "grad_norm": 1.2339022159576416, "learning_rate": 2.8870140448332875e-05, "loss": 0.0665, "step": 2192 }, { "epoch": 0.15070093457943926, "grad_norm": 1.529157280921936, "learning_rate": 2.886886892883956e-05, "loss": 0.0689, "step": 2193 }, { "epoch": 0.15076965365585487, "grad_norm": 2.188699722290039, "learning_rate": 2.886759672231098e-05, "loss": 0.0712, "step": 2194 }, { "epoch": 0.15083837273227046, "grad_norm": 3.1364896297454834, "learning_rate": 2.886632382881018e-05, "loss": 0.0817, "step": 2195 }, { "epoch": 0.15090709180868608, "grad_norm": 1.4323772192001343, "learning_rate": 2.88650502484002e-05, "loss": 0.0675, "step": 2196 }, { "epoch": 0.1509758108851017, "grad_norm": 1.4347110986709595, "learning_rate": 2.886377598114414e-05, "loss": 0.051, "step": 2197 }, { "epoch": 0.15104452996151732, "grad_norm": 1.3107575178146362, "learning_rate": 2.8862501027105118e-05, "loss": 0.0465, "step": 2198 }, { "epoch": 0.15111324903793294, "grad_norm": 1.5224318504333496, "learning_rate": 2.8861225386346295e-05, "loss": 0.0579, "step": 2199 }, { "epoch": 0.15118196811434853, "grad_norm": 1.2065796852111816, "learning_rate": 2.8859949058930862e-05, "loss": 0.0618, "step": 2200 }, { "epoch": 0.15125068719076415, "grad_norm": 1.164223074913025, "learning_rate": 2.8858672044922052e-05, "loss": 0.0526, "step": 2201 }, { "epoch": 0.15131940626717977, "grad_norm": 1.6633458137512207, "learning_rate": 2.885739434438312e-05, "loss": 0.051, "step": 2202 }, { "epoch": 0.1513881253435954, "grad_norm": 2.239851951599121, "learning_rate": 2.885611595737736e-05, "loss": 0.0554, "step": 2203 }, { "epoch": 0.151456844420011, "grad_norm": 1.0377275943756104, "learning_rate": 2.8854836883968102e-05, "loss": 0.0534, "step": 2204 }, { "epoch": 0.1515255634964266, "grad_norm": 1.5867291688919067, "learning_rate": 2.8853557124218713e-05, "loss": 0.0491, "step": 2205 }, { "epoch": 0.15159428257284221, "grad_norm": 1.5661451816558838, "learning_rate": 2.885227667819258e-05, "loss": 0.0622, "step": 2206 }, { "epoch": 0.15166300164925783, "grad_norm": 1.4598801136016846, "learning_rate": 2.8850995545953143e-05, "loss": 0.0497, "step": 2207 }, { "epoch": 0.15173172072567345, "grad_norm": 1.7402554750442505, "learning_rate": 2.8849713727563864e-05, "loss": 0.0645, "step": 2208 }, { "epoch": 0.15180043980208907, "grad_norm": 2.201716899871826, "learning_rate": 2.884843122308824e-05, "loss": 0.0564, "step": 2209 }, { "epoch": 0.15186915887850466, "grad_norm": 1.759162425994873, "learning_rate": 2.88471480325898e-05, "loss": 0.0673, "step": 2210 }, { "epoch": 0.15193787795492028, "grad_norm": 1.2635468244552612, "learning_rate": 2.884586415613212e-05, "loss": 0.0666, "step": 2211 }, { "epoch": 0.1520065970313359, "grad_norm": 1.2224125862121582, "learning_rate": 2.8844579593778794e-05, "loss": 0.0538, "step": 2212 }, { "epoch": 0.15207531610775152, "grad_norm": 2.608285903930664, "learning_rate": 2.884329434559346e-05, "loss": 0.0769, "step": 2213 }, { "epoch": 0.15214403518416714, "grad_norm": 1.6160084009170532, "learning_rate": 2.884200841163978e-05, "loss": 0.0685, "step": 2214 }, { "epoch": 0.15221275426058273, "grad_norm": 1.6150425672531128, "learning_rate": 2.8840721791981465e-05, "loss": 0.0525, "step": 2215 }, { "epoch": 0.15228147333699835, "grad_norm": 1.5361053943634033, "learning_rate": 2.8839434486682248e-05, "loss": 0.0747, "step": 2216 }, { "epoch": 0.15235019241341397, "grad_norm": 1.7924946546554565, "learning_rate": 2.8838146495805897e-05, "loss": 0.0547, "step": 2217 }, { "epoch": 0.15241891148982958, "grad_norm": 1.772573709487915, "learning_rate": 2.883685781941622e-05, "loss": 0.0596, "step": 2218 }, { "epoch": 0.1524876305662452, "grad_norm": 1.574181079864502, "learning_rate": 2.8835568457577052e-05, "loss": 0.0591, "step": 2219 }, { "epoch": 0.1525563496426608, "grad_norm": 0.938114583492279, "learning_rate": 2.8834278410352276e-05, "loss": 0.0614, "step": 2220 }, { "epoch": 0.1526250687190764, "grad_norm": 0.7490881681442261, "learning_rate": 2.8832987677805777e-05, "loss": 0.0439, "step": 2221 }, { "epoch": 0.15269378779549203, "grad_norm": 1.9691638946533203, "learning_rate": 2.8831696260001516e-05, "loss": 0.0544, "step": 2222 }, { "epoch": 0.15276250687190765, "grad_norm": 1.0839937925338745, "learning_rate": 2.8830404157003457e-05, "loss": 0.0474, "step": 2223 }, { "epoch": 0.15283122594832327, "grad_norm": 1.494070291519165, "learning_rate": 2.882911136887561e-05, "loss": 0.0585, "step": 2224 }, { "epoch": 0.15289994502473886, "grad_norm": 1.6315083503723145, "learning_rate": 2.882781789568202e-05, "loss": 0.043, "step": 2225 }, { "epoch": 0.15296866410115448, "grad_norm": 1.4774478673934937, "learning_rate": 2.8826523737486758e-05, "loss": 0.0389, "step": 2226 }, { "epoch": 0.1530373831775701, "grad_norm": 1.720868706703186, "learning_rate": 2.882522889435393e-05, "loss": 0.0455, "step": 2227 }, { "epoch": 0.15310610225398572, "grad_norm": 2.013134241104126, "learning_rate": 2.8823933366347693e-05, "loss": 0.09, "step": 2228 }, { "epoch": 0.1531748213304013, "grad_norm": 1.2414522171020508, "learning_rate": 2.8822637153532218e-05, "loss": 0.0815, "step": 2229 }, { "epoch": 0.15324354040681692, "grad_norm": 2.1612534523010254, "learning_rate": 2.8821340255971714e-05, "loss": 0.0494, "step": 2230 }, { "epoch": 0.15331225948323254, "grad_norm": 1.5297300815582275, "learning_rate": 2.882004267373043e-05, "loss": 0.0593, "step": 2231 }, { "epoch": 0.15338097855964816, "grad_norm": 1.7004082202911377, "learning_rate": 2.8818744406872645e-05, "loss": 0.0429, "step": 2232 }, { "epoch": 0.15344969763606378, "grad_norm": 1.803650140762329, "learning_rate": 2.8817445455462673e-05, "loss": 0.0627, "step": 2233 }, { "epoch": 0.15351841671247937, "grad_norm": 1.3079930543899536, "learning_rate": 2.881614581956486e-05, "loss": 0.0584, "step": 2234 }, { "epoch": 0.153587135788895, "grad_norm": 1.5118324756622314, "learning_rate": 2.8814845499243584e-05, "loss": 0.0458, "step": 2235 }, { "epoch": 0.1536558548653106, "grad_norm": 1.9267196655273438, "learning_rate": 2.8813544494563268e-05, "loss": 0.0678, "step": 2236 }, { "epoch": 0.15372457394172623, "grad_norm": 1.3803136348724365, "learning_rate": 2.881224280558836e-05, "loss": 0.0798, "step": 2237 }, { "epoch": 0.15379329301814185, "grad_norm": 1.235332727432251, "learning_rate": 2.8810940432383335e-05, "loss": 0.0479, "step": 2238 }, { "epoch": 0.15386201209455744, "grad_norm": 2.048523187637329, "learning_rate": 2.880963737501272e-05, "loss": 0.0715, "step": 2239 }, { "epoch": 0.15393073117097306, "grad_norm": 1.3941988945007324, "learning_rate": 2.8808333633541058e-05, "loss": 0.0674, "step": 2240 }, { "epoch": 0.15399945024738868, "grad_norm": 1.4165441989898682, "learning_rate": 2.8807029208032935e-05, "loss": 0.0564, "step": 2241 }, { "epoch": 0.1540681693238043, "grad_norm": 1.5214306116104126, "learning_rate": 2.880572409855297e-05, "loss": 0.06, "step": 2242 }, { "epoch": 0.1541368884002199, "grad_norm": 1.2815715074539185, "learning_rate": 2.8804418305165822e-05, "loss": 0.0468, "step": 2243 }, { "epoch": 0.1542056074766355, "grad_norm": 1.5407253503799438, "learning_rate": 2.880311182793617e-05, "loss": 0.0692, "step": 2244 }, { "epoch": 0.15427432655305112, "grad_norm": 1.5171946287155151, "learning_rate": 2.8801804666928735e-05, "loss": 0.0702, "step": 2245 }, { "epoch": 0.15434304562946674, "grad_norm": 1.8047358989715576, "learning_rate": 2.880049682220827e-05, "loss": 0.0535, "step": 2246 }, { "epoch": 0.15441176470588236, "grad_norm": 1.1791467666625977, "learning_rate": 2.8799188293839566e-05, "loss": 0.0541, "step": 2247 }, { "epoch": 0.15448048378229798, "grad_norm": 1.9701101779937744, "learning_rate": 2.8797879081887446e-05, "loss": 0.0439, "step": 2248 }, { "epoch": 0.15454920285871357, "grad_norm": 2.2973670959472656, "learning_rate": 2.8796569186416766e-05, "loss": 0.0684, "step": 2249 }, { "epoch": 0.1546179219351292, "grad_norm": 2.4792544841766357, "learning_rate": 2.8795258607492406e-05, "loss": 0.0639, "step": 2250 }, { "epoch": 0.1546866410115448, "grad_norm": 1.1943347454071045, "learning_rate": 2.8793947345179303e-05, "loss": 0.0569, "step": 2251 }, { "epoch": 0.15475536008796043, "grad_norm": 1.8835663795471191, "learning_rate": 2.87926353995424e-05, "loss": 0.0666, "step": 2252 }, { "epoch": 0.15482407916437604, "grad_norm": 1.519564151763916, "learning_rate": 2.8791322770646703e-05, "loss": 0.0486, "step": 2253 }, { "epoch": 0.15489279824079163, "grad_norm": 1.1404626369476318, "learning_rate": 2.879000945855723e-05, "loss": 0.0565, "step": 2254 }, { "epoch": 0.15496151731720725, "grad_norm": 1.1287766695022583, "learning_rate": 2.878869546333903e-05, "loss": 0.0609, "step": 2255 }, { "epoch": 0.15503023639362287, "grad_norm": 2.062087297439575, "learning_rate": 2.8787380785057214e-05, "loss": 0.0669, "step": 2256 }, { "epoch": 0.1550989554700385, "grad_norm": 1.6804454326629639, "learning_rate": 2.87860654237769e-05, "loss": 0.0767, "step": 2257 }, { "epoch": 0.15516767454645408, "grad_norm": 1.2964171171188354, "learning_rate": 2.8784749379563246e-05, "loss": 0.0513, "step": 2258 }, { "epoch": 0.1552363936228697, "grad_norm": 3.1481525897979736, "learning_rate": 2.8783432652481447e-05, "loss": 0.0693, "step": 2259 }, { "epoch": 0.15530511269928532, "grad_norm": 1.1857467889785767, "learning_rate": 2.878211524259673e-05, "loss": 0.0461, "step": 2260 }, { "epoch": 0.15537383177570094, "grad_norm": 1.502917766571045, "learning_rate": 2.8780797149974362e-05, "loss": 0.0481, "step": 2261 }, { "epoch": 0.15544255085211656, "grad_norm": 1.8867920637130737, "learning_rate": 2.8779478374679635e-05, "loss": 0.0664, "step": 2262 }, { "epoch": 0.15551126992853215, "grad_norm": 1.6967132091522217, "learning_rate": 2.8778158916777877e-05, "loss": 0.0737, "step": 2263 }, { "epoch": 0.15557998900494777, "grad_norm": 5.346114635467529, "learning_rate": 2.877683877633445e-05, "loss": 0.0596, "step": 2264 }, { "epoch": 0.15564870808136338, "grad_norm": 1.20924711227417, "learning_rate": 2.8775517953414762e-05, "loss": 0.0682, "step": 2265 }, { "epoch": 0.155717427157779, "grad_norm": 2.135068655014038, "learning_rate": 2.877419644808423e-05, "loss": 0.0471, "step": 2266 }, { "epoch": 0.15578614623419462, "grad_norm": 1.3549294471740723, "learning_rate": 2.877287426040833e-05, "loss": 0.0748, "step": 2267 }, { "epoch": 0.1558548653106102, "grad_norm": 0.9773874282836914, "learning_rate": 2.8771551390452546e-05, "loss": 0.0643, "step": 2268 }, { "epoch": 0.15592358438702583, "grad_norm": 1.4501596689224243, "learning_rate": 2.8770227838282424e-05, "loss": 0.0574, "step": 2269 }, { "epoch": 0.15599230346344145, "grad_norm": 1.2037228345870972, "learning_rate": 2.8768903603963528e-05, "loss": 0.0491, "step": 2270 }, { "epoch": 0.15606102253985707, "grad_norm": 2.184532880783081, "learning_rate": 2.8767578687561453e-05, "loss": 0.0484, "step": 2271 }, { "epoch": 0.1561297416162727, "grad_norm": 1.6102485656738281, "learning_rate": 2.8766253089141833e-05, "loss": 0.0624, "step": 2272 }, { "epoch": 0.15619846069268828, "grad_norm": 1.467916488647461, "learning_rate": 2.876492680877034e-05, "loss": 0.062, "step": 2273 }, { "epoch": 0.1562671797691039, "grad_norm": 1.124490737915039, "learning_rate": 2.876359984651267e-05, "loss": 0.0423, "step": 2274 }, { "epoch": 0.15633589884551952, "grad_norm": 1.5278801918029785, "learning_rate": 2.8762272202434563e-05, "loss": 0.0499, "step": 2275 }, { "epoch": 0.15640461792193514, "grad_norm": 0.9681172370910645, "learning_rate": 2.8760943876601782e-05, "loss": 0.05, "step": 2276 }, { "epoch": 0.15647333699835075, "grad_norm": 1.1920549869537354, "learning_rate": 2.8759614869080134e-05, "loss": 0.0608, "step": 2277 }, { "epoch": 0.15654205607476634, "grad_norm": 1.9556396007537842, "learning_rate": 2.8758285179935457e-05, "loss": 0.0643, "step": 2278 }, { "epoch": 0.15661077515118196, "grad_norm": 1.1685118675231934, "learning_rate": 2.8756954809233614e-05, "loss": 0.0508, "step": 2279 }, { "epoch": 0.15667949422759758, "grad_norm": 0.9836272597312927, "learning_rate": 2.8755623757040514e-05, "loss": 0.0437, "step": 2280 }, { "epoch": 0.1567482133040132, "grad_norm": 1.2834434509277344, "learning_rate": 2.8754292023422096e-05, "loss": 0.0507, "step": 2281 }, { "epoch": 0.15681693238042882, "grad_norm": 0.9241469502449036, "learning_rate": 2.8752959608444324e-05, "loss": 0.0464, "step": 2282 }, { "epoch": 0.1568856514568444, "grad_norm": 1.9604812860488892, "learning_rate": 2.875162651217321e-05, "loss": 0.0576, "step": 2283 }, { "epoch": 0.15695437053326003, "grad_norm": 1.4658551216125488, "learning_rate": 2.8750292734674792e-05, "loss": 0.049, "step": 2284 }, { "epoch": 0.15702308960967565, "grad_norm": 1.5987887382507324, "learning_rate": 2.874895827601514e-05, "loss": 0.0459, "step": 2285 }, { "epoch": 0.15709180868609127, "grad_norm": 1.808686375617981, "learning_rate": 2.874762313626036e-05, "loss": 0.0531, "step": 2286 }, { "epoch": 0.15716052776250689, "grad_norm": 1.9077062606811523, "learning_rate": 2.8746287315476595e-05, "loss": 0.0482, "step": 2287 }, { "epoch": 0.15722924683892248, "grad_norm": 1.573534369468689, "learning_rate": 2.8744950813730017e-05, "loss": 0.0493, "step": 2288 }, { "epoch": 0.1572979659153381, "grad_norm": 2.416792869567871, "learning_rate": 2.8743613631086837e-05, "loss": 0.0534, "step": 2289 }, { "epoch": 0.1573666849917537, "grad_norm": 2.173914909362793, "learning_rate": 2.874227576761329e-05, "loss": 0.0658, "step": 2290 }, { "epoch": 0.15743540406816933, "grad_norm": 0.730685830116272, "learning_rate": 2.8740937223375662e-05, "loss": 0.0581, "step": 2291 }, { "epoch": 0.15750412314458492, "grad_norm": 1.569545865058899, "learning_rate": 2.8739597998440246e-05, "loss": 0.0553, "step": 2292 }, { "epoch": 0.15757284222100054, "grad_norm": 1.3445775508880615, "learning_rate": 2.8738258092873392e-05, "loss": 0.0488, "step": 2293 }, { "epoch": 0.15764156129741616, "grad_norm": 0.5875515341758728, "learning_rate": 2.873691750674148e-05, "loss": 0.0509, "step": 2294 }, { "epoch": 0.15771028037383178, "grad_norm": 1.0125765800476074, "learning_rate": 2.8735576240110917e-05, "loss": 0.0625, "step": 2295 }, { "epoch": 0.1577789994502474, "grad_norm": 1.5505305528640747, "learning_rate": 2.8734234293048144e-05, "loss": 0.0421, "step": 2296 }, { "epoch": 0.157847718526663, "grad_norm": 1.7189295291900635, "learning_rate": 2.8732891665619645e-05, "loss": 0.0479, "step": 2297 }, { "epoch": 0.1579164376030786, "grad_norm": 1.240639328956604, "learning_rate": 2.873154835789192e-05, "loss": 0.0433, "step": 2298 }, { "epoch": 0.15798515667949423, "grad_norm": 1.0121511220932007, "learning_rate": 2.8730204369931525e-05, "loss": 0.0673, "step": 2299 }, { "epoch": 0.15805387575590985, "grad_norm": 1.3237980604171753, "learning_rate": 2.8728859701805032e-05, "loss": 0.0535, "step": 2300 }, { "epoch": 0.15812259483232546, "grad_norm": 0.9856956601142883, "learning_rate": 2.8727514353579058e-05, "loss": 0.0671, "step": 2301 }, { "epoch": 0.15819131390874105, "grad_norm": 1.8763090372085571, "learning_rate": 2.872616832532024e-05, "loss": 0.0543, "step": 2302 }, { "epoch": 0.15826003298515667, "grad_norm": 1.351618766784668, "learning_rate": 2.8724821617095266e-05, "loss": 0.0532, "step": 2303 }, { "epoch": 0.1583287520615723, "grad_norm": 1.7952522039413452, "learning_rate": 2.8723474228970843e-05, "loss": 0.0469, "step": 2304 }, { "epoch": 0.1583974711379879, "grad_norm": 1.3307902812957764, "learning_rate": 2.8722126161013725e-05, "loss": 0.0518, "step": 2305 }, { "epoch": 0.15846619021440353, "grad_norm": 1.3494476079940796, "learning_rate": 2.8720777413290687e-05, "loss": 0.0373, "step": 2306 }, { "epoch": 0.15853490929081912, "grad_norm": 1.373231291770935, "learning_rate": 2.8719427985868548e-05, "loss": 0.0505, "step": 2307 }, { "epoch": 0.15860362836723474, "grad_norm": 1.2239494323730469, "learning_rate": 2.8718077878814148e-05, "loss": 0.055, "step": 2308 }, { "epoch": 0.15867234744365036, "grad_norm": 0.8876920938491821, "learning_rate": 2.8716727092194372e-05, "loss": 0.0402, "step": 2309 }, { "epoch": 0.15874106652006598, "grad_norm": 169.12754821777344, "learning_rate": 2.871537562607614e-05, "loss": 0.1493, "step": 2310 }, { "epoch": 0.1588097855964816, "grad_norm": 5.327749729156494, "learning_rate": 2.871402348052639e-05, "loss": 0.2412, "step": 2311 }, { "epoch": 0.1588785046728972, "grad_norm": 25.493391036987305, "learning_rate": 2.8712670655612122e-05, "loss": 0.0811, "step": 2312 }, { "epoch": 0.1589472237493128, "grad_norm": 2.610781192779541, "learning_rate": 2.8711317151400337e-05, "loss": 0.1433, "step": 2313 }, { "epoch": 0.15901594282572842, "grad_norm": 3.013002872467041, "learning_rate": 2.870996296795809e-05, "loss": 0.0983, "step": 2314 }, { "epoch": 0.15908466190214404, "grad_norm": 1.861437439918518, "learning_rate": 2.8708608105352464e-05, "loss": 0.0807, "step": 2315 }, { "epoch": 0.15915338097855966, "grad_norm": 4.4711384773254395, "learning_rate": 2.8707252563650578e-05, "loss": 0.0858, "step": 2316 }, { "epoch": 0.15922210005497525, "grad_norm": 4.527608871459961, "learning_rate": 2.8705896342919576e-05, "loss": 0.1259, "step": 2317 }, { "epoch": 0.15929081913139087, "grad_norm": 2.0853662490844727, "learning_rate": 2.870453944322665e-05, "loss": 0.0902, "step": 2318 }, { "epoch": 0.1593595382078065, "grad_norm": 2.969926357269287, "learning_rate": 2.8703181864639013e-05, "loss": 0.085, "step": 2319 }, { "epoch": 0.1594282572842221, "grad_norm": 1.7045502662658691, "learning_rate": 2.8701823607223923e-05, "loss": 0.0609, "step": 2320 }, { "epoch": 0.15949697636063773, "grad_norm": 2.0901167392730713, "learning_rate": 2.8700464671048658e-05, "loss": 0.0786, "step": 2321 }, { "epoch": 0.15956569543705332, "grad_norm": 2.4593772888183594, "learning_rate": 2.8699105056180542e-05, "loss": 0.0588, "step": 2322 }, { "epoch": 0.15963441451346894, "grad_norm": 1.903010606765747, "learning_rate": 2.8697744762686926e-05, "loss": 0.0559, "step": 2323 }, { "epoch": 0.15970313358988455, "grad_norm": 2.6376190185546875, "learning_rate": 2.8696383790635195e-05, "loss": 0.0654, "step": 2324 }, { "epoch": 0.15977185266630017, "grad_norm": 1.041825771331787, "learning_rate": 2.869502214009277e-05, "loss": 0.0604, "step": 2325 }, { "epoch": 0.15984057174271576, "grad_norm": 2.3672287464141846, "learning_rate": 2.8693659811127103e-05, "loss": 0.0435, "step": 2326 }, { "epoch": 0.15990929081913138, "grad_norm": 1.794390320777893, "learning_rate": 2.8692296803805683e-05, "loss": 0.0783, "step": 2327 }, { "epoch": 0.159978009895547, "grad_norm": 2.56387996673584, "learning_rate": 2.869093311819603e-05, "loss": 0.0742, "step": 2328 }, { "epoch": 0.16004672897196262, "grad_norm": 0.9940693378448486, "learning_rate": 2.8689568754365696e-05, "loss": 0.0663, "step": 2329 }, { "epoch": 0.16011544804837824, "grad_norm": 1.7923877239227295, "learning_rate": 2.868820371238227e-05, "loss": 0.0628, "step": 2330 }, { "epoch": 0.16018416712479383, "grad_norm": 1.712680459022522, "learning_rate": 2.868683799231338e-05, "loss": 0.0733, "step": 2331 }, { "epoch": 0.16025288620120945, "grad_norm": 1.3782176971435547, "learning_rate": 2.868547159422667e-05, "loss": 0.0716, "step": 2332 }, { "epoch": 0.16032160527762507, "grad_norm": 1.7630497217178345, "learning_rate": 2.868410451818983e-05, "loss": 0.0396, "step": 2333 }, { "epoch": 0.1603903243540407, "grad_norm": 2.39050555229187, "learning_rate": 2.8682736764270592e-05, "loss": 0.0707, "step": 2334 }, { "epoch": 0.1604590434304563, "grad_norm": 1.7381501197814941, "learning_rate": 2.8681368332536697e-05, "loss": 0.0651, "step": 2335 }, { "epoch": 0.1605277625068719, "grad_norm": 3.0990357398986816, "learning_rate": 2.867999922305595e-05, "loss": 0.0767, "step": 2336 }, { "epoch": 0.16059648158328751, "grad_norm": 2.857360363006592, "learning_rate": 2.8678629435896165e-05, "loss": 0.0715, "step": 2337 }, { "epoch": 0.16066520065970313, "grad_norm": 1.5843474864959717, "learning_rate": 2.8677258971125198e-05, "loss": 0.074, "step": 2338 }, { "epoch": 0.16073391973611875, "grad_norm": 2.200925827026367, "learning_rate": 2.8675887828810943e-05, "loss": 0.0779, "step": 2339 }, { "epoch": 0.16080263881253437, "grad_norm": 1.6729859113693237, "learning_rate": 2.867451600902132e-05, "loss": 0.0581, "step": 2340 }, { "epoch": 0.16087135788894996, "grad_norm": 1.9765353202819824, "learning_rate": 2.8673143511824286e-05, "loss": 0.0694, "step": 2341 }, { "epoch": 0.16094007696536558, "grad_norm": 2.1574723720550537, "learning_rate": 2.8671770337287832e-05, "loss": 0.0856, "step": 2342 }, { "epoch": 0.1610087960417812, "grad_norm": 0.9851827025413513, "learning_rate": 2.8670396485479988e-05, "loss": 0.0452, "step": 2343 }, { "epoch": 0.16107751511819682, "grad_norm": 1.2553386688232422, "learning_rate": 2.8669021956468805e-05, "loss": 0.053, "step": 2344 }, { "epoch": 0.16114623419461244, "grad_norm": 1.6102678775787354, "learning_rate": 2.8667646750322377e-05, "loss": 0.0579, "step": 2345 }, { "epoch": 0.16121495327102803, "grad_norm": 1.0014368295669556, "learning_rate": 2.866627086710883e-05, "loss": 0.0422, "step": 2346 }, { "epoch": 0.16128367234744365, "grad_norm": 1.0707483291625977, "learning_rate": 2.8664894306896318e-05, "loss": 0.0454, "step": 2347 }, { "epoch": 0.16135239142385926, "grad_norm": 1.6361479759216309, "learning_rate": 2.866351706975304e-05, "loss": 0.0667, "step": 2348 }, { "epoch": 0.16142111050027488, "grad_norm": 1.77864408493042, "learning_rate": 2.8662139155747212e-05, "loss": 0.0566, "step": 2349 }, { "epoch": 0.1614898295766905, "grad_norm": 1.4065181016921997, "learning_rate": 2.8660760564947105e-05, "loss": 0.0526, "step": 2350 }, { "epoch": 0.1615585486531061, "grad_norm": 2.0849609375, "learning_rate": 2.8659381297421e-05, "loss": 0.0407, "step": 2351 }, { "epoch": 0.1616272677295217, "grad_norm": 1.2764136791229248, "learning_rate": 2.8658001353237227e-05, "loss": 0.0539, "step": 2352 }, { "epoch": 0.16169598680593733, "grad_norm": 1.5051246881484985, "learning_rate": 2.865662073246415e-05, "loss": 0.0543, "step": 2353 }, { "epoch": 0.16176470588235295, "grad_norm": 2.2590363025665283, "learning_rate": 2.8655239435170156e-05, "loss": 0.0483, "step": 2354 }, { "epoch": 0.16183342495876854, "grad_norm": 1.1077719926834106, "learning_rate": 2.8653857461423674e-05, "loss": 0.025, "step": 2355 }, { "epoch": 0.16190214403518416, "grad_norm": 1.2534583806991577, "learning_rate": 2.8652474811293164e-05, "loss": 0.0437, "step": 2356 }, { "epoch": 0.16197086311159978, "grad_norm": 2.022531747817993, "learning_rate": 2.8651091484847127e-05, "loss": 0.057, "step": 2357 }, { "epoch": 0.1620395821880154, "grad_norm": 1.7591077089309692, "learning_rate": 2.8649707482154078e-05, "loss": 0.0369, "step": 2358 }, { "epoch": 0.16210830126443102, "grad_norm": 1.1581658124923706, "learning_rate": 2.8648322803282583e-05, "loss": 0.0498, "step": 2359 }, { "epoch": 0.1621770203408466, "grad_norm": 1.5563384294509888, "learning_rate": 2.8646937448301237e-05, "loss": 0.0499, "step": 2360 }, { "epoch": 0.16224573941726222, "grad_norm": 1.2734442949295044, "learning_rate": 2.8645551417278667e-05, "loss": 0.0636, "step": 2361 }, { "epoch": 0.16231445849367784, "grad_norm": 1.1805646419525146, "learning_rate": 2.864416471028353e-05, "loss": 0.0558, "step": 2362 }, { "epoch": 0.16238317757009346, "grad_norm": 1.9616849422454834, "learning_rate": 2.864277732738453e-05, "loss": 0.059, "step": 2363 }, { "epoch": 0.16245189664650908, "grad_norm": 1.0121382474899292, "learning_rate": 2.8641389268650387e-05, "loss": 0.05, "step": 2364 }, { "epoch": 0.16252061572292467, "grad_norm": 1.494981050491333, "learning_rate": 2.864000053414987e-05, "loss": 0.0504, "step": 2365 }, { "epoch": 0.1625893347993403, "grad_norm": 2.151540994644165, "learning_rate": 2.863861112395176e-05, "loss": 0.0563, "step": 2366 }, { "epoch": 0.1626580538757559, "grad_norm": 1.574568748474121, "learning_rate": 2.86372210381249e-05, "loss": 0.0416, "step": 2367 }, { "epoch": 0.16272677295217153, "grad_norm": 1.1952404975891113, "learning_rate": 2.8635830276738146e-05, "loss": 0.0578, "step": 2368 }, { "epoch": 0.16279549202858715, "grad_norm": 0.8519805669784546, "learning_rate": 2.8634438839860396e-05, "loss": 0.0559, "step": 2369 }, { "epoch": 0.16286421110500274, "grad_norm": 2.478144884109497, "learning_rate": 2.8633046727560574e-05, "loss": 0.0569, "step": 2370 }, { "epoch": 0.16293293018141836, "grad_norm": 1.3771967887878418, "learning_rate": 2.863165393990765e-05, "loss": 0.05, "step": 2371 }, { "epoch": 0.16300164925783397, "grad_norm": 1.3667781352996826, "learning_rate": 2.8630260476970612e-05, "loss": 0.0377, "step": 2372 }, { "epoch": 0.1630703683342496, "grad_norm": 1.221164584159851, "learning_rate": 2.8628866338818488e-05, "loss": 0.0363, "step": 2373 }, { "epoch": 0.1631390874106652, "grad_norm": 1.5433571338653564, "learning_rate": 2.8627471525520353e-05, "loss": 0.0558, "step": 2374 }, { "epoch": 0.1632078064870808, "grad_norm": 1.2885921001434326, "learning_rate": 2.8626076037145293e-05, "loss": 0.0703, "step": 2375 }, { "epoch": 0.16327652556349642, "grad_norm": 1.5044835805892944, "learning_rate": 2.862467987376244e-05, "loss": 0.0492, "step": 2376 }, { "epoch": 0.16334524463991204, "grad_norm": 1.669851303100586, "learning_rate": 2.8623283035440957e-05, "loss": 0.0616, "step": 2377 }, { "epoch": 0.16341396371632766, "grad_norm": 0.9495413303375244, "learning_rate": 2.862188552225004e-05, "loss": 0.0558, "step": 2378 }, { "epoch": 0.16348268279274328, "grad_norm": 2.317208766937256, "learning_rate": 2.862048733425892e-05, "loss": 0.0426, "step": 2379 }, { "epoch": 0.16355140186915887, "grad_norm": 1.4432202577590942, "learning_rate": 2.861908847153686e-05, "loss": 0.0575, "step": 2380 }, { "epoch": 0.1636201209455745, "grad_norm": 2.668297290802002, "learning_rate": 2.8617688934153158e-05, "loss": 0.053, "step": 2381 }, { "epoch": 0.1636888400219901, "grad_norm": 1.4940309524536133, "learning_rate": 2.861628872217714e-05, "loss": 0.0602, "step": 2382 }, { "epoch": 0.16375755909840572, "grad_norm": 1.9810988903045654, "learning_rate": 2.8614887835678176e-05, "loss": 0.0583, "step": 2383 }, { "epoch": 0.16382627817482134, "grad_norm": 1.91853928565979, "learning_rate": 2.861348627472566e-05, "loss": 0.0563, "step": 2384 }, { "epoch": 0.16389499725123693, "grad_norm": 2.0192670822143555, "learning_rate": 2.8612084039389017e-05, "loss": 0.0599, "step": 2385 }, { "epoch": 0.16396371632765255, "grad_norm": 1.5149556398391724, "learning_rate": 2.861068112973772e-05, "loss": 0.0529, "step": 2386 }, { "epoch": 0.16403243540406817, "grad_norm": 1.2120085954666138, "learning_rate": 2.8609277545841263e-05, "loss": 0.0515, "step": 2387 }, { "epoch": 0.1641011544804838, "grad_norm": 1.5634443759918213, "learning_rate": 2.860787328776917e-05, "loss": 0.0567, "step": 2388 }, { "epoch": 0.16416987355689938, "grad_norm": 1.276517629623413, "learning_rate": 2.8606468355591018e-05, "loss": 0.0565, "step": 2389 }, { "epoch": 0.164238592633315, "grad_norm": 2.6884937286376953, "learning_rate": 2.8605062749376395e-05, "loss": 0.0694, "step": 2390 }, { "epoch": 0.16430731170973062, "grad_norm": 1.5503817796707153, "learning_rate": 2.8603656469194933e-05, "loss": 0.0649, "step": 2391 }, { "epoch": 0.16437603078614624, "grad_norm": 2.670335054397583, "learning_rate": 2.86022495151163e-05, "loss": 0.0629, "step": 2392 }, { "epoch": 0.16444474986256186, "grad_norm": 1.0490410327911377, "learning_rate": 2.8600841887210188e-05, "loss": 0.0448, "step": 2393 }, { "epoch": 0.16451346893897745, "grad_norm": 1.6027523279190063, "learning_rate": 2.8599433585546334e-05, "loss": 0.0423, "step": 2394 }, { "epoch": 0.16458218801539307, "grad_norm": 2.255716562271118, "learning_rate": 2.8598024610194497e-05, "loss": 0.0564, "step": 2395 }, { "epoch": 0.16465090709180868, "grad_norm": 1.8585389852523804, "learning_rate": 2.859661496122448e-05, "loss": 0.0495, "step": 2396 }, { "epoch": 0.1647196261682243, "grad_norm": 0.9734734892845154, "learning_rate": 2.859520463870611e-05, "loss": 0.0499, "step": 2397 }, { "epoch": 0.16478834524463992, "grad_norm": 1.5411605834960938, "learning_rate": 2.8593793642709252e-05, "loss": 0.0542, "step": 2398 }, { "epoch": 0.1648570643210555, "grad_norm": 1.1935160160064697, "learning_rate": 2.8592381973303803e-05, "loss": 0.0537, "step": 2399 }, { "epoch": 0.16492578339747113, "grad_norm": 1.2701102495193481, "learning_rate": 2.8590969630559698e-05, "loss": 0.0487, "step": 2400 }, { "epoch": 0.16499450247388675, "grad_norm": 0.9345831871032715, "learning_rate": 2.8589556614546897e-05, "loss": 0.0431, "step": 2401 }, { "epoch": 0.16506322155030237, "grad_norm": 1.2052041292190552, "learning_rate": 2.8588142925335404e-05, "loss": 0.0531, "step": 2402 }, { "epoch": 0.165131940626718, "grad_norm": 1.6038867235183716, "learning_rate": 2.8586728562995243e-05, "loss": 0.0554, "step": 2403 }, { "epoch": 0.16520065970313358, "grad_norm": 1.5455681085586548, "learning_rate": 2.8585313527596485e-05, "loss": 0.0765, "step": 2404 }, { "epoch": 0.1652693787795492, "grad_norm": 2.8005573749542236, "learning_rate": 2.858389781920922e-05, "loss": 0.0765, "step": 2405 }, { "epoch": 0.16533809785596482, "grad_norm": 2.0986459255218506, "learning_rate": 2.8582481437903593e-05, "loss": 0.056, "step": 2406 }, { "epoch": 0.16540681693238043, "grad_norm": 1.5509028434753418, "learning_rate": 2.8581064383749755e-05, "loss": 0.0402, "step": 2407 }, { "epoch": 0.16547553600879605, "grad_norm": 1.2756309509277344, "learning_rate": 2.857964665681791e-05, "loss": 0.0543, "step": 2408 }, { "epoch": 0.16554425508521164, "grad_norm": 1.0515289306640625, "learning_rate": 2.857822825717828e-05, "loss": 0.0571, "step": 2409 }, { "epoch": 0.16561297416162726, "grad_norm": 1.0506945848464966, "learning_rate": 2.8576809184901148e-05, "loss": 0.0479, "step": 2410 }, { "epoch": 0.16568169323804288, "grad_norm": 2.054029703140259, "learning_rate": 2.8575389440056802e-05, "loss": 0.0528, "step": 2411 }, { "epoch": 0.1657504123144585, "grad_norm": 1.9927273988723755, "learning_rate": 2.8573969022715567e-05, "loss": 0.0515, "step": 2412 }, { "epoch": 0.16581913139087412, "grad_norm": 1.7264896631240845, "learning_rate": 2.8572547932947817e-05, "loss": 0.0466, "step": 2413 }, { "epoch": 0.1658878504672897, "grad_norm": 1.9693235158920288, "learning_rate": 2.8571126170823946e-05, "loss": 0.0563, "step": 2414 }, { "epoch": 0.16595656954370533, "grad_norm": 1.5706254243850708, "learning_rate": 2.8569703736414385e-05, "loss": 0.0456, "step": 2415 }, { "epoch": 0.16602528862012095, "grad_norm": 2.009385824203491, "learning_rate": 2.8568280629789597e-05, "loss": 0.0546, "step": 2416 }, { "epoch": 0.16609400769653657, "grad_norm": 2.287937641143799, "learning_rate": 2.8566856851020086e-05, "loss": 0.0487, "step": 2417 }, { "epoch": 0.16616272677295219, "grad_norm": 1.9115254878997803, "learning_rate": 2.8565432400176376e-05, "loss": 0.0662, "step": 2418 }, { "epoch": 0.16623144584936778, "grad_norm": 1.414233922958374, "learning_rate": 2.8564007277329038e-05, "loss": 0.053, "step": 2419 }, { "epoch": 0.1663001649257834, "grad_norm": 1.7478210926055908, "learning_rate": 2.856258148254866e-05, "loss": 0.0791, "step": 2420 }, { "epoch": 0.166368884002199, "grad_norm": 1.3578397035598755, "learning_rate": 2.8561155015905883e-05, "loss": 0.0766, "step": 2421 }, { "epoch": 0.16643760307861463, "grad_norm": 1.7313642501831055, "learning_rate": 2.8559727877471365e-05, "loss": 0.0472, "step": 2422 }, { "epoch": 0.16650632215503022, "grad_norm": 0.6769095659255981, "learning_rate": 2.8558300067315806e-05, "loss": 0.0609, "step": 2423 }, { "epoch": 0.16657504123144584, "grad_norm": 2.0319619178771973, "learning_rate": 2.8556871585509937e-05, "loss": 0.0588, "step": 2424 }, { "epoch": 0.16664376030786146, "grad_norm": 1.2233242988586426, "learning_rate": 2.8555442432124523e-05, "loss": 0.0486, "step": 2425 }, { "epoch": 0.16671247938427708, "grad_norm": 1.50266695022583, "learning_rate": 2.855401260723036e-05, "loss": 0.0578, "step": 2426 }, { "epoch": 0.1667811984606927, "grad_norm": 1.3760615587234497, "learning_rate": 2.8552582110898272e-05, "loss": 0.055, "step": 2427 }, { "epoch": 0.1668499175371083, "grad_norm": 1.7462366819381714, "learning_rate": 2.8551150943199134e-05, "loss": 0.0458, "step": 2428 }, { "epoch": 0.1669186366135239, "grad_norm": 1.3657097816467285, "learning_rate": 2.8549719104203837e-05, "loss": 0.0607, "step": 2429 }, { "epoch": 0.16698735568993953, "grad_norm": 1.175055980682373, "learning_rate": 2.8548286593983312e-05, "loss": 0.0649, "step": 2430 }, { "epoch": 0.16705607476635514, "grad_norm": 1.3494353294372559, "learning_rate": 2.854685341260853e-05, "loss": 0.0457, "step": 2431 }, { "epoch": 0.16712479384277076, "grad_norm": 0.9975899457931519, "learning_rate": 2.8545419560150472e-05, "loss": 0.0441, "step": 2432 }, { "epoch": 0.16719351291918635, "grad_norm": 1.6036614179611206, "learning_rate": 2.854398503668018e-05, "loss": 0.0545, "step": 2433 }, { "epoch": 0.16726223199560197, "grad_norm": 1.2434595823287964, "learning_rate": 2.8542549842268716e-05, "loss": 0.05, "step": 2434 }, { "epoch": 0.1673309510720176, "grad_norm": 1.0802570581436157, "learning_rate": 2.8541113976987177e-05, "loss": 0.0409, "step": 2435 }, { "epoch": 0.1673996701484332, "grad_norm": 1.6062947511672974, "learning_rate": 2.8539677440906688e-05, "loss": 0.0512, "step": 2436 }, { "epoch": 0.16746838922484883, "grad_norm": 1.447535514831543, "learning_rate": 2.8538240234098416e-05, "loss": 0.0597, "step": 2437 }, { "epoch": 0.16753710830126442, "grad_norm": 1.2651633024215698, "learning_rate": 2.8536802356633556e-05, "loss": 0.0516, "step": 2438 }, { "epoch": 0.16760582737768004, "grad_norm": 2.2493977546691895, "learning_rate": 2.8535363808583333e-05, "loss": 0.0601, "step": 2439 }, { "epoch": 0.16767454645409566, "grad_norm": 0.9769768118858337, "learning_rate": 2.853392459001902e-05, "loss": 0.0591, "step": 2440 }, { "epoch": 0.16774326553051128, "grad_norm": 1.8461452722549438, "learning_rate": 2.853248470101191e-05, "loss": 0.0569, "step": 2441 }, { "epoch": 0.1678119846069269, "grad_norm": 1.246350646018982, "learning_rate": 2.8531044141633323e-05, "loss": 0.0679, "step": 2442 }, { "epoch": 0.16788070368334249, "grad_norm": 1.1850991249084473, "learning_rate": 2.8529602911954632e-05, "loss": 0.0662, "step": 2443 }, { "epoch": 0.1679494227597581, "grad_norm": 1.4179813861846924, "learning_rate": 2.8528161012047223e-05, "loss": 0.057, "step": 2444 }, { "epoch": 0.16801814183617372, "grad_norm": 1.2598098516464233, "learning_rate": 2.8526718441982536e-05, "loss": 0.0465, "step": 2445 }, { "epoch": 0.16808686091258934, "grad_norm": 1.2286758422851562, "learning_rate": 2.8525275201832025e-05, "loss": 0.0444, "step": 2446 }, { "epoch": 0.16815557998900496, "grad_norm": 1.5367430448532104, "learning_rate": 2.8523831291667187e-05, "loss": 0.0429, "step": 2447 }, { "epoch": 0.16822429906542055, "grad_norm": 1.6160272359848022, "learning_rate": 2.8522386711559553e-05, "loss": 0.0677, "step": 2448 }, { "epoch": 0.16829301814183617, "grad_norm": 1.288821816444397, "learning_rate": 2.852094146158068e-05, "loss": 0.0496, "step": 2449 }, { "epoch": 0.1683617372182518, "grad_norm": 2.077033758163452, "learning_rate": 2.851949554180216e-05, "loss": 0.0612, "step": 2450 }, { "epoch": 0.1684304562946674, "grad_norm": 1.3554037809371948, "learning_rate": 2.8518048952295634e-05, "loss": 0.0518, "step": 2451 }, { "epoch": 0.16849917537108303, "grad_norm": 1.6270078420639038, "learning_rate": 2.8516601693132756e-05, "loss": 0.0587, "step": 2452 }, { "epoch": 0.16856789444749862, "grad_norm": 2.16736102104187, "learning_rate": 2.851515376438521e-05, "loss": 0.0523, "step": 2453 }, { "epoch": 0.16863661352391424, "grad_norm": 1.0307199954986572, "learning_rate": 2.851370516612474e-05, "loss": 0.0409, "step": 2454 }, { "epoch": 0.16870533260032985, "grad_norm": 1.3237416744232178, "learning_rate": 2.8512255898423095e-05, "loss": 0.0435, "step": 2455 }, { "epoch": 0.16877405167674547, "grad_norm": 1.2728430032730103, "learning_rate": 2.8510805961352077e-05, "loss": 0.0566, "step": 2456 }, { "epoch": 0.16884277075316106, "grad_norm": 1.0377007722854614, "learning_rate": 2.8509355354983504e-05, "loss": 0.0529, "step": 2457 }, { "epoch": 0.16891148982957668, "grad_norm": 2.256324529647827, "learning_rate": 2.8507904079389243e-05, "loss": 0.0561, "step": 2458 }, { "epoch": 0.1689802089059923, "grad_norm": 1.9347243309020996, "learning_rate": 2.8506452134641187e-05, "loss": 0.049, "step": 2459 }, { "epoch": 0.16904892798240792, "grad_norm": 1.6224987506866455, "learning_rate": 2.8504999520811258e-05, "loss": 0.0585, "step": 2460 }, { "epoch": 0.16911764705882354, "grad_norm": 1.4748058319091797, "learning_rate": 2.8503546237971418e-05, "loss": 0.0496, "step": 2461 }, { "epoch": 0.16918636613523913, "grad_norm": 1.9489668607711792, "learning_rate": 2.850209228619366e-05, "loss": 0.0658, "step": 2462 }, { "epoch": 0.16925508521165475, "grad_norm": 1.641030192375183, "learning_rate": 2.850063766555001e-05, "loss": 0.0491, "step": 2463 }, { "epoch": 0.16932380428807037, "grad_norm": 1.1478350162506104, "learning_rate": 2.8499182376112523e-05, "loss": 0.0488, "step": 2464 }, { "epoch": 0.169392523364486, "grad_norm": 1.9606297016143799, "learning_rate": 2.8497726417953294e-05, "loss": 0.046, "step": 2465 }, { "epoch": 0.1694612424409016, "grad_norm": 1.962130069732666, "learning_rate": 2.8496269791144452e-05, "loss": 0.0467, "step": 2466 }, { "epoch": 0.1695299615173172, "grad_norm": 1.78419828414917, "learning_rate": 2.849481249575815e-05, "loss": 0.0471, "step": 2467 }, { "epoch": 0.16959868059373281, "grad_norm": 1.6752864122390747, "learning_rate": 2.8493354531866582e-05, "loss": 0.0555, "step": 2468 }, { "epoch": 0.16966739967014843, "grad_norm": 1.2210732698440552, "learning_rate": 2.849189589954197e-05, "loss": 0.0497, "step": 2469 }, { "epoch": 0.16973611874656405, "grad_norm": 1.4860366582870483, "learning_rate": 2.8490436598856576e-05, "loss": 0.055, "step": 2470 }, { "epoch": 0.16980483782297967, "grad_norm": 1.2764580249786377, "learning_rate": 2.8488976629882687e-05, "loss": 0.0656, "step": 2471 }, { "epoch": 0.16987355689939526, "grad_norm": 1.3965893983840942, "learning_rate": 2.8487515992692625e-05, "loss": 0.0422, "step": 2472 }, { "epoch": 0.16994227597581088, "grad_norm": 1.440246820449829, "learning_rate": 2.8486054687358752e-05, "loss": 0.0625, "step": 2473 }, { "epoch": 0.1700109950522265, "grad_norm": 1.6804922819137573, "learning_rate": 2.8484592713953456e-05, "loss": 0.0604, "step": 2474 }, { "epoch": 0.17007971412864212, "grad_norm": 1.0991102457046509, "learning_rate": 2.848313007254916e-05, "loss": 0.0517, "step": 2475 }, { "epoch": 0.17014843320505774, "grad_norm": 1.404159426689148, "learning_rate": 2.8481666763218318e-05, "loss": 0.0635, "step": 2476 }, { "epoch": 0.17021715228147333, "grad_norm": 1.4509804248809814, "learning_rate": 2.8480202786033427e-05, "loss": 0.0492, "step": 2477 }, { "epoch": 0.17028587135788895, "grad_norm": 1.3641279935836792, "learning_rate": 2.8478738141067e-05, "loss": 0.0509, "step": 2478 }, { "epoch": 0.17035459043430456, "grad_norm": 2.3660314083099365, "learning_rate": 2.84772728283916e-05, "loss": 0.055, "step": 2479 }, { "epoch": 0.17042330951072018, "grad_norm": 1.40339994430542, "learning_rate": 2.847580684807981e-05, "loss": 0.0477, "step": 2480 }, { "epoch": 0.1704920285871358, "grad_norm": 0.905389666557312, "learning_rate": 2.8474340200204252e-05, "loss": 0.0402, "step": 2481 }, { "epoch": 0.1705607476635514, "grad_norm": 1.5058300495147705, "learning_rate": 2.8472872884837586e-05, "loss": 0.0685, "step": 2482 }, { "epoch": 0.170629466739967, "grad_norm": 0.8937322497367859, "learning_rate": 2.8471404902052495e-05, "loss": 0.0352, "step": 2483 }, { "epoch": 0.17069818581638263, "grad_norm": 1.9246511459350586, "learning_rate": 2.84699362519217e-05, "loss": 0.0599, "step": 2484 }, { "epoch": 0.17076690489279825, "grad_norm": 1.1777251958847046, "learning_rate": 2.8468466934517956e-05, "loss": 0.049, "step": 2485 }, { "epoch": 0.17083562396921384, "grad_norm": 1.281383752822876, "learning_rate": 2.8466996949914053e-05, "loss": 0.0423, "step": 2486 }, { "epoch": 0.17090434304562946, "grad_norm": 1.4052027463912964, "learning_rate": 2.8465526298182806e-05, "loss": 0.0563, "step": 2487 }, { "epoch": 0.17097306212204508, "grad_norm": 2.5190680027008057, "learning_rate": 2.8464054979397073e-05, "loss": 0.0504, "step": 2488 }, { "epoch": 0.1710417811984607, "grad_norm": 2.777292490005493, "learning_rate": 2.846258299362973e-05, "loss": 0.0402, "step": 2489 }, { "epoch": 0.17111050027487631, "grad_norm": 1.509009599685669, "learning_rate": 2.846111034095371e-05, "loss": 0.0511, "step": 2490 }, { "epoch": 0.1711792193512919, "grad_norm": 0.8776708245277405, "learning_rate": 2.8459637021441957e-05, "loss": 0.057, "step": 2491 }, { "epoch": 0.17124793842770752, "grad_norm": 1.1101032495498657, "learning_rate": 2.8458163035167456e-05, "loss": 0.0573, "step": 2492 }, { "epoch": 0.17131665750412314, "grad_norm": 1.47970712184906, "learning_rate": 2.845668838220323e-05, "loss": 0.05, "step": 2493 }, { "epoch": 0.17138537658053876, "grad_norm": 0.793101966381073, "learning_rate": 2.8455213062622322e-05, "loss": 0.0462, "step": 2494 }, { "epoch": 0.17145409565695438, "grad_norm": 1.1261537075042725, "learning_rate": 2.845373707649783e-05, "loss": 0.0512, "step": 2495 }, { "epoch": 0.17152281473336997, "grad_norm": 1.5002083778381348, "learning_rate": 2.845226042390286e-05, "loss": 0.0535, "step": 2496 }, { "epoch": 0.1715915338097856, "grad_norm": 1.7616394758224487, "learning_rate": 2.8450783104910562e-05, "loss": 0.0534, "step": 2497 }, { "epoch": 0.1716602528862012, "grad_norm": 0.8341107964515686, "learning_rate": 2.8449305119594123e-05, "loss": 0.0537, "step": 2498 }, { "epoch": 0.17172897196261683, "grad_norm": 1.2910096645355225, "learning_rate": 2.844782646802676e-05, "loss": 0.0446, "step": 2499 }, { "epoch": 0.17179769103903245, "grad_norm": 1.6047402620315552, "learning_rate": 2.8446347150281718e-05, "loss": 0.0449, "step": 2500 }, { "epoch": 0.17186641011544804, "grad_norm": 1.7261555194854736, "learning_rate": 2.844486716643229e-05, "loss": 0.0669, "step": 2501 }, { "epoch": 0.17193512919186366, "grad_norm": 1.7162213325500488, "learning_rate": 2.844338651655178e-05, "loss": 0.0755, "step": 2502 }, { "epoch": 0.17200384826827927, "grad_norm": 1.1509658098220825, "learning_rate": 2.8441905200713537e-05, "loss": 0.05, "step": 2503 }, { "epoch": 0.1720725673446949, "grad_norm": 0.8025435209274292, "learning_rate": 2.8440423218990952e-05, "loss": 0.0499, "step": 2504 }, { "epoch": 0.1721412864211105, "grad_norm": 2.0023601055145264, "learning_rate": 2.843894057145743e-05, "loss": 0.0504, "step": 2505 }, { "epoch": 0.1722100054975261, "grad_norm": 1.5708531141281128, "learning_rate": 2.8437457258186422e-05, "loss": 0.0486, "step": 2506 }, { "epoch": 0.17227872457394172, "grad_norm": 1.7996795177459717, "learning_rate": 2.843597327925141e-05, "loss": 0.0517, "step": 2507 }, { "epoch": 0.17234744365035734, "grad_norm": 2.2775869369506836, "learning_rate": 2.84344886347259e-05, "loss": 0.0542, "step": 2508 }, { "epoch": 0.17241616272677296, "grad_norm": 1.497098445892334, "learning_rate": 2.8433003324683448e-05, "loss": 0.0581, "step": 2509 }, { "epoch": 0.17248488180318858, "grad_norm": 0.9437531232833862, "learning_rate": 2.8431517349197627e-05, "loss": 0.0607, "step": 2510 }, { "epoch": 0.17255360087960417, "grad_norm": 1.6448239088058472, "learning_rate": 2.843003070834205e-05, "loss": 0.0692, "step": 2511 }, { "epoch": 0.1726223199560198, "grad_norm": 1.796706199645996, "learning_rate": 2.8428543402190357e-05, "loss": 0.0414, "step": 2512 }, { "epoch": 0.1726910390324354, "grad_norm": 1.0917974710464478, "learning_rate": 2.842705543081624e-05, "loss": 0.0431, "step": 2513 }, { "epoch": 0.17275975810885102, "grad_norm": 1.0660126209259033, "learning_rate": 2.84255667942934e-05, "loss": 0.041, "step": 2514 }, { "epoch": 0.17282847718526664, "grad_norm": 1.6840413808822632, "learning_rate": 2.842407749269558e-05, "loss": 0.0442, "step": 2515 }, { "epoch": 0.17289719626168223, "grad_norm": 1.338223934173584, "learning_rate": 2.842258752609656e-05, "loss": 0.0618, "step": 2516 }, { "epoch": 0.17296591533809785, "grad_norm": 1.464109182357788, "learning_rate": 2.8421096894570154e-05, "loss": 0.0543, "step": 2517 }, { "epoch": 0.17303463441451347, "grad_norm": 1.9074554443359375, "learning_rate": 2.8419605598190197e-05, "loss": 0.0578, "step": 2518 }, { "epoch": 0.1731033534909291, "grad_norm": 1.393817663192749, "learning_rate": 2.8418113637030566e-05, "loss": 0.0354, "step": 2519 }, { "epoch": 0.17317207256734468, "grad_norm": 1.1391798257827759, "learning_rate": 2.8416621011165177e-05, "loss": 0.1007, "step": 2520 }, { "epoch": 0.1732407916437603, "grad_norm": 2.0665009021759033, "learning_rate": 2.8415127720667963e-05, "loss": 0.0473, "step": 2521 }, { "epoch": 0.17330951072017592, "grad_norm": 2.199066400527954, "learning_rate": 2.8413633765612907e-05, "loss": 0.0637, "step": 2522 }, { "epoch": 0.17337822979659154, "grad_norm": 1.1075739860534668, "learning_rate": 2.8412139146074005e-05, "loss": 0.0467, "step": 2523 }, { "epoch": 0.17344694887300716, "grad_norm": 1.4869062900543213, "learning_rate": 2.8410643862125305e-05, "loss": 0.0478, "step": 2524 }, { "epoch": 0.17351566794942275, "grad_norm": 0.9040416479110718, "learning_rate": 2.8409147913840883e-05, "loss": 0.052, "step": 2525 }, { "epoch": 0.17358438702583837, "grad_norm": 1.9513652324676514, "learning_rate": 2.840765130129484e-05, "loss": 0.0487, "step": 2526 }, { "epoch": 0.17365310610225398, "grad_norm": 2.0972001552581787, "learning_rate": 2.8406154024561316e-05, "loss": 0.0404, "step": 2527 }, { "epoch": 0.1737218251786696, "grad_norm": 0.9768392443656921, "learning_rate": 2.840465608371448e-05, "loss": 0.0446, "step": 2528 }, { "epoch": 0.17379054425508522, "grad_norm": 1.7431731224060059, "learning_rate": 2.8403157478828543e-05, "loss": 0.0571, "step": 2529 }, { "epoch": 0.1738592633315008, "grad_norm": 1.9271160364151, "learning_rate": 2.840165820997774e-05, "loss": 0.0448, "step": 2530 }, { "epoch": 0.17392798240791643, "grad_norm": 2.2114036083221436, "learning_rate": 2.8400158277236345e-05, "loss": 0.0559, "step": 2531 }, { "epoch": 0.17399670148433205, "grad_norm": 1.3325473070144653, "learning_rate": 2.8398657680678653e-05, "loss": 0.0545, "step": 2532 }, { "epoch": 0.17406542056074767, "grad_norm": 0.9459472894668579, "learning_rate": 2.8397156420379012e-05, "loss": 0.0366, "step": 2533 }, { "epoch": 0.1741341396371633, "grad_norm": 1.6811034679412842, "learning_rate": 2.839565449641178e-05, "loss": 0.0525, "step": 2534 }, { "epoch": 0.17420285871357888, "grad_norm": 1.1495102643966675, "learning_rate": 2.839415190885137e-05, "loss": 0.0394, "step": 2535 }, { "epoch": 0.1742715777899945, "grad_norm": 0.9033424258232117, "learning_rate": 2.8392648657772208e-05, "loss": 0.0501, "step": 2536 }, { "epoch": 0.17434029686641012, "grad_norm": 1.328315258026123, "learning_rate": 2.8391144743248765e-05, "loss": 0.0696, "step": 2537 }, { "epoch": 0.17440901594282573, "grad_norm": 0.7063829302787781, "learning_rate": 2.8389640165355545e-05, "loss": 0.0449, "step": 2538 }, { "epoch": 0.17447773501924135, "grad_norm": 0.954876184463501, "learning_rate": 2.8388134924167077e-05, "loss": 0.0345, "step": 2539 }, { "epoch": 0.17454645409565694, "grad_norm": 1.9202711582183838, "learning_rate": 2.8386629019757934e-05, "loss": 0.0571, "step": 2540 }, { "epoch": 0.17461517317207256, "grad_norm": 1.2248748540878296, "learning_rate": 2.8385122452202702e-05, "loss": 0.0648, "step": 2541 }, { "epoch": 0.17468389224848818, "grad_norm": 0.9495046138763428, "learning_rate": 2.8383615221576028e-05, "loss": 0.0408, "step": 2542 }, { "epoch": 0.1747526113249038, "grad_norm": 1.6703985929489136, "learning_rate": 2.8382107327952573e-05, "loss": 0.0586, "step": 2543 }, { "epoch": 0.17482133040131942, "grad_norm": 1.17648446559906, "learning_rate": 2.8380598771407035e-05, "loss": 0.0432, "step": 2544 }, { "epoch": 0.174890049477735, "grad_norm": 1.9053961038589478, "learning_rate": 2.8379089552014137e-05, "loss": 0.0473, "step": 2545 }, { "epoch": 0.17495876855415063, "grad_norm": 1.508416771888733, "learning_rate": 2.8377579669848654e-05, "loss": 0.0445, "step": 2546 }, { "epoch": 0.17502748763056625, "grad_norm": 1.7496038675308228, "learning_rate": 2.8376069124985376e-05, "loss": 0.0611, "step": 2547 }, { "epoch": 0.17509620670698187, "grad_norm": 1.283583164215088, "learning_rate": 2.8374557917499133e-05, "loss": 0.0392, "step": 2548 }, { "epoch": 0.17516492578339748, "grad_norm": 1.2316908836364746, "learning_rate": 2.837304604746479e-05, "loss": 0.0464, "step": 2549 }, { "epoch": 0.17523364485981308, "grad_norm": 1.064763069152832, "learning_rate": 2.8371533514957232e-05, "loss": 0.0367, "step": 2550 }, { "epoch": 0.1753023639362287, "grad_norm": 1.0443154573440552, "learning_rate": 2.83700203200514e-05, "loss": 0.0533, "step": 2551 }, { "epoch": 0.1753710830126443, "grad_norm": 1.5952423810958862, "learning_rate": 2.836850646282225e-05, "loss": 0.051, "step": 2552 }, { "epoch": 0.17543980208905993, "grad_norm": 0.859008252620697, "learning_rate": 2.836699194334477e-05, "loss": 0.0333, "step": 2553 }, { "epoch": 0.17550852116547552, "grad_norm": 1.130289912223816, "learning_rate": 2.8365476761693997e-05, "loss": 0.0519, "step": 2554 }, { "epoch": 0.17557724024189114, "grad_norm": 1.8279552459716797, "learning_rate": 2.836396091794498e-05, "loss": 0.0665, "step": 2555 }, { "epoch": 0.17564595931830676, "grad_norm": 1.2568095922470093, "learning_rate": 2.8362444412172814e-05, "loss": 0.0524, "step": 2556 }, { "epoch": 0.17571467839472238, "grad_norm": 1.8315249681472778, "learning_rate": 2.836092724445262e-05, "loss": 0.0803, "step": 2557 }, { "epoch": 0.175783397471138, "grad_norm": 1.7162638902664185, "learning_rate": 2.8359409414859563e-05, "loss": 0.0582, "step": 2558 }, { "epoch": 0.1758521165475536, "grad_norm": 2.049221992492676, "learning_rate": 2.8357890923468835e-05, "loss": 0.0501, "step": 2559 }, { "epoch": 0.1759208356239692, "grad_norm": 2.617950677871704, "learning_rate": 2.8356371770355645e-05, "loss": 0.0515, "step": 2560 }, { "epoch": 0.17598955470038483, "grad_norm": 1.9745759963989258, "learning_rate": 2.835485195559526e-05, "loss": 0.0697, "step": 2561 }, { "epoch": 0.17605827377680044, "grad_norm": 1.2252075672149658, "learning_rate": 2.8353331479262968e-05, "loss": 0.0635, "step": 2562 }, { "epoch": 0.17612699285321606, "grad_norm": 1.2563434839248657, "learning_rate": 2.8351810341434084e-05, "loss": 0.066, "step": 2563 }, { "epoch": 0.17619571192963165, "grad_norm": 1.1108770370483398, "learning_rate": 2.8350288542183966e-05, "loss": 0.0714, "step": 2564 }, { "epoch": 0.17626443100604727, "grad_norm": 1.5348811149597168, "learning_rate": 2.8348766081588e-05, "loss": 0.0468, "step": 2565 }, { "epoch": 0.1763331500824629, "grad_norm": 1.3077057600021362, "learning_rate": 2.834724295972161e-05, "loss": 0.0645, "step": 2566 }, { "epoch": 0.1764018691588785, "grad_norm": 1.4912943840026855, "learning_rate": 2.834571917666024e-05, "loss": 0.05, "step": 2567 }, { "epoch": 0.17647058823529413, "grad_norm": 1.0611827373504639, "learning_rate": 2.8344194732479384e-05, "loss": 0.0487, "step": 2568 }, { "epoch": 0.17653930731170972, "grad_norm": 1.3873305320739746, "learning_rate": 2.8342669627254553e-05, "loss": 0.054, "step": 2569 }, { "epoch": 0.17660802638812534, "grad_norm": 2.016460418701172, "learning_rate": 2.83411438610613e-05, "loss": 0.0557, "step": 2570 }, { "epoch": 0.17667674546454096, "grad_norm": 0.9113234877586365, "learning_rate": 2.833961743397521e-05, "loss": 0.0463, "step": 2571 }, { "epoch": 0.17674546454095658, "grad_norm": 1.5234220027923584, "learning_rate": 2.8338090346071894e-05, "loss": 0.068, "step": 2572 }, { "epoch": 0.1768141836173722, "grad_norm": 1.1417694091796875, "learning_rate": 2.833656259742701e-05, "loss": 0.0466, "step": 2573 }, { "epoch": 0.17688290269378779, "grad_norm": 1.4648734331130981, "learning_rate": 2.8335034188116228e-05, "loss": 0.0458, "step": 2574 }, { "epoch": 0.1769516217702034, "grad_norm": 1.645939588546753, "learning_rate": 2.8333505118215272e-05, "loss": 0.0501, "step": 2575 }, { "epoch": 0.17702034084661902, "grad_norm": 2.5709328651428223, "learning_rate": 2.833197538779988e-05, "loss": 0.0625, "step": 2576 }, { "epoch": 0.17708905992303464, "grad_norm": 1.1942400932312012, "learning_rate": 2.8330444996945842e-05, "loss": 0.05, "step": 2577 }, { "epoch": 0.17715777899945026, "grad_norm": 1.4276776313781738, "learning_rate": 2.8328913945728963e-05, "loss": 0.0471, "step": 2578 }, { "epoch": 0.17722649807586585, "grad_norm": 1.463727593421936, "learning_rate": 2.832738223422509e-05, "loss": 0.0566, "step": 2579 }, { "epoch": 0.17729521715228147, "grad_norm": 1.136858582496643, "learning_rate": 2.8325849862510102e-05, "loss": 0.0417, "step": 2580 }, { "epoch": 0.1773639362286971, "grad_norm": 1.3173140287399292, "learning_rate": 2.8324316830659905e-05, "loss": 0.0497, "step": 2581 }, { "epoch": 0.1774326553051127, "grad_norm": 0.8611404299736023, "learning_rate": 2.8322783138750447e-05, "loss": 0.0453, "step": 2582 }, { "epoch": 0.1775013743815283, "grad_norm": 2.218585252761841, "learning_rate": 2.8321248786857708e-05, "loss": 0.0378, "step": 2583 }, { "epoch": 0.17757009345794392, "grad_norm": 1.1088247299194336, "learning_rate": 2.8319713775057686e-05, "loss": 0.0389, "step": 2584 }, { "epoch": 0.17763881253435954, "grad_norm": 2.3781955242156982, "learning_rate": 2.831817810342643e-05, "loss": 0.0684, "step": 2585 }, { "epoch": 0.17770753161077515, "grad_norm": 1.7618805170059204, "learning_rate": 2.831664177204001e-05, "loss": 0.0611, "step": 2586 }, { "epoch": 0.17777625068719077, "grad_norm": 1.686086654663086, "learning_rate": 2.8315104780974535e-05, "loss": 0.078, "step": 2587 }, { "epoch": 0.17784496976360636, "grad_norm": 1.1207207441329956, "learning_rate": 2.831356713030615e-05, "loss": 0.0527, "step": 2588 }, { "epoch": 0.17791368884002198, "grad_norm": 1.8792016506195068, "learning_rate": 2.8312028820111014e-05, "loss": 0.0478, "step": 2589 }, { "epoch": 0.1779824079164376, "grad_norm": 1.2607672214508057, "learning_rate": 2.8310489850465338e-05, "loss": 0.0491, "step": 2590 }, { "epoch": 0.17805112699285322, "grad_norm": 1.9327634572982788, "learning_rate": 2.8308950221445362e-05, "loss": 0.0541, "step": 2591 }, { "epoch": 0.17811984606926884, "grad_norm": 0.7585175037384033, "learning_rate": 2.830740993312736e-05, "loss": 0.0436, "step": 2592 }, { "epoch": 0.17818856514568443, "grad_norm": 1.3587626218795776, "learning_rate": 2.830586898558762e-05, "loss": 0.0588, "step": 2593 }, { "epoch": 0.17825728422210005, "grad_norm": 1.5768011808395386, "learning_rate": 2.830432737890249e-05, "loss": 0.0437, "step": 2594 }, { "epoch": 0.17832600329851567, "grad_norm": 0.7820532917976379, "learning_rate": 2.8302785113148333e-05, "loss": 0.0401, "step": 2595 }, { "epoch": 0.17839472237493129, "grad_norm": 1.6546741724014282, "learning_rate": 2.8301242188401548e-05, "loss": 0.0477, "step": 2596 }, { "epoch": 0.1784634414513469, "grad_norm": 1.006256103515625, "learning_rate": 2.8299698604738577e-05, "loss": 0.0559, "step": 2597 }, { "epoch": 0.1785321605277625, "grad_norm": 0.9783194661140442, "learning_rate": 2.829815436223588e-05, "loss": 0.0526, "step": 2598 }, { "epoch": 0.17860087960417811, "grad_norm": 1.113821029663086, "learning_rate": 2.829660946096995e-05, "loss": 0.0402, "step": 2599 }, { "epoch": 0.17866959868059373, "grad_norm": 1.7552250623703003, "learning_rate": 2.8295063901017327e-05, "loss": 0.0352, "step": 2600 }, { "epoch": 0.17873831775700935, "grad_norm": 1.1229267120361328, "learning_rate": 2.8293517682454574e-05, "loss": 0.0284, "step": 2601 }, { "epoch": 0.17880703683342497, "grad_norm": 1.2388285398483276, "learning_rate": 2.8291970805358283e-05, "loss": 0.0517, "step": 2602 }, { "epoch": 0.17887575590984056, "grad_norm": 1.590406060218811, "learning_rate": 2.829042326980509e-05, "loss": 0.0437, "step": 2603 }, { "epoch": 0.17894447498625618, "grad_norm": 1.6707704067230225, "learning_rate": 2.8288875075871642e-05, "loss": 0.0452, "step": 2604 }, { "epoch": 0.1790131940626718, "grad_norm": 1.0263298749923706, "learning_rate": 2.8287326223634652e-05, "loss": 0.0374, "step": 2605 }, { "epoch": 0.17908191313908742, "grad_norm": 1.459311604499817, "learning_rate": 2.8285776713170836e-05, "loss": 0.0514, "step": 2606 }, { "epoch": 0.17915063221550304, "grad_norm": 1.4643193483352661, "learning_rate": 2.8284226544556957e-05, "loss": 0.0447, "step": 2607 }, { "epoch": 0.17921935129191863, "grad_norm": 1.1149705648422241, "learning_rate": 2.8282675717869806e-05, "loss": 0.0502, "step": 2608 }, { "epoch": 0.17928807036833425, "grad_norm": 1.8917876482009888, "learning_rate": 2.8281124233186204e-05, "loss": 0.0515, "step": 2609 }, { "epoch": 0.17935678944474986, "grad_norm": 2.7621474266052246, "learning_rate": 2.8279572090583017e-05, "loss": 0.0683, "step": 2610 }, { "epoch": 0.17942550852116548, "grad_norm": 1.1737645864486694, "learning_rate": 2.8278019290137127e-05, "loss": 0.0445, "step": 2611 }, { "epoch": 0.1794942275975811, "grad_norm": 1.6877175569534302, "learning_rate": 2.8276465831925464e-05, "loss": 0.0587, "step": 2612 }, { "epoch": 0.1795629466739967, "grad_norm": 1.9697258472442627, "learning_rate": 2.8274911716024976e-05, "loss": 0.0603, "step": 2613 }, { "epoch": 0.1796316657504123, "grad_norm": 1.1457939147949219, "learning_rate": 2.8273356942512655e-05, "loss": 0.053, "step": 2614 }, { "epoch": 0.17970038482682793, "grad_norm": 1.4554694890975952, "learning_rate": 2.827180151146552e-05, "loss": 0.0615, "step": 2615 }, { "epoch": 0.17976910390324355, "grad_norm": 1.7136398553848267, "learning_rate": 2.8270245422960622e-05, "loss": 0.0546, "step": 2616 }, { "epoch": 0.17983782297965914, "grad_norm": 1.5277879238128662, "learning_rate": 2.8268688677075048e-05, "loss": 0.0506, "step": 2617 }, { "epoch": 0.17990654205607476, "grad_norm": 0.9457663893699646, "learning_rate": 2.826713127388592e-05, "loss": 0.052, "step": 2618 }, { "epoch": 0.17997526113249038, "grad_norm": 2.2268404960632324, "learning_rate": 2.826557321347038e-05, "loss": 0.0414, "step": 2619 }, { "epoch": 0.180043980208906, "grad_norm": 1.852324366569519, "learning_rate": 2.8264014495905618e-05, "loss": 0.0545, "step": 2620 }, { "epoch": 0.18011269928532161, "grad_norm": 1.4424904584884644, "learning_rate": 2.8262455121268846e-05, "loss": 0.0678, "step": 2621 }, { "epoch": 0.1801814183617372, "grad_norm": 1.2051998376846313, "learning_rate": 2.8260895089637315e-05, "loss": 0.053, "step": 2622 }, { "epoch": 0.18025013743815282, "grad_norm": 1.5711363554000854, "learning_rate": 2.8259334401088307e-05, "loss": 0.0506, "step": 2623 }, { "epoch": 0.18031885651456844, "grad_norm": 1.2579636573791504, "learning_rate": 2.825777305569913e-05, "loss": 0.0739, "step": 2624 }, { "epoch": 0.18038757559098406, "grad_norm": 1.2767177820205688, "learning_rate": 2.8256211053547135e-05, "loss": 0.048, "step": 2625 }, { "epoch": 0.18045629466739968, "grad_norm": 1.29844069480896, "learning_rate": 2.8254648394709696e-05, "loss": 0.0529, "step": 2626 }, { "epoch": 0.18052501374381527, "grad_norm": 1.641297459602356, "learning_rate": 2.8253085079264226e-05, "loss": 0.0494, "step": 2627 }, { "epoch": 0.1805937328202309, "grad_norm": 1.2294716835021973, "learning_rate": 2.8251521107288167e-05, "loss": 0.0443, "step": 2628 }, { "epoch": 0.1806624518966465, "grad_norm": 1.3203009366989136, "learning_rate": 2.8249956478859e-05, "loss": 0.0634, "step": 2629 }, { "epoch": 0.18073117097306213, "grad_norm": 1.6198033094406128, "learning_rate": 2.8248391194054226e-05, "loss": 0.0568, "step": 2630 }, { "epoch": 0.18079989004947775, "grad_norm": 1.6927224397659302, "learning_rate": 2.8246825252951394e-05, "loss": 0.0512, "step": 2631 }, { "epoch": 0.18086860912589334, "grad_norm": 0.9959924221038818, "learning_rate": 2.824525865562807e-05, "loss": 0.0437, "step": 2632 }, { "epoch": 0.18093732820230896, "grad_norm": 1.8606584072113037, "learning_rate": 2.8243691402161866e-05, "loss": 0.0469, "step": 2633 }, { "epoch": 0.18100604727872457, "grad_norm": 1.4416385889053345, "learning_rate": 2.8242123492630413e-05, "loss": 0.0465, "step": 2634 }, { "epoch": 0.1810747663551402, "grad_norm": 1.1384180784225464, "learning_rate": 2.824055492711139e-05, "loss": 0.0375, "step": 2635 }, { "epoch": 0.1811434854315558, "grad_norm": 1.284522294998169, "learning_rate": 2.82389857056825e-05, "loss": 0.031, "step": 2636 }, { "epoch": 0.1812122045079714, "grad_norm": 2.322082996368408, "learning_rate": 2.823741582842147e-05, "loss": 0.0482, "step": 2637 }, { "epoch": 0.18128092358438702, "grad_norm": 1.7410976886749268, "learning_rate": 2.823584529540608e-05, "loss": 0.0498, "step": 2638 }, { "epoch": 0.18134964266080264, "grad_norm": 1.0764325857162476, "learning_rate": 2.823427410671412e-05, "loss": 0.0426, "step": 2639 }, { "epoch": 0.18141836173721826, "grad_norm": 0.8845528960227966, "learning_rate": 2.8232702262423435e-05, "loss": 0.0383, "step": 2640 }, { "epoch": 0.18148708081363388, "grad_norm": 1.414707064628601, "learning_rate": 2.8231129762611885e-05, "loss": 0.0532, "step": 2641 }, { "epoch": 0.18155579989004947, "grad_norm": 1.507010579109192, "learning_rate": 2.8229556607357366e-05, "loss": 0.046, "step": 2642 }, { "epoch": 0.1816245189664651, "grad_norm": 0.8197042346000671, "learning_rate": 2.8227982796737814e-05, "loss": 0.0599, "step": 2643 }, { "epoch": 0.1816932380428807, "grad_norm": 1.12967848777771, "learning_rate": 2.8226408330831188e-05, "loss": 0.0442, "step": 2644 }, { "epoch": 0.18176195711929632, "grad_norm": 1.1332042217254639, "learning_rate": 2.8224833209715486e-05, "loss": 0.0447, "step": 2645 }, { "epoch": 0.18183067619571194, "grad_norm": 1.4914582967758179, "learning_rate": 2.8223257433468735e-05, "loss": 0.0418, "step": 2646 }, { "epoch": 0.18189939527212753, "grad_norm": 0.9489706158638, "learning_rate": 2.8221681002169e-05, "loss": 0.0641, "step": 2647 }, { "epoch": 0.18196811434854315, "grad_norm": 0.999434232711792, "learning_rate": 2.8220103915894367e-05, "loss": 0.0506, "step": 2648 }, { "epoch": 0.18203683342495877, "grad_norm": 3.6180481910705566, "learning_rate": 2.8218526174722967e-05, "loss": 0.0537, "step": 2649 }, { "epoch": 0.1821055525013744, "grad_norm": 1.146721363067627, "learning_rate": 2.8216947778732953e-05, "loss": 0.0514, "step": 2650 }, { "epoch": 0.18217427157778998, "grad_norm": 0.9135676026344299, "learning_rate": 2.8215368728002524e-05, "loss": 0.0501, "step": 2651 }, { "epoch": 0.1822429906542056, "grad_norm": 1.106876254081726, "learning_rate": 2.8213789022609895e-05, "loss": 0.0425, "step": 2652 }, { "epoch": 0.18231170973062122, "grad_norm": 1.3161389827728271, "learning_rate": 2.8212208662633325e-05, "loss": 0.054, "step": 2653 }, { "epoch": 0.18238042880703684, "grad_norm": 1.2089182138442993, "learning_rate": 2.8210627648151103e-05, "loss": 0.0379, "step": 2654 }, { "epoch": 0.18244914788345246, "grad_norm": 1.0725709199905396, "learning_rate": 2.8209045979241546e-05, "loss": 0.0563, "step": 2655 }, { "epoch": 0.18251786695986805, "grad_norm": 1.3193258047103882, "learning_rate": 2.8207463655983005e-05, "loss": 0.0441, "step": 2656 }, { "epoch": 0.18258658603628367, "grad_norm": 1.7478872537612915, "learning_rate": 2.8205880678453874e-05, "loss": 0.0479, "step": 2657 }, { "epoch": 0.18265530511269928, "grad_norm": 0.8594749569892883, "learning_rate": 2.8204297046732558e-05, "loss": 0.0366, "step": 2658 }, { "epoch": 0.1827240241891149, "grad_norm": 1.4860345125198364, "learning_rate": 2.8202712760897516e-05, "loss": 0.0397, "step": 2659 }, { "epoch": 0.18279274326553052, "grad_norm": 1.1433123350143433, "learning_rate": 2.8201127821027227e-05, "loss": 0.0481, "step": 2660 }, { "epoch": 0.1828614623419461, "grad_norm": 1.3600730895996094, "learning_rate": 2.8199542227200205e-05, "loss": 0.0446, "step": 2661 }, { "epoch": 0.18293018141836173, "grad_norm": 1.551164984703064, "learning_rate": 2.8197955979495002e-05, "loss": 0.039, "step": 2662 }, { "epoch": 0.18299890049477735, "grad_norm": 1.2206683158874512, "learning_rate": 2.819636907799019e-05, "loss": 0.0487, "step": 2663 }, { "epoch": 0.18306761957119297, "grad_norm": 2.13161563873291, "learning_rate": 2.8194781522764382e-05, "loss": 0.051, "step": 2664 }, { "epoch": 0.1831363386476086, "grad_norm": 4.438136100769043, "learning_rate": 2.819319331389623e-05, "loss": 0.0484, "step": 2665 }, { "epoch": 0.18320505772402418, "grad_norm": 1.3154503107070923, "learning_rate": 2.8191604451464404e-05, "loss": 0.0532, "step": 2666 }, { "epoch": 0.1832737768004398, "grad_norm": 1.6589539051055908, "learning_rate": 2.8190014935547617e-05, "loss": 0.0853, "step": 2667 }, { "epoch": 0.18334249587685542, "grad_norm": 1.0012195110321045, "learning_rate": 2.81884247662246e-05, "loss": 0.0535, "step": 2668 }, { "epoch": 0.18341121495327103, "grad_norm": 2.2305760383605957, "learning_rate": 2.818683394357414e-05, "loss": 0.0653, "step": 2669 }, { "epoch": 0.18347993402968665, "grad_norm": 1.2993673086166382, "learning_rate": 2.818524246767504e-05, "loss": 0.0499, "step": 2670 }, { "epoch": 0.18354865310610224, "grad_norm": 1.0460138320922852, "learning_rate": 2.818365033860613e-05, "loss": 0.0569, "step": 2671 }, { "epoch": 0.18361737218251786, "grad_norm": 0.9497994184494019, "learning_rate": 2.818205755644629e-05, "loss": 0.0532, "step": 2672 }, { "epoch": 0.18368609125893348, "grad_norm": 1.857475757598877, "learning_rate": 2.8180464121274414e-05, "loss": 0.047, "step": 2673 }, { "epoch": 0.1837548103353491, "grad_norm": 0.8099668622016907, "learning_rate": 2.817887003316945e-05, "loss": 0.0409, "step": 2674 }, { "epoch": 0.18382352941176472, "grad_norm": 1.1752910614013672, "learning_rate": 2.8177275292210355e-05, "loss": 0.035, "step": 2675 }, { "epoch": 0.1838922484881803, "grad_norm": 1.0167908668518066, "learning_rate": 2.817567989847614e-05, "loss": 0.0361, "step": 2676 }, { "epoch": 0.18396096756459593, "grad_norm": 0.9725064039230347, "learning_rate": 2.8174083852045822e-05, "loss": 0.0316, "step": 2677 }, { "epoch": 0.18402968664101155, "grad_norm": 1.0772124528884888, "learning_rate": 2.817248715299848e-05, "loss": 0.0556, "step": 2678 }, { "epoch": 0.18409840571742717, "grad_norm": 1.4373725652694702, "learning_rate": 2.8170889801413208e-05, "loss": 0.055, "step": 2679 }, { "epoch": 0.18416712479384276, "grad_norm": 1.7471964359283447, "learning_rate": 2.8169291797369126e-05, "loss": 0.0424, "step": 2680 }, { "epoch": 0.18423584387025838, "grad_norm": 1.3210610151290894, "learning_rate": 2.816769314094541e-05, "loss": 0.0628, "step": 2681 }, { "epoch": 0.184304562946674, "grad_norm": 1.2861719131469727, "learning_rate": 2.8166093832221243e-05, "loss": 0.0508, "step": 2682 }, { "epoch": 0.1843732820230896, "grad_norm": 1.6732224225997925, "learning_rate": 2.816449387127586e-05, "loss": 0.0419, "step": 2683 }, { "epoch": 0.18444200109950523, "grad_norm": 1.5948072671890259, "learning_rate": 2.8162893258188512e-05, "loss": 0.0471, "step": 2684 }, { "epoch": 0.18451072017592082, "grad_norm": 1.1870203018188477, "learning_rate": 2.8161291993038496e-05, "loss": 0.0434, "step": 2685 }, { "epoch": 0.18457943925233644, "grad_norm": 1.4853284358978271, "learning_rate": 2.815969007590513e-05, "loss": 0.0437, "step": 2686 }, { "epoch": 0.18464815832875206, "grad_norm": 1.4348610639572144, "learning_rate": 2.8158087506867777e-05, "loss": 0.0547, "step": 2687 }, { "epoch": 0.18471687740516768, "grad_norm": 1.484602451324463, "learning_rate": 2.8156484286005818e-05, "loss": 0.0434, "step": 2688 }, { "epoch": 0.1847855964815833, "grad_norm": 1.1556872129440308, "learning_rate": 2.8154880413398674e-05, "loss": 0.0412, "step": 2689 }, { "epoch": 0.1848543155579989, "grad_norm": 3.3690853118896484, "learning_rate": 2.8153275889125804e-05, "loss": 0.0823, "step": 2690 }, { "epoch": 0.1849230346344145, "grad_norm": 1.4813392162322998, "learning_rate": 2.815167071326669e-05, "loss": 0.0458, "step": 2691 }, { "epoch": 0.18499175371083013, "grad_norm": 1.4152870178222656, "learning_rate": 2.8150064885900845e-05, "loss": 0.0636, "step": 2692 }, { "epoch": 0.18506047278724574, "grad_norm": 1.0994144678115845, "learning_rate": 2.8148458407107817e-05, "loss": 0.0411, "step": 2693 }, { "epoch": 0.18512919186366136, "grad_norm": 1.6438157558441162, "learning_rate": 2.8146851276967196e-05, "loss": 0.0564, "step": 2694 }, { "epoch": 0.18519791094007695, "grad_norm": 0.5090625286102295, "learning_rate": 2.8145243495558595e-05, "loss": 0.0387, "step": 2695 }, { "epoch": 0.18526663001649257, "grad_norm": 1.889111876487732, "learning_rate": 2.814363506296165e-05, "loss": 0.0538, "step": 2696 }, { "epoch": 0.1853353490929082, "grad_norm": 1.521178126335144, "learning_rate": 2.8142025979256048e-05, "loss": 0.0513, "step": 2697 }, { "epoch": 0.1854040681693238, "grad_norm": 1.379501461982727, "learning_rate": 2.8140416244521498e-05, "loss": 0.0386, "step": 2698 }, { "epoch": 0.18547278724573943, "grad_norm": 1.4741566181182861, "learning_rate": 2.8138805858837747e-05, "loss": 0.0405, "step": 2699 }, { "epoch": 0.18554150632215502, "grad_norm": 3.0834052562713623, "learning_rate": 2.8137194822284558e-05, "loss": 0.0639, "step": 2700 }, { "epoch": 0.18561022539857064, "grad_norm": 1.3211743831634521, "learning_rate": 2.8135583134941753e-05, "loss": 0.0538, "step": 2701 }, { "epoch": 0.18567894447498626, "grad_norm": 1.5536932945251465, "learning_rate": 2.8133970796889163e-05, "loss": 0.0498, "step": 2702 }, { "epoch": 0.18574766355140188, "grad_norm": 2.4124433994293213, "learning_rate": 2.813235780820666e-05, "loss": 0.0669, "step": 2703 }, { "epoch": 0.1858163826278175, "grad_norm": 0.8998525142669678, "learning_rate": 2.8130744168974154e-05, "loss": 0.0491, "step": 2704 }, { "epoch": 0.18588510170423309, "grad_norm": 1.452905297279358, "learning_rate": 2.8129129879271574e-05, "loss": 0.0422, "step": 2705 }, { "epoch": 0.1859538207806487, "grad_norm": 1.7121726274490356, "learning_rate": 2.812751493917889e-05, "loss": 0.0531, "step": 2706 }, { "epoch": 0.18602253985706432, "grad_norm": 0.9940103888511658, "learning_rate": 2.812589934877611e-05, "loss": 0.0424, "step": 2707 }, { "epoch": 0.18609125893347994, "grad_norm": 1.007059931755066, "learning_rate": 2.8124283108143262e-05, "loss": 0.0348, "step": 2708 }, { "epoch": 0.18615997800989556, "grad_norm": 1.61087167263031, "learning_rate": 2.812266621736041e-05, "loss": 0.0536, "step": 2709 }, { "epoch": 0.18622869708631115, "grad_norm": 1.4888988733291626, "learning_rate": 2.8121048676507653e-05, "loss": 0.0521, "step": 2710 }, { "epoch": 0.18629741616272677, "grad_norm": 1.155738353729248, "learning_rate": 2.811943048566512e-05, "loss": 0.0563, "step": 2711 }, { "epoch": 0.1863661352391424, "grad_norm": 1.340706467628479, "learning_rate": 2.8117811644912973e-05, "loss": 0.0565, "step": 2712 }, { "epoch": 0.186434854315558, "grad_norm": 0.9261972308158875, "learning_rate": 2.8116192154331404e-05, "loss": 0.0402, "step": 2713 }, { "epoch": 0.1865035733919736, "grad_norm": 1.4762072563171387, "learning_rate": 2.8114572014000645e-05, "loss": 0.0499, "step": 2714 }, { "epoch": 0.18657229246838922, "grad_norm": 2.255460262298584, "learning_rate": 2.811295122400095e-05, "loss": 0.0673, "step": 2715 }, { "epoch": 0.18664101154480484, "grad_norm": 1.5622743368148804, "learning_rate": 2.8111329784412607e-05, "loss": 0.0668, "step": 2716 }, { "epoch": 0.18670973062122045, "grad_norm": 1.7251051664352417, "learning_rate": 2.8109707695315945e-05, "loss": 0.0479, "step": 2717 }, { "epoch": 0.18677844969763607, "grad_norm": 1.3139456510543823, "learning_rate": 2.8108084956791316e-05, "loss": 0.045, "step": 2718 }, { "epoch": 0.18684716877405166, "grad_norm": 0.9839609861373901, "learning_rate": 2.8106461568919107e-05, "loss": 0.0562, "step": 2719 }, { "epoch": 0.18691588785046728, "grad_norm": 1.4574573040008545, "learning_rate": 2.810483753177974e-05, "loss": 0.0528, "step": 2720 }, { "epoch": 0.1869846069268829, "grad_norm": 1.5961198806762695, "learning_rate": 2.8103212845453663e-05, "loss": 0.0432, "step": 2721 }, { "epoch": 0.18705332600329852, "grad_norm": 1.495302438735962, "learning_rate": 2.810158751002136e-05, "loss": 0.0664, "step": 2722 }, { "epoch": 0.18712204507971414, "grad_norm": 0.8685864210128784, "learning_rate": 2.8099961525563346e-05, "loss": 0.0593, "step": 2723 }, { "epoch": 0.18719076415612973, "grad_norm": 1.2623121738433838, "learning_rate": 2.8098334892160174e-05, "loss": 0.0586, "step": 2724 }, { "epoch": 0.18725948323254535, "grad_norm": 1.387250304222107, "learning_rate": 2.8096707609892417e-05, "loss": 0.0561, "step": 2725 }, { "epoch": 0.18732820230896097, "grad_norm": 1.5538474321365356, "learning_rate": 2.809507967884069e-05, "loss": 0.0495, "step": 2726 }, { "epoch": 0.18739692138537659, "grad_norm": 0.9423860907554626, "learning_rate": 2.8093451099085644e-05, "loss": 0.0537, "step": 2727 }, { "epoch": 0.1874656404617922, "grad_norm": 0.9924835562705994, "learning_rate": 2.809182187070794e-05, "loss": 0.0525, "step": 2728 }, { "epoch": 0.1875343595382078, "grad_norm": 4.148291110992432, "learning_rate": 2.8090191993788307e-05, "loss": 0.0558, "step": 2729 }, { "epoch": 0.18760307861462341, "grad_norm": 1.8159362077713013, "learning_rate": 2.8088561468407465e-05, "loss": 0.0579, "step": 2730 }, { "epoch": 0.18767179769103903, "grad_norm": 1.1343436241149902, "learning_rate": 2.8086930294646204e-05, "loss": 0.0545, "step": 2731 }, { "epoch": 0.18774051676745465, "grad_norm": 3.5014989376068115, "learning_rate": 2.8085298472585317e-05, "loss": 0.0714, "step": 2732 }, { "epoch": 0.18780923584387027, "grad_norm": 2.5006141662597656, "learning_rate": 2.808366600230565e-05, "loss": 0.0715, "step": 2733 }, { "epoch": 0.18787795492028586, "grad_norm": 2.27016019821167, "learning_rate": 2.8082032883888063e-05, "loss": 0.0552, "step": 2734 }, { "epoch": 0.18794667399670148, "grad_norm": 1.346206784248352, "learning_rate": 2.808039911741347e-05, "loss": 0.0579, "step": 2735 }, { "epoch": 0.1880153930731171, "grad_norm": 1.726807951927185, "learning_rate": 2.807876470296279e-05, "loss": 0.0498, "step": 2736 }, { "epoch": 0.18808411214953272, "grad_norm": 1.9963957071304321, "learning_rate": 2.8077129640617e-05, "loss": 0.0681, "step": 2737 }, { "epoch": 0.18815283122594834, "grad_norm": 0.9310622215270996, "learning_rate": 2.8075493930457087e-05, "loss": 0.0509, "step": 2738 }, { "epoch": 0.18822155030236393, "grad_norm": 1.5692540407180786, "learning_rate": 2.8073857572564096e-05, "loss": 0.0435, "step": 2739 }, { "epoch": 0.18829026937877955, "grad_norm": 1.084006428718567, "learning_rate": 2.807222056701907e-05, "loss": 0.0392, "step": 2740 }, { "epoch": 0.18835898845519516, "grad_norm": 1.4301722049713135, "learning_rate": 2.807058291390312e-05, "loss": 0.0415, "step": 2741 }, { "epoch": 0.18842770753161078, "grad_norm": 2.738844633102417, "learning_rate": 2.8068944613297362e-05, "loss": 0.0538, "step": 2742 }, { "epoch": 0.1884964266080264, "grad_norm": 1.4317429065704346, "learning_rate": 2.8067305665282955e-05, "loss": 0.0362, "step": 2743 }, { "epoch": 0.188565145684442, "grad_norm": 1.65792977809906, "learning_rate": 2.806566606994109e-05, "loss": 0.0457, "step": 2744 }, { "epoch": 0.1886338647608576, "grad_norm": 0.9972598552703857, "learning_rate": 2.806402582735299e-05, "loss": 0.0382, "step": 2745 }, { "epoch": 0.18870258383727323, "grad_norm": 0.776827871799469, "learning_rate": 2.806238493759991e-05, "loss": 0.0395, "step": 2746 }, { "epoch": 0.18877130291368885, "grad_norm": 1.1877135038375854, "learning_rate": 2.8060743400763137e-05, "loss": 0.0274, "step": 2747 }, { "epoch": 0.18884002199010444, "grad_norm": 1.6914622783660889, "learning_rate": 2.8059101216923986e-05, "loss": 0.0533, "step": 2748 }, { "epoch": 0.18890874106652006, "grad_norm": 1.0999021530151367, "learning_rate": 2.8057458386163807e-05, "loss": 0.0384, "step": 2749 }, { "epoch": 0.18897746014293568, "grad_norm": 1.0830916166305542, "learning_rate": 2.805581490856399e-05, "loss": 0.0343, "step": 2750 }, { "epoch": 0.1890461792193513, "grad_norm": 1.2732949256896973, "learning_rate": 2.805417078420594e-05, "loss": 0.0414, "step": 2751 }, { "epoch": 0.18911489829576691, "grad_norm": 2.114631175994873, "learning_rate": 2.805252601317111e-05, "loss": 0.0739, "step": 2752 }, { "epoch": 0.1891836173721825, "grad_norm": 1.5732239484786987, "learning_rate": 2.8050880595540972e-05, "loss": 0.0513, "step": 2753 }, { "epoch": 0.18925233644859812, "grad_norm": 1.373989462852478, "learning_rate": 2.8049234531397045e-05, "loss": 0.0585, "step": 2754 }, { "epoch": 0.18932105552501374, "grad_norm": 1.5756549835205078, "learning_rate": 2.8047587820820866e-05, "loss": 0.0398, "step": 2755 }, { "epoch": 0.18938977460142936, "grad_norm": 1.7651283740997314, "learning_rate": 2.804594046389401e-05, "loss": 0.051, "step": 2756 }, { "epoch": 0.18945849367784498, "grad_norm": 1.4264687299728394, "learning_rate": 2.8044292460698083e-05, "loss": 0.0446, "step": 2757 }, { "epoch": 0.18952721275426057, "grad_norm": 0.9207376837730408, "learning_rate": 2.804264381131473e-05, "loss": 0.043, "step": 2758 }, { "epoch": 0.1895959318306762, "grad_norm": 1.1748982667922974, "learning_rate": 2.8040994515825614e-05, "loss": 0.0563, "step": 2759 }, { "epoch": 0.1896646509070918, "grad_norm": 2.5341858863830566, "learning_rate": 2.803934457431244e-05, "loss": 0.0519, "step": 2760 }, { "epoch": 0.18973336998350743, "grad_norm": 1.5277857780456543, "learning_rate": 2.8037693986856947e-05, "loss": 0.0552, "step": 2761 }, { "epoch": 0.18980208905992305, "grad_norm": 1.6985763311386108, "learning_rate": 2.80360427535409e-05, "loss": 0.0473, "step": 2762 }, { "epoch": 0.18987080813633864, "grad_norm": 0.8873284459114075, "learning_rate": 2.8034390874446088e-05, "loss": 0.0327, "step": 2763 }, { "epoch": 0.18993952721275426, "grad_norm": 1.5977280139923096, "learning_rate": 2.8032738349654354e-05, "loss": 0.0468, "step": 2764 }, { "epoch": 0.19000824628916987, "grad_norm": 1.0916187763214111, "learning_rate": 2.803108517924756e-05, "loss": 0.0536, "step": 2765 }, { "epoch": 0.1900769653655855, "grad_norm": 1.9659688472747803, "learning_rate": 2.802943136330759e-05, "loss": 0.061, "step": 2766 }, { "epoch": 0.1901456844420011, "grad_norm": 1.5000807046890259, "learning_rate": 2.8027776901916383e-05, "loss": 0.0626, "step": 2767 }, { "epoch": 0.1902144035184167, "grad_norm": 1.5913349390029907, "learning_rate": 2.8026121795155893e-05, "loss": 0.0686, "step": 2768 }, { "epoch": 0.19028312259483232, "grad_norm": 1.103074550628662, "learning_rate": 2.8024466043108105e-05, "loss": 0.0489, "step": 2769 }, { "epoch": 0.19035184167124794, "grad_norm": 1.136152982711792, "learning_rate": 2.802280964585505e-05, "loss": 0.0431, "step": 2770 }, { "epoch": 0.19042056074766356, "grad_norm": 1.6811342239379883, "learning_rate": 2.8021152603478775e-05, "loss": 0.0477, "step": 2771 }, { "epoch": 0.19048927982407918, "grad_norm": 2.4903390407562256, "learning_rate": 2.801949491606137e-05, "loss": 0.0401, "step": 2772 }, { "epoch": 0.19055799890049477, "grad_norm": 1.1215778589248657, "learning_rate": 2.801783658368496e-05, "loss": 0.049, "step": 2773 }, { "epoch": 0.1906267179769104, "grad_norm": 1.8102364540100098, "learning_rate": 2.801617760643169e-05, "loss": 0.053, "step": 2774 }, { "epoch": 0.190695437053326, "grad_norm": 2.081066131591797, "learning_rate": 2.8014517984383735e-05, "loss": 0.0854, "step": 2775 }, { "epoch": 0.19076415612974162, "grad_norm": 0.8615603446960449, "learning_rate": 2.8012857717623318e-05, "loss": 0.0337, "step": 2776 }, { "epoch": 0.19083287520615724, "grad_norm": 1.4687285423278809, "learning_rate": 2.8011196806232684e-05, "loss": 0.0593, "step": 2777 }, { "epoch": 0.19090159428257283, "grad_norm": 1.4346009492874146, "learning_rate": 2.800953525029411e-05, "loss": 0.0501, "step": 2778 }, { "epoch": 0.19097031335898845, "grad_norm": 1.250576376914978, "learning_rate": 2.80078730498899e-05, "loss": 0.0479, "step": 2779 }, { "epoch": 0.19103903243540407, "grad_norm": 1.3696398735046387, "learning_rate": 2.800621020510241e-05, "loss": 0.0559, "step": 2780 }, { "epoch": 0.1911077515118197, "grad_norm": 1.3293683528900146, "learning_rate": 2.8004546716014005e-05, "loss": 0.0367, "step": 2781 }, { "epoch": 0.19117647058823528, "grad_norm": 0.9935635924339294, "learning_rate": 2.8002882582707094e-05, "loss": 0.0438, "step": 2782 }, { "epoch": 0.1912451896646509, "grad_norm": 2.0375819206237793, "learning_rate": 2.8001217805264105e-05, "loss": 0.0459, "step": 2783 }, { "epoch": 0.19131390874106652, "grad_norm": 1.5687869787216187, "learning_rate": 2.7999552383767522e-05, "loss": 0.0506, "step": 2784 }, { "epoch": 0.19138262781748214, "grad_norm": 1.799679160118103, "learning_rate": 2.7997886318299833e-05, "loss": 0.0542, "step": 2785 }, { "epoch": 0.19145134689389776, "grad_norm": 0.6525881290435791, "learning_rate": 2.7996219608943587e-05, "loss": 0.0385, "step": 2786 }, { "epoch": 0.19152006597031335, "grad_norm": 2.1173107624053955, "learning_rate": 2.7994552255781334e-05, "loss": 0.0599, "step": 2787 }, { "epoch": 0.19158878504672897, "grad_norm": 1.4120008945465088, "learning_rate": 2.7992884258895682e-05, "loss": 0.045, "step": 2788 }, { "epoch": 0.19165750412314458, "grad_norm": 1.2091240882873535, "learning_rate": 2.7991215618369256e-05, "loss": 0.0497, "step": 2789 }, { "epoch": 0.1917262231995602, "grad_norm": 1.5922472476959229, "learning_rate": 2.7989546334284713e-05, "loss": 0.0486, "step": 2790 }, { "epoch": 0.19179494227597582, "grad_norm": 1.6790571212768555, "learning_rate": 2.7987876406724755e-05, "loss": 0.0676, "step": 2791 }, { "epoch": 0.1918636613523914, "grad_norm": 1.2557663917541504, "learning_rate": 2.7986205835772095e-05, "loss": 0.051, "step": 2792 }, { "epoch": 0.19193238042880703, "grad_norm": 1.2846593856811523, "learning_rate": 2.7984534621509503e-05, "loss": 0.0594, "step": 2793 }, { "epoch": 0.19200109950522265, "grad_norm": 1.5030763149261475, "learning_rate": 2.7982862764019756e-05, "loss": 0.0658, "step": 2794 }, { "epoch": 0.19206981858163827, "grad_norm": 1.0983269214630127, "learning_rate": 2.7981190263385683e-05, "loss": 0.0426, "step": 2795 }, { "epoch": 0.1921385376580539, "grad_norm": 0.8846912384033203, "learning_rate": 2.797951711969013e-05, "loss": 0.0458, "step": 2796 }, { "epoch": 0.19220725673446948, "grad_norm": 1.1353949308395386, "learning_rate": 2.7977843333015986e-05, "loss": 0.0361, "step": 2797 }, { "epoch": 0.1922759758108851, "grad_norm": 1.218963623046875, "learning_rate": 2.7976168903446166e-05, "loss": 0.0464, "step": 2798 }, { "epoch": 0.19234469488730072, "grad_norm": 1.805925726890564, "learning_rate": 2.7974493831063614e-05, "loss": 0.0543, "step": 2799 }, { "epoch": 0.19241341396371633, "grad_norm": 1.657373070716858, "learning_rate": 2.7972818115951313e-05, "loss": 0.0456, "step": 2800 }, { "epoch": 0.19248213304013195, "grad_norm": 1.1380534172058105, "learning_rate": 2.797114175819227e-05, "loss": 0.0563, "step": 2801 }, { "epoch": 0.19255085211654754, "grad_norm": 1.6655585765838623, "learning_rate": 2.7969464757869535e-05, "loss": 0.062, "step": 2802 }, { "epoch": 0.19261957119296316, "grad_norm": 0.916766345500946, "learning_rate": 2.796778711506618e-05, "loss": 0.0546, "step": 2803 }, { "epoch": 0.19268829026937878, "grad_norm": 1.3904056549072266, "learning_rate": 2.7966108829865312e-05, "loss": 0.0538, "step": 2804 }, { "epoch": 0.1927570093457944, "grad_norm": 1.5912481546401978, "learning_rate": 2.796442990235007e-05, "loss": 0.0368, "step": 2805 }, { "epoch": 0.19282572842221002, "grad_norm": 0.7163512110710144, "learning_rate": 2.796275033260362e-05, "loss": 0.0336, "step": 2806 }, { "epoch": 0.1928944474986256, "grad_norm": 1.9057904481887817, "learning_rate": 2.7961070120709175e-05, "loss": 0.0616, "step": 2807 }, { "epoch": 0.19296316657504123, "grad_norm": 1.4938651323318481, "learning_rate": 2.795938926674996e-05, "loss": 0.0548, "step": 2808 }, { "epoch": 0.19303188565145685, "grad_norm": 1.918710708618164, "learning_rate": 2.7957707770809243e-05, "loss": 0.0514, "step": 2809 }, { "epoch": 0.19310060472787247, "grad_norm": 1.5692543983459473, "learning_rate": 2.7956025632970327e-05, "loss": 0.0437, "step": 2810 }, { "epoch": 0.19316932380428806, "grad_norm": 1.3232886791229248, "learning_rate": 2.795434285331653e-05, "loss": 0.0549, "step": 2811 }, { "epoch": 0.19323804288070368, "grad_norm": 2.371333360671997, "learning_rate": 2.7952659431931226e-05, "loss": 0.0532, "step": 2812 }, { "epoch": 0.1933067619571193, "grad_norm": 1.128891110420227, "learning_rate": 2.795097536889781e-05, "loss": 0.0352, "step": 2813 }, { "epoch": 0.1933754810335349, "grad_norm": 1.4136401414871216, "learning_rate": 2.794929066429969e-05, "loss": 0.042, "step": 2814 }, { "epoch": 0.19344420010995053, "grad_norm": 0.9419192671775818, "learning_rate": 2.7947605318220334e-05, "loss": 0.0374, "step": 2815 }, { "epoch": 0.19351291918636612, "grad_norm": 1.7355178594589233, "learning_rate": 2.7945919330743237e-05, "loss": 0.0529, "step": 2816 }, { "epoch": 0.19358163826278174, "grad_norm": 1.853623390197754, "learning_rate": 2.7944232701951904e-05, "loss": 0.0524, "step": 2817 }, { "epoch": 0.19365035733919736, "grad_norm": 5.569084644317627, "learning_rate": 2.79425454319299e-05, "loss": 0.0655, "step": 2818 }, { "epoch": 0.19371907641561298, "grad_norm": 0.6515247225761414, "learning_rate": 2.79408575207608e-05, "loss": 0.0465, "step": 2819 }, { "epoch": 0.1937877954920286, "grad_norm": 0.7712896466255188, "learning_rate": 2.793916896852823e-05, "loss": 0.0509, "step": 2820 }, { "epoch": 0.1938565145684442, "grad_norm": 1.7619125843048096, "learning_rate": 2.7937479775315827e-05, "loss": 0.0581, "step": 2821 }, { "epoch": 0.1939252336448598, "grad_norm": 1.4849858283996582, "learning_rate": 2.7935789941207274e-05, "loss": 0.0528, "step": 2822 }, { "epoch": 0.19399395272127543, "grad_norm": 1.4215184450149536, "learning_rate": 2.7934099466286282e-05, "loss": 0.0466, "step": 2823 }, { "epoch": 0.19406267179769104, "grad_norm": 0.8397455215454102, "learning_rate": 2.7932408350636596e-05, "loss": 0.0361, "step": 2824 }, { "epoch": 0.19413139087410666, "grad_norm": 1.3845552206039429, "learning_rate": 2.7930716594341986e-05, "loss": 0.0515, "step": 2825 }, { "epoch": 0.19420010995052225, "grad_norm": 0.9140058755874634, "learning_rate": 2.7929024197486262e-05, "loss": 0.0502, "step": 2826 }, { "epoch": 0.19426882902693787, "grad_norm": 1.0483578443527222, "learning_rate": 2.7927331160153264e-05, "loss": 0.0455, "step": 2827 }, { "epoch": 0.1943375481033535, "grad_norm": 1.5823144912719727, "learning_rate": 2.792563748242685e-05, "loss": 0.0661, "step": 2828 }, { "epoch": 0.1944062671797691, "grad_norm": 1.3588758707046509, "learning_rate": 2.792394316439093e-05, "loss": 0.0646, "step": 2829 }, { "epoch": 0.19447498625618473, "grad_norm": 1.298277735710144, "learning_rate": 2.792224820612944e-05, "loss": 0.0419, "step": 2830 }, { "epoch": 0.19454370533260032, "grad_norm": 2.259984254837036, "learning_rate": 2.792055260772634e-05, "loss": 0.0572, "step": 2831 }, { "epoch": 0.19461242440901594, "grad_norm": 1.7871544361114502, "learning_rate": 2.7918856369265624e-05, "loss": 0.0521, "step": 2832 }, { "epoch": 0.19468114348543156, "grad_norm": 1.204180359840393, "learning_rate": 2.791715949083133e-05, "loss": 0.0425, "step": 2833 }, { "epoch": 0.19474986256184718, "grad_norm": 1.2916908264160156, "learning_rate": 2.7915461972507506e-05, "loss": 0.0454, "step": 2834 }, { "epoch": 0.1948185816382628, "grad_norm": 1.3598607778549194, "learning_rate": 2.791376381437825e-05, "loss": 0.0417, "step": 2835 }, { "epoch": 0.19488730071467839, "grad_norm": 2.095900535583496, "learning_rate": 2.7912065016527687e-05, "loss": 0.0614, "step": 2836 }, { "epoch": 0.194956019791094, "grad_norm": 1.1826343536376953, "learning_rate": 2.7910365579039963e-05, "loss": 0.0531, "step": 2837 }, { "epoch": 0.19502473886750962, "grad_norm": 1.3448370695114136, "learning_rate": 2.7908665501999277e-05, "loss": 0.0494, "step": 2838 }, { "epoch": 0.19509345794392524, "grad_norm": 1.4123295545578003, "learning_rate": 2.7906964785489838e-05, "loss": 0.0522, "step": 2839 }, { "epoch": 0.19516217702034086, "grad_norm": 2.447667121887207, "learning_rate": 2.79052634295959e-05, "loss": 0.0511, "step": 2840 }, { "epoch": 0.19523089609675645, "grad_norm": 1.0866961479187012, "learning_rate": 2.7903561434401743e-05, "loss": 0.0472, "step": 2841 }, { "epoch": 0.19529961517317207, "grad_norm": 1.2100462913513184, "learning_rate": 2.7901858799991685e-05, "loss": 0.0448, "step": 2842 }, { "epoch": 0.1953683342495877, "grad_norm": 0.8418573141098022, "learning_rate": 2.7900155526450063e-05, "loss": 0.0519, "step": 2843 }, { "epoch": 0.1954370533260033, "grad_norm": 1.5030604600906372, "learning_rate": 2.789845161386126e-05, "loss": 0.0813, "step": 2844 }, { "epoch": 0.1955057724024189, "grad_norm": 1.4559204578399658, "learning_rate": 2.789674706230968e-05, "loss": 0.0569, "step": 2845 }, { "epoch": 0.19557449147883452, "grad_norm": 1.3519104719161987, "learning_rate": 2.7895041871879766e-05, "loss": 0.0459, "step": 2846 }, { "epoch": 0.19564321055525014, "grad_norm": 1.1457395553588867, "learning_rate": 2.7893336042655988e-05, "loss": 0.0543, "step": 2847 }, { "epoch": 0.19571192963166575, "grad_norm": 1.677543044090271, "learning_rate": 2.789162957472285e-05, "loss": 0.045, "step": 2848 }, { "epoch": 0.19578064870808137, "grad_norm": 2.038386344909668, "learning_rate": 2.7889922468164887e-05, "loss": 0.0557, "step": 2849 }, { "epoch": 0.19584936778449696, "grad_norm": 1.004178762435913, "learning_rate": 2.7888214723066667e-05, "loss": 0.026, "step": 2850 }, { "epoch": 0.19591808686091258, "grad_norm": 1.2143994569778442, "learning_rate": 2.788650633951279e-05, "loss": 0.0526, "step": 2851 }, { "epoch": 0.1959868059373282, "grad_norm": 2.467621326446533, "learning_rate": 2.7884797317587878e-05, "loss": 0.0577, "step": 2852 }, { "epoch": 0.19605552501374382, "grad_norm": 1.7338758707046509, "learning_rate": 2.7883087657376597e-05, "loss": 0.0434, "step": 2853 }, { "epoch": 0.19612424409015944, "grad_norm": 1.2322067022323608, "learning_rate": 2.7881377358963636e-05, "loss": 0.0428, "step": 2854 }, { "epoch": 0.19619296316657503, "grad_norm": 1.0660768747329712, "learning_rate": 2.7879666422433732e-05, "loss": 0.0381, "step": 2855 }, { "epoch": 0.19626168224299065, "grad_norm": 1.28126859664917, "learning_rate": 2.787795484787163e-05, "loss": 0.0576, "step": 2856 }, { "epoch": 0.19633040131940627, "grad_norm": 1.3594201803207397, "learning_rate": 2.7876242635362118e-05, "loss": 0.0468, "step": 2857 }, { "epoch": 0.19639912039582189, "grad_norm": 1.3247209787368774, "learning_rate": 2.7874529784990018e-05, "loss": 0.0696, "step": 2858 }, { "epoch": 0.1964678394722375, "grad_norm": 1.3916456699371338, "learning_rate": 2.7872816296840185e-05, "loss": 0.0499, "step": 2859 }, { "epoch": 0.1965365585486531, "grad_norm": 1.1211087703704834, "learning_rate": 2.7871102170997497e-05, "loss": 0.0514, "step": 2860 }, { "epoch": 0.19660527762506871, "grad_norm": 1.7280906438827515, "learning_rate": 2.786938740754687e-05, "loss": 0.0781, "step": 2861 }, { "epoch": 0.19667399670148433, "grad_norm": 1.4553800821304321, "learning_rate": 2.7867672006573248e-05, "loss": 0.0621, "step": 2862 }, { "epoch": 0.19674271577789995, "grad_norm": 1.207790493965149, "learning_rate": 2.7865955968161607e-05, "loss": 0.0466, "step": 2863 }, { "epoch": 0.19681143485431557, "grad_norm": 1.1560167074203491, "learning_rate": 2.786423929239696e-05, "loss": 0.0485, "step": 2864 }, { "epoch": 0.19688015393073116, "grad_norm": 0.8572007417678833, "learning_rate": 2.786252197936435e-05, "loss": 0.0305, "step": 2865 }, { "epoch": 0.19694887300714678, "grad_norm": 1.2834111452102661, "learning_rate": 2.7860804029148843e-05, "loss": 0.0386, "step": 2866 }, { "epoch": 0.1970175920835624, "grad_norm": 1.0430840253829956, "learning_rate": 2.7859085441835544e-05, "loss": 0.0308, "step": 2867 }, { "epoch": 0.19708631115997802, "grad_norm": 1.3489151000976562, "learning_rate": 2.785736621750959e-05, "loss": 0.0433, "step": 2868 }, { "epoch": 0.19715503023639364, "grad_norm": 0.8494724035263062, "learning_rate": 2.785564635625615e-05, "loss": 0.0355, "step": 2869 }, { "epoch": 0.19722374931280923, "grad_norm": 0.883741557598114, "learning_rate": 2.7853925858160412e-05, "loss": 0.0381, "step": 2870 }, { "epoch": 0.19729246838922485, "grad_norm": 0.6257498264312744, "learning_rate": 2.785220472330762e-05, "loss": 0.0278, "step": 2871 }, { "epoch": 0.19736118746564046, "grad_norm": 1.6391472816467285, "learning_rate": 2.785048295178303e-05, "loss": 0.0451, "step": 2872 }, { "epoch": 0.19742990654205608, "grad_norm": 1.1828114986419678, "learning_rate": 2.784876054367193e-05, "loss": 0.0509, "step": 2873 }, { "epoch": 0.1974986256184717, "grad_norm": 0.9714270234107971, "learning_rate": 2.784703749905965e-05, "loss": 0.0502, "step": 2874 }, { "epoch": 0.1975673446948873, "grad_norm": 1.6825412511825562, "learning_rate": 2.7845313818031547e-05, "loss": 0.0529, "step": 2875 }, { "epoch": 0.1976360637713029, "grad_norm": 2.6297571659088135, "learning_rate": 2.7843589500673002e-05, "loss": 0.0482, "step": 2876 }, { "epoch": 0.19770478284771853, "grad_norm": 1.5387260913848877, "learning_rate": 2.784186454706944e-05, "loss": 0.0646, "step": 2877 }, { "epoch": 0.19777350192413415, "grad_norm": 1.7981590032577515, "learning_rate": 2.7840138957306312e-05, "loss": 0.0407, "step": 2878 }, { "epoch": 0.19784222100054974, "grad_norm": 1.6539702415466309, "learning_rate": 2.7838412731469095e-05, "loss": 0.0499, "step": 2879 }, { "epoch": 0.19791094007696536, "grad_norm": 1.124821662902832, "learning_rate": 2.783668586964331e-05, "loss": 0.0525, "step": 2880 }, { "epoch": 0.19797965915338098, "grad_norm": 1.5138590335845947, "learning_rate": 2.7834958371914493e-05, "loss": 0.0514, "step": 2881 }, { "epoch": 0.1980483782297966, "grad_norm": 1.4628499746322632, "learning_rate": 2.783323023836823e-05, "loss": 0.0546, "step": 2882 }, { "epoch": 0.19811709730621221, "grad_norm": 1.7432430982589722, "learning_rate": 2.7831501469090125e-05, "loss": 0.0389, "step": 2883 }, { "epoch": 0.1981858163826278, "grad_norm": 2.166149139404297, "learning_rate": 2.782977206416581e-05, "loss": 0.0555, "step": 2884 }, { "epoch": 0.19825453545904342, "grad_norm": 1.7486716508865356, "learning_rate": 2.7828042023680975e-05, "loss": 0.039, "step": 2885 }, { "epoch": 0.19832325453545904, "grad_norm": 1.2712571620941162, "learning_rate": 2.7826311347721304e-05, "loss": 0.042, "step": 2886 }, { "epoch": 0.19839197361187466, "grad_norm": 1.3035995960235596, "learning_rate": 2.7824580036372543e-05, "loss": 0.0417, "step": 2887 }, { "epoch": 0.19846069268829028, "grad_norm": 2.0640065670013428, "learning_rate": 2.7822848089720452e-05, "loss": 0.0521, "step": 2888 }, { "epoch": 0.19852941176470587, "grad_norm": 0.9260632395744324, "learning_rate": 2.7821115507850833e-05, "loss": 0.0509, "step": 2889 }, { "epoch": 0.1985981308411215, "grad_norm": 1.4319990873336792, "learning_rate": 2.7819382290849505e-05, "loss": 0.0558, "step": 2890 }, { "epoch": 0.1986668499175371, "grad_norm": 1.3664730787277222, "learning_rate": 2.7817648438802336e-05, "loss": 0.0396, "step": 2891 }, { "epoch": 0.19873556899395273, "grad_norm": 1.0387229919433594, "learning_rate": 2.7815913951795213e-05, "loss": 0.0566, "step": 2892 }, { "epoch": 0.19880428807036835, "grad_norm": 1.4242616891860962, "learning_rate": 2.7814178829914064e-05, "loss": 0.0617, "step": 2893 }, { "epoch": 0.19887300714678394, "grad_norm": 1.5795190334320068, "learning_rate": 2.781244307324484e-05, "loss": 0.0595, "step": 2894 }, { "epoch": 0.19894172622319956, "grad_norm": 0.9501801133155823, "learning_rate": 2.7810706681873528e-05, "loss": 0.0445, "step": 2895 }, { "epoch": 0.19901044529961517, "grad_norm": 1.289000391960144, "learning_rate": 2.7808969655886142e-05, "loss": 0.0396, "step": 2896 }, { "epoch": 0.1990791643760308, "grad_norm": 1.6067273616790771, "learning_rate": 2.7807231995368733e-05, "loss": 0.0705, "step": 2897 }, { "epoch": 0.1991478834524464, "grad_norm": 1.762702226638794, "learning_rate": 2.7805493700407384e-05, "loss": 0.0498, "step": 2898 }, { "epoch": 0.199216602528862, "grad_norm": 1.4548059701919556, "learning_rate": 2.7803754771088204e-05, "loss": 0.0502, "step": 2899 }, { "epoch": 0.19928532160527762, "grad_norm": 1.4017274379730225, "learning_rate": 2.780201520749734e-05, "loss": 0.052, "step": 2900 }, { "epoch": 0.19935404068169324, "grad_norm": 1.6662960052490234, "learning_rate": 2.7800275009720957e-05, "loss": 0.0555, "step": 2901 }, { "epoch": 0.19942275975810886, "grad_norm": 0.6574588418006897, "learning_rate": 2.7798534177845268e-05, "loss": 0.0593, "step": 2902 }, { "epoch": 0.19949147883452448, "grad_norm": 1.485642910003662, "learning_rate": 2.7796792711956508e-05, "loss": 0.0391, "step": 2903 }, { "epoch": 0.19956019791094007, "grad_norm": 1.9163695573806763, "learning_rate": 2.779505061214094e-05, "loss": 0.0545, "step": 2904 }, { "epoch": 0.1996289169873557, "grad_norm": 0.9202004671096802, "learning_rate": 2.779330787848488e-05, "loss": 0.0485, "step": 2905 }, { "epoch": 0.1996976360637713, "grad_norm": 1.2154772281646729, "learning_rate": 2.7791564511074647e-05, "loss": 0.0455, "step": 2906 }, { "epoch": 0.19976635514018692, "grad_norm": 0.7821887135505676, "learning_rate": 2.7789820509996604e-05, "loss": 0.0397, "step": 2907 }, { "epoch": 0.19983507421660252, "grad_norm": 1.2745529413223267, "learning_rate": 2.778807587533715e-05, "loss": 0.0364, "step": 2908 }, { "epoch": 0.19990379329301813, "grad_norm": 1.4560645818710327, "learning_rate": 2.7786330607182708e-05, "loss": 0.0528, "step": 2909 }, { "epoch": 0.19997251236943375, "grad_norm": 1.0830732583999634, "learning_rate": 2.7784584705619734e-05, "loss": 0.0439, "step": 2910 }, { "epoch": 0.20004123144584937, "grad_norm": 2.0548830032348633, "learning_rate": 2.778283817073472e-05, "loss": 0.0613, "step": 2911 }, { "epoch": 0.200109950522265, "grad_norm": 0.8740656971931458, "learning_rate": 2.778109100261418e-05, "loss": 0.0381, "step": 2912 }, { "epoch": 0.20017866959868058, "grad_norm": 1.2596404552459717, "learning_rate": 2.777934320134467e-05, "loss": 0.0418, "step": 2913 }, { "epoch": 0.2002473886750962, "grad_norm": 0.8852318525314331, "learning_rate": 2.7777594767012772e-05, "loss": 0.042, "step": 2914 }, { "epoch": 0.20031610775151182, "grad_norm": 1.465498447418213, "learning_rate": 2.7775845699705097e-05, "loss": 0.0375, "step": 2915 }, { "epoch": 0.20038482682792744, "grad_norm": 3.328040838241577, "learning_rate": 2.7774095999508295e-05, "loss": 0.0584, "step": 2916 }, { "epoch": 0.20045354590434306, "grad_norm": 0.9079594612121582, "learning_rate": 2.7772345666509035e-05, "loss": 0.0341, "step": 2917 }, { "epoch": 0.20052226498075865, "grad_norm": 1.5242109298706055, "learning_rate": 2.7770594700794035e-05, "loss": 0.062, "step": 2918 }, { "epoch": 0.20059098405717427, "grad_norm": 1.7743399143218994, "learning_rate": 2.776884310245003e-05, "loss": 0.0592, "step": 2919 }, { "epoch": 0.20065970313358988, "grad_norm": 1.4448293447494507, "learning_rate": 2.776709087156378e-05, "loss": 0.0561, "step": 2920 }, { "epoch": 0.2007284222100055, "grad_norm": 1.5567889213562012, "learning_rate": 2.7765338008222102e-05, "loss": 0.0665, "step": 2921 }, { "epoch": 0.20079714128642112, "grad_norm": 1.8061736822128296, "learning_rate": 2.7763584512511826e-05, "loss": 0.0485, "step": 2922 }, { "epoch": 0.2008658603628367, "grad_norm": 1.3064332008361816, "learning_rate": 2.776183038451981e-05, "loss": 0.0469, "step": 2923 }, { "epoch": 0.20093457943925233, "grad_norm": 1.0085961818695068, "learning_rate": 2.7760075624332956e-05, "loss": 0.0409, "step": 2924 }, { "epoch": 0.20100329851566795, "grad_norm": 1.432260274887085, "learning_rate": 2.7758320232038187e-05, "loss": 0.0525, "step": 2925 }, { "epoch": 0.20107201759208357, "grad_norm": 0.8552116751670837, "learning_rate": 2.7756564207722463e-05, "loss": 0.0423, "step": 2926 }, { "epoch": 0.2011407366684992, "grad_norm": 1.2558223009109497, "learning_rate": 2.775480755147278e-05, "loss": 0.0446, "step": 2927 }, { "epoch": 0.20120945574491478, "grad_norm": 1.788252830505371, "learning_rate": 2.775305026337615e-05, "loss": 0.0612, "step": 2928 }, { "epoch": 0.2012781748213304, "grad_norm": 1.2335370779037476, "learning_rate": 2.7751292343519627e-05, "loss": 0.0483, "step": 2929 }, { "epoch": 0.20134689389774602, "grad_norm": 1.3683905601501465, "learning_rate": 2.7749533791990303e-05, "loss": 0.0677, "step": 2930 }, { "epoch": 0.20141561297416163, "grad_norm": 1.2395776510238647, "learning_rate": 2.7747774608875283e-05, "loss": 0.0447, "step": 2931 }, { "epoch": 0.20148433205057725, "grad_norm": 1.6786447763442993, "learning_rate": 2.7746014794261717e-05, "loss": 0.0565, "step": 2932 }, { "epoch": 0.20155305112699284, "grad_norm": 0.9138020873069763, "learning_rate": 2.7744254348236782e-05, "loss": 0.0472, "step": 2933 }, { "epoch": 0.20162177020340846, "grad_norm": 6.282625198364258, "learning_rate": 2.7742493270887687e-05, "loss": 0.0372, "step": 2934 }, { "epoch": 0.20169048927982408, "grad_norm": 0.9724721312522888, "learning_rate": 2.7740731562301675e-05, "loss": 0.0338, "step": 2935 }, { "epoch": 0.2017592083562397, "grad_norm": 1.702470302581787, "learning_rate": 2.7738969222566016e-05, "loss": 0.0399, "step": 2936 }, { "epoch": 0.20182792743265532, "grad_norm": 1.1002123355865479, "learning_rate": 2.773720625176801e-05, "loss": 0.0506, "step": 2937 }, { "epoch": 0.2018966465090709, "grad_norm": 1.5089809894561768, "learning_rate": 2.7735442649994995e-05, "loss": 0.0445, "step": 2938 }, { "epoch": 0.20196536558548653, "grad_norm": 1.3767426013946533, "learning_rate": 2.7733678417334334e-05, "loss": 0.0564, "step": 2939 }, { "epoch": 0.20203408466190215, "grad_norm": 2.0182173252105713, "learning_rate": 2.773191355387342e-05, "loss": 0.0727, "step": 2940 }, { "epoch": 0.20210280373831777, "grad_norm": 0.8938625454902649, "learning_rate": 2.7730148059699687e-05, "loss": 0.0385, "step": 2941 }, { "epoch": 0.20217152281473336, "grad_norm": 1.0466605424880981, "learning_rate": 2.772838193490059e-05, "loss": 0.0343, "step": 2942 }, { "epoch": 0.20224024189114898, "grad_norm": 2.404172658920288, "learning_rate": 2.772661517956362e-05, "loss": 0.0602, "step": 2943 }, { "epoch": 0.2023089609675646, "grad_norm": 5.729150772094727, "learning_rate": 2.7724847793776304e-05, "loss": 0.0669, "step": 2944 }, { "epoch": 0.2023776800439802, "grad_norm": 2.6311357021331787, "learning_rate": 2.7723079777626185e-05, "loss": 0.0612, "step": 2945 }, { "epoch": 0.20244639912039583, "grad_norm": 1.0342434644699097, "learning_rate": 2.7721311131200852e-05, "loss": 0.0589, "step": 2946 }, { "epoch": 0.20251511819681142, "grad_norm": 1.495007038116455, "learning_rate": 2.771954185458792e-05, "loss": 0.0475, "step": 2947 }, { "epoch": 0.20258383727322704, "grad_norm": 0.8327227830886841, "learning_rate": 2.771777194787503e-05, "loss": 0.0347, "step": 2948 }, { "epoch": 0.20265255634964266, "grad_norm": 0.9888694286346436, "learning_rate": 2.771600141114987e-05, "loss": 0.0535, "step": 2949 }, { "epoch": 0.20272127542605828, "grad_norm": 0.9964407086372375, "learning_rate": 2.771423024450014e-05, "loss": 0.0355, "step": 2950 }, { "epoch": 0.2027899945024739, "grad_norm": 1.3185343742370605, "learning_rate": 2.7712458448013586e-05, "loss": 0.0357, "step": 2951 }, { "epoch": 0.2028587135788895, "grad_norm": 1.7315566539764404, "learning_rate": 2.7710686021777977e-05, "loss": 0.0502, "step": 2952 }, { "epoch": 0.2029274326553051, "grad_norm": 0.9898998141288757, "learning_rate": 2.770891296588111e-05, "loss": 0.0467, "step": 2953 }, { "epoch": 0.20299615173172073, "grad_norm": 1.1151460409164429, "learning_rate": 2.7707139280410824e-05, "loss": 0.0408, "step": 2954 }, { "epoch": 0.20306487080813634, "grad_norm": 1.0767301321029663, "learning_rate": 2.7705364965454986e-05, "loss": 0.0286, "step": 2955 }, { "epoch": 0.20313358988455196, "grad_norm": 1.3682947158813477, "learning_rate": 2.7703590021101488e-05, "loss": 0.0415, "step": 2956 }, { "epoch": 0.20320230896096755, "grad_norm": 1.6841999292373657, "learning_rate": 2.7701814447438254e-05, "loss": 0.0601, "step": 2957 }, { "epoch": 0.20327102803738317, "grad_norm": 2.2502329349517822, "learning_rate": 2.770003824455325e-05, "loss": 0.063, "step": 2958 }, { "epoch": 0.2033397471137988, "grad_norm": 1.0488457679748535, "learning_rate": 2.769826141253446e-05, "loss": 0.0426, "step": 2959 }, { "epoch": 0.2034084661902144, "grad_norm": 0.5995525121688843, "learning_rate": 2.7696483951469906e-05, "loss": 0.0376, "step": 2960 }, { "epoch": 0.20347718526663003, "grad_norm": 0.9621731042861938, "learning_rate": 2.769470586144764e-05, "loss": 0.0455, "step": 2961 }, { "epoch": 0.20354590434304562, "grad_norm": 1.1257984638214111, "learning_rate": 2.7692927142555748e-05, "loss": 0.0416, "step": 2962 }, { "epoch": 0.20361462341946124, "grad_norm": 0.6597365736961365, "learning_rate": 2.7691147794882337e-05, "loss": 0.0454, "step": 2963 }, { "epoch": 0.20368334249587686, "grad_norm": 1.335769772529602, "learning_rate": 2.7689367818515556e-05, "loss": 0.0376, "step": 2964 }, { "epoch": 0.20375206157229248, "grad_norm": 2.256180763244629, "learning_rate": 2.7687587213543586e-05, "loss": 0.0665, "step": 2965 }, { "epoch": 0.2038207806487081, "grad_norm": 1.7111778259277344, "learning_rate": 2.7685805980054625e-05, "loss": 0.0731, "step": 2966 }, { "epoch": 0.20388949972512369, "grad_norm": 1.0764331817626953, "learning_rate": 2.7684024118136915e-05, "loss": 0.0452, "step": 2967 }, { "epoch": 0.2039582188015393, "grad_norm": 1.7946921586990356, "learning_rate": 2.7682241627878733e-05, "loss": 0.0408, "step": 2968 }, { "epoch": 0.20402693787795492, "grad_norm": 2.132534980773926, "learning_rate": 2.7680458509368374e-05, "loss": 0.0478, "step": 2969 }, { "epoch": 0.20409565695437054, "grad_norm": 1.039939045906067, "learning_rate": 2.767867476269417e-05, "loss": 0.0297, "step": 2970 }, { "epoch": 0.20416437603078616, "grad_norm": 1.6966418027877808, "learning_rate": 2.7676890387944483e-05, "loss": 0.0519, "step": 2971 }, { "epoch": 0.20423309510720175, "grad_norm": 1.5374290943145752, "learning_rate": 2.767510538520771e-05, "loss": 0.0368, "step": 2972 }, { "epoch": 0.20430181418361737, "grad_norm": 0.9155722260475159, "learning_rate": 2.767331975457227e-05, "loss": 0.0466, "step": 2973 }, { "epoch": 0.204370533260033, "grad_norm": 1.738520860671997, "learning_rate": 2.767153349612663e-05, "loss": 0.0473, "step": 2974 }, { "epoch": 0.2044392523364486, "grad_norm": 1.3193416595458984, "learning_rate": 2.766974660995927e-05, "loss": 0.0496, "step": 2975 }, { "epoch": 0.2045079714128642, "grad_norm": 1.0111900568008423, "learning_rate": 2.7667959096158715e-05, "loss": 0.0362, "step": 2976 }, { "epoch": 0.20457669048927982, "grad_norm": 0.953757643699646, "learning_rate": 2.7666170954813503e-05, "loss": 0.0485, "step": 2977 }, { "epoch": 0.20464540956569544, "grad_norm": 2.0220606327056885, "learning_rate": 2.766438218601223e-05, "loss": 0.0521, "step": 2978 }, { "epoch": 0.20471412864211105, "grad_norm": 6.207425594329834, "learning_rate": 2.76625927898435e-05, "loss": 0.0479, "step": 2979 }, { "epoch": 0.20478284771852667, "grad_norm": 4.41954231262207, "learning_rate": 2.766080276639595e-05, "loss": 0.0584, "step": 2980 }, { "epoch": 0.20485156679494226, "grad_norm": 2.110267162322998, "learning_rate": 2.7659012115758264e-05, "loss": 0.0701, "step": 2981 }, { "epoch": 0.20492028587135788, "grad_norm": 1.3405876159667969, "learning_rate": 2.7657220838019143e-05, "loss": 0.0667, "step": 2982 }, { "epoch": 0.2049890049477735, "grad_norm": 1.1226431131362915, "learning_rate": 2.7655428933267325e-05, "loss": 0.0625, "step": 2983 }, { "epoch": 0.20505772402418912, "grad_norm": 1.928961992263794, "learning_rate": 2.7653636401591575e-05, "loss": 0.0719, "step": 2984 }, { "epoch": 0.20512644310060474, "grad_norm": 1.7448328733444214, "learning_rate": 2.7651843243080695e-05, "loss": 0.0448, "step": 2985 }, { "epoch": 0.20519516217702033, "grad_norm": 1.0952485799789429, "learning_rate": 2.7650049457823505e-05, "loss": 0.0378, "step": 2986 }, { "epoch": 0.20526388125343595, "grad_norm": 1.6320184469223022, "learning_rate": 2.764825504590888e-05, "loss": 0.0607, "step": 2987 }, { "epoch": 0.20533260032985157, "grad_norm": 1.4692379236221313, "learning_rate": 2.7646460007425698e-05, "loss": 0.0426, "step": 2988 }, { "epoch": 0.20540131940626719, "grad_norm": 1.0208767652511597, "learning_rate": 2.7644664342462893e-05, "loss": 0.0617, "step": 2989 }, { "epoch": 0.2054700384826828, "grad_norm": 1.64795982837677, "learning_rate": 2.7642868051109407e-05, "loss": 0.0487, "step": 2990 }, { "epoch": 0.2055387575590984, "grad_norm": 1.0811004638671875, "learning_rate": 2.7641071133454235e-05, "loss": 0.0421, "step": 2991 }, { "epoch": 0.205607476635514, "grad_norm": 1.1205908060073853, "learning_rate": 2.7639273589586388e-05, "loss": 0.0454, "step": 2992 }, { "epoch": 0.20567619571192963, "grad_norm": 1.2423207759857178, "learning_rate": 2.7637475419594907e-05, "loss": 0.0485, "step": 2993 }, { "epoch": 0.20574491478834525, "grad_norm": 0.7979393005371094, "learning_rate": 2.7635676623568883e-05, "loss": 0.0351, "step": 2994 }, { "epoch": 0.20581363386476087, "grad_norm": 1.7802578210830688, "learning_rate": 2.763387720159741e-05, "loss": 0.0517, "step": 2995 }, { "epoch": 0.20588235294117646, "grad_norm": 1.2652949094772339, "learning_rate": 2.763207715376964e-05, "loss": 0.045, "step": 2996 }, { "epoch": 0.20595107201759208, "grad_norm": 1.0428534746170044, "learning_rate": 2.7630276480174734e-05, "loss": 0.053, "step": 2997 }, { "epoch": 0.2060197910940077, "grad_norm": 0.8516495227813721, "learning_rate": 2.76284751809019e-05, "loss": 0.0329, "step": 2998 }, { "epoch": 0.20608851017042332, "grad_norm": 0.996779203414917, "learning_rate": 2.7626673256040368e-05, "loss": 0.0819, "step": 2999 }, { "epoch": 0.20615722924683894, "grad_norm": 1.1446577310562134, "learning_rate": 2.7624870705679403e-05, "loss": 0.0412, "step": 3000 }, { "epoch": 0.20622594832325453, "grad_norm": 0.8473184704780579, "learning_rate": 2.7623067529908294e-05, "loss": 0.0401, "step": 3001 }, { "epoch": 0.20629466739967015, "grad_norm": 2.090916395187378, "learning_rate": 2.7621263728816382e-05, "loss": 0.0707, "step": 3002 }, { "epoch": 0.20636338647608576, "grad_norm": 0.7315865755081177, "learning_rate": 2.7619459302493002e-05, "loss": 0.0464, "step": 3003 }, { "epoch": 0.20643210555250138, "grad_norm": 1.2782108783721924, "learning_rate": 2.7617654251027557e-05, "loss": 0.0505, "step": 3004 }, { "epoch": 0.20650082462891697, "grad_norm": 1.447389841079712, "learning_rate": 2.7615848574509463e-05, "loss": 0.0623, "step": 3005 }, { "epoch": 0.2065695437053326, "grad_norm": 1.1059045791625977, "learning_rate": 2.7614042273028167e-05, "loss": 0.0551, "step": 3006 }, { "epoch": 0.2066382627817482, "grad_norm": 0.942837655544281, "learning_rate": 2.761223534667315e-05, "loss": 0.035, "step": 3007 }, { "epoch": 0.20670698185816383, "grad_norm": 1.378036618232727, "learning_rate": 2.761042779553392e-05, "loss": 0.054, "step": 3008 }, { "epoch": 0.20677570093457945, "grad_norm": 1.4676960706710815, "learning_rate": 2.7608619619700028e-05, "loss": 0.0689, "step": 3009 }, { "epoch": 0.20684442001099504, "grad_norm": 1.2063815593719482, "learning_rate": 2.7606810819261043e-05, "loss": 0.0544, "step": 3010 }, { "epoch": 0.20691313908741066, "grad_norm": 0.8735116720199585, "learning_rate": 2.760500139430657e-05, "loss": 0.0554, "step": 3011 }, { "epoch": 0.20698185816382628, "grad_norm": 1.1505885124206543, "learning_rate": 2.7603191344926235e-05, "loss": 0.0511, "step": 3012 }, { "epoch": 0.2070505772402419, "grad_norm": 1.0799705982208252, "learning_rate": 2.7601380671209712e-05, "loss": 0.056, "step": 3013 }, { "epoch": 0.20711929631665751, "grad_norm": 1.0938384532928467, "learning_rate": 2.7599569373246702e-05, "loss": 0.053, "step": 3014 }, { "epoch": 0.2071880153930731, "grad_norm": 1.3319666385650635, "learning_rate": 2.759775745112693e-05, "loss": 0.0444, "step": 3015 }, { "epoch": 0.20725673446948872, "grad_norm": 1.6081442832946777, "learning_rate": 2.7595944904940153e-05, "loss": 0.0552, "step": 3016 }, { "epoch": 0.20732545354590434, "grad_norm": 0.9151066541671753, "learning_rate": 2.759413173477616e-05, "loss": 0.0518, "step": 3017 }, { "epoch": 0.20739417262231996, "grad_norm": 1.3169480562210083, "learning_rate": 2.7592317940724773e-05, "loss": 0.0497, "step": 3018 }, { "epoch": 0.20746289169873558, "grad_norm": 1.2270575761795044, "learning_rate": 2.759050352287585e-05, "loss": 0.0522, "step": 3019 }, { "epoch": 0.20753161077515117, "grad_norm": 3.180405855178833, "learning_rate": 2.758868848131926e-05, "loss": 0.0505, "step": 3020 }, { "epoch": 0.2076003298515668, "grad_norm": 1.0131627321243286, "learning_rate": 2.7586872816144927e-05, "loss": 0.0428, "step": 3021 }, { "epoch": 0.2076690489279824, "grad_norm": 0.9795428514480591, "learning_rate": 2.7585056527442786e-05, "loss": 0.0411, "step": 3022 }, { "epoch": 0.20773776800439803, "grad_norm": 1.913673758506775, "learning_rate": 2.7583239615302826e-05, "loss": 0.0629, "step": 3023 }, { "epoch": 0.20780648708081365, "grad_norm": 0.6829950213432312, "learning_rate": 2.7581422079815044e-05, "loss": 0.044, "step": 3024 }, { "epoch": 0.20787520615722924, "grad_norm": 1.343624234199524, "learning_rate": 2.7579603921069476e-05, "loss": 0.0478, "step": 3025 }, { "epoch": 0.20794392523364486, "grad_norm": 0.8270350694656372, "learning_rate": 2.7577785139156198e-05, "loss": 0.0345, "step": 3026 }, { "epoch": 0.20801264431006047, "grad_norm": 0.636345624923706, "learning_rate": 2.7575965734165298e-05, "loss": 0.0508, "step": 3027 }, { "epoch": 0.2080813633864761, "grad_norm": 1.0271865129470825, "learning_rate": 2.757414570618691e-05, "loss": 0.0453, "step": 3028 }, { "epoch": 0.2081500824628917, "grad_norm": 0.9893544316291809, "learning_rate": 2.7572325055311196e-05, "loss": 0.0459, "step": 3029 }, { "epoch": 0.2082188015393073, "grad_norm": 1.2790910005569458, "learning_rate": 2.7570503781628347e-05, "loss": 0.0327, "step": 3030 }, { "epoch": 0.20828752061572292, "grad_norm": 0.8714128136634827, "learning_rate": 2.7568681885228584e-05, "loss": 0.0364, "step": 3031 }, { "epoch": 0.20835623969213854, "grad_norm": 1.0849310159683228, "learning_rate": 2.7566859366202162e-05, "loss": 0.0402, "step": 3032 }, { "epoch": 0.20842495876855416, "grad_norm": 0.979627251625061, "learning_rate": 2.7565036224639365e-05, "loss": 0.0305, "step": 3033 }, { "epoch": 0.20849367784496978, "grad_norm": 1.5448232889175415, "learning_rate": 2.7563212460630503e-05, "loss": 0.0464, "step": 3034 }, { "epoch": 0.20856239692138537, "grad_norm": 1.6693735122680664, "learning_rate": 2.7561388074265925e-05, "loss": 0.0366, "step": 3035 }, { "epoch": 0.208631115997801, "grad_norm": 0.8998820185661316, "learning_rate": 2.7559563065636005e-05, "loss": 0.0328, "step": 3036 }, { "epoch": 0.2086998350742166, "grad_norm": 0.9318965673446655, "learning_rate": 2.7557737434831157e-05, "loss": 0.0422, "step": 3037 }, { "epoch": 0.20876855415063222, "grad_norm": 1.4904453754425049, "learning_rate": 2.755591118194181e-05, "loss": 0.0388, "step": 3038 }, { "epoch": 0.20883727322704781, "grad_norm": 0.828437864780426, "learning_rate": 2.755408430705844e-05, "loss": 0.0475, "step": 3039 }, { "epoch": 0.20890599230346343, "grad_norm": 1.273056149482727, "learning_rate": 2.7552256810271538e-05, "loss": 0.0383, "step": 3040 }, { "epoch": 0.20897471137987905, "grad_norm": 1.1445339918136597, "learning_rate": 2.7550428691671644e-05, "loss": 0.0532, "step": 3041 }, { "epoch": 0.20904343045629467, "grad_norm": 1.8509985208511353, "learning_rate": 2.7548599951349317e-05, "loss": 0.0422, "step": 3042 }, { "epoch": 0.2091121495327103, "grad_norm": 1.1770884990692139, "learning_rate": 2.7546770589395144e-05, "loss": 0.0426, "step": 3043 }, { "epoch": 0.20918086860912588, "grad_norm": 1.3674578666687012, "learning_rate": 2.7544940605899754e-05, "loss": 0.0439, "step": 3044 }, { "epoch": 0.2092495876855415, "grad_norm": 1.6288939714431763, "learning_rate": 2.7543110000953798e-05, "loss": 0.0581, "step": 3045 }, { "epoch": 0.20931830676195712, "grad_norm": 1.2283129692077637, "learning_rate": 2.7541278774647958e-05, "loss": 0.0355, "step": 3046 }, { "epoch": 0.20938702583837274, "grad_norm": 1.7731266021728516, "learning_rate": 2.7539446927072948e-05, "loss": 0.0768, "step": 3047 }, { "epoch": 0.20945574491478836, "grad_norm": 1.264345407485962, "learning_rate": 2.753761445831952e-05, "loss": 0.0469, "step": 3048 }, { "epoch": 0.20952446399120395, "grad_norm": 1.4973105192184448, "learning_rate": 2.753578136847845e-05, "loss": 0.0581, "step": 3049 }, { "epoch": 0.20959318306761957, "grad_norm": 1.4493516683578491, "learning_rate": 2.7533947657640544e-05, "loss": 0.0549, "step": 3050 }, { "epoch": 0.20966190214403518, "grad_norm": 1.2328370809555054, "learning_rate": 2.7532113325896637e-05, "loss": 0.0545, "step": 3051 }, { "epoch": 0.2097306212204508, "grad_norm": 1.0460234880447388, "learning_rate": 2.75302783733376e-05, "loss": 0.0365, "step": 3052 }, { "epoch": 0.20979934029686642, "grad_norm": 1.2802259922027588, "learning_rate": 2.7528442800054335e-05, "loss": 0.0428, "step": 3053 }, { "epoch": 0.209868059373282, "grad_norm": 0.8330867886543274, "learning_rate": 2.752660660613777e-05, "loss": 0.0355, "step": 3054 }, { "epoch": 0.20993677844969763, "grad_norm": 1.555443525314331, "learning_rate": 2.7524769791678866e-05, "loss": 0.0569, "step": 3055 }, { "epoch": 0.21000549752611325, "grad_norm": 0.9402831196784973, "learning_rate": 2.7522932356768625e-05, "loss": 0.038, "step": 3056 }, { "epoch": 0.21007421660252887, "grad_norm": 1.1171488761901855, "learning_rate": 2.752109430149805e-05, "loss": 0.0527, "step": 3057 }, { "epoch": 0.2101429356789445, "grad_norm": 1.0672588348388672, "learning_rate": 2.7519255625958215e-05, "loss": 0.0373, "step": 3058 }, { "epoch": 0.21021165475536008, "grad_norm": 0.7345482707023621, "learning_rate": 2.751741633024019e-05, "loss": 0.0345, "step": 3059 }, { "epoch": 0.2102803738317757, "grad_norm": 0.9882217049598694, "learning_rate": 2.7515576414435094e-05, "loss": 0.0388, "step": 3060 }, { "epoch": 0.21034909290819132, "grad_norm": 2.4023895263671875, "learning_rate": 2.7513735878634072e-05, "loss": 0.0547, "step": 3061 }, { "epoch": 0.21041781198460693, "grad_norm": 1.2425806522369385, "learning_rate": 2.751189472292831e-05, "loss": 0.0349, "step": 3062 }, { "epoch": 0.21048653106102255, "grad_norm": 1.8525137901306152, "learning_rate": 2.7510052947409e-05, "loss": 0.0456, "step": 3063 }, { "epoch": 0.21055525013743814, "grad_norm": 1.4328776597976685, "learning_rate": 2.7508210552167384e-05, "loss": 0.04, "step": 3064 }, { "epoch": 0.21062396921385376, "grad_norm": 0.9529500007629395, "learning_rate": 2.7506367537294738e-05, "loss": 0.061, "step": 3065 }, { "epoch": 0.21069268829026938, "grad_norm": 0.957883894443512, "learning_rate": 2.7504523902882353e-05, "loss": 0.0513, "step": 3066 }, { "epoch": 0.210761407366685, "grad_norm": 0.9474459290504456, "learning_rate": 2.7502679649021563e-05, "loss": 0.0427, "step": 3067 }, { "epoch": 0.21083012644310062, "grad_norm": 1.3442144393920898, "learning_rate": 2.750083477580373e-05, "loss": 0.0344, "step": 3068 }, { "epoch": 0.2108988455195162, "grad_norm": 1.6189687252044678, "learning_rate": 2.749898928332024e-05, "loss": 0.0538, "step": 3069 }, { "epoch": 0.21096756459593183, "grad_norm": 1.3843034505844116, "learning_rate": 2.7497143171662522e-05, "loss": 0.0571, "step": 3070 }, { "epoch": 0.21103628367234745, "grad_norm": 0.6813713312149048, "learning_rate": 2.749529644092202e-05, "loss": 0.0481, "step": 3071 }, { "epoch": 0.21110500274876307, "grad_norm": 1.7099080085754395, "learning_rate": 2.7493449091190223e-05, "loss": 0.0484, "step": 3072 }, { "epoch": 0.21117372182517866, "grad_norm": 1.4828253984451294, "learning_rate": 2.749160112255864e-05, "loss": 0.0459, "step": 3073 }, { "epoch": 0.21124244090159428, "grad_norm": 0.7831217050552368, "learning_rate": 2.7489752535118825e-05, "loss": 0.0365, "step": 3074 }, { "epoch": 0.2113111599780099, "grad_norm": 0.9743103981018066, "learning_rate": 2.748790332896234e-05, "loss": 0.039, "step": 3075 }, { "epoch": 0.2113798790544255, "grad_norm": 1.0750056505203247, "learning_rate": 2.7486053504180806e-05, "loss": 0.0376, "step": 3076 }, { "epoch": 0.21144859813084113, "grad_norm": 1.2795988321304321, "learning_rate": 2.748420306086585e-05, "loss": 0.0408, "step": 3077 }, { "epoch": 0.21151731720725672, "grad_norm": 0.5172474980354309, "learning_rate": 2.748235199910914e-05, "loss": 0.0302, "step": 3078 }, { "epoch": 0.21158603628367234, "grad_norm": 1.0438532829284668, "learning_rate": 2.748050031900237e-05, "loss": 0.0506, "step": 3079 }, { "epoch": 0.21165475536008796, "grad_norm": 1.1010041236877441, "learning_rate": 2.747864802063728e-05, "loss": 0.0408, "step": 3080 }, { "epoch": 0.21172347443650358, "grad_norm": 0.7648643851280212, "learning_rate": 2.747679510410562e-05, "loss": 0.0319, "step": 3081 }, { "epoch": 0.2117921935129192, "grad_norm": 2.1232244968414307, "learning_rate": 2.747494156949918e-05, "loss": 0.0332, "step": 3082 }, { "epoch": 0.2118609125893348, "grad_norm": 1.0001654624938965, "learning_rate": 2.7473087416909783e-05, "loss": 0.0395, "step": 3083 }, { "epoch": 0.2119296316657504, "grad_norm": 1.0357568264007568, "learning_rate": 2.7471232646429277e-05, "loss": 0.0429, "step": 3084 }, { "epoch": 0.21199835074216603, "grad_norm": 0.7517557144165039, "learning_rate": 2.7469377258149546e-05, "loss": 0.0368, "step": 3085 }, { "epoch": 0.21206706981858164, "grad_norm": 1.4018305540084839, "learning_rate": 2.7467521252162503e-05, "loss": 0.028, "step": 3086 }, { "epoch": 0.21213578889499726, "grad_norm": 1.0789833068847656, "learning_rate": 2.746566462856009e-05, "loss": 0.054, "step": 3087 }, { "epoch": 0.21220450797141285, "grad_norm": 1.0763070583343506, "learning_rate": 2.7463807387434278e-05, "loss": 0.0312, "step": 3088 }, { "epoch": 0.21227322704782847, "grad_norm": 0.9114823341369629, "learning_rate": 2.7461949528877072e-05, "loss": 0.0455, "step": 3089 }, { "epoch": 0.2123419461242441, "grad_norm": 4.11460018157959, "learning_rate": 2.746009105298051e-05, "loss": 0.0533, "step": 3090 }, { "epoch": 0.2124106652006597, "grad_norm": 1.3531087636947632, "learning_rate": 2.7458231959836654e-05, "loss": 0.0486, "step": 3091 }, { "epoch": 0.21247938427707533, "grad_norm": 2.6537625789642334, "learning_rate": 2.7456372249537597e-05, "loss": 0.0444, "step": 3092 }, { "epoch": 0.21254810335349092, "grad_norm": 1.870766282081604, "learning_rate": 2.7454511922175466e-05, "loss": 0.0484, "step": 3093 }, { "epoch": 0.21261682242990654, "grad_norm": 1.207628607749939, "learning_rate": 2.7452650977842426e-05, "loss": 0.0369, "step": 3094 }, { "epoch": 0.21268554150632216, "grad_norm": 1.3981472253799438, "learning_rate": 2.7450789416630653e-05, "loss": 0.0492, "step": 3095 }, { "epoch": 0.21275426058273778, "grad_norm": 1.383746862411499, "learning_rate": 2.744892723863237e-05, "loss": 0.0449, "step": 3096 }, { "epoch": 0.2128229796591534, "grad_norm": 2.010359287261963, "learning_rate": 2.7447064443939825e-05, "loss": 0.0406, "step": 3097 }, { "epoch": 0.21289169873556898, "grad_norm": 1.1299902200698853, "learning_rate": 2.7445201032645302e-05, "loss": 0.0328, "step": 3098 }, { "epoch": 0.2129604178119846, "grad_norm": 0.990871787071228, "learning_rate": 2.74433370048411e-05, "loss": 0.0481, "step": 3099 }, { "epoch": 0.21302913688840022, "grad_norm": 1.7450649738311768, "learning_rate": 2.7441472360619576e-05, "loss": 0.0584, "step": 3100 }, { "epoch": 0.21309785596481584, "grad_norm": 2.010403633117676, "learning_rate": 2.7439607100073083e-05, "loss": 0.058, "step": 3101 }, { "epoch": 0.21316657504123146, "grad_norm": 1.2988094091415405, "learning_rate": 2.7437741223294025e-05, "loss": 0.0468, "step": 3102 }, { "epoch": 0.21323529411764705, "grad_norm": 1.0751229524612427, "learning_rate": 2.7435874730374845e-05, "loss": 0.0325, "step": 3103 }, { "epoch": 0.21330401319406267, "grad_norm": 0.9470043182373047, "learning_rate": 2.7434007621407996e-05, "loss": 0.0387, "step": 3104 }, { "epoch": 0.2133727322704783, "grad_norm": 1.5195015668869019, "learning_rate": 2.743213989648597e-05, "loss": 0.0408, "step": 3105 }, { "epoch": 0.2134414513468939, "grad_norm": 0.5929848551750183, "learning_rate": 2.7430271555701296e-05, "loss": 0.0492, "step": 3106 }, { "epoch": 0.2135101704233095, "grad_norm": 1.7893264293670654, "learning_rate": 2.7428402599146518e-05, "loss": 0.0355, "step": 3107 }, { "epoch": 0.21357888949972512, "grad_norm": 1.0196316242218018, "learning_rate": 2.7426533026914234e-05, "loss": 0.0403, "step": 3108 }, { "epoch": 0.21364760857614074, "grad_norm": 0.9632266163825989, "learning_rate": 2.742466283909705e-05, "loss": 0.0354, "step": 3109 }, { "epoch": 0.21371632765255635, "grad_norm": 1.311361312866211, "learning_rate": 2.7422792035787612e-05, "loss": 0.0405, "step": 3110 }, { "epoch": 0.21378504672897197, "grad_norm": 0.9744505286216736, "learning_rate": 2.7420920617078595e-05, "loss": 0.0397, "step": 3111 }, { "epoch": 0.21385376580538756, "grad_norm": 1.1951196193695068, "learning_rate": 2.7419048583062713e-05, "loss": 0.0566, "step": 3112 }, { "epoch": 0.21392248488180318, "grad_norm": 2.0462372303009033, "learning_rate": 2.7417175933832693e-05, "loss": 0.0389, "step": 3113 }, { "epoch": 0.2139912039582188, "grad_norm": 1.2195264101028442, "learning_rate": 2.7415302669481303e-05, "loss": 0.0643, "step": 3114 }, { "epoch": 0.21405992303463442, "grad_norm": 0.9834181070327759, "learning_rate": 2.741342879010135e-05, "loss": 0.0433, "step": 3115 }, { "epoch": 0.21412864211105004, "grad_norm": 0.942222535610199, "learning_rate": 2.741155429578565e-05, "loss": 0.0442, "step": 3116 }, { "epoch": 0.21419736118746563, "grad_norm": 1.7656916379928589, "learning_rate": 2.740967918662707e-05, "loss": 0.0503, "step": 3117 }, { "epoch": 0.21426608026388125, "grad_norm": 1.1902153491973877, "learning_rate": 2.7407803462718497e-05, "loss": 0.0428, "step": 3118 }, { "epoch": 0.21433479934029687, "grad_norm": 0.8284803032875061, "learning_rate": 2.7405927124152846e-05, "loss": 0.0615, "step": 3119 }, { "epoch": 0.21440351841671249, "grad_norm": 0.9767499566078186, "learning_rate": 2.7404050171023075e-05, "loss": 0.0375, "step": 3120 }, { "epoch": 0.2144722374931281, "grad_norm": 1.096647024154663, "learning_rate": 2.7402172603422155e-05, "loss": 0.0441, "step": 3121 }, { "epoch": 0.2145409565695437, "grad_norm": 0.8211801648139954, "learning_rate": 2.7400294421443106e-05, "loss": 0.0278, "step": 3122 }, { "epoch": 0.2146096756459593, "grad_norm": 2.0783677101135254, "learning_rate": 2.7398415625178968e-05, "loss": 0.0516, "step": 3123 }, { "epoch": 0.21467839472237493, "grad_norm": 1.452165961265564, "learning_rate": 2.7396536214722804e-05, "loss": 0.0423, "step": 3124 }, { "epoch": 0.21474711379879055, "grad_norm": 1.458389401435852, "learning_rate": 2.7394656190167723e-05, "loss": 0.0416, "step": 3125 }, { "epoch": 0.21481583287520617, "grad_norm": 1.5500761270523071, "learning_rate": 2.739277555160686e-05, "loss": 0.0347, "step": 3126 }, { "epoch": 0.21488455195162176, "grad_norm": 1.5594626665115356, "learning_rate": 2.7390894299133374e-05, "loss": 0.0466, "step": 3127 }, { "epoch": 0.21495327102803738, "grad_norm": 1.7709678411483765, "learning_rate": 2.7389012432840457e-05, "loss": 0.0567, "step": 3128 }, { "epoch": 0.215021990104453, "grad_norm": 0.8056440353393555, "learning_rate": 2.738712995282134e-05, "loss": 0.0485, "step": 3129 }, { "epoch": 0.21509070918086862, "grad_norm": 0.7576759457588196, "learning_rate": 2.7385246859169265e-05, "loss": 0.0333, "step": 3130 }, { "epoch": 0.21515942825728424, "grad_norm": 0.8189021944999695, "learning_rate": 2.7383363151977526e-05, "loss": 0.0491, "step": 3131 }, { "epoch": 0.21522814733369983, "grad_norm": 1.5394901037216187, "learning_rate": 2.7381478831339434e-05, "loss": 0.0427, "step": 3132 }, { "epoch": 0.21529686641011545, "grad_norm": 1.1597689390182495, "learning_rate": 2.7379593897348343e-05, "loss": 0.0424, "step": 3133 }, { "epoch": 0.21536558548653106, "grad_norm": 0.7852223515510559, "learning_rate": 2.7377708350097615e-05, "loss": 0.0345, "step": 3134 }, { "epoch": 0.21543430456294668, "grad_norm": 0.7745534181594849, "learning_rate": 2.7375822189680662e-05, "loss": 0.042, "step": 3135 }, { "epoch": 0.21550302363936227, "grad_norm": 1.3566057682037354, "learning_rate": 2.7373935416190924e-05, "loss": 0.0449, "step": 3136 }, { "epoch": 0.2155717427157779, "grad_norm": 1.9621009826660156, "learning_rate": 2.7372048029721864e-05, "loss": 0.055, "step": 3137 }, { "epoch": 0.2156404617921935, "grad_norm": 1.2182985544204712, "learning_rate": 2.7370160030366983e-05, "loss": 0.0419, "step": 3138 }, { "epoch": 0.21570918086860913, "grad_norm": 1.8893773555755615, "learning_rate": 2.7368271418219805e-05, "loss": 0.0604, "step": 3139 }, { "epoch": 0.21577789994502475, "grad_norm": 1.0998183488845825, "learning_rate": 2.736638219337389e-05, "loss": 0.0533, "step": 3140 }, { "epoch": 0.21584661902144034, "grad_norm": 1.31802499294281, "learning_rate": 2.7364492355922824e-05, "loss": 0.0382, "step": 3141 }, { "epoch": 0.21591533809785596, "grad_norm": 1.8788797855377197, "learning_rate": 2.7362601905960224e-05, "loss": 0.0467, "step": 3142 }, { "epoch": 0.21598405717427158, "grad_norm": 1.0619666576385498, "learning_rate": 2.7360710843579748e-05, "loss": 0.0341, "step": 3143 }, { "epoch": 0.2160527762506872, "grad_norm": 1.110669732093811, "learning_rate": 2.7358819168875064e-05, "loss": 0.0394, "step": 3144 }, { "epoch": 0.2161214953271028, "grad_norm": 1.5849554538726807, "learning_rate": 2.7356926881939893e-05, "loss": 0.0542, "step": 3145 }, { "epoch": 0.2161902144035184, "grad_norm": 1.4868040084838867, "learning_rate": 2.735503398286796e-05, "loss": 0.0602, "step": 3146 }, { "epoch": 0.21625893347993402, "grad_norm": 1.1772488355636597, "learning_rate": 2.7353140471753052e-05, "loss": 0.0596, "step": 3147 }, { "epoch": 0.21632765255634964, "grad_norm": 1.7242767810821533, "learning_rate": 2.7351246348688962e-05, "loss": 0.0702, "step": 3148 }, { "epoch": 0.21639637163276526, "grad_norm": 1.1357096433639526, "learning_rate": 2.734935161376952e-05, "loss": 0.0664, "step": 3149 }, { "epoch": 0.21646509070918088, "grad_norm": 1.0662400722503662, "learning_rate": 2.7347456267088585e-05, "loss": 0.0468, "step": 3150 }, { "epoch": 0.21653380978559647, "grad_norm": 0.8211081624031067, "learning_rate": 2.7345560308740052e-05, "loss": 0.0343, "step": 3151 }, { "epoch": 0.2166025288620121, "grad_norm": 0.746798038482666, "learning_rate": 2.7343663738817847e-05, "loss": 0.049, "step": 3152 }, { "epoch": 0.2166712479384277, "grad_norm": 2.114423990249634, "learning_rate": 2.7341766557415913e-05, "loss": 0.0453, "step": 3153 }, { "epoch": 0.21673996701484333, "grad_norm": 1.2113323211669922, "learning_rate": 2.733986876462824e-05, "loss": 0.0525, "step": 3154 }, { "epoch": 0.21680868609125895, "grad_norm": 1.3494154214859009, "learning_rate": 2.7337970360548837e-05, "loss": 0.037, "step": 3155 }, { "epoch": 0.21687740516767454, "grad_norm": 1.5283114910125732, "learning_rate": 2.7336071345271752e-05, "loss": 0.0311, "step": 3156 }, { "epoch": 0.21694612424409015, "grad_norm": 0.9674059152603149, "learning_rate": 2.733417171889105e-05, "loss": 0.0388, "step": 3157 }, { "epoch": 0.21701484332050577, "grad_norm": 1.2749731540679932, "learning_rate": 2.7332271481500842e-05, "loss": 0.0391, "step": 3158 }, { "epoch": 0.2170835623969214, "grad_norm": 1.43831467628479, "learning_rate": 2.7330370633195258e-05, "loss": 0.0415, "step": 3159 }, { "epoch": 0.217152281473337, "grad_norm": 1.3091551065444946, "learning_rate": 2.732846917406846e-05, "loss": 0.0462, "step": 3160 }, { "epoch": 0.2172210005497526, "grad_norm": 2.5949907302856445, "learning_rate": 2.7326567104214648e-05, "loss": 0.0571, "step": 3161 }, { "epoch": 0.21728971962616822, "grad_norm": 1.848956823348999, "learning_rate": 2.7324664423728047e-05, "loss": 0.0513, "step": 3162 }, { "epoch": 0.21735843870258384, "grad_norm": 0.8890243768692017, "learning_rate": 2.73227611327029e-05, "loss": 0.0445, "step": 3163 }, { "epoch": 0.21742715777899946, "grad_norm": 1.2360142469406128, "learning_rate": 2.7320857231233507e-05, "loss": 0.0461, "step": 3164 }, { "epoch": 0.21749587685541508, "grad_norm": 1.7661575078964233, "learning_rate": 2.7318952719414175e-05, "loss": 0.0472, "step": 3165 }, { "epoch": 0.21756459593183067, "grad_norm": 1.676020860671997, "learning_rate": 2.7317047597339257e-05, "loss": 0.0391, "step": 3166 }, { "epoch": 0.2176333150082463, "grad_norm": 1.1465140581130981, "learning_rate": 2.7315141865103122e-05, "loss": 0.0471, "step": 3167 }, { "epoch": 0.2177020340846619, "grad_norm": 0.9036042094230652, "learning_rate": 2.7313235522800177e-05, "loss": 0.0584, "step": 3168 }, { "epoch": 0.21777075316107752, "grad_norm": 0.9857297539710999, "learning_rate": 2.731132857052486e-05, "loss": 0.0553, "step": 3169 }, { "epoch": 0.21783947223749311, "grad_norm": 1.3122167587280273, "learning_rate": 2.7309421008371634e-05, "loss": 0.043, "step": 3170 }, { "epoch": 0.21790819131390873, "grad_norm": 1.8827903270721436, "learning_rate": 2.7307512836434995e-05, "loss": 0.0439, "step": 3171 }, { "epoch": 0.21797691039032435, "grad_norm": 1.7097088098526, "learning_rate": 2.7305604054809482e-05, "loss": 0.062, "step": 3172 }, { "epoch": 0.21804562946673997, "grad_norm": 2.069528579711914, "learning_rate": 2.730369466358964e-05, "loss": 0.0573, "step": 3173 }, { "epoch": 0.2181143485431556, "grad_norm": 1.279284119606018, "learning_rate": 2.7301784662870058e-05, "loss": 0.0515, "step": 3174 }, { "epoch": 0.21818306761957118, "grad_norm": 1.3021090030670166, "learning_rate": 2.7299874052745356e-05, "loss": 0.0414, "step": 3175 }, { "epoch": 0.2182517866959868, "grad_norm": 1.214749813079834, "learning_rate": 2.7297962833310178e-05, "loss": 0.0327, "step": 3176 }, { "epoch": 0.21832050577240242, "grad_norm": 1.5386276245117188, "learning_rate": 2.729605100465921e-05, "loss": 0.0344, "step": 3177 }, { "epoch": 0.21838922484881804, "grad_norm": 1.0525346994400024, "learning_rate": 2.7294138566887155e-05, "loss": 0.0415, "step": 3178 }, { "epoch": 0.21845794392523366, "grad_norm": 1.4912073612213135, "learning_rate": 2.729222552008875e-05, "loss": 0.0454, "step": 3179 }, { "epoch": 0.21852666300164925, "grad_norm": 1.290885090827942, "learning_rate": 2.729031186435876e-05, "loss": 0.0354, "step": 3180 }, { "epoch": 0.21859538207806486, "grad_norm": 1.213274598121643, "learning_rate": 2.7288397599791988e-05, "loss": 0.0323, "step": 3181 }, { "epoch": 0.21866410115448048, "grad_norm": 1.528519630432129, "learning_rate": 2.728648272648327e-05, "loss": 0.0379, "step": 3182 }, { "epoch": 0.2187328202308961, "grad_norm": 0.8909437656402588, "learning_rate": 2.7284567244527457e-05, "loss": 0.0446, "step": 3183 }, { "epoch": 0.21880153930731172, "grad_norm": 1.0966384410858154, "learning_rate": 2.728265115401943e-05, "loss": 0.0286, "step": 3184 }, { "epoch": 0.2188702583837273, "grad_norm": 2.1298069953918457, "learning_rate": 2.7280734455054123e-05, "loss": 0.0502, "step": 3185 }, { "epoch": 0.21893897746014293, "grad_norm": 0.5854522585868835, "learning_rate": 2.727881714772648e-05, "loss": 0.0442, "step": 3186 }, { "epoch": 0.21900769653655855, "grad_norm": 1.149812936782837, "learning_rate": 2.7276899232131477e-05, "loss": 0.0475, "step": 3187 }, { "epoch": 0.21907641561297417, "grad_norm": 1.3229223489761353, "learning_rate": 2.7274980708364127e-05, "loss": 0.0514, "step": 3188 }, { "epoch": 0.2191451346893898, "grad_norm": 1.961092233657837, "learning_rate": 2.7273061576519472e-05, "loss": 0.0368, "step": 3189 }, { "epoch": 0.21921385376580538, "grad_norm": 0.9956019520759583, "learning_rate": 2.7271141836692575e-05, "loss": 0.0437, "step": 3190 }, { "epoch": 0.219282572842221, "grad_norm": 0.769594669342041, "learning_rate": 2.726922148897854e-05, "loss": 0.0432, "step": 3191 }, { "epoch": 0.21935129191863662, "grad_norm": 0.9625405669212341, "learning_rate": 2.7267300533472503e-05, "loss": 0.0449, "step": 3192 }, { "epoch": 0.21942001099505223, "grad_norm": 1.0529201030731201, "learning_rate": 2.7265378970269612e-05, "loss": 0.0462, "step": 3193 }, { "epoch": 0.21948873007146785, "grad_norm": 1.2228511571884155, "learning_rate": 2.726345679946506e-05, "loss": 0.0382, "step": 3194 }, { "epoch": 0.21955744914788344, "grad_norm": 1.1085925102233887, "learning_rate": 2.7261534021154078e-05, "loss": 0.0378, "step": 3195 }, { "epoch": 0.21962616822429906, "grad_norm": 1.2251794338226318, "learning_rate": 2.725961063543191e-05, "loss": 0.04, "step": 3196 }, { "epoch": 0.21969488730071468, "grad_norm": 0.6928436756134033, "learning_rate": 2.725768664239383e-05, "loss": 0.0348, "step": 3197 }, { "epoch": 0.2197636063771303, "grad_norm": 1.3141433000564575, "learning_rate": 2.7255762042135156e-05, "loss": 0.0448, "step": 3198 }, { "epoch": 0.21983232545354592, "grad_norm": 1.359401822090149, "learning_rate": 2.725383683475123e-05, "loss": 0.0473, "step": 3199 }, { "epoch": 0.2199010445299615, "grad_norm": 1.1182893514633179, "learning_rate": 2.7251911020337416e-05, "loss": 0.0466, "step": 3200 }, { "epoch": 0.21996976360637713, "grad_norm": 0.9471709728240967, "learning_rate": 2.7249984598989125e-05, "loss": 0.0339, "step": 3201 }, { "epoch": 0.22003848268279275, "grad_norm": 1.7448718547821045, "learning_rate": 2.7248057570801778e-05, "loss": 0.0417, "step": 3202 }, { "epoch": 0.22010720175920837, "grad_norm": 1.0808171033859253, "learning_rate": 2.7246129935870836e-05, "loss": 0.0421, "step": 3203 }, { "epoch": 0.22017592083562396, "grad_norm": 1.025705099105835, "learning_rate": 2.72442016942918e-05, "loss": 0.0523, "step": 3204 }, { "epoch": 0.22024463991203957, "grad_norm": 0.8548150062561035, "learning_rate": 2.7242272846160183e-05, "loss": 0.0436, "step": 3205 }, { "epoch": 0.2203133589884552, "grad_norm": 0.8444221019744873, "learning_rate": 2.7240343391571536e-05, "loss": 0.038, "step": 3206 }, { "epoch": 0.2203820780648708, "grad_norm": 1.1434239149093628, "learning_rate": 2.7238413330621444e-05, "loss": 0.0352, "step": 3207 }, { "epoch": 0.22045079714128643, "grad_norm": 0.8807713985443115, "learning_rate": 2.7236482663405515e-05, "loss": 0.0341, "step": 3208 }, { "epoch": 0.22051951621770202, "grad_norm": 0.9772242307662964, "learning_rate": 2.7234551390019395e-05, "loss": 0.0544, "step": 3209 }, { "epoch": 0.22058823529411764, "grad_norm": 0.7924942970275879, "learning_rate": 2.723261951055875e-05, "loss": 0.0357, "step": 3210 }, { "epoch": 0.22065695437053326, "grad_norm": 0.9278712868690491, "learning_rate": 2.723068702511929e-05, "loss": 0.0328, "step": 3211 }, { "epoch": 0.22072567344694888, "grad_norm": 1.2684375047683716, "learning_rate": 2.722875393379673e-05, "loss": 0.0326, "step": 3212 }, { "epoch": 0.2207943925233645, "grad_norm": 1.4467052221298218, "learning_rate": 2.722682023668685e-05, "loss": 0.0382, "step": 3213 }, { "epoch": 0.2208631115997801, "grad_norm": 1.548736572265625, "learning_rate": 2.722488593388543e-05, "loss": 0.0389, "step": 3214 }, { "epoch": 0.2209318306761957, "grad_norm": 1.4877725839614868, "learning_rate": 2.7222951025488295e-05, "loss": 0.0395, "step": 3215 }, { "epoch": 0.22100054975261132, "grad_norm": 2.1983072757720947, "learning_rate": 2.722101551159129e-05, "loss": 0.0434, "step": 3216 }, { "epoch": 0.22106926882902694, "grad_norm": 0.8509476780891418, "learning_rate": 2.721907939229031e-05, "loss": 0.0445, "step": 3217 }, { "epoch": 0.22113798790544256, "grad_norm": 0.9865605235099792, "learning_rate": 2.7217142667681248e-05, "loss": 0.0342, "step": 3218 }, { "epoch": 0.22120670698185815, "grad_norm": 1.20792818069458, "learning_rate": 2.7215205337860064e-05, "loss": 0.0647, "step": 3219 }, { "epoch": 0.22127542605827377, "grad_norm": 1.9122048616409302, "learning_rate": 2.721326740292272e-05, "loss": 0.0499, "step": 3220 }, { "epoch": 0.2213441451346894, "grad_norm": 0.7996428608894348, "learning_rate": 2.7211328862965215e-05, "loss": 0.0358, "step": 3221 }, { "epoch": 0.221412864211105, "grad_norm": 1.1536885499954224, "learning_rate": 2.7209389718083586e-05, "loss": 0.0471, "step": 3222 }, { "epoch": 0.22148158328752063, "grad_norm": 1.281006932258606, "learning_rate": 2.7207449968373894e-05, "loss": 0.0577, "step": 3223 }, { "epoch": 0.22155030236393622, "grad_norm": 0.8188950419425964, "learning_rate": 2.720550961393222e-05, "loss": 0.0382, "step": 3224 }, { "epoch": 0.22161902144035184, "grad_norm": 0.9659494161605835, "learning_rate": 2.72035686548547e-05, "loss": 0.0384, "step": 3225 }, { "epoch": 0.22168774051676746, "grad_norm": 1.247192144393921, "learning_rate": 2.720162709123748e-05, "loss": 0.0612, "step": 3226 }, { "epoch": 0.22175645959318308, "grad_norm": 1.1542953252792358, "learning_rate": 2.7199684923176733e-05, "loss": 0.0362, "step": 3227 }, { "epoch": 0.2218251786695987, "grad_norm": 1.3601678609848022, "learning_rate": 2.7197742150768684e-05, "loss": 0.0534, "step": 3228 }, { "epoch": 0.22189389774601428, "grad_norm": 1.4648994207382202, "learning_rate": 2.719579877410956e-05, "loss": 0.0581, "step": 3229 }, { "epoch": 0.2219626168224299, "grad_norm": 0.7330593466758728, "learning_rate": 2.719385479329564e-05, "loss": 0.0516, "step": 3230 }, { "epoch": 0.22203133589884552, "grad_norm": 1.0236200094223022, "learning_rate": 2.7191910208423224e-05, "loss": 0.0312, "step": 3231 }, { "epoch": 0.22210005497526114, "grad_norm": 1.1043821573257446, "learning_rate": 2.7189965019588643e-05, "loss": 0.0489, "step": 3232 }, { "epoch": 0.22216877405167673, "grad_norm": 1.3710819482803345, "learning_rate": 2.718801922688825e-05, "loss": 0.034, "step": 3233 }, { "epoch": 0.22223749312809235, "grad_norm": 1.0618040561676025, "learning_rate": 2.7186072830418447e-05, "loss": 0.0472, "step": 3234 }, { "epoch": 0.22230621220450797, "grad_norm": 0.9945096969604492, "learning_rate": 2.718412583027565e-05, "loss": 0.0604, "step": 3235 }, { "epoch": 0.2223749312809236, "grad_norm": 1.6910021305084229, "learning_rate": 2.7182178226556307e-05, "loss": 0.0589, "step": 3236 }, { "epoch": 0.2224436503573392, "grad_norm": 0.858880877494812, "learning_rate": 2.7180230019356898e-05, "loss": 0.04, "step": 3237 }, { "epoch": 0.2225123694337548, "grad_norm": 0.8068971037864685, "learning_rate": 2.717828120877394e-05, "loss": 0.0422, "step": 3238 }, { "epoch": 0.22258108851017042, "grad_norm": 1.7024118900299072, "learning_rate": 2.717633179490397e-05, "loss": 0.0538, "step": 3239 }, { "epoch": 0.22264980758658603, "grad_norm": 1.2728303670883179, "learning_rate": 2.7174381777843552e-05, "loss": 0.0449, "step": 3240 }, { "epoch": 0.22271852666300165, "grad_norm": 2.2945921421051025, "learning_rate": 2.717243115768929e-05, "loss": 0.0496, "step": 3241 }, { "epoch": 0.22278724573941727, "grad_norm": 1.3381421566009521, "learning_rate": 2.7170479934537816e-05, "loss": 0.0514, "step": 3242 }, { "epoch": 0.22285596481583286, "grad_norm": 1.8455127477645874, "learning_rate": 2.7168528108485787e-05, "loss": 0.0593, "step": 3243 }, { "epoch": 0.22292468389224848, "grad_norm": 1.3765519857406616, "learning_rate": 2.7166575679629892e-05, "loss": 0.0461, "step": 3244 }, { "epoch": 0.2229934029686641, "grad_norm": 1.2310959100723267, "learning_rate": 2.7164622648066854e-05, "loss": 0.0332, "step": 3245 }, { "epoch": 0.22306212204507972, "grad_norm": 3.651634931564331, "learning_rate": 2.716266901389342e-05, "loss": 0.0542, "step": 3246 }, { "epoch": 0.22313084112149534, "grad_norm": 1.3339638710021973, "learning_rate": 2.716071477720636e-05, "loss": 0.0587, "step": 3247 }, { "epoch": 0.22319956019791093, "grad_norm": 2.2297985553741455, "learning_rate": 2.71587599381025e-05, "loss": 0.0752, "step": 3248 }, { "epoch": 0.22326827927432655, "grad_norm": 0.6841418147087097, "learning_rate": 2.7156804496678665e-05, "loss": 0.0425, "step": 3249 }, { "epoch": 0.22333699835074217, "grad_norm": 1.89726722240448, "learning_rate": 2.7154848453031734e-05, "loss": 0.0686, "step": 3250 }, { "epoch": 0.22340571742715779, "grad_norm": 1.3189408779144287, "learning_rate": 2.7152891807258595e-05, "loss": 0.0628, "step": 3251 }, { "epoch": 0.2234744365035734, "grad_norm": 1.71876060962677, "learning_rate": 2.7150934559456182e-05, "loss": 0.0487, "step": 3252 }, { "epoch": 0.223543155579989, "grad_norm": 1.9255717992782593, "learning_rate": 2.7148976709721454e-05, "loss": 0.0582, "step": 3253 }, { "epoch": 0.2236118746564046, "grad_norm": 0.653676450252533, "learning_rate": 2.7147018258151392e-05, "loss": 0.0549, "step": 3254 }, { "epoch": 0.22368059373282023, "grad_norm": 1.0261456966400146, "learning_rate": 2.7145059204843022e-05, "loss": 0.052, "step": 3255 }, { "epoch": 0.22374931280923585, "grad_norm": 1.566209077835083, "learning_rate": 2.714309954989339e-05, "loss": 0.0536, "step": 3256 }, { "epoch": 0.22381803188565147, "grad_norm": 1.5968234539031982, "learning_rate": 2.7141139293399572e-05, "loss": 0.0432, "step": 3257 }, { "epoch": 0.22388675096206706, "grad_norm": 1.0007171630859375, "learning_rate": 2.713917843545867e-05, "loss": 0.0477, "step": 3258 }, { "epoch": 0.22395547003848268, "grad_norm": 1.2296082973480225, "learning_rate": 2.7137216976167828e-05, "loss": 0.0495, "step": 3259 }, { "epoch": 0.2240241891148983, "grad_norm": 0.6848344206809998, "learning_rate": 2.7135254915624213e-05, "loss": 0.0311, "step": 3260 }, { "epoch": 0.22409290819131392, "grad_norm": 0.8183065056800842, "learning_rate": 2.713329225392502e-05, "loss": 0.0383, "step": 3261 }, { "epoch": 0.22416162726772954, "grad_norm": 1.8142718076705933, "learning_rate": 2.7131328991167464e-05, "loss": 0.0584, "step": 3262 }, { "epoch": 0.22423034634414513, "grad_norm": 1.1571016311645508, "learning_rate": 2.7129365127448815e-05, "loss": 0.0419, "step": 3263 }, { "epoch": 0.22429906542056074, "grad_norm": 1.0168261528015137, "learning_rate": 2.7127400662866355e-05, "loss": 0.0395, "step": 3264 }, { "epoch": 0.22436778449697636, "grad_norm": 0.7115802764892578, "learning_rate": 2.7125435597517402e-05, "loss": 0.0382, "step": 3265 }, { "epoch": 0.22443650357339198, "grad_norm": 1.5999829769134521, "learning_rate": 2.7123469931499294e-05, "loss": 0.0411, "step": 3266 }, { "epoch": 0.22450522264980757, "grad_norm": 1.4033294916152954, "learning_rate": 2.7121503664909414e-05, "loss": 0.0456, "step": 3267 }, { "epoch": 0.2245739417262232, "grad_norm": 1.1362831592559814, "learning_rate": 2.7119536797845163e-05, "loss": 0.0371, "step": 3268 }, { "epoch": 0.2246426608026388, "grad_norm": 1.1019227504730225, "learning_rate": 2.7117569330403973e-05, "loss": 0.0341, "step": 3269 }, { "epoch": 0.22471137987905443, "grad_norm": 1.597404956817627, "learning_rate": 2.7115601262683312e-05, "loss": 0.045, "step": 3270 }, { "epoch": 0.22478009895547005, "grad_norm": 1.9643322229385376, "learning_rate": 2.7113632594780672e-05, "loss": 0.0488, "step": 3271 }, { "epoch": 0.22484881803188564, "grad_norm": 1.1240068674087524, "learning_rate": 2.7111663326793584e-05, "loss": 0.0484, "step": 3272 }, { "epoch": 0.22491753710830126, "grad_norm": 2.3377888202667236, "learning_rate": 2.7109693458819587e-05, "loss": 0.0469, "step": 3273 }, { "epoch": 0.22498625618471688, "grad_norm": 0.725979745388031, "learning_rate": 2.7107722990956278e-05, "loss": 0.0511, "step": 3274 }, { "epoch": 0.2250549752611325, "grad_norm": 1.0347405672073364, "learning_rate": 2.7105751923301264e-05, "loss": 0.0575, "step": 3275 }, { "epoch": 0.2251236943375481, "grad_norm": 1.2959707975387573, "learning_rate": 2.7103780255952186e-05, "loss": 0.0402, "step": 3276 }, { "epoch": 0.2251924134139637, "grad_norm": 0.9363229870796204, "learning_rate": 2.710180798900672e-05, "loss": 0.0377, "step": 3277 }, { "epoch": 0.22526113249037932, "grad_norm": 0.9590970277786255, "learning_rate": 2.7099835122562564e-05, "loss": 0.0525, "step": 3278 }, { "epoch": 0.22532985156679494, "grad_norm": 0.9008045792579651, "learning_rate": 2.7097861656717455e-05, "loss": 0.0333, "step": 3279 }, { "epoch": 0.22539857064321056, "grad_norm": 1.3644301891326904, "learning_rate": 2.709588759156915e-05, "loss": 0.0456, "step": 3280 }, { "epoch": 0.22546728971962618, "grad_norm": 1.2693053483963013, "learning_rate": 2.7093912927215442e-05, "loss": 0.0421, "step": 3281 }, { "epoch": 0.22553600879604177, "grad_norm": 1.524956226348877, "learning_rate": 2.7091937663754152e-05, "loss": 0.0606, "step": 3282 }, { "epoch": 0.2256047278724574, "grad_norm": 1.0346015691757202, "learning_rate": 2.7089961801283132e-05, "loss": 0.0465, "step": 3283 }, { "epoch": 0.225673446948873, "grad_norm": 1.2640851736068726, "learning_rate": 2.7087985339900255e-05, "loss": 0.0506, "step": 3284 }, { "epoch": 0.22574216602528863, "grad_norm": 1.0428715944290161, "learning_rate": 2.7086008279703442e-05, "loss": 0.0494, "step": 3285 }, { "epoch": 0.22581088510170425, "grad_norm": 1.1711864471435547, "learning_rate": 2.7084030620790622e-05, "loss": 0.0469, "step": 3286 }, { "epoch": 0.22587960417811984, "grad_norm": 1.1319031715393066, "learning_rate": 2.708205236325977e-05, "loss": 0.0492, "step": 3287 }, { "epoch": 0.22594832325453545, "grad_norm": 1.618682861328125, "learning_rate": 2.708007350720889e-05, "loss": 0.0543, "step": 3288 }, { "epoch": 0.22601704233095107, "grad_norm": 1.3776769638061523, "learning_rate": 2.7078094052735995e-05, "loss": 0.0589, "step": 3289 }, { "epoch": 0.2260857614073667, "grad_norm": 1.2162535190582275, "learning_rate": 2.7076113999939154e-05, "loss": 0.0492, "step": 3290 }, { "epoch": 0.2261544804837823, "grad_norm": 1.2585829496383667, "learning_rate": 2.7074133348916455e-05, "loss": 0.053, "step": 3291 }, { "epoch": 0.2262231995601979, "grad_norm": 1.686676025390625, "learning_rate": 2.707215209976601e-05, "loss": 0.0525, "step": 3292 }, { "epoch": 0.22629191863661352, "grad_norm": 1.9142059087753296, "learning_rate": 2.7070170252585972e-05, "loss": 0.0507, "step": 3293 }, { "epoch": 0.22636063771302914, "grad_norm": 1.2199969291687012, "learning_rate": 2.7068187807474517e-05, "loss": 0.0474, "step": 3294 }, { "epoch": 0.22642935678944476, "grad_norm": 1.3251670598983765, "learning_rate": 2.706620476452985e-05, "loss": 0.0419, "step": 3295 }, { "epoch": 0.22649807586586038, "grad_norm": 1.2522279024124146, "learning_rate": 2.70642211238502e-05, "loss": 0.0533, "step": 3296 }, { "epoch": 0.22656679494227597, "grad_norm": 1.1804466247558594, "learning_rate": 2.7062236885533842e-05, "loss": 0.0463, "step": 3297 }, { "epoch": 0.2266355140186916, "grad_norm": 1.2436082363128662, "learning_rate": 2.7060252049679066e-05, "loss": 0.0477, "step": 3298 }, { "epoch": 0.2267042330951072, "grad_norm": 0.899437427520752, "learning_rate": 2.70582666163842e-05, "loss": 0.0335, "step": 3299 }, { "epoch": 0.22677295217152282, "grad_norm": 0.9478182196617126, "learning_rate": 2.70562805857476e-05, "loss": 0.0462, "step": 3300 }, { "epoch": 0.22684167124793841, "grad_norm": 2.977025032043457, "learning_rate": 2.705429395786764e-05, "loss": 0.0426, "step": 3301 }, { "epoch": 0.22691039032435403, "grad_norm": 0.9013437628746033, "learning_rate": 2.7052306732842743e-05, "loss": 0.0337, "step": 3302 }, { "epoch": 0.22697910940076965, "grad_norm": 0.721847414970398, "learning_rate": 2.705031891077135e-05, "loss": 0.0457, "step": 3303 }, { "epoch": 0.22704782847718527, "grad_norm": 1.6962190866470337, "learning_rate": 2.704833049175193e-05, "loss": 0.054, "step": 3304 }, { "epoch": 0.2271165475536009, "grad_norm": 1.1617217063903809, "learning_rate": 2.704634147588299e-05, "loss": 0.0423, "step": 3305 }, { "epoch": 0.22718526663001648, "grad_norm": 1.0237623453140259, "learning_rate": 2.7044351863263055e-05, "loss": 0.0592, "step": 3306 }, { "epoch": 0.2272539857064321, "grad_norm": 1.4014972448349, "learning_rate": 2.7042361653990693e-05, "loss": 0.0487, "step": 3307 }, { "epoch": 0.22732270478284772, "grad_norm": 1.531789779663086, "learning_rate": 2.7040370848164496e-05, "loss": 0.0413, "step": 3308 }, { "epoch": 0.22739142385926334, "grad_norm": 1.0853909254074097, "learning_rate": 2.7038379445883078e-05, "loss": 0.0467, "step": 3309 }, { "epoch": 0.22746014293567896, "grad_norm": 0.9329037070274353, "learning_rate": 2.7036387447245094e-05, "loss": 0.0329, "step": 3310 }, { "epoch": 0.22752886201209455, "grad_norm": 1.8309367895126343, "learning_rate": 2.7034394852349218e-05, "loss": 0.0604, "step": 3311 }, { "epoch": 0.22759758108851016, "grad_norm": 2.2237839698791504, "learning_rate": 2.7032401661294165e-05, "loss": 0.036, "step": 3312 }, { "epoch": 0.22766630016492578, "grad_norm": 1.0009173154830933, "learning_rate": 2.7030407874178673e-05, "loss": 0.0364, "step": 3313 }, { "epoch": 0.2277350192413414, "grad_norm": 0.6675818562507629, "learning_rate": 2.7028413491101506e-05, "loss": 0.0379, "step": 3314 }, { "epoch": 0.22780373831775702, "grad_norm": 1.0151609182357788, "learning_rate": 2.702641851216147e-05, "loss": 0.0319, "step": 3315 }, { "epoch": 0.2278724573941726, "grad_norm": 1.5386906862258911, "learning_rate": 2.702442293745738e-05, "loss": 0.0359, "step": 3316 }, { "epoch": 0.22794117647058823, "grad_norm": 1.5931909084320068, "learning_rate": 2.70224267670881e-05, "loss": 0.0489, "step": 3317 }, { "epoch": 0.22800989554700385, "grad_norm": 1.4187982082366943, "learning_rate": 2.7020430001152515e-05, "loss": 0.0528, "step": 3318 }, { "epoch": 0.22807861462341947, "grad_norm": 1.2850663661956787, "learning_rate": 2.7018432639749538e-05, "loss": 0.057, "step": 3319 }, { "epoch": 0.2281473336998351, "grad_norm": 1.1146254539489746, "learning_rate": 2.7016434682978123e-05, "loss": 0.0294, "step": 3320 }, { "epoch": 0.22821605277625068, "grad_norm": 1.2762994766235352, "learning_rate": 2.7014436130937235e-05, "loss": 0.0505, "step": 3321 }, { "epoch": 0.2282847718526663, "grad_norm": 1.1489609479904175, "learning_rate": 2.701243698372588e-05, "loss": 0.0511, "step": 3322 }, { "epoch": 0.22835349092908191, "grad_norm": 0.8710072636604309, "learning_rate": 2.7010437241443095e-05, "loss": 0.0574, "step": 3323 }, { "epoch": 0.22842221000549753, "grad_norm": 1.0013772249221802, "learning_rate": 2.7008436904187938e-05, "loss": 0.0352, "step": 3324 }, { "epoch": 0.22849092908191315, "grad_norm": 1.7671576738357544, "learning_rate": 2.700643597205951e-05, "loss": 0.0685, "step": 3325 }, { "epoch": 0.22855964815832874, "grad_norm": 0.7607255578041077, "learning_rate": 2.7004434445156927e-05, "loss": 0.037, "step": 3326 }, { "epoch": 0.22862836723474436, "grad_norm": 1.3780465126037598, "learning_rate": 2.7002432323579344e-05, "loss": 0.0431, "step": 3327 }, { "epoch": 0.22869708631115998, "grad_norm": 0.5677679777145386, "learning_rate": 2.7000429607425937e-05, "loss": 0.0379, "step": 3328 }, { "epoch": 0.2287658053875756, "grad_norm": 0.7874839305877686, "learning_rate": 2.699842629679592e-05, "loss": 0.0458, "step": 3329 }, { "epoch": 0.2288345244639912, "grad_norm": 1.469158411026001, "learning_rate": 2.6996422391788532e-05, "loss": 0.0326, "step": 3330 }, { "epoch": 0.2289032435404068, "grad_norm": 1.5140693187713623, "learning_rate": 2.699441789250304e-05, "loss": 0.0434, "step": 3331 }, { "epoch": 0.22897196261682243, "grad_norm": 1.7320085763931274, "learning_rate": 2.6992412799038745e-05, "loss": 0.0516, "step": 3332 }, { "epoch": 0.22904068169323805, "grad_norm": 1.1043814420700073, "learning_rate": 2.6990407111494978e-05, "loss": 0.0527, "step": 3333 }, { "epoch": 0.22910940076965366, "grad_norm": 1.7433693408966064, "learning_rate": 2.6988400829971095e-05, "loss": 0.0481, "step": 3334 }, { "epoch": 0.22917811984606926, "grad_norm": 0.763104259967804, "learning_rate": 2.6986393954566477e-05, "loss": 0.0309, "step": 3335 }, { "epoch": 0.22924683892248487, "grad_norm": 1.4229085445404053, "learning_rate": 2.698438648538055e-05, "loss": 0.0349, "step": 3336 }, { "epoch": 0.2293155579989005, "grad_norm": 1.3736425638198853, "learning_rate": 2.6982378422512754e-05, "loss": 0.044, "step": 3337 }, { "epoch": 0.2293842770753161, "grad_norm": 1.132415771484375, "learning_rate": 2.6980369766062563e-05, "loss": 0.0411, "step": 3338 }, { "epoch": 0.22945299615173173, "grad_norm": 1.2832387685775757, "learning_rate": 2.6978360516129487e-05, "loss": 0.0419, "step": 3339 }, { "epoch": 0.22952171522814732, "grad_norm": 1.9902212619781494, "learning_rate": 2.6976350672813054e-05, "loss": 0.0572, "step": 3340 }, { "epoch": 0.22959043430456294, "grad_norm": 0.7028506398200989, "learning_rate": 2.6974340236212837e-05, "loss": 0.0389, "step": 3341 }, { "epoch": 0.22965915338097856, "grad_norm": 0.8543652892112732, "learning_rate": 2.697232920642842e-05, "loss": 0.0396, "step": 3342 }, { "epoch": 0.22972787245739418, "grad_norm": 1.8482680320739746, "learning_rate": 2.697031758355943e-05, "loss": 0.0533, "step": 3343 }, { "epoch": 0.2297965915338098, "grad_norm": 1.2596805095672607, "learning_rate": 2.6968305367705518e-05, "loss": 0.0495, "step": 3344 }, { "epoch": 0.2298653106102254, "grad_norm": 1.1405317783355713, "learning_rate": 2.6966292558966357e-05, "loss": 0.0454, "step": 3345 }, { "epoch": 0.229934029686641, "grad_norm": 1.1310549974441528, "learning_rate": 2.6964279157441668e-05, "loss": 0.0502, "step": 3346 }, { "epoch": 0.23000274876305662, "grad_norm": 1.7787922620773315, "learning_rate": 2.696226516323119e-05, "loss": 0.0485, "step": 3347 }, { "epoch": 0.23007146783947224, "grad_norm": 1.0825958251953125, "learning_rate": 2.6960250576434682e-05, "loss": 0.0438, "step": 3348 }, { "epoch": 0.23014018691588786, "grad_norm": 0.9598687291145325, "learning_rate": 2.6958235397151958e-05, "loss": 0.0417, "step": 3349 }, { "epoch": 0.23020890599230345, "grad_norm": 0.983622670173645, "learning_rate": 2.6956219625482833e-05, "loss": 0.0437, "step": 3350 }, { "epoch": 0.23027762506871907, "grad_norm": 0.7845152616500854, "learning_rate": 2.695420326152717e-05, "loss": 0.0422, "step": 3351 }, { "epoch": 0.2303463441451347, "grad_norm": 1.0630985498428345, "learning_rate": 2.6952186305384852e-05, "loss": 0.0412, "step": 3352 }, { "epoch": 0.2304150632215503, "grad_norm": 1.674176812171936, "learning_rate": 2.69501687571558e-05, "loss": 0.0461, "step": 3353 }, { "epoch": 0.23048378229796593, "grad_norm": 1.011092185974121, "learning_rate": 2.6948150616939958e-05, "loss": 0.05, "step": 3354 }, { "epoch": 0.23055250137438152, "grad_norm": 1.1294057369232178, "learning_rate": 2.6946131884837292e-05, "loss": 0.0356, "step": 3355 }, { "epoch": 0.23062122045079714, "grad_norm": 0.8776473999023438, "learning_rate": 2.6944112560947818e-05, "loss": 0.0339, "step": 3356 }, { "epoch": 0.23068993952721276, "grad_norm": 1.2965959310531616, "learning_rate": 2.694209264537156e-05, "loss": 0.0291, "step": 3357 }, { "epoch": 0.23075865860362837, "grad_norm": 1.042664647102356, "learning_rate": 2.6940072138208584e-05, "loss": 0.0438, "step": 3358 }, { "epoch": 0.230827377680044, "grad_norm": 1.4255906343460083, "learning_rate": 2.6938051039558988e-05, "loss": 0.0399, "step": 3359 }, { "epoch": 0.23089609675645958, "grad_norm": 0.6679789423942566, "learning_rate": 2.6936029349522884e-05, "loss": 0.0555, "step": 3360 }, { "epoch": 0.2309648158328752, "grad_norm": 0.8404951095581055, "learning_rate": 2.6934007068200422e-05, "loss": 0.0471, "step": 3361 }, { "epoch": 0.23103353490929082, "grad_norm": 1.529584288597107, "learning_rate": 2.693198419569179e-05, "loss": 0.0408, "step": 3362 }, { "epoch": 0.23110225398570644, "grad_norm": 0.9121772646903992, "learning_rate": 2.692996073209719e-05, "loss": 0.0502, "step": 3363 }, { "epoch": 0.23117097306212203, "grad_norm": 1.6593751907348633, "learning_rate": 2.6927936677516862e-05, "loss": 0.0461, "step": 3364 }, { "epoch": 0.23123969213853765, "grad_norm": 1.2124816179275513, "learning_rate": 2.6925912032051075e-05, "loss": 0.0373, "step": 3365 }, { "epoch": 0.23130841121495327, "grad_norm": 1.3111821413040161, "learning_rate": 2.6923886795800125e-05, "loss": 0.0574, "step": 3366 }, { "epoch": 0.2313771302913689, "grad_norm": 1.0426615476608276, "learning_rate": 2.6921860968864337e-05, "loss": 0.0329, "step": 3367 }, { "epoch": 0.2314458493677845, "grad_norm": 1.4851959943771362, "learning_rate": 2.6919834551344065e-05, "loss": 0.0465, "step": 3368 }, { "epoch": 0.2315145684442001, "grad_norm": 1.0753779411315918, "learning_rate": 2.69178075433397e-05, "loss": 0.0372, "step": 3369 }, { "epoch": 0.23158328752061572, "grad_norm": 1.52724289894104, "learning_rate": 2.6915779944951643e-05, "loss": 0.0535, "step": 3370 }, { "epoch": 0.23165200659703133, "grad_norm": 1.2024857997894287, "learning_rate": 2.6913751756280352e-05, "loss": 0.0391, "step": 3371 }, { "epoch": 0.23172072567344695, "grad_norm": 0.792665421962738, "learning_rate": 2.6911722977426296e-05, "loss": 0.0459, "step": 3372 }, { "epoch": 0.23178944474986257, "grad_norm": 1.7132411003112793, "learning_rate": 2.6909693608489965e-05, "loss": 0.0367, "step": 3373 }, { "epoch": 0.23185816382627816, "grad_norm": 1.1163429021835327, "learning_rate": 2.6907663649571904e-05, "loss": 0.0673, "step": 3374 }, { "epoch": 0.23192688290269378, "grad_norm": 1.0039167404174805, "learning_rate": 2.690563310077266e-05, "loss": 0.0482, "step": 3375 }, { "epoch": 0.2319956019791094, "grad_norm": 1.0846344232559204, "learning_rate": 2.6903601962192833e-05, "loss": 0.0515, "step": 3376 }, { "epoch": 0.23206432105552502, "grad_norm": 0.713559627532959, "learning_rate": 2.6901570233933042e-05, "loss": 0.0328, "step": 3377 }, { "epoch": 0.23213304013194064, "grad_norm": 1.4882242679595947, "learning_rate": 2.6899537916093922e-05, "loss": 0.0388, "step": 3378 }, { "epoch": 0.23220175920835623, "grad_norm": 0.8420224189758301, "learning_rate": 2.6897505008776164e-05, "loss": 0.0382, "step": 3379 }, { "epoch": 0.23227047828477185, "grad_norm": 2.2953715324401855, "learning_rate": 2.6895471512080464e-05, "loss": 0.0406, "step": 3380 }, { "epoch": 0.23233919736118747, "grad_norm": 1.0571568012237549, "learning_rate": 2.6893437426107565e-05, "loss": 0.0349, "step": 3381 }, { "epoch": 0.23240791643760308, "grad_norm": 1.1015164852142334, "learning_rate": 2.689140275095823e-05, "loss": 0.0303, "step": 3382 }, { "epoch": 0.2324766355140187, "grad_norm": 0.889146089553833, "learning_rate": 2.6889367486733243e-05, "loss": 0.043, "step": 3383 }, { "epoch": 0.2325453545904343, "grad_norm": 1.1008321046829224, "learning_rate": 2.688733163353344e-05, "loss": 0.0379, "step": 3384 }, { "epoch": 0.2326140736668499, "grad_norm": 0.8463855385780334, "learning_rate": 2.6885295191459663e-05, "loss": 0.0585, "step": 3385 }, { "epoch": 0.23268279274326553, "grad_norm": 1.1916927099227905, "learning_rate": 2.6883258160612797e-05, "loss": 0.0424, "step": 3386 }, { "epoch": 0.23275151181968115, "grad_norm": 0.9550451040267944, "learning_rate": 2.6881220541093758e-05, "loss": 0.0431, "step": 3387 }, { "epoch": 0.23282023089609677, "grad_norm": 0.5882170796394348, "learning_rate": 2.6879182333003477e-05, "loss": 0.0506, "step": 3388 }, { "epoch": 0.23288894997251236, "grad_norm": 2.0819759368896484, "learning_rate": 2.6877143536442928e-05, "loss": 0.0467, "step": 3389 }, { "epoch": 0.23295766904892798, "grad_norm": 1.4628489017486572, "learning_rate": 2.6875104151513104e-05, "loss": 0.0447, "step": 3390 }, { "epoch": 0.2330263881253436, "grad_norm": 1.5446300506591797, "learning_rate": 2.6873064178315035e-05, "loss": 0.0567, "step": 3391 }, { "epoch": 0.23309510720175922, "grad_norm": 1.0673247575759888, "learning_rate": 2.6871023616949776e-05, "loss": 0.0373, "step": 3392 }, { "epoch": 0.23316382627817483, "grad_norm": 0.7421342730522156, "learning_rate": 2.6868982467518413e-05, "loss": 0.034, "step": 3393 }, { "epoch": 0.23323254535459043, "grad_norm": 0.9652242064476013, "learning_rate": 2.686694073012206e-05, "loss": 0.0527, "step": 3394 }, { "epoch": 0.23330126443100604, "grad_norm": 1.0224541425704956, "learning_rate": 2.6864898404861863e-05, "loss": 0.041, "step": 3395 }, { "epoch": 0.23336998350742166, "grad_norm": 1.228825330734253, "learning_rate": 2.6862855491838988e-05, "loss": 0.0366, "step": 3396 }, { "epoch": 0.23343870258383728, "grad_norm": 1.831133484840393, "learning_rate": 2.6860811991154643e-05, "loss": 0.0468, "step": 3397 }, { "epoch": 0.23350742166025287, "grad_norm": 1.3088994026184082, "learning_rate": 2.6858767902910058e-05, "loss": 0.0415, "step": 3398 }, { "epoch": 0.2335761407366685, "grad_norm": 1.1916050910949707, "learning_rate": 2.6856723227206494e-05, "loss": 0.0352, "step": 3399 }, { "epoch": 0.2336448598130841, "grad_norm": 0.9975399374961853, "learning_rate": 2.685467796414523e-05, "loss": 0.046, "step": 3400 }, { "epoch": 0.23371357888949973, "grad_norm": 1.382873773574829, "learning_rate": 2.6852632113827603e-05, "loss": 0.0417, "step": 3401 }, { "epoch": 0.23378229796591535, "grad_norm": 1.2415794134140015, "learning_rate": 2.685058567635494e-05, "loss": 0.0387, "step": 3402 }, { "epoch": 0.23385101704233094, "grad_norm": 0.9668430089950562, "learning_rate": 2.684853865182863e-05, "loss": 0.0334, "step": 3403 }, { "epoch": 0.23391973611874656, "grad_norm": 1.4708129167556763, "learning_rate": 2.684649104035007e-05, "loss": 0.0234, "step": 3404 }, { "epoch": 0.23398845519516218, "grad_norm": 2.1337974071502686, "learning_rate": 2.6844442842020703e-05, "loss": 0.0479, "step": 3405 }, { "epoch": 0.2340571742715778, "grad_norm": 2.1491916179656982, "learning_rate": 2.6842394056941992e-05, "loss": 0.0295, "step": 3406 }, { "epoch": 0.2341258933479934, "grad_norm": 1.8248345851898193, "learning_rate": 2.684034468521542e-05, "loss": 0.0421, "step": 3407 }, { "epoch": 0.234194612424409, "grad_norm": 1.3283227682113647, "learning_rate": 2.683829472694252e-05, "loss": 0.0416, "step": 3408 }, { "epoch": 0.23426333150082462, "grad_norm": 0.9587522745132446, "learning_rate": 2.6836244182224837e-05, "loss": 0.0476, "step": 3409 }, { "epoch": 0.23433205057724024, "grad_norm": 1.0111411809921265, "learning_rate": 2.6834193051163947e-05, "loss": 0.029, "step": 3410 }, { "epoch": 0.23440076965365586, "grad_norm": 1.979854941368103, "learning_rate": 2.6832141333861467e-05, "loss": 0.0612, "step": 3411 }, { "epoch": 0.23446948873007148, "grad_norm": 2.5925230979919434, "learning_rate": 2.6830089030419032e-05, "loss": 0.0477, "step": 3412 }, { "epoch": 0.23453820780648707, "grad_norm": 1.2058298587799072, "learning_rate": 2.6828036140938307e-05, "loss": 0.0544, "step": 3413 }, { "epoch": 0.2346069268829027, "grad_norm": 1.1327462196350098, "learning_rate": 2.6825982665520984e-05, "loss": 0.0499, "step": 3414 }, { "epoch": 0.2346756459593183, "grad_norm": 1.3572797775268555, "learning_rate": 2.6823928604268804e-05, "loss": 0.0378, "step": 3415 }, { "epoch": 0.23474436503573393, "grad_norm": 0.8034027218818665, "learning_rate": 2.6821873957283502e-05, "loss": 0.0418, "step": 3416 }, { "epoch": 0.23481308411214954, "grad_norm": 0.7995656132698059, "learning_rate": 2.681981872466687e-05, "loss": 0.0531, "step": 3417 }, { "epoch": 0.23488180318856514, "grad_norm": 1.345102310180664, "learning_rate": 2.6817762906520717e-05, "loss": 0.0474, "step": 3418 }, { "epoch": 0.23495052226498075, "grad_norm": 1.7244292497634888, "learning_rate": 2.681570650294689e-05, "loss": 0.0552, "step": 3419 }, { "epoch": 0.23501924134139637, "grad_norm": 1.633962631225586, "learning_rate": 2.681364951404725e-05, "loss": 0.0727, "step": 3420 }, { "epoch": 0.235087960417812, "grad_norm": 1.1641005277633667, "learning_rate": 2.681159193992371e-05, "loss": 0.0413, "step": 3421 }, { "epoch": 0.2351566794942276, "grad_norm": 1.0254942178726196, "learning_rate": 2.680953378067818e-05, "loss": 0.0381, "step": 3422 }, { "epoch": 0.2352253985706432, "grad_norm": 1.0427378416061401, "learning_rate": 2.6807475036412627e-05, "loss": 0.0402, "step": 3423 }, { "epoch": 0.23529411764705882, "grad_norm": 0.8590850234031677, "learning_rate": 2.6805415707229038e-05, "loss": 0.0405, "step": 3424 }, { "epoch": 0.23536283672347444, "grad_norm": 0.7229647636413574, "learning_rate": 2.680335579322943e-05, "loss": 0.032, "step": 3425 }, { "epoch": 0.23543155579989006, "grad_norm": 0.7183938026428223, "learning_rate": 2.6801295294515836e-05, "loss": 0.0395, "step": 3426 }, { "epoch": 0.23550027487630565, "grad_norm": 0.6806249022483826, "learning_rate": 2.6799234211190334e-05, "loss": 0.0342, "step": 3427 }, { "epoch": 0.23556899395272127, "grad_norm": 1.6931871175765991, "learning_rate": 2.679717254335504e-05, "loss": 0.0426, "step": 3428 }, { "epoch": 0.23563771302913689, "grad_norm": 1.1083347797393799, "learning_rate": 2.679511029111206e-05, "loss": 0.0377, "step": 3429 }, { "epoch": 0.2357064321055525, "grad_norm": 0.8465394973754883, "learning_rate": 2.6793047454563566e-05, "loss": 0.0309, "step": 3430 }, { "epoch": 0.23577515118196812, "grad_norm": 1.024796485900879, "learning_rate": 2.679098403381175e-05, "loss": 0.0494, "step": 3431 }, { "epoch": 0.23584387025838371, "grad_norm": 1.859917163848877, "learning_rate": 2.6788920028958825e-05, "loss": 0.0443, "step": 3432 }, { "epoch": 0.23591258933479933, "grad_norm": 1.6482890844345093, "learning_rate": 2.6786855440107043e-05, "loss": 0.071, "step": 3433 }, { "epoch": 0.23598130841121495, "grad_norm": 2.1185829639434814, "learning_rate": 2.678479026735867e-05, "loss": 0.0396, "step": 3434 }, { "epoch": 0.23605002748763057, "grad_norm": 5.187160015106201, "learning_rate": 2.678272451081602e-05, "loss": 0.0562, "step": 3435 }, { "epoch": 0.2361187465640462, "grad_norm": 0.9464392066001892, "learning_rate": 2.678065817058142e-05, "loss": 0.0442, "step": 3436 }, { "epoch": 0.23618746564046178, "grad_norm": 1.105486512184143, "learning_rate": 2.6778591246757232e-05, "loss": 0.0361, "step": 3437 }, { "epoch": 0.2362561847168774, "grad_norm": 13.961713790893555, "learning_rate": 2.6776523739445848e-05, "loss": 0.0501, "step": 3438 }, { "epoch": 0.23632490379329302, "grad_norm": 1.644286036491394, "learning_rate": 2.6774455648749693e-05, "loss": 0.0367, "step": 3439 }, { "epoch": 0.23639362286970864, "grad_norm": 3.052438974380493, "learning_rate": 2.6772386974771218e-05, "loss": 0.0521, "step": 3440 }, { "epoch": 0.23646234194612425, "grad_norm": 1.2098075151443481, "learning_rate": 2.6770317717612886e-05, "loss": 0.0379, "step": 3441 }, { "epoch": 0.23653106102253985, "grad_norm": 1.6493767499923706, "learning_rate": 2.6768247877377216e-05, "loss": 0.0466, "step": 3442 }, { "epoch": 0.23659978009895546, "grad_norm": 0.8612551689147949, "learning_rate": 2.6766177454166742e-05, "loss": 0.0448, "step": 3443 }, { "epoch": 0.23666849917537108, "grad_norm": 1.4778406620025635, "learning_rate": 2.6764106448084024e-05, "loss": 0.0508, "step": 3444 }, { "epoch": 0.2367372182517867, "grad_norm": 1.0083292722702026, "learning_rate": 2.6762034859231664e-05, "loss": 0.0448, "step": 3445 }, { "epoch": 0.23680593732820232, "grad_norm": 1.647308349609375, "learning_rate": 2.6759962687712275e-05, "loss": 0.061, "step": 3446 }, { "epoch": 0.2368746564046179, "grad_norm": 2.0334014892578125, "learning_rate": 2.675788993362851e-05, "loss": 0.0671, "step": 3447 }, { "epoch": 0.23694337548103353, "grad_norm": 1.4953832626342773, "learning_rate": 2.6755816597083055e-05, "loss": 0.031, "step": 3448 }, { "epoch": 0.23701209455744915, "grad_norm": 0.9241452813148499, "learning_rate": 2.6753742678178616e-05, "loss": 0.0314, "step": 3449 }, { "epoch": 0.23708081363386477, "grad_norm": 0.6565074920654297, "learning_rate": 2.675166817701792e-05, "loss": 0.0388, "step": 3450 }, { "epoch": 0.2371495327102804, "grad_norm": 1.0892187356948853, "learning_rate": 2.6749593093703755e-05, "loss": 0.0347, "step": 3451 }, { "epoch": 0.23721825178669598, "grad_norm": 1.2880529165267944, "learning_rate": 2.6747517428338894e-05, "loss": 0.0354, "step": 3452 }, { "epoch": 0.2372869708631116, "grad_norm": 0.9543676972389221, "learning_rate": 2.674544118102618e-05, "loss": 0.0391, "step": 3453 }, { "epoch": 0.23735568993952721, "grad_norm": 1.6202291250228882, "learning_rate": 2.6743364351868454e-05, "loss": 0.0466, "step": 3454 }, { "epoch": 0.23742440901594283, "grad_norm": 0.8621169924736023, "learning_rate": 2.67412869409686e-05, "loss": 0.0348, "step": 3455 }, { "epoch": 0.23749312809235845, "grad_norm": 1.655892014503479, "learning_rate": 2.6739208948429528e-05, "loss": 0.0554, "step": 3456 }, { "epoch": 0.23756184716877404, "grad_norm": 0.8406842350959778, "learning_rate": 2.6737130374354187e-05, "loss": 0.0424, "step": 3457 }, { "epoch": 0.23763056624518966, "grad_norm": 0.9371564984321594, "learning_rate": 2.673505121884553e-05, "loss": 0.0413, "step": 3458 }, { "epoch": 0.23769928532160528, "grad_norm": 1.0553193092346191, "learning_rate": 2.6732971482006566e-05, "loss": 0.0249, "step": 3459 }, { "epoch": 0.2377680043980209, "grad_norm": 1.472728967666626, "learning_rate": 2.6730891163940317e-05, "loss": 0.0417, "step": 3460 }, { "epoch": 0.2378367234744365, "grad_norm": 0.8672513961791992, "learning_rate": 2.6728810264749832e-05, "loss": 0.0344, "step": 3461 }, { "epoch": 0.2379054425508521, "grad_norm": 1.1411393880844116, "learning_rate": 2.6726728784538205e-05, "loss": 0.0533, "step": 3462 }, { "epoch": 0.23797416162726773, "grad_norm": 0.7381301522254944, "learning_rate": 2.672464672340854e-05, "loss": 0.0276, "step": 3463 }, { "epoch": 0.23804288070368335, "grad_norm": 1.0828912258148193, "learning_rate": 2.672256408146399e-05, "loss": 0.0371, "step": 3464 }, { "epoch": 0.23811159978009896, "grad_norm": 1.171057939529419, "learning_rate": 2.6720480858807706e-05, "loss": 0.0448, "step": 3465 }, { "epoch": 0.23818031885651456, "grad_norm": 1.4548437595367432, "learning_rate": 2.67183970555429e-05, "loss": 0.0445, "step": 3466 }, { "epoch": 0.23824903793293017, "grad_norm": 2.612487316131592, "learning_rate": 2.6716312671772794e-05, "loss": 0.059, "step": 3467 }, { "epoch": 0.2383177570093458, "grad_norm": 1.1009297370910645, "learning_rate": 2.671422770760065e-05, "loss": 0.0395, "step": 3468 }, { "epoch": 0.2383864760857614, "grad_norm": 1.0248979330062866, "learning_rate": 2.6712142163129746e-05, "loss": 0.0382, "step": 3469 }, { "epoch": 0.23845519516217703, "grad_norm": 1.4514820575714111, "learning_rate": 2.67100560384634e-05, "loss": 0.052, "step": 3470 }, { "epoch": 0.23852391423859262, "grad_norm": 1.232442855834961, "learning_rate": 2.670796933370495e-05, "loss": 0.0353, "step": 3471 }, { "epoch": 0.23859263331500824, "grad_norm": 0.6079763770103455, "learning_rate": 2.670588204895777e-05, "loss": 0.0326, "step": 3472 }, { "epoch": 0.23866135239142386, "grad_norm": 0.6953364014625549, "learning_rate": 2.6703794184325266e-05, "loss": 0.0428, "step": 3473 }, { "epoch": 0.23873007146783948, "grad_norm": 1.4446016550064087, "learning_rate": 2.670170573991085e-05, "loss": 0.0426, "step": 3474 }, { "epoch": 0.2387987905442551, "grad_norm": 1.2258363962173462, "learning_rate": 2.6699616715817997e-05, "loss": 0.0395, "step": 3475 }, { "epoch": 0.2388675096206707, "grad_norm": 0.8373967409133911, "learning_rate": 2.669752711215018e-05, "loss": 0.0369, "step": 3476 }, { "epoch": 0.2389362286970863, "grad_norm": 0.856813371181488, "learning_rate": 2.6695436929010924e-05, "loss": 0.0339, "step": 3477 }, { "epoch": 0.23900494777350192, "grad_norm": 0.9379360675811768, "learning_rate": 2.6693346166503767e-05, "loss": 0.064, "step": 3478 }, { "epoch": 0.23907366684991754, "grad_norm": 1.7352955341339111, "learning_rate": 2.6691254824732278e-05, "loss": 0.0488, "step": 3479 }, { "epoch": 0.23914238592633316, "grad_norm": 0.6981686949729919, "learning_rate": 2.6689162903800063e-05, "loss": 0.0395, "step": 3480 }, { "epoch": 0.23921110500274875, "grad_norm": 0.6514850854873657, "learning_rate": 2.668707040381075e-05, "loss": 0.0331, "step": 3481 }, { "epoch": 0.23927982407916437, "grad_norm": 1.4007883071899414, "learning_rate": 2.6684977324867994e-05, "loss": 0.044, "step": 3482 }, { "epoch": 0.23934854315558, "grad_norm": 1.0245604515075684, "learning_rate": 2.6682883667075487e-05, "loss": 0.031, "step": 3483 }, { "epoch": 0.2394172622319956, "grad_norm": 1.0572984218597412, "learning_rate": 2.6680789430536944e-05, "loss": 0.0287, "step": 3484 }, { "epoch": 0.23948598130841123, "grad_norm": 1.0382745265960693, "learning_rate": 2.6678694615356103e-05, "loss": 0.0594, "step": 3485 }, { "epoch": 0.23955470038482682, "grad_norm": 1.348097562789917, "learning_rate": 2.6676599221636744e-05, "loss": 0.0315, "step": 3486 }, { "epoch": 0.23962341946124244, "grad_norm": 1.6734076738357544, "learning_rate": 2.6674503249482667e-05, "loss": 0.0464, "step": 3487 }, { "epoch": 0.23969213853765806, "grad_norm": 1.7583345174789429, "learning_rate": 2.6672406698997696e-05, "loss": 0.0412, "step": 3488 }, { "epoch": 0.23976085761407367, "grad_norm": 1.2997361421585083, "learning_rate": 2.6670309570285697e-05, "loss": 0.0467, "step": 3489 }, { "epoch": 0.2398295766904893, "grad_norm": 1.2513134479522705, "learning_rate": 2.6668211863450557e-05, "loss": 0.0395, "step": 3490 }, { "epoch": 0.23989829576690488, "grad_norm": 0.8543403744697571, "learning_rate": 2.6666113578596186e-05, "loss": 0.0391, "step": 3491 }, { "epoch": 0.2399670148433205, "grad_norm": 1.3727294206619263, "learning_rate": 2.666401471582654e-05, "loss": 0.0477, "step": 3492 }, { "epoch": 0.24003573391973612, "grad_norm": 1.4385799169540405, "learning_rate": 2.6661915275245576e-05, "loss": 0.0367, "step": 3493 }, { "epoch": 0.24010445299615174, "grad_norm": 2.0682718753814697, "learning_rate": 2.665981525695731e-05, "loss": 0.0667, "step": 3494 }, { "epoch": 0.24017317207256733, "grad_norm": 1.0285762548446655, "learning_rate": 2.6657714661065768e-05, "loss": 0.0532, "step": 3495 }, { "epoch": 0.24024189114898295, "grad_norm": 1.0623350143432617, "learning_rate": 2.665561348767501e-05, "loss": 0.054, "step": 3496 }, { "epoch": 0.24031061022539857, "grad_norm": 1.29618501663208, "learning_rate": 2.665351173688912e-05, "loss": 0.0407, "step": 3497 }, { "epoch": 0.2403793293018142, "grad_norm": 1.1610403060913086, "learning_rate": 2.6651409408812214e-05, "loss": 0.0332, "step": 3498 }, { "epoch": 0.2404480483782298, "grad_norm": 1.184424877166748, "learning_rate": 2.6649306503548445e-05, "loss": 0.0423, "step": 3499 }, { "epoch": 0.2405167674546454, "grad_norm": 1.0422778129577637, "learning_rate": 2.6647203021201983e-05, "loss": 0.0413, "step": 3500 }, { "epoch": 0.24058548653106102, "grad_norm": 0.7421925663948059, "learning_rate": 2.6645098961877026e-05, "loss": 0.0472, "step": 3501 }, { "epoch": 0.24065420560747663, "grad_norm": 1.3623101711273193, "learning_rate": 2.664299432567781e-05, "loss": 0.0457, "step": 3502 }, { "epoch": 0.24072292468389225, "grad_norm": 1.6322799921035767, "learning_rate": 2.6640889112708594e-05, "loss": 0.0555, "step": 3503 }, { "epoch": 0.24079164376030787, "grad_norm": 1.2968789339065552, "learning_rate": 2.6638783323073657e-05, "loss": 0.0356, "step": 3504 }, { "epoch": 0.24086036283672346, "grad_norm": 0.9763268232345581, "learning_rate": 2.6636676956877325e-05, "loss": 0.0463, "step": 3505 }, { "epoch": 0.24092908191313908, "grad_norm": 0.7684444189071655, "learning_rate": 2.6634570014223945e-05, "loss": 0.0356, "step": 3506 }, { "epoch": 0.2409978009895547, "grad_norm": 0.9432527422904968, "learning_rate": 2.6632462495217883e-05, "loss": 0.0408, "step": 3507 }, { "epoch": 0.24106652006597032, "grad_norm": 1.2716649770736694, "learning_rate": 2.6630354399963543e-05, "loss": 0.0449, "step": 3508 }, { "epoch": 0.24113523914238594, "grad_norm": 0.8808484673500061, "learning_rate": 2.6628245728565357e-05, "loss": 0.0405, "step": 3509 }, { "epoch": 0.24120395821880153, "grad_norm": 1.3317128419876099, "learning_rate": 2.662613648112779e-05, "loss": 0.0367, "step": 3510 }, { "epoch": 0.24127267729521715, "grad_norm": 0.9637055993080139, "learning_rate": 2.6624026657755316e-05, "loss": 0.0298, "step": 3511 }, { "epoch": 0.24134139637163277, "grad_norm": 0.6235661506652832, "learning_rate": 2.6621916258552464e-05, "loss": 0.0365, "step": 3512 }, { "epoch": 0.24141011544804838, "grad_norm": 1.8167471885681152, "learning_rate": 2.6619805283623774e-05, "loss": 0.0583, "step": 3513 }, { "epoch": 0.241478834524464, "grad_norm": 1.5216771364212036, "learning_rate": 2.661769373307382e-05, "loss": 0.0341, "step": 3514 }, { "epoch": 0.2415475536008796, "grad_norm": 1.2541297674179077, "learning_rate": 2.6615581607007203e-05, "loss": 0.024, "step": 3515 }, { "epoch": 0.2416162726772952, "grad_norm": 1.070551872253418, "learning_rate": 2.661346890552855e-05, "loss": 0.0361, "step": 3516 }, { "epoch": 0.24168499175371083, "grad_norm": 0.9652674794197083, "learning_rate": 2.661135562874253e-05, "loss": 0.0376, "step": 3517 }, { "epoch": 0.24175371083012645, "grad_norm": 1.814650297164917, "learning_rate": 2.660924177675382e-05, "loss": 0.0407, "step": 3518 }, { "epoch": 0.24182242990654207, "grad_norm": 1.1587191820144653, "learning_rate": 2.660712734966714e-05, "loss": 0.0536, "step": 3519 }, { "epoch": 0.24189114898295766, "grad_norm": 2.220903158187866, "learning_rate": 2.6605012347587234e-05, "loss": 0.0599, "step": 3520 }, { "epoch": 0.24195986805937328, "grad_norm": 1.415053129196167, "learning_rate": 2.6602896770618878e-05, "loss": 0.0433, "step": 3521 }, { "epoch": 0.2420285871357889, "grad_norm": 1.5671391487121582, "learning_rate": 2.660078061886687e-05, "loss": 0.0432, "step": 3522 }, { "epoch": 0.24209730621220452, "grad_norm": 0.8302785754203796, "learning_rate": 2.659866389243604e-05, "loss": 0.0298, "step": 3523 }, { "epoch": 0.24216602528862013, "grad_norm": 1.0281708240509033, "learning_rate": 2.6596546591431243e-05, "loss": 0.0397, "step": 3524 }, { "epoch": 0.24223474436503573, "grad_norm": 0.8743229508399963, "learning_rate": 2.659442871595737e-05, "loss": 0.0495, "step": 3525 }, { "epoch": 0.24230346344145134, "grad_norm": 1.0938001871109009, "learning_rate": 2.659231026611934e-05, "loss": 0.0442, "step": 3526 }, { "epoch": 0.24237218251786696, "grad_norm": 1.1111834049224854, "learning_rate": 2.6590191242022087e-05, "loss": 0.0539, "step": 3527 }, { "epoch": 0.24244090159428258, "grad_norm": 0.58870929479599, "learning_rate": 2.6588071643770594e-05, "loss": 0.0299, "step": 3528 }, { "epoch": 0.24250962067069817, "grad_norm": 2.468778610229492, "learning_rate": 2.6585951471469857e-05, "loss": 0.0582, "step": 3529 }, { "epoch": 0.2425783397471138, "grad_norm": 0.9963350892066956, "learning_rate": 2.6583830725224896e-05, "loss": 0.0394, "step": 3530 }, { "epoch": 0.2426470588235294, "grad_norm": 0.9710202813148499, "learning_rate": 2.6581709405140783e-05, "loss": 0.0356, "step": 3531 }, { "epoch": 0.24271577789994503, "grad_norm": 1.4320865869522095, "learning_rate": 2.657958751132259e-05, "loss": 0.0494, "step": 3532 }, { "epoch": 0.24278449697636065, "grad_norm": 1.5972603559494019, "learning_rate": 2.6577465043875445e-05, "loss": 0.041, "step": 3533 }, { "epoch": 0.24285321605277624, "grad_norm": 1.0529680252075195, "learning_rate": 2.657534200290448e-05, "loss": 0.0345, "step": 3534 }, { "epoch": 0.24292193512919186, "grad_norm": 1.1788201332092285, "learning_rate": 2.657321838851487e-05, "loss": 0.0378, "step": 3535 }, { "epoch": 0.24299065420560748, "grad_norm": 1.234550952911377, "learning_rate": 2.6571094200811817e-05, "loss": 0.0333, "step": 3536 }, { "epoch": 0.2430593732820231, "grad_norm": 1.3995232582092285, "learning_rate": 2.656896943990055e-05, "loss": 0.053, "step": 3537 }, { "epoch": 0.2431280923584387, "grad_norm": 0.8425830602645874, "learning_rate": 2.6566844105886313e-05, "loss": 0.0446, "step": 3538 }, { "epoch": 0.2431968114348543, "grad_norm": 1.0113554000854492, "learning_rate": 2.6564718198874404e-05, "loss": 0.0498, "step": 3539 }, { "epoch": 0.24326553051126992, "grad_norm": 1.15791916847229, "learning_rate": 2.656259171897013e-05, "loss": 0.042, "step": 3540 }, { "epoch": 0.24333424958768554, "grad_norm": 1.2563743591308594, "learning_rate": 2.6560464666278836e-05, "loss": 0.0452, "step": 3541 }, { "epoch": 0.24340296866410116, "grad_norm": 0.44328874349594116, "learning_rate": 2.655833704090589e-05, "loss": 0.0338, "step": 3542 }, { "epoch": 0.24347168774051678, "grad_norm": 1.4594476222991943, "learning_rate": 2.655620884295669e-05, "loss": 0.0481, "step": 3543 }, { "epoch": 0.24354040681693237, "grad_norm": 0.8151417970657349, "learning_rate": 2.655408007253666e-05, "loss": 0.0422, "step": 3544 }, { "epoch": 0.243609125893348, "grad_norm": 0.6919184327125549, "learning_rate": 2.655195072975126e-05, "loss": 0.027, "step": 3545 }, { "epoch": 0.2436778449697636, "grad_norm": 1.912481665611267, "learning_rate": 2.6549820814705964e-05, "loss": 0.0349, "step": 3546 }, { "epoch": 0.24374656404617923, "grad_norm": 1.0716092586517334, "learning_rate": 2.65476903275063e-05, "loss": 0.0315, "step": 3547 }, { "epoch": 0.24381528312259484, "grad_norm": 0.7391263842582703, "learning_rate": 2.6545559268257794e-05, "loss": 0.0315, "step": 3548 }, { "epoch": 0.24388400219901044, "grad_norm": 1.2362911701202393, "learning_rate": 2.654342763706602e-05, "loss": 0.0513, "step": 3549 }, { "epoch": 0.24395272127542605, "grad_norm": 0.6307518482208252, "learning_rate": 2.654129543403657e-05, "loss": 0.0324, "step": 3550 }, { "epoch": 0.24402144035184167, "grad_norm": 1.1600581407546997, "learning_rate": 2.653916265927507e-05, "loss": 0.0349, "step": 3551 }, { "epoch": 0.2440901594282573, "grad_norm": 1.2849853038787842, "learning_rate": 2.653702931288718e-05, "loss": 0.0359, "step": 3552 }, { "epoch": 0.2441588785046729, "grad_norm": 2.6241776943206787, "learning_rate": 2.6534895394978577e-05, "loss": 0.0595, "step": 3553 }, { "epoch": 0.2442275975810885, "grad_norm": 1.0987036228179932, "learning_rate": 2.6532760905654968e-05, "loss": 0.0398, "step": 3554 }, { "epoch": 0.24429631665750412, "grad_norm": 1.5810080766677856, "learning_rate": 2.6530625845022096e-05, "loss": 0.0556, "step": 3555 }, { "epoch": 0.24436503573391974, "grad_norm": 1.107401967048645, "learning_rate": 2.652849021318573e-05, "loss": 0.0383, "step": 3556 }, { "epoch": 0.24443375481033536, "grad_norm": 1.7518588304519653, "learning_rate": 2.6526354010251655e-05, "loss": 0.0587, "step": 3557 }, { "epoch": 0.24450247388675095, "grad_norm": 1.5177589654922485, "learning_rate": 2.6524217236325698e-05, "loss": 0.0345, "step": 3558 }, { "epoch": 0.24457119296316657, "grad_norm": 1.053487777709961, "learning_rate": 2.6522079891513714e-05, "loss": 0.032, "step": 3559 }, { "epoch": 0.24463991203958219, "grad_norm": 0.792023777961731, "learning_rate": 2.6519941975921582e-05, "loss": 0.0502, "step": 3560 }, { "epoch": 0.2447086311159978, "grad_norm": 0.7288333773612976, "learning_rate": 2.651780348965521e-05, "loss": 0.0301, "step": 3561 }, { "epoch": 0.24477735019241342, "grad_norm": 1.257631778717041, "learning_rate": 2.6515664432820527e-05, "loss": 0.0576, "step": 3562 }, { "epoch": 0.24484606926882901, "grad_norm": 1.148167371749878, "learning_rate": 2.6513524805523503e-05, "loss": 0.04, "step": 3563 }, { "epoch": 0.24491478834524463, "grad_norm": 0.549366295337677, "learning_rate": 2.6511384607870136e-05, "loss": 0.0389, "step": 3564 }, { "epoch": 0.24498350742166025, "grad_norm": 1.1003096103668213, "learning_rate": 2.650924383996644e-05, "loss": 0.043, "step": 3565 }, { "epoch": 0.24505222649807587, "grad_norm": 1.4113658666610718, "learning_rate": 2.6507102501918464e-05, "loss": 0.0504, "step": 3566 }, { "epoch": 0.2451209455744915, "grad_norm": 1.278032660484314, "learning_rate": 2.6504960593832294e-05, "loss": 0.0399, "step": 3567 }, { "epoch": 0.24518966465090708, "grad_norm": 1.7377204895019531, "learning_rate": 2.6502818115814028e-05, "loss": 0.0447, "step": 3568 }, { "epoch": 0.2452583837273227, "grad_norm": 1.1661198139190674, "learning_rate": 2.6500675067969793e-05, "loss": 0.0333, "step": 3569 }, { "epoch": 0.24532710280373832, "grad_norm": 1.1693980693817139, "learning_rate": 2.649853145040577e-05, "loss": 0.0423, "step": 3570 }, { "epoch": 0.24539582188015394, "grad_norm": 1.2374589443206787, "learning_rate": 2.6496387263228133e-05, "loss": 0.039, "step": 3571 }, { "epoch": 0.24546454095656955, "grad_norm": 1.3000059127807617, "learning_rate": 2.6494242506543107e-05, "loss": 0.0443, "step": 3572 }, { "epoch": 0.24553326003298515, "grad_norm": 0.9158942699432373, "learning_rate": 2.6492097180456936e-05, "loss": 0.0416, "step": 3573 }, { "epoch": 0.24560197910940076, "grad_norm": 0.9798029065132141, "learning_rate": 2.64899512850759e-05, "loss": 0.0566, "step": 3574 }, { "epoch": 0.24567069818581638, "grad_norm": 1.123939037322998, "learning_rate": 2.64878048205063e-05, "loss": 0.0405, "step": 3575 }, { "epoch": 0.245739417262232, "grad_norm": 0.910530149936676, "learning_rate": 2.648565778685447e-05, "loss": 0.0374, "step": 3576 }, { "epoch": 0.24580813633864762, "grad_norm": 1.2476131916046143, "learning_rate": 2.6483510184226765e-05, "loss": 0.048, "step": 3577 }, { "epoch": 0.2458768554150632, "grad_norm": 1.2681485414505005, "learning_rate": 2.6481362012729573e-05, "loss": 0.0474, "step": 3578 }, { "epoch": 0.24594557449147883, "grad_norm": 0.8290725946426392, "learning_rate": 2.6479213272469314e-05, "loss": 0.0433, "step": 3579 }, { "epoch": 0.24601429356789445, "grad_norm": 2.0898563861846924, "learning_rate": 2.6477063963552422e-05, "loss": 0.0414, "step": 3580 }, { "epoch": 0.24608301264431007, "grad_norm": 1.3019475936889648, "learning_rate": 2.6474914086085386e-05, "loss": 0.0667, "step": 3581 }, { "epoch": 0.24615173172072569, "grad_norm": 1.679219126701355, "learning_rate": 2.6472763640174695e-05, "loss": 0.0427, "step": 3582 }, { "epoch": 0.24622045079714128, "grad_norm": 1.2037090063095093, "learning_rate": 2.647061262592688e-05, "loss": 0.0633, "step": 3583 }, { "epoch": 0.2462891698735569, "grad_norm": 1.527783989906311, "learning_rate": 2.64684610434485e-05, "loss": 0.0445, "step": 3584 }, { "epoch": 0.24635788894997251, "grad_norm": 1.1000620126724243, "learning_rate": 2.6466308892846133e-05, "loss": 0.0502, "step": 3585 }, { "epoch": 0.24642660802638813, "grad_norm": 1.9235948324203491, "learning_rate": 2.6464156174226403e-05, "loss": 0.0336, "step": 3586 }, { "epoch": 0.24649532710280375, "grad_norm": 1.3979953527450562, "learning_rate": 2.646200288769594e-05, "loss": 0.034, "step": 3587 }, { "epoch": 0.24656404617921934, "grad_norm": 1.2103477716445923, "learning_rate": 2.645984903336142e-05, "loss": 0.0515, "step": 3588 }, { "epoch": 0.24663276525563496, "grad_norm": 1.6477079391479492, "learning_rate": 2.6457694611329542e-05, "loss": 0.0294, "step": 3589 }, { "epoch": 0.24670148433205058, "grad_norm": 1.396749496459961, "learning_rate": 2.645553962170703e-05, "loss": 0.0394, "step": 3590 }, { "epoch": 0.2467702034084662, "grad_norm": 1.1546066999435425, "learning_rate": 2.6453384064600634e-05, "loss": 0.0451, "step": 3591 }, { "epoch": 0.2468389224848818, "grad_norm": 1.6377793550491333, "learning_rate": 2.645122794011714e-05, "loss": 0.0406, "step": 3592 }, { "epoch": 0.2469076415612974, "grad_norm": 1.1886919736862183, "learning_rate": 2.6449071248363353e-05, "loss": 0.0472, "step": 3593 }, { "epoch": 0.24697636063771303, "grad_norm": 1.3046834468841553, "learning_rate": 2.644691398944612e-05, "loss": 0.0458, "step": 3594 }, { "epoch": 0.24704507971412865, "grad_norm": 0.9015142917633057, "learning_rate": 2.6444756163472298e-05, "loss": 0.0362, "step": 3595 }, { "epoch": 0.24711379879054426, "grad_norm": 1.0546848773956299, "learning_rate": 2.6442597770548788e-05, "loss": 0.0408, "step": 3596 }, { "epoch": 0.24718251786695986, "grad_norm": 1.4659045934677124, "learning_rate": 2.6440438810782504e-05, "loss": 0.0345, "step": 3597 }, { "epoch": 0.24725123694337547, "grad_norm": 0.8042749166488647, "learning_rate": 2.6438279284280408e-05, "loss": 0.0375, "step": 3598 }, { "epoch": 0.2473199560197911, "grad_norm": 1.0347256660461426, "learning_rate": 2.643611919114947e-05, "loss": 0.036, "step": 3599 }, { "epoch": 0.2473886750962067, "grad_norm": 1.2497986555099487, "learning_rate": 2.6433958531496694e-05, "loss": 0.0448, "step": 3600 }, { "epoch": 0.24745739417262233, "grad_norm": 1.9507478475570679, "learning_rate": 2.6431797305429125e-05, "loss": 0.054, "step": 3601 }, { "epoch": 0.24752611324903792, "grad_norm": 1.602587342262268, "learning_rate": 2.6429635513053824e-05, "loss": 0.052, "step": 3602 }, { "epoch": 0.24759483232545354, "grad_norm": 1.1974079608917236, "learning_rate": 2.642747315447787e-05, "loss": 0.0494, "step": 3603 }, { "epoch": 0.24766355140186916, "grad_norm": 1.5372898578643799, "learning_rate": 2.6425310229808397e-05, "loss": 0.0541, "step": 3604 }, { "epoch": 0.24773227047828478, "grad_norm": 1.0734481811523438, "learning_rate": 2.6423146739152538e-05, "loss": 0.0419, "step": 3605 }, { "epoch": 0.2478009895547004, "grad_norm": 0.7659348845481873, "learning_rate": 2.642098268261748e-05, "loss": 0.0383, "step": 3606 }, { "epoch": 0.247869708631116, "grad_norm": 0.9603962302207947, "learning_rate": 2.6418818060310417e-05, "loss": 0.0324, "step": 3607 }, { "epoch": 0.2479384277075316, "grad_norm": 0.7005270719528198, "learning_rate": 2.6416652872338584e-05, "loss": 0.0323, "step": 3608 }, { "epoch": 0.24800714678394722, "grad_norm": 1.4513927698135376, "learning_rate": 2.641448711880924e-05, "loss": 0.0422, "step": 3609 }, { "epoch": 0.24807586586036284, "grad_norm": 1.5816354751586914, "learning_rate": 2.641232079982967e-05, "loss": 0.0382, "step": 3610 }, { "epoch": 0.24814458493677846, "grad_norm": 0.6566861867904663, "learning_rate": 2.6410153915507194e-05, "loss": 0.0418, "step": 3611 }, { "epoch": 0.24821330401319405, "grad_norm": 1.0369348526000977, "learning_rate": 2.640798646594915e-05, "loss": 0.0274, "step": 3612 }, { "epoch": 0.24828202308960967, "grad_norm": 1.012439489364624, "learning_rate": 2.6405818451262908e-05, "loss": 0.038, "step": 3613 }, { "epoch": 0.2483507421660253, "grad_norm": 1.0690444707870483, "learning_rate": 2.640364987155587e-05, "loss": 0.0446, "step": 3614 }, { "epoch": 0.2484194612424409, "grad_norm": 1.3705788850784302, "learning_rate": 2.6401480726935462e-05, "loss": 0.0358, "step": 3615 }, { "epoch": 0.24848818031885653, "grad_norm": 1.0386368036270142, "learning_rate": 2.6399311017509137e-05, "loss": 0.0423, "step": 3616 }, { "epoch": 0.24855689939527212, "grad_norm": 1.071603536605835, "learning_rate": 2.6397140743384386e-05, "loss": 0.0287, "step": 3617 }, { "epoch": 0.24862561847168774, "grad_norm": 0.8353845477104187, "learning_rate": 2.639496990466871e-05, "loss": 0.0368, "step": 3618 }, { "epoch": 0.24869433754810336, "grad_norm": 1.2113312482833862, "learning_rate": 2.6392798501469655e-05, "loss": 0.0363, "step": 3619 }, { "epoch": 0.24876305662451897, "grad_norm": 1.0690999031066895, "learning_rate": 2.6390626533894783e-05, "loss": 0.0328, "step": 3620 }, { "epoch": 0.2488317757009346, "grad_norm": 1.2498360872268677, "learning_rate": 2.638845400205169e-05, "loss": 0.0349, "step": 3621 }, { "epoch": 0.24890049477735018, "grad_norm": 1.4364101886749268, "learning_rate": 2.6386280906048003e-05, "loss": 0.0358, "step": 3622 }, { "epoch": 0.2489692138537658, "grad_norm": 0.9191465377807617, "learning_rate": 2.6384107245991364e-05, "loss": 0.03, "step": 3623 }, { "epoch": 0.24903793293018142, "grad_norm": 0.9819169640541077, "learning_rate": 2.6381933021989462e-05, "loss": 0.0445, "step": 3624 }, { "epoch": 0.24910665200659704, "grad_norm": 0.81129390001297, "learning_rate": 2.6379758234149993e-05, "loss": 0.0411, "step": 3625 }, { "epoch": 0.24917537108301263, "grad_norm": 0.96062833070755, "learning_rate": 2.6377582882580694e-05, "loss": 0.0409, "step": 3626 }, { "epoch": 0.24924409015942825, "grad_norm": 0.7807425856590271, "learning_rate": 2.637540696738934e-05, "loss": 0.0291, "step": 3627 }, { "epoch": 0.24931280923584387, "grad_norm": 1.7052984237670898, "learning_rate": 2.6373230488683704e-05, "loss": 0.0466, "step": 3628 }, { "epoch": 0.2493815283122595, "grad_norm": 1.5360221862792969, "learning_rate": 2.6371053446571617e-05, "loss": 0.0327, "step": 3629 }, { "epoch": 0.2494502473886751, "grad_norm": 0.9463713765144348, "learning_rate": 2.6368875841160918e-05, "loss": 0.0442, "step": 3630 }, { "epoch": 0.2495189664650907, "grad_norm": 1.4151448011398315, "learning_rate": 2.6366697672559486e-05, "loss": 0.0433, "step": 3631 }, { "epoch": 0.24958768554150632, "grad_norm": 1.1383001804351807, "learning_rate": 2.636451894087522e-05, "loss": 0.036, "step": 3632 }, { "epoch": 0.24965640461792193, "grad_norm": 1.611820101737976, "learning_rate": 2.636233964621605e-05, "loss": 0.0534, "step": 3633 }, { "epoch": 0.24972512369433755, "grad_norm": 1.0475441217422485, "learning_rate": 2.6360159788689928e-05, "loss": 0.046, "step": 3634 }, { "epoch": 0.24979384277075317, "grad_norm": 0.9222829937934875, "learning_rate": 2.635797936840485e-05, "loss": 0.045, "step": 3635 }, { "epoch": 0.24986256184716876, "grad_norm": 0.7071849703788757, "learning_rate": 2.635579838546882e-05, "loss": 0.0551, "step": 3636 }, { "epoch": 0.24993128092358438, "grad_norm": 0.9199675917625427, "learning_rate": 2.635361683998989e-05, "loss": 0.0459, "step": 3637 }, { "epoch": 0.25, "grad_norm": 1.8916183710098267, "learning_rate": 2.6351434732076125e-05, "loss": 0.0364, "step": 3638 }, { "epoch": 0.2500687190764156, "grad_norm": 1.6800898313522339, "learning_rate": 2.6349252061835618e-05, "loss": 0.039, "step": 3639 }, { "epoch": 0.25013743815283124, "grad_norm": 1.326410174369812, "learning_rate": 2.6347068829376498e-05, "loss": 0.0452, "step": 3640 }, { "epoch": 0.25020615722924683, "grad_norm": 1.134949803352356, "learning_rate": 2.634488503480692e-05, "loss": 0.0417, "step": 3641 }, { "epoch": 0.2502748763056625, "grad_norm": 0.6389055848121643, "learning_rate": 2.6342700678235057e-05, "loss": 0.0339, "step": 3642 }, { "epoch": 0.25034359538207807, "grad_norm": 0.6054244637489319, "learning_rate": 2.634051575976912e-05, "loss": 0.0365, "step": 3643 }, { "epoch": 0.25041231445849366, "grad_norm": 0.9731985926628113, "learning_rate": 2.6338330279517353e-05, "loss": 0.0337, "step": 3644 }, { "epoch": 0.2504810335349093, "grad_norm": 1.2111555337905884, "learning_rate": 2.6336144237588012e-05, "loss": 0.0333, "step": 3645 }, { "epoch": 0.2505497526113249, "grad_norm": 1.3864878416061401, "learning_rate": 2.6333957634089387e-05, "loss": 0.0456, "step": 3646 }, { "epoch": 0.25061847168774054, "grad_norm": 2.022447109222412, "learning_rate": 2.6331770469129806e-05, "loss": 0.0423, "step": 3647 }, { "epoch": 0.25068719076415613, "grad_norm": 1.3110753297805786, "learning_rate": 2.6329582742817615e-05, "loss": 0.0448, "step": 3648 }, { "epoch": 0.2507559098405717, "grad_norm": 0.959936261177063, "learning_rate": 2.6327394455261186e-05, "loss": 0.0323, "step": 3649 }, { "epoch": 0.25082462891698737, "grad_norm": 1.4742975234985352, "learning_rate": 2.6325205606568927e-05, "loss": 0.0438, "step": 3650 }, { "epoch": 0.25089334799340296, "grad_norm": 0.756511926651001, "learning_rate": 2.632301619684926e-05, "loss": 0.0436, "step": 3651 }, { "epoch": 0.2509620670698186, "grad_norm": 1.0772312879562378, "learning_rate": 2.6320826226210653e-05, "loss": 0.0341, "step": 3652 }, { "epoch": 0.2510307861462342, "grad_norm": 0.5128979682922363, "learning_rate": 2.6318635694761592e-05, "loss": 0.0294, "step": 3653 }, { "epoch": 0.2510995052226498, "grad_norm": 0.6982438564300537, "learning_rate": 2.6316444602610587e-05, "loss": 0.0347, "step": 3654 }, { "epoch": 0.25116822429906543, "grad_norm": 1.5350651741027832, "learning_rate": 2.631425294986618e-05, "loss": 0.0425, "step": 3655 }, { "epoch": 0.251236943375481, "grad_norm": 1.3940132856369019, "learning_rate": 2.6312060736636947e-05, "loss": 0.0424, "step": 3656 }, { "epoch": 0.25130566245189667, "grad_norm": 1.1222182512283325, "learning_rate": 2.6309867963031477e-05, "loss": 0.0347, "step": 3657 }, { "epoch": 0.25137438152831226, "grad_norm": 1.465171217918396, "learning_rate": 2.6307674629158402e-05, "loss": 0.049, "step": 3658 }, { "epoch": 0.25144310060472785, "grad_norm": 0.925005316734314, "learning_rate": 2.6305480735126376e-05, "loss": 0.0405, "step": 3659 }, { "epoch": 0.2515118196811435, "grad_norm": 0.5889115929603577, "learning_rate": 2.6303286281044073e-05, "loss": 0.0295, "step": 3660 }, { "epoch": 0.2515805387575591, "grad_norm": 1.2423715591430664, "learning_rate": 2.630109126702021e-05, "loss": 0.0395, "step": 3661 }, { "epoch": 0.25164925783397474, "grad_norm": 0.9544276595115662, "learning_rate": 2.6298895693163513e-05, "loss": 0.039, "step": 3662 }, { "epoch": 0.25171797691039033, "grad_norm": 1.733527421951294, "learning_rate": 2.629669955958276e-05, "loss": 0.0481, "step": 3663 }, { "epoch": 0.2517866959868059, "grad_norm": 0.9996674060821533, "learning_rate": 2.6294502866386734e-05, "loss": 0.024, "step": 3664 }, { "epoch": 0.25185541506322157, "grad_norm": 2.21815824508667, "learning_rate": 2.6292305613684257e-05, "loss": 0.0365, "step": 3665 }, { "epoch": 0.25192413413963716, "grad_norm": 0.8204174637794495, "learning_rate": 2.6290107801584175e-05, "loss": 0.0451, "step": 3666 }, { "epoch": 0.2519928532160528, "grad_norm": 1.4897985458374023, "learning_rate": 2.628790943019536e-05, "loss": 0.0424, "step": 3667 }, { "epoch": 0.2520615722924684, "grad_norm": 0.5332175493240356, "learning_rate": 2.6285710499626726e-05, "loss": 0.0283, "step": 3668 }, { "epoch": 0.252130291368884, "grad_norm": 0.7390540242195129, "learning_rate": 2.6283511009987196e-05, "loss": 0.0396, "step": 3669 }, { "epoch": 0.25219901044529963, "grad_norm": 0.8582093119621277, "learning_rate": 2.6281310961385725e-05, "loss": 0.0293, "step": 3670 }, { "epoch": 0.2522677295217152, "grad_norm": 0.6751417517662048, "learning_rate": 2.6279110353931305e-05, "loss": 0.0423, "step": 3671 }, { "epoch": 0.2523364485981308, "grad_norm": 1.2024897336959839, "learning_rate": 2.627690918773294e-05, "loss": 0.0446, "step": 3672 }, { "epoch": 0.25240516767454646, "grad_norm": 0.9074002504348755, "learning_rate": 2.6274707462899684e-05, "loss": 0.0354, "step": 3673 }, { "epoch": 0.25247388675096205, "grad_norm": 1.2387642860412598, "learning_rate": 2.6272505179540597e-05, "loss": 0.0366, "step": 3674 }, { "epoch": 0.2525426058273777, "grad_norm": 1.4881514310836792, "learning_rate": 2.6270302337764783e-05, "loss": 0.0407, "step": 3675 }, { "epoch": 0.2526113249037933, "grad_norm": 0.7546269297599792, "learning_rate": 2.626809893768136e-05, "loss": 0.0513, "step": 3676 }, { "epoch": 0.2526800439802089, "grad_norm": 1.11362886428833, "learning_rate": 2.6265894979399483e-05, "loss": 0.0346, "step": 3677 }, { "epoch": 0.2527487630566245, "grad_norm": 1.262439489364624, "learning_rate": 2.6263690463028328e-05, "loss": 0.0383, "step": 3678 }, { "epoch": 0.2528174821330401, "grad_norm": 0.92835533618927, "learning_rate": 2.6261485388677105e-05, "loss": 0.0328, "step": 3679 }, { "epoch": 0.25288620120945576, "grad_norm": 0.7671427726745605, "learning_rate": 2.6259279756455052e-05, "loss": 0.0458, "step": 3680 }, { "epoch": 0.25295492028587135, "grad_norm": 0.8421193957328796, "learning_rate": 2.6257073566471426e-05, "loss": 0.0314, "step": 3681 }, { "epoch": 0.25302363936228695, "grad_norm": 0.40987083315849304, "learning_rate": 2.6254866818835517e-05, "loss": 0.0323, "step": 3682 }, { "epoch": 0.2530923584387026, "grad_norm": 0.8759261965751648, "learning_rate": 2.6252659513656647e-05, "loss": 0.0369, "step": 3683 }, { "epoch": 0.2531610775151182, "grad_norm": 1.051247000694275, "learning_rate": 2.6250451651044158e-05, "loss": 0.0251, "step": 3684 }, { "epoch": 0.25322979659153383, "grad_norm": 1.4937524795532227, "learning_rate": 2.6248243231107422e-05, "loss": 0.0459, "step": 3685 }, { "epoch": 0.2532985156679494, "grad_norm": 0.8494695425033569, "learning_rate": 2.6246034253955845e-05, "loss": 0.03, "step": 3686 }, { "epoch": 0.253367234744365, "grad_norm": 0.8814107179641724, "learning_rate": 2.624382471969885e-05, "loss": 0.045, "step": 3687 }, { "epoch": 0.25343595382078066, "grad_norm": 0.964643120765686, "learning_rate": 2.6241614628445898e-05, "loss": 0.0201, "step": 3688 }, { "epoch": 0.25350467289719625, "grad_norm": 1.6098231077194214, "learning_rate": 2.623940398030647e-05, "loss": 0.032, "step": 3689 }, { "epoch": 0.2535733919736119, "grad_norm": 1.1327710151672363, "learning_rate": 2.623719277539007e-05, "loss": 0.0346, "step": 3690 }, { "epoch": 0.2536421110500275, "grad_norm": 1.5714761018753052, "learning_rate": 2.6234981013806244e-05, "loss": 0.0363, "step": 3691 }, { "epoch": 0.2537108301264431, "grad_norm": 0.9204952716827393, "learning_rate": 2.6232768695664558e-05, "loss": 0.0351, "step": 3692 }, { "epoch": 0.2537795492028587, "grad_norm": 0.4561794102191925, "learning_rate": 2.6230555821074606e-05, "loss": 0.0263, "step": 3693 }, { "epoch": 0.2538482682792743, "grad_norm": 0.6958732604980469, "learning_rate": 2.6228342390146004e-05, "loss": 0.0253, "step": 3694 }, { "epoch": 0.25391698735568996, "grad_norm": 1.3525187969207764, "learning_rate": 2.622612840298841e-05, "loss": 0.0476, "step": 3695 }, { "epoch": 0.25398570643210555, "grad_norm": 1.0565900802612305, "learning_rate": 2.622391385971149e-05, "loss": 0.0287, "step": 3696 }, { "epoch": 0.25405442550852114, "grad_norm": 0.9029991626739502, "learning_rate": 2.622169876042495e-05, "loss": 0.036, "step": 3697 }, { "epoch": 0.2541231445849368, "grad_norm": 1.3783833980560303, "learning_rate": 2.6219483105238533e-05, "loss": 0.0509, "step": 3698 }, { "epoch": 0.2541918636613524, "grad_norm": 2.697124481201172, "learning_rate": 2.6217266894261984e-05, "loss": 0.0342, "step": 3699 }, { "epoch": 0.254260582737768, "grad_norm": 0.5757362246513367, "learning_rate": 2.62150501276051e-05, "loss": 0.0366, "step": 3700 }, { "epoch": 0.2543293018141836, "grad_norm": 1.008651614189148, "learning_rate": 2.6212832805377685e-05, "loss": 0.0367, "step": 3701 }, { "epoch": 0.2543980208905992, "grad_norm": 0.9804380536079407, "learning_rate": 2.6210614927689584e-05, "loss": 0.0264, "step": 3702 }, { "epoch": 0.25446673996701485, "grad_norm": 1.2107740640640259, "learning_rate": 2.620839649465067e-05, "loss": 0.0313, "step": 3703 }, { "epoch": 0.25453545904343045, "grad_norm": 0.9295381307601929, "learning_rate": 2.6206177506370835e-05, "loss": 0.0482, "step": 3704 }, { "epoch": 0.2546041781198461, "grad_norm": 1.9095877408981323, "learning_rate": 2.620395796296001e-05, "loss": 0.0451, "step": 3705 }, { "epoch": 0.2546728971962617, "grad_norm": 0.9215523600578308, "learning_rate": 2.6201737864528136e-05, "loss": 0.0337, "step": 3706 }, { "epoch": 0.2547416162726773, "grad_norm": 1.168399691581726, "learning_rate": 2.6199517211185204e-05, "loss": 0.0463, "step": 3707 }, { "epoch": 0.2548103353490929, "grad_norm": 1.1527329683303833, "learning_rate": 2.619729600304121e-05, "loss": 0.0459, "step": 3708 }, { "epoch": 0.2548790544255085, "grad_norm": 1.1566588878631592, "learning_rate": 2.6195074240206192e-05, "loss": 0.0304, "step": 3709 }, { "epoch": 0.25494777350192416, "grad_norm": 1.6258634328842163, "learning_rate": 2.6192851922790216e-05, "loss": 0.0456, "step": 3710 }, { "epoch": 0.25501649257833975, "grad_norm": 0.6479648351669312, "learning_rate": 2.619062905090337e-05, "loss": 0.045, "step": 3711 }, { "epoch": 0.25508521165475534, "grad_norm": 1.5108355283737183, "learning_rate": 2.618840562465576e-05, "loss": 0.0417, "step": 3712 }, { "epoch": 0.255153930731171, "grad_norm": 1.8306643962860107, "learning_rate": 2.618618164415754e-05, "loss": 0.0399, "step": 3713 }, { "epoch": 0.2552226498075866, "grad_norm": 9.714500427246094, "learning_rate": 2.618395710951888e-05, "loss": 0.0399, "step": 3714 }, { "epoch": 0.2552913688840022, "grad_norm": 2.0608277320861816, "learning_rate": 2.6181732020849978e-05, "loss": 0.0509, "step": 3715 }, { "epoch": 0.2553600879604178, "grad_norm": 2.727376937866211, "learning_rate": 2.6179506378261064e-05, "loss": 0.0511, "step": 3716 }, { "epoch": 0.2554288070368334, "grad_norm": 0.733939528465271, "learning_rate": 2.6177280181862385e-05, "loss": 0.0335, "step": 3717 }, { "epoch": 0.25549752611324905, "grad_norm": 1.1197946071624756, "learning_rate": 2.617505343176422e-05, "loss": 0.0489, "step": 3718 }, { "epoch": 0.25556624518966464, "grad_norm": 1.1101614236831665, "learning_rate": 2.617282612807689e-05, "loss": 0.0345, "step": 3719 }, { "epoch": 0.2556349642660803, "grad_norm": 1.5684669017791748, "learning_rate": 2.6170598270910723e-05, "loss": 0.0433, "step": 3720 }, { "epoch": 0.2557036833424959, "grad_norm": 1.2475439310073853, "learning_rate": 2.616836986037608e-05, "loss": 0.0439, "step": 3721 }, { "epoch": 0.25577240241891147, "grad_norm": 1.3404313325881958, "learning_rate": 2.6166140896583358e-05, "loss": 0.0422, "step": 3722 }, { "epoch": 0.2558411214953271, "grad_norm": 1.0021523237228394, "learning_rate": 2.6163911379642974e-05, "loss": 0.0545, "step": 3723 }, { "epoch": 0.2559098405717427, "grad_norm": 0.9611600041389465, "learning_rate": 2.6161681309665368e-05, "loss": 0.0552, "step": 3724 }, { "epoch": 0.25597855964815835, "grad_norm": 1.5627624988555908, "learning_rate": 2.615945068676102e-05, "loss": 0.0325, "step": 3725 }, { "epoch": 0.25604727872457395, "grad_norm": 0.9635158777236938, "learning_rate": 2.6157219511040428e-05, "loss": 0.0442, "step": 3726 }, { "epoch": 0.25611599780098954, "grad_norm": 1.068444013595581, "learning_rate": 2.615498778261412e-05, "loss": 0.0338, "step": 3727 }, { "epoch": 0.2561847168774052, "grad_norm": 0.8806588053703308, "learning_rate": 2.6152755501592652e-05, "loss": 0.0377, "step": 3728 }, { "epoch": 0.2562534359538208, "grad_norm": 0.7946787476539612, "learning_rate": 2.6150522668086602e-05, "loss": 0.0301, "step": 3729 }, { "epoch": 0.2563221550302364, "grad_norm": 2.711362361907959, "learning_rate": 2.614828928220659e-05, "loss": 0.0577, "step": 3730 }, { "epoch": 0.256390874106652, "grad_norm": 1.4399027824401855, "learning_rate": 2.614605534406324e-05, "loss": 0.0399, "step": 3731 }, { "epoch": 0.2564595931830676, "grad_norm": 1.639866590499878, "learning_rate": 2.6143820853767225e-05, "loss": 0.0434, "step": 3732 }, { "epoch": 0.25652831225948325, "grad_norm": 3.2274692058563232, "learning_rate": 2.6141585811429238e-05, "loss": 0.0608, "step": 3733 }, { "epoch": 0.25659703133589884, "grad_norm": 1.4373598098754883, "learning_rate": 2.6139350217159995e-05, "loss": 0.0495, "step": 3734 }, { "epoch": 0.2566657504123145, "grad_norm": 2.664033889770508, "learning_rate": 2.6137114071070244e-05, "loss": 0.0646, "step": 3735 }, { "epoch": 0.2567344694887301, "grad_norm": 1.1456208229064941, "learning_rate": 2.6134877373270757e-05, "loss": 0.0465, "step": 3736 }, { "epoch": 0.25680318856514567, "grad_norm": 1.8380717039108276, "learning_rate": 2.613264012387234e-05, "loss": 0.0507, "step": 3737 }, { "epoch": 0.2568719076415613, "grad_norm": 1.0180355310440063, "learning_rate": 2.6130402322985815e-05, "loss": 0.0492, "step": 3738 }, { "epoch": 0.2569406267179769, "grad_norm": 0.5768213272094727, "learning_rate": 2.6128163970722045e-05, "loss": 0.0363, "step": 3739 }, { "epoch": 0.2570093457943925, "grad_norm": 2.0813753604888916, "learning_rate": 2.612592506719191e-05, "loss": 0.0914, "step": 3740 }, { "epoch": 0.25707806487080814, "grad_norm": 1.4254993200302124, "learning_rate": 2.6123685612506323e-05, "loss": 0.048, "step": 3741 }, { "epoch": 0.25714678394722373, "grad_norm": 1.1094578504562378, "learning_rate": 2.6121445606776216e-05, "loss": 0.0339, "step": 3742 }, { "epoch": 0.2572155030236394, "grad_norm": 1.1328418254852295, "learning_rate": 2.6119205050112566e-05, "loss": 0.0391, "step": 3743 }, { "epoch": 0.25728422210005497, "grad_norm": 1.4136279821395874, "learning_rate": 2.6116963942626348e-05, "loss": 0.0474, "step": 3744 }, { "epoch": 0.25735294117647056, "grad_norm": 0.9956688284873962, "learning_rate": 2.6114722284428594e-05, "loss": 0.0267, "step": 3745 }, { "epoch": 0.2574216602528862, "grad_norm": 1.2676349878311157, "learning_rate": 2.6112480075630352e-05, "loss": 0.0361, "step": 3746 }, { "epoch": 0.2574903793293018, "grad_norm": 1.2518230676651, "learning_rate": 2.6110237316342693e-05, "loss": 0.0348, "step": 3747 }, { "epoch": 0.25755909840571745, "grad_norm": 1.102413535118103, "learning_rate": 2.610799400667672e-05, "loss": 0.0458, "step": 3748 }, { "epoch": 0.25762781748213304, "grad_norm": 0.664614200592041, "learning_rate": 2.610575014674356e-05, "loss": 0.0302, "step": 3749 }, { "epoch": 0.25769653655854863, "grad_norm": 1.8094286918640137, "learning_rate": 2.6103505736654368e-05, "loss": 0.0519, "step": 3750 }, { "epoch": 0.2577652556349643, "grad_norm": 2.6391234397888184, "learning_rate": 2.6101260776520335e-05, "loss": 0.0367, "step": 3751 }, { "epoch": 0.25783397471137987, "grad_norm": 1.1002360582351685, "learning_rate": 2.6099015266452658e-05, "loss": 0.0393, "step": 3752 }, { "epoch": 0.2579026937877955, "grad_norm": 2.5508315563201904, "learning_rate": 2.609676920656259e-05, "loss": 0.0488, "step": 3753 }, { "epoch": 0.2579714128642111, "grad_norm": 0.8775990009307861, "learning_rate": 2.609452259696139e-05, "loss": 0.0387, "step": 3754 }, { "epoch": 0.2580401319406267, "grad_norm": 1.0267599821090698, "learning_rate": 2.6092275437760345e-05, "loss": 0.0404, "step": 3755 }, { "epoch": 0.25810885101704234, "grad_norm": 1.138221025466919, "learning_rate": 2.6090027729070785e-05, "loss": 0.0388, "step": 3756 }, { "epoch": 0.25817757009345793, "grad_norm": 0.7736098170280457, "learning_rate": 2.6087779471004048e-05, "loss": 0.0408, "step": 3757 }, { "epoch": 0.2582462891698736, "grad_norm": 0.9158902168273926, "learning_rate": 2.6085530663671512e-05, "loss": 0.0394, "step": 3758 }, { "epoch": 0.25831500824628917, "grad_norm": 0.6266359090805054, "learning_rate": 2.6083281307184575e-05, "loss": 0.0354, "step": 3759 }, { "epoch": 0.25838372732270476, "grad_norm": 1.7694921493530273, "learning_rate": 2.6081031401654673e-05, "loss": 0.0544, "step": 3760 }, { "epoch": 0.2584524463991204, "grad_norm": 0.9624635577201843, "learning_rate": 2.6078780947193252e-05, "loss": 0.0478, "step": 3761 }, { "epoch": 0.258521165475536, "grad_norm": 0.5001574158668518, "learning_rate": 2.6076529943911808e-05, "loss": 0.0315, "step": 3762 }, { "epoch": 0.25858988455195164, "grad_norm": 0.8319532871246338, "learning_rate": 2.6074278391921835e-05, "loss": 0.0341, "step": 3763 }, { "epoch": 0.25865860362836723, "grad_norm": 1.0756474733352661, "learning_rate": 2.6072026291334883e-05, "loss": 0.0433, "step": 3764 }, { "epoch": 0.2587273227047828, "grad_norm": 1.005537748336792, "learning_rate": 2.6069773642262507e-05, "loss": 0.0334, "step": 3765 }, { "epoch": 0.25879604178119847, "grad_norm": 1.0548287630081177, "learning_rate": 2.606752044481631e-05, "loss": 0.0446, "step": 3766 }, { "epoch": 0.25886476085761406, "grad_norm": 0.6950393915176392, "learning_rate": 2.60652666991079e-05, "loss": 0.0324, "step": 3767 }, { "epoch": 0.2589334799340297, "grad_norm": 0.8179914951324463, "learning_rate": 2.606301240524892e-05, "loss": 0.0407, "step": 3768 }, { "epoch": 0.2590021990104453, "grad_norm": 1.0489648580551147, "learning_rate": 2.606075756335106e-05, "loss": 0.0432, "step": 3769 }, { "epoch": 0.2590709180868609, "grad_norm": 1.1946748495101929, "learning_rate": 2.605850217352601e-05, "loss": 0.0376, "step": 3770 }, { "epoch": 0.25913963716327654, "grad_norm": 1.1293976306915283, "learning_rate": 2.6056246235885495e-05, "loss": 0.0368, "step": 3771 }, { "epoch": 0.25920835623969213, "grad_norm": 1.771828055381775, "learning_rate": 2.6053989750541273e-05, "loss": 0.0629, "step": 3772 }, { "epoch": 0.2592770753161078, "grad_norm": 0.5973517298698425, "learning_rate": 2.605173271760512e-05, "loss": 0.0352, "step": 3773 }, { "epoch": 0.25934579439252337, "grad_norm": 1.0977510213851929, "learning_rate": 2.6049475137188855e-05, "loss": 0.0347, "step": 3774 }, { "epoch": 0.25941451346893896, "grad_norm": 0.8702496886253357, "learning_rate": 2.604721700940431e-05, "loss": 0.0599, "step": 3775 }, { "epoch": 0.2594832325453546, "grad_norm": 1.2025375366210938, "learning_rate": 2.6044958334363343e-05, "loss": 0.0403, "step": 3776 }, { "epoch": 0.2595519516217702, "grad_norm": 0.8739117383956909, "learning_rate": 2.6042699112177846e-05, "loss": 0.0307, "step": 3777 }, { "epoch": 0.25962067069818584, "grad_norm": 0.7321233153343201, "learning_rate": 2.604043934295974e-05, "loss": 0.0421, "step": 3778 }, { "epoch": 0.25968938977460143, "grad_norm": 0.5913907885551453, "learning_rate": 2.6038179026820966e-05, "loss": 0.0305, "step": 3779 }, { "epoch": 0.259758108851017, "grad_norm": 1.2478022575378418, "learning_rate": 2.6035918163873494e-05, "loss": 0.0459, "step": 3780 }, { "epoch": 0.25982682792743267, "grad_norm": 1.059373140335083, "learning_rate": 2.603365675422933e-05, "loss": 0.0419, "step": 3781 }, { "epoch": 0.25989554700384826, "grad_norm": 1.1535801887512207, "learning_rate": 2.6031394798000497e-05, "loss": 0.0506, "step": 3782 }, { "epoch": 0.2599642660802639, "grad_norm": 3.502826690673828, "learning_rate": 2.6029132295299042e-05, "loss": 0.0529, "step": 3783 }, { "epoch": 0.2600329851566795, "grad_norm": 1.6311826705932617, "learning_rate": 2.6026869246237047e-05, "loss": 0.0392, "step": 3784 }, { "epoch": 0.2601017042330951, "grad_norm": 1.461283802986145, "learning_rate": 2.602460565092662e-05, "loss": 0.0362, "step": 3785 }, { "epoch": 0.26017042330951073, "grad_norm": 0.7580013871192932, "learning_rate": 2.6022341509479895e-05, "loss": 0.032, "step": 3786 }, { "epoch": 0.2602391423859263, "grad_norm": 1.1719741821289062, "learning_rate": 2.6020076822009032e-05, "loss": 0.0332, "step": 3787 }, { "epoch": 0.26030786146234197, "grad_norm": 0.7196006178855896, "learning_rate": 2.6017811588626218e-05, "loss": 0.0372, "step": 3788 }, { "epoch": 0.26037658053875756, "grad_norm": 1.0078972578048706, "learning_rate": 2.6015545809443677e-05, "loss": 0.0409, "step": 3789 }, { "epoch": 0.26044529961517315, "grad_norm": 0.960383951663971, "learning_rate": 2.6013279484573637e-05, "loss": 0.0469, "step": 3790 }, { "epoch": 0.2605140186915888, "grad_norm": 1.4416903257369995, "learning_rate": 2.6011012614128372e-05, "loss": 0.0479, "step": 3791 }, { "epoch": 0.2605827377680044, "grad_norm": 0.8629319071769714, "learning_rate": 2.6008745198220184e-05, "loss": 0.037, "step": 3792 }, { "epoch": 0.26065145684442004, "grad_norm": 0.8390651345252991, "learning_rate": 2.600647723696139e-05, "loss": 0.039, "step": 3793 }, { "epoch": 0.26072017592083563, "grad_norm": 1.3074779510498047, "learning_rate": 2.600420873046434e-05, "loss": 0.0508, "step": 3794 }, { "epoch": 0.2607888949972512, "grad_norm": 1.7114907503128052, "learning_rate": 2.600193967884142e-05, "loss": 0.0484, "step": 3795 }, { "epoch": 0.26085761407366687, "grad_norm": 3.3916032314300537, "learning_rate": 2.599967008220502e-05, "loss": 0.0444, "step": 3796 }, { "epoch": 0.26092633315008246, "grad_norm": 0.9575221538543701, "learning_rate": 2.5997399940667584e-05, "loss": 0.0427, "step": 3797 }, { "epoch": 0.2609950522264981, "grad_norm": 1.3919674158096313, "learning_rate": 2.5995129254341562e-05, "loss": 0.0355, "step": 3798 }, { "epoch": 0.2610637713029137, "grad_norm": 1.0920668840408325, "learning_rate": 2.5992858023339442e-05, "loss": 0.0429, "step": 3799 }, { "epoch": 0.2611324903793293, "grad_norm": 0.794845700263977, "learning_rate": 2.5990586247773735e-05, "loss": 0.0392, "step": 3800 }, { "epoch": 0.26120120945574493, "grad_norm": 1.0555261373519897, "learning_rate": 2.598831392775698e-05, "loss": 0.0484, "step": 3801 }, { "epoch": 0.2612699285321605, "grad_norm": 1.8376502990722656, "learning_rate": 2.5986041063401748e-05, "loss": 0.0599, "step": 3802 }, { "epoch": 0.2613386476085761, "grad_norm": 0.97515469789505, "learning_rate": 2.5983767654820626e-05, "loss": 0.0461, "step": 3803 }, { "epoch": 0.26140736668499176, "grad_norm": 1.1096643209457397, "learning_rate": 2.5981493702126236e-05, "loss": 0.0491, "step": 3804 }, { "epoch": 0.26147608576140735, "grad_norm": 2.062908887863159, "learning_rate": 2.5979219205431228e-05, "loss": 0.0519, "step": 3805 }, { "epoch": 0.261544804837823, "grad_norm": 0.958598792552948, "learning_rate": 2.5976944164848267e-05, "loss": 0.0445, "step": 3806 }, { "epoch": 0.2616135239142386, "grad_norm": 1.4241454601287842, "learning_rate": 2.5974668580490066e-05, "loss": 0.0635, "step": 3807 }, { "epoch": 0.2616822429906542, "grad_norm": 1.0672564506530762, "learning_rate": 2.597239245246934e-05, "loss": 0.0386, "step": 3808 }, { "epoch": 0.2617509620670698, "grad_norm": 1.849552869796753, "learning_rate": 2.5970115780898853e-05, "loss": 0.0523, "step": 3809 }, { "epoch": 0.2618196811434854, "grad_norm": 1.5132235288619995, "learning_rate": 2.5967838565891385e-05, "loss": 0.0445, "step": 3810 }, { "epoch": 0.26188840021990106, "grad_norm": 1.1302508115768433, "learning_rate": 2.5965560807559743e-05, "loss": 0.0531, "step": 3811 }, { "epoch": 0.26195711929631665, "grad_norm": 1.33195161819458, "learning_rate": 2.5963282506016764e-05, "loss": 0.0402, "step": 3812 }, { "epoch": 0.26202583837273224, "grad_norm": 0.6736148595809937, "learning_rate": 2.5961003661375308e-05, "loss": 0.0174, "step": 3813 }, { "epoch": 0.2620945574491479, "grad_norm": 1.4068304300308228, "learning_rate": 2.5958724273748263e-05, "loss": 0.0414, "step": 3814 }, { "epoch": 0.2621632765255635, "grad_norm": 0.6597270369529724, "learning_rate": 2.595644434324855e-05, "loss": 0.0431, "step": 3815 }, { "epoch": 0.26223199560197913, "grad_norm": 0.7616124153137207, "learning_rate": 2.595416386998911e-05, "loss": 0.0402, "step": 3816 }, { "epoch": 0.2623007146783947, "grad_norm": 1.5528669357299805, "learning_rate": 2.5951882854082913e-05, "loss": 0.0704, "step": 3817 }, { "epoch": 0.2623694337548103, "grad_norm": 1.4020293951034546, "learning_rate": 2.5949601295642955e-05, "loss": 0.0427, "step": 3818 }, { "epoch": 0.26243815283122596, "grad_norm": 1.240494966506958, "learning_rate": 2.594731919478226e-05, "loss": 0.0347, "step": 3819 }, { "epoch": 0.26250687190764155, "grad_norm": 2.0570595264434814, "learning_rate": 2.5945036551613878e-05, "loss": 0.0462, "step": 3820 }, { "epoch": 0.2625755909840572, "grad_norm": 0.9097069501876831, "learning_rate": 2.5942753366250887e-05, "loss": 0.0374, "step": 3821 }, { "epoch": 0.2626443100604728, "grad_norm": 0.9041296243667603, "learning_rate": 2.5940469638806393e-05, "loss": 0.0354, "step": 3822 }, { "epoch": 0.2627130291368884, "grad_norm": 1.271438717842102, "learning_rate": 2.5938185369393525e-05, "loss": 0.0485, "step": 3823 }, { "epoch": 0.262781748213304, "grad_norm": 1.1868537664413452, "learning_rate": 2.5935900558125442e-05, "loss": 0.0428, "step": 3824 }, { "epoch": 0.2628504672897196, "grad_norm": 0.8780240416526794, "learning_rate": 2.593361520511533e-05, "loss": 0.0396, "step": 3825 }, { "epoch": 0.26291918636613526, "grad_norm": 1.7723747491836548, "learning_rate": 2.5931329310476396e-05, "loss": 0.0497, "step": 3826 }, { "epoch": 0.26298790544255085, "grad_norm": 11.210986137390137, "learning_rate": 2.5929042874321886e-05, "loss": 0.0469, "step": 3827 }, { "epoch": 0.26305662451896644, "grad_norm": 3.7004501819610596, "learning_rate": 2.5926755896765058e-05, "loss": 0.0367, "step": 3828 }, { "epoch": 0.2631253435953821, "grad_norm": 2.303565740585327, "learning_rate": 2.592446837791921e-05, "loss": 0.0632, "step": 3829 }, { "epoch": 0.2631940626717977, "grad_norm": 1.1462199687957764, "learning_rate": 2.5922180317897658e-05, "loss": 0.0603, "step": 3830 }, { "epoch": 0.2632627817482133, "grad_norm": 1.067769169807434, "learning_rate": 2.5919891716813745e-05, "loss": 0.0522, "step": 3831 }, { "epoch": 0.2633315008246289, "grad_norm": 2.1343770027160645, "learning_rate": 2.591760257478085e-05, "loss": 0.068, "step": 3832 }, { "epoch": 0.2634002199010445, "grad_norm": 1.1581683158874512, "learning_rate": 2.5915312891912363e-05, "loss": 0.0434, "step": 3833 }, { "epoch": 0.26346893897746015, "grad_norm": 0.8121318221092224, "learning_rate": 2.5913022668321726e-05, "loss": 0.0482, "step": 3834 }, { "epoch": 0.26353765805387575, "grad_norm": 1.7722805738449097, "learning_rate": 2.5910731904122372e-05, "loss": 0.0302, "step": 3835 }, { "epoch": 0.2636063771302914, "grad_norm": 1.1382235288619995, "learning_rate": 2.5908440599427802e-05, "loss": 0.0573, "step": 3836 }, { "epoch": 0.263675096206707, "grad_norm": 1.4233311414718628, "learning_rate": 2.59061487543515e-05, "loss": 0.0288, "step": 3837 }, { "epoch": 0.2637438152831226, "grad_norm": 0.8275502324104309, "learning_rate": 2.590385636900702e-05, "loss": 0.0392, "step": 3838 }, { "epoch": 0.2638125343595382, "grad_norm": 2.0971505641937256, "learning_rate": 2.590156344350791e-05, "loss": 0.0417, "step": 3839 }, { "epoch": 0.2638812534359538, "grad_norm": 0.7789492011070251, "learning_rate": 2.5899269977967754e-05, "loss": 0.0336, "step": 3840 }, { "epoch": 0.26394997251236946, "grad_norm": 1.1638267040252686, "learning_rate": 2.5896975972500178e-05, "loss": 0.0448, "step": 3841 }, { "epoch": 0.26401869158878505, "grad_norm": 1.0300565958023071, "learning_rate": 2.5894681427218813e-05, "loss": 0.0333, "step": 3842 }, { "epoch": 0.26408741066520064, "grad_norm": 0.7876601219177246, "learning_rate": 2.5892386342237328e-05, "loss": 0.0332, "step": 3843 }, { "epoch": 0.2641561297416163, "grad_norm": 0.6665454506874084, "learning_rate": 2.5890090717669424e-05, "loss": 0.0342, "step": 3844 }, { "epoch": 0.2642248488180319, "grad_norm": 1.6388527154922485, "learning_rate": 2.5887794553628805e-05, "loss": 0.0375, "step": 3845 }, { "epoch": 0.2642935678944475, "grad_norm": 0.7313417792320251, "learning_rate": 2.5885497850229236e-05, "loss": 0.0267, "step": 3846 }, { "epoch": 0.2643622869708631, "grad_norm": 1.467982292175293, "learning_rate": 2.5883200607584478e-05, "loss": 0.0458, "step": 3847 }, { "epoch": 0.2644310060472787, "grad_norm": 1.7689307928085327, "learning_rate": 2.588090282580834e-05, "loss": 0.0605, "step": 3848 }, { "epoch": 0.26449972512369435, "grad_norm": 1.5964088439941406, "learning_rate": 2.5878604505014644e-05, "loss": 0.0565, "step": 3849 }, { "epoch": 0.26456844420010994, "grad_norm": 1.039919376373291, "learning_rate": 2.5876305645317247e-05, "loss": 0.0379, "step": 3850 }, { "epoch": 0.2646371632765256, "grad_norm": 0.7455277442932129, "learning_rate": 2.587400624683003e-05, "loss": 0.0497, "step": 3851 }, { "epoch": 0.2647058823529412, "grad_norm": 1.0478588342666626, "learning_rate": 2.5871706309666895e-05, "loss": 0.05, "step": 3852 }, { "epoch": 0.26477460142935677, "grad_norm": 1.3613063097000122, "learning_rate": 2.586940583394179e-05, "loss": 0.0484, "step": 3853 }, { "epoch": 0.2648433205057724, "grad_norm": 1.4095921516418457, "learning_rate": 2.586710481976866e-05, "loss": 0.0387, "step": 3854 }, { "epoch": 0.264912039582188, "grad_norm": 0.6826646327972412, "learning_rate": 2.58648032672615e-05, "loss": 0.0354, "step": 3855 }, { "epoch": 0.26498075865860365, "grad_norm": 1.2437222003936768, "learning_rate": 2.5862501176534322e-05, "loss": 0.0473, "step": 3856 }, { "epoch": 0.26504947773501925, "grad_norm": 1.0972102880477905, "learning_rate": 2.586019854770117e-05, "loss": 0.042, "step": 3857 }, { "epoch": 0.26511819681143484, "grad_norm": 0.6790366768836975, "learning_rate": 2.5857895380876108e-05, "loss": 0.0504, "step": 3858 }, { "epoch": 0.2651869158878505, "grad_norm": 1.5907057523727417, "learning_rate": 2.5855591676173234e-05, "loss": 0.0387, "step": 3859 }, { "epoch": 0.2652556349642661, "grad_norm": 0.8917902708053589, "learning_rate": 2.5853287433706666e-05, "loss": 0.0405, "step": 3860 }, { "epoch": 0.2653243540406817, "grad_norm": 0.7806130051612854, "learning_rate": 2.585098265359055e-05, "loss": 0.0301, "step": 3861 }, { "epoch": 0.2653930731170973, "grad_norm": 1.048823356628418, "learning_rate": 2.5848677335939063e-05, "loss": 0.0435, "step": 3862 }, { "epoch": 0.2654617921935129, "grad_norm": 1.396857738494873, "learning_rate": 2.5846371480866406e-05, "loss": 0.0402, "step": 3863 }, { "epoch": 0.26553051126992855, "grad_norm": 1.224796175956726, "learning_rate": 2.5844065088486804e-05, "loss": 0.0316, "step": 3864 }, { "epoch": 0.26559923034634414, "grad_norm": 0.7023555636405945, "learning_rate": 2.584175815891451e-05, "loss": 0.035, "step": 3865 }, { "epoch": 0.26566794942275973, "grad_norm": 1.4167438745498657, "learning_rate": 2.5839450692263808e-05, "loss": 0.0429, "step": 3866 }, { "epoch": 0.2657366684991754, "grad_norm": 0.7810145020484924, "learning_rate": 2.5837142688649006e-05, "loss": 0.0313, "step": 3867 }, { "epoch": 0.26580538757559097, "grad_norm": 0.6672233939170837, "learning_rate": 2.583483414818443e-05, "loss": 0.0273, "step": 3868 }, { "epoch": 0.2658741066520066, "grad_norm": 1.170359492301941, "learning_rate": 2.5832525070984453e-05, "loss": 0.0308, "step": 3869 }, { "epoch": 0.2659428257284222, "grad_norm": 1.1550506353378296, "learning_rate": 2.583021545716345e-05, "loss": 0.0379, "step": 3870 }, { "epoch": 0.2660115448048378, "grad_norm": 1.4649704694747925, "learning_rate": 2.582790530683584e-05, "loss": 0.0373, "step": 3871 }, { "epoch": 0.26608026388125344, "grad_norm": 1.4049049615859985, "learning_rate": 2.5825594620116062e-05, "loss": 0.0449, "step": 3872 }, { "epoch": 0.26614898295766903, "grad_norm": 0.7791416049003601, "learning_rate": 2.5823283397118588e-05, "loss": 0.0373, "step": 3873 }, { "epoch": 0.2662177020340847, "grad_norm": 0.735333263874054, "learning_rate": 2.58209716379579e-05, "loss": 0.0321, "step": 3874 }, { "epoch": 0.26628642111050027, "grad_norm": 1.6504747867584229, "learning_rate": 2.5818659342748523e-05, "loss": 0.0443, "step": 3875 }, { "epoch": 0.26635514018691586, "grad_norm": 1.5271332263946533, "learning_rate": 2.581634651160501e-05, "loss": 0.0394, "step": 3876 }, { "epoch": 0.2664238592633315, "grad_norm": 0.7235339283943176, "learning_rate": 2.5814033144641925e-05, "loss": 0.0322, "step": 3877 }, { "epoch": 0.2664925783397471, "grad_norm": 1.1024158000946045, "learning_rate": 2.5811719241973868e-05, "loss": 0.038, "step": 3878 }, { "epoch": 0.26656129741616275, "grad_norm": 1.2624027729034424, "learning_rate": 2.5809404803715474e-05, "loss": 0.0314, "step": 3879 }, { "epoch": 0.26663001649257834, "grad_norm": 1.681154489517212, "learning_rate": 2.5807089829981383e-05, "loss": 0.0368, "step": 3880 }, { "epoch": 0.2666987355689939, "grad_norm": 0.8026517629623413, "learning_rate": 2.5804774320886285e-05, "loss": 0.0277, "step": 3881 }, { "epoch": 0.2667674546454096, "grad_norm": 1.519559383392334, "learning_rate": 2.580245827654488e-05, "loss": 0.0465, "step": 3882 }, { "epoch": 0.26683617372182517, "grad_norm": 1.4420437812805176, "learning_rate": 2.58001416970719e-05, "loss": 0.0468, "step": 3883 }, { "epoch": 0.2669048927982408, "grad_norm": 0.8522385358810425, "learning_rate": 2.5797824582582107e-05, "loss": 0.0258, "step": 3884 }, { "epoch": 0.2669736118746564, "grad_norm": 0.6230365633964539, "learning_rate": 2.579550693319028e-05, "loss": 0.0281, "step": 3885 }, { "epoch": 0.267042330951072, "grad_norm": 1.6777766942977905, "learning_rate": 2.5793188749011235e-05, "loss": 0.0506, "step": 3886 }, { "epoch": 0.26711105002748764, "grad_norm": 0.8073643445968628, "learning_rate": 2.5790870030159806e-05, "loss": 0.0391, "step": 3887 }, { "epoch": 0.26717976910390323, "grad_norm": 1.7411706447601318, "learning_rate": 2.5788550776750865e-05, "loss": 0.0521, "step": 3888 }, { "epoch": 0.2672484881803189, "grad_norm": 2.932605743408203, "learning_rate": 2.57862309888993e-05, "loss": 0.0366, "step": 3889 }, { "epoch": 0.26731720725673447, "grad_norm": 1.1807140111923218, "learning_rate": 2.5783910666720026e-05, "loss": 0.0387, "step": 3890 }, { "epoch": 0.26738592633315006, "grad_norm": 0.8083082437515259, "learning_rate": 2.5781589810327988e-05, "loss": 0.0448, "step": 3891 }, { "epoch": 0.2674546454095657, "grad_norm": 2.491929769515991, "learning_rate": 2.5779268419838157e-05, "loss": 0.0344, "step": 3892 }, { "epoch": 0.2675233644859813, "grad_norm": 1.8278437852859497, "learning_rate": 2.577694649536553e-05, "loss": 0.0432, "step": 3893 }, { "epoch": 0.26759208356239694, "grad_norm": 1.458390712738037, "learning_rate": 2.5774624037025126e-05, "loss": 0.0457, "step": 3894 }, { "epoch": 0.26766080263881253, "grad_norm": 1.4782012701034546, "learning_rate": 2.5772301044932004e-05, "loss": 0.0474, "step": 3895 }, { "epoch": 0.2677295217152281, "grad_norm": 1.053821325302124, "learning_rate": 2.5769977519201235e-05, "loss": 0.0286, "step": 3896 }, { "epoch": 0.26779824079164377, "grad_norm": 1.0639716386795044, "learning_rate": 2.576765345994792e-05, "loss": 0.029, "step": 3897 }, { "epoch": 0.26786695986805936, "grad_norm": 1.1689660549163818, "learning_rate": 2.576532886728719e-05, "loss": 0.0492, "step": 3898 }, { "epoch": 0.267935678944475, "grad_norm": 1.018217921257019, "learning_rate": 2.57630037413342e-05, "loss": 0.0341, "step": 3899 }, { "epoch": 0.2680043980208906, "grad_norm": 0.9025346040725708, "learning_rate": 2.5760678082204133e-05, "loss": 0.032, "step": 3900 }, { "epoch": 0.2680731170973062, "grad_norm": 0.810662567615509, "learning_rate": 2.57583518900122e-05, "loss": 0.0441, "step": 3901 }, { "epoch": 0.26814183617372184, "grad_norm": 1.9026546478271484, "learning_rate": 2.5756025164873625e-05, "loss": 0.0336, "step": 3902 }, { "epoch": 0.26821055525013743, "grad_norm": 1.6880996227264404, "learning_rate": 2.5753697906903684e-05, "loss": 0.0374, "step": 3903 }, { "epoch": 0.2682792743265531, "grad_norm": 2.938624620437622, "learning_rate": 2.5751370116217657e-05, "loss": 0.0467, "step": 3904 }, { "epoch": 0.26834799340296867, "grad_norm": 1.6462347507476807, "learning_rate": 2.574904179293085e-05, "loss": 0.0389, "step": 3905 }, { "epoch": 0.26841671247938426, "grad_norm": 1.193637490272522, "learning_rate": 2.5746712937158623e-05, "loss": 0.031, "step": 3906 }, { "epoch": 0.2684854315557999, "grad_norm": 0.6754775047302246, "learning_rate": 2.5744383549016325e-05, "loss": 0.0322, "step": 3907 }, { "epoch": 0.2685541506322155, "grad_norm": 2.6226131916046143, "learning_rate": 2.5742053628619356e-05, "loss": 0.034, "step": 3908 }, { "epoch": 0.26862286970863114, "grad_norm": 0.9025505185127258, "learning_rate": 2.5739723176083133e-05, "loss": 0.0365, "step": 3909 }, { "epoch": 0.26869158878504673, "grad_norm": 0.9819456934928894, "learning_rate": 2.5737392191523108e-05, "loss": 0.0386, "step": 3910 }, { "epoch": 0.2687603078614623, "grad_norm": 0.7418590784072876, "learning_rate": 2.573506067505475e-05, "loss": 0.0454, "step": 3911 }, { "epoch": 0.26882902693787797, "grad_norm": 2.179842472076416, "learning_rate": 2.5732728626793554e-05, "loss": 0.0423, "step": 3912 }, { "epoch": 0.26889774601429356, "grad_norm": 1.2080923318862915, "learning_rate": 2.573039604685505e-05, "loss": 0.0377, "step": 3913 }, { "epoch": 0.2689664650907092, "grad_norm": 0.8678542375564575, "learning_rate": 2.5728062935354778e-05, "loss": 0.0373, "step": 3914 }, { "epoch": 0.2690351841671248, "grad_norm": 2.2292184829711914, "learning_rate": 2.5725729292408336e-05, "loss": 0.0664, "step": 3915 }, { "epoch": 0.2691039032435404, "grad_norm": 1.204383373260498, "learning_rate": 2.5723395118131306e-05, "loss": 0.0434, "step": 3916 }, { "epoch": 0.26917262231995603, "grad_norm": 0.9011964797973633, "learning_rate": 2.572106041263933e-05, "loss": 0.0329, "step": 3917 }, { "epoch": 0.2692413413963716, "grad_norm": 1.2311664819717407, "learning_rate": 2.571872517604807e-05, "loss": 0.0348, "step": 3918 }, { "epoch": 0.26931006047278727, "grad_norm": 1.3244634866714478, "learning_rate": 2.5716389408473192e-05, "loss": 0.0507, "step": 3919 }, { "epoch": 0.26937877954920286, "grad_norm": 0.8640870451927185, "learning_rate": 2.571405311003042e-05, "loss": 0.0441, "step": 3920 }, { "epoch": 0.26944749862561845, "grad_norm": 1.472719430923462, "learning_rate": 2.5711716280835485e-05, "loss": 0.0462, "step": 3921 }, { "epoch": 0.2695162177020341, "grad_norm": 1.1646535396575928, "learning_rate": 2.5709378921004146e-05, "loss": 0.0413, "step": 3922 }, { "epoch": 0.2695849367784497, "grad_norm": 1.0141632556915283, "learning_rate": 2.5707041030652193e-05, "loss": 0.0312, "step": 3923 }, { "epoch": 0.26965365585486534, "grad_norm": 0.8981437683105469, "learning_rate": 2.570470260989544e-05, "loss": 0.032, "step": 3924 }, { "epoch": 0.26972237493128093, "grad_norm": 0.8580530881881714, "learning_rate": 2.5702363658849726e-05, "loss": 0.0422, "step": 3925 }, { "epoch": 0.2697910940076965, "grad_norm": 1.5661945343017578, "learning_rate": 2.570002417763092e-05, "loss": 0.0519, "step": 3926 }, { "epoch": 0.26985981308411217, "grad_norm": 2.1067135334014893, "learning_rate": 2.569768416635492e-05, "loss": 0.0447, "step": 3927 }, { "epoch": 0.26992853216052776, "grad_norm": 0.9290545582771301, "learning_rate": 2.5695343625137632e-05, "loss": 0.0426, "step": 3928 }, { "epoch": 0.2699972512369434, "grad_norm": 0.7105501294136047, "learning_rate": 2.5693002554095017e-05, "loss": 0.0416, "step": 3929 }, { "epoch": 0.270065970313359, "grad_norm": 1.129420280456543, "learning_rate": 2.569066095334303e-05, "loss": 0.038, "step": 3930 }, { "epoch": 0.2701346893897746, "grad_norm": 0.9721739292144775, "learning_rate": 2.5688318822997688e-05, "loss": 0.0418, "step": 3931 }, { "epoch": 0.27020340846619023, "grad_norm": 1.1257133483886719, "learning_rate": 2.5685976163175e-05, "loss": 0.0487, "step": 3932 }, { "epoch": 0.2702721275426058, "grad_norm": 0.8214741945266724, "learning_rate": 2.5683632973991026e-05, "loss": 0.0371, "step": 3933 }, { "epoch": 0.2703408466190214, "grad_norm": 0.9721624255180359, "learning_rate": 2.5681289255561835e-05, "loss": 0.0387, "step": 3934 }, { "epoch": 0.27040956569543706, "grad_norm": 0.5205824971199036, "learning_rate": 2.567894500800354e-05, "loss": 0.0292, "step": 3935 }, { "epoch": 0.27047828477185265, "grad_norm": 1.8307958841323853, "learning_rate": 2.567660023143226e-05, "loss": 0.0464, "step": 3936 }, { "epoch": 0.2705470038482683, "grad_norm": 1.1737353801727295, "learning_rate": 2.5674254925964153e-05, "loss": 0.0334, "step": 3937 }, { "epoch": 0.2706157229246839, "grad_norm": 0.7754996418952942, "learning_rate": 2.5671909091715405e-05, "loss": 0.0384, "step": 3938 }, { "epoch": 0.2706844420010995, "grad_norm": 2.124498128890991, "learning_rate": 2.566956272880222e-05, "loss": 0.0498, "step": 3939 }, { "epoch": 0.2707531610775151, "grad_norm": 1.1378966569900513, "learning_rate": 2.5667215837340835e-05, "loss": 0.0404, "step": 3940 }, { "epoch": 0.2708218801539307, "grad_norm": 1.02993905544281, "learning_rate": 2.5664868417447504e-05, "loss": 0.0335, "step": 3941 }, { "epoch": 0.27089059923034636, "grad_norm": 1.6227126121520996, "learning_rate": 2.5662520469238524e-05, "loss": 0.0525, "step": 3942 }, { "epoch": 0.27095931830676195, "grad_norm": 1.1592854261398315, "learning_rate": 2.5660171992830197e-05, "loss": 0.0395, "step": 3943 }, { "epoch": 0.27102803738317754, "grad_norm": 2.603849411010742, "learning_rate": 2.5657822988338863e-05, "loss": 0.0463, "step": 3944 }, { "epoch": 0.2710967564595932, "grad_norm": 2.1124911308288574, "learning_rate": 2.5655473455880896e-05, "loss": 0.047, "step": 3945 }, { "epoch": 0.2711654755360088, "grad_norm": 0.8248065710067749, "learning_rate": 2.565312339557268e-05, "loss": 0.0361, "step": 3946 }, { "epoch": 0.27123419461242443, "grad_norm": 1.042410135269165, "learning_rate": 2.565077280753063e-05, "loss": 0.042, "step": 3947 }, { "epoch": 0.27130291368884, "grad_norm": 1.6927156448364258, "learning_rate": 2.5648421691871193e-05, "loss": 0.0406, "step": 3948 }, { "epoch": 0.2713716327652556, "grad_norm": 1.2734525203704834, "learning_rate": 2.5646070048710838e-05, "loss": 0.0467, "step": 3949 }, { "epoch": 0.27144035184167126, "grad_norm": 1.1542577743530273, "learning_rate": 2.564371787816606e-05, "loss": 0.0391, "step": 3950 }, { "epoch": 0.27150907091808685, "grad_norm": 1.022966980934143, "learning_rate": 2.5641365180353378e-05, "loss": 0.0271, "step": 3951 }, { "epoch": 0.2715777899945025, "grad_norm": 1.0616121292114258, "learning_rate": 2.5639011955389345e-05, "loss": 0.0292, "step": 3952 }, { "epoch": 0.2716465090709181, "grad_norm": 0.9197938442230225, "learning_rate": 2.5636658203390534e-05, "loss": 0.0249, "step": 3953 }, { "epoch": 0.2717152281473337, "grad_norm": 0.8889080286026001, "learning_rate": 2.5634303924473542e-05, "loss": 0.0317, "step": 3954 }, { "epoch": 0.2717839472237493, "grad_norm": 1.2755491733551025, "learning_rate": 2.5631949118754997e-05, "loss": 0.041, "step": 3955 }, { "epoch": 0.2718526663001649, "grad_norm": 1.1795992851257324, "learning_rate": 2.5629593786351555e-05, "loss": 0.042, "step": 3956 }, { "epoch": 0.27192138537658056, "grad_norm": 0.7762389779090881, "learning_rate": 2.5627237927379886e-05, "loss": 0.0531, "step": 3957 }, { "epoch": 0.27199010445299615, "grad_norm": 1.6609524488449097, "learning_rate": 2.56248815419567e-05, "loss": 0.0337, "step": 3958 }, { "epoch": 0.27205882352941174, "grad_norm": 1.0768567323684692, "learning_rate": 2.5622524630198726e-05, "loss": 0.028, "step": 3959 }, { "epoch": 0.2721275426058274, "grad_norm": 1.5504323244094849, "learning_rate": 2.562016719222272e-05, "loss": 0.0394, "step": 3960 }, { "epoch": 0.272196261682243, "grad_norm": 1.31676185131073, "learning_rate": 2.5617809228145463e-05, "loss": 0.0428, "step": 3961 }, { "epoch": 0.2722649807586586, "grad_norm": 0.5338101983070374, "learning_rate": 2.561545073808377e-05, "loss": 0.0318, "step": 3962 }, { "epoch": 0.2723336998350742, "grad_norm": 1.408226728439331, "learning_rate": 2.5613091722154472e-05, "loss": 0.046, "step": 3963 }, { "epoch": 0.2724024189114898, "grad_norm": 0.6549201607704163, "learning_rate": 2.5610732180474426e-05, "loss": 0.0261, "step": 3964 }, { "epoch": 0.27247113798790545, "grad_norm": 0.9286736249923706, "learning_rate": 2.560837211316053e-05, "loss": 0.0227, "step": 3965 }, { "epoch": 0.27253985706432104, "grad_norm": 1.40995192527771, "learning_rate": 2.5606011520329682e-05, "loss": 0.0442, "step": 3966 }, { "epoch": 0.2726085761407367, "grad_norm": 1.1562647819519043, "learning_rate": 2.560365040209883e-05, "loss": 0.0413, "step": 3967 }, { "epoch": 0.2726772952171523, "grad_norm": 0.548936128616333, "learning_rate": 2.5601288758584942e-05, "loss": 0.0263, "step": 3968 }, { "epoch": 0.2727460142935679, "grad_norm": 0.9567570090293884, "learning_rate": 2.5598926589905e-05, "loss": 0.0384, "step": 3969 }, { "epoch": 0.2728147333699835, "grad_norm": 0.9307979345321655, "learning_rate": 2.5596563896176027e-05, "loss": 0.0315, "step": 3970 }, { "epoch": 0.2728834524463991, "grad_norm": 1.1145858764648438, "learning_rate": 2.5594200677515057e-05, "loss": 0.0394, "step": 3971 }, { "epoch": 0.27295217152281476, "grad_norm": 1.635739803314209, "learning_rate": 2.5591836934039175e-05, "loss": 0.0356, "step": 3972 }, { "epoch": 0.27302089059923035, "grad_norm": 0.8985871076583862, "learning_rate": 2.5589472665865468e-05, "loss": 0.0429, "step": 3973 }, { "epoch": 0.27308960967564594, "grad_norm": 0.42949149012565613, "learning_rate": 2.5587107873111052e-05, "loss": 0.0279, "step": 3974 }, { "epoch": 0.2731583287520616, "grad_norm": 1.2675877809524536, "learning_rate": 2.5584742555893074e-05, "loss": 0.0494, "step": 3975 }, { "epoch": 0.2732270478284772, "grad_norm": 0.7228192687034607, "learning_rate": 2.5582376714328717e-05, "loss": 0.0308, "step": 3976 }, { "epoch": 0.2732957669048928, "grad_norm": 1.4001771211624146, "learning_rate": 2.5580010348535172e-05, "loss": 0.046, "step": 3977 }, { "epoch": 0.2733644859813084, "grad_norm": 0.66737961769104, "learning_rate": 2.5577643458629667e-05, "loss": 0.0364, "step": 3978 }, { "epoch": 0.273433205057724, "grad_norm": 0.6889703869819641, "learning_rate": 2.5575276044729452e-05, "loss": 0.0253, "step": 3979 }, { "epoch": 0.27350192413413965, "grad_norm": 1.0531721115112305, "learning_rate": 2.5572908106951803e-05, "loss": 0.0465, "step": 3980 }, { "epoch": 0.27357064321055524, "grad_norm": 1.0506420135498047, "learning_rate": 2.5570539645414025e-05, "loss": 0.0227, "step": 3981 }, { "epoch": 0.2736393622869709, "grad_norm": 0.6932816505432129, "learning_rate": 2.5568170660233442e-05, "loss": 0.0228, "step": 3982 }, { "epoch": 0.2737080813633865, "grad_norm": 1.0659446716308594, "learning_rate": 2.5565801151527412e-05, "loss": 0.0338, "step": 3983 }, { "epoch": 0.27377680043980207, "grad_norm": 0.5317087769508362, "learning_rate": 2.5563431119413317e-05, "loss": 0.0266, "step": 3984 }, { "epoch": 0.2738455195162177, "grad_norm": 1.1689280271530151, "learning_rate": 2.5561060564008562e-05, "loss": 0.0408, "step": 3985 }, { "epoch": 0.2739142385926333, "grad_norm": 0.8676916360855103, "learning_rate": 2.555868948543058e-05, "loss": 0.0278, "step": 3986 }, { "epoch": 0.27398295766904895, "grad_norm": 1.0132766962051392, "learning_rate": 2.5556317883796824e-05, "loss": 0.0316, "step": 3987 }, { "epoch": 0.27405167674546455, "grad_norm": 0.9627946019172668, "learning_rate": 2.5553945759224788e-05, "loss": 0.0297, "step": 3988 }, { "epoch": 0.27412039582188014, "grad_norm": 0.8553344011306763, "learning_rate": 2.5551573111831975e-05, "loss": 0.0328, "step": 3989 }, { "epoch": 0.2741891148982958, "grad_norm": 0.43725675344467163, "learning_rate": 2.5549199941735927e-05, "loss": 0.0346, "step": 3990 }, { "epoch": 0.2742578339747114, "grad_norm": 0.7246650457382202, "learning_rate": 2.5546826249054194e-05, "loss": 0.0319, "step": 3991 }, { "epoch": 0.274326553051127, "grad_norm": 1.9293473958969116, "learning_rate": 2.5544452033904375e-05, "loss": 0.0347, "step": 3992 }, { "epoch": 0.2743952721275426, "grad_norm": 2.0992507934570312, "learning_rate": 2.5542077296404083e-05, "loss": 0.0388, "step": 3993 }, { "epoch": 0.2744639912039582, "grad_norm": 1.1319177150726318, "learning_rate": 2.5539702036670954e-05, "loss": 0.0353, "step": 3994 }, { "epoch": 0.27453271028037385, "grad_norm": 0.5504300594329834, "learning_rate": 2.5537326254822657e-05, "loss": 0.0439, "step": 3995 }, { "epoch": 0.27460142935678944, "grad_norm": 0.5596408843994141, "learning_rate": 2.553494995097688e-05, "loss": 0.0311, "step": 3996 }, { "epoch": 0.27467014843320503, "grad_norm": 1.3861404657363892, "learning_rate": 2.553257312525134e-05, "loss": 0.0275, "step": 3997 }, { "epoch": 0.2747388675096207, "grad_norm": 1.0561537742614746, "learning_rate": 2.553019577776378e-05, "loss": 0.0357, "step": 3998 }, { "epoch": 0.27480758658603627, "grad_norm": 1.0859485864639282, "learning_rate": 2.5527817908631975e-05, "loss": 0.043, "step": 3999 }, { "epoch": 0.2748763056624519, "grad_norm": 0.782302975654602, "learning_rate": 2.5525439517973713e-05, "loss": 0.0303, "step": 4000 }, { "epoch": 0.2749450247388675, "grad_norm": 0.8063561320304871, "learning_rate": 2.5523060605906815e-05, "loss": 0.0395, "step": 4001 }, { "epoch": 0.2750137438152831, "grad_norm": 1.0786701440811157, "learning_rate": 2.552068117254913e-05, "loss": 0.0363, "step": 4002 }, { "epoch": 0.27508246289169874, "grad_norm": 1.5371485948562622, "learning_rate": 2.551830121801853e-05, "loss": 0.0475, "step": 4003 }, { "epoch": 0.27515118196811433, "grad_norm": 2.2068464756011963, "learning_rate": 2.5515920742432912e-05, "loss": 0.0591, "step": 4004 }, { "epoch": 0.27521990104453, "grad_norm": 1.2712321281433105, "learning_rate": 2.5513539745910202e-05, "loss": 0.0562, "step": 4005 }, { "epoch": 0.27528862012094557, "grad_norm": 1.634768009185791, "learning_rate": 2.5511158228568347e-05, "loss": 0.0574, "step": 4006 }, { "epoch": 0.27535733919736116, "grad_norm": 3.4718363285064697, "learning_rate": 2.5508776190525323e-05, "loss": 0.0581, "step": 4007 }, { "epoch": 0.2754260582737768, "grad_norm": 1.0412044525146484, "learning_rate": 2.5506393631899133e-05, "loss": 0.0515, "step": 4008 }, { "epoch": 0.2754947773501924, "grad_norm": 1.1562058925628662, "learning_rate": 2.55040105528078e-05, "loss": 0.0449, "step": 4009 }, { "epoch": 0.27556349642660805, "grad_norm": 1.0071132183074951, "learning_rate": 2.5501626953369376e-05, "loss": 0.0327, "step": 4010 }, { "epoch": 0.27563221550302364, "grad_norm": 2.297761917114258, "learning_rate": 2.549924283370195e-05, "loss": 0.0418, "step": 4011 }, { "epoch": 0.2757009345794392, "grad_norm": 0.8844882845878601, "learning_rate": 2.5496858193923614e-05, "loss": 0.0435, "step": 4012 }, { "epoch": 0.2757696536558549, "grad_norm": 0.6844829320907593, "learning_rate": 2.5494473034152508e-05, "loss": 0.0387, "step": 4013 }, { "epoch": 0.27583837273227046, "grad_norm": 0.9944018125534058, "learning_rate": 2.5492087354506786e-05, "loss": 0.0345, "step": 4014 }, { "epoch": 0.2759070918086861, "grad_norm": 0.9528146982192993, "learning_rate": 2.548970115510462e-05, "loss": 0.0392, "step": 4015 }, { "epoch": 0.2759758108851017, "grad_norm": 1.4817920923233032, "learning_rate": 2.548731443606423e-05, "loss": 0.0384, "step": 4016 }, { "epoch": 0.2760445299615173, "grad_norm": 0.9914230108261108, "learning_rate": 2.5484927197503842e-05, "loss": 0.0395, "step": 4017 }, { "epoch": 0.27611324903793294, "grad_norm": 1.5293420553207397, "learning_rate": 2.5482539439541722e-05, "loss": 0.0276, "step": 4018 }, { "epoch": 0.27618196811434853, "grad_norm": 1.2625348567962646, "learning_rate": 2.548015116229614e-05, "loss": 0.0355, "step": 4019 }, { "epoch": 0.2762506871907642, "grad_norm": 1.9123718738555908, "learning_rate": 2.547776236588542e-05, "loss": 0.0434, "step": 4020 }, { "epoch": 0.27631940626717977, "grad_norm": 1.4670443534851074, "learning_rate": 2.5475373050427896e-05, "loss": 0.027, "step": 4021 }, { "epoch": 0.27638812534359536, "grad_norm": 1.0743019580841064, "learning_rate": 2.547298321604192e-05, "loss": 0.0417, "step": 4022 }, { "epoch": 0.276456844420011, "grad_norm": 0.879472553730011, "learning_rate": 2.5470592862845895e-05, "loss": 0.0404, "step": 4023 }, { "epoch": 0.2765255634964266, "grad_norm": 1.3199151754379272, "learning_rate": 2.5468201990958225e-05, "loss": 0.0411, "step": 4024 }, { "epoch": 0.27659428257284224, "grad_norm": 1.0672178268432617, "learning_rate": 2.5465810600497347e-05, "loss": 0.0311, "step": 4025 }, { "epoch": 0.27666300164925783, "grad_norm": 1.5680488348007202, "learning_rate": 2.5463418691581734e-05, "loss": 0.0582, "step": 4026 }, { "epoch": 0.2767317207256734, "grad_norm": 0.7696256041526794, "learning_rate": 2.5461026264329865e-05, "loss": 0.0318, "step": 4027 }, { "epoch": 0.27680043980208907, "grad_norm": 0.6583095192909241, "learning_rate": 2.545863331886026e-05, "loss": 0.035, "step": 4028 }, { "epoch": 0.27686915887850466, "grad_norm": 1.0718307495117188, "learning_rate": 2.545623985529147e-05, "loss": 0.0352, "step": 4029 }, { "epoch": 0.2769378779549203, "grad_norm": 0.940402090549469, "learning_rate": 2.545384587374205e-05, "loss": 0.0459, "step": 4030 }, { "epoch": 0.2770065970313359, "grad_norm": 1.0469095706939697, "learning_rate": 2.54514513743306e-05, "loss": 0.052, "step": 4031 }, { "epoch": 0.2770753161077515, "grad_norm": 1.7410166263580322, "learning_rate": 2.5449056357175735e-05, "loss": 0.0528, "step": 4032 }, { "epoch": 0.27714403518416714, "grad_norm": 1.0423057079315186, "learning_rate": 2.54466608223961e-05, "loss": 0.044, "step": 4033 }, { "epoch": 0.2772127542605827, "grad_norm": 0.984044075012207, "learning_rate": 2.5444264770110366e-05, "loss": 0.0283, "step": 4034 }, { "epoch": 0.2772814733369984, "grad_norm": 1.2646900415420532, "learning_rate": 2.544186820043723e-05, "loss": 0.0282, "step": 4035 }, { "epoch": 0.27735019241341397, "grad_norm": 0.9200401306152344, "learning_rate": 2.5439471113495408e-05, "loss": 0.0381, "step": 4036 }, { "epoch": 0.27741891148982956, "grad_norm": 1.3078868389129639, "learning_rate": 2.5437073509403646e-05, "loss": 0.0496, "step": 4037 }, { "epoch": 0.2774876305662452, "grad_norm": 0.731160581111908, "learning_rate": 2.5434675388280727e-05, "loss": 0.0242, "step": 4038 }, { "epoch": 0.2775563496426608, "grad_norm": 0.6704124212265015, "learning_rate": 2.5432276750245438e-05, "loss": 0.0317, "step": 4039 }, { "epoch": 0.27762506871907644, "grad_norm": 1.3114652633666992, "learning_rate": 2.542987759541661e-05, "loss": 0.044, "step": 4040 }, { "epoch": 0.27769378779549203, "grad_norm": 0.6234468221664429, "learning_rate": 2.5427477923913087e-05, "loss": 0.0249, "step": 4041 }, { "epoch": 0.2777625068719076, "grad_norm": 0.8592579364776611, "learning_rate": 2.5425077735853746e-05, "loss": 0.0477, "step": 4042 }, { "epoch": 0.27783122594832327, "grad_norm": 1.0963023900985718, "learning_rate": 2.5422677031357482e-05, "loss": 0.0324, "step": 4043 }, { "epoch": 0.27789994502473886, "grad_norm": 1.2098501920700073, "learning_rate": 2.542027581054323e-05, "loss": 0.0488, "step": 4044 }, { "epoch": 0.2779686641011545, "grad_norm": 1.3927910327911377, "learning_rate": 2.5417874073529942e-05, "loss": 0.0423, "step": 4045 }, { "epoch": 0.2780373831775701, "grad_norm": 1.4716086387634277, "learning_rate": 2.5415471820436585e-05, "loss": 0.0481, "step": 4046 }, { "epoch": 0.2781061022539857, "grad_norm": 1.3390395641326904, "learning_rate": 2.5413069051382168e-05, "loss": 0.0441, "step": 4047 }, { "epoch": 0.27817482133040133, "grad_norm": 1.175673007965088, "learning_rate": 2.5410665766485723e-05, "loss": 0.042, "step": 4048 }, { "epoch": 0.2782435404068169, "grad_norm": 2.279855489730835, "learning_rate": 2.540826196586629e-05, "loss": 0.0365, "step": 4049 }, { "epoch": 0.27831225948323257, "grad_norm": 0.8143685460090637, "learning_rate": 2.5405857649642967e-05, "loss": 0.0281, "step": 4050 }, { "epoch": 0.27838097855964816, "grad_norm": 0.7852559089660645, "learning_rate": 2.5403452817934845e-05, "loss": 0.0259, "step": 4051 }, { "epoch": 0.27844969763606375, "grad_norm": 0.6428440809249878, "learning_rate": 2.540104747086106e-05, "loss": 0.0303, "step": 4052 }, { "epoch": 0.2785184167124794, "grad_norm": 0.9189103245735168, "learning_rate": 2.5398641608540768e-05, "loss": 0.0408, "step": 4053 }, { "epoch": 0.278587135788895, "grad_norm": 0.7646641731262207, "learning_rate": 2.5396235231093148e-05, "loss": 0.03, "step": 4054 }, { "epoch": 0.27865585486531064, "grad_norm": 1.7374542951583862, "learning_rate": 2.5393828338637407e-05, "loss": 0.032, "step": 4055 }, { "epoch": 0.27872457394172623, "grad_norm": 1.5371627807617188, "learning_rate": 2.5391420931292783e-05, "loss": 0.0381, "step": 4056 }, { "epoch": 0.2787932930181418, "grad_norm": 1.1669200658798218, "learning_rate": 2.5389013009178526e-05, "loss": 0.0322, "step": 4057 }, { "epoch": 0.27886201209455747, "grad_norm": 0.8848616480827332, "learning_rate": 2.5386604572413928e-05, "loss": 0.0419, "step": 4058 }, { "epoch": 0.27893073117097306, "grad_norm": 1.1987621784210205, "learning_rate": 2.5384195621118286e-05, "loss": 0.0264, "step": 4059 }, { "epoch": 0.2789994502473887, "grad_norm": 0.896660327911377, "learning_rate": 2.538178615541095e-05, "loss": 0.0383, "step": 4060 }, { "epoch": 0.2790681693238043, "grad_norm": 0.999988317489624, "learning_rate": 2.5379376175411264e-05, "loss": 0.037, "step": 4061 }, { "epoch": 0.2791368884002199, "grad_norm": 1.309080958366394, "learning_rate": 2.5376965681238627e-05, "loss": 0.0277, "step": 4062 }, { "epoch": 0.27920560747663553, "grad_norm": 1.2730051279067993, "learning_rate": 2.5374554673012444e-05, "loss": 0.0377, "step": 4063 }, { "epoch": 0.2792743265530511, "grad_norm": 0.7163845896720886, "learning_rate": 2.5372143150852148e-05, "loss": 0.0284, "step": 4064 }, { "epoch": 0.2793430456294667, "grad_norm": 0.9202418923377991, "learning_rate": 2.536973111487721e-05, "loss": 0.0404, "step": 4065 }, { "epoch": 0.27941176470588236, "grad_norm": 3.4466960430145264, "learning_rate": 2.5367318565207108e-05, "loss": 0.0447, "step": 4066 }, { "epoch": 0.27948048378229795, "grad_norm": 1.455474853515625, "learning_rate": 2.536490550196136e-05, "loss": 0.0358, "step": 4067 }, { "epoch": 0.2795492028587136, "grad_norm": 1.7408902645111084, "learning_rate": 2.53624919252595e-05, "loss": 0.069, "step": 4068 }, { "epoch": 0.2796179219351292, "grad_norm": 1.5449835062026978, "learning_rate": 2.53600778352211e-05, "loss": 0.0385, "step": 4069 }, { "epoch": 0.2796866410115448, "grad_norm": 0.7168439626693726, "learning_rate": 2.535766323196574e-05, "loss": 0.0349, "step": 4070 }, { "epoch": 0.2797553600879604, "grad_norm": 0.8852766752243042, "learning_rate": 2.5355248115613037e-05, "loss": 0.041, "step": 4071 }, { "epoch": 0.279824079164376, "grad_norm": 1.051437258720398, "learning_rate": 2.5352832486282635e-05, "loss": 0.0439, "step": 4072 }, { "epoch": 0.27989279824079166, "grad_norm": 0.728021502494812, "learning_rate": 2.5350416344094193e-05, "loss": 0.0262, "step": 4073 }, { "epoch": 0.27996151731720725, "grad_norm": 0.9369743466377258, "learning_rate": 2.5347999689167408e-05, "loss": 0.0532, "step": 4074 }, { "epoch": 0.28003023639362284, "grad_norm": 2.3667566776275635, "learning_rate": 2.534558252162199e-05, "loss": 0.0431, "step": 4075 }, { "epoch": 0.2800989554700385, "grad_norm": 0.6553661823272705, "learning_rate": 2.5343164841577682e-05, "loss": 0.0436, "step": 4076 }, { "epoch": 0.2801676745464541, "grad_norm": 2.6710445880889893, "learning_rate": 2.5340746649154255e-05, "loss": 0.0371, "step": 4077 }, { "epoch": 0.28023639362286973, "grad_norm": 1.12867271900177, "learning_rate": 2.53383279444715e-05, "loss": 0.0344, "step": 4078 }, { "epoch": 0.2803051126992853, "grad_norm": 2.226604461669922, "learning_rate": 2.533590872764923e-05, "loss": 0.0518, "step": 4079 }, { "epoch": 0.2803738317757009, "grad_norm": 1.3804718255996704, "learning_rate": 2.5333488998807293e-05, "loss": 0.069, "step": 4080 }, { "epoch": 0.28044255085211656, "grad_norm": 1.1476808786392212, "learning_rate": 2.5331068758065558e-05, "loss": 0.0362, "step": 4081 }, { "epoch": 0.28051126992853215, "grad_norm": 0.7344386577606201, "learning_rate": 2.5328648005543913e-05, "loss": 0.042, "step": 4082 }, { "epoch": 0.2805799890049478, "grad_norm": 0.5992791056632996, "learning_rate": 2.5326226741362283e-05, "loss": 0.0358, "step": 4083 }, { "epoch": 0.2806487080813634, "grad_norm": 1.5424634218215942, "learning_rate": 2.5323804965640607e-05, "loss": 0.051, "step": 4084 }, { "epoch": 0.280717427157779, "grad_norm": 0.9313350319862366, "learning_rate": 2.532138267849886e-05, "loss": 0.0523, "step": 4085 }, { "epoch": 0.2807861462341946, "grad_norm": 1.4144752025604248, "learning_rate": 2.5318959880057034e-05, "loss": 0.0537, "step": 4086 }, { "epoch": 0.2808548653106102, "grad_norm": 1.8062644004821777, "learning_rate": 2.5316536570435154e-05, "loss": 0.0459, "step": 4087 }, { "epoch": 0.28092358438702586, "grad_norm": 1.2441155910491943, "learning_rate": 2.5314112749753258e-05, "loss": 0.047, "step": 4088 }, { "epoch": 0.28099230346344145, "grad_norm": 0.9121044278144836, "learning_rate": 2.5311688418131425e-05, "loss": 0.0335, "step": 4089 }, { "epoch": 0.28106102253985704, "grad_norm": 0.6931139230728149, "learning_rate": 2.5309263575689745e-05, "loss": 0.0252, "step": 4090 }, { "epoch": 0.2811297416162727, "grad_norm": 0.5735581517219543, "learning_rate": 2.5306838222548343e-05, "loss": 0.0256, "step": 4091 }, { "epoch": 0.2811984606926883, "grad_norm": 0.9807848334312439, "learning_rate": 2.5304412358827365e-05, "loss": 0.0297, "step": 4092 }, { "epoch": 0.2812671797691039, "grad_norm": 1.0626333951950073, "learning_rate": 2.5301985984646987e-05, "loss": 0.0309, "step": 4093 }, { "epoch": 0.2813358988455195, "grad_norm": 0.7942291498184204, "learning_rate": 2.5299559100127397e-05, "loss": 0.037, "step": 4094 }, { "epoch": 0.2814046179219351, "grad_norm": 1.9374693632125854, "learning_rate": 2.529713170538883e-05, "loss": 0.0432, "step": 4095 }, { "epoch": 0.28147333699835075, "grad_norm": 1.8082833290100098, "learning_rate": 2.5294703800551525e-05, "loss": 0.0409, "step": 4096 }, { "epoch": 0.28154205607476634, "grad_norm": 1.339490532875061, "learning_rate": 2.5292275385735758e-05, "loss": 0.0361, "step": 4097 }, { "epoch": 0.281610775151182, "grad_norm": 1.4122031927108765, "learning_rate": 2.528984646106183e-05, "loss": 0.0292, "step": 4098 }, { "epoch": 0.2816794942275976, "grad_norm": 0.6609857678413391, "learning_rate": 2.5287417026650067e-05, "loss": 0.0313, "step": 4099 }, { "epoch": 0.2817482133040132, "grad_norm": 1.047098994255066, "learning_rate": 2.5284987082620807e-05, "loss": 0.044, "step": 4100 }, { "epoch": 0.2818169323804288, "grad_norm": 1.3247005939483643, "learning_rate": 2.5282556629094437e-05, "loss": 0.0348, "step": 4101 }, { "epoch": 0.2818856514568444, "grad_norm": 0.9038993716239929, "learning_rate": 2.5280125666191353e-05, "loss": 0.0286, "step": 4102 }, { "epoch": 0.28195437053326006, "grad_norm": 1.6583893299102783, "learning_rate": 2.5277694194031972e-05, "loss": 0.0479, "step": 4103 }, { "epoch": 0.28202308960967565, "grad_norm": 0.8779585361480713, "learning_rate": 2.5275262212736758e-05, "loss": 0.0385, "step": 4104 }, { "epoch": 0.28209180868609124, "grad_norm": 1.6597325801849365, "learning_rate": 2.5272829722426173e-05, "loss": 0.0317, "step": 4105 }, { "epoch": 0.2821605277625069, "grad_norm": 5.001282215118408, "learning_rate": 2.5270396723220725e-05, "loss": 0.0422, "step": 4106 }, { "epoch": 0.2822292468389225, "grad_norm": 1.4960819482803345, "learning_rate": 2.5267963215240943e-05, "loss": 0.0317, "step": 4107 }, { "epoch": 0.2822979659153381, "grad_norm": 1.4904989004135132, "learning_rate": 2.5265529198607373e-05, "loss": 0.0352, "step": 4108 }, { "epoch": 0.2823666849917537, "grad_norm": 1.1138956546783447, "learning_rate": 2.5263094673440584e-05, "loss": 0.032, "step": 4109 }, { "epoch": 0.2824354040681693, "grad_norm": 1.411914587020874, "learning_rate": 2.526065963986119e-05, "loss": 0.032, "step": 4110 }, { "epoch": 0.28250412314458495, "grad_norm": 0.8664442896842957, "learning_rate": 2.5258224097989813e-05, "loss": 0.0295, "step": 4111 }, { "epoch": 0.28257284222100054, "grad_norm": 0.41906604170799255, "learning_rate": 2.5255788047947103e-05, "loss": 0.0354, "step": 4112 }, { "epoch": 0.2826415612974162, "grad_norm": 1.3489034175872803, "learning_rate": 2.525335148985374e-05, "loss": 0.0313, "step": 4113 }, { "epoch": 0.2827102803738318, "grad_norm": 1.3920570611953735, "learning_rate": 2.525091442383042e-05, "loss": 0.0491, "step": 4114 }, { "epoch": 0.28277899945024737, "grad_norm": 1.1376639604568481, "learning_rate": 2.524847684999788e-05, "loss": 0.0382, "step": 4115 }, { "epoch": 0.282847718526663, "grad_norm": 3.3906197547912598, "learning_rate": 2.524603876847686e-05, "loss": 0.0356, "step": 4116 }, { "epoch": 0.2829164376030786, "grad_norm": 1.6126635074615479, "learning_rate": 2.524360017938815e-05, "loss": 0.0307, "step": 4117 }, { "epoch": 0.28298515667949425, "grad_norm": 1.226855993270874, "learning_rate": 2.5241161082852543e-05, "loss": 0.0371, "step": 4118 }, { "epoch": 0.28305387575590985, "grad_norm": 0.6877408623695374, "learning_rate": 2.5238721478990877e-05, "loss": 0.029, "step": 4119 }, { "epoch": 0.28312259483232544, "grad_norm": 1.0222158432006836, "learning_rate": 2.5236281367923993e-05, "loss": 0.0327, "step": 4120 }, { "epoch": 0.2831913139087411, "grad_norm": 1.9888063669204712, "learning_rate": 2.5233840749772774e-05, "loss": 0.0643, "step": 4121 }, { "epoch": 0.2832600329851567, "grad_norm": 0.9954826235771179, "learning_rate": 2.5231399624658133e-05, "loss": 0.0391, "step": 4122 }, { "epoch": 0.2833287520615723, "grad_norm": 0.5209417343139648, "learning_rate": 2.5228957992700982e-05, "loss": 0.026, "step": 4123 }, { "epoch": 0.2833974711379879, "grad_norm": 0.6892924308776855, "learning_rate": 2.522651585402228e-05, "loss": 0.0252, "step": 4124 }, { "epoch": 0.2834661902144035, "grad_norm": 0.7015208005905151, "learning_rate": 2.5224073208743014e-05, "loss": 0.0244, "step": 4125 }, { "epoch": 0.28353490929081915, "grad_norm": 1.2230119705200195, "learning_rate": 2.522163005698418e-05, "loss": 0.0478, "step": 4126 }, { "epoch": 0.28360362836723474, "grad_norm": 1.441346287727356, "learning_rate": 2.52191863988668e-05, "loss": 0.0385, "step": 4127 }, { "epoch": 0.28367234744365033, "grad_norm": 0.7099036574363708, "learning_rate": 2.5216742234511944e-05, "loss": 0.0405, "step": 4128 }, { "epoch": 0.283741066520066, "grad_norm": 0.9949420690536499, "learning_rate": 2.5214297564040678e-05, "loss": 0.0239, "step": 4129 }, { "epoch": 0.28380978559648157, "grad_norm": 3.8704898357391357, "learning_rate": 2.521185238757411e-05, "loss": 0.0326, "step": 4130 }, { "epoch": 0.2838785046728972, "grad_norm": 1.9871032238006592, "learning_rate": 2.5209406705233373e-05, "loss": 0.0386, "step": 4131 }, { "epoch": 0.2839472237493128, "grad_norm": 0.6216509342193604, "learning_rate": 2.5206960517139613e-05, "loss": 0.0335, "step": 4132 }, { "epoch": 0.2840159428257284, "grad_norm": 0.6153659820556641, "learning_rate": 2.520451382341402e-05, "loss": 0.0422, "step": 4133 }, { "epoch": 0.28408466190214404, "grad_norm": 0.9981658458709717, "learning_rate": 2.520206662417779e-05, "loss": 0.0402, "step": 4134 }, { "epoch": 0.28415338097855963, "grad_norm": 1.7446868419647217, "learning_rate": 2.5199618919552152e-05, "loss": 0.0324, "step": 4135 }, { "epoch": 0.2842221000549753, "grad_norm": 1.1919034719467163, "learning_rate": 2.5197170709658363e-05, "loss": 0.0282, "step": 4136 }, { "epoch": 0.28429081913139087, "grad_norm": 1.6036736965179443, "learning_rate": 2.51947219946177e-05, "loss": 0.0424, "step": 4137 }, { "epoch": 0.28435953820780646, "grad_norm": 0.7942528128623962, "learning_rate": 2.519227277455147e-05, "loss": 0.0284, "step": 4138 }, { "epoch": 0.2844282572842221, "grad_norm": 1.3566969633102417, "learning_rate": 2.5189823049581e-05, "loss": 0.031, "step": 4139 }, { "epoch": 0.2844969763606377, "grad_norm": 1.0185226202011108, "learning_rate": 2.5187372819827647e-05, "loss": 0.0297, "step": 4140 }, { "epoch": 0.28456569543705335, "grad_norm": 0.7390490174293518, "learning_rate": 2.518492208541279e-05, "loss": 0.0275, "step": 4141 }, { "epoch": 0.28463441451346894, "grad_norm": 0.7402477860450745, "learning_rate": 2.5182470846457828e-05, "loss": 0.033, "step": 4142 }, { "epoch": 0.2847031335898845, "grad_norm": 2.785640239715576, "learning_rate": 2.51800191030842e-05, "loss": 0.0292, "step": 4143 }, { "epoch": 0.2847718526663002, "grad_norm": 0.6463701725006104, "learning_rate": 2.517756685541335e-05, "loss": 0.0252, "step": 4144 }, { "epoch": 0.28484057174271576, "grad_norm": 1.000035285949707, "learning_rate": 2.5175114103566766e-05, "loss": 0.0432, "step": 4145 }, { "epoch": 0.2849092908191314, "grad_norm": 1.0715278387069702, "learning_rate": 2.5172660847665943e-05, "loss": 0.0468, "step": 4146 }, { "epoch": 0.284978009895547, "grad_norm": 1.46059250831604, "learning_rate": 2.517020708783242e-05, "loss": 0.0406, "step": 4147 }, { "epoch": 0.2850467289719626, "grad_norm": 1.613158941268921, "learning_rate": 2.516775282418775e-05, "loss": 0.0266, "step": 4148 }, { "epoch": 0.28511544804837824, "grad_norm": 0.6334893107414246, "learning_rate": 2.51652980568535e-05, "loss": 0.0344, "step": 4149 }, { "epoch": 0.28518416712479383, "grad_norm": 1.0687010288238525, "learning_rate": 2.516284278595129e-05, "loss": 0.0343, "step": 4150 }, { "epoch": 0.2852528862012095, "grad_norm": 1.0244265794754028, "learning_rate": 2.5160387011602744e-05, "loss": 0.0253, "step": 4151 }, { "epoch": 0.28532160527762507, "grad_norm": 1.5038964748382568, "learning_rate": 2.515793073392951e-05, "loss": 0.0414, "step": 4152 }, { "epoch": 0.28539032435404066, "grad_norm": 1.1164042949676514, "learning_rate": 2.5155473953053272e-05, "loss": 0.0357, "step": 4153 }, { "epoch": 0.2854590434304563, "grad_norm": 0.7269356846809387, "learning_rate": 2.5153016669095735e-05, "loss": 0.0344, "step": 4154 }, { "epoch": 0.2855277625068719, "grad_norm": 0.9924625754356384, "learning_rate": 2.5150558882178626e-05, "loss": 0.0309, "step": 4155 }, { "epoch": 0.28559648158328754, "grad_norm": 0.7081835865974426, "learning_rate": 2.5148100592423694e-05, "loss": 0.0291, "step": 4156 }, { "epoch": 0.28566520065970313, "grad_norm": 1.141144871711731, "learning_rate": 2.5145641799952726e-05, "loss": 0.0458, "step": 4157 }, { "epoch": 0.2857339197361187, "grad_norm": 0.8809375166893005, "learning_rate": 2.5143182504887518e-05, "loss": 0.0282, "step": 4158 }, { "epoch": 0.28580263881253437, "grad_norm": 0.9185870885848999, "learning_rate": 2.5140722707349908e-05, "loss": 0.0385, "step": 4159 }, { "epoch": 0.28587135788894996, "grad_norm": 0.5377539396286011, "learning_rate": 2.5138262407461742e-05, "loss": 0.0418, "step": 4160 }, { "epoch": 0.2859400769653656, "grad_norm": 0.9478589296340942, "learning_rate": 2.51358016053449e-05, "loss": 0.0309, "step": 4161 }, { "epoch": 0.2860087960417812, "grad_norm": 0.8462845087051392, "learning_rate": 2.513334030112128e-05, "loss": 0.046, "step": 4162 }, { "epoch": 0.2860775151181968, "grad_norm": 1.017957329750061, "learning_rate": 2.513087849491282e-05, "loss": 0.0304, "step": 4163 }, { "epoch": 0.28614623419461244, "grad_norm": 1.4748847484588623, "learning_rate": 2.5128416186841467e-05, "loss": 0.0447, "step": 4164 }, { "epoch": 0.286214953271028, "grad_norm": 0.9697022438049316, "learning_rate": 2.5125953377029196e-05, "loss": 0.0418, "step": 4165 }, { "epoch": 0.2862836723474437, "grad_norm": 1.2251148223876953, "learning_rate": 2.5123490065598013e-05, "loss": 0.0382, "step": 4166 }, { "epoch": 0.28635239142385926, "grad_norm": 0.6852192282676697, "learning_rate": 2.512102625266995e-05, "loss": 0.0358, "step": 4167 }, { "epoch": 0.28642111050027486, "grad_norm": 1.1775156259536743, "learning_rate": 2.511856193836705e-05, "loss": 0.0381, "step": 4168 }, { "epoch": 0.2864898295766905, "grad_norm": 0.7679975032806396, "learning_rate": 2.51160971228114e-05, "loss": 0.0396, "step": 4169 }, { "epoch": 0.2865585486531061, "grad_norm": 1.0917201042175293, "learning_rate": 2.511363180612509e-05, "loss": 0.0317, "step": 4170 }, { "epoch": 0.28662726772952174, "grad_norm": 0.9730485081672668, "learning_rate": 2.5111165988430253e-05, "loss": 0.0241, "step": 4171 }, { "epoch": 0.28669598680593733, "grad_norm": 1.0000030994415283, "learning_rate": 2.5108699669849043e-05, "loss": 0.031, "step": 4172 }, { "epoch": 0.2867647058823529, "grad_norm": 0.7754005193710327, "learning_rate": 2.5106232850503632e-05, "loss": 0.0265, "step": 4173 }, { "epoch": 0.28683342495876857, "grad_norm": 1.26102876663208, "learning_rate": 2.510376553051623e-05, "loss": 0.0322, "step": 4174 }, { "epoch": 0.28690214403518416, "grad_norm": 0.6905608177185059, "learning_rate": 2.510129771000905e-05, "loss": 0.0493, "step": 4175 }, { "epoch": 0.2869708631115998, "grad_norm": 1.0485403537750244, "learning_rate": 2.509882938910435e-05, "loss": 0.0367, "step": 4176 }, { "epoch": 0.2870395821880154, "grad_norm": 1.101908802986145, "learning_rate": 2.5096360567924403e-05, "loss": 0.0406, "step": 4177 }, { "epoch": 0.287108301264431, "grad_norm": 1.0127311944961548, "learning_rate": 2.5093891246591514e-05, "loss": 0.0285, "step": 4178 }, { "epoch": 0.28717702034084663, "grad_norm": 1.136107325553894, "learning_rate": 2.5091421425228e-05, "loss": 0.0414, "step": 4179 }, { "epoch": 0.2872457394172622, "grad_norm": 0.8614048361778259, "learning_rate": 2.5088951103956217e-05, "loss": 0.0392, "step": 4180 }, { "epoch": 0.28731445849367787, "grad_norm": 0.7950728535652161, "learning_rate": 2.508648028289854e-05, "loss": 0.0432, "step": 4181 }, { "epoch": 0.28738317757009346, "grad_norm": 0.8484476208686829, "learning_rate": 2.508400896217737e-05, "loss": 0.033, "step": 4182 }, { "epoch": 0.28745189664650905, "grad_norm": 1.0956356525421143, "learning_rate": 2.5081537141915124e-05, "loss": 0.0415, "step": 4183 }, { "epoch": 0.2875206157229247, "grad_norm": 1.1833109855651855, "learning_rate": 2.5079064822234255e-05, "loss": 0.0306, "step": 4184 }, { "epoch": 0.2875893347993403, "grad_norm": 1.200393557548523, "learning_rate": 2.5076592003257233e-05, "loss": 0.0442, "step": 4185 }, { "epoch": 0.28765805387575594, "grad_norm": 1.3724993467330933, "learning_rate": 2.5074118685106567e-05, "loss": 0.0434, "step": 4186 }, { "epoch": 0.28772677295217153, "grad_norm": 1.3234421014785767, "learning_rate": 2.5071644867904767e-05, "loss": 0.0408, "step": 4187 }, { "epoch": 0.2877954920285871, "grad_norm": 1.4692814350128174, "learning_rate": 2.5069170551774387e-05, "loss": 0.0504, "step": 4188 }, { "epoch": 0.28786421110500277, "grad_norm": 1.2853485345840454, "learning_rate": 2.5066695736838e-05, "loss": 0.052, "step": 4189 }, { "epoch": 0.28793293018141836, "grad_norm": 1.3168995380401611, "learning_rate": 2.5064220423218207e-05, "loss": 0.0369, "step": 4190 }, { "epoch": 0.28800164925783395, "grad_norm": 2.065722942352295, "learning_rate": 2.506174461103762e-05, "loss": 0.0306, "step": 4191 }, { "epoch": 0.2880703683342496, "grad_norm": 0.8321943283081055, "learning_rate": 2.5059268300418895e-05, "loss": 0.0267, "step": 4192 }, { "epoch": 0.2881390874106652, "grad_norm": 0.48724353313446045, "learning_rate": 2.5056791491484697e-05, "loss": 0.0284, "step": 4193 }, { "epoch": 0.28820780648708083, "grad_norm": 1.3724126815795898, "learning_rate": 2.5054314184357724e-05, "loss": 0.0301, "step": 4194 }, { "epoch": 0.2882765255634964, "grad_norm": 0.5772931575775146, "learning_rate": 2.5051836379160697e-05, "loss": 0.029, "step": 4195 }, { "epoch": 0.288345244639912, "grad_norm": 0.6331813931465149, "learning_rate": 2.5049358076016365e-05, "loss": 0.0242, "step": 4196 }, { "epoch": 0.28841396371632766, "grad_norm": 0.7610636949539185, "learning_rate": 2.5046879275047487e-05, "loss": 0.0273, "step": 4197 }, { "epoch": 0.28848268279274325, "grad_norm": 1.0221295356750488, "learning_rate": 2.504439997637687e-05, "loss": 0.0385, "step": 4198 }, { "epoch": 0.2885514018691589, "grad_norm": 0.9951119422912598, "learning_rate": 2.504192018012733e-05, "loss": 0.0339, "step": 4199 }, { "epoch": 0.2886201209455745, "grad_norm": 1.1056987047195435, "learning_rate": 2.503943988642171e-05, "loss": 0.0388, "step": 4200 }, { "epoch": 0.2886888400219901, "grad_norm": 1.492545247077942, "learning_rate": 2.5036959095382875e-05, "loss": 0.0416, "step": 4201 }, { "epoch": 0.2887575590984057, "grad_norm": 0.7294974327087402, "learning_rate": 2.5034477807133724e-05, "loss": 0.0303, "step": 4202 }, { "epoch": 0.2888262781748213, "grad_norm": 1.0992279052734375, "learning_rate": 2.503199602179717e-05, "loss": 0.0368, "step": 4203 }, { "epoch": 0.28889499725123696, "grad_norm": 0.8886181712150574, "learning_rate": 2.502951373949616e-05, "loss": 0.0314, "step": 4204 }, { "epoch": 0.28896371632765255, "grad_norm": 2.0052952766418457, "learning_rate": 2.5027030960353655e-05, "loss": 0.0325, "step": 4205 }, { "epoch": 0.28903243540406814, "grad_norm": 1.1875895261764526, "learning_rate": 2.5024547684492653e-05, "loss": 0.0267, "step": 4206 }, { "epoch": 0.2891011544804838, "grad_norm": 0.9469020366668701, "learning_rate": 2.5022063912036174e-05, "loss": 0.0275, "step": 4207 }, { "epoch": 0.2891698735568994, "grad_norm": 1.1232671737670898, "learning_rate": 2.5019579643107247e-05, "loss": 0.0459, "step": 4208 }, { "epoch": 0.28923859263331503, "grad_norm": 0.4213894009590149, "learning_rate": 2.5017094877828946e-05, "loss": 0.0224, "step": 4209 }, { "epoch": 0.2893073117097306, "grad_norm": 1.051812767982483, "learning_rate": 2.5014609616324358e-05, "loss": 0.0384, "step": 4210 }, { "epoch": 0.2893760307861462, "grad_norm": 0.8707914352416992, "learning_rate": 2.50121238587166e-05, "loss": 0.025, "step": 4211 }, { "epoch": 0.28944474986256186, "grad_norm": 1.484114170074463, "learning_rate": 2.500963760512881e-05, "loss": 0.0459, "step": 4212 }, { "epoch": 0.28951346893897745, "grad_norm": 0.7471111416816711, "learning_rate": 2.5007150855684152e-05, "loss": 0.0344, "step": 4213 }, { "epoch": 0.2895821880153931, "grad_norm": 1.7168922424316406, "learning_rate": 2.5004663610505814e-05, "loss": 0.0421, "step": 4214 }, { "epoch": 0.2896509070918087, "grad_norm": 0.7934223413467407, "learning_rate": 2.500217586971701e-05, "loss": 0.0359, "step": 4215 }, { "epoch": 0.2897196261682243, "grad_norm": 1.2657620906829834, "learning_rate": 2.4999687633440977e-05, "loss": 0.0344, "step": 4216 }, { "epoch": 0.2897883452446399, "grad_norm": 0.9861454367637634, "learning_rate": 2.4997198901800976e-05, "loss": 0.033, "step": 4217 }, { "epoch": 0.2898570643210555, "grad_norm": 2.2182655334472656, "learning_rate": 2.4994709674920295e-05, "loss": 0.0474, "step": 4218 }, { "epoch": 0.28992578339747116, "grad_norm": 1.2109650373458862, "learning_rate": 2.499221995292225e-05, "loss": 0.0483, "step": 4219 }, { "epoch": 0.28999450247388675, "grad_norm": 1.3080830574035645, "learning_rate": 2.498972973593016e-05, "loss": 0.0331, "step": 4220 }, { "epoch": 0.29006322155030234, "grad_norm": 1.5259586572647095, "learning_rate": 2.4987239024067405e-05, "loss": 0.0344, "step": 4221 }, { "epoch": 0.290131940626718, "grad_norm": 0.8545582890510559, "learning_rate": 2.4984747817457366e-05, "loss": 0.0292, "step": 4222 }, { "epoch": 0.2902006597031336, "grad_norm": 0.6514385342597961, "learning_rate": 2.498225611622344e-05, "loss": 0.0409, "step": 4223 }, { "epoch": 0.2902693787795492, "grad_norm": 1.4042161703109741, "learning_rate": 2.497976392048907e-05, "loss": 0.051, "step": 4224 }, { "epoch": 0.2903380978559648, "grad_norm": 0.9979915618896484, "learning_rate": 2.4977271230377715e-05, "loss": 0.0422, "step": 4225 }, { "epoch": 0.2904068169323804, "grad_norm": 0.33675825595855713, "learning_rate": 2.4974778046012856e-05, "loss": 0.0398, "step": 4226 }, { "epoch": 0.29047553600879605, "grad_norm": 1.012134075164795, "learning_rate": 2.4972284367517998e-05, "loss": 0.0279, "step": 4227 }, { "epoch": 0.29054425508521164, "grad_norm": 1.046262502670288, "learning_rate": 2.4969790195016677e-05, "loss": 0.0345, "step": 4228 }, { "epoch": 0.2906129741616273, "grad_norm": 1.053539752960205, "learning_rate": 2.4967295528632446e-05, "loss": 0.0319, "step": 4229 }, { "epoch": 0.2906816932380429, "grad_norm": 0.8745731115341187, "learning_rate": 2.4964800368488882e-05, "loss": 0.0267, "step": 4230 }, { "epoch": 0.2907504123144585, "grad_norm": 0.8190634250640869, "learning_rate": 2.4962304714709597e-05, "loss": 0.0248, "step": 4231 }, { "epoch": 0.2908191313908741, "grad_norm": 0.8420481085777283, "learning_rate": 2.495980856741822e-05, "loss": 0.0318, "step": 4232 }, { "epoch": 0.2908878504672897, "grad_norm": 1.1399602890014648, "learning_rate": 2.49573119267384e-05, "loss": 0.0482, "step": 4233 }, { "epoch": 0.29095656954370536, "grad_norm": 1.035220980644226, "learning_rate": 2.495481479279382e-05, "loss": 0.0277, "step": 4234 }, { "epoch": 0.29102528862012095, "grad_norm": 1.4342918395996094, "learning_rate": 2.4952317165708184e-05, "loss": 0.0401, "step": 4235 }, { "epoch": 0.29109400769653654, "grad_norm": 0.8603370785713196, "learning_rate": 2.4949819045605214e-05, "loss": 0.0244, "step": 4236 }, { "epoch": 0.2911627267729522, "grad_norm": 0.8168255090713501, "learning_rate": 2.4947320432608663e-05, "loss": 0.0358, "step": 4237 }, { "epoch": 0.2912314458493678, "grad_norm": 0.7177402377128601, "learning_rate": 2.4944821326842316e-05, "loss": 0.0305, "step": 4238 }, { "epoch": 0.2913001649257834, "grad_norm": 1.0875072479248047, "learning_rate": 2.4942321728429955e-05, "loss": 0.0378, "step": 4239 }, { "epoch": 0.291368884002199, "grad_norm": 0.6419549584388733, "learning_rate": 2.4939821637495426e-05, "loss": 0.0392, "step": 4240 }, { "epoch": 0.2914376030786146, "grad_norm": 1.9359188079833984, "learning_rate": 2.4937321054162563e-05, "loss": 0.0333, "step": 4241 }, { "epoch": 0.29150632215503025, "grad_norm": 0.8468713760375977, "learning_rate": 2.493481997855525e-05, "loss": 0.0298, "step": 4242 }, { "epoch": 0.29157504123144584, "grad_norm": 0.9305140972137451, "learning_rate": 2.4932318410797374e-05, "loss": 0.0469, "step": 4243 }, { "epoch": 0.2916437603078615, "grad_norm": 0.7713059782981873, "learning_rate": 2.492981635101287e-05, "loss": 0.031, "step": 4244 }, { "epoch": 0.2917124793842771, "grad_norm": 1.8196579217910767, "learning_rate": 2.492731379932567e-05, "loss": 0.0528, "step": 4245 }, { "epoch": 0.29178119846069267, "grad_norm": 1.0923824310302734, "learning_rate": 2.492481075585976e-05, "loss": 0.0336, "step": 4246 }, { "epoch": 0.2918499175371083, "grad_norm": 1.0814937353134155, "learning_rate": 2.4922307220739137e-05, "loss": 0.0404, "step": 4247 }, { "epoch": 0.2919186366135239, "grad_norm": 0.8567585945129395, "learning_rate": 2.49198031940878e-05, "loss": 0.0343, "step": 4248 }, { "epoch": 0.29198735568993955, "grad_norm": 1.0912585258483887, "learning_rate": 2.4917298676029817e-05, "loss": 0.0542, "step": 4249 }, { "epoch": 0.29205607476635514, "grad_norm": 1.0169003009796143, "learning_rate": 2.4914793666689243e-05, "loss": 0.0339, "step": 4250 }, { "epoch": 0.29212479384277074, "grad_norm": 0.7181757688522339, "learning_rate": 2.491228816619017e-05, "loss": 0.0198, "step": 4251 }, { "epoch": 0.2921935129191864, "grad_norm": 0.47251421213150024, "learning_rate": 2.4909782174656724e-05, "loss": 0.0281, "step": 4252 }, { "epoch": 0.292262231995602, "grad_norm": 1.385504961013794, "learning_rate": 2.4907275692213047e-05, "loss": 0.0374, "step": 4253 }, { "epoch": 0.2923309510720176, "grad_norm": 0.8454645872116089, "learning_rate": 2.49047687189833e-05, "loss": 0.0249, "step": 4254 }, { "epoch": 0.2923996701484332, "grad_norm": 1.4349215030670166, "learning_rate": 2.490226125509167e-05, "loss": 0.0263, "step": 4255 }, { "epoch": 0.2924683892248488, "grad_norm": 1.3506442308425903, "learning_rate": 2.4899753300662378e-05, "loss": 0.0415, "step": 4256 }, { "epoch": 0.29253710830126445, "grad_norm": 1.5904946327209473, "learning_rate": 2.4897244855819657e-05, "loss": 0.0437, "step": 4257 }, { "epoch": 0.29260582737768004, "grad_norm": 0.8932166695594788, "learning_rate": 2.489473592068777e-05, "loss": 0.0485, "step": 4258 }, { "epoch": 0.29267454645409563, "grad_norm": 1.52115797996521, "learning_rate": 2.489222649539102e-05, "loss": 0.0286, "step": 4259 }, { "epoch": 0.2927432655305113, "grad_norm": 0.7795671224594116, "learning_rate": 2.48897165800537e-05, "loss": 0.0378, "step": 4260 }, { "epoch": 0.29281198460692687, "grad_norm": 0.3994980454444885, "learning_rate": 2.4887206174800154e-05, "loss": 0.0259, "step": 4261 }, { "epoch": 0.2928807036833425, "grad_norm": 0.7286141514778137, "learning_rate": 2.488469527975474e-05, "loss": 0.0298, "step": 4262 }, { "epoch": 0.2929494227597581, "grad_norm": 1.1237424612045288, "learning_rate": 2.4882183895041847e-05, "loss": 0.0362, "step": 4263 }, { "epoch": 0.2930181418361737, "grad_norm": 0.6358994245529175, "learning_rate": 2.487967202078588e-05, "loss": 0.0307, "step": 4264 }, { "epoch": 0.29308686091258934, "grad_norm": 1.0868165493011475, "learning_rate": 2.487715965711127e-05, "loss": 0.0329, "step": 4265 }, { "epoch": 0.29315557998900493, "grad_norm": 0.8955458998680115, "learning_rate": 2.487464680414248e-05, "loss": 0.0333, "step": 4266 }, { "epoch": 0.2932242990654206, "grad_norm": 0.9607861042022705, "learning_rate": 2.4872133462003986e-05, "loss": 0.0317, "step": 4267 }, { "epoch": 0.29329301814183617, "grad_norm": 0.6153407096862793, "learning_rate": 2.4869619630820298e-05, "loss": 0.0257, "step": 4268 }, { "epoch": 0.29336173721825176, "grad_norm": 1.0264294147491455, "learning_rate": 2.486710531071595e-05, "loss": 0.0441, "step": 4269 }, { "epoch": 0.2934304562946674, "grad_norm": 1.8128745555877686, "learning_rate": 2.4864590501815486e-05, "loss": 0.0431, "step": 4270 }, { "epoch": 0.293499175371083, "grad_norm": 0.8814264535903931, "learning_rate": 2.4862075204243488e-05, "loss": 0.0244, "step": 4271 }, { "epoch": 0.29356789444749865, "grad_norm": 1.0919057130813599, "learning_rate": 2.4859559418124556e-05, "loss": 0.0413, "step": 4272 }, { "epoch": 0.29363661352391424, "grad_norm": 0.8715153336524963, "learning_rate": 2.485704314358333e-05, "loss": 0.0331, "step": 4273 }, { "epoch": 0.2937053326003298, "grad_norm": 0.9607916474342346, "learning_rate": 2.4854526380744446e-05, "loss": 0.0277, "step": 4274 }, { "epoch": 0.2937740516767455, "grad_norm": 0.6526276469230652, "learning_rate": 2.485200912973259e-05, "loss": 0.0285, "step": 4275 }, { "epoch": 0.29384277075316106, "grad_norm": 1.7733561992645264, "learning_rate": 2.484949139067245e-05, "loss": 0.0488, "step": 4276 }, { "epoch": 0.2939114898295767, "grad_norm": 3.0440592765808105, "learning_rate": 2.4846973163688762e-05, "loss": 0.0279, "step": 4277 }, { "epoch": 0.2939802089059923, "grad_norm": 0.8131105899810791, "learning_rate": 2.4844454448906262e-05, "loss": 0.0283, "step": 4278 }, { "epoch": 0.2940489279824079, "grad_norm": 1.286487102508545, "learning_rate": 2.4841935246449734e-05, "loss": 0.0389, "step": 4279 }, { "epoch": 0.29411764705882354, "grad_norm": 0.8762199282646179, "learning_rate": 2.4839415556443965e-05, "loss": 0.0303, "step": 4280 }, { "epoch": 0.29418636613523913, "grad_norm": 1.0956112146377563, "learning_rate": 2.483689537901378e-05, "loss": 0.0388, "step": 4281 }, { "epoch": 0.2942550852116548, "grad_norm": 0.4742363393306732, "learning_rate": 2.4834374714284022e-05, "loss": 0.0348, "step": 4282 }, { "epoch": 0.29432380428807037, "grad_norm": 2.0234222412109375, "learning_rate": 2.4831853562379556e-05, "loss": 0.0376, "step": 4283 }, { "epoch": 0.29439252336448596, "grad_norm": 0.8812259435653687, "learning_rate": 2.4829331923425278e-05, "loss": 0.0408, "step": 4284 }, { "epoch": 0.2944612424409016, "grad_norm": 1.8441576957702637, "learning_rate": 2.482680979754611e-05, "loss": 0.0448, "step": 4285 }, { "epoch": 0.2945299615173172, "grad_norm": 1.5058075189590454, "learning_rate": 2.4824287184866984e-05, "loss": 0.0298, "step": 4286 }, { "epoch": 0.29459868059373284, "grad_norm": 1.2351893186569214, "learning_rate": 2.4821764085512867e-05, "loss": 0.0493, "step": 4287 }, { "epoch": 0.29466739967014843, "grad_norm": 1.7195539474487305, "learning_rate": 2.4819240499608754e-05, "loss": 0.0349, "step": 4288 }, { "epoch": 0.294736118746564, "grad_norm": 1.2840595245361328, "learning_rate": 2.4816716427279654e-05, "loss": 0.0365, "step": 4289 }, { "epoch": 0.29480483782297967, "grad_norm": 0.8250982165336609, "learning_rate": 2.48141918686506e-05, "loss": 0.0425, "step": 4290 }, { "epoch": 0.29487355689939526, "grad_norm": 0.947879433631897, "learning_rate": 2.481166682384666e-05, "loss": 0.0295, "step": 4291 }, { "epoch": 0.2949422759758109, "grad_norm": 0.9888718724250793, "learning_rate": 2.480914129299292e-05, "loss": 0.0256, "step": 4292 }, { "epoch": 0.2950109950522265, "grad_norm": 1.6396557092666626, "learning_rate": 2.4806615276214485e-05, "loss": 0.043, "step": 4293 }, { "epoch": 0.2950797141286421, "grad_norm": 1.302931785583496, "learning_rate": 2.4804088773636488e-05, "loss": 0.0378, "step": 4294 }, { "epoch": 0.29514843320505774, "grad_norm": 1.59483003616333, "learning_rate": 2.4801561785384097e-05, "loss": 0.0408, "step": 4295 }, { "epoch": 0.2952171522814733, "grad_norm": 0.7844313383102417, "learning_rate": 2.479903431158248e-05, "loss": 0.0287, "step": 4296 }, { "epoch": 0.295285871357889, "grad_norm": 1.1353585720062256, "learning_rate": 2.479650635235685e-05, "loss": 0.0497, "step": 4297 }, { "epoch": 0.29535459043430456, "grad_norm": 0.576242208480835, "learning_rate": 2.4793977907832437e-05, "loss": 0.0316, "step": 4298 }, { "epoch": 0.29542330951072016, "grad_norm": 0.955098032951355, "learning_rate": 2.4791448978134494e-05, "loss": 0.046, "step": 4299 }, { "epoch": 0.2954920285871358, "grad_norm": 1.5871199369430542, "learning_rate": 2.4788919563388296e-05, "loss": 0.037, "step": 4300 }, { "epoch": 0.2955607476635514, "grad_norm": 0.49684837460517883, "learning_rate": 2.4786389663719155e-05, "loss": 0.0261, "step": 4301 }, { "epoch": 0.29562946673996704, "grad_norm": 1.3235485553741455, "learning_rate": 2.4783859279252383e-05, "loss": 0.0412, "step": 4302 }, { "epoch": 0.29569818581638263, "grad_norm": 0.9046555161476135, "learning_rate": 2.4781328410113342e-05, "loss": 0.0334, "step": 4303 }, { "epoch": 0.2957669048927982, "grad_norm": 0.9650842547416687, "learning_rate": 2.4778797056427403e-05, "loss": 0.0266, "step": 4304 }, { "epoch": 0.29583562396921387, "grad_norm": 1.7667378187179565, "learning_rate": 2.4776265218319957e-05, "loss": 0.0423, "step": 4305 }, { "epoch": 0.29590434304562946, "grad_norm": 0.6601541638374329, "learning_rate": 2.4773732895916432e-05, "loss": 0.0238, "step": 4306 }, { "epoch": 0.2959730621220451, "grad_norm": 0.8498976826667786, "learning_rate": 2.477120008934227e-05, "loss": 0.0432, "step": 4307 }, { "epoch": 0.2960417811984607, "grad_norm": 1.05923330783844, "learning_rate": 2.4768666798722947e-05, "loss": 0.0448, "step": 4308 }, { "epoch": 0.2961105002748763, "grad_norm": 0.9019392132759094, "learning_rate": 2.476613302418396e-05, "loss": 0.0466, "step": 4309 }, { "epoch": 0.29617921935129193, "grad_norm": 1.201456069946289, "learning_rate": 2.4763598765850815e-05, "loss": 0.0389, "step": 4310 }, { "epoch": 0.2962479384277075, "grad_norm": 1.3970558643341064, "learning_rate": 2.476106402384906e-05, "loss": 0.0415, "step": 4311 }, { "epoch": 0.29631665750412317, "grad_norm": 0.777199923992157, "learning_rate": 2.4758528798304267e-05, "loss": 0.0327, "step": 4312 }, { "epoch": 0.29638537658053876, "grad_norm": 1.0598516464233398, "learning_rate": 2.4755993089342015e-05, "loss": 0.0418, "step": 4313 }, { "epoch": 0.29645409565695435, "grad_norm": 1.1354660987854004, "learning_rate": 2.475345689708792e-05, "loss": 0.029, "step": 4314 }, { "epoch": 0.29652281473337, "grad_norm": 1.0234839916229248, "learning_rate": 2.4750920221667633e-05, "loss": 0.0523, "step": 4315 }, { "epoch": 0.2965915338097856, "grad_norm": 1.2281343936920166, "learning_rate": 2.4748383063206796e-05, "loss": 0.0504, "step": 4316 }, { "epoch": 0.29666025288620124, "grad_norm": 1.0704659223556519, "learning_rate": 2.4745845421831108e-05, "loss": 0.046, "step": 4317 }, { "epoch": 0.2967289719626168, "grad_norm": 0.7057315111160278, "learning_rate": 2.4743307297666273e-05, "loss": 0.0293, "step": 4318 }, { "epoch": 0.2967976910390324, "grad_norm": 0.9832585453987122, "learning_rate": 2.4740768690838027e-05, "loss": 0.0414, "step": 4319 }, { "epoch": 0.29686641011544807, "grad_norm": 0.9386084079742432, "learning_rate": 2.473822960147212e-05, "loss": 0.0346, "step": 4320 }, { "epoch": 0.29693512919186366, "grad_norm": 1.0149174928665161, "learning_rate": 2.473569002969435e-05, "loss": 0.0373, "step": 4321 }, { "epoch": 0.29700384826827925, "grad_norm": 0.5550531148910522, "learning_rate": 2.473314997563051e-05, "loss": 0.0298, "step": 4322 }, { "epoch": 0.2970725673446949, "grad_norm": 0.74235999584198, "learning_rate": 2.4730609439406423e-05, "loss": 0.0379, "step": 4323 }, { "epoch": 0.2971412864211105, "grad_norm": 0.9820315837860107, "learning_rate": 2.4728068421147956e-05, "loss": 0.0253, "step": 4324 }, { "epoch": 0.29721000549752613, "grad_norm": 0.8142573833465576, "learning_rate": 2.4725526920980977e-05, "loss": 0.0328, "step": 4325 }, { "epoch": 0.2972787245739417, "grad_norm": 1.0791925191879272, "learning_rate": 2.4722984939031388e-05, "loss": 0.0395, "step": 4326 }, { "epoch": 0.2973474436503573, "grad_norm": 1.425566554069519, "learning_rate": 2.4720442475425115e-05, "loss": 0.0352, "step": 4327 }, { "epoch": 0.29741616272677296, "grad_norm": 1.0113478899002075, "learning_rate": 2.4717899530288107e-05, "loss": 0.0319, "step": 4328 }, { "epoch": 0.29748488180318855, "grad_norm": 1.6148637533187866, "learning_rate": 2.4715356103746337e-05, "loss": 0.0497, "step": 4329 }, { "epoch": 0.2975536008796042, "grad_norm": 2.1670143604278564, "learning_rate": 2.47128121959258e-05, "loss": 0.0448, "step": 4330 }, { "epoch": 0.2976223199560198, "grad_norm": 1.1133557558059692, "learning_rate": 2.471026780695251e-05, "loss": 0.0246, "step": 4331 }, { "epoch": 0.2976910390324354, "grad_norm": 1.288374662399292, "learning_rate": 2.470772293695252e-05, "loss": 0.0412, "step": 4332 }, { "epoch": 0.297759758108851, "grad_norm": 0.9604666829109192, "learning_rate": 2.4705177586051893e-05, "loss": 0.033, "step": 4333 }, { "epoch": 0.2978284771852666, "grad_norm": 0.8824843168258667, "learning_rate": 2.4702631754376726e-05, "loss": 0.0375, "step": 4334 }, { "epoch": 0.29789719626168226, "grad_norm": 1.2598743438720703, "learning_rate": 2.4700085442053125e-05, "loss": 0.0361, "step": 4335 }, { "epoch": 0.29796591533809785, "grad_norm": 1.2806298732757568, "learning_rate": 2.4697538649207236e-05, "loss": 0.0313, "step": 4336 }, { "epoch": 0.29803463441451344, "grad_norm": 1.0134998559951782, "learning_rate": 2.4694991375965218e-05, "loss": 0.0234, "step": 4337 }, { "epoch": 0.2981033534909291, "grad_norm": 0.982652485370636, "learning_rate": 2.469244362245326e-05, "loss": 0.0365, "step": 4338 }, { "epoch": 0.2981720725673447, "grad_norm": 0.9162691831588745, "learning_rate": 2.468989538879757e-05, "loss": 0.034, "step": 4339 }, { "epoch": 0.29824079164376033, "grad_norm": 0.8290132284164429, "learning_rate": 2.4687346675124388e-05, "loss": 0.0371, "step": 4340 }, { "epoch": 0.2983095107201759, "grad_norm": 1.122473120689392, "learning_rate": 2.468479748155996e-05, "loss": 0.0544, "step": 4341 }, { "epoch": 0.2983782297965915, "grad_norm": 1.0453894138336182, "learning_rate": 2.4682247808230586e-05, "loss": 0.0363, "step": 4342 }, { "epoch": 0.29844694887300716, "grad_norm": 1.374717116355896, "learning_rate": 2.467969765526256e-05, "loss": 0.027, "step": 4343 }, { "epoch": 0.29851566794942275, "grad_norm": 0.816713809967041, "learning_rate": 2.4677147022782205e-05, "loss": 0.0358, "step": 4344 }, { "epoch": 0.2985843870258384, "grad_norm": 1.3339853286743164, "learning_rate": 2.4674595910915886e-05, "loss": 0.0372, "step": 4345 }, { "epoch": 0.298653106102254, "grad_norm": 1.2025700807571411, "learning_rate": 2.4672044319789978e-05, "loss": 0.0327, "step": 4346 }, { "epoch": 0.2987218251786696, "grad_norm": 1.1982464790344238, "learning_rate": 2.4669492249530878e-05, "loss": 0.0324, "step": 4347 }, { "epoch": 0.2987905442550852, "grad_norm": 0.824195384979248, "learning_rate": 2.466693970026501e-05, "loss": 0.0379, "step": 4348 }, { "epoch": 0.2988592633315008, "grad_norm": 0.8427038788795471, "learning_rate": 2.4664386672118826e-05, "loss": 0.0392, "step": 4349 }, { "epoch": 0.29892798240791646, "grad_norm": 0.9993206858634949, "learning_rate": 2.4661833165218793e-05, "loss": 0.0297, "step": 4350 }, { "epoch": 0.29899670148433205, "grad_norm": 0.7292399406433105, "learning_rate": 2.465927917969141e-05, "loss": 0.0271, "step": 4351 }, { "epoch": 0.29906542056074764, "grad_norm": 0.902360200881958, "learning_rate": 2.46567247156632e-05, "loss": 0.0437, "step": 4352 }, { "epoch": 0.2991341396371633, "grad_norm": 1.0440995693206787, "learning_rate": 2.4654169773260694e-05, "loss": 0.0288, "step": 4353 }, { "epoch": 0.2992028587135789, "grad_norm": 1.2320656776428223, "learning_rate": 2.4651614352610466e-05, "loss": 0.0329, "step": 4354 }, { "epoch": 0.2992715777899945, "grad_norm": 1.9093981981277466, "learning_rate": 2.4649058453839107e-05, "loss": 0.0293, "step": 4355 }, { "epoch": 0.2993402968664101, "grad_norm": 1.1445673704147339, "learning_rate": 2.464650207707323e-05, "loss": 0.0282, "step": 4356 }, { "epoch": 0.2994090159428257, "grad_norm": 0.8115578889846802, "learning_rate": 2.4643945222439474e-05, "loss": 0.0503, "step": 4357 }, { "epoch": 0.29947773501924135, "grad_norm": 0.8704538941383362, "learning_rate": 2.46413878900645e-05, "loss": 0.0351, "step": 4358 }, { "epoch": 0.29954645409565694, "grad_norm": 1.0331752300262451, "learning_rate": 2.463883008007499e-05, "loss": 0.0521, "step": 4359 }, { "epoch": 0.2996151731720726, "grad_norm": 0.664036750793457, "learning_rate": 2.463627179259765e-05, "loss": 0.0307, "step": 4360 }, { "epoch": 0.2996838922484882, "grad_norm": 1.4989657402038574, "learning_rate": 2.463371302775922e-05, "loss": 0.0323, "step": 4361 }, { "epoch": 0.2997526113249038, "grad_norm": 1.3395041227340698, "learning_rate": 2.4631153785686454e-05, "loss": 0.0455, "step": 4362 }, { "epoch": 0.2998213304013194, "grad_norm": 0.9207924604415894, "learning_rate": 2.4628594066506127e-05, "loss": 0.0341, "step": 4363 }, { "epoch": 0.299890049477735, "grad_norm": 2.326653480529785, "learning_rate": 2.4626033870345047e-05, "loss": 0.0414, "step": 4364 }, { "epoch": 0.29995876855415066, "grad_norm": 1.1074773073196411, "learning_rate": 2.4623473197330045e-05, "loss": 0.0434, "step": 4365 }, { "epoch": 0.30002748763056625, "grad_norm": 0.8928117752075195, "learning_rate": 2.462091204758796e-05, "loss": 0.0354, "step": 4366 }, { "epoch": 0.30009620670698184, "grad_norm": 0.9314262270927429, "learning_rate": 2.4618350421245674e-05, "loss": 0.0296, "step": 4367 }, { "epoch": 0.3001649257833975, "grad_norm": 0.9825258255004883, "learning_rate": 2.4615788318430083e-05, "loss": 0.0312, "step": 4368 }, { "epoch": 0.3002336448598131, "grad_norm": 1.230583906173706, "learning_rate": 2.4613225739268105e-05, "loss": 0.0317, "step": 4369 }, { "epoch": 0.3003023639362287, "grad_norm": 1.0731925964355469, "learning_rate": 2.4610662683886688e-05, "loss": 0.0276, "step": 4370 }, { "epoch": 0.3003710830126443, "grad_norm": 0.7273163795471191, "learning_rate": 2.46080991524128e-05, "loss": 0.0275, "step": 4371 }, { "epoch": 0.3004398020890599, "grad_norm": 0.7570067644119263, "learning_rate": 2.4605535144973435e-05, "loss": 0.0285, "step": 4372 }, { "epoch": 0.30050852116547555, "grad_norm": 1.1220155954360962, "learning_rate": 2.460297066169561e-05, "loss": 0.0387, "step": 4373 }, { "epoch": 0.30057724024189114, "grad_norm": 1.105817437171936, "learning_rate": 2.460040570270636e-05, "loss": 0.0381, "step": 4374 }, { "epoch": 0.3006459593183068, "grad_norm": 0.8577398061752319, "learning_rate": 2.4597840268132748e-05, "loss": 0.0424, "step": 4375 }, { "epoch": 0.3007146783947224, "grad_norm": 0.8232551217079163, "learning_rate": 2.459527435810186e-05, "loss": 0.0398, "step": 4376 }, { "epoch": 0.30078339747113797, "grad_norm": 1.8707733154296875, "learning_rate": 2.459270797274081e-05, "loss": 0.0522, "step": 4377 }, { "epoch": 0.3008521165475536, "grad_norm": 1.3339722156524658, "learning_rate": 2.4590141112176723e-05, "loss": 0.041, "step": 4378 }, { "epoch": 0.3009208356239692, "grad_norm": 0.7614541053771973, "learning_rate": 2.4587573776536768e-05, "loss": 0.0553, "step": 4379 }, { "epoch": 0.30098955470038485, "grad_norm": 0.8194431066513062, "learning_rate": 2.458500596594812e-05, "loss": 0.0447, "step": 4380 }, { "epoch": 0.30105827377680044, "grad_norm": 1.2805417776107788, "learning_rate": 2.4582437680537978e-05, "loss": 0.0447, "step": 4381 }, { "epoch": 0.30112699285321604, "grad_norm": 0.8757633566856384, "learning_rate": 2.4579868920433574e-05, "loss": 0.0354, "step": 4382 }, { "epoch": 0.3011957119296317, "grad_norm": 1.1302672624588013, "learning_rate": 2.457729968576216e-05, "loss": 0.0339, "step": 4383 }, { "epoch": 0.3012644310060473, "grad_norm": 1.3882099390029907, "learning_rate": 2.457472997665101e-05, "loss": 0.0312, "step": 4384 }, { "epoch": 0.3013331500824629, "grad_norm": 1.2668566703796387, "learning_rate": 2.4572159793227422e-05, "loss": 0.03, "step": 4385 }, { "epoch": 0.3014018691588785, "grad_norm": 2.104750633239746, "learning_rate": 2.456958913561872e-05, "loss": 0.0399, "step": 4386 }, { "epoch": 0.3014705882352941, "grad_norm": 1.0440819263458252, "learning_rate": 2.4567018003952242e-05, "loss": 0.038, "step": 4387 }, { "epoch": 0.30153930731170975, "grad_norm": 1.3073804378509521, "learning_rate": 2.4564446398355364e-05, "loss": 0.0474, "step": 4388 }, { "epoch": 0.30160802638812534, "grad_norm": 1.0557893514633179, "learning_rate": 2.4561874318955476e-05, "loss": 0.0345, "step": 4389 }, { "epoch": 0.30167674546454093, "grad_norm": 1.4704259634017944, "learning_rate": 2.4559301765879988e-05, "loss": 0.0408, "step": 4390 }, { "epoch": 0.3017454645409566, "grad_norm": 1.3505572080612183, "learning_rate": 2.4556728739256348e-05, "loss": 0.0336, "step": 4391 }, { "epoch": 0.30181418361737217, "grad_norm": 1.3208385705947876, "learning_rate": 2.4554155239212014e-05, "loss": 0.0362, "step": 4392 }, { "epoch": 0.3018829026937878, "grad_norm": 1.147599458694458, "learning_rate": 2.455158126587447e-05, "loss": 0.0355, "step": 4393 }, { "epoch": 0.3019516217702034, "grad_norm": 1.1575018167495728, "learning_rate": 2.4549006819371224e-05, "loss": 0.0337, "step": 4394 }, { "epoch": 0.302020340846619, "grad_norm": 1.0371346473693848, "learning_rate": 2.4546431899829816e-05, "loss": 0.0337, "step": 4395 }, { "epoch": 0.30208905992303464, "grad_norm": 1.1395050287246704, "learning_rate": 2.4543856507377798e-05, "loss": 0.0358, "step": 4396 }, { "epoch": 0.30215777899945023, "grad_norm": 1.0014152526855469, "learning_rate": 2.4541280642142755e-05, "loss": 0.0336, "step": 4397 }, { "epoch": 0.3022264980758659, "grad_norm": 2.189958333969116, "learning_rate": 2.4538704304252277e-05, "loss": 0.0328, "step": 4398 }, { "epoch": 0.30229521715228147, "grad_norm": 0.9930230975151062, "learning_rate": 2.4536127493834e-05, "loss": 0.0299, "step": 4399 }, { "epoch": 0.30236393622869706, "grad_norm": 1.0303499698638916, "learning_rate": 2.4533550211015574e-05, "loss": 0.0281, "step": 4400 }, { "epoch": 0.3024326553051127, "grad_norm": 0.8656858801841736, "learning_rate": 2.4530972455924672e-05, "loss": 0.0318, "step": 4401 }, { "epoch": 0.3025013743815283, "grad_norm": 0.8057610988616943, "learning_rate": 2.4528394228688986e-05, "loss": 0.0313, "step": 4402 }, { "epoch": 0.30257009345794394, "grad_norm": 0.7082265615463257, "learning_rate": 2.4525815529436242e-05, "loss": 0.0257, "step": 4403 }, { "epoch": 0.30263881253435954, "grad_norm": 0.8740941882133484, "learning_rate": 2.4523236358294182e-05, "loss": 0.0339, "step": 4404 }, { "epoch": 0.3027075316107751, "grad_norm": 1.1025502681732178, "learning_rate": 2.452065671539057e-05, "loss": 0.0382, "step": 4405 }, { "epoch": 0.3027762506871908, "grad_norm": 0.9572781324386597, "learning_rate": 2.4518076600853198e-05, "loss": 0.0427, "step": 4406 }, { "epoch": 0.30284496976360636, "grad_norm": 1.217785358428955, "learning_rate": 2.451549601480988e-05, "loss": 0.0429, "step": 4407 }, { "epoch": 0.302913688840022, "grad_norm": 2.8801450729370117, "learning_rate": 2.4512914957388454e-05, "loss": 0.0339, "step": 4408 }, { "epoch": 0.3029824079164376, "grad_norm": 1.0584467649459839, "learning_rate": 2.4510333428716777e-05, "loss": 0.0424, "step": 4409 }, { "epoch": 0.3030511269928532, "grad_norm": 1.1143301725387573, "learning_rate": 2.450775142892273e-05, "loss": 0.0325, "step": 4410 }, { "epoch": 0.30311984606926884, "grad_norm": 1.2424463033676147, "learning_rate": 2.450516895813423e-05, "loss": 0.0301, "step": 4411 }, { "epoch": 0.30318856514568443, "grad_norm": 1.110543131828308, "learning_rate": 2.45025860164792e-05, "loss": 0.0328, "step": 4412 }, { "epoch": 0.3032572842221001, "grad_norm": 0.9442777633666992, "learning_rate": 2.4500002604085593e-05, "loss": 0.0372, "step": 4413 }, { "epoch": 0.30332600329851567, "grad_norm": 0.9003228545188904, "learning_rate": 2.4497418721081394e-05, "loss": 0.0388, "step": 4414 }, { "epoch": 0.30339472237493126, "grad_norm": 2.1309597492218018, "learning_rate": 2.4494834367594594e-05, "loss": 0.0448, "step": 4415 }, { "epoch": 0.3034634414513469, "grad_norm": 1.6176974773406982, "learning_rate": 2.4492249543753224e-05, "loss": 0.0311, "step": 4416 }, { "epoch": 0.3035321605277625, "grad_norm": 2.265378475189209, "learning_rate": 2.4489664249685318e-05, "loss": 0.04, "step": 4417 }, { "epoch": 0.30360087960417814, "grad_norm": 2.147263288497925, "learning_rate": 2.4487078485518958e-05, "loss": 0.0309, "step": 4418 }, { "epoch": 0.30366959868059373, "grad_norm": 1.307786464691162, "learning_rate": 2.448449225138224e-05, "loss": 0.0349, "step": 4419 }, { "epoch": 0.3037383177570093, "grad_norm": 0.6944969892501831, "learning_rate": 2.448190554740327e-05, "loss": 0.0292, "step": 4420 }, { "epoch": 0.30380703683342497, "grad_norm": 0.9597850441932678, "learning_rate": 2.4479318373710195e-05, "loss": 0.0305, "step": 4421 }, { "epoch": 0.30387575590984056, "grad_norm": 1.4111647605895996, "learning_rate": 2.4476730730431175e-05, "loss": 0.0474, "step": 4422 }, { "epoch": 0.3039444749862562, "grad_norm": 0.9325461983680725, "learning_rate": 2.4474142617694398e-05, "loss": 0.0365, "step": 4423 }, { "epoch": 0.3040131940626718, "grad_norm": 0.9283370971679688, "learning_rate": 2.4471554035628075e-05, "loss": 0.038, "step": 4424 }, { "epoch": 0.3040819131390874, "grad_norm": 0.8399947285652161, "learning_rate": 2.4468964984360438e-05, "loss": 0.0371, "step": 4425 }, { "epoch": 0.30415063221550304, "grad_norm": 0.9530466198921204, "learning_rate": 2.446637546401974e-05, "loss": 0.0365, "step": 4426 }, { "epoch": 0.3042193512919186, "grad_norm": 0.7936529517173767, "learning_rate": 2.446378547473427e-05, "loss": 0.0319, "step": 4427 }, { "epoch": 0.3042880703683343, "grad_norm": 0.8973393440246582, "learning_rate": 2.446119501663232e-05, "loss": 0.0366, "step": 4428 }, { "epoch": 0.30435678944474986, "grad_norm": 0.7015277147293091, "learning_rate": 2.445860408984222e-05, "loss": 0.0289, "step": 4429 }, { "epoch": 0.30442550852116546, "grad_norm": 0.46269553899765015, "learning_rate": 2.445601269449232e-05, "loss": 0.0257, "step": 4430 }, { "epoch": 0.3044942275975811, "grad_norm": 1.3398191928863525, "learning_rate": 2.4453420830710992e-05, "loss": 0.023, "step": 4431 }, { "epoch": 0.3045629466739967, "grad_norm": 1.3988704681396484, "learning_rate": 2.445082849862663e-05, "loss": 0.0264, "step": 4432 }, { "epoch": 0.30463166575041234, "grad_norm": 0.8901848793029785, "learning_rate": 2.4448235698367657e-05, "loss": 0.0386, "step": 4433 }, { "epoch": 0.30470038482682793, "grad_norm": 1.4964765310287476, "learning_rate": 2.4445642430062514e-05, "loss": 0.0429, "step": 4434 }, { "epoch": 0.3047691039032435, "grad_norm": 0.9503543376922607, "learning_rate": 2.4443048693839657e-05, "loss": 0.036, "step": 4435 }, { "epoch": 0.30483782297965917, "grad_norm": 0.8490949869155884, "learning_rate": 2.4440454489827585e-05, "loss": 0.0369, "step": 4436 }, { "epoch": 0.30490654205607476, "grad_norm": 0.6935877799987793, "learning_rate": 2.443785981815481e-05, "loss": 0.036, "step": 4437 }, { "epoch": 0.3049752611324904, "grad_norm": 1.7827162742614746, "learning_rate": 2.4435264678949858e-05, "loss": 0.0586, "step": 4438 }, { "epoch": 0.305043980208906, "grad_norm": 1.2629655599594116, "learning_rate": 2.4432669072341295e-05, "loss": 0.0456, "step": 4439 }, { "epoch": 0.3051126992853216, "grad_norm": 0.8567002415657043, "learning_rate": 2.44300729984577e-05, "loss": 0.0296, "step": 4440 }, { "epoch": 0.30518141836173723, "grad_norm": 0.8717685341835022, "learning_rate": 2.4427476457427672e-05, "loss": 0.0353, "step": 4441 }, { "epoch": 0.3052501374381528, "grad_norm": 0.8770506978034973, "learning_rate": 2.4424879449379845e-05, "loss": 0.0445, "step": 4442 }, { "epoch": 0.30531885651456847, "grad_norm": 0.7236136794090271, "learning_rate": 2.442228197444287e-05, "loss": 0.0389, "step": 4443 }, { "epoch": 0.30538757559098406, "grad_norm": 0.6899374127388, "learning_rate": 2.441968403274541e-05, "loss": 0.0429, "step": 4444 }, { "epoch": 0.30545629466739965, "grad_norm": 0.9937990307807922, "learning_rate": 2.4417085624416174e-05, "loss": 0.0377, "step": 4445 }, { "epoch": 0.3055250137438153, "grad_norm": 0.6740761399269104, "learning_rate": 2.4414486749583877e-05, "loss": 0.0351, "step": 4446 }, { "epoch": 0.3055937328202309, "grad_norm": 0.8598818182945251, "learning_rate": 2.4411887408377258e-05, "loss": 0.0318, "step": 4447 }, { "epoch": 0.30566245189664654, "grad_norm": 0.7739711403846741, "learning_rate": 2.4409287600925093e-05, "loss": 0.0304, "step": 4448 }, { "epoch": 0.3057311709730621, "grad_norm": 1.7168585062026978, "learning_rate": 2.440668732735616e-05, "loss": 0.0336, "step": 4449 }, { "epoch": 0.3057998900494777, "grad_norm": 0.7166334390640259, "learning_rate": 2.4404086587799273e-05, "loss": 0.0246, "step": 4450 }, { "epoch": 0.30586860912589336, "grad_norm": 1.1241697072982788, "learning_rate": 2.440148538238328e-05, "loss": 0.0304, "step": 4451 }, { "epoch": 0.30593732820230896, "grad_norm": 0.8118840456008911, "learning_rate": 2.4398883711237026e-05, "loss": 0.0363, "step": 4452 }, { "epoch": 0.30600604727872455, "grad_norm": 1.199484944343567, "learning_rate": 2.439628157448939e-05, "loss": 0.028, "step": 4453 }, { "epoch": 0.3060747663551402, "grad_norm": 0.8461492657661438, "learning_rate": 2.439367897226929e-05, "loss": 0.0444, "step": 4454 }, { "epoch": 0.3061434854315558, "grad_norm": 1.1080578565597534, "learning_rate": 2.4391075904705648e-05, "loss": 0.0322, "step": 4455 }, { "epoch": 0.30621220450797143, "grad_norm": 1.0247735977172852, "learning_rate": 2.4388472371927402e-05, "loss": 0.0299, "step": 4456 }, { "epoch": 0.306280923584387, "grad_norm": 1.0660605430603027, "learning_rate": 2.4385868374063545e-05, "loss": 0.0332, "step": 4457 }, { "epoch": 0.3063496426608026, "grad_norm": 0.6347733736038208, "learning_rate": 2.4383263911243065e-05, "loss": 0.028, "step": 4458 }, { "epoch": 0.30641836173721826, "grad_norm": 0.7161625027656555, "learning_rate": 2.4380658983594977e-05, "loss": 0.0258, "step": 4459 }, { "epoch": 0.30648708081363385, "grad_norm": 0.9348323345184326, "learning_rate": 2.4378053591248338e-05, "loss": 0.0337, "step": 4460 }, { "epoch": 0.3065557998900495, "grad_norm": 0.792452871799469, "learning_rate": 2.4375447734332197e-05, "loss": 0.0419, "step": 4461 }, { "epoch": 0.3066245189664651, "grad_norm": 1.1321381330490112, "learning_rate": 2.437284141297565e-05, "loss": 0.0429, "step": 4462 }, { "epoch": 0.3066932380428807, "grad_norm": 1.3342887163162231, "learning_rate": 2.4370234627307812e-05, "loss": 0.0309, "step": 4463 }, { "epoch": 0.3067619571192963, "grad_norm": 0.6589265465736389, "learning_rate": 2.4367627377457818e-05, "loss": 0.0231, "step": 4464 }, { "epoch": 0.3068306761957119, "grad_norm": 0.9029957056045532, "learning_rate": 2.4365019663554816e-05, "loss": 0.0267, "step": 4465 }, { "epoch": 0.30689939527212756, "grad_norm": 0.8323016166687012, "learning_rate": 2.4362411485727997e-05, "loss": 0.0331, "step": 4466 }, { "epoch": 0.30696811434854315, "grad_norm": 1.2826776504516602, "learning_rate": 2.4359802844106566e-05, "loss": 0.0263, "step": 4467 }, { "epoch": 0.30703683342495874, "grad_norm": 0.7041542530059814, "learning_rate": 2.4357193738819737e-05, "loss": 0.0282, "step": 4468 }, { "epoch": 0.3071055525013744, "grad_norm": 1.00362229347229, "learning_rate": 2.4354584169996773e-05, "loss": 0.0335, "step": 4469 }, { "epoch": 0.30717427157779, "grad_norm": 0.765821099281311, "learning_rate": 2.435197413776694e-05, "loss": 0.0397, "step": 4470 }, { "epoch": 0.3072429906542056, "grad_norm": 1.5167604684829712, "learning_rate": 2.4349363642259527e-05, "loss": 0.0403, "step": 4471 }, { "epoch": 0.3073117097306212, "grad_norm": 0.8604162335395813, "learning_rate": 2.434675268360387e-05, "loss": 0.0314, "step": 4472 }, { "epoch": 0.3073804288070368, "grad_norm": 0.6135399341583252, "learning_rate": 2.43441412619293e-05, "loss": 0.0272, "step": 4473 }, { "epoch": 0.30744914788345246, "grad_norm": 1.0768659114837646, "learning_rate": 2.4341529377365182e-05, "loss": 0.0354, "step": 4474 }, { "epoch": 0.30751786695986805, "grad_norm": 1.5709489583969116, "learning_rate": 2.4338917030040903e-05, "loss": 0.0439, "step": 4475 }, { "epoch": 0.3075865860362837, "grad_norm": 0.7651220560073853, "learning_rate": 2.4336304220085875e-05, "loss": 0.0268, "step": 4476 }, { "epoch": 0.3076553051126993, "grad_norm": 0.6473732590675354, "learning_rate": 2.433369094762953e-05, "loss": 0.0349, "step": 4477 }, { "epoch": 0.3077240241891149, "grad_norm": 1.1911001205444336, "learning_rate": 2.433107721280133e-05, "loss": 0.0357, "step": 4478 }, { "epoch": 0.3077927432655305, "grad_norm": 1.458088994026184, "learning_rate": 2.432846301573074e-05, "loss": 0.0397, "step": 4479 }, { "epoch": 0.3078614623419461, "grad_norm": 0.662499189376831, "learning_rate": 2.432584835654728e-05, "loss": 0.0434, "step": 4480 }, { "epoch": 0.30793018141836176, "grad_norm": 0.9436243176460266, "learning_rate": 2.4323233235380456e-05, "loss": 0.0273, "step": 4481 }, { "epoch": 0.30799890049477735, "grad_norm": 1.4357643127441406, "learning_rate": 2.432061765235983e-05, "loss": 0.0327, "step": 4482 }, { "epoch": 0.30806761957119294, "grad_norm": 0.8920970559120178, "learning_rate": 2.4318001607614967e-05, "loss": 0.0306, "step": 4483 }, { "epoch": 0.3081363386476086, "grad_norm": 0.6770675778388977, "learning_rate": 2.4315385101275464e-05, "loss": 0.0245, "step": 4484 }, { "epoch": 0.3082050577240242, "grad_norm": 1.0620826482772827, "learning_rate": 2.431276813347093e-05, "loss": 0.0273, "step": 4485 }, { "epoch": 0.3082737768004398, "grad_norm": 0.8409186005592346, "learning_rate": 2.4310150704331013e-05, "loss": 0.0251, "step": 4486 }, { "epoch": 0.3083424958768554, "grad_norm": 1.3583571910858154, "learning_rate": 2.430753281398537e-05, "loss": 0.0425, "step": 4487 }, { "epoch": 0.308411214953271, "grad_norm": 0.4405853748321533, "learning_rate": 2.4304914462563686e-05, "loss": 0.0455, "step": 4488 }, { "epoch": 0.30847993402968665, "grad_norm": 0.828704297542572, "learning_rate": 2.4302295650195666e-05, "loss": 0.0262, "step": 4489 }, { "epoch": 0.30854865310610224, "grad_norm": 1.1132659912109375, "learning_rate": 2.4299676377011053e-05, "loss": 0.0342, "step": 4490 }, { "epoch": 0.3086173721825179, "grad_norm": 0.8154954314231873, "learning_rate": 2.429705664313959e-05, "loss": 0.0251, "step": 4491 }, { "epoch": 0.3086860912589335, "grad_norm": 1.199497103691101, "learning_rate": 2.429443644871105e-05, "loss": 0.0414, "step": 4492 }, { "epoch": 0.3087548103353491, "grad_norm": 0.6459832191467285, "learning_rate": 2.4291815793855234e-05, "loss": 0.0343, "step": 4493 }, { "epoch": 0.3088235294117647, "grad_norm": 1.222449779510498, "learning_rate": 2.4289194678701972e-05, "loss": 0.0278, "step": 4494 }, { "epoch": 0.3088922484881803, "grad_norm": 2.458733081817627, "learning_rate": 2.42865731033811e-05, "loss": 0.0459, "step": 4495 }, { "epoch": 0.30896096756459596, "grad_norm": 1.4981296062469482, "learning_rate": 2.4283951068022495e-05, "loss": 0.0395, "step": 4496 }, { "epoch": 0.30902968664101155, "grad_norm": 1.0691473484039307, "learning_rate": 2.428132857275604e-05, "loss": 0.0288, "step": 4497 }, { "epoch": 0.30909840571742714, "grad_norm": 1.4309922456741333, "learning_rate": 2.4278705617711644e-05, "loss": 0.0461, "step": 4498 }, { "epoch": 0.3091671247938428, "grad_norm": 0.6715925335884094, "learning_rate": 2.427608220301925e-05, "loss": 0.0242, "step": 4499 }, { "epoch": 0.3092358438702584, "grad_norm": 1.8499869108200073, "learning_rate": 2.4273458328808818e-05, "loss": 0.028, "step": 4500 }, { "epoch": 0.309304562946674, "grad_norm": 0.9242340326309204, "learning_rate": 2.4270833995210318e-05, "loss": 0.0354, "step": 4501 }, { "epoch": 0.3093732820230896, "grad_norm": 1.0710142850875854, "learning_rate": 2.426820920235377e-05, "loss": 0.0327, "step": 4502 }, { "epoch": 0.3094420010995052, "grad_norm": 1.2545452117919922, "learning_rate": 2.426558395036919e-05, "loss": 0.0307, "step": 4503 }, { "epoch": 0.30951072017592085, "grad_norm": 0.6145105957984924, "learning_rate": 2.426295823938663e-05, "loss": 0.0269, "step": 4504 }, { "epoch": 0.30957943925233644, "grad_norm": 0.8335155248641968, "learning_rate": 2.4260332069536163e-05, "loss": 0.0272, "step": 4505 }, { "epoch": 0.3096481583287521, "grad_norm": 1.36825430393219, "learning_rate": 2.425770544094788e-05, "loss": 0.0484, "step": 4506 }, { "epoch": 0.3097168774051677, "grad_norm": 1.021874189376831, "learning_rate": 2.4255078353751907e-05, "loss": 0.0271, "step": 4507 }, { "epoch": 0.30978559648158327, "grad_norm": 0.7136865854263306, "learning_rate": 2.4252450808078378e-05, "loss": 0.0251, "step": 4508 }, { "epoch": 0.3098543155579989, "grad_norm": 0.6858087182044983, "learning_rate": 2.4249822804057463e-05, "loss": 0.0297, "step": 4509 }, { "epoch": 0.3099230346344145, "grad_norm": 0.6324765086174011, "learning_rate": 2.4247194341819337e-05, "loss": 0.0323, "step": 4510 }, { "epoch": 0.30999175371083015, "grad_norm": 0.5434035658836365, "learning_rate": 2.4244565421494212e-05, "loss": 0.0281, "step": 4511 }, { "epoch": 0.31006047278724574, "grad_norm": 2.8203978538513184, "learning_rate": 2.424193604321233e-05, "loss": 0.0384, "step": 4512 }, { "epoch": 0.31012919186366134, "grad_norm": 0.7015310525894165, "learning_rate": 2.4239306207103936e-05, "loss": 0.0348, "step": 4513 }, { "epoch": 0.310197910940077, "grad_norm": 1.9111135005950928, "learning_rate": 2.4236675913299302e-05, "loss": 0.0378, "step": 4514 }, { "epoch": 0.3102666300164926, "grad_norm": 0.7543812990188599, "learning_rate": 2.423404516192874e-05, "loss": 0.0424, "step": 4515 }, { "epoch": 0.31033534909290816, "grad_norm": 0.5834031105041504, "learning_rate": 2.423141395312256e-05, "loss": 0.0318, "step": 4516 }, { "epoch": 0.3104040681693238, "grad_norm": 1.126612901687622, "learning_rate": 2.4228782287011112e-05, "loss": 0.0366, "step": 4517 }, { "epoch": 0.3104727872457394, "grad_norm": 0.5059419274330139, "learning_rate": 2.4226150163724765e-05, "loss": 0.0264, "step": 4518 }, { "epoch": 0.31054150632215505, "grad_norm": 1.0450807809829712, "learning_rate": 2.4223517583393908e-05, "loss": 0.0327, "step": 4519 }, { "epoch": 0.31061022539857064, "grad_norm": 0.6760374903678894, "learning_rate": 2.4220884546148952e-05, "loss": 0.0287, "step": 4520 }, { "epoch": 0.31067894447498623, "grad_norm": 1.1723276376724243, "learning_rate": 2.421825105212033e-05, "loss": 0.0265, "step": 4521 }, { "epoch": 0.3107476635514019, "grad_norm": 0.7282127737998962, "learning_rate": 2.4215617101438503e-05, "loss": 0.05, "step": 4522 }, { "epoch": 0.31081638262781747, "grad_norm": 0.3663162589073181, "learning_rate": 2.4212982694233954e-05, "loss": 0.0231, "step": 4523 }, { "epoch": 0.3108851017042331, "grad_norm": 0.7053138017654419, "learning_rate": 2.4210347830637184e-05, "loss": 0.0299, "step": 4524 }, { "epoch": 0.3109538207806487, "grad_norm": 0.6457951664924622, "learning_rate": 2.4207712510778717e-05, "loss": 0.0327, "step": 4525 }, { "epoch": 0.3110225398570643, "grad_norm": 2.210186004638672, "learning_rate": 2.4205076734789102e-05, "loss": 0.0411, "step": 4526 }, { "epoch": 0.31109125893347994, "grad_norm": 0.7247411608695984, "learning_rate": 2.420244050279891e-05, "loss": 0.0291, "step": 4527 }, { "epoch": 0.31115997800989553, "grad_norm": 1.0967289209365845, "learning_rate": 2.4199803814938737e-05, "loss": 0.0334, "step": 4528 }, { "epoch": 0.3112286970863112, "grad_norm": 0.9835434556007385, "learning_rate": 2.4197166671339198e-05, "loss": 0.0299, "step": 4529 }, { "epoch": 0.31129741616272677, "grad_norm": 0.8648220896720886, "learning_rate": 2.4194529072130926e-05, "loss": 0.0377, "step": 4530 }, { "epoch": 0.31136613523914236, "grad_norm": 0.5911945104598999, "learning_rate": 2.4191891017444586e-05, "loss": 0.0209, "step": 4531 }, { "epoch": 0.311434854315558, "grad_norm": 1.245491862297058, "learning_rate": 2.418925250741087e-05, "loss": 0.0303, "step": 4532 }, { "epoch": 0.3115035733919736, "grad_norm": 0.8797070384025574, "learning_rate": 2.4186613542160472e-05, "loss": 0.038, "step": 4533 }, { "epoch": 0.31157229246838924, "grad_norm": 1.569041132926941, "learning_rate": 2.4183974121824127e-05, "loss": 0.0405, "step": 4534 }, { "epoch": 0.31164101154480484, "grad_norm": 0.43988367915153503, "learning_rate": 2.4181334246532583e-05, "loss": 0.0199, "step": 4535 }, { "epoch": 0.3117097306212204, "grad_norm": 0.7002723813056946, "learning_rate": 2.4178693916416625e-05, "loss": 0.0281, "step": 4536 }, { "epoch": 0.3117784496976361, "grad_norm": 0.6858330965042114, "learning_rate": 2.417605313160703e-05, "loss": 0.0292, "step": 4537 }, { "epoch": 0.31184716877405166, "grad_norm": 0.9860721826553345, "learning_rate": 2.417341189223463e-05, "loss": 0.0267, "step": 4538 }, { "epoch": 0.3119158878504673, "grad_norm": 1.0313198566436768, "learning_rate": 2.417077019843027e-05, "loss": 0.0241, "step": 4539 }, { "epoch": 0.3119846069268829, "grad_norm": 0.660496175289154, "learning_rate": 2.4168128050324806e-05, "loss": 0.0226, "step": 4540 }, { "epoch": 0.3120533260032985, "grad_norm": 1.0520824193954468, "learning_rate": 2.416548544804913e-05, "loss": 0.0282, "step": 4541 }, { "epoch": 0.31212204507971414, "grad_norm": 1.8328163623809814, "learning_rate": 2.4162842391734144e-05, "loss": 0.0361, "step": 4542 }, { "epoch": 0.31219076415612973, "grad_norm": 0.6225996613502502, "learning_rate": 2.4160198881510784e-05, "loss": 0.0315, "step": 4543 }, { "epoch": 0.3122594832325454, "grad_norm": 0.9258427023887634, "learning_rate": 2.4157554917510007e-05, "loss": 0.0232, "step": 4544 }, { "epoch": 0.31232820230896097, "grad_norm": 1.735343098640442, "learning_rate": 2.4154910499862784e-05, "loss": 0.0308, "step": 4545 }, { "epoch": 0.31239692138537656, "grad_norm": 0.5142659544944763, "learning_rate": 2.415226562870012e-05, "loss": 0.0371, "step": 4546 }, { "epoch": 0.3124656404617922, "grad_norm": 0.5462760925292969, "learning_rate": 2.414962030415303e-05, "loss": 0.0339, "step": 4547 }, { "epoch": 0.3125343595382078, "grad_norm": 0.8767867684364319, "learning_rate": 2.4146974526352562e-05, "loss": 0.0326, "step": 4548 }, { "epoch": 0.31260307861462344, "grad_norm": 0.846916139125824, "learning_rate": 2.414432829542978e-05, "loss": 0.0248, "step": 4549 }, { "epoch": 0.31267179769103903, "grad_norm": 1.405096173286438, "learning_rate": 2.4141681611515776e-05, "loss": 0.0401, "step": 4550 }, { "epoch": 0.3127405167674546, "grad_norm": 0.6573109030723572, "learning_rate": 2.413903447474166e-05, "loss": 0.0237, "step": 4551 }, { "epoch": 0.31280923584387027, "grad_norm": 1.6543763875961304, "learning_rate": 2.4136386885238558e-05, "loss": 0.0261, "step": 4552 }, { "epoch": 0.31287795492028586, "grad_norm": 1.2571921348571777, "learning_rate": 2.4133738843137646e-05, "loss": 0.0378, "step": 4553 }, { "epoch": 0.3129466739967015, "grad_norm": 1.1993567943572998, "learning_rate": 2.413109034857008e-05, "loss": 0.0317, "step": 4554 }, { "epoch": 0.3130153930731171, "grad_norm": 0.7738907933235168, "learning_rate": 2.4128441401667072e-05, "loss": 0.0327, "step": 4555 }, { "epoch": 0.3130841121495327, "grad_norm": 0.8300224542617798, "learning_rate": 2.412579200255985e-05, "loss": 0.0225, "step": 4556 }, { "epoch": 0.31315283122594834, "grad_norm": 0.7463357448577881, "learning_rate": 2.412314215137965e-05, "loss": 0.0319, "step": 4557 }, { "epoch": 0.3132215503023639, "grad_norm": 1.3293218612670898, "learning_rate": 2.412049184825774e-05, "loss": 0.048, "step": 4558 }, { "epoch": 0.3132902693787796, "grad_norm": 1.2306971549987793, "learning_rate": 2.411784109332542e-05, "loss": 0.0316, "step": 4559 }, { "epoch": 0.31335898845519516, "grad_norm": 0.7225078344345093, "learning_rate": 2.4115189886713997e-05, "loss": 0.0297, "step": 4560 }, { "epoch": 0.31342770753161076, "grad_norm": 1.0541245937347412, "learning_rate": 2.411253822855481e-05, "loss": 0.0398, "step": 4561 }, { "epoch": 0.3134964266080264, "grad_norm": 0.7866917848587036, "learning_rate": 2.410988611897921e-05, "loss": 0.0336, "step": 4562 }, { "epoch": 0.313565145684442, "grad_norm": 0.5729982256889343, "learning_rate": 2.4107233558118583e-05, "loss": 0.0289, "step": 4563 }, { "epoch": 0.31363386476085764, "grad_norm": 0.6723389625549316, "learning_rate": 2.4104580546104325e-05, "loss": 0.0216, "step": 4564 }, { "epoch": 0.31370258383727323, "grad_norm": 0.8254163861274719, "learning_rate": 2.4101927083067865e-05, "loss": 0.0402, "step": 4565 }, { "epoch": 0.3137713029136888, "grad_norm": 0.4597781002521515, "learning_rate": 2.4099273169140656e-05, "loss": 0.0306, "step": 4566 }, { "epoch": 0.31384002199010447, "grad_norm": 0.5501473546028137, "learning_rate": 2.4096618804454154e-05, "loss": 0.0225, "step": 4567 }, { "epoch": 0.31390874106652006, "grad_norm": 0.6364887952804565, "learning_rate": 2.4093963989139864e-05, "loss": 0.027, "step": 4568 }, { "epoch": 0.3139774601429357, "grad_norm": 0.8206453323364258, "learning_rate": 2.4091308723329294e-05, "loss": 0.0281, "step": 4569 }, { "epoch": 0.3140461792193513, "grad_norm": 1.4901467561721802, "learning_rate": 2.4088653007153983e-05, "loss": 0.0385, "step": 4570 }, { "epoch": 0.3141148982957669, "grad_norm": 0.9375789165496826, "learning_rate": 2.408599684074548e-05, "loss": 0.0344, "step": 4571 }, { "epoch": 0.31418361737218253, "grad_norm": 0.8117548823356628, "learning_rate": 2.408334022423538e-05, "loss": 0.0303, "step": 4572 }, { "epoch": 0.3142523364485981, "grad_norm": 0.683481752872467, "learning_rate": 2.4080683157755277e-05, "loss": 0.0278, "step": 4573 }, { "epoch": 0.31432105552501377, "grad_norm": 0.6708834767341614, "learning_rate": 2.40780256414368e-05, "loss": 0.0235, "step": 4574 }, { "epoch": 0.31438977460142936, "grad_norm": 1.2261558771133423, "learning_rate": 2.4075367675411597e-05, "loss": 0.0381, "step": 4575 }, { "epoch": 0.31445849367784495, "grad_norm": 1.223298192024231, "learning_rate": 2.4072709259811336e-05, "loss": 0.0444, "step": 4576 }, { "epoch": 0.3145272127542606, "grad_norm": 1.4789010286331177, "learning_rate": 2.4070050394767713e-05, "loss": 0.0316, "step": 4577 }, { "epoch": 0.3145959318306762, "grad_norm": 1.0631829500198364, "learning_rate": 2.4067391080412437e-05, "loss": 0.0436, "step": 4578 }, { "epoch": 0.31466465090709184, "grad_norm": 0.9297019839286804, "learning_rate": 2.406473131687725e-05, "loss": 0.0278, "step": 4579 }, { "epoch": 0.3147333699835074, "grad_norm": 1.0577741861343384, "learning_rate": 2.4062071104293915e-05, "loss": 0.0411, "step": 4580 }, { "epoch": 0.314802089059923, "grad_norm": 0.8734716773033142, "learning_rate": 2.4059410442794204e-05, "loss": 0.0372, "step": 4581 }, { "epoch": 0.31487080813633866, "grad_norm": 0.7135968804359436, "learning_rate": 2.4056749332509923e-05, "loss": 0.0277, "step": 4582 }, { "epoch": 0.31493952721275426, "grad_norm": 1.6167129278182983, "learning_rate": 2.4054087773572905e-05, "loss": 0.0412, "step": 4583 }, { "epoch": 0.31500824628916985, "grad_norm": 0.7779943943023682, "learning_rate": 2.4051425766114992e-05, "loss": 0.0292, "step": 4584 }, { "epoch": 0.3150769653655855, "grad_norm": 1.1647793054580688, "learning_rate": 2.404876331026805e-05, "loss": 0.0306, "step": 4585 }, { "epoch": 0.3151456844420011, "grad_norm": 0.8189385533332825, "learning_rate": 2.404610040616398e-05, "loss": 0.0353, "step": 4586 }, { "epoch": 0.31521440351841673, "grad_norm": 0.8965117931365967, "learning_rate": 2.4043437053934696e-05, "loss": 0.0236, "step": 4587 }, { "epoch": 0.3152831225948323, "grad_norm": 0.6878779530525208, "learning_rate": 2.4040773253712133e-05, "loss": 0.0184, "step": 4588 }, { "epoch": 0.3153518416712479, "grad_norm": 1.0213227272033691, "learning_rate": 2.4038109005628247e-05, "loss": 0.0338, "step": 4589 }, { "epoch": 0.31542056074766356, "grad_norm": 0.714451014995575, "learning_rate": 2.4035444309815024e-05, "loss": 0.0281, "step": 4590 }, { "epoch": 0.31548927982407915, "grad_norm": 0.9393527507781982, "learning_rate": 2.4032779166404468e-05, "loss": 0.0434, "step": 4591 }, { "epoch": 0.3155579989004948, "grad_norm": 1.1395394802093506, "learning_rate": 2.40301135755286e-05, "loss": 0.0309, "step": 4592 }, { "epoch": 0.3156267179769104, "grad_norm": 2.4868128299713135, "learning_rate": 2.402744753731947e-05, "loss": 0.0301, "step": 4593 }, { "epoch": 0.315695437053326, "grad_norm": 1.0335156917572021, "learning_rate": 2.4024781051909156e-05, "loss": 0.0282, "step": 4594 }, { "epoch": 0.3157641561297416, "grad_norm": 1.2920069694519043, "learning_rate": 2.402211411942973e-05, "loss": 0.0598, "step": 4595 }, { "epoch": 0.3158328752061572, "grad_norm": 1.4720065593719482, "learning_rate": 2.4019446740013328e-05, "loss": 0.0518, "step": 4596 }, { "epoch": 0.31590159428257286, "grad_norm": 0.7067310214042664, "learning_rate": 2.401677891379208e-05, "loss": 0.0435, "step": 4597 }, { "epoch": 0.31597031335898845, "grad_norm": 0.8595033288002014, "learning_rate": 2.4014110640898137e-05, "loss": 0.042, "step": 4598 }, { "epoch": 0.31603903243540404, "grad_norm": 0.7672697901725769, "learning_rate": 2.4011441921463685e-05, "loss": 0.0233, "step": 4599 }, { "epoch": 0.3161077515118197, "grad_norm": 0.6748706698417664, "learning_rate": 2.400877275562093e-05, "loss": 0.0342, "step": 4600 }, { "epoch": 0.3161764705882353, "grad_norm": 1.0115200281143188, "learning_rate": 2.400610314350209e-05, "loss": 0.0454, "step": 4601 }, { "epoch": 0.3162451896646509, "grad_norm": 1.0347422361373901, "learning_rate": 2.400343308523942e-05, "loss": 0.042, "step": 4602 }, { "epoch": 0.3163139087410665, "grad_norm": 1.1490774154663086, "learning_rate": 2.4000762580965182e-05, "loss": 0.0355, "step": 4603 }, { "epoch": 0.3163826278174821, "grad_norm": 0.33318498730659485, "learning_rate": 2.3998091630811675e-05, "loss": 0.0212, "step": 4604 }, { "epoch": 0.31645134689389776, "grad_norm": 1.3147850036621094, "learning_rate": 2.39954202349112e-05, "loss": 0.0384, "step": 4605 }, { "epoch": 0.31652006597031335, "grad_norm": 1.2927894592285156, "learning_rate": 2.3992748393396107e-05, "loss": 0.0368, "step": 4606 }, { "epoch": 0.316588785046729, "grad_norm": 0.6844911575317383, "learning_rate": 2.3990076106398746e-05, "loss": 0.0377, "step": 4607 }, { "epoch": 0.3166575041231446, "grad_norm": 1.0809459686279297, "learning_rate": 2.3987403374051496e-05, "loss": 0.0343, "step": 4608 }, { "epoch": 0.3167262231995602, "grad_norm": 0.8305254578590393, "learning_rate": 2.398473019648676e-05, "loss": 0.0234, "step": 4609 }, { "epoch": 0.3167949422759758, "grad_norm": 1.1559361219406128, "learning_rate": 2.3982056573836966e-05, "loss": 0.0325, "step": 4610 }, { "epoch": 0.3168636613523914, "grad_norm": 1.2449226379394531, "learning_rate": 2.397938250623455e-05, "loss": 0.0451, "step": 4611 }, { "epoch": 0.31693238042880706, "grad_norm": 1.878537893295288, "learning_rate": 2.397670799381199e-05, "loss": 0.0559, "step": 4612 }, { "epoch": 0.31700109950522265, "grad_norm": 1.1106325387954712, "learning_rate": 2.3974033036701766e-05, "loss": 0.0282, "step": 4613 }, { "epoch": 0.31706981858163824, "grad_norm": 0.7792430520057678, "learning_rate": 2.3971357635036397e-05, "loss": 0.0338, "step": 4614 }, { "epoch": 0.3171385376580539, "grad_norm": 1.1574673652648926, "learning_rate": 2.396868178894842e-05, "loss": 0.0371, "step": 4615 }, { "epoch": 0.3172072567344695, "grad_norm": 0.7897688746452332, "learning_rate": 2.396600549857038e-05, "loss": 0.0317, "step": 4616 }, { "epoch": 0.3172759758108851, "grad_norm": 1.0246185064315796, "learning_rate": 2.3963328764034863e-05, "loss": 0.0297, "step": 4617 }, { "epoch": 0.3173446948873007, "grad_norm": 1.2019842863082886, "learning_rate": 2.396065158547447e-05, "loss": 0.038, "step": 4618 }, { "epoch": 0.3174134139637163, "grad_norm": 1.3243736028671265, "learning_rate": 2.3957973963021817e-05, "loss": 0.037, "step": 4619 }, { "epoch": 0.31748213304013195, "grad_norm": 0.7695399522781372, "learning_rate": 2.3955295896809547e-05, "loss": 0.0507, "step": 4620 }, { "epoch": 0.31755085211654754, "grad_norm": 0.9364977478981018, "learning_rate": 2.3952617386970334e-05, "loss": 0.0445, "step": 4621 }, { "epoch": 0.3176195711929632, "grad_norm": 1.636489987373352, "learning_rate": 2.394993843363686e-05, "loss": 0.0407, "step": 4622 }, { "epoch": 0.3176882902693788, "grad_norm": 0.7596069574356079, "learning_rate": 2.394725903694184e-05, "loss": 0.0335, "step": 4623 }, { "epoch": 0.3177570093457944, "grad_norm": 0.6425881385803223, "learning_rate": 2.3944579197018e-05, "loss": 0.0366, "step": 4624 }, { "epoch": 0.31782572842221, "grad_norm": 0.8459494709968567, "learning_rate": 2.394189891399809e-05, "loss": 0.0477, "step": 4625 }, { "epoch": 0.3178944474986256, "grad_norm": 1.200473427772522, "learning_rate": 2.39392181880149e-05, "loss": 0.0276, "step": 4626 }, { "epoch": 0.31796316657504126, "grad_norm": 0.6743307709693909, "learning_rate": 2.3936537019201215e-05, "loss": 0.0329, "step": 4627 }, { "epoch": 0.31803188565145685, "grad_norm": 0.8654927611351013, "learning_rate": 2.393385540768986e-05, "loss": 0.03, "step": 4628 }, { "epoch": 0.31810060472787244, "grad_norm": 0.7058685421943665, "learning_rate": 2.3931173353613672e-05, "loss": 0.0209, "step": 4629 }, { "epoch": 0.3181693238042881, "grad_norm": 0.82504802942276, "learning_rate": 2.3928490857105527e-05, "loss": 0.029, "step": 4630 }, { "epoch": 0.3182380428807037, "grad_norm": 1.006279706954956, "learning_rate": 2.392580791829829e-05, "loss": 0.0313, "step": 4631 }, { "epoch": 0.3183067619571193, "grad_norm": 1.0133675336837769, "learning_rate": 2.3923124537324888e-05, "loss": 0.0365, "step": 4632 }, { "epoch": 0.3183754810335349, "grad_norm": 0.6320276260375977, "learning_rate": 2.3920440714318238e-05, "loss": 0.0258, "step": 4633 }, { "epoch": 0.3184442001099505, "grad_norm": 0.8259260058403015, "learning_rate": 2.391775644941129e-05, "loss": 0.0301, "step": 4634 }, { "epoch": 0.31851291918636615, "grad_norm": 0.8062256574630737, "learning_rate": 2.391507174273703e-05, "loss": 0.0238, "step": 4635 }, { "epoch": 0.31858163826278174, "grad_norm": 0.6725404262542725, "learning_rate": 2.391238659442844e-05, "loss": 0.0352, "step": 4636 }, { "epoch": 0.3186503573391974, "grad_norm": 1.5594534873962402, "learning_rate": 2.3909701004618537e-05, "loss": 0.0439, "step": 4637 }, { "epoch": 0.318719076415613, "grad_norm": 0.7353659868240356, "learning_rate": 2.390701497344037e-05, "loss": 0.0272, "step": 4638 }, { "epoch": 0.31878779549202857, "grad_norm": 1.3584157228469849, "learning_rate": 2.3904328501026992e-05, "loss": 0.0417, "step": 4639 }, { "epoch": 0.3188565145684442, "grad_norm": 0.7120450139045715, "learning_rate": 2.390164158751149e-05, "loss": 0.0405, "step": 4640 }, { "epoch": 0.3189252336448598, "grad_norm": 1.2915116548538208, "learning_rate": 2.3898954233026963e-05, "loss": 0.0326, "step": 4641 }, { "epoch": 0.31899395272127545, "grad_norm": 0.5733867883682251, "learning_rate": 2.3896266437706536e-05, "loss": 0.0289, "step": 4642 }, { "epoch": 0.31906267179769104, "grad_norm": 0.6833091378211975, "learning_rate": 2.3893578201683357e-05, "loss": 0.0288, "step": 4643 }, { "epoch": 0.31913139087410664, "grad_norm": 0.8997706174850464, "learning_rate": 2.38908895250906e-05, "loss": 0.019, "step": 4644 }, { "epoch": 0.3192001099505223, "grad_norm": 0.9889748096466064, "learning_rate": 2.388820040806146e-05, "loss": 0.0325, "step": 4645 }, { "epoch": 0.3192688290269379, "grad_norm": 0.832614541053772, "learning_rate": 2.388551085072914e-05, "loss": 0.0279, "step": 4646 }, { "epoch": 0.31933754810335346, "grad_norm": 0.9499784708023071, "learning_rate": 2.3882820853226885e-05, "loss": 0.0372, "step": 4647 }, { "epoch": 0.3194062671797691, "grad_norm": 0.8843510746955872, "learning_rate": 2.3880130415687945e-05, "loss": 0.0375, "step": 4648 }, { "epoch": 0.3194749862561847, "grad_norm": 0.7980295419692993, "learning_rate": 2.3877439538245595e-05, "loss": 0.0359, "step": 4649 }, { "epoch": 0.31954370533260035, "grad_norm": 1.2987178564071655, "learning_rate": 2.387474822103315e-05, "loss": 0.0368, "step": 4650 }, { "epoch": 0.31961242440901594, "grad_norm": 1.2646946907043457, "learning_rate": 2.3872056464183922e-05, "loss": 0.047, "step": 4651 }, { "epoch": 0.31968114348543153, "grad_norm": 0.8965786099433899, "learning_rate": 2.3869364267831256e-05, "loss": 0.0349, "step": 4652 }, { "epoch": 0.3197498625618472, "grad_norm": 0.985560417175293, "learning_rate": 2.3866671632108515e-05, "loss": 0.0281, "step": 4653 }, { "epoch": 0.31981858163826277, "grad_norm": 1.093263864517212, "learning_rate": 2.3863978557149094e-05, "loss": 0.0318, "step": 4654 }, { "epoch": 0.3198873007146784, "grad_norm": 0.7706771492958069, "learning_rate": 2.3861285043086398e-05, "loss": 0.0368, "step": 4655 }, { "epoch": 0.319956019791094, "grad_norm": 0.5005068182945251, "learning_rate": 2.385859109005386e-05, "loss": 0.0249, "step": 4656 }, { "epoch": 0.3200247388675096, "grad_norm": 1.4271494150161743, "learning_rate": 2.3855896698184937e-05, "loss": 0.0355, "step": 4657 }, { "epoch": 0.32009345794392524, "grad_norm": 0.9899484515190125, "learning_rate": 2.385320186761309e-05, "loss": 0.0278, "step": 4658 }, { "epoch": 0.32016217702034083, "grad_norm": 0.9814602732658386, "learning_rate": 2.3850506598471828e-05, "loss": 0.0485, "step": 4659 }, { "epoch": 0.3202308960967565, "grad_norm": 0.9527914524078369, "learning_rate": 2.3847810890894664e-05, "loss": 0.0296, "step": 4660 }, { "epoch": 0.32029961517317207, "grad_norm": 0.9542682766914368, "learning_rate": 2.3845114745015135e-05, "loss": 0.0178, "step": 4661 }, { "epoch": 0.32036833424958766, "grad_norm": 0.6504708528518677, "learning_rate": 2.3842418160966814e-05, "loss": 0.026, "step": 4662 }, { "epoch": 0.3204370533260033, "grad_norm": 0.7499579787254333, "learning_rate": 2.383972113888327e-05, "loss": 0.0266, "step": 4663 }, { "epoch": 0.3205057724024189, "grad_norm": 1.348079800605774, "learning_rate": 2.383702367889812e-05, "loss": 0.039, "step": 4664 }, { "epoch": 0.32057449147883454, "grad_norm": 0.6160635352134705, "learning_rate": 2.383432578114498e-05, "loss": 0.0236, "step": 4665 }, { "epoch": 0.32064321055525014, "grad_norm": 0.7280976176261902, "learning_rate": 2.3831627445757507e-05, "loss": 0.0188, "step": 4666 }, { "epoch": 0.3207119296316657, "grad_norm": 1.0468249320983887, "learning_rate": 2.3828928672869365e-05, "loss": 0.0229, "step": 4667 }, { "epoch": 0.3207806487080814, "grad_norm": 1.0828756093978882, "learning_rate": 2.3826229462614252e-05, "loss": 0.0438, "step": 4668 }, { "epoch": 0.32084936778449696, "grad_norm": 1.5186700820922852, "learning_rate": 2.3823529815125873e-05, "loss": 0.0384, "step": 4669 }, { "epoch": 0.3209180868609126, "grad_norm": 0.7679116725921631, "learning_rate": 2.3820829730537965e-05, "loss": 0.0342, "step": 4670 }, { "epoch": 0.3209868059373282, "grad_norm": 1.137500524520874, "learning_rate": 2.3818129208984293e-05, "loss": 0.0271, "step": 4671 }, { "epoch": 0.3210555250137438, "grad_norm": 0.9483278393745422, "learning_rate": 2.381542825059863e-05, "loss": 0.0441, "step": 4672 }, { "epoch": 0.32112424409015944, "grad_norm": 0.42354705929756165, "learning_rate": 2.3812726855514774e-05, "loss": 0.0302, "step": 4673 }, { "epoch": 0.32119296316657503, "grad_norm": 1.073598027229309, "learning_rate": 2.381002502386654e-05, "loss": 0.0369, "step": 4674 }, { "epoch": 0.3212616822429907, "grad_norm": 0.7051148414611816, "learning_rate": 2.380732275578779e-05, "loss": 0.0433, "step": 4675 }, { "epoch": 0.32133040131940627, "grad_norm": 1.1174181699752808, "learning_rate": 2.3804620051412373e-05, "loss": 0.0377, "step": 4676 }, { "epoch": 0.32139912039582186, "grad_norm": 1.061496376991272, "learning_rate": 2.3801916910874177e-05, "loss": 0.0451, "step": 4677 }, { "epoch": 0.3214678394722375, "grad_norm": 0.6012590527534485, "learning_rate": 2.3799213334307116e-05, "loss": 0.0223, "step": 4678 }, { "epoch": 0.3215365585486531, "grad_norm": 1.4867541790008545, "learning_rate": 2.3796509321845116e-05, "loss": 0.0408, "step": 4679 }, { "epoch": 0.32160527762506874, "grad_norm": 0.5532463788986206, "learning_rate": 2.379380487362213e-05, "loss": 0.0235, "step": 4680 }, { "epoch": 0.32167399670148433, "grad_norm": 0.9425316452980042, "learning_rate": 2.3791099989772135e-05, "loss": 0.0371, "step": 4681 }, { "epoch": 0.3217427157778999, "grad_norm": 0.5814061164855957, "learning_rate": 2.378839467042911e-05, "loss": 0.0247, "step": 4682 }, { "epoch": 0.32181143485431557, "grad_norm": 1.0461366176605225, "learning_rate": 2.378568891572709e-05, "loss": 0.0264, "step": 4683 }, { "epoch": 0.32188015393073116, "grad_norm": 1.0567454099655151, "learning_rate": 2.3782982725800103e-05, "loss": 0.0276, "step": 4684 }, { "epoch": 0.3219488730071468, "grad_norm": 0.9493486285209656, "learning_rate": 2.3780276100782204e-05, "loss": 0.0362, "step": 4685 }, { "epoch": 0.3220175920835624, "grad_norm": 0.8094517588615417, "learning_rate": 2.3777569040807482e-05, "loss": 0.0253, "step": 4686 }, { "epoch": 0.322086311159978, "grad_norm": 0.770602285861969, "learning_rate": 2.3774861546010033e-05, "loss": 0.0347, "step": 4687 }, { "epoch": 0.32215503023639364, "grad_norm": 1.2073321342468262, "learning_rate": 2.377215361652398e-05, "loss": 0.0284, "step": 4688 }, { "epoch": 0.3222237493128092, "grad_norm": 0.7677471041679382, "learning_rate": 2.3769445252483475e-05, "loss": 0.0276, "step": 4689 }, { "epoch": 0.3222924683892249, "grad_norm": 0.9683051109313965, "learning_rate": 2.3766736454022686e-05, "loss": 0.027, "step": 4690 }, { "epoch": 0.32236118746564046, "grad_norm": 1.2733875513076782, "learning_rate": 2.3764027221275796e-05, "loss": 0.0426, "step": 4691 }, { "epoch": 0.32242990654205606, "grad_norm": 1.243003249168396, "learning_rate": 2.3761317554377008e-05, "loss": 0.0315, "step": 4692 }, { "epoch": 0.3224986256184717, "grad_norm": 0.7711549997329712, "learning_rate": 2.3758607453460568e-05, "loss": 0.0389, "step": 4693 }, { "epoch": 0.3225673446948873, "grad_norm": 0.7576540112495422, "learning_rate": 2.375589691866072e-05, "loss": 0.038, "step": 4694 }, { "epoch": 0.32263606377130294, "grad_norm": 1.1255689859390259, "learning_rate": 2.3753185950111738e-05, "loss": 0.0349, "step": 4695 }, { "epoch": 0.32270478284771853, "grad_norm": 0.8242839574813843, "learning_rate": 2.375047454794792e-05, "loss": 0.0359, "step": 4696 }, { "epoch": 0.3227735019241341, "grad_norm": 0.9365834593772888, "learning_rate": 2.3747762712303584e-05, "loss": 0.0386, "step": 4697 }, { "epoch": 0.32284222100054977, "grad_norm": 1.1740885972976685, "learning_rate": 2.374505044331307e-05, "loss": 0.0291, "step": 4698 }, { "epoch": 0.32291094007696536, "grad_norm": 0.9290874004364014, "learning_rate": 2.3742337741110734e-05, "loss": 0.0353, "step": 4699 }, { "epoch": 0.322979659153381, "grad_norm": 0.5066708922386169, "learning_rate": 2.3739624605830962e-05, "loss": 0.0235, "step": 4700 }, { "epoch": 0.3230483782297966, "grad_norm": 0.9380823969841003, "learning_rate": 2.373691103760815e-05, "loss": 0.0399, "step": 4701 }, { "epoch": 0.3231170973062122, "grad_norm": 0.6725225448608398, "learning_rate": 2.373419703657673e-05, "loss": 0.0288, "step": 4702 }, { "epoch": 0.32318581638262783, "grad_norm": 0.8561004400253296, "learning_rate": 2.3731482602871145e-05, "loss": 0.0287, "step": 4703 }, { "epoch": 0.3232545354590434, "grad_norm": 0.6653369069099426, "learning_rate": 2.3728767736625864e-05, "loss": 0.0295, "step": 4704 }, { "epoch": 0.32332325453545907, "grad_norm": 0.7859249114990234, "learning_rate": 2.372605243797538e-05, "loss": 0.0266, "step": 4705 }, { "epoch": 0.32339197361187466, "grad_norm": 1.028084397315979, "learning_rate": 2.3723336707054192e-05, "loss": 0.0201, "step": 4706 }, { "epoch": 0.32346069268829025, "grad_norm": 1.0268981456756592, "learning_rate": 2.3720620543996837e-05, "loss": 0.0323, "step": 4707 }, { "epoch": 0.3235294117647059, "grad_norm": 0.7973083257675171, "learning_rate": 2.371790394893787e-05, "loss": 0.0364, "step": 4708 }, { "epoch": 0.3235981308411215, "grad_norm": 0.6097888350486755, "learning_rate": 2.3715186922011868e-05, "loss": 0.0317, "step": 4709 }, { "epoch": 0.3236668499175371, "grad_norm": 1.12998366355896, "learning_rate": 2.3712469463353418e-05, "loss": 0.0338, "step": 4710 }, { "epoch": 0.3237355689939527, "grad_norm": 0.6922215223312378, "learning_rate": 2.3709751573097153e-05, "loss": 0.0257, "step": 4711 }, { "epoch": 0.3238042880703683, "grad_norm": 0.8459084630012512, "learning_rate": 2.370703325137769e-05, "loss": 0.0216, "step": 4712 }, { "epoch": 0.32387300714678396, "grad_norm": 0.6234955787658691, "learning_rate": 2.3704314498329708e-05, "loss": 0.0202, "step": 4713 }, { "epoch": 0.32394172622319956, "grad_norm": 0.6219738125801086, "learning_rate": 2.3701595314087877e-05, "loss": 0.0214, "step": 4714 }, { "epoch": 0.32401044529961515, "grad_norm": 1.4453080892562866, "learning_rate": 2.3698875698786904e-05, "loss": 0.0341, "step": 4715 }, { "epoch": 0.3240791643760308, "grad_norm": 0.9181601405143738, "learning_rate": 2.3696155652561517e-05, "loss": 0.0347, "step": 4716 }, { "epoch": 0.3241478834524464, "grad_norm": 1.2415876388549805, "learning_rate": 2.3693435175546455e-05, "loss": 0.0447, "step": 4717 }, { "epoch": 0.32421660252886203, "grad_norm": 1.1406203508377075, "learning_rate": 2.3690714267876485e-05, "loss": 0.0309, "step": 4718 }, { "epoch": 0.3242853216052776, "grad_norm": 0.874035656452179, "learning_rate": 2.36879929296864e-05, "loss": 0.0279, "step": 4719 }, { "epoch": 0.3243540406816932, "grad_norm": 1.1976715326309204, "learning_rate": 2.368527116111101e-05, "loss": 0.04, "step": 4720 }, { "epoch": 0.32442275975810886, "grad_norm": 1.1978272199630737, "learning_rate": 2.3682548962285133e-05, "loss": 0.0357, "step": 4721 }, { "epoch": 0.32449147883452445, "grad_norm": 1.0972580909729004, "learning_rate": 2.367982633334364e-05, "loss": 0.0364, "step": 4722 }, { "epoch": 0.3245601979109401, "grad_norm": 1.3034894466400146, "learning_rate": 2.367710327442139e-05, "loss": 0.0492, "step": 4723 }, { "epoch": 0.3246289169873557, "grad_norm": 2.0769379138946533, "learning_rate": 2.367437978565329e-05, "loss": 0.038, "step": 4724 }, { "epoch": 0.3246976360637713, "grad_norm": 1.0820138454437256, "learning_rate": 2.367165586717424e-05, "loss": 0.0366, "step": 4725 }, { "epoch": 0.3247663551401869, "grad_norm": 0.7150774598121643, "learning_rate": 2.36689315191192e-05, "loss": 0.031, "step": 4726 }, { "epoch": 0.3248350742166025, "grad_norm": 0.6898335814476013, "learning_rate": 2.3666206741623102e-05, "loss": 0.0265, "step": 4727 }, { "epoch": 0.32490379329301816, "grad_norm": 0.5093157887458801, "learning_rate": 2.3663481534820942e-05, "loss": 0.0312, "step": 4728 }, { "epoch": 0.32497251236943375, "grad_norm": 1.1522350311279297, "learning_rate": 2.366075589884772e-05, "loss": 0.042, "step": 4729 }, { "epoch": 0.32504123144584934, "grad_norm": 0.8401603698730469, "learning_rate": 2.365802983383846e-05, "loss": 0.0295, "step": 4730 }, { "epoch": 0.325109950522265, "grad_norm": 0.520456850528717, "learning_rate": 2.3655303339928198e-05, "loss": 0.0275, "step": 4731 }, { "epoch": 0.3251786695986806, "grad_norm": 0.831757664680481, "learning_rate": 2.3652576417252007e-05, "loss": 0.0322, "step": 4732 }, { "epoch": 0.3252473886750962, "grad_norm": 0.5518521070480347, "learning_rate": 2.3649849065944963e-05, "loss": 0.0269, "step": 4733 }, { "epoch": 0.3253161077515118, "grad_norm": 0.7795596122741699, "learning_rate": 2.3647121286142183e-05, "loss": 0.0312, "step": 4734 }, { "epoch": 0.3253848268279274, "grad_norm": 0.6919777989387512, "learning_rate": 2.3644393077978793e-05, "loss": 0.0216, "step": 4735 }, { "epoch": 0.32545354590434306, "grad_norm": 1.8661317825317383, "learning_rate": 2.364166444158994e-05, "loss": 0.034, "step": 4736 }, { "epoch": 0.32552226498075865, "grad_norm": 1.4111266136169434, "learning_rate": 2.3638935377110795e-05, "loss": 0.0326, "step": 4737 }, { "epoch": 0.3255909840571743, "grad_norm": 0.872999370098114, "learning_rate": 2.3636205884676558e-05, "loss": 0.0299, "step": 4738 }, { "epoch": 0.3256597031335899, "grad_norm": 0.523194432258606, "learning_rate": 2.3633475964422435e-05, "loss": 0.0239, "step": 4739 }, { "epoch": 0.3257284222100055, "grad_norm": 0.5749959349632263, "learning_rate": 2.363074561648366e-05, "loss": 0.0315, "step": 4740 }, { "epoch": 0.3257971412864211, "grad_norm": 1.1595897674560547, "learning_rate": 2.3628014840995496e-05, "loss": 0.0366, "step": 4741 }, { "epoch": 0.3258658603628367, "grad_norm": 1.051889181137085, "learning_rate": 2.3625283638093208e-05, "loss": 0.0224, "step": 4742 }, { "epoch": 0.32593457943925236, "grad_norm": 2.179311752319336, "learning_rate": 2.3622552007912106e-05, "loss": 0.0331, "step": 4743 }, { "epoch": 0.32600329851566795, "grad_norm": 0.8546769022941589, "learning_rate": 2.3619819950587505e-05, "loss": 0.0365, "step": 4744 }, { "epoch": 0.32607201759208354, "grad_norm": 1.7914745807647705, "learning_rate": 2.3617087466254743e-05, "loss": 0.0574, "step": 4745 }, { "epoch": 0.3261407366684992, "grad_norm": 0.9705740213394165, "learning_rate": 2.3614354555049184e-05, "loss": 0.0335, "step": 4746 }, { "epoch": 0.3262094557449148, "grad_norm": 1.0948420763015747, "learning_rate": 2.3611621217106213e-05, "loss": 0.0272, "step": 4747 }, { "epoch": 0.3262781748213304, "grad_norm": 1.3508448600769043, "learning_rate": 2.360888745256123e-05, "loss": 0.0375, "step": 4748 }, { "epoch": 0.326346893897746, "grad_norm": 0.6317164897918701, "learning_rate": 2.3606153261549662e-05, "loss": 0.0401, "step": 4749 }, { "epoch": 0.3264156129741616, "grad_norm": 0.7318927049636841, "learning_rate": 2.360341864420696e-05, "loss": 0.0297, "step": 4750 }, { "epoch": 0.32648433205057725, "grad_norm": 0.9765655994415283, "learning_rate": 2.360068360066858e-05, "loss": 0.0366, "step": 4751 }, { "epoch": 0.32655305112699284, "grad_norm": 0.8197652101516724, "learning_rate": 2.3597948131070015e-05, "loss": 0.0273, "step": 4752 }, { "epoch": 0.3266217702034085, "grad_norm": 1.0274767875671387, "learning_rate": 2.3595212235546785e-05, "loss": 0.0285, "step": 4753 }, { "epoch": 0.3266904892798241, "grad_norm": 1.1289575099945068, "learning_rate": 2.3592475914234403e-05, "loss": 0.0301, "step": 4754 }, { "epoch": 0.32675920835623967, "grad_norm": 0.7872979640960693, "learning_rate": 2.3589739167268433e-05, "loss": 0.0333, "step": 4755 }, { "epoch": 0.3268279274326553, "grad_norm": 0.912166178226471, "learning_rate": 2.3587001994784444e-05, "loss": 0.0284, "step": 4756 }, { "epoch": 0.3268966465090709, "grad_norm": 0.6374949812889099, "learning_rate": 2.3584264396918032e-05, "loss": 0.0229, "step": 4757 }, { "epoch": 0.32696536558548656, "grad_norm": 0.7187755107879639, "learning_rate": 2.3581526373804807e-05, "loss": 0.0344, "step": 4758 }, { "epoch": 0.32703408466190215, "grad_norm": 1.0786895751953125, "learning_rate": 2.357878792558041e-05, "loss": 0.0214, "step": 4759 }, { "epoch": 0.32710280373831774, "grad_norm": 0.6808608770370483, "learning_rate": 2.3576049052380497e-05, "loss": 0.0207, "step": 4760 }, { "epoch": 0.3271715228147334, "grad_norm": 0.9485260248184204, "learning_rate": 2.3573309754340748e-05, "loss": 0.0233, "step": 4761 }, { "epoch": 0.327240241891149, "grad_norm": 0.7377134561538696, "learning_rate": 2.3570570031596855e-05, "loss": 0.0284, "step": 4762 }, { "epoch": 0.3273089609675646, "grad_norm": 1.187445044517517, "learning_rate": 2.3567829884284542e-05, "loss": 0.031, "step": 4763 }, { "epoch": 0.3273776800439802, "grad_norm": 1.0210727453231812, "learning_rate": 2.3565089312539554e-05, "loss": 0.029, "step": 4764 }, { "epoch": 0.3274463991203958, "grad_norm": 1.4629615545272827, "learning_rate": 2.3562348316497656e-05, "loss": 0.0227, "step": 4765 }, { "epoch": 0.32751511819681145, "grad_norm": 2.394411087036133, "learning_rate": 2.3559606896294615e-05, "loss": 0.0339, "step": 4766 }, { "epoch": 0.32758383727322704, "grad_norm": 1.5612484216690063, "learning_rate": 2.3556865052066257e-05, "loss": 0.0219, "step": 4767 }, { "epoch": 0.3276525563496427, "grad_norm": 2.639925718307495, "learning_rate": 2.355412278394839e-05, "loss": 0.0463, "step": 4768 }, { "epoch": 0.3277212754260583, "grad_norm": 8.24560546875, "learning_rate": 2.3551380092076864e-05, "loss": 0.0393, "step": 4769 }, { "epoch": 0.32778999450247387, "grad_norm": 0.611299991607666, "learning_rate": 2.3548636976587553e-05, "loss": 0.0221, "step": 4770 }, { "epoch": 0.3278587135788895, "grad_norm": 1.0928102731704712, "learning_rate": 2.3545893437616342e-05, "loss": 0.0413, "step": 4771 }, { "epoch": 0.3279274326553051, "grad_norm": 1.1595458984375, "learning_rate": 2.3543149475299137e-05, "loss": 0.0392, "step": 4772 }, { "epoch": 0.32799615173172075, "grad_norm": 2.063852071762085, "learning_rate": 2.3540405089771875e-05, "loss": 0.0326, "step": 4773 }, { "epoch": 0.32806487080813634, "grad_norm": 3.2576167583465576, "learning_rate": 2.3537660281170507e-05, "loss": 0.0286, "step": 4774 }, { "epoch": 0.32813358988455193, "grad_norm": 0.631758451461792, "learning_rate": 2.353491504963099e-05, "loss": 0.0207, "step": 4775 }, { "epoch": 0.3282023089609676, "grad_norm": 0.9903351068496704, "learning_rate": 2.3532169395289334e-05, "loss": 0.0403, "step": 4776 }, { "epoch": 0.3282710280373832, "grad_norm": 5.109817981719971, "learning_rate": 2.352942331828155e-05, "loss": 0.0338, "step": 4777 }, { "epoch": 0.32833974711379876, "grad_norm": 1.4499247074127197, "learning_rate": 2.3526676818743667e-05, "loss": 0.0576, "step": 4778 }, { "epoch": 0.3284084661902144, "grad_norm": 0.9265241026878357, "learning_rate": 2.352392989681175e-05, "loss": 0.031, "step": 4779 }, { "epoch": 0.32847718526663, "grad_norm": 2.4984920024871826, "learning_rate": 2.3521182552621864e-05, "loss": 0.0539, "step": 4780 }, { "epoch": 0.32854590434304565, "grad_norm": 2.4969050884246826, "learning_rate": 2.3518434786310112e-05, "loss": 0.036, "step": 4781 }, { "epoch": 0.32861462341946124, "grad_norm": 0.9662367105484009, "learning_rate": 2.351568659801262e-05, "loss": 0.034, "step": 4782 }, { "epoch": 0.32868334249587683, "grad_norm": 1.0299065113067627, "learning_rate": 2.3512937987865518e-05, "loss": 0.0385, "step": 4783 }, { "epoch": 0.3287520615722925, "grad_norm": 1.088032603263855, "learning_rate": 2.3510188956004974e-05, "loss": 0.0274, "step": 4784 }, { "epoch": 0.32882078064870807, "grad_norm": 0.5889924168586731, "learning_rate": 2.3507439502567156e-05, "loss": 0.0322, "step": 4785 }, { "epoch": 0.3288894997251237, "grad_norm": 0.9928608536720276, "learning_rate": 2.3504689627688286e-05, "loss": 0.0284, "step": 4786 }, { "epoch": 0.3289582188015393, "grad_norm": 0.9759211540222168, "learning_rate": 2.3501939331504568e-05, "loss": 0.0365, "step": 4787 }, { "epoch": 0.3290269378779549, "grad_norm": 0.8559666275978088, "learning_rate": 2.349918861415226e-05, "loss": 0.0326, "step": 4788 }, { "epoch": 0.32909565695437054, "grad_norm": 0.5461317300796509, "learning_rate": 2.3496437475767624e-05, "loss": 0.0409, "step": 4789 }, { "epoch": 0.32916437603078613, "grad_norm": 0.4613878130912781, "learning_rate": 2.3493685916486936e-05, "loss": 0.0485, "step": 4790 }, { "epoch": 0.3292330951072018, "grad_norm": 1.7549529075622559, "learning_rate": 2.3490933936446517e-05, "loss": 0.04, "step": 4791 }, { "epoch": 0.32930181418361737, "grad_norm": 0.9751548171043396, "learning_rate": 2.3488181535782686e-05, "loss": 0.0373, "step": 4792 }, { "epoch": 0.32937053326003296, "grad_norm": 0.7901479005813599, "learning_rate": 2.3485428714631788e-05, "loss": 0.0235, "step": 4793 }, { "epoch": 0.3294392523364486, "grad_norm": 0.7134853601455688, "learning_rate": 2.34826754731302e-05, "loss": 0.0356, "step": 4794 }, { "epoch": 0.3295079714128642, "grad_norm": 0.5330653786659241, "learning_rate": 2.347992181141431e-05, "loss": 0.0325, "step": 4795 }, { "epoch": 0.32957669048927984, "grad_norm": 1.1761386394500732, "learning_rate": 2.3477167729620526e-05, "loss": 0.0421, "step": 4796 }, { "epoch": 0.32964540956569544, "grad_norm": 0.6172162294387817, "learning_rate": 2.347441322788528e-05, "loss": 0.0286, "step": 4797 }, { "epoch": 0.329714128642111, "grad_norm": 1.8552454710006714, "learning_rate": 2.3471658306345026e-05, "loss": 0.0224, "step": 4798 }, { "epoch": 0.3297828477185267, "grad_norm": 1.430881381034851, "learning_rate": 2.3468902965136235e-05, "loss": 0.0416, "step": 4799 }, { "epoch": 0.32985156679494226, "grad_norm": 1.043426513671875, "learning_rate": 2.3466147204395407e-05, "loss": 0.03, "step": 4800 }, { "epoch": 0.3299202858713579, "grad_norm": 1.0758994817733765, "learning_rate": 2.3463391024259046e-05, "loss": 0.0345, "step": 4801 }, { "epoch": 0.3299890049477735, "grad_norm": 0.7084684371948242, "learning_rate": 2.34606344248637e-05, "loss": 0.0333, "step": 4802 }, { "epoch": 0.3300577240241891, "grad_norm": 1.099277138710022, "learning_rate": 2.3457877406345915e-05, "loss": 0.0293, "step": 4803 }, { "epoch": 0.33012644310060474, "grad_norm": 0.8571318984031677, "learning_rate": 2.3455119968842276e-05, "loss": 0.0264, "step": 4804 }, { "epoch": 0.33019516217702033, "grad_norm": 2.9933130741119385, "learning_rate": 2.345236211248937e-05, "loss": 0.0254, "step": 4805 }, { "epoch": 0.330263881253436, "grad_norm": 1.4304801225662231, "learning_rate": 2.3449603837423824e-05, "loss": 0.0364, "step": 4806 }, { "epoch": 0.33033260032985157, "grad_norm": 0.9089800119400024, "learning_rate": 2.344684514378228e-05, "loss": 0.025, "step": 4807 }, { "epoch": 0.33040131940626716, "grad_norm": 1.3470609188079834, "learning_rate": 2.3444086031701394e-05, "loss": 0.0294, "step": 4808 }, { "epoch": 0.3304700384826828, "grad_norm": 0.756610631942749, "learning_rate": 2.344132650131784e-05, "loss": 0.0241, "step": 4809 }, { "epoch": 0.3305387575590984, "grad_norm": 1.457380771636963, "learning_rate": 2.343856655276833e-05, "loss": 0.0402, "step": 4810 }, { "epoch": 0.33060747663551404, "grad_norm": 0.6341134905815125, "learning_rate": 2.3435806186189585e-05, "loss": 0.0388, "step": 4811 }, { "epoch": 0.33067619571192963, "grad_norm": 1.031564712524414, "learning_rate": 2.3433045401718335e-05, "loss": 0.0369, "step": 4812 }, { "epoch": 0.3307449147883452, "grad_norm": 0.32025155425071716, "learning_rate": 2.343028419949136e-05, "loss": 0.027, "step": 4813 }, { "epoch": 0.33081363386476087, "grad_norm": 1.3197519779205322, "learning_rate": 2.342752257964544e-05, "loss": 0.0333, "step": 4814 }, { "epoch": 0.33088235294117646, "grad_norm": 1.203812599182129, "learning_rate": 2.342476054231737e-05, "loss": 0.0329, "step": 4815 }, { "epoch": 0.3309510720175921, "grad_norm": 1.8655749559402466, "learning_rate": 2.342199808764399e-05, "loss": 0.0462, "step": 4816 }, { "epoch": 0.3310197910940077, "grad_norm": 0.8520341515541077, "learning_rate": 2.341923521576214e-05, "loss": 0.0319, "step": 4817 }, { "epoch": 0.3310885101704233, "grad_norm": 0.6222615838050842, "learning_rate": 2.341647192680868e-05, "loss": 0.0267, "step": 4818 }, { "epoch": 0.33115722924683894, "grad_norm": 0.5049178600311279, "learning_rate": 2.341370822092051e-05, "loss": 0.0198, "step": 4819 }, { "epoch": 0.3312259483232545, "grad_norm": 0.729983925819397, "learning_rate": 2.341094409823453e-05, "loss": 0.0295, "step": 4820 }, { "epoch": 0.3312946673996702, "grad_norm": 1.2631852626800537, "learning_rate": 2.3408179558887673e-05, "loss": 0.0312, "step": 4821 }, { "epoch": 0.33136338647608576, "grad_norm": 0.9223511219024658, "learning_rate": 2.340541460301689e-05, "loss": 0.0527, "step": 4822 }, { "epoch": 0.33143210555250135, "grad_norm": 1.47333562374115, "learning_rate": 2.340264923075914e-05, "loss": 0.0292, "step": 4823 }, { "epoch": 0.331500824628917, "grad_norm": 0.6360145807266235, "learning_rate": 2.339988344225143e-05, "loss": 0.0286, "step": 4824 }, { "epoch": 0.3315695437053326, "grad_norm": 0.9297571778297424, "learning_rate": 2.3397117237630763e-05, "loss": 0.0285, "step": 4825 }, { "epoch": 0.33163826278174824, "grad_norm": 0.7106973528862, "learning_rate": 2.339435061703417e-05, "loss": 0.0334, "step": 4826 }, { "epoch": 0.33170698185816383, "grad_norm": 0.7798745632171631, "learning_rate": 2.33915835805987e-05, "loss": 0.0307, "step": 4827 }, { "epoch": 0.3317757009345794, "grad_norm": 2.0804498195648193, "learning_rate": 2.338881612846144e-05, "loss": 0.0324, "step": 4828 }, { "epoch": 0.33184442001099507, "grad_norm": 1.2140181064605713, "learning_rate": 2.3386048260759473e-05, "loss": 0.0394, "step": 4829 }, { "epoch": 0.33191313908741066, "grad_norm": 5.53696346282959, "learning_rate": 2.3383279977629916e-05, "loss": 0.0209, "step": 4830 }, { "epoch": 0.3319818581638263, "grad_norm": 0.43822821974754333, "learning_rate": 2.3380511279209907e-05, "loss": 0.0235, "step": 4831 }, { "epoch": 0.3320505772402419, "grad_norm": 0.7061882615089417, "learning_rate": 2.3377742165636595e-05, "loss": 0.0265, "step": 4832 }, { "epoch": 0.3321192963166575, "grad_norm": 0.5520374774932861, "learning_rate": 2.337497263704716e-05, "loss": 0.0435, "step": 4833 }, { "epoch": 0.33218801539307313, "grad_norm": 1.319799780845642, "learning_rate": 2.33722026935788e-05, "loss": 0.046, "step": 4834 }, { "epoch": 0.3322567344694887, "grad_norm": 0.6607989072799683, "learning_rate": 2.336943233536873e-05, "loss": 0.0351, "step": 4835 }, { "epoch": 0.33232545354590437, "grad_norm": 1.6351158618927002, "learning_rate": 2.336666156255419e-05, "loss": 0.0498, "step": 4836 }, { "epoch": 0.33239417262231996, "grad_norm": 0.9724599123001099, "learning_rate": 2.3363890375272436e-05, "loss": 0.0293, "step": 4837 }, { "epoch": 0.33246289169873555, "grad_norm": 1.5298640727996826, "learning_rate": 2.3361118773660753e-05, "loss": 0.0358, "step": 4838 }, { "epoch": 0.3325316107751512, "grad_norm": 2.1132118701934814, "learning_rate": 2.335834675785643e-05, "loss": 0.038, "step": 4839 }, { "epoch": 0.3326003298515668, "grad_norm": 0.8016104698181152, "learning_rate": 2.3355574327996795e-05, "loss": 0.0336, "step": 4840 }, { "epoch": 0.3326690489279824, "grad_norm": 0.8292073607444763, "learning_rate": 2.3352801484219185e-05, "loss": 0.0389, "step": 4841 }, { "epoch": 0.332737768004398, "grad_norm": 1.4759844541549683, "learning_rate": 2.3350028226660963e-05, "loss": 0.0291, "step": 4842 }, { "epoch": 0.3328064870808136, "grad_norm": 0.9476250410079956, "learning_rate": 2.3347254555459507e-05, "loss": 0.0283, "step": 4843 }, { "epoch": 0.33287520615722926, "grad_norm": 1.225684404373169, "learning_rate": 2.3344480470752227e-05, "loss": 0.039, "step": 4844 }, { "epoch": 0.33294392523364486, "grad_norm": 0.8590390682220459, "learning_rate": 2.3341705972676533e-05, "loss": 0.025, "step": 4845 }, { "epoch": 0.33301264431006045, "grad_norm": 1.0779696702957153, "learning_rate": 2.333893106136988e-05, "loss": 0.0305, "step": 4846 }, { "epoch": 0.3330813633864761, "grad_norm": 0.8464577198028564, "learning_rate": 2.333615573696973e-05, "loss": 0.0446, "step": 4847 }, { "epoch": 0.3331500824628917, "grad_norm": 1.7926273345947266, "learning_rate": 2.333337999961355e-05, "loss": 0.0405, "step": 4848 }, { "epoch": 0.33321880153930733, "grad_norm": 0.7190973162651062, "learning_rate": 2.333060384943887e-05, "loss": 0.0203, "step": 4849 }, { "epoch": 0.3332875206157229, "grad_norm": 0.8627727031707764, "learning_rate": 2.33278272865832e-05, "loss": 0.0257, "step": 4850 }, { "epoch": 0.3333562396921385, "grad_norm": 0.9577804803848267, "learning_rate": 2.3325050311184083e-05, "loss": 0.0254, "step": 4851 }, { "epoch": 0.33342495876855416, "grad_norm": 1.4061869382858276, "learning_rate": 2.3322272923379092e-05, "loss": 0.0296, "step": 4852 }, { "epoch": 0.33349367784496975, "grad_norm": 1.0624310970306396, "learning_rate": 2.331949512330581e-05, "loss": 0.0386, "step": 4853 }, { "epoch": 0.3335623969213854, "grad_norm": 1.0843349695205688, "learning_rate": 2.3316716911101845e-05, "loss": 0.0343, "step": 4854 }, { "epoch": 0.333631115997801, "grad_norm": 0.9139512777328491, "learning_rate": 2.3313938286904822e-05, "loss": 0.0406, "step": 4855 }, { "epoch": 0.3336998350742166, "grad_norm": 0.7434367537498474, "learning_rate": 2.331115925085239e-05, "loss": 0.0257, "step": 4856 }, { "epoch": 0.3337685541506322, "grad_norm": 0.9853392839431763, "learning_rate": 2.330837980308221e-05, "loss": 0.0237, "step": 4857 }, { "epoch": 0.3338372732270478, "grad_norm": 0.7911674976348877, "learning_rate": 2.3305599943731983e-05, "loss": 0.0265, "step": 4858 }, { "epoch": 0.33390599230346346, "grad_norm": 1.9785075187683105, "learning_rate": 2.3302819672939408e-05, "loss": 0.0205, "step": 4859 }, { "epoch": 0.33397471137987905, "grad_norm": 1.4353643655776978, "learning_rate": 2.3300038990842212e-05, "loss": 0.0468, "step": 4860 }, { "epoch": 0.33404343045629464, "grad_norm": 1.6668567657470703, "learning_rate": 2.3297257897578155e-05, "loss": 0.0423, "step": 4861 }, { "epoch": 0.3341121495327103, "grad_norm": 1.2267664670944214, "learning_rate": 2.3294476393285e-05, "loss": 0.0326, "step": 4862 }, { "epoch": 0.3341808686091259, "grad_norm": 0.8712280988693237, "learning_rate": 2.3291694478100533e-05, "loss": 0.039, "step": 4863 }, { "epoch": 0.3342495876855415, "grad_norm": 0.703893780708313, "learning_rate": 2.328891215216257e-05, "loss": 0.031, "step": 4864 }, { "epoch": 0.3343183067619571, "grad_norm": 0.5530781149864197, "learning_rate": 2.3286129415608943e-05, "loss": 0.0378, "step": 4865 }, { "epoch": 0.3343870258383727, "grad_norm": 1.3740804195404053, "learning_rate": 2.3283346268577495e-05, "loss": 0.041, "step": 4866 }, { "epoch": 0.33445574491478836, "grad_norm": 0.8728175759315491, "learning_rate": 2.3280562711206106e-05, "loss": 0.03, "step": 4867 }, { "epoch": 0.33452446399120395, "grad_norm": 1.1615252494812012, "learning_rate": 2.3277778743632666e-05, "loss": 0.0395, "step": 4868 }, { "epoch": 0.3345931830676196, "grad_norm": 1.398271918296814, "learning_rate": 2.3274994365995083e-05, "loss": 0.0398, "step": 4869 }, { "epoch": 0.3346619021440352, "grad_norm": 1.0904191732406616, "learning_rate": 2.327220957843129e-05, "loss": 0.039, "step": 4870 }, { "epoch": 0.3347306212204508, "grad_norm": 1.5163298845291138, "learning_rate": 2.3269424381079246e-05, "loss": 0.0394, "step": 4871 }, { "epoch": 0.3347993402968664, "grad_norm": 1.1414377689361572, "learning_rate": 2.3266638774076914e-05, "loss": 0.0297, "step": 4872 }, { "epoch": 0.334868059373282, "grad_norm": 2.137805223464966, "learning_rate": 2.32638527575623e-05, "loss": 0.0511, "step": 4873 }, { "epoch": 0.33493677844969766, "grad_norm": 1.178149700164795, "learning_rate": 2.3261066331673403e-05, "loss": 0.0252, "step": 4874 }, { "epoch": 0.33500549752611325, "grad_norm": 0.8873534798622131, "learning_rate": 2.3258279496548267e-05, "loss": 0.0312, "step": 4875 }, { "epoch": 0.33507421660252884, "grad_norm": 1.6282713413238525, "learning_rate": 2.325549225232494e-05, "loss": 0.0334, "step": 4876 }, { "epoch": 0.3351429356789445, "grad_norm": 1.3137974739074707, "learning_rate": 2.3252704599141504e-05, "loss": 0.0286, "step": 4877 }, { "epoch": 0.3352116547553601, "grad_norm": 0.7005689144134521, "learning_rate": 2.3249916537136046e-05, "loss": 0.0325, "step": 4878 }, { "epoch": 0.3352803738317757, "grad_norm": 0.7461518049240112, "learning_rate": 2.3247128066446685e-05, "loss": 0.0298, "step": 4879 }, { "epoch": 0.3353490929081913, "grad_norm": 0.6212315559387207, "learning_rate": 2.3244339187211558e-05, "loss": 0.0313, "step": 4880 }, { "epoch": 0.3354178119846069, "grad_norm": 1.2172610759735107, "learning_rate": 2.3241549899568812e-05, "loss": 0.0477, "step": 4881 }, { "epoch": 0.33548653106102255, "grad_norm": 1.6598440408706665, "learning_rate": 2.323876020365663e-05, "loss": 0.0297, "step": 4882 }, { "epoch": 0.33555525013743814, "grad_norm": 0.9257919788360596, "learning_rate": 2.3235970099613206e-05, "loss": 0.0352, "step": 4883 }, { "epoch": 0.3356239692138538, "grad_norm": 1.0182973146438599, "learning_rate": 2.3233179587576756e-05, "loss": 0.0321, "step": 4884 }, { "epoch": 0.3356926882902694, "grad_norm": 0.4619996249675751, "learning_rate": 2.3230388667685516e-05, "loss": 0.03, "step": 4885 }, { "epoch": 0.33576140736668497, "grad_norm": 1.0998547077178955, "learning_rate": 2.322759734007774e-05, "loss": 0.0459, "step": 4886 }, { "epoch": 0.3358301264431006, "grad_norm": 1.6732616424560547, "learning_rate": 2.3224805604891712e-05, "loss": 0.0258, "step": 4887 }, { "epoch": 0.3358988455195162, "grad_norm": 0.5458085536956787, "learning_rate": 2.322201346226572e-05, "loss": 0.0237, "step": 4888 }, { "epoch": 0.33596756459593186, "grad_norm": 0.8232007026672363, "learning_rate": 2.3219220912338087e-05, "loss": 0.0315, "step": 4889 }, { "epoch": 0.33603628367234745, "grad_norm": 0.6676656603813171, "learning_rate": 2.3216427955247146e-05, "loss": 0.0212, "step": 4890 }, { "epoch": 0.33610500274876304, "grad_norm": 0.9747385382652283, "learning_rate": 2.321363459113126e-05, "loss": 0.0434, "step": 4891 }, { "epoch": 0.3361737218251787, "grad_norm": 0.8734063506126404, "learning_rate": 2.3210840820128796e-05, "loss": 0.0461, "step": 4892 }, { "epoch": 0.3362424409015943, "grad_norm": 0.8730322122573853, "learning_rate": 2.3208046642378164e-05, "loss": 0.0281, "step": 4893 }, { "epoch": 0.3363111599780099, "grad_norm": 0.6690266728401184, "learning_rate": 2.3205252058017774e-05, "loss": 0.0373, "step": 4894 }, { "epoch": 0.3363798790544255, "grad_norm": 1.0505865812301636, "learning_rate": 2.3202457067186066e-05, "loss": 0.0356, "step": 4895 }, { "epoch": 0.3364485981308411, "grad_norm": 0.8385086059570312, "learning_rate": 2.3199661670021497e-05, "loss": 0.0377, "step": 4896 }, { "epoch": 0.33651731720725675, "grad_norm": 1.5867185592651367, "learning_rate": 2.3196865866662552e-05, "loss": 0.0356, "step": 4897 }, { "epoch": 0.33658603628367234, "grad_norm": 1.1006869077682495, "learning_rate": 2.319406965724772e-05, "loss": 0.0297, "step": 4898 }, { "epoch": 0.336654755360088, "grad_norm": 1.1515332460403442, "learning_rate": 2.3191273041915526e-05, "loss": 0.0406, "step": 4899 }, { "epoch": 0.3367234744365036, "grad_norm": 0.7502903938293457, "learning_rate": 2.3188476020804506e-05, "loss": 0.0313, "step": 4900 }, { "epoch": 0.33679219351291917, "grad_norm": 0.9063253402709961, "learning_rate": 2.3185678594053216e-05, "loss": 0.0269, "step": 4901 }, { "epoch": 0.3368609125893348, "grad_norm": 0.6560551524162292, "learning_rate": 2.3182880761800236e-05, "loss": 0.0306, "step": 4902 }, { "epoch": 0.3369296316657504, "grad_norm": 0.7265577912330627, "learning_rate": 2.318008252418417e-05, "loss": 0.0277, "step": 4903 }, { "epoch": 0.33699835074216605, "grad_norm": 0.7882918119430542, "learning_rate": 2.3177283881343632e-05, "loss": 0.0286, "step": 4904 }, { "epoch": 0.33706706981858164, "grad_norm": 1.1626425981521606, "learning_rate": 2.3174484833417265e-05, "loss": 0.0558, "step": 4905 }, { "epoch": 0.33713578889499723, "grad_norm": 1.317177653312683, "learning_rate": 2.3171685380543723e-05, "loss": 0.0459, "step": 4906 }, { "epoch": 0.3372045079714129, "grad_norm": 1.2206957340240479, "learning_rate": 2.316888552286169e-05, "loss": 0.0405, "step": 4907 }, { "epoch": 0.33727322704782847, "grad_norm": 0.7770478129386902, "learning_rate": 2.3166085260509856e-05, "loss": 0.0376, "step": 4908 }, { "epoch": 0.33734194612424406, "grad_norm": 0.6316424012184143, "learning_rate": 2.3163284593626953e-05, "loss": 0.0243, "step": 4909 }, { "epoch": 0.3374106652006597, "grad_norm": 1.3271986246109009, "learning_rate": 2.3160483522351713e-05, "loss": 0.0448, "step": 4910 }, { "epoch": 0.3374793842770753, "grad_norm": 1.9219738245010376, "learning_rate": 2.3157682046822895e-05, "loss": 0.0323, "step": 4911 }, { "epoch": 0.33754810335349095, "grad_norm": 0.9670530557632446, "learning_rate": 2.3154880167179282e-05, "loss": 0.0343, "step": 4912 }, { "epoch": 0.33761682242990654, "grad_norm": 0.8281570076942444, "learning_rate": 2.315207788355967e-05, "loss": 0.0357, "step": 4913 }, { "epoch": 0.33768554150632213, "grad_norm": 1.3105677366256714, "learning_rate": 2.3149275196102883e-05, "loss": 0.0371, "step": 4914 }, { "epoch": 0.3377542605827378, "grad_norm": 0.7367860078811646, "learning_rate": 2.314647210494775e-05, "loss": 0.0339, "step": 4915 }, { "epoch": 0.33782297965915337, "grad_norm": 1.6730318069458008, "learning_rate": 2.3143668610233143e-05, "loss": 0.041, "step": 4916 }, { "epoch": 0.337891698735569, "grad_norm": 1.0215308666229248, "learning_rate": 2.314086471209793e-05, "loss": 0.0326, "step": 4917 }, { "epoch": 0.3379604178119846, "grad_norm": 0.5796467065811157, "learning_rate": 2.3138060410681023e-05, "loss": 0.0303, "step": 4918 }, { "epoch": 0.3380291368884002, "grad_norm": 0.9023404717445374, "learning_rate": 2.313525570612133e-05, "loss": 0.0185, "step": 4919 }, { "epoch": 0.33809785596481584, "grad_norm": 1.5163594484329224, "learning_rate": 2.3132450598557796e-05, "loss": 0.0236, "step": 4920 }, { "epoch": 0.33816657504123143, "grad_norm": 0.6421452164649963, "learning_rate": 2.312964508812938e-05, "loss": 0.034, "step": 4921 }, { "epoch": 0.3382352941176471, "grad_norm": 0.9805554151535034, "learning_rate": 2.3126839174975062e-05, "loss": 0.0288, "step": 4922 }, { "epoch": 0.33830401319406267, "grad_norm": 0.6391741037368774, "learning_rate": 2.3124032859233832e-05, "loss": 0.0311, "step": 4923 }, { "epoch": 0.33837273227047826, "grad_norm": 1.3718703985214233, "learning_rate": 2.3121226141044724e-05, "loss": 0.0363, "step": 4924 }, { "epoch": 0.3384414513468939, "grad_norm": 2.170156717300415, "learning_rate": 2.3118419020546767e-05, "loss": 0.0528, "step": 4925 }, { "epoch": 0.3385101704233095, "grad_norm": 0.7942155003547668, "learning_rate": 2.311561149787902e-05, "loss": 0.0565, "step": 4926 }, { "epoch": 0.33857888949972514, "grad_norm": 0.7283538579940796, "learning_rate": 2.3112803573180566e-05, "loss": 0.0376, "step": 4927 }, { "epoch": 0.33864760857614074, "grad_norm": 0.756744921207428, "learning_rate": 2.3109995246590505e-05, "loss": 0.033, "step": 4928 }, { "epoch": 0.3387163276525563, "grad_norm": 1.9919922351837158, "learning_rate": 2.3107186518247947e-05, "loss": 0.0364, "step": 4929 }, { "epoch": 0.338785046728972, "grad_norm": 1.5493937730789185, "learning_rate": 2.3104377388292045e-05, "loss": 0.0358, "step": 4930 }, { "epoch": 0.33885376580538756, "grad_norm": 0.6348230838775635, "learning_rate": 2.310156785686194e-05, "loss": 0.0202, "step": 4931 }, { "epoch": 0.3389224848818032, "grad_norm": 0.6862196922302246, "learning_rate": 2.3098757924096824e-05, "loss": 0.028, "step": 4932 }, { "epoch": 0.3389912039582188, "grad_norm": 0.5208510160446167, "learning_rate": 2.3095947590135893e-05, "loss": 0.0279, "step": 4933 }, { "epoch": 0.3390599230346344, "grad_norm": 0.8903704881668091, "learning_rate": 2.309313685511836e-05, "loss": 0.0332, "step": 4934 }, { "epoch": 0.33912864211105004, "grad_norm": 0.5192200541496277, "learning_rate": 2.3090325719183464e-05, "loss": 0.0293, "step": 4935 }, { "epoch": 0.33919736118746563, "grad_norm": 1.2439650297164917, "learning_rate": 2.308751418247047e-05, "loss": 0.0399, "step": 4936 }, { "epoch": 0.3392660802638813, "grad_norm": 0.36435171961784363, "learning_rate": 2.3084702245118645e-05, "loss": 0.0202, "step": 4937 }, { "epoch": 0.33933479934029687, "grad_norm": 0.9025247693061829, "learning_rate": 2.3081889907267293e-05, "loss": 0.0323, "step": 4938 }, { "epoch": 0.33940351841671246, "grad_norm": 0.8163657784461975, "learning_rate": 2.3079077169055736e-05, "loss": 0.0359, "step": 4939 }, { "epoch": 0.3394722374931281, "grad_norm": 1.0872079133987427, "learning_rate": 2.3076264030623303e-05, "loss": 0.0383, "step": 4940 }, { "epoch": 0.3395409565695437, "grad_norm": 1.9535350799560547, "learning_rate": 2.3073450492109353e-05, "loss": 0.0359, "step": 4941 }, { "epoch": 0.33960967564595934, "grad_norm": 0.5361445546150208, "learning_rate": 2.3070636553653265e-05, "loss": 0.0396, "step": 4942 }, { "epoch": 0.33967839472237493, "grad_norm": 1.3287007808685303, "learning_rate": 2.3067822215394434e-05, "loss": 0.0346, "step": 4943 }, { "epoch": 0.3397471137987905, "grad_norm": 0.7057675719261169, "learning_rate": 2.3065007477472272e-05, "loss": 0.0337, "step": 4944 }, { "epoch": 0.33981583287520617, "grad_norm": 0.6201030015945435, "learning_rate": 2.306219234002623e-05, "loss": 0.0299, "step": 4945 }, { "epoch": 0.33988455195162176, "grad_norm": 0.8205596804618835, "learning_rate": 2.3059376803195752e-05, "loss": 0.0468, "step": 4946 }, { "epoch": 0.3399532710280374, "grad_norm": 0.8916398882865906, "learning_rate": 2.3056560867120313e-05, "loss": 0.0264, "step": 4947 }, { "epoch": 0.340021990104453, "grad_norm": 0.8175399303436279, "learning_rate": 2.3053744531939418e-05, "loss": 0.0328, "step": 4948 }, { "epoch": 0.3400907091808686, "grad_norm": 0.8676018714904785, "learning_rate": 2.3050927797792574e-05, "loss": 0.0313, "step": 4949 }, { "epoch": 0.34015942825728424, "grad_norm": 0.4064609110355377, "learning_rate": 2.3048110664819316e-05, "loss": 0.0347, "step": 4950 }, { "epoch": 0.3402281473336998, "grad_norm": 0.7072848081588745, "learning_rate": 2.3045293133159202e-05, "loss": 0.0284, "step": 4951 }, { "epoch": 0.3402968664101155, "grad_norm": 1.0109226703643799, "learning_rate": 2.3042475202951808e-05, "loss": 0.0433, "step": 4952 }, { "epoch": 0.34036558548653106, "grad_norm": 1.1709297895431519, "learning_rate": 2.3039656874336728e-05, "loss": 0.0372, "step": 4953 }, { "epoch": 0.34043430456294665, "grad_norm": 0.47993019223213196, "learning_rate": 2.3036838147453576e-05, "loss": 0.0306, "step": 4954 }, { "epoch": 0.3405030236393623, "grad_norm": 0.9276830554008484, "learning_rate": 2.3034019022441984e-05, "loss": 0.0281, "step": 4955 }, { "epoch": 0.3405717427157779, "grad_norm": 0.5949503779411316, "learning_rate": 2.3031199499441605e-05, "loss": 0.0304, "step": 4956 }, { "epoch": 0.34064046179219354, "grad_norm": 1.5017262697219849, "learning_rate": 2.302837957859212e-05, "loss": 0.0278, "step": 4957 }, { "epoch": 0.34070918086860913, "grad_norm": 0.4933834671974182, "learning_rate": 2.3025559260033212e-05, "loss": 0.0235, "step": 4958 }, { "epoch": 0.3407778999450247, "grad_norm": 1.1305716037750244, "learning_rate": 2.3022738543904595e-05, "loss": 0.0358, "step": 4959 }, { "epoch": 0.34084661902144037, "grad_norm": 0.7242923378944397, "learning_rate": 2.3019917430346008e-05, "loss": 0.0348, "step": 4960 }, { "epoch": 0.34091533809785596, "grad_norm": 1.0236258506774902, "learning_rate": 2.3017095919497202e-05, "loss": 0.0372, "step": 4961 }, { "epoch": 0.3409840571742716, "grad_norm": 0.9230493903160095, "learning_rate": 2.3014274011497943e-05, "loss": 0.0377, "step": 4962 }, { "epoch": 0.3410527762506872, "grad_norm": 1.4651966094970703, "learning_rate": 2.3011451706488027e-05, "loss": 0.0464, "step": 4963 }, { "epoch": 0.3411214953271028, "grad_norm": 1.087470293045044, "learning_rate": 2.3008629004607264e-05, "loss": 0.0361, "step": 4964 }, { "epoch": 0.34119021440351843, "grad_norm": 0.638279139995575, "learning_rate": 2.300580590599549e-05, "loss": 0.0236, "step": 4965 }, { "epoch": 0.341258933479934, "grad_norm": 0.6208494901657104, "learning_rate": 2.300298241079255e-05, "loss": 0.0324, "step": 4966 }, { "epoch": 0.34132765255634967, "grad_norm": 1.75697660446167, "learning_rate": 2.3000158519138307e-05, "loss": 0.0266, "step": 4967 }, { "epoch": 0.34139637163276526, "grad_norm": 0.8791819214820862, "learning_rate": 2.2997334231172663e-05, "loss": 0.0239, "step": 4968 }, { "epoch": 0.34146509070918085, "grad_norm": 2.25395131111145, "learning_rate": 2.2994509547035524e-05, "loss": 0.0351, "step": 4969 }, { "epoch": 0.3415338097855965, "grad_norm": 0.8503202199935913, "learning_rate": 2.299168446686682e-05, "loss": 0.0301, "step": 4970 }, { "epoch": 0.3416025288620121, "grad_norm": 0.9504479169845581, "learning_rate": 2.298885899080649e-05, "loss": 0.0299, "step": 4971 }, { "epoch": 0.3416712479384277, "grad_norm": 0.8577936887741089, "learning_rate": 2.2986033118994516e-05, "loss": 0.0377, "step": 4972 }, { "epoch": 0.3417399670148433, "grad_norm": 0.9091536998748779, "learning_rate": 2.298320685157088e-05, "loss": 0.0333, "step": 4973 }, { "epoch": 0.3418086860912589, "grad_norm": 0.7950015664100647, "learning_rate": 2.298038018867559e-05, "loss": 0.0261, "step": 4974 }, { "epoch": 0.34187740516767456, "grad_norm": 0.8348479270935059, "learning_rate": 2.2977553130448667e-05, "loss": 0.0362, "step": 4975 }, { "epoch": 0.34194612424409015, "grad_norm": 1.300349235534668, "learning_rate": 2.297472567703017e-05, "loss": 0.0391, "step": 4976 }, { "epoch": 0.34201484332050575, "grad_norm": 1.515656590461731, "learning_rate": 2.2971897828560153e-05, "loss": 0.0431, "step": 4977 }, { "epoch": 0.3420835623969214, "grad_norm": 0.7569465637207031, "learning_rate": 2.2969069585178707e-05, "loss": 0.0328, "step": 4978 }, { "epoch": 0.342152281473337, "grad_norm": 3.4990932941436768, "learning_rate": 2.2966240947025942e-05, "loss": 0.0507, "step": 4979 }, { "epoch": 0.34222100054975263, "grad_norm": 7.244800567626953, "learning_rate": 2.296341191424197e-05, "loss": 0.0327, "step": 4980 }, { "epoch": 0.3422897196261682, "grad_norm": 1.0449254512786865, "learning_rate": 2.2960582486966952e-05, "loss": 0.0612, "step": 4981 }, { "epoch": 0.3423584387025838, "grad_norm": 0.754403829574585, "learning_rate": 2.2957752665341044e-05, "loss": 0.0316, "step": 4982 }, { "epoch": 0.34242715777899946, "grad_norm": 0.8294186592102051, "learning_rate": 2.295492244950443e-05, "loss": 0.0267, "step": 4983 }, { "epoch": 0.34249587685541505, "grad_norm": 1.7438735961914062, "learning_rate": 2.2952091839597308e-05, "loss": 0.04, "step": 4984 }, { "epoch": 0.3425645959318307, "grad_norm": 0.6185347437858582, "learning_rate": 2.294926083575991e-05, "loss": 0.0382, "step": 4985 }, { "epoch": 0.3426333150082463, "grad_norm": 1.360110878944397, "learning_rate": 2.2946429438132465e-05, "loss": 0.0464, "step": 4986 }, { "epoch": 0.3427020340846619, "grad_norm": 1.2878683805465698, "learning_rate": 2.294359764685525e-05, "loss": 0.0217, "step": 4987 }, { "epoch": 0.3427707531610775, "grad_norm": 0.7622864842414856, "learning_rate": 2.2940765462068534e-05, "loss": 0.0419, "step": 4988 }, { "epoch": 0.3428394722374931, "grad_norm": 1.5429198741912842, "learning_rate": 2.2937932883912623e-05, "loss": 0.0353, "step": 4989 }, { "epoch": 0.34290819131390876, "grad_norm": 1.266215205192566, "learning_rate": 2.2935099912527843e-05, "loss": 0.0367, "step": 4990 }, { "epoch": 0.34297691039032435, "grad_norm": 0.828387975692749, "learning_rate": 2.2932266548054523e-05, "loss": 0.0271, "step": 4991 }, { "epoch": 0.34304562946673994, "grad_norm": 0.5520569682121277, "learning_rate": 2.2929432790633023e-05, "loss": 0.0326, "step": 4992 }, { "epoch": 0.3431143485431556, "grad_norm": 0.6425572037696838, "learning_rate": 2.2926598640403728e-05, "loss": 0.044, "step": 4993 }, { "epoch": 0.3431830676195712, "grad_norm": 2.0079894065856934, "learning_rate": 2.2923764097507035e-05, "loss": 0.0482, "step": 4994 }, { "epoch": 0.3432517866959868, "grad_norm": 1.4251431226730347, "learning_rate": 2.2920929162083355e-05, "loss": 0.0508, "step": 4995 }, { "epoch": 0.3433205057724024, "grad_norm": 1.5468946695327759, "learning_rate": 2.2918093834273132e-05, "loss": 0.0424, "step": 4996 }, { "epoch": 0.343389224848818, "grad_norm": 0.8423623442649841, "learning_rate": 2.291525811421682e-05, "loss": 0.0341, "step": 4997 }, { "epoch": 0.34345794392523366, "grad_norm": 0.988700270652771, "learning_rate": 2.291242200205489e-05, "loss": 0.0368, "step": 4998 }, { "epoch": 0.34352666300164925, "grad_norm": 0.7265415191650391, "learning_rate": 2.2909585497927844e-05, "loss": 0.0356, "step": 4999 }, { "epoch": 0.3435953820780649, "grad_norm": 0.9278960824012756, "learning_rate": 2.2906748601976192e-05, "loss": 0.04, "step": 5000 }, { "epoch": 0.3436641011544805, "grad_norm": 0.8570512533187866, "learning_rate": 2.2903911314340472e-05, "loss": 0.0233, "step": 5001 }, { "epoch": 0.3437328202308961, "grad_norm": 1.0701779127120972, "learning_rate": 2.290107363516123e-05, "loss": 0.0241, "step": 5002 }, { "epoch": 0.3438015393073117, "grad_norm": 1.144607424736023, "learning_rate": 2.2898235564579053e-05, "loss": 0.0306, "step": 5003 }, { "epoch": 0.3438702583837273, "grad_norm": 1.109315276145935, "learning_rate": 2.2895397102734518e-05, "loss": 0.0346, "step": 5004 }, { "epoch": 0.34393897746014296, "grad_norm": 0.6904748678207397, "learning_rate": 2.289255824976825e-05, "loss": 0.0365, "step": 5005 }, { "epoch": 0.34400769653655855, "grad_norm": 0.7379464507102966, "learning_rate": 2.2889719005820865e-05, "loss": 0.0387, "step": 5006 }, { "epoch": 0.34407641561297414, "grad_norm": 0.8493260145187378, "learning_rate": 2.2886879371033023e-05, "loss": 0.0251, "step": 5007 }, { "epoch": 0.3441451346893898, "grad_norm": 1.1965227127075195, "learning_rate": 2.2884039345545393e-05, "loss": 0.0334, "step": 5008 }, { "epoch": 0.3442138537658054, "grad_norm": 0.9940168857574463, "learning_rate": 2.288119892949866e-05, "loss": 0.0405, "step": 5009 }, { "epoch": 0.344282572842221, "grad_norm": 1.1513397693634033, "learning_rate": 2.2878358123033534e-05, "loss": 0.0458, "step": 5010 }, { "epoch": 0.3443512919186366, "grad_norm": 2.01460599899292, "learning_rate": 2.2875516926290747e-05, "loss": 0.0257, "step": 5011 }, { "epoch": 0.3444200109950522, "grad_norm": 0.8850059509277344, "learning_rate": 2.2872675339411046e-05, "loss": 0.0261, "step": 5012 }, { "epoch": 0.34448873007146785, "grad_norm": 1.0464117527008057, "learning_rate": 2.2869833362535186e-05, "loss": 0.0407, "step": 5013 }, { "epoch": 0.34455744914788344, "grad_norm": 1.311571478843689, "learning_rate": 2.2866990995803967e-05, "loss": 0.0349, "step": 5014 }, { "epoch": 0.3446261682242991, "grad_norm": 0.9107946157455444, "learning_rate": 2.286414823935819e-05, "loss": 0.0311, "step": 5015 }, { "epoch": 0.3446948873007147, "grad_norm": 1.2418615818023682, "learning_rate": 2.286130509333867e-05, "loss": 0.0253, "step": 5016 }, { "epoch": 0.34476360637713027, "grad_norm": 0.8143390417098999, "learning_rate": 2.285846155788626e-05, "loss": 0.0402, "step": 5017 }, { "epoch": 0.3448323254535459, "grad_norm": 1.086355447769165, "learning_rate": 2.2855617633141826e-05, "loss": 0.0412, "step": 5018 }, { "epoch": 0.3449010445299615, "grad_norm": 1.5283886194229126, "learning_rate": 2.2852773319246242e-05, "loss": 0.0325, "step": 5019 }, { "epoch": 0.34496976360637716, "grad_norm": 1.3340612649917603, "learning_rate": 2.284992861634041e-05, "loss": 0.0455, "step": 5020 }, { "epoch": 0.34503848268279275, "grad_norm": 0.6182231903076172, "learning_rate": 2.2847083524565255e-05, "loss": 0.0355, "step": 5021 }, { "epoch": 0.34510720175920834, "grad_norm": 1.1501388549804688, "learning_rate": 2.2844238044061713e-05, "loss": 0.0484, "step": 5022 }, { "epoch": 0.345175920835624, "grad_norm": 1.049347996711731, "learning_rate": 2.2841392174970748e-05, "loss": 0.0351, "step": 5023 }, { "epoch": 0.3452446399120396, "grad_norm": 1.6862578392028809, "learning_rate": 2.2838545917433337e-05, "loss": 0.0434, "step": 5024 }, { "epoch": 0.3453133589884552, "grad_norm": 0.796220600605011, "learning_rate": 2.2835699271590475e-05, "loss": 0.0306, "step": 5025 }, { "epoch": 0.3453820780648708, "grad_norm": 0.7779503464698792, "learning_rate": 2.2832852237583182e-05, "loss": 0.0262, "step": 5026 }, { "epoch": 0.3454507971412864, "grad_norm": 1.714228868484497, "learning_rate": 2.2830004815552488e-05, "loss": 0.0488, "step": 5027 }, { "epoch": 0.34551951621770205, "grad_norm": 0.9438249468803406, "learning_rate": 2.2827157005639453e-05, "loss": 0.0291, "step": 5028 }, { "epoch": 0.34558823529411764, "grad_norm": 1.6046439409255981, "learning_rate": 2.2824308807985158e-05, "loss": 0.0404, "step": 5029 }, { "epoch": 0.3456569543705333, "grad_norm": 0.9860410094261169, "learning_rate": 2.282146022273069e-05, "loss": 0.042, "step": 5030 }, { "epoch": 0.3457256734469489, "grad_norm": 0.833490788936615, "learning_rate": 2.2818611250017162e-05, "loss": 0.0252, "step": 5031 }, { "epoch": 0.34579439252336447, "grad_norm": 1.3356053829193115, "learning_rate": 2.2815761889985703e-05, "loss": 0.0378, "step": 5032 }, { "epoch": 0.3458631115997801, "grad_norm": 0.879616379737854, "learning_rate": 2.2812912142777472e-05, "loss": 0.0353, "step": 5033 }, { "epoch": 0.3459318306761957, "grad_norm": 1.0741467475891113, "learning_rate": 2.2810062008533633e-05, "loss": 0.062, "step": 5034 }, { "epoch": 0.3460005497526113, "grad_norm": 0.9688892364501953, "learning_rate": 2.2807211487395383e-05, "loss": 0.0285, "step": 5035 }, { "epoch": 0.34606926882902694, "grad_norm": 0.7061722278594971, "learning_rate": 2.2804360579503925e-05, "loss": 0.024, "step": 5036 }, { "epoch": 0.34613798790544253, "grad_norm": 0.9235071539878845, "learning_rate": 2.280150928500049e-05, "loss": 0.0242, "step": 5037 }, { "epoch": 0.3462067069818582, "grad_norm": 0.6027148962020874, "learning_rate": 2.279865760402632e-05, "loss": 0.0256, "step": 5038 }, { "epoch": 0.34627542605827377, "grad_norm": 1.1428675651550293, "learning_rate": 2.2795805536722695e-05, "loss": 0.0277, "step": 5039 }, { "epoch": 0.34634414513468936, "grad_norm": 0.6449910402297974, "learning_rate": 2.2792953083230885e-05, "loss": 0.0288, "step": 5040 }, { "epoch": 0.346412864211105, "grad_norm": 1.223741888999939, "learning_rate": 2.27901002436922e-05, "loss": 0.0422, "step": 5041 }, { "epoch": 0.3464815832875206, "grad_norm": 0.546656608581543, "learning_rate": 2.2787247018247966e-05, "loss": 0.0176, "step": 5042 }, { "epoch": 0.34655030236393625, "grad_norm": 0.6701633334159851, "learning_rate": 2.2784393407039527e-05, "loss": 0.0342, "step": 5043 }, { "epoch": 0.34661902144035184, "grad_norm": 1.4380748271942139, "learning_rate": 2.2781539410208236e-05, "loss": 0.0463, "step": 5044 }, { "epoch": 0.34668774051676743, "grad_norm": 0.781696617603302, "learning_rate": 2.2778685027895483e-05, "loss": 0.0303, "step": 5045 }, { "epoch": 0.3467564595931831, "grad_norm": 0.9020368456840515, "learning_rate": 2.2775830260242672e-05, "loss": 0.0302, "step": 5046 }, { "epoch": 0.34682517866959867, "grad_norm": 0.725739061832428, "learning_rate": 2.277297510739121e-05, "loss": 0.0398, "step": 5047 }, { "epoch": 0.3468938977460143, "grad_norm": 1.9159636497497559, "learning_rate": 2.2770119569482547e-05, "loss": 0.0361, "step": 5048 }, { "epoch": 0.3469626168224299, "grad_norm": 1.051452398300171, "learning_rate": 2.276726364665813e-05, "loss": 0.0252, "step": 5049 }, { "epoch": 0.3470313358988455, "grad_norm": 1.0415652990341187, "learning_rate": 2.276440733905944e-05, "loss": 0.0264, "step": 5050 }, { "epoch": 0.34710005497526114, "grad_norm": 0.6769400238990784, "learning_rate": 2.276155064682798e-05, "loss": 0.0207, "step": 5051 }, { "epoch": 0.34716877405167673, "grad_norm": 0.9945383071899414, "learning_rate": 2.2758693570105253e-05, "loss": 0.0198, "step": 5052 }, { "epoch": 0.3472374931280924, "grad_norm": 0.5353222489356995, "learning_rate": 2.2755836109032797e-05, "loss": 0.0214, "step": 5053 }, { "epoch": 0.34730621220450797, "grad_norm": 0.876474916934967, "learning_rate": 2.2752978263752172e-05, "loss": 0.0251, "step": 5054 }, { "epoch": 0.34737493128092356, "grad_norm": 0.34123358130455017, "learning_rate": 2.2750120034404936e-05, "loss": 0.0272, "step": 5055 }, { "epoch": 0.3474436503573392, "grad_norm": 0.7524268627166748, "learning_rate": 2.2747261421132687e-05, "loss": 0.0187, "step": 5056 }, { "epoch": 0.3475123694337548, "grad_norm": 1.250976800918579, "learning_rate": 2.274440242407704e-05, "loss": 0.0324, "step": 5057 }, { "epoch": 0.34758108851017044, "grad_norm": 0.6325927972793579, "learning_rate": 2.2741543043379617e-05, "loss": 0.0386, "step": 5058 }, { "epoch": 0.34764980758658603, "grad_norm": 1.040928840637207, "learning_rate": 2.273868327918207e-05, "loss": 0.0225, "step": 5059 }, { "epoch": 0.3477185266630016, "grad_norm": 1.0259515047073364, "learning_rate": 2.273582313162606e-05, "loss": 0.0346, "step": 5060 }, { "epoch": 0.3477872457394173, "grad_norm": 0.9362322092056274, "learning_rate": 2.273296260085328e-05, "loss": 0.0332, "step": 5061 }, { "epoch": 0.34785596481583286, "grad_norm": 0.7273455858230591, "learning_rate": 2.2730101687005425e-05, "loss": 0.0237, "step": 5062 }, { "epoch": 0.3479246838922485, "grad_norm": 0.9397883415222168, "learning_rate": 2.2727240390224227e-05, "loss": 0.0352, "step": 5063 }, { "epoch": 0.3479934029686641, "grad_norm": 0.8324471712112427, "learning_rate": 2.272437871065143e-05, "loss": 0.0241, "step": 5064 }, { "epoch": 0.3480621220450797, "grad_norm": 0.7036795020103455, "learning_rate": 2.272151664842879e-05, "loss": 0.0276, "step": 5065 }, { "epoch": 0.34813084112149534, "grad_norm": 1.0000554323196411, "learning_rate": 2.2718654203698085e-05, "loss": 0.0273, "step": 5066 }, { "epoch": 0.34819956019791093, "grad_norm": 0.7464417815208435, "learning_rate": 2.2715791376601127e-05, "loss": 0.0323, "step": 5067 }, { "epoch": 0.3482682792743266, "grad_norm": 1.592100739479065, "learning_rate": 2.271292816727972e-05, "loss": 0.0325, "step": 5068 }, { "epoch": 0.34833699835074217, "grad_norm": 0.8699294924736023, "learning_rate": 2.2710064575875718e-05, "loss": 0.0362, "step": 5069 }, { "epoch": 0.34840571742715776, "grad_norm": 1.276870608329773, "learning_rate": 2.2707200602530963e-05, "loss": 0.031, "step": 5070 }, { "epoch": 0.3484744365035734, "grad_norm": 1.274898886680603, "learning_rate": 2.2704336247387333e-05, "loss": 0.0248, "step": 5071 }, { "epoch": 0.348543155579989, "grad_norm": 1.795464038848877, "learning_rate": 2.2701471510586727e-05, "loss": 0.0373, "step": 5072 }, { "epoch": 0.34861187465640464, "grad_norm": 1.328969955444336, "learning_rate": 2.2698606392271057e-05, "loss": 0.0324, "step": 5073 }, { "epoch": 0.34868059373282023, "grad_norm": 0.7898263335227966, "learning_rate": 2.2695740892582252e-05, "loss": 0.0289, "step": 5074 }, { "epoch": 0.3487493128092358, "grad_norm": 0.9384951591491699, "learning_rate": 2.2692875011662267e-05, "loss": 0.0212, "step": 5075 }, { "epoch": 0.34881803188565147, "grad_norm": 0.9171767234802246, "learning_rate": 2.269000874965307e-05, "loss": 0.02, "step": 5076 }, { "epoch": 0.34888675096206706, "grad_norm": 0.8775858879089355, "learning_rate": 2.268714210669665e-05, "loss": 0.0327, "step": 5077 }, { "epoch": 0.3489554700384827, "grad_norm": 0.9845703840255737, "learning_rate": 2.2684275082935016e-05, "loss": 0.0305, "step": 5078 }, { "epoch": 0.3490241891148983, "grad_norm": 0.782180666923523, "learning_rate": 2.2681407678510188e-05, "loss": 0.0303, "step": 5079 }, { "epoch": 0.3490929081913139, "grad_norm": 0.8085262775421143, "learning_rate": 2.2678539893564218e-05, "loss": 0.0295, "step": 5080 }, { "epoch": 0.34916162726772954, "grad_norm": 1.0573265552520752, "learning_rate": 2.2675671728239166e-05, "loss": 0.0294, "step": 5081 }, { "epoch": 0.3492303463441451, "grad_norm": 0.7611165046691895, "learning_rate": 2.2672803182677118e-05, "loss": 0.0259, "step": 5082 }, { "epoch": 0.3492990654205608, "grad_norm": 1.7926443815231323, "learning_rate": 2.2669934257020175e-05, "loss": 0.0304, "step": 5083 }, { "epoch": 0.34936778449697636, "grad_norm": 0.8189838528633118, "learning_rate": 2.266706495141046e-05, "loss": 0.0279, "step": 5084 }, { "epoch": 0.34943650357339195, "grad_norm": 1.454068660736084, "learning_rate": 2.2664195265990106e-05, "loss": 0.0435, "step": 5085 }, { "epoch": 0.3495052226498076, "grad_norm": 0.927561342716217, "learning_rate": 2.2661325200901278e-05, "loss": 0.03, "step": 5086 }, { "epoch": 0.3495739417262232, "grad_norm": 0.36832985281944275, "learning_rate": 2.265845475628615e-05, "loss": 0.0382, "step": 5087 }, { "epoch": 0.34964266080263884, "grad_norm": 0.5514177083969116, "learning_rate": 2.2655583932286922e-05, "loss": 0.0276, "step": 5088 }, { "epoch": 0.34971137987905443, "grad_norm": 0.8950256109237671, "learning_rate": 2.2652712729045798e-05, "loss": 0.0287, "step": 5089 }, { "epoch": 0.34978009895547, "grad_norm": 1.1067523956298828, "learning_rate": 2.2649841146705022e-05, "loss": 0.0245, "step": 5090 }, { "epoch": 0.34984881803188567, "grad_norm": 1.3380918502807617, "learning_rate": 2.2646969185406846e-05, "loss": 0.0351, "step": 5091 }, { "epoch": 0.34991753710830126, "grad_norm": 0.8087352514266968, "learning_rate": 2.2644096845293533e-05, "loss": 0.029, "step": 5092 }, { "epoch": 0.3499862561847169, "grad_norm": 0.8132447600364685, "learning_rate": 2.264122412650738e-05, "loss": 0.0292, "step": 5093 }, { "epoch": 0.3500549752611325, "grad_norm": 0.6200885772705078, "learning_rate": 2.2638351029190693e-05, "loss": 0.0222, "step": 5094 }, { "epoch": 0.3501236943375481, "grad_norm": 0.7168134450912476, "learning_rate": 2.2635477553485806e-05, "loss": 0.0256, "step": 5095 }, { "epoch": 0.35019241341396373, "grad_norm": 0.8137827515602112, "learning_rate": 2.2632603699535054e-05, "loss": 0.0224, "step": 5096 }, { "epoch": 0.3502611324903793, "grad_norm": 0.7393433451652527, "learning_rate": 2.2629729467480804e-05, "loss": 0.0292, "step": 5097 }, { "epoch": 0.35032985156679497, "grad_norm": 0.4255111515522003, "learning_rate": 2.262685485746544e-05, "loss": 0.0207, "step": 5098 }, { "epoch": 0.35039857064321056, "grad_norm": 0.4523084759712219, "learning_rate": 2.2623979869631374e-05, "loss": 0.0172, "step": 5099 }, { "epoch": 0.35046728971962615, "grad_norm": 0.8019487261772156, "learning_rate": 2.2621104504121017e-05, "loss": 0.0318, "step": 5100 }, { "epoch": 0.3505360087960418, "grad_norm": 0.5052773952484131, "learning_rate": 2.2618228761076813e-05, "loss": 0.023, "step": 5101 }, { "epoch": 0.3506047278724574, "grad_norm": 0.7283273339271545, "learning_rate": 2.2615352640641215e-05, "loss": 0.0181, "step": 5102 }, { "epoch": 0.350673446948873, "grad_norm": 0.8802670836448669, "learning_rate": 2.2612476142956707e-05, "loss": 0.022, "step": 5103 }, { "epoch": 0.3507421660252886, "grad_norm": 0.639573335647583, "learning_rate": 2.2609599268165778e-05, "loss": 0.0242, "step": 5104 }, { "epoch": 0.3508108851017042, "grad_norm": 0.17868861556053162, "learning_rate": 2.2606722016410947e-05, "loss": 0.0134, "step": 5105 }, { "epoch": 0.35087960417811986, "grad_norm": 2.08575701713562, "learning_rate": 2.260384438783475e-05, "loss": 0.0382, "step": 5106 }, { "epoch": 0.35094832325453545, "grad_norm": 0.6573291420936584, "learning_rate": 2.260096638257973e-05, "loss": 0.019, "step": 5107 }, { "epoch": 0.35101704233095105, "grad_norm": 0.7356423139572144, "learning_rate": 2.259808800078846e-05, "loss": 0.0211, "step": 5108 }, { "epoch": 0.3510857614073667, "grad_norm": 1.2694613933563232, "learning_rate": 2.2595209242603536e-05, "loss": 0.0304, "step": 5109 }, { "epoch": 0.3511544804837823, "grad_norm": 0.6944279670715332, "learning_rate": 2.2592330108167557e-05, "loss": 0.0361, "step": 5110 }, { "epoch": 0.35122319956019793, "grad_norm": 2.026047468185425, "learning_rate": 2.2589450597623162e-05, "loss": 0.0237, "step": 5111 }, { "epoch": 0.3512919186366135, "grad_norm": 0.930223286151886, "learning_rate": 2.258657071111298e-05, "loss": 0.0225, "step": 5112 }, { "epoch": 0.3513606377130291, "grad_norm": 1.1418654918670654, "learning_rate": 2.2583690448779687e-05, "loss": 0.0291, "step": 5113 }, { "epoch": 0.35142935678944476, "grad_norm": 0.7091242671012878, "learning_rate": 2.2580809810765958e-05, "loss": 0.0307, "step": 5114 }, { "epoch": 0.35149807586586035, "grad_norm": 1.7394134998321533, "learning_rate": 2.2577928797214497e-05, "loss": 0.0307, "step": 5115 }, { "epoch": 0.351566794942276, "grad_norm": 0.49227601289749146, "learning_rate": 2.2575047408268022e-05, "loss": 0.0201, "step": 5116 }, { "epoch": 0.3516355140186916, "grad_norm": 0.7120018005371094, "learning_rate": 2.2572165644069272e-05, "loss": 0.0333, "step": 5117 }, { "epoch": 0.3517042330951072, "grad_norm": 0.8519906401634216, "learning_rate": 2.256928350476101e-05, "loss": 0.0285, "step": 5118 }, { "epoch": 0.3517729521715228, "grad_norm": 0.7355486750602722, "learning_rate": 2.2566400990485995e-05, "loss": 0.0401, "step": 5119 }, { "epoch": 0.3518416712479384, "grad_norm": 2.992307186126709, "learning_rate": 2.256351810138704e-05, "loss": 0.0393, "step": 5120 }, { "epoch": 0.35191039032435406, "grad_norm": 0.7251071929931641, "learning_rate": 2.2560634837606943e-05, "loss": 0.0253, "step": 5121 }, { "epoch": 0.35197910940076965, "grad_norm": 0.8994451761245728, "learning_rate": 2.255775119928854e-05, "loss": 0.0241, "step": 5122 }, { "epoch": 0.35204782847718524, "grad_norm": 2.438382387161255, "learning_rate": 2.255486718657468e-05, "loss": 0.0291, "step": 5123 }, { "epoch": 0.3521165475536009, "grad_norm": 0.424709290266037, "learning_rate": 2.2551982799608234e-05, "loss": 0.0317, "step": 5124 }, { "epoch": 0.3521852666300165, "grad_norm": 0.7892211079597473, "learning_rate": 2.2549098038532083e-05, "loss": 0.0287, "step": 5125 }, { "epoch": 0.3522539857064321, "grad_norm": 0.6389082074165344, "learning_rate": 2.254621290348914e-05, "loss": 0.0427, "step": 5126 }, { "epoch": 0.3523227047828477, "grad_norm": 0.886664628982544, "learning_rate": 2.2543327394622322e-05, "loss": 0.0345, "step": 5127 }, { "epoch": 0.3523914238592633, "grad_norm": 0.44483962655067444, "learning_rate": 2.254044151207457e-05, "loss": 0.02, "step": 5128 }, { "epoch": 0.35246014293567896, "grad_norm": 0.8055991530418396, "learning_rate": 2.2537555255988853e-05, "loss": 0.0381, "step": 5129 }, { "epoch": 0.35252886201209455, "grad_norm": 0.7533624172210693, "learning_rate": 2.2534668626508145e-05, "loss": 0.0498, "step": 5130 }, { "epoch": 0.3525975810885102, "grad_norm": 1.1445242166519165, "learning_rate": 2.253178162377544e-05, "loss": 0.0445, "step": 5131 }, { "epoch": 0.3526663001649258, "grad_norm": 0.6377441883087158, "learning_rate": 2.252889424793376e-05, "loss": 0.0236, "step": 5132 }, { "epoch": 0.3527350192413414, "grad_norm": 1.1110221147537231, "learning_rate": 2.2526006499126142e-05, "loss": 0.0367, "step": 5133 }, { "epoch": 0.352803738317757, "grad_norm": 1.3759950399398804, "learning_rate": 2.252311837749563e-05, "loss": 0.0402, "step": 5134 }, { "epoch": 0.3528724573941726, "grad_norm": 1.067378044128418, "learning_rate": 2.2520229883185304e-05, "loss": 0.0335, "step": 5135 }, { "epoch": 0.35294117647058826, "grad_norm": 0.5049024820327759, "learning_rate": 2.2517341016338247e-05, "loss": 0.0317, "step": 5136 }, { "epoch": 0.35300989554700385, "grad_norm": 0.8329523205757141, "learning_rate": 2.2514451777097572e-05, "loss": 0.023, "step": 5137 }, { "epoch": 0.35307861462341944, "grad_norm": 1.987978458404541, "learning_rate": 2.2511562165606406e-05, "loss": 0.0437, "step": 5138 }, { "epoch": 0.3531473336998351, "grad_norm": 0.6834468245506287, "learning_rate": 2.2508672182007892e-05, "loss": 0.0202, "step": 5139 }, { "epoch": 0.3532160527762507, "grad_norm": 0.9263501763343811, "learning_rate": 2.2505781826445193e-05, "loss": 0.0274, "step": 5140 }, { "epoch": 0.3532847718526663, "grad_norm": 0.5551478862762451, "learning_rate": 2.25028910990615e-05, "loss": 0.0411, "step": 5141 }, { "epoch": 0.3533534909290819, "grad_norm": 1.0048725605010986, "learning_rate": 2.25e-05, "loss": 0.0264, "step": 5142 }, { "epoch": 0.3534222100054975, "grad_norm": 3.081834077835083, "learning_rate": 2.2497108529403924e-05, "loss": 0.0305, "step": 5143 }, { "epoch": 0.35349092908191315, "grad_norm": 1.1828173398971558, "learning_rate": 2.2494216687416503e-05, "loss": 0.0281, "step": 5144 }, { "epoch": 0.35355964815832874, "grad_norm": 0.8515859246253967, "learning_rate": 2.2491324474181e-05, "loss": 0.0251, "step": 5145 }, { "epoch": 0.3536283672347444, "grad_norm": 0.687949001789093, "learning_rate": 2.2488431889840672e-05, "loss": 0.0355, "step": 5146 }, { "epoch": 0.35369708631116, "grad_norm": 0.7853311896324158, "learning_rate": 2.2485538934538834e-05, "loss": 0.0388, "step": 5147 }, { "epoch": 0.35376580538757557, "grad_norm": 0.6980417966842651, "learning_rate": 2.2482645608418786e-05, "loss": 0.0313, "step": 5148 }, { "epoch": 0.3538345244639912, "grad_norm": 1.7911986112594604, "learning_rate": 2.2479751911623855e-05, "loss": 0.0343, "step": 5149 }, { "epoch": 0.3539032435404068, "grad_norm": 1.4066781997680664, "learning_rate": 2.2476857844297394e-05, "loss": 0.0322, "step": 5150 }, { "epoch": 0.35397196261682246, "grad_norm": 1.3686562776565552, "learning_rate": 2.2473963406582768e-05, "loss": 0.034, "step": 5151 }, { "epoch": 0.35404068169323805, "grad_norm": 1.343516230583191, "learning_rate": 2.247106859862336e-05, "loss": 0.0241, "step": 5152 }, { "epoch": 0.35410940076965364, "grad_norm": 1.1511834859848022, "learning_rate": 2.246817342056257e-05, "loss": 0.0455, "step": 5153 }, { "epoch": 0.3541781198460693, "grad_norm": 0.5204797983169556, "learning_rate": 2.2465277872543825e-05, "loss": 0.0339, "step": 5154 }, { "epoch": 0.3542468389224849, "grad_norm": 0.9162320494651794, "learning_rate": 2.2462381954710566e-05, "loss": 0.0311, "step": 5155 }, { "epoch": 0.3543155579989005, "grad_norm": 0.7987114787101746, "learning_rate": 2.2459485667206243e-05, "loss": 0.0396, "step": 5156 }, { "epoch": 0.3543842770753161, "grad_norm": 0.6754471659660339, "learning_rate": 2.2456589010174337e-05, "loss": 0.0224, "step": 5157 }, { "epoch": 0.3544529961517317, "grad_norm": 0.46400436758995056, "learning_rate": 2.2453691983758336e-05, "loss": 0.0201, "step": 5158 }, { "epoch": 0.35452171522814735, "grad_norm": 0.8544301390647888, "learning_rate": 2.2450794588101764e-05, "loss": 0.0288, "step": 5159 }, { "epoch": 0.35459043430456294, "grad_norm": 0.4682213068008423, "learning_rate": 2.244789682334815e-05, "loss": 0.0216, "step": 5160 }, { "epoch": 0.3546591533809786, "grad_norm": 0.7134612798690796, "learning_rate": 2.2444998689641033e-05, "loss": 0.0242, "step": 5161 }, { "epoch": 0.3547278724573942, "grad_norm": 1.1983134746551514, "learning_rate": 2.244210018712399e-05, "loss": 0.0311, "step": 5162 }, { "epoch": 0.35479659153380977, "grad_norm": 0.8649682402610779, "learning_rate": 2.2439201315940602e-05, "loss": 0.0288, "step": 5163 }, { "epoch": 0.3548653106102254, "grad_norm": 0.46805429458618164, "learning_rate": 2.2436302076234472e-05, "loss": 0.0217, "step": 5164 }, { "epoch": 0.354934029686641, "grad_norm": 0.5903185606002808, "learning_rate": 2.2433402468149235e-05, "loss": 0.0273, "step": 5165 }, { "epoch": 0.3550027487630566, "grad_norm": 1.0035350322723389, "learning_rate": 2.243050249182852e-05, "loss": 0.0255, "step": 5166 }, { "epoch": 0.35507146783947224, "grad_norm": 0.8867372870445251, "learning_rate": 2.2427602147415982e-05, "loss": 0.0234, "step": 5167 }, { "epoch": 0.35514018691588783, "grad_norm": 1.7540206909179688, "learning_rate": 2.2424701435055307e-05, "loss": 0.0287, "step": 5168 }, { "epoch": 0.3552089059923035, "grad_norm": 0.7469322681427002, "learning_rate": 2.242180035489019e-05, "loss": 0.0169, "step": 5169 }, { "epoch": 0.35527762506871907, "grad_norm": 0.8212186694145203, "learning_rate": 2.241889890706434e-05, "loss": 0.0307, "step": 5170 }, { "epoch": 0.35534634414513466, "grad_norm": 1.48502516746521, "learning_rate": 2.2415997091721493e-05, "loss": 0.0238, "step": 5171 }, { "epoch": 0.3554150632215503, "grad_norm": 0.9218698143959045, "learning_rate": 2.2413094909005392e-05, "loss": 0.0238, "step": 5172 }, { "epoch": 0.3554837822979659, "grad_norm": 0.757804811000824, "learning_rate": 2.241019235905981e-05, "loss": 0.0231, "step": 5173 }, { "epoch": 0.35555250137438155, "grad_norm": 0.6604236960411072, "learning_rate": 2.2407289442028536e-05, "loss": 0.0266, "step": 5174 }, { "epoch": 0.35562122045079714, "grad_norm": 1.0714616775512695, "learning_rate": 2.2404386158055372e-05, "loss": 0.028, "step": 5175 }, { "epoch": 0.35568993952721273, "grad_norm": 1.629124641418457, "learning_rate": 2.240148250728414e-05, "loss": 0.0411, "step": 5176 }, { "epoch": 0.3557586586036284, "grad_norm": 0.8550655245780945, "learning_rate": 2.2398578489858682e-05, "loss": 0.0255, "step": 5177 }, { "epoch": 0.35582737768004397, "grad_norm": 0.5155609846115112, "learning_rate": 2.2395674105922854e-05, "loss": 0.0269, "step": 5178 }, { "epoch": 0.3558960967564596, "grad_norm": 0.45996472239494324, "learning_rate": 2.2392769355620537e-05, "loss": 0.0331, "step": 5179 }, { "epoch": 0.3559648158328752, "grad_norm": 1.1170679330825806, "learning_rate": 2.2389864239095628e-05, "loss": 0.025, "step": 5180 }, { "epoch": 0.3560335349092908, "grad_norm": 0.5416746735572815, "learning_rate": 2.2386958756492033e-05, "loss": 0.021, "step": 5181 }, { "epoch": 0.35610225398570644, "grad_norm": 1.8416228294372559, "learning_rate": 2.2384052907953688e-05, "loss": 0.0367, "step": 5182 }, { "epoch": 0.35617097306212203, "grad_norm": 0.7140272259712219, "learning_rate": 2.2381146693624548e-05, "loss": 0.0245, "step": 5183 }, { "epoch": 0.3562396921385377, "grad_norm": 1.5764970779418945, "learning_rate": 2.237824011364857e-05, "loss": 0.036, "step": 5184 }, { "epoch": 0.35630841121495327, "grad_norm": 1.5365849733352661, "learning_rate": 2.2375333168169748e-05, "loss": 0.0336, "step": 5185 }, { "epoch": 0.35637713029136886, "grad_norm": 1.1152846813201904, "learning_rate": 2.2372425857332086e-05, "loss": 0.0374, "step": 5186 }, { "epoch": 0.3564458493677845, "grad_norm": 4.244269371032715, "learning_rate": 2.23695181812796e-05, "loss": 0.0267, "step": 5187 }, { "epoch": 0.3565145684442001, "grad_norm": 1.0908701419830322, "learning_rate": 2.236661014015634e-05, "loss": 0.0342, "step": 5188 }, { "epoch": 0.35658328752061574, "grad_norm": 0.9501123428344727, "learning_rate": 2.2363701734106356e-05, "loss": 0.0318, "step": 5189 }, { "epoch": 0.35665200659703133, "grad_norm": 1.085631012916565, "learning_rate": 2.2360792963273727e-05, "loss": 0.0325, "step": 5190 }, { "epoch": 0.3567207256734469, "grad_norm": 1.1677592992782593, "learning_rate": 2.2357883827802547e-05, "loss": 0.0316, "step": 5191 }, { "epoch": 0.35678944474986257, "grad_norm": 0.5951423048973083, "learning_rate": 2.2354974327836927e-05, "loss": 0.0337, "step": 5192 }, { "epoch": 0.35685816382627816, "grad_norm": 0.7674171924591064, "learning_rate": 2.2352064463521003e-05, "loss": 0.0307, "step": 5193 }, { "epoch": 0.3569268829026938, "grad_norm": 1.2425758838653564, "learning_rate": 2.2349154234998924e-05, "loss": 0.0489, "step": 5194 }, { "epoch": 0.3569956019791094, "grad_norm": 0.8853631615638733, "learning_rate": 2.2346243642414846e-05, "loss": 0.0344, "step": 5195 }, { "epoch": 0.357064321055525, "grad_norm": 1.1009118556976318, "learning_rate": 2.234333268591297e-05, "loss": 0.0306, "step": 5196 }, { "epoch": 0.35713304013194064, "grad_norm": 1.3621643781661987, "learning_rate": 2.234042136563748e-05, "loss": 0.0262, "step": 5197 }, { "epoch": 0.35720175920835623, "grad_norm": 0.586089551448822, "learning_rate": 2.2337509681732613e-05, "loss": 0.0196, "step": 5198 }, { "epoch": 0.3572704782847719, "grad_norm": 0.6805404424667358, "learning_rate": 2.23345976343426e-05, "loss": 0.0247, "step": 5199 }, { "epoch": 0.35733919736118747, "grad_norm": 0.4703271687030792, "learning_rate": 2.23316852236117e-05, "loss": 0.0253, "step": 5200 }, { "epoch": 0.35740791643760306, "grad_norm": 1.0758706331253052, "learning_rate": 2.232877244968419e-05, "loss": 0.0249, "step": 5201 }, { "epoch": 0.3574766355140187, "grad_norm": 1.2782442569732666, "learning_rate": 2.2325859312704358e-05, "loss": 0.0259, "step": 5202 }, { "epoch": 0.3575453545904343, "grad_norm": 0.8315905928611755, "learning_rate": 2.2322945812816512e-05, "loss": 0.0335, "step": 5203 }, { "epoch": 0.35761407366684994, "grad_norm": 1.1644526720046997, "learning_rate": 2.2320031950164993e-05, "loss": 0.0431, "step": 5204 }, { "epoch": 0.35768279274326553, "grad_norm": 0.8160808086395264, "learning_rate": 2.231711772489414e-05, "loss": 0.0302, "step": 5205 }, { "epoch": 0.3577515118196811, "grad_norm": 0.8718923330307007, "learning_rate": 2.2314203137148318e-05, "loss": 0.0496, "step": 5206 }, { "epoch": 0.35782023089609677, "grad_norm": 1.0809167623519897, "learning_rate": 2.2311288187071907e-05, "loss": 0.0477, "step": 5207 }, { "epoch": 0.35788894997251236, "grad_norm": 1.1850571632385254, "learning_rate": 2.230837287480931e-05, "loss": 0.0278, "step": 5208 }, { "epoch": 0.357957669048928, "grad_norm": 0.6699739694595337, "learning_rate": 2.2305457200504948e-05, "loss": 0.0396, "step": 5209 }, { "epoch": 0.3580263881253436, "grad_norm": 1.4957184791564941, "learning_rate": 2.2302541164303254e-05, "loss": 0.0436, "step": 5210 }, { "epoch": 0.3580951072017592, "grad_norm": 0.5197762250900269, "learning_rate": 2.229962476634869e-05, "loss": 0.029, "step": 5211 }, { "epoch": 0.35816382627817483, "grad_norm": 0.8631303310394287, "learning_rate": 2.2296708006785714e-05, "loss": 0.0378, "step": 5212 }, { "epoch": 0.3582325453545904, "grad_norm": 1.4233819246292114, "learning_rate": 2.229379088575883e-05, "loss": 0.0314, "step": 5213 }, { "epoch": 0.3583012644310061, "grad_norm": 0.48712557554244995, "learning_rate": 2.229087340341254e-05, "loss": 0.0305, "step": 5214 }, { "epoch": 0.35836998350742166, "grad_norm": 1.4098286628723145, "learning_rate": 2.228795555989137e-05, "loss": 0.0434, "step": 5215 }, { "epoch": 0.35843870258383725, "grad_norm": 0.9099448919296265, "learning_rate": 2.2285037355339865e-05, "loss": 0.0427, "step": 5216 }, { "epoch": 0.3585074216602529, "grad_norm": 0.9038567543029785, "learning_rate": 2.2282118789902588e-05, "loss": 0.0361, "step": 5217 }, { "epoch": 0.3585761407366685, "grad_norm": 1.0461622476577759, "learning_rate": 2.2279199863724117e-05, "loss": 0.0367, "step": 5218 }, { "epoch": 0.35864485981308414, "grad_norm": 0.5130905508995056, "learning_rate": 2.227628057694905e-05, "loss": 0.0197, "step": 5219 }, { "epoch": 0.35871357888949973, "grad_norm": 0.7319754362106323, "learning_rate": 2.2273360929722e-05, "loss": 0.0348, "step": 5220 }, { "epoch": 0.3587822979659153, "grad_norm": 1.0516588687896729, "learning_rate": 2.2270440922187608e-05, "loss": 0.0342, "step": 5221 }, { "epoch": 0.35885101704233097, "grad_norm": 0.7422266006469727, "learning_rate": 2.226752055449052e-05, "loss": 0.016, "step": 5222 }, { "epoch": 0.35891973611874656, "grad_norm": 1.268641710281372, "learning_rate": 2.2264599826775403e-05, "loss": 0.0509, "step": 5223 }, { "epoch": 0.3589884551951622, "grad_norm": 0.9358899593353271, "learning_rate": 2.2261678739186944e-05, "loss": 0.0269, "step": 5224 }, { "epoch": 0.3590571742715778, "grad_norm": 1.3451989889144897, "learning_rate": 2.225875729186985e-05, "loss": 0.0338, "step": 5225 }, { "epoch": 0.3591258933479934, "grad_norm": 1.2773244380950928, "learning_rate": 2.2255835484968846e-05, "loss": 0.04, "step": 5226 }, { "epoch": 0.35919461242440903, "grad_norm": 1.131766438484192, "learning_rate": 2.225291331862867e-05, "loss": 0.029, "step": 5227 }, { "epoch": 0.3592633315008246, "grad_norm": 0.9804988503456116, "learning_rate": 2.2249990792994076e-05, "loss": 0.0277, "step": 5228 }, { "epoch": 0.35933205057724027, "grad_norm": 0.9680737853050232, "learning_rate": 2.2247067908209847e-05, "loss": 0.0201, "step": 5229 }, { "epoch": 0.35940076965365586, "grad_norm": 0.5850586891174316, "learning_rate": 2.224414466442077e-05, "loss": 0.0332, "step": 5230 }, { "epoch": 0.35946948873007145, "grad_norm": 1.4773035049438477, "learning_rate": 2.2241221061771664e-05, "loss": 0.0269, "step": 5231 }, { "epoch": 0.3595382078064871, "grad_norm": 1.3881727457046509, "learning_rate": 2.2238297100407352e-05, "loss": 0.0308, "step": 5232 }, { "epoch": 0.3596069268829027, "grad_norm": 0.8395181894302368, "learning_rate": 2.2235372780472675e-05, "loss": 0.0358, "step": 5233 }, { "epoch": 0.3596756459593183, "grad_norm": 0.9685493111610413, "learning_rate": 2.2232448102112517e-05, "loss": 0.0216, "step": 5234 }, { "epoch": 0.3597443650357339, "grad_norm": 0.6253697872161865, "learning_rate": 2.2229523065471748e-05, "loss": 0.031, "step": 5235 }, { "epoch": 0.3598130841121495, "grad_norm": 0.8043612241744995, "learning_rate": 2.2226597670695263e-05, "loss": 0.0188, "step": 5236 }, { "epoch": 0.35988180318856516, "grad_norm": 1.1578423976898193, "learning_rate": 2.2223671917927988e-05, "loss": 0.0361, "step": 5237 }, { "epoch": 0.35995052226498075, "grad_norm": 1.4796619415283203, "learning_rate": 2.2220745807314863e-05, "loss": 0.029, "step": 5238 }, { "epoch": 0.36001924134139635, "grad_norm": 0.6881701350212097, "learning_rate": 2.2217819339000832e-05, "loss": 0.0294, "step": 5239 }, { "epoch": 0.360087960417812, "grad_norm": 0.5247688889503479, "learning_rate": 2.221489251313087e-05, "loss": 0.0215, "step": 5240 }, { "epoch": 0.3601566794942276, "grad_norm": 1.2864677906036377, "learning_rate": 2.221196532984997e-05, "loss": 0.0391, "step": 5241 }, { "epoch": 0.36022539857064323, "grad_norm": 0.8052838444709778, "learning_rate": 2.2209037789303127e-05, "loss": 0.022, "step": 5242 }, { "epoch": 0.3602941176470588, "grad_norm": 0.9362925291061401, "learning_rate": 2.2206109891635377e-05, "loss": 0.0513, "step": 5243 }, { "epoch": 0.3603628367234744, "grad_norm": 1.0432261228561401, "learning_rate": 2.220318163699176e-05, "loss": 0.0284, "step": 5244 }, { "epoch": 0.36043155579989006, "grad_norm": 1.093371868133545, "learning_rate": 2.220025302551733e-05, "loss": 0.0447, "step": 5245 }, { "epoch": 0.36050027487630565, "grad_norm": 0.7133861184120178, "learning_rate": 2.2197324057357176e-05, "loss": 0.037, "step": 5246 }, { "epoch": 0.3605689939527213, "grad_norm": 1.1992745399475098, "learning_rate": 2.219439473265638e-05, "loss": 0.0401, "step": 5247 }, { "epoch": 0.3606377130291369, "grad_norm": 1.1044821739196777, "learning_rate": 2.2191465051560054e-05, "loss": 0.0424, "step": 5248 }, { "epoch": 0.3607064321055525, "grad_norm": 0.6165369153022766, "learning_rate": 2.2188535014213343e-05, "loss": 0.0259, "step": 5249 }, { "epoch": 0.3607751511819681, "grad_norm": 0.8205263018608093, "learning_rate": 2.2185604620761384e-05, "loss": 0.0248, "step": 5250 }, { "epoch": 0.3608438702583837, "grad_norm": 1.2340978384017944, "learning_rate": 2.2182673871349345e-05, "loss": 0.0419, "step": 5251 }, { "epoch": 0.36091258933479936, "grad_norm": 0.9985992312431335, "learning_rate": 2.217974276612241e-05, "loss": 0.0244, "step": 5252 }, { "epoch": 0.36098130841121495, "grad_norm": 0.8494880199432373, "learning_rate": 2.217681130522578e-05, "loss": 0.0252, "step": 5253 }, { "epoch": 0.36105002748763054, "grad_norm": 1.145524024963379, "learning_rate": 2.2173879488804666e-05, "loss": 0.0402, "step": 5254 }, { "epoch": 0.3611187465640462, "grad_norm": 0.8993532061576843, "learning_rate": 2.217094731700432e-05, "loss": 0.0457, "step": 5255 }, { "epoch": 0.3611874656404618, "grad_norm": 0.6114439964294434, "learning_rate": 2.216801478996998e-05, "loss": 0.0353, "step": 5256 }, { "epoch": 0.3612561847168774, "grad_norm": 0.896490752696991, "learning_rate": 2.216508190784693e-05, "loss": 0.0295, "step": 5257 }, { "epoch": 0.361324903793293, "grad_norm": 0.9880701303482056, "learning_rate": 2.216214867078045e-05, "loss": 0.0308, "step": 5258 }, { "epoch": 0.3613936228697086, "grad_norm": 0.755406379699707, "learning_rate": 2.2159215078915848e-05, "loss": 0.0264, "step": 5259 }, { "epoch": 0.36146234194612425, "grad_norm": 0.8618478775024414, "learning_rate": 2.2156281132398458e-05, "loss": 0.0264, "step": 5260 }, { "epoch": 0.36153106102253985, "grad_norm": 1.4772566556930542, "learning_rate": 2.21533468313736e-05, "loss": 0.0269, "step": 5261 }, { "epoch": 0.3615997800989555, "grad_norm": 1.9763832092285156, "learning_rate": 2.2150412175986657e-05, "loss": 0.0339, "step": 5262 }, { "epoch": 0.3616684991753711, "grad_norm": 0.9633755087852478, "learning_rate": 2.214747716638299e-05, "loss": 0.0238, "step": 5263 }, { "epoch": 0.3617372182517867, "grad_norm": 0.6199625730514526, "learning_rate": 2.2144541802708004e-05, "loss": 0.0311, "step": 5264 }, { "epoch": 0.3618059373282023, "grad_norm": 1.014106273651123, "learning_rate": 2.2141606085107104e-05, "loss": 0.0285, "step": 5265 }, { "epoch": 0.3618746564046179, "grad_norm": 1.0246775150299072, "learning_rate": 2.2138670013725718e-05, "loss": 0.0266, "step": 5266 }, { "epoch": 0.36194337548103356, "grad_norm": 0.7788372039794922, "learning_rate": 2.2135733588709297e-05, "loss": 0.0342, "step": 5267 }, { "epoch": 0.36201209455744915, "grad_norm": 0.6002275943756104, "learning_rate": 2.2132796810203306e-05, "loss": 0.0247, "step": 5268 }, { "epoch": 0.36208081363386474, "grad_norm": 0.9120522737503052, "learning_rate": 2.2129859678353225e-05, "loss": 0.025, "step": 5269 }, { "epoch": 0.3621495327102804, "grad_norm": 0.8270391821861267, "learning_rate": 2.2126922193304553e-05, "loss": 0.0408, "step": 5270 }, { "epoch": 0.362218251786696, "grad_norm": 0.9139524698257446, "learning_rate": 2.212398435520281e-05, "loss": 0.0262, "step": 5271 }, { "epoch": 0.3622869708631116, "grad_norm": 0.915921688079834, "learning_rate": 2.2121046164193525e-05, "loss": 0.0203, "step": 5272 }, { "epoch": 0.3623556899395272, "grad_norm": 0.850766658782959, "learning_rate": 2.2118107620422252e-05, "loss": 0.0359, "step": 5273 }, { "epoch": 0.3624244090159428, "grad_norm": 0.8309542536735535, "learning_rate": 2.211516872403457e-05, "loss": 0.027, "step": 5274 }, { "epoch": 0.36249312809235845, "grad_norm": 0.9514709115028381, "learning_rate": 2.2112229475176048e-05, "loss": 0.0246, "step": 5275 }, { "epoch": 0.36256184716877404, "grad_norm": 1.1487317085266113, "learning_rate": 2.2109289873992304e-05, "loss": 0.0351, "step": 5276 }, { "epoch": 0.3626305662451897, "grad_norm": 1.1446868181228638, "learning_rate": 2.2106349920628956e-05, "loss": 0.0297, "step": 5277 }, { "epoch": 0.3626992853216053, "grad_norm": 0.722158670425415, "learning_rate": 2.210340961523164e-05, "loss": 0.0229, "step": 5278 }, { "epoch": 0.36276800439802087, "grad_norm": 1.0334491729736328, "learning_rate": 2.210046895794602e-05, "loss": 0.029, "step": 5279 }, { "epoch": 0.3628367234744365, "grad_norm": 1.0170894861221313, "learning_rate": 2.209752794891776e-05, "loss": 0.0267, "step": 5280 }, { "epoch": 0.3629054425508521, "grad_norm": 0.9041008353233337, "learning_rate": 2.2094586588292566e-05, "loss": 0.0314, "step": 5281 }, { "epoch": 0.36297416162726776, "grad_norm": 1.0008331537246704, "learning_rate": 2.2091644876216133e-05, "loss": 0.0356, "step": 5282 }, { "epoch": 0.36304288070368335, "grad_norm": 0.9390620589256287, "learning_rate": 2.2088702812834195e-05, "loss": 0.0199, "step": 5283 }, { "epoch": 0.36311159978009894, "grad_norm": 0.5188642740249634, "learning_rate": 2.208576039829249e-05, "loss": 0.0329, "step": 5284 }, { "epoch": 0.3631803188565146, "grad_norm": 0.6659818291664124, "learning_rate": 2.2082817632736788e-05, "loss": 0.0196, "step": 5285 }, { "epoch": 0.3632490379329302, "grad_norm": 1.1889146566390991, "learning_rate": 2.2079874516312857e-05, "loss": 0.0275, "step": 5286 }, { "epoch": 0.3633177570093458, "grad_norm": 0.5604586005210876, "learning_rate": 2.20769310491665e-05, "loss": 0.0308, "step": 5287 }, { "epoch": 0.3633864760857614, "grad_norm": 0.4923091232776642, "learning_rate": 2.2073987231443527e-05, "loss": 0.0242, "step": 5288 }, { "epoch": 0.363455195162177, "grad_norm": 1.160936713218689, "learning_rate": 2.207104306328977e-05, "loss": 0.0261, "step": 5289 }, { "epoch": 0.36352391423859265, "grad_norm": 0.9269687533378601, "learning_rate": 2.206809854485108e-05, "loss": 0.0252, "step": 5290 }, { "epoch": 0.36359263331500824, "grad_norm": 1.3165979385375977, "learning_rate": 2.2065153676273316e-05, "loss": 0.0254, "step": 5291 }, { "epoch": 0.3636613523914239, "grad_norm": 0.6909568309783936, "learning_rate": 2.2062208457702366e-05, "loss": 0.033, "step": 5292 }, { "epoch": 0.3637300714678395, "grad_norm": 0.5262306332588196, "learning_rate": 2.2059262889284126e-05, "loss": 0.0182, "step": 5293 }, { "epoch": 0.36379879054425507, "grad_norm": 0.8405232429504395, "learning_rate": 2.205631697116452e-05, "loss": 0.0334, "step": 5294 }, { "epoch": 0.3638675096206707, "grad_norm": 1.1138355731964111, "learning_rate": 2.2053370703489475e-05, "loss": 0.0339, "step": 5295 }, { "epoch": 0.3639362286970863, "grad_norm": 1.0299206972122192, "learning_rate": 2.2050424086404947e-05, "loss": 0.0284, "step": 5296 }, { "epoch": 0.3640049477735019, "grad_norm": 1.0935111045837402, "learning_rate": 2.2047477120056904e-05, "loss": 0.0358, "step": 5297 }, { "epoch": 0.36407366684991754, "grad_norm": 0.8762344717979431, "learning_rate": 2.2044529804591336e-05, "loss": 0.0285, "step": 5298 }, { "epoch": 0.36414238592633313, "grad_norm": 1.2949055433273315, "learning_rate": 2.2041582140154242e-05, "loss": 0.0283, "step": 5299 }, { "epoch": 0.3642111050027488, "grad_norm": 0.954279899597168, "learning_rate": 2.2038634126891647e-05, "loss": 0.02, "step": 5300 }, { "epoch": 0.36427982407916437, "grad_norm": 0.9383095502853394, "learning_rate": 2.2035685764949592e-05, "loss": 0.0268, "step": 5301 }, { "epoch": 0.36434854315557996, "grad_norm": 1.1764190196990967, "learning_rate": 2.2032737054474127e-05, "loss": 0.0299, "step": 5302 }, { "epoch": 0.3644172622319956, "grad_norm": 0.8703596591949463, "learning_rate": 2.202978799561132e-05, "loss": 0.0315, "step": 5303 }, { "epoch": 0.3644859813084112, "grad_norm": 0.7793524265289307, "learning_rate": 2.2026838588507275e-05, "loss": 0.0236, "step": 5304 }, { "epoch": 0.36455470038482685, "grad_norm": 1.1244152784347534, "learning_rate": 2.2023888833308094e-05, "loss": 0.026, "step": 5305 }, { "epoch": 0.36462341946124244, "grad_norm": 1.042677879333496, "learning_rate": 2.2020938730159895e-05, "loss": 0.0362, "step": 5306 }, { "epoch": 0.36469213853765803, "grad_norm": 0.5829599499702454, "learning_rate": 2.2017988279208828e-05, "loss": 0.0287, "step": 5307 }, { "epoch": 0.3647608576140737, "grad_norm": 0.9047175645828247, "learning_rate": 2.2015037480601053e-05, "loss": 0.032, "step": 5308 }, { "epoch": 0.36482957669048927, "grad_norm": 0.8354876637458801, "learning_rate": 2.2012086334482737e-05, "loss": 0.0365, "step": 5309 }, { "epoch": 0.3648982957669049, "grad_norm": 1.075865387916565, "learning_rate": 2.2009134841000086e-05, "loss": 0.0239, "step": 5310 }, { "epoch": 0.3649670148433205, "grad_norm": 0.7796506881713867, "learning_rate": 2.2006183000299307e-05, "loss": 0.0235, "step": 5311 }, { "epoch": 0.3650357339197361, "grad_norm": 0.5101402401924133, "learning_rate": 2.200323081252662e-05, "loss": 0.0179, "step": 5312 }, { "epoch": 0.36510445299615174, "grad_norm": 1.1516339778900146, "learning_rate": 2.200027827782828e-05, "loss": 0.0415, "step": 5313 }, { "epoch": 0.36517317207256733, "grad_norm": 0.9197160005569458, "learning_rate": 2.1997325396350547e-05, "loss": 0.0267, "step": 5314 }, { "epoch": 0.365241891148983, "grad_norm": 0.761136531829834, "learning_rate": 2.1994372168239697e-05, "loss": 0.0379, "step": 5315 }, { "epoch": 0.36531061022539857, "grad_norm": 1.0042182207107544, "learning_rate": 2.199141859364203e-05, "loss": 0.039, "step": 5316 }, { "epoch": 0.36537932930181416, "grad_norm": 0.9697871804237366, "learning_rate": 2.1988464672703863e-05, "loss": 0.0233, "step": 5317 }, { "epoch": 0.3654480483782298, "grad_norm": 0.7779854536056519, "learning_rate": 2.198551040557152e-05, "loss": 0.0295, "step": 5318 }, { "epoch": 0.3655167674546454, "grad_norm": 1.089115023612976, "learning_rate": 2.198255579239136e-05, "loss": 0.0202, "step": 5319 }, { "epoch": 0.36558548653106104, "grad_norm": 0.6652508974075317, "learning_rate": 2.1979600833309728e-05, "loss": 0.019, "step": 5320 }, { "epoch": 0.36565420560747663, "grad_norm": 0.6913716197013855, "learning_rate": 2.197664552847303e-05, "loss": 0.035, "step": 5321 }, { "epoch": 0.3657229246838922, "grad_norm": 0.9047980904579163, "learning_rate": 2.1973689878027657e-05, "loss": 0.03, "step": 5322 }, { "epoch": 0.36579164376030787, "grad_norm": 1.00358247756958, "learning_rate": 2.1970733882120023e-05, "loss": 0.0409, "step": 5323 }, { "epoch": 0.36586036283672346, "grad_norm": 0.7723384499549866, "learning_rate": 2.1967777540896562e-05, "loss": 0.0303, "step": 5324 }, { "epoch": 0.3659290819131391, "grad_norm": 1.1990242004394531, "learning_rate": 2.1964820854503725e-05, "loss": 0.0257, "step": 5325 }, { "epoch": 0.3659978009895547, "grad_norm": 0.8588818311691284, "learning_rate": 2.1961863823087986e-05, "loss": 0.0294, "step": 5326 }, { "epoch": 0.3660665200659703, "grad_norm": 1.0522356033325195, "learning_rate": 2.1958906446795823e-05, "loss": 0.0366, "step": 5327 }, { "epoch": 0.36613523914238594, "grad_norm": 0.8853012323379517, "learning_rate": 2.1955948725773742e-05, "loss": 0.0323, "step": 5328 }, { "epoch": 0.36620395821880153, "grad_norm": 0.7120217084884644, "learning_rate": 2.1952990660168267e-05, "loss": 0.0285, "step": 5329 }, { "epoch": 0.3662726772952172, "grad_norm": 0.5140645503997803, "learning_rate": 2.195003225012592e-05, "loss": 0.0227, "step": 5330 }, { "epoch": 0.36634139637163277, "grad_norm": 1.1380597352981567, "learning_rate": 2.194707349579327e-05, "loss": 0.0292, "step": 5331 }, { "epoch": 0.36641011544804836, "grad_norm": 1.0172706842422485, "learning_rate": 2.1944114397316884e-05, "loss": 0.0318, "step": 5332 }, { "epoch": 0.366478834524464, "grad_norm": 0.5312811136245728, "learning_rate": 2.194115495484334e-05, "loss": 0.0266, "step": 5333 }, { "epoch": 0.3665475536008796, "grad_norm": 1.424964189529419, "learning_rate": 2.1938195168519254e-05, "loss": 0.0284, "step": 5334 }, { "epoch": 0.36661627267729524, "grad_norm": 0.8558735251426697, "learning_rate": 2.1935235038491246e-05, "loss": 0.0247, "step": 5335 }, { "epoch": 0.36668499175371083, "grad_norm": 0.615300714969635, "learning_rate": 2.1932274564905946e-05, "loss": 0.0339, "step": 5336 }, { "epoch": 0.3667537108301264, "grad_norm": 0.5713469386100769, "learning_rate": 2.1929313747910027e-05, "loss": 0.0258, "step": 5337 }, { "epoch": 0.36682242990654207, "grad_norm": 0.9428964853286743, "learning_rate": 2.1926352587650143e-05, "loss": 0.0343, "step": 5338 }, { "epoch": 0.36689114898295766, "grad_norm": 1.2140616178512573, "learning_rate": 2.1923391084272995e-05, "loss": 0.0344, "step": 5339 }, { "epoch": 0.3669598680593733, "grad_norm": 2.1741793155670166, "learning_rate": 2.1920429237925284e-05, "loss": 0.0267, "step": 5340 }, { "epoch": 0.3670285871357889, "grad_norm": 1.1846357583999634, "learning_rate": 2.191746704875374e-05, "loss": 0.0403, "step": 5341 }, { "epoch": 0.3670973062122045, "grad_norm": 1.679924726486206, "learning_rate": 2.1914504516905094e-05, "loss": 0.0257, "step": 5342 }, { "epoch": 0.36716602528862013, "grad_norm": 1.1608197689056396, "learning_rate": 2.1911541642526113e-05, "loss": 0.0516, "step": 5343 }, { "epoch": 0.3672347443650357, "grad_norm": 1.2372227907180786, "learning_rate": 2.1908578425763573e-05, "loss": 0.0333, "step": 5344 }, { "epoch": 0.36730346344145137, "grad_norm": 1.033060908317566, "learning_rate": 2.1905614866764252e-05, "loss": 0.0328, "step": 5345 }, { "epoch": 0.36737218251786696, "grad_norm": 0.6546362638473511, "learning_rate": 2.1902650965674975e-05, "loss": 0.0297, "step": 5346 }, { "epoch": 0.36744090159428255, "grad_norm": 1.0748800039291382, "learning_rate": 2.189968672264256e-05, "loss": 0.0359, "step": 5347 }, { "epoch": 0.3675096206706982, "grad_norm": 0.8909060955047607, "learning_rate": 2.1896722137813843e-05, "loss": 0.0367, "step": 5348 }, { "epoch": 0.3675783397471138, "grad_norm": 0.4342830181121826, "learning_rate": 2.18937572113357e-05, "loss": 0.0278, "step": 5349 }, { "epoch": 0.36764705882352944, "grad_norm": 0.7207202911376953, "learning_rate": 2.1890791943354992e-05, "loss": 0.0234, "step": 5350 }, { "epoch": 0.36771577789994503, "grad_norm": 0.561337411403656, "learning_rate": 2.1887826334018615e-05, "loss": 0.018, "step": 5351 }, { "epoch": 0.3677844969763606, "grad_norm": 0.591552734375, "learning_rate": 2.1884860383473487e-05, "loss": 0.0364, "step": 5352 }, { "epoch": 0.36785321605277627, "grad_norm": 0.6818831562995911, "learning_rate": 2.1881894091866532e-05, "loss": 0.0268, "step": 5353 }, { "epoch": 0.36792193512919186, "grad_norm": 0.6113133430480957, "learning_rate": 2.1878927459344685e-05, "loss": 0.0382, "step": 5354 }, { "epoch": 0.3679906542056075, "grad_norm": 0.4762003421783447, "learning_rate": 2.1875960486054924e-05, "loss": 0.0269, "step": 5355 }, { "epoch": 0.3680593732820231, "grad_norm": 0.6854296922683716, "learning_rate": 2.1872993172144213e-05, "loss": 0.0285, "step": 5356 }, { "epoch": 0.3681280923584387, "grad_norm": 0.5610504150390625, "learning_rate": 2.187002551775955e-05, "loss": 0.0214, "step": 5357 }, { "epoch": 0.36819681143485433, "grad_norm": 0.7972062826156616, "learning_rate": 2.1867057523047946e-05, "loss": 0.0212, "step": 5358 }, { "epoch": 0.3682655305112699, "grad_norm": 1.1879876852035522, "learning_rate": 2.1864089188156432e-05, "loss": 0.0354, "step": 5359 }, { "epoch": 0.3683342495876855, "grad_norm": 0.6721494197845459, "learning_rate": 2.1861120513232046e-05, "loss": 0.0277, "step": 5360 }, { "epoch": 0.36840296866410116, "grad_norm": 0.8404914140701294, "learning_rate": 2.1858151498421863e-05, "loss": 0.0235, "step": 5361 }, { "epoch": 0.36847168774051675, "grad_norm": 0.726772129535675, "learning_rate": 2.1855182143872952e-05, "loss": 0.0289, "step": 5362 }, { "epoch": 0.3685404068169324, "grad_norm": 0.44359713792800903, "learning_rate": 2.1852212449732408e-05, "loss": 0.0293, "step": 5363 }, { "epoch": 0.368609125893348, "grad_norm": 0.6045411229133606, "learning_rate": 2.184924241614735e-05, "loss": 0.0212, "step": 5364 }, { "epoch": 0.3686778449697636, "grad_norm": 0.9121903777122498, "learning_rate": 2.1846272043264906e-05, "loss": 0.0246, "step": 5365 }, { "epoch": 0.3687465640461792, "grad_norm": 0.942191481590271, "learning_rate": 2.1843301331232215e-05, "loss": 0.0237, "step": 5366 }, { "epoch": 0.3688152831225948, "grad_norm": 0.802879273891449, "learning_rate": 2.1840330280196453e-05, "loss": 0.0385, "step": 5367 }, { "epoch": 0.36888400219901046, "grad_norm": 0.7734142541885376, "learning_rate": 2.1837358890304785e-05, "loss": 0.0223, "step": 5368 }, { "epoch": 0.36895272127542605, "grad_norm": 1.2960902452468872, "learning_rate": 2.1834387161704417e-05, "loss": 0.0216, "step": 5369 }, { "epoch": 0.36902144035184165, "grad_norm": 2.105404853820801, "learning_rate": 2.1831415094542558e-05, "loss": 0.0241, "step": 5370 }, { "epoch": 0.3690901594282573, "grad_norm": 1.0624889135360718, "learning_rate": 2.1828442688966446e-05, "loss": 0.0251, "step": 5371 }, { "epoch": 0.3691588785046729, "grad_norm": 1.7540675401687622, "learning_rate": 2.1825469945123315e-05, "loss": 0.0337, "step": 5372 }, { "epoch": 0.36922759758108853, "grad_norm": 1.5412360429763794, "learning_rate": 2.1822496863160437e-05, "loss": 0.0319, "step": 5373 }, { "epoch": 0.3692963166575041, "grad_norm": 0.9377455115318298, "learning_rate": 2.1819523443225094e-05, "loss": 0.0268, "step": 5374 }, { "epoch": 0.3693650357339197, "grad_norm": 0.9268890023231506, "learning_rate": 2.1816549685464578e-05, "loss": 0.0206, "step": 5375 }, { "epoch": 0.36943375481033536, "grad_norm": 0.8599278926849365, "learning_rate": 2.1813575590026206e-05, "loss": 0.0282, "step": 5376 }, { "epoch": 0.36950247388675095, "grad_norm": 0.346017986536026, "learning_rate": 2.1810601157057307e-05, "loss": 0.023, "step": 5377 }, { "epoch": 0.3695711929631666, "grad_norm": 0.5836716890335083, "learning_rate": 2.1807626386705227e-05, "loss": 0.0241, "step": 5378 }, { "epoch": 0.3696399120395822, "grad_norm": 1.235376000404358, "learning_rate": 2.1804651279117333e-05, "loss": 0.0358, "step": 5379 }, { "epoch": 0.3697086311159978, "grad_norm": 0.7742201089859009, "learning_rate": 2.1801675834441006e-05, "loss": 0.0252, "step": 5380 }, { "epoch": 0.3697773501924134, "grad_norm": 0.9490507245063782, "learning_rate": 2.179870005282364e-05, "loss": 0.0384, "step": 5381 }, { "epoch": 0.369846069268829, "grad_norm": 0.5667466521263123, "learning_rate": 2.1795723934412648e-05, "loss": 0.0259, "step": 5382 }, { "epoch": 0.36991478834524466, "grad_norm": 0.8629729151725769, "learning_rate": 2.1792747479355468e-05, "loss": 0.0314, "step": 5383 }, { "epoch": 0.36998350742166025, "grad_norm": 0.7983411550521851, "learning_rate": 2.178977068779954e-05, "loss": 0.0223, "step": 5384 }, { "epoch": 0.37005222649807584, "grad_norm": 0.8794796466827393, "learning_rate": 2.1786793559892333e-05, "loss": 0.0291, "step": 5385 }, { "epoch": 0.3701209455744915, "grad_norm": 1.0409647226333618, "learning_rate": 2.1783816095781327e-05, "loss": 0.0423, "step": 5386 }, { "epoch": 0.3701896646509071, "grad_norm": 1.3415741920471191, "learning_rate": 2.1780838295614015e-05, "loss": 0.0401, "step": 5387 }, { "epoch": 0.3702583837273227, "grad_norm": 0.8647997379302979, "learning_rate": 2.177786015953792e-05, "loss": 0.0248, "step": 5388 }, { "epoch": 0.3703271028037383, "grad_norm": 0.9737774729728699, "learning_rate": 2.1774881687700563e-05, "loss": 0.0352, "step": 5389 }, { "epoch": 0.3703958218801539, "grad_norm": 0.6319352388381958, "learning_rate": 2.17719028802495e-05, "loss": 0.0239, "step": 5390 }, { "epoch": 0.37046454095656955, "grad_norm": 1.1555352210998535, "learning_rate": 2.1768923737332288e-05, "loss": 0.0374, "step": 5391 }, { "epoch": 0.37053326003298515, "grad_norm": 1.2848793268203735, "learning_rate": 2.1765944259096508e-05, "loss": 0.0261, "step": 5392 }, { "epoch": 0.3706019791094008, "grad_norm": 0.3822612166404724, "learning_rate": 2.1762964445689766e-05, "loss": 0.027, "step": 5393 }, { "epoch": 0.3706706981858164, "grad_norm": 0.6779327392578125, "learning_rate": 2.1759984297259665e-05, "loss": 0.0257, "step": 5394 }, { "epoch": 0.370739417262232, "grad_norm": 1.1460927724838257, "learning_rate": 2.175700381395384e-05, "loss": 0.0184, "step": 5395 }, { "epoch": 0.3708081363386476, "grad_norm": 1.5845845937728882, "learning_rate": 2.175402299591994e-05, "loss": 0.0248, "step": 5396 }, { "epoch": 0.3708768554150632, "grad_norm": 0.8274095058441162, "learning_rate": 2.1751041843305624e-05, "loss": 0.0312, "step": 5397 }, { "epoch": 0.37094557449147886, "grad_norm": 1.024478554725647, "learning_rate": 2.1748060356258575e-05, "loss": 0.0299, "step": 5398 }, { "epoch": 0.37101429356789445, "grad_norm": 0.9827120304107666, "learning_rate": 2.174507853492649e-05, "loss": 0.0275, "step": 5399 }, { "epoch": 0.37108301264431004, "grad_norm": 0.655639111995697, "learning_rate": 2.1742096379457078e-05, "loss": 0.0198, "step": 5400 }, { "epoch": 0.3711517317207257, "grad_norm": 0.9007812738418579, "learning_rate": 2.173911388999808e-05, "loss": 0.0305, "step": 5401 }, { "epoch": 0.3712204507971413, "grad_norm": 0.611903965473175, "learning_rate": 2.1736131066697226e-05, "loss": 0.0157, "step": 5402 }, { "epoch": 0.3712891698735569, "grad_norm": 1.1438896656036377, "learning_rate": 2.1733147909702298e-05, "loss": 0.034, "step": 5403 }, { "epoch": 0.3713578889499725, "grad_norm": 1.2197388410568237, "learning_rate": 2.173016441916106e-05, "loss": 0.0306, "step": 5404 }, { "epoch": 0.3714266080263881, "grad_norm": 0.9157600402832031, "learning_rate": 2.172718059522131e-05, "loss": 0.0254, "step": 5405 }, { "epoch": 0.37149532710280375, "grad_norm": 0.40982702374458313, "learning_rate": 2.1724196438030868e-05, "loss": 0.0298, "step": 5406 }, { "epoch": 0.37156404617921934, "grad_norm": 0.6788689494132996, "learning_rate": 2.172121194773756e-05, "loss": 0.0356, "step": 5407 }, { "epoch": 0.371632765255635, "grad_norm": 0.6206318140029907, "learning_rate": 2.1718227124489227e-05, "loss": 0.0277, "step": 5408 }, { "epoch": 0.3717014843320506, "grad_norm": 0.8812406063079834, "learning_rate": 2.1715241968433738e-05, "loss": 0.0239, "step": 5409 }, { "epoch": 0.37177020340846617, "grad_norm": 1.1518393754959106, "learning_rate": 2.1712256479718965e-05, "loss": 0.0331, "step": 5410 }, { "epoch": 0.3718389224848818, "grad_norm": 0.8638670444488525, "learning_rate": 2.1709270658492805e-05, "loss": 0.0261, "step": 5411 }, { "epoch": 0.3719076415612974, "grad_norm": 1.0654444694519043, "learning_rate": 2.1706284504903172e-05, "loss": 0.0312, "step": 5412 }, { "epoch": 0.37197636063771305, "grad_norm": 1.9520034790039062, "learning_rate": 2.1703298019097995e-05, "loss": 0.0396, "step": 5413 }, { "epoch": 0.37204507971412865, "grad_norm": 0.7281249165534973, "learning_rate": 2.170031120122521e-05, "loss": 0.0218, "step": 5414 }, { "epoch": 0.37211379879054424, "grad_norm": 0.8658756017684937, "learning_rate": 2.1697324051432788e-05, "loss": 0.0308, "step": 5415 }, { "epoch": 0.3721825178669599, "grad_norm": 1.0287035703659058, "learning_rate": 2.16943365698687e-05, "loss": 0.0267, "step": 5416 }, { "epoch": 0.3722512369433755, "grad_norm": 0.5316194891929626, "learning_rate": 2.1691348756680937e-05, "loss": 0.0262, "step": 5417 }, { "epoch": 0.3723199560197911, "grad_norm": 1.2394503355026245, "learning_rate": 2.1688360612017518e-05, "loss": 0.0186, "step": 5418 }, { "epoch": 0.3723886750962067, "grad_norm": 0.7257567048072815, "learning_rate": 2.1685372136026465e-05, "loss": 0.0251, "step": 5419 }, { "epoch": 0.3724573941726223, "grad_norm": 0.8953027725219727, "learning_rate": 2.1682383328855816e-05, "loss": 0.0193, "step": 5420 }, { "epoch": 0.37252611324903795, "grad_norm": 0.907901406288147, "learning_rate": 2.167939419065364e-05, "loss": 0.0241, "step": 5421 }, { "epoch": 0.37259483232545354, "grad_norm": 0.8579717874526978, "learning_rate": 2.1676404721568007e-05, "loss": 0.0276, "step": 5422 }, { "epoch": 0.3726635514018692, "grad_norm": 0.6675554513931274, "learning_rate": 2.167341492174701e-05, "loss": 0.0257, "step": 5423 }, { "epoch": 0.3727322704782848, "grad_norm": 0.8568239212036133, "learning_rate": 2.167042479133876e-05, "loss": 0.0363, "step": 5424 }, { "epoch": 0.37280098955470037, "grad_norm": 1.3427457809448242, "learning_rate": 2.1667434330491377e-05, "loss": 0.028, "step": 5425 }, { "epoch": 0.372869708631116, "grad_norm": 0.609880268573761, "learning_rate": 2.1664443539353006e-05, "loss": 0.0154, "step": 5426 }, { "epoch": 0.3729384277075316, "grad_norm": 0.9952265024185181, "learning_rate": 2.16614524180718e-05, "loss": 0.0244, "step": 5427 }, { "epoch": 0.3730071467839472, "grad_norm": 1.4434964656829834, "learning_rate": 2.1658460966795942e-05, "loss": 0.0325, "step": 5428 }, { "epoch": 0.37307586586036284, "grad_norm": 0.6739010810852051, "learning_rate": 2.1655469185673618e-05, "loss": 0.0214, "step": 5429 }, { "epoch": 0.37314458493677843, "grad_norm": 1.8021934032440186, "learning_rate": 2.165247707485303e-05, "loss": 0.0326, "step": 5430 }, { "epoch": 0.3732133040131941, "grad_norm": 0.9130816459655762, "learning_rate": 2.1649484634482402e-05, "loss": 0.0309, "step": 5431 }, { "epoch": 0.37328202308960967, "grad_norm": 2.7017691135406494, "learning_rate": 2.1646491864709978e-05, "loss": 0.0436, "step": 5432 }, { "epoch": 0.37335074216602526, "grad_norm": 0.5002400875091553, "learning_rate": 2.1643498765684016e-05, "loss": 0.0232, "step": 5433 }, { "epoch": 0.3734194612424409, "grad_norm": 1.1612082719802856, "learning_rate": 2.164050533755278e-05, "loss": 0.0347, "step": 5434 }, { "epoch": 0.3734881803188565, "grad_norm": 0.7745426893234253, "learning_rate": 2.163751158046456e-05, "loss": 0.0221, "step": 5435 }, { "epoch": 0.37355689939527215, "grad_norm": 0.5063185095787048, "learning_rate": 2.1634517494567668e-05, "loss": 0.0256, "step": 5436 }, { "epoch": 0.37362561847168774, "grad_norm": 0.7191257476806641, "learning_rate": 2.1631523080010414e-05, "loss": 0.0279, "step": 5437 }, { "epoch": 0.37369433754810333, "grad_norm": 0.8312621712684631, "learning_rate": 2.162852833694114e-05, "loss": 0.0365, "step": 5438 }, { "epoch": 0.373763056624519, "grad_norm": 0.7573960423469543, "learning_rate": 2.1625533265508205e-05, "loss": 0.0299, "step": 5439 }, { "epoch": 0.37383177570093457, "grad_norm": 1.0515004396438599, "learning_rate": 2.1622537865859972e-05, "loss": 0.0236, "step": 5440 }, { "epoch": 0.3739004947773502, "grad_norm": 0.8434526324272156, "learning_rate": 2.1619542138144823e-05, "loss": 0.0247, "step": 5441 }, { "epoch": 0.3739692138537658, "grad_norm": 1.7516276836395264, "learning_rate": 2.161654608251117e-05, "loss": 0.0339, "step": 5442 }, { "epoch": 0.3740379329301814, "grad_norm": 1.2512214183807373, "learning_rate": 2.161354969910743e-05, "loss": 0.0247, "step": 5443 }, { "epoch": 0.37410665200659704, "grad_norm": 2.374812364578247, "learning_rate": 2.161055298808203e-05, "loss": 0.0349, "step": 5444 }, { "epoch": 0.37417537108301263, "grad_norm": 1.8529242277145386, "learning_rate": 2.160755594958343e-05, "loss": 0.0339, "step": 5445 }, { "epoch": 0.3742440901594283, "grad_norm": 0.6796116232872009, "learning_rate": 2.160455858376009e-05, "loss": 0.0241, "step": 5446 }, { "epoch": 0.37431280923584387, "grad_norm": 0.8544365763664246, "learning_rate": 2.1601560890760492e-05, "loss": 0.0312, "step": 5447 }, { "epoch": 0.37438152831225946, "grad_norm": 0.6206398606300354, "learning_rate": 2.1598562870733146e-05, "loss": 0.0171, "step": 5448 }, { "epoch": 0.3744502473886751, "grad_norm": 0.8523533344268799, "learning_rate": 2.1595564523826562e-05, "loss": 0.0382, "step": 5449 }, { "epoch": 0.3745189664650907, "grad_norm": 0.7785633206367493, "learning_rate": 2.1592565850189268e-05, "loss": 0.026, "step": 5450 }, { "epoch": 0.37458768554150634, "grad_norm": 1.0107609033584595, "learning_rate": 2.1589566849969818e-05, "loss": 0.0371, "step": 5451 }, { "epoch": 0.37465640461792193, "grad_norm": 1.8867695331573486, "learning_rate": 2.1586567523316773e-05, "loss": 0.0387, "step": 5452 }, { "epoch": 0.3747251236943375, "grad_norm": 0.9461925029754639, "learning_rate": 2.158356787037871e-05, "loss": 0.0427, "step": 5453 }, { "epoch": 0.37479384277075317, "grad_norm": 0.6285480856895447, "learning_rate": 2.1580567891304237e-05, "loss": 0.0255, "step": 5454 }, { "epoch": 0.37486256184716876, "grad_norm": 1.879655361175537, "learning_rate": 2.1577567586241956e-05, "loss": 0.0358, "step": 5455 }, { "epoch": 0.3749312809235844, "grad_norm": 1.2096000909805298, "learning_rate": 2.1574566955340498e-05, "loss": 0.0412, "step": 5456 }, { "epoch": 0.375, "grad_norm": 0.9470233917236328, "learning_rate": 2.157156599874851e-05, "loss": 0.0225, "step": 5457 }, { "epoch": 0.3750687190764156, "grad_norm": 0.9882782101631165, "learning_rate": 2.156856471661466e-05, "loss": 0.0244, "step": 5458 }, { "epoch": 0.37513743815283124, "grad_norm": 0.6149110198020935, "learning_rate": 2.156556310908761e-05, "loss": 0.0249, "step": 5459 }, { "epoch": 0.37520615722924683, "grad_norm": 1.245532751083374, "learning_rate": 2.1562561176316067e-05, "loss": 0.0388, "step": 5460 }, { "epoch": 0.3752748763056625, "grad_norm": 0.4865119457244873, "learning_rate": 2.1559558918448736e-05, "loss": 0.0236, "step": 5461 }, { "epoch": 0.37534359538207807, "grad_norm": 0.7498254776000977, "learning_rate": 2.155655633563433e-05, "loss": 0.0245, "step": 5462 }, { "epoch": 0.37541231445849366, "grad_norm": 0.5130119919776917, "learning_rate": 2.1553553428021616e-05, "loss": 0.0351, "step": 5463 }, { "epoch": 0.3754810335349093, "grad_norm": 0.9991954565048218, "learning_rate": 2.1550550195759336e-05, "loss": 0.0198, "step": 5464 }, { "epoch": 0.3755497526113249, "grad_norm": 0.6871040463447571, "learning_rate": 2.1547546638996264e-05, "loss": 0.028, "step": 5465 }, { "epoch": 0.37561847168774054, "grad_norm": 0.9715216159820557, "learning_rate": 2.1544542757881196e-05, "loss": 0.0315, "step": 5466 }, { "epoch": 0.37568719076415613, "grad_norm": 1.3887405395507812, "learning_rate": 2.154153855256293e-05, "loss": 0.0333, "step": 5467 }, { "epoch": 0.3757559098405717, "grad_norm": 0.6970604062080383, "learning_rate": 2.1538534023190302e-05, "loss": 0.036, "step": 5468 }, { "epoch": 0.37582462891698737, "grad_norm": 0.49125364422798157, "learning_rate": 2.1535529169912136e-05, "loss": 0.0251, "step": 5469 }, { "epoch": 0.37589334799340296, "grad_norm": 1.4630388021469116, "learning_rate": 2.153252399287729e-05, "loss": 0.0306, "step": 5470 }, { "epoch": 0.3759620670698186, "grad_norm": 1.3389286994934082, "learning_rate": 2.1529518492234634e-05, "loss": 0.0413, "step": 5471 }, { "epoch": 0.3760307861462342, "grad_norm": 0.7484680414199829, "learning_rate": 2.152651266813306e-05, "loss": 0.0362, "step": 5472 }, { "epoch": 0.3760995052226498, "grad_norm": 0.4900476038455963, "learning_rate": 2.1523506520721464e-05, "loss": 0.026, "step": 5473 }, { "epoch": 0.37616822429906543, "grad_norm": 0.8854719400405884, "learning_rate": 2.1520500050148764e-05, "loss": 0.0261, "step": 5474 }, { "epoch": 0.376236943375481, "grad_norm": 0.7371109127998352, "learning_rate": 2.1517493256563904e-05, "loss": 0.0181, "step": 5475 }, { "epoch": 0.37630566245189667, "grad_norm": 0.4422070384025574, "learning_rate": 2.1514486140115825e-05, "loss": 0.0165, "step": 5476 }, { "epoch": 0.37637438152831226, "grad_norm": 1.1205799579620361, "learning_rate": 2.151147870095349e-05, "loss": 0.0379, "step": 5477 }, { "epoch": 0.37644310060472785, "grad_norm": 0.6808006763458252, "learning_rate": 2.1508470939225894e-05, "loss": 0.0227, "step": 5478 }, { "epoch": 0.3765118196811435, "grad_norm": 1.1204878091812134, "learning_rate": 2.150546285508203e-05, "loss": 0.0208, "step": 5479 }, { "epoch": 0.3765805387575591, "grad_norm": 0.7570928335189819, "learning_rate": 2.1502454448670906e-05, "loss": 0.0318, "step": 5480 }, { "epoch": 0.37664925783397474, "grad_norm": 0.5727471113204956, "learning_rate": 2.1499445720141556e-05, "loss": 0.0212, "step": 5481 }, { "epoch": 0.37671797691039033, "grad_norm": 0.5564741492271423, "learning_rate": 2.149643666964303e-05, "loss": 0.0207, "step": 5482 }, { "epoch": 0.3767866959868059, "grad_norm": 0.45984676480293274, "learning_rate": 2.1493427297324388e-05, "loss": 0.0206, "step": 5483 }, { "epoch": 0.37685541506322157, "grad_norm": 1.0856926441192627, "learning_rate": 2.1490417603334704e-05, "loss": 0.0296, "step": 5484 }, { "epoch": 0.37692413413963716, "grad_norm": 0.940888524055481, "learning_rate": 2.148740758782308e-05, "loss": 0.0341, "step": 5485 }, { "epoch": 0.3769928532160528, "grad_norm": 0.6703576445579529, "learning_rate": 2.1484397250938622e-05, "loss": 0.0263, "step": 5486 }, { "epoch": 0.3770615722924684, "grad_norm": 0.8454474806785583, "learning_rate": 2.1481386592830454e-05, "loss": 0.0258, "step": 5487 }, { "epoch": 0.377130291368884, "grad_norm": 0.9350568652153015, "learning_rate": 2.147837561364772e-05, "loss": 0.0354, "step": 5488 }, { "epoch": 0.37719901044529963, "grad_norm": 0.45727893710136414, "learning_rate": 2.1475364313539575e-05, "loss": 0.0211, "step": 5489 }, { "epoch": 0.3772677295217152, "grad_norm": 0.9581918120384216, "learning_rate": 2.1472352692655196e-05, "loss": 0.0234, "step": 5490 }, { "epoch": 0.3773364485981308, "grad_norm": 0.502409815788269, "learning_rate": 2.1469340751143773e-05, "loss": 0.0239, "step": 5491 }, { "epoch": 0.37740516767454646, "grad_norm": 0.5083873867988586, "learning_rate": 2.1466328489154505e-05, "loss": 0.014, "step": 5492 }, { "epoch": 0.37747388675096205, "grad_norm": 0.6518163084983826, "learning_rate": 2.1463315906836617e-05, "loss": 0.0363, "step": 5493 }, { "epoch": 0.3775426058273777, "grad_norm": 0.6250990033149719, "learning_rate": 2.1460303004339353e-05, "loss": 0.0237, "step": 5494 }, { "epoch": 0.3776113249037933, "grad_norm": 1.0544703006744385, "learning_rate": 2.1457289781811958e-05, "loss": 0.0273, "step": 5495 }, { "epoch": 0.3776800439802089, "grad_norm": 0.738479495048523, "learning_rate": 2.14542762394037e-05, "loss": 0.0201, "step": 5496 }, { "epoch": 0.3777487630566245, "grad_norm": 0.4399369955062866, "learning_rate": 2.1451262377263872e-05, "loss": 0.0121, "step": 5497 }, { "epoch": 0.3778174821330401, "grad_norm": 0.7972975373268127, "learning_rate": 2.1448248195541764e-05, "loss": 0.0246, "step": 5498 }, { "epoch": 0.37788620120945576, "grad_norm": 0.295508474111557, "learning_rate": 2.14452336943867e-05, "loss": 0.0163, "step": 5499 }, { "epoch": 0.37795492028587135, "grad_norm": 0.4093400239944458, "learning_rate": 2.144221887394801e-05, "loss": 0.0144, "step": 5500 }, { "epoch": 0.37802363936228695, "grad_norm": 0.5883690714836121, "learning_rate": 2.1439203734375042e-05, "loss": 0.0194, "step": 5501 }, { "epoch": 0.3780923584387026, "grad_norm": 0.8922170400619507, "learning_rate": 2.1436188275817154e-05, "loss": 0.0287, "step": 5502 }, { "epoch": 0.3781610775151182, "grad_norm": 0.8438226580619812, "learning_rate": 2.1433172498423738e-05, "loss": 0.0233, "step": 5503 }, { "epoch": 0.37822979659153383, "grad_norm": 0.8021788597106934, "learning_rate": 2.1430156402344177e-05, "loss": 0.0191, "step": 5504 }, { "epoch": 0.3782985156679494, "grad_norm": 1.4746662378311157, "learning_rate": 2.142713998772789e-05, "loss": 0.0198, "step": 5505 }, { "epoch": 0.378367234744365, "grad_norm": 0.705558717250824, "learning_rate": 2.1424123254724305e-05, "loss": 0.0376, "step": 5506 }, { "epoch": 0.37843595382078066, "grad_norm": 1.2310789823532104, "learning_rate": 2.1421106203482857e-05, "loss": 0.0323, "step": 5507 }, { "epoch": 0.37850467289719625, "grad_norm": 1.0256059169769287, "learning_rate": 2.141808883415301e-05, "loss": 0.0212, "step": 5508 }, { "epoch": 0.3785733919736119, "grad_norm": 0.7489541172981262, "learning_rate": 2.141507114688424e-05, "loss": 0.0192, "step": 5509 }, { "epoch": 0.3786421110500275, "grad_norm": 0.8969508409500122, "learning_rate": 2.141205314182603e-05, "loss": 0.0258, "step": 5510 }, { "epoch": 0.3787108301264431, "grad_norm": 0.9775189757347107, "learning_rate": 2.1409034819127888e-05, "loss": 0.0343, "step": 5511 }, { "epoch": 0.3787795492028587, "grad_norm": 0.7263962626457214, "learning_rate": 2.1406016178939344e-05, "loss": 0.0261, "step": 5512 }, { "epoch": 0.3788482682792743, "grad_norm": 0.7558745741844177, "learning_rate": 2.1402997221409927e-05, "loss": 0.0288, "step": 5513 }, { "epoch": 0.37891698735568996, "grad_norm": 0.6909937262535095, "learning_rate": 2.1399977946689188e-05, "loss": 0.0322, "step": 5514 }, { "epoch": 0.37898570643210555, "grad_norm": 0.7841758131980896, "learning_rate": 2.1396958354926704e-05, "loss": 0.029, "step": 5515 }, { "epoch": 0.37905442550852114, "grad_norm": 0.903020977973938, "learning_rate": 2.139393844627205e-05, "loss": 0.0394, "step": 5516 }, { "epoch": 0.3791231445849368, "grad_norm": 0.9132198691368103, "learning_rate": 2.139091822087483e-05, "loss": 0.0304, "step": 5517 }, { "epoch": 0.3791918636613524, "grad_norm": 1.044950246810913, "learning_rate": 2.1387897678884667e-05, "loss": 0.0275, "step": 5518 }, { "epoch": 0.379260582737768, "grad_norm": 0.31975555419921875, "learning_rate": 2.1384876820451183e-05, "loss": 0.0184, "step": 5519 }, { "epoch": 0.3793293018141836, "grad_norm": 0.8644630908966064, "learning_rate": 2.1381855645724024e-05, "loss": 0.0214, "step": 5520 }, { "epoch": 0.3793980208905992, "grad_norm": 1.118111252784729, "learning_rate": 2.137883415485286e-05, "loss": 0.0278, "step": 5521 }, { "epoch": 0.37946673996701485, "grad_norm": 0.19199374318122864, "learning_rate": 2.1375812347987368e-05, "loss": 0.0144, "step": 5522 }, { "epoch": 0.37953545904343045, "grad_norm": 1.1544930934906006, "learning_rate": 2.137279022527724e-05, "loss": 0.0251, "step": 5523 }, { "epoch": 0.3796041781198461, "grad_norm": 0.8049165606498718, "learning_rate": 2.1369767786872188e-05, "loss": 0.0297, "step": 5524 }, { "epoch": 0.3796728971962617, "grad_norm": 0.8601580858230591, "learning_rate": 2.1366745032921934e-05, "loss": 0.0177, "step": 5525 }, { "epoch": 0.3797416162726773, "grad_norm": 0.8886396288871765, "learning_rate": 2.1363721963576215e-05, "loss": 0.0235, "step": 5526 }, { "epoch": 0.3798103353490929, "grad_norm": 1.0678675174713135, "learning_rate": 2.13606985789848e-05, "loss": 0.0291, "step": 5527 }, { "epoch": 0.3798790544255085, "grad_norm": 0.6001869440078735, "learning_rate": 2.135767487929745e-05, "loss": 0.0209, "step": 5528 }, { "epoch": 0.37994777350192416, "grad_norm": 1.1320438385009766, "learning_rate": 2.135465086466396e-05, "loss": 0.0209, "step": 5529 }, { "epoch": 0.38001649257833975, "grad_norm": 0.5469939708709717, "learning_rate": 2.135162653523413e-05, "loss": 0.0323, "step": 5530 }, { "epoch": 0.38008521165475534, "grad_norm": 1.5024629831314087, "learning_rate": 2.134860189115778e-05, "loss": 0.033, "step": 5531 }, { "epoch": 0.380153930731171, "grad_norm": 0.5963106155395508, "learning_rate": 2.1345576932584744e-05, "loss": 0.0227, "step": 5532 }, { "epoch": 0.3802226498075866, "grad_norm": 1.1078685522079468, "learning_rate": 2.1342551659664876e-05, "loss": 0.0335, "step": 5533 }, { "epoch": 0.3802913688840022, "grad_norm": 1.1956709623336792, "learning_rate": 2.1339526072548032e-05, "loss": 0.0358, "step": 5534 }, { "epoch": 0.3803600879604178, "grad_norm": 2.3323144912719727, "learning_rate": 2.1336500171384105e-05, "loss": 0.0294, "step": 5535 }, { "epoch": 0.3804288070368334, "grad_norm": 1.3025131225585938, "learning_rate": 2.1333473956322984e-05, "loss": 0.0494, "step": 5536 }, { "epoch": 0.38049752611324905, "grad_norm": 0.8425405025482178, "learning_rate": 2.1330447427514588e-05, "loss": 0.0252, "step": 5537 }, { "epoch": 0.38056624518966464, "grad_norm": 0.803982675075531, "learning_rate": 2.1327420585108835e-05, "loss": 0.0235, "step": 5538 }, { "epoch": 0.3806349642660803, "grad_norm": 0.7422731518745422, "learning_rate": 2.132439342925568e-05, "loss": 0.0346, "step": 5539 }, { "epoch": 0.3807036833424959, "grad_norm": 1.9366426467895508, "learning_rate": 2.1321365960105076e-05, "loss": 0.0421, "step": 5540 }, { "epoch": 0.38077240241891147, "grad_norm": 0.7737401723861694, "learning_rate": 2.1318338177807e-05, "loss": 0.0285, "step": 5541 }, { "epoch": 0.3808411214953271, "grad_norm": 0.6962565183639526, "learning_rate": 2.1315310082511434e-05, "loss": 0.0209, "step": 5542 }, { "epoch": 0.3809098405717427, "grad_norm": 1.1550006866455078, "learning_rate": 2.13122816743684e-05, "loss": 0.0348, "step": 5543 }, { "epoch": 0.38097855964815835, "grad_norm": 0.7380654215812683, "learning_rate": 2.13092529535279e-05, "loss": 0.0235, "step": 5544 }, { "epoch": 0.38104727872457395, "grad_norm": 0.8158470392227173, "learning_rate": 2.1306223920139987e-05, "loss": 0.0222, "step": 5545 }, { "epoch": 0.38111599780098954, "grad_norm": 1.1323961019515991, "learning_rate": 2.1303194574354708e-05, "loss": 0.0331, "step": 5546 }, { "epoch": 0.3811847168774052, "grad_norm": 0.6834646463394165, "learning_rate": 2.1300164916322117e-05, "loss": 0.0132, "step": 5547 }, { "epoch": 0.3812534359538208, "grad_norm": 0.8115816712379456, "learning_rate": 2.129713494619232e-05, "loss": 0.0198, "step": 5548 }, { "epoch": 0.3813221550302364, "grad_norm": 0.72989422082901, "learning_rate": 2.12941046641154e-05, "loss": 0.0248, "step": 5549 }, { "epoch": 0.381390874106652, "grad_norm": 0.9226238131523132, "learning_rate": 2.1291074070241473e-05, "loss": 0.0271, "step": 5550 }, { "epoch": 0.3814595931830676, "grad_norm": 1.489145278930664, "learning_rate": 2.1288043164720672e-05, "loss": 0.0234, "step": 5551 }, { "epoch": 0.38152831225948325, "grad_norm": 0.613585352897644, "learning_rate": 2.1285011947703142e-05, "loss": 0.0229, "step": 5552 }, { "epoch": 0.38159703133589884, "grad_norm": 0.6668891906738281, "learning_rate": 2.1281980419339035e-05, "loss": 0.0328, "step": 5553 }, { "epoch": 0.3816657504123145, "grad_norm": 0.9860650300979614, "learning_rate": 2.127894857977854e-05, "loss": 0.0181, "step": 5554 }, { "epoch": 0.3817344694887301, "grad_norm": 0.6171995997428894, "learning_rate": 2.1275916429171843e-05, "loss": 0.0271, "step": 5555 }, { "epoch": 0.38180318856514567, "grad_norm": 0.9577427506446838, "learning_rate": 2.1272883967669143e-05, "loss": 0.0343, "step": 5556 }, { "epoch": 0.3818719076415613, "grad_norm": 1.1210917234420776, "learning_rate": 2.1269851195420673e-05, "loss": 0.0298, "step": 5557 }, { "epoch": 0.3819406267179769, "grad_norm": 0.47283461689949036, "learning_rate": 2.126681811257666e-05, "loss": 0.0274, "step": 5558 }, { "epoch": 0.3820093457943925, "grad_norm": 0.7469847798347473, "learning_rate": 2.1263784719287362e-05, "loss": 0.028, "step": 5559 }, { "epoch": 0.38207806487080814, "grad_norm": 0.9363287687301636, "learning_rate": 2.126075101570305e-05, "loss": 0.019, "step": 5560 }, { "epoch": 0.38214678394722373, "grad_norm": 1.2539091110229492, "learning_rate": 2.1257717001974005e-05, "loss": 0.0389, "step": 5561 }, { "epoch": 0.3822155030236394, "grad_norm": 1.0189698934555054, "learning_rate": 2.1254682678250522e-05, "loss": 0.0271, "step": 5562 }, { "epoch": 0.38228422210005497, "grad_norm": 0.7465350031852722, "learning_rate": 2.1251648044682922e-05, "loss": 0.0399, "step": 5563 }, { "epoch": 0.38235294117647056, "grad_norm": 1.0285663604736328, "learning_rate": 2.1248613101421527e-05, "loss": 0.0244, "step": 5564 }, { "epoch": 0.3824216602528862, "grad_norm": 0.8736835718154907, "learning_rate": 2.1245577848616684e-05, "loss": 0.0295, "step": 5565 }, { "epoch": 0.3824903793293018, "grad_norm": 1.086834192276001, "learning_rate": 2.124254228641876e-05, "loss": 0.0389, "step": 5566 }, { "epoch": 0.38255909840571745, "grad_norm": 0.8535642027854919, "learning_rate": 2.1239506414978126e-05, "loss": 0.0248, "step": 5567 }, { "epoch": 0.38262781748213304, "grad_norm": 0.7462868690490723, "learning_rate": 2.123647023444516e-05, "loss": 0.0248, "step": 5568 }, { "epoch": 0.38269653655854863, "grad_norm": 1.4167978763580322, "learning_rate": 2.1233433744970294e-05, "loss": 0.0234, "step": 5569 }, { "epoch": 0.3827652556349643, "grad_norm": 2.0044443607330322, "learning_rate": 2.123039694670393e-05, "loss": 0.0455, "step": 5570 }, { "epoch": 0.38283397471137987, "grad_norm": 1.1865071058273315, "learning_rate": 2.1227359839796508e-05, "loss": 0.023, "step": 5571 }, { "epoch": 0.3829026937877955, "grad_norm": 0.8284863233566284, "learning_rate": 2.122432242439849e-05, "loss": 0.0293, "step": 5572 }, { "epoch": 0.3829714128642111, "grad_norm": 0.8539182543754578, "learning_rate": 2.1221284700660327e-05, "loss": 0.0285, "step": 5573 }, { "epoch": 0.3830401319406267, "grad_norm": 1.2572340965270996, "learning_rate": 2.1218246668732514e-05, "loss": 0.037, "step": 5574 }, { "epoch": 0.38310885101704234, "grad_norm": 0.5548664927482605, "learning_rate": 2.1215208328765547e-05, "loss": 0.0248, "step": 5575 }, { "epoch": 0.38317757009345793, "grad_norm": 0.7479032874107361, "learning_rate": 2.1212169680909937e-05, "loss": 0.0316, "step": 5576 }, { "epoch": 0.3832462891698736, "grad_norm": 1.8582311868667603, "learning_rate": 2.120913072531621e-05, "loss": 0.0406, "step": 5577 }, { "epoch": 0.38331500824628917, "grad_norm": 0.5711485147476196, "learning_rate": 2.1206091462134917e-05, "loss": 0.0218, "step": 5578 }, { "epoch": 0.38338372732270476, "grad_norm": 0.8142051696777344, "learning_rate": 2.1203051891516607e-05, "loss": 0.0247, "step": 5579 }, { "epoch": 0.3834524463991204, "grad_norm": 1.1082127094268799, "learning_rate": 2.1200012013611863e-05, "loss": 0.028, "step": 5580 }, { "epoch": 0.383521165475536, "grad_norm": 0.6999744176864624, "learning_rate": 2.119697182857127e-05, "loss": 0.0274, "step": 5581 }, { "epoch": 0.38358988455195164, "grad_norm": 0.5082419514656067, "learning_rate": 2.1193931336545433e-05, "loss": 0.0241, "step": 5582 }, { "epoch": 0.38365860362836723, "grad_norm": 0.6599092483520508, "learning_rate": 2.1190890537684965e-05, "loss": 0.0418, "step": 5583 }, { "epoch": 0.3837273227047828, "grad_norm": 1.021379828453064, "learning_rate": 2.1187849432140517e-05, "loss": 0.026, "step": 5584 }, { "epoch": 0.38379604178119847, "grad_norm": 0.6619403958320618, "learning_rate": 2.1184808020062727e-05, "loss": 0.0186, "step": 5585 }, { "epoch": 0.38386476085761406, "grad_norm": 0.6498317122459412, "learning_rate": 2.118176630160226e-05, "loss": 0.023, "step": 5586 }, { "epoch": 0.3839334799340297, "grad_norm": 1.5527935028076172, "learning_rate": 2.1178724276909802e-05, "loss": 0.0506, "step": 5587 }, { "epoch": 0.3840021990104453, "grad_norm": 0.6697441339492798, "learning_rate": 2.117568194613605e-05, "loss": 0.0202, "step": 5588 }, { "epoch": 0.3840709180868609, "grad_norm": 0.724082350730896, "learning_rate": 2.1172639309431704e-05, "loss": 0.0298, "step": 5589 }, { "epoch": 0.38413963716327654, "grad_norm": 0.5443500876426697, "learning_rate": 2.11695963669475e-05, "loss": 0.0229, "step": 5590 }, { "epoch": 0.38420835623969213, "grad_norm": 0.6699874401092529, "learning_rate": 2.116655311883418e-05, "loss": 0.0224, "step": 5591 }, { "epoch": 0.3842770753161078, "grad_norm": 0.9563148021697998, "learning_rate": 2.116350956524249e-05, "loss": 0.0354, "step": 5592 }, { "epoch": 0.38434579439252337, "grad_norm": 0.5269034504890442, "learning_rate": 2.116046570632321e-05, "loss": 0.0185, "step": 5593 }, { "epoch": 0.38441451346893896, "grad_norm": 1.4734925031661987, "learning_rate": 2.1157421542227127e-05, "loss": 0.0583, "step": 5594 }, { "epoch": 0.3844832325453546, "grad_norm": 0.7078639268875122, "learning_rate": 2.1154377073105034e-05, "loss": 0.0231, "step": 5595 }, { "epoch": 0.3845519516217702, "grad_norm": 0.8553537726402283, "learning_rate": 2.115133229910776e-05, "loss": 0.0207, "step": 5596 }, { "epoch": 0.38462067069818584, "grad_norm": 1.087604284286499, "learning_rate": 2.1148287220386128e-05, "loss": 0.0393, "step": 5597 }, { "epoch": 0.38468938977460143, "grad_norm": 0.9362315535545349, "learning_rate": 2.1145241837090987e-05, "loss": 0.0336, "step": 5598 }, { "epoch": 0.384758108851017, "grad_norm": 1.0192283391952515, "learning_rate": 2.11421961493732e-05, "loss": 0.0303, "step": 5599 }, { "epoch": 0.38482682792743267, "grad_norm": 1.0451674461364746, "learning_rate": 2.1139150157383644e-05, "loss": 0.0322, "step": 5600 }, { "epoch": 0.38489554700384826, "grad_norm": 1.5159544944763184, "learning_rate": 2.1136103861273208e-05, "loss": 0.027, "step": 5601 }, { "epoch": 0.3849642660802639, "grad_norm": 1.1115564107894897, "learning_rate": 2.1133057261192806e-05, "loss": 0.0337, "step": 5602 }, { "epoch": 0.3850329851566795, "grad_norm": 1.0373002290725708, "learning_rate": 2.1130010357293354e-05, "loss": 0.0291, "step": 5603 }, { "epoch": 0.3851017042330951, "grad_norm": 0.8103976249694824, "learning_rate": 2.1126963149725794e-05, "loss": 0.0289, "step": 5604 }, { "epoch": 0.38517042330951073, "grad_norm": 0.8641250133514404, "learning_rate": 2.1123915638641074e-05, "loss": 0.0318, "step": 5605 }, { "epoch": 0.3852391423859263, "grad_norm": 1.1548364162445068, "learning_rate": 2.1120867824190162e-05, "loss": 0.0417, "step": 5606 }, { "epoch": 0.38530786146234197, "grad_norm": 0.8881511688232422, "learning_rate": 2.1117819706524044e-05, "loss": 0.0355, "step": 5607 }, { "epoch": 0.38537658053875756, "grad_norm": 0.8017042875289917, "learning_rate": 2.1114771285793715e-05, "loss": 0.0317, "step": 5608 }, { "epoch": 0.38544529961517315, "grad_norm": 1.0193006992340088, "learning_rate": 2.111172256215019e-05, "loss": 0.0205, "step": 5609 }, { "epoch": 0.3855140186915888, "grad_norm": 0.47844481468200684, "learning_rate": 2.110867353574449e-05, "loss": 0.021, "step": 5610 }, { "epoch": 0.3855827377680044, "grad_norm": 0.8620095252990723, "learning_rate": 2.1105624206727663e-05, "loss": 0.031, "step": 5611 }, { "epoch": 0.38565145684442004, "grad_norm": 0.8061808347702026, "learning_rate": 2.110257457525077e-05, "loss": 0.0341, "step": 5612 }, { "epoch": 0.38572017592083563, "grad_norm": 0.7691959738731384, "learning_rate": 2.1099524641464874e-05, "loss": 0.0248, "step": 5613 }, { "epoch": 0.3857888949972512, "grad_norm": 1.1346696615219116, "learning_rate": 2.1096474405521068e-05, "loss": 0.0281, "step": 5614 }, { "epoch": 0.38585761407366687, "grad_norm": 0.6616771817207336, "learning_rate": 2.1093423867570457e-05, "loss": 0.0357, "step": 5615 }, { "epoch": 0.38592633315008246, "grad_norm": 0.2984117567539215, "learning_rate": 2.1090373027764154e-05, "loss": 0.0261, "step": 5616 }, { "epoch": 0.3859950522264981, "grad_norm": 0.70329350233078, "learning_rate": 2.108732188625329e-05, "loss": 0.0215, "step": 5617 }, { "epoch": 0.3860637713029137, "grad_norm": 1.0100117921829224, "learning_rate": 2.108427044318902e-05, "loss": 0.0302, "step": 5618 }, { "epoch": 0.3861324903793293, "grad_norm": 0.8252615928649902, "learning_rate": 2.1081218698722496e-05, "loss": 0.0414, "step": 5619 }, { "epoch": 0.38620120945574493, "grad_norm": 0.846275806427002, "learning_rate": 2.10781666530049e-05, "loss": 0.0258, "step": 5620 }, { "epoch": 0.3862699285321605, "grad_norm": 0.34864187240600586, "learning_rate": 2.1075114306187427e-05, "loss": 0.0171, "step": 5621 }, { "epoch": 0.3863386476085761, "grad_norm": 0.49658578634262085, "learning_rate": 2.1072061658421276e-05, "loss": 0.0219, "step": 5622 }, { "epoch": 0.38640736668499176, "grad_norm": 1.20703125, "learning_rate": 2.106900870985768e-05, "loss": 0.0308, "step": 5623 }, { "epoch": 0.38647608576140735, "grad_norm": 0.7748522162437439, "learning_rate": 2.106595546064787e-05, "loss": 0.0303, "step": 5624 }, { "epoch": 0.386544804837823, "grad_norm": 1.118436574935913, "learning_rate": 2.106290191094309e-05, "loss": 0.0312, "step": 5625 }, { "epoch": 0.3866135239142386, "grad_norm": 0.688827633857727, "learning_rate": 2.105984806089462e-05, "loss": 0.0293, "step": 5626 }, { "epoch": 0.3866822429906542, "grad_norm": 1.4598772525787354, "learning_rate": 2.1056793910653737e-05, "loss": 0.0275, "step": 5627 }, { "epoch": 0.3867509620670698, "grad_norm": 1.010524034500122, "learning_rate": 2.1053739460371736e-05, "loss": 0.028, "step": 5628 }, { "epoch": 0.3868196811434854, "grad_norm": 1.4996237754821777, "learning_rate": 2.1050684710199926e-05, "loss": 0.0235, "step": 5629 }, { "epoch": 0.38688840021990106, "grad_norm": 0.787100613117218, "learning_rate": 2.1047629660289637e-05, "loss": 0.0273, "step": 5630 }, { "epoch": 0.38695711929631665, "grad_norm": 0.7341209053993225, "learning_rate": 2.10445743107922e-05, "loss": 0.0366, "step": 5631 }, { "epoch": 0.38702583837273224, "grad_norm": 0.7387838959693909, "learning_rate": 2.1041518661858986e-05, "loss": 0.0282, "step": 5632 }, { "epoch": 0.3870945574491479, "grad_norm": 1.2035263776779175, "learning_rate": 2.1038462713641356e-05, "loss": 0.0372, "step": 5633 }, { "epoch": 0.3871632765255635, "grad_norm": 0.6386862993240356, "learning_rate": 2.10354064662907e-05, "loss": 0.024, "step": 5634 }, { "epoch": 0.38723199560197913, "grad_norm": 0.9567555785179138, "learning_rate": 2.1032349919958417e-05, "loss": 0.0296, "step": 5635 }, { "epoch": 0.3873007146783947, "grad_norm": 0.7846077084541321, "learning_rate": 2.1029293074795923e-05, "loss": 0.025, "step": 5636 }, { "epoch": 0.3873694337548103, "grad_norm": 0.6164476275444031, "learning_rate": 2.1026235930954635e-05, "loss": 0.0324, "step": 5637 }, { "epoch": 0.38743815283122596, "grad_norm": 1.7703629732131958, "learning_rate": 2.1023178488586018e-05, "loss": 0.0311, "step": 5638 }, { "epoch": 0.38750687190764155, "grad_norm": 1.2203271389007568, "learning_rate": 2.102012074784152e-05, "loss": 0.0326, "step": 5639 }, { "epoch": 0.3875755909840572, "grad_norm": 0.9482265114784241, "learning_rate": 2.1017062708872612e-05, "loss": 0.0292, "step": 5640 }, { "epoch": 0.3876443100604728, "grad_norm": 1.2768324613571167, "learning_rate": 2.1014004371830794e-05, "loss": 0.0308, "step": 5641 }, { "epoch": 0.3877130291368884, "grad_norm": 0.7234402298927307, "learning_rate": 2.1010945736867563e-05, "loss": 0.0322, "step": 5642 }, { "epoch": 0.387781748213304, "grad_norm": 1.5287593603134155, "learning_rate": 2.1007886804134433e-05, "loss": 0.0305, "step": 5643 }, { "epoch": 0.3878504672897196, "grad_norm": 0.9421746730804443, "learning_rate": 2.1004827573782947e-05, "loss": 0.0225, "step": 5644 }, { "epoch": 0.38791918636613526, "grad_norm": 1.3778700828552246, "learning_rate": 2.100176804596465e-05, "loss": 0.0281, "step": 5645 }, { "epoch": 0.38798790544255085, "grad_norm": 1.2307771444320679, "learning_rate": 2.0998708220831096e-05, "loss": 0.0261, "step": 5646 }, { "epoch": 0.38805662451896644, "grad_norm": 0.8766735792160034, "learning_rate": 2.0995648098533876e-05, "loss": 0.0355, "step": 5647 }, { "epoch": 0.3881253435953821, "grad_norm": 0.6087660789489746, "learning_rate": 2.0992587679224574e-05, "loss": 0.0304, "step": 5648 }, { "epoch": 0.3881940626717977, "grad_norm": 0.7193711996078491, "learning_rate": 2.0989526963054798e-05, "loss": 0.0266, "step": 5649 }, { "epoch": 0.3882627817482133, "grad_norm": 0.6476078629493713, "learning_rate": 2.098646595017617e-05, "loss": 0.0244, "step": 5650 }, { "epoch": 0.3883315008246289, "grad_norm": 1.2006696462631226, "learning_rate": 2.098340464074033e-05, "loss": 0.028, "step": 5651 }, { "epoch": 0.3884002199010445, "grad_norm": 0.7813234925270081, "learning_rate": 2.098034303489892e-05, "loss": 0.0212, "step": 5652 }, { "epoch": 0.38846893897746015, "grad_norm": 0.7497948408126831, "learning_rate": 2.097728113280361e-05, "loss": 0.0188, "step": 5653 }, { "epoch": 0.38853765805387575, "grad_norm": 0.7671800851821899, "learning_rate": 2.097421893460609e-05, "loss": 0.0232, "step": 5654 }, { "epoch": 0.3886063771302914, "grad_norm": 0.4715821146965027, "learning_rate": 2.0971156440458044e-05, "loss": 0.0243, "step": 5655 }, { "epoch": 0.388675096206707, "grad_norm": 0.9394846558570862, "learning_rate": 2.096809365051118e-05, "loss": 0.0348, "step": 5656 }, { "epoch": 0.3887438152831226, "grad_norm": 1.0703353881835938, "learning_rate": 2.096503056491723e-05, "loss": 0.0253, "step": 5657 }, { "epoch": 0.3888125343595382, "grad_norm": 0.9488407373428345, "learning_rate": 2.0961967183827925e-05, "loss": 0.0283, "step": 5658 }, { "epoch": 0.3888812534359538, "grad_norm": 1.0495612621307373, "learning_rate": 2.0958903507395028e-05, "loss": 0.0317, "step": 5659 }, { "epoch": 0.38894997251236946, "grad_norm": 0.5579923391342163, "learning_rate": 2.09558395357703e-05, "loss": 0.0252, "step": 5660 }, { "epoch": 0.38901869158878505, "grad_norm": 0.3127039074897766, "learning_rate": 2.0952775269105524e-05, "loss": 0.015, "step": 5661 }, { "epoch": 0.38908741066520064, "grad_norm": 0.8876565098762512, "learning_rate": 2.09497107075525e-05, "loss": 0.0169, "step": 5662 }, { "epoch": 0.3891561297416163, "grad_norm": 0.36683177947998047, "learning_rate": 2.0946645851263043e-05, "loss": 0.0384, "step": 5663 }, { "epoch": 0.3892248488180319, "grad_norm": 1.2429907321929932, "learning_rate": 2.0943580700388975e-05, "loss": 0.0267, "step": 5664 }, { "epoch": 0.3892935678944475, "grad_norm": 1.0842088460922241, "learning_rate": 2.0940515255082137e-05, "loss": 0.038, "step": 5665 }, { "epoch": 0.3893622869708631, "grad_norm": 1.0089380741119385, "learning_rate": 2.0937449515494387e-05, "loss": 0.0311, "step": 5666 }, { "epoch": 0.3894310060472787, "grad_norm": 0.8767892718315125, "learning_rate": 2.0934383481777593e-05, "loss": 0.036, "step": 5667 }, { "epoch": 0.38949972512369435, "grad_norm": 1.523781180381775, "learning_rate": 2.0931317154083645e-05, "loss": 0.0407, "step": 5668 }, { "epoch": 0.38956844420010994, "grad_norm": 0.6715369820594788, "learning_rate": 2.092825053256444e-05, "loss": 0.0538, "step": 5669 }, { "epoch": 0.3896371632765256, "grad_norm": 0.5348339080810547, "learning_rate": 2.0925183617371887e-05, "loss": 0.0234, "step": 5670 }, { "epoch": 0.3897058823529412, "grad_norm": 1.6460424661636353, "learning_rate": 2.0922116408657925e-05, "loss": 0.0364, "step": 5671 }, { "epoch": 0.38977460142935677, "grad_norm": 1.6402820348739624, "learning_rate": 2.0919048906574487e-05, "loss": 0.0361, "step": 5672 }, { "epoch": 0.3898433205057724, "grad_norm": 0.54383385181427, "learning_rate": 2.0915981111273538e-05, "loss": 0.0427, "step": 5673 }, { "epoch": 0.389912039582188, "grad_norm": 2.074882745742798, "learning_rate": 2.091291302290705e-05, "loss": 0.0333, "step": 5674 }, { "epoch": 0.38998075865860365, "grad_norm": 0.8610568642616272, "learning_rate": 2.0909844641627e-05, "loss": 0.0214, "step": 5675 }, { "epoch": 0.39004947773501925, "grad_norm": 0.4435202479362488, "learning_rate": 2.0906775967585403e-05, "loss": 0.0343, "step": 5676 }, { "epoch": 0.39011819681143484, "grad_norm": 1.1880555152893066, "learning_rate": 2.0903707000934267e-05, "loss": 0.027, "step": 5677 }, { "epoch": 0.3901869158878505, "grad_norm": 0.7528991103172302, "learning_rate": 2.0900637741825623e-05, "loss": 0.0237, "step": 5678 }, { "epoch": 0.3902556349642661, "grad_norm": 0.4160280227661133, "learning_rate": 2.0897568190411516e-05, "loss": 0.0243, "step": 5679 }, { "epoch": 0.3903243540406817, "grad_norm": 0.8985089063644409, "learning_rate": 2.0894498346844006e-05, "loss": 0.0285, "step": 5680 }, { "epoch": 0.3903930731170973, "grad_norm": 1.6423546075820923, "learning_rate": 2.0891428211275165e-05, "loss": 0.0429, "step": 5681 }, { "epoch": 0.3904617921935129, "grad_norm": 0.8303678631782532, "learning_rate": 2.0888357783857083e-05, "loss": 0.0187, "step": 5682 }, { "epoch": 0.39053051126992855, "grad_norm": 0.8457767963409424, "learning_rate": 2.088528706474186e-05, "loss": 0.0341, "step": 5683 }, { "epoch": 0.39059923034634414, "grad_norm": 0.6897199153900146, "learning_rate": 2.0882216054081626e-05, "loss": 0.0249, "step": 5684 }, { "epoch": 0.39066794942275973, "grad_norm": 0.8977672457695007, "learning_rate": 2.087914475202849e-05, "loss": 0.0222, "step": 5685 }, { "epoch": 0.3907366684991754, "grad_norm": 1.010912537574768, "learning_rate": 2.0876073158734616e-05, "loss": 0.0205, "step": 5686 }, { "epoch": 0.39080538757559097, "grad_norm": 0.6361986398696899, "learning_rate": 2.0873001274352157e-05, "loss": 0.0258, "step": 5687 }, { "epoch": 0.3908741066520066, "grad_norm": 0.7958696484565735, "learning_rate": 2.086992909903329e-05, "loss": 0.0245, "step": 5688 }, { "epoch": 0.3909428257284222, "grad_norm": 0.9541946053504944, "learning_rate": 2.0866856632930203e-05, "loss": 0.0246, "step": 5689 }, { "epoch": 0.3910115448048378, "grad_norm": 0.7317186594009399, "learning_rate": 2.0863783876195098e-05, "loss": 0.028, "step": 5690 }, { "epoch": 0.39108026388125344, "grad_norm": 1.0333523750305176, "learning_rate": 2.0860710828980197e-05, "loss": 0.0414, "step": 5691 }, { "epoch": 0.39114898295766903, "grad_norm": 0.6721642017364502, "learning_rate": 2.0857637491437734e-05, "loss": 0.0309, "step": 5692 }, { "epoch": 0.3912177020340847, "grad_norm": 0.5755858421325684, "learning_rate": 2.085456386371995e-05, "loss": 0.0243, "step": 5693 }, { "epoch": 0.39128642111050027, "grad_norm": 0.8722837567329407, "learning_rate": 2.0851489945979106e-05, "loss": 0.0299, "step": 5694 }, { "epoch": 0.39135514018691586, "grad_norm": 0.9266139268875122, "learning_rate": 2.084841573836748e-05, "loss": 0.0246, "step": 5695 }, { "epoch": 0.3914238592633315, "grad_norm": 0.6410175561904907, "learning_rate": 2.084534124103737e-05, "loss": 0.0193, "step": 5696 }, { "epoch": 0.3914925783397471, "grad_norm": 1.4265098571777344, "learning_rate": 2.084226645414106e-05, "loss": 0.0219, "step": 5697 }, { "epoch": 0.39156129741616275, "grad_norm": 0.3703326880931854, "learning_rate": 2.0839191377830892e-05, "loss": 0.0244, "step": 5698 }, { "epoch": 0.39163001649257834, "grad_norm": 0.8349830508232117, "learning_rate": 2.0836116012259185e-05, "loss": 0.0242, "step": 5699 }, { "epoch": 0.3916987355689939, "grad_norm": 0.4507155418395996, "learning_rate": 2.0833040357578284e-05, "loss": 0.0234, "step": 5700 }, { "epoch": 0.3917674546454096, "grad_norm": 0.6563207507133484, "learning_rate": 2.0829964413940563e-05, "loss": 0.0307, "step": 5701 }, { "epoch": 0.39183617372182517, "grad_norm": 0.9350759983062744, "learning_rate": 2.0826888181498387e-05, "loss": 0.0267, "step": 5702 }, { "epoch": 0.3919048927982408, "grad_norm": 0.5734410881996155, "learning_rate": 2.0823811660404154e-05, "loss": 0.0182, "step": 5703 }, { "epoch": 0.3919736118746564, "grad_norm": 0.913791835308075, "learning_rate": 2.082073485081026e-05, "loss": 0.0237, "step": 5704 }, { "epoch": 0.392042330951072, "grad_norm": 0.5097733736038208, "learning_rate": 2.081765775286913e-05, "loss": 0.0231, "step": 5705 }, { "epoch": 0.39211105002748764, "grad_norm": 0.7074660062789917, "learning_rate": 2.0814580366733195e-05, "loss": 0.0216, "step": 5706 }, { "epoch": 0.39217976910390323, "grad_norm": 1.0917130708694458, "learning_rate": 2.08115026925549e-05, "loss": 0.0212, "step": 5707 }, { "epoch": 0.3922484881803189, "grad_norm": 0.9038196206092834, "learning_rate": 2.0808424730486714e-05, "loss": 0.0273, "step": 5708 }, { "epoch": 0.39231720725673447, "grad_norm": 0.8278385400772095, "learning_rate": 2.080534648068111e-05, "loss": 0.0285, "step": 5709 }, { "epoch": 0.39238592633315006, "grad_norm": 0.8016522526741028, "learning_rate": 2.0802267943290576e-05, "loss": 0.0314, "step": 5710 }, { "epoch": 0.3924546454095657, "grad_norm": 1.3695553541183472, "learning_rate": 2.0799189118467613e-05, "loss": 0.0246, "step": 5711 }, { "epoch": 0.3925233644859813, "grad_norm": 0.43895459175109863, "learning_rate": 2.0796110006364748e-05, "loss": 0.0151, "step": 5712 }, { "epoch": 0.39259208356239694, "grad_norm": 0.6296882033348083, "learning_rate": 2.0793030607134505e-05, "loss": 0.0197, "step": 5713 }, { "epoch": 0.39266080263881253, "grad_norm": 1.0295095443725586, "learning_rate": 2.0789950920929442e-05, "loss": 0.0352, "step": 5714 }, { "epoch": 0.3927295217152281, "grad_norm": 0.8663643002510071, "learning_rate": 2.0786870947902107e-05, "loss": 0.0217, "step": 5715 }, { "epoch": 0.39279824079164377, "grad_norm": 0.8898918032646179, "learning_rate": 2.0783790688205086e-05, "loss": 0.0318, "step": 5716 }, { "epoch": 0.39286695986805936, "grad_norm": 0.5022631883621216, "learning_rate": 2.0780710141990965e-05, "loss": 0.019, "step": 5717 }, { "epoch": 0.392935678944475, "grad_norm": 0.5917985439300537, "learning_rate": 2.077762930941235e-05, "loss": 0.0288, "step": 5718 }, { "epoch": 0.3930043980208906, "grad_norm": 0.7274780869483948, "learning_rate": 2.0774548190621855e-05, "loss": 0.0214, "step": 5719 }, { "epoch": 0.3930731170973062, "grad_norm": 0.46557387709617615, "learning_rate": 2.0771466785772112e-05, "loss": 0.032, "step": 5720 }, { "epoch": 0.39314183617372184, "grad_norm": 0.5478065609931946, "learning_rate": 2.0768385095015773e-05, "loss": 0.0276, "step": 5721 }, { "epoch": 0.39321055525013743, "grad_norm": 0.6760910749435425, "learning_rate": 2.0765303118505494e-05, "loss": 0.0249, "step": 5722 }, { "epoch": 0.3932792743265531, "grad_norm": 0.6751284003257751, "learning_rate": 2.0762220856393956e-05, "loss": 0.0281, "step": 5723 }, { "epoch": 0.39334799340296867, "grad_norm": 1.0110011100769043, "learning_rate": 2.075913830883384e-05, "loss": 0.034, "step": 5724 }, { "epoch": 0.39341671247938426, "grad_norm": 0.7511439323425293, "learning_rate": 2.0756055475977852e-05, "loss": 0.0206, "step": 5725 }, { "epoch": 0.3934854315557999, "grad_norm": 0.6135537624359131, "learning_rate": 2.075297235797871e-05, "loss": 0.0197, "step": 5726 }, { "epoch": 0.3935541506322155, "grad_norm": 1.3524692058563232, "learning_rate": 2.0749888954989146e-05, "loss": 0.0235, "step": 5727 }, { "epoch": 0.39362286970863114, "grad_norm": 1.090073823928833, "learning_rate": 2.0746805267161906e-05, "loss": 0.0235, "step": 5728 }, { "epoch": 0.39369158878504673, "grad_norm": 0.6540341377258301, "learning_rate": 2.074372129464975e-05, "loss": 0.0307, "step": 5729 }, { "epoch": 0.3937603078614623, "grad_norm": 1.118303656578064, "learning_rate": 2.0740637037605444e-05, "loss": 0.0207, "step": 5730 }, { "epoch": 0.39382902693787797, "grad_norm": 0.5248265862464905, "learning_rate": 2.073755249618178e-05, "loss": 0.0285, "step": 5731 }, { "epoch": 0.39389774601429356, "grad_norm": 0.9089072346687317, "learning_rate": 2.073446767053157e-05, "loss": 0.0344, "step": 5732 }, { "epoch": 0.3939664650907092, "grad_norm": 1.2306021451950073, "learning_rate": 2.0731382560807616e-05, "loss": 0.026, "step": 5733 }, { "epoch": 0.3940351841671248, "grad_norm": 0.8965484499931335, "learning_rate": 2.072829716716275e-05, "loss": 0.0216, "step": 5734 }, { "epoch": 0.3941039032435404, "grad_norm": 0.954935610294342, "learning_rate": 2.072521148974982e-05, "loss": 0.0169, "step": 5735 }, { "epoch": 0.39417262231995603, "grad_norm": 0.9496985077857971, "learning_rate": 2.0722125528721687e-05, "loss": 0.0244, "step": 5736 }, { "epoch": 0.3942413413963716, "grad_norm": 0.6559740304946899, "learning_rate": 2.071903928423122e-05, "loss": 0.0428, "step": 5737 }, { "epoch": 0.39431006047278727, "grad_norm": 0.7300053834915161, "learning_rate": 2.07159527564313e-05, "loss": 0.0257, "step": 5738 }, { "epoch": 0.39437877954920286, "grad_norm": 0.6475920677185059, "learning_rate": 2.0712865945474836e-05, "loss": 0.0126, "step": 5739 }, { "epoch": 0.39444749862561845, "grad_norm": 0.4833623766899109, "learning_rate": 2.0709778851514733e-05, "loss": 0.0173, "step": 5740 }, { "epoch": 0.3945162177020341, "grad_norm": 0.4598204493522644, "learning_rate": 2.070669147470393e-05, "loss": 0.0243, "step": 5741 }, { "epoch": 0.3945849367784497, "grad_norm": 0.8968642354011536, "learning_rate": 2.070360381519536e-05, "loss": 0.0384, "step": 5742 }, { "epoch": 0.39465365585486534, "grad_norm": 0.7609917521476746, "learning_rate": 2.0700515873141986e-05, "loss": 0.0493, "step": 5743 }, { "epoch": 0.39472237493128093, "grad_norm": 0.5385901927947998, "learning_rate": 2.0697427648696778e-05, "loss": 0.0215, "step": 5744 }, { "epoch": 0.3947910940076965, "grad_norm": 0.8076997995376587, "learning_rate": 2.0694339142012713e-05, "loss": 0.0377, "step": 5745 }, { "epoch": 0.39485981308411217, "grad_norm": 0.6775421500205994, "learning_rate": 2.0691250353242798e-05, "loss": 0.0267, "step": 5746 }, { "epoch": 0.39492853216052776, "grad_norm": 0.7509361505508423, "learning_rate": 2.0688161282540036e-05, "loss": 0.0265, "step": 5747 }, { "epoch": 0.3949972512369434, "grad_norm": 0.6834538578987122, "learning_rate": 2.068507193005746e-05, "loss": 0.0236, "step": 5748 }, { "epoch": 0.395065970313359, "grad_norm": 0.8707718253135681, "learning_rate": 2.068198229594811e-05, "loss": 0.0286, "step": 5749 }, { "epoch": 0.3951346893897746, "grad_norm": 0.7247880101203918, "learning_rate": 2.0678892380365042e-05, "loss": 0.0297, "step": 5750 }, { "epoch": 0.39520340846619023, "grad_norm": 0.646868884563446, "learning_rate": 2.0675802183461314e-05, "loss": 0.0327, "step": 5751 }, { "epoch": 0.3952721275426058, "grad_norm": 1.1697134971618652, "learning_rate": 2.0672711705390017e-05, "loss": 0.0242, "step": 5752 }, { "epoch": 0.3953408466190214, "grad_norm": 0.6555367112159729, "learning_rate": 2.0669620946304247e-05, "loss": 0.028, "step": 5753 }, { "epoch": 0.39540956569543706, "grad_norm": 0.7139495015144348, "learning_rate": 2.066652990635711e-05, "loss": 0.0361, "step": 5754 }, { "epoch": 0.39547828477185265, "grad_norm": 0.5752590894699097, "learning_rate": 2.0663438585701728e-05, "loss": 0.0242, "step": 5755 }, { "epoch": 0.3955470038482683, "grad_norm": 1.3598778247833252, "learning_rate": 2.0660346984491248e-05, "loss": 0.0337, "step": 5756 }, { "epoch": 0.3956157229246839, "grad_norm": 1.5067421197891235, "learning_rate": 2.0657255102878816e-05, "loss": 0.0282, "step": 5757 }, { "epoch": 0.3956844420010995, "grad_norm": 0.8513872027397156, "learning_rate": 2.065416294101759e-05, "loss": 0.0319, "step": 5758 }, { "epoch": 0.3957531610775151, "grad_norm": 1.2639063596725464, "learning_rate": 2.065107049906076e-05, "loss": 0.0291, "step": 5759 }, { "epoch": 0.3958218801539307, "grad_norm": 0.6443334817886353, "learning_rate": 2.0647977777161518e-05, "loss": 0.0332, "step": 5760 }, { "epoch": 0.39589059923034636, "grad_norm": 1.045330286026001, "learning_rate": 2.0644884775473065e-05, "loss": 0.0232, "step": 5761 }, { "epoch": 0.39595931830676195, "grad_norm": 0.7121452689170837, "learning_rate": 2.0641791494148628e-05, "loss": 0.028, "step": 5762 }, { "epoch": 0.39602803738317754, "grad_norm": 0.5640497803688049, "learning_rate": 2.063869793334144e-05, "loss": 0.0213, "step": 5763 }, { "epoch": 0.3960967564595932, "grad_norm": 0.6632938385009766, "learning_rate": 2.0635604093204748e-05, "loss": 0.0301, "step": 5764 }, { "epoch": 0.3961654755360088, "grad_norm": 0.6437513828277588, "learning_rate": 2.0632509973891812e-05, "loss": 0.0257, "step": 5765 }, { "epoch": 0.39623419461242443, "grad_norm": 1.3905649185180664, "learning_rate": 2.0629415575555917e-05, "loss": 0.0169, "step": 5766 }, { "epoch": 0.39630291368884, "grad_norm": 1.198571801185608, "learning_rate": 2.0626320898350348e-05, "loss": 0.0237, "step": 5767 }, { "epoch": 0.3963716327652556, "grad_norm": 0.7058683633804321, "learning_rate": 2.0623225942428408e-05, "loss": 0.0221, "step": 5768 }, { "epoch": 0.39644035184167126, "grad_norm": 0.9913640022277832, "learning_rate": 2.0620130707943413e-05, "loss": 0.0214, "step": 5769 }, { "epoch": 0.39650907091808685, "grad_norm": 0.7137094736099243, "learning_rate": 2.0617035195048698e-05, "loss": 0.0219, "step": 5770 }, { "epoch": 0.3965777899945025, "grad_norm": 1.5065959692001343, "learning_rate": 2.0613939403897608e-05, "loss": 0.024, "step": 5771 }, { "epoch": 0.3966465090709181, "grad_norm": 0.34936249256134033, "learning_rate": 2.0610843334643504e-05, "loss": 0.0254, "step": 5772 }, { "epoch": 0.3967152281473337, "grad_norm": 0.7041373252868652, "learning_rate": 2.0607746987439753e-05, "loss": 0.0152, "step": 5773 }, { "epoch": 0.3967839472237493, "grad_norm": 0.4177579879760742, "learning_rate": 2.060465036243975e-05, "loss": 0.0221, "step": 5774 }, { "epoch": 0.3968526663001649, "grad_norm": 0.710776686668396, "learning_rate": 2.0601553459796884e-05, "loss": 0.0253, "step": 5775 }, { "epoch": 0.39692138537658056, "grad_norm": 0.3317088186740875, "learning_rate": 2.0598456279664583e-05, "loss": 0.0212, "step": 5776 }, { "epoch": 0.39699010445299615, "grad_norm": 0.7241785526275635, "learning_rate": 2.0595358822196262e-05, "loss": 0.0188, "step": 5777 }, { "epoch": 0.39705882352941174, "grad_norm": 1.2258261442184448, "learning_rate": 2.059226108754537e-05, "loss": 0.0224, "step": 5778 }, { "epoch": 0.3971275426058274, "grad_norm": 0.5152993202209473, "learning_rate": 2.058916307586536e-05, "loss": 0.023, "step": 5779 }, { "epoch": 0.397196261682243, "grad_norm": 1.1707383394241333, "learning_rate": 2.058606478730971e-05, "loss": 0.0322, "step": 5780 }, { "epoch": 0.3972649807586586, "grad_norm": 0.3385799825191498, "learning_rate": 2.0582966222031885e-05, "loss": 0.0243, "step": 5781 }, { "epoch": 0.3973336998350742, "grad_norm": 0.8102717995643616, "learning_rate": 2.0579867380185394e-05, "loss": 0.0353, "step": 5782 }, { "epoch": 0.3974024189114898, "grad_norm": 0.6502715349197388, "learning_rate": 2.0576768261923748e-05, "loss": 0.0179, "step": 5783 }, { "epoch": 0.39747113798790545, "grad_norm": 1.3721944093704224, "learning_rate": 2.0573668867400466e-05, "loss": 0.0287, "step": 5784 }, { "epoch": 0.39753985706432104, "grad_norm": 1.2673622369766235, "learning_rate": 2.057056919676909e-05, "loss": 0.0234, "step": 5785 }, { "epoch": 0.3976085761407367, "grad_norm": 1.8248502016067505, "learning_rate": 2.0567469250183162e-05, "loss": 0.0331, "step": 5786 }, { "epoch": 0.3976772952171523, "grad_norm": 0.6231353878974915, "learning_rate": 2.056436902779626e-05, "loss": 0.0292, "step": 5787 }, { "epoch": 0.3977460142935679, "grad_norm": 0.4572446048259735, "learning_rate": 2.0561268529761952e-05, "loss": 0.0192, "step": 5788 }, { "epoch": 0.3978147333699835, "grad_norm": 0.9801149368286133, "learning_rate": 2.0558167756233836e-05, "loss": 0.0269, "step": 5789 }, { "epoch": 0.3978834524463991, "grad_norm": 0.6612380146980286, "learning_rate": 2.0555066707365518e-05, "loss": 0.02, "step": 5790 }, { "epoch": 0.39795217152281476, "grad_norm": 1.1331303119659424, "learning_rate": 2.0551965383310615e-05, "loss": 0.0244, "step": 5791 }, { "epoch": 0.39802089059923035, "grad_norm": 1.0888255834579468, "learning_rate": 2.054886378422276e-05, "loss": 0.0454, "step": 5792 }, { "epoch": 0.39808960967564594, "grad_norm": 0.896975040435791, "learning_rate": 2.0545761910255604e-05, "loss": 0.0202, "step": 5793 }, { "epoch": 0.3981583287520616, "grad_norm": 0.30292147397994995, "learning_rate": 2.05426597615628e-05, "loss": 0.0432, "step": 5794 }, { "epoch": 0.3982270478284772, "grad_norm": 0.763591468334198, "learning_rate": 2.053955733829803e-05, "loss": 0.0324, "step": 5795 }, { "epoch": 0.3982957669048928, "grad_norm": 0.5725674629211426, "learning_rate": 2.0536454640614975e-05, "loss": 0.019, "step": 5796 }, { "epoch": 0.3983644859813084, "grad_norm": 0.7024863362312317, "learning_rate": 2.053335166866734e-05, "loss": 0.0253, "step": 5797 }, { "epoch": 0.398433205057724, "grad_norm": 0.7589266896247864, "learning_rate": 2.053024842260884e-05, "loss": 0.0309, "step": 5798 }, { "epoch": 0.39850192413413965, "grad_norm": 0.7923285365104675, "learning_rate": 2.052714490259321e-05, "loss": 0.0186, "step": 5799 }, { "epoch": 0.39857064321055524, "grad_norm": 0.602238655090332, "learning_rate": 2.0524041108774176e-05, "loss": 0.0349, "step": 5800 }, { "epoch": 0.3986393622869709, "grad_norm": 0.7745864987373352, "learning_rate": 2.0520937041305504e-05, "loss": 0.0162, "step": 5801 }, { "epoch": 0.3987080813633865, "grad_norm": 1.2313041687011719, "learning_rate": 2.0517832700340965e-05, "loss": 0.0224, "step": 5802 }, { "epoch": 0.39877680043980207, "grad_norm": 1.0300389528274536, "learning_rate": 2.0514728086034334e-05, "loss": 0.0203, "step": 5803 }, { "epoch": 0.3988455195162177, "grad_norm": 0.739368200302124, "learning_rate": 2.0511623198539415e-05, "loss": 0.0211, "step": 5804 }, { "epoch": 0.3989142385926333, "grad_norm": 0.8264457583427429, "learning_rate": 2.0508518038010015e-05, "loss": 0.0233, "step": 5805 }, { "epoch": 0.39898295766904895, "grad_norm": 0.9042609333992004, "learning_rate": 2.0505412604599952e-05, "loss": 0.018, "step": 5806 }, { "epoch": 0.39905167674546455, "grad_norm": 0.4567056894302368, "learning_rate": 2.050230689846307e-05, "loss": 0.0245, "step": 5807 }, { "epoch": 0.39912039582188014, "grad_norm": 1.1062026023864746, "learning_rate": 2.0499200919753218e-05, "loss": 0.031, "step": 5808 }, { "epoch": 0.3991891148982958, "grad_norm": 0.45550939440727234, "learning_rate": 2.0496094668624253e-05, "loss": 0.0158, "step": 5809 }, { "epoch": 0.3992578339747114, "grad_norm": 0.6067140698432922, "learning_rate": 2.0492988145230062e-05, "loss": 0.0259, "step": 5810 }, { "epoch": 0.399326553051127, "grad_norm": 0.7266212105751038, "learning_rate": 2.0489881349724532e-05, "loss": 0.0228, "step": 5811 }, { "epoch": 0.3993952721275426, "grad_norm": 0.9274150133132935, "learning_rate": 2.0486774282261564e-05, "loss": 0.0329, "step": 5812 }, { "epoch": 0.3994639912039582, "grad_norm": 0.874180257320404, "learning_rate": 2.048366694299508e-05, "loss": 0.0292, "step": 5813 }, { "epoch": 0.39953271028037385, "grad_norm": 0.7397989630699158, "learning_rate": 2.048055933207901e-05, "loss": 0.0213, "step": 5814 }, { "epoch": 0.39960142935678944, "grad_norm": 0.6581215858459473, "learning_rate": 2.0477451449667292e-05, "loss": 0.0255, "step": 5815 }, { "epoch": 0.39967014843320503, "grad_norm": 1.1004738807678223, "learning_rate": 2.04743432959139e-05, "loss": 0.0228, "step": 5816 }, { "epoch": 0.3997388675096207, "grad_norm": 1.7874854803085327, "learning_rate": 2.0471234870972793e-05, "loss": 0.0188, "step": 5817 }, { "epoch": 0.39980758658603627, "grad_norm": 0.43493691086769104, "learning_rate": 2.0468126174997955e-05, "loss": 0.02, "step": 5818 }, { "epoch": 0.3998763056624519, "grad_norm": 0.8275386095046997, "learning_rate": 2.046501720814339e-05, "loss": 0.0234, "step": 5819 }, { "epoch": 0.3999450247388675, "grad_norm": 0.5639863610267639, "learning_rate": 2.0461907970563108e-05, "loss": 0.0201, "step": 5820 }, { "epoch": 0.4000137438152831, "grad_norm": 1.3725595474243164, "learning_rate": 2.0458798462411135e-05, "loss": 0.0169, "step": 5821 }, { "epoch": 0.40008246289169874, "grad_norm": 0.7816755175590515, "learning_rate": 2.045568868384151e-05, "loss": 0.0233, "step": 5822 }, { "epoch": 0.40015118196811433, "grad_norm": 1.5555367469787598, "learning_rate": 2.0452578635008286e-05, "loss": 0.021, "step": 5823 }, { "epoch": 0.40021990104453, "grad_norm": 0.8266755938529968, "learning_rate": 2.044946831606552e-05, "loss": 0.0255, "step": 5824 }, { "epoch": 0.40028862012094557, "grad_norm": 0.503833532333374, "learning_rate": 2.0446357727167303e-05, "loss": 0.0253, "step": 5825 }, { "epoch": 0.40035733919736116, "grad_norm": 1.7212327718734741, "learning_rate": 2.044324686846772e-05, "loss": 0.0412, "step": 5826 }, { "epoch": 0.4004260582737768, "grad_norm": 1.1431450843811035, "learning_rate": 2.0440135740120876e-05, "loss": 0.0281, "step": 5827 }, { "epoch": 0.4004947773501924, "grad_norm": 0.801125705242157, "learning_rate": 2.043702434228089e-05, "loss": 0.0248, "step": 5828 }, { "epoch": 0.40056349642660805, "grad_norm": 1.000648021697998, "learning_rate": 2.0433912675101902e-05, "loss": 0.0391, "step": 5829 }, { "epoch": 0.40063221550302364, "grad_norm": 0.4917972981929779, "learning_rate": 2.0430800738738044e-05, "loss": 0.0243, "step": 5830 }, { "epoch": 0.4007009345794392, "grad_norm": 1.2352205514907837, "learning_rate": 2.042768853334349e-05, "loss": 0.0405, "step": 5831 }, { "epoch": 0.4007696536558549, "grad_norm": 0.915459156036377, "learning_rate": 2.04245760590724e-05, "loss": 0.03, "step": 5832 }, { "epoch": 0.40083837273227046, "grad_norm": 0.7020566463470459, "learning_rate": 2.042146331607897e-05, "loss": 0.0195, "step": 5833 }, { "epoch": 0.4009070918086861, "grad_norm": 0.9555028080940247, "learning_rate": 2.041835030451739e-05, "loss": 0.0251, "step": 5834 }, { "epoch": 0.4009758108851017, "grad_norm": 1.4153279066085815, "learning_rate": 2.0415237024541877e-05, "loss": 0.0324, "step": 5835 }, { "epoch": 0.4010445299615173, "grad_norm": 1.9468740224838257, "learning_rate": 2.041212347630665e-05, "loss": 0.03, "step": 5836 }, { "epoch": 0.40111324903793294, "grad_norm": 0.9197319746017456, "learning_rate": 2.0409009659965957e-05, "loss": 0.0208, "step": 5837 }, { "epoch": 0.40118196811434853, "grad_norm": 0.3251846134662628, "learning_rate": 2.0405895575674047e-05, "loss": 0.0291, "step": 5838 }, { "epoch": 0.4012506871907642, "grad_norm": 1.1327980756759644, "learning_rate": 2.0402781223585184e-05, "loss": 0.0272, "step": 5839 }, { "epoch": 0.40131940626717977, "grad_norm": 0.6359328031539917, "learning_rate": 2.0399666603853648e-05, "loss": 0.0258, "step": 5840 }, { "epoch": 0.40138812534359536, "grad_norm": 0.49216508865356445, "learning_rate": 2.039655171663373e-05, "loss": 0.0239, "step": 5841 }, { "epoch": 0.401456844420011, "grad_norm": 0.6740776300430298, "learning_rate": 2.0393436562079726e-05, "loss": 0.0203, "step": 5842 }, { "epoch": 0.4015255634964266, "grad_norm": 0.7449263334274292, "learning_rate": 2.039032114034597e-05, "loss": 0.03, "step": 5843 }, { "epoch": 0.40159428257284224, "grad_norm": 1.3620812892913818, "learning_rate": 2.038720545158679e-05, "loss": 0.0287, "step": 5844 }, { "epoch": 0.40166300164925783, "grad_norm": 0.30142590403556824, "learning_rate": 2.038408949595652e-05, "loss": 0.021, "step": 5845 }, { "epoch": 0.4017317207256734, "grad_norm": 1.2425180673599243, "learning_rate": 2.038097327360953e-05, "loss": 0.03, "step": 5846 }, { "epoch": 0.40180043980208907, "grad_norm": 0.6493044495582581, "learning_rate": 2.037785678470019e-05, "loss": 0.038, "step": 5847 }, { "epoch": 0.40186915887850466, "grad_norm": 0.4388674795627594, "learning_rate": 2.037474002938287e-05, "loss": 0.0234, "step": 5848 }, { "epoch": 0.4019378779549203, "grad_norm": 1.042931318283081, "learning_rate": 2.0371623007811988e-05, "loss": 0.0308, "step": 5849 }, { "epoch": 0.4020065970313359, "grad_norm": 0.7943124771118164, "learning_rate": 2.0368505720141947e-05, "loss": 0.0257, "step": 5850 }, { "epoch": 0.4020753161077515, "grad_norm": 1.9186862707138062, "learning_rate": 2.0365388166527162e-05, "loss": 0.0316, "step": 5851 }, { "epoch": 0.40214403518416714, "grad_norm": 0.4064747989177704, "learning_rate": 2.0362270347122085e-05, "loss": 0.0311, "step": 5852 }, { "epoch": 0.4022127542605827, "grad_norm": 0.9020446538925171, "learning_rate": 2.035915226208116e-05, "loss": 0.0299, "step": 5853 }, { "epoch": 0.4022814733369984, "grad_norm": 0.5195156931877136, "learning_rate": 2.0356033911558843e-05, "loss": 0.0217, "step": 5854 }, { "epoch": 0.40235019241341397, "grad_norm": 0.8486201763153076, "learning_rate": 2.035291529570962e-05, "loss": 0.0329, "step": 5855 }, { "epoch": 0.40241891148982956, "grad_norm": 0.5297520756721497, "learning_rate": 2.0349796414687987e-05, "loss": 0.0188, "step": 5856 }, { "epoch": 0.4024876305662452, "grad_norm": 0.5563024282455444, "learning_rate": 2.034667726864843e-05, "loss": 0.0243, "step": 5857 }, { "epoch": 0.4025563496426608, "grad_norm": 0.7190931439399719, "learning_rate": 2.0343557857745476e-05, "loss": 0.02, "step": 5858 }, { "epoch": 0.40262506871907644, "grad_norm": 1.2219210863113403, "learning_rate": 2.034043818213365e-05, "loss": 0.0385, "step": 5859 }, { "epoch": 0.40269378779549203, "grad_norm": 0.4407012462615967, "learning_rate": 2.0337318241967493e-05, "loss": 0.0233, "step": 5860 }, { "epoch": 0.4027625068719076, "grad_norm": 1.0627883672714233, "learning_rate": 2.033419803740157e-05, "loss": 0.0228, "step": 5861 }, { "epoch": 0.40283122594832327, "grad_norm": 1.2853480577468872, "learning_rate": 2.0331077568590442e-05, "loss": 0.0269, "step": 5862 }, { "epoch": 0.40289994502473886, "grad_norm": 0.7830513119697571, "learning_rate": 2.0327956835688685e-05, "loss": 0.0359, "step": 5863 }, { "epoch": 0.4029686641011545, "grad_norm": 0.7022579908370972, "learning_rate": 2.0324835838850905e-05, "loss": 0.0169, "step": 5864 }, { "epoch": 0.4030373831775701, "grad_norm": 0.867299497127533, "learning_rate": 2.0321714578231707e-05, "loss": 0.0229, "step": 5865 }, { "epoch": 0.4031061022539857, "grad_norm": 1.2178826332092285, "learning_rate": 2.0318593053985702e-05, "loss": 0.0296, "step": 5866 }, { "epoch": 0.40317482133040133, "grad_norm": 0.5302292704582214, "learning_rate": 2.0315471266267537e-05, "loss": 0.0185, "step": 5867 }, { "epoch": 0.4032435404068169, "grad_norm": 0.9582918286323547, "learning_rate": 2.0312349215231854e-05, "loss": 0.0215, "step": 5868 }, { "epoch": 0.40331225948323257, "grad_norm": 1.6157021522521973, "learning_rate": 2.030922690103331e-05, "loss": 0.0213, "step": 5869 }, { "epoch": 0.40338097855964816, "grad_norm": 0.5014550089836121, "learning_rate": 2.0306104323826587e-05, "loss": 0.0312, "step": 5870 }, { "epoch": 0.40344969763606375, "grad_norm": 1.0309901237487793, "learning_rate": 2.030298148376636e-05, "loss": 0.0297, "step": 5871 }, { "epoch": 0.4035184167124794, "grad_norm": 1.1988142728805542, "learning_rate": 2.029985838100733e-05, "loss": 0.027, "step": 5872 }, { "epoch": 0.403587135788895, "grad_norm": 0.42810380458831787, "learning_rate": 2.0296735015704215e-05, "loss": 0.0307, "step": 5873 }, { "epoch": 0.40365585486531064, "grad_norm": 0.9245587587356567, "learning_rate": 2.0293611388011738e-05, "loss": 0.0364, "step": 5874 }, { "epoch": 0.40372457394172623, "grad_norm": 0.7019236087799072, "learning_rate": 2.0290487498084636e-05, "loss": 0.0182, "step": 5875 }, { "epoch": 0.4037932930181418, "grad_norm": 0.7235502600669861, "learning_rate": 2.0287363346077656e-05, "loss": 0.02, "step": 5876 }, { "epoch": 0.40386201209455747, "grad_norm": 0.7867670059204102, "learning_rate": 2.028423893214557e-05, "loss": 0.0222, "step": 5877 }, { "epoch": 0.40393073117097306, "grad_norm": 0.5010905265808105, "learning_rate": 2.028111425644315e-05, "loss": 0.0251, "step": 5878 }, { "epoch": 0.4039994502473887, "grad_norm": 0.7147894501686096, "learning_rate": 2.0277989319125187e-05, "loss": 0.0205, "step": 5879 }, { "epoch": 0.4040681693238043, "grad_norm": 0.9648949503898621, "learning_rate": 2.027486412034649e-05, "loss": 0.0187, "step": 5880 }, { "epoch": 0.4041368884002199, "grad_norm": 0.6280426979064941, "learning_rate": 2.027173866026186e-05, "loss": 0.0202, "step": 5881 }, { "epoch": 0.40420560747663553, "grad_norm": 0.3543184995651245, "learning_rate": 2.026861293902614e-05, "loss": 0.0239, "step": 5882 }, { "epoch": 0.4042743265530511, "grad_norm": 0.7925532460212708, "learning_rate": 2.026548695679417e-05, "loss": 0.0304, "step": 5883 }, { "epoch": 0.4043430456294667, "grad_norm": 0.6826480031013489, "learning_rate": 2.0262360713720795e-05, "loss": 0.0201, "step": 5884 }, { "epoch": 0.40441176470588236, "grad_norm": 0.893330454826355, "learning_rate": 2.025923420996089e-05, "loss": 0.025, "step": 5885 }, { "epoch": 0.40448048378229795, "grad_norm": 1.0600008964538574, "learning_rate": 2.0256107445669336e-05, "loss": 0.0273, "step": 5886 }, { "epoch": 0.4045492028587136, "grad_norm": 0.6466431021690369, "learning_rate": 2.0252980421001022e-05, "loss": 0.0271, "step": 5887 }, { "epoch": 0.4046179219351292, "grad_norm": 0.5996161103248596, "learning_rate": 2.0249853136110865e-05, "loss": 0.0252, "step": 5888 }, { "epoch": 0.4046866410115448, "grad_norm": 0.7244311571121216, "learning_rate": 2.0246725591153773e-05, "loss": 0.0411, "step": 5889 }, { "epoch": 0.4047553600879604, "grad_norm": 0.8247092366218567, "learning_rate": 2.0243597786284678e-05, "loss": 0.0291, "step": 5890 }, { "epoch": 0.404824079164376, "grad_norm": 1.1990046501159668, "learning_rate": 2.0240469721658534e-05, "loss": 0.0353, "step": 5891 }, { "epoch": 0.40489279824079166, "grad_norm": 0.580717146396637, "learning_rate": 2.0237341397430297e-05, "loss": 0.0251, "step": 5892 }, { "epoch": 0.40496151731720725, "grad_norm": 1.4014698266983032, "learning_rate": 2.0234212813754927e-05, "loss": 0.0294, "step": 5893 }, { "epoch": 0.40503023639362284, "grad_norm": 0.5584294199943542, "learning_rate": 2.023108397078742e-05, "loss": 0.0151, "step": 5894 }, { "epoch": 0.4050989554700385, "grad_norm": 1.6363093852996826, "learning_rate": 2.0227954868682768e-05, "loss": 0.0383, "step": 5895 }, { "epoch": 0.4051676745464541, "grad_norm": 0.8982799053192139, "learning_rate": 2.0224825507595978e-05, "loss": 0.0332, "step": 5896 }, { "epoch": 0.40523639362286973, "grad_norm": 0.7582221627235413, "learning_rate": 2.0221695887682078e-05, "loss": 0.0333, "step": 5897 }, { "epoch": 0.4053051126992853, "grad_norm": 0.6446090340614319, "learning_rate": 2.0218566009096094e-05, "loss": 0.0229, "step": 5898 }, { "epoch": 0.4053738317757009, "grad_norm": 0.5109826326370239, "learning_rate": 2.021543587199308e-05, "loss": 0.0222, "step": 5899 }, { "epoch": 0.40544255085211656, "grad_norm": 0.8531697392463684, "learning_rate": 2.0212305476528095e-05, "loss": 0.0364, "step": 5900 }, { "epoch": 0.40551126992853215, "grad_norm": 1.0529158115386963, "learning_rate": 2.0209174822856216e-05, "loss": 0.0305, "step": 5901 }, { "epoch": 0.4055799890049478, "grad_norm": 0.7626609802246094, "learning_rate": 2.0206043911132525e-05, "loss": 0.0321, "step": 5902 }, { "epoch": 0.4056487080813634, "grad_norm": 0.539118230342865, "learning_rate": 2.020291274151212e-05, "loss": 0.0236, "step": 5903 }, { "epoch": 0.405717427157779, "grad_norm": 0.8011491894721985, "learning_rate": 2.0199781314150114e-05, "loss": 0.0275, "step": 5904 }, { "epoch": 0.4057861462341946, "grad_norm": 1.2320185899734497, "learning_rate": 2.019664962920163e-05, "loss": 0.0235, "step": 5905 }, { "epoch": 0.4058548653106102, "grad_norm": 1.0003550052642822, "learning_rate": 2.019351768682181e-05, "loss": 0.025, "step": 5906 }, { "epoch": 0.40592358438702586, "grad_norm": 0.6786086559295654, "learning_rate": 2.0190385487165804e-05, "loss": 0.0295, "step": 5907 }, { "epoch": 0.40599230346344145, "grad_norm": 0.7706063389778137, "learning_rate": 2.0187253030388768e-05, "loss": 0.024, "step": 5908 }, { "epoch": 0.40606102253985704, "grad_norm": 0.788479745388031, "learning_rate": 2.0184120316645883e-05, "loss": 0.0276, "step": 5909 }, { "epoch": 0.4061297416162727, "grad_norm": 0.3848187029361725, "learning_rate": 2.018098734609234e-05, "loss": 0.025, "step": 5910 }, { "epoch": 0.4061984606926883, "grad_norm": 0.41784772276878357, "learning_rate": 2.0177854118883327e-05, "loss": 0.0217, "step": 5911 }, { "epoch": 0.4062671797691039, "grad_norm": 1.0838127136230469, "learning_rate": 2.0174720635174073e-05, "loss": 0.0207, "step": 5912 }, { "epoch": 0.4063358988455195, "grad_norm": 0.28649893403053284, "learning_rate": 2.0171586895119794e-05, "loss": 0.0188, "step": 5913 }, { "epoch": 0.4064046179219351, "grad_norm": 0.4286256432533264, "learning_rate": 2.0168452898875735e-05, "loss": 0.0163, "step": 5914 }, { "epoch": 0.40647333699835075, "grad_norm": 0.41805288195610046, "learning_rate": 2.0165318646597143e-05, "loss": 0.0147, "step": 5915 }, { "epoch": 0.40654205607476634, "grad_norm": 0.8455459475517273, "learning_rate": 2.0162184138439286e-05, "loss": 0.0259, "step": 5916 }, { "epoch": 0.406610775151182, "grad_norm": 2.1351702213287354, "learning_rate": 2.015904937455744e-05, "loss": 0.0242, "step": 5917 }, { "epoch": 0.4066794942275976, "grad_norm": 0.4318152964115143, "learning_rate": 2.015591435510689e-05, "loss": 0.0202, "step": 5918 }, { "epoch": 0.4067482133040132, "grad_norm": 1.0654497146606445, "learning_rate": 2.015277908024295e-05, "loss": 0.0296, "step": 5919 }, { "epoch": 0.4068169323804288, "grad_norm": 0.6521778702735901, "learning_rate": 2.0149643550120918e-05, "loss": 0.031, "step": 5920 }, { "epoch": 0.4068856514568444, "grad_norm": 0.8103227615356445, "learning_rate": 2.0146507764896142e-05, "loss": 0.027, "step": 5921 }, { "epoch": 0.40695437053326006, "grad_norm": 0.707560658454895, "learning_rate": 2.014337172472395e-05, "loss": 0.0152, "step": 5922 }, { "epoch": 0.40702308960967565, "grad_norm": 1.4972681999206543, "learning_rate": 2.0140235429759696e-05, "loss": 0.0197, "step": 5923 }, { "epoch": 0.40709180868609124, "grad_norm": 0.9792076349258423, "learning_rate": 2.013709888015874e-05, "loss": 0.0226, "step": 5924 }, { "epoch": 0.4071605277625069, "grad_norm": 0.970737636089325, "learning_rate": 2.013396207607648e-05, "loss": 0.0361, "step": 5925 }, { "epoch": 0.4072292468389225, "grad_norm": 1.3144187927246094, "learning_rate": 2.0130825017668283e-05, "loss": 0.0286, "step": 5926 }, { "epoch": 0.4072979659153381, "grad_norm": 1.30510675907135, "learning_rate": 2.0127687705089566e-05, "loss": 0.0239, "step": 5927 }, { "epoch": 0.4073666849917537, "grad_norm": 0.39524251222610474, "learning_rate": 2.0124550138495745e-05, "loss": 0.0225, "step": 5928 }, { "epoch": 0.4074354040681693, "grad_norm": 0.531923234462738, "learning_rate": 2.0121412318042243e-05, "loss": 0.021, "step": 5929 }, { "epoch": 0.40750412314458495, "grad_norm": 1.6265422105789185, "learning_rate": 2.0118274243884505e-05, "loss": 0.0203, "step": 5930 }, { "epoch": 0.40757284222100054, "grad_norm": 0.44035041332244873, "learning_rate": 2.0115135916177984e-05, "loss": 0.0205, "step": 5931 }, { "epoch": 0.4076415612974162, "grad_norm": 1.1183065176010132, "learning_rate": 2.0111997335078148e-05, "loss": 0.0321, "step": 5932 }, { "epoch": 0.4077102803738318, "grad_norm": 0.9087429642677307, "learning_rate": 2.0108858500740472e-05, "loss": 0.0231, "step": 5933 }, { "epoch": 0.40777899945024737, "grad_norm": 0.5260049700737, "learning_rate": 2.0105719413320448e-05, "loss": 0.0253, "step": 5934 }, { "epoch": 0.407847718526663, "grad_norm": 1.1406121253967285, "learning_rate": 2.0102580072973584e-05, "loss": 0.028, "step": 5935 }, { "epoch": 0.4079164376030786, "grad_norm": 0.8268731832504272, "learning_rate": 2.0099440479855388e-05, "loss": 0.0211, "step": 5936 }, { "epoch": 0.40798515667949425, "grad_norm": 1.7980597019195557, "learning_rate": 2.00963006341214e-05, "loss": 0.0259, "step": 5937 }, { "epoch": 0.40805387575590985, "grad_norm": 0.6889641284942627, "learning_rate": 2.0093160535927154e-05, "loss": 0.0227, "step": 5938 }, { "epoch": 0.40812259483232544, "grad_norm": 0.7253490090370178, "learning_rate": 2.0090020185428205e-05, "loss": 0.0232, "step": 5939 }, { "epoch": 0.4081913139087411, "grad_norm": 0.6905197501182556, "learning_rate": 2.0086879582780123e-05, "loss": 0.0298, "step": 5940 }, { "epoch": 0.4082600329851567, "grad_norm": 0.5960486531257629, "learning_rate": 2.0083738728138483e-05, "loss": 0.0282, "step": 5941 }, { "epoch": 0.4083287520615723, "grad_norm": 0.7729674577713013, "learning_rate": 2.0080597621658877e-05, "loss": 0.0227, "step": 5942 }, { "epoch": 0.4083974711379879, "grad_norm": 0.8750134706497192, "learning_rate": 2.0077456263496912e-05, "loss": 0.0193, "step": 5943 }, { "epoch": 0.4084661902144035, "grad_norm": 0.9065575003623962, "learning_rate": 2.0074314653808205e-05, "loss": 0.0231, "step": 5944 }, { "epoch": 0.40853490929081915, "grad_norm": 1.216631531715393, "learning_rate": 2.0071172792748376e-05, "loss": 0.0353, "step": 5945 }, { "epoch": 0.40860362836723474, "grad_norm": 1.4130038022994995, "learning_rate": 2.0068030680473077e-05, "loss": 0.0363, "step": 5946 }, { "epoch": 0.40867234744365033, "grad_norm": 1.1708217859268188, "learning_rate": 2.006488831713796e-05, "loss": 0.0366, "step": 5947 }, { "epoch": 0.408741066520066, "grad_norm": 0.5447336435317993, "learning_rate": 2.006174570289868e-05, "loss": 0.0181, "step": 5948 }, { "epoch": 0.40880978559648157, "grad_norm": 0.5601971745491028, "learning_rate": 2.0058602837910935e-05, "loss": 0.031, "step": 5949 }, { "epoch": 0.4088785046728972, "grad_norm": 0.9646303057670593, "learning_rate": 2.00554597223304e-05, "loss": 0.0383, "step": 5950 }, { "epoch": 0.4089472237493128, "grad_norm": 0.6477227210998535, "learning_rate": 2.0052316356312786e-05, "loss": 0.0193, "step": 5951 }, { "epoch": 0.4090159428257284, "grad_norm": 0.6822782754898071, "learning_rate": 2.004917274001381e-05, "loss": 0.0217, "step": 5952 }, { "epoch": 0.40908466190214404, "grad_norm": 0.7861528992652893, "learning_rate": 2.004602887358919e-05, "loss": 0.0273, "step": 5953 }, { "epoch": 0.40915338097855963, "grad_norm": 0.7363292574882507, "learning_rate": 2.004288475719467e-05, "loss": 0.0162, "step": 5954 }, { "epoch": 0.4092221000549753, "grad_norm": 0.9387734532356262, "learning_rate": 2.0039740390986017e-05, "loss": 0.0232, "step": 5955 }, { "epoch": 0.40929081913139087, "grad_norm": 0.7444663643836975, "learning_rate": 2.0036595775118983e-05, "loss": 0.0192, "step": 5956 }, { "epoch": 0.40935953820780646, "grad_norm": 0.6779210567474365, "learning_rate": 2.0033450909749347e-05, "loss": 0.0194, "step": 5957 }, { "epoch": 0.4094282572842221, "grad_norm": 1.2114008665084839, "learning_rate": 2.0030305795032904e-05, "loss": 0.0294, "step": 5958 }, { "epoch": 0.4094969763606377, "grad_norm": 0.9485833048820496, "learning_rate": 2.0027160431125453e-05, "loss": 0.024, "step": 5959 }, { "epoch": 0.40956569543705335, "grad_norm": 0.6494934558868408, "learning_rate": 2.0024014818182807e-05, "loss": 0.0174, "step": 5960 }, { "epoch": 0.40963441451346894, "grad_norm": 1.0488569736480713, "learning_rate": 2.00208689563608e-05, "loss": 0.0261, "step": 5961 }, { "epoch": 0.4097031335898845, "grad_norm": 2.01082181930542, "learning_rate": 2.0017722845815266e-05, "loss": 0.0322, "step": 5962 }, { "epoch": 0.4097718526663002, "grad_norm": 0.605908215045929, "learning_rate": 2.0014576486702056e-05, "loss": 0.03, "step": 5963 }, { "epoch": 0.40984057174271576, "grad_norm": 0.7338370084762573, "learning_rate": 2.0011429879177037e-05, "loss": 0.0349, "step": 5964 }, { "epoch": 0.4099092908191314, "grad_norm": 0.8212536573410034, "learning_rate": 2.0008283023396086e-05, "loss": 0.0281, "step": 5965 }, { "epoch": 0.409978009895547, "grad_norm": 0.8731286525726318, "learning_rate": 2.0005135919515093e-05, "loss": 0.035, "step": 5966 }, { "epoch": 0.4100467289719626, "grad_norm": 0.5738387703895569, "learning_rate": 2.0001988567689955e-05, "loss": 0.0233, "step": 5967 }, { "epoch": 0.41011544804837824, "grad_norm": 0.650310218334198, "learning_rate": 1.9998840968076586e-05, "loss": 0.0208, "step": 5968 }, { "epoch": 0.41018416712479383, "grad_norm": 0.49125048518180847, "learning_rate": 1.9995693120830913e-05, "loss": 0.0208, "step": 5969 }, { "epoch": 0.4102528862012095, "grad_norm": 0.9937353730201721, "learning_rate": 1.9992545026108874e-05, "loss": 0.0299, "step": 5970 }, { "epoch": 0.41032160527762507, "grad_norm": 0.6762741804122925, "learning_rate": 1.9989396684066415e-05, "loss": 0.0262, "step": 5971 }, { "epoch": 0.41039032435404066, "grad_norm": 1.3929475545883179, "learning_rate": 1.9986248094859508e-05, "loss": 0.0327, "step": 5972 }, { "epoch": 0.4104590434304563, "grad_norm": 0.6928001642227173, "learning_rate": 1.998309925864412e-05, "loss": 0.0201, "step": 5973 }, { "epoch": 0.4105277625068719, "grad_norm": 1.3310248851776123, "learning_rate": 1.9979950175576237e-05, "loss": 0.048, "step": 5974 }, { "epoch": 0.41059648158328754, "grad_norm": 0.5651316046714783, "learning_rate": 1.9976800845811863e-05, "loss": 0.0207, "step": 5975 }, { "epoch": 0.41066520065970313, "grad_norm": 1.3193926811218262, "learning_rate": 1.9973651269507003e-05, "loss": 0.0325, "step": 5976 }, { "epoch": 0.4107339197361187, "grad_norm": 0.8755330443382263, "learning_rate": 1.9970501446817687e-05, "loss": 0.03, "step": 5977 }, { "epoch": 0.41080263881253437, "grad_norm": 1.4034745693206787, "learning_rate": 1.996735137789995e-05, "loss": 0.0289, "step": 5978 }, { "epoch": 0.41087135788894996, "grad_norm": 1.4839074611663818, "learning_rate": 1.9964201062909835e-05, "loss": 0.0336, "step": 5979 }, { "epoch": 0.4109400769653656, "grad_norm": 1.081175446510315, "learning_rate": 1.9961050502003408e-05, "loss": 0.0399, "step": 5980 }, { "epoch": 0.4110087960417812, "grad_norm": 1.2480562925338745, "learning_rate": 1.9957899695336734e-05, "loss": 0.024, "step": 5981 }, { "epoch": 0.4110775151181968, "grad_norm": 0.6615467667579651, "learning_rate": 1.9954748643065908e-05, "loss": 0.0265, "step": 5982 }, { "epoch": 0.41114623419461244, "grad_norm": 0.6529251933097839, "learning_rate": 1.995159734534702e-05, "loss": 0.0289, "step": 5983 }, { "epoch": 0.411214953271028, "grad_norm": 0.6915336847305298, "learning_rate": 1.9948445802336174e-05, "loss": 0.0422, "step": 5984 }, { "epoch": 0.4112836723474437, "grad_norm": 0.8663614988327026, "learning_rate": 1.9945294014189503e-05, "loss": 0.0201, "step": 5985 }, { "epoch": 0.41135239142385926, "grad_norm": 0.5249864459037781, "learning_rate": 1.9942141981063133e-05, "loss": 0.0252, "step": 5986 }, { "epoch": 0.41142111050027486, "grad_norm": 1.5455955266952515, "learning_rate": 1.9938989703113206e-05, "loss": 0.0252, "step": 5987 }, { "epoch": 0.4114898295766905, "grad_norm": 1.7763913869857788, "learning_rate": 1.9935837180495888e-05, "loss": 0.0353, "step": 5988 }, { "epoch": 0.4115585486531061, "grad_norm": 0.7551748752593994, "learning_rate": 1.993268441336734e-05, "loss": 0.0347, "step": 5989 }, { "epoch": 0.41162726772952174, "grad_norm": 1.1919324398040771, "learning_rate": 1.992953140188375e-05, "loss": 0.0249, "step": 5990 }, { "epoch": 0.41169598680593733, "grad_norm": 0.9734020233154297, "learning_rate": 1.9926378146201308e-05, "loss": 0.0336, "step": 5991 }, { "epoch": 0.4117647058823529, "grad_norm": 0.630888819694519, "learning_rate": 1.992322464647622e-05, "loss": 0.0398, "step": 5992 }, { "epoch": 0.41183342495876857, "grad_norm": 0.7818018794059753, "learning_rate": 1.9920070902864704e-05, "loss": 0.032, "step": 5993 }, { "epoch": 0.41190214403518416, "grad_norm": 0.8370647430419922, "learning_rate": 1.9916916915522994e-05, "loss": 0.0302, "step": 5994 }, { "epoch": 0.4119708631115998, "grad_norm": 0.6917720437049866, "learning_rate": 1.9913762684607327e-05, "loss": 0.0322, "step": 5995 }, { "epoch": 0.4120395821880154, "grad_norm": 1.4118050336837769, "learning_rate": 1.9910608210273958e-05, "loss": 0.0367, "step": 5996 }, { "epoch": 0.412108301264431, "grad_norm": 0.5522433519363403, "learning_rate": 1.9907453492679157e-05, "loss": 0.0183, "step": 5997 }, { "epoch": 0.41217702034084663, "grad_norm": 1.2583814859390259, "learning_rate": 1.9904298531979195e-05, "loss": 0.0286, "step": 5998 }, { "epoch": 0.4122457394172622, "grad_norm": 0.9168598055839539, "learning_rate": 1.9901143328330366e-05, "loss": 0.0308, "step": 5999 }, { "epoch": 0.41231445849367787, "grad_norm": 0.5104197263717651, "learning_rate": 1.9897987881888976e-05, "loss": 0.0228, "step": 6000 }, { "epoch": 0.41238317757009346, "grad_norm": 0.6710358262062073, "learning_rate": 1.9894832192811335e-05, "loss": 0.0239, "step": 6001 }, { "epoch": 0.41245189664650905, "grad_norm": 0.5759415626525879, "learning_rate": 1.989167626125377e-05, "loss": 0.026, "step": 6002 }, { "epoch": 0.4125206157229247, "grad_norm": 1.9952032566070557, "learning_rate": 1.988852008737262e-05, "loss": 0.0486, "step": 6003 }, { "epoch": 0.4125893347993403, "grad_norm": 0.5924586653709412, "learning_rate": 1.9885363671324236e-05, "loss": 0.0276, "step": 6004 }, { "epoch": 0.41265805387575594, "grad_norm": 1.1323636770248413, "learning_rate": 1.988220701326498e-05, "loss": 0.024, "step": 6005 }, { "epoch": 0.41272677295217153, "grad_norm": 0.4898436963558197, "learning_rate": 1.9879050113351224e-05, "loss": 0.0236, "step": 6006 }, { "epoch": 0.4127954920285871, "grad_norm": 0.8015350699424744, "learning_rate": 1.987589297173936e-05, "loss": 0.0251, "step": 6007 }, { "epoch": 0.41286421110500277, "grad_norm": 0.7963072061538696, "learning_rate": 1.987273558858578e-05, "loss": 0.0192, "step": 6008 }, { "epoch": 0.41293293018141836, "grad_norm": 0.4144088327884674, "learning_rate": 1.9869577964046897e-05, "loss": 0.0327, "step": 6009 }, { "epoch": 0.41300164925783395, "grad_norm": 1.1625502109527588, "learning_rate": 1.9866420098279134e-05, "loss": 0.0305, "step": 6010 }, { "epoch": 0.4130703683342496, "grad_norm": 0.7581436038017273, "learning_rate": 1.9863261991438922e-05, "loss": 0.0184, "step": 6011 }, { "epoch": 0.4131390874106652, "grad_norm": 0.6541237235069275, "learning_rate": 1.9860103643682713e-05, "loss": 0.0297, "step": 6012 }, { "epoch": 0.41320780648708083, "grad_norm": 0.9669374227523804, "learning_rate": 1.985694505516696e-05, "loss": 0.022, "step": 6013 }, { "epoch": 0.4132765255634964, "grad_norm": 0.7663697004318237, "learning_rate": 1.985378622604813e-05, "loss": 0.0321, "step": 6014 }, { "epoch": 0.413345244639912, "grad_norm": 0.7727714776992798, "learning_rate": 1.9850627156482716e-05, "loss": 0.0363, "step": 6015 }, { "epoch": 0.41341396371632766, "grad_norm": 1.2850440740585327, "learning_rate": 1.98474678466272e-05, "loss": 0.0383, "step": 6016 }, { "epoch": 0.41348268279274325, "grad_norm": 0.5567431449890137, "learning_rate": 1.9844308296638095e-05, "loss": 0.0329, "step": 6017 }, { "epoch": 0.4135514018691589, "grad_norm": 0.5337805151939392, "learning_rate": 1.984114850667192e-05, "loss": 0.0371, "step": 6018 }, { "epoch": 0.4136201209455745, "grad_norm": 1.1148592233657837, "learning_rate": 1.9837988476885197e-05, "loss": 0.0368, "step": 6019 }, { "epoch": 0.4136888400219901, "grad_norm": 1.1695783138275146, "learning_rate": 1.9834828207434468e-05, "loss": 0.0328, "step": 6020 }, { "epoch": 0.4137575590984057, "grad_norm": 0.5218837261199951, "learning_rate": 1.9831667698476296e-05, "loss": 0.0213, "step": 6021 }, { "epoch": 0.4138262781748213, "grad_norm": 0.6571851372718811, "learning_rate": 1.982850695016724e-05, "loss": 0.0171, "step": 6022 }, { "epoch": 0.41389499725123696, "grad_norm": 1.418282389640808, "learning_rate": 1.982534596266387e-05, "loss": 0.0244, "step": 6023 }, { "epoch": 0.41396371632765255, "grad_norm": 1.181222677230835, "learning_rate": 1.9822184736122786e-05, "loss": 0.0445, "step": 6024 }, { "epoch": 0.41403243540406814, "grad_norm": 0.3695487678050995, "learning_rate": 1.9819023270700584e-05, "loss": 0.0214, "step": 6025 }, { "epoch": 0.4141011544804838, "grad_norm": 0.8208456039428711, "learning_rate": 1.9815861566553878e-05, "loss": 0.0247, "step": 6026 }, { "epoch": 0.4141698735568994, "grad_norm": 0.5695506930351257, "learning_rate": 1.981269962383929e-05, "loss": 0.0276, "step": 6027 }, { "epoch": 0.41423859263331503, "grad_norm": 0.7003195881843567, "learning_rate": 1.9809537442713457e-05, "loss": 0.0193, "step": 6028 }, { "epoch": 0.4143073117097306, "grad_norm": 0.8881135582923889, "learning_rate": 1.9806375023333026e-05, "loss": 0.0307, "step": 6029 }, { "epoch": 0.4143760307861462, "grad_norm": 1.1388863325119019, "learning_rate": 1.9803212365854655e-05, "loss": 0.0351, "step": 6030 }, { "epoch": 0.41444474986256186, "grad_norm": 1.2649712562561035, "learning_rate": 1.9800049470435022e-05, "loss": 0.0523, "step": 6031 }, { "epoch": 0.41451346893897745, "grad_norm": 0.6427191495895386, "learning_rate": 1.9796886337230808e-05, "loss": 0.0202, "step": 6032 }, { "epoch": 0.4145821880153931, "grad_norm": 0.8563979268074036, "learning_rate": 1.9793722966398707e-05, "loss": 0.0333, "step": 6033 }, { "epoch": 0.4146509070918087, "grad_norm": 0.595766007900238, "learning_rate": 1.979055935809542e-05, "loss": 0.0249, "step": 6034 }, { "epoch": 0.4147196261682243, "grad_norm": 0.7686552405357361, "learning_rate": 1.9787395512477674e-05, "loss": 0.0179, "step": 6035 }, { "epoch": 0.4147883452446399, "grad_norm": 0.9204010963439941, "learning_rate": 1.9784231429702204e-05, "loss": 0.0408, "step": 6036 }, { "epoch": 0.4148570643210555, "grad_norm": 0.25206971168518066, "learning_rate": 1.9781067109925736e-05, "loss": 0.0141, "step": 6037 }, { "epoch": 0.41492578339747116, "grad_norm": 0.8849743008613586, "learning_rate": 1.9777902553305033e-05, "loss": 0.0198, "step": 6038 }, { "epoch": 0.41499450247388675, "grad_norm": 0.48457658290863037, "learning_rate": 1.9774737759996866e-05, "loss": 0.0193, "step": 6039 }, { "epoch": 0.41506322155030234, "grad_norm": 0.6239244937896729, "learning_rate": 1.9771572730158005e-05, "loss": 0.0291, "step": 6040 }, { "epoch": 0.415131940626718, "grad_norm": 0.916280210018158, "learning_rate": 1.9768407463945235e-05, "loss": 0.0203, "step": 6041 }, { "epoch": 0.4152006597031336, "grad_norm": 0.6448331475257874, "learning_rate": 1.976524196151537e-05, "loss": 0.0163, "step": 6042 }, { "epoch": 0.4152693787795492, "grad_norm": 0.9086791276931763, "learning_rate": 1.9762076223025214e-05, "loss": 0.0246, "step": 6043 }, { "epoch": 0.4153380978559648, "grad_norm": 1.1379843950271606, "learning_rate": 1.9758910248631592e-05, "loss": 0.0212, "step": 6044 }, { "epoch": 0.4154068169323804, "grad_norm": 0.6788846254348755, "learning_rate": 1.975574403849134e-05, "loss": 0.0242, "step": 6045 }, { "epoch": 0.41547553600879605, "grad_norm": 0.550553023815155, "learning_rate": 1.9752577592761305e-05, "loss": 0.0202, "step": 6046 }, { "epoch": 0.41554425508521164, "grad_norm": 0.8530102372169495, "learning_rate": 1.9749410911598344e-05, "loss": 0.0254, "step": 6047 }, { "epoch": 0.4156129741616273, "grad_norm": 0.6337419748306274, "learning_rate": 1.9746243995159333e-05, "loss": 0.0185, "step": 6048 }, { "epoch": 0.4156816932380429, "grad_norm": 0.6543214917182922, "learning_rate": 1.9743076843601158e-05, "loss": 0.0177, "step": 6049 }, { "epoch": 0.4157504123144585, "grad_norm": 0.7320078015327454, "learning_rate": 1.9739909457080698e-05, "loss": 0.0295, "step": 6050 }, { "epoch": 0.4158191313908741, "grad_norm": 0.9499195218086243, "learning_rate": 1.9736741835754875e-05, "loss": 0.0239, "step": 6051 }, { "epoch": 0.4158878504672897, "grad_norm": 1.4877561330795288, "learning_rate": 1.9733573979780598e-05, "loss": 0.033, "step": 6052 }, { "epoch": 0.41595656954370536, "grad_norm": 0.7315653562545776, "learning_rate": 1.9730405889314796e-05, "loss": 0.0099, "step": 6053 }, { "epoch": 0.41602528862012095, "grad_norm": 0.786589503288269, "learning_rate": 1.9727237564514413e-05, "loss": 0.0442, "step": 6054 }, { "epoch": 0.41609400769653654, "grad_norm": 0.62382972240448, "learning_rate": 1.9724069005536402e-05, "loss": 0.015, "step": 6055 }, { "epoch": 0.4161627267729522, "grad_norm": 0.8182750344276428, "learning_rate": 1.972090021253772e-05, "loss": 0.029, "step": 6056 }, { "epoch": 0.4162314458493678, "grad_norm": 0.3897152841091156, "learning_rate": 1.9717731185675352e-05, "loss": 0.0156, "step": 6057 }, { "epoch": 0.4163001649257834, "grad_norm": 0.9672663807868958, "learning_rate": 1.971456192510628e-05, "loss": 0.0262, "step": 6058 }, { "epoch": 0.416368884002199, "grad_norm": 1.1827784776687622, "learning_rate": 1.9711392430987498e-05, "loss": 0.0259, "step": 6059 }, { "epoch": 0.4164376030786146, "grad_norm": 0.9200922846794128, "learning_rate": 1.9708222703476027e-05, "loss": 0.0412, "step": 6060 }, { "epoch": 0.41650632215503025, "grad_norm": 0.5133207440376282, "learning_rate": 1.970505274272888e-05, "loss": 0.0182, "step": 6061 }, { "epoch": 0.41657504123144584, "grad_norm": 0.8863222599029541, "learning_rate": 1.9701882548903096e-05, "loss": 0.0212, "step": 6062 }, { "epoch": 0.4166437603078615, "grad_norm": 1.1221672296524048, "learning_rate": 1.9698712122155715e-05, "loss": 0.0265, "step": 6063 }, { "epoch": 0.4167124793842771, "grad_norm": 0.553817093372345, "learning_rate": 1.9695541462643795e-05, "loss": 0.0281, "step": 6064 }, { "epoch": 0.41678119846069267, "grad_norm": 0.771730899810791, "learning_rate": 1.9692370570524404e-05, "loss": 0.0222, "step": 6065 }, { "epoch": 0.4168499175371083, "grad_norm": 0.7387762665748596, "learning_rate": 1.9689199445954622e-05, "loss": 0.0191, "step": 6066 }, { "epoch": 0.4169186366135239, "grad_norm": 0.9469366669654846, "learning_rate": 1.9686028089091543e-05, "loss": 0.0264, "step": 6067 }, { "epoch": 0.41698735568993955, "grad_norm": 0.5702792406082153, "learning_rate": 1.968285650009226e-05, "loss": 0.0199, "step": 6068 }, { "epoch": 0.41705607476635514, "grad_norm": 0.6347740292549133, "learning_rate": 1.9679684679113904e-05, "loss": 0.0215, "step": 6069 }, { "epoch": 0.41712479384277074, "grad_norm": 0.48243239521980286, "learning_rate": 1.9676512626313583e-05, "loss": 0.011, "step": 6070 }, { "epoch": 0.4171935129191864, "grad_norm": 0.9007048010826111, "learning_rate": 1.967334034184844e-05, "loss": 0.0212, "step": 6071 }, { "epoch": 0.417262231995602, "grad_norm": 3.7751190662384033, "learning_rate": 1.9670167825875626e-05, "loss": 0.0338, "step": 6072 }, { "epoch": 0.4173309510720176, "grad_norm": 0.6784322261810303, "learning_rate": 1.9666995078552304e-05, "loss": 0.0235, "step": 6073 }, { "epoch": 0.4173996701484332, "grad_norm": 0.9205636382102966, "learning_rate": 1.9663822100035634e-05, "loss": 0.0285, "step": 6074 }, { "epoch": 0.4174683892248488, "grad_norm": 0.6925716996192932, "learning_rate": 1.9660648890482807e-05, "loss": 0.0301, "step": 6075 }, { "epoch": 0.41753710830126445, "grad_norm": 1.0564342737197876, "learning_rate": 1.965747545005102e-05, "loss": 0.0216, "step": 6076 }, { "epoch": 0.41760582737768004, "grad_norm": 0.7607070803642273, "learning_rate": 1.965430177889747e-05, "loss": 0.0171, "step": 6077 }, { "epoch": 0.41767454645409563, "grad_norm": 0.4441834092140198, "learning_rate": 1.965112787717938e-05, "loss": 0.0286, "step": 6078 }, { "epoch": 0.4177432655305113, "grad_norm": 1.0434155464172363, "learning_rate": 1.964795374505398e-05, "loss": 0.0308, "step": 6079 }, { "epoch": 0.41781198460692687, "grad_norm": 1.1293922662734985, "learning_rate": 1.9644779382678503e-05, "loss": 0.0423, "step": 6080 }, { "epoch": 0.4178807036833425, "grad_norm": 1.0858656167984009, "learning_rate": 1.964160479021021e-05, "loss": 0.04, "step": 6081 }, { "epoch": 0.4179494227597581, "grad_norm": 0.9406305551528931, "learning_rate": 1.963842996780635e-05, "loss": 0.0286, "step": 6082 }, { "epoch": 0.4180181418361737, "grad_norm": 0.7850381135940552, "learning_rate": 1.963525491562421e-05, "loss": 0.0233, "step": 6083 }, { "epoch": 0.41808686091258934, "grad_norm": 0.7035677433013916, "learning_rate": 1.9632079633821074e-05, "loss": 0.0224, "step": 6084 }, { "epoch": 0.41815557998900493, "grad_norm": 0.8814688324928284, "learning_rate": 1.9628904122554235e-05, "loss": 0.0274, "step": 6085 }, { "epoch": 0.4182242990654206, "grad_norm": 0.7162649035453796, "learning_rate": 1.9625728381981e-05, "loss": 0.0269, "step": 6086 }, { "epoch": 0.41829301814183617, "grad_norm": 2.85758113861084, "learning_rate": 1.9622552412258694e-05, "loss": 0.0222, "step": 6087 }, { "epoch": 0.41836173721825176, "grad_norm": 0.9674971103668213, "learning_rate": 1.961937621354464e-05, "loss": 0.0299, "step": 6088 }, { "epoch": 0.4184304562946674, "grad_norm": 0.389921098947525, "learning_rate": 1.961619978599619e-05, "loss": 0.0277, "step": 6089 }, { "epoch": 0.418499175371083, "grad_norm": 0.5721850395202637, "learning_rate": 1.9613023129770695e-05, "loss": 0.0181, "step": 6090 }, { "epoch": 0.41856789444749865, "grad_norm": 0.3668921887874603, "learning_rate": 1.9609846245025512e-05, "loss": 0.016, "step": 6091 }, { "epoch": 0.41863661352391424, "grad_norm": 2.6787993907928467, "learning_rate": 1.9606669131918028e-05, "loss": 0.028, "step": 6092 }, { "epoch": 0.4187053326003298, "grad_norm": 0.5518250465393066, "learning_rate": 1.9603491790605627e-05, "loss": 0.0274, "step": 6093 }, { "epoch": 0.4187740516767455, "grad_norm": 1.4945068359375, "learning_rate": 1.9600314221245705e-05, "loss": 0.0276, "step": 6094 }, { "epoch": 0.41884277075316106, "grad_norm": 0.8954306244850159, "learning_rate": 1.9597136423995668e-05, "loss": 0.0273, "step": 6095 }, { "epoch": 0.4189114898295767, "grad_norm": 1.2168266773223877, "learning_rate": 1.9593958399012953e-05, "loss": 0.0267, "step": 6096 }, { "epoch": 0.4189802089059923, "grad_norm": 0.606942355632782, "learning_rate": 1.9590780146454982e-05, "loss": 0.0187, "step": 6097 }, { "epoch": 0.4190489279824079, "grad_norm": 0.3794560432434082, "learning_rate": 1.9587601666479196e-05, "loss": 0.0207, "step": 6098 }, { "epoch": 0.41911764705882354, "grad_norm": 1.460953950881958, "learning_rate": 1.958442295924306e-05, "loss": 0.0182, "step": 6099 }, { "epoch": 0.41918636613523913, "grad_norm": 0.26254820823669434, "learning_rate": 1.958124402490403e-05, "loss": 0.0186, "step": 6100 }, { "epoch": 0.4192550852116548, "grad_norm": 0.5181288123130798, "learning_rate": 1.9578064863619597e-05, "loss": 0.0298, "step": 6101 }, { "epoch": 0.41932380428807037, "grad_norm": 0.7228912115097046, "learning_rate": 1.957488547554724e-05, "loss": 0.0245, "step": 6102 }, { "epoch": 0.41939252336448596, "grad_norm": 0.8792996406555176, "learning_rate": 1.9571705860844465e-05, "loss": 0.0366, "step": 6103 }, { "epoch": 0.4194612424409016, "grad_norm": 0.8652681708335876, "learning_rate": 1.9568526019668778e-05, "loss": 0.0198, "step": 6104 }, { "epoch": 0.4195299615173172, "grad_norm": 0.4879956543445587, "learning_rate": 1.9565345952177702e-05, "loss": 0.0431, "step": 6105 }, { "epoch": 0.41959868059373284, "grad_norm": 0.7966824769973755, "learning_rate": 1.956216565852878e-05, "loss": 0.021, "step": 6106 }, { "epoch": 0.41966739967014843, "grad_norm": 0.6601628661155701, "learning_rate": 1.955898513887954e-05, "loss": 0.023, "step": 6107 }, { "epoch": 0.419736118746564, "grad_norm": 0.4539530873298645, "learning_rate": 1.9555804393387557e-05, "loss": 0.0232, "step": 6108 }, { "epoch": 0.41980483782297967, "grad_norm": 0.5209857225418091, "learning_rate": 1.9552623422210392e-05, "loss": 0.0194, "step": 6109 }, { "epoch": 0.41987355689939526, "grad_norm": 0.41079169511795044, "learning_rate": 1.9549442225505613e-05, "loss": 0.0269, "step": 6110 }, { "epoch": 0.4199422759758109, "grad_norm": 2.922165870666504, "learning_rate": 1.9546260803430832e-05, "loss": 0.0342, "step": 6111 }, { "epoch": 0.4200109950522265, "grad_norm": 0.41004812717437744, "learning_rate": 1.954307915614363e-05, "loss": 0.0295, "step": 6112 }, { "epoch": 0.4200797141286421, "grad_norm": 1.011665940284729, "learning_rate": 1.9539897283801626e-05, "loss": 0.0413, "step": 6113 }, { "epoch": 0.42014843320505774, "grad_norm": 0.8145806789398193, "learning_rate": 1.9536715186562445e-05, "loss": 0.0177, "step": 6114 }, { "epoch": 0.4202171522814733, "grad_norm": 0.6499778628349304, "learning_rate": 1.9533532864583727e-05, "loss": 0.0186, "step": 6115 }, { "epoch": 0.420285871357889, "grad_norm": 0.6037622690200806, "learning_rate": 1.9530350318023103e-05, "loss": 0.0203, "step": 6116 }, { "epoch": 0.42035459043430456, "grad_norm": 0.5213770866394043, "learning_rate": 1.9527167547038243e-05, "loss": 0.0228, "step": 6117 }, { "epoch": 0.42042330951072016, "grad_norm": 0.8211660385131836, "learning_rate": 1.9523984551786807e-05, "loss": 0.0357, "step": 6118 }, { "epoch": 0.4204920285871358, "grad_norm": 0.5553194880485535, "learning_rate": 1.952080133242648e-05, "loss": 0.0223, "step": 6119 }, { "epoch": 0.4205607476635514, "grad_norm": 1.0630378723144531, "learning_rate": 1.951761788911495e-05, "loss": 0.028, "step": 6120 }, { "epoch": 0.42062946673996704, "grad_norm": 0.9927130341529846, "learning_rate": 1.9514434222009918e-05, "loss": 0.0253, "step": 6121 }, { "epoch": 0.42069818581638263, "grad_norm": 1.2103471755981445, "learning_rate": 1.951125033126909e-05, "loss": 0.0307, "step": 6122 }, { "epoch": 0.4207669048927982, "grad_norm": 0.6176916360855103, "learning_rate": 1.95080662170502e-05, "loss": 0.0184, "step": 6123 }, { "epoch": 0.42083562396921387, "grad_norm": 0.6393261551856995, "learning_rate": 1.950488187951098e-05, "loss": 0.0243, "step": 6124 }, { "epoch": 0.42090434304562946, "grad_norm": 0.6328718662261963, "learning_rate": 1.9501697318809166e-05, "loss": 0.0256, "step": 6125 }, { "epoch": 0.4209730621220451, "grad_norm": 0.524618923664093, "learning_rate": 1.949851253510253e-05, "loss": 0.0276, "step": 6126 }, { "epoch": 0.4210417811984607, "grad_norm": 0.5087026953697205, "learning_rate": 1.9495327528548827e-05, "loss": 0.0248, "step": 6127 }, { "epoch": 0.4211105002748763, "grad_norm": 0.7505837082862854, "learning_rate": 1.9492142299305842e-05, "loss": 0.0277, "step": 6128 }, { "epoch": 0.42117921935129193, "grad_norm": 0.5584938526153564, "learning_rate": 1.948895684753136e-05, "loss": 0.0293, "step": 6129 }, { "epoch": 0.4212479384277075, "grad_norm": 0.4726874530315399, "learning_rate": 1.948577117338319e-05, "loss": 0.0176, "step": 6130 }, { "epoch": 0.42131665750412317, "grad_norm": 2.2190678119659424, "learning_rate": 1.948258527701913e-05, "loss": 0.0361, "step": 6131 }, { "epoch": 0.42138537658053876, "grad_norm": 0.41910281777381897, "learning_rate": 1.947939915859702e-05, "loss": 0.0229, "step": 6132 }, { "epoch": 0.42145409565695435, "grad_norm": 1.1413638591766357, "learning_rate": 1.9476212818274682e-05, "loss": 0.0227, "step": 6133 }, { "epoch": 0.42152281473337, "grad_norm": 0.7027429342269897, "learning_rate": 1.947302625620996e-05, "loss": 0.0338, "step": 6134 }, { "epoch": 0.4215915338097856, "grad_norm": 0.8666156530380249, "learning_rate": 1.9469839472560716e-05, "loss": 0.0242, "step": 6135 }, { "epoch": 0.42166025288620124, "grad_norm": 1.132311224937439, "learning_rate": 1.9466652467484815e-05, "loss": 0.0346, "step": 6136 }, { "epoch": 0.4217289719626168, "grad_norm": 0.7419387698173523, "learning_rate": 1.9463465241140134e-05, "loss": 0.02, "step": 6137 }, { "epoch": 0.4217976910390324, "grad_norm": 0.8285346031188965, "learning_rate": 1.9460277793684564e-05, "loss": 0.0168, "step": 6138 }, { "epoch": 0.42186641011544807, "grad_norm": 0.43684181571006775, "learning_rate": 1.9457090125275996e-05, "loss": 0.0269, "step": 6139 }, { "epoch": 0.42193512919186366, "grad_norm": 1.3554713726043701, "learning_rate": 1.945390223607235e-05, "loss": 0.0242, "step": 6140 }, { "epoch": 0.42200384826827925, "grad_norm": 0.5828022956848145, "learning_rate": 1.9450714126231546e-05, "loss": 0.0169, "step": 6141 }, { "epoch": 0.4220725673446949, "grad_norm": 0.9832371473312378, "learning_rate": 1.9447525795911515e-05, "loss": 0.0188, "step": 6142 }, { "epoch": 0.4221412864211105, "grad_norm": 0.28891438245773315, "learning_rate": 1.9444337245270195e-05, "loss": 0.015, "step": 6143 }, { "epoch": 0.42221000549752613, "grad_norm": 1.0052059888839722, "learning_rate": 1.9441148474465547e-05, "loss": 0.0204, "step": 6144 }, { "epoch": 0.4222787245739417, "grad_norm": 0.8216879367828369, "learning_rate": 1.9437959483655532e-05, "loss": 0.0192, "step": 6145 }, { "epoch": 0.4223474436503573, "grad_norm": 0.7987598776817322, "learning_rate": 1.943477027299813e-05, "loss": 0.0389, "step": 6146 }, { "epoch": 0.42241616272677296, "grad_norm": 0.8334159851074219, "learning_rate": 1.9431580842651324e-05, "loss": 0.0247, "step": 6147 }, { "epoch": 0.42248488180318855, "grad_norm": 0.9125418066978455, "learning_rate": 1.9428391192773122e-05, "loss": 0.016, "step": 6148 }, { "epoch": 0.4225536008796042, "grad_norm": 0.8480127453804016, "learning_rate": 1.9425201323521515e-05, "loss": 0.0224, "step": 6149 }, { "epoch": 0.4226223199560198, "grad_norm": 0.6030243635177612, "learning_rate": 1.9422011235054538e-05, "loss": 0.0275, "step": 6150 }, { "epoch": 0.4226910390324354, "grad_norm": 0.4892015755176544, "learning_rate": 1.9418820927530214e-05, "loss": 0.0196, "step": 6151 }, { "epoch": 0.422759758108851, "grad_norm": 0.8801403045654297, "learning_rate": 1.9415630401106587e-05, "loss": 0.0224, "step": 6152 }, { "epoch": 0.4228284771852666, "grad_norm": 1.099768877029419, "learning_rate": 1.9412439655941705e-05, "loss": 0.026, "step": 6153 }, { "epoch": 0.42289719626168226, "grad_norm": 0.44902482628822327, "learning_rate": 1.9409248692193636e-05, "loss": 0.0156, "step": 6154 }, { "epoch": 0.42296591533809785, "grad_norm": 0.4701327681541443, "learning_rate": 1.9406057510020455e-05, "loss": 0.0216, "step": 6155 }, { "epoch": 0.42303463441451344, "grad_norm": 0.475847065448761, "learning_rate": 1.940286610958024e-05, "loss": 0.0227, "step": 6156 }, { "epoch": 0.4231033534909291, "grad_norm": 0.5879462361335754, "learning_rate": 1.939967449103109e-05, "loss": 0.0213, "step": 6157 }, { "epoch": 0.4231720725673447, "grad_norm": 0.8688351511955261, "learning_rate": 1.9396482654531114e-05, "loss": 0.0248, "step": 6158 }, { "epoch": 0.42324079164376033, "grad_norm": 1.1806944608688354, "learning_rate": 1.939329060023842e-05, "loss": 0.0257, "step": 6159 }, { "epoch": 0.4233095107201759, "grad_norm": 1.2881149053573608, "learning_rate": 1.9390098328311146e-05, "loss": 0.0191, "step": 6160 }, { "epoch": 0.4233782297965915, "grad_norm": 0.7825740575790405, "learning_rate": 1.9386905838907427e-05, "loss": 0.0259, "step": 6161 }, { "epoch": 0.42344694887300716, "grad_norm": 1.0181797742843628, "learning_rate": 1.9383713132185408e-05, "loss": 0.0309, "step": 6162 }, { "epoch": 0.42351566794942275, "grad_norm": 0.6731084585189819, "learning_rate": 1.9380520208303257e-05, "loss": 0.023, "step": 6163 }, { "epoch": 0.4235843870258384, "grad_norm": 0.42995086312294006, "learning_rate": 1.937732706741914e-05, "loss": 0.0227, "step": 6164 }, { "epoch": 0.423653106102254, "grad_norm": 0.3636332154273987, "learning_rate": 1.937413370969124e-05, "loss": 0.0289, "step": 6165 }, { "epoch": 0.4237218251786696, "grad_norm": 0.795788049697876, "learning_rate": 1.9370940135277745e-05, "loss": 0.0201, "step": 6166 }, { "epoch": 0.4237905442550852, "grad_norm": 1.435412049293518, "learning_rate": 1.9367746344336864e-05, "loss": 0.0316, "step": 6167 }, { "epoch": 0.4238592633315008, "grad_norm": 1.4327434301376343, "learning_rate": 1.936455233702681e-05, "loss": 0.0217, "step": 6168 }, { "epoch": 0.42392798240791646, "grad_norm": 0.6750890016555786, "learning_rate": 1.9361358113505806e-05, "loss": 0.0219, "step": 6169 }, { "epoch": 0.42399670148433205, "grad_norm": 1.3088346719741821, "learning_rate": 1.9358163673932087e-05, "loss": 0.0252, "step": 6170 }, { "epoch": 0.42406542056074764, "grad_norm": 0.7675501108169556, "learning_rate": 1.93549690184639e-05, "loss": 0.0193, "step": 6171 }, { "epoch": 0.4241341396371633, "grad_norm": 0.5338214039802551, "learning_rate": 1.9351774147259506e-05, "loss": 0.0195, "step": 6172 }, { "epoch": 0.4242028587135789, "grad_norm": 1.0735951662063599, "learning_rate": 1.934857906047716e-05, "loss": 0.0209, "step": 6173 }, { "epoch": 0.4242715777899945, "grad_norm": 0.8923740983009338, "learning_rate": 1.9345383758275153e-05, "loss": 0.0246, "step": 6174 }, { "epoch": 0.4243402968664101, "grad_norm": 0.9654259085655212, "learning_rate": 1.9342188240811768e-05, "loss": 0.0266, "step": 6175 }, { "epoch": 0.4244090159428257, "grad_norm": 0.5970586538314819, "learning_rate": 1.9338992508245302e-05, "loss": 0.0163, "step": 6176 }, { "epoch": 0.42447773501924135, "grad_norm": 0.8252625465393066, "learning_rate": 1.9335796560734067e-05, "loss": 0.0184, "step": 6177 }, { "epoch": 0.42454645409565694, "grad_norm": 0.815316379070282, "learning_rate": 1.9332600398436386e-05, "loss": 0.0232, "step": 6178 }, { "epoch": 0.4246151731720726, "grad_norm": 0.6878266334533691, "learning_rate": 1.932940402151059e-05, "loss": 0.0424, "step": 6179 }, { "epoch": 0.4246838922484882, "grad_norm": 0.7493141889572144, "learning_rate": 1.9326207430115013e-05, "loss": 0.0252, "step": 6180 }, { "epoch": 0.4247526113249038, "grad_norm": 0.8164727091789246, "learning_rate": 1.9323010624408025e-05, "loss": 0.0331, "step": 6181 }, { "epoch": 0.4248213304013194, "grad_norm": 0.9447790384292603, "learning_rate": 1.931981360454797e-05, "loss": 0.0294, "step": 6182 }, { "epoch": 0.424890049477735, "grad_norm": 0.7142377495765686, "learning_rate": 1.9316616370693234e-05, "loss": 0.0179, "step": 6183 }, { "epoch": 0.42495876855415066, "grad_norm": 0.47137582302093506, "learning_rate": 1.9313418923002197e-05, "loss": 0.0164, "step": 6184 }, { "epoch": 0.42502748763056625, "grad_norm": 0.9394602179527283, "learning_rate": 1.9310221261633256e-05, "loss": 0.0256, "step": 6185 }, { "epoch": 0.42509620670698184, "grad_norm": 1.079962134361267, "learning_rate": 1.9307023386744808e-05, "loss": 0.029, "step": 6186 }, { "epoch": 0.4251649257833975, "grad_norm": 0.9391701221466064, "learning_rate": 1.9303825298495283e-05, "loss": 0.0254, "step": 6187 }, { "epoch": 0.4252336448598131, "grad_norm": 0.6607189774513245, "learning_rate": 1.9300626997043103e-05, "loss": 0.0203, "step": 6188 }, { "epoch": 0.4253023639362287, "grad_norm": 0.7208991646766663, "learning_rate": 1.9297428482546694e-05, "loss": 0.0218, "step": 6189 }, { "epoch": 0.4253710830126443, "grad_norm": 0.22161385416984558, "learning_rate": 1.929422975516452e-05, "loss": 0.0254, "step": 6190 }, { "epoch": 0.4254398020890599, "grad_norm": 0.49820783734321594, "learning_rate": 1.9291030815055035e-05, "loss": 0.0254, "step": 6191 }, { "epoch": 0.42550852116547555, "grad_norm": 0.5542923212051392, "learning_rate": 1.9287831662376702e-05, "loss": 0.02, "step": 6192 }, { "epoch": 0.42557724024189114, "grad_norm": 0.922718346118927, "learning_rate": 1.9284632297288e-05, "loss": 0.0288, "step": 6193 }, { "epoch": 0.4256459593183068, "grad_norm": 0.5365665555000305, "learning_rate": 1.928143271994743e-05, "loss": 0.0196, "step": 6194 }, { "epoch": 0.4257146783947224, "grad_norm": 0.8478792905807495, "learning_rate": 1.927823293051348e-05, "loss": 0.0261, "step": 6195 }, { "epoch": 0.42578339747113797, "grad_norm": 0.9344770908355713, "learning_rate": 1.9275032929144672e-05, "loss": 0.0274, "step": 6196 }, { "epoch": 0.4258521165475536, "grad_norm": 1.0791294574737549, "learning_rate": 1.9271832715999517e-05, "loss": 0.026, "step": 6197 }, { "epoch": 0.4259208356239692, "grad_norm": 1.0006039142608643, "learning_rate": 1.9268632291236553e-05, "loss": 0.0299, "step": 6198 }, { "epoch": 0.42598955470038485, "grad_norm": 0.8752599358558655, "learning_rate": 1.926543165501432e-05, "loss": 0.0288, "step": 6199 }, { "epoch": 0.42605827377680044, "grad_norm": 1.571373701095581, "learning_rate": 1.9262230807491374e-05, "loss": 0.0339, "step": 6200 }, { "epoch": 0.42612699285321604, "grad_norm": 0.8288040161132812, "learning_rate": 1.9259029748826273e-05, "loss": 0.028, "step": 6201 }, { "epoch": 0.4261957119296317, "grad_norm": 0.8588808178901672, "learning_rate": 1.92558284791776e-05, "loss": 0.0208, "step": 6202 }, { "epoch": 0.4262644310060473, "grad_norm": 0.8346999287605286, "learning_rate": 1.9252626998703928e-05, "loss": 0.0229, "step": 6203 }, { "epoch": 0.4263331500824629, "grad_norm": 0.8210436701774597, "learning_rate": 1.924942530756386e-05, "loss": 0.0264, "step": 6204 }, { "epoch": 0.4264018691588785, "grad_norm": 0.8788115382194519, "learning_rate": 1.9246223405915998e-05, "loss": 0.0241, "step": 6205 }, { "epoch": 0.4264705882352941, "grad_norm": 1.2125059366226196, "learning_rate": 1.924302129391896e-05, "loss": 0.0289, "step": 6206 }, { "epoch": 0.42653930731170975, "grad_norm": 1.37289297580719, "learning_rate": 1.923981897173137e-05, "loss": 0.0309, "step": 6207 }, { "epoch": 0.42660802638812534, "grad_norm": 0.8541125059127808, "learning_rate": 1.9236616439511866e-05, "loss": 0.0216, "step": 6208 }, { "epoch": 0.42667674546454093, "grad_norm": 0.6233548521995544, "learning_rate": 1.9233413697419092e-05, "loss": 0.0184, "step": 6209 }, { "epoch": 0.4267454645409566, "grad_norm": 0.68228679895401, "learning_rate": 1.9230210745611707e-05, "loss": 0.0227, "step": 6210 }, { "epoch": 0.42681418361737217, "grad_norm": 0.885030210018158, "learning_rate": 1.922700758424838e-05, "loss": 0.0211, "step": 6211 }, { "epoch": 0.4268829026937878, "grad_norm": 0.3831174075603485, "learning_rate": 1.9223804213487785e-05, "loss": 0.0258, "step": 6212 }, { "epoch": 0.4269516217702034, "grad_norm": 0.6486735939979553, "learning_rate": 1.9220600633488613e-05, "loss": 0.0143, "step": 6213 }, { "epoch": 0.427020340846619, "grad_norm": 0.5615929365158081, "learning_rate": 1.921739684440957e-05, "loss": 0.0164, "step": 6214 }, { "epoch": 0.42708905992303464, "grad_norm": 0.8014920949935913, "learning_rate": 1.921419284640935e-05, "loss": 0.0243, "step": 6215 }, { "epoch": 0.42715777899945023, "grad_norm": 1.031854510307312, "learning_rate": 1.921098863964668e-05, "loss": 0.023, "step": 6216 }, { "epoch": 0.4272264980758659, "grad_norm": 0.9329953193664551, "learning_rate": 1.92077842242803e-05, "loss": 0.0216, "step": 6217 }, { "epoch": 0.42729521715228147, "grad_norm": 0.8237271904945374, "learning_rate": 1.920457960046893e-05, "loss": 0.0162, "step": 6218 }, { "epoch": 0.42736393622869706, "grad_norm": 1.0385825634002686, "learning_rate": 1.9201374768371333e-05, "loss": 0.0216, "step": 6219 }, { "epoch": 0.4274326553051127, "grad_norm": 0.6327087879180908, "learning_rate": 1.919816972814627e-05, "loss": 0.0204, "step": 6220 }, { "epoch": 0.4275013743815283, "grad_norm": 0.4853817820549011, "learning_rate": 1.9194964479952507e-05, "loss": 0.0223, "step": 6221 }, { "epoch": 0.42757009345794394, "grad_norm": 1.02158784866333, "learning_rate": 1.919175902394883e-05, "loss": 0.0296, "step": 6222 }, { "epoch": 0.42763881253435954, "grad_norm": 1.534348487854004, "learning_rate": 1.918855336029403e-05, "loss": 0.0279, "step": 6223 }, { "epoch": 0.4277075316107751, "grad_norm": 0.6131641268730164, "learning_rate": 1.9185347489146907e-05, "loss": 0.022, "step": 6224 }, { "epoch": 0.4277762506871908, "grad_norm": 1.065893530845642, "learning_rate": 1.9182141410666268e-05, "loss": 0.0301, "step": 6225 }, { "epoch": 0.42784496976360636, "grad_norm": 1.260945200920105, "learning_rate": 1.917893512501095e-05, "loss": 0.0328, "step": 6226 }, { "epoch": 0.427913688840022, "grad_norm": 0.602712094783783, "learning_rate": 1.9175728632339775e-05, "loss": 0.021, "step": 6227 }, { "epoch": 0.4279824079164376, "grad_norm": 0.8876877427101135, "learning_rate": 1.9172521932811585e-05, "loss": 0.0247, "step": 6228 }, { "epoch": 0.4280511269928532, "grad_norm": 0.9499471783638, "learning_rate": 1.916931502658524e-05, "loss": 0.0248, "step": 6229 }, { "epoch": 0.42811984606926884, "grad_norm": 0.5565263628959656, "learning_rate": 1.9166107913819603e-05, "loss": 0.0155, "step": 6230 }, { "epoch": 0.42818856514568443, "grad_norm": 0.6523725986480713, "learning_rate": 1.9162900594673544e-05, "loss": 0.0224, "step": 6231 }, { "epoch": 0.4282572842221001, "grad_norm": 1.3371254205703735, "learning_rate": 1.9159693069305946e-05, "loss": 0.0261, "step": 6232 }, { "epoch": 0.42832600329851567, "grad_norm": 0.7290859222412109, "learning_rate": 1.9156485337875708e-05, "loss": 0.0187, "step": 6233 }, { "epoch": 0.42839472237493126, "grad_norm": 2.511787176132202, "learning_rate": 1.9153277400541727e-05, "loss": 0.0211, "step": 6234 }, { "epoch": 0.4284634414513469, "grad_norm": 1.0428128242492676, "learning_rate": 1.9150069257462925e-05, "loss": 0.0321, "step": 6235 }, { "epoch": 0.4285321605277625, "grad_norm": 0.7852502465248108, "learning_rate": 1.9146860908798222e-05, "loss": 0.0315, "step": 6236 }, { "epoch": 0.42860087960417814, "grad_norm": 1.52170729637146, "learning_rate": 1.914365235470656e-05, "loss": 0.0264, "step": 6237 }, { "epoch": 0.42866959868059373, "grad_norm": 0.634638249874115, "learning_rate": 1.9140443595346873e-05, "loss": 0.024, "step": 6238 }, { "epoch": 0.4287383177570093, "grad_norm": 1.600730299949646, "learning_rate": 1.913723463087813e-05, "loss": 0.0253, "step": 6239 }, { "epoch": 0.42880703683342497, "grad_norm": 0.7573390603065491, "learning_rate": 1.9134025461459282e-05, "loss": 0.0176, "step": 6240 }, { "epoch": 0.42887575590984056, "grad_norm": 1.1872388124465942, "learning_rate": 1.9130816087249317e-05, "loss": 0.0243, "step": 6241 }, { "epoch": 0.4289444749862562, "grad_norm": 0.5821501016616821, "learning_rate": 1.9127606508407213e-05, "loss": 0.0176, "step": 6242 }, { "epoch": 0.4290131940626718, "grad_norm": 0.4335006773471832, "learning_rate": 1.912439672509197e-05, "loss": 0.0303, "step": 6243 }, { "epoch": 0.4290819131390874, "grad_norm": 0.7459515929222107, "learning_rate": 1.912118673746259e-05, "loss": 0.0188, "step": 6244 }, { "epoch": 0.42915063221550304, "grad_norm": 0.6885083913803101, "learning_rate": 1.9117976545678096e-05, "loss": 0.0293, "step": 6245 }, { "epoch": 0.4292193512919186, "grad_norm": 0.6430541276931763, "learning_rate": 1.9114766149897505e-05, "loss": 0.0268, "step": 6246 }, { "epoch": 0.4292880703683343, "grad_norm": 1.046411156654358, "learning_rate": 1.911155555027986e-05, "loss": 0.0203, "step": 6247 }, { "epoch": 0.42935678944474986, "grad_norm": 0.6403985619544983, "learning_rate": 1.9108344746984212e-05, "loss": 0.0252, "step": 6248 }, { "epoch": 0.42942550852116546, "grad_norm": 1.0281596183776855, "learning_rate": 1.910513374016961e-05, "loss": 0.0359, "step": 6249 }, { "epoch": 0.4294942275975811, "grad_norm": 0.6252575516700745, "learning_rate": 1.9101922529995116e-05, "loss": 0.0249, "step": 6250 }, { "epoch": 0.4295629466739967, "grad_norm": 0.9341225028038025, "learning_rate": 1.909871111661982e-05, "loss": 0.0244, "step": 6251 }, { "epoch": 0.42963166575041234, "grad_norm": 0.6517488360404968, "learning_rate": 1.9095499500202796e-05, "loss": 0.0286, "step": 6252 }, { "epoch": 0.42970038482682793, "grad_norm": 0.6124836802482605, "learning_rate": 1.909228768090315e-05, "loss": 0.0252, "step": 6253 }, { "epoch": 0.4297691039032435, "grad_norm": 0.505976140499115, "learning_rate": 1.9089075658879985e-05, "loss": 0.0226, "step": 6254 }, { "epoch": 0.42983782297965917, "grad_norm": 0.9622061848640442, "learning_rate": 1.9085863434292415e-05, "loss": 0.0188, "step": 6255 }, { "epoch": 0.42990654205607476, "grad_norm": 0.6354252099990845, "learning_rate": 1.9082651007299576e-05, "loss": 0.0187, "step": 6256 }, { "epoch": 0.4299752611324904, "grad_norm": 0.38275769352912903, "learning_rate": 1.9079438378060596e-05, "loss": 0.0215, "step": 6257 }, { "epoch": 0.430043980208906, "grad_norm": 1.0034459829330444, "learning_rate": 1.9076225546734628e-05, "loss": 0.0362, "step": 6258 }, { "epoch": 0.4301126992853216, "grad_norm": 0.9676159620285034, "learning_rate": 1.9073012513480827e-05, "loss": 0.0277, "step": 6259 }, { "epoch": 0.43018141836173723, "grad_norm": 1.1671419143676758, "learning_rate": 1.9069799278458357e-05, "loss": 0.021, "step": 6260 }, { "epoch": 0.4302501374381528, "grad_norm": 0.6343769431114197, "learning_rate": 1.90665858418264e-05, "loss": 0.0318, "step": 6261 }, { "epoch": 0.43031885651456847, "grad_norm": 0.7628983855247498, "learning_rate": 1.906337220374414e-05, "loss": 0.019, "step": 6262 }, { "epoch": 0.43038757559098406, "grad_norm": 0.9304102659225464, "learning_rate": 1.9060158364370778e-05, "loss": 0.0293, "step": 6263 }, { "epoch": 0.43045629466739965, "grad_norm": 1.015259861946106, "learning_rate": 1.9056944323865515e-05, "loss": 0.0273, "step": 6264 }, { "epoch": 0.4305250137438153, "grad_norm": 0.6265506744384766, "learning_rate": 1.9053730082387573e-05, "loss": 0.0193, "step": 6265 }, { "epoch": 0.4305937328202309, "grad_norm": 2.387255907058716, "learning_rate": 1.9050515640096175e-05, "loss": 0.0376, "step": 6266 }, { "epoch": 0.43066245189664654, "grad_norm": 0.898520827293396, "learning_rate": 1.9047300997150563e-05, "loss": 0.0184, "step": 6267 }, { "epoch": 0.4307311709730621, "grad_norm": 0.5738205909729004, "learning_rate": 1.904408615370998e-05, "loss": 0.0158, "step": 6268 }, { "epoch": 0.4307998900494777, "grad_norm": 0.44418999552726746, "learning_rate": 1.904087110993368e-05, "loss": 0.0208, "step": 6269 }, { "epoch": 0.43086860912589336, "grad_norm": 0.9405352473258972, "learning_rate": 1.903765586598093e-05, "loss": 0.0387, "step": 6270 }, { "epoch": 0.43093732820230896, "grad_norm": 1.2138607501983643, "learning_rate": 1.9034440422011013e-05, "loss": 0.0203, "step": 6271 }, { "epoch": 0.43100604727872455, "grad_norm": 0.5061847567558289, "learning_rate": 1.9031224778183217e-05, "loss": 0.027, "step": 6272 }, { "epoch": 0.4310747663551402, "grad_norm": 0.6084296107292175, "learning_rate": 1.9028008934656825e-05, "loss": 0.0218, "step": 6273 }, { "epoch": 0.4311434854315558, "grad_norm": 0.7001031637191772, "learning_rate": 1.9024792891591155e-05, "loss": 0.0235, "step": 6274 }, { "epoch": 0.43121220450797143, "grad_norm": 0.6391514539718628, "learning_rate": 1.902157664914552e-05, "loss": 0.0173, "step": 6275 }, { "epoch": 0.431280923584387, "grad_norm": 1.195818543434143, "learning_rate": 1.9018360207479245e-05, "loss": 0.0332, "step": 6276 }, { "epoch": 0.4313496426608026, "grad_norm": 0.46053725481033325, "learning_rate": 1.9015143566751667e-05, "loss": 0.0192, "step": 6277 }, { "epoch": 0.43141836173721826, "grad_norm": 1.0012142658233643, "learning_rate": 1.9011926727122133e-05, "loss": 0.03, "step": 6278 }, { "epoch": 0.43148708081363385, "grad_norm": 0.6011003255844116, "learning_rate": 1.9008709688749995e-05, "loss": 0.0124, "step": 6279 }, { "epoch": 0.4315557998900495, "grad_norm": 0.4968447983264923, "learning_rate": 1.9005492451794622e-05, "loss": 0.0212, "step": 6280 }, { "epoch": 0.4316245189664651, "grad_norm": 1.085568904876709, "learning_rate": 1.900227501641539e-05, "loss": 0.0263, "step": 6281 }, { "epoch": 0.4316932380428807, "grad_norm": 0.8213596343994141, "learning_rate": 1.8999057382771676e-05, "loss": 0.0282, "step": 6282 }, { "epoch": 0.4317619571192963, "grad_norm": 1.4027029275894165, "learning_rate": 1.8995839551022886e-05, "loss": 0.025, "step": 6283 }, { "epoch": 0.4318306761957119, "grad_norm": 1.4022191762924194, "learning_rate": 1.899262152132842e-05, "loss": 0.0341, "step": 6284 }, { "epoch": 0.43189939527212756, "grad_norm": 1.0384600162506104, "learning_rate": 1.8989403293847693e-05, "loss": 0.0231, "step": 6285 }, { "epoch": 0.43196811434854315, "grad_norm": 0.9832013845443726, "learning_rate": 1.8986184868740126e-05, "loss": 0.0256, "step": 6286 }, { "epoch": 0.43203683342495874, "grad_norm": 0.9783108234405518, "learning_rate": 1.8982966246165157e-05, "loss": 0.0189, "step": 6287 }, { "epoch": 0.4321055525013744, "grad_norm": 0.296305775642395, "learning_rate": 1.8979747426282225e-05, "loss": 0.0189, "step": 6288 }, { "epoch": 0.43217427157779, "grad_norm": 0.9479925036430359, "learning_rate": 1.897652840925079e-05, "loss": 0.0192, "step": 6289 }, { "epoch": 0.4322429906542056, "grad_norm": 1.0562978982925415, "learning_rate": 1.8973309195230315e-05, "loss": 0.0375, "step": 6290 }, { "epoch": 0.4323117097306212, "grad_norm": 0.6974268555641174, "learning_rate": 1.897008978438027e-05, "loss": 0.0245, "step": 6291 }, { "epoch": 0.4323804288070368, "grad_norm": 1.1285532712936401, "learning_rate": 1.8966870176860138e-05, "loss": 0.0229, "step": 6292 }, { "epoch": 0.43244914788345246, "grad_norm": 1.157407283782959, "learning_rate": 1.8963650372829415e-05, "loss": 0.0248, "step": 6293 }, { "epoch": 0.43251786695986805, "grad_norm": 0.649918258190155, "learning_rate": 1.89604303724476e-05, "loss": 0.0476, "step": 6294 }, { "epoch": 0.4325865860362837, "grad_norm": 0.5083963871002197, "learning_rate": 1.895721017587421e-05, "loss": 0.0226, "step": 6295 }, { "epoch": 0.4326553051126993, "grad_norm": 0.8113201856613159, "learning_rate": 1.8953989783268756e-05, "loss": 0.0253, "step": 6296 }, { "epoch": 0.4327240241891149, "grad_norm": 0.8684561848640442, "learning_rate": 1.8950769194790785e-05, "loss": 0.0183, "step": 6297 }, { "epoch": 0.4327927432655305, "grad_norm": 1.1582225561141968, "learning_rate": 1.8947548410599823e-05, "loss": 0.0271, "step": 6298 }, { "epoch": 0.4328614623419461, "grad_norm": 0.563555121421814, "learning_rate": 1.8944327430855436e-05, "loss": 0.0287, "step": 6299 }, { "epoch": 0.43293018141836176, "grad_norm": 0.6667567491531372, "learning_rate": 1.894110625571717e-05, "loss": 0.0215, "step": 6300 }, { "epoch": 0.43299890049477735, "grad_norm": 0.8078162670135498, "learning_rate": 1.8937884885344613e-05, "loss": 0.0273, "step": 6301 }, { "epoch": 0.43306761957119294, "grad_norm": 0.35289236903190613, "learning_rate": 1.8934663319897326e-05, "loss": 0.0149, "step": 6302 }, { "epoch": 0.4331363386476086, "grad_norm": 0.7839809060096741, "learning_rate": 1.893144155953491e-05, "loss": 0.0333, "step": 6303 }, { "epoch": 0.4332050577240242, "grad_norm": 0.5570720434188843, "learning_rate": 1.8928219604416963e-05, "loss": 0.0327, "step": 6304 }, { "epoch": 0.4332737768004398, "grad_norm": 0.6502417325973511, "learning_rate": 1.8924997454703093e-05, "loss": 0.0266, "step": 6305 }, { "epoch": 0.4333424958768554, "grad_norm": 0.7244076132774353, "learning_rate": 1.8921775110552918e-05, "loss": 0.0348, "step": 6306 }, { "epoch": 0.433411214953271, "grad_norm": 0.5089518427848816, "learning_rate": 1.8918552572126064e-05, "loss": 0.0327, "step": 6307 }, { "epoch": 0.43347993402968665, "grad_norm": 0.8237556219100952, "learning_rate": 1.8915329839582175e-05, "loss": 0.0327, "step": 6308 }, { "epoch": 0.43354865310610224, "grad_norm": 0.7432233095169067, "learning_rate": 1.8912106913080895e-05, "loss": 0.024, "step": 6309 }, { "epoch": 0.4336173721825179, "grad_norm": 1.3036913871765137, "learning_rate": 1.890888379278188e-05, "loss": 0.0325, "step": 6310 }, { "epoch": 0.4336860912589335, "grad_norm": 1.0200783014297485, "learning_rate": 1.89056604788448e-05, "loss": 0.0269, "step": 6311 }, { "epoch": 0.4337548103353491, "grad_norm": 0.7360101938247681, "learning_rate": 1.8902436971429323e-05, "loss": 0.0192, "step": 6312 }, { "epoch": 0.4338235294117647, "grad_norm": 0.986848771572113, "learning_rate": 1.889921327069515e-05, "loss": 0.0401, "step": 6313 }, { "epoch": 0.4338922484881803, "grad_norm": 0.7038630247116089, "learning_rate": 1.8895989376801966e-05, "loss": 0.0201, "step": 6314 }, { "epoch": 0.43396096756459596, "grad_norm": 0.33250293135643005, "learning_rate": 1.8892765289909475e-05, "loss": 0.0138, "step": 6315 }, { "epoch": 0.43402968664101155, "grad_norm": 0.729556143283844, "learning_rate": 1.88895410101774e-05, "loss": 0.0234, "step": 6316 }, { "epoch": 0.43409840571742714, "grad_norm": 0.42839375138282776, "learning_rate": 1.888631653776546e-05, "loss": 0.0147, "step": 6317 }, { "epoch": 0.4341671247938428, "grad_norm": 1.283673882484436, "learning_rate": 1.888309187283338e-05, "loss": 0.012, "step": 6318 }, { "epoch": 0.4342358438702584, "grad_norm": 0.8758820295333862, "learning_rate": 1.887986701554092e-05, "loss": 0.0182, "step": 6319 }, { "epoch": 0.434304562946674, "grad_norm": 0.3183218240737915, "learning_rate": 1.887664196604782e-05, "loss": 0.0171, "step": 6320 }, { "epoch": 0.4343732820230896, "grad_norm": 1.8228784799575806, "learning_rate": 1.887341672451385e-05, "loss": 0.0225, "step": 6321 }, { "epoch": 0.4344420010995052, "grad_norm": 0.4126022756099701, "learning_rate": 1.8870191291098776e-05, "loss": 0.0161, "step": 6322 }, { "epoch": 0.43451072017592085, "grad_norm": 0.8535840511322021, "learning_rate": 1.8866965665962387e-05, "loss": 0.0213, "step": 6323 }, { "epoch": 0.43457943925233644, "grad_norm": 0.4595167338848114, "learning_rate": 1.8863739849264466e-05, "loss": 0.0279, "step": 6324 }, { "epoch": 0.4346481583287521, "grad_norm": 0.8947141766548157, "learning_rate": 1.886051384116482e-05, "loss": 0.0186, "step": 6325 }, { "epoch": 0.4347168774051677, "grad_norm": 0.4886534512042999, "learning_rate": 1.8857287641823252e-05, "loss": 0.022, "step": 6326 }, { "epoch": 0.43478559648158327, "grad_norm": 0.7382788062095642, "learning_rate": 1.885406125139958e-05, "loss": 0.0196, "step": 6327 }, { "epoch": 0.4348543155579989, "grad_norm": 1.4343867301940918, "learning_rate": 1.885083467005364e-05, "loss": 0.0257, "step": 6328 }, { "epoch": 0.4349230346344145, "grad_norm": 1.0478850603103638, "learning_rate": 1.884760789794527e-05, "loss": 0.0208, "step": 6329 }, { "epoch": 0.43499175371083015, "grad_norm": 0.9588938355445862, "learning_rate": 1.8844380935234313e-05, "loss": 0.0179, "step": 6330 }, { "epoch": 0.43506047278724574, "grad_norm": 1.356038212776184, "learning_rate": 1.8841153782080635e-05, "loss": 0.027, "step": 6331 }, { "epoch": 0.43512919186366134, "grad_norm": 0.9405540823936462, "learning_rate": 1.8837926438644088e-05, "loss": 0.0255, "step": 6332 }, { "epoch": 0.435197910940077, "grad_norm": 0.9648782014846802, "learning_rate": 1.8834698905084553e-05, "loss": 0.032, "step": 6333 }, { "epoch": 0.4352666300164926, "grad_norm": 0.6494265198707581, "learning_rate": 1.8831471181561927e-05, "loss": 0.0199, "step": 6334 }, { "epoch": 0.43533534909290816, "grad_norm": 1.3610771894454956, "learning_rate": 1.882824326823609e-05, "loss": 0.02, "step": 6335 }, { "epoch": 0.4354040681693238, "grad_norm": 0.43049561977386475, "learning_rate": 1.8825015165266954e-05, "loss": 0.024, "step": 6336 }, { "epoch": 0.4354727872457394, "grad_norm": 0.7876593470573425, "learning_rate": 1.882178687281443e-05, "loss": 0.0277, "step": 6337 }, { "epoch": 0.43554150632215505, "grad_norm": 1.6083271503448486, "learning_rate": 1.881855839103844e-05, "loss": 0.0223, "step": 6338 }, { "epoch": 0.43561022539857064, "grad_norm": 0.7155066728591919, "learning_rate": 1.881532972009892e-05, "loss": 0.0239, "step": 6339 }, { "epoch": 0.43567894447498623, "grad_norm": 0.46586090326309204, "learning_rate": 1.881210086015581e-05, "loss": 0.0191, "step": 6340 }, { "epoch": 0.4357476635514019, "grad_norm": 0.4425276219844818, "learning_rate": 1.880887181136906e-05, "loss": 0.017, "step": 6341 }, { "epoch": 0.43581638262781747, "grad_norm": 0.856704831123352, "learning_rate": 1.8805642573898632e-05, "loss": 0.0464, "step": 6342 }, { "epoch": 0.4358851017042331, "grad_norm": 0.3887409269809723, "learning_rate": 1.8802413147904496e-05, "loss": 0.0153, "step": 6343 }, { "epoch": 0.4359538207806487, "grad_norm": 0.683906614780426, "learning_rate": 1.8799183533546627e-05, "loss": 0.0317, "step": 6344 }, { "epoch": 0.4360225398570643, "grad_norm": 0.5494898557662964, "learning_rate": 1.8795953730985014e-05, "loss": 0.0286, "step": 6345 }, { "epoch": 0.43609125893347994, "grad_norm": 0.42932844161987305, "learning_rate": 1.8792723740379665e-05, "loss": 0.0256, "step": 6346 }, { "epoch": 0.43615997800989553, "grad_norm": 0.9402649402618408, "learning_rate": 1.8789493561890576e-05, "loss": 0.024, "step": 6347 }, { "epoch": 0.4362286970863112, "grad_norm": 0.706299901008606, "learning_rate": 1.8786263195677766e-05, "loss": 0.0256, "step": 6348 }, { "epoch": 0.43629741616272677, "grad_norm": 0.558531641960144, "learning_rate": 1.8783032641901263e-05, "loss": 0.0163, "step": 6349 }, { "epoch": 0.43636613523914236, "grad_norm": 0.47435101866722107, "learning_rate": 1.87798019007211e-05, "loss": 0.0146, "step": 6350 }, { "epoch": 0.436434854315558, "grad_norm": 1.8219021558761597, "learning_rate": 1.877657097229732e-05, "loss": 0.0262, "step": 6351 }, { "epoch": 0.4365035733919736, "grad_norm": 1.6582576036453247, "learning_rate": 1.877333985678998e-05, "loss": 0.0329, "step": 6352 }, { "epoch": 0.43657229246838924, "grad_norm": 0.8894039392471313, "learning_rate": 1.877010855435914e-05, "loss": 0.018, "step": 6353 }, { "epoch": 0.43664101154480484, "grad_norm": 0.7411383390426636, "learning_rate": 1.8766877065164872e-05, "loss": 0.0128, "step": 6354 }, { "epoch": 0.4367097306212204, "grad_norm": 0.3793613910675049, "learning_rate": 1.876364538936726e-05, "loss": 0.0118, "step": 6355 }, { "epoch": 0.4367784496976361, "grad_norm": 0.9200184941291809, "learning_rate": 1.8760413527126394e-05, "loss": 0.0257, "step": 6356 }, { "epoch": 0.43684716877405166, "grad_norm": 1.445559024810791, "learning_rate": 1.8757181478602372e-05, "loss": 0.0205, "step": 6357 }, { "epoch": 0.4369158878504673, "grad_norm": 0.7756450176239014, "learning_rate": 1.8753949243955305e-05, "loss": 0.0243, "step": 6358 }, { "epoch": 0.4369846069268829, "grad_norm": 0.8565733432769775, "learning_rate": 1.8750716823345312e-05, "loss": 0.0335, "step": 6359 }, { "epoch": 0.4370533260032985, "grad_norm": 0.8320087194442749, "learning_rate": 1.8747484216932516e-05, "loss": 0.0313, "step": 6360 }, { "epoch": 0.43712204507971414, "grad_norm": 0.9134103059768677, "learning_rate": 1.8744251424877058e-05, "loss": 0.0237, "step": 6361 }, { "epoch": 0.43719076415612973, "grad_norm": 0.8414084315299988, "learning_rate": 1.874101844733908e-05, "loss": 0.0313, "step": 6362 }, { "epoch": 0.4372594832325454, "grad_norm": 1.11915922164917, "learning_rate": 1.873778528447874e-05, "loss": 0.0362, "step": 6363 }, { "epoch": 0.43732820230896097, "grad_norm": 1.58137845993042, "learning_rate": 1.8734551936456205e-05, "loss": 0.035, "step": 6364 }, { "epoch": 0.43739692138537656, "grad_norm": 0.6602029800415039, "learning_rate": 1.873131840343164e-05, "loss": 0.0269, "step": 6365 }, { "epoch": 0.4374656404617922, "grad_norm": 1.738732099533081, "learning_rate": 1.8728084685565232e-05, "loss": 0.0358, "step": 6366 }, { "epoch": 0.4375343595382078, "grad_norm": 0.8202759623527527, "learning_rate": 1.872485078301718e-05, "loss": 0.0302, "step": 6367 }, { "epoch": 0.43760307861462344, "grad_norm": 0.337387353181839, "learning_rate": 1.872161669594767e-05, "loss": 0.0169, "step": 6368 }, { "epoch": 0.43767179769103903, "grad_norm": 1.1130801439285278, "learning_rate": 1.8718382424516926e-05, "loss": 0.0203, "step": 6369 }, { "epoch": 0.4377405167674546, "grad_norm": 0.6196688413619995, "learning_rate": 1.871514796888516e-05, "loss": 0.0158, "step": 6370 }, { "epoch": 0.43780923584387027, "grad_norm": 0.7386893033981323, "learning_rate": 1.87119133292126e-05, "loss": 0.0229, "step": 6371 }, { "epoch": 0.43787795492028586, "grad_norm": 1.8654628992080688, "learning_rate": 1.8708678505659485e-05, "loss": 0.0284, "step": 6372 }, { "epoch": 0.4379466739967015, "grad_norm": 1.0113052129745483, "learning_rate": 1.8705443498386065e-05, "loss": 0.0236, "step": 6373 }, { "epoch": 0.4380153930731171, "grad_norm": 0.8102357387542725, "learning_rate": 1.8702208307552588e-05, "loss": 0.0231, "step": 6374 }, { "epoch": 0.4380841121495327, "grad_norm": 1.169347882270813, "learning_rate": 1.8698972933319328e-05, "loss": 0.0217, "step": 6375 }, { "epoch": 0.43815283122594834, "grad_norm": 1.263976812362671, "learning_rate": 1.869573737584655e-05, "loss": 0.0315, "step": 6376 }, { "epoch": 0.4382215503023639, "grad_norm": 0.8023220300674438, "learning_rate": 1.869250163529454e-05, "loss": 0.022, "step": 6377 }, { "epoch": 0.4382902693787796, "grad_norm": 1.4188463687896729, "learning_rate": 1.868926571182359e-05, "loss": 0.0273, "step": 6378 }, { "epoch": 0.43835898845519516, "grad_norm": 0.8351694941520691, "learning_rate": 1.8686029605594008e-05, "loss": 0.0202, "step": 6379 }, { "epoch": 0.43842770753161076, "grad_norm": 0.36551371216773987, "learning_rate": 1.8682793316766095e-05, "loss": 0.0136, "step": 6380 }, { "epoch": 0.4384964266080264, "grad_norm": 0.386138379573822, "learning_rate": 1.867955684550017e-05, "loss": 0.0225, "step": 6381 }, { "epoch": 0.438565145684442, "grad_norm": 0.3637685775756836, "learning_rate": 1.867632019195656e-05, "loss": 0.0103, "step": 6382 }, { "epoch": 0.43863386476085764, "grad_norm": 0.9191734790802002, "learning_rate": 1.867308335629561e-05, "loss": 0.0176, "step": 6383 }, { "epoch": 0.43870258383727323, "grad_norm": 0.8287572264671326, "learning_rate": 1.866984633867766e-05, "loss": 0.021, "step": 6384 }, { "epoch": 0.4387713029136888, "grad_norm": 0.40931618213653564, "learning_rate": 1.866660913926307e-05, "loss": 0.0175, "step": 6385 }, { "epoch": 0.43884002199010447, "grad_norm": 0.7199483513832092, "learning_rate": 1.86633717582122e-05, "loss": 0.0234, "step": 6386 }, { "epoch": 0.43890874106652006, "grad_norm": 0.4006880521774292, "learning_rate": 1.8660134195685428e-05, "loss": 0.0241, "step": 6387 }, { "epoch": 0.4389774601429357, "grad_norm": 0.8667550683021545, "learning_rate": 1.8656896451843126e-05, "loss": 0.0218, "step": 6388 }, { "epoch": 0.4390461792193513, "grad_norm": 1.0468027591705322, "learning_rate": 1.8653658526845697e-05, "loss": 0.0315, "step": 6389 }, { "epoch": 0.4391148982957669, "grad_norm": 0.4173083007335663, "learning_rate": 1.8650420420853536e-05, "loss": 0.0176, "step": 6390 }, { "epoch": 0.43918361737218253, "grad_norm": 2.1614739894866943, "learning_rate": 1.8647182134027047e-05, "loss": 0.0274, "step": 6391 }, { "epoch": 0.4392523364485981, "grad_norm": 2.621506929397583, "learning_rate": 1.864394366652666e-05, "loss": 0.0224, "step": 6392 }, { "epoch": 0.43932105552501377, "grad_norm": 0.973652184009552, "learning_rate": 1.8640705018512796e-05, "loss": 0.0228, "step": 6393 }, { "epoch": 0.43938977460142936, "grad_norm": 1.2220863103866577, "learning_rate": 1.8637466190145882e-05, "loss": 0.032, "step": 6394 }, { "epoch": 0.43945849367784495, "grad_norm": 1.0572112798690796, "learning_rate": 1.863422718158638e-05, "loss": 0.0324, "step": 6395 }, { "epoch": 0.4395272127542606, "grad_norm": 0.47293996810913086, "learning_rate": 1.8630987992994735e-05, "loss": 0.0328, "step": 6396 }, { "epoch": 0.4395959318306762, "grad_norm": 1.1800715923309326, "learning_rate": 1.8627748624531408e-05, "loss": 0.0293, "step": 6397 }, { "epoch": 0.43966465090709184, "grad_norm": 0.7434636950492859, "learning_rate": 1.862450907635687e-05, "loss": 0.0223, "step": 6398 }, { "epoch": 0.4397333699835074, "grad_norm": 0.7948897480964661, "learning_rate": 1.8621269348631605e-05, "loss": 0.0342, "step": 6399 }, { "epoch": 0.439802089059923, "grad_norm": 1.1277682781219482, "learning_rate": 1.86180294415161e-05, "loss": 0.0407, "step": 6400 }, { "epoch": 0.43987080813633866, "grad_norm": 0.7905375361442566, "learning_rate": 1.8614789355170856e-05, "loss": 0.0329, "step": 6401 }, { "epoch": 0.43993952721275426, "grad_norm": 0.3950742483139038, "learning_rate": 1.8611549089756382e-05, "loss": 0.0196, "step": 6402 }, { "epoch": 0.44000824628916985, "grad_norm": 0.8035916686058044, "learning_rate": 1.8608308645433188e-05, "loss": 0.0169, "step": 6403 }, { "epoch": 0.4400769653655855, "grad_norm": 0.4561399519443512, "learning_rate": 1.8605068022361804e-05, "loss": 0.0265, "step": 6404 }, { "epoch": 0.4401456844420011, "grad_norm": 1.18844473361969, "learning_rate": 1.860182722070276e-05, "loss": 0.0163, "step": 6405 }, { "epoch": 0.44021440351841673, "grad_norm": 0.5847489833831787, "learning_rate": 1.85985862406166e-05, "loss": 0.0251, "step": 6406 }, { "epoch": 0.4402831225948323, "grad_norm": 0.5418718457221985, "learning_rate": 1.8595345082263875e-05, "loss": 0.0193, "step": 6407 }, { "epoch": 0.4403518416712479, "grad_norm": 0.787833034992218, "learning_rate": 1.859210374580515e-05, "loss": 0.0189, "step": 6408 }, { "epoch": 0.44042056074766356, "grad_norm": 0.2693469822406769, "learning_rate": 1.858886223140099e-05, "loss": 0.0187, "step": 6409 }, { "epoch": 0.44048927982407915, "grad_norm": 2.97839617729187, "learning_rate": 1.858562053921197e-05, "loss": 0.0197, "step": 6410 }, { "epoch": 0.4405579989004948, "grad_norm": 0.6603730916976929, "learning_rate": 1.8582378669398683e-05, "loss": 0.023, "step": 6411 }, { "epoch": 0.4406267179769104, "grad_norm": 0.9299554824829102, "learning_rate": 1.8579136622121723e-05, "loss": 0.0203, "step": 6412 }, { "epoch": 0.440695437053326, "grad_norm": 1.177376627922058, "learning_rate": 1.8575894397541693e-05, "loss": 0.0277, "step": 6413 }, { "epoch": 0.4407641561297416, "grad_norm": 1.0068105459213257, "learning_rate": 1.8572651995819206e-05, "loss": 0.0272, "step": 6414 }, { "epoch": 0.4408328752061572, "grad_norm": 0.8627030253410339, "learning_rate": 1.856940941711488e-05, "loss": 0.0174, "step": 6415 }, { "epoch": 0.44090159428257286, "grad_norm": 0.5317160487174988, "learning_rate": 1.856616666158935e-05, "loss": 0.0375, "step": 6416 }, { "epoch": 0.44097031335898845, "grad_norm": 0.7281713485717773, "learning_rate": 1.856292372940326e-05, "loss": 0.0169, "step": 6417 }, { "epoch": 0.44103903243540404, "grad_norm": 1.2496709823608398, "learning_rate": 1.8559680620717253e-05, "loss": 0.0258, "step": 6418 }, { "epoch": 0.4411077515118197, "grad_norm": 0.6624453067779541, "learning_rate": 1.8556437335691986e-05, "loss": 0.0158, "step": 6419 }, { "epoch": 0.4411764705882353, "grad_norm": 0.4833611249923706, "learning_rate": 1.8553193874488126e-05, "loss": 0.0269, "step": 6420 }, { "epoch": 0.4412451896646509, "grad_norm": 0.8859610557556152, "learning_rate": 1.8549950237266343e-05, "loss": 0.0155, "step": 6421 }, { "epoch": 0.4413139087410665, "grad_norm": 0.7592828273773193, "learning_rate": 1.8546706424187327e-05, "loss": 0.0252, "step": 6422 }, { "epoch": 0.4413826278174821, "grad_norm": 0.6151525378227234, "learning_rate": 1.8543462435411763e-05, "loss": 0.025, "step": 6423 }, { "epoch": 0.44145134689389776, "grad_norm": 1.1096633672714233, "learning_rate": 1.8540218271100357e-05, "loss": 0.0174, "step": 6424 }, { "epoch": 0.44152006597031335, "grad_norm": 0.4873568117618561, "learning_rate": 1.853697393141382e-05, "loss": 0.0178, "step": 6425 }, { "epoch": 0.441588785046729, "grad_norm": 0.6694704294204712, "learning_rate": 1.853372941651286e-05, "loss": 0.0355, "step": 6426 }, { "epoch": 0.4416575041231446, "grad_norm": 1.2064440250396729, "learning_rate": 1.853048472655821e-05, "loss": 0.0282, "step": 6427 }, { "epoch": 0.4417262231995602, "grad_norm": 0.6329725384712219, "learning_rate": 1.8527239861710606e-05, "loss": 0.0236, "step": 6428 }, { "epoch": 0.4417949422759758, "grad_norm": 0.6349951028823853, "learning_rate": 1.8523994822130796e-05, "loss": 0.0351, "step": 6429 }, { "epoch": 0.4418636613523914, "grad_norm": 0.6736570000648499, "learning_rate": 1.852074960797952e-05, "loss": 0.0219, "step": 6430 }, { "epoch": 0.44193238042880706, "grad_norm": 0.8975229263305664, "learning_rate": 1.851750421941755e-05, "loss": 0.0292, "step": 6431 }, { "epoch": 0.44200109950522265, "grad_norm": 0.7114226222038269, "learning_rate": 1.8514258656605655e-05, "loss": 0.0183, "step": 6432 }, { "epoch": 0.44206981858163824, "grad_norm": 1.1855647563934326, "learning_rate": 1.8511012919704605e-05, "loss": 0.0234, "step": 6433 }, { "epoch": 0.4421385376580539, "grad_norm": 0.4740943908691406, "learning_rate": 1.85077670088752e-05, "loss": 0.0261, "step": 6434 }, { "epoch": 0.4422072567344695, "grad_norm": 0.6155036091804504, "learning_rate": 1.850452092427823e-05, "loss": 0.0193, "step": 6435 }, { "epoch": 0.4422759758108851, "grad_norm": 0.6131226420402527, "learning_rate": 1.8501274666074493e-05, "loss": 0.0217, "step": 6436 }, { "epoch": 0.4423446948873007, "grad_norm": 0.4305677115917206, "learning_rate": 1.849802823442481e-05, "loss": 0.016, "step": 6437 }, { "epoch": 0.4424134139637163, "grad_norm": 0.4682389199733734, "learning_rate": 1.849478162949e-05, "loss": 0.0167, "step": 6438 }, { "epoch": 0.44248213304013195, "grad_norm": 0.7763852477073669, "learning_rate": 1.8491534851430892e-05, "loss": 0.0241, "step": 6439 }, { "epoch": 0.44255085211654754, "grad_norm": 0.6493790745735168, "learning_rate": 1.8488287900408327e-05, "loss": 0.0336, "step": 6440 }, { "epoch": 0.4426195711929632, "grad_norm": 0.9856130480766296, "learning_rate": 1.848504077658315e-05, "loss": 0.0318, "step": 6441 }, { "epoch": 0.4426882902693788, "grad_norm": 0.7257274389266968, "learning_rate": 1.8481793480116215e-05, "loss": 0.0167, "step": 6442 }, { "epoch": 0.4427570093457944, "grad_norm": 1.4081038236618042, "learning_rate": 1.8478546011168392e-05, "loss": 0.0239, "step": 6443 }, { "epoch": 0.44282572842221, "grad_norm": 0.7749977111816406, "learning_rate": 1.8475298369900552e-05, "loss": 0.0171, "step": 6444 }, { "epoch": 0.4428944474986256, "grad_norm": 0.953887939453125, "learning_rate": 1.847205055647357e-05, "loss": 0.0289, "step": 6445 }, { "epoch": 0.44296316657504126, "grad_norm": 1.4235767126083374, "learning_rate": 1.846880257104835e-05, "loss": 0.0262, "step": 6446 }, { "epoch": 0.44303188565145685, "grad_norm": 0.8843665719032288, "learning_rate": 1.846555441378578e-05, "loss": 0.025, "step": 6447 }, { "epoch": 0.44310060472787244, "grad_norm": 0.8060528635978699, "learning_rate": 1.8462306084846766e-05, "loss": 0.0239, "step": 6448 }, { "epoch": 0.4431693238042881, "grad_norm": 0.9349725842475891, "learning_rate": 1.845905758439223e-05, "loss": 0.0207, "step": 6449 }, { "epoch": 0.4432380428807037, "grad_norm": 0.8830312490463257, "learning_rate": 1.8455808912583095e-05, "loss": 0.0316, "step": 6450 }, { "epoch": 0.4433067619571193, "grad_norm": 0.727379560470581, "learning_rate": 1.8452560069580285e-05, "loss": 0.0309, "step": 6451 }, { "epoch": 0.4433754810335349, "grad_norm": 0.5704207420349121, "learning_rate": 1.8449311055544748e-05, "loss": 0.0191, "step": 6452 }, { "epoch": 0.4434442001099505, "grad_norm": 0.7560985684394836, "learning_rate": 1.8446061870637437e-05, "loss": 0.0335, "step": 6453 }, { "epoch": 0.44351291918636615, "grad_norm": 0.6074033975601196, "learning_rate": 1.8442812515019303e-05, "loss": 0.0174, "step": 6454 }, { "epoch": 0.44358163826278174, "grad_norm": 0.7912928462028503, "learning_rate": 1.843956298885132e-05, "loss": 0.0239, "step": 6455 }, { "epoch": 0.4436503573391974, "grad_norm": 0.4924822747707367, "learning_rate": 1.8436313292294452e-05, "loss": 0.0242, "step": 6456 }, { "epoch": 0.443719076415613, "grad_norm": 0.8552762269973755, "learning_rate": 1.8433063425509688e-05, "loss": 0.0164, "step": 6457 }, { "epoch": 0.44378779549202857, "grad_norm": 0.9442865252494812, "learning_rate": 1.8429813388658022e-05, "loss": 0.0297, "step": 6458 }, { "epoch": 0.4438565145684442, "grad_norm": 0.5267126560211182, "learning_rate": 1.8426563181900454e-05, "loss": 0.0285, "step": 6459 }, { "epoch": 0.4439252336448598, "grad_norm": 1.36226224899292, "learning_rate": 1.8423312805397988e-05, "loss": 0.023, "step": 6460 }, { "epoch": 0.44399395272127545, "grad_norm": 0.7929458618164062, "learning_rate": 1.842006225931164e-05, "loss": 0.019, "step": 6461 }, { "epoch": 0.44406267179769104, "grad_norm": 0.47549405694007874, "learning_rate": 1.841681154380245e-05, "loss": 0.0226, "step": 6462 }, { "epoch": 0.44413139087410664, "grad_norm": 0.3946009874343872, "learning_rate": 1.8413560659031433e-05, "loss": 0.0431, "step": 6463 }, { "epoch": 0.4442001099505223, "grad_norm": 1.2550570964813232, "learning_rate": 1.841030960515964e-05, "loss": 0.0203, "step": 6464 }, { "epoch": 0.4442688290269379, "grad_norm": 0.4613020122051239, "learning_rate": 1.840705838234812e-05, "loss": 0.0169, "step": 6465 }, { "epoch": 0.44433754810335346, "grad_norm": 0.5042986273765564, "learning_rate": 1.840380699075793e-05, "loss": 0.0298, "step": 6466 }, { "epoch": 0.4444062671797691, "grad_norm": 0.856343686580658, "learning_rate": 1.8400555430550145e-05, "loss": 0.0233, "step": 6467 }, { "epoch": 0.4444749862561847, "grad_norm": 0.43853527307510376, "learning_rate": 1.839730370188583e-05, "loss": 0.0165, "step": 6468 }, { "epoch": 0.44454370533260035, "grad_norm": 0.8902866244316101, "learning_rate": 1.839405180492608e-05, "loss": 0.0445, "step": 6469 }, { "epoch": 0.44461242440901594, "grad_norm": 1.9184458255767822, "learning_rate": 1.8390799739831977e-05, "loss": 0.0248, "step": 6470 }, { "epoch": 0.44468114348543153, "grad_norm": 0.896924614906311, "learning_rate": 1.838754750676463e-05, "loss": 0.0294, "step": 6471 }, { "epoch": 0.4447498625618472, "grad_norm": 3.5827033519744873, "learning_rate": 1.838429510588514e-05, "loss": 0.0373, "step": 6472 }, { "epoch": 0.44481858163826277, "grad_norm": 0.8974266648292542, "learning_rate": 1.8381042537354633e-05, "loss": 0.0262, "step": 6473 }, { "epoch": 0.4448873007146784, "grad_norm": 0.813218891620636, "learning_rate": 1.8377789801334228e-05, "loss": 0.0217, "step": 6474 }, { "epoch": 0.444956019791094, "grad_norm": 0.4823078215122223, "learning_rate": 1.8374536897985056e-05, "loss": 0.0263, "step": 6475 }, { "epoch": 0.4450247388675096, "grad_norm": 1.2274198532104492, "learning_rate": 1.837128382746827e-05, "loss": 0.0266, "step": 6476 }, { "epoch": 0.44509345794392524, "grad_norm": 0.8035070300102234, "learning_rate": 1.8368030589945014e-05, "loss": 0.0299, "step": 6477 }, { "epoch": 0.44516217702034083, "grad_norm": 0.5197204947471619, "learning_rate": 1.836477718557644e-05, "loss": 0.0196, "step": 6478 }, { "epoch": 0.4452308960967565, "grad_norm": 0.20511779189109802, "learning_rate": 1.8361523614523728e-05, "loss": 0.0224, "step": 6479 }, { "epoch": 0.44529961517317207, "grad_norm": 0.9280948638916016, "learning_rate": 1.8358269876948045e-05, "loss": 0.0217, "step": 6480 }, { "epoch": 0.44536833424958766, "grad_norm": 1.293497085571289, "learning_rate": 1.8355015973010576e-05, "loss": 0.0285, "step": 6481 }, { "epoch": 0.4454370533260033, "grad_norm": 0.7755078673362732, "learning_rate": 1.8351761902872513e-05, "loss": 0.0213, "step": 6482 }, { "epoch": 0.4455057724024189, "grad_norm": 0.6306232213973999, "learning_rate": 1.834850766669506e-05, "loss": 0.0244, "step": 6483 }, { "epoch": 0.44557449147883454, "grad_norm": 0.8215628862380981, "learning_rate": 1.8345253264639415e-05, "loss": 0.0309, "step": 6484 }, { "epoch": 0.44564321055525014, "grad_norm": 0.5407748222351074, "learning_rate": 1.834199869686681e-05, "loss": 0.0218, "step": 6485 }, { "epoch": 0.4457119296316657, "grad_norm": 0.9222963452339172, "learning_rate": 1.8338743963538452e-05, "loss": 0.0183, "step": 6486 }, { "epoch": 0.4457806487080814, "grad_norm": 0.5848008990287781, "learning_rate": 1.8335489064815585e-05, "loss": 0.0258, "step": 6487 }, { "epoch": 0.44584936778449696, "grad_norm": 0.8860177397727966, "learning_rate": 1.833223400085945e-05, "loss": 0.0319, "step": 6488 }, { "epoch": 0.4459180868609126, "grad_norm": 0.8099243640899658, "learning_rate": 1.8328978771831292e-05, "loss": 0.0272, "step": 6489 }, { "epoch": 0.4459868059373282, "grad_norm": 0.6836031675338745, "learning_rate": 1.8325723377892368e-05, "loss": 0.0192, "step": 6490 }, { "epoch": 0.4460555250137438, "grad_norm": 0.5286601185798645, "learning_rate": 1.8322467819203947e-05, "loss": 0.0341, "step": 6491 }, { "epoch": 0.44612424409015944, "grad_norm": 0.856949508190155, "learning_rate": 1.8319212095927303e-05, "loss": 0.0329, "step": 6492 }, { "epoch": 0.44619296316657503, "grad_norm": 0.5322127342224121, "learning_rate": 1.831595620822371e-05, "loss": 0.0173, "step": 6493 }, { "epoch": 0.4462616822429907, "grad_norm": 1.2985175848007202, "learning_rate": 1.8312700156254467e-05, "loss": 0.0306, "step": 6494 }, { "epoch": 0.44633040131940627, "grad_norm": 0.8821560740470886, "learning_rate": 1.8309443940180868e-05, "loss": 0.0226, "step": 6495 }, { "epoch": 0.44639912039582186, "grad_norm": 0.5848165154457092, "learning_rate": 1.830618756016422e-05, "loss": 0.02, "step": 6496 }, { "epoch": 0.4464678394722375, "grad_norm": 0.9926678538322449, "learning_rate": 1.8302931016365837e-05, "loss": 0.0498, "step": 6497 }, { "epoch": 0.4465365585486531, "grad_norm": 0.48025453090667725, "learning_rate": 1.8299674308947047e-05, "loss": 0.0238, "step": 6498 }, { "epoch": 0.44660527762506874, "grad_norm": 1.4308528900146484, "learning_rate": 1.8296417438069167e-05, "loss": 0.0318, "step": 6499 }, { "epoch": 0.44667399670148433, "grad_norm": 1.1766254901885986, "learning_rate": 1.8293160403893553e-05, "loss": 0.0196, "step": 6500 }, { "epoch": 0.4467427157778999, "grad_norm": 0.6684561371803284, "learning_rate": 1.8289903206581538e-05, "loss": 0.0248, "step": 6501 }, { "epoch": 0.44681143485431557, "grad_norm": 0.9260309934616089, "learning_rate": 1.828664584629448e-05, "loss": 0.0266, "step": 6502 }, { "epoch": 0.44688015393073116, "grad_norm": 0.7835502624511719, "learning_rate": 1.8283388323193748e-05, "loss": 0.0233, "step": 6503 }, { "epoch": 0.4469488730071468, "grad_norm": 0.7867885828018188, "learning_rate": 1.8280130637440706e-05, "loss": 0.0185, "step": 6504 }, { "epoch": 0.4470175920835624, "grad_norm": 0.5536759495735168, "learning_rate": 1.827687278919674e-05, "loss": 0.0218, "step": 6505 }, { "epoch": 0.447086311159978, "grad_norm": 0.9111251831054688, "learning_rate": 1.8273614778623228e-05, "loss": 0.0259, "step": 6506 }, { "epoch": 0.44715503023639364, "grad_norm": 0.3095279932022095, "learning_rate": 1.8270356605881573e-05, "loss": 0.0235, "step": 6507 }, { "epoch": 0.4472237493128092, "grad_norm": 0.501181960105896, "learning_rate": 1.8267098271133177e-05, "loss": 0.0252, "step": 6508 }, { "epoch": 0.4472924683892249, "grad_norm": 1.8492029905319214, "learning_rate": 1.826383977453945e-05, "loss": 0.0207, "step": 6509 }, { "epoch": 0.44736118746564046, "grad_norm": 0.39492642879486084, "learning_rate": 1.8260581116261805e-05, "loss": 0.0153, "step": 6510 }, { "epoch": 0.44742990654205606, "grad_norm": 0.4871584475040436, "learning_rate": 1.8257322296461676e-05, "loss": 0.0171, "step": 6511 }, { "epoch": 0.4474986256184717, "grad_norm": 0.9518209099769592, "learning_rate": 1.8254063315300504e-05, "loss": 0.017, "step": 6512 }, { "epoch": 0.4475673446948873, "grad_norm": 1.576124906539917, "learning_rate": 1.8250804172939722e-05, "loss": 0.0249, "step": 6513 }, { "epoch": 0.44763606377130294, "grad_norm": 0.5014211535453796, "learning_rate": 1.824754486954078e-05, "loss": 0.0204, "step": 6514 }, { "epoch": 0.44770478284771853, "grad_norm": 1.4290326833724976, "learning_rate": 1.8244285405265143e-05, "loss": 0.0392, "step": 6515 }, { "epoch": 0.4477735019241341, "grad_norm": 1.5491771697998047, "learning_rate": 1.824102578027428e-05, "loss": 0.0268, "step": 6516 }, { "epoch": 0.44784222100054977, "grad_norm": 1.0732274055480957, "learning_rate": 1.8237765994729663e-05, "loss": 0.0279, "step": 6517 }, { "epoch": 0.44791094007696536, "grad_norm": 0.5198305249214172, "learning_rate": 1.8234506048792773e-05, "loss": 0.0309, "step": 6518 }, { "epoch": 0.447979659153381, "grad_norm": 0.5379177331924438, "learning_rate": 1.8231245942625107e-05, "loss": 0.0157, "step": 6519 }, { "epoch": 0.4480483782297966, "grad_norm": 0.5356467962265015, "learning_rate": 1.8227985676388156e-05, "loss": 0.0204, "step": 6520 }, { "epoch": 0.4481170973062122, "grad_norm": 0.7264551520347595, "learning_rate": 1.8224725250243433e-05, "loss": 0.034, "step": 6521 }, { "epoch": 0.44818581638262783, "grad_norm": 0.5898008942604065, "learning_rate": 1.8221464664352453e-05, "loss": 0.0164, "step": 6522 }, { "epoch": 0.4482545354590434, "grad_norm": 1.2242189645767212, "learning_rate": 1.821820391887673e-05, "loss": 0.0352, "step": 6523 }, { "epoch": 0.44832325453545907, "grad_norm": 0.5633395314216614, "learning_rate": 1.8214943013977807e-05, "loss": 0.0185, "step": 6524 }, { "epoch": 0.44839197361187466, "grad_norm": 0.7802043557167053, "learning_rate": 1.8211681949817216e-05, "loss": 0.023, "step": 6525 }, { "epoch": 0.44846069268829025, "grad_norm": 0.7539632320404053, "learning_rate": 1.8208420726556508e-05, "loss": 0.0298, "step": 6526 }, { "epoch": 0.4485294117647059, "grad_norm": 0.7808369994163513, "learning_rate": 1.8205159344357224e-05, "loss": 0.0177, "step": 6527 }, { "epoch": 0.4485981308411215, "grad_norm": 0.878775954246521, "learning_rate": 1.820189780338094e-05, "loss": 0.0216, "step": 6528 }, { "epoch": 0.4486668499175371, "grad_norm": 0.28980475664138794, "learning_rate": 1.8198636103789225e-05, "loss": 0.0215, "step": 6529 }, { "epoch": 0.4487355689939527, "grad_norm": 1.6182676553726196, "learning_rate": 1.819537424574365e-05, "loss": 0.0333, "step": 6530 }, { "epoch": 0.4488042880703683, "grad_norm": 0.7831836938858032, "learning_rate": 1.8192112229405807e-05, "loss": 0.0209, "step": 6531 }, { "epoch": 0.44887300714678396, "grad_norm": 0.3481959104537964, "learning_rate": 1.8188850054937284e-05, "loss": 0.0367, "step": 6532 }, { "epoch": 0.44894172622319956, "grad_norm": 0.4957466125488281, "learning_rate": 1.818558772249969e-05, "loss": 0.0155, "step": 6533 }, { "epoch": 0.44901044529961515, "grad_norm": 0.4045248031616211, "learning_rate": 1.8182325232254624e-05, "loss": 0.0208, "step": 6534 }, { "epoch": 0.4490791643760308, "grad_norm": 1.0897296667099, "learning_rate": 1.8179062584363707e-05, "loss": 0.0208, "step": 6535 }, { "epoch": 0.4491478834524464, "grad_norm": 1.4073889255523682, "learning_rate": 1.8175799778988568e-05, "loss": 0.0231, "step": 6536 }, { "epoch": 0.44921660252886203, "grad_norm": 0.4108937680721283, "learning_rate": 1.817253681629084e-05, "loss": 0.0538, "step": 6537 }, { "epoch": 0.4492853216052776, "grad_norm": 0.6992839574813843, "learning_rate": 1.8169273696432155e-05, "loss": 0.0205, "step": 6538 }, { "epoch": 0.4493540406816932, "grad_norm": 0.5299254059791565, "learning_rate": 1.816601041957417e-05, "loss": 0.0297, "step": 6539 }, { "epoch": 0.44942275975810886, "grad_norm": 0.6726096272468567, "learning_rate": 1.816274698587854e-05, "loss": 0.0254, "step": 6540 }, { "epoch": 0.44949147883452445, "grad_norm": 0.31517115235328674, "learning_rate": 1.815948339550692e-05, "loss": 0.0255, "step": 6541 }, { "epoch": 0.4495601979109401, "grad_norm": 0.41786086559295654, "learning_rate": 1.8156219648620998e-05, "loss": 0.0274, "step": 6542 }, { "epoch": 0.4496289169873557, "grad_norm": 0.34042155742645264, "learning_rate": 1.8152955745382438e-05, "loss": 0.0247, "step": 6543 }, { "epoch": 0.4496976360637713, "grad_norm": 0.3218092620372772, "learning_rate": 1.814969168595293e-05, "loss": 0.0163, "step": 6544 }, { "epoch": 0.4497663551401869, "grad_norm": 0.6746065020561218, "learning_rate": 1.8146427470494175e-05, "loss": 0.0232, "step": 6545 }, { "epoch": 0.4498350742166025, "grad_norm": 0.6471610069274902, "learning_rate": 1.814316309916787e-05, "loss": 0.0226, "step": 6546 }, { "epoch": 0.44990379329301816, "grad_norm": 0.9406033754348755, "learning_rate": 1.8139898572135733e-05, "loss": 0.0322, "step": 6547 }, { "epoch": 0.44997251236943375, "grad_norm": 1.14833664894104, "learning_rate": 1.8136633889559472e-05, "loss": 0.0233, "step": 6548 }, { "epoch": 0.45004123144584934, "grad_norm": 0.5535572171211243, "learning_rate": 1.8133369051600817e-05, "loss": 0.0236, "step": 6549 }, { "epoch": 0.450109950522265, "grad_norm": 2.104950428009033, "learning_rate": 1.8130104058421498e-05, "loss": 0.025, "step": 6550 }, { "epoch": 0.4501786695986806, "grad_norm": 0.37624886631965637, "learning_rate": 1.8126838910183263e-05, "loss": 0.0162, "step": 6551 }, { "epoch": 0.4502473886750962, "grad_norm": 0.47917237877845764, "learning_rate": 1.8123573607047856e-05, "loss": 0.0225, "step": 6552 }, { "epoch": 0.4503161077515118, "grad_norm": 0.7239161729812622, "learning_rate": 1.8120308149177032e-05, "loss": 0.0237, "step": 6553 }, { "epoch": 0.4503848268279274, "grad_norm": 0.5199211835861206, "learning_rate": 1.811704253673256e-05, "loss": 0.0108, "step": 6554 }, { "epoch": 0.45045354590434306, "grad_norm": 0.8075836896896362, "learning_rate": 1.8113776769876208e-05, "loss": 0.0133, "step": 6555 }, { "epoch": 0.45052226498075865, "grad_norm": 0.8975247740745544, "learning_rate": 1.8110510848769754e-05, "loss": 0.0235, "step": 6556 }, { "epoch": 0.4505909840571743, "grad_norm": 0.37047460675239563, "learning_rate": 1.810724477357499e-05, "loss": 0.0161, "step": 6557 }, { "epoch": 0.4506597031335899, "grad_norm": 0.8399550318717957, "learning_rate": 1.810397854445371e-05, "loss": 0.027, "step": 6558 }, { "epoch": 0.4507284222100055, "grad_norm": 0.6290421485900879, "learning_rate": 1.810071216156771e-05, "loss": 0.0208, "step": 6559 }, { "epoch": 0.4507971412864211, "grad_norm": 0.5522711873054504, "learning_rate": 1.809744562507881e-05, "loss": 0.0297, "step": 6560 }, { "epoch": 0.4508658603628367, "grad_norm": 0.5061098337173462, "learning_rate": 1.8094178935148816e-05, "loss": 0.0239, "step": 6561 }, { "epoch": 0.45093457943925236, "grad_norm": 0.36248862743377686, "learning_rate": 1.809091209193956e-05, "loss": 0.0161, "step": 6562 }, { "epoch": 0.45100329851566795, "grad_norm": 0.4347671568393707, "learning_rate": 1.808764509561288e-05, "loss": 0.0214, "step": 6563 }, { "epoch": 0.45107201759208354, "grad_norm": 0.6793220639228821, "learning_rate": 1.8084377946330606e-05, "loss": 0.0271, "step": 6564 }, { "epoch": 0.4511407366684992, "grad_norm": 1.0191657543182373, "learning_rate": 1.808111064425459e-05, "loss": 0.0212, "step": 6565 }, { "epoch": 0.4512094557449148, "grad_norm": 0.3915962278842926, "learning_rate": 1.8077843189546684e-05, "loss": 0.0224, "step": 6566 }, { "epoch": 0.4512781748213304, "grad_norm": 0.5273099541664124, "learning_rate": 1.8074575582368758e-05, "loss": 0.015, "step": 6567 }, { "epoch": 0.451346893897746, "grad_norm": 0.8845546841621399, "learning_rate": 1.8071307822882676e-05, "loss": 0.0282, "step": 6568 }, { "epoch": 0.4514156129741616, "grad_norm": 0.6703990697860718, "learning_rate": 1.806803991125032e-05, "loss": 0.0215, "step": 6569 }, { "epoch": 0.45148433205057725, "grad_norm": 1.0338460206985474, "learning_rate": 1.8064771847633576e-05, "loss": 0.0161, "step": 6570 }, { "epoch": 0.45155305112699284, "grad_norm": 0.8072637319564819, "learning_rate": 1.806150363219433e-05, "loss": 0.0224, "step": 6571 }, { "epoch": 0.4516217702034085, "grad_norm": 1.1203705072402954, "learning_rate": 1.8058235265094494e-05, "loss": 0.0162, "step": 6572 }, { "epoch": 0.4516904892798241, "grad_norm": 2.0544748306274414, "learning_rate": 1.805496674649597e-05, "loss": 0.0293, "step": 6573 }, { "epoch": 0.45175920835623967, "grad_norm": 0.8558187484741211, "learning_rate": 1.8051698076560672e-05, "loss": 0.0273, "step": 6574 }, { "epoch": 0.4518279274326553, "grad_norm": 0.5169885158538818, "learning_rate": 1.8048429255450525e-05, "loss": 0.0219, "step": 6575 }, { "epoch": 0.4518966465090709, "grad_norm": 0.8237066864967346, "learning_rate": 1.804516028332746e-05, "loss": 0.0262, "step": 6576 }, { "epoch": 0.45196536558548656, "grad_norm": 0.6153923273086548, "learning_rate": 1.8041891160353416e-05, "loss": 0.0166, "step": 6577 }, { "epoch": 0.45203408466190215, "grad_norm": 1.1367042064666748, "learning_rate": 1.8038621886690336e-05, "loss": 0.0255, "step": 6578 }, { "epoch": 0.45210280373831774, "grad_norm": 0.6383638978004456, "learning_rate": 1.8035352462500178e-05, "loss": 0.0227, "step": 6579 }, { "epoch": 0.4521715228147334, "grad_norm": 0.9862568378448486, "learning_rate": 1.8032082887944897e-05, "loss": 0.0242, "step": 6580 }, { "epoch": 0.452240241891149, "grad_norm": 0.44229578971862793, "learning_rate": 1.8028813163186465e-05, "loss": 0.0108, "step": 6581 }, { "epoch": 0.4523089609675646, "grad_norm": 0.5735453367233276, "learning_rate": 1.8025543288386858e-05, "loss": 0.0132, "step": 6582 }, { "epoch": 0.4523776800439802, "grad_norm": 0.49101167917251587, "learning_rate": 1.8022273263708056e-05, "loss": 0.018, "step": 6583 }, { "epoch": 0.4524463991203958, "grad_norm": 3.095377206802368, "learning_rate": 1.8019003089312048e-05, "loss": 0.019, "step": 6584 }, { "epoch": 0.45251511819681145, "grad_norm": 2.11552357673645, "learning_rate": 1.8015732765360833e-05, "loss": 0.0341, "step": 6585 }, { "epoch": 0.45258383727322704, "grad_norm": 1.4463483095169067, "learning_rate": 1.8012462292016415e-05, "loss": 0.0277, "step": 6586 }, { "epoch": 0.4526525563496427, "grad_norm": 0.9416343569755554, "learning_rate": 1.8009191669440813e-05, "loss": 0.0289, "step": 6587 }, { "epoch": 0.4527212754260583, "grad_norm": 0.77587890625, "learning_rate": 1.8005920897796043e-05, "loss": 0.0234, "step": 6588 }, { "epoch": 0.45278999450247387, "grad_norm": 1.052924394607544, "learning_rate": 1.8002649977244127e-05, "loss": 0.0221, "step": 6589 }, { "epoch": 0.4528587135788895, "grad_norm": 0.6632734537124634, "learning_rate": 1.7999378907947107e-05, "loss": 0.0261, "step": 6590 }, { "epoch": 0.4529274326553051, "grad_norm": 0.5871516466140747, "learning_rate": 1.7996107690067023e-05, "loss": 0.0245, "step": 6591 }, { "epoch": 0.45299615173172075, "grad_norm": 1.1665953397750854, "learning_rate": 1.799283632376592e-05, "loss": 0.0209, "step": 6592 }, { "epoch": 0.45306487080813634, "grad_norm": 0.8405221104621887, "learning_rate": 1.798956480920586e-05, "loss": 0.0203, "step": 6593 }, { "epoch": 0.45313358988455193, "grad_norm": 0.66644287109375, "learning_rate": 1.798629314654891e-05, "loss": 0.02, "step": 6594 }, { "epoch": 0.4532023089609676, "grad_norm": 1.4586552381515503, "learning_rate": 1.7983021335957134e-05, "loss": 0.0229, "step": 6595 }, { "epoch": 0.4532710280373832, "grad_norm": 0.747450053691864, "learning_rate": 1.797974937759261e-05, "loss": 0.032, "step": 6596 }, { "epoch": 0.45333974711379876, "grad_norm": 1.5826876163482666, "learning_rate": 1.7976477271617435e-05, "loss": 0.0189, "step": 6597 }, { "epoch": 0.4534084661902144, "grad_norm": 0.43186575174331665, "learning_rate": 1.797320501819369e-05, "loss": 0.0188, "step": 6598 }, { "epoch": 0.45347718526663, "grad_norm": 0.9687561392784119, "learning_rate": 1.7969932617483488e-05, "loss": 0.0311, "step": 6599 }, { "epoch": 0.45354590434304565, "grad_norm": 0.7029834389686584, "learning_rate": 1.7966660069648923e-05, "loss": 0.02, "step": 6600 }, { "epoch": 0.45361462341946124, "grad_norm": 1.060142159461975, "learning_rate": 1.796338737485212e-05, "loss": 0.022, "step": 6601 }, { "epoch": 0.45368334249587683, "grad_norm": 0.8528428077697754, "learning_rate": 1.7960114533255202e-05, "loss": 0.0207, "step": 6602 }, { "epoch": 0.4537520615722925, "grad_norm": 0.7069180607795715, "learning_rate": 1.795684154502029e-05, "loss": 0.0289, "step": 6603 }, { "epoch": 0.45382078064870807, "grad_norm": 0.6041733026504517, "learning_rate": 1.7953568410309532e-05, "loss": 0.0141, "step": 6604 }, { "epoch": 0.4538894997251237, "grad_norm": 0.991614043712616, "learning_rate": 1.795029512928506e-05, "loss": 0.0228, "step": 6605 }, { "epoch": 0.4539582188015393, "grad_norm": 1.0206879377365112, "learning_rate": 1.794702170210904e-05, "loss": 0.0199, "step": 6606 }, { "epoch": 0.4540269378779549, "grad_norm": 0.8992913365364075, "learning_rate": 1.7943748128943623e-05, "loss": 0.024, "step": 6607 }, { "epoch": 0.45409565695437054, "grad_norm": 1.2742730379104614, "learning_rate": 1.794047440995097e-05, "loss": 0.019, "step": 6608 }, { "epoch": 0.45416437603078613, "grad_norm": 0.4177382290363312, "learning_rate": 1.7937200545293264e-05, "loss": 0.0151, "step": 6609 }, { "epoch": 0.4542330951072018, "grad_norm": 0.8203365206718445, "learning_rate": 1.793392653513268e-05, "loss": 0.0151, "step": 6610 }, { "epoch": 0.45430181418361737, "grad_norm": 1.2606550455093384, "learning_rate": 1.7930652379631408e-05, "loss": 0.0373, "step": 6611 }, { "epoch": 0.45437053326003296, "grad_norm": 0.9367243051528931, "learning_rate": 1.7927378078951643e-05, "loss": 0.0213, "step": 6612 }, { "epoch": 0.4544392523364486, "grad_norm": 0.7460215091705322, "learning_rate": 1.792410363325559e-05, "loss": 0.0208, "step": 6613 }, { "epoch": 0.4545079714128642, "grad_norm": 1.0028984546661377, "learning_rate": 1.792082904270545e-05, "loss": 0.0235, "step": 6614 }, { "epoch": 0.45457669048927984, "grad_norm": 0.8578044176101685, "learning_rate": 1.7917554307463446e-05, "loss": 0.0308, "step": 6615 }, { "epoch": 0.45464540956569544, "grad_norm": 0.9598782658576965, "learning_rate": 1.79142794276918e-05, "loss": 0.0316, "step": 6616 }, { "epoch": 0.454714128642111, "grad_norm": 1.292378306388855, "learning_rate": 1.791100440355274e-05, "loss": 0.0337, "step": 6617 }, { "epoch": 0.4547828477185267, "grad_norm": 0.7945721745491028, "learning_rate": 1.7907729235208504e-05, "loss": 0.0231, "step": 6618 }, { "epoch": 0.45485156679494226, "grad_norm": 0.5233510136604309, "learning_rate": 1.7904453922821348e-05, "loss": 0.043, "step": 6619 }, { "epoch": 0.4549202858713579, "grad_norm": 0.9740257859230042, "learning_rate": 1.7901178466553513e-05, "loss": 0.0226, "step": 6620 }, { "epoch": 0.4549890049477735, "grad_norm": 0.7720537781715393, "learning_rate": 1.789790286656726e-05, "loss": 0.0301, "step": 6621 }, { "epoch": 0.4550577240241891, "grad_norm": 0.4557822048664093, "learning_rate": 1.7894627123024854e-05, "loss": 0.0204, "step": 6622 }, { "epoch": 0.45512644310060474, "grad_norm": 1.3946319818496704, "learning_rate": 1.789135123608858e-05, "loss": 0.0333, "step": 6623 }, { "epoch": 0.45519516217702033, "grad_norm": 0.5489616394042969, "learning_rate": 1.7888075205920703e-05, "loss": 0.0213, "step": 6624 }, { "epoch": 0.455263881253436, "grad_norm": 0.4355847239494324, "learning_rate": 1.788479903268352e-05, "loss": 0.0188, "step": 6625 }, { "epoch": 0.45533260032985157, "grad_norm": 0.19351769983768463, "learning_rate": 1.7881522716539324e-05, "loss": 0.0141, "step": 6626 }, { "epoch": 0.45540131940626716, "grad_norm": 0.5960472822189331, "learning_rate": 1.7878246257650414e-05, "loss": 0.0262, "step": 6627 }, { "epoch": 0.4554700384826828, "grad_norm": 0.8177894949913025, "learning_rate": 1.7874969656179107e-05, "loss": 0.0286, "step": 6628 }, { "epoch": 0.4555387575590984, "grad_norm": 0.9310884475708008, "learning_rate": 1.7871692912287707e-05, "loss": 0.0241, "step": 6629 }, { "epoch": 0.45560747663551404, "grad_norm": 0.8243967294692993, "learning_rate": 1.786841602613855e-05, "loss": 0.0238, "step": 6630 }, { "epoch": 0.45567619571192963, "grad_norm": 0.3868914544582367, "learning_rate": 1.786513899789396e-05, "loss": 0.028, "step": 6631 }, { "epoch": 0.4557449147883452, "grad_norm": 0.6705635190010071, "learning_rate": 1.7861861827716272e-05, "loss": 0.0272, "step": 6632 }, { "epoch": 0.45581363386476087, "grad_norm": 0.7419412136077881, "learning_rate": 1.7858584515767833e-05, "loss": 0.0196, "step": 6633 }, { "epoch": 0.45588235294117646, "grad_norm": 0.704796314239502, "learning_rate": 1.7855307062210996e-05, "loss": 0.0212, "step": 6634 }, { "epoch": 0.4559510720175921, "grad_norm": 0.9339787364006042, "learning_rate": 1.7852029467208114e-05, "loss": 0.0226, "step": 6635 }, { "epoch": 0.4560197910940077, "grad_norm": 0.7071428894996643, "learning_rate": 1.7848751730921556e-05, "loss": 0.0245, "step": 6636 }, { "epoch": 0.4560885101704233, "grad_norm": 0.4612237513065338, "learning_rate": 1.78454738535137e-05, "loss": 0.0165, "step": 6637 }, { "epoch": 0.45615722924683894, "grad_norm": 0.9679598808288574, "learning_rate": 1.7842195835146913e-05, "loss": 0.0249, "step": 6638 }, { "epoch": 0.4562259483232545, "grad_norm": 0.753287136554718, "learning_rate": 1.783891767598359e-05, "loss": 0.0152, "step": 6639 }, { "epoch": 0.4562946673996702, "grad_norm": 0.5508970618247986, "learning_rate": 1.7835639376186125e-05, "loss": 0.0238, "step": 6640 }, { "epoch": 0.45636338647608576, "grad_norm": 0.9266013503074646, "learning_rate": 1.783236093591691e-05, "loss": 0.0204, "step": 6641 }, { "epoch": 0.45643210555250135, "grad_norm": 2.174379587173462, "learning_rate": 1.782908235533836e-05, "loss": 0.0273, "step": 6642 }, { "epoch": 0.456500824628917, "grad_norm": 0.7228773832321167, "learning_rate": 1.782580363461289e-05, "loss": 0.0251, "step": 6643 }, { "epoch": 0.4565695437053326, "grad_norm": 1.4674338102340698, "learning_rate": 1.7822524773902907e-05, "loss": 0.0342, "step": 6644 }, { "epoch": 0.45663826278174824, "grad_norm": 0.5230618715286255, "learning_rate": 1.781924577337086e-05, "loss": 0.015, "step": 6645 }, { "epoch": 0.45670698185816383, "grad_norm": 0.7018682360649109, "learning_rate": 1.7815966633179172e-05, "loss": 0.023, "step": 6646 }, { "epoch": 0.4567757009345794, "grad_norm": 1.0440378189086914, "learning_rate": 1.781268735349028e-05, "loss": 0.0193, "step": 6647 }, { "epoch": 0.45684442001099507, "grad_norm": 1.129747986793518, "learning_rate": 1.7809407934466644e-05, "loss": 0.0334, "step": 6648 }, { "epoch": 0.45691313908741066, "grad_norm": 0.8910743594169617, "learning_rate": 1.7806128376270716e-05, "loss": 0.0203, "step": 6649 }, { "epoch": 0.4569818581638263, "grad_norm": 0.5198029279708862, "learning_rate": 1.7802848679064948e-05, "loss": 0.0288, "step": 6650 }, { "epoch": 0.4570505772402419, "grad_norm": 0.5388579368591309, "learning_rate": 1.779956884301183e-05, "loss": 0.0227, "step": 6651 }, { "epoch": 0.4571192963166575, "grad_norm": 0.49971485137939453, "learning_rate": 1.7796288868273825e-05, "loss": 0.0227, "step": 6652 }, { "epoch": 0.45718801539307313, "grad_norm": 1.644372820854187, "learning_rate": 1.7793008755013415e-05, "loss": 0.0315, "step": 6653 }, { "epoch": 0.4572567344694887, "grad_norm": 0.635850191116333, "learning_rate": 1.7789728503393093e-05, "loss": 0.0265, "step": 6654 }, { "epoch": 0.45732545354590437, "grad_norm": 0.9740034937858582, "learning_rate": 1.778644811357536e-05, "loss": 0.0258, "step": 6655 }, { "epoch": 0.45739417262231996, "grad_norm": 0.642876148223877, "learning_rate": 1.7783167585722713e-05, "loss": 0.0182, "step": 6656 }, { "epoch": 0.45746289169873555, "grad_norm": 0.7252311110496521, "learning_rate": 1.777988691999767e-05, "loss": 0.0227, "step": 6657 }, { "epoch": 0.4575316107751512, "grad_norm": 0.6153791546821594, "learning_rate": 1.777660611656274e-05, "loss": 0.0322, "step": 6658 }, { "epoch": 0.4576003298515668, "grad_norm": 0.29056575894355774, "learning_rate": 1.7773325175580453e-05, "loss": 0.0174, "step": 6659 }, { "epoch": 0.4576690489279824, "grad_norm": 1.2891309261322021, "learning_rate": 1.7770044097213342e-05, "loss": 0.0202, "step": 6660 }, { "epoch": 0.457737768004398, "grad_norm": 0.7189381122589111, "learning_rate": 1.7766762881623937e-05, "loss": 0.0413, "step": 6661 }, { "epoch": 0.4578064870808136, "grad_norm": 0.9151663780212402, "learning_rate": 1.7763481528974788e-05, "loss": 0.0217, "step": 6662 }, { "epoch": 0.45787520615722926, "grad_norm": 0.33913448452949524, "learning_rate": 1.7760200039428445e-05, "loss": 0.0132, "step": 6663 }, { "epoch": 0.45794392523364486, "grad_norm": 0.6143766641616821, "learning_rate": 1.775691841314747e-05, "loss": 0.0264, "step": 6664 }, { "epoch": 0.45801264431006045, "grad_norm": 1.1698354482650757, "learning_rate": 1.7753636650294418e-05, "loss": 0.022, "step": 6665 }, { "epoch": 0.4580813633864761, "grad_norm": 0.5349668860435486, "learning_rate": 1.7750354751031876e-05, "loss": 0.0377, "step": 6666 }, { "epoch": 0.4581500824628917, "grad_norm": 0.5308952927589417, "learning_rate": 1.7747072715522407e-05, "loss": 0.0293, "step": 6667 }, { "epoch": 0.45821880153930733, "grad_norm": 0.8953619599342346, "learning_rate": 1.7743790543928604e-05, "loss": 0.031, "step": 6668 }, { "epoch": 0.4582875206157229, "grad_norm": 0.6212732791900635, "learning_rate": 1.774050823641306e-05, "loss": 0.0199, "step": 6669 }, { "epoch": 0.4583562396921385, "grad_norm": 0.5957461595535278, "learning_rate": 1.7737225793138374e-05, "loss": 0.0191, "step": 6670 }, { "epoch": 0.45842495876855416, "grad_norm": 0.6073588132858276, "learning_rate": 1.7733943214267146e-05, "loss": 0.0235, "step": 6671 }, { "epoch": 0.45849367784496975, "grad_norm": 1.049234390258789, "learning_rate": 1.7730660499961997e-05, "loss": 0.0336, "step": 6672 }, { "epoch": 0.4585623969213854, "grad_norm": 0.5958689451217651, "learning_rate": 1.772737765038554e-05, "loss": 0.022, "step": 6673 }, { "epoch": 0.458631115997801, "grad_norm": 0.5150824785232544, "learning_rate": 1.7724094665700397e-05, "loss": 0.0269, "step": 6674 }, { "epoch": 0.4586998350742166, "grad_norm": 1.1210500001907349, "learning_rate": 1.7720811546069207e-05, "loss": 0.0224, "step": 6675 }, { "epoch": 0.4587685541506322, "grad_norm": 0.40654149651527405, "learning_rate": 1.771752829165461e-05, "loss": 0.0203, "step": 6676 }, { "epoch": 0.4588372732270478, "grad_norm": 0.698847770690918, "learning_rate": 1.7714244902619245e-05, "loss": 0.0229, "step": 6677 }, { "epoch": 0.45890599230346346, "grad_norm": 0.46224716305732727, "learning_rate": 1.771096137912577e-05, "loss": 0.0215, "step": 6678 }, { "epoch": 0.45897471137987905, "grad_norm": 0.317083477973938, "learning_rate": 1.7707677721336843e-05, "loss": 0.0181, "step": 6679 }, { "epoch": 0.45904343045629464, "grad_norm": 1.4650232791900635, "learning_rate": 1.7704393929415125e-05, "loss": 0.0257, "step": 6680 }, { "epoch": 0.4591121495327103, "grad_norm": 1.0305763483047485, "learning_rate": 1.770111000352329e-05, "loss": 0.0221, "step": 6681 }, { "epoch": 0.4591808686091259, "grad_norm": 0.4601043462753296, "learning_rate": 1.7697825943824022e-05, "loss": 0.0194, "step": 6682 }, { "epoch": 0.4592495876855415, "grad_norm": 0.549048900604248, "learning_rate": 1.769454175048e-05, "loss": 0.0235, "step": 6683 }, { "epoch": 0.4593183067619571, "grad_norm": 0.6570833921432495, "learning_rate": 1.7691257423653924e-05, "loss": 0.022, "step": 6684 }, { "epoch": 0.4593870258383727, "grad_norm": 0.9100042581558228, "learning_rate": 1.7687972963508486e-05, "loss": 0.0219, "step": 6685 }, { "epoch": 0.45945574491478836, "grad_norm": 1.074553370475769, "learning_rate": 1.7684688370206393e-05, "loss": 0.0203, "step": 6686 }, { "epoch": 0.45952446399120395, "grad_norm": 0.735831081867218, "learning_rate": 1.768140364391036e-05, "loss": 0.0177, "step": 6687 }, { "epoch": 0.4595931830676196, "grad_norm": 1.2534562349319458, "learning_rate": 1.76781187847831e-05, "loss": 0.0199, "step": 6688 }, { "epoch": 0.4596619021440352, "grad_norm": 0.6803508996963501, "learning_rate": 1.767483379298734e-05, "loss": 0.0242, "step": 6689 }, { "epoch": 0.4597306212204508, "grad_norm": 0.6180207133293152, "learning_rate": 1.7671548668685816e-05, "loss": 0.0216, "step": 6690 }, { "epoch": 0.4597993402968664, "grad_norm": 0.6257312893867493, "learning_rate": 1.7668263412041265e-05, "loss": 0.019, "step": 6691 }, { "epoch": 0.459868059373282, "grad_norm": 0.7378368973731995, "learning_rate": 1.766497802321643e-05, "loss": 0.0244, "step": 6692 }, { "epoch": 0.45993677844969766, "grad_norm": 1.0646781921386719, "learning_rate": 1.7661692502374057e-05, "loss": 0.0398, "step": 6693 }, { "epoch": 0.46000549752611325, "grad_norm": 0.9637964367866516, "learning_rate": 1.7658406849676916e-05, "loss": 0.0228, "step": 6694 }, { "epoch": 0.46007421660252884, "grad_norm": 0.996583104133606, "learning_rate": 1.7655121065287765e-05, "loss": 0.0226, "step": 6695 }, { "epoch": 0.4601429356789445, "grad_norm": 0.4168757200241089, "learning_rate": 1.7651835149369376e-05, "loss": 0.0181, "step": 6696 }, { "epoch": 0.4602116547553601, "grad_norm": 1.5531409978866577, "learning_rate": 1.7648549102084522e-05, "loss": 0.0232, "step": 6697 }, { "epoch": 0.4602803738317757, "grad_norm": 0.973668098449707, "learning_rate": 1.764526292359599e-05, "loss": 0.0214, "step": 6698 }, { "epoch": 0.4603490929081913, "grad_norm": 0.7092171907424927, "learning_rate": 1.7641976614066573e-05, "loss": 0.0447, "step": 6699 }, { "epoch": 0.4604178119846069, "grad_norm": 0.814677894115448, "learning_rate": 1.7638690173659068e-05, "loss": 0.0253, "step": 6700 }, { "epoch": 0.46048653106102255, "grad_norm": 0.45580536127090454, "learning_rate": 1.7635403602536273e-05, "loss": 0.0251, "step": 6701 }, { "epoch": 0.46055525013743814, "grad_norm": 1.0624988079071045, "learning_rate": 1.7632116900861e-05, "loss": 0.0263, "step": 6702 }, { "epoch": 0.4606239692138538, "grad_norm": 0.9880602955818176, "learning_rate": 1.7628830068796072e-05, "loss": 0.0358, "step": 6703 }, { "epoch": 0.4606926882902694, "grad_norm": 0.6761732697486877, "learning_rate": 1.7625543106504303e-05, "loss": 0.0165, "step": 6704 }, { "epoch": 0.46076140736668497, "grad_norm": 0.7129055857658386, "learning_rate": 1.7622256014148532e-05, "loss": 0.0164, "step": 6705 }, { "epoch": 0.4608301264431006, "grad_norm": 0.9955632090568542, "learning_rate": 1.7618968791891588e-05, "loss": 0.0276, "step": 6706 }, { "epoch": 0.4608988455195162, "grad_norm": 0.7925066947937012, "learning_rate": 1.7615681439896308e-05, "loss": 0.0238, "step": 6707 }, { "epoch": 0.46096756459593186, "grad_norm": 0.5737472176551819, "learning_rate": 1.7612393958325553e-05, "loss": 0.0203, "step": 6708 }, { "epoch": 0.46103628367234745, "grad_norm": 0.7682396769523621, "learning_rate": 1.760910634734217e-05, "loss": 0.0257, "step": 6709 }, { "epoch": 0.46110500274876304, "grad_norm": 0.6877076029777527, "learning_rate": 1.760581860710902e-05, "loss": 0.0267, "step": 6710 }, { "epoch": 0.4611737218251787, "grad_norm": 0.793763279914856, "learning_rate": 1.7602530737788982e-05, "loss": 0.0243, "step": 6711 }, { "epoch": 0.4612424409015943, "grad_norm": 0.93027263879776, "learning_rate": 1.7599242739544918e-05, "loss": 0.03, "step": 6712 }, { "epoch": 0.4613111599780099, "grad_norm": 0.6829739212989807, "learning_rate": 1.7595954612539712e-05, "loss": 0.0272, "step": 6713 }, { "epoch": 0.4613798790544255, "grad_norm": 0.6228693127632141, "learning_rate": 1.759266635693625e-05, "loss": 0.0224, "step": 6714 }, { "epoch": 0.4614485981308411, "grad_norm": 0.5596975088119507, "learning_rate": 1.758937797289743e-05, "loss": 0.0236, "step": 6715 }, { "epoch": 0.46151731720725675, "grad_norm": 0.554506242275238, "learning_rate": 1.7586089460586143e-05, "loss": 0.0266, "step": 6716 }, { "epoch": 0.46158603628367234, "grad_norm": 0.7540240287780762, "learning_rate": 1.7582800820165307e-05, "loss": 0.0282, "step": 6717 }, { "epoch": 0.461654755360088, "grad_norm": 0.6226708889007568, "learning_rate": 1.7579512051797826e-05, "loss": 0.0232, "step": 6718 }, { "epoch": 0.4617234744365036, "grad_norm": 0.8072658181190491, "learning_rate": 1.757622315564662e-05, "loss": 0.0272, "step": 6719 }, { "epoch": 0.46179219351291917, "grad_norm": 2.393232822418213, "learning_rate": 1.7572934131874616e-05, "loss": 0.0355, "step": 6720 }, { "epoch": 0.4618609125893348, "grad_norm": 1.1840380430221558, "learning_rate": 1.7569644980644747e-05, "loss": 0.0296, "step": 6721 }, { "epoch": 0.4619296316657504, "grad_norm": 0.9767747521400452, "learning_rate": 1.7566355702119944e-05, "loss": 0.0308, "step": 6722 }, { "epoch": 0.46199835074216605, "grad_norm": 0.399423748254776, "learning_rate": 1.756306629646316e-05, "loss": 0.021, "step": 6723 }, { "epoch": 0.46206706981858164, "grad_norm": 0.5162421464920044, "learning_rate": 1.755977676383734e-05, "loss": 0.0219, "step": 6724 }, { "epoch": 0.46213578889499723, "grad_norm": 0.7472082376480103, "learning_rate": 1.755648710440544e-05, "loss": 0.0229, "step": 6725 }, { "epoch": 0.4622045079714129, "grad_norm": 0.30411621928215027, "learning_rate": 1.7553197318330426e-05, "loss": 0.0152, "step": 6726 }, { "epoch": 0.46227322704782847, "grad_norm": 0.5965403318405151, "learning_rate": 1.754990740577527e-05, "loss": 0.0229, "step": 6727 }, { "epoch": 0.46234194612424406, "grad_norm": 0.7987880706787109, "learning_rate": 1.7546617366902935e-05, "loss": 0.0288, "step": 6728 }, { "epoch": 0.4624106652006597, "grad_norm": 1.3082809448242188, "learning_rate": 1.7543327201876417e-05, "loss": 0.0307, "step": 6729 }, { "epoch": 0.4624793842770753, "grad_norm": 0.7623061537742615, "learning_rate": 1.7540036910858696e-05, "loss": 0.0316, "step": 6730 }, { "epoch": 0.46254810335349095, "grad_norm": 1.0578523874282837, "learning_rate": 1.7536746494012773e-05, "loss": 0.0245, "step": 6731 }, { "epoch": 0.46261682242990654, "grad_norm": 0.8499635457992554, "learning_rate": 1.753345595150164e-05, "loss": 0.0168, "step": 6732 }, { "epoch": 0.46268554150632213, "grad_norm": 0.5378043055534363, "learning_rate": 1.753016528348831e-05, "loss": 0.0203, "step": 6733 }, { "epoch": 0.4627542605827378, "grad_norm": 1.2079180479049683, "learning_rate": 1.752687449013579e-05, "loss": 0.0303, "step": 6734 }, { "epoch": 0.46282297965915337, "grad_norm": 0.4561229646205902, "learning_rate": 1.7523583571607106e-05, "loss": 0.0207, "step": 6735 }, { "epoch": 0.462891698735569, "grad_norm": 0.7861579060554504, "learning_rate": 1.752029252806528e-05, "loss": 0.0266, "step": 6736 }, { "epoch": 0.4629604178119846, "grad_norm": 0.9431173801422119, "learning_rate": 1.7517001359673345e-05, "loss": 0.0279, "step": 6737 }, { "epoch": 0.4630291368884002, "grad_norm": 0.5791063904762268, "learning_rate": 1.7513710066594332e-05, "loss": 0.0226, "step": 6738 }, { "epoch": 0.46309785596481584, "grad_norm": 0.9973318576812744, "learning_rate": 1.7510418648991294e-05, "loss": 0.0303, "step": 6739 }, { "epoch": 0.46316657504123143, "grad_norm": 1.4116182327270508, "learning_rate": 1.7507127107027273e-05, "loss": 0.0235, "step": 6740 }, { "epoch": 0.4632352941176471, "grad_norm": 1.0946173667907715, "learning_rate": 1.750383544086534e-05, "loss": 0.0331, "step": 6741 }, { "epoch": 0.46330401319406267, "grad_norm": 0.5013731718063354, "learning_rate": 1.750054365066854e-05, "loss": 0.0214, "step": 6742 }, { "epoch": 0.46337273227047826, "grad_norm": 0.6073450446128845, "learning_rate": 1.7497251736599945e-05, "loss": 0.0226, "step": 6743 }, { "epoch": 0.4634414513468939, "grad_norm": 1.1130146980285645, "learning_rate": 1.7493959698822636e-05, "loss": 0.0233, "step": 6744 }, { "epoch": 0.4635101704233095, "grad_norm": 1.1053714752197266, "learning_rate": 1.7490667537499696e-05, "loss": 0.0232, "step": 6745 }, { "epoch": 0.46357888949972514, "grad_norm": 1.9508559703826904, "learning_rate": 1.7487375252794198e-05, "loss": 0.0208, "step": 6746 }, { "epoch": 0.46364760857614074, "grad_norm": 0.48045864701271057, "learning_rate": 1.748408284486925e-05, "loss": 0.0169, "step": 6747 }, { "epoch": 0.4637163276525563, "grad_norm": 0.9701398611068726, "learning_rate": 1.7480790313887943e-05, "loss": 0.0198, "step": 6748 }, { "epoch": 0.463785046728972, "grad_norm": 1.1884788274765015, "learning_rate": 1.7477497660013383e-05, "loss": 0.0303, "step": 6749 }, { "epoch": 0.46385376580538756, "grad_norm": 0.4942006766796112, "learning_rate": 1.7474204883408683e-05, "loss": 0.0284, "step": 6750 }, { "epoch": 0.4639224848818032, "grad_norm": 1.3736428022384644, "learning_rate": 1.7470911984236963e-05, "loss": 0.03, "step": 6751 }, { "epoch": 0.4639912039582188, "grad_norm": 0.6086597442626953, "learning_rate": 1.746761896266134e-05, "loss": 0.0156, "step": 6752 }, { "epoch": 0.4640599230346344, "grad_norm": 0.4534319043159485, "learning_rate": 1.746432581884495e-05, "loss": 0.0223, "step": 6753 }, { "epoch": 0.46412864211105004, "grad_norm": 1.6631520986557007, "learning_rate": 1.7461032552950923e-05, "loss": 0.0523, "step": 6754 }, { "epoch": 0.46419736118746563, "grad_norm": 0.7842380404472351, "learning_rate": 1.7457739165142403e-05, "loss": 0.017, "step": 6755 }, { "epoch": 0.4642660802638813, "grad_norm": 0.6542050838470459, "learning_rate": 1.745444565558254e-05, "loss": 0.0177, "step": 6756 }, { "epoch": 0.46433479934029687, "grad_norm": 0.8465451002120972, "learning_rate": 1.745115202443448e-05, "loss": 0.0271, "step": 6757 }, { "epoch": 0.46440351841671246, "grad_norm": 0.3075413107872009, "learning_rate": 1.7447858271861392e-05, "loss": 0.0214, "step": 6758 }, { "epoch": 0.4644722374931281, "grad_norm": 0.5962474942207336, "learning_rate": 1.7444564398026437e-05, "loss": 0.0242, "step": 6759 }, { "epoch": 0.4645409565695437, "grad_norm": 1.442559838294983, "learning_rate": 1.7441270403092788e-05, "loss": 0.0248, "step": 6760 }, { "epoch": 0.46460967564595934, "grad_norm": 0.5689430236816406, "learning_rate": 1.743797628722362e-05, "loss": 0.0152, "step": 6761 }, { "epoch": 0.46467839472237493, "grad_norm": 0.31237727403640747, "learning_rate": 1.7434682050582123e-05, "loss": 0.0166, "step": 6762 }, { "epoch": 0.4647471137987905, "grad_norm": 0.7390015721321106, "learning_rate": 1.7431387693331482e-05, "loss": 0.0188, "step": 6763 }, { "epoch": 0.46481583287520617, "grad_norm": 1.182281732559204, "learning_rate": 1.7428093215634887e-05, "loss": 0.0237, "step": 6764 }, { "epoch": 0.46488455195162176, "grad_norm": 0.4150468707084656, "learning_rate": 1.7424798617655552e-05, "loss": 0.0167, "step": 6765 }, { "epoch": 0.4649532710280374, "grad_norm": 0.5568429231643677, "learning_rate": 1.742150389955668e-05, "loss": 0.0184, "step": 6766 }, { "epoch": 0.465021990104453, "grad_norm": 0.6893922686576843, "learning_rate": 1.7418209061501477e-05, "loss": 0.0253, "step": 6767 }, { "epoch": 0.4650907091808686, "grad_norm": 0.8535215854644775, "learning_rate": 1.7414914103653173e-05, "loss": 0.0203, "step": 6768 }, { "epoch": 0.46515942825728424, "grad_norm": 0.4966682195663452, "learning_rate": 1.741161902617499e-05, "loss": 0.0201, "step": 6769 }, { "epoch": 0.4652281473336998, "grad_norm": 0.7481660842895508, "learning_rate": 1.7408323829230158e-05, "loss": 0.0167, "step": 6770 }, { "epoch": 0.4652968664101155, "grad_norm": 0.588727593421936, "learning_rate": 1.7405028512981913e-05, "loss": 0.0199, "step": 6771 }, { "epoch": 0.46536558548653106, "grad_norm": 0.9756424427032471, "learning_rate": 1.7401733077593503e-05, "loss": 0.0175, "step": 6772 }, { "epoch": 0.46543430456294665, "grad_norm": 1.058455228805542, "learning_rate": 1.739843752322817e-05, "loss": 0.0267, "step": 6773 }, { "epoch": 0.4655030236393623, "grad_norm": 0.8374657034873962, "learning_rate": 1.7395141850049175e-05, "loss": 0.0207, "step": 6774 }, { "epoch": 0.4655717427157779, "grad_norm": 0.8138098120689392, "learning_rate": 1.739184605821978e-05, "loss": 0.0278, "step": 6775 }, { "epoch": 0.46564046179219354, "grad_norm": 0.3858719766139984, "learning_rate": 1.7388550147903243e-05, "loss": 0.0259, "step": 6776 }, { "epoch": 0.46570918086860913, "grad_norm": 0.4774536192417145, "learning_rate": 1.7385254119262847e-05, "loss": 0.0228, "step": 6777 }, { "epoch": 0.4657778999450247, "grad_norm": 0.6800973415374756, "learning_rate": 1.7381957972461864e-05, "loss": 0.0288, "step": 6778 }, { "epoch": 0.46584661902144037, "grad_norm": 0.5546126961708069, "learning_rate": 1.737866170766358e-05, "loss": 0.0251, "step": 6779 }, { "epoch": 0.46591533809785596, "grad_norm": 0.6051273941993713, "learning_rate": 1.7375365325031286e-05, "loss": 0.0172, "step": 6780 }, { "epoch": 0.4659840571742716, "grad_norm": 0.6596538424491882, "learning_rate": 1.7372068824728274e-05, "loss": 0.0271, "step": 6781 }, { "epoch": 0.4660527762506872, "grad_norm": 0.5360535979270935, "learning_rate": 1.7368772206917852e-05, "loss": 0.0192, "step": 6782 }, { "epoch": 0.4661214953271028, "grad_norm": 1.0616341829299927, "learning_rate": 1.7365475471763325e-05, "loss": 0.0333, "step": 6783 }, { "epoch": 0.46619021440351843, "grad_norm": 0.5337667465209961, "learning_rate": 1.736217861942801e-05, "loss": 0.0214, "step": 6784 }, { "epoch": 0.466258933479934, "grad_norm": 1.0024895668029785, "learning_rate": 1.7358881650075214e-05, "loss": 0.0325, "step": 6785 }, { "epoch": 0.46632765255634967, "grad_norm": 1.5430738925933838, "learning_rate": 1.7355584563868274e-05, "loss": 0.0222, "step": 6786 }, { "epoch": 0.46639637163276526, "grad_norm": 0.5435633659362793, "learning_rate": 1.7352287360970525e-05, "loss": 0.0225, "step": 6787 }, { "epoch": 0.46646509070918085, "grad_norm": 0.4537675976753235, "learning_rate": 1.7348990041545288e-05, "loss": 0.0177, "step": 6788 }, { "epoch": 0.4665338097855965, "grad_norm": 0.7543362379074097, "learning_rate": 1.7345692605755918e-05, "loss": 0.0302, "step": 6789 }, { "epoch": 0.4666025288620121, "grad_norm": 0.3718132972717285, "learning_rate": 1.7342395053765757e-05, "loss": 0.021, "step": 6790 }, { "epoch": 0.4666712479384277, "grad_norm": 0.7487651705741882, "learning_rate": 1.733909738573816e-05, "loss": 0.0313, "step": 6791 }, { "epoch": 0.4667399670148433, "grad_norm": 0.8211844563484192, "learning_rate": 1.7335799601836487e-05, "loss": 0.0285, "step": 6792 }, { "epoch": 0.4668086860912589, "grad_norm": 0.4227069914340973, "learning_rate": 1.7332501702224106e-05, "loss": 0.0183, "step": 6793 }, { "epoch": 0.46687740516767456, "grad_norm": 0.6764168739318848, "learning_rate": 1.7329203687064383e-05, "loss": 0.0231, "step": 6794 }, { "epoch": 0.46694612424409015, "grad_norm": 0.7889308929443359, "learning_rate": 1.73259055565207e-05, "loss": 0.02, "step": 6795 }, { "epoch": 0.46701484332050575, "grad_norm": 0.6411511898040771, "learning_rate": 1.7322607310756437e-05, "loss": 0.0227, "step": 6796 }, { "epoch": 0.4670835623969214, "grad_norm": 0.9416409134864807, "learning_rate": 1.7319308949934983e-05, "loss": 0.0264, "step": 6797 }, { "epoch": 0.467152281473337, "grad_norm": 0.1733577698469162, "learning_rate": 1.7316010474219733e-05, "loss": 0.0149, "step": 6798 }, { "epoch": 0.46722100054975263, "grad_norm": 0.4921095073223114, "learning_rate": 1.7312711883774087e-05, "loss": 0.0211, "step": 6799 }, { "epoch": 0.4672897196261682, "grad_norm": 0.9530078172683716, "learning_rate": 1.7309413178761444e-05, "loss": 0.0208, "step": 6800 }, { "epoch": 0.4673584387025838, "grad_norm": 0.6839990019798279, "learning_rate": 1.7306114359345226e-05, "loss": 0.0152, "step": 6801 }, { "epoch": 0.46742715777899946, "grad_norm": 1.4388530254364014, "learning_rate": 1.7302815425688843e-05, "loss": 0.0276, "step": 6802 }, { "epoch": 0.46749587685541505, "grad_norm": 0.4246194660663605, "learning_rate": 1.7299516377955713e-05, "loss": 0.0261, "step": 6803 }, { "epoch": 0.4675645959318307, "grad_norm": 0.7094460129737854, "learning_rate": 1.7296217216309272e-05, "loss": 0.0132, "step": 6804 }, { "epoch": 0.4676333150082463, "grad_norm": 0.6276552677154541, "learning_rate": 1.729291794091295e-05, "loss": 0.0201, "step": 6805 }, { "epoch": 0.4677020340846619, "grad_norm": 0.5982655882835388, "learning_rate": 1.728961855193019e-05, "loss": 0.0178, "step": 6806 }, { "epoch": 0.4677707531610775, "grad_norm": 1.8448472023010254, "learning_rate": 1.7286319049524434e-05, "loss": 0.0326, "step": 6807 }, { "epoch": 0.4678394722374931, "grad_norm": 0.2703050673007965, "learning_rate": 1.7283019433859135e-05, "loss": 0.019, "step": 6808 }, { "epoch": 0.46790819131390876, "grad_norm": 0.6231561303138733, "learning_rate": 1.7279719705097742e-05, "loss": 0.0319, "step": 6809 }, { "epoch": 0.46797691039032435, "grad_norm": 0.4820935130119324, "learning_rate": 1.7276419863403724e-05, "loss": 0.02, "step": 6810 }, { "epoch": 0.46804562946673994, "grad_norm": 0.9847260117530823, "learning_rate": 1.7273119908940547e-05, "loss": 0.018, "step": 6811 }, { "epoch": 0.4681143485431556, "grad_norm": 0.6302261352539062, "learning_rate": 1.7269819841871678e-05, "loss": 0.0251, "step": 6812 }, { "epoch": 0.4681830676195712, "grad_norm": 0.6134756207466125, "learning_rate": 1.7266519662360605e-05, "loss": 0.0196, "step": 6813 }, { "epoch": 0.4682517866959868, "grad_norm": 0.7924410700798035, "learning_rate": 1.7263219370570805e-05, "loss": 0.0145, "step": 6814 }, { "epoch": 0.4683205057724024, "grad_norm": 0.8944349884986877, "learning_rate": 1.7259918966665773e-05, "loss": 0.0221, "step": 6815 }, { "epoch": 0.468389224848818, "grad_norm": 0.6755173206329346, "learning_rate": 1.7256618450809e-05, "loss": 0.0181, "step": 6816 }, { "epoch": 0.46845794392523366, "grad_norm": 1.2135109901428223, "learning_rate": 1.725331782316399e-05, "loss": 0.0285, "step": 6817 }, { "epoch": 0.46852666300164925, "grad_norm": 1.2775291204452515, "learning_rate": 1.7250017083894245e-05, "loss": 0.0244, "step": 6818 }, { "epoch": 0.4685953820780649, "grad_norm": 1.0568002462387085, "learning_rate": 1.724671623316328e-05, "loss": 0.0281, "step": 6819 }, { "epoch": 0.4686641011544805, "grad_norm": 0.8833407759666443, "learning_rate": 1.7243415271134616e-05, "loss": 0.025, "step": 6820 }, { "epoch": 0.4687328202308961, "grad_norm": 0.5207617878913879, "learning_rate": 1.7240114197971765e-05, "loss": 0.0184, "step": 6821 }, { "epoch": 0.4688015393073117, "grad_norm": 0.7623610496520996, "learning_rate": 1.7236813013838265e-05, "loss": 0.0281, "step": 6822 }, { "epoch": 0.4688702583837273, "grad_norm": 0.6077499985694885, "learning_rate": 1.723351171889765e-05, "loss": 0.0211, "step": 6823 }, { "epoch": 0.46893897746014296, "grad_norm": 0.908328652381897, "learning_rate": 1.7230210313313456e-05, "loss": 0.0173, "step": 6824 }, { "epoch": 0.46900769653655855, "grad_norm": 0.5729005932807922, "learning_rate": 1.7226908797249227e-05, "loss": 0.0216, "step": 6825 }, { "epoch": 0.46907641561297414, "grad_norm": 0.7143950462341309, "learning_rate": 1.7223607170868516e-05, "loss": 0.017, "step": 6826 }, { "epoch": 0.4691451346893898, "grad_norm": 1.0187034606933594, "learning_rate": 1.7220305434334875e-05, "loss": 0.0243, "step": 6827 }, { "epoch": 0.4692138537658054, "grad_norm": 0.6370376944541931, "learning_rate": 1.7217003587811866e-05, "loss": 0.0188, "step": 6828 }, { "epoch": 0.469282572842221, "grad_norm": 0.7745187878608704, "learning_rate": 1.721370163146306e-05, "loss": 0.0319, "step": 6829 }, { "epoch": 0.4693512919186366, "grad_norm": 1.0036050081253052, "learning_rate": 1.7210399565452024e-05, "loss": 0.0212, "step": 6830 }, { "epoch": 0.4694200109950522, "grad_norm": 0.2441488355398178, "learning_rate": 1.720709738994234e-05, "loss": 0.0197, "step": 6831 }, { "epoch": 0.46948873007146785, "grad_norm": 1.0741666555404663, "learning_rate": 1.7203795105097587e-05, "loss": 0.0271, "step": 6832 }, { "epoch": 0.46955744914788344, "grad_norm": 0.41988569498062134, "learning_rate": 1.7200492711081358e-05, "loss": 0.0178, "step": 6833 }, { "epoch": 0.4696261682242991, "grad_norm": 0.9572479724884033, "learning_rate": 1.719719020805724e-05, "loss": 0.0176, "step": 6834 }, { "epoch": 0.4696948873007147, "grad_norm": 0.35536855459213257, "learning_rate": 1.719388759618884e-05, "loss": 0.0328, "step": 6835 }, { "epoch": 0.46976360637713027, "grad_norm": 0.9980366230010986, "learning_rate": 1.7190584875639758e-05, "loss": 0.0255, "step": 6836 }, { "epoch": 0.4698323254535459, "grad_norm": 0.663776695728302, "learning_rate": 1.7187282046573603e-05, "loss": 0.0238, "step": 6837 }, { "epoch": 0.4699010445299615, "grad_norm": 1.0269968509674072, "learning_rate": 1.7183979109153998e-05, "loss": 0.0373, "step": 6838 }, { "epoch": 0.46996976360637716, "grad_norm": 0.7470809817314148, "learning_rate": 1.7180676063544555e-05, "loss": 0.02, "step": 6839 }, { "epoch": 0.47003848268279275, "grad_norm": 0.7796132564544678, "learning_rate": 1.7177372909908897e-05, "loss": 0.0225, "step": 6840 }, { "epoch": 0.47010720175920834, "grad_norm": 0.7154237627983093, "learning_rate": 1.7174069648410667e-05, "loss": 0.0268, "step": 6841 }, { "epoch": 0.470175920835624, "grad_norm": 0.6687478423118591, "learning_rate": 1.7170766279213497e-05, "loss": 0.023, "step": 6842 }, { "epoch": 0.4702446399120396, "grad_norm": 0.41248229146003723, "learning_rate": 1.716746280248103e-05, "loss": 0.0198, "step": 6843 }, { "epoch": 0.4703133589884552, "grad_norm": 1.0489709377288818, "learning_rate": 1.7164159218376903e-05, "loss": 0.0203, "step": 6844 }, { "epoch": 0.4703820780648708, "grad_norm": 0.43829864263534546, "learning_rate": 1.7160855527064788e-05, "loss": 0.023, "step": 6845 }, { "epoch": 0.4704507971412864, "grad_norm": 1.1778182983398438, "learning_rate": 1.715755172870833e-05, "loss": 0.0349, "step": 6846 }, { "epoch": 0.47051951621770205, "grad_norm": 0.9799864292144775, "learning_rate": 1.7154247823471195e-05, "loss": 0.0268, "step": 6847 }, { "epoch": 0.47058823529411764, "grad_norm": 0.7502113580703735, "learning_rate": 1.715094381151705e-05, "loss": 0.0215, "step": 6848 }, { "epoch": 0.4706569543705333, "grad_norm": 1.0252017974853516, "learning_rate": 1.7147639693009565e-05, "loss": 0.0325, "step": 6849 }, { "epoch": 0.4707256734469489, "grad_norm": 0.3127047121524811, "learning_rate": 1.7144335468112437e-05, "loss": 0.0136, "step": 6850 }, { "epoch": 0.47079439252336447, "grad_norm": 1.0242072343826294, "learning_rate": 1.7141031136989333e-05, "loss": 0.0339, "step": 6851 }, { "epoch": 0.4708631115997801, "grad_norm": 0.6590337753295898, "learning_rate": 1.7137726699803945e-05, "loss": 0.0222, "step": 6852 }, { "epoch": 0.4709318306761957, "grad_norm": 0.6218076944351196, "learning_rate": 1.7134422156719974e-05, "loss": 0.0116, "step": 6853 }, { "epoch": 0.4710005497526113, "grad_norm": 0.8155596256256104, "learning_rate": 1.7131117507901117e-05, "loss": 0.0276, "step": 6854 }, { "epoch": 0.47106926882902694, "grad_norm": 0.7234185338020325, "learning_rate": 1.7127812753511082e-05, "loss": 0.0111, "step": 6855 }, { "epoch": 0.47113798790544253, "grad_norm": 0.3595452308654785, "learning_rate": 1.7124507893713574e-05, "loss": 0.0262, "step": 6856 }, { "epoch": 0.4712067069818582, "grad_norm": 0.5000358819961548, "learning_rate": 1.712120292867232e-05, "loss": 0.021, "step": 6857 }, { "epoch": 0.47127542605827377, "grad_norm": 1.0542945861816406, "learning_rate": 1.7117897858551026e-05, "loss": 0.0244, "step": 6858 }, { "epoch": 0.47134414513468936, "grad_norm": 1.58937406539917, "learning_rate": 1.7114592683513432e-05, "loss": 0.0297, "step": 6859 }, { "epoch": 0.471412864211105, "grad_norm": 0.6525651812553406, "learning_rate": 1.7111287403723263e-05, "loss": 0.0155, "step": 6860 }, { "epoch": 0.4714815832875206, "grad_norm": 0.6637088656425476, "learning_rate": 1.710798201934425e-05, "loss": 0.0233, "step": 6861 }, { "epoch": 0.47155030236393625, "grad_norm": 1.320192813873291, "learning_rate": 1.7104676530540147e-05, "loss": 0.0351, "step": 6862 }, { "epoch": 0.47161902144035184, "grad_norm": 0.5732975006103516, "learning_rate": 1.7101370937474697e-05, "loss": 0.0167, "step": 6863 }, { "epoch": 0.47168774051676743, "grad_norm": 0.4491780698299408, "learning_rate": 1.709806524031165e-05, "loss": 0.026, "step": 6864 }, { "epoch": 0.4717564595931831, "grad_norm": 1.0290528535842896, "learning_rate": 1.7094759439214765e-05, "loss": 0.0217, "step": 6865 }, { "epoch": 0.47182517866959867, "grad_norm": 0.6852663159370422, "learning_rate": 1.7091453534347803e-05, "loss": 0.02, "step": 6866 }, { "epoch": 0.4718938977460143, "grad_norm": 0.8383033871650696, "learning_rate": 1.708814752587453e-05, "loss": 0.0355, "step": 6867 }, { "epoch": 0.4719626168224299, "grad_norm": 0.5755245685577393, "learning_rate": 1.7084841413958725e-05, "loss": 0.0139, "step": 6868 }, { "epoch": 0.4720313358988455, "grad_norm": 1.0343819856643677, "learning_rate": 1.7081535198764163e-05, "loss": 0.0221, "step": 6869 }, { "epoch": 0.47210005497526114, "grad_norm": 0.6671610474586487, "learning_rate": 1.7078228880454623e-05, "loss": 0.0202, "step": 6870 }, { "epoch": 0.47216877405167673, "grad_norm": 0.9315387606620789, "learning_rate": 1.7074922459193903e-05, "loss": 0.0175, "step": 6871 }, { "epoch": 0.4722374931280924, "grad_norm": 0.7382369041442871, "learning_rate": 1.7071615935145785e-05, "loss": 0.0282, "step": 6872 }, { "epoch": 0.47230621220450797, "grad_norm": 1.163941740989685, "learning_rate": 1.7068309308474073e-05, "loss": 0.0174, "step": 6873 }, { "epoch": 0.47237493128092356, "grad_norm": 0.6421414613723755, "learning_rate": 1.7065002579342573e-05, "loss": 0.0183, "step": 6874 }, { "epoch": 0.4724436503573392, "grad_norm": 0.5607871413230896, "learning_rate": 1.7061695747915086e-05, "loss": 0.0155, "step": 6875 }, { "epoch": 0.4725123694337548, "grad_norm": 0.48024412989616394, "learning_rate": 1.7058388814355432e-05, "loss": 0.0083, "step": 6876 }, { "epoch": 0.47258108851017044, "grad_norm": 1.4038163423538208, "learning_rate": 1.7055081778827432e-05, "loss": 0.0209, "step": 6877 }, { "epoch": 0.47264980758658603, "grad_norm": 0.477679580450058, "learning_rate": 1.7051774641494904e-05, "loss": 0.0144, "step": 6878 }, { "epoch": 0.4727185266630016, "grad_norm": 0.8033351302146912, "learning_rate": 1.7048467402521675e-05, "loss": 0.0181, "step": 6879 }, { "epoch": 0.4727872457394173, "grad_norm": 0.6568678021430969, "learning_rate": 1.7045160062071585e-05, "loss": 0.0128, "step": 6880 }, { "epoch": 0.47285596481583286, "grad_norm": 1.267724871635437, "learning_rate": 1.704185262030847e-05, "loss": 0.0238, "step": 6881 }, { "epoch": 0.4729246838922485, "grad_norm": 0.9083858728408813, "learning_rate": 1.7038545077396178e-05, "loss": 0.021, "step": 6882 }, { "epoch": 0.4729934029686641, "grad_norm": 0.6418748497962952, "learning_rate": 1.703523743349855e-05, "loss": 0.0262, "step": 6883 }, { "epoch": 0.4730621220450797, "grad_norm": 0.6866761445999146, "learning_rate": 1.703192968877945e-05, "loss": 0.0289, "step": 6884 }, { "epoch": 0.47313084112149534, "grad_norm": 0.6462776064872742, "learning_rate": 1.702862184340272e-05, "loss": 0.0148, "step": 6885 }, { "epoch": 0.47319956019791093, "grad_norm": 0.6977246403694153, "learning_rate": 1.702531389753224e-05, "loss": 0.0282, "step": 6886 }, { "epoch": 0.4732682792743266, "grad_norm": 0.43552640080451965, "learning_rate": 1.7022005851331874e-05, "loss": 0.0192, "step": 6887 }, { "epoch": 0.47333699835074217, "grad_norm": 0.9938355684280396, "learning_rate": 1.701869770496549e-05, "loss": 0.0261, "step": 6888 }, { "epoch": 0.47340571742715776, "grad_norm": 0.9350991249084473, "learning_rate": 1.7015389458596976e-05, "loss": 0.0163, "step": 6889 }, { "epoch": 0.4734744365035734, "grad_norm": 0.4611052870750427, "learning_rate": 1.7012081112390214e-05, "loss": 0.0245, "step": 6890 }, { "epoch": 0.473543155579989, "grad_norm": 0.4898296296596527, "learning_rate": 1.7008772666509083e-05, "loss": 0.0351, "step": 6891 }, { "epoch": 0.47361187465640464, "grad_norm": 0.623269259929657, "learning_rate": 1.7005464121117492e-05, "loss": 0.021, "step": 6892 }, { "epoch": 0.47368059373282023, "grad_norm": 0.32692384719848633, "learning_rate": 1.7002155476379327e-05, "loss": 0.0165, "step": 6893 }, { "epoch": 0.4737493128092358, "grad_norm": 0.7490169405937195, "learning_rate": 1.6998846732458497e-05, "loss": 0.0217, "step": 6894 }, { "epoch": 0.47381803188565147, "grad_norm": 1.6180800199508667, "learning_rate": 1.699553788951891e-05, "loss": 0.034, "step": 6895 }, { "epoch": 0.47388675096206706, "grad_norm": 0.6158643960952759, "learning_rate": 1.699222894772448e-05, "loss": 0.0225, "step": 6896 }, { "epoch": 0.4739554700384827, "grad_norm": 0.6348586678504944, "learning_rate": 1.698891990723912e-05, "loss": 0.0154, "step": 6897 }, { "epoch": 0.4740241891148983, "grad_norm": 0.577872633934021, "learning_rate": 1.698561076822676e-05, "loss": 0.0154, "step": 6898 }, { "epoch": 0.4740929081913139, "grad_norm": 0.5142132043838501, "learning_rate": 1.6982301530851324e-05, "loss": 0.0196, "step": 6899 }, { "epoch": 0.47416162726772954, "grad_norm": 0.6144260764122009, "learning_rate": 1.6978992195276744e-05, "loss": 0.0171, "step": 6900 }, { "epoch": 0.4742303463441451, "grad_norm": 0.4645150303840637, "learning_rate": 1.6975682761666963e-05, "loss": 0.0177, "step": 6901 }, { "epoch": 0.4742990654205608, "grad_norm": 0.6808245778083801, "learning_rate": 1.6972373230185916e-05, "loss": 0.0218, "step": 6902 }, { "epoch": 0.47436778449697636, "grad_norm": 1.0158605575561523, "learning_rate": 1.6969063600997556e-05, "loss": 0.0256, "step": 6903 }, { "epoch": 0.47443650357339195, "grad_norm": 0.5155683755874634, "learning_rate": 1.6965753874265833e-05, "loss": 0.0141, "step": 6904 }, { "epoch": 0.4745052226498076, "grad_norm": 0.719303548336029, "learning_rate": 1.6962444050154704e-05, "loss": 0.0188, "step": 6905 }, { "epoch": 0.4745739417262232, "grad_norm": 0.6303404569625854, "learning_rate": 1.695913412882813e-05, "loss": 0.02, "step": 6906 }, { "epoch": 0.47464266080263884, "grad_norm": 0.7775236368179321, "learning_rate": 1.6955824110450084e-05, "loss": 0.0186, "step": 6907 }, { "epoch": 0.47471137987905443, "grad_norm": 1.2448530197143555, "learning_rate": 1.6952513995184525e-05, "loss": 0.0219, "step": 6908 }, { "epoch": 0.47478009895547, "grad_norm": 0.5697184801101685, "learning_rate": 1.6949203783195444e-05, "loss": 0.0277, "step": 6909 }, { "epoch": 0.47484881803188567, "grad_norm": 0.45330971479415894, "learning_rate": 1.6945893474646812e-05, "loss": 0.0187, "step": 6910 }, { "epoch": 0.47491753710830126, "grad_norm": 0.44715815782546997, "learning_rate": 1.694258306970262e-05, "loss": 0.02, "step": 6911 }, { "epoch": 0.4749862561847169, "grad_norm": 0.760844349861145, "learning_rate": 1.693927256852685e-05, "loss": 0.0147, "step": 6912 }, { "epoch": 0.4750549752611325, "grad_norm": 1.0756431818008423, "learning_rate": 1.693596197128351e-05, "loss": 0.0264, "step": 6913 }, { "epoch": 0.4751236943375481, "grad_norm": 0.4931752681732178, "learning_rate": 1.6932651278136594e-05, "loss": 0.0201, "step": 6914 }, { "epoch": 0.47519241341396373, "grad_norm": 0.5134066343307495, "learning_rate": 1.6929340489250103e-05, "loss": 0.0254, "step": 6915 }, { "epoch": 0.4752611324903793, "grad_norm": 0.8925802707672119, "learning_rate": 1.6926029604788056e-05, "loss": 0.0201, "step": 6916 }, { "epoch": 0.47532985156679497, "grad_norm": 0.9903644919395447, "learning_rate": 1.692271862491446e-05, "loss": 0.0276, "step": 6917 }, { "epoch": 0.47539857064321056, "grad_norm": 0.7160903215408325, "learning_rate": 1.6919407549793336e-05, "loss": 0.0195, "step": 6918 }, { "epoch": 0.47546728971962615, "grad_norm": 0.9145036339759827, "learning_rate": 1.691609637958871e-05, "loss": 0.0197, "step": 6919 }, { "epoch": 0.4755360087960418, "grad_norm": 0.4218801259994507, "learning_rate": 1.6912785114464608e-05, "loss": 0.0138, "step": 6920 }, { "epoch": 0.4756047278724574, "grad_norm": 0.8675079345703125, "learning_rate": 1.6909473754585058e-05, "loss": 0.0279, "step": 6921 }, { "epoch": 0.475673446948873, "grad_norm": 0.648254930973053, "learning_rate": 1.6906162300114107e-05, "loss": 0.0261, "step": 6922 }, { "epoch": 0.4757421660252886, "grad_norm": 0.3509860932826996, "learning_rate": 1.6902850751215798e-05, "loss": 0.0218, "step": 6923 }, { "epoch": 0.4758108851017042, "grad_norm": 0.47432178258895874, "learning_rate": 1.6899539108054173e-05, "loss": 0.021, "step": 6924 }, { "epoch": 0.47587960417811986, "grad_norm": 0.7696526050567627, "learning_rate": 1.6896227370793284e-05, "loss": 0.0279, "step": 6925 }, { "epoch": 0.47594832325453545, "grad_norm": 0.5533406734466553, "learning_rate": 1.6892915539597192e-05, "loss": 0.0175, "step": 6926 }, { "epoch": 0.47601704233095105, "grad_norm": 0.47855958342552185, "learning_rate": 1.6889603614629956e-05, "loss": 0.0197, "step": 6927 }, { "epoch": 0.4760857614073667, "grad_norm": 1.0106438398361206, "learning_rate": 1.688629159605564e-05, "loss": 0.021, "step": 6928 }, { "epoch": 0.4761544804837823, "grad_norm": 0.8992735147476196, "learning_rate": 1.688297948403832e-05, "loss": 0.0167, "step": 6929 }, { "epoch": 0.47622319956019793, "grad_norm": 0.6462664008140564, "learning_rate": 1.687966727874206e-05, "loss": 0.0149, "step": 6930 }, { "epoch": 0.4762919186366135, "grad_norm": 0.44433343410491943, "learning_rate": 1.6876354980330957e-05, "loss": 0.0142, "step": 6931 }, { "epoch": 0.4763606377130291, "grad_norm": 0.6866230964660645, "learning_rate": 1.6873042588969084e-05, "loss": 0.0198, "step": 6932 }, { "epoch": 0.47642935678944476, "grad_norm": 0.6838734149932861, "learning_rate": 1.686973010482052e-05, "loss": 0.02, "step": 6933 }, { "epoch": 0.47649807586586035, "grad_norm": 0.7688747644424438, "learning_rate": 1.6866417528049384e-05, "loss": 0.0191, "step": 6934 }, { "epoch": 0.476566794942276, "grad_norm": 0.476921409368515, "learning_rate": 1.686310485881976e-05, "loss": 0.0079, "step": 6935 }, { "epoch": 0.4766355140186916, "grad_norm": 0.7459350824356079, "learning_rate": 1.6859792097295746e-05, "loss": 0.0306, "step": 6936 }, { "epoch": 0.4767042330951072, "grad_norm": 0.9813913702964783, "learning_rate": 1.685647924364146e-05, "loss": 0.0191, "step": 6937 }, { "epoch": 0.4767729521715228, "grad_norm": 0.5841884016990662, "learning_rate": 1.6853166298021007e-05, "loss": 0.0139, "step": 6938 }, { "epoch": 0.4768416712479384, "grad_norm": 0.3533121347427368, "learning_rate": 1.6849853260598503e-05, "loss": 0.0227, "step": 6939 }, { "epoch": 0.47691039032435406, "grad_norm": 0.7124837040901184, "learning_rate": 1.6846540131538072e-05, "loss": 0.0219, "step": 6940 }, { "epoch": 0.47697910940076965, "grad_norm": 0.7125833630561829, "learning_rate": 1.6843226911003843e-05, "loss": 0.024, "step": 6941 }, { "epoch": 0.47704782847718524, "grad_norm": 0.5868133902549744, "learning_rate": 1.683991359915994e-05, "loss": 0.0229, "step": 6942 }, { "epoch": 0.4771165475536009, "grad_norm": 0.5885576605796814, "learning_rate": 1.68366001961705e-05, "loss": 0.0254, "step": 6943 }, { "epoch": 0.4771852666300165, "grad_norm": 0.5071339011192322, "learning_rate": 1.6833286702199663e-05, "loss": 0.0223, "step": 6944 }, { "epoch": 0.4772539857064321, "grad_norm": 1.5187779664993286, "learning_rate": 1.682997311741157e-05, "loss": 0.0362, "step": 6945 }, { "epoch": 0.4773227047828477, "grad_norm": 0.37601086497306824, "learning_rate": 1.682665944197037e-05, "loss": 0.014, "step": 6946 }, { "epoch": 0.4773914238592633, "grad_norm": 0.6530755162239075, "learning_rate": 1.6823345676040223e-05, "loss": 0.0175, "step": 6947 }, { "epoch": 0.47746014293567896, "grad_norm": 0.9701388478279114, "learning_rate": 1.6820031819785273e-05, "loss": 0.0244, "step": 6948 }, { "epoch": 0.47752886201209455, "grad_norm": 1.6253901720046997, "learning_rate": 1.6816717873369693e-05, "loss": 0.0144, "step": 6949 }, { "epoch": 0.4775975810885102, "grad_norm": 0.43813198804855347, "learning_rate": 1.681340383695764e-05, "loss": 0.0183, "step": 6950 }, { "epoch": 0.4776663001649258, "grad_norm": 0.6522157788276672, "learning_rate": 1.681008971071329e-05, "loss": 0.0214, "step": 6951 }, { "epoch": 0.4777350192413414, "grad_norm": 0.98016357421875, "learning_rate": 1.6806775494800823e-05, "loss": 0.0256, "step": 6952 }, { "epoch": 0.477803738317757, "grad_norm": 1.1053204536437988, "learning_rate": 1.680346118938441e-05, "loss": 0.0234, "step": 6953 }, { "epoch": 0.4778724573941726, "grad_norm": 0.5409238338470459, "learning_rate": 1.6800146794628236e-05, "loss": 0.0219, "step": 6954 }, { "epoch": 0.47794117647058826, "grad_norm": 0.29655641317367554, "learning_rate": 1.679683231069649e-05, "loss": 0.0146, "step": 6955 }, { "epoch": 0.47800989554700385, "grad_norm": 0.6629018783569336, "learning_rate": 1.6793517737753365e-05, "loss": 0.0156, "step": 6956 }, { "epoch": 0.47807861462341944, "grad_norm": 0.5914449095726013, "learning_rate": 1.6790203075963062e-05, "loss": 0.0177, "step": 6957 }, { "epoch": 0.4781473336998351, "grad_norm": 0.8722125291824341, "learning_rate": 1.6786888325489778e-05, "loss": 0.0285, "step": 6958 }, { "epoch": 0.4782160527762507, "grad_norm": 0.4959457814693451, "learning_rate": 1.6783573486497716e-05, "loss": 0.0159, "step": 6959 }, { "epoch": 0.4782847718526663, "grad_norm": 1.265737533569336, "learning_rate": 1.6780258559151095e-05, "loss": 0.0227, "step": 6960 }, { "epoch": 0.4783534909290819, "grad_norm": 1.1319284439086914, "learning_rate": 1.6776943543614124e-05, "loss": 0.0238, "step": 6961 }, { "epoch": 0.4784222100054975, "grad_norm": 0.8833000659942627, "learning_rate": 1.6773628440051022e-05, "loss": 0.0206, "step": 6962 }, { "epoch": 0.47849092908191315, "grad_norm": 0.7379330992698669, "learning_rate": 1.6770313248626013e-05, "loss": 0.021, "step": 6963 }, { "epoch": 0.47855964815832874, "grad_norm": 0.47233375906944275, "learning_rate": 1.676699796950333e-05, "loss": 0.013, "step": 6964 }, { "epoch": 0.4786283672347444, "grad_norm": 0.4759550392627716, "learning_rate": 1.6763682602847197e-05, "loss": 0.0176, "step": 6965 }, { "epoch": 0.47869708631116, "grad_norm": 0.33201584219932556, "learning_rate": 1.676036714882185e-05, "loss": 0.017, "step": 6966 }, { "epoch": 0.47876580538757557, "grad_norm": 0.4132145941257477, "learning_rate": 1.6757051607591538e-05, "loss": 0.0129, "step": 6967 }, { "epoch": 0.4788345244639912, "grad_norm": 0.6881797909736633, "learning_rate": 1.67537359793205e-05, "loss": 0.0191, "step": 6968 }, { "epoch": 0.4789032435404068, "grad_norm": 0.48087629675865173, "learning_rate": 1.6750420264172987e-05, "loss": 0.0119, "step": 6969 }, { "epoch": 0.47897196261682246, "grad_norm": 0.812212347984314, "learning_rate": 1.6747104462313254e-05, "loss": 0.0171, "step": 6970 }, { "epoch": 0.47904068169323805, "grad_norm": 1.775980830192566, "learning_rate": 1.6743788573905555e-05, "loss": 0.0289, "step": 6971 }, { "epoch": 0.47910940076965364, "grad_norm": 0.7868577837944031, "learning_rate": 1.674047259911416e-05, "loss": 0.0226, "step": 6972 }, { "epoch": 0.4791781198460693, "grad_norm": 0.6731768846511841, "learning_rate": 1.6737156538103327e-05, "loss": 0.0159, "step": 6973 }, { "epoch": 0.4792468389224849, "grad_norm": 0.6197898387908936, "learning_rate": 1.6733840391037332e-05, "loss": 0.0203, "step": 6974 }, { "epoch": 0.4793155579989005, "grad_norm": 0.6948918700218201, "learning_rate": 1.673052415808045e-05, "loss": 0.029, "step": 6975 }, { "epoch": 0.4793842770753161, "grad_norm": 0.5054660439491272, "learning_rate": 1.672720783939696e-05, "loss": 0.0174, "step": 6976 }, { "epoch": 0.4794529961517317, "grad_norm": 0.37355726957321167, "learning_rate": 1.672389143515114e-05, "loss": 0.0237, "step": 6977 }, { "epoch": 0.47952171522814735, "grad_norm": 0.8206586837768555, "learning_rate": 1.6720574945507283e-05, "loss": 0.0217, "step": 6978 }, { "epoch": 0.47959043430456294, "grad_norm": 0.8698976039886475, "learning_rate": 1.6717258370629686e-05, "loss": 0.0112, "step": 6979 }, { "epoch": 0.4796591533809786, "grad_norm": 1.1911317110061646, "learning_rate": 1.6713941710682637e-05, "loss": 0.0166, "step": 6980 }, { "epoch": 0.4797278724573942, "grad_norm": 0.3145886957645416, "learning_rate": 1.6710624965830434e-05, "loss": 0.0115, "step": 6981 }, { "epoch": 0.47979659153380977, "grad_norm": 0.503339409828186, "learning_rate": 1.6707308136237396e-05, "loss": 0.0137, "step": 6982 }, { "epoch": 0.4798653106102254, "grad_norm": 0.6832849979400635, "learning_rate": 1.6703991222067817e-05, "loss": 0.0214, "step": 6983 }, { "epoch": 0.479934029686641, "grad_norm": 0.750397264957428, "learning_rate": 1.6700674223486014e-05, "loss": 0.0184, "step": 6984 }, { "epoch": 0.4800027487630566, "grad_norm": 0.9242172837257385, "learning_rate": 1.6697357140656312e-05, "loss": 0.021, "step": 6985 }, { "epoch": 0.48007146783947224, "grad_norm": 0.6681400537490845, "learning_rate": 1.6694039973743025e-05, "loss": 0.0206, "step": 6986 }, { "epoch": 0.48014018691588783, "grad_norm": 0.9515644311904907, "learning_rate": 1.669072272291048e-05, "loss": 0.0222, "step": 6987 }, { "epoch": 0.4802089059923035, "grad_norm": 0.3226674795150757, "learning_rate": 1.668740538832301e-05, "loss": 0.0192, "step": 6988 }, { "epoch": 0.48027762506871907, "grad_norm": 0.9334815740585327, "learning_rate": 1.6684087970144944e-05, "loss": 0.0247, "step": 6989 }, { "epoch": 0.48034634414513466, "grad_norm": 0.23823639750480652, "learning_rate": 1.668077046854062e-05, "loss": 0.0196, "step": 6990 }, { "epoch": 0.4804150632215503, "grad_norm": 0.40829893946647644, "learning_rate": 1.6677452883674386e-05, "loss": 0.0123, "step": 6991 }, { "epoch": 0.4804837822979659, "grad_norm": 0.9382230043411255, "learning_rate": 1.6674135215710583e-05, "loss": 0.0174, "step": 6992 }, { "epoch": 0.48055250137438155, "grad_norm": 1.1469299793243408, "learning_rate": 1.6670817464813565e-05, "loss": 0.0359, "step": 6993 }, { "epoch": 0.48062122045079714, "grad_norm": 0.5089296102523804, "learning_rate": 1.6667499631147685e-05, "loss": 0.0118, "step": 6994 }, { "epoch": 0.48068993952721273, "grad_norm": 0.7012317776679993, "learning_rate": 1.66641817148773e-05, "loss": 0.0143, "step": 6995 }, { "epoch": 0.4807586586036284, "grad_norm": 0.564393162727356, "learning_rate": 1.666086371616677e-05, "loss": 0.0176, "step": 6996 }, { "epoch": 0.48082737768004397, "grad_norm": 0.7028575539588928, "learning_rate": 1.6657545635180477e-05, "loss": 0.0292, "step": 6997 }, { "epoch": 0.4808960967564596, "grad_norm": 0.724671483039856, "learning_rate": 1.6654227472082776e-05, "loss": 0.0202, "step": 6998 }, { "epoch": 0.4809648158328752, "grad_norm": 0.43317490816116333, "learning_rate": 1.6650909227038042e-05, "loss": 0.0175, "step": 6999 }, { "epoch": 0.4810335349092908, "grad_norm": 1.1439034938812256, "learning_rate": 1.664759090021066e-05, "loss": 0.0104, "step": 7000 }, { "epoch": 0.48110225398570644, "grad_norm": 0.2697087526321411, "learning_rate": 1.6644272491765018e-05, "loss": 0.0217, "step": 7001 }, { "epoch": 0.48117097306212203, "grad_norm": 0.958336353302002, "learning_rate": 1.664095400186549e-05, "loss": 0.0281, "step": 7002 }, { "epoch": 0.4812396921385377, "grad_norm": 0.3711804449558258, "learning_rate": 1.663763543067648e-05, "loss": 0.0214, "step": 7003 }, { "epoch": 0.48130841121495327, "grad_norm": 1.1859804391860962, "learning_rate": 1.6634316778362378e-05, "loss": 0.02, "step": 7004 }, { "epoch": 0.48137713029136886, "grad_norm": 0.4427025318145752, "learning_rate": 1.663099804508758e-05, "loss": 0.0175, "step": 7005 }, { "epoch": 0.4814458493677845, "grad_norm": 1.9287704229354858, "learning_rate": 1.6627679231016492e-05, "loss": 0.0355, "step": 7006 }, { "epoch": 0.4815145684442001, "grad_norm": 1.1578835248947144, "learning_rate": 1.662436033631352e-05, "loss": 0.0244, "step": 7007 }, { "epoch": 0.48158328752061574, "grad_norm": 1.1920475959777832, "learning_rate": 1.662104136114308e-05, "loss": 0.0351, "step": 7008 }, { "epoch": 0.48165200659703133, "grad_norm": 0.7005287408828735, "learning_rate": 1.661772230566958e-05, "loss": 0.0256, "step": 7009 }, { "epoch": 0.4817207256734469, "grad_norm": 2.421257734298706, "learning_rate": 1.661440317005745e-05, "loss": 0.0358, "step": 7010 }, { "epoch": 0.48178944474986257, "grad_norm": 1.5786941051483154, "learning_rate": 1.6611083954471102e-05, "loss": 0.0229, "step": 7011 }, { "epoch": 0.48185816382627816, "grad_norm": 1.3154915571212769, "learning_rate": 1.6607764659074966e-05, "loss": 0.0262, "step": 7012 }, { "epoch": 0.4819268829026938, "grad_norm": 0.4375627636909485, "learning_rate": 1.6604445284033477e-05, "loss": 0.0207, "step": 7013 }, { "epoch": 0.4819956019791094, "grad_norm": 0.7871355414390564, "learning_rate": 1.6601125829511062e-05, "loss": 0.0254, "step": 7014 }, { "epoch": 0.482064321055525, "grad_norm": 1.71006441116333, "learning_rate": 1.6597806295672167e-05, "loss": 0.0293, "step": 7015 }, { "epoch": 0.48213304013194064, "grad_norm": 0.41505375504493713, "learning_rate": 1.6594486682681234e-05, "loss": 0.022, "step": 7016 }, { "epoch": 0.48220175920835623, "grad_norm": 0.4873087406158447, "learning_rate": 1.6591166990702707e-05, "loss": 0.0248, "step": 7017 }, { "epoch": 0.4822704782847719, "grad_norm": 0.24727904796600342, "learning_rate": 1.658784721990104e-05, "loss": 0.0198, "step": 7018 }, { "epoch": 0.48233919736118747, "grad_norm": 0.27986758947372437, "learning_rate": 1.658452737044069e-05, "loss": 0.0164, "step": 7019 }, { "epoch": 0.48240791643760306, "grad_norm": 1.4349186420440674, "learning_rate": 1.6581207442486104e-05, "loss": 0.0402, "step": 7020 }, { "epoch": 0.4824766355140187, "grad_norm": 0.28821176290512085, "learning_rate": 1.6577887436201758e-05, "loss": 0.0274, "step": 7021 }, { "epoch": 0.4825453545904343, "grad_norm": 1.9661815166473389, "learning_rate": 1.6574567351752114e-05, "loss": 0.0315, "step": 7022 }, { "epoch": 0.48261407366684994, "grad_norm": 0.8972253799438477, "learning_rate": 1.6571247189301635e-05, "loss": 0.0174, "step": 7023 }, { "epoch": 0.48268279274326553, "grad_norm": 0.487913578748703, "learning_rate": 1.6567926949014805e-05, "loss": 0.0277, "step": 7024 }, { "epoch": 0.4827515118196811, "grad_norm": 0.7218143939971924, "learning_rate": 1.6564606631056094e-05, "loss": 0.0156, "step": 7025 }, { "epoch": 0.48282023089609677, "grad_norm": 1.2374577522277832, "learning_rate": 1.6561286235589985e-05, "loss": 0.045, "step": 7026 }, { "epoch": 0.48288894997251236, "grad_norm": 0.7957454323768616, "learning_rate": 1.6557965762780973e-05, "loss": 0.0153, "step": 7027 }, { "epoch": 0.482957669048928, "grad_norm": 0.6331928372383118, "learning_rate": 1.6554645212793537e-05, "loss": 0.0241, "step": 7028 }, { "epoch": 0.4830263881253436, "grad_norm": 0.5280708074569702, "learning_rate": 1.6551324585792175e-05, "loss": 0.0246, "step": 7029 }, { "epoch": 0.4830951072017592, "grad_norm": 0.760355532169342, "learning_rate": 1.6548003881941377e-05, "loss": 0.0225, "step": 7030 }, { "epoch": 0.48316382627817483, "grad_norm": 0.6046249866485596, "learning_rate": 1.6544683101405654e-05, "loss": 0.0242, "step": 7031 }, { "epoch": 0.4832325453545904, "grad_norm": 0.7324709296226501, "learning_rate": 1.6541362244349503e-05, "loss": 0.0338, "step": 7032 }, { "epoch": 0.4833012644310061, "grad_norm": 0.2300824671983719, "learning_rate": 1.653804131093744e-05, "loss": 0.0158, "step": 7033 }, { "epoch": 0.48336998350742166, "grad_norm": 0.32322755455970764, "learning_rate": 1.653472030133397e-05, "loss": 0.0131, "step": 7034 }, { "epoch": 0.48343870258383725, "grad_norm": 0.8999955058097839, "learning_rate": 1.6531399215703607e-05, "loss": 0.024, "step": 7035 }, { "epoch": 0.4835074216602529, "grad_norm": 0.8348044753074646, "learning_rate": 1.652807805421088e-05, "loss": 0.0195, "step": 7036 }, { "epoch": 0.4835761407366685, "grad_norm": 0.42751672863960266, "learning_rate": 1.6524756817020308e-05, "loss": 0.0179, "step": 7037 }, { "epoch": 0.48364485981308414, "grad_norm": 0.6353347897529602, "learning_rate": 1.6521435504296417e-05, "loss": 0.0167, "step": 7038 }, { "epoch": 0.48371357888949973, "grad_norm": 0.6431682109832764, "learning_rate": 1.651811411620374e-05, "loss": 0.0164, "step": 7039 }, { "epoch": 0.4837822979659153, "grad_norm": 0.8506988883018494, "learning_rate": 1.6514792652906807e-05, "loss": 0.0185, "step": 7040 }, { "epoch": 0.48385101704233097, "grad_norm": 0.6625720262527466, "learning_rate": 1.651147111457016e-05, "loss": 0.0195, "step": 7041 }, { "epoch": 0.48391973611874656, "grad_norm": 0.5009884834289551, "learning_rate": 1.6508149501358347e-05, "loss": 0.0144, "step": 7042 }, { "epoch": 0.4839884551951622, "grad_norm": 0.29776236414909363, "learning_rate": 1.6504827813435906e-05, "loss": 0.0144, "step": 7043 }, { "epoch": 0.4840571742715778, "grad_norm": 3.338679790496826, "learning_rate": 1.650150605096739e-05, "loss": 0.014, "step": 7044 }, { "epoch": 0.4841258933479934, "grad_norm": 0.828987717628479, "learning_rate": 1.649818421411735e-05, "loss": 0.0203, "step": 7045 }, { "epoch": 0.48419461242440903, "grad_norm": 1.0378038883209229, "learning_rate": 1.6494862303050343e-05, "loss": 0.0196, "step": 7046 }, { "epoch": 0.4842633315008246, "grad_norm": 1.7483906745910645, "learning_rate": 1.6491540317930932e-05, "loss": 0.0396, "step": 7047 }, { "epoch": 0.48433205057724027, "grad_norm": 0.7734529376029968, "learning_rate": 1.6488218258923683e-05, "loss": 0.0243, "step": 7048 }, { "epoch": 0.48440076965365586, "grad_norm": 1.273748517036438, "learning_rate": 1.648489612619316e-05, "loss": 0.0392, "step": 7049 }, { "epoch": 0.48446948873007145, "grad_norm": 0.626262366771698, "learning_rate": 1.648157391990393e-05, "loss": 0.0128, "step": 7050 }, { "epoch": 0.4845382078064871, "grad_norm": 1.167648434638977, "learning_rate": 1.6478251640220577e-05, "loss": 0.0391, "step": 7051 }, { "epoch": 0.4846069268829027, "grad_norm": 0.9400454163551331, "learning_rate": 1.6474929287307684e-05, "loss": 0.0146, "step": 7052 }, { "epoch": 0.4846756459593183, "grad_norm": 0.9934290051460266, "learning_rate": 1.647160686132982e-05, "loss": 0.0256, "step": 7053 }, { "epoch": 0.4847443650357339, "grad_norm": 0.5034037828445435, "learning_rate": 1.6468284362451574e-05, "loss": 0.0131, "step": 7054 }, { "epoch": 0.4848130841121495, "grad_norm": 0.6981902122497559, "learning_rate": 1.6464961790837545e-05, "loss": 0.013, "step": 7055 }, { "epoch": 0.48488180318856516, "grad_norm": 2.6532998085021973, "learning_rate": 1.6461639146652318e-05, "loss": 0.0221, "step": 7056 }, { "epoch": 0.48495052226498075, "grad_norm": 0.7870790958404541, "learning_rate": 1.6458316430060493e-05, "loss": 0.0278, "step": 7057 }, { "epoch": 0.48501924134139635, "grad_norm": 0.9546906352043152, "learning_rate": 1.6454993641226674e-05, "loss": 0.0141, "step": 7058 }, { "epoch": 0.485087960417812, "grad_norm": 0.22626982629299164, "learning_rate": 1.6451670780315457e-05, "loss": 0.0206, "step": 7059 }, { "epoch": 0.4851566794942276, "grad_norm": 0.6735404133796692, "learning_rate": 1.644834784749145e-05, "loss": 0.0203, "step": 7060 }, { "epoch": 0.48522539857064323, "grad_norm": 0.5852230191230774, "learning_rate": 1.6445024842919276e-05, "loss": 0.0234, "step": 7061 }, { "epoch": 0.4852941176470588, "grad_norm": 1.0846738815307617, "learning_rate": 1.6441701766763543e-05, "loss": 0.0238, "step": 7062 }, { "epoch": 0.4853628367234744, "grad_norm": 0.7491770386695862, "learning_rate": 1.6438378619188862e-05, "loss": 0.0192, "step": 7063 }, { "epoch": 0.48543155579989006, "grad_norm": 0.9859089851379395, "learning_rate": 1.6435055400359865e-05, "loss": 0.018, "step": 7064 }, { "epoch": 0.48550027487630565, "grad_norm": 0.6417843699455261, "learning_rate": 1.6431732110441175e-05, "loss": 0.0228, "step": 7065 }, { "epoch": 0.4855689939527213, "grad_norm": 0.48783665895462036, "learning_rate": 1.6428408749597418e-05, "loss": 0.0288, "step": 7066 }, { "epoch": 0.4856377130291369, "grad_norm": 0.7149208784103394, "learning_rate": 1.6425085317993226e-05, "loss": 0.0345, "step": 7067 }, { "epoch": 0.4857064321055525, "grad_norm": 0.7862508296966553, "learning_rate": 1.642176181579324e-05, "loss": 0.0212, "step": 7068 }, { "epoch": 0.4857751511819681, "grad_norm": 0.8778502941131592, "learning_rate": 1.6418438243162098e-05, "loss": 0.0164, "step": 7069 }, { "epoch": 0.4858438702583837, "grad_norm": 1.1815699338912964, "learning_rate": 1.641511460026444e-05, "loss": 0.0266, "step": 7070 }, { "epoch": 0.48591258933479936, "grad_norm": 0.7294383645057678, "learning_rate": 1.6411790887264915e-05, "loss": 0.0233, "step": 7071 }, { "epoch": 0.48598130841121495, "grad_norm": 0.6873044371604919, "learning_rate": 1.6408467104328167e-05, "loss": 0.022, "step": 7072 }, { "epoch": 0.48605002748763054, "grad_norm": 0.7716236114501953, "learning_rate": 1.640514325161886e-05, "loss": 0.0245, "step": 7073 }, { "epoch": 0.4861187465640462, "grad_norm": 0.8138371706008911, "learning_rate": 1.6401819329301644e-05, "loss": 0.0143, "step": 7074 }, { "epoch": 0.4861874656404618, "grad_norm": 1.0277272462844849, "learning_rate": 1.6398495337541178e-05, "loss": 0.0124, "step": 7075 }, { "epoch": 0.4862561847168774, "grad_norm": 0.48618945479393005, "learning_rate": 1.639517127650213e-05, "loss": 0.0158, "step": 7076 }, { "epoch": 0.486324903793293, "grad_norm": 0.3890741169452667, "learning_rate": 1.6391847146349167e-05, "loss": 0.013, "step": 7077 }, { "epoch": 0.4863936228697086, "grad_norm": 0.6806914806365967, "learning_rate": 1.6388522947246954e-05, "loss": 0.0299, "step": 7078 }, { "epoch": 0.48646234194612425, "grad_norm": 0.4330016076564789, "learning_rate": 1.638519867936017e-05, "loss": 0.0233, "step": 7079 }, { "epoch": 0.48653106102253985, "grad_norm": 0.9708873629570007, "learning_rate": 1.6381874342853498e-05, "loss": 0.0156, "step": 7080 }, { "epoch": 0.4865997800989555, "grad_norm": 1.2264513969421387, "learning_rate": 1.63785499378916e-05, "loss": 0.0316, "step": 7081 }, { "epoch": 0.4866684991753711, "grad_norm": 0.681632399559021, "learning_rate": 1.637522546463918e-05, "loss": 0.0225, "step": 7082 }, { "epoch": 0.4867372182517867, "grad_norm": 1.311324954032898, "learning_rate": 1.637190092326092e-05, "loss": 0.0176, "step": 7083 }, { "epoch": 0.4868059373282023, "grad_norm": 0.7869303226470947, "learning_rate": 1.6368576313921502e-05, "loss": 0.0274, "step": 7084 }, { "epoch": 0.4868746564046179, "grad_norm": 0.7532734274864197, "learning_rate": 1.636525163678563e-05, "loss": 0.0187, "step": 7085 }, { "epoch": 0.48694337548103356, "grad_norm": 0.3501179814338684, "learning_rate": 1.6361926892018e-05, "loss": 0.0123, "step": 7086 }, { "epoch": 0.48701209455744915, "grad_norm": 0.5349522829055786, "learning_rate": 1.6358602079783312e-05, "loss": 0.023, "step": 7087 }, { "epoch": 0.48708081363386474, "grad_norm": 0.34008917212486267, "learning_rate": 1.6355277200246273e-05, "loss": 0.0227, "step": 7088 }, { "epoch": 0.4871495327102804, "grad_norm": 0.9803436398506165, "learning_rate": 1.6351952253571585e-05, "loss": 0.0279, "step": 7089 }, { "epoch": 0.487218251786696, "grad_norm": 0.7032672166824341, "learning_rate": 1.6348627239923954e-05, "loss": 0.0164, "step": 7090 }, { "epoch": 0.4872869708631116, "grad_norm": 0.5652562379837036, "learning_rate": 1.634530215946811e-05, "loss": 0.0174, "step": 7091 }, { "epoch": 0.4873556899395272, "grad_norm": 0.5112316012382507, "learning_rate": 1.6341977012368766e-05, "loss": 0.0228, "step": 7092 }, { "epoch": 0.4874244090159428, "grad_norm": 1.0257654190063477, "learning_rate": 1.633865179879063e-05, "loss": 0.0264, "step": 7093 }, { "epoch": 0.48749312809235845, "grad_norm": 0.7019971609115601, "learning_rate": 1.6335326518898442e-05, "loss": 0.0242, "step": 7094 }, { "epoch": 0.48756184716877404, "grad_norm": 0.721068263053894, "learning_rate": 1.6332001172856926e-05, "loss": 0.0206, "step": 7095 }, { "epoch": 0.4876305662451897, "grad_norm": 0.6191985011100769, "learning_rate": 1.6328675760830804e-05, "loss": 0.0143, "step": 7096 }, { "epoch": 0.4876992853216053, "grad_norm": 0.7634338736534119, "learning_rate": 1.6325350282984822e-05, "loss": 0.0258, "step": 7097 }, { "epoch": 0.48776800439802087, "grad_norm": 0.6895301342010498, "learning_rate": 1.632202473948371e-05, "loss": 0.0258, "step": 7098 }, { "epoch": 0.4878367234744365, "grad_norm": 0.3739456236362457, "learning_rate": 1.6318699130492206e-05, "loss": 0.0187, "step": 7099 }, { "epoch": 0.4879054425508521, "grad_norm": 0.7707299590110779, "learning_rate": 1.6315373456175068e-05, "loss": 0.0304, "step": 7100 }, { "epoch": 0.48797416162726776, "grad_norm": 0.5291299223899841, "learning_rate": 1.6312047716697027e-05, "loss": 0.0212, "step": 7101 }, { "epoch": 0.48804288070368335, "grad_norm": 0.3983880281448364, "learning_rate": 1.630872191222284e-05, "loss": 0.0226, "step": 7102 }, { "epoch": 0.48811159978009894, "grad_norm": 2.3185508251190186, "learning_rate": 1.630539604291726e-05, "loss": 0.0258, "step": 7103 }, { "epoch": 0.4881803188565146, "grad_norm": 0.4946253001689911, "learning_rate": 1.6302070108945045e-05, "loss": 0.0175, "step": 7104 }, { "epoch": 0.4882490379329302, "grad_norm": 0.4796317219734192, "learning_rate": 1.6298744110470952e-05, "loss": 0.0282, "step": 7105 }, { "epoch": 0.4883177570093458, "grad_norm": 0.4028768837451935, "learning_rate": 1.6295418047659748e-05, "loss": 0.0089, "step": 7106 }, { "epoch": 0.4883864760857614, "grad_norm": 0.5873129963874817, "learning_rate": 1.6292091920676195e-05, "loss": 0.017, "step": 7107 }, { "epoch": 0.488455195162177, "grad_norm": 0.6366205215454102, "learning_rate": 1.6288765729685064e-05, "loss": 0.0194, "step": 7108 }, { "epoch": 0.48852391423859265, "grad_norm": 0.6234961152076721, "learning_rate": 1.628543947485113e-05, "loss": 0.0173, "step": 7109 }, { "epoch": 0.48859263331500824, "grad_norm": 0.40015292167663574, "learning_rate": 1.6282113156339168e-05, "loss": 0.0212, "step": 7110 }, { "epoch": 0.4886613523914239, "grad_norm": 0.2998849153518677, "learning_rate": 1.627878677431395e-05, "loss": 0.0164, "step": 7111 }, { "epoch": 0.4887300714678395, "grad_norm": 0.6506744623184204, "learning_rate": 1.6275460328940272e-05, "loss": 0.0136, "step": 7112 }, { "epoch": 0.48879879054425507, "grad_norm": 1.0053712129592896, "learning_rate": 1.6272133820382912e-05, "loss": 0.0199, "step": 7113 }, { "epoch": 0.4888675096206707, "grad_norm": 1.349425196647644, "learning_rate": 1.626880724880665e-05, "loss": 0.0191, "step": 7114 }, { "epoch": 0.4889362286970863, "grad_norm": 0.792613685131073, "learning_rate": 1.6265480614376294e-05, "loss": 0.026, "step": 7115 }, { "epoch": 0.4890049477735019, "grad_norm": 0.9399861693382263, "learning_rate": 1.6262153917256627e-05, "loss": 0.013, "step": 7116 }, { "epoch": 0.48907366684991754, "grad_norm": 0.5326247811317444, "learning_rate": 1.6258827157612447e-05, "loss": 0.0161, "step": 7117 }, { "epoch": 0.48914238592633313, "grad_norm": 0.4080233573913574, "learning_rate": 1.625550033560856e-05, "loss": 0.0118, "step": 7118 }, { "epoch": 0.4892111050027488, "grad_norm": 0.4216197729110718, "learning_rate": 1.625217345140977e-05, "loss": 0.0256, "step": 7119 }, { "epoch": 0.48927982407916437, "grad_norm": 0.6079761981964111, "learning_rate": 1.6248846505180874e-05, "loss": 0.0232, "step": 7120 }, { "epoch": 0.48934854315557996, "grad_norm": 0.7994115352630615, "learning_rate": 1.6245519497086697e-05, "loss": 0.021, "step": 7121 }, { "epoch": 0.4894172622319956, "grad_norm": 1.3588709831237793, "learning_rate": 1.6242192427292045e-05, "loss": 0.0178, "step": 7122 }, { "epoch": 0.4894859813084112, "grad_norm": 1.0896461009979248, "learning_rate": 1.623886529596173e-05, "loss": 0.0225, "step": 7123 }, { "epoch": 0.48955470038482685, "grad_norm": 0.6841952204704285, "learning_rate": 1.623553810326058e-05, "loss": 0.0206, "step": 7124 }, { "epoch": 0.48962341946124244, "grad_norm": 0.6819902658462524, "learning_rate": 1.623221084935341e-05, "loss": 0.0253, "step": 7125 }, { "epoch": 0.48969213853765803, "grad_norm": 0.5734134912490845, "learning_rate": 1.6228883534405043e-05, "loss": 0.0185, "step": 7126 }, { "epoch": 0.4897608576140737, "grad_norm": 1.294704556465149, "learning_rate": 1.6225556158580318e-05, "loss": 0.0126, "step": 7127 }, { "epoch": 0.48982957669048927, "grad_norm": 1.0256041288375854, "learning_rate": 1.622222872204406e-05, "loss": 0.0264, "step": 7128 }, { "epoch": 0.4898982957669049, "grad_norm": 0.7591315507888794, "learning_rate": 1.6218901224961103e-05, "loss": 0.02, "step": 7129 }, { "epoch": 0.4899670148433205, "grad_norm": 0.6951004266738892, "learning_rate": 1.621557366749628e-05, "loss": 0.0173, "step": 7130 }, { "epoch": 0.4900357339197361, "grad_norm": 0.5085456967353821, "learning_rate": 1.6212246049814448e-05, "loss": 0.0259, "step": 7131 }, { "epoch": 0.49010445299615174, "grad_norm": 0.8555705547332764, "learning_rate": 1.6208918372080428e-05, "loss": 0.023, "step": 7132 }, { "epoch": 0.49017317207256733, "grad_norm": 1.0468450784683228, "learning_rate": 1.620559063445908e-05, "loss": 0.0249, "step": 7133 }, { "epoch": 0.490241891148983, "grad_norm": 0.8041121959686279, "learning_rate": 1.6202262837115257e-05, "loss": 0.023, "step": 7134 }, { "epoch": 0.49031061022539857, "grad_norm": 1.026503562927246, "learning_rate": 1.6198934980213796e-05, "loss": 0.03, "step": 7135 }, { "epoch": 0.49037932930181416, "grad_norm": 0.5585587620735168, "learning_rate": 1.619560706391957e-05, "loss": 0.0213, "step": 7136 }, { "epoch": 0.4904480483782298, "grad_norm": 0.9340463280677795, "learning_rate": 1.6192279088397422e-05, "loss": 0.0166, "step": 7137 }, { "epoch": 0.4905167674546454, "grad_norm": 0.2921997010707855, "learning_rate": 1.6188951053812216e-05, "loss": 0.0222, "step": 7138 }, { "epoch": 0.49058548653106104, "grad_norm": 0.6184999942779541, "learning_rate": 1.6185622960328825e-05, "loss": 0.0154, "step": 7139 }, { "epoch": 0.49065420560747663, "grad_norm": 0.5493099689483643, "learning_rate": 1.618229480811211e-05, "loss": 0.0199, "step": 7140 }, { "epoch": 0.4907229246838922, "grad_norm": 0.9473183751106262, "learning_rate": 1.6178966597326942e-05, "loss": 0.0217, "step": 7141 }, { "epoch": 0.49079164376030787, "grad_norm": 0.601729691028595, "learning_rate": 1.617563832813819e-05, "loss": 0.0177, "step": 7142 }, { "epoch": 0.49086036283672346, "grad_norm": 0.5483241081237793, "learning_rate": 1.6172310000710737e-05, "loss": 0.0207, "step": 7143 }, { "epoch": 0.4909290819131391, "grad_norm": 2.2149157524108887, "learning_rate": 1.6168981615209447e-05, "loss": 0.0146, "step": 7144 }, { "epoch": 0.4909978009895547, "grad_norm": 0.702116072177887, "learning_rate": 1.616565317179922e-05, "loss": 0.0177, "step": 7145 }, { "epoch": 0.4910665200659703, "grad_norm": 1.1500273942947388, "learning_rate": 1.6162324670644928e-05, "loss": 0.0277, "step": 7146 }, { "epoch": 0.49113523914238594, "grad_norm": 0.8790556192398071, "learning_rate": 1.6158996111911458e-05, "loss": 0.0142, "step": 7147 }, { "epoch": 0.49120395821880153, "grad_norm": 0.4971044659614563, "learning_rate": 1.615566749576371e-05, "loss": 0.0127, "step": 7148 }, { "epoch": 0.4912726772952172, "grad_norm": 1.2290741205215454, "learning_rate": 1.6152338822366564e-05, "loss": 0.0294, "step": 7149 }, { "epoch": 0.49134139637163277, "grad_norm": 0.4766221046447754, "learning_rate": 1.6149010091884923e-05, "loss": 0.0115, "step": 7150 }, { "epoch": 0.49141011544804836, "grad_norm": 0.3776588439941406, "learning_rate": 1.6145681304483687e-05, "loss": 0.0193, "step": 7151 }, { "epoch": 0.491478834524464, "grad_norm": 1.1234018802642822, "learning_rate": 1.6142352460327755e-05, "loss": 0.0354, "step": 7152 }, { "epoch": 0.4915475536008796, "grad_norm": 0.45639193058013916, "learning_rate": 1.6139023559582026e-05, "loss": 0.0173, "step": 7153 }, { "epoch": 0.49161627267729524, "grad_norm": 0.4305972754955292, "learning_rate": 1.6135694602411417e-05, "loss": 0.0133, "step": 7154 }, { "epoch": 0.49168499175371083, "grad_norm": 0.7831566333770752, "learning_rate": 1.6132365588980832e-05, "loss": 0.017, "step": 7155 }, { "epoch": 0.4917537108301264, "grad_norm": 0.5519199967384338, "learning_rate": 1.6129036519455175e-05, "loss": 0.0208, "step": 7156 }, { "epoch": 0.49182242990654207, "grad_norm": 0.7789026498794556, "learning_rate": 1.612570739399938e-05, "loss": 0.0154, "step": 7157 }, { "epoch": 0.49189114898295766, "grad_norm": 0.6323596835136414, "learning_rate": 1.6122378212778352e-05, "loss": 0.0155, "step": 7158 }, { "epoch": 0.4919598680593733, "grad_norm": 0.6891322731971741, "learning_rate": 1.6119048975957017e-05, "loss": 0.0156, "step": 7159 }, { "epoch": 0.4920285871357889, "grad_norm": 0.9714435338973999, "learning_rate": 1.6115719683700296e-05, "loss": 0.0233, "step": 7160 }, { "epoch": 0.4920973062122045, "grad_norm": 0.38442137837409973, "learning_rate": 1.611239033617311e-05, "loss": 0.0082, "step": 7161 }, { "epoch": 0.49216602528862013, "grad_norm": 0.5449204444885254, "learning_rate": 1.6109060933540397e-05, "loss": 0.0381, "step": 7162 }, { "epoch": 0.4922347443650357, "grad_norm": 0.8057053685188293, "learning_rate": 1.610573147596709e-05, "loss": 0.0168, "step": 7163 }, { "epoch": 0.49230346344145137, "grad_norm": 1.020734429359436, "learning_rate": 1.6102401963618118e-05, "loss": 0.0296, "step": 7164 }, { "epoch": 0.49237218251786696, "grad_norm": 0.6497492790222168, "learning_rate": 1.609907239665841e-05, "loss": 0.0193, "step": 7165 }, { "epoch": 0.49244090159428255, "grad_norm": 0.6700167655944824, "learning_rate": 1.6095742775252923e-05, "loss": 0.0165, "step": 7166 }, { "epoch": 0.4925096206706982, "grad_norm": 0.8305844068527222, "learning_rate": 1.609241309956659e-05, "loss": 0.0367, "step": 7167 }, { "epoch": 0.4925783397471138, "grad_norm": 1.2260380983352661, "learning_rate": 1.6089083369764357e-05, "loss": 0.031, "step": 7168 }, { "epoch": 0.49264705882352944, "grad_norm": 0.3880573809146881, "learning_rate": 1.6085753586011175e-05, "loss": 0.0193, "step": 7169 }, { "epoch": 0.49271577789994503, "grad_norm": 0.6026018857955933, "learning_rate": 1.6082423748471994e-05, "loss": 0.0244, "step": 7170 }, { "epoch": 0.4927844969763606, "grad_norm": 0.5536462068557739, "learning_rate": 1.607909385731176e-05, "loss": 0.0203, "step": 7171 }, { "epoch": 0.49285321605277627, "grad_norm": 0.356258749961853, "learning_rate": 1.6075763912695438e-05, "loss": 0.0139, "step": 7172 }, { "epoch": 0.49292193512919186, "grad_norm": 0.581512451171875, "learning_rate": 1.6072433914787986e-05, "loss": 0.0201, "step": 7173 }, { "epoch": 0.4929906542056075, "grad_norm": 0.6964277029037476, "learning_rate": 1.6069103863754357e-05, "loss": 0.0215, "step": 7174 }, { "epoch": 0.4930593732820231, "grad_norm": 0.5087518692016602, "learning_rate": 1.6065773759759525e-05, "loss": 0.016, "step": 7175 }, { "epoch": 0.4931280923584387, "grad_norm": 0.445036917924881, "learning_rate": 1.6062443602968452e-05, "loss": 0.0219, "step": 7176 }, { "epoch": 0.49319681143485433, "grad_norm": 0.6333205103874207, "learning_rate": 1.6059113393546106e-05, "loss": 0.0217, "step": 7177 }, { "epoch": 0.4932655305112699, "grad_norm": 0.9721063375473022, "learning_rate": 1.6055783131657458e-05, "loss": 0.025, "step": 7178 }, { "epoch": 0.4933342495876855, "grad_norm": 0.840988278388977, "learning_rate": 1.6052452817467487e-05, "loss": 0.018, "step": 7179 }, { "epoch": 0.49340296866410116, "grad_norm": 0.5304230451583862, "learning_rate": 1.6049122451141168e-05, "loss": 0.021, "step": 7180 }, { "epoch": 0.49347168774051675, "grad_norm": 0.5316306352615356, "learning_rate": 1.604579203284348e-05, "loss": 0.0218, "step": 7181 }, { "epoch": 0.4935404068169324, "grad_norm": 1.20958411693573, "learning_rate": 1.60424615627394e-05, "loss": 0.0209, "step": 7182 }, { "epoch": 0.493609125893348, "grad_norm": 0.5748409032821655, "learning_rate": 1.6039131040993923e-05, "loss": 0.0229, "step": 7183 }, { "epoch": 0.4936778449697636, "grad_norm": 0.9318127632141113, "learning_rate": 1.6035800467772027e-05, "loss": 0.0158, "step": 7184 }, { "epoch": 0.4937465640461792, "grad_norm": 0.8849233388900757, "learning_rate": 1.6032469843238705e-05, "loss": 0.0199, "step": 7185 }, { "epoch": 0.4938152831225948, "grad_norm": 1.0864875316619873, "learning_rate": 1.6029139167558952e-05, "loss": 0.0235, "step": 7186 }, { "epoch": 0.49388400219901046, "grad_norm": 0.2565767467021942, "learning_rate": 1.602580844089776e-05, "loss": 0.0127, "step": 7187 }, { "epoch": 0.49395272127542605, "grad_norm": 0.5504622459411621, "learning_rate": 1.6022477663420126e-05, "loss": 0.0176, "step": 7188 }, { "epoch": 0.49402144035184165, "grad_norm": 0.6638232469558716, "learning_rate": 1.6019146835291048e-05, "loss": 0.0192, "step": 7189 }, { "epoch": 0.4940901594282573, "grad_norm": 0.2846587598323822, "learning_rate": 1.6015815956675536e-05, "loss": 0.0152, "step": 7190 }, { "epoch": 0.4941588785046729, "grad_norm": 0.5947182178497314, "learning_rate": 1.601248502773859e-05, "loss": 0.0095, "step": 7191 }, { "epoch": 0.49422759758108853, "grad_norm": 0.46076148748397827, "learning_rate": 1.600915404864521e-05, "loss": 0.0198, "step": 7192 }, { "epoch": 0.4942963166575041, "grad_norm": 0.7727965116500854, "learning_rate": 1.6005823019560418e-05, "loss": 0.0197, "step": 7193 }, { "epoch": 0.4943650357339197, "grad_norm": 0.30881398916244507, "learning_rate": 1.6002491940649224e-05, "loss": 0.0096, "step": 7194 }, { "epoch": 0.49443375481033536, "grad_norm": 0.4363600015640259, "learning_rate": 1.5999160812076634e-05, "loss": 0.0102, "step": 7195 }, { "epoch": 0.49450247388675095, "grad_norm": 0.5156920552253723, "learning_rate": 1.5995829634007677e-05, "loss": 0.0204, "step": 7196 }, { "epoch": 0.4945711929631666, "grad_norm": 0.8506824374198914, "learning_rate": 1.599249840660737e-05, "loss": 0.0151, "step": 7197 }, { "epoch": 0.4946399120395822, "grad_norm": 0.8840373754501343, "learning_rate": 1.598916713004073e-05, "loss": 0.0187, "step": 7198 }, { "epoch": 0.4947086311159978, "grad_norm": 0.7764154672622681, "learning_rate": 1.5985835804472782e-05, "loss": 0.0175, "step": 7199 }, { "epoch": 0.4947773501924134, "grad_norm": 2.6126813888549805, "learning_rate": 1.5982504430068554e-05, "loss": 0.0364, "step": 7200 }, { "epoch": 0.494846069268829, "grad_norm": 0.6308138966560364, "learning_rate": 1.5979173006993077e-05, "loss": 0.0155, "step": 7201 }, { "epoch": 0.49491478834524466, "grad_norm": 0.7262395620346069, "learning_rate": 1.5975841535411385e-05, "loss": 0.015, "step": 7202 }, { "epoch": 0.49498350742166025, "grad_norm": 1.324209213256836, "learning_rate": 1.597251001548851e-05, "loss": 0.0194, "step": 7203 }, { "epoch": 0.49505222649807584, "grad_norm": 0.3268424868583679, "learning_rate": 1.5969178447389483e-05, "loss": 0.015, "step": 7204 }, { "epoch": 0.4951209455744915, "grad_norm": 0.8377305269241333, "learning_rate": 1.596584683127935e-05, "loss": 0.0233, "step": 7205 }, { "epoch": 0.4951896646509071, "grad_norm": 0.7872544527053833, "learning_rate": 1.5962515167323152e-05, "loss": 0.0173, "step": 7206 }, { "epoch": 0.4952583837273227, "grad_norm": 0.9847767949104309, "learning_rate": 1.595918345568593e-05, "loss": 0.0341, "step": 7207 }, { "epoch": 0.4953271028037383, "grad_norm": 0.8939310312271118, "learning_rate": 1.5955851696532736e-05, "loss": 0.0178, "step": 7208 }, { "epoch": 0.4953958218801539, "grad_norm": 0.592425525188446, "learning_rate": 1.595251989002861e-05, "loss": 0.03, "step": 7209 }, { "epoch": 0.49546454095656955, "grad_norm": 0.6972390413284302, "learning_rate": 1.5949188036338605e-05, "loss": 0.0167, "step": 7210 }, { "epoch": 0.49553326003298515, "grad_norm": 0.9623734354972839, "learning_rate": 1.5945856135627782e-05, "loss": 0.0223, "step": 7211 }, { "epoch": 0.4956019791094008, "grad_norm": 0.9977035522460938, "learning_rate": 1.5942524188061187e-05, "loss": 0.0186, "step": 7212 }, { "epoch": 0.4956706981858164, "grad_norm": 0.5900384783744812, "learning_rate": 1.5939192193803882e-05, "loss": 0.0165, "step": 7213 }, { "epoch": 0.495739417262232, "grad_norm": 0.2517099380493164, "learning_rate": 1.593586015302093e-05, "loss": 0.0257, "step": 7214 }, { "epoch": 0.4958081363386476, "grad_norm": 1.0875543355941772, "learning_rate": 1.5932528065877385e-05, "loss": 0.0325, "step": 7215 }, { "epoch": 0.4958768554150632, "grad_norm": 1.280861735343933, "learning_rate": 1.5929195932538317e-05, "loss": 0.0315, "step": 7216 }, { "epoch": 0.49594557449147886, "grad_norm": 0.7462977766990662, "learning_rate": 1.5925863753168796e-05, "loss": 0.0146, "step": 7217 }, { "epoch": 0.49601429356789445, "grad_norm": 0.6333251595497131, "learning_rate": 1.5922531527933888e-05, "loss": 0.018, "step": 7218 }, { "epoch": 0.49608301264431004, "grad_norm": 1.8360956907272339, "learning_rate": 1.5919199256998662e-05, "loss": 0.0256, "step": 7219 }, { "epoch": 0.4961517317207257, "grad_norm": 2.025785446166992, "learning_rate": 1.5915866940528198e-05, "loss": 0.0301, "step": 7220 }, { "epoch": 0.4962204507971413, "grad_norm": 0.32262635231018066, "learning_rate": 1.5912534578687567e-05, "loss": 0.0189, "step": 7221 }, { "epoch": 0.4962891698735569, "grad_norm": 0.5141324400901794, "learning_rate": 1.5909202171641847e-05, "loss": 0.027, "step": 7222 }, { "epoch": 0.4963578889499725, "grad_norm": 2.19427227973938, "learning_rate": 1.5905869719556123e-05, "loss": 0.0214, "step": 7223 }, { "epoch": 0.4964266080263881, "grad_norm": 0.6832473278045654, "learning_rate": 1.5902537222595478e-05, "loss": 0.0299, "step": 7224 }, { "epoch": 0.49649532710280375, "grad_norm": 0.5767144560813904, "learning_rate": 1.5899204680924988e-05, "loss": 0.0167, "step": 7225 }, { "epoch": 0.49656404617921934, "grad_norm": 0.7073248028755188, "learning_rate": 1.5895872094709754e-05, "loss": 0.0272, "step": 7226 }, { "epoch": 0.496632765255635, "grad_norm": 1.5157793760299683, "learning_rate": 1.589253946411485e-05, "loss": 0.0243, "step": 7227 }, { "epoch": 0.4967014843320506, "grad_norm": 0.6765666007995605, "learning_rate": 1.5889206789305382e-05, "loss": 0.0178, "step": 7228 }, { "epoch": 0.49677020340846617, "grad_norm": 3.211615562438965, "learning_rate": 1.5885874070446438e-05, "loss": 0.0291, "step": 7229 }, { "epoch": 0.4968389224848818, "grad_norm": 0.8529682755470276, "learning_rate": 1.588254130770311e-05, "loss": 0.0212, "step": 7230 }, { "epoch": 0.4969076415612974, "grad_norm": 1.0918769836425781, "learning_rate": 1.58792085012405e-05, "loss": 0.0219, "step": 7231 }, { "epoch": 0.49697636063771305, "grad_norm": 0.6151027679443359, "learning_rate": 1.587587565122371e-05, "loss": 0.0332, "step": 7232 }, { "epoch": 0.49704507971412865, "grad_norm": 0.8249033689498901, "learning_rate": 1.5872542757817843e-05, "loss": 0.0165, "step": 7233 }, { "epoch": 0.49711379879054424, "grad_norm": 0.5680723190307617, "learning_rate": 1.5869209821188002e-05, "loss": 0.0232, "step": 7234 }, { "epoch": 0.4971825178669599, "grad_norm": 1.0925487279891968, "learning_rate": 1.5865876841499292e-05, "loss": 0.0255, "step": 7235 }, { "epoch": 0.4972512369433755, "grad_norm": 0.7029844522476196, "learning_rate": 1.5862543818916827e-05, "loss": 0.0207, "step": 7236 }, { "epoch": 0.4973199560197911, "grad_norm": 0.4729362726211548, "learning_rate": 1.5859210753605707e-05, "loss": 0.0237, "step": 7237 }, { "epoch": 0.4973886750962067, "grad_norm": 0.8150231242179871, "learning_rate": 1.585587764573106e-05, "loss": 0.0237, "step": 7238 }, { "epoch": 0.4974573941726223, "grad_norm": 0.7016361951828003, "learning_rate": 1.585254449545799e-05, "loss": 0.0194, "step": 7239 }, { "epoch": 0.49752611324903795, "grad_norm": 0.8011831045150757, "learning_rate": 1.5849211302951617e-05, "loss": 0.0233, "step": 7240 }, { "epoch": 0.49759483232545354, "grad_norm": 0.9328784942626953, "learning_rate": 1.5845878068377067e-05, "loss": 0.0212, "step": 7241 }, { "epoch": 0.4976635514018692, "grad_norm": 0.5113463997840881, "learning_rate": 1.584254479189945e-05, "loss": 0.0208, "step": 7242 }, { "epoch": 0.4977322704782848, "grad_norm": 0.35928741097450256, "learning_rate": 1.58392114736839e-05, "loss": 0.0165, "step": 7243 }, { "epoch": 0.49780098955470037, "grad_norm": 0.5449796915054321, "learning_rate": 1.5835878113895538e-05, "loss": 0.0177, "step": 7244 }, { "epoch": 0.497869708631116, "grad_norm": 1.191433310508728, "learning_rate": 1.5832544712699493e-05, "loss": 0.0204, "step": 7245 }, { "epoch": 0.4979384277075316, "grad_norm": 0.9583770036697388, "learning_rate": 1.5829211270260896e-05, "loss": 0.0257, "step": 7246 }, { "epoch": 0.4980071467839472, "grad_norm": 0.6414771676063538, "learning_rate": 1.5825877786744873e-05, "loss": 0.0172, "step": 7247 }, { "epoch": 0.49807586586036284, "grad_norm": 0.866922914981842, "learning_rate": 1.582254426231657e-05, "loss": 0.029, "step": 7248 }, { "epoch": 0.49814458493677843, "grad_norm": 0.6963236927986145, "learning_rate": 1.581921069714111e-05, "loss": 0.0306, "step": 7249 }, { "epoch": 0.4982133040131941, "grad_norm": 0.46710875630378723, "learning_rate": 1.5815877091383636e-05, "loss": 0.0134, "step": 7250 }, { "epoch": 0.49828202308960967, "grad_norm": 0.4107604920864105, "learning_rate": 1.581254344520929e-05, "loss": 0.0269, "step": 7251 }, { "epoch": 0.49835074216602526, "grad_norm": 0.8996663093566895, "learning_rate": 1.5809209758783213e-05, "loss": 0.0223, "step": 7252 }, { "epoch": 0.4984194612424409, "grad_norm": 0.5223612785339355, "learning_rate": 1.5805876032270547e-05, "loss": 0.0161, "step": 7253 }, { "epoch": 0.4984881803188565, "grad_norm": 1.0309873819351196, "learning_rate": 1.5802542265836444e-05, "loss": 0.0294, "step": 7254 }, { "epoch": 0.49855689939527215, "grad_norm": 0.8675920963287354, "learning_rate": 1.579920845964604e-05, "loss": 0.0274, "step": 7255 }, { "epoch": 0.49862561847168774, "grad_norm": 0.7746303677558899, "learning_rate": 1.579587461386449e-05, "loss": 0.0212, "step": 7256 }, { "epoch": 0.49869433754810333, "grad_norm": 0.4943387508392334, "learning_rate": 1.5792540728656957e-05, "loss": 0.0198, "step": 7257 }, { "epoch": 0.498763056624519, "grad_norm": 0.5620134472846985, "learning_rate": 1.578920680418858e-05, "loss": 0.0201, "step": 7258 }, { "epoch": 0.49883177570093457, "grad_norm": 0.5681917071342468, "learning_rate": 1.5785872840624517e-05, "loss": 0.0222, "step": 7259 }, { "epoch": 0.4989004947773502, "grad_norm": 0.3735349178314209, "learning_rate": 1.5782538838129937e-05, "loss": 0.0278, "step": 7260 }, { "epoch": 0.4989692138537658, "grad_norm": 0.5121999382972717, "learning_rate": 1.5779204796869986e-05, "loss": 0.0228, "step": 7261 }, { "epoch": 0.4990379329301814, "grad_norm": 0.34302935004234314, "learning_rate": 1.5775870717009836e-05, "loss": 0.0136, "step": 7262 }, { "epoch": 0.49910665200659704, "grad_norm": 0.6728572845458984, "learning_rate": 1.5772536598714644e-05, "loss": 0.0237, "step": 7263 }, { "epoch": 0.49917537108301263, "grad_norm": 1.3060084581375122, "learning_rate": 1.5769202442149572e-05, "loss": 0.0283, "step": 7264 }, { "epoch": 0.4992440901594283, "grad_norm": 0.3084685206413269, "learning_rate": 1.5765868247479798e-05, "loss": 0.0105, "step": 7265 }, { "epoch": 0.49931280923584387, "grad_norm": 0.41319742798805237, "learning_rate": 1.5762534014870487e-05, "loss": 0.0135, "step": 7266 }, { "epoch": 0.49938152831225946, "grad_norm": 0.7658494710922241, "learning_rate": 1.5759199744486806e-05, "loss": 0.016, "step": 7267 }, { "epoch": 0.4994502473886751, "grad_norm": 0.6759844422340393, "learning_rate": 1.575586543649393e-05, "loss": 0.0194, "step": 7268 }, { "epoch": 0.4995189664650907, "grad_norm": 0.4220815598964691, "learning_rate": 1.5752531091057035e-05, "loss": 0.0156, "step": 7269 }, { "epoch": 0.49958768554150634, "grad_norm": 1.2763234376907349, "learning_rate": 1.5749196708341298e-05, "loss": 0.0249, "step": 7270 }, { "epoch": 0.49965640461792193, "grad_norm": 0.7819324135780334, "learning_rate": 1.5745862288511894e-05, "loss": 0.0184, "step": 7271 }, { "epoch": 0.4997251236943375, "grad_norm": 0.870124340057373, "learning_rate": 1.574252783173401e-05, "loss": 0.0267, "step": 7272 }, { "epoch": 0.49979384277075317, "grad_norm": 0.5863935351371765, "learning_rate": 1.573919333817282e-05, "loss": 0.0111, "step": 7273 }, { "epoch": 0.49986256184716876, "grad_norm": 0.7097975611686707, "learning_rate": 1.5735858807993512e-05, "loss": 0.018, "step": 7274 }, { "epoch": 0.4999312809235844, "grad_norm": 0.6881357431411743, "learning_rate": 1.5732524241361274e-05, "loss": 0.0124, "step": 7275 }, { "epoch": 0.5, "grad_norm": 0.7029273509979248, "learning_rate": 1.572918963844129e-05, "loss": 0.0209, "step": 7276 }, { "epoch": 0.5000687190764156, "grad_norm": 0.7684726119041443, "learning_rate": 1.5725854999398747e-05, "loss": 0.0242, "step": 7277 }, { "epoch": 0.5001374381528312, "grad_norm": 0.7067861557006836, "learning_rate": 1.5722520324398842e-05, "loss": 0.0127, "step": 7278 }, { "epoch": 0.5002061572292469, "grad_norm": 0.8725855946540833, "learning_rate": 1.5719185613606767e-05, "loss": 0.0287, "step": 7279 }, { "epoch": 0.5002748763056625, "grad_norm": 0.8638428449630737, "learning_rate": 1.5715850867187713e-05, "loss": 0.0236, "step": 7280 }, { "epoch": 0.5003435953820781, "grad_norm": 0.41563892364501953, "learning_rate": 1.571251608530688e-05, "loss": 0.0164, "step": 7281 }, { "epoch": 0.5004123144584937, "grad_norm": 0.40482857823371887, "learning_rate": 1.5709181268129465e-05, "loss": 0.0178, "step": 7282 }, { "epoch": 0.5004810335349092, "grad_norm": 0.6297497749328613, "learning_rate": 1.5705846415820665e-05, "loss": 0.0223, "step": 7283 }, { "epoch": 0.500549752611325, "grad_norm": 0.5932357907295227, "learning_rate": 1.5702511528545687e-05, "loss": 0.0297, "step": 7284 }, { "epoch": 0.5006184716877405, "grad_norm": 0.9874219298362732, "learning_rate": 1.5699176606469734e-05, "loss": 0.0244, "step": 7285 }, { "epoch": 0.5006871907641561, "grad_norm": 1.1175205707550049, "learning_rate": 1.5695841649758007e-05, "loss": 0.0272, "step": 7286 }, { "epoch": 0.5007559098405717, "grad_norm": 0.4721187651157379, "learning_rate": 1.569250665857572e-05, "loss": 0.0265, "step": 7287 }, { "epoch": 0.5008246289169873, "grad_norm": 0.4943926930427551, "learning_rate": 1.5689171633088074e-05, "loss": 0.022, "step": 7288 }, { "epoch": 0.500893347993403, "grad_norm": 0.986656129360199, "learning_rate": 1.568583657346028e-05, "loss": 0.0203, "step": 7289 }, { "epoch": 0.5009620670698186, "grad_norm": 0.6431788206100464, "learning_rate": 1.568250147985755e-05, "loss": 0.0202, "step": 7290 }, { "epoch": 0.5010307861462342, "grad_norm": 0.5728603005409241, "learning_rate": 1.5679166352445104e-05, "loss": 0.0133, "step": 7291 }, { "epoch": 0.5010995052226498, "grad_norm": 0.781996488571167, "learning_rate": 1.5675831191388153e-05, "loss": 0.0214, "step": 7292 }, { "epoch": 0.5011682242990654, "grad_norm": 1.1177603006362915, "learning_rate": 1.5672495996851916e-05, "loss": 0.02, "step": 7293 }, { "epoch": 0.5012369433754811, "grad_norm": 0.2075779139995575, "learning_rate": 1.566916076900161e-05, "loss": 0.0185, "step": 7294 }, { "epoch": 0.5013056624518967, "grad_norm": 0.9583064317703247, "learning_rate": 1.566582550800245e-05, "loss": 0.0193, "step": 7295 }, { "epoch": 0.5013743815283123, "grad_norm": 0.4764576852321625, "learning_rate": 1.566249021401967e-05, "loss": 0.0134, "step": 7296 }, { "epoch": 0.5014431006047279, "grad_norm": 0.7923205494880676, "learning_rate": 1.5659154887218482e-05, "loss": 0.016, "step": 7297 }, { "epoch": 0.5015118196811434, "grad_norm": 0.3080233335494995, "learning_rate": 1.5655819527764114e-05, "loss": 0.0158, "step": 7298 }, { "epoch": 0.5015805387575591, "grad_norm": 0.2745153307914734, "learning_rate": 1.5652484135821797e-05, "loss": 0.0203, "step": 7299 }, { "epoch": 0.5016492578339747, "grad_norm": 0.6409194469451904, "learning_rate": 1.564914871155676e-05, "loss": 0.015, "step": 7300 }, { "epoch": 0.5017179769103903, "grad_norm": 0.8887743353843689, "learning_rate": 1.564581325513423e-05, "loss": 0.0321, "step": 7301 }, { "epoch": 0.5017866959868059, "grad_norm": 0.46053433418273926, "learning_rate": 1.5642477766719437e-05, "loss": 0.0149, "step": 7302 }, { "epoch": 0.5018554150632215, "grad_norm": 1.1976779699325562, "learning_rate": 1.5639142246477618e-05, "loss": 0.0304, "step": 7303 }, { "epoch": 0.5019241341396372, "grad_norm": 0.3809315860271454, "learning_rate": 1.5635806694574e-05, "loss": 0.0113, "step": 7304 }, { "epoch": 0.5019928532160528, "grad_norm": 2.152700185775757, "learning_rate": 1.5632471111173834e-05, "loss": 0.0223, "step": 7305 }, { "epoch": 0.5020615722924684, "grad_norm": 0.9259541034698486, "learning_rate": 1.5629135496442346e-05, "loss": 0.0301, "step": 7306 }, { "epoch": 0.502130291368884, "grad_norm": 0.985745370388031, "learning_rate": 1.562579985054478e-05, "loss": 0.0204, "step": 7307 }, { "epoch": 0.5021990104452996, "grad_norm": 0.5891994833946228, "learning_rate": 1.5622464173646372e-05, "loss": 0.0175, "step": 7308 }, { "epoch": 0.5022677295217153, "grad_norm": 0.264594167470932, "learning_rate": 1.5619128465912374e-05, "loss": 0.0124, "step": 7309 }, { "epoch": 0.5023364485981309, "grad_norm": 0.3135203421115875, "learning_rate": 1.561579272750802e-05, "loss": 0.0212, "step": 7310 }, { "epoch": 0.5024051676745465, "grad_norm": 0.19872219860553741, "learning_rate": 1.5612456958598567e-05, "loss": 0.0089, "step": 7311 }, { "epoch": 0.502473886750962, "grad_norm": 0.5172043442726135, "learning_rate": 1.5609121159349252e-05, "loss": 0.0212, "step": 7312 }, { "epoch": 0.5025426058273776, "grad_norm": 1.468011736869812, "learning_rate": 1.5605785329925324e-05, "loss": 0.0233, "step": 7313 }, { "epoch": 0.5026113249037933, "grad_norm": 0.5105981230735779, "learning_rate": 1.560244947049204e-05, "loss": 0.0136, "step": 7314 }, { "epoch": 0.5026800439802089, "grad_norm": 0.2130676954984665, "learning_rate": 1.5599113581214645e-05, "loss": 0.0145, "step": 7315 }, { "epoch": 0.5027487630566245, "grad_norm": 0.4927222728729248, "learning_rate": 1.5595777662258393e-05, "loss": 0.0118, "step": 7316 }, { "epoch": 0.5028174821330401, "grad_norm": 0.9209648370742798, "learning_rate": 1.5592441713788548e-05, "loss": 0.02, "step": 7317 }, { "epoch": 0.5028862012094557, "grad_norm": 1.046216368675232, "learning_rate": 1.5589105735970356e-05, "loss": 0.0186, "step": 7318 }, { "epoch": 0.5029549202858714, "grad_norm": 0.9159384369850159, "learning_rate": 1.5585769728969068e-05, "loss": 0.0183, "step": 7319 }, { "epoch": 0.503023639362287, "grad_norm": 0.8695058822631836, "learning_rate": 1.5582433692949962e-05, "loss": 0.0173, "step": 7320 }, { "epoch": 0.5030923584387026, "grad_norm": 1.004364252090454, "learning_rate": 1.5579097628078284e-05, "loss": 0.0267, "step": 7321 }, { "epoch": 0.5031610775151182, "grad_norm": 0.5923169851303101, "learning_rate": 1.5575761534519302e-05, "loss": 0.0215, "step": 7322 }, { "epoch": 0.5032297965915338, "grad_norm": 0.950487494468689, "learning_rate": 1.5572425412438277e-05, "loss": 0.0138, "step": 7323 }, { "epoch": 0.5032985156679495, "grad_norm": 0.32161641120910645, "learning_rate": 1.5569089262000477e-05, "loss": 0.01, "step": 7324 }, { "epoch": 0.5033672347443651, "grad_norm": 0.3722822070121765, "learning_rate": 1.5565753083371157e-05, "loss": 0.0315, "step": 7325 }, { "epoch": 0.5034359538207807, "grad_norm": 1.1562978029251099, "learning_rate": 1.5562416876715596e-05, "loss": 0.0324, "step": 7326 }, { "epoch": 0.5035046728971962, "grad_norm": 0.4463619887828827, "learning_rate": 1.5559080642199063e-05, "loss": 0.017, "step": 7327 }, { "epoch": 0.5035733919736118, "grad_norm": 0.33552664518356323, "learning_rate": 1.555574437998682e-05, "loss": 0.0201, "step": 7328 }, { "epoch": 0.5036421110500275, "grad_norm": 0.35583314299583435, "learning_rate": 1.5552408090244146e-05, "loss": 0.0138, "step": 7329 }, { "epoch": 0.5037108301264431, "grad_norm": 0.5707559585571289, "learning_rate": 1.554907177313631e-05, "loss": 0.0246, "step": 7330 }, { "epoch": 0.5037795492028587, "grad_norm": 1.050150752067566, "learning_rate": 1.5545735428828585e-05, "loss": 0.0193, "step": 7331 }, { "epoch": 0.5038482682792743, "grad_norm": 1.1411479711532593, "learning_rate": 1.5542399057486252e-05, "loss": 0.032, "step": 7332 }, { "epoch": 0.5039169873556899, "grad_norm": 0.29639381170272827, "learning_rate": 1.5539062659274584e-05, "loss": 0.0173, "step": 7333 }, { "epoch": 0.5039857064321056, "grad_norm": 0.7025949954986572, "learning_rate": 1.5535726234358857e-05, "loss": 0.0246, "step": 7334 }, { "epoch": 0.5040544255085212, "grad_norm": 0.58307284116745, "learning_rate": 1.5532389782904358e-05, "loss": 0.0191, "step": 7335 }, { "epoch": 0.5041231445849368, "grad_norm": 0.4554502069950104, "learning_rate": 1.5529053305076363e-05, "loss": 0.0168, "step": 7336 }, { "epoch": 0.5041918636613524, "grad_norm": 1.378949761390686, "learning_rate": 1.5525716801040152e-05, "loss": 0.0241, "step": 7337 }, { "epoch": 0.504260582737768, "grad_norm": 0.9196227192878723, "learning_rate": 1.5522380270961013e-05, "loss": 0.0216, "step": 7338 }, { "epoch": 0.5043293018141837, "grad_norm": 0.861579179763794, "learning_rate": 1.551904371500423e-05, "loss": 0.0258, "step": 7339 }, { "epoch": 0.5043980208905993, "grad_norm": 1.6520494222640991, "learning_rate": 1.551570713333509e-05, "loss": 0.0197, "step": 7340 }, { "epoch": 0.5044667399670149, "grad_norm": 0.5698421001434326, "learning_rate": 1.5512370526118874e-05, "loss": 0.0127, "step": 7341 }, { "epoch": 0.5045354590434304, "grad_norm": 0.3606715500354767, "learning_rate": 1.550903389352088e-05, "loss": 0.0187, "step": 7342 }, { "epoch": 0.504604178119846, "grad_norm": 0.41841787099838257, "learning_rate": 1.5505697235706387e-05, "loss": 0.0147, "step": 7343 }, { "epoch": 0.5046728971962616, "grad_norm": 0.5420895218849182, "learning_rate": 1.55023605528407e-05, "loss": 0.0328, "step": 7344 }, { "epoch": 0.5047416162726773, "grad_norm": 0.7356393933296204, "learning_rate": 1.5499023845089106e-05, "loss": 0.0167, "step": 7345 }, { "epoch": 0.5048103353490929, "grad_norm": 0.7520543932914734, "learning_rate": 1.549568711261689e-05, "loss": 0.0197, "step": 7346 }, { "epoch": 0.5048790544255085, "grad_norm": 0.9888285994529724, "learning_rate": 1.549235035558936e-05, "loss": 0.0243, "step": 7347 }, { "epoch": 0.5049477735019241, "grad_norm": 0.5366081595420837, "learning_rate": 1.54890135741718e-05, "loss": 0.0113, "step": 7348 }, { "epoch": 0.5050164925783397, "grad_norm": 0.44648483395576477, "learning_rate": 1.548567676852951e-05, "loss": 0.0094, "step": 7349 }, { "epoch": 0.5050852116547554, "grad_norm": 0.524932861328125, "learning_rate": 1.5482339938827797e-05, "loss": 0.0243, "step": 7350 }, { "epoch": 0.505153930731171, "grad_norm": 0.6213334798812866, "learning_rate": 1.547900308523196e-05, "loss": 0.031, "step": 7351 }, { "epoch": 0.5052226498075866, "grad_norm": 0.7728666067123413, "learning_rate": 1.547566620790728e-05, "loss": 0.0271, "step": 7352 }, { "epoch": 0.5052913688840022, "grad_norm": 0.7586751580238342, "learning_rate": 1.547232930701909e-05, "loss": 0.0237, "step": 7353 }, { "epoch": 0.5053600879604178, "grad_norm": 0.8578246235847473, "learning_rate": 1.546899238273267e-05, "loss": 0.03, "step": 7354 }, { "epoch": 0.5054288070368335, "grad_norm": 0.7495983839035034, "learning_rate": 1.5465655435213325e-05, "loss": 0.0225, "step": 7355 }, { "epoch": 0.505497526113249, "grad_norm": 0.5553972721099854, "learning_rate": 1.5462318464626376e-05, "loss": 0.0145, "step": 7356 }, { "epoch": 0.5055662451896646, "grad_norm": 0.8040409088134766, "learning_rate": 1.545898147113712e-05, "loss": 0.0284, "step": 7357 }, { "epoch": 0.5056349642660802, "grad_norm": 0.7167613506317139, "learning_rate": 1.545564445491086e-05, "loss": 0.0264, "step": 7358 }, { "epoch": 0.5057036833424958, "grad_norm": 0.39707550406455994, "learning_rate": 1.5452307416112915e-05, "loss": 0.0237, "step": 7359 }, { "epoch": 0.5057724024189115, "grad_norm": 0.8196530342102051, "learning_rate": 1.5448970354908592e-05, "loss": 0.0245, "step": 7360 }, { "epoch": 0.5058411214953271, "grad_norm": 0.5204681158065796, "learning_rate": 1.5445633271463197e-05, "loss": 0.0263, "step": 7361 }, { "epoch": 0.5059098405717427, "grad_norm": 0.7894067764282227, "learning_rate": 1.544229616594205e-05, "loss": 0.0231, "step": 7362 }, { "epoch": 0.5059785596481583, "grad_norm": 0.6530416011810303, "learning_rate": 1.5438959038510455e-05, "loss": 0.0204, "step": 7363 }, { "epoch": 0.5060472787245739, "grad_norm": 0.9114983081817627, "learning_rate": 1.5435621889333735e-05, "loss": 0.0238, "step": 7364 }, { "epoch": 0.5061159978009896, "grad_norm": 0.8480647206306458, "learning_rate": 1.5432284718577203e-05, "loss": 0.0157, "step": 7365 }, { "epoch": 0.5061847168774052, "grad_norm": 0.7112551331520081, "learning_rate": 1.5428947526406175e-05, "loss": 0.0123, "step": 7366 }, { "epoch": 0.5062534359538208, "grad_norm": 0.6003066301345825, "learning_rate": 1.5425610312985966e-05, "loss": 0.0183, "step": 7367 }, { "epoch": 0.5063221550302364, "grad_norm": 0.6286323070526123, "learning_rate": 1.5422273078481898e-05, "loss": 0.0195, "step": 7368 }, { "epoch": 0.506390874106652, "grad_norm": 0.8419903516769409, "learning_rate": 1.541893582305929e-05, "loss": 0.0329, "step": 7369 }, { "epoch": 0.5064595931830677, "grad_norm": 1.0249011516571045, "learning_rate": 1.541559854688346e-05, "loss": 0.0384, "step": 7370 }, { "epoch": 0.5065283122594832, "grad_norm": 0.5245193243026733, "learning_rate": 1.5412261250119737e-05, "loss": 0.0168, "step": 7371 }, { "epoch": 0.5065970313358988, "grad_norm": 0.5745819211006165, "learning_rate": 1.5408923932933435e-05, "loss": 0.0185, "step": 7372 }, { "epoch": 0.5066657504123144, "grad_norm": 0.4210261106491089, "learning_rate": 1.5405586595489883e-05, "loss": 0.0142, "step": 7373 }, { "epoch": 0.50673446948873, "grad_norm": 0.6881447434425354, "learning_rate": 1.5402249237954408e-05, "loss": 0.0198, "step": 7374 }, { "epoch": 0.5068031885651457, "grad_norm": 0.3952907919883728, "learning_rate": 1.539891186049233e-05, "loss": 0.0183, "step": 7375 }, { "epoch": 0.5068719076415613, "grad_norm": 0.4657380282878876, "learning_rate": 1.5395574463268978e-05, "loss": 0.0164, "step": 7376 }, { "epoch": 0.5069406267179769, "grad_norm": 0.6533718109130859, "learning_rate": 1.5392237046449684e-05, "loss": 0.016, "step": 7377 }, { "epoch": 0.5070093457943925, "grad_norm": 0.46574538946151733, "learning_rate": 1.5388899610199772e-05, "loss": 0.0137, "step": 7378 }, { "epoch": 0.5070780648708081, "grad_norm": 1.1619240045547485, "learning_rate": 1.5385562154684567e-05, "loss": 0.0197, "step": 7379 }, { "epoch": 0.5071467839472238, "grad_norm": 0.2937425673007965, "learning_rate": 1.5382224680069413e-05, "loss": 0.0201, "step": 7380 }, { "epoch": 0.5072155030236394, "grad_norm": 0.6662659049034119, "learning_rate": 1.537888718651963e-05, "loss": 0.016, "step": 7381 }, { "epoch": 0.507284222100055, "grad_norm": 0.378167986869812, "learning_rate": 1.537554967420056e-05, "loss": 0.0208, "step": 7382 }, { "epoch": 0.5073529411764706, "grad_norm": 0.5473363995552063, "learning_rate": 1.5372212143277528e-05, "loss": 0.0129, "step": 7383 }, { "epoch": 0.5074216602528862, "grad_norm": 0.6308286786079407, "learning_rate": 1.536887459391587e-05, "loss": 0.0235, "step": 7384 }, { "epoch": 0.5074903793293019, "grad_norm": 0.7008001208305359, "learning_rate": 1.5365537026280924e-05, "loss": 0.0285, "step": 7385 }, { "epoch": 0.5075590984057174, "grad_norm": 1.238115906715393, "learning_rate": 1.5362199440538025e-05, "loss": 0.0331, "step": 7386 }, { "epoch": 0.507627817482133, "grad_norm": 0.6199084520339966, "learning_rate": 1.5358861836852512e-05, "loss": 0.021, "step": 7387 }, { "epoch": 0.5076965365585486, "grad_norm": 0.5579703450202942, "learning_rate": 1.535552421538972e-05, "loss": 0.0159, "step": 7388 }, { "epoch": 0.5077652556349642, "grad_norm": 0.47970592975616455, "learning_rate": 1.5352186576314987e-05, "loss": 0.0136, "step": 7389 }, { "epoch": 0.5078339747113799, "grad_norm": 0.6815887093544006, "learning_rate": 1.534884891979366e-05, "loss": 0.0296, "step": 7390 }, { "epoch": 0.5079026937877955, "grad_norm": 0.9300297498703003, "learning_rate": 1.5345511245991067e-05, "loss": 0.0207, "step": 7391 }, { "epoch": 0.5079714128642111, "grad_norm": 0.6039517521858215, "learning_rate": 1.5342173555072565e-05, "loss": 0.0236, "step": 7392 }, { "epoch": 0.5080401319406267, "grad_norm": 0.5755478739738464, "learning_rate": 1.5338835847203488e-05, "loss": 0.017, "step": 7393 }, { "epoch": 0.5081088510170423, "grad_norm": 0.7145316004753113, "learning_rate": 1.5335498122549174e-05, "loss": 0.0178, "step": 7394 }, { "epoch": 0.508177570093458, "grad_norm": 0.9381784796714783, "learning_rate": 1.5332160381274977e-05, "loss": 0.0252, "step": 7395 }, { "epoch": 0.5082462891698736, "grad_norm": 0.7680721879005432, "learning_rate": 1.532882262354624e-05, "loss": 0.0199, "step": 7396 }, { "epoch": 0.5083150082462892, "grad_norm": 0.6844429969787598, "learning_rate": 1.5325484849528297e-05, "loss": 0.022, "step": 7397 }, { "epoch": 0.5083837273227048, "grad_norm": 0.43583306670188904, "learning_rate": 1.5322147059386512e-05, "loss": 0.0277, "step": 7398 }, { "epoch": 0.5084524463991204, "grad_norm": 0.3655525743961334, "learning_rate": 1.5318809253286224e-05, "loss": 0.0115, "step": 7399 }, { "epoch": 0.508521165475536, "grad_norm": 0.5200888514518738, "learning_rate": 1.5315471431392775e-05, "loss": 0.0187, "step": 7400 }, { "epoch": 0.5085898845519516, "grad_norm": 0.4065651297569275, "learning_rate": 1.5312133593871525e-05, "loss": 0.0171, "step": 7401 }, { "epoch": 0.5086586036283672, "grad_norm": 0.5507276654243469, "learning_rate": 1.5308795740887818e-05, "loss": 0.0146, "step": 7402 }, { "epoch": 0.5087273227047828, "grad_norm": 0.7187018394470215, "learning_rate": 1.5305457872607007e-05, "loss": 0.021, "step": 7403 }, { "epoch": 0.5087960417811984, "grad_norm": 0.3289918601512909, "learning_rate": 1.5302119989194436e-05, "loss": 0.0166, "step": 7404 }, { "epoch": 0.5088647608576141, "grad_norm": 0.5402567982673645, "learning_rate": 1.5298782090815468e-05, "loss": 0.0262, "step": 7405 }, { "epoch": 0.5089334799340297, "grad_norm": 0.4385240077972412, "learning_rate": 1.5295444177635443e-05, "loss": 0.017, "step": 7406 }, { "epoch": 0.5090021990104453, "grad_norm": 0.5225775837898254, "learning_rate": 1.5292106249819725e-05, "loss": 0.0303, "step": 7407 }, { "epoch": 0.5090709180868609, "grad_norm": 0.6490488648414612, "learning_rate": 1.5288768307533662e-05, "loss": 0.0178, "step": 7408 }, { "epoch": 0.5091396371632765, "grad_norm": 0.6781805753707886, "learning_rate": 1.5285430350942606e-05, "loss": 0.0263, "step": 7409 }, { "epoch": 0.5092083562396922, "grad_norm": 0.7551093697547913, "learning_rate": 1.5282092380211926e-05, "loss": 0.0178, "step": 7410 }, { "epoch": 0.5092770753161078, "grad_norm": 1.0323781967163086, "learning_rate": 1.5278754395506965e-05, "loss": 0.027, "step": 7411 }, { "epoch": 0.5093457943925234, "grad_norm": 0.993118405342102, "learning_rate": 1.5275416396993087e-05, "loss": 0.0174, "step": 7412 }, { "epoch": 0.509414513468939, "grad_norm": 0.7416667342185974, "learning_rate": 1.5272078384835642e-05, "loss": 0.0124, "step": 7413 }, { "epoch": 0.5094832325453545, "grad_norm": 0.9467442035675049, "learning_rate": 1.5268740359199996e-05, "loss": 0.0202, "step": 7414 }, { "epoch": 0.5095519516217702, "grad_norm": 0.4080241024494171, "learning_rate": 1.5265402320251506e-05, "loss": 0.014, "step": 7415 }, { "epoch": 0.5096206706981858, "grad_norm": 0.6230353116989136, "learning_rate": 1.526206426815553e-05, "loss": 0.0195, "step": 7416 }, { "epoch": 0.5096893897746014, "grad_norm": 0.5210729837417603, "learning_rate": 1.5258726203077431e-05, "loss": 0.0164, "step": 7417 }, { "epoch": 0.509758108851017, "grad_norm": 0.36380934715270996, "learning_rate": 1.5255388125182562e-05, "loss": 0.0124, "step": 7418 }, { "epoch": 0.5098268279274326, "grad_norm": 1.0044798851013184, "learning_rate": 1.5252050034636296e-05, "loss": 0.0243, "step": 7419 }, { "epoch": 0.5098955470038483, "grad_norm": 0.9389017820358276, "learning_rate": 1.524871193160399e-05, "loss": 0.0143, "step": 7420 }, { "epoch": 0.5099642660802639, "grad_norm": 0.5588394403457642, "learning_rate": 1.5245373816251003e-05, "loss": 0.0214, "step": 7421 }, { "epoch": 0.5100329851566795, "grad_norm": 0.8049811124801636, "learning_rate": 1.5242035688742701e-05, "loss": 0.0264, "step": 7422 }, { "epoch": 0.5101017042330951, "grad_norm": 1.1521484851837158, "learning_rate": 1.5238697549244454e-05, "loss": 0.0201, "step": 7423 }, { "epoch": 0.5101704233095107, "grad_norm": 0.630156934261322, "learning_rate": 1.5235359397921614e-05, "loss": 0.025, "step": 7424 }, { "epoch": 0.5102391423859264, "grad_norm": 0.6099416613578796, "learning_rate": 1.5232021234939558e-05, "loss": 0.0179, "step": 7425 }, { "epoch": 0.510307861462342, "grad_norm": 0.23403812944889069, "learning_rate": 1.5228683060463647e-05, "loss": 0.0092, "step": 7426 }, { "epoch": 0.5103765805387576, "grad_norm": 0.816644012928009, "learning_rate": 1.5225344874659242e-05, "loss": 0.0167, "step": 7427 }, { "epoch": 0.5104452996151732, "grad_norm": 0.7644520401954651, "learning_rate": 1.5222006677691719e-05, "loss": 0.0077, "step": 7428 }, { "epoch": 0.5105140186915887, "grad_norm": 1.3075588941574097, "learning_rate": 1.5218668469726442e-05, "loss": 0.0327, "step": 7429 }, { "epoch": 0.5105827377680044, "grad_norm": 0.695414125919342, "learning_rate": 1.5215330250928778e-05, "loss": 0.0162, "step": 7430 }, { "epoch": 0.51065145684442, "grad_norm": 1.0737648010253906, "learning_rate": 1.5211992021464095e-05, "loss": 0.0296, "step": 7431 }, { "epoch": 0.5107201759208356, "grad_norm": 0.39155638217926025, "learning_rate": 1.5208653781497766e-05, "loss": 0.0208, "step": 7432 }, { "epoch": 0.5107888949972512, "grad_norm": 0.6302158236503601, "learning_rate": 1.5205315531195153e-05, "loss": 0.0173, "step": 7433 }, { "epoch": 0.5108576140736668, "grad_norm": 0.43852025270462036, "learning_rate": 1.5201977270721635e-05, "loss": 0.011, "step": 7434 }, { "epoch": 0.5109263331500825, "grad_norm": 0.44668522477149963, "learning_rate": 1.5198639000242577e-05, "loss": 0.024, "step": 7435 }, { "epoch": 0.5109950522264981, "grad_norm": 0.8523610830307007, "learning_rate": 1.5195300719923351e-05, "loss": 0.0151, "step": 7436 }, { "epoch": 0.5110637713029137, "grad_norm": 0.7467180490493774, "learning_rate": 1.5191962429929329e-05, "loss": 0.0154, "step": 7437 }, { "epoch": 0.5111324903793293, "grad_norm": 0.3578772246837616, "learning_rate": 1.5188624130425884e-05, "loss": 0.0273, "step": 7438 }, { "epoch": 0.5112012094557449, "grad_norm": 0.31761634349823, "learning_rate": 1.5185285821578388e-05, "loss": 0.0156, "step": 7439 }, { "epoch": 0.5112699285321606, "grad_norm": 1.4191662073135376, "learning_rate": 1.5181947503552212e-05, "loss": 0.0261, "step": 7440 }, { "epoch": 0.5113386476085762, "grad_norm": 0.57859206199646, "learning_rate": 1.517860917651273e-05, "loss": 0.016, "step": 7441 }, { "epoch": 0.5114073666849918, "grad_norm": 0.5810890793800354, "learning_rate": 1.5175270840625316e-05, "loss": 0.0326, "step": 7442 }, { "epoch": 0.5114760857614074, "grad_norm": 0.768435001373291, "learning_rate": 1.5171932496055347e-05, "loss": 0.0233, "step": 7443 }, { "epoch": 0.5115448048378229, "grad_norm": 0.47000303864479065, "learning_rate": 1.5168594142968197e-05, "loss": 0.0258, "step": 7444 }, { "epoch": 0.5116135239142386, "grad_norm": 0.6361526846885681, "learning_rate": 1.5165255781529232e-05, "loss": 0.0158, "step": 7445 }, { "epoch": 0.5116822429906542, "grad_norm": 0.4820300042629242, "learning_rate": 1.5161917411903845e-05, "loss": 0.0144, "step": 7446 }, { "epoch": 0.5117509620670698, "grad_norm": 0.7402446269989014, "learning_rate": 1.5158579034257399e-05, "loss": 0.0266, "step": 7447 }, { "epoch": 0.5118196811434854, "grad_norm": 0.500126838684082, "learning_rate": 1.515524064875527e-05, "loss": 0.0134, "step": 7448 }, { "epoch": 0.511888400219901, "grad_norm": 0.9199883937835693, "learning_rate": 1.5151902255562839e-05, "loss": 0.0248, "step": 7449 }, { "epoch": 0.5119571192963167, "grad_norm": 0.4571334719657898, "learning_rate": 1.514856385484549e-05, "loss": 0.0143, "step": 7450 }, { "epoch": 0.5120258383727323, "grad_norm": 0.7755193710327148, "learning_rate": 1.5145225446768582e-05, "loss": 0.0214, "step": 7451 }, { "epoch": 0.5120945574491479, "grad_norm": 0.7370942234992981, "learning_rate": 1.5141887031497512e-05, "loss": 0.0161, "step": 7452 }, { "epoch": 0.5121632765255635, "grad_norm": 1.1872450113296509, "learning_rate": 1.513854860919765e-05, "loss": 0.0173, "step": 7453 }, { "epoch": 0.5122319956019791, "grad_norm": 0.6802957653999329, "learning_rate": 1.5135210180034369e-05, "loss": 0.0181, "step": 7454 }, { "epoch": 0.5123007146783948, "grad_norm": 0.24915482103824615, "learning_rate": 1.5131871744173056e-05, "loss": 0.0123, "step": 7455 }, { "epoch": 0.5123694337548104, "grad_norm": 0.5653889775276184, "learning_rate": 1.512853330177909e-05, "loss": 0.0146, "step": 7456 }, { "epoch": 0.512438152831226, "grad_norm": 0.31798070669174194, "learning_rate": 1.512519485301785e-05, "loss": 0.0125, "step": 7457 }, { "epoch": 0.5125068719076415, "grad_norm": 0.7566570043563843, "learning_rate": 1.5121856398054712e-05, "loss": 0.0209, "step": 7458 }, { "epoch": 0.5125755909840571, "grad_norm": 0.5629528760910034, "learning_rate": 1.511851793705506e-05, "loss": 0.0224, "step": 7459 }, { "epoch": 0.5126443100604728, "grad_norm": 0.43931782245635986, "learning_rate": 1.5115179470184269e-05, "loss": 0.0239, "step": 7460 }, { "epoch": 0.5127130291368884, "grad_norm": 0.5007503628730774, "learning_rate": 1.5111840997607729e-05, "loss": 0.0245, "step": 7461 }, { "epoch": 0.512781748213304, "grad_norm": 0.7288240790367126, "learning_rate": 1.510850251949081e-05, "loss": 0.021, "step": 7462 }, { "epoch": 0.5128504672897196, "grad_norm": 0.6592163443565369, "learning_rate": 1.5105164035998902e-05, "loss": 0.0269, "step": 7463 }, { "epoch": 0.5129191863661352, "grad_norm": 1.1024775505065918, "learning_rate": 1.5101825547297386e-05, "loss": 0.0281, "step": 7464 }, { "epoch": 0.5129879054425509, "grad_norm": 0.5692812204360962, "learning_rate": 1.5098487053551643e-05, "loss": 0.02, "step": 7465 }, { "epoch": 0.5130566245189665, "grad_norm": 0.8227033019065857, "learning_rate": 1.5095148554927046e-05, "loss": 0.0228, "step": 7466 }, { "epoch": 0.5131253435953821, "grad_norm": 0.5691434144973755, "learning_rate": 1.5091810051588993e-05, "loss": 0.0173, "step": 7467 }, { "epoch": 0.5131940626717977, "grad_norm": 0.4231460392475128, "learning_rate": 1.5088471543702855e-05, "loss": 0.0161, "step": 7468 }, { "epoch": 0.5132627817482133, "grad_norm": 0.5076581835746765, "learning_rate": 1.508513303143402e-05, "loss": 0.0177, "step": 7469 }, { "epoch": 0.513331500824629, "grad_norm": 0.38464176654815674, "learning_rate": 1.5081794514947868e-05, "loss": 0.0205, "step": 7470 }, { "epoch": 0.5134002199010446, "grad_norm": 0.6611125469207764, "learning_rate": 1.5078455994409787e-05, "loss": 0.0163, "step": 7471 }, { "epoch": 0.5134689389774602, "grad_norm": 0.7649990320205688, "learning_rate": 1.5075117469985154e-05, "loss": 0.0194, "step": 7472 }, { "epoch": 0.5135376580538757, "grad_norm": 0.5496945381164551, "learning_rate": 1.5071778941839356e-05, "loss": 0.0266, "step": 7473 }, { "epoch": 0.5136063771302913, "grad_norm": 0.4798582196235657, "learning_rate": 1.5068440410137782e-05, "loss": 0.0253, "step": 7474 }, { "epoch": 0.5136750962067069, "grad_norm": 0.8841701745986938, "learning_rate": 1.5065101875045806e-05, "loss": 0.0188, "step": 7475 }, { "epoch": 0.5137438152831226, "grad_norm": 0.5608075857162476, "learning_rate": 1.506176333672882e-05, "loss": 0.0196, "step": 7476 }, { "epoch": 0.5138125343595382, "grad_norm": 0.7621526122093201, "learning_rate": 1.5058424795352205e-05, "loss": 0.0157, "step": 7477 }, { "epoch": 0.5138812534359538, "grad_norm": 1.3590549230575562, "learning_rate": 1.5055086251081341e-05, "loss": 0.0154, "step": 7478 }, { "epoch": 0.5139499725123694, "grad_norm": 0.6802467703819275, "learning_rate": 1.5051747704081621e-05, "loss": 0.016, "step": 7479 }, { "epoch": 0.514018691588785, "grad_norm": 0.292080819606781, "learning_rate": 1.5048409154518428e-05, "loss": 0.0242, "step": 7480 }, { "epoch": 0.5140874106652007, "grad_norm": 0.5524864792823792, "learning_rate": 1.504507060255714e-05, "loss": 0.0164, "step": 7481 }, { "epoch": 0.5141561297416163, "grad_norm": 0.8447694182395935, "learning_rate": 1.504173204836315e-05, "loss": 0.0287, "step": 7482 }, { "epoch": 0.5142248488180319, "grad_norm": 0.4564887285232544, "learning_rate": 1.5038393492101838e-05, "loss": 0.0269, "step": 7483 }, { "epoch": 0.5142935678944475, "grad_norm": 0.671941876411438, "learning_rate": 1.503505493393859e-05, "loss": 0.0177, "step": 7484 }, { "epoch": 0.5143622869708631, "grad_norm": 0.43976688385009766, "learning_rate": 1.50317163740388e-05, "loss": 0.0168, "step": 7485 }, { "epoch": 0.5144310060472788, "grad_norm": 0.513026773929596, "learning_rate": 1.5028377812567841e-05, "loss": 0.0288, "step": 7486 }, { "epoch": 0.5144997251236944, "grad_norm": 0.9037935733795166, "learning_rate": 1.5025039249691102e-05, "loss": 0.0221, "step": 7487 }, { "epoch": 0.5145684442001099, "grad_norm": 0.6680304408073425, "learning_rate": 1.5021700685573974e-05, "loss": 0.0171, "step": 7488 }, { "epoch": 0.5146371632765255, "grad_norm": 0.5393313765525818, "learning_rate": 1.5018362120381838e-05, "loss": 0.0148, "step": 7489 }, { "epoch": 0.5147058823529411, "grad_norm": 0.5753213763237, "learning_rate": 1.5015023554280076e-05, "loss": 0.0179, "step": 7490 }, { "epoch": 0.5147746014293568, "grad_norm": 0.7115525603294373, "learning_rate": 1.5011684987434083e-05, "loss": 0.022, "step": 7491 }, { "epoch": 0.5148433205057724, "grad_norm": 0.7464402914047241, "learning_rate": 1.5008346420009238e-05, "loss": 0.0185, "step": 7492 }, { "epoch": 0.514912039582188, "grad_norm": 0.6029486060142517, "learning_rate": 1.5005007852170929e-05, "loss": 0.0283, "step": 7493 }, { "epoch": 0.5149807586586036, "grad_norm": 1.0473500490188599, "learning_rate": 1.5001669284084543e-05, "loss": 0.0242, "step": 7494 }, { "epoch": 0.5150494777350192, "grad_norm": 0.601502537727356, "learning_rate": 1.4998330715915461e-05, "loss": 0.0235, "step": 7495 }, { "epoch": 0.5151181968114349, "grad_norm": 0.9956225156784058, "learning_rate": 1.4994992147829073e-05, "loss": 0.0211, "step": 7496 }, { "epoch": 0.5151869158878505, "grad_norm": 0.33013591170310974, "learning_rate": 1.4991653579990765e-05, "loss": 0.0171, "step": 7497 }, { "epoch": 0.5152556349642661, "grad_norm": 0.40935805439949036, "learning_rate": 1.4988315012565916e-05, "loss": 0.0142, "step": 7498 }, { "epoch": 0.5153243540406817, "grad_norm": 0.7383766174316406, "learning_rate": 1.4984976445719922e-05, "loss": 0.0262, "step": 7499 }, { "epoch": 0.5153930731170973, "grad_norm": 0.6516996026039124, "learning_rate": 1.4981637879618166e-05, "loss": 0.0171, "step": 7500 }, { "epoch": 0.515461792193513, "grad_norm": 0.6435000896453857, "learning_rate": 1.4978299314426029e-05, "loss": 0.0246, "step": 7501 }, { "epoch": 0.5155305112699285, "grad_norm": 0.3173366189002991, "learning_rate": 1.4974960750308898e-05, "loss": 0.0114, "step": 7502 }, { "epoch": 0.5155992303463441, "grad_norm": 0.6447160243988037, "learning_rate": 1.4971622187432165e-05, "loss": 0.0242, "step": 7503 }, { "epoch": 0.5156679494227597, "grad_norm": 0.655622661113739, "learning_rate": 1.4968283625961199e-05, "loss": 0.0186, "step": 7504 }, { "epoch": 0.5157366684991753, "grad_norm": 0.39139923453330994, "learning_rate": 1.4964945066061407e-05, "loss": 0.02, "step": 7505 }, { "epoch": 0.515805387575591, "grad_norm": 0.534816324710846, "learning_rate": 1.4961606507898162e-05, "loss": 0.028, "step": 7506 }, { "epoch": 0.5158741066520066, "grad_norm": 0.42062094807624817, "learning_rate": 1.4958267951636854e-05, "loss": 0.0216, "step": 7507 }, { "epoch": 0.5159428257284222, "grad_norm": 0.4308170676231384, "learning_rate": 1.4954929397442864e-05, "loss": 0.0199, "step": 7508 }, { "epoch": 0.5160115448048378, "grad_norm": 0.8903844356536865, "learning_rate": 1.495159084548158e-05, "loss": 0.0259, "step": 7509 }, { "epoch": 0.5160802638812534, "grad_norm": 0.5953107476234436, "learning_rate": 1.4948252295918381e-05, "loss": 0.0131, "step": 7510 }, { "epoch": 0.5161489829576691, "grad_norm": 0.7116311192512512, "learning_rate": 1.494491374891866e-05, "loss": 0.0122, "step": 7511 }, { "epoch": 0.5162177020340847, "grad_norm": 0.6654383540153503, "learning_rate": 1.49415752046478e-05, "loss": 0.0302, "step": 7512 }, { "epoch": 0.5162864211105003, "grad_norm": 0.46266070008277893, "learning_rate": 1.4938236663271183e-05, "loss": 0.0235, "step": 7513 }, { "epoch": 0.5163551401869159, "grad_norm": 0.43180426955223083, "learning_rate": 1.4934898124954196e-05, "loss": 0.0137, "step": 7514 }, { "epoch": 0.5164238592633315, "grad_norm": 0.9135953187942505, "learning_rate": 1.4931559589862222e-05, "loss": 0.0175, "step": 7515 }, { "epoch": 0.5164925783397472, "grad_norm": 0.7972155809402466, "learning_rate": 1.4928221058160641e-05, "loss": 0.0261, "step": 7516 }, { "epoch": 0.5165612974161627, "grad_norm": 0.5181431770324707, "learning_rate": 1.4924882530014847e-05, "loss": 0.0246, "step": 7517 }, { "epoch": 0.5166300164925783, "grad_norm": 0.20726744830608368, "learning_rate": 1.4921544005590216e-05, "loss": 0.0121, "step": 7518 }, { "epoch": 0.5166987355689939, "grad_norm": 0.5539357662200928, "learning_rate": 1.4918205485052132e-05, "loss": 0.0176, "step": 7519 }, { "epoch": 0.5167674546454095, "grad_norm": 1.1334959268569946, "learning_rate": 1.4914866968565984e-05, "loss": 0.0224, "step": 7520 }, { "epoch": 0.5168361737218252, "grad_norm": 0.738106369972229, "learning_rate": 1.4911528456297151e-05, "loss": 0.0155, "step": 7521 }, { "epoch": 0.5169048927982408, "grad_norm": 0.1442009061574936, "learning_rate": 1.4908189948411008e-05, "loss": 0.0149, "step": 7522 }, { "epoch": 0.5169736118746564, "grad_norm": 0.4745219051837921, "learning_rate": 1.4904851445072953e-05, "loss": 0.023, "step": 7523 }, { "epoch": 0.517042330951072, "grad_norm": 1.1630260944366455, "learning_rate": 1.4901512946448361e-05, "loss": 0.0239, "step": 7524 }, { "epoch": 0.5171110500274876, "grad_norm": 1.1322935819625854, "learning_rate": 1.4898174452702616e-05, "loss": 0.015, "step": 7525 }, { "epoch": 0.5171797691039033, "grad_norm": 0.579306423664093, "learning_rate": 1.48948359640011e-05, "loss": 0.022, "step": 7526 }, { "epoch": 0.5172484881803189, "grad_norm": 0.8286349773406982, "learning_rate": 1.4891497480509194e-05, "loss": 0.0238, "step": 7527 }, { "epoch": 0.5173172072567345, "grad_norm": 0.6744961142539978, "learning_rate": 1.4888159002392272e-05, "loss": 0.0327, "step": 7528 }, { "epoch": 0.5173859263331501, "grad_norm": 0.6541276574134827, "learning_rate": 1.488482052981573e-05, "loss": 0.0341, "step": 7529 }, { "epoch": 0.5174546454095657, "grad_norm": 0.5111405253410339, "learning_rate": 1.4881482062944942e-05, "loss": 0.02, "step": 7530 }, { "epoch": 0.5175233644859814, "grad_norm": 0.39608219265937805, "learning_rate": 1.487814360194529e-05, "loss": 0.0178, "step": 7531 }, { "epoch": 0.5175920835623969, "grad_norm": 0.6377713680267334, "learning_rate": 1.4874805146982153e-05, "loss": 0.0299, "step": 7532 }, { "epoch": 0.5176608026388125, "grad_norm": 0.45917126536369324, "learning_rate": 1.4871466698220912e-05, "loss": 0.017, "step": 7533 }, { "epoch": 0.5177295217152281, "grad_norm": 0.4943810999393463, "learning_rate": 1.4868128255826943e-05, "loss": 0.018, "step": 7534 }, { "epoch": 0.5177982407916437, "grad_norm": 0.5851079821586609, "learning_rate": 1.4864789819965632e-05, "loss": 0.0204, "step": 7535 }, { "epoch": 0.5178669598680594, "grad_norm": 0.7658225893974304, "learning_rate": 1.4861451390802354e-05, "loss": 0.0296, "step": 7536 }, { "epoch": 0.517935678944475, "grad_norm": 0.525958776473999, "learning_rate": 1.4858112968502492e-05, "loss": 0.017, "step": 7537 }, { "epoch": 0.5180043980208906, "grad_norm": 0.45383286476135254, "learning_rate": 1.4854774553231419e-05, "loss": 0.0148, "step": 7538 }, { "epoch": 0.5180731170973062, "grad_norm": 0.5097499489784241, "learning_rate": 1.4851436145154518e-05, "loss": 0.0157, "step": 7539 }, { "epoch": 0.5181418361737218, "grad_norm": 0.5062438249588013, "learning_rate": 1.4848097744437158e-05, "loss": 0.0158, "step": 7540 }, { "epoch": 0.5182105552501375, "grad_norm": 0.5153643488883972, "learning_rate": 1.4844759351244731e-05, "loss": 0.023, "step": 7541 }, { "epoch": 0.5182792743265531, "grad_norm": 0.44532346725463867, "learning_rate": 1.4841420965742605e-05, "loss": 0.0225, "step": 7542 }, { "epoch": 0.5183479934029687, "grad_norm": 0.6988255977630615, "learning_rate": 1.4838082588096158e-05, "loss": 0.0205, "step": 7543 }, { "epoch": 0.5184167124793843, "grad_norm": 0.6997541189193726, "learning_rate": 1.4834744218470769e-05, "loss": 0.0188, "step": 7544 }, { "epoch": 0.5184854315557998, "grad_norm": 0.4789767563343048, "learning_rate": 1.4831405857031809e-05, "loss": 0.0132, "step": 7545 }, { "epoch": 0.5185541506322155, "grad_norm": 0.2581000030040741, "learning_rate": 1.4828067503944652e-05, "loss": 0.0116, "step": 7546 }, { "epoch": 0.5186228697086311, "grad_norm": 0.40786558389663696, "learning_rate": 1.4824729159374682e-05, "loss": 0.016, "step": 7547 }, { "epoch": 0.5186915887850467, "grad_norm": 0.5017574429512024, "learning_rate": 1.4821390823487272e-05, "loss": 0.0081, "step": 7548 }, { "epoch": 0.5187603078614623, "grad_norm": 0.8934136033058167, "learning_rate": 1.4818052496447791e-05, "loss": 0.0209, "step": 7549 }, { "epoch": 0.5188290269378779, "grad_norm": 0.480999618768692, "learning_rate": 1.4814714178421616e-05, "loss": 0.0254, "step": 7550 }, { "epoch": 0.5188977460142936, "grad_norm": 0.5873470902442932, "learning_rate": 1.4811375869574119e-05, "loss": 0.0181, "step": 7551 }, { "epoch": 0.5189664650907092, "grad_norm": 0.6042795777320862, "learning_rate": 1.4808037570070669e-05, "loss": 0.0186, "step": 7552 }, { "epoch": 0.5190351841671248, "grad_norm": 0.5058210492134094, "learning_rate": 1.480469928007665e-05, "loss": 0.0235, "step": 7553 }, { "epoch": 0.5191039032435404, "grad_norm": 0.45276063680648804, "learning_rate": 1.4801360999757425e-05, "loss": 0.0149, "step": 7554 }, { "epoch": 0.519172622319956, "grad_norm": 0.9814419150352478, "learning_rate": 1.4798022729278367e-05, "loss": 0.0213, "step": 7555 }, { "epoch": 0.5192413413963717, "grad_norm": 0.32059675455093384, "learning_rate": 1.4794684468804847e-05, "loss": 0.0119, "step": 7556 }, { "epoch": 0.5193100604727873, "grad_norm": 0.5380330085754395, "learning_rate": 1.479134621850224e-05, "loss": 0.0158, "step": 7557 }, { "epoch": 0.5193787795492029, "grad_norm": 0.6043200492858887, "learning_rate": 1.4788007978535902e-05, "loss": 0.0115, "step": 7558 }, { "epoch": 0.5194474986256185, "grad_norm": 0.45402219891548157, "learning_rate": 1.4784669749071221e-05, "loss": 0.0143, "step": 7559 }, { "epoch": 0.519516217702034, "grad_norm": 0.8912310004234314, "learning_rate": 1.4781331530273559e-05, "loss": 0.0169, "step": 7560 }, { "epoch": 0.5195849367784497, "grad_norm": 0.24306130409240723, "learning_rate": 1.4777993322308283e-05, "loss": 0.0164, "step": 7561 }, { "epoch": 0.5196536558548653, "grad_norm": 0.9725320935249329, "learning_rate": 1.4774655125340762e-05, "loss": 0.0145, "step": 7562 }, { "epoch": 0.5197223749312809, "grad_norm": 1.5216501951217651, "learning_rate": 1.477131693953636e-05, "loss": 0.0353, "step": 7563 }, { "epoch": 0.5197910940076965, "grad_norm": 0.5697665810585022, "learning_rate": 1.4767978765060442e-05, "loss": 0.0196, "step": 7564 }, { "epoch": 0.5198598130841121, "grad_norm": 0.8413528800010681, "learning_rate": 1.4764640602078387e-05, "loss": 0.016, "step": 7565 }, { "epoch": 0.5199285321605278, "grad_norm": 0.4836428761482239, "learning_rate": 1.476130245075555e-05, "loss": 0.0229, "step": 7566 }, { "epoch": 0.5199972512369434, "grad_norm": 0.3043658435344696, "learning_rate": 1.47579643112573e-05, "loss": 0.0213, "step": 7567 }, { "epoch": 0.520065970313359, "grad_norm": 0.3859315514564514, "learning_rate": 1.4754626183749e-05, "loss": 0.0202, "step": 7568 }, { "epoch": 0.5201346893897746, "grad_norm": 0.41852307319641113, "learning_rate": 1.4751288068396014e-05, "loss": 0.0165, "step": 7569 }, { "epoch": 0.5202034084661902, "grad_norm": 0.4395546019077301, "learning_rate": 1.4747949965363705e-05, "loss": 0.0165, "step": 7570 }, { "epoch": 0.5202721275426059, "grad_norm": 0.6866450905799866, "learning_rate": 1.4744611874817437e-05, "loss": 0.0192, "step": 7571 }, { "epoch": 0.5203408466190215, "grad_norm": 0.38071557879447937, "learning_rate": 1.4741273796922573e-05, "loss": 0.0203, "step": 7572 }, { "epoch": 0.5204095656954371, "grad_norm": 1.0286637544631958, "learning_rate": 1.4737935731844472e-05, "loss": 0.0157, "step": 7573 }, { "epoch": 0.5204782847718527, "grad_norm": 0.4889727532863617, "learning_rate": 1.4734597679748499e-05, "loss": 0.0129, "step": 7574 }, { "epoch": 0.5205470038482682, "grad_norm": 1.0556998252868652, "learning_rate": 1.4731259640800007e-05, "loss": 0.0214, "step": 7575 }, { "epoch": 0.5206157229246839, "grad_norm": 0.7853670120239258, "learning_rate": 1.4727921615164357e-05, "loss": 0.0262, "step": 7576 }, { "epoch": 0.5206844420010995, "grad_norm": 1.0826152563095093, "learning_rate": 1.4724583603006916e-05, "loss": 0.0297, "step": 7577 }, { "epoch": 0.5207531610775151, "grad_norm": 0.583278238773346, "learning_rate": 1.4721245604493035e-05, "loss": 0.0215, "step": 7578 }, { "epoch": 0.5208218801539307, "grad_norm": 0.46708452701568604, "learning_rate": 1.4717907619788076e-05, "loss": 0.0144, "step": 7579 }, { "epoch": 0.5208905992303463, "grad_norm": 0.5542574524879456, "learning_rate": 1.4714569649057393e-05, "loss": 0.0183, "step": 7580 }, { "epoch": 0.520959318306762, "grad_norm": 0.8751152753829956, "learning_rate": 1.4711231692466343e-05, "loss": 0.0245, "step": 7581 }, { "epoch": 0.5210280373831776, "grad_norm": 1.0302335023880005, "learning_rate": 1.4707893750180276e-05, "loss": 0.0268, "step": 7582 }, { "epoch": 0.5210967564595932, "grad_norm": 0.8386752009391785, "learning_rate": 1.4704555822364558e-05, "loss": 0.0339, "step": 7583 }, { "epoch": 0.5211654755360088, "grad_norm": 0.47947749495506287, "learning_rate": 1.4701217909184534e-05, "loss": 0.0134, "step": 7584 }, { "epoch": 0.5212341946124244, "grad_norm": 0.6339591145515442, "learning_rate": 1.4697880010805563e-05, "loss": 0.0199, "step": 7585 }, { "epoch": 0.5213029136888401, "grad_norm": 0.733708381652832, "learning_rate": 1.4694542127392996e-05, "loss": 0.0261, "step": 7586 }, { "epoch": 0.5213716327652557, "grad_norm": 0.6541207432746887, "learning_rate": 1.4691204259112183e-05, "loss": 0.0171, "step": 7587 }, { "epoch": 0.5214403518416713, "grad_norm": 0.854131817817688, "learning_rate": 1.4687866406128474e-05, "loss": 0.019, "step": 7588 }, { "epoch": 0.5215090709180868, "grad_norm": 0.3537895977497101, "learning_rate": 1.4684528568607224e-05, "loss": 0.0129, "step": 7589 }, { "epoch": 0.5215777899945024, "grad_norm": 1.114561915397644, "learning_rate": 1.4681190746713779e-05, "loss": 0.0178, "step": 7590 }, { "epoch": 0.5216465090709181, "grad_norm": 0.9298565983772278, "learning_rate": 1.4677852940613489e-05, "loss": 0.0135, "step": 7591 }, { "epoch": 0.5217152281473337, "grad_norm": 0.5680503249168396, "learning_rate": 1.4674515150471704e-05, "loss": 0.0149, "step": 7592 }, { "epoch": 0.5217839472237493, "grad_norm": 0.5639333724975586, "learning_rate": 1.4671177376453768e-05, "loss": 0.0181, "step": 7593 }, { "epoch": 0.5218526663001649, "grad_norm": 0.8336085677146912, "learning_rate": 1.4667839618725022e-05, "loss": 0.02, "step": 7594 }, { "epoch": 0.5219213853765805, "grad_norm": 0.8143160939216614, "learning_rate": 1.4664501877450825e-05, "loss": 0.0151, "step": 7595 }, { "epoch": 0.5219901044529962, "grad_norm": 0.33051788806915283, "learning_rate": 1.4661164152796515e-05, "loss": 0.0107, "step": 7596 }, { "epoch": 0.5220588235294118, "grad_norm": 0.31980445981025696, "learning_rate": 1.4657826444927439e-05, "loss": 0.0134, "step": 7597 }, { "epoch": 0.5221275426058274, "grad_norm": 0.5280267000198364, "learning_rate": 1.4654488754008933e-05, "loss": 0.0171, "step": 7598 }, { "epoch": 0.522196261682243, "grad_norm": 0.35086071491241455, "learning_rate": 1.4651151080206348e-05, "loss": 0.017, "step": 7599 }, { "epoch": 0.5222649807586586, "grad_norm": 0.48146504163742065, "learning_rate": 1.464781342368501e-05, "loss": 0.0187, "step": 7600 }, { "epoch": 0.5223336998350742, "grad_norm": 0.658745527267456, "learning_rate": 1.4644475784610282e-05, "loss": 0.0164, "step": 7601 }, { "epoch": 0.5224024189114899, "grad_norm": 0.39595481753349304, "learning_rate": 1.464113816314749e-05, "loss": 0.0114, "step": 7602 }, { "epoch": 0.5224711379879055, "grad_norm": 0.4315882623195648, "learning_rate": 1.4637800559461978e-05, "loss": 0.0168, "step": 7603 }, { "epoch": 0.522539857064321, "grad_norm": 0.4881571829319, "learning_rate": 1.4634462973719079e-05, "loss": 0.0104, "step": 7604 }, { "epoch": 0.5226085761407366, "grad_norm": 0.5573217272758484, "learning_rate": 1.4631125406084135e-05, "loss": 0.0123, "step": 7605 }, { "epoch": 0.5226772952171522, "grad_norm": 0.5127530097961426, "learning_rate": 1.4627787856722474e-05, "loss": 0.0179, "step": 7606 }, { "epoch": 0.5227460142935679, "grad_norm": 1.0216089487075806, "learning_rate": 1.4624450325799445e-05, "loss": 0.0167, "step": 7607 }, { "epoch": 0.5228147333699835, "grad_norm": 0.38418257236480713, "learning_rate": 1.462111281348037e-05, "loss": 0.0123, "step": 7608 }, { "epoch": 0.5228834524463991, "grad_norm": 0.6525787711143494, "learning_rate": 1.461777531993059e-05, "loss": 0.0169, "step": 7609 }, { "epoch": 0.5229521715228147, "grad_norm": 0.7639651894569397, "learning_rate": 1.4614437845315434e-05, "loss": 0.0226, "step": 7610 }, { "epoch": 0.5230208905992303, "grad_norm": 0.7125699520111084, "learning_rate": 1.4611100389800234e-05, "loss": 0.0138, "step": 7611 }, { "epoch": 0.523089609675646, "grad_norm": 0.5297200679779053, "learning_rate": 1.4607762953550322e-05, "loss": 0.0138, "step": 7612 }, { "epoch": 0.5231583287520616, "grad_norm": 1.202640414237976, "learning_rate": 1.460442553673102e-05, "loss": 0.0242, "step": 7613 }, { "epoch": 0.5232270478284772, "grad_norm": 0.6942510008811951, "learning_rate": 1.460108813950767e-05, "loss": 0.0207, "step": 7614 }, { "epoch": 0.5232957669048928, "grad_norm": 0.5645651817321777, "learning_rate": 1.4597750762045594e-05, "loss": 0.0173, "step": 7615 }, { "epoch": 0.5233644859813084, "grad_norm": 0.5735659003257751, "learning_rate": 1.4594413404510117e-05, "loss": 0.0115, "step": 7616 }, { "epoch": 0.5234332050577241, "grad_norm": 0.6461012959480286, "learning_rate": 1.4591076067066567e-05, "loss": 0.0129, "step": 7617 }, { "epoch": 0.5235019241341397, "grad_norm": 0.7584621906280518, "learning_rate": 1.458773874988027e-05, "loss": 0.0169, "step": 7618 }, { "epoch": 0.5235706432105552, "grad_norm": 0.9643890857696533, "learning_rate": 1.458440145311654e-05, "loss": 0.0225, "step": 7619 }, { "epoch": 0.5236393622869708, "grad_norm": 0.4676927924156189, "learning_rate": 1.4581064176940712e-05, "loss": 0.0248, "step": 7620 }, { "epoch": 0.5237080813633864, "grad_norm": 0.9467179775238037, "learning_rate": 1.4577726921518105e-05, "loss": 0.0269, "step": 7621 }, { "epoch": 0.5237768004398021, "grad_norm": 0.3576768636703491, "learning_rate": 1.4574389687014038e-05, "loss": 0.0146, "step": 7622 }, { "epoch": 0.5238455195162177, "grad_norm": 0.5342651009559631, "learning_rate": 1.457105247359383e-05, "loss": 0.0163, "step": 7623 }, { "epoch": 0.5239142385926333, "grad_norm": 0.38850703835487366, "learning_rate": 1.4567715281422801e-05, "loss": 0.0148, "step": 7624 }, { "epoch": 0.5239829576690489, "grad_norm": 1.0091452598571777, "learning_rate": 1.4564378110666266e-05, "loss": 0.0184, "step": 7625 }, { "epoch": 0.5240516767454645, "grad_norm": 0.4667228162288666, "learning_rate": 1.4561040961489546e-05, "loss": 0.0182, "step": 7626 }, { "epoch": 0.5241203958218802, "grad_norm": 0.619541585445404, "learning_rate": 1.4557703834057955e-05, "loss": 0.0235, "step": 7627 }, { "epoch": 0.5241891148982958, "grad_norm": 1.1006344556808472, "learning_rate": 1.4554366728536807e-05, "loss": 0.0216, "step": 7628 }, { "epoch": 0.5242578339747114, "grad_norm": 0.8444328904151917, "learning_rate": 1.4551029645091413e-05, "loss": 0.0261, "step": 7629 }, { "epoch": 0.524326553051127, "grad_norm": 0.5232783555984497, "learning_rate": 1.454769258388709e-05, "loss": 0.0164, "step": 7630 }, { "epoch": 0.5243952721275426, "grad_norm": 1.035886287689209, "learning_rate": 1.4544355545089137e-05, "loss": 0.0222, "step": 7631 }, { "epoch": 0.5244639912039583, "grad_norm": 0.6439657211303711, "learning_rate": 1.4541018528862884e-05, "loss": 0.0118, "step": 7632 }, { "epoch": 0.5245327102803738, "grad_norm": 1.2041468620300293, "learning_rate": 1.4537681535373626e-05, "loss": 0.0216, "step": 7633 }, { "epoch": 0.5246014293567894, "grad_norm": 0.527309238910675, "learning_rate": 1.4534344564786676e-05, "loss": 0.0206, "step": 7634 }, { "epoch": 0.524670148433205, "grad_norm": 0.42794787883758545, "learning_rate": 1.4531007617267338e-05, "loss": 0.0143, "step": 7635 }, { "epoch": 0.5247388675096206, "grad_norm": 1.1311445236206055, "learning_rate": 1.4527670692980918e-05, "loss": 0.0184, "step": 7636 }, { "epoch": 0.5248075865860363, "grad_norm": 0.5674430131912231, "learning_rate": 1.4524333792092716e-05, "loss": 0.019, "step": 7637 }, { "epoch": 0.5248763056624519, "grad_norm": 0.4625653326511383, "learning_rate": 1.4520996914768047e-05, "loss": 0.0279, "step": 7638 }, { "epoch": 0.5249450247388675, "grad_norm": 0.9859650731086731, "learning_rate": 1.4517660061172203e-05, "loss": 0.0306, "step": 7639 }, { "epoch": 0.5250137438152831, "grad_norm": 0.77885901927948, "learning_rate": 1.451432323147049e-05, "loss": 0.0188, "step": 7640 }, { "epoch": 0.5250824628916987, "grad_norm": 0.8196964859962463, "learning_rate": 1.4510986425828207e-05, "loss": 0.024, "step": 7641 }, { "epoch": 0.5251511819681144, "grad_norm": 0.6006142497062683, "learning_rate": 1.4507649644410647e-05, "loss": 0.0204, "step": 7642 }, { "epoch": 0.52521990104453, "grad_norm": 1.0657306909561157, "learning_rate": 1.450431288738311e-05, "loss": 0.0176, "step": 7643 }, { "epoch": 0.5252886201209456, "grad_norm": 0.5075012445449829, "learning_rate": 1.45009761549109e-05, "loss": 0.0177, "step": 7644 }, { "epoch": 0.5253573391973612, "grad_norm": 1.2254321575164795, "learning_rate": 1.4497639447159301e-05, "loss": 0.0148, "step": 7645 }, { "epoch": 0.5254260582737768, "grad_norm": 0.43201959133148193, "learning_rate": 1.4494302764293612e-05, "loss": 0.0145, "step": 7646 }, { "epoch": 0.5254947773501925, "grad_norm": 0.4823366701602936, "learning_rate": 1.4490966106479125e-05, "loss": 0.0185, "step": 7647 }, { "epoch": 0.525563496426608, "grad_norm": 0.4879736304283142, "learning_rate": 1.4487629473881132e-05, "loss": 0.0226, "step": 7648 }, { "epoch": 0.5256322155030236, "grad_norm": 0.49998152256011963, "learning_rate": 1.4484292866664912e-05, "loss": 0.0248, "step": 7649 }, { "epoch": 0.5257009345794392, "grad_norm": 1.1598279476165771, "learning_rate": 1.448095628499577e-05, "loss": 0.0234, "step": 7650 }, { "epoch": 0.5257696536558548, "grad_norm": 0.5138770341873169, "learning_rate": 1.447761972903899e-05, "loss": 0.0116, "step": 7651 }, { "epoch": 0.5258383727322705, "grad_norm": 0.4820806086063385, "learning_rate": 1.4474283198959852e-05, "loss": 0.0224, "step": 7652 }, { "epoch": 0.5259070918086861, "grad_norm": 0.2460591048002243, "learning_rate": 1.4470946694923643e-05, "loss": 0.0075, "step": 7653 }, { "epoch": 0.5259758108851017, "grad_norm": 0.5466708540916443, "learning_rate": 1.4467610217095648e-05, "loss": 0.0197, "step": 7654 }, { "epoch": 0.5260445299615173, "grad_norm": 0.43357956409454346, "learning_rate": 1.446427376564114e-05, "loss": 0.0134, "step": 7655 }, { "epoch": 0.5261132490379329, "grad_norm": 0.46075817942619324, "learning_rate": 1.4460937340725417e-05, "loss": 0.0143, "step": 7656 }, { "epoch": 0.5261819681143486, "grad_norm": 1.1872060298919678, "learning_rate": 1.4457600942513749e-05, "loss": 0.017, "step": 7657 }, { "epoch": 0.5262506871907642, "grad_norm": 0.36771291494369507, "learning_rate": 1.4454264571171416e-05, "loss": 0.0196, "step": 7658 }, { "epoch": 0.5263194062671798, "grad_norm": 0.8283941149711609, "learning_rate": 1.4450928226863693e-05, "loss": 0.0172, "step": 7659 }, { "epoch": 0.5263881253435954, "grad_norm": 0.43650248646736145, "learning_rate": 1.444759190975586e-05, "loss": 0.0228, "step": 7660 }, { "epoch": 0.526456844420011, "grad_norm": 0.7604841589927673, "learning_rate": 1.4444255620013181e-05, "loss": 0.0144, "step": 7661 }, { "epoch": 0.5265255634964267, "grad_norm": 0.7899163961410522, "learning_rate": 1.444091935780094e-05, "loss": 0.0129, "step": 7662 }, { "epoch": 0.5265942825728422, "grad_norm": 0.6978721618652344, "learning_rate": 1.4437583123284404e-05, "loss": 0.0125, "step": 7663 }, { "epoch": 0.5266630016492578, "grad_norm": 0.6432932615280151, "learning_rate": 1.4434246916628846e-05, "loss": 0.016, "step": 7664 }, { "epoch": 0.5267317207256734, "grad_norm": 0.7592309713363647, "learning_rate": 1.4430910737999532e-05, "loss": 0.0245, "step": 7665 }, { "epoch": 0.526800439802089, "grad_norm": 0.37276652455329895, "learning_rate": 1.4427574587561729e-05, "loss": 0.021, "step": 7666 }, { "epoch": 0.5268691588785047, "grad_norm": 0.4133322536945343, "learning_rate": 1.4424238465480697e-05, "loss": 0.015, "step": 7667 }, { "epoch": 0.5269378779549203, "grad_norm": 0.4821239709854126, "learning_rate": 1.4420902371921716e-05, "loss": 0.0188, "step": 7668 }, { "epoch": 0.5270065970313359, "grad_norm": 0.6372405886650085, "learning_rate": 1.441756630705004e-05, "loss": 0.0127, "step": 7669 }, { "epoch": 0.5270753161077515, "grad_norm": 0.476060688495636, "learning_rate": 1.4414230271030933e-05, "loss": 0.0158, "step": 7670 }, { "epoch": 0.5271440351841671, "grad_norm": 0.7421441078186035, "learning_rate": 1.4410894264029653e-05, "loss": 0.0251, "step": 7671 }, { "epoch": 0.5272127542605828, "grad_norm": 0.829643189907074, "learning_rate": 1.440755828621146e-05, "loss": 0.0213, "step": 7672 }, { "epoch": 0.5272814733369984, "grad_norm": 0.738264799118042, "learning_rate": 1.4404222337741605e-05, "loss": 0.0244, "step": 7673 }, { "epoch": 0.527350192413414, "grad_norm": 0.26725298166275024, "learning_rate": 1.4400886418785356e-05, "loss": 0.0265, "step": 7674 }, { "epoch": 0.5274189114898296, "grad_norm": 0.41836121678352356, "learning_rate": 1.4397550529507965e-05, "loss": 0.0199, "step": 7675 }, { "epoch": 0.5274876305662451, "grad_norm": 0.6925947666168213, "learning_rate": 1.4394214670074678e-05, "loss": 0.0193, "step": 7676 }, { "epoch": 0.5275563496426608, "grad_norm": 0.6866093277931213, "learning_rate": 1.4390878840650753e-05, "loss": 0.0176, "step": 7677 }, { "epoch": 0.5276250687190764, "grad_norm": 0.42447444796562195, "learning_rate": 1.4387543041401439e-05, "loss": 0.0133, "step": 7678 }, { "epoch": 0.527693787795492, "grad_norm": 0.7286871075630188, "learning_rate": 1.438420727249198e-05, "loss": 0.023, "step": 7679 }, { "epoch": 0.5277625068719076, "grad_norm": 0.44360411167144775, "learning_rate": 1.4380871534087628e-05, "loss": 0.0162, "step": 7680 }, { "epoch": 0.5278312259483232, "grad_norm": 1.105909824371338, "learning_rate": 1.4377535826353629e-05, "loss": 0.0264, "step": 7681 }, { "epoch": 0.5278999450247389, "grad_norm": 0.8430176973342896, "learning_rate": 1.4374200149455225e-05, "loss": 0.0177, "step": 7682 }, { "epoch": 0.5279686641011545, "grad_norm": 0.5498858094215393, "learning_rate": 1.437086450355766e-05, "loss": 0.0155, "step": 7683 }, { "epoch": 0.5280373831775701, "grad_norm": 0.3223344683647156, "learning_rate": 1.4367528888826174e-05, "loss": 0.0157, "step": 7684 }, { "epoch": 0.5281061022539857, "grad_norm": 0.5193296670913696, "learning_rate": 1.4364193305425998e-05, "loss": 0.0208, "step": 7685 }, { "epoch": 0.5281748213304013, "grad_norm": 0.5276419520378113, "learning_rate": 1.4360857753522386e-05, "loss": 0.0122, "step": 7686 }, { "epoch": 0.528243540406817, "grad_norm": 0.4071061909198761, "learning_rate": 1.4357522233280568e-05, "loss": 0.0214, "step": 7687 }, { "epoch": 0.5283122594832326, "grad_norm": 0.639646589756012, "learning_rate": 1.4354186744865775e-05, "loss": 0.021, "step": 7688 }, { "epoch": 0.5283809785596482, "grad_norm": 0.4436134696006775, "learning_rate": 1.4350851288443246e-05, "loss": 0.016, "step": 7689 }, { "epoch": 0.5284496976360638, "grad_norm": 0.36768633127212524, "learning_rate": 1.434751586417821e-05, "loss": 0.0155, "step": 7690 }, { "epoch": 0.5285184167124793, "grad_norm": 0.50599205493927, "learning_rate": 1.4344180472235887e-05, "loss": 0.0251, "step": 7691 }, { "epoch": 0.528587135788895, "grad_norm": 0.41257867217063904, "learning_rate": 1.4340845112781522e-05, "loss": 0.0125, "step": 7692 }, { "epoch": 0.5286558548653106, "grad_norm": 0.5507121682167053, "learning_rate": 1.4337509785980336e-05, "loss": 0.0221, "step": 7693 }, { "epoch": 0.5287245739417262, "grad_norm": 0.6695601940155029, "learning_rate": 1.4334174491997554e-05, "loss": 0.0205, "step": 7694 }, { "epoch": 0.5287932930181418, "grad_norm": 0.29503533244132996, "learning_rate": 1.4330839230998395e-05, "loss": 0.0153, "step": 7695 }, { "epoch": 0.5288620120945574, "grad_norm": 0.5090546607971191, "learning_rate": 1.432750400314809e-05, "loss": 0.0186, "step": 7696 }, { "epoch": 0.5289307311709731, "grad_norm": 0.4741218388080597, "learning_rate": 1.4324168808611848e-05, "loss": 0.0198, "step": 7697 }, { "epoch": 0.5289994502473887, "grad_norm": 0.6250630021095276, "learning_rate": 1.4320833647554898e-05, "loss": 0.018, "step": 7698 }, { "epoch": 0.5290681693238043, "grad_norm": 0.6495524048805237, "learning_rate": 1.4317498520142453e-05, "loss": 0.0152, "step": 7699 }, { "epoch": 0.5291368884002199, "grad_norm": 0.3286147117614746, "learning_rate": 1.4314163426539727e-05, "loss": 0.0212, "step": 7700 }, { "epoch": 0.5292056074766355, "grad_norm": 0.8175265789031982, "learning_rate": 1.4310828366911932e-05, "loss": 0.0165, "step": 7701 }, { "epoch": 0.5292743265530512, "grad_norm": 0.364104688167572, "learning_rate": 1.4307493341424286e-05, "loss": 0.017, "step": 7702 }, { "epoch": 0.5293430456294668, "grad_norm": 0.8523363471031189, "learning_rate": 1.4304158350241992e-05, "loss": 0.0207, "step": 7703 }, { "epoch": 0.5294117647058824, "grad_norm": 0.4978674352169037, "learning_rate": 1.4300823393530267e-05, "loss": 0.0154, "step": 7704 }, { "epoch": 0.529480483782298, "grad_norm": 0.5770111083984375, "learning_rate": 1.4297488471454314e-05, "loss": 0.022, "step": 7705 }, { "epoch": 0.5295492028587135, "grad_norm": 1.1149271726608276, "learning_rate": 1.4294153584179338e-05, "loss": 0.0182, "step": 7706 }, { "epoch": 0.5296179219351292, "grad_norm": 0.5783482789993286, "learning_rate": 1.429081873187054e-05, "loss": 0.0212, "step": 7707 }, { "epoch": 0.5296866410115448, "grad_norm": 0.6338526010513306, "learning_rate": 1.4287483914693126e-05, "loss": 0.0203, "step": 7708 }, { "epoch": 0.5297553600879604, "grad_norm": 0.6193322539329529, "learning_rate": 1.4284149132812288e-05, "loss": 0.0166, "step": 7709 }, { "epoch": 0.529824079164376, "grad_norm": 0.5124592185020447, "learning_rate": 1.4280814386393234e-05, "loss": 0.0269, "step": 7710 }, { "epoch": 0.5298927982407916, "grad_norm": 0.7795156836509705, "learning_rate": 1.4277479675601159e-05, "loss": 0.011, "step": 7711 }, { "epoch": 0.5299615173172073, "grad_norm": 0.975982666015625, "learning_rate": 1.4274145000601257e-05, "loss": 0.0196, "step": 7712 }, { "epoch": 0.5300302363936229, "grad_norm": 0.4544611871242523, "learning_rate": 1.4270810361558717e-05, "loss": 0.0188, "step": 7713 }, { "epoch": 0.5300989554700385, "grad_norm": 0.8907821774482727, "learning_rate": 1.4267475758638732e-05, "loss": 0.0146, "step": 7714 }, { "epoch": 0.5301676745464541, "grad_norm": 0.6895731091499329, "learning_rate": 1.426414119200649e-05, "loss": 0.02, "step": 7715 }, { "epoch": 0.5302363936228697, "grad_norm": 0.6675856709480286, "learning_rate": 1.4260806661827183e-05, "loss": 0.0188, "step": 7716 }, { "epoch": 0.5303051126992854, "grad_norm": 0.9463203549385071, "learning_rate": 1.4257472168265992e-05, "loss": 0.0183, "step": 7717 }, { "epoch": 0.530373831775701, "grad_norm": 0.6571567058563232, "learning_rate": 1.4254137711488105e-05, "loss": 0.0362, "step": 7718 }, { "epoch": 0.5304425508521166, "grad_norm": 0.7386289238929749, "learning_rate": 1.4250803291658705e-05, "loss": 0.0176, "step": 7719 }, { "epoch": 0.5305112699285321, "grad_norm": 0.4539472758769989, "learning_rate": 1.4247468908942968e-05, "loss": 0.0139, "step": 7720 }, { "epoch": 0.5305799890049477, "grad_norm": 0.4545283317565918, "learning_rate": 1.424413456350607e-05, "loss": 0.0183, "step": 7721 }, { "epoch": 0.5306487080813634, "grad_norm": 0.4674491584300995, "learning_rate": 1.4240800255513195e-05, "loss": 0.0154, "step": 7722 }, { "epoch": 0.530717427157779, "grad_norm": 0.9978390336036682, "learning_rate": 1.4237465985129516e-05, "loss": 0.0169, "step": 7723 }, { "epoch": 0.5307861462341946, "grad_norm": 0.3120715916156769, "learning_rate": 1.4234131752520204e-05, "loss": 0.0281, "step": 7724 }, { "epoch": 0.5308548653106102, "grad_norm": 1.1266837120056152, "learning_rate": 1.423079755785043e-05, "loss": 0.0211, "step": 7725 }, { "epoch": 0.5309235843870258, "grad_norm": 0.6353321075439453, "learning_rate": 1.4227463401285363e-05, "loss": 0.0183, "step": 7726 }, { "epoch": 0.5309923034634415, "grad_norm": 0.7147700786590576, "learning_rate": 1.4224129282990165e-05, "loss": 0.0175, "step": 7727 }, { "epoch": 0.5310610225398571, "grad_norm": 0.8310998678207397, "learning_rate": 1.4220795203130014e-05, "loss": 0.0176, "step": 7728 }, { "epoch": 0.5311297416162727, "grad_norm": 0.7694449424743652, "learning_rate": 1.4217461161870065e-05, "loss": 0.0126, "step": 7729 }, { "epoch": 0.5311984606926883, "grad_norm": 0.6507358551025391, "learning_rate": 1.4214127159375484e-05, "loss": 0.0218, "step": 7730 }, { "epoch": 0.5312671797691039, "grad_norm": 0.8946315050125122, "learning_rate": 1.4210793195811426e-05, "loss": 0.0428, "step": 7731 }, { "epoch": 0.5313358988455195, "grad_norm": 1.0547932386398315, "learning_rate": 1.4207459271343051e-05, "loss": 0.0206, "step": 7732 }, { "epoch": 0.5314046179219352, "grad_norm": 1.0194629430770874, "learning_rate": 1.420412538613551e-05, "loss": 0.0227, "step": 7733 }, { "epoch": 0.5314733369983508, "grad_norm": 0.76162189245224, "learning_rate": 1.4200791540353962e-05, "loss": 0.0258, "step": 7734 }, { "epoch": 0.5315420560747663, "grad_norm": 0.5660056471824646, "learning_rate": 1.419745773416356e-05, "loss": 0.019, "step": 7735 }, { "epoch": 0.5316107751511819, "grad_norm": 0.44218093156814575, "learning_rate": 1.4194123967729455e-05, "loss": 0.0108, "step": 7736 }, { "epoch": 0.5316794942275975, "grad_norm": 1.521917700767517, "learning_rate": 1.419079024121679e-05, "loss": 0.0236, "step": 7737 }, { "epoch": 0.5317482133040132, "grad_norm": 0.5055264830589294, "learning_rate": 1.4187456554790713e-05, "loss": 0.0174, "step": 7738 }, { "epoch": 0.5318169323804288, "grad_norm": 0.6576114296913147, "learning_rate": 1.4184122908616363e-05, "loss": 0.018, "step": 7739 }, { "epoch": 0.5318856514568444, "grad_norm": 0.9149246215820312, "learning_rate": 1.4180789302858892e-05, "loss": 0.0197, "step": 7740 }, { "epoch": 0.53195437053326, "grad_norm": 0.3324533700942993, "learning_rate": 1.4177455737683435e-05, "loss": 0.0236, "step": 7741 }, { "epoch": 0.5320230896096756, "grad_norm": 0.536421000957489, "learning_rate": 1.4174122213255128e-05, "loss": 0.0246, "step": 7742 }, { "epoch": 0.5320918086860913, "grad_norm": 0.8142449855804443, "learning_rate": 1.417078872973911e-05, "loss": 0.0206, "step": 7743 }, { "epoch": 0.5321605277625069, "grad_norm": 0.35511067509651184, "learning_rate": 1.416745528730051e-05, "loss": 0.0122, "step": 7744 }, { "epoch": 0.5322292468389225, "grad_norm": 0.3632955849170685, "learning_rate": 1.4164121886104462e-05, "loss": 0.0177, "step": 7745 }, { "epoch": 0.5322979659153381, "grad_norm": 0.3547777831554413, "learning_rate": 1.4160788526316101e-05, "loss": 0.0179, "step": 7746 }, { "epoch": 0.5323666849917537, "grad_norm": 0.6086171865463257, "learning_rate": 1.4157455208100552e-05, "loss": 0.0149, "step": 7747 }, { "epoch": 0.5324354040681694, "grad_norm": 0.4966927170753479, "learning_rate": 1.4154121931622939e-05, "loss": 0.0165, "step": 7748 }, { "epoch": 0.532504123144585, "grad_norm": 0.5539606809616089, "learning_rate": 1.4150788697048387e-05, "loss": 0.0129, "step": 7749 }, { "epoch": 0.5325728422210005, "grad_norm": 0.5836220979690552, "learning_rate": 1.4147455504542016e-05, "loss": 0.0155, "step": 7750 }, { "epoch": 0.5326415612974161, "grad_norm": 0.6257336139678955, "learning_rate": 1.414412235426894e-05, "loss": 0.012, "step": 7751 }, { "epoch": 0.5327102803738317, "grad_norm": 1.4043411016464233, "learning_rate": 1.4140789246394292e-05, "loss": 0.0207, "step": 7752 }, { "epoch": 0.5327789994502474, "grad_norm": 0.7945519089698792, "learning_rate": 1.4137456181083177e-05, "loss": 0.0142, "step": 7753 }, { "epoch": 0.532847718526663, "grad_norm": 0.25020259618759155, "learning_rate": 1.4134123158500709e-05, "loss": 0.0106, "step": 7754 }, { "epoch": 0.5329164376030786, "grad_norm": 0.5227227807044983, "learning_rate": 1.4130790178811999e-05, "loss": 0.018, "step": 7755 }, { "epoch": 0.5329851566794942, "grad_norm": 0.6287157535552979, "learning_rate": 1.4127457242182157e-05, "loss": 0.0125, "step": 7756 }, { "epoch": 0.5330538757559098, "grad_norm": 0.38770952820777893, "learning_rate": 1.4124124348776287e-05, "loss": 0.0177, "step": 7757 }, { "epoch": 0.5331225948323255, "grad_norm": 0.7816397547721863, "learning_rate": 1.4120791498759497e-05, "loss": 0.017, "step": 7758 }, { "epoch": 0.5331913139087411, "grad_norm": 0.34177130460739136, "learning_rate": 1.411745869229689e-05, "loss": 0.013, "step": 7759 }, { "epoch": 0.5332600329851567, "grad_norm": 0.8096151947975159, "learning_rate": 1.4114125929553566e-05, "loss": 0.0162, "step": 7760 }, { "epoch": 0.5333287520615723, "grad_norm": 0.5021340250968933, "learning_rate": 1.4110793210694622e-05, "loss": 0.0097, "step": 7761 }, { "epoch": 0.5333974711379879, "grad_norm": 1.0585724115371704, "learning_rate": 1.4107460535885153e-05, "loss": 0.0134, "step": 7762 }, { "epoch": 0.5334661902144036, "grad_norm": 0.6438927054405212, "learning_rate": 1.4104127905290249e-05, "loss": 0.023, "step": 7763 }, { "epoch": 0.5335349092908191, "grad_norm": 0.7716566920280457, "learning_rate": 1.4100795319075013e-05, "loss": 0.0228, "step": 7764 }, { "epoch": 0.5336036283672347, "grad_norm": 1.1751197576522827, "learning_rate": 1.4097462777404526e-05, "loss": 0.0184, "step": 7765 }, { "epoch": 0.5336723474436503, "grad_norm": 0.6392040252685547, "learning_rate": 1.409413028044388e-05, "loss": 0.0174, "step": 7766 }, { "epoch": 0.5337410665200659, "grad_norm": 0.7518998384475708, "learning_rate": 1.4090797828358157e-05, "loss": 0.0104, "step": 7767 }, { "epoch": 0.5338097855964816, "grad_norm": 0.33671319484710693, "learning_rate": 1.4087465421312437e-05, "loss": 0.0164, "step": 7768 }, { "epoch": 0.5338785046728972, "grad_norm": 0.7167549729347229, "learning_rate": 1.4084133059471803e-05, "loss": 0.0198, "step": 7769 }, { "epoch": 0.5339472237493128, "grad_norm": 0.6044628620147705, "learning_rate": 1.4080800743001337e-05, "loss": 0.0356, "step": 7770 }, { "epoch": 0.5340159428257284, "grad_norm": 0.8867681622505188, "learning_rate": 1.4077468472066113e-05, "loss": 0.0225, "step": 7771 }, { "epoch": 0.534084661902144, "grad_norm": 0.6126719117164612, "learning_rate": 1.4074136246831204e-05, "loss": 0.0216, "step": 7772 }, { "epoch": 0.5341533809785597, "grad_norm": 0.5286262631416321, "learning_rate": 1.4070804067461683e-05, "loss": 0.0134, "step": 7773 }, { "epoch": 0.5342221000549753, "grad_norm": 0.24649442732334137, "learning_rate": 1.4067471934122618e-05, "loss": 0.0165, "step": 7774 }, { "epoch": 0.5342908191313909, "grad_norm": 0.5639219880104065, "learning_rate": 1.4064139846979072e-05, "loss": 0.0113, "step": 7775 }, { "epoch": 0.5343595382078065, "grad_norm": 0.41028594970703125, "learning_rate": 1.4060807806196118e-05, "loss": 0.0201, "step": 7776 }, { "epoch": 0.534428257284222, "grad_norm": 0.24824921786785126, "learning_rate": 1.4057475811938816e-05, "loss": 0.015, "step": 7777 }, { "epoch": 0.5344969763606378, "grad_norm": 0.48791924118995667, "learning_rate": 1.405414386437222e-05, "loss": 0.0233, "step": 7778 }, { "epoch": 0.5345656954370533, "grad_norm": 0.526194155216217, "learning_rate": 1.4050811963661396e-05, "loss": 0.0113, "step": 7779 }, { "epoch": 0.5346344145134689, "grad_norm": 0.5500326156616211, "learning_rate": 1.4047480109971396e-05, "loss": 0.0119, "step": 7780 }, { "epoch": 0.5347031335898845, "grad_norm": 0.48096123337745667, "learning_rate": 1.4044148303467266e-05, "loss": 0.0154, "step": 7781 }, { "epoch": 0.5347718526663001, "grad_norm": 0.433633953332901, "learning_rate": 1.4040816544314068e-05, "loss": 0.0177, "step": 7782 }, { "epoch": 0.5348405717427158, "grad_norm": 0.1567460298538208, "learning_rate": 1.4037484832676849e-05, "loss": 0.0131, "step": 7783 }, { "epoch": 0.5349092908191314, "grad_norm": 0.6378133893013, "learning_rate": 1.4034153168720652e-05, "loss": 0.0125, "step": 7784 }, { "epoch": 0.534978009895547, "grad_norm": 0.719546377658844, "learning_rate": 1.4030821552610523e-05, "loss": 0.017, "step": 7785 }, { "epoch": 0.5350467289719626, "grad_norm": 0.3383461833000183, "learning_rate": 1.40274899845115e-05, "loss": 0.0217, "step": 7786 }, { "epoch": 0.5351154480483782, "grad_norm": 0.5558542013168335, "learning_rate": 1.4024158464588615e-05, "loss": 0.0222, "step": 7787 }, { "epoch": 0.5351841671247939, "grad_norm": 0.4484025239944458, "learning_rate": 1.4020826993006923e-05, "loss": 0.0198, "step": 7788 }, { "epoch": 0.5352528862012095, "grad_norm": 0.2574426233768463, "learning_rate": 1.4017495569931447e-05, "loss": 0.0173, "step": 7789 }, { "epoch": 0.5353216052776251, "grad_norm": 0.2876730263233185, "learning_rate": 1.4014164195527222e-05, "loss": 0.0146, "step": 7790 }, { "epoch": 0.5353903243540407, "grad_norm": 0.8661143779754639, "learning_rate": 1.4010832869959275e-05, "loss": 0.0299, "step": 7791 }, { "epoch": 0.5354590434304562, "grad_norm": 0.36584603786468506, "learning_rate": 1.4007501593392635e-05, "loss": 0.0127, "step": 7792 }, { "epoch": 0.535527762506872, "grad_norm": 0.673799991607666, "learning_rate": 1.400417036599232e-05, "loss": 0.0187, "step": 7793 }, { "epoch": 0.5355964815832875, "grad_norm": 0.5875697135925293, "learning_rate": 1.4000839187923363e-05, "loss": 0.018, "step": 7794 }, { "epoch": 0.5356652006597031, "grad_norm": 0.5808505415916443, "learning_rate": 1.3997508059350778e-05, "loss": 0.0182, "step": 7795 }, { "epoch": 0.5357339197361187, "grad_norm": 0.6288681030273438, "learning_rate": 1.3994176980439583e-05, "loss": 0.0216, "step": 7796 }, { "epoch": 0.5358026388125343, "grad_norm": 0.5735710859298706, "learning_rate": 1.3990845951354792e-05, "loss": 0.0158, "step": 7797 }, { "epoch": 0.53587135788895, "grad_norm": 0.8573136329650879, "learning_rate": 1.398751497226142e-05, "loss": 0.0139, "step": 7798 }, { "epoch": 0.5359400769653656, "grad_norm": 0.4521178603172302, "learning_rate": 1.3984184043324465e-05, "loss": 0.0195, "step": 7799 }, { "epoch": 0.5360087960417812, "grad_norm": 0.6246184706687927, "learning_rate": 1.3980853164708952e-05, "loss": 0.0155, "step": 7800 }, { "epoch": 0.5360775151181968, "grad_norm": 0.911841630935669, "learning_rate": 1.3977522336579876e-05, "loss": 0.0157, "step": 7801 }, { "epoch": 0.5361462341946124, "grad_norm": 1.501779556274414, "learning_rate": 1.3974191559102243e-05, "loss": 0.0223, "step": 7802 }, { "epoch": 0.5362149532710281, "grad_norm": 0.2153538167476654, "learning_rate": 1.3970860832441052e-05, "loss": 0.012, "step": 7803 }, { "epoch": 0.5362836723474437, "grad_norm": 0.6416720151901245, "learning_rate": 1.3967530156761299e-05, "loss": 0.0226, "step": 7804 }, { "epoch": 0.5363523914238593, "grad_norm": 0.7670729160308838, "learning_rate": 1.3964199532227973e-05, "loss": 0.017, "step": 7805 }, { "epoch": 0.5364211105002749, "grad_norm": 0.4941485524177551, "learning_rate": 1.396086895900608e-05, "loss": 0.0233, "step": 7806 }, { "epoch": 0.5364898295766904, "grad_norm": 0.26862359046936035, "learning_rate": 1.39575384372606e-05, "loss": 0.0171, "step": 7807 }, { "epoch": 0.5365585486531061, "grad_norm": 0.4966219365596771, "learning_rate": 1.3954207967156522e-05, "loss": 0.0164, "step": 7808 }, { "epoch": 0.5366272677295217, "grad_norm": 0.8726699352264404, "learning_rate": 1.3950877548858835e-05, "loss": 0.0247, "step": 7809 }, { "epoch": 0.5366959868059373, "grad_norm": 0.4277726113796234, "learning_rate": 1.3947547182532515e-05, "loss": 0.0185, "step": 7810 }, { "epoch": 0.5367647058823529, "grad_norm": 0.48923051357269287, "learning_rate": 1.3944216868342541e-05, "loss": 0.0225, "step": 7811 }, { "epoch": 0.5368334249587685, "grad_norm": 0.39133110642433167, "learning_rate": 1.3940886606453897e-05, "loss": 0.022, "step": 7812 }, { "epoch": 0.5369021440351842, "grad_norm": 0.8793784379959106, "learning_rate": 1.393755639703155e-05, "loss": 0.024, "step": 7813 }, { "epoch": 0.5369708631115998, "grad_norm": 0.2865051329135895, "learning_rate": 1.3934226240240478e-05, "loss": 0.015, "step": 7814 }, { "epoch": 0.5370395821880154, "grad_norm": 0.8353723883628845, "learning_rate": 1.3930896136245646e-05, "loss": 0.0181, "step": 7815 }, { "epoch": 0.537108301264431, "grad_norm": 0.5431417226791382, "learning_rate": 1.3927566085212022e-05, "loss": 0.0137, "step": 7816 }, { "epoch": 0.5371770203408466, "grad_norm": 0.7064096331596375, "learning_rate": 1.3924236087304561e-05, "loss": 0.0277, "step": 7817 }, { "epoch": 0.5372457394172623, "grad_norm": 0.34991708397865295, "learning_rate": 1.392090614268824e-05, "loss": 0.0147, "step": 7818 }, { "epoch": 0.5373144584936779, "grad_norm": 0.8864082098007202, "learning_rate": 1.3917576251528009e-05, "loss": 0.0254, "step": 7819 }, { "epoch": 0.5373831775700935, "grad_norm": 0.5157490372657776, "learning_rate": 1.3914246413988827e-05, "loss": 0.015, "step": 7820 }, { "epoch": 0.537451896646509, "grad_norm": 0.6918784976005554, "learning_rate": 1.3910916630235645e-05, "loss": 0.0199, "step": 7821 }, { "epoch": 0.5375206157229246, "grad_norm": 0.6906389594078064, "learning_rate": 1.3907586900433415e-05, "loss": 0.0139, "step": 7822 }, { "epoch": 0.5375893347993403, "grad_norm": 0.6636772155761719, "learning_rate": 1.3904257224747074e-05, "loss": 0.0368, "step": 7823 }, { "epoch": 0.5376580538757559, "grad_norm": 0.6846679449081421, "learning_rate": 1.3900927603341589e-05, "loss": 0.0233, "step": 7824 }, { "epoch": 0.5377267729521715, "grad_norm": 0.8774909377098083, "learning_rate": 1.3897598036381888e-05, "loss": 0.019, "step": 7825 }, { "epoch": 0.5377954920285871, "grad_norm": 0.9153432250022888, "learning_rate": 1.3894268524032914e-05, "loss": 0.0204, "step": 7826 }, { "epoch": 0.5378642111050027, "grad_norm": 0.44468003511428833, "learning_rate": 1.3890939066459602e-05, "loss": 0.0178, "step": 7827 }, { "epoch": 0.5379329301814184, "grad_norm": 0.3847458064556122, "learning_rate": 1.3887609663826892e-05, "loss": 0.0177, "step": 7828 }, { "epoch": 0.538001649257834, "grad_norm": 0.6048357486724854, "learning_rate": 1.3884280316299708e-05, "loss": 0.0181, "step": 7829 }, { "epoch": 0.5380703683342496, "grad_norm": 0.46103915572166443, "learning_rate": 1.3880951024042985e-05, "loss": 0.0157, "step": 7830 }, { "epoch": 0.5381390874106652, "grad_norm": 0.4502897560596466, "learning_rate": 1.387762178722165e-05, "loss": 0.0158, "step": 7831 }, { "epoch": 0.5382078064870808, "grad_norm": 0.7347304224967957, "learning_rate": 1.3874292606000622e-05, "loss": 0.0322, "step": 7832 }, { "epoch": 0.5382765255634965, "grad_norm": 0.5393787026405334, "learning_rate": 1.3870963480544825e-05, "loss": 0.0239, "step": 7833 }, { "epoch": 0.5383452446399121, "grad_norm": 0.535487949848175, "learning_rate": 1.3867634411019177e-05, "loss": 0.0191, "step": 7834 }, { "epoch": 0.5384139637163277, "grad_norm": 0.8548043370246887, "learning_rate": 1.3864305397588591e-05, "loss": 0.0203, "step": 7835 }, { "epoch": 0.5384826827927433, "grad_norm": 0.8151935338973999, "learning_rate": 1.3860976440417974e-05, "loss": 0.0179, "step": 7836 }, { "epoch": 0.5385514018691588, "grad_norm": 0.8348183035850525, "learning_rate": 1.385764753967225e-05, "loss": 0.0272, "step": 7837 }, { "epoch": 0.5386201209455745, "grad_norm": 0.7705233097076416, "learning_rate": 1.3854318695516315e-05, "loss": 0.0142, "step": 7838 }, { "epoch": 0.5386888400219901, "grad_norm": 0.4167638421058655, "learning_rate": 1.385098990811508e-05, "loss": 0.0169, "step": 7839 }, { "epoch": 0.5387575590984057, "grad_norm": 0.6726976633071899, "learning_rate": 1.384766117763344e-05, "loss": 0.0156, "step": 7840 }, { "epoch": 0.5388262781748213, "grad_norm": 0.39397361874580383, "learning_rate": 1.38443325042363e-05, "loss": 0.0194, "step": 7841 }, { "epoch": 0.5388949972512369, "grad_norm": 0.35677653551101685, "learning_rate": 1.3841003888088543e-05, "loss": 0.016, "step": 7842 }, { "epoch": 0.5389637163276526, "grad_norm": 0.6225786209106445, "learning_rate": 1.3837675329355076e-05, "loss": 0.0231, "step": 7843 }, { "epoch": 0.5390324354040682, "grad_norm": 0.42383691668510437, "learning_rate": 1.3834346828200784e-05, "loss": 0.0159, "step": 7844 }, { "epoch": 0.5391011544804838, "grad_norm": 0.6689095497131348, "learning_rate": 1.3831018384790555e-05, "loss": 0.0166, "step": 7845 }, { "epoch": 0.5391698735568994, "grad_norm": 0.6193498373031616, "learning_rate": 1.382768999928927e-05, "loss": 0.0194, "step": 7846 }, { "epoch": 0.539238592633315, "grad_norm": 0.8740642070770264, "learning_rate": 1.3824361671861815e-05, "loss": 0.0167, "step": 7847 }, { "epoch": 0.5393073117097307, "grad_norm": 0.700461745262146, "learning_rate": 1.3821033402673062e-05, "loss": 0.0146, "step": 7848 }, { "epoch": 0.5393760307861463, "grad_norm": 0.31939855217933655, "learning_rate": 1.3817705191887891e-05, "loss": 0.0155, "step": 7849 }, { "epoch": 0.5394447498625619, "grad_norm": 1.2419553995132446, "learning_rate": 1.3814377039671176e-05, "loss": 0.0147, "step": 7850 }, { "epoch": 0.5395134689389774, "grad_norm": 0.843576192855835, "learning_rate": 1.3811048946187785e-05, "loss": 0.0141, "step": 7851 }, { "epoch": 0.539582188015393, "grad_norm": 0.46267497539520264, "learning_rate": 1.3807720911602584e-05, "loss": 0.0171, "step": 7852 }, { "epoch": 0.5396509070918087, "grad_norm": 0.5797869563102722, "learning_rate": 1.380439293608044e-05, "loss": 0.0148, "step": 7853 }, { "epoch": 0.5397196261682243, "grad_norm": 0.9439525604248047, "learning_rate": 1.3801065019786203e-05, "loss": 0.0228, "step": 7854 }, { "epoch": 0.5397883452446399, "grad_norm": 1.2987511157989502, "learning_rate": 1.3797737162884745e-05, "loss": 0.0235, "step": 7855 }, { "epoch": 0.5398570643210555, "grad_norm": 0.5470668077468872, "learning_rate": 1.3794409365540918e-05, "loss": 0.0134, "step": 7856 }, { "epoch": 0.5399257833974711, "grad_norm": 0.8656572103500366, "learning_rate": 1.3791081627919573e-05, "loss": 0.0216, "step": 7857 }, { "epoch": 0.5399945024738868, "grad_norm": 1.1317319869995117, "learning_rate": 1.378775395018556e-05, "loss": 0.0261, "step": 7858 }, { "epoch": 0.5400632215503024, "grad_norm": 0.38030391931533813, "learning_rate": 1.3784426332503723e-05, "loss": 0.0197, "step": 7859 }, { "epoch": 0.540131940626718, "grad_norm": 0.7116236090660095, "learning_rate": 1.37810987750389e-05, "loss": 0.0144, "step": 7860 }, { "epoch": 0.5402006597031336, "grad_norm": 0.5419328808784485, "learning_rate": 1.3777771277955943e-05, "loss": 0.0226, "step": 7861 }, { "epoch": 0.5402693787795492, "grad_norm": 0.8739895224571228, "learning_rate": 1.3774443841419684e-05, "loss": 0.0265, "step": 7862 }, { "epoch": 0.5403380978559648, "grad_norm": 0.6727621555328369, "learning_rate": 1.377111646559496e-05, "loss": 0.0154, "step": 7863 }, { "epoch": 0.5404068169323805, "grad_norm": 0.36720970273017883, "learning_rate": 1.3767789150646596e-05, "loss": 0.0114, "step": 7864 }, { "epoch": 0.540475536008796, "grad_norm": 0.5294450521469116, "learning_rate": 1.3764461896739426e-05, "loss": 0.0111, "step": 7865 }, { "epoch": 0.5405442550852116, "grad_norm": 0.7403645515441895, "learning_rate": 1.376113470403827e-05, "loss": 0.0222, "step": 7866 }, { "epoch": 0.5406129741616272, "grad_norm": 0.37003710865974426, "learning_rate": 1.3757807572707958e-05, "loss": 0.0129, "step": 7867 }, { "epoch": 0.5406816932380428, "grad_norm": 0.4825001060962677, "learning_rate": 1.3754480502913306e-05, "loss": 0.0102, "step": 7868 }, { "epoch": 0.5407504123144585, "grad_norm": 0.29278844594955444, "learning_rate": 1.3751153494819127e-05, "loss": 0.0161, "step": 7869 }, { "epoch": 0.5408191313908741, "grad_norm": 0.7179781794548035, "learning_rate": 1.3747826548590236e-05, "loss": 0.0164, "step": 7870 }, { "epoch": 0.5408878504672897, "grad_norm": 0.29467037320137024, "learning_rate": 1.3744499664391445e-05, "loss": 0.0126, "step": 7871 }, { "epoch": 0.5409565695437053, "grad_norm": 0.5298754572868347, "learning_rate": 1.374117284238755e-05, "loss": 0.0073, "step": 7872 }, { "epoch": 0.5410252886201209, "grad_norm": 0.3198765516281128, "learning_rate": 1.3737846082743376e-05, "loss": 0.0166, "step": 7873 }, { "epoch": 0.5410940076965366, "grad_norm": 1.9933326244354248, "learning_rate": 1.373451938562371e-05, "loss": 0.0143, "step": 7874 }, { "epoch": 0.5411627267729522, "grad_norm": 1.1271023750305176, "learning_rate": 1.373119275119335e-05, "loss": 0.0177, "step": 7875 }, { "epoch": 0.5412314458493678, "grad_norm": 0.8231607675552368, "learning_rate": 1.3727866179617094e-05, "loss": 0.0232, "step": 7876 }, { "epoch": 0.5413001649257834, "grad_norm": 0.2631441652774811, "learning_rate": 1.3724539671059732e-05, "loss": 0.0104, "step": 7877 }, { "epoch": 0.541368884002199, "grad_norm": 0.47931012511253357, "learning_rate": 1.3721213225686046e-05, "loss": 0.0187, "step": 7878 }, { "epoch": 0.5414376030786147, "grad_norm": 0.7140958905220032, "learning_rate": 1.3717886843660834e-05, "loss": 0.019, "step": 7879 }, { "epoch": 0.5415063221550303, "grad_norm": 0.46897679567337036, "learning_rate": 1.3714560525148872e-05, "loss": 0.0109, "step": 7880 }, { "epoch": 0.5415750412314458, "grad_norm": 0.7580159902572632, "learning_rate": 1.3711234270314939e-05, "loss": 0.0151, "step": 7881 }, { "epoch": 0.5416437603078614, "grad_norm": 0.2344335913658142, "learning_rate": 1.370790807932381e-05, "loss": 0.0132, "step": 7882 }, { "epoch": 0.541712479384277, "grad_norm": 0.3814675211906433, "learning_rate": 1.3704581952340258e-05, "loss": 0.0113, "step": 7883 }, { "epoch": 0.5417811984606927, "grad_norm": 0.708231508731842, "learning_rate": 1.3701255889529052e-05, "loss": 0.0219, "step": 7884 }, { "epoch": 0.5418499175371083, "grad_norm": 0.7202560305595398, "learning_rate": 1.369792989105496e-05, "loss": 0.0126, "step": 7885 }, { "epoch": 0.5419186366135239, "grad_norm": 0.3183887004852295, "learning_rate": 1.3694603957082745e-05, "loss": 0.0113, "step": 7886 }, { "epoch": 0.5419873556899395, "grad_norm": 0.6211197376251221, "learning_rate": 1.3691278087777166e-05, "loss": 0.0152, "step": 7887 }, { "epoch": 0.5420560747663551, "grad_norm": 0.5380750298500061, "learning_rate": 1.368795228330298e-05, "loss": 0.0137, "step": 7888 }, { "epoch": 0.5421247938427708, "grad_norm": 1.0713965892791748, "learning_rate": 1.3684626543824941e-05, "loss": 0.0148, "step": 7889 }, { "epoch": 0.5421935129191864, "grad_norm": 0.5300642848014832, "learning_rate": 1.3681300869507791e-05, "loss": 0.0289, "step": 7890 }, { "epoch": 0.542262231995602, "grad_norm": 1.9148049354553223, "learning_rate": 1.3677975260516293e-05, "loss": 0.0257, "step": 7891 }, { "epoch": 0.5423309510720176, "grad_norm": 1.4046930074691772, "learning_rate": 1.367464971701518e-05, "loss": 0.016, "step": 7892 }, { "epoch": 0.5423996701484332, "grad_norm": 0.5507751107215881, "learning_rate": 1.3671324239169197e-05, "loss": 0.0147, "step": 7893 }, { "epoch": 0.5424683892248489, "grad_norm": 1.4875688552856445, "learning_rate": 1.366799882714308e-05, "loss": 0.0249, "step": 7894 }, { "epoch": 0.5425371083012644, "grad_norm": 0.36984890699386597, "learning_rate": 1.3664673481101562e-05, "loss": 0.0196, "step": 7895 }, { "epoch": 0.54260582737768, "grad_norm": 0.5664859414100647, "learning_rate": 1.3661348201209367e-05, "loss": 0.0138, "step": 7896 }, { "epoch": 0.5426745464540956, "grad_norm": 0.9213985204696655, "learning_rate": 1.3658022987631239e-05, "loss": 0.0191, "step": 7897 }, { "epoch": 0.5427432655305112, "grad_norm": 0.32207557559013367, "learning_rate": 1.365469784053189e-05, "loss": 0.0202, "step": 7898 }, { "epoch": 0.5428119846069269, "grad_norm": 0.4078841805458069, "learning_rate": 1.3651372760076046e-05, "loss": 0.0134, "step": 7899 }, { "epoch": 0.5428807036833425, "grad_norm": 0.7256275415420532, "learning_rate": 1.3648047746428424e-05, "loss": 0.0162, "step": 7900 }, { "epoch": 0.5429494227597581, "grad_norm": 0.7615375518798828, "learning_rate": 1.3644722799753737e-05, "loss": 0.0158, "step": 7901 }, { "epoch": 0.5430181418361737, "grad_norm": 1.2288762331008911, "learning_rate": 1.364139792021669e-05, "loss": 0.02, "step": 7902 }, { "epoch": 0.5430868609125893, "grad_norm": 0.2825797200202942, "learning_rate": 1.3638073107982002e-05, "loss": 0.0088, "step": 7903 }, { "epoch": 0.543155579989005, "grad_norm": 0.8771367073059082, "learning_rate": 1.3634748363214372e-05, "loss": 0.0155, "step": 7904 }, { "epoch": 0.5432242990654206, "grad_norm": 1.146095871925354, "learning_rate": 1.36314236860785e-05, "loss": 0.0144, "step": 7905 }, { "epoch": 0.5432930181418362, "grad_norm": 1.4320815801620483, "learning_rate": 1.3628099076739088e-05, "loss": 0.0176, "step": 7906 }, { "epoch": 0.5433617372182518, "grad_norm": 0.718289852142334, "learning_rate": 1.362477453536082e-05, "loss": 0.0282, "step": 7907 }, { "epoch": 0.5434304562946674, "grad_norm": 0.6099740266799927, "learning_rate": 1.3621450062108398e-05, "loss": 0.02, "step": 7908 }, { "epoch": 0.543499175371083, "grad_norm": 1.0579872131347656, "learning_rate": 1.3618125657146507e-05, "loss": 0.0158, "step": 7909 }, { "epoch": 0.5435678944474986, "grad_norm": 0.6049785614013672, "learning_rate": 1.361480132063983e-05, "loss": 0.0196, "step": 7910 }, { "epoch": 0.5436366135239142, "grad_norm": 1.0430907011032104, "learning_rate": 1.3611477052753048e-05, "loss": 0.0204, "step": 7911 }, { "epoch": 0.5437053326003298, "grad_norm": 0.677400529384613, "learning_rate": 1.360815285365084e-05, "loss": 0.016, "step": 7912 }, { "epoch": 0.5437740516767454, "grad_norm": 0.7574402093887329, "learning_rate": 1.3604828723497875e-05, "loss": 0.0199, "step": 7913 }, { "epoch": 0.5438427707531611, "grad_norm": 1.3061593770980835, "learning_rate": 1.360150466245882e-05, "loss": 0.017, "step": 7914 }, { "epoch": 0.5439114898295767, "grad_norm": 0.4261288642883301, "learning_rate": 1.3598180670698357e-05, "loss": 0.0156, "step": 7915 }, { "epoch": 0.5439802089059923, "grad_norm": 0.4887692332267761, "learning_rate": 1.3594856748381141e-05, "loss": 0.0173, "step": 7916 }, { "epoch": 0.5440489279824079, "grad_norm": 0.8500397205352783, "learning_rate": 1.3591532895671833e-05, "loss": 0.0254, "step": 7917 }, { "epoch": 0.5441176470588235, "grad_norm": 0.6562191247940063, "learning_rate": 1.358820911273509e-05, "loss": 0.0185, "step": 7918 }, { "epoch": 0.5441863661352392, "grad_norm": 0.8675567507743835, "learning_rate": 1.3584885399735564e-05, "loss": 0.014, "step": 7919 }, { "epoch": 0.5442550852116548, "grad_norm": 0.5696389675140381, "learning_rate": 1.3581561756837905e-05, "loss": 0.0197, "step": 7920 }, { "epoch": 0.5443238042880704, "grad_norm": 0.5164708495140076, "learning_rate": 1.3578238184206761e-05, "loss": 0.0177, "step": 7921 }, { "epoch": 0.544392523364486, "grad_norm": 0.3594796657562256, "learning_rate": 1.3574914682006778e-05, "loss": 0.0174, "step": 7922 }, { "epoch": 0.5444612424409015, "grad_norm": 0.5724748373031616, "learning_rate": 1.3571591250402585e-05, "loss": 0.0145, "step": 7923 }, { "epoch": 0.5445299615173173, "grad_norm": 0.40134909749031067, "learning_rate": 1.356826788955883e-05, "loss": 0.0212, "step": 7924 }, { "epoch": 0.5445986805937328, "grad_norm": 0.8408058285713196, "learning_rate": 1.3564944599640138e-05, "loss": 0.0157, "step": 7925 }, { "epoch": 0.5446673996701484, "grad_norm": 0.44512686133384705, "learning_rate": 1.3561621380811137e-05, "loss": 0.0183, "step": 7926 }, { "epoch": 0.544736118746564, "grad_norm": 0.5128852725028992, "learning_rate": 1.3558298233236461e-05, "loss": 0.012, "step": 7927 }, { "epoch": 0.5448048378229796, "grad_norm": 0.5875217318534851, "learning_rate": 1.3554975157080727e-05, "loss": 0.0219, "step": 7928 }, { "epoch": 0.5448735568993953, "grad_norm": 0.40211305022239685, "learning_rate": 1.3551652152508551e-05, "loss": 0.0157, "step": 7929 }, { "epoch": 0.5449422759758109, "grad_norm": 0.8319861888885498, "learning_rate": 1.3548329219684549e-05, "loss": 0.0219, "step": 7930 }, { "epoch": 0.5450109950522265, "grad_norm": 0.6979824900627136, "learning_rate": 1.3545006358773335e-05, "loss": 0.0125, "step": 7931 }, { "epoch": 0.5450797141286421, "grad_norm": 0.30020672082901, "learning_rate": 1.3541683569939506e-05, "loss": 0.0141, "step": 7932 }, { "epoch": 0.5451484332050577, "grad_norm": 0.7673254609107971, "learning_rate": 1.3538360853347682e-05, "loss": 0.0206, "step": 7933 }, { "epoch": 0.5452171522814734, "grad_norm": 0.516434371471405, "learning_rate": 1.3535038209162458e-05, "loss": 0.0204, "step": 7934 }, { "epoch": 0.545285871357889, "grad_norm": 0.3225361406803131, "learning_rate": 1.3531715637548429e-05, "loss": 0.0117, "step": 7935 }, { "epoch": 0.5453545904343046, "grad_norm": 0.30026087164878845, "learning_rate": 1.3528393138670188e-05, "loss": 0.0157, "step": 7936 }, { "epoch": 0.5454233095107202, "grad_norm": 0.6860270500183105, "learning_rate": 1.3525070712692324e-05, "loss": 0.0267, "step": 7937 }, { "epoch": 0.5454920285871357, "grad_norm": 1.0010029077529907, "learning_rate": 1.3521748359779422e-05, "loss": 0.0211, "step": 7938 }, { "epoch": 0.5455607476635514, "grad_norm": 0.7158087491989136, "learning_rate": 1.3518426080096071e-05, "loss": 0.0208, "step": 7939 }, { "epoch": 0.545629466739967, "grad_norm": 0.39057987928390503, "learning_rate": 1.3515103873806844e-05, "loss": 0.0155, "step": 7940 }, { "epoch": 0.5456981858163826, "grad_norm": 0.574944019317627, "learning_rate": 1.3511781741076318e-05, "loss": 0.0183, "step": 7941 }, { "epoch": 0.5457669048927982, "grad_norm": 0.7104958891868591, "learning_rate": 1.3508459682069068e-05, "loss": 0.0135, "step": 7942 }, { "epoch": 0.5458356239692138, "grad_norm": 0.3461332619190216, "learning_rate": 1.3505137696949658e-05, "loss": 0.0086, "step": 7943 }, { "epoch": 0.5459043430456295, "grad_norm": 0.7040780186653137, "learning_rate": 1.350181578588265e-05, "loss": 0.0192, "step": 7944 }, { "epoch": 0.5459730621220451, "grad_norm": 0.3922024667263031, "learning_rate": 1.3498493949032611e-05, "loss": 0.02, "step": 7945 }, { "epoch": 0.5460417811984607, "grad_norm": 0.5519413948059082, "learning_rate": 1.3495172186564095e-05, "loss": 0.0118, "step": 7946 }, { "epoch": 0.5461105002748763, "grad_norm": 0.28282058238983154, "learning_rate": 1.3491850498641654e-05, "loss": 0.0352, "step": 7947 }, { "epoch": 0.5461792193512919, "grad_norm": 0.7095369100570679, "learning_rate": 1.348852888542984e-05, "loss": 0.0193, "step": 7948 }, { "epoch": 0.5462479384277076, "grad_norm": 1.0169793367385864, "learning_rate": 1.3485207347093197e-05, "loss": 0.0213, "step": 7949 }, { "epoch": 0.5463166575041232, "grad_norm": 0.5715327262878418, "learning_rate": 1.348188588379626e-05, "loss": 0.0133, "step": 7950 }, { "epoch": 0.5463853765805388, "grad_norm": 0.5364977121353149, "learning_rate": 1.3478564495703584e-05, "loss": 0.0185, "step": 7951 }, { "epoch": 0.5464540956569544, "grad_norm": 0.684868574142456, "learning_rate": 1.3475243182979693e-05, "loss": 0.0183, "step": 7952 }, { "epoch": 0.5465228147333699, "grad_norm": 0.47902804613113403, "learning_rate": 1.3471921945789123e-05, "loss": 0.0182, "step": 7953 }, { "epoch": 0.5465915338097856, "grad_norm": 1.4716089963912964, "learning_rate": 1.3468600784296395e-05, "loss": 0.0231, "step": 7954 }, { "epoch": 0.5466602528862012, "grad_norm": 0.7505061626434326, "learning_rate": 1.3465279698666038e-05, "loss": 0.0152, "step": 7955 }, { "epoch": 0.5467289719626168, "grad_norm": 0.45807722210884094, "learning_rate": 1.346195868906256e-05, "loss": 0.012, "step": 7956 }, { "epoch": 0.5467976910390324, "grad_norm": 0.32473331689834595, "learning_rate": 1.3458637755650496e-05, "loss": 0.0138, "step": 7957 }, { "epoch": 0.546866410115448, "grad_norm": 0.46159422397613525, "learning_rate": 1.3455316898594347e-05, "loss": 0.0187, "step": 7958 }, { "epoch": 0.5469351291918637, "grad_norm": 0.5391578674316406, "learning_rate": 1.3451996118058622e-05, "loss": 0.0212, "step": 7959 }, { "epoch": 0.5470038482682793, "grad_norm": 0.30469608306884766, "learning_rate": 1.3448675414207831e-05, "loss": 0.0112, "step": 7960 }, { "epoch": 0.5470725673446949, "grad_norm": 0.3046993017196655, "learning_rate": 1.3445354787206467e-05, "loss": 0.0121, "step": 7961 }, { "epoch": 0.5471412864211105, "grad_norm": 1.1684588193893433, "learning_rate": 1.3442034237219028e-05, "loss": 0.0167, "step": 7962 }, { "epoch": 0.5472100054975261, "grad_norm": 0.3776564300060272, "learning_rate": 1.3438713764410014e-05, "loss": 0.0113, "step": 7963 }, { "epoch": 0.5472787245739418, "grad_norm": 0.5204702615737915, "learning_rate": 1.3435393368943909e-05, "loss": 0.0224, "step": 7964 }, { "epoch": 0.5473474436503574, "grad_norm": 0.7339690923690796, "learning_rate": 1.3432073050985201e-05, "loss": 0.0176, "step": 7965 }, { "epoch": 0.547416162726773, "grad_norm": 0.6355395913124084, "learning_rate": 1.3428752810698369e-05, "loss": 0.0174, "step": 7966 }, { "epoch": 0.5474848818031886, "grad_norm": 0.508712887763977, "learning_rate": 1.3425432648247894e-05, "loss": 0.0223, "step": 7967 }, { "epoch": 0.5475536008796041, "grad_norm": 1.035267949104309, "learning_rate": 1.342211256379824e-05, "loss": 0.0191, "step": 7968 }, { "epoch": 0.5476223199560198, "grad_norm": 0.8546780347824097, "learning_rate": 1.3418792557513895e-05, "loss": 0.0216, "step": 7969 }, { "epoch": 0.5476910390324354, "grad_norm": 1.1177456378936768, "learning_rate": 1.3415472629559314e-05, "loss": 0.0207, "step": 7970 }, { "epoch": 0.547759758108851, "grad_norm": 0.4566617012023926, "learning_rate": 1.3412152780098961e-05, "loss": 0.014, "step": 7971 }, { "epoch": 0.5478284771852666, "grad_norm": 0.8776983618736267, "learning_rate": 1.3408833009297294e-05, "loss": 0.0204, "step": 7972 }, { "epoch": 0.5478971962616822, "grad_norm": 0.6491047739982605, "learning_rate": 1.340551331731877e-05, "loss": 0.0145, "step": 7973 }, { "epoch": 0.5479659153380979, "grad_norm": 0.9849351048469543, "learning_rate": 1.3402193704327832e-05, "loss": 0.0269, "step": 7974 }, { "epoch": 0.5480346344145135, "grad_norm": 0.5889410376548767, "learning_rate": 1.339887417048894e-05, "loss": 0.0177, "step": 7975 }, { "epoch": 0.5481033534909291, "grad_norm": 0.6092900037765503, "learning_rate": 1.3395554715966528e-05, "loss": 0.0102, "step": 7976 }, { "epoch": 0.5481720725673447, "grad_norm": 0.4942336976528168, "learning_rate": 1.3392235340925036e-05, "loss": 0.0193, "step": 7977 }, { "epoch": 0.5482407916437603, "grad_norm": 0.5754579901695251, "learning_rate": 1.3388916045528902e-05, "loss": 0.0261, "step": 7978 }, { "epoch": 0.548309510720176, "grad_norm": 0.5866357684135437, "learning_rate": 1.3385596829942553e-05, "loss": 0.0185, "step": 7979 }, { "epoch": 0.5483782297965916, "grad_norm": 0.742058277130127, "learning_rate": 1.3382277694330418e-05, "loss": 0.014, "step": 7980 }, { "epoch": 0.5484469488730072, "grad_norm": 0.39832374453544617, "learning_rate": 1.337895863885692e-05, "loss": 0.0147, "step": 7981 }, { "epoch": 0.5485156679494227, "grad_norm": 1.2911149263381958, "learning_rate": 1.337563966368648e-05, "loss": 0.0176, "step": 7982 }, { "epoch": 0.5485843870258383, "grad_norm": 0.706943154335022, "learning_rate": 1.3372320768983512e-05, "loss": 0.0154, "step": 7983 }, { "epoch": 0.548653106102254, "grad_norm": 0.6454925537109375, "learning_rate": 1.3369001954912425e-05, "loss": 0.017, "step": 7984 }, { "epoch": 0.5487218251786696, "grad_norm": 0.45506131649017334, "learning_rate": 1.336568322163763e-05, "loss": 0.0221, "step": 7985 }, { "epoch": 0.5487905442550852, "grad_norm": 0.8317420482635498, "learning_rate": 1.336236456932352e-05, "loss": 0.0171, "step": 7986 }, { "epoch": 0.5488592633315008, "grad_norm": 0.3680287003517151, "learning_rate": 1.3359045998134508e-05, "loss": 0.0244, "step": 7987 }, { "epoch": 0.5489279824079164, "grad_norm": 0.7287091612815857, "learning_rate": 1.3355727508234985e-05, "loss": 0.0172, "step": 7988 }, { "epoch": 0.5489967014843321, "grad_norm": 0.7025608420372009, "learning_rate": 1.335240909978934e-05, "loss": 0.0177, "step": 7989 }, { "epoch": 0.5490654205607477, "grad_norm": 0.6385213136672974, "learning_rate": 1.3349090772961962e-05, "loss": 0.0375, "step": 7990 }, { "epoch": 0.5491341396371633, "grad_norm": 0.24172458052635193, "learning_rate": 1.3345772527917234e-05, "loss": 0.0149, "step": 7991 }, { "epoch": 0.5492028587135789, "grad_norm": 1.2118098735809326, "learning_rate": 1.3342454364819525e-05, "loss": 0.0198, "step": 7992 }, { "epoch": 0.5492715777899945, "grad_norm": 0.8934622406959534, "learning_rate": 1.3339136283833225e-05, "loss": 0.0187, "step": 7993 }, { "epoch": 0.5493402968664101, "grad_norm": 0.8636633157730103, "learning_rate": 1.3335818285122702e-05, "loss": 0.0228, "step": 7994 }, { "epoch": 0.5494090159428258, "grad_norm": 0.6913759112358093, "learning_rate": 1.3332500368852318e-05, "loss": 0.0212, "step": 7995 }, { "epoch": 0.5494777350192414, "grad_norm": 0.6964544653892517, "learning_rate": 1.3329182535186438e-05, "loss": 0.0227, "step": 7996 }, { "epoch": 0.5495464540956569, "grad_norm": 0.768250048160553, "learning_rate": 1.332586478428942e-05, "loss": 0.0206, "step": 7997 }, { "epoch": 0.5496151731720725, "grad_norm": 0.3422018885612488, "learning_rate": 1.3322547116325615e-05, "loss": 0.0204, "step": 7998 }, { "epoch": 0.5496838922484881, "grad_norm": 1.4234991073608398, "learning_rate": 1.331922953145938e-05, "loss": 0.0394, "step": 7999 }, { "epoch": 0.5497526113249038, "grad_norm": 0.574441134929657, "learning_rate": 1.3315912029855059e-05, "loss": 0.0274, "step": 8000 }, { "epoch": 0.5498213304013194, "grad_norm": 0.31792590022087097, "learning_rate": 1.3312594611676995e-05, "loss": 0.0236, "step": 8001 }, { "epoch": 0.549890049477735, "grad_norm": 0.4082063138484955, "learning_rate": 1.3309277277089524e-05, "loss": 0.0177, "step": 8002 }, { "epoch": 0.5499587685541506, "grad_norm": 0.4841064512729645, "learning_rate": 1.330596002625698e-05, "loss": 0.0251, "step": 8003 }, { "epoch": 0.5500274876305662, "grad_norm": 0.9130215644836426, "learning_rate": 1.3302642859343687e-05, "loss": 0.0207, "step": 8004 }, { "epoch": 0.5500962067069819, "grad_norm": 1.4244987964630127, "learning_rate": 1.3299325776513983e-05, "loss": 0.0146, "step": 8005 }, { "epoch": 0.5501649257833975, "grad_norm": 0.4247429072856903, "learning_rate": 1.3296008777932185e-05, "loss": 0.0153, "step": 8006 }, { "epoch": 0.5502336448598131, "grad_norm": 0.5098201036453247, "learning_rate": 1.329269186376261e-05, "loss": 0.0136, "step": 8007 }, { "epoch": 0.5503023639362287, "grad_norm": 0.6214399337768555, "learning_rate": 1.3289375034169567e-05, "loss": 0.0211, "step": 8008 }, { "epoch": 0.5503710830126443, "grad_norm": 1.0056098699569702, "learning_rate": 1.3286058289317371e-05, "loss": 0.017, "step": 8009 }, { "epoch": 0.55043980208906, "grad_norm": 0.6963889598846436, "learning_rate": 1.3282741629370316e-05, "loss": 0.0199, "step": 8010 }, { "epoch": 0.5505085211654756, "grad_norm": 1.259087324142456, "learning_rate": 1.3279425054492715e-05, "loss": 0.0162, "step": 8011 }, { "epoch": 0.5505772402418911, "grad_norm": 0.5802698731422424, "learning_rate": 1.3276108564848859e-05, "loss": 0.0193, "step": 8012 }, { "epoch": 0.5506459593183067, "grad_norm": 0.806764543056488, "learning_rate": 1.3272792160603043e-05, "loss": 0.0155, "step": 8013 }, { "epoch": 0.5507146783947223, "grad_norm": 0.420687735080719, "learning_rate": 1.3269475841919552e-05, "loss": 0.0142, "step": 8014 }, { "epoch": 0.550783397471138, "grad_norm": 2.0072288513183594, "learning_rate": 1.326615960896267e-05, "loss": 0.0214, "step": 8015 }, { "epoch": 0.5508521165475536, "grad_norm": 0.7808833122253418, "learning_rate": 1.3262843461896672e-05, "loss": 0.019, "step": 8016 }, { "epoch": 0.5509208356239692, "grad_norm": 0.9095028638839722, "learning_rate": 1.3259527400885842e-05, "loss": 0.0175, "step": 8017 }, { "epoch": 0.5509895547003848, "grad_norm": 2.188769817352295, "learning_rate": 1.3256211426094444e-05, "loss": 0.0229, "step": 8018 }, { "epoch": 0.5510582737768004, "grad_norm": 0.6417564749717712, "learning_rate": 1.325289553768675e-05, "loss": 0.0212, "step": 8019 }, { "epoch": 0.5511269928532161, "grad_norm": 0.9206916093826294, "learning_rate": 1.3249579735827018e-05, "loss": 0.017, "step": 8020 }, { "epoch": 0.5511957119296317, "grad_norm": 1.8355612754821777, "learning_rate": 1.3246264020679507e-05, "loss": 0.0144, "step": 8021 }, { "epoch": 0.5512644310060473, "grad_norm": 0.7012558579444885, "learning_rate": 1.3242948392408463e-05, "loss": 0.0165, "step": 8022 }, { "epoch": 0.5513331500824629, "grad_norm": 2.0305283069610596, "learning_rate": 1.323963285117815e-05, "loss": 0.0159, "step": 8023 }, { "epoch": 0.5514018691588785, "grad_norm": 0.734563946723938, "learning_rate": 1.3236317397152809e-05, "loss": 0.0112, "step": 8024 }, { "epoch": 0.5514705882352942, "grad_norm": 0.40646860003471375, "learning_rate": 1.3233002030496677e-05, "loss": 0.0182, "step": 8025 }, { "epoch": 0.5515393073117097, "grad_norm": 0.6245179176330566, "learning_rate": 1.3229686751373991e-05, "loss": 0.0195, "step": 8026 }, { "epoch": 0.5516080263881253, "grad_norm": 0.7635485529899597, "learning_rate": 1.3226371559948982e-05, "loss": 0.0326, "step": 8027 }, { "epoch": 0.5516767454645409, "grad_norm": 0.6218582987785339, "learning_rate": 1.3223056456385875e-05, "loss": 0.0148, "step": 8028 }, { "epoch": 0.5517454645409565, "grad_norm": 0.25567054748535156, "learning_rate": 1.3219741440848906e-05, "loss": 0.0143, "step": 8029 }, { "epoch": 0.5518141836173722, "grad_norm": 1.0126780271530151, "learning_rate": 1.3216426513502283e-05, "loss": 0.0194, "step": 8030 }, { "epoch": 0.5518829026937878, "grad_norm": 0.4648224711418152, "learning_rate": 1.3213111674510224e-05, "loss": 0.0201, "step": 8031 }, { "epoch": 0.5519516217702034, "grad_norm": 1.0166140794754028, "learning_rate": 1.3209796924036942e-05, "loss": 0.0154, "step": 8032 }, { "epoch": 0.552020340846619, "grad_norm": 0.3944508135318756, "learning_rate": 1.3206482262246637e-05, "loss": 0.0117, "step": 8033 }, { "epoch": 0.5520890599230346, "grad_norm": 0.4783737361431122, "learning_rate": 1.320316768930351e-05, "loss": 0.0208, "step": 8034 }, { "epoch": 0.5521577789994503, "grad_norm": 0.965506911277771, "learning_rate": 1.3199853205371768e-05, "loss": 0.0277, "step": 8035 }, { "epoch": 0.5522264980758659, "grad_norm": 0.532882034778595, "learning_rate": 1.3196538810615594e-05, "loss": 0.0152, "step": 8036 }, { "epoch": 0.5522952171522815, "grad_norm": 0.8987716436386108, "learning_rate": 1.3193224505199181e-05, "loss": 0.0312, "step": 8037 }, { "epoch": 0.5523639362286971, "grad_norm": 0.42091405391693115, "learning_rate": 1.318991028928671e-05, "loss": 0.0161, "step": 8038 }, { "epoch": 0.5524326553051127, "grad_norm": 0.384124755859375, "learning_rate": 1.3186596163042362e-05, "loss": 0.0177, "step": 8039 }, { "epoch": 0.5525013743815284, "grad_norm": 0.6151126623153687, "learning_rate": 1.3183282126630308e-05, "loss": 0.0154, "step": 8040 }, { "epoch": 0.552570093457944, "grad_norm": 0.6655160784721375, "learning_rate": 1.3179968180214726e-05, "loss": 0.0268, "step": 8041 }, { "epoch": 0.5526388125343595, "grad_norm": 0.7563478350639343, "learning_rate": 1.3176654323959781e-05, "loss": 0.0202, "step": 8042 }, { "epoch": 0.5527075316107751, "grad_norm": 0.7578108310699463, "learning_rate": 1.317334055802963e-05, "loss": 0.0187, "step": 8043 }, { "epoch": 0.5527762506871907, "grad_norm": 0.43255648016929626, "learning_rate": 1.3170026882588435e-05, "loss": 0.0163, "step": 8044 }, { "epoch": 0.5528449697636064, "grad_norm": 1.1594346761703491, "learning_rate": 1.3166713297800343e-05, "loss": 0.0195, "step": 8045 }, { "epoch": 0.552913688840022, "grad_norm": 0.7268009185791016, "learning_rate": 1.3163399803829499e-05, "loss": 0.017, "step": 8046 }, { "epoch": 0.5529824079164376, "grad_norm": 0.6467212438583374, "learning_rate": 1.3160086400840061e-05, "loss": 0.0191, "step": 8047 }, { "epoch": 0.5530511269928532, "grad_norm": 0.5331384539604187, "learning_rate": 1.3156773088996158e-05, "loss": 0.0273, "step": 8048 }, { "epoch": 0.5531198460692688, "grad_norm": 0.7683979868888855, "learning_rate": 1.3153459868461929e-05, "loss": 0.0163, "step": 8049 }, { "epoch": 0.5531885651456845, "grad_norm": 0.4330410063266754, "learning_rate": 1.3150146739401501e-05, "loss": 0.0157, "step": 8050 }, { "epoch": 0.5532572842221001, "grad_norm": 0.7322490811347961, "learning_rate": 1.3146833701978998e-05, "loss": 0.012, "step": 8051 }, { "epoch": 0.5533260032985157, "grad_norm": 0.5200770497322083, "learning_rate": 1.3143520756358543e-05, "loss": 0.016, "step": 8052 }, { "epoch": 0.5533947223749313, "grad_norm": 1.0308276414871216, "learning_rate": 1.3140207902704255e-05, "loss": 0.018, "step": 8053 }, { "epoch": 0.5534634414513468, "grad_norm": 0.460972398519516, "learning_rate": 1.3136895141180246e-05, "loss": 0.0174, "step": 8054 }, { "epoch": 0.5535321605277626, "grad_norm": 1.0277892351150513, "learning_rate": 1.3133582471950619e-05, "loss": 0.0227, "step": 8055 }, { "epoch": 0.5536008796041781, "grad_norm": 0.8010485768318176, "learning_rate": 1.3130269895179478e-05, "loss": 0.0173, "step": 8056 }, { "epoch": 0.5536695986805937, "grad_norm": 0.7413132190704346, "learning_rate": 1.3126957411030926e-05, "loss": 0.027, "step": 8057 }, { "epoch": 0.5537383177570093, "grad_norm": 0.7436158657073975, "learning_rate": 1.3123645019669051e-05, "loss": 0.0182, "step": 8058 }, { "epoch": 0.5538070368334249, "grad_norm": 0.6140785217285156, "learning_rate": 1.3120332721257938e-05, "loss": 0.0196, "step": 8059 }, { "epoch": 0.5538757559098406, "grad_norm": 0.42181462049484253, "learning_rate": 1.3117020515961683e-05, "loss": 0.0153, "step": 8060 }, { "epoch": 0.5539444749862562, "grad_norm": 0.6976031064987183, "learning_rate": 1.311370840394436e-05, "loss": 0.0218, "step": 8061 }, { "epoch": 0.5540131940626718, "grad_norm": 0.49450555443763733, "learning_rate": 1.3110396385370048e-05, "loss": 0.0158, "step": 8062 }, { "epoch": 0.5540819131390874, "grad_norm": 0.26604145765304565, "learning_rate": 1.310708446040281e-05, "loss": 0.011, "step": 8063 }, { "epoch": 0.554150632215503, "grad_norm": 0.34611648321151733, "learning_rate": 1.3103772629206719e-05, "loss": 0.0102, "step": 8064 }, { "epoch": 0.5542193512919187, "grad_norm": 0.9022730588912964, "learning_rate": 1.3100460891945826e-05, "loss": 0.0264, "step": 8065 }, { "epoch": 0.5542880703683343, "grad_norm": 0.5390937924385071, "learning_rate": 1.3097149248784203e-05, "loss": 0.0146, "step": 8066 }, { "epoch": 0.5543567894447499, "grad_norm": 0.26260945200920105, "learning_rate": 1.3093837699885892e-05, "loss": 0.0113, "step": 8067 }, { "epoch": 0.5544255085211655, "grad_norm": 0.7487532496452332, "learning_rate": 1.3090526245414944e-05, "loss": 0.0126, "step": 8068 }, { "epoch": 0.554494227597581, "grad_norm": 0.47375407814979553, "learning_rate": 1.30872148855354e-05, "loss": 0.0205, "step": 8069 }, { "epoch": 0.5545629466739967, "grad_norm": 0.28687798976898193, "learning_rate": 1.3083903620411297e-05, "loss": 0.0138, "step": 8070 }, { "epoch": 0.5546316657504123, "grad_norm": 0.5192314386367798, "learning_rate": 1.3080592450206668e-05, "loss": 0.0115, "step": 8071 }, { "epoch": 0.5547003848268279, "grad_norm": 0.46563246846199036, "learning_rate": 1.3077281375085543e-05, "loss": 0.0175, "step": 8072 }, { "epoch": 0.5547691039032435, "grad_norm": 0.6101728081703186, "learning_rate": 1.3073970395211948e-05, "loss": 0.0171, "step": 8073 }, { "epoch": 0.5548378229796591, "grad_norm": 0.23634247481822968, "learning_rate": 1.30706595107499e-05, "loss": 0.0132, "step": 8074 }, { "epoch": 0.5549065420560748, "grad_norm": 0.6563560962677002, "learning_rate": 1.3067348721863413e-05, "loss": 0.0213, "step": 8075 }, { "epoch": 0.5549752611324904, "grad_norm": 0.6897582411766052, "learning_rate": 1.3064038028716497e-05, "loss": 0.0124, "step": 8076 }, { "epoch": 0.555043980208906, "grad_norm": 0.3421960175037384, "learning_rate": 1.3060727431473148e-05, "loss": 0.0131, "step": 8077 }, { "epoch": 0.5551126992853216, "grad_norm": 0.12817630171775818, "learning_rate": 1.3057416930297384e-05, "loss": 0.0121, "step": 8078 }, { "epoch": 0.5551814183617372, "grad_norm": 0.23512744903564453, "learning_rate": 1.3054106525353192e-05, "loss": 0.0253, "step": 8079 }, { "epoch": 0.5552501374381529, "grad_norm": 0.5500817894935608, "learning_rate": 1.305079621680456e-05, "loss": 0.0125, "step": 8080 }, { "epoch": 0.5553188565145685, "grad_norm": 0.7817851305007935, "learning_rate": 1.3047486004815477e-05, "loss": 0.0139, "step": 8081 }, { "epoch": 0.5553875755909841, "grad_norm": 0.562034547328949, "learning_rate": 1.3044175889549925e-05, "loss": 0.0215, "step": 8082 }, { "epoch": 0.5554562946673997, "grad_norm": 0.4118027091026306, "learning_rate": 1.3040865871171868e-05, "loss": 0.0222, "step": 8083 }, { "epoch": 0.5555250137438152, "grad_norm": 0.8436376452445984, "learning_rate": 1.3037555949845299e-05, "loss": 0.0119, "step": 8084 }, { "epoch": 0.555593732820231, "grad_norm": 0.5712012648582458, "learning_rate": 1.3034246125734171e-05, "loss": 0.0147, "step": 8085 }, { "epoch": 0.5556624518966465, "grad_norm": 0.6448320746421814, "learning_rate": 1.3030936399002448e-05, "loss": 0.0115, "step": 8086 }, { "epoch": 0.5557311709730621, "grad_norm": 0.5160427093505859, "learning_rate": 1.3027626769814088e-05, "loss": 0.02, "step": 8087 }, { "epoch": 0.5557998900494777, "grad_norm": 0.5716456770896912, "learning_rate": 1.3024317238333045e-05, "loss": 0.0145, "step": 8088 }, { "epoch": 0.5558686091258933, "grad_norm": 0.35462895035743713, "learning_rate": 1.3021007804723257e-05, "loss": 0.0162, "step": 8089 }, { "epoch": 0.555937328202309, "grad_norm": 1.11104416847229, "learning_rate": 1.301769846914868e-05, "loss": 0.0177, "step": 8090 }, { "epoch": 0.5560060472787246, "grad_norm": 0.4596874713897705, "learning_rate": 1.3014389231773244e-05, "loss": 0.0104, "step": 8091 }, { "epoch": 0.5560747663551402, "grad_norm": 0.732096254825592, "learning_rate": 1.3011080092760885e-05, "loss": 0.0158, "step": 8092 }, { "epoch": 0.5561434854315558, "grad_norm": 0.9161426424980164, "learning_rate": 1.3007771052275528e-05, "loss": 0.0209, "step": 8093 }, { "epoch": 0.5562122045079714, "grad_norm": 1.0813369750976562, "learning_rate": 1.3004462110481096e-05, "loss": 0.0409, "step": 8094 }, { "epoch": 0.5562809235843871, "grad_norm": 0.9370751976966858, "learning_rate": 1.3001153267541502e-05, "loss": 0.0115, "step": 8095 }, { "epoch": 0.5563496426608027, "grad_norm": 1.3157014846801758, "learning_rate": 1.2997844523620675e-05, "loss": 0.0308, "step": 8096 }, { "epoch": 0.5564183617372183, "grad_norm": 0.4572758376598358, "learning_rate": 1.299453587888251e-05, "loss": 0.013, "step": 8097 }, { "epoch": 0.5564870808136338, "grad_norm": 0.9800923466682434, "learning_rate": 1.299122733349092e-05, "loss": 0.0206, "step": 8098 }, { "epoch": 0.5565557998900494, "grad_norm": 0.4094153642654419, "learning_rate": 1.2987918887609792e-05, "loss": 0.0227, "step": 8099 }, { "epoch": 0.5566245189664651, "grad_norm": 0.2414274662733078, "learning_rate": 1.298461054140303e-05, "loss": 0.0133, "step": 8100 }, { "epoch": 0.5566932380428807, "grad_norm": 0.3904472887516022, "learning_rate": 1.298130229503451e-05, "loss": 0.0143, "step": 8101 }, { "epoch": 0.5567619571192963, "grad_norm": 0.5643033385276794, "learning_rate": 1.2977994148668132e-05, "loss": 0.0127, "step": 8102 }, { "epoch": 0.5568306761957119, "grad_norm": 1.0329145193099976, "learning_rate": 1.2974686102467764e-05, "loss": 0.0166, "step": 8103 }, { "epoch": 0.5568993952721275, "grad_norm": 0.7723889946937561, "learning_rate": 1.2971378156597284e-05, "loss": 0.0091, "step": 8104 }, { "epoch": 0.5569681143485432, "grad_norm": 0.4570957124233246, "learning_rate": 1.296807031122056e-05, "loss": 0.022, "step": 8105 }, { "epoch": 0.5570368334249588, "grad_norm": 0.8616910576820374, "learning_rate": 1.2964762566501455e-05, "loss": 0.0186, "step": 8106 }, { "epoch": 0.5571055525013744, "grad_norm": 0.5649195909500122, "learning_rate": 1.2961454922603824e-05, "loss": 0.0114, "step": 8107 }, { "epoch": 0.55717427157779, "grad_norm": 1.0991907119750977, "learning_rate": 1.295814737969153e-05, "loss": 0.0229, "step": 8108 }, { "epoch": 0.5572429906542056, "grad_norm": 0.7384955286979675, "learning_rate": 1.2954839937928415e-05, "loss": 0.0304, "step": 8109 }, { "epoch": 0.5573117097306213, "grad_norm": 0.4805757403373718, "learning_rate": 1.2951532597478325e-05, "loss": 0.0162, "step": 8110 }, { "epoch": 0.5573804288070369, "grad_norm": 0.460610032081604, "learning_rate": 1.2948225358505102e-05, "loss": 0.0192, "step": 8111 }, { "epoch": 0.5574491478834525, "grad_norm": 0.6330872774124146, "learning_rate": 1.2944918221172574e-05, "loss": 0.0119, "step": 8112 }, { "epoch": 0.557517866959868, "grad_norm": 0.7466796040534973, "learning_rate": 1.2941611185644565e-05, "loss": 0.0272, "step": 8113 }, { "epoch": 0.5575865860362836, "grad_norm": 0.48878931999206543, "learning_rate": 1.2938304252084913e-05, "loss": 0.011, "step": 8114 }, { "epoch": 0.5576553051126993, "grad_norm": 0.4504428207874298, "learning_rate": 1.2934997420657432e-05, "loss": 0.0209, "step": 8115 }, { "epoch": 0.5577240241891149, "grad_norm": 1.1015422344207764, "learning_rate": 1.2931690691525929e-05, "loss": 0.0232, "step": 8116 }, { "epoch": 0.5577927432655305, "grad_norm": 0.269503116607666, "learning_rate": 1.292838406485422e-05, "loss": 0.0221, "step": 8117 }, { "epoch": 0.5578614623419461, "grad_norm": 0.43293318152427673, "learning_rate": 1.2925077540806105e-05, "loss": 0.0128, "step": 8118 }, { "epoch": 0.5579301814183617, "grad_norm": 0.47204867005348206, "learning_rate": 1.2921771119545374e-05, "loss": 0.0213, "step": 8119 }, { "epoch": 0.5579989004947774, "grad_norm": 1.0331388711929321, "learning_rate": 1.2918464801235839e-05, "loss": 0.0222, "step": 8120 }, { "epoch": 0.558067619571193, "grad_norm": 0.5558371543884277, "learning_rate": 1.2915158586041276e-05, "loss": 0.0217, "step": 8121 }, { "epoch": 0.5581363386476086, "grad_norm": 0.806133508682251, "learning_rate": 1.2911852474125472e-05, "loss": 0.0173, "step": 8122 }, { "epoch": 0.5582050577240242, "grad_norm": 0.46993929147720337, "learning_rate": 1.2908546465652201e-05, "loss": 0.0118, "step": 8123 }, { "epoch": 0.5582737768004398, "grad_norm": 0.23785613477230072, "learning_rate": 1.2905240560785241e-05, "loss": 0.0186, "step": 8124 }, { "epoch": 0.5583424958768554, "grad_norm": 0.40706902742385864, "learning_rate": 1.2901934759688352e-05, "loss": 0.018, "step": 8125 }, { "epoch": 0.5584112149532711, "grad_norm": 0.5280455350875854, "learning_rate": 1.2898629062525306e-05, "loss": 0.0271, "step": 8126 }, { "epoch": 0.5584799340296867, "grad_norm": 0.5927839279174805, "learning_rate": 1.2895323469459855e-05, "loss": 0.0306, "step": 8127 }, { "epoch": 0.5585486531061022, "grad_norm": 0.7964165210723877, "learning_rate": 1.2892017980655752e-05, "loss": 0.0231, "step": 8128 }, { "epoch": 0.5586173721825178, "grad_norm": 1.3407114744186401, "learning_rate": 1.2888712596276743e-05, "loss": 0.0328, "step": 8129 }, { "epoch": 0.5586860912589334, "grad_norm": 0.22354741394519806, "learning_rate": 1.2885407316486572e-05, "loss": 0.0161, "step": 8130 }, { "epoch": 0.5587548103353491, "grad_norm": 0.5556638240814209, "learning_rate": 1.2882102141448975e-05, "loss": 0.0182, "step": 8131 }, { "epoch": 0.5588235294117647, "grad_norm": 1.7920172214508057, "learning_rate": 1.2878797071327685e-05, "loss": 0.0133, "step": 8132 }, { "epoch": 0.5588922484881803, "grad_norm": 0.33293846249580383, "learning_rate": 1.2875492106286425e-05, "loss": 0.014, "step": 8133 }, { "epoch": 0.5589609675645959, "grad_norm": 0.3185051679611206, "learning_rate": 1.2872187246488922e-05, "loss": 0.0219, "step": 8134 }, { "epoch": 0.5590296866410115, "grad_norm": 1.3218733072280884, "learning_rate": 1.2868882492098885e-05, "loss": 0.0175, "step": 8135 }, { "epoch": 0.5590984057174272, "grad_norm": 0.9510100483894348, "learning_rate": 1.286557784328003e-05, "loss": 0.0153, "step": 8136 }, { "epoch": 0.5591671247938428, "grad_norm": 0.374631404876709, "learning_rate": 1.2862273300196054e-05, "loss": 0.018, "step": 8137 }, { "epoch": 0.5592358438702584, "grad_norm": 0.9867918491363525, "learning_rate": 1.285896886301067e-05, "loss": 0.0144, "step": 8138 }, { "epoch": 0.559304562946674, "grad_norm": 0.5818644762039185, "learning_rate": 1.2855664531887567e-05, "loss": 0.0174, "step": 8139 }, { "epoch": 0.5593732820230896, "grad_norm": 0.4549063444137573, "learning_rate": 1.2852360306990434e-05, "loss": 0.0097, "step": 8140 }, { "epoch": 0.5594420010995053, "grad_norm": 1.1545072793960571, "learning_rate": 1.2849056188482956e-05, "loss": 0.0191, "step": 8141 }, { "epoch": 0.5595107201759209, "grad_norm": 0.34110063314437866, "learning_rate": 1.2845752176528814e-05, "loss": 0.0097, "step": 8142 }, { "epoch": 0.5595794392523364, "grad_norm": 0.49183419346809387, "learning_rate": 1.2842448271291675e-05, "loss": 0.0139, "step": 8143 }, { "epoch": 0.559648158328752, "grad_norm": 0.31831595301628113, "learning_rate": 1.2839144472935216e-05, "loss": 0.0145, "step": 8144 }, { "epoch": 0.5597168774051676, "grad_norm": 0.8531498312950134, "learning_rate": 1.2835840781623096e-05, "loss": 0.0277, "step": 8145 }, { "epoch": 0.5597855964815833, "grad_norm": 0.5025055408477783, "learning_rate": 1.2832537197518975e-05, "loss": 0.0161, "step": 8146 }, { "epoch": 0.5598543155579989, "grad_norm": 0.38461580872535706, "learning_rate": 1.2829233720786506e-05, "loss": 0.0095, "step": 8147 }, { "epoch": 0.5599230346344145, "grad_norm": 0.37618061900138855, "learning_rate": 1.2825930351589335e-05, "loss": 0.0114, "step": 8148 }, { "epoch": 0.5599917537108301, "grad_norm": 0.6126326322555542, "learning_rate": 1.28226270900911e-05, "loss": 0.025, "step": 8149 }, { "epoch": 0.5600604727872457, "grad_norm": 0.2863016724586487, "learning_rate": 1.2819323936455451e-05, "loss": 0.017, "step": 8150 }, { "epoch": 0.5601291918636614, "grad_norm": 0.45612862706184387, "learning_rate": 1.2816020890846008e-05, "loss": 0.0129, "step": 8151 }, { "epoch": 0.560197910940077, "grad_norm": 1.1669265031814575, "learning_rate": 1.2812717953426398e-05, "loss": 0.0244, "step": 8152 }, { "epoch": 0.5602666300164926, "grad_norm": 0.9032773971557617, "learning_rate": 1.2809415124360246e-05, "loss": 0.016, "step": 8153 }, { "epoch": 0.5603353490929082, "grad_norm": 0.7990554571151733, "learning_rate": 1.2806112403811166e-05, "loss": 0.0202, "step": 8154 }, { "epoch": 0.5604040681693238, "grad_norm": 0.7461264729499817, "learning_rate": 1.2802809791942758e-05, "loss": 0.0209, "step": 8155 }, { "epoch": 0.5604727872457395, "grad_norm": 0.5526897311210632, "learning_rate": 1.2799507288918645e-05, "loss": 0.0232, "step": 8156 }, { "epoch": 0.560541506322155, "grad_norm": 0.7752214074134827, "learning_rate": 1.2796204894902415e-05, "loss": 0.0184, "step": 8157 }, { "epoch": 0.5606102253985706, "grad_norm": 0.4419275224208832, "learning_rate": 1.2792902610057663e-05, "loss": 0.0243, "step": 8158 }, { "epoch": 0.5606789444749862, "grad_norm": 0.706123948097229, "learning_rate": 1.2789600434547979e-05, "loss": 0.0223, "step": 8159 }, { "epoch": 0.5607476635514018, "grad_norm": 0.3743863105773926, "learning_rate": 1.2786298368536947e-05, "loss": 0.0156, "step": 8160 }, { "epoch": 0.5608163826278175, "grad_norm": 0.6541187167167664, "learning_rate": 1.2782996412188137e-05, "loss": 0.0262, "step": 8161 }, { "epoch": 0.5608851017042331, "grad_norm": 0.7408274412155151, "learning_rate": 1.2779694565665127e-05, "loss": 0.0149, "step": 8162 }, { "epoch": 0.5609538207806487, "grad_norm": 0.7760313153266907, "learning_rate": 1.2776392829131487e-05, "loss": 0.0252, "step": 8163 }, { "epoch": 0.5610225398570643, "grad_norm": 0.9895880818367004, "learning_rate": 1.2773091202750773e-05, "loss": 0.0287, "step": 8164 }, { "epoch": 0.5610912589334799, "grad_norm": 0.5082370042800903, "learning_rate": 1.2769789686686547e-05, "loss": 0.0207, "step": 8165 }, { "epoch": 0.5611599780098956, "grad_norm": 0.21407687664031982, "learning_rate": 1.276648828110235e-05, "loss": 0.02, "step": 8166 }, { "epoch": 0.5612286970863112, "grad_norm": 0.5147190093994141, "learning_rate": 1.276318698616173e-05, "loss": 0.0196, "step": 8167 }, { "epoch": 0.5612974161627268, "grad_norm": 0.4439380466938019, "learning_rate": 1.2759885802028234e-05, "loss": 0.0166, "step": 8168 }, { "epoch": 0.5613661352391424, "grad_norm": 0.8666623830795288, "learning_rate": 1.2756584728865389e-05, "loss": 0.0197, "step": 8169 }, { "epoch": 0.561434854315558, "grad_norm": 0.46905797719955444, "learning_rate": 1.2753283766836722e-05, "loss": 0.017, "step": 8170 }, { "epoch": 0.5615035733919737, "grad_norm": 0.3146718144416809, "learning_rate": 1.2749982916105759e-05, "loss": 0.0153, "step": 8171 }, { "epoch": 0.5615722924683892, "grad_norm": 0.4749532639980316, "learning_rate": 1.2746682176836016e-05, "loss": 0.0151, "step": 8172 }, { "epoch": 0.5616410115448048, "grad_norm": 0.767555296421051, "learning_rate": 1.2743381549191e-05, "loss": 0.026, "step": 8173 }, { "epoch": 0.5617097306212204, "grad_norm": 0.7849367260932922, "learning_rate": 1.2740081033334228e-05, "loss": 0.013, "step": 8174 }, { "epoch": 0.561778449697636, "grad_norm": 0.4808240234851837, "learning_rate": 1.2736780629429196e-05, "loss": 0.0288, "step": 8175 }, { "epoch": 0.5618471687740517, "grad_norm": 0.3805106282234192, "learning_rate": 1.27334803376394e-05, "loss": 0.0261, "step": 8176 }, { "epoch": 0.5619158878504673, "grad_norm": 0.3195075988769531, "learning_rate": 1.2730180158128326e-05, "loss": 0.0213, "step": 8177 }, { "epoch": 0.5619846069268829, "grad_norm": 0.5759930610656738, "learning_rate": 1.272688009105946e-05, "loss": 0.0204, "step": 8178 }, { "epoch": 0.5620533260032985, "grad_norm": 0.8043424487113953, "learning_rate": 1.2723580136596277e-05, "loss": 0.0354, "step": 8179 }, { "epoch": 0.5621220450797141, "grad_norm": 0.33404508233070374, "learning_rate": 1.2720280294902259e-05, "loss": 0.0125, "step": 8180 }, { "epoch": 0.5621907641561298, "grad_norm": 0.6405314803123474, "learning_rate": 1.2716980566140868e-05, "loss": 0.0164, "step": 8181 }, { "epoch": 0.5622594832325454, "grad_norm": 0.6646043658256531, "learning_rate": 1.2713680950475565e-05, "loss": 0.0183, "step": 8182 }, { "epoch": 0.562328202308961, "grad_norm": 0.8336702585220337, "learning_rate": 1.271038144806981e-05, "loss": 0.0142, "step": 8183 }, { "epoch": 0.5623969213853766, "grad_norm": 0.4197000563144684, "learning_rate": 1.270708205908705e-05, "loss": 0.0152, "step": 8184 }, { "epoch": 0.5624656404617921, "grad_norm": 0.4665102958679199, "learning_rate": 1.2703782783690727e-05, "loss": 0.0196, "step": 8185 }, { "epoch": 0.5625343595382079, "grad_norm": 0.6309810280799866, "learning_rate": 1.2700483622044288e-05, "loss": 0.018, "step": 8186 }, { "epoch": 0.5626030786146234, "grad_norm": 0.6223663091659546, "learning_rate": 1.2697184574311161e-05, "loss": 0.0175, "step": 8187 }, { "epoch": 0.562671797691039, "grad_norm": 0.23077575862407684, "learning_rate": 1.2693885640654777e-05, "loss": 0.0122, "step": 8188 }, { "epoch": 0.5627405167674546, "grad_norm": 0.5943346619606018, "learning_rate": 1.2690586821238556e-05, "loss": 0.0151, "step": 8189 }, { "epoch": 0.5628092358438702, "grad_norm": 0.5121400952339172, "learning_rate": 1.2687288116225919e-05, "loss": 0.0166, "step": 8190 }, { "epoch": 0.5628779549202859, "grad_norm": 1.014772891998291, "learning_rate": 1.2683989525780265e-05, "loss": 0.0141, "step": 8191 }, { "epoch": 0.5629466739967015, "grad_norm": 0.9086053371429443, "learning_rate": 1.2680691050065016e-05, "loss": 0.0213, "step": 8192 }, { "epoch": 0.5630153930731171, "grad_norm": 0.6771764755249023, "learning_rate": 1.2677392689243562e-05, "loss": 0.016, "step": 8193 }, { "epoch": 0.5630841121495327, "grad_norm": 0.5037415027618408, "learning_rate": 1.26740944434793e-05, "loss": 0.0117, "step": 8194 }, { "epoch": 0.5631528312259483, "grad_norm": 0.7914149165153503, "learning_rate": 1.2670796312935619e-05, "loss": 0.014, "step": 8195 }, { "epoch": 0.563221550302364, "grad_norm": 0.9949803352355957, "learning_rate": 1.26674982977759e-05, "loss": 0.0229, "step": 8196 }, { "epoch": 0.5632902693787796, "grad_norm": 0.6475616097450256, "learning_rate": 1.2664200398163512e-05, "loss": 0.0158, "step": 8197 }, { "epoch": 0.5633589884551952, "grad_norm": 0.9137738347053528, "learning_rate": 1.266090261426184e-05, "loss": 0.0259, "step": 8198 }, { "epoch": 0.5634277075316108, "grad_norm": 0.6158434748649597, "learning_rate": 1.2657604946234244e-05, "loss": 0.0117, "step": 8199 }, { "epoch": 0.5634964266080263, "grad_norm": 0.9639079570770264, "learning_rate": 1.2654307394244085e-05, "loss": 0.0146, "step": 8200 }, { "epoch": 0.563565145684442, "grad_norm": 1.134150505065918, "learning_rate": 1.2651009958454713e-05, "loss": 0.0188, "step": 8201 }, { "epoch": 0.5636338647608576, "grad_norm": 0.5132061243057251, "learning_rate": 1.264771263902948e-05, "loss": 0.0126, "step": 8202 }, { "epoch": 0.5637025838372732, "grad_norm": 0.4757193922996521, "learning_rate": 1.2644415436131721e-05, "loss": 0.0142, "step": 8203 }, { "epoch": 0.5637713029136888, "grad_norm": 0.8038146495819092, "learning_rate": 1.2641118349924785e-05, "loss": 0.0153, "step": 8204 }, { "epoch": 0.5638400219901044, "grad_norm": 0.4424901306629181, "learning_rate": 1.2637821380571995e-05, "loss": 0.0127, "step": 8205 }, { "epoch": 0.5639087410665201, "grad_norm": 0.5031557083129883, "learning_rate": 1.2634524528236676e-05, "loss": 0.0166, "step": 8206 }, { "epoch": 0.5639774601429357, "grad_norm": 0.4657725393772125, "learning_rate": 1.2631227793082152e-05, "loss": 0.0146, "step": 8207 }, { "epoch": 0.5640461792193513, "grad_norm": 0.28187817335128784, "learning_rate": 1.262793117527173e-05, "loss": 0.0117, "step": 8208 }, { "epoch": 0.5641148982957669, "grad_norm": 0.41523975133895874, "learning_rate": 1.2624634674968716e-05, "loss": 0.0167, "step": 8209 }, { "epoch": 0.5641836173721825, "grad_norm": 0.8781020641326904, "learning_rate": 1.2621338292336421e-05, "loss": 0.02, "step": 8210 }, { "epoch": 0.5642523364485982, "grad_norm": 0.4557887017726898, "learning_rate": 1.2618042027538139e-05, "loss": 0.0163, "step": 8211 }, { "epoch": 0.5643210555250138, "grad_norm": 0.3370606005191803, "learning_rate": 1.2614745880737155e-05, "loss": 0.0131, "step": 8212 }, { "epoch": 0.5643897746014294, "grad_norm": 0.8647406697273254, "learning_rate": 1.261144985209676e-05, "loss": 0.0206, "step": 8213 }, { "epoch": 0.564458493677845, "grad_norm": 0.5800062417984009, "learning_rate": 1.2608153941780225e-05, "loss": 0.0114, "step": 8214 }, { "epoch": 0.5645272127542605, "grad_norm": 0.6388435959815979, "learning_rate": 1.2604858149950822e-05, "loss": 0.014, "step": 8215 }, { "epoch": 0.5645959318306762, "grad_norm": 0.19835153222084045, "learning_rate": 1.2601562476771828e-05, "loss": 0.0123, "step": 8216 }, { "epoch": 0.5646646509070918, "grad_norm": 0.4893008768558502, "learning_rate": 1.2598266922406497e-05, "loss": 0.0131, "step": 8217 }, { "epoch": 0.5647333699835074, "grad_norm": 0.45745664834976196, "learning_rate": 1.2594971487018086e-05, "loss": 0.0142, "step": 8218 }, { "epoch": 0.564802089059923, "grad_norm": 0.3131279945373535, "learning_rate": 1.2591676170769844e-05, "loss": 0.0172, "step": 8219 }, { "epoch": 0.5648708081363386, "grad_norm": 0.7570332288742065, "learning_rate": 1.2588380973825013e-05, "loss": 0.0131, "step": 8220 }, { "epoch": 0.5649395272127543, "grad_norm": 0.4289928674697876, "learning_rate": 1.2585085896346826e-05, "loss": 0.0228, "step": 8221 }, { "epoch": 0.5650082462891699, "grad_norm": 0.8630871176719666, "learning_rate": 1.2581790938498522e-05, "loss": 0.0209, "step": 8222 }, { "epoch": 0.5650769653655855, "grad_norm": 1.018538236618042, "learning_rate": 1.2578496100443322e-05, "loss": 0.0226, "step": 8223 }, { "epoch": 0.5651456844420011, "grad_norm": 0.6713200807571411, "learning_rate": 1.257520138234445e-05, "loss": 0.0128, "step": 8224 }, { "epoch": 0.5652144035184167, "grad_norm": 0.8863258361816406, "learning_rate": 1.2571906784365115e-05, "loss": 0.0207, "step": 8225 }, { "epoch": 0.5652831225948324, "grad_norm": 1.3612167835235596, "learning_rate": 1.2568612306668527e-05, "loss": 0.0166, "step": 8226 }, { "epoch": 0.565351841671248, "grad_norm": 1.077353835105896, "learning_rate": 1.2565317949417878e-05, "loss": 0.023, "step": 8227 }, { "epoch": 0.5654205607476636, "grad_norm": 0.5502393841743469, "learning_rate": 1.256202371277638e-05, "loss": 0.0176, "step": 8228 }, { "epoch": 0.5654892798240791, "grad_norm": 1.0617477893829346, "learning_rate": 1.2558729596907215e-05, "loss": 0.0197, "step": 8229 }, { "epoch": 0.5655579989004947, "grad_norm": 0.5589233040809631, "learning_rate": 1.2555435601973565e-05, "loss": 0.02, "step": 8230 }, { "epoch": 0.5656267179769104, "grad_norm": 0.6403236985206604, "learning_rate": 1.2552141728138612e-05, "loss": 0.013, "step": 8231 }, { "epoch": 0.565695437053326, "grad_norm": 0.6713885068893433, "learning_rate": 1.2548847975565522e-05, "loss": 0.0116, "step": 8232 }, { "epoch": 0.5657641561297416, "grad_norm": 0.41371479630470276, "learning_rate": 1.254555434441746e-05, "loss": 0.0159, "step": 8233 }, { "epoch": 0.5658328752061572, "grad_norm": 1.2207510471343994, "learning_rate": 1.2542260834857598e-05, "loss": 0.0168, "step": 8234 }, { "epoch": 0.5659015942825728, "grad_norm": 0.16783051192760468, "learning_rate": 1.2538967447049078e-05, "loss": 0.0241, "step": 8235 }, { "epoch": 0.5659703133589885, "grad_norm": 0.4347195625305176, "learning_rate": 1.2535674181155051e-05, "loss": 0.0186, "step": 8236 }, { "epoch": 0.5660390324354041, "grad_norm": 1.0087209939956665, "learning_rate": 1.2532381037338661e-05, "loss": 0.0169, "step": 8237 }, { "epoch": 0.5661077515118197, "grad_norm": 0.5808965563774109, "learning_rate": 1.2529088015763038e-05, "loss": 0.0116, "step": 8238 }, { "epoch": 0.5661764705882353, "grad_norm": 0.37843650579452515, "learning_rate": 1.2525795116591315e-05, "loss": 0.0106, "step": 8239 }, { "epoch": 0.5662451896646509, "grad_norm": 1.484012246131897, "learning_rate": 1.2522502339986616e-05, "loss": 0.0297, "step": 8240 }, { "epoch": 0.5663139087410666, "grad_norm": 0.7433571815490723, "learning_rate": 1.251920968611206e-05, "loss": 0.0198, "step": 8241 }, { "epoch": 0.5663826278174822, "grad_norm": 0.36308351159095764, "learning_rate": 1.2515917155130753e-05, "loss": 0.0175, "step": 8242 }, { "epoch": 0.5664513468938978, "grad_norm": 0.2903159260749817, "learning_rate": 1.2512624747205805e-05, "loss": 0.0165, "step": 8243 }, { "epoch": 0.5665200659703133, "grad_norm": 0.5619956851005554, "learning_rate": 1.2509332462500311e-05, "loss": 0.0169, "step": 8244 }, { "epoch": 0.5665887850467289, "grad_norm": 1.5077234506607056, "learning_rate": 1.2506040301177362e-05, "loss": 0.0194, "step": 8245 }, { "epoch": 0.5666575041231446, "grad_norm": 0.22648093104362488, "learning_rate": 1.2502748263400055e-05, "loss": 0.0272, "step": 8246 }, { "epoch": 0.5667262231995602, "grad_norm": 0.9608548283576965, "learning_rate": 1.2499456349331464e-05, "loss": 0.0206, "step": 8247 }, { "epoch": 0.5667949422759758, "grad_norm": 0.4863940477371216, "learning_rate": 1.2496164559134665e-05, "loss": 0.0238, "step": 8248 }, { "epoch": 0.5668636613523914, "grad_norm": 0.49284109473228455, "learning_rate": 1.2492872892972726e-05, "loss": 0.0159, "step": 8249 }, { "epoch": 0.566932380428807, "grad_norm": 0.5694804787635803, "learning_rate": 1.248958135100871e-05, "loss": 0.0098, "step": 8250 }, { "epoch": 0.5670010995052226, "grad_norm": 0.8238516449928284, "learning_rate": 1.2486289933405667e-05, "loss": 0.0231, "step": 8251 }, { "epoch": 0.5670698185816383, "grad_norm": 0.40347006916999817, "learning_rate": 1.2482998640326657e-05, "loss": 0.0154, "step": 8252 }, { "epoch": 0.5671385376580539, "grad_norm": 0.7319072484970093, "learning_rate": 1.2479707471934721e-05, "loss": 0.02, "step": 8253 }, { "epoch": 0.5672072567344695, "grad_norm": 0.5175944566726685, "learning_rate": 1.2476416428392895e-05, "loss": 0.0201, "step": 8254 }, { "epoch": 0.5672759758108851, "grad_norm": 0.4141180217266083, "learning_rate": 1.2473125509864211e-05, "loss": 0.0227, "step": 8255 }, { "epoch": 0.5673446948873007, "grad_norm": 0.4453912377357483, "learning_rate": 1.2469834716511696e-05, "loss": 0.0209, "step": 8256 }, { "epoch": 0.5674134139637164, "grad_norm": 0.4026612937450409, "learning_rate": 1.2466544048498363e-05, "loss": 0.0166, "step": 8257 }, { "epoch": 0.567482133040132, "grad_norm": 0.7726255655288696, "learning_rate": 1.246325350598723e-05, "loss": 0.0186, "step": 8258 }, { "epoch": 0.5675508521165475, "grad_norm": 0.3077832758426666, "learning_rate": 1.2459963089141303e-05, "loss": 0.0109, "step": 8259 }, { "epoch": 0.5676195711929631, "grad_norm": 0.32197654247283936, "learning_rate": 1.2456672798123585e-05, "loss": 0.0143, "step": 8260 }, { "epoch": 0.5676882902693787, "grad_norm": 0.5230309963226318, "learning_rate": 1.2453382633097067e-05, "loss": 0.0212, "step": 8261 }, { "epoch": 0.5677570093457944, "grad_norm": 0.4210575520992279, "learning_rate": 1.2450092594224739e-05, "loss": 0.0165, "step": 8262 }, { "epoch": 0.56782572842221, "grad_norm": 0.4478507936000824, "learning_rate": 1.2446802681669573e-05, "loss": 0.0127, "step": 8263 }, { "epoch": 0.5678944474986256, "grad_norm": 0.6301462650299072, "learning_rate": 1.244351289559456e-05, "loss": 0.0192, "step": 8264 }, { "epoch": 0.5679631665750412, "grad_norm": 0.5564376711845398, "learning_rate": 1.2440223236162663e-05, "loss": 0.0334, "step": 8265 }, { "epoch": 0.5680318856514568, "grad_norm": 0.3509990870952606, "learning_rate": 1.2436933703536843e-05, "loss": 0.0202, "step": 8266 }, { "epoch": 0.5681006047278725, "grad_norm": 0.9406622052192688, "learning_rate": 1.2433644297880059e-05, "loss": 0.0166, "step": 8267 }, { "epoch": 0.5681693238042881, "grad_norm": 1.1252906322479248, "learning_rate": 1.243035501935526e-05, "loss": 0.0153, "step": 8268 }, { "epoch": 0.5682380428807037, "grad_norm": 0.7869469523429871, "learning_rate": 1.2427065868125383e-05, "loss": 0.0209, "step": 8269 }, { "epoch": 0.5683067619571193, "grad_norm": 0.4660276174545288, "learning_rate": 1.2423776844353381e-05, "loss": 0.0103, "step": 8270 }, { "epoch": 0.5683754810335349, "grad_norm": 0.9649559259414673, "learning_rate": 1.2420487948202176e-05, "loss": 0.0273, "step": 8271 }, { "epoch": 0.5684442001099506, "grad_norm": 0.713481605052948, "learning_rate": 1.2417199179834698e-05, "loss": 0.0164, "step": 8272 }, { "epoch": 0.5685129191863662, "grad_norm": 0.6427595615386963, "learning_rate": 1.241391053941386e-05, "loss": 0.013, "step": 8273 }, { "epoch": 0.5685816382627817, "grad_norm": 0.7864263653755188, "learning_rate": 1.2410622027102577e-05, "loss": 0.0173, "step": 8274 }, { "epoch": 0.5686503573391973, "grad_norm": 0.26219695806503296, "learning_rate": 1.2407333643063752e-05, "loss": 0.0134, "step": 8275 }, { "epoch": 0.5687190764156129, "grad_norm": 0.6029027700424194, "learning_rate": 1.2404045387460292e-05, "loss": 0.0086, "step": 8276 }, { "epoch": 0.5687877954920286, "grad_norm": 0.40555498003959656, "learning_rate": 1.2400757260455085e-05, "loss": 0.0188, "step": 8277 }, { "epoch": 0.5688565145684442, "grad_norm": 0.35862937569618225, "learning_rate": 1.2397469262211022e-05, "loss": 0.0151, "step": 8278 }, { "epoch": 0.5689252336448598, "grad_norm": 0.5535497665405273, "learning_rate": 1.2394181392890978e-05, "loss": 0.0136, "step": 8279 }, { "epoch": 0.5689939527212754, "grad_norm": 0.633061945438385, "learning_rate": 1.2390893652657833e-05, "loss": 0.0221, "step": 8280 }, { "epoch": 0.569062671797691, "grad_norm": 0.34479761123657227, "learning_rate": 1.2387606041674453e-05, "loss": 0.0189, "step": 8281 }, { "epoch": 0.5691313908741067, "grad_norm": 0.7995500564575195, "learning_rate": 1.2384318560103691e-05, "loss": 0.0146, "step": 8282 }, { "epoch": 0.5692001099505223, "grad_norm": 0.5325205326080322, "learning_rate": 1.2381031208108416e-05, "loss": 0.0245, "step": 8283 }, { "epoch": 0.5692688290269379, "grad_norm": 0.8772236704826355, "learning_rate": 1.2377743985851472e-05, "loss": 0.0166, "step": 8284 }, { "epoch": 0.5693375481033535, "grad_norm": 0.28902631998062134, "learning_rate": 1.2374456893495697e-05, "loss": 0.0099, "step": 8285 }, { "epoch": 0.569406267179769, "grad_norm": 0.5817995071411133, "learning_rate": 1.2371169931203932e-05, "loss": 0.0153, "step": 8286 }, { "epoch": 0.5694749862561848, "grad_norm": 0.5641685724258423, "learning_rate": 1.2367883099139004e-05, "loss": 0.015, "step": 8287 }, { "epoch": 0.5695437053326003, "grad_norm": 0.7297329306602478, "learning_rate": 1.2364596397463728e-05, "loss": 0.0183, "step": 8288 }, { "epoch": 0.5696124244090159, "grad_norm": 0.24855566024780273, "learning_rate": 1.2361309826340936e-05, "loss": 0.0134, "step": 8289 }, { "epoch": 0.5696811434854315, "grad_norm": 0.4899178445339203, "learning_rate": 1.2358023385933429e-05, "loss": 0.0136, "step": 8290 }, { "epoch": 0.5697498625618471, "grad_norm": 0.4420184791088104, "learning_rate": 1.2354737076404013e-05, "loss": 0.0186, "step": 8291 }, { "epoch": 0.5698185816382628, "grad_norm": 0.5757070779800415, "learning_rate": 1.2351450897915484e-05, "loss": 0.0242, "step": 8292 }, { "epoch": 0.5698873007146784, "grad_norm": 0.6013815999031067, "learning_rate": 1.2348164850630633e-05, "loss": 0.0147, "step": 8293 }, { "epoch": 0.569956019791094, "grad_norm": 0.7179673314094543, "learning_rate": 1.2344878934712239e-05, "loss": 0.0169, "step": 8294 }, { "epoch": 0.5700247388675096, "grad_norm": 0.5490151643753052, "learning_rate": 1.2341593150323084e-05, "loss": 0.0206, "step": 8295 }, { "epoch": 0.5700934579439252, "grad_norm": 0.35784801840782166, "learning_rate": 1.2338307497625943e-05, "loss": 0.0161, "step": 8296 }, { "epoch": 0.5701621770203409, "grad_norm": 0.4592772424221039, "learning_rate": 1.2335021976783577e-05, "loss": 0.0204, "step": 8297 }, { "epoch": 0.5702308960967565, "grad_norm": 0.4049617052078247, "learning_rate": 1.233173658795874e-05, "loss": 0.0138, "step": 8298 }, { "epoch": 0.5702996151731721, "grad_norm": 1.1616382598876953, "learning_rate": 1.2328451331314188e-05, "loss": 0.0204, "step": 8299 }, { "epoch": 0.5703683342495877, "grad_norm": 0.6409207582473755, "learning_rate": 1.2325166207012659e-05, "loss": 0.0141, "step": 8300 }, { "epoch": 0.5704370533260033, "grad_norm": 0.7637501358985901, "learning_rate": 1.2321881215216902e-05, "loss": 0.0164, "step": 8301 }, { "epoch": 0.570505772402419, "grad_norm": 1.353110909461975, "learning_rate": 1.2318596356089646e-05, "loss": 0.0248, "step": 8302 }, { "epoch": 0.5705744914788345, "grad_norm": 0.3612572252750397, "learning_rate": 1.2315311629793612e-05, "loss": 0.0176, "step": 8303 }, { "epoch": 0.5706432105552501, "grad_norm": 0.3657352030277252, "learning_rate": 1.231202703649152e-05, "loss": 0.0123, "step": 8304 }, { "epoch": 0.5707119296316657, "grad_norm": 0.19285473227500916, "learning_rate": 1.2308742576346084e-05, "loss": 0.0132, "step": 8305 }, { "epoch": 0.5707806487080813, "grad_norm": 0.6197502613067627, "learning_rate": 1.2305458249519998e-05, "loss": 0.0197, "step": 8306 }, { "epoch": 0.570849367784497, "grad_norm": 0.6104012131690979, "learning_rate": 1.230217405617598e-05, "loss": 0.0154, "step": 8307 }, { "epoch": 0.5709180868609126, "grad_norm": 0.36829259991645813, "learning_rate": 1.2298889996476713e-05, "loss": 0.01, "step": 8308 }, { "epoch": 0.5709868059373282, "grad_norm": 0.9544649720191956, "learning_rate": 1.2295606070584881e-05, "loss": 0.0202, "step": 8309 }, { "epoch": 0.5710555250137438, "grad_norm": 0.5237520337104797, "learning_rate": 1.2292322278663165e-05, "loss": 0.025, "step": 8310 }, { "epoch": 0.5711242440901594, "grad_norm": 1.4079581499099731, "learning_rate": 1.2289038620874237e-05, "loss": 0.0327, "step": 8311 }, { "epoch": 0.5711929631665751, "grad_norm": 0.47016772627830505, "learning_rate": 1.2285755097380758e-05, "loss": 0.0204, "step": 8312 }, { "epoch": 0.5712616822429907, "grad_norm": 0.5250635743141174, "learning_rate": 1.2282471708345394e-05, "loss": 0.0144, "step": 8313 }, { "epoch": 0.5713304013194063, "grad_norm": 0.8963950872421265, "learning_rate": 1.2279188453930794e-05, "loss": 0.0161, "step": 8314 }, { "epoch": 0.5713991203958219, "grad_norm": 0.7877215147018433, "learning_rate": 1.2275905334299605e-05, "loss": 0.0247, "step": 8315 }, { "epoch": 0.5714678394722374, "grad_norm": 0.5555488467216492, "learning_rate": 1.2272622349614467e-05, "loss": 0.0152, "step": 8316 }, { "epoch": 0.5715365585486532, "grad_norm": 0.7257394194602966, "learning_rate": 1.2269339500038009e-05, "loss": 0.017, "step": 8317 }, { "epoch": 0.5716052776250687, "grad_norm": 0.7372589111328125, "learning_rate": 1.2266056785732852e-05, "loss": 0.0131, "step": 8318 }, { "epoch": 0.5716739967014843, "grad_norm": 1.180911660194397, "learning_rate": 1.2262774206861627e-05, "loss": 0.0179, "step": 8319 }, { "epoch": 0.5717427157778999, "grad_norm": 0.7198290228843689, "learning_rate": 1.2259491763586941e-05, "loss": 0.0136, "step": 8320 }, { "epoch": 0.5718114348543155, "grad_norm": 0.7679315805435181, "learning_rate": 1.2256209456071398e-05, "loss": 0.0186, "step": 8321 }, { "epoch": 0.5718801539307312, "grad_norm": 0.3680988848209381, "learning_rate": 1.2252927284477599e-05, "loss": 0.0182, "step": 8322 }, { "epoch": 0.5719488730071468, "grad_norm": 0.38009029626846313, "learning_rate": 1.2249645248968133e-05, "loss": 0.0133, "step": 8323 }, { "epoch": 0.5720175920835624, "grad_norm": 0.5153495073318481, "learning_rate": 1.2246363349705581e-05, "loss": 0.0165, "step": 8324 }, { "epoch": 0.572086311159978, "grad_norm": 1.767237663269043, "learning_rate": 1.2243081586852535e-05, "loss": 0.0222, "step": 8325 }, { "epoch": 0.5721550302363936, "grad_norm": 0.6634926795959473, "learning_rate": 1.2239799960571557e-05, "loss": 0.0147, "step": 8326 }, { "epoch": 0.5722237493128093, "grad_norm": 0.9387981295585632, "learning_rate": 1.2236518471025216e-05, "loss": 0.018, "step": 8327 }, { "epoch": 0.5722924683892249, "grad_norm": 0.45575109124183655, "learning_rate": 1.2233237118376068e-05, "loss": 0.0196, "step": 8328 }, { "epoch": 0.5723611874656405, "grad_norm": 0.40048566460609436, "learning_rate": 1.2229955902786665e-05, "loss": 0.018, "step": 8329 }, { "epoch": 0.572429906542056, "grad_norm": 0.47784098982810974, "learning_rate": 1.2226674824419548e-05, "loss": 0.0108, "step": 8330 }, { "epoch": 0.5724986256184716, "grad_norm": 0.5317391157150269, "learning_rate": 1.2223393883437263e-05, "loss": 0.015, "step": 8331 }, { "epoch": 0.5725673446948873, "grad_norm": 0.5787709951400757, "learning_rate": 1.2220113080002335e-05, "loss": 0.0132, "step": 8332 }, { "epoch": 0.5726360637713029, "grad_norm": 0.7921081185340881, "learning_rate": 1.221683241427729e-05, "loss": 0.0138, "step": 8333 }, { "epoch": 0.5727047828477185, "grad_norm": 0.8230943083763123, "learning_rate": 1.2213551886424643e-05, "loss": 0.0145, "step": 8334 }, { "epoch": 0.5727735019241341, "grad_norm": 0.45392438769340515, "learning_rate": 1.2210271496606907e-05, "loss": 0.0208, "step": 8335 }, { "epoch": 0.5728422210005497, "grad_norm": 0.3552117645740509, "learning_rate": 1.2206991244986586e-05, "loss": 0.0174, "step": 8336 }, { "epoch": 0.5729109400769654, "grad_norm": 0.3667294979095459, "learning_rate": 1.220371113172618e-05, "loss": 0.0168, "step": 8337 }, { "epoch": 0.572979659153381, "grad_norm": 0.9776783585548401, "learning_rate": 1.2200431156988172e-05, "loss": 0.0254, "step": 8338 }, { "epoch": 0.5730483782297966, "grad_norm": 0.823885977268219, "learning_rate": 1.2197151320935051e-05, "loss": 0.017, "step": 8339 }, { "epoch": 0.5731170973062122, "grad_norm": 1.0136890411376953, "learning_rate": 1.2193871623729293e-05, "loss": 0.0226, "step": 8340 }, { "epoch": 0.5731858163826278, "grad_norm": 0.9811099171638489, "learning_rate": 1.2190592065533362e-05, "loss": 0.0183, "step": 8341 }, { "epoch": 0.5732545354590435, "grad_norm": 0.3994423449039459, "learning_rate": 1.218731264650972e-05, "loss": 0.0168, "step": 8342 }, { "epoch": 0.5733232545354591, "grad_norm": 1.0005499124526978, "learning_rate": 1.2184033366820832e-05, "loss": 0.0146, "step": 8343 }, { "epoch": 0.5733919736118747, "grad_norm": 0.3212784230709076, "learning_rate": 1.2180754226629143e-05, "loss": 0.0151, "step": 8344 }, { "epoch": 0.5734606926882903, "grad_norm": 0.8003919720649719, "learning_rate": 1.2177475226097092e-05, "loss": 0.018, "step": 8345 }, { "epoch": 0.5735294117647058, "grad_norm": 0.2726030647754669, "learning_rate": 1.2174196365387117e-05, "loss": 0.0116, "step": 8346 }, { "epoch": 0.5735981308411215, "grad_norm": 0.35308465361595154, "learning_rate": 1.2170917644661644e-05, "loss": 0.0137, "step": 8347 }, { "epoch": 0.5736668499175371, "grad_norm": 0.42796117067337036, "learning_rate": 1.2167639064083091e-05, "loss": 0.0113, "step": 8348 }, { "epoch": 0.5737355689939527, "grad_norm": 0.826021134853363, "learning_rate": 1.216436062381388e-05, "loss": 0.0171, "step": 8349 }, { "epoch": 0.5738042880703683, "grad_norm": 0.9429311156272888, "learning_rate": 1.2161082324016411e-05, "loss": 0.0222, "step": 8350 }, { "epoch": 0.5738730071467839, "grad_norm": 0.5336331725120544, "learning_rate": 1.215780416485309e-05, "loss": 0.0118, "step": 8351 }, { "epoch": 0.5739417262231996, "grad_norm": 0.8189389705657959, "learning_rate": 1.2154526146486303e-05, "loss": 0.014, "step": 8352 }, { "epoch": 0.5740104452996152, "grad_norm": 0.7900857329368591, "learning_rate": 1.2151248269078443e-05, "loss": 0.0289, "step": 8353 }, { "epoch": 0.5740791643760308, "grad_norm": 0.5204759836196899, "learning_rate": 1.2147970532791885e-05, "loss": 0.0181, "step": 8354 }, { "epoch": 0.5741478834524464, "grad_norm": 0.5925857424736023, "learning_rate": 1.2144692937789008e-05, "loss": 0.0118, "step": 8355 }, { "epoch": 0.574216602528862, "grad_norm": 0.7115470170974731, "learning_rate": 1.214141548423217e-05, "loss": 0.013, "step": 8356 }, { "epoch": 0.5742853216052777, "grad_norm": 0.7453801035881042, "learning_rate": 1.2138138172283732e-05, "loss": 0.0108, "step": 8357 }, { "epoch": 0.5743540406816933, "grad_norm": 0.5349806547164917, "learning_rate": 1.2134861002106045e-05, "loss": 0.0143, "step": 8358 }, { "epoch": 0.5744227597581089, "grad_norm": 0.54154372215271, "learning_rate": 1.2131583973861455e-05, "loss": 0.0129, "step": 8359 }, { "epoch": 0.5744914788345244, "grad_norm": 0.31399303674697876, "learning_rate": 1.212830708771229e-05, "loss": 0.0188, "step": 8360 }, { "epoch": 0.57456019791094, "grad_norm": 0.8561282753944397, "learning_rate": 1.2125030343820896e-05, "loss": 0.0175, "step": 8361 }, { "epoch": 0.5746289169873557, "grad_norm": 0.7224067449569702, "learning_rate": 1.2121753742349586e-05, "loss": 0.0126, "step": 8362 }, { "epoch": 0.5746976360637713, "grad_norm": 0.5670844912528992, "learning_rate": 1.211847728346068e-05, "loss": 0.0108, "step": 8363 }, { "epoch": 0.5747663551401869, "grad_norm": 0.23140452802181244, "learning_rate": 1.2115200967316485e-05, "loss": 0.0179, "step": 8364 }, { "epoch": 0.5748350742166025, "grad_norm": 1.627331018447876, "learning_rate": 1.2111924794079303e-05, "loss": 0.0136, "step": 8365 }, { "epoch": 0.5749037932930181, "grad_norm": 0.2638988792896271, "learning_rate": 1.2108648763911425e-05, "loss": 0.0198, "step": 8366 }, { "epoch": 0.5749725123694338, "grad_norm": 0.2760718762874603, "learning_rate": 1.2105372876975146e-05, "loss": 0.0197, "step": 8367 }, { "epoch": 0.5750412314458494, "grad_norm": 0.3292677700519562, "learning_rate": 1.2102097133432742e-05, "loss": 0.0165, "step": 8368 }, { "epoch": 0.575109950522265, "grad_norm": 0.3155944347381592, "learning_rate": 1.2098821533446489e-05, "loss": 0.0102, "step": 8369 }, { "epoch": 0.5751786695986806, "grad_norm": 0.958402156829834, "learning_rate": 1.2095546077178654e-05, "loss": 0.0129, "step": 8370 }, { "epoch": 0.5752473886750962, "grad_norm": 0.6301891207695007, "learning_rate": 1.2092270764791495e-05, "loss": 0.0145, "step": 8371 }, { "epoch": 0.5753161077515119, "grad_norm": 0.8118458390235901, "learning_rate": 1.2088995596447263e-05, "loss": 0.0097, "step": 8372 }, { "epoch": 0.5753848268279275, "grad_norm": 1.384749412536621, "learning_rate": 1.2085720572308205e-05, "loss": 0.017, "step": 8373 }, { "epoch": 0.5754535459043431, "grad_norm": 0.4898311495780945, "learning_rate": 1.208244569253656e-05, "loss": 0.0158, "step": 8374 }, { "epoch": 0.5755222649807586, "grad_norm": 0.654291570186615, "learning_rate": 1.2079170957294554e-05, "loss": 0.0228, "step": 8375 }, { "epoch": 0.5755909840571742, "grad_norm": 0.4067382216453552, "learning_rate": 1.2075896366744417e-05, "loss": 0.0499, "step": 8376 }, { "epoch": 0.5756597031335899, "grad_norm": 1.0864237546920776, "learning_rate": 1.2072621921048363e-05, "loss": 0.0164, "step": 8377 }, { "epoch": 0.5757284222100055, "grad_norm": 0.45147302746772766, "learning_rate": 1.206934762036859e-05, "loss": 0.0164, "step": 8378 }, { "epoch": 0.5757971412864211, "grad_norm": 0.308692991733551, "learning_rate": 1.206607346486732e-05, "loss": 0.0167, "step": 8379 }, { "epoch": 0.5758658603628367, "grad_norm": 0.3780401051044464, "learning_rate": 1.2062799454706739e-05, "loss": 0.0153, "step": 8380 }, { "epoch": 0.5759345794392523, "grad_norm": 0.3850342929363251, "learning_rate": 1.2059525590049035e-05, "loss": 0.0069, "step": 8381 }, { "epoch": 0.5760032985156679, "grad_norm": 0.5201504826545715, "learning_rate": 1.2056251871056385e-05, "loss": 0.0154, "step": 8382 }, { "epoch": 0.5760720175920836, "grad_norm": 0.41436269879341125, "learning_rate": 1.2052978297890967e-05, "loss": 0.0112, "step": 8383 }, { "epoch": 0.5761407366684992, "grad_norm": 0.4244861602783203, "learning_rate": 1.2049704870714942e-05, "loss": 0.0228, "step": 8384 }, { "epoch": 0.5762094557449148, "grad_norm": 0.49397799372673035, "learning_rate": 1.2046431589690472e-05, "loss": 0.0164, "step": 8385 }, { "epoch": 0.5762781748213304, "grad_norm": 0.7166316509246826, "learning_rate": 1.2043158454979713e-05, "loss": 0.0181, "step": 8386 }, { "epoch": 0.576346893897746, "grad_norm": 0.3697710633277893, "learning_rate": 1.2039885466744804e-05, "loss": 0.0249, "step": 8387 }, { "epoch": 0.5764156129741617, "grad_norm": 0.5229414701461792, "learning_rate": 1.2036612625147881e-05, "loss": 0.0142, "step": 8388 }, { "epoch": 0.5764843320505773, "grad_norm": 0.44380155205726624, "learning_rate": 1.203333993035108e-05, "loss": 0.0112, "step": 8389 }, { "epoch": 0.5765530511269928, "grad_norm": 0.5099351406097412, "learning_rate": 1.2030067382516515e-05, "loss": 0.0092, "step": 8390 }, { "epoch": 0.5766217702034084, "grad_norm": 0.2927647829055786, "learning_rate": 1.202679498180631e-05, "loss": 0.0122, "step": 8391 }, { "epoch": 0.576690489279824, "grad_norm": 0.3636585772037506, "learning_rate": 1.2023522728382565e-05, "loss": 0.0069, "step": 8392 }, { "epoch": 0.5767592083562397, "grad_norm": 0.5252871513366699, "learning_rate": 1.202025062240739e-05, "loss": 0.0209, "step": 8393 }, { "epoch": 0.5768279274326553, "grad_norm": 0.42963355779647827, "learning_rate": 1.201697866404287e-05, "loss": 0.0153, "step": 8394 }, { "epoch": 0.5768966465090709, "grad_norm": 0.7114558815956116, "learning_rate": 1.2013706853451093e-05, "loss": 0.0136, "step": 8395 }, { "epoch": 0.5769653655854865, "grad_norm": 0.7392740249633789, "learning_rate": 1.2010435190794136e-05, "loss": 0.0179, "step": 8396 }, { "epoch": 0.5770340846619021, "grad_norm": 0.7836253643035889, "learning_rate": 1.2007163676234078e-05, "loss": 0.0144, "step": 8397 }, { "epoch": 0.5771028037383178, "grad_norm": 0.35438916087150574, "learning_rate": 1.2003892309932981e-05, "loss": 0.0097, "step": 8398 }, { "epoch": 0.5771715228147334, "grad_norm": 0.3004276752471924, "learning_rate": 1.2000621092052895e-05, "loss": 0.0123, "step": 8399 }, { "epoch": 0.577240241891149, "grad_norm": 0.6647160649299622, "learning_rate": 1.1997350022755875e-05, "loss": 0.0173, "step": 8400 }, { "epoch": 0.5773089609675646, "grad_norm": 0.37321504950523376, "learning_rate": 1.1994079102203961e-05, "loss": 0.0116, "step": 8401 }, { "epoch": 0.5773776800439802, "grad_norm": 0.8904221653938293, "learning_rate": 1.1990808330559184e-05, "loss": 0.015, "step": 8402 }, { "epoch": 0.5774463991203959, "grad_norm": 0.8773605823516846, "learning_rate": 1.1987537707983581e-05, "loss": 0.0234, "step": 8403 }, { "epoch": 0.5775151181968114, "grad_norm": 0.4414599537849426, "learning_rate": 1.1984267234639168e-05, "loss": 0.0101, "step": 8404 }, { "epoch": 0.577583837273227, "grad_norm": 0.4985118508338928, "learning_rate": 1.1980996910687954e-05, "loss": 0.0152, "step": 8405 }, { "epoch": 0.5776525563496426, "grad_norm": 0.7211898565292358, "learning_rate": 1.1977726736291948e-05, "loss": 0.0158, "step": 8406 }, { "epoch": 0.5777212754260582, "grad_norm": 0.2462007850408554, "learning_rate": 1.1974456711613145e-05, "loss": 0.0181, "step": 8407 }, { "epoch": 0.5777899945024739, "grad_norm": 0.6525473594665527, "learning_rate": 1.1971186836813534e-05, "loss": 0.0113, "step": 8408 }, { "epoch": 0.5778587135788895, "grad_norm": 0.6144498586654663, "learning_rate": 1.1967917112055102e-05, "loss": 0.0315, "step": 8409 }, { "epoch": 0.5779274326553051, "grad_norm": 0.5561482310295105, "learning_rate": 1.1964647537499823e-05, "loss": 0.0201, "step": 8410 }, { "epoch": 0.5779961517317207, "grad_norm": 1.203456997871399, "learning_rate": 1.1961378113309664e-05, "loss": 0.0122, "step": 8411 }, { "epoch": 0.5780648708081363, "grad_norm": 0.5134778618812561, "learning_rate": 1.1958108839646588e-05, "loss": 0.0165, "step": 8412 }, { "epoch": 0.578133589884552, "grad_norm": 0.8000384569168091, "learning_rate": 1.1954839716672545e-05, "loss": 0.015, "step": 8413 }, { "epoch": 0.5782023089609676, "grad_norm": 0.6842458844184875, "learning_rate": 1.1951570744549474e-05, "loss": 0.0086, "step": 8414 }, { "epoch": 0.5782710280373832, "grad_norm": 0.37553486227989197, "learning_rate": 1.194830192343933e-05, "loss": 0.0149, "step": 8415 }, { "epoch": 0.5783397471137988, "grad_norm": 0.48639413714408875, "learning_rate": 1.1945033253504034e-05, "loss": 0.0153, "step": 8416 }, { "epoch": 0.5784084661902144, "grad_norm": 0.41047215461730957, "learning_rate": 1.1941764734905509e-05, "loss": 0.0173, "step": 8417 }, { "epoch": 0.5784771852666301, "grad_norm": 0.3138134479522705, "learning_rate": 1.1938496367805672e-05, "loss": 0.0123, "step": 8418 }, { "epoch": 0.5785459043430456, "grad_norm": 0.47507908940315247, "learning_rate": 1.193522815236643e-05, "loss": 0.0169, "step": 8419 }, { "epoch": 0.5786146234194612, "grad_norm": 0.43309128284454346, "learning_rate": 1.1931960088749679e-05, "loss": 0.0187, "step": 8420 }, { "epoch": 0.5786833424958768, "grad_norm": 0.4524582624435425, "learning_rate": 1.1928692177117323e-05, "loss": 0.0125, "step": 8421 }, { "epoch": 0.5787520615722924, "grad_norm": 0.5478574633598328, "learning_rate": 1.1925424417631243e-05, "loss": 0.0186, "step": 8422 }, { "epoch": 0.5788207806487081, "grad_norm": 0.39887312054634094, "learning_rate": 1.1922156810453317e-05, "loss": 0.0201, "step": 8423 }, { "epoch": 0.5788894997251237, "grad_norm": 0.19333867728710175, "learning_rate": 1.1918889355745415e-05, "loss": 0.014, "step": 8424 }, { "epoch": 0.5789582188015393, "grad_norm": 0.6413007974624634, "learning_rate": 1.1915622053669398e-05, "loss": 0.0161, "step": 8425 }, { "epoch": 0.5790269378779549, "grad_norm": 0.3812849819660187, "learning_rate": 1.191235490438712e-05, "loss": 0.0108, "step": 8426 }, { "epoch": 0.5790956569543705, "grad_norm": 0.7317949533462524, "learning_rate": 1.1909087908060436e-05, "loss": 0.0098, "step": 8427 }, { "epoch": 0.5791643760307862, "grad_norm": 0.9320334792137146, "learning_rate": 1.1905821064851183e-05, "loss": 0.018, "step": 8428 }, { "epoch": 0.5792330951072018, "grad_norm": 0.7763376235961914, "learning_rate": 1.1902554374921193e-05, "loss": 0.029, "step": 8429 }, { "epoch": 0.5793018141836174, "grad_norm": 0.7043965458869934, "learning_rate": 1.189928783843229e-05, "loss": 0.0161, "step": 8430 }, { "epoch": 0.579370533260033, "grad_norm": 0.4614019989967346, "learning_rate": 1.1896021455546296e-05, "loss": 0.0148, "step": 8431 }, { "epoch": 0.5794392523364486, "grad_norm": 1.5206125974655151, "learning_rate": 1.1892755226425008e-05, "loss": 0.0193, "step": 8432 }, { "epoch": 0.5795079714128643, "grad_norm": 0.7510109543800354, "learning_rate": 1.1889489151230243e-05, "loss": 0.0156, "step": 8433 }, { "epoch": 0.5795766904892798, "grad_norm": 0.6129646301269531, "learning_rate": 1.1886223230123794e-05, "loss": 0.024, "step": 8434 }, { "epoch": 0.5796454095656954, "grad_norm": 0.8578460812568665, "learning_rate": 1.1882957463267444e-05, "loss": 0.0195, "step": 8435 }, { "epoch": 0.579714128642111, "grad_norm": 0.7953170537948608, "learning_rate": 1.1879691850822972e-05, "loss": 0.0235, "step": 8436 }, { "epoch": 0.5797828477185266, "grad_norm": 0.4554215669631958, "learning_rate": 1.1876426392952152e-05, "loss": 0.0185, "step": 8437 }, { "epoch": 0.5798515667949423, "grad_norm": 0.2766353487968445, "learning_rate": 1.1873161089816738e-05, "loss": 0.0203, "step": 8438 }, { "epoch": 0.5799202858713579, "grad_norm": 0.3797537088394165, "learning_rate": 1.1869895941578503e-05, "loss": 0.0195, "step": 8439 }, { "epoch": 0.5799890049477735, "grad_norm": 0.6327986121177673, "learning_rate": 1.1866630948399187e-05, "loss": 0.0218, "step": 8440 }, { "epoch": 0.5800577240241891, "grad_norm": 0.5434760451316833, "learning_rate": 1.186336611044053e-05, "loss": 0.0148, "step": 8441 }, { "epoch": 0.5801264431006047, "grad_norm": 0.8459399342536926, "learning_rate": 1.1860101427864271e-05, "loss": 0.0182, "step": 8442 }, { "epoch": 0.5801951621770204, "grad_norm": 0.46800917387008667, "learning_rate": 1.1856836900832128e-05, "loss": 0.0107, "step": 8443 }, { "epoch": 0.580263881253436, "grad_norm": 1.3346445560455322, "learning_rate": 1.1853572529505823e-05, "loss": 0.0182, "step": 8444 }, { "epoch": 0.5803326003298516, "grad_norm": 0.6081026792526245, "learning_rate": 1.1850308314047068e-05, "loss": 0.0301, "step": 8445 }, { "epoch": 0.5804013194062672, "grad_norm": 0.8008357882499695, "learning_rate": 1.1847044254617566e-05, "loss": 0.0144, "step": 8446 }, { "epoch": 0.5804700384826827, "grad_norm": 0.4899239242076874, "learning_rate": 1.1843780351379006e-05, "loss": 0.0167, "step": 8447 }, { "epoch": 0.5805387575590985, "grad_norm": 0.5661494731903076, "learning_rate": 1.1840516604493079e-05, "loss": 0.021, "step": 8448 }, { "epoch": 0.580607476635514, "grad_norm": 0.563307523727417, "learning_rate": 1.1837253014121465e-05, "loss": 0.0167, "step": 8449 }, { "epoch": 0.5806761957119296, "grad_norm": 0.4515797793865204, "learning_rate": 1.1833989580425828e-05, "loss": 0.0133, "step": 8450 }, { "epoch": 0.5807449147883452, "grad_norm": 0.2263542264699936, "learning_rate": 1.1830726303567844e-05, "loss": 0.0142, "step": 8451 }, { "epoch": 0.5808136338647608, "grad_norm": 0.4675089120864868, "learning_rate": 1.1827463183709165e-05, "loss": 0.0186, "step": 8452 }, { "epoch": 0.5808823529411765, "grad_norm": 0.3701145648956299, "learning_rate": 1.1824200221011434e-05, "loss": 0.0119, "step": 8453 }, { "epoch": 0.5809510720175921, "grad_norm": 0.49052947759628296, "learning_rate": 1.1820937415636296e-05, "loss": 0.0148, "step": 8454 }, { "epoch": 0.5810197910940077, "grad_norm": 0.5885790586471558, "learning_rate": 1.1817674767745384e-05, "loss": 0.0109, "step": 8455 }, { "epoch": 0.5810885101704233, "grad_norm": 0.7488649487495422, "learning_rate": 1.1814412277500314e-05, "loss": 0.0165, "step": 8456 }, { "epoch": 0.5811572292468389, "grad_norm": 1.013636589050293, "learning_rate": 1.1811149945062717e-05, "loss": 0.0223, "step": 8457 }, { "epoch": 0.5812259483232546, "grad_norm": 0.6015159487724304, "learning_rate": 1.1807887770594196e-05, "loss": 0.0106, "step": 8458 }, { "epoch": 0.5812946673996702, "grad_norm": 0.32053330540657043, "learning_rate": 1.180462575425635e-05, "loss": 0.0086, "step": 8459 }, { "epoch": 0.5813633864760858, "grad_norm": 0.8793719410896301, "learning_rate": 1.1801363896210778e-05, "loss": 0.0186, "step": 8460 }, { "epoch": 0.5814321055525014, "grad_norm": 0.5219104886054993, "learning_rate": 1.1798102196619061e-05, "loss": 0.0143, "step": 8461 }, { "epoch": 0.581500824628917, "grad_norm": 0.38055557012557983, "learning_rate": 1.1794840655642775e-05, "loss": 0.0106, "step": 8462 }, { "epoch": 0.5815695437053326, "grad_norm": 1.26130211353302, "learning_rate": 1.1791579273443498e-05, "loss": 0.0119, "step": 8463 }, { "epoch": 0.5816382627817482, "grad_norm": 0.32438039779663086, "learning_rate": 1.1788318050182785e-05, "loss": 0.0134, "step": 8464 }, { "epoch": 0.5817069818581638, "grad_norm": 0.5504477620124817, "learning_rate": 1.1785056986022194e-05, "loss": 0.0179, "step": 8465 }, { "epoch": 0.5817757009345794, "grad_norm": 0.29085013270378113, "learning_rate": 1.1781796081123271e-05, "loss": 0.0177, "step": 8466 }, { "epoch": 0.581844420010995, "grad_norm": 0.49901050329208374, "learning_rate": 1.1778535335647554e-05, "loss": 0.0101, "step": 8467 }, { "epoch": 0.5819131390874107, "grad_norm": 0.47392183542251587, "learning_rate": 1.1775274749756564e-05, "loss": 0.0094, "step": 8468 }, { "epoch": 0.5819818581638263, "grad_norm": 0.5230134129524231, "learning_rate": 1.1772014323611843e-05, "loss": 0.0155, "step": 8469 }, { "epoch": 0.5820505772402419, "grad_norm": 0.667624294757843, "learning_rate": 1.1768754057374894e-05, "loss": 0.0221, "step": 8470 }, { "epoch": 0.5821192963166575, "grad_norm": 0.6537376642227173, "learning_rate": 1.1765493951207227e-05, "loss": 0.0271, "step": 8471 }, { "epoch": 0.5821880153930731, "grad_norm": 0.2526964545249939, "learning_rate": 1.176223400527034e-05, "loss": 0.0122, "step": 8472 }, { "epoch": 0.5822567344694888, "grad_norm": 0.32676658034324646, "learning_rate": 1.1758974219725723e-05, "loss": 0.018, "step": 8473 }, { "epoch": 0.5823254535459044, "grad_norm": 0.7719643712043762, "learning_rate": 1.1755714594734853e-05, "loss": 0.0089, "step": 8474 }, { "epoch": 0.58239417262232, "grad_norm": 0.46977153420448303, "learning_rate": 1.175245513045922e-05, "loss": 0.0107, "step": 8475 }, { "epoch": 0.5824628916987356, "grad_norm": 0.3320413827896118, "learning_rate": 1.1749195827060284e-05, "loss": 0.0135, "step": 8476 }, { "epoch": 0.5825316107751511, "grad_norm": 0.31912583112716675, "learning_rate": 1.17459366846995e-05, "loss": 0.0083, "step": 8477 }, { "epoch": 0.5826003298515668, "grad_norm": 0.822087287902832, "learning_rate": 1.1742677703538325e-05, "loss": 0.0118, "step": 8478 }, { "epoch": 0.5826690489279824, "grad_norm": 0.4844628870487213, "learning_rate": 1.17394188837382e-05, "loss": 0.0205, "step": 8479 }, { "epoch": 0.582737768004398, "grad_norm": 0.4116036295890808, "learning_rate": 1.1736160225460555e-05, "loss": 0.0083, "step": 8480 }, { "epoch": 0.5828064870808136, "grad_norm": 0.24904392659664154, "learning_rate": 1.1732901728866827e-05, "loss": 0.0145, "step": 8481 }, { "epoch": 0.5828752061572292, "grad_norm": 0.7958173751831055, "learning_rate": 1.1729643394118428e-05, "loss": 0.0227, "step": 8482 }, { "epoch": 0.5829439252336449, "grad_norm": 0.5814969539642334, "learning_rate": 1.1726385221376774e-05, "loss": 0.0261, "step": 8483 }, { "epoch": 0.5830126443100605, "grad_norm": 0.4890790283679962, "learning_rate": 1.1723127210803265e-05, "loss": 0.0091, "step": 8484 }, { "epoch": 0.5830813633864761, "grad_norm": 0.6016812324523926, "learning_rate": 1.1719869362559297e-05, "loss": 0.012, "step": 8485 }, { "epoch": 0.5831500824628917, "grad_norm": 0.9666731953620911, "learning_rate": 1.1716611676806251e-05, "loss": 0.0128, "step": 8486 }, { "epoch": 0.5832188015393073, "grad_norm": 0.6280749440193176, "learning_rate": 1.1713354153705519e-05, "loss": 0.0295, "step": 8487 }, { "epoch": 0.583287520615723, "grad_norm": 0.222176194190979, "learning_rate": 1.1710096793418464e-05, "loss": 0.013, "step": 8488 }, { "epoch": 0.5833562396921386, "grad_norm": 0.42421525716781616, "learning_rate": 1.1706839596106453e-05, "loss": 0.0119, "step": 8489 }, { "epoch": 0.5834249587685542, "grad_norm": 0.5898285508155823, "learning_rate": 1.1703582561930836e-05, "loss": 0.0239, "step": 8490 }, { "epoch": 0.5834936778449697, "grad_norm": 0.36977243423461914, "learning_rate": 1.170032569105296e-05, "loss": 0.0162, "step": 8491 }, { "epoch": 0.5835623969213853, "grad_norm": 0.9842788577079773, "learning_rate": 1.1697068983634162e-05, "loss": 0.0235, "step": 8492 }, { "epoch": 0.583631115997801, "grad_norm": 0.6097045540809631, "learning_rate": 1.169381243983578e-05, "loss": 0.0148, "step": 8493 }, { "epoch": 0.5836998350742166, "grad_norm": 0.36934152245521545, "learning_rate": 1.1690556059819133e-05, "loss": 0.0175, "step": 8494 }, { "epoch": 0.5837685541506322, "grad_norm": 0.7243824601173401, "learning_rate": 1.1687299843745536e-05, "loss": 0.0186, "step": 8495 }, { "epoch": 0.5838372732270478, "grad_norm": 0.7713558673858643, "learning_rate": 1.1684043791776295e-05, "loss": 0.0191, "step": 8496 }, { "epoch": 0.5839059923034634, "grad_norm": 0.6670237183570862, "learning_rate": 1.1680787904072704e-05, "loss": 0.0121, "step": 8497 }, { "epoch": 0.5839747113798791, "grad_norm": 1.3667974472045898, "learning_rate": 1.1677532180796054e-05, "loss": 0.0172, "step": 8498 }, { "epoch": 0.5840434304562947, "grad_norm": 0.25099384784698486, "learning_rate": 1.1674276622107633e-05, "loss": 0.0107, "step": 8499 }, { "epoch": 0.5841121495327103, "grad_norm": 0.561526894569397, "learning_rate": 1.1671021228168712e-05, "loss": 0.0154, "step": 8500 }, { "epoch": 0.5841808686091259, "grad_norm": 0.35976627469062805, "learning_rate": 1.1667765999140554e-05, "loss": 0.0124, "step": 8501 }, { "epoch": 0.5842495876855415, "grad_norm": 0.5831795930862427, "learning_rate": 1.1664510935184417e-05, "loss": 0.014, "step": 8502 }, { "epoch": 0.5843183067619572, "grad_norm": 0.4265791177749634, "learning_rate": 1.1661256036461552e-05, "loss": 0.0147, "step": 8503 }, { "epoch": 0.5843870258383728, "grad_norm": 0.5270523428916931, "learning_rate": 1.1658001303133199e-05, "loss": 0.0171, "step": 8504 }, { "epoch": 0.5844557449147884, "grad_norm": 0.3262160122394562, "learning_rate": 1.1654746735360582e-05, "loss": 0.0129, "step": 8505 }, { "epoch": 0.584524463991204, "grad_norm": 0.25984102487564087, "learning_rate": 1.1651492333304942e-05, "loss": 0.0124, "step": 8506 }, { "epoch": 0.5845931830676195, "grad_norm": 1.7176036834716797, "learning_rate": 1.1648238097127488e-05, "loss": 0.0144, "step": 8507 }, { "epoch": 0.5846619021440352, "grad_norm": 0.7760698795318604, "learning_rate": 1.1644984026989428e-05, "loss": 0.0178, "step": 8508 }, { "epoch": 0.5847306212204508, "grad_norm": 0.3617539405822754, "learning_rate": 1.1641730123051962e-05, "loss": 0.0058, "step": 8509 }, { "epoch": 0.5847993402968664, "grad_norm": 0.4834834933280945, "learning_rate": 1.163847638547628e-05, "loss": 0.016, "step": 8510 }, { "epoch": 0.584868059373282, "grad_norm": 0.8205797672271729, "learning_rate": 1.163522281442356e-05, "loss": 0.0191, "step": 8511 }, { "epoch": 0.5849367784496976, "grad_norm": 0.3363450765609741, "learning_rate": 1.163196941005499e-05, "loss": 0.0079, "step": 8512 }, { "epoch": 0.5850054975261132, "grad_norm": 0.42573654651641846, "learning_rate": 1.1628716172531734e-05, "loss": 0.0179, "step": 8513 }, { "epoch": 0.5850742166025289, "grad_norm": 0.3436484932899475, "learning_rate": 1.1625463102014945e-05, "loss": 0.0134, "step": 8514 }, { "epoch": 0.5851429356789445, "grad_norm": 0.8609461784362793, "learning_rate": 1.1622210198665778e-05, "loss": 0.0146, "step": 8515 }, { "epoch": 0.5852116547553601, "grad_norm": 1.0328454971313477, "learning_rate": 1.1618957462645373e-05, "loss": 0.0167, "step": 8516 }, { "epoch": 0.5852803738317757, "grad_norm": 0.2758745551109314, "learning_rate": 1.1615704894114859e-05, "loss": 0.0138, "step": 8517 }, { "epoch": 0.5853490929081913, "grad_norm": 0.09216754883527756, "learning_rate": 1.1612452493235372e-05, "loss": 0.0101, "step": 8518 }, { "epoch": 0.585417811984607, "grad_norm": 0.22508607804775238, "learning_rate": 1.1609200260168024e-05, "loss": 0.013, "step": 8519 }, { "epoch": 0.5854865310610226, "grad_norm": 0.47191429138183594, "learning_rate": 1.1605948195073924e-05, "loss": 0.018, "step": 8520 }, { "epoch": 0.5855552501374381, "grad_norm": 0.3621039092540741, "learning_rate": 1.1602696298114171e-05, "loss": 0.0121, "step": 8521 }, { "epoch": 0.5856239692138537, "grad_norm": 0.24171702563762665, "learning_rate": 1.159944456944986e-05, "loss": 0.0129, "step": 8522 }, { "epoch": 0.5856926882902693, "grad_norm": 0.24824407696723938, "learning_rate": 1.1596193009242068e-05, "loss": 0.0087, "step": 8523 }, { "epoch": 0.585761407366685, "grad_norm": 0.295978307723999, "learning_rate": 1.1592941617651881e-05, "loss": 0.0119, "step": 8524 }, { "epoch": 0.5858301264431006, "grad_norm": 0.5324736833572388, "learning_rate": 1.1589690394840362e-05, "loss": 0.0121, "step": 8525 }, { "epoch": 0.5858988455195162, "grad_norm": 0.8215931057929993, "learning_rate": 1.1586439340968573e-05, "loss": 0.0124, "step": 8526 }, { "epoch": 0.5859675645959318, "grad_norm": 0.4020112454891205, "learning_rate": 1.1583188456197558e-05, "loss": 0.0118, "step": 8527 }, { "epoch": 0.5860362836723474, "grad_norm": 0.5457566976547241, "learning_rate": 1.157993774068836e-05, "loss": 0.0173, "step": 8528 }, { "epoch": 0.5861050027487631, "grad_norm": 0.2559633255004883, "learning_rate": 1.1576687194602011e-05, "loss": 0.0238, "step": 8529 }, { "epoch": 0.5861737218251787, "grad_norm": 0.25543341040611267, "learning_rate": 1.1573436818099549e-05, "loss": 0.0078, "step": 8530 }, { "epoch": 0.5862424409015943, "grad_norm": 0.7102159857749939, "learning_rate": 1.1570186611341979e-05, "loss": 0.018, "step": 8531 }, { "epoch": 0.5863111599780099, "grad_norm": 0.2816590666770935, "learning_rate": 1.1566936574490314e-05, "loss": 0.0119, "step": 8532 }, { "epoch": 0.5863798790544255, "grad_norm": 0.2726306617259979, "learning_rate": 1.1563686707705554e-05, "loss": 0.0083, "step": 8533 }, { "epoch": 0.5864485981308412, "grad_norm": 0.605180561542511, "learning_rate": 1.1560437011148688e-05, "loss": 0.0207, "step": 8534 }, { "epoch": 0.5865173172072567, "grad_norm": 0.6008281707763672, "learning_rate": 1.1557187484980699e-05, "loss": 0.0171, "step": 8535 }, { "epoch": 0.5865860362836723, "grad_norm": 0.8247111439704895, "learning_rate": 1.1553938129362567e-05, "loss": 0.0134, "step": 8536 }, { "epoch": 0.5866547553600879, "grad_norm": 0.6297664642333984, "learning_rate": 1.1550688944455253e-05, "loss": 0.0112, "step": 8537 }, { "epoch": 0.5867234744365035, "grad_norm": 0.6886032819747925, "learning_rate": 1.1547439930419721e-05, "loss": 0.0216, "step": 8538 }, { "epoch": 0.5867921935129192, "grad_norm": 0.44243136048316956, "learning_rate": 1.1544191087416915e-05, "loss": 0.0182, "step": 8539 }, { "epoch": 0.5868609125893348, "grad_norm": 0.5184319615364075, "learning_rate": 1.1540942415607778e-05, "loss": 0.01, "step": 8540 }, { "epoch": 0.5869296316657504, "grad_norm": 0.5155888199806213, "learning_rate": 1.1537693915153233e-05, "loss": 0.021, "step": 8541 }, { "epoch": 0.586998350742166, "grad_norm": 0.6775436401367188, "learning_rate": 1.1534445586214222e-05, "loss": 0.0169, "step": 8542 }, { "epoch": 0.5870670698185816, "grad_norm": 0.7840059399604797, "learning_rate": 1.1531197428951653e-05, "loss": 0.0236, "step": 8543 }, { "epoch": 0.5871357888949973, "grad_norm": 0.6749730706214905, "learning_rate": 1.1527949443526429e-05, "loss": 0.0258, "step": 8544 }, { "epoch": 0.5872045079714129, "grad_norm": 0.7544586658477783, "learning_rate": 1.1524701630099454e-05, "loss": 0.0232, "step": 8545 }, { "epoch": 0.5872732270478285, "grad_norm": 0.4830838143825531, "learning_rate": 1.1521453988831614e-05, "loss": 0.0192, "step": 8546 }, { "epoch": 0.5873419461242441, "grad_norm": 0.16271798312664032, "learning_rate": 1.1518206519883785e-05, "loss": 0.0106, "step": 8547 }, { "epoch": 0.5874106652006597, "grad_norm": 0.25959938764572144, "learning_rate": 1.1514959223416854e-05, "loss": 0.0084, "step": 8548 }, { "epoch": 0.5874793842770754, "grad_norm": 0.5461520552635193, "learning_rate": 1.1511712099591677e-05, "loss": 0.0184, "step": 8549 }, { "epoch": 0.587548103353491, "grad_norm": 0.5263417959213257, "learning_rate": 1.1508465148569112e-05, "loss": 0.0204, "step": 8550 }, { "epoch": 0.5876168224299065, "grad_norm": 0.8358708024024963, "learning_rate": 1.1505218370510006e-05, "loss": 0.0235, "step": 8551 }, { "epoch": 0.5876855415063221, "grad_norm": 0.6949005722999573, "learning_rate": 1.1501971765575196e-05, "loss": 0.0168, "step": 8552 }, { "epoch": 0.5877542605827377, "grad_norm": 0.645214855670929, "learning_rate": 1.149872533392551e-05, "loss": 0.018, "step": 8553 }, { "epoch": 0.5878229796591534, "grad_norm": 0.7112931609153748, "learning_rate": 1.1495479075721777e-05, "loss": 0.0153, "step": 8554 }, { "epoch": 0.587891698735569, "grad_norm": 0.6870256662368774, "learning_rate": 1.1492232991124804e-05, "loss": 0.0159, "step": 8555 }, { "epoch": 0.5879604178119846, "grad_norm": 0.2280021756887436, "learning_rate": 1.1488987080295395e-05, "loss": 0.0122, "step": 8556 }, { "epoch": 0.5880291368884002, "grad_norm": 1.914247751235962, "learning_rate": 1.1485741343394351e-05, "loss": 0.0217, "step": 8557 }, { "epoch": 0.5880978559648158, "grad_norm": 0.9292582273483276, "learning_rate": 1.1482495780582452e-05, "loss": 0.0179, "step": 8558 }, { "epoch": 0.5881665750412315, "grad_norm": 0.4250456988811493, "learning_rate": 1.1479250392020479e-05, "loss": 0.0197, "step": 8559 }, { "epoch": 0.5882352941176471, "grad_norm": 0.6197047829627991, "learning_rate": 1.147600517786921e-05, "loss": 0.0139, "step": 8560 }, { "epoch": 0.5883040131940627, "grad_norm": 0.8023362159729004, "learning_rate": 1.1472760138289396e-05, "loss": 0.0203, "step": 8561 }, { "epoch": 0.5883727322704783, "grad_norm": 0.8000980615615845, "learning_rate": 1.1469515273441794e-05, "loss": 0.0114, "step": 8562 }, { "epoch": 0.5884414513468939, "grad_norm": 0.7929165959358215, "learning_rate": 1.1466270583487147e-05, "loss": 0.0179, "step": 8563 }, { "epoch": 0.5885101704233096, "grad_norm": 0.6894000172615051, "learning_rate": 1.146302606858619e-05, "loss": 0.0197, "step": 8564 }, { "epoch": 0.5885788894997251, "grad_norm": 0.5378397703170776, "learning_rate": 1.1459781728899644e-05, "loss": 0.0156, "step": 8565 }, { "epoch": 0.5886476085761407, "grad_norm": 1.1732913255691528, "learning_rate": 1.1456537564588238e-05, "loss": 0.03, "step": 8566 }, { "epoch": 0.5887163276525563, "grad_norm": 0.5882346034049988, "learning_rate": 1.1453293575812677e-05, "loss": 0.0141, "step": 8567 }, { "epoch": 0.5887850467289719, "grad_norm": 0.7682070732116699, "learning_rate": 1.145004976273366e-05, "loss": 0.0239, "step": 8568 }, { "epoch": 0.5888537658053876, "grad_norm": 1.0274235010147095, "learning_rate": 1.144680612551188e-05, "loss": 0.0258, "step": 8569 }, { "epoch": 0.5889224848818032, "grad_norm": 0.5106025338172913, "learning_rate": 1.1443562664308018e-05, "loss": 0.0185, "step": 8570 }, { "epoch": 0.5889912039582188, "grad_norm": 0.7120288610458374, "learning_rate": 1.144031937928275e-05, "loss": 0.0171, "step": 8571 }, { "epoch": 0.5890599230346344, "grad_norm": 0.46847689151763916, "learning_rate": 1.1437076270596742e-05, "loss": 0.0118, "step": 8572 }, { "epoch": 0.58912864211105, "grad_norm": 0.27596086263656616, "learning_rate": 1.1433833338410651e-05, "loss": 0.0128, "step": 8573 }, { "epoch": 0.5891973611874657, "grad_norm": 0.8276534080505371, "learning_rate": 1.1430590582885121e-05, "loss": 0.0182, "step": 8574 }, { "epoch": 0.5892660802638813, "grad_norm": 0.7065615653991699, "learning_rate": 1.14273480041808e-05, "loss": 0.0162, "step": 8575 }, { "epoch": 0.5893347993402969, "grad_norm": 0.40505409240722656, "learning_rate": 1.1424105602458313e-05, "loss": 0.0133, "step": 8576 }, { "epoch": 0.5894035184167125, "grad_norm": 0.5521584153175354, "learning_rate": 1.1420863377878279e-05, "loss": 0.0181, "step": 8577 }, { "epoch": 0.589472237493128, "grad_norm": 0.3487301468849182, "learning_rate": 1.1417621330601318e-05, "loss": 0.012, "step": 8578 }, { "epoch": 0.5895409565695438, "grad_norm": 0.33686718344688416, "learning_rate": 1.1414379460788032e-05, "loss": 0.0097, "step": 8579 }, { "epoch": 0.5896096756459593, "grad_norm": 0.2270665466785431, "learning_rate": 1.1411137768599015e-05, "loss": 0.0214, "step": 8580 }, { "epoch": 0.5896783947223749, "grad_norm": 0.13242268562316895, "learning_rate": 1.1407896254194854e-05, "loss": 0.0154, "step": 8581 }, { "epoch": 0.5897471137987905, "grad_norm": 0.5361332893371582, "learning_rate": 1.140465491773613e-05, "loss": 0.0161, "step": 8582 }, { "epoch": 0.5898158328752061, "grad_norm": 0.6777111887931824, "learning_rate": 1.14014137593834e-05, "loss": 0.0214, "step": 8583 }, { "epoch": 0.5898845519516218, "grad_norm": 0.3061177730560303, "learning_rate": 1.1398172779297244e-05, "loss": 0.0134, "step": 8584 }, { "epoch": 0.5899532710280374, "grad_norm": 0.4518295228481293, "learning_rate": 1.13949319776382e-05, "loss": 0.0128, "step": 8585 }, { "epoch": 0.590021990104453, "grad_norm": 0.6895666122436523, "learning_rate": 1.1391691354566816e-05, "loss": 0.0168, "step": 8586 }, { "epoch": 0.5900907091808686, "grad_norm": 0.6264123916625977, "learning_rate": 1.1388450910243624e-05, "loss": 0.0139, "step": 8587 }, { "epoch": 0.5901594282572842, "grad_norm": 0.43848514556884766, "learning_rate": 1.1385210644829149e-05, "loss": 0.0171, "step": 8588 }, { "epoch": 0.5902281473336999, "grad_norm": 0.6361192464828491, "learning_rate": 1.1381970558483903e-05, "loss": 0.0128, "step": 8589 }, { "epoch": 0.5902968664101155, "grad_norm": 0.7260279059410095, "learning_rate": 1.1378730651368401e-05, "loss": 0.0115, "step": 8590 }, { "epoch": 0.5903655854865311, "grad_norm": 0.5646765828132629, "learning_rate": 1.1375490923643134e-05, "loss": 0.0222, "step": 8591 }, { "epoch": 0.5904343045629467, "grad_norm": 0.2980772852897644, "learning_rate": 1.1372251375468598e-05, "loss": 0.0151, "step": 8592 }, { "epoch": 0.5905030236393622, "grad_norm": 0.4421333074569702, "learning_rate": 1.1369012007005271e-05, "loss": 0.0118, "step": 8593 }, { "epoch": 0.590571742715778, "grad_norm": 0.23855996131896973, "learning_rate": 1.1365772818413622e-05, "loss": 0.0129, "step": 8594 }, { "epoch": 0.5906404617921935, "grad_norm": 0.29783114790916443, "learning_rate": 1.1362533809854113e-05, "loss": 0.0207, "step": 8595 }, { "epoch": 0.5907091808686091, "grad_norm": 0.5784353017807007, "learning_rate": 1.1359294981487208e-05, "loss": 0.0106, "step": 8596 }, { "epoch": 0.5907778999450247, "grad_norm": 0.4229256212711334, "learning_rate": 1.1356056333473342e-05, "loss": 0.0164, "step": 8597 }, { "epoch": 0.5908466190214403, "grad_norm": 0.4150140583515167, "learning_rate": 1.1352817865972954e-05, "loss": 0.0103, "step": 8598 }, { "epoch": 0.590915338097856, "grad_norm": 0.3911094069480896, "learning_rate": 1.134957957914647e-05, "loss": 0.0115, "step": 8599 }, { "epoch": 0.5909840571742716, "grad_norm": 1.0656936168670654, "learning_rate": 1.134634147315431e-05, "loss": 0.0157, "step": 8600 }, { "epoch": 0.5910527762506872, "grad_norm": 0.9728391766548157, "learning_rate": 1.1343103548156873e-05, "loss": 0.0287, "step": 8601 }, { "epoch": 0.5911214953271028, "grad_norm": 0.7000800967216492, "learning_rate": 1.1339865804314576e-05, "loss": 0.0177, "step": 8602 }, { "epoch": 0.5911902144035184, "grad_norm": 0.5954800844192505, "learning_rate": 1.1336628241787804e-05, "loss": 0.014, "step": 8603 }, { "epoch": 0.5912589334799341, "grad_norm": 0.6394275426864624, "learning_rate": 1.1333390860736934e-05, "loss": 0.0148, "step": 8604 }, { "epoch": 0.5913276525563497, "grad_norm": 0.6872566938400269, "learning_rate": 1.1330153661322344e-05, "loss": 0.0127, "step": 8605 }, { "epoch": 0.5913963716327653, "grad_norm": 1.045274257659912, "learning_rate": 1.1326916643704396e-05, "loss": 0.0166, "step": 8606 }, { "epoch": 0.5914650907091809, "grad_norm": 0.6973972916603088, "learning_rate": 1.132367980804344e-05, "loss": 0.0211, "step": 8607 }, { "epoch": 0.5915338097855964, "grad_norm": 0.44323787093162537, "learning_rate": 1.1320443154499835e-05, "loss": 0.0111, "step": 8608 }, { "epoch": 0.5916025288620121, "grad_norm": 0.43062880635261536, "learning_rate": 1.1317206683233911e-05, "loss": 0.0198, "step": 8609 }, { "epoch": 0.5916712479384277, "grad_norm": 0.8716225624084473, "learning_rate": 1.1313970394405998e-05, "loss": 0.0145, "step": 8610 }, { "epoch": 0.5917399670148433, "grad_norm": 0.2969462275505066, "learning_rate": 1.131073428817641e-05, "loss": 0.0072, "step": 8611 }, { "epoch": 0.5918086860912589, "grad_norm": 0.7109808325767517, "learning_rate": 1.1307498364705463e-05, "loss": 0.0225, "step": 8612 }, { "epoch": 0.5918774051676745, "grad_norm": 0.22538025677204132, "learning_rate": 1.1304262624153451e-05, "loss": 0.0157, "step": 8613 }, { "epoch": 0.5919461242440902, "grad_norm": 0.23771461844444275, "learning_rate": 1.1301027066680674e-05, "loss": 0.0218, "step": 8614 }, { "epoch": 0.5920148433205058, "grad_norm": 0.3740137815475464, "learning_rate": 1.1297791692447413e-05, "loss": 0.0352, "step": 8615 }, { "epoch": 0.5920835623969214, "grad_norm": 0.3989505171775818, "learning_rate": 1.1294556501613938e-05, "loss": 0.0155, "step": 8616 }, { "epoch": 0.592152281473337, "grad_norm": 0.5794773101806641, "learning_rate": 1.1291321494340517e-05, "loss": 0.0133, "step": 8617 }, { "epoch": 0.5922210005497526, "grad_norm": 0.26762256026268005, "learning_rate": 1.1288086670787403e-05, "loss": 0.0096, "step": 8618 }, { "epoch": 0.5922897196261683, "grad_norm": 0.497604638338089, "learning_rate": 1.128485203111484e-05, "loss": 0.0197, "step": 8619 }, { "epoch": 0.5923584387025839, "grad_norm": 0.3262140154838562, "learning_rate": 1.1281617575483075e-05, "loss": 0.0118, "step": 8620 }, { "epoch": 0.5924271577789995, "grad_norm": 0.8107247352600098, "learning_rate": 1.127838330405233e-05, "loss": 0.021, "step": 8621 }, { "epoch": 0.592495876855415, "grad_norm": 0.5822196006774902, "learning_rate": 1.1275149216982825e-05, "loss": 0.0177, "step": 8622 }, { "epoch": 0.5925645959318306, "grad_norm": 0.3343489170074463, "learning_rate": 1.127191531443477e-05, "loss": 0.0125, "step": 8623 }, { "epoch": 0.5926333150082463, "grad_norm": 0.31038737297058105, "learning_rate": 1.1268681596568365e-05, "loss": 0.0159, "step": 8624 }, { "epoch": 0.5927020340846619, "grad_norm": 0.48983234167099, "learning_rate": 1.1265448063543797e-05, "loss": 0.0154, "step": 8625 }, { "epoch": 0.5927707531610775, "grad_norm": 0.6068115830421448, "learning_rate": 1.1262214715521259e-05, "loss": 0.0165, "step": 8626 }, { "epoch": 0.5928394722374931, "grad_norm": 0.39975637197494507, "learning_rate": 1.125898155266092e-05, "loss": 0.0169, "step": 8627 }, { "epoch": 0.5929081913139087, "grad_norm": 0.39763879776000977, "learning_rate": 1.1255748575122944e-05, "loss": 0.0191, "step": 8628 }, { "epoch": 0.5929769103903244, "grad_norm": 0.951220691204071, "learning_rate": 1.1252515783067486e-05, "loss": 0.0164, "step": 8629 }, { "epoch": 0.59304562946674, "grad_norm": 0.7866215705871582, "learning_rate": 1.1249283176654692e-05, "loss": 0.0156, "step": 8630 }, { "epoch": 0.5931143485431556, "grad_norm": 0.26401782035827637, "learning_rate": 1.1246050756044694e-05, "loss": 0.0088, "step": 8631 }, { "epoch": 0.5931830676195712, "grad_norm": 0.4547126591205597, "learning_rate": 1.1242818521397629e-05, "loss": 0.015, "step": 8632 }, { "epoch": 0.5932517866959868, "grad_norm": 0.5039782524108887, "learning_rate": 1.1239586472873607e-05, "loss": 0.0151, "step": 8633 }, { "epoch": 0.5933205057724025, "grad_norm": 0.3111855685710907, "learning_rate": 1.1236354610632742e-05, "loss": 0.0094, "step": 8634 }, { "epoch": 0.5933892248488181, "grad_norm": 0.9224246144294739, "learning_rate": 1.123312293483513e-05, "loss": 0.012, "step": 8635 }, { "epoch": 0.5934579439252337, "grad_norm": 0.6146516799926758, "learning_rate": 1.1229891445640865e-05, "loss": 0.0142, "step": 8636 }, { "epoch": 0.5935266630016492, "grad_norm": 0.3946419656276703, "learning_rate": 1.1226660143210021e-05, "loss": 0.0125, "step": 8637 }, { "epoch": 0.5935953820780648, "grad_norm": 1.3610018491744995, "learning_rate": 1.1223429027702681e-05, "loss": 0.0298, "step": 8638 }, { "epoch": 0.5936641011544805, "grad_norm": 1.2636195421218872, "learning_rate": 1.1220198099278903e-05, "loss": 0.0193, "step": 8639 }, { "epoch": 0.5937328202308961, "grad_norm": 0.38092851638793945, "learning_rate": 1.1216967358098741e-05, "loss": 0.0121, "step": 8640 }, { "epoch": 0.5938015393073117, "grad_norm": 1.4860403537750244, "learning_rate": 1.1213736804322237e-05, "loss": 0.0333, "step": 8641 }, { "epoch": 0.5938702583837273, "grad_norm": 0.8166409134864807, "learning_rate": 1.1210506438109429e-05, "loss": 0.0123, "step": 8642 }, { "epoch": 0.5939389774601429, "grad_norm": 0.5787167549133301, "learning_rate": 1.1207276259620333e-05, "loss": 0.0225, "step": 8643 }, { "epoch": 0.5940076965365585, "grad_norm": 0.5492544174194336, "learning_rate": 1.1204046269014981e-05, "loss": 0.0167, "step": 8644 }, { "epoch": 0.5940764156129742, "grad_norm": 0.5583983063697815, "learning_rate": 1.1200816466453373e-05, "loss": 0.0181, "step": 8645 }, { "epoch": 0.5941451346893898, "grad_norm": 0.5982725620269775, "learning_rate": 1.1197586852095507e-05, "loss": 0.0157, "step": 8646 }, { "epoch": 0.5942138537658054, "grad_norm": 0.3785465955734253, "learning_rate": 1.1194357426101369e-05, "loss": 0.0136, "step": 8647 }, { "epoch": 0.594282572842221, "grad_norm": 0.40465742349624634, "learning_rate": 1.1191128188630943e-05, "loss": 0.0209, "step": 8648 }, { "epoch": 0.5943512919186366, "grad_norm": 0.5016432404518127, "learning_rate": 1.118789913984419e-05, "loss": 0.0176, "step": 8649 }, { "epoch": 0.5944200109950523, "grad_norm": 0.3667669892311096, "learning_rate": 1.118467027990108e-05, "loss": 0.0176, "step": 8650 }, { "epoch": 0.5944887300714679, "grad_norm": 0.41380104422569275, "learning_rate": 1.118144160896156e-05, "loss": 0.0143, "step": 8651 }, { "epoch": 0.5945574491478834, "grad_norm": 0.8163902759552002, "learning_rate": 1.1178213127185573e-05, "loss": 0.0125, "step": 8652 }, { "epoch": 0.594626168224299, "grad_norm": 0.49204909801483154, "learning_rate": 1.117498483473305e-05, "loss": 0.0115, "step": 8653 }, { "epoch": 0.5946948873007146, "grad_norm": 1.1345198154449463, "learning_rate": 1.1171756731763913e-05, "loss": 0.0185, "step": 8654 }, { "epoch": 0.5947636063771303, "grad_norm": 0.725720226764679, "learning_rate": 1.1168528818438074e-05, "loss": 0.0198, "step": 8655 }, { "epoch": 0.5948323254535459, "grad_norm": 0.5344561338424683, "learning_rate": 1.1165301094915443e-05, "loss": 0.0165, "step": 8656 }, { "epoch": 0.5949010445299615, "grad_norm": 0.9068637490272522, "learning_rate": 1.1162073561355913e-05, "loss": 0.0178, "step": 8657 }, { "epoch": 0.5949697636063771, "grad_norm": 1.2940748929977417, "learning_rate": 1.1158846217919371e-05, "loss": 0.0217, "step": 8658 }, { "epoch": 0.5950384826827927, "grad_norm": 1.3054132461547852, "learning_rate": 1.1155619064765688e-05, "loss": 0.015, "step": 8659 }, { "epoch": 0.5951072017592084, "grad_norm": 0.7374447584152222, "learning_rate": 1.1152392102054732e-05, "loss": 0.0196, "step": 8660 }, { "epoch": 0.595175920835624, "grad_norm": 0.4541606307029724, "learning_rate": 1.1149165329946354e-05, "loss": 0.0195, "step": 8661 }, { "epoch": 0.5952446399120396, "grad_norm": 0.5928442478179932, "learning_rate": 1.1145938748600419e-05, "loss": 0.0206, "step": 8662 }, { "epoch": 0.5953133589884552, "grad_norm": 0.36549055576324463, "learning_rate": 1.1142712358176752e-05, "loss": 0.0161, "step": 8663 }, { "epoch": 0.5953820780648708, "grad_norm": 0.43175557255744934, "learning_rate": 1.1139486158835185e-05, "loss": 0.0208, "step": 8664 }, { "epoch": 0.5954507971412865, "grad_norm": 0.39238908886909485, "learning_rate": 1.1136260150735536e-05, "loss": 0.0215, "step": 8665 }, { "epoch": 0.595519516217702, "grad_norm": 0.4746069610118866, "learning_rate": 1.1133034334037617e-05, "loss": 0.0132, "step": 8666 }, { "epoch": 0.5955882352941176, "grad_norm": 0.5678092241287231, "learning_rate": 1.1129808708901223e-05, "loss": 0.0101, "step": 8667 }, { "epoch": 0.5956569543705332, "grad_norm": 0.46989941596984863, "learning_rate": 1.1126583275486151e-05, "loss": 0.0172, "step": 8668 }, { "epoch": 0.5957256734469488, "grad_norm": 0.37821850180625916, "learning_rate": 1.112335803395218e-05, "loss": 0.0081, "step": 8669 }, { "epoch": 0.5957943925233645, "grad_norm": 0.6935087442398071, "learning_rate": 1.1120132984459084e-05, "loss": 0.0178, "step": 8670 }, { "epoch": 0.5958631115997801, "grad_norm": 0.3616655170917511, "learning_rate": 1.1116908127166622e-05, "loss": 0.0152, "step": 8671 }, { "epoch": 0.5959318306761957, "grad_norm": 0.723774254322052, "learning_rate": 1.111368346223455e-05, "loss": 0.017, "step": 8672 }, { "epoch": 0.5960005497526113, "grad_norm": 0.41049259901046753, "learning_rate": 1.11104589898226e-05, "loss": 0.0207, "step": 8673 }, { "epoch": 0.5960692688290269, "grad_norm": 0.48369839787483215, "learning_rate": 1.1107234710090522e-05, "loss": 0.0222, "step": 8674 }, { "epoch": 0.5961379879054426, "grad_norm": 0.7182322144508362, "learning_rate": 1.1104010623198035e-05, "loss": 0.012, "step": 8675 }, { "epoch": 0.5962067069818582, "grad_norm": 0.5809246897697449, "learning_rate": 1.1100786729304852e-05, "loss": 0.0136, "step": 8676 }, { "epoch": 0.5962754260582738, "grad_norm": 0.6718233227729797, "learning_rate": 1.1097563028570676e-05, "loss": 0.0142, "step": 8677 }, { "epoch": 0.5963441451346894, "grad_norm": 0.8521907925605774, "learning_rate": 1.1094339521155205e-05, "loss": 0.0208, "step": 8678 }, { "epoch": 0.596412864211105, "grad_norm": 0.743757426738739, "learning_rate": 1.109111620721812e-05, "loss": 0.0166, "step": 8679 }, { "epoch": 0.5964815832875207, "grad_norm": 0.15793859958648682, "learning_rate": 1.1087893086919106e-05, "loss": 0.0217, "step": 8680 }, { "epoch": 0.5965503023639362, "grad_norm": 0.9642031192779541, "learning_rate": 1.1084670160417827e-05, "loss": 0.015, "step": 8681 }, { "epoch": 0.5966190214403518, "grad_norm": 0.3104581832885742, "learning_rate": 1.1081447427873937e-05, "loss": 0.0159, "step": 8682 }, { "epoch": 0.5966877405167674, "grad_norm": 0.8133047223091125, "learning_rate": 1.1078224889447088e-05, "loss": 0.0163, "step": 8683 }, { "epoch": 0.596756459593183, "grad_norm": 0.23252953588962555, "learning_rate": 1.1075002545296911e-05, "loss": 0.0108, "step": 8684 }, { "epoch": 0.5968251786695987, "grad_norm": 0.4865802526473999, "learning_rate": 1.107178039558304e-05, "loss": 0.0155, "step": 8685 }, { "epoch": 0.5968938977460143, "grad_norm": 0.43597862124443054, "learning_rate": 1.106855844046509e-05, "loss": 0.0152, "step": 8686 }, { "epoch": 0.5969626168224299, "grad_norm": 0.5253945589065552, "learning_rate": 1.1065336680102675e-05, "loss": 0.0167, "step": 8687 }, { "epoch": 0.5970313358988455, "grad_norm": 0.926864743232727, "learning_rate": 1.1062115114655393e-05, "loss": 0.0143, "step": 8688 }, { "epoch": 0.5971000549752611, "grad_norm": 0.6126595735549927, "learning_rate": 1.105889374428283e-05, "loss": 0.0145, "step": 8689 }, { "epoch": 0.5971687740516768, "grad_norm": 0.6494622230529785, "learning_rate": 1.105567256914457e-05, "loss": 0.0095, "step": 8690 }, { "epoch": 0.5972374931280924, "grad_norm": 0.4086195230484009, "learning_rate": 1.1052451589400174e-05, "loss": 0.0215, "step": 8691 }, { "epoch": 0.597306212204508, "grad_norm": 0.9297314882278442, "learning_rate": 1.1049230805209218e-05, "loss": 0.0174, "step": 8692 }, { "epoch": 0.5973749312809236, "grad_norm": 0.23248505592346191, "learning_rate": 1.1046010216731243e-05, "loss": 0.0152, "step": 8693 }, { "epoch": 0.5974436503573392, "grad_norm": 0.6114768385887146, "learning_rate": 1.1042789824125795e-05, "loss": 0.0201, "step": 8694 }, { "epoch": 0.5975123694337549, "grad_norm": 0.4516347646713257, "learning_rate": 1.1039569627552405e-05, "loss": 0.0136, "step": 8695 }, { "epoch": 0.5975810885101704, "grad_norm": 0.579677164554596, "learning_rate": 1.103634962717059e-05, "loss": 0.0117, "step": 8696 }, { "epoch": 0.597649807586586, "grad_norm": 0.3810961842536926, "learning_rate": 1.103312982313986e-05, "loss": 0.0108, "step": 8697 }, { "epoch": 0.5977185266630016, "grad_norm": 0.4591599702835083, "learning_rate": 1.1029910215619731e-05, "loss": 0.0192, "step": 8698 }, { "epoch": 0.5977872457394172, "grad_norm": 1.1815611124038696, "learning_rate": 1.1026690804769686e-05, "loss": 0.0119, "step": 8699 }, { "epoch": 0.5978559648158329, "grad_norm": 0.8994127511978149, "learning_rate": 1.102347159074921e-05, "loss": 0.0153, "step": 8700 }, { "epoch": 0.5979246838922485, "grad_norm": 0.339577317237854, "learning_rate": 1.1020252573717776e-05, "loss": 0.0125, "step": 8701 }, { "epoch": 0.5979934029686641, "grad_norm": 0.6901583671569824, "learning_rate": 1.101703375383485e-05, "loss": 0.0204, "step": 8702 }, { "epoch": 0.5980621220450797, "grad_norm": 0.7050746083259583, "learning_rate": 1.1013815131259877e-05, "loss": 0.0203, "step": 8703 }, { "epoch": 0.5981308411214953, "grad_norm": 0.4718201458454132, "learning_rate": 1.1010596706152312e-05, "loss": 0.0122, "step": 8704 }, { "epoch": 0.598199560197911, "grad_norm": 0.7847493290901184, "learning_rate": 1.1007378478671584e-05, "loss": 0.0167, "step": 8705 }, { "epoch": 0.5982682792743266, "grad_norm": 0.3125937581062317, "learning_rate": 1.1004160448977114e-05, "loss": 0.0126, "step": 8706 }, { "epoch": 0.5983369983507422, "grad_norm": 0.38607874512672424, "learning_rate": 1.1000942617228326e-05, "loss": 0.0208, "step": 8707 }, { "epoch": 0.5984057174271578, "grad_norm": 0.20766986906528473, "learning_rate": 1.0997724983584616e-05, "loss": 0.0143, "step": 8708 }, { "epoch": 0.5984744365035733, "grad_norm": 0.4093456566333771, "learning_rate": 1.0994507548205375e-05, "loss": 0.0201, "step": 8709 }, { "epoch": 0.598543155579989, "grad_norm": 0.3251420557498932, "learning_rate": 1.0991290311250005e-05, "loss": 0.0184, "step": 8710 }, { "epoch": 0.5986118746564046, "grad_norm": 0.7743217945098877, "learning_rate": 1.0988073272877868e-05, "loss": 0.0251, "step": 8711 }, { "epoch": 0.5986805937328202, "grad_norm": 0.5772688984870911, "learning_rate": 1.0984856433248333e-05, "loss": 0.012, "step": 8712 }, { "epoch": 0.5987493128092358, "grad_norm": 1.0989007949829102, "learning_rate": 1.0981639792520758e-05, "loss": 0.0198, "step": 8713 }, { "epoch": 0.5988180318856514, "grad_norm": 0.447503000497818, "learning_rate": 1.0978423350854485e-05, "loss": 0.0113, "step": 8714 }, { "epoch": 0.5988867509620671, "grad_norm": 0.21299071609973907, "learning_rate": 1.0975207108408844e-05, "loss": 0.0104, "step": 8715 }, { "epoch": 0.5989554700384827, "grad_norm": 0.6267138719558716, "learning_rate": 1.0971991065343176e-05, "loss": 0.0118, "step": 8716 }, { "epoch": 0.5990241891148983, "grad_norm": 1.204521656036377, "learning_rate": 1.0968775221816788e-05, "loss": 0.0251, "step": 8717 }, { "epoch": 0.5990929081913139, "grad_norm": 0.4567081332206726, "learning_rate": 1.0965559577988988e-05, "loss": 0.0151, "step": 8718 }, { "epoch": 0.5991616272677295, "grad_norm": 0.3231123089790344, "learning_rate": 1.0962344134019073e-05, "loss": 0.0087, "step": 8719 }, { "epoch": 0.5992303463441452, "grad_norm": 0.4800390303134918, "learning_rate": 1.0959128890066327e-05, "loss": 0.0137, "step": 8720 }, { "epoch": 0.5992990654205608, "grad_norm": 0.5776491761207581, "learning_rate": 1.0955913846290027e-05, "loss": 0.0201, "step": 8721 }, { "epoch": 0.5993677844969764, "grad_norm": 1.1946918964385986, "learning_rate": 1.0952699002849442e-05, "loss": 0.0225, "step": 8722 }, { "epoch": 0.599436503573392, "grad_norm": 0.5211188793182373, "learning_rate": 1.0949484359903829e-05, "loss": 0.0104, "step": 8723 }, { "epoch": 0.5995052226498075, "grad_norm": 0.7391829490661621, "learning_rate": 1.094626991761243e-05, "loss": 0.0108, "step": 8724 }, { "epoch": 0.5995739417262232, "grad_norm": 0.2375679463148117, "learning_rate": 1.094305567613449e-05, "loss": 0.0109, "step": 8725 }, { "epoch": 0.5996426608026388, "grad_norm": 0.5781627893447876, "learning_rate": 1.0939841635629228e-05, "loss": 0.0163, "step": 8726 }, { "epoch": 0.5997113798790544, "grad_norm": 0.7148758769035339, "learning_rate": 1.0936627796255865e-05, "loss": 0.0121, "step": 8727 }, { "epoch": 0.59978009895547, "grad_norm": 0.46975064277648926, "learning_rate": 1.0933414158173602e-05, "loss": 0.0278, "step": 8728 }, { "epoch": 0.5998488180318856, "grad_norm": 0.37531089782714844, "learning_rate": 1.0930200721541644e-05, "loss": 0.0122, "step": 8729 }, { "epoch": 0.5999175371083013, "grad_norm": 0.5859968662261963, "learning_rate": 1.0926987486519177e-05, "loss": 0.0144, "step": 8730 }, { "epoch": 0.5999862561847169, "grad_norm": 0.29827749729156494, "learning_rate": 1.0923774453265378e-05, "loss": 0.0116, "step": 8731 }, { "epoch": 0.6000549752611325, "grad_norm": 1.0569512844085693, "learning_rate": 1.0920561621939408e-05, "loss": 0.0088, "step": 8732 }, { "epoch": 0.6001236943375481, "grad_norm": 0.29422527551651, "learning_rate": 1.0917348992700431e-05, "loss": 0.0144, "step": 8733 }, { "epoch": 0.6001924134139637, "grad_norm": 0.4243060350418091, "learning_rate": 1.0914136565707584e-05, "loss": 0.0158, "step": 8734 }, { "epoch": 0.6002611324903794, "grad_norm": 0.5196083188056946, "learning_rate": 1.091092434112002e-05, "loss": 0.0115, "step": 8735 }, { "epoch": 0.600329851566795, "grad_norm": 0.5789592266082764, "learning_rate": 1.0907712319096856e-05, "loss": 0.0105, "step": 8736 }, { "epoch": 0.6003985706432106, "grad_norm": 0.34987643361091614, "learning_rate": 1.0904500499797209e-05, "loss": 0.0074, "step": 8737 }, { "epoch": 0.6004672897196262, "grad_norm": 0.7905526161193848, "learning_rate": 1.0901288883380187e-05, "loss": 0.0133, "step": 8738 }, { "epoch": 0.6005360087960417, "grad_norm": 0.6621198654174805, "learning_rate": 1.089807747000489e-05, "loss": 0.018, "step": 8739 }, { "epoch": 0.6006047278724574, "grad_norm": 0.524035632610321, "learning_rate": 1.0894866259830397e-05, "loss": 0.0117, "step": 8740 }, { "epoch": 0.600673446948873, "grad_norm": 0.7340638637542725, "learning_rate": 1.0891655253015793e-05, "loss": 0.0151, "step": 8741 }, { "epoch": 0.6007421660252886, "grad_norm": 0.7671294808387756, "learning_rate": 1.088844444972014e-05, "loss": 0.0222, "step": 8742 }, { "epoch": 0.6008108851017042, "grad_norm": 0.5218846797943115, "learning_rate": 1.0885233850102499e-05, "loss": 0.0129, "step": 8743 }, { "epoch": 0.6008796041781198, "grad_norm": 0.3744852542877197, "learning_rate": 1.0882023454321911e-05, "loss": 0.013, "step": 8744 }, { "epoch": 0.6009483232545355, "grad_norm": 0.45905056595802307, "learning_rate": 1.0878813262537415e-05, "loss": 0.0098, "step": 8745 }, { "epoch": 0.6010170423309511, "grad_norm": 0.4178258776664734, "learning_rate": 1.0875603274908032e-05, "loss": 0.0149, "step": 8746 }, { "epoch": 0.6010857614073667, "grad_norm": 0.5908185243606567, "learning_rate": 1.087239349159279e-05, "loss": 0.0142, "step": 8747 }, { "epoch": 0.6011544804837823, "grad_norm": 0.7651347517967224, "learning_rate": 1.0869183912750687e-05, "loss": 0.0199, "step": 8748 }, { "epoch": 0.6012231995601979, "grad_norm": 0.4939075708389282, "learning_rate": 1.0865974538540722e-05, "loss": 0.0081, "step": 8749 }, { "epoch": 0.6012919186366136, "grad_norm": 0.45149484276771545, "learning_rate": 1.0862765369121876e-05, "loss": 0.0161, "step": 8750 }, { "epoch": 0.6013606377130292, "grad_norm": 0.6562961935997009, "learning_rate": 1.0859556404653131e-05, "loss": 0.0218, "step": 8751 }, { "epoch": 0.6014293567894448, "grad_norm": 0.9535791277885437, "learning_rate": 1.0856347645293443e-05, "loss": 0.0159, "step": 8752 }, { "epoch": 0.6014980758658603, "grad_norm": 0.605548083782196, "learning_rate": 1.0853139091201778e-05, "loss": 0.0163, "step": 8753 }, { "epoch": 0.6015667949422759, "grad_norm": 0.3038853704929352, "learning_rate": 1.0849930742537077e-05, "loss": 0.0151, "step": 8754 }, { "epoch": 0.6016355140186916, "grad_norm": 0.506121039390564, "learning_rate": 1.0846722599458277e-05, "loss": 0.0115, "step": 8755 }, { "epoch": 0.6017042330951072, "grad_norm": 0.960088849067688, "learning_rate": 1.08435146621243e-05, "loss": 0.0172, "step": 8756 }, { "epoch": 0.6017729521715228, "grad_norm": 0.5989469885826111, "learning_rate": 1.084030693069406e-05, "loss": 0.0161, "step": 8757 }, { "epoch": 0.6018416712479384, "grad_norm": 0.818200945854187, "learning_rate": 1.0837099405326459e-05, "loss": 0.0192, "step": 8758 }, { "epoch": 0.601910390324354, "grad_norm": 0.39185068011283875, "learning_rate": 1.08338920861804e-05, "loss": 0.0105, "step": 8759 }, { "epoch": 0.6019791094007697, "grad_norm": 0.4848270118236542, "learning_rate": 1.0830684973414761e-05, "loss": 0.0109, "step": 8760 }, { "epoch": 0.6020478284771853, "grad_norm": 0.595197319984436, "learning_rate": 1.0827478067188416e-05, "loss": 0.0169, "step": 8761 }, { "epoch": 0.6021165475536009, "grad_norm": 0.32634907960891724, "learning_rate": 1.082427136766023e-05, "loss": 0.0123, "step": 8762 }, { "epoch": 0.6021852666300165, "grad_norm": 0.5178066492080688, "learning_rate": 1.0821064874989055e-05, "loss": 0.0095, "step": 8763 }, { "epoch": 0.6022539857064321, "grad_norm": 0.3365226089954376, "learning_rate": 1.0817858589333729e-05, "loss": 0.009, "step": 8764 }, { "epoch": 0.6023227047828478, "grad_norm": 0.466705858707428, "learning_rate": 1.0814652510853098e-05, "loss": 0.0209, "step": 8765 }, { "epoch": 0.6023914238592634, "grad_norm": 0.42113280296325684, "learning_rate": 1.0811446639705976e-05, "loss": 0.0104, "step": 8766 }, { "epoch": 0.602460142935679, "grad_norm": 0.718657910823822, "learning_rate": 1.0808240976051175e-05, "loss": 0.012, "step": 8767 }, { "epoch": 0.6025288620120945, "grad_norm": 0.34333646297454834, "learning_rate": 1.0805035520047497e-05, "loss": 0.014, "step": 8768 }, { "epoch": 0.6025975810885101, "grad_norm": 0.4749920964241028, "learning_rate": 1.0801830271853736e-05, "loss": 0.0096, "step": 8769 }, { "epoch": 0.6026663001649258, "grad_norm": 0.7980682849884033, "learning_rate": 1.0798625231628668e-05, "loss": 0.0201, "step": 8770 }, { "epoch": 0.6027350192413414, "grad_norm": 0.5642944574356079, "learning_rate": 1.0795420399531072e-05, "loss": 0.006, "step": 8771 }, { "epoch": 0.602803738317757, "grad_norm": 0.778969943523407, "learning_rate": 1.0792215775719706e-05, "loss": 0.0171, "step": 8772 }, { "epoch": 0.6028724573941726, "grad_norm": 0.7809565663337708, "learning_rate": 1.078901136035332e-05, "loss": 0.0214, "step": 8773 }, { "epoch": 0.6029411764705882, "grad_norm": 0.347134530544281, "learning_rate": 1.0785807153590653e-05, "loss": 0.009, "step": 8774 }, { "epoch": 0.6030098955470038, "grad_norm": 0.6294769048690796, "learning_rate": 1.0782603155590437e-05, "loss": 0.018, "step": 8775 }, { "epoch": 0.6030786146234195, "grad_norm": 0.9573007225990295, "learning_rate": 1.0779399366511386e-05, "loss": 0.0147, "step": 8776 }, { "epoch": 0.6031473336998351, "grad_norm": 0.17649337649345398, "learning_rate": 1.0776195786512216e-05, "loss": 0.0092, "step": 8777 }, { "epoch": 0.6032160527762507, "grad_norm": 0.40042245388031006, "learning_rate": 1.0772992415751624e-05, "loss": 0.0115, "step": 8778 }, { "epoch": 0.6032847718526663, "grad_norm": 0.24746042490005493, "learning_rate": 1.0769789254388299e-05, "loss": 0.0117, "step": 8779 }, { "epoch": 0.6033534909290819, "grad_norm": 0.6286851167678833, "learning_rate": 1.076658630258091e-05, "loss": 0.0139, "step": 8780 }, { "epoch": 0.6034222100054976, "grad_norm": 0.5181041955947876, "learning_rate": 1.0763383560488138e-05, "loss": 0.0202, "step": 8781 }, { "epoch": 0.6034909290819132, "grad_norm": 0.2367807775735855, "learning_rate": 1.0760181028268629e-05, "loss": 0.0106, "step": 8782 }, { "epoch": 0.6035596481583287, "grad_norm": 0.34100621938705444, "learning_rate": 1.075697870608104e-05, "loss": 0.02, "step": 8783 }, { "epoch": 0.6036283672347443, "grad_norm": 0.5078058838844299, "learning_rate": 1.0753776594084003e-05, "loss": 0.0246, "step": 8784 }, { "epoch": 0.6036970863111599, "grad_norm": 0.5561378598213196, "learning_rate": 1.0750574692436143e-05, "loss": 0.0144, "step": 8785 }, { "epoch": 0.6037658053875756, "grad_norm": 0.594185471534729, "learning_rate": 1.0747373001296077e-05, "loss": 0.0214, "step": 8786 }, { "epoch": 0.6038345244639912, "grad_norm": 0.5401326417922974, "learning_rate": 1.0744171520822408e-05, "loss": 0.0114, "step": 8787 }, { "epoch": 0.6039032435404068, "grad_norm": 0.730383574962616, "learning_rate": 1.0740970251173728e-05, "loss": 0.0128, "step": 8788 }, { "epoch": 0.6039719626168224, "grad_norm": 0.3612273633480072, "learning_rate": 1.073776919250863e-05, "loss": 0.0168, "step": 8789 }, { "epoch": 0.604040681693238, "grad_norm": 0.5266174674034119, "learning_rate": 1.0734568344985682e-05, "loss": 0.0118, "step": 8790 }, { "epoch": 0.6041094007696537, "grad_norm": 0.4732780456542969, "learning_rate": 1.0731367708763452e-05, "loss": 0.0104, "step": 8791 }, { "epoch": 0.6041781198460693, "grad_norm": 0.9615870118141174, "learning_rate": 1.0728167284000489e-05, "loss": 0.0215, "step": 8792 }, { "epoch": 0.6042468389224849, "grad_norm": 0.5959092974662781, "learning_rate": 1.0724967070855336e-05, "loss": 0.0123, "step": 8793 }, { "epoch": 0.6043155579989005, "grad_norm": 1.2765401601791382, "learning_rate": 1.0721767069486522e-05, "loss": 0.019, "step": 8794 }, { "epoch": 0.6043842770753161, "grad_norm": 0.4796428680419922, "learning_rate": 1.0718567280052575e-05, "loss": 0.0132, "step": 8795 }, { "epoch": 0.6044529961517318, "grad_norm": 0.38221997022628784, "learning_rate": 1.0715367702712001e-05, "loss": 0.0201, "step": 8796 }, { "epoch": 0.6045217152281473, "grad_norm": 2.2998673915863037, "learning_rate": 1.07121683376233e-05, "loss": 0.0112, "step": 8797 }, { "epoch": 0.6045904343045629, "grad_norm": 0.849675714969635, "learning_rate": 1.0708969184944969e-05, "loss": 0.0155, "step": 8798 }, { "epoch": 0.6046591533809785, "grad_norm": 0.9445217251777649, "learning_rate": 1.0705770244835481e-05, "loss": 0.0182, "step": 8799 }, { "epoch": 0.6047278724573941, "grad_norm": 1.1906858682632446, "learning_rate": 1.0702571517453304e-05, "loss": 0.0159, "step": 8800 }, { "epoch": 0.6047965915338098, "grad_norm": 0.5345550179481506, "learning_rate": 1.0699373002956903e-05, "loss": 0.0167, "step": 8801 }, { "epoch": 0.6048653106102254, "grad_norm": 0.8616076111793518, "learning_rate": 1.0696174701504721e-05, "loss": 0.0152, "step": 8802 }, { "epoch": 0.604934029686641, "grad_norm": 1.367872714996338, "learning_rate": 1.0692976613255195e-05, "loss": 0.0208, "step": 8803 }, { "epoch": 0.6050027487630566, "grad_norm": 0.627142071723938, "learning_rate": 1.0689778738366754e-05, "loss": 0.0185, "step": 8804 }, { "epoch": 0.6050714678394722, "grad_norm": 0.2295726239681244, "learning_rate": 1.068658107699781e-05, "loss": 0.0152, "step": 8805 }, { "epoch": 0.6051401869158879, "grad_norm": 0.434969425201416, "learning_rate": 1.0683383629306767e-05, "loss": 0.0154, "step": 8806 }, { "epoch": 0.6052089059923035, "grad_norm": 0.26974430680274963, "learning_rate": 1.068018639545203e-05, "loss": 0.0105, "step": 8807 }, { "epoch": 0.6052776250687191, "grad_norm": 0.31240367889404297, "learning_rate": 1.0676989375591981e-05, "loss": 0.0127, "step": 8808 }, { "epoch": 0.6053463441451347, "grad_norm": 0.500381588935852, "learning_rate": 1.0673792569884987e-05, "loss": 0.0109, "step": 8809 }, { "epoch": 0.6054150632215503, "grad_norm": 0.3153393268585205, "learning_rate": 1.0670595978489417e-05, "loss": 0.0105, "step": 8810 }, { "epoch": 0.605483782297966, "grad_norm": 1.5619562864303589, "learning_rate": 1.066739960156362e-05, "loss": 0.0146, "step": 8811 }, { "epoch": 0.6055525013743815, "grad_norm": 0.6755164265632629, "learning_rate": 1.0664203439265937e-05, "loss": 0.0156, "step": 8812 }, { "epoch": 0.6056212204507971, "grad_norm": 0.46912360191345215, "learning_rate": 1.0661007491754702e-05, "loss": 0.0125, "step": 8813 }, { "epoch": 0.6056899395272127, "grad_norm": 0.35238558053970337, "learning_rate": 1.0657811759188238e-05, "loss": 0.0222, "step": 8814 }, { "epoch": 0.6057586586036283, "grad_norm": 0.5917509198188782, "learning_rate": 1.065461624172485e-05, "loss": 0.0194, "step": 8815 }, { "epoch": 0.605827377680044, "grad_norm": 0.5654718279838562, "learning_rate": 1.065142093952284e-05, "loss": 0.0141, "step": 8816 }, { "epoch": 0.6058960967564596, "grad_norm": 0.6514397859573364, "learning_rate": 1.06482258527405e-05, "loss": 0.0188, "step": 8817 }, { "epoch": 0.6059648158328752, "grad_norm": 0.5632936954498291, "learning_rate": 1.06450309815361e-05, "loss": 0.021, "step": 8818 }, { "epoch": 0.6060335349092908, "grad_norm": 0.8451827168464661, "learning_rate": 1.0641836326067914e-05, "loss": 0.0129, "step": 8819 }, { "epoch": 0.6061022539857064, "grad_norm": 0.3212079405784607, "learning_rate": 1.0638641886494195e-05, "loss": 0.0102, "step": 8820 }, { "epoch": 0.6061709730621221, "grad_norm": 0.19675347208976746, "learning_rate": 1.0635447662973193e-05, "loss": 0.0107, "step": 8821 }, { "epoch": 0.6062396921385377, "grad_norm": 0.5045435428619385, "learning_rate": 1.0632253655663139e-05, "loss": 0.0206, "step": 8822 }, { "epoch": 0.6063084112149533, "grad_norm": 0.5529835820198059, "learning_rate": 1.062905986472226e-05, "loss": 0.0131, "step": 8823 }, { "epoch": 0.6063771302913689, "grad_norm": 0.4536241590976715, "learning_rate": 1.0625866290308762e-05, "loss": 0.0221, "step": 8824 }, { "epoch": 0.6064458493677845, "grad_norm": 0.5764299631118774, "learning_rate": 1.0622672932580862e-05, "loss": 0.0174, "step": 8825 }, { "epoch": 0.6065145684442002, "grad_norm": 0.5003480315208435, "learning_rate": 1.0619479791696745e-05, "loss": 0.0142, "step": 8826 }, { "epoch": 0.6065832875206157, "grad_norm": 0.8351168632507324, "learning_rate": 1.0616286867814593e-05, "loss": 0.0153, "step": 8827 }, { "epoch": 0.6066520065970313, "grad_norm": 0.56130450963974, "learning_rate": 1.0613094161092577e-05, "loss": 0.014, "step": 8828 }, { "epoch": 0.6067207256734469, "grad_norm": 0.9469110369682312, "learning_rate": 1.0609901671688858e-05, "loss": 0.0255, "step": 8829 }, { "epoch": 0.6067894447498625, "grad_norm": 0.7320516705513, "learning_rate": 1.0606709399761578e-05, "loss": 0.0181, "step": 8830 }, { "epoch": 0.6068581638262782, "grad_norm": 0.6262848377227783, "learning_rate": 1.0603517345468889e-05, "loss": 0.0192, "step": 8831 }, { "epoch": 0.6069268829026938, "grad_norm": 0.5705540180206299, "learning_rate": 1.060032550896891e-05, "loss": 0.0107, "step": 8832 }, { "epoch": 0.6069956019791094, "grad_norm": 0.5071391463279724, "learning_rate": 1.0597133890419763e-05, "loss": 0.0302, "step": 8833 }, { "epoch": 0.607064321055525, "grad_norm": 0.493894100189209, "learning_rate": 1.0593942489979548e-05, "loss": 0.0125, "step": 8834 }, { "epoch": 0.6071330401319406, "grad_norm": 0.6699485778808594, "learning_rate": 1.0590751307806365e-05, "loss": 0.0208, "step": 8835 }, { "epoch": 0.6072017592083563, "grad_norm": 1.0408926010131836, "learning_rate": 1.0587560344058294e-05, "loss": 0.0172, "step": 8836 }, { "epoch": 0.6072704782847719, "grad_norm": 0.4131181538105011, "learning_rate": 1.0584369598893415e-05, "loss": 0.0128, "step": 8837 }, { "epoch": 0.6073391973611875, "grad_norm": 0.645583987236023, "learning_rate": 1.0581179072469788e-05, "loss": 0.0188, "step": 8838 }, { "epoch": 0.6074079164376031, "grad_norm": 0.5545335412025452, "learning_rate": 1.0577988764945465e-05, "loss": 0.0078, "step": 8839 }, { "epoch": 0.6074766355140186, "grad_norm": 0.23490287363529205, "learning_rate": 1.0574798676478488e-05, "loss": 0.009, "step": 8840 }, { "epoch": 0.6075453545904343, "grad_norm": 0.5382471680641174, "learning_rate": 1.0571608807226887e-05, "loss": 0.0176, "step": 8841 }, { "epoch": 0.6076140736668499, "grad_norm": 0.584071159362793, "learning_rate": 1.0568419157348673e-05, "loss": 0.01, "step": 8842 }, { "epoch": 0.6076827927432655, "grad_norm": 0.6718289256095886, "learning_rate": 1.0565229727001871e-05, "loss": 0.0123, "step": 8843 }, { "epoch": 0.6077515118196811, "grad_norm": 0.24745769798755646, "learning_rate": 1.0562040516344469e-05, "loss": 0.0081, "step": 8844 }, { "epoch": 0.6078202308960967, "grad_norm": 1.1921262741088867, "learning_rate": 1.0558851525534458e-05, "loss": 0.0257, "step": 8845 }, { "epoch": 0.6078889499725124, "grad_norm": 0.43009501695632935, "learning_rate": 1.055566275472981e-05, "loss": 0.0137, "step": 8846 }, { "epoch": 0.607957669048928, "grad_norm": 0.31688839197158813, "learning_rate": 1.0552474204088494e-05, "loss": 0.0099, "step": 8847 }, { "epoch": 0.6080263881253436, "grad_norm": 0.4906037151813507, "learning_rate": 1.0549285873768453e-05, "loss": 0.0098, "step": 8848 }, { "epoch": 0.6080951072017592, "grad_norm": 0.20773422718048096, "learning_rate": 1.0546097763927649e-05, "loss": 0.0124, "step": 8849 }, { "epoch": 0.6081638262781748, "grad_norm": 0.3562263250350952, "learning_rate": 1.0542909874724003e-05, "loss": 0.015, "step": 8850 }, { "epoch": 0.6082325453545905, "grad_norm": 1.12571382522583, "learning_rate": 1.0539722206315439e-05, "loss": 0.0115, "step": 8851 }, { "epoch": 0.6083012644310061, "grad_norm": 0.5725511312484741, "learning_rate": 1.0536534758859866e-05, "loss": 0.0157, "step": 8852 }, { "epoch": 0.6083699835074217, "grad_norm": 0.33467769622802734, "learning_rate": 1.0533347532515185e-05, "loss": 0.0184, "step": 8853 }, { "epoch": 0.6084387025838373, "grad_norm": 0.5861769914627075, "learning_rate": 1.0530160527439282e-05, "loss": 0.0183, "step": 8854 }, { "epoch": 0.6085074216602528, "grad_norm": 0.40990370512008667, "learning_rate": 1.0526973743790038e-05, "loss": 0.0128, "step": 8855 }, { "epoch": 0.6085761407366685, "grad_norm": 0.1813099980354309, "learning_rate": 1.052378718172532e-05, "loss": 0.0133, "step": 8856 }, { "epoch": 0.6086448598130841, "grad_norm": 0.30976787209510803, "learning_rate": 1.0520600841402984e-05, "loss": 0.0153, "step": 8857 }, { "epoch": 0.6087135788894997, "grad_norm": 0.47575056552886963, "learning_rate": 1.0517414722980871e-05, "loss": 0.0155, "step": 8858 }, { "epoch": 0.6087822979659153, "grad_norm": 0.754787027835846, "learning_rate": 1.0514228826616816e-05, "loss": 0.0223, "step": 8859 }, { "epoch": 0.6088510170423309, "grad_norm": 0.7745071649551392, "learning_rate": 1.0511043152468638e-05, "loss": 0.0217, "step": 8860 }, { "epoch": 0.6089197361187466, "grad_norm": 0.6877559423446655, "learning_rate": 1.050785770069416e-05, "loss": 0.018, "step": 8861 }, { "epoch": 0.6089884551951622, "grad_norm": 0.3018608093261719, "learning_rate": 1.0504672471451174e-05, "loss": 0.0082, "step": 8862 }, { "epoch": 0.6090571742715778, "grad_norm": 0.5474700331687927, "learning_rate": 1.0501487464897474e-05, "loss": 0.013, "step": 8863 }, { "epoch": 0.6091258933479934, "grad_norm": 0.6683222651481628, "learning_rate": 1.0498302681190835e-05, "loss": 0.0158, "step": 8864 }, { "epoch": 0.609194612424409, "grad_norm": 1.021156668663025, "learning_rate": 1.0495118120489027e-05, "loss": 0.0213, "step": 8865 }, { "epoch": 0.6092633315008247, "grad_norm": 0.2329753190279007, "learning_rate": 1.0491933782949798e-05, "loss": 0.0175, "step": 8866 }, { "epoch": 0.6093320505772403, "grad_norm": 0.3385348320007324, "learning_rate": 1.0488749668730907e-05, "loss": 0.0105, "step": 8867 }, { "epoch": 0.6094007696536559, "grad_norm": 0.3122129738330841, "learning_rate": 1.0485565777990087e-05, "loss": 0.0094, "step": 8868 }, { "epoch": 0.6094694887300715, "grad_norm": 0.46328842639923096, "learning_rate": 1.0482382110885052e-05, "loss": 0.0192, "step": 8869 }, { "epoch": 0.609538207806487, "grad_norm": 0.32042446732521057, "learning_rate": 1.047919866757352e-05, "loss": 0.0168, "step": 8870 }, { "epoch": 0.6096069268829027, "grad_norm": 0.192041277885437, "learning_rate": 1.0476015448213193e-05, "loss": 0.0135, "step": 8871 }, { "epoch": 0.6096756459593183, "grad_norm": 0.6929438710212708, "learning_rate": 1.047283245296176e-05, "loss": 0.0151, "step": 8872 }, { "epoch": 0.6097443650357339, "grad_norm": 0.5500849485397339, "learning_rate": 1.0469649681976898e-05, "loss": 0.0152, "step": 8873 }, { "epoch": 0.6098130841121495, "grad_norm": 0.47921159863471985, "learning_rate": 1.0466467135416279e-05, "loss": 0.0125, "step": 8874 }, { "epoch": 0.6098818031885651, "grad_norm": 0.5396959185600281, "learning_rate": 1.0463284813437557e-05, "loss": 0.0099, "step": 8875 }, { "epoch": 0.6099505222649808, "grad_norm": 0.50193852186203, "learning_rate": 1.0460102716198377e-05, "loss": 0.0102, "step": 8876 }, { "epoch": 0.6100192413413964, "grad_norm": 0.6629091501235962, "learning_rate": 1.0456920843856377e-05, "loss": 0.014, "step": 8877 }, { "epoch": 0.610087960417812, "grad_norm": 0.6364204287528992, "learning_rate": 1.045373919656917e-05, "loss": 0.0204, "step": 8878 }, { "epoch": 0.6101566794942276, "grad_norm": 0.8082881569862366, "learning_rate": 1.0450557774494384e-05, "loss": 0.0115, "step": 8879 }, { "epoch": 0.6102253985706432, "grad_norm": 0.5552789568901062, "learning_rate": 1.0447376577789614e-05, "loss": 0.0147, "step": 8880 }, { "epoch": 0.6102941176470589, "grad_norm": 0.44562995433807373, "learning_rate": 1.0444195606612446e-05, "loss": 0.0132, "step": 8881 }, { "epoch": 0.6103628367234745, "grad_norm": 0.7858907580375671, "learning_rate": 1.0441014861120461e-05, "loss": 0.0144, "step": 8882 }, { "epoch": 0.6104315557998901, "grad_norm": 0.5319052338600159, "learning_rate": 1.043783434147123e-05, "loss": 0.0164, "step": 8883 }, { "epoch": 0.6105002748763056, "grad_norm": 0.9214735627174377, "learning_rate": 1.0434654047822297e-05, "loss": 0.0202, "step": 8884 }, { "epoch": 0.6105689939527212, "grad_norm": 0.37430524826049805, "learning_rate": 1.0431473980331225e-05, "loss": 0.0131, "step": 8885 }, { "epoch": 0.6106377130291369, "grad_norm": 1.014591932296753, "learning_rate": 1.0428294139155536e-05, "loss": 0.0156, "step": 8886 }, { "epoch": 0.6107064321055525, "grad_norm": 0.5576303601264954, "learning_rate": 1.042511452445276e-05, "loss": 0.0228, "step": 8887 }, { "epoch": 0.6107751511819681, "grad_norm": 0.37989985942840576, "learning_rate": 1.0421935136380402e-05, "loss": 0.0154, "step": 8888 }, { "epoch": 0.6108438702583837, "grad_norm": 0.7807921171188354, "learning_rate": 1.0418755975095968e-05, "loss": 0.0191, "step": 8889 }, { "epoch": 0.6109125893347993, "grad_norm": 1.425379991531372, "learning_rate": 1.041557704075694e-05, "loss": 0.0204, "step": 8890 }, { "epoch": 0.610981308411215, "grad_norm": 0.49493035674095154, "learning_rate": 1.0412398333520804e-05, "loss": 0.0158, "step": 8891 }, { "epoch": 0.6110500274876306, "grad_norm": 0.3171307444572449, "learning_rate": 1.0409219853545022e-05, "loss": 0.0132, "step": 8892 }, { "epoch": 0.6111187465640462, "grad_norm": 0.5070885419845581, "learning_rate": 1.0406041600987051e-05, "loss": 0.0166, "step": 8893 }, { "epoch": 0.6111874656404618, "grad_norm": 0.17874236404895782, "learning_rate": 1.0402863576004333e-05, "loss": 0.009, "step": 8894 }, { "epoch": 0.6112561847168774, "grad_norm": 0.2925414443016052, "learning_rate": 1.0399685778754304e-05, "loss": 0.0128, "step": 8895 }, { "epoch": 0.6113249037932931, "grad_norm": 0.4712205231189728, "learning_rate": 1.0396508209394376e-05, "loss": 0.0165, "step": 8896 }, { "epoch": 0.6113936228697087, "grad_norm": 1.9592540264129639, "learning_rate": 1.0393330868081973e-05, "loss": 0.0136, "step": 8897 }, { "epoch": 0.6114623419461243, "grad_norm": 0.8252732157707214, "learning_rate": 1.0390153754974487e-05, "loss": 0.0143, "step": 8898 }, { "epoch": 0.6115310610225398, "grad_norm": 1.0607717037200928, "learning_rate": 1.038697687022931e-05, "loss": 0.0192, "step": 8899 }, { "epoch": 0.6115997800989554, "grad_norm": 0.27704471349716187, "learning_rate": 1.0383800214003813e-05, "loss": 0.0155, "step": 8900 }, { "epoch": 0.611668499175371, "grad_norm": 0.6366090774536133, "learning_rate": 1.0380623786455361e-05, "loss": 0.0218, "step": 8901 }, { "epoch": 0.6117372182517867, "grad_norm": 0.42214974761009216, "learning_rate": 1.0377447587741307e-05, "loss": 0.0238, "step": 8902 }, { "epoch": 0.6118059373282023, "grad_norm": 0.3615800738334656, "learning_rate": 1.0374271618019e-05, "loss": 0.0157, "step": 8903 }, { "epoch": 0.6118746564046179, "grad_norm": 0.4516857862472534, "learning_rate": 1.0371095877445766e-05, "loss": 0.0207, "step": 8904 }, { "epoch": 0.6119433754810335, "grad_norm": 0.4666425883769989, "learning_rate": 1.0367920366178927e-05, "loss": 0.0111, "step": 8905 }, { "epoch": 0.6120120945574491, "grad_norm": 0.3355186879634857, "learning_rate": 1.036474508437579e-05, "loss": 0.0107, "step": 8906 }, { "epoch": 0.6120808136338648, "grad_norm": 0.39366260170936584, "learning_rate": 1.0361570032193653e-05, "loss": 0.0195, "step": 8907 }, { "epoch": 0.6121495327102804, "grad_norm": 0.2822781801223755, "learning_rate": 1.0358395209789795e-05, "loss": 0.0118, "step": 8908 }, { "epoch": 0.612218251786696, "grad_norm": 0.9423218369483948, "learning_rate": 1.0355220617321498e-05, "loss": 0.01, "step": 8909 }, { "epoch": 0.6122869708631116, "grad_norm": 0.5240446925163269, "learning_rate": 1.0352046254946025e-05, "loss": 0.0122, "step": 8910 }, { "epoch": 0.6123556899395272, "grad_norm": 0.6805936694145203, "learning_rate": 1.0348872122820625e-05, "loss": 0.0159, "step": 8911 }, { "epoch": 0.6124244090159429, "grad_norm": 0.5806146264076233, "learning_rate": 1.0345698221102536e-05, "loss": 0.0101, "step": 8912 }, { "epoch": 0.6124931280923585, "grad_norm": 0.7063677906990051, "learning_rate": 1.0342524549948987e-05, "loss": 0.0105, "step": 8913 }, { "epoch": 0.612561847168774, "grad_norm": 0.2599778473377228, "learning_rate": 1.0339351109517192e-05, "loss": 0.0141, "step": 8914 }, { "epoch": 0.6126305662451896, "grad_norm": 0.44757550954818726, "learning_rate": 1.0336177899964368e-05, "loss": 0.0202, "step": 8915 }, { "epoch": 0.6126992853216052, "grad_norm": 0.32854920625686646, "learning_rate": 1.03330049214477e-05, "loss": 0.0149, "step": 8916 }, { "epoch": 0.6127680043980209, "grad_norm": 0.649581789970398, "learning_rate": 1.0329832174124375e-05, "loss": 0.0122, "step": 8917 }, { "epoch": 0.6128367234744365, "grad_norm": 0.5024107694625854, "learning_rate": 1.0326659658151563e-05, "loss": 0.0179, "step": 8918 }, { "epoch": 0.6129054425508521, "grad_norm": 0.14372873306274414, "learning_rate": 1.0323487373686425e-05, "loss": 0.0168, "step": 8919 }, { "epoch": 0.6129741616272677, "grad_norm": 0.36403802037239075, "learning_rate": 1.0320315320886097e-05, "loss": 0.0101, "step": 8920 }, { "epoch": 0.6130428807036833, "grad_norm": 0.6305118203163147, "learning_rate": 1.0317143499907737e-05, "loss": 0.0202, "step": 8921 }, { "epoch": 0.613111599780099, "grad_norm": 0.43944811820983887, "learning_rate": 1.0313971910908461e-05, "loss": 0.0148, "step": 8922 }, { "epoch": 0.6131803188565146, "grad_norm": 0.4648667573928833, "learning_rate": 1.031080055404538e-05, "loss": 0.0127, "step": 8923 }, { "epoch": 0.6132490379329302, "grad_norm": 0.513197124004364, "learning_rate": 1.03076294294756e-05, "loss": 0.0088, "step": 8924 }, { "epoch": 0.6133177570093458, "grad_norm": 0.5178117752075195, "learning_rate": 1.030445853735621e-05, "loss": 0.0115, "step": 8925 }, { "epoch": 0.6133864760857614, "grad_norm": 0.47156012058258057, "learning_rate": 1.030128787784429e-05, "loss": 0.013, "step": 8926 }, { "epoch": 0.6134551951621771, "grad_norm": 0.9236434698104858, "learning_rate": 1.029811745109691e-05, "loss": 0.0106, "step": 8927 }, { "epoch": 0.6135239142385926, "grad_norm": 0.845412015914917, "learning_rate": 1.0294947257271125e-05, "loss": 0.0147, "step": 8928 }, { "epoch": 0.6135926333150082, "grad_norm": 0.9553526043891907, "learning_rate": 1.0291777296523979e-05, "loss": 0.0199, "step": 8929 }, { "epoch": 0.6136613523914238, "grad_norm": 0.9141058921813965, "learning_rate": 1.0288607569012505e-05, "loss": 0.0252, "step": 8930 }, { "epoch": 0.6137300714678394, "grad_norm": 0.7908550500869751, "learning_rate": 1.0285438074893727e-05, "loss": 0.0132, "step": 8931 }, { "epoch": 0.6137987905442551, "grad_norm": 0.7094090580940247, "learning_rate": 1.0282268814324647e-05, "loss": 0.0218, "step": 8932 }, { "epoch": 0.6138675096206707, "grad_norm": 0.7141279578208923, "learning_rate": 1.0279099787462279e-05, "loss": 0.0173, "step": 8933 }, { "epoch": 0.6139362286970863, "grad_norm": 0.9101760983467102, "learning_rate": 1.0275930994463599e-05, "loss": 0.0179, "step": 8934 }, { "epoch": 0.6140049477735019, "grad_norm": 0.7138628363609314, "learning_rate": 1.0272762435485588e-05, "loss": 0.0233, "step": 8935 }, { "epoch": 0.6140736668499175, "grad_norm": 0.6191034913063049, "learning_rate": 1.0269594110685206e-05, "loss": 0.0289, "step": 8936 }, { "epoch": 0.6141423859263332, "grad_norm": 0.756602942943573, "learning_rate": 1.0266426020219407e-05, "loss": 0.0234, "step": 8937 }, { "epoch": 0.6142111050027488, "grad_norm": 0.7062939405441284, "learning_rate": 1.0263258164245124e-05, "loss": 0.0238, "step": 8938 }, { "epoch": 0.6142798240791644, "grad_norm": 0.5109886527061462, "learning_rate": 1.02600905429193e-05, "loss": 0.012, "step": 8939 }, { "epoch": 0.61434854315558, "grad_norm": 0.925849199295044, "learning_rate": 1.0256923156398846e-05, "loss": 0.013, "step": 8940 }, { "epoch": 0.6144172622319956, "grad_norm": 0.569409191608429, "learning_rate": 1.0253756004840666e-05, "loss": 0.0114, "step": 8941 }, { "epoch": 0.6144859813084113, "grad_norm": 0.18331676721572876, "learning_rate": 1.0250589088401655e-05, "loss": 0.0069, "step": 8942 }, { "epoch": 0.6145547003848268, "grad_norm": 0.4406130015850067, "learning_rate": 1.02474224072387e-05, "loss": 0.0087, "step": 8943 }, { "epoch": 0.6146234194612424, "grad_norm": 0.47976163029670715, "learning_rate": 1.0244255961508662e-05, "loss": 0.0136, "step": 8944 }, { "epoch": 0.614692138537658, "grad_norm": 0.6303184032440186, "learning_rate": 1.0241089751368412e-05, "loss": 0.0237, "step": 8945 }, { "epoch": 0.6147608576140736, "grad_norm": 0.5885294675827026, "learning_rate": 1.023792377697479e-05, "loss": 0.0128, "step": 8946 }, { "epoch": 0.6148295766904893, "grad_norm": 0.551857054233551, "learning_rate": 1.0234758038484631e-05, "loss": 0.014, "step": 8947 }, { "epoch": 0.6148982957669049, "grad_norm": 0.5529122352600098, "learning_rate": 1.0231592536054767e-05, "loss": 0.0144, "step": 8948 }, { "epoch": 0.6149670148433205, "grad_norm": 0.5755861401557922, "learning_rate": 1.0228427269842002e-05, "loss": 0.0162, "step": 8949 }, { "epoch": 0.6150357339197361, "grad_norm": 0.7132168412208557, "learning_rate": 1.0225262240003136e-05, "loss": 0.0125, "step": 8950 }, { "epoch": 0.6151044529961517, "grad_norm": 0.5317896008491516, "learning_rate": 1.0222097446694966e-05, "loss": 0.0119, "step": 8951 }, { "epoch": 0.6151731720725674, "grad_norm": 0.39690735936164856, "learning_rate": 1.0218932890074268e-05, "loss": 0.0109, "step": 8952 }, { "epoch": 0.615241891148983, "grad_norm": 0.9994815587997437, "learning_rate": 1.0215768570297802e-05, "loss": 0.0146, "step": 8953 }, { "epoch": 0.6153106102253986, "grad_norm": 0.6120741367340088, "learning_rate": 1.0212604487522327e-05, "loss": 0.0184, "step": 8954 }, { "epoch": 0.6153793293018142, "grad_norm": 0.41261476278305054, "learning_rate": 1.0209440641904583e-05, "loss": 0.0142, "step": 8955 }, { "epoch": 0.6154480483782298, "grad_norm": 0.34861594438552856, "learning_rate": 1.0206277033601302e-05, "loss": 0.0133, "step": 8956 }, { "epoch": 0.6155167674546455, "grad_norm": 0.3876745104789734, "learning_rate": 1.0203113662769193e-05, "loss": 0.0065, "step": 8957 }, { "epoch": 0.615585486531061, "grad_norm": 0.5466233491897583, "learning_rate": 1.0199950529564977e-05, "loss": 0.0137, "step": 8958 }, { "epoch": 0.6156542056074766, "grad_norm": 0.7670456171035767, "learning_rate": 1.0196787634145344e-05, "loss": 0.0142, "step": 8959 }, { "epoch": 0.6157229246838922, "grad_norm": 0.6715137362480164, "learning_rate": 1.0193624976666978e-05, "loss": 0.0362, "step": 8960 }, { "epoch": 0.6157916437603078, "grad_norm": 0.4005168676376343, "learning_rate": 1.0190462557286548e-05, "loss": 0.0123, "step": 8961 }, { "epoch": 0.6158603628367235, "grad_norm": 1.3249131441116333, "learning_rate": 1.0187300376160715e-05, "loss": 0.0237, "step": 8962 }, { "epoch": 0.6159290819131391, "grad_norm": 0.6577938199043274, "learning_rate": 1.0184138433446124e-05, "loss": 0.019, "step": 8963 }, { "epoch": 0.6159978009895547, "grad_norm": 0.5496118068695068, "learning_rate": 1.0180976729299419e-05, "loss": 0.0188, "step": 8964 }, { "epoch": 0.6160665200659703, "grad_norm": 0.7132799029350281, "learning_rate": 1.0177815263877216e-05, "loss": 0.0214, "step": 8965 }, { "epoch": 0.6161352391423859, "grad_norm": 0.2773102819919586, "learning_rate": 1.0174654037336132e-05, "loss": 0.0097, "step": 8966 }, { "epoch": 0.6162039582188016, "grad_norm": 0.420615553855896, "learning_rate": 1.0171493049832768e-05, "loss": 0.0127, "step": 8967 }, { "epoch": 0.6162726772952172, "grad_norm": 0.3316452205181122, "learning_rate": 1.016833230152371e-05, "loss": 0.0285, "step": 8968 }, { "epoch": 0.6163413963716328, "grad_norm": 0.5804142355918884, "learning_rate": 1.016517179256553e-05, "loss": 0.017, "step": 8969 }, { "epoch": 0.6164101154480484, "grad_norm": 0.49575746059417725, "learning_rate": 1.0162011523114806e-05, "loss": 0.0137, "step": 8970 }, { "epoch": 0.616478834524464, "grad_norm": 0.5429571270942688, "learning_rate": 1.0158851493328083e-05, "loss": 0.0118, "step": 8971 }, { "epoch": 0.6165475536008796, "grad_norm": 0.23327745497226715, "learning_rate": 1.0155691703361906e-05, "loss": 0.01, "step": 8972 }, { "epoch": 0.6166162726772952, "grad_norm": 0.3894199728965759, "learning_rate": 1.0152532153372802e-05, "loss": 0.0153, "step": 8973 }, { "epoch": 0.6166849917537108, "grad_norm": 0.1918589025735855, "learning_rate": 1.014937284351729e-05, "loss": 0.014, "step": 8974 }, { "epoch": 0.6167537108301264, "grad_norm": 0.9031983017921448, "learning_rate": 1.0146213773951868e-05, "loss": 0.0287, "step": 8975 }, { "epoch": 0.616822429906542, "grad_norm": 0.39494726061820984, "learning_rate": 1.0143054944833042e-05, "loss": 0.0132, "step": 8976 }, { "epoch": 0.6168911489829577, "grad_norm": 0.8531885147094727, "learning_rate": 1.013989635631729e-05, "loss": 0.026, "step": 8977 }, { "epoch": 0.6169598680593733, "grad_norm": 0.8938316106796265, "learning_rate": 1.013673800856108e-05, "loss": 0.0155, "step": 8978 }, { "epoch": 0.6170285871357889, "grad_norm": 0.5275007486343384, "learning_rate": 1.013357990172087e-05, "loss": 0.0108, "step": 8979 }, { "epoch": 0.6170973062122045, "grad_norm": 0.35613346099853516, "learning_rate": 1.0130422035953107e-05, "loss": 0.0169, "step": 8980 }, { "epoch": 0.6171660252886201, "grad_norm": 0.1366165727376938, "learning_rate": 1.0127264411414221e-05, "loss": 0.01, "step": 8981 }, { "epoch": 0.6172347443650358, "grad_norm": 0.36891627311706543, "learning_rate": 1.0124107028260644e-05, "loss": 0.0123, "step": 8982 }, { "epoch": 0.6173034634414514, "grad_norm": 0.7501509785652161, "learning_rate": 1.0120949886648777e-05, "loss": 0.0176, "step": 8983 }, { "epoch": 0.617372182517867, "grad_norm": 0.21528160572052002, "learning_rate": 1.0117792986735024e-05, "loss": 0.0142, "step": 8984 }, { "epoch": 0.6174409015942826, "grad_norm": 0.42263615131378174, "learning_rate": 1.0114636328675768e-05, "loss": 0.0109, "step": 8985 }, { "epoch": 0.6175096206706981, "grad_norm": 0.37521693110466003, "learning_rate": 1.0111479912627383e-05, "loss": 0.0212, "step": 8986 }, { "epoch": 0.6175783397471138, "grad_norm": 0.4907248914241791, "learning_rate": 1.010832373874623e-05, "loss": 0.0126, "step": 8987 }, { "epoch": 0.6176470588235294, "grad_norm": 0.11235475540161133, "learning_rate": 1.0105167807188667e-05, "loss": 0.0115, "step": 8988 }, { "epoch": 0.617715777899945, "grad_norm": 0.3001403510570526, "learning_rate": 1.0102012118111028e-05, "loss": 0.0098, "step": 8989 }, { "epoch": 0.6177844969763606, "grad_norm": 0.5159761309623718, "learning_rate": 1.0098856671669638e-05, "loss": 0.0109, "step": 8990 }, { "epoch": 0.6178532160527762, "grad_norm": 0.6389658451080322, "learning_rate": 1.0095701468020813e-05, "loss": 0.0152, "step": 8991 }, { "epoch": 0.6179219351291919, "grad_norm": 0.9017230272293091, "learning_rate": 1.0092546507320852e-05, "loss": 0.0199, "step": 8992 }, { "epoch": 0.6179906542056075, "grad_norm": 0.37597954273223877, "learning_rate": 1.0089391789726043e-05, "loss": 0.0131, "step": 8993 }, { "epoch": 0.6180593732820231, "grad_norm": 0.31941524147987366, "learning_rate": 1.0086237315392675e-05, "loss": 0.0153, "step": 8994 }, { "epoch": 0.6181280923584387, "grad_norm": 0.3607167601585388, "learning_rate": 1.008308308447701e-05, "loss": 0.0104, "step": 8995 }, { "epoch": 0.6181968114348543, "grad_norm": 0.39186179637908936, "learning_rate": 1.0079929097135298e-05, "loss": 0.0088, "step": 8996 }, { "epoch": 0.61826553051127, "grad_norm": 0.7557143568992615, "learning_rate": 1.0076775353523785e-05, "loss": 0.0196, "step": 8997 }, { "epoch": 0.6183342495876856, "grad_norm": 0.673448920249939, "learning_rate": 1.0073621853798698e-05, "loss": 0.015, "step": 8998 }, { "epoch": 0.6184029686641012, "grad_norm": 0.47389423847198486, "learning_rate": 1.0070468598116253e-05, "loss": 0.0145, "step": 8999 }, { "epoch": 0.6184716877405168, "grad_norm": 0.7130323052406311, "learning_rate": 1.0067315586632661e-05, "loss": 0.0086, "step": 9000 }, { "epoch": 0.6185404068169323, "grad_norm": 0.17488642036914825, "learning_rate": 1.0064162819504118e-05, "loss": 0.0071, "step": 9001 }, { "epoch": 0.618609125893348, "grad_norm": 0.5694195032119751, "learning_rate": 1.0061010296886796e-05, "loss": 0.0122, "step": 9002 }, { "epoch": 0.6186778449697636, "grad_norm": 1.3353179693222046, "learning_rate": 1.0057858018936869e-05, "loss": 0.0263, "step": 9003 }, { "epoch": 0.6187465640461792, "grad_norm": 0.5285196900367737, "learning_rate": 1.0054705985810498e-05, "loss": 0.0113, "step": 9004 }, { "epoch": 0.6188152831225948, "grad_norm": 0.5973756909370422, "learning_rate": 1.0051554197663823e-05, "loss": 0.0198, "step": 9005 }, { "epoch": 0.6188840021990104, "grad_norm": 0.48651617765426636, "learning_rate": 1.0048402654652984e-05, "loss": 0.0129, "step": 9006 }, { "epoch": 0.6189527212754261, "grad_norm": 0.21045264601707458, "learning_rate": 1.0045251356934093e-05, "loss": 0.014, "step": 9007 }, { "epoch": 0.6190214403518417, "grad_norm": 0.3342340588569641, "learning_rate": 1.0042100304663265e-05, "loss": 0.0148, "step": 9008 }, { "epoch": 0.6190901594282573, "grad_norm": 0.6010879278182983, "learning_rate": 1.0038949497996596e-05, "loss": 0.0109, "step": 9009 }, { "epoch": 0.6191588785046729, "grad_norm": 0.6002990007400513, "learning_rate": 1.0035798937090169e-05, "loss": 0.0164, "step": 9010 }, { "epoch": 0.6192275975810885, "grad_norm": 0.4346487820148468, "learning_rate": 1.003264862210005e-05, "loss": 0.0132, "step": 9011 }, { "epoch": 0.6192963166575042, "grad_norm": 1.0257230997085571, "learning_rate": 1.0029498553182312e-05, "loss": 0.0131, "step": 9012 }, { "epoch": 0.6193650357339198, "grad_norm": 0.7417117357254028, "learning_rate": 1.0026348730492998e-05, "loss": 0.0134, "step": 9013 }, { "epoch": 0.6194337548103354, "grad_norm": 0.913718044757843, "learning_rate": 1.0023199154188143e-05, "loss": 0.0153, "step": 9014 }, { "epoch": 0.619502473886751, "grad_norm": 0.2632337510585785, "learning_rate": 1.0020049824423769e-05, "loss": 0.0095, "step": 9015 }, { "epoch": 0.6195711929631665, "grad_norm": 0.6533312797546387, "learning_rate": 1.0016900741355887e-05, "loss": 0.0169, "step": 9016 }, { "epoch": 0.6196399120395822, "grad_norm": 0.6653249859809875, "learning_rate": 1.0013751905140497e-05, "loss": 0.0179, "step": 9017 }, { "epoch": 0.6197086311159978, "grad_norm": 0.3049449920654297, "learning_rate": 1.0010603315933588e-05, "loss": 0.0184, "step": 9018 }, { "epoch": 0.6197773501924134, "grad_norm": 0.2614418566226959, "learning_rate": 1.000745497389113e-05, "loss": 0.0116, "step": 9019 }, { "epoch": 0.619846069268829, "grad_norm": 0.8189747333526611, "learning_rate": 1.000430687916909e-05, "loss": 0.0148, "step": 9020 }, { "epoch": 0.6199147883452446, "grad_norm": 0.6198720932006836, "learning_rate": 1.0001159031923418e-05, "loss": 0.0144, "step": 9021 }, { "epoch": 0.6199835074216603, "grad_norm": 0.6006848216056824, "learning_rate": 9.99801143231005e-06, "loss": 0.0253, "step": 9022 }, { "epoch": 0.6200522264980759, "grad_norm": 0.628059983253479, "learning_rate": 9.99486408048491e-06, "loss": 0.0159, "step": 9023 }, { "epoch": 0.6201209455744915, "grad_norm": 0.3018515706062317, "learning_rate": 9.991716976603913e-06, "loss": 0.012, "step": 9024 }, { "epoch": 0.6201896646509071, "grad_norm": 0.9549418091773987, "learning_rate": 9.988570120822964e-06, "loss": 0.0134, "step": 9025 }, { "epoch": 0.6202583837273227, "grad_norm": 0.4124979078769684, "learning_rate": 9.985423513297948e-06, "loss": 0.0169, "step": 9026 }, { "epoch": 0.6203271028037384, "grad_norm": 0.9367163777351379, "learning_rate": 9.98227715418474e-06, "loss": 0.0144, "step": 9027 }, { "epoch": 0.620395821880154, "grad_norm": 0.7926726341247559, "learning_rate": 9.979131043639207e-06, "loss": 0.0237, "step": 9028 }, { "epoch": 0.6204645409565696, "grad_norm": 0.5072227716445923, "learning_rate": 9.975985181817192e-06, "loss": 0.0105, "step": 9029 }, { "epoch": 0.6205332600329851, "grad_norm": 0.5121964812278748, "learning_rate": 9.97283956887455e-06, "loss": 0.0168, "step": 9030 }, { "epoch": 0.6206019791094007, "grad_norm": 0.6779036521911621, "learning_rate": 9.969694204967098e-06, "loss": 0.0149, "step": 9031 }, { "epoch": 0.6206706981858163, "grad_norm": 0.994191586971283, "learning_rate": 9.966549090250657e-06, "loss": 0.0138, "step": 9032 }, { "epoch": 0.620739417262232, "grad_norm": 0.4643368422985077, "learning_rate": 9.963404224881021e-06, "loss": 0.012, "step": 9033 }, { "epoch": 0.6208081363386476, "grad_norm": 1.0856887102127075, "learning_rate": 9.960259609013988e-06, "loss": 0.0152, "step": 9034 }, { "epoch": 0.6208768554150632, "grad_norm": 0.4269302189350128, "learning_rate": 9.957115242805328e-06, "loss": 0.0134, "step": 9035 }, { "epoch": 0.6209455744914788, "grad_norm": 0.8235107660293579, "learning_rate": 9.953971126410813e-06, "loss": 0.0158, "step": 9036 }, { "epoch": 0.6210142935678944, "grad_norm": 0.3387930393218994, "learning_rate": 9.950827259986197e-06, "loss": 0.01, "step": 9037 }, { "epoch": 0.6210830126443101, "grad_norm": 0.19249151647090912, "learning_rate": 9.947683643687218e-06, "loss": 0.0118, "step": 9038 }, { "epoch": 0.6211517317207257, "grad_norm": 0.30405303835868835, "learning_rate": 9.944540277669601e-06, "loss": 0.0081, "step": 9039 }, { "epoch": 0.6212204507971413, "grad_norm": 1.5219154357910156, "learning_rate": 9.941397162089068e-06, "loss": 0.0198, "step": 9040 }, { "epoch": 0.6212891698735569, "grad_norm": 0.3984534442424774, "learning_rate": 9.938254297101316e-06, "loss": 0.0182, "step": 9041 }, { "epoch": 0.6213578889499725, "grad_norm": 0.7101995944976807, "learning_rate": 9.935111682862044e-06, "loss": 0.0165, "step": 9042 }, { "epoch": 0.6214266080263882, "grad_norm": 0.4144481420516968, "learning_rate": 9.931969319526924e-06, "loss": 0.0103, "step": 9043 }, { "epoch": 0.6214953271028038, "grad_norm": 0.5314571857452393, "learning_rate": 9.928827207251626e-06, "loss": 0.0207, "step": 9044 }, { "epoch": 0.6215640461792193, "grad_norm": 0.19055010378360748, "learning_rate": 9.925685346191801e-06, "loss": 0.0147, "step": 9045 }, { "epoch": 0.6216327652556349, "grad_norm": 0.5757684111595154, "learning_rate": 9.922543736503092e-06, "loss": 0.0162, "step": 9046 }, { "epoch": 0.6217014843320505, "grad_norm": 1.380478024482727, "learning_rate": 9.919402378341122e-06, "loss": 0.0129, "step": 9047 }, { "epoch": 0.6217702034084662, "grad_norm": 0.7036469578742981, "learning_rate": 9.91626127186152e-06, "loss": 0.0251, "step": 9048 }, { "epoch": 0.6218389224848818, "grad_norm": 0.6562557816505432, "learning_rate": 9.91312041721988e-06, "loss": 0.0159, "step": 9049 }, { "epoch": 0.6219076415612974, "grad_norm": 0.30464139580726624, "learning_rate": 9.9099798145718e-06, "loss": 0.0183, "step": 9050 }, { "epoch": 0.621976360637713, "grad_norm": 0.5379188656806946, "learning_rate": 9.906839464072851e-06, "loss": 0.009, "step": 9051 }, { "epoch": 0.6220450797141286, "grad_norm": 0.4265499711036682, "learning_rate": 9.903699365878606e-06, "loss": 0.0167, "step": 9052 }, { "epoch": 0.6221137987905443, "grad_norm": 0.7222778797149658, "learning_rate": 9.900559520144612e-06, "loss": 0.0164, "step": 9053 }, { "epoch": 0.6221825178669599, "grad_norm": 0.4353315234184265, "learning_rate": 9.89741992702642e-06, "loss": 0.0075, "step": 9054 }, { "epoch": 0.6222512369433755, "grad_norm": 0.524644136428833, "learning_rate": 9.894280586679554e-06, "loss": 0.0182, "step": 9055 }, { "epoch": 0.6223199560197911, "grad_norm": 1.8180694580078125, "learning_rate": 9.891141499259533e-06, "loss": 0.0165, "step": 9056 }, { "epoch": 0.6223886750962067, "grad_norm": 0.3864923417568207, "learning_rate": 9.888002664921858e-06, "loss": 0.0103, "step": 9057 }, { "epoch": 0.6224573941726224, "grad_norm": 0.3687126934528351, "learning_rate": 9.884864083822019e-06, "loss": 0.0148, "step": 9058 }, { "epoch": 0.622526113249038, "grad_norm": 0.5458483695983887, "learning_rate": 9.881725756115494e-06, "loss": 0.0113, "step": 9059 }, { "epoch": 0.6225948323254535, "grad_norm": 0.840772271156311, "learning_rate": 9.878587681957757e-06, "loss": 0.0237, "step": 9060 }, { "epoch": 0.6226635514018691, "grad_norm": 0.34427177906036377, "learning_rate": 9.875449861504257e-06, "loss": 0.0157, "step": 9061 }, { "epoch": 0.6227322704782847, "grad_norm": 0.40954166650772095, "learning_rate": 9.872312294910436e-06, "loss": 0.025, "step": 9062 }, { "epoch": 0.6228009895547004, "grad_norm": 0.33163484930992126, "learning_rate": 9.86917498233172e-06, "loss": 0.0085, "step": 9063 }, { "epoch": 0.622869708631116, "grad_norm": 0.8261091113090515, "learning_rate": 9.866037923923528e-06, "loss": 0.0193, "step": 9064 }, { "epoch": 0.6229384277075316, "grad_norm": 0.44826334714889526, "learning_rate": 9.862901119841255e-06, "loss": 0.0125, "step": 9065 }, { "epoch": 0.6230071467839472, "grad_norm": 0.3326023519039154, "learning_rate": 9.859764570240307e-06, "loss": 0.0105, "step": 9066 }, { "epoch": 0.6230758658603628, "grad_norm": 1.2733023166656494, "learning_rate": 9.856628275276053e-06, "loss": 0.0374, "step": 9067 }, { "epoch": 0.6231445849367785, "grad_norm": 1.5533404350280762, "learning_rate": 9.853492235103862e-06, "loss": 0.0166, "step": 9068 }, { "epoch": 0.6232133040131941, "grad_norm": 0.3693157434463501, "learning_rate": 9.850356449879081e-06, "loss": 0.0119, "step": 9069 }, { "epoch": 0.6232820230896097, "grad_norm": 0.2622916102409363, "learning_rate": 9.847220919757056e-06, "loss": 0.0113, "step": 9070 }, { "epoch": 0.6233507421660253, "grad_norm": 0.4437004029750824, "learning_rate": 9.84408564489311e-06, "loss": 0.0147, "step": 9071 }, { "epoch": 0.6234194612424409, "grad_norm": 1.257594108581543, "learning_rate": 9.840950625442563e-06, "loss": 0.0141, "step": 9072 }, { "epoch": 0.6234881803188566, "grad_norm": 1.0364922285079956, "learning_rate": 9.837815861560716e-06, "loss": 0.0139, "step": 9073 }, { "epoch": 0.6235568993952721, "grad_norm": 0.3853611350059509, "learning_rate": 9.83468135340286e-06, "loss": 0.0129, "step": 9074 }, { "epoch": 0.6236256184716877, "grad_norm": 0.7352455854415894, "learning_rate": 9.83154710112427e-06, "loss": 0.0133, "step": 9075 }, { "epoch": 0.6236943375481033, "grad_norm": 1.765066385269165, "learning_rate": 9.82841310488021e-06, "loss": 0.0132, "step": 9076 }, { "epoch": 0.6237630566245189, "grad_norm": 0.5067014694213867, "learning_rate": 9.82527936482593e-06, "loss": 0.0134, "step": 9077 }, { "epoch": 0.6238317757009346, "grad_norm": 0.49423548579216003, "learning_rate": 9.822145881116674e-06, "loss": 0.0148, "step": 9078 }, { "epoch": 0.6239004947773502, "grad_norm": 0.5169110298156738, "learning_rate": 9.819012653907665e-06, "loss": 0.0159, "step": 9079 }, { "epoch": 0.6239692138537658, "grad_norm": 1.0134295225143433, "learning_rate": 9.81587968335412e-06, "loss": 0.0138, "step": 9080 }, { "epoch": 0.6240379329301814, "grad_norm": 0.17001192271709442, "learning_rate": 9.812746969611233e-06, "loss": 0.0092, "step": 9081 }, { "epoch": 0.624106652006597, "grad_norm": 0.7072017788887024, "learning_rate": 9.8096145128342e-06, "loss": 0.0179, "step": 9082 }, { "epoch": 0.6241753710830127, "grad_norm": 0.9405797719955444, "learning_rate": 9.806482313178187e-06, "loss": 0.0171, "step": 9083 }, { "epoch": 0.6242440901594283, "grad_norm": 0.5771417021751404, "learning_rate": 9.803350370798368e-06, "loss": 0.0108, "step": 9084 }, { "epoch": 0.6243128092358439, "grad_norm": 0.2839851975440979, "learning_rate": 9.800218685849888e-06, "loss": 0.0117, "step": 9085 }, { "epoch": 0.6243815283122595, "grad_norm": 0.6886816024780273, "learning_rate": 9.797087258487883e-06, "loss": 0.0166, "step": 9086 }, { "epoch": 0.624450247388675, "grad_norm": 0.5283339023590088, "learning_rate": 9.793956088867481e-06, "loss": 0.0111, "step": 9087 }, { "epoch": 0.6245189664650908, "grad_norm": 0.8714007139205933, "learning_rate": 9.79082517714379e-06, "loss": 0.0118, "step": 9088 }, { "epoch": 0.6245876855415063, "grad_norm": 0.5023754239082336, "learning_rate": 9.787694523471902e-06, "loss": 0.0106, "step": 9089 }, { "epoch": 0.6246564046179219, "grad_norm": 0.7823940515518188, "learning_rate": 9.78456412800692e-06, "loss": 0.0178, "step": 9090 }, { "epoch": 0.6247251236943375, "grad_norm": 0.456384539604187, "learning_rate": 9.781433990903909e-06, "loss": 0.0114, "step": 9091 }, { "epoch": 0.6247938427707531, "grad_norm": 0.6783839464187622, "learning_rate": 9.778304112317928e-06, "loss": 0.0202, "step": 9092 }, { "epoch": 0.6248625618471688, "grad_norm": 0.632279634475708, "learning_rate": 9.775174492404025e-06, "loss": 0.0122, "step": 9093 }, { "epoch": 0.6249312809235844, "grad_norm": 0.3136228919029236, "learning_rate": 9.772045131317236e-06, "loss": 0.0119, "step": 9094 }, { "epoch": 0.625, "grad_norm": 0.6209571957588196, "learning_rate": 9.76891602921258e-06, "loss": 0.0188, "step": 9095 }, { "epoch": 0.6250687190764156, "grad_norm": 0.49179884791374207, "learning_rate": 9.765787186245073e-06, "loss": 0.015, "step": 9096 }, { "epoch": 0.6251374381528312, "grad_norm": 0.7010318040847778, "learning_rate": 9.762658602569709e-06, "loss": 0.018, "step": 9097 }, { "epoch": 0.6252061572292469, "grad_norm": 0.8303945064544678, "learning_rate": 9.759530278341468e-06, "loss": 0.0094, "step": 9098 }, { "epoch": 0.6252748763056625, "grad_norm": 0.5286616086959839, "learning_rate": 9.756402213715324e-06, "loss": 0.0177, "step": 9099 }, { "epoch": 0.6253435953820781, "grad_norm": 0.454606294631958, "learning_rate": 9.753274408846235e-06, "loss": 0.0118, "step": 9100 }, { "epoch": 0.6254123144584937, "grad_norm": 0.8443323373794556, "learning_rate": 9.750146863889136e-06, "loss": 0.0366, "step": 9101 }, { "epoch": 0.6254810335349092, "grad_norm": 0.6097177267074585, "learning_rate": 9.747019578998977e-06, "loss": 0.0103, "step": 9102 }, { "epoch": 0.625549752611325, "grad_norm": 0.458815336227417, "learning_rate": 9.743892554330668e-06, "loss": 0.0135, "step": 9103 }, { "epoch": 0.6256184716877405, "grad_norm": 0.4044078588485718, "learning_rate": 9.740765790039115e-06, "loss": 0.0269, "step": 9104 }, { "epoch": 0.6256871907641561, "grad_norm": 0.6886166930198669, "learning_rate": 9.737639286279211e-06, "loss": 0.0093, "step": 9105 }, { "epoch": 0.6257559098405717, "grad_norm": 0.6951286792755127, "learning_rate": 9.73451304320584e-06, "loss": 0.0172, "step": 9106 }, { "epoch": 0.6258246289169873, "grad_norm": 1.4170589447021484, "learning_rate": 9.731387060973859e-06, "loss": 0.0154, "step": 9107 }, { "epoch": 0.625893347993403, "grad_norm": 0.7256481647491455, "learning_rate": 9.72826133973814e-06, "loss": 0.0144, "step": 9108 }, { "epoch": 0.6259620670698186, "grad_norm": 0.6932698488235474, "learning_rate": 9.725135879653514e-06, "loss": 0.0154, "step": 9109 }, { "epoch": 0.6260307861462342, "grad_norm": 0.3899860680103302, "learning_rate": 9.722010680874814e-06, "loss": 0.0122, "step": 9110 }, { "epoch": 0.6260995052226498, "grad_norm": 0.3960842192173004, "learning_rate": 9.718885743556853e-06, "loss": 0.0112, "step": 9111 }, { "epoch": 0.6261682242990654, "grad_norm": 0.17722533643245697, "learning_rate": 9.715761067854431e-06, "loss": 0.0124, "step": 9112 }, { "epoch": 0.6262369433754811, "grad_norm": 0.2932622730731964, "learning_rate": 9.712636653922345e-06, "loss": 0.0111, "step": 9113 }, { "epoch": 0.6263056624518967, "grad_norm": 0.5521051287651062, "learning_rate": 9.709512501915367e-06, "loss": 0.0147, "step": 9114 }, { "epoch": 0.6263743815283123, "grad_norm": 0.43762463331222534, "learning_rate": 9.706388611988267e-06, "loss": 0.0123, "step": 9115 }, { "epoch": 0.6264431006047279, "grad_norm": 0.8348477482795715, "learning_rate": 9.703264984295787e-06, "loss": 0.0188, "step": 9116 }, { "epoch": 0.6265118196811434, "grad_norm": 0.40297070145606995, "learning_rate": 9.700141618992673e-06, "loss": 0.0147, "step": 9117 }, { "epoch": 0.6265805387575591, "grad_norm": 0.5125356912612915, "learning_rate": 9.697018516233648e-06, "loss": 0.0147, "step": 9118 }, { "epoch": 0.6266492578339747, "grad_norm": 0.5607818365097046, "learning_rate": 9.693895676173414e-06, "loss": 0.0127, "step": 9119 }, { "epoch": 0.6267179769103903, "grad_norm": 0.39568567276000977, "learning_rate": 9.690773098966687e-06, "loss": 0.0101, "step": 9120 }, { "epoch": 0.6267866959868059, "grad_norm": 0.6343554258346558, "learning_rate": 9.687650784768147e-06, "loss": 0.0152, "step": 9121 }, { "epoch": 0.6268554150632215, "grad_norm": 0.20325510203838348, "learning_rate": 9.684528733732465e-06, "loss": 0.0067, "step": 9122 }, { "epoch": 0.6269241341396372, "grad_norm": 0.81354159116745, "learning_rate": 9.6814069460143e-06, "loss": 0.0203, "step": 9123 }, { "epoch": 0.6269928532160528, "grad_norm": 0.36240652203559875, "learning_rate": 9.6782854217683e-06, "loss": 0.013, "step": 9124 }, { "epoch": 0.6270615722924684, "grad_norm": 0.5569586753845215, "learning_rate": 9.675164161149096e-06, "loss": 0.0174, "step": 9125 }, { "epoch": 0.627130291368884, "grad_norm": 0.8149144649505615, "learning_rate": 9.672043164311314e-06, "loss": 0.0163, "step": 9126 }, { "epoch": 0.6271990104452996, "grad_norm": 0.5716319680213928, "learning_rate": 9.668922431409564e-06, "loss": 0.0164, "step": 9127 }, { "epoch": 0.6272677295217153, "grad_norm": 1.1160814762115479, "learning_rate": 9.665801962598434e-06, "loss": 0.0195, "step": 9128 }, { "epoch": 0.6273364485981309, "grad_norm": 0.996172308921814, "learning_rate": 9.662681758032507e-06, "loss": 0.0144, "step": 9129 }, { "epoch": 0.6274051676745465, "grad_norm": 0.9179590344429016, "learning_rate": 9.659561817866355e-06, "loss": 0.0117, "step": 9130 }, { "epoch": 0.627473886750962, "grad_norm": 0.7449242472648621, "learning_rate": 9.656442142254528e-06, "loss": 0.0194, "step": 9131 }, { "epoch": 0.6275426058273776, "grad_norm": 1.3435392379760742, "learning_rate": 9.653322731351572e-06, "loss": 0.0192, "step": 9132 }, { "epoch": 0.6276113249037933, "grad_norm": 0.342902272939682, "learning_rate": 9.650203585312019e-06, "loss": 0.0137, "step": 9133 }, { "epoch": 0.6276800439802089, "grad_norm": 0.47243404388427734, "learning_rate": 9.647084704290378e-06, "loss": 0.0112, "step": 9134 }, { "epoch": 0.6277487630566245, "grad_norm": 0.8251209259033203, "learning_rate": 9.64396608844116e-06, "loss": 0.0151, "step": 9135 }, { "epoch": 0.6278174821330401, "grad_norm": 0.6340909004211426, "learning_rate": 9.640847737918847e-06, "loss": 0.0146, "step": 9136 }, { "epoch": 0.6278862012094557, "grad_norm": 0.531957745552063, "learning_rate": 9.637729652877912e-06, "loss": 0.0164, "step": 9137 }, { "epoch": 0.6279549202858714, "grad_norm": 0.6633660197257996, "learning_rate": 9.634611833472835e-06, "loss": 0.0166, "step": 9138 }, { "epoch": 0.628023639362287, "grad_norm": 0.47340312600135803, "learning_rate": 9.631494279858056e-06, "loss": 0.0133, "step": 9139 }, { "epoch": 0.6280923584387026, "grad_norm": 0.6624581813812256, "learning_rate": 9.628376992188013e-06, "loss": 0.0145, "step": 9140 }, { "epoch": 0.6281610775151182, "grad_norm": 0.13369965553283691, "learning_rate": 9.62525997061713e-06, "loss": 0.01, "step": 9141 }, { "epoch": 0.6282297965915338, "grad_norm": 0.8750709295272827, "learning_rate": 9.622143215299819e-06, "loss": 0.0138, "step": 9142 }, { "epoch": 0.6282985156679495, "grad_norm": 0.7603495717048645, "learning_rate": 9.61902672639047e-06, "loss": 0.0128, "step": 9143 }, { "epoch": 0.6283672347443651, "grad_norm": 0.2880096137523651, "learning_rate": 9.615910504043479e-06, "loss": 0.0142, "step": 9144 }, { "epoch": 0.6284359538207807, "grad_norm": 1.02448570728302, "learning_rate": 9.612794548413214e-06, "loss": 0.0178, "step": 9145 }, { "epoch": 0.6285046728971962, "grad_norm": 0.3776891231536865, "learning_rate": 9.60967885965403e-06, "loss": 0.0175, "step": 9146 }, { "epoch": 0.6285733919736118, "grad_norm": 0.6118411421775818, "learning_rate": 9.606563437920275e-06, "loss": 0.0199, "step": 9147 }, { "epoch": 0.6286421110500275, "grad_norm": 0.4522716999053955, "learning_rate": 9.603448283366278e-06, "loss": 0.0139, "step": 9148 }, { "epoch": 0.6287108301264431, "grad_norm": 0.4995255768299103, "learning_rate": 9.600333396146356e-06, "loss": 0.0084, "step": 9149 }, { "epoch": 0.6287795492028587, "grad_norm": 0.30656400322914124, "learning_rate": 9.597218776414817e-06, "loss": 0.0096, "step": 9150 }, { "epoch": 0.6288482682792743, "grad_norm": 0.42401471734046936, "learning_rate": 9.594104424325954e-06, "loss": 0.008, "step": 9151 }, { "epoch": 0.6289169873556899, "grad_norm": 0.55208820104599, "learning_rate": 9.590990340034044e-06, "loss": 0.018, "step": 9152 }, { "epoch": 0.6289857064321056, "grad_norm": 0.6303206086158752, "learning_rate": 9.58787652369335e-06, "loss": 0.0194, "step": 9153 }, { "epoch": 0.6290544255085212, "grad_norm": 0.9689244627952576, "learning_rate": 9.58476297545813e-06, "loss": 0.013, "step": 9154 }, { "epoch": 0.6291231445849368, "grad_norm": 0.47406280040740967, "learning_rate": 9.58164969548261e-06, "loss": 0.0146, "step": 9155 }, { "epoch": 0.6291918636613524, "grad_norm": 2.4899497032165527, "learning_rate": 9.578536683921032e-06, "loss": 0.0151, "step": 9156 }, { "epoch": 0.629260582737768, "grad_norm": 0.20168037712574005, "learning_rate": 9.575423940927597e-06, "loss": 0.0109, "step": 9157 }, { "epoch": 0.6293293018141837, "grad_norm": 0.2791864573955536, "learning_rate": 9.572311466656512e-06, "loss": 0.0177, "step": 9158 }, { "epoch": 0.6293980208905993, "grad_norm": 0.31199270486831665, "learning_rate": 9.569199261261956e-06, "loss": 0.0101, "step": 9159 }, { "epoch": 0.6294667399670149, "grad_norm": 0.6478461027145386, "learning_rate": 9.566087324898106e-06, "loss": 0.0173, "step": 9160 }, { "epoch": 0.6295354590434304, "grad_norm": 0.3934778869152069, "learning_rate": 9.562975657719108e-06, "loss": 0.0172, "step": 9161 }, { "epoch": 0.629604178119846, "grad_norm": 0.6588971614837646, "learning_rate": 9.559864259879127e-06, "loss": 0.0118, "step": 9162 }, { "epoch": 0.6296728971962616, "grad_norm": 0.2020416110754013, "learning_rate": 9.556753131532283e-06, "loss": 0.0104, "step": 9163 }, { "epoch": 0.6297416162726773, "grad_norm": 0.204739049077034, "learning_rate": 9.5536422728327e-06, "loss": 0.0104, "step": 9164 }, { "epoch": 0.6298103353490929, "grad_norm": 0.6116618514060974, "learning_rate": 9.550531683934484e-06, "loss": 0.0119, "step": 9165 }, { "epoch": 0.6298790544255085, "grad_norm": 0.5687517523765564, "learning_rate": 9.54742136499172e-06, "loss": 0.0197, "step": 9166 }, { "epoch": 0.6299477735019241, "grad_norm": 0.8327916264533997, "learning_rate": 9.544311316158491e-06, "loss": 0.0202, "step": 9167 }, { "epoch": 0.6300164925783397, "grad_norm": 0.7636848092079163, "learning_rate": 9.541201537588866e-06, "loss": 0.0135, "step": 9168 }, { "epoch": 0.6300852116547554, "grad_norm": 0.4212782084941864, "learning_rate": 9.538092029436893e-06, "loss": 0.0084, "step": 9169 }, { "epoch": 0.630153930731171, "grad_norm": 0.5667003393173218, "learning_rate": 9.534982791856615e-06, "loss": 0.0061, "step": 9170 }, { "epoch": 0.6302226498075866, "grad_norm": 0.6705639958381653, "learning_rate": 9.53187382500205e-06, "loss": 0.0144, "step": 9171 }, { "epoch": 0.6302913688840022, "grad_norm": 0.3482069671154022, "learning_rate": 9.528765129027216e-06, "loss": 0.0097, "step": 9172 }, { "epoch": 0.6303600879604178, "grad_norm": 0.7365918755531311, "learning_rate": 9.5256567040861e-06, "loss": 0.0201, "step": 9173 }, { "epoch": 0.6304288070368335, "grad_norm": 0.571771502494812, "learning_rate": 9.522548550332705e-06, "loss": 0.0111, "step": 9174 }, { "epoch": 0.630497526113249, "grad_norm": 0.45698660612106323, "learning_rate": 9.519440667920995e-06, "loss": 0.0123, "step": 9175 }, { "epoch": 0.6305662451896646, "grad_norm": 0.6243259310722351, "learning_rate": 9.516333057004924e-06, "loss": 0.0097, "step": 9176 }, { "epoch": 0.6306349642660802, "grad_norm": 0.2826496660709381, "learning_rate": 9.513225717738439e-06, "loss": 0.0084, "step": 9177 }, { "epoch": 0.6307036833424958, "grad_norm": 0.8318489789962769, "learning_rate": 9.510118650275474e-06, "loss": 0.017, "step": 9178 }, { "epoch": 0.6307724024189115, "grad_norm": 1.6736854314804077, "learning_rate": 9.507011854769942e-06, "loss": 0.0117, "step": 9179 }, { "epoch": 0.6308411214953271, "grad_norm": 0.4579734802246094, "learning_rate": 9.503905331375746e-06, "loss": 0.018, "step": 9180 }, { "epoch": 0.6309098405717427, "grad_norm": 0.4141102731227875, "learning_rate": 9.500799080246785e-06, "loss": 0.0141, "step": 9181 }, { "epoch": 0.6309785596481583, "grad_norm": 0.4682782292366028, "learning_rate": 9.49769310153693e-06, "loss": 0.0116, "step": 9182 }, { "epoch": 0.6310472787245739, "grad_norm": 0.5519015789031982, "learning_rate": 9.49458739540005e-06, "loss": 0.0138, "step": 9183 }, { "epoch": 0.6311159978009896, "grad_norm": 0.4926134943962097, "learning_rate": 9.491481961989989e-06, "loss": 0.0103, "step": 9184 }, { "epoch": 0.6311847168774052, "grad_norm": 0.9769462943077087, "learning_rate": 9.488376801460587e-06, "loss": 0.0183, "step": 9185 }, { "epoch": 0.6312534359538208, "grad_norm": 0.9891361594200134, "learning_rate": 9.485271913965665e-06, "loss": 0.0135, "step": 9186 }, { "epoch": 0.6313221550302364, "grad_norm": 0.44871458411216736, "learning_rate": 9.482167299659039e-06, "loss": 0.0161, "step": 9187 }, { "epoch": 0.631390874106652, "grad_norm": 0.4535170793533325, "learning_rate": 9.479062958694497e-06, "loss": 0.0091, "step": 9188 }, { "epoch": 0.6314595931830677, "grad_norm": 0.6509043574333191, "learning_rate": 9.475958891225828e-06, "loss": 0.0224, "step": 9189 }, { "epoch": 0.6315283122594832, "grad_norm": 0.6705689430236816, "learning_rate": 9.472855097406799e-06, "loss": 0.0221, "step": 9190 }, { "epoch": 0.6315970313358988, "grad_norm": 0.3184114992618561, "learning_rate": 9.469751577391162e-06, "loss": 0.0138, "step": 9191 }, { "epoch": 0.6316657504123144, "grad_norm": 1.806372880935669, "learning_rate": 9.466648331332658e-06, "loss": 0.0256, "step": 9192 }, { "epoch": 0.63173446948873, "grad_norm": 0.5204300284385681, "learning_rate": 9.463545359385027e-06, "loss": 0.0279, "step": 9193 }, { "epoch": 0.6318031885651457, "grad_norm": 1.717298150062561, "learning_rate": 9.460442661701974e-06, "loss": 0.0147, "step": 9194 }, { "epoch": 0.6318719076415613, "grad_norm": 0.6454567313194275, "learning_rate": 9.457340238437206e-06, "loss": 0.011, "step": 9195 }, { "epoch": 0.6319406267179769, "grad_norm": 0.22878921031951904, "learning_rate": 9.454238089744405e-06, "loss": 0.0102, "step": 9196 }, { "epoch": 0.6320093457943925, "grad_norm": 0.45732271671295166, "learning_rate": 9.451136215777246e-06, "loss": 0.0154, "step": 9197 }, { "epoch": 0.6320780648708081, "grad_norm": 0.8811895847320557, "learning_rate": 9.448034616689388e-06, "loss": 0.0203, "step": 9198 }, { "epoch": 0.6321467839472238, "grad_norm": 0.9259005188941956, "learning_rate": 9.444933292634484e-06, "loss": 0.0186, "step": 9199 }, { "epoch": 0.6322155030236394, "grad_norm": 0.8162262439727783, "learning_rate": 9.441832243766166e-06, "loss": 0.0136, "step": 9200 }, { "epoch": 0.632284222100055, "grad_norm": 0.5797758102416992, "learning_rate": 9.43873147023805e-06, "loss": 0.0245, "step": 9201 }, { "epoch": 0.6323529411764706, "grad_norm": 0.5726160407066345, "learning_rate": 9.435630972203746e-06, "loss": 0.0127, "step": 9202 }, { "epoch": 0.6324216602528862, "grad_norm": 0.4900174140930176, "learning_rate": 9.43253074981684e-06, "loss": 0.0098, "step": 9203 }, { "epoch": 0.6324903793293019, "grad_norm": 0.5873688459396362, "learning_rate": 9.429430803230914e-06, "loss": 0.0117, "step": 9204 }, { "epoch": 0.6325590984057174, "grad_norm": 0.43507876992225647, "learning_rate": 9.426331132599536e-06, "loss": 0.0092, "step": 9205 }, { "epoch": 0.632627817482133, "grad_norm": 0.4479186236858368, "learning_rate": 9.423231738076255e-06, "loss": 0.0136, "step": 9206 }, { "epoch": 0.6326965365585486, "grad_norm": 2.090020179748535, "learning_rate": 9.420132619814608e-06, "loss": 0.0281, "step": 9207 }, { "epoch": 0.6327652556349642, "grad_norm": 0.7031213045120239, "learning_rate": 9.417033777968119e-06, "loss": 0.0129, "step": 9208 }, { "epoch": 0.6328339747113799, "grad_norm": 0.36676186323165894, "learning_rate": 9.413935212690295e-06, "loss": 0.0167, "step": 9209 }, { "epoch": 0.6329026937877955, "grad_norm": 0.5675680637359619, "learning_rate": 9.410836924134638e-06, "loss": 0.0169, "step": 9210 }, { "epoch": 0.6329714128642111, "grad_norm": 1.2056550979614258, "learning_rate": 9.40773891245463e-06, "loss": 0.0207, "step": 9211 }, { "epoch": 0.6330401319406267, "grad_norm": 0.49899739027023315, "learning_rate": 9.40464117780374e-06, "loss": 0.0175, "step": 9212 }, { "epoch": 0.6331088510170423, "grad_norm": 0.44832441210746765, "learning_rate": 9.401543720335423e-06, "loss": 0.0189, "step": 9213 }, { "epoch": 0.633177570093458, "grad_norm": 1.0179743766784668, "learning_rate": 9.398446540203118e-06, "loss": 0.0168, "step": 9214 }, { "epoch": 0.6332462891698736, "grad_norm": 0.568623960018158, "learning_rate": 9.395349637560258e-06, "loss": 0.0134, "step": 9215 }, { "epoch": 0.6333150082462892, "grad_norm": 0.835952877998352, "learning_rate": 9.392253012560246e-06, "loss": 0.0174, "step": 9216 }, { "epoch": 0.6333837273227048, "grad_norm": 0.4466508626937866, "learning_rate": 9.389156665356499e-06, "loss": 0.011, "step": 9217 }, { "epoch": 0.6334524463991204, "grad_norm": 0.5354678630828857, "learning_rate": 9.386060596102393e-06, "loss": 0.0126, "step": 9218 }, { "epoch": 0.633521165475536, "grad_norm": 0.7256777286529541, "learning_rate": 9.382964804951303e-06, "loss": 0.0095, "step": 9219 }, { "epoch": 0.6335898845519516, "grad_norm": 0.5331791043281555, "learning_rate": 9.379869292056591e-06, "loss": 0.0149, "step": 9220 }, { "epoch": 0.6336586036283672, "grad_norm": 0.6005426645278931, "learning_rate": 9.3767740575716e-06, "loss": 0.0155, "step": 9221 }, { "epoch": 0.6337273227047828, "grad_norm": 0.5707093477249146, "learning_rate": 9.373679101649655e-06, "loss": 0.0157, "step": 9222 }, { "epoch": 0.6337960417811984, "grad_norm": 0.4885811507701874, "learning_rate": 9.370584424444086e-06, "loss": 0.0165, "step": 9223 }, { "epoch": 0.6338647608576141, "grad_norm": 0.3624812364578247, "learning_rate": 9.367490026108189e-06, "loss": 0.0151, "step": 9224 }, { "epoch": 0.6339334799340297, "grad_norm": 0.40452414751052856, "learning_rate": 9.364395906795256e-06, "loss": 0.0133, "step": 9225 }, { "epoch": 0.6340021990104453, "grad_norm": 0.11826340854167938, "learning_rate": 9.361302066658562e-06, "loss": 0.0062, "step": 9226 }, { "epoch": 0.6340709180868609, "grad_norm": 0.5899996757507324, "learning_rate": 9.358208505851374e-06, "loss": 0.0127, "step": 9227 }, { "epoch": 0.6341396371632765, "grad_norm": 0.7033888101577759, "learning_rate": 9.355115224526936e-06, "loss": 0.0135, "step": 9228 }, { "epoch": 0.6342083562396922, "grad_norm": 0.34655335545539856, "learning_rate": 9.352022222838486e-06, "loss": 0.0097, "step": 9229 }, { "epoch": 0.6342770753161078, "grad_norm": 0.5828232169151306, "learning_rate": 9.348929500939241e-06, "loss": 0.0213, "step": 9230 }, { "epoch": 0.6343457943925234, "grad_norm": 0.49783793091773987, "learning_rate": 9.345837058982412e-06, "loss": 0.0114, "step": 9231 }, { "epoch": 0.634414513468939, "grad_norm": 0.3037165701389313, "learning_rate": 9.342744897121192e-06, "loss": 0.0164, "step": 9232 }, { "epoch": 0.6344832325453545, "grad_norm": 1.1113985776901245, "learning_rate": 9.339653015508758e-06, "loss": 0.0216, "step": 9233 }, { "epoch": 0.6345519516217702, "grad_norm": 0.5665590763092041, "learning_rate": 9.33656141429827e-06, "loss": 0.0167, "step": 9234 }, { "epoch": 0.6346206706981858, "grad_norm": 0.5855709314346313, "learning_rate": 9.333470093642892e-06, "loss": 0.017, "step": 9235 }, { "epoch": 0.6346893897746014, "grad_norm": 0.525178849697113, "learning_rate": 9.330379053695757e-06, "loss": 0.0132, "step": 9236 }, { "epoch": 0.634758108851017, "grad_norm": 0.541129469871521, "learning_rate": 9.327288294609985e-06, "loss": 0.0119, "step": 9237 }, { "epoch": 0.6348268279274326, "grad_norm": 0.5190395712852478, "learning_rate": 9.324197816538689e-06, "loss": 0.0143, "step": 9238 }, { "epoch": 0.6348955470038483, "grad_norm": 0.8197696805000305, "learning_rate": 9.321107619634966e-06, "loss": 0.0197, "step": 9239 }, { "epoch": 0.6349642660802639, "grad_norm": 0.4853947162628174, "learning_rate": 9.318017704051891e-06, "loss": 0.0159, "step": 9240 }, { "epoch": 0.6350329851566795, "grad_norm": 0.5752642750740051, "learning_rate": 9.314928069942544e-06, "loss": 0.0147, "step": 9241 }, { "epoch": 0.6351017042330951, "grad_norm": 0.42006388306617737, "learning_rate": 9.311838717459964e-06, "loss": 0.0165, "step": 9242 }, { "epoch": 0.6351704233095107, "grad_norm": 0.7114643454551697, "learning_rate": 9.308749646757206e-06, "loss": 0.0125, "step": 9243 }, { "epoch": 0.6352391423859264, "grad_norm": 0.7460581064224243, "learning_rate": 9.305660857987288e-06, "loss": 0.0133, "step": 9244 }, { "epoch": 0.635307861462342, "grad_norm": 0.8732593059539795, "learning_rate": 9.302572351303227e-06, "loss": 0.0113, "step": 9245 }, { "epoch": 0.6353765805387576, "grad_norm": 0.4816739559173584, "learning_rate": 9.299484126858013e-06, "loss": 0.0088, "step": 9246 }, { "epoch": 0.6354452996151732, "grad_norm": 0.3001287579536438, "learning_rate": 9.296396184804639e-06, "loss": 0.0113, "step": 9247 }, { "epoch": 0.6355140186915887, "grad_norm": 0.41169512271881104, "learning_rate": 9.293308525296073e-06, "loss": 0.0126, "step": 9248 }, { "epoch": 0.6355827377680044, "grad_norm": 0.5542407035827637, "learning_rate": 9.290221148485267e-06, "loss": 0.0183, "step": 9249 }, { "epoch": 0.63565145684442, "grad_norm": 0.33031561970710754, "learning_rate": 9.28713405452517e-06, "loss": 0.0101, "step": 9250 }, { "epoch": 0.6357201759208356, "grad_norm": 0.5396732687950134, "learning_rate": 9.284047243568705e-06, "loss": 0.0121, "step": 9251 }, { "epoch": 0.6357888949972512, "grad_norm": 0.5689143538475037, "learning_rate": 9.280960715768782e-06, "loss": 0.0116, "step": 9252 }, { "epoch": 0.6358576140736668, "grad_norm": 0.49671831727027893, "learning_rate": 9.277874471278314e-06, "loss": 0.0116, "step": 9253 }, { "epoch": 0.6359263331500825, "grad_norm": 1.0761367082595825, "learning_rate": 9.27478851025018e-06, "loss": 0.016, "step": 9254 }, { "epoch": 0.6359950522264981, "grad_norm": 0.4295491874217987, "learning_rate": 9.271702832837255e-06, "loss": 0.0168, "step": 9255 }, { "epoch": 0.6360637713029137, "grad_norm": 0.53962242603302, "learning_rate": 9.268617439192392e-06, "loss": 0.0133, "step": 9256 }, { "epoch": 0.6361324903793293, "grad_norm": 0.7224099636077881, "learning_rate": 9.265532329468438e-06, "loss": 0.0134, "step": 9257 }, { "epoch": 0.6362012094557449, "grad_norm": 0.8885031938552856, "learning_rate": 9.262447503818217e-06, "loss": 0.0188, "step": 9258 }, { "epoch": 0.6362699285321606, "grad_norm": 0.31082913279533386, "learning_rate": 9.259362962394556e-06, "loss": 0.0156, "step": 9259 }, { "epoch": 0.6363386476085762, "grad_norm": 0.5062836408615112, "learning_rate": 9.256278705350254e-06, "loss": 0.0124, "step": 9260 }, { "epoch": 0.6364073666849918, "grad_norm": 0.340507447719574, "learning_rate": 9.253194732838095e-06, "loss": 0.0116, "step": 9261 }, { "epoch": 0.6364760857614074, "grad_norm": 0.8228510618209839, "learning_rate": 9.250111045010855e-06, "loss": 0.0122, "step": 9262 }, { "epoch": 0.6365448048378229, "grad_norm": 0.2755267918109894, "learning_rate": 9.24702764202129e-06, "loss": 0.0113, "step": 9263 }, { "epoch": 0.6366135239142386, "grad_norm": 0.75605309009552, "learning_rate": 9.243944524022147e-06, "loss": 0.0212, "step": 9264 }, { "epoch": 0.6366822429906542, "grad_norm": 0.5824092626571655, "learning_rate": 9.240861691166162e-06, "loss": 0.0131, "step": 9265 }, { "epoch": 0.6367509620670698, "grad_norm": 0.37320834398269653, "learning_rate": 9.237779143606048e-06, "loss": 0.0104, "step": 9266 }, { "epoch": 0.6368196811434854, "grad_norm": 0.4461609423160553, "learning_rate": 9.234696881494507e-06, "loss": 0.0106, "step": 9267 }, { "epoch": 0.636888400219901, "grad_norm": 0.34805622696876526, "learning_rate": 9.23161490498423e-06, "loss": 0.0153, "step": 9268 }, { "epoch": 0.6369571192963167, "grad_norm": 0.736404299736023, "learning_rate": 9.228533214227894e-06, "loss": 0.0142, "step": 9269 }, { "epoch": 0.6370258383727323, "grad_norm": 0.5220001935958862, "learning_rate": 9.225451809378148e-06, "loss": 0.016, "step": 9270 }, { "epoch": 0.6370945574491479, "grad_norm": 0.5144782662391663, "learning_rate": 9.222370690587652e-06, "loss": 0.0159, "step": 9271 }, { "epoch": 0.6371632765255635, "grad_norm": 0.6121453642845154, "learning_rate": 9.219289858009037e-06, "loss": 0.0201, "step": 9272 }, { "epoch": 0.6372319956019791, "grad_norm": 0.386849045753479, "learning_rate": 9.216209311794918e-06, "loss": 0.019, "step": 9273 }, { "epoch": 0.6373007146783948, "grad_norm": 0.6664405465126038, "learning_rate": 9.213129052097896e-06, "loss": 0.0141, "step": 9274 }, { "epoch": 0.6373694337548104, "grad_norm": 0.227992981672287, "learning_rate": 9.210049079070565e-06, "loss": 0.0102, "step": 9275 }, { "epoch": 0.637438152831226, "grad_norm": 0.1651877462863922, "learning_rate": 9.206969392865494e-06, "loss": 0.0086, "step": 9276 }, { "epoch": 0.6375068719076415, "grad_norm": 0.4631311893463135, "learning_rate": 9.203889993635255e-06, "loss": 0.0141, "step": 9277 }, { "epoch": 0.6375755909840571, "grad_norm": 0.6532123684883118, "learning_rate": 9.200810881532387e-06, "loss": 0.0192, "step": 9278 }, { "epoch": 0.6376443100604728, "grad_norm": 0.3614610433578491, "learning_rate": 9.197732056709428e-06, "loss": 0.0201, "step": 9279 }, { "epoch": 0.6377130291368884, "grad_norm": 0.4238967299461365, "learning_rate": 9.194653519318892e-06, "loss": 0.0097, "step": 9280 }, { "epoch": 0.637781748213304, "grad_norm": 0.4565095901489258, "learning_rate": 9.191575269513285e-06, "loss": 0.0101, "step": 9281 }, { "epoch": 0.6378504672897196, "grad_norm": 0.16946303844451904, "learning_rate": 9.188497307445098e-06, "loss": 0.0075, "step": 9282 }, { "epoch": 0.6379191863661352, "grad_norm": 0.6371262073516846, "learning_rate": 9.185419633266806e-06, "loss": 0.0162, "step": 9283 }, { "epoch": 0.6379879054425509, "grad_norm": 0.4752219021320343, "learning_rate": 9.182342247130873e-06, "loss": 0.0176, "step": 9284 }, { "epoch": 0.6380566245189665, "grad_norm": 0.6887412071228027, "learning_rate": 9.179265149189744e-06, "loss": 0.0171, "step": 9285 }, { "epoch": 0.6381253435953821, "grad_norm": 0.2758811116218567, "learning_rate": 9.176188339595853e-06, "loss": 0.0124, "step": 9286 }, { "epoch": 0.6381940626717977, "grad_norm": 0.9652655720710754, "learning_rate": 9.173111818501617e-06, "loss": 0.021, "step": 9287 }, { "epoch": 0.6382627817482133, "grad_norm": 0.5956900715827942, "learning_rate": 9.170035586059438e-06, "loss": 0.016, "step": 9288 }, { "epoch": 0.638331500824629, "grad_norm": 0.2377857118844986, "learning_rate": 9.166959642421714e-06, "loss": 0.015, "step": 9289 }, { "epoch": 0.6384002199010446, "grad_norm": 0.44332975149154663, "learning_rate": 9.163883987740818e-06, "loss": 0.0114, "step": 9290 }, { "epoch": 0.6384689389774602, "grad_norm": 0.5021483302116394, "learning_rate": 9.16080862216911e-06, "loss": 0.013, "step": 9291 }, { "epoch": 0.6385376580538757, "grad_norm": 1.010164499282837, "learning_rate": 9.15773354585894e-06, "loss": 0.0105, "step": 9292 }, { "epoch": 0.6386063771302913, "grad_norm": 0.3767792880535126, "learning_rate": 9.154658758962638e-06, "loss": 0.0172, "step": 9293 }, { "epoch": 0.6386750962067069, "grad_norm": 0.791232705116272, "learning_rate": 9.151584261632516e-06, "loss": 0.0156, "step": 9294 }, { "epoch": 0.6387438152831226, "grad_norm": 0.39705798029899597, "learning_rate": 9.148510054020895e-06, "loss": 0.0129, "step": 9295 }, { "epoch": 0.6388125343595382, "grad_norm": 0.2831593155860901, "learning_rate": 9.145436136280053e-06, "loss": 0.0119, "step": 9296 }, { "epoch": 0.6388812534359538, "grad_norm": 0.5533684492111206, "learning_rate": 9.142362508562268e-06, "loss": 0.0106, "step": 9297 }, { "epoch": 0.6389499725123694, "grad_norm": 0.2717392146587372, "learning_rate": 9.139289171019802e-06, "loss": 0.0109, "step": 9298 }, { "epoch": 0.639018691588785, "grad_norm": 0.21971853077411652, "learning_rate": 9.136216123804903e-06, "loss": 0.0115, "step": 9299 }, { "epoch": 0.6390874106652007, "grad_norm": 0.7453992962837219, "learning_rate": 9.133143367069798e-06, "loss": 0.0121, "step": 9300 }, { "epoch": 0.6391561297416163, "grad_norm": 1.1270394325256348, "learning_rate": 9.130070900966712e-06, "loss": 0.0183, "step": 9301 }, { "epoch": 0.6392248488180319, "grad_norm": 8.02987003326416, "learning_rate": 9.126998725647846e-06, "loss": 0.0115, "step": 9302 }, { "epoch": 0.6392935678944475, "grad_norm": 0.43354761600494385, "learning_rate": 9.123926841265388e-06, "loss": 0.0165, "step": 9303 }, { "epoch": 0.6393622869708631, "grad_norm": 0.5206308364868164, "learning_rate": 9.120855247971512e-06, "loss": 0.015, "step": 9304 }, { "epoch": 0.6394310060472788, "grad_norm": 0.5057122707366943, "learning_rate": 9.117783945918383e-06, "loss": 0.0168, "step": 9305 }, { "epoch": 0.6394997251236944, "grad_norm": 0.6135451197624207, "learning_rate": 9.114712935258134e-06, "loss": 0.0156, "step": 9306 }, { "epoch": 0.6395684442001099, "grad_norm": 0.4923582375049591, "learning_rate": 9.111642216142917e-06, "loss": 0.0109, "step": 9307 }, { "epoch": 0.6396371632765255, "grad_norm": 0.1829913854598999, "learning_rate": 9.108571788724838e-06, "loss": 0.0119, "step": 9308 }, { "epoch": 0.6397058823529411, "grad_norm": 0.3055824339389801, "learning_rate": 9.105501653155998e-06, "loss": 0.0224, "step": 9309 }, { "epoch": 0.6397746014293568, "grad_norm": 0.4432598352432251, "learning_rate": 9.10243180958849e-06, "loss": 0.014, "step": 9310 }, { "epoch": 0.6398433205057724, "grad_norm": 0.673116147518158, "learning_rate": 9.099362258174383e-06, "loss": 0.0111, "step": 9311 }, { "epoch": 0.639912039582188, "grad_norm": 0.5345456600189209, "learning_rate": 9.096292999065732e-06, "loss": 0.0161, "step": 9312 }, { "epoch": 0.6399807586586036, "grad_norm": 0.6965786814689636, "learning_rate": 9.093224032414596e-06, "loss": 0.0144, "step": 9313 }, { "epoch": 0.6400494777350192, "grad_norm": 0.46413668990135193, "learning_rate": 9.090155358372998e-06, "loss": 0.0136, "step": 9314 }, { "epoch": 0.6401181968114349, "grad_norm": 0.2417011857032776, "learning_rate": 9.087086977092954e-06, "loss": 0.0182, "step": 9315 }, { "epoch": 0.6401869158878505, "grad_norm": 0.6867009401321411, "learning_rate": 9.084018888726463e-06, "loss": 0.0165, "step": 9316 }, { "epoch": 0.6402556349642661, "grad_norm": 0.5346134305000305, "learning_rate": 9.080951093425512e-06, "loss": 0.0167, "step": 9317 }, { "epoch": 0.6403243540406817, "grad_norm": 0.5374290347099304, "learning_rate": 9.077883591342074e-06, "loss": 0.0189, "step": 9318 }, { "epoch": 0.6403930731170973, "grad_norm": 0.520128607749939, "learning_rate": 9.074816382628112e-06, "loss": 0.0099, "step": 9319 }, { "epoch": 0.640461792193513, "grad_norm": 0.2740228772163391, "learning_rate": 9.071749467435563e-06, "loss": 0.0119, "step": 9320 }, { "epoch": 0.6405305112699285, "grad_norm": 0.6475782990455627, "learning_rate": 9.068682845916357e-06, "loss": 0.0152, "step": 9321 }, { "epoch": 0.6405992303463441, "grad_norm": 0.25596117973327637, "learning_rate": 9.065616518222408e-06, "loss": 0.0184, "step": 9322 }, { "epoch": 0.6406679494227597, "grad_norm": 0.47025266289711, "learning_rate": 9.062550484505617e-06, "loss": 0.0115, "step": 9323 }, { "epoch": 0.6407366684991753, "grad_norm": 0.37268468737602234, "learning_rate": 9.059484744917864e-06, "loss": 0.0104, "step": 9324 }, { "epoch": 0.640805387575591, "grad_norm": 0.5497572422027588, "learning_rate": 9.056419299611028e-06, "loss": 0.0111, "step": 9325 }, { "epoch": 0.6408741066520066, "grad_norm": 0.4796712398529053, "learning_rate": 9.053354148736959e-06, "loss": 0.019, "step": 9326 }, { "epoch": 0.6409428257284222, "grad_norm": 0.6119176149368286, "learning_rate": 9.0502892924475e-06, "loss": 0.0151, "step": 9327 }, { "epoch": 0.6410115448048378, "grad_norm": 0.3976374864578247, "learning_rate": 9.04722473089448e-06, "loss": 0.0088, "step": 9328 }, { "epoch": 0.6410802638812534, "grad_norm": 0.637499988079071, "learning_rate": 9.044160464229706e-06, "loss": 0.0107, "step": 9329 }, { "epoch": 0.6411489829576691, "grad_norm": 0.34113579988479614, "learning_rate": 9.041096492604973e-06, "loss": 0.0195, "step": 9330 }, { "epoch": 0.6412177020340847, "grad_norm": 0.24441049993038177, "learning_rate": 9.038032816172074e-06, "loss": 0.0225, "step": 9331 }, { "epoch": 0.6412864211105003, "grad_norm": 0.30501788854599, "learning_rate": 9.034969435082774e-06, "loss": 0.014, "step": 9332 }, { "epoch": 0.6413551401869159, "grad_norm": 0.32471963763237, "learning_rate": 9.031906349488822e-06, "loss": 0.0134, "step": 9333 }, { "epoch": 0.6414238592633315, "grad_norm": 0.7124269008636475, "learning_rate": 9.02884355954196e-06, "loss": 0.0144, "step": 9334 }, { "epoch": 0.6414925783397472, "grad_norm": 0.5717567205429077, "learning_rate": 9.025781065393912e-06, "loss": 0.0172, "step": 9335 }, { "epoch": 0.6415612974161627, "grad_norm": 0.6498004794120789, "learning_rate": 9.022718867196385e-06, "loss": 0.0209, "step": 9336 }, { "epoch": 0.6416300164925783, "grad_norm": 0.4577302932739258, "learning_rate": 9.019656965101079e-06, "loss": 0.0234, "step": 9337 }, { "epoch": 0.6416987355689939, "grad_norm": 0.6422202587127686, "learning_rate": 9.016595359259674e-06, "loss": 0.0127, "step": 9338 }, { "epoch": 0.6417674546454095, "grad_norm": 0.36209991574287415, "learning_rate": 9.01353404982383e-06, "loss": 0.0144, "step": 9339 }, { "epoch": 0.6418361737218252, "grad_norm": 0.35182029008865356, "learning_rate": 9.010473036945204e-06, "loss": 0.0088, "step": 9340 }, { "epoch": 0.6419048927982408, "grad_norm": 0.5687143206596375, "learning_rate": 9.00741232077543e-06, "loss": 0.0109, "step": 9341 }, { "epoch": 0.6419736118746564, "grad_norm": 0.819861114025116, "learning_rate": 9.004351901466123e-06, "loss": 0.0154, "step": 9342 }, { "epoch": 0.642042330951072, "grad_norm": 0.6616982221603394, "learning_rate": 9.001291779168903e-06, "loss": 0.0126, "step": 9343 }, { "epoch": 0.6421110500274876, "grad_norm": 0.5349470376968384, "learning_rate": 8.998231954035354e-06, "loss": 0.014, "step": 9344 }, { "epoch": 0.6421797691039033, "grad_norm": 0.5061554312705994, "learning_rate": 8.995172426217055e-06, "loss": 0.0171, "step": 9345 }, { "epoch": 0.6422484881803189, "grad_norm": 0.6344165802001953, "learning_rate": 8.99211319586557e-06, "loss": 0.0075, "step": 9346 }, { "epoch": 0.6423172072567345, "grad_norm": 0.5578762888908386, "learning_rate": 8.989054263132443e-06, "loss": 0.0171, "step": 9347 }, { "epoch": 0.6423859263331501, "grad_norm": 0.5171419382095337, "learning_rate": 8.985995628169207e-06, "loss": 0.0119, "step": 9348 }, { "epoch": 0.6424546454095657, "grad_norm": 0.4395202398300171, "learning_rate": 8.982937291127385e-06, "loss": 0.021, "step": 9349 }, { "epoch": 0.6425233644859814, "grad_norm": 0.26011401414871216, "learning_rate": 8.979879252158482e-06, "loss": 0.0101, "step": 9350 }, { "epoch": 0.6425920835623969, "grad_norm": 1.3176831007003784, "learning_rate": 8.976821511413985e-06, "loss": 0.0155, "step": 9351 }, { "epoch": 0.6426608026388125, "grad_norm": 0.34138691425323486, "learning_rate": 8.973764069045364e-06, "loss": 0.0092, "step": 9352 }, { "epoch": 0.6427295217152281, "grad_norm": 1.5187586545944214, "learning_rate": 8.970706925204085e-06, "loss": 0.0206, "step": 9353 }, { "epoch": 0.6427982407916437, "grad_norm": 0.7413549423217773, "learning_rate": 8.967650080041585e-06, "loss": 0.0232, "step": 9354 }, { "epoch": 0.6428669598680594, "grad_norm": 0.5823287963867188, "learning_rate": 8.9645935337093e-06, "loss": 0.0198, "step": 9355 }, { "epoch": 0.642935678944475, "grad_norm": 0.5897897481918335, "learning_rate": 8.961537286358643e-06, "loss": 0.0114, "step": 9356 }, { "epoch": 0.6430043980208906, "grad_norm": 0.526878833770752, "learning_rate": 8.958481338141016e-06, "loss": 0.016, "step": 9357 }, { "epoch": 0.6430731170973062, "grad_norm": 0.28796470165252686, "learning_rate": 8.955425689207801e-06, "loss": 0.0121, "step": 9358 }, { "epoch": 0.6431418361737218, "grad_norm": 0.5503102540969849, "learning_rate": 8.952370339710372e-06, "loss": 0.0166, "step": 9359 }, { "epoch": 0.6432105552501375, "grad_norm": 0.3563409149646759, "learning_rate": 8.949315289800075e-06, "loss": 0.0148, "step": 9360 }, { "epoch": 0.6432792743265531, "grad_norm": 0.4651847183704376, "learning_rate": 8.946260539628267e-06, "loss": 0.0094, "step": 9361 }, { "epoch": 0.6433479934029687, "grad_norm": 0.6657014489173889, "learning_rate": 8.943206089346264e-06, "loss": 0.0198, "step": 9362 }, { "epoch": 0.6434167124793843, "grad_norm": 0.6962487101554871, "learning_rate": 8.94015193910538e-06, "loss": 0.0195, "step": 9363 }, { "epoch": 0.6434854315557998, "grad_norm": 0.8595312833786011, "learning_rate": 8.937098089056911e-06, "loss": 0.0163, "step": 9364 }, { "epoch": 0.6435541506322155, "grad_norm": 0.34084445238113403, "learning_rate": 8.934044539352137e-06, "loss": 0.0143, "step": 9365 }, { "epoch": 0.6436228697086311, "grad_norm": 0.6504918932914734, "learning_rate": 8.930991290142319e-06, "loss": 0.0116, "step": 9366 }, { "epoch": 0.6436915887850467, "grad_norm": 0.48625293374061584, "learning_rate": 8.927938341578721e-06, "loss": 0.0106, "step": 9367 }, { "epoch": 0.6437603078614623, "grad_norm": 0.6483991146087646, "learning_rate": 8.924885693812575e-06, "loss": 0.0214, "step": 9368 }, { "epoch": 0.6438290269378779, "grad_norm": 0.3994959592819214, "learning_rate": 8.9218333469951e-06, "loss": 0.0157, "step": 9369 }, { "epoch": 0.6438977460142936, "grad_norm": 0.2696358263492584, "learning_rate": 8.918781301277508e-06, "loss": 0.011, "step": 9370 }, { "epoch": 0.6439664650907092, "grad_norm": 0.41077926754951477, "learning_rate": 8.915729556810987e-06, "loss": 0.0076, "step": 9371 }, { "epoch": 0.6440351841671248, "grad_norm": 0.47686251997947693, "learning_rate": 8.91267811374671e-06, "loss": 0.0181, "step": 9372 }, { "epoch": 0.6441039032435404, "grad_norm": 0.4964199960231781, "learning_rate": 8.909626972235848e-06, "loss": 0.0125, "step": 9373 }, { "epoch": 0.644172622319956, "grad_norm": 0.22725407779216766, "learning_rate": 8.906576132429545e-06, "loss": 0.008, "step": 9374 }, { "epoch": 0.6442413413963717, "grad_norm": 0.9921162128448486, "learning_rate": 8.903525594478933e-06, "loss": 0.0119, "step": 9375 }, { "epoch": 0.6443100604727873, "grad_norm": 0.8446808457374573, "learning_rate": 8.90047535853513e-06, "loss": 0.0122, "step": 9376 }, { "epoch": 0.6443787795492029, "grad_norm": 0.7176409363746643, "learning_rate": 8.897425424749236e-06, "loss": 0.0117, "step": 9377 }, { "epoch": 0.6444474986256185, "grad_norm": 0.7093494534492493, "learning_rate": 8.894375793272336e-06, "loss": 0.0128, "step": 9378 }, { "epoch": 0.644516217702034, "grad_norm": 0.1603967547416687, "learning_rate": 8.89132646425551e-06, "loss": 0.0123, "step": 9379 }, { "epoch": 0.6445849367784497, "grad_norm": 0.45922935009002686, "learning_rate": 8.888277437849814e-06, "loss": 0.0175, "step": 9380 }, { "epoch": 0.6446536558548653, "grad_norm": 0.197319895029068, "learning_rate": 8.885228714206288e-06, "loss": 0.0094, "step": 9381 }, { "epoch": 0.6447223749312809, "grad_norm": 0.21577569842338562, "learning_rate": 8.88218029347596e-06, "loss": 0.01, "step": 9382 }, { "epoch": 0.6447910940076965, "grad_norm": 0.5731900334358215, "learning_rate": 8.879132175809842e-06, "loss": 0.0178, "step": 9383 }, { "epoch": 0.6448598130841121, "grad_norm": 0.8384027481079102, "learning_rate": 8.876084361358927e-06, "loss": 0.0196, "step": 9384 }, { "epoch": 0.6449285321605278, "grad_norm": 0.26766523718833923, "learning_rate": 8.873036850274208e-06, "loss": 0.0205, "step": 9385 }, { "epoch": 0.6449972512369434, "grad_norm": 1.2356467247009277, "learning_rate": 8.869989642706647e-06, "loss": 0.0191, "step": 9386 }, { "epoch": 0.645065970313359, "grad_norm": 1.2615141868591309, "learning_rate": 8.866942738807195e-06, "loss": 0.0215, "step": 9387 }, { "epoch": 0.6451346893897746, "grad_norm": 0.19037188589572906, "learning_rate": 8.863896138726793e-06, "loss": 0.0075, "step": 9388 }, { "epoch": 0.6452034084661902, "grad_norm": 0.35372409224510193, "learning_rate": 8.86084984261636e-06, "loss": 0.0113, "step": 9389 }, { "epoch": 0.6452721275426059, "grad_norm": 0.42356565594673157, "learning_rate": 8.857803850626801e-06, "loss": 0.0154, "step": 9390 }, { "epoch": 0.6453408466190215, "grad_norm": 0.3313135504722595, "learning_rate": 8.854758162909015e-06, "loss": 0.0121, "step": 9391 }, { "epoch": 0.6454095656954371, "grad_norm": 0.7552188038825989, "learning_rate": 8.851712779613876e-06, "loss": 0.0149, "step": 9392 }, { "epoch": 0.6454782847718527, "grad_norm": 0.26288044452667236, "learning_rate": 8.848667700892244e-06, "loss": 0.009, "step": 9393 }, { "epoch": 0.6455470038482682, "grad_norm": 0.8026725053787231, "learning_rate": 8.845622926894968e-06, "loss": 0.01, "step": 9394 }, { "epoch": 0.6456157229246839, "grad_norm": 0.44310829043388367, "learning_rate": 8.84257845777288e-06, "loss": 0.0126, "step": 9395 }, { "epoch": 0.6456844420010995, "grad_norm": 0.36756661534309387, "learning_rate": 8.83953429367679e-06, "loss": 0.0116, "step": 9396 }, { "epoch": 0.6457531610775151, "grad_norm": 0.578450083732605, "learning_rate": 8.83649043475751e-06, "loss": 0.0119, "step": 9397 }, { "epoch": 0.6458218801539307, "grad_norm": 0.37221479415893555, "learning_rate": 8.833446881165825e-06, "loss": 0.0174, "step": 9398 }, { "epoch": 0.6458905992303463, "grad_norm": 0.9895715117454529, "learning_rate": 8.830403633052502e-06, "loss": 0.0274, "step": 9399 }, { "epoch": 0.645959318306762, "grad_norm": 1.6410248279571533, "learning_rate": 8.8273606905683e-06, "loss": 0.0294, "step": 9400 }, { "epoch": 0.6460280373831776, "grad_norm": 0.5106614828109741, "learning_rate": 8.824318053863957e-06, "loss": 0.0169, "step": 9401 }, { "epoch": 0.6460967564595932, "grad_norm": 0.4808150827884674, "learning_rate": 8.821275723090202e-06, "loss": 0.0158, "step": 9402 }, { "epoch": 0.6461654755360088, "grad_norm": 0.5627532005310059, "learning_rate": 8.818233698397738e-06, "loss": 0.015, "step": 9403 }, { "epoch": 0.6462341946124244, "grad_norm": 0.3708231747150421, "learning_rate": 8.815191979937276e-06, "loss": 0.0151, "step": 9404 }, { "epoch": 0.6463029136888401, "grad_norm": 0.45038262009620667, "learning_rate": 8.812150567859485e-06, "loss": 0.0116, "step": 9405 }, { "epoch": 0.6463716327652557, "grad_norm": 0.8536729216575623, "learning_rate": 8.809109462315034e-06, "loss": 0.0164, "step": 9406 }, { "epoch": 0.6464403518416713, "grad_norm": 0.33176544308662415, "learning_rate": 8.806068663454573e-06, "loss": 0.0081, "step": 9407 }, { "epoch": 0.6465090709180868, "grad_norm": 0.47850146889686584, "learning_rate": 8.803028171428737e-06, "loss": 0.0201, "step": 9408 }, { "epoch": 0.6465777899945024, "grad_norm": 0.23216663300991058, "learning_rate": 8.799987986388139e-06, "loss": 0.0116, "step": 9409 }, { "epoch": 0.6466465090709181, "grad_norm": 0.5859963893890381, "learning_rate": 8.796948108483396e-06, "loss": 0.0093, "step": 9410 }, { "epoch": 0.6467152281473337, "grad_norm": 0.8058140277862549, "learning_rate": 8.793908537865089e-06, "loss": 0.0188, "step": 9411 }, { "epoch": 0.6467839472237493, "grad_norm": 0.3067159652709961, "learning_rate": 8.790869274683792e-06, "loss": 0.0082, "step": 9412 }, { "epoch": 0.6468526663001649, "grad_norm": 0.6426392197608948, "learning_rate": 8.787830319090069e-06, "loss": 0.012, "step": 9413 }, { "epoch": 0.6469213853765805, "grad_norm": 0.5284608602523804, "learning_rate": 8.784791671234459e-06, "loss": 0.0102, "step": 9414 }, { "epoch": 0.6469901044529962, "grad_norm": 0.33928653597831726, "learning_rate": 8.781753331267485e-06, "loss": 0.0117, "step": 9415 }, { "epoch": 0.6470588235294118, "grad_norm": 0.9717537760734558, "learning_rate": 8.778715299339672e-06, "loss": 0.0171, "step": 9416 }, { "epoch": 0.6471275426058274, "grad_norm": 0.46539169549942017, "learning_rate": 8.775677575601516e-06, "loss": 0.0139, "step": 9417 }, { "epoch": 0.647196261682243, "grad_norm": 0.7752729654312134, "learning_rate": 8.772640160203493e-06, "loss": 0.0094, "step": 9418 }, { "epoch": 0.6472649807586586, "grad_norm": 0.38527965545654297, "learning_rate": 8.769603053296075e-06, "loss": 0.0089, "step": 9419 }, { "epoch": 0.6473336998350742, "grad_norm": 0.6359352469444275, "learning_rate": 8.766566255029713e-06, "loss": 0.0102, "step": 9420 }, { "epoch": 0.6474024189114899, "grad_norm": 0.25979286432266235, "learning_rate": 8.763529765554835e-06, "loss": 0.014, "step": 9421 }, { "epoch": 0.6474711379879055, "grad_norm": 0.5908741354942322, "learning_rate": 8.76049358502188e-06, "loss": 0.0237, "step": 9422 }, { "epoch": 0.647539857064321, "grad_norm": 0.6947521567344666, "learning_rate": 8.757457713581246e-06, "loss": 0.0203, "step": 9423 }, { "epoch": 0.6476085761407366, "grad_norm": 0.23411449790000916, "learning_rate": 8.754422151383318e-06, "loss": 0.0199, "step": 9424 }, { "epoch": 0.6476772952171522, "grad_norm": 0.42483291029930115, "learning_rate": 8.75138689857848e-06, "loss": 0.0105, "step": 9425 }, { "epoch": 0.6477460142935679, "grad_norm": 0.36372265219688416, "learning_rate": 8.748351955317084e-06, "loss": 0.0125, "step": 9426 }, { "epoch": 0.6478147333699835, "grad_norm": 0.487432599067688, "learning_rate": 8.74531732174948e-06, "loss": 0.0178, "step": 9427 }, { "epoch": 0.6478834524463991, "grad_norm": 0.7940252423286438, "learning_rate": 8.742282998025996e-06, "loss": 0.0141, "step": 9428 }, { "epoch": 0.6479521715228147, "grad_norm": 0.37213751673698425, "learning_rate": 8.739248984296953e-06, "loss": 0.0105, "step": 9429 }, { "epoch": 0.6480208905992303, "grad_norm": 0.3444279432296753, "learning_rate": 8.736215280712637e-06, "loss": 0.0083, "step": 9430 }, { "epoch": 0.648089609675646, "grad_norm": 0.25231537222862244, "learning_rate": 8.733181887423343e-06, "loss": 0.0081, "step": 9431 }, { "epoch": 0.6481583287520616, "grad_norm": 0.4069301187992096, "learning_rate": 8.730148804579332e-06, "loss": 0.0121, "step": 9432 }, { "epoch": 0.6482270478284772, "grad_norm": 0.8692221641540527, "learning_rate": 8.727116032330857e-06, "loss": 0.0221, "step": 9433 }, { "epoch": 0.6482957669048928, "grad_norm": 0.29412856698036194, "learning_rate": 8.724083570828156e-06, "loss": 0.0159, "step": 9434 }, { "epoch": 0.6483644859813084, "grad_norm": 0.49970030784606934, "learning_rate": 8.721051420221459e-06, "loss": 0.0156, "step": 9435 }, { "epoch": 0.6484332050577241, "grad_norm": 0.6284779906272888, "learning_rate": 8.718019580660962e-06, "loss": 0.0135, "step": 9436 }, { "epoch": 0.6485019241341397, "grad_norm": 1.1197161674499512, "learning_rate": 8.714988052296862e-06, "loss": 0.0175, "step": 9437 }, { "epoch": 0.6485706432105552, "grad_norm": 0.2500998079776764, "learning_rate": 8.71195683527933e-06, "loss": 0.0126, "step": 9438 }, { "epoch": 0.6486393622869708, "grad_norm": 0.6021116375923157, "learning_rate": 8.708925929758527e-06, "loss": 0.0125, "step": 9439 }, { "epoch": 0.6487080813633864, "grad_norm": 0.44485539197921753, "learning_rate": 8.705895335884605e-06, "loss": 0.0198, "step": 9440 }, { "epoch": 0.6487768004398021, "grad_norm": 0.09226354956626892, "learning_rate": 8.702865053807684e-06, "loss": 0.0192, "step": 9441 }, { "epoch": 0.6488455195162177, "grad_norm": 1.304926872253418, "learning_rate": 8.699835083677887e-06, "loss": 0.0121, "step": 9442 }, { "epoch": 0.6489142385926333, "grad_norm": 0.5650606155395508, "learning_rate": 8.6968054256453e-06, "loss": 0.0103, "step": 9443 }, { "epoch": 0.6489829576690489, "grad_norm": 0.7488411068916321, "learning_rate": 8.69377607986002e-06, "loss": 0.0127, "step": 9444 }, { "epoch": 0.6490516767454645, "grad_norm": 0.45211929082870483, "learning_rate": 8.690747046472095e-06, "loss": 0.0146, "step": 9445 }, { "epoch": 0.6491203958218802, "grad_norm": 0.8572999835014343, "learning_rate": 8.687718325631604e-06, "loss": 0.0183, "step": 9446 }, { "epoch": 0.6491891148982958, "grad_norm": 0.3605786859989166, "learning_rate": 8.684689917488562e-06, "loss": 0.0211, "step": 9447 }, { "epoch": 0.6492578339747114, "grad_norm": 0.6537038683891296, "learning_rate": 8.681661822193005e-06, "loss": 0.0179, "step": 9448 }, { "epoch": 0.649326553051127, "grad_norm": 0.21106821298599243, "learning_rate": 8.678634039894925e-06, "loss": 0.007, "step": 9449 }, { "epoch": 0.6493952721275426, "grad_norm": 0.5985227823257446, "learning_rate": 8.675606570744325e-06, "loss": 0.0146, "step": 9450 }, { "epoch": 0.6494639912039583, "grad_norm": 0.5243404507637024, "learning_rate": 8.67257941489116e-06, "loss": 0.0156, "step": 9451 }, { "epoch": 0.6495327102803738, "grad_norm": 0.5572054386138916, "learning_rate": 8.669552572485416e-06, "loss": 0.0108, "step": 9452 }, { "epoch": 0.6496014293567894, "grad_norm": 0.654907763004303, "learning_rate": 8.666526043677015e-06, "loss": 0.0184, "step": 9453 }, { "epoch": 0.649670148433205, "grad_norm": 0.7845885157585144, "learning_rate": 8.663499828615898e-06, "loss": 0.0184, "step": 9454 }, { "epoch": 0.6497388675096206, "grad_norm": 0.22129184007644653, "learning_rate": 8.660473927451969e-06, "loss": 0.0096, "step": 9455 }, { "epoch": 0.6498075865860363, "grad_norm": 0.401904821395874, "learning_rate": 8.657448340335132e-06, "loss": 0.0105, "step": 9456 }, { "epoch": 0.6498763056624519, "grad_norm": 0.2819434106349945, "learning_rate": 8.654423067415258e-06, "loss": 0.0054, "step": 9457 }, { "epoch": 0.6499450247388675, "grad_norm": 1.2483668327331543, "learning_rate": 8.65139810884222e-06, "loss": 0.0111, "step": 9458 }, { "epoch": 0.6500137438152831, "grad_norm": 0.5551494359970093, "learning_rate": 8.648373464765875e-06, "loss": 0.0068, "step": 9459 }, { "epoch": 0.6500824628916987, "grad_norm": 0.6339166164398193, "learning_rate": 8.645349135336042e-06, "loss": 0.0149, "step": 9460 }, { "epoch": 0.6501511819681144, "grad_norm": 0.3794237971305847, "learning_rate": 8.642325120702554e-06, "loss": 0.0146, "step": 9461 }, { "epoch": 0.65021990104453, "grad_norm": 0.38380488753318787, "learning_rate": 8.639301421015206e-06, "loss": 0.0081, "step": 9462 }, { "epoch": 0.6502886201209456, "grad_norm": 0.2827836573123932, "learning_rate": 8.636278036423786e-06, "loss": 0.0144, "step": 9463 }, { "epoch": 0.6503573391973612, "grad_norm": 0.2726637125015259, "learning_rate": 8.63325496707807e-06, "loss": 0.009, "step": 9464 }, { "epoch": 0.6504260582737768, "grad_norm": 0.47354555130004883, "learning_rate": 8.630232213127818e-06, "loss": 0.0135, "step": 9465 }, { "epoch": 0.6504947773501925, "grad_norm": 0.6486855745315552, "learning_rate": 8.627209774722762e-06, "loss": 0.0078, "step": 9466 }, { "epoch": 0.650563496426608, "grad_norm": 0.2273540198802948, "learning_rate": 8.624187652012634e-06, "loss": 0.0067, "step": 9467 }, { "epoch": 0.6506322155030236, "grad_norm": 0.5067611336708069, "learning_rate": 8.621165845147141e-06, "loss": 0.0148, "step": 9468 }, { "epoch": 0.6507009345794392, "grad_norm": 1.2812354564666748, "learning_rate": 8.618144354275975e-06, "loss": 0.0172, "step": 9469 }, { "epoch": 0.6507696536558548, "grad_norm": 0.47021764516830444, "learning_rate": 8.615123179548817e-06, "loss": 0.0113, "step": 9470 }, { "epoch": 0.6508383727322705, "grad_norm": 0.5746809244155884, "learning_rate": 8.612102321115336e-06, "loss": 0.0133, "step": 9471 }, { "epoch": 0.6509070918086861, "grad_norm": 0.21026362478733063, "learning_rate": 8.609081779125167e-06, "loss": 0.0143, "step": 9472 }, { "epoch": 0.6509758108851017, "grad_norm": 0.8082496523857117, "learning_rate": 8.606061553727954e-06, "loss": 0.0185, "step": 9473 }, { "epoch": 0.6510445299615173, "grad_norm": 0.4083186388015747, "learning_rate": 8.603041645073302e-06, "loss": 0.0093, "step": 9474 }, { "epoch": 0.6511132490379329, "grad_norm": 0.41984620690345764, "learning_rate": 8.600022053310813e-06, "loss": 0.0141, "step": 9475 }, { "epoch": 0.6511819681143486, "grad_norm": 0.36499685049057007, "learning_rate": 8.597002778590081e-06, "loss": 0.0106, "step": 9476 }, { "epoch": 0.6512506871907642, "grad_norm": 0.40912869572639465, "learning_rate": 8.593983821060658e-06, "loss": 0.0119, "step": 9477 }, { "epoch": 0.6513194062671798, "grad_norm": 0.611579418182373, "learning_rate": 8.590965180872116e-06, "loss": 0.01, "step": 9478 }, { "epoch": 0.6513881253435954, "grad_norm": 0.2280515432357788, "learning_rate": 8.587946858173975e-06, "loss": 0.009, "step": 9479 }, { "epoch": 0.651456844420011, "grad_norm": 0.6289471983909607, "learning_rate": 8.58492885311577e-06, "loss": 0.0108, "step": 9480 }, { "epoch": 0.6515255634964267, "grad_norm": 0.5107221007347107, "learning_rate": 8.581911165846991e-06, "loss": 0.0155, "step": 9481 }, { "epoch": 0.6515942825728422, "grad_norm": 0.5046976208686829, "learning_rate": 8.578893796517147e-06, "loss": 0.0101, "step": 9482 }, { "epoch": 0.6516630016492578, "grad_norm": 0.47406676411628723, "learning_rate": 8.575876745275698e-06, "loss": 0.0191, "step": 9483 }, { "epoch": 0.6517317207256734, "grad_norm": 0.2348344624042511, "learning_rate": 8.572860012272113e-06, "loss": 0.0075, "step": 9484 }, { "epoch": 0.651800439802089, "grad_norm": 0.6156405210494995, "learning_rate": 8.569843597655824e-06, "loss": 0.0112, "step": 9485 }, { "epoch": 0.6518691588785047, "grad_norm": 0.4082467555999756, "learning_rate": 8.566827501576272e-06, "loss": 0.0107, "step": 9486 }, { "epoch": 0.6519378779549203, "grad_norm": 0.39723554253578186, "learning_rate": 8.563811724182842e-06, "loss": 0.0152, "step": 9487 }, { "epoch": 0.6520065970313359, "grad_norm": 0.2976984977722168, "learning_rate": 8.560796265624964e-06, "loss": 0.0206, "step": 9488 }, { "epoch": 0.6520753161077515, "grad_norm": 0.665152907371521, "learning_rate": 8.557781126051992e-06, "loss": 0.0102, "step": 9489 }, { "epoch": 0.6521440351841671, "grad_norm": 1.5709497928619385, "learning_rate": 8.554766305613304e-06, "loss": 0.0147, "step": 9490 }, { "epoch": 0.6522127542605828, "grad_norm": 0.4129549562931061, "learning_rate": 8.551751804458235e-06, "loss": 0.022, "step": 9491 }, { "epoch": 0.6522814733369984, "grad_norm": 0.30758821964263916, "learning_rate": 8.548737622736134e-06, "loss": 0.0118, "step": 9492 }, { "epoch": 0.652350192413414, "grad_norm": 0.3240562677383423, "learning_rate": 8.5457237605963e-06, "loss": 0.0079, "step": 9493 }, { "epoch": 0.6524189114898296, "grad_norm": 0.49052202701568604, "learning_rate": 8.542710218188043e-06, "loss": 0.0178, "step": 9494 }, { "epoch": 0.6524876305662451, "grad_norm": 0.4843105673789978, "learning_rate": 8.539696995660651e-06, "loss": 0.0071, "step": 9495 }, { "epoch": 0.6525563496426608, "grad_norm": 0.8115957975387573, "learning_rate": 8.536684093163382e-06, "loss": 0.0184, "step": 9496 }, { "epoch": 0.6526250687190764, "grad_norm": 0.3076939582824707, "learning_rate": 8.533671510845503e-06, "loss": 0.0185, "step": 9497 }, { "epoch": 0.652693787795492, "grad_norm": 0.922324001789093, "learning_rate": 8.530659248856236e-06, "loss": 0.0096, "step": 9498 }, { "epoch": 0.6527625068719076, "grad_norm": 0.4442616105079651, "learning_rate": 8.527647307344805e-06, "loss": 0.0155, "step": 9499 }, { "epoch": 0.6528312259483232, "grad_norm": 0.49495214223861694, "learning_rate": 8.524635686460424e-06, "loss": 0.0174, "step": 9500 }, { "epoch": 0.6528999450247389, "grad_norm": 0.6881579756736755, "learning_rate": 8.521624386352285e-06, "loss": 0.014, "step": 9501 }, { "epoch": 0.6529686641011545, "grad_norm": 0.7011836171150208, "learning_rate": 8.518613407169548e-06, "loss": 0.0259, "step": 9502 }, { "epoch": 0.6530373831775701, "grad_norm": 0.2957806885242462, "learning_rate": 8.515602749061384e-06, "loss": 0.0096, "step": 9503 }, { "epoch": 0.6531061022539857, "grad_norm": 0.2731021046638489, "learning_rate": 8.512592412176922e-06, "loss": 0.0183, "step": 9504 }, { "epoch": 0.6531748213304013, "grad_norm": 0.19892385601997375, "learning_rate": 8.509582396665293e-06, "loss": 0.0092, "step": 9505 }, { "epoch": 0.653243540406817, "grad_norm": 0.959913969039917, "learning_rate": 8.50657270267561e-06, "loss": 0.0164, "step": 9506 }, { "epoch": 0.6533122594832326, "grad_norm": 0.7006710767745972, "learning_rate": 8.503563330356971e-06, "loss": 0.0147, "step": 9507 }, { "epoch": 0.6533809785596482, "grad_norm": 0.3545069694519043, "learning_rate": 8.500554279858444e-06, "loss": 0.0089, "step": 9508 }, { "epoch": 0.6534496976360638, "grad_norm": 0.8509413599967957, "learning_rate": 8.4975455513291e-06, "loss": 0.0213, "step": 9509 }, { "epoch": 0.6535184167124793, "grad_norm": 0.7644885778427124, "learning_rate": 8.494537144917975e-06, "loss": 0.0178, "step": 9510 }, { "epoch": 0.653587135788895, "grad_norm": 1.3726799488067627, "learning_rate": 8.491529060774105e-06, "loss": 0.0301, "step": 9511 }, { "epoch": 0.6536558548653106, "grad_norm": 0.5733005404472351, "learning_rate": 8.488521299046511e-06, "loss": 0.0176, "step": 9512 }, { "epoch": 0.6537245739417262, "grad_norm": 0.3248387575149536, "learning_rate": 8.485513859884178e-06, "loss": 0.0211, "step": 9513 }, { "epoch": 0.6537932930181418, "grad_norm": 0.48673897981643677, "learning_rate": 8.482506743436102e-06, "loss": 0.0173, "step": 9514 }, { "epoch": 0.6538620120945574, "grad_norm": 0.5315321683883667, "learning_rate": 8.479499949851237e-06, "loss": 0.0133, "step": 9515 }, { "epoch": 0.6539307311709731, "grad_norm": 0.5014433860778809, "learning_rate": 8.476493479278543e-06, "loss": 0.0127, "step": 9516 }, { "epoch": 0.6539994502473887, "grad_norm": 0.3537399470806122, "learning_rate": 8.47348733186694e-06, "loss": 0.0108, "step": 9517 }, { "epoch": 0.6540681693238043, "grad_norm": 0.3790101110935211, "learning_rate": 8.47048150776537e-06, "loss": 0.0161, "step": 9518 }, { "epoch": 0.6541368884002199, "grad_norm": 0.3926980793476105, "learning_rate": 8.467476007122713e-06, "loss": 0.0141, "step": 9519 }, { "epoch": 0.6542056074766355, "grad_norm": 0.658176839351654, "learning_rate": 8.464470830087871e-06, "loss": 0.015, "step": 9520 }, { "epoch": 0.6542743265530512, "grad_norm": 0.21452859044075012, "learning_rate": 8.461465976809702e-06, "loss": 0.0139, "step": 9521 }, { "epoch": 0.6543430456294668, "grad_norm": 0.25654909014701843, "learning_rate": 8.458461447437072e-06, "loss": 0.0155, "step": 9522 }, { "epoch": 0.6544117647058824, "grad_norm": 0.6802899241447449, "learning_rate": 8.455457242118801e-06, "loss": 0.0173, "step": 9523 }, { "epoch": 0.654480483782298, "grad_norm": 0.4007052481174469, "learning_rate": 8.452453361003735e-06, "loss": 0.0139, "step": 9524 }, { "epoch": 0.6545492028587135, "grad_norm": 0.39553123712539673, "learning_rate": 8.449449804240663e-06, "loss": 0.0124, "step": 9525 }, { "epoch": 0.6546179219351292, "grad_norm": 0.6958163976669312, "learning_rate": 8.446446571978386e-06, "loss": 0.0271, "step": 9526 }, { "epoch": 0.6546866410115448, "grad_norm": 0.3100587725639343, "learning_rate": 8.443443664365666e-06, "loss": 0.0108, "step": 9527 }, { "epoch": 0.6547553600879604, "grad_norm": 0.3311315178871155, "learning_rate": 8.440441081551274e-06, "loss": 0.0118, "step": 9528 }, { "epoch": 0.654824079164376, "grad_norm": 0.6038370132446289, "learning_rate": 8.437438823683937e-06, "loss": 0.0189, "step": 9529 }, { "epoch": 0.6548927982407916, "grad_norm": 1.6755226850509644, "learning_rate": 8.43443689091239e-06, "loss": 0.0262, "step": 9530 }, { "epoch": 0.6549615173172073, "grad_norm": 0.5716504454612732, "learning_rate": 8.431435283385346e-06, "loss": 0.0153, "step": 9531 }, { "epoch": 0.6550302363936229, "grad_norm": 0.2660711109638214, "learning_rate": 8.428434001251488e-06, "loss": 0.0135, "step": 9532 }, { "epoch": 0.6550989554700385, "grad_norm": 0.47128480672836304, "learning_rate": 8.425433044659506e-06, "loss": 0.018, "step": 9533 }, { "epoch": 0.6551676745464541, "grad_norm": 1.1178773641586304, "learning_rate": 8.42243241375805e-06, "loss": 0.0239, "step": 9534 }, { "epoch": 0.6552363936228697, "grad_norm": 0.4178517162799835, "learning_rate": 8.419432108695767e-06, "loss": 0.0107, "step": 9535 }, { "epoch": 0.6553051126992854, "grad_norm": 0.27082711458206177, "learning_rate": 8.416432129621288e-06, "loss": 0.0228, "step": 9536 }, { "epoch": 0.655373831775701, "grad_norm": 0.4601750373840332, "learning_rate": 8.413432476683231e-06, "loss": 0.0127, "step": 9537 }, { "epoch": 0.6554425508521166, "grad_norm": 0.517726480960846, "learning_rate": 8.410433150030184e-06, "loss": 0.0097, "step": 9538 }, { "epoch": 0.6555112699285321, "grad_norm": 0.4404760003089905, "learning_rate": 8.407434149810736e-06, "loss": 0.019, "step": 9539 }, { "epoch": 0.6555799890049477, "grad_norm": 0.40005069971084595, "learning_rate": 8.404435476173442e-06, "loss": 0.0114, "step": 9540 }, { "epoch": 0.6556487080813634, "grad_norm": 0.8465903997421265, "learning_rate": 8.401437129266854e-06, "loss": 0.0119, "step": 9541 }, { "epoch": 0.655717427157779, "grad_norm": 0.35792261362075806, "learning_rate": 8.398439109239504e-06, "loss": 0.0148, "step": 9542 }, { "epoch": 0.6557861462341946, "grad_norm": 0.5296646356582642, "learning_rate": 8.395441416239913e-06, "loss": 0.0113, "step": 9543 }, { "epoch": 0.6558548653106102, "grad_norm": 0.36227941513061523, "learning_rate": 8.392444050416571e-06, "loss": 0.0143, "step": 9544 }, { "epoch": 0.6559235843870258, "grad_norm": 0.49318963289260864, "learning_rate": 8.389447011917974e-06, "loss": 0.0129, "step": 9545 }, { "epoch": 0.6559923034634415, "grad_norm": 0.1868397444486618, "learning_rate": 8.386450300892573e-06, "loss": 0.0093, "step": 9546 }, { "epoch": 0.6560610225398571, "grad_norm": 0.44758570194244385, "learning_rate": 8.383453917488828e-06, "loss": 0.0175, "step": 9547 }, { "epoch": 0.6561297416162727, "grad_norm": 0.39245763421058655, "learning_rate": 8.380457861855178e-06, "loss": 0.01, "step": 9548 }, { "epoch": 0.6561984606926883, "grad_norm": 0.32117730379104614, "learning_rate": 8.37746213414003e-06, "loss": 0.0072, "step": 9549 }, { "epoch": 0.6562671797691039, "grad_norm": 0.5671300292015076, "learning_rate": 8.3744667344918e-06, "loss": 0.0165, "step": 9550 }, { "epoch": 0.6563358988455195, "grad_norm": 0.33934128284454346, "learning_rate": 8.37147166305886e-06, "loss": 0.0083, "step": 9551 }, { "epoch": 0.6564046179219352, "grad_norm": 0.5776668787002563, "learning_rate": 8.368476919989588e-06, "loss": 0.0153, "step": 9552 }, { "epoch": 0.6564733369983508, "grad_norm": 0.6471682190895081, "learning_rate": 8.365482505432334e-06, "loss": 0.0149, "step": 9553 }, { "epoch": 0.6565420560747663, "grad_norm": 1.0680302381515503, "learning_rate": 8.362488419535441e-06, "loss": 0.0126, "step": 9554 }, { "epoch": 0.6566107751511819, "grad_norm": 0.47263655066490173, "learning_rate": 8.359494662447223e-06, "loss": 0.0139, "step": 9555 }, { "epoch": 0.6566794942275975, "grad_norm": 0.18599776923656464, "learning_rate": 8.356501234315988e-06, "loss": 0.0086, "step": 9556 }, { "epoch": 0.6567482133040132, "grad_norm": 1.1066561937332153, "learning_rate": 8.35350813529002e-06, "loss": 0.0181, "step": 9557 }, { "epoch": 0.6568169323804288, "grad_norm": 0.586256742477417, "learning_rate": 8.350515365517604e-06, "loss": 0.0145, "step": 9558 }, { "epoch": 0.6568856514568444, "grad_norm": 0.202071413397789, "learning_rate": 8.347522925146969e-06, "loss": 0.0101, "step": 9559 }, { "epoch": 0.65695437053326, "grad_norm": 0.5712620615959167, "learning_rate": 8.344530814326386e-06, "loss": 0.0122, "step": 9560 }, { "epoch": 0.6570230896096756, "grad_norm": 0.3776453137397766, "learning_rate": 8.341539033204057e-06, "loss": 0.0109, "step": 9561 }, { "epoch": 0.6570918086860913, "grad_norm": 0.40744346380233765, "learning_rate": 8.338547581928201e-06, "loss": 0.0129, "step": 9562 }, { "epoch": 0.6571605277625069, "grad_norm": 0.7130798101425171, "learning_rate": 8.335556460646996e-06, "loss": 0.0191, "step": 9563 }, { "epoch": 0.6572292468389225, "grad_norm": 1.0082460641860962, "learning_rate": 8.332565669508628e-06, "loss": 0.0186, "step": 9564 }, { "epoch": 0.6572979659153381, "grad_norm": 0.1940140724182129, "learning_rate": 8.329575208661244e-06, "loss": 0.0187, "step": 9565 }, { "epoch": 0.6573666849917537, "grad_norm": 0.30642011761665344, "learning_rate": 8.32658507825299e-06, "loss": 0.0141, "step": 9566 }, { "epoch": 0.6574354040681694, "grad_norm": 0.8693123459815979, "learning_rate": 8.32359527843199e-06, "loss": 0.0195, "step": 9567 }, { "epoch": 0.657504123144585, "grad_norm": 0.42211052775382996, "learning_rate": 8.320605809346361e-06, "loss": 0.0091, "step": 9568 }, { "epoch": 0.6575728422210005, "grad_norm": 0.40981626510620117, "learning_rate": 8.317616671144183e-06, "loss": 0.0145, "step": 9569 }, { "epoch": 0.6576415612974161, "grad_norm": 0.16189679503440857, "learning_rate": 8.314627863973542e-06, "loss": 0.0065, "step": 9570 }, { "epoch": 0.6577102803738317, "grad_norm": 0.3322274386882782, "learning_rate": 8.311639387982485e-06, "loss": 0.0096, "step": 9571 }, { "epoch": 0.6577789994502474, "grad_norm": 0.7029440999031067, "learning_rate": 8.308651243319064e-06, "loss": 0.0109, "step": 9572 }, { "epoch": 0.657847718526663, "grad_norm": 0.6087700128555298, "learning_rate": 8.305663430131308e-06, "loss": 0.0097, "step": 9573 }, { "epoch": 0.6579164376030786, "grad_norm": 0.42111310362815857, "learning_rate": 8.302675948567218e-06, "loss": 0.0094, "step": 9574 }, { "epoch": 0.6579851566794942, "grad_norm": 0.5866957306861877, "learning_rate": 8.299688798774797e-06, "loss": 0.0174, "step": 9575 }, { "epoch": 0.6580538757559098, "grad_norm": 0.23359540104866028, "learning_rate": 8.296701980902013e-06, "loss": 0.0096, "step": 9576 }, { "epoch": 0.6581225948323255, "grad_norm": 0.8785355091094971, "learning_rate": 8.29371549509683e-06, "loss": 0.015, "step": 9577 }, { "epoch": 0.6581913139087411, "grad_norm": 0.48825129866600037, "learning_rate": 8.290729341507195e-06, "loss": 0.0118, "step": 9578 }, { "epoch": 0.6582600329851567, "grad_norm": 0.5399473309516907, "learning_rate": 8.28774352028104e-06, "loss": 0.0115, "step": 9579 }, { "epoch": 0.6583287520615723, "grad_norm": 0.5074483752250671, "learning_rate": 8.284758031566264e-06, "loss": 0.0096, "step": 9580 }, { "epoch": 0.6583974711379879, "grad_norm": 0.34699901938438416, "learning_rate": 8.281772875510779e-06, "loss": 0.0129, "step": 9581 }, { "epoch": 0.6584661902144036, "grad_norm": 0.6254207491874695, "learning_rate": 8.278788052262444e-06, "loss": 0.011, "step": 9582 }, { "epoch": 0.6585349092908191, "grad_norm": 0.6807108521461487, "learning_rate": 8.275803561969131e-06, "loss": 0.0299, "step": 9583 }, { "epoch": 0.6586036283672347, "grad_norm": 0.8721994161605835, "learning_rate": 8.272819404778685e-06, "loss": 0.0209, "step": 9584 }, { "epoch": 0.6586723474436503, "grad_norm": 0.4268324375152588, "learning_rate": 8.269835580838943e-06, "loss": 0.0114, "step": 9585 }, { "epoch": 0.6587410665200659, "grad_norm": 0.7381414175033569, "learning_rate": 8.266852090297702e-06, "loss": 0.0139, "step": 9586 }, { "epoch": 0.6588097855964816, "grad_norm": 0.5440401434898376, "learning_rate": 8.263868933302773e-06, "loss": 0.0148, "step": 9587 }, { "epoch": 0.6588785046728972, "grad_norm": 0.7892276644706726, "learning_rate": 8.260886110001923e-06, "loss": 0.0163, "step": 9588 }, { "epoch": 0.6589472237493128, "grad_norm": 0.511937141418457, "learning_rate": 8.257903620542918e-06, "loss": 0.0119, "step": 9589 }, { "epoch": 0.6590159428257284, "grad_norm": 0.6854370832443237, "learning_rate": 8.254921465073515e-06, "loss": 0.0126, "step": 9590 }, { "epoch": 0.659084661902144, "grad_norm": 0.3287400007247925, "learning_rate": 8.251939643741426e-06, "loss": 0.007, "step": 9591 }, { "epoch": 0.6591533809785597, "grad_norm": 0.1935833990573883, "learning_rate": 8.248958156694382e-06, "loss": 0.0074, "step": 9592 }, { "epoch": 0.6592221000549753, "grad_norm": 0.42551809549331665, "learning_rate": 8.245977004080065e-06, "loss": 0.0141, "step": 9593 }, { "epoch": 0.6592908191313909, "grad_norm": 0.5711674094200134, "learning_rate": 8.242996186046167e-06, "loss": 0.0145, "step": 9594 }, { "epoch": 0.6593595382078065, "grad_norm": 0.43443554639816284, "learning_rate": 8.240015702740333e-06, "loss": 0.0122, "step": 9595 }, { "epoch": 0.659428257284222, "grad_norm": 0.6477835178375244, "learning_rate": 8.237035554310238e-06, "loss": 0.0181, "step": 9596 }, { "epoch": 0.6594969763606378, "grad_norm": 0.5562692284584045, "learning_rate": 8.234055740903488e-06, "loss": 0.0119, "step": 9597 }, { "epoch": 0.6595656954370533, "grad_norm": 0.2856224775314331, "learning_rate": 8.231076262667716e-06, "loss": 0.0089, "step": 9598 }, { "epoch": 0.6596344145134689, "grad_norm": 1.0805332660675049, "learning_rate": 8.228097119750501e-06, "loss": 0.0198, "step": 9599 }, { "epoch": 0.6597031335898845, "grad_norm": 0.43533506989479065, "learning_rate": 8.225118312299441e-06, "loss": 0.0117, "step": 9600 }, { "epoch": 0.6597718526663001, "grad_norm": 0.91420578956604, "learning_rate": 8.222139840462078e-06, "loss": 0.0174, "step": 9601 }, { "epoch": 0.6598405717427158, "grad_norm": 0.5298722982406616, "learning_rate": 8.219161704385983e-06, "loss": 0.013, "step": 9602 }, { "epoch": 0.6599092908191314, "grad_norm": 0.6703181862831116, "learning_rate": 8.216183904218672e-06, "loss": 0.0151, "step": 9603 }, { "epoch": 0.659978009895547, "grad_norm": 0.48421385884284973, "learning_rate": 8.213206440107668e-06, "loss": 0.0086, "step": 9604 }, { "epoch": 0.6600467289719626, "grad_norm": 0.5761468410491943, "learning_rate": 8.21022931220046e-06, "loss": 0.0133, "step": 9605 }, { "epoch": 0.6601154480483782, "grad_norm": 0.5215937495231628, "learning_rate": 8.207252520644538e-06, "loss": 0.0168, "step": 9606 }, { "epoch": 0.6601841671247939, "grad_norm": 0.4256506860256195, "learning_rate": 8.204276065587353e-06, "loss": 0.018, "step": 9607 }, { "epoch": 0.6602528862012095, "grad_norm": 0.9221482872962952, "learning_rate": 8.201299947176363e-06, "loss": 0.0127, "step": 9608 }, { "epoch": 0.6603216052776251, "grad_norm": 0.27992573380470276, "learning_rate": 8.198324165559e-06, "loss": 0.0083, "step": 9609 }, { "epoch": 0.6603903243540407, "grad_norm": 0.889348030090332, "learning_rate": 8.195348720882669e-06, "loss": 0.0118, "step": 9610 }, { "epoch": 0.6604590434304562, "grad_norm": 0.4582669734954834, "learning_rate": 8.192373613294779e-06, "loss": 0.0097, "step": 9611 }, { "epoch": 0.660527762506872, "grad_norm": 0.423681378364563, "learning_rate": 8.189398842942697e-06, "loss": 0.0113, "step": 9612 }, { "epoch": 0.6605964815832875, "grad_norm": 0.35926854610443115, "learning_rate": 8.186424409973797e-06, "loss": 0.0073, "step": 9613 }, { "epoch": 0.6606652006597031, "grad_norm": 0.994178831577301, "learning_rate": 8.183450314535421e-06, "loss": 0.0153, "step": 9614 }, { "epoch": 0.6607339197361187, "grad_norm": 0.3601260781288147, "learning_rate": 8.180476556774907e-06, "loss": 0.0087, "step": 9615 }, { "epoch": 0.6608026388125343, "grad_norm": 0.3868720233440399, "learning_rate": 8.177503136839563e-06, "loss": 0.0125, "step": 9616 }, { "epoch": 0.66087135788895, "grad_norm": 1.7978452444076538, "learning_rate": 8.17453005487669e-06, "loss": 0.0117, "step": 9617 }, { "epoch": 0.6609400769653656, "grad_norm": 0.48237818479537964, "learning_rate": 8.171557311033558e-06, "loss": 0.0127, "step": 9618 }, { "epoch": 0.6610087960417812, "grad_norm": 0.3959561586380005, "learning_rate": 8.16858490545744e-06, "loss": 0.0121, "step": 9619 }, { "epoch": 0.6610775151181968, "grad_norm": 0.3816784918308258, "learning_rate": 8.16561283829558e-06, "loss": 0.0148, "step": 9620 }, { "epoch": 0.6611462341946124, "grad_norm": 0.3495751619338989, "learning_rate": 8.162641109695216e-06, "loss": 0.0125, "step": 9621 }, { "epoch": 0.6612149532710281, "grad_norm": 0.5435428023338318, "learning_rate": 8.159669719803548e-06, "loss": 0.0089, "step": 9622 }, { "epoch": 0.6612836723474437, "grad_norm": 0.6372271776199341, "learning_rate": 8.156698668767786e-06, "loss": 0.012, "step": 9623 }, { "epoch": 0.6613523914238593, "grad_norm": 0.48625487089157104, "learning_rate": 8.153727956735096e-06, "loss": 0.0165, "step": 9624 }, { "epoch": 0.6614211105002749, "grad_norm": 0.420404314994812, "learning_rate": 8.150757583852654e-06, "loss": 0.0109, "step": 9625 }, { "epoch": 0.6614898295766904, "grad_norm": 0.2963430881500244, "learning_rate": 8.147787550267593e-06, "loss": 0.0176, "step": 9626 }, { "epoch": 0.6615585486531061, "grad_norm": 0.5913437008857727, "learning_rate": 8.14481785612705e-06, "loss": 0.0204, "step": 9627 }, { "epoch": 0.6616272677295217, "grad_norm": 0.34263983368873596, "learning_rate": 8.141848501578143e-06, "loss": 0.0141, "step": 9628 }, { "epoch": 0.6616959868059373, "grad_norm": 0.9261531233787537, "learning_rate": 8.138879486767954e-06, "loss": 0.0089, "step": 9629 }, { "epoch": 0.6617647058823529, "grad_norm": 0.47137588262557983, "learning_rate": 8.135910811843577e-06, "loss": 0.0113, "step": 9630 }, { "epoch": 0.6618334249587685, "grad_norm": 0.37276491522789, "learning_rate": 8.132942476952058e-06, "loss": 0.0069, "step": 9631 }, { "epoch": 0.6619021440351842, "grad_norm": 0.7269026637077332, "learning_rate": 8.129974482240454e-06, "loss": 0.0181, "step": 9632 }, { "epoch": 0.6619708631115998, "grad_norm": 0.35048162937164307, "learning_rate": 8.12700682785579e-06, "loss": 0.0103, "step": 9633 }, { "epoch": 0.6620395821880154, "grad_norm": 0.13306766748428345, "learning_rate": 8.124039513945081e-06, "loss": 0.0133, "step": 9634 }, { "epoch": 0.662108301264431, "grad_norm": 0.5169276595115662, "learning_rate": 8.12107254065531e-06, "loss": 0.0065, "step": 9635 }, { "epoch": 0.6621770203408466, "grad_norm": 0.40877866744995117, "learning_rate": 8.118105908133474e-06, "loss": 0.02, "step": 9636 }, { "epoch": 0.6622457394172623, "grad_norm": 0.28233036398887634, "learning_rate": 8.115139616526513e-06, "loss": 0.0125, "step": 9637 }, { "epoch": 0.6623144584936779, "grad_norm": 0.4251050353050232, "learning_rate": 8.112173665981384e-06, "loss": 0.019, "step": 9638 }, { "epoch": 0.6623831775700935, "grad_norm": 0.33809661865234375, "learning_rate": 8.109208056645009e-06, "loss": 0.0143, "step": 9639 }, { "epoch": 0.662451896646509, "grad_norm": 0.7415963411331177, "learning_rate": 8.106242788664307e-06, "loss": 0.0163, "step": 9640 }, { "epoch": 0.6625206157229246, "grad_norm": 0.6079705953598022, "learning_rate": 8.103277862186154e-06, "loss": 0.0105, "step": 9641 }, { "epoch": 0.6625893347993403, "grad_norm": 0.29598936438560486, "learning_rate": 8.100313277357449e-06, "loss": 0.0125, "step": 9642 }, { "epoch": 0.6626580538757559, "grad_norm": 0.6803948879241943, "learning_rate": 8.097349034325028e-06, "loss": 0.0193, "step": 9643 }, { "epoch": 0.6627267729521715, "grad_norm": 0.7786543369293213, "learning_rate": 8.094385133235747e-06, "loss": 0.0164, "step": 9644 }, { "epoch": 0.6627954920285871, "grad_norm": 0.21325737237930298, "learning_rate": 8.091421574236435e-06, "loss": 0.006, "step": 9645 }, { "epoch": 0.6628642111050027, "grad_norm": 1.9344613552093506, "learning_rate": 8.088458357473888e-06, "loss": 0.0168, "step": 9646 }, { "epoch": 0.6629329301814184, "grad_norm": 0.19804242253303528, "learning_rate": 8.085495483094913e-06, "loss": 0.0091, "step": 9647 }, { "epoch": 0.663001649257834, "grad_norm": 0.31689080595970154, "learning_rate": 8.082532951246267e-06, "loss": 0.009, "step": 9648 }, { "epoch": 0.6630703683342496, "grad_norm": 0.5471302270889282, "learning_rate": 8.079570762074722e-06, "loss": 0.0199, "step": 9649 }, { "epoch": 0.6631390874106652, "grad_norm": 0.4099847078323364, "learning_rate": 8.076608915727006e-06, "loss": 0.0225, "step": 9650 }, { "epoch": 0.6632078064870808, "grad_norm": 0.4071512222290039, "learning_rate": 8.07364741234986e-06, "loss": 0.0137, "step": 9651 }, { "epoch": 0.6632765255634965, "grad_norm": 0.4250440299510956, "learning_rate": 8.070686252089975e-06, "loss": 0.0128, "step": 9652 }, { "epoch": 0.6633452446399121, "grad_norm": 0.3920762836933136, "learning_rate": 8.067725435094053e-06, "loss": 0.0138, "step": 9653 }, { "epoch": 0.6634139637163277, "grad_norm": 1.192657709121704, "learning_rate": 8.064764961508755e-06, "loss": 0.0153, "step": 9654 }, { "epoch": 0.6634826827927433, "grad_norm": 0.6613327264785767, "learning_rate": 8.061804831480749e-06, "loss": 0.0169, "step": 9655 }, { "epoch": 0.6635514018691588, "grad_norm": 0.6621221899986267, "learning_rate": 8.058845045156656e-06, "loss": 0.017, "step": 9656 }, { "epoch": 0.6636201209455745, "grad_norm": 0.7679387331008911, "learning_rate": 8.05588560268312e-06, "loss": 0.021, "step": 9657 }, { "epoch": 0.6636888400219901, "grad_norm": 0.5736732482910156, "learning_rate": 8.052926504206728e-06, "loss": 0.0121, "step": 9658 }, { "epoch": 0.6637575590984057, "grad_norm": 0.7266834378242493, "learning_rate": 8.04996774987408e-06, "loss": 0.0179, "step": 9659 }, { "epoch": 0.6638262781748213, "grad_norm": 0.5641521215438843, "learning_rate": 8.047009339831738e-06, "loss": 0.0135, "step": 9660 }, { "epoch": 0.6638949972512369, "grad_norm": 0.3848123252391815, "learning_rate": 8.044051274226262e-06, "loss": 0.0142, "step": 9661 }, { "epoch": 0.6639637163276526, "grad_norm": 0.6007639169692993, "learning_rate": 8.041093553204181e-06, "loss": 0.015, "step": 9662 }, { "epoch": 0.6640324354040682, "grad_norm": 0.7375457286834717, "learning_rate": 8.038136176912017e-06, "loss": 0.0086, "step": 9663 }, { "epoch": 0.6641011544804838, "grad_norm": 0.47160497307777405, "learning_rate": 8.035179145496279e-06, "loss": 0.0077, "step": 9664 }, { "epoch": 0.6641698735568994, "grad_norm": 0.4995347559452057, "learning_rate": 8.032222459103442e-06, "loss": 0.0148, "step": 9665 }, { "epoch": 0.664238592633315, "grad_norm": 0.553365170955658, "learning_rate": 8.029266117879986e-06, "loss": 0.0101, "step": 9666 }, { "epoch": 0.6643073117097307, "grad_norm": 0.3178567588329315, "learning_rate": 8.026310121972348e-06, "loss": 0.0137, "step": 9667 }, { "epoch": 0.6643760307861463, "grad_norm": 0.6472551822662354, "learning_rate": 8.02335447152697e-06, "loss": 0.0234, "step": 9668 }, { "epoch": 0.6644447498625619, "grad_norm": 0.7308430075645447, "learning_rate": 8.020399166690268e-06, "loss": 0.0167, "step": 9669 }, { "epoch": 0.6645134689389774, "grad_norm": 0.4860689640045166, "learning_rate": 8.01744420760865e-06, "loss": 0.0192, "step": 9670 }, { "epoch": 0.664582188015393, "grad_norm": 1.9968525171279907, "learning_rate": 8.01448959442848e-06, "loss": 0.0122, "step": 9671 }, { "epoch": 0.6646509070918087, "grad_norm": 0.3515027165412903, "learning_rate": 8.011535327296142e-06, "loss": 0.0182, "step": 9672 }, { "epoch": 0.6647196261682243, "grad_norm": 0.5548986792564392, "learning_rate": 8.008581406357973e-06, "loss": 0.0133, "step": 9673 }, { "epoch": 0.6647883452446399, "grad_norm": 0.30323249101638794, "learning_rate": 8.005627831760303e-06, "loss": 0.018, "step": 9674 }, { "epoch": 0.6648570643210555, "grad_norm": 0.1401240974664688, "learning_rate": 8.002674603649454e-06, "loss": 0.008, "step": 9675 }, { "epoch": 0.6649257833974711, "grad_norm": 0.9544025659561157, "learning_rate": 7.99972172217172e-06, "loss": 0.0122, "step": 9676 }, { "epoch": 0.6649945024738868, "grad_norm": 0.4874904751777649, "learning_rate": 7.99676918747338e-06, "loss": 0.0199, "step": 9677 }, { "epoch": 0.6650632215503024, "grad_norm": 0.5411961078643799, "learning_rate": 7.993816999700699e-06, "loss": 0.0151, "step": 9678 }, { "epoch": 0.665131940626718, "grad_norm": 0.26780784130096436, "learning_rate": 7.990865158999913e-06, "loss": 0.0074, "step": 9679 }, { "epoch": 0.6652006597031336, "grad_norm": 0.689263641834259, "learning_rate": 7.98791366551726e-06, "loss": 0.0114, "step": 9680 }, { "epoch": 0.6652693787795492, "grad_norm": 0.35201677680015564, "learning_rate": 7.984962519398952e-06, "loss": 0.016, "step": 9681 }, { "epoch": 0.6653380978559648, "grad_norm": 0.5288193821907043, "learning_rate": 7.982011720791173e-06, "loss": 0.0168, "step": 9682 }, { "epoch": 0.6654068169323805, "grad_norm": 0.5193855166435242, "learning_rate": 7.97906126984011e-06, "loss": 0.0131, "step": 9683 }, { "epoch": 0.665475536008796, "grad_norm": 0.39763906598091125, "learning_rate": 7.976111166691913e-06, "loss": 0.0123, "step": 9684 }, { "epoch": 0.6655442550852116, "grad_norm": 0.32928410172462463, "learning_rate": 7.973161411492733e-06, "loss": 0.0118, "step": 9685 }, { "epoch": 0.6656129741616272, "grad_norm": 0.34245026111602783, "learning_rate": 7.970212004388677e-06, "loss": 0.012, "step": 9686 }, { "epoch": 0.6656816932380428, "grad_norm": 0.4077739715576172, "learning_rate": 7.96726294552588e-06, "loss": 0.0157, "step": 9687 }, { "epoch": 0.6657504123144585, "grad_norm": 0.30337029695510864, "learning_rate": 7.96431423505041e-06, "loss": 0.0109, "step": 9688 }, { "epoch": 0.6658191313908741, "grad_norm": 0.25924110412597656, "learning_rate": 7.961365873108356e-06, "loss": 0.0102, "step": 9689 }, { "epoch": 0.6658878504672897, "grad_norm": 0.7004212737083435, "learning_rate": 7.958417859845759e-06, "loss": 0.0198, "step": 9690 }, { "epoch": 0.6659565695437053, "grad_norm": 0.32673436403274536, "learning_rate": 7.95547019540867e-06, "loss": 0.0104, "step": 9691 }, { "epoch": 0.6660252886201209, "grad_norm": 0.5098075270652771, "learning_rate": 7.952522879943093e-06, "loss": 0.01, "step": 9692 }, { "epoch": 0.6660940076965366, "grad_norm": 1.0995594263076782, "learning_rate": 7.949575913595054e-06, "loss": 0.0116, "step": 9693 }, { "epoch": 0.6661627267729522, "grad_norm": 0.21952030062675476, "learning_rate": 7.946629296510526e-06, "loss": 0.0101, "step": 9694 }, { "epoch": 0.6662314458493678, "grad_norm": 0.4926830232143402, "learning_rate": 7.943683028835485e-06, "loss": 0.0127, "step": 9695 }, { "epoch": 0.6663001649257834, "grad_norm": 0.6145357489585876, "learning_rate": 7.940737110715874e-06, "loss": 0.0151, "step": 9696 }, { "epoch": 0.666368884002199, "grad_norm": 1.2205469608306885, "learning_rate": 7.937791542297641e-06, "loss": 0.0209, "step": 9697 }, { "epoch": 0.6664376030786147, "grad_norm": 0.684316098690033, "learning_rate": 7.934846323726688e-06, "loss": 0.0109, "step": 9698 }, { "epoch": 0.6665063221550303, "grad_norm": 0.7017850279808044, "learning_rate": 7.931901455148923e-06, "loss": 0.0055, "step": 9699 }, { "epoch": 0.6665750412314458, "grad_norm": 0.7260716557502747, "learning_rate": 7.928956936710236e-06, "loss": 0.0087, "step": 9700 }, { "epoch": 0.6666437603078614, "grad_norm": 0.272028386592865, "learning_rate": 7.926012768556478e-06, "loss": 0.008, "step": 9701 }, { "epoch": 0.666712479384277, "grad_norm": 0.7257969975471497, "learning_rate": 7.923068950833509e-06, "loss": 0.016, "step": 9702 }, { "epoch": 0.6667811984606927, "grad_norm": 0.46013757586479187, "learning_rate": 7.920125483687149e-06, "loss": 0.0134, "step": 9703 }, { "epoch": 0.6668499175371083, "grad_norm": 0.651644766330719, "learning_rate": 7.917182367263218e-06, "loss": 0.0109, "step": 9704 }, { "epoch": 0.6669186366135239, "grad_norm": 1.0731722116470337, "learning_rate": 7.91423960170751e-06, "loss": 0.0072, "step": 9705 }, { "epoch": 0.6669873556899395, "grad_norm": 0.4755766987800598, "learning_rate": 7.911297187165811e-06, "loss": 0.0121, "step": 9706 }, { "epoch": 0.6670560747663551, "grad_norm": 0.679779589176178, "learning_rate": 7.908355123783867e-06, "loss": 0.0123, "step": 9707 }, { "epoch": 0.6671247938427708, "grad_norm": 0.48769906163215637, "learning_rate": 7.90541341170744e-06, "loss": 0.0115, "step": 9708 }, { "epoch": 0.6671935129191864, "grad_norm": 0.4084290564060211, "learning_rate": 7.902472051082239e-06, "loss": 0.0093, "step": 9709 }, { "epoch": 0.667262231995602, "grad_norm": 1.2659564018249512, "learning_rate": 7.89953104205398e-06, "loss": 0.0268, "step": 9710 }, { "epoch": 0.6673309510720176, "grad_norm": 0.2863510251045227, "learning_rate": 7.896590384768356e-06, "loss": 0.0103, "step": 9711 }, { "epoch": 0.6673996701484332, "grad_norm": 0.8542715311050415, "learning_rate": 7.893650079371048e-06, "loss": 0.0125, "step": 9712 }, { "epoch": 0.6674683892248489, "grad_norm": 0.4109908938407898, "learning_rate": 7.890710126007697e-06, "loss": 0.0158, "step": 9713 }, { "epoch": 0.6675371083012644, "grad_norm": 0.4437732398509979, "learning_rate": 7.887770524823956e-06, "loss": 0.0196, "step": 9714 }, { "epoch": 0.66760582737768, "grad_norm": 0.7549803853034973, "learning_rate": 7.884831275965437e-06, "loss": 0.0126, "step": 9715 }, { "epoch": 0.6676745464540956, "grad_norm": 0.3275461196899414, "learning_rate": 7.881892379577747e-06, "loss": 0.0086, "step": 9716 }, { "epoch": 0.6677432655305112, "grad_norm": 0.5059723854064941, "learning_rate": 7.878953835806481e-06, "loss": 0.02, "step": 9717 }, { "epoch": 0.6678119846069269, "grad_norm": 0.3282387852668762, "learning_rate": 7.876015644797197e-06, "loss": 0.0077, "step": 9718 }, { "epoch": 0.6678807036833425, "grad_norm": 0.4491584300994873, "learning_rate": 7.873077806695454e-06, "loss": 0.012, "step": 9719 }, { "epoch": 0.6679494227597581, "grad_norm": 0.3524303436279297, "learning_rate": 7.870140321646781e-06, "loss": 0.02, "step": 9720 }, { "epoch": 0.6680181418361737, "grad_norm": 0.527556836605072, "learning_rate": 7.867203189796703e-06, "loss": 0.0156, "step": 9721 }, { "epoch": 0.6680868609125893, "grad_norm": 0.52299565076828, "learning_rate": 7.864266411290704e-06, "loss": 0.0117, "step": 9722 }, { "epoch": 0.668155579989005, "grad_norm": 0.46311435103416443, "learning_rate": 7.861329986274286e-06, "loss": 0.0105, "step": 9723 }, { "epoch": 0.6682242990654206, "grad_norm": 0.26810476183891296, "learning_rate": 7.858393914892899e-06, "loss": 0.0083, "step": 9724 }, { "epoch": 0.6682930181418362, "grad_norm": 0.2966035306453705, "learning_rate": 7.855458197292003e-06, "loss": 0.0096, "step": 9725 }, { "epoch": 0.6683617372182518, "grad_norm": 0.4344950020313263, "learning_rate": 7.852522833617011e-06, "loss": 0.0116, "step": 9726 }, { "epoch": 0.6684304562946674, "grad_norm": 3.2295074462890625, "learning_rate": 7.84958782401335e-06, "loss": 0.0292, "step": 9727 }, { "epoch": 0.668499175371083, "grad_norm": 0.23424771428108215, "learning_rate": 7.846653168626395e-06, "loss": 0.0093, "step": 9728 }, { "epoch": 0.6685678944474986, "grad_norm": 0.2590954601764679, "learning_rate": 7.843718867601548e-06, "loss": 0.0115, "step": 9729 }, { "epoch": 0.6686366135239142, "grad_norm": 0.47672852873802185, "learning_rate": 7.840784921084148e-06, "loss": 0.0118, "step": 9730 }, { "epoch": 0.6687053326003298, "grad_norm": 0.8823180794715881, "learning_rate": 7.837851329219553e-06, "loss": 0.0123, "step": 9731 }, { "epoch": 0.6687740516767454, "grad_norm": 0.2719578742980957, "learning_rate": 7.834918092153071e-06, "loss": 0.0129, "step": 9732 }, { "epoch": 0.6688427707531611, "grad_norm": 1.569079041481018, "learning_rate": 7.831985210030021e-06, "loss": 0.0206, "step": 9733 }, { "epoch": 0.6689114898295767, "grad_norm": 0.3363645374774933, "learning_rate": 7.829052682995682e-06, "loss": 0.0182, "step": 9734 }, { "epoch": 0.6689802089059923, "grad_norm": 0.40262895822525024, "learning_rate": 7.82612051119533e-06, "loss": 0.0142, "step": 9735 }, { "epoch": 0.6690489279824079, "grad_norm": 0.7032443284988403, "learning_rate": 7.823188694774227e-06, "loss": 0.0147, "step": 9736 }, { "epoch": 0.6691176470588235, "grad_norm": 0.8515105843544006, "learning_rate": 7.820257233877592e-06, "loss": 0.0236, "step": 9737 }, { "epoch": 0.6691863661352392, "grad_norm": 0.4937129616737366, "learning_rate": 7.81732612865066e-06, "loss": 0.0088, "step": 9738 }, { "epoch": 0.6692550852116548, "grad_norm": 0.4659033715724945, "learning_rate": 7.814395379238618e-06, "loss": 0.013, "step": 9739 }, { "epoch": 0.6693238042880704, "grad_norm": 0.20621509850025177, "learning_rate": 7.811464985786657e-06, "loss": 0.0154, "step": 9740 }, { "epoch": 0.669392523364486, "grad_norm": 0.902945876121521, "learning_rate": 7.808534948439941e-06, "loss": 0.007, "step": 9741 }, { "epoch": 0.6694612424409015, "grad_norm": 0.46572378277778625, "learning_rate": 7.805605267343624e-06, "loss": 0.0109, "step": 9742 }, { "epoch": 0.6695299615173173, "grad_norm": 0.49577364325523376, "learning_rate": 7.802675942642829e-06, "loss": 0.0139, "step": 9743 }, { "epoch": 0.6695986805937328, "grad_norm": 0.5629599094390869, "learning_rate": 7.799746974482673e-06, "loss": 0.0184, "step": 9744 }, { "epoch": 0.6696673996701484, "grad_norm": 1.084640622138977, "learning_rate": 7.796818363008244e-06, "loss": 0.0372, "step": 9745 }, { "epoch": 0.669736118746564, "grad_norm": 0.2998821437358856, "learning_rate": 7.793890108364622e-06, "loss": 0.0127, "step": 9746 }, { "epoch": 0.6698048378229796, "grad_norm": 0.3190207779407501, "learning_rate": 7.790962210696872e-06, "loss": 0.01, "step": 9747 }, { "epoch": 0.6698735568993953, "grad_norm": 0.40247124433517456, "learning_rate": 7.788034670150035e-06, "loss": 0.0152, "step": 9748 }, { "epoch": 0.6699422759758109, "grad_norm": 0.5769086480140686, "learning_rate": 7.78510748686913e-06, "loss": 0.0123, "step": 9749 }, { "epoch": 0.6700109950522265, "grad_norm": 0.43759918212890625, "learning_rate": 7.782180660999172e-06, "loss": 0.0098, "step": 9750 }, { "epoch": 0.6700797141286421, "grad_norm": 0.37949201464653015, "learning_rate": 7.77925419268514e-06, "loss": 0.0162, "step": 9751 }, { "epoch": 0.6701484332050577, "grad_norm": 0.5320408940315247, "learning_rate": 7.77632808207201e-06, "loss": 0.0229, "step": 9752 }, { "epoch": 0.6702171522814734, "grad_norm": 1.0315608978271484, "learning_rate": 7.773402329304742e-06, "loss": 0.0123, "step": 9753 }, { "epoch": 0.670285871357889, "grad_norm": 0.23435404896736145, "learning_rate": 7.770476934528258e-06, "loss": 0.0175, "step": 9754 }, { "epoch": 0.6703545904343046, "grad_norm": 0.8899797797203064, "learning_rate": 7.767551897887488e-06, "loss": 0.0219, "step": 9755 }, { "epoch": 0.6704233095107202, "grad_norm": 0.8214015960693359, "learning_rate": 7.764627219527324e-06, "loss": 0.0262, "step": 9756 }, { "epoch": 0.6704920285871357, "grad_norm": 0.12886932492256165, "learning_rate": 7.761702899592652e-06, "loss": 0.006, "step": 9757 }, { "epoch": 0.6705607476635514, "grad_norm": 0.3913183808326721, "learning_rate": 7.758778938228337e-06, "loss": 0.0098, "step": 9758 }, { "epoch": 0.670629466739967, "grad_norm": 0.2502438724040985, "learning_rate": 7.755855335579232e-06, "loss": 0.0119, "step": 9759 }, { "epoch": 0.6706981858163826, "grad_norm": 0.1525695025920868, "learning_rate": 7.752932091790154e-06, "loss": 0.0116, "step": 9760 }, { "epoch": 0.6707669048927982, "grad_norm": 0.5009323358535767, "learning_rate": 7.750009207005926e-06, "loss": 0.0174, "step": 9761 }, { "epoch": 0.6708356239692138, "grad_norm": 0.3951858580112457, "learning_rate": 7.747086681371334e-06, "loss": 0.0127, "step": 9762 }, { "epoch": 0.6709043430456295, "grad_norm": 0.710311233997345, "learning_rate": 7.744164515031158e-06, "loss": 0.0182, "step": 9763 }, { "epoch": 0.6709730621220451, "grad_norm": 0.31801265478134155, "learning_rate": 7.741242708130143e-06, "loss": 0.013, "step": 9764 }, { "epoch": 0.6710417811984607, "grad_norm": 0.5460506081581116, "learning_rate": 7.738321260813056e-06, "loss": 0.0119, "step": 9765 }, { "epoch": 0.6711105002748763, "grad_norm": 0.25777116417884827, "learning_rate": 7.735400173224598e-06, "loss": 0.0191, "step": 9766 }, { "epoch": 0.6711792193512919, "grad_norm": 0.3340612053871155, "learning_rate": 7.732479445509484e-06, "loss": 0.0165, "step": 9767 }, { "epoch": 0.6712479384277076, "grad_norm": 2.300513982772827, "learning_rate": 7.729559077812391e-06, "loss": 0.0201, "step": 9768 }, { "epoch": 0.6713166575041232, "grad_norm": 0.6147249341011047, "learning_rate": 7.726639070278002e-06, "loss": 0.0111, "step": 9769 }, { "epoch": 0.6713853765805388, "grad_norm": 0.3344257175922394, "learning_rate": 7.723719423050952e-06, "loss": 0.0149, "step": 9770 }, { "epoch": 0.6714540956569544, "grad_norm": 0.07992038130760193, "learning_rate": 7.720800136275884e-06, "loss": 0.0102, "step": 9771 }, { "epoch": 0.6715228147333699, "grad_norm": 0.6985715627670288, "learning_rate": 7.717881210097414e-06, "loss": 0.0108, "step": 9772 }, { "epoch": 0.6715915338097856, "grad_norm": 0.3829280138015747, "learning_rate": 7.714962644660136e-06, "loss": 0.0096, "step": 9773 }, { "epoch": 0.6716602528862012, "grad_norm": 0.5853068232536316, "learning_rate": 7.71204444010863e-06, "loss": 0.02, "step": 9774 }, { "epoch": 0.6717289719626168, "grad_norm": 0.20304177701473236, "learning_rate": 7.709126596587463e-06, "loss": 0.0172, "step": 9775 }, { "epoch": 0.6717976910390324, "grad_norm": 3.4859848022460938, "learning_rate": 7.706209114241172e-06, "loss": 0.0242, "step": 9776 }, { "epoch": 0.671866410115448, "grad_norm": 0.3706836700439453, "learning_rate": 7.703291993214285e-06, "loss": 0.0118, "step": 9777 }, { "epoch": 0.6719351291918637, "grad_norm": 0.4914858043193817, "learning_rate": 7.700375233651316e-06, "loss": 0.0165, "step": 9778 }, { "epoch": 0.6720038482682793, "grad_norm": 2.666346311569214, "learning_rate": 7.697458835696745e-06, "loss": 0.0113, "step": 9779 }, { "epoch": 0.6720725673446949, "grad_norm": 0.469862699508667, "learning_rate": 7.694542799495057e-06, "loss": 0.008, "step": 9780 }, { "epoch": 0.6721412864211105, "grad_norm": 1.757548213005066, "learning_rate": 7.691627125190694e-06, "loss": 0.016, "step": 9781 }, { "epoch": 0.6722100054975261, "grad_norm": 0.6291857957839966, "learning_rate": 7.688711812928097e-06, "loss": 0.0137, "step": 9782 }, { "epoch": 0.6722787245739418, "grad_norm": 0.4276408553123474, "learning_rate": 7.685796862851683e-06, "loss": 0.0113, "step": 9783 }, { "epoch": 0.6723474436503574, "grad_norm": 0.5637755990028381, "learning_rate": 7.682882275105864e-06, "loss": 0.0109, "step": 9784 }, { "epoch": 0.672416162726773, "grad_norm": 0.898105263710022, "learning_rate": 7.679968049835007e-06, "loss": 0.0179, "step": 9785 }, { "epoch": 0.6724848818031886, "grad_norm": 0.6573602557182312, "learning_rate": 7.677054187183487e-06, "loss": 0.0157, "step": 9786 }, { "epoch": 0.6725536008796041, "grad_norm": 0.21082495152950287, "learning_rate": 7.674140687295645e-06, "loss": 0.0059, "step": 9787 }, { "epoch": 0.6726223199560198, "grad_norm": 0.3561513125896454, "learning_rate": 7.671227550315811e-06, "loss": 0.01, "step": 9788 }, { "epoch": 0.6726910390324354, "grad_norm": 0.5151116847991943, "learning_rate": 7.668314776388301e-06, "loss": 0.0184, "step": 9789 }, { "epoch": 0.672759758108851, "grad_norm": 0.44943660497665405, "learning_rate": 7.6654023656574e-06, "loss": 0.0155, "step": 9790 }, { "epoch": 0.6728284771852666, "grad_norm": 0.5966516137123108, "learning_rate": 7.662490318267384e-06, "loss": 0.0341, "step": 9791 }, { "epoch": 0.6728971962616822, "grad_norm": 0.2619277834892273, "learning_rate": 7.65957863436252e-06, "loss": 0.0092, "step": 9792 }, { "epoch": 0.6729659153380979, "grad_norm": 0.3029983341693878, "learning_rate": 7.656667314087034e-06, "loss": 0.0111, "step": 9793 }, { "epoch": 0.6730346344145135, "grad_norm": 0.719899594783783, "learning_rate": 7.65375635758515e-06, "loss": 0.0192, "step": 9794 }, { "epoch": 0.6731033534909291, "grad_norm": 0.46006345748901367, "learning_rate": 7.65084576500108e-06, "loss": 0.0132, "step": 9795 }, { "epoch": 0.6731720725673447, "grad_norm": 0.3693127930164337, "learning_rate": 7.647935536478996e-06, "loss": 0.0147, "step": 9796 }, { "epoch": 0.6732407916437603, "grad_norm": 0.15557053685188293, "learning_rate": 7.645025672163075e-06, "loss": 0.0089, "step": 9797 }, { "epoch": 0.673309510720176, "grad_norm": 0.4003460109233856, "learning_rate": 7.642116172197456e-06, "loss": 0.0237, "step": 9798 }, { "epoch": 0.6733782297965916, "grad_norm": 0.17215219140052795, "learning_rate": 7.639207036726282e-06, "loss": 0.0063, "step": 9799 }, { "epoch": 0.6734469488730072, "grad_norm": 0.3622604310512543, "learning_rate": 7.636298265893643e-06, "loss": 0.0084, "step": 9800 }, { "epoch": 0.6735156679494227, "grad_norm": 1.2058545351028442, "learning_rate": 7.633389859843664e-06, "loss": 0.0221, "step": 9801 }, { "epoch": 0.6735843870258383, "grad_norm": 0.3141258955001831, "learning_rate": 7.630481818720397e-06, "loss": 0.0132, "step": 9802 }, { "epoch": 0.673653106102254, "grad_norm": 1.205323576927185, "learning_rate": 7.627574142667918e-06, "loss": 0.0227, "step": 9803 }, { "epoch": 0.6737218251786696, "grad_norm": 0.3562140464782715, "learning_rate": 7.624666831830252e-06, "loss": 0.0147, "step": 9804 }, { "epoch": 0.6737905442550852, "grad_norm": 0.44714871048927307, "learning_rate": 7.621759886351432e-06, "loss": 0.0174, "step": 9805 }, { "epoch": 0.6738592633315008, "grad_norm": 0.8491315841674805, "learning_rate": 7.61885330637545e-06, "loss": 0.0222, "step": 9806 }, { "epoch": 0.6739279824079164, "grad_norm": 0.9828423261642456, "learning_rate": 7.615947092046311e-06, "loss": 0.0152, "step": 9807 }, { "epoch": 0.6739967014843321, "grad_norm": 0.5267623066902161, "learning_rate": 7.613041243507967e-06, "loss": 0.0103, "step": 9808 }, { "epoch": 0.6740654205607477, "grad_norm": 0.591110110282898, "learning_rate": 7.6101357609043766e-06, "loss": 0.0149, "step": 9809 }, { "epoch": 0.6741341396371633, "grad_norm": 0.18396708369255066, "learning_rate": 7.6072306443794635e-06, "loss": 0.0124, "step": 9810 }, { "epoch": 0.6742028587135789, "grad_norm": 0.5004593133926392, "learning_rate": 7.60432589407715e-06, "loss": 0.0186, "step": 9811 }, { "epoch": 0.6742715777899945, "grad_norm": 0.46159279346466064, "learning_rate": 7.601421510141322e-06, "loss": 0.005, "step": 9812 }, { "epoch": 0.6743402968664101, "grad_norm": 0.5004802942276001, "learning_rate": 7.59851749271586e-06, "loss": 0.0117, "step": 9813 }, { "epoch": 0.6744090159428258, "grad_norm": 0.6857544183731079, "learning_rate": 7.595613841944631e-06, "loss": 0.012, "step": 9814 }, { "epoch": 0.6744777350192414, "grad_norm": 0.7578995823860168, "learning_rate": 7.592710557971465e-06, "loss": 0.0191, "step": 9815 }, { "epoch": 0.6745464540956569, "grad_norm": 1.1987037658691406, "learning_rate": 7.5898076409401925e-06, "loss": 0.0189, "step": 9816 }, { "epoch": 0.6746151731720725, "grad_norm": 0.4701751172542572, "learning_rate": 7.586905090994612e-06, "loss": 0.0122, "step": 9817 }, { "epoch": 0.6746838922484881, "grad_norm": 0.7536119818687439, "learning_rate": 7.584002908278513e-06, "loss": 0.0217, "step": 9818 }, { "epoch": 0.6747526113249038, "grad_norm": 0.4013589918613434, "learning_rate": 7.581101092935661e-06, "loss": 0.0095, "step": 9819 }, { "epoch": 0.6748213304013194, "grad_norm": 0.7047732472419739, "learning_rate": 7.578199645109814e-06, "loss": 0.014, "step": 9820 }, { "epoch": 0.674890049477735, "grad_norm": 0.4269149899482727, "learning_rate": 7.575298564944693e-06, "loss": 0.0116, "step": 9821 }, { "epoch": 0.6749587685541506, "grad_norm": 0.6965612769126892, "learning_rate": 7.5723978525840215e-06, "loss": 0.0166, "step": 9822 }, { "epoch": 0.6750274876305662, "grad_norm": 0.20691576600074768, "learning_rate": 7.569497508171486e-06, "loss": 0.0147, "step": 9823 }, { "epoch": 0.6750962067069819, "grad_norm": 0.3925575613975525, "learning_rate": 7.566597531850765e-06, "loss": 0.0186, "step": 9824 }, { "epoch": 0.6751649257833975, "grad_norm": 0.2978757321834564, "learning_rate": 7.56369792376552e-06, "loss": 0.0142, "step": 9825 }, { "epoch": 0.6752336448598131, "grad_norm": 0.4008124768733978, "learning_rate": 7.560798684059399e-06, "loss": 0.0072, "step": 9826 }, { "epoch": 0.6753023639362287, "grad_norm": 0.3217986226081848, "learning_rate": 7.55789981287601e-06, "loss": 0.0151, "step": 9827 }, { "epoch": 0.6753710830126443, "grad_norm": 0.4871467053890228, "learning_rate": 7.555001310358972e-06, "loss": 0.0101, "step": 9828 }, { "epoch": 0.67543980208906, "grad_norm": 0.49690884351730347, "learning_rate": 7.5521031766518534e-06, "loss": 0.0072, "step": 9829 }, { "epoch": 0.6755085211654756, "grad_norm": 0.458782821893692, "learning_rate": 7.549205411898234e-06, "loss": 0.0127, "step": 9830 }, { "epoch": 0.6755772402418911, "grad_norm": 0.9879658818244934, "learning_rate": 7.546308016241664e-06, "loss": 0.0167, "step": 9831 }, { "epoch": 0.6756459593183067, "grad_norm": 0.11984898895025253, "learning_rate": 7.543410989825668e-06, "loss": 0.0126, "step": 9832 }, { "epoch": 0.6757146783947223, "grad_norm": 0.355919748544693, "learning_rate": 7.540514332793764e-06, "loss": 0.0139, "step": 9833 }, { "epoch": 0.675783397471138, "grad_norm": 0.3467872738838196, "learning_rate": 7.537618045289439e-06, "loss": 0.0101, "step": 9834 }, { "epoch": 0.6758521165475536, "grad_norm": 0.4219316840171814, "learning_rate": 7.534722127456182e-06, "loss": 0.013, "step": 9835 }, { "epoch": 0.6759208356239692, "grad_norm": 0.5126397609710693, "learning_rate": 7.531826579437428e-06, "loss": 0.0124, "step": 9836 }, { "epoch": 0.6759895547003848, "grad_norm": 0.684059739112854, "learning_rate": 7.528931401376645e-06, "loss": 0.0182, "step": 9837 }, { "epoch": 0.6760582737768004, "grad_norm": 0.2334469109773636, "learning_rate": 7.526036593417234e-06, "loss": 0.0152, "step": 9838 }, { "epoch": 0.6761269928532161, "grad_norm": 0.4770699739456177, "learning_rate": 7.52314215570261e-06, "loss": 0.0146, "step": 9839 }, { "epoch": 0.6761957119296317, "grad_norm": 0.5161649584770203, "learning_rate": 7.520248088376147e-06, "loss": 0.0137, "step": 9840 }, { "epoch": 0.6762644310060473, "grad_norm": 0.45688924193382263, "learning_rate": 7.517354391581219e-06, "loss": 0.0123, "step": 9841 }, { "epoch": 0.6763331500824629, "grad_norm": 0.19637589156627655, "learning_rate": 7.514461065461162e-06, "loss": 0.0122, "step": 9842 }, { "epoch": 0.6764018691588785, "grad_norm": 0.21899978816509247, "learning_rate": 7.511568110159325e-06, "loss": 0.0135, "step": 9843 }, { "epoch": 0.6764705882352942, "grad_norm": 0.8876237869262695, "learning_rate": 7.508675525819002e-06, "loss": 0.0223, "step": 9844 }, { "epoch": 0.6765393073117097, "grad_norm": 0.5670899748802185, "learning_rate": 7.5057833125835e-06, "loss": 0.015, "step": 9845 }, { "epoch": 0.6766080263881253, "grad_norm": 0.5619555711746216, "learning_rate": 7.502891470596076e-06, "loss": 0.0111, "step": 9846 }, { "epoch": 0.6766767454645409, "grad_norm": 1.5026049613952637, "learning_rate": 7.500000000000004e-06, "loss": 0.0171, "step": 9847 }, { "epoch": 0.6767454645409565, "grad_norm": 0.058869216591119766, "learning_rate": 7.497108900938504e-06, "loss": 0.006, "step": 9848 }, { "epoch": 0.6768141836173722, "grad_norm": 0.6937903165817261, "learning_rate": 7.4942181735548065e-06, "loss": 0.0132, "step": 9849 }, { "epoch": 0.6768829026937878, "grad_norm": 0.5914576053619385, "learning_rate": 7.491327817992113e-06, "loss": 0.0107, "step": 9850 }, { "epoch": 0.6769516217702034, "grad_norm": 0.6092701554298401, "learning_rate": 7.4884378343935976e-06, "loss": 0.012, "step": 9851 }, { "epoch": 0.677020340846619, "grad_norm": 0.6689354181289673, "learning_rate": 7.485548222902434e-06, "loss": 0.013, "step": 9852 }, { "epoch": 0.6770890599230346, "grad_norm": 1.7911152839660645, "learning_rate": 7.4826589836617575e-06, "loss": 0.0219, "step": 9853 }, { "epoch": 0.6771577789994503, "grad_norm": 0.9766186475753784, "learning_rate": 7.479770116814705e-06, "loss": 0.0112, "step": 9854 }, { "epoch": 0.6772264980758659, "grad_norm": 3.1253550052642822, "learning_rate": 7.476881622504368e-06, "loss": 0.0422, "step": 9855 }, { "epoch": 0.6772952171522815, "grad_norm": 0.23328591883182526, "learning_rate": 7.473993500873861e-06, "loss": 0.0143, "step": 9856 }, { "epoch": 0.6773639362286971, "grad_norm": 1.8017873764038086, "learning_rate": 7.471105752066237e-06, "loss": 0.0102, "step": 9857 }, { "epoch": 0.6774326553051127, "grad_norm": 0.44437217712402344, "learning_rate": 7.468218376224562e-06, "loss": 0.02, "step": 9858 }, { "epoch": 0.6775013743815284, "grad_norm": 0.5813695192337036, "learning_rate": 7.465331373491857e-06, "loss": 0.0167, "step": 9859 }, { "epoch": 0.677570093457944, "grad_norm": 0.6704213619232178, "learning_rate": 7.462444744011152e-06, "loss": 0.0137, "step": 9860 }, { "epoch": 0.6776388125343595, "grad_norm": 0.5111820697784424, "learning_rate": 7.4595584879254246e-06, "loss": 0.0131, "step": 9861 }, { "epoch": 0.6777075316107751, "grad_norm": 0.7870299220085144, "learning_rate": 7.456672605377679e-06, "loss": 0.0156, "step": 9862 }, { "epoch": 0.6777762506871907, "grad_norm": 0.34970664978027344, "learning_rate": 7.45378709651086e-06, "loss": 0.0128, "step": 9863 }, { "epoch": 0.6778449697636064, "grad_norm": 0.7434683442115784, "learning_rate": 7.450901961467918e-06, "loss": 0.0095, "step": 9864 }, { "epoch": 0.677913688840022, "grad_norm": 0.4730492830276489, "learning_rate": 7.448017200391768e-06, "loss": 0.0083, "step": 9865 }, { "epoch": 0.6779824079164376, "grad_norm": 1.226835012435913, "learning_rate": 7.445132813425324e-06, "loss": 0.0159, "step": 9866 }, { "epoch": 0.6780511269928532, "grad_norm": 0.7112879753112793, "learning_rate": 7.442248800711465e-06, "loss": 0.0193, "step": 9867 }, { "epoch": 0.6781198460692688, "grad_norm": 0.34968337416648865, "learning_rate": 7.4393651623930605e-06, "loss": 0.0152, "step": 9868 }, { "epoch": 0.6781885651456845, "grad_norm": 0.8761312961578369, "learning_rate": 7.4364818986129665e-06, "loss": 0.0098, "step": 9869 }, { "epoch": 0.6782572842221001, "grad_norm": 0.34794244170188904, "learning_rate": 7.4335990095140046e-06, "loss": 0.0066, "step": 9870 }, { "epoch": 0.6783260032985157, "grad_norm": 0.4548076391220093, "learning_rate": 7.430716495238999e-06, "loss": 0.0083, "step": 9871 }, { "epoch": 0.6783947223749313, "grad_norm": 0.33256444334983826, "learning_rate": 7.427834355930729e-06, "loss": 0.019, "step": 9872 }, { "epoch": 0.6784634414513468, "grad_norm": 0.47669413685798645, "learning_rate": 7.4249525917319785e-06, "loss": 0.0091, "step": 9873 }, { "epoch": 0.6785321605277626, "grad_norm": 0.2809053659439087, "learning_rate": 7.422071202785501e-06, "loss": 0.0198, "step": 9874 }, { "epoch": 0.6786008796041781, "grad_norm": 0.37077876925468445, "learning_rate": 7.4191901892340444e-06, "loss": 0.0156, "step": 9875 }, { "epoch": 0.6786695986805937, "grad_norm": 0.6797707080841064, "learning_rate": 7.416309551220314e-06, "loss": 0.0134, "step": 9876 }, { "epoch": 0.6787383177570093, "grad_norm": 0.3471727669239044, "learning_rate": 7.4134292888870235e-06, "loss": 0.0085, "step": 9877 }, { "epoch": 0.6788070368334249, "grad_norm": 0.5590721964836121, "learning_rate": 7.4105494023768414e-06, "loss": 0.013, "step": 9878 }, { "epoch": 0.6788757559098406, "grad_norm": 0.43274015188217163, "learning_rate": 7.407669891832439e-06, "loss": 0.0116, "step": 9879 }, { "epoch": 0.6789444749862562, "grad_norm": 0.419630765914917, "learning_rate": 7.4047907573964615e-06, "loss": 0.0097, "step": 9880 }, { "epoch": 0.6790131940626718, "grad_norm": 0.29190438985824585, "learning_rate": 7.401911999211542e-06, "loss": 0.0117, "step": 9881 }, { "epoch": 0.6790819131390874, "grad_norm": 0.33449289202690125, "learning_rate": 7.3990336174202724e-06, "loss": 0.007, "step": 9882 }, { "epoch": 0.679150632215503, "grad_norm": 0.2599082887172699, "learning_rate": 7.396155612165257e-06, "loss": 0.0064, "step": 9883 }, { "epoch": 0.6792193512919187, "grad_norm": 0.2630963921546936, "learning_rate": 7.393277983589055e-06, "loss": 0.0067, "step": 9884 }, { "epoch": 0.6792880703683343, "grad_norm": 0.22611111402511597, "learning_rate": 7.390400731834221e-06, "loss": 0.0122, "step": 9885 }, { "epoch": 0.6793567894447499, "grad_norm": 0.5934851169586182, "learning_rate": 7.387523857043298e-06, "loss": 0.0204, "step": 9886 }, { "epoch": 0.6794255085211655, "grad_norm": 0.6482414603233337, "learning_rate": 7.384647359358786e-06, "loss": 0.0176, "step": 9887 }, { "epoch": 0.679494227597581, "grad_norm": 0.48545560240745544, "learning_rate": 7.381771238923193e-06, "loss": 0.0131, "step": 9888 }, { "epoch": 0.6795629466739967, "grad_norm": 0.6097899079322815, "learning_rate": 7.378895495878985e-06, "loss": 0.0156, "step": 9889 }, { "epoch": 0.6796316657504123, "grad_norm": 0.45884594321250916, "learning_rate": 7.376020130368631e-06, "loss": 0.0103, "step": 9890 }, { "epoch": 0.6797003848268279, "grad_norm": 0.4056099057197571, "learning_rate": 7.373145142534555e-06, "loss": 0.008, "step": 9891 }, { "epoch": 0.6797691039032435, "grad_norm": 0.4139333665370941, "learning_rate": 7.3702705325192e-06, "loss": 0.0072, "step": 9892 }, { "epoch": 0.6798378229796591, "grad_norm": 0.6619556546211243, "learning_rate": 7.367396300464949e-06, "loss": 0.0171, "step": 9893 }, { "epoch": 0.6799065420560748, "grad_norm": 0.27625343203544617, "learning_rate": 7.3645224465142005e-06, "loss": 0.0074, "step": 9894 }, { "epoch": 0.6799752611324904, "grad_norm": 0.31757235527038574, "learning_rate": 7.361648970809307e-06, "loss": 0.0122, "step": 9895 }, { "epoch": 0.680043980208906, "grad_norm": 0.5969223380088806, "learning_rate": 7.358775873492624e-06, "loss": 0.0103, "step": 9896 }, { "epoch": 0.6801126992853216, "grad_norm": 0.09898697584867477, "learning_rate": 7.355903154706465e-06, "loss": 0.0098, "step": 9897 }, { "epoch": 0.6801814183617372, "grad_norm": 0.3478938639163971, "learning_rate": 7.353030814593157e-06, "loss": 0.0159, "step": 9898 }, { "epoch": 0.6802501374381529, "grad_norm": 0.48519471287727356, "learning_rate": 7.350158853294977e-06, "loss": 0.0141, "step": 9899 }, { "epoch": 0.6803188565145685, "grad_norm": 0.26935917139053345, "learning_rate": 7.347287270954205e-06, "loss": 0.0101, "step": 9900 }, { "epoch": 0.6803875755909841, "grad_norm": 0.3881343901157379, "learning_rate": 7.3444160677130835e-06, "loss": 0.0097, "step": 9901 }, { "epoch": 0.6804562946673997, "grad_norm": 0.25015172362327576, "learning_rate": 7.341545243713855e-06, "loss": 0.0132, "step": 9902 }, { "epoch": 0.6805250137438152, "grad_norm": 0.5083296895027161, "learning_rate": 7.338674799098725e-06, "loss": 0.0113, "step": 9903 }, { "epoch": 0.680593732820231, "grad_norm": 0.20026838779449463, "learning_rate": 7.335804734009895e-06, "loss": 0.0108, "step": 9904 }, { "epoch": 0.6806624518966465, "grad_norm": 0.5929752588272095, "learning_rate": 7.332935048589547e-06, "loss": 0.0117, "step": 9905 }, { "epoch": 0.6807311709730621, "grad_norm": 0.4186732769012451, "learning_rate": 7.330065742979827e-06, "loss": 0.0076, "step": 9906 }, { "epoch": 0.6807998900494777, "grad_norm": 0.4277409315109253, "learning_rate": 7.327196817322889e-06, "loss": 0.0116, "step": 9907 }, { "epoch": 0.6808686091258933, "grad_norm": 0.29384738206863403, "learning_rate": 7.324328271760839e-06, "loss": 0.0123, "step": 9908 }, { "epoch": 0.680937328202309, "grad_norm": 0.13674752414226532, "learning_rate": 7.321460106435786e-06, "loss": 0.0168, "step": 9909 }, { "epoch": 0.6810060472787246, "grad_norm": 0.8301714658737183, "learning_rate": 7.318592321489815e-06, "loss": 0.014, "step": 9910 }, { "epoch": 0.6810747663551402, "grad_norm": 0.2702575623989105, "learning_rate": 7.315724917064992e-06, "loss": 0.0045, "step": 9911 }, { "epoch": 0.6811434854315558, "grad_norm": 0.26830342411994934, "learning_rate": 7.3128578933033525e-06, "loss": 0.0075, "step": 9912 }, { "epoch": 0.6812122045079714, "grad_norm": 0.9234970211982727, "learning_rate": 7.309991250346934e-06, "loss": 0.0103, "step": 9913 }, { "epoch": 0.6812809235843871, "grad_norm": 0.30789414048194885, "learning_rate": 7.307124988337734e-06, "loss": 0.0147, "step": 9914 }, { "epoch": 0.6813496426608027, "grad_norm": 1.0714313983917236, "learning_rate": 7.304259107417747e-06, "loss": 0.0204, "step": 9915 }, { "epoch": 0.6814183617372183, "grad_norm": 0.5263210535049438, "learning_rate": 7.301393607728941e-06, "loss": 0.0097, "step": 9916 }, { "epoch": 0.6814870808136338, "grad_norm": 0.3799906075000763, "learning_rate": 7.298528489413274e-06, "loss": 0.0135, "step": 9917 }, { "epoch": 0.6815557998900494, "grad_norm": 0.08019450306892395, "learning_rate": 7.2956637526126675e-06, "loss": 0.0172, "step": 9918 }, { "epoch": 0.6816245189664651, "grad_norm": 0.5317438840866089, "learning_rate": 7.292799397469043e-06, "loss": 0.0048, "step": 9919 }, { "epoch": 0.6816932380428807, "grad_norm": 0.15159879624843597, "learning_rate": 7.289935424124287e-06, "loss": 0.0113, "step": 9920 }, { "epoch": 0.6817619571192963, "grad_norm": 0.3116050958633423, "learning_rate": 7.287071832720277e-06, "loss": 0.0098, "step": 9921 }, { "epoch": 0.6818306761957119, "grad_norm": 0.35027778148651123, "learning_rate": 7.284208623398879e-06, "loss": 0.0121, "step": 9922 }, { "epoch": 0.6818993952721275, "grad_norm": 0.7176771759986877, "learning_rate": 7.281345796301915e-06, "loss": 0.0075, "step": 9923 }, { "epoch": 0.6819681143485432, "grad_norm": 0.36392122507095337, "learning_rate": 7.278483351571219e-06, "loss": 0.0081, "step": 9924 }, { "epoch": 0.6820368334249588, "grad_norm": 0.8207331299781799, "learning_rate": 7.275621289348576e-06, "loss": 0.0198, "step": 9925 }, { "epoch": 0.6821055525013744, "grad_norm": 0.19391880929470062, "learning_rate": 7.27275960977578e-06, "loss": 0.0104, "step": 9926 }, { "epoch": 0.68217427157779, "grad_norm": 0.13717958331108093, "learning_rate": 7.2698983129945745e-06, "loss": 0.0162, "step": 9927 }, { "epoch": 0.6822429906542056, "grad_norm": 0.5180914402008057, "learning_rate": 7.267037399146727e-06, "loss": 0.0117, "step": 9928 }, { "epoch": 0.6823117097306213, "grad_norm": 0.4145212769508362, "learning_rate": 7.2641768683739415e-06, "loss": 0.011, "step": 9929 }, { "epoch": 0.6823804288070369, "grad_norm": 1.1886887550354004, "learning_rate": 7.261316720817936e-06, "loss": 0.0144, "step": 9930 }, { "epoch": 0.6824491478834525, "grad_norm": 0.730144739151001, "learning_rate": 7.258456956620383e-06, "loss": 0.0113, "step": 9931 }, { "epoch": 0.682517866959868, "grad_norm": 0.3008885085582733, "learning_rate": 7.255597575922963e-06, "loss": 0.009, "step": 9932 }, { "epoch": 0.6825865860362836, "grad_norm": 0.27037516236305237, "learning_rate": 7.252738578867306e-06, "loss": 0.0121, "step": 9933 }, { "epoch": 0.6826553051126993, "grad_norm": 0.5669350624084473, "learning_rate": 7.249879965595063e-06, "loss": 0.0122, "step": 9934 }, { "epoch": 0.6827240241891149, "grad_norm": 0.44643837213516235, "learning_rate": 7.2470217362478294e-06, "loss": 0.0056, "step": 9935 }, { "epoch": 0.6827927432655305, "grad_norm": 0.6233859658241272, "learning_rate": 7.244163890967204e-06, "loss": 0.0153, "step": 9936 }, { "epoch": 0.6828614623419461, "grad_norm": 0.49950146675109863, "learning_rate": 7.241306429894747e-06, "loss": 0.0124, "step": 9937 }, { "epoch": 0.6829301814183617, "grad_norm": 0.299837589263916, "learning_rate": 7.238449353172025e-06, "loss": 0.0126, "step": 9938 }, { "epoch": 0.6829989004947774, "grad_norm": 0.38669124245643616, "learning_rate": 7.23559266094056e-06, "loss": 0.0118, "step": 9939 }, { "epoch": 0.683067619571193, "grad_norm": 0.7076166868209839, "learning_rate": 7.232736353341871e-06, "loss": 0.0182, "step": 9940 }, { "epoch": 0.6831363386476086, "grad_norm": 0.9599664807319641, "learning_rate": 7.22988043051746e-06, "loss": 0.0123, "step": 9941 }, { "epoch": 0.6832050577240242, "grad_norm": 0.30905628204345703, "learning_rate": 7.227024892608791e-06, "loss": 0.0111, "step": 9942 }, { "epoch": 0.6832737768004398, "grad_norm": 0.9027562737464905, "learning_rate": 7.224169739757335e-06, "loss": 0.0165, "step": 9943 }, { "epoch": 0.6833424958768554, "grad_norm": 0.537750244140625, "learning_rate": 7.221314972104519e-06, "loss": 0.01, "step": 9944 }, { "epoch": 0.6834112149532711, "grad_norm": 0.31681522727012634, "learning_rate": 7.218460589791766e-06, "loss": 0.0136, "step": 9945 }, { "epoch": 0.6834799340296867, "grad_norm": 0.3344227969646454, "learning_rate": 7.215606592960476e-06, "loss": 0.0118, "step": 9946 }, { "epoch": 0.6835486531061022, "grad_norm": 0.8383356928825378, "learning_rate": 7.2127529817520384e-06, "loss": 0.0118, "step": 9947 }, { "epoch": 0.6836173721825178, "grad_norm": 0.3039382994174957, "learning_rate": 7.209899756307803e-06, "loss": 0.0221, "step": 9948 }, { "epoch": 0.6836860912589334, "grad_norm": 0.21739411354064941, "learning_rate": 7.207046916769122e-06, "loss": 0.0062, "step": 9949 }, { "epoch": 0.6837548103353491, "grad_norm": 0.5547558069229126, "learning_rate": 7.2041944632773105e-06, "loss": 0.0102, "step": 9950 }, { "epoch": 0.6838235294117647, "grad_norm": 0.7637404799461365, "learning_rate": 7.2013423959736785e-06, "loss": 0.0117, "step": 9951 }, { "epoch": 0.6838922484881803, "grad_norm": 0.3084503710269928, "learning_rate": 7.198490714999508e-06, "loss": 0.0134, "step": 9952 }, { "epoch": 0.6839609675645959, "grad_norm": 0.5691920518875122, "learning_rate": 7.195639420496077e-06, "loss": 0.0099, "step": 9953 }, { "epoch": 0.6840296866410115, "grad_norm": 0.31276723742485046, "learning_rate": 7.192788512604618e-06, "loss": 0.0044, "step": 9954 }, { "epoch": 0.6840984057174272, "grad_norm": 0.39120030403137207, "learning_rate": 7.18993799146637e-06, "loss": 0.0118, "step": 9955 }, { "epoch": 0.6841671247938428, "grad_norm": 0.4373928904533386, "learning_rate": 7.1870878572225305e-06, "loss": 0.0099, "step": 9956 }, { "epoch": 0.6842358438702584, "grad_norm": 0.42725202441215515, "learning_rate": 7.184238110014297e-06, "loss": 0.0143, "step": 9957 }, { "epoch": 0.684304562946674, "grad_norm": 0.15428973734378815, "learning_rate": 7.181388749982844e-06, "loss": 0.0116, "step": 9958 }, { "epoch": 0.6843732820230896, "grad_norm": 0.45556971430778503, "learning_rate": 7.178539777269312e-06, "loss": 0.0089, "step": 9959 }, { "epoch": 0.6844420010995053, "grad_norm": 0.5129189491271973, "learning_rate": 7.175691192014846e-06, "loss": 0.0095, "step": 9960 }, { "epoch": 0.6845107201759209, "grad_norm": 0.4391069710254669, "learning_rate": 7.172842994360546e-06, "loss": 0.0086, "step": 9961 }, { "epoch": 0.6845794392523364, "grad_norm": 0.24696077406406403, "learning_rate": 7.169995184447518e-06, "loss": 0.013, "step": 9962 }, { "epoch": 0.684648158328752, "grad_norm": 0.29379284381866455, "learning_rate": 7.1671477624168205e-06, "loss": 0.0087, "step": 9963 }, { "epoch": 0.6847168774051676, "grad_norm": 0.3990410566329956, "learning_rate": 7.164300728409529e-06, "loss": 0.0136, "step": 9964 }, { "epoch": 0.6847855964815833, "grad_norm": 0.6142302751541138, "learning_rate": 7.161454082566665e-06, "loss": 0.0104, "step": 9965 }, { "epoch": 0.6848543155579989, "grad_norm": 0.28991052508354187, "learning_rate": 7.158607825029255e-06, "loss": 0.0128, "step": 9966 }, { "epoch": 0.6849230346344145, "grad_norm": 0.4058040976524353, "learning_rate": 7.1557619559382875e-06, "loss": 0.0071, "step": 9967 }, { "epoch": 0.6849917537108301, "grad_norm": 0.43250754475593567, "learning_rate": 7.152916475434751e-06, "loss": 0.0143, "step": 9968 }, { "epoch": 0.6850604727872457, "grad_norm": 1.3237091302871704, "learning_rate": 7.1500713836595875e-06, "loss": 0.0149, "step": 9969 }, { "epoch": 0.6851291918636614, "grad_norm": 0.5104677081108093, "learning_rate": 7.147226680753762e-06, "loss": 0.0193, "step": 9970 }, { "epoch": 0.685197910940077, "grad_norm": 0.43392595648765564, "learning_rate": 7.1443823668581735e-06, "loss": 0.0119, "step": 9971 }, { "epoch": 0.6852666300164926, "grad_norm": 0.719843864440918, "learning_rate": 7.14153844211374e-06, "loss": 0.0147, "step": 9972 }, { "epoch": 0.6853353490929082, "grad_norm": 0.6170533895492554, "learning_rate": 7.1386949066613285e-06, "loss": 0.0096, "step": 9973 }, { "epoch": 0.6854040681693238, "grad_norm": 0.50593501329422, "learning_rate": 7.135851760641816e-06, "loss": 0.0196, "step": 9974 }, { "epoch": 0.6854727872457395, "grad_norm": 0.1615656614303589, "learning_rate": 7.133009004196034e-06, "loss": 0.0166, "step": 9975 }, { "epoch": 0.685541506322155, "grad_norm": 0.37284770607948303, "learning_rate": 7.130166637464811e-06, "loss": 0.0137, "step": 9976 }, { "epoch": 0.6856102253985706, "grad_norm": 0.2833777070045471, "learning_rate": 7.127324660588961e-06, "loss": 0.0068, "step": 9977 }, { "epoch": 0.6856789444749862, "grad_norm": 0.6486603021621704, "learning_rate": 7.124483073709255e-06, "loss": 0.0096, "step": 9978 }, { "epoch": 0.6857476635514018, "grad_norm": 0.4994679391384125, "learning_rate": 7.121641876966463e-06, "loss": 0.0108, "step": 9979 }, { "epoch": 0.6858163826278175, "grad_norm": 0.6213764548301697, "learning_rate": 7.1188010705013455e-06, "loss": 0.0122, "step": 9980 }, { "epoch": 0.6858851017042331, "grad_norm": 0.494913250207901, "learning_rate": 7.115960654454611e-06, "loss": 0.0127, "step": 9981 }, { "epoch": 0.6859538207806487, "grad_norm": 0.23089183866977692, "learning_rate": 7.113120628966978e-06, "loss": 0.006, "step": 9982 }, { "epoch": 0.6860225398570643, "grad_norm": 0.15092454850673676, "learning_rate": 7.11028099417914e-06, "loss": 0.0107, "step": 9983 }, { "epoch": 0.6860912589334799, "grad_norm": 0.9771289229393005, "learning_rate": 7.107441750231755e-06, "loss": 0.0136, "step": 9984 }, { "epoch": 0.6861599780098956, "grad_norm": 0.35247886180877686, "learning_rate": 7.104602897265485e-06, "loss": 0.0061, "step": 9985 }, { "epoch": 0.6862286970863112, "grad_norm": 0.38487404584884644, "learning_rate": 7.101764435420949e-06, "loss": 0.009, "step": 9986 }, { "epoch": 0.6862974161627268, "grad_norm": 0.217181995511055, "learning_rate": 7.098926364838766e-06, "loss": 0.0088, "step": 9987 }, { "epoch": 0.6863661352391424, "grad_norm": 0.9401357769966125, "learning_rate": 7.096088685659528e-06, "loss": 0.0113, "step": 9988 }, { "epoch": 0.686434854315558, "grad_norm": 0.6023740768432617, "learning_rate": 7.09325139802381e-06, "loss": 0.0062, "step": 9989 }, { "epoch": 0.6865035733919737, "grad_norm": 0.5748042464256287, "learning_rate": 7.090414502072158e-06, "loss": 0.018, "step": 9990 }, { "epoch": 0.6865722924683892, "grad_norm": 0.25158876180648804, "learning_rate": 7.087577997945115e-06, "loss": 0.0118, "step": 9991 }, { "epoch": 0.6866410115448048, "grad_norm": 0.642560601234436, "learning_rate": 7.084741885783184e-06, "loss": 0.0106, "step": 9992 }, { "epoch": 0.6867097306212204, "grad_norm": 1.0907522439956665, "learning_rate": 7.08190616572687e-06, "loss": 0.0145, "step": 9993 }, { "epoch": 0.686778449697636, "grad_norm": 0.2940458059310913, "learning_rate": 7.079070837916649e-06, "loss": 0.0142, "step": 9994 }, { "epoch": 0.6868471687740517, "grad_norm": 0.683525562286377, "learning_rate": 7.076235902492968e-06, "loss": 0.0132, "step": 9995 }, { "epoch": 0.6869158878504673, "grad_norm": 0.551711916923523, "learning_rate": 7.073401359596271e-06, "loss": 0.0171, "step": 9996 }, { "epoch": 0.6869846069268829, "grad_norm": 0.21257805824279785, "learning_rate": 7.070567209366979e-06, "loss": 0.0106, "step": 9997 }, { "epoch": 0.6870533260032985, "grad_norm": 0.6125190258026123, "learning_rate": 7.06773345194548e-06, "loss": 0.0147, "step": 9998 }, { "epoch": 0.6871220450797141, "grad_norm": 0.37945395708084106, "learning_rate": 7.064900087472159e-06, "loss": 0.0075, "step": 9999 }, { "epoch": 0.6871907641561298, "grad_norm": 0.462520569562912, "learning_rate": 7.062067116087378e-06, "loss": 0.0094, "step": 10000 }, { "epoch": 0.6872594832325454, "grad_norm": 0.35426944494247437, "learning_rate": 7.059234537931466e-06, "loss": 0.0207, "step": 10001 }, { "epoch": 0.687328202308961, "grad_norm": 0.26105406880378723, "learning_rate": 7.056402353144756e-06, "loss": 0.0105, "step": 10002 }, { "epoch": 0.6873969213853766, "grad_norm": 0.427939236164093, "learning_rate": 7.053570561867537e-06, "loss": 0.0156, "step": 10003 }, { "epoch": 0.6874656404617921, "grad_norm": 0.22567395865917206, "learning_rate": 7.0507391642401e-06, "loss": 0.0115, "step": 10004 }, { "epoch": 0.6875343595382079, "grad_norm": 0.863467276096344, "learning_rate": 7.047908160402691e-06, "loss": 0.0193, "step": 10005 }, { "epoch": 0.6876030786146234, "grad_norm": 0.5436360239982605, "learning_rate": 7.045077550495574e-06, "loss": 0.0155, "step": 10006 }, { "epoch": 0.687671797691039, "grad_norm": 0.9585868716239929, "learning_rate": 7.0422473346589544e-06, "loss": 0.0132, "step": 10007 }, { "epoch": 0.6877405167674546, "grad_norm": 0.42125269770622253, "learning_rate": 7.039417513033047e-06, "loss": 0.011, "step": 10008 }, { "epoch": 0.6878092358438702, "grad_norm": 0.3011593818664551, "learning_rate": 7.0365880857580254e-06, "loss": 0.0133, "step": 10009 }, { "epoch": 0.6878779549202859, "grad_norm": 0.6578947305679321, "learning_rate": 7.033759052974064e-06, "loss": 0.0121, "step": 10010 }, { "epoch": 0.6879466739967015, "grad_norm": 0.3577144742012024, "learning_rate": 7.0309304148212945e-06, "loss": 0.009, "step": 10011 }, { "epoch": 0.6880153930731171, "grad_norm": 0.6297920346260071, "learning_rate": 7.028102171439847e-06, "loss": 0.0122, "step": 10012 }, { "epoch": 0.6880841121495327, "grad_norm": 0.47985008358955383, "learning_rate": 7.025274322969831e-06, "loss": 0.0111, "step": 10013 }, { "epoch": 0.6881528312259483, "grad_norm": 0.48247143626213074, "learning_rate": 7.022446869551334e-06, "loss": 0.0132, "step": 10014 }, { "epoch": 0.688221550302364, "grad_norm": 0.46835821866989136, "learning_rate": 7.019619811324413e-06, "loss": 0.0065, "step": 10015 }, { "epoch": 0.6882902693787796, "grad_norm": 0.7612631320953369, "learning_rate": 7.016793148429125e-06, "loss": 0.0158, "step": 10016 }, { "epoch": 0.6883589884551952, "grad_norm": 0.4011608362197876, "learning_rate": 7.013966881005485e-06, "loss": 0.0076, "step": 10017 }, { "epoch": 0.6884277075316108, "grad_norm": 0.5241909027099609, "learning_rate": 7.011141009193508e-06, "loss": 0.0127, "step": 10018 }, { "epoch": 0.6884964266080263, "grad_norm": 0.8432974219322205, "learning_rate": 7.008315533133186e-06, "loss": 0.0119, "step": 10019 }, { "epoch": 0.688565145684442, "grad_norm": 0.5527579188346863, "learning_rate": 7.005490452964477e-06, "loss": 0.0172, "step": 10020 }, { "epoch": 0.6886338647608576, "grad_norm": 0.4586922824382782, "learning_rate": 7.002665768827341e-06, "loss": 0.01, "step": 10021 }, { "epoch": 0.6887025838372732, "grad_norm": 0.27107173204421997, "learning_rate": 6.999841480861696e-06, "loss": 0.0147, "step": 10022 }, { "epoch": 0.6887713029136888, "grad_norm": 0.5101038217544556, "learning_rate": 6.997017589207455e-06, "loss": 0.0155, "step": 10023 }, { "epoch": 0.6888400219901044, "grad_norm": 0.5456153750419617, "learning_rate": 6.9941940940045094e-06, "loss": 0.0088, "step": 10024 }, { "epoch": 0.6889087410665201, "grad_norm": 0.6427392959594727, "learning_rate": 6.991370995392737e-06, "loss": 0.0176, "step": 10025 }, { "epoch": 0.6889774601429357, "grad_norm": 0.23919837176799774, "learning_rate": 6.988548293511971e-06, "loss": 0.011, "step": 10026 }, { "epoch": 0.6890461792193513, "grad_norm": 0.12775996327400208, "learning_rate": 6.98572598850206e-06, "loss": 0.007, "step": 10027 }, { "epoch": 0.6891148982957669, "grad_norm": 0.3700442612171173, "learning_rate": 6.9829040805028005e-06, "loss": 0.0111, "step": 10028 }, { "epoch": 0.6891836173721825, "grad_norm": 0.6269422173500061, "learning_rate": 6.98008256965399e-06, "loss": 0.0133, "step": 10029 }, { "epoch": 0.6892523364485982, "grad_norm": 0.37617939710617065, "learning_rate": 6.977261456095401e-06, "loss": 0.0072, "step": 10030 }, { "epoch": 0.6893210555250138, "grad_norm": 0.2561146914958954, "learning_rate": 6.974440739966791e-06, "loss": 0.0273, "step": 10031 }, { "epoch": 0.6893897746014294, "grad_norm": 0.4634717106819153, "learning_rate": 6.9716204214078825e-06, "loss": 0.0105, "step": 10032 }, { "epoch": 0.689458493677845, "grad_norm": 0.1243944764137268, "learning_rate": 6.968800500558397e-06, "loss": 0.0072, "step": 10033 }, { "epoch": 0.6895272127542605, "grad_norm": 0.4697626829147339, "learning_rate": 6.9659809775580175e-06, "loss": 0.0107, "step": 10034 }, { "epoch": 0.6895959318306762, "grad_norm": 0.3389628827571869, "learning_rate": 6.963161852546423e-06, "loss": 0.0155, "step": 10035 }, { "epoch": 0.6896646509070918, "grad_norm": 0.9809905886650085, "learning_rate": 6.960343125663274e-06, "loss": 0.0131, "step": 10036 }, { "epoch": 0.6897333699835074, "grad_norm": 0.17782524228096008, "learning_rate": 6.957524797048191e-06, "loss": 0.0099, "step": 10037 }, { "epoch": 0.689802089059923, "grad_norm": 0.15689964592456818, "learning_rate": 6.954706866840801e-06, "loss": 0.0049, "step": 10038 }, { "epoch": 0.6898708081363386, "grad_norm": 0.36766183376312256, "learning_rate": 6.951889335180688e-06, "loss": 0.0072, "step": 10039 }, { "epoch": 0.6899395272127543, "grad_norm": 0.25361019372940063, "learning_rate": 6.949072202207436e-06, "loss": 0.0145, "step": 10040 }, { "epoch": 0.6900082462891699, "grad_norm": 0.9054688811302185, "learning_rate": 6.946255468060582e-06, "loss": 0.0213, "step": 10041 }, { "epoch": 0.6900769653655855, "grad_norm": 0.4101465940475464, "learning_rate": 6.943439132879687e-06, "loss": 0.01, "step": 10042 }, { "epoch": 0.6901456844420011, "grad_norm": 0.25588908791542053, "learning_rate": 6.940623196804248e-06, "loss": 0.014, "step": 10043 }, { "epoch": 0.6902144035184167, "grad_norm": 0.2826049029827118, "learning_rate": 6.937807659973773e-06, "loss": 0.0114, "step": 10044 }, { "epoch": 0.6902831225948324, "grad_norm": 0.4198368787765503, "learning_rate": 6.934992522527725e-06, "loss": 0.0115, "step": 10045 }, { "epoch": 0.690351841671248, "grad_norm": 0.47537946701049805, "learning_rate": 6.932177784605572e-06, "loss": 0.014, "step": 10046 }, { "epoch": 0.6904205607476636, "grad_norm": 0.1806497871875763, "learning_rate": 6.929363446346733e-06, "loss": 0.0087, "step": 10047 }, { "epoch": 0.6904892798240791, "grad_norm": 0.3217126131057739, "learning_rate": 6.926549507890649e-06, "loss": 0.016, "step": 10048 }, { "epoch": 0.6905579989004947, "grad_norm": 0.3345690667629242, "learning_rate": 6.923735969376697e-06, "loss": 0.016, "step": 10049 }, { "epoch": 0.6906267179769104, "grad_norm": 0.3128930628299713, "learning_rate": 6.9209228309442665e-06, "loss": 0.0105, "step": 10050 }, { "epoch": 0.690695437053326, "grad_norm": 0.19197188317775726, "learning_rate": 6.918110092732704e-06, "loss": 0.0095, "step": 10051 }, { "epoch": 0.6907641561297416, "grad_norm": 0.4305137097835541, "learning_rate": 6.915297754881359e-06, "loss": 0.0077, "step": 10052 }, { "epoch": 0.6908328752061572, "grad_norm": 0.5641666054725647, "learning_rate": 6.912485817529534e-06, "loss": 0.0277, "step": 10053 }, { "epoch": 0.6909015942825728, "grad_norm": 0.41001442074775696, "learning_rate": 6.9096742808165355e-06, "loss": 0.0118, "step": 10054 }, { "epoch": 0.6909703133589885, "grad_norm": 0.5250855088233948, "learning_rate": 6.906863144881646e-06, "loss": 0.0123, "step": 10055 }, { "epoch": 0.6910390324354041, "grad_norm": 0.19185657799243927, "learning_rate": 6.904052409864111e-06, "loss": 0.0076, "step": 10056 }, { "epoch": 0.6911077515118197, "grad_norm": 1.2027596235275269, "learning_rate": 6.901242075903181e-06, "loss": 0.0113, "step": 10057 }, { "epoch": 0.6911764705882353, "grad_norm": 0.27031952142715454, "learning_rate": 6.898432143138062e-06, "loss": 0.0116, "step": 10058 }, { "epoch": 0.6912451896646509, "grad_norm": 0.3130469024181366, "learning_rate": 6.89562261170796e-06, "loss": 0.0086, "step": 10059 }, { "epoch": 0.6913139087410666, "grad_norm": 0.6376173496246338, "learning_rate": 6.8928134817520505e-06, "loss": 0.0184, "step": 10060 }, { "epoch": 0.6913826278174822, "grad_norm": 0.40285515785217285, "learning_rate": 6.8900047534095005e-06, "loss": 0.0095, "step": 10061 }, { "epoch": 0.6914513468938978, "grad_norm": 1.150472640991211, "learning_rate": 6.887196426819434e-06, "loss": 0.0201, "step": 10062 }, { "epoch": 0.6915200659703133, "grad_norm": 0.23224429786205292, "learning_rate": 6.884388502120983e-06, "loss": 0.0096, "step": 10063 }, { "epoch": 0.6915887850467289, "grad_norm": 0.7479248046875, "learning_rate": 6.881580979453236e-06, "loss": 0.0182, "step": 10064 }, { "epoch": 0.6916575041231446, "grad_norm": 0.842423677444458, "learning_rate": 6.878773858955279e-06, "loss": 0.02, "step": 10065 }, { "epoch": 0.6917262231995602, "grad_norm": 0.24753624200820923, "learning_rate": 6.875967140766165e-06, "loss": 0.0105, "step": 10066 }, { "epoch": 0.6917949422759758, "grad_norm": 0.1620262712240219, "learning_rate": 6.873160825024944e-06, "loss": 0.0079, "step": 10067 }, { "epoch": 0.6918636613523914, "grad_norm": 0.24571593105793, "learning_rate": 6.87035491187062e-06, "loss": 0.0109, "step": 10068 }, { "epoch": 0.691932380428807, "grad_norm": 0.4544217586517334, "learning_rate": 6.867549401442207e-06, "loss": 0.0113, "step": 10069 }, { "epoch": 0.6920010995052226, "grad_norm": 0.6069482564926147, "learning_rate": 6.8647442938786715e-06, "loss": 0.024, "step": 10070 }, { "epoch": 0.6920698185816383, "grad_norm": 0.47602853178977966, "learning_rate": 6.861939589318976e-06, "loss": 0.0069, "step": 10071 }, { "epoch": 0.6921385376580539, "grad_norm": 0.2433076649904251, "learning_rate": 6.8591352879020705e-06, "loss": 0.0113, "step": 10072 }, { "epoch": 0.6922072567344695, "grad_norm": 0.45193588733673096, "learning_rate": 6.856331389766859e-06, "loss": 0.0146, "step": 10073 }, { "epoch": 0.6922759758108851, "grad_norm": 0.8771893382072449, "learning_rate": 6.853527895052255e-06, "loss": 0.0151, "step": 10074 }, { "epoch": 0.6923446948873007, "grad_norm": 0.3211638927459717, "learning_rate": 6.850724803897123e-06, "loss": 0.0146, "step": 10075 }, { "epoch": 0.6924134139637164, "grad_norm": 0.7415811419487, "learning_rate": 6.847922116440337e-06, "loss": 0.0146, "step": 10076 }, { "epoch": 0.692482133040132, "grad_norm": 0.6465764045715332, "learning_rate": 6.845119832820722e-06, "loss": 0.0079, "step": 10077 }, { "epoch": 0.6925508521165475, "grad_norm": 0.16188140213489532, "learning_rate": 6.842317953177106e-06, "loss": 0.0065, "step": 10078 }, { "epoch": 0.6926195711929631, "grad_norm": 0.3615105450153351, "learning_rate": 6.8395164776482875e-06, "loss": 0.0083, "step": 10079 }, { "epoch": 0.6926882902693787, "grad_norm": 0.49733051657676697, "learning_rate": 6.8367154063730506e-06, "loss": 0.0081, "step": 10080 }, { "epoch": 0.6927570093457944, "grad_norm": 0.3731285333633423, "learning_rate": 6.833914739490145e-06, "loss": 0.0071, "step": 10081 }, { "epoch": 0.69282572842221, "grad_norm": 0.40496569871902466, "learning_rate": 6.8311144771383175e-06, "loss": 0.0077, "step": 10082 }, { "epoch": 0.6928944474986256, "grad_norm": 0.3492131233215332, "learning_rate": 6.82831461945628e-06, "loss": 0.017, "step": 10083 }, { "epoch": 0.6929631665750412, "grad_norm": 0.3052932918071747, "learning_rate": 6.825515166582738e-06, "loss": 0.0089, "step": 10084 }, { "epoch": 0.6930318856514568, "grad_norm": 0.6681231260299683, "learning_rate": 6.822716118656366e-06, "loss": 0.0197, "step": 10085 }, { "epoch": 0.6931006047278725, "grad_norm": 0.5146694183349609, "learning_rate": 6.819917475815832e-06, "loss": 0.0123, "step": 10086 }, { "epoch": 0.6931693238042881, "grad_norm": 0.6121962070465088, "learning_rate": 6.817119238199763e-06, "loss": 0.0148, "step": 10087 }, { "epoch": 0.6932380428807037, "grad_norm": 0.29850664734840393, "learning_rate": 6.8143214059467895e-06, "loss": 0.0073, "step": 10088 }, { "epoch": 0.6933067619571193, "grad_norm": 0.17849640548229218, "learning_rate": 6.811523979195498e-06, "loss": 0.0097, "step": 10089 }, { "epoch": 0.6933754810335349, "grad_norm": 0.3257967233657837, "learning_rate": 6.808726958084476e-06, "loss": 0.0186, "step": 10090 }, { "epoch": 0.6934442001099506, "grad_norm": 0.18308769166469574, "learning_rate": 6.8059303427522835e-06, "loss": 0.0086, "step": 10091 }, { "epoch": 0.6935129191863662, "grad_norm": 0.15995058417320251, "learning_rate": 6.8031341333374505e-06, "loss": 0.0155, "step": 10092 }, { "epoch": 0.6935816382627817, "grad_norm": 0.36625853180885315, "learning_rate": 6.8003383299785055e-06, "loss": 0.0141, "step": 10093 }, { "epoch": 0.6936503573391973, "grad_norm": 0.5105263590812683, "learning_rate": 6.797542932813937e-06, "loss": 0.0384, "step": 10094 }, { "epoch": 0.6937190764156129, "grad_norm": 0.639045000076294, "learning_rate": 6.7947479419822335e-06, "loss": 0.0133, "step": 10095 }, { "epoch": 0.6937877954920286, "grad_norm": 0.48926618695259094, "learning_rate": 6.791953357621837e-06, "loss": 0.0217, "step": 10096 }, { "epoch": 0.6938565145684442, "grad_norm": 0.2783852815628052, "learning_rate": 6.789159179871208e-06, "loss": 0.0061, "step": 10097 }, { "epoch": 0.6939252336448598, "grad_norm": 0.3443451225757599, "learning_rate": 6.786365408868746e-06, "loss": 0.0118, "step": 10098 }, { "epoch": 0.6939939527212754, "grad_norm": 0.2766292691230774, "learning_rate": 6.783572044752859e-06, "loss": 0.0212, "step": 10099 }, { "epoch": 0.694062671797691, "grad_norm": 0.37727612257003784, "learning_rate": 6.780779087661917e-06, "loss": 0.0114, "step": 10100 }, { "epoch": 0.6941313908741067, "grad_norm": 0.60476154088974, "learning_rate": 6.777986537734286e-06, "loss": 0.011, "step": 10101 }, { "epoch": 0.6942001099505223, "grad_norm": 0.29966339468955994, "learning_rate": 6.775194395108287e-06, "loss": 0.011, "step": 10102 }, { "epoch": 0.6942688290269379, "grad_norm": 0.2873048782348633, "learning_rate": 6.772402659922259e-06, "loss": 0.0204, "step": 10103 }, { "epoch": 0.6943375481033535, "grad_norm": 0.7037249207496643, "learning_rate": 6.769611332314483e-06, "loss": 0.0232, "step": 10104 }, { "epoch": 0.694406267179769, "grad_norm": 0.4007885456085205, "learning_rate": 6.766820412423246e-06, "loss": 0.0076, "step": 10105 }, { "epoch": 0.6944749862561848, "grad_norm": 0.5831739902496338, "learning_rate": 6.764029900386795e-06, "loss": 0.0076, "step": 10106 }, { "epoch": 0.6945437053326003, "grad_norm": 0.4476044774055481, "learning_rate": 6.761239796343373e-06, "loss": 0.0232, "step": 10107 }, { "epoch": 0.6946124244090159, "grad_norm": 0.4892513155937195, "learning_rate": 6.75845010043119e-06, "loss": 0.0119, "step": 10108 }, { "epoch": 0.6946811434854315, "grad_norm": 0.25937652587890625, "learning_rate": 6.755660812788445e-06, "loss": 0.0128, "step": 10109 }, { "epoch": 0.6947498625618471, "grad_norm": 1.2445882558822632, "learning_rate": 6.7528719335533166e-06, "loss": 0.0281, "step": 10110 }, { "epoch": 0.6948185816382628, "grad_norm": 0.3503074645996094, "learning_rate": 6.7500834628639545e-06, "loss": 0.0088, "step": 10111 }, { "epoch": 0.6948873007146784, "grad_norm": 2.7707326412200928, "learning_rate": 6.747295400858501e-06, "loss": 0.0116, "step": 10112 }, { "epoch": 0.694956019791094, "grad_norm": 0.4152211546897888, "learning_rate": 6.744507747675061e-06, "loss": 0.0117, "step": 10113 }, { "epoch": 0.6950247388675096, "grad_norm": 0.28824082016944885, "learning_rate": 6.741720503451735e-06, "loss": 0.0109, "step": 10114 }, { "epoch": 0.6950934579439252, "grad_norm": 0.31084442138671875, "learning_rate": 6.738933668326597e-06, "loss": 0.0069, "step": 10115 }, { "epoch": 0.6951621770203409, "grad_norm": 0.28493714332580566, "learning_rate": 6.736147242437707e-06, "loss": 0.0107, "step": 10116 }, { "epoch": 0.6952308960967565, "grad_norm": 0.892302930355072, "learning_rate": 6.733361225923086e-06, "loss": 0.0178, "step": 10117 }, { "epoch": 0.6952996151731721, "grad_norm": 0.4612181484699249, "learning_rate": 6.73057561892076e-06, "loss": 0.0105, "step": 10118 }, { "epoch": 0.6953683342495877, "grad_norm": 0.536923348903656, "learning_rate": 6.727790421568712e-06, "loss": 0.0135, "step": 10119 }, { "epoch": 0.6954370533260033, "grad_norm": 0.4649023711681366, "learning_rate": 6.7250056340049195e-06, "loss": 0.0164, "step": 10120 }, { "epoch": 0.695505772402419, "grad_norm": 0.23089323937892914, "learning_rate": 6.722221256367335e-06, "loss": 0.0119, "step": 10121 }, { "epoch": 0.6955744914788345, "grad_norm": 0.3414234220981598, "learning_rate": 6.719437288793896e-06, "loss": 0.0045, "step": 10122 }, { "epoch": 0.6956432105552501, "grad_norm": 0.5678264498710632, "learning_rate": 6.7166537314225055e-06, "loss": 0.0131, "step": 10123 }, { "epoch": 0.6957119296316657, "grad_norm": 0.7538627982139587, "learning_rate": 6.713870584391063e-06, "loss": 0.0157, "step": 10124 }, { "epoch": 0.6957806487080813, "grad_norm": 0.8458392024040222, "learning_rate": 6.7110878478374315e-06, "loss": 0.0051, "step": 10125 }, { "epoch": 0.695849367784497, "grad_norm": 0.4736734628677368, "learning_rate": 6.7083055218994675e-06, "loss": 0.0116, "step": 10126 }, { "epoch": 0.6959180868609126, "grad_norm": 0.19123458862304688, "learning_rate": 6.705523606715006e-06, "loss": 0.0072, "step": 10127 }, { "epoch": 0.6959868059373282, "grad_norm": 0.6953057646751404, "learning_rate": 6.702742102421847e-06, "loss": 0.0107, "step": 10128 }, { "epoch": 0.6960555250137438, "grad_norm": 0.5302728414535522, "learning_rate": 6.69996100915779e-06, "loss": 0.0095, "step": 10129 }, { "epoch": 0.6961242440901594, "grad_norm": 0.39474043250083923, "learning_rate": 6.697180327060595e-06, "loss": 0.0189, "step": 10130 }, { "epoch": 0.6961929631665751, "grad_norm": 0.5626105070114136, "learning_rate": 6.694400056268023e-06, "loss": 0.0208, "step": 10131 }, { "epoch": 0.6962616822429907, "grad_norm": 0.4024795889854431, "learning_rate": 6.691620196917787e-06, "loss": 0.0104, "step": 10132 }, { "epoch": 0.6963304013194063, "grad_norm": 0.5462505221366882, "learning_rate": 6.688840749147615e-06, "loss": 0.009, "step": 10133 }, { "epoch": 0.6963991203958219, "grad_norm": 0.5370432734489441, "learning_rate": 6.686061713095179e-06, "loss": 0.0149, "step": 10134 }, { "epoch": 0.6964678394722374, "grad_norm": 0.8976731300354004, "learning_rate": 6.683283088898159e-06, "loss": 0.0149, "step": 10135 }, { "epoch": 0.6965365585486532, "grad_norm": 0.1847529262304306, "learning_rate": 6.68050487669419e-06, "loss": 0.0068, "step": 10136 }, { "epoch": 0.6966052776250687, "grad_norm": 0.3223782479763031, "learning_rate": 6.677727076620912e-06, "loss": 0.015, "step": 10137 }, { "epoch": 0.6966739967014843, "grad_norm": 0.45594435930252075, "learning_rate": 6.674949688815914e-06, "loss": 0.0106, "step": 10138 }, { "epoch": 0.6967427157778999, "grad_norm": 0.27391621470451355, "learning_rate": 6.672172713416803e-06, "loss": 0.0148, "step": 10139 }, { "epoch": 0.6968114348543155, "grad_norm": 0.7298282384872437, "learning_rate": 6.66939615056113e-06, "loss": 0.0155, "step": 10140 }, { "epoch": 0.6968801539307312, "grad_norm": 0.6729704141616821, "learning_rate": 6.66662000038645e-06, "loss": 0.0157, "step": 10141 }, { "epoch": 0.6969488730071468, "grad_norm": 0.4317208230495453, "learning_rate": 6.663844263030275e-06, "loss": 0.0065, "step": 10142 }, { "epoch": 0.6970175920835624, "grad_norm": 0.40678083896636963, "learning_rate": 6.661068938630124e-06, "loss": 0.01, "step": 10143 }, { "epoch": 0.697086311159978, "grad_norm": 0.41256874799728394, "learning_rate": 6.658294027323468e-06, "loss": 0.0047, "step": 10144 }, { "epoch": 0.6971550302363936, "grad_norm": 0.6226008534431458, "learning_rate": 6.655519529247776e-06, "loss": 0.0056, "step": 10145 }, { "epoch": 0.6972237493128093, "grad_norm": 0.5942531228065491, "learning_rate": 6.652745444540498e-06, "loss": 0.0177, "step": 10146 }, { "epoch": 0.6972924683892249, "grad_norm": 0.5446844100952148, "learning_rate": 6.64997177333904e-06, "loss": 0.0112, "step": 10147 }, { "epoch": 0.6973611874656405, "grad_norm": 0.471390038728714, "learning_rate": 6.647198515780823e-06, "loss": 0.0154, "step": 10148 }, { "epoch": 0.697429906542056, "grad_norm": 0.3911469876766205, "learning_rate": 6.6444256720032105e-06, "loss": 0.0093, "step": 10149 }, { "epoch": 0.6974986256184716, "grad_norm": 0.40608376264572144, "learning_rate": 6.6416532421435745e-06, "loss": 0.0138, "step": 10150 }, { "epoch": 0.6975673446948873, "grad_norm": 0.31057533621788025, "learning_rate": 6.63888122633925e-06, "loss": 0.0094, "step": 10151 }, { "epoch": 0.6976360637713029, "grad_norm": 0.9572612643241882, "learning_rate": 6.6361096247275675e-06, "loss": 0.0157, "step": 10152 }, { "epoch": 0.6977047828477185, "grad_norm": 0.40905460715293884, "learning_rate": 6.633338437445812e-06, "loss": 0.0151, "step": 10153 }, { "epoch": 0.6977735019241341, "grad_norm": 0.4421028196811676, "learning_rate": 6.630567664631276e-06, "loss": 0.0117, "step": 10154 }, { "epoch": 0.6978422210005497, "grad_norm": 0.25188541412353516, "learning_rate": 6.6277973064212046e-06, "loss": 0.0071, "step": 10155 }, { "epoch": 0.6979109400769654, "grad_norm": 0.6196262240409851, "learning_rate": 6.625027362952842e-06, "loss": 0.0123, "step": 10156 }, { "epoch": 0.697979659153381, "grad_norm": 0.6302770972251892, "learning_rate": 6.622257834363406e-06, "loss": 0.0106, "step": 10157 }, { "epoch": 0.6980483782297966, "grad_norm": 0.7319104075431824, "learning_rate": 6.6194887207900976e-06, "loss": 0.012, "step": 10158 }, { "epoch": 0.6981170973062122, "grad_norm": 0.4360291361808777, "learning_rate": 6.6167200223700845e-06, "loss": 0.0099, "step": 10159 }, { "epoch": 0.6981858163826278, "grad_norm": 0.43945246934890747, "learning_rate": 6.613951739240531e-06, "loss": 0.0071, "step": 10160 }, { "epoch": 0.6982545354590435, "grad_norm": 0.16823190450668335, "learning_rate": 6.611183871538561e-06, "loss": 0.0128, "step": 10161 }, { "epoch": 0.6983232545354591, "grad_norm": 0.5632128119468689, "learning_rate": 6.608416419401297e-06, "loss": 0.0138, "step": 10162 }, { "epoch": 0.6983919736118747, "grad_norm": 0.27441683411598206, "learning_rate": 6.605649382965836e-06, "loss": 0.0046, "step": 10163 }, { "epoch": 0.6984606926882903, "grad_norm": 0.7512964010238647, "learning_rate": 6.602882762369241e-06, "loss": 0.0116, "step": 10164 }, { "epoch": 0.6985294117647058, "grad_norm": 0.6393246054649353, "learning_rate": 6.600116557748576e-06, "loss": 0.0172, "step": 10165 }, { "epoch": 0.6985981308411215, "grad_norm": 0.6494079828262329, "learning_rate": 6.59735076924086e-06, "loss": 0.0102, "step": 10166 }, { "epoch": 0.6986668499175371, "grad_norm": 0.3097570538520813, "learning_rate": 6.5945853969831184e-06, "loss": 0.0081, "step": 10167 }, { "epoch": 0.6987355689939527, "grad_norm": 0.4025824964046478, "learning_rate": 6.591820441112325e-06, "loss": 0.0093, "step": 10168 }, { "epoch": 0.6988042880703683, "grad_norm": 0.6964141130447388, "learning_rate": 6.589055901765471e-06, "loss": 0.0078, "step": 10169 }, { "epoch": 0.6988730071467839, "grad_norm": 0.5765654444694519, "learning_rate": 6.5862917790794905e-06, "loss": 0.0117, "step": 10170 }, { "epoch": 0.6989417262231996, "grad_norm": 0.7023218274116516, "learning_rate": 6.583528073191323e-06, "loss": 0.0094, "step": 10171 }, { "epoch": 0.6990104452996152, "grad_norm": 0.32412800192832947, "learning_rate": 6.580764784237864e-06, "loss": 0.0103, "step": 10172 }, { "epoch": 0.6990791643760308, "grad_norm": 0.7626737952232361, "learning_rate": 6.578001912356015e-06, "loss": 0.016, "step": 10173 }, { "epoch": 0.6991478834524464, "grad_norm": 1.9804795980453491, "learning_rate": 6.575239457682626e-06, "loss": 0.0123, "step": 10174 }, { "epoch": 0.699216602528862, "grad_norm": 0.8491575121879578, "learning_rate": 6.572477420354564e-06, "loss": 0.0116, "step": 10175 }, { "epoch": 0.6992853216052777, "grad_norm": 0.3971167504787445, "learning_rate": 6.569715800508639e-06, "loss": 0.0116, "step": 10176 }, { "epoch": 0.6993540406816933, "grad_norm": 0.4141696095466614, "learning_rate": 6.566954598281666e-06, "loss": 0.0131, "step": 10177 }, { "epoch": 0.6994227597581089, "grad_norm": 0.42266911268234253, "learning_rate": 6.564193813810421e-06, "loss": 0.0197, "step": 10178 }, { "epoch": 0.6994914788345244, "grad_norm": 0.3330332934856415, "learning_rate": 6.561433447231676e-06, "loss": 0.0079, "step": 10179 }, { "epoch": 0.69956019791094, "grad_norm": 0.30929237604141235, "learning_rate": 6.558673498682162e-06, "loss": 0.0129, "step": 10180 }, { "epoch": 0.6996289169873557, "grad_norm": 0.11728893220424652, "learning_rate": 6.55591396829861e-06, "loss": 0.0144, "step": 10181 }, { "epoch": 0.6996976360637713, "grad_norm": 0.32633084058761597, "learning_rate": 6.553154856217724e-06, "loss": 0.0081, "step": 10182 }, { "epoch": 0.6997663551401869, "grad_norm": 0.2672635614871979, "learning_rate": 6.550396162576176e-06, "loss": 0.0069, "step": 10183 }, { "epoch": 0.6998350742166025, "grad_norm": 0.585950493812561, "learning_rate": 6.547637887510634e-06, "loss": 0.0118, "step": 10184 }, { "epoch": 0.6999037932930181, "grad_norm": 0.5171961188316345, "learning_rate": 6.54488003115773e-06, "loss": 0.0137, "step": 10185 }, { "epoch": 0.6999725123694338, "grad_norm": 0.43951958417892456, "learning_rate": 6.542122593654085e-06, "loss": 0.0121, "step": 10186 }, { "epoch": 0.7000412314458494, "grad_norm": 0.885029137134552, "learning_rate": 6.539365575136301e-06, "loss": 0.0146, "step": 10187 }, { "epoch": 0.700109950522265, "grad_norm": 0.2849781811237335, "learning_rate": 6.536608975740956e-06, "loss": 0.0259, "step": 10188 }, { "epoch": 0.7001786695986806, "grad_norm": 0.8120207786560059, "learning_rate": 6.533852795604595e-06, "loss": 0.015, "step": 10189 }, { "epoch": 0.7002473886750962, "grad_norm": 0.5077491402626038, "learning_rate": 6.531097034863768e-06, "loss": 0.0114, "step": 10190 }, { "epoch": 0.7003161077515119, "grad_norm": 0.44824081659317017, "learning_rate": 6.528341693654978e-06, "loss": 0.0126, "step": 10191 }, { "epoch": 0.7003848268279275, "grad_norm": 0.822319507598877, "learning_rate": 6.525586772114721e-06, "loss": 0.0119, "step": 10192 }, { "epoch": 0.7004535459043431, "grad_norm": 0.3957124650478363, "learning_rate": 6.522832270379475e-06, "loss": 0.016, "step": 10193 }, { "epoch": 0.7005222649807586, "grad_norm": 0.3014368712902069, "learning_rate": 6.520078188585694e-06, "loss": 0.0098, "step": 10194 }, { "epoch": 0.7005909840571742, "grad_norm": 0.3166975975036621, "learning_rate": 6.517324526869801e-06, "loss": 0.0095, "step": 10195 }, { "epoch": 0.7006597031335899, "grad_norm": 0.4959771931171417, "learning_rate": 6.5145712853682156e-06, "loss": 0.0102, "step": 10196 }, { "epoch": 0.7007284222100055, "grad_norm": 0.29366591572761536, "learning_rate": 6.511818464217319e-06, "loss": 0.006, "step": 10197 }, { "epoch": 0.7007971412864211, "grad_norm": 0.3292320668697357, "learning_rate": 6.509066063553484e-06, "loss": 0.0173, "step": 10198 }, { "epoch": 0.7008658603628367, "grad_norm": 0.3008494973182678, "learning_rate": 6.506314083513066e-06, "loss": 0.0269, "step": 10199 }, { "epoch": 0.7009345794392523, "grad_norm": 0.240242600440979, "learning_rate": 6.50356252423238e-06, "loss": 0.0164, "step": 10200 }, { "epoch": 0.7010032985156679, "grad_norm": 0.6614487767219543, "learning_rate": 6.500811385847743e-06, "loss": 0.0071, "step": 10201 }, { "epoch": 0.7010720175920836, "grad_norm": 0.4131113588809967, "learning_rate": 6.498060668495433e-06, "loss": 0.01, "step": 10202 }, { "epoch": 0.7011407366684992, "grad_norm": 0.2876932621002197, "learning_rate": 6.495310372311718e-06, "loss": 0.0072, "step": 10203 }, { "epoch": 0.7012094557449148, "grad_norm": 0.41407114267349243, "learning_rate": 6.492560497432841e-06, "loss": 0.0142, "step": 10204 }, { "epoch": 0.7012781748213304, "grad_norm": 0.4882214069366455, "learning_rate": 6.489811043995032e-06, "loss": 0.0112, "step": 10205 }, { "epoch": 0.701346893897746, "grad_norm": 0.31212425231933594, "learning_rate": 6.4870620121344825e-06, "loss": 0.0112, "step": 10206 }, { "epoch": 0.7014156129741617, "grad_norm": 0.430901437997818, "learning_rate": 6.484313401987386e-06, "loss": 0.0158, "step": 10207 }, { "epoch": 0.7014843320505773, "grad_norm": 0.6486769318580627, "learning_rate": 6.481565213689889e-06, "loss": 0.015, "step": 10208 }, { "epoch": 0.7015530511269928, "grad_norm": 0.16117477416992188, "learning_rate": 6.4788174473781435e-06, "loss": 0.0118, "step": 10209 }, { "epoch": 0.7016217702034084, "grad_norm": 0.27070510387420654, "learning_rate": 6.4760701031882524e-06, "loss": 0.0103, "step": 10210 }, { "epoch": 0.701690489279824, "grad_norm": 0.3879643976688385, "learning_rate": 6.473323181256333e-06, "loss": 0.0117, "step": 10211 }, { "epoch": 0.7017592083562397, "grad_norm": 0.20970045030117035, "learning_rate": 6.47057668171845e-06, "loss": 0.0054, "step": 10212 }, { "epoch": 0.7018279274326553, "grad_norm": 0.29429009556770325, "learning_rate": 6.467830604710666e-06, "loss": 0.0167, "step": 10213 }, { "epoch": 0.7018966465090709, "grad_norm": 0.42429181933403015, "learning_rate": 6.46508495036901e-06, "loss": 0.0161, "step": 10214 }, { "epoch": 0.7019653655854865, "grad_norm": 0.4391263723373413, "learning_rate": 6.462339718829502e-06, "loss": 0.0101, "step": 10215 }, { "epoch": 0.7020340846619021, "grad_norm": 0.18397285044193268, "learning_rate": 6.459594910228127e-06, "loss": 0.0075, "step": 10216 }, { "epoch": 0.7021028037383178, "grad_norm": 0.3922026753425598, "learning_rate": 6.45685052470086e-06, "loss": 0.0157, "step": 10217 }, { "epoch": 0.7021715228147334, "grad_norm": 0.599658191204071, "learning_rate": 6.454106562383656e-06, "loss": 0.0088, "step": 10218 }, { "epoch": 0.702240241891149, "grad_norm": 0.43164029717445374, "learning_rate": 6.451363023412448e-06, "loss": 0.0057, "step": 10219 }, { "epoch": 0.7023089609675646, "grad_norm": 0.772879421710968, "learning_rate": 6.448619907923135e-06, "loss": 0.0117, "step": 10220 }, { "epoch": 0.7023776800439802, "grad_norm": 0.6622151136398315, "learning_rate": 6.445877216051616e-06, "loss": 0.0162, "step": 10221 }, { "epoch": 0.7024463991203959, "grad_norm": 0.5563493371009827, "learning_rate": 6.443134947933748e-06, "loss": 0.0095, "step": 10222 }, { "epoch": 0.7025151181968114, "grad_norm": 0.38652288913726807, "learning_rate": 6.440393103705382e-06, "loss": 0.0079, "step": 10223 }, { "epoch": 0.702583837273227, "grad_norm": 0.15959347784519196, "learning_rate": 6.437651683502351e-06, "loss": 0.0072, "step": 10224 }, { "epoch": 0.7026525563496426, "grad_norm": 0.645456075668335, "learning_rate": 6.434910687460445e-06, "loss": 0.0075, "step": 10225 }, { "epoch": 0.7027212754260582, "grad_norm": 0.34799131751060486, "learning_rate": 6.432170115715461e-06, "loss": 0.0117, "step": 10226 }, { "epoch": 0.7027899945024739, "grad_norm": 0.6888665556907654, "learning_rate": 6.429429968403148e-06, "loss": 0.0137, "step": 10227 }, { "epoch": 0.7028587135788895, "grad_norm": 0.58621746301651, "learning_rate": 6.426690245659257e-06, "loss": 0.0125, "step": 10228 }, { "epoch": 0.7029274326553051, "grad_norm": 0.3839234709739685, "learning_rate": 6.423950947619501e-06, "loss": 0.0122, "step": 10229 }, { "epoch": 0.7029961517317207, "grad_norm": 0.5909634232521057, "learning_rate": 6.421212074419591e-06, "loss": 0.0154, "step": 10230 }, { "epoch": 0.7030648708081363, "grad_norm": 0.4876726269721985, "learning_rate": 6.418473626195192e-06, "loss": 0.0078, "step": 10231 }, { "epoch": 0.703133589884552, "grad_norm": 0.7285220623016357, "learning_rate": 6.415735603081973e-06, "loss": 0.0158, "step": 10232 }, { "epoch": 0.7032023089609676, "grad_norm": 0.6646744608879089, "learning_rate": 6.412998005215558e-06, "loss": 0.0148, "step": 10233 }, { "epoch": 0.7032710280373832, "grad_norm": 0.3305065929889679, "learning_rate": 6.410260832731567e-06, "loss": 0.016, "step": 10234 }, { "epoch": 0.7033397471137988, "grad_norm": 0.4371705651283264, "learning_rate": 6.407524085765594e-06, "loss": 0.0077, "step": 10235 }, { "epoch": 0.7034084661902144, "grad_norm": 0.5157927870750427, "learning_rate": 6.40478776445322e-06, "loss": 0.0241, "step": 10236 }, { "epoch": 0.7034771852666301, "grad_norm": 0.4323478043079376, "learning_rate": 6.4020518689299824e-06, "loss": 0.0056, "step": 10237 }, { "epoch": 0.7035459043430456, "grad_norm": 0.6512686014175415, "learning_rate": 6.3993163993314226e-06, "loss": 0.0097, "step": 10238 }, { "epoch": 0.7036146234194612, "grad_norm": 0.2845352292060852, "learning_rate": 6.396581355793042e-06, "loss": 0.0068, "step": 10239 }, { "epoch": 0.7036833424958768, "grad_norm": 0.328009694814682, "learning_rate": 6.393846738450334e-06, "loss": 0.0093, "step": 10240 }, { "epoch": 0.7037520615722924, "grad_norm": 0.22738653421401978, "learning_rate": 6.391112547438771e-06, "loss": 0.0133, "step": 10241 }, { "epoch": 0.7038207806487081, "grad_norm": 0.5936235785484314, "learning_rate": 6.388378782893787e-06, "loss": 0.0097, "step": 10242 }, { "epoch": 0.7038894997251237, "grad_norm": 0.5948896408081055, "learning_rate": 6.385645444950818e-06, "loss": 0.0135, "step": 10243 }, { "epoch": 0.7039582188015393, "grad_norm": 0.6316744685173035, "learning_rate": 6.382912533745258e-06, "loss": 0.0155, "step": 10244 }, { "epoch": 0.7040269378779549, "grad_norm": 1.0634634494781494, "learning_rate": 6.380180049412501e-06, "loss": 0.0119, "step": 10245 }, { "epoch": 0.7040956569543705, "grad_norm": 0.7576687932014465, "learning_rate": 6.37744799208789e-06, "loss": 0.0189, "step": 10246 }, { "epoch": 0.7041643760307862, "grad_norm": 0.6160529851913452, "learning_rate": 6.374716361906792e-06, "loss": 0.013, "step": 10247 }, { "epoch": 0.7042330951072018, "grad_norm": 0.6165721416473389, "learning_rate": 6.3719851590045056e-06, "loss": 0.014, "step": 10248 }, { "epoch": 0.7043018141836174, "grad_norm": 0.39244556427001953, "learning_rate": 6.3692543835163415e-06, "loss": 0.0119, "step": 10249 }, { "epoch": 0.704370533260033, "grad_norm": 0.38914012908935547, "learning_rate": 6.366524035577566e-06, "loss": 0.0194, "step": 10250 }, { "epoch": 0.7044392523364486, "grad_norm": 0.7634445428848267, "learning_rate": 6.363794115323446e-06, "loss": 0.0152, "step": 10251 }, { "epoch": 0.7045079714128643, "grad_norm": 0.24523480236530304, "learning_rate": 6.361064622889199e-06, "loss": 0.0125, "step": 10252 }, { "epoch": 0.7045766904892798, "grad_norm": 0.4729175567626953, "learning_rate": 6.358335558410061e-06, "loss": 0.0077, "step": 10253 }, { "epoch": 0.7046454095656954, "grad_norm": 0.7536904811859131, "learning_rate": 6.355606922021206e-06, "loss": 0.0117, "step": 10254 }, { "epoch": 0.704714128642111, "grad_norm": 0.25525903701782227, "learning_rate": 6.352878713857819e-06, "loss": 0.0077, "step": 10255 }, { "epoch": 0.7047828477185266, "grad_norm": 0.348685085773468, "learning_rate": 6.350150934055036e-06, "loss": 0.0092, "step": 10256 }, { "epoch": 0.7048515667949423, "grad_norm": 0.4707692265510559, "learning_rate": 6.347423582747999e-06, "loss": 0.0103, "step": 10257 }, { "epoch": 0.7049202858713579, "grad_norm": 0.7948170304298401, "learning_rate": 6.3446966600718036e-06, "loss": 0.0136, "step": 10258 }, { "epoch": 0.7049890049477735, "grad_norm": 0.6397384405136108, "learning_rate": 6.34197016616154e-06, "loss": 0.0196, "step": 10259 }, { "epoch": 0.7050577240241891, "grad_norm": 0.5231978297233582, "learning_rate": 6.33924410115228e-06, "loss": 0.0145, "step": 10260 }, { "epoch": 0.7051264431006047, "grad_norm": 0.34170371294021606, "learning_rate": 6.336518465179056e-06, "loss": 0.0063, "step": 10261 }, { "epoch": 0.7051951621770204, "grad_norm": 0.9787119626998901, "learning_rate": 6.333793258376901e-06, "loss": 0.0179, "step": 10262 }, { "epoch": 0.705263881253436, "grad_norm": 0.7068516612052917, "learning_rate": 6.331068480880807e-06, "loss": 0.013, "step": 10263 }, { "epoch": 0.7053326003298516, "grad_norm": 0.2657037675380707, "learning_rate": 6.328344132825757e-06, "loss": 0.0094, "step": 10264 }, { "epoch": 0.7054013194062672, "grad_norm": 0.49556228518486023, "learning_rate": 6.32562021434671e-06, "loss": 0.0112, "step": 10265 }, { "epoch": 0.7054700384826827, "grad_norm": 0.7208208441734314, "learning_rate": 6.322896725578608e-06, "loss": 0.0109, "step": 10266 }, { "epoch": 0.7055387575590985, "grad_norm": 0.31943854689598083, "learning_rate": 6.320173666656361e-06, "loss": 0.0103, "step": 10267 }, { "epoch": 0.705607476635514, "grad_norm": 0.6218613982200623, "learning_rate": 6.317451037714867e-06, "loss": 0.0111, "step": 10268 }, { "epoch": 0.7056761957119296, "grad_norm": 0.38072898983955383, "learning_rate": 6.3147288388889955e-06, "loss": 0.0092, "step": 10269 }, { "epoch": 0.7057449147883452, "grad_norm": 0.3626660108566284, "learning_rate": 6.312007070313601e-06, "loss": 0.0111, "step": 10270 }, { "epoch": 0.7058136338647608, "grad_norm": 0.34151962399482727, "learning_rate": 6.309285732123513e-06, "loss": 0.0164, "step": 10271 }, { "epoch": 0.7058823529411765, "grad_norm": 0.4684062898159027, "learning_rate": 6.306564824453548e-06, "loss": 0.0148, "step": 10272 }, { "epoch": 0.7059510720175921, "grad_norm": 0.47985923290252686, "learning_rate": 6.303844347438486e-06, "loss": 0.0133, "step": 10273 }, { "epoch": 0.7060197910940077, "grad_norm": 0.5124370455741882, "learning_rate": 6.301124301213098e-06, "loss": 0.0103, "step": 10274 }, { "epoch": 0.7060885101704233, "grad_norm": 0.6374444365501404, "learning_rate": 6.298404685912125e-06, "loss": 0.0105, "step": 10275 }, { "epoch": 0.7061572292468389, "grad_norm": 0.411544531583786, "learning_rate": 6.295685501670294e-06, "loss": 0.0146, "step": 10276 }, { "epoch": 0.7062259483232546, "grad_norm": 0.4505685865879059, "learning_rate": 6.2929667486223115e-06, "loss": 0.0077, "step": 10277 }, { "epoch": 0.7062946673996702, "grad_norm": 0.3107737898826599, "learning_rate": 6.290248426902853e-06, "loss": 0.0149, "step": 10278 }, { "epoch": 0.7063633864760858, "grad_norm": 0.5452378988265991, "learning_rate": 6.287530536646584e-06, "loss": 0.0087, "step": 10279 }, { "epoch": 0.7064321055525014, "grad_norm": 0.0883873701095581, "learning_rate": 6.284813077988135e-06, "loss": 0.0072, "step": 10280 }, { "epoch": 0.706500824628917, "grad_norm": 0.33082687854766846, "learning_rate": 6.282096051062135e-06, "loss": 0.0126, "step": 10281 }, { "epoch": 0.7065695437053326, "grad_norm": 0.22417302429676056, "learning_rate": 6.279379456003162e-06, "loss": 0.0138, "step": 10282 }, { "epoch": 0.7066382627817482, "grad_norm": 0.4488595724105835, "learning_rate": 6.276663292945812e-06, "loss": 0.012, "step": 10283 }, { "epoch": 0.7067069818581638, "grad_norm": 0.6650792360305786, "learning_rate": 6.273947562024623e-06, "loss": 0.0175, "step": 10284 }, { "epoch": 0.7067757009345794, "grad_norm": 0.5617098808288574, "learning_rate": 6.271232263374138e-06, "loss": 0.0121, "step": 10285 }, { "epoch": 0.706844420010995, "grad_norm": 0.2695590853691101, "learning_rate": 6.268517397128855e-06, "loss": 0.0094, "step": 10286 }, { "epoch": 0.7069131390874107, "grad_norm": 0.3680937588214874, "learning_rate": 6.265802963423275e-06, "loss": 0.0113, "step": 10287 }, { "epoch": 0.7069818581638263, "grad_norm": 0.7538487315177917, "learning_rate": 6.263088962391848e-06, "loss": 0.0235, "step": 10288 }, { "epoch": 0.7070505772402419, "grad_norm": 0.13040655851364136, "learning_rate": 6.260375394169042e-06, "loss": 0.0055, "step": 10289 }, { "epoch": 0.7071192963166575, "grad_norm": 0.5385776162147522, "learning_rate": 6.2576622588892655e-06, "loss": 0.0181, "step": 10290 }, { "epoch": 0.7071880153930731, "grad_norm": 0.37937864661216736, "learning_rate": 6.254949556686934e-06, "loss": 0.0088, "step": 10291 }, { "epoch": 0.7072567344694888, "grad_norm": 0.27871668338775635, "learning_rate": 6.252237287696416e-06, "loss": 0.0102, "step": 10292 }, { "epoch": 0.7073254535459044, "grad_norm": 0.2629517614841461, "learning_rate": 6.249525452052084e-06, "loss": 0.0152, "step": 10293 }, { "epoch": 0.70739417262232, "grad_norm": 0.3802239000797272, "learning_rate": 6.2468140498882645e-06, "loss": 0.0109, "step": 10294 }, { "epoch": 0.7074628916987356, "grad_norm": 0.39177069067955017, "learning_rate": 6.244103081339281e-06, "loss": 0.0127, "step": 10295 }, { "epoch": 0.7075316107751511, "grad_norm": 0.3884953558444977, "learning_rate": 6.241392546539436e-06, "loss": 0.0099, "step": 10296 }, { "epoch": 0.7076003298515668, "grad_norm": 0.399621844291687, "learning_rate": 6.238682445622992e-06, "loss": 0.0158, "step": 10297 }, { "epoch": 0.7076690489279824, "grad_norm": 0.43649569153785706, "learning_rate": 6.2359727787242126e-06, "loss": 0.0084, "step": 10298 }, { "epoch": 0.707737768004398, "grad_norm": 0.30448010563850403, "learning_rate": 6.233263545977318e-06, "loss": 0.01, "step": 10299 }, { "epoch": 0.7078064870808136, "grad_norm": 0.5197250247001648, "learning_rate": 6.230554747516529e-06, "loss": 0.0109, "step": 10300 }, { "epoch": 0.7078752061572292, "grad_norm": 0.6416959762573242, "learning_rate": 6.227846383476017e-06, "loss": 0.0095, "step": 10301 }, { "epoch": 0.7079439252336449, "grad_norm": 0.8023585081100464, "learning_rate": 6.225138453989972e-06, "loss": 0.0064, "step": 10302 }, { "epoch": 0.7080126443100605, "grad_norm": 0.8282856941223145, "learning_rate": 6.222430959192521e-06, "loss": 0.0084, "step": 10303 }, { "epoch": 0.7080813633864761, "grad_norm": 0.587250292301178, "learning_rate": 6.219723899217802e-06, "loss": 0.0184, "step": 10304 }, { "epoch": 0.7081500824628917, "grad_norm": 0.5680186748504639, "learning_rate": 6.217017274199904e-06, "loss": 0.0098, "step": 10305 }, { "epoch": 0.7082188015393073, "grad_norm": 0.5625176429748535, "learning_rate": 6.2143110842729165e-06, "loss": 0.0154, "step": 10306 }, { "epoch": 0.708287520615723, "grad_norm": 0.267744243144989, "learning_rate": 6.211605329570886e-06, "loss": 0.0137, "step": 10307 }, { "epoch": 0.7083562396921386, "grad_norm": 0.42324134707450867, "learning_rate": 6.20890001022787e-06, "loss": 0.0152, "step": 10308 }, { "epoch": 0.7084249587685542, "grad_norm": 0.34906426072120667, "learning_rate": 6.2061951263778685e-06, "loss": 0.0154, "step": 10309 }, { "epoch": 0.7084936778449697, "grad_norm": 0.32746872305870056, "learning_rate": 6.203490678154885e-06, "loss": 0.0076, "step": 10310 }, { "epoch": 0.7085623969213853, "grad_norm": 0.47040286660194397, "learning_rate": 6.200786665692885e-06, "loss": 0.0102, "step": 10311 }, { "epoch": 0.708631115997801, "grad_norm": 0.3847824037075043, "learning_rate": 6.198083089125827e-06, "loss": 0.0086, "step": 10312 }, { "epoch": 0.7086998350742166, "grad_norm": 0.16137109696865082, "learning_rate": 6.195379948587633e-06, "loss": 0.0079, "step": 10313 }, { "epoch": 0.7087685541506322, "grad_norm": 0.2490687519311905, "learning_rate": 6.192677244212213e-06, "loss": 0.0073, "step": 10314 }, { "epoch": 0.7088372732270478, "grad_norm": 0.427664577960968, "learning_rate": 6.189974976133463e-06, "loss": 0.0114, "step": 10315 }, { "epoch": 0.7089059923034634, "grad_norm": 0.18159155547618866, "learning_rate": 6.187273144485231e-06, "loss": 0.0071, "step": 10316 }, { "epoch": 0.7089747113798791, "grad_norm": 0.534015953540802, "learning_rate": 6.184571749401377e-06, "loss": 0.017, "step": 10317 }, { "epoch": 0.7090434304562947, "grad_norm": 0.7080742120742798, "learning_rate": 6.1818707910157104e-06, "loss": 0.013, "step": 10318 }, { "epoch": 0.7091121495327103, "grad_norm": 0.4083153009414673, "learning_rate": 6.179170269462034e-06, "loss": 0.0116, "step": 10319 }, { "epoch": 0.7091808686091259, "grad_norm": 0.5657504200935364, "learning_rate": 6.176470184874128e-06, "loss": 0.0127, "step": 10320 }, { "epoch": 0.7092495876855415, "grad_norm": 0.30040016770362854, "learning_rate": 6.1737705373857535e-06, "loss": 0.0168, "step": 10321 }, { "epoch": 0.7093183067619572, "grad_norm": 0.12480570375919342, "learning_rate": 6.171071327130635e-06, "loss": 0.0104, "step": 10322 }, { "epoch": 0.7093870258383728, "grad_norm": 0.7662149667739868, "learning_rate": 6.168372554242498e-06, "loss": 0.0115, "step": 10323 }, { "epoch": 0.7094557449147884, "grad_norm": 0.618686854839325, "learning_rate": 6.165674218855022e-06, "loss": 0.0168, "step": 10324 }, { "epoch": 0.709524463991204, "grad_norm": 0.19700659811496735, "learning_rate": 6.162976321101882e-06, "loss": 0.0088, "step": 10325 }, { "epoch": 0.7095931830676195, "grad_norm": 0.32705289125442505, "learning_rate": 6.1602788611167275e-06, "loss": 0.0171, "step": 10326 }, { "epoch": 0.7096619021440352, "grad_norm": 0.3580344617366791, "learning_rate": 6.15758183903319e-06, "loss": 0.0081, "step": 10327 }, { "epoch": 0.7097306212204508, "grad_norm": 0.5275037288665771, "learning_rate": 6.154885254984862e-06, "loss": 0.0103, "step": 10328 }, { "epoch": 0.7097993402968664, "grad_norm": 0.592427134513855, "learning_rate": 6.152189109105341e-06, "loss": 0.0127, "step": 10329 }, { "epoch": 0.709868059373282, "grad_norm": 0.15878653526306152, "learning_rate": 6.149493401528175e-06, "loss": 0.0135, "step": 10330 }, { "epoch": 0.7099367784496976, "grad_norm": 0.4346598982810974, "learning_rate": 6.146798132386911e-06, "loss": 0.0078, "step": 10331 }, { "epoch": 0.7100054975261132, "grad_norm": 0.05814524367451668, "learning_rate": 6.14410330181507e-06, "loss": 0.0094, "step": 10332 }, { "epoch": 0.7100742166025289, "grad_norm": 0.7890493273735046, "learning_rate": 6.14140890994614e-06, "loss": 0.0211, "step": 10333 }, { "epoch": 0.7101429356789445, "grad_norm": 0.3154886066913605, "learning_rate": 6.1387149569136045e-06, "loss": 0.0082, "step": 10334 }, { "epoch": 0.7102116547553601, "grad_norm": 0.4921967685222626, "learning_rate": 6.136021442850907e-06, "loss": 0.0156, "step": 10335 }, { "epoch": 0.7102803738317757, "grad_norm": 0.3608103394508362, "learning_rate": 6.133328367891491e-06, "loss": 0.0082, "step": 10336 }, { "epoch": 0.7103490929081913, "grad_norm": 0.17276622354984283, "learning_rate": 6.130635732168746e-06, "loss": 0.0057, "step": 10337 }, { "epoch": 0.710417811984607, "grad_norm": 0.34253448247909546, "learning_rate": 6.127943535816083e-06, "loss": 0.0087, "step": 10338 }, { "epoch": 0.7104865310610226, "grad_norm": 0.20173510909080505, "learning_rate": 6.125251778966851e-06, "loss": 0.015, "step": 10339 }, { "epoch": 0.7105552501374381, "grad_norm": 0.29449009895324707, "learning_rate": 6.122560461754407e-06, "loss": 0.0102, "step": 10340 }, { "epoch": 0.7106239692138537, "grad_norm": 0.3494407534599304, "learning_rate": 6.11986958431206e-06, "loss": 0.0122, "step": 10341 }, { "epoch": 0.7106926882902693, "grad_norm": 0.4759292006492615, "learning_rate": 6.117179146773122e-06, "loss": 0.0123, "step": 10342 }, { "epoch": 0.710761407366685, "grad_norm": 0.7135450839996338, "learning_rate": 6.114489149270857e-06, "loss": 0.0105, "step": 10343 }, { "epoch": 0.7108301264431006, "grad_norm": 0.364449143409729, "learning_rate": 6.1117995919385425e-06, "loss": 0.0168, "step": 10344 }, { "epoch": 0.7108988455195162, "grad_norm": 0.7618456482887268, "learning_rate": 6.109110474909398e-06, "loss": 0.0247, "step": 10345 }, { "epoch": 0.7109675645959318, "grad_norm": 0.7763140797615051, "learning_rate": 6.106421798316645e-06, "loss": 0.0183, "step": 10346 }, { "epoch": 0.7110362836723474, "grad_norm": 0.4263259172439575, "learning_rate": 6.103733562293468e-06, "loss": 0.0063, "step": 10347 }, { "epoch": 0.7111050027487631, "grad_norm": 0.6374548673629761, "learning_rate": 6.101045766973046e-06, "loss": 0.0102, "step": 10348 }, { "epoch": 0.7111737218251787, "grad_norm": 0.455556184053421, "learning_rate": 6.098358412488515e-06, "loss": 0.0091, "step": 10349 }, { "epoch": 0.7112424409015943, "grad_norm": 0.8242130279541016, "learning_rate": 6.095671498973008e-06, "loss": 0.0114, "step": 10350 }, { "epoch": 0.7113111599780099, "grad_norm": 0.8469464182853699, "learning_rate": 6.092985026559633e-06, "loss": 0.0112, "step": 10351 }, { "epoch": 0.7113798790544255, "grad_norm": 0.31875351071357727, "learning_rate": 6.090298995381463e-06, "loss": 0.0152, "step": 10352 }, { "epoch": 0.7114485981308412, "grad_norm": 0.41861188411712646, "learning_rate": 6.087613405571568e-06, "loss": 0.0183, "step": 10353 }, { "epoch": 0.7115173172072567, "grad_norm": 0.4426966607570648, "learning_rate": 6.084928257262977e-06, "loss": 0.0169, "step": 10354 }, { "epoch": 0.7115860362836723, "grad_norm": 0.23069843649864197, "learning_rate": 6.08224355058871e-06, "loss": 0.0084, "step": 10355 }, { "epoch": 0.7116547553600879, "grad_norm": 0.5272266268730164, "learning_rate": 6.079559285681764e-06, "loss": 0.012, "step": 10356 }, { "epoch": 0.7117234744365035, "grad_norm": 0.23232747614383698, "learning_rate": 6.076875462675118e-06, "loss": 0.0088, "step": 10357 }, { "epoch": 0.7117921935129192, "grad_norm": 0.3099657893180847, "learning_rate": 6.07419208170171e-06, "loss": 0.0152, "step": 10358 }, { "epoch": 0.7118609125893348, "grad_norm": 0.3414161503314972, "learning_rate": 6.071509142894481e-06, "loss": 0.0109, "step": 10359 }, { "epoch": 0.7119296316657504, "grad_norm": 0.40601810812950134, "learning_rate": 6.068826646386327e-06, "loss": 0.0095, "step": 10360 }, { "epoch": 0.711998350742166, "grad_norm": 0.4687701463699341, "learning_rate": 6.066144592310141e-06, "loss": 0.0112, "step": 10361 }, { "epoch": 0.7120670698185816, "grad_norm": 0.7212871313095093, "learning_rate": 6.063462980798784e-06, "loss": 0.0233, "step": 10362 }, { "epoch": 0.7121357888949973, "grad_norm": 0.3300362825393677, "learning_rate": 6.060781811985103e-06, "loss": 0.0116, "step": 10363 }, { "epoch": 0.7122045079714129, "grad_norm": 0.6959275603294373, "learning_rate": 6.058101086001906e-06, "loss": 0.0106, "step": 10364 }, { "epoch": 0.7122732270478285, "grad_norm": 0.34093478322029114, "learning_rate": 6.0554208029820055e-06, "loss": 0.0117, "step": 10365 }, { "epoch": 0.7123419461242441, "grad_norm": 0.2257094383239746, "learning_rate": 6.052740963058163e-06, "loss": 0.0111, "step": 10366 }, { "epoch": 0.7124106652006597, "grad_norm": 0.5760868191719055, "learning_rate": 6.050061566363138e-06, "loss": 0.0071, "step": 10367 }, { "epoch": 0.7124793842770754, "grad_norm": 0.2600669860839844, "learning_rate": 6.047382613029668e-06, "loss": 0.0178, "step": 10368 }, { "epoch": 0.712548103353491, "grad_norm": 0.5720370411872864, "learning_rate": 6.044704103190453e-06, "loss": 0.017, "step": 10369 }, { "epoch": 0.7126168224299065, "grad_norm": 0.7061932682991028, "learning_rate": 6.042026036978189e-06, "loss": 0.0123, "step": 10370 }, { "epoch": 0.7126855415063221, "grad_norm": 0.4915921688079834, "learning_rate": 6.039348414525535e-06, "loss": 0.0149, "step": 10371 }, { "epoch": 0.7127542605827377, "grad_norm": 0.15386073291301727, "learning_rate": 6.0366712359651425e-06, "loss": 0.0057, "step": 10372 }, { "epoch": 0.7128229796591534, "grad_norm": 0.27021297812461853, "learning_rate": 6.033994501429618e-06, "loss": 0.0109, "step": 10373 }, { "epoch": 0.712891698735569, "grad_norm": 0.21563637256622314, "learning_rate": 6.031318211051584e-06, "loss": 0.0203, "step": 10374 }, { "epoch": 0.7129604178119846, "grad_norm": 0.451237291097641, "learning_rate": 6.028642364963602e-06, "loss": 0.0206, "step": 10375 }, { "epoch": 0.7130291368884002, "grad_norm": 0.2338564544916153, "learning_rate": 6.025966963298237e-06, "loss": 0.0093, "step": 10376 }, { "epoch": 0.7130978559648158, "grad_norm": 0.7456673383712769, "learning_rate": 6.023292006188015e-06, "loss": 0.0135, "step": 10377 }, { "epoch": 0.7131665750412315, "grad_norm": 0.36511003971099854, "learning_rate": 6.0206174937654555e-06, "loss": 0.0076, "step": 10378 }, { "epoch": 0.7132352941176471, "grad_norm": 0.6525375843048096, "learning_rate": 6.017943426163035e-06, "loss": 0.0124, "step": 10379 }, { "epoch": 0.7133040131940627, "grad_norm": 0.4319116473197937, "learning_rate": 6.0152698035132415e-06, "loss": 0.016, "step": 10380 }, { "epoch": 0.7133727322704783, "grad_norm": 0.77357017993927, "learning_rate": 6.012596625948503e-06, "loss": 0.0133, "step": 10381 }, { "epoch": 0.7134414513468939, "grad_norm": 0.6020447015762329, "learning_rate": 6.009923893601258e-06, "loss": 0.0142, "step": 10382 }, { "epoch": 0.7135101704233096, "grad_norm": 0.5132613182067871, "learning_rate": 6.007251606603895e-06, "loss": 0.0128, "step": 10383 }, { "epoch": 0.7135788894997251, "grad_norm": 0.26269403100013733, "learning_rate": 6.004579765088802e-06, "loss": 0.0107, "step": 10384 }, { "epoch": 0.7136476085761407, "grad_norm": 0.5575762391090393, "learning_rate": 6.00190836918833e-06, "loss": 0.0057, "step": 10385 }, { "epoch": 0.7137163276525563, "grad_norm": 0.33098146319389343, "learning_rate": 5.999237419034819e-06, "loss": 0.0124, "step": 10386 }, { "epoch": 0.7137850467289719, "grad_norm": 0.4126240611076355, "learning_rate": 5.996566914760585e-06, "loss": 0.0086, "step": 10387 }, { "epoch": 0.7138537658053876, "grad_norm": 0.27826592326164246, "learning_rate": 5.993896856497912e-06, "loss": 0.0203, "step": 10388 }, { "epoch": 0.7139224848818032, "grad_norm": 0.7895912528038025, "learning_rate": 5.991227244379077e-06, "loss": 0.0185, "step": 10389 }, { "epoch": 0.7139912039582188, "grad_norm": 0.4217384159564972, "learning_rate": 5.988558078536319e-06, "loss": 0.013, "step": 10390 }, { "epoch": 0.7140599230346344, "grad_norm": 0.8656349778175354, "learning_rate": 5.985889359101866e-06, "loss": 0.0097, "step": 10391 }, { "epoch": 0.71412864211105, "grad_norm": 0.12017090618610382, "learning_rate": 5.983221086207921e-06, "loss": 0.012, "step": 10392 }, { "epoch": 0.7141973611874657, "grad_norm": 0.8553171753883362, "learning_rate": 5.980553259986674e-06, "loss": 0.0168, "step": 10393 }, { "epoch": 0.7142660802638813, "grad_norm": 0.2511441111564636, "learning_rate": 5.977885880570269e-06, "loss": 0.0152, "step": 10394 }, { "epoch": 0.7143347993402969, "grad_norm": 0.26348692178726196, "learning_rate": 5.9752189480908524e-06, "loss": 0.0156, "step": 10395 }, { "epoch": 0.7144035184167125, "grad_norm": 0.3939489722251892, "learning_rate": 5.97255246268053e-06, "loss": 0.0143, "step": 10396 }, { "epoch": 0.714472237493128, "grad_norm": 0.975219190120697, "learning_rate": 5.9698864244714e-06, "loss": 0.0177, "step": 10397 }, { "epoch": 0.7145409565695438, "grad_norm": 0.357418954372406, "learning_rate": 5.967220833595531e-06, "loss": 0.0189, "step": 10398 }, { "epoch": 0.7146096756459593, "grad_norm": 0.4057054817676544, "learning_rate": 5.964555690184976e-06, "loss": 0.0114, "step": 10399 }, { "epoch": 0.7146783947223749, "grad_norm": 0.20936621725559235, "learning_rate": 5.961890994371752e-06, "loss": 0.0192, "step": 10400 }, { "epoch": 0.7147471137987905, "grad_norm": 0.6967874765396118, "learning_rate": 5.959226746287871e-06, "loss": 0.0203, "step": 10401 }, { "epoch": 0.7148158328752061, "grad_norm": 0.2273905873298645, "learning_rate": 5.956562946065305e-06, "loss": 0.0075, "step": 10402 }, { "epoch": 0.7148845519516218, "grad_norm": 0.7685882449150085, "learning_rate": 5.953899593836018e-06, "loss": 0.0187, "step": 10403 }, { "epoch": 0.7149532710280374, "grad_norm": 0.6321727633476257, "learning_rate": 5.951236689731953e-06, "loss": 0.0108, "step": 10404 }, { "epoch": 0.715021990104453, "grad_norm": 0.20191454887390137, "learning_rate": 5.9485742338850115e-06, "loss": 0.0168, "step": 10405 }, { "epoch": 0.7150907091808686, "grad_norm": 0.6834419369697571, "learning_rate": 5.945912226427101e-06, "loss": 0.0111, "step": 10406 }, { "epoch": 0.7151594282572842, "grad_norm": 0.42583534121513367, "learning_rate": 5.943250667490078e-06, "loss": 0.0132, "step": 10407 }, { "epoch": 0.7152281473336999, "grad_norm": 0.23513945937156677, "learning_rate": 5.940589557205798e-06, "loss": 0.0087, "step": 10408 }, { "epoch": 0.7152968664101155, "grad_norm": 0.4811619520187378, "learning_rate": 5.9379288957060865e-06, "loss": 0.0203, "step": 10409 }, { "epoch": 0.7153655854865311, "grad_norm": 0.2577616274356842, "learning_rate": 5.93526868312275e-06, "loss": 0.0098, "step": 10410 }, { "epoch": 0.7154343045629467, "grad_norm": 0.23735643923282623, "learning_rate": 5.932608919587562e-06, "loss": 0.0128, "step": 10411 }, { "epoch": 0.7155030236393622, "grad_norm": 0.21744410693645477, "learning_rate": 5.929949605232293e-06, "loss": 0.0059, "step": 10412 }, { "epoch": 0.715571742715778, "grad_norm": 1.0026628971099854, "learning_rate": 5.927290740188666e-06, "loss": 0.0092, "step": 10413 }, { "epoch": 0.7156404617921935, "grad_norm": 0.4689527153968811, "learning_rate": 5.924632324588409e-06, "loss": 0.0119, "step": 10414 }, { "epoch": 0.7157091808686091, "grad_norm": 1.5047175884246826, "learning_rate": 5.921974358563199e-06, "loss": 0.0182, "step": 10415 }, { "epoch": 0.7157778999450247, "grad_norm": 0.2693462371826172, "learning_rate": 5.919316842244725e-06, "loss": 0.0106, "step": 10416 }, { "epoch": 0.7158466190214403, "grad_norm": 0.4823143482208252, "learning_rate": 5.91665977576462e-06, "loss": 0.0161, "step": 10417 }, { "epoch": 0.715915338097856, "grad_norm": 0.277993768453598, "learning_rate": 5.914003159254521e-06, "loss": 0.0116, "step": 10418 }, { "epoch": 0.7159840571742716, "grad_norm": 0.5686436891555786, "learning_rate": 5.91134699284602e-06, "loss": 0.0133, "step": 10419 }, { "epoch": 0.7160527762506872, "grad_norm": 0.28681740164756775, "learning_rate": 5.9086912766707095e-06, "loss": 0.0106, "step": 10420 }, { "epoch": 0.7161214953271028, "grad_norm": 0.5735793113708496, "learning_rate": 5.906036010860136e-06, "loss": 0.0077, "step": 10421 }, { "epoch": 0.7161902144035184, "grad_norm": 0.5150352120399475, "learning_rate": 5.903381195545843e-06, "loss": 0.0168, "step": 10422 }, { "epoch": 0.7162589334799341, "grad_norm": 0.6440798044204712, "learning_rate": 5.900726830859349e-06, "loss": 0.0194, "step": 10423 }, { "epoch": 0.7163276525563497, "grad_norm": 1.2053124904632568, "learning_rate": 5.898072916932134e-06, "loss": 0.02, "step": 10424 }, { "epoch": 0.7163963716327653, "grad_norm": 0.2691040635108948, "learning_rate": 5.895419453895676e-06, "loss": 0.0121, "step": 10425 }, { "epoch": 0.7164650907091809, "grad_norm": 0.761017918586731, "learning_rate": 5.892766441881424e-06, "loss": 0.0106, "step": 10426 }, { "epoch": 0.7165338097855964, "grad_norm": 0.31048598885536194, "learning_rate": 5.8901138810207925e-06, "loss": 0.0145, "step": 10427 }, { "epoch": 0.7166025288620121, "grad_norm": 1.0342974662780762, "learning_rate": 5.8874617714451915e-06, "loss": 0.0214, "step": 10428 }, { "epoch": 0.7166712479384277, "grad_norm": 0.4222634732723236, "learning_rate": 5.8848101132860055e-06, "loss": 0.0115, "step": 10429 }, { "epoch": 0.7167399670148433, "grad_norm": 0.42119836807250977, "learning_rate": 5.88215890667458e-06, "loss": 0.0069, "step": 10430 }, { "epoch": 0.7168086860912589, "grad_norm": 0.5458720326423645, "learning_rate": 5.879508151742263e-06, "loss": 0.0224, "step": 10431 }, { "epoch": 0.7168774051676745, "grad_norm": 0.284507155418396, "learning_rate": 5.876857848620355e-06, "loss": 0.0097, "step": 10432 }, { "epoch": 0.7169461242440902, "grad_norm": 0.3299528658390045, "learning_rate": 5.8742079974401544e-06, "loss": 0.0127, "step": 10433 }, { "epoch": 0.7170148433205058, "grad_norm": 0.47147658467292786, "learning_rate": 5.871558598332925e-06, "loss": 0.011, "step": 10434 }, { "epoch": 0.7170835623969214, "grad_norm": 0.6345648169517517, "learning_rate": 5.868909651429923e-06, "loss": 0.0123, "step": 10435 }, { "epoch": 0.717152281473337, "grad_norm": 0.3525659143924713, "learning_rate": 5.866261156862358e-06, "loss": 0.0124, "step": 10436 }, { "epoch": 0.7172210005497526, "grad_norm": 0.5229187607765198, "learning_rate": 5.863613114761443e-06, "loss": 0.0086, "step": 10437 }, { "epoch": 0.7172897196261683, "grad_norm": 0.44445285201072693, "learning_rate": 5.860965525258345e-06, "loss": 0.0178, "step": 10438 }, { "epoch": 0.7173584387025839, "grad_norm": 0.6973019242286682, "learning_rate": 5.858318388484225e-06, "loss": 0.0103, "step": 10439 }, { "epoch": 0.7174271577789995, "grad_norm": 0.3152945339679718, "learning_rate": 5.855671704570224e-06, "loss": 0.0069, "step": 10440 }, { "epoch": 0.717495876855415, "grad_norm": 0.942308783531189, "learning_rate": 5.853025473647442e-06, "loss": 0.011, "step": 10441 }, { "epoch": 0.7175645959318306, "grad_norm": 0.329057902097702, "learning_rate": 5.850379695846971e-06, "loss": 0.0142, "step": 10442 }, { "epoch": 0.7176333150082463, "grad_norm": 0.4405422508716583, "learning_rate": 5.847734371299885e-06, "loss": 0.0121, "step": 10443 }, { "epoch": 0.7177020340846619, "grad_norm": 0.4847664535045624, "learning_rate": 5.845089500137216e-06, "loss": 0.0106, "step": 10444 }, { "epoch": 0.7177707531610775, "grad_norm": 0.21815325319766998, "learning_rate": 5.842445082489993e-06, "loss": 0.0141, "step": 10445 }, { "epoch": 0.7178394722374931, "grad_norm": 0.09572727233171463, "learning_rate": 5.839801118489218e-06, "loss": 0.0096, "step": 10446 }, { "epoch": 0.7179081913139087, "grad_norm": 0.3830336928367615, "learning_rate": 5.837157608265857e-06, "loss": 0.0126, "step": 10447 }, { "epoch": 0.7179769103903244, "grad_norm": 0.8898686170578003, "learning_rate": 5.834514551950876e-06, "loss": 0.0136, "step": 10448 }, { "epoch": 0.71804562946674, "grad_norm": 0.4043830335140228, "learning_rate": 5.831871949675195e-06, "loss": 0.0127, "step": 10449 }, { "epoch": 0.7181143485431556, "grad_norm": 0.2560460865497589, "learning_rate": 5.8292298015697356e-06, "loss": 0.0127, "step": 10450 }, { "epoch": 0.7181830676195712, "grad_norm": 0.9581272602081299, "learning_rate": 5.826588107765365e-06, "loss": 0.0132, "step": 10451 }, { "epoch": 0.7182517866959868, "grad_norm": 0.40281015634536743, "learning_rate": 5.823946868392973e-06, "loss": 0.0094, "step": 10452 }, { "epoch": 0.7183205057724025, "grad_norm": 0.21974708139896393, "learning_rate": 5.821306083583379e-06, "loss": 0.0063, "step": 10453 }, { "epoch": 0.7183892248488181, "grad_norm": 0.212550550699234, "learning_rate": 5.818665753467419e-06, "loss": 0.008, "step": 10454 }, { "epoch": 0.7184579439252337, "grad_norm": 0.67671799659729, "learning_rate": 5.816025878175876e-06, "loss": 0.012, "step": 10455 }, { "epoch": 0.7185266630016492, "grad_norm": 0.5683655738830566, "learning_rate": 5.813386457839534e-06, "loss": 0.019, "step": 10456 }, { "epoch": 0.7185953820780648, "grad_norm": 0.5117276310920715, "learning_rate": 5.810747492589129e-06, "loss": 0.0206, "step": 10457 }, { "epoch": 0.7186641011544805, "grad_norm": 0.4125891625881195, "learning_rate": 5.808108982555413e-06, "loss": 0.0099, "step": 10458 }, { "epoch": 0.7187328202308961, "grad_norm": 0.6028867959976196, "learning_rate": 5.805470927869074e-06, "loss": 0.01, "step": 10459 }, { "epoch": 0.7188015393073117, "grad_norm": 0.40897461771965027, "learning_rate": 5.802833328660807e-06, "loss": 0.0134, "step": 10460 }, { "epoch": 0.7188702583837273, "grad_norm": 0.9843996167182922, "learning_rate": 5.800196185061264e-06, "loss": 0.018, "step": 10461 }, { "epoch": 0.7189389774601429, "grad_norm": 0.6753772497177124, "learning_rate": 5.797559497201093e-06, "loss": 0.0133, "step": 10462 }, { "epoch": 0.7190076965365585, "grad_norm": 0.737552285194397, "learning_rate": 5.7949232652109e-06, "loss": 0.0131, "step": 10463 }, { "epoch": 0.7190764156129742, "grad_norm": 0.38545194268226624, "learning_rate": 5.792287489221284e-06, "loss": 0.0156, "step": 10464 }, { "epoch": 0.7191451346893898, "grad_norm": 0.43099749088287354, "learning_rate": 5.789652169362819e-06, "loss": 0.0128, "step": 10465 }, { "epoch": 0.7192138537658054, "grad_norm": 0.5666557550430298, "learning_rate": 5.787017305766046e-06, "loss": 0.0071, "step": 10466 }, { "epoch": 0.719282572842221, "grad_norm": 0.46749022603034973, "learning_rate": 5.7843828985615e-06, "loss": 0.0153, "step": 10467 }, { "epoch": 0.7193512919186366, "grad_norm": 0.6863183379173279, "learning_rate": 5.781748947879672e-06, "loss": 0.0154, "step": 10468 }, { "epoch": 0.7194200109950523, "grad_norm": 0.34494200348854065, "learning_rate": 5.779115453851051e-06, "loss": 0.0094, "step": 10469 }, { "epoch": 0.7194887300714679, "grad_norm": 0.44863763451576233, "learning_rate": 5.776482416606092e-06, "loss": 0.0132, "step": 10470 }, { "epoch": 0.7195574491478834, "grad_norm": 0.43389764428138733, "learning_rate": 5.773849836275237e-06, "loss": 0.013, "step": 10471 }, { "epoch": 0.719626168224299, "grad_norm": 0.31327858567237854, "learning_rate": 5.771217712988888e-06, "loss": 0.0082, "step": 10472 }, { "epoch": 0.7196948873007146, "grad_norm": 0.42958372831344604, "learning_rate": 5.7685860468774445e-06, "loss": 0.0119, "step": 10473 }, { "epoch": 0.7197636063771303, "grad_norm": 0.7098437547683716, "learning_rate": 5.7659548380712645e-06, "loss": 0.0134, "step": 10474 }, { "epoch": 0.7198323254535459, "grad_norm": 0.1740584522485733, "learning_rate": 5.763324086700698e-06, "loss": 0.009, "step": 10475 }, { "epoch": 0.7199010445299615, "grad_norm": 0.14886534214019775, "learning_rate": 5.760693792896064e-06, "loss": 0.0047, "step": 10476 }, { "epoch": 0.7199697636063771, "grad_norm": 0.3345790207386017, "learning_rate": 5.758063956787671e-06, "loss": 0.02, "step": 10477 }, { "epoch": 0.7200384826827927, "grad_norm": 0.3756401240825653, "learning_rate": 5.7554345785057835e-06, "loss": 0.0137, "step": 10478 }, { "epoch": 0.7201072017592084, "grad_norm": 0.5409951210021973, "learning_rate": 5.752805658180667e-06, "loss": 0.0114, "step": 10479 }, { "epoch": 0.720175920835624, "grad_norm": 0.6853671669960022, "learning_rate": 5.750177195942542e-06, "loss": 0.0124, "step": 10480 }, { "epoch": 0.7202446399120396, "grad_norm": 0.303301066160202, "learning_rate": 5.747549191921621e-06, "loss": 0.0112, "step": 10481 }, { "epoch": 0.7203133589884552, "grad_norm": 0.5586467385292053, "learning_rate": 5.744921646248095e-06, "loss": 0.0171, "step": 10482 }, { "epoch": 0.7203820780648708, "grad_norm": 0.7520442605018616, "learning_rate": 5.74229455905212e-06, "loss": 0.0105, "step": 10483 }, { "epoch": 0.7204507971412865, "grad_norm": 0.29098930954933167, "learning_rate": 5.7396679304638435e-06, "loss": 0.025, "step": 10484 }, { "epoch": 0.720519516217702, "grad_norm": 0.5013243556022644, "learning_rate": 5.737041760613373e-06, "loss": 0.01, "step": 10485 }, { "epoch": 0.7205882352941176, "grad_norm": 0.46692854166030884, "learning_rate": 5.7344160496308165e-06, "loss": 0.0098, "step": 10486 }, { "epoch": 0.7206569543705332, "grad_norm": 0.20916903018951416, "learning_rate": 5.73179079764623e-06, "loss": 0.011, "step": 10487 }, { "epoch": 0.7207256734469488, "grad_norm": 0.324709415435791, "learning_rate": 5.729166004789681e-06, "loss": 0.0081, "step": 10488 }, { "epoch": 0.7207943925233645, "grad_norm": 0.3441244959831238, "learning_rate": 5.726541671191185e-06, "loss": 0.008, "step": 10489 }, { "epoch": 0.7208631115997801, "grad_norm": 0.4903099536895752, "learning_rate": 5.7239177969807525e-06, "loss": 0.0101, "step": 10490 }, { "epoch": 0.7209318306761957, "grad_norm": 0.23494049906730652, "learning_rate": 5.721294382288357e-06, "loss": 0.0081, "step": 10491 }, { "epoch": 0.7210005497526113, "grad_norm": 0.48439615964889526, "learning_rate": 5.718671427243968e-06, "loss": 0.0165, "step": 10492 }, { "epoch": 0.7210692688290269, "grad_norm": 0.5904365181922913, "learning_rate": 5.716048931977504e-06, "loss": 0.0081, "step": 10493 }, { "epoch": 0.7211379879054426, "grad_norm": 0.6822746396064758, "learning_rate": 5.7134268966188986e-06, "loss": 0.0202, "step": 10494 }, { "epoch": 0.7212067069818582, "grad_norm": 0.3302032947540283, "learning_rate": 5.710805321298027e-06, "loss": 0.0131, "step": 10495 }, { "epoch": 0.7212754260582738, "grad_norm": 0.916759192943573, "learning_rate": 5.708184206144767e-06, "loss": 0.0121, "step": 10496 }, { "epoch": 0.7213441451346894, "grad_norm": 0.5112080574035645, "learning_rate": 5.705563551288954e-06, "loss": 0.0095, "step": 10497 }, { "epoch": 0.721412864211105, "grad_norm": 0.6437450647354126, "learning_rate": 5.70294335686042e-06, "loss": 0.0108, "step": 10498 }, { "epoch": 0.7214815832875207, "grad_norm": 0.5840981602668762, "learning_rate": 5.7003236229889506e-06, "loss": 0.0165, "step": 10499 }, { "epoch": 0.7215503023639362, "grad_norm": 0.3079482614994049, "learning_rate": 5.69770434980433e-06, "loss": 0.0106, "step": 10500 }, { "epoch": 0.7216190214403518, "grad_norm": 0.3004269003868103, "learning_rate": 5.695085537436318e-06, "loss": 0.0082, "step": 10501 }, { "epoch": 0.7216877405167674, "grad_norm": 0.30685657262802124, "learning_rate": 5.692467186014632e-06, "loss": 0.01, "step": 10502 }, { "epoch": 0.721756459593183, "grad_norm": 0.513639509677887, "learning_rate": 5.689849295668992e-06, "loss": 0.0148, "step": 10503 }, { "epoch": 0.7218251786695987, "grad_norm": 0.37895461916923523, "learning_rate": 5.687231866529072e-06, "loss": 0.0126, "step": 10504 }, { "epoch": 0.7218938977460143, "grad_norm": 0.7226516604423523, "learning_rate": 5.684614898724539e-06, "loss": 0.0103, "step": 10505 }, { "epoch": 0.7219626168224299, "grad_norm": 0.2670770585536957, "learning_rate": 5.681998392385031e-06, "loss": 0.0135, "step": 10506 }, { "epoch": 0.7220313358988455, "grad_norm": 0.20464469492435455, "learning_rate": 5.679382347640173e-06, "loss": 0.0093, "step": 10507 }, { "epoch": 0.7221000549752611, "grad_norm": 0.5008825063705444, "learning_rate": 5.676766764619543e-06, "loss": 0.0067, "step": 10508 }, { "epoch": 0.7221687740516768, "grad_norm": 0.6000626087188721, "learning_rate": 5.674151643452728e-06, "loss": 0.013, "step": 10509 }, { "epoch": 0.7222374931280924, "grad_norm": 0.29387959837913513, "learning_rate": 5.67153698426926e-06, "loss": 0.0084, "step": 10510 }, { "epoch": 0.722306212204508, "grad_norm": 0.41541656851768494, "learning_rate": 5.668922787198672e-06, "loss": 0.0166, "step": 10511 }, { "epoch": 0.7223749312809236, "grad_norm": 0.375164270401001, "learning_rate": 5.666309052370467e-06, "loss": 0.0113, "step": 10512 }, { "epoch": 0.7224436503573392, "grad_norm": 0.4342944025993347, "learning_rate": 5.663695779914126e-06, "loss": 0.0098, "step": 10513 }, { "epoch": 0.7225123694337549, "grad_norm": 0.37281379103660583, "learning_rate": 5.661082969959096e-06, "loss": 0.0114, "step": 10514 }, { "epoch": 0.7225810885101704, "grad_norm": 0.2991156280040741, "learning_rate": 5.658470622634821e-06, "loss": 0.0107, "step": 10515 }, { "epoch": 0.722649807586586, "grad_norm": 0.13968314230442047, "learning_rate": 5.655858738070701e-06, "loss": 0.0077, "step": 10516 }, { "epoch": 0.7227185266630016, "grad_norm": 0.492512971162796, "learning_rate": 5.653247316396128e-06, "loss": 0.019, "step": 10517 }, { "epoch": 0.7227872457394172, "grad_norm": 0.7352394461631775, "learning_rate": 5.650636357740473e-06, "loss": 0.0246, "step": 10518 }, { "epoch": 0.7228559648158329, "grad_norm": 0.19763989746570587, "learning_rate": 5.6480258622330645e-06, "loss": 0.012, "step": 10519 }, { "epoch": 0.7229246838922485, "grad_norm": 0.2840673625469208, "learning_rate": 5.645415830003234e-06, "loss": 0.0123, "step": 10520 }, { "epoch": 0.7229934029686641, "grad_norm": 0.48435667157173157, "learning_rate": 5.642806261180266e-06, "loss": 0.0231, "step": 10521 }, { "epoch": 0.7230621220450797, "grad_norm": 0.23886753618717194, "learning_rate": 5.640197155893444e-06, "loss": 0.0082, "step": 10522 }, { "epoch": 0.7231308411214953, "grad_norm": 0.5737016797065735, "learning_rate": 5.637588514272006e-06, "loss": 0.0095, "step": 10523 }, { "epoch": 0.723199560197911, "grad_norm": 0.45144781470298767, "learning_rate": 5.6349803364451845e-06, "loss": 0.0141, "step": 10524 }, { "epoch": 0.7232682792743266, "grad_norm": 0.5816461443901062, "learning_rate": 5.632372622542185e-06, "loss": 0.0161, "step": 10525 }, { "epoch": 0.7233369983507422, "grad_norm": 0.6957719922065735, "learning_rate": 5.62976537269219e-06, "loss": 0.0288, "step": 10526 }, { "epoch": 0.7234057174271578, "grad_norm": 0.47763410210609436, "learning_rate": 5.6271585870243496e-06, "loss": 0.0121, "step": 10527 }, { "epoch": 0.7234744365035733, "grad_norm": 0.31933191418647766, "learning_rate": 5.624552265667809e-06, "loss": 0.0093, "step": 10528 }, { "epoch": 0.723543155579989, "grad_norm": 0.46443161368370056, "learning_rate": 5.621946408751668e-06, "loss": 0.0123, "step": 10529 }, { "epoch": 0.7236118746564046, "grad_norm": 0.405879944562912, "learning_rate": 5.619341016405022e-06, "loss": 0.0169, "step": 10530 }, { "epoch": 0.7236805937328202, "grad_norm": 0.4034070074558258, "learning_rate": 5.616736088756935e-06, "loss": 0.0101, "step": 10531 }, { "epoch": 0.7237493128092358, "grad_norm": 0.3333277404308319, "learning_rate": 5.614131625936457e-06, "loss": 0.0096, "step": 10532 }, { "epoch": 0.7238180318856514, "grad_norm": 0.4234551787376404, "learning_rate": 5.611527628072596e-06, "loss": 0.0093, "step": 10533 }, { "epoch": 0.7238867509620671, "grad_norm": 0.3855198323726654, "learning_rate": 5.608924095294361e-06, "loss": 0.0088, "step": 10534 }, { "epoch": 0.7239554700384827, "grad_norm": 0.5566185116767883, "learning_rate": 5.606321027730713e-06, "loss": 0.012, "step": 10535 }, { "epoch": 0.7240241891148983, "grad_norm": 1.2090023756027222, "learning_rate": 5.603718425510608e-06, "loss": 0.0158, "step": 10536 }, { "epoch": 0.7240929081913139, "grad_norm": 0.12158458679914474, "learning_rate": 5.6011162887629804e-06, "loss": 0.0097, "step": 10537 }, { "epoch": 0.7241616272677295, "grad_norm": 0.43009307980537415, "learning_rate": 5.598514617616724e-06, "loss": 0.0069, "step": 10538 }, { "epoch": 0.7242303463441452, "grad_norm": 0.5177533030509949, "learning_rate": 5.595913412200729e-06, "loss": 0.012, "step": 10539 }, { "epoch": 0.7242990654205608, "grad_norm": 0.453869104385376, "learning_rate": 5.593312672643845e-06, "loss": 0.0071, "step": 10540 }, { "epoch": 0.7243677844969764, "grad_norm": 0.1062479168176651, "learning_rate": 5.590712399074915e-06, "loss": 0.0108, "step": 10541 }, { "epoch": 0.724436503573392, "grad_norm": 0.537409782409668, "learning_rate": 5.588112591622739e-06, "loss": 0.0109, "step": 10542 }, { "epoch": 0.7245052226498075, "grad_norm": 0.548851728439331, "learning_rate": 5.5855132504161275e-06, "loss": 0.0124, "step": 10543 }, { "epoch": 0.7245739417262232, "grad_norm": 0.19314055144786835, "learning_rate": 5.582914375583827e-06, "loss": 0.0104, "step": 10544 }, { "epoch": 0.7246426608026388, "grad_norm": 0.4985891878604889, "learning_rate": 5.580315967254594e-06, "loss": 0.0121, "step": 10545 }, { "epoch": 0.7247113798790544, "grad_norm": 0.37827467918395996, "learning_rate": 5.5777180255571355e-06, "loss": 0.0078, "step": 10546 }, { "epoch": 0.72478009895547, "grad_norm": 0.8024190664291382, "learning_rate": 5.57512055062016e-06, "loss": 0.0139, "step": 10547 }, { "epoch": 0.7248488180318856, "grad_norm": 0.552195131778717, "learning_rate": 5.572523542572326e-06, "loss": 0.0141, "step": 10548 }, { "epoch": 0.7249175371083013, "grad_norm": 0.4459037482738495, "learning_rate": 5.5699270015423025e-06, "loss": 0.0151, "step": 10549 }, { "epoch": 0.7249862561847169, "grad_norm": 0.34609436988830566, "learning_rate": 5.567330927658705e-06, "loss": 0.0085, "step": 10550 }, { "epoch": 0.7250549752611325, "grad_norm": 0.7856890559196472, "learning_rate": 5.564735321050144e-06, "loss": 0.0079, "step": 10551 }, { "epoch": 0.7251236943375481, "grad_norm": 0.4516204595565796, "learning_rate": 5.562140181845192e-06, "loss": 0.0097, "step": 10552 }, { "epoch": 0.7251924134139637, "grad_norm": 0.3886930048465729, "learning_rate": 5.559545510172418e-06, "loss": 0.008, "step": 10553 }, { "epoch": 0.7252611324903794, "grad_norm": 0.24765713512897491, "learning_rate": 5.5569513061603455e-06, "loss": 0.0101, "step": 10554 }, { "epoch": 0.725329851566795, "grad_norm": 0.39180293679237366, "learning_rate": 5.55435756993749e-06, "loss": 0.0087, "step": 10555 }, { "epoch": 0.7253985706432106, "grad_norm": 0.29946884512901306, "learning_rate": 5.551764301632348e-06, "loss": 0.0125, "step": 10556 }, { "epoch": 0.7254672897196262, "grad_norm": 0.20234598219394684, "learning_rate": 5.549171501373371e-06, "loss": 0.0087, "step": 10557 }, { "epoch": 0.7255360087960417, "grad_norm": 0.27965492010116577, "learning_rate": 5.546579169289015e-06, "loss": 0.0167, "step": 10558 }, { "epoch": 0.7256047278724574, "grad_norm": 0.378780335187912, "learning_rate": 5.543987305507685e-06, "loss": 0.0123, "step": 10559 }, { "epoch": 0.725673446948873, "grad_norm": 0.27937576174736023, "learning_rate": 5.5413959101577825e-06, "loss": 0.0144, "step": 10560 }, { "epoch": 0.7257421660252886, "grad_norm": 0.6939685344696045, "learning_rate": 5.538804983367682e-06, "loss": 0.0138, "step": 10561 }, { "epoch": 0.7258108851017042, "grad_norm": 0.8819000720977783, "learning_rate": 5.536214525265735e-06, "loss": 0.0207, "step": 10562 }, { "epoch": 0.7258796041781198, "grad_norm": 0.5838645100593567, "learning_rate": 5.533624535980258e-06, "loss": 0.0087, "step": 10563 }, { "epoch": 0.7259483232545355, "grad_norm": 0.2296055406332016, "learning_rate": 5.531035015639566e-06, "loss": 0.0089, "step": 10564 }, { "epoch": 0.7260170423309511, "grad_norm": 0.26366016268730164, "learning_rate": 5.528445964371927e-06, "loss": 0.0095, "step": 10565 }, { "epoch": 0.7260857614073667, "grad_norm": 0.37540411949157715, "learning_rate": 5.5258573823056015e-06, "loss": 0.0087, "step": 10566 }, { "epoch": 0.7261544804837823, "grad_norm": 0.5716826915740967, "learning_rate": 5.523269269568824e-06, "loss": 0.0148, "step": 10567 }, { "epoch": 0.7262231995601979, "grad_norm": 0.9015947580337524, "learning_rate": 5.520681626289808e-06, "loss": 0.0087, "step": 10568 }, { "epoch": 0.7262919186366136, "grad_norm": 0.6144538521766663, "learning_rate": 5.518094452596731e-06, "loss": 0.0153, "step": 10569 }, { "epoch": 0.7263606377130292, "grad_norm": 0.6775543689727783, "learning_rate": 5.515507748617766e-06, "loss": 0.0099, "step": 10570 }, { "epoch": 0.7264293567894448, "grad_norm": 0.8539833426475525, "learning_rate": 5.512921514481042e-06, "loss": 0.0133, "step": 10571 }, { "epoch": 0.7264980758658603, "grad_norm": 0.4292084276676178, "learning_rate": 5.510335750314682e-06, "loss": 0.0101, "step": 10572 }, { "epoch": 0.7265667949422759, "grad_norm": 0.7246969938278198, "learning_rate": 5.5077504562467855e-06, "loss": 0.0162, "step": 10573 }, { "epoch": 0.7266355140186916, "grad_norm": 0.34722763299942017, "learning_rate": 5.505165632405408e-06, "loss": 0.0114, "step": 10574 }, { "epoch": 0.7267042330951072, "grad_norm": 0.4911036491394043, "learning_rate": 5.502581278918613e-06, "loss": 0.0125, "step": 10575 }, { "epoch": 0.7267729521715228, "grad_norm": 0.8426854610443115, "learning_rate": 5.499997395914407e-06, "loss": 0.0149, "step": 10576 }, { "epoch": 0.7268416712479384, "grad_norm": 0.0940634086728096, "learning_rate": 5.497413983520806e-06, "loss": 0.0065, "step": 10577 }, { "epoch": 0.726910390324354, "grad_norm": 0.6989172101020813, "learning_rate": 5.494831041865769e-06, "loss": 0.0135, "step": 10578 }, { "epoch": 0.7269791094007697, "grad_norm": 0.38315629959106445, "learning_rate": 5.492248571077271e-06, "loss": 0.0107, "step": 10579 }, { "epoch": 0.7270478284771853, "grad_norm": 0.45619213581085205, "learning_rate": 5.489666571283227e-06, "loss": 0.0136, "step": 10580 }, { "epoch": 0.7271165475536009, "grad_norm": 0.4000439941883087, "learning_rate": 5.487085042611553e-06, "loss": 0.0115, "step": 10581 }, { "epoch": 0.7271852666300165, "grad_norm": 0.30312609672546387, "learning_rate": 5.4845039851901234e-06, "loss": 0.006, "step": 10582 }, { "epoch": 0.7272539857064321, "grad_norm": 0.33631956577301025, "learning_rate": 5.481923399146807e-06, "loss": 0.0094, "step": 10583 }, { "epoch": 0.7273227047828478, "grad_norm": 0.3748225271701813, "learning_rate": 5.4793432846094284e-06, "loss": 0.0119, "step": 10584 }, { "epoch": 0.7273914238592634, "grad_norm": 0.661393940448761, "learning_rate": 5.476763641705819e-06, "loss": 0.0117, "step": 10585 }, { "epoch": 0.727460142935679, "grad_norm": 0.3484888970851898, "learning_rate": 5.474184470563758e-06, "loss": 0.0124, "step": 10586 }, { "epoch": 0.7275288620120945, "grad_norm": 0.7210265398025513, "learning_rate": 5.471605771311015e-06, "loss": 0.0201, "step": 10587 }, { "epoch": 0.7275975810885101, "grad_norm": 0.2946484088897705, "learning_rate": 5.46902754407533e-06, "loss": 0.0065, "step": 10588 }, { "epoch": 0.7276663001649258, "grad_norm": 0.2957562506198883, "learning_rate": 5.466449788984431e-06, "loss": 0.0069, "step": 10589 }, { "epoch": 0.7277350192413414, "grad_norm": 0.5910353660583496, "learning_rate": 5.463872506166003e-06, "loss": 0.0072, "step": 10590 }, { "epoch": 0.727803738317757, "grad_norm": 0.1718416064977646, "learning_rate": 5.461295695747726e-06, "loss": 0.0177, "step": 10591 }, { "epoch": 0.7278724573941726, "grad_norm": 0.3923473358154297, "learning_rate": 5.458719357857254e-06, "loss": 0.0095, "step": 10592 }, { "epoch": 0.7279411764705882, "grad_norm": 0.5931156277656555, "learning_rate": 5.456143492622203e-06, "loss": 0.0099, "step": 10593 }, { "epoch": 0.7280098955470038, "grad_norm": 0.7629978656768799, "learning_rate": 5.453568100170189e-06, "loss": 0.0157, "step": 10594 }, { "epoch": 0.7280786146234195, "grad_norm": 0.33630961179733276, "learning_rate": 5.450993180628778e-06, "loss": 0.0112, "step": 10595 }, { "epoch": 0.7281473336998351, "grad_norm": 0.7284420132637024, "learning_rate": 5.448418734125535e-06, "loss": 0.0114, "step": 10596 }, { "epoch": 0.7282160527762507, "grad_norm": 0.40897518396377563, "learning_rate": 5.44584476078799e-06, "loss": 0.0066, "step": 10597 }, { "epoch": 0.7282847718526663, "grad_norm": 0.8196414113044739, "learning_rate": 5.4432712607436555e-06, "loss": 0.0086, "step": 10598 }, { "epoch": 0.7283534909290819, "grad_norm": 0.0964834913611412, "learning_rate": 5.440698234120013e-06, "loss": 0.0105, "step": 10599 }, { "epoch": 0.7284222100054976, "grad_norm": 0.33800384402275085, "learning_rate": 5.438125681044531e-06, "loss": 0.0103, "step": 10600 }, { "epoch": 0.7284909290819132, "grad_norm": 0.7664501070976257, "learning_rate": 5.4355536016446385e-06, "loss": 0.0143, "step": 10601 }, { "epoch": 0.7285596481583287, "grad_norm": 0.08623761683702469, "learning_rate": 5.432981996047759e-06, "loss": 0.006, "step": 10602 }, { "epoch": 0.7286283672347443, "grad_norm": 0.21112124621868134, "learning_rate": 5.43041086438128e-06, "loss": 0.013, "step": 10603 }, { "epoch": 0.7286970863111599, "grad_norm": 0.4235098958015442, "learning_rate": 5.42784020677258e-06, "loss": 0.0061, "step": 10604 }, { "epoch": 0.7287658053875756, "grad_norm": 0.5739336609840393, "learning_rate": 5.425270023348989e-06, "loss": 0.0117, "step": 10605 }, { "epoch": 0.7288345244639912, "grad_norm": 0.852222740650177, "learning_rate": 5.4227003142378425e-06, "loss": 0.0131, "step": 10606 }, { "epoch": 0.7289032435404068, "grad_norm": 0.2958965301513672, "learning_rate": 5.4201310795664285e-06, "loss": 0.0104, "step": 10607 }, { "epoch": 0.7289719626168224, "grad_norm": 0.328508198261261, "learning_rate": 5.417562319462024e-06, "loss": 0.0128, "step": 10608 }, { "epoch": 0.729040681693238, "grad_norm": 0.6857260465621948, "learning_rate": 5.4149940340518875e-06, "loss": 0.0077, "step": 10609 }, { "epoch": 0.7291094007696537, "grad_norm": 0.3016790747642517, "learning_rate": 5.412426223463234e-06, "loss": 0.0082, "step": 10610 }, { "epoch": 0.7291781198460693, "grad_norm": 0.3129637837409973, "learning_rate": 5.4098588878232806e-06, "loss": 0.0128, "step": 10611 }, { "epoch": 0.7292468389224849, "grad_norm": 0.529735803604126, "learning_rate": 5.407292027259195e-06, "loss": 0.0115, "step": 10612 }, { "epoch": 0.7293155579989005, "grad_norm": 0.5557640790939331, "learning_rate": 5.404725641898147e-06, "loss": 0.0155, "step": 10613 }, { "epoch": 0.7293842770753161, "grad_norm": 0.7045699954032898, "learning_rate": 5.4021597318672544e-06, "loss": 0.0156, "step": 10614 }, { "epoch": 0.7294529961517318, "grad_norm": 0.24192361533641815, "learning_rate": 5.399594297293645e-06, "loss": 0.0087, "step": 10615 }, { "epoch": 0.7295217152281473, "grad_norm": 0.18983232975006104, "learning_rate": 5.397029338304391e-06, "loss": 0.0076, "step": 10616 }, { "epoch": 0.7295904343045629, "grad_norm": 0.3063275218009949, "learning_rate": 5.394464855026567e-06, "loss": 0.0067, "step": 10617 }, { "epoch": 0.7296591533809785, "grad_norm": 0.2902451753616333, "learning_rate": 5.391900847587201e-06, "loss": 0.0081, "step": 10618 }, { "epoch": 0.7297278724573941, "grad_norm": 0.31079044938087463, "learning_rate": 5.389337316113317e-06, "loss": 0.0068, "step": 10619 }, { "epoch": 0.7297965915338098, "grad_norm": 0.25333157181739807, "learning_rate": 5.386774260731895e-06, "loss": 0.008, "step": 10620 }, { "epoch": 0.7298653106102254, "grad_norm": 0.4694253206253052, "learning_rate": 5.3842116815699214e-06, "loss": 0.0145, "step": 10621 }, { "epoch": 0.729934029686641, "grad_norm": 0.38811156153678894, "learning_rate": 5.381649578754328e-06, "loss": 0.0171, "step": 10622 }, { "epoch": 0.7300027487630566, "grad_norm": 0.5519775152206421, "learning_rate": 5.379087952412044e-06, "loss": 0.0139, "step": 10623 }, { "epoch": 0.7300714678394722, "grad_norm": 0.2133069932460785, "learning_rate": 5.376526802669958e-06, "loss": 0.0082, "step": 10624 }, { "epoch": 0.7301401869158879, "grad_norm": 0.37439560890197754, "learning_rate": 5.373966129654954e-06, "loss": 0.0114, "step": 10625 }, { "epoch": 0.7302089059923035, "grad_norm": 0.6576544046401978, "learning_rate": 5.371405933493874e-06, "loss": 0.0159, "step": 10626 }, { "epoch": 0.7302776250687191, "grad_norm": 0.6014607548713684, "learning_rate": 5.3688462143135465e-06, "loss": 0.0186, "step": 10627 }, { "epoch": 0.7303463441451347, "grad_norm": 0.6054367423057556, "learning_rate": 5.366286972240784e-06, "loss": 0.0132, "step": 10628 }, { "epoch": 0.7304150632215503, "grad_norm": 0.18338283896446228, "learning_rate": 5.363728207402353e-06, "loss": 0.0034, "step": 10629 }, { "epoch": 0.730483782297966, "grad_norm": 0.49190497398376465, "learning_rate": 5.3611699199250145e-06, "loss": 0.015, "step": 10630 }, { "epoch": 0.7305525013743815, "grad_norm": 0.18291006982326508, "learning_rate": 5.358612109935506e-06, "loss": 0.0131, "step": 10631 }, { "epoch": 0.7306212204507971, "grad_norm": 0.8373483419418335, "learning_rate": 5.356054777560529e-06, "loss": 0.0128, "step": 10632 }, { "epoch": 0.7306899395272127, "grad_norm": 0.4882737696170807, "learning_rate": 5.35349792292677e-06, "loss": 0.0122, "step": 10633 }, { "epoch": 0.7307586586036283, "grad_norm": 0.6199215650558472, "learning_rate": 5.350941546160896e-06, "loss": 0.0157, "step": 10634 }, { "epoch": 0.730827377680044, "grad_norm": 0.353059858083725, "learning_rate": 5.348385647389535e-06, "loss": 0.0152, "step": 10635 }, { "epoch": 0.7308960967564596, "grad_norm": 0.39270585775375366, "learning_rate": 5.345830226739313e-06, "loss": 0.0112, "step": 10636 }, { "epoch": 0.7309648158328752, "grad_norm": 0.15583091974258423, "learning_rate": 5.343275284336807e-06, "loss": 0.0084, "step": 10637 }, { "epoch": 0.7310335349092908, "grad_norm": 0.24885490536689758, "learning_rate": 5.34072082030859e-06, "loss": 0.0166, "step": 10638 }, { "epoch": 0.7311022539857064, "grad_norm": 0.6202653646469116, "learning_rate": 5.3381668347812055e-06, "loss": 0.0171, "step": 10639 }, { "epoch": 0.7311709730621221, "grad_norm": 0.44765910506248474, "learning_rate": 5.335613327881177e-06, "loss": 0.0099, "step": 10640 }, { "epoch": 0.7312396921385377, "grad_norm": 0.32688555121421814, "learning_rate": 5.333060299734989e-06, "loss": 0.0144, "step": 10641 }, { "epoch": 0.7313084112149533, "grad_norm": 0.2666964828968048, "learning_rate": 5.330507750469125e-06, "loss": 0.0096, "step": 10642 }, { "epoch": 0.7313771302913689, "grad_norm": 0.4460969567298889, "learning_rate": 5.327955680210023e-06, "loss": 0.0114, "step": 10643 }, { "epoch": 0.7314458493677845, "grad_norm": 0.2919495403766632, "learning_rate": 5.325404089084112e-06, "loss": 0.0134, "step": 10644 }, { "epoch": 0.7315145684442002, "grad_norm": 0.43678346276283264, "learning_rate": 5.322852977217796e-06, "loss": 0.0187, "step": 10645 }, { "epoch": 0.7315832875206157, "grad_norm": 0.4587351977825165, "learning_rate": 5.320302344737445e-06, "loss": 0.0073, "step": 10646 }, { "epoch": 0.7316520065970313, "grad_norm": 0.4286903738975525, "learning_rate": 5.3177521917694134e-06, "loss": 0.0082, "step": 10647 }, { "epoch": 0.7317207256734469, "grad_norm": 0.30915093421936035, "learning_rate": 5.315202518440039e-06, "loss": 0.0093, "step": 10648 }, { "epoch": 0.7317894447498625, "grad_norm": 0.21467018127441406, "learning_rate": 5.312653324875614e-06, "loss": 0.0061, "step": 10649 }, { "epoch": 0.7318581638262782, "grad_norm": 0.24701344966888428, "learning_rate": 5.310104611202427e-06, "loss": 0.0091, "step": 10650 }, { "epoch": 0.7319268829026938, "grad_norm": 0.6158477663993835, "learning_rate": 5.307556377546743e-06, "loss": 0.0171, "step": 10651 }, { "epoch": 0.7319956019791094, "grad_norm": 0.3656156659126282, "learning_rate": 5.305008624034782e-06, "loss": 0.0084, "step": 10652 }, { "epoch": 0.732064321055525, "grad_norm": 0.326288640499115, "learning_rate": 5.302461350792768e-06, "loss": 0.011, "step": 10653 }, { "epoch": 0.7321330401319406, "grad_norm": 0.20373481512069702, "learning_rate": 5.299914557946877e-06, "loss": 0.0042, "step": 10654 }, { "epoch": 0.7322017592083563, "grad_norm": 0.2617114186286926, "learning_rate": 5.297368245623279e-06, "loss": 0.0084, "step": 10655 }, { "epoch": 0.7322704782847719, "grad_norm": 0.41383010149002075, "learning_rate": 5.294822413948101e-06, "loss": 0.0092, "step": 10656 }, { "epoch": 0.7323391973611875, "grad_norm": 0.369058758020401, "learning_rate": 5.29227706304748e-06, "loss": 0.0195, "step": 10657 }, { "epoch": 0.7324079164376031, "grad_norm": 0.32412704825401306, "learning_rate": 5.289732193047488e-06, "loss": 0.0181, "step": 10658 }, { "epoch": 0.7324766355140186, "grad_norm": 0.35648313164711, "learning_rate": 5.2871878040742045e-06, "loss": 0.0119, "step": 10659 }, { "epoch": 0.7325453545904343, "grad_norm": 0.14656184613704681, "learning_rate": 5.284643896253664e-06, "loss": 0.0069, "step": 10660 }, { "epoch": 0.7326140736668499, "grad_norm": 0.23208948969841003, "learning_rate": 5.282100469711896e-06, "loss": 0.018, "step": 10661 }, { "epoch": 0.7326827927432655, "grad_norm": 0.26248636841773987, "learning_rate": 5.279557524574887e-06, "loss": 0.0075, "step": 10662 }, { "epoch": 0.7327515118196811, "grad_norm": 0.15558183193206787, "learning_rate": 5.277015060968612e-06, "loss": 0.0078, "step": 10663 }, { "epoch": 0.7328202308960967, "grad_norm": 0.1333087682723999, "learning_rate": 5.274473079019025e-06, "loss": 0.0124, "step": 10664 }, { "epoch": 0.7328889499725124, "grad_norm": 0.26997286081314087, "learning_rate": 5.2719315788520485e-06, "loss": 0.0078, "step": 10665 }, { "epoch": 0.732957669048928, "grad_norm": 0.7430102825164795, "learning_rate": 5.269390560593577e-06, "loss": 0.0213, "step": 10666 }, { "epoch": 0.7330263881253436, "grad_norm": 0.31447306275367737, "learning_rate": 5.266850024369498e-06, "loss": 0.0175, "step": 10667 }, { "epoch": 0.7330951072017592, "grad_norm": 0.3561668395996094, "learning_rate": 5.264309970305653e-06, "loss": 0.0131, "step": 10668 }, { "epoch": 0.7331638262781748, "grad_norm": 0.31483888626098633, "learning_rate": 5.261770398527876e-06, "loss": 0.0131, "step": 10669 }, { "epoch": 0.7332325453545905, "grad_norm": 0.3679930865764618, "learning_rate": 5.259231309161977e-06, "loss": 0.0075, "step": 10670 }, { "epoch": 0.7333012644310061, "grad_norm": 0.5386756062507629, "learning_rate": 5.256692702333728e-06, "loss": 0.0126, "step": 10671 }, { "epoch": 0.7333699835074217, "grad_norm": 0.06336624920368195, "learning_rate": 5.254154578168896e-06, "loss": 0.0155, "step": 10672 }, { "epoch": 0.7334387025838373, "grad_norm": 0.23724885284900665, "learning_rate": 5.251616936793205e-06, "loss": 0.007, "step": 10673 }, { "epoch": 0.7335074216602528, "grad_norm": 0.18608486652374268, "learning_rate": 5.24907977833237e-06, "loss": 0.0077, "step": 10674 }, { "epoch": 0.7335761407366685, "grad_norm": 0.2238232046365738, "learning_rate": 5.2465431029120744e-06, "loss": 0.0104, "step": 10675 }, { "epoch": 0.7336448598130841, "grad_norm": 0.15855655074119568, "learning_rate": 5.244006910657986e-06, "loss": 0.0145, "step": 10676 }, { "epoch": 0.7337135788894997, "grad_norm": 0.13243962824344635, "learning_rate": 5.241471201695735e-06, "loss": 0.0044, "step": 10677 }, { "epoch": 0.7337822979659153, "grad_norm": 0.270853728055954, "learning_rate": 5.23893597615094e-06, "loss": 0.0099, "step": 10678 }, { "epoch": 0.7338510170423309, "grad_norm": 0.3006693422794342, "learning_rate": 5.236401234149185e-06, "loss": 0.0099, "step": 10679 }, { "epoch": 0.7339197361187466, "grad_norm": 0.3472692370414734, "learning_rate": 5.23386697581604e-06, "loss": 0.0115, "step": 10680 }, { "epoch": 0.7339884551951622, "grad_norm": 0.33515334129333496, "learning_rate": 5.231333201277048e-06, "loss": 0.0103, "step": 10681 }, { "epoch": 0.7340571742715778, "grad_norm": 0.14585474133491516, "learning_rate": 5.228799910657729e-06, "loss": 0.0088, "step": 10682 }, { "epoch": 0.7341258933479934, "grad_norm": 0.49427786469459534, "learning_rate": 5.226267104083569e-06, "loss": 0.0174, "step": 10683 }, { "epoch": 0.734194612424409, "grad_norm": 0.20314081013202667, "learning_rate": 5.223734781680048e-06, "loss": 0.0108, "step": 10684 }, { "epoch": 0.7342633315008247, "grad_norm": 0.139018252491951, "learning_rate": 5.221202943572601e-06, "loss": 0.0035, "step": 10685 }, { "epoch": 0.7343320505772403, "grad_norm": 0.35020169615745544, "learning_rate": 5.218671589886656e-06, "loss": 0.0135, "step": 10686 }, { "epoch": 0.7344007696536559, "grad_norm": 0.5970134735107422, "learning_rate": 5.216140720747616e-06, "loss": 0.0084, "step": 10687 }, { "epoch": 0.7344694887300715, "grad_norm": 0.49287787079811096, "learning_rate": 5.213610336280845e-06, "loss": 0.013, "step": 10688 }, { "epoch": 0.734538207806487, "grad_norm": 0.13121791183948517, "learning_rate": 5.211080436611704e-06, "loss": 0.0057, "step": 10689 }, { "epoch": 0.7346069268829027, "grad_norm": 0.6357287168502808, "learning_rate": 5.208551021865507e-06, "loss": 0.0126, "step": 10690 }, { "epoch": 0.7346756459593183, "grad_norm": 0.16281478106975555, "learning_rate": 5.206022092167568e-06, "loss": 0.0117, "step": 10691 }, { "epoch": 0.7347443650357339, "grad_norm": 0.6371064186096191, "learning_rate": 5.203493647643148e-06, "loss": 0.0158, "step": 10692 }, { "epoch": 0.7348130841121495, "grad_norm": 0.4357433021068573, "learning_rate": 5.200965688417522e-06, "loss": 0.011, "step": 10693 }, { "epoch": 0.7348818031885651, "grad_norm": 0.31614482402801514, "learning_rate": 5.198438214615905e-06, "loss": 0.016, "step": 10694 }, { "epoch": 0.7349505222649808, "grad_norm": 0.37258002161979675, "learning_rate": 5.195911226363512e-06, "loss": 0.0119, "step": 10695 }, { "epoch": 0.7350192413413964, "grad_norm": 1.0289783477783203, "learning_rate": 5.193384723785517e-06, "loss": 0.009, "step": 10696 }, { "epoch": 0.735087960417812, "grad_norm": 0.5116332173347473, "learning_rate": 5.190858707007085e-06, "loss": 0.0117, "step": 10697 }, { "epoch": 0.7351566794942276, "grad_norm": 0.3215431869029999, "learning_rate": 5.188333176153337e-06, "loss": 0.0094, "step": 10698 }, { "epoch": 0.7352253985706432, "grad_norm": 0.8167309165000916, "learning_rate": 5.1858081313494014e-06, "loss": 0.015, "step": 10699 }, { "epoch": 0.7352941176470589, "grad_norm": 0.28997015953063965, "learning_rate": 5.1832835727203474e-06, "loss": 0.0054, "step": 10700 }, { "epoch": 0.7353628367234745, "grad_norm": 0.3100409209728241, "learning_rate": 5.180759500391249e-06, "loss": 0.0082, "step": 10701 }, { "epoch": 0.7354315557998901, "grad_norm": 0.4088754951953888, "learning_rate": 5.1782359144871315e-06, "loss": 0.0089, "step": 10702 }, { "epoch": 0.7355002748763056, "grad_norm": 0.30714574456214905, "learning_rate": 5.1757128151330204e-06, "loss": 0.0087, "step": 10703 }, { "epoch": 0.7355689939527212, "grad_norm": 0.5066428780555725, "learning_rate": 5.173190202453893e-06, "loss": 0.012, "step": 10704 }, { "epoch": 0.7356377130291369, "grad_norm": 0.14743836224079132, "learning_rate": 5.17066807657472e-06, "loss": 0.0091, "step": 10705 }, { "epoch": 0.7357064321055525, "grad_norm": 0.2770969867706299, "learning_rate": 5.168146437620447e-06, "loss": 0.0092, "step": 10706 }, { "epoch": 0.7357751511819681, "grad_norm": 0.18642884492874146, "learning_rate": 5.165625285715981e-06, "loss": 0.0107, "step": 10707 }, { "epoch": 0.7358438702583837, "grad_norm": 0.22161640226840973, "learning_rate": 5.163104620986225e-06, "loss": 0.0069, "step": 10708 }, { "epoch": 0.7359125893347993, "grad_norm": 0.34870558977127075, "learning_rate": 5.160584443556037e-06, "loss": 0.0071, "step": 10709 }, { "epoch": 0.735981308411215, "grad_norm": 0.6035820841789246, "learning_rate": 5.158064753550267e-06, "loss": 0.023, "step": 10710 }, { "epoch": 0.7360500274876306, "grad_norm": 0.3888990879058838, "learning_rate": 5.155545551093736e-06, "loss": 0.0075, "step": 10711 }, { "epoch": 0.7361187465640462, "grad_norm": 0.5674468278884888, "learning_rate": 5.153026836311242e-06, "loss": 0.0135, "step": 10712 }, { "epoch": 0.7361874656404618, "grad_norm": 0.37244129180908203, "learning_rate": 5.15050860932755e-06, "loss": 0.0115, "step": 10713 }, { "epoch": 0.7362561847168774, "grad_norm": 0.32729408144950867, "learning_rate": 5.147990870267416e-06, "loss": 0.0118, "step": 10714 }, { "epoch": 0.7363249037932931, "grad_norm": 0.5717990398406982, "learning_rate": 5.145473619255556e-06, "loss": 0.0131, "step": 10715 }, { "epoch": 0.7363936228697087, "grad_norm": 0.3723392188549042, "learning_rate": 5.142956856416671e-06, "loss": 0.0106, "step": 10716 }, { "epoch": 0.7364623419461243, "grad_norm": 0.4526835083961487, "learning_rate": 5.1404405818754396e-06, "loss": 0.011, "step": 10717 }, { "epoch": 0.7365310610225398, "grad_norm": 0.4974680542945862, "learning_rate": 5.137924795756516e-06, "loss": 0.0083, "step": 10718 }, { "epoch": 0.7365997800989554, "grad_norm": 0.274350643157959, "learning_rate": 5.135409498184517e-06, "loss": 0.0077, "step": 10719 }, { "epoch": 0.736668499175371, "grad_norm": 0.8660722970962524, "learning_rate": 5.132894689284055e-06, "loss": 0.0135, "step": 10720 }, { "epoch": 0.7367372182517867, "grad_norm": 0.5151031613349915, "learning_rate": 5.130380369179702e-06, "loss": 0.0153, "step": 10721 }, { "epoch": 0.7368059373282023, "grad_norm": 1.2813845872879028, "learning_rate": 5.127866537996011e-06, "loss": 0.015, "step": 10722 }, { "epoch": 0.7368746564046179, "grad_norm": 0.6957752704620361, "learning_rate": 5.125353195857522e-06, "loss": 0.0086, "step": 10723 }, { "epoch": 0.7369433754810335, "grad_norm": 0.4000546336174011, "learning_rate": 5.12284034288873e-06, "loss": 0.0081, "step": 10724 }, { "epoch": 0.7370120945574491, "grad_norm": 0.1279028207063675, "learning_rate": 5.120327979214125e-06, "loss": 0.0069, "step": 10725 }, { "epoch": 0.7370808136338648, "grad_norm": 0.5421532392501831, "learning_rate": 5.117816104958155e-06, "loss": 0.014, "step": 10726 }, { "epoch": 0.7371495327102804, "grad_norm": 0.4993618428707123, "learning_rate": 5.1153047202452645e-06, "loss": 0.0145, "step": 10727 }, { "epoch": 0.737218251786696, "grad_norm": 0.428122341632843, "learning_rate": 5.112793825199844e-06, "loss": 0.0183, "step": 10728 }, { "epoch": 0.7372869708631116, "grad_norm": 1.0191570520401, "learning_rate": 5.110283419946302e-06, "loss": 0.0163, "step": 10729 }, { "epoch": 0.7373556899395272, "grad_norm": 0.3154015839099884, "learning_rate": 5.107773504608981e-06, "loss": 0.0067, "step": 10730 }, { "epoch": 0.7374244090159429, "grad_norm": 0.33595481514930725, "learning_rate": 5.105264079312228e-06, "loss": 0.0127, "step": 10731 }, { "epoch": 0.7374931280923585, "grad_norm": 0.40232032537460327, "learning_rate": 5.1027551441803464e-06, "loss": 0.0048, "step": 10732 }, { "epoch": 0.737561847168774, "grad_norm": 0.31445491313934326, "learning_rate": 5.10024669933763e-06, "loss": 0.0087, "step": 10733 }, { "epoch": 0.7376305662451896, "grad_norm": 0.5517775416374207, "learning_rate": 5.09773874490833e-06, "loss": 0.0129, "step": 10734 }, { "epoch": 0.7376992853216052, "grad_norm": 0.5596057176589966, "learning_rate": 5.095231281016705e-06, "loss": 0.0112, "step": 10735 }, { "epoch": 0.7377680043980209, "grad_norm": 0.2152872085571289, "learning_rate": 5.0927243077869535e-06, "loss": 0.0133, "step": 10736 }, { "epoch": 0.7378367234744365, "grad_norm": 0.3289262652397156, "learning_rate": 5.090217825343275e-06, "loss": 0.009, "step": 10737 }, { "epoch": 0.7379054425508521, "grad_norm": 0.21268007159233093, "learning_rate": 5.0877118338098275e-06, "loss": 0.0082, "step": 10738 }, { "epoch": 0.7379741616272677, "grad_norm": 0.34858670830726624, "learning_rate": 5.085206333310764e-06, "loss": 0.0108, "step": 10739 }, { "epoch": 0.7380428807036833, "grad_norm": 0.573783278465271, "learning_rate": 5.0827013239701885e-06, "loss": 0.0101, "step": 10740 }, { "epoch": 0.738111599780099, "grad_norm": 0.4017810821533203, "learning_rate": 5.080196805912199e-06, "loss": 0.0133, "step": 10741 }, { "epoch": 0.7381803188565146, "grad_norm": 0.6892701387405396, "learning_rate": 5.077692779260872e-06, "loss": 0.0143, "step": 10742 }, { "epoch": 0.7382490379329302, "grad_norm": 0.37458908557891846, "learning_rate": 5.0751892441402395e-06, "loss": 0.0122, "step": 10743 }, { "epoch": 0.7383177570093458, "grad_norm": 1.6971641778945923, "learning_rate": 5.072686200674332e-06, "loss": 0.0172, "step": 10744 }, { "epoch": 0.7383864760857614, "grad_norm": 0.4091992676258087, "learning_rate": 5.070183648987136e-06, "loss": 0.0091, "step": 10745 }, { "epoch": 0.7384551951621771, "grad_norm": 0.12045387923717499, "learning_rate": 5.067681589202631e-06, "loss": 0.0118, "step": 10746 }, { "epoch": 0.7385239142385926, "grad_norm": 0.2877022325992584, "learning_rate": 5.065180021444753e-06, "loss": 0.0155, "step": 10747 }, { "epoch": 0.7385926333150082, "grad_norm": 0.3127121925354004, "learning_rate": 5.062678945837439e-06, "loss": 0.0193, "step": 10748 }, { "epoch": 0.7386613523914238, "grad_norm": 0.33980825543403625, "learning_rate": 5.060178362504575e-06, "loss": 0.0098, "step": 10749 }, { "epoch": 0.7387300714678394, "grad_norm": 0.5534871220588684, "learning_rate": 5.057678271570045e-06, "loss": 0.0177, "step": 10750 }, { "epoch": 0.7387987905442551, "grad_norm": 0.44336098432540894, "learning_rate": 5.055178673157689e-06, "loss": 0.0091, "step": 10751 }, { "epoch": 0.7388675096206707, "grad_norm": 1.0120195150375366, "learning_rate": 5.052679567391339e-06, "loss": 0.0167, "step": 10752 }, { "epoch": 0.7389362286970863, "grad_norm": 0.45107734203338623, "learning_rate": 5.050180954394784e-06, "loss": 0.0147, "step": 10753 }, { "epoch": 0.7390049477735019, "grad_norm": 0.443053662776947, "learning_rate": 5.047682834291818e-06, "loss": 0.0076, "step": 10754 }, { "epoch": 0.7390736668499175, "grad_norm": 0.31838148832321167, "learning_rate": 5.045185207206177e-06, "loss": 0.0095, "step": 10755 }, { "epoch": 0.7391423859263332, "grad_norm": 0.2953384816646576, "learning_rate": 5.042688073261602e-06, "loss": 0.008, "step": 10756 }, { "epoch": 0.7392111050027488, "grad_norm": 0.4071003496646881, "learning_rate": 5.040191432581781e-06, "loss": 0.0108, "step": 10757 }, { "epoch": 0.7392798240791644, "grad_norm": 0.5846370458602905, "learning_rate": 5.037695285290407e-06, "loss": 0.0141, "step": 10758 }, { "epoch": 0.73934854315558, "grad_norm": 0.5206644535064697, "learning_rate": 5.0351996315111206e-06, "loss": 0.0121, "step": 10759 }, { "epoch": 0.7394172622319956, "grad_norm": 0.46113818883895874, "learning_rate": 5.032704471367558e-06, "loss": 0.0078, "step": 10760 }, { "epoch": 0.7394859813084113, "grad_norm": 0.49926894903182983, "learning_rate": 5.030209804983329e-06, "loss": 0.011, "step": 10761 }, { "epoch": 0.7395547003848268, "grad_norm": 0.2560030519962311, "learning_rate": 5.027715632482004e-06, "loss": 0.0096, "step": 10762 }, { "epoch": 0.7396234194612424, "grad_norm": 0.6038016676902771, "learning_rate": 5.025221953987151e-06, "loss": 0.0061, "step": 10763 }, { "epoch": 0.739692138537658, "grad_norm": 0.41781702637672424, "learning_rate": 5.022728769622289e-06, "loss": 0.0146, "step": 10764 }, { "epoch": 0.7397608576140736, "grad_norm": 0.6266093850135803, "learning_rate": 5.02023607951093e-06, "loss": 0.014, "step": 10765 }, { "epoch": 0.7398295766904893, "grad_norm": 0.7527603507041931, "learning_rate": 5.017743883776561e-06, "loss": 0.017, "step": 10766 }, { "epoch": 0.7398982957669049, "grad_norm": 0.5813724994659424, "learning_rate": 5.01525218254264e-06, "loss": 0.0068, "step": 10767 }, { "epoch": 0.7399670148433205, "grad_norm": 0.4349771738052368, "learning_rate": 5.0127609759325935e-06, "loss": 0.0177, "step": 10768 }, { "epoch": 0.7400357339197361, "grad_norm": 0.3868774175643921, "learning_rate": 5.010270264069842e-06, "loss": 0.0174, "step": 10769 }, { "epoch": 0.7401044529961517, "grad_norm": 0.39606723189353943, "learning_rate": 5.007780047077756e-06, "loss": 0.0087, "step": 10770 }, { "epoch": 0.7401731720725674, "grad_norm": 0.19036883115768433, "learning_rate": 5.0052903250797055e-06, "loss": 0.0089, "step": 10771 }, { "epoch": 0.740241891148983, "grad_norm": 0.34567025303840637, "learning_rate": 5.002801098199022e-06, "loss": 0.0208, "step": 10772 }, { "epoch": 0.7403106102253986, "grad_norm": 0.9124990105628967, "learning_rate": 5.000312366559026e-06, "loss": 0.0143, "step": 10773 }, { "epoch": 0.7403793293018142, "grad_norm": 0.3898269236087799, "learning_rate": 4.997824130282991e-06, "loss": 0.0125, "step": 10774 }, { "epoch": 0.7404480483782298, "grad_norm": 0.18859592080116272, "learning_rate": 4.9953363894941895e-06, "loss": 0.0083, "step": 10775 }, { "epoch": 0.7405167674546455, "grad_norm": 0.3345506191253662, "learning_rate": 4.99284914431585e-06, "loss": 0.0088, "step": 10776 }, { "epoch": 0.740585486531061, "grad_norm": 0.3528837263584137, "learning_rate": 4.99036239487119e-06, "loss": 0.0116, "step": 10777 }, { "epoch": 0.7406542056074766, "grad_norm": 0.22873125970363617, "learning_rate": 4.987876141283403e-06, "loss": 0.0113, "step": 10778 }, { "epoch": 0.7407229246838922, "grad_norm": 0.2684648036956787, "learning_rate": 4.985390383675642e-06, "loss": 0.0143, "step": 10779 }, { "epoch": 0.7407916437603078, "grad_norm": 0.33908095955848694, "learning_rate": 4.982905122171057e-06, "loss": 0.0094, "step": 10780 }, { "epoch": 0.7408603628367235, "grad_norm": 0.4107434153556824, "learning_rate": 4.980420356892755e-06, "loss": 0.0138, "step": 10781 }, { "epoch": 0.7409290819131391, "grad_norm": 0.29648783802986145, "learning_rate": 4.977936087963833e-06, "loss": 0.0087, "step": 10782 }, { "epoch": 0.7409978009895547, "grad_norm": 0.5765300393104553, "learning_rate": 4.975452315507344e-06, "loss": 0.0117, "step": 10783 }, { "epoch": 0.7410665200659703, "grad_norm": 0.20950359106063843, "learning_rate": 4.972969039646346e-06, "loss": 0.0077, "step": 10784 }, { "epoch": 0.7411352391423859, "grad_norm": 0.3272867202758789, "learning_rate": 4.970486260503843e-06, "loss": 0.0073, "step": 10785 }, { "epoch": 0.7412039582188016, "grad_norm": 0.49771934747695923, "learning_rate": 4.968003978202835e-06, "loss": 0.0103, "step": 10786 }, { "epoch": 0.7412726772952172, "grad_norm": 0.23230646550655365, "learning_rate": 4.965522192866281e-06, "loss": 0.0099, "step": 10787 }, { "epoch": 0.7413413963716328, "grad_norm": 0.2907831072807312, "learning_rate": 4.963040904617131e-06, "loss": 0.0098, "step": 10788 }, { "epoch": 0.7414101154480484, "grad_norm": 0.8818403482437134, "learning_rate": 4.960560113578291e-06, "loss": 0.0099, "step": 10789 }, { "epoch": 0.741478834524464, "grad_norm": 0.25782540440559387, "learning_rate": 4.9580798198726715e-06, "loss": 0.0099, "step": 10790 }, { "epoch": 0.7415475536008796, "grad_norm": 0.3594714403152466, "learning_rate": 4.955600023623128e-06, "loss": 0.0075, "step": 10791 }, { "epoch": 0.7416162726772952, "grad_norm": 0.15450885891914368, "learning_rate": 4.953120724952514e-06, "loss": 0.0101, "step": 10792 }, { "epoch": 0.7416849917537108, "grad_norm": 0.6469739675521851, "learning_rate": 4.95064192398364e-06, "loss": 0.0144, "step": 10793 }, { "epoch": 0.7417537108301264, "grad_norm": 0.5639315247535706, "learning_rate": 4.948163620839307e-06, "loss": 0.0086, "step": 10794 }, { "epoch": 0.741822429906542, "grad_norm": 0.39293184876441956, "learning_rate": 4.9456858156422795e-06, "loss": 0.0157, "step": 10795 }, { "epoch": 0.7418911489829577, "grad_norm": 0.5143484473228455, "learning_rate": 4.9432085085153055e-06, "loss": 0.0116, "step": 10796 }, { "epoch": 0.7419598680593733, "grad_norm": 0.657874345779419, "learning_rate": 4.94073169958111e-06, "loss": 0.0166, "step": 10797 }, { "epoch": 0.7420285871357889, "grad_norm": 0.38429349660873413, "learning_rate": 4.938255388962382e-06, "loss": 0.0091, "step": 10798 }, { "epoch": 0.7420973062122045, "grad_norm": 0.7313346266746521, "learning_rate": 4.935779576781799e-06, "loss": 0.0112, "step": 10799 }, { "epoch": 0.7421660252886201, "grad_norm": 0.4519657790660858, "learning_rate": 4.9333042631620016e-06, "loss": 0.0128, "step": 10800 }, { "epoch": 0.7422347443650358, "grad_norm": 0.5109586715698242, "learning_rate": 4.9308294482256135e-06, "loss": 0.0117, "step": 10801 }, { "epoch": 0.7423034634414514, "grad_norm": 0.27172040939331055, "learning_rate": 4.928355132095233e-06, "loss": 0.0059, "step": 10802 }, { "epoch": 0.742372182517867, "grad_norm": 0.5927114486694336, "learning_rate": 4.925881314893438e-06, "loss": 0.0101, "step": 10803 }, { "epoch": 0.7424409015942826, "grad_norm": 0.4356885552406311, "learning_rate": 4.9234079967427665e-06, "loss": 0.0069, "step": 10804 }, { "epoch": 0.7425096206706981, "grad_norm": 0.43069401383399963, "learning_rate": 4.920935177765752e-06, "loss": 0.0155, "step": 10805 }, { "epoch": 0.7425783397471138, "grad_norm": 0.5754079818725586, "learning_rate": 4.918462858084881e-06, "loss": 0.0103, "step": 10806 }, { "epoch": 0.7426470588235294, "grad_norm": 0.343708336353302, "learning_rate": 4.915991037822633e-06, "loss": 0.0098, "step": 10807 }, { "epoch": 0.742715777899945, "grad_norm": 0.29541242122650146, "learning_rate": 4.913519717101458e-06, "loss": 0.0085, "step": 10808 }, { "epoch": 0.7427844969763606, "grad_norm": 0.9322421550750732, "learning_rate": 4.911048896043784e-06, "loss": 0.0206, "step": 10809 }, { "epoch": 0.7428532160527762, "grad_norm": 0.18769405782222748, "learning_rate": 4.908578574772001e-06, "loss": 0.0078, "step": 10810 }, { "epoch": 0.7429219351291919, "grad_norm": 0.3517433702945709, "learning_rate": 4.906108753408493e-06, "loss": 0.0058, "step": 10811 }, { "epoch": 0.7429906542056075, "grad_norm": 0.16710318624973297, "learning_rate": 4.9036394320756e-06, "loss": 0.0112, "step": 10812 }, { "epoch": 0.7430593732820231, "grad_norm": 0.48702508211135864, "learning_rate": 4.901170610895653e-06, "loss": 0.0078, "step": 10813 }, { "epoch": 0.7431280923584387, "grad_norm": 1.1155050992965698, "learning_rate": 4.898702289990955e-06, "loss": 0.0128, "step": 10814 }, { "epoch": 0.7431968114348543, "grad_norm": 0.3356733024120331, "learning_rate": 4.896234469483774e-06, "loss": 0.0111, "step": 10815 }, { "epoch": 0.74326553051127, "grad_norm": 0.5120883584022522, "learning_rate": 4.89376714949637e-06, "loss": 0.0102, "step": 10816 }, { "epoch": 0.7433342495876856, "grad_norm": 0.5925986170768738, "learning_rate": 4.891300330150959e-06, "loss": 0.0126, "step": 10817 }, { "epoch": 0.7434029686641012, "grad_norm": 0.292210191488266, "learning_rate": 4.888834011569752e-06, "loss": 0.0121, "step": 10818 }, { "epoch": 0.7434716877405168, "grad_norm": 0.7979742884635925, "learning_rate": 4.88636819387491e-06, "loss": 0.0085, "step": 10819 }, { "epoch": 0.7435404068169323, "grad_norm": 0.21161511540412903, "learning_rate": 4.883902877188607e-06, "loss": 0.0047, "step": 10820 }, { "epoch": 0.743609125893348, "grad_norm": 0.9382020235061646, "learning_rate": 4.88143806163295e-06, "loss": 0.0103, "step": 10821 }, { "epoch": 0.7436778449697636, "grad_norm": 0.39330098032951355, "learning_rate": 4.878973747330055e-06, "loss": 0.0067, "step": 10822 }, { "epoch": 0.7437465640461792, "grad_norm": 0.06689146906137466, "learning_rate": 4.876509934401987e-06, "loss": 0.005, "step": 10823 }, { "epoch": 0.7438152831225948, "grad_norm": 0.43597570061683655, "learning_rate": 4.874046622970808e-06, "loss": 0.015, "step": 10824 }, { "epoch": 0.7438840021990104, "grad_norm": 0.21393464505672455, "learning_rate": 4.871583813158533e-06, "loss": 0.0065, "step": 10825 }, { "epoch": 0.7439527212754261, "grad_norm": 0.7633403539657593, "learning_rate": 4.869121505087182e-06, "loss": 0.0103, "step": 10826 }, { "epoch": 0.7440214403518417, "grad_norm": 0.421207070350647, "learning_rate": 4.866659698878718e-06, "loss": 0.0145, "step": 10827 }, { "epoch": 0.7440901594282573, "grad_norm": 0.4556259214878082, "learning_rate": 4.864198394655105e-06, "loss": 0.0205, "step": 10828 }, { "epoch": 0.7441588785046729, "grad_norm": 0.5008546113967896, "learning_rate": 4.861737592538259e-06, "loss": 0.0085, "step": 10829 }, { "epoch": 0.7442275975810885, "grad_norm": 0.6662501692771912, "learning_rate": 4.8592772926500965e-06, "loss": 0.0096, "step": 10830 }, { "epoch": 0.7442963166575042, "grad_norm": 0.38220617175102234, "learning_rate": 4.856817495112482e-06, "loss": 0.0121, "step": 10831 }, { "epoch": 0.7443650357339198, "grad_norm": 0.2758048176765442, "learning_rate": 4.854358200047275e-06, "loss": 0.0139, "step": 10832 }, { "epoch": 0.7444337548103354, "grad_norm": 1.1009886264801025, "learning_rate": 4.8518994075763095e-06, "loss": 0.0093, "step": 10833 }, { "epoch": 0.744502473886751, "grad_norm": 0.6410748362541199, "learning_rate": 4.849441117821379e-06, "loss": 0.0101, "step": 10834 }, { "epoch": 0.7445711929631665, "grad_norm": 0.14266645908355713, "learning_rate": 4.846983330904272e-06, "loss": 0.009, "step": 10835 }, { "epoch": 0.7446399120395822, "grad_norm": 0.10635591298341751, "learning_rate": 4.844526046946732e-06, "loss": 0.0109, "step": 10836 }, { "epoch": 0.7447086311159978, "grad_norm": 0.3989696204662323, "learning_rate": 4.8420692660704925e-06, "loss": 0.0088, "step": 10837 }, { "epoch": 0.7447773501924134, "grad_norm": 1.1090404987335205, "learning_rate": 4.839612988397259e-06, "loss": 0.01, "step": 10838 }, { "epoch": 0.744846069268829, "grad_norm": 0.6096281409263611, "learning_rate": 4.837157214048713e-06, "loss": 0.0182, "step": 10839 }, { "epoch": 0.7449147883452446, "grad_norm": 0.19160674512386322, "learning_rate": 4.834701943146499e-06, "loss": 0.0056, "step": 10840 }, { "epoch": 0.7449835074216603, "grad_norm": 0.4038386940956116, "learning_rate": 4.832247175812257e-06, "loss": 0.0197, "step": 10841 }, { "epoch": 0.7450522264980759, "grad_norm": 0.11714262515306473, "learning_rate": 4.8297929121675825e-06, "loss": 0.0056, "step": 10842 }, { "epoch": 0.7451209455744915, "grad_norm": 0.7918432950973511, "learning_rate": 4.827339152334056e-06, "loss": 0.0107, "step": 10843 }, { "epoch": 0.7451896646509071, "grad_norm": 0.8673344850540161, "learning_rate": 4.824885896433235e-06, "loss": 0.0128, "step": 10844 }, { "epoch": 0.7452583837273227, "grad_norm": 0.6159682273864746, "learning_rate": 4.822433144586653e-06, "loss": 0.0209, "step": 10845 }, { "epoch": 0.7453271028037384, "grad_norm": 0.8211299777030945, "learning_rate": 4.8199808969158025e-06, "loss": 0.0133, "step": 10846 }, { "epoch": 0.745395821880154, "grad_norm": 0.14784307777881622, "learning_rate": 4.817529153542176e-06, "loss": 0.0215, "step": 10847 }, { "epoch": 0.7454645409565696, "grad_norm": 0.21699005365371704, "learning_rate": 4.815077914587215e-06, "loss": 0.0054, "step": 10848 }, { "epoch": 0.7455332600329851, "grad_norm": 0.6479623913764954, "learning_rate": 4.812627180172353e-06, "loss": 0.0097, "step": 10849 }, { "epoch": 0.7456019791094007, "grad_norm": 0.40569278597831726, "learning_rate": 4.810176950419004e-06, "loss": 0.0098, "step": 10850 }, { "epoch": 0.7456706981858163, "grad_norm": 0.5684661269187927, "learning_rate": 4.8077272254485324e-06, "loss": 0.0113, "step": 10851 }, { "epoch": 0.745739417262232, "grad_norm": 0.4886167049407959, "learning_rate": 4.805278005382305e-06, "loss": 0.0083, "step": 10852 }, { "epoch": 0.7458081363386476, "grad_norm": 0.44127529859542847, "learning_rate": 4.8028292903416416e-06, "loss": 0.0172, "step": 10853 }, { "epoch": 0.7458768554150632, "grad_norm": 0.33180901408195496, "learning_rate": 4.800381080447851e-06, "loss": 0.0068, "step": 10854 }, { "epoch": 0.7459455744914788, "grad_norm": 0.42105960845947266, "learning_rate": 4.797933375822211e-06, "loss": 0.0136, "step": 10855 }, { "epoch": 0.7460142935678944, "grad_norm": 0.5239546298980713, "learning_rate": 4.795486176585983e-06, "loss": 0.0121, "step": 10856 }, { "epoch": 0.7460830126443101, "grad_norm": 0.43201059103012085, "learning_rate": 4.793039482860384e-06, "loss": 0.0109, "step": 10857 }, { "epoch": 0.7461517317207257, "grad_norm": 0.21200144290924072, "learning_rate": 4.790593294766629e-06, "loss": 0.0097, "step": 10858 }, { "epoch": 0.7462204507971413, "grad_norm": 0.20281420648097992, "learning_rate": 4.78814761242589e-06, "loss": 0.0082, "step": 10859 }, { "epoch": 0.7462891698735569, "grad_norm": 0.36064019799232483, "learning_rate": 4.785702435959327e-06, "loss": 0.0142, "step": 10860 }, { "epoch": 0.7463578889499725, "grad_norm": 0.32710957527160645, "learning_rate": 4.783257765488056e-06, "loss": 0.0106, "step": 10861 }, { "epoch": 0.7464266080263882, "grad_norm": 0.7749799489974976, "learning_rate": 4.7808136011332e-06, "loss": 0.0132, "step": 10862 }, { "epoch": 0.7464953271028038, "grad_norm": 0.3329983353614807, "learning_rate": 4.778369943015823e-06, "loss": 0.0173, "step": 10863 }, { "epoch": 0.7465640461792193, "grad_norm": 0.46212947368621826, "learning_rate": 4.775926791256991e-06, "loss": 0.0073, "step": 10864 }, { "epoch": 0.7466327652556349, "grad_norm": 0.942200243473053, "learning_rate": 4.773484145977719e-06, "loss": 0.0143, "step": 10865 }, { "epoch": 0.7467014843320505, "grad_norm": 0.32153695821762085, "learning_rate": 4.771042007299023e-06, "loss": 0.0147, "step": 10866 }, { "epoch": 0.7467702034084662, "grad_norm": 0.33768317103385925, "learning_rate": 4.768600375341872e-06, "loss": 0.0094, "step": 10867 }, { "epoch": 0.7468389224848818, "grad_norm": 0.7449252605438232, "learning_rate": 4.766159250227223e-06, "loss": 0.0085, "step": 10868 }, { "epoch": 0.7469076415612974, "grad_norm": 0.6226500272750854, "learning_rate": 4.7637186320760064e-06, "loss": 0.0149, "step": 10869 }, { "epoch": 0.746976360637713, "grad_norm": 0.4394127130508423, "learning_rate": 4.761278521009127e-06, "loss": 0.0096, "step": 10870 }, { "epoch": 0.7470450797141286, "grad_norm": 0.2933552861213684, "learning_rate": 4.758838917147455e-06, "loss": 0.0059, "step": 10871 }, { "epoch": 0.7471137987905443, "grad_norm": 0.32215696573257446, "learning_rate": 4.756399820611854e-06, "loss": 0.0127, "step": 10872 }, { "epoch": 0.7471825178669599, "grad_norm": 0.31646373867988586, "learning_rate": 4.7539612315231405e-06, "loss": 0.0068, "step": 10873 }, { "epoch": 0.7472512369433755, "grad_norm": 0.2889191210269928, "learning_rate": 4.7515231500021215e-06, "loss": 0.0155, "step": 10874 }, { "epoch": 0.7473199560197911, "grad_norm": 0.706778883934021, "learning_rate": 4.749085576169582e-06, "loss": 0.0096, "step": 10875 }, { "epoch": 0.7473886750962067, "grad_norm": 0.6600204110145569, "learning_rate": 4.746648510146263e-06, "loss": 0.0194, "step": 10876 }, { "epoch": 0.7474573941726224, "grad_norm": 0.33523687720298767, "learning_rate": 4.744211952052901e-06, "loss": 0.013, "step": 10877 }, { "epoch": 0.747526113249038, "grad_norm": 0.5340520739555359, "learning_rate": 4.741775902010189e-06, "loss": 0.0102, "step": 10878 }, { "epoch": 0.7475948323254535, "grad_norm": 0.2628023326396942, "learning_rate": 4.7393403601388105e-06, "loss": 0.0095, "step": 10879 }, { "epoch": 0.7476635514018691, "grad_norm": 0.3497781455516815, "learning_rate": 4.736905326559414e-06, "loss": 0.0096, "step": 10880 }, { "epoch": 0.7477322704782847, "grad_norm": 0.8372683525085449, "learning_rate": 4.734470801392633e-06, "loss": 0.0114, "step": 10881 }, { "epoch": 0.7478009895547004, "grad_norm": 0.46230077743530273, "learning_rate": 4.732036784759057e-06, "loss": 0.0097, "step": 10882 }, { "epoch": 0.747869708631116, "grad_norm": 0.21466664969921112, "learning_rate": 4.729603276779275e-06, "loss": 0.0137, "step": 10883 }, { "epoch": 0.7479384277075316, "grad_norm": 0.1293552964925766, "learning_rate": 4.727170277573827e-06, "loss": 0.0093, "step": 10884 }, { "epoch": 0.7480071467839472, "grad_norm": 0.46879804134368896, "learning_rate": 4.724737787263244e-06, "loss": 0.0186, "step": 10885 }, { "epoch": 0.7480758658603628, "grad_norm": 0.34255197644233704, "learning_rate": 4.722305805968024e-06, "loss": 0.0116, "step": 10886 }, { "epoch": 0.7481445849367785, "grad_norm": 0.4733238220214844, "learning_rate": 4.71987433380865e-06, "loss": 0.0085, "step": 10887 }, { "epoch": 0.7482133040131941, "grad_norm": 0.22204625606536865, "learning_rate": 4.717443370905563e-06, "loss": 0.0087, "step": 10888 }, { "epoch": 0.7482820230896097, "grad_norm": 0.4662605822086334, "learning_rate": 4.715012917379195e-06, "loss": 0.0162, "step": 10889 }, { "epoch": 0.7483507421660253, "grad_norm": 0.5062786936759949, "learning_rate": 4.712582973349937e-06, "loss": 0.0078, "step": 10890 }, { "epoch": 0.7484194612424409, "grad_norm": 0.6049954295158386, "learning_rate": 4.710153538938168e-06, "loss": 0.0181, "step": 10891 }, { "epoch": 0.7484881803188566, "grad_norm": 0.2772281765937805, "learning_rate": 4.7077246142642425e-06, "loss": 0.0072, "step": 10892 }, { "epoch": 0.7485568993952721, "grad_norm": 0.6565437316894531, "learning_rate": 4.705296199448477e-06, "loss": 0.0108, "step": 10893 }, { "epoch": 0.7486256184716877, "grad_norm": 0.40923577547073364, "learning_rate": 4.702868294611175e-06, "loss": 0.0113, "step": 10894 }, { "epoch": 0.7486943375481033, "grad_norm": 0.6054365634918213, "learning_rate": 4.700440899872603e-06, "loss": 0.0112, "step": 10895 }, { "epoch": 0.7487630566245189, "grad_norm": 0.3480254113674164, "learning_rate": 4.698014015353021e-06, "loss": 0.008, "step": 10896 }, { "epoch": 0.7488317757009346, "grad_norm": 0.4860157072544098, "learning_rate": 4.695587641172634e-06, "loss": 0.0118, "step": 10897 }, { "epoch": 0.7489004947773502, "grad_norm": 0.33570367097854614, "learning_rate": 4.693161777451659e-06, "loss": 0.01, "step": 10898 }, { "epoch": 0.7489692138537658, "grad_norm": 0.6430409550666809, "learning_rate": 4.690736424310256e-06, "loss": 0.0186, "step": 10899 }, { "epoch": 0.7490379329301814, "grad_norm": 0.5059111714363098, "learning_rate": 4.688311581868578e-06, "loss": 0.0108, "step": 10900 }, { "epoch": 0.749106652006597, "grad_norm": 0.36655130982398987, "learning_rate": 4.685887250246741e-06, "loss": 0.0102, "step": 10901 }, { "epoch": 0.7491753710830127, "grad_norm": 0.8081387877464294, "learning_rate": 4.68346342956485e-06, "loss": 0.0186, "step": 10902 }, { "epoch": 0.7492440901594283, "grad_norm": 0.30146560072898865, "learning_rate": 4.68104011994296e-06, "loss": 0.0089, "step": 10903 }, { "epoch": 0.7493128092358439, "grad_norm": 0.4275646209716797, "learning_rate": 4.678617321501139e-06, "loss": 0.0127, "step": 10904 }, { "epoch": 0.7493815283122595, "grad_norm": 0.5203796625137329, "learning_rate": 4.676195034359391e-06, "loss": 0.014, "step": 10905 }, { "epoch": 0.749450247388675, "grad_norm": 0.26557657122612, "learning_rate": 4.6737732586377195e-06, "loss": 0.0144, "step": 10906 }, { "epoch": 0.7495189664650908, "grad_norm": 0.34797629714012146, "learning_rate": 4.671351994456087e-06, "loss": 0.0084, "step": 10907 }, { "epoch": 0.7495876855415063, "grad_norm": 0.3832136392593384, "learning_rate": 4.668931241934446e-06, "loss": 0.0074, "step": 10908 }, { "epoch": 0.7496564046179219, "grad_norm": 0.34502604603767395, "learning_rate": 4.666511001192707e-06, "loss": 0.0069, "step": 10909 }, { "epoch": 0.7497251236943375, "grad_norm": 0.48742708563804626, "learning_rate": 4.66409127235077e-06, "loss": 0.0112, "step": 10910 }, { "epoch": 0.7497938427707531, "grad_norm": 0.4201144874095917, "learning_rate": 4.661672055528504e-06, "loss": 0.0071, "step": 10911 }, { "epoch": 0.7498625618471688, "grad_norm": 0.6498249769210815, "learning_rate": 4.6592533508457454e-06, "loss": 0.0121, "step": 10912 }, { "epoch": 0.7499312809235844, "grad_norm": 0.27863284945487976, "learning_rate": 4.656835158422322e-06, "loss": 0.0106, "step": 10913 }, { "epoch": 0.75, "grad_norm": 2.132357597351074, "learning_rate": 4.654417478378015e-06, "loss": 0.0106, "step": 10914 }, { "epoch": 0.7500687190764156, "grad_norm": 0.36157476902008057, "learning_rate": 4.6520003108325954e-06, "loss": 0.0129, "step": 10915 }, { "epoch": 0.7501374381528312, "grad_norm": 0.38108283281326294, "learning_rate": 4.649583655905807e-06, "loss": 0.0126, "step": 10916 }, { "epoch": 0.7502061572292469, "grad_norm": 0.45021554827690125, "learning_rate": 4.647167513717369e-06, "loss": 0.0128, "step": 10917 }, { "epoch": 0.7502748763056625, "grad_norm": 0.6605406999588013, "learning_rate": 4.644751884386963e-06, "loss": 0.0194, "step": 10918 }, { "epoch": 0.7503435953820781, "grad_norm": 0.22481581568717957, "learning_rate": 4.642336768034264e-06, "loss": 0.0112, "step": 10919 }, { "epoch": 0.7504123144584937, "grad_norm": 0.6196793913841248, "learning_rate": 4.639922164778903e-06, "loss": 0.0159, "step": 10920 }, { "epoch": 0.7504810335349092, "grad_norm": 0.7485296726226807, "learning_rate": 4.6375080747404986e-06, "loss": 0.0122, "step": 10921 }, { "epoch": 0.750549752611325, "grad_norm": 0.3626459240913391, "learning_rate": 4.635094498038638e-06, "loss": 0.012, "step": 10922 }, { "epoch": 0.7506184716877405, "grad_norm": 0.0758429691195488, "learning_rate": 4.632681434792892e-06, "loss": 0.008, "step": 10923 }, { "epoch": 0.7506871907641561, "grad_norm": 0.6595293283462524, "learning_rate": 4.63026888512279e-06, "loss": 0.0133, "step": 10924 }, { "epoch": 0.7507559098405717, "grad_norm": 1.345484972000122, "learning_rate": 4.627856849147853e-06, "loss": 0.0128, "step": 10925 }, { "epoch": 0.7508246289169873, "grad_norm": 0.2222047597169876, "learning_rate": 4.625445326987558e-06, "loss": 0.0106, "step": 10926 }, { "epoch": 0.750893347993403, "grad_norm": 0.46253180503845215, "learning_rate": 4.6230343187613715e-06, "loss": 0.0089, "step": 10927 }, { "epoch": 0.7509620670698186, "grad_norm": 0.37414535880088806, "learning_rate": 4.6206238245887365e-06, "loss": 0.0199, "step": 10928 }, { "epoch": 0.7510307861462342, "grad_norm": 0.3362387716770172, "learning_rate": 4.618213844589053e-06, "loss": 0.0071, "step": 10929 }, { "epoch": 0.7510995052226498, "grad_norm": 0.1421588957309723, "learning_rate": 4.615804378881716e-06, "loss": 0.0077, "step": 10930 }, { "epoch": 0.7511682242990654, "grad_norm": 0.2524327337741852, "learning_rate": 4.613395427586077e-06, "loss": 0.0148, "step": 10931 }, { "epoch": 0.7512369433754811, "grad_norm": 0.1812783032655716, "learning_rate": 4.610986990821479e-06, "loss": 0.0096, "step": 10932 }, { "epoch": 0.7513056624518967, "grad_norm": 0.3079816699028015, "learning_rate": 4.608579068707217e-06, "loss": 0.0074, "step": 10933 }, { "epoch": 0.7513743815283123, "grad_norm": 0.7923927307128906, "learning_rate": 4.606171661362594e-06, "loss": 0.0148, "step": 10934 }, { "epoch": 0.7514431006047279, "grad_norm": 0.41790255904197693, "learning_rate": 4.603764768906854e-06, "loss": 0.0095, "step": 10935 }, { "epoch": 0.7515118196811434, "grad_norm": 0.19795779883861542, "learning_rate": 4.601358391459236e-06, "loss": 0.0077, "step": 10936 }, { "epoch": 0.7515805387575591, "grad_norm": 0.5249770879745483, "learning_rate": 4.598952529138941e-06, "loss": 0.0105, "step": 10937 }, { "epoch": 0.7516492578339747, "grad_norm": 0.21603842079639435, "learning_rate": 4.59654718206516e-06, "loss": 0.0062, "step": 10938 }, { "epoch": 0.7517179769103903, "grad_norm": 0.44871124625205994, "learning_rate": 4.594142350357032e-06, "loss": 0.0109, "step": 10939 }, { "epoch": 0.7517866959868059, "grad_norm": 0.4124520719051361, "learning_rate": 4.591738034133708e-06, "loss": 0.0111, "step": 10940 }, { "epoch": 0.7518554150632215, "grad_norm": 0.375907838344574, "learning_rate": 4.58933423351428e-06, "loss": 0.0153, "step": 10941 }, { "epoch": 0.7519241341396372, "grad_norm": 0.45224279165267944, "learning_rate": 4.586930948617834e-06, "loss": 0.0105, "step": 10942 }, { "epoch": 0.7519928532160528, "grad_norm": 0.5251612663269043, "learning_rate": 4.5845281795634156e-06, "loss": 0.0104, "step": 10943 }, { "epoch": 0.7520615722924684, "grad_norm": 0.3900006115436554, "learning_rate": 4.5821259264700634e-06, "loss": 0.0122, "step": 10944 }, { "epoch": 0.752130291368884, "grad_norm": 0.6406199336051941, "learning_rate": 4.579724189456769e-06, "loss": 0.0144, "step": 10945 }, { "epoch": 0.7521990104452996, "grad_norm": 0.3273100256919861, "learning_rate": 4.577322968642516e-06, "loss": 0.0102, "step": 10946 }, { "epoch": 0.7522677295217153, "grad_norm": 0.2505418062210083, "learning_rate": 4.574922264146258e-06, "loss": 0.0104, "step": 10947 }, { "epoch": 0.7523364485981309, "grad_norm": 0.2733020484447479, "learning_rate": 4.5725220760869165e-06, "loss": 0.0113, "step": 10948 }, { "epoch": 0.7524051676745465, "grad_norm": 0.26302850246429443, "learning_rate": 4.570122404583396e-06, "loss": 0.0073, "step": 10949 }, { "epoch": 0.752473886750962, "grad_norm": 0.5175628662109375, "learning_rate": 4.567723249754563e-06, "loss": 0.0066, "step": 10950 }, { "epoch": 0.7525426058273776, "grad_norm": 1.016950011253357, "learning_rate": 4.565324611719275e-06, "loss": 0.0162, "step": 10951 }, { "epoch": 0.7526113249037933, "grad_norm": 0.3117407262325287, "learning_rate": 4.5629264905963494e-06, "loss": 0.0146, "step": 10952 }, { "epoch": 0.7526800439802089, "grad_norm": 0.4505428969860077, "learning_rate": 4.560528886504596e-06, "loss": 0.0093, "step": 10953 }, { "epoch": 0.7527487630566245, "grad_norm": 0.8002416491508484, "learning_rate": 4.558131799562773e-06, "loss": 0.0141, "step": 10954 }, { "epoch": 0.7528174821330401, "grad_norm": 0.5247658491134644, "learning_rate": 4.555735229889636e-06, "loss": 0.0169, "step": 10955 }, { "epoch": 0.7528862012094557, "grad_norm": 0.10495718568563461, "learning_rate": 4.553339177603902e-06, "loss": 0.0059, "step": 10956 }, { "epoch": 0.7529549202858714, "grad_norm": 0.3896172344684601, "learning_rate": 4.550943642824264e-06, "loss": 0.0137, "step": 10957 }, { "epoch": 0.753023639362287, "grad_norm": 0.2930671274662018, "learning_rate": 4.548548625669398e-06, "loss": 0.0062, "step": 10958 }, { "epoch": 0.7530923584387026, "grad_norm": 0.1615600734949112, "learning_rate": 4.54615412625795e-06, "loss": 0.0051, "step": 10959 }, { "epoch": 0.7531610775151182, "grad_norm": 0.42514538764953613, "learning_rate": 4.543760144708529e-06, "loss": 0.0112, "step": 10960 }, { "epoch": 0.7532297965915338, "grad_norm": 0.5017193555831909, "learning_rate": 4.541366681139738e-06, "loss": 0.0054, "step": 10961 }, { "epoch": 0.7532985156679495, "grad_norm": 0.12502190470695496, "learning_rate": 4.538973735670135e-06, "loss": 0.0085, "step": 10962 }, { "epoch": 0.7533672347443651, "grad_norm": 0.3785174489021301, "learning_rate": 4.536581308418267e-06, "loss": 0.0082, "step": 10963 }, { "epoch": 0.7534359538207807, "grad_norm": 0.8652437329292297, "learning_rate": 4.534189399502654e-06, "loss": 0.0088, "step": 10964 }, { "epoch": 0.7535046728971962, "grad_norm": 0.4370625615119934, "learning_rate": 4.531798009041776e-06, "loss": 0.0093, "step": 10965 }, { "epoch": 0.7535733919736118, "grad_norm": 0.5425922870635986, "learning_rate": 4.529407137154108e-06, "loss": 0.0159, "step": 10966 }, { "epoch": 0.7536421110500275, "grad_norm": 0.20256860554218292, "learning_rate": 4.527016783958078e-06, "loss": 0.0099, "step": 10967 }, { "epoch": 0.7537108301264431, "grad_norm": 0.5208959579467773, "learning_rate": 4.524626949572111e-06, "loss": 0.0157, "step": 10968 }, { "epoch": 0.7537795492028587, "grad_norm": 0.2978464961051941, "learning_rate": 4.522237634114583e-06, "loss": 0.0156, "step": 10969 }, { "epoch": 0.7538482682792743, "grad_norm": 0.3192078471183777, "learning_rate": 4.519848837703862e-06, "loss": 0.0098, "step": 10970 }, { "epoch": 0.7539169873556899, "grad_norm": 0.5052134394645691, "learning_rate": 4.517460560458282e-06, "loss": 0.0112, "step": 10971 }, { "epoch": 0.7539857064321056, "grad_norm": 0.27382516860961914, "learning_rate": 4.515072802496161e-06, "loss": 0.0045, "step": 10972 }, { "epoch": 0.7540544255085212, "grad_norm": 0.11899454146623611, "learning_rate": 4.51268556393577e-06, "loss": 0.0077, "step": 10973 }, { "epoch": 0.7541231445849368, "grad_norm": 0.3322581350803375, "learning_rate": 4.510298844895383e-06, "loss": 0.0118, "step": 10974 }, { "epoch": 0.7541918636613524, "grad_norm": 0.5433316230773926, "learning_rate": 4.5079126454932185e-06, "loss": 0.0082, "step": 10975 }, { "epoch": 0.754260582737768, "grad_norm": 0.1687699407339096, "learning_rate": 4.505526965847492e-06, "loss": 0.0072, "step": 10976 }, { "epoch": 0.7543293018141837, "grad_norm": 0.33376964926719666, "learning_rate": 4.5031418060763816e-06, "loss": 0.0097, "step": 10977 }, { "epoch": 0.7543980208905993, "grad_norm": 0.19816550612449646, "learning_rate": 4.500757166298052e-06, "loss": 0.0042, "step": 10978 }, { "epoch": 0.7544667399670149, "grad_norm": 0.6889228820800781, "learning_rate": 4.498373046630622e-06, "loss": 0.0139, "step": 10979 }, { "epoch": 0.7545354590434304, "grad_norm": 0.13903161883354187, "learning_rate": 4.495989447192207e-06, "loss": 0.012, "step": 10980 }, { "epoch": 0.754604178119846, "grad_norm": 0.3891623914241791, "learning_rate": 4.493606368100872e-06, "loss": 0.0122, "step": 10981 }, { "epoch": 0.7546728971962616, "grad_norm": 0.7596697807312012, "learning_rate": 4.491223809474678e-06, "loss": 0.0079, "step": 10982 }, { "epoch": 0.7547416162726773, "grad_norm": 0.45808908343315125, "learning_rate": 4.4888417714316574e-06, "loss": 0.0114, "step": 10983 }, { "epoch": 0.7548103353490929, "grad_norm": 0.30729198455810547, "learning_rate": 4.486460254089801e-06, "loss": 0.0055, "step": 10984 }, { "epoch": 0.7548790544255085, "grad_norm": 0.12359623610973358, "learning_rate": 4.4840792575670915e-06, "loss": 0.0093, "step": 10985 }, { "epoch": 0.7549477735019241, "grad_norm": 0.7117589116096497, "learning_rate": 4.481698781981472e-06, "loss": 0.0056, "step": 10986 }, { "epoch": 0.7550164925783397, "grad_norm": 0.5370805263519287, "learning_rate": 4.479318827450875e-06, "loss": 0.0137, "step": 10987 }, { "epoch": 0.7550852116547554, "grad_norm": 0.5150166749954224, "learning_rate": 4.476939394093184e-06, "loss": 0.0118, "step": 10988 }, { "epoch": 0.755153930731171, "grad_norm": 0.4676421284675598, "learning_rate": 4.474560482026289e-06, "loss": 0.0102, "step": 10989 }, { "epoch": 0.7552226498075866, "grad_norm": 0.6003144979476929, "learning_rate": 4.472182091368026e-06, "loss": 0.0203, "step": 10990 }, { "epoch": 0.7552913688840022, "grad_norm": 0.3523618280887604, "learning_rate": 4.469804222236222e-06, "loss": 0.0115, "step": 10991 }, { "epoch": 0.7553600879604178, "grad_norm": 0.4061020612716675, "learning_rate": 4.4674268747486616e-06, "loss": 0.0083, "step": 10992 }, { "epoch": 0.7554288070368335, "grad_norm": 0.3495617210865021, "learning_rate": 4.465050049023126e-06, "loss": 0.0092, "step": 10993 }, { "epoch": 0.755497526113249, "grad_norm": 0.17786361277103424, "learning_rate": 4.462673745177342e-06, "loss": 0.0072, "step": 10994 }, { "epoch": 0.7555662451896646, "grad_norm": 0.3477441668510437, "learning_rate": 4.460297963329047e-06, "loss": 0.0061, "step": 10995 }, { "epoch": 0.7556349642660802, "grad_norm": 0.21080322563648224, "learning_rate": 4.4579227035959165e-06, "loss": 0.0088, "step": 10996 }, { "epoch": 0.7557036833424958, "grad_norm": 0.3122667670249939, "learning_rate": 4.455547966095626e-06, "loss": 0.0075, "step": 10997 }, { "epoch": 0.7557724024189115, "grad_norm": 0.42304471135139465, "learning_rate": 4.453173750945807e-06, "loss": 0.0075, "step": 10998 }, { "epoch": 0.7558411214953271, "grad_norm": 0.2916405498981476, "learning_rate": 4.450800058264082e-06, "loss": 0.0173, "step": 10999 }, { "epoch": 0.7559098405717427, "grad_norm": 0.5477606654167175, "learning_rate": 4.448426888168028e-06, "loss": 0.0088, "step": 11000 }, { "epoch": 0.7559785596481583, "grad_norm": 0.41204601526260376, "learning_rate": 4.446054240775213e-06, "loss": 0.0103, "step": 11001 }, { "epoch": 0.7560472787245739, "grad_norm": 0.3946537375450134, "learning_rate": 4.443682116203178e-06, "loss": 0.0112, "step": 11002 }, { "epoch": 0.7561159978009896, "grad_norm": 0.14049720764160156, "learning_rate": 4.441310514569425e-06, "loss": 0.0086, "step": 11003 }, { "epoch": 0.7561847168774052, "grad_norm": 0.3055465817451477, "learning_rate": 4.438939435991444e-06, "loss": 0.0066, "step": 11004 }, { "epoch": 0.7562534359538208, "grad_norm": 0.6017889976501465, "learning_rate": 4.436568880586685e-06, "loss": 0.0146, "step": 11005 }, { "epoch": 0.7563221550302364, "grad_norm": 0.36538225412368774, "learning_rate": 4.434198848472589e-06, "loss": 0.0084, "step": 11006 }, { "epoch": 0.756390874106652, "grad_norm": 0.3908340036869049, "learning_rate": 4.431829339766558e-06, "loss": 0.013, "step": 11007 }, { "epoch": 0.7564595931830677, "grad_norm": 0.22830067574977875, "learning_rate": 4.4294603545859804e-06, "loss": 0.009, "step": 11008 }, { "epoch": 0.7565283122594832, "grad_norm": 0.2593824863433838, "learning_rate": 4.427091893048198e-06, "loss": 0.0086, "step": 11009 }, { "epoch": 0.7565970313358988, "grad_norm": 0.6026803255081177, "learning_rate": 4.424723955270552e-06, "loss": 0.0144, "step": 11010 }, { "epoch": 0.7566657504123144, "grad_norm": 0.8575921654701233, "learning_rate": 4.4223565413703355e-06, "loss": 0.0191, "step": 11011 }, { "epoch": 0.75673446948873, "grad_norm": 0.18177062273025513, "learning_rate": 4.419989651464828e-06, "loss": 0.0188, "step": 11012 }, { "epoch": 0.7568031885651457, "grad_norm": 0.5627562999725342, "learning_rate": 4.417623285671282e-06, "loss": 0.0152, "step": 11013 }, { "epoch": 0.7568719076415613, "grad_norm": 0.3738047480583191, "learning_rate": 4.415257444106926e-06, "loss": 0.0133, "step": 11014 }, { "epoch": 0.7569406267179769, "grad_norm": 0.2796686589717865, "learning_rate": 4.412892126888952e-06, "loss": 0.0092, "step": 11015 }, { "epoch": 0.7570093457943925, "grad_norm": 0.32862094044685364, "learning_rate": 4.410527334134538e-06, "loss": 0.0093, "step": 11016 }, { "epoch": 0.7570780648708081, "grad_norm": 0.6234142780303955, "learning_rate": 4.4081630659608265e-06, "loss": 0.0197, "step": 11017 }, { "epoch": 0.7571467839472238, "grad_norm": 0.10050836950540543, "learning_rate": 4.405799322484939e-06, "loss": 0.0168, "step": 11018 }, { "epoch": 0.7572155030236394, "grad_norm": 0.5830414891242981, "learning_rate": 4.403436103823978e-06, "loss": 0.0127, "step": 11019 }, { "epoch": 0.757284222100055, "grad_norm": 0.35987579822540283, "learning_rate": 4.401073410095003e-06, "loss": 0.0108, "step": 11020 }, { "epoch": 0.7573529411764706, "grad_norm": 0.4382718503475189, "learning_rate": 4.3987112414150634e-06, "loss": 0.0126, "step": 11021 }, { "epoch": 0.7574216602528862, "grad_norm": 1.2208386659622192, "learning_rate": 4.39634959790117e-06, "loss": 0.0139, "step": 11022 }, { "epoch": 0.7574903793293019, "grad_norm": 0.6818383932113647, "learning_rate": 4.393988479670322e-06, "loss": 0.0093, "step": 11023 }, { "epoch": 0.7575590984057174, "grad_norm": 0.3779360055923462, "learning_rate": 4.391627886839471e-06, "loss": 0.0083, "step": 11024 }, { "epoch": 0.757627817482133, "grad_norm": 1.2476911544799805, "learning_rate": 4.389267819525573e-06, "loss": 0.0132, "step": 11025 }, { "epoch": 0.7576965365585486, "grad_norm": 0.3265463709831238, "learning_rate": 4.386908277845528e-06, "loss": 0.0131, "step": 11026 }, { "epoch": 0.7577652556349642, "grad_norm": 0.12861116230487823, "learning_rate": 4.384549261916231e-06, "loss": 0.0106, "step": 11027 }, { "epoch": 0.7578339747113799, "grad_norm": 0.5032350420951843, "learning_rate": 4.3821907718545375e-06, "loss": 0.0112, "step": 11028 }, { "epoch": 0.7579026937877955, "grad_norm": 0.4694538712501526, "learning_rate": 4.379832807777286e-06, "loss": 0.0163, "step": 11029 }, { "epoch": 0.7579714128642111, "grad_norm": 0.212490051984787, "learning_rate": 4.377475369801274e-06, "loss": 0.0053, "step": 11030 }, { "epoch": 0.7580401319406267, "grad_norm": 0.5309178233146667, "learning_rate": 4.3751184580433025e-06, "loss": 0.0081, "step": 11031 }, { "epoch": 0.7581088510170423, "grad_norm": 0.9232907295227051, "learning_rate": 4.372762072620115e-06, "loss": 0.0161, "step": 11032 }, { "epoch": 0.758177570093458, "grad_norm": 0.09724525362253189, "learning_rate": 4.37040621364845e-06, "loss": 0.0066, "step": 11033 }, { "epoch": 0.7582462891698736, "grad_norm": 0.4571557939052582, "learning_rate": 4.368050881245002e-06, "loss": 0.0109, "step": 11034 }, { "epoch": 0.7583150082462892, "grad_norm": 0.2619738280773163, "learning_rate": 4.3656960755264604e-06, "loss": 0.0056, "step": 11035 }, { "epoch": 0.7583837273227048, "grad_norm": 0.29653218388557434, "learning_rate": 4.363341796609468e-06, "loss": 0.0062, "step": 11036 }, { "epoch": 0.7584524463991204, "grad_norm": 0.4635404944419861, "learning_rate": 4.3609880446106545e-06, "loss": 0.0186, "step": 11037 }, { "epoch": 0.758521165475536, "grad_norm": 0.09261088073253632, "learning_rate": 4.358634819646624e-06, "loss": 0.0129, "step": 11038 }, { "epoch": 0.7585898845519516, "grad_norm": 0.6370511054992676, "learning_rate": 4.3562821218339445e-06, "loss": 0.0104, "step": 11039 }, { "epoch": 0.7586586036283672, "grad_norm": 0.18967102468013763, "learning_rate": 4.3539299512891684e-06, "loss": 0.0125, "step": 11040 }, { "epoch": 0.7587273227047828, "grad_norm": 0.28821972012519836, "learning_rate": 4.351578308128812e-06, "loss": 0.0069, "step": 11041 }, { "epoch": 0.7587960417811984, "grad_norm": 0.6335291266441345, "learning_rate": 4.349227192469373e-06, "loss": 0.032, "step": 11042 }, { "epoch": 0.7588647608576141, "grad_norm": 0.28249672055244446, "learning_rate": 4.346876604427323e-06, "loss": 0.0064, "step": 11043 }, { "epoch": 0.7589334799340297, "grad_norm": 1.014888048171997, "learning_rate": 4.3445265441191075e-06, "loss": 0.0137, "step": 11044 }, { "epoch": 0.7590021990104453, "grad_norm": 0.08508094400167465, "learning_rate": 4.342177011661136e-06, "loss": 0.0108, "step": 11045 }, { "epoch": 0.7590709180868609, "grad_norm": 0.3913714587688446, "learning_rate": 4.339828007169809e-06, "loss": 0.0144, "step": 11046 }, { "epoch": 0.7591396371632765, "grad_norm": 0.330854594707489, "learning_rate": 4.33747953076148e-06, "loss": 0.0107, "step": 11047 }, { "epoch": 0.7592083562396922, "grad_norm": 0.16119244694709778, "learning_rate": 4.335131582552495e-06, "loss": 0.0116, "step": 11048 }, { "epoch": 0.7592770753161078, "grad_norm": 0.30441346764564514, "learning_rate": 4.3327841626591655e-06, "loss": 0.0069, "step": 11049 }, { "epoch": 0.7593457943925234, "grad_norm": 0.2378556728363037, "learning_rate": 4.330437271197782e-06, "loss": 0.0074, "step": 11050 }, { "epoch": 0.759414513468939, "grad_norm": 0.2716217041015625, "learning_rate": 4.328090908284597e-06, "loss": 0.0077, "step": 11051 }, { "epoch": 0.7594832325453545, "grad_norm": 0.26155155897140503, "learning_rate": 4.32574507403585e-06, "loss": 0.009, "step": 11052 }, { "epoch": 0.7595519516217702, "grad_norm": 0.6397151947021484, "learning_rate": 4.323399768567744e-06, "loss": 0.0211, "step": 11053 }, { "epoch": 0.7596206706981858, "grad_norm": 0.17507624626159668, "learning_rate": 4.321054991996462e-06, "loss": 0.0125, "step": 11054 }, { "epoch": 0.7596893897746014, "grad_norm": 0.7124530673027039, "learning_rate": 4.318710744438167e-06, "loss": 0.0081, "step": 11055 }, { "epoch": 0.759758108851017, "grad_norm": 0.2751106917858124, "learning_rate": 4.316367026008976e-06, "loss": 0.0047, "step": 11056 }, { "epoch": 0.7598268279274326, "grad_norm": 0.6229171752929688, "learning_rate": 4.3140238368250035e-06, "loss": 0.0114, "step": 11057 }, { "epoch": 0.7598955470038483, "grad_norm": 0.46671032905578613, "learning_rate": 4.311681177002315e-06, "loss": 0.0092, "step": 11058 }, { "epoch": 0.7599642660802639, "grad_norm": 0.1478615701198578, "learning_rate": 4.309339046656966e-06, "loss": 0.0106, "step": 11059 }, { "epoch": 0.7600329851566795, "grad_norm": 0.5908471941947937, "learning_rate": 4.306997445904983e-06, "loss": 0.0087, "step": 11060 }, { "epoch": 0.7601017042330951, "grad_norm": 0.32305705547332764, "learning_rate": 4.304656374862368e-06, "loss": 0.0076, "step": 11061 }, { "epoch": 0.7601704233095107, "grad_norm": 0.5363233089447021, "learning_rate": 4.302315833645081e-06, "loss": 0.0184, "step": 11062 }, { "epoch": 0.7602391423859264, "grad_norm": 0.5291239619255066, "learning_rate": 4.29997582236908e-06, "loss": 0.0087, "step": 11063 }, { "epoch": 0.760307861462342, "grad_norm": 1.4319977760314941, "learning_rate": 4.297636341150274e-06, "loss": 0.0088, "step": 11064 }, { "epoch": 0.7603765805387576, "grad_norm": 0.2875942885875702, "learning_rate": 4.295297390104564e-06, "loss": 0.0144, "step": 11065 }, { "epoch": 0.7604452996151732, "grad_norm": 0.6710678935050964, "learning_rate": 4.292958969347806e-06, "loss": 0.0146, "step": 11066 }, { "epoch": 0.7605140186915887, "grad_norm": 0.2739047706127167, "learning_rate": 4.290621078995855e-06, "loss": 0.0074, "step": 11067 }, { "epoch": 0.7605827377680044, "grad_norm": 0.6747468113899231, "learning_rate": 4.288283719164515e-06, "loss": 0.0121, "step": 11068 }, { "epoch": 0.76065145684442, "grad_norm": 0.37473028898239136, "learning_rate": 4.285946889969581e-06, "loss": 0.0107, "step": 11069 }, { "epoch": 0.7607201759208356, "grad_norm": 0.70103919506073, "learning_rate": 4.283610591526806e-06, "loss": 0.0122, "step": 11070 }, { "epoch": 0.7607888949972512, "grad_norm": 0.26036038994789124, "learning_rate": 4.281274823951935e-06, "loss": 0.0184, "step": 11071 }, { "epoch": 0.7608576140736668, "grad_norm": 0.19228234887123108, "learning_rate": 4.278939587360669e-06, "loss": 0.0079, "step": 11072 }, { "epoch": 0.7609263331500825, "grad_norm": 0.23837250471115112, "learning_rate": 4.276604881868693e-06, "loss": 0.0089, "step": 11073 }, { "epoch": 0.7609950522264981, "grad_norm": 0.25733286142349243, "learning_rate": 4.2742707075916725e-06, "loss": 0.0074, "step": 11074 }, { "epoch": 0.7610637713029137, "grad_norm": 0.18092870712280273, "learning_rate": 4.271937064645222e-06, "loss": 0.0074, "step": 11075 }, { "epoch": 0.7611324903793293, "grad_norm": 0.3940427601337433, "learning_rate": 4.269603953144954e-06, "loss": 0.0081, "step": 11076 }, { "epoch": 0.7612012094557449, "grad_norm": 0.2570336163043976, "learning_rate": 4.267271373206449e-06, "loss": 0.0059, "step": 11077 }, { "epoch": 0.7612699285321606, "grad_norm": 0.3800840973854065, "learning_rate": 4.264939324945253e-06, "loss": 0.0077, "step": 11078 }, { "epoch": 0.7613386476085762, "grad_norm": 0.2207387089729309, "learning_rate": 4.262607808476891e-06, "loss": 0.0075, "step": 11079 }, { "epoch": 0.7614073666849918, "grad_norm": 0.35702720284461975, "learning_rate": 4.2602768239168666e-06, "loss": 0.0146, "step": 11080 }, { "epoch": 0.7614760857614074, "grad_norm": 0.2210330367088318, "learning_rate": 4.257946371380646e-06, "loss": 0.0111, "step": 11081 }, { "epoch": 0.7615448048378229, "grad_norm": 0.0977528914809227, "learning_rate": 4.2556164509836786e-06, "loss": 0.0063, "step": 11082 }, { "epoch": 0.7616135239142386, "grad_norm": 0.16140107810497284, "learning_rate": 4.253287062841381e-06, "loss": 0.0042, "step": 11083 }, { "epoch": 0.7616822429906542, "grad_norm": 0.6419059634208679, "learning_rate": 4.250958207069147e-06, "loss": 0.016, "step": 11084 }, { "epoch": 0.7617509620670698, "grad_norm": 0.29037946462631226, "learning_rate": 4.2486298837823446e-06, "loss": 0.0054, "step": 11085 }, { "epoch": 0.7618196811434854, "grad_norm": 0.2696983516216278, "learning_rate": 4.246302093096319e-06, "loss": 0.0177, "step": 11086 }, { "epoch": 0.761888400219901, "grad_norm": 0.6881622672080994, "learning_rate": 4.243974835126372e-06, "loss": 0.0117, "step": 11087 }, { "epoch": 0.7619571192963167, "grad_norm": 0.3556789755821228, "learning_rate": 4.241648109987805e-06, "loss": 0.0113, "step": 11088 }, { "epoch": 0.7620258383727323, "grad_norm": 0.17998798191547394, "learning_rate": 4.239321917795868e-06, "loss": 0.0106, "step": 11089 }, { "epoch": 0.7620945574491479, "grad_norm": 0.758729100227356, "learning_rate": 4.236996258665799e-06, "loss": 0.0151, "step": 11090 }, { "epoch": 0.7621632765255635, "grad_norm": 0.2872949242591858, "learning_rate": 4.234671132712814e-06, "loss": 0.0129, "step": 11091 }, { "epoch": 0.7622319956019791, "grad_norm": 0.6397547721862793, "learning_rate": 4.232346540052083e-06, "loss": 0.0058, "step": 11092 }, { "epoch": 0.7623007146783948, "grad_norm": 0.4350166916847229, "learning_rate": 4.230022480798766e-06, "loss": 0.0093, "step": 11093 }, { "epoch": 0.7623694337548104, "grad_norm": 0.9020873308181763, "learning_rate": 4.227698955067998e-06, "loss": 0.0103, "step": 11094 }, { "epoch": 0.762438152831226, "grad_norm": 0.3523632884025574, "learning_rate": 4.225375962974873e-06, "loss": 0.016, "step": 11095 }, { "epoch": 0.7625068719076415, "grad_norm": 0.353203684091568, "learning_rate": 4.223053504634471e-06, "loss": 0.017, "step": 11096 }, { "epoch": 0.7625755909840571, "grad_norm": 0.3243188261985779, "learning_rate": 4.220731580161847e-06, "loss": 0.0098, "step": 11097 }, { "epoch": 0.7626443100604728, "grad_norm": 0.22704574465751648, "learning_rate": 4.218410189672014e-06, "loss": 0.0063, "step": 11098 }, { "epoch": 0.7627130291368884, "grad_norm": 0.8514379858970642, "learning_rate": 4.216089333279979e-06, "loss": 0.0106, "step": 11099 }, { "epoch": 0.762781748213304, "grad_norm": 0.641425371170044, "learning_rate": 4.213769011100703e-06, "loss": 0.0083, "step": 11100 }, { "epoch": 0.7628504672897196, "grad_norm": 0.5255029201507568, "learning_rate": 4.211449223249138e-06, "loss": 0.0074, "step": 11101 }, { "epoch": 0.7629191863661352, "grad_norm": 0.4642984867095947, "learning_rate": 4.20912996984019e-06, "loss": 0.0089, "step": 11102 }, { "epoch": 0.7629879054425509, "grad_norm": 1.1964970827102661, "learning_rate": 4.206811250988767e-06, "loss": 0.0129, "step": 11103 }, { "epoch": 0.7630566245189665, "grad_norm": 0.36996227502822876, "learning_rate": 4.204493066809721e-06, "loss": 0.009, "step": 11104 }, { "epoch": 0.7631253435953821, "grad_norm": 0.11391294747591019, "learning_rate": 4.202175417417898e-06, "loss": 0.0085, "step": 11105 }, { "epoch": 0.7631940626717977, "grad_norm": 0.33593347668647766, "learning_rate": 4.1998583029281016e-06, "loss": 0.0123, "step": 11106 }, { "epoch": 0.7632627817482133, "grad_norm": 0.39183202385902405, "learning_rate": 4.197541723455125e-06, "loss": 0.0166, "step": 11107 }, { "epoch": 0.763331500824629, "grad_norm": 0.15757307410240173, "learning_rate": 4.195225679113713e-06, "loss": 0.0158, "step": 11108 }, { "epoch": 0.7634002199010446, "grad_norm": 1.22898268699646, "learning_rate": 4.1929101700186165e-06, "loss": 0.0144, "step": 11109 }, { "epoch": 0.7634689389774602, "grad_norm": 0.15723660588264465, "learning_rate": 4.190595196284527e-06, "loss": 0.0077, "step": 11110 }, { "epoch": 0.7635376580538757, "grad_norm": 0.31922289729118347, "learning_rate": 4.188280758026131e-06, "loss": 0.012, "step": 11111 }, { "epoch": 0.7636063771302913, "grad_norm": 0.7942368388175964, "learning_rate": 4.185966855358076e-06, "loss": 0.0095, "step": 11112 }, { "epoch": 0.7636750962067069, "grad_norm": 0.25685545802116394, "learning_rate": 4.1836534883949945e-06, "loss": 0.0044, "step": 11113 }, { "epoch": 0.7637438152831226, "grad_norm": 0.31876325607299805, "learning_rate": 4.181340657251477e-06, "loss": 0.0125, "step": 11114 }, { "epoch": 0.7638125343595382, "grad_norm": 0.11726253479719162, "learning_rate": 4.179028362042101e-06, "loss": 0.0095, "step": 11115 }, { "epoch": 0.7638812534359538, "grad_norm": 0.3944268524646759, "learning_rate": 4.176716602881419e-06, "loss": 0.0047, "step": 11116 }, { "epoch": 0.7639499725123694, "grad_norm": 0.23151446878910065, "learning_rate": 4.174405379883938e-06, "loss": 0.0095, "step": 11117 }, { "epoch": 0.764018691588785, "grad_norm": 0.26074790954589844, "learning_rate": 4.172094693164164e-06, "loss": 0.0041, "step": 11118 }, { "epoch": 0.7640874106652007, "grad_norm": 0.20020072162151337, "learning_rate": 4.169784542836551e-06, "loss": 0.0086, "step": 11119 }, { "epoch": 0.7641561297416163, "grad_norm": 1.3525162935256958, "learning_rate": 4.167474929015549e-06, "loss": 0.0115, "step": 11120 }, { "epoch": 0.7642248488180319, "grad_norm": 0.4083735942840576, "learning_rate": 4.165165851815566e-06, "loss": 0.0088, "step": 11121 }, { "epoch": 0.7642935678944475, "grad_norm": 0.5194078683853149, "learning_rate": 4.162857311350996e-06, "loss": 0.0183, "step": 11122 }, { "epoch": 0.7643622869708631, "grad_norm": 0.24766752123832703, "learning_rate": 4.160549307736191e-06, "loss": 0.0101, "step": 11123 }, { "epoch": 0.7644310060472788, "grad_norm": 0.24421918392181396, "learning_rate": 4.158241841085493e-06, "loss": 0.0099, "step": 11124 }, { "epoch": 0.7644997251236944, "grad_norm": 0.4000333249568939, "learning_rate": 4.155934911513198e-06, "loss": 0.0103, "step": 11125 }, { "epoch": 0.7645684442001099, "grad_norm": 0.3227810859680176, "learning_rate": 4.153628519133594e-06, "loss": 0.0102, "step": 11126 }, { "epoch": 0.7646371632765255, "grad_norm": 0.7504376769065857, "learning_rate": 4.151322664060936e-06, "loss": 0.0064, "step": 11127 }, { "epoch": 0.7647058823529411, "grad_norm": 0.33269888162612915, "learning_rate": 4.149017346409452e-06, "loss": 0.0083, "step": 11128 }, { "epoch": 0.7647746014293568, "grad_norm": 0.2972147762775421, "learning_rate": 4.146712566293336e-06, "loss": 0.0096, "step": 11129 }, { "epoch": 0.7648433205057724, "grad_norm": 0.6338471174240112, "learning_rate": 4.144408323826769e-06, "loss": 0.0084, "step": 11130 }, { "epoch": 0.764912039582188, "grad_norm": 0.30216193199157715, "learning_rate": 4.142104619123893e-06, "loss": 0.0069, "step": 11131 }, { "epoch": 0.7649807586586036, "grad_norm": 0.7477177977561951, "learning_rate": 4.1398014522988285e-06, "loss": 0.018, "step": 11132 }, { "epoch": 0.7650494777350192, "grad_norm": 0.6267127394676208, "learning_rate": 4.1374988234656795e-06, "loss": 0.011, "step": 11133 }, { "epoch": 0.7651181968114349, "grad_norm": 1.0499193668365479, "learning_rate": 4.135196732738501e-06, "loss": 0.0079, "step": 11134 }, { "epoch": 0.7651869158878505, "grad_norm": 0.3581017255783081, "learning_rate": 4.132895180231345e-06, "loss": 0.0096, "step": 11135 }, { "epoch": 0.7652556349642661, "grad_norm": 0.27274712920188904, "learning_rate": 4.130594166058214e-06, "loss": 0.01, "step": 11136 }, { "epoch": 0.7653243540406817, "grad_norm": 0.6441911458969116, "learning_rate": 4.128293690333106e-06, "loss": 0.0143, "step": 11137 }, { "epoch": 0.7653930731170973, "grad_norm": 0.17927823960781097, "learning_rate": 4.125993753169969e-06, "loss": 0.0138, "step": 11138 }, { "epoch": 0.765461792193513, "grad_norm": 0.37835270166397095, "learning_rate": 4.123694354682755e-06, "loss": 0.013, "step": 11139 }, { "epoch": 0.7655305112699285, "grad_norm": 0.373073548078537, "learning_rate": 4.121395494985355e-06, "loss": 0.0175, "step": 11140 }, { "epoch": 0.7655992303463441, "grad_norm": 0.28558579087257385, "learning_rate": 4.119097174191664e-06, "loss": 0.0084, "step": 11141 }, { "epoch": 0.7656679494227597, "grad_norm": 0.8345673084259033, "learning_rate": 4.116799392415523e-06, "loss": 0.0222, "step": 11142 }, { "epoch": 0.7657366684991753, "grad_norm": 0.4162716269493103, "learning_rate": 4.11450214977077e-06, "loss": 0.0199, "step": 11143 }, { "epoch": 0.765805387575591, "grad_norm": 0.5160726308822632, "learning_rate": 4.112205446371191e-06, "loss": 0.0123, "step": 11144 }, { "epoch": 0.7658741066520066, "grad_norm": 0.5627215504646301, "learning_rate": 4.109909282330579e-06, "loss": 0.0079, "step": 11145 }, { "epoch": 0.7659428257284222, "grad_norm": 0.24441814422607422, "learning_rate": 4.107613657762669e-06, "loss": 0.008, "step": 11146 }, { "epoch": 0.7660115448048378, "grad_norm": 0.280112624168396, "learning_rate": 4.1053185727811875e-06, "loss": 0.0092, "step": 11147 }, { "epoch": 0.7660802638812534, "grad_norm": 0.4240284562110901, "learning_rate": 4.103024027499823e-06, "loss": 0.0159, "step": 11148 }, { "epoch": 0.7661489829576691, "grad_norm": 0.35292765498161316, "learning_rate": 4.100730022032246e-06, "loss": 0.005, "step": 11149 }, { "epoch": 0.7662177020340847, "grad_norm": 0.26110410690307617, "learning_rate": 4.0984365564920944e-06, "loss": 0.0082, "step": 11150 }, { "epoch": 0.7662864211105003, "grad_norm": 0.6841548085212708, "learning_rate": 4.0961436309929815e-06, "loss": 0.0089, "step": 11151 }, { "epoch": 0.7663551401869159, "grad_norm": 0.43654534220695496, "learning_rate": 4.0938512456485e-06, "loss": 0.008, "step": 11152 }, { "epoch": 0.7664238592633315, "grad_norm": 0.142215296626091, "learning_rate": 4.091559400572204e-06, "loss": 0.0188, "step": 11153 }, { "epoch": 0.7664925783397472, "grad_norm": 0.1349441409111023, "learning_rate": 4.089268095877631e-06, "loss": 0.0188, "step": 11154 }, { "epoch": 0.7665612974161627, "grad_norm": 0.3453144133090973, "learning_rate": 4.08697733167828e-06, "loss": 0.0097, "step": 11155 }, { "epoch": 0.7666300164925783, "grad_norm": 0.37191149592399597, "learning_rate": 4.084687108087635e-06, "loss": 0.0153, "step": 11156 }, { "epoch": 0.7666987355689939, "grad_norm": 0.7910171151161194, "learning_rate": 4.082397425219151e-06, "loss": 0.0126, "step": 11157 }, { "epoch": 0.7667674546454095, "grad_norm": 0.40832602977752686, "learning_rate": 4.080108283186258e-06, "loss": 0.0072, "step": 11158 }, { "epoch": 0.7668361737218252, "grad_norm": 0.6758573651313782, "learning_rate": 4.0778196821023455e-06, "loss": 0.0116, "step": 11159 }, { "epoch": 0.7669048927982408, "grad_norm": 1.2704672813415527, "learning_rate": 4.075531622080795e-06, "loss": 0.0164, "step": 11160 }, { "epoch": 0.7669736118746564, "grad_norm": 0.3066873550415039, "learning_rate": 4.073244103234943e-06, "loss": 0.0089, "step": 11161 }, { "epoch": 0.767042330951072, "grad_norm": 0.25757932662963867, "learning_rate": 4.070957125678115e-06, "loss": 0.0075, "step": 11162 }, { "epoch": 0.7671110500274876, "grad_norm": 0.3912476599216461, "learning_rate": 4.068670689523601e-06, "loss": 0.0146, "step": 11163 }, { "epoch": 0.7671797691039033, "grad_norm": 0.47854095697402954, "learning_rate": 4.066384794884673e-06, "loss": 0.0099, "step": 11164 }, { "epoch": 0.7672484881803189, "grad_norm": 0.3615367114543915, "learning_rate": 4.064099441874557e-06, "loss": 0.0064, "step": 11165 }, { "epoch": 0.7673172072567345, "grad_norm": 0.40222424268722534, "learning_rate": 4.061814630606477e-06, "loss": 0.0111, "step": 11166 }, { "epoch": 0.7673859263331501, "grad_norm": 0.7860857844352722, "learning_rate": 4.059530361193608e-06, "loss": 0.0081, "step": 11167 }, { "epoch": 0.7674546454095657, "grad_norm": 0.3984980881214142, "learning_rate": 4.057246633749112e-06, "loss": 0.0161, "step": 11168 }, { "epoch": 0.7675233644859814, "grad_norm": 0.5940678119659424, "learning_rate": 4.054963448386124e-06, "loss": 0.0089, "step": 11169 }, { "epoch": 0.7675920835623969, "grad_norm": 0.5739973783493042, "learning_rate": 4.0526808052177424e-06, "loss": 0.0076, "step": 11170 }, { "epoch": 0.7676608026388125, "grad_norm": 0.1394316405057907, "learning_rate": 4.05039870435705e-06, "loss": 0.0069, "step": 11171 }, { "epoch": 0.7677295217152281, "grad_norm": 0.09413712471723557, "learning_rate": 4.04811714591709e-06, "loss": 0.0062, "step": 11172 }, { "epoch": 0.7677982407916437, "grad_norm": 0.7209336757659912, "learning_rate": 4.045836130010894e-06, "loss": 0.0122, "step": 11173 }, { "epoch": 0.7678669598680594, "grad_norm": 0.2526978552341461, "learning_rate": 4.0435556567514485e-06, "loss": 0.0116, "step": 11174 }, { "epoch": 0.767935678944475, "grad_norm": 0.38334402441978455, "learning_rate": 4.041275726251738e-06, "loss": 0.0072, "step": 11175 }, { "epoch": 0.7680043980208906, "grad_norm": 0.4681323766708374, "learning_rate": 4.038996338624693e-06, "loss": 0.0092, "step": 11176 }, { "epoch": 0.7680731170973062, "grad_norm": 0.06315590441226959, "learning_rate": 4.0367174939832406e-06, "loss": 0.0037, "step": 11177 }, { "epoch": 0.7681418361737218, "grad_norm": 0.19706803560256958, "learning_rate": 4.034439192440258e-06, "loss": 0.0121, "step": 11178 }, { "epoch": 0.7682105552501375, "grad_norm": 0.5863069295883179, "learning_rate": 4.032161434108618e-06, "loss": 0.0194, "step": 11179 }, { "epoch": 0.7682792743265531, "grad_norm": 0.22099636495113373, "learning_rate": 4.029884219101142e-06, "loss": 0.006, "step": 11180 }, { "epoch": 0.7683479934029687, "grad_norm": 0.10690096020698547, "learning_rate": 4.027607547530659e-06, "loss": 0.0057, "step": 11181 }, { "epoch": 0.7684167124793843, "grad_norm": 0.31031250953674316, "learning_rate": 4.025331419509936e-06, "loss": 0.0155, "step": 11182 }, { "epoch": 0.7684854315557998, "grad_norm": 0.281292587518692, "learning_rate": 4.023055835151734e-06, "loss": 0.0046, "step": 11183 }, { "epoch": 0.7685541506322155, "grad_norm": 0.27244827151298523, "learning_rate": 4.020780794568773e-06, "loss": 0.0082, "step": 11184 }, { "epoch": 0.7686228697086311, "grad_norm": 0.4787614643573761, "learning_rate": 4.018506297873766e-06, "loss": 0.0213, "step": 11185 }, { "epoch": 0.7686915887850467, "grad_norm": 0.49620574712753296, "learning_rate": 4.016232345179375e-06, "loss": 0.0062, "step": 11186 }, { "epoch": 0.7687603078614623, "grad_norm": 0.44961321353912354, "learning_rate": 4.01395893659825e-06, "loss": 0.0125, "step": 11187 }, { "epoch": 0.7688290269378779, "grad_norm": 0.6979424953460693, "learning_rate": 4.01168607224302e-06, "loss": 0.0121, "step": 11188 }, { "epoch": 0.7688977460142936, "grad_norm": 0.5097019076347351, "learning_rate": 4.009413752226265e-06, "loss": 0.0152, "step": 11189 }, { "epoch": 0.7689664650907092, "grad_norm": 0.4397869408130646, "learning_rate": 4.007141976660562e-06, "loss": 0.0087, "step": 11190 }, { "epoch": 0.7690351841671248, "grad_norm": 1.6783649921417236, "learning_rate": 4.004870745658441e-06, "loss": 0.0066, "step": 11191 }, { "epoch": 0.7691039032435404, "grad_norm": 0.299203485250473, "learning_rate": 4.002600059332422e-06, "loss": 0.0093, "step": 11192 }, { "epoch": 0.769172622319956, "grad_norm": 0.54754638671875, "learning_rate": 4.000329917794977e-06, "loss": 0.0128, "step": 11193 }, { "epoch": 0.7692413413963717, "grad_norm": 0.12185747176408768, "learning_rate": 3.998060321158584e-06, "loss": 0.0063, "step": 11194 }, { "epoch": 0.7693100604727873, "grad_norm": 0.5880566835403442, "learning_rate": 3.995791269535659e-06, "loss": 0.0111, "step": 11195 }, { "epoch": 0.7693787795492029, "grad_norm": 0.6542816758155823, "learning_rate": 3.993522763038614e-06, "loss": 0.0131, "step": 11196 }, { "epoch": 0.7694474986256185, "grad_norm": 0.26102444529533386, "learning_rate": 3.991254801779819e-06, "loss": 0.0052, "step": 11197 }, { "epoch": 0.769516217702034, "grad_norm": 0.14068756997585297, "learning_rate": 3.988987385871632e-06, "loss": 0.0075, "step": 11198 }, { "epoch": 0.7695849367784497, "grad_norm": 0.14180132746696472, "learning_rate": 3.9867205154263645e-06, "loss": 0.0075, "step": 11199 }, { "epoch": 0.7696536558548653, "grad_norm": 0.639733076095581, "learning_rate": 3.984454190556329e-06, "loss": 0.0125, "step": 11200 }, { "epoch": 0.7697223749312809, "grad_norm": 0.6481134295463562, "learning_rate": 3.9821884113737806e-06, "loss": 0.0142, "step": 11201 }, { "epoch": 0.7697910940076965, "grad_norm": 0.625598669052124, "learning_rate": 3.979923177990972e-06, "loss": 0.0117, "step": 11202 }, { "epoch": 0.7698598130841121, "grad_norm": 0.45024028420448303, "learning_rate": 3.977658490520108e-06, "loss": 0.0121, "step": 11203 }, { "epoch": 0.7699285321605278, "grad_norm": 0.5049562454223633, "learning_rate": 3.975394349073386e-06, "loss": 0.014, "step": 11204 }, { "epoch": 0.7699972512369434, "grad_norm": 0.24078942835330963, "learning_rate": 3.9731307537629575e-06, "loss": 0.0068, "step": 11205 }, { "epoch": 0.770065970313359, "grad_norm": 0.46210339665412903, "learning_rate": 3.970867704700961e-06, "loss": 0.0125, "step": 11206 }, { "epoch": 0.7701346893897746, "grad_norm": 0.39184504747390747, "learning_rate": 3.9686052019995075e-06, "loss": 0.0081, "step": 11207 }, { "epoch": 0.7702034084661902, "grad_norm": 0.43782782554626465, "learning_rate": 3.966343245770669e-06, "loss": 0.0064, "step": 11208 }, { "epoch": 0.7702721275426059, "grad_norm": 1.302216649055481, "learning_rate": 3.9640818361265065e-06, "loss": 0.0087, "step": 11209 }, { "epoch": 0.7703408466190215, "grad_norm": 0.486625611782074, "learning_rate": 3.961820973179037e-06, "loss": 0.0147, "step": 11210 }, { "epoch": 0.7704095656954371, "grad_norm": 0.21207916736602783, "learning_rate": 3.959560657040263e-06, "loss": 0.0086, "step": 11211 }, { "epoch": 0.7704782847718527, "grad_norm": 0.16350020468235016, "learning_rate": 3.957300887822155e-06, "loss": 0.0049, "step": 11212 }, { "epoch": 0.7705470038482682, "grad_norm": 0.4598483145236969, "learning_rate": 3.955041665636663e-06, "loss": 0.0141, "step": 11213 }, { "epoch": 0.7706157229246839, "grad_norm": 0.5907942056655884, "learning_rate": 3.9527829905956946e-06, "loss": 0.0105, "step": 11214 }, { "epoch": 0.7706844420010995, "grad_norm": 0.3286445140838623, "learning_rate": 3.9505248628111496e-06, "loss": 0.0089, "step": 11215 }, { "epoch": 0.7707531610775151, "grad_norm": 0.344190776348114, "learning_rate": 3.948267282394882e-06, "loss": 0.0111, "step": 11216 }, { "epoch": 0.7708218801539307, "grad_norm": 0.3320111632347107, "learning_rate": 3.94601024945873e-06, "loss": 0.0094, "step": 11217 }, { "epoch": 0.7708905992303463, "grad_norm": 0.39281967282295227, "learning_rate": 3.943753764114505e-06, "loss": 0.0041, "step": 11218 }, { "epoch": 0.770959318306762, "grad_norm": 0.31794631481170654, "learning_rate": 3.9414978264739935e-06, "loss": 0.0121, "step": 11219 }, { "epoch": 0.7710280373831776, "grad_norm": 0.2098056524991989, "learning_rate": 3.9392424366489404e-06, "loss": 0.0084, "step": 11220 }, { "epoch": 0.7710967564595932, "grad_norm": 0.11630231142044067, "learning_rate": 3.93698759475108e-06, "loss": 0.0089, "step": 11221 }, { "epoch": 0.7711654755360088, "grad_norm": 0.3017328083515167, "learning_rate": 3.934733300892105e-06, "loss": 0.0092, "step": 11222 }, { "epoch": 0.7712341946124244, "grad_norm": 0.49843645095825195, "learning_rate": 3.932479555183693e-06, "loss": 0.0061, "step": 11223 }, { "epoch": 0.7713029136888401, "grad_norm": 0.09866171330213547, "learning_rate": 3.930226357737496e-06, "loss": 0.0131, "step": 11224 }, { "epoch": 0.7713716327652557, "grad_norm": 0.21346160769462585, "learning_rate": 3.927973708665121e-06, "loss": 0.0076, "step": 11225 }, { "epoch": 0.7714403518416713, "grad_norm": 0.32229211926460266, "learning_rate": 3.925721608078168e-06, "loss": 0.0101, "step": 11226 }, { "epoch": 0.7715090709180868, "grad_norm": 0.2544470727443695, "learning_rate": 3.923470056088198e-06, "loss": 0.0098, "step": 11227 }, { "epoch": 0.7715777899945024, "grad_norm": 0.40333685278892517, "learning_rate": 3.921219052806751e-06, "loss": 0.0097, "step": 11228 }, { "epoch": 0.7716465090709181, "grad_norm": 0.5705260038375854, "learning_rate": 3.9189685983453284e-06, "loss": 0.0116, "step": 11229 }, { "epoch": 0.7717152281473337, "grad_norm": 0.3047007620334625, "learning_rate": 3.916718692815427e-06, "loss": 0.0137, "step": 11230 }, { "epoch": 0.7717839472237493, "grad_norm": 0.18746855854988098, "learning_rate": 3.914469336328491e-06, "loss": 0.0114, "step": 11231 }, { "epoch": 0.7718526663001649, "grad_norm": 0.43708914518356323, "learning_rate": 3.912220528995959e-06, "loss": 0.0102, "step": 11232 }, { "epoch": 0.7719213853765805, "grad_norm": 1.0544867515563965, "learning_rate": 3.90997227092922e-06, "loss": 0.0104, "step": 11233 }, { "epoch": 0.7719901044529962, "grad_norm": 0.22373177111148834, "learning_rate": 3.907724562239659e-06, "loss": 0.0057, "step": 11234 }, { "epoch": 0.7720588235294118, "grad_norm": 0.31043675541877747, "learning_rate": 3.905477403038611e-06, "loss": 0.0059, "step": 11235 }, { "epoch": 0.7721275426058274, "grad_norm": 0.4381760358810425, "learning_rate": 3.903230793437412e-06, "loss": 0.0099, "step": 11236 }, { "epoch": 0.772196261682243, "grad_norm": 0.30263984203338623, "learning_rate": 3.90098473354734e-06, "loss": 0.0072, "step": 11237 }, { "epoch": 0.7722649807586586, "grad_norm": 0.25966861844062805, "learning_rate": 3.898739223479671e-06, "loss": 0.0081, "step": 11238 }, { "epoch": 0.7723336998350742, "grad_norm": 0.10182180255651474, "learning_rate": 3.8964942633456326e-06, "loss": 0.0041, "step": 11239 }, { "epoch": 0.7724024189114899, "grad_norm": 0.35018298029899597, "learning_rate": 3.894249853256445e-06, "loss": 0.0109, "step": 11240 }, { "epoch": 0.7724711379879055, "grad_norm": 0.5755449533462524, "learning_rate": 3.892005993323282e-06, "loss": 0.0141, "step": 11241 }, { "epoch": 0.772539857064321, "grad_norm": 0.4390212893486023, "learning_rate": 3.889762683657308e-06, "loss": 0.0129, "step": 11242 }, { "epoch": 0.7726085761407366, "grad_norm": 1.2043092250823975, "learning_rate": 3.887519924369651e-06, "loss": 0.017, "step": 11243 }, { "epoch": 0.7726772952171522, "grad_norm": 0.30411043763160706, "learning_rate": 3.885277715571406e-06, "loss": 0.0151, "step": 11244 }, { "epoch": 0.7727460142935679, "grad_norm": 0.21905885636806488, "learning_rate": 3.883036057373657e-06, "loss": 0.0083, "step": 11245 }, { "epoch": 0.7728147333699835, "grad_norm": 0.6389356255531311, "learning_rate": 3.880794949887441e-06, "loss": 0.007, "step": 11246 }, { "epoch": 0.7728834524463991, "grad_norm": 0.21461467444896698, "learning_rate": 3.878554393223784e-06, "loss": 0.0082, "step": 11247 }, { "epoch": 0.7729521715228147, "grad_norm": 0.31881317496299744, "learning_rate": 3.876314387493678e-06, "loss": 0.0098, "step": 11248 }, { "epoch": 0.7730208905992303, "grad_norm": 0.37145063281059265, "learning_rate": 3.874074932808093e-06, "loss": 0.0198, "step": 11249 }, { "epoch": 0.773089609675646, "grad_norm": 0.42422497272491455, "learning_rate": 3.8718360292779565e-06, "loss": 0.0048, "step": 11250 }, { "epoch": 0.7731583287520616, "grad_norm": 0.42003315687179565, "learning_rate": 3.869597677014189e-06, "loss": 0.0081, "step": 11251 }, { "epoch": 0.7732270478284772, "grad_norm": 0.3423506021499634, "learning_rate": 3.867359876127663e-06, "loss": 0.0062, "step": 11252 }, { "epoch": 0.7732957669048928, "grad_norm": 0.6640177369117737, "learning_rate": 3.865122626729243e-06, "loss": 0.0181, "step": 11253 }, { "epoch": 0.7733644859813084, "grad_norm": 0.4672119617462158, "learning_rate": 3.862885928929756e-06, "loss": 0.012, "step": 11254 }, { "epoch": 0.7734332050577241, "grad_norm": 0.9498290419578552, "learning_rate": 3.860649782840007e-06, "loss": 0.0167, "step": 11255 }, { "epoch": 0.7735019241341397, "grad_norm": 0.27025750279426575, "learning_rate": 3.8584141885707626e-06, "loss": 0.0065, "step": 11256 }, { "epoch": 0.7735706432105552, "grad_norm": 0.2856435477733612, "learning_rate": 3.8561791462327765e-06, "loss": 0.0067, "step": 11257 }, { "epoch": 0.7736393622869708, "grad_norm": 0.5525385737419128, "learning_rate": 3.853944655936761e-06, "loss": 0.0074, "step": 11258 }, { "epoch": 0.7737080813633864, "grad_norm": 0.2452784925699234, "learning_rate": 3.851710717793413e-06, "loss": 0.0073, "step": 11259 }, { "epoch": 0.7737768004398021, "grad_norm": 0.16786523163318634, "learning_rate": 3.8494773319134e-06, "loss": 0.0103, "step": 11260 }, { "epoch": 0.7738455195162177, "grad_norm": 0.4163820743560791, "learning_rate": 3.84724449840735e-06, "loss": 0.0102, "step": 11261 }, { "epoch": 0.7739142385926333, "grad_norm": 0.9045299291610718, "learning_rate": 3.845012217385883e-06, "loss": 0.0152, "step": 11262 }, { "epoch": 0.7739829576690489, "grad_norm": 0.3444535434246063, "learning_rate": 3.842780488959572e-06, "loss": 0.0128, "step": 11263 }, { "epoch": 0.7740516767454645, "grad_norm": 0.23464596271514893, "learning_rate": 3.840549313238984e-06, "loss": 0.0124, "step": 11264 }, { "epoch": 0.7741203958218802, "grad_norm": 0.5312763452529907, "learning_rate": 3.838318690334629e-06, "loss": 0.0064, "step": 11265 }, { "epoch": 0.7741891148982958, "grad_norm": 0.13209626078605652, "learning_rate": 3.836088620357028e-06, "loss": 0.0095, "step": 11266 }, { "epoch": 0.7742578339747114, "grad_norm": 1.2803025245666504, "learning_rate": 3.833859103416642e-06, "loss": 0.0074, "step": 11267 }, { "epoch": 0.774326553051127, "grad_norm": 0.14887481927871704, "learning_rate": 3.831630139623922e-06, "loss": 0.008, "step": 11268 }, { "epoch": 0.7743952721275426, "grad_norm": 0.4420778751373291, "learning_rate": 3.829401729089279e-06, "loss": 0.0173, "step": 11269 }, { "epoch": 0.7744639912039583, "grad_norm": 0.43077629804611206, "learning_rate": 3.8271738719231145e-06, "loss": 0.0094, "step": 11270 }, { "epoch": 0.7745327102803738, "grad_norm": 0.3013891577720642, "learning_rate": 3.824946568235777e-06, "loss": 0.0072, "step": 11271 }, { "epoch": 0.7746014293567894, "grad_norm": 0.5796229839324951, "learning_rate": 3.8227198181376195e-06, "loss": 0.0096, "step": 11272 }, { "epoch": 0.774670148433205, "grad_norm": 0.8389753103256226, "learning_rate": 3.820493621738938e-06, "loss": 0.012, "step": 11273 }, { "epoch": 0.7747388675096206, "grad_norm": 0.056614410132169724, "learning_rate": 3.818267979150024e-06, "loss": 0.0119, "step": 11274 }, { "epoch": 0.7748075865860363, "grad_norm": 0.08514644205570221, "learning_rate": 3.816042890481121e-06, "loss": 0.0057, "step": 11275 }, { "epoch": 0.7748763056624519, "grad_norm": 0.07183797657489777, "learning_rate": 3.8138183558424645e-06, "loss": 0.0085, "step": 11276 }, { "epoch": 0.7749450247388675, "grad_norm": 0.189842090010643, "learning_rate": 3.8115943753442444e-06, "loss": 0.0108, "step": 11277 }, { "epoch": 0.7750137438152831, "grad_norm": 0.6074939966201782, "learning_rate": 3.809370949096635e-06, "loss": 0.0114, "step": 11278 }, { "epoch": 0.7750824628916987, "grad_norm": 0.41413620114326477, "learning_rate": 3.8071480772097878e-06, "loss": 0.0054, "step": 11279 }, { "epoch": 0.7751511819681144, "grad_norm": 0.566831648349762, "learning_rate": 3.804925759793808e-06, "loss": 0.0078, "step": 11280 }, { "epoch": 0.77521990104453, "grad_norm": 0.35951438546180725, "learning_rate": 3.802703996958791e-06, "loss": 0.0069, "step": 11281 }, { "epoch": 0.7752886201209456, "grad_norm": 0.2425236850976944, "learning_rate": 3.800482788814801e-06, "loss": 0.0126, "step": 11282 }, { "epoch": 0.7753573391973612, "grad_norm": 0.3295031487941742, "learning_rate": 3.7982621354718656e-06, "loss": 0.006, "step": 11283 }, { "epoch": 0.7754260582737768, "grad_norm": 0.6483626365661621, "learning_rate": 3.7960420370399923e-06, "loss": 0.0113, "step": 11284 }, { "epoch": 0.7754947773501925, "grad_norm": 0.11980243027210236, "learning_rate": 3.7938224936291665e-06, "loss": 0.0106, "step": 11285 }, { "epoch": 0.775563496426608, "grad_norm": 0.673629879951477, "learning_rate": 3.791603505349332e-06, "loss": 0.0056, "step": 11286 }, { "epoch": 0.7756322155030236, "grad_norm": 0.6280516982078552, "learning_rate": 3.7893850723104208e-06, "loss": 0.0107, "step": 11287 }, { "epoch": 0.7757009345794392, "grad_norm": 0.5122708082199097, "learning_rate": 3.7871671946223195e-06, "loss": 0.0111, "step": 11288 }, { "epoch": 0.7757696536558548, "grad_norm": 0.31020182371139526, "learning_rate": 3.7849498723949046e-06, "loss": 0.0149, "step": 11289 }, { "epoch": 0.7758383727322705, "grad_norm": 0.601762592792511, "learning_rate": 3.782733105738014e-06, "loss": 0.0135, "step": 11290 }, { "epoch": 0.7759070918086861, "grad_norm": 0.5224605202674866, "learning_rate": 3.7805168947614684e-06, "loss": 0.0104, "step": 11291 }, { "epoch": 0.7759758108851017, "grad_norm": 0.10898788273334503, "learning_rate": 3.7783012395750467e-06, "loss": 0.0077, "step": 11292 }, { "epoch": 0.7760445299615173, "grad_norm": 0.40128958225250244, "learning_rate": 3.776086140288513e-06, "loss": 0.0096, "step": 11293 }, { "epoch": 0.7761132490379329, "grad_norm": 0.3292415738105774, "learning_rate": 3.773871597011594e-06, "loss": 0.0075, "step": 11294 }, { "epoch": 0.7761819681143486, "grad_norm": 0.14855337142944336, "learning_rate": 3.771657609853994e-06, "loss": 0.0065, "step": 11295 }, { "epoch": 0.7762506871907642, "grad_norm": 0.9921184778213501, "learning_rate": 3.7694441789253986e-06, "loss": 0.0076, "step": 11296 }, { "epoch": 0.7763194062671798, "grad_norm": 0.6095848679542542, "learning_rate": 3.7672313043354434e-06, "loss": 0.0114, "step": 11297 }, { "epoch": 0.7763881253435954, "grad_norm": 0.35660308599472046, "learning_rate": 3.765018986193756e-06, "loss": 0.0053, "step": 11298 }, { "epoch": 0.776456844420011, "grad_norm": 0.38685816526412964, "learning_rate": 3.762807224609934e-06, "loss": 0.0074, "step": 11299 }, { "epoch": 0.7765255634964267, "grad_norm": 0.5145121812820435, "learning_rate": 3.760596019693535e-06, "loss": 0.0096, "step": 11300 }, { "epoch": 0.7765942825728422, "grad_norm": 0.3673219084739685, "learning_rate": 3.758385371554102e-06, "loss": 0.0159, "step": 11301 }, { "epoch": 0.7766630016492578, "grad_norm": 0.7041835784912109, "learning_rate": 3.7561752803011514e-06, "loss": 0.013, "step": 11302 }, { "epoch": 0.7767317207256734, "grad_norm": 0.6777515411376953, "learning_rate": 3.7539657460441563e-06, "loss": 0.0076, "step": 11303 }, { "epoch": 0.776800439802089, "grad_norm": 0.2794390320777893, "learning_rate": 3.7517567688925814e-06, "loss": 0.0072, "step": 11304 }, { "epoch": 0.7768691588785047, "grad_norm": 0.41820546984672546, "learning_rate": 3.7495483489558457e-06, "loss": 0.0085, "step": 11305 }, { "epoch": 0.7769378779549203, "grad_norm": 0.1789134293794632, "learning_rate": 3.7473404863433603e-06, "loss": 0.0066, "step": 11306 }, { "epoch": 0.7770065970313359, "grad_norm": 0.2605903744697571, "learning_rate": 3.7451331811644823e-06, "loss": 0.0069, "step": 11307 }, { "epoch": 0.7770753161077515, "grad_norm": 2.121007204055786, "learning_rate": 3.742926433528578e-06, "loss": 0.0185, "step": 11308 }, { "epoch": 0.7771440351841671, "grad_norm": 0.44426360726356506, "learning_rate": 3.7407202435449495e-06, "loss": 0.0105, "step": 11309 }, { "epoch": 0.7772127542605828, "grad_norm": 0.393000066280365, "learning_rate": 3.738514611322896e-06, "loss": 0.0067, "step": 11310 }, { "epoch": 0.7772814733369984, "grad_norm": 0.6918566823005676, "learning_rate": 3.7363095369716726e-06, "loss": 0.0154, "step": 11311 }, { "epoch": 0.777350192413414, "grad_norm": 0.3617876470088959, "learning_rate": 3.734105020600523e-06, "loss": 0.0114, "step": 11312 }, { "epoch": 0.7774189114898296, "grad_norm": 0.42466190457344055, "learning_rate": 3.731901062318643e-06, "loss": 0.0141, "step": 11313 }, { "epoch": 0.7774876305662451, "grad_norm": 0.6948705315589905, "learning_rate": 3.729697662235217e-06, "loss": 0.0174, "step": 11314 }, { "epoch": 0.7775563496426608, "grad_norm": 0.1914249211549759, "learning_rate": 3.7274948204593995e-06, "loss": 0.0058, "step": 11315 }, { "epoch": 0.7776250687190764, "grad_norm": 0.22699199616909027, "learning_rate": 3.7252925371003184e-06, "loss": 0.0095, "step": 11316 }, { "epoch": 0.777693787795492, "grad_norm": 0.1944505125284195, "learning_rate": 3.7230908122670594e-06, "loss": 0.0066, "step": 11317 }, { "epoch": 0.7777625068719076, "grad_norm": 0.9294106960296631, "learning_rate": 3.720889646068702e-06, "loss": 0.0137, "step": 11318 }, { "epoch": 0.7778312259483232, "grad_norm": 0.4242706298828125, "learning_rate": 3.7186890386142792e-06, "loss": 0.0143, "step": 11319 }, { "epoch": 0.7778999450247389, "grad_norm": 0.8280565142631531, "learning_rate": 3.7164889900128067e-06, "loss": 0.0129, "step": 11320 }, { "epoch": 0.7779686641011545, "grad_norm": 0.5460629463195801, "learning_rate": 3.714289500373276e-06, "loss": 0.0133, "step": 11321 }, { "epoch": 0.7780373831775701, "grad_norm": 0.4789382517337799, "learning_rate": 3.7120905698046385e-06, "loss": 0.0138, "step": 11322 }, { "epoch": 0.7781061022539857, "grad_norm": 0.39038747549057007, "learning_rate": 3.70989219841583e-06, "loss": 0.0152, "step": 11323 }, { "epoch": 0.7781748213304013, "grad_norm": 0.30548638105392456, "learning_rate": 3.707694386315747e-06, "loss": 0.0081, "step": 11324 }, { "epoch": 0.778243540406817, "grad_norm": 0.7444175481796265, "learning_rate": 3.705497133613267e-06, "loss": 0.017, "step": 11325 }, { "epoch": 0.7783122594832326, "grad_norm": 0.5021594762802124, "learning_rate": 3.70330044041724e-06, "loss": 0.0096, "step": 11326 }, { "epoch": 0.7783809785596482, "grad_norm": 0.379558265209198, "learning_rate": 3.7011043068364867e-06, "loss": 0.0214, "step": 11327 }, { "epoch": 0.7784496976360638, "grad_norm": 0.4134705364704132, "learning_rate": 3.698908732979793e-06, "loss": 0.0091, "step": 11328 }, { "epoch": 0.7785184167124793, "grad_norm": 0.8213833570480347, "learning_rate": 3.696713718955932e-06, "loss": 0.0136, "step": 11329 }, { "epoch": 0.778587135788895, "grad_norm": 0.2432960420846939, "learning_rate": 3.6945192648736293e-06, "loss": 0.0073, "step": 11330 }, { "epoch": 0.7786558548653106, "grad_norm": 0.3896231949329376, "learning_rate": 3.6923253708415994e-06, "loss": 0.0156, "step": 11331 }, { "epoch": 0.7787245739417262, "grad_norm": 1.5182656049728394, "learning_rate": 3.6901320369685222e-06, "loss": 0.0085, "step": 11332 }, { "epoch": 0.7787932930181418, "grad_norm": 0.7659145593643188, "learning_rate": 3.687939263363057e-06, "loss": 0.0061, "step": 11333 }, { "epoch": 0.7788620120945574, "grad_norm": 0.28580111265182495, "learning_rate": 3.6857470501338193e-06, "loss": 0.0097, "step": 11334 }, { "epoch": 0.7789307311709731, "grad_norm": 0.6590680480003357, "learning_rate": 3.6835553973894167e-06, "loss": 0.0122, "step": 11335 }, { "epoch": 0.7789994502473887, "grad_norm": 0.7994101047515869, "learning_rate": 3.681364305238409e-06, "loss": 0.012, "step": 11336 }, { "epoch": 0.7790681693238043, "grad_norm": 0.1042463555932045, "learning_rate": 3.6791737737893444e-06, "loss": 0.0064, "step": 11337 }, { "epoch": 0.7791368884002199, "grad_norm": 0.4575138986110687, "learning_rate": 3.6769838031507396e-06, "loss": 0.0089, "step": 11338 }, { "epoch": 0.7792056074766355, "grad_norm": 0.46745726466178894, "learning_rate": 3.6747943934310744e-06, "loss": 0.0078, "step": 11339 }, { "epoch": 0.7792743265530512, "grad_norm": 0.232539564371109, "learning_rate": 3.6726055447388156e-06, "loss": 0.0083, "step": 11340 }, { "epoch": 0.7793430456294668, "grad_norm": 0.5019736289978027, "learning_rate": 3.6704172571823864e-06, "loss": 0.0103, "step": 11341 }, { "epoch": 0.7794117647058824, "grad_norm": 0.34837397933006287, "learning_rate": 3.6682295308701964e-06, "loss": 0.0134, "step": 11342 }, { "epoch": 0.779480483782298, "grad_norm": 0.4183451235294342, "learning_rate": 3.666042365910611e-06, "loss": 0.0092, "step": 11343 }, { "epoch": 0.7795492028587135, "grad_norm": 0.4469273090362549, "learning_rate": 3.663855762411993e-06, "loss": 0.0093, "step": 11344 }, { "epoch": 0.7796179219351292, "grad_norm": 0.9722895622253418, "learning_rate": 3.6616697204826494e-06, "loss": 0.0174, "step": 11345 }, { "epoch": 0.7796866410115448, "grad_norm": 0.3115934729576111, "learning_rate": 3.6594842402308837e-06, "loss": 0.008, "step": 11346 }, { "epoch": 0.7797553600879604, "grad_norm": 0.349656343460083, "learning_rate": 3.6572993217649475e-06, "loss": 0.0086, "step": 11347 }, { "epoch": 0.779824079164376, "grad_norm": 0.377773255109787, "learning_rate": 3.6551149651930874e-06, "loss": 0.0121, "step": 11348 }, { "epoch": 0.7798927982407916, "grad_norm": 0.6039007902145386, "learning_rate": 3.652931170623501e-06, "loss": 0.0181, "step": 11349 }, { "epoch": 0.7799615173172073, "grad_norm": 0.2721669673919678, "learning_rate": 3.6507479381643838e-06, "loss": 0.0074, "step": 11350 }, { "epoch": 0.7800302363936229, "grad_norm": 0.4958149194717407, "learning_rate": 3.648565267923875e-06, "loss": 0.0121, "step": 11351 }, { "epoch": 0.7800989554700385, "grad_norm": 0.4171113669872284, "learning_rate": 3.6463831600101117e-06, "loss": 0.0139, "step": 11352 }, { "epoch": 0.7801676745464541, "grad_norm": 0.2625882625579834, "learning_rate": 3.6442016145311785e-06, "loss": 0.0074, "step": 11353 }, { "epoch": 0.7802363936228697, "grad_norm": 0.5440624952316284, "learning_rate": 3.642020631595156e-06, "loss": 0.0135, "step": 11354 }, { "epoch": 0.7803051126992854, "grad_norm": 0.8725361824035645, "learning_rate": 3.639840211310076e-06, "loss": 0.0127, "step": 11355 }, { "epoch": 0.780373831775701, "grad_norm": 0.2879338264465332, "learning_rate": 3.6376603537839552e-06, "loss": 0.0115, "step": 11356 }, { "epoch": 0.7804425508521166, "grad_norm": 0.14815175533294678, "learning_rate": 3.635481059124786e-06, "loss": 0.0105, "step": 11357 }, { "epoch": 0.7805112699285321, "grad_norm": 0.5138179659843445, "learning_rate": 3.633302327440517e-06, "loss": 0.0079, "step": 11358 }, { "epoch": 0.7805799890049477, "grad_norm": 0.4011783301830292, "learning_rate": 3.631124158839085e-06, "loss": 0.012, "step": 11359 }, { "epoch": 0.7806487080813634, "grad_norm": 0.7975739240646362, "learning_rate": 3.628946553428385e-06, "loss": 0.0078, "step": 11360 }, { "epoch": 0.780717427157779, "grad_norm": 0.08837856352329254, "learning_rate": 3.6267695113162946e-06, "loss": 0.0033, "step": 11361 }, { "epoch": 0.7807861462341946, "grad_norm": 0.4481869637966156, "learning_rate": 3.624593032610661e-06, "loss": 0.0124, "step": 11362 }, { "epoch": 0.7808548653106102, "grad_norm": 0.4152451157569885, "learning_rate": 3.6224171174193037e-06, "loss": 0.0077, "step": 11363 }, { "epoch": 0.7809235843870258, "grad_norm": 0.4209355413913727, "learning_rate": 3.6202417658500093e-06, "loss": 0.0118, "step": 11364 }, { "epoch": 0.7809923034634415, "grad_norm": 0.6820225715637207, "learning_rate": 3.6180669780105456e-06, "loss": 0.0113, "step": 11365 }, { "epoch": 0.7810610225398571, "grad_norm": 0.756874144077301, "learning_rate": 3.615892754008639e-06, "loss": 0.0074, "step": 11366 }, { "epoch": 0.7811297416162727, "grad_norm": 0.4371494650840759, "learning_rate": 3.6137190939519993e-06, "loss": 0.0113, "step": 11367 }, { "epoch": 0.7811984606926883, "grad_norm": 0.2895753085613251, "learning_rate": 3.611545997948309e-06, "loss": 0.007, "step": 11368 }, { "epoch": 0.7812671797691039, "grad_norm": 0.27531617879867554, "learning_rate": 3.6093734661052187e-06, "loss": 0.0146, "step": 11369 }, { "epoch": 0.7813358988455195, "grad_norm": 0.49065980315208435, "learning_rate": 3.607201498530345e-06, "loss": 0.0168, "step": 11370 }, { "epoch": 0.7814046179219352, "grad_norm": 0.1910727471113205, "learning_rate": 3.6050300953312913e-06, "loss": 0.0066, "step": 11371 }, { "epoch": 0.7814733369983508, "grad_norm": 0.5387780666351318, "learning_rate": 3.6028592566156145e-06, "loss": 0.0127, "step": 11372 }, { "epoch": 0.7815420560747663, "grad_norm": 0.3286267817020416, "learning_rate": 3.6006889824908596e-06, "loss": 0.0063, "step": 11373 }, { "epoch": 0.7816107751511819, "grad_norm": 0.8504583835601807, "learning_rate": 3.5985192730645395e-06, "loss": 0.009, "step": 11374 }, { "epoch": 0.7816794942275975, "grad_norm": 0.5436093807220459, "learning_rate": 3.5963501284441315e-06, "loss": 0.0172, "step": 11375 }, { "epoch": 0.7817482133040132, "grad_norm": 0.3065655827522278, "learning_rate": 3.594181548737096e-06, "loss": 0.0106, "step": 11376 }, { "epoch": 0.7818169323804288, "grad_norm": 0.28336620330810547, "learning_rate": 3.592013534050854e-06, "loss": 0.01, "step": 11377 }, { "epoch": 0.7818856514568444, "grad_norm": 0.3512483239173889, "learning_rate": 3.5898460844928116e-06, "loss": 0.0113, "step": 11378 }, { "epoch": 0.78195437053326, "grad_norm": 0.4248833656311035, "learning_rate": 3.5876792001703272e-06, "loss": 0.0118, "step": 11379 }, { "epoch": 0.7820230896096756, "grad_norm": 1.5738155841827393, "learning_rate": 3.5855128811907606e-06, "loss": 0.0211, "step": 11380 }, { "epoch": 0.7820918086860913, "grad_norm": 0.33082401752471924, "learning_rate": 3.5833471276614154e-06, "loss": 0.0089, "step": 11381 }, { "epoch": 0.7821605277625069, "grad_norm": 0.5152547955513, "learning_rate": 3.5811819396895864e-06, "loss": 0.0136, "step": 11382 }, { "epoch": 0.7822292468389225, "grad_norm": 0.21163754165172577, "learning_rate": 3.5790173173825218e-06, "loss": 0.0076, "step": 11383 }, { "epoch": 0.7822979659153381, "grad_norm": 0.19990170001983643, "learning_rate": 3.5768532608474652e-06, "loss": 0.0086, "step": 11384 }, { "epoch": 0.7823666849917537, "grad_norm": 0.5367249250411987, "learning_rate": 3.5746897701916037e-06, "loss": 0.0081, "step": 11385 }, { "epoch": 0.7824354040681694, "grad_norm": 0.19552533328533173, "learning_rate": 3.57252684552213e-06, "loss": 0.0159, "step": 11386 }, { "epoch": 0.782504123144585, "grad_norm": 0.14031782746315002, "learning_rate": 3.5703644869461767e-06, "loss": 0.0088, "step": 11387 }, { "epoch": 0.7825728422210005, "grad_norm": 0.8110173344612122, "learning_rate": 3.5682026945708746e-06, "loss": 0.0105, "step": 11388 }, { "epoch": 0.7826415612974161, "grad_norm": 0.5321976542472839, "learning_rate": 3.566041468503303e-06, "loss": 0.0125, "step": 11389 }, { "epoch": 0.7827102803738317, "grad_norm": 1.6899995803833008, "learning_rate": 3.563880808850535e-06, "loss": 0.0095, "step": 11390 }, { "epoch": 0.7827789994502474, "grad_norm": 0.4008420705795288, "learning_rate": 3.561720715719595e-06, "loss": 0.0084, "step": 11391 }, { "epoch": 0.782847718526663, "grad_norm": 0.3342422842979431, "learning_rate": 3.5595611892174955e-06, "loss": 0.0082, "step": 11392 }, { "epoch": 0.7829164376030786, "grad_norm": 0.8694480657577515, "learning_rate": 3.557402229451218e-06, "loss": 0.0137, "step": 11393 }, { "epoch": 0.7829851566794942, "grad_norm": 0.2967895269393921, "learning_rate": 3.555243836527704e-06, "loss": 0.0114, "step": 11394 }, { "epoch": 0.7830538757559098, "grad_norm": 0.3347141742706299, "learning_rate": 3.553086010553886e-06, "loss": 0.0094, "step": 11395 }, { "epoch": 0.7831225948323255, "grad_norm": 0.22782564163208008, "learning_rate": 3.5509287516366494e-06, "loss": 0.0101, "step": 11396 }, { "epoch": 0.7831913139087411, "grad_norm": 0.1683044582605362, "learning_rate": 3.5487720598828626e-06, "loss": 0.0057, "step": 11397 }, { "epoch": 0.7832600329851567, "grad_norm": 0.49002522230148315, "learning_rate": 3.5466159353993667e-06, "loss": 0.009, "step": 11398 }, { "epoch": 0.7833287520615723, "grad_norm": 0.1561795026063919, "learning_rate": 3.544460378292974e-06, "loss": 0.0111, "step": 11399 }, { "epoch": 0.7833974711379879, "grad_norm": 0.41977953910827637, "learning_rate": 3.5423053886704584e-06, "loss": 0.0091, "step": 11400 }, { "epoch": 0.7834661902144036, "grad_norm": 0.3710181415081024, "learning_rate": 3.5401509666385817e-06, "loss": 0.0072, "step": 11401 }, { "epoch": 0.7835349092908191, "grad_norm": 0.4154720604419708, "learning_rate": 3.537997112304061e-06, "loss": 0.0097, "step": 11402 }, { "epoch": 0.7836036283672347, "grad_norm": 0.5037109851837158, "learning_rate": 3.535843825773599e-06, "loss": 0.0101, "step": 11403 }, { "epoch": 0.7836723474436503, "grad_norm": 0.6919224262237549, "learning_rate": 3.533691107153866e-06, "loss": 0.0117, "step": 11404 }, { "epoch": 0.7837410665200659, "grad_norm": 0.29358336329460144, "learning_rate": 3.531538956551504e-06, "loss": 0.0082, "step": 11405 }, { "epoch": 0.7838097855964816, "grad_norm": 0.40310409665107727, "learning_rate": 3.52938737407312e-06, "loss": 0.0126, "step": 11406 }, { "epoch": 0.7838785046728972, "grad_norm": 0.39806684851646423, "learning_rate": 3.5272363598253084e-06, "loss": 0.0107, "step": 11407 }, { "epoch": 0.7839472237493128, "grad_norm": 0.15379327535629272, "learning_rate": 3.5250859139146156e-06, "loss": 0.0152, "step": 11408 }, { "epoch": 0.7840159428257284, "grad_norm": 0.3308407664299011, "learning_rate": 3.522936036447574e-06, "loss": 0.0069, "step": 11409 }, { "epoch": 0.784084661902144, "grad_norm": 0.5024449825286865, "learning_rate": 3.5207867275306903e-06, "loss": 0.013, "step": 11410 }, { "epoch": 0.7841533809785597, "grad_norm": 0.9837303757667542, "learning_rate": 3.5186379872704287e-06, "loss": 0.0076, "step": 11411 }, { "epoch": 0.7842221000549753, "grad_norm": 0.6858192682266235, "learning_rate": 3.5164898157732404e-06, "loss": 0.0179, "step": 11412 }, { "epoch": 0.7842908191313909, "grad_norm": 0.4112613797187805, "learning_rate": 3.5143422131455322e-06, "loss": 0.011, "step": 11413 }, { "epoch": 0.7843595382078065, "grad_norm": 0.45275336503982544, "learning_rate": 3.512195179493702e-06, "loss": 0.0093, "step": 11414 }, { "epoch": 0.784428257284222, "grad_norm": 0.2149670273065567, "learning_rate": 3.510048714924096e-06, "loss": 0.0064, "step": 11415 }, { "epoch": 0.7844969763606378, "grad_norm": 0.34168046712875366, "learning_rate": 3.5079028195430634e-06, "loss": 0.009, "step": 11416 }, { "epoch": 0.7845656954370533, "grad_norm": 0.39293017983436584, "learning_rate": 3.5057574934568936e-06, "loss": 0.0081, "step": 11417 }, { "epoch": 0.7846344145134689, "grad_norm": 0.6064407229423523, "learning_rate": 3.5036127367718717e-06, "loss": 0.0117, "step": 11418 }, { "epoch": 0.7847031335898845, "grad_norm": 0.3811741769313812, "learning_rate": 3.5014685495942343e-06, "loss": 0.0073, "step": 11419 }, { "epoch": 0.7847718526663001, "grad_norm": 0.4862760603427887, "learning_rate": 3.49932493203021e-06, "loss": 0.011, "step": 11420 }, { "epoch": 0.7848405717427158, "grad_norm": 0.1376698911190033, "learning_rate": 3.497181884185979e-06, "loss": 0.0044, "step": 11421 }, { "epoch": 0.7849092908191314, "grad_norm": 0.2705504894256592, "learning_rate": 3.4950394061677087e-06, "loss": 0.0112, "step": 11422 }, { "epoch": 0.784978009895547, "grad_norm": 0.5871448516845703, "learning_rate": 3.492897498081532e-06, "loss": 0.0092, "step": 11423 }, { "epoch": 0.7850467289719626, "grad_norm": 0.12175253033638, "learning_rate": 3.490756160033562e-06, "loss": 0.0063, "step": 11424 }, { "epoch": 0.7851154480483782, "grad_norm": 0.1398317664861679, "learning_rate": 3.4886153921298636e-06, "loss": 0.0065, "step": 11425 }, { "epoch": 0.7851841671247939, "grad_norm": 0.3701801896095276, "learning_rate": 3.486475194476499e-06, "loss": 0.0126, "step": 11426 }, { "epoch": 0.7852528862012095, "grad_norm": 0.4204169511795044, "learning_rate": 3.4843355671794763e-06, "loss": 0.0069, "step": 11427 }, { "epoch": 0.7853216052776251, "grad_norm": 0.30290210247039795, "learning_rate": 3.4821965103447946e-06, "loss": 0.0072, "step": 11428 }, { "epoch": 0.7853903243540407, "grad_norm": 0.3434712588787079, "learning_rate": 3.480058024078422e-06, "loss": 0.0109, "step": 11429 }, { "epoch": 0.7854590434304562, "grad_norm": 0.2927514612674713, "learning_rate": 3.477920108486287e-06, "loss": 0.007, "step": 11430 }, { "epoch": 0.785527762506872, "grad_norm": 0.18011474609375, "learning_rate": 3.4757827636743073e-06, "loss": 0.016, "step": 11431 }, { "epoch": 0.7855964815832875, "grad_norm": 0.15249624848365784, "learning_rate": 3.473645989748351e-06, "loss": 0.004, "step": 11432 }, { "epoch": 0.7856652006597031, "grad_norm": 0.20502690970897675, "learning_rate": 3.471509786814278e-06, "loss": 0.0069, "step": 11433 }, { "epoch": 0.7857339197361187, "grad_norm": 0.2383318394422531, "learning_rate": 3.4693741549779027e-06, "loss": 0.0087, "step": 11434 }, { "epoch": 0.7858026388125343, "grad_norm": 0.38940203189849854, "learning_rate": 3.467239094345034e-06, "loss": 0.0115, "step": 11435 }, { "epoch": 0.78587135788895, "grad_norm": 0.30362120270729065, "learning_rate": 3.4651046050214246e-06, "loss": 0.0105, "step": 11436 }, { "epoch": 0.7859400769653656, "grad_norm": 0.47081372141838074, "learning_rate": 3.4629706871128225e-06, "loss": 0.0086, "step": 11437 }, { "epoch": 0.7860087960417812, "grad_norm": 0.5742455720901489, "learning_rate": 3.4608373407249292e-06, "loss": 0.009, "step": 11438 }, { "epoch": 0.7860775151181968, "grad_norm": 0.5120972394943237, "learning_rate": 3.458704565963436e-06, "loss": 0.0069, "step": 11439 }, { "epoch": 0.7861462341946124, "grad_norm": 0.4974120259284973, "learning_rate": 3.4565723629339826e-06, "loss": 0.0102, "step": 11440 }, { "epoch": 0.7862149532710281, "grad_norm": 0.3046230673789978, "learning_rate": 3.4544407317422096e-06, "loss": 0.0086, "step": 11441 }, { "epoch": 0.7862836723474437, "grad_norm": 0.14551520347595215, "learning_rate": 3.452309672493701e-06, "loss": 0.0102, "step": 11442 }, { "epoch": 0.7863523914238593, "grad_norm": 0.2446744441986084, "learning_rate": 3.450179185294036e-06, "loss": 0.0128, "step": 11443 }, { "epoch": 0.7864211105002749, "grad_norm": 0.3109356164932251, "learning_rate": 3.4480492702487427e-06, "loss": 0.0083, "step": 11444 }, { "epoch": 0.7864898295766904, "grad_norm": 0.9065811038017273, "learning_rate": 3.4459199274633443e-06, "loss": 0.0116, "step": 11445 }, { "epoch": 0.7865585486531061, "grad_norm": 3.000899314880371, "learning_rate": 3.443791157043314e-06, "loss": 0.019, "step": 11446 }, { "epoch": 0.7866272677295217, "grad_norm": 0.45995059609413147, "learning_rate": 3.4416629590941113e-06, "loss": 0.0118, "step": 11447 }, { "epoch": 0.7866959868059373, "grad_norm": 0.4294390082359314, "learning_rate": 3.4395353337211664e-06, "loss": 0.0179, "step": 11448 }, { "epoch": 0.7867647058823529, "grad_norm": 0.3732753396034241, "learning_rate": 3.437408281029871e-06, "loss": 0.0066, "step": 11449 }, { "epoch": 0.7868334249587685, "grad_norm": 0.8402151465415955, "learning_rate": 3.435281801125601e-06, "loss": 0.0126, "step": 11450 }, { "epoch": 0.7869021440351842, "grad_norm": 0.27674123644828796, "learning_rate": 3.43315589411369e-06, "loss": 0.0141, "step": 11451 }, { "epoch": 0.7869708631115998, "grad_norm": 0.2657843828201294, "learning_rate": 3.4310305600994562e-06, "loss": 0.0064, "step": 11452 }, { "epoch": 0.7870395821880154, "grad_norm": 0.35628166794776917, "learning_rate": 3.4289057991881826e-06, "loss": 0.0115, "step": 11453 }, { "epoch": 0.787108301264431, "grad_norm": 0.5244269371032715, "learning_rate": 3.426781611485132e-06, "loss": 0.0158, "step": 11454 }, { "epoch": 0.7871770203408466, "grad_norm": 0.07253473997116089, "learning_rate": 3.424657997095521e-06, "loss": 0.0067, "step": 11455 }, { "epoch": 0.7872457394172623, "grad_norm": 0.38480496406555176, "learning_rate": 3.4225349561245613e-06, "loss": 0.0092, "step": 11456 }, { "epoch": 0.7873144584936779, "grad_norm": 0.8544758558273315, "learning_rate": 3.420412488677411e-06, "loss": 0.0146, "step": 11457 }, { "epoch": 0.7873831775700935, "grad_norm": 0.6421570777893066, "learning_rate": 3.418290594859221e-06, "loss": 0.0177, "step": 11458 }, { "epoch": 0.787451896646509, "grad_norm": 0.43571537733078003, "learning_rate": 3.416169274775104e-06, "loss": 0.0103, "step": 11459 }, { "epoch": 0.7875206157229246, "grad_norm": 0.3755229711532593, "learning_rate": 3.41404852853015e-06, "loss": 0.0087, "step": 11460 }, { "epoch": 0.7875893347993403, "grad_norm": 0.4376313388347626, "learning_rate": 3.4119283562294064e-06, "loss": 0.0088, "step": 11461 }, { "epoch": 0.7876580538757559, "grad_norm": 0.6706584095954895, "learning_rate": 3.409808757977914e-06, "loss": 0.0118, "step": 11462 }, { "epoch": 0.7877267729521715, "grad_norm": 0.48867255449295044, "learning_rate": 3.407689733880663e-06, "loss": 0.0138, "step": 11463 }, { "epoch": 0.7877954920285871, "grad_norm": 0.4641880393028259, "learning_rate": 3.4055712840426283e-06, "loss": 0.016, "step": 11464 }, { "epoch": 0.7878642111050027, "grad_norm": 0.43025732040405273, "learning_rate": 3.4034534085687603e-06, "loss": 0.0155, "step": 11465 }, { "epoch": 0.7879329301814184, "grad_norm": 0.6739760041236877, "learning_rate": 3.4013361075639644e-06, "loss": 0.019, "step": 11466 }, { "epoch": 0.788001649257834, "grad_norm": 0.11260046064853668, "learning_rate": 3.399219381133137e-06, "loss": 0.0105, "step": 11467 }, { "epoch": 0.7880703683342496, "grad_norm": 0.21144525706768036, "learning_rate": 3.3971032293811257e-06, "loss": 0.0036, "step": 11468 }, { "epoch": 0.7881390874106652, "grad_norm": 0.31635528802871704, "learning_rate": 3.3949876524127716e-06, "loss": 0.0073, "step": 11469 }, { "epoch": 0.7882078064870808, "grad_norm": 0.5171484351158142, "learning_rate": 3.3928726503328592e-06, "loss": 0.0053, "step": 11470 }, { "epoch": 0.7882765255634965, "grad_norm": 0.7699390053749084, "learning_rate": 3.3907582232461837e-06, "loss": 0.0133, "step": 11471 }, { "epoch": 0.7883452446399121, "grad_norm": 0.33771559596061707, "learning_rate": 3.388644371257474e-06, "loss": 0.0189, "step": 11472 }, { "epoch": 0.7884139637163277, "grad_norm": 0.5481458306312561, "learning_rate": 3.3865310944714526e-06, "loss": 0.0128, "step": 11473 }, { "epoch": 0.7884826827927433, "grad_norm": 0.37168067693710327, "learning_rate": 3.3844183929928018e-06, "loss": 0.0081, "step": 11474 }, { "epoch": 0.7885514018691588, "grad_norm": 0.5375357866287231, "learning_rate": 3.3823062669261855e-06, "loss": 0.0065, "step": 11475 }, { "epoch": 0.7886201209455745, "grad_norm": 0.3873007595539093, "learning_rate": 3.3801947163762255e-06, "loss": 0.0171, "step": 11476 }, { "epoch": 0.7886888400219901, "grad_norm": 0.39584049582481384, "learning_rate": 3.3780837414475376e-06, "loss": 0.0162, "step": 11477 }, { "epoch": 0.7887575590984057, "grad_norm": 0.6305251121520996, "learning_rate": 3.3759733422446833e-06, "loss": 0.0169, "step": 11478 }, { "epoch": 0.7888262781748213, "grad_norm": 0.27540117502212524, "learning_rate": 3.3738635188722157e-06, "loss": 0.0107, "step": 11479 }, { "epoch": 0.7888949972512369, "grad_norm": 0.363434761762619, "learning_rate": 3.371754271434643e-06, "loss": 0.0081, "step": 11480 }, { "epoch": 0.7889637163276526, "grad_norm": 0.35460638999938965, "learning_rate": 3.369645600036461e-06, "loss": 0.0069, "step": 11481 }, { "epoch": 0.7890324354040682, "grad_norm": 0.8889690041542053, "learning_rate": 3.3675375047821206e-06, "loss": 0.0101, "step": 11482 }, { "epoch": 0.7891011544804838, "grad_norm": 0.2961060404777527, "learning_rate": 3.3654299857760573e-06, "loss": 0.0048, "step": 11483 }, { "epoch": 0.7891698735568994, "grad_norm": 0.3723559081554413, "learning_rate": 3.3633230431226765e-06, "loss": 0.0048, "step": 11484 }, { "epoch": 0.789238592633315, "grad_norm": 0.18513032793998718, "learning_rate": 3.361216676926344e-06, "loss": 0.01, "step": 11485 }, { "epoch": 0.7893073117097307, "grad_norm": 0.35119563341140747, "learning_rate": 3.359110887291414e-06, "loss": 0.0123, "step": 11486 }, { "epoch": 0.7893760307861463, "grad_norm": 0.22990025579929352, "learning_rate": 3.357005674322194e-06, "loss": 0.0109, "step": 11487 }, { "epoch": 0.7894447498625619, "grad_norm": 0.4969092011451721, "learning_rate": 3.354901038122974e-06, "loss": 0.0095, "step": 11488 }, { "epoch": 0.7895134689389774, "grad_norm": 0.36710125207901, "learning_rate": 3.3527969787980174e-06, "loss": 0.0072, "step": 11489 }, { "epoch": 0.789582188015393, "grad_norm": 0.3436810374259949, "learning_rate": 3.3506934964515555e-06, "loss": 0.0067, "step": 11490 }, { "epoch": 0.7896509070918087, "grad_norm": 0.18279540538787842, "learning_rate": 3.348590591187784e-06, "loss": 0.0039, "step": 11491 }, { "epoch": 0.7897196261682243, "grad_norm": 0.3893723487854004, "learning_rate": 3.3464882631108855e-06, "loss": 0.0057, "step": 11492 }, { "epoch": 0.7897883452446399, "grad_norm": 0.9475578665733337, "learning_rate": 3.3443865123249937e-06, "loss": 0.0172, "step": 11493 }, { "epoch": 0.7898570643210555, "grad_norm": 0.30237337946891785, "learning_rate": 3.342285338934234e-06, "loss": 0.0143, "step": 11494 }, { "epoch": 0.7899257833974711, "grad_norm": 0.39031991362571716, "learning_rate": 3.3401847430426892e-06, "loss": 0.0355, "step": 11495 }, { "epoch": 0.7899945024738868, "grad_norm": 0.32098326086997986, "learning_rate": 3.338084724754426e-06, "loss": 0.0139, "step": 11496 }, { "epoch": 0.7900632215503024, "grad_norm": 0.3739348351955414, "learning_rate": 3.3359852841734633e-06, "loss": 0.0096, "step": 11497 }, { "epoch": 0.790131940626718, "grad_norm": 1.718303918838501, "learning_rate": 3.3338864214038163e-06, "loss": 0.019, "step": 11498 }, { "epoch": 0.7902006597031336, "grad_norm": 0.48792359232902527, "learning_rate": 3.3317881365494463e-06, "loss": 0.0113, "step": 11499 }, { "epoch": 0.7902693787795492, "grad_norm": 0.36092498898506165, "learning_rate": 3.329690429714302e-06, "loss": 0.0128, "step": 11500 }, { "epoch": 0.7903380978559648, "grad_norm": 0.3059792220592499, "learning_rate": 3.327593301002306e-06, "loss": 0.0109, "step": 11501 }, { "epoch": 0.7904068169323805, "grad_norm": 0.7213882207870483, "learning_rate": 3.3254967505173373e-06, "loss": 0.013, "step": 11502 }, { "epoch": 0.790475536008796, "grad_norm": 0.7892105579376221, "learning_rate": 3.32340077836326e-06, "loss": 0.0108, "step": 11503 }, { "epoch": 0.7905442550852116, "grad_norm": 0.11254411190748215, "learning_rate": 3.3213053846438983e-06, "loss": 0.0072, "step": 11504 }, { "epoch": 0.7906129741616272, "grad_norm": 0.10303350538015366, "learning_rate": 3.3192105694630585e-06, "loss": 0.0072, "step": 11505 }, { "epoch": 0.7906816932380428, "grad_norm": 0.28061628341674805, "learning_rate": 3.3171163329245114e-06, "loss": 0.008, "step": 11506 }, { "epoch": 0.7907504123144585, "grad_norm": 0.1567605435848236, "learning_rate": 3.315022675132007e-06, "loss": 0.0093, "step": 11507 }, { "epoch": 0.7908191313908741, "grad_norm": 0.29601433873176575, "learning_rate": 3.312929596189252e-06, "loss": 0.012, "step": 11508 }, { "epoch": 0.7908878504672897, "grad_norm": 0.25473377108573914, "learning_rate": 3.310837096199941e-06, "loss": 0.0088, "step": 11509 }, { "epoch": 0.7909565695437053, "grad_norm": 0.35589468479156494, "learning_rate": 3.3087451752677232e-06, "loss": 0.0099, "step": 11510 }, { "epoch": 0.7910252886201209, "grad_norm": 0.24193675816059113, "learning_rate": 3.3066538334962387e-06, "loss": 0.0082, "step": 11511 }, { "epoch": 0.7910940076965366, "grad_norm": 0.3585406243801117, "learning_rate": 3.3045630709890738e-06, "loss": 0.0093, "step": 11512 }, { "epoch": 0.7911627267729522, "grad_norm": 0.32360199093818665, "learning_rate": 3.3024728878498194e-06, "loss": 0.0137, "step": 11513 }, { "epoch": 0.7912314458493678, "grad_norm": 0.3336751461029053, "learning_rate": 3.3003832841820025e-06, "loss": 0.0109, "step": 11514 }, { "epoch": 0.7913001649257834, "grad_norm": 0.17867399752140045, "learning_rate": 3.2982942600891504e-06, "loss": 0.0102, "step": 11515 }, { "epoch": 0.791368884002199, "grad_norm": 0.20059111714363098, "learning_rate": 3.2962058156747383e-06, "loss": 0.0134, "step": 11516 }, { "epoch": 0.7914376030786147, "grad_norm": 0.5531704425811768, "learning_rate": 3.294117951042232e-06, "loss": 0.0115, "step": 11517 }, { "epoch": 0.7915063221550303, "grad_norm": 0.24744650721549988, "learning_rate": 3.2920306662950504e-06, "loss": 0.0057, "step": 11518 }, { "epoch": 0.7915750412314458, "grad_norm": 0.43894973397254944, "learning_rate": 3.289943961536601e-06, "loss": 0.0091, "step": 11519 }, { "epoch": 0.7916437603078614, "grad_norm": 0.41330698132514954, "learning_rate": 3.287857836870253e-06, "loss": 0.0064, "step": 11520 }, { "epoch": 0.791712479384277, "grad_norm": 0.23516862094402313, "learning_rate": 3.2857722923993526e-06, "loss": 0.0147, "step": 11521 }, { "epoch": 0.7917811984606927, "grad_norm": 0.20009461045265198, "learning_rate": 3.283687328227205e-06, "loss": 0.0085, "step": 11522 }, { "epoch": 0.7918499175371083, "grad_norm": 0.473018616437912, "learning_rate": 3.281602944457104e-06, "loss": 0.008, "step": 11523 }, { "epoch": 0.7919186366135239, "grad_norm": 0.3085099756717682, "learning_rate": 3.2795191411922947e-06, "loss": 0.0119, "step": 11524 }, { "epoch": 0.7919873556899395, "grad_norm": 0.5708168745040894, "learning_rate": 3.2774359185360137e-06, "loss": 0.0074, "step": 11525 }, { "epoch": 0.7920560747663551, "grad_norm": 0.401213139295578, "learning_rate": 3.2753532765914597e-06, "loss": 0.0122, "step": 11526 }, { "epoch": 0.7921247938427708, "grad_norm": 0.42825984954833984, "learning_rate": 3.273271215461795e-06, "loss": 0.0088, "step": 11527 }, { "epoch": 0.7921935129191864, "grad_norm": 0.3502293825149536, "learning_rate": 3.2711897352501706e-06, "loss": 0.0093, "step": 11528 }, { "epoch": 0.792262231995602, "grad_norm": 0.33711522817611694, "learning_rate": 3.2691088360596875e-06, "loss": 0.0179, "step": 11529 }, { "epoch": 0.7923309510720176, "grad_norm": 0.2576650381088257, "learning_rate": 3.267028517993437e-06, "loss": 0.0075, "step": 11530 }, { "epoch": 0.7923996701484332, "grad_norm": 0.4914400577545166, "learning_rate": 3.2649487811544705e-06, "loss": 0.0118, "step": 11531 }, { "epoch": 0.7924683892248489, "grad_norm": 0.6772419810295105, "learning_rate": 3.262869625645818e-06, "loss": 0.0075, "step": 11532 }, { "epoch": 0.7925371083012644, "grad_norm": 0.14387857913970947, "learning_rate": 3.260791051570472e-06, "loss": 0.0055, "step": 11533 }, { "epoch": 0.79260582737768, "grad_norm": 0.43251001834869385, "learning_rate": 3.2587130590314053e-06, "loss": 0.0069, "step": 11534 }, { "epoch": 0.7926745464540956, "grad_norm": 0.5375955700874329, "learning_rate": 3.25663564813155e-06, "loss": 0.0099, "step": 11535 }, { "epoch": 0.7927432655305112, "grad_norm": 0.40243035554885864, "learning_rate": 3.254558818973823e-06, "loss": 0.0103, "step": 11536 }, { "epoch": 0.7928119846069269, "grad_norm": 0.23367264866828918, "learning_rate": 3.252482571661103e-06, "loss": 0.0053, "step": 11537 }, { "epoch": 0.7928807036833425, "grad_norm": 0.23036690056324005, "learning_rate": 3.2504069062962495e-06, "loss": 0.0098, "step": 11538 }, { "epoch": 0.7929494227597581, "grad_norm": 0.4630607068538666, "learning_rate": 3.248331822982077e-06, "loss": 0.01, "step": 11539 }, { "epoch": 0.7930181418361737, "grad_norm": 0.33812105655670166, "learning_rate": 3.2462573218213897e-06, "loss": 0.0079, "step": 11540 }, { "epoch": 0.7930868609125893, "grad_norm": 0.6152692437171936, "learning_rate": 3.2441834029169474e-06, "loss": 0.0087, "step": 11541 }, { "epoch": 0.793155579989005, "grad_norm": 0.8189640045166016, "learning_rate": 3.242110066371488e-06, "loss": 0.0201, "step": 11542 }, { "epoch": 0.7932242990654206, "grad_norm": 0.6325665712356567, "learning_rate": 3.240037312287729e-06, "loss": 0.012, "step": 11543 }, { "epoch": 0.7932930181418362, "grad_norm": 0.40888896584510803, "learning_rate": 3.2379651407683384e-06, "loss": 0.0127, "step": 11544 }, { "epoch": 0.7933617372182518, "grad_norm": 0.3137046992778778, "learning_rate": 3.235893551915978e-06, "loss": 0.0126, "step": 11545 }, { "epoch": 0.7934304562946674, "grad_norm": 0.5381375551223755, "learning_rate": 3.233822545833261e-06, "loss": 0.0131, "step": 11546 }, { "epoch": 0.793499175371083, "grad_norm": 0.15454316139221191, "learning_rate": 3.231752122622788e-06, "loss": 0.0082, "step": 11547 }, { "epoch": 0.7935678944474986, "grad_norm": 0.36972880363464355, "learning_rate": 3.2296822823871136e-06, "loss": 0.0096, "step": 11548 }, { "epoch": 0.7936366135239142, "grad_norm": 0.1848204880952835, "learning_rate": 3.2276130252287874e-06, "loss": 0.005, "step": 11549 }, { "epoch": 0.7937053326003298, "grad_norm": 0.35665199160575867, "learning_rate": 3.2255443512503054e-06, "loss": 0.0133, "step": 11550 }, { "epoch": 0.7937740516767454, "grad_norm": 0.2893107831478119, "learning_rate": 3.2234762605541522e-06, "loss": 0.0069, "step": 11551 }, { "epoch": 0.7938427707531611, "grad_norm": 1.0672537088394165, "learning_rate": 3.22140875324277e-06, "loss": 0.0063, "step": 11552 }, { "epoch": 0.7939114898295767, "grad_norm": 0.24598579108715057, "learning_rate": 3.219341829418586e-06, "loss": 0.0072, "step": 11553 }, { "epoch": 0.7939802089059923, "grad_norm": 0.46961742639541626, "learning_rate": 3.2172754891839807e-06, "loss": 0.0148, "step": 11554 }, { "epoch": 0.7940489279824079, "grad_norm": 0.4091308116912842, "learning_rate": 3.21520973264133e-06, "loss": 0.0118, "step": 11555 }, { "epoch": 0.7941176470588235, "grad_norm": 0.1798621118068695, "learning_rate": 3.213144559892959e-06, "loss": 0.0075, "step": 11556 }, { "epoch": 0.7941863661352392, "grad_norm": 0.6447299718856812, "learning_rate": 3.2110799710411755e-06, "loss": 0.0105, "step": 11557 }, { "epoch": 0.7942550852116548, "grad_norm": 0.45354902744293213, "learning_rate": 3.2090159661882495e-06, "loss": 0.0096, "step": 11558 }, { "epoch": 0.7943238042880704, "grad_norm": 0.23335957527160645, "learning_rate": 3.2069525454364363e-06, "loss": 0.0101, "step": 11559 }, { "epoch": 0.794392523364486, "grad_norm": 0.5422605872154236, "learning_rate": 3.2048897088879437e-06, "loss": 0.0114, "step": 11560 }, { "epoch": 0.7944612424409015, "grad_norm": 0.4352622628211975, "learning_rate": 3.202827456644965e-06, "loss": 0.0114, "step": 11561 }, { "epoch": 0.7945299615173173, "grad_norm": 0.22157683968544006, "learning_rate": 3.200765788809665e-06, "loss": 0.0091, "step": 11562 }, { "epoch": 0.7945986805937328, "grad_norm": 0.3858073055744171, "learning_rate": 3.198704705484166e-06, "loss": 0.0129, "step": 11563 }, { "epoch": 0.7946673996701484, "grad_norm": 0.2994324862957001, "learning_rate": 3.196644206770576e-06, "loss": 0.0093, "step": 11564 }, { "epoch": 0.794736118746564, "grad_norm": 0.3891906440258026, "learning_rate": 3.1945842927709627e-06, "loss": 0.0089, "step": 11565 }, { "epoch": 0.7948048378229796, "grad_norm": 0.4864308536052704, "learning_rate": 3.192524963587372e-06, "loss": 0.0068, "step": 11566 }, { "epoch": 0.7948735568993953, "grad_norm": 0.28003937005996704, "learning_rate": 3.190466219321819e-06, "loss": 0.0069, "step": 11567 }, { "epoch": 0.7949422759758109, "grad_norm": 0.8743689060211182, "learning_rate": 3.1884080600762942e-06, "loss": 0.017, "step": 11568 }, { "epoch": 0.7950109950522265, "grad_norm": 0.4698391258716583, "learning_rate": 3.1863504859527477e-06, "loss": 0.024, "step": 11569 }, { "epoch": 0.7950797141286421, "grad_norm": 0.378019779920578, "learning_rate": 3.1842934970531134e-06, "loss": 0.0076, "step": 11570 }, { "epoch": 0.7951484332050577, "grad_norm": 0.4206852316856384, "learning_rate": 3.182237093479284e-06, "loss": 0.0063, "step": 11571 }, { "epoch": 0.7952171522814734, "grad_norm": 0.10911117494106293, "learning_rate": 3.1801812753331304e-06, "loss": 0.0146, "step": 11572 }, { "epoch": 0.795285871357889, "grad_norm": 0.08500435948371887, "learning_rate": 3.1781260427164987e-06, "loss": 0.007, "step": 11573 }, { "epoch": 0.7953545904343046, "grad_norm": 0.5130225419998169, "learning_rate": 3.1760713957312004e-06, "loss": 0.0041, "step": 11574 }, { "epoch": 0.7954233095107202, "grad_norm": 1.1475365161895752, "learning_rate": 3.1740173344790124e-06, "loss": 0.0124, "step": 11575 }, { "epoch": 0.7954920285871357, "grad_norm": 0.4487590193748474, "learning_rate": 3.1719638590616968e-06, "loss": 0.0134, "step": 11576 }, { "epoch": 0.7955607476635514, "grad_norm": 0.37749531865119934, "learning_rate": 3.1699109695809704e-06, "loss": 0.0074, "step": 11577 }, { "epoch": 0.795629466739967, "grad_norm": 0.2928255498409271, "learning_rate": 3.1678586661385318e-06, "loss": 0.01, "step": 11578 }, { "epoch": 0.7956981858163826, "grad_norm": 0.3089973032474518, "learning_rate": 3.1658069488360546e-06, "loss": 0.0085, "step": 11579 }, { "epoch": 0.7957669048927982, "grad_norm": 0.2933674454689026, "learning_rate": 3.1637558177751662e-06, "loss": 0.0047, "step": 11580 }, { "epoch": 0.7958356239692138, "grad_norm": 0.3758344352245331, "learning_rate": 3.1617052730574853e-06, "loss": 0.0085, "step": 11581 }, { "epoch": 0.7959043430456295, "grad_norm": 0.22556762397289276, "learning_rate": 3.1596553147845803e-06, "loss": 0.0084, "step": 11582 }, { "epoch": 0.7959730621220451, "grad_norm": 0.2585517168045044, "learning_rate": 3.157605943058015e-06, "loss": 0.0045, "step": 11583 }, { "epoch": 0.7960417811984607, "grad_norm": 0.6586741209030151, "learning_rate": 3.1555571579792953e-06, "loss": 0.0137, "step": 11584 }, { "epoch": 0.7961105002748763, "grad_norm": 0.7663415670394897, "learning_rate": 3.1535089596499296e-06, "loss": 0.0126, "step": 11585 }, { "epoch": 0.7961792193512919, "grad_norm": 0.45348164439201355, "learning_rate": 3.1514613481713715e-06, "loss": 0.0067, "step": 11586 }, { "epoch": 0.7962479384277076, "grad_norm": 0.6468613743782043, "learning_rate": 3.149414323645064e-06, "loss": 0.0141, "step": 11587 }, { "epoch": 0.7963166575041232, "grad_norm": 0.8648176193237305, "learning_rate": 3.147367886172402e-06, "loss": 0.0216, "step": 11588 }, { "epoch": 0.7963853765805388, "grad_norm": 0.3055426776409149, "learning_rate": 3.145322035854771e-06, "loss": 0.0103, "step": 11589 }, { "epoch": 0.7964540956569544, "grad_norm": 0.11718937009572983, "learning_rate": 3.1432767727935056e-06, "loss": 0.0063, "step": 11590 }, { "epoch": 0.7965228147333699, "grad_norm": 0.300014853477478, "learning_rate": 3.141232097089941e-06, "loss": 0.0056, "step": 11591 }, { "epoch": 0.7965915338097856, "grad_norm": 0.4989040791988373, "learning_rate": 3.1391880088453546e-06, "loss": 0.0083, "step": 11592 }, { "epoch": 0.7966602528862012, "grad_norm": 0.30981844663619995, "learning_rate": 3.1371445081610117e-06, "loss": 0.0118, "step": 11593 }, { "epoch": 0.7967289719626168, "grad_norm": 0.3454674184322357, "learning_rate": 3.1351015951381386e-06, "loss": 0.009, "step": 11594 }, { "epoch": 0.7967976910390324, "grad_norm": 0.2625567317008972, "learning_rate": 3.133059269877942e-06, "loss": 0.0176, "step": 11595 }, { "epoch": 0.796866410115448, "grad_norm": 0.30322301387786865, "learning_rate": 3.131017532481588e-06, "loss": 0.0105, "step": 11596 }, { "epoch": 0.7969351291918637, "grad_norm": 0.13801346719264984, "learning_rate": 3.1289763830502244e-06, "loss": 0.0104, "step": 11597 }, { "epoch": 0.7970038482682793, "grad_norm": 0.63626629114151, "learning_rate": 3.1269358216849696e-06, "loss": 0.0145, "step": 11598 }, { "epoch": 0.7970725673446949, "grad_norm": 0.4733811616897583, "learning_rate": 3.1248958484868984e-06, "loss": 0.0051, "step": 11599 }, { "epoch": 0.7971412864211105, "grad_norm": 0.2073608636856079, "learning_rate": 3.122856463557078e-06, "loss": 0.0073, "step": 11600 }, { "epoch": 0.7972100054975261, "grad_norm": 0.6724640727043152, "learning_rate": 3.1208176669965255e-06, "loss": 0.0073, "step": 11601 }, { "epoch": 0.7972787245739418, "grad_norm": 0.6703370213508606, "learning_rate": 3.118779458906244e-06, "loss": 0.0111, "step": 11602 }, { "epoch": 0.7973474436503574, "grad_norm": 0.3266650140285492, "learning_rate": 3.116741839387201e-06, "loss": 0.0056, "step": 11603 }, { "epoch": 0.797416162726773, "grad_norm": 0.20602254569530487, "learning_rate": 3.114704808540339e-06, "loss": 0.0066, "step": 11604 }, { "epoch": 0.7974848818031886, "grad_norm": 0.18683013319969177, "learning_rate": 3.1126683664665624e-06, "loss": 0.0074, "step": 11605 }, { "epoch": 0.7975536008796041, "grad_norm": 0.585519015789032, "learning_rate": 3.1106325132667603e-06, "loss": 0.0051, "step": 11606 }, { "epoch": 0.7976223199560198, "grad_norm": 0.342934787273407, "learning_rate": 3.108597249041775e-06, "loss": 0.0116, "step": 11607 }, { "epoch": 0.7976910390324354, "grad_norm": 0.8730826377868652, "learning_rate": 3.1065625738924353e-06, "loss": 0.0126, "step": 11608 }, { "epoch": 0.797759758108851, "grad_norm": 0.37379196286201477, "learning_rate": 3.104528487919533e-06, "loss": 0.0035, "step": 11609 }, { "epoch": 0.7978284771852666, "grad_norm": 0.17142020165920258, "learning_rate": 3.1024949912238374e-06, "loss": 0.0055, "step": 11610 }, { "epoch": 0.7978971962616822, "grad_norm": 0.4228051006793976, "learning_rate": 3.100462083906077e-06, "loss": 0.0136, "step": 11611 }, { "epoch": 0.7979659153380979, "grad_norm": 0.4928126335144043, "learning_rate": 3.0984297660669643e-06, "loss": 0.0108, "step": 11612 }, { "epoch": 0.7980346344145135, "grad_norm": 0.4149508774280548, "learning_rate": 3.0963980378071666e-06, "loss": 0.0091, "step": 11613 }, { "epoch": 0.7981033534909291, "grad_norm": 0.555641233921051, "learning_rate": 3.094366899227339e-06, "loss": 0.0141, "step": 11614 }, { "epoch": 0.7981720725673447, "grad_norm": 0.4499942660331726, "learning_rate": 3.0923363504281023e-06, "loss": 0.0092, "step": 11615 }, { "epoch": 0.7982407916437603, "grad_norm": 0.46982672810554504, "learning_rate": 3.090306391510036e-06, "loss": 0.0069, "step": 11616 }, { "epoch": 0.798309510720176, "grad_norm": 0.3587166368961334, "learning_rate": 3.0882770225737115e-06, "loss": 0.0085, "step": 11617 }, { "epoch": 0.7983782297965916, "grad_norm": 0.7310749888420105, "learning_rate": 3.086248243719649e-06, "loss": 0.0127, "step": 11618 }, { "epoch": 0.7984469488730072, "grad_norm": 0.33699581027030945, "learning_rate": 3.0842200550483584e-06, "loss": 0.0166, "step": 11619 }, { "epoch": 0.7985156679494227, "grad_norm": 0.18923376500606537, "learning_rate": 3.082192456660302e-06, "loss": 0.0089, "step": 11620 }, { "epoch": 0.7985843870258383, "grad_norm": 0.3152071535587311, "learning_rate": 3.080165448655936e-06, "loss": 0.0091, "step": 11621 }, { "epoch": 0.798653106102254, "grad_norm": 0.3621912896633148, "learning_rate": 3.0781390311356634e-06, "loss": 0.0105, "step": 11622 }, { "epoch": 0.7987218251786696, "grad_norm": 0.34743139147758484, "learning_rate": 3.0761132041998787e-06, "loss": 0.0075, "step": 11623 }, { "epoch": 0.7987905442550852, "grad_norm": 0.46330660581588745, "learning_rate": 3.0740879679489247e-06, "loss": 0.009, "step": 11624 }, { "epoch": 0.7988592633315008, "grad_norm": 0.8678496479988098, "learning_rate": 3.0720633224831403e-06, "loss": 0.0141, "step": 11625 }, { "epoch": 0.7989279824079164, "grad_norm": 0.3076026141643524, "learning_rate": 3.0700392679028076e-06, "loss": 0.006, "step": 11626 }, { "epoch": 0.7989967014843321, "grad_norm": 0.08381961286067963, "learning_rate": 3.0680158043082106e-06, "loss": 0.007, "step": 11627 }, { "epoch": 0.7990654205607477, "grad_norm": 0.15288056433200836, "learning_rate": 3.0659929317995756e-06, "loss": 0.0115, "step": 11628 }, { "epoch": 0.7991341396371633, "grad_norm": 0.38853442668914795, "learning_rate": 3.0639706504771187e-06, "loss": 0.0092, "step": 11629 }, { "epoch": 0.7992028587135789, "grad_norm": 0.3293461799621582, "learning_rate": 3.0619489604410132e-06, "loss": 0.0084, "step": 11630 }, { "epoch": 0.7992715777899945, "grad_norm": 0.7496787905693054, "learning_rate": 3.059927861791416e-06, "loss": 0.0163, "step": 11631 }, { "epoch": 0.7993402968664101, "grad_norm": 0.10901151597499847, "learning_rate": 3.05790735462844e-06, "loss": 0.0042, "step": 11632 }, { "epoch": 0.7994090159428258, "grad_norm": 0.4035789370536804, "learning_rate": 3.055887439052184e-06, "loss": 0.0082, "step": 11633 }, { "epoch": 0.7994777350192414, "grad_norm": 0.37435853481292725, "learning_rate": 3.0538681151627105e-06, "loss": 0.0366, "step": 11634 }, { "epoch": 0.7995464540956569, "grad_norm": 0.5116621851921082, "learning_rate": 3.051849383060047e-06, "loss": 0.0142, "step": 11635 }, { "epoch": 0.7996151731720725, "grad_norm": 0.38429728150367737, "learning_rate": 3.0498312428442043e-06, "loss": 0.0146, "step": 11636 }, { "epoch": 0.7996838922484881, "grad_norm": 0.3148825764656067, "learning_rate": 3.04781369461515e-06, "loss": 0.0077, "step": 11637 }, { "epoch": 0.7997526113249038, "grad_norm": 0.9613494873046875, "learning_rate": 3.0457967384728332e-06, "loss": 0.0082, "step": 11638 }, { "epoch": 0.7998213304013194, "grad_norm": 0.19283431768417358, "learning_rate": 3.0437803745171675e-06, "loss": 0.0084, "step": 11639 }, { "epoch": 0.799890049477735, "grad_norm": 0.3774162232875824, "learning_rate": 3.0417646028480455e-06, "loss": 0.0048, "step": 11640 }, { "epoch": 0.7999587685541506, "grad_norm": 0.4092766344547272, "learning_rate": 3.0397494235653167e-06, "loss": 0.0153, "step": 11641 }, { "epoch": 0.8000274876305662, "grad_norm": 0.2998722493648529, "learning_rate": 3.037734836768816e-06, "loss": 0.0094, "step": 11642 }, { "epoch": 0.8000962067069819, "grad_norm": 0.7511789202690125, "learning_rate": 3.035720842558333e-06, "loss": 0.0065, "step": 11643 }, { "epoch": 0.8001649257833975, "grad_norm": 0.3918653428554535, "learning_rate": 3.033707441033648e-06, "loss": 0.0102, "step": 11644 }, { "epoch": 0.8002336448598131, "grad_norm": 0.5160775184631348, "learning_rate": 3.0316946322944855e-06, "loss": 0.01, "step": 11645 }, { "epoch": 0.8003023639362287, "grad_norm": 0.23454275727272034, "learning_rate": 3.0296824164405725e-06, "loss": 0.0094, "step": 11646 }, { "epoch": 0.8003710830126443, "grad_norm": 0.45854976773262024, "learning_rate": 3.0276707935715788e-06, "loss": 0.0072, "step": 11647 }, { "epoch": 0.80043980208906, "grad_norm": 0.26625385880470276, "learning_rate": 3.0256597637871646e-06, "loss": 0.0099, "step": 11648 }, { "epoch": 0.8005085211654756, "grad_norm": 0.34192922711372375, "learning_rate": 3.023649327186944e-06, "loss": 0.0109, "step": 11649 }, { "epoch": 0.8005772402418911, "grad_norm": 0.20812322199344635, "learning_rate": 3.021639483870516e-06, "loss": 0.01, "step": 11650 }, { "epoch": 0.8006459593183067, "grad_norm": 0.24900691211223602, "learning_rate": 3.019630233937438e-06, "loss": 0.0128, "step": 11651 }, { "epoch": 0.8007146783947223, "grad_norm": 0.248188316822052, "learning_rate": 3.0176215774872484e-06, "loss": 0.0099, "step": 11652 }, { "epoch": 0.800783397471138, "grad_norm": 0.801315426826477, "learning_rate": 3.015613514619455e-06, "loss": 0.0091, "step": 11653 }, { "epoch": 0.8008521165475536, "grad_norm": 0.2938556969165802, "learning_rate": 3.0136060454335242e-06, "loss": 0.0091, "step": 11654 }, { "epoch": 0.8009208356239692, "grad_norm": 0.14243699610233307, "learning_rate": 3.011599170028912e-06, "loss": 0.0074, "step": 11655 }, { "epoch": 0.8009895547003848, "grad_norm": 0.3001088500022888, "learning_rate": 3.009592888505026e-06, "loss": 0.0081, "step": 11656 }, { "epoch": 0.8010582737768004, "grad_norm": 0.2361498773097992, "learning_rate": 3.0075872009612553e-06, "loss": 0.0141, "step": 11657 }, { "epoch": 0.8011269928532161, "grad_norm": 0.20394982397556305, "learning_rate": 3.005582107496962e-06, "loss": 0.0117, "step": 11658 }, { "epoch": 0.8011957119296317, "grad_norm": 0.38480451703071594, "learning_rate": 3.003577608211473e-06, "loss": 0.0048, "step": 11659 }, { "epoch": 0.8012644310060473, "grad_norm": 0.491985559463501, "learning_rate": 3.0015737032040826e-06, "loss": 0.0119, "step": 11660 }, { "epoch": 0.8013331500824629, "grad_norm": 0.20756655931472778, "learning_rate": 2.9995703925740675e-06, "loss": 0.0054, "step": 11661 }, { "epoch": 0.8014018691588785, "grad_norm": 0.7173035740852356, "learning_rate": 2.997567676420659e-06, "loss": 0.0092, "step": 11662 }, { "epoch": 0.8014705882352942, "grad_norm": 0.3812156319618225, "learning_rate": 2.995565554843072e-06, "loss": 0.0114, "step": 11663 }, { "epoch": 0.8015393073117097, "grad_norm": 0.684987485408783, "learning_rate": 2.993564027940488e-06, "loss": 0.0134, "step": 11664 }, { "epoch": 0.8016080263881253, "grad_norm": 0.29277101159095764, "learning_rate": 2.9915630958120623e-06, "loss": 0.009, "step": 11665 }, { "epoch": 0.8016767454645409, "grad_norm": 0.48030737042427063, "learning_rate": 2.9895627585569067e-06, "loss": 0.012, "step": 11666 }, { "epoch": 0.8017454645409565, "grad_norm": 0.7168188095092773, "learning_rate": 2.9875630162741245e-06, "loss": 0.0098, "step": 11667 }, { "epoch": 0.8018141836173722, "grad_norm": 0.4946304261684418, "learning_rate": 2.9855638690627698e-06, "loss": 0.0054, "step": 11668 }, { "epoch": 0.8018829026937878, "grad_norm": 0.18668578565120697, "learning_rate": 2.9835653170218805e-06, "loss": 0.0092, "step": 11669 }, { "epoch": 0.8019516217702034, "grad_norm": 0.30449122190475464, "learning_rate": 2.981567360250464e-06, "loss": 0.0154, "step": 11670 }, { "epoch": 0.802020340846619, "grad_norm": 0.2190774381160736, "learning_rate": 2.9795699988474883e-06, "loss": 0.0039, "step": 11671 }, { "epoch": 0.8020890599230346, "grad_norm": 0.3592826724052429, "learning_rate": 2.9775732329119063e-06, "loss": 0.0175, "step": 11672 }, { "epoch": 0.8021577789994503, "grad_norm": 0.26828134059906006, "learning_rate": 2.9755770625426245e-06, "loss": 0.0065, "step": 11673 }, { "epoch": 0.8022264980758659, "grad_norm": 0.27100425958633423, "learning_rate": 2.9735814878385394e-06, "loss": 0.0079, "step": 11674 }, { "epoch": 0.8022952171522815, "grad_norm": 0.17244966328144073, "learning_rate": 2.971586508898492e-06, "loss": 0.0086, "step": 11675 }, { "epoch": 0.8023639362286971, "grad_norm": 0.14136594533920288, "learning_rate": 2.9695921258213293e-06, "loss": 0.0067, "step": 11676 }, { "epoch": 0.8024326553051127, "grad_norm": 0.30096113681793213, "learning_rate": 2.9675983387058333e-06, "loss": 0.0083, "step": 11677 }, { "epoch": 0.8025013743815284, "grad_norm": 1.3239121437072754, "learning_rate": 2.9656051476507828e-06, "loss": 0.0096, "step": 11678 }, { "epoch": 0.802570093457944, "grad_norm": 0.6778292059898376, "learning_rate": 2.963612552754908e-06, "loss": 0.0078, "step": 11679 }, { "epoch": 0.8026388125343595, "grad_norm": 0.40479785203933716, "learning_rate": 2.961620554116926e-06, "loss": 0.0192, "step": 11680 }, { "epoch": 0.8027075316107751, "grad_norm": 0.4580206573009491, "learning_rate": 2.9596291518355035e-06, "loss": 0.0153, "step": 11681 }, { "epoch": 0.8027762506871907, "grad_norm": 0.11930663883686066, "learning_rate": 2.9576383460093075e-06, "loss": 0.0042, "step": 11682 }, { "epoch": 0.8028449697636064, "grad_norm": 0.37560734152793884, "learning_rate": 2.9556481367369448e-06, "loss": 0.0057, "step": 11683 }, { "epoch": 0.802913688840022, "grad_norm": 0.6085273027420044, "learning_rate": 2.9536585241170143e-06, "loss": 0.0137, "step": 11684 }, { "epoch": 0.8029824079164376, "grad_norm": 0.13349996507167816, "learning_rate": 2.9516695082480726e-06, "loss": 0.0112, "step": 11685 }, { "epoch": 0.8030511269928532, "grad_norm": 0.4777126908302307, "learning_rate": 2.949681089228656e-06, "loss": 0.007, "step": 11686 }, { "epoch": 0.8031198460692688, "grad_norm": 0.24739797413349152, "learning_rate": 2.9476932671572597e-06, "loss": 0.0045, "step": 11687 }, { "epoch": 0.8031885651456845, "grad_norm": 0.6543446183204651, "learning_rate": 2.9457060421323613e-06, "loss": 0.0088, "step": 11688 }, { "epoch": 0.8032572842221001, "grad_norm": 0.4270671010017395, "learning_rate": 2.943719414252406e-06, "loss": 0.0135, "step": 11689 }, { "epoch": 0.8033260032985157, "grad_norm": 0.25936388969421387, "learning_rate": 2.9417333836158004e-06, "loss": 0.0097, "step": 11690 }, { "epoch": 0.8033947223749313, "grad_norm": 0.25966209173202515, "learning_rate": 2.939747950320936e-06, "loss": 0.0111, "step": 11691 }, { "epoch": 0.8034634414513468, "grad_norm": 0.4179360270500183, "learning_rate": 2.937763114466161e-06, "loss": 0.0093, "step": 11692 }, { "epoch": 0.8035321605277626, "grad_norm": 0.7512919306755066, "learning_rate": 2.935778876149801e-06, "loss": 0.0133, "step": 11693 }, { "epoch": 0.8036008796041781, "grad_norm": 0.22661393880844116, "learning_rate": 2.9337952354701515e-06, "loss": 0.0065, "step": 11694 }, { "epoch": 0.8036695986805937, "grad_norm": 0.2259044349193573, "learning_rate": 2.9318121925254846e-06, "loss": 0.0077, "step": 11695 }, { "epoch": 0.8037383177570093, "grad_norm": 0.33968815207481384, "learning_rate": 2.9298297474140262e-06, "loss": 0.0084, "step": 11696 }, { "epoch": 0.8038070368334249, "grad_norm": 1.0001730918884277, "learning_rate": 2.92784790023399e-06, "loss": 0.0075, "step": 11697 }, { "epoch": 0.8038757559098406, "grad_norm": 0.1839321106672287, "learning_rate": 2.9258666510835467e-06, "loss": 0.0096, "step": 11698 }, { "epoch": 0.8039444749862562, "grad_norm": 0.7300636172294617, "learning_rate": 2.923886000060846e-06, "loss": 0.0101, "step": 11699 }, { "epoch": 0.8040131940626718, "grad_norm": 0.25675931572914124, "learning_rate": 2.9219059472640054e-06, "loss": 0.0127, "step": 11700 }, { "epoch": 0.8040819131390874, "grad_norm": 0.18402278423309326, "learning_rate": 2.9199264927911167e-06, "loss": 0.0074, "step": 11701 }, { "epoch": 0.804150632215503, "grad_norm": 0.6212694644927979, "learning_rate": 2.9179476367402285e-06, "loss": 0.0167, "step": 11702 }, { "epoch": 0.8042193512919187, "grad_norm": 0.3374074101448059, "learning_rate": 2.9159693792093807e-06, "loss": 0.0053, "step": 11703 }, { "epoch": 0.8042880703683343, "grad_norm": 0.8338297605514526, "learning_rate": 2.913991720296561e-06, "loss": 0.0127, "step": 11704 }, { "epoch": 0.8043567894447499, "grad_norm": 0.6101974248886108, "learning_rate": 2.912014660099745e-06, "loss": 0.0114, "step": 11705 }, { "epoch": 0.8044255085211655, "grad_norm": 0.6490024328231812, "learning_rate": 2.9100381987168723e-06, "loss": 0.0147, "step": 11706 }, { "epoch": 0.804494227597581, "grad_norm": 0.5144279599189758, "learning_rate": 2.9080623362458495e-06, "loss": 0.0139, "step": 11707 }, { "epoch": 0.8045629466739967, "grad_norm": 0.29473087191581726, "learning_rate": 2.9060870727845623e-06, "loss": 0.0093, "step": 11708 }, { "epoch": 0.8046316657504123, "grad_norm": 0.15439167618751526, "learning_rate": 2.9041124084308534e-06, "loss": 0.0061, "step": 11709 }, { "epoch": 0.8047003848268279, "grad_norm": 0.292513906955719, "learning_rate": 2.9021383432825465e-06, "loss": 0.0128, "step": 11710 }, { "epoch": 0.8047691039032435, "grad_norm": 0.3404369056224823, "learning_rate": 2.900164877437435e-06, "loss": 0.0109, "step": 11711 }, { "epoch": 0.8048378229796591, "grad_norm": 0.6778784990310669, "learning_rate": 2.8981920109932843e-06, "loss": 0.0128, "step": 11712 }, { "epoch": 0.8049065420560748, "grad_norm": 0.6277775764465332, "learning_rate": 2.8962197440478154e-06, "loss": 0.0114, "step": 11713 }, { "epoch": 0.8049752611324904, "grad_norm": 0.33421486616134644, "learning_rate": 2.8942480766987413e-06, "loss": 0.007, "step": 11714 }, { "epoch": 0.805043980208906, "grad_norm": 0.4716896414756775, "learning_rate": 2.8922770090437244e-06, "loss": 0.0088, "step": 11715 }, { "epoch": 0.8051126992853216, "grad_norm": 0.5536317229270935, "learning_rate": 2.890306541180416e-06, "loss": 0.008, "step": 11716 }, { "epoch": 0.8051814183617372, "grad_norm": 0.37380141019821167, "learning_rate": 2.888336673206417e-06, "loss": 0.0057, "step": 11717 }, { "epoch": 0.8052501374381529, "grad_norm": 0.2860040068626404, "learning_rate": 2.8863674052193267e-06, "loss": 0.0124, "step": 11718 }, { "epoch": 0.8053188565145685, "grad_norm": 0.46443113684654236, "learning_rate": 2.884398737316687e-06, "loss": 0.009, "step": 11719 }, { "epoch": 0.8053875755909841, "grad_norm": 0.21908992528915405, "learning_rate": 2.8824306695960295e-06, "loss": 0.0134, "step": 11720 }, { "epoch": 0.8054562946673997, "grad_norm": 0.029080793261528015, "learning_rate": 2.8804632021548384e-06, "loss": 0.0056, "step": 11721 }, { "epoch": 0.8055250137438152, "grad_norm": 0.3741222620010376, "learning_rate": 2.8784963350905885e-06, "loss": 0.0222, "step": 11722 }, { "epoch": 0.805593732820231, "grad_norm": 0.46606457233428955, "learning_rate": 2.8765300685007054e-06, "loss": 0.0075, "step": 11723 }, { "epoch": 0.8056624518966465, "grad_norm": 0.5395549535751343, "learning_rate": 2.8745644024825985e-06, "loss": 0.0107, "step": 11724 }, { "epoch": 0.8057311709730621, "grad_norm": 0.12787257134914398, "learning_rate": 2.8725993371336456e-06, "loss": 0.0084, "step": 11725 }, { "epoch": 0.8057998900494777, "grad_norm": 0.15845629572868347, "learning_rate": 2.870634872551186e-06, "loss": 0.0072, "step": 11726 }, { "epoch": 0.8058686091258933, "grad_norm": 0.510643720626831, "learning_rate": 2.868671008832537e-06, "loss": 0.0073, "step": 11727 }, { "epoch": 0.805937328202309, "grad_norm": 0.20167280733585358, "learning_rate": 2.8667077460749886e-06, "loss": 0.0106, "step": 11728 }, { "epoch": 0.8060060472787246, "grad_norm": 0.31398072838783264, "learning_rate": 2.86474508437579e-06, "loss": 0.0057, "step": 11729 }, { "epoch": 0.8060747663551402, "grad_norm": 0.4590373635292053, "learning_rate": 2.86278302383217e-06, "loss": 0.0098, "step": 11730 }, { "epoch": 0.8061434854315558, "grad_norm": 0.499659925699234, "learning_rate": 2.8608215645413312e-06, "loss": 0.008, "step": 11731 }, { "epoch": 0.8062122045079714, "grad_norm": 0.4245963394641876, "learning_rate": 2.85886070660043e-06, "loss": 0.0089, "step": 11732 }, { "epoch": 0.8062809235843871, "grad_norm": 0.33314767479896545, "learning_rate": 2.8569004501066133e-06, "loss": 0.0082, "step": 11733 }, { "epoch": 0.8063496426608027, "grad_norm": 0.3206422030925751, "learning_rate": 2.8549407951569788e-06, "loss": 0.0116, "step": 11734 }, { "epoch": 0.8064183617372183, "grad_norm": 0.2588420808315277, "learning_rate": 2.852981741848608e-06, "loss": 0.0096, "step": 11735 }, { "epoch": 0.8064870808136338, "grad_norm": 0.2260838896036148, "learning_rate": 2.8510232902785484e-06, "loss": 0.0064, "step": 11736 }, { "epoch": 0.8065557998900494, "grad_norm": 0.1662013828754425, "learning_rate": 2.8490654405438214e-06, "loss": 0.0092, "step": 11737 }, { "epoch": 0.8066245189664651, "grad_norm": 0.4730151295661926, "learning_rate": 2.8471081927414074e-06, "loss": 0.0057, "step": 11738 }, { "epoch": 0.8066932380428807, "grad_norm": 0.6712153553962708, "learning_rate": 2.8451515469682727e-06, "loss": 0.0116, "step": 11739 }, { "epoch": 0.8067619571192963, "grad_norm": 0.2795875668525696, "learning_rate": 2.843195503321337e-06, "loss": 0.0092, "step": 11740 }, { "epoch": 0.8068306761957119, "grad_norm": 0.07576721906661987, "learning_rate": 2.8412400618975025e-06, "loss": 0.0046, "step": 11741 }, { "epoch": 0.8068993952721275, "grad_norm": 0.3061615526676178, "learning_rate": 2.8392852227936424e-06, "loss": 0.0066, "step": 11742 }, { "epoch": 0.8069681143485432, "grad_norm": 0.3011729419231415, "learning_rate": 2.8373309861065867e-06, "loss": 0.014, "step": 11743 }, { "epoch": 0.8070368334249588, "grad_norm": 0.5142695307731628, "learning_rate": 2.8353773519331483e-06, "loss": 0.0123, "step": 11744 }, { "epoch": 0.8071055525013744, "grad_norm": 0.5687927603721619, "learning_rate": 2.8334243203701097e-06, "loss": 0.0133, "step": 11745 }, { "epoch": 0.80717427157779, "grad_norm": 0.44960829615592957, "learning_rate": 2.831471891514214e-06, "loss": 0.0066, "step": 11746 }, { "epoch": 0.8072429906542056, "grad_norm": 0.7568857073783875, "learning_rate": 2.8295200654621835e-06, "loss": 0.0082, "step": 11747 }, { "epoch": 0.8073117097306213, "grad_norm": 0.2761005163192749, "learning_rate": 2.827568842310712e-06, "loss": 0.0055, "step": 11748 }, { "epoch": 0.8073804288070369, "grad_norm": 0.5663027167320251, "learning_rate": 2.82561822215645e-06, "loss": 0.0065, "step": 11749 }, { "epoch": 0.8074491478834525, "grad_norm": 0.3697671592235565, "learning_rate": 2.8236682050960357e-06, "loss": 0.0076, "step": 11750 }, { "epoch": 0.807517866959868, "grad_norm": 0.40445172786712646, "learning_rate": 2.8217187912260605e-06, "loss": 0.0075, "step": 11751 }, { "epoch": 0.8075865860362836, "grad_norm": 0.28030478954315186, "learning_rate": 2.8197699806431026e-06, "loss": 0.0092, "step": 11752 }, { "epoch": 0.8076553051126993, "grad_norm": 0.2045709192752838, "learning_rate": 2.817821773443691e-06, "loss": 0.0057, "step": 11753 }, { "epoch": 0.8077240241891149, "grad_norm": 0.3714984953403473, "learning_rate": 2.8158741697243523e-06, "loss": 0.0075, "step": 11754 }, { "epoch": 0.8077927432655305, "grad_norm": 0.2536352276802063, "learning_rate": 2.813927169581552e-06, "loss": 0.0127, "step": 11755 }, { "epoch": 0.8078614623419461, "grad_norm": 0.37984171509742737, "learning_rate": 2.811980773111752e-06, "loss": 0.012, "step": 11756 }, { "epoch": 0.8079301814183617, "grad_norm": 0.37116193771362305, "learning_rate": 2.810034980411361e-06, "loss": 0.0059, "step": 11757 }, { "epoch": 0.8079989004947774, "grad_norm": 0.33140820264816284, "learning_rate": 2.8080897915767813e-06, "loss": 0.008, "step": 11758 }, { "epoch": 0.808067619571193, "grad_norm": 0.3866419196128845, "learning_rate": 2.806145206704358e-06, "loss": 0.0149, "step": 11759 }, { "epoch": 0.8081363386476086, "grad_norm": 1.2745075225830078, "learning_rate": 2.8042012258904416e-06, "loss": 0.0046, "step": 11760 }, { "epoch": 0.8082050577240242, "grad_norm": 0.33097702264785767, "learning_rate": 2.8022578492313176e-06, "loss": 0.0128, "step": 11761 }, { "epoch": 0.8082737768004398, "grad_norm": 0.31193241477012634, "learning_rate": 2.8003150768232666e-06, "loss": 0.0063, "step": 11762 }, { "epoch": 0.8083424958768554, "grad_norm": 0.0943271592259407, "learning_rate": 2.7983729087625225e-06, "loss": 0.0055, "step": 11763 }, { "epoch": 0.8084112149532711, "grad_norm": 0.399429053068161, "learning_rate": 2.7964313451453006e-06, "loss": 0.0115, "step": 11764 }, { "epoch": 0.8084799340296867, "grad_norm": 0.3129880130290985, "learning_rate": 2.7944903860677785e-06, "loss": 0.0066, "step": 11765 }, { "epoch": 0.8085486531061022, "grad_norm": 0.41567498445510864, "learning_rate": 2.7925500316261075e-06, "loss": 0.0072, "step": 11766 }, { "epoch": 0.8086173721825178, "grad_norm": 0.32476410269737244, "learning_rate": 2.7906102819164163e-06, "loss": 0.0059, "step": 11767 }, { "epoch": 0.8086860912589334, "grad_norm": 0.21795059740543365, "learning_rate": 2.7886711370347846e-06, "loss": 0.0064, "step": 11768 }, { "epoch": 0.8087548103353491, "grad_norm": 0.5959479212760925, "learning_rate": 2.786732597077285e-06, "loss": 0.0116, "step": 11769 }, { "epoch": 0.8088235294117647, "grad_norm": 0.25509676337242126, "learning_rate": 2.784794662139938e-06, "loss": 0.0077, "step": 11770 }, { "epoch": 0.8088922484881803, "grad_norm": 0.20471756160259247, "learning_rate": 2.7828573323187494e-06, "loss": 0.0063, "step": 11771 }, { "epoch": 0.8089609675645959, "grad_norm": 0.29427099227905273, "learning_rate": 2.780920607709693e-06, "loss": 0.0079, "step": 11772 }, { "epoch": 0.8090296866410115, "grad_norm": 0.18856899440288544, "learning_rate": 2.778984488408711e-06, "loss": 0.0066, "step": 11773 }, { "epoch": 0.8090984057174272, "grad_norm": 0.5364164113998413, "learning_rate": 2.7770489745117085e-06, "loss": 0.0091, "step": 11774 }, { "epoch": 0.8091671247938428, "grad_norm": 0.11244312673807144, "learning_rate": 2.7751140661145747e-06, "loss": 0.01, "step": 11775 }, { "epoch": 0.8092358438702584, "grad_norm": 0.2231421321630478, "learning_rate": 2.7731797633131527e-06, "loss": 0.0108, "step": 11776 }, { "epoch": 0.809304562946674, "grad_norm": 0.4078375995159149, "learning_rate": 2.771246066203268e-06, "loss": 0.0087, "step": 11777 }, { "epoch": 0.8093732820230896, "grad_norm": 0.558715283870697, "learning_rate": 2.7693129748807116e-06, "loss": 0.0089, "step": 11778 }, { "epoch": 0.8094420010995053, "grad_norm": 0.8231844902038574, "learning_rate": 2.767380489441249e-06, "loss": 0.0103, "step": 11779 }, { "epoch": 0.8095107201759209, "grad_norm": 0.24654217064380646, "learning_rate": 2.7654486099806035e-06, "loss": 0.0121, "step": 11780 }, { "epoch": 0.8095794392523364, "grad_norm": 0.29010456800460815, "learning_rate": 2.7635173365944855e-06, "loss": 0.0065, "step": 11781 }, { "epoch": 0.809648158328752, "grad_norm": 0.4336717426776886, "learning_rate": 2.761586669378558e-06, "loss": 0.016, "step": 11782 }, { "epoch": 0.8097168774051676, "grad_norm": 0.544600248336792, "learning_rate": 2.7596566084284646e-06, "loss": 0.0132, "step": 11783 }, { "epoch": 0.8097855964815833, "grad_norm": 0.41756629943847656, "learning_rate": 2.7577271538398226e-06, "loss": 0.0076, "step": 11784 }, { "epoch": 0.8098543155579989, "grad_norm": 0.20457959175109863, "learning_rate": 2.755798305708203e-06, "loss": 0.0072, "step": 11785 }, { "epoch": 0.8099230346344145, "grad_norm": 0.15885299444198608, "learning_rate": 2.753870064129167e-06, "loss": 0.0072, "step": 11786 }, { "epoch": 0.8099917537108301, "grad_norm": 0.05611056089401245, "learning_rate": 2.7519424291982277e-06, "loss": 0.0101, "step": 11787 }, { "epoch": 0.8100604727872457, "grad_norm": 0.7128401398658752, "learning_rate": 2.7500154010108837e-06, "loss": 0.0088, "step": 11788 }, { "epoch": 0.8101291918636614, "grad_norm": 0.3308528661727905, "learning_rate": 2.7480889796625817e-06, "loss": 0.0099, "step": 11789 }, { "epoch": 0.810197910940077, "grad_norm": 0.2831469774246216, "learning_rate": 2.746163165248773e-06, "loss": 0.0133, "step": 11790 }, { "epoch": 0.8102666300164926, "grad_norm": 0.3473883271217346, "learning_rate": 2.7442379578648434e-06, "loss": 0.0133, "step": 11791 }, { "epoch": 0.8103353490929082, "grad_norm": 0.4684189558029175, "learning_rate": 2.742313357606173e-06, "loss": 0.0107, "step": 11792 }, { "epoch": 0.8104040681693238, "grad_norm": 0.7243918776512146, "learning_rate": 2.740389364568094e-06, "loss": 0.0174, "step": 11793 }, { "epoch": 0.8104727872457395, "grad_norm": 0.5310693979263306, "learning_rate": 2.7384659788459243e-06, "loss": 0.009, "step": 11794 }, { "epoch": 0.810541506322155, "grad_norm": 0.38229286670684814, "learning_rate": 2.7365432005349343e-06, "loss": 0.0137, "step": 11795 }, { "epoch": 0.8106102253985706, "grad_norm": 0.09701742231845856, "learning_rate": 2.7346210297303897e-06, "loss": 0.0089, "step": 11796 }, { "epoch": 0.8106789444749862, "grad_norm": 0.32833802700042725, "learning_rate": 2.732699466527499e-06, "loss": 0.0113, "step": 11797 }, { "epoch": 0.8107476635514018, "grad_norm": 0.2919418513774872, "learning_rate": 2.7307785110214594e-06, "loss": 0.0131, "step": 11798 }, { "epoch": 0.8108163826278175, "grad_norm": 0.3608528673648834, "learning_rate": 2.728858163307425e-06, "loss": 0.0068, "step": 11799 }, { "epoch": 0.8108851017042331, "grad_norm": 0.2250213474035263, "learning_rate": 2.726938423480531e-06, "loss": 0.0071, "step": 11800 }, { "epoch": 0.8109538207806487, "grad_norm": 1.1207451820373535, "learning_rate": 2.725019291635872e-06, "loss": 0.0152, "step": 11801 }, { "epoch": 0.8110225398570643, "grad_norm": 0.5084227919578552, "learning_rate": 2.723100767868521e-06, "loss": 0.0109, "step": 11802 }, { "epoch": 0.8110912589334799, "grad_norm": 0.37300464510917664, "learning_rate": 2.7211828522735225e-06, "loss": 0.0132, "step": 11803 }, { "epoch": 0.8111599780098956, "grad_norm": 0.2190195918083191, "learning_rate": 2.7192655449458757e-06, "loss": 0.0094, "step": 11804 }, { "epoch": 0.8112286970863112, "grad_norm": 0.31777748465538025, "learning_rate": 2.7173488459805714e-06, "loss": 0.016, "step": 11805 }, { "epoch": 0.8112974161627268, "grad_norm": 0.2940387427806854, "learning_rate": 2.7154327554725476e-06, "loss": 0.0123, "step": 11806 }, { "epoch": 0.8113661352391424, "grad_norm": 0.238676980137825, "learning_rate": 2.7135172735167314e-06, "loss": 0.007, "step": 11807 }, { "epoch": 0.811434854315558, "grad_norm": 0.2829681932926178, "learning_rate": 2.7116024002080074e-06, "loss": 0.0041, "step": 11808 }, { "epoch": 0.8115035733919737, "grad_norm": 0.48284298181533813, "learning_rate": 2.7096881356412407e-06, "loss": 0.0086, "step": 11809 }, { "epoch": 0.8115722924683892, "grad_norm": 0.34670427441596985, "learning_rate": 2.707774479911253e-06, "loss": 0.0071, "step": 11810 }, { "epoch": 0.8116410115448048, "grad_norm": 0.17597761750221252, "learning_rate": 2.705861433112849e-06, "loss": 0.0086, "step": 11811 }, { "epoch": 0.8117097306212204, "grad_norm": 0.42549383640289307, "learning_rate": 2.703948995340791e-06, "loss": 0.0181, "step": 11812 }, { "epoch": 0.811778449697636, "grad_norm": 0.5848963856697083, "learning_rate": 2.70203716668982e-06, "loss": 0.0198, "step": 11813 }, { "epoch": 0.8118471687740517, "grad_norm": 0.25426703691482544, "learning_rate": 2.700125947254643e-06, "loss": 0.007, "step": 11814 }, { "epoch": 0.8119158878504673, "grad_norm": 0.2754823863506317, "learning_rate": 2.6982153371299432e-06, "loss": 0.0064, "step": 11815 }, { "epoch": 0.8119846069268829, "grad_norm": 0.3435717225074768, "learning_rate": 2.696305336410362e-06, "loss": 0.0099, "step": 11816 }, { "epoch": 0.8120533260032985, "grad_norm": 0.4858701527118683, "learning_rate": 2.694395945190521e-06, "loss": 0.0073, "step": 11817 }, { "epoch": 0.8121220450797141, "grad_norm": 0.10075640678405762, "learning_rate": 2.6924871635650033e-06, "loss": 0.0163, "step": 11818 }, { "epoch": 0.8121907641561298, "grad_norm": 0.3220650553703308, "learning_rate": 2.690578991628367e-06, "loss": 0.0107, "step": 11819 }, { "epoch": 0.8122594832325454, "grad_norm": 0.17394395172595978, "learning_rate": 2.688671429475146e-06, "loss": 0.0072, "step": 11820 }, { "epoch": 0.812328202308961, "grad_norm": 0.427822083234787, "learning_rate": 2.686764477199826e-06, "loss": 0.0122, "step": 11821 }, { "epoch": 0.8123969213853766, "grad_norm": 0.18474295735359192, "learning_rate": 2.6848581348968824e-06, "loss": 0.011, "step": 11822 }, { "epoch": 0.8124656404617921, "grad_norm": 0.8346499800682068, "learning_rate": 2.6829524026607445e-06, "loss": 0.0097, "step": 11823 }, { "epoch": 0.8125343595382079, "grad_norm": 0.4451303482055664, "learning_rate": 2.681047280585826e-06, "loss": 0.0076, "step": 11824 }, { "epoch": 0.8126030786146234, "grad_norm": 0.43159031867980957, "learning_rate": 2.6791427687664903e-06, "loss": 0.0096, "step": 11825 }, { "epoch": 0.812671797691039, "grad_norm": 0.27441394329071045, "learning_rate": 2.6772388672971002e-06, "loss": 0.0105, "step": 11826 }, { "epoch": 0.8127405167674546, "grad_norm": 0.23527003824710846, "learning_rate": 2.675335576271957e-06, "loss": 0.0103, "step": 11827 }, { "epoch": 0.8128092358438702, "grad_norm": 0.2757912576198578, "learning_rate": 2.6734328957853555e-06, "loss": 0.0062, "step": 11828 }, { "epoch": 0.8128779549202859, "grad_norm": 0.3547064960002899, "learning_rate": 2.6715308259315415e-06, "loss": 0.0072, "step": 11829 }, { "epoch": 0.8129466739967015, "grad_norm": 0.07255176454782486, "learning_rate": 2.6696293668047476e-06, "loss": 0.0053, "step": 11830 }, { "epoch": 0.8130153930731171, "grad_norm": 0.1744225025177002, "learning_rate": 2.667728518499158e-06, "loss": 0.0102, "step": 11831 }, { "epoch": 0.8130841121495327, "grad_norm": 0.34076055884361267, "learning_rate": 2.6658282811089513e-06, "loss": 0.0074, "step": 11832 }, { "epoch": 0.8131528312259483, "grad_norm": 0.4385744035243988, "learning_rate": 2.6639286547282493e-06, "loss": 0.0124, "step": 11833 }, { "epoch": 0.813221550302364, "grad_norm": 0.2868601083755493, "learning_rate": 2.6620296394511633e-06, "loss": 0.0111, "step": 11834 }, { "epoch": 0.8132902693787796, "grad_norm": 0.499379962682724, "learning_rate": 2.66013123537176e-06, "loss": 0.0085, "step": 11835 }, { "epoch": 0.8133589884551952, "grad_norm": 0.6257550716400146, "learning_rate": 2.6582334425840887e-06, "loss": 0.0057, "step": 11836 }, { "epoch": 0.8134277075316108, "grad_norm": 0.25787708163261414, "learning_rate": 2.656336261182156e-06, "loss": 0.0087, "step": 11837 }, { "epoch": 0.8134964266080263, "grad_norm": 0.5062836408615112, "learning_rate": 2.6544396912599472e-06, "loss": 0.0116, "step": 11838 }, { "epoch": 0.813565145684442, "grad_norm": 0.20780129730701447, "learning_rate": 2.6525437329114183e-06, "loss": 0.0088, "step": 11839 }, { "epoch": 0.8136338647608576, "grad_norm": 0.20793962478637695, "learning_rate": 2.650648386230484e-06, "loss": 0.0096, "step": 11840 }, { "epoch": 0.8137025838372732, "grad_norm": 0.19854317605495453, "learning_rate": 2.6487536513110433e-06, "loss": 0.0025, "step": 11841 }, { "epoch": 0.8137713029136888, "grad_norm": 0.307787150144577, "learning_rate": 2.6468595282469505e-06, "loss": 0.0093, "step": 11842 }, { "epoch": 0.8138400219901044, "grad_norm": 0.24411295354366302, "learning_rate": 2.644966017132039e-06, "loss": 0.0057, "step": 11843 }, { "epoch": 0.8139087410665201, "grad_norm": 1.0994981527328491, "learning_rate": 2.6430731180601087e-06, "loss": 0.017, "step": 11844 }, { "epoch": 0.8139774601429357, "grad_norm": 0.28767573833465576, "learning_rate": 2.6411808311249363e-06, "loss": 0.0071, "step": 11845 }, { "epoch": 0.8140461792193513, "grad_norm": 0.1640925109386444, "learning_rate": 2.6392891564202536e-06, "loss": 0.0056, "step": 11846 }, { "epoch": 0.8141148982957669, "grad_norm": 0.3400316834449768, "learning_rate": 2.6373980940397775e-06, "loss": 0.0086, "step": 11847 }, { "epoch": 0.8141836173721825, "grad_norm": 0.36593496799468994, "learning_rate": 2.6355076440771793e-06, "loss": 0.009, "step": 11848 }, { "epoch": 0.8142523364485982, "grad_norm": 0.21658110618591309, "learning_rate": 2.6336178066261124e-06, "loss": 0.0083, "step": 11849 }, { "epoch": 0.8143210555250138, "grad_norm": 0.1672147959470749, "learning_rate": 2.631728581780194e-06, "loss": 0.0073, "step": 11850 }, { "epoch": 0.8143897746014294, "grad_norm": 0.33358463644981384, "learning_rate": 2.629839969633017e-06, "loss": 0.0149, "step": 11851 }, { "epoch": 0.814458493677845, "grad_norm": 0.0804285928606987, "learning_rate": 2.627951970278134e-06, "loss": 0.0049, "step": 11852 }, { "epoch": 0.8145272127542605, "grad_norm": 0.3712337911128998, "learning_rate": 2.626064583809078e-06, "loss": 0.0103, "step": 11853 }, { "epoch": 0.8145959318306762, "grad_norm": 0.44025886058807373, "learning_rate": 2.6241778103193378e-06, "loss": 0.0059, "step": 11854 }, { "epoch": 0.8146646509070918, "grad_norm": 0.7239102721214294, "learning_rate": 2.6222916499023865e-06, "loss": 0.0121, "step": 11855 }, { "epoch": 0.8147333699835074, "grad_norm": 0.636823832988739, "learning_rate": 2.620406102651663e-06, "loss": 0.0105, "step": 11856 }, { "epoch": 0.814802089059923, "grad_norm": 0.5324360132217407, "learning_rate": 2.6185211686605643e-06, "loss": 0.0079, "step": 11857 }, { "epoch": 0.8148708081363386, "grad_norm": 0.9908967018127441, "learning_rate": 2.616636848022477e-06, "loss": 0.013, "step": 11858 }, { "epoch": 0.8149395272127543, "grad_norm": 0.24886776506900787, "learning_rate": 2.6147531408307375e-06, "loss": 0.0081, "step": 11859 }, { "epoch": 0.8150082462891699, "grad_norm": 0.2632580101490021, "learning_rate": 2.6128700471786683e-06, "loss": 0.0112, "step": 11860 }, { "epoch": 0.8150769653655855, "grad_norm": 0.4491496980190277, "learning_rate": 2.6109875671595416e-06, "loss": 0.0107, "step": 11861 }, { "epoch": 0.8151456844420011, "grad_norm": 2.0900990962982178, "learning_rate": 2.6091057008666287e-06, "loss": 0.0144, "step": 11862 }, { "epoch": 0.8152144035184167, "grad_norm": 0.3042865991592407, "learning_rate": 2.607224448393139e-06, "loss": 0.0108, "step": 11863 }, { "epoch": 0.8152831225948324, "grad_norm": 0.8896772265434265, "learning_rate": 2.6053438098322766e-06, "loss": 0.0108, "step": 11864 }, { "epoch": 0.815351841671248, "grad_norm": 0.24330659210681915, "learning_rate": 2.6034637852771974e-06, "loss": 0.0062, "step": 11865 }, { "epoch": 0.8154205607476636, "grad_norm": 0.3169647753238678, "learning_rate": 2.6015843748210384e-06, "loss": 0.0079, "step": 11866 }, { "epoch": 0.8154892798240791, "grad_norm": 0.5340826511383057, "learning_rate": 2.5997055785568945e-06, "loss": 0.0105, "step": 11867 }, { "epoch": 0.8155579989004947, "grad_norm": 0.3515458405017853, "learning_rate": 2.5978273965778435e-06, "loss": 0.0058, "step": 11868 }, { "epoch": 0.8156267179769104, "grad_norm": 0.2772911787033081, "learning_rate": 2.595949828976926e-06, "loss": 0.0094, "step": 11869 }, { "epoch": 0.815695437053326, "grad_norm": 0.288655161857605, "learning_rate": 2.594072875847156e-06, "loss": 0.0025, "step": 11870 }, { "epoch": 0.8157641561297416, "grad_norm": 0.30893248319625854, "learning_rate": 2.5921965372815054e-06, "loss": 0.0085, "step": 11871 }, { "epoch": 0.8158328752061572, "grad_norm": 0.4246234893798828, "learning_rate": 2.5903208133729334e-06, "loss": 0.0112, "step": 11872 }, { "epoch": 0.8159015942825728, "grad_norm": 0.5681286454200745, "learning_rate": 2.588445704214352e-06, "loss": 0.0141, "step": 11873 }, { "epoch": 0.8159703133589885, "grad_norm": 0.4030972421169281, "learning_rate": 2.586571209898653e-06, "loss": 0.0095, "step": 11874 }, { "epoch": 0.8160390324354041, "grad_norm": 0.42221662402153015, "learning_rate": 2.5846973305186987e-06, "loss": 0.0102, "step": 11875 }, { "epoch": 0.8161077515118197, "grad_norm": 0.5205326676368713, "learning_rate": 2.5828240661673097e-06, "loss": 0.0104, "step": 11876 }, { "epoch": 0.8161764705882353, "grad_norm": 0.0859297513961792, "learning_rate": 2.5809514169372927e-06, "loss": 0.004, "step": 11877 }, { "epoch": 0.8162451896646509, "grad_norm": 0.3330255448818207, "learning_rate": 2.5790793829214054e-06, "loss": 0.0106, "step": 11878 }, { "epoch": 0.8163139087410666, "grad_norm": 0.19777962565422058, "learning_rate": 2.577207964212392e-06, "loss": 0.0059, "step": 11879 }, { "epoch": 0.8163826278174822, "grad_norm": 0.19818313419818878, "learning_rate": 2.5753371609029497e-06, "loss": 0.0068, "step": 11880 }, { "epoch": 0.8164513468938978, "grad_norm": 0.3573434352874756, "learning_rate": 2.573466973085768e-06, "loss": 0.0055, "step": 11881 }, { "epoch": 0.8165200659703133, "grad_norm": 0.49621570110321045, "learning_rate": 2.571597400853481e-06, "loss": 0.0073, "step": 11882 }, { "epoch": 0.8165887850467289, "grad_norm": 0.5061037540435791, "learning_rate": 2.5697284442987095e-06, "loss": 0.0081, "step": 11883 }, { "epoch": 0.8166575041231446, "grad_norm": 0.5597586631774902, "learning_rate": 2.567860103514033e-06, "loss": 0.0164, "step": 11884 }, { "epoch": 0.8167262231995602, "grad_norm": 0.22974538803100586, "learning_rate": 2.565992378592011e-06, "loss": 0.0069, "step": 11885 }, { "epoch": 0.8167949422759758, "grad_norm": 0.6925187110900879, "learning_rate": 2.5641252696251554e-06, "loss": 0.0108, "step": 11886 }, { "epoch": 0.8168636613523914, "grad_norm": 0.5963859558105469, "learning_rate": 2.562258776705975e-06, "loss": 0.0076, "step": 11887 }, { "epoch": 0.816932380428807, "grad_norm": 0.34710216522216797, "learning_rate": 2.5603928999269194e-06, "loss": 0.0054, "step": 11888 }, { "epoch": 0.8170010995052226, "grad_norm": 0.3505125641822815, "learning_rate": 2.5585276393804276e-06, "loss": 0.0065, "step": 11889 }, { "epoch": 0.8170698185816383, "grad_norm": 0.7705103158950806, "learning_rate": 2.5566629951588967e-06, "loss": 0.0095, "step": 11890 }, { "epoch": 0.8171385376580539, "grad_norm": 0.6166166663169861, "learning_rate": 2.5547989673547012e-06, "loss": 0.0079, "step": 11891 }, { "epoch": 0.8172072567344695, "grad_norm": 0.29673564434051514, "learning_rate": 2.552935556060174e-06, "loss": 0.006, "step": 11892 }, { "epoch": 0.8172759758108851, "grad_norm": 0.28454574942588806, "learning_rate": 2.55107276136763e-06, "loss": 0.0142, "step": 11893 }, { "epoch": 0.8173446948873007, "grad_norm": 0.16039502620697021, "learning_rate": 2.5492105833693513e-06, "loss": 0.0068, "step": 11894 }, { "epoch": 0.8174134139637164, "grad_norm": 0.28385961055755615, "learning_rate": 2.547349022157578e-06, "loss": 0.0092, "step": 11895 }, { "epoch": 0.817482133040132, "grad_norm": 0.049376230686903, "learning_rate": 2.5454880778245377e-06, "loss": 0.0021, "step": 11896 }, { "epoch": 0.8175508521165475, "grad_norm": 0.7647057175636292, "learning_rate": 2.543627750462408e-06, "loss": 0.0098, "step": 11897 }, { "epoch": 0.8176195711929631, "grad_norm": 0.9412760138511658, "learning_rate": 2.5417680401633512e-06, "loss": 0.0109, "step": 11898 }, { "epoch": 0.8176882902693787, "grad_norm": 0.5542225241661072, "learning_rate": 2.5399089470194915e-06, "loss": 0.0072, "step": 11899 }, { "epoch": 0.8177570093457944, "grad_norm": 0.21526512503623962, "learning_rate": 2.53805047112293e-06, "loss": 0.008, "step": 11900 }, { "epoch": 0.81782572842221, "grad_norm": 0.16046005487442017, "learning_rate": 2.5361926125657243e-06, "loss": 0.009, "step": 11901 }, { "epoch": 0.8178944474986256, "grad_norm": 0.3157505393028259, "learning_rate": 2.5343353714399147e-06, "loss": 0.0108, "step": 11902 }, { "epoch": 0.8179631665750412, "grad_norm": 0.28912585973739624, "learning_rate": 2.5324787478374996e-06, "loss": 0.0085, "step": 11903 }, { "epoch": 0.8180318856514568, "grad_norm": 0.24548766016960144, "learning_rate": 2.5306227418504547e-06, "loss": 0.0069, "step": 11904 }, { "epoch": 0.8181006047278725, "grad_norm": 0.3542698323726654, "learning_rate": 2.5287673535707235e-06, "loss": 0.0099, "step": 11905 }, { "epoch": 0.8181693238042881, "grad_norm": 0.5794983506202698, "learning_rate": 2.5269125830902208e-06, "loss": 0.0113, "step": 11906 }, { "epoch": 0.8182380428807037, "grad_norm": 0.3455837368965149, "learning_rate": 2.525058430500823e-06, "loss": 0.0053, "step": 11907 }, { "epoch": 0.8183067619571193, "grad_norm": 0.8851084113121033, "learning_rate": 2.523204895894385e-06, "loss": 0.0153, "step": 11908 }, { "epoch": 0.8183754810335349, "grad_norm": 0.19529728591442108, "learning_rate": 2.5213519793627233e-06, "loss": 0.0072, "step": 11909 }, { "epoch": 0.8184442001099506, "grad_norm": 0.4065278172492981, "learning_rate": 2.5194996809976277e-06, "loss": 0.0079, "step": 11910 }, { "epoch": 0.8185129191863662, "grad_norm": 0.233648881316185, "learning_rate": 2.517648000890864e-06, "loss": 0.0108, "step": 11911 }, { "epoch": 0.8185816382627817, "grad_norm": 0.1029965803027153, "learning_rate": 2.5157969391341524e-06, "loss": 0.0086, "step": 11912 }, { "epoch": 0.8186503573391973, "grad_norm": 0.5889387726783752, "learning_rate": 2.5139464958191973e-06, "loss": 0.0099, "step": 11913 }, { "epoch": 0.8187190764156129, "grad_norm": 0.7834978103637695, "learning_rate": 2.512096671037658e-06, "loss": 0.0088, "step": 11914 }, { "epoch": 0.8187877954920286, "grad_norm": 0.5128000974655151, "learning_rate": 2.51024746488118e-06, "loss": 0.0122, "step": 11915 }, { "epoch": 0.8188565145684442, "grad_norm": 0.16585350036621094, "learning_rate": 2.5083988774413567e-06, "loss": 0.0096, "step": 11916 }, { "epoch": 0.8189252336448598, "grad_norm": 0.8225013017654419, "learning_rate": 2.50655090880978e-06, "loss": 0.0045, "step": 11917 }, { "epoch": 0.8189939527212754, "grad_norm": 0.6328451633453369, "learning_rate": 2.504703559077981e-06, "loss": 0.0101, "step": 11918 }, { "epoch": 0.819062671797691, "grad_norm": 0.586467981338501, "learning_rate": 2.5028568283374836e-06, "loss": 0.0094, "step": 11919 }, { "epoch": 0.8191313908741067, "grad_norm": 0.6079575419425964, "learning_rate": 2.5010107166797615e-06, "loss": 0.0132, "step": 11920 }, { "epoch": 0.8192001099505223, "grad_norm": 0.654723048210144, "learning_rate": 2.4991652241962743e-06, "loss": 0.0083, "step": 11921 }, { "epoch": 0.8192688290269379, "grad_norm": 0.22245821356773376, "learning_rate": 2.497320350978434e-06, "loss": 0.0077, "step": 11922 }, { "epoch": 0.8193375481033535, "grad_norm": 0.3922485113143921, "learning_rate": 2.4954760971176483e-06, "loss": 0.0121, "step": 11923 }, { "epoch": 0.819406267179769, "grad_norm": 0.4072451889514923, "learning_rate": 2.4936324627052633e-06, "loss": 0.0132, "step": 11924 }, { "epoch": 0.8194749862561848, "grad_norm": 0.6087939739227295, "learning_rate": 2.4917894478326177e-06, "loss": 0.0076, "step": 11925 }, { "epoch": 0.8195437053326003, "grad_norm": 0.6432779431343079, "learning_rate": 2.489947052591005e-06, "loss": 0.0085, "step": 11926 }, { "epoch": 0.8196124244090159, "grad_norm": 0.6310824751853943, "learning_rate": 2.488105277071698e-06, "loss": 0.0152, "step": 11927 }, { "epoch": 0.8196811434854315, "grad_norm": 0.4712921380996704, "learning_rate": 2.4862641213659275e-06, "loss": 0.0159, "step": 11928 }, { "epoch": 0.8197498625618471, "grad_norm": 0.1586407870054245, "learning_rate": 2.484423585564907e-06, "loss": 0.0066, "step": 11929 }, { "epoch": 0.8198185816382628, "grad_norm": 0.23330549895763397, "learning_rate": 2.4825836697598153e-06, "loss": 0.0096, "step": 11930 }, { "epoch": 0.8198873007146784, "grad_norm": 0.49768632650375366, "learning_rate": 2.480744374041789e-06, "loss": 0.018, "step": 11931 }, { "epoch": 0.819956019791094, "grad_norm": 0.6281014084815979, "learning_rate": 2.478905698501948e-06, "loss": 0.0164, "step": 11932 }, { "epoch": 0.8200247388675096, "grad_norm": 0.30589789152145386, "learning_rate": 2.4770676432313798e-06, "loss": 0.0113, "step": 11933 }, { "epoch": 0.8200934579439252, "grad_norm": 0.45606762170791626, "learning_rate": 2.4752302083211313e-06, "loss": 0.0167, "step": 11934 }, { "epoch": 0.8201621770203409, "grad_norm": 0.4975365400314331, "learning_rate": 2.4733933938622283e-06, "loss": 0.01, "step": 11935 }, { "epoch": 0.8202308960967565, "grad_norm": 0.12114493548870087, "learning_rate": 2.4715571999456674e-06, "loss": 0.0046, "step": 11936 }, { "epoch": 0.8202996151731721, "grad_norm": 0.35592857003211975, "learning_rate": 2.469721626662402e-06, "loss": 0.0121, "step": 11937 }, { "epoch": 0.8203683342495877, "grad_norm": 0.37850815057754517, "learning_rate": 2.467886674103368e-06, "loss": 0.0124, "step": 11938 }, { "epoch": 0.8204370533260033, "grad_norm": 0.26833146810531616, "learning_rate": 2.46605234235946e-06, "loss": 0.0078, "step": 11939 }, { "epoch": 0.820505772402419, "grad_norm": 0.27578073740005493, "learning_rate": 2.4642186315215504e-06, "loss": 0.011, "step": 11940 }, { "epoch": 0.8205744914788345, "grad_norm": 0.419038325548172, "learning_rate": 2.462385541680478e-06, "loss": 0.0057, "step": 11941 }, { "epoch": 0.8206432105552501, "grad_norm": 0.3409339189529419, "learning_rate": 2.4605530729270537e-06, "loss": 0.0164, "step": 11942 }, { "epoch": 0.8207119296316657, "grad_norm": 0.369230717420578, "learning_rate": 2.4587212253520448e-06, "loss": 0.0114, "step": 11943 }, { "epoch": 0.8207806487080813, "grad_norm": 0.18715597689151764, "learning_rate": 2.456889999046207e-06, "loss": 0.008, "step": 11944 }, { "epoch": 0.820849367784497, "grad_norm": 0.18908949196338654, "learning_rate": 2.4550593941002475e-06, "loss": 0.0035, "step": 11945 }, { "epoch": 0.8209180868609126, "grad_norm": 0.2241092324256897, "learning_rate": 2.453229410604856e-06, "loss": 0.0093, "step": 11946 }, { "epoch": 0.8209868059373282, "grad_norm": 0.5435548424720764, "learning_rate": 2.451400048650686e-06, "loss": 0.0064, "step": 11947 }, { "epoch": 0.8210555250137438, "grad_norm": 0.28212547302246094, "learning_rate": 2.4495713083283556e-06, "loss": 0.0042, "step": 11948 }, { "epoch": 0.8211242440901594, "grad_norm": 0.612919270992279, "learning_rate": 2.4477431897284604e-06, "loss": 0.016, "step": 11949 }, { "epoch": 0.8211929631665751, "grad_norm": 0.23156462609767914, "learning_rate": 2.4459156929415643e-06, "loss": 0.0077, "step": 11950 }, { "epoch": 0.8212616822429907, "grad_norm": 0.19779175519943237, "learning_rate": 2.4440888180581907e-06, "loss": 0.0127, "step": 11951 }, { "epoch": 0.8213304013194063, "grad_norm": 0.3860393166542053, "learning_rate": 2.442262565168843e-06, "loss": 0.0058, "step": 11952 }, { "epoch": 0.8213991203958219, "grad_norm": 0.3005969226360321, "learning_rate": 2.440436934363994e-06, "loss": 0.0076, "step": 11953 }, { "epoch": 0.8214678394722374, "grad_norm": 0.3408506214618683, "learning_rate": 2.4386119257340754e-06, "loss": 0.0061, "step": 11954 }, { "epoch": 0.8215365585486532, "grad_norm": 0.4853295683860779, "learning_rate": 2.436787539369501e-06, "loss": 0.0154, "step": 11955 }, { "epoch": 0.8216052776250687, "grad_norm": 0.4114140570163727, "learning_rate": 2.4349637753606386e-06, "loss": 0.0165, "step": 11956 }, { "epoch": 0.8216739967014843, "grad_norm": 0.6257152557373047, "learning_rate": 2.4331406337978406e-06, "loss": 0.0118, "step": 11957 }, { "epoch": 0.8217427157778999, "grad_norm": 0.29393520951271057, "learning_rate": 2.431318114771414e-06, "loss": 0.0091, "step": 11958 }, { "epoch": 0.8218114348543155, "grad_norm": 0.4131087064743042, "learning_rate": 2.4294962183716533e-06, "loss": 0.0271, "step": 11959 }, { "epoch": 0.8218801539307312, "grad_norm": 0.2505446970462799, "learning_rate": 2.427674944688803e-06, "loss": 0.0043, "step": 11960 }, { "epoch": 0.8219488730071468, "grad_norm": 0.8028570413589478, "learning_rate": 2.425854293813092e-06, "loss": 0.0107, "step": 11961 }, { "epoch": 0.8220175920835624, "grad_norm": 0.43108633160591125, "learning_rate": 2.424034265834704e-06, "loss": 0.009, "step": 11962 }, { "epoch": 0.822086311159978, "grad_norm": 0.2806892693042755, "learning_rate": 2.4222148608438083e-06, "loss": 0.006, "step": 11963 }, { "epoch": 0.8221550302363936, "grad_norm": 0.24362823367118835, "learning_rate": 2.420396078930525e-06, "loss": 0.0106, "step": 11964 }, { "epoch": 0.8222237493128093, "grad_norm": 0.35629966855049133, "learning_rate": 2.4185779201849577e-06, "loss": 0.0125, "step": 11965 }, { "epoch": 0.8222924683892249, "grad_norm": 0.25311142206192017, "learning_rate": 2.4167603846971727e-06, "loss": 0.0099, "step": 11966 }, { "epoch": 0.8223611874656405, "grad_norm": 0.2630347013473511, "learning_rate": 2.414943472557214e-06, "loss": 0.0106, "step": 11967 }, { "epoch": 0.822429906542056, "grad_norm": 0.13073308765888214, "learning_rate": 2.4131271838550766e-06, "loss": 0.0135, "step": 11968 }, { "epoch": 0.8224986256184716, "grad_norm": 0.4869557023048401, "learning_rate": 2.4113115186807453e-06, "loss": 0.0101, "step": 11969 }, { "epoch": 0.8225673446948873, "grad_norm": 0.20273952186107635, "learning_rate": 2.4094964771241566e-06, "loss": 0.0048, "step": 11970 }, { "epoch": 0.8226360637713029, "grad_norm": 0.5246686339378357, "learning_rate": 2.407682059275227e-06, "loss": 0.0183, "step": 11971 }, { "epoch": 0.8227047828477185, "grad_norm": 0.47951430082321167, "learning_rate": 2.405868265223842e-06, "loss": 0.0126, "step": 11972 }, { "epoch": 0.8227735019241341, "grad_norm": 0.4576793313026428, "learning_rate": 2.4040550950598495e-06, "loss": 0.0186, "step": 11973 }, { "epoch": 0.8228422210005497, "grad_norm": 0.4087304472923279, "learning_rate": 2.402242548873073e-06, "loss": 0.006, "step": 11974 }, { "epoch": 0.8229109400769654, "grad_norm": 0.2458461970090866, "learning_rate": 2.4004306267532984e-06, "loss": 0.0067, "step": 11975 }, { "epoch": 0.822979659153381, "grad_norm": 0.17569851875305176, "learning_rate": 2.3986193287902864e-06, "loss": 0.0036, "step": 11976 }, { "epoch": 0.8230483782297966, "grad_norm": 0.49705687165260315, "learning_rate": 2.3968086550737654e-06, "loss": 0.0119, "step": 11977 }, { "epoch": 0.8231170973062122, "grad_norm": 0.2431900054216385, "learning_rate": 2.3949986056934374e-06, "loss": 0.0055, "step": 11978 }, { "epoch": 0.8231858163826278, "grad_norm": 0.10554991662502289, "learning_rate": 2.3931891807389587e-06, "loss": 0.002, "step": 11979 }, { "epoch": 0.8232545354590435, "grad_norm": 0.3632693886756897, "learning_rate": 2.391380380299973e-06, "loss": 0.0087, "step": 11980 }, { "epoch": 0.8233232545354591, "grad_norm": 0.35925111174583435, "learning_rate": 2.3895722044660777e-06, "loss": 0.0126, "step": 11981 }, { "epoch": 0.8233919736118747, "grad_norm": 0.264813631772995, "learning_rate": 2.3877646533268504e-06, "loss": 0.0098, "step": 11982 }, { "epoch": 0.8234606926882903, "grad_norm": 0.29115748405456543, "learning_rate": 2.385957726971831e-06, "loss": 0.0087, "step": 11983 }, { "epoch": 0.8235294117647058, "grad_norm": 0.2855355143547058, "learning_rate": 2.3841514254905374e-06, "loss": 0.0102, "step": 11984 }, { "epoch": 0.8235981308411215, "grad_norm": 0.23859746754169464, "learning_rate": 2.3823457489724417e-06, "loss": 0.0182, "step": 11985 }, { "epoch": 0.8236668499175371, "grad_norm": 0.44921600818634033, "learning_rate": 2.380540697506999e-06, "loss": 0.0084, "step": 11986 }, { "epoch": 0.8237355689939527, "grad_norm": 0.2914586365222931, "learning_rate": 2.3787362711836225e-06, "loss": 0.0094, "step": 11987 }, { "epoch": 0.8238042880703683, "grad_norm": 0.27903130650520325, "learning_rate": 2.3769324700917026e-06, "loss": 0.0056, "step": 11988 }, { "epoch": 0.8238730071467839, "grad_norm": 0.3914128839969635, "learning_rate": 2.3751292943206e-06, "loss": 0.0041, "step": 11989 }, { "epoch": 0.8239417262231996, "grad_norm": 0.08660057932138443, "learning_rate": 2.373326743959634e-06, "loss": 0.0054, "step": 11990 }, { "epoch": 0.8240104452996152, "grad_norm": 0.25931069254875183, "learning_rate": 2.3715248190981038e-06, "loss": 0.0087, "step": 11991 }, { "epoch": 0.8240791643760308, "grad_norm": 0.21709074079990387, "learning_rate": 2.369723519825269e-06, "loss": 0.0038, "step": 11992 }, { "epoch": 0.8241478834524464, "grad_norm": 0.9403628706932068, "learning_rate": 2.3679228462303674e-06, "loss": 0.0061, "step": 11993 }, { "epoch": 0.824216602528862, "grad_norm": 0.49213549494743347, "learning_rate": 2.3661227984025893e-06, "loss": 0.0138, "step": 11994 }, { "epoch": 0.8242853216052777, "grad_norm": 0.20628924667835236, "learning_rate": 2.3643233764311225e-06, "loss": 0.0082, "step": 11995 }, { "epoch": 0.8243540406816933, "grad_norm": 0.19564701616764069, "learning_rate": 2.3625245804050927e-06, "loss": 0.0047, "step": 11996 }, { "epoch": 0.8244227597581089, "grad_norm": 0.23409582674503326, "learning_rate": 2.360726410413617e-06, "loss": 0.0049, "step": 11997 }, { "epoch": 0.8244914788345244, "grad_norm": 0.9171900749206543, "learning_rate": 2.3589288665457665e-06, "loss": 0.0133, "step": 11998 }, { "epoch": 0.82456019791094, "grad_norm": 0.37446528673171997, "learning_rate": 2.3571319488905955e-06, "loss": 0.0104, "step": 11999 }, { "epoch": 0.8246289169873557, "grad_norm": 0.23716050386428833, "learning_rate": 2.355335657537107e-06, "loss": 0.0119, "step": 12000 }, { "epoch": 0.8246976360637713, "grad_norm": 0.2888587713241577, "learning_rate": 2.3535399925743013e-06, "loss": 0.0094, "step": 12001 }, { "epoch": 0.8247663551401869, "grad_norm": 0.37413719296455383, "learning_rate": 2.3517449540911218e-06, "loss": 0.0079, "step": 12002 }, { "epoch": 0.8248350742166025, "grad_norm": 0.722893476486206, "learning_rate": 2.3499505421764945e-06, "loss": 0.0075, "step": 12003 }, { "epoch": 0.8249037932930181, "grad_norm": 0.5353007912635803, "learning_rate": 2.348156756919308e-06, "loss": 0.0089, "step": 12004 }, { "epoch": 0.8249725123694338, "grad_norm": 0.4151449203491211, "learning_rate": 2.3463635984084283e-06, "loss": 0.0131, "step": 12005 }, { "epoch": 0.8250412314458494, "grad_norm": 0.8346454501152039, "learning_rate": 2.3445710667326787e-06, "loss": 0.0175, "step": 12006 }, { "epoch": 0.825109950522265, "grad_norm": 0.906040608882904, "learning_rate": 2.3427791619808586e-06, "loss": 0.0107, "step": 12007 }, { "epoch": 0.8251786695986806, "grad_norm": 0.21424244344234467, "learning_rate": 2.3409878842417405e-06, "loss": 0.0124, "step": 12008 }, { "epoch": 0.8252473886750962, "grad_norm": 0.2290688157081604, "learning_rate": 2.339197233604053e-06, "loss": 0.0143, "step": 12009 }, { "epoch": 0.8253161077515119, "grad_norm": 0.4756471514701843, "learning_rate": 2.3374072101565088e-06, "loss": 0.0127, "step": 12010 }, { "epoch": 0.8253848268279275, "grad_norm": 0.663055956363678, "learning_rate": 2.3356178139877732e-06, "loss": 0.0093, "step": 12011 }, { "epoch": 0.8254535459043431, "grad_norm": 0.2779981791973114, "learning_rate": 2.3338290451864957e-06, "loss": 0.0079, "step": 12012 }, { "epoch": 0.8255222649807586, "grad_norm": 0.9976113438606262, "learning_rate": 2.3320409038412872e-06, "loss": 0.0143, "step": 12013 }, { "epoch": 0.8255909840571742, "grad_norm": 0.5379055142402649, "learning_rate": 2.3302533900407313e-06, "loss": 0.0114, "step": 12014 }, { "epoch": 0.8256597031335899, "grad_norm": 0.46677032113075256, "learning_rate": 2.328466503873371e-06, "loss": 0.0062, "step": 12015 }, { "epoch": 0.8257284222100055, "grad_norm": 0.7060167789459229, "learning_rate": 2.326680245427732e-06, "loss": 0.015, "step": 12016 }, { "epoch": 0.8257971412864211, "grad_norm": 0.4170282781124115, "learning_rate": 2.3248946147922936e-06, "loss": 0.0094, "step": 12017 }, { "epoch": 0.8258658603628367, "grad_norm": 0.19765034317970276, "learning_rate": 2.323109612055518e-06, "loss": 0.0106, "step": 12018 }, { "epoch": 0.8259345794392523, "grad_norm": 0.35752734541893005, "learning_rate": 2.32132523730583e-06, "loss": 0.0084, "step": 12019 }, { "epoch": 0.8260032985156679, "grad_norm": 0.46479079127311707, "learning_rate": 2.319541490631628e-06, "loss": 0.0111, "step": 12020 }, { "epoch": 0.8260720175920836, "grad_norm": 0.5336742401123047, "learning_rate": 2.3177583721212657e-06, "loss": 0.012, "step": 12021 }, { "epoch": 0.8261407366684992, "grad_norm": 0.41752681136131287, "learning_rate": 2.315975881863084e-06, "loss": 0.0065, "step": 12022 }, { "epoch": 0.8262094557449148, "grad_norm": 0.5831420421600342, "learning_rate": 2.3141940199453767e-06, "loss": 0.0149, "step": 12023 }, { "epoch": 0.8262781748213304, "grad_norm": 1.4067274332046509, "learning_rate": 2.312412786456415e-06, "loss": 0.0204, "step": 12024 }, { "epoch": 0.826346893897746, "grad_norm": 0.19370479881763458, "learning_rate": 2.310632181484445e-06, "loss": 0.0104, "step": 12025 }, { "epoch": 0.8264156129741617, "grad_norm": 0.28468358516693115, "learning_rate": 2.3088522051176632e-06, "loss": 0.0103, "step": 12026 }, { "epoch": 0.8264843320505773, "grad_norm": 0.09789973497390747, "learning_rate": 2.3070728574442552e-06, "loss": 0.006, "step": 12027 }, { "epoch": 0.8265530511269928, "grad_norm": 0.31733429431915283, "learning_rate": 2.305294138552359e-06, "loss": 0.0105, "step": 12028 }, { "epoch": 0.8266217702034084, "grad_norm": 0.2681722342967987, "learning_rate": 2.303516048530096e-06, "loss": 0.0082, "step": 12029 }, { "epoch": 0.826690489279824, "grad_norm": 0.4573723375797272, "learning_rate": 2.301738587465539e-06, "loss": 0.0079, "step": 12030 }, { "epoch": 0.8267592083562397, "grad_norm": 0.14052018523216248, "learning_rate": 2.2999617554467505e-06, "loss": 0.003, "step": 12031 }, { "epoch": 0.8268279274326553, "grad_norm": 0.4017796814441681, "learning_rate": 2.298185552561744e-06, "loss": 0.0083, "step": 12032 }, { "epoch": 0.8268966465090709, "grad_norm": 0.5010396242141724, "learning_rate": 2.2964099788985153e-06, "loss": 0.0086, "step": 12033 }, { "epoch": 0.8269653655854865, "grad_norm": 0.43968212604522705, "learning_rate": 2.2946350345450147e-06, "loss": 0.0054, "step": 12034 }, { "epoch": 0.8270340846619021, "grad_norm": 0.32544034719467163, "learning_rate": 2.2928607195891775e-06, "loss": 0.0103, "step": 12035 }, { "epoch": 0.8271028037383178, "grad_norm": 0.48107829689979553, "learning_rate": 2.2910870341188876e-06, "loss": 0.0145, "step": 12036 }, { "epoch": 0.8271715228147334, "grad_norm": 0.2800135910511017, "learning_rate": 2.2893139782220263e-06, "loss": 0.008, "step": 12037 }, { "epoch": 0.827240241891149, "grad_norm": 0.5522658824920654, "learning_rate": 2.287541551986413e-06, "loss": 0.0091, "step": 12038 }, { "epoch": 0.8273089609675646, "grad_norm": 0.3637412190437317, "learning_rate": 2.285769755499859e-06, "loss": 0.01, "step": 12039 }, { "epoch": 0.8273776800439802, "grad_norm": 0.3094404637813568, "learning_rate": 2.2839985888501297e-06, "loss": 0.0042, "step": 12040 }, { "epoch": 0.8274463991203959, "grad_norm": 0.6476417779922485, "learning_rate": 2.2822280521249713e-06, "loss": 0.0145, "step": 12041 }, { "epoch": 0.8275151181968114, "grad_norm": 0.3986979126930237, "learning_rate": 2.2804581454120833e-06, "loss": 0.0073, "step": 12042 }, { "epoch": 0.827583837273227, "grad_norm": 0.23205527663230896, "learning_rate": 2.278688868799149e-06, "loss": 0.0086, "step": 12043 }, { "epoch": 0.8276525563496426, "grad_norm": 0.8067317008972168, "learning_rate": 2.276920222373819e-06, "loss": 0.0123, "step": 12044 }, { "epoch": 0.8277212754260582, "grad_norm": 0.21912741661071777, "learning_rate": 2.275152206223699e-06, "loss": 0.0062, "step": 12045 }, { "epoch": 0.8277899945024739, "grad_norm": 0.465888112783432, "learning_rate": 2.273384820436381e-06, "loss": 0.0083, "step": 12046 }, { "epoch": 0.8278587135788895, "grad_norm": 0.41757506132125854, "learning_rate": 2.27161806509941e-06, "loss": 0.0079, "step": 12047 }, { "epoch": 0.8279274326553051, "grad_norm": 0.5097485780715942, "learning_rate": 2.2698519403003128e-06, "loss": 0.0088, "step": 12048 }, { "epoch": 0.8279961517317207, "grad_norm": 0.3975870609283447, "learning_rate": 2.2680864461265786e-06, "loss": 0.0077, "step": 12049 }, { "epoch": 0.8280648708081363, "grad_norm": 0.16505001485347748, "learning_rate": 2.2663215826656687e-06, "loss": 0.0152, "step": 12050 }, { "epoch": 0.828133589884552, "grad_norm": 0.2868420481681824, "learning_rate": 2.2645573500050055e-06, "loss": 0.0066, "step": 12051 }, { "epoch": 0.8282023089609676, "grad_norm": 0.4200372099876404, "learning_rate": 2.262793748231991e-06, "loss": 0.01, "step": 12052 }, { "epoch": 0.8282710280373832, "grad_norm": 0.20571401715278625, "learning_rate": 2.2610307774339856e-06, "loss": 0.0065, "step": 12053 }, { "epoch": 0.8283397471137988, "grad_norm": 0.6429787874221802, "learning_rate": 2.2592684376983243e-06, "loss": 0.011, "step": 12054 }, { "epoch": 0.8284084661902144, "grad_norm": 0.20214685797691345, "learning_rate": 2.2575067291123104e-06, "loss": 0.0087, "step": 12055 }, { "epoch": 0.8284771852666301, "grad_norm": 0.37716490030288696, "learning_rate": 2.255745651763219e-06, "loss": 0.0113, "step": 12056 }, { "epoch": 0.8285459043430456, "grad_norm": 0.09326102584600449, "learning_rate": 2.2539852057382845e-06, "loss": 0.0119, "step": 12057 }, { "epoch": 0.8286146234194612, "grad_norm": 0.3997423052787781, "learning_rate": 2.252225391124721e-06, "loss": 0.0082, "step": 12058 }, { "epoch": 0.8286833424958768, "grad_norm": 0.28744757175445557, "learning_rate": 2.2504662080096995e-06, "loss": 0.0085, "step": 12059 }, { "epoch": 0.8287520615722924, "grad_norm": 0.2386256754398346, "learning_rate": 2.2487076564803714e-06, "loss": 0.0048, "step": 12060 }, { "epoch": 0.8288207806487081, "grad_norm": 0.3214719891548157, "learning_rate": 2.246949736623853e-06, "loss": 0.0088, "step": 12061 }, { "epoch": 0.8288894997251237, "grad_norm": 0.45597192645072937, "learning_rate": 2.245192448527222e-06, "loss": 0.0113, "step": 12062 }, { "epoch": 0.8289582188015393, "grad_norm": 0.3641814887523651, "learning_rate": 2.2434357922775373e-06, "loss": 0.0059, "step": 12063 }, { "epoch": 0.8290269378779549, "grad_norm": 0.4144116938114166, "learning_rate": 2.2416797679618144e-06, "loss": 0.0071, "step": 12064 }, { "epoch": 0.8290956569543705, "grad_norm": 0.7684323191642761, "learning_rate": 2.239924375667049e-06, "loss": 0.0131, "step": 12065 }, { "epoch": 0.8291643760307862, "grad_norm": 0.28816917538642883, "learning_rate": 2.2381696154801893e-06, "loss": 0.0078, "step": 12066 }, { "epoch": 0.8292330951072018, "grad_norm": 0.2934402823448181, "learning_rate": 2.2364154874881767e-06, "loss": 0.007, "step": 12067 }, { "epoch": 0.8293018141836174, "grad_norm": 0.27472636103630066, "learning_rate": 2.234661991777897e-06, "loss": 0.005, "step": 12068 }, { "epoch": 0.829370533260033, "grad_norm": 0.282313734292984, "learning_rate": 2.232909128436222e-06, "loss": 0.0112, "step": 12069 }, { "epoch": 0.8294392523364486, "grad_norm": 0.443368136882782, "learning_rate": 2.231156897549976e-06, "loss": 0.0136, "step": 12070 }, { "epoch": 0.8295079714128643, "grad_norm": 0.5808558464050293, "learning_rate": 2.2294052992059692e-06, "loss": 0.0139, "step": 12071 }, { "epoch": 0.8295766904892798, "grad_norm": 0.3392496407032013, "learning_rate": 2.2276543334909616e-06, "loss": 0.0101, "step": 12072 }, { "epoch": 0.8296454095656954, "grad_norm": 0.22879932820796967, "learning_rate": 2.225904000491707e-06, "loss": 0.0152, "step": 12073 }, { "epoch": 0.829714128642111, "grad_norm": 0.6461589336395264, "learning_rate": 2.224154300294901e-06, "loss": 0.0122, "step": 12074 }, { "epoch": 0.8297828477185266, "grad_norm": 0.06277560442686081, "learning_rate": 2.2224052329872296e-06, "loss": 0.0112, "step": 12075 }, { "epoch": 0.8298515667949423, "grad_norm": 0.20421701669692993, "learning_rate": 2.22065679865533e-06, "loss": 0.0111, "step": 12076 }, { "epoch": 0.8299202858713579, "grad_norm": 0.8932066559791565, "learning_rate": 2.218908997385823e-06, "loss": 0.0097, "step": 12077 }, { "epoch": 0.8299890049477735, "grad_norm": 0.33367156982421875, "learning_rate": 2.217161829265284e-06, "loss": 0.0089, "step": 12078 }, { "epoch": 0.8300577240241891, "grad_norm": 0.18358609080314636, "learning_rate": 2.2154152943802676e-06, "loss": 0.0071, "step": 12079 }, { "epoch": 0.8301264431006047, "grad_norm": 0.4404505789279938, "learning_rate": 2.2136693928172958e-06, "loss": 0.0072, "step": 12080 }, { "epoch": 0.8301951621770204, "grad_norm": 0.5232340693473816, "learning_rate": 2.2119241246628514e-06, "loss": 0.0088, "step": 12081 }, { "epoch": 0.830263881253436, "grad_norm": 0.6753098368644714, "learning_rate": 2.2101794900033983e-06, "loss": 0.0102, "step": 12082 }, { "epoch": 0.8303326003298516, "grad_norm": 0.2452061027288437, "learning_rate": 2.2084354889253563e-06, "loss": 0.0105, "step": 12083 }, { "epoch": 0.8304013194062672, "grad_norm": 0.2755034863948822, "learning_rate": 2.2066921215151202e-06, "loss": 0.0088, "step": 12084 }, { "epoch": 0.8304700384826827, "grad_norm": 0.5065986514091492, "learning_rate": 2.2049493878590553e-06, "loss": 0.0061, "step": 12085 }, { "epoch": 0.8305387575590985, "grad_norm": 0.8437554240226746, "learning_rate": 2.2032072880434974e-06, "loss": 0.01, "step": 12086 }, { "epoch": 0.830607476635514, "grad_norm": 0.09092814475297928, "learning_rate": 2.201465822154735e-06, "loss": 0.0064, "step": 12087 }, { "epoch": 0.8306761957119296, "grad_norm": 0.1208312064409256, "learning_rate": 2.1997249902790483e-06, "loss": 0.0084, "step": 12088 }, { "epoch": 0.8307449147883452, "grad_norm": 0.2250375747680664, "learning_rate": 2.1979847925026648e-06, "loss": 0.0065, "step": 12089 }, { "epoch": 0.8308136338647608, "grad_norm": 0.1872669756412506, "learning_rate": 2.1962452289117983e-06, "loss": 0.0101, "step": 12090 }, { "epoch": 0.8308823529411765, "grad_norm": 0.5111324787139893, "learning_rate": 2.194506299592614e-06, "loss": 0.0076, "step": 12091 }, { "epoch": 0.8309510720175921, "grad_norm": 0.16476070880889893, "learning_rate": 2.192768004631268e-06, "loss": 0.0061, "step": 12092 }, { "epoch": 0.8310197910940077, "grad_norm": 0.13692635297775269, "learning_rate": 2.1910303441138584e-06, "loss": 0.0052, "step": 12093 }, { "epoch": 0.8310885101704233, "grad_norm": 0.4916313588619232, "learning_rate": 2.189293318126476e-06, "loss": 0.0081, "step": 12094 }, { "epoch": 0.8311572292468389, "grad_norm": 0.3737350106239319, "learning_rate": 2.187556926755162e-06, "loss": 0.0132, "step": 12095 }, { "epoch": 0.8312259483232546, "grad_norm": 0.19674569368362427, "learning_rate": 2.18582117008594e-06, "loss": 0.0048, "step": 12096 }, { "epoch": 0.8312946673996702, "grad_norm": 0.31177788972854614, "learning_rate": 2.184086048204789e-06, "loss": 0.0076, "step": 12097 }, { "epoch": 0.8313633864760858, "grad_norm": 0.29753759503364563, "learning_rate": 2.182351561197666e-06, "loss": 0.0079, "step": 12098 }, { "epoch": 0.8314321055525014, "grad_norm": 0.5267885327339172, "learning_rate": 2.1806177091504997e-06, "loss": 0.0124, "step": 12099 }, { "epoch": 0.831500824628917, "grad_norm": 0.3493962287902832, "learning_rate": 2.1788844921491723e-06, "loss": 0.0093, "step": 12100 }, { "epoch": 0.8315695437053326, "grad_norm": 0.13516737520694733, "learning_rate": 2.1771519102795504e-06, "loss": 0.0063, "step": 12101 }, { "epoch": 0.8316382627817482, "grad_norm": 0.4458673298358917, "learning_rate": 2.175419963627459e-06, "loss": 0.0067, "step": 12102 }, { "epoch": 0.8317069818581638, "grad_norm": 0.4483608603477478, "learning_rate": 2.173688652278694e-06, "loss": 0.0112, "step": 12103 }, { "epoch": 0.8317757009345794, "grad_norm": 0.35756853222846985, "learning_rate": 2.171957976319026e-06, "loss": 0.0071, "step": 12104 }, { "epoch": 0.831844420010995, "grad_norm": 0.994254469871521, "learning_rate": 2.170227935834188e-06, "loss": 0.0132, "step": 12105 }, { "epoch": 0.8319131390874107, "grad_norm": 0.2598543167114258, "learning_rate": 2.168498530909878e-06, "loss": 0.0096, "step": 12106 }, { "epoch": 0.8319818581638263, "grad_norm": 0.14857250452041626, "learning_rate": 2.166769761631774e-06, "loss": 0.0089, "step": 12107 }, { "epoch": 0.8320505772402419, "grad_norm": 0.1374557465314865, "learning_rate": 2.1650416280855082e-06, "loss": 0.0087, "step": 12108 }, { "epoch": 0.8321192963166575, "grad_norm": 0.5187878012657166, "learning_rate": 2.1633141303566935e-06, "loss": 0.0118, "step": 12109 }, { "epoch": 0.8321880153930731, "grad_norm": 0.8044307827949524, "learning_rate": 2.161587268530905e-06, "loss": 0.0112, "step": 12110 }, { "epoch": 0.8322567344694888, "grad_norm": 0.44588959217071533, "learning_rate": 2.1598610426936922e-06, "loss": 0.0116, "step": 12111 }, { "epoch": 0.8323254535459044, "grad_norm": 0.3609510660171509, "learning_rate": 2.158135452930561e-06, "loss": 0.0074, "step": 12112 }, { "epoch": 0.83239417262232, "grad_norm": 0.5988863110542297, "learning_rate": 2.156410499327003e-06, "loss": 0.0084, "step": 12113 }, { "epoch": 0.8324628916987356, "grad_norm": 0.1796993762254715, "learning_rate": 2.154686181968458e-06, "loss": 0.0079, "step": 12114 }, { "epoch": 0.8325316107751511, "grad_norm": 0.21846485137939453, "learning_rate": 2.152962500940352e-06, "loss": 0.0053, "step": 12115 }, { "epoch": 0.8326003298515668, "grad_norm": 0.49459606409072876, "learning_rate": 2.151239456328074e-06, "loss": 0.0103, "step": 12116 }, { "epoch": 0.8326690489279824, "grad_norm": 0.46552035212516785, "learning_rate": 2.1495170482169746e-06, "loss": 0.011, "step": 12117 }, { "epoch": 0.832737768004398, "grad_norm": 0.35915014147758484, "learning_rate": 2.1477952766923836e-06, "loss": 0.0105, "step": 12118 }, { "epoch": 0.8328064870808136, "grad_norm": 0.8182192444801331, "learning_rate": 2.1460741418395894e-06, "loss": 0.0142, "step": 12119 }, { "epoch": 0.8328752061572292, "grad_norm": 0.5302290916442871, "learning_rate": 2.144353643743858e-06, "loss": 0.0091, "step": 12120 }, { "epoch": 0.8329439252336449, "grad_norm": 0.6983209252357483, "learning_rate": 2.1426337824904113e-06, "loss": 0.013, "step": 12121 }, { "epoch": 0.8330126443100605, "grad_norm": 0.09697701036930084, "learning_rate": 2.140914558164459e-06, "loss": 0.0071, "step": 12122 }, { "epoch": 0.8330813633864761, "grad_norm": 0.3302399814128876, "learning_rate": 2.1391959708511595e-06, "loss": 0.0106, "step": 12123 }, { "epoch": 0.8331500824628917, "grad_norm": 0.3274960517883301, "learning_rate": 2.1374780206356544e-06, "loss": 0.0072, "step": 12124 }, { "epoch": 0.8332188015393073, "grad_norm": 0.307056188583374, "learning_rate": 2.1357607076030407e-06, "loss": 0.0058, "step": 12125 }, { "epoch": 0.833287520615723, "grad_norm": 0.7105454206466675, "learning_rate": 2.134044031838397e-06, "loss": 0.0117, "step": 12126 }, { "epoch": 0.8333562396921386, "grad_norm": 0.5684340596199036, "learning_rate": 2.1323279934267535e-06, "loss": 0.0107, "step": 12127 }, { "epoch": 0.8334249587685542, "grad_norm": 0.5341896414756775, "learning_rate": 2.130612592453133e-06, "loss": 0.0107, "step": 12128 }, { "epoch": 0.8334936778449697, "grad_norm": 0.10788816958665848, "learning_rate": 2.128897829002505e-06, "loss": 0.0099, "step": 12129 }, { "epoch": 0.8335623969213853, "grad_norm": 0.36944326758384705, "learning_rate": 2.1271837031598178e-06, "loss": 0.0065, "step": 12130 }, { "epoch": 0.833631115997801, "grad_norm": 0.44443151354789734, "learning_rate": 2.1254702150099813e-06, "loss": 0.01, "step": 12131 }, { "epoch": 0.8336998350742166, "grad_norm": 0.45426082611083984, "learning_rate": 2.1237573646378855e-06, "loss": 0.0088, "step": 12132 }, { "epoch": 0.8337685541506322, "grad_norm": 0.2840079665184021, "learning_rate": 2.122045152128374e-06, "loss": 0.0125, "step": 12133 }, { "epoch": 0.8338372732270478, "grad_norm": 0.46091312170028687, "learning_rate": 2.120333577566269e-06, "loss": 0.0133, "step": 12134 }, { "epoch": 0.8339059923034634, "grad_norm": 0.034170493483543396, "learning_rate": 2.1186226410363646e-06, "loss": 0.0052, "step": 12135 }, { "epoch": 0.8339747113798791, "grad_norm": 0.5674079656600952, "learning_rate": 2.116912342623406e-06, "loss": 0.008, "step": 12136 }, { "epoch": 0.8340434304562947, "grad_norm": 0.23070932924747467, "learning_rate": 2.115202682412127e-06, "loss": 0.0052, "step": 12137 }, { "epoch": 0.8341121495327103, "grad_norm": 0.445430725812912, "learning_rate": 2.1134936604872155e-06, "loss": 0.0066, "step": 12138 }, { "epoch": 0.8341808686091259, "grad_norm": 0.527741551399231, "learning_rate": 2.1117852769333335e-06, "loss": 0.0128, "step": 12139 }, { "epoch": 0.8342495876855415, "grad_norm": 0.5189682841300964, "learning_rate": 2.1100775318351112e-06, "loss": 0.0077, "step": 12140 }, { "epoch": 0.8343183067619572, "grad_norm": 0.8747290372848511, "learning_rate": 2.1083704252771515e-06, "loss": 0.0086, "step": 12141 }, { "epoch": 0.8343870258383728, "grad_norm": 0.4867289066314697, "learning_rate": 2.106663957344013e-06, "loss": 0.0071, "step": 12142 }, { "epoch": 0.8344557449147884, "grad_norm": 0.4184374511241913, "learning_rate": 2.104958128120238e-06, "loss": 0.0112, "step": 12143 }, { "epoch": 0.834524463991204, "grad_norm": 0.17693763971328735, "learning_rate": 2.1032529376903226e-06, "loss": 0.0046, "step": 12144 }, { "epoch": 0.8345931830676195, "grad_norm": 0.2737603783607483, "learning_rate": 2.1015483861387424e-06, "loss": 0.0063, "step": 12145 }, { "epoch": 0.8346619021440352, "grad_norm": 0.1420619934797287, "learning_rate": 2.0998444735499366e-06, "loss": 0.0066, "step": 12146 }, { "epoch": 0.8347306212204508, "grad_norm": 0.4369572699069977, "learning_rate": 2.0981412000083177e-06, "loss": 0.011, "step": 12147 }, { "epoch": 0.8347993402968664, "grad_norm": 0.454643577337265, "learning_rate": 2.0964385655982564e-06, "loss": 0.009, "step": 12148 }, { "epoch": 0.834868059373282, "grad_norm": 0.47803568840026855, "learning_rate": 2.094736570404102e-06, "loss": 0.0117, "step": 12149 }, { "epoch": 0.8349367784496976, "grad_norm": 0.38781610131263733, "learning_rate": 2.093035214510162e-06, "loss": 0.0085, "step": 12150 }, { "epoch": 0.8350054975261132, "grad_norm": 0.289662629365921, "learning_rate": 2.0913344980007227e-06, "loss": 0.0092, "step": 12151 }, { "epoch": 0.8350742166025289, "grad_norm": 0.2512855529785156, "learning_rate": 2.0896344209600375e-06, "loss": 0.0062, "step": 12152 }, { "epoch": 0.8351429356789445, "grad_norm": 0.13753749430179596, "learning_rate": 2.0879349834723165e-06, "loss": 0.0102, "step": 12153 }, { "epoch": 0.8352116547553601, "grad_norm": 0.11378282308578491, "learning_rate": 2.086236185621754e-06, "loss": 0.0049, "step": 12154 }, { "epoch": 0.8352803738317757, "grad_norm": 0.15524110198020935, "learning_rate": 2.084538027492497e-06, "loss": 0.0051, "step": 12155 }, { "epoch": 0.8353490929081913, "grad_norm": 0.27045026421546936, "learning_rate": 2.082840509168672e-06, "loss": 0.0051, "step": 12156 }, { "epoch": 0.835417811984607, "grad_norm": 0.3092353641986847, "learning_rate": 2.0811436307343733e-06, "loss": 0.0178, "step": 12157 }, { "epoch": 0.8354865310610226, "grad_norm": 0.08119073510169983, "learning_rate": 2.0794473922736623e-06, "loss": 0.0078, "step": 12158 }, { "epoch": 0.8355552501374381, "grad_norm": 0.9607276320457458, "learning_rate": 2.0777517938705616e-06, "loss": 0.0085, "step": 12159 }, { "epoch": 0.8356239692138537, "grad_norm": 0.3532869219779968, "learning_rate": 2.076056835609072e-06, "loss": 0.0074, "step": 12160 }, { "epoch": 0.8356926882902693, "grad_norm": 0.4664154350757599, "learning_rate": 2.0743625175731536e-06, "loss": 0.0098, "step": 12161 }, { "epoch": 0.835761407366685, "grad_norm": 0.4079289138317108, "learning_rate": 2.0726688398467435e-06, "loss": 0.0187, "step": 12162 }, { "epoch": 0.8358301264431006, "grad_norm": 0.2137080729007721, "learning_rate": 2.070975802513737e-06, "loss": 0.01, "step": 12163 }, { "epoch": 0.8358988455195162, "grad_norm": 0.3324662446975708, "learning_rate": 2.069283405658015e-06, "loss": 0.0104, "step": 12164 }, { "epoch": 0.8359675645959318, "grad_norm": 0.4788472354412079, "learning_rate": 2.067591649363404e-06, "loss": 0.0058, "step": 12165 }, { "epoch": 0.8360362836723474, "grad_norm": 0.4570927023887634, "learning_rate": 2.065900533713718e-06, "loss": 0.013, "step": 12166 }, { "epoch": 0.8361050027487631, "grad_norm": 0.35534700751304626, "learning_rate": 2.0642100587927258e-06, "loss": 0.0061, "step": 12167 }, { "epoch": 0.8361737218251787, "grad_norm": 0.26672258973121643, "learning_rate": 2.0625202246841767e-06, "loss": 0.0091, "step": 12168 }, { "epoch": 0.8362424409015943, "grad_norm": 0.5430437326431274, "learning_rate": 2.060831031471772e-06, "loss": 0.0105, "step": 12169 }, { "epoch": 0.8363111599780099, "grad_norm": 0.37490078806877136, "learning_rate": 2.0591424792391982e-06, "loss": 0.0052, "step": 12170 }, { "epoch": 0.8363798790544255, "grad_norm": 0.1386977583169937, "learning_rate": 2.0574545680701e-06, "loss": 0.0047, "step": 12171 }, { "epoch": 0.8364485981308412, "grad_norm": 0.1931304633617401, "learning_rate": 2.055767298048096e-06, "loss": 0.0099, "step": 12172 }, { "epoch": 0.8365173172072567, "grad_norm": 0.5663473010063171, "learning_rate": 2.054080669256765e-06, "loss": 0.0113, "step": 12173 }, { "epoch": 0.8365860362836723, "grad_norm": 0.18683871626853943, "learning_rate": 2.0523946817796656e-06, "loss": 0.0084, "step": 12174 }, { "epoch": 0.8366547553600879, "grad_norm": 0.6165608763694763, "learning_rate": 2.0507093357003114e-06, "loss": 0.0097, "step": 12175 }, { "epoch": 0.8367234744365035, "grad_norm": 0.3992578685283661, "learning_rate": 2.0490246311021935e-06, "loss": 0.0066, "step": 12176 }, { "epoch": 0.8367921935129192, "grad_norm": 0.3278273642063141, "learning_rate": 2.047340568068774e-06, "loss": 0.0124, "step": 12177 }, { "epoch": 0.8368609125893348, "grad_norm": 0.3769926130771637, "learning_rate": 2.045657146683469e-06, "loss": 0.0097, "step": 12178 }, { "epoch": 0.8369296316657504, "grad_norm": 0.43029356002807617, "learning_rate": 2.0439743670296785e-06, "loss": 0.0097, "step": 12179 }, { "epoch": 0.836998350742166, "grad_norm": 0.23946240544319153, "learning_rate": 2.042292229190759e-06, "loss": 0.0087, "step": 12180 }, { "epoch": 0.8370670698185816, "grad_norm": 0.5291372537612915, "learning_rate": 2.040610733250041e-06, "loss": 0.0156, "step": 12181 }, { "epoch": 0.8371357888949973, "grad_norm": 0.45384979248046875, "learning_rate": 2.0389298792908264e-06, "loss": 0.0094, "step": 12182 }, { "epoch": 0.8372045079714129, "grad_norm": 0.38208356499671936, "learning_rate": 2.037249667396381e-06, "loss": 0.0073, "step": 12183 }, { "epoch": 0.8372732270478285, "grad_norm": 0.4371812045574188, "learning_rate": 2.035570097649932e-06, "loss": 0.012, "step": 12184 }, { "epoch": 0.8373419461242441, "grad_norm": 0.2863311469554901, "learning_rate": 2.0338911701346914e-06, "loss": 0.0096, "step": 12185 }, { "epoch": 0.8374106652006597, "grad_norm": 0.17295441031455994, "learning_rate": 2.0322128849338224e-06, "loss": 0.0071, "step": 12186 }, { "epoch": 0.8374793842770754, "grad_norm": 0.3046601116657257, "learning_rate": 2.030535242130465e-06, "loss": 0.0071, "step": 12187 }, { "epoch": 0.837548103353491, "grad_norm": 0.9415353536605835, "learning_rate": 2.0288582418077283e-06, "loss": 0.0139, "step": 12188 }, { "epoch": 0.8376168224299065, "grad_norm": 0.23377585411071777, "learning_rate": 2.0271818840486906e-06, "loss": 0.0139, "step": 12189 }, { "epoch": 0.8376855415063221, "grad_norm": 0.23642775416374207, "learning_rate": 2.0255061689363867e-06, "loss": 0.0103, "step": 12190 }, { "epoch": 0.8377542605827377, "grad_norm": 0.3557060658931732, "learning_rate": 2.0238310965538377e-06, "loss": 0.0132, "step": 12191 }, { "epoch": 0.8378229796591534, "grad_norm": 0.31256017088890076, "learning_rate": 2.022156666984013e-06, "loss": 0.0106, "step": 12192 }, { "epoch": 0.837891698735569, "grad_norm": 0.6784510016441345, "learning_rate": 2.020482880309867e-06, "loss": 0.0104, "step": 12193 }, { "epoch": 0.8379604178119846, "grad_norm": 0.45126456022262573, "learning_rate": 2.0188097366143176e-06, "loss": 0.0135, "step": 12194 }, { "epoch": 0.8380291368884002, "grad_norm": 0.4012160301208496, "learning_rate": 2.0171372359802426e-06, "loss": 0.0057, "step": 12195 }, { "epoch": 0.8380978559648158, "grad_norm": 0.30621328949928284, "learning_rate": 2.0154653784905003e-06, "loss": 0.0077, "step": 12196 }, { "epoch": 0.8381665750412315, "grad_norm": 0.4196527898311615, "learning_rate": 2.0137941642279045e-06, "loss": 0.0085, "step": 12197 }, { "epoch": 0.8382352941176471, "grad_norm": 0.14046767354011536, "learning_rate": 2.0121235932752507e-06, "loss": 0.007, "step": 12198 }, { "epoch": 0.8383040131940627, "grad_norm": 0.8114938735961914, "learning_rate": 2.0104536657152863e-06, "loss": 0.0179, "step": 12199 }, { "epoch": 0.8383727322704783, "grad_norm": 0.15168112516403198, "learning_rate": 2.008784381630748e-06, "loss": 0.0098, "step": 12200 }, { "epoch": 0.8384414513468939, "grad_norm": 0.31987446546554565, "learning_rate": 2.0071157411043177e-06, "loss": 0.0111, "step": 12201 }, { "epoch": 0.8385101704233096, "grad_norm": 0.7745029330253601, "learning_rate": 2.005447744218667e-06, "loss": 0.0213, "step": 12202 }, { "epoch": 0.8385788894997251, "grad_norm": 0.21808958053588867, "learning_rate": 2.0037803910564157e-06, "loss": 0.0065, "step": 12203 }, { "epoch": 0.8386476085761407, "grad_norm": 0.6837995052337646, "learning_rate": 2.002113681700168e-06, "loss": 0.0095, "step": 12204 }, { "epoch": 0.8387163276525563, "grad_norm": 0.5135723352432251, "learning_rate": 2.0004476162324774e-06, "loss": 0.0092, "step": 12205 }, { "epoch": 0.8387850467289719, "grad_norm": 0.42585334181785583, "learning_rate": 1.9987821947358953e-06, "loss": 0.015, "step": 12206 }, { "epoch": 0.8388537658053876, "grad_norm": 0.18239720165729523, "learning_rate": 1.9971174172929085e-06, "loss": 0.0083, "step": 12207 }, { "epoch": 0.8389224848818032, "grad_norm": 0.19312672317028046, "learning_rate": 1.995453283985996e-06, "loss": 0.0057, "step": 12208 }, { "epoch": 0.8389912039582188, "grad_norm": 0.6393465995788574, "learning_rate": 1.9937897948975893e-06, "loss": 0.0052, "step": 12209 }, { "epoch": 0.8390599230346344, "grad_norm": 0.2642441391944885, "learning_rate": 1.9921269501100996e-06, "loss": 0.0111, "step": 12210 }, { "epoch": 0.83912864211105, "grad_norm": 0.16502155363559723, "learning_rate": 1.9904647497058937e-06, "loss": 0.0052, "step": 12211 }, { "epoch": 0.8391973611874657, "grad_norm": 0.16529051959514618, "learning_rate": 1.988803193767317e-06, "loss": 0.0112, "step": 12212 }, { "epoch": 0.8392660802638813, "grad_norm": 0.5284433960914612, "learning_rate": 1.987142282376685e-06, "loss": 0.0075, "step": 12213 }, { "epoch": 0.8393347993402969, "grad_norm": 0.3423559069633484, "learning_rate": 1.9854820156162673e-06, "loss": 0.0108, "step": 12214 }, { "epoch": 0.8394035184167125, "grad_norm": 0.6848679780960083, "learning_rate": 1.9838223935683165e-06, "loss": 0.0077, "step": 12215 }, { "epoch": 0.839472237493128, "grad_norm": 0.5222849249839783, "learning_rate": 1.9821634163150413e-06, "loss": 0.0097, "step": 12216 }, { "epoch": 0.8395409565695438, "grad_norm": 0.3864537179470062, "learning_rate": 1.9805050839386267e-06, "loss": 0.013, "step": 12217 }, { "epoch": 0.8396096756459593, "grad_norm": 0.619316041469574, "learning_rate": 1.9788473965212234e-06, "loss": 0.0097, "step": 12218 }, { "epoch": 0.8396783947223749, "grad_norm": 0.3168436586856842, "learning_rate": 1.977190354144954e-06, "loss": 0.0086, "step": 12219 }, { "epoch": 0.8397471137987905, "grad_norm": 0.27273398637771606, "learning_rate": 1.9755339568918955e-06, "loss": 0.0088, "step": 12220 }, { "epoch": 0.8398158328752061, "grad_norm": 0.34202900528907776, "learning_rate": 1.973878204844112e-06, "loss": 0.0141, "step": 12221 }, { "epoch": 0.8398845519516218, "grad_norm": 0.4289775490760803, "learning_rate": 1.972223098083618e-06, "loss": 0.0046, "step": 12222 }, { "epoch": 0.8399532710280374, "grad_norm": 0.25112035870552063, "learning_rate": 1.970568636692409e-06, "loss": 0.0073, "step": 12223 }, { "epoch": 0.840021990104453, "grad_norm": 0.2888973355293274, "learning_rate": 1.968914820752441e-06, "loss": 0.0083, "step": 12224 }, { "epoch": 0.8400907091808686, "grad_norm": 1.3750368356704712, "learning_rate": 1.9672616503456447e-06, "loss": 0.0042, "step": 12225 }, { "epoch": 0.8401594282572842, "grad_norm": 0.4412050247192383, "learning_rate": 1.96560912555391e-06, "loss": 0.0135, "step": 12226 }, { "epoch": 0.8402281473336999, "grad_norm": 0.18572252988815308, "learning_rate": 1.963957246459106e-06, "loss": 0.0063, "step": 12227 }, { "epoch": 0.8402968664101155, "grad_norm": 1.0554214715957642, "learning_rate": 1.962306013143054e-06, "loss": 0.0126, "step": 12228 }, { "epoch": 0.8403655854865311, "grad_norm": 0.2574043273925781, "learning_rate": 1.9606554256875575e-06, "loss": 0.0065, "step": 12229 }, { "epoch": 0.8404343045629467, "grad_norm": 0.25279685854911804, "learning_rate": 1.959005484174387e-06, "loss": 0.01, "step": 12230 }, { "epoch": 0.8405030236393622, "grad_norm": 0.7771399021148682, "learning_rate": 1.957356188685271e-06, "loss": 0.0112, "step": 12231 }, { "epoch": 0.840571742715778, "grad_norm": 0.2097267210483551, "learning_rate": 1.9557075393019186e-06, "loss": 0.0077, "step": 12232 }, { "epoch": 0.8406404617921935, "grad_norm": 0.18252992630004883, "learning_rate": 1.9540595361059933e-06, "loss": 0.0112, "step": 12233 }, { "epoch": 0.8407091808686091, "grad_norm": 0.3029268980026245, "learning_rate": 1.9524121791791394e-06, "loss": 0.0056, "step": 12234 }, { "epoch": 0.8407778999450247, "grad_norm": 0.13854791224002838, "learning_rate": 1.9507654686029554e-06, "loss": 0.0088, "step": 12235 }, { "epoch": 0.8408466190214403, "grad_norm": 0.46160221099853516, "learning_rate": 1.949119404459029e-06, "loss": 0.0064, "step": 12236 }, { "epoch": 0.840915338097856, "grad_norm": 0.17427000403404236, "learning_rate": 1.947473986828893e-06, "loss": 0.0105, "step": 12237 }, { "epoch": 0.8409840571742716, "grad_norm": 0.27763602137565613, "learning_rate": 1.945829215794063e-06, "loss": 0.0079, "step": 12238 }, { "epoch": 0.8410527762506872, "grad_norm": 0.6080803871154785, "learning_rate": 1.944185091436012e-06, "loss": 0.0124, "step": 12239 }, { "epoch": 0.8411214953271028, "grad_norm": 0.2229430228471756, "learning_rate": 1.9425416138361944e-06, "loss": 0.0086, "step": 12240 }, { "epoch": 0.8411902144035184, "grad_norm": 0.33842459321022034, "learning_rate": 1.940898783076012e-06, "loss": 0.0072, "step": 12241 }, { "epoch": 0.8412589334799341, "grad_norm": 0.441832035779953, "learning_rate": 1.9392565992368647e-06, "loss": 0.0066, "step": 12242 }, { "epoch": 0.8413276525563497, "grad_norm": 0.20196503400802612, "learning_rate": 1.9376150624000876e-06, "loss": 0.006, "step": 12243 }, { "epoch": 0.8413963716327653, "grad_norm": 0.3504546880722046, "learning_rate": 1.9359741726470104e-06, "loss": 0.0082, "step": 12244 }, { "epoch": 0.8414650907091809, "grad_norm": 0.07733214646577835, "learning_rate": 1.9343339300589098e-06, "loss": 0.0044, "step": 12245 }, { "epoch": 0.8415338097855964, "grad_norm": 0.6051375269889832, "learning_rate": 1.932694334717049e-06, "loss": 0.01, "step": 12246 }, { "epoch": 0.8416025288620121, "grad_norm": 0.48963677883148193, "learning_rate": 1.9310553867026415e-06, "loss": 0.0099, "step": 12247 }, { "epoch": 0.8416712479384277, "grad_norm": 0.515716016292572, "learning_rate": 1.9294170860968807e-06, "loss": 0.0095, "step": 12248 }, { "epoch": 0.8417399670148433, "grad_norm": 0.3045530915260315, "learning_rate": 1.92777943298093e-06, "loss": 0.0026, "step": 12249 }, { "epoch": 0.8418086860912589, "grad_norm": 0.2368389517068863, "learning_rate": 1.926142427435908e-06, "loss": 0.0088, "step": 12250 }, { "epoch": 0.8418774051676745, "grad_norm": 0.33263063430786133, "learning_rate": 1.924506069542913e-06, "loss": 0.0104, "step": 12251 }, { "epoch": 0.8419461242440902, "grad_norm": 0.6204851865768433, "learning_rate": 1.9228703593830033e-06, "loss": 0.01, "step": 12252 }, { "epoch": 0.8420148433205058, "grad_norm": 0.20709377527236938, "learning_rate": 1.9212352970372093e-06, "loss": 0.0089, "step": 12253 }, { "epoch": 0.8420835623969214, "grad_norm": 0.1911853402853012, "learning_rate": 1.919600882586531e-06, "loss": 0.0125, "step": 12254 }, { "epoch": 0.842152281473337, "grad_norm": 0.3700016438961029, "learning_rate": 1.917967116111936e-06, "loss": 0.0083, "step": 12255 }, { "epoch": 0.8422210005497526, "grad_norm": 0.1944735050201416, "learning_rate": 1.9163339976943508e-06, "loss": 0.0112, "step": 12256 }, { "epoch": 0.8422897196261683, "grad_norm": 0.28900495171546936, "learning_rate": 1.9147015274146834e-06, "loss": 0.0078, "step": 12257 }, { "epoch": 0.8423584387025839, "grad_norm": 0.5270363092422485, "learning_rate": 1.913069705353798e-06, "loss": 0.0178, "step": 12258 }, { "epoch": 0.8424271577789995, "grad_norm": 1.1433312892913818, "learning_rate": 1.9114385315925336e-06, "loss": 0.0181, "step": 12259 }, { "epoch": 0.842495876855415, "grad_norm": 0.39319121837615967, "learning_rate": 1.909808006211695e-06, "loss": 0.0081, "step": 12260 }, { "epoch": 0.8425645959318306, "grad_norm": 0.37447357177734375, "learning_rate": 1.908178129292058e-06, "loss": 0.0043, "step": 12261 }, { "epoch": 0.8426333150082463, "grad_norm": 0.3186081647872925, "learning_rate": 1.9065489009143594e-06, "loss": 0.0073, "step": 12262 }, { "epoch": 0.8427020340846619, "grad_norm": 0.40562158823013306, "learning_rate": 1.9049203211593113e-06, "loss": 0.0129, "step": 12263 }, { "epoch": 0.8427707531610775, "grad_norm": 0.14571578800678253, "learning_rate": 1.9032923901075844e-06, "loss": 0.0109, "step": 12264 }, { "epoch": 0.8428394722374931, "grad_norm": 0.5239770412445068, "learning_rate": 1.9016651078398273e-06, "loss": 0.0234, "step": 12265 }, { "epoch": 0.8429081913139087, "grad_norm": 0.18579359352588654, "learning_rate": 1.9000384744366555e-06, "loss": 0.0082, "step": 12266 }, { "epoch": 0.8429769103903244, "grad_norm": 0.27459821105003357, "learning_rate": 1.8984124899786415e-06, "loss": 0.0106, "step": 12267 }, { "epoch": 0.84304562946674, "grad_norm": 0.29900118708610535, "learning_rate": 1.8967871545463405e-06, "loss": 0.0071, "step": 12268 }, { "epoch": 0.8431143485431556, "grad_norm": 0.4166187644004822, "learning_rate": 1.8951624682202617e-06, "loss": 0.0098, "step": 12269 }, { "epoch": 0.8431830676195712, "grad_norm": 0.3118983209133148, "learning_rate": 1.8935384310808956e-06, "loss": 0.0101, "step": 12270 }, { "epoch": 0.8432517866959868, "grad_norm": 0.30162230134010315, "learning_rate": 1.8919150432086813e-06, "loss": 0.01, "step": 12271 }, { "epoch": 0.8433205057724025, "grad_norm": 0.2382272183895111, "learning_rate": 1.890292304684056e-06, "loss": 0.0082, "step": 12272 }, { "epoch": 0.8433892248488181, "grad_norm": 0.4583856165409088, "learning_rate": 1.888670215587392e-06, "loss": 0.0097, "step": 12273 }, { "epoch": 0.8434579439252337, "grad_norm": 0.330640971660614, "learning_rate": 1.8870487759990534e-06, "loss": 0.0085, "step": 12274 }, { "epoch": 0.8435266630016492, "grad_norm": 0.11798464506864548, "learning_rate": 1.8854279859993561e-06, "loss": 0.0049, "step": 12275 }, { "epoch": 0.8435953820780648, "grad_norm": 0.09533250331878662, "learning_rate": 1.8838078456685975e-06, "loss": 0.0074, "step": 12276 }, { "epoch": 0.8436641011544805, "grad_norm": 0.371036559343338, "learning_rate": 1.8821883550870267e-06, "loss": 0.0059, "step": 12277 }, { "epoch": 0.8437328202308961, "grad_norm": 0.24796907603740692, "learning_rate": 1.880569514334881e-06, "loss": 0.0189, "step": 12278 }, { "epoch": 0.8438015393073117, "grad_norm": 0.3260277509689331, "learning_rate": 1.8789513234923466e-06, "loss": 0.0058, "step": 12279 }, { "epoch": 0.8438702583837273, "grad_norm": 0.33224159479141235, "learning_rate": 1.877333782639591e-06, "loss": 0.0124, "step": 12280 }, { "epoch": 0.8439389774601429, "grad_norm": 0.2425028681755066, "learning_rate": 1.8757168918567381e-06, "loss": 0.0119, "step": 12281 }, { "epoch": 0.8440076965365585, "grad_norm": 0.17567192018032074, "learning_rate": 1.8741006512238906e-06, "loss": 0.0126, "step": 12282 }, { "epoch": 0.8440764156129742, "grad_norm": 0.36291903257369995, "learning_rate": 1.8724850608211081e-06, "loss": 0.0059, "step": 12283 }, { "epoch": 0.8441451346893898, "grad_norm": 0.3372202515602112, "learning_rate": 1.8708701207284263e-06, "loss": 0.0043, "step": 12284 }, { "epoch": 0.8442138537658054, "grad_norm": 0.316849946975708, "learning_rate": 1.8692558310258496e-06, "loss": 0.0092, "step": 12285 }, { "epoch": 0.844282572842221, "grad_norm": 0.42391520738601685, "learning_rate": 1.8676421917933406e-06, "loss": 0.0125, "step": 12286 }, { "epoch": 0.8443512919186366, "grad_norm": 0.5043982863426208, "learning_rate": 1.8660292031108423e-06, "loss": 0.0114, "step": 12287 }, { "epoch": 0.8444200109950523, "grad_norm": 0.36914145946502686, "learning_rate": 1.8644168650582504e-06, "loss": 0.0134, "step": 12288 }, { "epoch": 0.8444887300714679, "grad_norm": 0.19208231568336487, "learning_rate": 1.8628051777154415e-06, "loss": 0.0069, "step": 12289 }, { "epoch": 0.8445574491478834, "grad_norm": 0.5566819310188293, "learning_rate": 1.8611941411622562e-06, "loss": 0.007, "step": 12290 }, { "epoch": 0.844626168224299, "grad_norm": 0.6078192591667175, "learning_rate": 1.8595837554785029e-06, "loss": 0.0107, "step": 12291 }, { "epoch": 0.8446948873007146, "grad_norm": 0.17618584632873535, "learning_rate": 1.8579740207439522e-06, "loss": 0.0047, "step": 12292 }, { "epoch": 0.8447636063771303, "grad_norm": 0.2504737079143524, "learning_rate": 1.8563649370383523e-06, "loss": 0.004, "step": 12293 }, { "epoch": 0.8448323254535459, "grad_norm": 0.3529260754585266, "learning_rate": 1.8547565044414094e-06, "loss": 0.0049, "step": 12294 }, { "epoch": 0.8449010445299615, "grad_norm": 0.5212513208389282, "learning_rate": 1.8531487230328032e-06, "loss": 0.0084, "step": 12295 }, { "epoch": 0.8449697636063771, "grad_norm": 0.49724358320236206, "learning_rate": 1.8515415928921797e-06, "loss": 0.007, "step": 12296 }, { "epoch": 0.8450384826827927, "grad_norm": 0.08647368103265762, "learning_rate": 1.8499351140991572e-06, "loss": 0.0047, "step": 12297 }, { "epoch": 0.8451072017592084, "grad_norm": 0.12002862244844437, "learning_rate": 1.8483292867333119e-06, "loss": 0.0103, "step": 12298 }, { "epoch": 0.845175920835624, "grad_norm": 0.29763272404670715, "learning_rate": 1.8467241108741967e-06, "loss": 0.009, "step": 12299 }, { "epoch": 0.8452446399120396, "grad_norm": 0.18670958280563354, "learning_rate": 1.8451195866013231e-06, "loss": 0.011, "step": 12300 }, { "epoch": 0.8453133589884552, "grad_norm": 0.237504243850708, "learning_rate": 1.8435157139941811e-06, "loss": 0.0064, "step": 12301 }, { "epoch": 0.8453820780648708, "grad_norm": 0.17124034464359283, "learning_rate": 1.841912493132225e-06, "loss": 0.0114, "step": 12302 }, { "epoch": 0.8454507971412865, "grad_norm": 0.494829922914505, "learning_rate": 1.84030992409487e-06, "loss": 0.0049, "step": 12303 }, { "epoch": 0.845519516217702, "grad_norm": 0.3887522220611572, "learning_rate": 1.8387080069615074e-06, "loss": 0.009, "step": 12304 }, { "epoch": 0.8455882352941176, "grad_norm": 0.43084079027175903, "learning_rate": 1.8371067418114905e-06, "loss": 0.0141, "step": 12305 }, { "epoch": 0.8456569543705332, "grad_norm": 0.5947561264038086, "learning_rate": 1.8355061287241459e-06, "loss": 0.0136, "step": 12306 }, { "epoch": 0.8457256734469488, "grad_norm": 0.37557363510131836, "learning_rate": 1.8339061677787565e-06, "loss": 0.0119, "step": 12307 }, { "epoch": 0.8457943925233645, "grad_norm": 0.17341460287570953, "learning_rate": 1.8323068590545928e-06, "loss": 0.0127, "step": 12308 }, { "epoch": 0.8458631115997801, "grad_norm": 0.21828988194465637, "learning_rate": 1.8307082026308742e-06, "loss": 0.0062, "step": 12309 }, { "epoch": 0.8459318306761957, "grad_norm": 0.1693117469549179, "learning_rate": 1.8291101985867975e-06, "loss": 0.0072, "step": 12310 }, { "epoch": 0.8460005497526113, "grad_norm": 0.3999119997024536, "learning_rate": 1.8275128470015213e-06, "loss": 0.0103, "step": 12311 }, { "epoch": 0.8460692688290269, "grad_norm": 0.22014440596103668, "learning_rate": 1.8259161479541803e-06, "loss": 0.0158, "step": 12312 }, { "epoch": 0.8461379879054426, "grad_norm": 0.5065652132034302, "learning_rate": 1.8243201015238647e-06, "loss": 0.0059, "step": 12313 }, { "epoch": 0.8462067069818582, "grad_norm": 0.6965992450714111, "learning_rate": 1.8227247077896446e-06, "loss": 0.014, "step": 12314 }, { "epoch": 0.8462754260582738, "grad_norm": 0.20681223273277283, "learning_rate": 1.8211299668305504e-06, "loss": 0.0089, "step": 12315 }, { "epoch": 0.8463441451346894, "grad_norm": 0.35581541061401367, "learning_rate": 1.8195358787255867e-06, "loss": 0.0082, "step": 12316 }, { "epoch": 0.846412864211105, "grad_norm": 0.3224051296710968, "learning_rate": 1.817942443553714e-06, "loss": 0.0047, "step": 12317 }, { "epoch": 0.8464815832875207, "grad_norm": 0.15742053091526031, "learning_rate": 1.816349661393874e-06, "loss": 0.0064, "step": 12318 }, { "epoch": 0.8465503023639362, "grad_norm": 0.1567104458808899, "learning_rate": 1.8147575323249655e-06, "loss": 0.007, "step": 12319 }, { "epoch": 0.8466190214403518, "grad_norm": 0.4456072449684143, "learning_rate": 1.8131660564258602e-06, "loss": 0.0096, "step": 12320 }, { "epoch": 0.8466877405167674, "grad_norm": 0.34216398000717163, "learning_rate": 1.8115752337754e-06, "loss": 0.0104, "step": 12321 }, { "epoch": 0.846756459593183, "grad_norm": 0.2823280394077301, "learning_rate": 1.8099850644523868e-06, "loss": 0.0066, "step": 12322 }, { "epoch": 0.8468251786695987, "grad_norm": 1.647424340248108, "learning_rate": 1.808395548535598e-06, "loss": 0.0145, "step": 12323 }, { "epoch": 0.8468938977460143, "grad_norm": 0.493647038936615, "learning_rate": 1.8068066861037701e-06, "loss": 0.0157, "step": 12324 }, { "epoch": 0.8469626168224299, "grad_norm": 0.3746022880077362, "learning_rate": 1.805218477235619e-06, "loss": 0.006, "step": 12325 }, { "epoch": 0.8470313358988455, "grad_norm": 0.24580524861812592, "learning_rate": 1.8036309220098096e-06, "loss": 0.0091, "step": 12326 }, { "epoch": 0.8471000549752611, "grad_norm": 0.48132678866386414, "learning_rate": 1.802044020505001e-06, "loss": 0.0081, "step": 12327 }, { "epoch": 0.8471687740516768, "grad_norm": 1.334751009941101, "learning_rate": 1.800457772799795e-06, "loss": 0.0056, "step": 12328 }, { "epoch": 0.8472374931280924, "grad_norm": 0.20546776056289673, "learning_rate": 1.7988721789727758e-06, "loss": 0.0077, "step": 12329 }, { "epoch": 0.847306212204508, "grad_norm": 0.25110387802124023, "learning_rate": 1.7972872391024852e-06, "loss": 0.0063, "step": 12330 }, { "epoch": 0.8473749312809236, "grad_norm": 0.2614049017429352, "learning_rate": 1.7957029532674456e-06, "loss": 0.0072, "step": 12331 }, { "epoch": 0.8474436503573392, "grad_norm": 0.15386272966861725, "learning_rate": 1.7941193215461277e-06, "loss": 0.0118, "step": 12332 }, { "epoch": 0.8475123694337549, "grad_norm": 0.25614234805107117, "learning_rate": 1.792536344016995e-06, "loss": 0.0053, "step": 12333 }, { "epoch": 0.8475810885101704, "grad_norm": 1.27078378200531, "learning_rate": 1.790954020758455e-06, "loss": 0.0133, "step": 12334 }, { "epoch": 0.847649807586586, "grad_norm": 0.5865470170974731, "learning_rate": 1.7893723518488985e-06, "loss": 0.0112, "step": 12335 }, { "epoch": 0.8477185266630016, "grad_norm": 0.48602157831192017, "learning_rate": 1.7877913373666743e-06, "loss": 0.0068, "step": 12336 }, { "epoch": 0.8477872457394172, "grad_norm": 0.2873017489910126, "learning_rate": 1.7862109773901064e-06, "loss": 0.0154, "step": 12337 }, { "epoch": 0.8478559648158329, "grad_norm": 0.47788965702056885, "learning_rate": 1.7846312719974773e-06, "loss": 0.0065, "step": 12338 }, { "epoch": 0.8479246838922485, "grad_norm": 0.4837457239627838, "learning_rate": 1.783052221267046e-06, "loss": 0.0082, "step": 12339 }, { "epoch": 0.8479934029686641, "grad_norm": 0.4003351926803589, "learning_rate": 1.7814738252770363e-06, "loss": 0.0074, "step": 12340 }, { "epoch": 0.8480621220450797, "grad_norm": 0.12511883676052094, "learning_rate": 1.7798960841056344e-06, "loss": 0.0071, "step": 12341 }, { "epoch": 0.8481308411214953, "grad_norm": 0.5136227607727051, "learning_rate": 1.778318997831006e-06, "loss": 0.0088, "step": 12342 }, { "epoch": 0.848199560197911, "grad_norm": 0.3887403905391693, "learning_rate": 1.776742566531267e-06, "loss": 0.0059, "step": 12343 }, { "epoch": 0.8482682792743266, "grad_norm": 0.27913936972618103, "learning_rate": 1.7751667902845164e-06, "loss": 0.0038, "step": 12344 }, { "epoch": 0.8483369983507422, "grad_norm": 0.2564239203929901, "learning_rate": 1.773591669168812e-06, "loss": 0.0071, "step": 12345 }, { "epoch": 0.8484057174271578, "grad_norm": 0.17288601398468018, "learning_rate": 1.7720172032621895e-06, "loss": 0.0057, "step": 12346 }, { "epoch": 0.8484744365035733, "grad_norm": 0.2594221532344818, "learning_rate": 1.770443392642635e-06, "loss": 0.0101, "step": 12347 }, { "epoch": 0.848543155579989, "grad_norm": 0.37700244784355164, "learning_rate": 1.7688702373881194e-06, "loss": 0.0106, "step": 12348 }, { "epoch": 0.8486118746564046, "grad_norm": 0.2620767652988434, "learning_rate": 1.7672977375765653e-06, "loss": 0.0058, "step": 12349 }, { "epoch": 0.8486805937328202, "grad_norm": 0.362438440322876, "learning_rate": 1.7657258932858788e-06, "loss": 0.006, "step": 12350 }, { "epoch": 0.8487493128092358, "grad_norm": 0.5217657685279846, "learning_rate": 1.764154704593921e-06, "loss": 0.0093, "step": 12351 }, { "epoch": 0.8488180318856514, "grad_norm": 0.3944885730743408, "learning_rate": 1.7625841715785313e-06, "loss": 0.0087, "step": 12352 }, { "epoch": 0.8488867509620671, "grad_norm": 0.1910446584224701, "learning_rate": 1.7610142943175039e-06, "loss": 0.0065, "step": 12353 }, { "epoch": 0.8489554700384827, "grad_norm": 0.16684146225452423, "learning_rate": 1.7594450728886136e-06, "loss": 0.004, "step": 12354 }, { "epoch": 0.8490241891148983, "grad_norm": 0.33402174711227417, "learning_rate": 1.757876507369588e-06, "loss": 0.0151, "step": 12355 }, { "epoch": 0.8490929081913139, "grad_norm": 0.3887512981891632, "learning_rate": 1.7563085978381365e-06, "loss": 0.0088, "step": 12356 }, { "epoch": 0.8491616272677295, "grad_norm": 0.2920181453227997, "learning_rate": 1.754741344371934e-06, "loss": 0.0057, "step": 12357 }, { "epoch": 0.8492303463441452, "grad_norm": 0.21233420073986053, "learning_rate": 1.7531747470486082e-06, "loss": 0.0063, "step": 12358 }, { "epoch": 0.8492990654205608, "grad_norm": 0.30468663573265076, "learning_rate": 1.7516088059457768e-06, "loss": 0.0075, "step": 12359 }, { "epoch": 0.8493677844969764, "grad_norm": 0.34494853019714355, "learning_rate": 1.7500435211410032e-06, "loss": 0.0057, "step": 12360 }, { "epoch": 0.849436503573392, "grad_norm": 0.6349226832389832, "learning_rate": 1.7484788927118333e-06, "loss": 0.0065, "step": 12361 }, { "epoch": 0.8495052226498075, "grad_norm": 0.1753288209438324, "learning_rate": 1.7469149207357738e-06, "loss": 0.0063, "step": 12362 }, { "epoch": 0.8495739417262232, "grad_norm": 0.3449757695198059, "learning_rate": 1.7453516052903057e-06, "loss": 0.0078, "step": 12363 }, { "epoch": 0.8496426608026388, "grad_norm": 0.5045890808105469, "learning_rate": 1.7437889464528673e-06, "loss": 0.0097, "step": 12364 }, { "epoch": 0.8497113798790544, "grad_norm": 0.2939526438713074, "learning_rate": 1.7422269443008715e-06, "loss": 0.0065, "step": 12365 }, { "epoch": 0.84978009895547, "grad_norm": 0.13615483045578003, "learning_rate": 1.7406655989116949e-06, "loss": 0.004, "step": 12366 }, { "epoch": 0.8498488180318856, "grad_norm": 0.632976233959198, "learning_rate": 1.7391049103626856e-06, "loss": 0.0139, "step": 12367 }, { "epoch": 0.8499175371083013, "grad_norm": 0.4829828441143036, "learning_rate": 1.7375448787311515e-06, "loss": 0.0063, "step": 12368 }, { "epoch": 0.8499862561847169, "grad_norm": 0.3470917344093323, "learning_rate": 1.7359855040943829e-06, "loss": 0.0077, "step": 12369 }, { "epoch": 0.8500549752611325, "grad_norm": 0.40925800800323486, "learning_rate": 1.7344267865296193e-06, "loss": 0.0051, "step": 12370 }, { "epoch": 0.8501236943375481, "grad_norm": 0.54404217004776, "learning_rate": 1.7328687261140841e-06, "loss": 0.0096, "step": 12371 }, { "epoch": 0.8501924134139637, "grad_norm": 0.2886030673980713, "learning_rate": 1.7313113229249522e-06, "loss": 0.01, "step": 12372 }, { "epoch": 0.8502611324903794, "grad_norm": 0.5389566421508789, "learning_rate": 1.729754577039382e-06, "loss": 0.0112, "step": 12373 }, { "epoch": 0.850329851566795, "grad_norm": 0.4716413617134094, "learning_rate": 1.7281984885344831e-06, "loss": 0.0105, "step": 12374 }, { "epoch": 0.8503985706432106, "grad_norm": 0.25222498178482056, "learning_rate": 1.726643057487346e-06, "loss": 0.0047, "step": 12375 }, { "epoch": 0.8504672897196262, "grad_norm": 0.5530714988708496, "learning_rate": 1.725088283975027e-06, "loss": 0.0081, "step": 12376 }, { "epoch": 0.8505360087960417, "grad_norm": 0.6439020037651062, "learning_rate": 1.7235341680745382e-06, "loss": 0.0143, "step": 12377 }, { "epoch": 0.8506047278724574, "grad_norm": 0.32231301069259644, "learning_rate": 1.7219807098628727e-06, "loss": 0.007, "step": 12378 }, { "epoch": 0.850673446948873, "grad_norm": 0.3310530483722687, "learning_rate": 1.7204279094169856e-06, "loss": 0.0135, "step": 12379 }, { "epoch": 0.8507421660252886, "grad_norm": 0.13182556629180908, "learning_rate": 1.7188757668137955e-06, "loss": 0.006, "step": 12380 }, { "epoch": 0.8508108851017042, "grad_norm": 0.2566964328289032, "learning_rate": 1.717324282130196e-06, "loss": 0.0079, "step": 12381 }, { "epoch": 0.8508796041781198, "grad_norm": 0.08997149765491486, "learning_rate": 1.715773455443047e-06, "loss": 0.0062, "step": 12382 }, { "epoch": 0.8509483232545355, "grad_norm": 0.2044503092765808, "learning_rate": 1.7142232868291656e-06, "loss": 0.0175, "step": 12383 }, { "epoch": 0.8510170423309511, "grad_norm": 0.29163748025894165, "learning_rate": 1.712673776365352e-06, "loss": 0.0053, "step": 12384 }, { "epoch": 0.8510857614073667, "grad_norm": 0.4226992428302765, "learning_rate": 1.7111249241283583e-06, "loss": 0.0106, "step": 12385 }, { "epoch": 0.8511544804837823, "grad_norm": 0.29622548818588257, "learning_rate": 1.709576730194916e-06, "loss": 0.0092, "step": 12386 }, { "epoch": 0.8512231995601979, "grad_norm": 0.6086899042129517, "learning_rate": 1.7080291946417177e-06, "loss": 0.0118, "step": 12387 }, { "epoch": 0.8512919186366136, "grad_norm": 0.3488825857639313, "learning_rate": 1.706482317545428e-06, "loss": 0.0043, "step": 12388 }, { "epoch": 0.8513606377130292, "grad_norm": 0.4566682279109955, "learning_rate": 1.7049360989826728e-06, "loss": 0.0106, "step": 12389 }, { "epoch": 0.8514293567894448, "grad_norm": 0.25598815083503723, "learning_rate": 1.7033905390300525e-06, "loss": 0.0062, "step": 12390 }, { "epoch": 0.8514980758658603, "grad_norm": 0.4775940477848053, "learning_rate": 1.7018456377641239e-06, "loss": 0.008, "step": 12391 }, { "epoch": 0.8515667949422759, "grad_norm": 0.2405065894126892, "learning_rate": 1.7003013952614244e-06, "loss": 0.0147, "step": 12392 }, { "epoch": 0.8516355140186916, "grad_norm": 0.6182851791381836, "learning_rate": 1.6987578115984526e-06, "loss": 0.0149, "step": 12393 }, { "epoch": 0.8517042330951072, "grad_norm": 0.21793803572654724, "learning_rate": 1.697214886851669e-06, "loss": 0.006, "step": 12394 }, { "epoch": 0.8517729521715228, "grad_norm": 0.24634619057178497, "learning_rate": 1.6956726210975109e-06, "loss": 0.0064, "step": 12395 }, { "epoch": 0.8518416712479384, "grad_norm": 0.552375316619873, "learning_rate": 1.6941310144123823e-06, "loss": 0.0064, "step": 12396 }, { "epoch": 0.851910390324354, "grad_norm": 0.5978779792785645, "learning_rate": 1.6925900668726452e-06, "loss": 0.0125, "step": 12397 }, { "epoch": 0.8519791094007697, "grad_norm": 0.08502507954835892, "learning_rate": 1.6910497785546369e-06, "loss": 0.0046, "step": 12398 }, { "epoch": 0.8520478284771853, "grad_norm": 0.42835021018981934, "learning_rate": 1.6895101495346632e-06, "loss": 0.0073, "step": 12399 }, { "epoch": 0.8521165475536009, "grad_norm": 0.22836990654468536, "learning_rate": 1.687971179888988e-06, "loss": 0.0078, "step": 12400 }, { "epoch": 0.8521852666300165, "grad_norm": 0.2570568323135376, "learning_rate": 1.6864328696938553e-06, "loss": 0.007, "step": 12401 }, { "epoch": 0.8522539857064321, "grad_norm": 0.16063034534454346, "learning_rate": 1.684895219025464e-06, "loss": 0.0081, "step": 12402 }, { "epoch": 0.8523227047828478, "grad_norm": 0.1581067591905594, "learning_rate": 1.6833582279599918e-06, "loss": 0.0081, "step": 12403 }, { "epoch": 0.8523914238592634, "grad_norm": 0.11529760807752609, "learning_rate": 1.6818218965735694e-06, "loss": 0.0026, "step": 12404 }, { "epoch": 0.852460142935679, "grad_norm": 0.44473716616630554, "learning_rate": 1.680286224942314e-06, "loss": 0.0087, "step": 12405 }, { "epoch": 0.8525288620120945, "grad_norm": 0.19783779978752136, "learning_rate": 1.6787512131422934e-06, "loss": 0.0103, "step": 12406 }, { "epoch": 0.8525975810885101, "grad_norm": 0.5784906148910522, "learning_rate": 1.6772168612495531e-06, "loss": 0.0094, "step": 12407 }, { "epoch": 0.8526663001649258, "grad_norm": 0.6298657059669495, "learning_rate": 1.6756831693400943e-06, "loss": 0.0119, "step": 12408 }, { "epoch": 0.8527350192413414, "grad_norm": 0.12346528470516205, "learning_rate": 1.6741501374899027e-06, "loss": 0.0046, "step": 12409 }, { "epoch": 0.852803738317757, "grad_norm": 0.5354821681976318, "learning_rate": 1.6726177657749092e-06, "loss": 0.0098, "step": 12410 }, { "epoch": 0.8528724573941726, "grad_norm": 0.621368944644928, "learning_rate": 1.6710860542710383e-06, "loss": 0.012, "step": 12411 }, { "epoch": 0.8529411764705882, "grad_norm": 0.3711054027080536, "learning_rate": 1.669555003054159e-06, "loss": 0.0069, "step": 12412 }, { "epoch": 0.8530098955470038, "grad_norm": 0.14316782355308533, "learning_rate": 1.6680246122001207e-06, "loss": 0.0118, "step": 12413 }, { "epoch": 0.8530786146234195, "grad_norm": 0.4559236168861389, "learning_rate": 1.6664948817847314e-06, "loss": 0.013, "step": 12414 }, { "epoch": 0.8531473336998351, "grad_norm": 0.4968894124031067, "learning_rate": 1.6649658118837752e-06, "loss": 0.0062, "step": 12415 }, { "epoch": 0.8532160527762507, "grad_norm": 0.35584136843681335, "learning_rate": 1.6634374025729932e-06, "loss": 0.0043, "step": 12416 }, { "epoch": 0.8532847718526663, "grad_norm": 0.7061353325843811, "learning_rate": 1.6619096539281047e-06, "loss": 0.0113, "step": 12417 }, { "epoch": 0.8533534909290819, "grad_norm": 0.14349178969860077, "learning_rate": 1.660382566024793e-06, "loss": 0.0069, "step": 12418 }, { "epoch": 0.8534222100054976, "grad_norm": 0.41240954399108887, "learning_rate": 1.6588561389387019e-06, "loss": 0.0109, "step": 12419 }, { "epoch": 0.8534909290819132, "grad_norm": 0.17902228236198425, "learning_rate": 1.6573303727454513e-06, "loss": 0.0061, "step": 12420 }, { "epoch": 0.8535596481583287, "grad_norm": 0.4253329038619995, "learning_rate": 1.6558052675206193e-06, "loss": 0.0081, "step": 12421 }, { "epoch": 0.8536283672347443, "grad_norm": 0.342617928981781, "learning_rate": 1.65428082333976e-06, "loss": 0.0057, "step": 12422 }, { "epoch": 0.8536970863111599, "grad_norm": 0.6059120297431946, "learning_rate": 1.6527570402783915e-06, "loss": 0.0087, "step": 12423 }, { "epoch": 0.8537658053875756, "grad_norm": 0.5450890064239502, "learning_rate": 1.6512339184120019e-06, "loss": 0.0094, "step": 12424 }, { "epoch": 0.8538345244639912, "grad_norm": 0.549378514289856, "learning_rate": 1.6497114578160356e-06, "loss": 0.006, "step": 12425 }, { "epoch": 0.8539032435404068, "grad_norm": 0.23373787105083466, "learning_rate": 1.6481896585659206e-06, "loss": 0.0047, "step": 12426 }, { "epoch": 0.8539719626168224, "grad_norm": 0.16268670558929443, "learning_rate": 1.6466685207370368e-06, "loss": 0.0032, "step": 12427 }, { "epoch": 0.854040681693238, "grad_norm": 0.7287465333938599, "learning_rate": 1.6451480444047406e-06, "loss": 0.0055, "step": 12428 }, { "epoch": 0.8541094007696537, "grad_norm": 0.1246381625533104, "learning_rate": 1.643628229644355e-06, "loss": 0.0168, "step": 12429 }, { "epoch": 0.8541781198460693, "grad_norm": 0.2094612717628479, "learning_rate": 1.6421090765311697e-06, "loss": 0.0082, "step": 12430 }, { "epoch": 0.8542468389224849, "grad_norm": 0.37545347213745117, "learning_rate": 1.6405905851404345e-06, "loss": 0.007, "step": 12431 }, { "epoch": 0.8543155579989005, "grad_norm": 0.23893658816814423, "learning_rate": 1.6390727555473794e-06, "loss": 0.0067, "step": 12432 }, { "epoch": 0.8543842770753161, "grad_norm": 0.7680916786193848, "learning_rate": 1.6375555878271892e-06, "loss": 0.0115, "step": 12433 }, { "epoch": 0.8544529961517318, "grad_norm": 0.3499032258987427, "learning_rate": 1.636039082055022e-06, "loss": 0.0055, "step": 12434 }, { "epoch": 0.8545217152281473, "grad_norm": 0.6929242610931396, "learning_rate": 1.634523238306006e-06, "loss": 0.0084, "step": 12435 }, { "epoch": 0.8545904343045629, "grad_norm": 0.23942267894744873, "learning_rate": 1.633008056655228e-06, "loss": 0.0072, "step": 12436 }, { "epoch": 0.8546591533809785, "grad_norm": 0.1249174103140831, "learning_rate": 1.6314935371777527e-06, "loss": 0.0106, "step": 12437 }, { "epoch": 0.8547278724573941, "grad_norm": 0.1359485387802124, "learning_rate": 1.6299796799486006e-06, "loss": 0.0088, "step": 12438 }, { "epoch": 0.8547965915338098, "grad_norm": 0.6425939798355103, "learning_rate": 1.6284664850427695e-06, "loss": 0.0099, "step": 12439 }, { "epoch": 0.8548653106102254, "grad_norm": 1.0893244743347168, "learning_rate": 1.6269539525352129e-06, "loss": 0.0139, "step": 12440 }, { "epoch": 0.854934029686641, "grad_norm": 0.29053276777267456, "learning_rate": 1.6254420825008693e-06, "loss": 0.0246, "step": 12441 }, { "epoch": 0.8550027487630566, "grad_norm": 0.32841894030570984, "learning_rate": 1.6239308750146254e-06, "loss": 0.0105, "step": 12442 }, { "epoch": 0.8550714678394722, "grad_norm": 0.31663620471954346, "learning_rate": 1.622420330151348e-06, "loss": 0.0049, "step": 12443 }, { "epoch": 0.8551401869158879, "grad_norm": 0.6546825766563416, "learning_rate": 1.6209104479858622e-06, "loss": 0.0095, "step": 12444 }, { "epoch": 0.8552089059923035, "grad_norm": 0.10431020706892014, "learning_rate": 1.61940122859297e-06, "loss": 0.0054, "step": 12445 }, { "epoch": 0.8552776250687191, "grad_norm": 0.28980207443237305, "learning_rate": 1.6178926720474246e-06, "loss": 0.0083, "step": 12446 }, { "epoch": 0.8553463441451347, "grad_norm": 0.4808967113494873, "learning_rate": 1.61638477842397e-06, "loss": 0.0137, "step": 12447 }, { "epoch": 0.8554150632215503, "grad_norm": 1.6499426364898682, "learning_rate": 1.6148775477972943e-06, "loss": 0.0161, "step": 12448 }, { "epoch": 0.855483782297966, "grad_norm": 0.7075579166412354, "learning_rate": 1.6133709802420698e-06, "loss": 0.0102, "step": 12449 }, { "epoch": 0.8555525013743815, "grad_norm": 0.4007588326931, "learning_rate": 1.6118650758329235e-06, "loss": 0.0106, "step": 12450 }, { "epoch": 0.8556212204507971, "grad_norm": 0.3817482590675354, "learning_rate": 1.610359834644457e-06, "loss": 0.0159, "step": 12451 }, { "epoch": 0.8556899395272127, "grad_norm": 0.4025731086730957, "learning_rate": 1.608855256751236e-06, "loss": 0.0086, "step": 12452 }, { "epoch": 0.8557586586036283, "grad_norm": 0.5235246419906616, "learning_rate": 1.6073513422277925e-06, "loss": 0.0113, "step": 12453 }, { "epoch": 0.855827377680044, "grad_norm": 0.42898234724998474, "learning_rate": 1.6058480911486333e-06, "loss": 0.0071, "step": 12454 }, { "epoch": 0.8558960967564596, "grad_norm": 1.326453685760498, "learning_rate": 1.6043455035882193e-06, "loss": 0.0092, "step": 12455 }, { "epoch": 0.8559648158328752, "grad_norm": 0.5627119541168213, "learning_rate": 1.6028435796209922e-06, "loss": 0.0104, "step": 12456 }, { "epoch": 0.8560335349092908, "grad_norm": 0.6086945533752441, "learning_rate": 1.6013423193213462e-06, "loss": 0.0071, "step": 12457 }, { "epoch": 0.8561022539857064, "grad_norm": 0.6001411080360413, "learning_rate": 1.5998417227636564e-06, "loss": 0.0155, "step": 12458 }, { "epoch": 0.8561709730621221, "grad_norm": 0.11652777343988419, "learning_rate": 1.5983417900222585e-06, "loss": 0.0056, "step": 12459 }, { "epoch": 0.8562396921385377, "grad_norm": 0.5153899192810059, "learning_rate": 1.5968425211714582e-06, "loss": 0.0167, "step": 12460 }, { "epoch": 0.8563084112149533, "grad_norm": 0.3007946312427521, "learning_rate": 1.595343916285521e-06, "loss": 0.0056, "step": 12461 }, { "epoch": 0.8563771302913689, "grad_norm": 1.4797323942184448, "learning_rate": 1.593845975438689e-06, "loss": 0.0179, "step": 12462 }, { "epoch": 0.8564458493677845, "grad_norm": 0.19005174934864044, "learning_rate": 1.5923486987051632e-06, "loss": 0.0092, "step": 12463 }, { "epoch": 0.8565145684442002, "grad_norm": 0.3850124478340149, "learning_rate": 1.590852086159119e-06, "loss": 0.008, "step": 12464 }, { "epoch": 0.8565832875206157, "grad_norm": 0.4724330008029938, "learning_rate": 1.589356137874694e-06, "loss": 0.0076, "step": 12465 }, { "epoch": 0.8566520065970313, "grad_norm": 0.24093246459960938, "learning_rate": 1.587860853925997e-06, "loss": 0.0093, "step": 12466 }, { "epoch": 0.8567207256734469, "grad_norm": 0.18809035420417786, "learning_rate": 1.5863662343870971e-06, "loss": 0.0122, "step": 12467 }, { "epoch": 0.8567894447498625, "grad_norm": 0.4613986909389496, "learning_rate": 1.5848722793320386e-06, "loss": 0.0128, "step": 12468 }, { "epoch": 0.8568581638262782, "grad_norm": 0.3211672604084015, "learning_rate": 1.5833789888348255e-06, "loss": 0.0126, "step": 12469 }, { "epoch": 0.8569268829026938, "grad_norm": 0.2784567177295685, "learning_rate": 1.5818863629694319e-06, "loss": 0.0063, "step": 12470 }, { "epoch": 0.8569956019791094, "grad_norm": 0.7317799925804138, "learning_rate": 1.5803944018098055e-06, "loss": 0.0131, "step": 12471 }, { "epoch": 0.857064321055525, "grad_norm": 0.1787731647491455, "learning_rate": 1.5789031054298487e-06, "loss": 0.0044, "step": 12472 }, { "epoch": 0.8571330401319406, "grad_norm": 0.21182502806186676, "learning_rate": 1.5774124739034406e-06, "loss": 0.0107, "step": 12473 }, { "epoch": 0.8572017592083563, "grad_norm": 0.48734548687934875, "learning_rate": 1.575922507304421e-06, "loss": 0.0071, "step": 12474 }, { "epoch": 0.8572704782847719, "grad_norm": 0.3046519458293915, "learning_rate": 1.5744332057066052e-06, "loss": 0.0125, "step": 12475 }, { "epoch": 0.8573391973611875, "grad_norm": 0.4554113745689392, "learning_rate": 1.5729445691837596e-06, "loss": 0.0097, "step": 12476 }, { "epoch": 0.8574079164376031, "grad_norm": 0.614946722984314, "learning_rate": 1.5714565978096419e-06, "loss": 0.0127, "step": 12477 }, { "epoch": 0.8574766355140186, "grad_norm": 0.25797170400619507, "learning_rate": 1.5699692916579528e-06, "loss": 0.0066, "step": 12478 }, { "epoch": 0.8575453545904343, "grad_norm": 0.35244473814964294, "learning_rate": 1.568482650802377e-06, "loss": 0.0105, "step": 12479 }, { "epoch": 0.8576140736668499, "grad_norm": 0.12572969496250153, "learning_rate": 1.5669966753165537e-06, "loss": 0.0079, "step": 12480 }, { "epoch": 0.8576827927432655, "grad_norm": 0.15529665350914001, "learning_rate": 1.5655113652741027e-06, "loss": 0.006, "step": 12481 }, { "epoch": 0.8577515118196811, "grad_norm": 0.16159240901470184, "learning_rate": 1.5640267207485898e-06, "loss": 0.0064, "step": 12482 }, { "epoch": 0.8578202308960967, "grad_norm": 0.1680341213941574, "learning_rate": 1.562542741813578e-06, "loss": 0.0047, "step": 12483 }, { "epoch": 0.8578889499725124, "grad_norm": 0.19307973980903625, "learning_rate": 1.5610594285425684e-06, "loss": 0.0034, "step": 12484 }, { "epoch": 0.857957669048928, "grad_norm": 0.26029402017593384, "learning_rate": 1.5595767810090489e-06, "loss": 0.0046, "step": 12485 }, { "epoch": 0.8580263881253436, "grad_norm": 0.3731531798839569, "learning_rate": 1.5580947992864609e-06, "loss": 0.0147, "step": 12486 }, { "epoch": 0.8580951072017592, "grad_norm": 0.7863641977310181, "learning_rate": 1.5566134834482237e-06, "loss": 0.0207, "step": 12487 }, { "epoch": 0.8581638262781748, "grad_norm": 0.24797962605953217, "learning_rate": 1.5551328335677123e-06, "loss": 0.0111, "step": 12488 }, { "epoch": 0.8582325453545905, "grad_norm": 0.27991896867752075, "learning_rate": 1.5536528497182795e-06, "loss": 0.0101, "step": 12489 }, { "epoch": 0.8583012644310061, "grad_norm": 0.5453490614891052, "learning_rate": 1.5521735319732434e-06, "loss": 0.0108, "step": 12490 }, { "epoch": 0.8583699835074217, "grad_norm": 0.2916255593299866, "learning_rate": 1.5506948804058784e-06, "loss": 0.0154, "step": 12491 }, { "epoch": 0.8584387025838373, "grad_norm": 0.18577182292938232, "learning_rate": 1.549216895089443e-06, "loss": 0.0064, "step": 12492 }, { "epoch": 0.8585074216602528, "grad_norm": 0.200535848736763, "learning_rate": 1.5477395760971452e-06, "loss": 0.0082, "step": 12493 }, { "epoch": 0.8585761407366685, "grad_norm": 0.23737189173698425, "learning_rate": 1.546262923502173e-06, "loss": 0.0075, "step": 12494 }, { "epoch": 0.8586448598130841, "grad_norm": 0.5213960409164429, "learning_rate": 1.5447869373776746e-06, "loss": 0.0208, "step": 12495 }, { "epoch": 0.8587135788894997, "grad_norm": 0.783566415309906, "learning_rate": 1.5433116177967732e-06, "loss": 0.0095, "step": 12496 }, { "epoch": 0.8587822979659153, "grad_norm": 0.39626461267471313, "learning_rate": 1.5418369648325436e-06, "loss": 0.0115, "step": 12497 }, { "epoch": 0.8588510170423309, "grad_norm": 0.20670853555202484, "learning_rate": 1.540362978558046e-06, "loss": 0.0113, "step": 12498 }, { "epoch": 0.8589197361187466, "grad_norm": 0.1955881416797638, "learning_rate": 1.5388896590462915e-06, "loss": 0.0058, "step": 12499 }, { "epoch": 0.8589884551951622, "grad_norm": 0.22766242921352386, "learning_rate": 1.5374170063702685e-06, "loss": 0.0099, "step": 12500 }, { "epoch": 0.8590571742715778, "grad_norm": 0.2786252498626709, "learning_rate": 1.535945020602929e-06, "loss": 0.007, "step": 12501 }, { "epoch": 0.8591258933479934, "grad_norm": 0.44269922375679016, "learning_rate": 1.534473701817196e-06, "loss": 0.0076, "step": 12502 }, { "epoch": 0.859194612424409, "grad_norm": 0.14387638866901398, "learning_rate": 1.5330030500859482e-06, "loss": 0.0117, "step": 12503 }, { "epoch": 0.8592633315008247, "grad_norm": 0.5771119594573975, "learning_rate": 1.5315330654820452e-06, "loss": 0.0101, "step": 12504 }, { "epoch": 0.8593320505772403, "grad_norm": 0.4215834140777588, "learning_rate": 1.5300637480783008e-06, "loss": 0.0086, "step": 12505 }, { "epoch": 0.8594007696536559, "grad_norm": 0.17748618125915527, "learning_rate": 1.528595097947505e-06, "loss": 0.0076, "step": 12506 }, { "epoch": 0.8594694887300715, "grad_norm": 0.2647343575954437, "learning_rate": 1.5271271151624162e-06, "loss": 0.0132, "step": 12507 }, { "epoch": 0.859538207806487, "grad_norm": 0.17811614274978638, "learning_rate": 1.5256597997957477e-06, "loss": 0.005, "step": 12508 }, { "epoch": 0.8596069268829027, "grad_norm": 0.22917777299880981, "learning_rate": 1.5241931519201935e-06, "loss": 0.012, "step": 12509 }, { "epoch": 0.8596756459593183, "grad_norm": 0.18656125664710999, "learning_rate": 1.5227271716084035e-06, "loss": 0.0074, "step": 12510 }, { "epoch": 0.8597443650357339, "grad_norm": 0.41604936122894287, "learning_rate": 1.521261858933003e-06, "loss": 0.0101, "step": 12511 }, { "epoch": 0.8598130841121495, "grad_norm": 0.07363121956586838, "learning_rate": 1.5197972139665722e-06, "loss": 0.0052, "step": 12512 }, { "epoch": 0.8598818031885651, "grad_norm": 0.2652797996997833, "learning_rate": 1.5183332367816816e-06, "loss": 0.0064, "step": 12513 }, { "epoch": 0.8599505222649808, "grad_norm": 0.271678626537323, "learning_rate": 1.5168699274508395e-06, "loss": 0.0078, "step": 12514 }, { "epoch": 0.8600192413413964, "grad_norm": 0.46382489800453186, "learning_rate": 1.5154072860465467e-06, "loss": 0.0086, "step": 12515 }, { "epoch": 0.860087960417812, "grad_norm": 0.3399040997028351, "learning_rate": 1.51394531264125e-06, "loss": 0.0048, "step": 12516 }, { "epoch": 0.8601566794942276, "grad_norm": 0.639305830001831, "learning_rate": 1.5124840073073781e-06, "loss": 0.0067, "step": 12517 }, { "epoch": 0.8602253985706432, "grad_norm": 0.9223431348800659, "learning_rate": 1.5110233701173148e-06, "loss": 0.0044, "step": 12518 }, { "epoch": 0.8602941176470589, "grad_norm": 0.10898499935865402, "learning_rate": 1.5095634011434273e-06, "loss": 0.0069, "step": 12519 }, { "epoch": 0.8603628367234745, "grad_norm": 0.3745884895324707, "learning_rate": 1.508104100458031e-06, "loss": 0.0087, "step": 12520 }, { "epoch": 0.8604315557998901, "grad_norm": 0.38916751742362976, "learning_rate": 1.5066454681334213e-06, "loss": 0.0061, "step": 12521 }, { "epoch": 0.8605002748763056, "grad_norm": 0.3360093832015991, "learning_rate": 1.5051875042418506e-06, "loss": 0.0068, "step": 12522 }, { "epoch": 0.8605689939527212, "grad_norm": 0.6482900977134705, "learning_rate": 1.503730208855551e-06, "loss": 0.0112, "step": 12523 }, { "epoch": 0.8606377130291369, "grad_norm": 0.373146653175354, "learning_rate": 1.5022735820467065e-06, "loss": 0.0097, "step": 12524 }, { "epoch": 0.8607064321055525, "grad_norm": 0.30039235949516296, "learning_rate": 1.5008176238874794e-06, "loss": 0.009, "step": 12525 }, { "epoch": 0.8607751511819681, "grad_norm": 0.18483075499534607, "learning_rate": 1.499362334449995e-06, "loss": 0.0061, "step": 12526 }, { "epoch": 0.8608438702583837, "grad_norm": 0.4575609266757965, "learning_rate": 1.497907713806343e-06, "loss": 0.0094, "step": 12527 }, { "epoch": 0.8609125893347993, "grad_norm": 0.4861074686050415, "learning_rate": 1.4964537620285867e-06, "loss": 0.0151, "step": 12528 }, { "epoch": 0.860981308411215, "grad_norm": 0.17744769155979156, "learning_rate": 1.4950004791887456e-06, "loss": 0.008, "step": 12529 }, { "epoch": 0.8610500274876306, "grad_norm": 0.3602364957332611, "learning_rate": 1.4935478653588153e-06, "loss": 0.0128, "step": 12530 }, { "epoch": 0.8611187465640462, "grad_norm": 0.4506692886352539, "learning_rate": 1.4920959206107566e-06, "loss": 0.0074, "step": 12531 }, { "epoch": 0.8611874656404618, "grad_norm": 0.2585325241088867, "learning_rate": 1.4906446450164968e-06, "loss": 0.0124, "step": 12532 }, { "epoch": 0.8612561847168774, "grad_norm": 0.5327944159507751, "learning_rate": 1.4891940386479252e-06, "loss": 0.0135, "step": 12533 }, { "epoch": 0.8613249037932931, "grad_norm": 0.09472187608480453, "learning_rate": 1.487744101576906e-06, "loss": 0.0157, "step": 12534 }, { "epoch": 0.8613936228697087, "grad_norm": 0.13575661182403564, "learning_rate": 1.4862948338752619e-06, "loss": 0.0088, "step": 12535 }, { "epoch": 0.8614623419461243, "grad_norm": 0.21780887246131897, "learning_rate": 1.4848462356147901e-06, "loss": 0.007, "step": 12536 }, { "epoch": 0.8615310610225398, "grad_norm": 0.15131281316280365, "learning_rate": 1.4833983068672451e-06, "loss": 0.0044, "step": 12537 }, { "epoch": 0.8615997800989554, "grad_norm": 0.5934381484985352, "learning_rate": 1.4819510477043663e-06, "loss": 0.0094, "step": 12538 }, { "epoch": 0.861668499175371, "grad_norm": 0.49582111835479736, "learning_rate": 1.4805044581978361e-06, "loss": 0.0061, "step": 12539 }, { "epoch": 0.8617372182517867, "grad_norm": 0.3067210614681244, "learning_rate": 1.4790585384193223e-06, "loss": 0.0074, "step": 12540 }, { "epoch": 0.8618059373282023, "grad_norm": 0.5762063264846802, "learning_rate": 1.4776132884404493e-06, "loss": 0.0126, "step": 12541 }, { "epoch": 0.8618746564046179, "grad_norm": 0.0751076191663742, "learning_rate": 1.4761687083328146e-06, "loss": 0.0068, "step": 12542 }, { "epoch": 0.8619433754810335, "grad_norm": 0.25243842601776123, "learning_rate": 1.4747247981679762e-06, "loss": 0.0083, "step": 12543 }, { "epoch": 0.8620120945574491, "grad_norm": 0.3395489752292633, "learning_rate": 1.4732815580174651e-06, "loss": 0.0079, "step": 12544 }, { "epoch": 0.8620808136338648, "grad_norm": 0.26583459973335266, "learning_rate": 1.4718389879527777e-06, "loss": 0.0057, "step": 12545 }, { "epoch": 0.8621495327102804, "grad_norm": 0.1463886797428131, "learning_rate": 1.4703970880453715e-06, "loss": 0.0097, "step": 12546 }, { "epoch": 0.862218251786696, "grad_norm": 0.3315894901752472, "learning_rate": 1.4689558583666812e-06, "loss": 0.0117, "step": 12547 }, { "epoch": 0.8622869708631116, "grad_norm": 0.23027051985263824, "learning_rate": 1.4675152989880963e-06, "loss": 0.0205, "step": 12548 }, { "epoch": 0.8623556899395272, "grad_norm": 0.41742610931396484, "learning_rate": 1.4660754099809813e-06, "loss": 0.0193, "step": 12549 }, { "epoch": 0.8624244090159429, "grad_norm": 0.09555072337388992, "learning_rate": 1.4646361914166661e-06, "loss": 0.0086, "step": 12550 }, { "epoch": 0.8624931280923585, "grad_norm": 0.10483045876026154, "learning_rate": 1.4631976433664485e-06, "loss": 0.0105, "step": 12551 }, { "epoch": 0.862561847168774, "grad_norm": 0.14155490696430206, "learning_rate": 1.461759765901588e-06, "loss": 0.0088, "step": 12552 }, { "epoch": 0.8626305662451896, "grad_norm": 0.17977212369441986, "learning_rate": 1.4603225590933161e-06, "loss": 0.0088, "step": 12553 }, { "epoch": 0.8626992853216052, "grad_norm": 0.43286848068237305, "learning_rate": 1.4588860230128276e-06, "loss": 0.012, "step": 12554 }, { "epoch": 0.8627680043980209, "grad_norm": 0.3108847737312317, "learning_rate": 1.4574501577312837e-06, "loss": 0.0059, "step": 12555 }, { "epoch": 0.8628367234744365, "grad_norm": 0.5358332991600037, "learning_rate": 1.4560149633198177e-06, "loss": 0.0118, "step": 12556 }, { "epoch": 0.8629054425508521, "grad_norm": 0.2628234922885895, "learning_rate": 1.4545804398495295e-06, "loss": 0.0046, "step": 12557 }, { "epoch": 0.8629741616272677, "grad_norm": 0.5215633511543274, "learning_rate": 1.4531465873914734e-06, "loss": 0.0085, "step": 12558 }, { "epoch": 0.8630428807036833, "grad_norm": 0.116534024477005, "learning_rate": 1.4517134060166882e-06, "loss": 0.0074, "step": 12559 }, { "epoch": 0.863111599780099, "grad_norm": 0.22405774891376495, "learning_rate": 1.4502808957961633e-06, "loss": 0.0085, "step": 12560 }, { "epoch": 0.8631803188565146, "grad_norm": 0.2059665024280548, "learning_rate": 1.4488490568008655e-06, "loss": 0.0063, "step": 12561 }, { "epoch": 0.8632490379329302, "grad_norm": 0.1714085042476654, "learning_rate": 1.4474178891017298e-06, "loss": 0.0078, "step": 12562 }, { "epoch": 0.8633177570093458, "grad_norm": 0.3522123694419861, "learning_rate": 1.4459873927696442e-06, "loss": 0.0082, "step": 12563 }, { "epoch": 0.8633864760857614, "grad_norm": 0.2733985185623169, "learning_rate": 1.4445575678754808e-06, "loss": 0.0083, "step": 12564 }, { "epoch": 0.8634551951621771, "grad_norm": 0.2510354518890381, "learning_rate": 1.4431284144900642e-06, "loss": 0.009, "step": 12565 }, { "epoch": 0.8635239142385926, "grad_norm": 0.5218505859375, "learning_rate": 1.4416999326841962e-06, "loss": 0.0183, "step": 12566 }, { "epoch": 0.8635926333150082, "grad_norm": 0.1055777445435524, "learning_rate": 1.4402721225286336e-06, "loss": 0.0056, "step": 12567 }, { "epoch": 0.8636613523914238, "grad_norm": 0.05128004029393196, "learning_rate": 1.4388449840941198e-06, "loss": 0.0056, "step": 12568 }, { "epoch": 0.8637300714678394, "grad_norm": 0.20049068331718445, "learning_rate": 1.43741851745134e-06, "loss": 0.0091, "step": 12569 }, { "epoch": 0.8637987905442551, "grad_norm": 0.29332807660102844, "learning_rate": 1.4359927226709674e-06, "loss": 0.0048, "step": 12570 }, { "epoch": 0.8638675096206707, "grad_norm": 0.5085837244987488, "learning_rate": 1.434567599823624e-06, "loss": 0.016, "step": 12571 }, { "epoch": 0.8639362286970863, "grad_norm": 0.21690841019153595, "learning_rate": 1.4331431489799169e-06, "loss": 0.008, "step": 12572 }, { "epoch": 0.8640049477735019, "grad_norm": 0.24262909591197968, "learning_rate": 1.4317193702104008e-06, "loss": 0.0429, "step": 12573 }, { "epoch": 0.8640736668499175, "grad_norm": 0.8258937001228333, "learning_rate": 1.4302962635856165e-06, "loss": 0.0101, "step": 12574 }, { "epoch": 0.8641423859263332, "grad_norm": 0.32028812170028687, "learning_rate": 1.428873829176054e-06, "loss": 0.007, "step": 12575 }, { "epoch": 0.8642111050027488, "grad_norm": 0.3952160179615021, "learning_rate": 1.427452067052185e-06, "loss": 0.0122, "step": 12576 }, { "epoch": 0.8642798240791644, "grad_norm": 0.2875087559223175, "learning_rate": 1.4260309772844337e-06, "loss": 0.0096, "step": 12577 }, { "epoch": 0.86434854315558, "grad_norm": 0.6317941546440125, "learning_rate": 1.4246105599432019e-06, "loss": 0.0047, "step": 12578 }, { "epoch": 0.8644172622319956, "grad_norm": 0.36232689023017883, "learning_rate": 1.4231908150988532e-06, "loss": 0.011, "step": 12579 }, { "epoch": 0.8644859813084113, "grad_norm": 0.3979775309562683, "learning_rate": 1.4217717428217165e-06, "loss": 0.0041, "step": 12580 }, { "epoch": 0.8645547003848268, "grad_norm": 0.5133607387542725, "learning_rate": 1.4203533431820954e-06, "loss": 0.0111, "step": 12581 }, { "epoch": 0.8646234194612424, "grad_norm": 0.5069577693939209, "learning_rate": 1.4189356162502488e-06, "loss": 0.0069, "step": 12582 }, { "epoch": 0.864692138537658, "grad_norm": 0.1328304558992386, "learning_rate": 1.4175185620964086e-06, "loss": 0.0056, "step": 12583 }, { "epoch": 0.8647608576140736, "grad_norm": 0.5558283925056458, "learning_rate": 1.4161021807907793e-06, "loss": 0.0114, "step": 12584 }, { "epoch": 0.8648295766904893, "grad_norm": 0.22407038509845734, "learning_rate": 1.4146864724035174e-06, "loss": 0.0115, "step": 12585 }, { "epoch": 0.8648982957669049, "grad_norm": 0.3637426495552063, "learning_rate": 1.4132714370047573e-06, "loss": 0.0094, "step": 12586 }, { "epoch": 0.8649670148433205, "grad_norm": 0.2826021611690521, "learning_rate": 1.4118570746645993e-06, "loss": 0.007, "step": 12587 }, { "epoch": 0.8650357339197361, "grad_norm": 0.31902486085891724, "learning_rate": 1.410443385453104e-06, "loss": 0.0063, "step": 12588 }, { "epoch": 0.8651044529961517, "grad_norm": 0.25766029953956604, "learning_rate": 1.4090303694403056e-06, "loss": 0.0047, "step": 12589 }, { "epoch": 0.8651731720725674, "grad_norm": 0.2378843128681183, "learning_rate": 1.4076180266961997e-06, "loss": 0.0131, "step": 12590 }, { "epoch": 0.865241891148983, "grad_norm": 0.3172190487384796, "learning_rate": 1.406206357290752e-06, "loss": 0.0103, "step": 12591 }, { "epoch": 0.8653106102253986, "grad_norm": 0.24946174025535583, "learning_rate": 1.404795361293893e-06, "loss": 0.0156, "step": 12592 }, { "epoch": 0.8653793293018142, "grad_norm": 0.36646902561187744, "learning_rate": 1.4033850387755238e-06, "loss": 0.0077, "step": 12593 }, { "epoch": 0.8654480483782298, "grad_norm": 0.1981210559606552, "learning_rate": 1.4019753898055032e-06, "loss": 0.0094, "step": 12594 }, { "epoch": 0.8655167674546455, "grad_norm": 0.13062584400177002, "learning_rate": 1.4005664144536689e-06, "loss": 0.0093, "step": 12595 }, { "epoch": 0.865585486531061, "grad_norm": 0.5114433765411377, "learning_rate": 1.3991581127898135e-06, "loss": 0.0082, "step": 12596 }, { "epoch": 0.8656542056074766, "grad_norm": 0.5137307047843933, "learning_rate": 1.3977504848837007e-06, "loss": 0.0082, "step": 12597 }, { "epoch": 0.8657229246838922, "grad_norm": 0.5129841566085815, "learning_rate": 1.3963435308050686e-06, "loss": 0.0171, "step": 12598 }, { "epoch": 0.8657916437603078, "grad_norm": 1.0109626054763794, "learning_rate": 1.3949372506236064e-06, "loss": 0.0062, "step": 12599 }, { "epoch": 0.8658603628367235, "grad_norm": 0.29458338022232056, "learning_rate": 1.3935316444089813e-06, "loss": 0.0043, "step": 12600 }, { "epoch": 0.8659290819131391, "grad_norm": 0.1574632078409195, "learning_rate": 1.3921267122308295e-06, "loss": 0.0061, "step": 12601 }, { "epoch": 0.8659978009895547, "grad_norm": 0.3444423973560333, "learning_rate": 1.3907224541587389e-06, "loss": 0.0067, "step": 12602 }, { "epoch": 0.8660665200659703, "grad_norm": 0.5050787925720215, "learning_rate": 1.38931887026228e-06, "loss": 0.0143, "step": 12603 }, { "epoch": 0.8661352391423859, "grad_norm": 0.13689494132995605, "learning_rate": 1.3879159606109843e-06, "loss": 0.0043, "step": 12604 }, { "epoch": 0.8662039582188016, "grad_norm": 0.4516492187976837, "learning_rate": 1.3865137252743443e-06, "loss": 0.0129, "step": 12605 }, { "epoch": 0.8662726772952172, "grad_norm": 1.0650591850280762, "learning_rate": 1.3851121643218278e-06, "loss": 0.0156, "step": 12606 }, { "epoch": 0.8663413963716328, "grad_norm": 0.29520267248153687, "learning_rate": 1.383711277822861e-06, "loss": 0.0065, "step": 12607 }, { "epoch": 0.8664101154480484, "grad_norm": 1.0923982858657837, "learning_rate": 1.382311065846847e-06, "loss": 0.016, "step": 12608 }, { "epoch": 0.866478834524464, "grad_norm": 0.1350756138563156, "learning_rate": 1.3809115284631397e-06, "loss": 0.0054, "step": 12609 }, { "epoch": 0.8665475536008796, "grad_norm": 0.4323195815086365, "learning_rate": 1.3795126657410822e-06, "loss": 0.0093, "step": 12610 }, { "epoch": 0.8666162726772952, "grad_norm": 0.09214089065790176, "learning_rate": 1.378114477749961e-06, "loss": 0.0101, "step": 12611 }, { "epoch": 0.8666849917537108, "grad_norm": 0.3685402274131775, "learning_rate": 1.3767169645590473e-06, "loss": 0.0163, "step": 12612 }, { "epoch": 0.8667537108301264, "grad_norm": 0.5896568894386292, "learning_rate": 1.3753201262375637e-06, "loss": 0.0091, "step": 12613 }, { "epoch": 0.866822429906542, "grad_norm": 0.4021422564983368, "learning_rate": 1.3739239628547117e-06, "loss": 0.0048, "step": 12614 }, { "epoch": 0.8668911489829577, "grad_norm": 0.31902316212654114, "learning_rate": 1.3725284744796495e-06, "loss": 0.0071, "step": 12615 }, { "epoch": 0.8669598680593733, "grad_norm": 0.3901505768299103, "learning_rate": 1.3711336611815117e-06, "loss": 0.0136, "step": 12616 }, { "epoch": 0.8670285871357889, "grad_norm": 0.5272403359413147, "learning_rate": 1.3697395230293913e-06, "loss": 0.0117, "step": 12617 }, { "epoch": 0.8670973062122045, "grad_norm": 0.7750481367111206, "learning_rate": 1.3683460600923546e-06, "loss": 0.0142, "step": 12618 }, { "epoch": 0.8671660252886201, "grad_norm": 0.484984427690506, "learning_rate": 1.3669532724394268e-06, "loss": 0.0069, "step": 12619 }, { "epoch": 0.8672347443650358, "grad_norm": 0.22165817022323608, "learning_rate": 1.3655611601396074e-06, "loss": 0.0051, "step": 12620 }, { "epoch": 0.8673034634414514, "grad_norm": 0.23236343264579773, "learning_rate": 1.3641697232618545e-06, "loss": 0.0105, "step": 12621 }, { "epoch": 0.867372182517867, "grad_norm": 0.29833984375, "learning_rate": 1.3627789618750997e-06, "loss": 0.0081, "step": 12622 }, { "epoch": 0.8674409015942826, "grad_norm": 0.36749738454818726, "learning_rate": 1.3613888760482412e-06, "loss": 0.0072, "step": 12623 }, { "epoch": 0.8675096206706981, "grad_norm": 0.22207628190517426, "learning_rate": 1.3599994658501341e-06, "loss": 0.0068, "step": 12624 }, { "epoch": 0.8675783397471138, "grad_norm": 0.40428462624549866, "learning_rate": 1.3586107313496148e-06, "loss": 0.0105, "step": 12625 }, { "epoch": 0.8676470588235294, "grad_norm": 0.3399849832057953, "learning_rate": 1.357222672615472e-06, "loss": 0.0042, "step": 12626 }, { "epoch": 0.867715777899945, "grad_norm": 0.4496205151081085, "learning_rate": 1.3558352897164683e-06, "loss": 0.0109, "step": 12627 }, { "epoch": 0.8677844969763606, "grad_norm": 0.42104417085647583, "learning_rate": 1.3544485827213326e-06, "loss": 0.0124, "step": 12628 }, { "epoch": 0.8678532160527762, "grad_norm": 0.17182496190071106, "learning_rate": 1.353062551698765e-06, "loss": 0.0068, "step": 12629 }, { "epoch": 0.8679219351291919, "grad_norm": 0.285044401884079, "learning_rate": 1.3516771967174168e-06, "loss": 0.0073, "step": 12630 }, { "epoch": 0.8679906542056075, "grad_norm": 0.2321881353855133, "learning_rate": 1.3502925178459253e-06, "loss": 0.0045, "step": 12631 }, { "epoch": 0.8680593732820231, "grad_norm": 0.2819637060165405, "learning_rate": 1.3489085151528752e-06, "loss": 0.009, "step": 12632 }, { "epoch": 0.8681280923584387, "grad_norm": 0.480286568403244, "learning_rate": 1.3475251887068317e-06, "loss": 0.0174, "step": 12633 }, { "epoch": 0.8681968114348543, "grad_norm": 0.20612619817256927, "learning_rate": 1.3461425385763238e-06, "loss": 0.0075, "step": 12634 }, { "epoch": 0.86826553051127, "grad_norm": 0.21424344182014465, "learning_rate": 1.3447605648298461e-06, "loss": 0.0085, "step": 12635 }, { "epoch": 0.8683342495876856, "grad_norm": 0.21674074232578278, "learning_rate": 1.3433792675358525e-06, "loss": 0.0102, "step": 12636 }, { "epoch": 0.8684029686641012, "grad_norm": 0.21767005324363708, "learning_rate": 1.3419986467627748e-06, "loss": 0.0063, "step": 12637 }, { "epoch": 0.8684716877405168, "grad_norm": 0.6348352432250977, "learning_rate": 1.3406187025790033e-06, "loss": 0.0137, "step": 12638 }, { "epoch": 0.8685404068169323, "grad_norm": 0.34494587779045105, "learning_rate": 1.3392394350528985e-06, "loss": 0.0068, "step": 12639 }, { "epoch": 0.868609125893348, "grad_norm": 0.11767837405204773, "learning_rate": 1.337860844252789e-06, "loss": 0.0059, "step": 12640 }, { "epoch": 0.8686778449697636, "grad_norm": 0.2516583800315857, "learning_rate": 1.3364829302469634e-06, "loss": 0.01, "step": 12641 }, { "epoch": 0.8687465640461792, "grad_norm": 0.28666552901268005, "learning_rate": 1.3351056931036838e-06, "loss": 0.0076, "step": 12642 }, { "epoch": 0.8688152831225948, "grad_norm": 0.20680201053619385, "learning_rate": 1.333729132891172e-06, "loss": 0.005, "step": 12643 }, { "epoch": 0.8688840021990104, "grad_norm": 0.27415481209754944, "learning_rate": 1.3323532496776258e-06, "loss": 0.0108, "step": 12644 }, { "epoch": 0.8689527212754261, "grad_norm": 0.15351654589176178, "learning_rate": 1.3309780435311935e-06, "loss": 0.0168, "step": 12645 }, { "epoch": 0.8690214403518417, "grad_norm": 0.5796772837638855, "learning_rate": 1.3296035145200126e-06, "loss": 0.0085, "step": 12646 }, { "epoch": 0.8690901594282573, "grad_norm": 0.7211951017379761, "learning_rate": 1.3282296627121665e-06, "loss": 0.0047, "step": 12647 }, { "epoch": 0.8691588785046729, "grad_norm": 0.25162529945373535, "learning_rate": 1.3268564881757162e-06, "loss": 0.01, "step": 12648 }, { "epoch": 0.8692275975810885, "grad_norm": 0.39326316118240356, "learning_rate": 1.3254839909786837e-06, "loss": 0.0081, "step": 12649 }, { "epoch": 0.8692963166575042, "grad_norm": 0.3398292064666748, "learning_rate": 1.3241121711890613e-06, "loss": 0.0085, "step": 12650 }, { "epoch": 0.8693650357339198, "grad_norm": 0.4503498375415802, "learning_rate": 1.3227410288748015e-06, "loss": 0.0096, "step": 12651 }, { "epoch": 0.8694337548103354, "grad_norm": 0.18908114731311798, "learning_rate": 1.3213705641038364e-06, "loss": 0.0059, "step": 12652 }, { "epoch": 0.869502473886751, "grad_norm": 0.44382911920547485, "learning_rate": 1.3200007769440503e-06, "loss": 0.0099, "step": 12653 }, { "epoch": 0.8695711929631665, "grad_norm": 0.25099000334739685, "learning_rate": 1.318631667463302e-06, "loss": 0.0061, "step": 12654 }, { "epoch": 0.8696399120395822, "grad_norm": 0.32339558005332947, "learning_rate": 1.3172632357294106e-06, "loss": 0.0087, "step": 12655 }, { "epoch": 0.8697086311159978, "grad_norm": 1.1520116329193115, "learning_rate": 1.315895481810172e-06, "loss": 0.0226, "step": 12656 }, { "epoch": 0.8697773501924134, "grad_norm": 0.58079993724823, "learning_rate": 1.3145284057733338e-06, "loss": 0.0064, "step": 12657 }, { "epoch": 0.869846069268829, "grad_norm": 0.48128998279571533, "learning_rate": 1.313162007686623e-06, "loss": 0.0093, "step": 12658 }, { "epoch": 0.8699147883452446, "grad_norm": 0.11329279839992523, "learning_rate": 1.3117962876177308e-06, "loss": 0.0031, "step": 12659 }, { "epoch": 0.8699835074216603, "grad_norm": 0.16208404302597046, "learning_rate": 1.3104312456343048e-06, "loss": 0.0063, "step": 12660 }, { "epoch": 0.8700522264980759, "grad_norm": 0.35863226652145386, "learning_rate": 1.3090668818039725e-06, "loss": 0.0058, "step": 12661 }, { "epoch": 0.8701209455744915, "grad_norm": 0.1900389939546585, "learning_rate": 1.3077031961943182e-06, "loss": 0.0068, "step": 12662 }, { "epoch": 0.8701896646509071, "grad_norm": 0.31294235587120056, "learning_rate": 1.3063401888728977e-06, "loss": 0.0027, "step": 12663 }, { "epoch": 0.8702583837273227, "grad_norm": 0.49936115741729736, "learning_rate": 1.3049778599072304e-06, "loss": 0.0104, "step": 12664 }, { "epoch": 0.8703271028037384, "grad_norm": 1.0681668519973755, "learning_rate": 1.3036162093648058e-06, "loss": 0.0092, "step": 12665 }, { "epoch": 0.870395821880154, "grad_norm": 0.1547306627035141, "learning_rate": 1.3022552373130748e-06, "loss": 0.0096, "step": 12666 }, { "epoch": 0.8704645409565696, "grad_norm": 0.37727001309394836, "learning_rate": 1.30089494381946e-06, "loss": 0.0118, "step": 12667 }, { "epoch": 0.8705332600329851, "grad_norm": 0.6421685814857483, "learning_rate": 1.2995353289513428e-06, "loss": 0.0103, "step": 12668 }, { "epoch": 0.8706019791094007, "grad_norm": 0.16012567281723022, "learning_rate": 1.2981763927760777e-06, "loss": 0.0054, "step": 12669 }, { "epoch": 0.8706706981858163, "grad_norm": 0.3054578900337219, "learning_rate": 1.2968181353609854e-06, "loss": 0.0085, "step": 12670 }, { "epoch": 0.870739417262232, "grad_norm": 1.1504805088043213, "learning_rate": 1.2954605567733525e-06, "loss": 0.0092, "step": 12671 }, { "epoch": 0.8708081363386476, "grad_norm": 0.28950443863868713, "learning_rate": 1.294103657080425e-06, "loss": 0.0126, "step": 12672 }, { "epoch": 0.8708768554150632, "grad_norm": 0.3883584439754486, "learning_rate": 1.2927474363494274e-06, "loss": 0.0094, "step": 12673 }, { "epoch": 0.8709455744914788, "grad_norm": 0.2358672171831131, "learning_rate": 1.2913918946475394e-06, "loss": 0.0152, "step": 12674 }, { "epoch": 0.8710142935678944, "grad_norm": 0.19469831883907318, "learning_rate": 1.2900370320419108e-06, "loss": 0.0141, "step": 12675 }, { "epoch": 0.8710830126443101, "grad_norm": 0.21191608905792236, "learning_rate": 1.288682848599666e-06, "loss": 0.0065, "step": 12676 }, { "epoch": 0.8711517317207257, "grad_norm": 0.20839688181877136, "learning_rate": 1.2873293443878798e-06, "loss": 0.0077, "step": 12677 }, { "epoch": 0.8712204507971413, "grad_norm": 0.2518344521522522, "learning_rate": 1.2859765194736084e-06, "loss": 0.0087, "step": 12678 }, { "epoch": 0.8712891698735569, "grad_norm": 0.18017008900642395, "learning_rate": 1.2846243739238632e-06, "loss": 0.0065, "step": 12679 }, { "epoch": 0.8713578889499725, "grad_norm": 0.5768264532089233, "learning_rate": 1.283272907805631e-06, "loss": 0.0052, "step": 12680 }, { "epoch": 0.8714266080263882, "grad_norm": 0.1712486892938614, "learning_rate": 1.2819221211858545e-06, "loss": 0.0098, "step": 12681 }, { "epoch": 0.8714953271028038, "grad_norm": 0.19053329527378082, "learning_rate": 1.280572014131457e-06, "loss": 0.0051, "step": 12682 }, { "epoch": 0.8715640461792193, "grad_norm": 0.2212485373020172, "learning_rate": 1.2792225867093132e-06, "loss": 0.0062, "step": 12683 }, { "epoch": 0.8716327652556349, "grad_norm": 0.4544523060321808, "learning_rate": 1.2778738389862765e-06, "loss": 0.0093, "step": 12684 }, { "epoch": 0.8717014843320505, "grad_norm": 0.3042454123497009, "learning_rate": 1.2765257710291567e-06, "loss": 0.0085, "step": 12685 }, { "epoch": 0.8717702034084662, "grad_norm": 0.2342834174633026, "learning_rate": 1.2751783829047386e-06, "loss": 0.0082, "step": 12686 }, { "epoch": 0.8718389224848818, "grad_norm": 0.39674508571624756, "learning_rate": 1.273831674679759e-06, "loss": 0.0128, "step": 12687 }, { "epoch": 0.8719076415612974, "grad_norm": 0.21302634477615356, "learning_rate": 1.272485646420946e-06, "loss": 0.005, "step": 12688 }, { "epoch": 0.871976360637713, "grad_norm": 0.3075675368309021, "learning_rate": 1.2711402981949682e-06, "loss": 0.0139, "step": 12689 }, { "epoch": 0.8720450797141286, "grad_norm": 0.6140947341918945, "learning_rate": 1.269795630068477e-06, "loss": 0.0107, "step": 12690 }, { "epoch": 0.8721137987905443, "grad_norm": 0.0825919434428215, "learning_rate": 1.2684516421080794e-06, "loss": 0.0056, "step": 12691 }, { "epoch": 0.8721825178669599, "grad_norm": 0.8309152722358704, "learning_rate": 1.2671083343803602e-06, "loss": 0.0092, "step": 12692 }, { "epoch": 0.8722512369433755, "grad_norm": 0.4806015193462372, "learning_rate": 1.2657657069518564e-06, "loss": 0.0082, "step": 12693 }, { "epoch": 0.8723199560197911, "grad_norm": 0.46939578652381897, "learning_rate": 1.2644237598890846e-06, "loss": 0.0147, "step": 12694 }, { "epoch": 0.8723886750962067, "grad_norm": 0.7493168711662292, "learning_rate": 1.2630824932585216e-06, "loss": 0.0315, "step": 12695 }, { "epoch": 0.8724573941726224, "grad_norm": 0.5621792674064636, "learning_rate": 1.261741907126608e-06, "loss": 0.0088, "step": 12696 }, { "epoch": 0.872526113249038, "grad_norm": 0.17297542095184326, "learning_rate": 1.2604020015597583e-06, "loss": 0.0039, "step": 12697 }, { "epoch": 0.8725948323254535, "grad_norm": 0.28162118792533875, "learning_rate": 1.2590627766243433e-06, "loss": 0.0048, "step": 12698 }, { "epoch": 0.8726635514018691, "grad_norm": 0.10776287317276001, "learning_rate": 1.2577242323867083e-06, "loss": 0.0025, "step": 12699 }, { "epoch": 0.8727322704782847, "grad_norm": 0.2004672735929489, "learning_rate": 1.2563863689131632e-06, "loss": 0.0077, "step": 12700 }, { "epoch": 0.8728009895547004, "grad_norm": 0.38507384061813354, "learning_rate": 1.255049186269982e-06, "loss": 0.008, "step": 12701 }, { "epoch": 0.872869708631116, "grad_norm": 0.29049694538116455, "learning_rate": 1.2537126845234048e-06, "loss": 0.0064, "step": 12702 }, { "epoch": 0.8729384277075316, "grad_norm": 0.4371885359287262, "learning_rate": 1.2523768637396422e-06, "loss": 0.0081, "step": 12703 }, { "epoch": 0.8730071467839472, "grad_norm": 0.1996772438287735, "learning_rate": 1.2510417239848626e-06, "loss": 0.0121, "step": 12704 }, { "epoch": 0.8730758658603628, "grad_norm": 0.09262481331825256, "learning_rate": 1.2497072653252101e-06, "loss": 0.006, "step": 12705 }, { "epoch": 0.8731445849367785, "grad_norm": 0.3158881664276123, "learning_rate": 1.24837348782679e-06, "loss": 0.0097, "step": 12706 }, { "epoch": 0.8732133040131941, "grad_norm": 0.7865718007087708, "learning_rate": 1.2470403915556777e-06, "loss": 0.0098, "step": 12707 }, { "epoch": 0.8732820230896097, "grad_norm": 0.48514696955680847, "learning_rate": 1.2457079765779072e-06, "loss": 0.0098, "step": 12708 }, { "epoch": 0.8733507421660253, "grad_norm": 0.4366108477115631, "learning_rate": 1.244376242959489e-06, "loss": 0.0172, "step": 12709 }, { "epoch": 0.8734194612424409, "grad_norm": 0.35194867849349976, "learning_rate": 1.2430451907663887e-06, "loss": 0.0102, "step": 12710 }, { "epoch": 0.8734881803188566, "grad_norm": 0.22529540956020355, "learning_rate": 1.241714820064545e-06, "loss": 0.0057, "step": 12711 }, { "epoch": 0.8735568993952721, "grad_norm": 0.3963584899902344, "learning_rate": 1.2403851309198673e-06, "loss": 0.0185, "step": 12712 }, { "epoch": 0.8736256184716877, "grad_norm": 0.29631462693214417, "learning_rate": 1.2390561233982189e-06, "loss": 0.0055, "step": 12713 }, { "epoch": 0.8736943375481033, "grad_norm": 0.26933878660202026, "learning_rate": 1.237727797565441e-06, "loss": 0.0049, "step": 12714 }, { "epoch": 0.8737630566245189, "grad_norm": 0.14608755707740784, "learning_rate": 1.2364001534873308e-06, "loss": 0.0028, "step": 12715 }, { "epoch": 0.8738317757009346, "grad_norm": 0.23323842883110046, "learning_rate": 1.2350731912296642e-06, "loss": 0.005, "step": 12716 }, { "epoch": 0.8739004947773502, "grad_norm": 0.37323859333992004, "learning_rate": 1.2337469108581651e-06, "loss": 0.0101, "step": 12717 }, { "epoch": 0.8739692138537658, "grad_norm": 0.27537664771080017, "learning_rate": 1.2324213124385493e-06, "loss": 0.0085, "step": 12718 }, { "epoch": 0.8740379329301814, "grad_norm": 0.17232906818389893, "learning_rate": 1.2310963960364724e-06, "loss": 0.0129, "step": 12719 }, { "epoch": 0.874106652006597, "grad_norm": 0.71640944480896, "learning_rate": 1.2297721617175756e-06, "loss": 0.0159, "step": 12720 }, { "epoch": 0.8741753710830127, "grad_norm": 0.18075285851955414, "learning_rate": 1.2284486095474529e-06, "loss": 0.0086, "step": 12721 }, { "epoch": 0.8742440901594283, "grad_norm": 0.409352570772171, "learning_rate": 1.2271257395916768e-06, "loss": 0.0123, "step": 12722 }, { "epoch": 0.8743128092358439, "grad_norm": 0.04962962120771408, "learning_rate": 1.2258035519157684e-06, "loss": 0.0057, "step": 12723 }, { "epoch": 0.8743815283122595, "grad_norm": 0.43682366609573364, "learning_rate": 1.2244820465852402e-06, "loss": 0.0064, "step": 12724 }, { "epoch": 0.874450247388675, "grad_norm": 0.22815954685211182, "learning_rate": 1.2231612236655464e-06, "loss": 0.0101, "step": 12725 }, { "epoch": 0.8745189664650908, "grad_norm": 0.24322712421417236, "learning_rate": 1.221841083222125e-06, "loss": 0.0069, "step": 12726 }, { "epoch": 0.8745876855415063, "grad_norm": 0.22999180853366852, "learning_rate": 1.2205216253203667e-06, "loss": 0.0107, "step": 12727 }, { "epoch": 0.8746564046179219, "grad_norm": 0.15014319121837616, "learning_rate": 1.2192028500256408e-06, "loss": 0.0084, "step": 12728 }, { "epoch": 0.8747251236943375, "grad_norm": 0.262175977230072, "learning_rate": 1.2178847574032704e-06, "loss": 0.0079, "step": 12729 }, { "epoch": 0.8747938427707531, "grad_norm": 0.5304322242736816, "learning_rate": 1.2165673475185546e-06, "loss": 0.0065, "step": 12730 }, { "epoch": 0.8748625618471688, "grad_norm": 0.3608775734901428, "learning_rate": 1.215250620436758e-06, "loss": 0.0096, "step": 12731 }, { "epoch": 0.8749312809235844, "grad_norm": 0.4725969433784485, "learning_rate": 1.213934576223102e-06, "loss": 0.0123, "step": 12732 }, { "epoch": 0.875, "grad_norm": 0.39757487177848816, "learning_rate": 1.2126192149427873e-06, "loss": 0.0045, "step": 12733 }, { "epoch": 0.8750687190764156, "grad_norm": 0.1681712567806244, "learning_rate": 1.2113045366609687e-06, "loss": 0.0082, "step": 12734 }, { "epoch": 0.8751374381528312, "grad_norm": 0.42985451221466064, "learning_rate": 1.2099905414427741e-06, "loss": 0.0051, "step": 12735 }, { "epoch": 0.8752061572292469, "grad_norm": 0.4669380784034729, "learning_rate": 1.2086772293532966e-06, "loss": 0.0092, "step": 12736 }, { "epoch": 0.8752748763056625, "grad_norm": 0.27572008967399597, "learning_rate": 1.207364600457599e-06, "loss": 0.0065, "step": 12737 }, { "epoch": 0.8753435953820781, "grad_norm": 0.5663964152336121, "learning_rate": 1.2060526548207008e-06, "loss": 0.0066, "step": 12738 }, { "epoch": 0.8754123144584937, "grad_norm": 0.19186626374721527, "learning_rate": 1.204741392507595e-06, "loss": 0.0104, "step": 12739 }, { "epoch": 0.8754810335349092, "grad_norm": 0.27027252316474915, "learning_rate": 1.2034308135832384e-06, "loss": 0.0182, "step": 12740 }, { "epoch": 0.875549752611325, "grad_norm": 0.3892441689968109, "learning_rate": 1.2021209181125537e-06, "loss": 0.0072, "step": 12741 }, { "epoch": 0.8756184716877405, "grad_norm": 0.16597163677215576, "learning_rate": 1.2008117061604324e-06, "loss": 0.0062, "step": 12742 }, { "epoch": 0.8756871907641561, "grad_norm": 0.5426186323165894, "learning_rate": 1.199503177791731e-06, "loss": 0.0047, "step": 12743 }, { "epoch": 0.8757559098405717, "grad_norm": 0.21231955289840698, "learning_rate": 1.1981953330712676e-06, "loss": 0.0082, "step": 12744 }, { "epoch": 0.8758246289169873, "grad_norm": 0.24158838391304016, "learning_rate": 1.1968881720638337e-06, "loss": 0.0083, "step": 12745 }, { "epoch": 0.875893347993403, "grad_norm": 1.3354450464248657, "learning_rate": 1.1955816948341791e-06, "loss": 0.0083, "step": 12746 }, { "epoch": 0.8759620670698186, "grad_norm": 0.11927381902933121, "learning_rate": 1.1942759014470272e-06, "loss": 0.006, "step": 12747 }, { "epoch": 0.8760307861462342, "grad_norm": 0.5534182190895081, "learning_rate": 1.192970791967068e-06, "loss": 0.0065, "step": 12748 }, { "epoch": 0.8760995052226498, "grad_norm": 0.5540397763252258, "learning_rate": 1.1916663664589444e-06, "loss": 0.0145, "step": 12749 }, { "epoch": 0.8761682242990654, "grad_norm": 0.2433490753173828, "learning_rate": 1.190362624987285e-06, "loss": 0.0107, "step": 12750 }, { "epoch": 0.8762369433754811, "grad_norm": 0.1949215680360794, "learning_rate": 1.189059567616665e-06, "loss": 0.0104, "step": 12751 }, { "epoch": 0.8763056624518967, "grad_norm": 0.22689110040664673, "learning_rate": 1.1877571944116439e-06, "loss": 0.0076, "step": 12752 }, { "epoch": 0.8763743815283123, "grad_norm": 0.21240468323230743, "learning_rate": 1.1864555054367304e-06, "loss": 0.007, "step": 12753 }, { "epoch": 0.8764431006047279, "grad_norm": 0.49706214666366577, "learning_rate": 1.1851545007564146e-06, "loss": 0.0049, "step": 12754 }, { "epoch": 0.8765118196811434, "grad_norm": 0.15115535259246826, "learning_rate": 1.1838541804351416e-06, "loss": 0.0059, "step": 12755 }, { "epoch": 0.8765805387575591, "grad_norm": 0.36404699087142944, "learning_rate": 1.1825545445373297e-06, "loss": 0.0084, "step": 12756 }, { "epoch": 0.8766492578339747, "grad_norm": 0.22870808839797974, "learning_rate": 1.181255593127356e-06, "loss": 0.0069, "step": 12757 }, { "epoch": 0.8767179769103903, "grad_norm": 0.10622592270374298, "learning_rate": 1.1799573262695724e-06, "loss": 0.0086, "step": 12758 }, { "epoch": 0.8767866959868059, "grad_norm": 0.7657299041748047, "learning_rate": 1.1786597440282853e-06, "loss": 0.0157, "step": 12759 }, { "epoch": 0.8768554150632215, "grad_norm": 0.4760562479496002, "learning_rate": 1.1773628464677837e-06, "loss": 0.007, "step": 12760 }, { "epoch": 0.8769241341396372, "grad_norm": 0.35592231154441833, "learning_rate": 1.1760666336523063e-06, "loss": 0.0071, "step": 12761 }, { "epoch": 0.8769928532160528, "grad_norm": 0.2868415117263794, "learning_rate": 1.1747711056460698e-06, "loss": 0.013, "step": 12762 }, { "epoch": 0.8770615722924684, "grad_norm": 0.5171084403991699, "learning_rate": 1.1734762625132462e-06, "loss": 0.0151, "step": 12763 }, { "epoch": 0.877130291368884, "grad_norm": 0.23669567704200745, "learning_rate": 1.172182104317986e-06, "loss": 0.0102, "step": 12764 }, { "epoch": 0.8771990104452996, "grad_norm": 0.2653319239616394, "learning_rate": 1.1708886311243928e-06, "loss": 0.0077, "step": 12765 }, { "epoch": 0.8772677295217153, "grad_norm": 0.27718082070350647, "learning_rate": 1.1695958429965437e-06, "loss": 0.0091, "step": 12766 }, { "epoch": 0.8773364485981309, "grad_norm": 0.21607400476932526, "learning_rate": 1.1683037399984858e-06, "loss": 0.0112, "step": 12767 }, { "epoch": 0.8774051676745465, "grad_norm": 0.27051228284835815, "learning_rate": 1.1670123221942213e-06, "loss": 0.0082, "step": 12768 }, { "epoch": 0.877473886750962, "grad_norm": 0.2712872326374054, "learning_rate": 1.1657215896477307e-06, "loss": 0.0078, "step": 12769 }, { "epoch": 0.8775426058273776, "grad_norm": 0.19869239628314972, "learning_rate": 1.1644315424229463e-06, "loss": 0.0123, "step": 12770 }, { "epoch": 0.8776113249037933, "grad_norm": 0.43916353583335876, "learning_rate": 1.1631421805837817e-06, "loss": 0.0129, "step": 12771 }, { "epoch": 0.8776800439802089, "grad_norm": 0.4235467314720154, "learning_rate": 1.1618535041941026e-06, "loss": 0.0074, "step": 12772 }, { "epoch": 0.8777487630566245, "grad_norm": 0.2042032927274704, "learning_rate": 1.1605655133177529e-06, "loss": 0.0055, "step": 12773 }, { "epoch": 0.8778174821330401, "grad_norm": 0.21227556467056274, "learning_rate": 1.159278208018535e-06, "loss": 0.0155, "step": 12774 }, { "epoch": 0.8778862012094557, "grad_norm": 0.24835911393165588, "learning_rate": 1.157991588360221e-06, "loss": 0.008, "step": 12775 }, { "epoch": 0.8779549202858714, "grad_norm": 0.9353353977203369, "learning_rate": 1.1567056544065435e-06, "loss": 0.0185, "step": 12776 }, { "epoch": 0.878023639362287, "grad_norm": 0.1430206298828125, "learning_rate": 1.155420406221208e-06, "loss": 0.0068, "step": 12777 }, { "epoch": 0.8780923584387026, "grad_norm": 0.14021018147468567, "learning_rate": 1.154135843867879e-06, "loss": 0.0079, "step": 12778 }, { "epoch": 0.8781610775151182, "grad_norm": 0.09115543961524963, "learning_rate": 1.1528519674102e-06, "loss": 0.0076, "step": 12779 }, { "epoch": 0.8782297965915338, "grad_norm": 0.27810022234916687, "learning_rate": 1.1515687769117623e-06, "loss": 0.0064, "step": 12780 }, { "epoch": 0.8782985156679495, "grad_norm": 0.187382310628891, "learning_rate": 1.1502862724361396e-06, "loss": 0.0074, "step": 12781 }, { "epoch": 0.8783672347443651, "grad_norm": 0.547356128692627, "learning_rate": 1.149004454046858e-06, "loss": 0.0063, "step": 12782 }, { "epoch": 0.8784359538207807, "grad_norm": 0.2902079224586487, "learning_rate": 1.1477233218074217e-06, "loss": 0.0084, "step": 12783 }, { "epoch": 0.8785046728971962, "grad_norm": 0.41731205582618713, "learning_rate": 1.14644287578129e-06, "loss": 0.0079, "step": 12784 }, { "epoch": 0.8785733919736118, "grad_norm": 0.46014976501464844, "learning_rate": 1.1451631160318987e-06, "loss": 0.0061, "step": 12785 }, { "epoch": 0.8786421110500275, "grad_norm": 0.08571106940507889, "learning_rate": 1.1438840426226422e-06, "loss": 0.005, "step": 12786 }, { "epoch": 0.8787108301264431, "grad_norm": 0.2303869128227234, "learning_rate": 1.1426056556168828e-06, "loss": 0.0098, "step": 12787 }, { "epoch": 0.8787795492028587, "grad_norm": 0.3759342133998871, "learning_rate": 1.1413279550779504e-06, "loss": 0.0083, "step": 12788 }, { "epoch": 0.8788482682792743, "grad_norm": 0.18716607987880707, "learning_rate": 1.1400509410691373e-06, "loss": 0.0123, "step": 12789 }, { "epoch": 0.8789169873556899, "grad_norm": 0.24890731275081635, "learning_rate": 1.1387746136537063e-06, "loss": 0.0055, "step": 12790 }, { "epoch": 0.8789857064321056, "grad_norm": 0.7106258869171143, "learning_rate": 1.137498972894882e-06, "loss": 0.016, "step": 12791 }, { "epoch": 0.8790544255085212, "grad_norm": 0.3009225130081177, "learning_rate": 1.1362240188558621e-06, "loss": 0.015, "step": 12792 }, { "epoch": 0.8791231445849368, "grad_norm": 0.23516342043876648, "learning_rate": 1.1349497515997991e-06, "loss": 0.0112, "step": 12793 }, { "epoch": 0.8791918636613524, "grad_norm": 0.3700469434261322, "learning_rate": 1.133676171189823e-06, "loss": 0.0105, "step": 12794 }, { "epoch": 0.879260582737768, "grad_norm": 0.3199313282966614, "learning_rate": 1.132403277689018e-06, "loss": 0.0106, "step": 12795 }, { "epoch": 0.8793293018141837, "grad_norm": 0.3406464457511902, "learning_rate": 1.131131071160444e-06, "loss": 0.0129, "step": 12796 }, { "epoch": 0.8793980208905993, "grad_norm": 0.7131968140602112, "learning_rate": 1.1298595516671235e-06, "loss": 0.0139, "step": 12797 }, { "epoch": 0.8794667399670149, "grad_norm": 0.840303897857666, "learning_rate": 1.1285887192720479e-06, "loss": 0.0103, "step": 12798 }, { "epoch": 0.8795354590434304, "grad_norm": 0.3451957404613495, "learning_rate": 1.1273185740381652e-06, "loss": 0.0074, "step": 12799 }, { "epoch": 0.879604178119846, "grad_norm": 0.4279353618621826, "learning_rate": 1.1260491160284037e-06, "loss": 0.0076, "step": 12800 }, { "epoch": 0.8796728971962616, "grad_norm": 0.35885247588157654, "learning_rate": 1.124780345305641e-06, "loss": 0.0112, "step": 12801 }, { "epoch": 0.8797416162726773, "grad_norm": 0.38329124450683594, "learning_rate": 1.1235122619327353e-06, "loss": 0.012, "step": 12802 }, { "epoch": 0.8798103353490929, "grad_norm": 0.2084348201751709, "learning_rate": 1.1222448659725048e-06, "loss": 0.0084, "step": 12803 }, { "epoch": 0.8798790544255085, "grad_norm": 0.3348241448402405, "learning_rate": 1.120978157487731e-06, "loss": 0.0105, "step": 12804 }, { "epoch": 0.8799477735019241, "grad_norm": 0.4868318438529968, "learning_rate": 1.1197121365411666e-06, "loss": 0.0117, "step": 12805 }, { "epoch": 0.8800164925783397, "grad_norm": 0.25900790095329285, "learning_rate": 1.1184468031955236e-06, "loss": 0.0046, "step": 12806 }, { "epoch": 0.8800852116547554, "grad_norm": 0.8705852627754211, "learning_rate": 1.1171821575134867e-06, "loss": 0.0048, "step": 12807 }, { "epoch": 0.880153930731171, "grad_norm": 0.27105581760406494, "learning_rate": 1.1159181995577039e-06, "loss": 0.0046, "step": 12808 }, { "epoch": 0.8802226498075866, "grad_norm": 0.21303540468215942, "learning_rate": 1.114654929390792e-06, "loss": 0.0038, "step": 12809 }, { "epoch": 0.8802913688840022, "grad_norm": 0.34842610359191895, "learning_rate": 1.113392347075326e-06, "loss": 0.0057, "step": 12810 }, { "epoch": 0.8803600879604178, "grad_norm": 0.0843309760093689, "learning_rate": 1.1121304526738556e-06, "loss": 0.0046, "step": 12811 }, { "epoch": 0.8804288070368335, "grad_norm": 0.3825000822544098, "learning_rate": 1.1108692462488863e-06, "loss": 0.0094, "step": 12812 }, { "epoch": 0.880497526113249, "grad_norm": 0.19337473809719086, "learning_rate": 1.1096087278629042e-06, "loss": 0.0044, "step": 12813 }, { "epoch": 0.8805662451896646, "grad_norm": 0.4842570722103119, "learning_rate": 1.1083488975783412e-06, "loss": 0.0089, "step": 12814 }, { "epoch": 0.8806349642660802, "grad_norm": 0.11704646050930023, "learning_rate": 1.107089755457621e-06, "loss": 0.0041, "step": 12815 }, { "epoch": 0.8807036833424958, "grad_norm": 0.8610244393348694, "learning_rate": 1.1058313015631084e-06, "loss": 0.0118, "step": 12816 }, { "epoch": 0.8807724024189115, "grad_norm": 0.4946258068084717, "learning_rate": 1.1045735359571502e-06, "loss": 0.0099, "step": 12817 }, { "epoch": 0.8808411214953271, "grad_norm": 0.16007541120052338, "learning_rate": 1.1033164587020483e-06, "loss": 0.0079, "step": 12818 }, { "epoch": 0.8809098405717427, "grad_norm": 0.4732763171195984, "learning_rate": 1.1020600698600814e-06, "loss": 0.0105, "step": 12819 }, { "epoch": 0.8809785596481583, "grad_norm": 0.33077380061149597, "learning_rate": 1.1008043694934826e-06, "loss": 0.0146, "step": 12820 }, { "epoch": 0.8810472787245739, "grad_norm": 0.16291461884975433, "learning_rate": 1.0995493576644612e-06, "loss": 0.0019, "step": 12821 }, { "epoch": 0.8811159978009896, "grad_norm": 0.8485091328620911, "learning_rate": 1.0982950344351866e-06, "loss": 0.013, "step": 12822 }, { "epoch": 0.8811847168774052, "grad_norm": 0.2090265303850174, "learning_rate": 1.0970413998677965e-06, "loss": 0.0079, "step": 12823 }, { "epoch": 0.8812534359538208, "grad_norm": 0.10007616877555847, "learning_rate": 1.0957884540243907e-06, "loss": 0.0117, "step": 12824 }, { "epoch": 0.8813221550302364, "grad_norm": 0.27030107378959656, "learning_rate": 1.0945361969670398e-06, "loss": 0.0066, "step": 12825 }, { "epoch": 0.881390874106652, "grad_norm": 0.2183801829814911, "learning_rate": 1.0932846287577774e-06, "loss": 0.0063, "step": 12826 }, { "epoch": 0.8814595931830677, "grad_norm": 0.1426522582769394, "learning_rate": 1.0920337494586008e-06, "loss": 0.0121, "step": 12827 }, { "epoch": 0.8815283122594832, "grad_norm": 0.4899061620235443, "learning_rate": 1.0907835591314836e-06, "loss": 0.0044, "step": 12828 }, { "epoch": 0.8815970313358988, "grad_norm": 0.23551355302333832, "learning_rate": 1.0895340578383495e-06, "loss": 0.009, "step": 12829 }, { "epoch": 0.8816657504123144, "grad_norm": 0.21088165044784546, "learning_rate": 1.0882852456411008e-06, "loss": 0.006, "step": 12830 }, { "epoch": 0.88173446948873, "grad_norm": 0.40873390436172485, "learning_rate": 1.0870371226015979e-06, "loss": 0.0088, "step": 12831 }, { "epoch": 0.8818031885651457, "grad_norm": 0.19457565248012543, "learning_rate": 1.0857896887816733e-06, "loss": 0.0068, "step": 12832 }, { "epoch": 0.8818719076415613, "grad_norm": 0.06422916054725647, "learning_rate": 1.0845429442431204e-06, "loss": 0.0056, "step": 12833 }, { "epoch": 0.8819406267179769, "grad_norm": 0.3679114580154419, "learning_rate": 1.0832968890477018e-06, "loss": 0.0121, "step": 12834 }, { "epoch": 0.8820093457943925, "grad_norm": 0.1486632376909256, "learning_rate": 1.0820515232571431e-06, "loss": 0.0061, "step": 12835 }, { "epoch": 0.8820780648708081, "grad_norm": 0.6235010027885437, "learning_rate": 1.08080684693314e-06, "loss": 0.0087, "step": 12836 }, { "epoch": 0.8821467839472238, "grad_norm": 0.2964731454849243, "learning_rate": 1.0795628601373464e-06, "loss": 0.0132, "step": 12837 }, { "epoch": 0.8822155030236394, "grad_norm": 0.5362824201583862, "learning_rate": 1.0783195629313897e-06, "loss": 0.0121, "step": 12838 }, { "epoch": 0.882284222100055, "grad_norm": 0.49881044030189514, "learning_rate": 1.0770769553768606e-06, "loss": 0.0086, "step": 12839 }, { "epoch": 0.8823529411764706, "grad_norm": 0.2756277024745941, "learning_rate": 1.0758350375353164e-06, "loss": 0.0082, "step": 12840 }, { "epoch": 0.8824216602528862, "grad_norm": 0.5718286037445068, "learning_rate": 1.0745938094682766e-06, "loss": 0.018, "step": 12841 }, { "epoch": 0.8824903793293019, "grad_norm": 0.5759645700454712, "learning_rate": 1.0733532712372319e-06, "loss": 0.007, "step": 12842 }, { "epoch": 0.8825590984057174, "grad_norm": 0.2562035322189331, "learning_rate": 1.0721134229036311e-06, "loss": 0.0096, "step": 12843 }, { "epoch": 0.882627817482133, "grad_norm": 0.31487494707107544, "learning_rate": 1.0708742645288954e-06, "loss": 0.0097, "step": 12844 }, { "epoch": 0.8826965365585486, "grad_norm": 0.08513403683900833, "learning_rate": 1.0696357961744157e-06, "loss": 0.0081, "step": 12845 }, { "epoch": 0.8827652556349642, "grad_norm": 0.39417603611946106, "learning_rate": 1.0683980179015363e-06, "loss": 0.008, "step": 12846 }, { "epoch": 0.8828339747113799, "grad_norm": 0.21639429032802582, "learning_rate": 1.0671609297715794e-06, "loss": 0.0056, "step": 12847 }, { "epoch": 0.8829026937877955, "grad_norm": 0.6028327345848083, "learning_rate": 1.0659245318458234e-06, "loss": 0.0123, "step": 12848 }, { "epoch": 0.8829714128642111, "grad_norm": 0.36131811141967773, "learning_rate": 1.0646888241855219e-06, "loss": 0.0038, "step": 12849 }, { "epoch": 0.8830401319406267, "grad_norm": 0.48781779408454895, "learning_rate": 1.0634538068518795e-06, "loss": 0.0094, "step": 12850 }, { "epoch": 0.8831088510170423, "grad_norm": 0.19169016182422638, "learning_rate": 1.0622194799060892e-06, "loss": 0.006, "step": 12851 }, { "epoch": 0.883177570093458, "grad_norm": 0.16922342777252197, "learning_rate": 1.0609858434092883e-06, "loss": 0.0059, "step": 12852 }, { "epoch": 0.8832462891698736, "grad_norm": 0.29411524534225464, "learning_rate": 1.0597528974225933e-06, "loss": 0.0036, "step": 12853 }, { "epoch": 0.8833150082462892, "grad_norm": 0.16592241823673248, "learning_rate": 1.0585206420070787e-06, "loss": 0.0085, "step": 12854 }, { "epoch": 0.8833837273227048, "grad_norm": 0.48242488503456116, "learning_rate": 1.0572890772237902e-06, "loss": 0.0103, "step": 12855 }, { "epoch": 0.8834524463991204, "grad_norm": 0.25172916054725647, "learning_rate": 1.0560582031337313e-06, "loss": 0.0074, "step": 12856 }, { "epoch": 0.883521165475536, "grad_norm": 0.3146742880344391, "learning_rate": 1.054828019797886e-06, "loss": 0.0089, "step": 12857 }, { "epoch": 0.8835898845519516, "grad_norm": 0.2553166151046753, "learning_rate": 1.0535985272771892e-06, "loss": 0.011, "step": 12858 }, { "epoch": 0.8836586036283672, "grad_norm": 0.253880113363266, "learning_rate": 1.0523697256325505e-06, "loss": 0.0164, "step": 12859 }, { "epoch": 0.8837273227047828, "grad_norm": 0.7379122972488403, "learning_rate": 1.0511416149248375e-06, "loss": 0.0149, "step": 12860 }, { "epoch": 0.8837960417811984, "grad_norm": 0.14900176227092743, "learning_rate": 1.0499141952148955e-06, "loss": 0.007, "step": 12861 }, { "epoch": 0.8838647608576141, "grad_norm": 0.2691930830478668, "learning_rate": 1.0486874665635205e-06, "loss": 0.01, "step": 12862 }, { "epoch": 0.8839334799340297, "grad_norm": 0.2978416383266449, "learning_rate": 1.0474614290314854e-06, "loss": 0.0053, "step": 12863 }, { "epoch": 0.8840021990104453, "grad_norm": 0.25201213359832764, "learning_rate": 1.0462360826795286e-06, "loss": 0.0088, "step": 12864 }, { "epoch": 0.8840709180868609, "grad_norm": 0.3079017996788025, "learning_rate": 1.0450114275683464e-06, "loss": 0.0097, "step": 12865 }, { "epoch": 0.8841396371632765, "grad_norm": 0.5500103831291199, "learning_rate": 1.0437874637586087e-06, "loss": 0.0078, "step": 12866 }, { "epoch": 0.8842083562396922, "grad_norm": 0.4502878785133362, "learning_rate": 1.0425641913109468e-06, "loss": 0.0095, "step": 12867 }, { "epoch": 0.8842770753161078, "grad_norm": 0.25304606556892395, "learning_rate": 1.041341610285959e-06, "loss": 0.007, "step": 12868 }, { "epoch": 0.8843457943925234, "grad_norm": 0.29561975598335266, "learning_rate": 1.0401197207442103e-06, "loss": 0.0106, "step": 12869 }, { "epoch": 0.884414513468939, "grad_norm": 0.4148175120353699, "learning_rate": 1.0388985227462323e-06, "loss": 0.0104, "step": 12870 }, { "epoch": 0.8844832325453545, "grad_norm": 0.5052017569541931, "learning_rate": 1.0376780163525162e-06, "loss": 0.0053, "step": 12871 }, { "epoch": 0.8845519516217702, "grad_norm": 0.2990131974220276, "learning_rate": 1.0364582016235292e-06, "loss": 0.0113, "step": 12872 }, { "epoch": 0.8846206706981858, "grad_norm": 0.2968856990337372, "learning_rate": 1.0352390786196907e-06, "loss": 0.0062, "step": 12873 }, { "epoch": 0.8846893897746014, "grad_norm": 0.22952553629875183, "learning_rate": 1.0340206474013996e-06, "loss": 0.0094, "step": 12874 }, { "epoch": 0.884758108851017, "grad_norm": 0.6413889527320862, "learning_rate": 1.0328029080290124e-06, "loss": 0.0117, "step": 12875 }, { "epoch": 0.8848268279274326, "grad_norm": 0.2979257106781006, "learning_rate": 1.0315858605628559e-06, "loss": 0.0097, "step": 12876 }, { "epoch": 0.8848955470038483, "grad_norm": 0.3325539827346802, "learning_rate": 1.0303695050632168e-06, "loss": 0.0034, "step": 12877 }, { "epoch": 0.8849642660802639, "grad_norm": 0.9829308986663818, "learning_rate": 1.0291538415903534e-06, "loss": 0.0112, "step": 12878 }, { "epoch": 0.8850329851566795, "grad_norm": 0.15462574362754822, "learning_rate": 1.0279388702044828e-06, "loss": 0.0027, "step": 12879 }, { "epoch": 0.8851017042330951, "grad_norm": 0.41293635964393616, "learning_rate": 1.0267245909657968e-06, "loss": 0.0118, "step": 12880 }, { "epoch": 0.8851704233095107, "grad_norm": 0.3339191675186157, "learning_rate": 1.0255110039344473e-06, "loss": 0.0069, "step": 12881 }, { "epoch": 0.8852391423859264, "grad_norm": 0.2708374559879303, "learning_rate": 1.024298109170551e-06, "loss": 0.0085, "step": 12882 }, { "epoch": 0.885307861462342, "grad_norm": 0.1695244014263153, "learning_rate": 1.0230859067341968e-06, "loss": 0.0035, "step": 12883 }, { "epoch": 0.8853765805387576, "grad_norm": 0.3468039333820343, "learning_rate": 1.0218743966854282e-06, "loss": 0.0113, "step": 12884 }, { "epoch": 0.8854452996151732, "grad_norm": 1.2732781171798706, "learning_rate": 1.0206635790842655e-06, "loss": 0.0209, "step": 12885 }, { "epoch": 0.8855140186915887, "grad_norm": 0.7082263827323914, "learning_rate": 1.0194534539906859e-06, "loss": 0.0063, "step": 12886 }, { "epoch": 0.8855827377680044, "grad_norm": 0.316860169172287, "learning_rate": 1.0182440214646426e-06, "loss": 0.0069, "step": 12887 }, { "epoch": 0.88565145684442, "grad_norm": 0.40790656208992004, "learning_rate": 1.0170352815660433e-06, "loss": 0.0151, "step": 12888 }, { "epoch": 0.8857201759208356, "grad_norm": 0.3188481330871582, "learning_rate": 1.0158272343547715e-06, "loss": 0.0117, "step": 12889 }, { "epoch": 0.8857888949972512, "grad_norm": 0.3074805438518524, "learning_rate": 1.014619879890666e-06, "loss": 0.0084, "step": 12890 }, { "epoch": 0.8858576140736668, "grad_norm": 0.38042694330215454, "learning_rate": 1.0134132182335403e-06, "loss": 0.0118, "step": 12891 }, { "epoch": 0.8859263331500825, "grad_norm": 0.3229607343673706, "learning_rate": 1.012207249443164e-06, "loss": 0.0082, "step": 12892 }, { "epoch": 0.8859950522264981, "grad_norm": 0.3932640552520752, "learning_rate": 1.0110019735792886e-06, "loss": 0.0071, "step": 12893 }, { "epoch": 0.8860637713029137, "grad_norm": 0.24752697348594666, "learning_rate": 1.0097973907016117e-06, "loss": 0.0068, "step": 12894 }, { "epoch": 0.8861324903793293, "grad_norm": 0.05446576327085495, "learning_rate": 1.008593500869812e-06, "loss": 0.0031, "step": 12895 }, { "epoch": 0.8862012094557449, "grad_norm": 0.41712936758995056, "learning_rate": 1.0073903041435224e-06, "loss": 0.003, "step": 12896 }, { "epoch": 0.8862699285321606, "grad_norm": 0.35560372471809387, "learning_rate": 1.0061878005823527e-06, "loss": 0.0047, "step": 12897 }, { "epoch": 0.8863386476085762, "grad_norm": 0.26934105157852173, "learning_rate": 1.0049859902458675e-06, "loss": 0.0113, "step": 12898 }, { "epoch": 0.8864073666849918, "grad_norm": 0.25799572467803955, "learning_rate": 1.0037848731936022e-06, "loss": 0.0092, "step": 12899 }, { "epoch": 0.8864760857614074, "grad_norm": 0.14597591757774353, "learning_rate": 1.0025844494850627e-06, "loss": 0.0086, "step": 12900 }, { "epoch": 0.8865448048378229, "grad_norm": 0.538382887840271, "learning_rate": 1.00138471917971e-06, "loss": 0.0138, "step": 12901 }, { "epoch": 0.8866135239142386, "grad_norm": 0.2741731107234955, "learning_rate": 1.000185682336981e-06, "loss": 0.0083, "step": 12902 }, { "epoch": 0.8866822429906542, "grad_norm": 0.8097614645957947, "learning_rate": 9.989873390162669e-07, "loss": 0.0094, "step": 12903 }, { "epoch": 0.8867509620670698, "grad_norm": 0.2661722004413605, "learning_rate": 9.977896892769368e-07, "loss": 0.0049, "step": 12904 }, { "epoch": 0.8868196811434854, "grad_norm": 0.26183342933654785, "learning_rate": 9.965927331783164e-07, "loss": 0.0079, "step": 12905 }, { "epoch": 0.886888400219901, "grad_norm": 0.22107601165771484, "learning_rate": 9.953964707797069e-07, "loss": 0.0061, "step": 12906 }, { "epoch": 0.8869571192963167, "grad_norm": 0.22693948447704315, "learning_rate": 9.942009021403608e-07, "loss": 0.012, "step": 12907 }, { "epoch": 0.8870258383727323, "grad_norm": 0.3006129562854767, "learning_rate": 9.93006027319509e-07, "loss": 0.0111, "step": 12908 }, { "epoch": 0.8870945574491479, "grad_norm": 0.5832507014274597, "learning_rate": 9.918118463763392e-07, "loss": 0.0134, "step": 12909 }, { "epoch": 0.8871632765255635, "grad_norm": 0.1122809499502182, "learning_rate": 9.906183593700108e-07, "loss": 0.0045, "step": 12910 }, { "epoch": 0.8872319956019791, "grad_norm": 0.413222998380661, "learning_rate": 9.894255663596464e-07, "loss": 0.0138, "step": 12911 }, { "epoch": 0.8873007146783948, "grad_norm": 0.2436075657606125, "learning_rate": 9.882334674043387e-07, "loss": 0.0149, "step": 12912 }, { "epoch": 0.8873694337548104, "grad_norm": 0.19097968935966492, "learning_rate": 9.870420625631354e-07, "loss": 0.0262, "step": 12913 }, { "epoch": 0.887438152831226, "grad_norm": 0.25169289112091064, "learning_rate": 9.858513518950597e-07, "loss": 0.0061, "step": 12914 }, { "epoch": 0.8875068719076415, "grad_norm": 0.5574100613594055, "learning_rate": 9.846613354590956e-07, "loss": 0.0153, "step": 12915 }, { "epoch": 0.8875755909840571, "grad_norm": 0.46217989921569824, "learning_rate": 9.834720133141944e-07, "loss": 0.0072, "step": 12916 }, { "epoch": 0.8876443100604728, "grad_norm": 0.2788239121437073, "learning_rate": 9.822833855192743e-07, "loss": 0.007, "step": 12917 }, { "epoch": 0.8877130291368884, "grad_norm": 0.4775969982147217, "learning_rate": 9.810954521332144e-07, "loss": 0.0089, "step": 12918 }, { "epoch": 0.887781748213304, "grad_norm": 0.4873567223548889, "learning_rate": 9.799082132148662e-07, "loss": 0.0073, "step": 12919 }, { "epoch": 0.8878504672897196, "grad_norm": 0.24413664638996124, "learning_rate": 9.787216688230393e-07, "loss": 0.0061, "step": 12920 }, { "epoch": 0.8879191863661352, "grad_norm": 0.34490400552749634, "learning_rate": 9.775358190165152e-07, "loss": 0.0121, "step": 12921 }, { "epoch": 0.8879879054425509, "grad_norm": 0.31218019127845764, "learning_rate": 9.763506638540349e-07, "loss": 0.0088, "step": 12922 }, { "epoch": 0.8880566245189665, "grad_norm": 0.31752821803092957, "learning_rate": 9.75166203394315e-07, "loss": 0.0056, "step": 12923 }, { "epoch": 0.8881253435953821, "grad_norm": 0.27020972967147827, "learning_rate": 9.739824376960248e-07, "loss": 0.0072, "step": 12924 }, { "epoch": 0.8881940626717977, "grad_norm": 0.2785663306713104, "learning_rate": 9.727993668178114e-07, "loss": 0.0095, "step": 12925 }, { "epoch": 0.8882627817482133, "grad_norm": 0.34993061423301697, "learning_rate": 9.71616990818276e-07, "loss": 0.0069, "step": 12926 }, { "epoch": 0.888331500824629, "grad_norm": 0.4354965388774872, "learning_rate": 9.704353097559965e-07, "loss": 0.0093, "step": 12927 }, { "epoch": 0.8884002199010446, "grad_norm": 0.715351402759552, "learning_rate": 9.69254323689503e-07, "loss": 0.0125, "step": 12928 }, { "epoch": 0.8884689389774602, "grad_norm": 0.393398642539978, "learning_rate": 9.680740326773101e-07, "loss": 0.008, "step": 12929 }, { "epoch": 0.8885376580538757, "grad_norm": 0.4148755669593811, "learning_rate": 9.668944367778782e-07, "loss": 0.0115, "step": 12930 }, { "epoch": 0.8886063771302913, "grad_norm": 0.5170483589172363, "learning_rate": 9.657155360496467e-07, "loss": 0.0085, "step": 12931 }, { "epoch": 0.8886750962067069, "grad_norm": 0.08622397482395172, "learning_rate": 9.645373305510124e-07, "loss": 0.0049, "step": 12932 }, { "epoch": 0.8887438152831226, "grad_norm": 0.21064019203186035, "learning_rate": 9.633598203403454e-07, "loss": 0.0115, "step": 12933 }, { "epoch": 0.8888125343595382, "grad_norm": 0.37861713767051697, "learning_rate": 9.621830054759723e-07, "loss": 0.0076, "step": 12934 }, { "epoch": 0.8888812534359538, "grad_norm": 0.3211149275302887, "learning_rate": 9.61006886016193e-07, "loss": 0.0044, "step": 12935 }, { "epoch": 0.8889499725123694, "grad_norm": 0.4072679281234741, "learning_rate": 9.598314620192712e-07, "loss": 0.0086, "step": 12936 }, { "epoch": 0.889018691588785, "grad_norm": 1.2815518379211426, "learning_rate": 9.586567335434314e-07, "loss": 0.0109, "step": 12937 }, { "epoch": 0.8890874106652007, "grad_norm": 0.3008289039134979, "learning_rate": 9.574827006468728e-07, "loss": 0.007, "step": 12938 }, { "epoch": 0.8891561297416163, "grad_norm": 0.6930274367332458, "learning_rate": 9.563093633877467e-07, "loss": 0.0097, "step": 12939 }, { "epoch": 0.8892248488180319, "grad_norm": 0.3631626069545746, "learning_rate": 9.551367218241835e-07, "loss": 0.0067, "step": 12940 }, { "epoch": 0.8892935678944475, "grad_norm": 0.37376293540000916, "learning_rate": 9.539647760142717e-07, "loss": 0.0063, "step": 12941 }, { "epoch": 0.8893622869708631, "grad_norm": 0.23080851137638092, "learning_rate": 9.527935260160697e-07, "loss": 0.0066, "step": 12942 }, { "epoch": 0.8894310060472788, "grad_norm": 0.2211480438709259, "learning_rate": 9.516229718875946e-07, "loss": 0.0048, "step": 12943 }, { "epoch": 0.8894997251236944, "grad_norm": 0.23083262145519257, "learning_rate": 9.504531136868366e-07, "loss": 0.0089, "step": 12944 }, { "epoch": 0.8895684442001099, "grad_norm": 0.18827436864376068, "learning_rate": 9.492839514717461e-07, "loss": 0.0064, "step": 12945 }, { "epoch": 0.8896371632765255, "grad_norm": 0.17658475041389465, "learning_rate": 9.4811548530024e-07, "loss": 0.0074, "step": 12946 }, { "epoch": 0.8897058823529411, "grad_norm": 0.15115617215633392, "learning_rate": 9.469477152302036e-07, "loss": 0.0059, "step": 12947 }, { "epoch": 0.8897746014293568, "grad_norm": 0.478581964969635, "learning_rate": 9.457806413194892e-07, "loss": 0.014, "step": 12948 }, { "epoch": 0.8898433205057724, "grad_norm": 0.15980572998523712, "learning_rate": 9.446142636259037e-07, "loss": 0.0093, "step": 12949 }, { "epoch": 0.889912039582188, "grad_norm": 0.4206671416759491, "learning_rate": 9.434485822072342e-07, "loss": 0.0063, "step": 12950 }, { "epoch": 0.8899807586586036, "grad_norm": 0.2768923044204712, "learning_rate": 9.422835971212196e-07, "loss": 0.0069, "step": 12951 }, { "epoch": 0.8900494777350192, "grad_norm": 0.15924417972564697, "learning_rate": 9.411193084255753e-07, "loss": 0.0038, "step": 12952 }, { "epoch": 0.8901181968114349, "grad_norm": 0.25569671392440796, "learning_rate": 9.399557161779771e-07, "loss": 0.0088, "step": 12953 }, { "epoch": 0.8901869158878505, "grad_norm": 0.2862330377101898, "learning_rate": 9.387928204360669e-07, "loss": 0.0056, "step": 12954 }, { "epoch": 0.8902556349642661, "grad_norm": 0.37977805733680725, "learning_rate": 9.37630621257452e-07, "loss": 0.0101, "step": 12955 }, { "epoch": 0.8903243540406817, "grad_norm": 0.3123561441898346, "learning_rate": 9.364691186997032e-07, "loss": 0.0067, "step": 12956 }, { "epoch": 0.8903930731170973, "grad_norm": 0.11716071516275406, "learning_rate": 9.353083128203627e-07, "loss": 0.0063, "step": 12957 }, { "epoch": 0.890461792193513, "grad_norm": 0.5266055464744568, "learning_rate": 9.341482036769294e-07, "loss": 0.0101, "step": 12958 }, { "epoch": 0.8905305112699285, "grad_norm": 0.39931830763816833, "learning_rate": 9.329887913268792e-07, "loss": 0.0108, "step": 12959 }, { "epoch": 0.8905992303463441, "grad_norm": 0.22127008438110352, "learning_rate": 9.318300758276427e-07, "loss": 0.0076, "step": 12960 }, { "epoch": 0.8906679494227597, "grad_norm": 0.5156329274177551, "learning_rate": 9.306720572366223e-07, "loss": 0.0099, "step": 12961 }, { "epoch": 0.8907366684991753, "grad_norm": 0.556527316570282, "learning_rate": 9.295147356111822e-07, "loss": 0.012, "step": 12962 }, { "epoch": 0.890805387575591, "grad_norm": 0.2605803906917572, "learning_rate": 9.283581110086547e-07, "loss": 0.0072, "step": 12963 }, { "epoch": 0.8908741066520066, "grad_norm": 0.5081904530525208, "learning_rate": 9.272021834863326e-07, "loss": 0.0134, "step": 12964 }, { "epoch": 0.8909428257284222, "grad_norm": 0.20049971342086792, "learning_rate": 9.260469531014865e-07, "loss": 0.0053, "step": 12965 }, { "epoch": 0.8910115448048378, "grad_norm": 0.42095914483070374, "learning_rate": 9.248924199113373e-07, "loss": 0.0058, "step": 12966 }, { "epoch": 0.8910802638812534, "grad_norm": 0.19218707084655762, "learning_rate": 9.237385839730827e-07, "loss": 0.0081, "step": 12967 }, { "epoch": 0.8911489829576691, "grad_norm": 0.9191632270812988, "learning_rate": 9.225854453438771e-07, "loss": 0.0062, "step": 12968 }, { "epoch": 0.8912177020340847, "grad_norm": 0.3059893548488617, "learning_rate": 9.214330040808495e-07, "loss": 0.0114, "step": 12969 }, { "epoch": 0.8912864211105003, "grad_norm": 0.34257322549819946, "learning_rate": 9.202812602410826e-07, "loss": 0.0062, "step": 12970 }, { "epoch": 0.8913551401869159, "grad_norm": 0.04056984558701515, "learning_rate": 9.19130213881636e-07, "loss": 0.009, "step": 12971 }, { "epoch": 0.8914238592633315, "grad_norm": 0.5270019769668579, "learning_rate": 9.179798650595323e-07, "loss": 0.0116, "step": 12972 }, { "epoch": 0.8914925783397472, "grad_norm": 0.1800113320350647, "learning_rate": 9.168302138317524e-07, "loss": 0.0036, "step": 12973 }, { "epoch": 0.8915612974161627, "grad_norm": 0.17826171219348907, "learning_rate": 9.156812602552511e-07, "loss": 0.0076, "step": 12974 }, { "epoch": 0.8916300164925783, "grad_norm": 0.2880685031414032, "learning_rate": 9.145330043869426e-07, "loss": 0.0111, "step": 12975 }, { "epoch": 0.8916987355689939, "grad_norm": 0.22428153455257416, "learning_rate": 9.133854462837115e-07, "loss": 0.0082, "step": 12976 }, { "epoch": 0.8917674546454095, "grad_norm": 0.4073185920715332, "learning_rate": 9.12238586002404e-07, "loss": 0.015, "step": 12977 }, { "epoch": 0.8918361737218252, "grad_norm": 0.16820380091667175, "learning_rate": 9.110924235998347e-07, "loss": 0.0083, "step": 12978 }, { "epoch": 0.8919048927982408, "grad_norm": 0.18965482711791992, "learning_rate": 9.099469591327797e-07, "loss": 0.0153, "step": 12979 }, { "epoch": 0.8919736118746564, "grad_norm": 0.2796861529350281, "learning_rate": 9.08802192657987e-07, "loss": 0.0076, "step": 12980 }, { "epoch": 0.892042330951072, "grad_norm": 0.30685150623321533, "learning_rate": 9.076581242321613e-07, "loss": 0.0126, "step": 12981 }, { "epoch": 0.8921110500274876, "grad_norm": 0.1393190622329712, "learning_rate": 9.06514753911979e-07, "loss": 0.0029, "step": 12982 }, { "epoch": 0.8921797691039033, "grad_norm": 0.23280182480812073, "learning_rate": 9.053720817540811e-07, "loss": 0.0037, "step": 12983 }, { "epoch": 0.8922484881803189, "grad_norm": 0.1941029280424118, "learning_rate": 9.042301078150744e-07, "loss": 0.0078, "step": 12984 }, { "epoch": 0.8923172072567345, "grad_norm": 0.21216462552547455, "learning_rate": 9.030888321515268e-07, "loss": 0.0075, "step": 12985 }, { "epoch": 0.8923859263331501, "grad_norm": 0.46729958057403564, "learning_rate": 9.019482548199798e-07, "loss": 0.0158, "step": 12986 }, { "epoch": 0.8924546454095657, "grad_norm": 0.2512142062187195, "learning_rate": 9.008083758769281e-07, "loss": 0.0058, "step": 12987 }, { "epoch": 0.8925233644859814, "grad_norm": 0.2063068002462387, "learning_rate": 8.996691953788449e-07, "loss": 0.0069, "step": 12988 }, { "epoch": 0.8925920835623969, "grad_norm": 0.5651814341545105, "learning_rate": 8.985307133821602e-07, "loss": 0.0085, "step": 12989 }, { "epoch": 0.8926608026388125, "grad_norm": 0.4705970883369446, "learning_rate": 8.973929299432703e-07, "loss": 0.0114, "step": 12990 }, { "epoch": 0.8927295217152281, "grad_norm": 0.3578107953071594, "learning_rate": 8.962558451185455e-07, "loss": 0.0137, "step": 12991 }, { "epoch": 0.8927982407916437, "grad_norm": 0.37182992696762085, "learning_rate": 8.95119458964307e-07, "loss": 0.006, "step": 12992 }, { "epoch": 0.8928669598680594, "grad_norm": 0.29541176557540894, "learning_rate": 8.939837715368532e-07, "loss": 0.0055, "step": 12993 }, { "epoch": 0.892935678944475, "grad_norm": 0.37347063422203064, "learning_rate": 8.928487828924409e-07, "loss": 0.007, "step": 12994 }, { "epoch": 0.8930043980208906, "grad_norm": 0.39357084035873413, "learning_rate": 8.917144930872967e-07, "loss": 0.0079, "step": 12995 }, { "epoch": 0.8930731170973062, "grad_norm": 0.24839971959590912, "learning_rate": 8.905809021776124e-07, "loss": 0.008, "step": 12996 }, { "epoch": 0.8931418361737218, "grad_norm": 0.267062783241272, "learning_rate": 8.894480102195429e-07, "loss": 0.0079, "step": 12997 }, { "epoch": 0.8932105552501375, "grad_norm": 0.4383123517036438, "learning_rate": 8.883158172692085e-07, "loss": 0.0125, "step": 12998 }, { "epoch": 0.8932792743265531, "grad_norm": 0.2829793095588684, "learning_rate": 8.871843233826976e-07, "loss": 0.0118, "step": 12999 }, { "epoch": 0.8933479934029687, "grad_norm": 0.25740429759025574, "learning_rate": 8.860535286160571e-07, "loss": 0.0062, "step": 13000 }, { "epoch": 0.8934167124793843, "grad_norm": 0.5065832138061523, "learning_rate": 8.849234330253087e-07, "loss": 0.0121, "step": 13001 }, { "epoch": 0.8934854315557998, "grad_norm": 0.36711546778678894, "learning_rate": 8.837940366664343e-07, "loss": 0.01, "step": 13002 }, { "epoch": 0.8935541506322155, "grad_norm": 0.2151545286178589, "learning_rate": 8.826653395953843e-07, "loss": 0.0067, "step": 13003 }, { "epoch": 0.8936228697086311, "grad_norm": 0.2011214941740036, "learning_rate": 8.815373418680672e-07, "loss": 0.0068, "step": 13004 }, { "epoch": 0.8936915887850467, "grad_norm": 0.538405179977417, "learning_rate": 8.804100435403651e-07, "loss": 0.0084, "step": 13005 }, { "epoch": 0.8937603078614623, "grad_norm": 0.2159041464328766, "learning_rate": 8.792834446681181e-07, "loss": 0.0121, "step": 13006 }, { "epoch": 0.8938290269378779, "grad_norm": 0.24530461430549622, "learning_rate": 8.7815754530714e-07, "loss": 0.0066, "step": 13007 }, { "epoch": 0.8938977460142936, "grad_norm": 0.600953221321106, "learning_rate": 8.770323455132045e-07, "loss": 0.0064, "step": 13008 }, { "epoch": 0.8939664650907092, "grad_norm": 0.6281388998031616, "learning_rate": 8.759078453420505e-07, "loss": 0.0072, "step": 13009 }, { "epoch": 0.8940351841671248, "grad_norm": 0.3851979970932007, "learning_rate": 8.74784044849385e-07, "loss": 0.0093, "step": 13010 }, { "epoch": 0.8941039032435404, "grad_norm": 0.2036723792552948, "learning_rate": 8.73660944090875e-07, "loss": 0.0055, "step": 13011 }, { "epoch": 0.894172622319956, "grad_norm": 0.32915717363357544, "learning_rate": 8.725385431221594e-07, "loss": 0.0073, "step": 13012 }, { "epoch": 0.8942413413963717, "grad_norm": 0.32734912633895874, "learning_rate": 8.714168419988406e-07, "loss": 0.0033, "step": 13013 }, { "epoch": 0.8943100604727873, "grad_norm": 0.25894659757614136, "learning_rate": 8.702958407764855e-07, "loss": 0.0146, "step": 13014 }, { "epoch": 0.8943787795492029, "grad_norm": 0.4725984036922455, "learning_rate": 8.691755395106232e-07, "loss": 0.0156, "step": 13015 }, { "epoch": 0.8944474986256185, "grad_norm": 0.18874070048332214, "learning_rate": 8.68055938256756e-07, "loss": 0.0064, "step": 13016 }, { "epoch": 0.894516217702034, "grad_norm": 0.4495847821235657, "learning_rate": 8.669370370703411e-07, "loss": 0.0058, "step": 13017 }, { "epoch": 0.8945849367784497, "grad_norm": 0.41708534955978394, "learning_rate": 8.658188360068093e-07, "loss": 0.0114, "step": 13018 }, { "epoch": 0.8946536558548653, "grad_norm": 0.09762921184301376, "learning_rate": 8.647013351215511e-07, "loss": 0.005, "step": 13019 }, { "epoch": 0.8947223749312809, "grad_norm": 0.21319963037967682, "learning_rate": 8.635845344699323e-07, "loss": 0.0075, "step": 13020 }, { "epoch": 0.8947910940076965, "grad_norm": 0.37086111307144165, "learning_rate": 8.624684341072703e-07, "loss": 0.0076, "step": 13021 }, { "epoch": 0.8948598130841121, "grad_norm": 0.32526254653930664, "learning_rate": 8.613530340888576e-07, "loss": 0.0127, "step": 13022 }, { "epoch": 0.8949285321605278, "grad_norm": 0.1146899163722992, "learning_rate": 8.602383344699466e-07, "loss": 0.0081, "step": 13023 }, { "epoch": 0.8949972512369434, "grad_norm": 0.25993990898132324, "learning_rate": 8.591243353057599e-07, "loss": 0.0049, "step": 13024 }, { "epoch": 0.895065970313359, "grad_norm": 0.33374860882759094, "learning_rate": 8.580110366514798e-07, "loss": 0.0072, "step": 13025 }, { "epoch": 0.8951346893897746, "grad_norm": 0.571873128414154, "learning_rate": 8.568984385622575e-07, "loss": 0.005, "step": 13026 }, { "epoch": 0.8952034084661902, "grad_norm": 0.13426648080348969, "learning_rate": 8.55786541093212e-07, "loss": 0.0072, "step": 13027 }, { "epoch": 0.8952721275426059, "grad_norm": 0.4919394552707672, "learning_rate": 8.546753442994193e-07, "loss": 0.0061, "step": 13028 }, { "epoch": 0.8953408466190215, "grad_norm": 0.3891497254371643, "learning_rate": 8.535648482359287e-07, "loss": 0.0072, "step": 13029 }, { "epoch": 0.8954095656954371, "grad_norm": 0.5474555492401123, "learning_rate": 8.524550529577547e-07, "loss": 0.0201, "step": 13030 }, { "epoch": 0.8954782847718527, "grad_norm": 0.15069104731082916, "learning_rate": 8.513459585198679e-07, "loss": 0.0077, "step": 13031 }, { "epoch": 0.8955470038482682, "grad_norm": 0.3202832341194153, "learning_rate": 8.502375649772148e-07, "loss": 0.0086, "step": 13032 }, { "epoch": 0.8956157229246839, "grad_norm": 0.06676751375198364, "learning_rate": 8.491298723847047e-07, "loss": 0.0092, "step": 13033 }, { "epoch": 0.8956844420010995, "grad_norm": 0.1818484216928482, "learning_rate": 8.480228807972051e-07, "loss": 0.0081, "step": 13034 }, { "epoch": 0.8957531610775151, "grad_norm": 0.2248843014240265, "learning_rate": 8.469165902695575e-07, "loss": 0.0067, "step": 13035 }, { "epoch": 0.8958218801539307, "grad_norm": 0.15747575461864471, "learning_rate": 8.458110008565645e-07, "loss": 0.0091, "step": 13036 }, { "epoch": 0.8958905992303463, "grad_norm": 0.22932130098342896, "learning_rate": 8.44706112612994e-07, "loss": 0.0096, "step": 13037 }, { "epoch": 0.895959318306762, "grad_norm": 0.19905641674995422, "learning_rate": 8.436019255935789e-07, "loss": 0.0082, "step": 13038 }, { "epoch": 0.8960280373831776, "grad_norm": 0.2812066078186035, "learning_rate": 8.424984398530238e-07, "loss": 0.0115, "step": 13039 }, { "epoch": 0.8960967564595932, "grad_norm": 0.1518275886774063, "learning_rate": 8.413956554459884e-07, "loss": 0.0155, "step": 13040 }, { "epoch": 0.8961654755360088, "grad_norm": 0.4646545350551605, "learning_rate": 8.402935724271038e-07, "loss": 0.0184, "step": 13041 }, { "epoch": 0.8962341946124244, "grad_norm": 0.05994431674480438, "learning_rate": 8.391921908509631e-07, "loss": 0.0105, "step": 13042 }, { "epoch": 0.8963029136888401, "grad_norm": 0.2901466190814972, "learning_rate": 8.380915107721277e-07, "loss": 0.0104, "step": 13043 }, { "epoch": 0.8963716327652557, "grad_norm": 0.27002301812171936, "learning_rate": 8.369915322451255e-07, "loss": 0.0048, "step": 13044 }, { "epoch": 0.8964403518416713, "grad_norm": 1.034883975982666, "learning_rate": 8.358922553244413e-07, "loss": 0.0126, "step": 13045 }, { "epoch": 0.8965090709180868, "grad_norm": 0.19699198007583618, "learning_rate": 8.347936800645367e-07, "loss": 0.0119, "step": 13046 }, { "epoch": 0.8965777899945024, "grad_norm": 0.445499062538147, "learning_rate": 8.336958065198313e-07, "loss": 0.0059, "step": 13047 }, { "epoch": 0.8966465090709181, "grad_norm": 0.6807352900505066, "learning_rate": 8.325986347447101e-07, "loss": 0.0134, "step": 13048 }, { "epoch": 0.8967152281473337, "grad_norm": 0.7678900361061096, "learning_rate": 8.315021647935245e-07, "loss": 0.0077, "step": 13049 }, { "epoch": 0.8967839472237493, "grad_norm": 0.4666875898838043, "learning_rate": 8.304063967205961e-07, "loss": 0.0086, "step": 13050 }, { "epoch": 0.8968526663001649, "grad_norm": 0.43324097990989685, "learning_rate": 8.293113305801997e-07, "loss": 0.0099, "step": 13051 }, { "epoch": 0.8969213853765805, "grad_norm": 0.08630569279193878, "learning_rate": 8.282169664265888e-07, "loss": 0.0062, "step": 13052 }, { "epoch": 0.8969901044529962, "grad_norm": 0.42318063974380493, "learning_rate": 8.271233043139714e-07, "loss": 0.0076, "step": 13053 }, { "epoch": 0.8970588235294118, "grad_norm": 0.28297528624534607, "learning_rate": 8.260303442965295e-07, "loss": 0.0072, "step": 13054 }, { "epoch": 0.8971275426058274, "grad_norm": 0.4931715726852417, "learning_rate": 8.249380864283995e-07, "loss": 0.0074, "step": 13055 }, { "epoch": 0.897196261682243, "grad_norm": 0.36357367038726807, "learning_rate": 8.238465307636983e-07, "loss": 0.0105, "step": 13056 }, { "epoch": 0.8972649807586586, "grad_norm": 0.5497000217437744, "learning_rate": 8.227556773564943e-07, "loss": 0.0042, "step": 13057 }, { "epoch": 0.8973336998350742, "grad_norm": 0.23142941296100616, "learning_rate": 8.216655262608275e-07, "loss": 0.0082, "step": 13058 }, { "epoch": 0.8974024189114899, "grad_norm": 0.27401697635650635, "learning_rate": 8.205760775306981e-07, "loss": 0.0054, "step": 13059 }, { "epoch": 0.8974711379879055, "grad_norm": 0.32205718755722046, "learning_rate": 8.194873312200812e-07, "loss": 0.0094, "step": 13060 }, { "epoch": 0.897539857064321, "grad_norm": 0.2581603527069092, "learning_rate": 8.183992873829055e-07, "loss": 0.0103, "step": 13061 }, { "epoch": 0.8976085761407366, "grad_norm": 0.35325971245765686, "learning_rate": 8.173119460730744e-07, "loss": 0.0103, "step": 13062 }, { "epoch": 0.8976772952171522, "grad_norm": 0.4758574068546295, "learning_rate": 8.162253073444515e-07, "loss": 0.0086, "step": 13063 }, { "epoch": 0.8977460142935679, "grad_norm": 0.413568377494812, "learning_rate": 8.15139371250867e-07, "loss": 0.0064, "step": 13064 }, { "epoch": 0.8978147333699835, "grad_norm": 0.3294241726398468, "learning_rate": 8.14054137846113e-07, "loss": 0.0074, "step": 13065 }, { "epoch": 0.8978834524463991, "grad_norm": 0.24541251361370087, "learning_rate": 8.129696071839532e-07, "loss": 0.0109, "step": 13066 }, { "epoch": 0.8979521715228147, "grad_norm": 0.20812416076660156, "learning_rate": 8.118857793181112e-07, "loss": 0.0085, "step": 13067 }, { "epoch": 0.8980208905992303, "grad_norm": 0.47409817576408386, "learning_rate": 8.108026543022773e-07, "loss": 0.0153, "step": 13068 }, { "epoch": 0.898089609675646, "grad_norm": 0.49015018343925476, "learning_rate": 8.097202321901121e-07, "loss": 0.0066, "step": 13069 }, { "epoch": 0.8981583287520616, "grad_norm": 0.2566419839859009, "learning_rate": 8.086385130352275e-07, "loss": 0.0094, "step": 13070 }, { "epoch": 0.8982270478284772, "grad_norm": 0.11823805421590805, "learning_rate": 8.075574968912191e-07, "loss": 0.0068, "step": 13071 }, { "epoch": 0.8982957669048928, "grad_norm": 0.45228472352027893, "learning_rate": 8.064771838116308e-07, "loss": 0.014, "step": 13072 }, { "epoch": 0.8983644859813084, "grad_norm": 0.4272971451282501, "learning_rate": 8.05397573849983e-07, "loss": 0.0121, "step": 13073 }, { "epoch": 0.8984332050577241, "grad_norm": 0.14955727756023407, "learning_rate": 8.043186670597563e-07, "loss": 0.0056, "step": 13074 }, { "epoch": 0.8985019241341397, "grad_norm": 0.2586532235145569, "learning_rate": 8.032404634943996e-07, "loss": 0.0054, "step": 13075 }, { "epoch": 0.8985706432105552, "grad_norm": 0.37732943892478943, "learning_rate": 8.021629632073202e-07, "loss": 0.0099, "step": 13076 }, { "epoch": 0.8986393622869708, "grad_norm": 0.30472829937934875, "learning_rate": 8.010861662519003e-07, "loss": 0.0168, "step": 13077 }, { "epoch": 0.8987080813633864, "grad_norm": 0.3052287995815277, "learning_rate": 8.000100726814791e-07, "loss": 0.0073, "step": 13078 }, { "epoch": 0.8987768004398021, "grad_norm": 0.5476858615875244, "learning_rate": 7.989346825493638e-07, "loss": 0.0135, "step": 13079 }, { "epoch": 0.8988455195162177, "grad_norm": 0.19136832654476166, "learning_rate": 7.978599959088301e-07, "loss": 0.009, "step": 13080 }, { "epoch": 0.8989142385926333, "grad_norm": 0.11287039518356323, "learning_rate": 7.967860128131138e-07, "loss": 0.0078, "step": 13081 }, { "epoch": 0.8989829576690489, "grad_norm": 0.1486455202102661, "learning_rate": 7.957127333154157e-07, "loss": 0.0069, "step": 13082 }, { "epoch": 0.8990516767454645, "grad_norm": 0.08341033011674881, "learning_rate": 7.9464015746891e-07, "loss": 0.0035, "step": 13083 }, { "epoch": 0.8991203958218802, "grad_norm": 0.31177836656570435, "learning_rate": 7.935682853267224e-07, "loss": 0.0065, "step": 13084 }, { "epoch": 0.8991891148982958, "grad_norm": 0.3768352270126343, "learning_rate": 7.924971169419553e-07, "loss": 0.0128, "step": 13085 }, { "epoch": 0.8992578339747114, "grad_norm": 0.268371045589447, "learning_rate": 7.914266523676734e-07, "loss": 0.0139, "step": 13086 }, { "epoch": 0.899326553051127, "grad_norm": 0.5041705965995789, "learning_rate": 7.903568916569021e-07, "loss": 0.0053, "step": 13087 }, { "epoch": 0.8993952721275426, "grad_norm": 0.687309741973877, "learning_rate": 7.892878348626375e-07, "loss": 0.0072, "step": 13088 }, { "epoch": 0.8994639912039583, "grad_norm": 0.3252819776535034, "learning_rate": 7.882194820378358e-07, "loss": 0.0045, "step": 13089 }, { "epoch": 0.8995327102803738, "grad_norm": 0.28894126415252686, "learning_rate": 7.871518332354261e-07, "loss": 0.0089, "step": 13090 }, { "epoch": 0.8996014293567894, "grad_norm": 0.10650533437728882, "learning_rate": 7.860848885082894e-07, "loss": 0.0058, "step": 13091 }, { "epoch": 0.899670148433205, "grad_norm": 0.17385810613632202, "learning_rate": 7.850186479092886e-07, "loss": 0.0078, "step": 13092 }, { "epoch": 0.8997388675096206, "grad_norm": 0.36016184091567993, "learning_rate": 7.839531114912363e-07, "loss": 0.0037, "step": 13093 }, { "epoch": 0.8998075865860363, "grad_norm": 0.06933005899190903, "learning_rate": 7.828882793069237e-07, "loss": 0.0042, "step": 13094 }, { "epoch": 0.8998763056624519, "grad_norm": 0.5998312830924988, "learning_rate": 7.818241514090934e-07, "loss": 0.0102, "step": 13095 }, { "epoch": 0.8999450247388675, "grad_norm": 0.1392359435558319, "learning_rate": 7.807607278504653e-07, "loss": 0.0078, "step": 13096 }, { "epoch": 0.9000137438152831, "grad_norm": 0.03215426951646805, "learning_rate": 7.796980086837136e-07, "loss": 0.0033, "step": 13097 }, { "epoch": 0.9000824628916987, "grad_norm": 0.2225831001996994, "learning_rate": 7.786359939614912e-07, "loss": 0.0082, "step": 13098 }, { "epoch": 0.9001511819681144, "grad_norm": 0.4203369915485382, "learning_rate": 7.775746837364012e-07, "loss": 0.01, "step": 13099 }, { "epoch": 0.90021990104453, "grad_norm": 0.4341168999671936, "learning_rate": 7.76514078061023e-07, "loss": 0.0084, "step": 13100 }, { "epoch": 0.9002886201209456, "grad_norm": 0.4576689302921295, "learning_rate": 7.754541769878948e-07, "loss": 0.0077, "step": 13101 }, { "epoch": 0.9003573391973612, "grad_norm": 0.4988428056240082, "learning_rate": 7.743949805695228e-07, "loss": 0.0118, "step": 13102 }, { "epoch": 0.9004260582737768, "grad_norm": 0.26400256156921387, "learning_rate": 7.733364888583749e-07, "loss": 0.0074, "step": 13103 }, { "epoch": 0.9004947773501925, "grad_norm": 0.1341518610715866, "learning_rate": 7.722787019068894e-07, "loss": 0.0052, "step": 13104 }, { "epoch": 0.900563496426608, "grad_norm": 0.4181578755378723, "learning_rate": 7.712216197674659e-07, "loss": 0.0104, "step": 13105 }, { "epoch": 0.9006322155030236, "grad_norm": 0.41108036041259766, "learning_rate": 7.70165242492471e-07, "loss": 0.0101, "step": 13106 }, { "epoch": 0.9007009345794392, "grad_norm": 0.24208325147628784, "learning_rate": 7.691095701342344e-07, "loss": 0.0097, "step": 13107 }, { "epoch": 0.9007696536558548, "grad_norm": 0.4424259662628174, "learning_rate": 7.680546027450525e-07, "loss": 0.0119, "step": 13108 }, { "epoch": 0.9008383727322705, "grad_norm": 0.10521597415208817, "learning_rate": 7.670003403771835e-07, "loss": 0.0042, "step": 13109 }, { "epoch": 0.9009070918086861, "grad_norm": 0.46074557304382324, "learning_rate": 7.659467830828576e-07, "loss": 0.0084, "step": 13110 }, { "epoch": 0.9009758108851017, "grad_norm": 0.1603003442287445, "learning_rate": 7.648939309142661e-07, "loss": 0.0119, "step": 13111 }, { "epoch": 0.9010445299615173, "grad_norm": 0.13961569964885712, "learning_rate": 7.638417839235606e-07, "loss": 0.0058, "step": 13112 }, { "epoch": 0.9011132490379329, "grad_norm": 0.3673543930053711, "learning_rate": 7.627903421628663e-07, "loss": 0.0065, "step": 13113 }, { "epoch": 0.9011819681143486, "grad_norm": 0.23977357149124146, "learning_rate": 7.617396056842663e-07, "loss": 0.0075, "step": 13114 }, { "epoch": 0.9012506871907642, "grad_norm": 0.20630142092704773, "learning_rate": 7.60689574539814e-07, "loss": 0.0122, "step": 13115 }, { "epoch": 0.9013194062671798, "grad_norm": 0.24180755019187927, "learning_rate": 7.596402487815246e-07, "loss": 0.0091, "step": 13116 }, { "epoch": 0.9013881253435954, "grad_norm": 0.3423822522163391, "learning_rate": 7.585916284613831e-07, "loss": 0.0076, "step": 13117 }, { "epoch": 0.901456844420011, "grad_norm": 0.16780094802379608, "learning_rate": 7.575437136313313e-07, "loss": 0.0084, "step": 13118 }, { "epoch": 0.9015255634964267, "grad_norm": 0.06964746117591858, "learning_rate": 7.564965043432842e-07, "loss": 0.0044, "step": 13119 }, { "epoch": 0.9015942825728422, "grad_norm": 0.4464329779148102, "learning_rate": 7.554500006491156e-07, "loss": 0.015, "step": 13120 }, { "epoch": 0.9016630016492578, "grad_norm": 0.3074657618999481, "learning_rate": 7.54404202600667e-07, "loss": 0.0047, "step": 13121 }, { "epoch": 0.9017317207256734, "grad_norm": 0.1263430267572403, "learning_rate": 7.533591102497506e-07, "loss": 0.0063, "step": 13122 }, { "epoch": 0.901800439802089, "grad_norm": 0.28523606061935425, "learning_rate": 7.523147236481298e-07, "loss": 0.0124, "step": 13123 }, { "epoch": 0.9018691588785047, "grad_norm": 0.17161904275417328, "learning_rate": 7.5127104284755e-07, "loss": 0.0068, "step": 13124 }, { "epoch": 0.9019378779549203, "grad_norm": 0.24250799417495728, "learning_rate": 7.502280678997047e-07, "loss": 0.0097, "step": 13125 }, { "epoch": 0.9020065970313359, "grad_norm": 0.26805275678634644, "learning_rate": 7.491857988562661e-07, "loss": 0.0037, "step": 13126 }, { "epoch": 0.9020753161077515, "grad_norm": 0.3529370129108429, "learning_rate": 7.481442357688629e-07, "loss": 0.0069, "step": 13127 }, { "epoch": 0.9021440351841671, "grad_norm": 0.30676597356796265, "learning_rate": 7.471033786890952e-07, "loss": 0.0096, "step": 13128 }, { "epoch": 0.9022127542605828, "grad_norm": 0.3117307722568512, "learning_rate": 7.460632276685237e-07, "loss": 0.0083, "step": 13129 }, { "epoch": 0.9022814733369984, "grad_norm": 0.3319771885871887, "learning_rate": 7.450237827586753e-07, "loss": 0.0079, "step": 13130 }, { "epoch": 0.902350192413414, "grad_norm": 0.4457821249961853, "learning_rate": 7.439850440110407e-07, "loss": 0.0078, "step": 13131 }, { "epoch": 0.9024189114898296, "grad_norm": 0.5446968674659729, "learning_rate": 7.429470114770804e-07, "loss": 0.0126, "step": 13132 }, { "epoch": 0.9024876305662451, "grad_norm": 0.22064408659934998, "learning_rate": 7.41909685208208e-07, "loss": 0.0038, "step": 13133 }, { "epoch": 0.9025563496426608, "grad_norm": 0.3975922763347626, "learning_rate": 7.408730652558227e-07, "loss": 0.0136, "step": 13134 }, { "epoch": 0.9026250687190764, "grad_norm": 0.3423210084438324, "learning_rate": 7.398371516712665e-07, "loss": 0.0101, "step": 13135 }, { "epoch": 0.902693787795492, "grad_norm": 0.5636551976203918, "learning_rate": 7.388019445058619e-07, "loss": 0.0066, "step": 13136 }, { "epoch": 0.9027625068719076, "grad_norm": 0.0740671157836914, "learning_rate": 7.377674438108878e-07, "loss": 0.0059, "step": 13137 }, { "epoch": 0.9028312259483232, "grad_norm": 0.3058578073978424, "learning_rate": 7.367336496375948e-07, "loss": 0.0141, "step": 13138 }, { "epoch": 0.9028999450247389, "grad_norm": 0.1793983280658722, "learning_rate": 7.357005620371921e-07, "loss": 0.0064, "step": 13139 }, { "epoch": 0.9029686641011545, "grad_norm": 0.37315797805786133, "learning_rate": 7.346681810608552e-07, "loss": 0.0074, "step": 13140 }, { "epoch": 0.9030373831775701, "grad_norm": 0.17744846642017365, "learning_rate": 7.336365067597317e-07, "loss": 0.0082, "step": 13141 }, { "epoch": 0.9031061022539857, "grad_norm": 0.2660124897956848, "learning_rate": 7.326055391849224e-07, "loss": 0.0063, "step": 13142 }, { "epoch": 0.9031748213304013, "grad_norm": 0.42937496304512024, "learning_rate": 7.315752783875046e-07, "loss": 0.0072, "step": 13143 }, { "epoch": 0.903243540406817, "grad_norm": 0.7999011874198914, "learning_rate": 7.30545724418511e-07, "loss": 0.0172, "step": 13144 }, { "epoch": 0.9033122594832326, "grad_norm": 0.12488484382629395, "learning_rate": 7.29516877328944e-07, "loss": 0.0051, "step": 13145 }, { "epoch": 0.9033809785596482, "grad_norm": 0.22271449863910675, "learning_rate": 7.284887371697746e-07, "loss": 0.0103, "step": 13146 }, { "epoch": 0.9034496976360638, "grad_norm": 0.47679194808006287, "learning_rate": 7.27461303991932e-07, "loss": 0.0084, "step": 13147 }, { "epoch": 0.9035184167124793, "grad_norm": 0.5175712704658508, "learning_rate": 7.264345778463121e-07, "loss": 0.0089, "step": 13148 }, { "epoch": 0.903587135788895, "grad_norm": 0.20020325481891632, "learning_rate": 7.254085587837811e-07, "loss": 0.0086, "step": 13149 }, { "epoch": 0.9036558548653106, "grad_norm": 0.24942372739315033, "learning_rate": 7.243832468551581e-07, "loss": 0.0044, "step": 13150 }, { "epoch": 0.9037245739417262, "grad_norm": 0.4147699177265167, "learning_rate": 7.233586421112426e-07, "loss": 0.0069, "step": 13151 }, { "epoch": 0.9037932930181418, "grad_norm": 0.25431588292121887, "learning_rate": 7.223347446027873e-07, "loss": 0.0079, "step": 13152 }, { "epoch": 0.9038620120945574, "grad_norm": 0.5423434376716614, "learning_rate": 7.213115543805166e-07, "loss": 0.0113, "step": 13153 }, { "epoch": 0.9039307311709731, "grad_norm": 0.25477632880210876, "learning_rate": 7.202890714951133e-07, "loss": 0.0106, "step": 13154 }, { "epoch": 0.9039994502473887, "grad_norm": 0.2783794105052948, "learning_rate": 7.192672959972352e-07, "loss": 0.0052, "step": 13155 }, { "epoch": 0.9040681693238043, "grad_norm": 0.3530685305595398, "learning_rate": 7.182462279374919e-07, "loss": 0.0091, "step": 13156 }, { "epoch": 0.9041368884002199, "grad_norm": 0.35679128766059875, "learning_rate": 7.172258673664695e-07, "loss": 0.0091, "step": 13157 }, { "epoch": 0.9042056074766355, "grad_norm": 0.22982308268547058, "learning_rate": 7.162062143347142e-07, "loss": 0.0091, "step": 13158 }, { "epoch": 0.9042743265530512, "grad_norm": 0.31355899572372437, "learning_rate": 7.151872688927358e-07, "loss": 0.0073, "step": 13159 }, { "epoch": 0.9043430456294668, "grad_norm": 0.14086204767227173, "learning_rate": 7.141690310910138e-07, "loss": 0.0049, "step": 13160 }, { "epoch": 0.9044117647058824, "grad_norm": 0.7658814787864685, "learning_rate": 7.131515009799843e-07, "loss": 0.0114, "step": 13161 }, { "epoch": 0.904480483782298, "grad_norm": 0.22395184636116028, "learning_rate": 7.121346786100608e-07, "loss": 0.0098, "step": 13162 }, { "epoch": 0.9045492028587135, "grad_norm": 0.14078985154628754, "learning_rate": 7.111185640316059e-07, "loss": 0.0106, "step": 13163 }, { "epoch": 0.9046179219351292, "grad_norm": 0.38624516129493713, "learning_rate": 7.101031572949645e-07, "loss": 0.008, "step": 13164 }, { "epoch": 0.9046866410115448, "grad_norm": 0.3945952355861664, "learning_rate": 7.090884584504314e-07, "loss": 0.0069, "step": 13165 }, { "epoch": 0.9047553600879604, "grad_norm": 0.13403838872909546, "learning_rate": 7.080744675482781e-07, "loss": 0.0053, "step": 13166 }, { "epoch": 0.904824079164376, "grad_norm": 0.20883136987686157, "learning_rate": 7.070611846387309e-07, "loss": 0.0049, "step": 13167 }, { "epoch": 0.9048927982407916, "grad_norm": 0.21816354990005493, "learning_rate": 7.060486097719882e-07, "loss": 0.0074, "step": 13168 }, { "epoch": 0.9049615173172073, "grad_norm": 0.4521578252315521, "learning_rate": 7.05036742998208e-07, "loss": 0.011, "step": 13169 }, { "epoch": 0.9050302363936229, "grad_norm": 0.2123541533946991, "learning_rate": 7.040255843675203e-07, "loss": 0.0036, "step": 13170 }, { "epoch": 0.9050989554700385, "grad_norm": 0.15895554423332214, "learning_rate": 7.030151339300134e-07, "loss": 0.0102, "step": 13171 }, { "epoch": 0.9051676745464541, "grad_norm": 0.4591062068939209, "learning_rate": 7.020053917357455e-07, "loss": 0.008, "step": 13172 }, { "epoch": 0.9052363936228697, "grad_norm": 0.4027290940284729, "learning_rate": 7.009963578347334e-07, "loss": 0.0057, "step": 13173 }, { "epoch": 0.9053051126992854, "grad_norm": 0.3370678722858429, "learning_rate": 6.999880322769652e-07, "loss": 0.0091, "step": 13174 }, { "epoch": 0.905373831775701, "grad_norm": 0.16998615860939026, "learning_rate": 6.989804151123879e-07, "loss": 0.0054, "step": 13175 }, { "epoch": 0.9054425508521166, "grad_norm": 0.19019541144371033, "learning_rate": 6.979735063909198e-07, "loss": 0.013, "step": 13176 }, { "epoch": 0.9055112699285321, "grad_norm": 0.48801884055137634, "learning_rate": 6.969673061624427e-07, "loss": 0.0062, "step": 13177 }, { "epoch": 0.9055799890049477, "grad_norm": 0.517869234085083, "learning_rate": 6.959618144767965e-07, "loss": 0.0097, "step": 13178 }, { "epoch": 0.9056487080813634, "grad_norm": 0.4807986617088318, "learning_rate": 6.949570313837983e-07, "loss": 0.007, "step": 13179 }, { "epoch": 0.905717427157779, "grad_norm": 0.46912121772766113, "learning_rate": 6.939529569332148e-07, "loss": 0.0076, "step": 13180 }, { "epoch": 0.9057861462341946, "grad_norm": 0.10416805744171143, "learning_rate": 6.929495911747913e-07, "loss": 0.0067, "step": 13181 }, { "epoch": 0.9058548653106102, "grad_norm": 1.3389136791229248, "learning_rate": 6.919469341582297e-07, "loss": 0.0077, "step": 13182 }, { "epoch": 0.9059235843870258, "grad_norm": 0.6815277338027954, "learning_rate": 6.909449859332035e-07, "loss": 0.0101, "step": 13183 }, { "epoch": 0.9059923034634415, "grad_norm": 0.34078559279441833, "learning_rate": 6.899437465493413e-07, "loss": 0.0104, "step": 13184 }, { "epoch": 0.9060610225398571, "grad_norm": 0.23485854268074036, "learning_rate": 6.889432160562503e-07, "loss": 0.0065, "step": 13185 }, { "epoch": 0.9061297416162727, "grad_norm": 0.2769046127796173, "learning_rate": 6.879433945034858e-07, "loss": 0.0117, "step": 13186 }, { "epoch": 0.9061984606926883, "grad_norm": 0.2640126943588257, "learning_rate": 6.869442819405813e-07, "loss": 0.0146, "step": 13187 }, { "epoch": 0.9062671797691039, "grad_norm": 0.35992398858070374, "learning_rate": 6.859458784170325e-07, "loss": 0.0093, "step": 13188 }, { "epoch": 0.9063358988455195, "grad_norm": 0.3704383671283722, "learning_rate": 6.849481839822964e-07, "loss": 0.0147, "step": 13189 }, { "epoch": 0.9064046179219352, "grad_norm": 0.17493802309036255, "learning_rate": 6.839511986857966e-07, "loss": 0.0047, "step": 13190 }, { "epoch": 0.9064733369983508, "grad_norm": 0.4901149570941925, "learning_rate": 6.829549225769221e-07, "loss": 0.0059, "step": 13191 }, { "epoch": 0.9065420560747663, "grad_norm": 0.45909711718559265, "learning_rate": 6.819593557050252e-07, "loss": 0.0148, "step": 13192 }, { "epoch": 0.9066107751511819, "grad_norm": 0.3596133291721344, "learning_rate": 6.809644981194246e-07, "loss": 0.007, "step": 13193 }, { "epoch": 0.9066794942275975, "grad_norm": 0.2866167426109314, "learning_rate": 6.79970349869406e-07, "loss": 0.0047, "step": 13194 }, { "epoch": 0.9067482133040132, "grad_norm": 0.1159641295671463, "learning_rate": 6.789769110042149e-07, "loss": 0.0033, "step": 13195 }, { "epoch": 0.9068169323804288, "grad_norm": 0.2304387241601944, "learning_rate": 6.779841815730653e-07, "loss": 0.0084, "step": 13196 }, { "epoch": 0.9068856514568444, "grad_norm": 0.4179770350456238, "learning_rate": 6.769921616251328e-07, "loss": 0.0109, "step": 13197 }, { "epoch": 0.90695437053326, "grad_norm": 0.3424050807952881, "learning_rate": 6.76000851209565e-07, "loss": 0.0157, "step": 13198 }, { "epoch": 0.9070230896096756, "grad_norm": 0.2772350311279297, "learning_rate": 6.750102503754607e-07, "loss": 0.0049, "step": 13199 }, { "epoch": 0.9070918086860913, "grad_norm": 0.5262355208396912, "learning_rate": 6.740203591719025e-07, "loss": 0.0065, "step": 13200 }, { "epoch": 0.9071605277625069, "grad_norm": 0.8299558758735657, "learning_rate": 6.730311776479209e-07, "loss": 0.0152, "step": 13201 }, { "epoch": 0.9072292468389225, "grad_norm": 0.43880394101142883, "learning_rate": 6.720427058525202e-07, "loss": 0.0094, "step": 13202 }, { "epoch": 0.9072979659153381, "grad_norm": 0.515922486782074, "learning_rate": 6.710549438346663e-07, "loss": 0.0152, "step": 13203 }, { "epoch": 0.9073666849917537, "grad_norm": 0.8164964318275452, "learning_rate": 6.700678916432934e-07, "loss": 0.0191, "step": 13204 }, { "epoch": 0.9074354040681694, "grad_norm": 0.6544370651245117, "learning_rate": 6.690815493272906e-07, "loss": 0.009, "step": 13205 }, { "epoch": 0.907504123144585, "grad_norm": 0.21163202822208405, "learning_rate": 6.680959169355305e-07, "loss": 0.0079, "step": 13206 }, { "epoch": 0.9075728422210005, "grad_norm": 1.2320283651351929, "learning_rate": 6.671109945168308e-07, "loss": 0.0149, "step": 13207 }, { "epoch": 0.9076415612974161, "grad_norm": 0.33077120780944824, "learning_rate": 6.661267821199873e-07, "loss": 0.0065, "step": 13208 }, { "epoch": 0.9077102803738317, "grad_norm": 0.057382840663194656, "learning_rate": 6.651432797937513e-07, "loss": 0.0118, "step": 13209 }, { "epoch": 0.9077789994502474, "grad_norm": 0.2650778293609619, "learning_rate": 6.641604875868484e-07, "loss": 0.0077, "step": 13210 }, { "epoch": 0.907847718526663, "grad_norm": 0.39324989914894104, "learning_rate": 6.631784055479601e-07, "loss": 0.0077, "step": 13211 }, { "epoch": 0.9079164376030786, "grad_norm": 0.24952608346939087, "learning_rate": 6.621970337257388e-07, "loss": 0.0113, "step": 13212 }, { "epoch": 0.9079851566794942, "grad_norm": 0.2225870043039322, "learning_rate": 6.612163721687992e-07, "loss": 0.0057, "step": 13213 }, { "epoch": 0.9080538757559098, "grad_norm": 0.46244290471076965, "learning_rate": 6.602364209257205e-07, "loss": 0.01, "step": 13214 }, { "epoch": 0.9081225948323255, "grad_norm": 0.23832271993160248, "learning_rate": 6.592571800450509e-07, "loss": 0.0035, "step": 13215 }, { "epoch": 0.9081913139087411, "grad_norm": 0.07796455174684525, "learning_rate": 6.582786495752946e-07, "loss": 0.004, "step": 13216 }, { "epoch": 0.9082600329851567, "grad_norm": 0.38070163130760193, "learning_rate": 6.573008295649297e-07, "loss": 0.0063, "step": 13217 }, { "epoch": 0.9083287520615723, "grad_norm": 0.4014522135257721, "learning_rate": 6.563237200623923e-07, "loss": 0.0096, "step": 13218 }, { "epoch": 0.9083974711379879, "grad_norm": 0.5079295635223389, "learning_rate": 6.553473211160904e-07, "loss": 0.0097, "step": 13219 }, { "epoch": 0.9084661902144036, "grad_norm": 0.38012465834617615, "learning_rate": 6.543716327743904e-07, "loss": 0.0088, "step": 13220 }, { "epoch": 0.9085349092908191, "grad_norm": 0.480770081281662, "learning_rate": 6.533966550856285e-07, "loss": 0.0099, "step": 13221 }, { "epoch": 0.9086036283672347, "grad_norm": 0.2668619453907013, "learning_rate": 6.524223880980978e-07, "loss": 0.0088, "step": 13222 }, { "epoch": 0.9086723474436503, "grad_norm": 0.36360564827919006, "learning_rate": 6.514488318600665e-07, "loss": 0.0124, "step": 13223 }, { "epoch": 0.9087410665200659, "grad_norm": 0.29207631945610046, "learning_rate": 6.504759864197573e-07, "loss": 0.0078, "step": 13224 }, { "epoch": 0.9088097855964816, "grad_norm": 0.3164203464984894, "learning_rate": 6.495038518253704e-07, "loss": 0.0058, "step": 13225 }, { "epoch": 0.9088785046728972, "grad_norm": 0.69917893409729, "learning_rate": 6.485324281250587e-07, "loss": 0.0122, "step": 13226 }, { "epoch": 0.9089472237493128, "grad_norm": 0.2571260631084442, "learning_rate": 6.475617153669456e-07, "loss": 0.0074, "step": 13227 }, { "epoch": 0.9090159428257284, "grad_norm": 0.03790688142180443, "learning_rate": 6.465917135991156e-07, "loss": 0.0036, "step": 13228 }, { "epoch": 0.909084661902144, "grad_norm": 1.5540434122085571, "learning_rate": 6.456224228696239e-07, "loss": 0.0116, "step": 13229 }, { "epoch": 0.9091533809785597, "grad_norm": 0.5922518372535706, "learning_rate": 6.446538432264854e-07, "loss": 0.0075, "step": 13230 }, { "epoch": 0.9092221000549753, "grad_norm": 0.5580807328224182, "learning_rate": 6.436859747176832e-07, "loss": 0.0094, "step": 13231 }, { "epoch": 0.9092908191313909, "grad_norm": 0.2550564110279083, "learning_rate": 6.427188173911624e-07, "loss": 0.0097, "step": 13232 }, { "epoch": 0.9093595382078065, "grad_norm": 0.6158192157745361, "learning_rate": 6.417523712948331e-07, "loss": 0.0073, "step": 13233 }, { "epoch": 0.909428257284222, "grad_norm": 0.34391507506370544, "learning_rate": 6.407866364765718e-07, "loss": 0.0165, "step": 13234 }, { "epoch": 0.9094969763606378, "grad_norm": 0.2647404372692108, "learning_rate": 6.398216129842205e-07, "loss": 0.0087, "step": 13235 }, { "epoch": 0.9095656954370533, "grad_norm": 0.29494595527648926, "learning_rate": 6.388573008655824e-07, "loss": 0.0064, "step": 13236 }, { "epoch": 0.9096344145134689, "grad_norm": 0.24574796855449677, "learning_rate": 6.378937001684276e-07, "loss": 0.009, "step": 13237 }, { "epoch": 0.9097031335898845, "grad_norm": 0.3974160850048065, "learning_rate": 6.369308109404947e-07, "loss": 0.0076, "step": 13238 }, { "epoch": 0.9097718526663001, "grad_norm": 0.2450568675994873, "learning_rate": 6.359686332294773e-07, "loss": 0.0047, "step": 13239 }, { "epoch": 0.9098405717427158, "grad_norm": 0.1188022792339325, "learning_rate": 6.350071670830438e-07, "loss": 0.0071, "step": 13240 }, { "epoch": 0.9099092908191314, "grad_norm": 0.6419755816459656, "learning_rate": 6.34046412548821e-07, "loss": 0.0088, "step": 13241 }, { "epoch": 0.909978009895547, "grad_norm": 0.503341794013977, "learning_rate": 6.330863696744044e-07, "loss": 0.0085, "step": 13242 }, { "epoch": 0.9100467289719626, "grad_norm": 0.0931999459862709, "learning_rate": 6.321270385073508e-07, "loss": 0.0078, "step": 13243 }, { "epoch": 0.9101154480483782, "grad_norm": 0.2541629672050476, "learning_rate": 6.311684190951872e-07, "loss": 0.0073, "step": 13244 }, { "epoch": 0.9101841671247939, "grad_norm": 0.19678856432437897, "learning_rate": 6.302105114853973e-07, "loss": 0.0179, "step": 13245 }, { "epoch": 0.9102528862012095, "grad_norm": 0.42068442702293396, "learning_rate": 6.292533157254382e-07, "loss": 0.0111, "step": 13246 }, { "epoch": 0.9103216052776251, "grad_norm": 0.8242626190185547, "learning_rate": 6.282968318627219e-07, "loss": 0.0101, "step": 13247 }, { "epoch": 0.9103903243540407, "grad_norm": 0.875428318977356, "learning_rate": 6.273410599446339e-07, "loss": 0.0132, "step": 13248 }, { "epoch": 0.9104590434304562, "grad_norm": 0.28798267245292664, "learning_rate": 6.263860000185212e-07, "loss": 0.0101, "step": 13249 }, { "epoch": 0.910527762506872, "grad_norm": 0.2383233904838562, "learning_rate": 6.254316521316944e-07, "loss": 0.0122, "step": 13250 }, { "epoch": 0.9105964815832875, "grad_norm": 0.2528364658355713, "learning_rate": 6.244780163314323e-07, "loss": 0.0051, "step": 13251 }, { "epoch": 0.9106652006597031, "grad_norm": 0.17328186333179474, "learning_rate": 6.235250926649738e-07, "loss": 0.0092, "step": 13252 }, { "epoch": 0.9107339197361187, "grad_norm": 0.382350355386734, "learning_rate": 6.225728811795245e-07, "loss": 0.006, "step": 13253 }, { "epoch": 0.9108026388125343, "grad_norm": 0.03729463741183281, "learning_rate": 6.216213819222566e-07, "loss": 0.0114, "step": 13254 }, { "epoch": 0.91087135788895, "grad_norm": 0.2516075372695923, "learning_rate": 6.206705949403058e-07, "loss": 0.0057, "step": 13255 }, { "epoch": 0.9109400769653656, "grad_norm": 0.4617561101913452, "learning_rate": 6.197205202807693e-07, "loss": 0.0141, "step": 13256 }, { "epoch": 0.9110087960417812, "grad_norm": 0.251233845949173, "learning_rate": 6.187711579907163e-07, "loss": 0.0097, "step": 13257 }, { "epoch": 0.9110775151181968, "grad_norm": 0.2760695219039917, "learning_rate": 6.178225081171707e-07, "loss": 0.0067, "step": 13258 }, { "epoch": 0.9111462341946124, "grad_norm": 0.24781836569309235, "learning_rate": 6.168745707071316e-07, "loss": 0.0065, "step": 13259 }, { "epoch": 0.9112149532710281, "grad_norm": 0.5892231464385986, "learning_rate": 6.159273458075532e-07, "loss": 0.0137, "step": 13260 }, { "epoch": 0.9112836723474437, "grad_norm": 0.3178514540195465, "learning_rate": 6.149808334653645e-07, "loss": 0.006, "step": 13261 }, { "epoch": 0.9113523914238593, "grad_norm": 0.5193593502044678, "learning_rate": 6.140350337274498e-07, "loss": 0.0098, "step": 13262 }, { "epoch": 0.9114211105002749, "grad_norm": 0.27770134806632996, "learning_rate": 6.130899466406631e-07, "loss": 0.0106, "step": 13263 }, { "epoch": 0.9114898295766904, "grad_norm": 0.2659350633621216, "learning_rate": 6.121455722518221e-07, "loss": 0.0046, "step": 13264 }, { "epoch": 0.9115585486531061, "grad_norm": 0.16900070011615753, "learning_rate": 6.11201910607711e-07, "loss": 0.0092, "step": 13265 }, { "epoch": 0.9116272677295217, "grad_norm": 0.18994303047657013, "learning_rate": 6.102589617550725e-07, "loss": 0.007, "step": 13266 }, { "epoch": 0.9116959868059373, "grad_norm": 0.35216987133026123, "learning_rate": 6.093167257406224e-07, "loss": 0.0072, "step": 13267 }, { "epoch": 0.9117647058823529, "grad_norm": 0.42695459723472595, "learning_rate": 6.083752026110351e-07, "loss": 0.0112, "step": 13268 }, { "epoch": 0.9118334249587685, "grad_norm": 0.132485494017601, "learning_rate": 6.074343924129533e-07, "loss": 0.0036, "step": 13269 }, { "epoch": 0.9119021440351842, "grad_norm": 0.42932143807411194, "learning_rate": 6.064942951929797e-07, "loss": 0.0055, "step": 13270 }, { "epoch": 0.9119708631115998, "grad_norm": 0.27940812706947327, "learning_rate": 6.055549109976905e-07, "loss": 0.0058, "step": 13271 }, { "epoch": 0.9120395821880154, "grad_norm": 0.2227117270231247, "learning_rate": 6.046162398736149e-07, "loss": 0.0065, "step": 13272 }, { "epoch": 0.912108301264431, "grad_norm": 0.05162012204527855, "learning_rate": 6.036782818672543e-07, "loss": 0.0073, "step": 13273 }, { "epoch": 0.9121770203408466, "grad_norm": 0.3031352460384369, "learning_rate": 6.027410370250764e-07, "loss": 0.0076, "step": 13274 }, { "epoch": 0.9122457394172623, "grad_norm": 0.29148709774017334, "learning_rate": 6.018045053935056e-07, "loss": 0.0094, "step": 13275 }, { "epoch": 0.9123144584936779, "grad_norm": 0.4441894590854645, "learning_rate": 6.008686870189417e-07, "loss": 0.0103, "step": 13276 }, { "epoch": 0.9123831775700935, "grad_norm": 0.23981249332427979, "learning_rate": 5.999335819477359e-07, "loss": 0.009, "step": 13277 }, { "epoch": 0.912451896646509, "grad_norm": 0.6074041724205017, "learning_rate": 5.98999190226216e-07, "loss": 0.0082, "step": 13278 }, { "epoch": 0.9125206157229246, "grad_norm": 0.38491857051849365, "learning_rate": 5.980655119006684e-07, "loss": 0.0068, "step": 13279 }, { "epoch": 0.9125893347993403, "grad_norm": 0.1923244446516037, "learning_rate": 5.971325470173494e-07, "loss": 0.0058, "step": 13280 }, { "epoch": 0.9126580538757559, "grad_norm": 0.8915987610816956, "learning_rate": 5.962002956224704e-07, "loss": 0.0091, "step": 13281 }, { "epoch": 0.9127267729521715, "grad_norm": 0.21959850192070007, "learning_rate": 5.952687577622178e-07, "loss": 0.0094, "step": 13282 }, { "epoch": 0.9127954920285871, "grad_norm": 0.9506725668907166, "learning_rate": 5.943379334827331e-07, "loss": 0.0107, "step": 13283 }, { "epoch": 0.9128642111050027, "grad_norm": 0.7948821187019348, "learning_rate": 5.934078228301326e-07, "loss": 0.0103, "step": 13284 }, { "epoch": 0.9129329301814184, "grad_norm": 0.2632899284362793, "learning_rate": 5.924784258504878e-07, "loss": 0.0042, "step": 13285 }, { "epoch": 0.913001649257834, "grad_norm": 0.31886202096939087, "learning_rate": 5.915497425898437e-07, "loss": 0.0087, "step": 13286 }, { "epoch": 0.9130703683342496, "grad_norm": 0.25523167848587036, "learning_rate": 5.906217730942015e-07, "loss": 0.0094, "step": 13287 }, { "epoch": 0.9131390874106652, "grad_norm": 0.6416545510292053, "learning_rate": 5.896945174095331e-07, "loss": 0.0114, "step": 13288 }, { "epoch": 0.9132078064870808, "grad_norm": 0.6522623300552368, "learning_rate": 5.887679755817715e-07, "loss": 0.0111, "step": 13289 }, { "epoch": 0.9132765255634965, "grad_norm": 0.3221060633659363, "learning_rate": 5.878421476568152e-07, "loss": 0.0046, "step": 13290 }, { "epoch": 0.9133452446399121, "grad_norm": 0.5612629055976868, "learning_rate": 5.869170336805307e-07, "loss": 0.0076, "step": 13291 }, { "epoch": 0.9134139637163277, "grad_norm": 0.23192693293094635, "learning_rate": 5.859926336987431e-07, "loss": 0.0042, "step": 13292 }, { "epoch": 0.9134826827927433, "grad_norm": 0.24696297943592072, "learning_rate": 5.850689477572474e-07, "loss": 0.0074, "step": 13293 }, { "epoch": 0.9135514018691588, "grad_norm": 0.6546118855476379, "learning_rate": 5.841459759017987e-07, "loss": 0.0085, "step": 13294 }, { "epoch": 0.9136201209455745, "grad_norm": 0.09474553912878036, "learning_rate": 5.83223718178122e-07, "loss": 0.0047, "step": 13295 }, { "epoch": 0.9136888400219901, "grad_norm": 0.28440847992897034, "learning_rate": 5.82302174631899e-07, "loss": 0.0049, "step": 13296 }, { "epoch": 0.9137575590984057, "grad_norm": 0.2807562053203583, "learning_rate": 5.813813453087901e-07, "loss": 0.0035, "step": 13297 }, { "epoch": 0.9138262781748213, "grad_norm": 0.26576581597328186, "learning_rate": 5.804612302544021e-07, "loss": 0.007, "step": 13298 }, { "epoch": 0.9138949972512369, "grad_norm": 0.3329010605812073, "learning_rate": 5.795418295143218e-07, "loss": 0.0087, "step": 13299 }, { "epoch": 0.9139637163276526, "grad_norm": 0.24112994968891144, "learning_rate": 5.786231431340911e-07, "loss": 0.0067, "step": 13300 }, { "epoch": 0.9140324354040682, "grad_norm": 0.3491952121257782, "learning_rate": 5.777051711592219e-07, "loss": 0.011, "step": 13301 }, { "epoch": 0.9141011544804838, "grad_norm": 0.13740508258342743, "learning_rate": 5.767879136351844e-07, "loss": 0.0042, "step": 13302 }, { "epoch": 0.9141698735568994, "grad_norm": 0.3960467576980591, "learning_rate": 5.758713706074242e-07, "loss": 0.0074, "step": 13303 }, { "epoch": 0.914238592633315, "grad_norm": 0.0941186249256134, "learning_rate": 5.749555421213382e-07, "loss": 0.0054, "step": 13304 }, { "epoch": 0.9143073117097307, "grad_norm": 0.45319029688835144, "learning_rate": 5.740404282223016e-07, "loss": 0.0057, "step": 13305 }, { "epoch": 0.9143760307861463, "grad_norm": 0.27826523780822754, "learning_rate": 5.731260289556417e-07, "loss": 0.0065, "step": 13306 }, { "epoch": 0.9144447498625619, "grad_norm": 0.6160431504249573, "learning_rate": 5.722123443666589e-07, "loss": 0.0149, "step": 13307 }, { "epoch": 0.9145134689389774, "grad_norm": 0.06359932571649551, "learning_rate": 5.712993745006134e-07, "loss": 0.0053, "step": 13308 }, { "epoch": 0.914582188015393, "grad_norm": 0.20497000217437744, "learning_rate": 5.703871194027327e-07, "loss": 0.0074, "step": 13309 }, { "epoch": 0.9146509070918087, "grad_norm": 0.10183021426200867, "learning_rate": 5.694755791182088e-07, "loss": 0.005, "step": 13310 }, { "epoch": 0.9147196261682243, "grad_norm": 0.4975825548171997, "learning_rate": 5.685647536921956e-07, "loss": 0.0076, "step": 13311 }, { "epoch": 0.9147883452446399, "grad_norm": 0.11873053014278412, "learning_rate": 5.676546431698171e-07, "loss": 0.005, "step": 13312 }, { "epoch": 0.9148570643210555, "grad_norm": 0.4673478603363037, "learning_rate": 5.667452475961538e-07, "loss": 0.0134, "step": 13313 }, { "epoch": 0.9149257833974711, "grad_norm": 0.5482662320137024, "learning_rate": 5.65836567016258e-07, "loss": 0.008, "step": 13314 }, { "epoch": 0.9149945024738868, "grad_norm": 0.3488185703754425, "learning_rate": 5.64928601475142e-07, "loss": 0.0056, "step": 13315 }, { "epoch": 0.9150632215503024, "grad_norm": 0.1301947683095932, "learning_rate": 5.640213510177883e-07, "loss": 0.0104, "step": 13316 }, { "epoch": 0.915131940626718, "grad_norm": 0.07432878017425537, "learning_rate": 5.631148156891359e-07, "loss": 0.005, "step": 13317 }, { "epoch": 0.9152006597031336, "grad_norm": 0.32853004336357117, "learning_rate": 5.62208995534097e-07, "loss": 0.0078, "step": 13318 }, { "epoch": 0.9152693787795492, "grad_norm": 0.9955095648765564, "learning_rate": 5.613038905975392e-07, "loss": 0.0118, "step": 13319 }, { "epoch": 0.9153380978559648, "grad_norm": 0.42129480838775635, "learning_rate": 5.603995009243018e-07, "loss": 0.0078, "step": 13320 }, { "epoch": 0.9154068169323805, "grad_norm": 0.8740114569664001, "learning_rate": 5.594958265591871e-07, "loss": 0.009, "step": 13321 }, { "epoch": 0.915475536008796, "grad_norm": 0.22171452641487122, "learning_rate": 5.585928675469609e-07, "loss": 0.0102, "step": 13322 }, { "epoch": 0.9155442550852116, "grad_norm": 0.4199698865413666, "learning_rate": 5.576906239323543e-07, "loss": 0.0079, "step": 13323 }, { "epoch": 0.9156129741616272, "grad_norm": 0.7691105008125305, "learning_rate": 5.567890957600629e-07, "loss": 0.0082, "step": 13324 }, { "epoch": 0.9156816932380428, "grad_norm": 0.4391474723815918, "learning_rate": 5.55888283074743e-07, "loss": 0.009, "step": 13325 }, { "epoch": 0.9157504123144585, "grad_norm": 0.41282016038894653, "learning_rate": 5.549881859210221e-07, "loss": 0.0063, "step": 13326 }, { "epoch": 0.9158191313908741, "grad_norm": 0.2314271777868271, "learning_rate": 5.540888043434911e-07, "loss": 0.0041, "step": 13327 }, { "epoch": 0.9158878504672897, "grad_norm": 0.3916042149066925, "learning_rate": 5.531901383866994e-07, "loss": 0.0133, "step": 13328 }, { "epoch": 0.9159565695437053, "grad_norm": 0.23988138139247894, "learning_rate": 5.522921880951681e-07, "loss": 0.0069, "step": 13329 }, { "epoch": 0.9160252886201209, "grad_norm": 0.2589050829410553, "learning_rate": 5.513949535133783e-07, "loss": 0.0136, "step": 13330 }, { "epoch": 0.9160940076965366, "grad_norm": 0.28220346570014954, "learning_rate": 5.504984346857778e-07, "loss": 0.0044, "step": 13331 }, { "epoch": 0.9161627267729522, "grad_norm": 0.3166608512401581, "learning_rate": 5.496026316567742e-07, "loss": 0.0055, "step": 13332 }, { "epoch": 0.9162314458493678, "grad_norm": 0.29689982533454895, "learning_rate": 5.487075444707523e-07, "loss": 0.0109, "step": 13333 }, { "epoch": 0.9163001649257834, "grad_norm": 0.15977545082569122, "learning_rate": 5.478131731720448e-07, "loss": 0.0039, "step": 13334 }, { "epoch": 0.916368884002199, "grad_norm": 0.1530294120311737, "learning_rate": 5.46919517804963e-07, "loss": 0.0049, "step": 13335 }, { "epoch": 0.9164376030786147, "grad_norm": 0.2779513895511627, "learning_rate": 5.460265784137731e-07, "loss": 0.0052, "step": 13336 }, { "epoch": 0.9165063221550303, "grad_norm": 0.2657786011695862, "learning_rate": 5.451343550427096e-07, "loss": 0.006, "step": 13337 }, { "epoch": 0.9165750412314458, "grad_norm": 0.5462045073509216, "learning_rate": 5.442428477359706e-07, "loss": 0.0082, "step": 13338 }, { "epoch": 0.9166437603078614, "grad_norm": 0.258314847946167, "learning_rate": 5.433520565377243e-07, "loss": 0.0053, "step": 13339 }, { "epoch": 0.916712479384277, "grad_norm": 0.2780725657939911, "learning_rate": 5.424619814920934e-07, "loss": 0.0084, "step": 13340 }, { "epoch": 0.9167811984606927, "grad_norm": 0.26637694239616394, "learning_rate": 5.415726226431744e-07, "loss": 0.0061, "step": 13341 }, { "epoch": 0.9168499175371083, "grad_norm": 0.5535226464271545, "learning_rate": 5.406839800350206e-07, "loss": 0.008, "step": 13342 }, { "epoch": 0.9169186366135239, "grad_norm": 0.509395956993103, "learning_rate": 5.397960537116581e-07, "loss": 0.0107, "step": 13343 }, { "epoch": 0.9169873556899395, "grad_norm": 0.27171674370765686, "learning_rate": 5.389088437170669e-07, "loss": 0.0054, "step": 13344 }, { "epoch": 0.9170560747663551, "grad_norm": 0.12880124151706696, "learning_rate": 5.38022350095202e-07, "loss": 0.0103, "step": 13345 }, { "epoch": 0.9171247938427708, "grad_norm": 0.12282201647758484, "learning_rate": 5.37136572889978e-07, "loss": 0.0059, "step": 13346 }, { "epoch": 0.9171935129191864, "grad_norm": 0.297278493642807, "learning_rate": 5.362515121452733e-07, "loss": 0.0167, "step": 13347 }, { "epoch": 0.917262231995602, "grad_norm": 0.34955331683158875, "learning_rate": 5.353671679049326e-07, "loss": 0.0103, "step": 13348 }, { "epoch": 0.9173309510720176, "grad_norm": 0.19789491593837738, "learning_rate": 5.344835402127645e-07, "loss": 0.0071, "step": 13349 }, { "epoch": 0.9173996701484332, "grad_norm": 0.3704374432563782, "learning_rate": 5.336006291125406e-07, "loss": 0.0092, "step": 13350 }, { "epoch": 0.9174683892248489, "grad_norm": 0.3021175265312195, "learning_rate": 5.327184346480007e-07, "loss": 0.0046, "step": 13351 }, { "epoch": 0.9175371083012644, "grad_norm": 0.5109718441963196, "learning_rate": 5.318369568628467e-07, "loss": 0.0121, "step": 13352 }, { "epoch": 0.91760582737768, "grad_norm": 0.2280900478363037, "learning_rate": 5.309561958007437e-07, "loss": 0.0109, "step": 13353 }, { "epoch": 0.9176745464540956, "grad_norm": 0.383666068315506, "learning_rate": 5.300761515053265e-07, "loss": 0.0119, "step": 13354 }, { "epoch": 0.9177432655305112, "grad_norm": 0.18677642941474915, "learning_rate": 5.291968240201839e-07, "loss": 0.0028, "step": 13355 }, { "epoch": 0.9178119846069269, "grad_norm": 0.3637087643146515, "learning_rate": 5.283182133888809e-07, "loss": 0.0108, "step": 13356 }, { "epoch": 0.9178807036833425, "grad_norm": 0.23620502650737762, "learning_rate": 5.27440319654941e-07, "loss": 0.0059, "step": 13357 }, { "epoch": 0.9179494227597581, "grad_norm": 0.42627573013305664, "learning_rate": 5.265631428618561e-07, "loss": 0.0085, "step": 13358 }, { "epoch": 0.9180181418361737, "grad_norm": 0.498643696308136, "learning_rate": 5.256866830530733e-07, "loss": 0.0111, "step": 13359 }, { "epoch": 0.9180868609125893, "grad_norm": 0.21745312213897705, "learning_rate": 5.248109402720175e-07, "loss": 0.0118, "step": 13360 }, { "epoch": 0.918155579989005, "grad_norm": 0.06668317317962646, "learning_rate": 5.239359145620659e-07, "loss": 0.0043, "step": 13361 }, { "epoch": 0.9182242990654206, "grad_norm": 0.10340645164251328, "learning_rate": 5.230616059665672e-07, "loss": 0.0033, "step": 13362 }, { "epoch": 0.9182930181418362, "grad_norm": 0.5383269190788269, "learning_rate": 5.22188014528835e-07, "loss": 0.0123, "step": 13363 }, { "epoch": 0.9183617372182518, "grad_norm": 0.2751392424106598, "learning_rate": 5.213151402921429e-07, "loss": 0.0064, "step": 13364 }, { "epoch": 0.9184304562946674, "grad_norm": 0.3722374439239502, "learning_rate": 5.204429832997315e-07, "loss": 0.0065, "step": 13365 }, { "epoch": 0.918499175371083, "grad_norm": 0.5035587549209595, "learning_rate": 5.195715435948062e-07, "loss": 0.011, "step": 13366 }, { "epoch": 0.9185678944474986, "grad_norm": 1.534548282623291, "learning_rate": 5.187008212205374e-07, "loss": 0.0101, "step": 13367 }, { "epoch": 0.9186366135239142, "grad_norm": 0.32989802956581116, "learning_rate": 5.17830816220054e-07, "loss": 0.0119, "step": 13368 }, { "epoch": 0.9187053326003298, "grad_norm": 0.5364248752593994, "learning_rate": 5.169615286364599e-07, "loss": 0.0072, "step": 13369 }, { "epoch": 0.9187740516767454, "grad_norm": 0.20435801148414612, "learning_rate": 5.160929585128154e-07, "loss": 0.0059, "step": 13370 }, { "epoch": 0.9188427707531611, "grad_norm": 0.3474862575531006, "learning_rate": 5.152251058921498e-07, "loss": 0.0142, "step": 13371 }, { "epoch": 0.9189114898295767, "grad_norm": 0.2749911844730377, "learning_rate": 5.143579708174517e-07, "loss": 0.012, "step": 13372 }, { "epoch": 0.9189802089059923, "grad_norm": 0.2416527420282364, "learning_rate": 5.134915533316803e-07, "loss": 0.0044, "step": 13373 }, { "epoch": 0.9190489279824079, "grad_norm": 0.4854244589805603, "learning_rate": 5.126258534777495e-07, "loss": 0.0137, "step": 13374 }, { "epoch": 0.9191176470588235, "grad_norm": 0.720761239528656, "learning_rate": 5.117608712985533e-07, "loss": 0.0088, "step": 13375 }, { "epoch": 0.9191863661352392, "grad_norm": 0.6095474362373352, "learning_rate": 5.108966068369358e-07, "loss": 0.0045, "step": 13376 }, { "epoch": 0.9192550852116548, "grad_norm": 0.2499518245458603, "learning_rate": 5.100330601357145e-07, "loss": 0.0101, "step": 13377 }, { "epoch": 0.9193238042880704, "grad_norm": 0.12014304101467133, "learning_rate": 5.091702312376617e-07, "loss": 0.003, "step": 13378 }, { "epoch": 0.919392523364486, "grad_norm": 0.2811969220638275, "learning_rate": 5.083081201855267e-07, "loss": 0.0047, "step": 13379 }, { "epoch": 0.9194612424409015, "grad_norm": 0.2872636318206787, "learning_rate": 5.074467270220117e-07, "loss": 0.0102, "step": 13380 }, { "epoch": 0.9195299615173173, "grad_norm": 0.31717801094055176, "learning_rate": 5.065860517897897e-07, "loss": 0.0114, "step": 13381 }, { "epoch": 0.9195986805937328, "grad_norm": 0.4109775424003601, "learning_rate": 5.057260945314995e-07, "loss": 0.0076, "step": 13382 }, { "epoch": 0.9196673996701484, "grad_norm": 0.20606428384780884, "learning_rate": 5.04866855289739e-07, "loss": 0.0063, "step": 13383 }, { "epoch": 0.919736118746564, "grad_norm": 0.38275429606437683, "learning_rate": 5.040083341070739e-07, "loss": 0.0132, "step": 13384 }, { "epoch": 0.9198048378229796, "grad_norm": 0.7853319048881531, "learning_rate": 5.031505310260337e-07, "loss": 0.0107, "step": 13385 }, { "epoch": 0.9198735568993953, "grad_norm": 0.2919691801071167, "learning_rate": 5.022934460891093e-07, "loss": 0.011, "step": 13386 }, { "epoch": 0.9199422759758109, "grad_norm": 0.9878631234169006, "learning_rate": 5.014370793387634e-07, "loss": 0.01, "step": 13387 }, { "epoch": 0.9200109950522265, "grad_norm": 0.28150513768196106, "learning_rate": 5.005814308174189e-07, "loss": 0.0071, "step": 13388 }, { "epoch": 0.9200797141286421, "grad_norm": 0.31825852394104004, "learning_rate": 4.997265005674583e-07, "loss": 0.0037, "step": 13389 }, { "epoch": 0.9201484332050577, "grad_norm": 0.29404377937316895, "learning_rate": 4.988722886312363e-07, "loss": 0.0065, "step": 13390 }, { "epoch": 0.9202171522814734, "grad_norm": 0.40340614318847656, "learning_rate": 4.980187950510673e-07, "loss": 0.0084, "step": 13391 }, { "epoch": 0.920285871357889, "grad_norm": 0.20076574385166168, "learning_rate": 4.971660198692324e-07, "loss": 0.0039, "step": 13392 }, { "epoch": 0.9203545904343046, "grad_norm": 0.3174871802330017, "learning_rate": 4.963139631279761e-07, "loss": 0.0083, "step": 13393 }, { "epoch": 0.9204233095107202, "grad_norm": 0.1851765662431717, "learning_rate": 4.954626248695099e-07, "loss": 0.0073, "step": 13394 }, { "epoch": 0.9204920285871357, "grad_norm": 0.05232643336057663, "learning_rate": 4.946120051360031e-07, "loss": 0.0055, "step": 13395 }, { "epoch": 0.9205607476635514, "grad_norm": 0.6696624159812927, "learning_rate": 4.937621039695988e-07, "loss": 0.0094, "step": 13396 }, { "epoch": 0.920629466739967, "grad_norm": 0.3862568140029907, "learning_rate": 4.929129214123934e-07, "loss": 0.0108, "step": 13397 }, { "epoch": 0.9206981858163826, "grad_norm": 0.3210022449493408, "learning_rate": 4.920644575064565e-07, "loss": 0.0058, "step": 13398 }, { "epoch": 0.9207669048927982, "grad_norm": 0.41991689801216125, "learning_rate": 4.91216712293821e-07, "loss": 0.0056, "step": 13399 }, { "epoch": 0.9208356239692138, "grad_norm": 0.15969707071781158, "learning_rate": 4.903696858164786e-07, "loss": 0.0058, "step": 13400 }, { "epoch": 0.9209043430456295, "grad_norm": 0.3439215421676636, "learning_rate": 4.895233781163938e-07, "loss": 0.0109, "step": 13401 }, { "epoch": 0.9209730621220451, "grad_norm": 0.2664441764354706, "learning_rate": 4.886777892354866e-07, "loss": 0.0063, "step": 13402 }, { "epoch": 0.9210417811984607, "grad_norm": 0.7696765661239624, "learning_rate": 4.8783291921565e-07, "loss": 0.012, "step": 13403 }, { "epoch": 0.9211105002748763, "grad_norm": 0.4960530996322632, "learning_rate": 4.869887680987306e-07, "loss": 0.0127, "step": 13404 }, { "epoch": 0.9211792193512919, "grad_norm": 0.29258614778518677, "learning_rate": 4.86145335926555e-07, "loss": 0.0134, "step": 13405 }, { "epoch": 0.9212479384277076, "grad_norm": 0.22718536853790283, "learning_rate": 4.853026227408964e-07, "loss": 0.0048, "step": 13406 }, { "epoch": 0.9213166575041232, "grad_norm": 0.41492411494255066, "learning_rate": 4.84460628583508e-07, "loss": 0.0063, "step": 13407 }, { "epoch": 0.9213853765805388, "grad_norm": 0.4577721655368805, "learning_rate": 4.836193534960964e-07, "loss": 0.0075, "step": 13408 }, { "epoch": 0.9214540956569544, "grad_norm": 0.3600590229034424, "learning_rate": 4.827787975203385e-07, "loss": 0.013, "step": 13409 }, { "epoch": 0.9215228147333699, "grad_norm": 0.338153600692749, "learning_rate": 4.819389606978691e-07, "loss": 0.011, "step": 13410 }, { "epoch": 0.9215915338097856, "grad_norm": 0.31076884269714355, "learning_rate": 4.810998430702984e-07, "loss": 0.006, "step": 13411 }, { "epoch": 0.9216602528862012, "grad_norm": 0.20356148481369019, "learning_rate": 4.802614446791914e-07, "loss": 0.0048, "step": 13412 }, { "epoch": 0.9217289719626168, "grad_norm": 0.27229705452919006, "learning_rate": 4.794237655660816e-07, "loss": 0.0109, "step": 13413 }, { "epoch": 0.9217976910390324, "grad_norm": 0.6501989960670471, "learning_rate": 4.785868057724657e-07, "loss": 0.0074, "step": 13414 }, { "epoch": 0.921866410115448, "grad_norm": 0.6265588998794556, "learning_rate": 4.777505653398056e-07, "loss": 0.0127, "step": 13415 }, { "epoch": 0.9219351291918637, "grad_norm": 0.2899837791919708, "learning_rate": 4.769150443095233e-07, "loss": 0.0131, "step": 13416 }, { "epoch": 0.9220038482682793, "grad_norm": 0.47732359170913696, "learning_rate": 4.7608024272301207e-07, "loss": 0.0105, "step": 13417 }, { "epoch": 0.9220725673446949, "grad_norm": 0.4494415521621704, "learning_rate": 4.752461606216274e-07, "loss": 0.0051, "step": 13418 }, { "epoch": 0.9221412864211105, "grad_norm": 0.4395017623901367, "learning_rate": 4.744127980466845e-07, "loss": 0.0057, "step": 13419 }, { "epoch": 0.9222100054975261, "grad_norm": 0.5946404933929443, "learning_rate": 4.7358015503947027e-07, "loss": 0.0113, "step": 13420 }, { "epoch": 0.9222787245739418, "grad_norm": 0.3840616047382355, "learning_rate": 4.727482316412285e-07, "loss": 0.0087, "step": 13421 }, { "epoch": 0.9223474436503574, "grad_norm": 0.2629031240940094, "learning_rate": 4.719170278931728e-07, "loss": 0.005, "step": 13422 }, { "epoch": 0.922416162726773, "grad_norm": 0.17946547269821167, "learning_rate": 4.710865438364786e-07, "loss": 0.0038, "step": 13423 }, { "epoch": 0.9224848818031886, "grad_norm": 0.2747802138328552, "learning_rate": 4.7025677951228806e-07, "loss": 0.0112, "step": 13424 }, { "epoch": 0.9225536008796041, "grad_norm": 0.1672193855047226, "learning_rate": 4.6942773496170324e-07, "loss": 0.0072, "step": 13425 }, { "epoch": 0.9226223199560198, "grad_norm": 0.6914846897125244, "learning_rate": 4.685994102257979e-07, "loss": 0.0108, "step": 13426 }, { "epoch": 0.9226910390324354, "grad_norm": 0.6882466077804565, "learning_rate": 4.677718053455993e-07, "loss": 0.0134, "step": 13427 }, { "epoch": 0.922759758108851, "grad_norm": 0.2523997724056244, "learning_rate": 4.669449203621096e-07, "loss": 0.0106, "step": 13428 }, { "epoch": 0.9228284771852666, "grad_norm": 0.4572626054286957, "learning_rate": 4.6611875531628933e-07, "loss": 0.0121, "step": 13429 }, { "epoch": 0.9228971962616822, "grad_norm": 0.7738760113716125, "learning_rate": 4.6529331024906585e-07, "loss": 0.0114, "step": 13430 }, { "epoch": 0.9229659153380979, "grad_norm": 0.33508577942848206, "learning_rate": 4.6446858520132976e-07, "loss": 0.0094, "step": 13431 }, { "epoch": 0.9230346344145135, "grad_norm": 0.32129529118537903, "learning_rate": 4.636445802139366e-07, "loss": 0.0078, "step": 13432 }, { "epoch": 0.9231033534909291, "grad_norm": 0.27785158157348633, "learning_rate": 4.6282129532770556e-07, "loss": 0.0067, "step": 13433 }, { "epoch": 0.9231720725673447, "grad_norm": 0.278642863035202, "learning_rate": 4.6199873058342055e-07, "loss": 0.0065, "step": 13434 }, { "epoch": 0.9232407916437603, "grad_norm": 0.3493213951587677, "learning_rate": 4.611768860218291e-07, "loss": 0.0059, "step": 13435 }, { "epoch": 0.923309510720176, "grad_norm": 0.3691302239894867, "learning_rate": 4.6035576168364347e-07, "loss": 0.0084, "step": 13436 }, { "epoch": 0.9233782297965916, "grad_norm": 0.3061949610710144, "learning_rate": 4.5953535760954126e-07, "loss": 0.0069, "step": 13437 }, { "epoch": 0.9234469488730072, "grad_norm": 0.2911039888858795, "learning_rate": 4.587156738401649e-07, "loss": 0.008, "step": 13438 }, { "epoch": 0.9235156679494227, "grad_norm": 0.20506824553012848, "learning_rate": 4.578967104161169e-07, "loss": 0.0072, "step": 13439 }, { "epoch": 0.9235843870258383, "grad_norm": 0.5460783839225769, "learning_rate": 4.570784673779699e-07, "loss": 0.0064, "step": 13440 }, { "epoch": 0.923653106102254, "grad_norm": 0.4339662790298462, "learning_rate": 4.5626094476625466e-07, "loss": 0.0106, "step": 13441 }, { "epoch": 0.9237218251786696, "grad_norm": 0.21351751685142517, "learning_rate": 4.5544414262147225e-07, "loss": 0.0066, "step": 13442 }, { "epoch": 0.9237905442550852, "grad_norm": 0.3153928220272064, "learning_rate": 4.5462806098408695e-07, "loss": 0.009, "step": 13443 }, { "epoch": 0.9238592633315008, "grad_norm": 0.24568688869476318, "learning_rate": 4.5381269989452133e-07, "loss": 0.0046, "step": 13444 }, { "epoch": 0.9239279824079164, "grad_norm": 0.3127921521663666, "learning_rate": 4.5299805939316987e-07, "loss": 0.0133, "step": 13445 }, { "epoch": 0.9239967014843321, "grad_norm": 0.23846906423568726, "learning_rate": 4.5218413952038517e-07, "loss": 0.0059, "step": 13446 }, { "epoch": 0.9240654205607477, "grad_norm": 0.3999791741371155, "learning_rate": 4.5137094031649007e-07, "loss": 0.0046, "step": 13447 }, { "epoch": 0.9241341396371633, "grad_norm": 0.2946036756038666, "learning_rate": 4.50558461821769e-07, "loss": 0.0065, "step": 13448 }, { "epoch": 0.9242028587135789, "grad_norm": 0.2182173877954483, "learning_rate": 4.497467040764697e-07, "loss": 0.0078, "step": 13449 }, { "epoch": 0.9242715777899945, "grad_norm": 0.2361857295036316, "learning_rate": 4.489356671208017e-07, "loss": 0.0056, "step": 13450 }, { "epoch": 0.9243402968664101, "grad_norm": 0.6471993923187256, "learning_rate": 4.4812535099494957e-07, "loss": 0.0103, "step": 13451 }, { "epoch": 0.9244090159428258, "grad_norm": 0.46267417073249817, "learning_rate": 4.4731575573904616e-07, "loss": 0.0075, "step": 13452 }, { "epoch": 0.9244777350192414, "grad_norm": 0.3329100012779236, "learning_rate": 4.465068813932027e-07, "loss": 0.007, "step": 13453 }, { "epoch": 0.9245464540956569, "grad_norm": 0.12991811335086823, "learning_rate": 4.4569872799748724e-07, "loss": 0.0056, "step": 13454 }, { "epoch": 0.9246151731720725, "grad_norm": 0.40666720271110535, "learning_rate": 4.448912955919343e-07, "loss": 0.0035, "step": 13455 }, { "epoch": 0.9246838922484881, "grad_norm": 0.43320295214653015, "learning_rate": 4.4408458421654365e-07, "loss": 0.0084, "step": 13456 }, { "epoch": 0.9247526113249038, "grad_norm": 0.07890186458826065, "learning_rate": 4.4327859391127655e-07, "loss": 0.0067, "step": 13457 }, { "epoch": 0.9248213304013194, "grad_norm": 0.1451471596956253, "learning_rate": 4.424733247160595e-07, "loss": 0.0102, "step": 13458 }, { "epoch": 0.924890049477735, "grad_norm": 0.39020001888275146, "learning_rate": 4.416687766707855e-07, "loss": 0.0104, "step": 13459 }, { "epoch": 0.9249587685541506, "grad_norm": 0.3273162543773651, "learning_rate": 4.4086494981531114e-07, "loss": 0.0118, "step": 13460 }, { "epoch": 0.9250274876305662, "grad_norm": 0.29806870222091675, "learning_rate": 4.4006184418945274e-07, "loss": 0.0071, "step": 13461 }, { "epoch": 0.9250962067069819, "grad_norm": 0.211365208029747, "learning_rate": 4.3925945983299854e-07, "loss": 0.0054, "step": 13462 }, { "epoch": 0.9251649257833975, "grad_norm": 0.3085293173789978, "learning_rate": 4.3845779678569344e-07, "loss": 0.0094, "step": 13463 }, { "epoch": 0.9252336448598131, "grad_norm": 0.09952221065759659, "learning_rate": 4.3765685508725397e-07, "loss": 0.0077, "step": 13464 }, { "epoch": 0.9253023639362287, "grad_norm": 0.1271185725927353, "learning_rate": 4.3685663477735006e-07, "loss": 0.0045, "step": 13465 }, { "epoch": 0.9253710830126443, "grad_norm": 0.3739553391933441, "learning_rate": 4.3605713589563335e-07, "loss": 0.0103, "step": 13466 }, { "epoch": 0.92543980208906, "grad_norm": 0.18195296823978424, "learning_rate": 4.3525835848170046e-07, "loss": 0.0037, "step": 13467 }, { "epoch": 0.9255085211654756, "grad_norm": 0.26401856541633606, "learning_rate": 4.344603025751265e-07, "loss": 0.0064, "step": 13468 }, { "epoch": 0.9255772402418911, "grad_norm": 0.31802091002464294, "learning_rate": 4.3366296821544304e-07, "loss": 0.0081, "step": 13469 }, { "epoch": 0.9256459593183067, "grad_norm": 0.47251659631729126, "learning_rate": 4.3286635544215024e-07, "loss": 0.0093, "step": 13470 }, { "epoch": 0.9257146783947223, "grad_norm": 0.21449212729930878, "learning_rate": 4.320704642947082e-07, "loss": 0.0061, "step": 13471 }, { "epoch": 0.925783397471138, "grad_norm": 0.2571961283683777, "learning_rate": 4.312752948125437e-07, "loss": 0.0086, "step": 13472 }, { "epoch": 0.9258521165475536, "grad_norm": 0.355159193277359, "learning_rate": 4.304808470350502e-07, "loss": 0.0186, "step": 13473 }, { "epoch": 0.9259208356239692, "grad_norm": 0.3363438844680786, "learning_rate": 4.296871210015829e-07, "loss": 0.0079, "step": 13474 }, { "epoch": 0.9259895547003848, "grad_norm": 0.1394023448228836, "learning_rate": 4.288941167514604e-07, "loss": 0.009, "step": 13475 }, { "epoch": 0.9260582737768004, "grad_norm": 0.3838941752910614, "learning_rate": 4.2810183432396624e-07, "loss": 0.0053, "step": 13476 }, { "epoch": 0.9261269928532161, "grad_norm": 0.22901806235313416, "learning_rate": 4.273102737583473e-07, "loss": 0.0069, "step": 13477 }, { "epoch": 0.9261957119296317, "grad_norm": 0.6215418577194214, "learning_rate": 4.265194350938173e-07, "loss": 0.0067, "step": 13478 }, { "epoch": 0.9262644310060473, "grad_norm": 0.43793442845344543, "learning_rate": 4.2572931836955487e-07, "loss": 0.0104, "step": 13479 }, { "epoch": 0.9263331500824629, "grad_norm": 0.2707720100879669, "learning_rate": 4.2493992362469535e-07, "loss": 0.0059, "step": 13480 }, { "epoch": 0.9264018691588785, "grad_norm": 0.34452107548713684, "learning_rate": 4.2415125089835084e-07, "loss": 0.0103, "step": 13481 }, { "epoch": 0.9264705882352942, "grad_norm": 0.10037211328744888, "learning_rate": 4.2336330022958336e-07, "loss": 0.005, "step": 13482 }, { "epoch": 0.9265393073117097, "grad_norm": 0.398427277803421, "learning_rate": 4.225760716574301e-07, "loss": 0.0076, "step": 13483 }, { "epoch": 0.9266080263881253, "grad_norm": 0.4820472002029419, "learning_rate": 4.2178956522088806e-07, "loss": 0.009, "step": 13484 }, { "epoch": 0.9266767454645409, "grad_norm": 0.2324477732181549, "learning_rate": 4.210037809589229e-07, "loss": 0.0074, "step": 13485 }, { "epoch": 0.9267454645409565, "grad_norm": 0.5239942073822021, "learning_rate": 4.2021871891045337e-07, "loss": 0.0084, "step": 13486 }, { "epoch": 0.9268141836173722, "grad_norm": 0.35099220275878906, "learning_rate": 4.1943437911437676e-07, "loss": 0.0081, "step": 13487 }, { "epoch": 0.9268829026937878, "grad_norm": 0.47104373574256897, "learning_rate": 4.1865076160954197e-07, "loss": 0.0072, "step": 13488 }, { "epoch": 0.9269516217702034, "grad_norm": 0.3888222575187683, "learning_rate": 4.178678664347713e-07, "loss": 0.012, "step": 13489 }, { "epoch": 0.927020340846619, "grad_norm": 0.5770229697227478, "learning_rate": 4.17085693628847e-07, "loss": 0.0075, "step": 13490 }, { "epoch": 0.9270890599230346, "grad_norm": 0.1797952950000763, "learning_rate": 4.163042432305164e-07, "loss": 0.0078, "step": 13491 }, { "epoch": 0.9271577789994503, "grad_norm": 0.36485764384269714, "learning_rate": 4.155235152784903e-07, "loss": 0.0082, "step": 13492 }, { "epoch": 0.9272264980758659, "grad_norm": 0.23732490837574005, "learning_rate": 4.147435098114477e-07, "loss": 0.0055, "step": 13493 }, { "epoch": 0.9272952171522815, "grad_norm": 0.04984110966324806, "learning_rate": 4.139642268680227e-07, "loss": 0.0069, "step": 13494 }, { "epoch": 0.9273639362286971, "grad_norm": 0.19730715453624725, "learning_rate": 4.131856664868211e-07, "loss": 0.0054, "step": 13495 }, { "epoch": 0.9274326553051127, "grad_norm": 0.4620633125305176, "learning_rate": 4.1240782870641535e-07, "loss": 0.0063, "step": 13496 }, { "epoch": 0.9275013743815284, "grad_norm": 0.3413732051849365, "learning_rate": 4.11630713565333e-07, "loss": 0.0082, "step": 13497 }, { "epoch": 0.927570093457944, "grad_norm": 0.4499368965625763, "learning_rate": 4.108543211020749e-07, "loss": 0.0076, "step": 13498 }, { "epoch": 0.9276388125343595, "grad_norm": 0.18265630304813385, "learning_rate": 4.100786513550969e-07, "loss": 0.0065, "step": 13499 }, { "epoch": 0.9277075316107751, "grad_norm": 0.30942314863204956, "learning_rate": 4.0930370436283004e-07, "loss": 0.0069, "step": 13500 }, { "epoch": 0.9277762506871907, "grad_norm": 0.6930037140846252, "learning_rate": 4.0852948016365686e-07, "loss": 0.0127, "step": 13501 }, { "epoch": 0.9278449697636064, "grad_norm": 0.1995535045862198, "learning_rate": 4.0775597879593676e-07, "loss": 0.0052, "step": 13502 }, { "epoch": 0.927913688840022, "grad_norm": 0.30683019757270813, "learning_rate": 4.06983200297984e-07, "loss": 0.0074, "step": 13503 }, { "epoch": 0.9279824079164376, "grad_norm": 0.14823321998119354, "learning_rate": 4.0621114470808297e-07, "loss": 0.0037, "step": 13504 }, { "epoch": 0.9280511269928532, "grad_norm": 0.42200589179992676, "learning_rate": 4.0543981206447643e-07, "loss": 0.007, "step": 13505 }, { "epoch": 0.9281198460692688, "grad_norm": 0.5135058760643005, "learning_rate": 4.0466920240537885e-07, "loss": 0.0133, "step": 13506 }, { "epoch": 0.9281885651456845, "grad_norm": 0.24068228900432587, "learning_rate": 4.0389931576895966e-07, "loss": 0.0113, "step": 13507 }, { "epoch": 0.9282572842221001, "grad_norm": 0.6256840825080872, "learning_rate": 4.031301521933617e-07, "loss": 0.008, "step": 13508 }, { "epoch": 0.9283260032985157, "grad_norm": 0.29958781599998474, "learning_rate": 4.0236171171668445e-07, "loss": 0.0081, "step": 13509 }, { "epoch": 0.9283947223749313, "grad_norm": 0.40344467759132385, "learning_rate": 4.015939943769992e-07, "loss": 0.0083, "step": 13510 }, { "epoch": 0.9284634414513468, "grad_norm": 0.22026395797729492, "learning_rate": 4.008270002123321e-07, "loss": 0.0055, "step": 13511 }, { "epoch": 0.9285321605277626, "grad_norm": 0.44909927248954773, "learning_rate": 4.000607292606828e-07, "loss": 0.0114, "step": 13512 }, { "epoch": 0.9286008796041781, "grad_norm": 0.47337648272514343, "learning_rate": 3.9929518156000765e-07, "loss": 0.0142, "step": 13513 }, { "epoch": 0.9286695986805937, "grad_norm": 0.4487047493457794, "learning_rate": 3.985303571482296e-07, "loss": 0.007, "step": 13514 }, { "epoch": 0.9287383177570093, "grad_norm": 0.2541626989841461, "learning_rate": 3.9776625606324167e-07, "loss": 0.0047, "step": 13515 }, { "epoch": 0.9288070368334249, "grad_norm": 0.16693812608718872, "learning_rate": 3.970028783428903e-07, "loss": 0.004, "step": 13516 }, { "epoch": 0.9288757559098406, "grad_norm": 0.2246236652135849, "learning_rate": 3.9624022402499517e-07, "loss": 0.0042, "step": 13517 }, { "epoch": 0.9289444749862562, "grad_norm": 0.25290077924728394, "learning_rate": 3.9547829314733444e-07, "loss": 0.0114, "step": 13518 }, { "epoch": 0.9290131940626718, "grad_norm": 0.5275204181671143, "learning_rate": 3.9471708574765284e-07, "loss": 0.0146, "step": 13519 }, { "epoch": 0.9290819131390874, "grad_norm": 0.5648801326751709, "learning_rate": 3.9395660186366033e-07, "loss": 0.0071, "step": 13520 }, { "epoch": 0.929150632215503, "grad_norm": 0.5535132884979248, "learning_rate": 3.9319684153303003e-07, "loss": 0.0104, "step": 13521 }, { "epoch": 0.9292193512919187, "grad_norm": 0.26759031414985657, "learning_rate": 3.924378047933952e-07, "loss": 0.0031, "step": 13522 }, { "epoch": 0.9292880703683343, "grad_norm": 0.3180738687515259, "learning_rate": 3.916794916823624e-07, "loss": 0.0081, "step": 13523 }, { "epoch": 0.9293567894447499, "grad_norm": 0.23422928154468536, "learning_rate": 3.9092190223749324e-07, "loss": 0.0062, "step": 13524 }, { "epoch": 0.9294255085211655, "grad_norm": 0.1962740123271942, "learning_rate": 3.901650364963161e-07, "loss": 0.0048, "step": 13525 }, { "epoch": 0.929494227597581, "grad_norm": 0.3988744914531708, "learning_rate": 3.8940889449632923e-07, "loss": 0.0116, "step": 13526 }, { "epoch": 0.9295629466739967, "grad_norm": 0.15708886086940765, "learning_rate": 3.8865347627498606e-07, "loss": 0.0067, "step": 13527 }, { "epoch": 0.9296316657504123, "grad_norm": 0.2316945344209671, "learning_rate": 3.878987818697116e-07, "loss": 0.0068, "step": 13528 }, { "epoch": 0.9297003848268279, "grad_norm": 0.2171354740858078, "learning_rate": 3.871448113178894e-07, "loss": 0.0076, "step": 13529 }, { "epoch": 0.9297691039032435, "grad_norm": 0.4539062976837158, "learning_rate": 3.863915646568694e-07, "loss": 0.0089, "step": 13530 }, { "epoch": 0.9298378229796591, "grad_norm": 0.42855560779571533, "learning_rate": 3.856390419239686e-07, "loss": 0.0076, "step": 13531 }, { "epoch": 0.9299065420560748, "grad_norm": 0.301419198513031, "learning_rate": 3.848872431564637e-07, "loss": 0.0082, "step": 13532 }, { "epoch": 0.9299752611324904, "grad_norm": 0.29258161783218384, "learning_rate": 3.841361683915967e-07, "loss": 0.0173, "step": 13533 }, { "epoch": 0.930043980208906, "grad_norm": 0.22598816454410553, "learning_rate": 3.833858176665761e-07, "loss": 0.0108, "step": 13534 }, { "epoch": 0.9301126992853216, "grad_norm": 0.6432376503944397, "learning_rate": 3.8263619101856886e-07, "loss": 0.0101, "step": 13535 }, { "epoch": 0.9301814183617372, "grad_norm": 0.09572703391313553, "learning_rate": 3.818872884847169e-07, "loss": 0.0077, "step": 13536 }, { "epoch": 0.9302501374381529, "grad_norm": 0.1285056173801422, "learning_rate": 3.8113911010211055e-07, "loss": 0.0045, "step": 13537 }, { "epoch": 0.9303188565145685, "grad_norm": 0.4884638786315918, "learning_rate": 3.803916559078202e-07, "loss": 0.0048, "step": 13538 }, { "epoch": 0.9303875755909841, "grad_norm": 0.5290927290916443, "learning_rate": 3.796449259388696e-07, "loss": 0.0086, "step": 13539 }, { "epoch": 0.9304562946673997, "grad_norm": 0.30726316571235657, "learning_rate": 3.7889892023225234e-07, "loss": 0.0049, "step": 13540 }, { "epoch": 0.9305250137438152, "grad_norm": 0.21121227741241455, "learning_rate": 3.7815363882492237e-07, "loss": 0.0064, "step": 13541 }, { "epoch": 0.930593732820231, "grad_norm": 0.6451824903488159, "learning_rate": 3.774090817538001e-07, "loss": 0.0118, "step": 13542 }, { "epoch": 0.9306624518966465, "grad_norm": 0.10213493555784225, "learning_rate": 3.766652490557659e-07, "loss": 0.0067, "step": 13543 }, { "epoch": 0.9307311709730621, "grad_norm": 0.2178085595369339, "learning_rate": 3.759221407676755e-07, "loss": 0.0062, "step": 13544 }, { "epoch": 0.9307998900494777, "grad_norm": 0.36455675959587097, "learning_rate": 3.751797569263343e-07, "loss": 0.0103, "step": 13545 }, { "epoch": 0.9308686091258933, "grad_norm": 0.4812619388103485, "learning_rate": 3.744380975685213e-07, "loss": 0.0092, "step": 13546 }, { "epoch": 0.930937328202309, "grad_norm": 0.5017748475074768, "learning_rate": 3.7369716273097545e-07, "loss": 0.0056, "step": 13547 }, { "epoch": 0.9310060472787246, "grad_norm": 0.14913885295391083, "learning_rate": 3.72956952450404e-07, "loss": 0.0088, "step": 13548 }, { "epoch": 0.9310747663551402, "grad_norm": 0.08633074164390564, "learning_rate": 3.722174667634709e-07, "loss": 0.0095, "step": 13549 }, { "epoch": 0.9311434854315558, "grad_norm": 0.06152874603867531, "learning_rate": 3.714787057068103e-07, "loss": 0.0041, "step": 13550 }, { "epoch": 0.9312122045079714, "grad_norm": 0.17868313193321228, "learning_rate": 3.7074066931702123e-07, "loss": 0.0065, "step": 13551 }, { "epoch": 0.9312809235843871, "grad_norm": 0.45193296670913696, "learning_rate": 3.7000335763066273e-07, "loss": 0.0063, "step": 13552 }, { "epoch": 0.9313496426608027, "grad_norm": 0.36385101079940796, "learning_rate": 3.6926677068426064e-07, "loss": 0.0083, "step": 13553 }, { "epoch": 0.9314183617372183, "grad_norm": 0.24339251220226288, "learning_rate": 3.685309085143007e-07, "loss": 0.0097, "step": 13554 }, { "epoch": 0.9314870808136338, "grad_norm": 0.020942572504281998, "learning_rate": 3.677957711572405e-07, "loss": 0.0086, "step": 13555 }, { "epoch": 0.9315557998900494, "grad_norm": 0.24463728070259094, "learning_rate": 3.670613586494942e-07, "loss": 0.0118, "step": 13556 }, { "epoch": 0.9316245189664651, "grad_norm": 0.48370540142059326, "learning_rate": 3.6632767102744603e-07, "loss": 0.0099, "step": 13557 }, { "epoch": 0.9316932380428807, "grad_norm": 0.8826115727424622, "learning_rate": 3.655947083274369e-07, "loss": 0.0066, "step": 13558 }, { "epoch": 0.9317619571192963, "grad_norm": 0.3506598174571991, "learning_rate": 3.6486247058578116e-07, "loss": 0.0081, "step": 13559 }, { "epoch": 0.9318306761957119, "grad_norm": 0.1744268387556076, "learning_rate": 3.6413095783874806e-07, "loss": 0.0063, "step": 13560 }, { "epoch": 0.9318993952721275, "grad_norm": 0.6691838502883911, "learning_rate": 3.6340017012257866e-07, "loss": 0.01, "step": 13561 }, { "epoch": 0.9319681143485432, "grad_norm": 0.39854857325553894, "learning_rate": 3.626701074734723e-07, "loss": 0.0065, "step": 13562 }, { "epoch": 0.9320368334249588, "grad_norm": 0.35324734449386597, "learning_rate": 3.6194076992759675e-07, "loss": 0.0088, "step": 13563 }, { "epoch": 0.9321055525013744, "grad_norm": 0.2672683894634247, "learning_rate": 3.612121575210797e-07, "loss": 0.0094, "step": 13564 }, { "epoch": 0.93217427157779, "grad_norm": 0.5879917144775391, "learning_rate": 3.604842702900191e-07, "loss": 0.0085, "step": 13565 }, { "epoch": 0.9322429906542056, "grad_norm": 0.29145973920822144, "learning_rate": 3.597571082704676e-07, "loss": 0.0059, "step": 13566 }, { "epoch": 0.9323117097306213, "grad_norm": 0.5274559855461121, "learning_rate": 3.5903067149844983e-07, "loss": 0.0113, "step": 13567 }, { "epoch": 0.9323804288070369, "grad_norm": 0.29618072509765625, "learning_rate": 3.583049600099536e-07, "loss": 0.0074, "step": 13568 }, { "epoch": 0.9324491478834525, "grad_norm": 0.5263134241104126, "learning_rate": 3.575799738409252e-07, "loss": 0.0079, "step": 13569 }, { "epoch": 0.932517866959868, "grad_norm": 0.3540875315666199, "learning_rate": 3.5685571302728425e-07, "loss": 0.0139, "step": 13570 }, { "epoch": 0.9325865860362836, "grad_norm": 0.25444504618644714, "learning_rate": 3.5613217760490366e-07, "loss": 0.0033, "step": 13571 }, { "epoch": 0.9326553051126993, "grad_norm": 0.5449948310852051, "learning_rate": 3.5540936760962985e-07, "loss": 0.0107, "step": 13572 }, { "epoch": 0.9327240241891149, "grad_norm": 0.08586228638887405, "learning_rate": 3.546872830772657e-07, "loss": 0.0071, "step": 13573 }, { "epoch": 0.9327927432655305, "grad_norm": 0.5527995228767395, "learning_rate": 3.5396592404358776e-07, "loss": 0.0079, "step": 13574 }, { "epoch": 0.9328614623419461, "grad_norm": 0.12568369507789612, "learning_rate": 3.53245290544324e-07, "loss": 0.0064, "step": 13575 }, { "epoch": 0.9329301814183617, "grad_norm": 0.24613459408283234, "learning_rate": 3.525253826151792e-07, "loss": 0.0039, "step": 13576 }, { "epoch": 0.9329989004947774, "grad_norm": 0.2992902100086212, "learning_rate": 3.518062002918099e-07, "loss": 0.0127, "step": 13577 }, { "epoch": 0.933067619571193, "grad_norm": 0.18589679896831512, "learning_rate": 3.510877436098492e-07, "loss": 0.0104, "step": 13578 }, { "epoch": 0.9331363386476086, "grad_norm": 0.08583872765302658, "learning_rate": 3.5037001260488033e-07, "loss": 0.0071, "step": 13579 }, { "epoch": 0.9332050577240242, "grad_norm": 0.28233224153518677, "learning_rate": 3.496530073124665e-07, "loss": 0.0096, "step": 13580 }, { "epoch": 0.9332737768004398, "grad_norm": 0.4187141954898834, "learning_rate": 3.48936727768121e-07, "loss": 0.0095, "step": 13581 }, { "epoch": 0.9333424958768554, "grad_norm": 0.16928179562091827, "learning_rate": 3.4822117400733044e-07, "loss": 0.0065, "step": 13582 }, { "epoch": 0.9334112149532711, "grad_norm": 0.2998638451099396, "learning_rate": 3.475063460655381e-07, "loss": 0.0113, "step": 13583 }, { "epoch": 0.9334799340296867, "grad_norm": 0.21290215849876404, "learning_rate": 3.4679224397815903e-07, "loss": 0.0094, "step": 13584 }, { "epoch": 0.9335486531061022, "grad_norm": 0.27085208892822266, "learning_rate": 3.4607886778056485e-07, "loss": 0.0116, "step": 13585 }, { "epoch": 0.9336173721825178, "grad_norm": 0.4530967175960541, "learning_rate": 3.453662175080957e-07, "loss": 0.0108, "step": 13586 }, { "epoch": 0.9336860912589334, "grad_norm": 0.3435389995574951, "learning_rate": 3.4465429319605834e-07, "loss": 0.0136, "step": 13587 }, { "epoch": 0.9337548103353491, "grad_norm": 0.31403231620788574, "learning_rate": 3.4394309487971454e-07, "loss": 0.007, "step": 13588 }, { "epoch": 0.9338235294117647, "grad_norm": 0.13458187878131866, "learning_rate": 3.432326225943011e-07, "loss": 0.0045, "step": 13589 }, { "epoch": 0.9338922484881803, "grad_norm": 0.4530683755874634, "learning_rate": 3.4252287637500825e-07, "loss": 0.0099, "step": 13590 }, { "epoch": 0.9339609675645959, "grad_norm": 0.41057834029197693, "learning_rate": 3.418138562569978e-07, "loss": 0.0103, "step": 13591 }, { "epoch": 0.9340296866410115, "grad_norm": 0.19182829558849335, "learning_rate": 3.411055622753917e-07, "loss": 0.005, "step": 13592 }, { "epoch": 0.9340984057174272, "grad_norm": 0.11463779956102371, "learning_rate": 3.4039799446528185e-07, "loss": 0.0082, "step": 13593 }, { "epoch": 0.9341671247938428, "grad_norm": 0.2163136601448059, "learning_rate": 3.396911528617136e-07, "loss": 0.008, "step": 13594 }, { "epoch": 0.9342358438702584, "grad_norm": 0.21846957504749298, "learning_rate": 3.3898503749970724e-07, "loss": 0.0055, "step": 13595 }, { "epoch": 0.934304562946674, "grad_norm": 0.08385887742042542, "learning_rate": 3.3827964841423974e-07, "loss": 0.0075, "step": 13596 }, { "epoch": 0.9343732820230896, "grad_norm": 0.507124662399292, "learning_rate": 3.375749856402549e-07, "loss": 0.0098, "step": 13597 }, { "epoch": 0.9344420010995053, "grad_norm": 0.20859017968177795, "learning_rate": 3.3687104921265976e-07, "loss": 0.0086, "step": 13598 }, { "epoch": 0.9345107201759209, "grad_norm": 0.5154660940170288, "learning_rate": 3.3616783916632807e-07, "loss": 0.004, "step": 13599 }, { "epoch": 0.9345794392523364, "grad_norm": 0.31941288709640503, "learning_rate": 3.354653555360937e-07, "loss": 0.0097, "step": 13600 }, { "epoch": 0.934648158328752, "grad_norm": 0.3357228934764862, "learning_rate": 3.347635983567571e-07, "loss": 0.0098, "step": 13601 }, { "epoch": 0.9347168774051676, "grad_norm": 0.245613694190979, "learning_rate": 3.3406256766308043e-07, "loss": 0.01, "step": 13602 }, { "epoch": 0.9347855964815833, "grad_norm": 0.35716164112091064, "learning_rate": 3.333622634897926e-07, "loss": 0.0122, "step": 13603 }, { "epoch": 0.9348543155579989, "grad_norm": 0.18123535811901093, "learning_rate": 3.326626858715859e-07, "loss": 0.008, "step": 13604 }, { "epoch": 0.9349230346344145, "grad_norm": 0.23788031935691833, "learning_rate": 3.319638348431142e-07, "loss": 0.0051, "step": 13605 }, { "epoch": 0.9349917537108301, "grad_norm": 0.2909822463989258, "learning_rate": 3.312657104389982e-07, "loss": 0.0106, "step": 13606 }, { "epoch": 0.9350604727872457, "grad_norm": 0.37929099798202515, "learning_rate": 3.3056831269382194e-07, "loss": 0.0071, "step": 13607 }, { "epoch": 0.9351291918636614, "grad_norm": 0.277159720659256, "learning_rate": 3.29871641642131e-07, "loss": 0.0086, "step": 13608 }, { "epoch": 0.935197910940077, "grad_norm": 0.33752334117889404, "learning_rate": 3.291756973184379e-07, "loss": 0.0103, "step": 13609 }, { "epoch": 0.9352666300164926, "grad_norm": 0.23082220554351807, "learning_rate": 3.2848047975721996e-07, "loss": 0.0042, "step": 13610 }, { "epoch": 0.9353353490929082, "grad_norm": 0.17473042011260986, "learning_rate": 3.277859889929147e-07, "loss": 0.0031, "step": 13611 }, { "epoch": 0.9354040681693238, "grad_norm": 0.525973916053772, "learning_rate": 3.270922250599295e-07, "loss": 0.0091, "step": 13612 }, { "epoch": 0.9354727872457395, "grad_norm": 0.36988452076911926, "learning_rate": 3.2639918799262703e-07, "loss": 0.0047, "step": 13613 }, { "epoch": 0.935541506322155, "grad_norm": 0.20269761979579926, "learning_rate": 3.2570687782534136e-07, "loss": 0.0071, "step": 13614 }, { "epoch": 0.9356102253985706, "grad_norm": 0.8792168498039246, "learning_rate": 3.250152945923668e-07, "loss": 0.0125, "step": 13615 }, { "epoch": 0.9356789444749862, "grad_norm": 0.23320791125297546, "learning_rate": 3.243244383279659e-07, "loss": 0.0074, "step": 13616 }, { "epoch": 0.9357476635514018, "grad_norm": 0.45897942781448364, "learning_rate": 3.236343090663596e-07, "loss": 0.0088, "step": 13617 }, { "epoch": 0.9358163826278175, "grad_norm": 0.27339741587638855, "learning_rate": 3.2294490684173726e-07, "loss": 0.0109, "step": 13618 }, { "epoch": 0.9358851017042331, "grad_norm": 0.29087477922439575, "learning_rate": 3.2225623168824827e-07, "loss": 0.0052, "step": 13619 }, { "epoch": 0.9359538207806487, "grad_norm": 0.365413635969162, "learning_rate": 3.215682836400102e-07, "loss": 0.0064, "step": 13620 }, { "epoch": 0.9360225398570643, "grad_norm": 0.1886383444070816, "learning_rate": 3.2088106273110096e-07, "loss": 0.0044, "step": 13621 }, { "epoch": 0.9360912589334799, "grad_norm": 1.0894157886505127, "learning_rate": 3.201945689955649e-07, "loss": 0.0132, "step": 13622 }, { "epoch": 0.9361599780098956, "grad_norm": 0.38550639152526855, "learning_rate": 3.1950880246740975e-07, "loss": 0.0067, "step": 13623 }, { "epoch": 0.9362286970863112, "grad_norm": 0.3478456735610962, "learning_rate": 3.188237631806068e-07, "loss": 0.0086, "step": 13624 }, { "epoch": 0.9362974161627268, "grad_norm": 0.7258639335632324, "learning_rate": 3.1813945116909216e-07, "loss": 0.0097, "step": 13625 }, { "epoch": 0.9363661352391424, "grad_norm": 0.1531635820865631, "learning_rate": 3.174558664667637e-07, "loss": 0.0053, "step": 13626 }, { "epoch": 0.936434854315558, "grad_norm": 0.6006904244422913, "learning_rate": 3.167730091074844e-07, "loss": 0.0082, "step": 13627 }, { "epoch": 0.9365035733919737, "grad_norm": 0.27017301321029663, "learning_rate": 3.160908791250838e-07, "loss": 0.0064, "step": 13628 }, { "epoch": 0.9365722924683892, "grad_norm": 0.29207342863082886, "learning_rate": 3.154094765533533e-07, "loss": 0.0061, "step": 13629 }, { "epoch": 0.9366410115448048, "grad_norm": 0.6334842443466187, "learning_rate": 3.1472880142604577e-07, "loss": 0.0102, "step": 13630 }, { "epoch": 0.9367097306212204, "grad_norm": 0.170387864112854, "learning_rate": 3.1404885377688433e-07, "loss": 0.0138, "step": 13631 }, { "epoch": 0.936778449697636, "grad_norm": 0.32662561535835266, "learning_rate": 3.13369633639547e-07, "loss": 0.0082, "step": 13632 }, { "epoch": 0.9368471687740517, "grad_norm": 0.3105301558971405, "learning_rate": 3.126911410476835e-07, "loss": 0.0067, "step": 13633 }, { "epoch": 0.9369158878504673, "grad_norm": 0.30918627977371216, "learning_rate": 3.120133760349053e-07, "loss": 0.0081, "step": 13634 }, { "epoch": 0.9369846069268829, "grad_norm": 0.242163747549057, "learning_rate": 3.113363386347873e-07, "loss": 0.0083, "step": 13635 }, { "epoch": 0.9370533260032985, "grad_norm": 0.6980826258659363, "learning_rate": 3.106600288808692e-07, "loss": 0.0055, "step": 13636 }, { "epoch": 0.9371220450797141, "grad_norm": 0.3539723753929138, "learning_rate": 3.0998444680665426e-07, "loss": 0.0099, "step": 13637 }, { "epoch": 0.9371907641561298, "grad_norm": 0.19811327755451202, "learning_rate": 3.0930959244560573e-07, "loss": 0.0045, "step": 13638 }, { "epoch": 0.9372594832325454, "grad_norm": 0.3096749186515808, "learning_rate": 3.0863546583115855e-07, "loss": 0.0125, "step": 13639 }, { "epoch": 0.937328202308961, "grad_norm": 0.1808672547340393, "learning_rate": 3.0796206699670594e-07, "loss": 0.0052, "step": 13640 }, { "epoch": 0.9373969213853766, "grad_norm": 0.6011495590209961, "learning_rate": 3.072893959756079e-07, "loss": 0.0074, "step": 13641 }, { "epoch": 0.9374656404617921, "grad_norm": 0.13803315162658691, "learning_rate": 3.066174528011861e-07, "loss": 0.0071, "step": 13642 }, { "epoch": 0.9375343595382079, "grad_norm": 0.1810043454170227, "learning_rate": 3.0594623750672567e-07, "loss": 0.0046, "step": 13643 }, { "epoch": 0.9376030786146234, "grad_norm": 0.5019959807395935, "learning_rate": 3.052757501254816e-07, "loss": 0.0091, "step": 13644 }, { "epoch": 0.937671797691039, "grad_norm": 0.22914539277553558, "learning_rate": 3.046059906906623e-07, "loss": 0.0122, "step": 13645 }, { "epoch": 0.9377405167674546, "grad_norm": 0.2895650267601013, "learning_rate": 3.0393695923545297e-07, "loss": 0.0068, "step": 13646 }, { "epoch": 0.9378092358438702, "grad_norm": 0.29965490102767944, "learning_rate": 3.032686557929903e-07, "loss": 0.0082, "step": 13647 }, { "epoch": 0.9378779549202859, "grad_norm": 0.18827834725379944, "learning_rate": 3.0260108039638635e-07, "loss": 0.0061, "step": 13648 }, { "epoch": 0.9379466739967015, "grad_norm": 0.2912387251853943, "learning_rate": 3.0193423307870615e-07, "loss": 0.0103, "step": 13649 }, { "epoch": 0.9380153930731171, "grad_norm": 0.2595105469226837, "learning_rate": 3.012681138729867e-07, "loss": 0.0051, "step": 13650 }, { "epoch": 0.9380841121495327, "grad_norm": 0.15751540660858154, "learning_rate": 3.006027228122249e-07, "loss": 0.0048, "step": 13651 }, { "epoch": 0.9381528312259483, "grad_norm": 0.1563486009836197, "learning_rate": 2.9993805992938437e-07, "loss": 0.0044, "step": 13652 }, { "epoch": 0.938221550302364, "grad_norm": 0.5330849289894104, "learning_rate": 2.9927412525738884e-07, "loss": 0.0104, "step": 13653 }, { "epoch": 0.9382902693787796, "grad_norm": 0.31248733401298523, "learning_rate": 2.9861091882913196e-07, "loss": 0.0081, "step": 13654 }, { "epoch": 0.9383589884551952, "grad_norm": 0.44830062985420227, "learning_rate": 2.979484406774641e-07, "loss": 0.0148, "step": 13655 }, { "epoch": 0.9384277075316108, "grad_norm": 0.32545241713523865, "learning_rate": 2.9728669083520566e-07, "loss": 0.0064, "step": 13656 }, { "epoch": 0.9384964266080263, "grad_norm": 0.38862207531929016, "learning_rate": 2.9662566933513546e-07, "loss": 0.0115, "step": 13657 }, { "epoch": 0.938565145684442, "grad_norm": 0.26562702655792236, "learning_rate": 2.959653762100006e-07, "loss": 0.0047, "step": 13658 }, { "epoch": 0.9386338647608576, "grad_norm": 0.2030244767665863, "learning_rate": 2.9530581149251166e-07, "loss": 0.007, "step": 13659 }, { "epoch": 0.9387025838372732, "grad_norm": 0.11683307588100433, "learning_rate": 2.946469752153408e-07, "loss": 0.0061, "step": 13660 }, { "epoch": 0.9387713029136888, "grad_norm": 0.3329239785671234, "learning_rate": 2.9398886741112684e-07, "loss": 0.0075, "step": 13661 }, { "epoch": 0.9388400219901044, "grad_norm": 0.6210997104644775, "learning_rate": 2.933314881124688e-07, "loss": 0.0076, "step": 13662 }, { "epoch": 0.9389087410665201, "grad_norm": 0.289510577917099, "learning_rate": 2.926748373519339e-07, "loss": 0.0133, "step": 13663 }, { "epoch": 0.9389774601429357, "grad_norm": 0.22347430884838104, "learning_rate": 2.920189151620495e-07, "loss": 0.0062, "step": 13664 }, { "epoch": 0.9390461792193513, "grad_norm": 0.3328389823436737, "learning_rate": 2.913637215753112e-07, "loss": 0.0056, "step": 13665 }, { "epoch": 0.9391148982957669, "grad_norm": 0.5999282002449036, "learning_rate": 2.907092566241731e-07, "loss": 0.0145, "step": 13666 }, { "epoch": 0.9391836173721825, "grad_norm": 0.40306344628334045, "learning_rate": 2.900555203410593e-07, "loss": 0.0079, "step": 13667 }, { "epoch": 0.9392523364485982, "grad_norm": 1.1000726222991943, "learning_rate": 2.894025127583505e-07, "loss": 0.0187, "step": 13668 }, { "epoch": 0.9393210555250138, "grad_norm": 0.20695842802524567, "learning_rate": 2.8875023390840084e-07, "loss": 0.0117, "step": 13669 }, { "epoch": 0.9393897746014294, "grad_norm": 0.16292381286621094, "learning_rate": 2.880986838235145e-07, "loss": 0.0048, "step": 13670 }, { "epoch": 0.939458493677845, "grad_norm": 0.45755448937416077, "learning_rate": 2.874478625359772e-07, "loss": 0.0064, "step": 13671 }, { "epoch": 0.9395272127542605, "grad_norm": 0.24636448919773102, "learning_rate": 2.8679777007802334e-07, "loss": 0.0124, "step": 13672 }, { "epoch": 0.9395959318306762, "grad_norm": 0.1415424346923828, "learning_rate": 2.861484064818587e-07, "loss": 0.0046, "step": 13673 }, { "epoch": 0.9396646509070918, "grad_norm": 0.21764594316482544, "learning_rate": 2.8549977177965256e-07, "loss": 0.0043, "step": 13674 }, { "epoch": 0.9397333699835074, "grad_norm": 0.10833780467510223, "learning_rate": 2.8485186600353594e-07, "loss": 0.0043, "step": 13675 }, { "epoch": 0.939802089059923, "grad_norm": 0.13154222071170807, "learning_rate": 2.8420468918560316e-07, "loss": 0.0056, "step": 13676 }, { "epoch": 0.9398708081363386, "grad_norm": 0.06273692846298218, "learning_rate": 2.8355824135791685e-07, "loss": 0.0065, "step": 13677 }, { "epoch": 0.9399395272127543, "grad_norm": 0.26275408267974854, "learning_rate": 2.8291252255250146e-07, "loss": 0.0085, "step": 13678 }, { "epoch": 0.9400082462891699, "grad_norm": 0.27153703570365906, "learning_rate": 2.8226753280133964e-07, "loss": 0.0082, "step": 13679 }, { "epoch": 0.9400769653655855, "grad_norm": 0.05460895970463753, "learning_rate": 2.816232721363859e-07, "loss": 0.0287, "step": 13680 }, { "epoch": 0.9401456844420011, "grad_norm": 0.5617883205413818, "learning_rate": 2.809797405895581e-07, "loss": 0.0057, "step": 13681 }, { "epoch": 0.9402144035184167, "grad_norm": 0.4648374319076538, "learning_rate": 2.8033693819273064e-07, "loss": 0.0103, "step": 13682 }, { "epoch": 0.9402831225948324, "grad_norm": 0.1581858992576599, "learning_rate": 2.7969486497774977e-07, "loss": 0.0057, "step": 13683 }, { "epoch": 0.940351841671248, "grad_norm": 0.6080924272537231, "learning_rate": 2.790535209764217e-07, "loss": 0.0097, "step": 13684 }, { "epoch": 0.9404205607476636, "grad_norm": 0.9017139673233032, "learning_rate": 2.7841290622051764e-07, "loss": 0.0067, "step": 13685 }, { "epoch": 0.9404892798240791, "grad_norm": 0.29187101125717163, "learning_rate": 2.7777302074177224e-07, "loss": 0.009, "step": 13686 }, { "epoch": 0.9405579989004947, "grad_norm": 0.6425854563713074, "learning_rate": 2.7713386457188185e-07, "loss": 0.0062, "step": 13687 }, { "epoch": 0.9406267179769104, "grad_norm": 0.2940348982810974, "learning_rate": 2.764954377425111e-07, "loss": 0.0047, "step": 13688 }, { "epoch": 0.940695437053326, "grad_norm": 0.2022000104188919, "learning_rate": 2.75857740285288e-07, "loss": 0.0114, "step": 13689 }, { "epoch": 0.9407641561297416, "grad_norm": 0.17208269238471985, "learning_rate": 2.752207722318006e-07, "loss": 0.0031, "step": 13690 }, { "epoch": 0.9408328752061572, "grad_norm": 0.5662787556648254, "learning_rate": 2.7458453361360215e-07, "loss": 0.0096, "step": 13691 }, { "epoch": 0.9409015942825728, "grad_norm": 0.12352833151817322, "learning_rate": 2.7394902446221227e-07, "loss": 0.0097, "step": 13692 }, { "epoch": 0.9409703133589885, "grad_norm": 1.0074549913406372, "learning_rate": 2.733142448091108e-07, "loss": 0.0072, "step": 13693 }, { "epoch": 0.9410390324354041, "grad_norm": 0.17582930624485016, "learning_rate": 2.72680194685746e-07, "loss": 0.005, "step": 13694 }, { "epoch": 0.9411077515118197, "grad_norm": 0.31647780537605286, "learning_rate": 2.720468741235277e-07, "loss": 0.0059, "step": 13695 }, { "epoch": 0.9411764705882353, "grad_norm": 0.23916791379451752, "learning_rate": 2.714142831538258e-07, "loss": 0.0091, "step": 13696 }, { "epoch": 0.9412451896646509, "grad_norm": 0.14593873918056488, "learning_rate": 2.7078242180798016e-07, "loss": 0.0046, "step": 13697 }, { "epoch": 0.9413139087410666, "grad_norm": 0.25226688385009766, "learning_rate": 2.7015129011729256e-07, "loss": 0.0057, "step": 13698 }, { "epoch": 0.9413826278174822, "grad_norm": 0.3843546211719513, "learning_rate": 2.695208881130262e-07, "loss": 0.0078, "step": 13699 }, { "epoch": 0.9414513468938978, "grad_norm": 0.35669779777526855, "learning_rate": 2.6889121582640953e-07, "loss": 0.0073, "step": 13700 }, { "epoch": 0.9415200659703133, "grad_norm": 0.3833012580871582, "learning_rate": 2.682622732886375e-07, "loss": 0.0069, "step": 13701 }, { "epoch": 0.9415887850467289, "grad_norm": 0.39985933899879456, "learning_rate": 2.676340605308636e-07, "loss": 0.0076, "step": 13702 }, { "epoch": 0.9416575041231446, "grad_norm": 0.2590758204460144, "learning_rate": 2.670065775842112e-07, "loss": 0.0088, "step": 13703 }, { "epoch": 0.9417262231995602, "grad_norm": 0.16501447558403015, "learning_rate": 2.663798244797622e-07, "loss": 0.0035, "step": 13704 }, { "epoch": 0.9417949422759758, "grad_norm": 0.4100697636604309, "learning_rate": 2.657538012485683e-07, "loss": 0.0107, "step": 13705 }, { "epoch": 0.9418636613523914, "grad_norm": 0.7545502781867981, "learning_rate": 2.6512850792163477e-07, "loss": 0.0189, "step": 13706 }, { "epoch": 0.941932380428807, "grad_norm": 0.2053503841161728, "learning_rate": 2.6450394452994355e-07, "loss": 0.0101, "step": 13707 }, { "epoch": 0.9420010995052226, "grad_norm": 0.16653303802013397, "learning_rate": 2.638801111044314e-07, "loss": 0.008, "step": 13708 }, { "epoch": 0.9420698185816383, "grad_norm": 0.12821266055107117, "learning_rate": 2.632570076760038e-07, "loss": 0.0065, "step": 13709 }, { "epoch": 0.9421385376580539, "grad_norm": 0.48749810457229614, "learning_rate": 2.6263463427552424e-07, "loss": 0.0083, "step": 13710 }, { "epoch": 0.9422072567344695, "grad_norm": 0.5016224980354309, "learning_rate": 2.6201299093382814e-07, "loss": 0.008, "step": 13711 }, { "epoch": 0.9422759758108851, "grad_norm": 0.17491865158081055, "learning_rate": 2.6139207768170583e-07, "loss": 0.0055, "step": 13712 }, { "epoch": 0.9423446948873007, "grad_norm": 0.17115561664104462, "learning_rate": 2.6077189454992265e-07, "loss": 0.0093, "step": 13713 }, { "epoch": 0.9424134139637164, "grad_norm": 0.2372978776693344, "learning_rate": 2.6015244156919415e-07, "loss": 0.0057, "step": 13714 }, { "epoch": 0.942482133040132, "grad_norm": 0.12369297444820404, "learning_rate": 2.595337187702107e-07, "loss": 0.0077, "step": 13715 }, { "epoch": 0.9425508521165475, "grad_norm": 0.21685834228992462, "learning_rate": 2.589157261836228e-07, "loss": 0.0074, "step": 13716 }, { "epoch": 0.9426195711929631, "grad_norm": 0.39538419246673584, "learning_rate": 2.582984638400426e-07, "loss": 0.0096, "step": 13717 }, { "epoch": 0.9426882902693787, "grad_norm": 0.485442578792572, "learning_rate": 2.5768193177004907e-07, "loss": 0.0169, "step": 13718 }, { "epoch": 0.9427570093457944, "grad_norm": 0.15962554514408112, "learning_rate": 2.5706613000418265e-07, "loss": 0.0061, "step": 13719 }, { "epoch": 0.94282572842221, "grad_norm": 0.435630738735199, "learning_rate": 2.564510585729507e-07, "loss": 0.0075, "step": 13720 }, { "epoch": 0.9428944474986256, "grad_norm": 0.16719099879264832, "learning_rate": 2.5583671750682216e-07, "loss": 0.0078, "step": 13721 }, { "epoch": 0.9429631665750412, "grad_norm": 0.36712953448295593, "learning_rate": 2.5522310683623097e-07, "loss": 0.0063, "step": 13722 }, { "epoch": 0.9430318856514568, "grad_norm": 0.4275752604007721, "learning_rate": 2.546102265915712e-07, "loss": 0.0068, "step": 13723 }, { "epoch": 0.9431006047278725, "grad_norm": 0.254611611366272, "learning_rate": 2.5399807680320517e-07, "loss": 0.0063, "step": 13724 }, { "epoch": 0.9431693238042881, "grad_norm": 0.4232203960418701, "learning_rate": 2.533866575014587e-07, "loss": 0.0095, "step": 13725 }, { "epoch": 0.9432380428807037, "grad_norm": 0.4367760419845581, "learning_rate": 2.527759687166209e-07, "loss": 0.01, "step": 13726 }, { "epoch": 0.9433067619571193, "grad_norm": 0.2973985970020294, "learning_rate": 2.5216601047894085e-07, "loss": 0.0056, "step": 13727 }, { "epoch": 0.9433754810335349, "grad_norm": 0.33978405594825745, "learning_rate": 2.515567828186377e-07, "loss": 0.0079, "step": 13728 }, { "epoch": 0.9434442001099506, "grad_norm": 0.6211637258529663, "learning_rate": 2.5094828576588733e-07, "loss": 0.0088, "step": 13729 }, { "epoch": 0.9435129191863662, "grad_norm": 0.6476461291313171, "learning_rate": 2.50340519350839e-07, "loss": 0.0123, "step": 13730 }, { "epoch": 0.9435816382627817, "grad_norm": 0.2791706919670105, "learning_rate": 2.497334836035953e-07, "loss": 0.0072, "step": 13731 }, { "epoch": 0.9436503573391973, "grad_norm": 0.24475207924842834, "learning_rate": 2.491271785542304e-07, "loss": 0.0081, "step": 13732 }, { "epoch": 0.9437190764156129, "grad_norm": 0.13380102813243866, "learning_rate": 2.485216042327787e-07, "loss": 0.0031, "step": 13733 }, { "epoch": 0.9437877954920286, "grad_norm": 0.310315877199173, "learning_rate": 2.479167606692395e-07, "loss": 0.0042, "step": 13734 }, { "epoch": 0.9438565145684442, "grad_norm": 0.15399692952632904, "learning_rate": 2.4731264789357553e-07, "loss": 0.0155, "step": 13735 }, { "epoch": 0.9439252336448598, "grad_norm": 0.3602693974971771, "learning_rate": 2.467092659357112e-07, "loss": 0.0048, "step": 13736 }, { "epoch": 0.9439939527212754, "grad_norm": 0.12477459758520126, "learning_rate": 2.461066148255392e-07, "loss": 0.004, "step": 13737 }, { "epoch": 0.944062671797691, "grad_norm": 0.3597106337547302, "learning_rate": 2.4550469459291237e-07, "loss": 0.0057, "step": 13738 }, { "epoch": 0.9441313908741067, "grad_norm": 0.3506683111190796, "learning_rate": 2.4490350526764847e-07, "loss": 0.008, "step": 13739 }, { "epoch": 0.9442001099505223, "grad_norm": 0.2577536404132843, "learning_rate": 2.4430304687953045e-07, "loss": 0.0064, "step": 13740 }, { "epoch": 0.9442688290269379, "grad_norm": 0.501133918762207, "learning_rate": 2.437033194583044e-07, "loss": 0.0069, "step": 13741 }, { "epoch": 0.9443375481033535, "grad_norm": 0.3098453879356384, "learning_rate": 2.431043230336749e-07, "loss": 0.012, "step": 13742 }, { "epoch": 0.944406267179769, "grad_norm": 0.6630953550338745, "learning_rate": 2.4250605763532e-07, "loss": 0.0081, "step": 13743 }, { "epoch": 0.9444749862561848, "grad_norm": 0.12061759829521179, "learning_rate": 2.4190852329287585e-07, "loss": 0.0019, "step": 13744 }, { "epoch": 0.9445437053326003, "grad_norm": 0.13367986679077148, "learning_rate": 2.413117200359422e-07, "loss": 0.0093, "step": 13745 }, { "epoch": 0.9446124244090159, "grad_norm": 0.34042325615882874, "learning_rate": 2.4071564789408196e-07, "loss": 0.0045, "step": 13746 }, { "epoch": 0.9446811434854315, "grad_norm": 0.3444574773311615, "learning_rate": 2.4012030689682494e-07, "loss": 0.0271, "step": 13747 }, { "epoch": 0.9447498625618471, "grad_norm": 0.22765150666236877, "learning_rate": 2.395256970736626e-07, "loss": 0.0122, "step": 13748 }, { "epoch": 0.9448185816382628, "grad_norm": 0.07800406217575073, "learning_rate": 2.3893181845405287e-07, "loss": 0.0055, "step": 13749 }, { "epoch": 0.9448873007146784, "grad_norm": 0.3341149091720581, "learning_rate": 2.383386710674107e-07, "loss": 0.0055, "step": 13750 }, { "epoch": 0.944956019791094, "grad_norm": 0.33397260308265686, "learning_rate": 2.377462549431242e-07, "loss": 0.0046, "step": 13751 }, { "epoch": 0.9450247388675096, "grad_norm": 0.3103964328765869, "learning_rate": 2.371545701105382e-07, "loss": 0.008, "step": 13752 }, { "epoch": 0.9450934579439252, "grad_norm": 0.1358519047498703, "learning_rate": 2.3656361659896263e-07, "loss": 0.0074, "step": 13753 }, { "epoch": 0.9451621770203409, "grad_norm": 0.49723392724990845, "learning_rate": 2.3597339443767406e-07, "loss": 0.0055, "step": 13754 }, { "epoch": 0.9452308960967565, "grad_norm": 0.4432712495326996, "learning_rate": 2.3538390365590913e-07, "loss": 0.0096, "step": 13755 }, { "epoch": 0.9452996151731721, "grad_norm": 1.1561975479125977, "learning_rate": 2.3479514428287107e-07, "loss": 0.0102, "step": 13756 }, { "epoch": 0.9453683342495877, "grad_norm": 0.5908234119415283, "learning_rate": 2.3420711634772664e-07, "loss": 0.0074, "step": 13757 }, { "epoch": 0.9454370533260033, "grad_norm": 0.3679724335670471, "learning_rate": 2.3361981987960412e-07, "loss": 0.006, "step": 13758 }, { "epoch": 0.945505772402419, "grad_norm": 0.6092364192008972, "learning_rate": 2.3303325490759698e-07, "loss": 0.009, "step": 13759 }, { "epoch": 0.9455744914788345, "grad_norm": 0.29103440046310425, "learning_rate": 2.3244742146076192e-07, "loss": 0.0111, "step": 13760 }, { "epoch": 0.9456432105552501, "grad_norm": 0.3293628394603729, "learning_rate": 2.3186231956812075e-07, "loss": 0.0073, "step": 13761 }, { "epoch": 0.9457119296316657, "grad_norm": 0.5646068453788757, "learning_rate": 2.312779492586603e-07, "loss": 0.0072, "step": 13762 }, { "epoch": 0.9457806487080813, "grad_norm": 0.04886654391884804, "learning_rate": 2.3069431056132405e-07, "loss": 0.0069, "step": 13763 }, { "epoch": 0.945849367784497, "grad_norm": 0.2630916237831116, "learning_rate": 2.3011140350503058e-07, "loss": 0.0068, "step": 13764 }, { "epoch": 0.9459180868609126, "grad_norm": 0.1839185506105423, "learning_rate": 2.2952922811865007e-07, "loss": 0.0062, "step": 13765 }, { "epoch": 0.9459868059373282, "grad_norm": 0.2709934413433075, "learning_rate": 2.2894778443102617e-07, "loss": 0.004, "step": 13766 }, { "epoch": 0.9460555250137438, "grad_norm": 0.23057250678539276, "learning_rate": 2.2836707247095913e-07, "loss": 0.0057, "step": 13767 }, { "epoch": 0.9461242440901594, "grad_norm": 0.35109788179397583, "learning_rate": 2.2778709226722094e-07, "loss": 0.0078, "step": 13768 }, { "epoch": 0.9461929631665751, "grad_norm": 0.2886403203010559, "learning_rate": 2.2720784384853865e-07, "loss": 0.0094, "step": 13769 }, { "epoch": 0.9462616822429907, "grad_norm": 0.3122682571411133, "learning_rate": 2.2662932724360763e-07, "loss": 0.0075, "step": 13770 }, { "epoch": 0.9463304013194063, "grad_norm": 0.27991417050361633, "learning_rate": 2.2605154248108663e-07, "loss": 0.0071, "step": 13771 }, { "epoch": 0.9463991203958219, "grad_norm": 0.07657133042812347, "learning_rate": 2.2547448958959938e-07, "loss": 0.0064, "step": 13772 }, { "epoch": 0.9464678394722374, "grad_norm": 0.23624981939792633, "learning_rate": 2.2489816859773139e-07, "loss": 0.0056, "step": 13773 }, { "epoch": 0.9465365585486532, "grad_norm": 0.4068703055381775, "learning_rate": 2.2432257953402979e-07, "loss": 0.0064, "step": 13774 }, { "epoch": 0.9466052776250687, "grad_norm": 0.6940404176712036, "learning_rate": 2.237477224270118e-07, "loss": 0.0086, "step": 13775 }, { "epoch": 0.9466739967014843, "grad_norm": 0.4846518933773041, "learning_rate": 2.231735973051513e-07, "loss": 0.0121, "step": 13776 }, { "epoch": 0.9467427157778999, "grad_norm": 0.15160998702049255, "learning_rate": 2.2260020419689386e-07, "loss": 0.007, "step": 13777 }, { "epoch": 0.9468114348543155, "grad_norm": 0.6571817398071289, "learning_rate": 2.2202754313063677e-07, "loss": 0.0099, "step": 13778 }, { "epoch": 0.9468801539307312, "grad_norm": 0.44110581278800964, "learning_rate": 2.2145561413475735e-07, "loss": 0.0039, "step": 13779 }, { "epoch": 0.9469488730071468, "grad_norm": 0.1920446753501892, "learning_rate": 2.2088441723757957e-07, "loss": 0.0094, "step": 13780 }, { "epoch": 0.9470175920835624, "grad_norm": 0.7706817984580994, "learning_rate": 2.203139524674058e-07, "loss": 0.0103, "step": 13781 }, { "epoch": 0.947086311159978, "grad_norm": 0.36199161410331726, "learning_rate": 2.197442198524935e-07, "loss": 0.0072, "step": 13782 }, { "epoch": 0.9471550302363936, "grad_norm": 0.629549503326416, "learning_rate": 2.19175219421065e-07, "loss": 0.0089, "step": 13783 }, { "epoch": 0.9472237493128093, "grad_norm": 0.2625589668750763, "learning_rate": 2.1860695120130614e-07, "loss": 0.0094, "step": 13784 }, { "epoch": 0.9472924683892249, "grad_norm": 0.07537581771612167, "learning_rate": 2.1803941522137106e-07, "loss": 0.0079, "step": 13785 }, { "epoch": 0.9473611874656405, "grad_norm": 0.07700857520103455, "learning_rate": 2.1747261150937226e-07, "loss": 0.0036, "step": 13786 }, { "epoch": 0.947429906542056, "grad_norm": 0.26073023676872253, "learning_rate": 2.1690654009339062e-07, "loss": 0.0033, "step": 13787 }, { "epoch": 0.9474986256184716, "grad_norm": 2.041771411895752, "learning_rate": 2.1634120100146537e-07, "loss": 0.0188, "step": 13788 }, { "epoch": 0.9475673446948873, "grad_norm": 0.21610230207443237, "learning_rate": 2.1577659426160244e-07, "loss": 0.0073, "step": 13789 }, { "epoch": 0.9476360637713029, "grad_norm": 0.5190749168395996, "learning_rate": 2.1521271990177282e-07, "loss": 0.0061, "step": 13790 }, { "epoch": 0.9477047828477185, "grad_norm": 0.2812754809856415, "learning_rate": 2.1464957794990746e-07, "loss": 0.008, "step": 13791 }, { "epoch": 0.9477735019241341, "grad_norm": 0.7189648747444153, "learning_rate": 2.1408716843390573e-07, "loss": 0.0108, "step": 13792 }, { "epoch": 0.9478422210005497, "grad_norm": 0.5380106568336487, "learning_rate": 2.1352549138162702e-07, "loss": 0.0114, "step": 13793 }, { "epoch": 0.9479109400769654, "grad_norm": 0.14130419492721558, "learning_rate": 2.1296454682089573e-07, "loss": 0.0071, "step": 13794 }, { "epoch": 0.947979659153381, "grad_norm": 0.2219991832971573, "learning_rate": 2.1240433477949961e-07, "loss": 0.0134, "step": 13795 }, { "epoch": 0.9480483782297966, "grad_norm": 0.10764750838279724, "learning_rate": 2.1184485528518983e-07, "loss": 0.006, "step": 13796 }, { "epoch": 0.9481170973062122, "grad_norm": 0.5843668580055237, "learning_rate": 2.1128610836568419e-07, "loss": 0.0147, "step": 13797 }, { "epoch": 0.9481858163826278, "grad_norm": 0.35373231768608093, "learning_rate": 2.1072809404866056e-07, "loss": 0.0071, "step": 13798 }, { "epoch": 0.9482545354590435, "grad_norm": 0.45219895243644714, "learning_rate": 2.1017081236176183e-07, "loss": 0.0058, "step": 13799 }, { "epoch": 0.9483232545354591, "grad_norm": 0.42346808314323425, "learning_rate": 2.0961426333259425e-07, "loss": 0.0084, "step": 13800 }, { "epoch": 0.9483919736118747, "grad_norm": 0.44558510184288025, "learning_rate": 2.0905844698872745e-07, "loss": 0.0071, "step": 13801 }, { "epoch": 0.9484606926882903, "grad_norm": 0.5317187905311584, "learning_rate": 2.0850336335769605e-07, "loss": 0.0064, "step": 13802 }, { "epoch": 0.9485294117647058, "grad_norm": 0.24734683334827423, "learning_rate": 2.0794901246699806e-07, "loss": 0.0064, "step": 13803 }, { "epoch": 0.9485981308411215, "grad_norm": 0.7321276664733887, "learning_rate": 2.0739539434409648e-07, "loss": 0.0064, "step": 13804 }, { "epoch": 0.9486668499175371, "grad_norm": 0.6377944946289062, "learning_rate": 2.0684250901641279e-07, "loss": 0.0165, "step": 13805 }, { "epoch": 0.9487355689939527, "grad_norm": 0.22518552839756012, "learning_rate": 2.0629035651134e-07, "loss": 0.0042, "step": 13806 }, { "epoch": 0.9488042880703683, "grad_norm": 0.2933584153652191, "learning_rate": 2.0573893685622624e-07, "loss": 0.006, "step": 13807 }, { "epoch": 0.9488730071467839, "grad_norm": 0.17524778842926025, "learning_rate": 2.0518825007839137e-07, "loss": 0.0108, "step": 13808 }, { "epoch": 0.9489417262231996, "grad_norm": 0.24949584901332855, "learning_rate": 2.0463829620511355e-07, "loss": 0.0148, "step": 13809 }, { "epoch": 0.9490104452996152, "grad_norm": 0.17212378978729248, "learning_rate": 2.0408907526363596e-07, "loss": 0.0076, "step": 13810 }, { "epoch": 0.9490791643760308, "grad_norm": 0.2795780301094055, "learning_rate": 2.0354058728116686e-07, "loss": 0.0087, "step": 13811 }, { "epoch": 0.9491478834524464, "grad_norm": 0.37061190605163574, "learning_rate": 2.029928322848762e-07, "loss": 0.0083, "step": 13812 }, { "epoch": 0.949216602528862, "grad_norm": 0.6426830291748047, "learning_rate": 2.0244581030190058e-07, "loss": 0.0085, "step": 13813 }, { "epoch": 0.9492853216052777, "grad_norm": 0.3528953194618225, "learning_rate": 2.018995213593333e-07, "loss": 0.0049, "step": 13814 }, { "epoch": 0.9493540406816933, "grad_norm": 0.5263612866401672, "learning_rate": 2.0135396548424444e-07, "loss": 0.0088, "step": 13815 }, { "epoch": 0.9494227597581089, "grad_norm": 0.2488842010498047, "learning_rate": 2.0080914270365403e-07, "loss": 0.0076, "step": 13816 }, { "epoch": 0.9494914788345244, "grad_norm": 0.2899072468280792, "learning_rate": 2.0026505304455211e-07, "loss": 0.0207, "step": 13817 }, { "epoch": 0.94956019791094, "grad_norm": 0.24767738580703735, "learning_rate": 1.9972169653389382e-07, "loss": 0.0075, "step": 13818 }, { "epoch": 0.9496289169873557, "grad_norm": 0.45458224415779114, "learning_rate": 1.991790731985943e-07, "loss": 0.0065, "step": 13819 }, { "epoch": 0.9496976360637713, "grad_norm": 0.4509918987751007, "learning_rate": 1.9863718306553204e-07, "loss": 0.0079, "step": 13820 }, { "epoch": 0.9497663551401869, "grad_norm": 0.5402864813804626, "learning_rate": 1.9809602616155554e-07, "loss": 0.0075, "step": 13821 }, { "epoch": 0.9498350742166025, "grad_norm": 0.6461368203163147, "learning_rate": 1.9755560251347003e-07, "loss": 0.0113, "step": 13822 }, { "epoch": 0.9499037932930181, "grad_norm": 0.3719472885131836, "learning_rate": 1.9701591214804915e-07, "loss": 0.0091, "step": 13823 }, { "epoch": 0.9499725123694338, "grad_norm": 0.18273375928401947, "learning_rate": 1.9647695509202313e-07, "loss": 0.0068, "step": 13824 }, { "epoch": 0.9500412314458494, "grad_norm": 0.4415549039840698, "learning_rate": 1.9593873137209728e-07, "loss": 0.0072, "step": 13825 }, { "epoch": 0.950109950522265, "grad_norm": 0.2924000322818756, "learning_rate": 1.9540124101492863e-07, "loss": 0.0048, "step": 13826 }, { "epoch": 0.9501786695986806, "grad_norm": 0.290527880191803, "learning_rate": 1.9486448404714585e-07, "loss": 0.0053, "step": 13827 }, { "epoch": 0.9502473886750962, "grad_norm": 0.4053552746772766, "learning_rate": 1.9432846049533935e-07, "loss": 0.0073, "step": 13828 }, { "epoch": 0.9503161077515119, "grad_norm": 0.3014475107192993, "learning_rate": 1.9379317038606126e-07, "loss": 0.0078, "step": 13829 }, { "epoch": 0.9503848268279275, "grad_norm": 0.5754907727241516, "learning_rate": 1.9325861374582865e-07, "loss": 0.0082, "step": 13830 }, { "epoch": 0.9504535459043431, "grad_norm": 0.34822651743888855, "learning_rate": 1.927247906011237e-07, "loss": 0.0097, "step": 13831 }, { "epoch": 0.9505222649807586, "grad_norm": 0.09432534873485565, "learning_rate": 1.9219170097838857e-07, "loss": 0.0042, "step": 13832 }, { "epoch": 0.9505909840571742, "grad_norm": 0.3194836974143982, "learning_rate": 1.916593449040338e-07, "loss": 0.0073, "step": 13833 }, { "epoch": 0.9506597031335899, "grad_norm": 0.0697643980383873, "learning_rate": 1.9112772240443167e-07, "loss": 0.0047, "step": 13834 }, { "epoch": 0.9507284222100055, "grad_norm": 0.45697805285453796, "learning_rate": 1.9059683350591605e-07, "loss": 0.009, "step": 13835 }, { "epoch": 0.9507971412864211, "grad_norm": 0.1644899845123291, "learning_rate": 1.9006667823478597e-07, "loss": 0.0102, "step": 13836 }, { "epoch": 0.9508658603628367, "grad_norm": 0.5966219305992126, "learning_rate": 1.8953725661730537e-07, "loss": 0.0069, "step": 13837 }, { "epoch": 0.9509345794392523, "grad_norm": 0.3469698429107666, "learning_rate": 1.890085686796983e-07, "loss": 0.0107, "step": 13838 }, { "epoch": 0.9510032985156679, "grad_norm": 0.27669084072113037, "learning_rate": 1.8848061444815712e-07, "loss": 0.0061, "step": 13839 }, { "epoch": 0.9510720175920836, "grad_norm": 0.1651138961315155, "learning_rate": 1.8795339394883758e-07, "loss": 0.0047, "step": 13840 }, { "epoch": 0.9511407366684992, "grad_norm": 1.2335443496704102, "learning_rate": 1.8742690720785215e-07, "loss": 0.0173, "step": 13841 }, { "epoch": 0.9512094557449148, "grad_norm": 1.0851097106933594, "learning_rate": 1.8690115425128497e-07, "loss": 0.0053, "step": 13842 }, { "epoch": 0.9512781748213304, "grad_norm": 0.3165965974330902, "learning_rate": 1.8637613510518014e-07, "loss": 0.0061, "step": 13843 }, { "epoch": 0.951346893897746, "grad_norm": 0.1613205075263977, "learning_rate": 1.8585184979554525e-07, "loss": 0.0036, "step": 13844 }, { "epoch": 0.9514156129741617, "grad_norm": 0.3769141137599945, "learning_rate": 1.853282983483562e-07, "loss": 0.0099, "step": 13845 }, { "epoch": 0.9514843320505773, "grad_norm": 0.5212210416793823, "learning_rate": 1.848054807895422e-07, "loss": 0.007, "step": 13846 }, { "epoch": 0.9515530511269928, "grad_norm": 0.3759664297103882, "learning_rate": 1.8428339714500753e-07, "loss": 0.0111, "step": 13847 }, { "epoch": 0.9516217702034084, "grad_norm": 0.04473366588354111, "learning_rate": 1.837620474406132e-07, "loss": 0.0042, "step": 13848 }, { "epoch": 0.951690489279824, "grad_norm": 0.21748881042003632, "learning_rate": 1.8324143170218687e-07, "loss": 0.0071, "step": 13849 }, { "epoch": 0.9517592083562397, "grad_norm": 0.42316076159477234, "learning_rate": 1.8272154995551626e-07, "loss": 0.0116, "step": 13850 }, { "epoch": 0.9518279274326553, "grad_norm": 0.46114593744277954, "learning_rate": 1.8220240222635908e-07, "loss": 0.0062, "step": 13851 }, { "epoch": 0.9518966465090709, "grad_norm": 0.3957432806491852, "learning_rate": 1.8168398854042813e-07, "loss": 0.0087, "step": 13852 }, { "epoch": 0.9519653655854865, "grad_norm": 0.7284746170043945, "learning_rate": 1.8116630892340946e-07, "loss": 0.0081, "step": 13853 }, { "epoch": 0.9520340846619021, "grad_norm": 0.14430475234985352, "learning_rate": 1.8064936340094594e-07, "loss": 0.0136, "step": 13854 }, { "epoch": 0.9521028037383178, "grad_norm": 0.46582624316215515, "learning_rate": 1.8013315199864545e-07, "loss": 0.0103, "step": 13855 }, { "epoch": 0.9521715228147334, "grad_norm": 0.3023558259010315, "learning_rate": 1.7961767474207746e-07, "loss": 0.0097, "step": 13856 }, { "epoch": 0.952240241891149, "grad_norm": 0.21386860311031342, "learning_rate": 1.7910293165678326e-07, "loss": 0.0069, "step": 13857 }, { "epoch": 0.9523089609675646, "grad_norm": 0.30582326650619507, "learning_rate": 1.7858892276825745e-07, "loss": 0.0123, "step": 13858 }, { "epoch": 0.9523776800439802, "grad_norm": 0.4422227740287781, "learning_rate": 1.7807564810196798e-07, "loss": 0.0098, "step": 13859 }, { "epoch": 0.9524463991203959, "grad_norm": 0.4114679992198944, "learning_rate": 1.7756310768333616e-07, "loss": 0.0063, "step": 13860 }, { "epoch": 0.9525151181968114, "grad_norm": 0.1530548334121704, "learning_rate": 1.7705130153775506e-07, "loss": 0.0066, "step": 13861 }, { "epoch": 0.952583837273227, "grad_norm": 0.49085795879364014, "learning_rate": 1.7654022969057604e-07, "loss": 0.008, "step": 13862 }, { "epoch": 0.9526525563496426, "grad_norm": 0.33327943086624146, "learning_rate": 1.760298921671205e-07, "loss": 0.005, "step": 13863 }, { "epoch": 0.9527212754260582, "grad_norm": 0.07118391245603561, "learning_rate": 1.755202889926666e-07, "loss": 0.0049, "step": 13864 }, { "epoch": 0.9527899945024739, "grad_norm": 0.1820707768201828, "learning_rate": 1.750114201924591e-07, "loss": 0.0097, "step": 13865 }, { "epoch": 0.9528587135788895, "grad_norm": 0.2387109100818634, "learning_rate": 1.745032857917078e-07, "loss": 0.0127, "step": 13866 }, { "epoch": 0.9529274326553051, "grad_norm": 0.27168652415275574, "learning_rate": 1.7399588581558267e-07, "loss": 0.0064, "step": 13867 }, { "epoch": 0.9529961517317207, "grad_norm": 0.9613514542579651, "learning_rate": 1.7348922028922186e-07, "loss": 0.0051, "step": 13868 }, { "epoch": 0.9530648708081363, "grad_norm": 0.3806745409965515, "learning_rate": 1.7298328923772032e-07, "loss": 0.0083, "step": 13869 }, { "epoch": 0.953133589884552, "grad_norm": 0.3058290183544159, "learning_rate": 1.7247809268614633e-07, "loss": 0.0064, "step": 13870 }, { "epoch": 0.9532023089609676, "grad_norm": 0.7352340817451477, "learning_rate": 1.719736306595232e-07, "loss": 0.0089, "step": 13871 }, { "epoch": 0.9532710280373832, "grad_norm": 0.2401300072669983, "learning_rate": 1.7146990318284096e-07, "loss": 0.0082, "step": 13872 }, { "epoch": 0.9533397471137988, "grad_norm": 0.25174039602279663, "learning_rate": 1.7096691028105293e-07, "loss": 0.0046, "step": 13873 }, { "epoch": 0.9534084661902144, "grad_norm": 0.3427140712738037, "learning_rate": 1.7046465197907756e-07, "loss": 0.0044, "step": 13874 }, { "epoch": 0.9534771852666301, "grad_norm": 0.5871682167053223, "learning_rate": 1.6996312830179494e-07, "loss": 0.0066, "step": 13875 }, { "epoch": 0.9535459043430456, "grad_norm": 0.1556503176689148, "learning_rate": 1.694623392740502e-07, "loss": 0.0045, "step": 13876 }, { "epoch": 0.9536146234194612, "grad_norm": 0.17530396580696106, "learning_rate": 1.6896228492065014e-07, "loss": 0.014, "step": 13877 }, { "epoch": 0.9536833424958768, "grad_norm": 0.3354606330394745, "learning_rate": 1.6846296526636827e-07, "loss": 0.0146, "step": 13878 }, { "epoch": 0.9537520615722924, "grad_norm": 0.38956019282341003, "learning_rate": 1.6796438033593653e-07, "loss": 0.009, "step": 13879 }, { "epoch": 0.9538207806487081, "grad_norm": 0.36708468198776245, "learning_rate": 1.6746653015405845e-07, "loss": 0.0059, "step": 13880 }, { "epoch": 0.9538894997251237, "grad_norm": 0.1381441205739975, "learning_rate": 1.6696941474539263e-07, "loss": 0.0066, "step": 13881 }, { "epoch": 0.9539582188015393, "grad_norm": 0.29563578963279724, "learning_rate": 1.664730341345677e-07, "loss": 0.0165, "step": 13882 }, { "epoch": 0.9540269378779549, "grad_norm": 0.18560472130775452, "learning_rate": 1.659773883461707e-07, "loss": 0.0071, "step": 13883 }, { "epoch": 0.9540956569543705, "grad_norm": 0.2147994488477707, "learning_rate": 1.6548247740475698e-07, "loss": 0.0148, "step": 13884 }, { "epoch": 0.9541643760307862, "grad_norm": 0.24422886967658997, "learning_rate": 1.6498830133484356e-07, "loss": 0.0052, "step": 13885 }, { "epoch": 0.9542330951072018, "grad_norm": 0.2751796245574951, "learning_rate": 1.6449486016091086e-07, "loss": 0.0064, "step": 13886 }, { "epoch": 0.9543018141836174, "grad_norm": 0.05804269388318062, "learning_rate": 1.6400215390739936e-07, "loss": 0.0035, "step": 13887 }, { "epoch": 0.954370533260033, "grad_norm": 0.3705456852912903, "learning_rate": 1.635101825987212e-07, "loss": 0.0072, "step": 13888 }, { "epoch": 0.9544392523364486, "grad_norm": 0.08123715221881866, "learning_rate": 1.6301894625924685e-07, "loss": 0.0063, "step": 13889 }, { "epoch": 0.9545079714128643, "grad_norm": 0.1870078593492508, "learning_rate": 1.625284449133102e-07, "loss": 0.008, "step": 13890 }, { "epoch": 0.9545766904892798, "grad_norm": 0.23252923786640167, "learning_rate": 1.6203867858521015e-07, "loss": 0.0069, "step": 13891 }, { "epoch": 0.9546454095656954, "grad_norm": 0.4047139883041382, "learning_rate": 1.615496472992073e-07, "loss": 0.0119, "step": 13892 }, { "epoch": 0.954714128642111, "grad_norm": 0.1898462027311325, "learning_rate": 1.610613510795289e-07, "loss": 0.005, "step": 13893 }, { "epoch": 0.9547828477185266, "grad_norm": 0.5349994897842407, "learning_rate": 1.605737899503623e-07, "loss": 0.0126, "step": 13894 }, { "epoch": 0.9548515667949423, "grad_norm": 0.294563353061676, "learning_rate": 1.600869639358632e-07, "loss": 0.007, "step": 13895 }, { "epoch": 0.9549202858713579, "grad_norm": 0.3067221939563751, "learning_rate": 1.596008730601456e-07, "loss": 0.0063, "step": 13896 }, { "epoch": 0.9549890049477735, "grad_norm": 1.0318739414215088, "learning_rate": 1.591155173472919e-07, "loss": 0.0134, "step": 13897 }, { "epoch": 0.9550577240241891, "grad_norm": 0.2385593056678772, "learning_rate": 1.5863089682134125e-07, "loss": 0.0086, "step": 13898 }, { "epoch": 0.9551264431006047, "grad_norm": 0.23671287298202515, "learning_rate": 1.581470115063044e-07, "loss": 0.0136, "step": 13899 }, { "epoch": 0.9551951621770204, "grad_norm": 0.18780352175235748, "learning_rate": 1.576638614261522e-07, "loss": 0.0131, "step": 13900 }, { "epoch": 0.955263881253436, "grad_norm": 0.2253895103931427, "learning_rate": 1.5718144660481714e-07, "loss": 0.0064, "step": 13901 }, { "epoch": 0.9553326003298516, "grad_norm": 0.2656814754009247, "learning_rate": 1.566997670661968e-07, "loss": 0.0035, "step": 13902 }, { "epoch": 0.9554013194062672, "grad_norm": 0.24179084599018097, "learning_rate": 1.5621882283415378e-07, "loss": 0.005, "step": 13903 }, { "epoch": 0.9554700384826827, "grad_norm": 0.2693294286727905, "learning_rate": 1.55738613932514e-07, "loss": 0.011, "step": 13904 }, { "epoch": 0.9555387575590985, "grad_norm": 0.5419481992721558, "learning_rate": 1.5525914038506338e-07, "loss": 0.0089, "step": 13905 }, { "epoch": 0.955607476635514, "grad_norm": 3.167097806930542, "learning_rate": 1.5478040221555635e-07, "loss": 0.0164, "step": 13906 }, { "epoch": 0.9556761957119296, "grad_norm": 0.24305039644241333, "learning_rate": 1.543023994477072e-07, "loss": 0.0087, "step": 13907 }, { "epoch": 0.9557449147883452, "grad_norm": 0.29117119312286377, "learning_rate": 1.5382513210519534e-07, "loss": 0.0098, "step": 13908 }, { "epoch": 0.9558136338647608, "grad_norm": 0.3497585952281952, "learning_rate": 1.5334860021166354e-07, "loss": 0.0092, "step": 13909 }, { "epoch": 0.9558823529411765, "grad_norm": 0.27097052335739136, "learning_rate": 1.528728037907212e-07, "loss": 0.0105, "step": 13910 }, { "epoch": 0.9559510720175921, "grad_norm": 0.1890692412853241, "learning_rate": 1.5239774286593288e-07, "loss": 0.0033, "step": 13911 }, { "epoch": 0.9560197910940077, "grad_norm": 0.2106005698442459, "learning_rate": 1.5192341746083638e-07, "loss": 0.0058, "step": 13912 }, { "epoch": 0.9560885101704233, "grad_norm": 0.45949968695640564, "learning_rate": 1.5144982759892623e-07, "loss": 0.0076, "step": 13913 }, { "epoch": 0.9561572292468389, "grad_norm": 0.17152182757854462, "learning_rate": 1.50976973303667e-07, "loss": 0.009, "step": 13914 }, { "epoch": 0.9562259483232546, "grad_norm": 0.1262931078672409, "learning_rate": 1.5050485459847829e-07, "loss": 0.0051, "step": 13915 }, { "epoch": 0.9562946673996702, "grad_norm": 0.34859874844551086, "learning_rate": 1.5003347150674973e-07, "loss": 0.0123, "step": 13916 }, { "epoch": 0.9563633864760858, "grad_norm": 0.12290313094854355, "learning_rate": 1.4956282405183264e-07, "loss": 0.0041, "step": 13917 }, { "epoch": 0.9564321055525014, "grad_norm": 0.3455813229084015, "learning_rate": 1.4909291225704337e-07, "loss": 0.0069, "step": 13918 }, { "epoch": 0.956500824628917, "grad_norm": 0.23113727569580078, "learning_rate": 1.486237361456566e-07, "loss": 0.0071, "step": 13919 }, { "epoch": 0.9565695437053326, "grad_norm": 0.7009484767913818, "learning_rate": 1.481552957409188e-07, "loss": 0.0075, "step": 13920 }, { "epoch": 0.9566382627817482, "grad_norm": 0.10993922501802444, "learning_rate": 1.47687591066033e-07, "loss": 0.0098, "step": 13921 }, { "epoch": 0.9567069818581638, "grad_norm": 0.3718353509902954, "learning_rate": 1.4722062214417075e-07, "loss": 0.0057, "step": 13922 }, { "epoch": 0.9567757009345794, "grad_norm": 0.3246757686138153, "learning_rate": 1.4675438899846017e-07, "loss": 0.0089, "step": 13923 }, { "epoch": 0.956844420010995, "grad_norm": 0.46739017963409424, "learning_rate": 1.462888916520011e-07, "loss": 0.0054, "step": 13924 }, { "epoch": 0.9569131390874107, "grad_norm": 0.26716580986976624, "learning_rate": 1.4582413012785345e-07, "loss": 0.0058, "step": 13925 }, { "epoch": 0.9569818581638263, "grad_norm": 0.15381062030792236, "learning_rate": 1.4536010444903713e-07, "loss": 0.0127, "step": 13926 }, { "epoch": 0.9570505772402419, "grad_norm": 0.21931034326553345, "learning_rate": 1.4489681463854377e-07, "loss": 0.0068, "step": 13927 }, { "epoch": 0.9571192963166575, "grad_norm": 0.2496665120124817, "learning_rate": 1.4443426071931996e-07, "loss": 0.0061, "step": 13928 }, { "epoch": 0.9571880153930731, "grad_norm": 0.44571465253829956, "learning_rate": 1.4397244271428078e-07, "loss": 0.0106, "step": 13929 }, { "epoch": 0.9572567344694888, "grad_norm": 0.6067138910293579, "learning_rate": 1.4351136064630454e-07, "loss": 0.0104, "step": 13930 }, { "epoch": 0.9573254535459044, "grad_norm": 0.38547882437705994, "learning_rate": 1.43051014538233e-07, "loss": 0.0091, "step": 13931 }, { "epoch": 0.95739417262232, "grad_norm": 0.04413706809282303, "learning_rate": 1.4259140441286789e-07, "loss": 0.0056, "step": 13932 }, { "epoch": 0.9574628916987356, "grad_norm": 0.3233895003795624, "learning_rate": 1.42132530292981e-07, "loss": 0.0054, "step": 13933 }, { "epoch": 0.9575316107751511, "grad_norm": 0.3231221139431, "learning_rate": 1.4167439220130084e-07, "loss": 0.0042, "step": 13934 }, { "epoch": 0.9576003298515668, "grad_norm": 0.25886571407318115, "learning_rate": 1.4121699016052426e-07, "loss": 0.0086, "step": 13935 }, { "epoch": 0.9576690489279824, "grad_norm": 0.5031871199607849, "learning_rate": 1.407603241933081e-07, "loss": 0.0064, "step": 13936 }, { "epoch": 0.957737768004398, "grad_norm": 0.2701965570449829, "learning_rate": 1.4030439432227926e-07, "loss": 0.0095, "step": 13937 }, { "epoch": 0.9578064870808136, "grad_norm": 0.43061429262161255, "learning_rate": 1.3984920057001806e-07, "loss": 0.0154, "step": 13938 }, { "epoch": 0.9578752061572292, "grad_norm": 0.46799489855766296, "learning_rate": 1.393947429590764e-07, "loss": 0.0069, "step": 13939 }, { "epoch": 0.9579439252336449, "grad_norm": 0.29868611693382263, "learning_rate": 1.3894102151196797e-07, "loss": 0.0064, "step": 13940 }, { "epoch": 0.9580126443100605, "grad_norm": 0.3053569197654724, "learning_rate": 1.3848803625116813e-07, "loss": 0.0114, "step": 13941 }, { "epoch": 0.9580813633864761, "grad_norm": 0.40918785333633423, "learning_rate": 1.3803578719911725e-07, "loss": 0.0071, "step": 13942 }, { "epoch": 0.9581500824628917, "grad_norm": 0.09881671518087387, "learning_rate": 1.375842743782174e-07, "loss": 0.0084, "step": 13943 }, { "epoch": 0.9582188015393073, "grad_norm": 0.6071538925170898, "learning_rate": 1.371334978108374e-07, "loss": 0.008, "step": 13944 }, { "epoch": 0.958287520615723, "grad_norm": 0.285526841878891, "learning_rate": 1.3668345751930767e-07, "loss": 0.0065, "step": 13945 }, { "epoch": 0.9583562396921386, "grad_norm": 0.3848590552806854, "learning_rate": 1.3623415352592207e-07, "loss": 0.0088, "step": 13946 }, { "epoch": 0.9584249587685542, "grad_norm": 0.3449151813983917, "learning_rate": 1.357855858529361e-07, "loss": 0.0106, "step": 13947 }, { "epoch": 0.9584936778449697, "grad_norm": 0.12765108048915863, "learning_rate": 1.353377545225737e-07, "loss": 0.0124, "step": 13948 }, { "epoch": 0.9585623969213853, "grad_norm": 0.21554262936115265, "learning_rate": 1.3489065955701873e-07, "loss": 0.0095, "step": 13949 }, { "epoch": 0.958631115997801, "grad_norm": 0.188398078083992, "learning_rate": 1.3444430097841854e-07, "loss": 0.0087, "step": 13950 }, { "epoch": 0.9586998350742166, "grad_norm": 0.23857243359088898, "learning_rate": 1.3399867880888706e-07, "loss": 0.0025, "step": 13951 }, { "epoch": 0.9587685541506322, "grad_norm": 0.22053934633731842, "learning_rate": 1.3355379307049664e-07, "loss": 0.0042, "step": 13952 }, { "epoch": 0.9588372732270478, "grad_norm": 0.15361101925373077, "learning_rate": 1.3310964378528634e-07, "loss": 0.0051, "step": 13953 }, { "epoch": 0.9589059923034634, "grad_norm": 0.3937588930130005, "learning_rate": 1.326662309752602e-07, "loss": 0.0111, "step": 13954 }, { "epoch": 0.9589747113798791, "grad_norm": 0.13849885761737823, "learning_rate": 1.3222355466238234e-07, "loss": 0.0048, "step": 13955 }, { "epoch": 0.9590434304562947, "grad_norm": 0.45901429653167725, "learning_rate": 1.3178161486858354e-07, "loss": 0.0092, "step": 13956 }, { "epoch": 0.9591121495327103, "grad_norm": 0.15415740013122559, "learning_rate": 1.3134041161575462e-07, "loss": 0.0056, "step": 13957 }, { "epoch": 0.9591808686091259, "grad_norm": 0.09336946904659271, "learning_rate": 1.3089994492575476e-07, "loss": 0.0095, "step": 13958 }, { "epoch": 0.9592495876855415, "grad_norm": 0.2227473109960556, "learning_rate": 1.304602148204015e-07, "loss": 0.0109, "step": 13959 }, { "epoch": 0.9593183067619572, "grad_norm": 0.24934181571006775, "learning_rate": 1.3002122132147744e-07, "loss": 0.0074, "step": 13960 }, { "epoch": 0.9593870258383728, "grad_norm": 0.14859318733215332, "learning_rate": 1.295829644507318e-07, "loss": 0.011, "step": 13961 }, { "epoch": 0.9594557449147884, "grad_norm": 0.5114365220069885, "learning_rate": 1.2914544422987395e-07, "loss": 0.0093, "step": 13962 }, { "epoch": 0.959524463991204, "grad_norm": 0.33352118730545044, "learning_rate": 1.287086606805782e-07, "loss": 0.0082, "step": 13963 }, { "epoch": 0.9595931830676195, "grad_norm": 0.4430077075958252, "learning_rate": 1.2827261382448054e-07, "loss": 0.0083, "step": 13964 }, { "epoch": 0.9596619021440352, "grad_norm": 0.28033822774887085, "learning_rate": 1.2783730368318202e-07, "loss": 0.0107, "step": 13965 }, { "epoch": 0.9597306212204508, "grad_norm": 0.5001900792121887, "learning_rate": 1.274027302782471e-07, "loss": 0.012, "step": 13966 }, { "epoch": 0.9597993402968664, "grad_norm": 0.5720798373222351, "learning_rate": 1.269688936312069e-07, "loss": 0.0161, "step": 13967 }, { "epoch": 0.959868059373282, "grad_norm": 0.16782088577747345, "learning_rate": 1.265357937635475e-07, "loss": 0.0049, "step": 13968 }, { "epoch": 0.9599367784496976, "grad_norm": 1.1027342081069946, "learning_rate": 1.2610343069672682e-07, "loss": 0.0092, "step": 13969 }, { "epoch": 0.9600054975261132, "grad_norm": 0.245461568236351, "learning_rate": 1.256718044521643e-07, "loss": 0.0055, "step": 13970 }, { "epoch": 0.9600742166025289, "grad_norm": 0.23422352969646454, "learning_rate": 1.252409150512379e-07, "loss": 0.0103, "step": 13971 }, { "epoch": 0.9601429356789445, "grad_norm": 0.40572649240493774, "learning_rate": 1.2481076251529554e-07, "loss": 0.007, "step": 13972 }, { "epoch": 0.9602116547553601, "grad_norm": 0.19169150292873383, "learning_rate": 1.243813468656485e-07, "loss": 0.0131, "step": 13973 }, { "epoch": 0.9602803738317757, "grad_norm": 0.3490772247314453, "learning_rate": 1.2395266812356475e-07, "loss": 0.0112, "step": 13974 }, { "epoch": 0.9603490929081913, "grad_norm": 0.3326198160648346, "learning_rate": 1.2352472631028233e-07, "loss": 0.0071, "step": 13975 }, { "epoch": 0.960417811984607, "grad_norm": 0.1285531371831894, "learning_rate": 1.2309752144699926e-07, "loss": 0.003, "step": 13976 }, { "epoch": 0.9604865310610226, "grad_norm": 0.4037383496761322, "learning_rate": 1.2267105355488028e-07, "loss": 0.0106, "step": 13977 }, { "epoch": 0.9605552501374381, "grad_norm": 0.4483399987220764, "learning_rate": 1.2224532265505184e-07, "loss": 0.0089, "step": 13978 }, { "epoch": 0.9606239692138537, "grad_norm": 0.6996666789054871, "learning_rate": 1.2182032876860038e-07, "loss": 0.0099, "step": 13979 }, { "epoch": 0.9606926882902693, "grad_norm": 0.3777105510234833, "learning_rate": 1.213960719165841e-07, "loss": 0.0117, "step": 13980 }, { "epoch": 0.960761407366685, "grad_norm": 0.2386520355939865, "learning_rate": 1.2097255212001612e-07, "loss": 0.0067, "step": 13981 }, { "epoch": 0.9608301264431006, "grad_norm": 0.09409263730049133, "learning_rate": 1.2054976939987805e-07, "loss": 0.0065, "step": 13982 }, { "epoch": 0.9608988455195162, "grad_norm": 0.5413909554481506, "learning_rate": 1.2012772377711147e-07, "loss": 0.0095, "step": 13983 }, { "epoch": 0.9609675645959318, "grad_norm": 0.47173115611076355, "learning_rate": 1.1970641527262793e-07, "loss": 0.008, "step": 13984 }, { "epoch": 0.9610362836723474, "grad_norm": 0.06121854484081268, "learning_rate": 1.192858439072958e-07, "loss": 0.0046, "step": 13985 }, { "epoch": 0.9611050027487631, "grad_norm": 0.3119690418243408, "learning_rate": 1.1886600970195005e-07, "loss": 0.01, "step": 13986 }, { "epoch": 0.9611737218251787, "grad_norm": 0.18006131052970886, "learning_rate": 1.1844691267738739e-07, "loss": 0.0106, "step": 13987 }, { "epoch": 0.9612424409015943, "grad_norm": 0.26119837164878845, "learning_rate": 1.1802855285436953e-07, "loss": 0.0071, "step": 13988 }, { "epoch": 0.9613111599780099, "grad_norm": 0.5493496656417847, "learning_rate": 1.176109302536199e-07, "loss": 0.0078, "step": 13989 }, { "epoch": 0.9613798790544255, "grad_norm": 0.18152232468128204, "learning_rate": 1.1719404489582863e-07, "loss": 0.0122, "step": 13990 }, { "epoch": 0.9614485981308412, "grad_norm": 0.5094819664955139, "learning_rate": 1.1677789680164753e-07, "loss": 0.0173, "step": 13991 }, { "epoch": 0.9615173172072567, "grad_norm": 0.14440399408340454, "learning_rate": 1.1636248599169009e-07, "loss": 0.0071, "step": 13992 }, { "epoch": 0.9615860362836723, "grad_norm": 0.2571380138397217, "learning_rate": 1.1594781248653486e-07, "loss": 0.0136, "step": 13993 }, { "epoch": 0.9616547553600879, "grad_norm": 0.3306888937950134, "learning_rate": 1.1553387630672708e-07, "loss": 0.007, "step": 13994 }, { "epoch": 0.9617234744365035, "grad_norm": 0.34904608130455017, "learning_rate": 1.15120677472767e-07, "loss": 0.0128, "step": 13995 }, { "epoch": 0.9617921935129192, "grad_norm": 0.2709844708442688, "learning_rate": 1.1470821600512827e-07, "loss": 0.0088, "step": 13996 }, { "epoch": 0.9618609125893348, "grad_norm": 0.2853911519050598, "learning_rate": 1.1429649192424119e-07, "loss": 0.0053, "step": 13997 }, { "epoch": 0.9619296316657504, "grad_norm": 0.3786919414997101, "learning_rate": 1.1388550525050279e-07, "loss": 0.0072, "step": 13998 }, { "epoch": 0.961998350742166, "grad_norm": 0.4087586998939514, "learning_rate": 1.1347525600427178e-07, "loss": 0.0074, "step": 13999 }, { "epoch": 0.9620670698185816, "grad_norm": 0.4201715290546417, "learning_rate": 1.1306574420587023e-07, "loss": 0.008, "step": 14000 }, { "epoch": 0.9621357888949973, "grad_norm": 0.1901262104511261, "learning_rate": 1.1265696987558527e-07, "loss": 0.0075, "step": 14001 }, { "epoch": 0.9622045079714129, "grad_norm": 0.4025326669216156, "learning_rate": 1.1224893303366735e-07, "loss": 0.0073, "step": 14002 }, { "epoch": 0.9622732270478285, "grad_norm": 0.646705687046051, "learning_rate": 1.1184163370033029e-07, "loss": 0.0079, "step": 14003 }, { "epoch": 0.9623419461242441, "grad_norm": 0.3420480489730835, "learning_rate": 1.1143507189574963e-07, "loss": 0.0093, "step": 14004 }, { "epoch": 0.9624106652006597, "grad_norm": 0.6193325519561768, "learning_rate": 1.1102924764006594e-07, "loss": 0.0106, "step": 14005 }, { "epoch": 0.9624793842770754, "grad_norm": 0.11122245341539383, "learning_rate": 1.1062416095338312e-07, "loss": 0.0075, "step": 14006 }, { "epoch": 0.962548103353491, "grad_norm": 0.3527015149593353, "learning_rate": 1.1021981185576679e-07, "loss": 0.0097, "step": 14007 }, { "epoch": 0.9626168224299065, "grad_norm": 0.6263560652732849, "learning_rate": 1.0981620036724926e-07, "loss": 0.0127, "step": 14008 }, { "epoch": 0.9626855415063221, "grad_norm": 0.42106837034225464, "learning_rate": 1.0941332650782454e-07, "loss": 0.0054, "step": 14009 }, { "epoch": 0.9627542605827377, "grad_norm": 0.2300490289926529, "learning_rate": 1.0901119029744999e-07, "loss": 0.0154, "step": 14010 }, { "epoch": 0.9628229796591534, "grad_norm": 0.031118309125304222, "learning_rate": 1.0860979175604802e-07, "loss": 0.0081, "step": 14011 }, { "epoch": 0.962891698735569, "grad_norm": 0.3768966495990753, "learning_rate": 1.0820913090349937e-07, "loss": 0.0087, "step": 14012 }, { "epoch": 0.9629604178119846, "grad_norm": 0.16947025060653687, "learning_rate": 1.0780920775965486e-07, "loss": 0.0078, "step": 14013 }, { "epoch": 0.9630291368884002, "grad_norm": 0.10575324296951294, "learning_rate": 1.0741002234432529e-07, "loss": 0.0043, "step": 14014 }, { "epoch": 0.9630978559648158, "grad_norm": 0.5261629223823547, "learning_rate": 1.0701157467728484e-07, "loss": 0.0087, "step": 14015 }, { "epoch": 0.9631665750412315, "grad_norm": 0.48204272985458374, "learning_rate": 1.0661386477827273e-07, "loss": 0.0113, "step": 14016 }, { "epoch": 0.9632352941176471, "grad_norm": 0.1822884976863861, "learning_rate": 1.0621689266698986e-07, "loss": 0.0048, "step": 14017 }, { "epoch": 0.9633040131940627, "grad_norm": 0.36292558908462524, "learning_rate": 1.0582065836310384e-07, "loss": 0.0075, "step": 14018 }, { "epoch": 0.9633727322704783, "grad_norm": 0.4610322117805481, "learning_rate": 1.0542516188623729e-07, "loss": 0.0098, "step": 14019 }, { "epoch": 0.9634414513468939, "grad_norm": 0.6885429620742798, "learning_rate": 1.0503040325598956e-07, "loss": 0.015, "step": 14020 }, { "epoch": 0.9635101704233096, "grad_norm": 0.412722647190094, "learning_rate": 1.0463638249191165e-07, "loss": 0.0082, "step": 14021 }, { "epoch": 0.9635788894997251, "grad_norm": 0.25986018776893616, "learning_rate": 1.0424309961352463e-07, "loss": 0.0072, "step": 14022 }, { "epoch": 0.9636476085761407, "grad_norm": 0.480996698141098, "learning_rate": 1.0385055464030957e-07, "loss": 0.0095, "step": 14023 }, { "epoch": 0.9637163276525563, "grad_norm": 0.5504000782966614, "learning_rate": 1.0345874759171258e-07, "loss": 0.0115, "step": 14024 }, { "epoch": 0.9637850467289719, "grad_norm": 0.3764633536338806, "learning_rate": 1.0306767848714316e-07, "loss": 0.0051, "step": 14025 }, { "epoch": 0.9638537658053876, "grad_norm": 0.36829572916030884, "learning_rate": 1.0267734734597412e-07, "loss": 0.0091, "step": 14026 }, { "epoch": 0.9639224848818032, "grad_norm": 0.6146916747093201, "learning_rate": 1.0228775418754166e-07, "loss": 0.009, "step": 14027 }, { "epoch": 0.9639912039582188, "grad_norm": 0.06840222328901291, "learning_rate": 1.0189889903114703e-07, "loss": 0.0051, "step": 14028 }, { "epoch": 0.9640599230346344, "grad_norm": 0.34740039706230164, "learning_rate": 1.0151078189604979e-07, "loss": 0.0068, "step": 14029 }, { "epoch": 0.96412864211105, "grad_norm": 0.130975604057312, "learning_rate": 1.0112340280147958e-07, "loss": 0.0059, "step": 14030 }, { "epoch": 0.9641973611874657, "grad_norm": 0.22072811424732208, "learning_rate": 1.0073676176662439e-07, "loss": 0.007, "step": 14031 }, { "epoch": 0.9642660802638813, "grad_norm": 0.3886997103691101, "learning_rate": 1.0035085881063888e-07, "loss": 0.0093, "step": 14032 }, { "epoch": 0.9643347993402969, "grad_norm": 0.2398255169391632, "learning_rate": 9.996569395263943e-08, "loss": 0.0062, "step": 14033 }, { "epoch": 0.9644035184167125, "grad_norm": 0.12830573320388794, "learning_rate": 9.95812672117058e-08, "loss": 0.0055, "step": 14034 }, { "epoch": 0.964472237493128, "grad_norm": 0.25737398862838745, "learning_rate": 9.919757860688273e-08, "loss": 0.0096, "step": 14035 }, { "epoch": 0.9645409565695438, "grad_norm": 0.13273264467716217, "learning_rate": 9.881462815717668e-08, "loss": 0.0082, "step": 14036 }, { "epoch": 0.9646096756459593, "grad_norm": 0.41866207122802734, "learning_rate": 9.84324158815575e-08, "loss": 0.0058, "step": 14037 }, { "epoch": 0.9646783947223749, "grad_norm": 0.15354950726032257, "learning_rate": 9.805094179896168e-08, "loss": 0.0059, "step": 14038 }, { "epoch": 0.9647471137987905, "grad_norm": 0.35722675919532776, "learning_rate": 9.767020592828579e-08, "loss": 0.0112, "step": 14039 }, { "epoch": 0.9648158328752061, "grad_norm": 0.6436026096343994, "learning_rate": 9.729020828838808e-08, "loss": 0.009, "step": 14040 }, { "epoch": 0.9648845519516218, "grad_norm": 0.31038615107536316, "learning_rate": 9.691094889809515e-08, "loss": 0.0056, "step": 14041 }, { "epoch": 0.9649532710280374, "grad_norm": 0.6646554470062256, "learning_rate": 9.653242777619531e-08, "loss": 0.0103, "step": 14042 }, { "epoch": 0.965021990104453, "grad_norm": 0.4042338728904724, "learning_rate": 9.615464494143688e-08, "loss": 0.0041, "step": 14043 }, { "epoch": 0.9650907091808686, "grad_norm": 0.5288040637969971, "learning_rate": 9.577760041253824e-08, "loss": 0.0198, "step": 14044 }, { "epoch": 0.9651594282572842, "grad_norm": 0.6000759601593018, "learning_rate": 9.540129420817446e-08, "loss": 0.0046, "step": 14045 }, { "epoch": 0.9652281473336999, "grad_norm": 0.3197172284126282, "learning_rate": 9.502572634698892e-08, "loss": 0.0053, "step": 14046 }, { "epoch": 0.9652968664101155, "grad_norm": 0.11601202934980392, "learning_rate": 9.46508968475851e-08, "loss": 0.004, "step": 14047 }, { "epoch": 0.9653655854865311, "grad_norm": 0.4665939211845398, "learning_rate": 9.427680572853147e-08, "loss": 0.008, "step": 14048 }, { "epoch": 0.9654343045629467, "grad_norm": 0.22476623952388763, "learning_rate": 9.390345300835989e-08, "loss": 0.0057, "step": 14049 }, { "epoch": 0.9655030236393622, "grad_norm": 0.07359462976455688, "learning_rate": 9.353083870556722e-08, "loss": 0.0073, "step": 14050 }, { "epoch": 0.965571742715778, "grad_norm": 0.2855078876018524, "learning_rate": 9.315896283860869e-08, "loss": 0.0072, "step": 14051 }, { "epoch": 0.9656404617921935, "grad_norm": 0.39500415325164795, "learning_rate": 9.278782542590791e-08, "loss": 0.0125, "step": 14052 }, { "epoch": 0.9657091808686091, "grad_norm": 0.16740112006664276, "learning_rate": 9.241742648585183e-08, "loss": 0.0043, "step": 14053 }, { "epoch": 0.9657778999450247, "grad_norm": 0.18535900115966797, "learning_rate": 9.204776603678577e-08, "loss": 0.0044, "step": 14054 }, { "epoch": 0.9658466190214403, "grad_norm": 0.4998791217803955, "learning_rate": 9.16788440970251e-08, "loss": 0.0127, "step": 14055 }, { "epoch": 0.965915338097856, "grad_norm": 0.33617839217185974, "learning_rate": 9.13106606848435e-08, "loss": 0.0122, "step": 14056 }, { "epoch": 0.9659840571742716, "grad_norm": 0.5568981170654297, "learning_rate": 9.094321581848142e-08, "loss": 0.0114, "step": 14057 }, { "epoch": 0.9660527762506872, "grad_norm": 0.798211395740509, "learning_rate": 9.057650951614094e-08, "loss": 0.012, "step": 14058 }, { "epoch": 0.9661214953271028, "grad_norm": 0.2893957197666168, "learning_rate": 9.021054179598753e-08, "loss": 0.0055, "step": 14059 }, { "epoch": 0.9661902144035184, "grad_norm": 0.659067690372467, "learning_rate": 8.98453126761517e-08, "loss": 0.0079, "step": 14060 }, { "epoch": 0.9662589334799341, "grad_norm": 0.24627414345741272, "learning_rate": 8.948082217472231e-08, "loss": 0.012, "step": 14061 }, { "epoch": 0.9663276525563497, "grad_norm": 0.7441434264183044, "learning_rate": 8.91170703097599e-08, "loss": 0.0139, "step": 14062 }, { "epoch": 0.9663963716327653, "grad_norm": 0.2656635344028473, "learning_rate": 8.875405709928342e-08, "loss": 0.0077, "step": 14063 }, { "epoch": 0.9664650907091809, "grad_norm": 0.3447295129299164, "learning_rate": 8.839178256127344e-08, "loss": 0.0054, "step": 14064 }, { "epoch": 0.9665338097855964, "grad_norm": 0.05025707557797432, "learning_rate": 8.803024671367732e-08, "loss": 0.0047, "step": 14065 }, { "epoch": 0.9666025288620121, "grad_norm": 0.34936416149139404, "learning_rate": 8.766944957440571e-08, "loss": 0.0103, "step": 14066 }, { "epoch": 0.9666712479384277, "grad_norm": 0.1963787078857422, "learning_rate": 8.730939116133097e-08, "loss": 0.0054, "step": 14067 }, { "epoch": 0.9667399670148433, "grad_norm": 0.33910828828811646, "learning_rate": 8.695007149229051e-08, "loss": 0.0141, "step": 14068 }, { "epoch": 0.9668086860912589, "grad_norm": 0.10830020904541016, "learning_rate": 8.659149058508342e-08, "loss": 0.0043, "step": 14069 }, { "epoch": 0.9668774051676745, "grad_norm": 0.9385190010070801, "learning_rate": 8.623364845747217e-08, "loss": 0.0115, "step": 14070 }, { "epoch": 0.9669461242440902, "grad_norm": 0.538861870765686, "learning_rate": 8.58765451271859e-08, "loss": 0.0132, "step": 14071 }, { "epoch": 0.9670148433205058, "grad_norm": 0.3085494339466095, "learning_rate": 8.552018061191213e-08, "loss": 0.0276, "step": 14072 }, { "epoch": 0.9670835623969214, "grad_norm": 1.4863426685333252, "learning_rate": 8.516455492930508e-08, "loss": 0.0047, "step": 14073 }, { "epoch": 0.967152281473337, "grad_norm": 0.49683448672294617, "learning_rate": 8.480966809698398e-08, "loss": 0.0065, "step": 14074 }, { "epoch": 0.9672210005497526, "grad_norm": 0.09776987880468369, "learning_rate": 8.445552013252645e-08, "loss": 0.0035, "step": 14075 }, { "epoch": 0.9672897196261683, "grad_norm": 0.5504260659217834, "learning_rate": 8.410211105347843e-08, "loss": 0.0102, "step": 14076 }, { "epoch": 0.9673584387025839, "grad_norm": 0.40250763297080994, "learning_rate": 8.374944087734593e-08, "loss": 0.0054, "step": 14077 }, { "epoch": 0.9674271577789995, "grad_norm": 0.4453107714653015, "learning_rate": 8.339750962159998e-08, "loss": 0.0068, "step": 14078 }, { "epoch": 0.967495876855415, "grad_norm": 0.3779599368572235, "learning_rate": 8.304631730367329e-08, "loss": 0.0069, "step": 14079 }, { "epoch": 0.9675645959318306, "grad_norm": 0.25858649611473083, "learning_rate": 8.269586394096363e-08, "loss": 0.0061, "step": 14080 }, { "epoch": 0.9676333150082463, "grad_norm": 0.41305649280548096, "learning_rate": 8.234614955083374e-08, "loss": 0.0082, "step": 14081 }, { "epoch": 0.9677020340846619, "grad_norm": 0.33204975724220276, "learning_rate": 8.199717415060481e-08, "loss": 0.0069, "step": 14082 }, { "epoch": 0.9677707531610775, "grad_norm": 0.4557028114795685, "learning_rate": 8.164893775756632e-08, "loss": 0.0068, "step": 14083 }, { "epoch": 0.9678394722374931, "grad_norm": 0.1153300553560257, "learning_rate": 8.130144038896947e-08, "loss": 0.0088, "step": 14084 }, { "epoch": 0.9679081913139087, "grad_norm": 0.40031638741493225, "learning_rate": 8.09546820620255e-08, "loss": 0.0089, "step": 14085 }, { "epoch": 0.9679769103903244, "grad_norm": 0.12963570654392242, "learning_rate": 8.060866279391732e-08, "loss": 0.004, "step": 14086 }, { "epoch": 0.96804562946674, "grad_norm": 0.334527850151062, "learning_rate": 8.026338260178123e-08, "loss": 0.0049, "step": 14087 }, { "epoch": 0.9681143485431556, "grad_norm": 0.011484880931675434, "learning_rate": 7.991884150272355e-08, "loss": 0.0033, "step": 14088 }, { "epoch": 0.9681830676195712, "grad_norm": 0.14947445690631866, "learning_rate": 7.957503951381228e-08, "loss": 0.0063, "step": 14089 }, { "epoch": 0.9682517866959868, "grad_norm": 0.18837101757526398, "learning_rate": 7.923197665207882e-08, "loss": 0.0051, "step": 14090 }, { "epoch": 0.9683205057724025, "grad_norm": 0.23924843966960907, "learning_rate": 7.888965293451623e-08, "loss": 0.0061, "step": 14091 }, { "epoch": 0.9683892248488181, "grad_norm": 0.43556198477745056, "learning_rate": 7.854806837808426e-08, "loss": 0.0079, "step": 14092 }, { "epoch": 0.9684579439252337, "grad_norm": 0.2496856302022934, "learning_rate": 7.82072229997044e-08, "loss": 0.0127, "step": 14093 }, { "epoch": 0.9685266630016492, "grad_norm": 0.5713144540786743, "learning_rate": 7.786711681625979e-08, "loss": 0.0088, "step": 14094 }, { "epoch": 0.9685953820780648, "grad_norm": 0.9495430588722229, "learning_rate": 7.752774984460032e-08, "loss": 0.0116, "step": 14095 }, { "epoch": 0.9686641011544805, "grad_norm": 0.2205791175365448, "learning_rate": 7.71891221015375e-08, "loss": 0.007, "step": 14096 }, { "epoch": 0.9687328202308961, "grad_norm": 0.3586415946483612, "learning_rate": 7.685123360384293e-08, "loss": 0.0072, "step": 14097 }, { "epoch": 0.9688015393073117, "grad_norm": 0.4687873423099518, "learning_rate": 7.65140843682599e-08, "loss": 0.0105, "step": 14098 }, { "epoch": 0.9688702583837273, "grad_norm": 0.17323394119739532, "learning_rate": 7.617767441148837e-08, "loss": 0.0072, "step": 14099 }, { "epoch": 0.9689389774601429, "grad_norm": 0.31237295269966125, "learning_rate": 7.584200375019168e-08, "loss": 0.0065, "step": 14100 }, { "epoch": 0.9690076965365585, "grad_norm": 0.21979813277721405, "learning_rate": 7.55070724009982e-08, "loss": 0.0114, "step": 14101 }, { "epoch": 0.9690764156129742, "grad_norm": 0.463932067155838, "learning_rate": 7.517288038050296e-08, "loss": 0.0079, "step": 14102 }, { "epoch": 0.9691451346893898, "grad_norm": 0.3017047047615051, "learning_rate": 7.483942770525776e-08, "loss": 0.0079, "step": 14103 }, { "epoch": 0.9692138537658054, "grad_norm": 0.2775270640850067, "learning_rate": 7.450671439178269e-08, "loss": 0.0063, "step": 14104 }, { "epoch": 0.969282572842221, "grad_norm": 0.4053260087966919, "learning_rate": 7.417474045655958e-08, "loss": 0.013, "step": 14105 }, { "epoch": 0.9693512919186366, "grad_norm": 0.652066707611084, "learning_rate": 7.384350591603361e-08, "loss": 0.0084, "step": 14106 }, { "epoch": 0.9694200109950523, "grad_norm": 0.26130637526512146, "learning_rate": 7.351301078661499e-08, "loss": 0.0057, "step": 14107 }, { "epoch": 0.9694887300714679, "grad_norm": 0.3009113371372223, "learning_rate": 7.318325508467227e-08, "loss": 0.0071, "step": 14108 }, { "epoch": 0.9695574491478834, "grad_norm": 0.3890013098716736, "learning_rate": 7.285423882654407e-08, "loss": 0.0154, "step": 14109 }, { "epoch": 0.969626168224299, "grad_norm": 0.1480707824230194, "learning_rate": 7.252596202852568e-08, "loss": 0.0079, "step": 14110 }, { "epoch": 0.9696948873007146, "grad_norm": 0.09043846279382706, "learning_rate": 7.219842470688409e-08, "loss": 0.0033, "step": 14111 }, { "epoch": 0.9697636063771303, "grad_norm": 0.2361191213130951, "learning_rate": 7.187162687784133e-08, "loss": 0.0057, "step": 14112 }, { "epoch": 0.9698323254535459, "grad_norm": 0.0464053750038147, "learning_rate": 7.154556855758777e-08, "loss": 0.0138, "step": 14113 }, { "epoch": 0.9699010445299615, "grad_norm": 0.23887746036052704, "learning_rate": 7.122024976227382e-08, "loss": 0.0071, "step": 14114 }, { "epoch": 0.9699697636063771, "grad_norm": 0.323075532913208, "learning_rate": 7.089567050801826e-08, "loss": 0.0079, "step": 14115 }, { "epoch": 0.9700384826827927, "grad_norm": 0.9052006602287292, "learning_rate": 7.057183081089657e-08, "loss": 0.0095, "step": 14116 }, { "epoch": 0.9701072017592084, "grad_norm": 0.4985120892524719, "learning_rate": 7.024873068695425e-08, "loss": 0.0051, "step": 14117 }, { "epoch": 0.970175920835624, "grad_norm": 0.12118110805749893, "learning_rate": 6.99263701521935e-08, "loss": 0.013, "step": 14118 }, { "epoch": 0.9702446399120396, "grad_norm": 0.4241660237312317, "learning_rate": 6.96047492225882e-08, "loss": 0.0062, "step": 14119 }, { "epoch": 0.9703133589884552, "grad_norm": 0.3803121745586395, "learning_rate": 6.928386791406561e-08, "loss": 0.0059, "step": 14120 }, { "epoch": 0.9703820780648708, "grad_norm": 0.23145781457424164, "learning_rate": 6.896372624252633e-08, "loss": 0.0079, "step": 14121 }, { "epoch": 0.9704507971412865, "grad_norm": 0.11307032406330109, "learning_rate": 6.864432422382438e-08, "loss": 0.0117, "step": 14122 }, { "epoch": 0.970519516217702, "grad_norm": 0.3610993027687073, "learning_rate": 6.832566187378708e-08, "loss": 0.0134, "step": 14123 }, { "epoch": 0.9705882352941176, "grad_norm": 0.243340402841568, "learning_rate": 6.800773920819681e-08, "loss": 0.0063, "step": 14124 }, { "epoch": 0.9706569543705332, "grad_norm": 0.3356097638607025, "learning_rate": 6.769055624280596e-08, "loss": 0.0083, "step": 14125 }, { "epoch": 0.9707256734469488, "grad_norm": 0.1911696046590805, "learning_rate": 6.737411299332364e-08, "loss": 0.0069, "step": 14126 }, { "epoch": 0.9707943925233645, "grad_norm": 0.35756081342697144, "learning_rate": 6.705840947542896e-08, "loss": 0.0048, "step": 14127 }, { "epoch": 0.9708631115997801, "grad_norm": 0.12655633687973022, "learning_rate": 6.674344570475943e-08, "loss": 0.0143, "step": 14128 }, { "epoch": 0.9709318306761957, "grad_norm": 0.4617542028427124, "learning_rate": 6.642922169691756e-08, "loss": 0.0135, "step": 14129 }, { "epoch": 0.9710005497526113, "grad_norm": 0.3762321174144745, "learning_rate": 6.61157374674709e-08, "loss": 0.0076, "step": 14130 }, { "epoch": 0.9710692688290269, "grad_norm": 0.9306253790855408, "learning_rate": 6.580299303194704e-08, "loss": 0.0097, "step": 14131 }, { "epoch": 0.9711379879054426, "grad_norm": 0.5664963722229004, "learning_rate": 6.549098840584023e-08, "loss": 0.0137, "step": 14132 }, { "epoch": 0.9712067069818582, "grad_norm": 0.3715532720088959, "learning_rate": 6.517972360460478e-08, "loss": 0.0073, "step": 14133 }, { "epoch": 0.9712754260582738, "grad_norm": 0.31133773922920227, "learning_rate": 6.48691986436617e-08, "loss": 0.0061, "step": 14134 }, { "epoch": 0.9713441451346894, "grad_norm": 0.2640293538570404, "learning_rate": 6.455941353839201e-08, "loss": 0.0071, "step": 14135 }, { "epoch": 0.971412864211105, "grad_norm": 0.1131751760840416, "learning_rate": 6.425036830414511e-08, "loss": 0.0075, "step": 14136 }, { "epoch": 0.9714815832875207, "grad_norm": 0.14586521685123444, "learning_rate": 6.394206295622707e-08, "loss": 0.005, "step": 14137 }, { "epoch": 0.9715503023639362, "grad_norm": 0.2880341410636902, "learning_rate": 6.363449750991235e-08, "loss": 0.0091, "step": 14138 }, { "epoch": 0.9716190214403518, "grad_norm": 0.22984038293361664, "learning_rate": 6.33276719804371e-08, "loss": 0.0074, "step": 14139 }, { "epoch": 0.9716877405167674, "grad_norm": 0.5808408260345459, "learning_rate": 6.302158638300082e-08, "loss": 0.0079, "step": 14140 }, { "epoch": 0.971756459593183, "grad_norm": 0.31665319204330444, "learning_rate": 6.271624073276472e-08, "loss": 0.0082, "step": 14141 }, { "epoch": 0.9718251786695987, "grad_norm": 0.299969345331192, "learning_rate": 6.24116350448567e-08, "loss": 0.005, "step": 14142 }, { "epoch": 0.9718938977460143, "grad_norm": 0.8151790499687195, "learning_rate": 6.210776933436635e-08, "loss": 0.008, "step": 14143 }, { "epoch": 0.9719626168224299, "grad_norm": 0.255341112613678, "learning_rate": 6.180464361634664e-08, "loss": 0.0104, "step": 14144 }, { "epoch": 0.9720313358988455, "grad_norm": 0.4486164450645447, "learning_rate": 6.150225790581221e-08, "loss": 0.0092, "step": 14145 }, { "epoch": 0.9721000549752611, "grad_norm": 0.23171895742416382, "learning_rate": 6.120061221774275e-08, "loss": 0.006, "step": 14146 }, { "epoch": 0.9721687740516768, "grad_norm": 0.2648046314716339, "learning_rate": 6.089970656708466e-08, "loss": 0.0056, "step": 14147 }, { "epoch": 0.9722374931280924, "grad_norm": 0.47809290885925293, "learning_rate": 6.059954096873933e-08, "loss": 0.0105, "step": 14148 }, { "epoch": 0.972306212204508, "grad_norm": 0.4723505973815918, "learning_rate": 6.030011543757985e-08, "loss": 0.0136, "step": 14149 }, { "epoch": 0.9723749312809236, "grad_norm": 0.22581122815608978, "learning_rate": 6.000142998843772e-08, "loss": 0.0041, "step": 14150 }, { "epoch": 0.9724436503573392, "grad_norm": 0.27108657360076904, "learning_rate": 5.970348463610942e-08, "loss": 0.0049, "step": 14151 }, { "epoch": 0.9725123694337549, "grad_norm": 0.5139565467834473, "learning_rate": 5.9406279395354814e-08, "loss": 0.0107, "step": 14152 }, { "epoch": 0.9725810885101704, "grad_norm": 0.17440439760684967, "learning_rate": 5.910981428089546e-08, "loss": 0.0124, "step": 14153 }, { "epoch": 0.972649807586586, "grad_norm": 0.5187050700187683, "learning_rate": 5.881408930741961e-08, "loss": 0.0045, "step": 14154 }, { "epoch": 0.9727185266630016, "grad_norm": 0.3939760625362396, "learning_rate": 5.8519104489577205e-08, "loss": 0.0097, "step": 14155 }, { "epoch": 0.9727872457394172, "grad_norm": 0.6235361099243164, "learning_rate": 5.822485984197823e-08, "loss": 0.0096, "step": 14156 }, { "epoch": 0.9728559648158329, "grad_norm": 0.32816410064697266, "learning_rate": 5.793135537920102e-08, "loss": 0.0052, "step": 14157 }, { "epoch": 0.9729246838922485, "grad_norm": 0.08139362931251526, "learning_rate": 5.7638591115783954e-08, "loss": 0.0117, "step": 14158 }, { "epoch": 0.9729934029686641, "grad_norm": 0.16885386407375336, "learning_rate": 5.734656706623209e-08, "loss": 0.0064, "step": 14159 }, { "epoch": 0.9730621220450797, "grad_norm": 0.3636701703071594, "learning_rate": 5.7055283245010525e-08, "loss": 0.0047, "step": 14160 }, { "epoch": 0.9731308411214953, "grad_norm": 0.2605801224708557, "learning_rate": 5.676473966654938e-08, "loss": 0.0084, "step": 14161 }, { "epoch": 0.973199560197911, "grad_norm": 0.5148698687553406, "learning_rate": 5.647493634524048e-08, "loss": 0.0085, "step": 14162 }, { "epoch": 0.9732682792743266, "grad_norm": 0.4045543670654297, "learning_rate": 5.618587329544067e-08, "loss": 0.0102, "step": 14163 }, { "epoch": 0.9733369983507422, "grad_norm": 0.38250812888145447, "learning_rate": 5.58975505314685e-08, "loss": 0.0073, "step": 14164 }, { "epoch": 0.9734057174271578, "grad_norm": 0.4491306245326996, "learning_rate": 5.5609968067609206e-08, "loss": 0.0097, "step": 14165 }, { "epoch": 0.9734744365035733, "grad_norm": 0.4496057331562042, "learning_rate": 5.53231259181064e-08, "loss": 0.0051, "step": 14166 }, { "epoch": 0.973543155579989, "grad_norm": 0.254846453666687, "learning_rate": 5.503702409717204e-08, "loss": 0.0193, "step": 14167 }, { "epoch": 0.9736118746564046, "grad_norm": 0.31179162859916687, "learning_rate": 5.4751662618978125e-08, "loss": 0.0072, "step": 14168 }, { "epoch": 0.9736805937328202, "grad_norm": 0.4002635180950165, "learning_rate": 5.446704149766002e-08, "loss": 0.0071, "step": 14169 }, { "epoch": 0.9737493128092358, "grad_norm": 0.5022642612457275, "learning_rate": 5.4183160747318105e-08, "loss": 0.0074, "step": 14170 }, { "epoch": 0.9738180318856514, "grad_norm": 0.3004939556121826, "learning_rate": 5.3900020382016133e-08, "loss": 0.0054, "step": 14171 }, { "epoch": 0.9738867509620671, "grad_norm": 0.20171506702899933, "learning_rate": 5.361762041577789e-08, "loss": 0.0065, "step": 14172 }, { "epoch": 0.9739554700384827, "grad_norm": 0.32323211431503296, "learning_rate": 5.333596086259551e-08, "loss": 0.0037, "step": 14173 }, { "epoch": 0.9740241891148983, "grad_norm": 0.285293847322464, "learning_rate": 5.305504173641951e-08, "loss": 0.0062, "step": 14174 }, { "epoch": 0.9740929081913139, "grad_norm": 0.6249011754989624, "learning_rate": 5.277486305116708e-08, "loss": 0.0096, "step": 14175 }, { "epoch": 0.9741616272677295, "grad_norm": 0.2575412690639496, "learning_rate": 5.249542482071879e-08, "loss": 0.0099, "step": 14176 }, { "epoch": 0.9742303463441452, "grad_norm": 0.5738356113433838, "learning_rate": 5.2216727058915224e-08, "loss": 0.0073, "step": 14177 }, { "epoch": 0.9742990654205608, "grad_norm": 0.18561352789402008, "learning_rate": 5.193876977956535e-08, "loss": 0.0109, "step": 14178 }, { "epoch": 0.9743677844969764, "grad_norm": 0.6711993217468262, "learning_rate": 5.166155299643482e-08, "loss": 0.0079, "step": 14179 }, { "epoch": 0.974436503573392, "grad_norm": 0.19905605912208557, "learning_rate": 5.1385076723259316e-08, "loss": 0.0058, "step": 14180 }, { "epoch": 0.9745052226498075, "grad_norm": 0.24219217896461487, "learning_rate": 5.1109340973734545e-08, "loss": 0.0104, "step": 14181 }, { "epoch": 0.9745739417262232, "grad_norm": 0.1696764975786209, "learning_rate": 5.0834345761519576e-08, "loss": 0.0164, "step": 14182 }, { "epoch": 0.9746426608026388, "grad_norm": 0.11349941045045853, "learning_rate": 5.056009110023518e-08, "loss": 0.0054, "step": 14183 }, { "epoch": 0.9747113798790544, "grad_norm": 0.2311321496963501, "learning_rate": 5.02865770034705e-08, "loss": 0.005, "step": 14184 }, { "epoch": 0.97478009895547, "grad_norm": 0.3520490825176239, "learning_rate": 5.001380348477469e-08, "loss": 0.0079, "step": 14185 }, { "epoch": 0.9748488180318856, "grad_norm": 0.910908579826355, "learning_rate": 4.974177055765694e-08, "loss": 0.0117, "step": 14186 }, { "epoch": 0.9749175371083013, "grad_norm": 0.22421282529830933, "learning_rate": 4.947047823559647e-08, "loss": 0.0032, "step": 14187 }, { "epoch": 0.9749862561847169, "grad_norm": 0.2405622899532318, "learning_rate": 4.919992653203087e-08, "loss": 0.0108, "step": 14188 }, { "epoch": 0.9750549752611325, "grad_norm": 0.2308940291404724, "learning_rate": 4.893011546036441e-08, "loss": 0.0088, "step": 14189 }, { "epoch": 0.9751236943375481, "grad_norm": 0.3139554262161255, "learning_rate": 4.86610450339614e-08, "loss": 0.0109, "step": 14190 }, { "epoch": 0.9751924134139637, "grad_norm": 0.10858292877674103, "learning_rate": 4.8392715266151186e-08, "loss": 0.0073, "step": 14191 }, { "epoch": 0.9752611324903794, "grad_norm": 0.11416313052177429, "learning_rate": 4.812512617022813e-08, "loss": 0.0076, "step": 14192 }, { "epoch": 0.975329851566795, "grad_norm": 0.15138089656829834, "learning_rate": 4.785827775944496e-08, "loss": 0.0098, "step": 14193 }, { "epoch": 0.9753985706432106, "grad_norm": 0.2684805691242218, "learning_rate": 4.7592170047022764e-08, "loss": 0.0051, "step": 14194 }, { "epoch": 0.9754672897196262, "grad_norm": 0.4980331361293793, "learning_rate": 4.732680304614434e-08, "loss": 0.0106, "step": 14195 }, { "epoch": 0.9755360087960417, "grad_norm": 0.19891951978206635, "learning_rate": 4.7062176769952504e-08, "loss": 0.0055, "step": 14196 }, { "epoch": 0.9756047278724574, "grad_norm": 0.18842418491840363, "learning_rate": 4.6798291231561765e-08, "loss": 0.0119, "step": 14197 }, { "epoch": 0.975673446948873, "grad_norm": 0.4660918116569519, "learning_rate": 4.6535146444038335e-08, "loss": 0.0063, "step": 14198 }, { "epoch": 0.9757421660252886, "grad_norm": 0.21170265972614288, "learning_rate": 4.627274242042345e-08, "loss": 0.0083, "step": 14199 }, { "epoch": 0.9758108851017042, "grad_norm": 0.4961073100566864, "learning_rate": 4.6011079173711723e-08, "loss": 0.0102, "step": 14200 }, { "epoch": 0.9758796041781198, "grad_norm": 0.32674282789230347, "learning_rate": 4.575015671686611e-08, "loss": 0.0072, "step": 14201 }, { "epoch": 0.9759483232545355, "grad_norm": 0.3180598020553589, "learning_rate": 4.54899750628146e-08, "loss": 0.0059, "step": 14202 }, { "epoch": 0.9760170423309511, "grad_norm": 0.5556667447090149, "learning_rate": 4.523053422444357e-08, "loss": 0.0098, "step": 14203 }, { "epoch": 0.9760857614073667, "grad_norm": 0.5225342512130737, "learning_rate": 4.497183421460771e-08, "loss": 0.0084, "step": 14204 }, { "epoch": 0.9761544804837823, "grad_norm": 0.21374550461769104, "learning_rate": 4.4713875046120126e-08, "loss": 0.0031, "step": 14205 }, { "epoch": 0.9762231995601979, "grad_norm": 0.08236923813819885, "learning_rate": 4.445665673175892e-08, "loss": 0.0084, "step": 14206 }, { "epoch": 0.9762919186366136, "grad_norm": 0.5286702513694763, "learning_rate": 4.4200179284268895e-08, "loss": 0.0117, "step": 14207 }, { "epoch": 0.9763606377130292, "grad_norm": 0.5851877331733704, "learning_rate": 4.3944442716354896e-08, "loss": 0.0099, "step": 14208 }, { "epoch": 0.9764293567894448, "grad_norm": 0.4079389274120331, "learning_rate": 4.368944704068345e-08, "loss": 0.0103, "step": 14209 }, { "epoch": 0.9764980758658603, "grad_norm": 0.12745918333530426, "learning_rate": 4.343519226988946e-08, "loss": 0.0042, "step": 14210 }, { "epoch": 0.9765667949422759, "grad_norm": 0.35872364044189453, "learning_rate": 4.318167841656617e-08, "loss": 0.0101, "step": 14211 }, { "epoch": 0.9766355140186916, "grad_norm": 0.3303634822368622, "learning_rate": 4.2928905493271886e-08, "loss": 0.0111, "step": 14212 }, { "epoch": 0.9767042330951072, "grad_norm": 0.1886001080274582, "learning_rate": 4.267687351252991e-08, "loss": 0.0124, "step": 14213 }, { "epoch": 0.9767729521715228, "grad_norm": 0.12620487809181213, "learning_rate": 4.24255824868236e-08, "loss": 0.0047, "step": 14214 }, { "epoch": 0.9768416712479384, "grad_norm": 0.38650771975517273, "learning_rate": 4.217503242860299e-08, "loss": 0.0042, "step": 14215 }, { "epoch": 0.976910390324354, "grad_norm": 0.6801593899726868, "learning_rate": 4.1925223350279816e-08, "loss": 0.0225, "step": 14216 }, { "epoch": 0.9769791094007697, "grad_norm": 0.3197319507598877, "learning_rate": 4.167615526422919e-08, "loss": 0.0098, "step": 14217 }, { "epoch": 0.9770478284771853, "grad_norm": 0.2214440107345581, "learning_rate": 4.1427828182787894e-08, "loss": 0.007, "step": 14218 }, { "epoch": 0.9771165475536009, "grad_norm": 0.14134538173675537, "learning_rate": 4.118024211825777e-08, "loss": 0.0027, "step": 14219 }, { "epoch": 0.9771852666300165, "grad_norm": 0.17024075984954834, "learning_rate": 4.0933397082905663e-08, "loss": 0.012, "step": 14220 }, { "epoch": 0.9772539857064321, "grad_norm": 0.21308377385139465, "learning_rate": 4.068729308895847e-08, "loss": 0.0029, "step": 14221 }, { "epoch": 0.9773227047828478, "grad_norm": 0.26339370012283325, "learning_rate": 4.044193014860642e-08, "loss": 0.0092, "step": 14222 }, { "epoch": 0.9773914238592634, "grad_norm": 0.41397371888160706, "learning_rate": 4.0197308274006474e-08, "loss": 0.0055, "step": 14223 }, { "epoch": 0.977460142935679, "grad_norm": 0.2866309583187103, "learning_rate": 3.99534274772756e-08, "loss": 0.0085, "step": 14224 }, { "epoch": 0.9775288620120945, "grad_norm": 0.38297080993652344, "learning_rate": 3.971028777049579e-08, "loss": 0.0077, "step": 14225 }, { "epoch": 0.9775975810885101, "grad_norm": 0.16603383421897888, "learning_rate": 3.946788916571076e-08, "loss": 0.0066, "step": 14226 }, { "epoch": 0.9776663001649258, "grad_norm": 0.34014198184013367, "learning_rate": 3.922623167492923e-08, "loss": 0.0055, "step": 14227 }, { "epoch": 0.9777350192413414, "grad_norm": 0.15539582073688507, "learning_rate": 3.898531531012328e-08, "loss": 0.0159, "step": 14228 }, { "epoch": 0.977803738317757, "grad_norm": 0.2519164979457855, "learning_rate": 3.874514008322505e-08, "loss": 0.0056, "step": 14229 }, { "epoch": 0.9778724573941726, "grad_norm": 0.5467711091041565, "learning_rate": 3.8505706006133345e-08, "loss": 0.0071, "step": 14230 }, { "epoch": 0.9779411764705882, "grad_norm": 0.2555706799030304, "learning_rate": 3.826701309071035e-08, "loss": 0.0046, "step": 14231 }, { "epoch": 0.9780098955470038, "grad_norm": 0.6789894700050354, "learning_rate": 3.802906134877826e-08, "loss": 0.0084, "step": 14232 }, { "epoch": 0.9780786146234195, "grad_norm": 0.15409016609191895, "learning_rate": 3.779185079212599e-08, "loss": 0.0155, "step": 14233 }, { "epoch": 0.9781473336998351, "grad_norm": 0.6193872690200806, "learning_rate": 3.755538143250581e-08, "loss": 0.014, "step": 14234 }, { "epoch": 0.9782160527762507, "grad_norm": 0.5247287750244141, "learning_rate": 3.731965328163001e-08, "loss": 0.0049, "step": 14235 }, { "epoch": 0.9782847718526663, "grad_norm": 0.2979632616043091, "learning_rate": 3.708466635117591e-08, "loss": 0.0083, "step": 14236 }, { "epoch": 0.9783534909290819, "grad_norm": 0.1360139548778534, "learning_rate": 3.6850420652785876e-08, "loss": 0.0064, "step": 14237 }, { "epoch": 0.9784222100054976, "grad_norm": 0.35678961873054504, "learning_rate": 3.661691619806229e-08, "loss": 0.0084, "step": 14238 }, { "epoch": 0.9784909290819132, "grad_norm": 0.1702159345149994, "learning_rate": 3.638415299857423e-08, "loss": 0.0043, "step": 14239 }, { "epoch": 0.9785596481583287, "grad_norm": 0.5170734524726868, "learning_rate": 3.61521310658508e-08, "loss": 0.0106, "step": 14240 }, { "epoch": 0.9786283672347443, "grad_norm": 0.32570603489875793, "learning_rate": 3.5920850411386154e-08, "loss": 0.0064, "step": 14241 }, { "epoch": 0.9786970863111599, "grad_norm": 0.5394492149353027, "learning_rate": 3.569031104663778e-08, "loss": 0.0131, "step": 14242 }, { "epoch": 0.9787658053875756, "grad_norm": 0.27712276577949524, "learning_rate": 3.5460512983024887e-08, "loss": 0.0054, "step": 14243 }, { "epoch": 0.9788345244639912, "grad_norm": 0.5861419439315796, "learning_rate": 3.5231456231935024e-08, "loss": 0.015, "step": 14244 }, { "epoch": 0.9789032435404068, "grad_norm": 0.4068189561367035, "learning_rate": 3.5003140804710785e-08, "loss": 0.0054, "step": 14245 }, { "epoch": 0.9789719626168224, "grad_norm": 0.4839479625225067, "learning_rate": 3.4775566712664796e-08, "loss": 0.0086, "step": 14246 }, { "epoch": 0.979040681693238, "grad_norm": 0.9309731125831604, "learning_rate": 3.4548733967069703e-08, "loss": 0.0129, "step": 14247 }, { "epoch": 0.9791094007696537, "grad_norm": 0.21298794448375702, "learning_rate": 3.432264257916318e-08, "loss": 0.0063, "step": 14248 }, { "epoch": 0.9791781198460693, "grad_norm": 0.15335826575756073, "learning_rate": 3.40972925601446e-08, "loss": 0.0046, "step": 14249 }, { "epoch": 0.9792468389224849, "grad_norm": 0.22766625881195068, "learning_rate": 3.387268392117837e-08, "loss": 0.0062, "step": 14250 }, { "epoch": 0.9793155579989005, "grad_norm": 0.09797488898038864, "learning_rate": 3.364881667338893e-08, "loss": 0.0098, "step": 14251 }, { "epoch": 0.9793842770753161, "grad_norm": 0.2404298484325409, "learning_rate": 3.3425690827867396e-08, "loss": 0.0071, "step": 14252 }, { "epoch": 0.9794529961517318, "grad_norm": 0.4132828414440155, "learning_rate": 3.320330639566826e-08, "loss": 0.0101, "step": 14253 }, { "epoch": 0.9795217152281473, "grad_norm": 0.2581992447376251, "learning_rate": 3.298166338780606e-08, "loss": 0.0053, "step": 14254 }, { "epoch": 0.9795904343045629, "grad_norm": 0.3727468252182007, "learning_rate": 3.2760761815261997e-08, "loss": 0.0103, "step": 14255 }, { "epoch": 0.9796591533809785, "grad_norm": 0.29372337460517883, "learning_rate": 3.254060168897899e-08, "loss": 0.0057, "step": 14256 }, { "epoch": 0.9797278724573941, "grad_norm": 0.6119934916496277, "learning_rate": 3.232118301986164e-08, "loss": 0.0166, "step": 14257 }, { "epoch": 0.9797965915338098, "grad_norm": 0.13806796073913574, "learning_rate": 3.210250581878127e-08, "loss": 0.0054, "step": 14258 }, { "epoch": 0.9798653106102254, "grad_norm": 0.019418364390730858, "learning_rate": 3.188457009657087e-08, "loss": 0.0037, "step": 14259 }, { "epoch": 0.979934029686641, "grad_norm": 0.26937395334243774, "learning_rate": 3.166737586402346e-08, "loss": 0.0037, "step": 14260 }, { "epoch": 0.9800027487630566, "grad_norm": 0.27202704548835754, "learning_rate": 3.145092313190212e-08, "loss": 0.0117, "step": 14261 }, { "epoch": 0.9800714678394722, "grad_norm": 0.1410110741853714, "learning_rate": 3.123521191092826e-08, "loss": 0.004, "step": 14262 }, { "epoch": 0.9801401869158879, "grad_norm": 0.5755008459091187, "learning_rate": 3.102024221178668e-08, "loss": 0.0091, "step": 14263 }, { "epoch": 0.9802089059923035, "grad_norm": 0.2989680767059326, "learning_rate": 3.080601404512884e-08, "loss": 0.0084, "step": 14264 }, { "epoch": 0.9802776250687191, "grad_norm": 0.19576126337051392, "learning_rate": 3.05925274215646e-08, "loss": 0.0048, "step": 14265 }, { "epoch": 0.9803463441451347, "grad_norm": 0.21549370884895325, "learning_rate": 3.037978235167049e-08, "loss": 0.0104, "step": 14266 }, { "epoch": 0.9804150632215503, "grad_norm": 1.0258985757827759, "learning_rate": 3.016777884598643e-08, "loss": 0.0076, "step": 14267 }, { "epoch": 0.980483782297966, "grad_norm": 0.7088649868965149, "learning_rate": 2.995651691501233e-08, "loss": 0.0099, "step": 14268 }, { "epoch": 0.9805525013743815, "grad_norm": 0.3921816945075989, "learning_rate": 2.974599656921817e-08, "loss": 0.0141, "step": 14269 }, { "epoch": 0.9806212204507971, "grad_norm": 0.3896625339984894, "learning_rate": 2.9536217819027266e-08, "loss": 0.0085, "step": 14270 }, { "epoch": 0.9806899395272127, "grad_norm": 0.30969226360321045, "learning_rate": 2.9327180674834643e-08, "loss": 0.0066, "step": 14271 }, { "epoch": 0.9807586586036283, "grad_norm": 0.7861739993095398, "learning_rate": 2.9118885146995346e-08, "loss": 0.019, "step": 14272 }, { "epoch": 0.980827377680044, "grad_norm": 0.13528423011302948, "learning_rate": 2.8911331245827788e-08, "loss": 0.0071, "step": 14273 }, { "epoch": 0.9808960967564596, "grad_norm": 0.19080394506454468, "learning_rate": 2.8704518981613747e-08, "loss": 0.005, "step": 14274 }, { "epoch": 0.9809648158328752, "grad_norm": 0.20213286578655243, "learning_rate": 2.8498448364598363e-08, "loss": 0.0124, "step": 14275 }, { "epoch": 0.9810335349092908, "grad_norm": 0.31636327505111694, "learning_rate": 2.829311940499013e-08, "loss": 0.0096, "step": 14276 }, { "epoch": 0.9811022539857064, "grad_norm": 0.04937789961695671, "learning_rate": 2.8088532112959252e-08, "loss": 0.0153, "step": 14277 }, { "epoch": 0.9811709730621221, "grad_norm": 0.32506680488586426, "learning_rate": 2.7884686498640954e-08, "loss": 0.0073, "step": 14278 }, { "epoch": 0.9812396921385377, "grad_norm": 0.21863268315792084, "learning_rate": 2.768158257213549e-08, "loss": 0.0106, "step": 14279 }, { "epoch": 0.9813084112149533, "grad_norm": 0.31958961486816406, "learning_rate": 2.7479220343501478e-08, "loss": 0.013, "step": 14280 }, { "epoch": 0.9813771302913689, "grad_norm": 0.4745994806289673, "learning_rate": 2.7277599822765896e-08, "loss": 0.0105, "step": 14281 }, { "epoch": 0.9814458493677845, "grad_norm": 0.5252524018287659, "learning_rate": 2.7076721019914097e-08, "loss": 0.0096, "step": 14282 }, { "epoch": 0.9815145684442002, "grad_norm": 0.2844780683517456, "learning_rate": 2.6876583944899778e-08, "loss": 0.0117, "step": 14283 }, { "epoch": 0.9815832875206157, "grad_norm": 0.32570338249206543, "learning_rate": 2.6677188607635018e-08, "loss": 0.0114, "step": 14284 }, { "epoch": 0.9816520065970313, "grad_norm": 0.15592066943645477, "learning_rate": 2.6478535017998572e-08, "loss": 0.011, "step": 14285 }, { "epoch": 0.9817207256734469, "grad_norm": 0.1792418360710144, "learning_rate": 2.6280623185832574e-08, "loss": 0.0081, "step": 14286 }, { "epoch": 0.9817894447498625, "grad_norm": 0.22540295124053955, "learning_rate": 2.6083453120937516e-08, "loss": 0.0057, "step": 14287 }, { "epoch": 0.9818581638262782, "grad_norm": 0.7862467765808105, "learning_rate": 2.5887024833085583e-08, "loss": 0.01, "step": 14288 }, { "epoch": 0.9819268829026938, "grad_norm": 0.5074895620346069, "learning_rate": 2.5691338332003987e-08, "loss": 0.01, "step": 14289 }, { "epoch": 0.9819956019791094, "grad_norm": 0.29428932070732117, "learning_rate": 2.5496393627386648e-08, "loss": 0.0092, "step": 14290 }, { "epoch": 0.982064321055525, "grad_norm": 0.23549391329288483, "learning_rate": 2.5302190728892506e-08, "loss": 0.005, "step": 14291 }, { "epoch": 0.9821330401319406, "grad_norm": 0.30700263381004333, "learning_rate": 2.5108729646140528e-08, "loss": 0.0109, "step": 14292 }, { "epoch": 0.9822017592083563, "grad_norm": 0.35977140069007874, "learning_rate": 2.4916010388716383e-08, "loss": 0.0096, "step": 14293 }, { "epoch": 0.9822704782847719, "grad_norm": 0.2190253734588623, "learning_rate": 2.4724032966164102e-08, "loss": 0.0106, "step": 14294 }, { "epoch": 0.9823391973611875, "grad_norm": 0.3417361378669739, "learning_rate": 2.4532797387996074e-08, "loss": 0.0079, "step": 14295 }, { "epoch": 0.9824079164376031, "grad_norm": 0.25332918763160706, "learning_rate": 2.434230366368473e-08, "loss": 0.005, "step": 14296 }, { "epoch": 0.9824766355140186, "grad_norm": 0.3958773910999298, "learning_rate": 2.415255180266751e-08, "loss": 0.0072, "step": 14297 }, { "epoch": 0.9825453545904343, "grad_norm": 0.3111390769481659, "learning_rate": 2.396354181434357e-08, "loss": 0.0065, "step": 14298 }, { "epoch": 0.9826140736668499, "grad_norm": 0.2977413237094879, "learning_rate": 2.3775273708077084e-08, "loss": 0.0079, "step": 14299 }, { "epoch": 0.9826827927432655, "grad_norm": 0.21483619511127472, "learning_rate": 2.3587747493192257e-08, "loss": 0.0064, "step": 14300 }, { "epoch": 0.9827515118196811, "grad_norm": 0.27419888973236084, "learning_rate": 2.340096317898166e-08, "loss": 0.0061, "step": 14301 }, { "epoch": 0.9828202308960967, "grad_norm": 0.29815009236335754, "learning_rate": 2.3214920774696225e-08, "loss": 0.006, "step": 14302 }, { "epoch": 0.9828889499725124, "grad_norm": 0.5228893756866455, "learning_rate": 2.3029620289551913e-08, "loss": 0.0058, "step": 14303 }, { "epoch": 0.982957669048928, "grad_norm": 0.17892713844776154, "learning_rate": 2.2845061732729712e-08, "loss": 0.0051, "step": 14304 }, { "epoch": 0.9830263881253436, "grad_norm": 0.5608581304550171, "learning_rate": 2.2661245113372308e-08, "loss": 0.0087, "step": 14305 }, { "epoch": 0.9830951072017592, "grad_norm": 0.3532474935054779, "learning_rate": 2.2478170440582423e-08, "loss": 0.0113, "step": 14306 }, { "epoch": 0.9831638262781748, "grad_norm": 0.44164684414863586, "learning_rate": 2.22958377234328e-08, "loss": 0.0089, "step": 14307 }, { "epoch": 0.9832325453545905, "grad_norm": 0.15453366935253143, "learning_rate": 2.2114246970954543e-08, "loss": 0.0089, "step": 14308 }, { "epoch": 0.9833012644310061, "grad_norm": 0.32708483934402466, "learning_rate": 2.1933398192143796e-08, "loss": 0.0098, "step": 14309 }, { "epoch": 0.9833699835074217, "grad_norm": 0.2263118475675583, "learning_rate": 2.1753291395958385e-08, "loss": 0.0085, "step": 14310 }, { "epoch": 0.9834387025838373, "grad_norm": 0.2067241668701172, "learning_rate": 2.1573926591319515e-08, "loss": 0.0043, "step": 14311 }, { "epoch": 0.9835074216602528, "grad_norm": 0.29232847690582275, "learning_rate": 2.1395303787116736e-08, "loss": 0.0058, "step": 14312 }, { "epoch": 0.9835761407366685, "grad_norm": 0.3337899446487427, "learning_rate": 2.121742299219298e-08, "loss": 0.0054, "step": 14313 }, { "epoch": 0.9836448598130841, "grad_norm": 0.11922284960746765, "learning_rate": 2.1040284215364526e-08, "loss": 0.0062, "step": 14314 }, { "epoch": 0.9837135788894997, "grad_norm": 0.1031898558139801, "learning_rate": 2.0863887465406016e-08, "loss": 0.0076, "step": 14315 }, { "epoch": 0.9837822979659153, "grad_norm": 0.5146011710166931, "learning_rate": 2.0688232751053805e-08, "loss": 0.0087, "step": 14316 }, { "epoch": 0.9838510170423309, "grad_norm": 0.2763393819332123, "learning_rate": 2.0513320081009258e-08, "loss": 0.0078, "step": 14317 }, { "epoch": 0.9839197361187466, "grad_norm": 0.13830837607383728, "learning_rate": 2.033914946394044e-08, "loss": 0.0117, "step": 14318 }, { "epoch": 0.9839884551951622, "grad_norm": 0.12003021687269211, "learning_rate": 2.0165720908470464e-08, "loss": 0.0039, "step": 14319 }, { "epoch": 0.9840571742715778, "grad_norm": 0.39532050490379333, "learning_rate": 1.9993034423195776e-08, "loss": 0.0069, "step": 14320 }, { "epoch": 0.9841258933479934, "grad_norm": 0.2509765028953552, "learning_rate": 1.9821090016667876e-08, "loss": 0.0048, "step": 14321 }, { "epoch": 0.984194612424409, "grad_norm": 0.2551174461841583, "learning_rate": 1.9649887697404945e-08, "loss": 0.0075, "step": 14322 }, { "epoch": 0.9842633315008247, "grad_norm": 0.2643279731273651, "learning_rate": 1.9479427473888533e-08, "loss": 0.0111, "step": 14323 }, { "epoch": 0.9843320505772403, "grad_norm": 0.6222530007362366, "learning_rate": 1.9309709354563553e-08, "loss": 0.0097, "step": 14324 }, { "epoch": 0.9844007696536559, "grad_norm": 0.38740313053131104, "learning_rate": 1.914073334783495e-08, "loss": 0.0063, "step": 14325 }, { "epoch": 0.9844694887300715, "grad_norm": 0.3688744902610779, "learning_rate": 1.897249946207602e-08, "loss": 0.0101, "step": 14326 }, { "epoch": 0.984538207806487, "grad_norm": 0.3231142461299896, "learning_rate": 1.8805007705620103e-08, "loss": 0.0084, "step": 14327 }, { "epoch": 0.9846069268829027, "grad_norm": 0.1775401532649994, "learning_rate": 1.8638258086762227e-08, "loss": 0.0096, "step": 14328 }, { "epoch": 0.9846756459593183, "grad_norm": 0.2620523273944855, "learning_rate": 1.847225061376745e-08, "loss": 0.0061, "step": 14329 }, { "epoch": 0.9847443650357339, "grad_norm": 0.25671157240867615, "learning_rate": 1.8306985294854195e-08, "loss": 0.0097, "step": 14330 }, { "epoch": 0.9848130841121495, "grad_norm": 0.1807742565870285, "learning_rate": 1.814246213821258e-08, "loss": 0.01, "step": 14331 }, { "epoch": 0.9848818031885651, "grad_norm": 0.4821779727935791, "learning_rate": 1.797868115199275e-08, "loss": 0.0135, "step": 14332 }, { "epoch": 0.9849505222649808, "grad_norm": 0.19893383979797363, "learning_rate": 1.781564234430655e-08, "loss": 0.0067, "step": 14333 }, { "epoch": 0.9850192413413964, "grad_norm": 0.23508034646511078, "learning_rate": 1.765334572323085e-08, "loss": 0.0081, "step": 14334 }, { "epoch": 0.985087960417812, "grad_norm": 0.3736628592014313, "learning_rate": 1.749179129680756e-08, "loss": 0.0141, "step": 14335 }, { "epoch": 0.9851566794942276, "grad_norm": 0.19573712348937988, "learning_rate": 1.7330979073036936e-08, "loss": 0.0037, "step": 14336 }, { "epoch": 0.9852253985706432, "grad_norm": 0.17781712114810944, "learning_rate": 1.7170909059887608e-08, "loss": 0.0106, "step": 14337 }, { "epoch": 0.9852941176470589, "grad_norm": 0.29450538754463196, "learning_rate": 1.7011581265288233e-08, "loss": 0.0077, "step": 14338 }, { "epoch": 0.9853628367234745, "grad_norm": 0.44260725378990173, "learning_rate": 1.685299569712917e-08, "loss": 0.0074, "step": 14339 }, { "epoch": 0.9854315557998901, "grad_norm": 0.13133753836154938, "learning_rate": 1.6695152363270792e-08, "loss": 0.0053, "step": 14340 }, { "epoch": 0.9855002748763056, "grad_norm": 0.5241823792457581, "learning_rate": 1.6538051271530184e-08, "loss": 0.0173, "step": 14341 }, { "epoch": 0.9855689939527212, "grad_norm": 0.32150888442993164, "learning_rate": 1.638169242968779e-08, "loss": 0.0077, "step": 14342 }, { "epoch": 0.9856377130291369, "grad_norm": 0.15561944246292114, "learning_rate": 1.622607584549407e-08, "loss": 0.008, "step": 14343 }, { "epoch": 0.9857064321055525, "grad_norm": 0.24330253899097443, "learning_rate": 1.6071201526652866e-08, "loss": 0.0098, "step": 14344 }, { "epoch": 0.9857751511819681, "grad_norm": 0.2544255554676056, "learning_rate": 1.5917069480839708e-08, "loss": 0.0047, "step": 14345 }, { "epoch": 0.9858438702583837, "grad_norm": 0.7736639380455017, "learning_rate": 1.576367971569015e-08, "loss": 0.008, "step": 14346 }, { "epoch": 0.9859125893347993, "grad_norm": 0.5826893448829651, "learning_rate": 1.5611032238799784e-08, "loss": 0.01, "step": 14347 }, { "epoch": 0.985981308411215, "grad_norm": 0.3296586573123932, "learning_rate": 1.545912705773256e-08, "loss": 0.0124, "step": 14348 }, { "epoch": 0.9860500274876306, "grad_norm": 0.49034053087234497, "learning_rate": 1.530796418001412e-08, "loss": 0.0074, "step": 14349 }, { "epoch": 0.9861187465640462, "grad_norm": 0.18742844462394714, "learning_rate": 1.515754361313182e-08, "loss": 0.007, "step": 14350 }, { "epoch": 0.9861874656404618, "grad_norm": 0.18220201134681702, "learning_rate": 1.5007865364536356e-08, "loss": 0.0091, "step": 14351 }, { "epoch": 0.9862561847168774, "grad_norm": 0.35870322585105896, "learning_rate": 1.4858929441643465e-08, "loss": 0.0142, "step": 14352 }, { "epoch": 0.9863249037932931, "grad_norm": 0.20208019018173218, "learning_rate": 1.4710735851832246e-08, "loss": 0.0095, "step": 14353 }, { "epoch": 0.9863936228697087, "grad_norm": 0.7711992263793945, "learning_rate": 1.4563284602441828e-08, "loss": 0.0066, "step": 14354 }, { "epoch": 0.9864623419461243, "grad_norm": 0.36030343174934387, "learning_rate": 1.4416575700778034e-08, "loss": 0.0136, "step": 14355 }, { "epoch": 0.9865310610225398, "grad_norm": 0.2473917305469513, "learning_rate": 1.4270609154106718e-08, "loss": 0.0062, "step": 14356 }, { "epoch": 0.9865997800989554, "grad_norm": 0.2955692410469055, "learning_rate": 1.4125384969660426e-08, "loss": 0.0139, "step": 14357 }, { "epoch": 0.986668499175371, "grad_norm": 0.32709601521492004, "learning_rate": 1.3980903154635071e-08, "loss": 0.0081, "step": 14358 }, { "epoch": 0.9867372182517867, "grad_norm": 0.2942088544368744, "learning_rate": 1.3837163716183265e-08, "loss": 0.0113, "step": 14359 }, { "epoch": 0.9868059373282023, "grad_norm": 0.46567830443382263, "learning_rate": 1.3694166661429308e-08, "loss": 0.007, "step": 14360 }, { "epoch": 0.9868746564046179, "grad_norm": 0.3456980884075165, "learning_rate": 1.3551911997455868e-08, "loss": 0.0054, "step": 14361 }, { "epoch": 0.9869433754810335, "grad_norm": 0.3530398905277252, "learning_rate": 1.3410399731310641e-08, "loss": 0.0097, "step": 14362 }, { "epoch": 0.9870120945574491, "grad_norm": 0.3685750365257263, "learning_rate": 1.3269629870001354e-08, "loss": 0.0058, "step": 14363 }, { "epoch": 0.9870808136338648, "grad_norm": 0.21222586929798126, "learning_rate": 1.3129602420504095e-08, "loss": 0.0046, "step": 14364 }, { "epoch": 0.9871495327102804, "grad_norm": 0.24476994574069977, "learning_rate": 1.2990317389753315e-08, "loss": 0.0091, "step": 14365 }, { "epoch": 0.987218251786696, "grad_norm": 0.44527074694633484, "learning_rate": 1.2851774784651827e-08, "loss": 0.0106, "step": 14366 }, { "epoch": 0.9872869708631116, "grad_norm": 0.24922214448451996, "learning_rate": 1.2713974612060809e-08, "loss": 0.0049, "step": 14367 }, { "epoch": 0.9873556899395272, "grad_norm": 0.5879213809967041, "learning_rate": 1.2576916878806465e-08, "loss": 0.019, "step": 14368 }, { "epoch": 0.9874244090159429, "grad_norm": 0.20861828327178955, "learning_rate": 1.2440601591680035e-08, "loss": 0.0076, "step": 14369 }, { "epoch": 0.9874931280923585, "grad_norm": 0.20570380985736847, "learning_rate": 1.2305028757431114e-08, "loss": 0.0056, "step": 14370 }, { "epoch": 0.987561847168774, "grad_norm": 0.15420717000961304, "learning_rate": 1.217019838277933e-08, "loss": 0.0074, "step": 14371 }, { "epoch": 0.9876305662451896, "grad_norm": 0.20397567749023438, "learning_rate": 1.2036110474402672e-08, "loss": 0.0047, "step": 14372 }, { "epoch": 0.9876992853216052, "grad_norm": 0.19825711846351624, "learning_rate": 1.1902765038940833e-08, "loss": 0.0087, "step": 14373 }, { "epoch": 0.9877680043980209, "grad_norm": 0.5462325811386108, "learning_rate": 1.1770162083003522e-08, "loss": 0.0077, "step": 14374 }, { "epoch": 0.9878367234744365, "grad_norm": 0.11052720993757248, "learning_rate": 1.1638301613157155e-08, "loss": 0.0049, "step": 14375 }, { "epoch": 0.9879054425508521, "grad_norm": 0.2598378658294678, "learning_rate": 1.1507183635934837e-08, "loss": 0.01, "step": 14376 }, { "epoch": 0.9879741616272677, "grad_norm": 1.0662930011749268, "learning_rate": 1.137680815783304e-08, "loss": 0.0136, "step": 14377 }, { "epoch": 0.9880428807036833, "grad_norm": 0.5462879538536072, "learning_rate": 1.1247175185306602e-08, "loss": 0.0077, "step": 14378 }, { "epoch": 0.988111599780099, "grad_norm": 0.4077637195587158, "learning_rate": 1.1118284724780381e-08, "loss": 0.0078, "step": 14379 }, { "epoch": 0.9881803188565146, "grad_norm": 0.0711456760764122, "learning_rate": 1.099013678263927e-08, "loss": 0.0086, "step": 14380 }, { "epoch": 0.9882490379329302, "grad_norm": 0.13952048122882843, "learning_rate": 1.0862731365229861e-08, "loss": 0.0035, "step": 14381 }, { "epoch": 0.9883177570093458, "grad_norm": 0.32908210158348083, "learning_rate": 1.073606847886377e-08, "loss": 0.007, "step": 14382 }, { "epoch": 0.9883864760857614, "grad_norm": 0.24328483641147614, "learning_rate": 1.0610148129817642e-08, "loss": 0.0061, "step": 14383 }, { "epoch": 0.9884551951621771, "grad_norm": 0.12437866628170013, "learning_rate": 1.0484970324326493e-08, "loss": 0.0031, "step": 14384 }, { "epoch": 0.9885239142385926, "grad_norm": 0.36529839038848877, "learning_rate": 1.0360535068593691e-08, "loss": 0.0062, "step": 14385 }, { "epoch": 0.9885926333150082, "grad_norm": 0.09367792308330536, "learning_rate": 1.0236842368782639e-08, "loss": 0.0077, "step": 14386 }, { "epoch": 0.9886613523914238, "grad_norm": 0.1140633150935173, "learning_rate": 1.0113892231020106e-08, "loss": 0.007, "step": 14387 }, { "epoch": 0.9887300714678394, "grad_norm": 0.3940421938896179, "learning_rate": 9.991684661397881e-09, "loss": 0.0069, "step": 14388 }, { "epoch": 0.9887987905442551, "grad_norm": 0.057047683745622635, "learning_rate": 9.87021966596946e-09, "loss": 0.0036, "step": 14389 }, { "epoch": 0.9888675096206707, "grad_norm": 0.2723960876464844, "learning_rate": 9.749497250751693e-09, "loss": 0.0088, "step": 14390 }, { "epoch": 0.9889362286970863, "grad_norm": 0.29896852374076843, "learning_rate": 9.629517421726464e-09, "loss": 0.005, "step": 14391 }, { "epoch": 0.9890049477735019, "grad_norm": 0.5231678485870361, "learning_rate": 9.51028018483402e-09, "loss": 0.0172, "step": 14392 }, { "epoch": 0.9890736668499175, "grad_norm": 0.24307706952095032, "learning_rate": 9.391785545984633e-09, "loss": 0.0069, "step": 14393 }, { "epoch": 0.9891423859263332, "grad_norm": 0.6733373999595642, "learning_rate": 9.274033511046942e-09, "loss": 0.0061, "step": 14394 }, { "epoch": 0.9892111050027488, "grad_norm": 0.29033565521240234, "learning_rate": 9.15702408585295e-09, "loss": 0.0105, "step": 14395 }, { "epoch": 0.9892798240791644, "grad_norm": 0.20203226804733276, "learning_rate": 9.040757276199685e-09, "loss": 0.0075, "step": 14396 }, { "epoch": 0.98934854315558, "grad_norm": 0.2380916178226471, "learning_rate": 8.925233087849205e-09, "loss": 0.0123, "step": 14397 }, { "epoch": 0.9894172622319956, "grad_norm": 0.19583559036254883, "learning_rate": 8.810451526520268e-09, "loss": 0.0046, "step": 14398 }, { "epoch": 0.9894859813084113, "grad_norm": 0.23883335292339325, "learning_rate": 8.69641259790166e-09, "loss": 0.0038, "step": 14399 }, { "epoch": 0.9895547003848268, "grad_norm": 0.543258011341095, "learning_rate": 8.583116307642191e-09, "loss": 0.0051, "step": 14400 }, { "epoch": 0.9896234194612424, "grad_norm": 0.12874147295951843, "learning_rate": 8.470562661354042e-09, "loss": 0.0109, "step": 14401 }, { "epoch": 0.989692138537658, "grad_norm": 0.21883243322372437, "learning_rate": 8.358751664612751e-09, "loss": 0.0057, "step": 14402 }, { "epoch": 0.9897608576140736, "grad_norm": 0.5634143948554993, "learning_rate": 8.247683322958888e-09, "loss": 0.0127, "step": 14403 }, { "epoch": 0.9898295766904893, "grad_norm": 0.5783631205558777, "learning_rate": 8.137357641891386e-09, "loss": 0.0145, "step": 14404 }, { "epoch": 0.9898982957669049, "grad_norm": 0.3954181373119354, "learning_rate": 8.027774626877538e-09, "loss": 0.0102, "step": 14405 }, { "epoch": 0.9899670148433205, "grad_norm": 0.23844772577285767, "learning_rate": 7.91893428334467e-09, "loss": 0.0084, "step": 14406 }, { "epoch": 0.9900357339197361, "grad_norm": 0.24655978381633759, "learning_rate": 7.810836616686801e-09, "loss": 0.005, "step": 14407 }, { "epoch": 0.9901044529961517, "grad_norm": 0.4380119740962982, "learning_rate": 7.70348163225798e-09, "loss": 0.0118, "step": 14408 }, { "epoch": 0.9901731720725674, "grad_norm": 0.05716921016573906, "learning_rate": 7.596869335373957e-09, "loss": 0.0075, "step": 14409 }, { "epoch": 0.990241891148983, "grad_norm": 0.5385076403617859, "learning_rate": 7.490999731318837e-09, "loss": 0.0086, "step": 14410 }, { "epoch": 0.9903106102253986, "grad_norm": 0.5646872520446777, "learning_rate": 7.385872825336759e-09, "loss": 0.0084, "step": 14411 }, { "epoch": 0.9903793293018142, "grad_norm": 0.21224157512187958, "learning_rate": 7.2814886226335585e-09, "loss": 0.0035, "step": 14412 }, { "epoch": 0.9904480483782298, "grad_norm": 0.11077943444252014, "learning_rate": 7.17784712838343e-09, "loss": 0.0074, "step": 14413 }, { "epoch": 0.9905167674546455, "grad_norm": 0.2799651026725769, "learning_rate": 7.074948347718935e-09, "loss": 0.0064, "step": 14414 }, { "epoch": 0.990585486531061, "grad_norm": 0.37871548533439636, "learning_rate": 6.972792285735996e-09, "loss": 0.0052, "step": 14415 }, { "epoch": 0.9906542056074766, "grad_norm": 0.41512930393218994, "learning_rate": 6.871378947497231e-09, "loss": 0.0052, "step": 14416 }, { "epoch": 0.9907229246838922, "grad_norm": 0.40673166513442993, "learning_rate": 6.770708338025289e-09, "loss": 0.0075, "step": 14417 }, { "epoch": 0.9907916437603078, "grad_norm": 0.3443223237991333, "learning_rate": 6.670780462307846e-09, "loss": 0.0089, "step": 14418 }, { "epoch": 0.9908603628367235, "grad_norm": 0.30687081813812256, "learning_rate": 6.5715953252959426e-09, "loss": 0.0096, "step": 14419 }, { "epoch": 0.9909290819131391, "grad_norm": 0.22643327713012695, "learning_rate": 6.47315293190065e-09, "loss": 0.009, "step": 14420 }, { "epoch": 0.9909978009895547, "grad_norm": 0.3007238209247589, "learning_rate": 6.375453287001398e-09, "loss": 0.003, "step": 14421 }, { "epoch": 0.9910665200659703, "grad_norm": 0.32226499915122986, "learning_rate": 6.278496395434319e-09, "loss": 0.0132, "step": 14422 }, { "epoch": 0.9911352391423859, "grad_norm": 0.3084299862384796, "learning_rate": 6.182282262007233e-09, "loss": 0.0081, "step": 14423 }, { "epoch": 0.9912039582188016, "grad_norm": 0.3146580159664154, "learning_rate": 6.086810891482997e-09, "loss": 0.006, "step": 14424 }, { "epoch": 0.9912726772952172, "grad_norm": 0.5513904690742493, "learning_rate": 5.992082288591161e-09, "loss": 0.0109, "step": 14425 }, { "epoch": 0.9913413963716328, "grad_norm": 0.16634099185466766, "learning_rate": 5.898096458026303e-09, "loss": 0.0054, "step": 14426 }, { "epoch": 0.9914101154480484, "grad_norm": 1.3052905797958374, "learning_rate": 5.804853404443034e-09, "loss": 0.0164, "step": 14427 }, { "epoch": 0.991478834524464, "grad_norm": 0.20932334661483765, "learning_rate": 5.712353132459325e-09, "loss": 0.0049, "step": 14428 }, { "epoch": 0.9915475536008796, "grad_norm": 0.274100124835968, "learning_rate": 5.620595646659843e-09, "loss": 0.0157, "step": 14429 }, { "epoch": 0.9916162726772952, "grad_norm": 0.42596131563186646, "learning_rate": 5.529580951589286e-09, "loss": 0.0095, "step": 14430 }, { "epoch": 0.9916849917537108, "grad_norm": 0.4384574890136719, "learning_rate": 5.439309051754049e-09, "loss": 0.014, "step": 14431 }, { "epoch": 0.9917537108301264, "grad_norm": 0.37358444929122925, "learning_rate": 5.349779951630551e-09, "loss": 0.009, "step": 14432 }, { "epoch": 0.991822429906542, "grad_norm": 0.28826385736465454, "learning_rate": 5.260993655650248e-09, "loss": 0.0094, "step": 14433 }, { "epoch": 0.9918911489829577, "grad_norm": 0.037494201213121414, "learning_rate": 5.172950168211288e-09, "loss": 0.0038, "step": 14434 }, { "epoch": 0.9919598680593733, "grad_norm": 0.5698632001876831, "learning_rate": 5.085649493676847e-09, "loss": 0.0116, "step": 14435 }, { "epoch": 0.9920285871357889, "grad_norm": 0.30205127596855164, "learning_rate": 4.9990916363718e-09, "loss": 0.0225, "step": 14436 }, { "epoch": 0.9920973062122045, "grad_norm": 0.217315673828125, "learning_rate": 4.913276600582717e-09, "loss": 0.0068, "step": 14437 }, { "epoch": 0.9921660252886201, "grad_norm": 0.2693471610546112, "learning_rate": 4.828204390561198e-09, "loss": 0.0115, "step": 14438 }, { "epoch": 0.9922347443650358, "grad_norm": 0.4457927346229553, "learning_rate": 4.743875010522203e-09, "loss": 0.0086, "step": 14439 }, { "epoch": 0.9923034634414514, "grad_norm": 0.22682058811187744, "learning_rate": 4.660288464644058e-09, "loss": 0.0111, "step": 14440 }, { "epoch": 0.992372182517867, "grad_norm": 0.19518624246120453, "learning_rate": 4.577444757063454e-09, "loss": 0.013, "step": 14441 }, { "epoch": 0.9924409015942826, "grad_norm": 0.4904019236564636, "learning_rate": 4.49534389188877e-09, "loss": 0.0074, "step": 14442 }, { "epoch": 0.9925096206706981, "grad_norm": 0.31501856446266174, "learning_rate": 4.413985873185089e-09, "loss": 0.0083, "step": 14443 }, { "epoch": 0.9925783397471138, "grad_norm": 0.5317071676254272, "learning_rate": 4.3333707049808545e-09, "loss": 0.0088, "step": 14444 }, { "epoch": 0.9926470588235294, "grad_norm": 0.36361315846443176, "learning_rate": 4.253498391274535e-09, "loss": 0.0087, "step": 14445 }, { "epoch": 0.992715777899945, "grad_norm": 0.19320018589496613, "learning_rate": 4.174368936017969e-09, "loss": 0.0067, "step": 14446 }, { "epoch": 0.9927844969763606, "grad_norm": 0.6726513504981995, "learning_rate": 4.0959823431346855e-09, "loss": 0.0108, "step": 14447 }, { "epoch": 0.9928532160527762, "grad_norm": 0.2735706865787506, "learning_rate": 4.018338616504913e-09, "loss": 0.0066, "step": 14448 }, { "epoch": 0.9929219351291919, "grad_norm": 0.08923723548650742, "learning_rate": 3.941437759975575e-09, "loss": 0.0063, "step": 14449 }, { "epoch": 0.9929906542056075, "grad_norm": 0.38749563694000244, "learning_rate": 3.865279777358622e-09, "loss": 0.0188, "step": 14450 }, { "epoch": 0.9930593732820231, "grad_norm": 0.3888357877731323, "learning_rate": 3.78986467242437e-09, "loss": 0.0054, "step": 14451 }, { "epoch": 0.9931280923584387, "grad_norm": 0.45619726181030273, "learning_rate": 3.715192448908167e-09, "loss": 0.0117, "step": 14452 }, { "epoch": 0.9931968114348543, "grad_norm": 0.5054832696914673, "learning_rate": 3.64126311051205e-09, "loss": 0.006, "step": 14453 }, { "epoch": 0.99326553051127, "grad_norm": 0.2077638804912567, "learning_rate": 3.568076660894759e-09, "loss": 0.0067, "step": 14454 }, { "epoch": 0.9933342495876856, "grad_norm": 0.5123234987258911, "learning_rate": 3.4956331036850586e-09, "loss": 0.005, "step": 14455 }, { "epoch": 0.9934029686641012, "grad_norm": 0.4950799345970154, "learning_rate": 3.423932442470079e-09, "loss": 0.0051, "step": 14456 }, { "epoch": 0.9934716877405168, "grad_norm": 0.4630815088748932, "learning_rate": 3.352974680801979e-09, "loss": 0.0099, "step": 14457 }, { "epoch": 0.9935404068169323, "grad_norm": 0.09965628385543823, "learning_rate": 3.2827598221946144e-09, "loss": 0.0062, "step": 14458 }, { "epoch": 0.993609125893348, "grad_norm": 0.5317139625549316, "learning_rate": 3.213287870126869e-09, "loss": 0.009, "step": 14459 }, { "epoch": 0.9936778449697636, "grad_norm": 0.43983495235443115, "learning_rate": 3.144558828042654e-09, "loss": 0.0093, "step": 14460 }, { "epoch": 0.9937465640461792, "grad_norm": 0.15147081017494202, "learning_rate": 3.0765726993442487e-09, "loss": 0.0074, "step": 14461 }, { "epoch": 0.9938152831225948, "grad_norm": 0.16159702837467194, "learning_rate": 3.0093294873989595e-09, "loss": 0.0043, "step": 14462 }, { "epoch": 0.9938840021990104, "grad_norm": 0.3933204412460327, "learning_rate": 2.942829195540786e-09, "loss": 0.0065, "step": 14463 }, { "epoch": 0.9939527212754261, "grad_norm": 0.25907227396965027, "learning_rate": 2.877071827060429e-09, "loss": 0.0049, "step": 14464 }, { "epoch": 0.9940214403518417, "grad_norm": 0.1666921228170395, "learning_rate": 2.8120573852186136e-09, "loss": 0.0049, "step": 14465 }, { "epoch": 0.9940901594282573, "grad_norm": 0.6106996536254883, "learning_rate": 2.7477858732327667e-09, "loss": 0.0091, "step": 14466 }, { "epoch": 0.9941588785046729, "grad_norm": 0.24916180968284607, "learning_rate": 2.684257294290338e-09, "loss": 0.0107, "step": 14467 }, { "epoch": 0.9942275975810885, "grad_norm": 0.4187367558479309, "learning_rate": 2.6214716515354786e-09, "loss": 0.0088, "step": 14468 }, { "epoch": 0.9942963166575042, "grad_norm": 0.11016031354665756, "learning_rate": 2.5594289480806996e-09, "loss": 0.0069, "step": 14469 }, { "epoch": 0.9943650357339198, "grad_norm": 0.4965266287326813, "learning_rate": 2.498129186996878e-09, "loss": 0.0099, "step": 14470 }, { "epoch": 0.9944337548103354, "grad_norm": 0.891008734703064, "learning_rate": 2.437572371323249e-09, "loss": 0.011, "step": 14471 }, { "epoch": 0.994502473886751, "grad_norm": 0.32837557792663574, "learning_rate": 2.3777585040590797e-09, "loss": 0.0084, "step": 14472 }, { "epoch": 0.9945711929631665, "grad_norm": 0.05057568475604057, "learning_rate": 2.318687588165336e-09, "loss": 0.0161, "step": 14473 }, { "epoch": 0.9946399120395822, "grad_norm": 0.05264825001358986, "learning_rate": 2.26035962657134e-09, "loss": 0.007, "step": 14474 }, { "epoch": 0.9947086311159978, "grad_norm": 0.48510295152664185, "learning_rate": 2.2027746221647827e-09, "loss": 0.0073, "step": 14475 }, { "epoch": 0.9947773501924134, "grad_norm": 0.4016864597797394, "learning_rate": 2.1459325777983816e-09, "loss": 0.0132, "step": 14476 }, { "epoch": 0.994846069268829, "grad_norm": 0.47609269618988037, "learning_rate": 2.0898334962882184e-09, "loss": 0.0105, "step": 14477 }, { "epoch": 0.9949147883452446, "grad_norm": 0.30076244473457336, "learning_rate": 2.034477380413735e-09, "loss": 0.0041, "step": 14478 }, { "epoch": 0.9949835074216603, "grad_norm": 0.15688244998455048, "learning_rate": 1.9798642329160733e-09, "loss": 0.0067, "step": 14479 }, { "epoch": 0.9950522264980759, "grad_norm": 0.5123544335365295, "learning_rate": 1.9259940565014013e-09, "loss": 0.0068, "step": 14480 }, { "epoch": 0.9951209455744915, "grad_norm": 0.2297620326280594, "learning_rate": 1.8728668538375847e-09, "loss": 0.0058, "step": 14481 }, { "epoch": 0.9951896646509071, "grad_norm": 0.4650304615497589, "learning_rate": 1.8204826275591834e-09, "loss": 0.0152, "step": 14482 }, { "epoch": 0.9952583837273227, "grad_norm": 1.157126784324646, "learning_rate": 1.7688413802574577e-09, "loss": 0.0068, "step": 14483 }, { "epoch": 0.9953271028037384, "grad_norm": 0.4730969965457916, "learning_rate": 1.7179431144920266e-09, "loss": 0.011, "step": 14484 }, { "epoch": 0.995395821880154, "grad_norm": 0.16945527493953705, "learning_rate": 1.667787832785872e-09, "loss": 0.0058, "step": 14485 }, { "epoch": 0.9954645409565696, "grad_norm": 0.2756410539150238, "learning_rate": 1.6183755376220077e-09, "loss": 0.0129, "step": 14486 }, { "epoch": 0.9955332600329851, "grad_norm": 0.4026450216770172, "learning_rate": 1.5697062314468103e-09, "loss": 0.0082, "step": 14487 }, { "epoch": 0.9956019791094007, "grad_norm": 0.11058033257722855, "learning_rate": 1.5217799166750145e-09, "loss": 0.0102, "step": 14488 }, { "epoch": 0.9956706981858163, "grad_norm": 0.3055325150489807, "learning_rate": 1.4745965956763919e-09, "loss": 0.0078, "step": 14489 }, { "epoch": 0.995739417262232, "grad_norm": 0.1244020015001297, "learning_rate": 1.4281562707924023e-09, "loss": 0.0029, "step": 14490 }, { "epoch": 0.9958081363386476, "grad_norm": 0.5338013172149658, "learning_rate": 1.3824589443212077e-09, "loss": 0.0104, "step": 14491 }, { "epoch": 0.9958768554150632, "grad_norm": 0.121617890894413, "learning_rate": 1.3375046185276629e-09, "loss": 0.0056, "step": 14492 }, { "epoch": 0.9959455744914788, "grad_norm": 0.5239899754524231, "learning_rate": 1.2932932956383203e-09, "loss": 0.0139, "step": 14493 }, { "epoch": 0.9960142935678944, "grad_norm": 0.12996205687522888, "learning_rate": 1.2498249778414294e-09, "loss": 0.0056, "step": 14494 }, { "epoch": 0.9960830126443101, "grad_norm": 0.1530291587114334, "learning_rate": 1.2070996672935985e-09, "loss": 0.0062, "step": 14495 }, { "epoch": 0.9961517317207257, "grad_norm": 0.3903881311416626, "learning_rate": 1.1651173661098025e-09, "loss": 0.0099, "step": 14496 }, { "epoch": 0.9962204507971413, "grad_norm": 0.25936391949653625, "learning_rate": 1.123878076370044e-09, "loss": 0.0059, "step": 14497 }, { "epoch": 0.9962891698735569, "grad_norm": 0.4300267696380615, "learning_rate": 1.0833818001160234e-09, "loss": 0.0101, "step": 14498 }, { "epoch": 0.9963578889499725, "grad_norm": 0.6984431743621826, "learning_rate": 1.043628539356134e-09, "loss": 0.0128, "step": 14499 }, { "epoch": 0.9964266080263882, "grad_norm": 0.5273237824440002, "learning_rate": 1.004618296057136e-09, "loss": 0.0081, "step": 14500 }, { "epoch": 0.9964953271028038, "grad_norm": 0.10392564535140991, "learning_rate": 9.663510721524827e-10, "loss": 0.0038, "step": 14501 }, { "epoch": 0.9965640461792193, "grad_norm": 0.2036413848400116, "learning_rate": 9.288268695389901e-10, "loss": 0.0041, "step": 14502 }, { "epoch": 0.9966327652556349, "grad_norm": 0.31895318627357483, "learning_rate": 8.920456900735063e-10, "loss": 0.0033, "step": 14503 }, { "epoch": 0.9967014843320505, "grad_norm": 0.19308733940124512, "learning_rate": 8.56007535581238e-10, "loss": 0.0029, "step": 14504 }, { "epoch": 0.9967702034084662, "grad_norm": 0.5487720370292664, "learning_rate": 8.20712407844093e-10, "loss": 0.0055, "step": 14505 }, { "epoch": 0.9968389224848818, "grad_norm": 0.5711811184883118, "learning_rate": 7.861603086123381e-10, "loss": 0.0143, "step": 14506 }, { "epoch": 0.9969076415612974, "grad_norm": 0.1987064927816391, "learning_rate": 7.523512395962717e-10, "loss": 0.0015, "step": 14507 }, { "epoch": 0.996976360637713, "grad_norm": 0.3324078917503357, "learning_rate": 7.192852024712204e-10, "loss": 0.0082, "step": 14508 }, { "epoch": 0.9970450797141286, "grad_norm": 0.2978596091270447, "learning_rate": 6.869621988775387e-10, "loss": 0.0087, "step": 14509 }, { "epoch": 0.9971137987905443, "grad_norm": 0.23362278938293457, "learning_rate": 6.553822304122825e-10, "loss": 0.0089, "step": 14510 }, { "epoch": 0.9971825178669599, "grad_norm": 0.3136037290096283, "learning_rate": 6.245452986441968e-10, "loss": 0.006, "step": 14511 }, { "epoch": 0.9972512369433755, "grad_norm": 0.23589277267456055, "learning_rate": 5.944514050970628e-10, "loss": 0.0048, "step": 14512 }, { "epoch": 0.9973199560197911, "grad_norm": 0.28577321767807007, "learning_rate": 5.651005512646857e-10, "loss": 0.0055, "step": 14513 }, { "epoch": 0.9973886750962067, "grad_norm": 0.34328919649124146, "learning_rate": 5.364927385975715e-10, "loss": 0.0065, "step": 14514 }, { "epoch": 0.9974573941726224, "grad_norm": 0.13081258535385132, "learning_rate": 5.08627968516251e-10, "loss": 0.0047, "step": 14515 }, { "epoch": 0.997526113249038, "grad_norm": 0.4522911608219147, "learning_rate": 4.815062424012861e-10, "loss": 0.0089, "step": 14516 }, { "epoch": 0.9975948323254535, "grad_norm": 0.5691702961921692, "learning_rate": 4.551275615932715e-10, "loss": 0.0098, "step": 14517 }, { "epoch": 0.9976635514018691, "grad_norm": 0.15578283369541168, "learning_rate": 4.294919273994946e-10, "loss": 0.0092, "step": 14518 }, { "epoch": 0.9977322704782847, "grad_norm": 0.1874019205570221, "learning_rate": 4.0459934109227105e-10, "loss": 0.0069, "step": 14519 }, { "epoch": 0.9978009895547004, "grad_norm": 0.21617119014263153, "learning_rate": 3.8044980390394834e-10, "loss": 0.0052, "step": 14520 }, { "epoch": 0.997869708631116, "grad_norm": 0.22669832408428192, "learning_rate": 3.570433170285714e-10, "loss": 0.0088, "step": 14521 }, { "epoch": 0.9979384277075316, "grad_norm": 0.4199737310409546, "learning_rate": 3.343798816285437e-10, "loss": 0.0116, "step": 14522 }, { "epoch": 0.9980071467839472, "grad_norm": 0.20458726584911346, "learning_rate": 3.124594988246354e-10, "loss": 0.0087, "step": 14523 }, { "epoch": 0.9980758658603628, "grad_norm": 0.18045039474964142, "learning_rate": 2.9128216970430997e-10, "loss": 0.0074, "step": 14524 }, { "epoch": 0.9981445849367785, "grad_norm": 0.24747410416603088, "learning_rate": 2.7084789531506284e-10, "loss": 0.0181, "step": 14525 }, { "epoch": 0.9982133040131941, "grad_norm": 0.44659850001335144, "learning_rate": 2.511566766694173e-10, "loss": 0.0108, "step": 14526 }, { "epoch": 0.9982820230896097, "grad_norm": 0.17283327877521515, "learning_rate": 2.322085147449249e-10, "loss": 0.0069, "step": 14527 }, { "epoch": 0.9983507421660253, "grad_norm": 0.2535530924797058, "learning_rate": 2.140034104775035e-10, "loss": 0.0118, "step": 14528 }, { "epoch": 0.9984194612424409, "grad_norm": 1.0457992553710938, "learning_rate": 1.9654136476976447e-10, "loss": 0.0181, "step": 14529 }, { "epoch": 0.9984881803188566, "grad_norm": 0.4604516923427582, "learning_rate": 1.798223784876818e-10, "loss": 0.0089, "step": 14530 }, { "epoch": 0.9985568993952721, "grad_norm": 0.4236147701740265, "learning_rate": 1.638464524589267e-10, "loss": 0.0113, "step": 14531 }, { "epoch": 0.9986256184716877, "grad_norm": 0.09073265641927719, "learning_rate": 1.4861358747453312e-10, "loss": 0.005, "step": 14532 }, { "epoch": 0.9986943375481033, "grad_norm": 0.40783026814460754, "learning_rate": 1.341237842888976e-10, "loss": 0.0055, "step": 14533 }, { "epoch": 0.9987630566245189, "grad_norm": 0.32299020886421204, "learning_rate": 1.2037704362144465e-10, "loss": 0.0063, "step": 14534 }, { "epoch": 0.9988317757009346, "grad_norm": 0.3452908992767334, "learning_rate": 1.0737336615163074e-10, "loss": 0.0065, "step": 14535 }, { "epoch": 0.9989004947773502, "grad_norm": 0.549677312374115, "learning_rate": 9.511275252394036e-11, "loss": 0.0142, "step": 14536 }, { "epoch": 0.9989692138537658, "grad_norm": 0.12337131798267365, "learning_rate": 8.359520334622061e-11, "loss": 0.007, "step": 14537 }, { "epoch": 0.9990379329301814, "grad_norm": 0.4231027364730835, "learning_rate": 7.282071918801591e-11, "loss": 0.0096, "step": 14538 }, { "epoch": 0.999106652006597, "grad_norm": 0.44383031129837036, "learning_rate": 6.278930058389865e-11, "loss": 0.0062, "step": 14539 }, { "epoch": 0.9991753710830127, "grad_norm": 0.5645372867584229, "learning_rate": 5.350094803180383e-11, "loss": 0.0081, "step": 14540 }, { "epoch": 0.9992440901594283, "grad_norm": 0.11077044904232025, "learning_rate": 4.495566198969847e-11, "loss": 0.0108, "step": 14541 }, { "epoch": 0.9993128092358439, "grad_norm": 0.38008296489715576, "learning_rate": 3.715344288057754e-11, "loss": 0.0094, "step": 14542 }, { "epoch": 0.9993815283122595, "grad_norm": 0.3427509367465973, "learning_rate": 3.0094291094129313e-11, "loss": 0.0077, "step": 14543 }, { "epoch": 0.999450247388675, "grad_norm": 0.1021290197968483, "learning_rate": 2.3778206978408712e-11, "loss": 0.0175, "step": 14544 }, { "epoch": 0.9995189664650908, "grad_norm": 0.24342679977416992, "learning_rate": 1.8205190844833296e-11, "loss": 0.0054, "step": 14545 }, { "epoch": 0.9995876855415063, "grad_norm": 0.48648136854171753, "learning_rate": 1.3375242971513934e-11, "loss": 0.0131, "step": 14546 }, { "epoch": 0.9996564046179219, "grad_norm": 0.19782815873622894, "learning_rate": 9.288363596593463e-12, "loss": 0.0051, "step": 14547 }, { "epoch": 0.9997251236943375, "grad_norm": 0.16356055438518524, "learning_rate": 5.944552921577362e-12, "loss": 0.0053, "step": 14548 }, { "epoch": 0.9997938427707531, "grad_norm": 0.5616030097007751, "learning_rate": 3.343811114664419e-12, "loss": 0.0106, "step": 14549 }, { "epoch": 0.9998625618471688, "grad_norm": 0.46813011169433594, "learning_rate": 1.4861383040853937e-12, "loss": 0.0064, "step": 14550 }, { "epoch": 0.9999312809235844, "grad_norm": 0.16222643852233887, "learning_rate": 3.715345814336857e-13, "loss": 0.0067, "step": 14551 }, { "epoch": 1.0, "grad_norm": 0.35146573185920715, "learning_rate": 0.0, "loss": 0.0089, "step": 14552 }, { "epoch": 1.0, "step": 14552, "total_flos": 8.822204449989591e+17, "train_loss": 0.04244820932215008, "train_runtime": 60713.6691, "train_samples_per_second": 15.34, "train_steps_per_second": 0.24 } ], "logging_steps": 1.0, "max_steps": 14552, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 200, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 8.822204449989591e+17, "train_batch_size": 8, "trial_name": null, "trial_params": null }