{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 5.0, "eval_steps": 500, "global_step": 1490, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0336417157275021, "grad_norm": 146.7955780029297, "learning_rate": 6.04026845637584e-07, "loss": 2.8527, "step": 10 }, { "epoch": 0.0672834314550042, "grad_norm": 71.41879272460938, "learning_rate": 1.2751677852348996e-06, "loss": 2.335, "step": 20 }, { "epoch": 0.10092514718250631, "grad_norm": 34.02959442138672, "learning_rate": 1.9463087248322147e-06, "loss": 1.254, "step": 30 }, { "epoch": 0.1345668629100084, "grad_norm": 60.82749938964844, "learning_rate": 2.6174496644295307e-06, "loss": 0.6452, "step": 40 }, { "epoch": 0.16820857863751051, "grad_norm": 17.56026268005371, "learning_rate": 3.2885906040268462e-06, "loss": 0.5344, "step": 50 }, { "epoch": 0.20185029436501262, "grad_norm": 26.253686904907227, "learning_rate": 3.959731543624161e-06, "loss": 0.43, "step": 60 }, { "epoch": 0.23549201009251472, "grad_norm": 20.974191665649414, "learning_rate": 4.6308724832214765e-06, "loss": 0.4868, "step": 70 }, { "epoch": 0.2691337258200168, "grad_norm": 31.288042068481445, "learning_rate": 5.302013422818793e-06, "loss": 0.4761, "step": 80 }, { "epoch": 0.3027754415475189, "grad_norm": 27.679262161254883, "learning_rate": 5.973154362416108e-06, "loss": 0.491, "step": 90 }, { "epoch": 0.33641715727502103, "grad_norm": 19.251558303833008, "learning_rate": 6.644295302013424e-06, "loss": 0.3406, "step": 100 }, { "epoch": 0.3700588730025231, "grad_norm": 22.788070678710938, "learning_rate": 7.3154362416107395e-06, "loss": 0.4966, "step": 110 }, { "epoch": 0.40370058873002523, "grad_norm": 17.20550537109375, "learning_rate": 7.986577181208053e-06, "loss": 0.3959, "step": 120 }, { "epoch": 0.4373423044575273, "grad_norm": 9.727134704589844, "learning_rate": 8.65771812080537e-06, "loss": 0.4267, "step": 130 }, { "epoch": 0.47098402018502944, "grad_norm": 21.130931854248047, "learning_rate": 9.328859060402684e-06, "loss": 0.4692, "step": 140 }, { "epoch": 0.5046257359125316, "grad_norm": 22.301837921142578, "learning_rate": 1e-05, "loss": 0.4706, "step": 150 }, { "epoch": 0.5382674516400336, "grad_norm": 17.202621459960938, "learning_rate": 9.998627974019322e-06, "loss": 0.4097, "step": 160 }, { "epoch": 0.5719091673675357, "grad_norm": 14.249327659606934, "learning_rate": 9.994512649059401e-06, "loss": 0.5236, "step": 170 }, { "epoch": 0.6055508830950378, "grad_norm": 17.96388816833496, "learning_rate": 9.987656283653344e-06, "loss": 0.4191, "step": 180 }, { "epoch": 0.63919259882254, "grad_norm": 19.75460433959961, "learning_rate": 9.978062640645737e-06, "loss": 0.5392, "step": 190 }, { "epoch": 0.6728343145500421, "grad_norm": 16.74843406677246, "learning_rate": 9.965736985127568e-06, "loss": 0.472, "step": 200 }, { "epoch": 0.7064760302775441, "grad_norm": 20.544544219970703, "learning_rate": 9.95068608154667e-06, "loss": 0.4416, "step": 210 }, { "epoch": 0.7401177460050462, "grad_norm": 7.9023542404174805, "learning_rate": 9.932918189995345e-06, "loss": 0.3824, "step": 220 }, { "epoch": 0.7737594617325484, "grad_norm": 18.007862091064453, "learning_rate": 9.912443061677125e-06, "loss": 0.4249, "step": 230 }, { "epoch": 0.8074011774600505, "grad_norm": 13.94369125366211, "learning_rate": 9.889271933555214e-06, "loss": 0.4905, "step": 240 }, { "epoch": 0.8410428931875525, "grad_norm": 20.00733757019043, "learning_rate": 9.863417522185525e-06, "loss": 0.3883, "step": 250 }, { "epoch": 0.8746846089150546, "grad_norm": 13.405632972717285, "learning_rate": 9.834894016737705e-06, "loss": 0.3839, "step": 260 }, { "epoch": 0.9083263246425568, "grad_norm": 24.92371940612793, "learning_rate": 9.803717071207965e-06, "loss": 0.4281, "step": 270 }, { "epoch": 0.9419680403700589, "grad_norm": 24.29949951171875, "learning_rate": 9.769903795828016e-06, "loss": 0.4549, "step": 280 }, { "epoch": 0.975609756097561, "grad_norm": 15.265240669250488, "learning_rate": 9.733472747674779e-06, "loss": 0.4838, "step": 290 }, { "epoch": 1.0067283431455005, "grad_norm": 16.928268432617188, "learning_rate": 9.694443920486083e-06, "loss": 0.5004, "step": 300 }, { "epoch": 1.0403700588730025, "grad_norm": 14.954211235046387, "learning_rate": 9.652838733687888e-06, "loss": 0.2062, "step": 310 }, { "epoch": 1.0740117746005047, "grad_norm": 13.043270111083984, "learning_rate": 9.608680020639081e-06, "loss": 0.2032, "step": 320 }, { "epoch": 1.1076534903280066, "grad_norm": 15.098630905151367, "learning_rate": 9.561992016100293e-06, "loss": 0.1926, "step": 330 }, { "epoch": 1.1412952060555088, "grad_norm": 16.94780731201172, "learning_rate": 9.512800342933608e-06, "loss": 0.1832, "step": 340 }, { "epoch": 1.174936921783011, "grad_norm": 16.518619537353516, "learning_rate": 9.461131998040473e-06, "loss": 0.1924, "step": 350 }, { "epoch": 1.208578637510513, "grad_norm": 13.654525756835938, "learning_rate": 9.40701533754552e-06, "loss": 0.1868, "step": 360 }, { "epoch": 1.2422203532380152, "grad_norm": 7.158348560333252, "learning_rate": 9.350480061234419e-06, "loss": 0.1929, "step": 370 }, { "epoch": 1.2758620689655173, "grad_norm": 15.472833633422852, "learning_rate": 9.291557196254342e-06, "loss": 0.1947, "step": 380 }, { "epoch": 1.3095037846930193, "grad_norm": 9.724027633666992, "learning_rate": 9.230279080085933e-06, "loss": 0.2044, "step": 390 }, { "epoch": 1.3431455004205215, "grad_norm": 3.915374755859375, "learning_rate": 9.166679342796162e-06, "loss": 0.1592, "step": 400 }, { "epoch": 1.3767872161480237, "grad_norm": 15.35324764251709, "learning_rate": 9.100792888581803e-06, "loss": 0.2242, "step": 410 }, { "epoch": 1.4104289318755256, "grad_norm": 11.980818748474121, "learning_rate": 9.032655876613636e-06, "loss": 0.2397, "step": 420 }, { "epoch": 1.4440706476030278, "grad_norm": 7.560662746429443, "learning_rate": 8.962305701191927e-06, "loss": 0.225, "step": 430 }, { "epoch": 1.4777123633305298, "grad_norm": 9.49743938446045, "learning_rate": 8.889780971224047e-06, "loss": 0.2332, "step": 440 }, { "epoch": 1.511354079058032, "grad_norm": 12.313446998596191, "learning_rate": 8.8151214890355e-06, "loss": 0.1712, "step": 450 }, { "epoch": 1.544995794785534, "grad_norm": 10.388385772705078, "learning_rate": 8.738368228525988e-06, "loss": 0.2514, "step": 460 }, { "epoch": 1.5786375105130361, "grad_norm": 22.984180450439453, "learning_rate": 8.659563312682524e-06, "loss": 0.1734, "step": 470 }, { "epoch": 1.6122792262405383, "grad_norm": 11.20322036743164, "learning_rate": 8.578749990461884e-06, "loss": 0.2604, "step": 480 }, { "epoch": 1.6459209419680403, "grad_norm": 13.423975944519043, "learning_rate": 8.495972613055137e-06, "loss": 0.1816, "step": 490 }, { "epoch": 1.6795626576955425, "grad_norm": 12.367583274841309, "learning_rate": 8.411276609547246e-06, "loss": 0.2098, "step": 500 }, { "epoch": 1.7132043734230447, "grad_norm": 22.17727279663086, "learning_rate": 8.324708461985124e-06, "loss": 0.2962, "step": 510 }, { "epoch": 1.7468460891505466, "grad_norm": 19.474199295043945, "learning_rate": 8.23631567986779e-06, "loss": 0.2505, "step": 520 }, { "epoch": 1.7804878048780488, "grad_norm": 11.064474105834961, "learning_rate": 8.146146774072674e-06, "loss": 0.2569, "step": 530 }, { "epoch": 1.814129520605551, "grad_norm": 13.412657737731934, "learning_rate": 8.054251230232333e-06, "loss": 0.2279, "step": 540 }, { "epoch": 1.847771236333053, "grad_norm": 13.497424125671387, "learning_rate": 7.960679481576233e-06, "loss": 0.2533, "step": 550 }, { "epoch": 1.8814129520605551, "grad_norm": 14.042634010314941, "learning_rate": 7.86548288125246e-06, "loss": 0.19, "step": 560 }, { "epoch": 1.9150546677880573, "grad_norm": 11.462506294250488, "learning_rate": 7.768713674144578e-06, "loss": 0.2569, "step": 570 }, { "epoch": 1.9486963835155593, "grad_norm": 11.845952033996582, "learning_rate": 7.670424968199099e-06, "loss": 0.1922, "step": 580 }, { "epoch": 1.9823380992430613, "grad_norm": 13.49905776977539, "learning_rate": 7.570670705279291e-06, "loss": 0.1995, "step": 590 }, { "epoch": 2.013456686291001, "grad_norm": 10.430764198303223, "learning_rate": 7.469505631561318e-06, "loss": 0.1615, "step": 600 }, { "epoch": 2.047098402018503, "grad_norm": 7.39661979675293, "learning_rate": 7.366985267488971e-06, "loss": 0.0886, "step": 610 }, { "epoch": 2.080740117746005, "grad_norm": 13.615009307861328, "learning_rate": 7.2631658773034715e-06, "loss": 0.0593, "step": 620 }, { "epoch": 2.114381833473507, "grad_norm": 14.060647964477539, "learning_rate": 7.1581044381650735e-06, "loss": 0.085, "step": 630 }, { "epoch": 2.1480235492010094, "grad_norm": 13.222198486328125, "learning_rate": 7.051858608883404e-06, "loss": 0.0654, "step": 640 }, { "epoch": 2.1816652649285113, "grad_norm": 7.685726642608643, "learning_rate": 6.944486698273704e-06, "loss": 0.0683, "step": 650 }, { "epoch": 2.2153069806560133, "grad_norm": 10.78395938873291, "learning_rate": 6.836047633156361e-06, "loss": 0.0836, "step": 660 }, { "epoch": 2.2489486963835157, "grad_norm": 3.9530932903289795, "learning_rate": 6.726600926017234e-06, "loss": 0.0701, "step": 670 }, { "epoch": 2.2825904121110177, "grad_norm": 9.838655471801758, "learning_rate": 6.616206642346603e-06, "loss": 0.0709, "step": 680 }, { "epoch": 2.3162321278385196, "grad_norm": 9.633439064025879, "learning_rate": 6.504925367674595e-06, "loss": 0.0422, "step": 690 }, { "epoch": 2.349873843566022, "grad_norm": 4.185202121734619, "learning_rate": 6.392818174321213e-06, "loss": 0.0708, "step": 700 }, { "epoch": 2.383515559293524, "grad_norm": 4.698387622833252, "learning_rate": 6.279946587879216e-06, "loss": 0.0356, "step": 710 }, { "epoch": 2.417157275021026, "grad_norm": 7.418368816375732, "learning_rate": 6.166372553448241e-06, "loss": 0.0342, "step": 720 }, { "epoch": 2.4507989907485284, "grad_norm": 5.932042121887207, "learning_rate": 6.0521584016386735e-06, "loss": 0.0615, "step": 730 }, { "epoch": 2.4844407064760303, "grad_norm": 2.541346549987793, "learning_rate": 5.9373668143639694e-06, "loss": 0.0882, "step": 740 }, { "epoch": 2.5180824222035323, "grad_norm": 6.172818183898926, "learning_rate": 5.8220607904401725e-06, "loss": 0.0727, "step": 750 }, { "epoch": 2.5517241379310347, "grad_norm": 8.212017059326172, "learning_rate": 5.706303611011502e-06, "loss": 0.0645, "step": 760 }, { "epoch": 2.5853658536585367, "grad_norm": 20.28338050842285, "learning_rate": 5.590158804821011e-06, "loss": 0.0528, "step": 770 }, { "epoch": 2.6190075693860386, "grad_norm": 12.685500144958496, "learning_rate": 5.473690113345343e-06, "loss": 0.0759, "step": 780 }, { "epoch": 2.652649285113541, "grad_norm": 2.813051700592041, "learning_rate": 5.356961455812754e-06, "loss": 0.0676, "step": 790 }, { "epoch": 2.686291000841043, "grad_norm": 4.1951165199279785, "learning_rate": 5.2400368941235745e-06, "loss": 0.0517, "step": 800 }, { "epoch": 2.719932716568545, "grad_norm": 10.7334566116333, "learning_rate": 5.122980597692372e-06, "loss": 0.0854, "step": 810 }, { "epoch": 2.7535744322960474, "grad_norm": 6.649056911468506, "learning_rate": 5.005856808231108e-06, "loss": 0.0929, "step": 820 }, { "epoch": 2.7872161480235493, "grad_norm": 7.63721227645874, "learning_rate": 4.88872980449262e-06, "loss": 0.0668, "step": 830 }, { "epoch": 2.8208578637510513, "grad_norm": 6.333030700683594, "learning_rate": 4.7716638669937784e-06, "loss": 0.0984, "step": 840 }, { "epoch": 2.8544995794785533, "grad_norm": 9.704402923583984, "learning_rate": 4.654723242737661e-06, "loss": 0.0842, "step": 850 }, { "epoch": 2.8881412952060557, "grad_norm": 9.808692932128906, "learning_rate": 4.5379721099541385e-06, "loss": 0.0847, "step": 860 }, { "epoch": 2.9217830109335576, "grad_norm": 4.981376647949219, "learning_rate": 4.4214745428781946e-06, "loss": 0.0637, "step": 870 }, { "epoch": 2.9554247266610596, "grad_norm": 18.70244026184082, "learning_rate": 4.305294476585312e-06, "loss": 0.0545, "step": 880 }, { "epoch": 2.9890664423885616, "grad_norm": 15.018904685974121, "learning_rate": 4.189495671903246e-06, "loss": 0.0736, "step": 890 }, { "epoch": 3.0201850294365014, "grad_norm": 3.326235294342041, "learning_rate": 4.074141680419422e-06, "loss": 0.0658, "step": 900 }, { "epoch": 3.0538267451640033, "grad_norm": 4.905099868774414, "learning_rate": 3.959295809603155e-06, "loss": 0.0231, "step": 910 }, { "epoch": 3.0874684608915053, "grad_norm": 11.493444442749023, "learning_rate": 3.845021088061858e-06, "loss": 0.0342, "step": 920 }, { "epoch": 3.1211101766190077, "grad_norm": 0.6767758131027222, "learning_rate": 3.731380230950288e-06, "loss": 0.0293, "step": 930 }, { "epoch": 3.1547518923465097, "grad_norm": 8.130093574523926, "learning_rate": 3.6184356055518143e-06, "loss": 0.0263, "step": 940 }, { "epoch": 3.1883936080740116, "grad_norm": 2.4990627765655518, "learning_rate": 3.506249197050604e-06, "loss": 0.0161, "step": 950 }, { "epoch": 3.222035323801514, "grad_norm": 8.8124418258667, "learning_rate": 3.3948825745135196e-06, "loss": 0.027, "step": 960 }, { "epoch": 3.255677039529016, "grad_norm": 9.777318000793457, "learning_rate": 3.284396857100357e-06, "loss": 0.0405, "step": 970 }, { "epoch": 3.289318755256518, "grad_norm": 8.85478687286377, "learning_rate": 3.174852680521032e-06, "loss": 0.0208, "step": 980 }, { "epoch": 3.3229604709840204, "grad_norm": 5.0749688148498535, "learning_rate": 3.0663101637580626e-06, "loss": 0.0302, "step": 990 }, { "epoch": 3.3566021867115223, "grad_norm": 18.818824768066406, "learning_rate": 2.958828876072654e-06, "loss": 0.0477, "step": 1000 }, { "epoch": 3.3902439024390243, "grad_norm": 6.966226100921631, "learning_rate": 2.852467804312463e-06, "loss": 0.025, "step": 1010 }, { "epoch": 3.4238856181665263, "grad_norm": 11.965907096862793, "learning_rate": 2.7472853205389997e-06, "loss": 0.0133, "step": 1020 }, { "epoch": 3.4575273338940287, "grad_norm": 6.451248645782471, "learning_rate": 2.64333914999245e-06, "loss": 0.022, "step": 1030 }, { "epoch": 3.4911690496215306, "grad_norm": 0.45957621932029724, "learning_rate": 2.540686339411446e-06, "loss": 0.0136, "step": 1040 }, { "epoch": 3.5248107653490326, "grad_norm": 11.877004623413086, "learning_rate": 2.4393832257252253e-06, "loss": 0.0186, "step": 1050 }, { "epoch": 3.558452481076535, "grad_norm": 10.027815818786621, "learning_rate": 2.3394854051353534e-06, "loss": 0.015, "step": 1060 }, { "epoch": 3.592094196804037, "grad_norm": 4.314270973205566, "learning_rate": 2.2410477026039335e-06, "loss": 0.0189, "step": 1070 }, { "epoch": 3.625735912531539, "grad_norm": 12.965841293334961, "learning_rate": 2.1441241417651072e-06, "loss": 0.0091, "step": 1080 }, { "epoch": 3.6593776282590413, "grad_norm": 5.011205196380615, "learning_rate": 2.0487679152763173e-06, "loss": 0.0081, "step": 1090 }, { "epoch": 3.6930193439865433, "grad_norm": 9.740224838256836, "learning_rate": 1.9550313556256294e-06, "loss": 0.0237, "step": 1100 }, { "epoch": 3.7266610597140453, "grad_norm": 0.04536225646734238, "learning_rate": 1.8629659064111138e-06, "loss": 0.0336, "step": 1110 }, { "epoch": 3.7603027754415477, "grad_norm": 5.120383262634277, "learning_rate": 1.772622094108074e-06, "loss": 0.0153, "step": 1120 }, { "epoch": 3.7939444911690496, "grad_norm": 7.686666011810303, "learning_rate": 1.6840495003395741e-06, "loss": 0.022, "step": 1130 }, { "epoch": 3.8275862068965516, "grad_norm": 0.6258940100669861, "learning_rate": 1.5972967346655449e-06, "loss": 0.0094, "step": 1140 }, { "epoch": 3.861227922624054, "grad_norm": 4.576685905456543, "learning_rate": 1.5124114079053492e-06, "loss": 0.0186, "step": 1150 }, { "epoch": 3.894869638351556, "grad_norm": 0.3393450081348419, "learning_rate": 1.4294401060084634e-06, "loss": 0.0473, "step": 1160 }, { "epoch": 3.928511354079058, "grad_norm": 7.82568359375, "learning_rate": 1.3484283644876289e-06, "loss": 0.0305, "step": 1170 }, { "epoch": 3.9621530698065603, "grad_norm": 2.9642977714538574, "learning_rate": 1.2694206434284878e-06, "loss": 0.0187, "step": 1180 }, { "epoch": 3.9957947855340623, "grad_norm": 1.896406650543213, "learning_rate": 1.1924603030894277e-06, "loss": 0.017, "step": 1190 }, { "epoch": 4.026913372582002, "grad_norm": 1.2707265615463257, "learning_rate": 1.1175895801050185e-06, "loss": 0.0044, "step": 1200 }, { "epoch": 4.060555088309504, "grad_norm": 0.6478537917137146, "learning_rate": 1.044849564306112e-06, "loss": 0.0012, "step": 1210 }, { "epoch": 4.094196804037006, "grad_norm": 0.07563937455415726, "learning_rate": 9.742801761693122e-07, "loss": 0.0057, "step": 1220 }, { "epoch": 4.1278385197645076, "grad_norm": 1.501904845237732, "learning_rate": 9.059201449082045e-07, "loss": 0.0089, "step": 1230 }, { "epoch": 4.16148023549201, "grad_norm": 0.10858506709337234, "learning_rate": 8.398069872183607e-07, "loss": 0.0008, "step": 1240 }, { "epoch": 4.195121951219512, "grad_norm": 0.3397045135498047, "learning_rate": 7.759769866877892e-07, "loss": 0.0017, "step": 1250 }, { "epoch": 4.228763666947014, "grad_norm": 0.1667356938123703, "learning_rate": 7.144651738841174e-07, "loss": 0.0069, "step": 1260 }, { "epoch": 4.262405382674516, "grad_norm": 0.09766753017902374, "learning_rate": 6.55305307129459e-07, "loss": 0.0037, "step": 1270 }, { "epoch": 4.296047098402019, "grad_norm": 0.07076102495193481, "learning_rate": 5.985298539734973e-07, "loss": 0.0011, "step": 1280 }, { "epoch": 4.32968881412952, "grad_norm": 0.09349212050437927, "learning_rate": 5.441699733749479e-07, "loss": 0.0022, "step": 1290 }, { "epoch": 4.363330529857023, "grad_norm": 1.491163730621338, "learning_rate": 4.922554986012068e-07, "loss": 0.0052, "step": 1300 }, { "epoch": 4.396972245584525, "grad_norm": 0.2895282506942749, "learning_rate": 4.428149208555388e-07, "loss": 0.0041, "step": 1310 }, { "epoch": 4.430613961312027, "grad_norm": 0.34550341963768005, "learning_rate": 3.958753736408105e-07, "loss": 0.0016, "step": 1320 }, { "epoch": 4.464255677039529, "grad_norm": 0.2858336865901947, "learning_rate": 3.5146261786834225e-07, "loss": 0.0019, "step": 1330 }, { "epoch": 4.497897392767031, "grad_norm": 2.363687515258789, "learning_rate": 3.0960102772005174e-07, "loss": 0.0036, "step": 1340 }, { "epoch": 4.531539108494533, "grad_norm": 0.8231300115585327, "learning_rate": 2.7031357727164865e-07, "loss": 0.001, "step": 1350 }, { "epoch": 4.565180824222035, "grad_norm": 0.06915020197629929, "learning_rate": 2.3362182788422395e-07, "loss": 0.0016, "step": 1360 }, { "epoch": 4.598822539949538, "grad_norm": 0.3646424412727356, "learning_rate": 1.9954591637115495e-07, "loss": 0.0031, "step": 1370 }, { "epoch": 4.632464255677039, "grad_norm": 8.59700870513916, "learning_rate": 1.6810454394680431e-07, "loss": 0.0035, "step": 1380 }, { "epoch": 4.666105971404542, "grad_norm": 0.7782748341560364, "learning_rate": 1.3931496596310545e-07, "loss": 0.0081, "step": 1390 }, { "epoch": 4.699747687132044, "grad_norm": 0.10166136920452118, "learning_rate": 1.1319298243964549e-07, "loss": 0.004, "step": 1400 }, { "epoch": 4.733389402859546, "grad_norm": 0.7652943134307861, "learning_rate": 8.975292939244928e-08, "loss": 0.0022, "step": 1410 }, { "epoch": 4.767031118587048, "grad_norm": 0.9027767181396484, "learning_rate": 6.900767096622352e-08, "loss": 0.0017, "step": 1420 }, { "epoch": 4.80067283431455, "grad_norm": 0.052162300795316696, "learning_rate": 5.0968592374384116e-08, "loss": 0.0027, "step": 1430 }, { "epoch": 4.834314550042052, "grad_norm": 1.7110869884490967, "learning_rate": 3.5645593650728284e-08, "loss": 0.0079, "step": 1440 }, { "epoch": 4.867956265769554, "grad_norm": 0.23056207597255707, "learning_rate": 2.3047084216196724e-08, "loss": 0.0131, "step": 1450 }, { "epoch": 4.901597981497057, "grad_norm": 2.3443868160247803, "learning_rate": 1.3179978263694326e-08, "loss": 0.0083, "step": 1460 }, { "epoch": 4.935239697224558, "grad_norm": 0.14247232675552368, "learning_rate": 6.04969096350938e-09, "loss": 0.0094, "step": 1470 }, { "epoch": 4.968881412952061, "grad_norm": 0.3539099395275116, "learning_rate": 1.660135491411974e-09, "loss": 0.0032, "step": 1480 }, { "epoch": 5.0, "grad_norm": 0.3966084122657776, "learning_rate": 1.3720881062440073e-11, "loss": 0.0083, "step": 1490 } ], "logging_steps": 10, "max_steps": 1490, "num_input_tokens_seen": 0, "num_train_epochs": 5, "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": 6911780082155520.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }