{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "eval_steps": 500, "global_step": 3468, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.00865051903114187, "grad_norm": 19.23911476135254, "learning_rate": 4.9913494809688585e-05, "loss": 3.5987, "step": 10 }, { "epoch": 0.01730103806228374, "grad_norm": 38.634117126464844, "learning_rate": 4.9769319492502886e-05, "loss": 3.3494, "step": 20 }, { "epoch": 0.025951557093425604, "grad_norm": 16.30203628540039, "learning_rate": 4.962514417531719e-05, "loss": 2.7171, "step": 30 }, { "epoch": 0.03460207612456748, "grad_norm": 0.0, "learning_rate": 4.949538638985006e-05, "loss": 3.1395, "step": 40 }, { "epoch": 0.04325259515570934, "grad_norm": 19.528146743774414, "learning_rate": 4.935121107266436e-05, "loss": 3.1388, "step": 50 }, { "epoch": 0.05190311418685121, "grad_norm": 44.69698715209961, "learning_rate": 4.920703575547866e-05, "loss": 3.1337, "step": 60 }, { "epoch": 0.06055363321799308, "grad_norm": 19.81340980529785, "learning_rate": 4.9077277970011535e-05, "loss": 3.2477, "step": 70 }, { "epoch": 0.06920415224913495, "grad_norm": 11.779070854187012, "learning_rate": 4.893310265282584e-05, "loss": 3.1636, "step": 80 }, { "epoch": 0.07785467128027682, "grad_norm": 15.74388599395752, "learning_rate": 4.8788927335640136e-05, "loss": 3.0162, "step": 90 }, { "epoch": 0.08650519031141868, "grad_norm": 18.48145866394043, "learning_rate": 4.864475201845444e-05, "loss": 3.1576, "step": 100 }, { "epoch": 0.09515570934256055, "grad_norm": 13.73051643371582, "learning_rate": 4.8500576701268744e-05, "loss": 2.8694, "step": 110 }, { "epoch": 0.10380622837370242, "grad_norm": 14.890731811523438, "learning_rate": 4.835640138408305e-05, "loss": 3.267, "step": 120 }, { "epoch": 0.11245674740484429, "grad_norm": 15.936467170715332, "learning_rate": 4.821222606689735e-05, "loss": 2.827, "step": 130 }, { "epoch": 0.12110726643598616, "grad_norm": 16.747299194335938, "learning_rate": 4.806805074971165e-05, "loss": 3.0362, "step": 140 }, { "epoch": 0.12975778546712802, "grad_norm": 19.833837509155273, "learning_rate": 4.792387543252596e-05, "loss": 3.0446, "step": 150 }, { "epoch": 0.1384083044982699, "grad_norm": 21.07550048828125, "learning_rate": 4.777970011534025e-05, "loss": 2.9943, "step": 160 }, { "epoch": 0.14705882352941177, "grad_norm": 0.0, "learning_rate": 4.763552479815456e-05, "loss": 2.9449, "step": 170 }, { "epoch": 0.15570934256055363, "grad_norm": 25.829200744628906, "learning_rate": 4.749134948096886e-05, "loss": 3.1063, "step": 180 }, { "epoch": 0.1643598615916955, "grad_norm": 16.745203018188477, "learning_rate": 4.734717416378317e-05, "loss": 3.0687, "step": 190 }, { "epoch": 0.17301038062283736, "grad_norm": 19.552099227905273, "learning_rate": 4.720299884659746e-05, "loss": 3.0418, "step": 200 }, { "epoch": 0.18166089965397925, "grad_norm": 24.787086486816406, "learning_rate": 4.705882352941177e-05, "loss": 3.3645, "step": 210 }, { "epoch": 0.1903114186851211, "grad_norm": 22.9111385345459, "learning_rate": 4.691464821222607e-05, "loss": 2.7289, "step": 220 }, { "epoch": 0.19896193771626297, "grad_norm": 0.0, "learning_rate": 4.6770472895040375e-05, "loss": 3.1899, "step": 230 }, { "epoch": 0.20761245674740483, "grad_norm": 17.025423049926758, "learning_rate": 4.6626297577854675e-05, "loss": 3.0219, "step": 240 }, { "epoch": 0.21626297577854672, "grad_norm": 23.269264221191406, "learning_rate": 4.6482122260668976e-05, "loss": 3.0619, "step": 250 }, { "epoch": 0.22491349480968859, "grad_norm": 12.333551406860352, "learning_rate": 4.6337946943483276e-05, "loss": 3.0275, "step": 260 }, { "epoch": 0.23356401384083045, "grad_norm": 9.938949584960938, "learning_rate": 4.619377162629758e-05, "loss": 2.9617, "step": 270 }, { "epoch": 0.2422145328719723, "grad_norm": 11.523890495300293, "learning_rate": 4.604959630911188e-05, "loss": 3.2321, "step": 280 }, { "epoch": 0.2508650519031142, "grad_norm": 11.910848617553711, "learning_rate": 4.5905420991926184e-05, "loss": 2.6359, "step": 290 }, { "epoch": 0.25951557093425603, "grad_norm": 11.915596008300781, "learning_rate": 4.576124567474049e-05, "loss": 3.018, "step": 300 }, { "epoch": 0.2681660899653979, "grad_norm": 25.279979705810547, "learning_rate": 4.5617070357554784e-05, "loss": 2.9482, "step": 310 }, { "epoch": 0.2768166089965398, "grad_norm": 19.945743560791016, "learning_rate": 4.547289504036909e-05, "loss": 3.1401, "step": 320 }, { "epoch": 0.28546712802768165, "grad_norm": 15.122772216796875, "learning_rate": 4.532871972318339e-05, "loss": 2.903, "step": 330 }, { "epoch": 0.29411764705882354, "grad_norm": 10.13607406616211, "learning_rate": 4.51845444059977e-05, "loss": 3.1216, "step": 340 }, { "epoch": 0.3027681660899654, "grad_norm": 12.250367164611816, "learning_rate": 4.504036908881199e-05, "loss": 3.1527, "step": 350 }, { "epoch": 0.31141868512110726, "grad_norm": 8.994067192077637, "learning_rate": 4.48961937716263e-05, "loss": 2.931, "step": 360 }, { "epoch": 0.32006920415224915, "grad_norm": 9.081120491027832, "learning_rate": 4.47520184544406e-05, "loss": 2.9034, "step": 370 }, { "epoch": 0.328719723183391, "grad_norm": 10.299670219421387, "learning_rate": 4.460784313725491e-05, "loss": 3.0039, "step": 380 }, { "epoch": 0.3373702422145329, "grad_norm": 53.3034553527832, "learning_rate": 4.446366782006921e-05, "loss": 2.9574, "step": 390 }, { "epoch": 0.3460207612456747, "grad_norm": 10.064446449279785, "learning_rate": 4.431949250288351e-05, "loss": 2.914, "step": 400 }, { "epoch": 0.3546712802768166, "grad_norm": 10.004849433898926, "learning_rate": 4.4175317185697815e-05, "loss": 2.9598, "step": 410 }, { "epoch": 0.3633217993079585, "grad_norm": 9.381330490112305, "learning_rate": 4.403114186851211e-05, "loss": 2.964, "step": 420 }, { "epoch": 0.3719723183391003, "grad_norm": 6.36283016204834, "learning_rate": 4.3886966551326416e-05, "loss": 2.8286, "step": 430 }, { "epoch": 0.3806228373702422, "grad_norm": 9.161887168884277, "learning_rate": 4.3742791234140716e-05, "loss": 3.1334, "step": 440 }, { "epoch": 0.3892733564013841, "grad_norm": 7.230731964111328, "learning_rate": 4.359861591695502e-05, "loss": 3.0322, "step": 450 }, { "epoch": 0.39792387543252594, "grad_norm": 17.26264762878418, "learning_rate": 4.345444059976932e-05, "loss": 3.3673, "step": 460 }, { "epoch": 0.40657439446366783, "grad_norm": 11.14217758178711, "learning_rate": 4.3310265282583624e-05, "loss": 2.9077, "step": 470 }, { "epoch": 0.41522491349480967, "grad_norm": 18.250459671020508, "learning_rate": 4.3166089965397924e-05, "loss": 2.8501, "step": 480 }, { "epoch": 0.42387543252595156, "grad_norm": 9.473771095275879, "learning_rate": 4.302191464821223e-05, "loss": 2.6472, "step": 490 }, { "epoch": 0.43252595155709345, "grad_norm": 10.694830894470215, "learning_rate": 4.287773933102653e-05, "loss": 2.9812, "step": 500 }, { "epoch": 0.4411764705882353, "grad_norm": 14.740256309509277, "learning_rate": 4.273356401384083e-05, "loss": 2.7586, "step": 510 }, { "epoch": 0.44982698961937717, "grad_norm": 13.539804458618164, "learning_rate": 4.258938869665514e-05, "loss": 2.9536, "step": 520 }, { "epoch": 0.458477508650519, "grad_norm": 9.57966423034668, "learning_rate": 4.244521337946944e-05, "loss": 2.5686, "step": 530 }, { "epoch": 0.4671280276816609, "grad_norm": 5.894380569458008, "learning_rate": 4.230103806228374e-05, "loss": 2.9487, "step": 540 }, { "epoch": 0.4757785467128028, "grad_norm": 6.372859954833984, "learning_rate": 4.215686274509804e-05, "loss": 3.2582, "step": 550 }, { "epoch": 0.4844290657439446, "grad_norm": 7.707736015319824, "learning_rate": 4.201268742791235e-05, "loss": 3.036, "step": 560 }, { "epoch": 0.4930795847750865, "grad_norm": 12.041340827941895, "learning_rate": 4.186851211072664e-05, "loss": 3.0721, "step": 570 }, { "epoch": 0.5017301038062284, "grad_norm": 7.917064666748047, "learning_rate": 4.172433679354095e-05, "loss": 2.8503, "step": 580 }, { "epoch": 0.5103806228373703, "grad_norm": 10.359112739562988, "learning_rate": 4.158016147635525e-05, "loss": 2.7367, "step": 590 }, { "epoch": 0.5190311418685121, "grad_norm": 27.387939453125, "learning_rate": 4.1435986159169555e-05, "loss": 2.6254, "step": 600 }, { "epoch": 0.527681660899654, "grad_norm": 6.714576244354248, "learning_rate": 4.1291810841983856e-05, "loss": 2.7978, "step": 610 }, { "epoch": 0.5363321799307958, "grad_norm": 9.878549575805664, "learning_rate": 4.1147635524798156e-05, "loss": 2.9624, "step": 620 }, { "epoch": 0.5449826989619377, "grad_norm": 6.7428131103515625, "learning_rate": 4.1003460207612456e-05, "loss": 2.9623, "step": 630 }, { "epoch": 0.5536332179930796, "grad_norm": 13.340039253234863, "learning_rate": 4.0859284890426763e-05, "loss": 3.0023, "step": 640 }, { "epoch": 0.5622837370242214, "grad_norm": 9.870621681213379, "learning_rate": 4.0715109573241064e-05, "loss": 2.8046, "step": 650 }, { "epoch": 0.5709342560553633, "grad_norm": 6.491202354431152, "learning_rate": 4.0570934256055364e-05, "loss": 3.1365, "step": 660 }, { "epoch": 0.5795847750865052, "grad_norm": 12.942968368530273, "learning_rate": 4.042675893886967e-05, "loss": 2.9377, "step": 670 }, { "epoch": 0.5882352941176471, "grad_norm": 7.937127590179443, "learning_rate": 4.0282583621683965e-05, "loss": 2.9007, "step": 680 }, { "epoch": 0.596885813148789, "grad_norm": 6.2800421714782715, "learning_rate": 4.013840830449827e-05, "loss": 2.9504, "step": 690 }, { "epoch": 0.6055363321799307, "grad_norm": 8.236554145812988, "learning_rate": 3.999423298731257e-05, "loss": 2.9998, "step": 700 }, { "epoch": 0.6141868512110726, "grad_norm": 6.5029215812683105, "learning_rate": 3.985005767012688e-05, "loss": 2.9348, "step": 710 }, { "epoch": 0.6228373702422145, "grad_norm": 7.156394004821777, "learning_rate": 3.970588235294117e-05, "loss": 2.5296, "step": 720 }, { "epoch": 0.6314878892733564, "grad_norm": 6.988163948059082, "learning_rate": 3.956170703575548e-05, "loss": 2.9235, "step": 730 }, { "epoch": 0.6401384083044983, "grad_norm": 4.899734020233154, "learning_rate": 3.941753171856978e-05, "loss": 3.0278, "step": 740 }, { "epoch": 0.6487889273356401, "grad_norm": 7.569077014923096, "learning_rate": 3.927335640138409e-05, "loss": 2.8099, "step": 750 }, { "epoch": 0.657439446366782, "grad_norm": 10.045110702514648, "learning_rate": 3.912918108419839e-05, "loss": 2.6677, "step": 760 }, { "epoch": 0.6660899653979239, "grad_norm": 7.325290679931641, "learning_rate": 3.898500576701269e-05, "loss": 3.009, "step": 770 }, { "epoch": 0.6747404844290658, "grad_norm": 4.040036678314209, "learning_rate": 3.8840830449826995e-05, "loss": 2.8662, "step": 780 }, { "epoch": 0.6833910034602076, "grad_norm": 24.85966682434082, "learning_rate": 3.8696655132641296e-05, "loss": 3.0819, "step": 790 }, { "epoch": 0.6920415224913494, "grad_norm": 7.573343753814697, "learning_rate": 3.8552479815455596e-05, "loss": 3.1691, "step": 800 }, { "epoch": 0.7006920415224913, "grad_norm": 5.489699840545654, "learning_rate": 3.8408304498269896e-05, "loss": 2.7598, "step": 810 }, { "epoch": 0.7093425605536332, "grad_norm": 6.8325090408325195, "learning_rate": 3.8264129181084204e-05, "loss": 2.7702, "step": 820 }, { "epoch": 0.7179930795847751, "grad_norm": 14.738582611083984, "learning_rate": 3.81199538638985e-05, "loss": 2.5193, "step": 830 }, { "epoch": 0.726643598615917, "grad_norm": 5.153214454650879, "learning_rate": 3.7975778546712804e-05, "loss": 2.6036, "step": 840 }, { "epoch": 0.7352941176470589, "grad_norm": 5.4641499519348145, "learning_rate": 3.7831603229527105e-05, "loss": 2.8178, "step": 850 }, { "epoch": 0.7439446366782007, "grad_norm": 10.48047924041748, "learning_rate": 3.768742791234141e-05, "loss": 2.4638, "step": 860 }, { "epoch": 0.7525951557093425, "grad_norm": 6.817728042602539, "learning_rate": 3.754325259515571e-05, "loss": 2.7103, "step": 870 }, { "epoch": 0.7612456747404844, "grad_norm": 11.84115982055664, "learning_rate": 3.739907727797001e-05, "loss": 2.8477, "step": 880 }, { "epoch": 0.7698961937716263, "grad_norm": 8.383696556091309, "learning_rate": 3.725490196078432e-05, "loss": 2.8378, "step": 890 }, { "epoch": 0.7785467128027682, "grad_norm": 6.298950672149658, "learning_rate": 3.711072664359862e-05, "loss": 2.968, "step": 900 }, { "epoch": 0.78719723183391, "grad_norm": 4.829587936401367, "learning_rate": 3.696655132641292e-05, "loss": 3.096, "step": 910 }, { "epoch": 0.7958477508650519, "grad_norm": 4.4047746658325195, "learning_rate": 3.682237600922722e-05, "loss": 2.7938, "step": 920 }, { "epoch": 0.8044982698961938, "grad_norm": 6.090274810791016, "learning_rate": 3.667820069204153e-05, "loss": 2.6475, "step": 930 }, { "epoch": 0.8131487889273357, "grad_norm": 6.233980178833008, "learning_rate": 3.653402537485582e-05, "loss": 3.033, "step": 940 }, { "epoch": 0.8217993079584776, "grad_norm": 6.986254692077637, "learning_rate": 3.638985005767013e-05, "loss": 2.8964, "step": 950 }, { "epoch": 0.8304498269896193, "grad_norm": 6.642579555511475, "learning_rate": 3.624567474048443e-05, "loss": 2.4588, "step": 960 }, { "epoch": 0.8391003460207612, "grad_norm": 4.945983409881592, "learning_rate": 3.6101499423298736e-05, "loss": 2.5413, "step": 970 }, { "epoch": 0.8477508650519031, "grad_norm": 5.664891719818115, "learning_rate": 3.5957324106113036e-05, "loss": 2.6916, "step": 980 }, { "epoch": 0.856401384083045, "grad_norm": 6.726191520690918, "learning_rate": 3.5813148788927336e-05, "loss": 2.7688, "step": 990 }, { "epoch": 0.8650519031141869, "grad_norm": 7.181143760681152, "learning_rate": 3.566897347174164e-05, "loss": 2.8977, "step": 1000 }, { "epoch": 0.8737024221453287, "grad_norm": 6.649637222290039, "learning_rate": 3.5524798154555944e-05, "loss": 2.9901, "step": 1010 }, { "epoch": 0.8823529411764706, "grad_norm": 0.0, "learning_rate": 3.5380622837370244e-05, "loss": 2.6497, "step": 1020 }, { "epoch": 0.8910034602076125, "grad_norm": 8.609131813049316, "learning_rate": 3.5236447520184545e-05, "loss": 3.1123, "step": 1030 }, { "epoch": 0.8996539792387543, "grad_norm": 6.280732154846191, "learning_rate": 3.509227220299885e-05, "loss": 3.0722, "step": 1040 }, { "epoch": 0.9083044982698962, "grad_norm": 5.108139991760254, "learning_rate": 3.494809688581315e-05, "loss": 2.8024, "step": 1050 }, { "epoch": 0.916955017301038, "grad_norm": 6.271091461181641, "learning_rate": 3.480392156862745e-05, "loss": 3.0201, "step": 1060 }, { "epoch": 0.9256055363321799, "grad_norm": 6.791084289550781, "learning_rate": 3.465974625144175e-05, "loss": 2.7122, "step": 1070 }, { "epoch": 0.9342560553633218, "grad_norm": 14.90443229675293, "learning_rate": 3.451557093425606e-05, "loss": 2.7771, "step": 1080 }, { "epoch": 0.9429065743944637, "grad_norm": 7.935113906860352, "learning_rate": 3.4371395617070353e-05, "loss": 2.8365, "step": 1090 }, { "epoch": 0.9515570934256056, "grad_norm": 10.18995189666748, "learning_rate": 3.422722029988466e-05, "loss": 2.8855, "step": 1100 }, { "epoch": 0.9602076124567474, "grad_norm": 7.072559356689453, "learning_rate": 3.408304498269896e-05, "loss": 2.7104, "step": 1110 }, { "epoch": 0.9688581314878892, "grad_norm": 4.9115400314331055, "learning_rate": 3.393886966551327e-05, "loss": 2.847, "step": 1120 }, { "epoch": 0.9775086505190311, "grad_norm": 6.05402135848999, "learning_rate": 3.379469434832757e-05, "loss": 2.4883, "step": 1130 }, { "epoch": 0.986159169550173, "grad_norm": 6.126327991485596, "learning_rate": 3.365051903114187e-05, "loss": 2.7995, "step": 1140 }, { "epoch": 0.9948096885813149, "grad_norm": 8.214347839355469, "learning_rate": 3.3506343713956176e-05, "loss": 2.8806, "step": 1150 }, { "epoch": 1.0034602076124568, "grad_norm": 6.953145980834961, "learning_rate": 3.3362168396770476e-05, "loss": 2.6513, "step": 1160 }, { "epoch": 1.0121107266435987, "grad_norm": 7.279242992401123, "learning_rate": 3.3217993079584777e-05, "loss": 2.5831, "step": 1170 }, { "epoch": 1.0207612456747406, "grad_norm": 6.322566032409668, "learning_rate": 3.307381776239908e-05, "loss": 2.5757, "step": 1180 }, { "epoch": 1.0294117647058822, "grad_norm": 5.4494547843933105, "learning_rate": 3.2929642445213384e-05, "loss": 2.3202, "step": 1190 }, { "epoch": 1.0380622837370241, "grad_norm": 0.0, "learning_rate": 3.2785467128027684e-05, "loss": 2.1757, "step": 1200 }, { "epoch": 1.046712802768166, "grad_norm": 6.834903717041016, "learning_rate": 3.2641291810841985e-05, "loss": 2.4916, "step": 1210 }, { "epoch": 1.055363321799308, "grad_norm": 7.15425443649292, "learning_rate": 3.2497116493656285e-05, "loss": 2.5786, "step": 1220 }, { "epoch": 1.0640138408304498, "grad_norm": 25.43375587463379, "learning_rate": 3.235294117647059e-05, "loss": 2.457, "step": 1230 }, { "epoch": 1.0726643598615917, "grad_norm": 6.561356067657471, "learning_rate": 3.220876585928489e-05, "loss": 2.3607, "step": 1240 }, { "epoch": 1.0813148788927336, "grad_norm": 11.72268009185791, "learning_rate": 3.206459054209919e-05, "loss": 2.3018, "step": 1250 }, { "epoch": 1.0899653979238755, "grad_norm": 0.0, "learning_rate": 3.19204152249135e-05, "loss": 2.4857, "step": 1260 }, { "epoch": 1.0986159169550174, "grad_norm": 5.507035255432129, "learning_rate": 3.17762399077278e-05, "loss": 2.4546, "step": 1270 }, { "epoch": 1.1072664359861593, "grad_norm": 8.205038070678711, "learning_rate": 3.16320645905421e-05, "loss": 2.5896, "step": 1280 }, { "epoch": 1.1159169550173011, "grad_norm": 6.336702346801758, "learning_rate": 3.14878892733564e-05, "loss": 2.6643, "step": 1290 }, { "epoch": 1.1245674740484428, "grad_norm": 6.227362155914307, "learning_rate": 3.134371395617071e-05, "loss": 2.4297, "step": 1300 }, { "epoch": 1.1332179930795847, "grad_norm": 7.871458053588867, "learning_rate": 3.119953863898501e-05, "loss": 2.5482, "step": 1310 }, { "epoch": 1.1418685121107266, "grad_norm": 5.489798545837402, "learning_rate": 3.105536332179931e-05, "loss": 2.3449, "step": 1320 }, { "epoch": 1.1505190311418685, "grad_norm": 4.873505592346191, "learning_rate": 3.091118800461361e-05, "loss": 2.3388, "step": 1330 }, { "epoch": 1.1591695501730104, "grad_norm": 7.125423431396484, "learning_rate": 3.0767012687427916e-05, "loss": 2.3973, "step": 1340 }, { "epoch": 1.1678200692041523, "grad_norm": 8.925414085388184, "learning_rate": 3.062283737024222e-05, "loss": 2.3922, "step": 1350 }, { "epoch": 1.1764705882352942, "grad_norm": 11.737508773803711, "learning_rate": 3.0478662053056517e-05, "loss": 2.4235, "step": 1360 }, { "epoch": 1.185121107266436, "grad_norm": 7.375851154327393, "learning_rate": 3.033448673587082e-05, "loss": 2.3064, "step": 1370 }, { "epoch": 1.193771626297578, "grad_norm": 9.243279457092285, "learning_rate": 3.0190311418685124e-05, "loss": 2.2488, "step": 1380 }, { "epoch": 1.2024221453287196, "grad_norm": 8.473196983337402, "learning_rate": 3.004613610149942e-05, "loss": 2.4362, "step": 1390 }, { "epoch": 1.2110726643598615, "grad_norm": 4.9002790451049805, "learning_rate": 2.9901960784313725e-05, "loss": 2.5061, "step": 1400 }, { "epoch": 1.2197231833910034, "grad_norm": 8.8832368850708, "learning_rate": 2.975778546712803e-05, "loss": 2.3609, "step": 1410 }, { "epoch": 1.2283737024221453, "grad_norm": 9.568990707397461, "learning_rate": 2.9613610149942333e-05, "loss": 2.4853, "step": 1420 }, { "epoch": 1.2370242214532872, "grad_norm": 10.570289611816406, "learning_rate": 2.9469434832756633e-05, "loss": 2.3944, "step": 1430 }, { "epoch": 1.245674740484429, "grad_norm": 10.139781951904297, "learning_rate": 2.9325259515570937e-05, "loss": 2.4187, "step": 1440 }, { "epoch": 1.254325259515571, "grad_norm": 12.260625839233398, "learning_rate": 2.918108419838524e-05, "loss": 2.3519, "step": 1450 }, { "epoch": 1.2629757785467128, "grad_norm": 7.253377437591553, "learning_rate": 2.9036908881199544e-05, "loss": 2.4657, "step": 1460 }, { "epoch": 1.2716262975778547, "grad_norm": 4.9923553466796875, "learning_rate": 2.889273356401384e-05, "loss": 2.4163, "step": 1470 }, { "epoch": 1.2802768166089966, "grad_norm": 9.419482231140137, "learning_rate": 2.8748558246828145e-05, "loss": 2.3895, "step": 1480 }, { "epoch": 1.2889273356401385, "grad_norm": 5.093005180358887, "learning_rate": 2.860438292964245e-05, "loss": 2.3857, "step": 1490 }, { "epoch": 1.2975778546712804, "grad_norm": 14.546469688415527, "learning_rate": 2.8460207612456746e-05, "loss": 2.3892, "step": 1500 }, { "epoch": 1.306228373702422, "grad_norm": 4.272151470184326, "learning_rate": 2.831603229527105e-05, "loss": 2.7683, "step": 1510 }, { "epoch": 1.314878892733564, "grad_norm": 6.617842197418213, "learning_rate": 2.8171856978085353e-05, "loss": 2.3034, "step": 1520 }, { "epoch": 1.3235294117647058, "grad_norm": 5.439128398895264, "learning_rate": 2.8027681660899657e-05, "loss": 2.6287, "step": 1530 }, { "epoch": 1.3321799307958477, "grad_norm": 6.48740816116333, "learning_rate": 2.7883506343713957e-05, "loss": 2.5707, "step": 1540 }, { "epoch": 1.3408304498269896, "grad_norm": 6.10065221786499, "learning_rate": 2.773933102652826e-05, "loss": 2.3402, "step": 1550 }, { "epoch": 1.3494809688581315, "grad_norm": 5.562924861907959, "learning_rate": 2.759515570934256e-05, "loss": 2.3861, "step": 1560 }, { "epoch": 1.3581314878892734, "grad_norm": 8.798916816711426, "learning_rate": 2.7450980392156865e-05, "loss": 2.3875, "step": 1570 }, { "epoch": 1.3667820069204153, "grad_norm": 13.19672679901123, "learning_rate": 2.7306805074971165e-05, "loss": 2.1105, "step": 1580 }, { "epoch": 1.375432525951557, "grad_norm": 7.34409236907959, "learning_rate": 2.716262975778547e-05, "loss": 2.3937, "step": 1590 }, { "epoch": 1.3840830449826989, "grad_norm": 6.497256278991699, "learning_rate": 2.7018454440599773e-05, "loss": 2.3291, "step": 1600 }, { "epoch": 1.3927335640138407, "grad_norm": 4.751974582672119, "learning_rate": 2.687427912341407e-05, "loss": 2.528, "step": 1610 }, { "epoch": 1.4013840830449826, "grad_norm": 12.393454551696777, "learning_rate": 2.6730103806228373e-05, "loss": 2.5473, "step": 1620 }, { "epoch": 1.4100346020761245, "grad_norm": 5.408919811248779, "learning_rate": 2.6585928489042677e-05, "loss": 2.3739, "step": 1630 }, { "epoch": 1.4186851211072664, "grad_norm": 6.614011287689209, "learning_rate": 2.644175317185698e-05, "loss": 2.3951, "step": 1640 }, { "epoch": 1.4273356401384083, "grad_norm": 6.63184928894043, "learning_rate": 2.6297577854671278e-05, "loss": 2.3318, "step": 1650 }, { "epoch": 1.4359861591695502, "grad_norm": 6.916550636291504, "learning_rate": 2.615340253748558e-05, "loss": 2.5457, "step": 1660 }, { "epoch": 1.444636678200692, "grad_norm": 7.225638389587402, "learning_rate": 2.6009227220299885e-05, "loss": 2.4932, "step": 1670 }, { "epoch": 1.453287197231834, "grad_norm": 7.826704978942871, "learning_rate": 2.586505190311419e-05, "loss": 2.3264, "step": 1680 }, { "epoch": 1.4619377162629759, "grad_norm": 12.383480072021484, "learning_rate": 2.572087658592849e-05, "loss": 2.3894, "step": 1690 }, { "epoch": 1.4705882352941178, "grad_norm": 6.668808937072754, "learning_rate": 2.5576701268742793e-05, "loss": 2.4778, "step": 1700 }, { "epoch": 1.4792387543252596, "grad_norm": 7.074410915374756, "learning_rate": 2.5432525951557097e-05, "loss": 2.166, "step": 1710 }, { "epoch": 1.4878892733564013, "grad_norm": 9.129437446594238, "learning_rate": 2.52883506343714e-05, "loss": 2.2821, "step": 1720 }, { "epoch": 1.4965397923875432, "grad_norm": 7.847931385040283, "learning_rate": 2.5144175317185697e-05, "loss": 2.2413, "step": 1730 }, { "epoch": 1.505190311418685, "grad_norm": 6.493180274963379, "learning_rate": 2.5e-05, "loss": 2.3985, "step": 1740 }, { "epoch": 1.513840830449827, "grad_norm": 8.56859016418457, "learning_rate": 2.48558246828143e-05, "loss": 2.4284, "step": 1750 }, { "epoch": 1.5224913494809689, "grad_norm": 0.0, "learning_rate": 2.4711649365628605e-05, "loss": 2.5876, "step": 1760 }, { "epoch": 1.5311418685121108, "grad_norm": 7.028433799743652, "learning_rate": 2.4567474048442906e-05, "loss": 2.4459, "step": 1770 }, { "epoch": 1.5397923875432526, "grad_norm": 5.476517200469971, "learning_rate": 2.442329873125721e-05, "loss": 2.3663, "step": 1780 }, { "epoch": 1.5484429065743943, "grad_norm": 7.162520408630371, "learning_rate": 2.427912341407151e-05, "loss": 2.2765, "step": 1790 }, { "epoch": 1.5570934256055362, "grad_norm": 13.420760154724121, "learning_rate": 2.4134948096885813e-05, "loss": 2.2691, "step": 1800 }, { "epoch": 1.565743944636678, "grad_norm": 7.58514928817749, "learning_rate": 2.3990772779700117e-05, "loss": 2.4497, "step": 1810 }, { "epoch": 1.57439446366782, "grad_norm": 5.244626998901367, "learning_rate": 2.384659746251442e-05, "loss": 2.2665, "step": 1820 }, { "epoch": 1.5830449826989619, "grad_norm": 8.593738555908203, "learning_rate": 2.370242214532872e-05, "loss": 2.5301, "step": 1830 }, { "epoch": 1.5916955017301038, "grad_norm": 5.1958231925964355, "learning_rate": 2.3558246828143025e-05, "loss": 2.2558, "step": 1840 }, { "epoch": 1.6003460207612457, "grad_norm": 5.891276836395264, "learning_rate": 2.3414071510957325e-05, "loss": 2.437, "step": 1850 }, { "epoch": 1.6089965397923875, "grad_norm": 8.773421287536621, "learning_rate": 2.326989619377163e-05, "loss": 2.2866, "step": 1860 }, { "epoch": 1.6176470588235294, "grad_norm": 5.609074592590332, "learning_rate": 2.312572087658593e-05, "loss": 2.3159, "step": 1870 }, { "epoch": 1.6262975778546713, "grad_norm": 7.042026519775391, "learning_rate": 2.298154555940023e-05, "loss": 2.4329, "step": 1880 }, { "epoch": 1.6349480968858132, "grad_norm": 5.367044925689697, "learning_rate": 2.2837370242214533e-05, "loss": 2.2781, "step": 1890 }, { "epoch": 1.643598615916955, "grad_norm": 5.378485202789307, "learning_rate": 2.2693194925028834e-05, "loss": 2.3279, "step": 1900 }, { "epoch": 1.652249134948097, "grad_norm": 8.597953796386719, "learning_rate": 2.2549019607843138e-05, "loss": 2.5955, "step": 1910 }, { "epoch": 1.6608996539792389, "grad_norm": 7.4986572265625, "learning_rate": 2.240484429065744e-05, "loss": 2.25, "step": 1920 }, { "epoch": 1.6695501730103808, "grad_norm": 12.225547790527344, "learning_rate": 2.226066897347174e-05, "loss": 2.1325, "step": 1930 }, { "epoch": 1.6782006920415224, "grad_norm": 25.887487411499023, "learning_rate": 2.2116493656286045e-05, "loss": 2.3929, "step": 1940 }, { "epoch": 1.6868512110726643, "grad_norm": 13.56473445892334, "learning_rate": 2.197231833910035e-05, "loss": 2.3572, "step": 1950 }, { "epoch": 1.6955017301038062, "grad_norm": 10.653745651245117, "learning_rate": 2.182814302191465e-05, "loss": 2.4658, "step": 1960 }, { "epoch": 1.7041522491349481, "grad_norm": 9.479430198669434, "learning_rate": 2.1683967704728953e-05, "loss": 2.4963, "step": 1970 }, { "epoch": 1.71280276816609, "grad_norm": 12.149545669555664, "learning_rate": 2.1539792387543253e-05, "loss": 2.2464, "step": 1980 }, { "epoch": 1.7214532871972317, "grad_norm": 8.545853614807129, "learning_rate": 2.1395617070357557e-05, "loss": 2.3789, "step": 1990 }, { "epoch": 1.7301038062283736, "grad_norm": 10.555283546447754, "learning_rate": 2.1251441753171858e-05, "loss": 2.2933, "step": 2000 }, { "epoch": 1.7387543252595155, "grad_norm": 10.031829833984375, "learning_rate": 2.1107266435986158e-05, "loss": 2.3333, "step": 2010 }, { "epoch": 1.7474048442906573, "grad_norm": 0.0, "learning_rate": 2.096309111880046e-05, "loss": 2.7103, "step": 2020 }, { "epoch": 1.7560553633217992, "grad_norm": 7.1533522605896, "learning_rate": 2.0818915801614762e-05, "loss": 2.1076, "step": 2030 }, { "epoch": 1.7647058823529411, "grad_norm": 11.710586547851562, "learning_rate": 2.0674740484429066e-05, "loss": 2.352, "step": 2040 }, { "epoch": 1.773356401384083, "grad_norm": 10.389070510864258, "learning_rate": 2.053056516724337e-05, "loss": 2.5781, "step": 2050 }, { "epoch": 1.782006920415225, "grad_norm": 10.627307891845703, "learning_rate": 2.0386389850057673e-05, "loss": 2.4335, "step": 2060 }, { "epoch": 1.7906574394463668, "grad_norm": 7.498691082000732, "learning_rate": 2.0242214532871974e-05, "loss": 2.4912, "step": 2070 }, { "epoch": 1.7993079584775087, "grad_norm": 7.602413654327393, "learning_rate": 2.0098039215686277e-05, "loss": 2.396, "step": 2080 }, { "epoch": 1.8079584775086506, "grad_norm": 5.651869297027588, "learning_rate": 1.9953863898500578e-05, "loss": 2.2124, "step": 2090 }, { "epoch": 1.8166089965397925, "grad_norm": 5.149964809417725, "learning_rate": 1.980968858131488e-05, "loss": 2.47, "step": 2100 }, { "epoch": 1.8252595155709344, "grad_norm": 5.61548376083374, "learning_rate": 1.966551326412918e-05, "loss": 2.3634, "step": 2110 }, { "epoch": 1.8339100346020762, "grad_norm": 6.321987628936768, "learning_rate": 1.9521337946943485e-05, "loss": 2.3837, "step": 2120 }, { "epoch": 1.8425605536332181, "grad_norm": 17.681230545043945, "learning_rate": 1.9377162629757786e-05, "loss": 2.5887, "step": 2130 }, { "epoch": 1.85121107266436, "grad_norm": 9.560349464416504, "learning_rate": 1.923298731257209e-05, "loss": 2.6304, "step": 2140 }, { "epoch": 1.8598615916955017, "grad_norm": 4.678035736083984, "learning_rate": 1.908881199538639e-05, "loss": 2.3547, "step": 2150 }, { "epoch": 1.8685121107266436, "grad_norm": 6.619549751281738, "learning_rate": 1.894463667820069e-05, "loss": 2.4747, "step": 2160 }, { "epoch": 1.8771626297577855, "grad_norm": 12.084946632385254, "learning_rate": 1.8800461361014994e-05, "loss": 2.46, "step": 2170 }, { "epoch": 1.8858131487889274, "grad_norm": 8.923602104187012, "learning_rate": 1.8656286043829298e-05, "loss": 2.3937, "step": 2180 }, { "epoch": 1.8944636678200693, "grad_norm": 9.561132431030273, "learning_rate": 1.85121107266436e-05, "loss": 2.1683, "step": 2190 }, { "epoch": 1.903114186851211, "grad_norm": 6.697421073913574, "learning_rate": 1.8367935409457902e-05, "loss": 2.3719, "step": 2200 }, { "epoch": 1.9117647058823528, "grad_norm": 6.0673322677612305, "learning_rate": 1.8223760092272205e-05, "loss": 2.1131, "step": 2210 }, { "epoch": 1.9204152249134947, "grad_norm": 6.221293926239014, "learning_rate": 1.8079584775086506e-05, "loss": 2.3373, "step": 2220 }, { "epoch": 1.9290657439446366, "grad_norm": 19.771453857421875, "learning_rate": 1.793540945790081e-05, "loss": 2.1739, "step": 2230 }, { "epoch": 1.9377162629757785, "grad_norm": 4.692147731781006, "learning_rate": 1.779123414071511e-05, "loss": 2.3212, "step": 2240 }, { "epoch": 1.9463667820069204, "grad_norm": 13.161782264709473, "learning_rate": 1.7647058823529414e-05, "loss": 2.3698, "step": 2250 }, { "epoch": 1.9550173010380623, "grad_norm": 7.448920726776123, "learning_rate": 1.7502883506343714e-05, "loss": 2.0373, "step": 2260 }, { "epoch": 1.9636678200692042, "grad_norm": 8.207565307617188, "learning_rate": 1.7358708189158018e-05, "loss": 2.4333, "step": 2270 }, { "epoch": 1.972318339100346, "grad_norm": 7.432470798492432, "learning_rate": 1.7214532871972318e-05, "loss": 2.1247, "step": 2280 }, { "epoch": 1.980968858131488, "grad_norm": 8.950566291809082, "learning_rate": 1.7070357554786622e-05, "loss": 2.2152, "step": 2290 }, { "epoch": 1.9896193771626298, "grad_norm": 8.370878219604492, "learning_rate": 1.6926182237600925e-05, "loss": 2.3552, "step": 2300 }, { "epoch": 1.9982698961937717, "grad_norm": 9.798127174377441, "learning_rate": 1.6782006920415226e-05, "loss": 2.5178, "step": 2310 }, { "epoch": 2.0069204152249136, "grad_norm": 4.229007244110107, "learning_rate": 1.663783160322953e-05, "loss": 2.1752, "step": 2320 }, { "epoch": 2.0155709342560555, "grad_norm": 6.5957255363464355, "learning_rate": 1.649365628604383e-05, "loss": 2.2877, "step": 2330 }, { "epoch": 2.0242214532871974, "grad_norm": 6.140663146972656, "learning_rate": 1.6349480968858134e-05, "loss": 2.2057, "step": 2340 }, { "epoch": 2.0328719723183393, "grad_norm": 11.750750541687012, "learning_rate": 1.6205305651672434e-05, "loss": 2.1969, "step": 2350 }, { "epoch": 2.041522491349481, "grad_norm": 8.632787704467773, "learning_rate": 1.6061130334486738e-05, "loss": 2.1248, "step": 2360 }, { "epoch": 2.050173010380623, "grad_norm": 13.063236236572266, "learning_rate": 1.5916955017301038e-05, "loss": 2.056, "step": 2370 }, { "epoch": 2.0588235294117645, "grad_norm": 16.210952758789062, "learning_rate": 1.5772779700115342e-05, "loss": 1.8912, "step": 2380 }, { "epoch": 2.0674740484429064, "grad_norm": 13.75781536102295, "learning_rate": 1.5628604382929642e-05, "loss": 2.0829, "step": 2390 }, { "epoch": 2.0761245674740483, "grad_norm": 6.415799617767334, "learning_rate": 1.5484429065743946e-05, "loss": 2.0569, "step": 2400 }, { "epoch": 2.08477508650519, "grad_norm": 7.936450004577637, "learning_rate": 1.5340253748558246e-05, "loss": 2.1501, "step": 2410 }, { "epoch": 2.093425605536332, "grad_norm": 7.677123069763184, "learning_rate": 1.5196078431372548e-05, "loss": 1.9873, "step": 2420 }, { "epoch": 2.102076124567474, "grad_norm": 7.399821758270264, "learning_rate": 1.5051903114186852e-05, "loss": 2.078, "step": 2430 }, { "epoch": 2.110726643598616, "grad_norm": 9.003458976745605, "learning_rate": 1.4907727797001152e-05, "loss": 1.9666, "step": 2440 }, { "epoch": 2.1193771626297577, "grad_norm": 5.406769752502441, "learning_rate": 1.4763552479815456e-05, "loss": 2.1621, "step": 2450 }, { "epoch": 2.1280276816608996, "grad_norm": 9.71753978729248, "learning_rate": 1.4619377162629758e-05, "loss": 2.3213, "step": 2460 }, { "epoch": 2.1366782006920415, "grad_norm": 7.547745227813721, "learning_rate": 1.4475201845444062e-05, "loss": 1.981, "step": 2470 }, { "epoch": 2.1453287197231834, "grad_norm": 6.105127811431885, "learning_rate": 1.4331026528258362e-05, "loss": 2.0367, "step": 2480 }, { "epoch": 2.1539792387543253, "grad_norm": 19.412294387817383, "learning_rate": 1.4186851211072666e-05, "loss": 2.0869, "step": 2490 }, { "epoch": 2.162629757785467, "grad_norm": 7.70440149307251, "learning_rate": 1.4042675893886966e-05, "loss": 2.1693, "step": 2500 }, { "epoch": 2.171280276816609, "grad_norm": 7.652112007141113, "learning_rate": 1.389850057670127e-05, "loss": 2.1045, "step": 2510 }, { "epoch": 2.179930795847751, "grad_norm": 11.779850959777832, "learning_rate": 1.3754325259515572e-05, "loss": 2.0855, "step": 2520 }, { "epoch": 2.188581314878893, "grad_norm": 0.0, "learning_rate": 1.3610149942329876e-05, "loss": 2.1312, "step": 2530 }, { "epoch": 2.1972318339100347, "grad_norm": 6.408977031707764, "learning_rate": 1.3465974625144176e-05, "loss": 2.1698, "step": 2540 }, { "epoch": 2.2058823529411766, "grad_norm": 8.69074821472168, "learning_rate": 1.3321799307958476e-05, "loss": 2.0696, "step": 2550 }, { "epoch": 2.2145328719723185, "grad_norm": 8.960659980773926, "learning_rate": 1.317762399077278e-05, "loss": 1.73, "step": 2560 }, { "epoch": 2.2231833910034604, "grad_norm": 8.215168952941895, "learning_rate": 1.3033448673587082e-05, "loss": 2.0223, "step": 2570 }, { "epoch": 2.2318339100346023, "grad_norm": 6.832982063293457, "learning_rate": 1.2889273356401386e-05, "loss": 2.2219, "step": 2580 }, { "epoch": 2.2404844290657437, "grad_norm": 15.58371639251709, "learning_rate": 1.2745098039215686e-05, "loss": 2.1099, "step": 2590 }, { "epoch": 2.2491349480968856, "grad_norm": 5.2607197761535645, "learning_rate": 1.260092272202999e-05, "loss": 2.1367, "step": 2600 }, { "epoch": 2.2577854671280275, "grad_norm": 5.739188194274902, "learning_rate": 1.245674740484429e-05, "loss": 2.2481, "step": 2610 }, { "epoch": 2.2664359861591694, "grad_norm": 0.0, "learning_rate": 1.2312572087658592e-05, "loss": 2.1261, "step": 2620 }, { "epoch": 2.2750865051903113, "grad_norm": 15.587763786315918, "learning_rate": 1.2168396770472894e-05, "loss": 2.301, "step": 2630 }, { "epoch": 2.283737024221453, "grad_norm": 4.138315200805664, "learning_rate": 1.2024221453287198e-05, "loss": 2.2832, "step": 2640 }, { "epoch": 2.292387543252595, "grad_norm": 7.0834784507751465, "learning_rate": 1.18800461361015e-05, "loss": 2.2382, "step": 2650 }, { "epoch": 2.301038062283737, "grad_norm": 7.361220359802246, "learning_rate": 1.1735870818915802e-05, "loss": 2.0817, "step": 2660 }, { "epoch": 2.309688581314879, "grad_norm": 15.36158275604248, "learning_rate": 1.1591695501730104e-05, "loss": 1.8927, "step": 2670 }, { "epoch": 2.3183391003460208, "grad_norm": 5.309439659118652, "learning_rate": 1.1447520184544406e-05, "loss": 1.9435, "step": 2680 }, { "epoch": 2.3269896193771626, "grad_norm": 0.0, "learning_rate": 1.1303344867358708e-05, "loss": 1.9827, "step": 2690 }, { "epoch": 2.3356401384083045, "grad_norm": 5.341144561767578, "learning_rate": 1.115916955017301e-05, "loss": 2.2153, "step": 2700 }, { "epoch": 2.3442906574394464, "grad_norm": 7.391653537750244, "learning_rate": 1.1014994232987314e-05, "loss": 2.0618, "step": 2710 }, { "epoch": 2.3529411764705883, "grad_norm": 11.139851570129395, "learning_rate": 1.0870818915801616e-05, "loss": 2.02, "step": 2720 }, { "epoch": 2.36159169550173, "grad_norm": 6.582757472991943, "learning_rate": 1.0726643598615918e-05, "loss": 2.2848, "step": 2730 }, { "epoch": 2.370242214532872, "grad_norm": 8.014312744140625, "learning_rate": 1.058246828143022e-05, "loss": 2.3369, "step": 2740 }, { "epoch": 2.378892733564014, "grad_norm": 17.67845344543457, "learning_rate": 1.043829296424452e-05, "loss": 2.1467, "step": 2750 }, { "epoch": 2.387543252595156, "grad_norm": 4.640291690826416, "learning_rate": 1.0294117647058824e-05, "loss": 2.2248, "step": 2760 }, { "epoch": 2.3961937716262978, "grad_norm": 10.063308715820312, "learning_rate": 1.0149942329873126e-05, "loss": 2.1502, "step": 2770 }, { "epoch": 2.404844290657439, "grad_norm": 10.428187370300293, "learning_rate": 1.0005767012687428e-05, "loss": 2.1736, "step": 2780 }, { "epoch": 2.413494809688581, "grad_norm": 5.396498680114746, "learning_rate": 9.86159169550173e-06, "loss": 1.9383, "step": 2790 }, { "epoch": 2.422145328719723, "grad_norm": 0.0, "learning_rate": 9.717416378316032e-06, "loss": 1.9406, "step": 2800 }, { "epoch": 2.430795847750865, "grad_norm": 8.664689064025879, "learning_rate": 9.573241061130335e-06, "loss": 2.3603, "step": 2810 }, { "epoch": 2.4394463667820068, "grad_norm": 13.918140411376953, "learning_rate": 9.429065743944637e-06, "loss": 2.2744, "step": 2820 }, { "epoch": 2.4480968858131487, "grad_norm": 0.0, "learning_rate": 9.28489042675894e-06, "loss": 2.1881, "step": 2830 }, { "epoch": 2.4567474048442905, "grad_norm": 5.737022876739502, "learning_rate": 9.140715109573242e-06, "loss": 2.0455, "step": 2840 }, { "epoch": 2.4653979238754324, "grad_norm": 6.97428560256958, "learning_rate": 8.996539792387544e-06, "loss": 1.9571, "step": 2850 }, { "epoch": 2.4740484429065743, "grad_norm": 5.972537994384766, "learning_rate": 8.852364475201846e-06, "loss": 2.2661, "step": 2860 }, { "epoch": 2.482698961937716, "grad_norm": 8.718277931213379, "learning_rate": 8.708189158016148e-06, "loss": 2.1354, "step": 2870 }, { "epoch": 2.491349480968858, "grad_norm": 15.07697868347168, "learning_rate": 8.564013840830449e-06, "loss": 2.1308, "step": 2880 }, { "epoch": 2.5, "grad_norm": 13.866670608520508, "learning_rate": 8.419838523644753e-06, "loss": 2.1775, "step": 2890 }, { "epoch": 2.508650519031142, "grad_norm": 13.32663631439209, "learning_rate": 8.275663206459055e-06, "loss": 1.7741, "step": 2900 }, { "epoch": 2.517301038062284, "grad_norm": 0.0, "learning_rate": 8.131487889273357e-06, "loss": 1.8375, "step": 2910 }, { "epoch": 2.5259515570934257, "grad_norm": 8.219161987304688, "learning_rate": 7.987312572087659e-06, "loss": 2.088, "step": 2920 }, { "epoch": 2.5346020761245676, "grad_norm": 6.607839107513428, "learning_rate": 7.84313725490196e-06, "loss": 2.2074, "step": 2930 }, { "epoch": 2.5432525951557095, "grad_norm": 5.926668643951416, "learning_rate": 7.698961937716263e-06, "loss": 2.0621, "step": 2940 }, { "epoch": 2.5519031141868513, "grad_norm": 14.596614837646484, "learning_rate": 7.554786620530566e-06, "loss": 2.1123, "step": 2950 }, { "epoch": 2.5605536332179932, "grad_norm": 5.22907829284668, "learning_rate": 7.410611303344868e-06, "loss": 2.1252, "step": 2960 }, { "epoch": 2.569204152249135, "grad_norm": 9.819774627685547, "learning_rate": 7.2664359861591705e-06, "loss": 2.069, "step": 2970 }, { "epoch": 2.577854671280277, "grad_norm": 7.364692211151123, "learning_rate": 7.1222606689734725e-06, "loss": 2.2576, "step": 2980 }, { "epoch": 2.586505190311419, "grad_norm": 7.494932651519775, "learning_rate": 6.978085351787775e-06, "loss": 2.2005, "step": 2990 }, { "epoch": 2.595155709342561, "grad_norm": 7.424341678619385, "learning_rate": 6.8339100346020775e-06, "loss": 1.9901, "step": 3000 }, { "epoch": 2.6038062283737027, "grad_norm": 7.437248706817627, "learning_rate": 6.689734717416378e-06, "loss": 2.1721, "step": 3010 }, { "epoch": 2.612456747404844, "grad_norm": 9.077981948852539, "learning_rate": 6.54555940023068e-06, "loss": 2.2581, "step": 3020 }, { "epoch": 2.621107266435986, "grad_norm": 13.281733512878418, "learning_rate": 6.401384083044983e-06, "loss": 2.2352, "step": 3030 }, { "epoch": 2.629757785467128, "grad_norm": 7.700297832489014, "learning_rate": 6.257208765859285e-06, "loss": 2.0725, "step": 3040 }, { "epoch": 2.63840830449827, "grad_norm": 10.37764835357666, "learning_rate": 6.113033448673587e-06, "loss": 2.0559, "step": 3050 }, { "epoch": 2.6470588235294117, "grad_norm": 7.580658912658691, "learning_rate": 5.96885813148789e-06, "loss": 2.103, "step": 3060 }, { "epoch": 2.6557093425605536, "grad_norm": 6.464138984680176, "learning_rate": 5.824682814302192e-06, "loss": 1.7107, "step": 3070 }, { "epoch": 2.6643598615916955, "grad_norm": 9.265501022338867, "learning_rate": 5.680507497116494e-06, "loss": 2.2707, "step": 3080 }, { "epoch": 2.6730103806228374, "grad_norm": 13.189889907836914, "learning_rate": 5.536332179930797e-06, "loss": 2.1762, "step": 3090 }, { "epoch": 2.6816608996539792, "grad_norm": 7.203210830688477, "learning_rate": 5.392156862745099e-06, "loss": 2.2239, "step": 3100 }, { "epoch": 2.690311418685121, "grad_norm": 4.9819560050964355, "learning_rate": 5.2479815455594e-06, "loss": 2.1301, "step": 3110 }, { "epoch": 2.698961937716263, "grad_norm": 9.288585662841797, "learning_rate": 5.103806228373703e-06, "loss": 2.062, "step": 3120 }, { "epoch": 2.707612456747405, "grad_norm": 0.0, "learning_rate": 4.959630911188005e-06, "loss": 2.1536, "step": 3130 }, { "epoch": 2.716262975778547, "grad_norm": 6.519169807434082, "learning_rate": 4.815455594002307e-06, "loss": 2.1112, "step": 3140 }, { "epoch": 2.7249134948096887, "grad_norm": 6.174961566925049, "learning_rate": 4.67128027681661e-06, "loss": 2.2762, "step": 3150 }, { "epoch": 2.7335640138408306, "grad_norm": 7.6661553382873535, "learning_rate": 4.527104959630912e-06, "loss": 2.2072, "step": 3160 }, { "epoch": 2.7422145328719725, "grad_norm": 13.350886344909668, "learning_rate": 4.382929642445214e-06, "loss": 2.0755, "step": 3170 }, { "epoch": 2.750865051903114, "grad_norm": 12.666411399841309, "learning_rate": 4.238754325259516e-06, "loss": 2.2341, "step": 3180 }, { "epoch": 2.759515570934256, "grad_norm": 7.005519390106201, "learning_rate": 4.094579008073818e-06, "loss": 1.9337, "step": 3190 }, { "epoch": 2.7681660899653977, "grad_norm": 9.67486572265625, "learning_rate": 3.95040369088812e-06, "loss": 2.1086, "step": 3200 }, { "epoch": 2.7768166089965396, "grad_norm": 7.333971977233887, "learning_rate": 3.8062283737024224e-06, "loss": 2.227, "step": 3210 }, { "epoch": 2.7854671280276815, "grad_norm": 7.098743915557861, "learning_rate": 3.662053056516725e-06, "loss": 1.983, "step": 3220 }, { "epoch": 2.7941176470588234, "grad_norm": 6.220783233642578, "learning_rate": 3.517877739331027e-06, "loss": 2.0352, "step": 3230 }, { "epoch": 2.8027681660899653, "grad_norm": 4.763909816741943, "learning_rate": 3.3737024221453293e-06, "loss": 2.0064, "step": 3240 }, { "epoch": 2.811418685121107, "grad_norm": 7.206717014312744, "learning_rate": 3.229527104959631e-06, "loss": 2.3018, "step": 3250 }, { "epoch": 2.820069204152249, "grad_norm": 6.354711055755615, "learning_rate": 3.085351787773933e-06, "loss": 2.0381, "step": 3260 }, { "epoch": 2.828719723183391, "grad_norm": 5.708442211151123, "learning_rate": 2.9411764705882355e-06, "loss": 2.027, "step": 3270 }, { "epoch": 2.837370242214533, "grad_norm": 6.7003960609436035, "learning_rate": 2.7970011534025375e-06, "loss": 1.9782, "step": 3280 }, { "epoch": 2.8460207612456747, "grad_norm": 9.009029388427734, "learning_rate": 2.65282583621684e-06, "loss": 1.8581, "step": 3290 }, { "epoch": 2.8546712802768166, "grad_norm": 17.01302719116211, "learning_rate": 2.508650519031142e-06, "loss": 2.1368, "step": 3300 }, { "epoch": 2.8633217993079585, "grad_norm": 9.710434913635254, "learning_rate": 2.364475201845444e-06, "loss": 2.1412, "step": 3310 }, { "epoch": 2.8719723183391004, "grad_norm": 5.8521552085876465, "learning_rate": 2.2202998846597465e-06, "loss": 1.8587, "step": 3320 }, { "epoch": 2.8806228373702423, "grad_norm": 5.487595558166504, "learning_rate": 2.0761245674740485e-06, "loss": 2.3644, "step": 3330 }, { "epoch": 2.889273356401384, "grad_norm": 15.505248069763184, "learning_rate": 1.9319492502883506e-06, "loss": 2.4036, "step": 3340 }, { "epoch": 2.897923875432526, "grad_norm": 12.557036399841309, "learning_rate": 1.787773933102653e-06, "loss": 2.117, "step": 3350 }, { "epoch": 2.906574394463668, "grad_norm": 6.89960241317749, "learning_rate": 1.6435986159169549e-06, "loss": 1.7789, "step": 3360 }, { "epoch": 2.91522491349481, "grad_norm": 16.061315536499023, "learning_rate": 1.4994232987312573e-06, "loss": 2.2278, "step": 3370 }, { "epoch": 2.9238754325259517, "grad_norm": 7.849757194519043, "learning_rate": 1.3552479815455596e-06, "loss": 2.0531, "step": 3380 }, { "epoch": 2.9325259515570936, "grad_norm": 6.261150360107422, "learning_rate": 1.2110726643598616e-06, "loss": 2.2677, "step": 3390 }, { "epoch": 2.9411764705882355, "grad_norm": 8.204355239868164, "learning_rate": 1.0668973471741639e-06, "loss": 2.1763, "step": 3400 }, { "epoch": 2.9498269896193774, "grad_norm": 7.603878021240234, "learning_rate": 9.22722029988466e-07, "loss": 2.1721, "step": 3410 }, { "epoch": 2.9584775086505193, "grad_norm": 10.588805198669434, "learning_rate": 7.785467128027681e-07, "loss": 2.1428, "step": 3420 }, { "epoch": 2.967128027681661, "grad_norm": 6.211885452270508, "learning_rate": 6.343713956170704e-07, "loss": 2.2216, "step": 3430 }, { "epoch": 2.9757785467128026, "grad_norm": 10.20269775390625, "learning_rate": 4.901960784313725e-07, "loss": 2.0611, "step": 3440 }, { "epoch": 2.9844290657439445, "grad_norm": 5.556583881378174, "learning_rate": 3.460207612456748e-07, "loss": 2.3122, "step": 3450 }, { "epoch": 2.9930795847750864, "grad_norm": 15.4011812210083, "learning_rate": 2.0184544405997694e-07, "loss": 2.142, "step": 3460 } ], "logging_steps": 10, "max_steps": 3468, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 3623859191808000.0, "train_batch_size": 4, "trial_name": null, "trial_params": null }