{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.07031576171746248, "eval_steps": 500, "global_step": 3000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0002343858723915416, "grad_norm": Infinity, "learning_rate": 1.40625e-07, "loss": 0.6092071533203125, "step": 10 }, { "epoch": 0.0004687717447830832, "grad_norm": Infinity, "learning_rate": 2.96875e-07, "loss": 0.6080734252929687, "step": 20 }, { "epoch": 0.0007031576171746248, "grad_norm": Infinity, "learning_rate": 4.53125e-07, "loss": 0.6008880615234375, "step": 30 }, { "epoch": 0.0009375434895661664, "grad_norm": Infinity, "learning_rate": 6.09375e-07, "loss": 0.6059707641601563, "step": 40 }, { "epoch": 0.001171929361957708, "grad_norm": Infinity, "learning_rate": 7.656250000000001e-07, "loss": 0.615313720703125, "step": 50 }, { "epoch": 0.0014063152343492496, "grad_norm": Infinity, "learning_rate": 9.218750000000002e-07, "loss": 0.6074554443359375, "step": 60 }, { "epoch": 0.0016407011067407912, "grad_norm": Infinity, "learning_rate": 1.0781250000000002e-06, "loss": 0.605792236328125, "step": 70 }, { "epoch": 0.0018750869791323327, "grad_norm": Infinity, "learning_rate": 1.2343750000000001e-06, "loss": 0.6068206787109375, "step": 80 }, { "epoch": 0.0021094728515238742, "grad_norm": Infinity, "learning_rate": 1.3906250000000001e-06, "loss": 0.6020843505859375, "step": 90 }, { "epoch": 0.002343858723915416, "grad_norm": Infinity, "learning_rate": 1.5468750000000001e-06, "loss": 0.59473876953125, "step": 100 }, { "epoch": 0.0025782445963069577, "grad_norm": Infinity, "learning_rate": 1.703125e-06, "loss": 0.6020477294921875, "step": 110 }, { "epoch": 0.0028126304686984993, "grad_norm": Infinity, "learning_rate": 1.8593750000000003e-06, "loss": 0.596124267578125, "step": 120 }, { "epoch": 0.003047016341090041, "grad_norm": Infinity, "learning_rate": 2.0156250000000003e-06, "loss": 0.6029708862304688, "step": 130 }, { "epoch": 0.0032814022134815823, "grad_norm": Infinity, "learning_rate": 2.1718750000000003e-06, "loss": 0.5990798950195313, "step": 140 }, { "epoch": 0.003515788085873124, "grad_norm": Infinity, "learning_rate": 2.3281250000000003e-06, "loss": 0.5956100463867188, "step": 150 }, { "epoch": 0.0037501739582646654, "grad_norm": Infinity, "learning_rate": 2.4843750000000002e-06, "loss": 0.59793701171875, "step": 160 }, { "epoch": 0.003984559830656207, "grad_norm": Infinity, "learning_rate": 2.640625e-06, "loss": 0.6025466918945312, "step": 170 }, { "epoch": 0.0042189457030477485, "grad_norm": Infinity, "learning_rate": 2.796875e-06, "loss": 0.6034042358398437, "step": 180 }, { "epoch": 0.0044533315754392904, "grad_norm": Infinity, "learning_rate": 2.953125e-06, "loss": 0.6109519958496094, "step": 190 }, { "epoch": 0.004687717447830832, "grad_norm": Infinity, "learning_rate": 3.109375e-06, "loss": 0.6124420166015625, "step": 200 }, { "epoch": 0.0049221033202223735, "grad_norm": Infinity, "learning_rate": 3.265625e-06, "loss": 0.5962295532226562, "step": 210 }, { "epoch": 0.0051564891926139155, "grad_norm": Infinity, "learning_rate": 3.421875e-06, "loss": 0.5996864318847657, "step": 220 }, { "epoch": 0.005390875065005457, "grad_norm": Infinity, "learning_rate": 3.578125e-06, "loss": 0.6116897583007812, "step": 230 }, { "epoch": 0.0056252609373969985, "grad_norm": Infinity, "learning_rate": 3.734375e-06, "loss": 0.6195701599121094, "step": 240 }, { "epoch": 0.00585964680978854, "grad_norm": Infinity, "learning_rate": 3.890625e-06, "loss": 0.6111648559570313, "step": 250 }, { "epoch": 0.006094032682180082, "grad_norm": Infinity, "learning_rate": 4.046875e-06, "loss": 0.6055526733398438, "step": 260 }, { "epoch": 0.006328418554571624, "grad_norm": Infinity, "learning_rate": 4.2031250000000005e-06, "loss": 0.5984527587890625, "step": 270 }, { "epoch": 0.006562804426963165, "grad_norm": Infinity, "learning_rate": 4.359375e-06, "loss": 0.6135429382324219, "step": 280 }, { "epoch": 0.006797190299354707, "grad_norm": Infinity, "learning_rate": 4.5156250000000005e-06, "loss": 0.5952308654785157, "step": 290 }, { "epoch": 0.007031576171746248, "grad_norm": Infinity, "learning_rate": 4.671875e-06, "loss": 0.6037254333496094, "step": 300 }, { "epoch": 0.00726596204413779, "grad_norm": Infinity, "learning_rate": 4.8281250000000005e-06, "loss": 0.5975120544433594, "step": 310 }, { "epoch": 0.007500347916529331, "grad_norm": Infinity, "learning_rate": 4.984375e-06, "loss": 0.5918365478515625, "step": 320 }, { "epoch": 0.007734733788920873, "grad_norm": Infinity, "learning_rate": 5.1406250000000004e-06, "loss": 0.6080474853515625, "step": 330 }, { "epoch": 0.007969119661312415, "grad_norm": Infinity, "learning_rate": 5.296875e-06, "loss": 0.6048797607421875, "step": 340 }, { "epoch": 0.008203505533703956, "grad_norm": Infinity, "learning_rate": 5.453125e-06, "loss": 0.593927001953125, "step": 350 }, { "epoch": 0.008437891406095497, "grad_norm": Infinity, "learning_rate": 5.609375e-06, "loss": 0.6036277770996094, "step": 360 }, { "epoch": 0.00867227727848704, "grad_norm": Infinity, "learning_rate": 5.765625e-06, "loss": 0.604638671875, "step": 370 }, { "epoch": 0.008906663150878581, "grad_norm": Infinity, "learning_rate": 5.921875e-06, "loss": 0.605169677734375, "step": 380 }, { "epoch": 0.009141049023270122, "grad_norm": Infinity, "learning_rate": 6.078125e-06, "loss": 0.613519287109375, "step": 390 }, { "epoch": 0.009375434895661665, "grad_norm": Infinity, "learning_rate": 6.234375e-06, "loss": 0.6017379760742188, "step": 400 }, { "epoch": 0.009609820768053206, "grad_norm": Infinity, "learning_rate": 6.390625e-06, "loss": 0.6115753173828125, "step": 410 }, { "epoch": 0.009844206640444747, "grad_norm": Infinity, "learning_rate": 6.546875e-06, "loss": 0.6092376708984375, "step": 420 }, { "epoch": 0.010078592512836288, "grad_norm": Infinity, "learning_rate": 6.703125e-06, "loss": 0.6099075317382813, "step": 430 }, { "epoch": 0.010312978385227831, "grad_norm": Infinity, "learning_rate": 6.859375000000001e-06, "loss": 0.61324462890625, "step": 440 }, { "epoch": 0.010547364257619372, "grad_norm": Infinity, "learning_rate": 7.015625e-06, "loss": 0.6022865295410156, "step": 450 }, { "epoch": 0.010781750130010913, "grad_norm": Infinity, "learning_rate": 7.171875000000001e-06, "loss": 0.6093353271484375, "step": 460 }, { "epoch": 0.011016136002402456, "grad_norm": Infinity, "learning_rate": 7.328125e-06, "loss": 0.62122802734375, "step": 470 }, { "epoch": 0.011250521874793997, "grad_norm": Infinity, "learning_rate": 7.484375000000001e-06, "loss": 0.6011276245117188, "step": 480 }, { "epoch": 0.011484907747185538, "grad_norm": Infinity, "learning_rate": 7.640625000000001e-06, "loss": 0.6096969604492187, "step": 490 }, { "epoch": 0.01171929361957708, "grad_norm": Infinity, "learning_rate": 7.796875e-06, "loss": 0.6085403442382813, "step": 500 }, { "epoch": 0.011953679491968622, "grad_norm": Infinity, "learning_rate": 7.953125e-06, "loss": 0.6013351440429687, "step": 510 }, { "epoch": 0.012188065364360163, "grad_norm": Infinity, "learning_rate": 8.109375e-06, "loss": 0.6108566284179687, "step": 520 }, { "epoch": 0.012422451236751704, "grad_norm": Infinity, "learning_rate": 8.265625000000001e-06, "loss": 0.6119140625, "step": 530 }, { "epoch": 0.012656837109143247, "grad_norm": Infinity, "learning_rate": 8.421875e-06, "loss": 0.6089179992675782, "step": 540 }, { "epoch": 0.012891222981534788, "grad_norm": Infinity, "learning_rate": 8.578125e-06, "loss": 0.6096832275390625, "step": 550 }, { "epoch": 0.01312560885392633, "grad_norm": Infinity, "learning_rate": 8.734375e-06, "loss": 0.6111366271972656, "step": 560 }, { "epoch": 0.01335999472631787, "grad_norm": Infinity, "learning_rate": 8.890625000000001e-06, "loss": 0.6114044189453125, "step": 570 }, { "epoch": 0.013594380598709413, "grad_norm": Infinity, "learning_rate": 9.046875e-06, "loss": 0.6101348876953125, "step": 580 }, { "epoch": 0.013828766471100954, "grad_norm": Infinity, "learning_rate": 9.203125e-06, "loss": 0.6197845458984375, "step": 590 }, { "epoch": 0.014063152343492496, "grad_norm": Infinity, "learning_rate": 9.359375e-06, "loss": 0.615380859375, "step": 600 }, { "epoch": 0.014297538215884038, "grad_norm": Infinity, "learning_rate": 9.515625000000001e-06, "loss": 0.606341552734375, "step": 610 }, { "epoch": 0.01453192408827558, "grad_norm": Infinity, "learning_rate": 9.671875000000001e-06, "loss": 0.6048248291015625, "step": 620 }, { "epoch": 0.01476630996066712, "grad_norm": Infinity, "learning_rate": 9.828125e-06, "loss": 0.6092742919921875, "step": 630 }, { "epoch": 0.015000695833058662, "grad_norm": Infinity, "learning_rate": 9.984375e-06, "loss": 0.6080833435058594, "step": 640 }, { "epoch": 0.015235081705450204, "grad_norm": Infinity, "learning_rate": 1.0140625000000003e-05, "loss": 0.6076614379882812, "step": 650 }, { "epoch": 0.015469467577841746, "grad_norm": Infinity, "learning_rate": 1.0296875000000001e-05, "loss": 0.6062957763671875, "step": 660 }, { "epoch": 0.01570385345023329, "grad_norm": Infinity, "learning_rate": 1.0453125000000002e-05, "loss": 0.6010467529296875, "step": 670 }, { "epoch": 0.01593823932262483, "grad_norm": Infinity, "learning_rate": 1.0609375000000002e-05, "loss": 0.6066925048828125, "step": 680 }, { "epoch": 0.01617262519501637, "grad_norm": Infinity, "learning_rate": 1.0765625000000002e-05, "loss": 0.6043624877929688, "step": 690 }, { "epoch": 0.01640701106740791, "grad_norm": Infinity, "learning_rate": 1.0921875000000001e-05, "loss": 0.6023162841796875, "step": 700 }, { "epoch": 0.016641396939799453, "grad_norm": Infinity, "learning_rate": 1.1078125000000002e-05, "loss": 0.6094833374023437, "step": 710 }, { "epoch": 0.016875782812190994, "grad_norm": Infinity, "learning_rate": 1.1234375000000002e-05, "loss": 0.6054359436035156, "step": 720 }, { "epoch": 0.01711016868458254, "grad_norm": Infinity, "learning_rate": 1.1390625000000002e-05, "loss": 0.6059349060058594, "step": 730 }, { "epoch": 0.01734455455697408, "grad_norm": Infinity, "learning_rate": 1.1546875000000001e-05, "loss": 0.60126953125, "step": 740 }, { "epoch": 0.01757894042936562, "grad_norm": Infinity, "learning_rate": 1.1703125000000002e-05, "loss": 0.6094718933105469, "step": 750 }, { "epoch": 0.017813326301757162, "grad_norm": Infinity, "learning_rate": 1.1859375000000002e-05, "loss": 0.6063560485839844, "step": 760 }, { "epoch": 0.018047712174148703, "grad_norm": Infinity, "learning_rate": 1.2015625000000002e-05, "loss": 0.6024421691894531, "step": 770 }, { "epoch": 0.018282098046540244, "grad_norm": Infinity, "learning_rate": 1.2171875000000001e-05, "loss": 0.6170989990234375, "step": 780 }, { "epoch": 0.018516483918931785, "grad_norm": Infinity, "learning_rate": 1.2328125000000002e-05, "loss": 0.5938522338867187, "step": 790 }, { "epoch": 0.01875086979132333, "grad_norm": Infinity, "learning_rate": 1.2484375000000002e-05, "loss": 0.6142776489257813, "step": 800 }, { "epoch": 0.01898525566371487, "grad_norm": Infinity, "learning_rate": 1.2640625000000002e-05, "loss": 0.6042465209960938, "step": 810 }, { "epoch": 0.019219641536106412, "grad_norm": Infinity, "learning_rate": 1.2796875000000003e-05, "loss": 0.6070510864257812, "step": 820 }, { "epoch": 0.019454027408497953, "grad_norm": Infinity, "learning_rate": 1.2953125000000001e-05, "loss": 0.6068756103515625, "step": 830 }, { "epoch": 0.019688413280889494, "grad_norm": Infinity, "learning_rate": 1.3109375000000002e-05, "loss": 0.5897056579589843, "step": 840 }, { "epoch": 0.019922799153281035, "grad_norm": Infinity, "learning_rate": 1.3265625000000002e-05, "loss": 0.6050689697265625, "step": 850 }, { "epoch": 0.020157185025672576, "grad_norm": Infinity, "learning_rate": 1.3421875000000003e-05, "loss": 0.590875244140625, "step": 860 }, { "epoch": 0.02039157089806412, "grad_norm": Infinity, "learning_rate": 1.3578125000000001e-05, "loss": 0.6160301208496094, "step": 870 }, { "epoch": 0.020625956770455662, "grad_norm": Infinity, "learning_rate": 1.3734375000000002e-05, "loss": 0.6036376953125, "step": 880 }, { "epoch": 0.020860342642847203, "grad_norm": Infinity, "learning_rate": 1.3890625000000002e-05, "loss": 0.610693359375, "step": 890 }, { "epoch": 0.021094728515238744, "grad_norm": Infinity, "learning_rate": 1.4046875000000003e-05, "loss": 0.6084335327148438, "step": 900 }, { "epoch": 0.021329114387630285, "grad_norm": Infinity, "learning_rate": 1.4203125000000001e-05, "loss": 0.6077606201171875, "step": 910 }, { "epoch": 0.021563500260021826, "grad_norm": Infinity, "learning_rate": 1.4359375000000002e-05, "loss": 0.6037399291992187, "step": 920 }, { "epoch": 0.021797886132413367, "grad_norm": Infinity, "learning_rate": 1.4515625000000002e-05, "loss": 0.6144996643066406, "step": 930 }, { "epoch": 0.022032272004804912, "grad_norm": Infinity, "learning_rate": 1.4671875000000003e-05, "loss": 0.5966384887695313, "step": 940 }, { "epoch": 0.022266657877196453, "grad_norm": Infinity, "learning_rate": 1.4828125000000001e-05, "loss": 0.603125, "step": 950 }, { "epoch": 0.022501043749587994, "grad_norm": Infinity, "learning_rate": 1.4984375000000002e-05, "loss": 0.612744140625, "step": 960 }, { "epoch": 0.022735429621979535, "grad_norm": Infinity, "learning_rate": 1.5140625000000002e-05, "loss": 0.6046875, "step": 970 }, { "epoch": 0.022969815494371076, "grad_norm": Infinity, "learning_rate": 1.5296875e-05, "loss": 0.6094146728515625, "step": 980 }, { "epoch": 0.023204201366762617, "grad_norm": Infinity, "learning_rate": 1.5453125e-05, "loss": 0.610845947265625, "step": 990 }, { "epoch": 0.02343858723915416, "grad_norm": Infinity, "learning_rate": 1.5609375e-05, "loss": 0.5980422973632813, "step": 1000 }, { "epoch": 0.023672973111545703, "grad_norm": Infinity, "learning_rate": 1.5765625000000002e-05, "loss": 0.6019882202148438, "step": 1010 }, { "epoch": 0.023907358983937244, "grad_norm": Infinity, "learning_rate": 1.5921875000000002e-05, "loss": 0.6057106018066406, "step": 1020 }, { "epoch": 0.024141744856328785, "grad_norm": Infinity, "learning_rate": 1.6078125000000003e-05, "loss": 0.6211753845214844, "step": 1030 }, { "epoch": 0.024376130728720326, "grad_norm": Infinity, "learning_rate": 1.6234375000000003e-05, "loss": 0.6060012817382813, "step": 1040 }, { "epoch": 0.024610516601111868, "grad_norm": Infinity, "learning_rate": 1.6390625000000004e-05, "loss": 0.5908248901367188, "step": 1050 }, { "epoch": 0.02484490247350341, "grad_norm": Infinity, "learning_rate": 1.6546875e-05, "loss": 0.5960098266601562, "step": 1060 }, { "epoch": 0.02507928834589495, "grad_norm": Infinity, "learning_rate": 1.6703125e-05, "loss": 0.5927337646484375, "step": 1070 }, { "epoch": 0.025313674218286494, "grad_norm": Infinity, "learning_rate": 1.6859375e-05, "loss": 0.6080078125, "step": 1080 }, { "epoch": 0.025548060090678035, "grad_norm": Infinity, "learning_rate": 1.7015625000000002e-05, "loss": 0.6029281616210938, "step": 1090 }, { "epoch": 0.025782445963069577, "grad_norm": Infinity, "learning_rate": 1.7171875000000002e-05, "loss": 0.6014022827148438, "step": 1100 }, { "epoch": 0.026016831835461118, "grad_norm": Infinity, "learning_rate": 1.7328125000000003e-05, "loss": 0.6068984985351562, "step": 1110 }, { "epoch": 0.02625121770785266, "grad_norm": Infinity, "learning_rate": 1.7484375000000003e-05, "loss": 0.611175537109375, "step": 1120 }, { "epoch": 0.0264856035802442, "grad_norm": Infinity, "learning_rate": 1.7640625000000004e-05, "loss": 0.6017318725585937, "step": 1130 }, { "epoch": 0.02671998945263574, "grad_norm": Infinity, "learning_rate": 1.7796875000000004e-05, "loss": 0.6082473754882812, "step": 1140 }, { "epoch": 0.026954375325027286, "grad_norm": Infinity, "learning_rate": 1.7953125e-05, "loss": 0.6051132202148437, "step": 1150 }, { "epoch": 0.027188761197418827, "grad_norm": Infinity, "learning_rate": 1.8109375e-05, "loss": 0.6071746826171875, "step": 1160 }, { "epoch": 0.027423147069810368, "grad_norm": Infinity, "learning_rate": 1.8265625000000002e-05, "loss": 0.6165802001953125, "step": 1170 }, { "epoch": 0.02765753294220191, "grad_norm": Infinity, "learning_rate": 1.8421875000000002e-05, "loss": 0.5984237670898438, "step": 1180 }, { "epoch": 0.02789191881459345, "grad_norm": Infinity, "learning_rate": 1.8578125000000003e-05, "loss": 0.6050758361816406, "step": 1190 }, { "epoch": 0.02812630468698499, "grad_norm": Infinity, "learning_rate": 1.8734375000000003e-05, "loss": 0.598980712890625, "step": 1200 }, { "epoch": 0.028360690559376532, "grad_norm": Infinity, "learning_rate": 1.8890625000000003e-05, "loss": 0.5932998657226562, "step": 1210 }, { "epoch": 0.028595076431768077, "grad_norm": Infinity, "learning_rate": 1.9046875000000004e-05, "loss": 0.6028633117675781, "step": 1220 }, { "epoch": 0.028829462304159618, "grad_norm": Infinity, "learning_rate": 1.9203125e-05, "loss": 0.6024459838867188, "step": 1230 }, { "epoch": 0.02906384817655116, "grad_norm": Infinity, "learning_rate": 1.9359375e-05, "loss": 0.6077346801757812, "step": 1240 }, { "epoch": 0.0292982340489427, "grad_norm": Infinity, "learning_rate": 1.9515625000000002e-05, "loss": 0.6018325805664062, "step": 1250 }, { "epoch": 0.02953261992133424, "grad_norm": Infinity, "learning_rate": 1.9671875000000002e-05, "loss": 0.6083480834960937, "step": 1260 }, { "epoch": 0.029767005793725782, "grad_norm": Infinity, "learning_rate": 1.9828125000000003e-05, "loss": 0.6122581481933593, "step": 1270 }, { "epoch": 0.030001391666117323, "grad_norm": Infinity, "learning_rate": 1.9984375000000003e-05, "loss": 0.6014678955078125, "step": 1280 }, { "epoch": 0.030235777538508868, "grad_norm": Infinity, "learning_rate": 1.9999997666172074e-05, "loss": 0.6058135986328125, "step": 1290 }, { "epoch": 0.03047016341090041, "grad_norm": Infinity, "learning_rate": 1.9999989598620137e-05, "loss": 0.6135894775390625, "step": 1300 }, { "epoch": 0.03070454928329195, "grad_norm": Infinity, "learning_rate": 1.9999975768536158e-05, "loss": 0.6187896728515625, "step": 1310 }, { "epoch": 0.03093893515568349, "grad_norm": Infinity, "learning_rate": 1.9999956175928097e-05, "loss": 0.60782470703125, "step": 1320 }, { "epoch": 0.031173321028075032, "grad_norm": Infinity, "learning_rate": 1.9999930820807245e-05, "loss": 0.6168212890625, "step": 1330 }, { "epoch": 0.03140770690046658, "grad_norm": Infinity, "learning_rate": 1.999989970318822e-05, "loss": 0.5995613098144531, "step": 1340 }, { "epoch": 0.031642092772858114, "grad_norm": Infinity, "learning_rate": 1.9999862823088946e-05, "loss": 0.6088516235351562, "step": 1350 }, { "epoch": 0.03187647864524966, "grad_norm": Infinity, "learning_rate": 1.999982018053068e-05, "loss": 0.6085662841796875, "step": 1360 }, { "epoch": 0.0321108645176412, "grad_norm": Infinity, "learning_rate": 1.999977177553799e-05, "loss": 0.6037391662597656, "step": 1370 }, { "epoch": 0.03234525039003274, "grad_norm": Infinity, "learning_rate": 1.9999717608138778e-05, "loss": 0.6000991821289062, "step": 1380 }, { "epoch": 0.032579636262424286, "grad_norm": Infinity, "learning_rate": 1.9999657678364252e-05, "loss": 0.5998687744140625, "step": 1390 }, { "epoch": 0.03281402213481582, "grad_norm": Infinity, "learning_rate": 1.9999591986248947e-05, "loss": 0.5970962524414063, "step": 1400 }, { "epoch": 0.03304840800720737, "grad_norm": Infinity, "learning_rate": 1.999952053183072e-05, "loss": 0.6116119384765625, "step": 1410 }, { "epoch": 0.033282793879598906, "grad_norm": Infinity, "learning_rate": 1.9999443315150742e-05, "loss": 0.602606201171875, "step": 1420 }, { "epoch": 0.03351717975199045, "grad_norm": Infinity, "learning_rate": 1.999936033625352e-05, "loss": 0.6079139709472656, "step": 1430 }, { "epoch": 0.03375156562438199, "grad_norm": Infinity, "learning_rate": 1.999927159518686e-05, "loss": 0.6049301147460937, "step": 1440 }, { "epoch": 0.03398595149677353, "grad_norm": Infinity, "learning_rate": 1.9999177092001907e-05, "loss": 0.6092140197753906, "step": 1450 }, { "epoch": 0.03422033736916508, "grad_norm": Infinity, "learning_rate": 1.9999076826753116e-05, "loss": 0.5999214172363281, "step": 1460 }, { "epoch": 0.034454723241556615, "grad_norm": Infinity, "learning_rate": 1.9998970799498262e-05, "loss": 0.6099578857421875, "step": 1470 }, { "epoch": 0.03468910911394816, "grad_norm": Infinity, "learning_rate": 1.9998859010298447e-05, "loss": 0.6190902709960937, "step": 1480 }, { "epoch": 0.0349234949863397, "grad_norm": Infinity, "learning_rate": 1.9998741459218093e-05, "loss": 0.6074859619140625, "step": 1490 }, { "epoch": 0.03515788085873124, "grad_norm": Infinity, "learning_rate": 1.999861814632493e-05, "loss": 0.6160079956054687, "step": 1500 }, { "epoch": 0.03539226673112278, "grad_norm": Infinity, "learning_rate": 1.9998489071690024e-05, "loss": 0.5973175048828125, "step": 1510 }, { "epoch": 0.035626652603514324, "grad_norm": Infinity, "learning_rate": 1.9998354235387758e-05, "loss": 0.6043167114257812, "step": 1520 }, { "epoch": 0.03586103847590587, "grad_norm": Infinity, "learning_rate": 1.9998213637495824e-05, "loss": 0.6068344116210938, "step": 1530 }, { "epoch": 0.036095424348297406, "grad_norm": Infinity, "learning_rate": 1.999806727809525e-05, "loss": 0.6181640625, "step": 1540 }, { "epoch": 0.03632981022068895, "grad_norm": Infinity, "learning_rate": 1.9997915157270367e-05, "loss": 0.6040145874023437, "step": 1550 }, { "epoch": 0.03656419609308049, "grad_norm": Infinity, "learning_rate": 1.9997757275108847e-05, "loss": 0.5977676391601563, "step": 1560 }, { "epoch": 0.03679858196547203, "grad_norm": Infinity, "learning_rate": 1.999759363170166e-05, "loss": 0.6115097045898438, "step": 1570 }, { "epoch": 0.03703296783786357, "grad_norm": Infinity, "learning_rate": 1.9997424227143108e-05, "loss": 0.6066993713378906, "step": 1580 }, { "epoch": 0.037267353710255115, "grad_norm": Infinity, "learning_rate": 1.9997249061530814e-05, "loss": 0.6025985717773438, "step": 1590 }, { "epoch": 0.03750173958264666, "grad_norm": Infinity, "learning_rate": 1.999706813496572e-05, "loss": 0.598687744140625, "step": 1600 }, { "epoch": 0.0377361254550382, "grad_norm": Infinity, "learning_rate": 1.9996881447552077e-05, "loss": 0.5959098815917969, "step": 1610 }, { "epoch": 0.03797051132742974, "grad_norm": Infinity, "learning_rate": 1.9996688999397473e-05, "loss": 0.6068992614746094, "step": 1620 }, { "epoch": 0.03820489719982128, "grad_norm": Infinity, "learning_rate": 1.99964907906128e-05, "loss": 0.6098403930664062, "step": 1630 }, { "epoch": 0.038439283072212824, "grad_norm": Infinity, "learning_rate": 1.9996286821312282e-05, "loss": 0.6098678588867188, "step": 1640 }, { "epoch": 0.03867366894460436, "grad_norm": Infinity, "learning_rate": 1.9996077091613454e-05, "loss": 0.60889892578125, "step": 1650 }, { "epoch": 0.038908054816995906, "grad_norm": Infinity, "learning_rate": 1.9995861601637175e-05, "loss": 0.6116622924804688, "step": 1660 }, { "epoch": 0.03914244068938745, "grad_norm": Infinity, "learning_rate": 1.9995640351507623e-05, "loss": 0.6043411254882812, "step": 1670 }, { "epoch": 0.03937682656177899, "grad_norm": Infinity, "learning_rate": 1.9995413341352293e-05, "loss": 0.6063034057617187, "step": 1680 }, { "epoch": 0.03961121243417053, "grad_norm": Infinity, "learning_rate": 1.9995180571301998e-05, "loss": 0.6143798828125, "step": 1690 }, { "epoch": 0.03984559830656207, "grad_norm": Infinity, "learning_rate": 1.9994942041490878e-05, "loss": 0.602484130859375, "step": 1700 }, { "epoch": 0.040079984178953615, "grad_norm": Infinity, "learning_rate": 1.9994697752056385e-05, "loss": 0.6013954162597657, "step": 1710 }, { "epoch": 0.04031437005134515, "grad_norm": Infinity, "learning_rate": 1.9994447703139286e-05, "loss": 0.5985885620117187, "step": 1720 }, { "epoch": 0.0405487559237367, "grad_norm": Infinity, "learning_rate": 1.999419189488368e-05, "loss": 0.6090835571289063, "step": 1730 }, { "epoch": 0.04078314179612824, "grad_norm": Infinity, "learning_rate": 1.999393032743697e-05, "loss": 0.6157844543457032, "step": 1740 }, { "epoch": 0.04101752766851978, "grad_norm": Infinity, "learning_rate": 1.9993663000949893e-05, "loss": 0.6032363891601562, "step": 1750 }, { "epoch": 0.041251913540911324, "grad_norm": Infinity, "learning_rate": 1.9993389915576494e-05, "loss": 0.6065597534179688, "step": 1760 }, { "epoch": 0.04148629941330286, "grad_norm": Infinity, "learning_rate": 1.9993111071474138e-05, "loss": 0.60250244140625, "step": 1770 }, { "epoch": 0.041720685285694406, "grad_norm": Infinity, "learning_rate": 1.999282646880351e-05, "loss": 0.5997932434082032, "step": 1780 }, { "epoch": 0.041955071158085944, "grad_norm": Infinity, "learning_rate": 1.999253610772862e-05, "loss": 0.6047691345214844, "step": 1790 }, { "epoch": 0.04218945703047749, "grad_norm": Infinity, "learning_rate": 1.9992239988416778e-05, "loss": 0.5975669860839844, "step": 1800 }, { "epoch": 0.04242384290286903, "grad_norm": Infinity, "learning_rate": 1.999193811103863e-05, "loss": 0.600311279296875, "step": 1810 }, { "epoch": 0.04265822877526057, "grad_norm": Infinity, "learning_rate": 1.9991630475768136e-05, "loss": 0.588726806640625, "step": 1820 }, { "epoch": 0.042892614647652115, "grad_norm": Infinity, "learning_rate": 1.9991317082782572e-05, "loss": 0.6026519775390625, "step": 1830 }, { "epoch": 0.04312700052004365, "grad_norm": Infinity, "learning_rate": 1.9990997932262525e-05, "loss": 0.5979637145996094, "step": 1840 }, { "epoch": 0.0433613863924352, "grad_norm": Infinity, "learning_rate": 1.9990673024391914e-05, "loss": 0.6101821899414063, "step": 1850 }, { "epoch": 0.043595772264826735, "grad_norm": Infinity, "learning_rate": 1.999034235935797e-05, "loss": 0.6061965942382812, "step": 1860 }, { "epoch": 0.04383015813721828, "grad_norm": Infinity, "learning_rate": 1.999000593735123e-05, "loss": 0.5992431640625, "step": 1870 }, { "epoch": 0.044064544009609824, "grad_norm": Infinity, "learning_rate": 1.9989663758565565e-05, "loss": 0.6042930603027343, "step": 1880 }, { "epoch": 0.04429892988200136, "grad_norm": Infinity, "learning_rate": 1.9989315823198158e-05, "loss": 0.6003738403320312, "step": 1890 }, { "epoch": 0.044533315754392906, "grad_norm": Infinity, "learning_rate": 1.998896213144951e-05, "loss": 0.6165328979492187, "step": 1900 }, { "epoch": 0.044767701626784444, "grad_norm": Infinity, "learning_rate": 1.9988602683523427e-05, "loss": 0.6063156127929688, "step": 1910 }, { "epoch": 0.04500208749917599, "grad_norm": Infinity, "learning_rate": 1.998823747962705e-05, "loss": 0.6078697204589844, "step": 1920 }, { "epoch": 0.045236473371567526, "grad_norm": Infinity, "learning_rate": 1.9987866519970832e-05, "loss": 0.6019554138183594, "step": 1930 }, { "epoch": 0.04547085924395907, "grad_norm": Infinity, "learning_rate": 1.9987489804768532e-05, "loss": 0.6099159240722656, "step": 1940 }, { "epoch": 0.045705245116350615, "grad_norm": Infinity, "learning_rate": 1.998710733423724e-05, "loss": 0.6168869018554688, "step": 1950 }, { "epoch": 0.04593963098874215, "grad_norm": Infinity, "learning_rate": 1.9986719108597354e-05, "loss": 0.6052276611328125, "step": 1960 }, { "epoch": 0.0461740168611337, "grad_norm": Infinity, "learning_rate": 1.9986325128072585e-05, "loss": 0.6113533020019531, "step": 1970 }, { "epoch": 0.046408402733525235, "grad_norm": Infinity, "learning_rate": 1.9985925392889974e-05, "loss": 0.5971221923828125, "step": 1980 }, { "epoch": 0.04664278860591678, "grad_norm": Infinity, "learning_rate": 1.9985519903279868e-05, "loss": 0.6067138671875, "step": 1990 }, { "epoch": 0.04687717447830832, "grad_norm": Infinity, "learning_rate": 1.998510865947593e-05, "loss": 0.6020065307617187, "step": 2000 }, { "epoch": 0.04711156035069986, "grad_norm": Infinity, "learning_rate": 1.998469166171514e-05, "loss": 0.6013572692871094, "step": 2010 }, { "epoch": 0.047345946223091406, "grad_norm": Infinity, "learning_rate": 1.9984268910237797e-05, "loss": 0.609564208984375, "step": 2020 }, { "epoch": 0.047580332095482944, "grad_norm": Infinity, "learning_rate": 1.9983840405287507e-05, "loss": 0.6112464904785156, "step": 2030 }, { "epoch": 0.04781471796787449, "grad_norm": Infinity, "learning_rate": 1.998340614711121e-05, "loss": 0.6053474426269532, "step": 2040 }, { "epoch": 0.048049103840266026, "grad_norm": Infinity, "learning_rate": 1.998296613595914e-05, "loss": 0.615625, "step": 2050 }, { "epoch": 0.04828348971265757, "grad_norm": Infinity, "learning_rate": 1.9982520372084855e-05, "loss": 0.6065841674804687, "step": 2060 }, { "epoch": 0.04851787558504911, "grad_norm": Infinity, "learning_rate": 1.9982068855745232e-05, "loss": 0.6024673461914063, "step": 2070 }, { "epoch": 0.04875226145744065, "grad_norm": Infinity, "learning_rate": 1.9981611587200453e-05, "loss": 0.6060653686523437, "step": 2080 }, { "epoch": 0.0489866473298322, "grad_norm": Infinity, "learning_rate": 1.998114856671403e-05, "loss": 0.5970504760742188, "step": 2090 }, { "epoch": 0.049221033202223735, "grad_norm": Infinity, "learning_rate": 1.998067979455277e-05, "loss": 0.6126579284667969, "step": 2100 }, { "epoch": 0.04945541907461528, "grad_norm": Infinity, "learning_rate": 1.998020527098681e-05, "loss": 0.5988540649414062, "step": 2110 }, { "epoch": 0.04968980494700682, "grad_norm": Infinity, "learning_rate": 1.99797249962896e-05, "loss": 0.6030006408691406, "step": 2120 }, { "epoch": 0.04992419081939836, "grad_norm": Infinity, "learning_rate": 1.997923897073789e-05, "loss": 0.6037818908691406, "step": 2130 }, { "epoch": 0.0501585766917899, "grad_norm": Infinity, "learning_rate": 1.9978747194611767e-05, "loss": 0.6064773559570312, "step": 2140 }, { "epoch": 0.050392962564181444, "grad_norm": Infinity, "learning_rate": 1.9978249668194607e-05, "loss": 0.59366455078125, "step": 2150 }, { "epoch": 0.05062734843657299, "grad_norm": Infinity, "learning_rate": 1.9977746391773117e-05, "loss": 0.605047607421875, "step": 2160 }, { "epoch": 0.050861734308964526, "grad_norm": Infinity, "learning_rate": 1.9977237365637313e-05, "loss": 0.6047653198242188, "step": 2170 }, { "epoch": 0.05109612018135607, "grad_norm": Infinity, "learning_rate": 1.997672259008052e-05, "loss": 0.6039505004882812, "step": 2180 }, { "epoch": 0.05133050605374761, "grad_norm": Infinity, "learning_rate": 1.997620206539938e-05, "loss": 0.6057266235351563, "step": 2190 }, { "epoch": 0.05156489192613915, "grad_norm": Infinity, "learning_rate": 1.9975675791893848e-05, "loss": 0.6024742126464844, "step": 2200 }, { "epoch": 0.05179927779853069, "grad_norm": Infinity, "learning_rate": 1.997514376986719e-05, "loss": 0.6101814270019531, "step": 2210 }, { "epoch": 0.052033663670922235, "grad_norm": Infinity, "learning_rate": 1.997460599962599e-05, "loss": 0.60872802734375, "step": 2220 }, { "epoch": 0.05226804954331378, "grad_norm": Infinity, "learning_rate": 1.9974062481480136e-05, "loss": 0.6200424194335937, "step": 2230 }, { "epoch": 0.05250243541570532, "grad_norm": Infinity, "learning_rate": 1.9973513215742836e-05, "loss": 0.6001205444335938, "step": 2240 }, { "epoch": 0.05273682128809686, "grad_norm": Infinity, "learning_rate": 1.9972958202730602e-05, "loss": 0.6031242370605469, "step": 2250 }, { "epoch": 0.0529712071604884, "grad_norm": Infinity, "learning_rate": 1.997239744276326e-05, "loss": 0.6068923950195313, "step": 2260 }, { "epoch": 0.053205593032879944, "grad_norm": Infinity, "learning_rate": 1.9971830936163964e-05, "loss": 0.5976806640625, "step": 2270 }, { "epoch": 0.05343997890527148, "grad_norm": Infinity, "learning_rate": 1.997125868325915e-05, "loss": 0.59666748046875, "step": 2280 }, { "epoch": 0.053674364777663026, "grad_norm": Infinity, "learning_rate": 1.9970680684378586e-05, "loss": 0.616412353515625, "step": 2290 }, { "epoch": 0.05390875065005457, "grad_norm": Infinity, "learning_rate": 1.9970096939855347e-05, "loss": 0.6073104858398437, "step": 2300 }, { "epoch": 0.05414313652244611, "grad_norm": Infinity, "learning_rate": 1.996950745002582e-05, "loss": 0.6183563232421875, "step": 2310 }, { "epoch": 0.05437752239483765, "grad_norm": Infinity, "learning_rate": 1.9968912215229697e-05, "loss": 0.609417724609375, "step": 2320 }, { "epoch": 0.05461190826722919, "grad_norm": Infinity, "learning_rate": 1.9968311235809988e-05, "loss": 0.6044479370117187, "step": 2330 }, { "epoch": 0.054846294139620735, "grad_norm": Infinity, "learning_rate": 1.9967704512113006e-05, "loss": 0.6251556396484375, "step": 2340 }, { "epoch": 0.05508068001201227, "grad_norm": Infinity, "learning_rate": 1.996709204448838e-05, "loss": 0.6080245971679688, "step": 2350 }, { "epoch": 0.05531506588440382, "grad_norm": Infinity, "learning_rate": 1.9966473833289047e-05, "loss": 0.6170455932617187, "step": 2360 }, { "epoch": 0.05554945175679536, "grad_norm": Infinity, "learning_rate": 1.9965849878871254e-05, "loss": 0.6163551330566406, "step": 2370 }, { "epoch": 0.0557838376291869, "grad_norm": Infinity, "learning_rate": 1.9965220181594555e-05, "loss": 0.6092254638671875, "step": 2380 }, { "epoch": 0.056018223501578444, "grad_norm": Infinity, "learning_rate": 1.9964584741821814e-05, "loss": 0.612530517578125, "step": 2390 }, { "epoch": 0.05625260937396998, "grad_norm": Infinity, "learning_rate": 1.996394355991921e-05, "loss": 0.6067825317382812, "step": 2400 }, { "epoch": 0.05648699524636153, "grad_norm": Infinity, "learning_rate": 1.9963296636256223e-05, "loss": 0.6045364379882813, "step": 2410 }, { "epoch": 0.056721381118753064, "grad_norm": Infinity, "learning_rate": 1.996264397120565e-05, "loss": 0.6018569946289063, "step": 2420 }, { "epoch": 0.05695576699114461, "grad_norm": Infinity, "learning_rate": 1.9961985565143586e-05, "loss": 0.6091682434082031, "step": 2430 }, { "epoch": 0.05719015286353615, "grad_norm": Infinity, "learning_rate": 1.9961321418449443e-05, "loss": 0.6118743896484375, "step": 2440 }, { "epoch": 0.05742453873592769, "grad_norm": Infinity, "learning_rate": 1.996065153150594e-05, "loss": 0.6104141235351562, "step": 2450 }, { "epoch": 0.057658924608319236, "grad_norm": Infinity, "learning_rate": 1.9959975904699096e-05, "loss": 0.61741943359375, "step": 2460 }, { "epoch": 0.05789331048071077, "grad_norm": Infinity, "learning_rate": 1.995929453841825e-05, "loss": 0.5939064025878906, "step": 2470 }, { "epoch": 0.05812769635310232, "grad_norm": Infinity, "learning_rate": 1.9958607433056035e-05, "loss": 0.60416259765625, "step": 2480 }, { "epoch": 0.058362082225493855, "grad_norm": Infinity, "learning_rate": 1.9957914589008405e-05, "loss": 0.601788330078125, "step": 2490 }, { "epoch": 0.0585964680978854, "grad_norm": Infinity, "learning_rate": 1.995721600667461e-05, "loss": 0.5972190856933594, "step": 2500 }, { "epoch": 0.058830853970276945, "grad_norm": Infinity, "learning_rate": 1.9956511686457213e-05, "loss": 0.611767578125, "step": 2510 }, { "epoch": 0.05906523984266848, "grad_norm": Infinity, "learning_rate": 1.9955801628762076e-05, "loss": 0.5967109680175782, "step": 2520 }, { "epoch": 0.05929962571506003, "grad_norm": Infinity, "learning_rate": 1.995508583399838e-05, "loss": 0.60712890625, "step": 2530 }, { "epoch": 0.059534011587451564, "grad_norm": Infinity, "learning_rate": 1.9954364302578598e-05, "loss": 0.6026504516601563, "step": 2540 }, { "epoch": 0.05976839745984311, "grad_norm": Infinity, "learning_rate": 1.9953637034918523e-05, "loss": 0.5859962463378906, "step": 2550 }, { "epoch": 0.06000278333223465, "grad_norm": Infinity, "learning_rate": 1.9952904031437236e-05, "loss": 0.6039627075195313, "step": 2560 }, { "epoch": 0.06023716920462619, "grad_norm": Infinity, "learning_rate": 1.9952165292557143e-05, "loss": 0.6072914123535156, "step": 2570 }, { "epoch": 0.060471555077017736, "grad_norm": Infinity, "learning_rate": 1.9951420818703935e-05, "loss": 0.6082473754882812, "step": 2580 }, { "epoch": 0.06070594094940927, "grad_norm": Infinity, "learning_rate": 1.9950670610306625e-05, "loss": 0.6094711303710938, "step": 2590 }, { "epoch": 0.06094032682180082, "grad_norm": Infinity, "learning_rate": 1.994991466779752e-05, "loss": 0.5962432861328125, "step": 2600 }, { "epoch": 0.061174712694192356, "grad_norm": Infinity, "learning_rate": 1.9949152991612236e-05, "loss": 0.6050712585449218, "step": 2610 }, { "epoch": 0.0614090985665839, "grad_norm": Infinity, "learning_rate": 1.9948385582189695e-05, "loss": 0.6007720947265625, "step": 2620 }, { "epoch": 0.06164348443897544, "grad_norm": Infinity, "learning_rate": 1.994761243997211e-05, "loss": 0.5951507568359375, "step": 2630 }, { "epoch": 0.06187787031136698, "grad_norm": Infinity, "learning_rate": 1.994683356540502e-05, "loss": 0.6106948852539062, "step": 2640 }, { "epoch": 0.06211225618375853, "grad_norm": Infinity, "learning_rate": 1.9946048958937245e-05, "loss": 0.610369873046875, "step": 2650 }, { "epoch": 0.062346642056150064, "grad_norm": Infinity, "learning_rate": 1.994525862102092e-05, "loss": 0.6162002563476563, "step": 2660 }, { "epoch": 0.0625810279285416, "grad_norm": Infinity, "learning_rate": 1.9944462552111482e-05, "loss": 0.6124046325683594, "step": 2670 }, { "epoch": 0.06281541380093315, "grad_norm": Infinity, "learning_rate": 1.9943660752667667e-05, "loss": 0.6085456848144531, "step": 2680 }, { "epoch": 0.06304979967332469, "grad_norm": Infinity, "learning_rate": 1.9942853223151514e-05, "loss": 0.6119369506835938, "step": 2690 }, { "epoch": 0.06328418554571623, "grad_norm": Infinity, "learning_rate": 1.994203996402837e-05, "loss": 0.5981208801269531, "step": 2700 }, { "epoch": 0.06351857141810778, "grad_norm": Infinity, "learning_rate": 1.9941220975766874e-05, "loss": 0.6123336791992188, "step": 2710 }, { "epoch": 0.06375295729049932, "grad_norm": Infinity, "learning_rate": 1.9940396258838966e-05, "loss": 0.6030326843261719, "step": 2720 }, { "epoch": 0.06398734316289086, "grad_norm": Infinity, "learning_rate": 1.9939565813719906e-05, "loss": 0.5945320129394531, "step": 2730 }, { "epoch": 0.0642217290352824, "grad_norm": Infinity, "learning_rate": 1.993872964088823e-05, "loss": 0.6185462951660157, "step": 2740 }, { "epoch": 0.06445611490767394, "grad_norm": Infinity, "learning_rate": 1.993788774082579e-05, "loss": 0.6118087768554688, "step": 2750 }, { "epoch": 0.06469050078006548, "grad_norm": Infinity, "learning_rate": 1.9937040114017735e-05, "loss": 0.5965530395507812, "step": 2760 }, { "epoch": 0.06492488665245702, "grad_norm": Infinity, "learning_rate": 1.9936186760952508e-05, "loss": 0.6073684692382812, "step": 2770 }, { "epoch": 0.06515927252484857, "grad_norm": Infinity, "learning_rate": 1.9935327682121866e-05, "loss": 0.6033073425292969, "step": 2780 }, { "epoch": 0.06539365839724011, "grad_norm": Infinity, "learning_rate": 1.9934462878020844e-05, "loss": 0.6004119873046875, "step": 2790 }, { "epoch": 0.06562804426963165, "grad_norm": Infinity, "learning_rate": 1.99335923491478e-05, "loss": 0.6079216003417969, "step": 2800 }, { "epoch": 0.06586243014202318, "grad_norm": Infinity, "learning_rate": 1.9932716096004375e-05, "loss": 0.6035270690917969, "step": 2810 }, { "epoch": 0.06609681601441474, "grad_norm": Infinity, "learning_rate": 1.9931834119095513e-05, "loss": 0.5924835205078125, "step": 2820 }, { "epoch": 0.06633120188680627, "grad_norm": Infinity, "learning_rate": 1.9930946418929457e-05, "loss": 0.6025344848632812, "step": 2830 }, { "epoch": 0.06656558775919781, "grad_norm": Infinity, "learning_rate": 1.9930052996017748e-05, "loss": 0.5995651245117187, "step": 2840 }, { "epoch": 0.06679997363158936, "grad_norm": Infinity, "learning_rate": 1.9929153850875225e-05, "loss": 0.6150283813476562, "step": 2850 }, { "epoch": 0.0670343595039809, "grad_norm": Infinity, "learning_rate": 1.992824898402002e-05, "loss": 0.6117012023925781, "step": 2860 }, { "epoch": 0.06726874537637244, "grad_norm": Infinity, "learning_rate": 1.992733839597357e-05, "loss": 0.6016891479492188, "step": 2870 }, { "epoch": 0.06750313124876398, "grad_norm": Infinity, "learning_rate": 1.9926422087260605e-05, "loss": 0.6109222412109375, "step": 2880 }, { "epoch": 0.06773751712115553, "grad_norm": Infinity, "learning_rate": 1.9925500058409147e-05, "loss": 0.6073577880859375, "step": 2890 }, { "epoch": 0.06797190299354706, "grad_norm": Infinity, "learning_rate": 1.9924572309950523e-05, "loss": 0.6049072265625, "step": 2900 }, { "epoch": 0.0682062888659386, "grad_norm": Infinity, "learning_rate": 1.992363884241935e-05, "loss": 0.6114601135253906, "step": 2910 }, { "epoch": 0.06844067473833015, "grad_norm": Infinity, "learning_rate": 1.9922699656353544e-05, "loss": 0.5975456237792969, "step": 2920 }, { "epoch": 0.06867506061072169, "grad_norm": Infinity, "learning_rate": 1.992175475229431e-05, "loss": 0.6088493347167969, "step": 2930 }, { "epoch": 0.06890944648311323, "grad_norm": Infinity, "learning_rate": 1.9920804130786154e-05, "loss": 0.6201919555664063, "step": 2940 }, { "epoch": 0.06914383235550477, "grad_norm": Infinity, "learning_rate": 1.9919847792376883e-05, "loss": 0.606353759765625, "step": 2950 }, { "epoch": 0.06937821822789632, "grad_norm": Infinity, "learning_rate": 1.991888573761758e-05, "loss": 0.6163558959960938, "step": 2960 }, { "epoch": 0.06961260410028786, "grad_norm": Infinity, "learning_rate": 1.991791796706264e-05, "loss": 0.6057174682617188, "step": 2970 }, { "epoch": 0.0698469899726794, "grad_norm": Infinity, "learning_rate": 1.9916944481269737e-05, "loss": 0.6093704223632812, "step": 2980 }, { "epoch": 0.07008137584507095, "grad_norm": Infinity, "learning_rate": 1.9915965280799853e-05, "loss": 0.615911865234375, "step": 2990 }, { "epoch": 0.07031576171746248, "grad_norm": Infinity, "learning_rate": 1.9914980366217253e-05, "loss": 0.6008621215820312, "step": 3000 } ], "logging_steps": 10, "max_steps": 42665, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 1.3152695723130618e+19, "train_batch_size": 1, "trial_name": null, "trial_params": null }