{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 17.457084666860634, "eval_steps": 500, "global_step": 60000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.002909514111143439, "grad_norm": 5.517890930175781, "learning_rate": 9e-08, "loss": 0.5714, "step": 10 }, { "epoch": 0.005819028222286878, "grad_norm": 5.328445911407471, "learning_rate": 1.9e-07, "loss": 0.5387, "step": 20 }, { "epoch": 0.008728542333430317, "grad_norm": 5.409641742706299, "learning_rate": 2.9000000000000003e-07, "loss": 0.5693, "step": 30 }, { "epoch": 0.011638056444573755, "grad_norm": 5.405916690826416, "learning_rate": 3.8999999999999997e-07, "loss": 0.5621, "step": 40 }, { "epoch": 0.014547570555717195, "grad_norm": 5.031722545623779, "learning_rate": 4.9e-07, "loss": 0.5143, "step": 50 }, { "epoch": 0.017457084666860634, "grad_norm": 4.626715183258057, "learning_rate": 5.9e-07, "loss": 0.4931, "step": 60 }, { "epoch": 0.020366598778004074, "grad_norm": 4.105497360229492, "learning_rate": 6.9e-07, "loss": 0.4371, "step": 70 }, { "epoch": 0.02327611288914751, "grad_norm": 3.1085870265960693, "learning_rate": 7.900000000000001e-07, "loss": 0.4027, "step": 80 }, { "epoch": 0.02618562700029095, "grad_norm": 1.8463505506515503, "learning_rate": 8.900000000000001e-07, "loss": 0.268, "step": 90 }, { "epoch": 0.02909514111143439, "grad_norm": 1.047481894493103, "learning_rate": 9.9e-07, "loss": 0.233, "step": 100 }, { "epoch": 0.03200465522257783, "grad_norm": 0.7391247749328613, "learning_rate": 1.0900000000000002e-06, "loss": 0.2102, "step": 110 }, { "epoch": 0.03491416933372127, "grad_norm": 0.6765145659446716, "learning_rate": 1.19e-06, "loss": 0.2009, "step": 120 }, { "epoch": 0.037823683444864704, "grad_norm": 0.48983868956565857, "learning_rate": 1.29e-06, "loss": 0.171, "step": 130 }, { "epoch": 0.04073319755600815, "grad_norm": 0.5598902106285095, "learning_rate": 1.39e-06, "loss": 0.1617, "step": 140 }, { "epoch": 0.043642711667151585, "grad_norm": 0.5578387379646301, "learning_rate": 1.49e-06, "loss": 0.1619, "step": 150 }, { "epoch": 0.04655222577829502, "grad_norm": 0.46492764353752136, "learning_rate": 1.59e-06, "loss": 0.1519, "step": 160 }, { "epoch": 0.049461739889438465, "grad_norm": 0.2962999641895294, "learning_rate": 1.69e-06, "loss": 0.1531, "step": 170 }, { "epoch": 0.0523712540005819, "grad_norm": 0.41179507970809937, "learning_rate": 1.79e-06, "loss": 0.1415, "step": 180 }, { "epoch": 0.055280768111725345, "grad_norm": 0.36194145679473877, "learning_rate": 1.8900000000000001e-06, "loss": 0.135, "step": 190 }, { "epoch": 0.05819028222286878, "grad_norm": 0.36306336522102356, "learning_rate": 1.99e-06, "loss": 0.1305, "step": 200 }, { "epoch": 0.06109979633401222, "grad_norm": 0.40288788080215454, "learning_rate": 2.09e-06, "loss": 0.13, "step": 210 }, { "epoch": 0.06400931044515566, "grad_norm": 0.4192882776260376, "learning_rate": 2.1899999999999998e-06, "loss": 0.1365, "step": 220 }, { "epoch": 0.06691882455629909, "grad_norm": 0.4094921052455902, "learning_rate": 2.29e-06, "loss": 0.1277, "step": 230 }, { "epoch": 0.06982833866744254, "grad_norm": 0.38068294525146484, "learning_rate": 2.39e-06, "loss": 0.1225, "step": 240 }, { "epoch": 0.07273785277858598, "grad_norm": 0.36606907844543457, "learning_rate": 2.4900000000000003e-06, "loss": 0.1204, "step": 250 }, { "epoch": 0.07564736688972941, "grad_norm": 0.33838096261024475, "learning_rate": 2.59e-06, "loss": 0.1194, "step": 260 }, { "epoch": 0.07855688100087285, "grad_norm": 0.33277052640914917, "learning_rate": 2.69e-06, "loss": 0.123, "step": 270 }, { "epoch": 0.0814663951120163, "grad_norm": 0.3669566512107849, "learning_rate": 2.79e-06, "loss": 0.1146, "step": 280 }, { "epoch": 0.08437590922315973, "grad_norm": 0.31503039598464966, "learning_rate": 2.8900000000000003e-06, "loss": 0.1174, "step": 290 }, { "epoch": 0.08728542333430317, "grad_norm": 0.44869911670684814, "learning_rate": 2.99e-06, "loss": 0.1106, "step": 300 }, { "epoch": 0.09019493744544661, "grad_norm": 0.3286576569080353, "learning_rate": 3.09e-06, "loss": 0.1078, "step": 310 }, { "epoch": 0.09310445155659004, "grad_norm": 0.4365527331829071, "learning_rate": 3.19e-06, "loss": 0.1105, "step": 320 }, { "epoch": 0.09601396566773349, "grad_norm": 0.4132164418697357, "learning_rate": 3.29e-06, "loss": 0.111, "step": 330 }, { "epoch": 0.09892347977887693, "grad_norm": 0.38617241382598877, "learning_rate": 3.39e-06, "loss": 0.1023, "step": 340 }, { "epoch": 0.10183299389002037, "grad_norm": 0.39552491903305054, "learning_rate": 3.49e-06, "loss": 0.1004, "step": 350 }, { "epoch": 0.1047425080011638, "grad_norm": 0.5855547785758972, "learning_rate": 3.5900000000000004e-06, "loss": 0.1041, "step": 360 }, { "epoch": 0.10765202211230725, "grad_norm": 0.3020547330379486, "learning_rate": 3.6900000000000002e-06, "loss": 0.1018, "step": 370 }, { "epoch": 0.11056153622345069, "grad_norm": 0.47132548689842224, "learning_rate": 3.7899999999999997e-06, "loss": 0.1046, "step": 380 }, { "epoch": 0.11347105033459412, "grad_norm": 0.5439547896385193, "learning_rate": 3.890000000000001e-06, "loss": 0.1062, "step": 390 }, { "epoch": 0.11638056444573756, "grad_norm": 0.41928598284721375, "learning_rate": 3.99e-06, "loss": 0.1005, "step": 400 }, { "epoch": 0.11929007855688101, "grad_norm": 0.4094143211841583, "learning_rate": 4.09e-06, "loss": 0.096, "step": 410 }, { "epoch": 0.12219959266802444, "grad_norm": 0.4680388569831848, "learning_rate": 4.19e-06, "loss": 0.0957, "step": 420 }, { "epoch": 0.12510910677916787, "grad_norm": 0.4811868369579315, "learning_rate": 4.29e-06, "loss": 0.099, "step": 430 }, { "epoch": 0.12801862089031132, "grad_norm": 0.4884318709373474, "learning_rate": 4.39e-06, "loss": 0.0973, "step": 440 }, { "epoch": 0.13092813500145475, "grad_norm": 0.565917432308197, "learning_rate": 4.49e-06, "loss": 0.0935, "step": 450 }, { "epoch": 0.13383764911259818, "grad_norm": 0.35832539200782776, "learning_rate": 4.59e-06, "loss": 0.0927, "step": 460 }, { "epoch": 0.13674716322374164, "grad_norm": 0.37413477897644043, "learning_rate": 4.69e-06, "loss": 0.0935, "step": 470 }, { "epoch": 0.13965667733488507, "grad_norm": 0.4468092620372772, "learning_rate": 4.790000000000001e-06, "loss": 0.0908, "step": 480 }, { "epoch": 0.1425661914460285, "grad_norm": 0.3869225084781647, "learning_rate": 4.890000000000001e-06, "loss": 0.0893, "step": 490 }, { "epoch": 0.14547570555717196, "grad_norm": 0.29299911856651306, "learning_rate": 4.9900000000000005e-06, "loss": 0.0947, "step": 500 }, { "epoch": 0.1483852196683154, "grad_norm": 0.39316126704216003, "learning_rate": 5.0899999999999995e-06, "loss": 0.089, "step": 510 }, { "epoch": 0.15129473377945882, "grad_norm": 0.34199607372283936, "learning_rate": 5.1899999999999994e-06, "loss": 0.0827, "step": 520 }, { "epoch": 0.15420424789060228, "grad_norm": 0.4122438430786133, "learning_rate": 5.29e-06, "loss": 0.0822, "step": 530 }, { "epoch": 0.1571137620017457, "grad_norm": 0.3321143686771393, "learning_rate": 5.39e-06, "loss": 0.0881, "step": 540 }, { "epoch": 0.16002327611288913, "grad_norm": 0.41549941897392273, "learning_rate": 5.49e-06, "loss": 0.0856, "step": 550 }, { "epoch": 0.1629327902240326, "grad_norm": 0.36289432644844055, "learning_rate": 5.59e-06, "loss": 0.0832, "step": 560 }, { "epoch": 0.16584230433517602, "grad_norm": 0.3584991693496704, "learning_rate": 5.690000000000001e-06, "loss": 0.0895, "step": 570 }, { "epoch": 0.16875181844631945, "grad_norm": 0.39339351654052734, "learning_rate": 5.7900000000000005e-06, "loss": 0.0806, "step": 580 }, { "epoch": 0.1716613325574629, "grad_norm": 0.3806701600551605, "learning_rate": 5.89e-06, "loss": 0.0871, "step": 590 }, { "epoch": 0.17457084666860634, "grad_norm": 0.3422127366065979, "learning_rate": 5.99e-06, "loss": 0.0847, "step": 600 }, { "epoch": 0.17748036077974977, "grad_norm": 0.28945115208625793, "learning_rate": 6.090000000000001e-06, "loss": 0.0797, "step": 610 }, { "epoch": 0.18038987489089323, "grad_norm": 0.3700658977031708, "learning_rate": 6.19e-06, "loss": 0.0834, "step": 620 }, { "epoch": 0.18329938900203666, "grad_norm": 0.34670495986938477, "learning_rate": 6.29e-06, "loss": 0.0788, "step": 630 }, { "epoch": 0.18620890311318009, "grad_norm": 0.3959951102733612, "learning_rate": 6.39e-06, "loss": 0.0803, "step": 640 }, { "epoch": 0.18911841722432354, "grad_norm": 0.26546362042427063, "learning_rate": 6.49e-06, "loss": 0.0795, "step": 650 }, { "epoch": 0.19202793133546697, "grad_norm": 0.5597039461135864, "learning_rate": 6.5900000000000004e-06, "loss": 0.0744, "step": 660 }, { "epoch": 0.1949374454466104, "grad_norm": 0.38635051250457764, "learning_rate": 6.69e-06, "loss": 0.0816, "step": 670 }, { "epoch": 0.19784695955775386, "grad_norm": 0.3380928635597229, "learning_rate": 6.79e-06, "loss": 0.0798, "step": 680 }, { "epoch": 0.2007564736688973, "grad_norm": 0.4169606864452362, "learning_rate": 6.89e-06, "loss": 0.0751, "step": 690 }, { "epoch": 0.20366598778004075, "grad_norm": 0.5294784307479858, "learning_rate": 6.990000000000001e-06, "loss": 0.0742, "step": 700 }, { "epoch": 0.20657550189118418, "grad_norm": 0.37085092067718506, "learning_rate": 7.090000000000001e-06, "loss": 0.0767, "step": 710 }, { "epoch": 0.2094850160023276, "grad_norm": 0.3811768591403961, "learning_rate": 7.19e-06, "loss": 0.0728, "step": 720 }, { "epoch": 0.21239453011347106, "grad_norm": 0.4438173770904541, "learning_rate": 7.29e-06, "loss": 0.0759, "step": 730 }, { "epoch": 0.2153040442246145, "grad_norm": 0.4014468491077423, "learning_rate": 7.3899999999999995e-06, "loss": 0.0734, "step": 740 }, { "epoch": 0.21821355833575792, "grad_norm": 0.3995489180088043, "learning_rate": 7.49e-06, "loss": 0.0755, "step": 750 }, { "epoch": 0.22112307244690138, "grad_norm": 0.35601791739463806, "learning_rate": 7.59e-06, "loss": 0.0743, "step": 760 }, { "epoch": 0.2240325865580448, "grad_norm": 0.5403217673301697, "learning_rate": 7.690000000000001e-06, "loss": 0.0751, "step": 770 }, { "epoch": 0.22694210066918824, "grad_norm": 0.45096123218536377, "learning_rate": 7.79e-06, "loss": 0.0754, "step": 780 }, { "epoch": 0.2298516147803317, "grad_norm": 0.29106324911117554, "learning_rate": 7.89e-06, "loss": 0.0731, "step": 790 }, { "epoch": 0.23276112889147513, "grad_norm": 0.2719796299934387, "learning_rate": 7.99e-06, "loss": 0.0734, "step": 800 }, { "epoch": 0.23567064300261856, "grad_norm": 0.3572279214859009, "learning_rate": 8.09e-06, "loss": 0.07, "step": 810 }, { "epoch": 0.23858015711376201, "grad_norm": 0.31937333941459656, "learning_rate": 8.190000000000001e-06, "loss": 0.0718, "step": 820 }, { "epoch": 0.24148967122490544, "grad_norm": 0.2887006402015686, "learning_rate": 8.29e-06, "loss": 0.0679, "step": 830 }, { "epoch": 0.24439918533604887, "grad_norm": 0.24929392337799072, "learning_rate": 8.390000000000001e-06, "loss": 0.0698, "step": 840 }, { "epoch": 0.24730869944719233, "grad_norm": 0.2879127860069275, "learning_rate": 8.49e-06, "loss": 0.0747, "step": 850 }, { "epoch": 0.25021821355833573, "grad_norm": 0.2651780843734741, "learning_rate": 8.59e-06, "loss": 0.0717, "step": 860 }, { "epoch": 0.2531277276694792, "grad_norm": 0.2889194190502167, "learning_rate": 8.690000000000002e-06, "loss": 0.0684, "step": 870 }, { "epoch": 0.25603724178062265, "grad_norm": 0.28309887647628784, "learning_rate": 8.79e-06, "loss": 0.0706, "step": 880 }, { "epoch": 0.2589467558917661, "grad_norm": 0.43961426615715027, "learning_rate": 8.89e-06, "loss": 0.07, "step": 890 }, { "epoch": 0.2618562700029095, "grad_norm": 0.35015812516212463, "learning_rate": 8.99e-06, "loss": 0.0708, "step": 900 }, { "epoch": 0.26476578411405294, "grad_norm": 0.4506077468395233, "learning_rate": 9.09e-06, "loss": 0.0675, "step": 910 }, { "epoch": 0.26767529822519637, "grad_norm": 0.3825240135192871, "learning_rate": 9.19e-06, "loss": 0.0688, "step": 920 }, { "epoch": 0.27058481233633985, "grad_norm": 0.27108654379844666, "learning_rate": 9.289999999999999e-06, "loss": 0.0704, "step": 930 }, { "epoch": 0.2734943264474833, "grad_norm": 0.383567750453949, "learning_rate": 9.39e-06, "loss": 0.0698, "step": 940 }, { "epoch": 0.2764038405586267, "grad_norm": 0.40884849429130554, "learning_rate": 9.49e-06, "loss": 0.0695, "step": 950 }, { "epoch": 0.27931335466977014, "grad_norm": 0.3573484420776367, "learning_rate": 9.59e-06, "loss": 0.0696, "step": 960 }, { "epoch": 0.28222286878091357, "grad_norm": 0.2750702202320099, "learning_rate": 9.69e-06, "loss": 0.0676, "step": 970 }, { "epoch": 0.285132382892057, "grad_norm": 0.35659530758857727, "learning_rate": 9.79e-06, "loss": 0.066, "step": 980 }, { "epoch": 0.2880418970032005, "grad_norm": 0.34817540645599365, "learning_rate": 9.89e-06, "loss": 0.0682, "step": 990 }, { "epoch": 0.2909514111143439, "grad_norm": 0.22359809279441833, "learning_rate": 9.990000000000001e-06, "loss": 0.0663, "step": 1000 }, { "epoch": 0.29386092522548735, "grad_norm": 0.3841455280780792, "learning_rate": 1.009e-05, "loss": 0.0688, "step": 1010 }, { "epoch": 0.2967704393366308, "grad_norm": 0.3679603338241577, "learning_rate": 1.019e-05, "loss": 0.0661, "step": 1020 }, { "epoch": 0.2996799534477742, "grad_norm": 0.22877363860607147, "learning_rate": 1.0290000000000001e-05, "loss": 0.065, "step": 1030 }, { "epoch": 0.30258946755891764, "grad_norm": 0.33624470233917236, "learning_rate": 1.039e-05, "loss": 0.066, "step": 1040 }, { "epoch": 0.3054989816700611, "grad_norm": 0.3473489582538605, "learning_rate": 1.0490000000000001e-05, "loss": 0.0671, "step": 1050 }, { "epoch": 0.30840849578120455, "grad_norm": 0.3758750855922699, "learning_rate": 1.059e-05, "loss": 0.0649, "step": 1060 }, { "epoch": 0.311318009892348, "grad_norm": 0.28316089510917664, "learning_rate": 1.0690000000000001e-05, "loss": 0.0652, "step": 1070 }, { "epoch": 0.3142275240034914, "grad_norm": 0.3210902214050293, "learning_rate": 1.0790000000000002e-05, "loss": 0.0612, "step": 1080 }, { "epoch": 0.31713703811463484, "grad_norm": 0.4336917996406555, "learning_rate": 1.089e-05, "loss": 0.0636, "step": 1090 }, { "epoch": 0.32004655222577827, "grad_norm": 0.37814727425575256, "learning_rate": 1.099e-05, "loss": 0.0671, "step": 1100 }, { "epoch": 0.32295606633692175, "grad_norm": 0.3199159502983093, "learning_rate": 1.1089999999999999e-05, "loss": 0.0642, "step": 1110 }, { "epoch": 0.3258655804480652, "grad_norm": 0.31301024556159973, "learning_rate": 1.119e-05, "loss": 0.0625, "step": 1120 }, { "epoch": 0.3287750945592086, "grad_norm": 0.33653488755226135, "learning_rate": 1.129e-05, "loss": 0.0655, "step": 1130 }, { "epoch": 0.33168460867035204, "grad_norm": 0.37012001872062683, "learning_rate": 1.139e-05, "loss": 0.0608, "step": 1140 }, { "epoch": 0.3345941227814955, "grad_norm": 0.44625362753868103, "learning_rate": 1.149e-05, "loss": 0.0644, "step": 1150 }, { "epoch": 0.3375036368926389, "grad_norm": 0.3204994201660156, "learning_rate": 1.1589999999999999e-05, "loss": 0.0664, "step": 1160 }, { "epoch": 0.3404131510037824, "grad_norm": 0.4755965769290924, "learning_rate": 1.169e-05, "loss": 0.0633, "step": 1170 }, { "epoch": 0.3433226651149258, "grad_norm": 0.32921740412712097, "learning_rate": 1.179e-05, "loss": 0.0654, "step": 1180 }, { "epoch": 0.34623217922606925, "grad_norm": 0.25423866510391235, "learning_rate": 1.189e-05, "loss": 0.0612, "step": 1190 }, { "epoch": 0.3491416933372127, "grad_norm": 0.3001701533794403, "learning_rate": 1.199e-05, "loss": 0.0637, "step": 1200 }, { "epoch": 0.3520512074483561, "grad_norm": 0.33879053592681885, "learning_rate": 1.2090000000000001e-05, "loss": 0.0621, "step": 1210 }, { "epoch": 0.35496072155949954, "grad_norm": 0.3057038486003876, "learning_rate": 1.219e-05, "loss": 0.0614, "step": 1220 }, { "epoch": 0.357870235670643, "grad_norm": 0.25112810730934143, "learning_rate": 1.2290000000000001e-05, "loss": 0.062, "step": 1230 }, { "epoch": 0.36077974978178645, "grad_norm": 0.2705489695072174, "learning_rate": 1.239e-05, "loss": 0.0618, "step": 1240 }, { "epoch": 0.3636892638929299, "grad_norm": 0.2748051881790161, "learning_rate": 1.249e-05, "loss": 0.0629, "step": 1250 }, { "epoch": 0.3665987780040733, "grad_norm": 0.2777414321899414, "learning_rate": 1.2590000000000001e-05, "loss": 0.0627, "step": 1260 }, { "epoch": 0.36950829211521674, "grad_norm": 0.2427351325750351, "learning_rate": 1.269e-05, "loss": 0.0606, "step": 1270 }, { "epoch": 0.37241780622636017, "grad_norm": 0.2646106779575348, "learning_rate": 1.2790000000000001e-05, "loss": 0.0658, "step": 1280 }, { "epoch": 0.37532732033750366, "grad_norm": 0.24326178431510925, "learning_rate": 1.289e-05, "loss": 0.0584, "step": 1290 }, { "epoch": 0.3782368344486471, "grad_norm": 0.3852478861808777, "learning_rate": 1.2990000000000001e-05, "loss": 0.0605, "step": 1300 }, { "epoch": 0.3811463485597905, "grad_norm": 0.24830475449562073, "learning_rate": 1.309e-05, "loss": 0.06, "step": 1310 }, { "epoch": 0.38405586267093395, "grad_norm": 0.3178030550479889, "learning_rate": 1.3189999999999999e-05, "loss": 0.0613, "step": 1320 }, { "epoch": 0.3869653767820774, "grad_norm": 0.29359832406044006, "learning_rate": 1.329e-05, "loss": 0.0623, "step": 1330 }, { "epoch": 0.3898748908932208, "grad_norm": 0.2708103656768799, "learning_rate": 1.339e-05, "loss": 0.0589, "step": 1340 }, { "epoch": 0.3927844050043643, "grad_norm": 0.2994367480278015, "learning_rate": 1.349e-05, "loss": 0.0637, "step": 1350 }, { "epoch": 0.3956939191155077, "grad_norm": 0.3507819175720215, "learning_rate": 1.359e-05, "loss": 0.0621, "step": 1360 }, { "epoch": 0.39860343322665115, "grad_norm": 0.26022544503211975, "learning_rate": 1.369e-05, "loss": 0.0607, "step": 1370 }, { "epoch": 0.4015129473377946, "grad_norm": 0.3776799142360687, "learning_rate": 1.379e-05, "loss": 0.0575, "step": 1380 }, { "epoch": 0.404422461448938, "grad_norm": 0.26443183422088623, "learning_rate": 1.389e-05, "loss": 0.0603, "step": 1390 }, { "epoch": 0.4073319755600815, "grad_norm": 0.2691558301448822, "learning_rate": 1.399e-05, "loss": 0.0587, "step": 1400 }, { "epoch": 0.4102414896712249, "grad_norm": 0.27407073974609375, "learning_rate": 1.409e-05, "loss": 0.0629, "step": 1410 }, { "epoch": 0.41315100378236835, "grad_norm": 0.38113343715667725, "learning_rate": 1.419e-05, "loss": 0.0575, "step": 1420 }, { "epoch": 0.4160605178935118, "grad_norm": 0.2773817181587219, "learning_rate": 1.429e-05, "loss": 0.06, "step": 1430 }, { "epoch": 0.4189700320046552, "grad_norm": 0.24384526908397675, "learning_rate": 1.4390000000000001e-05, "loss": 0.0572, "step": 1440 }, { "epoch": 0.42187954611579864, "grad_norm": 0.3030109405517578, "learning_rate": 1.449e-05, "loss": 0.0598, "step": 1450 }, { "epoch": 0.42478906022694213, "grad_norm": 0.2166089415550232, "learning_rate": 1.4590000000000001e-05, "loss": 0.059, "step": 1460 }, { "epoch": 0.42769857433808556, "grad_norm": 0.29942983388900757, "learning_rate": 1.469e-05, "loss": 0.061, "step": 1470 }, { "epoch": 0.430608088449229, "grad_norm": 0.225367933511734, "learning_rate": 1.479e-05, "loss": 0.0575, "step": 1480 }, { "epoch": 0.4335176025603724, "grad_norm": 0.23812495172023773, "learning_rate": 1.4890000000000001e-05, "loss": 0.056, "step": 1490 }, { "epoch": 0.43642711667151585, "grad_norm": 0.25749635696411133, "learning_rate": 1.499e-05, "loss": 0.0573, "step": 1500 }, { "epoch": 0.4393366307826593, "grad_norm": 0.30408936738967896, "learning_rate": 1.5090000000000001e-05, "loss": 0.0577, "step": 1510 }, { "epoch": 0.44224614489380276, "grad_norm": 0.3318593502044678, "learning_rate": 1.519e-05, "loss": 0.0595, "step": 1520 }, { "epoch": 0.4451556590049462, "grad_norm": 0.23338373005390167, "learning_rate": 1.529e-05, "loss": 0.058, "step": 1530 }, { "epoch": 0.4480651731160896, "grad_norm": 0.23514990508556366, "learning_rate": 1.539e-05, "loss": 0.0564, "step": 1540 }, { "epoch": 0.45097468722723305, "grad_norm": 0.24689163267612457, "learning_rate": 1.549e-05, "loss": 0.0593, "step": 1550 }, { "epoch": 0.4538842013383765, "grad_norm": 0.24164582788944244, "learning_rate": 1.559e-05, "loss": 0.0581, "step": 1560 }, { "epoch": 0.4567937154495199, "grad_norm": 0.23757606744766235, "learning_rate": 1.569e-05, "loss": 0.0582, "step": 1570 }, { "epoch": 0.4597032295606634, "grad_norm": 0.3562818169593811, "learning_rate": 1.579e-05, "loss": 0.0556, "step": 1580 }, { "epoch": 0.4626127436718068, "grad_norm": 0.2803914546966553, "learning_rate": 1.589e-05, "loss": 0.0546, "step": 1590 }, { "epoch": 0.46552225778295026, "grad_norm": 0.32448405027389526, "learning_rate": 1.599e-05, "loss": 0.0573, "step": 1600 }, { "epoch": 0.4684317718940937, "grad_norm": 0.4047822952270508, "learning_rate": 1.609e-05, "loss": 0.0567, "step": 1610 }, { "epoch": 0.4713412860052371, "grad_norm": 0.261622816324234, "learning_rate": 1.619e-05, "loss": 0.0531, "step": 1620 }, { "epoch": 0.47425080011638054, "grad_norm": 0.28579971194267273, "learning_rate": 1.629e-05, "loss": 0.0566, "step": 1630 }, { "epoch": 0.47716031422752403, "grad_norm": 0.2897278964519501, "learning_rate": 1.639e-05, "loss": 0.0575, "step": 1640 }, { "epoch": 0.48006982833866746, "grad_norm": 0.31465643644332886, "learning_rate": 1.649e-05, "loss": 0.0607, "step": 1650 }, { "epoch": 0.4829793424498109, "grad_norm": 0.3291495442390442, "learning_rate": 1.6590000000000002e-05, "loss": 0.0562, "step": 1660 }, { "epoch": 0.4858888565609543, "grad_norm": 0.3070833683013916, "learning_rate": 1.669e-05, "loss": 0.0579, "step": 1670 }, { "epoch": 0.48879837067209775, "grad_norm": 0.23706436157226562, "learning_rate": 1.679e-05, "loss": 0.0538, "step": 1680 }, { "epoch": 0.4917078847832412, "grad_norm": 0.34867486357688904, "learning_rate": 1.689e-05, "loss": 0.0581, "step": 1690 }, { "epoch": 0.49461739889438466, "grad_norm": 0.23711511492729187, "learning_rate": 1.699e-05, "loss": 0.0567, "step": 1700 }, { "epoch": 0.4975269130055281, "grad_norm": 0.2632048726081848, "learning_rate": 1.709e-05, "loss": 0.0505, "step": 1710 }, { "epoch": 0.5004364271166715, "grad_norm": 0.3280385136604309, "learning_rate": 1.719e-05, "loss": 0.0593, "step": 1720 }, { "epoch": 0.503345941227815, "grad_norm": 0.23751583695411682, "learning_rate": 1.7290000000000002e-05, "loss": 0.0577, "step": 1730 }, { "epoch": 0.5062554553389584, "grad_norm": 0.24453946948051453, "learning_rate": 1.739e-05, "loss": 0.056, "step": 1740 }, { "epoch": 0.5091649694501018, "grad_norm": 0.3070123791694641, "learning_rate": 1.749e-05, "loss": 0.0553, "step": 1750 }, { "epoch": 0.5120744835612453, "grad_norm": 0.3147856593132019, "learning_rate": 1.7590000000000003e-05, "loss": 0.0567, "step": 1760 }, { "epoch": 0.5149839976723887, "grad_norm": 0.21862418949604034, "learning_rate": 1.7690000000000002e-05, "loss": 0.0552, "step": 1770 }, { "epoch": 0.5178935117835322, "grad_norm": 0.19423121213912964, "learning_rate": 1.779e-05, "loss": 0.0548, "step": 1780 }, { "epoch": 0.5208030258946756, "grad_norm": 0.2867182195186615, "learning_rate": 1.7890000000000003e-05, "loss": 0.0559, "step": 1790 }, { "epoch": 0.523712540005819, "grad_norm": 0.2617549002170563, "learning_rate": 1.7990000000000002e-05, "loss": 0.0582, "step": 1800 }, { "epoch": 0.5266220541169625, "grad_norm": 0.3316369354724884, "learning_rate": 1.809e-05, "loss": 0.055, "step": 1810 }, { "epoch": 0.5295315682281059, "grad_norm": 0.21327628195285797, "learning_rate": 1.819e-05, "loss": 0.0536, "step": 1820 }, { "epoch": 0.5324410823392494, "grad_norm": 0.21412546932697296, "learning_rate": 1.8290000000000003e-05, "loss": 0.0563, "step": 1830 }, { "epoch": 0.5353505964503927, "grad_norm": 0.2152285873889923, "learning_rate": 1.8390000000000002e-05, "loss": 0.0535, "step": 1840 }, { "epoch": 0.5382601105615362, "grad_norm": 0.3087080121040344, "learning_rate": 1.8489999999999997e-05, "loss": 0.055, "step": 1850 }, { "epoch": 0.5411696246726797, "grad_norm": 0.25082120299339294, "learning_rate": 1.859e-05, "loss": 0.0566, "step": 1860 }, { "epoch": 0.5440791387838231, "grad_norm": 0.19333535432815552, "learning_rate": 1.869e-05, "loss": 0.0529, "step": 1870 }, { "epoch": 0.5469886528949666, "grad_norm": 0.22656840085983276, "learning_rate": 1.8789999999999998e-05, "loss": 0.0554, "step": 1880 }, { "epoch": 0.5498981670061099, "grad_norm": 0.3016730844974518, "learning_rate": 1.889e-05, "loss": 0.054, "step": 1890 }, { "epoch": 0.5528076811172534, "grad_norm": 0.3086705803871155, "learning_rate": 1.899e-05, "loss": 0.0555, "step": 1900 }, { "epoch": 0.5557171952283969, "grad_norm": 0.26778391003608704, "learning_rate": 1.909e-05, "loss": 0.0544, "step": 1910 }, { "epoch": 0.5586267093395403, "grad_norm": 0.36830881237983704, "learning_rate": 1.919e-05, "loss": 0.052, "step": 1920 }, { "epoch": 0.5615362234506838, "grad_norm": 0.20201042294502258, "learning_rate": 1.929e-05, "loss": 0.0539, "step": 1930 }, { "epoch": 0.5644457375618271, "grad_norm": 0.2952854335308075, "learning_rate": 1.939e-05, "loss": 0.0531, "step": 1940 }, { "epoch": 0.5673552516729706, "grad_norm": 0.20744408667087555, "learning_rate": 1.9489999999999998e-05, "loss": 0.0533, "step": 1950 }, { "epoch": 0.570264765784114, "grad_norm": 0.2691434919834137, "learning_rate": 1.959e-05, "loss": 0.0576, "step": 1960 }, { "epoch": 0.5731742798952575, "grad_norm": 0.22738133370876312, "learning_rate": 1.969e-05, "loss": 0.0522, "step": 1970 }, { "epoch": 0.576083794006401, "grad_norm": 0.23058085143566132, "learning_rate": 1.979e-05, "loss": 0.0546, "step": 1980 }, { "epoch": 0.5789933081175443, "grad_norm": 0.2745729684829712, "learning_rate": 1.989e-05, "loss": 0.0586, "step": 1990 }, { "epoch": 0.5819028222286878, "grad_norm": 0.28763699531555176, "learning_rate": 1.999e-05, "loss": 0.0552, "step": 2000 }, { "epoch": 0.5848123363398312, "grad_norm": 0.24988853931427002, "learning_rate": 2.009e-05, "loss": 0.0544, "step": 2010 }, { "epoch": 0.5877218504509747, "grad_norm": 0.23085060715675354, "learning_rate": 2.019e-05, "loss": 0.0517, "step": 2020 }, { "epoch": 0.5906313645621182, "grad_norm": 0.2869768738746643, "learning_rate": 2.029e-05, "loss": 0.0514, "step": 2030 }, { "epoch": 0.5935408786732616, "grad_norm": 0.342989444732666, "learning_rate": 2.039e-05, "loss": 0.0542, "step": 2040 }, { "epoch": 0.596450392784405, "grad_norm": 0.2237582504749298, "learning_rate": 2.0490000000000002e-05, "loss": 0.0535, "step": 2050 }, { "epoch": 0.5993599068955484, "grad_norm": 0.2374129444360733, "learning_rate": 2.059e-05, "loss": 0.0523, "step": 2060 }, { "epoch": 0.6022694210066919, "grad_norm": 0.20966170728206635, "learning_rate": 2.069e-05, "loss": 0.0542, "step": 2070 }, { "epoch": 0.6051789351178353, "grad_norm": 0.2770410180091858, "learning_rate": 2.079e-05, "loss": 0.0541, "step": 2080 }, { "epoch": 0.6080884492289788, "grad_norm": 0.24118001759052277, "learning_rate": 2.089e-05, "loss": 0.0528, "step": 2090 }, { "epoch": 0.6109979633401222, "grad_norm": 0.2423812299966812, "learning_rate": 2.099e-05, "loss": 0.0504, "step": 2100 }, { "epoch": 0.6139074774512656, "grad_norm": 0.3434012234210968, "learning_rate": 2.109e-05, "loss": 0.0526, "step": 2110 }, { "epoch": 0.6168169915624091, "grad_norm": 0.23631402850151062, "learning_rate": 2.1190000000000002e-05, "loss": 0.0516, "step": 2120 }, { "epoch": 0.6197265056735525, "grad_norm": 0.2768946588039398, "learning_rate": 2.129e-05, "loss": 0.0502, "step": 2130 }, { "epoch": 0.622636019784696, "grad_norm": 0.22398480772972107, "learning_rate": 2.139e-05, "loss": 0.0513, "step": 2140 }, { "epoch": 0.6255455338958394, "grad_norm": 0.24525921046733856, "learning_rate": 2.1490000000000003e-05, "loss": 0.0552, "step": 2150 }, { "epoch": 0.6284550480069828, "grad_norm": 0.22910070419311523, "learning_rate": 2.159e-05, "loss": 0.0518, "step": 2160 }, { "epoch": 0.6313645621181263, "grad_norm": 0.3465941846370697, "learning_rate": 2.169e-05, "loss": 0.0553, "step": 2170 }, { "epoch": 0.6342740762292697, "grad_norm": 0.24853204190731049, "learning_rate": 2.1790000000000003e-05, "loss": 0.0524, "step": 2180 }, { "epoch": 0.6371835903404132, "grad_norm": 0.20722408592700958, "learning_rate": 2.1890000000000002e-05, "loss": 0.0526, "step": 2190 }, { "epoch": 0.6400931044515565, "grad_norm": 0.2644493579864502, "learning_rate": 2.199e-05, "loss": 0.0546, "step": 2200 }, { "epoch": 0.6430026185627, "grad_norm": 0.34095296263694763, "learning_rate": 2.209e-05, "loss": 0.0535, "step": 2210 }, { "epoch": 0.6459121326738435, "grad_norm": 0.33414334058761597, "learning_rate": 2.2190000000000003e-05, "loss": 0.0526, "step": 2220 }, { "epoch": 0.6488216467849869, "grad_norm": 0.25128844380378723, "learning_rate": 2.2290000000000002e-05, "loss": 0.0506, "step": 2230 }, { "epoch": 0.6517311608961304, "grad_norm": 0.2133583426475525, "learning_rate": 2.239e-05, "loss": 0.0513, "step": 2240 }, { "epoch": 0.6546406750072737, "grad_norm": 0.265137642621994, "learning_rate": 2.2490000000000003e-05, "loss": 0.0534, "step": 2250 }, { "epoch": 0.6575501891184172, "grad_norm": 0.18463392555713654, "learning_rate": 2.2590000000000002e-05, "loss": 0.0491, "step": 2260 }, { "epoch": 0.6604597032295607, "grad_norm": 0.201565220952034, "learning_rate": 2.269e-05, "loss": 0.0512, "step": 2270 }, { "epoch": 0.6633692173407041, "grad_norm": 0.34099382162094116, "learning_rate": 2.279e-05, "loss": 0.0526, "step": 2280 }, { "epoch": 0.6662787314518476, "grad_norm": 0.2508865296840668, "learning_rate": 2.289e-05, "loss": 0.0505, "step": 2290 }, { "epoch": 0.669188245562991, "grad_norm": 0.287776917219162, "learning_rate": 2.299e-05, "loss": 0.0523, "step": 2300 }, { "epoch": 0.6720977596741344, "grad_norm": 0.1943676620721817, "learning_rate": 2.309e-05, "loss": 0.0537, "step": 2310 }, { "epoch": 0.6750072737852778, "grad_norm": 0.23578791320323944, "learning_rate": 2.319e-05, "loss": 0.0502, "step": 2320 }, { "epoch": 0.6779167878964213, "grad_norm": 0.2711820900440216, "learning_rate": 2.329e-05, "loss": 0.0505, "step": 2330 }, { "epoch": 0.6808263020075648, "grad_norm": 0.2522619366645813, "learning_rate": 2.3389999999999998e-05, "loss": 0.0515, "step": 2340 }, { "epoch": 0.6837358161187082, "grad_norm": 0.20830851793289185, "learning_rate": 2.349e-05, "loss": 0.0555, "step": 2350 }, { "epoch": 0.6866453302298516, "grad_norm": 0.2329421490430832, "learning_rate": 2.359e-05, "loss": 0.0507, "step": 2360 }, { "epoch": 0.689554844340995, "grad_norm": 0.26514333486557007, "learning_rate": 2.369e-05, "loss": 0.0516, "step": 2370 }, { "epoch": 0.6924643584521385, "grad_norm": 0.2102152705192566, "learning_rate": 2.379e-05, "loss": 0.0531, "step": 2380 }, { "epoch": 0.695373872563282, "grad_norm": 0.201460063457489, "learning_rate": 2.389e-05, "loss": 0.0534, "step": 2390 }, { "epoch": 0.6982833866744254, "grad_norm": 0.2215193212032318, "learning_rate": 2.399e-05, "loss": 0.0499, "step": 2400 }, { "epoch": 0.7011929007855688, "grad_norm": 0.21887636184692383, "learning_rate": 2.409e-05, "loss": 0.0551, "step": 2410 }, { "epoch": 0.7041024148967122, "grad_norm": 0.2104310393333435, "learning_rate": 2.419e-05, "loss": 0.0529, "step": 2420 }, { "epoch": 0.7070119290078557, "grad_norm": 0.19673478603363037, "learning_rate": 2.429e-05, "loss": 0.0499, "step": 2430 }, { "epoch": 0.7099214431189991, "grad_norm": 0.24917297065258026, "learning_rate": 2.439e-05, "loss": 0.0518, "step": 2440 }, { "epoch": 0.7128309572301426, "grad_norm": 0.177176371216774, "learning_rate": 2.449e-05, "loss": 0.0477, "step": 2450 }, { "epoch": 0.715740471341286, "grad_norm": 0.2412344366312027, "learning_rate": 2.459e-05, "loss": 0.0534, "step": 2460 }, { "epoch": 0.7186499854524294, "grad_norm": 0.3411995470523834, "learning_rate": 2.469e-05, "loss": 0.05, "step": 2470 }, { "epoch": 0.7215594995635729, "grad_norm": 0.2410188764333725, "learning_rate": 2.479e-05, "loss": 0.0524, "step": 2480 }, { "epoch": 0.7244690136747163, "grad_norm": 0.20879635214805603, "learning_rate": 2.489e-05, "loss": 0.0519, "step": 2490 }, { "epoch": 0.7273785277858598, "grad_norm": 0.24046729505062103, "learning_rate": 2.499e-05, "loss": 0.0497, "step": 2500 }, { "epoch": 0.7302880418970032, "grad_norm": 0.1911788135766983, "learning_rate": 2.5090000000000002e-05, "loss": 0.0516, "step": 2510 }, { "epoch": 0.7331975560081466, "grad_norm": 0.23389431834220886, "learning_rate": 2.519e-05, "loss": 0.0506, "step": 2520 }, { "epoch": 0.7361070701192901, "grad_norm": 0.1990044265985489, "learning_rate": 2.529e-05, "loss": 0.0487, "step": 2530 }, { "epoch": 0.7390165842304335, "grad_norm": 0.19921153783798218, "learning_rate": 2.5390000000000003e-05, "loss": 0.048, "step": 2540 }, { "epoch": 0.741926098341577, "grad_norm": 0.2802295684814453, "learning_rate": 2.549e-05, "loss": 0.051, "step": 2550 }, { "epoch": 0.7448356124527203, "grad_norm": 0.2800270915031433, "learning_rate": 2.559e-05, "loss": 0.0508, "step": 2560 }, { "epoch": 0.7477451265638638, "grad_norm": 0.28072237968444824, "learning_rate": 2.569e-05, "loss": 0.0531, "step": 2570 }, { "epoch": 0.7506546406750073, "grad_norm": 0.3373444378376007, "learning_rate": 2.5790000000000002e-05, "loss": 0.0513, "step": 2580 }, { "epoch": 0.7535641547861507, "grad_norm": 0.19341887533664703, "learning_rate": 2.589e-05, "loss": 0.0478, "step": 2590 }, { "epoch": 0.7564736688972942, "grad_norm": 0.22359783947467804, "learning_rate": 2.599e-05, "loss": 0.0501, "step": 2600 }, { "epoch": 0.7593831830084375, "grad_norm": 0.17980048060417175, "learning_rate": 2.6090000000000003e-05, "loss": 0.0476, "step": 2610 }, { "epoch": 0.762292697119581, "grad_norm": 0.2999880611896515, "learning_rate": 2.619e-05, "loss": 0.0518, "step": 2620 }, { "epoch": 0.7652022112307245, "grad_norm": 0.2509192228317261, "learning_rate": 2.629e-05, "loss": 0.0517, "step": 2630 }, { "epoch": 0.7681117253418679, "grad_norm": 0.18714138865470886, "learning_rate": 2.6390000000000003e-05, "loss": 0.0482, "step": 2640 }, { "epoch": 0.7710212394530114, "grad_norm": 0.21703273057937622, "learning_rate": 2.6490000000000002e-05, "loss": 0.0488, "step": 2650 }, { "epoch": 0.7739307535641547, "grad_norm": 0.22573548555374146, "learning_rate": 2.659e-05, "loss": 0.048, "step": 2660 }, { "epoch": 0.7768402676752982, "grad_norm": 0.18755318224430084, "learning_rate": 2.6690000000000004e-05, "loss": 0.0474, "step": 2670 }, { "epoch": 0.7797497817864416, "grad_norm": 0.2154727578163147, "learning_rate": 2.6790000000000003e-05, "loss": 0.0478, "step": 2680 }, { "epoch": 0.7826592958975851, "grad_norm": 0.1945435255765915, "learning_rate": 2.6890000000000002e-05, "loss": 0.0477, "step": 2690 }, { "epoch": 0.7855688100087286, "grad_norm": 0.18407292664051056, "learning_rate": 2.699e-05, "loss": 0.0491, "step": 2700 }, { "epoch": 0.788478324119872, "grad_norm": 0.24937784671783447, "learning_rate": 2.709e-05, "loss": 0.0489, "step": 2710 }, { "epoch": 0.7913878382310154, "grad_norm": 0.168559268116951, "learning_rate": 2.719e-05, "loss": 0.0449, "step": 2720 }, { "epoch": 0.7942973523421588, "grad_norm": 0.19336377084255219, "learning_rate": 2.7289999999999998e-05, "loss": 0.0462, "step": 2730 }, { "epoch": 0.7972068664533023, "grad_norm": 0.2569170296192169, "learning_rate": 2.739e-05, "loss": 0.049, "step": 2740 }, { "epoch": 0.8001163805644458, "grad_norm": 0.19858111441135406, "learning_rate": 2.749e-05, "loss": 0.05, "step": 2750 }, { "epoch": 0.8030258946755892, "grad_norm": 0.178179070353508, "learning_rate": 2.759e-05, "loss": 0.0487, "step": 2760 }, { "epoch": 0.8059354087867326, "grad_norm": 0.22307711839675903, "learning_rate": 2.769e-05, "loss": 0.049, "step": 2770 }, { "epoch": 0.808844922897876, "grad_norm": 0.23509469628334045, "learning_rate": 2.779e-05, "loss": 0.0515, "step": 2780 }, { "epoch": 0.8117544370090195, "grad_norm": 0.2781869173049927, "learning_rate": 2.789e-05, "loss": 0.0502, "step": 2790 }, { "epoch": 0.814663951120163, "grad_norm": 0.19866441190242767, "learning_rate": 2.799e-05, "loss": 0.0506, "step": 2800 }, { "epoch": 0.8175734652313064, "grad_norm": 0.26628172397613525, "learning_rate": 2.809e-05, "loss": 0.0485, "step": 2810 }, { "epoch": 0.8204829793424498, "grad_norm": 0.22351808845996857, "learning_rate": 2.819e-05, "loss": 0.0484, "step": 2820 }, { "epoch": 0.8233924934535932, "grad_norm": 0.18317529559135437, "learning_rate": 2.829e-05, "loss": 0.0492, "step": 2830 }, { "epoch": 0.8263020075647367, "grad_norm": 0.1944294571876526, "learning_rate": 2.839e-05, "loss": 0.0491, "step": 2840 }, { "epoch": 0.8292115216758801, "grad_norm": 0.25529685616493225, "learning_rate": 2.849e-05, "loss": 0.0479, "step": 2850 }, { "epoch": 0.8321210357870236, "grad_norm": 0.2421206682920456, "learning_rate": 2.859e-05, "loss": 0.0495, "step": 2860 }, { "epoch": 0.835030549898167, "grad_norm": 0.19663067162036896, "learning_rate": 2.869e-05, "loss": 0.0512, "step": 2870 }, { "epoch": 0.8379400640093104, "grad_norm": 0.24166126549243927, "learning_rate": 2.879e-05, "loss": 0.0464, "step": 2880 }, { "epoch": 0.8408495781204539, "grad_norm": 0.2629076838493347, "learning_rate": 2.889e-05, "loss": 0.0474, "step": 2890 }, { "epoch": 0.8437590922315973, "grad_norm": 0.18241138756275177, "learning_rate": 2.8990000000000002e-05, "loss": 0.0473, "step": 2900 }, { "epoch": 0.8466686063427408, "grad_norm": 0.20032690465450287, "learning_rate": 2.909e-05, "loss": 0.0481, "step": 2910 }, { "epoch": 0.8495781204538843, "grad_norm": 0.27909034490585327, "learning_rate": 2.919e-05, "loss": 0.0468, "step": 2920 }, { "epoch": 0.8524876345650276, "grad_norm": 0.20450608432292938, "learning_rate": 2.9290000000000002e-05, "loss": 0.0494, "step": 2930 }, { "epoch": 0.8553971486761711, "grad_norm": 0.3491055965423584, "learning_rate": 2.939e-05, "loss": 0.0482, "step": 2940 }, { "epoch": 0.8583066627873145, "grad_norm": 0.2021041065454483, "learning_rate": 2.949e-05, "loss": 0.046, "step": 2950 }, { "epoch": 0.861216176898458, "grad_norm": 0.21508947014808655, "learning_rate": 2.959e-05, "loss": 0.0483, "step": 2960 }, { "epoch": 0.8641256910096013, "grad_norm": 0.21039338409900665, "learning_rate": 2.9690000000000002e-05, "loss": 0.0497, "step": 2970 }, { "epoch": 0.8670352051207448, "grad_norm": 0.2893143892288208, "learning_rate": 2.979e-05, "loss": 0.0492, "step": 2980 }, { "epoch": 0.8699447192318883, "grad_norm": 0.2327040284872055, "learning_rate": 2.989e-05, "loss": 0.0462, "step": 2990 }, { "epoch": 0.8728542333430317, "grad_norm": 0.28080353140830994, "learning_rate": 2.9990000000000003e-05, "loss": 0.0486, "step": 3000 }, { "epoch": 0.8757637474541752, "grad_norm": 0.24165192246437073, "learning_rate": 2.9999998154575393e-05, "loss": 0.0498, "step": 3010 }, { "epoch": 0.8786732615653186, "grad_norm": 0.21650919318199158, "learning_rate": 2.999999177533042e-05, "loss": 0.0472, "step": 3020 }, { "epoch": 0.881582775676462, "grad_norm": 0.28231731057167053, "learning_rate": 2.9999980839483992e-05, "loss": 0.0471, "step": 3030 }, { "epoch": 0.8844922897876055, "grad_norm": 0.23491334915161133, "learning_rate": 2.999996534703944e-05, "loss": 0.0511, "step": 3040 }, { "epoch": 0.8874018038987489, "grad_norm": 0.273527592420578, "learning_rate": 2.9999945298001468e-05, "loss": 0.0468, "step": 3050 }, { "epoch": 0.8903113180098924, "grad_norm": 0.2047262042760849, "learning_rate": 2.9999920692376165e-05, "loss": 0.0502, "step": 3060 }, { "epoch": 0.8932208321210358, "grad_norm": 0.32708296179771423, "learning_rate": 2.9999891530171002e-05, "loss": 0.047, "step": 3070 }, { "epoch": 0.8961303462321792, "grad_norm": 0.23661379516124725, "learning_rate": 2.9999857811394845e-05, "loss": 0.0459, "step": 3080 }, { "epoch": 0.8990398603433226, "grad_norm": 0.18705163896083832, "learning_rate": 2.9999819536057935e-05, "loss": 0.0446, "step": 3090 }, { "epoch": 0.9019493744544661, "grad_norm": 0.1884559988975525, "learning_rate": 2.9999776704171894e-05, "loss": 0.0455, "step": 3100 }, { "epoch": 0.9048588885656096, "grad_norm": 0.22597476840019226, "learning_rate": 2.999972931574973e-05, "loss": 0.0478, "step": 3110 }, { "epoch": 0.907768402676753, "grad_norm": 0.179483562707901, "learning_rate": 2.9999677370805858e-05, "loss": 0.0455, "step": 3120 }, { "epoch": 0.9106779167878964, "grad_norm": 0.1774996966123581, "learning_rate": 2.9999620869356034e-05, "loss": 0.0452, "step": 3130 }, { "epoch": 0.9135874308990398, "grad_norm": 0.19030524790287018, "learning_rate": 2.9999559811417436e-05, "loss": 0.0443, "step": 3140 }, { "epoch": 0.9164969450101833, "grad_norm": 0.2090316265821457, "learning_rate": 2.999949419700861e-05, "loss": 0.0472, "step": 3150 }, { "epoch": 0.9194064591213268, "grad_norm": 0.2114928513765335, "learning_rate": 2.9999424026149484e-05, "loss": 0.0458, "step": 3160 }, { "epoch": 0.9223159732324702, "grad_norm": 0.21484243869781494, "learning_rate": 2.9999349298861376e-05, "loss": 0.0458, "step": 3170 }, { "epoch": 0.9252254873436137, "grad_norm": 0.19960108399391174, "learning_rate": 2.9999270015166983e-05, "loss": 0.0486, "step": 3180 }, { "epoch": 0.928135001454757, "grad_norm": 0.2324952632188797, "learning_rate": 2.999918617509039e-05, "loss": 0.0449, "step": 3190 }, { "epoch": 0.9310445155659005, "grad_norm": 0.20183566212654114, "learning_rate": 2.9999097778657074e-05, "loss": 0.047, "step": 3200 }, { "epoch": 0.9339540296770439, "grad_norm": 0.2038273960351944, "learning_rate": 2.9999004825893878e-05, "loss": 0.0483, "step": 3210 }, { "epoch": 0.9368635437881874, "grad_norm": 0.15796031057834625, "learning_rate": 2.9998907316829043e-05, "loss": 0.0459, "step": 3220 }, { "epoch": 0.9397730578993309, "grad_norm": 0.2669369876384735, "learning_rate": 2.999880525149219e-05, "loss": 0.0447, "step": 3230 }, { "epoch": 0.9426825720104742, "grad_norm": 0.18033458292484283, "learning_rate": 2.999869862991432e-05, "loss": 0.0441, "step": 3240 }, { "epoch": 0.9455920861216177, "grad_norm": 0.22309815883636475, "learning_rate": 2.9998587452127822e-05, "loss": 0.0472, "step": 3250 }, { "epoch": 0.9485016002327611, "grad_norm": 0.20341287553310394, "learning_rate": 2.999847171816647e-05, "loss": 0.0455, "step": 3260 }, { "epoch": 0.9514111143439046, "grad_norm": 0.20571516454219818, "learning_rate": 2.999835142806543e-05, "loss": 0.0442, "step": 3270 }, { "epoch": 0.9543206284550481, "grad_norm": 0.19627153873443604, "learning_rate": 2.9998226581861227e-05, "loss": 0.0464, "step": 3280 }, { "epoch": 0.9572301425661914, "grad_norm": 0.2241826057434082, "learning_rate": 2.9998097179591793e-05, "loss": 0.0442, "step": 3290 }, { "epoch": 0.9601396566773349, "grad_norm": 0.22111670672893524, "learning_rate": 2.9997963221296443e-05, "loss": 0.0473, "step": 3300 }, { "epoch": 0.9630491707884783, "grad_norm": 0.26421985030174255, "learning_rate": 2.9997824707015858e-05, "loss": 0.0461, "step": 3310 }, { "epoch": 0.9659586848996218, "grad_norm": 0.25288140773773193, "learning_rate": 2.9997681636792126e-05, "loss": 0.0479, "step": 3320 }, { "epoch": 0.9688681990107652, "grad_norm": 0.21757011115550995, "learning_rate": 2.9997534010668707e-05, "loss": 0.0466, "step": 3330 }, { "epoch": 0.9717777131219086, "grad_norm": 0.23207587003707886, "learning_rate": 2.9997381828690435e-05, "loss": 0.0448, "step": 3340 }, { "epoch": 0.9746872272330521, "grad_norm": 0.16811656951904297, "learning_rate": 2.9997225090903552e-05, "loss": 0.0442, "step": 3350 }, { "epoch": 0.9775967413441955, "grad_norm": 0.22009889781475067, "learning_rate": 2.9997063797355664e-05, "loss": 0.0506, "step": 3360 }, { "epoch": 0.980506255455339, "grad_norm": 0.17711181938648224, "learning_rate": 2.9996897948095768e-05, "loss": 0.0456, "step": 3370 }, { "epoch": 0.9834157695664824, "grad_norm": 0.16644759476184845, "learning_rate": 2.9996727543174244e-05, "loss": 0.0482, "step": 3380 }, { "epoch": 0.9863252836776258, "grad_norm": 0.21965470910072327, "learning_rate": 2.9996552582642857e-05, "loss": 0.0462, "step": 3390 }, { "epoch": 0.9892347977887693, "grad_norm": 0.20110289752483368, "learning_rate": 2.9996373066554763e-05, "loss": 0.0452, "step": 3400 }, { "epoch": 0.9921443118999127, "grad_norm": 0.18359749019145966, "learning_rate": 2.999618899496448e-05, "loss": 0.0513, "step": 3410 }, { "epoch": 0.9950538260110562, "grad_norm": 0.24697566032409668, "learning_rate": 2.9996000367927936e-05, "loss": 0.0459, "step": 3420 }, { "epoch": 0.9979633401221996, "grad_norm": 0.2103855460882187, "learning_rate": 2.9995807185502425e-05, "loss": 0.0472, "step": 3430 }, { "epoch": 1.000872854233343, "grad_norm": 0.16177618503570557, "learning_rate": 2.9995609447746636e-05, "loss": 0.0445, "step": 3440 }, { "epoch": 1.0037823683444864, "grad_norm": 0.17136389017105103, "learning_rate": 2.999540715472063e-05, "loss": 0.0453, "step": 3450 }, { "epoch": 1.00669188245563, "grad_norm": 0.16502591967582703, "learning_rate": 2.999520030648586e-05, "loss": 0.0464, "step": 3460 }, { "epoch": 1.0096013965667734, "grad_norm": 0.18952810764312744, "learning_rate": 2.9994988903105163e-05, "loss": 0.0437, "step": 3470 }, { "epoch": 1.0125109106779169, "grad_norm": 0.15830445289611816, "learning_rate": 2.999477294464276e-05, "loss": 0.0446, "step": 3480 }, { "epoch": 1.0154204247890601, "grad_norm": 0.18032631278038025, "learning_rate": 2.9994552431164247e-05, "loss": 0.0452, "step": 3490 }, { "epoch": 1.0183299389002036, "grad_norm": 0.19681726396083832, "learning_rate": 2.9994327362736617e-05, "loss": 0.0442, "step": 3500 }, { "epoch": 1.021239453011347, "grad_norm": 0.19932463765144348, "learning_rate": 2.9994097739428233e-05, "loss": 0.0444, "step": 3510 }, { "epoch": 1.0241489671224906, "grad_norm": 0.22861382365226746, "learning_rate": 2.9993863561308856e-05, "loss": 0.0433, "step": 3520 }, { "epoch": 1.027058481233634, "grad_norm": 0.20167602598667145, "learning_rate": 2.9993624828449616e-05, "loss": 0.0449, "step": 3530 }, { "epoch": 1.0299679953447773, "grad_norm": 0.22683775424957275, "learning_rate": 2.9993381540923037e-05, "loss": 0.0447, "step": 3540 }, { "epoch": 1.0328775094559208, "grad_norm": 0.19140365719795227, "learning_rate": 2.9993133698803025e-05, "loss": 0.0445, "step": 3550 }, { "epoch": 1.0357870235670643, "grad_norm": 0.187163844704628, "learning_rate": 2.9992881302164862e-05, "loss": 0.0431, "step": 3560 }, { "epoch": 1.0386965376782078, "grad_norm": 0.1882355511188507, "learning_rate": 2.9992624351085226e-05, "loss": 0.0439, "step": 3570 }, { "epoch": 1.0416060517893513, "grad_norm": 0.21659860014915466, "learning_rate": 2.9992362845642167e-05, "loss": 0.0443, "step": 3580 }, { "epoch": 1.0445155659004945, "grad_norm": 0.17048729956150055, "learning_rate": 2.9992096785915132e-05, "loss": 0.0447, "step": 3590 }, { "epoch": 1.047425080011638, "grad_norm": 0.21387426555156708, "learning_rate": 2.999182617198493e-05, "loss": 0.0466, "step": 3600 }, { "epoch": 1.0503345941227815, "grad_norm": 0.1604042649269104, "learning_rate": 2.999155100393378e-05, "loss": 0.0412, "step": 3610 }, { "epoch": 1.053244108233925, "grad_norm": 0.23740248382091522, "learning_rate": 2.9991271281845256e-05, "loss": 0.0469, "step": 3620 }, { "epoch": 1.0561536223450685, "grad_norm": 0.16189396381378174, "learning_rate": 2.9990987005804347e-05, "loss": 0.0429, "step": 3630 }, { "epoch": 1.0590631364562118, "grad_norm": 0.24757643043994904, "learning_rate": 2.999069817589739e-05, "loss": 0.0444, "step": 3640 }, { "epoch": 1.0619726505673552, "grad_norm": 0.1901935487985611, "learning_rate": 2.9990404792212142e-05, "loss": 0.0447, "step": 3650 }, { "epoch": 1.0648821646784987, "grad_norm": 0.2800537347793579, "learning_rate": 2.9990106854837715e-05, "loss": 0.0445, "step": 3660 }, { "epoch": 1.0677916787896422, "grad_norm": 0.24123845994472504, "learning_rate": 2.9989804363864615e-05, "loss": 0.0442, "step": 3670 }, { "epoch": 1.0707011929007855, "grad_norm": 0.21424239873886108, "learning_rate": 2.998949731938473e-05, "loss": 0.0456, "step": 3680 }, { "epoch": 1.073610707011929, "grad_norm": 0.15927696228027344, "learning_rate": 2.9989185721491338e-05, "loss": 0.0456, "step": 3690 }, { "epoch": 1.0765202211230724, "grad_norm": 0.2342563420534134, "learning_rate": 2.998886957027909e-05, "loss": 0.045, "step": 3700 }, { "epoch": 1.079429735234216, "grad_norm": 0.17592190206050873, "learning_rate": 2.998854886584402e-05, "loss": 0.0406, "step": 3710 }, { "epoch": 1.0823392493453594, "grad_norm": 0.1893317997455597, "learning_rate": 2.9988223608283557e-05, "loss": 0.0427, "step": 3720 }, { "epoch": 1.0852487634565027, "grad_norm": 0.18561504781246185, "learning_rate": 2.9987893797696502e-05, "loss": 0.0441, "step": 3730 }, { "epoch": 1.0881582775676462, "grad_norm": 0.18621112406253815, "learning_rate": 2.9987559434183046e-05, "loss": 0.0424, "step": 3740 }, { "epoch": 1.0910677916787896, "grad_norm": 0.2157931923866272, "learning_rate": 2.998722051784476e-05, "loss": 0.0437, "step": 3750 }, { "epoch": 1.0939773057899331, "grad_norm": 0.18780076503753662, "learning_rate": 2.9986877048784594e-05, "loss": 0.0438, "step": 3760 }, { "epoch": 1.0968868199010766, "grad_norm": 0.1958472728729248, "learning_rate": 2.9986529027106885e-05, "loss": 0.0431, "step": 3770 }, { "epoch": 1.0997963340122199, "grad_norm": 0.18335162103176117, "learning_rate": 2.998617645291735e-05, "loss": 0.0443, "step": 3780 }, { "epoch": 1.1027058481233634, "grad_norm": 0.20056045055389404, "learning_rate": 2.9985819326323097e-05, "loss": 0.041, "step": 3790 }, { "epoch": 1.1056153622345068, "grad_norm": 0.19871258735656738, "learning_rate": 2.9985457647432613e-05, "loss": 0.0411, "step": 3800 }, { "epoch": 1.1085248763456503, "grad_norm": 0.1713438630104065, "learning_rate": 2.998509141635576e-05, "loss": 0.0417, "step": 3810 }, { "epoch": 1.1114343904567938, "grad_norm": 0.15562936663627625, "learning_rate": 2.99847206332038e-05, "loss": 0.0434, "step": 3820 }, { "epoch": 1.114343904567937, "grad_norm": 0.18083824217319489, "learning_rate": 2.9984345298089356e-05, "loss": 0.0446, "step": 3830 }, { "epoch": 1.1172534186790806, "grad_norm": 0.14352670311927795, "learning_rate": 2.998396541112645e-05, "loss": 0.043, "step": 3840 }, { "epoch": 1.120162932790224, "grad_norm": 0.17940504848957062, "learning_rate": 2.998358097243048e-05, "loss": 0.0436, "step": 3850 }, { "epoch": 1.1230724469013675, "grad_norm": 0.23755693435668945, "learning_rate": 2.998319198211823e-05, "loss": 0.0419, "step": 3860 }, { "epoch": 1.1259819610125108, "grad_norm": 0.21266232430934906, "learning_rate": 2.998279844030786e-05, "loss": 0.0407, "step": 3870 }, { "epoch": 1.1288914751236543, "grad_norm": 0.1911960393190384, "learning_rate": 2.9982400347118926e-05, "loss": 0.0435, "step": 3880 }, { "epoch": 1.1318009892347978, "grad_norm": 0.18037854135036469, "learning_rate": 2.9981997702672353e-05, "loss": 0.045, "step": 3890 }, { "epoch": 1.1347105033459413, "grad_norm": 0.16058249771595, "learning_rate": 2.9981590507090454e-05, "loss": 0.0422, "step": 3900 }, { "epoch": 1.1376200174570847, "grad_norm": 0.17884571850299835, "learning_rate": 2.9981178760496927e-05, "loss": 0.0394, "step": 3910 }, { "epoch": 1.140529531568228, "grad_norm": 0.20452144742012024, "learning_rate": 2.9980762463016848e-05, "loss": 0.0432, "step": 3920 }, { "epoch": 1.1434390456793715, "grad_norm": 0.21047839522361755, "learning_rate": 2.9980341614776673e-05, "loss": 0.0418, "step": 3930 }, { "epoch": 1.146348559790515, "grad_norm": 0.2097507268190384, "learning_rate": 2.9979916215904247e-05, "loss": 0.0431, "step": 3940 }, { "epoch": 1.1492580739016585, "grad_norm": 0.1427605003118515, "learning_rate": 2.99794862665288e-05, "loss": 0.0428, "step": 3950 }, { "epoch": 1.152167588012802, "grad_norm": 0.23656804859638214, "learning_rate": 2.9979051766780938e-05, "loss": 0.0422, "step": 3960 }, { "epoch": 1.1550771021239452, "grad_norm": 0.2294016033411026, "learning_rate": 2.9978612716792647e-05, "loss": 0.0427, "step": 3970 }, { "epoch": 1.1579866162350887, "grad_norm": 0.22043481469154358, "learning_rate": 2.99781691166973e-05, "loss": 0.0444, "step": 3980 }, { "epoch": 1.1608961303462322, "grad_norm": 0.189597949385643, "learning_rate": 2.9977720966629646e-05, "loss": 0.0459, "step": 3990 }, { "epoch": 1.1638056444573757, "grad_norm": 0.18280912935733795, "learning_rate": 2.997726826672583e-05, "loss": 0.0411, "step": 4000 }, { "epoch": 1.1667151585685192, "grad_norm": 0.14928048849105835, "learning_rate": 2.9976811017123368e-05, "loss": 0.0404, "step": 4010 }, { "epoch": 1.1696246726796624, "grad_norm": 0.22554485499858856, "learning_rate": 2.9976349217961158e-05, "loss": 0.0427, "step": 4020 }, { "epoch": 1.172534186790806, "grad_norm": 0.23622818291187286, "learning_rate": 2.9975882869379482e-05, "loss": 0.0415, "step": 4030 }, { "epoch": 1.1754437009019494, "grad_norm": 0.1972798854112625, "learning_rate": 2.997541197152001e-05, "loss": 0.0425, "step": 4040 }, { "epoch": 1.1783532150130929, "grad_norm": 0.193428635597229, "learning_rate": 2.9974936524525774e-05, "loss": 0.0433, "step": 4050 }, { "epoch": 1.1812627291242364, "grad_norm": 0.2364550232887268, "learning_rate": 2.9974456528541222e-05, "loss": 0.0439, "step": 4060 }, { "epoch": 1.1841722432353796, "grad_norm": 0.209161177277565, "learning_rate": 2.9973971983712147e-05, "loss": 0.0417, "step": 4070 }, { "epoch": 1.187081757346523, "grad_norm": 0.20617970824241638, "learning_rate": 2.9973482890185753e-05, "loss": 0.0432, "step": 4080 }, { "epoch": 1.1899912714576666, "grad_norm": 0.17226345837116241, "learning_rate": 2.9972989248110606e-05, "loss": 0.0424, "step": 4090 }, { "epoch": 1.19290078556881, "grad_norm": 0.18333429098129272, "learning_rate": 2.9972491057636665e-05, "loss": 0.0432, "step": 4100 }, { "epoch": 1.1958102996799536, "grad_norm": 0.225136861205101, "learning_rate": 2.9971988318915268e-05, "loss": 0.0409, "step": 4110 }, { "epoch": 1.1987198137910968, "grad_norm": 0.24684658646583557, "learning_rate": 2.997148103209913e-05, "loss": 0.0426, "step": 4120 }, { "epoch": 1.2016293279022403, "grad_norm": 0.18139322102069855, "learning_rate": 2.9970969197342356e-05, "loss": 0.043, "step": 4130 }, { "epoch": 1.2045388420133838, "grad_norm": 0.26455745100975037, "learning_rate": 2.9970452814800422e-05, "loss": 0.0419, "step": 4140 }, { "epoch": 1.2074483561245273, "grad_norm": 0.1870831996202469, "learning_rate": 2.99699318846302e-05, "loss": 0.0425, "step": 4150 }, { "epoch": 1.2103578702356708, "grad_norm": 0.1658840924501419, "learning_rate": 2.9969406406989927e-05, "loss": 0.0405, "step": 4160 }, { "epoch": 1.213267384346814, "grad_norm": 0.15646706521511078, "learning_rate": 2.996887638203923e-05, "loss": 0.0424, "step": 4170 }, { "epoch": 1.2161768984579575, "grad_norm": 0.22191469371318817, "learning_rate": 2.9968341809939116e-05, "loss": 0.0436, "step": 4180 }, { "epoch": 1.219086412569101, "grad_norm": 0.17340469360351562, "learning_rate": 2.9967802690851985e-05, "loss": 0.0387, "step": 4190 }, { "epoch": 1.2219959266802445, "grad_norm": 0.21366280317306519, "learning_rate": 2.9967259024941595e-05, "loss": 0.0425, "step": 4200 }, { "epoch": 1.2249054407913877, "grad_norm": 0.16080045700073242, "learning_rate": 2.9966710812373097e-05, "loss": 0.041, "step": 4210 }, { "epoch": 1.2278149549025312, "grad_norm": 0.1619821935892105, "learning_rate": 2.996615805331303e-05, "loss": 0.0417, "step": 4220 }, { "epoch": 1.2307244690136747, "grad_norm": 0.18752031028270721, "learning_rate": 2.99656007479293e-05, "loss": 0.0425, "step": 4230 }, { "epoch": 1.2336339831248182, "grad_norm": 0.16798673570156097, "learning_rate": 2.9965038896391213e-05, "loss": 0.042, "step": 4240 }, { "epoch": 1.2365434972359617, "grad_norm": 0.19619518518447876, "learning_rate": 2.9964472498869434e-05, "loss": 0.0422, "step": 4250 }, { "epoch": 1.239453011347105, "grad_norm": 0.16482631862163544, "learning_rate": 2.996390155553603e-05, "loss": 0.0399, "step": 4260 }, { "epoch": 1.2423625254582484, "grad_norm": 0.15593254566192627, "learning_rate": 2.9963326066564426e-05, "loss": 0.0424, "step": 4270 }, { "epoch": 1.245272039569392, "grad_norm": 0.1608118861913681, "learning_rate": 2.996274603212945e-05, "loss": 0.0405, "step": 4280 }, { "epoch": 1.2481815536805354, "grad_norm": 0.14939048886299133, "learning_rate": 2.9962161452407296e-05, "loss": 0.0427, "step": 4290 }, { "epoch": 1.2510910677916787, "grad_norm": 0.16675253212451935, "learning_rate": 2.996157232757555e-05, "loss": 0.0408, "step": 4300 }, { "epoch": 1.2540005819028222, "grad_norm": 0.16277901828289032, "learning_rate": 2.996097865781316e-05, "loss": 0.0414, "step": 4310 }, { "epoch": 1.2569100960139656, "grad_norm": 0.17267730832099915, "learning_rate": 2.996038044330048e-05, "loss": 0.0418, "step": 4320 }, { "epoch": 1.2598196101251091, "grad_norm": 0.20022255182266235, "learning_rate": 2.9959777684219233e-05, "loss": 0.043, "step": 4330 }, { "epoch": 1.2627291242362526, "grad_norm": 0.16030709445476532, "learning_rate": 2.9959170380752508e-05, "loss": 0.0423, "step": 4340 }, { "epoch": 1.2656386383473959, "grad_norm": 0.17309041321277618, "learning_rate": 2.99585585330848e-05, "loss": 0.0378, "step": 4350 }, { "epoch": 1.2685481524585394, "grad_norm": 0.15266495943069458, "learning_rate": 2.9957942141401967e-05, "loss": 0.0429, "step": 4360 }, { "epoch": 1.2714576665696828, "grad_norm": 0.15711914002895355, "learning_rate": 2.995732120589125e-05, "loss": 0.0405, "step": 4370 }, { "epoch": 1.2743671806808263, "grad_norm": 0.14205870032310486, "learning_rate": 2.9956695726741276e-05, "loss": 0.0392, "step": 4380 }, { "epoch": 1.2772766947919698, "grad_norm": 0.17807990312576294, "learning_rate": 2.995606570414205e-05, "loss": 0.0423, "step": 4390 }, { "epoch": 1.280186208903113, "grad_norm": 0.24167108535766602, "learning_rate": 2.9955431138284955e-05, "loss": 0.0414, "step": 4400 }, { "epoch": 1.2830957230142566, "grad_norm": 0.1606665700674057, "learning_rate": 2.9954792029362754e-05, "loss": 0.0419, "step": 4410 }, { "epoch": 1.2860052371254, "grad_norm": 0.17613933980464935, "learning_rate": 2.9954148377569598e-05, "loss": 0.0423, "step": 4420 }, { "epoch": 1.2889147512365435, "grad_norm": 0.17932747304439545, "learning_rate": 2.9953500183101002e-05, "loss": 0.0414, "step": 4430 }, { "epoch": 1.291824265347687, "grad_norm": 0.23386242985725403, "learning_rate": 2.9952847446153877e-05, "loss": 0.043, "step": 4440 }, { "epoch": 1.2947337794588303, "grad_norm": 0.17253951728343964, "learning_rate": 2.9952190166926498e-05, "loss": 0.0423, "step": 4450 }, { "epoch": 1.2976432935699738, "grad_norm": 0.18018461763858795, "learning_rate": 2.9951528345618546e-05, "loss": 0.0418, "step": 4460 }, { "epoch": 1.3005528076811173, "grad_norm": 0.18746407330036163, "learning_rate": 2.995086198243105e-05, "loss": 0.0436, "step": 4470 }, { "epoch": 1.3034623217922607, "grad_norm": 0.17705684900283813, "learning_rate": 2.9950191077566434e-05, "loss": 0.0431, "step": 4480 }, { "epoch": 1.3063718359034042, "grad_norm": 0.19179247319698334, "learning_rate": 2.9949515631228515e-05, "loss": 0.0402, "step": 4490 }, { "epoch": 1.3092813500145475, "grad_norm": 0.14404326677322388, "learning_rate": 2.994883564362246e-05, "loss": 0.0417, "step": 4500 }, { "epoch": 1.312190864125691, "grad_norm": 0.17527277767658234, "learning_rate": 2.994815111495484e-05, "loss": 0.0432, "step": 4510 }, { "epoch": 1.3151003782368345, "grad_norm": 0.17367956042289734, "learning_rate": 2.9947462045433594e-05, "loss": 0.0423, "step": 4520 }, { "epoch": 1.318009892347978, "grad_norm": 0.17037038505077362, "learning_rate": 2.9946768435268045e-05, "loss": 0.0413, "step": 4530 }, { "epoch": 1.3209194064591214, "grad_norm": 0.19695104658603668, "learning_rate": 2.9946070284668884e-05, "loss": 0.0419, "step": 4540 }, { "epoch": 1.3238289205702647, "grad_norm": 0.1620352417230606, "learning_rate": 2.994536759384821e-05, "loss": 0.0415, "step": 4550 }, { "epoch": 1.3267384346814082, "grad_norm": 0.1603638231754303, "learning_rate": 2.994466036301946e-05, "loss": 0.0399, "step": 4560 }, { "epoch": 1.3296479487925517, "grad_norm": 0.23713287711143494, "learning_rate": 2.9943948592397494e-05, "loss": 0.0398, "step": 4570 }, { "epoch": 1.3325574629036951, "grad_norm": 0.16987746953964233, "learning_rate": 2.994323228219851e-05, "loss": 0.0416, "step": 4580 }, { "epoch": 1.3354669770148386, "grad_norm": 0.16761204600334167, "learning_rate": 2.9942511432640113e-05, "loss": 0.0404, "step": 4590 }, { "epoch": 1.338376491125982, "grad_norm": 0.22855566442012787, "learning_rate": 2.9941786043941278e-05, "loss": 0.0422, "step": 4600 }, { "epoch": 1.3412860052371254, "grad_norm": 0.1689605861902237, "learning_rate": 2.9941056116322357e-05, "loss": 0.041, "step": 4610 }, { "epoch": 1.3441955193482689, "grad_norm": 0.15265201032161713, "learning_rate": 2.994032165000508e-05, "loss": 0.0409, "step": 4620 }, { "epoch": 1.3471050334594123, "grad_norm": 0.15963773429393768, "learning_rate": 2.9939582645212566e-05, "loss": 0.0389, "step": 4630 }, { "epoch": 1.3500145475705558, "grad_norm": 0.20251666009426117, "learning_rate": 2.9938839102169303e-05, "loss": 0.0403, "step": 4640 }, { "epoch": 1.352924061681699, "grad_norm": 0.12910820543766022, "learning_rate": 2.993809102110116e-05, "loss": 0.0396, "step": 4650 }, { "epoch": 1.3558335757928426, "grad_norm": 0.15399503707885742, "learning_rate": 2.9937338402235373e-05, "loss": 0.0395, "step": 4660 }, { "epoch": 1.358743089903986, "grad_norm": 0.1942742019891739, "learning_rate": 2.993658124580058e-05, "loss": 0.0428, "step": 4670 }, { "epoch": 1.3616526040151296, "grad_norm": 0.17447015643119812, "learning_rate": 2.9935819552026785e-05, "loss": 0.04, "step": 4680 }, { "epoch": 1.364562118126273, "grad_norm": 0.18135948479175568, "learning_rate": 2.9935053321145368e-05, "loss": 0.0354, "step": 4690 }, { "epoch": 1.3674716322374163, "grad_norm": 0.16358478367328644, "learning_rate": 2.9934282553389088e-05, "loss": 0.0412, "step": 4700 }, { "epoch": 1.3703811463485598, "grad_norm": 0.21508686244487762, "learning_rate": 2.9933507248992084e-05, "loss": 0.0393, "step": 4710 }, { "epoch": 1.3732906604597033, "grad_norm": 0.14794011414051056, "learning_rate": 2.9932727408189876e-05, "loss": 0.0381, "step": 4720 }, { "epoch": 1.3762001745708465, "grad_norm": 0.22785253822803497, "learning_rate": 2.9931943031219356e-05, "loss": 0.041, "step": 4730 }, { "epoch": 1.3791096886819902, "grad_norm": 0.17595769464969635, "learning_rate": 2.9931154118318803e-05, "loss": 0.0411, "step": 4740 }, { "epoch": 1.3820192027931335, "grad_norm": 0.16630840301513672, "learning_rate": 2.9930360669727858e-05, "loss": 0.039, "step": 4750 }, { "epoch": 1.384928716904277, "grad_norm": 0.1494436413049698, "learning_rate": 2.992956268568755e-05, "loss": 0.0403, "step": 4760 }, { "epoch": 1.3878382310154205, "grad_norm": 0.21429456770420074, "learning_rate": 2.99287601664403e-05, "loss": 0.0377, "step": 4770 }, { "epoch": 1.3907477451265637, "grad_norm": 0.21724587678909302, "learning_rate": 2.992795311222988e-05, "loss": 0.038, "step": 4780 }, { "epoch": 1.3936572592377072, "grad_norm": 0.20693016052246094, "learning_rate": 2.9927141523301453e-05, "loss": 0.0402, "step": 4790 }, { "epoch": 1.3965667733488507, "grad_norm": 0.1503114402294159, "learning_rate": 2.992632539990156e-05, "loss": 0.0411, "step": 4800 }, { "epoch": 1.3994762874599942, "grad_norm": 0.18108297884464264, "learning_rate": 2.9925504742278117e-05, "loss": 0.0378, "step": 4810 }, { "epoch": 1.4023858015711377, "grad_norm": 0.14698949456214905, "learning_rate": 2.9924679550680416e-05, "loss": 0.0409, "step": 4820 }, { "epoch": 1.405295315682281, "grad_norm": 0.1856032907962799, "learning_rate": 2.9923849825359133e-05, "loss": 0.0386, "step": 4830 }, { "epoch": 1.4082048297934244, "grad_norm": 0.18419429659843445, "learning_rate": 2.9923015566566317e-05, "loss": 0.0378, "step": 4840 }, { "epoch": 1.411114343904568, "grad_norm": 0.15882720053195953, "learning_rate": 2.992217677455539e-05, "loss": 0.0398, "step": 4850 }, { "epoch": 1.4140238580157114, "grad_norm": 0.19209088385105133, "learning_rate": 2.9921333449581153e-05, "loss": 0.0401, "step": 4860 }, { "epoch": 1.4169333721268549, "grad_norm": 0.19405464828014374, "learning_rate": 2.9920485591899795e-05, "loss": 0.0388, "step": 4870 }, { "epoch": 1.4198428862379981, "grad_norm": 0.17170898616313934, "learning_rate": 2.991963320176886e-05, "loss": 0.0379, "step": 4880 }, { "epoch": 1.4227524003491416, "grad_norm": 0.21294911205768585, "learning_rate": 2.991877627944729e-05, "loss": 0.0396, "step": 4890 }, { "epoch": 1.4256619144602851, "grad_norm": 0.14839968085289001, "learning_rate": 2.9917914825195393e-05, "loss": 0.0404, "step": 4900 }, { "epoch": 1.4285714285714286, "grad_norm": 0.1844959706068039, "learning_rate": 2.991704883927486e-05, "loss": 0.0377, "step": 4910 }, { "epoch": 1.431480942682572, "grad_norm": 0.20516429841518402, "learning_rate": 2.991617832194875e-05, "loss": 0.039, "step": 4920 }, { "epoch": 1.4343904567937154, "grad_norm": 0.14610500633716583, "learning_rate": 2.9915303273481504e-05, "loss": 0.0411, "step": 4930 }, { "epoch": 1.4372999709048588, "grad_norm": 0.1418379843235016, "learning_rate": 2.9914423694138942e-05, "loss": 0.0409, "step": 4940 }, { "epoch": 1.4402094850160023, "grad_norm": 0.19109126925468445, "learning_rate": 2.9913539584188253e-05, "loss": 0.0402, "step": 4950 }, { "epoch": 1.4431189991271458, "grad_norm": 0.16192501783370972, "learning_rate": 2.9912650943898008e-05, "loss": 0.0395, "step": 4960 }, { "epoch": 1.4460285132382893, "grad_norm": 0.21190033853054047, "learning_rate": 2.9911757773538148e-05, "loss": 0.0403, "step": 4970 }, { "epoch": 1.4489380273494326, "grad_norm": 0.16605347394943237, "learning_rate": 2.9910860073380005e-05, "loss": 0.0374, "step": 4980 }, { "epoch": 1.451847541460576, "grad_norm": 0.18824052810668945, "learning_rate": 2.9909957843696267e-05, "loss": 0.0373, "step": 4990 }, { "epoch": 1.4547570555717195, "grad_norm": 0.20194631814956665, "learning_rate": 2.990905108476101e-05, "loss": 0.0379, "step": 5000 }, { "epoch": 1.457666569682863, "grad_norm": 0.19760699570178986, "learning_rate": 2.9908139796849683e-05, "loss": 0.0387, "step": 5010 }, { "epoch": 1.4605760837940065, "grad_norm": 0.2167540341615677, "learning_rate": 2.9907223980239114e-05, "loss": 0.0374, "step": 5020 }, { "epoch": 1.4634855979051498, "grad_norm": 0.190390944480896, "learning_rate": 2.99063036352075e-05, "loss": 0.0373, "step": 5030 }, { "epoch": 1.4663951120162932, "grad_norm": 0.16583938896656036, "learning_rate": 2.9905378762034424e-05, "loss": 0.0402, "step": 5040 }, { "epoch": 1.4693046261274367, "grad_norm": 0.1623157113790512, "learning_rate": 2.9904449361000833e-05, "loss": 0.0398, "step": 5050 }, { "epoch": 1.4722141402385802, "grad_norm": 0.16936975717544556, "learning_rate": 2.9903515432389056e-05, "loss": 0.0398, "step": 5060 }, { "epoch": 1.4751236543497237, "grad_norm": 0.15530578792095184, "learning_rate": 2.990257697648279e-05, "loss": 0.038, "step": 5070 }, { "epoch": 1.478033168460867, "grad_norm": 0.14610524475574493, "learning_rate": 2.9901633993567125e-05, "loss": 0.0362, "step": 5080 }, { "epoch": 1.4809426825720104, "grad_norm": 0.1737520843744278, "learning_rate": 2.99006864839285e-05, "loss": 0.0396, "step": 5090 }, { "epoch": 1.483852196683154, "grad_norm": 0.14880512654781342, "learning_rate": 2.989973444785476e-05, "loss": 0.0374, "step": 5100 }, { "epoch": 1.4867617107942974, "grad_norm": 0.1690666526556015, "learning_rate": 2.989877788563509e-05, "loss": 0.0358, "step": 5110 }, { "epoch": 1.489671224905441, "grad_norm": 0.13936589658260345, "learning_rate": 2.9897816797560085e-05, "loss": 0.0377, "step": 5120 }, { "epoch": 1.4925807390165842, "grad_norm": 0.16280968487262726, "learning_rate": 2.989685118392169e-05, "loss": 0.0357, "step": 5130 }, { "epoch": 1.4954902531277277, "grad_norm": 0.2037695348262787, "learning_rate": 2.9895881045013232e-05, "loss": 0.0384, "step": 5140 }, { "epoch": 1.4983997672388711, "grad_norm": 0.15884585678577423, "learning_rate": 2.9894906381129414e-05, "loss": 0.0381, "step": 5150 }, { "epoch": 1.5013092813500144, "grad_norm": 0.16072319447994232, "learning_rate": 2.9893927192566316e-05, "loss": 0.0376, "step": 5160 }, { "epoch": 1.504218795461158, "grad_norm": 0.15692652761936188, "learning_rate": 2.989294347962139e-05, "loss": 0.0381, "step": 5170 }, { "epoch": 1.5071283095723014, "grad_norm": 0.14910051226615906, "learning_rate": 2.989195524259346e-05, "loss": 0.0398, "step": 5180 }, { "epoch": 1.5100378236834449, "grad_norm": 0.12884077429771423, "learning_rate": 2.9890962481782723e-05, "loss": 0.0365, "step": 5190 }, { "epoch": 1.5129473377945883, "grad_norm": 0.12772971391677856, "learning_rate": 2.9889965197490757e-05, "loss": 0.0405, "step": 5200 }, { "epoch": 1.5158568519057316, "grad_norm": 0.1521998941898346, "learning_rate": 2.988896339002052e-05, "loss": 0.0385, "step": 5210 }, { "epoch": 1.5187663660168753, "grad_norm": 0.11846956610679626, "learning_rate": 2.9887957059676315e-05, "loss": 0.0367, "step": 5220 }, { "epoch": 1.5216758801280186, "grad_norm": 0.20640972256660461, "learning_rate": 2.9886946206763855e-05, "loss": 0.0366, "step": 5230 }, { "epoch": 1.524585394239162, "grad_norm": 0.17890691757202148, "learning_rate": 2.9885930831590202e-05, "loss": 0.0366, "step": 5240 }, { "epoch": 1.5274949083503055, "grad_norm": 0.17437580227851868, "learning_rate": 2.98849109344638e-05, "loss": 0.0405, "step": 5250 }, { "epoch": 1.5304044224614488, "grad_norm": 0.1998003125190735, "learning_rate": 2.9883886515694474e-05, "loss": 0.0368, "step": 5260 }, { "epoch": 1.5333139365725925, "grad_norm": 0.17039676010608673, "learning_rate": 2.988285757559341e-05, "loss": 0.0407, "step": 5270 }, { "epoch": 1.5362234506837358, "grad_norm": 0.23711930215358734, "learning_rate": 2.988182411447317e-05, "loss": 0.0395, "step": 5280 }, { "epoch": 1.5391329647948793, "grad_norm": 0.17751912772655487, "learning_rate": 2.98807861326477e-05, "loss": 0.0375, "step": 5290 }, { "epoch": 1.5420424789060228, "grad_norm": 0.16246521472930908, "learning_rate": 2.9879743630432307e-05, "loss": 0.0368, "step": 5300 }, { "epoch": 1.544951993017166, "grad_norm": 0.17187853157520294, "learning_rate": 2.9878696608143677e-05, "loss": 0.0376, "step": 5310 }, { "epoch": 1.5478615071283097, "grad_norm": 0.1903022676706314, "learning_rate": 2.9877645066099862e-05, "loss": 0.0403, "step": 5320 }, { "epoch": 1.550771021239453, "grad_norm": 0.15912504494190216, "learning_rate": 2.9876589004620298e-05, "loss": 0.0372, "step": 5330 }, { "epoch": 1.5536805353505965, "grad_norm": 0.19590941071510315, "learning_rate": 2.987552842402579e-05, "loss": 0.0391, "step": 5340 }, { "epoch": 1.55659004946174, "grad_norm": 0.248128741979599, "learning_rate": 2.9874463324638515e-05, "loss": 0.04, "step": 5350 }, { "epoch": 1.5594995635728832, "grad_norm": 0.20805510878562927, "learning_rate": 2.9873393706782016e-05, "loss": 0.0389, "step": 5360 }, { "epoch": 1.562409077684027, "grad_norm": 0.17698654532432556, "learning_rate": 2.987231957078122e-05, "loss": 0.0395, "step": 5370 }, { "epoch": 1.5653185917951702, "grad_norm": 0.20228178799152374, "learning_rate": 2.9871240916962414e-05, "loss": 0.0385, "step": 5380 }, { "epoch": 1.5682281059063137, "grad_norm": 0.1758001148700714, "learning_rate": 2.9870157745653274e-05, "loss": 0.0352, "step": 5390 }, { "epoch": 1.5711376200174572, "grad_norm": 0.1761690080165863, "learning_rate": 2.9869070057182838e-05, "loss": 0.0388, "step": 5400 }, { "epoch": 1.5740471341286004, "grad_norm": 0.1587902158498764, "learning_rate": 2.986797785188151e-05, "loss": 0.0369, "step": 5410 }, { "epoch": 1.576956648239744, "grad_norm": 0.16415439546108246, "learning_rate": 2.9866881130081074e-05, "loss": 0.0371, "step": 5420 }, { "epoch": 1.5798661623508874, "grad_norm": 0.21838489174842834, "learning_rate": 2.986577989211469e-05, "loss": 0.0376, "step": 5430 }, { "epoch": 1.5827756764620309, "grad_norm": 0.14505290985107422, "learning_rate": 2.9864674138316887e-05, "loss": 0.0375, "step": 5440 }, { "epoch": 1.5856851905731744, "grad_norm": 0.21441583335399628, "learning_rate": 2.9863563869023554e-05, "loss": 0.0363, "step": 5450 }, { "epoch": 1.5885947046843176, "grad_norm": 0.1701727956533432, "learning_rate": 2.9862449084571972e-05, "loss": 0.0373, "step": 5460 }, { "epoch": 1.5915042187954611, "grad_norm": 0.20329073071479797, "learning_rate": 2.9861329785300774e-05, "loss": 0.0393, "step": 5470 }, { "epoch": 1.5944137329066046, "grad_norm": 0.13727979362010956, "learning_rate": 2.9860205971549978e-05, "loss": 0.0368, "step": 5480 }, { "epoch": 1.597323247017748, "grad_norm": 0.15854553878307343, "learning_rate": 2.9859077643660974e-05, "loss": 0.0394, "step": 5490 }, { "epoch": 1.6002327611288916, "grad_norm": 0.18291640281677246, "learning_rate": 2.985794480197651e-05, "loss": 0.0384, "step": 5500 }, { "epoch": 1.6031422752400348, "grad_norm": 0.17231933772563934, "learning_rate": 2.9856807446840716e-05, "loss": 0.0382, "step": 5510 }, { "epoch": 1.6060517893511783, "grad_norm": 0.17351198196411133, "learning_rate": 2.9855665578599093e-05, "loss": 0.0374, "step": 5520 }, { "epoch": 1.6089613034623218, "grad_norm": 0.23560835421085358, "learning_rate": 2.9854519197598504e-05, "loss": 0.0388, "step": 5530 }, { "epoch": 1.6118708175734653, "grad_norm": 0.15193478763103485, "learning_rate": 2.9853368304187197e-05, "loss": 0.0382, "step": 5540 }, { "epoch": 1.6147803316846088, "grad_norm": 0.18336932361125946, "learning_rate": 2.9852212898714783e-05, "loss": 0.0375, "step": 5550 }, { "epoch": 1.617689845795752, "grad_norm": 0.1863551288843155, "learning_rate": 2.9851052981532233e-05, "loss": 0.0377, "step": 5560 }, { "epoch": 1.6205993599068955, "grad_norm": 0.15525485575199127, "learning_rate": 2.9849888552991908e-05, "loss": 0.04, "step": 5570 }, { "epoch": 1.623508874018039, "grad_norm": 0.1792917400598526, "learning_rate": 2.9848719613447535e-05, "loss": 0.0358, "step": 5580 }, { "epoch": 1.6264183881291823, "grad_norm": 0.15848185122013092, "learning_rate": 2.9847546163254194e-05, "loss": 0.0385, "step": 5590 }, { "epoch": 1.629327902240326, "grad_norm": 0.12040398269891739, "learning_rate": 2.984636820276836e-05, "loss": 0.0387, "step": 5600 }, { "epoch": 1.6322374163514692, "grad_norm": 0.14530591666698456, "learning_rate": 2.984518573234786e-05, "loss": 0.0364, "step": 5610 }, { "epoch": 1.6351469304626127, "grad_norm": 0.14632205665111542, "learning_rate": 2.98439987523519e-05, "loss": 0.0403, "step": 5620 }, { "epoch": 1.6380564445737562, "grad_norm": 0.15447041392326355, "learning_rate": 2.9842807263141052e-05, "loss": 0.0384, "step": 5630 }, { "epoch": 1.6409659586848995, "grad_norm": 0.1574178785085678, "learning_rate": 2.9841611265077256e-05, "loss": 0.0361, "step": 5640 }, { "epoch": 1.6438754727960432, "grad_norm": 0.16078434884548187, "learning_rate": 2.9840410758523832e-05, "loss": 0.0344, "step": 5650 }, { "epoch": 1.6467849869071864, "grad_norm": 0.1994800865650177, "learning_rate": 2.9839205743845453e-05, "loss": 0.036, "step": 5660 }, { "epoch": 1.64969450101833, "grad_norm": 0.21108701825141907, "learning_rate": 2.983799622140818e-05, "loss": 0.0376, "step": 5670 }, { "epoch": 1.6526040151294734, "grad_norm": 0.20461377501487732, "learning_rate": 2.9836782191579425e-05, "loss": 0.0406, "step": 5680 }, { "epoch": 1.6555135292406167, "grad_norm": 0.14715217053890228, "learning_rate": 2.983556365472799e-05, "loss": 0.0381, "step": 5690 }, { "epoch": 1.6584230433517604, "grad_norm": 0.19061970710754395, "learning_rate": 2.9834340611224022e-05, "loss": 0.0392, "step": 5700 }, { "epoch": 1.6613325574629036, "grad_norm": 0.14467044174671173, "learning_rate": 2.9833113061439057e-05, "loss": 0.036, "step": 5710 }, { "epoch": 1.6642420715740471, "grad_norm": 0.2011120766401291, "learning_rate": 2.9831881005745986e-05, "loss": 0.0352, "step": 5720 }, { "epoch": 1.6671515856851906, "grad_norm": 0.18218469619750977, "learning_rate": 2.983064444451908e-05, "loss": 0.0362, "step": 5730 }, { "epoch": 1.6700610997963339, "grad_norm": 0.14674565196037292, "learning_rate": 2.9829403378133974e-05, "loss": 0.0356, "step": 5740 }, { "epoch": 1.6729706139074776, "grad_norm": 0.17365239560604095, "learning_rate": 2.9828157806967668e-05, "loss": 0.0349, "step": 5750 }, { "epoch": 1.6758801280186209, "grad_norm": 0.1751367151737213, "learning_rate": 2.9826907731398533e-05, "loss": 0.0383, "step": 5760 }, { "epoch": 1.6787896421297643, "grad_norm": 0.19567906856536865, "learning_rate": 2.9825653151806315e-05, "loss": 0.0399, "step": 5770 }, { "epoch": 1.6816991562409078, "grad_norm": 0.18082977831363678, "learning_rate": 2.9824394068572114e-05, "loss": 0.0392, "step": 5780 }, { "epoch": 1.684608670352051, "grad_norm": 0.1459677368402481, "learning_rate": 2.9823130482078415e-05, "loss": 0.0369, "step": 5790 }, { "epoch": 1.6875181844631948, "grad_norm": 0.1880727857351303, "learning_rate": 2.9821862392709054e-05, "loss": 0.0367, "step": 5800 }, { "epoch": 1.690427698574338, "grad_norm": 0.1275312900543213, "learning_rate": 2.982058980084925e-05, "loss": 0.0398, "step": 5810 }, { "epoch": 1.6933372126854815, "grad_norm": 0.17978420853614807, "learning_rate": 2.9819312706885577e-05, "loss": 0.0388, "step": 5820 }, { "epoch": 1.696246726796625, "grad_norm": 0.199208065867424, "learning_rate": 2.9818031111205986e-05, "loss": 0.0356, "step": 5830 }, { "epoch": 1.6991562409077683, "grad_norm": 0.13447484374046326, "learning_rate": 2.981674501419979e-05, "loss": 0.0345, "step": 5840 }, { "epoch": 1.702065755018912, "grad_norm": 0.16746380925178528, "learning_rate": 2.981545441625767e-05, "loss": 0.0364, "step": 5850 }, { "epoch": 1.7049752691300553, "grad_norm": 0.15336604416370392, "learning_rate": 2.9814159317771682e-05, "loss": 0.0381, "step": 5860 }, { "epoch": 1.7078847832411987, "grad_norm": 0.17565390467643738, "learning_rate": 2.9812859719135236e-05, "loss": 0.0381, "step": 5870 }, { "epoch": 1.7107942973523422, "grad_norm": 0.16336455941200256, "learning_rate": 2.981155562074312e-05, "loss": 0.0375, "step": 5880 }, { "epoch": 1.7137038114634855, "grad_norm": 0.17726610600948334, "learning_rate": 2.9810247022991483e-05, "loss": 0.036, "step": 5890 }, { "epoch": 1.7166133255746292, "grad_norm": 0.15183603763580322, "learning_rate": 2.9808933926277842e-05, "loss": 0.0388, "step": 5900 }, { "epoch": 1.7195228396857725, "grad_norm": 0.16755259037017822, "learning_rate": 2.9807616331001078e-05, "loss": 0.038, "step": 5910 }, { "epoch": 1.722432353796916, "grad_norm": 0.19688290357589722, "learning_rate": 2.9806294237561452e-05, "loss": 0.0378, "step": 5920 }, { "epoch": 1.7253418679080594, "grad_norm": 0.15453219413757324, "learning_rate": 2.980496764636057e-05, "loss": 0.0387, "step": 5930 }, { "epoch": 1.7282513820192027, "grad_norm": 0.2036299705505371, "learning_rate": 2.9803636557801417e-05, "loss": 0.0368, "step": 5940 }, { "epoch": 1.7311608961303462, "grad_norm": 0.1515842229127884, "learning_rate": 2.980230097228835e-05, "loss": 0.0371, "step": 5950 }, { "epoch": 1.7340704102414897, "grad_norm": 0.18470697104930878, "learning_rate": 2.9800960890227076e-05, "loss": 0.0377, "step": 5960 }, { "epoch": 1.7369799243526332, "grad_norm": 0.16464018821716309, "learning_rate": 2.979961631202468e-05, "loss": 0.0393, "step": 5970 }, { "epoch": 1.7398894384637766, "grad_norm": 0.13872914016246796, "learning_rate": 2.9798267238089613e-05, "loss": 0.0387, "step": 5980 }, { "epoch": 1.74279895257492, "grad_norm": 0.1568777710199356, "learning_rate": 2.9796913668831684e-05, "loss": 0.0367, "step": 5990 }, { "epoch": 1.7457084666860634, "grad_norm": 0.18474642932415009, "learning_rate": 2.979555560466207e-05, "loss": 0.0373, "step": 6000 }, { "epoch": 1.7486179807972069, "grad_norm": 0.1459096372127533, "learning_rate": 2.9794193045993317e-05, "loss": 0.0394, "step": 6010 }, { "epoch": 1.7515274949083504, "grad_norm": 0.13968819379806519, "learning_rate": 2.9792825993239337e-05, "loss": 0.0366, "step": 6020 }, { "epoch": 1.7544370090194938, "grad_norm": 0.15066953003406525, "learning_rate": 2.9791454446815404e-05, "loss": 0.0352, "step": 6030 }, { "epoch": 1.757346523130637, "grad_norm": 0.19506770372390747, "learning_rate": 2.9790078407138154e-05, "loss": 0.0368, "step": 6040 }, { "epoch": 1.7602560372417806, "grad_norm": 0.1775238811969757, "learning_rate": 2.9788697874625588e-05, "loss": 0.0382, "step": 6050 }, { "epoch": 1.763165551352924, "grad_norm": 0.21141788363456726, "learning_rate": 2.9787312849697088e-05, "loss": 0.0364, "step": 6060 }, { "epoch": 1.7660750654640673, "grad_norm": 0.1449492871761322, "learning_rate": 2.978592333277338e-05, "loss": 0.0369, "step": 6070 }, { "epoch": 1.768984579575211, "grad_norm": 0.18315206468105316, "learning_rate": 2.9784529324276557e-05, "loss": 0.0372, "step": 6080 }, { "epoch": 1.7718940936863543, "grad_norm": 0.13627880811691284, "learning_rate": 2.9783130824630093e-05, "loss": 0.0362, "step": 6090 }, { "epoch": 1.7748036077974978, "grad_norm": 0.18635053932666779, "learning_rate": 2.9781727834258806e-05, "loss": 0.039, "step": 6100 }, { "epoch": 1.7777131219086413, "grad_norm": 0.146100252866745, "learning_rate": 2.97803203535889e-05, "loss": 0.0359, "step": 6110 }, { "epoch": 1.7806226360197845, "grad_norm": 0.157785564661026, "learning_rate": 2.977890838304792e-05, "loss": 0.0363, "step": 6120 }, { "epoch": 1.7835321501309283, "grad_norm": 0.15798725187778473, "learning_rate": 2.9777491923064782e-05, "loss": 0.0355, "step": 6130 }, { "epoch": 1.7864416642420715, "grad_norm": 0.16318079829216003, "learning_rate": 2.977607097406978e-05, "loss": 0.0365, "step": 6140 }, { "epoch": 1.789351178353215, "grad_norm": 0.16623836755752563, "learning_rate": 2.9774645536494552e-05, "loss": 0.0366, "step": 6150 }, { "epoch": 1.7922606924643585, "grad_norm": 0.13428665697574615, "learning_rate": 2.9773215610772117e-05, "loss": 0.0356, "step": 6160 }, { "epoch": 1.7951702065755017, "grad_norm": 0.16602857410907745, "learning_rate": 2.9771781197336837e-05, "loss": 0.0359, "step": 6170 }, { "epoch": 1.7980797206866455, "grad_norm": 0.14112763106822968, "learning_rate": 2.977034229662446e-05, "loss": 0.0359, "step": 6180 }, { "epoch": 1.8009892347977887, "grad_norm": 0.18402452766895294, "learning_rate": 2.976889890907208e-05, "loss": 0.0347, "step": 6190 }, { "epoch": 1.8038987489089322, "grad_norm": 0.16556233167648315, "learning_rate": 2.9767451035118164e-05, "loss": 0.0346, "step": 6200 }, { "epoch": 1.8068082630200757, "grad_norm": 0.1676390916109085, "learning_rate": 2.9765998675202536e-05, "loss": 0.0332, "step": 6210 }, { "epoch": 1.809717777131219, "grad_norm": 0.13689298927783966, "learning_rate": 2.9764541829766382e-05, "loss": 0.0374, "step": 6220 }, { "epoch": 1.8126272912423627, "grad_norm": 0.11842726171016693, "learning_rate": 2.9763080499252253e-05, "loss": 0.0348, "step": 6230 }, { "epoch": 1.815536805353506, "grad_norm": 0.11891444772481918, "learning_rate": 2.9761614684104064e-05, "loss": 0.0361, "step": 6240 }, { "epoch": 1.8184463194646494, "grad_norm": 0.1468084156513214, "learning_rate": 2.9760144384767095e-05, "loss": 0.0357, "step": 6250 }, { "epoch": 1.821355833575793, "grad_norm": 0.11740945279598236, "learning_rate": 2.9758669601687986e-05, "loss": 0.0351, "step": 6260 }, { "epoch": 1.8242653476869362, "grad_norm": 0.146686390042305, "learning_rate": 2.9757190335314722e-05, "loss": 0.0357, "step": 6270 }, { "epoch": 1.8271748617980799, "grad_norm": 0.1618788242340088, "learning_rate": 2.975570658609668e-05, "loss": 0.0362, "step": 6280 }, { "epoch": 1.8300843759092231, "grad_norm": 0.1638021171092987, "learning_rate": 2.9754218354484582e-05, "loss": 0.0355, "step": 6290 }, { "epoch": 1.8329938900203666, "grad_norm": 0.11893215030431747, "learning_rate": 2.9752725640930504e-05, "loss": 0.0344, "step": 6300 }, { "epoch": 1.83590340413151, "grad_norm": 0.16369390487670898, "learning_rate": 2.9751228445887905e-05, "loss": 0.0353, "step": 6310 }, { "epoch": 1.8388129182426534, "grad_norm": 0.19052161276340485, "learning_rate": 2.9749726769811588e-05, "loss": 0.0367, "step": 6320 }, { "epoch": 1.841722432353797, "grad_norm": 0.16650381684303284, "learning_rate": 2.9748220613157717e-05, "loss": 0.0331, "step": 6330 }, { "epoch": 1.8446319464649403, "grad_norm": 0.10830635577440262, "learning_rate": 2.9746709976383832e-05, "loss": 0.0343, "step": 6340 }, { "epoch": 1.8475414605760838, "grad_norm": 0.15280932188034058, "learning_rate": 2.974519485994882e-05, "loss": 0.0341, "step": 6350 }, { "epoch": 1.8504509746872273, "grad_norm": 0.20866042375564575, "learning_rate": 2.9743675264312934e-05, "loss": 0.0371, "step": 6360 }, { "epoch": 1.8533604887983706, "grad_norm": 0.15468186140060425, "learning_rate": 2.9742151189937784e-05, "loss": 0.0357, "step": 6370 }, { "epoch": 1.8562700029095143, "grad_norm": 0.18982955813407898, "learning_rate": 2.974062263728635e-05, "loss": 0.0382, "step": 6380 }, { "epoch": 1.8591795170206575, "grad_norm": 0.22638069093227386, "learning_rate": 2.973908960682296e-05, "loss": 0.0363, "step": 6390 }, { "epoch": 1.862089031131801, "grad_norm": 0.14602981507778168, "learning_rate": 2.9737552099013313e-05, "loss": 0.0389, "step": 6400 }, { "epoch": 1.8649985452429445, "grad_norm": 0.18092665076255798, "learning_rate": 2.9736010114324462e-05, "loss": 0.0362, "step": 6410 }, { "epoch": 1.8679080593540878, "grad_norm": 0.17147880792617798, "learning_rate": 2.9734463653224813e-05, "loss": 0.0352, "step": 6420 }, { "epoch": 1.8708175734652313, "grad_norm": 0.2115580141544342, "learning_rate": 2.9732912716184155e-05, "loss": 0.0368, "step": 6430 }, { "epoch": 1.8737270875763747, "grad_norm": 0.16797061264514923, "learning_rate": 2.9731357303673607e-05, "loss": 0.0349, "step": 6440 }, { "epoch": 1.8766366016875182, "grad_norm": 0.1289326250553131, "learning_rate": 2.9729797416165677e-05, "loss": 0.036, "step": 6450 }, { "epoch": 1.8795461157986617, "grad_norm": 0.12207240611314774, "learning_rate": 2.9728233054134202e-05, "loss": 0.0335, "step": 6460 }, { "epoch": 1.882455629909805, "grad_norm": 0.1424020528793335, "learning_rate": 2.9726664218054405e-05, "loss": 0.0358, "step": 6470 }, { "epoch": 1.8853651440209485, "grad_norm": 0.17096497118473053, "learning_rate": 2.9725090908402856e-05, "loss": 0.0364, "step": 6480 }, { "epoch": 1.888274658132092, "grad_norm": 0.15180440247058868, "learning_rate": 2.972351312565748e-05, "loss": 0.034, "step": 6490 }, { "epoch": 1.8911841722432354, "grad_norm": 0.15747098624706268, "learning_rate": 2.9721930870297567e-05, "loss": 0.0349, "step": 6500 }, { "epoch": 1.894093686354379, "grad_norm": 0.162339448928833, "learning_rate": 2.9720344142803766e-05, "loss": 0.035, "step": 6510 }, { "epoch": 1.8970032004655222, "grad_norm": 0.1611018180847168, "learning_rate": 2.9718752943658085e-05, "loss": 0.0336, "step": 6520 }, { "epoch": 1.8999127145766657, "grad_norm": 0.13288071751594543, "learning_rate": 2.9717157273343885e-05, "loss": 0.0342, "step": 6530 }, { "epoch": 1.9028222286878091, "grad_norm": 0.15235835313796997, "learning_rate": 2.971555713234589e-05, "loss": 0.0351, "step": 6540 }, { "epoch": 1.9057317427989524, "grad_norm": 0.14269982278347015, "learning_rate": 2.9713952521150176e-05, "loss": 0.0352, "step": 6550 }, { "epoch": 1.9086412569100961, "grad_norm": 0.1280871331691742, "learning_rate": 2.971234344024419e-05, "loss": 0.0364, "step": 6560 }, { "epoch": 1.9115507710212394, "grad_norm": 0.1298246830701828, "learning_rate": 2.971072989011672e-05, "loss": 0.0329, "step": 6570 }, { "epoch": 1.9144602851323829, "grad_norm": 0.17855219542980194, "learning_rate": 2.9709111871257927e-05, "loss": 0.0337, "step": 6580 }, { "epoch": 1.9173697992435264, "grad_norm": 0.1425316482782364, "learning_rate": 2.9707489384159318e-05, "loss": 0.0356, "step": 6590 }, { "epoch": 1.9202793133546696, "grad_norm": 0.144452765583992, "learning_rate": 2.9705862429313763e-05, "loss": 0.0343, "step": 6600 }, { "epoch": 1.9231888274658133, "grad_norm": 0.14052753150463104, "learning_rate": 2.9704231007215488e-05, "loss": 0.0341, "step": 6610 }, { "epoch": 1.9260983415769566, "grad_norm": 0.12434771656990051, "learning_rate": 2.9702595118360077e-05, "loss": 0.034, "step": 6620 }, { "epoch": 1.9290078556881, "grad_norm": 0.1613381803035736, "learning_rate": 2.9700954763244465e-05, "loss": 0.034, "step": 6630 }, { "epoch": 1.9319173697992436, "grad_norm": 0.13137777149677277, "learning_rate": 2.9699309942366953e-05, "loss": 0.033, "step": 6640 }, { "epoch": 1.9348268839103868, "grad_norm": 0.15815593302249908, "learning_rate": 2.969766065622719e-05, "loss": 0.0345, "step": 6650 }, { "epoch": 1.9377363980215305, "grad_norm": 0.19146351516246796, "learning_rate": 2.96960069053262e-05, "loss": 0.0375, "step": 6660 }, { "epoch": 1.9406459121326738, "grad_norm": 0.19818228483200073, "learning_rate": 2.9694348690166327e-05, "loss": 0.0365, "step": 6670 }, { "epoch": 1.9435554262438173, "grad_norm": 0.17502936720848083, "learning_rate": 2.969268601125131e-05, "loss": 0.0372, "step": 6680 }, { "epoch": 1.9464649403549608, "grad_norm": 0.18091125786304474, "learning_rate": 2.969101886908622e-05, "loss": 0.0327, "step": 6690 }, { "epoch": 1.949374454466104, "grad_norm": 0.17635329067707062, "learning_rate": 2.968934726417749e-05, "loss": 0.0351, "step": 6700 }, { "epoch": 1.9522839685772477, "grad_norm": 0.12751898169517517, "learning_rate": 2.968767119703291e-05, "loss": 0.0362, "step": 6710 }, { "epoch": 1.955193482688391, "grad_norm": 0.15480786561965942, "learning_rate": 2.9685990668161628e-05, "loss": 0.0373, "step": 6720 }, { "epoch": 1.9581029967995345, "grad_norm": 0.1508558839559555, "learning_rate": 2.9684305678074137e-05, "loss": 0.0327, "step": 6730 }, { "epoch": 1.961012510910678, "grad_norm": 0.17679965496063232, "learning_rate": 2.9682616227282304e-05, "loss": 0.0352, "step": 6740 }, { "epoch": 1.9639220250218212, "grad_norm": 0.19855165481567383, "learning_rate": 2.968092231629933e-05, "loss": 0.0359, "step": 6750 }, { "epoch": 1.966831539132965, "grad_norm": 0.1665288358926773, "learning_rate": 2.967922394563978e-05, "loss": 0.0343, "step": 6760 }, { "epoch": 1.9697410532441082, "grad_norm": 0.1367957592010498, "learning_rate": 2.9677521115819585e-05, "loss": 0.0355, "step": 6770 }, { "epoch": 1.9726505673552517, "grad_norm": 0.16411623358726501, "learning_rate": 2.967581382735601e-05, "loss": 0.0356, "step": 6780 }, { "epoch": 1.9755600814663952, "grad_norm": 0.15048205852508545, "learning_rate": 2.9674102080767685e-05, "loss": 0.0373, "step": 6790 }, { "epoch": 1.9784695955775384, "grad_norm": 0.1672627180814743, "learning_rate": 2.9672385876574597e-05, "loss": 0.0362, "step": 6800 }, { "epoch": 1.9813791096886821, "grad_norm": 0.16154631972312927, "learning_rate": 2.967066521529808e-05, "loss": 0.0344, "step": 6810 }, { "epoch": 1.9842886237998254, "grad_norm": 0.1306646764278412, "learning_rate": 2.966894009746083e-05, "loss": 0.0338, "step": 6820 }, { "epoch": 1.9871981379109689, "grad_norm": 0.12356775254011154, "learning_rate": 2.9667210523586888e-05, "loss": 0.0329, "step": 6830 }, { "epoch": 1.9901076520221124, "grad_norm": 0.11662308126688004, "learning_rate": 2.966547649420165e-05, "loss": 0.0349, "step": 6840 }, { "epoch": 1.9930171661332556, "grad_norm": 0.1261843591928482, "learning_rate": 2.9663738009831877e-05, "loss": 0.0352, "step": 6850 }, { "epoch": 1.9959266802443993, "grad_norm": 0.15809977054595947, "learning_rate": 2.966199507100567e-05, "loss": 0.0354, "step": 6860 }, { "epoch": 1.9988361943555426, "grad_norm": 0.12787644565105438, "learning_rate": 2.9660247678252484e-05, "loss": 0.0352, "step": 6870 }, { "epoch": 2.001745708466686, "grad_norm": 0.17359381914138794, "learning_rate": 2.965849583210314e-05, "loss": 0.0365, "step": 6880 }, { "epoch": 2.0046552225778296, "grad_norm": 0.1549927294254303, "learning_rate": 2.9656739533089794e-05, "loss": 0.0334, "step": 6890 }, { "epoch": 2.007564736688973, "grad_norm": 0.1583806425333023, "learning_rate": 2.9654978781745968e-05, "loss": 0.0336, "step": 6900 }, { "epoch": 2.0104742508001165, "grad_norm": 0.15174974501132965, "learning_rate": 2.9653213578606534e-05, "loss": 0.0339, "step": 6910 }, { "epoch": 2.01338376491126, "grad_norm": 0.18720115721225739, "learning_rate": 2.9651443924207704e-05, "loss": 0.0368, "step": 6920 }, { "epoch": 2.016293279022403, "grad_norm": 0.15250587463378906, "learning_rate": 2.964966981908706e-05, "loss": 0.032, "step": 6930 }, { "epoch": 2.019202793133547, "grad_norm": 0.1256183683872223, "learning_rate": 2.964789126378353e-05, "loss": 0.0358, "step": 6940 }, { "epoch": 2.02211230724469, "grad_norm": 0.18715889751911163, "learning_rate": 2.9646108258837388e-05, "loss": 0.035, "step": 6950 }, { "epoch": 2.0250218213558338, "grad_norm": 0.13545945286750793, "learning_rate": 2.9644320804790267e-05, "loss": 0.0347, "step": 6960 }, { "epoch": 2.027931335466977, "grad_norm": 0.18264997005462646, "learning_rate": 2.9642528902185142e-05, "loss": 0.0321, "step": 6970 }, { "epoch": 2.0308408495781203, "grad_norm": 0.15477201342582703, "learning_rate": 2.9640732551566354e-05, "loss": 0.0348, "step": 6980 }, { "epoch": 2.033750363689264, "grad_norm": 0.19302043318748474, "learning_rate": 2.9638931753479578e-05, "loss": 0.0354, "step": 6990 }, { "epoch": 2.0366598778004072, "grad_norm": 0.19930045306682587, "learning_rate": 2.9637126508471858e-05, "loss": 0.0351, "step": 7000 }, { "epoch": 2.039569391911551, "grad_norm": 0.16138190031051636, "learning_rate": 2.9635316817091577e-05, "loss": 0.034, "step": 7010 }, { "epoch": 2.042478906022694, "grad_norm": 0.16918005049228668, "learning_rate": 2.9633502679888467e-05, "loss": 0.0342, "step": 7020 }, { "epoch": 2.0453884201338375, "grad_norm": 0.15915723145008087, "learning_rate": 2.963168409741362e-05, "loss": 0.0351, "step": 7030 }, { "epoch": 2.048297934244981, "grad_norm": 0.1261761337518692, "learning_rate": 2.9629861070219467e-05, "loss": 0.0329, "step": 7040 }, { "epoch": 2.0512074483561245, "grad_norm": 0.21282994747161865, "learning_rate": 2.9628033598859807e-05, "loss": 0.0348, "step": 7050 }, { "epoch": 2.054116962467268, "grad_norm": 0.15933465957641602, "learning_rate": 2.9626201683889764e-05, "loss": 0.0344, "step": 7060 }, { "epoch": 2.0570264765784114, "grad_norm": 0.1144433245062828, "learning_rate": 2.9624365325865835e-05, "loss": 0.0333, "step": 7070 }, { "epoch": 2.0599359906895547, "grad_norm": 0.16010084748268127, "learning_rate": 2.962252452534585e-05, "loss": 0.0317, "step": 7080 }, { "epoch": 2.0628455048006984, "grad_norm": 0.13850802183151245, "learning_rate": 2.9620679282889006e-05, "loss": 0.0329, "step": 7090 }, { "epoch": 2.0657550189118417, "grad_norm": 0.1412331461906433, "learning_rate": 2.9618829599055833e-05, "loss": 0.0339, "step": 7100 }, { "epoch": 2.0686645330229854, "grad_norm": 0.17299632728099823, "learning_rate": 2.961697547440821e-05, "loss": 0.032, "step": 7110 }, { "epoch": 2.0715740471341286, "grad_norm": 0.15454614162445068, "learning_rate": 2.9615116909509383e-05, "loss": 0.0348, "step": 7120 }, { "epoch": 2.074483561245272, "grad_norm": 0.15919791162014008, "learning_rate": 2.9613253904923924e-05, "loss": 0.0354, "step": 7130 }, { "epoch": 2.0773930753564156, "grad_norm": 0.1735571175813675, "learning_rate": 2.9611386461217772e-05, "loss": 0.0356, "step": 7140 }, { "epoch": 2.080302589467559, "grad_norm": 0.23105452954769135, "learning_rate": 2.96095145789582e-05, "loss": 0.0346, "step": 7150 }, { "epoch": 2.0832121035787026, "grad_norm": 0.1713782548904419, "learning_rate": 2.9607638258713845e-05, "loss": 0.0347, "step": 7160 }, { "epoch": 2.086121617689846, "grad_norm": 0.22632606327533722, "learning_rate": 2.960575750105468e-05, "loss": 0.0368, "step": 7170 }, { "epoch": 2.089031131800989, "grad_norm": 0.11418648064136505, "learning_rate": 2.9603872306552025e-05, "loss": 0.0339, "step": 7180 }, { "epoch": 2.091940645912133, "grad_norm": 0.1641159951686859, "learning_rate": 2.9601982675778562e-05, "loss": 0.0327, "step": 7190 }, { "epoch": 2.094850160023276, "grad_norm": 0.17741326987743378, "learning_rate": 2.9600088609308294e-05, "loss": 0.035, "step": 7200 }, { "epoch": 2.0977596741344193, "grad_norm": 0.15380780398845673, "learning_rate": 2.959819010771661e-05, "loss": 0.033, "step": 7210 }, { "epoch": 2.100669188245563, "grad_norm": 0.16107186675071716, "learning_rate": 2.959628717158021e-05, "loss": 0.0348, "step": 7220 }, { "epoch": 2.1035787023567063, "grad_norm": 0.15895412862300873, "learning_rate": 2.959437980147716e-05, "loss": 0.0329, "step": 7230 }, { "epoch": 2.10648821646785, "grad_norm": 0.1337195485830307, "learning_rate": 2.959246799798687e-05, "loss": 0.031, "step": 7240 }, { "epoch": 2.1093977305789933, "grad_norm": 0.09901538491249084, "learning_rate": 2.9590551761690095e-05, "loss": 0.0325, "step": 7250 }, { "epoch": 2.112307244690137, "grad_norm": 0.20116236805915833, "learning_rate": 2.9588631093168936e-05, "loss": 0.0335, "step": 7260 }, { "epoch": 2.1152167588012802, "grad_norm": 0.121772900223732, "learning_rate": 2.9586705993006837e-05, "loss": 0.0326, "step": 7270 }, { "epoch": 2.1181262729124235, "grad_norm": 0.15257829427719116, "learning_rate": 2.95847764617886e-05, "loss": 0.0367, "step": 7280 }, { "epoch": 2.121035787023567, "grad_norm": 0.1460602879524231, "learning_rate": 2.9582842500100364e-05, "loss": 0.0327, "step": 7290 }, { "epoch": 2.1239453011347105, "grad_norm": 0.14754845201969147, "learning_rate": 2.958090410852961e-05, "loss": 0.0338, "step": 7300 }, { "epoch": 2.1268548152458537, "grad_norm": 0.15481463074684143, "learning_rate": 2.9578961287665175e-05, "loss": 0.0332, "step": 7310 }, { "epoch": 2.1297643293569974, "grad_norm": 0.13423536717891693, "learning_rate": 2.9577014038097238e-05, "loss": 0.0332, "step": 7320 }, { "epoch": 2.1326738434681407, "grad_norm": 0.13352298736572266, "learning_rate": 2.9575062360417324e-05, "loss": 0.0328, "step": 7330 }, { "epoch": 2.1355833575792844, "grad_norm": 0.1643143594264984, "learning_rate": 2.957310625521829e-05, "loss": 0.0357, "step": 7340 }, { "epoch": 2.1384928716904277, "grad_norm": 0.19456438720226288, "learning_rate": 2.957114572309436e-05, "loss": 0.0335, "step": 7350 }, { "epoch": 2.141402385801571, "grad_norm": 0.15606363117694855, "learning_rate": 2.956918076464109e-05, "loss": 0.0337, "step": 7360 }, { "epoch": 2.1443118999127146, "grad_norm": 0.1480807662010193, "learning_rate": 2.9567211380455376e-05, "loss": 0.0342, "step": 7370 }, { "epoch": 2.147221414023858, "grad_norm": 0.1102185994386673, "learning_rate": 2.9565237571135472e-05, "loss": 0.0317, "step": 7380 }, { "epoch": 2.1501309281350016, "grad_norm": 0.1528310924768448, "learning_rate": 2.9563259337280967e-05, "loss": 0.033, "step": 7390 }, { "epoch": 2.153040442246145, "grad_norm": 0.1480344831943512, "learning_rate": 2.9561276679492794e-05, "loss": 0.0319, "step": 7400 }, { "epoch": 2.155949956357288, "grad_norm": 0.14290300011634827, "learning_rate": 2.9559289598373236e-05, "loss": 0.0327, "step": 7410 }, { "epoch": 2.158859470468432, "grad_norm": 0.1603696644306183, "learning_rate": 2.9557298094525913e-05, "loss": 0.0354, "step": 7420 }, { "epoch": 2.161768984579575, "grad_norm": 0.14182749390602112, "learning_rate": 2.955530216855579e-05, "loss": 0.0316, "step": 7430 }, { "epoch": 2.164678498690719, "grad_norm": 0.16920924186706543, "learning_rate": 2.955330182106918e-05, "loss": 0.0348, "step": 7440 }, { "epoch": 2.167588012801862, "grad_norm": 0.1681007593870163, "learning_rate": 2.9551297052673734e-05, "loss": 0.0333, "step": 7450 }, { "epoch": 2.1704975269130053, "grad_norm": 0.1210852861404419, "learning_rate": 2.954928786397845e-05, "loss": 0.0344, "step": 7460 }, { "epoch": 2.173407041024149, "grad_norm": 0.17389433085918427, "learning_rate": 2.9547274255593665e-05, "loss": 0.0372, "step": 7470 }, { "epoch": 2.1763165551352923, "grad_norm": 0.15677979588508606, "learning_rate": 2.9545256228131058e-05, "loss": 0.0305, "step": 7480 }, { "epoch": 2.179226069246436, "grad_norm": 0.1417347490787506, "learning_rate": 2.9543233782203653e-05, "loss": 0.0346, "step": 7490 }, { "epoch": 2.1821355833575793, "grad_norm": 0.14048030972480774, "learning_rate": 2.954120691842582e-05, "loss": 0.0312, "step": 7500 }, { "epoch": 2.1850450974687226, "grad_norm": 0.22249263525009155, "learning_rate": 2.9539175637413264e-05, "loss": 0.0333, "step": 7510 }, { "epoch": 2.1879546115798663, "grad_norm": 0.135530024766922, "learning_rate": 2.9537139939783033e-05, "loss": 0.0326, "step": 7520 }, { "epoch": 2.1908641256910095, "grad_norm": 0.12757661938667297, "learning_rate": 2.9535099826153516e-05, "loss": 0.0332, "step": 7530 }, { "epoch": 2.1937736398021532, "grad_norm": 0.15437163412570953, "learning_rate": 2.9533055297144454e-05, "loss": 0.0361, "step": 7540 }, { "epoch": 2.1966831539132965, "grad_norm": 0.1461186558008194, "learning_rate": 2.953100635337691e-05, "loss": 0.0351, "step": 7550 }, { "epoch": 2.1995926680244398, "grad_norm": 0.13212771713733673, "learning_rate": 2.952895299547331e-05, "loss": 0.0337, "step": 7560 }, { "epoch": 2.2025021821355835, "grad_norm": 0.16748753190040588, "learning_rate": 2.9526895224057402e-05, "loss": 0.0333, "step": 7570 }, { "epoch": 2.2054116962467267, "grad_norm": 0.13756662607192993, "learning_rate": 2.9524833039754284e-05, "loss": 0.0317, "step": 7580 }, { "epoch": 2.2083212103578704, "grad_norm": 0.16727396845817566, "learning_rate": 2.9522766443190398e-05, "loss": 0.033, "step": 7590 }, { "epoch": 2.2112307244690137, "grad_norm": 0.129810631275177, "learning_rate": 2.9520695434993516e-05, "loss": 0.0364, "step": 7600 }, { "epoch": 2.214140238580157, "grad_norm": 0.17120836675167084, "learning_rate": 2.9518620015792757e-05, "loss": 0.0326, "step": 7610 }, { "epoch": 2.2170497526913007, "grad_norm": 0.13351064920425415, "learning_rate": 2.951654018621858e-05, "loss": 0.0327, "step": 7620 }, { "epoch": 2.219959266802444, "grad_norm": 0.1571197211742401, "learning_rate": 2.951445594690278e-05, "loss": 0.0345, "step": 7630 }, { "epoch": 2.2228687809135876, "grad_norm": 0.12039355933666229, "learning_rate": 2.95123672984785e-05, "loss": 0.034, "step": 7640 }, { "epoch": 2.225778295024731, "grad_norm": 0.10816852748394012, "learning_rate": 2.9510274241580207e-05, "loss": 0.0333, "step": 7650 }, { "epoch": 2.228687809135874, "grad_norm": 0.16954249143600464, "learning_rate": 2.9508176776843726e-05, "loss": 0.0319, "step": 7660 }, { "epoch": 2.231597323247018, "grad_norm": 0.13359926640987396, "learning_rate": 2.95060749049062e-05, "loss": 0.0318, "step": 7670 }, { "epoch": 2.234506837358161, "grad_norm": 0.1733793020248413, "learning_rate": 2.9503968626406133e-05, "loss": 0.0338, "step": 7680 }, { "epoch": 2.237416351469305, "grad_norm": 0.1378943771123886, "learning_rate": 2.9501857941983354e-05, "loss": 0.0334, "step": 7690 }, { "epoch": 2.240325865580448, "grad_norm": 0.13209664821624756, "learning_rate": 2.9499742852279025e-05, "loss": 0.0338, "step": 7700 }, { "epoch": 2.2432353796915914, "grad_norm": 0.13200686872005463, "learning_rate": 2.9497623357935666e-05, "loss": 0.0324, "step": 7710 }, { "epoch": 2.246144893802735, "grad_norm": 0.1332215517759323, "learning_rate": 2.9495499459597116e-05, "loss": 0.0336, "step": 7720 }, { "epoch": 2.2490544079138783, "grad_norm": 0.17743521928787231, "learning_rate": 2.9493371157908563e-05, "loss": 0.0321, "step": 7730 }, { "epoch": 2.2519639220250216, "grad_norm": 0.16496984660625458, "learning_rate": 2.9491238453516524e-05, "loss": 0.0321, "step": 7740 }, { "epoch": 2.2548734361361653, "grad_norm": 0.17292296886444092, "learning_rate": 2.9489101347068868e-05, "loss": 0.0338, "step": 7750 }, { "epoch": 2.2577829502473086, "grad_norm": 0.1320175677537918, "learning_rate": 2.948695983921478e-05, "loss": 0.0337, "step": 7760 }, { "epoch": 2.2606924643584523, "grad_norm": 0.14238597452640533, "learning_rate": 2.9484813930604804e-05, "loss": 0.034, "step": 7770 }, { "epoch": 2.2636019784695955, "grad_norm": 0.14998053014278412, "learning_rate": 2.9482663621890804e-05, "loss": 0.0349, "step": 7780 }, { "epoch": 2.2665114925807393, "grad_norm": 0.18344110250473022, "learning_rate": 2.9480508913725986e-05, "loss": 0.0348, "step": 7790 }, { "epoch": 2.2694210066918825, "grad_norm": 0.19543252885341644, "learning_rate": 2.9478349806764895e-05, "loss": 0.0324, "step": 7800 }, { "epoch": 2.2723305208030258, "grad_norm": 0.13014459609985352, "learning_rate": 2.9476186301663414e-05, "loss": 0.0315, "step": 7810 }, { "epoch": 2.2752400349141695, "grad_norm": 0.15764831006526947, "learning_rate": 2.9474018399078752e-05, "loss": 0.0312, "step": 7820 }, { "epoch": 2.2781495490253127, "grad_norm": 0.13951200246810913, "learning_rate": 2.947184609966947e-05, "loss": 0.0322, "step": 7830 }, { "epoch": 2.281059063136456, "grad_norm": 0.17071855068206787, "learning_rate": 2.9469669404095447e-05, "loss": 0.0314, "step": 7840 }, { "epoch": 2.2839685772475997, "grad_norm": 0.13564947247505188, "learning_rate": 2.9467488313017908e-05, "loss": 0.033, "step": 7850 }, { "epoch": 2.286878091358743, "grad_norm": 0.20172208547592163, "learning_rate": 2.9465302827099412e-05, "loss": 0.0319, "step": 7860 }, { "epoch": 2.2897876054698867, "grad_norm": 0.14083461463451385, "learning_rate": 2.9463112947003854e-05, "loss": 0.0303, "step": 7870 }, { "epoch": 2.29269711958103, "grad_norm": 0.15372303128242493, "learning_rate": 2.9460918673396455e-05, "loss": 0.0339, "step": 7880 }, { "epoch": 2.295606633692173, "grad_norm": 0.12249718606472015, "learning_rate": 2.9458720006943784e-05, "loss": 0.0325, "step": 7890 }, { "epoch": 2.298516147803317, "grad_norm": 0.17456722259521484, "learning_rate": 2.9456516948313736e-05, "loss": 0.0333, "step": 7900 }, { "epoch": 2.30142566191446, "grad_norm": 0.17923133075237274, "learning_rate": 2.9454309498175538e-05, "loss": 0.0333, "step": 7910 }, { "epoch": 2.304335176025604, "grad_norm": 0.15975257754325867, "learning_rate": 2.9452097657199762e-05, "loss": 0.0327, "step": 7920 }, { "epoch": 2.307244690136747, "grad_norm": 0.17091336846351624, "learning_rate": 2.9449881426058303e-05, "loss": 0.0343, "step": 7930 }, { "epoch": 2.3101542042478904, "grad_norm": 0.1289815604686737, "learning_rate": 2.9447660805424388e-05, "loss": 0.0314, "step": 7940 }, { "epoch": 2.313063718359034, "grad_norm": 0.183475062251091, "learning_rate": 2.94454357959726e-05, "loss": 0.034, "step": 7950 }, { "epoch": 2.3159732324701774, "grad_norm": 0.12243068963289261, "learning_rate": 2.9443206398378815e-05, "loss": 0.0324, "step": 7960 }, { "epoch": 2.318882746581321, "grad_norm": 0.2028583288192749, "learning_rate": 2.9440972613320282e-05, "loss": 0.033, "step": 7970 }, { "epoch": 2.3217922606924644, "grad_norm": 0.11873102188110352, "learning_rate": 2.943873444147556e-05, "loss": 0.0321, "step": 7980 }, { "epoch": 2.3247017748036076, "grad_norm": 0.11970862746238708, "learning_rate": 2.9436491883524547e-05, "loss": 0.0316, "step": 7990 }, { "epoch": 2.3276112889147513, "grad_norm": 0.1745399534702301, "learning_rate": 2.9434244940148472e-05, "loss": 0.0319, "step": 8000 }, { "epoch": 2.3305208030258946, "grad_norm": 0.14404653012752533, "learning_rate": 2.9431993612029903e-05, "loss": 0.0333, "step": 8010 }, { "epoch": 2.3334303171370383, "grad_norm": 0.15173417329788208, "learning_rate": 2.9429737899852728e-05, "loss": 0.0301, "step": 8020 }, { "epoch": 2.3363398312481816, "grad_norm": 0.15871521830558777, "learning_rate": 2.9427477804302174e-05, "loss": 0.034, "step": 8030 }, { "epoch": 2.339249345359325, "grad_norm": 0.14498674869537354, "learning_rate": 2.9425213326064797e-05, "loss": 0.0336, "step": 8040 }, { "epoch": 2.3421588594704685, "grad_norm": 0.176997572183609, "learning_rate": 2.942294446582849e-05, "loss": 0.0311, "step": 8050 }, { "epoch": 2.345068373581612, "grad_norm": 0.17124775052070618, "learning_rate": 2.9420671224282468e-05, "loss": 0.0336, "step": 8060 }, { "epoch": 2.347977887692755, "grad_norm": 0.17821529507637024, "learning_rate": 2.9418393602117284e-05, "loss": 0.0327, "step": 8070 }, { "epoch": 2.3508874018038988, "grad_norm": 0.1480492353439331, "learning_rate": 2.9416111600024816e-05, "loss": 0.0321, "step": 8080 }, { "epoch": 2.353796915915042, "grad_norm": 0.1611628532409668, "learning_rate": 2.9413825218698284e-05, "loss": 0.0316, "step": 8090 }, { "epoch": 2.3567064300261857, "grad_norm": 0.1709814965724945, "learning_rate": 2.9411534458832222e-05, "loss": 0.0359, "step": 8100 }, { "epoch": 2.359615944137329, "grad_norm": 0.16058418154716492, "learning_rate": 2.9409239321122513e-05, "loss": 0.0337, "step": 8110 }, { "epoch": 2.3625254582484727, "grad_norm": 0.16840115189552307, "learning_rate": 2.940693980626634e-05, "loss": 0.0317, "step": 8120 }, { "epoch": 2.365434972359616, "grad_norm": 0.14692817628383636, "learning_rate": 2.9404635914962258e-05, "loss": 0.0324, "step": 8130 }, { "epoch": 2.3683444864707592, "grad_norm": 0.1333029419183731, "learning_rate": 2.9402327647910113e-05, "loss": 0.0339, "step": 8140 }, { "epoch": 2.371254000581903, "grad_norm": 0.13448825478553772, "learning_rate": 2.94000150058111e-05, "loss": 0.0352, "step": 8150 }, { "epoch": 2.374163514693046, "grad_norm": 0.12746772170066833, "learning_rate": 2.939769798936774e-05, "loss": 0.0302, "step": 8160 }, { "epoch": 2.3770730288041895, "grad_norm": 0.13816919922828674, "learning_rate": 2.9395376599283878e-05, "loss": 0.0336, "step": 8170 }, { "epoch": 2.379982542915333, "grad_norm": 0.1413239687681198, "learning_rate": 2.9393050836264697e-05, "loss": 0.0328, "step": 8180 }, { "epoch": 2.3828920570264764, "grad_norm": 0.1272432655096054, "learning_rate": 2.9390720701016693e-05, "loss": 0.0315, "step": 8190 }, { "epoch": 2.38580157113762, "grad_norm": 0.13505901396274567, "learning_rate": 2.938838619424771e-05, "loss": 0.0316, "step": 8200 }, { "epoch": 2.3887110852487634, "grad_norm": 0.18118710815906525, "learning_rate": 2.93860473166669e-05, "loss": 0.0332, "step": 8210 }, { "epoch": 2.391620599359907, "grad_norm": 0.13766096532344818, "learning_rate": 2.938370406898476e-05, "loss": 0.0304, "step": 8220 }, { "epoch": 2.3945301134710504, "grad_norm": 0.14413727819919586, "learning_rate": 2.9381356451913103e-05, "loss": 0.0306, "step": 8230 }, { "epoch": 2.3974396275821936, "grad_norm": 0.19745385646820068, "learning_rate": 2.9379004466165072e-05, "loss": 0.0323, "step": 8240 }, { "epoch": 2.4003491416933374, "grad_norm": 0.16178420186042786, "learning_rate": 2.937664811245514e-05, "loss": 0.0309, "step": 8250 }, { "epoch": 2.4032586558044806, "grad_norm": 0.17761139571666718, "learning_rate": 2.9374287391499108e-05, "loss": 0.0327, "step": 8260 }, { "epoch": 2.406168169915624, "grad_norm": 0.16738849878311157, "learning_rate": 2.9371922304014093e-05, "loss": 0.0316, "step": 8270 }, { "epoch": 2.4090776840267676, "grad_norm": 0.16695638000965118, "learning_rate": 2.9369552850718557e-05, "loss": 0.0312, "step": 8280 }, { "epoch": 2.411987198137911, "grad_norm": 0.15322062373161316, "learning_rate": 2.9367179032332262e-05, "loss": 0.032, "step": 8290 }, { "epoch": 2.4148967122490546, "grad_norm": 0.20747016370296478, "learning_rate": 2.936480084957633e-05, "loss": 0.0336, "step": 8300 }, { "epoch": 2.417806226360198, "grad_norm": 0.14201144874095917, "learning_rate": 2.9362418303173174e-05, "loss": 0.0305, "step": 8310 }, { "epoch": 2.4207157404713415, "grad_norm": 0.13845960795879364, "learning_rate": 2.9360031393846558e-05, "loss": 0.0346, "step": 8320 }, { "epoch": 2.423625254582485, "grad_norm": 0.1346966028213501, "learning_rate": 2.9357640122321557e-05, "loss": 0.0316, "step": 8330 }, { "epoch": 2.426534768693628, "grad_norm": 0.14122803509235382, "learning_rate": 2.935524448932458e-05, "loss": 0.0324, "step": 8340 }, { "epoch": 2.4294442828047718, "grad_norm": 0.11656798422336578, "learning_rate": 2.9352844495583356e-05, "loss": 0.0337, "step": 8350 }, { "epoch": 2.432353796915915, "grad_norm": 0.16169032454490662, "learning_rate": 2.9350440141826938e-05, "loss": 0.0311, "step": 8360 }, { "epoch": 2.4352633110270583, "grad_norm": 0.1480330526828766, "learning_rate": 2.9348031428785705e-05, "loss": 0.0314, "step": 8370 }, { "epoch": 2.438172825138202, "grad_norm": 0.1501837819814682, "learning_rate": 2.9345618357191363e-05, "loss": 0.0316, "step": 8380 }, { "epoch": 2.4410823392493453, "grad_norm": 0.1852465271949768, "learning_rate": 2.934320092777694e-05, "loss": 0.0318, "step": 8390 }, { "epoch": 2.443991853360489, "grad_norm": 0.15626195073127747, "learning_rate": 2.9340779141276782e-05, "loss": 0.0325, "step": 8400 }, { "epoch": 2.4469013674716322, "grad_norm": 0.12046386301517487, "learning_rate": 2.933835299842657e-05, "loss": 0.032, "step": 8410 }, { "epoch": 2.4498108815827755, "grad_norm": 0.1801455020904541, "learning_rate": 2.93359224999633e-05, "loss": 0.0322, "step": 8420 }, { "epoch": 2.452720395693919, "grad_norm": 0.11043783277273178, "learning_rate": 2.933348764662529e-05, "loss": 0.0304, "step": 8430 }, { "epoch": 2.4556299098050625, "grad_norm": 0.15612240135669708, "learning_rate": 2.933104843915219e-05, "loss": 0.0312, "step": 8440 }, { "epoch": 2.458539423916206, "grad_norm": 0.16851073503494263, "learning_rate": 2.9328604878284963e-05, "loss": 0.031, "step": 8450 }, { "epoch": 2.4614489380273494, "grad_norm": 0.15544092655181885, "learning_rate": 2.9326156964765892e-05, "loss": 0.0316, "step": 8460 }, { "epoch": 2.4643584521384927, "grad_norm": 0.15462881326675415, "learning_rate": 2.9323704699338603e-05, "loss": 0.0295, "step": 8470 }, { "epoch": 2.4672679662496364, "grad_norm": 0.18113327026367188, "learning_rate": 2.932124808274802e-05, "loss": 0.0311, "step": 8480 }, { "epoch": 2.4701774803607797, "grad_norm": 0.12762995064258575, "learning_rate": 2.93187871157404e-05, "loss": 0.0308, "step": 8490 }, { "epoch": 2.4730869944719234, "grad_norm": 0.1414715200662613, "learning_rate": 2.931632179906332e-05, "loss": 0.0323, "step": 8500 }, { "epoch": 2.4759965085830666, "grad_norm": 0.14066524803638458, "learning_rate": 2.931385213346568e-05, "loss": 0.0345, "step": 8510 }, { "epoch": 2.47890602269421, "grad_norm": 0.1703260838985443, "learning_rate": 2.9311378119697694e-05, "loss": 0.0314, "step": 8520 }, { "epoch": 2.4818155368053536, "grad_norm": 0.1251901388168335, "learning_rate": 2.9308899758510903e-05, "loss": 0.0321, "step": 8530 }, { "epoch": 2.484725050916497, "grad_norm": 0.12031096965074539, "learning_rate": 2.9306417050658172e-05, "loss": 0.0306, "step": 8540 }, { "epoch": 2.4876345650276406, "grad_norm": 0.1343660056591034, "learning_rate": 2.9303929996893687e-05, "loss": 0.0316, "step": 8550 }, { "epoch": 2.490544079138784, "grad_norm": 0.11691556870937347, "learning_rate": 2.9301438597972932e-05, "loss": 0.0325, "step": 8560 }, { "epoch": 2.493453593249927, "grad_norm": 0.12919439375400543, "learning_rate": 2.9298942854652744e-05, "loss": 0.0329, "step": 8570 }, { "epoch": 2.496363107361071, "grad_norm": 0.12246531248092651, "learning_rate": 2.9296442767691257e-05, "loss": 0.0312, "step": 8580 }, { "epoch": 2.499272621472214, "grad_norm": 0.15629276633262634, "learning_rate": 2.9293938337847936e-05, "loss": 0.0309, "step": 8590 }, { "epoch": 2.5021821355833573, "grad_norm": 0.15894944965839386, "learning_rate": 2.9291429565883554e-05, "loss": 0.0315, "step": 8600 }, { "epoch": 2.505091649694501, "grad_norm": 0.12116479128599167, "learning_rate": 2.9288916452560214e-05, "loss": 0.0324, "step": 8610 }, { "epoch": 2.5080011638056443, "grad_norm": 0.14647002518177032, "learning_rate": 2.9286398998641334e-05, "loss": 0.0333, "step": 8620 }, { "epoch": 2.510910677916788, "grad_norm": 0.1668260544538498, "learning_rate": 2.9283877204891645e-05, "loss": 0.0321, "step": 8630 }, { "epoch": 2.5138201920279313, "grad_norm": 0.1632433980703354, "learning_rate": 2.9281351072077208e-05, "loss": 0.0294, "step": 8640 }, { "epoch": 2.516729706139075, "grad_norm": 0.12833356857299805, "learning_rate": 2.9278820600965393e-05, "loss": 0.0312, "step": 8650 }, { "epoch": 2.5196392202502182, "grad_norm": 0.12732158601284027, "learning_rate": 2.9276285792324887e-05, "loss": 0.0317, "step": 8660 }, { "epoch": 2.5225487343613615, "grad_norm": 0.15002202987670898, "learning_rate": 2.9273746646925703e-05, "loss": 0.0301, "step": 8670 }, { "epoch": 2.525458248472505, "grad_norm": 0.12578223645687103, "learning_rate": 2.9271203165539166e-05, "loss": 0.03, "step": 8680 }, { "epoch": 2.5283677625836485, "grad_norm": 0.16266442835330963, "learning_rate": 2.9268655348937913e-05, "loss": 0.0323, "step": 8690 }, { "epoch": 2.5312772766947917, "grad_norm": 0.12486927956342697, "learning_rate": 2.926610319789591e-05, "loss": 0.0328, "step": 8700 }, { "epoch": 2.5341867908059355, "grad_norm": 0.12862664461135864, "learning_rate": 2.926354671318843e-05, "loss": 0.0305, "step": 8710 }, { "epoch": 2.5370963049170787, "grad_norm": 0.13086695969104767, "learning_rate": 2.926098589559206e-05, "loss": 0.029, "step": 8720 }, { "epoch": 2.5400058190282224, "grad_norm": 0.14192815124988556, "learning_rate": 2.925842074588472e-05, "loss": 0.0295, "step": 8730 }, { "epoch": 2.5429153331393657, "grad_norm": 0.12796060740947723, "learning_rate": 2.925585126484563e-05, "loss": 0.0329, "step": 8740 }, { "epoch": 2.5458248472505094, "grad_norm": 0.13268138468265533, "learning_rate": 2.9253277453255326e-05, "loss": 0.031, "step": 8750 }, { "epoch": 2.5487343613616527, "grad_norm": 0.16064266860485077, "learning_rate": 2.925069931189567e-05, "loss": 0.0316, "step": 8760 }, { "epoch": 2.551643875472796, "grad_norm": 0.11417713016271591, "learning_rate": 2.924811684154983e-05, "loss": 0.0334, "step": 8770 }, { "epoch": 2.5545533895839396, "grad_norm": 0.13250337541103363, "learning_rate": 2.9245530043002293e-05, "loss": 0.0314, "step": 8780 }, { "epoch": 2.557462903695083, "grad_norm": 0.14566652476787567, "learning_rate": 2.9242938917038863e-05, "loss": 0.032, "step": 8790 }, { "epoch": 2.560372417806226, "grad_norm": 0.1269494593143463, "learning_rate": 2.9240343464446647e-05, "loss": 0.0334, "step": 8800 }, { "epoch": 2.56328193191737, "grad_norm": 0.15472553670406342, "learning_rate": 2.923774368601409e-05, "loss": 0.0309, "step": 8810 }, { "epoch": 2.566191446028513, "grad_norm": 0.12065412104129791, "learning_rate": 2.9235139582530918e-05, "loss": 0.0336, "step": 8820 }, { "epoch": 2.569100960139657, "grad_norm": 0.17062567174434662, "learning_rate": 2.9232531154788203e-05, "loss": 0.0339, "step": 8830 }, { "epoch": 2.5720104742508, "grad_norm": 0.1592220962047577, "learning_rate": 2.9229918403578313e-05, "loss": 0.0297, "step": 8840 }, { "epoch": 2.574919988361944, "grad_norm": 0.11564536392688751, "learning_rate": 2.922730132969493e-05, "loss": 0.0316, "step": 8850 }, { "epoch": 2.577829502473087, "grad_norm": 0.13951526582241058, "learning_rate": 2.922467993393305e-05, "loss": 0.0283, "step": 8860 }, { "epoch": 2.5807390165842303, "grad_norm": 0.1073501855134964, "learning_rate": 2.9222054217088985e-05, "loss": 0.0298, "step": 8870 }, { "epoch": 2.583648530695374, "grad_norm": 0.1426246613264084, "learning_rate": 2.921942417996037e-05, "loss": 0.0299, "step": 8880 }, { "epoch": 2.5865580448065173, "grad_norm": 0.12150508165359497, "learning_rate": 2.921678982334612e-05, "loss": 0.0309, "step": 8890 }, { "epoch": 2.5894675589176606, "grad_norm": 0.12546797096729279, "learning_rate": 2.92141511480465e-05, "loss": 0.0308, "step": 8900 }, { "epoch": 2.5923770730288043, "grad_norm": 0.12227417528629303, "learning_rate": 2.9211508154863066e-05, "loss": 0.0311, "step": 8910 }, { "epoch": 2.5952865871399475, "grad_norm": 0.14789937436580658, "learning_rate": 2.9208860844598684e-05, "loss": 0.035, "step": 8920 }, { "epoch": 2.598196101251091, "grad_norm": 0.13640856742858887, "learning_rate": 2.9206209218057543e-05, "loss": 0.0313, "step": 8930 }, { "epoch": 2.6011056153622345, "grad_norm": 0.10541826486587524, "learning_rate": 2.9203553276045135e-05, "loss": 0.0301, "step": 8940 }, { "epoch": 2.604015129473378, "grad_norm": 0.17734338343143463, "learning_rate": 2.9200893019368263e-05, "loss": 0.0295, "step": 8950 }, { "epoch": 2.6069246435845215, "grad_norm": 0.11170995980501175, "learning_rate": 2.9198228448835045e-05, "loss": 0.0311, "step": 8960 }, { "epoch": 2.6098341576956647, "grad_norm": 0.1468552201986313, "learning_rate": 2.9195559565254908e-05, "loss": 0.0283, "step": 8970 }, { "epoch": 2.6127436718068084, "grad_norm": 0.14143064618110657, "learning_rate": 2.919288636943858e-05, "loss": 0.031, "step": 8980 }, { "epoch": 2.6156531859179517, "grad_norm": 0.11491545289754868, "learning_rate": 2.9190208862198126e-05, "loss": 0.0316, "step": 8990 }, { "epoch": 2.618562700029095, "grad_norm": 0.15750983357429504, "learning_rate": 2.9187527044346883e-05, "loss": 0.0305, "step": 9000 }, { "epoch": 2.6214722141402387, "grad_norm": 0.13084937632083893, "learning_rate": 2.918484091669953e-05, "loss": 0.0306, "step": 9010 }, { "epoch": 2.624381728251382, "grad_norm": 0.13166417181491852, "learning_rate": 2.9182150480072026e-05, "loss": 0.0299, "step": 9020 }, { "epoch": 2.627291242362525, "grad_norm": 0.1139231026172638, "learning_rate": 2.917945573528167e-05, "loss": 0.0297, "step": 9030 }, { "epoch": 2.630200756473669, "grad_norm": 0.1031503826379776, "learning_rate": 2.9176756683147052e-05, "loss": 0.03, "step": 9040 }, { "epoch": 2.633110270584812, "grad_norm": 0.19221612811088562, "learning_rate": 2.9174053324488063e-05, "loss": 0.0315, "step": 9050 }, { "epoch": 2.636019784695956, "grad_norm": 0.136548712849617, "learning_rate": 2.917134566012592e-05, "loss": 0.0293, "step": 9060 }, { "epoch": 2.638929298807099, "grad_norm": 0.14730225503444672, "learning_rate": 2.916863369088314e-05, "loss": 0.031, "step": 9070 }, { "epoch": 2.641838812918243, "grad_norm": 0.11878341436386108, "learning_rate": 2.9165917417583544e-05, "loss": 0.0293, "step": 9080 }, { "epoch": 2.644748327029386, "grad_norm": 0.15415222942829132, "learning_rate": 2.916319684105227e-05, "loss": 0.0325, "step": 9090 }, { "epoch": 2.6476578411405294, "grad_norm": 0.14222325384616852, "learning_rate": 2.916047196211575e-05, "loss": 0.029, "step": 9100 }, { "epoch": 2.650567355251673, "grad_norm": 0.1337117701768875, "learning_rate": 2.915774278160173e-05, "loss": 0.0319, "step": 9110 }, { "epoch": 2.6534768693628163, "grad_norm": 0.12834376096725464, "learning_rate": 2.9155009300339268e-05, "loss": 0.0313, "step": 9120 }, { "epoch": 2.6563863834739596, "grad_norm": 0.1518581062555313, "learning_rate": 2.9152271519158724e-05, "loss": 0.0308, "step": 9130 }, { "epoch": 2.6592958975851033, "grad_norm": 0.16327142715454102, "learning_rate": 2.9149529438891764e-05, "loss": 0.0315, "step": 9140 }, { "epoch": 2.6622054116962466, "grad_norm": 0.12837600708007812, "learning_rate": 2.9146783060371352e-05, "loss": 0.0308, "step": 9150 }, { "epoch": 2.6651149258073903, "grad_norm": 0.1625407636165619, "learning_rate": 2.914403238443177e-05, "loss": 0.03, "step": 9160 }, { "epoch": 2.6680244399185336, "grad_norm": 0.13595837354660034, "learning_rate": 2.9141277411908607e-05, "loss": 0.0333, "step": 9170 }, { "epoch": 2.6709339540296773, "grad_norm": 0.17140863835811615, "learning_rate": 2.9138518143638744e-05, "loss": 0.0296, "step": 9180 }, { "epoch": 2.6738434681408205, "grad_norm": 0.13536491990089417, "learning_rate": 2.9135754580460373e-05, "loss": 0.0328, "step": 9190 }, { "epoch": 2.676752982251964, "grad_norm": 0.13154666125774384, "learning_rate": 2.9132986723212992e-05, "loss": 0.0322, "step": 9200 }, { "epoch": 2.6796624963631075, "grad_norm": 0.11681555211544037, "learning_rate": 2.9130214572737403e-05, "loss": 0.0318, "step": 9210 }, { "epoch": 2.6825720104742508, "grad_norm": 0.14743953943252563, "learning_rate": 2.9127438129875717e-05, "loss": 0.0309, "step": 9220 }, { "epoch": 2.685481524585394, "grad_norm": 0.16170328855514526, "learning_rate": 2.912465739547134e-05, "loss": 0.0315, "step": 9230 }, { "epoch": 2.6883910386965377, "grad_norm": 0.1185581386089325, "learning_rate": 2.9121872370368988e-05, "loss": 0.029, "step": 9240 }, { "epoch": 2.691300552807681, "grad_norm": 0.16426295042037964, "learning_rate": 2.911908305541467e-05, "loss": 0.0315, "step": 9250 }, { "epoch": 2.6942100669188247, "grad_norm": 0.16451504826545715, "learning_rate": 2.9116289451455717e-05, "loss": 0.0318, "step": 9260 }, { "epoch": 2.697119581029968, "grad_norm": 0.11905410140752792, "learning_rate": 2.911349155934075e-05, "loss": 0.0305, "step": 9270 }, { "epoch": 2.7000290951411117, "grad_norm": 0.13563641905784607, "learning_rate": 2.9110689379919687e-05, "loss": 0.0323, "step": 9280 }, { "epoch": 2.702938609252255, "grad_norm": 0.1242348924279213, "learning_rate": 2.9107882914043767e-05, "loss": 0.031, "step": 9290 }, { "epoch": 2.705848123363398, "grad_norm": 0.13210289180278778, "learning_rate": 2.910507216256551e-05, "loss": 0.03, "step": 9300 }, { "epoch": 2.708757637474542, "grad_norm": 0.1273929923772812, "learning_rate": 2.910225712633876e-05, "loss": 0.0309, "step": 9310 }, { "epoch": 2.711667151585685, "grad_norm": 0.1313595473766327, "learning_rate": 2.9099437806218637e-05, "loss": 0.0295, "step": 9320 }, { "epoch": 2.7145766656968284, "grad_norm": 0.15506014227867126, "learning_rate": 2.9096614203061584e-05, "loss": 0.0276, "step": 9330 }, { "epoch": 2.717486179807972, "grad_norm": 0.13925109803676605, "learning_rate": 2.9093786317725336e-05, "loss": 0.0294, "step": 9340 }, { "epoch": 2.7203956939191154, "grad_norm": 0.12995412945747375, "learning_rate": 2.909095415106893e-05, "loss": 0.0311, "step": 9350 }, { "epoch": 2.723305208030259, "grad_norm": 0.14978477358818054, "learning_rate": 2.9088117703952698e-05, "loss": 0.0305, "step": 9360 }, { "epoch": 2.7262147221414024, "grad_norm": 0.14907871186733246, "learning_rate": 2.908527697723829e-05, "loss": 0.0302, "step": 9370 }, { "epoch": 2.729124236252546, "grad_norm": 0.14872828125953674, "learning_rate": 2.9082431971788635e-05, "loss": 0.0306, "step": 9380 }, { "epoch": 2.7320337503636893, "grad_norm": 0.12671366333961487, "learning_rate": 2.9079582688467972e-05, "loss": 0.0283, "step": 9390 }, { "epoch": 2.7349432644748326, "grad_norm": 0.16726766526699066, "learning_rate": 2.9076729128141833e-05, "loss": 0.0307, "step": 9400 }, { "epoch": 2.7378527785859763, "grad_norm": 0.12228977680206299, "learning_rate": 2.9073871291677063e-05, "loss": 0.0293, "step": 9410 }, { "epoch": 2.7407622926971196, "grad_norm": 0.12649738788604736, "learning_rate": 2.9071009179941797e-05, "loss": 0.0318, "step": 9420 }, { "epoch": 2.743671806808263, "grad_norm": 0.11510005593299866, "learning_rate": 2.906814279380546e-05, "loss": 0.0312, "step": 9430 }, { "epoch": 2.7465813209194065, "grad_norm": 0.13456125557422638, "learning_rate": 2.9065272134138795e-05, "loss": 0.0291, "step": 9440 }, { "epoch": 2.74949083503055, "grad_norm": 0.1536651849746704, "learning_rate": 2.9062397201813823e-05, "loss": 0.0292, "step": 9450 }, { "epoch": 2.752400349141693, "grad_norm": 0.11861708015203476, "learning_rate": 2.905951799770388e-05, "loss": 0.0298, "step": 9460 }, { "epoch": 2.7553098632528368, "grad_norm": 0.09680566936731339, "learning_rate": 2.9056634522683586e-05, "loss": 0.0296, "step": 9470 }, { "epoch": 2.7582193773639805, "grad_norm": 0.13542033731937408, "learning_rate": 2.9053746777628872e-05, "loss": 0.0291, "step": 9480 }, { "epoch": 2.7611288914751237, "grad_norm": 0.13086313009262085, "learning_rate": 2.9050854763416952e-05, "loss": 0.0302, "step": 9490 }, { "epoch": 2.764038405586267, "grad_norm": 0.11816871911287308, "learning_rate": 2.9047958480926346e-05, "loss": 0.0298, "step": 9500 }, { "epoch": 2.7669479196974107, "grad_norm": 0.1468793749809265, "learning_rate": 2.9045057931036874e-05, "loss": 0.0278, "step": 9510 }, { "epoch": 2.769857433808554, "grad_norm": 0.11239022016525269, "learning_rate": 2.9042153114629636e-05, "loss": 0.0319, "step": 9520 }, { "epoch": 2.7727669479196972, "grad_norm": 0.18405501544475555, "learning_rate": 2.9039244032587043e-05, "loss": 0.0329, "step": 9530 }, { "epoch": 2.775676462030841, "grad_norm": 0.16124384105205536, "learning_rate": 2.90363306857928e-05, "loss": 0.0319, "step": 9540 }, { "epoch": 2.778585976141984, "grad_norm": 0.15229517221450806, "learning_rate": 2.90334130751319e-05, "loss": 0.0304, "step": 9550 }, { "epoch": 2.7814954902531275, "grad_norm": 0.09558228403329849, "learning_rate": 2.903049120149064e-05, "loss": 0.032, "step": 9560 }, { "epoch": 2.784405004364271, "grad_norm": 0.12323088198900223, "learning_rate": 2.9027565065756608e-05, "loss": 0.0316, "step": 9570 }, { "epoch": 2.7873145184754144, "grad_norm": 0.1205105260014534, "learning_rate": 2.9024634668818682e-05, "loss": 0.0341, "step": 9580 }, { "epoch": 2.790224032586558, "grad_norm": 0.12612877786159515, "learning_rate": 2.9021700011567045e-05, "loss": 0.0277, "step": 9590 }, { "epoch": 2.7931335466977014, "grad_norm": 0.13928359746932983, "learning_rate": 2.9018761094893166e-05, "loss": 0.0337, "step": 9600 }, { "epoch": 2.796043060808845, "grad_norm": 0.1780450940132141, "learning_rate": 2.9015817919689808e-05, "loss": 0.0278, "step": 9610 }, { "epoch": 2.7989525749199884, "grad_norm": 0.12098591029644012, "learning_rate": 2.9012870486851034e-05, "loss": 0.0306, "step": 9620 }, { "epoch": 2.8018620890311317, "grad_norm": 0.1437063366174698, "learning_rate": 2.9009918797272195e-05, "loss": 0.0336, "step": 9630 }, { "epoch": 2.8047716031422754, "grad_norm": 0.13368991017341614, "learning_rate": 2.9006962851849933e-05, "loss": 0.0306, "step": 9640 }, { "epoch": 2.8076811172534186, "grad_norm": 0.14518561959266663, "learning_rate": 2.9004002651482192e-05, "loss": 0.0311, "step": 9650 }, { "epoch": 2.810590631364562, "grad_norm": 0.14251528680324554, "learning_rate": 2.9001038197068198e-05, "loss": 0.0314, "step": 9660 }, { "epoch": 2.8135001454757056, "grad_norm": 0.1652197539806366, "learning_rate": 2.899806948950848e-05, "loss": 0.0292, "step": 9670 }, { "epoch": 2.816409659586849, "grad_norm": 0.216352179646492, "learning_rate": 2.8995096529704846e-05, "loss": 0.0292, "step": 9680 }, { "epoch": 2.8193191736979926, "grad_norm": 0.1418590396642685, "learning_rate": 2.8992119318560403e-05, "loss": 0.0298, "step": 9690 }, { "epoch": 2.822228687809136, "grad_norm": 0.15420220792293549, "learning_rate": 2.8989137856979555e-05, "loss": 0.032, "step": 9700 }, { "epoch": 2.8251382019202795, "grad_norm": 0.141637921333313, "learning_rate": 2.8986152145867983e-05, "loss": 0.0302, "step": 9710 }, { "epoch": 2.828047716031423, "grad_norm": 0.12551352381706238, "learning_rate": 2.8983162186132678e-05, "loss": 0.0304, "step": 9720 }, { "epoch": 2.830957230142566, "grad_norm": 0.12498537451028824, "learning_rate": 2.8980167978681905e-05, "loss": 0.0284, "step": 9730 }, { "epoch": 2.8338667442537098, "grad_norm": 0.15078409016132355, "learning_rate": 2.897716952442522e-05, "loss": 0.0299, "step": 9740 }, { "epoch": 2.836776258364853, "grad_norm": 0.15481024980545044, "learning_rate": 2.8974166824273482e-05, "loss": 0.0289, "step": 9750 }, { "epoch": 2.8396857724759963, "grad_norm": 0.14183157682418823, "learning_rate": 2.897115987913883e-05, "loss": 0.0322, "step": 9760 }, { "epoch": 2.84259528658714, "grad_norm": 0.12618082761764526, "learning_rate": 2.896814868993469e-05, "loss": 0.0317, "step": 9770 }, { "epoch": 2.8455048006982833, "grad_norm": 0.14413438737392426, "learning_rate": 2.896513325757579e-05, "loss": 0.0308, "step": 9780 }, { "epoch": 2.848414314809427, "grad_norm": 0.14696811139583588, "learning_rate": 2.896211358297813e-05, "loss": 0.0301, "step": 9790 }, { "epoch": 2.8513238289205702, "grad_norm": 0.17287464439868927, "learning_rate": 2.8959089667059014e-05, "loss": 0.0292, "step": 9800 }, { "epoch": 2.854233343031714, "grad_norm": 0.18278531730175018, "learning_rate": 2.8956061510737027e-05, "loss": 0.029, "step": 9810 }, { "epoch": 2.857142857142857, "grad_norm": 0.13076996803283691, "learning_rate": 2.8953029114932042e-05, "loss": 0.0307, "step": 9820 }, { "epoch": 2.8600523712540005, "grad_norm": 0.16651052236557007, "learning_rate": 2.8949992480565214e-05, "loss": 0.031, "step": 9830 }, { "epoch": 2.862961885365144, "grad_norm": 0.12112797796726227, "learning_rate": 2.894695160855901e-05, "loss": 0.0307, "step": 9840 }, { "epoch": 2.8658713994762874, "grad_norm": 0.11327257007360458, "learning_rate": 2.8943906499837146e-05, "loss": 0.0283, "step": 9850 }, { "epoch": 2.8687809135874307, "grad_norm": 0.12342819571495056, "learning_rate": 2.8940857155324655e-05, "loss": 0.0288, "step": 9860 }, { "epoch": 2.8716904276985744, "grad_norm": 0.11777912825345993, "learning_rate": 2.893780357594785e-05, "loss": 0.0297, "step": 9870 }, { "epoch": 2.8745999418097177, "grad_norm": 0.12374354898929596, "learning_rate": 2.8934745762634326e-05, "loss": 0.0285, "step": 9880 }, { "epoch": 2.8775094559208614, "grad_norm": 0.17270581424236298, "learning_rate": 2.8931683716312964e-05, "loss": 0.0313, "step": 9890 }, { "epoch": 2.8804189700320046, "grad_norm": 0.13099779188632965, "learning_rate": 2.8928617437913932e-05, "loss": 0.0303, "step": 9900 }, { "epoch": 2.8833284841431484, "grad_norm": 0.1262933760881424, "learning_rate": 2.8925546928368688e-05, "loss": 0.0286, "step": 9910 }, { "epoch": 2.8862379982542916, "grad_norm": 0.10170990973711014, "learning_rate": 2.8922472188609968e-05, "loss": 0.0292, "step": 9920 }, { "epoch": 2.889147512365435, "grad_norm": 0.10240564495325089, "learning_rate": 2.8919393219571805e-05, "loss": 0.03, "step": 9930 }, { "epoch": 2.8920570264765786, "grad_norm": 0.1455734670162201, "learning_rate": 2.8916310022189495e-05, "loss": 0.0291, "step": 9940 }, { "epoch": 2.894966540587722, "grad_norm": 0.11619999259710312, "learning_rate": 2.891322259739964e-05, "loss": 0.0302, "step": 9950 }, { "epoch": 2.897876054698865, "grad_norm": 0.10984021425247192, "learning_rate": 2.891013094614012e-05, "loss": 0.0315, "step": 9960 }, { "epoch": 2.900785568810009, "grad_norm": 0.16140763461589813, "learning_rate": 2.8907035069350093e-05, "loss": 0.0329, "step": 9970 }, { "epoch": 2.903695082921152, "grad_norm": 0.17907917499542236, "learning_rate": 2.8903934967970007e-05, "loss": 0.0307, "step": 9980 }, { "epoch": 2.9066045970322953, "grad_norm": 0.16508905589580536, "learning_rate": 2.8900830642941583e-05, "loss": 0.0293, "step": 9990 }, { "epoch": 2.909514111143439, "grad_norm": 0.14679917693138123, "learning_rate": 2.889772209520785e-05, "loss": 0.0269, "step": 10000 }, { "epoch": 2.9124236252545828, "grad_norm": 0.13242611289024353, "learning_rate": 2.8894609325713087e-05, "loss": 0.0279, "step": 10010 }, { "epoch": 2.915333139365726, "grad_norm": 0.09829951822757721, "learning_rate": 2.889149233540288e-05, "loss": 0.0288, "step": 10020 }, { "epoch": 2.9182426534768693, "grad_norm": 0.11715073883533478, "learning_rate": 2.8888371125224084e-05, "loss": 0.0275, "step": 10030 }, { "epoch": 2.921152167588013, "grad_norm": 0.10065282881259918, "learning_rate": 2.8885245696124835e-05, "loss": 0.0283, "step": 10040 }, { "epoch": 2.9240616816991563, "grad_norm": 0.12710267305374146, "learning_rate": 2.888211604905457e-05, "loss": 0.0284, "step": 10050 }, { "epoch": 2.9269711958102995, "grad_norm": 0.14784790575504303, "learning_rate": 2.8878982184963986e-05, "loss": 0.0308, "step": 10060 }, { "epoch": 2.9298807099214432, "grad_norm": 0.16373197734355927, "learning_rate": 2.8875844104805066e-05, "loss": 0.0286, "step": 10070 }, { "epoch": 2.9327902240325865, "grad_norm": 0.12714089453220367, "learning_rate": 2.887270180953107e-05, "loss": 0.029, "step": 10080 }, { "epoch": 2.9356997381437298, "grad_norm": 0.18822553753852844, "learning_rate": 2.886955530009656e-05, "loss": 0.0295, "step": 10090 }, { "epoch": 2.9386092522548735, "grad_norm": 0.13622671365737915, "learning_rate": 2.8866404577457352e-05, "loss": 0.03, "step": 10100 }, { "epoch": 2.9415187663660167, "grad_norm": 0.13835632801055908, "learning_rate": 2.8863249642570557e-05, "loss": 0.0314, "step": 10110 }, { "epoch": 2.9444282804771604, "grad_norm": 0.1521265059709549, "learning_rate": 2.8860090496394552e-05, "loss": 0.0311, "step": 10120 }, { "epoch": 2.9473377945883037, "grad_norm": 0.1288456916809082, "learning_rate": 2.8856927139889015e-05, "loss": 0.0306, "step": 10130 }, { "epoch": 2.9502473086994474, "grad_norm": 0.131577730178833, "learning_rate": 2.8853759574014878e-05, "loss": 0.0288, "step": 10140 }, { "epoch": 2.9531568228105907, "grad_norm": 0.12542694807052612, "learning_rate": 2.8850587799734372e-05, "loss": 0.0292, "step": 10150 }, { "epoch": 2.956066336921734, "grad_norm": 0.11534543335437775, "learning_rate": 2.8847411818011e-05, "loss": 0.0282, "step": 10160 }, { "epoch": 2.9589758510328776, "grad_norm": 0.14110003411769867, "learning_rate": 2.8844231629809534e-05, "loss": 0.0273, "step": 10170 }, { "epoch": 2.961885365144021, "grad_norm": 0.1625298112630844, "learning_rate": 2.8841047236096037e-05, "loss": 0.0289, "step": 10180 }, { "epoch": 2.964794879255164, "grad_norm": 0.12163512408733368, "learning_rate": 2.8837858637837838e-05, "loss": 0.0311, "step": 10190 }, { "epoch": 2.967704393366308, "grad_norm": 0.12158892303705215, "learning_rate": 2.8834665836003556e-05, "loss": 0.0296, "step": 10200 }, { "epoch": 2.970613907477451, "grad_norm": 0.12870322167873383, "learning_rate": 2.8831468831563075e-05, "loss": 0.0275, "step": 10210 }, { "epoch": 2.973523421588595, "grad_norm": 0.14774636924266815, "learning_rate": 2.882826762548756e-05, "loss": 0.0278, "step": 10220 }, { "epoch": 2.976432935699738, "grad_norm": 0.1374322474002838, "learning_rate": 2.8825062218749456e-05, "loss": 0.0289, "step": 10230 }, { "epoch": 2.979342449810882, "grad_norm": 0.12369602918624878, "learning_rate": 2.8821852612322477e-05, "loss": 0.0289, "step": 10240 }, { "epoch": 2.982251963922025, "grad_norm": 0.14562387764453888, "learning_rate": 2.8818638807181622e-05, "loss": 0.0296, "step": 10250 }, { "epoch": 2.9851614780331683, "grad_norm": 0.1630023568868637, "learning_rate": 2.8815420804303154e-05, "loss": 0.0296, "step": 10260 }, { "epoch": 2.988070992144312, "grad_norm": 0.12456399947404861, "learning_rate": 2.8812198604664627e-05, "loss": 0.0284, "step": 10270 }, { "epoch": 2.9909805062554553, "grad_norm": 0.1596379429101944, "learning_rate": 2.8808972209244848e-05, "loss": 0.0286, "step": 10280 }, { "epoch": 2.9938900203665986, "grad_norm": 0.12833824753761292, "learning_rate": 2.8805741619023922e-05, "loss": 0.0296, "step": 10290 }, { "epoch": 2.9967995344777423, "grad_norm": 0.13623017072677612, "learning_rate": 2.880250683498321e-05, "loss": 0.029, "step": 10300 }, { "epoch": 2.9997090485888855, "grad_norm": 0.15461139380931854, "learning_rate": 2.879926785810535e-05, "loss": 0.0294, "step": 10310 }, { "epoch": 3.0026185627000292, "grad_norm": 0.12167926132678986, "learning_rate": 2.8796024689374267e-05, "loss": 0.0297, "step": 10320 }, { "epoch": 3.0055280768111725, "grad_norm": 0.13306094706058502, "learning_rate": 2.8792777329775148e-05, "loss": 0.0278, "step": 10330 }, { "epoch": 3.0084375909223158, "grad_norm": 0.11599764227867126, "learning_rate": 2.878952578029445e-05, "loss": 0.0289, "step": 10340 }, { "epoch": 3.0113471050334595, "grad_norm": 0.1475159078836441, "learning_rate": 2.8786270041919915e-05, "loss": 0.0305, "step": 10350 }, { "epoch": 3.0142566191446027, "grad_norm": 0.14915412664413452, "learning_rate": 2.8783010115640546e-05, "loss": 0.0305, "step": 10360 }, { "epoch": 3.0171661332557465, "grad_norm": 0.11961477994918823, "learning_rate": 2.877974600244662e-05, "loss": 0.0291, "step": 10370 }, { "epoch": 3.0200756473668897, "grad_norm": 0.11666947603225708, "learning_rate": 2.8776477703329696e-05, "loss": 0.0271, "step": 10380 }, { "epoch": 3.022985161478033, "grad_norm": 0.12436296790838242, "learning_rate": 2.877320521928259e-05, "loss": 0.0276, "step": 10390 }, { "epoch": 3.0258946755891767, "grad_norm": 0.13979658484458923, "learning_rate": 2.87699285512994e-05, "loss": 0.027, "step": 10400 }, { "epoch": 3.02880418970032, "grad_norm": 0.13885529339313507, "learning_rate": 2.876664770037549e-05, "loss": 0.029, "step": 10410 }, { "epoch": 3.0317137038114637, "grad_norm": 0.12012404203414917, "learning_rate": 2.8763362667507495e-05, "loss": 0.0289, "step": 10420 }, { "epoch": 3.034623217922607, "grad_norm": 0.1317286491394043, "learning_rate": 2.8760073453693322e-05, "loss": 0.029, "step": 10430 }, { "epoch": 3.03753273203375, "grad_norm": 0.1589413285255432, "learning_rate": 2.8756780059932146e-05, "loss": 0.0279, "step": 10440 }, { "epoch": 3.040442246144894, "grad_norm": 0.1381264626979828, "learning_rate": 2.875348248722442e-05, "loss": 0.028, "step": 10450 }, { "epoch": 3.043351760256037, "grad_norm": 0.15459340810775757, "learning_rate": 2.8750180736571848e-05, "loss": 0.0312, "step": 10460 }, { "epoch": 3.046261274367181, "grad_norm": 0.13758540153503418, "learning_rate": 2.8746874808977424e-05, "loss": 0.0288, "step": 10470 }, { "epoch": 3.049170788478324, "grad_norm": 0.1193523108959198, "learning_rate": 2.87435647054454e-05, "loss": 0.0281, "step": 10480 }, { "epoch": 3.0520803025894674, "grad_norm": 0.16423770785331726, "learning_rate": 2.8740250426981292e-05, "loss": 0.0307, "step": 10490 }, { "epoch": 3.054989816700611, "grad_norm": 0.13016556203365326, "learning_rate": 2.8736931974591895e-05, "loss": 0.0275, "step": 10500 }, { "epoch": 3.0578993308117544, "grad_norm": 0.12015879899263382, "learning_rate": 2.8733609349285268e-05, "loss": 0.0288, "step": 10510 }, { "epoch": 3.060808844922898, "grad_norm": 0.1601400524377823, "learning_rate": 2.8730282552070734e-05, "loss": 0.0297, "step": 10520 }, { "epoch": 3.0637183590340413, "grad_norm": 0.13188587129116058, "learning_rate": 2.872695158395889e-05, "loss": 0.0281, "step": 10530 }, { "epoch": 3.0666278731451846, "grad_norm": 0.16666565835475922, "learning_rate": 2.872361644596159e-05, "loss": 0.0292, "step": 10540 }, { "epoch": 3.0695373872563283, "grad_norm": 0.12138117849826813, "learning_rate": 2.8720277139091972e-05, "loss": 0.0273, "step": 10550 }, { "epoch": 3.0724469013674716, "grad_norm": 0.13666509091854095, "learning_rate": 2.8716933664364417e-05, "loss": 0.0277, "step": 10560 }, { "epoch": 3.0753564154786153, "grad_norm": 0.11743183434009552, "learning_rate": 2.871358602279459e-05, "loss": 0.03, "step": 10570 }, { "epoch": 3.0782659295897585, "grad_norm": 0.11408600211143494, "learning_rate": 2.871023421539942e-05, "loss": 0.0281, "step": 10580 }, { "epoch": 3.081175443700902, "grad_norm": 0.13617490231990814, "learning_rate": 2.8706878243197085e-05, "loss": 0.0279, "step": 10590 }, { "epoch": 3.0840849578120455, "grad_norm": 0.13240878283977509, "learning_rate": 2.8703518107207052e-05, "loss": 0.0304, "step": 10600 }, { "epoch": 3.0869944719231888, "grad_norm": 0.1590077131986618, "learning_rate": 2.870015380845004e-05, "loss": 0.0277, "step": 10610 }, { "epoch": 3.0899039860343325, "grad_norm": 0.121014803647995, "learning_rate": 2.8696785347948035e-05, "loss": 0.029, "step": 10620 }, { "epoch": 3.0928135001454757, "grad_norm": 0.14649103581905365, "learning_rate": 2.8693412726724282e-05, "loss": 0.0281, "step": 10630 }, { "epoch": 3.095723014256619, "grad_norm": 0.15487143397331238, "learning_rate": 2.869003594580329e-05, "loss": 0.0272, "step": 10640 }, { "epoch": 3.0986325283677627, "grad_norm": 0.13007690012454987, "learning_rate": 2.868665500621085e-05, "loss": 0.0287, "step": 10650 }, { "epoch": 3.101542042478906, "grad_norm": 0.12130315601825714, "learning_rate": 2.8683269908973992e-05, "loss": 0.028, "step": 10660 }, { "epoch": 3.1044515565900497, "grad_norm": 0.16366417706012726, "learning_rate": 2.867988065512102e-05, "loss": 0.0292, "step": 10670 }, { "epoch": 3.107361070701193, "grad_norm": 0.17484162747859955, "learning_rate": 2.8676487245681507e-05, "loss": 0.029, "step": 10680 }, { "epoch": 3.110270584812336, "grad_norm": 0.12066026777029037, "learning_rate": 2.867308968168627e-05, "loss": 0.0289, "step": 10690 }, { "epoch": 3.11318009892348, "grad_norm": 0.12418416887521744, "learning_rate": 2.8669687964167404e-05, "loss": 0.0286, "step": 10700 }, { "epoch": 3.116089613034623, "grad_norm": 0.10211862623691559, "learning_rate": 2.8666282094158264e-05, "loss": 0.0296, "step": 10710 }, { "epoch": 3.1189991271457664, "grad_norm": 0.12534357607364655, "learning_rate": 2.866287207269346e-05, "loss": 0.0279, "step": 10720 }, { "epoch": 3.12190864125691, "grad_norm": 0.1166292279958725, "learning_rate": 2.8659457900808868e-05, "loss": 0.0257, "step": 10730 }, { "epoch": 3.1248181553680534, "grad_norm": 0.11348787695169449, "learning_rate": 2.8656039579541628e-05, "loss": 0.03, "step": 10740 }, { "epoch": 3.127727669479197, "grad_norm": 0.12063264846801758, "learning_rate": 2.865261710993012e-05, "loss": 0.0272, "step": 10750 }, { "epoch": 3.1306371835903404, "grad_norm": 0.15239644050598145, "learning_rate": 2.864919049301402e-05, "loss": 0.0303, "step": 10760 }, { "epoch": 3.133546697701484, "grad_norm": 0.14733676612377167, "learning_rate": 2.8645759729834227e-05, "loss": 0.0282, "step": 10770 }, { "epoch": 3.1364562118126273, "grad_norm": 0.13395655155181885, "learning_rate": 2.864232482143293e-05, "loss": 0.0286, "step": 10780 }, { "epoch": 3.1393657259237706, "grad_norm": 0.12341427057981491, "learning_rate": 2.8638885768853548e-05, "loss": 0.0281, "step": 10790 }, { "epoch": 3.1422752400349143, "grad_norm": 0.14871230721473694, "learning_rate": 2.8635442573140793e-05, "loss": 0.0285, "step": 10800 }, { "epoch": 3.1451847541460576, "grad_norm": 0.1301719695329666, "learning_rate": 2.8631995235340606e-05, "loss": 0.0286, "step": 10810 }, { "epoch": 3.148094268257201, "grad_norm": 0.11442147195339203, "learning_rate": 2.8628543756500192e-05, "loss": 0.0265, "step": 10820 }, { "epoch": 3.1510037823683446, "grad_norm": 0.12251599133014679, "learning_rate": 2.862508813766803e-05, "loss": 0.0286, "step": 10830 }, { "epoch": 3.153913296479488, "grad_norm": 0.12235421687364578, "learning_rate": 2.8621628379893837e-05, "loss": 0.0281, "step": 10840 }, { "epoch": 3.1568228105906315, "grad_norm": 0.11969001591205597, "learning_rate": 2.861816448422861e-05, "loss": 0.0272, "step": 10850 }, { "epoch": 3.159732324701775, "grad_norm": 0.13792578876018524, "learning_rate": 2.8614696451724573e-05, "loss": 0.0272, "step": 10860 }, { "epoch": 3.162641838812918, "grad_norm": 0.15461276471614838, "learning_rate": 2.861122428343523e-05, "loss": 0.0283, "step": 10870 }, { "epoch": 3.1655513529240618, "grad_norm": 0.13415735960006714, "learning_rate": 2.8607747980415333e-05, "loss": 0.0276, "step": 10880 }, { "epoch": 3.168460867035205, "grad_norm": 0.13249056041240692, "learning_rate": 2.860426754372089e-05, "loss": 0.0278, "step": 10890 }, { "epoch": 3.1713703811463487, "grad_norm": 0.13522744178771973, "learning_rate": 2.860078297440917e-05, "loss": 0.0269, "step": 10900 }, { "epoch": 3.174279895257492, "grad_norm": 0.14336545765399933, "learning_rate": 2.8597294273538687e-05, "loss": 0.0289, "step": 10910 }, { "epoch": 3.1771894093686353, "grad_norm": 0.15356457233428955, "learning_rate": 2.8593801442169223e-05, "loss": 0.0281, "step": 10920 }, { "epoch": 3.180098923479779, "grad_norm": 0.10890817642211914, "learning_rate": 2.85903044813618e-05, "loss": 0.0292, "step": 10930 }, { "epoch": 3.1830084375909222, "grad_norm": 0.12318896502256393, "learning_rate": 2.8586803392178715e-05, "loss": 0.0301, "step": 10940 }, { "epoch": 3.185917951702066, "grad_norm": 0.12389013171195984, "learning_rate": 2.858329817568349e-05, "loss": 0.028, "step": 10950 }, { "epoch": 3.188827465813209, "grad_norm": 0.12211494892835617, "learning_rate": 2.8579788832940925e-05, "loss": 0.0269, "step": 10960 }, { "epoch": 3.1917369799243525, "grad_norm": 0.12043635547161102, "learning_rate": 2.8576275365017073e-05, "loss": 0.0301, "step": 10970 }, { "epoch": 3.194646494035496, "grad_norm": 0.1670045703649521, "learning_rate": 2.857275777297922e-05, "loss": 0.0261, "step": 10980 }, { "epoch": 3.1975560081466394, "grad_norm": 0.13966591656208038, "learning_rate": 2.8569236057895924e-05, "loss": 0.0298, "step": 10990 }, { "epoch": 3.200465522257783, "grad_norm": 0.1167115718126297, "learning_rate": 2.8565710220836994e-05, "loss": 0.0296, "step": 11000 }, { "epoch": 3.2033750363689264, "grad_norm": 0.14605388045310974, "learning_rate": 2.8562180262873474e-05, "loss": 0.0281, "step": 11010 }, { "epoch": 3.2062845504800697, "grad_norm": 0.10636798292398453, "learning_rate": 2.8558646185077682e-05, "loss": 0.0275, "step": 11020 }, { "epoch": 3.2091940645912134, "grad_norm": 0.13726139068603516, "learning_rate": 2.8555107988523177e-05, "loss": 0.0261, "step": 11030 }, { "epoch": 3.2121035787023566, "grad_norm": 0.1576090157032013, "learning_rate": 2.8551565674284762e-05, "loss": 0.0272, "step": 11040 }, { "epoch": 3.2150130928135003, "grad_norm": 0.11097870767116547, "learning_rate": 2.8548019243438503e-05, "loss": 0.0267, "step": 11050 }, { "epoch": 3.2179226069246436, "grad_norm": 0.12444623559713364, "learning_rate": 2.8544468697061715e-05, "loss": 0.0278, "step": 11060 }, { "epoch": 3.220832121035787, "grad_norm": 0.11397341638803482, "learning_rate": 2.854091403623296e-05, "loss": 0.0279, "step": 11070 }, { "epoch": 3.2237416351469306, "grad_norm": 0.12060768902301788, "learning_rate": 2.8537355262032047e-05, "loss": 0.031, "step": 11080 }, { "epoch": 3.226651149258074, "grad_norm": 0.13762804865837097, "learning_rate": 2.853379237554004e-05, "loss": 0.0284, "step": 11090 }, { "epoch": 3.2295606633692175, "grad_norm": 0.10569200664758682, "learning_rate": 2.853022537783924e-05, "loss": 0.0265, "step": 11100 }, { "epoch": 3.232470177480361, "grad_norm": 0.12382914870977402, "learning_rate": 2.8526654270013223e-05, "loss": 0.0277, "step": 11110 }, { "epoch": 3.235379691591504, "grad_norm": 0.0867050290107727, "learning_rate": 2.8523079053146785e-05, "loss": 0.0262, "step": 11120 }, { "epoch": 3.2382892057026478, "grad_norm": 0.12963789701461792, "learning_rate": 2.851949972832599e-05, "loss": 0.0265, "step": 11130 }, { "epoch": 3.241198719813791, "grad_norm": 0.14934073388576508, "learning_rate": 2.851591629663814e-05, "loss": 0.028, "step": 11140 }, { "epoch": 3.2441082339249343, "grad_norm": 0.118750661611557, "learning_rate": 2.8512328759171783e-05, "loss": 0.0295, "step": 11150 }, { "epoch": 3.247017748036078, "grad_norm": 0.11710761487483978, "learning_rate": 2.8508737117016722e-05, "loss": 0.0268, "step": 11160 }, { "epoch": 3.2499272621472213, "grad_norm": 0.10127806663513184, "learning_rate": 2.8505141371264005e-05, "loss": 0.0292, "step": 11170 }, { "epoch": 3.252836776258365, "grad_norm": 0.12428110837936401, "learning_rate": 2.8501541523005918e-05, "loss": 0.0293, "step": 11180 }, { "epoch": 3.2557462903695082, "grad_norm": 0.15430067479610443, "learning_rate": 2.8497937573336007e-05, "loss": 0.0283, "step": 11190 }, { "epoch": 3.258655804480652, "grad_norm": 0.16434088349342346, "learning_rate": 2.8494329523349053e-05, "loss": 0.0289, "step": 11200 }, { "epoch": 3.261565318591795, "grad_norm": 0.12495582550764084, "learning_rate": 2.8490717374141088e-05, "loss": 0.0302, "step": 11210 }, { "epoch": 3.2644748327029385, "grad_norm": 0.12893100082874298, "learning_rate": 2.8487101126809384e-05, "loss": 0.0275, "step": 11220 }, { "epoch": 3.267384346814082, "grad_norm": 0.13258342444896698, "learning_rate": 2.8483480782452463e-05, "loss": 0.0271, "step": 11230 }, { "epoch": 3.2702938609252254, "grad_norm": 0.13396413624286652, "learning_rate": 2.8479856342170097e-05, "loss": 0.0274, "step": 11240 }, { "epoch": 3.2732033750363687, "grad_norm": 0.1298154592514038, "learning_rate": 2.8476227807063283e-05, "loss": 0.0267, "step": 11250 }, { "epoch": 3.2761128891475124, "grad_norm": 0.12277962267398834, "learning_rate": 2.8472595178234284e-05, "loss": 0.029, "step": 11260 }, { "epoch": 3.2790224032586557, "grad_norm": 0.13881248235702515, "learning_rate": 2.846895845678659e-05, "loss": 0.0275, "step": 11270 }, { "epoch": 3.2819319173697994, "grad_norm": 0.13396276533603668, "learning_rate": 2.8465317643824945e-05, "loss": 0.0273, "step": 11280 }, { "epoch": 3.2848414314809427, "grad_norm": 0.1215229406952858, "learning_rate": 2.8461672740455334e-05, "loss": 0.0285, "step": 11290 }, { "epoch": 3.2877509455920864, "grad_norm": 0.15068842470645905, "learning_rate": 2.8458023747784976e-05, "loss": 0.0292, "step": 11300 }, { "epoch": 3.2906604597032296, "grad_norm": 0.11691996455192566, "learning_rate": 2.8454370666922343e-05, "loss": 0.0284, "step": 11310 }, { "epoch": 3.293569973814373, "grad_norm": 0.1727527678012848, "learning_rate": 2.8450713498977145e-05, "loss": 0.029, "step": 11320 }, { "epoch": 3.2964794879255166, "grad_norm": 0.1544932723045349, "learning_rate": 2.844705224506033e-05, "loss": 0.0283, "step": 11330 }, { "epoch": 3.29938900203666, "grad_norm": 0.14729885756969452, "learning_rate": 2.844338690628409e-05, "loss": 0.0292, "step": 11340 }, { "epoch": 3.302298516147803, "grad_norm": 0.17118807137012482, "learning_rate": 2.8439717483761863e-05, "loss": 0.0291, "step": 11350 }, { "epoch": 3.305208030258947, "grad_norm": 0.11290440708398819, "learning_rate": 2.843604397860832e-05, "loss": 0.0305, "step": 11360 }, { "epoch": 3.30811754437009, "grad_norm": 0.13652624189853668, "learning_rate": 2.8432366391939373e-05, "loss": 0.0286, "step": 11370 }, { "epoch": 3.311027058481234, "grad_norm": 0.1257334053516388, "learning_rate": 2.8428684724872176e-05, "loss": 0.0286, "step": 11380 }, { "epoch": 3.313936572592377, "grad_norm": 0.13800649344921112, "learning_rate": 2.8424998978525123e-05, "loss": 0.0276, "step": 11390 }, { "epoch": 3.3168460867035203, "grad_norm": 0.11345656216144562, "learning_rate": 2.842130915401785e-05, "loss": 0.0266, "step": 11400 }, { "epoch": 3.319755600814664, "grad_norm": 0.13930955529212952, "learning_rate": 2.8417615252471226e-05, "loss": 0.027, "step": 11410 }, { "epoch": 3.3226651149258073, "grad_norm": 0.12716282904148102, "learning_rate": 2.841391727500735e-05, "loss": 0.025, "step": 11420 }, { "epoch": 3.325574629036951, "grad_norm": 0.10190656036138535, "learning_rate": 2.8410215222749593e-05, "loss": 0.0268, "step": 11430 }, { "epoch": 3.3284841431480943, "grad_norm": 0.12910278141498566, "learning_rate": 2.840650909682252e-05, "loss": 0.0285, "step": 11440 }, { "epoch": 3.3313936572592375, "grad_norm": 0.11500681936740875, "learning_rate": 2.8402798898351963e-05, "loss": 0.0273, "step": 11450 }, { "epoch": 3.3343031713703812, "grad_norm": 0.14396315813064575, "learning_rate": 2.8399084628464986e-05, "loss": 0.0278, "step": 11460 }, { "epoch": 3.3372126854815245, "grad_norm": 0.11684903502464294, "learning_rate": 2.8395366288289876e-05, "loss": 0.0275, "step": 11470 }, { "epoch": 3.340122199592668, "grad_norm": 0.15330836176872253, "learning_rate": 2.839164387895617e-05, "loss": 0.0255, "step": 11480 }, { "epoch": 3.3430317137038115, "grad_norm": 0.13458935916423798, "learning_rate": 2.8387917401594644e-05, "loss": 0.0271, "step": 11490 }, { "epoch": 3.3459412278149547, "grad_norm": 0.15964213013648987, "learning_rate": 2.8384186857337296e-05, "loss": 0.0285, "step": 11500 }, { "epoch": 3.3488507419260984, "grad_norm": 0.1029202863574028, "learning_rate": 2.838045224731737e-05, "loss": 0.0278, "step": 11510 }, { "epoch": 3.3517602560372417, "grad_norm": 0.12762852013111115, "learning_rate": 2.8376713572669348e-05, "loss": 0.028, "step": 11520 }, { "epoch": 3.3546697701483854, "grad_norm": 0.11296931654214859, "learning_rate": 2.8372970834528924e-05, "loss": 0.0296, "step": 11530 }, { "epoch": 3.3575792842595287, "grad_norm": 0.13491900265216827, "learning_rate": 2.836922403403306e-05, "loss": 0.0261, "step": 11540 }, { "epoch": 3.360488798370672, "grad_norm": 0.19459140300750732, "learning_rate": 2.8365473172319926e-05, "loss": 0.0289, "step": 11550 }, { "epoch": 3.3633983124818156, "grad_norm": 0.16170580685138702, "learning_rate": 2.8361718250528936e-05, "loss": 0.0311, "step": 11560 }, { "epoch": 3.366307826592959, "grad_norm": 0.12149515748023987, "learning_rate": 2.835795926980074e-05, "loss": 0.0258, "step": 11570 }, { "epoch": 3.369217340704102, "grad_norm": 0.13237299025058746, "learning_rate": 2.8354196231277213e-05, "loss": 0.0295, "step": 11580 }, { "epoch": 3.372126854815246, "grad_norm": 0.12777532637119293, "learning_rate": 2.835042913610147e-05, "loss": 0.0277, "step": 11590 }, { "epoch": 3.375036368926389, "grad_norm": 0.130171000957489, "learning_rate": 2.8346657985417846e-05, "loss": 0.0272, "step": 11600 }, { "epoch": 3.377945883037533, "grad_norm": 0.13349699974060059, "learning_rate": 2.8342882780371932e-05, "loss": 0.0263, "step": 11610 }, { "epoch": 3.380855397148676, "grad_norm": 0.12418384850025177, "learning_rate": 2.833910352211053e-05, "loss": 0.0267, "step": 11620 }, { "epoch": 3.38376491125982, "grad_norm": 0.13155825436115265, "learning_rate": 2.8335320211781667e-05, "loss": 0.0287, "step": 11630 }, { "epoch": 3.386674425370963, "grad_norm": 0.11999624222517014, "learning_rate": 2.8331532850534628e-05, "loss": 0.0274, "step": 11640 }, { "epoch": 3.3895839394821063, "grad_norm": 0.13655851781368256, "learning_rate": 2.8327741439519906e-05, "loss": 0.0281, "step": 11650 }, { "epoch": 3.39249345359325, "grad_norm": 0.12974338233470917, "learning_rate": 2.8323945979889233e-05, "loss": 0.0277, "step": 11660 }, { "epoch": 3.3954029677043933, "grad_norm": 0.10902276635169983, "learning_rate": 2.8320146472795572e-05, "loss": 0.0281, "step": 11670 }, { "epoch": 3.3983124818155366, "grad_norm": 0.1244804859161377, "learning_rate": 2.831634291939311e-05, "loss": 0.0274, "step": 11680 }, { "epoch": 3.4012219959266803, "grad_norm": 0.10782346874475479, "learning_rate": 2.8312535320837266e-05, "loss": 0.0289, "step": 11690 }, { "epoch": 3.4041315100378235, "grad_norm": 0.12305128574371338, "learning_rate": 2.8308723678284687e-05, "loss": 0.0272, "step": 11700 }, { "epoch": 3.4070410241489673, "grad_norm": 0.12572233378887177, "learning_rate": 2.8304907992893254e-05, "loss": 0.0281, "step": 11710 }, { "epoch": 3.4099505382601105, "grad_norm": 0.1368945688009262, "learning_rate": 2.8301088265822066e-05, "loss": 0.0273, "step": 11720 }, { "epoch": 3.4128600523712542, "grad_norm": 0.1048337072134018, "learning_rate": 2.829726449823146e-05, "loss": 0.0259, "step": 11730 }, { "epoch": 3.4157695664823975, "grad_norm": 0.12540952861309052, "learning_rate": 2.8293436691282993e-05, "loss": 0.03, "step": 11740 }, { "epoch": 3.4186790805935408, "grad_norm": 0.09521406888961792, "learning_rate": 2.828960484613945e-05, "loss": 0.0267, "step": 11750 }, { "epoch": 3.4215885947046845, "grad_norm": 0.13416974246501923, "learning_rate": 2.828576896396485e-05, "loss": 0.0278, "step": 11760 }, { "epoch": 3.4244981088158277, "grad_norm": 0.11029139161109924, "learning_rate": 2.828192904592443e-05, "loss": 0.0273, "step": 11770 }, { "epoch": 3.427407622926971, "grad_norm": 0.13870088756084442, "learning_rate": 2.8278085093184648e-05, "loss": 0.0303, "step": 11780 }, { "epoch": 3.4303171370381147, "grad_norm": 0.11576035618782043, "learning_rate": 2.827423710691321e-05, "loss": 0.0275, "step": 11790 }, { "epoch": 3.433226651149258, "grad_norm": 0.12159353494644165, "learning_rate": 2.827038508827902e-05, "loss": 0.0288, "step": 11800 }, { "epoch": 3.4361361652604017, "grad_norm": 0.1256970465183258, "learning_rate": 2.8266529038452223e-05, "loss": 0.0281, "step": 11810 }, { "epoch": 3.439045679371545, "grad_norm": 0.10995139181613922, "learning_rate": 2.826266895860419e-05, "loss": 0.0255, "step": 11820 }, { "epoch": 3.4419551934826886, "grad_norm": 0.11078328639268875, "learning_rate": 2.8258804849907508e-05, "loss": 0.0275, "step": 11830 }, { "epoch": 3.444864707593832, "grad_norm": 0.133528470993042, "learning_rate": 2.825493671353599e-05, "loss": 0.0281, "step": 11840 }, { "epoch": 3.447774221704975, "grad_norm": 0.1149422824382782, "learning_rate": 2.8251064550664673e-05, "loss": 0.0255, "step": 11850 }, { "epoch": 3.450683735816119, "grad_norm": 0.11697466671466827, "learning_rate": 2.8247188362469823e-05, "loss": 0.0286, "step": 11860 }, { "epoch": 3.453593249927262, "grad_norm": 0.12574753165245056, "learning_rate": 2.824330815012892e-05, "loss": 0.0261, "step": 11870 }, { "epoch": 3.4565027640384054, "grad_norm": 0.11880913376808167, "learning_rate": 2.8239423914820668e-05, "loss": 0.0268, "step": 11880 }, { "epoch": 3.459412278149549, "grad_norm": 0.1575860232114792, "learning_rate": 2.8235535657724997e-05, "loss": 0.0279, "step": 11890 }, { "epoch": 3.4623217922606924, "grad_norm": 0.11217334866523743, "learning_rate": 2.823164338002306e-05, "loss": 0.0262, "step": 11900 }, { "epoch": 3.465231306371836, "grad_norm": 0.1231905072927475, "learning_rate": 2.8227747082897223e-05, "loss": 0.0281, "step": 11910 }, { "epoch": 3.4681408204829793, "grad_norm": 0.18468314409255981, "learning_rate": 2.8223846767531084e-05, "loss": 0.0286, "step": 11920 }, { "epoch": 3.4710503345941226, "grad_norm": 0.14336468279361725, "learning_rate": 2.821994243510945e-05, "loss": 0.029, "step": 11930 }, { "epoch": 3.4739598487052663, "grad_norm": 0.14995542168617249, "learning_rate": 2.821603408681835e-05, "loss": 0.0261, "step": 11940 }, { "epoch": 3.4768693628164096, "grad_norm": 0.11554647237062454, "learning_rate": 2.8212121723845054e-05, "loss": 0.0273, "step": 11950 }, { "epoch": 3.4797788769275533, "grad_norm": 0.11874719709157944, "learning_rate": 2.820820534737802e-05, "loss": 0.026, "step": 11960 }, { "epoch": 3.4826883910386965, "grad_norm": 0.11162230372428894, "learning_rate": 2.8204284958606946e-05, "loss": 0.0272, "step": 11970 }, { "epoch": 3.48559790514984, "grad_norm": 0.15260232985019684, "learning_rate": 2.8200360558722738e-05, "loss": 0.0291, "step": 11980 }, { "epoch": 3.4885074192609835, "grad_norm": 0.10330313444137573, "learning_rate": 2.819643214891753e-05, "loss": 0.0298, "step": 11990 }, { "epoch": 3.4914169333721268, "grad_norm": 0.1137683093547821, "learning_rate": 2.8192499730384668e-05, "loss": 0.026, "step": 12000 }, { "epoch": 3.4943264474832705, "grad_norm": 0.12400386482477188, "learning_rate": 2.818856330431871e-05, "loss": 0.0283, "step": 12010 }, { "epoch": 3.4972359615944137, "grad_norm": 0.11482469737529755, "learning_rate": 2.8184622871915446e-05, "loss": 0.0261, "step": 12020 }, { "epoch": 3.500145475705557, "grad_norm": 0.11155331879854202, "learning_rate": 2.8180678434371874e-05, "loss": 0.0253, "step": 12030 }, { "epoch": 3.5030549898167007, "grad_norm": 0.1156400591135025, "learning_rate": 2.817672999288621e-05, "loss": 0.0272, "step": 12040 }, { "epoch": 3.505964503927844, "grad_norm": 0.14951151609420776, "learning_rate": 2.8172777548657886e-05, "loss": 0.0264, "step": 12050 }, { "epoch": 3.5088740180389877, "grad_norm": 0.14436353743076324, "learning_rate": 2.8168821102887545e-05, "loss": 0.0265, "step": 12060 }, { "epoch": 3.511783532150131, "grad_norm": 0.11616861820220947, "learning_rate": 2.816486065677706e-05, "loss": 0.0261, "step": 12070 }, { "epoch": 3.514693046261274, "grad_norm": 0.12046486884355545, "learning_rate": 2.8160896211529506e-05, "loss": 0.0264, "step": 12080 }, { "epoch": 3.517602560372418, "grad_norm": 0.10865577310323715, "learning_rate": 2.8156927768349175e-05, "loss": 0.0264, "step": 12090 }, { "epoch": 3.520512074483561, "grad_norm": 0.11852222681045532, "learning_rate": 2.815295532844158e-05, "loss": 0.0273, "step": 12100 }, { "epoch": 3.5234215885947044, "grad_norm": 0.14301274716854095, "learning_rate": 2.8148978893013434e-05, "loss": 0.0266, "step": 12110 }, { "epoch": 3.526331102705848, "grad_norm": 0.12682390213012695, "learning_rate": 2.814499846327268e-05, "loss": 0.0254, "step": 12120 }, { "epoch": 3.5292406168169914, "grad_norm": 0.11050483584403992, "learning_rate": 2.8141014040428468e-05, "loss": 0.0261, "step": 12130 }, { "epoch": 3.532150130928135, "grad_norm": 0.12568919360637665, "learning_rate": 2.8137025625691163e-05, "loss": 0.0256, "step": 12140 }, { "epoch": 3.5350596450392784, "grad_norm": 0.12422506511211395, "learning_rate": 2.813303322027233e-05, "loss": 0.0278, "step": 12150 }, { "epoch": 3.537969159150422, "grad_norm": 0.12206327170133591, "learning_rate": 2.8129036825384763e-05, "loss": 0.0288, "step": 12160 }, { "epoch": 3.5408786732615654, "grad_norm": 0.12994138896465302, "learning_rate": 2.8125036442242466e-05, "loss": 0.0257, "step": 12170 }, { "epoch": 3.5437881873727086, "grad_norm": 0.11494376510381699, "learning_rate": 2.8121032072060635e-05, "loss": 0.0281, "step": 12180 }, { "epoch": 3.5466977014838523, "grad_norm": 0.11661399900913239, "learning_rate": 2.811702371605571e-05, "loss": 0.0253, "step": 12190 }, { "epoch": 3.5496072155949956, "grad_norm": 0.11625202745199203, "learning_rate": 2.8113011375445308e-05, "loss": 0.027, "step": 12200 }, { "epoch": 3.552516729706139, "grad_norm": 0.12599530816078186, "learning_rate": 2.8108995051448284e-05, "loss": 0.0296, "step": 12210 }, { "epoch": 3.5554262438172826, "grad_norm": 0.11620306968688965, "learning_rate": 2.8104974745284685e-05, "loss": 0.0284, "step": 12220 }, { "epoch": 3.558335757928426, "grad_norm": 0.15182681381702423, "learning_rate": 2.8100950458175775e-05, "loss": 0.0288, "step": 12230 }, { "epoch": 3.5612452720395695, "grad_norm": 0.1296600103378296, "learning_rate": 2.8096922191344027e-05, "loss": 0.0257, "step": 12240 }, { "epoch": 3.564154786150713, "grad_norm": 0.12585945427417755, "learning_rate": 2.809288994601312e-05, "loss": 0.0264, "step": 12250 }, { "epoch": 3.5670643002618565, "grad_norm": 0.12512606382369995, "learning_rate": 2.8088853723407946e-05, "loss": 0.0256, "step": 12260 }, { "epoch": 3.5699738143729998, "grad_norm": 0.1273498386144638, "learning_rate": 2.8084813524754602e-05, "loss": 0.0257, "step": 12270 }, { "epoch": 3.572883328484143, "grad_norm": 0.11875646561384201, "learning_rate": 2.8080769351280395e-05, "loss": 0.0265, "step": 12280 }, { "epoch": 3.5757928425952867, "grad_norm": 0.08654177188873291, "learning_rate": 2.8076721204213835e-05, "loss": 0.0269, "step": 12290 }, { "epoch": 3.57870235670643, "grad_norm": 0.1258412003517151, "learning_rate": 2.8072669084784642e-05, "loss": 0.0266, "step": 12300 }, { "epoch": 3.5816118708175733, "grad_norm": 0.12934604287147522, "learning_rate": 2.8068612994223747e-05, "loss": 0.0257, "step": 12310 }, { "epoch": 3.584521384928717, "grad_norm": 0.13409367203712463, "learning_rate": 2.8064552933763284e-05, "loss": 0.0256, "step": 12320 }, { "epoch": 3.5874308990398602, "grad_norm": 0.115351103246212, "learning_rate": 2.8060488904636586e-05, "loss": 0.025, "step": 12330 }, { "epoch": 3.590340413151004, "grad_norm": 0.10890772938728333, "learning_rate": 2.80564209080782e-05, "loss": 0.0259, "step": 12340 }, { "epoch": 3.593249927262147, "grad_norm": 0.12439223378896713, "learning_rate": 2.8052348945323877e-05, "loss": 0.0275, "step": 12350 }, { "epoch": 3.596159441373291, "grad_norm": 0.10644928365945816, "learning_rate": 2.8048273017610574e-05, "loss": 0.028, "step": 12360 }, { "epoch": 3.599068955484434, "grad_norm": 0.11788830161094666, "learning_rate": 2.804419312617645e-05, "loss": 0.0251, "step": 12370 }, { "epoch": 3.6019784695955774, "grad_norm": 0.14131319522857666, "learning_rate": 2.8040109272260862e-05, "loss": 0.026, "step": 12380 }, { "epoch": 3.604887983706721, "grad_norm": 0.09557460248470306, "learning_rate": 2.8036021457104384e-05, "loss": 0.0264, "step": 12390 }, { "epoch": 3.6077974978178644, "grad_norm": 0.13270138204097748, "learning_rate": 2.803192968194878e-05, "loss": 0.0273, "step": 12400 }, { "epoch": 3.6107070119290077, "grad_norm": 0.10490912199020386, "learning_rate": 2.8027833948037032e-05, "loss": 0.0262, "step": 12410 }, { "epoch": 3.6136165260401514, "grad_norm": 0.13148964941501617, "learning_rate": 2.802373425661331e-05, "loss": 0.0257, "step": 12420 }, { "epoch": 3.6165260401512946, "grad_norm": 0.12241486459970474, "learning_rate": 2.8019630608922988e-05, "loss": 0.0279, "step": 12430 }, { "epoch": 3.619435554262438, "grad_norm": 0.16500583291053772, "learning_rate": 2.801552300621265e-05, "loss": 0.028, "step": 12440 }, { "epoch": 3.6223450683735816, "grad_norm": 0.130830317735672, "learning_rate": 2.8011411449730084e-05, "loss": 0.0265, "step": 12450 }, { "epoch": 3.6252545824847253, "grad_norm": 0.12825210392475128, "learning_rate": 2.800729594072426e-05, "loss": 0.0282, "step": 12460 }, { "epoch": 3.6281640965958686, "grad_norm": 0.1227913498878479, "learning_rate": 2.8003176480445373e-05, "loss": 0.0285, "step": 12470 }, { "epoch": 3.631073610707012, "grad_norm": 0.11301539838314056, "learning_rate": 2.7999053070144793e-05, "loss": 0.0272, "step": 12480 }, { "epoch": 3.6339831248181556, "grad_norm": 0.13646025955677032, "learning_rate": 2.7994925711075117e-05, "loss": 0.0278, "step": 12490 }, { "epoch": 3.636892638929299, "grad_norm": 0.10524403303861618, "learning_rate": 2.7990794404490118e-05, "loss": 0.0263, "step": 12500 }, { "epoch": 3.639802153040442, "grad_norm": 0.13608072698116302, "learning_rate": 2.798665915164478e-05, "loss": 0.0275, "step": 12510 }, { "epoch": 3.642711667151586, "grad_norm": 0.10926099121570587, "learning_rate": 2.7982519953795284e-05, "loss": 0.0269, "step": 12520 }, { "epoch": 3.645621181262729, "grad_norm": 0.09832017123699188, "learning_rate": 2.7978376812199013e-05, "loss": 0.0266, "step": 12530 }, { "epoch": 3.6485306953738723, "grad_norm": 0.13045886158943176, "learning_rate": 2.797422972811454e-05, "loss": 0.0269, "step": 12540 }, { "epoch": 3.651440209485016, "grad_norm": 0.12867344915866852, "learning_rate": 2.7970078702801635e-05, "loss": 0.0255, "step": 12550 }, { "epoch": 3.6543497235961593, "grad_norm": 0.12875770032405853, "learning_rate": 2.7965923737521283e-05, "loss": 0.0278, "step": 12560 }, { "epoch": 3.657259237707303, "grad_norm": 0.14519724249839783, "learning_rate": 2.796176483353564e-05, "loss": 0.0272, "step": 12570 }, { "epoch": 3.6601687518184463, "grad_norm": 0.16276080906391144, "learning_rate": 2.795760199210808e-05, "loss": 0.027, "step": 12580 }, { "epoch": 3.66307826592959, "grad_norm": 0.1294969916343689, "learning_rate": 2.7953435214503157e-05, "loss": 0.0267, "step": 12590 }, { "epoch": 3.6659877800407332, "grad_norm": 0.13526184856891632, "learning_rate": 2.7949264501986634e-05, "loss": 0.0276, "step": 12600 }, { "epoch": 3.6688972941518765, "grad_norm": 0.1408282071352005, "learning_rate": 2.794508985582546e-05, "loss": 0.0275, "step": 12610 }, { "epoch": 3.67180680826302, "grad_norm": 0.1426941603422165, "learning_rate": 2.7940911277287777e-05, "loss": 0.0268, "step": 12620 }, { "epoch": 3.6747163223741635, "grad_norm": 0.1492464542388916, "learning_rate": 2.793672876764294e-05, "loss": 0.0277, "step": 12630 }, { "epoch": 3.6776258364853067, "grad_norm": 0.13133876025676727, "learning_rate": 2.7932542328161478e-05, "loss": 0.0253, "step": 12640 }, { "epoch": 3.6805353505964504, "grad_norm": 0.14460037648677826, "learning_rate": 2.792835196011512e-05, "loss": 0.0259, "step": 12650 }, { "epoch": 3.6834448647075937, "grad_norm": 0.11711481213569641, "learning_rate": 2.792415766477679e-05, "loss": 0.0253, "step": 12660 }, { "epoch": 3.6863543788187374, "grad_norm": 0.12339013069868088, "learning_rate": 2.79199594434206e-05, "loss": 0.0268, "step": 12670 }, { "epoch": 3.6892638929298807, "grad_norm": 0.13033753633499146, "learning_rate": 2.7915757297321867e-05, "loss": 0.0256, "step": 12680 }, { "epoch": 3.6921734070410244, "grad_norm": 0.151859313249588, "learning_rate": 2.7911551227757085e-05, "loss": 0.0256, "step": 12690 }, { "epoch": 3.6950829211521676, "grad_norm": 0.13655301928520203, "learning_rate": 2.790734123600395e-05, "loss": 0.0253, "step": 12700 }, { "epoch": 3.697992435263311, "grad_norm": 0.13339833915233612, "learning_rate": 2.7903127323341347e-05, "loss": 0.027, "step": 12710 }, { "epoch": 3.7009019493744546, "grad_norm": 0.1128537580370903, "learning_rate": 2.7898909491049353e-05, "loss": 0.0246, "step": 12720 }, { "epoch": 3.703811463485598, "grad_norm": 0.11258917301893234, "learning_rate": 2.789468774040923e-05, "loss": 0.0262, "step": 12730 }, { "epoch": 3.706720977596741, "grad_norm": 0.12364691495895386, "learning_rate": 2.7890462072703433e-05, "loss": 0.0274, "step": 12740 }, { "epoch": 3.709630491707885, "grad_norm": 0.14339150488376617, "learning_rate": 2.7886232489215617e-05, "loss": 0.0286, "step": 12750 }, { "epoch": 3.712540005819028, "grad_norm": 0.12888184189796448, "learning_rate": 2.7881998991230606e-05, "loss": 0.027, "step": 12760 }, { "epoch": 3.715449519930172, "grad_norm": 0.1296333521604538, "learning_rate": 2.7877761580034436e-05, "loss": 0.0265, "step": 12770 }, { "epoch": 3.718359034041315, "grad_norm": 0.10719702392816544, "learning_rate": 2.787352025691431e-05, "loss": 0.0268, "step": 12780 }, { "epoch": 3.7212685481524588, "grad_norm": 0.15584217011928558, "learning_rate": 2.7869275023158642e-05, "loss": 0.0261, "step": 12790 }, { "epoch": 3.724178062263602, "grad_norm": 0.11867398023605347, "learning_rate": 2.7865025880057014e-05, "loss": 0.0284, "step": 12800 }, { "epoch": 3.7270875763747453, "grad_norm": 0.11274160444736481, "learning_rate": 2.78607728289002e-05, "loss": 0.0269, "step": 12810 }, { "epoch": 3.729997090485889, "grad_norm": 0.15163150429725647, "learning_rate": 2.7856515870980176e-05, "loss": 0.0258, "step": 12820 }, { "epoch": 3.7329066045970323, "grad_norm": 0.12092813849449158, "learning_rate": 2.785225500759008e-05, "loss": 0.0257, "step": 12830 }, { "epoch": 3.7358161187081755, "grad_norm": 0.1089605912566185, "learning_rate": 2.7847990240024263e-05, "loss": 0.0256, "step": 12840 }, { "epoch": 3.7387256328193192, "grad_norm": 0.1256360411643982, "learning_rate": 2.784372156957824e-05, "loss": 0.0276, "step": 12850 }, { "epoch": 3.7416351469304625, "grad_norm": 0.12180452048778534, "learning_rate": 2.7839448997548728e-05, "loss": 0.027, "step": 12860 }, { "epoch": 3.744544661041606, "grad_norm": 0.11620897054672241, "learning_rate": 2.783517252523361e-05, "loss": 0.0252, "step": 12870 }, { "epoch": 3.7474541751527495, "grad_norm": 0.10638363659381866, "learning_rate": 2.7830892153931975e-05, "loss": 0.0248, "step": 12880 }, { "epoch": 3.750363689263893, "grad_norm": 0.14196443557739258, "learning_rate": 2.7826607884944083e-05, "loss": 0.0261, "step": 12890 }, { "epoch": 3.7532732033750364, "grad_norm": 0.12689955532550812, "learning_rate": 2.782231971957138e-05, "loss": 0.0277, "step": 12900 }, { "epoch": 3.7561827174861797, "grad_norm": 0.13536278903484344, "learning_rate": 2.78180276591165e-05, "loss": 0.0259, "step": 12910 }, { "epoch": 3.7590922315973234, "grad_norm": 0.11821288615465164, "learning_rate": 2.781373170488326e-05, "loss": 0.0264, "step": 12920 }, { "epoch": 3.7620017457084667, "grad_norm": 0.10341784358024597, "learning_rate": 2.7809431858176655e-05, "loss": 0.0262, "step": 12930 }, { "epoch": 3.76491125981961, "grad_norm": 0.11881068348884583, "learning_rate": 2.7805128120302865e-05, "loss": 0.0273, "step": 12940 }, { "epoch": 3.7678207739307537, "grad_norm": 0.11699987947940826, "learning_rate": 2.7800820492569254e-05, "loss": 0.0268, "step": 12950 }, { "epoch": 3.770730288041897, "grad_norm": 0.1168835461139679, "learning_rate": 2.779650897628436e-05, "loss": 0.0244, "step": 12960 }, { "epoch": 3.77363980215304, "grad_norm": 0.08170627802610397, "learning_rate": 2.7792193572757915e-05, "loss": 0.0256, "step": 12970 }, { "epoch": 3.776549316264184, "grad_norm": 0.13656464219093323, "learning_rate": 2.7787874283300812e-05, "loss": 0.0256, "step": 12980 }, { "epoch": 3.7794588303753276, "grad_norm": 0.12222883105278015, "learning_rate": 2.7783551109225155e-05, "loss": 0.0261, "step": 12990 }, { "epoch": 3.782368344486471, "grad_norm": 0.11611222475767136, "learning_rate": 2.7779224051844195e-05, "loss": 0.0266, "step": 13000 }, { "epoch": 3.785277858597614, "grad_norm": 0.13990092277526855, "learning_rate": 2.777489311247239e-05, "loss": 0.0267, "step": 13010 }, { "epoch": 3.788187372708758, "grad_norm": 0.11403480917215347, "learning_rate": 2.7770558292425355e-05, "loss": 0.0268, "step": 13020 }, { "epoch": 3.791096886819901, "grad_norm": 0.12104486674070358, "learning_rate": 2.77662195930199e-05, "loss": 0.026, "step": 13030 }, { "epoch": 3.7940064009310444, "grad_norm": 0.10701839625835419, "learning_rate": 2.7761877015574005e-05, "loss": 0.0262, "step": 13040 }, { "epoch": 3.796915915042188, "grad_norm": 0.12315713614225388, "learning_rate": 2.775753056140683e-05, "loss": 0.0244, "step": 13050 }, { "epoch": 3.7998254291533313, "grad_norm": 0.1499435305595398, "learning_rate": 2.775318023183871e-05, "loss": 0.0276, "step": 13060 }, { "epoch": 3.8027349432644746, "grad_norm": 0.11804681271314621, "learning_rate": 2.774882602819117e-05, "loss": 0.0269, "step": 13070 }, { "epoch": 3.8056444573756183, "grad_norm": 0.1200079619884491, "learning_rate": 2.7744467951786892e-05, "loss": 0.0262, "step": 13080 }, { "epoch": 3.8085539714867616, "grad_norm": 0.16637791693210602, "learning_rate": 2.7740106003949747e-05, "loss": 0.0254, "step": 13090 }, { "epoch": 3.8114634855979053, "grad_norm": 0.1458464413881302, "learning_rate": 2.7735740186004785e-05, "loss": 0.0247, "step": 13100 }, { "epoch": 3.8143729997090485, "grad_norm": 0.131627157330513, "learning_rate": 2.7731370499278222e-05, "loss": 0.0261, "step": 13110 }, { "epoch": 3.8172825138201922, "grad_norm": 0.12577767670154572, "learning_rate": 2.772699694509745e-05, "loss": 0.0274, "step": 13120 }, { "epoch": 3.8201920279313355, "grad_norm": 0.13190332055091858, "learning_rate": 2.7722619524791046e-05, "loss": 0.0257, "step": 13130 }, { "epoch": 3.8231015420424788, "grad_norm": 0.1422741860151291, "learning_rate": 2.7718238239688747e-05, "loss": 0.0262, "step": 13140 }, { "epoch": 3.8260110561536225, "grad_norm": 0.12308301776647568, "learning_rate": 2.7713853091121484e-05, "loss": 0.0258, "step": 13150 }, { "epoch": 3.8289205702647657, "grad_norm": 0.11852675676345825, "learning_rate": 2.7709464080421338e-05, "loss": 0.0266, "step": 13160 }, { "epoch": 3.831830084375909, "grad_norm": 0.12704512476921082, "learning_rate": 2.770507120892158e-05, "loss": 0.0289, "step": 13170 }, { "epoch": 3.8347395984870527, "grad_norm": 0.12465476244688034, "learning_rate": 2.7700674477956643e-05, "loss": 0.0265, "step": 13180 }, { "epoch": 3.837649112598196, "grad_norm": 0.12637312710285187, "learning_rate": 2.7696273888862145e-05, "loss": 0.0258, "step": 13190 }, { "epoch": 3.8405586267093397, "grad_norm": 0.10013556480407715, "learning_rate": 2.7691869442974864e-05, "loss": 0.0274, "step": 13200 }, { "epoch": 3.843468140820483, "grad_norm": 0.11476576328277588, "learning_rate": 2.7687461141632754e-05, "loss": 0.0252, "step": 13210 }, { "epoch": 3.8463776549316266, "grad_norm": 0.1421649307012558, "learning_rate": 2.7683048986174945e-05, "loss": 0.0267, "step": 13220 }, { "epoch": 3.84928716904277, "grad_norm": 0.11735635995864868, "learning_rate": 2.7678632977941727e-05, "loss": 0.0242, "step": 13230 }, { "epoch": 3.852196683153913, "grad_norm": 0.13946533203125, "learning_rate": 2.767421311827457e-05, "loss": 0.0258, "step": 13240 }, { "epoch": 3.855106197265057, "grad_norm": 0.13182789087295532, "learning_rate": 2.7669789408516117e-05, "loss": 0.0263, "step": 13250 }, { "epoch": 3.8580157113762, "grad_norm": 0.12900641560554504, "learning_rate": 2.766536185001016e-05, "loss": 0.0265, "step": 13260 }, { "epoch": 3.8609252254873434, "grad_norm": 0.13949275016784668, "learning_rate": 2.7660930444101692e-05, "loss": 0.0252, "step": 13270 }, { "epoch": 3.863834739598487, "grad_norm": 0.14166583120822906, "learning_rate": 2.7656495192136842e-05, "loss": 0.0256, "step": 13280 }, { "epoch": 3.8667442537096304, "grad_norm": 0.10012108087539673, "learning_rate": 2.7652056095462926e-05, "loss": 0.0287, "step": 13290 }, { "epoch": 3.869653767820774, "grad_norm": 0.12432703375816345, "learning_rate": 2.764761315542843e-05, "loss": 0.0267, "step": 13300 }, { "epoch": 3.8725632819319173, "grad_norm": 0.11181257665157318, "learning_rate": 2.7643166373383e-05, "loss": 0.0244, "step": 13310 }, { "epoch": 3.875472796043061, "grad_norm": 0.14148472249507904, "learning_rate": 2.763871575067744e-05, "loss": 0.0285, "step": 13320 }, { "epoch": 3.8783823101542043, "grad_norm": 0.1097230389714241, "learning_rate": 2.763426128866375e-05, "loss": 0.025, "step": 13330 }, { "epoch": 3.8812918242653476, "grad_norm": 0.12473341077566147, "learning_rate": 2.7629802988695064e-05, "loss": 0.0272, "step": 13340 }, { "epoch": 3.8842013383764913, "grad_norm": 0.10761269927024841, "learning_rate": 2.7625340852125703e-05, "loss": 0.0255, "step": 13350 }, { "epoch": 3.8871108524876345, "grad_norm": 0.14157076179981232, "learning_rate": 2.7620874880311148e-05, "loss": 0.0251, "step": 13360 }, { "epoch": 3.890020366598778, "grad_norm": 0.10901161283254623, "learning_rate": 2.7616405074608033e-05, "loss": 0.0261, "step": 13370 }, { "epoch": 3.8929298807099215, "grad_norm": 0.1149931401014328, "learning_rate": 2.7611931436374173e-05, "loss": 0.0267, "step": 13380 }, { "epoch": 3.895839394821065, "grad_norm": 0.1168234571814537, "learning_rate": 2.7607453966968545e-05, "loss": 0.026, "step": 13390 }, { "epoch": 3.898748908932208, "grad_norm": 0.15341296792030334, "learning_rate": 2.760297266775128e-05, "loss": 0.0273, "step": 13400 }, { "epoch": 3.9016584230433518, "grad_norm": 0.11903547495603561, "learning_rate": 2.759848754008368e-05, "loss": 0.0251, "step": 13410 }, { "epoch": 3.9045679371544955, "grad_norm": 0.1320924311876297, "learning_rate": 2.759399858532821e-05, "loss": 0.0264, "step": 13420 }, { "epoch": 3.9074774512656387, "grad_norm": 0.12105493247509003, "learning_rate": 2.75895058048485e-05, "loss": 0.0279, "step": 13430 }, { "epoch": 3.910386965376782, "grad_norm": 0.18074245750904083, "learning_rate": 2.7585009200009326e-05, "loss": 0.0256, "step": 13440 }, { "epoch": 3.9132964794879257, "grad_norm": 0.11697207391262054, "learning_rate": 2.7580508772176648e-05, "loss": 0.0242, "step": 13450 }, { "epoch": 3.916205993599069, "grad_norm": 0.11348891258239746, "learning_rate": 2.7576004522717577e-05, "loss": 0.0251, "step": 13460 }, { "epoch": 3.919115507710212, "grad_norm": 0.1279630810022354, "learning_rate": 2.7571496453000377e-05, "loss": 0.0263, "step": 13470 }, { "epoch": 3.922025021821356, "grad_norm": 0.10640689730644226, "learning_rate": 2.7566984564394487e-05, "loss": 0.0245, "step": 13480 }, { "epoch": 3.924934535932499, "grad_norm": 0.15012714266777039, "learning_rate": 2.7562468858270498e-05, "loss": 0.0248, "step": 13490 }, { "epoch": 3.9278440500436425, "grad_norm": 0.1300770491361618, "learning_rate": 2.7557949336000165e-05, "loss": 0.0267, "step": 13500 }, { "epoch": 3.930753564154786, "grad_norm": 0.12519508600234985, "learning_rate": 2.7553425998956395e-05, "loss": 0.0239, "step": 13510 }, { "epoch": 3.9336630782659294, "grad_norm": 0.10229005664587021, "learning_rate": 2.7548898848513264e-05, "loss": 0.0262, "step": 13520 }, { "epoch": 3.936572592377073, "grad_norm": 0.1338396817445755, "learning_rate": 2.754436788604599e-05, "loss": 0.0258, "step": 13530 }, { "epoch": 3.9394821064882164, "grad_norm": 0.10600780695676804, "learning_rate": 2.7539833112930973e-05, "loss": 0.0266, "step": 13540 }, { "epoch": 3.94239162059936, "grad_norm": 0.12628477811813354, "learning_rate": 2.753529453054575e-05, "loss": 0.0243, "step": 13550 }, { "epoch": 3.9453011347105034, "grad_norm": 0.12775465846061707, "learning_rate": 2.7530752140269028e-05, "loss": 0.0259, "step": 13560 }, { "epoch": 3.9482106488216466, "grad_norm": 0.12155082076787949, "learning_rate": 2.7526205943480657e-05, "loss": 0.0252, "step": 13570 }, { "epoch": 3.9511201629327903, "grad_norm": 0.12523499131202698, "learning_rate": 2.7521655941561653e-05, "loss": 0.0231, "step": 13580 }, { "epoch": 3.9540296770439336, "grad_norm": 0.10882124304771423, "learning_rate": 2.751710213589419e-05, "loss": 0.0248, "step": 13590 }, { "epoch": 3.956939191155077, "grad_norm": 0.11985442787408829, "learning_rate": 2.7512544527861597e-05, "loss": 0.0263, "step": 13600 }, { "epoch": 3.9598487052662206, "grad_norm": 0.11696766316890717, "learning_rate": 2.7507983118848353e-05, "loss": 0.0261, "step": 13610 }, { "epoch": 3.962758219377364, "grad_norm": 0.1605997085571289, "learning_rate": 2.7503417910240085e-05, "loss": 0.0253, "step": 13620 }, { "epoch": 3.9656677334885075, "grad_norm": 0.10662023723125458, "learning_rate": 2.7498848903423593e-05, "loss": 0.0272, "step": 13630 }, { "epoch": 3.968577247599651, "grad_norm": 0.1303427815437317, "learning_rate": 2.7494276099786817e-05, "loss": 0.0232, "step": 13640 }, { "epoch": 3.9714867617107945, "grad_norm": 0.12200242280960083, "learning_rate": 2.7489699500718854e-05, "loss": 0.0261, "step": 13650 }, { "epoch": 3.9743962758219378, "grad_norm": 0.11656424403190613, "learning_rate": 2.7485119107609953e-05, "loss": 0.0253, "step": 13660 }, { "epoch": 3.977305789933081, "grad_norm": 0.1298699826002121, "learning_rate": 2.748053492185152e-05, "loss": 0.0253, "step": 13670 }, { "epoch": 3.9802153040442247, "grad_norm": 0.126081645488739, "learning_rate": 2.7475946944836103e-05, "loss": 0.0236, "step": 13680 }, { "epoch": 3.983124818155368, "grad_norm": 0.13921399414539337, "learning_rate": 2.7471355177957414e-05, "loss": 0.0246, "step": 13690 }, { "epoch": 3.9860343322665113, "grad_norm": 0.12696388363838196, "learning_rate": 2.746675962261031e-05, "loss": 0.024, "step": 13700 }, { "epoch": 3.988943846377655, "grad_norm": 0.10440041869878769, "learning_rate": 2.74621602801908e-05, "loss": 0.0251, "step": 13710 }, { "epoch": 3.9918533604887982, "grad_norm": 0.12466151267290115, "learning_rate": 2.745755715209604e-05, "loss": 0.0267, "step": 13720 }, { "epoch": 3.994762874599942, "grad_norm": 0.12947039306163788, "learning_rate": 2.7452950239724337e-05, "loss": 0.0265, "step": 13730 }, { "epoch": 3.997672388711085, "grad_norm": 0.14005382359027863, "learning_rate": 2.7448339544475156e-05, "loss": 0.0256, "step": 13740 }, { "epoch": 4.000581902822229, "grad_norm": 0.1517648696899414, "learning_rate": 2.7443725067749098e-05, "loss": 0.0235, "step": 13750 }, { "epoch": 4.003491416933372, "grad_norm": 0.12960092723369598, "learning_rate": 2.7439106810947923e-05, "loss": 0.0256, "step": 13760 }, { "epoch": 4.0064009310445154, "grad_norm": 0.10935242474079132, "learning_rate": 2.7434484775474532e-05, "loss": 0.0253, "step": 13770 }, { "epoch": 4.009310445155659, "grad_norm": 0.12054923921823502, "learning_rate": 2.7429858962732984e-05, "loss": 0.0244, "step": 13780 }, { "epoch": 4.012219959266803, "grad_norm": 0.10846009105443954, "learning_rate": 2.7425229374128475e-05, "loss": 0.0264, "step": 13790 }, { "epoch": 4.015129473377946, "grad_norm": 0.12531796097755432, "learning_rate": 2.742059601106735e-05, "loss": 0.0245, "step": 13800 }, { "epoch": 4.018038987489089, "grad_norm": 0.12270881235599518, "learning_rate": 2.74159588749571e-05, "loss": 0.0232, "step": 13810 }, { "epoch": 4.020948501600233, "grad_norm": 0.1299131065607071, "learning_rate": 2.7411317967206373e-05, "loss": 0.0272, "step": 13820 }, { "epoch": 4.023858015711376, "grad_norm": 0.13932688534259796, "learning_rate": 2.7406673289224945e-05, "loss": 0.0266, "step": 13830 }, { "epoch": 4.02676752982252, "grad_norm": 0.12068501859903336, "learning_rate": 2.740202484242376e-05, "loss": 0.0256, "step": 13840 }, { "epoch": 4.029677043933663, "grad_norm": 0.11375617235898972, "learning_rate": 2.7397372628214873e-05, "loss": 0.0271, "step": 13850 }, { "epoch": 4.032586558044806, "grad_norm": 0.12344492971897125, "learning_rate": 2.739271664801152e-05, "loss": 0.0269, "step": 13860 }, { "epoch": 4.03549607215595, "grad_norm": 0.11604185402393341, "learning_rate": 2.738805690322806e-05, "loss": 0.0264, "step": 13870 }, { "epoch": 4.038405586267094, "grad_norm": 0.11813008785247803, "learning_rate": 2.7383393395280004e-05, "loss": 0.026, "step": 13880 }, { "epoch": 4.041315100378237, "grad_norm": 0.11945053189992905, "learning_rate": 2.7378726125584e-05, "loss": 0.025, "step": 13890 }, { "epoch": 4.04422461448938, "grad_norm": 0.11909756809473038, "learning_rate": 2.7374055095557832e-05, "loss": 0.0264, "step": 13900 }, { "epoch": 4.047134128600524, "grad_norm": 0.1512121707201004, "learning_rate": 2.7369380306620446e-05, "loss": 0.0255, "step": 13910 }, { "epoch": 4.0500436427116675, "grad_norm": 0.10909899324178696, "learning_rate": 2.7364701760191923e-05, "loss": 0.0263, "step": 13920 }, { "epoch": 4.05295315682281, "grad_norm": 0.11786768585443497, "learning_rate": 2.7360019457693468e-05, "loss": 0.0244, "step": 13930 }, { "epoch": 4.055862670933954, "grad_norm": 0.13229550421237946, "learning_rate": 2.7355333400547457e-05, "loss": 0.026, "step": 13940 }, { "epoch": 4.058772185045098, "grad_norm": 0.1467132717370987, "learning_rate": 2.7350643590177378e-05, "loss": 0.0274, "step": 13950 }, { "epoch": 4.0616816991562406, "grad_norm": 0.1308804452419281, "learning_rate": 2.7345950028007878e-05, "loss": 0.027, "step": 13960 }, { "epoch": 4.064591213267384, "grad_norm": 0.12300308793783188, "learning_rate": 2.7341252715464736e-05, "loss": 0.0241, "step": 13970 }, { "epoch": 4.067500727378528, "grad_norm": 0.12026852369308472, "learning_rate": 2.7336551653974868e-05, "loss": 0.0255, "step": 13980 }, { "epoch": 4.070410241489671, "grad_norm": 0.11192727833986282, "learning_rate": 2.733184684496634e-05, "loss": 0.0268, "step": 13990 }, { "epoch": 4.0733197556008145, "grad_norm": 0.08777649700641632, "learning_rate": 2.7327138289868337e-05, "loss": 0.0263, "step": 14000 }, { "epoch": 4.076229269711958, "grad_norm": 0.14284144341945648, "learning_rate": 2.7322425990111208e-05, "loss": 0.0252, "step": 14010 }, { "epoch": 4.079138783823102, "grad_norm": 0.09247416257858276, "learning_rate": 2.7317709947126416e-05, "loss": 0.0253, "step": 14020 }, { "epoch": 4.082048297934245, "grad_norm": 0.1215178519487381, "learning_rate": 2.731299016234657e-05, "loss": 0.0277, "step": 14030 }, { "epoch": 4.084957812045388, "grad_norm": 0.14825576543807983, "learning_rate": 2.730826663720542e-05, "loss": 0.0259, "step": 14040 }, { "epoch": 4.087867326156532, "grad_norm": 0.11010383069515228, "learning_rate": 2.730353937313785e-05, "loss": 0.0266, "step": 14050 }, { "epoch": 4.090776840267675, "grad_norm": 0.11698488891124725, "learning_rate": 2.729880837157987e-05, "loss": 0.0267, "step": 14060 }, { "epoch": 4.093686354378819, "grad_norm": 0.13126420974731445, "learning_rate": 2.729407363396864e-05, "loss": 0.0269, "step": 14070 }, { "epoch": 4.096595868489962, "grad_norm": 0.11659082770347595, "learning_rate": 2.7289335161742444e-05, "loss": 0.0258, "step": 14080 }, { "epoch": 4.099505382601105, "grad_norm": 0.12358759343624115, "learning_rate": 2.7284592956340707e-05, "loss": 0.0259, "step": 14090 }, { "epoch": 4.102414896712249, "grad_norm": 0.10540125519037247, "learning_rate": 2.727984701920399e-05, "loss": 0.0243, "step": 14100 }, { "epoch": 4.105324410823393, "grad_norm": 0.12213299423456192, "learning_rate": 2.7275097351773977e-05, "loss": 0.0231, "step": 14110 }, { "epoch": 4.108233924934536, "grad_norm": 0.15919367969036102, "learning_rate": 2.727034395549349e-05, "loss": 0.0232, "step": 14120 }, { "epoch": 4.111143439045679, "grad_norm": 0.1253902018070221, "learning_rate": 2.726558683180649e-05, "loss": 0.0271, "step": 14130 }, { "epoch": 4.114052953156823, "grad_norm": 0.09258479624986649, "learning_rate": 2.7260825982158067e-05, "loss": 0.0276, "step": 14140 }, { "epoch": 4.1169624672679666, "grad_norm": 0.10731474310159683, "learning_rate": 2.725606140799444e-05, "loss": 0.0255, "step": 14150 }, { "epoch": 4.119871981379109, "grad_norm": 0.11090464144945145, "learning_rate": 2.7251293110762957e-05, "loss": 0.0262, "step": 14160 }, { "epoch": 4.122781495490253, "grad_norm": 0.14142276346683502, "learning_rate": 2.7246521091912106e-05, "loss": 0.0261, "step": 14170 }, { "epoch": 4.125691009601397, "grad_norm": 0.11770179122686386, "learning_rate": 2.7241745352891497e-05, "loss": 0.0251, "step": 14180 }, { "epoch": 4.12860052371254, "grad_norm": 0.10821583867073059, "learning_rate": 2.7236965895151874e-05, "loss": 0.0245, "step": 14190 }, { "epoch": 4.131510037823683, "grad_norm": 0.10436071455478668, "learning_rate": 2.7232182720145105e-05, "loss": 0.0255, "step": 14200 }, { "epoch": 4.134419551934827, "grad_norm": 0.08868594467639923, "learning_rate": 2.7227395829324208e-05, "loss": 0.0237, "step": 14210 }, { "epoch": 4.137329066045971, "grad_norm": 0.10813303291797638, "learning_rate": 2.7222605224143298e-05, "loss": 0.0227, "step": 14220 }, { "epoch": 4.1402385801571135, "grad_norm": 0.11614318937063217, "learning_rate": 2.721781090605764e-05, "loss": 0.0245, "step": 14230 }, { "epoch": 4.143148094268257, "grad_norm": 0.11284636706113815, "learning_rate": 2.721301287652362e-05, "loss": 0.0249, "step": 14240 }, { "epoch": 4.146057608379401, "grad_norm": 0.09370618313550949, "learning_rate": 2.7208211136998758e-05, "loss": 0.0253, "step": 14250 }, { "epoch": 4.148967122490544, "grad_norm": 0.11440402269363403, "learning_rate": 2.7203405688941693e-05, "loss": 0.0251, "step": 14260 }, { "epoch": 4.1518766366016875, "grad_norm": 0.11432331055402756, "learning_rate": 2.719859653381219e-05, "loss": 0.022, "step": 14270 }, { "epoch": 4.154786150712831, "grad_norm": 0.13261118531227112, "learning_rate": 2.719378367307114e-05, "loss": 0.024, "step": 14280 }, { "epoch": 4.157695664823974, "grad_norm": 0.14612187445163727, "learning_rate": 2.7188967108180574e-05, "loss": 0.0257, "step": 14290 }, { "epoch": 4.160605178935118, "grad_norm": 0.15001417696475983, "learning_rate": 2.7184146840603634e-05, "loss": 0.0253, "step": 14300 }, { "epoch": 4.163514693046261, "grad_norm": 0.12831546366214752, "learning_rate": 2.7179322871804584e-05, "loss": 0.0243, "step": 14310 }, { "epoch": 4.166424207157405, "grad_norm": 0.13748694956302643, "learning_rate": 2.717449520324882e-05, "loss": 0.0251, "step": 14320 }, { "epoch": 4.169333721268548, "grad_norm": 0.11847007274627686, "learning_rate": 2.7169663836402864e-05, "loss": 0.0254, "step": 14330 }, { "epoch": 4.172243235379692, "grad_norm": 0.15918590128421783, "learning_rate": 2.7164828772734352e-05, "loss": 0.0244, "step": 14340 }, { "epoch": 4.175152749490835, "grad_norm": 0.12226768583059311, "learning_rate": 2.7159990013712057e-05, "loss": 0.0255, "step": 14350 }, { "epoch": 4.178062263601978, "grad_norm": 0.13599678874015808, "learning_rate": 2.7155147560805863e-05, "loss": 0.0247, "step": 14360 }, { "epoch": 4.180971777713122, "grad_norm": 0.13958293199539185, "learning_rate": 2.7150301415486777e-05, "loss": 0.0258, "step": 14370 }, { "epoch": 4.183881291824266, "grad_norm": 0.12577730417251587, "learning_rate": 2.7145451579226932e-05, "loss": 0.0253, "step": 14380 }, { "epoch": 4.186790805935408, "grad_norm": 0.10721693933010101, "learning_rate": 2.7140598053499577e-05, "loss": 0.0251, "step": 14390 }, { "epoch": 4.189700320046552, "grad_norm": 0.11048572510480881, "learning_rate": 2.7135740839779087e-05, "loss": 0.0252, "step": 14400 }, { "epoch": 4.192609834157696, "grad_norm": 0.11912602186203003, "learning_rate": 2.7130879939540957e-05, "loss": 0.0246, "step": 14410 }, { "epoch": 4.195519348268839, "grad_norm": 0.13067324459552765, "learning_rate": 2.7126015354261798e-05, "loss": 0.025, "step": 14420 }, { "epoch": 4.198428862379982, "grad_norm": 0.1139615923166275, "learning_rate": 2.712114708541934e-05, "loss": 0.0258, "step": 14430 }, { "epoch": 4.201338376491126, "grad_norm": 0.11214298754930496, "learning_rate": 2.7116275134492445e-05, "loss": 0.0263, "step": 14440 }, { "epoch": 4.20424789060227, "grad_norm": 0.11281604319810867, "learning_rate": 2.7111399502961073e-05, "loss": 0.0262, "step": 14450 }, { "epoch": 4.207157404713413, "grad_norm": 0.11715683341026306, "learning_rate": 2.710652019230631e-05, "loss": 0.0267, "step": 14460 }, { "epoch": 4.210066918824556, "grad_norm": 0.13563761115074158, "learning_rate": 2.710163720401037e-05, "loss": 0.0253, "step": 14470 }, { "epoch": 4.2129764329357, "grad_norm": 0.10548754036426544, "learning_rate": 2.709675053955658e-05, "loss": 0.0255, "step": 14480 }, { "epoch": 4.215885947046843, "grad_norm": 0.10857761651277542, "learning_rate": 2.7091860200429362e-05, "loss": 0.0249, "step": 14490 }, { "epoch": 4.2187954611579865, "grad_norm": 0.14305821061134338, "learning_rate": 2.7086966188114285e-05, "loss": 0.0249, "step": 14500 }, { "epoch": 4.22170497526913, "grad_norm": 0.09745027869939804, "learning_rate": 2.708206850409802e-05, "loss": 0.023, "step": 14510 }, { "epoch": 4.224614489380274, "grad_norm": 0.13074645400047302, "learning_rate": 2.707716714986835e-05, "loss": 0.0256, "step": 14520 }, { "epoch": 4.227524003491417, "grad_norm": 0.12332235276699066, "learning_rate": 2.707226212691418e-05, "loss": 0.0252, "step": 14530 }, { "epoch": 4.2304335176025605, "grad_norm": 0.16582684218883514, "learning_rate": 2.7067353436725525e-05, "loss": 0.0252, "step": 14540 }, { "epoch": 4.233343031713704, "grad_norm": 0.13545599579811096, "learning_rate": 2.706244108079352e-05, "loss": 0.0245, "step": 14550 }, { "epoch": 4.236252545824847, "grad_norm": 0.17190690338611603, "learning_rate": 2.7057525060610396e-05, "loss": 0.0261, "step": 14560 }, { "epoch": 4.239162059935991, "grad_norm": 0.1255744993686676, "learning_rate": 2.7052605377669527e-05, "loss": 0.0258, "step": 14570 }, { "epoch": 4.242071574047134, "grad_norm": 0.12895460426807404, "learning_rate": 2.704768203346537e-05, "loss": 0.0265, "step": 14580 }, { "epoch": 4.244981088158277, "grad_norm": 0.15400294959545135, "learning_rate": 2.7042755029493513e-05, "loss": 0.0264, "step": 14590 }, { "epoch": 4.247890602269421, "grad_norm": 0.1275034099817276, "learning_rate": 2.703782436725065e-05, "loss": 0.0245, "step": 14600 }, { "epoch": 4.250800116380565, "grad_norm": 0.11625692248344421, "learning_rate": 2.7032890048234585e-05, "loss": 0.0237, "step": 14610 }, { "epoch": 4.2537096304917075, "grad_norm": 0.10397017002105713, "learning_rate": 2.702795207394423e-05, "loss": 0.0259, "step": 14620 }, { "epoch": 4.256619144602851, "grad_norm": 0.1202000230550766, "learning_rate": 2.702301044587962e-05, "loss": 0.0245, "step": 14630 }, { "epoch": 4.259528658713995, "grad_norm": 0.09448982775211334, "learning_rate": 2.7018065165541885e-05, "loss": 0.0244, "step": 14640 }, { "epoch": 4.262438172825139, "grad_norm": 0.13024188578128815, "learning_rate": 2.701311623443327e-05, "loss": 0.0251, "step": 14650 }, { "epoch": 4.265347686936281, "grad_norm": 0.1147368997335434, "learning_rate": 2.700816365405713e-05, "loss": 0.0252, "step": 14660 }, { "epoch": 4.268257201047425, "grad_norm": 0.13647782802581787, "learning_rate": 2.7003207425917926e-05, "loss": 0.0236, "step": 14670 }, { "epoch": 4.271166715158569, "grad_norm": 0.11814698576927185, "learning_rate": 2.6998247551521238e-05, "loss": 0.0262, "step": 14680 }, { "epoch": 4.274076229269712, "grad_norm": 0.11014021933078766, "learning_rate": 2.6993284032373732e-05, "loss": 0.0244, "step": 14690 }, { "epoch": 4.276985743380855, "grad_norm": 0.13047325611114502, "learning_rate": 2.6988316869983208e-05, "loss": 0.0245, "step": 14700 }, { "epoch": 4.279895257491999, "grad_norm": 0.10857924073934555, "learning_rate": 2.698334606585855e-05, "loss": 0.0244, "step": 14710 }, { "epoch": 4.282804771603142, "grad_norm": 0.1003536656498909, "learning_rate": 2.6978371621509757e-05, "loss": 0.027, "step": 14720 }, { "epoch": 4.285714285714286, "grad_norm": 0.12625855207443237, "learning_rate": 2.697339353844794e-05, "loss": 0.0244, "step": 14730 }, { "epoch": 4.288623799825429, "grad_norm": 0.09122280776500702, "learning_rate": 2.6968411818185303e-05, "loss": 0.0264, "step": 14740 }, { "epoch": 4.291533313936573, "grad_norm": 0.127949520945549, "learning_rate": 2.6963426462235167e-05, "loss": 0.0262, "step": 14750 }, { "epoch": 4.294442828047716, "grad_norm": 0.13052774965763092, "learning_rate": 2.6958437472111947e-05, "loss": 0.0265, "step": 14760 }, { "epoch": 4.2973523421588595, "grad_norm": 0.13659770786762238, "learning_rate": 2.695344484933116e-05, "loss": 0.0235, "step": 14770 }, { "epoch": 4.300261856270003, "grad_norm": 0.13372528553009033, "learning_rate": 2.6948448595409443e-05, "loss": 0.0247, "step": 14780 }, { "epoch": 4.303171370381146, "grad_norm": 0.13439680635929108, "learning_rate": 2.694344871186453e-05, "loss": 0.0254, "step": 14790 }, { "epoch": 4.30608088449229, "grad_norm": 0.11588787287473679, "learning_rate": 2.6938445200215237e-05, "loss": 0.0252, "step": 14800 }, { "epoch": 4.3089903986034335, "grad_norm": 0.13188688457012177, "learning_rate": 2.6933438061981513e-05, "loss": 0.0247, "step": 14810 }, { "epoch": 4.311899912714576, "grad_norm": 0.12196939438581467, "learning_rate": 2.6928427298684395e-05, "loss": 0.0263, "step": 14820 }, { "epoch": 4.31480942682572, "grad_norm": 0.103696309030056, "learning_rate": 2.6923412911846008e-05, "loss": 0.0239, "step": 14830 }, { "epoch": 4.317718940936864, "grad_norm": 0.1396871656179428, "learning_rate": 2.6918394902989604e-05, "loss": 0.0234, "step": 14840 }, { "epoch": 4.320628455048007, "grad_norm": 0.10123049467802048, "learning_rate": 2.6913373273639508e-05, "loss": 0.0256, "step": 14850 }, { "epoch": 4.32353796915915, "grad_norm": 0.14893856644630432, "learning_rate": 2.690834802532117e-05, "loss": 0.0229, "step": 14860 }, { "epoch": 4.326447483270294, "grad_norm": 0.10221580415964127, "learning_rate": 2.6903319159561125e-05, "loss": 0.0234, "step": 14870 }, { "epoch": 4.329356997381438, "grad_norm": 0.13062062859535217, "learning_rate": 2.6898286677887005e-05, "loss": 0.0255, "step": 14880 }, { "epoch": 4.3322665114925805, "grad_norm": 0.1272597312927246, "learning_rate": 2.689325058182755e-05, "loss": 0.0252, "step": 14890 }, { "epoch": 4.335176025603724, "grad_norm": 0.11900710314512253, "learning_rate": 2.6888210872912594e-05, "loss": 0.0251, "step": 14900 }, { "epoch": 4.338085539714868, "grad_norm": 0.13248082995414734, "learning_rate": 2.688316755267306e-05, "loss": 0.0249, "step": 14910 }, { "epoch": 4.340995053826011, "grad_norm": 0.12057030200958252, "learning_rate": 2.6878120622640988e-05, "loss": 0.0244, "step": 14920 }, { "epoch": 4.343904567937154, "grad_norm": 0.10824505239725113, "learning_rate": 2.6873070084349486e-05, "loss": 0.0253, "step": 14930 }, { "epoch": 4.346814082048298, "grad_norm": 0.13594679534435272, "learning_rate": 2.6868015939332788e-05, "loss": 0.025, "step": 14940 }, { "epoch": 4.349723596159441, "grad_norm": 0.10376486927270889, "learning_rate": 2.68629581891262e-05, "loss": 0.0247, "step": 14950 }, { "epoch": 4.352633110270585, "grad_norm": 0.10825996100902557, "learning_rate": 2.685789683526614e-05, "loss": 0.025, "step": 14960 }, { "epoch": 4.355542624381728, "grad_norm": 0.15123404562473297, "learning_rate": 2.685283187929011e-05, "loss": 0.0268, "step": 14970 }, { "epoch": 4.358452138492872, "grad_norm": 0.11526373773813248, "learning_rate": 2.6847763322736713e-05, "loss": 0.0259, "step": 14980 }, { "epoch": 4.361361652604015, "grad_norm": 0.11731050908565521, "learning_rate": 2.684269116714564e-05, "loss": 0.0238, "step": 14990 }, { "epoch": 4.364271166715159, "grad_norm": 0.09724122285842896, "learning_rate": 2.6837615414057677e-05, "loss": 0.0236, "step": 15000 }, { "epoch": 4.367180680826302, "grad_norm": 0.10560443252325058, "learning_rate": 2.6832536065014708e-05, "loss": 0.0243, "step": 15010 }, { "epoch": 4.370090194937445, "grad_norm": 0.10960055142641068, "learning_rate": 2.68274531215597e-05, "loss": 0.0258, "step": 15020 }, { "epoch": 4.372999709048589, "grad_norm": 0.1165262907743454, "learning_rate": 2.6822366585236716e-05, "loss": 0.0235, "step": 15030 }, { "epoch": 4.3759092231597325, "grad_norm": 0.13830050826072693, "learning_rate": 2.6817276457590924e-05, "loss": 0.0268, "step": 15040 }, { "epoch": 4.378818737270876, "grad_norm": 0.10577180236577988, "learning_rate": 2.6812182740168555e-05, "loss": 0.025, "step": 15050 }, { "epoch": 4.381728251382019, "grad_norm": 0.13166093826293945, "learning_rate": 2.6807085434516953e-05, "loss": 0.0234, "step": 15060 }, { "epoch": 4.384637765493163, "grad_norm": 0.11998872458934784, "learning_rate": 2.6801984542184544e-05, "loss": 0.0272, "step": 15070 }, { "epoch": 4.3875472796043065, "grad_norm": 0.11889805644750595, "learning_rate": 2.6796880064720845e-05, "loss": 0.0259, "step": 15080 }, { "epoch": 4.390456793715449, "grad_norm": 0.11419632285833359, "learning_rate": 2.6791772003676462e-05, "loss": 0.0233, "step": 15090 }, { "epoch": 4.393366307826593, "grad_norm": 0.13548928499221802, "learning_rate": 2.6786660360603087e-05, "loss": 0.0238, "step": 15100 }, { "epoch": 4.396275821937737, "grad_norm": 0.12927013635635376, "learning_rate": 2.6781545137053503e-05, "loss": 0.0238, "step": 15110 }, { "epoch": 4.3991853360488795, "grad_norm": 0.13256949186325073, "learning_rate": 2.677642633458158e-05, "loss": 0.0244, "step": 15120 }, { "epoch": 4.402094850160023, "grad_norm": 0.1079435870051384, "learning_rate": 2.6771303954742274e-05, "loss": 0.024, "step": 15130 }, { "epoch": 4.405004364271167, "grad_norm": 0.1883859783411026, "learning_rate": 2.6766177999091633e-05, "loss": 0.0248, "step": 15140 }, { "epoch": 4.40791387838231, "grad_norm": 0.12561070919036865, "learning_rate": 2.676104846918678e-05, "loss": 0.0238, "step": 15150 }, { "epoch": 4.4108233924934535, "grad_norm": 0.08564610034227371, "learning_rate": 2.675591536658594e-05, "loss": 0.0242, "step": 15160 }, { "epoch": 4.413732906604597, "grad_norm": 0.0792727917432785, "learning_rate": 2.6750778692848405e-05, "loss": 0.025, "step": 15170 }, { "epoch": 4.416642420715741, "grad_norm": 0.10235551744699478, "learning_rate": 2.6745638449534562e-05, "loss": 0.0247, "step": 15180 }, { "epoch": 4.419551934826884, "grad_norm": 0.13966532051563263, "learning_rate": 2.6740494638205888e-05, "loss": 0.0234, "step": 15190 }, { "epoch": 4.422461448938027, "grad_norm": 0.11886624991893768, "learning_rate": 2.6735347260424928e-05, "loss": 0.0241, "step": 15200 }, { "epoch": 4.425370963049171, "grad_norm": 0.12305711954832077, "learning_rate": 2.6730196317755326e-05, "loss": 0.0279, "step": 15210 }, { "epoch": 4.428280477160314, "grad_norm": 0.12275588512420654, "learning_rate": 2.6725041811761805e-05, "loss": 0.0233, "step": 15220 }, { "epoch": 4.431189991271458, "grad_norm": 0.16699235141277313, "learning_rate": 2.671988374401016e-05, "loss": 0.024, "step": 15230 }, { "epoch": 4.434099505382601, "grad_norm": 0.11265746504068375, "learning_rate": 2.671472211606728e-05, "loss": 0.0226, "step": 15240 }, { "epoch": 4.437009019493744, "grad_norm": 0.11105114221572876, "learning_rate": 2.6709556929501128e-05, "loss": 0.0227, "step": 15250 }, { "epoch": 4.439918533604888, "grad_norm": 0.13808885216712952, "learning_rate": 2.670438818588076e-05, "loss": 0.0231, "step": 15260 }, { "epoch": 4.442828047716032, "grad_norm": 0.11429394036531448, "learning_rate": 2.6699215886776287e-05, "loss": 0.0241, "step": 15270 }, { "epoch": 4.445737561827175, "grad_norm": 0.12478657066822052, "learning_rate": 2.6694040033758933e-05, "loss": 0.0238, "step": 15280 }, { "epoch": 4.448647075938318, "grad_norm": 0.10802243649959564, "learning_rate": 2.6688860628400982e-05, "loss": 0.0237, "step": 15290 }, { "epoch": 4.451556590049462, "grad_norm": 0.11436960846185684, "learning_rate": 2.6683677672275797e-05, "loss": 0.0235, "step": 15300 }, { "epoch": 4.4544661041606055, "grad_norm": 0.145559161901474, "learning_rate": 2.6678491166957825e-05, "loss": 0.0254, "step": 15310 }, { "epoch": 4.457375618271748, "grad_norm": 0.1230301782488823, "learning_rate": 2.6673301114022592e-05, "loss": 0.0221, "step": 15320 }, { "epoch": 4.460285132382892, "grad_norm": 0.0915931835770607, "learning_rate": 2.666810751504669e-05, "loss": 0.0235, "step": 15330 }, { "epoch": 4.463194646494036, "grad_norm": 0.1394130289554596, "learning_rate": 2.6662910371607807e-05, "loss": 0.0245, "step": 15340 }, { "epoch": 4.466104160605179, "grad_norm": 0.12163946032524109, "learning_rate": 2.6657709685284696e-05, "loss": 0.0243, "step": 15350 }, { "epoch": 4.469013674716322, "grad_norm": 0.12433195859193802, "learning_rate": 2.6652505457657183e-05, "loss": 0.0265, "step": 15360 }, { "epoch": 4.471923188827466, "grad_norm": 0.12456993758678436, "learning_rate": 2.664729769030618e-05, "loss": 0.0234, "step": 15370 }, { "epoch": 4.47483270293861, "grad_norm": 0.11860794574022293, "learning_rate": 2.6642086384813667e-05, "loss": 0.0264, "step": 15380 }, { "epoch": 4.4777422170497525, "grad_norm": 0.1157834529876709, "learning_rate": 2.6636871542762703e-05, "loss": 0.0236, "step": 15390 }, { "epoch": 4.480651731160896, "grad_norm": 0.1046522855758667, "learning_rate": 2.6631653165737418e-05, "loss": 0.024, "step": 15400 }, { "epoch": 4.48356124527204, "grad_norm": 0.11108940094709396, "learning_rate": 2.662643125532302e-05, "loss": 0.0254, "step": 15410 }, { "epoch": 4.486470759383183, "grad_norm": 0.14857666194438934, "learning_rate": 2.6621205813105777e-05, "loss": 0.0246, "step": 15420 }, { "epoch": 4.4893802734943264, "grad_norm": 0.08830012381076813, "learning_rate": 2.6615976840673056e-05, "loss": 0.0221, "step": 15430 }, { "epoch": 4.49228978760547, "grad_norm": 0.12634865939617157, "learning_rate": 2.6610744339613265e-05, "loss": 0.0228, "step": 15440 }, { "epoch": 4.495199301716613, "grad_norm": 0.12961223721504211, "learning_rate": 2.6605508311515916e-05, "loss": 0.0265, "step": 15450 }, { "epoch": 4.498108815827757, "grad_norm": 0.1250443011522293, "learning_rate": 2.6600268757971566e-05, "loss": 0.0228, "step": 15460 }, { "epoch": 4.5010183299389, "grad_norm": 0.11888087540864944, "learning_rate": 2.659502568057185e-05, "loss": 0.0239, "step": 15470 }, { "epoch": 4.503927844050043, "grad_norm": 0.12149249017238617, "learning_rate": 2.658977908090949e-05, "loss": 0.025, "step": 15480 }, { "epoch": 4.506837358161187, "grad_norm": 0.09852369129657745, "learning_rate": 2.6584528960578245e-05, "loss": 0.0229, "step": 15490 }, { "epoch": 4.509746872272331, "grad_norm": 0.09380757808685303, "learning_rate": 2.657927532117298e-05, "loss": 0.0242, "step": 15500 }, { "epoch": 4.512656386383474, "grad_norm": 0.12444166839122772, "learning_rate": 2.6574018164289605e-05, "loss": 0.0247, "step": 15510 }, { "epoch": 4.515565900494617, "grad_norm": 0.12764182686805725, "learning_rate": 2.6568757491525103e-05, "loss": 0.0237, "step": 15520 }, { "epoch": 4.518475414605761, "grad_norm": 0.1082252562046051, "learning_rate": 2.6563493304477534e-05, "loss": 0.026, "step": 15530 }, { "epoch": 4.521384928716905, "grad_norm": 0.114132821559906, "learning_rate": 2.655822560474601e-05, "loss": 0.0253, "step": 15540 }, { "epoch": 4.524294442828047, "grad_norm": 0.09896709769964218, "learning_rate": 2.655295439393073e-05, "loss": 0.0237, "step": 15550 }, { "epoch": 4.527203956939191, "grad_norm": 0.10797465592622757, "learning_rate": 2.654767967363294e-05, "loss": 0.0261, "step": 15560 }, { "epoch": 4.530113471050335, "grad_norm": 0.1021900624036789, "learning_rate": 2.6542401445454967e-05, "loss": 0.024, "step": 15570 }, { "epoch": 4.5330229851614785, "grad_norm": 0.1602030247449875, "learning_rate": 2.6537119711000187e-05, "loss": 0.0235, "step": 15580 }, { "epoch": 4.535932499272621, "grad_norm": 0.11139824241399765, "learning_rate": 2.6531834471873068e-05, "loss": 0.0235, "step": 15590 }, { "epoch": 4.538842013383765, "grad_norm": 0.10680071264505386, "learning_rate": 2.6526545729679115e-05, "loss": 0.0255, "step": 15600 }, { "epoch": 4.541751527494909, "grad_norm": 0.14584244787693024, "learning_rate": 2.6521253486024908e-05, "loss": 0.0238, "step": 15610 }, { "epoch": 4.5446610416060516, "grad_norm": 0.15090355277061462, "learning_rate": 2.651595774251809e-05, "loss": 0.0221, "step": 15620 }, { "epoch": 4.547570555717195, "grad_norm": 0.13442876935005188, "learning_rate": 2.6510658500767378e-05, "loss": 0.0243, "step": 15630 }, { "epoch": 4.550480069828339, "grad_norm": 0.11132744699716568, "learning_rate": 2.650535576238253e-05, "loss": 0.025, "step": 15640 }, { "epoch": 4.553389583939482, "grad_norm": 0.10832029581069946, "learning_rate": 2.650004952897438e-05, "loss": 0.024, "step": 15650 }, { "epoch": 4.5562990980506255, "grad_norm": 0.11227482557296753, "learning_rate": 2.649473980215483e-05, "loss": 0.0229, "step": 15660 }, { "epoch": 4.559208612161769, "grad_norm": 0.08507394790649414, "learning_rate": 2.648942658353683e-05, "loss": 0.0242, "step": 15670 }, { "epoch": 4.562118126272912, "grad_norm": 0.13298965990543365, "learning_rate": 2.648410987473439e-05, "loss": 0.0235, "step": 15680 }, { "epoch": 4.565027640384056, "grad_norm": 0.13807477056980133, "learning_rate": 2.6478789677362596e-05, "loss": 0.0238, "step": 15690 }, { "epoch": 4.567937154495199, "grad_norm": 0.1342650055885315, "learning_rate": 2.6473465993037572e-05, "loss": 0.025, "step": 15700 }, { "epoch": 4.570846668606343, "grad_norm": 0.10691782087087631, "learning_rate": 2.6468138823376522e-05, "loss": 0.0251, "step": 15710 }, { "epoch": 4.573756182717486, "grad_norm": 0.16872960329055786, "learning_rate": 2.6462808169997695e-05, "loss": 0.0261, "step": 15720 }, { "epoch": 4.57666569682863, "grad_norm": 0.12431442737579346, "learning_rate": 2.6457474034520402e-05, "loss": 0.0241, "step": 15730 }, { "epoch": 4.579575210939773, "grad_norm": 0.11066822707653046, "learning_rate": 2.6452136418565016e-05, "loss": 0.023, "step": 15740 }, { "epoch": 4.582484725050916, "grad_norm": 0.10822102427482605, "learning_rate": 2.6446795323752962e-05, "loss": 0.0237, "step": 15750 }, { "epoch": 4.58539423916206, "grad_norm": 0.11289121955633163, "learning_rate": 2.6441450751706725e-05, "loss": 0.0234, "step": 15760 }, { "epoch": 4.588303753273204, "grad_norm": 0.1313984990119934, "learning_rate": 2.6436102704049843e-05, "loss": 0.0239, "step": 15770 }, { "epoch": 4.591213267384346, "grad_norm": 0.13679704070091248, "learning_rate": 2.6430751182406913e-05, "loss": 0.0231, "step": 15780 }, { "epoch": 4.59412278149549, "grad_norm": 0.13039728999137878, "learning_rate": 2.6425396188403585e-05, "loss": 0.024, "step": 15790 }, { "epoch": 4.597032295606634, "grad_norm": 0.11365078389644623, "learning_rate": 2.6420037723666565e-05, "loss": 0.0227, "step": 15800 }, { "epoch": 4.599941809717777, "grad_norm": 0.1297876387834549, "learning_rate": 2.641467578982361e-05, "loss": 0.0247, "step": 15810 }, { "epoch": 4.60285132382892, "grad_norm": 0.15118847787380219, "learning_rate": 2.6409310388503542e-05, "loss": 0.0247, "step": 15820 }, { "epoch": 4.605760837940064, "grad_norm": 0.11209290474653244, "learning_rate": 2.640394152133622e-05, "loss": 0.0235, "step": 15830 }, { "epoch": 4.608670352051208, "grad_norm": 0.11775553971529007, "learning_rate": 2.6398569189952573e-05, "loss": 0.0228, "step": 15840 }, { "epoch": 4.611579866162351, "grad_norm": 0.15107856690883636, "learning_rate": 2.6393193395984563e-05, "loss": 0.0242, "step": 15850 }, { "epoch": 4.614489380273494, "grad_norm": 0.10574696958065033, "learning_rate": 2.6387814141065216e-05, "loss": 0.0235, "step": 15860 }, { "epoch": 4.617398894384638, "grad_norm": 0.10095727443695068, "learning_rate": 2.6382431426828615e-05, "loss": 0.023, "step": 15870 }, { "epoch": 4.620308408495781, "grad_norm": 0.08201318234205246, "learning_rate": 2.6377045254909885e-05, "loss": 0.0229, "step": 15880 }, { "epoch": 4.6232179226069245, "grad_norm": 0.13268642127513885, "learning_rate": 2.6371655626945196e-05, "loss": 0.0248, "step": 15890 }, { "epoch": 4.626127436718068, "grad_norm": 0.12042821198701859, "learning_rate": 2.6366262544571778e-05, "loss": 0.026, "step": 15900 }, { "epoch": 4.629036950829212, "grad_norm": 0.10729409009218216, "learning_rate": 2.636086600942791e-05, "loss": 0.0228, "step": 15910 }, { "epoch": 4.631946464940355, "grad_norm": 0.15185776352882385, "learning_rate": 2.6355466023152913e-05, "loss": 0.0236, "step": 15920 }, { "epoch": 4.6348559790514985, "grad_norm": 0.1233530268073082, "learning_rate": 2.6350062587387164e-05, "loss": 0.0257, "step": 15930 }, { "epoch": 4.637765493162642, "grad_norm": 0.11047647893428802, "learning_rate": 2.634465570377208e-05, "loss": 0.0248, "step": 15940 }, { "epoch": 4.640675007273785, "grad_norm": 0.10388235002756119, "learning_rate": 2.633924537395013e-05, "loss": 0.0252, "step": 15950 }, { "epoch": 4.643584521384929, "grad_norm": 0.10441145300865173, "learning_rate": 2.6333831599564835e-05, "loss": 0.0237, "step": 15960 }, { "epoch": 4.646494035496072, "grad_norm": 0.12902361154556274, "learning_rate": 2.632841438226075e-05, "loss": 0.0221, "step": 15970 }, { "epoch": 4.649403549607215, "grad_norm": 0.11829404532909393, "learning_rate": 2.6322993723683486e-05, "loss": 0.0274, "step": 15980 }, { "epoch": 4.652313063718359, "grad_norm": 0.1318294256925583, "learning_rate": 2.6317569625479695e-05, "loss": 0.0253, "step": 15990 }, { "epoch": 4.655222577829503, "grad_norm": 0.10689719766378403, "learning_rate": 2.6312142089297076e-05, "loss": 0.0222, "step": 16000 }, { "epoch": 4.6581320919406455, "grad_norm": 0.09516586363315582, "learning_rate": 2.6306711116784366e-05, "loss": 0.0235, "step": 16010 }, { "epoch": 4.661041606051789, "grad_norm": 0.1499670296907425, "learning_rate": 2.6301276709591358e-05, "loss": 0.0263, "step": 16020 }, { "epoch": 4.663951120162933, "grad_norm": 0.11875282227993011, "learning_rate": 2.6295838869368878e-05, "loss": 0.0242, "step": 16030 }, { "epoch": 4.666860634274077, "grad_norm": 0.14864400029182434, "learning_rate": 2.6290397597768803e-05, "loss": 0.0217, "step": 16040 }, { "epoch": 4.669770148385219, "grad_norm": 0.10844101011753082, "learning_rate": 2.6284952896444037e-05, "loss": 0.0249, "step": 16050 }, { "epoch": 4.672679662496363, "grad_norm": 0.0979212298989296, "learning_rate": 2.6279504767048547e-05, "loss": 0.0244, "step": 16060 }, { "epoch": 4.675589176607507, "grad_norm": 0.12195292115211487, "learning_rate": 2.6274053211237323e-05, "loss": 0.0234, "step": 16070 }, { "epoch": 4.67849869071865, "grad_norm": 0.12838438153266907, "learning_rate": 2.626859823066641e-05, "loss": 0.0249, "step": 16080 }, { "epoch": 4.681408204829793, "grad_norm": 0.11526485532522202, "learning_rate": 2.6263139826992886e-05, "loss": 0.0236, "step": 16090 }, { "epoch": 4.684317718940937, "grad_norm": 0.09500087052583694, "learning_rate": 2.625767800187487e-05, "loss": 0.0229, "step": 16100 }, { "epoch": 4.687227233052081, "grad_norm": 0.14061810076236725, "learning_rate": 2.6252212756971514e-05, "loss": 0.0235, "step": 16110 }, { "epoch": 4.690136747163224, "grad_norm": 0.10854833573102951, "learning_rate": 2.6246744093943022e-05, "loss": 0.0244, "step": 16120 }, { "epoch": 4.693046261274367, "grad_norm": 0.0929989144206047, "learning_rate": 2.6241272014450625e-05, "loss": 0.0235, "step": 16130 }, { "epoch": 4.69595577538551, "grad_norm": 0.13274160027503967, "learning_rate": 2.6235796520156603e-05, "loss": 0.0233, "step": 16140 }, { "epoch": 4.698865289496654, "grad_norm": 0.117539182305336, "learning_rate": 2.623031761272426e-05, "loss": 0.0234, "step": 16150 }, { "epoch": 4.7017748036077975, "grad_norm": 0.11598224192857742, "learning_rate": 2.6224835293817948e-05, "loss": 0.026, "step": 16160 }, { "epoch": 4.704684317718941, "grad_norm": 0.13489404320716858, "learning_rate": 2.6219349565103044e-05, "loss": 0.023, "step": 16170 }, { "epoch": 4.707593831830084, "grad_norm": 0.12100228667259216, "learning_rate": 2.6213860428245974e-05, "loss": 0.024, "step": 16180 }, { "epoch": 4.710503345941228, "grad_norm": 0.10805671662092209, "learning_rate": 2.620836788491419e-05, "loss": 0.023, "step": 16190 }, { "epoch": 4.7134128600523715, "grad_norm": 0.11745242029428482, "learning_rate": 2.6202871936776183e-05, "loss": 0.0246, "step": 16200 }, { "epoch": 4.716322374163514, "grad_norm": 0.14072878658771515, "learning_rate": 2.6197372585501477e-05, "loss": 0.0249, "step": 16210 }, { "epoch": 4.719231888274658, "grad_norm": 0.14129173755645752, "learning_rate": 2.6191869832760626e-05, "loss": 0.0259, "step": 16220 }, { "epoch": 4.722141402385802, "grad_norm": 0.11228679120540619, "learning_rate": 2.618636368022523e-05, "loss": 0.0241, "step": 16230 }, { "epoch": 4.725050916496945, "grad_norm": 0.10488570481538773, "learning_rate": 2.6180854129567902e-05, "loss": 0.0235, "step": 16240 }, { "epoch": 4.727960430608088, "grad_norm": 0.12238582968711853, "learning_rate": 2.6175341182462303e-05, "loss": 0.0244, "step": 16250 }, { "epoch": 4.730869944719232, "grad_norm": 0.09338736534118652, "learning_rate": 2.6169824840583124e-05, "loss": 0.0246, "step": 16260 }, { "epoch": 4.733779458830376, "grad_norm": 0.1517498642206192, "learning_rate": 2.6164305105606076e-05, "loss": 0.023, "step": 16270 }, { "epoch": 4.7366889729415185, "grad_norm": 0.08880232274532318, "learning_rate": 2.6158781979207918e-05, "loss": 0.0235, "step": 16280 }, { "epoch": 4.739598487052662, "grad_norm": 0.15238291025161743, "learning_rate": 2.615325546306642e-05, "loss": 0.0252, "step": 16290 }, { "epoch": 4.742508001163806, "grad_norm": 0.10570890456438065, "learning_rate": 2.61477255588604e-05, "loss": 0.0238, "step": 16300 }, { "epoch": 4.745417515274949, "grad_norm": 0.12834681570529938, "learning_rate": 2.614219226826969e-05, "loss": 0.0241, "step": 16310 }, { "epoch": 4.748327029386092, "grad_norm": 0.1398264318704605, "learning_rate": 2.6136655592975166e-05, "loss": 0.0223, "step": 16320 }, { "epoch": 4.751236543497236, "grad_norm": 0.1324148029088974, "learning_rate": 2.613111553465872e-05, "loss": 0.0254, "step": 16330 }, { "epoch": 4.754146057608379, "grad_norm": 0.10227368026971817, "learning_rate": 2.6125572095003276e-05, "loss": 0.0224, "step": 16340 }, { "epoch": 4.757055571719523, "grad_norm": 0.11849241703748703, "learning_rate": 2.6120025275692777e-05, "loss": 0.024, "step": 16350 }, { "epoch": 4.759965085830666, "grad_norm": 0.10472945868968964, "learning_rate": 2.6114475078412212e-05, "loss": 0.0232, "step": 16360 }, { "epoch": 4.76287459994181, "grad_norm": 0.10311471670866013, "learning_rate": 2.6108921504847575e-05, "loss": 0.0244, "step": 16370 }, { "epoch": 4.765784114052953, "grad_norm": 0.11233075708150864, "learning_rate": 2.6103364556685902e-05, "loss": 0.0226, "step": 16380 }, { "epoch": 4.768693628164097, "grad_norm": 0.12555596232414246, "learning_rate": 2.6097804235615242e-05, "loss": 0.0255, "step": 16390 }, { "epoch": 4.77160314227524, "grad_norm": 0.11746930330991745, "learning_rate": 2.6092240543324676e-05, "loss": 0.0235, "step": 16400 }, { "epoch": 4.774512656386383, "grad_norm": 0.12132561951875687, "learning_rate": 2.6086673481504303e-05, "loss": 0.0223, "step": 16410 }, { "epoch": 4.777422170497527, "grad_norm": 0.09909696131944656, "learning_rate": 2.608110305184526e-05, "loss": 0.0216, "step": 16420 }, { "epoch": 4.7803316846086705, "grad_norm": 0.09402447193861008, "learning_rate": 2.6075529256039687e-05, "loss": 0.0231, "step": 16430 }, { "epoch": 4.783241198719814, "grad_norm": 0.1464047133922577, "learning_rate": 2.6069952095780756e-05, "loss": 0.0244, "step": 16440 }, { "epoch": 4.786150712830957, "grad_norm": 0.09869758784770966, "learning_rate": 2.6064371572762667e-05, "loss": 0.0234, "step": 16450 }, { "epoch": 4.789060226942101, "grad_norm": 0.1172235906124115, "learning_rate": 2.605878768868063e-05, "loss": 0.0253, "step": 16460 }, { "epoch": 4.7919697410532445, "grad_norm": 0.10808856785297394, "learning_rate": 2.605320044523088e-05, "loss": 0.0223, "step": 16470 }, { "epoch": 4.794879255164387, "grad_norm": 0.10708021372556686, "learning_rate": 2.6047609844110683e-05, "loss": 0.0223, "step": 16480 }, { "epoch": 4.797788769275531, "grad_norm": 0.11365750432014465, "learning_rate": 2.604201588701831e-05, "loss": 0.0237, "step": 16490 }, { "epoch": 4.800698283386675, "grad_norm": 0.10638071596622467, "learning_rate": 2.6036418575653057e-05, "loss": 0.023, "step": 16500 }, { "epoch": 4.8036077974978175, "grad_norm": 0.10186021029949188, "learning_rate": 2.603081791171524e-05, "loss": 0.023, "step": 16510 }, { "epoch": 4.806517311608961, "grad_norm": 0.09727506339550018, "learning_rate": 2.602521389690619e-05, "loss": 0.022, "step": 16520 }, { "epoch": 4.809426825720105, "grad_norm": 0.11883401870727539, "learning_rate": 2.601960653292827e-05, "loss": 0.0229, "step": 16530 }, { "epoch": 4.812336339831248, "grad_norm": 0.0955570787191391, "learning_rate": 2.6013995821484833e-05, "loss": 0.0227, "step": 16540 }, { "epoch": 4.8152458539423915, "grad_norm": 0.0980711355805397, "learning_rate": 2.600838176428028e-05, "loss": 0.0238, "step": 16550 }, { "epoch": 4.818155368053535, "grad_norm": 0.13324925303459167, "learning_rate": 2.6002764363020004e-05, "loss": 0.0212, "step": 16560 }, { "epoch": 4.821064882164679, "grad_norm": 0.1273586004972458, "learning_rate": 2.5997143619410427e-05, "loss": 0.0226, "step": 16570 }, { "epoch": 4.823974396275822, "grad_norm": 0.09948864579200745, "learning_rate": 2.5991519535158976e-05, "loss": 0.0212, "step": 16580 }, { "epoch": 4.826883910386965, "grad_norm": 0.11843887716531754, "learning_rate": 2.598589211197411e-05, "loss": 0.0223, "step": 16590 }, { "epoch": 4.829793424498109, "grad_norm": 0.112971231341362, "learning_rate": 2.598026135156528e-05, "loss": 0.0231, "step": 16600 }, { "epoch": 4.832702938609252, "grad_norm": 0.10154061764478683, "learning_rate": 2.5974627255642974e-05, "loss": 0.0256, "step": 16610 }, { "epoch": 4.835612452720396, "grad_norm": 0.1058439239859581, "learning_rate": 2.5968989825918674e-05, "loss": 0.0235, "step": 16620 }, { "epoch": 4.838521966831539, "grad_norm": 0.11332020163536072, "learning_rate": 2.5963349064104887e-05, "loss": 0.0251, "step": 16630 }, { "epoch": 4.841431480942683, "grad_norm": 0.0951303094625473, "learning_rate": 2.5957704971915116e-05, "loss": 0.0207, "step": 16640 }, { "epoch": 4.844340995053826, "grad_norm": 0.09341806918382645, "learning_rate": 2.5952057551063905e-05, "loss": 0.0237, "step": 16650 }, { "epoch": 4.84725050916497, "grad_norm": 0.1426696479320526, "learning_rate": 2.594640680326678e-05, "loss": 0.024, "step": 16660 }, { "epoch": 4.850160023276112, "grad_norm": 0.14358799159526825, "learning_rate": 2.594075273024029e-05, "loss": 0.0231, "step": 16670 }, { "epoch": 4.853069537387256, "grad_norm": 0.11796397715806961, "learning_rate": 2.5935095333701994e-05, "loss": 0.0236, "step": 16680 }, { "epoch": 4.8559790514984, "grad_norm": 0.1402348428964615, "learning_rate": 2.5929434615370462e-05, "loss": 0.0227, "step": 16690 }, { "epoch": 4.8588885656095435, "grad_norm": 0.08074359595775604, "learning_rate": 2.592377057696527e-05, "loss": 0.0204, "step": 16700 }, { "epoch": 4.861798079720686, "grad_norm": 0.10314544290304184, "learning_rate": 2.5918103220206997e-05, "loss": 0.0224, "step": 16710 }, { "epoch": 4.86470759383183, "grad_norm": 0.11419067531824112, "learning_rate": 2.5912432546817247e-05, "loss": 0.0238, "step": 16720 }, { "epoch": 4.867617107942974, "grad_norm": 0.12157788127660751, "learning_rate": 2.5906758558518613e-05, "loss": 0.0246, "step": 16730 }, { "epoch": 4.870526622054117, "grad_norm": 0.14602088928222656, "learning_rate": 2.5901081257034706e-05, "loss": 0.0234, "step": 16740 }, { "epoch": 4.87343613616526, "grad_norm": 0.1312021017074585, "learning_rate": 2.5895400644090138e-05, "loss": 0.0239, "step": 16750 }, { "epoch": 4.876345650276404, "grad_norm": 0.11484456062316895, "learning_rate": 2.5889716721410535e-05, "loss": 0.0241, "step": 16760 }, { "epoch": 4.879255164387548, "grad_norm": 0.10238681733608246, "learning_rate": 2.5884029490722515e-05, "loss": 0.0246, "step": 16770 }, { "epoch": 4.8821646784986905, "grad_norm": 0.12901952862739563, "learning_rate": 2.587833895375371e-05, "loss": 0.0228, "step": 16780 }, { "epoch": 4.885074192609834, "grad_norm": 0.14987236261367798, "learning_rate": 2.587264511223276e-05, "loss": 0.023, "step": 16790 }, { "epoch": 4.887983706720978, "grad_norm": 0.11273784190416336, "learning_rate": 2.5866947967889295e-05, "loss": 0.0228, "step": 16800 }, { "epoch": 4.890893220832121, "grad_norm": 0.10874190926551819, "learning_rate": 2.586124752245397e-05, "loss": 0.023, "step": 16810 }, { "epoch": 4.8938027349432645, "grad_norm": 0.1384638547897339, "learning_rate": 2.585554377765842e-05, "loss": 0.0247, "step": 16820 }, { "epoch": 4.896712249054408, "grad_norm": 0.09790349006652832, "learning_rate": 2.584983673523529e-05, "loss": 0.0238, "step": 16830 }, { "epoch": 4.899621763165551, "grad_norm": 0.10293148458003998, "learning_rate": 2.5844126396918236e-05, "loss": 0.0235, "step": 16840 }, { "epoch": 4.902531277276695, "grad_norm": 0.13687148690223694, "learning_rate": 2.5838412764441905e-05, "loss": 0.0237, "step": 16850 }, { "epoch": 4.905440791387838, "grad_norm": 0.13764417171478271, "learning_rate": 2.583269583954195e-05, "loss": 0.0237, "step": 16860 }, { "epoch": 4.908350305498981, "grad_norm": 0.10363858938217163, "learning_rate": 2.5826975623955017e-05, "loss": 0.0232, "step": 16870 }, { "epoch": 4.911259819610125, "grad_norm": 0.09102467447519302, "learning_rate": 2.5821252119418762e-05, "loss": 0.0229, "step": 16880 }, { "epoch": 4.914169333721269, "grad_norm": 0.11551088839769363, "learning_rate": 2.581552532767183e-05, "loss": 0.0236, "step": 16890 }, { "epoch": 4.917078847832412, "grad_norm": 0.12942752242088318, "learning_rate": 2.580979525045387e-05, "loss": 0.0232, "step": 16900 }, { "epoch": 4.919988361943555, "grad_norm": 0.12019187957048416, "learning_rate": 2.580406188950553e-05, "loss": 0.0219, "step": 16910 }, { "epoch": 4.922897876054699, "grad_norm": 0.1059793159365654, "learning_rate": 2.5798325246568453e-05, "loss": 0.0228, "step": 16920 }, { "epoch": 4.925807390165843, "grad_norm": 0.11691360175609589, "learning_rate": 2.579258532338528e-05, "loss": 0.0233, "step": 16930 }, { "epoch": 4.928716904276985, "grad_norm": 0.12322987616062164, "learning_rate": 2.5786842121699644e-05, "loss": 0.0231, "step": 16940 }, { "epoch": 4.931626418388129, "grad_norm": 0.13790662586688995, "learning_rate": 2.5781095643256186e-05, "loss": 0.0261, "step": 16950 }, { "epoch": 4.934535932499273, "grad_norm": 0.12958472967147827, "learning_rate": 2.577534588980053e-05, "loss": 0.0215, "step": 16960 }, { "epoch": 4.9374454466104165, "grad_norm": 0.11769662797451019, "learning_rate": 2.5769592863079298e-05, "loss": 0.0225, "step": 16970 }, { "epoch": 4.940354960721559, "grad_norm": 0.12834292650222778, "learning_rate": 2.576383656484011e-05, "loss": 0.0248, "step": 16980 }, { "epoch": 4.943264474832703, "grad_norm": 0.1086018830537796, "learning_rate": 2.5758076996831572e-05, "loss": 0.0227, "step": 16990 }, { "epoch": 4.946173988943847, "grad_norm": 0.10665929317474365, "learning_rate": 2.5752314160803295e-05, "loss": 0.0218, "step": 17000 }, { "epoch": 4.94908350305499, "grad_norm": 0.12056791037321091, "learning_rate": 2.5746548058505873e-05, "loss": 0.0229, "step": 17010 }, { "epoch": 4.951993017166133, "grad_norm": 0.12055988609790802, "learning_rate": 2.57407786916909e-05, "loss": 0.0233, "step": 17020 }, { "epoch": 4.954902531277277, "grad_norm": 0.13528527319431305, "learning_rate": 2.573500606211095e-05, "loss": 0.0254, "step": 17030 }, { "epoch": 4.95781204538842, "grad_norm": 0.12316254526376724, "learning_rate": 2.57292301715196e-05, "loss": 0.0207, "step": 17040 }, { "epoch": 4.9607215594995635, "grad_norm": 0.09055925160646439, "learning_rate": 2.572345102167142e-05, "loss": 0.0229, "step": 17050 }, { "epoch": 4.963631073610707, "grad_norm": 0.13299013674259186, "learning_rate": 2.571766861432195e-05, "loss": 0.0227, "step": 17060 }, { "epoch": 4.96654058772185, "grad_norm": 0.12128204107284546, "learning_rate": 2.5711882951227738e-05, "loss": 0.023, "step": 17070 }, { "epoch": 4.969450101832994, "grad_norm": 0.09710124880075455, "learning_rate": 2.570609403414632e-05, "loss": 0.0243, "step": 17080 }, { "epoch": 4.9723596159441374, "grad_norm": 0.13376401364803314, "learning_rate": 2.5700301864836212e-05, "loss": 0.0233, "step": 17090 }, { "epoch": 4.975269130055281, "grad_norm": 0.0970648005604744, "learning_rate": 2.569450644505692e-05, "loss": 0.023, "step": 17100 }, { "epoch": 4.978178644166424, "grad_norm": 0.13333186507225037, "learning_rate": 2.5688707776568943e-05, "loss": 0.0235, "step": 17110 }, { "epoch": 4.981088158277568, "grad_norm": 0.1061614453792572, "learning_rate": 2.5682905861133764e-05, "loss": 0.0227, "step": 17120 }, { "epoch": 4.983997672388711, "grad_norm": 0.11227385699748993, "learning_rate": 2.567710070051385e-05, "loss": 0.0223, "step": 17130 }, { "epoch": 4.986907186499854, "grad_norm": 0.10608243942260742, "learning_rate": 2.5671292296472652e-05, "loss": 0.0243, "step": 17140 }, { "epoch": 4.989816700610998, "grad_norm": 0.11702915281057358, "learning_rate": 2.5665480650774616e-05, "loss": 0.0234, "step": 17150 }, { "epoch": 4.992726214722142, "grad_norm": 0.11741728335618973, "learning_rate": 2.565966576518517e-05, "loss": 0.0239, "step": 17160 }, { "epoch": 4.995635728833285, "grad_norm": 0.13534197211265564, "learning_rate": 2.5653847641470706e-05, "loss": 0.0262, "step": 17170 }, { "epoch": 4.998545242944428, "grad_norm": 0.137279212474823, "learning_rate": 2.564802628139863e-05, "loss": 0.0253, "step": 17180 }, { "epoch": 5.001454757055572, "grad_norm": 0.12472286820411682, "learning_rate": 2.5642201686737318e-05, "loss": 0.0216, "step": 17190 }, { "epoch": 5.004364271166716, "grad_norm": 0.10858817398548126, "learning_rate": 2.563637385925612e-05, "loss": 0.024, "step": 17200 }, { "epoch": 5.007273785277858, "grad_norm": 0.09950491786003113, "learning_rate": 2.5630542800725388e-05, "loss": 0.0235, "step": 17210 }, { "epoch": 5.010183299389002, "grad_norm": 0.11959999799728394, "learning_rate": 2.562470851291643e-05, "loss": 0.022, "step": 17220 }, { "epoch": 5.013092813500146, "grad_norm": 0.1284840852022171, "learning_rate": 2.561887099760156e-05, "loss": 0.0228, "step": 17230 }, { "epoch": 5.016002327611289, "grad_norm": 0.1281498372554779, "learning_rate": 2.5613030256554054e-05, "loss": 0.0236, "step": 17240 }, { "epoch": 5.018911841722432, "grad_norm": 0.1225350871682167, "learning_rate": 2.5607186291548174e-05, "loss": 0.0228, "step": 17250 }, { "epoch": 5.021821355833576, "grad_norm": 0.10205882042646408, "learning_rate": 2.5601339104359167e-05, "loss": 0.0253, "step": 17260 }, { "epoch": 5.024730869944719, "grad_norm": 0.11626902967691422, "learning_rate": 2.559548869676325e-05, "loss": 0.0245, "step": 17270 }, { "epoch": 5.0276403840558626, "grad_norm": 0.11376036703586578, "learning_rate": 2.558963507053763e-05, "loss": 0.023, "step": 17280 }, { "epoch": 5.030549898167006, "grad_norm": 0.11477601528167725, "learning_rate": 2.5583778227460476e-05, "loss": 0.0237, "step": 17290 }, { "epoch": 5.03345941227815, "grad_norm": 0.11458522826433182, "learning_rate": 2.5577918169310943e-05, "loss": 0.0234, "step": 17300 }, { "epoch": 5.036368926389293, "grad_norm": 0.1211061030626297, "learning_rate": 2.5572054897869167e-05, "loss": 0.0241, "step": 17310 }, { "epoch": 5.0392784405004365, "grad_norm": 0.1009163036942482, "learning_rate": 2.556618841491625e-05, "loss": 0.0213, "step": 17320 }, { "epoch": 5.04218795461158, "grad_norm": 0.09195863455533981, "learning_rate": 2.5560318722234276e-05, "loss": 0.0236, "step": 17330 }, { "epoch": 5.045097468722723, "grad_norm": 0.11850956082344055, "learning_rate": 2.5554445821606306e-05, "loss": 0.0219, "step": 17340 }, { "epoch": 5.048006982833867, "grad_norm": 0.10593444108963013, "learning_rate": 2.554856971481637e-05, "loss": 0.0212, "step": 17350 }, { "epoch": 5.05091649694501, "grad_norm": 0.11250682175159454, "learning_rate": 2.554269040364947e-05, "loss": 0.0218, "step": 17360 }, { "epoch": 5.053826011056153, "grad_norm": 0.12421586364507675, "learning_rate": 2.553680788989159e-05, "loss": 0.024, "step": 17370 }, { "epoch": 5.056735525167297, "grad_norm": 0.13303717970848083, "learning_rate": 2.5530922175329677e-05, "loss": 0.024, "step": 17380 }, { "epoch": 5.059645039278441, "grad_norm": 0.14700360596179962, "learning_rate": 2.5525033261751667e-05, "loss": 0.0236, "step": 17390 }, { "epoch": 5.0625545533895835, "grad_norm": 0.09999045729637146, "learning_rate": 2.5519141150946445e-05, "loss": 0.0235, "step": 17400 }, { "epoch": 5.065464067500727, "grad_norm": 0.10118665546178818, "learning_rate": 2.551324584470389e-05, "loss": 0.0234, "step": 17410 }, { "epoch": 5.068373581611871, "grad_norm": 0.1135183647274971, "learning_rate": 2.5507347344814825e-05, "loss": 0.0222, "step": 17420 }, { "epoch": 5.071283095723015, "grad_norm": 0.1404920518398285, "learning_rate": 2.5501445653071072e-05, "loss": 0.0216, "step": 17430 }, { "epoch": 5.074192609834157, "grad_norm": 0.12492448836565018, "learning_rate": 2.5495540771265406e-05, "loss": 0.0223, "step": 17440 }, { "epoch": 5.077102123945301, "grad_norm": 0.09294672310352325, "learning_rate": 2.5489632701191565e-05, "loss": 0.0231, "step": 17450 }, { "epoch": 5.080011638056445, "grad_norm": 0.19052383303642273, "learning_rate": 2.5483721444644276e-05, "loss": 0.0227, "step": 17460 }, { "epoch": 5.082921152167588, "grad_norm": 0.09998326748609543, "learning_rate": 2.5477807003419222e-05, "loss": 0.0248, "step": 17470 }, { "epoch": 5.085830666278731, "grad_norm": 0.0924016609787941, "learning_rate": 2.547188937931305e-05, "loss": 0.0225, "step": 17480 }, { "epoch": 5.088740180389875, "grad_norm": 0.12066630274057388, "learning_rate": 2.5465968574123372e-05, "loss": 0.0231, "step": 17490 }, { "epoch": 5.091649694501018, "grad_norm": 0.08731024712324142, "learning_rate": 2.5460044589648787e-05, "loss": 0.0225, "step": 17500 }, { "epoch": 5.094559208612162, "grad_norm": 0.11754007637500763, "learning_rate": 2.5454117427688833e-05, "loss": 0.0222, "step": 17510 }, { "epoch": 5.097468722723305, "grad_norm": 0.08964215219020844, "learning_rate": 2.5448187090044033e-05, "loss": 0.0231, "step": 17520 }, { "epoch": 5.100378236834449, "grad_norm": 0.10676191002130508, "learning_rate": 2.544225357851586e-05, "loss": 0.0233, "step": 17530 }, { "epoch": 5.103287750945592, "grad_norm": 0.11840490251779556, "learning_rate": 2.543631689490676e-05, "loss": 0.0229, "step": 17540 }, { "epoch": 5.1061972650567355, "grad_norm": 0.11481478065252304, "learning_rate": 2.543037704102015e-05, "loss": 0.0221, "step": 17550 }, { "epoch": 5.109106779167879, "grad_norm": 0.1325782686471939, "learning_rate": 2.5424434018660386e-05, "loss": 0.0235, "step": 17560 }, { "epoch": 5.112016293279022, "grad_norm": 0.1287343055009842, "learning_rate": 2.541848782963281e-05, "loss": 0.0209, "step": 17570 }, { "epoch": 5.114925807390166, "grad_norm": 0.09805801510810852, "learning_rate": 2.5412538475743714e-05, "loss": 0.0204, "step": 17580 }, { "epoch": 5.1178353215013095, "grad_norm": 0.09005285054445267, "learning_rate": 2.540658595880036e-05, "loss": 0.0206, "step": 17590 }, { "epoch": 5.120744835612452, "grad_norm": 0.13665364682674408, "learning_rate": 2.540063028061096e-05, "loss": 0.021, "step": 17600 }, { "epoch": 5.123654349723596, "grad_norm": 0.10920724272727966, "learning_rate": 2.5394671442984692e-05, "loss": 0.0234, "step": 17610 }, { "epoch": 5.12656386383474, "grad_norm": 0.11692824959754944, "learning_rate": 2.5388709447731696e-05, "loss": 0.0228, "step": 17620 }, { "epoch": 5.129473377945883, "grad_norm": 0.09697020053863525, "learning_rate": 2.5382744296663068e-05, "loss": 0.0201, "step": 17630 }, { "epoch": 5.132382892057026, "grad_norm": 0.1316155344247818, "learning_rate": 2.5376775991590865e-05, "loss": 0.0208, "step": 17640 }, { "epoch": 5.13529240616817, "grad_norm": 0.10532210022211075, "learning_rate": 2.5370804534328097e-05, "loss": 0.0231, "step": 17650 }, { "epoch": 5.138201920279314, "grad_norm": 0.1068422719836235, "learning_rate": 2.5364829926688736e-05, "loss": 0.0234, "step": 17660 }, { "epoch": 5.1411114343904565, "grad_norm": 0.11350494623184204, "learning_rate": 2.5358852170487713e-05, "loss": 0.0231, "step": 17670 }, { "epoch": 5.1440209485016, "grad_norm": 0.08918315917253494, "learning_rate": 2.5352871267540906e-05, "loss": 0.0219, "step": 17680 }, { "epoch": 5.146930462612744, "grad_norm": 0.10094509273767471, "learning_rate": 2.534688721966516e-05, "loss": 0.0229, "step": 17690 }, { "epoch": 5.149839976723887, "grad_norm": 0.1044062003493309, "learning_rate": 2.5340900028678276e-05, "loss": 0.0223, "step": 17700 }, { "epoch": 5.15274949083503, "grad_norm": 0.10863516479730606, "learning_rate": 2.5334909696398998e-05, "loss": 0.0239, "step": 17710 }, { "epoch": 5.155659004946174, "grad_norm": 0.1080537810921669, "learning_rate": 2.5328916224647027e-05, "loss": 0.0221, "step": 17720 }, { "epoch": 5.158568519057318, "grad_norm": 0.10965254157781601, "learning_rate": 2.532291961524303e-05, "loss": 0.0226, "step": 17730 }, { "epoch": 5.161478033168461, "grad_norm": 0.11571719497442245, "learning_rate": 2.531691987000861e-05, "loss": 0.0207, "step": 17740 }, { "epoch": 5.164387547279604, "grad_norm": 0.11477227509021759, "learning_rate": 2.5310916990766343e-05, "loss": 0.022, "step": 17750 }, { "epoch": 5.167297061390748, "grad_norm": 0.1298733949661255, "learning_rate": 2.5304910979339734e-05, "loss": 0.0225, "step": 17760 }, { "epoch": 5.170206575501891, "grad_norm": 0.10531269758939743, "learning_rate": 2.5298901837553255e-05, "loss": 0.024, "step": 17770 }, { "epoch": 5.173116089613035, "grad_norm": 0.09876066446304321, "learning_rate": 2.5292889567232326e-05, "loss": 0.0208, "step": 17780 }, { "epoch": 5.176025603724178, "grad_norm": 0.10749529302120209, "learning_rate": 2.528687417020331e-05, "loss": 0.0247, "step": 17790 }, { "epoch": 5.178935117835321, "grad_norm": 0.10950575768947601, "learning_rate": 2.5280855648293536e-05, "loss": 0.0241, "step": 17800 }, { "epoch": 5.181844631946465, "grad_norm": 0.12738098204135895, "learning_rate": 2.5274834003331266e-05, "loss": 0.0228, "step": 17810 }, { "epoch": 5.1847541460576085, "grad_norm": 0.13477015495300293, "learning_rate": 2.5268809237145717e-05, "loss": 0.0224, "step": 17820 }, { "epoch": 5.187663660168752, "grad_norm": 0.14440742135047913, "learning_rate": 2.5262781351567052e-05, "loss": 0.0218, "step": 17830 }, { "epoch": 5.190573174279895, "grad_norm": 0.1141875609755516, "learning_rate": 2.525675034842638e-05, "loss": 0.0248, "step": 17840 }, { "epoch": 5.193482688391039, "grad_norm": 0.12209905683994293, "learning_rate": 2.5250716229555774e-05, "loss": 0.0218, "step": 17850 }, { "epoch": 5.1963922025021825, "grad_norm": 0.12666279077529907, "learning_rate": 2.524467899678823e-05, "loss": 0.0214, "step": 17860 }, { "epoch": 5.199301716613325, "grad_norm": 0.10332920402288437, "learning_rate": 2.52386386519577e-05, "loss": 0.0225, "step": 17870 }, { "epoch": 5.202211230724469, "grad_norm": 0.11479515582323074, "learning_rate": 2.5232595196899086e-05, "loss": 0.0211, "step": 17880 }, { "epoch": 5.205120744835613, "grad_norm": 0.10716737806797028, "learning_rate": 2.5226548633448224e-05, "loss": 0.0219, "step": 17890 }, { "epoch": 5.2080302589467555, "grad_norm": 0.09877078980207443, "learning_rate": 2.5220498963441906e-05, "loss": 0.0206, "step": 17900 }, { "epoch": 5.210939773057899, "grad_norm": 0.11911492049694061, "learning_rate": 2.521444618871786e-05, "loss": 0.0227, "step": 17910 }, { "epoch": 5.213849287169043, "grad_norm": 0.09329527616500854, "learning_rate": 2.5208390311114758e-05, "loss": 0.0212, "step": 17920 }, { "epoch": 5.216758801280186, "grad_norm": 0.12297383695840836, "learning_rate": 2.520233133247221e-05, "loss": 0.0221, "step": 17930 }, { "epoch": 5.2196683153913295, "grad_norm": 0.14302369952201843, "learning_rate": 2.519626925463079e-05, "loss": 0.0228, "step": 17940 }, { "epoch": 5.222577829502473, "grad_norm": 0.09925561398267746, "learning_rate": 2.519020407943198e-05, "loss": 0.0212, "step": 17950 }, { "epoch": 5.225487343613617, "grad_norm": 0.11279819160699844, "learning_rate": 2.518413580871823e-05, "loss": 0.0218, "step": 17960 }, { "epoch": 5.22839685772476, "grad_norm": 0.10017169266939163, "learning_rate": 2.5178064444332922e-05, "loss": 0.0209, "step": 17970 }, { "epoch": 5.231306371835903, "grad_norm": 0.1183178499341011, "learning_rate": 2.5171989988120368e-05, "loss": 0.0225, "step": 17980 }, { "epoch": 5.234215885947047, "grad_norm": 0.1355547457933426, "learning_rate": 2.5165912441925832e-05, "loss": 0.0233, "step": 17990 }, { "epoch": 5.23712540005819, "grad_norm": 0.10517433285713196, "learning_rate": 2.515983180759551e-05, "loss": 0.0239, "step": 18000 }, { "epoch": 5.240034914169334, "grad_norm": 0.12211483716964722, "learning_rate": 2.515374808697654e-05, "loss": 0.0225, "step": 18010 }, { "epoch": 5.242944428280477, "grad_norm": 0.12246674299240112, "learning_rate": 2.5147661281916996e-05, "loss": 0.0224, "step": 18020 }, { "epoch": 5.24585394239162, "grad_norm": 0.1592627465724945, "learning_rate": 2.5141571394265892e-05, "loss": 0.0223, "step": 18030 }, { "epoch": 5.248763456502764, "grad_norm": 0.11148598790168762, "learning_rate": 2.513547842587317e-05, "loss": 0.0217, "step": 18040 }, { "epoch": 5.251672970613908, "grad_norm": 0.09984174370765686, "learning_rate": 2.5129382378589708e-05, "loss": 0.0219, "step": 18050 }, { "epoch": 5.254582484725051, "grad_norm": 0.10041368007659912, "learning_rate": 2.512328325426733e-05, "loss": 0.0225, "step": 18060 }, { "epoch": 5.257491998836194, "grad_norm": 0.09902539849281311, "learning_rate": 2.5117181054758798e-05, "loss": 0.0205, "step": 18070 }, { "epoch": 5.260401512947338, "grad_norm": 0.10929208993911743, "learning_rate": 2.5111075781917783e-05, "loss": 0.0205, "step": 18080 }, { "epoch": 5.2633110270584815, "grad_norm": 0.12558868527412415, "learning_rate": 2.510496743759892e-05, "loss": 0.0215, "step": 18090 }, { "epoch": 5.266220541169624, "grad_norm": 0.1219436451792717, "learning_rate": 2.509885602365775e-05, "loss": 0.0213, "step": 18100 }, { "epoch": 5.269130055280768, "grad_norm": 0.11566095799207687, "learning_rate": 2.509274154195076e-05, "loss": 0.0223, "step": 18110 }, { "epoch": 5.272039569391912, "grad_norm": 0.10054291039705276, "learning_rate": 2.5086623994335383e-05, "loss": 0.0229, "step": 18120 }, { "epoch": 5.274949083503055, "grad_norm": 0.10469702631235123, "learning_rate": 2.5080503382669953e-05, "loss": 0.0231, "step": 18130 }, { "epoch": 5.277858597614198, "grad_norm": 0.09474682062864304, "learning_rate": 2.507437970881376e-05, "loss": 0.0233, "step": 18140 }, { "epoch": 5.280768111725342, "grad_norm": 0.1289101243019104, "learning_rate": 2.5068252974627003e-05, "loss": 0.0225, "step": 18150 }, { "epoch": 5.283677625836486, "grad_norm": 0.118758924305439, "learning_rate": 2.5062123181970834e-05, "loss": 0.0227, "step": 18160 }, { "epoch": 5.2865871399476285, "grad_norm": 0.09184300154447556, "learning_rate": 2.5055990332707316e-05, "loss": 0.0221, "step": 18170 }, { "epoch": 5.289496654058772, "grad_norm": 0.11647159606218338, "learning_rate": 2.5049854428699444e-05, "loss": 0.0209, "step": 18180 }, { "epoch": 5.292406168169916, "grad_norm": 0.08780943602323532, "learning_rate": 2.5043715471811158e-05, "loss": 0.0222, "step": 18190 }, { "epoch": 5.295315682281059, "grad_norm": 0.13437910377979279, "learning_rate": 2.5037573463907296e-05, "loss": 0.0237, "step": 18200 }, { "epoch": 5.2982251963922025, "grad_norm": 0.1367574781179428, "learning_rate": 2.5031428406853637e-05, "loss": 0.0239, "step": 18210 }, { "epoch": 5.301134710503346, "grad_norm": 0.12975043058395386, "learning_rate": 2.5025280302516897e-05, "loss": 0.0213, "step": 18220 }, { "epoch": 5.304044224614489, "grad_norm": 0.11579491943120956, "learning_rate": 2.5019129152764698e-05, "loss": 0.0232, "step": 18230 }, { "epoch": 5.306953738725633, "grad_norm": 0.10027477145195007, "learning_rate": 2.5012974959465612e-05, "loss": 0.0212, "step": 18240 }, { "epoch": 5.309863252836776, "grad_norm": 0.13028517365455627, "learning_rate": 2.5006817724489105e-05, "loss": 0.0237, "step": 18250 }, { "epoch": 5.31277276694792, "grad_norm": 0.11037512868642807, "learning_rate": 2.500065744970559e-05, "loss": 0.0222, "step": 18260 }, { "epoch": 5.315682281059063, "grad_norm": 0.10128932446241379, "learning_rate": 2.499449413698639e-05, "loss": 0.0237, "step": 18270 }, { "epoch": 5.318591795170207, "grad_norm": 0.12954263389110565, "learning_rate": 2.4988327788203764e-05, "loss": 0.0226, "step": 18280 }, { "epoch": 5.32150130928135, "grad_norm": 0.12659166753292084, "learning_rate": 2.498215840523088e-05, "loss": 0.0217, "step": 18290 }, { "epoch": 5.324410823392493, "grad_norm": 0.0972171276807785, "learning_rate": 2.4975985989941837e-05, "loss": 0.0227, "step": 18300 }, { "epoch": 5.327320337503637, "grad_norm": 0.11303168535232544, "learning_rate": 2.4969810544211652e-05, "loss": 0.0201, "step": 18310 }, { "epoch": 5.330229851614781, "grad_norm": 0.11963445693254471, "learning_rate": 2.4963632069916258e-05, "loss": 0.0261, "step": 18320 }, { "epoch": 5.333139365725923, "grad_norm": 0.1006234660744667, "learning_rate": 2.495745056893252e-05, "loss": 0.0224, "step": 18330 }, { "epoch": 5.336048879837067, "grad_norm": 0.12198198586702347, "learning_rate": 2.495126604313821e-05, "loss": 0.0231, "step": 18340 }, { "epoch": 5.338958393948211, "grad_norm": 0.09884251654148102, "learning_rate": 2.4945078494412023e-05, "loss": 0.0202, "step": 18350 }, { "epoch": 5.3418679080593545, "grad_norm": 0.11003100126981735, "learning_rate": 2.493888792463357e-05, "loss": 0.0234, "step": 18360 }, { "epoch": 5.344777422170497, "grad_norm": 0.10472878813743591, "learning_rate": 2.4932694335683395e-05, "loss": 0.0224, "step": 18370 }, { "epoch": 5.347686936281641, "grad_norm": 0.10448338091373444, "learning_rate": 2.4926497729442932e-05, "loss": 0.0212, "step": 18380 }, { "epoch": 5.350596450392785, "grad_norm": 0.11575742065906525, "learning_rate": 2.4920298107794555e-05, "loss": 0.0219, "step": 18390 }, { "epoch": 5.353505964503928, "grad_norm": 0.10967526584863663, "learning_rate": 2.4914095472621544e-05, "loss": 0.0216, "step": 18400 }, { "epoch": 5.356415478615071, "grad_norm": 0.12339939922094345, "learning_rate": 2.4907889825808093e-05, "loss": 0.0232, "step": 18410 }, { "epoch": 5.359324992726215, "grad_norm": 0.11566894501447678, "learning_rate": 2.4901681169239314e-05, "loss": 0.0216, "step": 18420 }, { "epoch": 5.362234506837358, "grad_norm": 0.09581127017736435, "learning_rate": 2.4895469504801236e-05, "loss": 0.0218, "step": 18430 }, { "epoch": 5.3651440209485015, "grad_norm": 0.10950949788093567, "learning_rate": 2.4889254834380802e-05, "loss": 0.0229, "step": 18440 }, { "epoch": 5.368053535059645, "grad_norm": 0.09732433408498764, "learning_rate": 2.488303715986585e-05, "loss": 0.022, "step": 18450 }, { "epoch": 5.370963049170788, "grad_norm": 0.10392498224973679, "learning_rate": 2.4876816483145166e-05, "loss": 0.0218, "step": 18460 }, { "epoch": 5.373872563281932, "grad_norm": 0.10491091758012772, "learning_rate": 2.487059280610841e-05, "loss": 0.0217, "step": 18470 }, { "epoch": 5.3767820773930755, "grad_norm": 0.12266457080841064, "learning_rate": 2.4864366130646178e-05, "loss": 0.0225, "step": 18480 }, { "epoch": 5.379691591504219, "grad_norm": 0.10814981907606125, "learning_rate": 2.485813645864997e-05, "loss": 0.0222, "step": 18490 }, { "epoch": 5.382601105615362, "grad_norm": 0.1265314370393753, "learning_rate": 2.4851903792012198e-05, "loss": 0.0207, "step": 18500 }, { "epoch": 5.385510619726506, "grad_norm": 0.10011819005012512, "learning_rate": 2.4845668132626173e-05, "loss": 0.0213, "step": 18510 }, { "epoch": 5.388420133837649, "grad_norm": 0.13495488464832306, "learning_rate": 2.483942948238613e-05, "loss": 0.0225, "step": 18520 }, { "epoch": 5.391329647948792, "grad_norm": 0.1309993714094162, "learning_rate": 2.4833187843187207e-05, "loss": 0.0237, "step": 18530 }, { "epoch": 5.394239162059936, "grad_norm": 0.10196097195148468, "learning_rate": 2.4826943216925448e-05, "loss": 0.0242, "step": 18540 }, { "epoch": 5.39714867617108, "grad_norm": 0.1291700005531311, "learning_rate": 2.4820695605497807e-05, "loss": 0.0216, "step": 18550 }, { "epoch": 5.400058190282223, "grad_norm": 0.13126219809055328, "learning_rate": 2.481444501080214e-05, "loss": 0.0211, "step": 18560 }, { "epoch": 5.402967704393366, "grad_norm": 0.11050731688737869, "learning_rate": 2.4808191434737217e-05, "loss": 0.0232, "step": 18570 }, { "epoch": 5.40587721850451, "grad_norm": 0.09819452464580536, "learning_rate": 2.4801934879202696e-05, "loss": 0.0208, "step": 18580 }, { "epoch": 5.408786732615654, "grad_norm": 0.14122559130191803, "learning_rate": 2.4795675346099172e-05, "loss": 0.0229, "step": 18590 }, { "epoch": 5.411696246726796, "grad_norm": 0.11032779514789581, "learning_rate": 2.4789412837328114e-05, "loss": 0.0231, "step": 18600 }, { "epoch": 5.41460576083794, "grad_norm": 0.10477471351623535, "learning_rate": 2.4783147354791915e-05, "loss": 0.0212, "step": 18610 }, { "epoch": 5.417515274949084, "grad_norm": 0.11098092049360275, "learning_rate": 2.4776878900393858e-05, "loss": 0.0216, "step": 18620 }, { "epoch": 5.420424789060227, "grad_norm": 0.10923383384943008, "learning_rate": 2.477060747603813e-05, "loss": 0.022, "step": 18630 }, { "epoch": 5.42333430317137, "grad_norm": 0.11181803792715073, "learning_rate": 2.4764333083629833e-05, "loss": 0.0228, "step": 18640 }, { "epoch": 5.426243817282514, "grad_norm": 0.10138531774282455, "learning_rate": 2.4758055725074954e-05, "loss": 0.0235, "step": 18650 }, { "epoch": 5.429153331393657, "grad_norm": 0.11970719695091248, "learning_rate": 2.4751775402280396e-05, "loss": 0.0215, "step": 18660 }, { "epoch": 5.432062845504801, "grad_norm": 0.09340296685695648, "learning_rate": 2.4745492117153947e-05, "loss": 0.0212, "step": 18670 }, { "epoch": 5.434972359615944, "grad_norm": 0.1021743193268776, "learning_rate": 2.473920587160431e-05, "loss": 0.0224, "step": 18680 }, { "epoch": 5.437881873727088, "grad_norm": 0.1193508580327034, "learning_rate": 2.473291666754108e-05, "loss": 0.0207, "step": 18690 }, { "epoch": 5.440791387838231, "grad_norm": 0.08901523053646088, "learning_rate": 2.4726624506874748e-05, "loss": 0.0213, "step": 18700 }, { "epoch": 5.4437009019493745, "grad_norm": 0.09710301458835602, "learning_rate": 2.4720329391516708e-05, "loss": 0.0218, "step": 18710 }, { "epoch": 5.446610416060518, "grad_norm": 0.10067304223775864, "learning_rate": 2.4714031323379248e-05, "loss": 0.0215, "step": 18720 }, { "epoch": 5.449519930171661, "grad_norm": 0.10819956660270691, "learning_rate": 2.4707730304375556e-05, "loss": 0.024, "step": 18730 }, { "epoch": 5.452429444282805, "grad_norm": 0.10419230163097382, "learning_rate": 2.4701426336419713e-05, "loss": 0.0225, "step": 18740 }, { "epoch": 5.4553389583939484, "grad_norm": 0.1245461031794548, "learning_rate": 2.4695119421426707e-05, "loss": 0.0229, "step": 18750 }, { "epoch": 5.458248472505091, "grad_norm": 0.15667939186096191, "learning_rate": 2.46888095613124e-05, "loss": 0.0248, "step": 18760 }, { "epoch": 5.461157986616235, "grad_norm": 0.12210971117019653, "learning_rate": 2.468249675799357e-05, "loss": 0.0242, "step": 18770 }, { "epoch": 5.464067500727379, "grad_norm": 0.12055803835391998, "learning_rate": 2.467618101338787e-05, "loss": 0.0217, "step": 18780 }, { "epoch": 5.4669770148385215, "grad_norm": 0.1551360785961151, "learning_rate": 2.466986232941387e-05, "loss": 0.0238, "step": 18790 }, { "epoch": 5.469886528949665, "grad_norm": 0.10157369822263718, "learning_rate": 2.466354070799101e-05, "loss": 0.023, "step": 18800 }, { "epoch": 5.472796043060809, "grad_norm": 0.10720960050821304, "learning_rate": 2.4657216151039634e-05, "loss": 0.0227, "step": 18810 }, { "epoch": 5.475705557171953, "grad_norm": 0.10409092158079147, "learning_rate": 2.4650888660480976e-05, "loss": 0.022, "step": 18820 }, { "epoch": 5.478615071283095, "grad_norm": 0.13718651235103607, "learning_rate": 2.464455823823716e-05, "loss": 0.0224, "step": 18830 }, { "epoch": 5.481524585394239, "grad_norm": 0.32791173458099365, "learning_rate": 2.4638224886231196e-05, "loss": 0.0213, "step": 18840 }, { "epoch": 5.484434099505383, "grad_norm": 0.10290338844060898, "learning_rate": 2.4631888606387e-05, "loss": 0.0209, "step": 18850 }, { "epoch": 5.487343613616526, "grad_norm": 0.15345759689807892, "learning_rate": 2.4625549400629356e-05, "loss": 0.0228, "step": 18860 }, { "epoch": 5.490253127727669, "grad_norm": 0.10803928971290588, "learning_rate": 2.4619207270883958e-05, "loss": 0.0235, "step": 18870 }, { "epoch": 5.493162641838813, "grad_norm": 0.09784426540136337, "learning_rate": 2.461286221907737e-05, "loss": 0.0241, "step": 18880 }, { "epoch": 5.496072155949957, "grad_norm": 0.09362995624542236, "learning_rate": 2.460651424713705e-05, "loss": 0.0228, "step": 18890 }, { "epoch": 5.4989816700611, "grad_norm": 0.09733206033706665, "learning_rate": 2.4600163356991347e-05, "loss": 0.0208, "step": 18900 }, { "epoch": 5.501891184172243, "grad_norm": 0.11500963568687439, "learning_rate": 2.4593809550569498e-05, "loss": 0.0214, "step": 18910 }, { "epoch": 5.504800698283387, "grad_norm": 0.10689786076545715, "learning_rate": 2.4587452829801614e-05, "loss": 0.0225, "step": 18920 }, { "epoch": 5.50771021239453, "grad_norm": 0.10755900293588638, "learning_rate": 2.45810931966187e-05, "loss": 0.023, "step": 18930 }, { "epoch": 5.5106197265056736, "grad_norm": 0.12297265976667404, "learning_rate": 2.457473065295265e-05, "loss": 0.0238, "step": 18940 }, { "epoch": 5.513529240616817, "grad_norm": 0.11310580372810364, "learning_rate": 2.4568365200736232e-05, "loss": 0.022, "step": 18950 }, { "epoch": 5.51643875472796, "grad_norm": 0.10812629759311676, "learning_rate": 2.4561996841903104e-05, "loss": 0.0211, "step": 18960 }, { "epoch": 5.519348268839104, "grad_norm": 0.1094415932893753, "learning_rate": 2.455562557838781e-05, "loss": 0.025, "step": 18970 }, { "epoch": 5.5222577829502475, "grad_norm": 0.10732744634151459, "learning_rate": 2.4549251412125762e-05, "loss": 0.0214, "step": 18980 }, { "epoch": 5.52516729706139, "grad_norm": 0.1050729900598526, "learning_rate": 2.454287434505327e-05, "loss": 0.0219, "step": 18990 }, { "epoch": 5.528076811172534, "grad_norm": 0.12559734284877777, "learning_rate": 2.4536494379107514e-05, "loss": 0.0212, "step": 19000 }, { "epoch": 5.530986325283678, "grad_norm": 0.10116450488567352, "learning_rate": 2.453011151622657e-05, "loss": 0.0209, "step": 19010 }, { "epoch": 5.533895839394821, "grad_norm": 0.11806429177522659, "learning_rate": 2.452372575834937e-05, "loss": 0.0203, "step": 19020 }, { "epoch": 5.536805353505964, "grad_norm": 0.11008771508932114, "learning_rate": 2.4517337107415742e-05, "loss": 0.0207, "step": 19030 }, { "epoch": 5.539714867617108, "grad_norm": 0.1425953060388565, "learning_rate": 2.4510945565366397e-05, "loss": 0.0235, "step": 19040 }, { "epoch": 5.542624381728252, "grad_norm": 0.14673905074596405, "learning_rate": 2.4504551134142905e-05, "loss": 0.0211, "step": 19050 }, { "epoch": 5.5455338958393945, "grad_norm": 0.13407665491104126, "learning_rate": 2.4498153815687738e-05, "loss": 0.0219, "step": 19060 }, { "epoch": 5.548443409950538, "grad_norm": 0.12001321464776993, "learning_rate": 2.4491753611944224e-05, "loss": 0.0227, "step": 19070 }, { "epoch": 5.551352924061682, "grad_norm": 0.10617246478796005, "learning_rate": 2.4485350524856577e-05, "loss": 0.0226, "step": 19080 }, { "epoch": 5.554262438172826, "grad_norm": 0.12438152730464935, "learning_rate": 2.4478944556369886e-05, "loss": 0.0236, "step": 19090 }, { "epoch": 5.557171952283968, "grad_norm": 0.09816859662532806, "learning_rate": 2.4472535708430116e-05, "loss": 0.0216, "step": 19100 }, { "epoch": 5.560081466395112, "grad_norm": 0.11019152402877808, "learning_rate": 2.4466123982984103e-05, "loss": 0.0226, "step": 19110 }, { "epoch": 5.562990980506255, "grad_norm": 0.13255555927753448, "learning_rate": 2.445970938197957e-05, "loss": 0.0211, "step": 19120 }, { "epoch": 5.565900494617399, "grad_norm": 0.10383117198944092, "learning_rate": 2.4453291907365092e-05, "loss": 0.0201, "step": 19130 }, { "epoch": 5.568810008728542, "grad_norm": 0.10719206929206848, "learning_rate": 2.444687156109013e-05, "loss": 0.0223, "step": 19140 }, { "epoch": 5.571719522839686, "grad_norm": 0.1358424723148346, "learning_rate": 2.444044834510502e-05, "loss": 0.0224, "step": 19150 }, { "epoch": 5.574629036950829, "grad_norm": 0.10057475417852402, "learning_rate": 2.443402226136096e-05, "loss": 0.0211, "step": 19160 }, { "epoch": 5.577538551061973, "grad_norm": 0.10789661109447479, "learning_rate": 2.442759331181003e-05, "loss": 0.0215, "step": 19170 }, { "epoch": 5.580448065173116, "grad_norm": 0.10993610322475433, "learning_rate": 2.4421161498405176e-05, "loss": 0.0218, "step": 19180 }, { "epoch": 5.583357579284259, "grad_norm": 0.10448245704174042, "learning_rate": 2.4414726823100207e-05, "loss": 0.02, "step": 19190 }, { "epoch": 5.586267093395403, "grad_norm": 0.13635551929473877, "learning_rate": 2.4408289287849813e-05, "loss": 0.0219, "step": 19200 }, { "epoch": 5.5891766075065465, "grad_norm": 0.1214231625199318, "learning_rate": 2.4401848894609543e-05, "loss": 0.0206, "step": 19210 }, { "epoch": 5.59208612161769, "grad_norm": 0.09590977430343628, "learning_rate": 2.439540564533582e-05, "loss": 0.0219, "step": 19220 }, { "epoch": 5.594995635728833, "grad_norm": 0.10882826894521713, "learning_rate": 2.4388959541985935e-05, "loss": 0.0234, "step": 19230 }, { "epoch": 5.597905149839977, "grad_norm": 0.10367783904075623, "learning_rate": 2.438251058651804e-05, "loss": 0.021, "step": 19240 }, { "epoch": 5.6008146639511205, "grad_norm": 0.10162529349327087, "learning_rate": 2.437605878089116e-05, "loss": 0.0227, "step": 19250 }, { "epoch": 5.603724178062263, "grad_norm": 0.0825289711356163, "learning_rate": 2.4369604127065183e-05, "loss": 0.0214, "step": 19260 }, { "epoch": 5.606633692173407, "grad_norm": 0.09235069155693054, "learning_rate": 2.4363146627000862e-05, "loss": 0.0209, "step": 19270 }, { "epoch": 5.609543206284551, "grad_norm": 0.1202160120010376, "learning_rate": 2.435668628265982e-05, "loss": 0.0219, "step": 19280 }, { "epoch": 5.6124527203956935, "grad_norm": 0.10592128336429596, "learning_rate": 2.4350223096004533e-05, "loss": 0.0227, "step": 19290 }, { "epoch": 5.615362234506837, "grad_norm": 0.1099410355091095, "learning_rate": 2.4343757068998343e-05, "loss": 0.021, "step": 19300 }, { "epoch": 5.618271748617981, "grad_norm": 0.11937104165554047, "learning_rate": 2.4337288203605465e-05, "loss": 0.0243, "step": 19310 }, { "epoch": 5.621181262729124, "grad_norm": 0.13182665407657623, "learning_rate": 2.433081650179097e-05, "loss": 0.0197, "step": 19320 }, { "epoch": 5.6240907768402675, "grad_norm": 0.13434793055057526, "learning_rate": 2.4324341965520787e-05, "loss": 0.0225, "step": 19330 }, { "epoch": 5.627000290951411, "grad_norm": 0.1276320070028305, "learning_rate": 2.4317864596761706e-05, "loss": 0.0211, "step": 19340 }, { "epoch": 5.629909805062555, "grad_norm": 0.11137967556715012, "learning_rate": 2.4311384397481387e-05, "loss": 0.0196, "step": 19350 }, { "epoch": 5.632819319173698, "grad_norm": 0.07638765126466751, "learning_rate": 2.4304901369648344e-05, "loss": 0.0213, "step": 19360 }, { "epoch": 5.635728833284841, "grad_norm": 0.12820173799991608, "learning_rate": 2.429841551523194e-05, "loss": 0.0209, "step": 19370 }, { "epoch": 5.638638347395985, "grad_norm": 0.10267675668001175, "learning_rate": 2.4291926836202416e-05, "loss": 0.023, "step": 19380 }, { "epoch": 5.641547861507128, "grad_norm": 0.10062888264656067, "learning_rate": 2.428543533453086e-05, "loss": 0.0218, "step": 19390 }, { "epoch": 5.644457375618272, "grad_norm": 0.1165318638086319, "learning_rate": 2.4278941012189215e-05, "loss": 0.0203, "step": 19400 }, { "epoch": 5.647366889729415, "grad_norm": 0.10744383931159973, "learning_rate": 2.427244387115029e-05, "loss": 0.0215, "step": 19410 }, { "epoch": 5.650276403840559, "grad_norm": 0.11200229823589325, "learning_rate": 2.4265943913387738e-05, "loss": 0.0227, "step": 19420 }, { "epoch": 5.653185917951702, "grad_norm": 0.10568483918905258, "learning_rate": 2.425944114087608e-05, "loss": 0.0219, "step": 19430 }, { "epoch": 5.656095432062846, "grad_norm": 0.09231296181678772, "learning_rate": 2.4252935555590684e-05, "loss": 0.0218, "step": 19440 }, { "epoch": 5.659004946173989, "grad_norm": 0.11547335982322693, "learning_rate": 2.4246427159507772e-05, "loss": 0.0225, "step": 19450 }, { "epoch": 5.661914460285132, "grad_norm": 0.10239502042531967, "learning_rate": 2.4239915954604433e-05, "loss": 0.0201, "step": 19460 }, { "epoch": 5.664823974396276, "grad_norm": 0.09482552856206894, "learning_rate": 2.4233401942858595e-05, "loss": 0.0221, "step": 19470 }, { "epoch": 5.6677334885074195, "grad_norm": 0.089571014046669, "learning_rate": 2.4226885126249033e-05, "loss": 0.0206, "step": 19480 }, { "epoch": 5.670643002618562, "grad_norm": 0.09377489984035492, "learning_rate": 2.4220365506755394e-05, "loss": 0.0203, "step": 19490 }, { "epoch": 5.673552516729706, "grad_norm": 0.102546326816082, "learning_rate": 2.4213843086358166e-05, "loss": 0.0227, "step": 19500 }, { "epoch": 5.67646203084085, "grad_norm": 0.09388752281665802, "learning_rate": 2.4207317867038684e-05, "loss": 0.0201, "step": 19510 }, { "epoch": 5.679371544951993, "grad_norm": 0.10808558762073517, "learning_rate": 2.4200789850779137e-05, "loss": 0.0208, "step": 19520 }, { "epoch": 5.682281059063136, "grad_norm": 0.15512563288211823, "learning_rate": 2.419425903956257e-05, "loss": 0.0209, "step": 19530 }, { "epoch": 5.68519057317428, "grad_norm": 0.14795003831386566, "learning_rate": 2.4187725435372865e-05, "loss": 0.0207, "step": 19540 }, { "epoch": 5.688100087285424, "grad_norm": 0.07875760644674301, "learning_rate": 2.418118904019476e-05, "loss": 0.0208, "step": 19550 }, { "epoch": 5.6910096013965665, "grad_norm": 0.1265588253736496, "learning_rate": 2.4174649856013842e-05, "loss": 0.0212, "step": 19560 }, { "epoch": 5.69391911550771, "grad_norm": 0.08128771185874939, "learning_rate": 2.416810788481654e-05, "loss": 0.0197, "step": 19570 }, { "epoch": 5.696828629618854, "grad_norm": 0.09865407645702362, "learning_rate": 2.416156312859013e-05, "loss": 0.0212, "step": 19580 }, { "epoch": 5.699738143729997, "grad_norm": 0.10590136796236038, "learning_rate": 2.4155015589322742e-05, "loss": 0.022, "step": 19590 }, { "epoch": 5.7026476578411405, "grad_norm": 0.12336152046918869, "learning_rate": 2.4148465269003338e-05, "loss": 0.0208, "step": 19600 }, { "epoch": 5.705557171952284, "grad_norm": 0.12687188386917114, "learning_rate": 2.4141912169621742e-05, "loss": 0.0228, "step": 19610 }, { "epoch": 5.708466686063428, "grad_norm": 0.10035014152526855, "learning_rate": 2.4135356293168602e-05, "loss": 0.0212, "step": 19620 }, { "epoch": 5.711376200174571, "grad_norm": 0.1252957284450531, "learning_rate": 2.4128797641635427e-05, "loss": 0.0215, "step": 19630 }, { "epoch": 5.714285714285714, "grad_norm": 0.11611350625753403, "learning_rate": 2.412223621701456e-05, "loss": 0.0225, "step": 19640 }, { "epoch": 5.717195228396857, "grad_norm": 0.10469160974025726, "learning_rate": 2.4115672021299193e-05, "loss": 0.0206, "step": 19650 }, { "epoch": 5.720104742508001, "grad_norm": 0.12744568288326263, "learning_rate": 2.4109105056483345e-05, "loss": 0.0246, "step": 19660 }, { "epoch": 5.723014256619145, "grad_norm": 0.09618661552667618, "learning_rate": 2.4102535324561898e-05, "loss": 0.0215, "step": 19670 }, { "epoch": 5.725923770730288, "grad_norm": 0.10235495865345001, "learning_rate": 2.4095962827530556e-05, "loss": 0.0219, "step": 19680 }, { "epoch": 5.728833284841431, "grad_norm": 0.1345948874950409, "learning_rate": 2.408938756738587e-05, "loss": 0.0226, "step": 19690 }, { "epoch": 5.731742798952575, "grad_norm": 0.1066831648349762, "learning_rate": 2.4082809546125235e-05, "loss": 0.0218, "step": 19700 }, { "epoch": 5.734652313063719, "grad_norm": 0.10172862559556961, "learning_rate": 2.4076228765746876e-05, "loss": 0.0217, "step": 19710 }, { "epoch": 5.737561827174861, "grad_norm": 0.10184410214424133, "learning_rate": 2.4069645228249864e-05, "loss": 0.0219, "step": 19720 }, { "epoch": 5.740471341286005, "grad_norm": 0.10576258599758148, "learning_rate": 2.40630589356341e-05, "loss": 0.023, "step": 19730 }, { "epoch": 5.743380855397149, "grad_norm": 0.10303689539432526, "learning_rate": 2.4056469889900327e-05, "loss": 0.0211, "step": 19740 }, { "epoch": 5.7462903695082925, "grad_norm": 0.12038400024175644, "learning_rate": 2.404987809305012e-05, "loss": 0.0193, "step": 19750 }, { "epoch": 5.749199883619435, "grad_norm": 0.08597660064697266, "learning_rate": 2.4043283547085903e-05, "loss": 0.0214, "step": 19760 }, { "epoch": 5.752109397730579, "grad_norm": 0.12709520757198334, "learning_rate": 2.403668625401092e-05, "loss": 0.0214, "step": 19770 }, { "epoch": 5.755018911841723, "grad_norm": 0.09546703100204468, "learning_rate": 2.4030086215829247e-05, "loss": 0.0211, "step": 19780 }, { "epoch": 5.757928425952866, "grad_norm": 0.09822896122932434, "learning_rate": 2.4023483434545806e-05, "loss": 0.0238, "step": 19790 }, { "epoch": 5.760837940064009, "grad_norm": 0.10230816155672073, "learning_rate": 2.4016877912166356e-05, "loss": 0.0203, "step": 19800 }, { "epoch": 5.763747454175153, "grad_norm": 0.06920212507247925, "learning_rate": 2.401026965069747e-05, "loss": 0.022, "step": 19810 }, { "epoch": 5.766656968286296, "grad_norm": 0.1012442335486412, "learning_rate": 2.4003658652146564e-05, "loss": 0.022, "step": 19820 }, { "epoch": 5.7695664823974395, "grad_norm": 0.10575774312019348, "learning_rate": 2.3997044918521896e-05, "loss": 0.021, "step": 19830 }, { "epoch": 5.772475996508583, "grad_norm": 0.0988878682255745, "learning_rate": 2.3990428451832524e-05, "loss": 0.0207, "step": 19840 }, { "epoch": 5.775385510619726, "grad_norm": 0.13361051678657532, "learning_rate": 2.398380925408837e-05, "loss": 0.0228, "step": 19850 }, { "epoch": 5.77829502473087, "grad_norm": 0.10976801812648773, "learning_rate": 2.3977187327300174e-05, "loss": 0.0208, "step": 19860 }, { "epoch": 5.7812045388420135, "grad_norm": 0.1063256785273552, "learning_rate": 2.397056267347949e-05, "loss": 0.0215, "step": 19870 }, { "epoch": 5.784114052953157, "grad_norm": 0.14997762441635132, "learning_rate": 2.3963935294638725e-05, "loss": 0.0215, "step": 19880 }, { "epoch": 5.7870235670643, "grad_norm": 0.12962572276592255, "learning_rate": 2.395730519279109e-05, "loss": 0.0215, "step": 19890 }, { "epoch": 5.789933081175444, "grad_norm": 0.11693678796291351, "learning_rate": 2.3950672369950646e-05, "loss": 0.0224, "step": 19900 }, { "epoch": 5.792842595286587, "grad_norm": 0.1401456594467163, "learning_rate": 2.3944036828132266e-05, "loss": 0.0226, "step": 19910 }, { "epoch": 5.79575210939773, "grad_norm": 0.09929713606834412, "learning_rate": 2.3937398569351647e-05, "loss": 0.0219, "step": 19920 }, { "epoch": 5.798661623508874, "grad_norm": 0.0896625891327858, "learning_rate": 2.3930757595625326e-05, "loss": 0.0207, "step": 19930 }, { "epoch": 5.801571137620018, "grad_norm": 0.08721157908439636, "learning_rate": 2.3924113908970645e-05, "loss": 0.0201, "step": 19940 }, { "epoch": 5.804480651731161, "grad_norm": 0.14388063549995422, "learning_rate": 2.391746751140579e-05, "loss": 0.0217, "step": 19950 }, { "epoch": 5.807390165842304, "grad_norm": 0.139926016330719, "learning_rate": 2.3910818404949756e-05, "loss": 0.0222, "step": 19960 }, { "epoch": 5.810299679953448, "grad_norm": 0.09211308509111404, "learning_rate": 2.3904166591622367e-05, "loss": 0.0202, "step": 19970 }, { "epoch": 5.813209194064592, "grad_norm": 0.12001873552799225, "learning_rate": 2.3897512073444267e-05, "loss": 0.0207, "step": 19980 }, { "epoch": 5.816118708175734, "grad_norm": 0.11641892045736313, "learning_rate": 2.389085485243693e-05, "loss": 0.0212, "step": 19990 }, { "epoch": 5.819028222286878, "grad_norm": 0.10335983335971832, "learning_rate": 2.3884194930622632e-05, "loss": 0.0216, "step": 20000 }, { "epoch": 5.821937736398022, "grad_norm": 0.11729051172733307, "learning_rate": 2.3877532310024488e-05, "loss": 0.0224, "step": 20010 }, { "epoch": 5.824847250509165, "grad_norm": 0.13603191077709198, "learning_rate": 2.3870866992666428e-05, "loss": 0.023, "step": 20020 }, { "epoch": 5.827756764620308, "grad_norm": 0.11245884746313095, "learning_rate": 2.3864198980573196e-05, "loss": 0.0228, "step": 20030 }, { "epoch": 5.830666278731452, "grad_norm": 0.1188894659280777, "learning_rate": 2.385752827577036e-05, "loss": 0.0217, "step": 20040 }, { "epoch": 5.833575792842595, "grad_norm": 0.12821471691131592, "learning_rate": 2.38508548802843e-05, "loss": 0.0222, "step": 20050 }, { "epoch": 5.836485306953739, "grad_norm": 0.08568267524242401, "learning_rate": 2.3844178796142226e-05, "loss": 0.0213, "step": 20060 }, { "epoch": 5.839394821064882, "grad_norm": 0.1300572156906128, "learning_rate": 2.383750002537215e-05, "loss": 0.0202, "step": 20070 }, { "epoch": 5.842304335176026, "grad_norm": 0.1401849240064621, "learning_rate": 2.383081857000291e-05, "loss": 0.0215, "step": 20080 }, { "epoch": 5.845213849287169, "grad_norm": 0.1126788929104805, "learning_rate": 2.3824134432064145e-05, "loss": 0.0222, "step": 20090 }, { "epoch": 5.8481233633983125, "grad_norm": 0.09344836324453354, "learning_rate": 2.3817447613586332e-05, "loss": 0.0197, "step": 20100 }, { "epoch": 5.851032877509456, "grad_norm": 0.10900556296110153, "learning_rate": 2.381075811660074e-05, "loss": 0.0208, "step": 20110 }, { "epoch": 5.853942391620599, "grad_norm": 0.14074698090553284, "learning_rate": 2.3804065943139473e-05, "loss": 0.0205, "step": 20120 }, { "epoch": 5.856851905731743, "grad_norm": 0.1115136370062828, "learning_rate": 2.379737109523543e-05, "loss": 0.0236, "step": 20130 }, { "epoch": 5.8597614198428865, "grad_norm": 0.12694020569324493, "learning_rate": 2.3790673574922324e-05, "loss": 0.0209, "step": 20140 }, { "epoch": 5.86267093395403, "grad_norm": 0.12446530908346176, "learning_rate": 2.3783973384234695e-05, "loss": 0.0226, "step": 20150 }, { "epoch": 5.865580448065173, "grad_norm": 0.10285195708274841, "learning_rate": 2.3777270525207876e-05, "loss": 0.0221, "step": 20160 }, { "epoch": 5.868489962176317, "grad_norm": 0.11291944235563278, "learning_rate": 2.3770564999878024e-05, "loss": 0.0205, "step": 20170 }, { "epoch": 5.8713994762874595, "grad_norm": 0.13431550562381744, "learning_rate": 2.37638568102821e-05, "loss": 0.021, "step": 20180 }, { "epoch": 5.874308990398603, "grad_norm": 0.11405289173126221, "learning_rate": 2.3757145958457868e-05, "loss": 0.0212, "step": 20190 }, { "epoch": 5.877218504509747, "grad_norm": 0.1295577734708786, "learning_rate": 2.375043244644392e-05, "loss": 0.0203, "step": 20200 }, { "epoch": 5.880128018620891, "grad_norm": 0.09470809996128082, "learning_rate": 2.374371627627963e-05, "loss": 0.0207, "step": 20210 }, { "epoch": 5.8830375327320334, "grad_norm": 0.12260182946920395, "learning_rate": 2.37369974500052e-05, "loss": 0.0226, "step": 20220 }, { "epoch": 5.885947046843177, "grad_norm": 0.131659135222435, "learning_rate": 2.373027596966164e-05, "loss": 0.0207, "step": 20230 }, { "epoch": 5.888856560954321, "grad_norm": 0.10421070456504822, "learning_rate": 2.3723551837290744e-05, "loss": 0.0209, "step": 20240 }, { "epoch": 5.891766075065464, "grad_norm": 0.12435156852006912, "learning_rate": 2.3716825054935133e-05, "loss": 0.021, "step": 20250 }, { "epoch": 5.894675589176607, "grad_norm": 0.10925376415252686, "learning_rate": 2.371009562463822e-05, "loss": 0.021, "step": 20260 }, { "epoch": 5.897585103287751, "grad_norm": 0.11871691048145294, "learning_rate": 2.3703363548444238e-05, "loss": 0.022, "step": 20270 }, { "epoch": 5.900494617398895, "grad_norm": 0.08715866506099701, "learning_rate": 2.3696628828398208e-05, "loss": 0.0212, "step": 20280 }, { "epoch": 5.903404131510038, "grad_norm": 0.13650278747081757, "learning_rate": 2.368989146654596e-05, "loss": 0.0217, "step": 20290 }, { "epoch": 5.906313645621181, "grad_norm": 0.10350251197814941, "learning_rate": 2.368315146493413e-05, "loss": 0.0207, "step": 20300 }, { "epoch": 5.909223159732325, "grad_norm": 0.08650841563940048, "learning_rate": 2.367640882561015e-05, "loss": 0.0227, "step": 20310 }, { "epoch": 5.912132673843468, "grad_norm": 0.11526601016521454, "learning_rate": 2.3669663550622255e-05, "loss": 0.022, "step": 20320 }, { "epoch": 5.915042187954612, "grad_norm": 0.13224925100803375, "learning_rate": 2.3662915642019482e-05, "loss": 0.0212, "step": 20330 }, { "epoch": 5.917951702065755, "grad_norm": 0.12140834331512451, "learning_rate": 2.365616510185167e-05, "loss": 0.0213, "step": 20340 }, { "epoch": 5.920861216176898, "grad_norm": 0.12528009712696075, "learning_rate": 2.364941193216945e-05, "loss": 0.021, "step": 20350 }, { "epoch": 5.923770730288042, "grad_norm": 0.1132027730345726, "learning_rate": 2.364265613502426e-05, "loss": 0.0232, "step": 20360 }, { "epoch": 5.9266802443991855, "grad_norm": 0.11376345902681351, "learning_rate": 2.3635897712468333e-05, "loss": 0.0221, "step": 20370 }, { "epoch": 5.929589758510328, "grad_norm": 0.1109917163848877, "learning_rate": 2.36291366665547e-05, "loss": 0.0214, "step": 20380 }, { "epoch": 5.932499272621472, "grad_norm": 0.10381016135215759, "learning_rate": 2.362237299933719e-05, "loss": 0.0227, "step": 20390 }, { "epoch": 5.935408786732616, "grad_norm": 0.11122079193592072, "learning_rate": 2.3615606712870414e-05, "loss": 0.0204, "step": 20400 }, { "epoch": 5.9383183008437594, "grad_norm": 0.11776449531316757, "learning_rate": 2.3608837809209807e-05, "loss": 0.0217, "step": 20410 }, { "epoch": 5.941227814954902, "grad_norm": 0.11827163398265839, "learning_rate": 2.360206629041157e-05, "loss": 0.0215, "step": 20420 }, { "epoch": 5.944137329066046, "grad_norm": 0.10690202564001083, "learning_rate": 2.359529215853273e-05, "loss": 0.0201, "step": 20430 }, { "epoch": 5.94704684317719, "grad_norm": 0.1325589120388031, "learning_rate": 2.358851541563107e-05, "loss": 0.0226, "step": 20440 }, { "epoch": 5.9499563572883325, "grad_norm": 0.11982657015323639, "learning_rate": 2.3581736063765193e-05, "loss": 0.0222, "step": 20450 }, { "epoch": 5.952865871399476, "grad_norm": 0.09334678947925568, "learning_rate": 2.3574954104994492e-05, "loss": 0.0206, "step": 20460 }, { "epoch": 5.95577538551062, "grad_norm": 0.11099380999803543, "learning_rate": 2.3568169541379137e-05, "loss": 0.0222, "step": 20470 }, { "epoch": 5.958684899621764, "grad_norm": 0.10145154595375061, "learning_rate": 2.356138237498011e-05, "loss": 0.0211, "step": 20480 }, { "epoch": 5.961594413732906, "grad_norm": 0.09437248855829239, "learning_rate": 2.3554592607859166e-05, "loss": 0.0214, "step": 20490 }, { "epoch": 5.96450392784405, "grad_norm": 0.14014390110969543, "learning_rate": 2.3547800242078857e-05, "loss": 0.0205, "step": 20500 }, { "epoch": 5.967413441955194, "grad_norm": 0.11256888508796692, "learning_rate": 2.3541005279702528e-05, "loss": 0.0211, "step": 20510 }, { "epoch": 5.970322956066337, "grad_norm": 0.0927610918879509, "learning_rate": 2.3534207722794307e-05, "loss": 0.022, "step": 20520 }, { "epoch": 5.97323247017748, "grad_norm": 0.13880126178264618, "learning_rate": 2.352740757341911e-05, "loss": 0.0207, "step": 20530 }, { "epoch": 5.976141984288624, "grad_norm": 0.09728730469942093, "learning_rate": 2.3520604833642645e-05, "loss": 0.0205, "step": 20540 }, { "epoch": 5.979051498399767, "grad_norm": 0.10015565901994705, "learning_rate": 2.3513799505531407e-05, "loss": 0.02, "step": 20550 }, { "epoch": 5.981961012510911, "grad_norm": 0.13638345897197723, "learning_rate": 2.350699159115268e-05, "loss": 0.0219, "step": 20560 }, { "epoch": 5.984870526622054, "grad_norm": 0.1001751497387886, "learning_rate": 2.350018109257451e-05, "loss": 0.021, "step": 20570 }, { "epoch": 5.987780040733197, "grad_norm": 0.09473789483308792, "learning_rate": 2.3493368011865766e-05, "loss": 0.0197, "step": 20580 }, { "epoch": 5.990689554844341, "grad_norm": 0.11235690116882324, "learning_rate": 2.348655235109607e-05, "loss": 0.0209, "step": 20590 }, { "epoch": 5.9935990689554846, "grad_norm": 0.13160079717636108, "learning_rate": 2.3479734112335845e-05, "loss": 0.0232, "step": 20600 }, { "epoch": 5.996508583066628, "grad_norm": 0.0986599251627922, "learning_rate": 2.347291329765629e-05, "loss": 0.0197, "step": 20610 }, { "epoch": 5.999418097177771, "grad_norm": 0.09850507229566574, "learning_rate": 2.3466089909129396e-05, "loss": 0.0206, "step": 20620 }, { "epoch": 6.002327611288915, "grad_norm": 0.1354484260082245, "learning_rate": 2.3459263948827916e-05, "loss": 0.0204, "step": 20630 }, { "epoch": 6.0052371254000585, "grad_norm": 0.127113476395607, "learning_rate": 2.3452435418825404e-05, "loss": 0.0225, "step": 20640 }, { "epoch": 6.008146639511201, "grad_norm": 0.12571996450424194, "learning_rate": 2.344560432119618e-05, "loss": 0.0221, "step": 20650 }, { "epoch": 6.011056153622345, "grad_norm": 0.08561748266220093, "learning_rate": 2.3438770658015368e-05, "loss": 0.021, "step": 20660 }, { "epoch": 6.013965667733489, "grad_norm": 0.11601262539625168, "learning_rate": 2.3431934431358835e-05, "loss": 0.0217, "step": 20670 }, { "epoch": 6.0168751818446315, "grad_norm": 0.09429620206356049, "learning_rate": 2.342509564330325e-05, "loss": 0.0199, "step": 20680 }, { "epoch": 6.019784695955775, "grad_norm": 0.10286743938922882, "learning_rate": 2.3418254295926068e-05, "loss": 0.0215, "step": 20690 }, { "epoch": 6.022694210066919, "grad_norm": 0.13044972717761993, "learning_rate": 2.34114103913055e-05, "loss": 0.0223, "step": 20700 }, { "epoch": 6.025603724178063, "grad_norm": 0.13120004534721375, "learning_rate": 2.3404563931520545e-05, "loss": 0.0213, "step": 20710 }, { "epoch": 6.0285132382892055, "grad_norm": 0.13119274377822876, "learning_rate": 2.3397714918650982e-05, "loss": 0.02, "step": 20720 }, { "epoch": 6.031422752400349, "grad_norm": 0.08077781647443771, "learning_rate": 2.3390863354777346e-05, "loss": 0.0214, "step": 20730 }, { "epoch": 6.034332266511493, "grad_norm": 0.1015126034617424, "learning_rate": 2.338400924198098e-05, "loss": 0.0216, "step": 20740 }, { "epoch": 6.037241780622636, "grad_norm": 0.09870992600917816, "learning_rate": 2.3377152582343963e-05, "loss": 0.0206, "step": 20750 }, { "epoch": 6.040151294733779, "grad_norm": 0.12077158689498901, "learning_rate": 2.3370293377949184e-05, "loss": 0.0209, "step": 20760 }, { "epoch": 6.043060808844923, "grad_norm": 0.09527082741260529, "learning_rate": 2.3363431630880282e-05, "loss": 0.0214, "step": 20770 }, { "epoch": 6.045970322956066, "grad_norm": 0.10566988587379456, "learning_rate": 2.3356567343221676e-05, "loss": 0.0203, "step": 20780 }, { "epoch": 6.04887983706721, "grad_norm": 0.14433713257312775, "learning_rate": 2.3349700517058548e-05, "loss": 0.0208, "step": 20790 }, { "epoch": 6.051789351178353, "grad_norm": 0.12006741017103195, "learning_rate": 2.3342831154476866e-05, "loss": 0.0198, "step": 20800 }, { "epoch": 6.054698865289497, "grad_norm": 0.11391939222812653, "learning_rate": 2.3335959257563362e-05, "loss": 0.0194, "step": 20810 }, { "epoch": 6.05760837940064, "grad_norm": 0.1575349122285843, "learning_rate": 2.332908482840553e-05, "loss": 0.0231, "step": 20820 }, { "epoch": 6.060517893511784, "grad_norm": 0.1413782387971878, "learning_rate": 2.3322207869091648e-05, "loss": 0.02, "step": 20830 }, { "epoch": 6.063427407622927, "grad_norm": 0.10006045550107956, "learning_rate": 2.3315328381710743e-05, "loss": 0.0209, "step": 20840 }, { "epoch": 6.06633692173407, "grad_norm": 0.11049094051122665, "learning_rate": 2.3308446368352634e-05, "loss": 0.0198, "step": 20850 }, { "epoch": 6.069246435845214, "grad_norm": 0.1116638109087944, "learning_rate": 2.3301561831107888e-05, "loss": 0.0214, "step": 20860 }, { "epoch": 6.0721559499563575, "grad_norm": 0.11951228976249695, "learning_rate": 2.3294674772067853e-05, "loss": 0.0202, "step": 20870 }, { "epoch": 6.0750654640675, "grad_norm": 0.11191160976886749, "learning_rate": 2.3287785193324628e-05, "loss": 0.0203, "step": 20880 }, { "epoch": 6.077974978178644, "grad_norm": 0.131102055311203, "learning_rate": 2.3280893096971087e-05, "loss": 0.0218, "step": 20890 }, { "epoch": 6.080884492289788, "grad_norm": 0.12255993485450745, "learning_rate": 2.3273998485100867e-05, "loss": 0.022, "step": 20900 }, { "epoch": 6.083794006400931, "grad_norm": 0.11248026043176651, "learning_rate": 2.326710135980837e-05, "loss": 0.0199, "step": 20910 }, { "epoch": 6.086703520512074, "grad_norm": 0.10214821249246597, "learning_rate": 2.326020172318876e-05, "loss": 0.0204, "step": 20920 }, { "epoch": 6.089613034623218, "grad_norm": 0.13013996183872223, "learning_rate": 2.3253299577337966e-05, "loss": 0.0204, "step": 20930 }, { "epoch": 6.092522548734362, "grad_norm": 0.10841581225395203, "learning_rate": 2.324639492435267e-05, "loss": 0.0209, "step": 20940 }, { "epoch": 6.0954320628455045, "grad_norm": 0.10486618429422379, "learning_rate": 2.3239487766330336e-05, "loss": 0.0205, "step": 20950 }, { "epoch": 6.098341576956648, "grad_norm": 0.10091666877269745, "learning_rate": 2.323257810536917e-05, "loss": 0.0207, "step": 20960 }, { "epoch": 6.101251091067792, "grad_norm": 0.11515850573778152, "learning_rate": 2.3225665943568138e-05, "loss": 0.0218, "step": 20970 }, { "epoch": 6.104160605178935, "grad_norm": 0.14356887340545654, "learning_rate": 2.3218751283026984e-05, "loss": 0.0219, "step": 20980 }, { "epoch": 6.1070701192900785, "grad_norm": 0.13108421862125397, "learning_rate": 2.321183412584619e-05, "loss": 0.0221, "step": 20990 }, { "epoch": 6.109979633401222, "grad_norm": 0.11922410130500793, "learning_rate": 2.320491447412701e-05, "loss": 0.02, "step": 21000 }, { "epoch": 6.112889147512365, "grad_norm": 0.10598036646842957, "learning_rate": 2.3197992329971455e-05, "loss": 0.0197, "step": 21010 }, { "epoch": 6.115798661623509, "grad_norm": 0.08583904057741165, "learning_rate": 2.3191067695482285e-05, "loss": 0.0192, "step": 21020 }, { "epoch": 6.118708175734652, "grad_norm": 0.12081202864646912, "learning_rate": 2.3184140572763018e-05, "loss": 0.0209, "step": 21030 }, { "epoch": 6.121617689845796, "grad_norm": 0.10003688186407089, "learning_rate": 2.3177210963917937e-05, "loss": 0.0229, "step": 21040 }, { "epoch": 6.124527203956939, "grad_norm": 0.1155533716082573, "learning_rate": 2.317027887105207e-05, "loss": 0.021, "step": 21050 }, { "epoch": 6.127436718068083, "grad_norm": 0.09578444808721542, "learning_rate": 2.3163344296271206e-05, "loss": 0.0203, "step": 21060 }, { "epoch": 6.130346232179226, "grad_norm": 0.1077202558517456, "learning_rate": 2.3156407241681888e-05, "loss": 0.0198, "step": 21070 }, { "epoch": 6.133255746290369, "grad_norm": 0.09997234493494034, "learning_rate": 2.3149467709391407e-05, "loss": 0.0199, "step": 21080 }, { "epoch": 6.136165260401513, "grad_norm": 0.10205390304327011, "learning_rate": 2.3142525701507807e-05, "loss": 0.0214, "step": 21090 }, { "epoch": 6.139074774512657, "grad_norm": 0.1072118952870369, "learning_rate": 2.3135581220139893e-05, "loss": 0.0195, "step": 21100 }, { "epoch": 6.141984288623799, "grad_norm": 0.11831562966108322, "learning_rate": 2.3128634267397213e-05, "loss": 0.0203, "step": 21110 }, { "epoch": 6.144893802734943, "grad_norm": 0.0873606950044632, "learning_rate": 2.3121684845390065e-05, "loss": 0.0196, "step": 21120 }, { "epoch": 6.147803316846087, "grad_norm": 0.10435103625059128, "learning_rate": 2.3114732956229496e-05, "loss": 0.0201, "step": 21130 }, { "epoch": 6.1507128309572305, "grad_norm": 0.09189043194055557, "learning_rate": 2.3107778602027318e-05, "loss": 0.0215, "step": 21140 }, { "epoch": 6.153622345068373, "grad_norm": 0.0972067192196846, "learning_rate": 2.3100821784896072e-05, "loss": 0.0196, "step": 21150 }, { "epoch": 6.156531859179517, "grad_norm": 0.08695201575756073, "learning_rate": 2.309386250694906e-05, "loss": 0.0202, "step": 21160 }, { "epoch": 6.159441373290661, "grad_norm": 0.11174771189689636, "learning_rate": 2.3086900770300325e-05, "loss": 0.0215, "step": 21170 }, { "epoch": 6.162350887401804, "grad_norm": 0.10954684764146805, "learning_rate": 2.3079936577064656e-05, "loss": 0.0183, "step": 21180 }, { "epoch": 6.165260401512947, "grad_norm": 0.11549319326877594, "learning_rate": 2.3072969929357593e-05, "loss": 0.0218, "step": 21190 }, { "epoch": 6.168169915624091, "grad_norm": 0.11196195334196091, "learning_rate": 2.3066000829295422e-05, "loss": 0.0206, "step": 21200 }, { "epoch": 6.171079429735234, "grad_norm": 0.11247599124908447, "learning_rate": 2.3059029278995167e-05, "loss": 0.0221, "step": 21210 }, { "epoch": 6.1739889438463775, "grad_norm": 0.10039642453193665, "learning_rate": 2.3052055280574605e-05, "loss": 0.0218, "step": 21220 }, { "epoch": 6.176898457957521, "grad_norm": 0.08539005368947983, "learning_rate": 2.3045078836152254e-05, "loss": 0.0205, "step": 21230 }, { "epoch": 6.179807972068665, "grad_norm": 0.1062358021736145, "learning_rate": 2.3038099947847367e-05, "loss": 0.022, "step": 21240 }, { "epoch": 6.182717486179808, "grad_norm": 0.1283537745475769, "learning_rate": 2.3031118617779946e-05, "loss": 0.0227, "step": 21250 }, { "epoch": 6.1856270002909515, "grad_norm": 0.10774364322423935, "learning_rate": 2.3024134848070744e-05, "loss": 0.0195, "step": 21260 }, { "epoch": 6.188536514402095, "grad_norm": 0.12864592671394348, "learning_rate": 2.301714864084124e-05, "loss": 0.0216, "step": 21270 }, { "epoch": 6.191446028513238, "grad_norm": 0.11454860866069794, "learning_rate": 2.3010159998213654e-05, "loss": 0.0211, "step": 21280 }, { "epoch": 6.194355542624382, "grad_norm": 0.10628743469715118, "learning_rate": 2.300316892231096e-05, "loss": 0.0206, "step": 21290 }, { "epoch": 6.197265056735525, "grad_norm": 0.11614257842302322, "learning_rate": 2.299617541525686e-05, "loss": 0.0223, "step": 21300 }, { "epoch": 6.200174570846668, "grad_norm": 0.1356704980134964, "learning_rate": 2.2989179479175794e-05, "loss": 0.0215, "step": 21310 }, { "epoch": 6.203084084957812, "grad_norm": 0.11391148716211319, "learning_rate": 2.2982181116192944e-05, "loss": 0.0209, "step": 21320 }, { "epoch": 6.205993599068956, "grad_norm": 0.11709043383598328, "learning_rate": 2.297518032843423e-05, "loss": 0.0189, "step": 21330 }, { "epoch": 6.208903113180099, "grad_norm": 0.10637588053941727, "learning_rate": 2.2968177118026305e-05, "loss": 0.0221, "step": 21340 }, { "epoch": 6.211812627291242, "grad_norm": 0.11402124166488647, "learning_rate": 2.296117148709656e-05, "loss": 0.0214, "step": 21350 }, { "epoch": 6.214722141402386, "grad_norm": 0.0790625736117363, "learning_rate": 2.295416343777312e-05, "loss": 0.0195, "step": 21360 }, { "epoch": 6.21763165551353, "grad_norm": 0.11507515609264374, "learning_rate": 2.2947152972184838e-05, "loss": 0.0196, "step": 21370 }, { "epoch": 6.220541169624672, "grad_norm": 0.10207062214612961, "learning_rate": 2.294014009246132e-05, "loss": 0.0213, "step": 21380 }, { "epoch": 6.223450683735816, "grad_norm": 0.11755503714084625, "learning_rate": 2.293312480073289e-05, "loss": 0.0195, "step": 21390 }, { "epoch": 6.22636019784696, "grad_norm": 0.11242105066776276, "learning_rate": 2.2926107099130607e-05, "loss": 0.0199, "step": 21400 }, { "epoch": 6.229269711958103, "grad_norm": 0.13147595524787903, "learning_rate": 2.2919086989786267e-05, "loss": 0.0204, "step": 21410 }, { "epoch": 6.232179226069246, "grad_norm": 0.10578632354736328, "learning_rate": 2.2912064474832384e-05, "loss": 0.0198, "step": 21420 }, { "epoch": 6.23508874018039, "grad_norm": 0.0959605798125267, "learning_rate": 2.2905039556402223e-05, "loss": 0.0204, "step": 21430 }, { "epoch": 6.237998254291533, "grad_norm": 0.10849829018115997, "learning_rate": 2.2898012236629766e-05, "loss": 0.0198, "step": 21440 }, { "epoch": 6.240907768402677, "grad_norm": 0.10403690487146378, "learning_rate": 2.2890982517649728e-05, "loss": 0.0207, "step": 21450 }, { "epoch": 6.24381728251382, "grad_norm": 0.10882092267274857, "learning_rate": 2.2883950401597548e-05, "loss": 0.0215, "step": 21460 }, { "epoch": 6.246726796624964, "grad_norm": 0.13299259543418884, "learning_rate": 2.2876915890609396e-05, "loss": 0.0199, "step": 21470 }, { "epoch": 6.249636310736107, "grad_norm": 0.09925740957260132, "learning_rate": 2.2869878986822182e-05, "loss": 0.0208, "step": 21480 }, { "epoch": 6.2525458248472505, "grad_norm": 0.1484501212835312, "learning_rate": 2.286283969237352e-05, "loss": 0.0201, "step": 21490 }, { "epoch": 6.255455338958394, "grad_norm": 0.10799816250801086, "learning_rate": 2.2855798009401768e-05, "loss": 0.0207, "step": 21500 }, { "epoch": 6.258364853069537, "grad_norm": 0.13651156425476074, "learning_rate": 2.2848753940045996e-05, "loss": 0.021, "step": 21510 }, { "epoch": 6.261274367180681, "grad_norm": 0.12488546222448349, "learning_rate": 2.284170748644602e-05, "loss": 0.0214, "step": 21520 }, { "epoch": 6.2641838812918245, "grad_norm": 0.09791239351034164, "learning_rate": 2.283465865074235e-05, "loss": 0.022, "step": 21530 }, { "epoch": 6.267093395402968, "grad_norm": 0.10569383949041367, "learning_rate": 2.2827607435076245e-05, "loss": 0.0211, "step": 21540 }, { "epoch": 6.270002909514111, "grad_norm": 0.1363755464553833, "learning_rate": 2.282055384158968e-05, "loss": 0.0203, "step": 21550 }, { "epoch": 6.272912423625255, "grad_norm": 0.0948423221707344, "learning_rate": 2.2813497872425348e-05, "loss": 0.02, "step": 21560 }, { "epoch": 6.275821937736398, "grad_norm": 0.09378861635923386, "learning_rate": 2.2806439529726662e-05, "loss": 0.0215, "step": 21570 }, { "epoch": 6.278731451847541, "grad_norm": 0.11390753835439682, "learning_rate": 2.2799378815637766e-05, "loss": 0.0194, "step": 21580 }, { "epoch": 6.281640965958685, "grad_norm": 0.10171394050121307, "learning_rate": 2.2792315732303517e-05, "loss": 0.021, "step": 21590 }, { "epoch": 6.284550480069829, "grad_norm": 0.11087194085121155, "learning_rate": 2.2785250281869492e-05, "loss": 0.0211, "step": 21600 }, { "epoch": 6.2874599941809715, "grad_norm": 0.093198262155056, "learning_rate": 2.277818246648199e-05, "loss": 0.0208, "step": 21610 }, { "epoch": 6.290369508292115, "grad_norm": 0.12430267035961151, "learning_rate": 2.2771112288288025e-05, "loss": 0.0221, "step": 21620 }, { "epoch": 6.293279022403259, "grad_norm": 0.09510017931461334, "learning_rate": 2.276403974943533e-05, "loss": 0.0198, "step": 21630 }, { "epoch": 6.296188536514402, "grad_norm": 0.11176985502243042, "learning_rate": 2.2756964852072357e-05, "loss": 0.0221, "step": 21640 }, { "epoch": 6.299098050625545, "grad_norm": 0.1568373292684555, "learning_rate": 2.2749887598348273e-05, "loss": 0.0205, "step": 21650 }, { "epoch": 6.302007564736689, "grad_norm": 0.0978565663099289, "learning_rate": 2.2742807990412963e-05, "loss": 0.0198, "step": 21660 }, { "epoch": 6.304917078847833, "grad_norm": 0.10002394020557404, "learning_rate": 2.2735726030417022e-05, "loss": 0.0212, "step": 21670 }, { "epoch": 6.307826592958976, "grad_norm": 0.13950487971305847, "learning_rate": 2.272864172051176e-05, "loss": 0.0199, "step": 21680 }, { "epoch": 6.310736107070119, "grad_norm": 0.10636593401432037, "learning_rate": 2.272155506284921e-05, "loss": 0.0205, "step": 21690 }, { "epoch": 6.313645621181263, "grad_norm": 0.09939748793840408, "learning_rate": 2.2714466059582105e-05, "loss": 0.0214, "step": 21700 }, { "epoch": 6.316555135292406, "grad_norm": 0.09461610019207001, "learning_rate": 2.2707374712863902e-05, "loss": 0.0198, "step": 21710 }, { "epoch": 6.31946464940355, "grad_norm": 0.1058908998966217, "learning_rate": 2.270028102484876e-05, "loss": 0.0189, "step": 21720 }, { "epoch": 6.322374163514693, "grad_norm": 0.10487693548202515, "learning_rate": 2.2693184997691558e-05, "loss": 0.0185, "step": 21730 }, { "epoch": 6.325283677625836, "grad_norm": 0.11427279561758041, "learning_rate": 2.268608663354788e-05, "loss": 0.019, "step": 21740 }, { "epoch": 6.32819319173698, "grad_norm": 0.13885833323001862, "learning_rate": 2.2678985934574022e-05, "loss": 0.0199, "step": 21750 }, { "epoch": 6.3311027058481235, "grad_norm": 0.09479173272848129, "learning_rate": 2.267188290292699e-05, "loss": 0.0194, "step": 21760 }, { "epoch": 6.334012219959266, "grad_norm": 0.12189827114343643, "learning_rate": 2.2664777540764496e-05, "loss": 0.0211, "step": 21770 }, { "epoch": 6.33692173407041, "grad_norm": 0.08791027963161469, "learning_rate": 2.265766985024496e-05, "loss": 0.0208, "step": 21780 }, { "epoch": 6.339831248181554, "grad_norm": 0.09014482796192169, "learning_rate": 2.2650559833527516e-05, "loss": 0.0214, "step": 21790 }, { "epoch": 6.3427407622926975, "grad_norm": 0.08705073595046997, "learning_rate": 2.2643447492771998e-05, "loss": 0.0212, "step": 21800 }, { "epoch": 6.34565027640384, "grad_norm": 0.09731854498386383, "learning_rate": 2.2636332830138943e-05, "loss": 0.0206, "step": 21810 }, { "epoch": 6.348559790514984, "grad_norm": 0.13399814069271088, "learning_rate": 2.2629215847789596e-05, "loss": 0.0218, "step": 21820 }, { "epoch": 6.351469304626128, "grad_norm": 0.13006490468978882, "learning_rate": 2.2622096547885912e-05, "loss": 0.0213, "step": 21830 }, { "epoch": 6.3543788187372705, "grad_norm": 0.09823071211576462, "learning_rate": 2.261497493259055e-05, "loss": 0.0189, "step": 21840 }, { "epoch": 6.357288332848414, "grad_norm": 0.085336834192276, "learning_rate": 2.2607851004066863e-05, "loss": 0.0216, "step": 21850 }, { "epoch": 6.360197846959558, "grad_norm": 0.11223438382148743, "learning_rate": 2.2600724764478915e-05, "loss": 0.0199, "step": 21860 }, { "epoch": 6.363107361070702, "grad_norm": 0.11656636744737625, "learning_rate": 2.2593596215991473e-05, "loss": 0.0215, "step": 21870 }, { "epoch": 6.3660168751818444, "grad_norm": 0.08770813047885895, "learning_rate": 2.2586465360769996e-05, "loss": 0.0209, "step": 21880 }, { "epoch": 6.368926389292988, "grad_norm": 0.15145522356033325, "learning_rate": 2.257933220098065e-05, "loss": 0.0214, "step": 21890 }, { "epoch": 6.371835903404132, "grad_norm": 0.14294521510601044, "learning_rate": 2.2572196738790298e-05, "loss": 0.0189, "step": 21900 }, { "epoch": 6.374745417515275, "grad_norm": 0.13285815715789795, "learning_rate": 2.2565058976366517e-05, "loss": 0.0228, "step": 21910 }, { "epoch": 6.377654931626418, "grad_norm": 0.11303704231977463, "learning_rate": 2.2557918915877558e-05, "loss": 0.0204, "step": 21920 }, { "epoch": 6.380564445737562, "grad_norm": 0.12078692764043808, "learning_rate": 2.2550776559492392e-05, "loss": 0.0217, "step": 21930 }, { "epoch": 6.383473959848705, "grad_norm": 0.1123318076133728, "learning_rate": 2.254363190938067e-05, "loss": 0.0203, "step": 21940 }, { "epoch": 6.386383473959849, "grad_norm": 0.11393281817436218, "learning_rate": 2.2536484967712753e-05, "loss": 0.0203, "step": 21950 }, { "epoch": 6.389292988070992, "grad_norm": 0.10338664799928665, "learning_rate": 2.252933573665969e-05, "loss": 0.0219, "step": 21960 }, { "epoch": 6.392202502182135, "grad_norm": 0.09253668785095215, "learning_rate": 2.2522184218393234e-05, "loss": 0.0202, "step": 21970 }, { "epoch": 6.395112016293279, "grad_norm": 0.14781570434570312, "learning_rate": 2.2515030415085822e-05, "loss": 0.0203, "step": 21980 }, { "epoch": 6.398021530404423, "grad_norm": 0.10294511169195175, "learning_rate": 2.2507874328910592e-05, "loss": 0.0209, "step": 21990 }, { "epoch": 6.400931044515566, "grad_norm": 0.0920216515660286, "learning_rate": 2.250071596204138e-05, "loss": 0.0207, "step": 22000 }, { "epoch": 6.403840558626709, "grad_norm": 0.08825372904539108, "learning_rate": 2.2493555316652698e-05, "loss": 0.0194, "step": 22010 }, { "epoch": 6.406750072737853, "grad_norm": 0.0902145653963089, "learning_rate": 2.248639239491977e-05, "loss": 0.0197, "step": 22020 }, { "epoch": 6.4096595868489965, "grad_norm": 0.12914282083511353, "learning_rate": 2.24792271990185e-05, "loss": 0.0211, "step": 22030 }, { "epoch": 6.412569100960139, "grad_norm": 0.11312973499298096, "learning_rate": 2.2472059731125483e-05, "loss": 0.021, "step": 22040 }, { "epoch": 6.415478615071283, "grad_norm": 0.15021757781505585, "learning_rate": 2.2464889993418003e-05, "loss": 0.0203, "step": 22050 }, { "epoch": 6.418388129182427, "grad_norm": 0.11169181019067764, "learning_rate": 2.2457717988074047e-05, "loss": 0.0213, "step": 22060 }, { "epoch": 6.4212976432935704, "grad_norm": 0.0856681615114212, "learning_rate": 2.245054371727228e-05, "loss": 0.0206, "step": 22070 }, { "epoch": 6.424207157404713, "grad_norm": 0.1236334890127182, "learning_rate": 2.2443367183192046e-05, "loss": 0.022, "step": 22080 }, { "epoch": 6.427116671515857, "grad_norm": 0.10077019035816193, "learning_rate": 2.2436188388013394e-05, "loss": 0.0214, "step": 22090 }, { "epoch": 6.430026185627001, "grad_norm": 0.09543675929307938, "learning_rate": 2.2429007333917048e-05, "loss": 0.0212, "step": 22100 }, { "epoch": 6.4329356997381435, "grad_norm": 0.09564119577407837, "learning_rate": 2.242182402308443e-05, "loss": 0.019, "step": 22110 }, { "epoch": 6.435845213849287, "grad_norm": 0.11371074616909027, "learning_rate": 2.2414638457697632e-05, "loss": 0.0219, "step": 22120 }, { "epoch": 6.438754727960431, "grad_norm": 0.14590699970722198, "learning_rate": 2.2407450639939448e-05, "loss": 0.0233, "step": 22130 }, { "epoch": 6.441664242071574, "grad_norm": 0.11232102662324905, "learning_rate": 2.2400260571993333e-05, "loss": 0.0202, "step": 22140 }, { "epoch": 6.444573756182717, "grad_norm": 0.08333616703748703, "learning_rate": 2.2393068256043457e-05, "loss": 0.0209, "step": 22150 }, { "epoch": 6.447483270293861, "grad_norm": 0.10611465573310852, "learning_rate": 2.2385873694274646e-05, "loss": 0.0189, "step": 22160 }, { "epoch": 6.450392784405004, "grad_norm": 0.09230098128318787, "learning_rate": 2.2378676888872418e-05, "loss": 0.0206, "step": 22170 }, { "epoch": 6.453302298516148, "grad_norm": 0.10343257337808609, "learning_rate": 2.2371477842022976e-05, "loss": 0.019, "step": 22180 }, { "epoch": 6.456211812627291, "grad_norm": 0.136274054646492, "learning_rate": 2.2364276555913195e-05, "loss": 0.0216, "step": 22190 }, { "epoch": 6.459121326738435, "grad_norm": 0.09907646477222443, "learning_rate": 2.235707303273064e-05, "loss": 0.02, "step": 22200 }, { "epoch": 6.462030840849578, "grad_norm": 0.11895360052585602, "learning_rate": 2.234986727466355e-05, "loss": 0.0192, "step": 22210 }, { "epoch": 6.464940354960722, "grad_norm": 0.09484484046697617, "learning_rate": 2.2342659283900843e-05, "loss": 0.0201, "step": 22220 }, { "epoch": 6.467849869071865, "grad_norm": 0.09374019503593445, "learning_rate": 2.2335449062632117e-05, "loss": 0.0217, "step": 22230 }, { "epoch": 6.470759383183008, "grad_norm": 0.11515262722969055, "learning_rate": 2.232823661304765e-05, "loss": 0.0193, "step": 22240 }, { "epoch": 6.473668897294152, "grad_norm": 0.1528579592704773, "learning_rate": 2.2321021937338392e-05, "loss": 0.0198, "step": 22250 }, { "epoch": 6.4765784114052956, "grad_norm": 0.10098729282617569, "learning_rate": 2.2313805037695965e-05, "loss": 0.0209, "step": 22260 }, { "epoch": 6.479487925516438, "grad_norm": 0.1285858005285263, "learning_rate": 2.230658591631268e-05, "loss": 0.0218, "step": 22270 }, { "epoch": 6.482397439627582, "grad_norm": 0.10189543664455414, "learning_rate": 2.2299364575381512e-05, "loss": 0.021, "step": 22280 }, { "epoch": 6.485306953738726, "grad_norm": 0.11776158213615417, "learning_rate": 2.2292141017096115e-05, "loss": 0.0212, "step": 22290 }, { "epoch": 6.488216467849869, "grad_norm": 0.13440774381160736, "learning_rate": 2.228491524365081e-05, "loss": 0.0181, "step": 22300 }, { "epoch": 6.491125981961012, "grad_norm": 0.11073185503482819, "learning_rate": 2.227768725724061e-05, "loss": 0.0197, "step": 22310 }, { "epoch": 6.494035496072156, "grad_norm": 0.14421415328979492, "learning_rate": 2.227045706006117e-05, "loss": 0.0204, "step": 22320 }, { "epoch": 6.4969450101833, "grad_norm": 0.12198848277330399, "learning_rate": 2.2263224654308845e-05, "loss": 0.0204, "step": 22330 }, { "epoch": 6.4998545242944425, "grad_norm": 0.09919540584087372, "learning_rate": 2.2255990042180646e-05, "loss": 0.02, "step": 22340 }, { "epoch": 6.502764038405586, "grad_norm": 0.08060254901647568, "learning_rate": 2.2248753225874252e-05, "loss": 0.0219, "step": 22350 }, { "epoch": 6.50567355251673, "grad_norm": 0.10738559812307358, "learning_rate": 2.2241514207588018e-05, "loss": 0.0217, "step": 22360 }, { "epoch": 6.508583066627873, "grad_norm": 0.1082891896367073, "learning_rate": 2.223427298952097e-05, "loss": 0.0223, "step": 22370 }, { "epoch": 6.5114925807390165, "grad_norm": 0.1344405859708786, "learning_rate": 2.2227029573872803e-05, "loss": 0.0197, "step": 22380 }, { "epoch": 6.51440209485016, "grad_norm": 0.0986073687672615, "learning_rate": 2.2219783962843866e-05, "loss": 0.0185, "step": 22390 }, { "epoch": 6.517311608961304, "grad_norm": 0.09313423186540604, "learning_rate": 2.221253615863519e-05, "loss": 0.0183, "step": 22400 }, { "epoch": 6.520221123072447, "grad_norm": 0.11206455528736115, "learning_rate": 2.2205286163448466e-05, "loss": 0.0203, "step": 22410 }, { "epoch": 6.52313063718359, "grad_norm": 0.11964265257120132, "learning_rate": 2.219803397948604e-05, "loss": 0.0197, "step": 22420 }, { "epoch": 6.526040151294734, "grad_norm": 0.11830508708953857, "learning_rate": 2.2190779608950953e-05, "loss": 0.0198, "step": 22430 }, { "epoch": 6.528949665405877, "grad_norm": 0.11615486443042755, "learning_rate": 2.218352305404688e-05, "loss": 0.0184, "step": 22440 }, { "epoch": 6.531859179517021, "grad_norm": 0.1190621629357338, "learning_rate": 2.217626431697818e-05, "loss": 0.0203, "step": 22450 }, { "epoch": 6.534768693628164, "grad_norm": 0.11171691864728928, "learning_rate": 2.216900339994985e-05, "loss": 0.0204, "step": 22460 }, { "epoch": 6.537678207739307, "grad_norm": 0.11144464462995529, "learning_rate": 2.2161740305167574e-05, "loss": 0.0204, "step": 22470 }, { "epoch": 6.540587721850451, "grad_norm": 0.12287883460521698, "learning_rate": 2.215447503483769e-05, "loss": 0.0202, "step": 22480 }, { "epoch": 6.543497235961595, "grad_norm": 0.12262041866779327, "learning_rate": 2.21472075911672e-05, "loss": 0.0186, "step": 22490 }, { "epoch": 6.546406750072737, "grad_norm": 0.10708487033843994, "learning_rate": 2.2139937976363747e-05, "loss": 0.0184, "step": 22500 }, { "epoch": 6.549316264183881, "grad_norm": 0.12064751982688904, "learning_rate": 2.2132666192635658e-05, "loss": 0.0212, "step": 22510 }, { "epoch": 6.552225778295025, "grad_norm": 0.0957719013094902, "learning_rate": 2.2125392242191905e-05, "loss": 0.0197, "step": 22520 }, { "epoch": 6.5551352924061685, "grad_norm": 0.11841316521167755, "learning_rate": 2.2118116127242128e-05, "loss": 0.0203, "step": 22530 }, { "epoch": 6.558044806517311, "grad_norm": 0.09903433918952942, "learning_rate": 2.2110837849996614e-05, "loss": 0.0195, "step": 22540 }, { "epoch": 6.560954320628455, "grad_norm": 0.10468652099370956, "learning_rate": 2.210355741266631e-05, "loss": 0.0197, "step": 22550 }, { "epoch": 6.563863834739599, "grad_norm": 0.09275902062654495, "learning_rate": 2.2096274817462823e-05, "loss": 0.0208, "step": 22560 }, { "epoch": 6.566773348850742, "grad_norm": 0.10246459394693375, "learning_rate": 2.208899006659841e-05, "loss": 0.0199, "step": 22570 }, { "epoch": 6.569682862961885, "grad_norm": 0.1095660850405693, "learning_rate": 2.2081703162285994e-05, "loss": 0.0193, "step": 22580 }, { "epoch": 6.572592377073029, "grad_norm": 0.08657128363847733, "learning_rate": 2.2074414106739133e-05, "loss": 0.0191, "step": 22590 }, { "epoch": 6.575501891184173, "grad_norm": 0.09939993917942047, "learning_rate": 2.2067122902172062e-05, "loss": 0.0191, "step": 22600 }, { "epoch": 6.5784114052953155, "grad_norm": 0.12540853023529053, "learning_rate": 2.2059829550799645e-05, "loss": 0.0217, "step": 22610 }, { "epoch": 6.581320919406459, "grad_norm": 0.09227637946605682, "learning_rate": 2.2052534054837417e-05, "loss": 0.0194, "step": 22620 }, { "epoch": 6.584230433517602, "grad_norm": 0.11694002896547318, "learning_rate": 2.2045236416501552e-05, "loss": 0.0216, "step": 22630 }, { "epoch": 6.587139947628746, "grad_norm": 0.10319943726062775, "learning_rate": 2.2037936638008885e-05, "loss": 0.0211, "step": 22640 }, { "epoch": 6.5900494617398895, "grad_norm": 0.10177615284919739, "learning_rate": 2.203063472157689e-05, "loss": 0.021, "step": 22650 }, { "epoch": 6.592958975851033, "grad_norm": 0.13315680623054504, "learning_rate": 2.2023330669423703e-05, "loss": 0.021, "step": 22660 }, { "epoch": 6.595868489962176, "grad_norm": 0.12178623676300049, "learning_rate": 2.20160244837681e-05, "loss": 0.0176, "step": 22670 }, { "epoch": 6.59877800407332, "grad_norm": 0.0905139297246933, "learning_rate": 2.20087161668295e-05, "loss": 0.0197, "step": 22680 }, { "epoch": 6.601687518184463, "grad_norm": 0.08592864871025085, "learning_rate": 2.200140572082799e-05, "loss": 0.0188, "step": 22690 }, { "epoch": 6.604597032295606, "grad_norm": 0.10971427708864212, "learning_rate": 2.1994093147984283e-05, "loss": 0.0193, "step": 22700 }, { "epoch": 6.60750654640675, "grad_norm": 0.0941147431731224, "learning_rate": 2.1986778450519742e-05, "loss": 0.0184, "step": 22710 }, { "epoch": 6.610416060517894, "grad_norm": 0.11209166049957275, "learning_rate": 2.1979461630656384e-05, "loss": 0.0187, "step": 22720 }, { "epoch": 6.613325574629037, "grad_norm": 0.10277436673641205, "learning_rate": 2.1972142690616866e-05, "loss": 0.0205, "step": 22730 }, { "epoch": 6.61623508874018, "grad_norm": 0.10214804857969284, "learning_rate": 2.1964821632624484e-05, "loss": 0.0218, "step": 22740 }, { "epoch": 6.619144602851324, "grad_norm": 0.12494538724422455, "learning_rate": 2.1957498458903188e-05, "loss": 0.0195, "step": 22750 }, { "epoch": 6.622054116962468, "grad_norm": 0.0991630107164383, "learning_rate": 2.195017317167756e-05, "loss": 0.0187, "step": 22760 }, { "epoch": 6.62496363107361, "grad_norm": 0.1034352257847786, "learning_rate": 2.1942845773172836e-05, "loss": 0.0205, "step": 22770 }, { "epoch": 6.627873145184754, "grad_norm": 0.12175639718770981, "learning_rate": 2.193551626561488e-05, "loss": 0.0209, "step": 22780 }, { "epoch": 6.630782659295898, "grad_norm": 0.10018133372068405, "learning_rate": 2.19281846512302e-05, "loss": 0.0191, "step": 22790 }, { "epoch": 6.633692173407041, "grad_norm": 0.09653471410274506, "learning_rate": 2.1920850932245955e-05, "loss": 0.0191, "step": 22800 }, { "epoch": 6.636601687518184, "grad_norm": 0.1009698137640953, "learning_rate": 2.1913515110889932e-05, "loss": 0.0185, "step": 22810 }, { "epoch": 6.639511201629328, "grad_norm": 0.09841009974479675, "learning_rate": 2.1906177189390557e-05, "loss": 0.02, "step": 22820 }, { "epoch": 6.642420715740471, "grad_norm": 0.16801340878009796, "learning_rate": 2.18988371699769e-05, "loss": 0.0211, "step": 22830 }, { "epoch": 6.645330229851615, "grad_norm": 0.10179883986711502, "learning_rate": 2.189149505487866e-05, "loss": 0.0194, "step": 22840 }, { "epoch": 6.648239743962758, "grad_norm": 0.10458178073167801, "learning_rate": 2.1884150846326186e-05, "loss": 0.0202, "step": 22850 }, { "epoch": 6.651149258073902, "grad_norm": 0.09762511402368546, "learning_rate": 2.187680454655045e-05, "loss": 0.0203, "step": 22860 }, { "epoch": 6.654058772185045, "grad_norm": 0.09681999683380127, "learning_rate": 2.186945615778306e-05, "loss": 0.0188, "step": 22870 }, { "epoch": 6.6569682862961885, "grad_norm": 0.09533578902482986, "learning_rate": 2.186210568225627e-05, "loss": 0.0206, "step": 22880 }, { "epoch": 6.659877800407332, "grad_norm": 0.09981118887662888, "learning_rate": 2.1854753122202955e-05, "loss": 0.0203, "step": 22890 }, { "epoch": 6.662787314518475, "grad_norm": 0.11007751524448395, "learning_rate": 2.1847398479856638e-05, "loss": 0.0186, "step": 22900 }, { "epoch": 6.665696828629619, "grad_norm": 0.09610297530889511, "learning_rate": 2.184004175745145e-05, "loss": 0.0188, "step": 22910 }, { "epoch": 6.6686063427407625, "grad_norm": 0.09750641137361526, "learning_rate": 2.1832682957222182e-05, "loss": 0.0188, "step": 22920 }, { "epoch": 6.671515856851906, "grad_norm": 0.11555935442447662, "learning_rate": 2.1825322081404236e-05, "loss": 0.0193, "step": 22930 }, { "epoch": 6.674425370963049, "grad_norm": 0.09487448632717133, "learning_rate": 2.1817959132233655e-05, "loss": 0.0191, "step": 22940 }, { "epoch": 6.677334885074193, "grad_norm": 0.10622060298919678, "learning_rate": 2.1810594111947104e-05, "loss": 0.0188, "step": 22950 }, { "epoch": 6.680244399185336, "grad_norm": 0.09781807661056519, "learning_rate": 2.1803227022781892e-05, "loss": 0.0192, "step": 22960 }, { "epoch": 6.683153913296479, "grad_norm": 0.09409067779779434, "learning_rate": 2.1795857866975933e-05, "loss": 0.0191, "step": 22970 }, { "epoch": 6.686063427407623, "grad_norm": 0.11616773903369904, "learning_rate": 2.178848664676779e-05, "loss": 0.0195, "step": 22980 }, { "epoch": 6.688972941518767, "grad_norm": 0.12740176916122437, "learning_rate": 2.1781113364396643e-05, "loss": 0.0213, "step": 22990 }, { "epoch": 6.6918824556299095, "grad_norm": 0.11053642630577087, "learning_rate": 2.17737380221023e-05, "loss": 0.0212, "step": 23000 }, { "epoch": 6.694791969741053, "grad_norm": 0.10421355068683624, "learning_rate": 2.1766360622125198e-05, "loss": 0.0197, "step": 23010 }, { "epoch": 6.697701483852197, "grad_norm": 0.1002814918756485, "learning_rate": 2.175898116670639e-05, "loss": 0.0197, "step": 23020 }, { "epoch": 6.70061099796334, "grad_norm": 0.12352363765239716, "learning_rate": 2.175159965808757e-05, "loss": 0.0202, "step": 23030 }, { "epoch": 6.703520512074483, "grad_norm": 0.10269822925329208, "learning_rate": 2.1744216098511038e-05, "loss": 0.0203, "step": 23040 }, { "epoch": 6.706430026185627, "grad_norm": 0.09019002318382263, "learning_rate": 2.1736830490219722e-05, "loss": 0.0182, "step": 23050 }, { "epoch": 6.709339540296771, "grad_norm": 0.08781816810369492, "learning_rate": 2.1729442835457182e-05, "loss": 0.0202, "step": 23060 }, { "epoch": 6.712249054407914, "grad_norm": 0.10688629746437073, "learning_rate": 2.1722053136467594e-05, "loss": 0.0179, "step": 23070 }, { "epoch": 6.715158568519057, "grad_norm": 0.11180438101291656, "learning_rate": 2.1714661395495743e-05, "loss": 0.02, "step": 23080 }, { "epoch": 6.718068082630201, "grad_norm": 0.131430983543396, "learning_rate": 2.170726761478705e-05, "loss": 0.0202, "step": 23090 }, { "epoch": 6.720977596741344, "grad_norm": 0.09592291712760925, "learning_rate": 2.1699871796587554e-05, "loss": 0.0179, "step": 23100 }, { "epoch": 6.723887110852488, "grad_norm": 0.10900662839412689, "learning_rate": 2.1692473943143904e-05, "loss": 0.0189, "step": 23110 }, { "epoch": 6.726796624963631, "grad_norm": 0.1134210154414177, "learning_rate": 2.1685074056703373e-05, "loss": 0.0186, "step": 23120 }, { "epoch": 6.729706139074775, "grad_norm": 0.0768001452088356, "learning_rate": 2.167767213951386e-05, "loss": 0.0182, "step": 23130 }, { "epoch": 6.732615653185918, "grad_norm": 0.08840926736593246, "learning_rate": 2.1670268193823853e-05, "loss": 0.0194, "step": 23140 }, { "epoch": 6.7355251672970615, "grad_norm": 0.10497499257326126, "learning_rate": 2.1662862221882495e-05, "loss": 0.018, "step": 23150 }, { "epoch": 6.738434681408204, "grad_norm": 0.09230361878871918, "learning_rate": 2.1655454225939514e-05, "loss": 0.0205, "step": 23160 }, { "epoch": 6.741344195519348, "grad_norm": 0.09922000020742416, "learning_rate": 2.1648044208245266e-05, "loss": 0.0193, "step": 23170 }, { "epoch": 6.744253709630492, "grad_norm": 0.09365326911211014, "learning_rate": 2.164063217105072e-05, "loss": 0.0199, "step": 23180 }, { "epoch": 6.7471632237416355, "grad_norm": 0.09993220865726471, "learning_rate": 2.163321811660746e-05, "loss": 0.0201, "step": 23190 }, { "epoch": 6.750072737852778, "grad_norm": 0.08118276298046112, "learning_rate": 2.162580204716767e-05, "loss": 0.0195, "step": 23200 }, { "epoch": 6.752982251963922, "grad_norm": 0.0858805775642395, "learning_rate": 2.1618383964984162e-05, "loss": 0.0204, "step": 23210 }, { "epoch": 6.755891766075066, "grad_norm": 0.08742697536945343, "learning_rate": 2.1610963872310358e-05, "loss": 0.0203, "step": 23220 }, { "epoch": 6.7588012801862085, "grad_norm": 0.1314486712217331, "learning_rate": 2.160354177140028e-05, "loss": 0.02, "step": 23230 }, { "epoch": 6.761710794297352, "grad_norm": 0.09897257387638092, "learning_rate": 2.159611766450857e-05, "loss": 0.0216, "step": 23240 }, { "epoch": 6.764620308408496, "grad_norm": 0.09430516511201859, "learning_rate": 2.1588691553890475e-05, "loss": 0.0198, "step": 23250 }, { "epoch": 6.76752982251964, "grad_norm": 0.10173292458057404, "learning_rate": 2.1581263441801846e-05, "loss": 0.0188, "step": 23260 }, { "epoch": 6.7704393366307825, "grad_norm": 0.11377286165952682, "learning_rate": 2.157383333049915e-05, "loss": 0.0188, "step": 23270 }, { "epoch": 6.773348850741926, "grad_norm": 0.10178735107183456, "learning_rate": 2.156640122223947e-05, "loss": 0.021, "step": 23280 }, { "epoch": 6.77625836485307, "grad_norm": 0.09535397589206696, "learning_rate": 2.1558967119280468e-05, "loss": 0.0189, "step": 23290 }, { "epoch": 6.779167878964213, "grad_norm": 0.10208355635404587, "learning_rate": 2.155153102388043e-05, "loss": 0.0206, "step": 23300 }, { "epoch": 6.782077393075356, "grad_norm": 0.09824355691671371, "learning_rate": 2.1544092938298258e-05, "loss": 0.0194, "step": 23310 }, { "epoch": 6.7849869071865, "grad_norm": 0.11235559731721878, "learning_rate": 2.153665286479343e-05, "loss": 0.0212, "step": 23320 }, { "epoch": 6.787896421297643, "grad_norm": 0.10265575349330902, "learning_rate": 2.1529210805626058e-05, "loss": 0.0184, "step": 23330 }, { "epoch": 6.790805935408787, "grad_norm": 0.10564474761486053, "learning_rate": 2.1521766763056835e-05, "loss": 0.0206, "step": 23340 }, { "epoch": 6.79371544951993, "grad_norm": 0.12021390348672867, "learning_rate": 2.151432073934707e-05, "loss": 0.0186, "step": 23350 }, { "epoch": 6.796624963631073, "grad_norm": 0.116246797144413, "learning_rate": 2.1506872736758665e-05, "loss": 0.019, "step": 23360 }, { "epoch": 6.799534477742217, "grad_norm": 0.11407195776700974, "learning_rate": 2.149942275755412e-05, "loss": 0.0198, "step": 23370 }, { "epoch": 6.802443991853361, "grad_norm": 0.11101358383893967, "learning_rate": 2.1491970803996557e-05, "loss": 0.0187, "step": 23380 }, { "epoch": 6.805353505964504, "grad_norm": 0.13592465221881866, "learning_rate": 2.1484516878349667e-05, "loss": 0.0198, "step": 23390 }, { "epoch": 6.808263020075647, "grad_norm": 0.10590767860412598, "learning_rate": 2.147706098287777e-05, "loss": 0.0208, "step": 23400 }, { "epoch": 6.811172534186791, "grad_norm": 0.0887679010629654, "learning_rate": 2.146960311984576e-05, "loss": 0.0198, "step": 23410 }, { "epoch": 6.8140820482979345, "grad_norm": 0.10118834674358368, "learning_rate": 2.146214329151914e-05, "loss": 0.0197, "step": 23420 }, { "epoch": 6.816991562409077, "grad_norm": 0.09704320877790451, "learning_rate": 2.1454681500164012e-05, "loss": 0.0196, "step": 23430 }, { "epoch": 6.819901076520221, "grad_norm": 0.10656221956014633, "learning_rate": 2.144721774804707e-05, "loss": 0.0187, "step": 23440 }, { "epoch": 6.822810590631365, "grad_norm": 0.12544046342372894, "learning_rate": 2.1439752037435607e-05, "loss": 0.0197, "step": 23450 }, { "epoch": 6.8257201047425085, "grad_norm": 0.09381880611181259, "learning_rate": 2.1432284370597505e-05, "loss": 0.019, "step": 23460 }, { "epoch": 6.828629618853651, "grad_norm": 0.15262269973754883, "learning_rate": 2.1424814749801246e-05, "loss": 0.0215, "step": 23470 }, { "epoch": 6.831539132964795, "grad_norm": 0.11551612615585327, "learning_rate": 2.1417343177315905e-05, "loss": 0.0191, "step": 23480 }, { "epoch": 6.834448647075939, "grad_norm": 0.10624992847442627, "learning_rate": 2.1409869655411145e-05, "loss": 0.0195, "step": 23490 }, { "epoch": 6.8373581611870815, "grad_norm": 0.11075709015130997, "learning_rate": 2.140239418635723e-05, "loss": 0.0208, "step": 23500 }, { "epoch": 6.840267675298225, "grad_norm": 0.09582839161157608, "learning_rate": 2.1394916772425002e-05, "loss": 0.0189, "step": 23510 }, { "epoch": 6.843177189409369, "grad_norm": 0.09931399673223495, "learning_rate": 2.138743741588591e-05, "loss": 0.0195, "step": 23520 }, { "epoch": 6.846086703520512, "grad_norm": 0.09637772291898727, "learning_rate": 2.137995611901198e-05, "loss": 0.0166, "step": 23530 }, { "epoch": 6.8489962176316554, "grad_norm": 0.16436272859573364, "learning_rate": 2.1372472884075835e-05, "loss": 0.0209, "step": 23540 }, { "epoch": 6.851905731742799, "grad_norm": 0.09652041643857956, "learning_rate": 2.136498771335068e-05, "loss": 0.0188, "step": 23550 }, { "epoch": 6.854815245853942, "grad_norm": 0.11709102243185043, "learning_rate": 2.1357500609110324e-05, "loss": 0.0185, "step": 23560 }, { "epoch": 6.857724759965086, "grad_norm": 0.09928011149168015, "learning_rate": 2.1350011573629134e-05, "loss": 0.0192, "step": 23570 }, { "epoch": 6.860634274076229, "grad_norm": 0.133616641163826, "learning_rate": 2.1342520609182093e-05, "loss": 0.0193, "step": 23580 }, { "epoch": 6.863543788187373, "grad_norm": 0.10664752870798111, "learning_rate": 2.1335027718044753e-05, "loss": 0.0208, "step": 23590 }, { "epoch": 6.866453302298516, "grad_norm": 0.1334938108921051, "learning_rate": 2.1327532902493263e-05, "loss": 0.019, "step": 23600 }, { "epoch": 6.86936281640966, "grad_norm": 0.11474815756082535, "learning_rate": 2.1320036164804345e-05, "loss": 0.02, "step": 23610 }, { "epoch": 6.872272330520803, "grad_norm": 0.14754600822925568, "learning_rate": 2.131253750725531e-05, "loss": 0.0207, "step": 23620 }, { "epoch": 6.875181844631946, "grad_norm": 0.13784460723400116, "learning_rate": 2.1305036932124053e-05, "loss": 0.0186, "step": 23630 }, { "epoch": 6.87809135874309, "grad_norm": 0.09012959152460098, "learning_rate": 2.1297534441689054e-05, "loss": 0.0201, "step": 23640 }, { "epoch": 6.881000872854234, "grad_norm": 0.09882214665412903, "learning_rate": 2.1290030038229365e-05, "loss": 0.0202, "step": 23650 }, { "epoch": 6.883910386965377, "grad_norm": 0.11772932857275009, "learning_rate": 2.1282523724024632e-05, "loss": 0.0174, "step": 23660 }, { "epoch": 6.88681990107652, "grad_norm": 0.1083313524723053, "learning_rate": 2.127501550135507e-05, "loss": 0.021, "step": 23670 }, { "epoch": 6.889729415187664, "grad_norm": 0.10789720714092255, "learning_rate": 2.126750537250148e-05, "loss": 0.0187, "step": 23680 }, { "epoch": 6.892638929298807, "grad_norm": 0.1245412677526474, "learning_rate": 2.1259993339745244e-05, "loss": 0.0191, "step": 23690 }, { "epoch": 6.89554844340995, "grad_norm": 0.11686645448207855, "learning_rate": 2.1252479405368314e-05, "loss": 0.02, "step": 23700 }, { "epoch": 6.898457957521094, "grad_norm": 0.10994692146778107, "learning_rate": 2.1244963571653232e-05, "loss": 0.0188, "step": 23710 }, { "epoch": 6.901367471632238, "grad_norm": 0.10372791439294815, "learning_rate": 2.1237445840883103e-05, "loss": 0.018, "step": 23720 }, { "epoch": 6.904276985743381, "grad_norm": 0.11426715552806854, "learning_rate": 2.1229926215341616e-05, "loss": 0.0191, "step": 23730 }, { "epoch": 6.907186499854524, "grad_norm": 0.11876816302537918, "learning_rate": 2.1222404697313036e-05, "loss": 0.0199, "step": 23740 }, { "epoch": 6.910096013965668, "grad_norm": 0.1371668428182602, "learning_rate": 2.1214881289082205e-05, "loss": 0.0199, "step": 23750 }, { "epoch": 6.913005528076811, "grad_norm": 0.1165928766131401, "learning_rate": 2.1207355992934535e-05, "loss": 0.0215, "step": 23760 }, { "epoch": 6.9159150421879545, "grad_norm": 0.09593873471021652, "learning_rate": 2.119982881115601e-05, "loss": 0.0196, "step": 23770 }, { "epoch": 6.918824556299098, "grad_norm": 0.12225604802370071, "learning_rate": 2.1192299746033187e-05, "loss": 0.0206, "step": 23780 }, { "epoch": 6.921734070410242, "grad_norm": 0.0957389697432518, "learning_rate": 2.1184768799853195e-05, "loss": 0.0182, "step": 23790 }, { "epoch": 6.924643584521385, "grad_norm": 0.120449960231781, "learning_rate": 2.1177235974903748e-05, "loss": 0.0225, "step": 23800 }, { "epoch": 6.927553098632528, "grad_norm": 0.07683759182691574, "learning_rate": 2.116970127347311e-05, "loss": 0.0194, "step": 23810 }, { "epoch": 6.930462612743672, "grad_norm": 0.09860112518072128, "learning_rate": 2.1162164697850127e-05, "loss": 0.0204, "step": 23820 }, { "epoch": 6.933372126854815, "grad_norm": 0.10845496505498886, "learning_rate": 2.1154626250324206e-05, "loss": 0.0182, "step": 23830 }, { "epoch": 6.936281640965959, "grad_norm": 0.09923604130744934, "learning_rate": 2.1147085933185338e-05, "loss": 0.0198, "step": 23840 }, { "epoch": 6.939191155077102, "grad_norm": 0.1051173135638237, "learning_rate": 2.113954374872407e-05, "loss": 0.0196, "step": 23850 }, { "epoch": 6.942100669188245, "grad_norm": 0.11479814350605011, "learning_rate": 2.1131999699231513e-05, "loss": 0.0182, "step": 23860 }, { "epoch": 6.945010183299389, "grad_norm": 0.11135349422693253, "learning_rate": 2.1124453786999354e-05, "loss": 0.0202, "step": 23870 }, { "epoch": 6.947919697410533, "grad_norm": 0.11025597900152206, "learning_rate": 2.111690601431984e-05, "loss": 0.0188, "step": 23880 }, { "epoch": 6.950829211521675, "grad_norm": 0.11984822899103165, "learning_rate": 2.110935638348579e-05, "loss": 0.0208, "step": 23890 }, { "epoch": 6.953738725632819, "grad_norm": 0.08436216413974762, "learning_rate": 2.1101804896790576e-05, "loss": 0.0192, "step": 23900 }, { "epoch": 6.956648239743963, "grad_norm": 0.11427998542785645, "learning_rate": 2.1094251556528143e-05, "loss": 0.0195, "step": 23910 }, { "epoch": 6.9595577538551066, "grad_norm": 0.09425127506256104, "learning_rate": 2.1086696364993e-05, "loss": 0.0188, "step": 23920 }, { "epoch": 6.962467267966249, "grad_norm": 0.0901489183306694, "learning_rate": 2.107913932448021e-05, "loss": 0.018, "step": 23930 }, { "epoch": 6.965376782077393, "grad_norm": 0.09428131580352783, "learning_rate": 2.10715804372854e-05, "loss": 0.0199, "step": 23940 }, { "epoch": 6.968286296188537, "grad_norm": 0.10735876858234406, "learning_rate": 2.1064019705704767e-05, "loss": 0.0203, "step": 23950 }, { "epoch": 6.97119581029968, "grad_norm": 0.10122586786746979, "learning_rate": 2.1056457132035062e-05, "loss": 0.02, "step": 23960 }, { "epoch": 6.974105324410823, "grad_norm": 0.11119381338357925, "learning_rate": 2.1048892718573586e-05, "loss": 0.0206, "step": 23970 }, { "epoch": 6.977014838521967, "grad_norm": 0.0949196144938469, "learning_rate": 2.1041326467618215e-05, "loss": 0.0199, "step": 23980 }, { "epoch": 6.979924352633111, "grad_norm": 0.12454386055469513, "learning_rate": 2.103375838146738e-05, "loss": 0.02, "step": 23990 }, { "epoch": 6.9828338667442535, "grad_norm": 0.11052357405424118, "learning_rate": 2.1026188462420052e-05, "loss": 0.0178, "step": 24000 }, { "epoch": 6.985743380855397, "grad_norm": 0.08788308501243591, "learning_rate": 2.1018616712775792e-05, "loss": 0.0189, "step": 24010 }, { "epoch": 6.988652894966541, "grad_norm": 0.10402873903512955, "learning_rate": 2.1011043134834676e-05, "loss": 0.0186, "step": 24020 }, { "epoch": 6.991562409077684, "grad_norm": 0.09489603340625763, "learning_rate": 2.1003467730897374e-05, "loss": 0.0183, "step": 24030 }, { "epoch": 6.9944719231888275, "grad_norm": 0.09075503051280975, "learning_rate": 2.0995890503265086e-05, "loss": 0.02, "step": 24040 }, { "epoch": 6.997381437299971, "grad_norm": 0.13187959790229797, "learning_rate": 2.0988311454239575e-05, "loss": 0.0193, "step": 24050 }, { "epoch": 7.000290951411114, "grad_norm": 0.10344602167606354, "learning_rate": 2.098073058612316e-05, "loss": 0.02, "step": 24060 }, { "epoch": 7.003200465522258, "grad_norm": 0.10758036375045776, "learning_rate": 2.0973147901218705e-05, "loss": 0.0182, "step": 24070 }, { "epoch": 7.006109979633401, "grad_norm": 0.10706723481416702, "learning_rate": 2.0965563401829625e-05, "loss": 0.0201, "step": 24080 }, { "epoch": 7.009019493744544, "grad_norm": 0.1371404528617859, "learning_rate": 2.0957977090259897e-05, "loss": 0.0199, "step": 24090 }, { "epoch": 7.011929007855688, "grad_norm": 0.09371301531791687, "learning_rate": 2.095038896881404e-05, "loss": 0.0183, "step": 24100 }, { "epoch": 7.014838521966832, "grad_norm": 0.092584528028965, "learning_rate": 2.0942799039797128e-05, "loss": 0.0179, "step": 24110 }, { "epoch": 7.017748036077975, "grad_norm": 0.09726285189390182, "learning_rate": 2.0935207305514777e-05, "loss": 0.0187, "step": 24120 }, { "epoch": 7.020657550189118, "grad_norm": 0.12334126979112625, "learning_rate": 2.0927613768273156e-05, "loss": 0.02, "step": 24130 }, { "epoch": 7.023567064300262, "grad_norm": 0.09692539274692535, "learning_rate": 2.0920018430378984e-05, "loss": 0.02, "step": 24140 }, { "epoch": 7.026476578411406, "grad_norm": 0.11387176066637039, "learning_rate": 2.0912421294139517e-05, "loss": 0.02, "step": 24150 }, { "epoch": 7.029386092522548, "grad_norm": 0.0969589576125145, "learning_rate": 2.0904822361862572e-05, "loss": 0.0192, "step": 24160 }, { "epoch": 7.032295606633692, "grad_norm": 0.09889660030603409, "learning_rate": 2.0897221635856498e-05, "loss": 0.0189, "step": 24170 }, { "epoch": 7.035205120744836, "grad_norm": 0.12938618659973145, "learning_rate": 2.08896191184302e-05, "loss": 0.0182, "step": 24180 }, { "epoch": 7.038114634855979, "grad_norm": 0.09848909825086594, "learning_rate": 2.0882014811893124e-05, "loss": 0.0192, "step": 24190 }, { "epoch": 7.041024148967122, "grad_norm": 0.09474027901887894, "learning_rate": 2.0874408718555254e-05, "loss": 0.018, "step": 24200 }, { "epoch": 7.043933663078266, "grad_norm": 0.10977398604154587, "learning_rate": 2.0866800840727114e-05, "loss": 0.0199, "step": 24210 }, { "epoch": 7.04684317718941, "grad_norm": 0.10740996897220612, "learning_rate": 2.0859191180719786e-05, "loss": 0.0187, "step": 24220 }, { "epoch": 7.049752691300553, "grad_norm": 0.10197170823812485, "learning_rate": 2.085157974084488e-05, "loss": 0.0197, "step": 24230 }, { "epoch": 7.052662205411696, "grad_norm": 0.10517723113298416, "learning_rate": 2.0843966523414553e-05, "loss": 0.0179, "step": 24240 }, { "epoch": 7.05557171952284, "grad_norm": 0.11692392826080322, "learning_rate": 2.0836351530741492e-05, "loss": 0.0182, "step": 24250 }, { "epoch": 7.058481233633983, "grad_norm": 0.11020547896623611, "learning_rate": 2.082873476513894e-05, "loss": 0.0205, "step": 24260 }, { "epoch": 7.0613907477451265, "grad_norm": 0.09561160951852798, "learning_rate": 2.0821116228920663e-05, "loss": 0.0193, "step": 24270 }, { "epoch": 7.06430026185627, "grad_norm": 0.100818932056427, "learning_rate": 2.0813495924400977e-05, "loss": 0.0198, "step": 24280 }, { "epoch": 7.067209775967413, "grad_norm": 0.12434681504964828, "learning_rate": 2.0805873853894723e-05, "loss": 0.0207, "step": 24290 }, { "epoch": 7.070119290078557, "grad_norm": 0.11496193706989288, "learning_rate": 2.0798250019717284e-05, "loss": 0.0183, "step": 24300 }, { "epoch": 7.0730288041897005, "grad_norm": 0.11122062057256699, "learning_rate": 2.079062442418459e-05, "loss": 0.0207, "step": 24310 }, { "epoch": 7.075938318300844, "grad_norm": 0.14532941579818726, "learning_rate": 2.078299706961308e-05, "loss": 0.0206, "step": 24320 }, { "epoch": 7.078847832411987, "grad_norm": 0.10467599332332611, "learning_rate": 2.077536795831976e-05, "loss": 0.0209, "step": 24330 }, { "epoch": 7.081757346523131, "grad_norm": 0.08560017496347427, "learning_rate": 2.0767737092622137e-05, "loss": 0.0177, "step": 24340 }, { "epoch": 7.084666860634274, "grad_norm": 0.10783320665359497, "learning_rate": 2.076010447483828e-05, "loss": 0.0189, "step": 24350 }, { "epoch": 7.087576374745417, "grad_norm": 0.09925302863121033, "learning_rate": 2.0752470107286765e-05, "loss": 0.0185, "step": 24360 }, { "epoch": 7.090485888856561, "grad_norm": 0.09750796854496002, "learning_rate": 2.074483399228672e-05, "loss": 0.0203, "step": 24370 }, { "epoch": 7.093395402967705, "grad_norm": 0.11073056608438492, "learning_rate": 2.0737196132157788e-05, "loss": 0.019, "step": 24380 }, { "epoch": 7.0963049170788475, "grad_norm": 0.09801289439201355, "learning_rate": 2.0729556529220156e-05, "loss": 0.0192, "step": 24390 }, { "epoch": 7.099214431189991, "grad_norm": 0.09735113382339478, "learning_rate": 2.0721915185794532e-05, "loss": 0.0206, "step": 24400 }, { "epoch": 7.102123945301135, "grad_norm": 0.10600923746824265, "learning_rate": 2.071427210420215e-05, "loss": 0.0211, "step": 24410 }, { "epoch": 7.105033459412278, "grad_norm": 0.11624863743782043, "learning_rate": 2.070662728676478e-05, "loss": 0.0191, "step": 24420 }, { "epoch": 7.107942973523421, "grad_norm": 0.11389003694057465, "learning_rate": 2.0698980735804712e-05, "loss": 0.0188, "step": 24430 }, { "epoch": 7.110852487634565, "grad_norm": 0.10453390330076218, "learning_rate": 2.0691332453644775e-05, "loss": 0.0186, "step": 24440 }, { "epoch": 7.113762001745709, "grad_norm": 0.11993682384490967, "learning_rate": 2.0683682442608306e-05, "loss": 0.0197, "step": 24450 }, { "epoch": 7.116671515856852, "grad_norm": 0.10036895424127579, "learning_rate": 2.0676030705019182e-05, "loss": 0.0176, "step": 24460 }, { "epoch": 7.119581029967995, "grad_norm": 0.11464079469442368, "learning_rate": 2.066837724320179e-05, "loss": 0.0196, "step": 24470 }, { "epoch": 7.122490544079139, "grad_norm": 0.10519436746835709, "learning_rate": 2.0660722059481066e-05, "loss": 0.0188, "step": 24480 }, { "epoch": 7.125400058190282, "grad_norm": 0.13564078509807587, "learning_rate": 2.065306515618244e-05, "loss": 0.0185, "step": 24490 }, { "epoch": 7.128309572301426, "grad_norm": 0.11096461862325668, "learning_rate": 2.064540653563188e-05, "loss": 0.0204, "step": 24500 }, { "epoch": 7.131219086412569, "grad_norm": 0.13614259660243988, "learning_rate": 2.063774620015587e-05, "loss": 0.0199, "step": 24510 }, { "epoch": 7.134128600523712, "grad_norm": 0.11100830882787704, "learning_rate": 2.0630084152081422e-05, "loss": 0.0209, "step": 24520 }, { "epoch": 7.137038114634856, "grad_norm": 0.10330361872911453, "learning_rate": 2.0622420393736065e-05, "loss": 0.0182, "step": 24530 }, { "epoch": 7.1399476287459995, "grad_norm": 0.10812036693096161, "learning_rate": 2.061475492744784e-05, "loss": 0.0205, "step": 24540 }, { "epoch": 7.142857142857143, "grad_norm": 0.1002824455499649, "learning_rate": 2.060708775554532e-05, "loss": 0.0198, "step": 24550 }, { "epoch": 7.145766656968286, "grad_norm": 0.10758275538682938, "learning_rate": 2.0599418880357586e-05, "loss": 0.0192, "step": 24560 }, { "epoch": 7.14867617107943, "grad_norm": 0.12285541743040085, "learning_rate": 2.059174830421424e-05, "loss": 0.0184, "step": 24570 }, { "epoch": 7.1515856851905735, "grad_norm": 0.10481039434671402, "learning_rate": 2.05840760294454e-05, "loss": 0.0188, "step": 24580 }, { "epoch": 7.154495199301716, "grad_norm": 0.10197117924690247, "learning_rate": 2.05764020583817e-05, "loss": 0.0202, "step": 24590 }, { "epoch": 7.15740471341286, "grad_norm": 0.12099257856607437, "learning_rate": 2.056872639335429e-05, "loss": 0.0194, "step": 24600 }, { "epoch": 7.160314227524004, "grad_norm": 0.10770238935947418, "learning_rate": 2.056104903669484e-05, "loss": 0.0208, "step": 24610 }, { "epoch": 7.1632237416351465, "grad_norm": 0.10795408487319946, "learning_rate": 2.055336999073552e-05, "loss": 0.0182, "step": 24620 }, { "epoch": 7.16613325574629, "grad_norm": 0.10599682480096817, "learning_rate": 2.0545689257809028e-05, "loss": 0.0195, "step": 24630 }, { "epoch": 7.169042769857434, "grad_norm": 0.12991070747375488, "learning_rate": 2.053800684024856e-05, "loss": 0.0206, "step": 24640 }, { "epoch": 7.171952283968578, "grad_norm": 0.1013123169541359, "learning_rate": 2.0530322740387842e-05, "loss": 0.019, "step": 24650 }, { "epoch": 7.1748617980797205, "grad_norm": 0.10942579060792923, "learning_rate": 2.0522636960561096e-05, "loss": 0.0178, "step": 24660 }, { "epoch": 7.177771312190864, "grad_norm": 0.0710269883275032, "learning_rate": 2.0514949503103056e-05, "loss": 0.0188, "step": 24670 }, { "epoch": 7.180680826302008, "grad_norm": 0.0852358266711235, "learning_rate": 2.0507260370348968e-05, "loss": 0.0183, "step": 24680 }, { "epoch": 7.183590340413151, "grad_norm": 0.10589047521352768, "learning_rate": 2.049956956463459e-05, "loss": 0.0191, "step": 24690 }, { "epoch": 7.186499854524294, "grad_norm": 0.09973011165857315, "learning_rate": 2.049187708829619e-05, "loss": 0.0177, "step": 24700 }, { "epoch": 7.189409368635438, "grad_norm": 0.10767164826393127, "learning_rate": 2.0484182943670537e-05, "loss": 0.0201, "step": 24710 }, { "epoch": 7.192318882746581, "grad_norm": 0.08694391697645187, "learning_rate": 2.04764871330949e-05, "loss": 0.0192, "step": 24720 }, { "epoch": 7.195228396857725, "grad_norm": 0.12235952168703079, "learning_rate": 2.0468789658907075e-05, "loss": 0.0187, "step": 24730 }, { "epoch": 7.198137910968868, "grad_norm": 0.12291170656681061, "learning_rate": 2.0461090523445344e-05, "loss": 0.0177, "step": 24740 }, { "epoch": 7.201047425080012, "grad_norm": 0.08456093817949295, "learning_rate": 2.04533897290485e-05, "loss": 0.0209, "step": 24750 }, { "epoch": 7.203956939191155, "grad_norm": 0.13150067627429962, "learning_rate": 2.0445687278055852e-05, "loss": 0.0182, "step": 24760 }, { "epoch": 7.206866453302299, "grad_norm": 0.13984283804893494, "learning_rate": 2.0437983172807196e-05, "loss": 0.0173, "step": 24770 }, { "epoch": 7.209775967413442, "grad_norm": 0.08309306204319, "learning_rate": 2.043027741564283e-05, "loss": 0.0206, "step": 24780 }, { "epoch": 7.212685481524585, "grad_norm": 0.16594147682189941, "learning_rate": 2.042257000890356e-05, "loss": 0.0199, "step": 24790 }, { "epoch": 7.215594995635729, "grad_norm": 0.12454362213611603, "learning_rate": 2.0414860954930698e-05, "loss": 0.0197, "step": 24800 }, { "epoch": 7.2185045097468725, "grad_norm": 0.1188468262553215, "learning_rate": 2.0407150256066052e-05, "loss": 0.0184, "step": 24810 }, { "epoch": 7.221414023858015, "grad_norm": 0.10458897799253464, "learning_rate": 2.0399437914651927e-05, "loss": 0.0188, "step": 24820 }, { "epoch": 7.224323537969159, "grad_norm": 0.09962938725948334, "learning_rate": 2.039172393303112e-05, "loss": 0.0181, "step": 24830 }, { "epoch": 7.227233052080303, "grad_norm": 0.11371888965368271, "learning_rate": 2.0384008313546947e-05, "loss": 0.0187, "step": 24840 }, { "epoch": 7.2301425661914465, "grad_norm": 0.0958237573504448, "learning_rate": 2.0376291058543206e-05, "loss": 0.0192, "step": 24850 }, { "epoch": 7.233052080302589, "grad_norm": 0.08964123576879501, "learning_rate": 2.036857217036419e-05, "loss": 0.0172, "step": 24860 }, { "epoch": 7.235961594413733, "grad_norm": 0.10206729918718338, "learning_rate": 2.0360851651354702e-05, "loss": 0.0201, "step": 24870 }, { "epoch": 7.238871108524877, "grad_norm": 0.11407425999641418, "learning_rate": 2.0353129503860025e-05, "loss": 0.0186, "step": 24880 }, { "epoch": 7.2417806226360195, "grad_norm": 0.11346901208162308, "learning_rate": 2.034540573022595e-05, "loss": 0.0178, "step": 24890 }, { "epoch": 7.244690136747163, "grad_norm": 0.09775830060243607, "learning_rate": 2.0337680332798742e-05, "loss": 0.0189, "step": 24900 }, { "epoch": 7.247599650858307, "grad_norm": 0.10518741607666016, "learning_rate": 2.0329953313925188e-05, "loss": 0.0188, "step": 24910 }, { "epoch": 7.25050916496945, "grad_norm": 0.10111989825963974, "learning_rate": 2.032222467595255e-05, "loss": 0.0192, "step": 24920 }, { "epoch": 7.2534186790805935, "grad_norm": 0.09030096977949142, "learning_rate": 2.0314494421228577e-05, "loss": 0.0191, "step": 24930 }, { "epoch": 7.256328193191737, "grad_norm": 0.10715708881616592, "learning_rate": 2.0306762552101516e-05, "loss": 0.0198, "step": 24940 }, { "epoch": 7.25923770730288, "grad_norm": 0.12882623076438904, "learning_rate": 2.0299029070920113e-05, "loss": 0.0183, "step": 24950 }, { "epoch": 7.262147221414024, "grad_norm": 0.1085454598069191, "learning_rate": 2.0291293980033594e-05, "loss": 0.0186, "step": 24960 }, { "epoch": 7.265056735525167, "grad_norm": 0.10229602456092834, "learning_rate": 2.028355728179167e-05, "loss": 0.0196, "step": 24970 }, { "epoch": 7.267966249636311, "grad_norm": 0.11528395116329193, "learning_rate": 2.0275818978544546e-05, "loss": 0.0215, "step": 24980 }, { "epoch": 7.270875763747454, "grad_norm": 0.08701801300048828, "learning_rate": 2.026807907264292e-05, "loss": 0.0173, "step": 24990 }, { "epoch": 7.273785277858598, "grad_norm": 0.11366569995880127, "learning_rate": 2.026033756643796e-05, "loss": 0.0195, "step": 25000 }, { "epoch": 7.276694791969741, "grad_norm": 0.12425609678030014, "learning_rate": 2.0252594462281347e-05, "loss": 0.0185, "step": 25010 }, { "epoch": 7.279604306080884, "grad_norm": 0.09321565181016922, "learning_rate": 2.024484976252522e-05, "loss": 0.018, "step": 25020 }, { "epoch": 7.282513820192028, "grad_norm": 0.10275041311979294, "learning_rate": 2.0237103469522212e-05, "loss": 0.0188, "step": 25030 }, { "epoch": 7.285423334303172, "grad_norm": 0.10539267957210541, "learning_rate": 2.022935558562545e-05, "loss": 0.0202, "step": 25040 }, { "epoch": 7.288332848414315, "grad_norm": 0.08356344699859619, "learning_rate": 2.0221606113188538e-05, "loss": 0.0179, "step": 25050 }, { "epoch": 7.291242362525458, "grad_norm": 0.09129975736141205, "learning_rate": 2.0213855054565554e-05, "loss": 0.0186, "step": 25060 }, { "epoch": 7.294151876636602, "grad_norm": 0.09115494042634964, "learning_rate": 2.020610241211107e-05, "loss": 0.0192, "step": 25070 }, { "epoch": 7.2970613907477455, "grad_norm": 0.11133257299661636, "learning_rate": 2.0198348188180137e-05, "loss": 0.0198, "step": 25080 }, { "epoch": 7.299970904858888, "grad_norm": 0.10973386466503143, "learning_rate": 2.0190592385128274e-05, "loss": 0.0187, "step": 25090 }, { "epoch": 7.302880418970032, "grad_norm": 0.0959266647696495, "learning_rate": 2.018283500531149e-05, "loss": 0.0181, "step": 25100 }, { "epoch": 7.305789933081176, "grad_norm": 0.10357888042926788, "learning_rate": 2.0175076051086282e-05, "loss": 0.0189, "step": 25110 }, { "epoch": 7.308699447192319, "grad_norm": 0.09865142405033112, "learning_rate": 2.016731552480961e-05, "loss": 0.0196, "step": 25120 }, { "epoch": 7.311608961303462, "grad_norm": 0.08667062968015671, "learning_rate": 2.0159553428838914e-05, "loss": 0.016, "step": 25130 }, { "epoch": 7.314518475414606, "grad_norm": 0.09895775467157364, "learning_rate": 2.0151789765532117e-05, "loss": 0.0197, "step": 25140 }, { "epoch": 7.317427989525749, "grad_norm": 0.08041193336248398, "learning_rate": 2.0144024537247612e-05, "loss": 0.0193, "step": 25150 }, { "epoch": 7.3203375036368925, "grad_norm": 0.10360544174909592, "learning_rate": 2.013625774634428e-05, "loss": 0.0179, "step": 25160 }, { "epoch": 7.323247017748036, "grad_norm": 0.11236684024333954, "learning_rate": 2.0128489395181455e-05, "loss": 0.0176, "step": 25170 }, { "epoch": 7.32615653185918, "grad_norm": 0.083212710916996, "learning_rate": 2.0120719486118958e-05, "loss": 0.0186, "step": 25180 }, { "epoch": 7.329066045970323, "grad_norm": 0.10758993029594421, "learning_rate": 2.0112948021517093e-05, "loss": 0.0191, "step": 25190 }, { "epoch": 7.3319755600814664, "grad_norm": 0.11085674166679382, "learning_rate": 2.0105175003736615e-05, "loss": 0.018, "step": 25200 }, { "epoch": 7.33488507419261, "grad_norm": 0.10827530920505524, "learning_rate": 2.0097400435138764e-05, "loss": 0.0196, "step": 25210 }, { "epoch": 7.337794588303753, "grad_norm": 0.1022036001086235, "learning_rate": 2.0089624318085253e-05, "loss": 0.0189, "step": 25220 }, { "epoch": 7.340704102414897, "grad_norm": 0.10726906359195709, "learning_rate": 2.0081846654938256e-05, "loss": 0.018, "step": 25230 }, { "epoch": 7.34361361652604, "grad_norm": 0.1105571761727333, "learning_rate": 2.0074067448060423e-05, "loss": 0.02, "step": 25240 }, { "epoch": 7.346523130637183, "grad_norm": 0.08883800357580185, "learning_rate": 2.0066286699814873e-05, "loss": 0.0182, "step": 25250 }, { "epoch": 7.349432644748327, "grad_norm": 0.11015619337558746, "learning_rate": 2.005850441256519e-05, "loss": 0.0178, "step": 25260 }, { "epoch": 7.352342158859471, "grad_norm": 0.09130851924419403, "learning_rate": 2.0050720588675432e-05, "loss": 0.0168, "step": 25270 }, { "epoch": 7.355251672970613, "grad_norm": 0.09137305617332458, "learning_rate": 2.0042935230510113e-05, "loss": 0.0177, "step": 25280 }, { "epoch": 7.358161187081757, "grad_norm": 0.1053197905421257, "learning_rate": 2.0035148340434222e-05, "loss": 0.0188, "step": 25290 }, { "epoch": 7.361070701192901, "grad_norm": 0.10026387870311737, "learning_rate": 2.0027359920813212e-05, "loss": 0.0188, "step": 25300 }, { "epoch": 7.363980215304045, "grad_norm": 0.0875556617975235, "learning_rate": 2.0019569974012994e-05, "loss": 0.0198, "step": 25310 }, { "epoch": 7.366889729415187, "grad_norm": 0.1140732690691948, "learning_rate": 2.0011778502399954e-05, "loss": 0.0187, "step": 25320 }, { "epoch": 7.369799243526331, "grad_norm": 0.09862842410802841, "learning_rate": 2.0003985508340937e-05, "loss": 0.0208, "step": 25330 }, { "epoch": 7.372708757637475, "grad_norm": 0.11322524398565292, "learning_rate": 1.9996190994203244e-05, "loss": 0.0175, "step": 25340 }, { "epoch": 7.375618271748618, "grad_norm": 0.12777313590049744, "learning_rate": 1.998839496235464e-05, "loss": 0.0184, "step": 25350 }, { "epoch": 7.378527785859761, "grad_norm": 0.0951438695192337, "learning_rate": 1.9980597415163358e-05, "loss": 0.0193, "step": 25360 }, { "epoch": 7.381437299970905, "grad_norm": 0.0880478024482727, "learning_rate": 1.997279835499809e-05, "loss": 0.0183, "step": 25370 }, { "epoch": 7.384346814082049, "grad_norm": 0.1232195794582367, "learning_rate": 1.9964997784227976e-05, "loss": 0.0183, "step": 25380 }, { "epoch": 7.387256328193192, "grad_norm": 0.11889614909887314, "learning_rate": 1.9957195705222633e-05, "loss": 0.0194, "step": 25390 }, { "epoch": 7.390165842304335, "grad_norm": 0.09937863796949387, "learning_rate": 1.994939212035212e-05, "loss": 0.0202, "step": 25400 }, { "epoch": 7.393075356415479, "grad_norm": 0.10639068484306335, "learning_rate": 1.9941587031986963e-05, "loss": 0.0184, "step": 25410 }, { "epoch": 7.395984870526622, "grad_norm": 0.1132659912109375, "learning_rate": 1.9933780442498137e-05, "loss": 0.0187, "step": 25420 }, { "epoch": 7.3988943846377655, "grad_norm": 0.08871228247880936, "learning_rate": 1.9925972354257085e-05, "loss": 0.0191, "step": 25430 }, { "epoch": 7.401803898748909, "grad_norm": 0.10161938518285751, "learning_rate": 1.9918162769635696e-05, "loss": 0.0173, "step": 25440 }, { "epoch": 7.404713412860052, "grad_norm": 0.1108233630657196, "learning_rate": 1.9910351691006308e-05, "loss": 0.0189, "step": 25450 }, { "epoch": 7.407622926971196, "grad_norm": 0.08256623893976212, "learning_rate": 1.990253912074173e-05, "loss": 0.0181, "step": 25460 }, { "epoch": 7.410532441082339, "grad_norm": 0.10507344454526901, "learning_rate": 1.989472506121521e-05, "loss": 0.0192, "step": 25470 }, { "epoch": 7.413441955193482, "grad_norm": 0.07838868349790573, "learning_rate": 1.9886909514800454e-05, "loss": 0.018, "step": 25480 }, { "epoch": 7.416351469304626, "grad_norm": 0.08510671555995941, "learning_rate": 1.9879092483871622e-05, "loss": 0.0181, "step": 25490 }, { "epoch": 7.41926098341577, "grad_norm": 0.10052032768726349, "learning_rate": 1.987127397080332e-05, "loss": 0.0191, "step": 25500 }, { "epoch": 7.422170497526913, "grad_norm": 0.08545302599668503, "learning_rate": 1.98634539779706e-05, "loss": 0.0171, "step": 25510 }, { "epoch": 7.425080011638056, "grad_norm": 0.1106206476688385, "learning_rate": 1.985563250774897e-05, "loss": 0.0179, "step": 25520 }, { "epoch": 7.4279895257492, "grad_norm": 0.09382063150405884, "learning_rate": 1.98478095625144e-05, "loss": 0.0198, "step": 25530 }, { "epoch": 7.430899039860344, "grad_norm": 0.08904961496591568, "learning_rate": 1.9839985144643275e-05, "loss": 0.0185, "step": 25540 }, { "epoch": 7.433808553971486, "grad_norm": 0.1173681691288948, "learning_rate": 1.9832159256512463e-05, "loss": 0.0176, "step": 25550 }, { "epoch": 7.43671806808263, "grad_norm": 0.10650038719177246, "learning_rate": 1.982433190049925e-05, "loss": 0.0191, "step": 25560 }, { "epoch": 7.439627582193774, "grad_norm": 0.09013915061950684, "learning_rate": 1.981650307898138e-05, "loss": 0.0183, "step": 25570 }, { "epoch": 7.442537096304917, "grad_norm": 0.0968073308467865, "learning_rate": 1.980867279433705e-05, "loss": 0.0197, "step": 25580 }, { "epoch": 7.44544661041606, "grad_norm": 0.16721849143505096, "learning_rate": 1.98008410489449e-05, "loss": 0.021, "step": 25590 }, { "epoch": 7.448356124527204, "grad_norm": 0.08595552295446396, "learning_rate": 1.9793007845183987e-05, "loss": 0.0187, "step": 25600 }, { "epoch": 7.451265638638348, "grad_norm": 0.10311555862426758, "learning_rate": 1.9785173185433846e-05, "loss": 0.0188, "step": 25610 }, { "epoch": 7.454175152749491, "grad_norm": 0.10193890333175659, "learning_rate": 1.977733707207444e-05, "loss": 0.0201, "step": 25620 }, { "epoch": 7.457084666860634, "grad_norm": 0.08833041042089462, "learning_rate": 1.976949950748616e-05, "loss": 0.0183, "step": 25630 }, { "epoch": 7.459994180971778, "grad_norm": 0.09415630996227264, "learning_rate": 1.976166049404987e-05, "loss": 0.0179, "step": 25640 }, { "epoch": 7.462903695082921, "grad_norm": 0.08074906468391418, "learning_rate": 1.9753820034146848e-05, "loss": 0.0186, "step": 25650 }, { "epoch": 7.4658132091940645, "grad_norm": 0.12050198018550873, "learning_rate": 1.9745978130158813e-05, "loss": 0.0193, "step": 25660 }, { "epoch": 7.468722723305208, "grad_norm": 0.07507319003343582, "learning_rate": 1.9738134784467932e-05, "loss": 0.0169, "step": 25670 }, { "epoch": 7.471632237416351, "grad_norm": 0.10570202022790909, "learning_rate": 1.9730289999456804e-05, "loss": 0.0187, "step": 25680 }, { "epoch": 7.474541751527495, "grad_norm": 0.08785482496023178, "learning_rate": 1.9722443777508473e-05, "loss": 0.0182, "step": 25690 }, { "epoch": 7.4774512656386385, "grad_norm": 0.08601397275924683, "learning_rate": 1.9714596121006414e-05, "loss": 0.018, "step": 25700 }, { "epoch": 7.480360779749782, "grad_norm": 0.09405910968780518, "learning_rate": 1.9706747032334535e-05, "loss": 0.0172, "step": 25710 }, { "epoch": 7.483270293860925, "grad_norm": 0.10846952348947525, "learning_rate": 1.9698896513877177e-05, "loss": 0.0189, "step": 25720 }, { "epoch": 7.486179807972069, "grad_norm": 0.08860670775175095, "learning_rate": 1.9691044568019128e-05, "loss": 0.0182, "step": 25730 }, { "epoch": 7.489089322083212, "grad_norm": 0.10647547990083694, "learning_rate": 1.96831911971456e-05, "loss": 0.0182, "step": 25740 }, { "epoch": 7.491998836194355, "grad_norm": 0.1002979502081871, "learning_rate": 1.9675336403642234e-05, "loss": 0.0194, "step": 25750 }, { "epoch": 7.494908350305499, "grad_norm": 0.107967808842659, "learning_rate": 1.966748018989512e-05, "loss": 0.0184, "step": 25760 }, { "epoch": 7.497817864416643, "grad_norm": 0.09870414435863495, "learning_rate": 1.965962255829076e-05, "loss": 0.0184, "step": 25770 }, { "epoch": 7.5007273785277855, "grad_norm": 0.09527278691530228, "learning_rate": 1.9651763511216096e-05, "loss": 0.0187, "step": 25780 }, { "epoch": 7.503636892638929, "grad_norm": 0.09178827702999115, "learning_rate": 1.96439030510585e-05, "loss": 0.0168, "step": 25790 }, { "epoch": 7.506546406750073, "grad_norm": 0.11523718386888504, "learning_rate": 1.9636041180205774e-05, "loss": 0.0182, "step": 25800 }, { "epoch": 7.509455920861216, "grad_norm": 0.11796889454126358, "learning_rate": 1.9628177901046144e-05, "loss": 0.0187, "step": 25810 }, { "epoch": 7.512365434972359, "grad_norm": 0.10771070420742035, "learning_rate": 1.962031321596827e-05, "loss": 0.02, "step": 25820 }, { "epoch": 7.515274949083503, "grad_norm": 0.10231661796569824, "learning_rate": 1.961244712736123e-05, "loss": 0.0205, "step": 25830 }, { "epoch": 7.518184463194647, "grad_norm": 0.12648135423660278, "learning_rate": 1.9604579637614535e-05, "loss": 0.0181, "step": 25840 }, { "epoch": 7.52109397730579, "grad_norm": 0.118966244161129, "learning_rate": 1.9596710749118126e-05, "loss": 0.018, "step": 25850 }, { "epoch": 7.524003491416933, "grad_norm": 0.11914217472076416, "learning_rate": 1.9588840464262364e-05, "loss": 0.0188, "step": 25860 }, { "epoch": 7.526913005528077, "grad_norm": 0.09693727642297745, "learning_rate": 1.9580968785438026e-05, "loss": 0.0185, "step": 25870 }, { "epoch": 7.52982251963922, "grad_norm": 0.07961001992225647, "learning_rate": 1.9573095715036324e-05, "loss": 0.0181, "step": 25880 }, { "epoch": 7.532732033750364, "grad_norm": 0.09534532576799393, "learning_rate": 1.9565221255448892e-05, "loss": 0.0175, "step": 25890 }, { "epoch": 7.535641547861507, "grad_norm": 0.0854903906583786, "learning_rate": 1.9557345409067775e-05, "loss": 0.0182, "step": 25900 }, { "epoch": 7.538551061972651, "grad_norm": 0.09124147891998291, "learning_rate": 1.954946817828546e-05, "loss": 0.0156, "step": 25910 }, { "epoch": 7.541460576083794, "grad_norm": 0.08880707621574402, "learning_rate": 1.954158956549483e-05, "loss": 0.0179, "step": 25920 }, { "epoch": 7.5443700901949375, "grad_norm": 0.0873529314994812, "learning_rate": 1.9533709573089204e-05, "loss": 0.0189, "step": 25930 }, { "epoch": 7.547279604306081, "grad_norm": 0.0961451455950737, "learning_rate": 1.9525828203462313e-05, "loss": 0.0195, "step": 25940 }, { "epoch": 7.550189118417224, "grad_norm": 0.1216452494263649, "learning_rate": 1.9517945459008318e-05, "loss": 0.0167, "step": 25950 }, { "epoch": 7.553098632528368, "grad_norm": 0.0919605940580368, "learning_rate": 1.9510061342121776e-05, "loss": 0.0178, "step": 25960 }, { "epoch": 7.5560081466395115, "grad_norm": 0.09581374377012253, "learning_rate": 1.950217585519768e-05, "loss": 0.0181, "step": 25970 }, { "epoch": 7.558917660750654, "grad_norm": 0.09534551203250885, "learning_rate": 1.9494289000631437e-05, "loss": 0.0178, "step": 25980 }, { "epoch": 7.561827174861798, "grad_norm": 0.09521900117397308, "learning_rate": 1.9486400780818857e-05, "loss": 0.02, "step": 25990 }, { "epoch": 7.564736688972942, "grad_norm": 0.10976839810609818, "learning_rate": 1.9478511198156175e-05, "loss": 0.0193, "step": 26000 }, { "epoch": 7.5676462030840845, "grad_norm": 0.08325490355491638, "learning_rate": 1.9470620255040043e-05, "loss": 0.0178, "step": 26010 }, { "epoch": 7.570555717195228, "grad_norm": 0.07300552725791931, "learning_rate": 1.9462727953867515e-05, "loss": 0.0163, "step": 26020 }, { "epoch": 7.573465231306372, "grad_norm": 0.1195712685585022, "learning_rate": 1.9454834297036066e-05, "loss": 0.0183, "step": 26030 }, { "epoch": 7.576374745417516, "grad_norm": 0.12315302342176437, "learning_rate": 1.944693928694358e-05, "loss": 0.0189, "step": 26040 }, { "epoch": 7.5792842595286585, "grad_norm": 0.10115044564008713, "learning_rate": 1.943904292598835e-05, "loss": 0.0174, "step": 26050 }, { "epoch": 7.582193773639802, "grad_norm": 0.10658083856105804, "learning_rate": 1.9431145216569093e-05, "loss": 0.0181, "step": 26060 }, { "epoch": 7.585103287750946, "grad_norm": 0.109320729970932, "learning_rate": 1.942324616108491e-05, "loss": 0.0188, "step": 26070 }, { "epoch": 7.588012801862089, "grad_norm": 0.12155945599079132, "learning_rate": 1.9415345761935342e-05, "loss": 0.0175, "step": 26080 }, { "epoch": 7.590922315973232, "grad_norm": 0.10471302270889282, "learning_rate": 1.9407444021520302e-05, "loss": 0.017, "step": 26090 }, { "epoch": 7.593831830084376, "grad_norm": 0.0891629010438919, "learning_rate": 1.9399540942240144e-05, "loss": 0.017, "step": 26100 }, { "epoch": 7.59674134419552, "grad_norm": 0.09724890440702438, "learning_rate": 1.9391636526495613e-05, "loss": 0.0198, "step": 26110 }, { "epoch": 7.599650858306663, "grad_norm": 0.10163082182407379, "learning_rate": 1.9383730776687858e-05, "loss": 0.0196, "step": 26120 }, { "epoch": 7.602560372417806, "grad_norm": 0.10360337793827057, "learning_rate": 1.9375823695218437e-05, "loss": 0.0171, "step": 26130 }, { "epoch": 7.605469886528949, "grad_norm": 0.12194203585386276, "learning_rate": 1.9367915284489316e-05, "loss": 0.0169, "step": 26140 }, { "epoch": 7.608379400640093, "grad_norm": 0.1073961853981018, "learning_rate": 1.9360005546902856e-05, "loss": 0.018, "step": 26150 }, { "epoch": 7.611288914751237, "grad_norm": 0.09700575470924377, "learning_rate": 1.9352094484861836e-05, "loss": 0.0182, "step": 26160 }, { "epoch": 7.61419842886238, "grad_norm": 0.10252813249826431, "learning_rate": 1.934418210076942e-05, "loss": 0.0178, "step": 26170 }, { "epoch": 7.617107942973523, "grad_norm": 0.08260110765695572, "learning_rate": 1.9336268397029176e-05, "loss": 0.0176, "step": 26180 }, { "epoch": 7.620017457084667, "grad_norm": 0.09535125643014908, "learning_rate": 1.9328353376045096e-05, "loss": 0.0185, "step": 26190 }, { "epoch": 7.6229269711958105, "grad_norm": 0.12318351864814758, "learning_rate": 1.932043704022154e-05, "loss": 0.0197, "step": 26200 }, { "epoch": 7.625836485306953, "grad_norm": 0.09070868790149689, "learning_rate": 1.9312519391963284e-05, "loss": 0.0194, "step": 26210 }, { "epoch": 7.628745999418097, "grad_norm": 0.09664326906204224, "learning_rate": 1.9304600433675504e-05, "loss": 0.0193, "step": 26220 }, { "epoch": 7.631655513529241, "grad_norm": 0.08787550032138824, "learning_rate": 1.929668016776377e-05, "loss": 0.0186, "step": 26230 }, { "epoch": 7.6345650276403845, "grad_norm": 0.10859256982803345, "learning_rate": 1.9288758596634046e-05, "loss": 0.0172, "step": 26240 }, { "epoch": 7.637474541751527, "grad_norm": 0.11597400903701782, "learning_rate": 1.9280835722692698e-05, "loss": 0.0173, "step": 26250 }, { "epoch": 7.640384055862671, "grad_norm": 0.0858125314116478, "learning_rate": 1.9272911548346487e-05, "loss": 0.0178, "step": 26260 }, { "epoch": 7.643293569973815, "grad_norm": 0.0935552641749382, "learning_rate": 1.926498607600257e-05, "loss": 0.0179, "step": 26270 }, { "epoch": 7.6462030840849575, "grad_norm": 0.10270800441503525, "learning_rate": 1.9257059308068492e-05, "loss": 0.0178, "step": 26280 }, { "epoch": 7.649112598196101, "grad_norm": 0.0962609052658081, "learning_rate": 1.9249131246952195e-05, "loss": 0.0183, "step": 26290 }, { "epoch": 7.652022112307245, "grad_norm": 0.12079180032014847, "learning_rate": 1.924120189506202e-05, "loss": 0.0184, "step": 26300 }, { "epoch": 7.654931626418388, "grad_norm": 0.11815265566110611, "learning_rate": 1.9233271254806686e-05, "loss": 0.018, "step": 26310 }, { "epoch": 7.6578411405295315, "grad_norm": 0.13469508290290833, "learning_rate": 1.9225339328595323e-05, "loss": 0.0175, "step": 26320 }, { "epoch": 7.660750654640675, "grad_norm": 0.11682644486427307, "learning_rate": 1.9217406118837435e-05, "loss": 0.0193, "step": 26330 }, { "epoch": 7.663660168751818, "grad_norm": 0.09458595514297485, "learning_rate": 1.9209471627942922e-05, "loss": 0.0202, "step": 26340 }, { "epoch": 7.666569682862962, "grad_norm": 0.07918424159288406, "learning_rate": 1.9201535858322072e-05, "loss": 0.0192, "step": 26350 }, { "epoch": 7.669479196974105, "grad_norm": 0.13778094947338104, "learning_rate": 1.9193598812385562e-05, "loss": 0.0182, "step": 26360 }, { "epoch": 7.672388711085249, "grad_norm": 0.0996975526213646, "learning_rate": 1.9185660492544464e-05, "loss": 0.0184, "step": 26370 }, { "epoch": 7.675298225196392, "grad_norm": 0.08905719220638275, "learning_rate": 1.917772090121022e-05, "loss": 0.0181, "step": 26380 }, { "epoch": 7.678207739307536, "grad_norm": 0.11509138345718384, "learning_rate": 1.9169780040794678e-05, "loss": 0.0174, "step": 26390 }, { "epoch": 7.681117253418679, "grad_norm": 0.09209135919809341, "learning_rate": 1.916183791371005e-05, "loss": 0.0183, "step": 26400 }, { "epoch": 7.684026767529822, "grad_norm": 0.08833374828100204, "learning_rate": 1.9153894522368952e-05, "loss": 0.0172, "step": 26410 }, { "epoch": 7.686936281640966, "grad_norm": 0.09190213680267334, "learning_rate": 1.9145949869184377e-05, "loss": 0.0187, "step": 26420 }, { "epoch": 7.68984579575211, "grad_norm": 0.11465497314929962, "learning_rate": 1.91380039565697e-05, "loss": 0.0184, "step": 26430 }, { "epoch": 7.692755309863253, "grad_norm": 0.0879593938589096, "learning_rate": 1.9130056786938678e-05, "loss": 0.0176, "step": 26440 }, { "epoch": 7.695664823974396, "grad_norm": 0.11328631639480591, "learning_rate": 1.9122108362705447e-05, "loss": 0.0187, "step": 26450 }, { "epoch": 7.69857433808554, "grad_norm": 0.09368794411420822, "learning_rate": 1.9114158686284538e-05, "loss": 0.0199, "step": 26460 }, { "epoch": 7.7014838521966835, "grad_norm": 0.08923807740211487, "learning_rate": 1.910620776009085e-05, "loss": 0.016, "step": 26470 }, { "epoch": 7.704393366307826, "grad_norm": 0.119949571788311, "learning_rate": 1.9098255586539653e-05, "loss": 0.0198, "step": 26480 }, { "epoch": 7.70730288041897, "grad_norm": 0.096442811191082, "learning_rate": 1.909030216804663e-05, "loss": 0.018, "step": 26490 }, { "epoch": 7.710212394530114, "grad_norm": 0.09488483518362045, "learning_rate": 1.9082347507027797e-05, "loss": 0.0182, "step": 26500 }, { "epoch": 7.713121908641257, "grad_norm": 0.10375423729419708, "learning_rate": 1.9074391605899584e-05, "loss": 0.016, "step": 26510 }, { "epoch": 7.7160314227524, "grad_norm": 0.10804838687181473, "learning_rate": 1.906643446707878e-05, "loss": 0.0186, "step": 26520 }, { "epoch": 7.718940936863544, "grad_norm": 0.09792845696210861, "learning_rate": 1.9058476092982552e-05, "loss": 0.0198, "step": 26530 }, { "epoch": 7.721850450974687, "grad_norm": 0.12364465743303299, "learning_rate": 1.9050516486028444e-05, "loss": 0.018, "step": 26540 }, { "epoch": 7.7247599650858305, "grad_norm": 0.16547024250030518, "learning_rate": 1.9042555648634375e-05, "loss": 0.0163, "step": 26550 }, { "epoch": 7.727669479196974, "grad_norm": 0.12086207419633865, "learning_rate": 1.903459358321864e-05, "loss": 0.0182, "step": 26560 }, { "epoch": 7.730578993308118, "grad_norm": 0.10768904536962509, "learning_rate": 1.9026630292199896e-05, "loss": 0.019, "step": 26570 }, { "epoch": 7.733488507419261, "grad_norm": 0.0929616242647171, "learning_rate": 1.901866577799719e-05, "loss": 0.0171, "step": 26580 }, { "epoch": 7.7363980215304045, "grad_norm": 0.10247590392827988, "learning_rate": 1.901070004302993e-05, "loss": 0.0182, "step": 26590 }, { "epoch": 7.739307535641548, "grad_norm": 0.11909624934196472, "learning_rate": 1.900273308971789e-05, "loss": 0.0179, "step": 26600 }, { "epoch": 7.742217049752691, "grad_norm": 0.11363153159618378, "learning_rate": 1.899476492048122e-05, "loss": 0.0185, "step": 26610 }, { "epoch": 7.745126563863835, "grad_norm": 0.0806778147816658, "learning_rate": 1.8986795537740448e-05, "loss": 0.0165, "step": 26620 }, { "epoch": 7.748036077974978, "grad_norm": 0.14872755110263824, "learning_rate": 1.897882494391645e-05, "loss": 0.0193, "step": 26630 }, { "epoch": 7.750945592086122, "grad_norm": 0.11240088194608688, "learning_rate": 1.8970853141430497e-05, "loss": 0.018, "step": 26640 }, { "epoch": 7.753855106197265, "grad_norm": 0.12709923088550568, "learning_rate": 1.89628801327042e-05, "loss": 0.0184, "step": 26650 }, { "epoch": 7.756764620308409, "grad_norm": 0.12162839621305466, "learning_rate": 1.8954905920159555e-05, "loss": 0.02, "step": 26660 }, { "epoch": 7.7596741344195515, "grad_norm": 0.1070476695895195, "learning_rate": 1.8946930506218916e-05, "loss": 0.0202, "step": 26670 }, { "epoch": 7.762583648530695, "grad_norm": 0.08539938926696777, "learning_rate": 1.8938953893305e-05, "loss": 0.0179, "step": 26680 }, { "epoch": 7.765493162641839, "grad_norm": 0.08093903213739395, "learning_rate": 1.893097608384089e-05, "loss": 0.0186, "step": 26690 }, { "epoch": 7.768402676752983, "grad_norm": 0.10279754549264908, "learning_rate": 1.892299708025004e-05, "loss": 0.017, "step": 26700 }, { "epoch": 7.771312190864125, "grad_norm": 0.11166024208068848, "learning_rate": 1.891501688495626e-05, "loss": 0.019, "step": 26710 }, { "epoch": 7.774221704975269, "grad_norm": 0.10509093850851059, "learning_rate": 1.8907035500383717e-05, "loss": 0.0181, "step": 26720 }, { "epoch": 7.777131219086413, "grad_norm": 0.09372425079345703, "learning_rate": 1.889905292895695e-05, "loss": 0.0177, "step": 26730 }, { "epoch": 7.780040733197556, "grad_norm": 0.08886196464300156, "learning_rate": 1.889106917310085e-05, "loss": 0.0172, "step": 26740 }, { "epoch": 7.782950247308699, "grad_norm": 0.1343131810426712, "learning_rate": 1.8883084235240672e-05, "loss": 0.0175, "step": 26750 }, { "epoch": 7.785859761419843, "grad_norm": 0.09447109699249268, "learning_rate": 1.887509811780203e-05, "loss": 0.0184, "step": 26760 }, { "epoch": 7.788769275530987, "grad_norm": 0.0941811203956604, "learning_rate": 1.88671108232109e-05, "loss": 0.0175, "step": 26770 }, { "epoch": 7.79167878964213, "grad_norm": 0.0986538678407669, "learning_rate": 1.88591223538936e-05, "loss": 0.0185, "step": 26780 }, { "epoch": 7.794588303753273, "grad_norm": 0.0838734582066536, "learning_rate": 1.885113271227683e-05, "loss": 0.0185, "step": 26790 }, { "epoch": 7.797497817864417, "grad_norm": 0.0863511711359024, "learning_rate": 1.8843141900787625e-05, "loss": 0.0192, "step": 26800 }, { "epoch": 7.80040733197556, "grad_norm": 0.15601208806037903, "learning_rate": 1.8835149921853383e-05, "loss": 0.018, "step": 26810 }, { "epoch": 7.8033168460867035, "grad_norm": 0.11191179603338242, "learning_rate": 1.8827156777901853e-05, "loss": 0.019, "step": 26820 }, { "epoch": 7.806226360197847, "grad_norm": 0.09862468391656876, "learning_rate": 1.8819162471361144e-05, "loss": 0.017, "step": 26830 }, { "epoch": 7.80913587430899, "grad_norm": 0.10074210166931152, "learning_rate": 1.8811167004659715e-05, "loss": 0.0188, "step": 26840 }, { "epoch": 7.812045388420134, "grad_norm": 0.1048240140080452, "learning_rate": 1.8803170380226383e-05, "loss": 0.0172, "step": 26850 }, { "epoch": 7.8149549025312774, "grad_norm": 0.11028654128313065, "learning_rate": 1.8795172600490302e-05, "loss": 0.0178, "step": 26860 }, { "epoch": 7.81786441664242, "grad_norm": 0.088680200278759, "learning_rate": 1.878717366788099e-05, "loss": 0.0181, "step": 26870 }, { "epoch": 7.820773930753564, "grad_norm": 0.09233999252319336, "learning_rate": 1.8779173584828316e-05, "loss": 0.0169, "step": 26880 }, { "epoch": 7.823683444864708, "grad_norm": 0.10843590646982193, "learning_rate": 1.8771172353762488e-05, "loss": 0.0195, "step": 26890 }, { "epoch": 7.826592958975851, "grad_norm": 0.08610621839761734, "learning_rate": 1.876316997711407e-05, "loss": 0.0168, "step": 26900 }, { "epoch": 7.829502473086994, "grad_norm": 0.11166226118803024, "learning_rate": 1.8755166457313972e-05, "loss": 0.0182, "step": 26910 }, { "epoch": 7.832411987198138, "grad_norm": 0.10293586552143097, "learning_rate": 1.8747161796793457e-05, "loss": 0.0164, "step": 26920 }, { "epoch": 7.835321501309282, "grad_norm": 0.07917263358831406, "learning_rate": 1.8739155997984124e-05, "loss": 0.017, "step": 26930 }, { "epoch": 7.838231015420424, "grad_norm": 0.09843166917562485, "learning_rate": 1.8731149063317927e-05, "loss": 0.0184, "step": 26940 }, { "epoch": 7.841140529531568, "grad_norm": 0.12075354903936386, "learning_rate": 1.872314099522716e-05, "loss": 0.0183, "step": 26950 }, { "epoch": 7.844050043642712, "grad_norm": 0.13660551607608795, "learning_rate": 1.8715131796144466e-05, "loss": 0.0194, "step": 26960 }, { "epoch": 7.846959557753856, "grad_norm": 0.0972733423113823, "learning_rate": 1.870712146850282e-05, "loss": 0.0187, "step": 26970 }, { "epoch": 7.849869071864998, "grad_norm": 0.11214591562747955, "learning_rate": 1.8699110014735554e-05, "loss": 0.0176, "step": 26980 }, { "epoch": 7.852778585976142, "grad_norm": 0.13233350217342377, "learning_rate": 1.8691097437276334e-05, "loss": 0.0195, "step": 26990 }, { "epoch": 7.855688100087286, "grad_norm": 0.11640062183141708, "learning_rate": 1.868308373855917e-05, "loss": 0.0184, "step": 27000 }, { "epoch": 7.858597614198429, "grad_norm": 0.09987052530050278, "learning_rate": 1.8675068921018415e-05, "loss": 0.0183, "step": 27010 }, { "epoch": 7.861507128309572, "grad_norm": 0.1399647295475006, "learning_rate": 1.8667052987088756e-05, "loss": 0.0181, "step": 27020 }, { "epoch": 7.864416642420716, "grad_norm": 0.11728379130363464, "learning_rate": 1.8659035939205224e-05, "loss": 0.0175, "step": 27030 }, { "epoch": 7.867326156531859, "grad_norm": 0.10932213813066483, "learning_rate": 1.8651017779803178e-05, "loss": 0.0185, "step": 27040 }, { "epoch": 7.870235670643003, "grad_norm": 0.08787412196397781, "learning_rate": 1.8642998511318336e-05, "loss": 0.0184, "step": 27050 }, { "epoch": 7.873145184754146, "grad_norm": 0.0981983095407486, "learning_rate": 1.863497813618673e-05, "loss": 0.0181, "step": 27060 }, { "epoch": 7.876054698865289, "grad_norm": 0.10203482955694199, "learning_rate": 1.8626956656844746e-05, "loss": 0.0204, "step": 27070 }, { "epoch": 7.878964212976433, "grad_norm": 0.0976995974779129, "learning_rate": 1.861893407572909e-05, "loss": 0.017, "step": 27080 }, { "epoch": 7.8818737270875765, "grad_norm": 0.10040596127510071, "learning_rate": 1.861091039527682e-05, "loss": 0.0187, "step": 27090 }, { "epoch": 7.88478324119872, "grad_norm": 0.09434287250041962, "learning_rate": 1.8602885617925307e-05, "loss": 0.0165, "step": 27100 }, { "epoch": 7.887692755309863, "grad_norm": 0.11277969926595688, "learning_rate": 1.8594859746112273e-05, "loss": 0.0176, "step": 27110 }, { "epoch": 7.890602269421007, "grad_norm": 0.09640557318925858, "learning_rate": 1.8586832782275768e-05, "loss": 0.0171, "step": 27120 }, { "epoch": 7.89351178353215, "grad_norm": 0.09062476456165314, "learning_rate": 1.857880472885417e-05, "loss": 0.0193, "step": 27130 }, { "epoch": 7.896421297643293, "grad_norm": 0.10349342972040176, "learning_rate": 1.8570775588286185e-05, "loss": 0.0176, "step": 27140 }, { "epoch": 7.899330811754437, "grad_norm": 0.08524607867002487, "learning_rate": 1.8562745363010854e-05, "loss": 0.0174, "step": 27150 }, { "epoch": 7.902240325865581, "grad_norm": 0.08830724656581879, "learning_rate": 1.8554714055467555e-05, "loss": 0.0174, "step": 27160 }, { "epoch": 7.905149839976724, "grad_norm": 0.10808826237916946, "learning_rate": 1.8546681668095978e-05, "loss": 0.0191, "step": 27170 }, { "epoch": 7.908059354087867, "grad_norm": 0.09322279691696167, "learning_rate": 1.853864820333616e-05, "loss": 0.0187, "step": 27180 }, { "epoch": 7.910968868199011, "grad_norm": 0.09323714673519135, "learning_rate": 1.8530613663628448e-05, "loss": 0.0169, "step": 27190 }, { "epoch": 7.913878382310154, "grad_norm": 0.11480075120925903, "learning_rate": 1.8522578051413525e-05, "loss": 0.0166, "step": 27200 }, { "epoch": 7.916787896421297, "grad_norm": 0.1001242846250534, "learning_rate": 1.85145413691324e-05, "loss": 0.0181, "step": 27210 }, { "epoch": 7.919697410532441, "grad_norm": 0.0847363993525505, "learning_rate": 1.8506503619226402e-05, "loss": 0.0173, "step": 27220 }, { "epoch": 7.922606924643585, "grad_norm": 0.0959823727607727, "learning_rate": 1.849846480413719e-05, "loss": 0.0192, "step": 27230 }, { "epoch": 7.925516438754728, "grad_norm": 0.10589879006147385, "learning_rate": 1.8490424926306745e-05, "loss": 0.0178, "step": 27240 }, { "epoch": 7.928425952865871, "grad_norm": 0.08337391167879105, "learning_rate": 1.8482383988177362e-05, "loss": 0.0181, "step": 27250 }, { "epoch": 7.931335466977015, "grad_norm": 0.09116851538419724, "learning_rate": 1.8474341992191666e-05, "loss": 0.0167, "step": 27260 }, { "epoch": 7.934244981088158, "grad_norm": 0.08363956212997437, "learning_rate": 1.8466298940792614e-05, "loss": 0.0165, "step": 27270 }, { "epoch": 7.937154495199302, "grad_norm": 0.10729952901601791, "learning_rate": 1.8458254836423467e-05, "loss": 0.0185, "step": 27280 }, { "epoch": 7.940064009310445, "grad_norm": 0.08711615949869156, "learning_rate": 1.8450209681527808e-05, "loss": 0.0172, "step": 27290 }, { "epoch": 7.942973523421589, "grad_norm": 0.08852975070476532, "learning_rate": 1.8442163478549546e-05, "loss": 0.0173, "step": 27300 }, { "epoch": 7.945883037532732, "grad_norm": 0.08591364324092865, "learning_rate": 1.8434116229932905e-05, "loss": 0.0184, "step": 27310 }, { "epoch": 7.9487925516438755, "grad_norm": 0.08399909734725952, "learning_rate": 1.8426067938122423e-05, "loss": 0.0176, "step": 27320 }, { "epoch": 7.951702065755019, "grad_norm": 0.13749806582927704, "learning_rate": 1.841801860556296e-05, "loss": 0.0181, "step": 27330 }, { "epoch": 7.954611579866162, "grad_norm": 0.0959157794713974, "learning_rate": 1.8409968234699698e-05, "loss": 0.0169, "step": 27340 }, { "epoch": 7.957521093977306, "grad_norm": 0.09669391810894012, "learning_rate": 1.840191682797812e-05, "loss": 0.0184, "step": 27350 }, { "epoch": 7.9604306080884495, "grad_norm": 0.09930824488401413, "learning_rate": 1.839386438784403e-05, "loss": 0.0172, "step": 27360 }, { "epoch": 7.963340122199592, "grad_norm": 0.10460752993822098, "learning_rate": 1.838581091674355e-05, "loss": 0.0171, "step": 27370 }, { "epoch": 7.966249636310736, "grad_norm": 0.13087524473667145, "learning_rate": 1.8377756417123108e-05, "loss": 0.0182, "step": 27380 }, { "epoch": 7.96915915042188, "grad_norm": 0.14603254199028015, "learning_rate": 1.8369700891429455e-05, "loss": 0.0189, "step": 27390 }, { "epoch": 7.9720686645330225, "grad_norm": 0.09328558295965195, "learning_rate": 1.8361644342109642e-05, "loss": 0.0166, "step": 27400 }, { "epoch": 7.974978178644166, "grad_norm": 0.11740151792764664, "learning_rate": 1.8353586771611038e-05, "loss": 0.0182, "step": 27410 }, { "epoch": 7.97788769275531, "grad_norm": 0.10162795335054398, "learning_rate": 1.8345528182381314e-05, "loss": 0.0188, "step": 27420 }, { "epoch": 7.980797206866454, "grad_norm": 0.11756797879934311, "learning_rate": 1.8337468576868466e-05, "loss": 0.017, "step": 27430 }, { "epoch": 7.9837067209775965, "grad_norm": 0.11663879454135895, "learning_rate": 1.8329407957520786e-05, "loss": 0.0182, "step": 27440 }, { "epoch": 7.98661623508874, "grad_norm": 0.0960971936583519, "learning_rate": 1.8321346326786872e-05, "loss": 0.0189, "step": 27450 }, { "epoch": 7.989525749199884, "grad_norm": 0.0808597281575203, "learning_rate": 1.831328368711564e-05, "loss": 0.0171, "step": 27460 }, { "epoch": 7.992435263311027, "grad_norm": 0.1209394559264183, "learning_rate": 1.830522004095631e-05, "loss": 0.0188, "step": 27470 }, { "epoch": 7.99534477742217, "grad_norm": 0.09016517549753189, "learning_rate": 1.8297155390758394e-05, "loss": 0.018, "step": 27480 }, { "epoch": 7.998254291533314, "grad_norm": 0.09578903764486313, "learning_rate": 1.828908973897173e-05, "loss": 0.0166, "step": 27490 }, { "epoch": 8.001163805644458, "grad_norm": 0.08453914523124695, "learning_rate": 1.828102308804645e-05, "loss": 0.0155, "step": 27500 }, { "epoch": 8.004073319755602, "grad_norm": 0.11035636812448502, "learning_rate": 1.827295544043299e-05, "loss": 0.0184, "step": 27510 }, { "epoch": 8.006982833866743, "grad_norm": 0.09150513261556625, "learning_rate": 1.826488679858207e-05, "loss": 0.0184, "step": 27520 }, { "epoch": 8.009892347977887, "grad_norm": 0.09739593416452408, "learning_rate": 1.8256817164944756e-05, "loss": 0.0187, "step": 27530 }, { "epoch": 8.012801862089031, "grad_norm": 0.09586352854967117, "learning_rate": 1.8248746541972377e-05, "loss": 0.0183, "step": 27540 }, { "epoch": 8.015711376200175, "grad_norm": 0.14155267179012299, "learning_rate": 1.8240674932116575e-05, "loss": 0.0187, "step": 27550 }, { "epoch": 8.018620890311318, "grad_norm": 0.09532133489847183, "learning_rate": 1.8232602337829297e-05, "loss": 0.0163, "step": 27560 }, { "epoch": 8.021530404422462, "grad_norm": 0.09516549110412598, "learning_rate": 1.8224528761562777e-05, "loss": 0.0186, "step": 27570 }, { "epoch": 8.024439918533606, "grad_norm": 0.10749347507953644, "learning_rate": 1.821645420576956e-05, "loss": 0.0183, "step": 27580 }, { "epoch": 8.027349432644748, "grad_norm": 0.10674332082271576, "learning_rate": 1.820837867290248e-05, "loss": 0.0168, "step": 27590 }, { "epoch": 8.030258946755891, "grad_norm": 0.1167605072259903, "learning_rate": 1.820030216541467e-05, "loss": 0.0195, "step": 27600 }, { "epoch": 8.033168460867035, "grad_norm": 0.09266718477010727, "learning_rate": 1.8192224685759562e-05, "loss": 0.0177, "step": 27610 }, { "epoch": 8.036077974978179, "grad_norm": 0.10169215500354767, "learning_rate": 1.8184146236390883e-05, "loss": 0.0183, "step": 27620 }, { "epoch": 8.038987489089322, "grad_norm": 0.092675119638443, "learning_rate": 1.8176066819762644e-05, "loss": 0.0167, "step": 27630 }, { "epoch": 8.041897003200466, "grad_norm": 0.10127676278352737, "learning_rate": 1.8167986438329167e-05, "loss": 0.0166, "step": 27640 }, { "epoch": 8.044806517311608, "grad_norm": 0.11780785024166107, "learning_rate": 1.815990509454506e-05, "loss": 0.0181, "step": 27650 }, { "epoch": 8.047716031422752, "grad_norm": 0.10046444833278656, "learning_rate": 1.8151822790865218e-05, "loss": 0.0194, "step": 27660 }, { "epoch": 8.050625545533896, "grad_norm": 0.08564100414514542, "learning_rate": 1.8143739529744828e-05, "loss": 0.0172, "step": 27670 }, { "epoch": 8.05353505964504, "grad_norm": 0.11875355988740921, "learning_rate": 1.813565531363937e-05, "loss": 0.0181, "step": 27680 }, { "epoch": 8.056444573756183, "grad_norm": 0.10112699121236801, "learning_rate": 1.8127570145004628e-05, "loss": 0.0178, "step": 27690 }, { "epoch": 8.059354087867327, "grad_norm": 0.09303989261388779, "learning_rate": 1.811948402629665e-05, "loss": 0.0187, "step": 27700 }, { "epoch": 8.06226360197847, "grad_norm": 0.11208964884281158, "learning_rate": 1.811139695997179e-05, "loss": 0.0197, "step": 27710 }, { "epoch": 8.065173116089612, "grad_norm": 0.1026201918721199, "learning_rate": 1.810330894848669e-05, "loss": 0.0166, "step": 27720 }, { "epoch": 8.068082630200756, "grad_norm": 0.1011146828532219, "learning_rate": 1.8095219994298266e-05, "loss": 0.0165, "step": 27730 }, { "epoch": 8.0709921443119, "grad_norm": 0.10572229325771332, "learning_rate": 1.8087130099863728e-05, "loss": 0.0179, "step": 27740 }, { "epoch": 8.073901658423043, "grad_norm": 0.1282251924276352, "learning_rate": 1.8079039267640583e-05, "loss": 0.0178, "step": 27750 }, { "epoch": 8.076811172534187, "grad_norm": 0.11979338526725769, "learning_rate": 1.80709475000866e-05, "loss": 0.0188, "step": 27760 }, { "epoch": 8.07972068664533, "grad_norm": 0.09913320094347, "learning_rate": 1.8062854799659855e-05, "loss": 0.0183, "step": 27770 }, { "epoch": 8.082630200756475, "grad_norm": 0.1112154871225357, "learning_rate": 1.805476116881869e-05, "loss": 0.0175, "step": 27780 }, { "epoch": 8.085539714867616, "grad_norm": 0.11100530624389648, "learning_rate": 1.804666661002174e-05, "loss": 0.0184, "step": 27790 }, { "epoch": 8.08844922897876, "grad_norm": 0.09311731904745102, "learning_rate": 1.8038571125727917e-05, "loss": 0.0174, "step": 27800 }, { "epoch": 8.091358743089904, "grad_norm": 0.10160970687866211, "learning_rate": 1.8030474718396414e-05, "loss": 0.0187, "step": 27810 }, { "epoch": 8.094268257201048, "grad_norm": 0.1231425553560257, "learning_rate": 1.8022377390486704e-05, "loss": 0.0193, "step": 27820 }, { "epoch": 8.097177771312191, "grad_norm": 0.1304127275943756, "learning_rate": 1.8014279144458546e-05, "loss": 0.0176, "step": 27830 }, { "epoch": 8.100087285423335, "grad_norm": 0.10114669799804688, "learning_rate": 1.800617998277197e-05, "loss": 0.0177, "step": 27840 }, { "epoch": 8.102996799534477, "grad_norm": 0.10051921755075455, "learning_rate": 1.799807990788729e-05, "loss": 0.0178, "step": 27850 }, { "epoch": 8.10590631364562, "grad_norm": 0.09664566069841385, "learning_rate": 1.7989978922265094e-05, "loss": 0.0172, "step": 27860 }, { "epoch": 8.108815827756764, "grad_norm": 0.0945870578289032, "learning_rate": 1.7981877028366248e-05, "loss": 0.0189, "step": 27870 }, { "epoch": 8.111725341867908, "grad_norm": 0.09279078245162964, "learning_rate": 1.797377422865189e-05, "loss": 0.0173, "step": 27880 }, { "epoch": 8.114634855979052, "grad_norm": 0.15204831957817078, "learning_rate": 1.796567052558344e-05, "loss": 0.0177, "step": 27890 }, { "epoch": 8.117544370090195, "grad_norm": 0.11262819170951843, "learning_rate": 1.795756592162259e-05, "loss": 0.0172, "step": 27900 }, { "epoch": 8.12045388420134, "grad_norm": 0.09878537803888321, "learning_rate": 1.79494604192313e-05, "loss": 0.0168, "step": 27910 }, { "epoch": 8.123363398312481, "grad_norm": 0.12260572612285614, "learning_rate": 1.7941354020871815e-05, "loss": 0.0173, "step": 27920 }, { "epoch": 8.126272912423625, "grad_norm": 0.09950241446495056, "learning_rate": 1.793324672900664e-05, "loss": 0.017, "step": 27930 }, { "epoch": 8.129182426534769, "grad_norm": 0.09466820955276489, "learning_rate": 1.7925138546098553e-05, "loss": 0.018, "step": 27940 }, { "epoch": 8.132091940645912, "grad_norm": 0.07884930819272995, "learning_rate": 1.791702947461061e-05, "loss": 0.0156, "step": 27950 }, { "epoch": 8.135001454757056, "grad_norm": 0.1018683910369873, "learning_rate": 1.7908919517006135e-05, "loss": 0.0175, "step": 27960 }, { "epoch": 8.1379109688682, "grad_norm": 0.1448395699262619, "learning_rate": 1.7900808675748714e-05, "loss": 0.0177, "step": 27970 }, { "epoch": 8.140820482979342, "grad_norm": 0.12960192561149597, "learning_rate": 1.7892696953302214e-05, "loss": 0.0172, "step": 27980 }, { "epoch": 8.143729997090485, "grad_norm": 0.0836184173822403, "learning_rate": 1.7884584352130747e-05, "loss": 0.0178, "step": 27990 }, { "epoch": 8.146639511201629, "grad_norm": 0.11783985048532486, "learning_rate": 1.7876470874698722e-05, "loss": 0.0185, "step": 28000 }, { "epoch": 8.149549025312773, "grad_norm": 0.10058765858411789, "learning_rate": 1.7868356523470796e-05, "loss": 0.0173, "step": 28010 }, { "epoch": 8.152458539423916, "grad_norm": 0.11390881985425949, "learning_rate": 1.7860241300911895e-05, "loss": 0.0171, "step": 28020 }, { "epoch": 8.15536805353506, "grad_norm": 0.10457662492990494, "learning_rate": 1.7852125209487202e-05, "loss": 0.0191, "step": 28030 }, { "epoch": 8.158277567646204, "grad_norm": 0.10573966056108475, "learning_rate": 1.7844008251662175e-05, "loss": 0.0196, "step": 28040 }, { "epoch": 8.161187081757346, "grad_norm": 0.11177290230989456, "learning_rate": 1.7835890429902535e-05, "loss": 0.0188, "step": 28050 }, { "epoch": 8.16409659586849, "grad_norm": 0.15932805836200714, "learning_rate": 1.7827771746674257e-05, "loss": 0.0183, "step": 28060 }, { "epoch": 8.167006109979633, "grad_norm": 0.08148003369569778, "learning_rate": 1.7819652204443593e-05, "loss": 0.0181, "step": 28070 }, { "epoch": 8.169915624090777, "grad_norm": 0.12358573079109192, "learning_rate": 1.7811531805677033e-05, "loss": 0.0189, "step": 28080 }, { "epoch": 8.17282513820192, "grad_norm": 0.07478731870651245, "learning_rate": 1.7803410552841346e-05, "loss": 0.0154, "step": 28090 }, { "epoch": 8.175734652313064, "grad_norm": 0.09586742520332336, "learning_rate": 1.7795288448403548e-05, "loss": 0.018, "step": 28100 }, { "epoch": 8.178644166424208, "grad_norm": 0.08029922097921371, "learning_rate": 1.7787165494830928e-05, "loss": 0.0177, "step": 28110 }, { "epoch": 8.18155368053535, "grad_norm": 0.1057315543293953, "learning_rate": 1.7779041694591024e-05, "loss": 0.02, "step": 28120 }, { "epoch": 8.184463194646494, "grad_norm": 0.09866666048765182, "learning_rate": 1.7770917050151628e-05, "loss": 0.0172, "step": 28130 }, { "epoch": 8.187372708757637, "grad_norm": 0.12594203650951385, "learning_rate": 1.7762791563980795e-05, "loss": 0.0188, "step": 28140 }, { "epoch": 8.190282222868781, "grad_norm": 0.08693578839302063, "learning_rate": 1.7754665238546828e-05, "loss": 0.0177, "step": 28150 }, { "epoch": 8.193191736979925, "grad_norm": 0.09945548325777054, "learning_rate": 1.77465380763183e-05, "loss": 0.0185, "step": 28160 }, { "epoch": 8.196101251091068, "grad_norm": 0.08432228863239288, "learning_rate": 1.773841007976402e-05, "loss": 0.0193, "step": 28170 }, { "epoch": 8.19901076520221, "grad_norm": 0.09602271765470505, "learning_rate": 1.773028125135306e-05, "loss": 0.0173, "step": 28180 }, { "epoch": 8.201920279313354, "grad_norm": 0.11894898116588593, "learning_rate": 1.7722151593554743e-05, "loss": 0.0201, "step": 28190 }, { "epoch": 8.204829793424498, "grad_norm": 0.09142159670591354, "learning_rate": 1.7714021108838652e-05, "loss": 0.0162, "step": 28200 }, { "epoch": 8.207739307535642, "grad_norm": 0.10460410267114639, "learning_rate": 1.77058897996746e-05, "loss": 0.0177, "step": 28210 }, { "epoch": 8.210648821646785, "grad_norm": 0.12631194293498993, "learning_rate": 1.7697757668532677e-05, "loss": 0.0191, "step": 28220 }, { "epoch": 8.213558335757929, "grad_norm": 0.08267771452665329, "learning_rate": 1.7689624717883204e-05, "loss": 0.0179, "step": 28230 }, { "epoch": 8.216467849869073, "grad_norm": 0.08747711777687073, "learning_rate": 1.768149095019676e-05, "loss": 0.0175, "step": 28240 }, { "epoch": 8.219377363980215, "grad_norm": 0.08848028630018234, "learning_rate": 1.7673356367944162e-05, "loss": 0.0166, "step": 28250 }, { "epoch": 8.222286878091358, "grad_norm": 0.0906011089682579, "learning_rate": 1.7665220973596484e-05, "loss": 0.0153, "step": 28260 }, { "epoch": 8.225196392202502, "grad_norm": 0.08501096814870834, "learning_rate": 1.7657084769625046e-05, "loss": 0.0185, "step": 28270 }, { "epoch": 8.228105906313646, "grad_norm": 0.10151166468858719, "learning_rate": 1.7648947758501414e-05, "loss": 0.0182, "step": 28280 }, { "epoch": 8.23101542042479, "grad_norm": 0.09232232719659805, "learning_rate": 1.7640809942697394e-05, "loss": 0.0164, "step": 28290 }, { "epoch": 8.233924934535933, "grad_norm": 0.12740649282932281, "learning_rate": 1.7632671324685037e-05, "loss": 0.018, "step": 28300 }, { "epoch": 8.236834448647077, "grad_norm": 0.11008419841527939, "learning_rate": 1.7624531906936647e-05, "loss": 0.0167, "step": 28310 }, { "epoch": 8.239743962758219, "grad_norm": 0.11592330038547516, "learning_rate": 1.761639169192476e-05, "loss": 0.0186, "step": 28320 }, { "epoch": 8.242653476869362, "grad_norm": 0.08655289560556412, "learning_rate": 1.7608250682122157e-05, "loss": 0.0171, "step": 28330 }, { "epoch": 8.245562990980506, "grad_norm": 0.10038162767887115, "learning_rate": 1.760010888000186e-05, "loss": 0.0185, "step": 28340 }, { "epoch": 8.24847250509165, "grad_norm": 0.12287943065166473, "learning_rate": 1.7591966288037142e-05, "loss": 0.0162, "step": 28350 }, { "epoch": 8.251382019202794, "grad_norm": 0.08714531362056732, "learning_rate": 1.75838229087015e-05, "loss": 0.0175, "step": 28360 }, { "epoch": 8.254291533313937, "grad_norm": 0.15841227769851685, "learning_rate": 1.7575678744468684e-05, "loss": 0.0189, "step": 28370 }, { "epoch": 8.25720104742508, "grad_norm": 0.10444694757461548, "learning_rate": 1.756753379781267e-05, "loss": 0.0178, "step": 28380 }, { "epoch": 8.260110561536223, "grad_norm": 0.10371960699558258, "learning_rate": 1.7559388071207682e-05, "loss": 0.0174, "step": 28390 }, { "epoch": 8.263020075647367, "grad_norm": 0.09994588047266006, "learning_rate": 1.7551241567128173e-05, "loss": 0.0159, "step": 28400 }, { "epoch": 8.26592958975851, "grad_norm": 0.10984156280755997, "learning_rate": 1.7543094288048836e-05, "loss": 0.0181, "step": 28410 }, { "epoch": 8.268839103869654, "grad_norm": 0.11108235269784927, "learning_rate": 1.7534946236444602e-05, "loss": 0.0169, "step": 28420 }, { "epoch": 8.271748617980798, "grad_norm": 0.09497318416833878, "learning_rate": 1.7526797414790633e-05, "loss": 0.0182, "step": 28430 }, { "epoch": 8.274658132091941, "grad_norm": 0.09293806552886963, "learning_rate": 1.751864782556233e-05, "loss": 0.0172, "step": 28440 }, { "epoch": 8.277567646203083, "grad_norm": 0.09661752730607986, "learning_rate": 1.7510497471235314e-05, "loss": 0.0161, "step": 28450 }, { "epoch": 8.280477160314227, "grad_norm": 0.10834655910730362, "learning_rate": 1.7502346354285455e-05, "loss": 0.0176, "step": 28460 }, { "epoch": 8.28338667442537, "grad_norm": 0.09892911463975906, "learning_rate": 1.7494194477188838e-05, "loss": 0.017, "step": 28470 }, { "epoch": 8.286296188536515, "grad_norm": 0.11716172099113464, "learning_rate": 1.74860418424218e-05, "loss": 0.0186, "step": 28480 }, { "epoch": 8.289205702647658, "grad_norm": 0.1172880306839943, "learning_rate": 1.7477888452460885e-05, "loss": 0.0171, "step": 28490 }, { "epoch": 8.292115216758802, "grad_norm": 0.10042128711938858, "learning_rate": 1.746973430978289e-05, "loss": 0.0175, "step": 28500 }, { "epoch": 8.295024730869944, "grad_norm": 0.1140235960483551, "learning_rate": 1.7461579416864814e-05, "loss": 0.0173, "step": 28510 }, { "epoch": 8.297934244981088, "grad_norm": 0.10567872226238251, "learning_rate": 1.7453423776183907e-05, "loss": 0.0168, "step": 28520 }, { "epoch": 8.300843759092231, "grad_norm": 0.08637983351945877, "learning_rate": 1.744526739021764e-05, "loss": 0.0177, "step": 28530 }, { "epoch": 8.303753273203375, "grad_norm": 0.09941034018993378, "learning_rate": 1.74371102614437e-05, "loss": 0.0172, "step": 28540 }, { "epoch": 8.306662787314519, "grad_norm": 0.10909367352724075, "learning_rate": 1.742895239234001e-05, "loss": 0.0162, "step": 28550 }, { "epoch": 8.309572301425662, "grad_norm": 0.08929133415222168, "learning_rate": 1.742079378538472e-05, "loss": 0.0187, "step": 28560 }, { "epoch": 8.312481815536806, "grad_norm": 0.10003723204135895, "learning_rate": 1.7412634443056192e-05, "loss": 0.0178, "step": 28570 }, { "epoch": 8.315391329647948, "grad_norm": 0.13390213251113892, "learning_rate": 1.7404474367833024e-05, "loss": 0.0168, "step": 28580 }, { "epoch": 8.318300843759092, "grad_norm": 0.11792588979005814, "learning_rate": 1.739631356219403e-05, "loss": 0.0184, "step": 28590 }, { "epoch": 8.321210357870235, "grad_norm": 0.09419289976358414, "learning_rate": 1.7388152028618244e-05, "loss": 0.0169, "step": 28600 }, { "epoch": 8.32411987198138, "grad_norm": 0.1102866604924202, "learning_rate": 1.737998976958493e-05, "loss": 0.0173, "step": 28610 }, { "epoch": 8.327029386092523, "grad_norm": 0.11945035308599472, "learning_rate": 1.7371826787573562e-05, "loss": 0.0198, "step": 28620 }, { "epoch": 8.329938900203667, "grad_norm": 0.12031900137662888, "learning_rate": 1.736366308506384e-05, "loss": 0.0164, "step": 28630 }, { "epoch": 8.33284841431481, "grad_norm": 0.10384706407785416, "learning_rate": 1.7355498664535682e-05, "loss": 0.0176, "step": 28640 }, { "epoch": 8.335757928425952, "grad_norm": 0.13887743651866913, "learning_rate": 1.7347333528469225e-05, "loss": 0.018, "step": 28650 }, { "epoch": 8.338667442537096, "grad_norm": 0.10491052269935608, "learning_rate": 1.733916767934482e-05, "loss": 0.0182, "step": 28660 }, { "epoch": 8.34157695664824, "grad_norm": 0.09485246241092682, "learning_rate": 1.7331001119643034e-05, "loss": 0.0173, "step": 28670 }, { "epoch": 8.344486470759383, "grad_norm": 0.10756167769432068, "learning_rate": 1.7322833851844653e-05, "loss": 0.0184, "step": 28680 }, { "epoch": 8.347395984870527, "grad_norm": 0.15517820417881012, "learning_rate": 1.731466587843068e-05, "loss": 0.021, "step": 28690 }, { "epoch": 8.35030549898167, "grad_norm": 0.09946390241384506, "learning_rate": 1.7306497201882334e-05, "loss": 0.0166, "step": 28700 }, { "epoch": 8.353215013092813, "grad_norm": 0.11172281205654144, "learning_rate": 1.7298327824681032e-05, "loss": 0.0188, "step": 28710 }, { "epoch": 8.356124527203956, "grad_norm": 0.11355141550302505, "learning_rate": 1.729015774930842e-05, "loss": 0.0193, "step": 28720 }, { "epoch": 8.3590340413151, "grad_norm": 0.10448677092790604, "learning_rate": 1.7281986978246353e-05, "loss": 0.0173, "step": 28730 }, { "epoch": 8.361943555426244, "grad_norm": 0.09884058684110641, "learning_rate": 1.7273815513976894e-05, "loss": 0.0169, "step": 28740 }, { "epoch": 8.364853069537387, "grad_norm": 0.1113496869802475, "learning_rate": 1.7265643358982318e-05, "loss": 0.0171, "step": 28750 }, { "epoch": 8.367762583648531, "grad_norm": 0.09735320508480072, "learning_rate": 1.7257470515745117e-05, "loss": 0.0171, "step": 28760 }, { "epoch": 8.370672097759675, "grad_norm": 0.10322939604520798, "learning_rate": 1.724929698674797e-05, "loss": 0.0167, "step": 28770 }, { "epoch": 8.373581611870817, "grad_norm": 0.09596314281225204, "learning_rate": 1.7241122774473788e-05, "loss": 0.0175, "step": 28780 }, { "epoch": 8.37649112598196, "grad_norm": 0.10719845443964005, "learning_rate": 1.7232947881405684e-05, "loss": 0.0171, "step": 28790 }, { "epoch": 8.379400640093104, "grad_norm": 0.11237397789955139, "learning_rate": 1.722477231002697e-05, "loss": 0.0165, "step": 28800 }, { "epoch": 8.382310154204248, "grad_norm": 0.09617278724908829, "learning_rate": 1.721659606282117e-05, "loss": 0.0178, "step": 28810 }, { "epoch": 8.385219668315392, "grad_norm": 0.09570842981338501, "learning_rate": 1.7208419142272012e-05, "loss": 0.0165, "step": 28820 }, { "epoch": 8.388129182426535, "grad_norm": 0.12101088464260101, "learning_rate": 1.720024155086343e-05, "loss": 0.017, "step": 28830 }, { "epoch": 8.391038696537677, "grad_norm": 0.12604600191116333, "learning_rate": 1.7192063291079555e-05, "loss": 0.0166, "step": 28840 }, { "epoch": 8.393948210648821, "grad_norm": 0.10699638724327087, "learning_rate": 1.7183884365404738e-05, "loss": 0.018, "step": 28850 }, { "epoch": 8.396857724759965, "grad_norm": 0.11520282924175262, "learning_rate": 1.7175704776323504e-05, "loss": 0.0169, "step": 28860 }, { "epoch": 8.399767238871108, "grad_norm": 0.084729865193367, "learning_rate": 1.7167524526320612e-05, "loss": 0.0167, "step": 28870 }, { "epoch": 8.402676752982252, "grad_norm": 0.1119435578584671, "learning_rate": 1.7159343617881e-05, "loss": 0.0168, "step": 28880 }, { "epoch": 8.405586267093396, "grad_norm": 0.08900485187768936, "learning_rate": 1.71511620534898e-05, "loss": 0.0175, "step": 28890 }, { "epoch": 8.40849578120454, "grad_norm": 0.11573904752731323, "learning_rate": 1.7142979835632376e-05, "loss": 0.0172, "step": 28900 }, { "epoch": 8.411405295315681, "grad_norm": 0.10745099931955338, "learning_rate": 1.7134796966794254e-05, "loss": 0.0173, "step": 28910 }, { "epoch": 8.414314809426825, "grad_norm": 0.10541543364524841, "learning_rate": 1.7126613449461186e-05, "loss": 0.0155, "step": 28920 }, { "epoch": 8.417224323537969, "grad_norm": 0.08400427550077438, "learning_rate": 1.7118429286119098e-05, "loss": 0.0186, "step": 28930 }, { "epoch": 8.420133837649113, "grad_norm": 0.0855042040348053, "learning_rate": 1.7110244479254126e-05, "loss": 0.0163, "step": 28940 }, { "epoch": 8.423043351760256, "grad_norm": 0.10352940857410431, "learning_rate": 1.71020590313526e-05, "loss": 0.0178, "step": 28950 }, { "epoch": 8.4259528658714, "grad_norm": 0.09399323165416718, "learning_rate": 1.7093872944901044e-05, "loss": 0.0174, "step": 28960 }, { "epoch": 8.428862379982544, "grad_norm": 0.08805204182863235, "learning_rate": 1.7085686222386172e-05, "loss": 0.0174, "step": 28970 }, { "epoch": 8.431771894093686, "grad_norm": 0.09544476121664047, "learning_rate": 1.7077498866294895e-05, "loss": 0.016, "step": 28980 }, { "epoch": 8.43468140820483, "grad_norm": 0.1097244918346405, "learning_rate": 1.7069310879114313e-05, "loss": 0.0165, "step": 28990 }, { "epoch": 8.437590922315973, "grad_norm": 0.08736522495746613, "learning_rate": 1.7061122263331727e-05, "loss": 0.017, "step": 29000 }, { "epoch": 8.440500436427117, "grad_norm": 0.07053357362747192, "learning_rate": 1.7052933021434617e-05, "loss": 0.0166, "step": 29010 }, { "epoch": 8.44340995053826, "grad_norm": 0.09817615896463394, "learning_rate": 1.7044743155910658e-05, "loss": 0.016, "step": 29020 }, { "epoch": 8.446319464649404, "grad_norm": 0.11119817197322845, "learning_rate": 1.703655266924772e-05, "loss": 0.0174, "step": 29030 }, { "epoch": 8.449228978760548, "grad_norm": 0.12285135686397552, "learning_rate": 1.7028361563933846e-05, "loss": 0.0184, "step": 29040 }, { "epoch": 8.45213849287169, "grad_norm": 0.10609748959541321, "learning_rate": 1.702016984245729e-05, "loss": 0.0166, "step": 29050 }, { "epoch": 8.455048006982834, "grad_norm": 0.11744695156812668, "learning_rate": 1.701197750730647e-05, "loss": 0.0177, "step": 29060 }, { "epoch": 8.457957521093977, "grad_norm": 0.09522219747304916, "learning_rate": 1.7003784560970013e-05, "loss": 0.0177, "step": 29070 }, { "epoch": 8.460867035205121, "grad_norm": 0.09666924178600311, "learning_rate": 1.6995591005936714e-05, "loss": 0.0167, "step": 29080 }, { "epoch": 8.463776549316265, "grad_norm": 0.09344863146543503, "learning_rate": 1.6987396844695555e-05, "loss": 0.0178, "step": 29090 }, { "epoch": 8.466686063427408, "grad_norm": 0.10635789483785629, "learning_rate": 1.6979202079735708e-05, "loss": 0.0182, "step": 29100 }, { "epoch": 8.46959557753855, "grad_norm": 0.1037282943725586, "learning_rate": 1.697100671354653e-05, "loss": 0.0201, "step": 29110 }, { "epoch": 8.472505091649694, "grad_norm": 0.09013007581233978, "learning_rate": 1.6962810748617548e-05, "loss": 0.017, "step": 29120 }, { "epoch": 8.475414605760838, "grad_norm": 0.08699395507574081, "learning_rate": 1.695461418743849e-05, "loss": 0.017, "step": 29130 }, { "epoch": 8.478324119871981, "grad_norm": 0.09325745701789856, "learning_rate": 1.6946417032499252e-05, "loss": 0.0155, "step": 29140 }, { "epoch": 8.481233633983125, "grad_norm": 0.09630192816257477, "learning_rate": 1.6938219286289908e-05, "loss": 0.0161, "step": 29150 }, { "epoch": 8.484143148094269, "grad_norm": 0.10527335852384567, "learning_rate": 1.693002095130072e-05, "loss": 0.0188, "step": 29160 }, { "epoch": 8.487052662205413, "grad_norm": 0.08811891078948975, "learning_rate": 1.6921822030022132e-05, "loss": 0.0165, "step": 29170 }, { "epoch": 8.489962176316554, "grad_norm": 0.10814167559146881, "learning_rate": 1.691362252494475e-05, "loss": 0.0173, "step": 29180 }, { "epoch": 8.492871690427698, "grad_norm": 0.10360515117645264, "learning_rate": 1.690542243855937e-05, "loss": 0.0166, "step": 29190 }, { "epoch": 8.495781204538842, "grad_norm": 0.09477024525403976, "learning_rate": 1.6897221773356963e-05, "loss": 0.0162, "step": 29200 }, { "epoch": 8.498690718649986, "grad_norm": 0.09053520113229752, "learning_rate": 1.6889020531828676e-05, "loss": 0.0166, "step": 29210 }, { "epoch": 8.50160023276113, "grad_norm": 0.09922651201486588, "learning_rate": 1.688081871646583e-05, "loss": 0.0172, "step": 29220 }, { "epoch": 8.504509746872273, "grad_norm": 0.09253700822591782, "learning_rate": 1.6872616329759917e-05, "loss": 0.0166, "step": 29230 }, { "epoch": 8.507419260983415, "grad_norm": 0.15311360359191895, "learning_rate": 1.6864413374202605e-05, "loss": 0.0182, "step": 29240 }, { "epoch": 8.510328775094559, "grad_norm": 0.09038402885198593, "learning_rate": 1.685620985228574e-05, "loss": 0.0163, "step": 29250 }, { "epoch": 8.513238289205702, "grad_norm": 0.09080081433057785, "learning_rate": 1.6848005766501326e-05, "loss": 0.0174, "step": 29260 }, { "epoch": 8.516147803316846, "grad_norm": 0.10919585078954697, "learning_rate": 1.683980111934156e-05, "loss": 0.0172, "step": 29270 }, { "epoch": 8.51905731742799, "grad_norm": 0.10501942783594131, "learning_rate": 1.6831595913298787e-05, "loss": 0.0193, "step": 29280 }, { "epoch": 8.521966831539133, "grad_norm": 0.0816054418683052, "learning_rate": 1.682339015086554e-05, "loss": 0.018, "step": 29290 }, { "epoch": 8.524876345650277, "grad_norm": 0.10103871673345566, "learning_rate": 1.6815183834534507e-05, "loss": 0.0176, "step": 29300 }, { "epoch": 8.52778585976142, "grad_norm": 0.14923755824565887, "learning_rate": 1.680697696679855e-05, "loss": 0.0178, "step": 29310 }, { "epoch": 8.530695373872563, "grad_norm": 0.10571178048849106, "learning_rate": 1.6798769550150702e-05, "loss": 0.0174, "step": 29320 }, { "epoch": 8.533604887983707, "grad_norm": 0.08827657997608185, "learning_rate": 1.679056158708416e-05, "loss": 0.0155, "step": 29330 }, { "epoch": 8.53651440209485, "grad_norm": 0.12748943269252777, "learning_rate": 1.6782353080092283e-05, "loss": 0.0191, "step": 29340 }, { "epoch": 8.539423916205994, "grad_norm": 0.10538510233163834, "learning_rate": 1.6774144031668603e-05, "loss": 0.0182, "step": 29350 }, { "epoch": 8.542333430317138, "grad_norm": 0.10263239592313766, "learning_rate": 1.6765934444306804e-05, "loss": 0.0172, "step": 29360 }, { "epoch": 8.545242944428281, "grad_norm": 0.08704989403486252, "learning_rate": 1.6757724320500753e-05, "loss": 0.0153, "step": 29370 }, { "epoch": 8.548152458539423, "grad_norm": 0.12051400542259216, "learning_rate": 1.674951366274447e-05, "loss": 0.0187, "step": 29380 }, { "epoch": 8.551061972650567, "grad_norm": 0.10462487488985062, "learning_rate": 1.674130247353212e-05, "loss": 0.018, "step": 29390 }, { "epoch": 8.55397148676171, "grad_norm": 0.08737163990736008, "learning_rate": 1.6733090755358056e-05, "loss": 0.0174, "step": 29400 }, { "epoch": 8.556881000872854, "grad_norm": 0.1047951951622963, "learning_rate": 1.672487851071678e-05, "loss": 0.0201, "step": 29410 }, { "epoch": 8.559790514983998, "grad_norm": 0.09411831200122833, "learning_rate": 1.6716665742102956e-05, "loss": 0.0166, "step": 29420 }, { "epoch": 8.562700029095142, "grad_norm": 0.10019273310899734, "learning_rate": 1.670845245201141e-05, "loss": 0.0174, "step": 29430 }, { "epoch": 8.565609543206284, "grad_norm": 0.09283831715583801, "learning_rate": 1.670023864293711e-05, "loss": 0.0146, "step": 29440 }, { "epoch": 8.568519057317427, "grad_norm": 0.0970662459731102, "learning_rate": 1.669202431737521e-05, "loss": 0.0179, "step": 29450 }, { "epoch": 8.571428571428571, "grad_norm": 0.09185145050287247, "learning_rate": 1.6683809477820995e-05, "loss": 0.0169, "step": 29460 }, { "epoch": 8.574338085539715, "grad_norm": 0.09494022279977798, "learning_rate": 1.6675594126769916e-05, "loss": 0.0175, "step": 29470 }, { "epoch": 8.577247599650859, "grad_norm": 0.10700459778308868, "learning_rate": 1.6667378266717586e-05, "loss": 0.0186, "step": 29480 }, { "epoch": 8.580157113762002, "grad_norm": 0.12561559677124023, "learning_rate": 1.665916190015976e-05, "loss": 0.0156, "step": 29490 }, { "epoch": 8.583066627873146, "grad_norm": 0.1144491583108902, "learning_rate": 1.665094502959236e-05, "loss": 0.0166, "step": 29500 }, { "epoch": 8.585976141984288, "grad_norm": 0.10869542509317398, "learning_rate": 1.664272765751145e-05, "loss": 0.0162, "step": 29510 }, { "epoch": 8.588885656095432, "grad_norm": 0.11053754389286041, "learning_rate": 1.6634509786413248e-05, "loss": 0.0161, "step": 29520 }, { "epoch": 8.591795170206575, "grad_norm": 0.10235197842121124, "learning_rate": 1.6626291418794135e-05, "loss": 0.0181, "step": 29530 }, { "epoch": 8.594704684317719, "grad_norm": 0.1229064017534256, "learning_rate": 1.6618072557150625e-05, "loss": 0.0171, "step": 29540 }, { "epoch": 8.597614198428863, "grad_norm": 0.10184184461832047, "learning_rate": 1.6609853203979398e-05, "loss": 0.0185, "step": 29550 }, { "epoch": 8.600523712540006, "grad_norm": 0.1006816029548645, "learning_rate": 1.6601633361777273e-05, "loss": 0.016, "step": 29560 }, { "epoch": 8.603433226651148, "grad_norm": 0.12387163937091827, "learning_rate": 1.6593413033041215e-05, "loss": 0.0158, "step": 29570 }, { "epoch": 8.606342740762292, "grad_norm": 0.09408186376094818, "learning_rate": 1.6585192220268356e-05, "loss": 0.0167, "step": 29580 }, { "epoch": 8.609252254873436, "grad_norm": 0.08306333422660828, "learning_rate": 1.6576970925955956e-05, "loss": 0.0178, "step": 29590 }, { "epoch": 8.61216176898458, "grad_norm": 0.11640579998493195, "learning_rate": 1.6568749152601422e-05, "loss": 0.0173, "step": 29600 }, { "epoch": 8.615071283095723, "grad_norm": 0.10855811834335327, "learning_rate": 1.6560526902702317e-05, "loss": 0.0191, "step": 29610 }, { "epoch": 8.617980797206867, "grad_norm": 0.09687766432762146, "learning_rate": 1.6552304178756337e-05, "loss": 0.0167, "step": 29620 }, { "epoch": 8.62089031131801, "grad_norm": 0.09434355050325394, "learning_rate": 1.654408098326134e-05, "loss": 0.0156, "step": 29630 }, { "epoch": 8.623799825429153, "grad_norm": 0.10106755048036575, "learning_rate": 1.6535857318715307e-05, "loss": 0.0155, "step": 29640 }, { "epoch": 8.626709339540296, "grad_norm": 0.08595811575651169, "learning_rate": 1.6527633187616376e-05, "loss": 0.0179, "step": 29650 }, { "epoch": 8.62961885365144, "grad_norm": 0.1160951629281044, "learning_rate": 1.6519408592462818e-05, "loss": 0.0155, "step": 29660 }, { "epoch": 8.632528367762584, "grad_norm": 0.07656998932361603, "learning_rate": 1.6511183535753044e-05, "loss": 0.0165, "step": 29670 }, { "epoch": 8.635437881873727, "grad_norm": 0.07752007246017456, "learning_rate": 1.650295801998561e-05, "loss": 0.0173, "step": 29680 }, { "epoch": 8.638347395984871, "grad_norm": 0.09991763532161713, "learning_rate": 1.6494732047659217e-05, "loss": 0.0174, "step": 29690 }, { "epoch": 8.641256910096015, "grad_norm": 0.10601641237735748, "learning_rate": 1.6486505621272693e-05, "loss": 0.0153, "step": 29700 }, { "epoch": 8.644166424207157, "grad_norm": 0.1432541012763977, "learning_rate": 1.647827874332501e-05, "loss": 0.0194, "step": 29710 }, { "epoch": 8.6470759383183, "grad_norm": 0.12575966119766235, "learning_rate": 1.6470051416315277e-05, "loss": 0.0167, "step": 29720 }, { "epoch": 8.649985452429444, "grad_norm": 0.08693546056747437, "learning_rate": 1.6461823642742734e-05, "loss": 0.0165, "step": 29730 }, { "epoch": 8.652894966540588, "grad_norm": 0.10007554292678833, "learning_rate": 1.645359542510677e-05, "loss": 0.0176, "step": 29740 }, { "epoch": 8.655804480651732, "grad_norm": 0.138940691947937, "learning_rate": 1.6445366765906897e-05, "loss": 0.0174, "step": 29750 }, { "epoch": 8.658713994762875, "grad_norm": 0.10737534612417221, "learning_rate": 1.643713766764276e-05, "loss": 0.0165, "step": 29760 }, { "epoch": 8.661623508874017, "grad_norm": 0.08922506868839264, "learning_rate": 1.6428908132814146e-05, "loss": 0.018, "step": 29770 }, { "epoch": 8.664533022985161, "grad_norm": 0.1204441711306572, "learning_rate": 1.642067816392097e-05, "loss": 0.018, "step": 29780 }, { "epoch": 8.667442537096305, "grad_norm": 0.1444484144449234, "learning_rate": 1.6412447763463275e-05, "loss": 0.0184, "step": 29790 }, { "epoch": 8.670352051207448, "grad_norm": 0.09622640162706375, "learning_rate": 1.6404216933941247e-05, "loss": 0.0174, "step": 29800 }, { "epoch": 8.673261565318592, "grad_norm": 0.08645924180746078, "learning_rate": 1.639598567785519e-05, "loss": 0.0167, "step": 29810 }, { "epoch": 8.676171079429736, "grad_norm": 0.08727701753377914, "learning_rate": 1.6387753997705547e-05, "loss": 0.0185, "step": 29820 }, { "epoch": 8.67908059354088, "grad_norm": 0.09176316857337952, "learning_rate": 1.637952189599287e-05, "loss": 0.0167, "step": 29830 }, { "epoch": 8.681990107652021, "grad_norm": 0.07497666031122208, "learning_rate": 1.6371289375217873e-05, "loss": 0.0166, "step": 29840 }, { "epoch": 8.684899621763165, "grad_norm": 0.10308926552534103, "learning_rate": 1.6363056437881372e-05, "loss": 0.0158, "step": 29850 }, { "epoch": 8.687809135874309, "grad_norm": 0.09883914887905121, "learning_rate": 1.6354823086484315e-05, "loss": 0.0164, "step": 29860 }, { "epoch": 8.690718649985453, "grad_norm": 0.08740632236003876, "learning_rate": 1.634658932352777e-05, "loss": 0.0162, "step": 29870 }, { "epoch": 8.693628164096596, "grad_norm": 0.0668332502245903, "learning_rate": 1.633835515151295e-05, "loss": 0.0161, "step": 29880 }, { "epoch": 8.69653767820774, "grad_norm": 0.10671298205852509, "learning_rate": 1.6330120572941166e-05, "loss": 0.0179, "step": 29890 }, { "epoch": 8.699447192318882, "grad_norm": 0.08093543350696564, "learning_rate": 1.6321885590313876e-05, "loss": 0.0171, "step": 29900 }, { "epoch": 8.702356706430026, "grad_norm": 0.08570899814367294, "learning_rate": 1.6313650206132643e-05, "loss": 0.016, "step": 29910 }, { "epoch": 8.70526622054117, "grad_norm": 0.09928283840417862, "learning_rate": 1.6305414422899163e-05, "loss": 0.0177, "step": 29920 }, { "epoch": 8.708175734652313, "grad_norm": 0.08390014618635178, "learning_rate": 1.629717824311525e-05, "loss": 0.0158, "step": 29930 }, { "epoch": 8.711085248763457, "grad_norm": 0.09207816421985626, "learning_rate": 1.6288941669282834e-05, "loss": 0.0175, "step": 29940 }, { "epoch": 8.7139947628746, "grad_norm": 0.11274168640375137, "learning_rate": 1.628070470390398e-05, "loss": 0.0158, "step": 29950 }, { "epoch": 8.716904276985744, "grad_norm": 0.11588722467422485, "learning_rate": 1.627246734948084e-05, "loss": 0.0167, "step": 29960 }, { "epoch": 8.719813791096886, "grad_norm": 0.09726651012897491, "learning_rate": 1.6264229608515725e-05, "loss": 0.0173, "step": 29970 }, { "epoch": 8.72272330520803, "grad_norm": 0.10415154695510864, "learning_rate": 1.6255991483511035e-05, "loss": 0.0172, "step": 29980 }, { "epoch": 8.725632819319173, "grad_norm": 0.09471286833286285, "learning_rate": 1.6247752976969288e-05, "loss": 0.017, "step": 29990 }, { "epoch": 8.728542333430317, "grad_norm": 0.0886528417468071, "learning_rate": 1.6239514091393138e-05, "loss": 0.0166, "step": 30000 }, { "epoch": 8.73145184754146, "grad_norm": 0.09792597591876984, "learning_rate": 1.6231274829285334e-05, "loss": 0.0174, "step": 30010 }, { "epoch": 8.734361361652605, "grad_norm": 0.10862288624048233, "learning_rate": 1.6223035193148746e-05, "loss": 0.0176, "step": 30020 }, { "epoch": 8.737270875763748, "grad_norm": 0.08853109180927277, "learning_rate": 1.6214795185486357e-05, "loss": 0.0165, "step": 30030 }, { "epoch": 8.74018038987489, "grad_norm": 0.0963480994105339, "learning_rate": 1.6206554808801264e-05, "loss": 0.0177, "step": 30040 }, { "epoch": 8.743089903986034, "grad_norm": 0.12176165729761124, "learning_rate": 1.619831406559668e-05, "loss": 0.0158, "step": 30050 }, { "epoch": 8.745999418097178, "grad_norm": 0.10843916982412338, "learning_rate": 1.6190072958375926e-05, "loss": 0.0178, "step": 30060 }, { "epoch": 8.748908932208321, "grad_norm": 0.09111883491277695, "learning_rate": 1.618183148964242e-05, "loss": 0.0154, "step": 30070 }, { "epoch": 8.751818446319465, "grad_norm": 0.0965135395526886, "learning_rate": 1.617358966189972e-05, "loss": 0.0175, "step": 30080 }, { "epoch": 8.754727960430609, "grad_norm": 0.10057961195707321, "learning_rate": 1.616534747765147e-05, "loss": 0.0158, "step": 30090 }, { "epoch": 8.757637474541752, "grad_norm": 0.10434307903051376, "learning_rate": 1.6157104939401415e-05, "loss": 0.0157, "step": 30100 }, { "epoch": 8.760546988652894, "grad_norm": 0.09930958598852158, "learning_rate": 1.6148862049653437e-05, "loss": 0.0169, "step": 30110 }, { "epoch": 8.763456502764038, "grad_norm": 0.10317490249872208, "learning_rate": 1.6140618810911504e-05, "loss": 0.0175, "step": 30120 }, { "epoch": 8.766366016875182, "grad_norm": 0.09836270660161972, "learning_rate": 1.613237522567969e-05, "loss": 0.018, "step": 30130 }, { "epoch": 8.769275530986326, "grad_norm": 0.120427206158638, "learning_rate": 1.6124131296462184e-05, "loss": 0.019, "step": 30140 }, { "epoch": 8.77218504509747, "grad_norm": 0.08872213959693909, "learning_rate": 1.6115887025763266e-05, "loss": 0.0156, "step": 30150 }, { "epoch": 8.775094559208613, "grad_norm": 0.07524844259023666, "learning_rate": 1.6107642416087334e-05, "loss": 0.0163, "step": 30160 }, { "epoch": 8.778004073319755, "grad_norm": 0.08319266140460968, "learning_rate": 1.6099397469938886e-05, "loss": 0.0171, "step": 30170 }, { "epoch": 8.780913587430899, "grad_norm": 0.09882451593875885, "learning_rate": 1.609115218982251e-05, "loss": 0.0167, "step": 30180 }, { "epoch": 8.783823101542042, "grad_norm": 0.0927443727850914, "learning_rate": 1.6082906578242913e-05, "loss": 0.0181, "step": 30190 }, { "epoch": 8.786732615653186, "grad_norm": 0.1093512624502182, "learning_rate": 1.6074660637704883e-05, "loss": 0.0172, "step": 30200 }, { "epoch": 8.78964212976433, "grad_norm": 0.10663340240716934, "learning_rate": 1.606641437071333e-05, "loss": 0.016, "step": 30210 }, { "epoch": 8.792551643875473, "grad_norm": 0.09308840334415436, "learning_rate": 1.6058167779773247e-05, "loss": 0.0174, "step": 30220 }, { "epoch": 8.795461157986615, "grad_norm": 0.08348508179187775, "learning_rate": 1.6049920867389736e-05, "loss": 0.0164, "step": 30230 }, { "epoch": 8.798370672097759, "grad_norm": 0.10282351076602936, "learning_rate": 1.6041673636067986e-05, "loss": 0.0177, "step": 30240 }, { "epoch": 8.801280186208903, "grad_norm": 0.0828281044960022, "learning_rate": 1.6033426088313284e-05, "loss": 0.0164, "step": 30250 }, { "epoch": 8.804189700320046, "grad_norm": 0.07915310561656952, "learning_rate": 1.6025178226631028e-05, "loss": 0.0179, "step": 30260 }, { "epoch": 8.80709921443119, "grad_norm": 0.08894212543964386, "learning_rate": 1.6016930053526696e-05, "loss": 0.016, "step": 30270 }, { "epoch": 8.810008728542334, "grad_norm": 0.07823186367750168, "learning_rate": 1.600868157150586e-05, "loss": 0.0164, "step": 30280 }, { "epoch": 8.812918242653478, "grad_norm": 0.10721119493246078, "learning_rate": 1.60004327830742e-05, "loss": 0.0188, "step": 30290 }, { "epoch": 8.81582775676462, "grad_norm": 0.09032375365495682, "learning_rate": 1.599218369073747e-05, "loss": 0.0171, "step": 30300 }, { "epoch": 8.818737270875763, "grad_norm": 0.10912773013114929, "learning_rate": 1.5983934297001533e-05, "loss": 0.0155, "step": 30310 }, { "epoch": 8.821646784986907, "grad_norm": 0.11118059605360031, "learning_rate": 1.597568460437234e-05, "loss": 0.0177, "step": 30320 }, { "epoch": 8.82455629909805, "grad_norm": 0.11479590833187103, "learning_rate": 1.5967434615355923e-05, "loss": 0.0186, "step": 30330 }, { "epoch": 8.827465813209194, "grad_norm": 0.12562790513038635, "learning_rate": 1.5959184332458413e-05, "loss": 0.0156, "step": 30340 }, { "epoch": 8.830375327320338, "grad_norm": 0.09045041352510452, "learning_rate": 1.5950933758186025e-05, "loss": 0.0152, "step": 30350 }, { "epoch": 8.833284841431482, "grad_norm": 0.12910300493240356, "learning_rate": 1.594268289504507e-05, "loss": 0.0159, "step": 30360 }, { "epoch": 8.836194355542624, "grad_norm": 0.08833429962396622, "learning_rate": 1.5934431745541947e-05, "loss": 0.0162, "step": 30370 }, { "epoch": 8.839103869653767, "grad_norm": 0.08514991402626038, "learning_rate": 1.592618031218313e-05, "loss": 0.0184, "step": 30380 }, { "epoch": 8.842013383764911, "grad_norm": 0.09880506992340088, "learning_rate": 1.5917928597475186e-05, "loss": 0.0159, "step": 30390 }, { "epoch": 8.844922897876055, "grad_norm": 0.12425553053617477, "learning_rate": 1.5909676603924773e-05, "loss": 0.0181, "step": 30400 }, { "epoch": 8.847832411987198, "grad_norm": 0.08630908280611038, "learning_rate": 1.5901424334038626e-05, "loss": 0.0168, "step": 30410 }, { "epoch": 8.850741926098342, "grad_norm": 0.1158103495836258, "learning_rate": 1.5893171790323566e-05, "loss": 0.0162, "step": 30420 }, { "epoch": 8.853651440209486, "grad_norm": 0.1324843466281891, "learning_rate": 1.5884918975286502e-05, "loss": 0.0168, "step": 30430 }, { "epoch": 8.856560954320628, "grad_norm": 0.08504466712474823, "learning_rate": 1.5876665891434415e-05, "loss": 0.0162, "step": 30440 }, { "epoch": 8.859470468431772, "grad_norm": 0.0847587063908577, "learning_rate": 1.586841254127438e-05, "loss": 0.0173, "step": 30450 }, { "epoch": 8.862379982542915, "grad_norm": 0.11054817587137222, "learning_rate": 1.5860158927313537e-05, "loss": 0.0171, "step": 30460 }, { "epoch": 8.865289496654059, "grad_norm": 0.08623109757900238, "learning_rate": 1.585190505205913e-05, "loss": 0.0158, "step": 30470 }, { "epoch": 8.868199010765203, "grad_norm": 0.09587983042001724, "learning_rate": 1.5843650918018464e-05, "loss": 0.0186, "step": 30480 }, { "epoch": 8.871108524876346, "grad_norm": 0.09202545136213303, "learning_rate": 1.5835396527698917e-05, "loss": 0.0174, "step": 30490 }, { "epoch": 8.874018038987488, "grad_norm": 0.10838089138269424, "learning_rate": 1.5827141883607964e-05, "loss": 0.0152, "step": 30500 }, { "epoch": 8.876927553098632, "grad_norm": 0.08772590756416321, "learning_rate": 1.581888698825315e-05, "loss": 0.0166, "step": 30510 }, { "epoch": 8.879837067209776, "grad_norm": 0.0992419570684433, "learning_rate": 1.5810631844142084e-05, "loss": 0.0183, "step": 30520 }, { "epoch": 8.88274658132092, "grad_norm": 0.11444392800331116, "learning_rate": 1.5802376453782473e-05, "loss": 0.0188, "step": 30530 }, { "epoch": 8.885656095432063, "grad_norm": 0.08558442443609238, "learning_rate": 1.5794120819682077e-05, "loss": 0.0148, "step": 30540 }, { "epoch": 8.888565609543207, "grad_norm": 0.07748161256313324, "learning_rate": 1.578586494434875e-05, "loss": 0.0167, "step": 30550 }, { "epoch": 8.89147512365435, "grad_norm": 0.08078838884830475, "learning_rate": 1.5777608830290397e-05, "loss": 0.0152, "step": 30560 }, { "epoch": 8.894384637765492, "grad_norm": 0.08774082362651825, "learning_rate": 1.5769352480015015e-05, "loss": 0.0157, "step": 30570 }, { "epoch": 8.897294151876636, "grad_norm": 0.11314518749713898, "learning_rate": 1.5761095896030663e-05, "loss": 0.0152, "step": 30580 }, { "epoch": 8.90020366598778, "grad_norm": 0.11280682682991028, "learning_rate": 1.5752839080845472e-05, "loss": 0.0169, "step": 30590 }, { "epoch": 8.903113180098924, "grad_norm": 0.0885385274887085, "learning_rate": 1.5744582036967648e-05, "loss": 0.0184, "step": 30600 }, { "epoch": 8.906022694210067, "grad_norm": 0.09441486746072769, "learning_rate": 1.573632476690546e-05, "loss": 0.0175, "step": 30610 }, { "epoch": 8.908932208321211, "grad_norm": 0.09146057814359665, "learning_rate": 1.5728067273167244e-05, "loss": 0.0167, "step": 30620 }, { "epoch": 8.911841722432353, "grad_norm": 0.09979455918073654, "learning_rate": 1.5719809558261417e-05, "loss": 0.0163, "step": 30630 }, { "epoch": 8.914751236543497, "grad_norm": 0.09296651929616928, "learning_rate": 1.5711551624696455e-05, "loss": 0.0186, "step": 30640 }, { "epoch": 8.91766075065464, "grad_norm": 0.11508355289697647, "learning_rate": 1.5703293474980895e-05, "loss": 0.0171, "step": 30650 }, { "epoch": 8.920570264765784, "grad_norm": 0.11698789894580841, "learning_rate": 1.5695035111623347e-05, "loss": 0.0167, "step": 30660 }, { "epoch": 8.923479778876928, "grad_norm": 0.1144503578543663, "learning_rate": 1.5686776537132484e-05, "loss": 0.0166, "step": 30670 }, { "epoch": 8.926389292988071, "grad_norm": 0.1035073772072792, "learning_rate": 1.567851775401704e-05, "loss": 0.0177, "step": 30680 }, { "epoch": 8.929298807099215, "grad_norm": 0.09733280539512634, "learning_rate": 1.5670258764785817e-05, "loss": 0.0166, "step": 30690 }, { "epoch": 8.932208321210357, "grad_norm": 0.10633537173271179, "learning_rate": 1.5661999571947682e-05, "loss": 0.0157, "step": 30700 }, { "epoch": 8.9351178353215, "grad_norm": 0.11255999654531479, "learning_rate": 1.5653740178011555e-05, "loss": 0.018, "step": 30710 }, { "epoch": 8.938027349432645, "grad_norm": 0.09471182525157928, "learning_rate": 1.5645480585486422e-05, "loss": 0.016, "step": 30720 }, { "epoch": 8.940936863543788, "grad_norm": 0.08501791208982468, "learning_rate": 1.5637220796881327e-05, "loss": 0.0172, "step": 30730 }, { "epoch": 8.943846377654932, "grad_norm": 0.09249471127986908, "learning_rate": 1.5628960814705385e-05, "loss": 0.0172, "step": 30740 }, { "epoch": 8.946755891766076, "grad_norm": 0.12050576508045197, "learning_rate": 1.5620700641467752e-05, "loss": 0.0177, "step": 30750 }, { "epoch": 8.94966540587722, "grad_norm": 0.10131388902664185, "learning_rate": 1.561244027967765e-05, "loss": 0.0157, "step": 30760 }, { "epoch": 8.952574919988361, "grad_norm": 0.09519189596176147, "learning_rate": 1.5604179731844366e-05, "loss": 0.0161, "step": 30770 }, { "epoch": 8.955484434099505, "grad_norm": 0.10121957957744598, "learning_rate": 1.5595919000477228e-05, "loss": 0.0166, "step": 30780 }, { "epoch": 8.958393948210649, "grad_norm": 0.1020519956946373, "learning_rate": 1.558765808808563e-05, "loss": 0.0151, "step": 30790 }, { "epoch": 8.961303462321792, "grad_norm": 0.08015932887792587, "learning_rate": 1.557939699717903e-05, "loss": 0.0167, "step": 30800 }, { "epoch": 8.964212976432936, "grad_norm": 0.1316869705915451, "learning_rate": 1.5571135730266914e-05, "loss": 0.0141, "step": 30810 }, { "epoch": 8.96712249054408, "grad_norm": 0.07814919203519821, "learning_rate": 1.5562874289858844e-05, "loss": 0.015, "step": 30820 }, { "epoch": 8.970032004655222, "grad_norm": 0.09315983206033707, "learning_rate": 1.5554612678464424e-05, "loss": 0.0163, "step": 30830 }, { "epoch": 8.972941518766365, "grad_norm": 0.09830046445131302, "learning_rate": 1.5546350898593317e-05, "loss": 0.0162, "step": 30840 }, { "epoch": 8.97585103287751, "grad_norm": 0.11497214436531067, "learning_rate": 1.553808895275524e-05, "loss": 0.0169, "step": 30850 }, { "epoch": 8.978760546988653, "grad_norm": 0.08234383910894394, "learning_rate": 1.5529826843459938e-05, "loss": 0.0156, "step": 30860 }, { "epoch": 8.981670061099797, "grad_norm": 0.11260825395584106, "learning_rate": 1.5521564573217237e-05, "loss": 0.0174, "step": 30870 }, { "epoch": 8.98457957521094, "grad_norm": 0.11173421144485474, "learning_rate": 1.5513302144536985e-05, "loss": 0.0172, "step": 30880 }, { "epoch": 8.987489089322084, "grad_norm": 0.09455826878547668, "learning_rate": 1.5505039559929095e-05, "loss": 0.0164, "step": 30890 }, { "epoch": 8.990398603433226, "grad_norm": 0.10820209980010986, "learning_rate": 1.5496776821903517e-05, "loss": 0.0158, "step": 30900 }, { "epoch": 8.99330811754437, "grad_norm": 0.1100122258067131, "learning_rate": 1.548851393297027e-05, "loss": 0.0179, "step": 30910 }, { "epoch": 8.996217631655513, "grad_norm": 0.13570743799209595, "learning_rate": 1.5480250895639375e-05, "loss": 0.0194, "step": 30920 }, { "epoch": 8.999127145766657, "grad_norm": 0.10978685319423676, "learning_rate": 1.5471987712420945e-05, "loss": 0.0159, "step": 30930 }, { "epoch": 9.0020366598778, "grad_norm": 0.08922844380140305, "learning_rate": 1.5463724385825108e-05, "loss": 0.0165, "step": 30940 }, { "epoch": 9.004946173988944, "grad_norm": 0.08478621393442154, "learning_rate": 1.5455460918362044e-05, "loss": 0.0165, "step": 30950 }, { "epoch": 9.007855688100086, "grad_norm": 0.09722667187452316, "learning_rate": 1.5447197312541983e-05, "loss": 0.0162, "step": 30960 }, { "epoch": 9.01076520221123, "grad_norm": 0.09194847196340561, "learning_rate": 1.5438933570875188e-05, "loss": 0.0149, "step": 30970 }, { "epoch": 9.013674716322374, "grad_norm": 0.12740276753902435, "learning_rate": 1.5430669695871962e-05, "loss": 0.0167, "step": 30980 }, { "epoch": 9.016584230433518, "grad_norm": 0.08779609948396683, "learning_rate": 1.5422405690042648e-05, "loss": 0.0161, "step": 30990 }, { "epoch": 9.019493744544661, "grad_norm": 0.09765806049108505, "learning_rate": 1.541414155589764e-05, "loss": 0.0162, "step": 31000 }, { "epoch": 9.022403258655805, "grad_norm": 0.08300325274467468, "learning_rate": 1.540587729594737e-05, "loss": 0.0161, "step": 31010 }, { "epoch": 9.025312772766949, "grad_norm": 0.0895213782787323, "learning_rate": 1.539761291270229e-05, "loss": 0.0161, "step": 31020 }, { "epoch": 9.02822228687809, "grad_norm": 0.09828859567642212, "learning_rate": 1.5389348408672902e-05, "loss": 0.0166, "step": 31030 }, { "epoch": 9.031131800989234, "grad_norm": 0.11076074093580246, "learning_rate": 1.5381083786369746e-05, "loss": 0.0152, "step": 31040 }, { "epoch": 9.034041315100378, "grad_norm": 0.09090074151754379, "learning_rate": 1.5372819048303407e-05, "loss": 0.0167, "step": 31050 }, { "epoch": 9.036950829211522, "grad_norm": 0.06592404097318649, "learning_rate": 1.5364554196984477e-05, "loss": 0.0156, "step": 31060 }, { "epoch": 9.039860343322665, "grad_norm": 0.09133961796760559, "learning_rate": 1.535628923492361e-05, "loss": 0.0151, "step": 31070 }, { "epoch": 9.04276985743381, "grad_norm": 0.10742321610450745, "learning_rate": 1.5348024164631477e-05, "loss": 0.015, "step": 31080 }, { "epoch": 9.045679371544953, "grad_norm": 0.1014154925942421, "learning_rate": 1.5339758988618797e-05, "loss": 0.0156, "step": 31090 }, { "epoch": 9.048588885656095, "grad_norm": 0.12236484885215759, "learning_rate": 1.5331493709396302e-05, "loss": 0.0172, "step": 31100 }, { "epoch": 9.051498399767238, "grad_norm": 0.09822987020015717, "learning_rate": 1.5323228329474775e-05, "loss": 0.0175, "step": 31110 }, { "epoch": 9.054407913878382, "grad_norm": 0.09457488358020782, "learning_rate": 1.5314962851365017e-05, "loss": 0.016, "step": 31120 }, { "epoch": 9.057317427989526, "grad_norm": 0.08477060496807098, "learning_rate": 1.530669727757786e-05, "loss": 0.0164, "step": 31130 }, { "epoch": 9.06022694210067, "grad_norm": 0.08491060882806778, "learning_rate": 1.529843161062417e-05, "loss": 0.0169, "step": 31140 }, { "epoch": 9.063136456211813, "grad_norm": 0.08499537408351898, "learning_rate": 1.529016585301483e-05, "loss": 0.0173, "step": 31150 }, { "epoch": 9.066045970322955, "grad_norm": 0.09279286861419678, "learning_rate": 1.5281900007260767e-05, "loss": 0.0162, "step": 31160 }, { "epoch": 9.068955484434099, "grad_norm": 0.1137106865644455, "learning_rate": 1.5273634075872927e-05, "loss": 0.0159, "step": 31170 }, { "epoch": 9.071864998545243, "grad_norm": 0.10329840332269669, "learning_rate": 1.526536806136228e-05, "loss": 0.0172, "step": 31180 }, { "epoch": 9.074774512656386, "grad_norm": 0.10017069429159164, "learning_rate": 1.525710196623982e-05, "loss": 0.0168, "step": 31190 }, { "epoch": 9.07768402676753, "grad_norm": 0.08410672843456268, "learning_rate": 1.5248835793016572e-05, "loss": 0.018, "step": 31200 }, { "epoch": 9.080593540878674, "grad_norm": 0.09177308529615402, "learning_rate": 1.524056954420358e-05, "loss": 0.0168, "step": 31210 }, { "epoch": 9.083503054989817, "grad_norm": 0.1099018082022667, "learning_rate": 1.5232303222311911e-05, "loss": 0.0175, "step": 31220 }, { "epoch": 9.08641256910096, "grad_norm": 0.09297297894954681, "learning_rate": 1.522403682985266e-05, "loss": 0.0186, "step": 31230 }, { "epoch": 9.089322083212103, "grad_norm": 0.14073243737220764, "learning_rate": 1.5215770369336935e-05, "loss": 0.0156, "step": 31240 }, { "epoch": 9.092231597323247, "grad_norm": 0.11431795358657837, "learning_rate": 1.5207503843275864e-05, "loss": 0.0164, "step": 31250 }, { "epoch": 9.09514111143439, "grad_norm": 0.1336662918329239, "learning_rate": 1.519923725418061e-05, "loss": 0.0174, "step": 31260 }, { "epoch": 9.098050625545534, "grad_norm": 0.09683118015527725, "learning_rate": 1.5190970604562338e-05, "loss": 0.0147, "step": 31270 }, { "epoch": 9.100960139656678, "grad_norm": 0.10886288434267044, "learning_rate": 1.5182703896932239e-05, "loss": 0.0173, "step": 31280 }, { "epoch": 9.103869653767822, "grad_norm": 0.11830322444438934, "learning_rate": 1.5174437133801518e-05, "loss": 0.0167, "step": 31290 }, { "epoch": 9.106779167878964, "grad_norm": 0.08365700393915176, "learning_rate": 1.5166170317681405e-05, "loss": 0.0164, "step": 31300 }, { "epoch": 9.109688681990107, "grad_norm": 0.10499875247478485, "learning_rate": 1.5157903451083131e-05, "loss": 0.0156, "step": 31310 }, { "epoch": 9.112598196101251, "grad_norm": 0.08995556831359863, "learning_rate": 1.5149636536517962e-05, "loss": 0.018, "step": 31320 }, { "epoch": 9.115507710212395, "grad_norm": 0.12407133728265762, "learning_rate": 1.5141369576497164e-05, "loss": 0.0166, "step": 31330 }, { "epoch": 9.118417224323538, "grad_norm": 0.10636167973279953, "learning_rate": 1.513310257353202e-05, "loss": 0.0164, "step": 31340 }, { "epoch": 9.121326738434682, "grad_norm": 0.10387950390577316, "learning_rate": 1.5124835530133827e-05, "loss": 0.0154, "step": 31350 }, { "epoch": 9.124236252545824, "grad_norm": 0.11897458136081696, "learning_rate": 1.5116568448813898e-05, "loss": 0.0181, "step": 31360 }, { "epoch": 9.127145766656968, "grad_norm": 0.10472101718187332, "learning_rate": 1.5108301332083551e-05, "loss": 0.0177, "step": 31370 }, { "epoch": 9.130055280768111, "grad_norm": 0.11261878907680511, "learning_rate": 1.510003418245412e-05, "loss": 0.0189, "step": 31380 }, { "epoch": 9.132964794879255, "grad_norm": 0.10814837366342545, "learning_rate": 1.5091767002436944e-05, "loss": 0.0177, "step": 31390 }, { "epoch": 9.135874308990399, "grad_norm": 0.10157189518213272, "learning_rate": 1.5083499794543379e-05, "loss": 0.0171, "step": 31400 }, { "epoch": 9.138783823101543, "grad_norm": 0.09802704304456711, "learning_rate": 1.5075232561284775e-05, "loss": 0.0168, "step": 31410 }, { "epoch": 9.141693337212686, "grad_norm": 0.10090295225381851, "learning_rate": 1.5066965305172512e-05, "loss": 0.0152, "step": 31420 }, { "epoch": 9.144602851323828, "grad_norm": 0.09862040728330612, "learning_rate": 1.5058698028717954e-05, "loss": 0.0165, "step": 31430 }, { "epoch": 9.147512365434972, "grad_norm": 0.08251466602087021, "learning_rate": 1.5050430734432486e-05, "loss": 0.0154, "step": 31440 }, { "epoch": 9.150421879546116, "grad_norm": 0.09701664000749588, "learning_rate": 1.5042163424827493e-05, "loss": 0.0163, "step": 31450 }, { "epoch": 9.15333139365726, "grad_norm": 0.08388002961874008, "learning_rate": 1.5033896102414362e-05, "loss": 0.0153, "step": 31460 }, { "epoch": 9.156240907768403, "grad_norm": 0.0812537744641304, "learning_rate": 1.5025628769704493e-05, "loss": 0.0178, "step": 31470 }, { "epoch": 9.159150421879547, "grad_norm": 0.11311346292495728, "learning_rate": 1.501736142920928e-05, "loss": 0.0168, "step": 31480 }, { "epoch": 9.162059935990689, "grad_norm": 0.10741401463747025, "learning_rate": 1.5009094083440125e-05, "loss": 0.0168, "step": 31490 }, { "epoch": 9.164969450101832, "grad_norm": 0.11226187646389008, "learning_rate": 1.5000826734908423e-05, "loss": 0.017, "step": 31500 }, { "epoch": 9.167878964212976, "grad_norm": 0.09450846910476685, "learning_rate": 1.4992559386125582e-05, "loss": 0.0188, "step": 31510 }, { "epoch": 9.17078847832412, "grad_norm": 0.11123664677143097, "learning_rate": 1.4984292039603005e-05, "loss": 0.0186, "step": 31520 }, { "epoch": 9.173697992435264, "grad_norm": 0.10033498704433441, "learning_rate": 1.4976024697852088e-05, "loss": 0.0161, "step": 31530 }, { "epoch": 9.176607506546407, "grad_norm": 0.12739883363246918, "learning_rate": 1.4967757363384234e-05, "loss": 0.0155, "step": 31540 }, { "epoch": 9.179517020657551, "grad_norm": 0.10952523350715637, "learning_rate": 1.495949003871085e-05, "loss": 0.0162, "step": 31550 }, { "epoch": 9.182426534768693, "grad_norm": 0.09909851104021072, "learning_rate": 1.4951222726343314e-05, "loss": 0.0175, "step": 31560 }, { "epoch": 9.185336048879837, "grad_norm": 0.09468593448400497, "learning_rate": 1.4942955428793034e-05, "loss": 0.0165, "step": 31570 }, { "epoch": 9.18824556299098, "grad_norm": 0.09938393533229828, "learning_rate": 1.4934688148571381e-05, "loss": 0.0176, "step": 31580 }, { "epoch": 9.191155077102124, "grad_norm": 0.1450936198234558, "learning_rate": 1.4926420888189753e-05, "loss": 0.0163, "step": 31590 }, { "epoch": 9.194064591213268, "grad_norm": 0.10892076045274734, "learning_rate": 1.4918153650159507e-05, "loss": 0.0182, "step": 31600 }, { "epoch": 9.196974105324411, "grad_norm": 0.12183618545532227, "learning_rate": 1.4909886436992022e-05, "loss": 0.0162, "step": 31610 }, { "epoch": 9.199883619435555, "grad_norm": 0.08936697244644165, "learning_rate": 1.4901619251198664e-05, "loss": 0.0166, "step": 31620 }, { "epoch": 9.202793133546697, "grad_norm": 0.10480847209692001, "learning_rate": 1.4893352095290774e-05, "loss": 0.017, "step": 31630 }, { "epoch": 9.20570264765784, "grad_norm": 0.0972481369972229, "learning_rate": 1.4885084971779706e-05, "loss": 0.0166, "step": 31640 }, { "epoch": 9.208612161768984, "grad_norm": 0.0826442763209343, "learning_rate": 1.4876817883176784e-05, "loss": 0.0156, "step": 31650 }, { "epoch": 9.211521675880128, "grad_norm": 0.09436764568090439, "learning_rate": 1.486855083199334e-05, "loss": 0.0161, "step": 31660 }, { "epoch": 9.214431189991272, "grad_norm": 0.11004743725061417, "learning_rate": 1.4860283820740685e-05, "loss": 0.0161, "step": 31670 }, { "epoch": 9.217340704102416, "grad_norm": 0.08758048713207245, "learning_rate": 1.4852016851930112e-05, "loss": 0.015, "step": 31680 }, { "epoch": 9.220250218213558, "grad_norm": 0.11005257070064545, "learning_rate": 1.4843749928072917e-05, "loss": 0.0174, "step": 31690 }, { "epoch": 9.223159732324701, "grad_norm": 0.08796889334917068, "learning_rate": 1.4835483051680362e-05, "loss": 0.0173, "step": 31700 }, { "epoch": 9.226069246435845, "grad_norm": 0.08185217529535294, "learning_rate": 1.4827216225263713e-05, "loss": 0.0166, "step": 31710 }, { "epoch": 9.228978760546989, "grad_norm": 0.08719804883003235, "learning_rate": 1.481894945133422e-05, "loss": 0.015, "step": 31720 }, { "epoch": 9.231888274658132, "grad_norm": 0.08325599879026413, "learning_rate": 1.4810682732403095e-05, "loss": 0.0157, "step": 31730 }, { "epoch": 9.234797788769276, "grad_norm": 0.08453196287155151, "learning_rate": 1.4802416070981565e-05, "loss": 0.0164, "step": 31740 }, { "epoch": 9.23770730288042, "grad_norm": 0.09091883152723312, "learning_rate": 1.4794149469580809e-05, "loss": 0.0161, "step": 31750 }, { "epoch": 9.240616816991562, "grad_norm": 0.11416573077440262, "learning_rate": 1.4785882930712009e-05, "loss": 0.0161, "step": 31760 }, { "epoch": 9.243526331102705, "grad_norm": 0.14469878375530243, "learning_rate": 1.4777616456886327e-05, "loss": 0.0154, "step": 31770 }, { "epoch": 9.246435845213849, "grad_norm": 0.11334595084190369, "learning_rate": 1.4769350050614889e-05, "loss": 0.0158, "step": 31780 }, { "epoch": 9.249345359324993, "grad_norm": 0.10551277548074722, "learning_rate": 1.4761083714408818e-05, "loss": 0.0159, "step": 31790 }, { "epoch": 9.252254873436137, "grad_norm": 0.11369174718856812, "learning_rate": 1.4752817450779203e-05, "loss": 0.0174, "step": 31800 }, { "epoch": 9.25516438754728, "grad_norm": 0.12304098159074783, "learning_rate": 1.4744551262237122e-05, "loss": 0.0171, "step": 31810 }, { "epoch": 9.258073901658424, "grad_norm": 0.10936714708805084, "learning_rate": 1.4736285151293624e-05, "loss": 0.0163, "step": 31820 }, { "epoch": 9.260983415769566, "grad_norm": 0.08585745841264725, "learning_rate": 1.472801912045973e-05, "loss": 0.0165, "step": 31830 }, { "epoch": 9.26389292988071, "grad_norm": 0.12529027462005615, "learning_rate": 1.4719753172246454e-05, "loss": 0.0168, "step": 31840 }, { "epoch": 9.266802443991853, "grad_norm": 0.10473466664552689, "learning_rate": 1.4711487309164758e-05, "loss": 0.0179, "step": 31850 }, { "epoch": 9.269711958102997, "grad_norm": 0.0974607989192009, "learning_rate": 1.4703221533725598e-05, "loss": 0.017, "step": 31860 }, { "epoch": 9.27262147221414, "grad_norm": 0.11276291310787201, "learning_rate": 1.4694955848439909e-05, "loss": 0.0165, "step": 31870 }, { "epoch": 9.275530986325284, "grad_norm": 0.10199007391929626, "learning_rate": 1.468669025581857e-05, "loss": 0.0174, "step": 31880 }, { "epoch": 9.278440500436426, "grad_norm": 0.1025705561041832, "learning_rate": 1.4678424758372468e-05, "loss": 0.0167, "step": 31890 }, { "epoch": 9.28135001454757, "grad_norm": 0.0988611951470375, "learning_rate": 1.4670159358612425e-05, "loss": 0.016, "step": 31900 }, { "epoch": 9.284259528658714, "grad_norm": 0.10873701423406601, "learning_rate": 1.466189405904926e-05, "loss": 0.0171, "step": 31910 }, { "epoch": 9.287169042769857, "grad_norm": 0.08469417691230774, "learning_rate": 1.4653628862193761e-05, "loss": 0.0168, "step": 31920 }, { "epoch": 9.290078556881001, "grad_norm": 0.11030808091163635, "learning_rate": 1.464536377055666e-05, "loss": 0.0167, "step": 31930 }, { "epoch": 9.292988070992145, "grad_norm": 0.1013009324669838, "learning_rate": 1.4637098786648687e-05, "loss": 0.0163, "step": 31940 }, { "epoch": 9.295897585103289, "grad_norm": 0.09646546095609665, "learning_rate": 1.4628833912980515e-05, "loss": 0.0152, "step": 31950 }, { "epoch": 9.29880709921443, "grad_norm": 0.08359614759683609, "learning_rate": 1.4620569152062804e-05, "loss": 0.0163, "step": 31960 }, { "epoch": 9.301716613325574, "grad_norm": 0.09089069068431854, "learning_rate": 1.4612304506406169e-05, "loss": 0.0167, "step": 31970 }, { "epoch": 9.304626127436718, "grad_norm": 0.10842322558164597, "learning_rate": 1.4604039978521183e-05, "loss": 0.0147, "step": 31980 }, { "epoch": 9.307535641547862, "grad_norm": 0.10259244590997696, "learning_rate": 1.4595775570918402e-05, "loss": 0.0161, "step": 31990 }, { "epoch": 9.310445155659005, "grad_norm": 0.095454141497612, "learning_rate": 1.4587511286108325e-05, "loss": 0.0172, "step": 32000 }, { "epoch": 9.313354669770149, "grad_norm": 0.09404460340738297, "learning_rate": 1.4579247126601435e-05, "loss": 0.017, "step": 32010 }, { "epoch": 9.316264183881291, "grad_norm": 0.08882125467061996, "learning_rate": 1.4570983094908151e-05, "loss": 0.0162, "step": 32020 }, { "epoch": 9.319173697992435, "grad_norm": 0.08877360820770264, "learning_rate": 1.4562719193538877e-05, "loss": 0.0176, "step": 32030 }, { "epoch": 9.322083212103578, "grad_norm": 0.09526687860488892, "learning_rate": 1.455445542500397e-05, "loss": 0.0152, "step": 32040 }, { "epoch": 9.324992726214722, "grad_norm": 0.08932705968618393, "learning_rate": 1.4546191791813736e-05, "loss": 0.0164, "step": 32050 }, { "epoch": 9.327902240325866, "grad_norm": 0.11368057876825333, "learning_rate": 1.4537928296478457e-05, "loss": 0.0152, "step": 32060 }, { "epoch": 9.33081175443701, "grad_norm": 0.09258165955543518, "learning_rate": 1.452966494150836e-05, "loss": 0.0162, "step": 32070 }, { "epoch": 9.333721268548153, "grad_norm": 0.08413653075695038, "learning_rate": 1.4521401729413633e-05, "loss": 0.0156, "step": 32080 }, { "epoch": 9.336630782659295, "grad_norm": 0.10732846707105637, "learning_rate": 1.4513138662704428e-05, "loss": 0.0174, "step": 32090 }, { "epoch": 9.339540296770439, "grad_norm": 0.07841528952121735, "learning_rate": 1.4504875743890839e-05, "loss": 0.0154, "step": 32100 }, { "epoch": 9.342449810881583, "grad_norm": 0.09964699298143387, "learning_rate": 1.4496612975482931e-05, "loss": 0.0172, "step": 32110 }, { "epoch": 9.345359324992726, "grad_norm": 0.10013481229543686, "learning_rate": 1.4488350359990705e-05, "loss": 0.0151, "step": 32120 }, { "epoch": 9.34826883910387, "grad_norm": 0.09279962629079819, "learning_rate": 1.4480087899924129e-05, "loss": 0.0163, "step": 32130 }, { "epoch": 9.351178353215014, "grad_norm": 0.08478882908821106, "learning_rate": 1.4471825597793129e-05, "loss": 0.0154, "step": 32140 }, { "epoch": 9.354087867326157, "grad_norm": 0.09082057327032089, "learning_rate": 1.4463563456107559e-05, "loss": 0.0166, "step": 32150 }, { "epoch": 9.3569973814373, "grad_norm": 0.09623721987009048, "learning_rate": 1.4455301477377255e-05, "loss": 0.0166, "step": 32160 }, { "epoch": 9.359906895548443, "grad_norm": 0.09782104194164276, "learning_rate": 1.4447039664111973e-05, "loss": 0.0154, "step": 32170 }, { "epoch": 9.362816409659587, "grad_norm": 0.09183979779481888, "learning_rate": 1.4438778018821441e-05, "loss": 0.0153, "step": 32180 }, { "epoch": 9.36572592377073, "grad_norm": 0.08120181411504745, "learning_rate": 1.4430516544015336e-05, "loss": 0.0185, "step": 32190 }, { "epoch": 9.368635437881874, "grad_norm": 0.0873057022690773, "learning_rate": 1.442225524220326e-05, "loss": 0.016, "step": 32200 }, { "epoch": 9.371544951993018, "grad_norm": 0.14605535566806793, "learning_rate": 1.4413994115894792e-05, "loss": 0.0171, "step": 32210 }, { "epoch": 9.37445446610416, "grad_norm": 0.09121130406856537, "learning_rate": 1.440573316759944e-05, "loss": 0.0165, "step": 32220 }, { "epoch": 9.377363980215303, "grad_norm": 0.11727946996688843, "learning_rate": 1.4397472399826657e-05, "loss": 0.0165, "step": 32230 }, { "epoch": 9.380273494326447, "grad_norm": 0.08797433972358704, "learning_rate": 1.4389211815085852e-05, "loss": 0.0151, "step": 32240 }, { "epoch": 9.383183008437591, "grad_norm": 0.0869416892528534, "learning_rate": 1.4380951415886368e-05, "loss": 0.0159, "step": 32250 }, { "epoch": 9.386092522548735, "grad_norm": 0.10696879029273987, "learning_rate": 1.4372691204737505e-05, "loss": 0.0149, "step": 32260 }, { "epoch": 9.389002036659878, "grad_norm": 0.09406039118766785, "learning_rate": 1.4364431184148485e-05, "loss": 0.0161, "step": 32270 }, { "epoch": 9.391911550771022, "grad_norm": 0.10025027394294739, "learning_rate": 1.435617135662849e-05, "loss": 0.0169, "step": 32280 }, { "epoch": 9.394821064882164, "grad_norm": 0.10094588994979858, "learning_rate": 1.4347911724686645e-05, "loss": 0.0148, "step": 32290 }, { "epoch": 9.397730578993308, "grad_norm": 0.09354131668806076, "learning_rate": 1.4339652290831991e-05, "loss": 0.0153, "step": 32300 }, { "epoch": 9.400640093104451, "grad_norm": 0.08467317372560501, "learning_rate": 1.4331393057573547e-05, "loss": 0.0147, "step": 32310 }, { "epoch": 9.403549607215595, "grad_norm": 0.0949355810880661, "learning_rate": 1.4323134027420227e-05, "loss": 0.0154, "step": 32320 }, { "epoch": 9.406459121326739, "grad_norm": 0.0924554169178009, "learning_rate": 1.431487520288092e-05, "loss": 0.0148, "step": 32330 }, { "epoch": 9.409368635437882, "grad_norm": 0.09319765865802765, "learning_rate": 1.4306616586464444e-05, "loss": 0.0166, "step": 32340 }, { "epoch": 9.412278149549024, "grad_norm": 0.09759948402643204, "learning_rate": 1.4298358180679534e-05, "loss": 0.0159, "step": 32350 }, { "epoch": 9.415187663660168, "grad_norm": 0.11083949357271194, "learning_rate": 1.4290099988034888e-05, "loss": 0.0168, "step": 32360 }, { "epoch": 9.418097177771312, "grad_norm": 0.11062198132276535, "learning_rate": 1.4281842011039119e-05, "loss": 0.0147, "step": 32370 }, { "epoch": 9.421006691882456, "grad_norm": 0.08929240703582764, "learning_rate": 1.4273584252200783e-05, "loss": 0.0157, "step": 32380 }, { "epoch": 9.4239162059936, "grad_norm": 0.09916242957115173, "learning_rate": 1.4265326714028374e-05, "loss": 0.0163, "step": 32390 }, { "epoch": 9.426825720104743, "grad_norm": 0.09114837646484375, "learning_rate": 1.4257069399030308e-05, "loss": 0.0155, "step": 32400 }, { "epoch": 9.429735234215887, "grad_norm": 0.08501191437244415, "learning_rate": 1.424881230971495e-05, "loss": 0.0163, "step": 32410 }, { "epoch": 9.432644748327029, "grad_norm": 0.10000558942556381, "learning_rate": 1.4240555448590572e-05, "loss": 0.0177, "step": 32420 }, { "epoch": 9.435554262438172, "grad_norm": 0.11119459569454193, "learning_rate": 1.4232298818165404e-05, "loss": 0.0164, "step": 32430 }, { "epoch": 9.438463776549316, "grad_norm": 0.08875039964914322, "learning_rate": 1.4224042420947576e-05, "loss": 0.0152, "step": 32440 }, { "epoch": 9.44137329066046, "grad_norm": 0.09303630888462067, "learning_rate": 1.4215786259445176e-05, "loss": 0.015, "step": 32450 }, { "epoch": 9.444282804771603, "grad_norm": 0.12328226864337921, "learning_rate": 1.4207530336166209e-05, "loss": 0.015, "step": 32460 }, { "epoch": 9.447192318882747, "grad_norm": 0.10749973356723785, "learning_rate": 1.4199274653618593e-05, "loss": 0.0161, "step": 32470 }, { "epoch": 9.45010183299389, "grad_norm": 0.10049011558294296, "learning_rate": 1.4191019214310205e-05, "loss": 0.015, "step": 32480 }, { "epoch": 9.453011347105033, "grad_norm": 0.09012830257415771, "learning_rate": 1.4182764020748813e-05, "loss": 0.0147, "step": 32490 }, { "epoch": 9.455920861216176, "grad_norm": 0.09323642402887344, "learning_rate": 1.4174509075442132e-05, "loss": 0.0179, "step": 32500 }, { "epoch": 9.45883037532732, "grad_norm": 0.10227085649967194, "learning_rate": 1.4166254380897804e-05, "loss": 0.0167, "step": 32510 }, { "epoch": 9.461739889438464, "grad_norm": 0.08929472416639328, "learning_rate": 1.4157999939623377e-05, "loss": 0.0167, "step": 32520 }, { "epoch": 9.464649403549608, "grad_norm": 0.0779775083065033, "learning_rate": 1.4149745754126341e-05, "loss": 0.0162, "step": 32530 }, { "epoch": 9.467558917660751, "grad_norm": 0.10675351321697235, "learning_rate": 1.414149182691409e-05, "loss": 0.0155, "step": 32540 }, { "epoch": 9.470468431771893, "grad_norm": 0.1099664717912674, "learning_rate": 1.4133238160493956e-05, "loss": 0.0157, "step": 32550 }, { "epoch": 9.473377945883037, "grad_norm": 0.07509981095790863, "learning_rate": 1.4124984757373187e-05, "loss": 0.0145, "step": 32560 }, { "epoch": 9.47628745999418, "grad_norm": 0.08473297953605652, "learning_rate": 1.411673162005894e-05, "loss": 0.0149, "step": 32570 }, { "epoch": 9.479196974105324, "grad_norm": 0.111446812748909, "learning_rate": 1.4108478751058313e-05, "loss": 0.0154, "step": 32580 }, { "epoch": 9.482106488216468, "grad_norm": 0.11063645035028458, "learning_rate": 1.4100226152878295e-05, "loss": 0.015, "step": 32590 }, { "epoch": 9.485016002327612, "grad_norm": 0.09954550862312317, "learning_rate": 1.4091973828025814e-05, "loss": 0.0159, "step": 32600 }, { "epoch": 9.487925516438755, "grad_norm": 0.08591616898775101, "learning_rate": 1.4083721779007717e-05, "loss": 0.0155, "step": 32610 }, { "epoch": 9.490835030549897, "grad_norm": 0.0902068093419075, "learning_rate": 1.4075470008330742e-05, "loss": 0.0167, "step": 32620 }, { "epoch": 9.493744544661041, "grad_norm": 0.09544514119625092, "learning_rate": 1.4067218518501573e-05, "loss": 0.0156, "step": 32630 }, { "epoch": 9.496654058772185, "grad_norm": 0.0885588601231575, "learning_rate": 1.4058967312026789e-05, "loss": 0.0182, "step": 32640 }, { "epoch": 9.499563572883329, "grad_norm": 0.10400356352329254, "learning_rate": 1.4050716391412884e-05, "loss": 0.016, "step": 32650 }, { "epoch": 9.502473086994472, "grad_norm": 0.08483724296092987, "learning_rate": 1.4042465759166282e-05, "loss": 0.0168, "step": 32660 }, { "epoch": 9.505382601105616, "grad_norm": 0.09947464615106583, "learning_rate": 1.4034215417793292e-05, "loss": 0.017, "step": 32670 }, { "epoch": 9.508292115216758, "grad_norm": 0.11442410945892334, "learning_rate": 1.4025965369800167e-05, "loss": 0.0159, "step": 32680 }, { "epoch": 9.511201629327902, "grad_norm": 0.09014861285686493, "learning_rate": 1.4017715617693038e-05, "loss": 0.0162, "step": 32690 }, { "epoch": 9.514111143439045, "grad_norm": 0.09995733201503754, "learning_rate": 1.400946616397797e-05, "loss": 0.0164, "step": 32700 }, { "epoch": 9.517020657550189, "grad_norm": 0.09463772177696228, "learning_rate": 1.4001217011160932e-05, "loss": 0.0165, "step": 32710 }, { "epoch": 9.519930171661333, "grad_norm": 0.08570963144302368, "learning_rate": 1.399296816174779e-05, "loss": 0.0164, "step": 32720 }, { "epoch": 9.522839685772476, "grad_norm": 0.09251432120800018, "learning_rate": 1.3984719618244336e-05, "loss": 0.0146, "step": 32730 }, { "epoch": 9.52574919988362, "grad_norm": 0.06468988209962845, "learning_rate": 1.397647138315625e-05, "loss": 0.0166, "step": 32740 }, { "epoch": 9.528658713994762, "grad_norm": 0.096041239798069, "learning_rate": 1.3968223458989133e-05, "loss": 0.0164, "step": 32750 }, { "epoch": 9.531568228105906, "grad_norm": 0.09566929936408997, "learning_rate": 1.3959975848248492e-05, "loss": 0.0149, "step": 32760 }, { "epoch": 9.53447774221705, "grad_norm": 0.10482487827539444, "learning_rate": 1.3951728553439724e-05, "loss": 0.0156, "step": 32770 }, { "epoch": 9.537387256328193, "grad_norm": 0.08667465299367905, "learning_rate": 1.3943481577068145e-05, "loss": 0.0142, "step": 32780 }, { "epoch": 9.540296770439337, "grad_norm": 0.12307582795619965, "learning_rate": 1.3935234921638967e-05, "loss": 0.0149, "step": 32790 }, { "epoch": 9.54320628455048, "grad_norm": 0.10636352002620697, "learning_rate": 1.3926988589657303e-05, "loss": 0.0148, "step": 32800 }, { "epoch": 9.546115798661624, "grad_norm": 0.09259574860334396, "learning_rate": 1.3918742583628179e-05, "loss": 0.0153, "step": 32810 }, { "epoch": 9.549025312772766, "grad_norm": 0.10641936957836151, "learning_rate": 1.3910496906056504e-05, "loss": 0.0162, "step": 32820 }, { "epoch": 9.55193482688391, "grad_norm": 0.08956436067819595, "learning_rate": 1.390225155944711e-05, "loss": 0.0153, "step": 32830 }, { "epoch": 9.554844340995054, "grad_norm": 0.11302214860916138, "learning_rate": 1.38940065463047e-05, "loss": 0.0157, "step": 32840 }, { "epoch": 9.557753855106197, "grad_norm": 0.09487991780042648, "learning_rate": 1.3885761869133907e-05, "loss": 0.0147, "step": 32850 }, { "epoch": 9.560663369217341, "grad_norm": 0.07373865693807602, "learning_rate": 1.387751753043923e-05, "loss": 0.0158, "step": 32860 }, { "epoch": 9.563572883328485, "grad_norm": 0.10602781921625137, "learning_rate": 1.3869273532725091e-05, "loss": 0.0171, "step": 32870 }, { "epoch": 9.566482397439628, "grad_norm": 0.11242394894361496, "learning_rate": 1.3861029878495807e-05, "loss": 0.0152, "step": 32880 }, { "epoch": 9.56939191155077, "grad_norm": 0.10261470079421997, "learning_rate": 1.3852786570255562e-05, "loss": 0.0169, "step": 32890 }, { "epoch": 9.572301425661914, "grad_norm": 0.08747435361146927, "learning_rate": 1.3844543610508475e-05, "loss": 0.0174, "step": 32900 }, { "epoch": 9.575210939773058, "grad_norm": 0.10400641709566116, "learning_rate": 1.3836301001758526e-05, "loss": 0.0152, "step": 32910 }, { "epoch": 9.578120453884202, "grad_norm": 0.08401966840028763, "learning_rate": 1.3828058746509607e-05, "loss": 0.0149, "step": 32920 }, { "epoch": 9.581029967995345, "grad_norm": 0.08842923492193222, "learning_rate": 1.3819816847265498e-05, "loss": 0.0153, "step": 32930 }, { "epoch": 9.583939482106489, "grad_norm": 0.1164076030254364, "learning_rate": 1.3811575306529872e-05, "loss": 0.0154, "step": 32940 }, { "epoch": 9.58684899621763, "grad_norm": 0.09576044231653214, "learning_rate": 1.3803334126806289e-05, "loss": 0.016, "step": 32950 }, { "epoch": 9.589758510328775, "grad_norm": 0.08378253877162933, "learning_rate": 1.3795093310598198e-05, "loss": 0.0153, "step": 32960 }, { "epoch": 9.592668024439918, "grad_norm": 0.08375579863786697, "learning_rate": 1.3786852860408946e-05, "loss": 0.0167, "step": 32970 }, { "epoch": 9.595577538551062, "grad_norm": 0.08165834099054337, "learning_rate": 1.377861277874177e-05, "loss": 0.0155, "step": 32980 }, { "epoch": 9.598487052662206, "grad_norm": 0.10487306118011475, "learning_rate": 1.3770373068099778e-05, "loss": 0.0153, "step": 32990 }, { "epoch": 9.60139656677335, "grad_norm": 0.07995253056287766, "learning_rate": 1.3762133730985985e-05, "loss": 0.0158, "step": 33000 }, { "epoch": 9.604306080884493, "grad_norm": 0.08219297975301743, "learning_rate": 1.3753894769903276e-05, "loss": 0.0157, "step": 33010 }, { "epoch": 9.607215594995635, "grad_norm": 0.09429360926151276, "learning_rate": 1.3745656187354434e-05, "loss": 0.0155, "step": 33020 }, { "epoch": 9.610125109106779, "grad_norm": 0.09566749632358551, "learning_rate": 1.373741798584213e-05, "loss": 0.016, "step": 33030 }, { "epoch": 9.613034623217922, "grad_norm": 0.11219678819179535, "learning_rate": 1.3729180167868899e-05, "loss": 0.0151, "step": 33040 }, { "epoch": 9.615944137329066, "grad_norm": 0.09844381362199783, "learning_rate": 1.3720942735937184e-05, "loss": 0.0161, "step": 33050 }, { "epoch": 9.61885365144021, "grad_norm": 0.10449547320604324, "learning_rate": 1.371270569254929e-05, "loss": 0.0168, "step": 33060 }, { "epoch": 9.621763165551354, "grad_norm": 0.10430438071489334, "learning_rate": 1.3704469040207417e-05, "loss": 0.0167, "step": 33070 }, { "epoch": 9.624672679662496, "grad_norm": 0.10212738066911697, "learning_rate": 1.3696232781413646e-05, "loss": 0.0154, "step": 33080 }, { "epoch": 9.62758219377364, "grad_norm": 0.1005038172006607, "learning_rate": 1.3687996918669933e-05, "loss": 0.0154, "step": 33090 }, { "epoch": 9.630491707884783, "grad_norm": 0.11224083602428436, "learning_rate": 1.3679761454478112e-05, "loss": 0.0149, "step": 33100 }, { "epoch": 9.633401221995927, "grad_norm": 0.09418526291847229, "learning_rate": 1.36715263913399e-05, "loss": 0.0155, "step": 33110 }, { "epoch": 9.63631073610707, "grad_norm": 0.13841359317302704, "learning_rate": 1.3663291731756895e-05, "loss": 0.0152, "step": 33120 }, { "epoch": 9.639220250218214, "grad_norm": 0.09929917752742767, "learning_rate": 1.3655057478230573e-05, "loss": 0.0165, "step": 33130 }, { "epoch": 9.642129764329358, "grad_norm": 0.12795515358448029, "learning_rate": 1.3646823633262274e-05, "loss": 0.0152, "step": 33140 }, { "epoch": 9.6450392784405, "grad_norm": 0.10107401758432388, "learning_rate": 1.3638590199353232e-05, "loss": 0.0168, "step": 33150 }, { "epoch": 9.647948792551643, "grad_norm": 0.09627226740121841, "learning_rate": 1.3630357179004535e-05, "loss": 0.0164, "step": 33160 }, { "epoch": 9.650858306662787, "grad_norm": 0.0939495861530304, "learning_rate": 1.3622124574717164e-05, "loss": 0.017, "step": 33170 }, { "epoch": 9.65376782077393, "grad_norm": 0.08529265969991684, "learning_rate": 1.3613892388991971e-05, "loss": 0.0161, "step": 33180 }, { "epoch": 9.656677334885075, "grad_norm": 0.09984976053237915, "learning_rate": 1.3605660624329667e-05, "loss": 0.0164, "step": 33190 }, { "epoch": 9.659586848996218, "grad_norm": 0.1161544993519783, "learning_rate": 1.3597429283230857e-05, "loss": 0.0152, "step": 33200 }, { "epoch": 9.662496363107362, "grad_norm": 0.11846214532852173, "learning_rate": 1.3589198368195993e-05, "loss": 0.0172, "step": 33210 }, { "epoch": 9.665405877218504, "grad_norm": 0.08730077743530273, "learning_rate": 1.3580967881725415e-05, "loss": 0.0148, "step": 33220 }, { "epoch": 9.668315391329648, "grad_norm": 0.09149779379367828, "learning_rate": 1.3572737826319327e-05, "loss": 0.0146, "step": 33230 }, { "epoch": 9.671224905440791, "grad_norm": 0.10046450048685074, "learning_rate": 1.3564508204477803e-05, "loss": 0.016, "step": 33240 }, { "epoch": 9.674134419551935, "grad_norm": 0.11730194836854935, "learning_rate": 1.3556279018700785e-05, "loss": 0.0151, "step": 33250 }, { "epoch": 9.677043933663079, "grad_norm": 0.09765754640102386, "learning_rate": 1.354805027148808e-05, "loss": 0.0149, "step": 33260 }, { "epoch": 9.679953447774222, "grad_norm": 0.09615422040224075, "learning_rate": 1.353982196533937e-05, "loss": 0.0167, "step": 33270 }, { "epoch": 9.682862961885364, "grad_norm": 0.08882171660661697, "learning_rate": 1.3531594102754188e-05, "loss": 0.0157, "step": 33280 }, { "epoch": 9.685772475996508, "grad_norm": 0.08728436380624771, "learning_rate": 1.3523366686231949e-05, "loss": 0.0155, "step": 33290 }, { "epoch": 9.688681990107652, "grad_norm": 0.10506363213062286, "learning_rate": 1.3515139718271925e-05, "loss": 0.0161, "step": 33300 }, { "epoch": 9.691591504218795, "grad_norm": 0.09096923470497131, "learning_rate": 1.3506913201373245e-05, "loss": 0.0149, "step": 33310 }, { "epoch": 9.69450101832994, "grad_norm": 0.10697800666093826, "learning_rate": 1.3498687138034918e-05, "loss": 0.0165, "step": 33320 }, { "epoch": 9.697410532441083, "grad_norm": 0.10379841923713684, "learning_rate": 1.3490461530755792e-05, "loss": 0.0184, "step": 33330 }, { "epoch": 9.700320046552227, "grad_norm": 0.08163736760616302, "learning_rate": 1.3482236382034595e-05, "loss": 0.0159, "step": 33340 }, { "epoch": 9.703229560663369, "grad_norm": 0.08713623881340027, "learning_rate": 1.3474011694369917e-05, "loss": 0.0144, "step": 33350 }, { "epoch": 9.706139074774512, "grad_norm": 0.12183257192373276, "learning_rate": 1.346578747026019e-05, "loss": 0.0154, "step": 33360 }, { "epoch": 9.709048588885656, "grad_norm": 0.11318934708833694, "learning_rate": 1.3457563712203725e-05, "loss": 0.0166, "step": 33370 }, { "epoch": 9.7119581029968, "grad_norm": 0.07194750756025314, "learning_rate": 1.3449340422698673e-05, "loss": 0.0149, "step": 33380 }, { "epoch": 9.714867617107943, "grad_norm": 0.10939237475395203, "learning_rate": 1.3441117604243063e-05, "loss": 0.0157, "step": 33390 }, { "epoch": 9.717777131219087, "grad_norm": 0.11201560497283936, "learning_rate": 1.3432895259334764e-05, "loss": 0.0151, "step": 33400 }, { "epoch": 9.720686645330229, "grad_norm": 0.11187876760959625, "learning_rate": 1.3424673390471504e-05, "loss": 0.0164, "step": 33410 }, { "epoch": 9.723596159441373, "grad_norm": 0.1006801649928093, "learning_rate": 1.3416452000150883e-05, "loss": 0.0164, "step": 33420 }, { "epoch": 9.726505673552516, "grad_norm": 0.11661835014820099, "learning_rate": 1.3408231090870324e-05, "loss": 0.0172, "step": 33430 }, { "epoch": 9.72941518766366, "grad_norm": 0.08843022584915161, "learning_rate": 1.340001066512713e-05, "loss": 0.0152, "step": 33440 }, { "epoch": 9.732324701774804, "grad_norm": 0.08717059344053268, "learning_rate": 1.3391790725418463e-05, "loss": 0.0158, "step": 33450 }, { "epoch": 9.735234215885948, "grad_norm": 0.09202715009450912, "learning_rate": 1.33835712742413e-05, "loss": 0.0152, "step": 33460 }, { "epoch": 9.738143729997091, "grad_norm": 0.08966484665870667, "learning_rate": 1.3375352314092513e-05, "loss": 0.0169, "step": 33470 }, { "epoch": 9.741053244108233, "grad_norm": 0.08308511972427368, "learning_rate": 1.336713384746879e-05, "loss": 0.0141, "step": 33480 }, { "epoch": 9.743962758219377, "grad_norm": 0.11721006780862808, "learning_rate": 1.335891587686669e-05, "loss": 0.016, "step": 33490 }, { "epoch": 9.74687227233052, "grad_norm": 0.10250768065452576, "learning_rate": 1.3350698404782624e-05, "loss": 0.0161, "step": 33500 }, { "epoch": 9.749781786441664, "grad_norm": 0.08308625966310501, "learning_rate": 1.3342481433712827e-05, "loss": 0.0155, "step": 33510 }, { "epoch": 9.752691300552808, "grad_norm": 0.09661974757909775, "learning_rate": 1.333426496615341e-05, "loss": 0.0138, "step": 33520 }, { "epoch": 9.755600814663952, "grad_norm": 0.09068183600902557, "learning_rate": 1.3326049004600312e-05, "loss": 0.0151, "step": 33530 }, { "epoch": 9.758510328775095, "grad_norm": 0.08463626354932785, "learning_rate": 1.3317833551549334e-05, "loss": 0.015, "step": 33540 }, { "epoch": 9.761419842886237, "grad_norm": 0.08884880691766739, "learning_rate": 1.3309618609496104e-05, "loss": 0.0145, "step": 33550 }, { "epoch": 9.764329356997381, "grad_norm": 0.11835556477308273, "learning_rate": 1.3301404180936108e-05, "loss": 0.0155, "step": 33560 }, { "epoch": 9.767238871108525, "grad_norm": 0.0976843610405922, "learning_rate": 1.3293190268364679e-05, "loss": 0.0147, "step": 33570 }, { "epoch": 9.770148385219668, "grad_norm": 0.07689375430345535, "learning_rate": 1.3284976874276973e-05, "loss": 0.0157, "step": 33580 }, { "epoch": 9.773057899330812, "grad_norm": 0.08720552176237106, "learning_rate": 1.327676400116801e-05, "loss": 0.0179, "step": 33590 }, { "epoch": 9.775967413441956, "grad_norm": 0.09446143358945847, "learning_rate": 1.3268551651532652e-05, "loss": 0.0156, "step": 33600 }, { "epoch": 9.7788769275531, "grad_norm": 0.0927794799208641, "learning_rate": 1.326033982786558e-05, "loss": 0.0151, "step": 33610 }, { "epoch": 9.781786441664241, "grad_norm": 0.09271299093961716, "learning_rate": 1.325212853266134e-05, "loss": 0.0159, "step": 33620 }, { "epoch": 9.784695955775385, "grad_norm": 0.08499102294445038, "learning_rate": 1.3243917768414297e-05, "loss": 0.0153, "step": 33630 }, { "epoch": 9.787605469886529, "grad_norm": 0.11990822106599808, "learning_rate": 1.3235707537618671e-05, "loss": 0.0158, "step": 33640 }, { "epoch": 9.790514983997673, "grad_norm": 0.0971517488360405, "learning_rate": 1.3227497842768513e-05, "loss": 0.0156, "step": 33650 }, { "epoch": 9.793424498108816, "grad_norm": 0.06969733536243439, "learning_rate": 1.3219288686357705e-05, "loss": 0.016, "step": 33660 }, { "epoch": 9.79633401221996, "grad_norm": 0.09533010423183441, "learning_rate": 1.3211080070879983e-05, "loss": 0.0147, "step": 33670 }, { "epoch": 9.799243526331102, "grad_norm": 0.11747326701879501, "learning_rate": 1.3202871998828896e-05, "loss": 0.016, "step": 33680 }, { "epoch": 9.802153040442246, "grad_norm": 0.09582477062940598, "learning_rate": 1.3194664472697852e-05, "loss": 0.0136, "step": 33690 }, { "epoch": 9.80506255455339, "grad_norm": 0.12323487550020218, "learning_rate": 1.318645749498007e-05, "loss": 0.0159, "step": 33700 }, { "epoch": 9.807972068664533, "grad_norm": 0.13887648284435272, "learning_rate": 1.3178251068168615e-05, "loss": 0.0145, "step": 33710 }, { "epoch": 9.810881582775677, "grad_norm": 0.08977581560611725, "learning_rate": 1.3170045194756392e-05, "loss": 0.0169, "step": 33720 }, { "epoch": 9.81379109688682, "grad_norm": 0.10027999430894852, "learning_rate": 1.3161839877236114e-05, "loss": 0.0171, "step": 33730 }, { "epoch": 9.816700610997962, "grad_norm": 0.08590882271528244, "learning_rate": 1.3153635118100357e-05, "loss": 0.0154, "step": 33740 }, { "epoch": 9.819610125109106, "grad_norm": 0.09235526621341705, "learning_rate": 1.3145430919841494e-05, "loss": 0.0147, "step": 33750 }, { "epoch": 9.82251963922025, "grad_norm": 0.12803344428539276, "learning_rate": 1.3137227284951749e-05, "loss": 0.0167, "step": 33760 }, { "epoch": 9.825429153331394, "grad_norm": 0.12629485130310059, "learning_rate": 1.3129024215923177e-05, "loss": 0.0151, "step": 33770 }, { "epoch": 9.828338667442537, "grad_norm": 0.1280989944934845, "learning_rate": 1.3120821715247642e-05, "loss": 0.0156, "step": 33780 }, { "epoch": 9.831248181553681, "grad_norm": 0.09280861169099808, "learning_rate": 1.3112619785416856e-05, "loss": 0.0144, "step": 33790 }, { "epoch": 9.834157695664825, "grad_norm": 0.10100240260362625, "learning_rate": 1.3104418428922347e-05, "loss": 0.0163, "step": 33800 }, { "epoch": 9.837067209775967, "grad_norm": 0.0925745889544487, "learning_rate": 1.3096217648255465e-05, "loss": 0.0153, "step": 33810 }, { "epoch": 9.83997672388711, "grad_norm": 0.09926179051399231, "learning_rate": 1.3088017445907399e-05, "loss": 0.0166, "step": 33820 }, { "epoch": 9.842886237998254, "grad_norm": 0.12345223128795624, "learning_rate": 1.3079817824369145e-05, "loss": 0.0153, "step": 33830 }, { "epoch": 9.845795752109398, "grad_norm": 0.12536561489105225, "learning_rate": 1.3071618786131544e-05, "loss": 0.0148, "step": 33840 }, { "epoch": 9.848705266220541, "grad_norm": 0.08646860718727112, "learning_rate": 1.3063420333685235e-05, "loss": 0.0161, "step": 33850 }, { "epoch": 9.851614780331685, "grad_norm": 0.1236172541975975, "learning_rate": 1.3055222469520693e-05, "loss": 0.0164, "step": 33860 }, { "epoch": 9.854524294442829, "grad_norm": 0.09142203629016876, "learning_rate": 1.3047025196128226e-05, "loss": 0.0144, "step": 33870 }, { "epoch": 9.85743380855397, "grad_norm": 0.12573851644992828, "learning_rate": 1.303882851599793e-05, "loss": 0.0176, "step": 33880 }, { "epoch": 9.860343322665114, "grad_norm": 0.11355161666870117, "learning_rate": 1.3030632431619758e-05, "loss": 0.0165, "step": 33890 }, { "epoch": 9.863252836776258, "grad_norm": 0.09283009171485901, "learning_rate": 1.3022436945483447e-05, "loss": 0.0159, "step": 33900 }, { "epoch": 9.866162350887402, "grad_norm": 0.12050113081932068, "learning_rate": 1.3014242060078578e-05, "loss": 0.0151, "step": 33910 }, { "epoch": 9.869071864998546, "grad_norm": 0.1004074439406395, "learning_rate": 1.3006047777894544e-05, "loss": 0.0155, "step": 33920 }, { "epoch": 9.87198137910969, "grad_norm": 0.09962358325719833, "learning_rate": 1.2997854101420544e-05, "loss": 0.0152, "step": 33930 }, { "epoch": 9.874890893220833, "grad_norm": 0.08196268230676651, "learning_rate": 1.2989661033145607e-05, "loss": 0.0151, "step": 33940 }, { "epoch": 9.877800407331975, "grad_norm": 0.10013247281312943, "learning_rate": 1.298146857555857e-05, "loss": 0.0161, "step": 33950 }, { "epoch": 9.880709921443119, "grad_norm": 0.09801299124956131, "learning_rate": 1.2973276731148077e-05, "loss": 0.0156, "step": 33960 }, { "epoch": 9.883619435554262, "grad_norm": 0.1282014548778534, "learning_rate": 1.2965085502402607e-05, "loss": 0.0169, "step": 33970 }, { "epoch": 9.886528949665406, "grad_norm": 0.12919244170188904, "learning_rate": 1.2956894891810427e-05, "loss": 0.0169, "step": 33980 }, { "epoch": 9.88943846377655, "grad_norm": 0.08060052990913391, "learning_rate": 1.2948704901859641e-05, "loss": 0.0153, "step": 33990 }, { "epoch": 9.892347977887693, "grad_norm": 0.14466504752635956, "learning_rate": 1.2940515535038138e-05, "loss": 0.0148, "step": 34000 }, { "epoch": 9.895257491998835, "grad_norm": 0.0952676460146904, "learning_rate": 1.2932326793833638e-05, "loss": 0.0159, "step": 34010 }, { "epoch": 9.89816700610998, "grad_norm": 0.10485552996397018, "learning_rate": 1.292413868073367e-05, "loss": 0.0156, "step": 34020 }, { "epoch": 9.901076520221123, "grad_norm": 0.11441043019294739, "learning_rate": 1.2915951198225555e-05, "loss": 0.0155, "step": 34030 }, { "epoch": 9.903986034332267, "grad_norm": 0.12439761310815811, "learning_rate": 1.2907764348796447e-05, "loss": 0.0165, "step": 34040 }, { "epoch": 9.90689554844341, "grad_norm": 0.10323179513216019, "learning_rate": 1.2899578134933283e-05, "loss": 0.015, "step": 34050 }, { "epoch": 9.909805062554554, "grad_norm": 0.09257055819034576, "learning_rate": 1.2891392559122826e-05, "loss": 0.0177, "step": 34060 }, { "epoch": 9.912714576665696, "grad_norm": 0.1187189444899559, "learning_rate": 1.2883207623851629e-05, "loss": 0.0158, "step": 34070 }, { "epoch": 9.91562409077684, "grad_norm": 0.10886043310165405, "learning_rate": 1.2875023331606066e-05, "loss": 0.0142, "step": 34080 }, { "epoch": 9.918533604887983, "grad_norm": 0.19693371653556824, "learning_rate": 1.2866839684872313e-05, "loss": 0.0171, "step": 34090 }, { "epoch": 9.921443118999127, "grad_norm": 0.09205162525177002, "learning_rate": 1.285865668613634e-05, "loss": 0.017, "step": 34100 }, { "epoch": 9.92435263311027, "grad_norm": 0.10505953431129456, "learning_rate": 1.2850474337883923e-05, "loss": 0.0166, "step": 34110 }, { "epoch": 9.927262147221414, "grad_norm": 0.12160027027130127, "learning_rate": 1.2842292642600646e-05, "loss": 0.0154, "step": 34120 }, { "epoch": 9.930171661332558, "grad_norm": 0.07192245870828629, "learning_rate": 1.2834111602771896e-05, "loss": 0.0145, "step": 34130 }, { "epoch": 9.9330811754437, "grad_norm": 0.1014755442738533, "learning_rate": 1.2825931220882859e-05, "loss": 0.0147, "step": 34140 }, { "epoch": 9.935990689554844, "grad_norm": 0.12656651437282562, "learning_rate": 1.2817751499418507e-05, "loss": 0.0153, "step": 34150 }, { "epoch": 9.938900203665987, "grad_norm": 0.1067795380949974, "learning_rate": 1.2809572440863637e-05, "loss": 0.0169, "step": 34160 }, { "epoch": 9.941809717777131, "grad_norm": 0.0928855836391449, "learning_rate": 1.280139404770282e-05, "loss": 0.0166, "step": 34170 }, { "epoch": 9.944719231888275, "grad_norm": 0.0902293249964714, "learning_rate": 1.2793216322420439e-05, "loss": 0.0156, "step": 34180 }, { "epoch": 9.947628745999419, "grad_norm": 0.11537790298461914, "learning_rate": 1.278503926750068e-05, "loss": 0.0153, "step": 34190 }, { "epoch": 9.950538260110562, "grad_norm": 0.09744393825531006, "learning_rate": 1.27768628854275e-05, "loss": 0.0149, "step": 34200 }, { "epoch": 9.953447774221704, "grad_norm": 0.12423749268054962, "learning_rate": 1.2768687178684685e-05, "loss": 0.0147, "step": 34210 }, { "epoch": 9.956357288332848, "grad_norm": 0.11928030103445053, "learning_rate": 1.2760512149755783e-05, "loss": 0.0179, "step": 34220 }, { "epoch": 9.959266802443992, "grad_norm": 0.0996769592165947, "learning_rate": 1.275233780112416e-05, "loss": 0.0173, "step": 34230 }, { "epoch": 9.962176316555135, "grad_norm": 0.11459759622812271, "learning_rate": 1.2744164135272972e-05, "loss": 0.0148, "step": 34240 }, { "epoch": 9.965085830666279, "grad_norm": 0.0893368199467659, "learning_rate": 1.2735991154685154e-05, "loss": 0.0159, "step": 34250 }, { "epoch": 9.967995344777423, "grad_norm": 0.1105501651763916, "learning_rate": 1.2727818861843452e-05, "loss": 0.0151, "step": 34260 }, { "epoch": 9.970904858888566, "grad_norm": 0.10569296777248383, "learning_rate": 1.2719647259230377e-05, "loss": 0.0161, "step": 34270 }, { "epoch": 9.973814372999708, "grad_norm": 0.09742768853902817, "learning_rate": 1.271147634932826e-05, "loss": 0.0147, "step": 34280 }, { "epoch": 9.976723887110852, "grad_norm": 0.08313369750976562, "learning_rate": 1.2703306134619206e-05, "loss": 0.0139, "step": 34290 }, { "epoch": 9.979633401221996, "grad_norm": 0.10876228660345078, "learning_rate": 1.2695136617585103e-05, "loss": 0.0153, "step": 34300 }, { "epoch": 9.98254291533314, "grad_norm": 0.09303877502679825, "learning_rate": 1.2686967800707647e-05, "loss": 0.0169, "step": 34310 }, { "epoch": 9.985452429444283, "grad_norm": 0.092074453830719, "learning_rate": 1.2678799686468291e-05, "loss": 0.0148, "step": 34320 }, { "epoch": 9.988361943555427, "grad_norm": 0.13298399746418, "learning_rate": 1.2670632277348304e-05, "loss": 0.015, "step": 34330 }, { "epoch": 9.991271457666569, "grad_norm": 0.13734176754951477, "learning_rate": 1.2662465575828734e-05, "loss": 0.0162, "step": 34340 }, { "epoch": 9.994180971777713, "grad_norm": 0.12225615978240967, "learning_rate": 1.2654299584390398e-05, "loss": 0.0132, "step": 34350 }, { "epoch": 9.997090485888856, "grad_norm": 0.08960353583097458, "learning_rate": 1.2646134305513915e-05, "loss": 0.0151, "step": 34360 }, { "epoch": 10.0, "grad_norm": 0.14757779240608215, "learning_rate": 1.263796974167968e-05, "loss": 0.0161, "step": 34370 }, { "epoch": 10.002909514111144, "grad_norm": 0.1208169236779213, "learning_rate": 1.262980589536787e-05, "loss": 0.0149, "step": 34380 }, { "epoch": 10.005819028222287, "grad_norm": 0.08430039137601852, "learning_rate": 1.262164276905845e-05, "loss": 0.0154, "step": 34390 }, { "epoch": 10.008728542333431, "grad_norm": 0.1002969965338707, "learning_rate": 1.2613480365231156e-05, "loss": 0.0154, "step": 34400 }, { "epoch": 10.011638056444573, "grad_norm": 0.08896370977163315, "learning_rate": 1.260531868636552e-05, "loss": 0.0159, "step": 34410 }, { "epoch": 10.014547570555717, "grad_norm": 0.12590551376342773, "learning_rate": 1.2597157734940838e-05, "loss": 0.0168, "step": 34420 }, { "epoch": 10.01745708466686, "grad_norm": 0.17466114461421967, "learning_rate": 1.2588997513436188e-05, "loss": 0.0152, "step": 34430 }, { "epoch": 10.020366598778004, "grad_norm": 0.11771948635578156, "learning_rate": 1.2580838024330441e-05, "loss": 0.0154, "step": 34440 }, { "epoch": 10.023276112889148, "grad_norm": 0.09655945003032684, "learning_rate": 1.2572679270102224e-05, "loss": 0.0152, "step": 34450 }, { "epoch": 10.026185627000292, "grad_norm": 0.0991717129945755, "learning_rate": 1.256452125322996e-05, "loss": 0.0161, "step": 34460 }, { "epoch": 10.029095141111434, "grad_norm": 0.10942744463682175, "learning_rate": 1.2556363976191827e-05, "loss": 0.0154, "step": 34470 }, { "epoch": 10.032004655222577, "grad_norm": 0.07649759203195572, "learning_rate": 1.2548207441465805e-05, "loss": 0.0162, "step": 34480 }, { "epoch": 10.034914169333721, "grad_norm": 0.10878745466470718, "learning_rate": 1.254005165152962e-05, "loss": 0.0149, "step": 34490 }, { "epoch": 10.037823683444865, "grad_norm": 0.06751321256160736, "learning_rate": 1.2531896608860793e-05, "loss": 0.0163, "step": 34500 }, { "epoch": 10.040733197556008, "grad_norm": 0.09665802866220474, "learning_rate": 1.2523742315936611e-05, "loss": 0.0158, "step": 34510 }, { "epoch": 10.043642711667152, "grad_norm": 0.08707582950592041, "learning_rate": 1.2515588775234133e-05, "loss": 0.0149, "step": 34520 }, { "epoch": 10.046552225778296, "grad_norm": 0.08871655911207199, "learning_rate": 1.2507435989230188e-05, "loss": 0.0167, "step": 34530 }, { "epoch": 10.049461739889438, "grad_norm": 0.09902302175760269, "learning_rate": 1.2499283960401373e-05, "loss": 0.0163, "step": 34540 }, { "epoch": 10.052371254000581, "grad_norm": 0.08293027430772781, "learning_rate": 1.2491132691224063e-05, "loss": 0.0137, "step": 34550 }, { "epoch": 10.055280768111725, "grad_norm": 0.113865427672863, "learning_rate": 1.2482982184174407e-05, "loss": 0.0159, "step": 34560 }, { "epoch": 10.058190282222869, "grad_norm": 0.12404094636440277, "learning_rate": 1.2474832441728297e-05, "loss": 0.0157, "step": 34570 }, { "epoch": 10.061099796334013, "grad_norm": 0.08664334565401077, "learning_rate": 1.2466683466361427e-05, "loss": 0.016, "step": 34580 }, { "epoch": 10.064009310445156, "grad_norm": 0.09562128782272339, "learning_rate": 1.245853526054922e-05, "loss": 0.0176, "step": 34590 }, { "epoch": 10.0669188245563, "grad_norm": 0.10057607293128967, "learning_rate": 1.2450387826766903e-05, "loss": 0.0161, "step": 34600 }, { "epoch": 10.069828338667442, "grad_norm": 0.07667047530412674, "learning_rate": 1.2442241167489449e-05, "loss": 0.0151, "step": 34610 }, { "epoch": 10.072737852778586, "grad_norm": 0.08184117823839188, "learning_rate": 1.243409528519159e-05, "loss": 0.0161, "step": 34620 }, { "epoch": 10.07564736688973, "grad_norm": 0.083738312125206, "learning_rate": 1.242595018234784e-05, "loss": 0.0157, "step": 34630 }, { "epoch": 10.078556881000873, "grad_norm": 0.10139529407024384, "learning_rate": 1.2417805861432453e-05, "loss": 0.0145, "step": 34640 }, { "epoch": 10.081466395112017, "grad_norm": 0.12752458453178406, "learning_rate": 1.2409662324919469e-05, "loss": 0.0159, "step": 34650 }, { "epoch": 10.08437590922316, "grad_norm": 0.09586697816848755, "learning_rate": 1.2401519575282679e-05, "loss": 0.0142, "step": 34660 }, { "epoch": 10.087285423334302, "grad_norm": 0.09728216379880905, "learning_rate": 1.2393377614995631e-05, "loss": 0.0147, "step": 34670 }, { "epoch": 10.090194937445446, "grad_norm": 0.10366188734769821, "learning_rate": 1.2385236446531643e-05, "loss": 0.0169, "step": 34680 }, { "epoch": 10.09310445155659, "grad_norm": 0.10635070502758026, "learning_rate": 1.2377096072363778e-05, "loss": 0.0144, "step": 34690 }, { "epoch": 10.096013965667733, "grad_norm": 0.20137330889701843, "learning_rate": 1.2368956494964873e-05, "loss": 0.017, "step": 34700 }, { "epoch": 10.098923479778877, "grad_norm": 0.09880179166793823, "learning_rate": 1.2360817716807524e-05, "loss": 0.0156, "step": 34710 }, { "epoch": 10.10183299389002, "grad_norm": 0.136089026927948, "learning_rate": 1.235267974036406e-05, "loss": 0.0161, "step": 34720 }, { "epoch": 10.104742508001165, "grad_norm": 0.08691886067390442, "learning_rate": 1.2344542568106602e-05, "loss": 0.0169, "step": 34730 }, { "epoch": 10.107652022112307, "grad_norm": 0.09398937225341797, "learning_rate": 1.2336406202506991e-05, "loss": 0.0152, "step": 34740 }, { "epoch": 10.11056153622345, "grad_norm": 0.10491388291120529, "learning_rate": 1.2328270646036847e-05, "loss": 0.0159, "step": 34750 }, { "epoch": 10.113471050334594, "grad_norm": 0.11541780829429626, "learning_rate": 1.2320135901167543e-05, "loss": 0.0159, "step": 34760 }, { "epoch": 10.116380564445738, "grad_norm": 0.10831478983163834, "learning_rate": 1.231200197037019e-05, "loss": 0.0153, "step": 34770 }, { "epoch": 10.119290078556881, "grad_norm": 0.1119464859366417, "learning_rate": 1.2303868856115671e-05, "loss": 0.0161, "step": 34780 }, { "epoch": 10.122199592668025, "grad_norm": 0.11420103907585144, "learning_rate": 1.2295736560874603e-05, "loss": 0.0165, "step": 34790 }, { "epoch": 10.125109106779167, "grad_norm": 0.09391137212514877, "learning_rate": 1.2287605087117362e-05, "loss": 0.0154, "step": 34800 }, { "epoch": 10.12801862089031, "grad_norm": 0.07968925684690475, "learning_rate": 1.2279474437314083e-05, "loss": 0.0155, "step": 34810 }, { "epoch": 10.130928135001454, "grad_norm": 0.09956901520490646, "learning_rate": 1.227134461393464e-05, "loss": 0.0159, "step": 34820 }, { "epoch": 10.133837649112598, "grad_norm": 0.09373705834150314, "learning_rate": 1.2263215619448658e-05, "loss": 0.0154, "step": 34830 }, { "epoch": 10.136747163223742, "grad_norm": 0.09158887714147568, "learning_rate": 1.2255087456325504e-05, "loss": 0.0141, "step": 34840 }, { "epoch": 10.139656677334886, "grad_norm": 0.08239975571632385, "learning_rate": 1.2246960127034314e-05, "loss": 0.0145, "step": 34850 }, { "epoch": 10.14256619144603, "grad_norm": 0.16433411836624146, "learning_rate": 1.2238833634043942e-05, "loss": 0.0146, "step": 34860 }, { "epoch": 10.145475705557171, "grad_norm": 0.09598816186189651, "learning_rate": 1.2230707979823006e-05, "loss": 0.0158, "step": 34870 }, { "epoch": 10.148385219668315, "grad_norm": 0.12415501475334167, "learning_rate": 1.2222583166839874e-05, "loss": 0.018, "step": 34880 }, { "epoch": 10.151294733779459, "grad_norm": 0.13987956941127777, "learning_rate": 1.2214459197562635e-05, "loss": 0.0153, "step": 34890 }, { "epoch": 10.154204247890602, "grad_norm": 0.07477002590894699, "learning_rate": 1.220633607445915e-05, "loss": 0.0134, "step": 34900 }, { "epoch": 10.157113762001746, "grad_norm": 0.10524139553308487, "learning_rate": 1.2198213799997e-05, "loss": 0.0157, "step": 34910 }, { "epoch": 10.16002327611289, "grad_norm": 0.09182346612215042, "learning_rate": 1.2190092376643519e-05, "loss": 0.0161, "step": 34920 }, { "epoch": 10.162932790224033, "grad_norm": 0.07552211731672287, "learning_rate": 1.2181971806865786e-05, "loss": 0.0164, "step": 34930 }, { "epoch": 10.165842304335175, "grad_norm": 0.08606097102165222, "learning_rate": 1.217385209313061e-05, "loss": 0.015, "step": 34940 }, { "epoch": 10.168751818446319, "grad_norm": 0.09699784964323044, "learning_rate": 1.2165733237904548e-05, "loss": 0.0164, "step": 34950 }, { "epoch": 10.171661332557463, "grad_norm": 0.10827870666980743, "learning_rate": 1.2157615243653894e-05, "loss": 0.0155, "step": 34960 }, { "epoch": 10.174570846668606, "grad_norm": 0.10360084474086761, "learning_rate": 1.214949811284468e-05, "loss": 0.0161, "step": 34970 }, { "epoch": 10.17748036077975, "grad_norm": 0.07909736782312393, "learning_rate": 1.2141381847942678e-05, "loss": 0.0143, "step": 34980 }, { "epoch": 10.180389874890894, "grad_norm": 0.1106911227107048, "learning_rate": 1.213326645141339e-05, "loss": 0.0157, "step": 34990 }, { "epoch": 10.183299389002036, "grad_norm": 0.11404693126678467, "learning_rate": 1.2125151925722069e-05, "loss": 0.0162, "step": 35000 }, { "epoch": 10.18620890311318, "grad_norm": 0.16025769710540771, "learning_rate": 1.2117038273333678e-05, "loss": 0.0151, "step": 35010 }, { "epoch": 10.189118417224323, "grad_norm": 0.08144821226596832, "learning_rate": 1.2108925496712943e-05, "loss": 0.0145, "step": 35020 }, { "epoch": 10.192027931335467, "grad_norm": 0.08869004994630814, "learning_rate": 1.2100813598324311e-05, "loss": 0.0147, "step": 35030 }, { "epoch": 10.19493744544661, "grad_norm": 0.11569684743881226, "learning_rate": 1.2092702580631955e-05, "loss": 0.0154, "step": 35040 }, { "epoch": 10.197846959557754, "grad_norm": 0.09455376863479614, "learning_rate": 1.20845924460998e-05, "loss": 0.0153, "step": 35050 }, { "epoch": 10.200756473668898, "grad_norm": 0.08442956954240799, "learning_rate": 1.2076483197191473e-05, "loss": 0.0152, "step": 35060 }, { "epoch": 10.20366598778004, "grad_norm": 0.08609243482351303, "learning_rate": 1.2068374836370361e-05, "loss": 0.0135, "step": 35070 }, { "epoch": 10.206575501891184, "grad_norm": 0.08675917983055115, "learning_rate": 1.2060267366099576e-05, "loss": 0.0166, "step": 35080 }, { "epoch": 10.209485016002327, "grad_norm": 0.09467640519142151, "learning_rate": 1.205216078884194e-05, "loss": 0.0152, "step": 35090 }, { "epoch": 10.212394530113471, "grad_norm": 0.0949053093791008, "learning_rate": 1.2044055107060025e-05, "loss": 0.0144, "step": 35100 }, { "epoch": 10.215304044224615, "grad_norm": 0.11596895754337311, "learning_rate": 1.2035950323216117e-05, "loss": 0.0157, "step": 35110 }, { "epoch": 10.218213558335759, "grad_norm": 0.12141550332307816, "learning_rate": 1.2027846439772245e-05, "loss": 0.0153, "step": 35120 }, { "epoch": 10.221123072446902, "grad_norm": 0.0939219519495964, "learning_rate": 1.2019743459190148e-05, "loss": 0.0161, "step": 35130 }, { "epoch": 10.224032586558044, "grad_norm": 0.09886018931865692, "learning_rate": 1.2011641383931296e-05, "loss": 0.015, "step": 35140 }, { "epoch": 10.226942100669188, "grad_norm": 0.11648042500019073, "learning_rate": 1.2003540216456895e-05, "loss": 0.0164, "step": 35150 }, { "epoch": 10.229851614780332, "grad_norm": 0.08748338371515274, "learning_rate": 1.1995439959227854e-05, "loss": 0.0145, "step": 35160 }, { "epoch": 10.232761128891475, "grad_norm": 0.1104946956038475, "learning_rate": 1.1987340614704823e-05, "loss": 0.016, "step": 35170 }, { "epoch": 10.235670643002619, "grad_norm": 0.08077836781740189, "learning_rate": 1.1979242185348173e-05, "loss": 0.0142, "step": 35180 }, { "epoch": 10.238580157113763, "grad_norm": 0.07871329039335251, "learning_rate": 1.1971144673617983e-05, "loss": 0.0152, "step": 35190 }, { "epoch": 10.241489671224905, "grad_norm": 0.09163115918636322, "learning_rate": 1.1963048081974075e-05, "loss": 0.0155, "step": 35200 }, { "epoch": 10.244399185336048, "grad_norm": 0.09416773170232773, "learning_rate": 1.1954952412875968e-05, "loss": 0.0154, "step": 35210 }, { "epoch": 10.247308699447192, "grad_norm": 0.11236676573753357, "learning_rate": 1.1946857668782919e-05, "loss": 0.0144, "step": 35220 }, { "epoch": 10.250218213558336, "grad_norm": 0.07436349987983704, "learning_rate": 1.19387638521539e-05, "loss": 0.016, "step": 35230 }, { "epoch": 10.25312772766948, "grad_norm": 0.09788785129785538, "learning_rate": 1.1930670965447591e-05, "loss": 0.0162, "step": 35240 }, { "epoch": 10.256037241780623, "grad_norm": 0.09364727139472961, "learning_rate": 1.1922579011122404e-05, "loss": 0.0147, "step": 35250 }, { "epoch": 10.258946755891767, "grad_norm": 0.09167074412107468, "learning_rate": 1.1914487991636457e-05, "loss": 0.0145, "step": 35260 }, { "epoch": 10.261856270002909, "grad_norm": 0.12550802528858185, "learning_rate": 1.1906397909447597e-05, "loss": 0.0146, "step": 35270 }, { "epoch": 10.264765784114052, "grad_norm": 0.10150057822465897, "learning_rate": 1.1898308767013361e-05, "loss": 0.0163, "step": 35280 }, { "epoch": 10.267675298225196, "grad_norm": 0.08853165805339813, "learning_rate": 1.1890220566791028e-05, "loss": 0.0159, "step": 35290 }, { "epoch": 10.27058481233634, "grad_norm": 0.12165136635303497, "learning_rate": 1.1882133311237584e-05, "loss": 0.0162, "step": 35300 }, { "epoch": 10.273494326447484, "grad_norm": 0.09263758361339569, "learning_rate": 1.187404700280971e-05, "loss": 0.0157, "step": 35310 }, { "epoch": 10.276403840558627, "grad_norm": 0.07818451523780823, "learning_rate": 1.1865961643963828e-05, "loss": 0.0149, "step": 35320 }, { "epoch": 10.279313354669771, "grad_norm": 0.11976674199104309, "learning_rate": 1.1857877237156043e-05, "loss": 0.014, "step": 35330 }, { "epoch": 10.282222868780913, "grad_norm": 0.09423913806676865, "learning_rate": 1.1849793784842191e-05, "loss": 0.0164, "step": 35340 }, { "epoch": 10.285132382892057, "grad_norm": 0.10256023705005646, "learning_rate": 1.1841711289477815e-05, "loss": 0.0161, "step": 35350 }, { "epoch": 10.2880418970032, "grad_norm": 0.08567488938570023, "learning_rate": 1.1833629753518156e-05, "loss": 0.0141, "step": 35360 }, { "epoch": 10.290951411114344, "grad_norm": 0.06838906556367874, "learning_rate": 1.1825549179418176e-05, "loss": 0.016, "step": 35370 }, { "epoch": 10.293860925225488, "grad_norm": 0.09711728245019913, "learning_rate": 1.1817469569632539e-05, "loss": 0.0152, "step": 35380 }, { "epoch": 10.296770439336631, "grad_norm": 0.08849100023508072, "learning_rate": 1.1809390926615615e-05, "loss": 0.0145, "step": 35390 }, { "epoch": 10.299679953447773, "grad_norm": 0.11297854036092758, "learning_rate": 1.1801313252821485e-05, "loss": 0.0155, "step": 35400 }, { "epoch": 10.302589467558917, "grad_norm": 0.1334563046693802, "learning_rate": 1.179323655070393e-05, "loss": 0.0165, "step": 35410 }, { "epoch": 10.30549898167006, "grad_norm": 0.0998210608959198, "learning_rate": 1.1785160822716448e-05, "loss": 0.0144, "step": 35420 }, { "epoch": 10.308408495781205, "grad_norm": 0.08092465996742249, "learning_rate": 1.177708607131222e-05, "loss": 0.0155, "step": 35430 }, { "epoch": 10.311318009892348, "grad_norm": 0.10362548381090164, "learning_rate": 1.1769012298944146e-05, "loss": 0.0153, "step": 35440 }, { "epoch": 10.314227524003492, "grad_norm": 0.09302632510662079, "learning_rate": 1.1760939508064834e-05, "loss": 0.0147, "step": 35450 }, { "epoch": 10.317137038114636, "grad_norm": 0.10669826716184616, "learning_rate": 1.1752867701126569e-05, "loss": 0.0146, "step": 35460 }, { "epoch": 10.320046552225778, "grad_norm": 0.10647069662809372, "learning_rate": 1.1744796880581368e-05, "loss": 0.0159, "step": 35470 }, { "epoch": 10.322956066336921, "grad_norm": 0.09637600928544998, "learning_rate": 1.1736727048880917e-05, "loss": 0.0147, "step": 35480 }, { "epoch": 10.325865580448065, "grad_norm": 0.10294894129037857, "learning_rate": 1.172865820847663e-05, "loss": 0.0147, "step": 35490 }, { "epoch": 10.328775094559209, "grad_norm": 0.09562772512435913, "learning_rate": 1.172059036181961e-05, "loss": 0.0158, "step": 35500 }, { "epoch": 10.331684608670352, "grad_norm": 0.10532674938440323, "learning_rate": 1.1712523511360644e-05, "loss": 0.016, "step": 35510 }, { "epoch": 10.334594122781496, "grad_norm": 0.10284417867660522, "learning_rate": 1.170445765955024e-05, "loss": 0.015, "step": 35520 }, { "epoch": 10.337503636892638, "grad_norm": 0.0837864950299263, "learning_rate": 1.169639280883858e-05, "loss": 0.0143, "step": 35530 }, { "epoch": 10.340413151003782, "grad_norm": 0.11155769228935242, "learning_rate": 1.1688328961675558e-05, "loss": 0.0159, "step": 35540 }, { "epoch": 10.343322665114925, "grad_norm": 0.08966422826051712, "learning_rate": 1.1680266120510763e-05, "loss": 0.0156, "step": 35550 }, { "epoch": 10.34623217922607, "grad_norm": 0.10138235986232758, "learning_rate": 1.1672204287793461e-05, "loss": 0.0155, "step": 35560 }, { "epoch": 10.349141693337213, "grad_norm": 0.10875502228736877, "learning_rate": 1.166414346597264e-05, "loss": 0.0152, "step": 35570 }, { "epoch": 10.352051207448357, "grad_norm": 0.1120242029428482, "learning_rate": 1.1656083657496949e-05, "loss": 0.015, "step": 35580 }, { "epoch": 10.3549607215595, "grad_norm": 0.08817871659994125, "learning_rate": 1.164802486481475e-05, "loss": 0.0139, "step": 35590 }, { "epoch": 10.357870235670642, "grad_norm": 0.08604110032320023, "learning_rate": 1.16399670903741e-05, "loss": 0.0166, "step": 35600 }, { "epoch": 10.360779749781786, "grad_norm": 0.09564799070358276, "learning_rate": 1.1631910336622728e-05, "loss": 0.0145, "step": 35610 }, { "epoch": 10.36368926389293, "grad_norm": 0.2588071823120117, "learning_rate": 1.1623854606008072e-05, "loss": 0.016, "step": 35620 }, { "epoch": 10.366598778004073, "grad_norm": 0.09795548021793365, "learning_rate": 1.1615799900977235e-05, "loss": 0.0151, "step": 35630 }, { "epoch": 10.369508292115217, "grad_norm": 0.09771189093589783, "learning_rate": 1.1607746223977036e-05, "loss": 0.015, "step": 35640 }, { "epoch": 10.37241780622636, "grad_norm": 0.09339005500078201, "learning_rate": 1.1599693577453974e-05, "loss": 0.014, "step": 35650 }, { "epoch": 10.375327320337504, "grad_norm": 0.10958296805620193, "learning_rate": 1.1591641963854217e-05, "loss": 0.0148, "step": 35660 }, { "epoch": 10.378236834448646, "grad_norm": 0.0821002721786499, "learning_rate": 1.158359138562364e-05, "loss": 0.0138, "step": 35670 }, { "epoch": 10.38114634855979, "grad_norm": 0.12050667405128479, "learning_rate": 1.1575541845207796e-05, "loss": 0.0156, "step": 35680 }, { "epoch": 10.384055862670934, "grad_norm": 0.14755946397781372, "learning_rate": 1.1567493345051922e-05, "loss": 0.0156, "step": 35690 }, { "epoch": 10.386965376782078, "grad_norm": 0.12763963639736176, "learning_rate": 1.1559445887600936e-05, "loss": 0.0154, "step": 35700 }, { "epoch": 10.389874890893221, "grad_norm": 0.08766680955886841, "learning_rate": 1.1551399475299446e-05, "loss": 0.0157, "step": 35710 }, { "epoch": 10.392784405004365, "grad_norm": 0.10374492406845093, "learning_rate": 1.1543354110591748e-05, "loss": 0.0164, "step": 35720 }, { "epoch": 10.395693919115507, "grad_norm": 0.10568471252918243, "learning_rate": 1.1535309795921798e-05, "loss": 0.0146, "step": 35730 }, { "epoch": 10.39860343322665, "grad_norm": 0.10017837584018707, "learning_rate": 1.1527266533733259e-05, "loss": 0.0155, "step": 35740 }, { "epoch": 10.401512947337794, "grad_norm": 0.12239962816238403, "learning_rate": 1.1519224326469448e-05, "loss": 0.0162, "step": 35750 }, { "epoch": 10.404422461448938, "grad_norm": 0.1089867427945137, "learning_rate": 1.151118317657338e-05, "loss": 0.0139, "step": 35760 }, { "epoch": 10.407331975560082, "grad_norm": 0.08851274847984314, "learning_rate": 1.1503143086487754e-05, "loss": 0.0146, "step": 35770 }, { "epoch": 10.410241489671225, "grad_norm": 0.08922601491212845, "learning_rate": 1.1495104058654923e-05, "loss": 0.0152, "step": 35780 }, { "epoch": 10.41315100378237, "grad_norm": 0.09337562322616577, "learning_rate": 1.1487066095516943e-05, "loss": 0.0138, "step": 35790 }, { "epoch": 10.416060517893511, "grad_norm": 0.10565630346536636, "learning_rate": 1.1479029199515523e-05, "loss": 0.0156, "step": 35800 }, { "epoch": 10.418970032004655, "grad_norm": 0.10536327958106995, "learning_rate": 1.1470993373092067e-05, "loss": 0.0137, "step": 35810 }, { "epoch": 10.421879546115798, "grad_norm": 0.11189640313386917, "learning_rate": 1.146295861868765e-05, "loss": 0.0132, "step": 35820 }, { "epoch": 10.424789060226942, "grad_norm": 0.09211520850658417, "learning_rate": 1.1454924938743014e-05, "loss": 0.0142, "step": 35830 }, { "epoch": 10.427698574338086, "grad_norm": 0.08889251202344894, "learning_rate": 1.144689233569858e-05, "loss": 0.0145, "step": 35840 }, { "epoch": 10.43060808844923, "grad_norm": 0.09009508043527603, "learning_rate": 1.1438860811994435e-05, "loss": 0.0148, "step": 35850 }, { "epoch": 10.433517602560372, "grad_norm": 0.0979972630739212, "learning_rate": 1.1430830370070351e-05, "loss": 0.0156, "step": 35860 }, { "epoch": 10.436427116671515, "grad_norm": 0.12512478232383728, "learning_rate": 1.1422801012365767e-05, "loss": 0.0157, "step": 35870 }, { "epoch": 10.439336630782659, "grad_norm": 0.12028149515390396, "learning_rate": 1.141477274131978e-05, "loss": 0.0139, "step": 35880 }, { "epoch": 10.442246144893803, "grad_norm": 0.08263573050498962, "learning_rate": 1.1406745559371177e-05, "loss": 0.0142, "step": 35890 }, { "epoch": 10.445155659004946, "grad_norm": 0.10141216963529587, "learning_rate": 1.139871946895839e-05, "loss": 0.0153, "step": 35900 }, { "epoch": 10.44806517311609, "grad_norm": 0.14719435572624207, "learning_rate": 1.1390694472519547e-05, "loss": 0.015, "step": 35910 }, { "epoch": 10.450974687227234, "grad_norm": 0.09647219628095627, "learning_rate": 1.1382670572492428e-05, "loss": 0.015, "step": 35920 }, { "epoch": 10.453884201338376, "grad_norm": 0.11242697387933731, "learning_rate": 1.1374647771314473e-05, "loss": 0.0164, "step": 35930 }, { "epoch": 10.45679371544952, "grad_norm": 0.08565375208854675, "learning_rate": 1.1366626071422811e-05, "loss": 0.0148, "step": 35940 }, { "epoch": 10.459703229560663, "grad_norm": 0.06827718019485474, "learning_rate": 1.1358605475254207e-05, "loss": 0.0154, "step": 35950 }, { "epoch": 10.462612743671807, "grad_norm": 0.08952511847019196, "learning_rate": 1.1350585985245116e-05, "loss": 0.0131, "step": 35960 }, { "epoch": 10.46552225778295, "grad_norm": 0.07509234547615051, "learning_rate": 1.1342567603831647e-05, "loss": 0.0138, "step": 35970 }, { "epoch": 10.468431771894094, "grad_norm": 0.10649361461400986, "learning_rate": 1.1334550333449568e-05, "loss": 0.0146, "step": 35980 }, { "epoch": 10.471341286005238, "grad_norm": 0.1295204758644104, "learning_rate": 1.1326534176534325e-05, "loss": 0.0164, "step": 35990 }, { "epoch": 10.47425080011638, "grad_norm": 0.08855697512626648, "learning_rate": 1.1318519135520998e-05, "loss": 0.0148, "step": 36000 }, { "epoch": 10.477160314227524, "grad_norm": 0.10287550836801529, "learning_rate": 1.1310505212844355e-05, "loss": 0.0151, "step": 36010 }, { "epoch": 10.480069828338667, "grad_norm": 0.2061672955751419, "learning_rate": 1.130249241093882e-05, "loss": 0.0167, "step": 36020 }, { "epoch": 10.482979342449811, "grad_norm": 0.09035919606685638, "learning_rate": 1.1294480732238455e-05, "loss": 0.0137, "step": 36030 }, { "epoch": 10.485888856560955, "grad_norm": 0.10960910469293594, "learning_rate": 1.1286470179177014e-05, "loss": 0.0157, "step": 36040 }, { "epoch": 10.488798370672098, "grad_norm": 0.10113932937383652, "learning_rate": 1.1278460754187875e-05, "loss": 0.0158, "step": 36050 }, { "epoch": 10.49170788478324, "grad_norm": 0.07951643317937851, "learning_rate": 1.1270452459704098e-05, "loss": 0.0147, "step": 36060 }, { "epoch": 10.494617398894384, "grad_norm": 0.08954338729381561, "learning_rate": 1.1262445298158397e-05, "loss": 0.0154, "step": 36070 }, { "epoch": 10.497526913005528, "grad_norm": 0.09609285742044449, "learning_rate": 1.1254439271983122e-05, "loss": 0.0142, "step": 36080 }, { "epoch": 10.500436427116671, "grad_norm": 0.11744078248739243, "learning_rate": 1.1246434383610305e-05, "loss": 0.0148, "step": 36090 }, { "epoch": 10.503345941227815, "grad_norm": 0.1415107399225235, "learning_rate": 1.1238430635471615e-05, "loss": 0.0134, "step": 36100 }, { "epoch": 10.506255455338959, "grad_norm": 0.08252239227294922, "learning_rate": 1.1230428029998378e-05, "loss": 0.0163, "step": 36110 }, { "epoch": 10.509164969450103, "grad_norm": 0.10660441219806671, "learning_rate": 1.1222426569621574e-05, "loss": 0.0148, "step": 36120 }, { "epoch": 10.512074483561245, "grad_norm": 0.09621665626764297, "learning_rate": 1.1214426256771838e-05, "loss": 0.0171, "step": 36130 }, { "epoch": 10.514983997672388, "grad_norm": 0.08960185199975967, "learning_rate": 1.1206427093879457e-05, "loss": 0.0157, "step": 36140 }, { "epoch": 10.517893511783532, "grad_norm": 0.10523191094398499, "learning_rate": 1.1198429083374356e-05, "loss": 0.0146, "step": 36150 }, { "epoch": 10.520803025894676, "grad_norm": 0.10870489478111267, "learning_rate": 1.119043222768613e-05, "loss": 0.0159, "step": 36160 }, { "epoch": 10.52371254000582, "grad_norm": 0.07768254727125168, "learning_rate": 1.1182436529244004e-05, "loss": 0.0155, "step": 36170 }, { "epoch": 10.526622054116963, "grad_norm": 0.07567271590232849, "learning_rate": 1.117444199047686e-05, "loss": 0.0135, "step": 36180 }, { "epoch": 10.529531568228105, "grad_norm": 0.07565609365701675, "learning_rate": 1.116644861381324e-05, "loss": 0.0137, "step": 36190 }, { "epoch": 10.532441082339249, "grad_norm": 0.10845106840133667, "learning_rate": 1.1158456401681306e-05, "loss": 0.0149, "step": 36200 }, { "epoch": 10.535350596450392, "grad_norm": 0.08554939925670624, "learning_rate": 1.1150465356508892e-05, "loss": 0.0159, "step": 36210 }, { "epoch": 10.538260110561536, "grad_norm": 0.09088730812072754, "learning_rate": 1.1142475480723456e-05, "loss": 0.0159, "step": 36220 }, { "epoch": 10.54116962467268, "grad_norm": 0.11690647155046463, "learning_rate": 1.1134486776752116e-05, "loss": 0.016, "step": 36230 }, { "epoch": 10.544079138783824, "grad_norm": 0.08450573682785034, "learning_rate": 1.1126499247021635e-05, "loss": 0.0157, "step": 36240 }, { "epoch": 10.546988652894967, "grad_norm": 0.09295015037059784, "learning_rate": 1.111851289395841e-05, "loss": 0.0134, "step": 36250 }, { "epoch": 10.54989816700611, "grad_norm": 0.13299918174743652, "learning_rate": 1.111052771998848e-05, "loss": 0.0151, "step": 36260 }, { "epoch": 10.552807681117253, "grad_norm": 0.12077877670526505, "learning_rate": 1.110254372753753e-05, "loss": 0.0152, "step": 36270 }, { "epoch": 10.555717195228397, "grad_norm": 0.09214090555906296, "learning_rate": 1.1094560919030891e-05, "loss": 0.0145, "step": 36280 }, { "epoch": 10.55862670933954, "grad_norm": 0.09252152591943741, "learning_rate": 1.1086579296893533e-05, "loss": 0.0147, "step": 36290 }, { "epoch": 10.561536223450684, "grad_norm": 0.09547659009695053, "learning_rate": 1.1078598863550051e-05, "loss": 0.015, "step": 36300 }, { "epoch": 10.564445737561828, "grad_norm": 0.09344267100095749, "learning_rate": 1.1070619621424703e-05, "loss": 0.0157, "step": 36310 }, { "epoch": 10.567355251672971, "grad_norm": 0.08558107912540436, "learning_rate": 1.1062641572941355e-05, "loss": 0.0142, "step": 36320 }, { "epoch": 10.570264765784113, "grad_norm": 0.12179996818304062, "learning_rate": 1.105466472052354e-05, "loss": 0.017, "step": 36330 }, { "epoch": 10.573174279895257, "grad_norm": 0.10081776231527328, "learning_rate": 1.1046689066594418e-05, "loss": 0.0167, "step": 36340 }, { "epoch": 10.5760837940064, "grad_norm": 0.09950615465641022, "learning_rate": 1.103871461357677e-05, "loss": 0.0141, "step": 36350 }, { "epoch": 10.578993308117544, "grad_norm": 0.09534846991300583, "learning_rate": 1.1030741363893036e-05, "loss": 0.0132, "step": 36360 }, { "epoch": 10.581902822228688, "grad_norm": 0.10275480896234512, "learning_rate": 1.102276931996527e-05, "loss": 0.0147, "step": 36370 }, { "epoch": 10.584812336339832, "grad_norm": 0.11043201386928558, "learning_rate": 1.1014798484215171e-05, "loss": 0.0145, "step": 36380 }, { "epoch": 10.587721850450976, "grad_norm": 0.09707465767860413, "learning_rate": 1.100682885906407e-05, "loss": 0.0139, "step": 36390 }, { "epoch": 10.590631364562118, "grad_norm": 0.14055053889751434, "learning_rate": 1.099886044693293e-05, "loss": 0.016, "step": 36400 }, { "epoch": 10.593540878673261, "grad_norm": 0.1157311350107193, "learning_rate": 1.0990893250242343e-05, "loss": 0.0148, "step": 36410 }, { "epoch": 10.596450392784405, "grad_norm": 0.1164625734090805, "learning_rate": 1.0982927271412528e-05, "loss": 0.0137, "step": 36420 }, { "epoch": 10.599359906895549, "grad_norm": 0.08050373196601868, "learning_rate": 1.0974962512863342e-05, "loss": 0.0136, "step": 36430 }, { "epoch": 10.602269421006692, "grad_norm": 0.07727055251598358, "learning_rate": 1.096699897701428e-05, "loss": 0.0144, "step": 36440 }, { "epoch": 10.605178935117836, "grad_norm": 0.1681963950395584, "learning_rate": 1.095903666628443e-05, "loss": 0.0144, "step": 36450 }, { "epoch": 10.608088449228978, "grad_norm": 0.09375076740980148, "learning_rate": 1.0951075583092559e-05, "loss": 0.0151, "step": 36460 }, { "epoch": 10.610997963340122, "grad_norm": 0.1879110336303711, "learning_rate": 1.094311572985701e-05, "loss": 0.0159, "step": 36470 }, { "epoch": 10.613907477451265, "grad_norm": 0.09749863296747208, "learning_rate": 1.0935157108995795e-05, "loss": 0.0145, "step": 36480 }, { "epoch": 10.616816991562409, "grad_norm": 0.09685482829809189, "learning_rate": 1.0927199722926517e-05, "loss": 0.0163, "step": 36490 }, { "epoch": 10.619726505673553, "grad_norm": 0.1407281905412674, "learning_rate": 1.0919243574066427e-05, "loss": 0.0153, "step": 36500 }, { "epoch": 10.622636019784697, "grad_norm": 0.11934120953083038, "learning_rate": 1.0911288664832398e-05, "loss": 0.0152, "step": 36510 }, { "epoch": 10.62554553389584, "grad_norm": 0.14438070356845856, "learning_rate": 1.0903334997640912e-05, "loss": 0.0141, "step": 36520 }, { "epoch": 10.628455048006982, "grad_norm": 0.14536763727664948, "learning_rate": 1.089538257490809e-05, "loss": 0.0156, "step": 36530 }, { "epoch": 10.631364562118126, "grad_norm": 0.10027530044317245, "learning_rate": 1.0887431399049663e-05, "loss": 0.0154, "step": 36540 }, { "epoch": 10.63427407622927, "grad_norm": 0.09925452619791031, "learning_rate": 1.0879481472480991e-05, "loss": 0.0146, "step": 36550 }, { "epoch": 10.637183590340413, "grad_norm": 0.09951187670230865, "learning_rate": 1.0871532797617055e-05, "loss": 0.0147, "step": 36560 }, { "epoch": 10.640093104451557, "grad_norm": 0.14122872054576874, "learning_rate": 1.0863585376872446e-05, "loss": 0.018, "step": 36570 }, { "epoch": 10.6430026185627, "grad_norm": 0.09553996473550797, "learning_rate": 1.0855639212661389e-05, "loss": 0.015, "step": 36580 }, { "epoch": 10.645912132673843, "grad_norm": 0.0943167433142662, "learning_rate": 1.084769430739771e-05, "loss": 0.0162, "step": 36590 }, { "epoch": 10.648821646784986, "grad_norm": 0.0930362418293953, "learning_rate": 1.0839750663494862e-05, "loss": 0.0145, "step": 36600 }, { "epoch": 10.65173116089613, "grad_norm": 0.15825410187244415, "learning_rate": 1.0831808283365925e-05, "loss": 0.0162, "step": 36610 }, { "epoch": 10.654640675007274, "grad_norm": 0.08395988494157791, "learning_rate": 1.0823867169423569e-05, "loss": 0.0165, "step": 36620 }, { "epoch": 10.657550189118417, "grad_norm": 0.07618142664432526, "learning_rate": 1.0815927324080111e-05, "loss": 0.013, "step": 36630 }, { "epoch": 10.660459703229561, "grad_norm": 0.11685293167829514, "learning_rate": 1.080798874974745e-05, "loss": 0.0138, "step": 36640 }, { "epoch": 10.663369217340705, "grad_norm": 0.09392520040273666, "learning_rate": 1.0800051448837124e-05, "loss": 0.0156, "step": 36650 }, { "epoch": 10.666278731451847, "grad_norm": 0.11508429050445557, "learning_rate": 1.079211542376028e-05, "loss": 0.0144, "step": 36660 }, { "epoch": 10.66918824556299, "grad_norm": 0.1980699747800827, "learning_rate": 1.0784180676927664e-05, "loss": 0.0153, "step": 36670 }, { "epoch": 10.672097759674134, "grad_norm": 0.09659045189619064, "learning_rate": 1.0776247210749646e-05, "loss": 0.0144, "step": 36680 }, { "epoch": 10.675007273785278, "grad_norm": 0.08895978331565857, "learning_rate": 1.07683150276362e-05, "loss": 0.0147, "step": 36690 }, { "epoch": 10.677916787896422, "grad_norm": 0.07753913104534149, "learning_rate": 1.0760384129996925e-05, "loss": 0.0151, "step": 36700 }, { "epoch": 10.680826302007565, "grad_norm": 0.0714288279414177, "learning_rate": 1.0752454520241006e-05, "loss": 0.0149, "step": 36710 }, { "epoch": 10.683735816118709, "grad_norm": 0.10206925868988037, "learning_rate": 1.0744526200777254e-05, "loss": 0.0142, "step": 36720 }, { "epoch": 10.686645330229851, "grad_norm": 0.10317527502775192, "learning_rate": 1.0736599174014088e-05, "loss": 0.0164, "step": 36730 }, { "epoch": 10.689554844340995, "grad_norm": 0.11388915777206421, "learning_rate": 1.0728673442359517e-05, "loss": 0.015, "step": 36740 }, { "epoch": 10.692464358452138, "grad_norm": 0.10488435626029968, "learning_rate": 1.0720749008221178e-05, "loss": 0.0148, "step": 36750 }, { "epoch": 10.695373872563282, "grad_norm": 0.1356169581413269, "learning_rate": 1.0712825874006308e-05, "loss": 0.0147, "step": 36760 }, { "epoch": 10.698283386674426, "grad_norm": 0.09249447286128998, "learning_rate": 1.0704904042121736e-05, "loss": 0.0157, "step": 36770 }, { "epoch": 10.70119290078557, "grad_norm": 0.10785356163978577, "learning_rate": 1.0696983514973917e-05, "loss": 0.0136, "step": 36780 }, { "epoch": 10.704102414896711, "grad_norm": 0.08728450536727905, "learning_rate": 1.0689064294968884e-05, "loss": 0.014, "step": 36790 }, { "epoch": 10.707011929007855, "grad_norm": 0.08920876681804657, "learning_rate": 1.0681146384512296e-05, "loss": 0.0155, "step": 36800 }, { "epoch": 10.709921443118999, "grad_norm": 0.10783623903989792, "learning_rate": 1.0673229786009411e-05, "loss": 0.0159, "step": 36810 }, { "epoch": 10.712830957230143, "grad_norm": 0.0857483446598053, "learning_rate": 1.0665314501865066e-05, "loss": 0.0152, "step": 36820 }, { "epoch": 10.715740471341286, "grad_norm": 0.12671604752540588, "learning_rate": 1.0657400534483728e-05, "loss": 0.0158, "step": 36830 }, { "epoch": 10.71864998545243, "grad_norm": 0.09468656778335571, "learning_rate": 1.0649487886269448e-05, "loss": 0.0142, "step": 36840 }, { "epoch": 10.721559499563574, "grad_norm": 0.09732194989919662, "learning_rate": 1.064157655962588e-05, "loss": 0.0162, "step": 36850 }, { "epoch": 10.724469013674716, "grad_norm": 0.09370232373476028, "learning_rate": 1.0633666556956275e-05, "loss": 0.0155, "step": 36860 }, { "epoch": 10.72737852778586, "grad_norm": 0.1102123111486435, "learning_rate": 1.0625757880663482e-05, "loss": 0.0155, "step": 36870 }, { "epoch": 10.730288041897003, "grad_norm": 0.12913334369659424, "learning_rate": 1.0617850533149956e-05, "loss": 0.0153, "step": 36880 }, { "epoch": 10.733197556008147, "grad_norm": 0.07495704293251038, "learning_rate": 1.0609944516817727e-05, "loss": 0.014, "step": 36890 }, { "epoch": 10.73610707011929, "grad_norm": 0.08875039219856262, "learning_rate": 1.0602039834068448e-05, "loss": 0.0149, "step": 36900 }, { "epoch": 10.739016584230434, "grad_norm": 0.09446416795253754, "learning_rate": 1.0594136487303341e-05, "loss": 0.0141, "step": 36910 }, { "epoch": 10.741926098341576, "grad_norm": 0.09482160955667496, "learning_rate": 1.0586234478923237e-05, "loss": 0.0143, "step": 36920 }, { "epoch": 10.74483561245272, "grad_norm": 0.07614340633153915, "learning_rate": 1.0578333811328567e-05, "loss": 0.0136, "step": 36930 }, { "epoch": 10.747745126563863, "grad_norm": 0.08324717730283737, "learning_rate": 1.057043448691933e-05, "loss": 0.0117, "step": 36940 }, { "epoch": 10.750654640675007, "grad_norm": 0.0828099250793457, "learning_rate": 1.0562536508095144e-05, "loss": 0.0136, "step": 36950 }, { "epoch": 10.753564154786151, "grad_norm": 0.06751711666584015, "learning_rate": 1.0554639877255202e-05, "loss": 0.0147, "step": 36960 }, { "epoch": 10.756473668897295, "grad_norm": 0.09534574300050735, "learning_rate": 1.054674459679829e-05, "loss": 0.0133, "step": 36970 }, { "epoch": 10.759383183008438, "grad_norm": 0.09000837057828903, "learning_rate": 1.053885066912279e-05, "loss": 0.0128, "step": 36980 }, { "epoch": 10.76229269711958, "grad_norm": 0.1025284007191658, "learning_rate": 1.0530958096626666e-05, "loss": 0.0162, "step": 36990 }, { "epoch": 10.765202211230724, "grad_norm": 0.08418624848127365, "learning_rate": 1.0523066881707479e-05, "loss": 0.014, "step": 37000 }, { "epoch": 10.768111725341868, "grad_norm": 0.09062622487545013, "learning_rate": 1.0515177026762361e-05, "loss": 0.0147, "step": 37010 }, { "epoch": 10.771021239453011, "grad_norm": 0.10437601059675217, "learning_rate": 1.0507288534188048e-05, "loss": 0.0148, "step": 37020 }, { "epoch": 10.773930753564155, "grad_norm": 0.10624193400144577, "learning_rate": 1.0499401406380859e-05, "loss": 0.0148, "step": 37030 }, { "epoch": 10.776840267675299, "grad_norm": 0.07623979449272156, "learning_rate": 1.0491515645736688e-05, "loss": 0.0161, "step": 37040 }, { "epoch": 10.779749781786442, "grad_norm": 0.07739309221506119, "learning_rate": 1.0483631254651025e-05, "loss": 0.0142, "step": 37050 }, { "epoch": 10.782659295897584, "grad_norm": 0.08628606796264648, "learning_rate": 1.0475748235518939e-05, "loss": 0.0152, "step": 37060 }, { "epoch": 10.785568810008728, "grad_norm": 0.08160300552845001, "learning_rate": 1.0467866590735081e-05, "loss": 0.0147, "step": 37070 }, { "epoch": 10.788478324119872, "grad_norm": 0.12814727425575256, "learning_rate": 1.0459986322693693e-05, "loss": 0.0149, "step": 37080 }, { "epoch": 10.791387838231016, "grad_norm": 0.11924653500318527, "learning_rate": 1.0452107433788583e-05, "loss": 0.0146, "step": 37090 }, { "epoch": 10.79429735234216, "grad_norm": 0.10276668518781662, "learning_rate": 1.044422992641316e-05, "loss": 0.0148, "step": 37100 }, { "epoch": 10.797206866453303, "grad_norm": 0.08539284765720367, "learning_rate": 1.0436353802960395e-05, "loss": 0.0142, "step": 37110 }, { "epoch": 10.800116380564447, "grad_norm": 0.10364104807376862, "learning_rate": 1.0428479065822845e-05, "loss": 0.0162, "step": 37120 }, { "epoch": 10.803025894675589, "grad_norm": 0.08964911103248596, "learning_rate": 1.0420605717392656e-05, "loss": 0.0141, "step": 37130 }, { "epoch": 10.805935408786732, "grad_norm": 0.131947860121727, "learning_rate": 1.0412733760061533e-05, "loss": 0.0149, "step": 37140 }, { "epoch": 10.808844922897876, "grad_norm": 0.12229155004024506, "learning_rate": 1.040486319622078e-05, "loss": 0.0151, "step": 37150 }, { "epoch": 10.81175443700902, "grad_norm": 0.11652657389640808, "learning_rate": 1.0396994028261253e-05, "loss": 0.014, "step": 37160 }, { "epoch": 10.814663951120163, "grad_norm": 0.09795548766851425, "learning_rate": 1.0389126258573409e-05, "loss": 0.0142, "step": 37170 }, { "epoch": 10.817573465231307, "grad_norm": 0.08014168590307236, "learning_rate": 1.0381259889547269e-05, "loss": 0.014, "step": 37180 }, { "epoch": 10.820482979342449, "grad_norm": 0.0770418643951416, "learning_rate": 1.0373394923572418e-05, "loss": 0.0134, "step": 37190 }, { "epoch": 10.823392493453593, "grad_norm": 0.09422784298658371, "learning_rate": 1.0365531363038034e-05, "loss": 0.0161, "step": 37200 }, { "epoch": 10.826302007564736, "grad_norm": 0.11326318979263306, "learning_rate": 1.0357669210332854e-05, "loss": 0.0139, "step": 37210 }, { "epoch": 10.82921152167588, "grad_norm": 0.12693659961223602, "learning_rate": 1.0349808467845191e-05, "loss": 0.0146, "step": 37220 }, { "epoch": 10.832121035787024, "grad_norm": 0.0945097953081131, "learning_rate": 1.0341949137962942e-05, "loss": 0.0159, "step": 37230 }, { "epoch": 10.835030549898168, "grad_norm": 0.10594253242015839, "learning_rate": 1.033409122307355e-05, "loss": 0.0149, "step": 37240 }, { "epoch": 10.83794006400931, "grad_norm": 0.08840081095695496, "learning_rate": 1.0326234725564051e-05, "loss": 0.0135, "step": 37250 }, { "epoch": 10.840849578120453, "grad_norm": 0.0825161337852478, "learning_rate": 1.0318379647821037e-05, "loss": 0.0143, "step": 37260 }, { "epoch": 10.843759092231597, "grad_norm": 0.09024710208177567, "learning_rate": 1.0310525992230675e-05, "loss": 0.0163, "step": 37270 }, { "epoch": 10.84666860634274, "grad_norm": 0.08284886926412582, "learning_rate": 1.03026737611787e-05, "loss": 0.0147, "step": 37280 }, { "epoch": 10.849578120453884, "grad_norm": 0.10375133901834488, "learning_rate": 1.0294822957050406e-05, "loss": 0.0149, "step": 37290 }, { "epoch": 10.852487634565028, "grad_norm": 0.12657174468040466, "learning_rate": 1.0286973582230673e-05, "loss": 0.0146, "step": 37300 }, { "epoch": 10.855397148676172, "grad_norm": 0.09608585387468338, "learning_rate": 1.0279125639103917e-05, "loss": 0.0163, "step": 37310 }, { "epoch": 10.858306662787314, "grad_norm": 0.09503819793462753, "learning_rate": 1.0271279130054154e-05, "loss": 0.0158, "step": 37320 }, { "epoch": 10.861216176898457, "grad_norm": 0.11298179626464844, "learning_rate": 1.0263434057464928e-05, "loss": 0.0144, "step": 37330 }, { "epoch": 10.864125691009601, "grad_norm": 0.10068558156490326, "learning_rate": 1.0255590423719373e-05, "loss": 0.015, "step": 37340 }, { "epoch": 10.867035205120745, "grad_norm": 0.09400478005409241, "learning_rate": 1.0247748231200186e-05, "loss": 0.016, "step": 37350 }, { "epoch": 10.869944719231889, "grad_norm": 0.08888796716928482, "learning_rate": 1.0239907482289605e-05, "loss": 0.0145, "step": 37360 }, { "epoch": 10.872854233343032, "grad_norm": 0.07691303640604019, "learning_rate": 1.0232068179369453e-05, "loss": 0.0144, "step": 37370 }, { "epoch": 10.875763747454176, "grad_norm": 0.09504391252994537, "learning_rate": 1.0224230324821092e-05, "loss": 0.0165, "step": 37380 }, { "epoch": 10.878673261565318, "grad_norm": 0.06884073466062546, "learning_rate": 1.021639392102546e-05, "loss": 0.0132, "step": 37390 }, { "epoch": 10.881582775676462, "grad_norm": 0.0789293423295021, "learning_rate": 1.0208558970363056e-05, "loss": 0.0145, "step": 37400 }, { "epoch": 10.884492289787605, "grad_norm": 0.10011166334152222, "learning_rate": 1.0200725475213926e-05, "loss": 0.0132, "step": 37410 }, { "epoch": 10.887401803898749, "grad_norm": 0.14037154614925385, "learning_rate": 1.019289343795768e-05, "loss": 0.014, "step": 37420 }, { "epoch": 10.890311318009893, "grad_norm": 0.09769529849290848, "learning_rate": 1.0185062860973478e-05, "loss": 0.0143, "step": 37430 }, { "epoch": 10.893220832121036, "grad_norm": 0.08997710049152374, "learning_rate": 1.017723374664005e-05, "loss": 0.0143, "step": 37440 }, { "epoch": 10.89613034623218, "grad_norm": 0.0624091662466526, "learning_rate": 1.0169406097335676e-05, "loss": 0.0141, "step": 37450 }, { "epoch": 10.899039860343322, "grad_norm": 0.11657574027776718, "learning_rate": 1.0161579915438178e-05, "loss": 0.0162, "step": 37460 }, { "epoch": 10.901949374454466, "grad_norm": 0.10669881850481033, "learning_rate": 1.0153755203324961e-05, "loss": 0.015, "step": 37470 }, { "epoch": 10.90485888856561, "grad_norm": 0.10233265906572342, "learning_rate": 1.0145931963372947e-05, "loss": 0.0142, "step": 37480 }, { "epoch": 10.907768402676753, "grad_norm": 0.10131096094846725, "learning_rate": 1.0138110197958638e-05, "loss": 0.0139, "step": 37490 }, { "epoch": 10.910677916787897, "grad_norm": 0.0965999886393547, "learning_rate": 1.0130289909458085e-05, "loss": 0.0145, "step": 37500 }, { "epoch": 10.91358743089904, "grad_norm": 0.08164308965206146, "learning_rate": 1.0122471100246877e-05, "loss": 0.0139, "step": 37510 }, { "epoch": 10.916496945010183, "grad_norm": 0.09501001238822937, "learning_rate": 1.0114653772700168e-05, "loss": 0.0141, "step": 37520 }, { "epoch": 10.919406459121326, "grad_norm": 0.1401377171278, "learning_rate": 1.010683792919265e-05, "loss": 0.0147, "step": 37530 }, { "epoch": 10.92231597323247, "grad_norm": 0.08946261554956436, "learning_rate": 1.009902357209857e-05, "loss": 0.0134, "step": 37540 }, { "epoch": 10.925225487343614, "grad_norm": 0.08074209094047546, "learning_rate": 1.009121070379173e-05, "loss": 0.016, "step": 37550 }, { "epoch": 10.928135001454757, "grad_norm": 0.08395368605852127, "learning_rate": 1.0083399326645469e-05, "loss": 0.0149, "step": 37560 }, { "epoch": 10.931044515565901, "grad_norm": 0.10483583807945251, "learning_rate": 1.007558944303268e-05, "loss": 0.0145, "step": 37570 }, { "epoch": 10.933954029677043, "grad_norm": 0.08954320102930069, "learning_rate": 1.006778105532579e-05, "loss": 0.0145, "step": 37580 }, { "epoch": 10.936863543788187, "grad_norm": 0.0935533419251442, "learning_rate": 1.0059974165896787e-05, "loss": 0.0134, "step": 37590 }, { "epoch": 10.93977305789933, "grad_norm": 0.09866839647293091, "learning_rate": 1.0052168777117207e-05, "loss": 0.0142, "step": 37600 }, { "epoch": 10.942682572010474, "grad_norm": 0.0661332979798317, "learning_rate": 1.0044364891358108e-05, "loss": 0.0136, "step": 37610 }, { "epoch": 10.945592086121618, "grad_norm": 0.10428694635629654, "learning_rate": 1.0036562510990112e-05, "loss": 0.0143, "step": 37620 }, { "epoch": 10.948501600232762, "grad_norm": 0.12269623577594757, "learning_rate": 1.002876163838337e-05, "loss": 0.0148, "step": 37630 }, { "epoch": 10.951411114343905, "grad_norm": 0.0768473744392395, "learning_rate": 1.0020962275907583e-05, "loss": 0.0136, "step": 37640 }, { "epoch": 10.954320628455047, "grad_norm": 0.09551824629306793, "learning_rate": 1.0013164425931997e-05, "loss": 0.0145, "step": 37650 }, { "epoch": 10.95723014256619, "grad_norm": 0.10475314408540726, "learning_rate": 1.0005368090825385e-05, "loss": 0.015, "step": 37660 }, { "epoch": 10.960139656677335, "grad_norm": 0.1276400089263916, "learning_rate": 9.997573272956076e-06, "loss": 0.0143, "step": 37670 }, { "epoch": 10.963049170788478, "grad_norm": 0.08758185058832169, "learning_rate": 9.98977997469192e-06, "loss": 0.0137, "step": 37680 }, { "epoch": 10.965958684899622, "grad_norm": 0.09311164915561676, "learning_rate": 9.981988198400323e-06, "loss": 0.0138, "step": 37690 }, { "epoch": 10.968868199010766, "grad_norm": 0.12297255545854568, "learning_rate": 9.974197946448215e-06, "loss": 0.0134, "step": 37700 }, { "epoch": 10.97177771312191, "grad_norm": 0.10171477496623993, "learning_rate": 9.966409221202073e-06, "loss": 0.0134, "step": 37710 }, { "epoch": 10.974687227233051, "grad_norm": 0.09270523488521576, "learning_rate": 9.958622025027911e-06, "loss": 0.0155, "step": 37720 }, { "epoch": 10.977596741344195, "grad_norm": 0.10562647134065628, "learning_rate": 9.950836360291261e-06, "loss": 0.0154, "step": 37730 }, { "epoch": 10.980506255455339, "grad_norm": 0.09713825583457947, "learning_rate": 9.943052229357217e-06, "loss": 0.0146, "step": 37740 }, { "epoch": 10.983415769566482, "grad_norm": 0.09030510485172272, "learning_rate": 9.935269634590377e-06, "loss": 0.0148, "step": 37750 }, { "epoch": 10.986325283677626, "grad_norm": 0.09164624661207199, "learning_rate": 9.927488578354898e-06, "loss": 0.0157, "step": 37760 }, { "epoch": 10.98923479778877, "grad_norm": 0.1254998743534088, "learning_rate": 9.919709063014461e-06, "loss": 0.014, "step": 37770 }, { "epoch": 10.992144311899914, "grad_norm": 0.08720699697732925, "learning_rate": 9.911931090932269e-06, "loss": 0.014, "step": 37780 }, { "epoch": 10.995053826011056, "grad_norm": 0.10781741887331009, "learning_rate": 9.904154664471072e-06, "loss": 0.0142, "step": 37790 }, { "epoch": 10.9979633401222, "grad_norm": 0.13143250346183777, "learning_rate": 9.896379785993137e-06, "loss": 0.0134, "step": 37800 }, { "epoch": 11.000872854233343, "grad_norm": 0.09512372314929962, "learning_rate": 9.888606457860268e-06, "loss": 0.0143, "step": 37810 }, { "epoch": 11.003782368344487, "grad_norm": 0.1218508780002594, "learning_rate": 9.880834682433803e-06, "loss": 0.0161, "step": 37820 }, { "epoch": 11.00669188245563, "grad_norm": 0.10203303396701813, "learning_rate": 9.873064462074596e-06, "loss": 0.0152, "step": 37830 }, { "epoch": 11.009601396566774, "grad_norm": 0.07687167078256607, "learning_rate": 9.86529579914304e-06, "loss": 0.0148, "step": 37840 }, { "epoch": 11.012510910677916, "grad_norm": 0.08961066603660583, "learning_rate": 9.857528695999038e-06, "loss": 0.0155, "step": 37850 }, { "epoch": 11.01542042478906, "grad_norm": 0.09871958941221237, "learning_rate": 9.849763155002043e-06, "loss": 0.0155, "step": 37860 }, { "epoch": 11.018329938900203, "grad_norm": 0.09229865670204163, "learning_rate": 9.841999178511022e-06, "loss": 0.0142, "step": 37870 }, { "epoch": 11.021239453011347, "grad_norm": 0.08366646617650986, "learning_rate": 9.834236768884454e-06, "loss": 0.0146, "step": 37880 }, { "epoch": 11.02414896712249, "grad_norm": 0.08240217715501785, "learning_rate": 9.826475928480369e-06, "loss": 0.0152, "step": 37890 }, { "epoch": 11.027058481233635, "grad_norm": 0.1109221875667572, "learning_rate": 9.818716659656293e-06, "loss": 0.0133, "step": 37900 }, { "epoch": 11.029967995344778, "grad_norm": 0.08579573780298233, "learning_rate": 9.81095896476929e-06, "loss": 0.0143, "step": 37910 }, { "epoch": 11.03287750945592, "grad_norm": 0.09820345044136047, "learning_rate": 9.803202846175953e-06, "loss": 0.0145, "step": 37920 }, { "epoch": 11.035787023567064, "grad_norm": 0.0944761410355568, "learning_rate": 9.79544830623237e-06, "loss": 0.0158, "step": 37930 }, { "epoch": 11.038696537678208, "grad_norm": 0.08937295526266098, "learning_rate": 9.78769534729418e-06, "loss": 0.0137, "step": 37940 }, { "epoch": 11.041606051789351, "grad_norm": 0.10514526814222336, "learning_rate": 9.779943971716514e-06, "loss": 0.0128, "step": 37950 }, { "epoch": 11.044515565900495, "grad_norm": 0.07782764732837677, "learning_rate": 9.772194181854043e-06, "loss": 0.0141, "step": 37960 }, { "epoch": 11.047425080011639, "grad_norm": 0.0895526334643364, "learning_rate": 9.76444598006095e-06, "loss": 0.0135, "step": 37970 }, { "epoch": 11.05033459412278, "grad_norm": 0.10054484009742737, "learning_rate": 9.756699368690931e-06, "loss": 0.0147, "step": 37980 }, { "epoch": 11.053244108233924, "grad_norm": 0.08189494907855988, "learning_rate": 9.748954350097201e-06, "loss": 0.0152, "step": 37990 }, { "epoch": 11.056153622345068, "grad_norm": 0.11105866730213165, "learning_rate": 9.741210926632494e-06, "loss": 0.0153, "step": 38000 }, { "epoch": 11.059063136456212, "grad_norm": 0.08028875291347504, "learning_rate": 9.733469100649055e-06, "loss": 0.0141, "step": 38010 }, { "epoch": 11.061972650567355, "grad_norm": 0.08379866182804108, "learning_rate": 9.725728874498655e-06, "loss": 0.0153, "step": 38020 }, { "epoch": 11.0648821646785, "grad_norm": 0.09854670614004135, "learning_rate": 9.717990250532556e-06, "loss": 0.0144, "step": 38030 }, { "epoch": 11.067791678789643, "grad_norm": 0.08286099135875702, "learning_rate": 9.710253231101565e-06, "loss": 0.0146, "step": 38040 }, { "epoch": 11.070701192900785, "grad_norm": 0.07409900426864624, "learning_rate": 9.702517818555964e-06, "loss": 0.0152, "step": 38050 }, { "epoch": 11.073610707011929, "grad_norm": 0.0944238007068634, "learning_rate": 9.694784015245584e-06, "loss": 0.0151, "step": 38060 }, { "epoch": 11.076520221123072, "grad_norm": 0.10517814755439758, "learning_rate": 9.687051823519745e-06, "loss": 0.0154, "step": 38070 }, { "epoch": 11.079429735234216, "grad_norm": 0.10949978232383728, "learning_rate": 9.679321245727278e-06, "loss": 0.0151, "step": 38080 }, { "epoch": 11.08233924934536, "grad_norm": 0.11059867590665817, "learning_rate": 9.671592284216537e-06, "loss": 0.0155, "step": 38090 }, { "epoch": 11.085248763456503, "grad_norm": 0.07232819497585297, "learning_rate": 9.663864941335367e-06, "loss": 0.0129, "step": 38100 }, { "epoch": 11.088158277567647, "grad_norm": 0.0877014771103859, "learning_rate": 9.656139219431138e-06, "loss": 0.0156, "step": 38110 }, { "epoch": 11.091067791678789, "grad_norm": 0.08368748426437378, "learning_rate": 9.648415120850714e-06, "loss": 0.0139, "step": 38120 }, { "epoch": 11.093977305789933, "grad_norm": 0.21955864131450653, "learning_rate": 9.64069264794048e-06, "loss": 0.0153, "step": 38130 }, { "epoch": 11.096886819901076, "grad_norm": 0.09028193354606628, "learning_rate": 9.632971803046313e-06, "loss": 0.0136, "step": 38140 }, { "epoch": 11.09979633401222, "grad_norm": 0.11953037977218628, "learning_rate": 9.625252588513604e-06, "loss": 0.0153, "step": 38150 }, { "epoch": 11.102705848123364, "grad_norm": 0.1056593805551529, "learning_rate": 9.61753500668725e-06, "loss": 0.0138, "step": 38160 }, { "epoch": 11.105615362234508, "grad_norm": 0.09274862706661224, "learning_rate": 9.609819059911643e-06, "loss": 0.0151, "step": 38170 }, { "epoch": 11.10852487634565, "grad_norm": 0.08286253362894058, "learning_rate": 9.602104750530682e-06, "loss": 0.0144, "step": 38180 }, { "epoch": 11.111434390456793, "grad_norm": 0.08249843865633011, "learning_rate": 9.594392080887783e-06, "loss": 0.0138, "step": 38190 }, { "epoch": 11.114343904567937, "grad_norm": 0.11271748691797256, "learning_rate": 9.586681053325836e-06, "loss": 0.0134, "step": 38200 }, { "epoch": 11.11725341867908, "grad_norm": 0.09240779280662537, "learning_rate": 9.578971670187259e-06, "loss": 0.016, "step": 38210 }, { "epoch": 11.120162932790224, "grad_norm": 0.09276774525642395, "learning_rate": 9.57126393381395e-06, "loss": 0.0128, "step": 38220 }, { "epoch": 11.123072446901368, "grad_norm": 0.09380434453487396, "learning_rate": 9.563557846547321e-06, "loss": 0.0143, "step": 38230 }, { "epoch": 11.125981961012512, "grad_norm": 0.0895095095038414, "learning_rate": 9.555853410728283e-06, "loss": 0.0125, "step": 38240 }, { "epoch": 11.128891475123654, "grad_norm": 0.10246508568525314, "learning_rate": 9.54815062869723e-06, "loss": 0.0145, "step": 38250 }, { "epoch": 11.131800989234797, "grad_norm": 0.11109674721956253, "learning_rate": 9.540449502794068e-06, "loss": 0.0154, "step": 38260 }, { "epoch": 11.134710503345941, "grad_norm": 0.11765151470899582, "learning_rate": 9.532750035358194e-06, "loss": 0.0159, "step": 38270 }, { "epoch": 11.137620017457085, "grad_norm": 0.07767774909734726, "learning_rate": 9.52505222872851e-06, "loss": 0.0134, "step": 38280 }, { "epoch": 11.140529531568228, "grad_norm": 0.0766606479883194, "learning_rate": 9.5173560852434e-06, "loss": 0.0139, "step": 38290 }, { "epoch": 11.143439045679372, "grad_norm": 0.09885755926370621, "learning_rate": 9.509661607240749e-06, "loss": 0.0141, "step": 38300 }, { "epoch": 11.146348559790514, "grad_norm": 0.09214988350868225, "learning_rate": 9.50196879705794e-06, "loss": 0.0142, "step": 38310 }, { "epoch": 11.149258073901658, "grad_norm": 0.10933130234479904, "learning_rate": 9.49427765703184e-06, "loss": 0.015, "step": 38320 }, { "epoch": 11.152167588012802, "grad_norm": 0.1002318263053894, "learning_rate": 9.486588189498816e-06, "loss": 0.0134, "step": 38330 }, { "epoch": 11.155077102123945, "grad_norm": 0.07138784229755402, "learning_rate": 9.478900396794736e-06, "loss": 0.0128, "step": 38340 }, { "epoch": 11.157986616235089, "grad_norm": 0.12528975307941437, "learning_rate": 9.47121428125493e-06, "loss": 0.0136, "step": 38350 }, { "epoch": 11.160896130346233, "grad_norm": 0.10921413451433182, "learning_rate": 9.463529845214253e-06, "loss": 0.0146, "step": 38360 }, { "epoch": 11.163805644457376, "grad_norm": 0.08588913083076477, "learning_rate": 9.455847091007023e-06, "loss": 0.0134, "step": 38370 }, { "epoch": 11.166715158568518, "grad_norm": 0.11919644474983215, "learning_rate": 9.44816602096706e-06, "loss": 0.0149, "step": 38380 }, { "epoch": 11.169624672679662, "grad_norm": 0.08496907353401184, "learning_rate": 9.440486637427677e-06, "loss": 0.0135, "step": 38390 }, { "epoch": 11.172534186790806, "grad_norm": 0.0957072377204895, "learning_rate": 9.432808942721659e-06, "loss": 0.0148, "step": 38400 }, { "epoch": 11.17544370090195, "grad_norm": 0.13940319418907166, "learning_rate": 9.425132939181294e-06, "loss": 0.0137, "step": 38410 }, { "epoch": 11.178353215013093, "grad_norm": 0.17926816642284393, "learning_rate": 9.417458629138342e-06, "loss": 0.0127, "step": 38420 }, { "epoch": 11.181262729124237, "grad_norm": 0.06420166045427322, "learning_rate": 9.409786014924064e-06, "loss": 0.0137, "step": 38430 }, { "epoch": 11.18417224323538, "grad_norm": 0.11133631318807602, "learning_rate": 9.402115098869193e-06, "loss": 0.0141, "step": 38440 }, { "epoch": 11.187081757346522, "grad_norm": 0.08205176144838333, "learning_rate": 9.394445883303951e-06, "loss": 0.0129, "step": 38450 }, { "epoch": 11.189991271457666, "grad_norm": 0.07839523255825043, "learning_rate": 9.386778370558047e-06, "loss": 0.0156, "step": 38460 }, { "epoch": 11.19290078556881, "grad_norm": 0.07007651031017303, "learning_rate": 9.37911256296066e-06, "loss": 0.0146, "step": 38470 }, { "epoch": 11.195810299679954, "grad_norm": 0.12113568186759949, "learning_rate": 9.371448462840464e-06, "loss": 0.0138, "step": 38480 }, { "epoch": 11.198719813791097, "grad_norm": 0.0639490932226181, "learning_rate": 9.363786072525619e-06, "loss": 0.0133, "step": 38490 }, { "epoch": 11.201629327902241, "grad_norm": 0.07300303131341934, "learning_rate": 9.356125394343744e-06, "loss": 0.0158, "step": 38500 }, { "epoch": 11.204538842013383, "grad_norm": 0.11332804709672928, "learning_rate": 9.348466430621963e-06, "loss": 0.0148, "step": 38510 }, { "epoch": 11.207448356124527, "grad_norm": 0.08533833920955658, "learning_rate": 9.34080918368685e-06, "loss": 0.0148, "step": 38520 }, { "epoch": 11.21035787023567, "grad_norm": 0.1230078712105751, "learning_rate": 9.333153655864493e-06, "loss": 0.016, "step": 38530 }, { "epoch": 11.213267384346814, "grad_norm": 0.08690423518419266, "learning_rate": 9.325499849480428e-06, "loss": 0.0157, "step": 38540 }, { "epoch": 11.216176898457958, "grad_norm": 0.1132742166519165, "learning_rate": 9.317847766859679e-06, "loss": 0.0145, "step": 38550 }, { "epoch": 11.219086412569101, "grad_norm": 0.08520562201738358, "learning_rate": 9.310197410326756e-06, "loss": 0.0138, "step": 38560 }, { "epoch": 11.221995926680245, "grad_norm": 0.10632943361997604, "learning_rate": 9.302548782205624e-06, "loss": 0.0139, "step": 38570 }, { "epoch": 11.224905440791387, "grad_norm": 0.0892411544919014, "learning_rate": 9.294901884819745e-06, "loss": 0.0139, "step": 38580 }, { "epoch": 11.22781495490253, "grad_norm": 0.0855162963271141, "learning_rate": 9.287256720492035e-06, "loss": 0.0128, "step": 38590 }, { "epoch": 11.230724469013674, "grad_norm": 0.09132663905620575, "learning_rate": 9.279613291544892e-06, "loss": 0.0137, "step": 38600 }, { "epoch": 11.233633983124818, "grad_norm": 0.12896917760372162, "learning_rate": 9.271971600300204e-06, "loss": 0.0157, "step": 38610 }, { "epoch": 11.236543497235962, "grad_norm": 0.08528541773557663, "learning_rate": 9.264331649079298e-06, "loss": 0.0144, "step": 38620 }, { "epoch": 11.239453011347106, "grad_norm": 0.1090557649731636, "learning_rate": 9.256693440202999e-06, "loss": 0.015, "step": 38630 }, { "epoch": 11.24236252545825, "grad_norm": 0.15471409261226654, "learning_rate": 9.249056975991586e-06, "loss": 0.0142, "step": 38640 }, { "epoch": 11.245272039569391, "grad_norm": 0.1556185483932495, "learning_rate": 9.241422258764816e-06, "loss": 0.0143, "step": 38650 }, { "epoch": 11.248181553680535, "grad_norm": 0.09330505132675171, "learning_rate": 9.233789290841927e-06, "loss": 0.0149, "step": 38660 }, { "epoch": 11.251091067791679, "grad_norm": 0.0960702896118164, "learning_rate": 9.226158074541593e-06, "loss": 0.0131, "step": 38670 }, { "epoch": 11.254000581902822, "grad_norm": 0.09929781407117844, "learning_rate": 9.218528612181993e-06, "loss": 0.0134, "step": 38680 }, { "epoch": 11.256910096013966, "grad_norm": 0.08031968027353287, "learning_rate": 9.210900906080748e-06, "loss": 0.0149, "step": 38690 }, { "epoch": 11.25981961012511, "grad_norm": 0.10858916491270065, "learning_rate": 9.203274958554954e-06, "loss": 0.0143, "step": 38700 }, { "epoch": 11.262729124236252, "grad_norm": 0.11520040035247803, "learning_rate": 9.195650771921175e-06, "loss": 0.0138, "step": 38710 }, { "epoch": 11.265638638347395, "grad_norm": 0.10159282386302948, "learning_rate": 9.188028348495435e-06, "loss": 0.0143, "step": 38720 }, { "epoch": 11.26854815245854, "grad_norm": 0.10454180836677551, "learning_rate": 9.180407690593231e-06, "loss": 0.0133, "step": 38730 }, { "epoch": 11.271457666569683, "grad_norm": 0.11830858886241913, "learning_rate": 9.17278880052951e-06, "loss": 0.0147, "step": 38740 }, { "epoch": 11.274367180680827, "grad_norm": 0.11102917045354843, "learning_rate": 9.16517168061869e-06, "loss": 0.0132, "step": 38750 }, { "epoch": 11.27727669479197, "grad_norm": 0.11111327260732651, "learning_rate": 9.157556333174662e-06, "loss": 0.0136, "step": 38760 }, { "epoch": 11.280186208903114, "grad_norm": 0.0964566096663475, "learning_rate": 9.14994276051075e-06, "loss": 0.0143, "step": 38770 }, { "epoch": 11.283095723014256, "grad_norm": 0.14637917280197144, "learning_rate": 9.142330964939774e-06, "loss": 0.0147, "step": 38780 }, { "epoch": 11.2860052371254, "grad_norm": 0.0818556472659111, "learning_rate": 9.134720948773982e-06, "loss": 0.0147, "step": 38790 }, { "epoch": 11.288914751236543, "grad_norm": 0.07059363275766373, "learning_rate": 9.127112714325101e-06, "loss": 0.0138, "step": 38800 }, { "epoch": 11.291824265347687, "grad_norm": 0.09245149791240692, "learning_rate": 9.11950626390432e-06, "loss": 0.013, "step": 38810 }, { "epoch": 11.29473377945883, "grad_norm": 0.09246931225061417, "learning_rate": 9.111901599822263e-06, "loss": 0.0133, "step": 38820 }, { "epoch": 11.297643293569974, "grad_norm": 0.1513374000787735, "learning_rate": 9.104298724389039e-06, "loss": 0.0142, "step": 38830 }, { "epoch": 11.300552807681118, "grad_norm": 0.12376555055379868, "learning_rate": 9.096697639914192e-06, "loss": 0.0151, "step": 38840 }, { "epoch": 11.30346232179226, "grad_norm": 0.09588395059108734, "learning_rate": 9.089098348706734e-06, "loss": 0.0129, "step": 38850 }, { "epoch": 11.306371835903404, "grad_norm": 0.10057131201028824, "learning_rate": 9.081500853075132e-06, "loss": 0.0153, "step": 38860 }, { "epoch": 11.309281350014547, "grad_norm": 0.07577554881572723, "learning_rate": 9.073905155327298e-06, "loss": 0.0141, "step": 38870 }, { "epoch": 11.312190864125691, "grad_norm": 0.1192447692155838, "learning_rate": 9.066311257770615e-06, "loss": 0.013, "step": 38880 }, { "epoch": 11.315100378236835, "grad_norm": 0.07083728909492493, "learning_rate": 9.058719162711897e-06, "loss": 0.0147, "step": 38890 }, { "epoch": 11.318009892347979, "grad_norm": 0.0937337875366211, "learning_rate": 9.051128872457424e-06, "loss": 0.0148, "step": 38900 }, { "epoch": 11.32091940645912, "grad_norm": 0.09815724939107895, "learning_rate": 9.043540389312938e-06, "loss": 0.0149, "step": 38910 }, { "epoch": 11.323828920570264, "grad_norm": 0.10292569547891617, "learning_rate": 9.035953715583604e-06, "loss": 0.014, "step": 38920 }, { "epoch": 11.326738434681408, "grad_norm": 0.09467686712741852, "learning_rate": 9.028368853574065e-06, "loss": 0.0142, "step": 38930 }, { "epoch": 11.329647948792552, "grad_norm": 0.07676172256469727, "learning_rate": 9.02078580558839e-06, "loss": 0.0147, "step": 38940 }, { "epoch": 11.332557462903695, "grad_norm": 0.09559540450572968, "learning_rate": 9.013204573930123e-06, "loss": 0.0136, "step": 38950 }, { "epoch": 11.335466977014839, "grad_norm": 0.10769447684288025, "learning_rate": 9.005625160902227e-06, "loss": 0.0132, "step": 38960 }, { "epoch": 11.338376491125983, "grad_norm": 0.07769870012998581, "learning_rate": 8.998047568807133e-06, "loss": 0.0145, "step": 38970 }, { "epoch": 11.341286005237125, "grad_norm": 0.11738584190607071, "learning_rate": 8.99047179994672e-06, "loss": 0.0143, "step": 38980 }, { "epoch": 11.344195519348268, "grad_norm": 0.08603871613740921, "learning_rate": 8.9828978566223e-06, "loss": 0.0142, "step": 38990 }, { "epoch": 11.347105033459412, "grad_norm": 0.09421418607234955, "learning_rate": 8.975325741134644e-06, "loss": 0.016, "step": 39000 }, { "epoch": 11.350014547570556, "grad_norm": 0.12646488845348358, "learning_rate": 8.967755455783947e-06, "loss": 0.0157, "step": 39010 }, { "epoch": 11.3529240616817, "grad_norm": 0.09341306984424591, "learning_rate": 8.960187002869872e-06, "loss": 0.0155, "step": 39020 }, { "epoch": 11.355833575792843, "grad_norm": 0.0810171440243721, "learning_rate": 8.95262038469152e-06, "loss": 0.0117, "step": 39030 }, { "epoch": 11.358743089903985, "grad_norm": 0.0906781479716301, "learning_rate": 8.945055603547418e-06, "loss": 0.0152, "step": 39040 }, { "epoch": 11.361652604015129, "grad_norm": 0.08890226483345032, "learning_rate": 8.937492661735557e-06, "loss": 0.0137, "step": 39050 }, { "epoch": 11.364562118126273, "grad_norm": 0.10958243161439896, "learning_rate": 8.929931561553349e-06, "loss": 0.0131, "step": 39060 }, { "epoch": 11.367471632237416, "grad_norm": 0.11536101996898651, "learning_rate": 8.922372305297662e-06, "loss": 0.0147, "step": 39070 }, { "epoch": 11.37038114634856, "grad_norm": 0.11802159249782562, "learning_rate": 8.9148148952648e-06, "loss": 0.0153, "step": 39080 }, { "epoch": 11.373290660459704, "grad_norm": 0.08316300809383392, "learning_rate": 8.907259333750503e-06, "loss": 0.0154, "step": 39090 }, { "epoch": 11.376200174570847, "grad_norm": 0.13496091961860657, "learning_rate": 8.899705623049952e-06, "loss": 0.0142, "step": 39100 }, { "epoch": 11.37910968868199, "grad_norm": 0.09680332243442535, "learning_rate": 8.89215376545776e-06, "loss": 0.0152, "step": 39110 }, { "epoch": 11.382019202793133, "grad_norm": 0.08937074989080429, "learning_rate": 8.884603763267985e-06, "loss": 0.014, "step": 39120 }, { "epoch": 11.384928716904277, "grad_norm": 0.11322426050901413, "learning_rate": 8.877055618774124e-06, "loss": 0.0149, "step": 39130 }, { "epoch": 11.38783823101542, "grad_norm": 0.08948460966348648, "learning_rate": 8.869509334269096e-06, "loss": 0.0154, "step": 39140 }, { "epoch": 11.390747745126564, "grad_norm": 0.10604261606931686, "learning_rate": 8.861964912045268e-06, "loss": 0.0151, "step": 39150 }, { "epoch": 11.393657259237708, "grad_norm": 0.10246381163597107, "learning_rate": 8.854422354394433e-06, "loss": 0.0132, "step": 39160 }, { "epoch": 11.396566773348852, "grad_norm": 0.08818592876195908, "learning_rate": 8.84688166360782e-06, "loss": 0.0144, "step": 39170 }, { "epoch": 11.399476287459994, "grad_norm": 0.07884637266397476, "learning_rate": 8.8393428419761e-06, "loss": 0.013, "step": 39180 }, { "epoch": 11.402385801571137, "grad_norm": 0.11325618624687195, "learning_rate": 8.831805891789354e-06, "loss": 0.0142, "step": 39190 }, { "epoch": 11.405295315682281, "grad_norm": 0.09341298043727875, "learning_rate": 8.824270815337126e-06, "loss": 0.0138, "step": 39200 }, { "epoch": 11.408204829793425, "grad_norm": 0.10538271814584732, "learning_rate": 8.816737614908355e-06, "loss": 0.0138, "step": 39210 }, { "epoch": 11.411114343904568, "grad_norm": 0.10461614280939102, "learning_rate": 8.809206292791438e-06, "loss": 0.0134, "step": 39220 }, { "epoch": 11.414023858015712, "grad_norm": 0.10234629362821579, "learning_rate": 8.801676851274197e-06, "loss": 0.0137, "step": 39230 }, { "epoch": 11.416933372126854, "grad_norm": 0.1109677255153656, "learning_rate": 8.794149292643864e-06, "loss": 0.014, "step": 39240 }, { "epoch": 11.419842886237998, "grad_norm": 0.10326403379440308, "learning_rate": 8.786623619187128e-06, "loss": 0.0132, "step": 39250 }, { "epoch": 11.422752400349141, "grad_norm": 0.08908642083406448, "learning_rate": 8.779099833190073e-06, "loss": 0.0128, "step": 39260 }, { "epoch": 11.425661914460285, "grad_norm": 0.09716662764549255, "learning_rate": 8.771577936938239e-06, "loss": 0.0133, "step": 39270 }, { "epoch": 11.428571428571429, "grad_norm": 0.10227945446968079, "learning_rate": 8.764057932716575e-06, "loss": 0.0148, "step": 39280 }, { "epoch": 11.431480942682573, "grad_norm": 0.08592981845140457, "learning_rate": 8.756539822809463e-06, "loss": 0.0137, "step": 39290 }, { "epoch": 11.434390456793716, "grad_norm": 0.08206906914710999, "learning_rate": 8.749023609500704e-06, "loss": 0.0144, "step": 39300 }, { "epoch": 11.437299970904858, "grad_norm": 0.10388447344303131, "learning_rate": 8.741509295073525e-06, "loss": 0.0155, "step": 39310 }, { "epoch": 11.440209485016002, "grad_norm": 0.09029947966337204, "learning_rate": 8.733996881810574e-06, "loss": 0.0136, "step": 39320 }, { "epoch": 11.443118999127146, "grad_norm": 0.09899389743804932, "learning_rate": 8.726486371993928e-06, "loss": 0.0136, "step": 39330 }, { "epoch": 11.44602851323829, "grad_norm": 0.09811296314001083, "learning_rate": 8.718977767905072e-06, "loss": 0.0143, "step": 39340 }, { "epoch": 11.448938027349433, "grad_norm": 0.0862271636724472, "learning_rate": 8.711471071824938e-06, "loss": 0.0136, "step": 39350 }, { "epoch": 11.451847541460577, "grad_norm": 0.09839482605457306, "learning_rate": 8.703966286033852e-06, "loss": 0.0153, "step": 39360 }, { "epoch": 11.454757055571719, "grad_norm": 0.10118918120861053, "learning_rate": 8.696463412811569e-06, "loss": 0.0129, "step": 39370 }, { "epoch": 11.457666569682862, "grad_norm": 0.09223374724388123, "learning_rate": 8.68896245443727e-06, "loss": 0.0147, "step": 39380 }, { "epoch": 11.460576083794006, "grad_norm": 0.10537391155958176, "learning_rate": 8.681463413189535e-06, "loss": 0.0136, "step": 39390 }, { "epoch": 11.46348559790515, "grad_norm": 0.1327352523803711, "learning_rate": 8.673966291346395e-06, "loss": 0.014, "step": 39400 }, { "epoch": 11.466395112016293, "grad_norm": 0.07973611354827881, "learning_rate": 8.666471091185258e-06, "loss": 0.0126, "step": 39410 }, { "epoch": 11.469304626127437, "grad_norm": 0.0885097086429596, "learning_rate": 8.658977814982986e-06, "loss": 0.0153, "step": 39420 }, { "epoch": 11.47221414023858, "grad_norm": 0.11188068240880966, "learning_rate": 8.651486465015818e-06, "loss": 0.0139, "step": 39430 }, { "epoch": 11.475123654349723, "grad_norm": 0.07267429679632187, "learning_rate": 8.643997043559446e-06, "loss": 0.0145, "step": 39440 }, { "epoch": 11.478033168460867, "grad_norm": 0.13680976629257202, "learning_rate": 8.636509552888948e-06, "loss": 0.0166, "step": 39450 }, { "epoch": 11.48094268257201, "grad_norm": 0.08268503099679947, "learning_rate": 8.629023995278834e-06, "loss": 0.0128, "step": 39460 }, { "epoch": 11.483852196683154, "grad_norm": 0.15283744037151337, "learning_rate": 8.621540373003011e-06, "loss": 0.0134, "step": 39470 }, { "epoch": 11.486761710794298, "grad_norm": 0.14111609756946564, "learning_rate": 8.614058688334809e-06, "loss": 0.0125, "step": 39480 }, { "epoch": 11.489671224905441, "grad_norm": 0.09916407614946365, "learning_rate": 8.606578943546962e-06, "loss": 0.0153, "step": 39490 }, { "epoch": 11.492580739016585, "grad_norm": 0.12193623930215836, "learning_rate": 8.59910114091163e-06, "loss": 0.0149, "step": 39500 }, { "epoch": 11.495490253127727, "grad_norm": 0.10696612298488617, "learning_rate": 8.591625282700364e-06, "loss": 0.016, "step": 39510 }, { "epoch": 11.49839976723887, "grad_norm": 0.08190912008285522, "learning_rate": 8.584151371184135e-06, "loss": 0.0138, "step": 39520 }, { "epoch": 11.501309281350014, "grad_norm": 0.08488373458385468, "learning_rate": 8.576679408633318e-06, "loss": 0.0126, "step": 39530 }, { "epoch": 11.504218795461158, "grad_norm": 0.08789441734552383, "learning_rate": 8.569209397317694e-06, "loss": 0.0153, "step": 39540 }, { "epoch": 11.507128309572302, "grad_norm": 0.09683622419834137, "learning_rate": 8.561741339506469e-06, "loss": 0.0137, "step": 39550 }, { "epoch": 11.510037823683446, "grad_norm": 0.1353137344121933, "learning_rate": 8.554275237468225e-06, "loss": 0.0133, "step": 39560 }, { "epoch": 11.51294733779459, "grad_norm": 0.07956407219171524, "learning_rate": 8.546811093470981e-06, "loss": 0.0126, "step": 39570 }, { "epoch": 11.515856851905731, "grad_norm": 0.08537231385707855, "learning_rate": 8.539348909782132e-06, "loss": 0.0137, "step": 39580 }, { "epoch": 11.518766366016875, "grad_norm": 0.08107581734657288, "learning_rate": 8.531888688668507e-06, "loss": 0.0137, "step": 39590 }, { "epoch": 11.521675880128019, "grad_norm": 0.09153013676404953, "learning_rate": 8.524430432396316e-06, "loss": 0.0143, "step": 39600 }, { "epoch": 11.524585394239162, "grad_norm": 0.09536144882440567, "learning_rate": 8.516974143231181e-06, "loss": 0.0127, "step": 39610 }, { "epoch": 11.527494908350306, "grad_norm": 0.1304817497730255, "learning_rate": 8.50951982343813e-06, "loss": 0.0136, "step": 39620 }, { "epoch": 11.53040442246145, "grad_norm": 0.1039382815361023, "learning_rate": 8.502067475281585e-06, "loss": 0.0132, "step": 39630 }, { "epoch": 11.533313936572592, "grad_norm": 0.09742418676614761, "learning_rate": 8.494617101025365e-06, "loss": 0.0144, "step": 39640 }, { "epoch": 11.536223450683735, "grad_norm": 0.08618821203708649, "learning_rate": 8.487168702932712e-06, "loss": 0.0148, "step": 39650 }, { "epoch": 11.539132964794879, "grad_norm": 0.08493015170097351, "learning_rate": 8.479722283266245e-06, "loss": 0.0148, "step": 39660 }, { "epoch": 11.542042478906023, "grad_norm": 0.07741054147481918, "learning_rate": 8.472277844287994e-06, "loss": 0.0141, "step": 39670 }, { "epoch": 11.544951993017166, "grad_norm": 0.09583154320716858, "learning_rate": 8.464835388259376e-06, "loss": 0.015, "step": 39680 }, { "epoch": 11.54786150712831, "grad_norm": 0.14925795793533325, "learning_rate": 8.45739491744121e-06, "loss": 0.0138, "step": 39690 }, { "epoch": 11.550771021239452, "grad_norm": 0.10317162424325943, "learning_rate": 8.449956434093733e-06, "loss": 0.015, "step": 39700 }, { "epoch": 11.553680535350596, "grad_norm": 0.09326215088367462, "learning_rate": 8.442519940476534e-06, "loss": 0.0147, "step": 39710 }, { "epoch": 11.55659004946174, "grad_norm": 0.07784201204776764, "learning_rate": 8.435085438848651e-06, "loss": 0.0139, "step": 39720 }, { "epoch": 11.559499563572883, "grad_norm": 0.10405431687831879, "learning_rate": 8.427652931468464e-06, "loss": 0.0144, "step": 39730 }, { "epoch": 11.562409077684027, "grad_norm": 0.09093012660741806, "learning_rate": 8.420222420593793e-06, "loss": 0.0122, "step": 39740 }, { "epoch": 11.56531859179517, "grad_norm": 0.09221100062131882, "learning_rate": 8.412793908481812e-06, "loss": 0.0133, "step": 39750 }, { "epoch": 11.568228105906314, "grad_norm": 0.10054229199886322, "learning_rate": 8.405367397389121e-06, "loss": 0.0149, "step": 39760 }, { "epoch": 11.571137620017456, "grad_norm": 0.09542777389287949, "learning_rate": 8.397942889571694e-06, "loss": 0.0136, "step": 39770 }, { "epoch": 11.5740471341286, "grad_norm": 0.10570696741342545, "learning_rate": 8.390520387284899e-06, "loss": 0.0148, "step": 39780 }, { "epoch": 11.576956648239744, "grad_norm": 0.09465997666120529, "learning_rate": 8.383099892783497e-06, "loss": 0.0144, "step": 39790 }, { "epoch": 11.579866162350887, "grad_norm": 0.10178295522928238, "learning_rate": 8.375681408321634e-06, "loss": 0.0148, "step": 39800 }, { "epoch": 11.582775676462031, "grad_norm": 0.09059750288724899, "learning_rate": 8.36826493615286e-06, "loss": 0.0136, "step": 39810 }, { "epoch": 11.585685190573175, "grad_norm": 0.08750323951244354, "learning_rate": 8.360850478530096e-06, "loss": 0.0143, "step": 39820 }, { "epoch": 11.588594704684319, "grad_norm": 0.09284038096666336, "learning_rate": 8.353438037705662e-06, "loss": 0.0139, "step": 39830 }, { "epoch": 11.59150421879546, "grad_norm": 0.09329604357481003, "learning_rate": 8.34602761593126e-06, "loss": 0.0153, "step": 39840 }, { "epoch": 11.594413732906604, "grad_norm": 0.10390728712081909, "learning_rate": 8.33861921545798e-06, "loss": 0.0148, "step": 39850 }, { "epoch": 11.597323247017748, "grad_norm": 0.09063324332237244, "learning_rate": 8.33121283853629e-06, "loss": 0.0148, "step": 39860 }, { "epoch": 11.600232761128892, "grad_norm": 0.10495299845933914, "learning_rate": 8.323808487416075e-06, "loss": 0.0122, "step": 39870 }, { "epoch": 11.603142275240035, "grad_norm": 0.1131146028637886, "learning_rate": 8.316406164346556e-06, "loss": 0.0159, "step": 39880 }, { "epoch": 11.606051789351179, "grad_norm": 0.08886486291885376, "learning_rate": 8.309005871576382e-06, "loss": 0.0128, "step": 39890 }, { "epoch": 11.608961303462323, "grad_norm": 0.1029951199889183, "learning_rate": 8.301607611353552e-06, "loss": 0.0129, "step": 39900 }, { "epoch": 11.611870817573465, "grad_norm": 0.08350777626037598, "learning_rate": 8.294211385925471e-06, "loss": 0.0145, "step": 39910 }, { "epoch": 11.614780331684608, "grad_norm": 0.08639242500066757, "learning_rate": 8.286817197538917e-06, "loss": 0.0134, "step": 39920 }, { "epoch": 11.617689845795752, "grad_norm": 0.09578525274991989, "learning_rate": 8.279425048440046e-06, "loss": 0.0139, "step": 39930 }, { "epoch": 11.620599359906896, "grad_norm": 0.08232537657022476, "learning_rate": 8.272034940874399e-06, "loss": 0.0158, "step": 39940 }, { "epoch": 11.62350887401804, "grad_norm": 0.10689658671617508, "learning_rate": 8.264646877086889e-06, "loss": 0.0156, "step": 39950 }, { "epoch": 11.626418388129183, "grad_norm": 0.12978020310401917, "learning_rate": 8.257260859321827e-06, "loss": 0.0153, "step": 39960 }, { "epoch": 11.629327902240325, "grad_norm": 0.0921076089143753, "learning_rate": 8.249876889822885e-06, "loss": 0.014, "step": 39970 }, { "epoch": 11.632237416351469, "grad_norm": 0.081091970205307, "learning_rate": 8.242494970833118e-06, "loss": 0.0138, "step": 39980 }, { "epoch": 11.635146930462613, "grad_norm": 0.11966104060411453, "learning_rate": 8.235115104594955e-06, "loss": 0.0143, "step": 39990 }, { "epoch": 11.638056444573756, "grad_norm": 0.12092694640159607, "learning_rate": 8.227737293350209e-06, "loss": 0.0137, "step": 40000 }, { "epoch": 11.6409659586849, "grad_norm": 0.08557713031768799, "learning_rate": 8.220361539340054e-06, "loss": 0.0137, "step": 40010 }, { "epoch": 11.643875472796044, "grad_norm": 0.20606933534145355, "learning_rate": 8.212987844805071e-06, "loss": 0.0149, "step": 40020 }, { "epoch": 11.646784986907187, "grad_norm": 0.13242487609386444, "learning_rate": 8.205616211985167e-06, "loss": 0.0132, "step": 40030 }, { "epoch": 11.64969450101833, "grad_norm": 0.08625209331512451, "learning_rate": 8.198246643119675e-06, "loss": 0.0137, "step": 40040 }, { "epoch": 11.652604015129473, "grad_norm": 0.09880979359149933, "learning_rate": 8.190879140447252e-06, "loss": 0.0121, "step": 40050 }, { "epoch": 11.655513529240617, "grad_norm": 0.09613361209630966, "learning_rate": 8.183513706205965e-06, "loss": 0.0138, "step": 40060 }, { "epoch": 11.65842304335176, "grad_norm": 0.08356622606515884, "learning_rate": 8.176150342633236e-06, "loss": 0.0123, "step": 40070 }, { "epoch": 11.661332557462904, "grad_norm": 0.09014326333999634, "learning_rate": 8.168789051965861e-06, "loss": 0.0147, "step": 40080 }, { "epoch": 11.664242071574048, "grad_norm": 0.08260239660739899, "learning_rate": 8.161429836440003e-06, "loss": 0.012, "step": 40090 }, { "epoch": 11.66715158568519, "grad_norm": 0.07658157497644424, "learning_rate": 8.154072698291196e-06, "loss": 0.013, "step": 40100 }, { "epoch": 11.670061099796333, "grad_norm": 0.08686750382184982, "learning_rate": 8.146717639754347e-06, "loss": 0.0134, "step": 40110 }, { "epoch": 11.672970613907477, "grad_norm": 0.09473700076341629, "learning_rate": 8.139364663063734e-06, "loss": 0.0128, "step": 40120 }, { "epoch": 11.67588012801862, "grad_norm": 0.1326497346162796, "learning_rate": 8.132013770452988e-06, "loss": 0.0134, "step": 40130 }, { "epoch": 11.678789642129765, "grad_norm": 0.08199939131736755, "learning_rate": 8.124664964155122e-06, "loss": 0.0136, "step": 40140 }, { "epoch": 11.681699156240908, "grad_norm": 0.07249289751052856, "learning_rate": 8.117318246402506e-06, "loss": 0.0131, "step": 40150 }, { "epoch": 11.684608670352052, "grad_norm": 0.08438624441623688, "learning_rate": 8.109973619426881e-06, "loss": 0.0117, "step": 40160 }, { "epoch": 11.687518184463194, "grad_norm": 0.10521124303340912, "learning_rate": 8.10263108545935e-06, "loss": 0.0144, "step": 40170 }, { "epoch": 11.690427698574338, "grad_norm": 0.09190700203180313, "learning_rate": 8.095290646730375e-06, "loss": 0.0132, "step": 40180 }, { "epoch": 11.693337212685481, "grad_norm": 0.08540832996368408, "learning_rate": 8.087952305469805e-06, "loss": 0.014, "step": 40190 }, { "epoch": 11.696246726796625, "grad_norm": 0.1423284113407135, "learning_rate": 8.08061606390681e-06, "loss": 0.0136, "step": 40200 }, { "epoch": 11.699156240907769, "grad_norm": 0.08525705337524414, "learning_rate": 8.073281924269966e-06, "loss": 0.0147, "step": 40210 }, { "epoch": 11.702065755018912, "grad_norm": 0.10264676064252853, "learning_rate": 8.065949888787181e-06, "loss": 0.014, "step": 40220 }, { "epoch": 11.704975269130056, "grad_norm": 0.08661758154630661, "learning_rate": 8.05861995968574e-06, "loss": 0.0149, "step": 40230 }, { "epoch": 11.707884783241198, "grad_norm": 0.09622340649366379, "learning_rate": 8.051292139192279e-06, "loss": 0.0137, "step": 40240 }, { "epoch": 11.710794297352342, "grad_norm": 0.08817204087972641, "learning_rate": 8.043966429532788e-06, "loss": 0.0142, "step": 40250 }, { "epoch": 11.713703811463485, "grad_norm": 0.09988410025835037, "learning_rate": 8.036642832932646e-06, "loss": 0.0141, "step": 40260 }, { "epoch": 11.71661332557463, "grad_norm": 0.09430195391178131, "learning_rate": 8.029321351616541e-06, "loss": 0.012, "step": 40270 }, { "epoch": 11.719522839685773, "grad_norm": 0.09238729625940323, "learning_rate": 8.022001987808568e-06, "loss": 0.0129, "step": 40280 }, { "epoch": 11.722432353796917, "grad_norm": 0.08273492008447647, "learning_rate": 8.014684743732144e-06, "loss": 0.014, "step": 40290 }, { "epoch": 11.725341867908059, "grad_norm": 0.08237381279468536, "learning_rate": 8.00736962161006e-06, "loss": 0.0126, "step": 40300 }, { "epoch": 11.728251382019202, "grad_norm": 0.09624475985765457, "learning_rate": 8.000056623664456e-06, "loss": 0.0152, "step": 40310 }, { "epoch": 11.731160896130346, "grad_norm": 0.0923820436000824, "learning_rate": 7.992745752116826e-06, "loss": 0.0142, "step": 40320 }, { "epoch": 11.73407041024149, "grad_norm": 0.08647571504116058, "learning_rate": 7.985437009188013e-06, "loss": 0.014, "step": 40330 }, { "epoch": 11.736979924352633, "grad_norm": 0.09780749678611755, "learning_rate": 7.978130397098244e-06, "loss": 0.0131, "step": 40340 }, { "epoch": 11.739889438463777, "grad_norm": 0.11616078019142151, "learning_rate": 7.970825918067043e-06, "loss": 0.0151, "step": 40350 }, { "epoch": 11.74279895257492, "grad_norm": 0.0954664871096611, "learning_rate": 7.96352357431334e-06, "loss": 0.0131, "step": 40360 }, { "epoch": 11.745708466686063, "grad_norm": 0.0903848260641098, "learning_rate": 7.956223368055389e-06, "loss": 0.0141, "step": 40370 }, { "epoch": 11.748617980797206, "grad_norm": 0.07621830701828003, "learning_rate": 7.948925301510793e-06, "loss": 0.0123, "step": 40380 }, { "epoch": 11.75152749490835, "grad_norm": 0.10498376935720444, "learning_rate": 7.941629376896527e-06, "loss": 0.0125, "step": 40390 }, { "epoch": 11.754437009019494, "grad_norm": 0.09775484353303909, "learning_rate": 7.934335596428882e-06, "loss": 0.0143, "step": 40400 }, { "epoch": 11.757346523130638, "grad_norm": 0.07940182089805603, "learning_rate": 7.927043962323535e-06, "loss": 0.0142, "step": 40410 }, { "epoch": 11.760256037241781, "grad_norm": 0.14191007614135742, "learning_rate": 7.91975447679547e-06, "loss": 0.0139, "step": 40420 }, { "epoch": 11.763165551352923, "grad_norm": 0.09486910700798035, "learning_rate": 7.912467142059057e-06, "loss": 0.014, "step": 40430 }, { "epoch": 11.766075065464067, "grad_norm": 0.11662834137678146, "learning_rate": 7.905181960327993e-06, "loss": 0.0161, "step": 40440 }, { "epoch": 11.76898457957521, "grad_norm": 0.08875726163387299, "learning_rate": 7.897898933815319e-06, "loss": 0.0142, "step": 40450 }, { "epoch": 11.771894093686354, "grad_norm": 0.12296207994222641, "learning_rate": 7.89061806473343e-06, "loss": 0.0146, "step": 40460 }, { "epoch": 11.774803607797498, "grad_norm": 0.06233605742454529, "learning_rate": 7.88333935529406e-06, "loss": 0.0141, "step": 40470 }, { "epoch": 11.777713121908642, "grad_norm": 0.10007920116186142, "learning_rate": 7.876062807708286e-06, "loss": 0.0134, "step": 40480 }, { "epoch": 11.780622636019785, "grad_norm": 0.0929766595363617, "learning_rate": 7.868788424186536e-06, "loss": 0.0131, "step": 40490 }, { "epoch": 11.783532150130927, "grad_norm": 0.06573706865310669, "learning_rate": 7.861516206938578e-06, "loss": 0.0131, "step": 40500 }, { "epoch": 11.786441664242071, "grad_norm": 0.06786996871232986, "learning_rate": 7.854246158173512e-06, "loss": 0.0119, "step": 40510 }, { "epoch": 11.789351178353215, "grad_norm": 0.07777401059865952, "learning_rate": 7.846978280099794e-06, "loss": 0.0132, "step": 40520 }, { "epoch": 11.792260692464358, "grad_norm": 0.09261398762464523, "learning_rate": 7.839712574925202e-06, "loss": 0.0123, "step": 40530 }, { "epoch": 11.795170206575502, "grad_norm": 0.09484419226646423, "learning_rate": 7.832449044856885e-06, "loss": 0.0156, "step": 40540 }, { "epoch": 11.798079720686646, "grad_norm": 0.07971522957086563, "learning_rate": 7.825187692101289e-06, "loss": 0.0139, "step": 40550 }, { "epoch": 11.80098923479779, "grad_norm": 0.0943145900964737, "learning_rate": 7.817928518864245e-06, "loss": 0.014, "step": 40560 }, { "epoch": 11.803898748908932, "grad_norm": 0.08274882286787033, "learning_rate": 7.810671527350872e-06, "loss": 0.0129, "step": 40570 }, { "epoch": 11.806808263020075, "grad_norm": 0.08408793061971664, "learning_rate": 7.803416719765677e-06, "loss": 0.0144, "step": 40580 }, { "epoch": 11.809717777131219, "grad_norm": 0.10255514085292816, "learning_rate": 7.796164098312456e-06, "loss": 0.0138, "step": 40590 }, { "epoch": 11.812627291242363, "grad_norm": 0.07898738235235214, "learning_rate": 7.788913665194384e-06, "loss": 0.0134, "step": 40600 }, { "epoch": 11.815536805353506, "grad_norm": 0.08450685441493988, "learning_rate": 7.78166542261394e-06, "loss": 0.013, "step": 40610 }, { "epoch": 11.81844631946465, "grad_norm": 0.09401392936706543, "learning_rate": 7.774419372772952e-06, "loss": 0.0131, "step": 40620 }, { "epoch": 11.821355833575794, "grad_norm": 0.08786585181951523, "learning_rate": 7.76717551787258e-06, "loss": 0.0138, "step": 40630 }, { "epoch": 11.824265347686936, "grad_norm": 0.0673438236117363, "learning_rate": 7.759933860113306e-06, "loss": 0.0142, "step": 40640 }, { "epoch": 11.82717486179808, "grad_norm": 0.09622155874967575, "learning_rate": 7.752694401694968e-06, "loss": 0.0129, "step": 40650 }, { "epoch": 11.830084375909223, "grad_norm": 0.08889530599117279, "learning_rate": 7.745457144816718e-06, "loss": 0.0137, "step": 40660 }, { "epoch": 11.832993890020367, "grad_norm": 0.1960521638393402, "learning_rate": 7.738222091677042e-06, "loss": 0.0131, "step": 40670 }, { "epoch": 11.83590340413151, "grad_norm": 0.0960463434457779, "learning_rate": 7.730989244473759e-06, "loss": 0.0146, "step": 40680 }, { "epoch": 11.838812918242654, "grad_norm": 0.10286245495080948, "learning_rate": 7.723758605404016e-06, "loss": 0.0127, "step": 40690 }, { "epoch": 11.841722432353796, "grad_norm": 0.09792003035545349, "learning_rate": 7.716530176664285e-06, "loss": 0.014, "step": 40700 }, { "epoch": 11.84463194646494, "grad_norm": 0.08508423715829849, "learning_rate": 7.70930396045039e-06, "loss": 0.0154, "step": 40710 }, { "epoch": 11.847541460576084, "grad_norm": 0.096847765147686, "learning_rate": 7.70207995895744e-06, "loss": 0.0132, "step": 40720 }, { "epoch": 11.850450974687227, "grad_norm": 0.09209644794464111, "learning_rate": 7.694858174379918e-06, "loss": 0.0146, "step": 40730 }, { "epoch": 11.853360488798371, "grad_norm": 0.10056664794683456, "learning_rate": 7.68763860891159e-06, "loss": 0.0133, "step": 40740 }, { "epoch": 11.856270002909515, "grad_norm": 0.11061692237854004, "learning_rate": 7.680421264745587e-06, "loss": 0.0134, "step": 40750 }, { "epoch": 11.859179517020657, "grad_norm": 0.12747888267040253, "learning_rate": 7.673206144074344e-06, "loss": 0.0139, "step": 40760 }, { "epoch": 11.8620890311318, "grad_norm": 0.10884598642587662, "learning_rate": 7.66599324908962e-06, "loss": 0.013, "step": 40770 }, { "epoch": 11.864998545242944, "grad_norm": 0.12290549278259277, "learning_rate": 7.658782581982504e-06, "loss": 0.0129, "step": 40780 }, { "epoch": 11.867908059354088, "grad_norm": 0.09661205857992172, "learning_rate": 7.651574144943399e-06, "loss": 0.0137, "step": 40790 }, { "epoch": 11.870817573465231, "grad_norm": 0.14145176112651825, "learning_rate": 7.64436794016205e-06, "loss": 0.0141, "step": 40800 }, { "epoch": 11.873727087576375, "grad_norm": 0.08190681785345078, "learning_rate": 7.63716396982751e-06, "loss": 0.0129, "step": 40810 }, { "epoch": 11.876636601687519, "grad_norm": 0.07990626245737076, "learning_rate": 7.62996223612815e-06, "loss": 0.0132, "step": 40820 }, { "epoch": 11.87954611579866, "grad_norm": 0.07998479902744293, "learning_rate": 7.622762741251666e-06, "loss": 0.0126, "step": 40830 }, { "epoch": 11.882455629909805, "grad_norm": 0.0824938714504242, "learning_rate": 7.615565487385079e-06, "loss": 0.0137, "step": 40840 }, { "epoch": 11.885365144020948, "grad_norm": 0.10518500953912735, "learning_rate": 7.6083704767147145e-06, "loss": 0.0129, "step": 40850 }, { "epoch": 11.888274658132092, "grad_norm": 0.09570855647325516, "learning_rate": 7.601177711426247e-06, "loss": 0.0135, "step": 40860 }, { "epoch": 11.891184172243236, "grad_norm": 0.09554045647382736, "learning_rate": 7.593987193704626e-06, "loss": 0.0154, "step": 40870 }, { "epoch": 11.89409368635438, "grad_norm": 0.09036668390035629, "learning_rate": 7.58679892573416e-06, "loss": 0.0122, "step": 40880 }, { "epoch": 11.897003200465523, "grad_norm": 0.0891854390501976, "learning_rate": 7.579612909698436e-06, "loss": 0.015, "step": 40890 }, { "epoch": 11.899912714576665, "grad_norm": 0.1014237031340599, "learning_rate": 7.572429147780391e-06, "loss": 0.0142, "step": 40900 }, { "epoch": 11.902822228687809, "grad_norm": 0.11025429517030716, "learning_rate": 7.56524764216226e-06, "loss": 0.0145, "step": 40910 }, { "epoch": 11.905731742798952, "grad_norm": 0.0988166481256485, "learning_rate": 7.55806839502559e-06, "loss": 0.0132, "step": 40920 }, { "epoch": 11.908641256910096, "grad_norm": 0.08387865871191025, "learning_rate": 7.550891408551248e-06, "loss": 0.0149, "step": 40930 }, { "epoch": 11.91155077102124, "grad_norm": 0.09168742597103119, "learning_rate": 7.54371668491941e-06, "loss": 0.0143, "step": 40940 }, { "epoch": 11.914460285132384, "grad_norm": 0.06988797336816788, "learning_rate": 7.536544226309581e-06, "loss": 0.013, "step": 40950 }, { "epoch": 11.917369799243527, "grad_norm": 0.11291385442018509, "learning_rate": 7.529374034900544e-06, "loss": 0.0135, "step": 40960 }, { "epoch": 11.92027931335467, "grad_norm": 0.1035306304693222, "learning_rate": 7.5222061128704304e-06, "loss": 0.0134, "step": 40970 }, { "epoch": 11.923188827465813, "grad_norm": 0.09752239286899567, "learning_rate": 7.51504046239666e-06, "loss": 0.0148, "step": 40980 }, { "epoch": 11.926098341576957, "grad_norm": 0.10889366269111633, "learning_rate": 7.50787708565597e-06, "loss": 0.0137, "step": 40990 }, { "epoch": 11.9290078556881, "grad_norm": 0.14295870065689087, "learning_rate": 7.500715984824401e-06, "loss": 0.0141, "step": 41000 }, { "epoch": 11.931917369799244, "grad_norm": 0.11799327284097672, "learning_rate": 7.4935571620773095e-06, "loss": 0.0161, "step": 41010 }, { "epoch": 11.934826883910388, "grad_norm": 0.10258566588163376, "learning_rate": 7.486400619589349e-06, "loss": 0.0143, "step": 41020 }, { "epoch": 11.93773639802153, "grad_norm": 0.12322388589382172, "learning_rate": 7.479246359534509e-06, "loss": 0.0146, "step": 41030 }, { "epoch": 11.940645912132673, "grad_norm": 0.0861552283167839, "learning_rate": 7.472094384086038e-06, "loss": 0.0137, "step": 41040 }, { "epoch": 11.943555426243817, "grad_norm": 0.09515087306499481, "learning_rate": 7.464944695416536e-06, "loss": 0.0133, "step": 41050 }, { "epoch": 11.94646494035496, "grad_norm": 0.08768843859434128, "learning_rate": 7.457797295697885e-06, "loss": 0.0138, "step": 41060 }, { "epoch": 11.949374454466104, "grad_norm": 0.0888480618596077, "learning_rate": 7.4506521871012755e-06, "loss": 0.0143, "step": 41070 }, { "epoch": 11.952283968577248, "grad_norm": 0.099757120013237, "learning_rate": 7.443509371797201e-06, "loss": 0.0127, "step": 41080 }, { "epoch": 11.95519348268839, "grad_norm": 0.10089496523141861, "learning_rate": 7.436368851955456e-06, "loss": 0.0143, "step": 41090 }, { "epoch": 11.958102996799534, "grad_norm": 0.08180245012044907, "learning_rate": 7.4292306297451594e-06, "loss": 0.0129, "step": 41100 }, { "epoch": 11.961012510910678, "grad_norm": 0.11631730943918228, "learning_rate": 7.422094707334693e-06, "loss": 0.0135, "step": 41110 }, { "epoch": 11.963922025021821, "grad_norm": 0.08550785481929779, "learning_rate": 7.414961086891776e-06, "loss": 0.0124, "step": 41120 }, { "epoch": 11.966831539132965, "grad_norm": 0.10386592149734497, "learning_rate": 7.407829770583409e-06, "loss": 0.0144, "step": 41130 }, { "epoch": 11.969741053244109, "grad_norm": 0.09377942979335785, "learning_rate": 7.4007007605759e-06, "loss": 0.0136, "step": 41140 }, { "epoch": 11.972650567355252, "grad_norm": 0.08924377709627151, "learning_rate": 7.39357405903485e-06, "loss": 0.0136, "step": 41150 }, { "epoch": 11.975560081466394, "grad_norm": 0.0962611511349678, "learning_rate": 7.386449668125165e-06, "loss": 0.0147, "step": 41160 }, { "epoch": 11.978469595577538, "grad_norm": 0.0910572037100792, "learning_rate": 7.379327590011041e-06, "loss": 0.0134, "step": 41170 }, { "epoch": 11.981379109688682, "grad_norm": 0.0774419903755188, "learning_rate": 7.372207826855996e-06, "loss": 0.0154, "step": 41180 }, { "epoch": 11.984288623799825, "grad_norm": 0.10280915349721909, "learning_rate": 7.3650903808228e-06, "loss": 0.0149, "step": 41190 }, { "epoch": 11.987198137910969, "grad_norm": 0.08324434608221054, "learning_rate": 7.357975254073565e-06, "loss": 0.0131, "step": 41200 }, { "epoch": 11.990107652022113, "grad_norm": 0.09394155442714691, "learning_rate": 7.35086244876967e-06, "loss": 0.0126, "step": 41210 }, { "epoch": 11.993017166133257, "grad_norm": 0.11579308658838272, "learning_rate": 7.343751967071801e-06, "loss": 0.0144, "step": 41220 }, { "epoch": 11.995926680244398, "grad_norm": 0.08948614448308945, "learning_rate": 7.336643811139933e-06, "loss": 0.0142, "step": 41230 }, { "epoch": 11.998836194355542, "grad_norm": 0.09407734125852585, "learning_rate": 7.329537983133329e-06, "loss": 0.0126, "step": 41240 }, { "epoch": 12.001745708466686, "grad_norm": 0.105417400598526, "learning_rate": 7.3224344852105716e-06, "loss": 0.0138, "step": 41250 }, { "epoch": 12.00465522257783, "grad_norm": 0.11377424001693726, "learning_rate": 7.31533331952949e-06, "loss": 0.0154, "step": 41260 }, { "epoch": 12.007564736688973, "grad_norm": 0.09292905032634735, "learning_rate": 7.308234488247251e-06, "loss": 0.0132, "step": 41270 }, { "epoch": 12.010474250800117, "grad_norm": 0.09045904129743576, "learning_rate": 7.301137993520286e-06, "loss": 0.0142, "step": 41280 }, { "epoch": 12.01338376491126, "grad_norm": 0.09862980991601944, "learning_rate": 7.294043837504321e-06, "loss": 0.0138, "step": 41290 }, { "epoch": 12.016293279022403, "grad_norm": 0.11348935216665268, "learning_rate": 7.286952022354374e-06, "loss": 0.0132, "step": 41300 }, { "epoch": 12.019202793133546, "grad_norm": 0.09238149225711823, "learning_rate": 7.279862550224751e-06, "loss": 0.0153, "step": 41310 }, { "epoch": 12.02211230724469, "grad_norm": 0.08561956882476807, "learning_rate": 7.272775423269042e-06, "loss": 0.0135, "step": 41320 }, { "epoch": 12.025021821355834, "grad_norm": 0.09609067440032959, "learning_rate": 7.265690643640146e-06, "loss": 0.0139, "step": 41330 }, { "epoch": 12.027931335466977, "grad_norm": 0.09047139436006546, "learning_rate": 7.258608213490205e-06, "loss": 0.0131, "step": 41340 }, { "epoch": 12.030840849578121, "grad_norm": 0.12792937457561493, "learning_rate": 7.2515281349706995e-06, "loss": 0.0145, "step": 41350 }, { "epoch": 12.033750363689263, "grad_norm": 0.11939319968223572, "learning_rate": 7.244450410232356e-06, "loss": 0.0152, "step": 41360 }, { "epoch": 12.036659877800407, "grad_norm": 0.07787299901247025, "learning_rate": 7.237375041425206e-06, "loss": 0.0109, "step": 41370 }, { "epoch": 12.03956939191155, "grad_norm": 0.11134544759988785, "learning_rate": 7.23030203069856e-06, "loss": 0.0139, "step": 41380 }, { "epoch": 12.042478906022694, "grad_norm": 0.09481623768806458, "learning_rate": 7.223231380201001e-06, "loss": 0.0136, "step": 41390 }, { "epoch": 12.045388420133838, "grad_norm": 0.10703476518392563, "learning_rate": 7.216163092080429e-06, "loss": 0.0122, "step": 41400 }, { "epoch": 12.048297934244982, "grad_norm": 0.10679551959037781, "learning_rate": 7.209097168483978e-06, "loss": 0.0151, "step": 41410 }, { "epoch": 12.051207448356125, "grad_norm": 0.08268868923187256, "learning_rate": 7.20203361155811e-06, "loss": 0.0135, "step": 41420 }, { "epoch": 12.054116962467267, "grad_norm": 0.0886596292257309, "learning_rate": 7.194972423448526e-06, "loss": 0.0137, "step": 41430 }, { "epoch": 12.057026476578411, "grad_norm": 0.11376915127038956, "learning_rate": 7.187913606300247e-06, "loss": 0.0139, "step": 41440 }, { "epoch": 12.059935990689555, "grad_norm": 0.10014203935861588, "learning_rate": 7.180857162257548e-06, "loss": 0.0143, "step": 41450 }, { "epoch": 12.062845504800698, "grad_norm": 0.0909474790096283, "learning_rate": 7.17380309346399e-06, "loss": 0.0142, "step": 41460 }, { "epoch": 12.065755018911842, "grad_norm": 0.10142254084348679, "learning_rate": 7.166751402062415e-06, "loss": 0.0154, "step": 41470 }, { "epoch": 12.068664533022986, "grad_norm": 0.0903758555650711, "learning_rate": 7.159702090194938e-06, "loss": 0.013, "step": 41480 }, { "epoch": 12.071574047134128, "grad_norm": 0.06854785978794098, "learning_rate": 7.152655160002948e-06, "loss": 0.0126, "step": 41490 }, { "epoch": 12.074483561245271, "grad_norm": 0.08968209475278854, "learning_rate": 7.145610613627131e-06, "loss": 0.0137, "step": 41500 }, { "epoch": 12.077393075356415, "grad_norm": 0.08497679233551025, "learning_rate": 7.138568453207424e-06, "loss": 0.0137, "step": 41510 }, { "epoch": 12.080302589467559, "grad_norm": 0.08566275984048843, "learning_rate": 7.131528680883054e-06, "loss": 0.0133, "step": 41520 }, { "epoch": 12.083212103578703, "grad_norm": 0.1820719987154007, "learning_rate": 7.124491298792515e-06, "loss": 0.0127, "step": 41530 }, { "epoch": 12.086121617689846, "grad_norm": 0.10246255993843079, "learning_rate": 7.117456309073574e-06, "loss": 0.013, "step": 41540 }, { "epoch": 12.08903113180099, "grad_norm": 0.09309858083724976, "learning_rate": 7.110423713863289e-06, "loss": 0.013, "step": 41550 }, { "epoch": 12.091940645912132, "grad_norm": 0.10912955552339554, "learning_rate": 7.103393515297958e-06, "loss": 0.0137, "step": 41560 }, { "epoch": 12.094850160023276, "grad_norm": 0.09311892837285995, "learning_rate": 7.096365715513189e-06, "loss": 0.0144, "step": 41570 }, { "epoch": 12.09775967413442, "grad_norm": 0.09339810907840729, "learning_rate": 7.089340316643822e-06, "loss": 0.0132, "step": 41580 }, { "epoch": 12.100669188245563, "grad_norm": 0.11135534197092056, "learning_rate": 7.082317320824005e-06, "loss": 0.0141, "step": 41590 }, { "epoch": 12.103578702356707, "grad_norm": 0.08415772765874863, "learning_rate": 7.07529673018713e-06, "loss": 0.0132, "step": 41600 }, { "epoch": 12.10648821646785, "grad_norm": 0.10854849964380264, "learning_rate": 7.068278546865869e-06, "loss": 0.0122, "step": 41610 }, { "epoch": 12.109397730578994, "grad_norm": 0.09282209724187851, "learning_rate": 7.061262772992159e-06, "loss": 0.0125, "step": 41620 }, { "epoch": 12.112307244690136, "grad_norm": 0.09090126305818558, "learning_rate": 7.054249410697211e-06, "loss": 0.0133, "step": 41630 }, { "epoch": 12.11521675880128, "grad_norm": 0.09652746468782425, "learning_rate": 7.047238462111489e-06, "loss": 0.0129, "step": 41640 }, { "epoch": 12.118126272912424, "grad_norm": 0.18810635805130005, "learning_rate": 7.040229929364749e-06, "loss": 0.013, "step": 41650 }, { "epoch": 12.121035787023567, "grad_norm": 0.12495067715644836, "learning_rate": 7.033223814585989e-06, "loss": 0.016, "step": 41660 }, { "epoch": 12.123945301134711, "grad_norm": 0.09589889645576477, "learning_rate": 7.0262201199034856e-06, "loss": 0.0146, "step": 41670 }, { "epoch": 12.126854815245855, "grad_norm": 0.13149601221084595, "learning_rate": 7.019218847444773e-06, "loss": 0.0134, "step": 41680 }, { "epoch": 12.129764329356997, "grad_norm": 0.08258403837680817, "learning_rate": 7.012219999336649e-06, "loss": 0.0129, "step": 41690 }, { "epoch": 12.13267384346814, "grad_norm": 0.10569573938846588, "learning_rate": 7.0052235777051945e-06, "loss": 0.0136, "step": 41700 }, { "epoch": 12.135583357579284, "grad_norm": 0.08067364990711212, "learning_rate": 6.998229584675718e-06, "loss": 0.0138, "step": 41710 }, { "epoch": 12.138492871690428, "grad_norm": 0.0966428741812706, "learning_rate": 6.991238022372828e-06, "loss": 0.0135, "step": 41720 }, { "epoch": 12.141402385801571, "grad_norm": 0.15219855308532715, "learning_rate": 6.984248892920357e-06, "loss": 0.0125, "step": 41730 }, { "epoch": 12.144311899912715, "grad_norm": 0.10126567631959915, "learning_rate": 6.977262198441435e-06, "loss": 0.0132, "step": 41740 }, { "epoch": 12.147221414023859, "grad_norm": 0.11037992686033249, "learning_rate": 6.970277941058429e-06, "loss": 0.0122, "step": 41750 }, { "epoch": 12.150130928135, "grad_norm": 0.08930518478155136, "learning_rate": 6.963296122892973e-06, "loss": 0.0131, "step": 41760 }, { "epoch": 12.153040442246144, "grad_norm": 0.08850938081741333, "learning_rate": 6.956316746065956e-06, "loss": 0.0141, "step": 41770 }, { "epoch": 12.155949956357288, "grad_norm": 0.08561238646507263, "learning_rate": 6.949339812697531e-06, "loss": 0.0132, "step": 41780 }, { "epoch": 12.158859470468432, "grad_norm": 0.09585575759410858, "learning_rate": 6.942365324907106e-06, "loss": 0.0139, "step": 41790 }, { "epoch": 12.161768984579576, "grad_norm": 0.08807472139596939, "learning_rate": 6.93539328481334e-06, "loss": 0.0123, "step": 41800 }, { "epoch": 12.16467849869072, "grad_norm": 0.11291614174842834, "learning_rate": 6.9284236945341654e-06, "loss": 0.0142, "step": 41810 }, { "epoch": 12.167588012801861, "grad_norm": 0.09196051210165024, "learning_rate": 6.921456556186756e-06, "loss": 0.0139, "step": 41820 }, { "epoch": 12.170497526913005, "grad_norm": 0.10357771068811417, "learning_rate": 6.9144918718875434e-06, "loss": 0.0118, "step": 41830 }, { "epoch": 12.173407041024149, "grad_norm": 0.09561555832624435, "learning_rate": 6.907529643752213e-06, "loss": 0.0136, "step": 41840 }, { "epoch": 12.176316555135292, "grad_norm": 0.11043541878461838, "learning_rate": 6.900569873895709e-06, "loss": 0.0133, "step": 41850 }, { "epoch": 12.179226069246436, "grad_norm": 0.10018054395914078, "learning_rate": 6.893612564432217e-06, "loss": 0.0129, "step": 41860 }, { "epoch": 12.18213558335758, "grad_norm": 0.07830996811389923, "learning_rate": 6.8866577174752044e-06, "loss": 0.0145, "step": 41870 }, { "epoch": 12.185045097468723, "grad_norm": 0.13348819315433502, "learning_rate": 6.879705335137344e-06, "loss": 0.0129, "step": 41880 }, { "epoch": 12.187954611579865, "grad_norm": 0.11415565013885498, "learning_rate": 6.87275541953061e-06, "loss": 0.0125, "step": 41890 }, { "epoch": 12.190864125691009, "grad_norm": 0.10629294067621231, "learning_rate": 6.865807972766183e-06, "loss": 0.0131, "step": 41900 }, { "epoch": 12.193773639802153, "grad_norm": 0.07622118294239044, "learning_rate": 6.858862996954527e-06, "loss": 0.0124, "step": 41910 }, { "epoch": 12.196683153913296, "grad_norm": 0.09192676842212677, "learning_rate": 6.851920494205339e-06, "loss": 0.0126, "step": 41920 }, { "epoch": 12.19959266802444, "grad_norm": 0.08104871213436127, "learning_rate": 6.8449804666275665e-06, "loss": 0.0128, "step": 41930 }, { "epoch": 12.202502182135584, "grad_norm": 0.09730176627635956, "learning_rate": 6.838042916329407e-06, "loss": 0.0139, "step": 41940 }, { "epoch": 12.205411696246728, "grad_norm": 0.09710240364074707, "learning_rate": 6.8311078454183e-06, "loss": 0.0128, "step": 41950 }, { "epoch": 12.20832121035787, "grad_norm": 0.08054017275571823, "learning_rate": 6.824175256000949e-06, "loss": 0.0138, "step": 41960 }, { "epoch": 12.211230724469013, "grad_norm": 0.07155817747116089, "learning_rate": 6.817245150183284e-06, "loss": 0.0124, "step": 41970 }, { "epoch": 12.214140238580157, "grad_norm": 0.10688310861587524, "learning_rate": 6.81031753007049e-06, "loss": 0.0135, "step": 41980 }, { "epoch": 12.2170497526913, "grad_norm": 0.09836956858634949, "learning_rate": 6.8033923977669945e-06, "loss": 0.0142, "step": 41990 }, { "epoch": 12.219959266802444, "grad_norm": 0.0920526385307312, "learning_rate": 6.79646975537647e-06, "loss": 0.0135, "step": 42000 }, { "epoch": 12.222868780913588, "grad_norm": 0.10715300589799881, "learning_rate": 6.7895496050018264e-06, "loss": 0.0131, "step": 42010 }, { "epoch": 12.22577829502473, "grad_norm": 0.08548543602228165, "learning_rate": 6.782631948745238e-06, "loss": 0.014, "step": 42020 }, { "epoch": 12.228687809135874, "grad_norm": 0.09103671461343765, "learning_rate": 6.775716788708087e-06, "loss": 0.0136, "step": 42030 }, { "epoch": 12.231597323247017, "grad_norm": 0.11288245022296906, "learning_rate": 6.768804126991036e-06, "loss": 0.0148, "step": 42040 }, { "epoch": 12.234506837358161, "grad_norm": 0.08884479850530624, "learning_rate": 6.761893965693949e-06, "loss": 0.0114, "step": 42050 }, { "epoch": 12.237416351469305, "grad_norm": 0.08744005113840103, "learning_rate": 6.754986306915966e-06, "loss": 0.0136, "step": 42060 }, { "epoch": 12.240325865580449, "grad_norm": 0.07872378081083298, "learning_rate": 6.748081152755445e-06, "loss": 0.0129, "step": 42070 }, { "epoch": 12.243235379691592, "grad_norm": 0.10129423439502716, "learning_rate": 6.741178505309992e-06, "loss": 0.0141, "step": 42080 }, { "epoch": 12.246144893802734, "grad_norm": 0.07754047214984894, "learning_rate": 6.734278366676447e-06, "loss": 0.0126, "step": 42090 }, { "epoch": 12.249054407913878, "grad_norm": 0.13950707018375397, "learning_rate": 6.727380738950884e-06, "loss": 0.0145, "step": 42100 }, { "epoch": 12.251963922025022, "grad_norm": 0.11444208770990372, "learning_rate": 6.720485624228633e-06, "loss": 0.0134, "step": 42110 }, { "epoch": 12.254873436136165, "grad_norm": 0.09897226095199585, "learning_rate": 6.713593024604242e-06, "loss": 0.0122, "step": 42120 }, { "epoch": 12.257782950247309, "grad_norm": 0.10216664522886276, "learning_rate": 6.706702942171498e-06, "loss": 0.0148, "step": 42130 }, { "epoch": 12.260692464358453, "grad_norm": 0.07656121999025345, "learning_rate": 6.699815379023432e-06, "loss": 0.0129, "step": 42140 }, { "epoch": 12.263601978469596, "grad_norm": 0.08347392827272415, "learning_rate": 6.692930337252297e-06, "loss": 0.0118, "step": 42150 }, { "epoch": 12.266511492580738, "grad_norm": 0.07953644543886185, "learning_rate": 6.686047818949586e-06, "loss": 0.0134, "step": 42160 }, { "epoch": 12.269421006691882, "grad_norm": 0.09643151611089706, "learning_rate": 6.679167826206041e-06, "loss": 0.0145, "step": 42170 }, { "epoch": 12.272330520803026, "grad_norm": 0.13268496096134186, "learning_rate": 6.6722903611116005e-06, "loss": 0.0122, "step": 42180 }, { "epoch": 12.27524003491417, "grad_norm": 0.0907733365893364, "learning_rate": 6.66541542575548e-06, "loss": 0.0134, "step": 42190 }, { "epoch": 12.278149549025313, "grad_norm": 0.12686558067798615, "learning_rate": 6.658543022226078e-06, "loss": 0.0131, "step": 42200 }, { "epoch": 12.281059063136457, "grad_norm": 0.08508040010929108, "learning_rate": 6.651673152611069e-06, "loss": 0.0107, "step": 42210 }, { "epoch": 12.283968577247599, "grad_norm": 0.06993691623210907, "learning_rate": 6.644805818997332e-06, "loss": 0.0117, "step": 42220 }, { "epoch": 12.286878091358743, "grad_norm": 0.10084322839975357, "learning_rate": 6.6379410234709815e-06, "loss": 0.0152, "step": 42230 }, { "epoch": 12.289787605469886, "grad_norm": 0.09634705632925034, "learning_rate": 6.63107876811736e-06, "loss": 0.0127, "step": 42240 }, { "epoch": 12.29269711958103, "grad_norm": 0.07779264450073242, "learning_rate": 6.624219055021034e-06, "loss": 0.0137, "step": 42250 }, { "epoch": 12.295606633692174, "grad_norm": 0.09173823148012161, "learning_rate": 6.617361886265821e-06, "loss": 0.0134, "step": 42260 }, { "epoch": 12.298516147803317, "grad_norm": 0.12143653631210327, "learning_rate": 6.610507263934726e-06, "loss": 0.0131, "step": 42270 }, { "epoch": 12.301425661914461, "grad_norm": 0.08716940134763718, "learning_rate": 6.603655190110017e-06, "loss": 0.0138, "step": 42280 }, { "epoch": 12.304335176025603, "grad_norm": 0.09127604961395264, "learning_rate": 6.596805666873167e-06, "loss": 0.0123, "step": 42290 }, { "epoch": 12.307244690136747, "grad_norm": 0.09228435903787613, "learning_rate": 6.589958696304885e-06, "loss": 0.0119, "step": 42300 }, { "epoch": 12.31015420424789, "grad_norm": 0.10782895237207413, "learning_rate": 6.583114280485095e-06, "loss": 0.0137, "step": 42310 }, { "epoch": 12.313063718359034, "grad_norm": 0.09590618312358856, "learning_rate": 6.576272421492951e-06, "loss": 0.0134, "step": 42320 }, { "epoch": 12.315973232470178, "grad_norm": 0.08928563445806503, "learning_rate": 6.569433121406826e-06, "loss": 0.0131, "step": 42330 }, { "epoch": 12.318882746581322, "grad_norm": 0.07802315801382065, "learning_rate": 6.5625963823043324e-06, "loss": 0.0149, "step": 42340 }, { "epoch": 12.321792260692465, "grad_norm": 0.09090308099985123, "learning_rate": 6.5557622062622705e-06, "loss": 0.0123, "step": 42350 }, { "epoch": 12.324701774803607, "grad_norm": 0.08031400293111801, "learning_rate": 6.548930595356699e-06, "loss": 0.0147, "step": 42360 }, { "epoch": 12.32761128891475, "grad_norm": 0.10122358798980713, "learning_rate": 6.542101551662874e-06, "loss": 0.0145, "step": 42370 }, { "epoch": 12.330520803025895, "grad_norm": 0.08444145321846008, "learning_rate": 6.5352750772552805e-06, "loss": 0.0129, "step": 42380 }, { "epoch": 12.333430317137038, "grad_norm": 0.11234021186828613, "learning_rate": 6.528451174207622e-06, "loss": 0.0139, "step": 42390 }, { "epoch": 12.336339831248182, "grad_norm": 0.08999954909086227, "learning_rate": 6.521629844592811e-06, "loss": 0.0143, "step": 42400 }, { "epoch": 12.339249345359326, "grad_norm": 0.08991101384162903, "learning_rate": 6.5148110904830045e-06, "loss": 0.0136, "step": 42410 }, { "epoch": 12.342158859470468, "grad_norm": 0.1468673050403595, "learning_rate": 6.507994913949542e-06, "loss": 0.0138, "step": 42420 }, { "epoch": 12.345068373581611, "grad_norm": 0.11368099600076675, "learning_rate": 6.50118131706301e-06, "loss": 0.0128, "step": 42430 }, { "epoch": 12.347977887692755, "grad_norm": 0.10203392058610916, "learning_rate": 6.494370301893197e-06, "loss": 0.0131, "step": 42440 }, { "epoch": 12.350887401803899, "grad_norm": 0.10264330357313156, "learning_rate": 6.487561870509106e-06, "loss": 0.012, "step": 42450 }, { "epoch": 12.353796915915042, "grad_norm": 0.11911541223526001, "learning_rate": 6.480756024978963e-06, "loss": 0.0131, "step": 42460 }, { "epoch": 12.356706430026186, "grad_norm": 0.09877748787403107, "learning_rate": 6.473952767370201e-06, "loss": 0.0135, "step": 42470 }, { "epoch": 12.35961594413733, "grad_norm": 0.1147301197052002, "learning_rate": 6.467152099749466e-06, "loss": 0.0129, "step": 42480 }, { "epoch": 12.362525458248472, "grad_norm": 0.07914730161428452, "learning_rate": 6.460354024182636e-06, "loss": 0.0124, "step": 42490 }, { "epoch": 12.365434972359616, "grad_norm": 0.08288632333278656, "learning_rate": 6.453558542734767e-06, "loss": 0.0139, "step": 42500 }, { "epoch": 12.36834448647076, "grad_norm": 0.08845669031143188, "learning_rate": 6.4467656574701635e-06, "loss": 0.0127, "step": 42510 }, { "epoch": 12.371254000581903, "grad_norm": 0.14476169645786285, "learning_rate": 6.439975370452318e-06, "loss": 0.0143, "step": 42520 }, { "epoch": 12.374163514693047, "grad_norm": 0.12362048774957657, "learning_rate": 6.433187683743939e-06, "loss": 0.0144, "step": 42530 }, { "epoch": 12.37707302880419, "grad_norm": 0.17326776683330536, "learning_rate": 6.426402599406949e-06, "loss": 0.0131, "step": 42540 }, { "epoch": 12.379982542915332, "grad_norm": 0.09449182450771332, "learning_rate": 6.419620119502472e-06, "loss": 0.013, "step": 42550 }, { "epoch": 12.382892057026476, "grad_norm": 0.07714490592479706, "learning_rate": 6.41284024609086e-06, "loss": 0.0129, "step": 42560 }, { "epoch": 12.38580157113762, "grad_norm": 0.09968701004981995, "learning_rate": 6.4060629812316405e-06, "loss": 0.0131, "step": 42570 }, { "epoch": 12.388711085248763, "grad_norm": 0.0859520211815834, "learning_rate": 6.399288326983588e-06, "loss": 0.0118, "step": 42580 }, { "epoch": 12.391620599359907, "grad_norm": 0.09723910689353943, "learning_rate": 6.392516285404644e-06, "loss": 0.013, "step": 42590 }, { "epoch": 12.39453011347105, "grad_norm": 0.08895329385995865, "learning_rate": 6.385746858551989e-06, "loss": 0.0118, "step": 42600 }, { "epoch": 12.397439627582195, "grad_norm": 0.09987135231494904, "learning_rate": 6.378980048481993e-06, "loss": 0.0143, "step": 42610 }, { "epoch": 12.400349141693336, "grad_norm": 0.09602128714323044, "learning_rate": 6.372215857250232e-06, "loss": 0.0116, "step": 42620 }, { "epoch": 12.40325865580448, "grad_norm": 0.10816957801580429, "learning_rate": 6.36545428691149e-06, "loss": 0.0147, "step": 42630 }, { "epoch": 12.406168169915624, "grad_norm": 0.10676535218954086, "learning_rate": 6.3586953395197554e-06, "loss": 0.0143, "step": 42640 }, { "epoch": 12.409077684026768, "grad_norm": 0.09502795338630676, "learning_rate": 6.351939017128209e-06, "loss": 0.0126, "step": 42650 }, { "epoch": 12.411987198137911, "grad_norm": 0.09177447855472565, "learning_rate": 6.345185321789259e-06, "loss": 0.0136, "step": 42660 }, { "epoch": 12.414896712249055, "grad_norm": 0.14646582305431366, "learning_rate": 6.338434255554487e-06, "loss": 0.0147, "step": 42670 }, { "epoch": 12.417806226360199, "grad_norm": 0.07953815907239914, "learning_rate": 6.331685820474695e-06, "loss": 0.0134, "step": 42680 }, { "epoch": 12.42071574047134, "grad_norm": 0.09862860292196274, "learning_rate": 6.3249400185998775e-06, "loss": 0.0127, "step": 42690 }, { "epoch": 12.423625254582484, "grad_norm": 0.08682037144899368, "learning_rate": 6.318196851979225e-06, "loss": 0.0118, "step": 42700 }, { "epoch": 12.426534768693628, "grad_norm": 0.07968766987323761, "learning_rate": 6.311456322661148e-06, "loss": 0.0131, "step": 42710 }, { "epoch": 12.429444282804772, "grad_norm": 0.1072552278637886, "learning_rate": 6.3047184326932235e-06, "loss": 0.0131, "step": 42720 }, { "epoch": 12.432353796915915, "grad_norm": 0.0962810218334198, "learning_rate": 6.297983184122264e-06, "loss": 0.0129, "step": 42730 }, { "epoch": 12.43526331102706, "grad_norm": 0.10312101244926453, "learning_rate": 6.291250578994239e-06, "loss": 0.0142, "step": 42740 }, { "epoch": 12.438172825138201, "grad_norm": 0.06876495480537415, "learning_rate": 6.284520619354353e-06, "loss": 0.0127, "step": 42750 }, { "epoch": 12.441082339249345, "grad_norm": 0.12864181399345398, "learning_rate": 6.277793307246982e-06, "loss": 0.0131, "step": 42760 }, { "epoch": 12.443991853360489, "grad_norm": 0.08553463220596313, "learning_rate": 6.271068644715709e-06, "loss": 0.0119, "step": 42770 }, { "epoch": 12.446901367471632, "grad_norm": 0.0892707034945488, "learning_rate": 6.264346633803307e-06, "loss": 0.0122, "step": 42780 }, { "epoch": 12.449810881582776, "grad_norm": 0.0696263536810875, "learning_rate": 6.2576272765517465e-06, "loss": 0.0128, "step": 42790 }, { "epoch": 12.45272039569392, "grad_norm": 0.10620126873254776, "learning_rate": 6.250910575002186e-06, "loss": 0.0148, "step": 42800 }, { "epoch": 12.455629909805063, "grad_norm": 0.16430769860744476, "learning_rate": 6.244196531194992e-06, "loss": 0.0112, "step": 42810 }, { "epoch": 12.458539423916205, "grad_norm": 0.09429623186588287, "learning_rate": 6.237485147169707e-06, "loss": 0.0153, "step": 42820 }, { "epoch": 12.461448938027349, "grad_norm": 0.08527855575084686, "learning_rate": 6.230776424965073e-06, "loss": 0.0134, "step": 42830 }, { "epoch": 12.464358452138493, "grad_norm": 0.0907105952501297, "learning_rate": 6.224070366619024e-06, "loss": 0.0143, "step": 42840 }, { "epoch": 12.467267966249636, "grad_norm": 0.06848391145467758, "learning_rate": 6.217366974168678e-06, "loss": 0.0124, "step": 42850 }, { "epoch": 12.47017748036078, "grad_norm": 0.1117527186870575, "learning_rate": 6.210666249650362e-06, "loss": 0.0128, "step": 42860 }, { "epoch": 12.473086994471924, "grad_norm": 0.09193270653486252, "learning_rate": 6.203968195099561e-06, "loss": 0.0114, "step": 42870 }, { "epoch": 12.475996508583066, "grad_norm": 0.09016552567481995, "learning_rate": 6.197272812550987e-06, "loss": 0.0135, "step": 42880 }, { "epoch": 12.47890602269421, "grad_norm": 0.09789007157087326, "learning_rate": 6.1905801040385005e-06, "loss": 0.0125, "step": 42890 }, { "epoch": 12.481815536805353, "grad_norm": 0.09310886263847351, "learning_rate": 6.183890071595183e-06, "loss": 0.0116, "step": 42900 }, { "epoch": 12.484725050916497, "grad_norm": 0.10270377993583679, "learning_rate": 6.177202717253286e-06, "loss": 0.0144, "step": 42910 }, { "epoch": 12.48763456502764, "grad_norm": 0.08947062492370605, "learning_rate": 6.170518043044251e-06, "loss": 0.0143, "step": 42920 }, { "epoch": 12.490544079138784, "grad_norm": 0.15187117457389832, "learning_rate": 6.1638360509987055e-06, "loss": 0.0139, "step": 42930 }, { "epoch": 12.493453593249928, "grad_norm": 0.10407720506191254, "learning_rate": 6.1571567431464634e-06, "loss": 0.0117, "step": 42940 }, { "epoch": 12.49636310736107, "grad_norm": 0.0947345420718193, "learning_rate": 6.150480121516515e-06, "loss": 0.0113, "step": 42950 }, { "epoch": 12.499272621472214, "grad_norm": 0.10774438083171844, "learning_rate": 6.143806188137053e-06, "loss": 0.0126, "step": 42960 }, { "epoch": 12.502182135583357, "grad_norm": 0.13094985485076904, "learning_rate": 6.1371349450354366e-06, "loss": 0.0127, "step": 42970 }, { "epoch": 12.505091649694501, "grad_norm": 0.11143464595079422, "learning_rate": 6.130466394238213e-06, "loss": 0.0131, "step": 42980 }, { "epoch": 12.508001163805645, "grad_norm": 0.07427125424146652, "learning_rate": 6.123800537771113e-06, "loss": 0.0127, "step": 42990 }, { "epoch": 12.510910677916788, "grad_norm": 0.09562034159898758, "learning_rate": 6.117137377659048e-06, "loss": 0.0121, "step": 43000 }, { "epoch": 12.513820192027932, "grad_norm": 0.07485993206501007, "learning_rate": 6.110476915926107e-06, "loss": 0.0139, "step": 43010 }, { "epoch": 12.516729706139074, "grad_norm": 0.08827158063650131, "learning_rate": 6.1038191545955604e-06, "loss": 0.0126, "step": 43020 }, { "epoch": 12.519639220250218, "grad_norm": 0.096913181245327, "learning_rate": 6.097164095689876e-06, "loss": 0.0116, "step": 43030 }, { "epoch": 12.522548734361362, "grad_norm": 0.09201185405254364, "learning_rate": 6.090511741230663e-06, "loss": 0.0133, "step": 43040 }, { "epoch": 12.525458248472505, "grad_norm": 0.09249391406774521, "learning_rate": 6.0838620932387506e-06, "loss": 0.0138, "step": 43050 }, { "epoch": 12.528367762583649, "grad_norm": 0.08539712429046631, "learning_rate": 6.077215153734109e-06, "loss": 0.0148, "step": 43060 }, { "epoch": 12.531277276694793, "grad_norm": 0.07465074211359024, "learning_rate": 6.070570924735916e-06, "loss": 0.0144, "step": 43070 }, { "epoch": 12.534186790805936, "grad_norm": 0.08125561475753784, "learning_rate": 6.063929408262506e-06, "loss": 0.012, "step": 43080 }, { "epoch": 12.537096304917078, "grad_norm": 0.09427566826343536, "learning_rate": 6.0572906063313995e-06, "loss": 0.0147, "step": 43090 }, { "epoch": 12.540005819028222, "grad_norm": 0.08066245913505554, "learning_rate": 6.050654520959289e-06, "loss": 0.0134, "step": 43100 }, { "epoch": 12.542915333139366, "grad_norm": 0.09289716929197311, "learning_rate": 6.044021154162036e-06, "loss": 0.012, "step": 43110 }, { "epoch": 12.54582484725051, "grad_norm": 0.11108581721782684, "learning_rate": 6.037390507954693e-06, "loss": 0.0126, "step": 43120 }, { "epoch": 12.548734361361653, "grad_norm": 0.11933474987745285, "learning_rate": 6.0307625843514676e-06, "loss": 0.0137, "step": 43130 }, { "epoch": 12.551643875472797, "grad_norm": 0.10216743499040604, "learning_rate": 6.024137385365751e-06, "loss": 0.0134, "step": 43140 }, { "epoch": 12.554553389583939, "grad_norm": 0.08709380030632019, "learning_rate": 6.017514913010106e-06, "loss": 0.0133, "step": 43150 }, { "epoch": 12.557462903695082, "grad_norm": 0.07294430583715439, "learning_rate": 6.010895169296258e-06, "loss": 0.0119, "step": 43160 }, { "epoch": 12.560372417806226, "grad_norm": 0.08676115423440933, "learning_rate": 6.004278156235109e-06, "loss": 0.0129, "step": 43170 }, { "epoch": 12.56328193191737, "grad_norm": 0.10843116044998169, "learning_rate": 5.99766387583675e-06, "loss": 0.013, "step": 43180 }, { "epoch": 12.566191446028514, "grad_norm": 0.08801031112670898, "learning_rate": 5.9910523301103995e-06, "loss": 0.0137, "step": 43190 }, { "epoch": 12.569100960139657, "grad_norm": 0.08170392364263535, "learning_rate": 5.984443521064497e-06, "loss": 0.0119, "step": 43200 }, { "epoch": 12.5720104742508, "grad_norm": 0.110316202044487, "learning_rate": 5.977837450706599e-06, "loss": 0.0125, "step": 43210 }, { "epoch": 12.574919988361943, "grad_norm": 0.08897393196821213, "learning_rate": 5.971234121043472e-06, "loss": 0.0142, "step": 43220 }, { "epoch": 12.577829502473087, "grad_norm": 0.09123928844928741, "learning_rate": 5.964633534081028e-06, "loss": 0.0141, "step": 43230 }, { "epoch": 12.58073901658423, "grad_norm": 0.09801651537418365, "learning_rate": 5.958035691824353e-06, "loss": 0.0122, "step": 43240 }, { "epoch": 12.583648530695374, "grad_norm": 0.12021128088235855, "learning_rate": 5.9514405962776944e-06, "loss": 0.0151, "step": 43250 }, { "epoch": 12.586558044806518, "grad_norm": 0.08644226938486099, "learning_rate": 5.944848249444465e-06, "loss": 0.0121, "step": 43260 }, { "epoch": 12.589467558917661, "grad_norm": 0.09788432717323303, "learning_rate": 5.938258653327254e-06, "loss": 0.0138, "step": 43270 }, { "epoch": 12.592377073028803, "grad_norm": 0.09356756508350372, "learning_rate": 5.931671809927803e-06, "loss": 0.0144, "step": 43280 }, { "epoch": 12.595286587139947, "grad_norm": 0.08926738798618317, "learning_rate": 5.925087721247022e-06, "loss": 0.0139, "step": 43290 }, { "epoch": 12.59819610125109, "grad_norm": 0.10992760956287384, "learning_rate": 5.91850638928498e-06, "loss": 0.0156, "step": 43300 }, { "epoch": 12.601105615362235, "grad_norm": 0.07932299375534058, "learning_rate": 5.911927816040915e-06, "loss": 0.0132, "step": 43310 }, { "epoch": 12.604015129473378, "grad_norm": 0.10516732931137085, "learning_rate": 5.905352003513219e-06, "loss": 0.0141, "step": 43320 }, { "epoch": 12.606924643584522, "grad_norm": 0.08104826509952545, "learning_rate": 5.898778953699463e-06, "loss": 0.0128, "step": 43330 }, { "epoch": 12.609834157695666, "grad_norm": 0.10309336334466934, "learning_rate": 5.892208668596346e-06, "loss": 0.0124, "step": 43340 }, { "epoch": 12.612743671806808, "grad_norm": 0.09432721138000488, "learning_rate": 5.88564115019977e-06, "loss": 0.0135, "step": 43350 }, { "epoch": 12.615653185917951, "grad_norm": 0.110113225877285, "learning_rate": 5.879076400504751e-06, "loss": 0.0131, "step": 43360 }, { "epoch": 12.618562700029095, "grad_norm": 0.11901184916496277, "learning_rate": 5.872514421505502e-06, "loss": 0.0129, "step": 43370 }, { "epoch": 12.621472214140239, "grad_norm": 0.1166120395064354, "learning_rate": 5.865955215195376e-06, "loss": 0.0124, "step": 43380 }, { "epoch": 12.624381728251382, "grad_norm": 0.10612849146127701, "learning_rate": 5.859398783566884e-06, "loss": 0.0126, "step": 43390 }, { "epoch": 12.627291242362526, "grad_norm": 0.09254978597164154, "learning_rate": 5.852845128611698e-06, "loss": 0.0135, "step": 43400 }, { "epoch": 12.63020075647367, "grad_norm": 0.08650772273540497, "learning_rate": 5.84629425232064e-06, "loss": 0.0127, "step": 43410 }, { "epoch": 12.633110270584812, "grad_norm": 0.11293607950210571, "learning_rate": 5.839746156683709e-06, "loss": 0.013, "step": 43420 }, { "epoch": 12.636019784695955, "grad_norm": 0.1365441232919693, "learning_rate": 5.833200843690023e-06, "loss": 0.014, "step": 43430 }, { "epoch": 12.6389292988071, "grad_norm": 0.08672510832548141, "learning_rate": 5.82665831532789e-06, "loss": 0.0124, "step": 43440 }, { "epoch": 12.641838812918243, "grad_norm": 0.07865484058856964, "learning_rate": 5.820118573584754e-06, "loss": 0.0116, "step": 43450 }, { "epoch": 12.644748327029387, "grad_norm": 0.14570216834545135, "learning_rate": 5.8135816204472134e-06, "loss": 0.0136, "step": 43460 }, { "epoch": 12.64765784114053, "grad_norm": 0.08072727918624878, "learning_rate": 5.807047457901024e-06, "loss": 0.0124, "step": 43470 }, { "epoch": 12.650567355251672, "grad_norm": 0.1143265888094902, "learning_rate": 5.80051608793109e-06, "loss": 0.0134, "step": 43480 }, { "epoch": 12.653476869362816, "grad_norm": 0.11280538141727448, "learning_rate": 5.793987512521465e-06, "loss": 0.0137, "step": 43490 }, { "epoch": 12.65638638347396, "grad_norm": 0.09785706549882889, "learning_rate": 5.787461733655376e-06, "loss": 0.0123, "step": 43500 }, { "epoch": 12.659295897585103, "grad_norm": 0.0861981064081192, "learning_rate": 5.780938753315156e-06, "loss": 0.0137, "step": 43510 }, { "epoch": 12.662205411696247, "grad_norm": 0.09711551666259766, "learning_rate": 5.774418573482333e-06, "loss": 0.0122, "step": 43520 }, { "epoch": 12.66511492580739, "grad_norm": 0.0979224219918251, "learning_rate": 5.767901196137561e-06, "loss": 0.014, "step": 43530 }, { "epoch": 12.668024439918533, "grad_norm": 0.10264666378498077, "learning_rate": 5.761386623260646e-06, "loss": 0.0122, "step": 43540 }, { "epoch": 12.670933954029676, "grad_norm": 0.06645476073026657, "learning_rate": 5.754874856830544e-06, "loss": 0.0138, "step": 43550 }, { "epoch": 12.67384346814082, "grad_norm": 0.09328227490186691, "learning_rate": 5.74836589882535e-06, "loss": 0.0132, "step": 43560 }, { "epoch": 12.676752982251964, "grad_norm": 0.09186181426048279, "learning_rate": 5.741859751222333e-06, "loss": 0.0147, "step": 43570 }, { "epoch": 12.679662496363107, "grad_norm": 0.10458057373762131, "learning_rate": 5.735356415997865e-06, "loss": 0.0125, "step": 43580 }, { "epoch": 12.682572010474251, "grad_norm": 0.09312666207551956, "learning_rate": 5.728855895127506e-06, "loss": 0.0121, "step": 43590 }, { "epoch": 12.685481524585395, "grad_norm": 0.09856464713811874, "learning_rate": 5.722358190585935e-06, "loss": 0.0143, "step": 43600 }, { "epoch": 12.688391038696537, "grad_norm": 0.09291768819093704, "learning_rate": 5.715863304346982e-06, "loss": 0.0148, "step": 43610 }, { "epoch": 12.69130055280768, "grad_norm": 0.08511671423912048, "learning_rate": 5.709371238383625e-06, "loss": 0.0119, "step": 43620 }, { "epoch": 12.694210066918824, "grad_norm": 0.08641469478607178, "learning_rate": 5.702881994667981e-06, "loss": 0.0135, "step": 43630 }, { "epoch": 12.697119581029968, "grad_norm": 0.09798933565616608, "learning_rate": 5.696395575171305e-06, "loss": 0.0146, "step": 43640 }, { "epoch": 12.700029095141112, "grad_norm": 0.12425646185874939, "learning_rate": 5.6899119818640165e-06, "loss": 0.0146, "step": 43650 }, { "epoch": 12.702938609252255, "grad_norm": 0.1108761802315712, "learning_rate": 5.683431216715639e-06, "loss": 0.0129, "step": 43660 }, { "epoch": 12.705848123363399, "grad_norm": 0.09585796296596527, "learning_rate": 5.676953281694873e-06, "loss": 0.0144, "step": 43670 }, { "epoch": 12.708757637474541, "grad_norm": 0.08958464860916138, "learning_rate": 5.670478178769541e-06, "loss": 0.0128, "step": 43680 }, { "epoch": 12.711667151585685, "grad_norm": 0.07304947078227997, "learning_rate": 5.664005909906605e-06, "loss": 0.0129, "step": 43690 }, { "epoch": 12.714576665696828, "grad_norm": 0.13333235681056976, "learning_rate": 5.657536477072171e-06, "loss": 0.0147, "step": 43700 }, { "epoch": 12.717486179807972, "grad_norm": 0.08803168684244156, "learning_rate": 5.651069882231475e-06, "loss": 0.013, "step": 43710 }, { "epoch": 12.720395693919116, "grad_norm": 0.11799080669879913, "learning_rate": 5.644606127348917e-06, "loss": 0.0134, "step": 43720 }, { "epoch": 12.72330520803026, "grad_norm": 0.09095441550016403, "learning_rate": 5.638145214387993e-06, "loss": 0.0119, "step": 43730 }, { "epoch": 12.726214722141403, "grad_norm": 0.08015689998865128, "learning_rate": 5.63168714531137e-06, "loss": 0.0159, "step": 43740 }, { "epoch": 12.729124236252545, "grad_norm": 0.099899061024189, "learning_rate": 5.625231922080839e-06, "loss": 0.0138, "step": 43750 }, { "epoch": 12.732033750363689, "grad_norm": 0.09972963482141495, "learning_rate": 5.618779546657322e-06, "loss": 0.0119, "step": 43760 }, { "epoch": 12.734943264474833, "grad_norm": 0.09005188941955566, "learning_rate": 5.612330021000882e-06, "loss": 0.0144, "step": 43770 }, { "epoch": 12.737852778585976, "grad_norm": 0.16727708280086517, "learning_rate": 5.605883347070716e-06, "loss": 0.0131, "step": 43780 }, { "epoch": 12.74076229269712, "grad_norm": 0.12443412840366364, "learning_rate": 5.599439526825145e-06, "loss": 0.0128, "step": 43790 }, { "epoch": 12.743671806808264, "grad_norm": 0.07327979803085327, "learning_rate": 5.592998562221649e-06, "loss": 0.0119, "step": 43800 }, { "epoch": 12.746581320919406, "grad_norm": 0.09407670050859451, "learning_rate": 5.586560455216802e-06, "loss": 0.014, "step": 43810 }, { "epoch": 12.74949083503055, "grad_norm": 0.08906875550746918, "learning_rate": 5.580125207766347e-06, "loss": 0.0135, "step": 43820 }, { "epoch": 12.752400349141693, "grad_norm": 0.10013342648744583, "learning_rate": 5.573692821825136e-06, "loss": 0.014, "step": 43830 }, { "epoch": 12.755309863252837, "grad_norm": 0.10640765726566315, "learning_rate": 5.567263299347158e-06, "loss": 0.0139, "step": 43840 }, { "epoch": 12.75821937736398, "grad_norm": 0.08003807812929153, "learning_rate": 5.560836642285535e-06, "loss": 0.0115, "step": 43850 }, { "epoch": 12.761128891475124, "grad_norm": 0.10034050792455673, "learning_rate": 5.554412852592506e-06, "loss": 0.0131, "step": 43860 }, { "epoch": 12.764038405586268, "grad_norm": 0.08678188174962997, "learning_rate": 5.547991932219466e-06, "loss": 0.0123, "step": 43870 }, { "epoch": 12.76694791969741, "grad_norm": 0.0926811546087265, "learning_rate": 5.541573883116904e-06, "loss": 0.0127, "step": 43880 }, { "epoch": 12.769857433808554, "grad_norm": 0.0939018502831459, "learning_rate": 5.535158707234468e-06, "loss": 0.0127, "step": 43890 }, { "epoch": 12.772766947919697, "grad_norm": 0.09938536584377289, "learning_rate": 5.528746406520903e-06, "loss": 0.0128, "step": 43900 }, { "epoch": 12.775676462030841, "grad_norm": 0.10456690937280655, "learning_rate": 5.522336982924111e-06, "loss": 0.0136, "step": 43910 }, { "epoch": 12.778585976141985, "grad_norm": 0.08320068567991257, "learning_rate": 5.515930438391101e-06, "loss": 0.0133, "step": 43920 }, { "epoch": 12.781495490253128, "grad_norm": 0.08950196206569672, "learning_rate": 5.509526774868011e-06, "loss": 0.0128, "step": 43930 }, { "epoch": 12.78440500436427, "grad_norm": 0.08219548314809799, "learning_rate": 5.503125994300105e-06, "loss": 0.0125, "step": 43940 }, { "epoch": 12.787314518475414, "grad_norm": 0.07873041182756424, "learning_rate": 5.49672809863177e-06, "loss": 0.0131, "step": 43950 }, { "epoch": 12.790224032586558, "grad_norm": 0.08124048262834549, "learning_rate": 5.490333089806517e-06, "loss": 0.0128, "step": 43960 }, { "epoch": 12.793133546697701, "grad_norm": 0.10492229461669922, "learning_rate": 5.4839409697669875e-06, "loss": 0.0123, "step": 43970 }, { "epoch": 12.796043060808845, "grad_norm": 0.08003266155719757, "learning_rate": 5.4775517404549335e-06, "loss": 0.0124, "step": 43980 }, { "epoch": 12.798952574919989, "grad_norm": 0.06956302374601364, "learning_rate": 5.471165403811235e-06, "loss": 0.0128, "step": 43990 }, { "epoch": 12.801862089031133, "grad_norm": 0.12458837032318115, "learning_rate": 5.464781961775894e-06, "loss": 0.013, "step": 44000 }, { "epoch": 12.804771603142274, "grad_norm": 0.08977669477462769, "learning_rate": 5.458401416288023e-06, "loss": 0.0133, "step": 44010 }, { "epoch": 12.807681117253418, "grad_norm": 0.09268458932638168, "learning_rate": 5.452023769285882e-06, "loss": 0.0146, "step": 44020 }, { "epoch": 12.810590631364562, "grad_norm": 0.0986364483833313, "learning_rate": 5.445649022706809e-06, "loss": 0.012, "step": 44030 }, { "epoch": 12.813500145475706, "grad_norm": 0.0992216169834137, "learning_rate": 5.439277178487306e-06, "loss": 0.0133, "step": 44040 }, { "epoch": 12.81640965958685, "grad_norm": 0.10179655998945236, "learning_rate": 5.432908238562949e-06, "loss": 0.0138, "step": 44050 }, { "epoch": 12.819319173697993, "grad_norm": 0.08116674423217773, "learning_rate": 5.42654220486847e-06, "loss": 0.0137, "step": 44060 }, { "epoch": 12.822228687809137, "grad_norm": 0.09469372779130936, "learning_rate": 5.420179079337698e-06, "loss": 0.0135, "step": 44070 }, { "epoch": 12.825138201920279, "grad_norm": 0.08667029440402985, "learning_rate": 5.413818863903579e-06, "loss": 0.0126, "step": 44080 }, { "epoch": 12.828047716031422, "grad_norm": 0.08200781792402267, "learning_rate": 5.407461560498182e-06, "loss": 0.0116, "step": 44090 }, { "epoch": 12.830957230142566, "grad_norm": 0.08658450841903687, "learning_rate": 5.401107171052682e-06, "loss": 0.0126, "step": 44100 }, { "epoch": 12.83386674425371, "grad_norm": 0.07709836214780807, "learning_rate": 5.394755697497383e-06, "loss": 0.0124, "step": 44110 }, { "epoch": 12.836776258364853, "grad_norm": 0.13626979291439056, "learning_rate": 5.388407141761692e-06, "loss": 0.0141, "step": 44120 }, { "epoch": 12.839685772475997, "grad_norm": 0.12366174161434174, "learning_rate": 5.382061505774132e-06, "loss": 0.0129, "step": 44130 }, { "epoch": 12.842595286587141, "grad_norm": 0.1106986254453659, "learning_rate": 5.37571879146234e-06, "loss": 0.0115, "step": 44140 }, { "epoch": 12.845504800698283, "grad_norm": 0.09932457655668259, "learning_rate": 5.369379000753065e-06, "loss": 0.0133, "step": 44150 }, { "epoch": 12.848414314809427, "grad_norm": 0.09156982600688934, "learning_rate": 5.363042135572161e-06, "loss": 0.0143, "step": 44160 }, { "epoch": 12.85132382892057, "grad_norm": 0.09077192842960358, "learning_rate": 5.356708197844619e-06, "loss": 0.013, "step": 44170 }, { "epoch": 12.854233343031714, "grad_norm": 0.08846408873796463, "learning_rate": 5.3503771894944985e-06, "loss": 0.0134, "step": 44180 }, { "epoch": 12.857142857142858, "grad_norm": 0.10716366022825241, "learning_rate": 5.344049112445015e-06, "loss": 0.0128, "step": 44190 }, { "epoch": 12.860052371254001, "grad_norm": 0.07778974622488022, "learning_rate": 5.337723968618451e-06, "loss": 0.0137, "step": 44200 }, { "epoch": 12.862961885365143, "grad_norm": 0.09440521150827408, "learning_rate": 5.331401759936232e-06, "loss": 0.0148, "step": 44210 }, { "epoch": 12.865871399476287, "grad_norm": 0.10684272646903992, "learning_rate": 5.325082488318874e-06, "loss": 0.0106, "step": 44220 }, { "epoch": 12.86878091358743, "grad_norm": 0.1124066561460495, "learning_rate": 5.318766155686004e-06, "loss": 0.0146, "step": 44230 }, { "epoch": 12.871690427698574, "grad_norm": 0.09468758851289749, "learning_rate": 5.31245276395636e-06, "loss": 0.0121, "step": 44240 }, { "epoch": 12.874599941809718, "grad_norm": 0.09290508180856705, "learning_rate": 5.306142315047774e-06, "loss": 0.0129, "step": 44250 }, { "epoch": 12.877509455920862, "grad_norm": 0.08823568373918533, "learning_rate": 5.299834810877212e-06, "loss": 0.0118, "step": 44260 }, { "epoch": 12.880418970032004, "grad_norm": 0.12346681952476501, "learning_rate": 5.293530253360707e-06, "loss": 0.0134, "step": 44270 }, { "epoch": 12.883328484143147, "grad_norm": 0.09235703945159912, "learning_rate": 5.28722864441343e-06, "loss": 0.0144, "step": 44280 }, { "epoch": 12.886237998254291, "grad_norm": 0.10005355626344681, "learning_rate": 5.280929985949642e-06, "loss": 0.0136, "step": 44290 }, { "epoch": 12.889147512365435, "grad_norm": 0.11406692862510681, "learning_rate": 5.2746342798827055e-06, "loss": 0.0124, "step": 44300 }, { "epoch": 12.892057026476579, "grad_norm": 0.07244039326906204, "learning_rate": 5.268341528125092e-06, "loss": 0.012, "step": 44310 }, { "epoch": 12.894966540587722, "grad_norm": 0.11147821694612503, "learning_rate": 5.2620517325883735e-06, "loss": 0.0145, "step": 44320 }, { "epoch": 12.897876054698866, "grad_norm": 0.09417087584733963, "learning_rate": 5.255764895183216e-06, "loss": 0.0121, "step": 44330 }, { "epoch": 12.900785568810008, "grad_norm": 0.07218325138092041, "learning_rate": 5.249481017819414e-06, "loss": 0.0121, "step": 44340 }, { "epoch": 12.903695082921152, "grad_norm": 0.10638504475355148, "learning_rate": 5.243200102405819e-06, "loss": 0.0129, "step": 44350 }, { "epoch": 12.906604597032295, "grad_norm": 0.08009985834360123, "learning_rate": 5.236922150850426e-06, "loss": 0.0135, "step": 44360 }, { "epoch": 12.909514111143439, "grad_norm": 0.08622291684150696, "learning_rate": 5.230647165060305e-06, "loss": 0.0128, "step": 44370 }, { "epoch": 12.912423625254583, "grad_norm": 0.10078838467597961, "learning_rate": 5.224375146941629e-06, "loss": 0.0136, "step": 44380 }, { "epoch": 12.915333139365726, "grad_norm": 0.07858460396528244, "learning_rate": 5.218106098399674e-06, "loss": 0.0126, "step": 44390 }, { "epoch": 12.91824265347687, "grad_norm": 0.11563673615455627, "learning_rate": 5.211840021338804e-06, "loss": 0.013, "step": 44400 }, { "epoch": 12.921152167588012, "grad_norm": 0.09595376253128052, "learning_rate": 5.205576917662506e-06, "loss": 0.0133, "step": 44410 }, { "epoch": 12.924061681699156, "grad_norm": 0.09235931932926178, "learning_rate": 5.1993167892733224e-06, "loss": 0.0119, "step": 44420 }, { "epoch": 12.9269711958103, "grad_norm": 0.1004432663321495, "learning_rate": 5.193059638072931e-06, "loss": 0.0133, "step": 44430 }, { "epoch": 12.929880709921443, "grad_norm": 0.11126980930566788, "learning_rate": 5.186805465962085e-06, "loss": 0.0136, "step": 44440 }, { "epoch": 12.932790224032587, "grad_norm": 0.11166372150182724, "learning_rate": 5.180554274840636e-06, "loss": 0.0144, "step": 44450 }, { "epoch": 12.93569973814373, "grad_norm": 0.12881411612033844, "learning_rate": 5.17430606660753e-06, "loss": 0.0133, "step": 44460 }, { "epoch": 12.938609252254874, "grad_norm": 0.09583310037851334, "learning_rate": 5.168060843160808e-06, "loss": 0.0132, "step": 44470 }, { "epoch": 12.941518766366016, "grad_norm": 0.09325210750102997, "learning_rate": 5.161818606397601e-06, "loss": 0.0133, "step": 44480 }, { "epoch": 12.94442828047716, "grad_norm": 0.09334293752908707, "learning_rate": 5.15557935821415e-06, "loss": 0.0134, "step": 44490 }, { "epoch": 12.947337794588304, "grad_norm": 0.08900540322065353, "learning_rate": 5.149343100505752e-06, "loss": 0.0115, "step": 44500 }, { "epoch": 12.950247308699447, "grad_norm": 0.11307711154222488, "learning_rate": 5.1431098351668334e-06, "loss": 0.0118, "step": 44510 }, { "epoch": 12.953156822810591, "grad_norm": 0.12939687073230743, "learning_rate": 5.13687956409089e-06, "loss": 0.0146, "step": 44520 }, { "epoch": 12.956066336921735, "grad_norm": 0.0948772132396698, "learning_rate": 5.130652289170515e-06, "loss": 0.014, "step": 44530 }, { "epoch": 12.958975851032877, "grad_norm": 0.110581174492836, "learning_rate": 5.1244280122973916e-06, "loss": 0.0122, "step": 44540 }, { "epoch": 12.96188536514402, "grad_norm": 0.13728763163089752, "learning_rate": 5.118206735362282e-06, "loss": 0.0152, "step": 44550 }, { "epoch": 12.964794879255164, "grad_norm": 0.09254664927721024, "learning_rate": 5.111988460255062e-06, "loss": 0.0144, "step": 44560 }, { "epoch": 12.967704393366308, "grad_norm": 0.12332849949598312, "learning_rate": 5.105773188864661e-06, "loss": 0.0147, "step": 44570 }, { "epoch": 12.970613907477452, "grad_norm": 0.0752645954489708, "learning_rate": 5.0995609230791265e-06, "loss": 0.0115, "step": 44580 }, { "epoch": 12.973523421588595, "grad_norm": 0.0928787812590599, "learning_rate": 5.093351664785578e-06, "loss": 0.0118, "step": 44590 }, { "epoch": 12.976432935699737, "grad_norm": 0.09049023687839508, "learning_rate": 5.0871454158702265e-06, "loss": 0.0138, "step": 44600 }, { "epoch": 12.979342449810881, "grad_norm": 0.1022428646683693, "learning_rate": 5.080942178218361e-06, "loss": 0.012, "step": 44610 }, { "epoch": 12.982251963922025, "grad_norm": 0.10021836310625076, "learning_rate": 5.074741953714366e-06, "loss": 0.0127, "step": 44620 }, { "epoch": 12.985161478033168, "grad_norm": 0.0906120166182518, "learning_rate": 5.068544744241703e-06, "loss": 0.0132, "step": 44630 }, { "epoch": 12.988070992144312, "grad_norm": 0.11120517551898956, "learning_rate": 5.062350551682922e-06, "loss": 0.0138, "step": 44640 }, { "epoch": 12.990980506255456, "grad_norm": 0.11053928732872009, "learning_rate": 5.056159377919651e-06, "loss": 0.0127, "step": 44650 }, { "epoch": 12.9938900203666, "grad_norm": 0.06414616852998734, "learning_rate": 5.049971224832613e-06, "loss": 0.0114, "step": 44660 }, { "epoch": 12.996799534477741, "grad_norm": 0.15824849903583527, "learning_rate": 5.043786094301602e-06, "loss": 0.0115, "step": 44670 }, { "epoch": 12.999709048588885, "grad_norm": 0.09293278306722641, "learning_rate": 5.037603988205498e-06, "loss": 0.0121, "step": 44680 }, { "epoch": 13.002618562700029, "grad_norm": 0.10180754959583282, "learning_rate": 5.031424908422258e-06, "loss": 0.0112, "step": 44690 }, { "epoch": 13.005528076811173, "grad_norm": 0.09188362956047058, "learning_rate": 5.025248856828923e-06, "loss": 0.0123, "step": 44700 }, { "epoch": 13.008437590922316, "grad_norm": 0.10185271501541138, "learning_rate": 5.019075835301627e-06, "loss": 0.0126, "step": 44710 }, { "epoch": 13.01134710503346, "grad_norm": 0.09703028947114944, "learning_rate": 5.01290584571555e-06, "loss": 0.0133, "step": 44720 }, { "epoch": 13.014256619144604, "grad_norm": 0.11196744441986084, "learning_rate": 5.006738889944994e-06, "loss": 0.0139, "step": 44730 }, { "epoch": 13.017166133255746, "grad_norm": 0.08452003449201584, "learning_rate": 5.000574969863296e-06, "loss": 0.0123, "step": 44740 }, { "epoch": 13.02007564736689, "grad_norm": 0.08098533004522324, "learning_rate": 4.994414087342909e-06, "loss": 0.0119, "step": 44750 }, { "epoch": 13.022985161478033, "grad_norm": 0.1275739222764969, "learning_rate": 4.988256244255338e-06, "loss": 0.0127, "step": 44760 }, { "epoch": 13.025894675589177, "grad_norm": 0.10864962637424469, "learning_rate": 4.982101442471177e-06, "loss": 0.0113, "step": 44770 }, { "epoch": 13.02880418970032, "grad_norm": 0.11127828806638718, "learning_rate": 4.975949683860091e-06, "loss": 0.0141, "step": 44780 }, { "epoch": 13.031713703811464, "grad_norm": 0.09527474641799927, "learning_rate": 4.969800970290819e-06, "loss": 0.0122, "step": 44790 }, { "epoch": 13.034623217922608, "grad_norm": 0.0967836081981659, "learning_rate": 4.9636553036311765e-06, "loss": 0.0131, "step": 44800 }, { "epoch": 13.03753273203375, "grad_norm": 0.10994989424943924, "learning_rate": 4.957512685748063e-06, "loss": 0.0128, "step": 44810 }, { "epoch": 13.040442246144893, "grad_norm": 0.07904589921236038, "learning_rate": 4.95137311850744e-06, "loss": 0.0124, "step": 44820 }, { "epoch": 13.043351760256037, "grad_norm": 0.07745049893856049, "learning_rate": 4.945236603774343e-06, "loss": 0.0113, "step": 44830 }, { "epoch": 13.04626127436718, "grad_norm": 0.08225405961275101, "learning_rate": 4.939103143412886e-06, "loss": 0.0128, "step": 44840 }, { "epoch": 13.049170788478325, "grad_norm": 0.11890552937984467, "learning_rate": 4.932972739286248e-06, "loss": 0.0123, "step": 44850 }, { "epoch": 13.052080302589468, "grad_norm": 0.08898553252220154, "learning_rate": 4.926845393256696e-06, "loss": 0.0118, "step": 44860 }, { "epoch": 13.05498981670061, "grad_norm": 0.10022666305303574, "learning_rate": 4.920721107185536e-06, "loss": 0.0138, "step": 44870 }, { "epoch": 13.057899330811754, "grad_norm": 0.0944943055510521, "learning_rate": 4.914599882933188e-06, "loss": 0.0128, "step": 44880 }, { "epoch": 13.060808844922898, "grad_norm": 0.09423062950372696, "learning_rate": 4.908481722359095e-06, "loss": 0.0122, "step": 44890 }, { "epoch": 13.063718359034041, "grad_norm": 0.10573592036962509, "learning_rate": 4.90236662732181e-06, "loss": 0.0138, "step": 44900 }, { "epoch": 13.066627873145185, "grad_norm": 0.10861211270093918, "learning_rate": 4.89625459967893e-06, "loss": 0.0128, "step": 44910 }, { "epoch": 13.069537387256329, "grad_norm": 0.11340077221393585, "learning_rate": 4.890145641287128e-06, "loss": 0.0122, "step": 44920 }, { "epoch": 13.072446901367472, "grad_norm": 0.14071427285671234, "learning_rate": 4.884039754002146e-06, "loss": 0.0122, "step": 44930 }, { "epoch": 13.075356415478614, "grad_norm": 0.10103024542331696, "learning_rate": 4.877936939678789e-06, "loss": 0.0117, "step": 44940 }, { "epoch": 13.078265929589758, "grad_norm": 0.08692694455385208, "learning_rate": 4.871837200170927e-06, "loss": 0.0133, "step": 44950 }, { "epoch": 13.081175443700902, "grad_norm": 0.10656991600990295, "learning_rate": 4.865740537331512e-06, "loss": 0.0118, "step": 44960 }, { "epoch": 13.084084957812046, "grad_norm": 0.08822862803936005, "learning_rate": 4.85964695301254e-06, "loss": 0.0123, "step": 44970 }, { "epoch": 13.08699447192319, "grad_norm": 0.06783229857683182, "learning_rate": 4.853556449065084e-06, "loss": 0.0128, "step": 44980 }, { "epoch": 13.089903986034333, "grad_norm": 0.08261586725711823, "learning_rate": 4.847469027339278e-06, "loss": 0.0143, "step": 44990 }, { "epoch": 13.092813500145475, "grad_norm": 0.1081249862909317, "learning_rate": 4.841384689684314e-06, "loss": 0.0122, "step": 45000 }, { "epoch": 13.095723014256619, "grad_norm": 0.0823172852396965, "learning_rate": 4.835303437948468e-06, "loss": 0.012, "step": 45010 }, { "epoch": 13.098632528367762, "grad_norm": 0.11790493875741959, "learning_rate": 4.8292252739790455e-06, "loss": 0.012, "step": 45020 }, { "epoch": 13.101542042478906, "grad_norm": 0.08918595314025879, "learning_rate": 4.82315019962245e-06, "loss": 0.0118, "step": 45030 }, { "epoch": 13.10445155659005, "grad_norm": 0.09644203633069992, "learning_rate": 4.8170782167241104e-06, "loss": 0.0125, "step": 45040 }, { "epoch": 13.107361070701193, "grad_norm": 0.16972611844539642, "learning_rate": 4.811009327128552e-06, "loss": 0.0115, "step": 45050 }, { "epoch": 13.110270584812337, "grad_norm": 0.1064622551202774, "learning_rate": 4.8049435326793275e-06, "loss": 0.0132, "step": 45060 }, { "epoch": 13.113180098923479, "grad_norm": 0.08741249889135361, "learning_rate": 4.798880835219078e-06, "loss": 0.0139, "step": 45070 }, { "epoch": 13.116089613034623, "grad_norm": 0.08980703353881836, "learning_rate": 4.792821236589485e-06, "loss": 0.0127, "step": 45080 }, { "epoch": 13.118999127145766, "grad_norm": 0.10667525231838226, "learning_rate": 4.7867647386312965e-06, "loss": 0.0135, "step": 45090 }, { "epoch": 13.12190864125691, "grad_norm": 0.12355615198612213, "learning_rate": 4.780711343184314e-06, "loss": 0.0115, "step": 45100 }, { "epoch": 13.124818155368054, "grad_norm": 0.10572559386491776, "learning_rate": 4.774661052087397e-06, "loss": 0.0137, "step": 45110 }, { "epoch": 13.127727669479198, "grad_norm": 0.08551068603992462, "learning_rate": 4.768613867178472e-06, "loss": 0.0126, "step": 45120 }, { "epoch": 13.130637183590341, "grad_norm": 0.12383793294429779, "learning_rate": 4.762569790294511e-06, "loss": 0.0149, "step": 45130 }, { "epoch": 13.133546697701483, "grad_norm": 0.09306840598583221, "learning_rate": 4.756528823271544e-06, "loss": 0.0118, "step": 45140 }, { "epoch": 13.136456211812627, "grad_norm": 0.1141437366604805, "learning_rate": 4.7504909679446565e-06, "loss": 0.0123, "step": 45150 }, { "epoch": 13.13936572592377, "grad_norm": 0.10036598891019821, "learning_rate": 4.7444562261479905e-06, "loss": 0.0133, "step": 45160 }, { "epoch": 13.142275240034914, "grad_norm": 0.09194821864366531, "learning_rate": 4.738424599714736e-06, "loss": 0.0136, "step": 45170 }, { "epoch": 13.145184754146058, "grad_norm": 0.08310922980308533, "learning_rate": 4.732396090477157e-06, "loss": 0.0135, "step": 45180 }, { "epoch": 13.148094268257202, "grad_norm": 0.12680011987686157, "learning_rate": 4.726370700266536e-06, "loss": 0.0121, "step": 45190 }, { "epoch": 13.151003782368344, "grad_norm": 0.07568854093551636, "learning_rate": 4.720348430913245e-06, "loss": 0.0142, "step": 45200 }, { "epoch": 13.153913296479487, "grad_norm": 0.09605418145656586, "learning_rate": 4.714329284246674e-06, "loss": 0.0123, "step": 45210 }, { "epoch": 13.156822810590631, "grad_norm": 0.09386259317398071, "learning_rate": 4.708313262095293e-06, "loss": 0.0135, "step": 45220 }, { "epoch": 13.159732324701775, "grad_norm": 0.10547700524330139, "learning_rate": 4.7023003662866085e-06, "loss": 0.0117, "step": 45230 }, { "epoch": 13.162641838812918, "grad_norm": 0.08597496896982193, "learning_rate": 4.696290598647175e-06, "loss": 0.0136, "step": 45240 }, { "epoch": 13.165551352924062, "grad_norm": 0.09204474091529846, "learning_rate": 4.690283961002605e-06, "loss": 0.0145, "step": 45250 }, { "epoch": 13.168460867035206, "grad_norm": 0.07643705606460571, "learning_rate": 4.684280455177552e-06, "loss": 0.0111, "step": 45260 }, { "epoch": 13.171370381146348, "grad_norm": 0.07834848016500473, "learning_rate": 4.678280082995729e-06, "loss": 0.0122, "step": 45270 }, { "epoch": 13.174279895257492, "grad_norm": 0.10469163209199905, "learning_rate": 4.672282846279888e-06, "loss": 0.0151, "step": 45280 }, { "epoch": 13.177189409368635, "grad_norm": 0.09700009226799011, "learning_rate": 4.6662887468518315e-06, "loss": 0.0144, "step": 45290 }, { "epoch": 13.180098923479779, "grad_norm": 0.10220013558864594, "learning_rate": 4.660297786532408e-06, "loss": 0.0114, "step": 45300 }, { "epoch": 13.183008437590923, "grad_norm": 0.117092564702034, "learning_rate": 4.654309967141511e-06, "loss": 0.0136, "step": 45310 }, { "epoch": 13.185917951702066, "grad_norm": 0.10224407911300659, "learning_rate": 4.648325290498081e-06, "loss": 0.0126, "step": 45320 }, { "epoch": 13.188827465813208, "grad_norm": 0.10794214904308319, "learning_rate": 4.642343758420116e-06, "loss": 0.0122, "step": 45330 }, { "epoch": 13.191736979924352, "grad_norm": 0.10264508426189423, "learning_rate": 4.6363653727246294e-06, "loss": 0.0135, "step": 45340 }, { "epoch": 13.194646494035496, "grad_norm": 0.08198370784521103, "learning_rate": 4.630390135227716e-06, "loss": 0.0126, "step": 45350 }, { "epoch": 13.19755600814664, "grad_norm": 0.09545069932937622, "learning_rate": 4.6244180477444785e-06, "loss": 0.0142, "step": 45360 }, { "epoch": 13.200465522257783, "grad_norm": 0.09637915343046188, "learning_rate": 4.618449112089089e-06, "loss": 0.0123, "step": 45370 }, { "epoch": 13.203375036368927, "grad_norm": 0.08843624591827393, "learning_rate": 4.612483330074751e-06, "loss": 0.0128, "step": 45380 }, { "epoch": 13.20628455048007, "grad_norm": 0.15693442523479462, "learning_rate": 4.606520703513713e-06, "loss": 0.0127, "step": 45390 }, { "epoch": 13.209194064591212, "grad_norm": 0.08045376092195511, "learning_rate": 4.600561234217258e-06, "loss": 0.0117, "step": 45400 }, { "epoch": 13.212103578702356, "grad_norm": 0.10595377534627914, "learning_rate": 4.594604923995715e-06, "loss": 0.0118, "step": 45410 }, { "epoch": 13.2150130928135, "grad_norm": 0.11041507869958878, "learning_rate": 4.5886517746584635e-06, "loss": 0.0123, "step": 45420 }, { "epoch": 13.217922606924644, "grad_norm": 0.09869062900543213, "learning_rate": 4.582701788013907e-06, "loss": 0.0129, "step": 45430 }, { "epoch": 13.220832121035787, "grad_norm": 0.10580651462078094, "learning_rate": 4.576754965869493e-06, "loss": 0.0123, "step": 45440 }, { "epoch": 13.223741635146931, "grad_norm": 0.16476714611053467, "learning_rate": 4.570811310031713e-06, "loss": 0.0122, "step": 45450 }, { "epoch": 13.226651149258075, "grad_norm": 0.10373087227344513, "learning_rate": 4.5648708223060884e-06, "loss": 0.0123, "step": 45460 }, { "epoch": 13.229560663369217, "grad_norm": 0.08573351800441742, "learning_rate": 4.558933504497184e-06, "loss": 0.0125, "step": 45470 }, { "epoch": 13.23247017748036, "grad_norm": 0.18537551164627075, "learning_rate": 4.552999358408602e-06, "loss": 0.0146, "step": 45480 }, { "epoch": 13.235379691591504, "grad_norm": 0.07779243588447571, "learning_rate": 4.547068385842972e-06, "loss": 0.0138, "step": 45490 }, { "epoch": 13.238289205702648, "grad_norm": 0.0984557643532753, "learning_rate": 4.541140588601981e-06, "loss": 0.0141, "step": 45500 }, { "epoch": 13.241198719813791, "grad_norm": 0.08746114373207092, "learning_rate": 4.535215968486321e-06, "loss": 0.0139, "step": 45510 }, { "epoch": 13.244108233924935, "grad_norm": 0.12299273163080215, "learning_rate": 4.52929452729575e-06, "loss": 0.0124, "step": 45520 }, { "epoch": 13.247017748036077, "grad_norm": 0.06674238294363022, "learning_rate": 4.523376266829036e-06, "loss": 0.0143, "step": 45530 }, { "epoch": 13.24992726214722, "grad_norm": 0.10791859030723572, "learning_rate": 4.517461188883993e-06, "loss": 0.0138, "step": 45540 }, { "epoch": 13.252836776258365, "grad_norm": 0.12293585389852524, "learning_rate": 4.5115492952574655e-06, "loss": 0.0137, "step": 45550 }, { "epoch": 13.255746290369508, "grad_norm": 0.06335607916116714, "learning_rate": 4.505640587745327e-06, "loss": 0.0122, "step": 45560 }, { "epoch": 13.258655804480652, "grad_norm": 0.10242882370948792, "learning_rate": 4.499735068142499e-06, "loss": 0.0111, "step": 45570 }, { "epoch": 13.261565318591796, "grad_norm": 0.1332336962223053, "learning_rate": 4.493832738242905e-06, "loss": 0.013, "step": 45580 }, { "epoch": 13.26447483270294, "grad_norm": 0.1432618647813797, "learning_rate": 4.4879335998395285e-06, "loss": 0.014, "step": 45590 }, { "epoch": 13.267384346814081, "grad_norm": 0.08122864365577698, "learning_rate": 4.482037654724372e-06, "loss": 0.0125, "step": 45600 }, { "epoch": 13.270293860925225, "grad_norm": 0.07944197952747345, "learning_rate": 4.476144904688464e-06, "loss": 0.0108, "step": 45610 }, { "epoch": 13.273203375036369, "grad_norm": 0.0838153138756752, "learning_rate": 4.470255351521868e-06, "loss": 0.0111, "step": 45620 }, { "epoch": 13.276112889147512, "grad_norm": 0.10167590528726578, "learning_rate": 4.4643689970136746e-06, "loss": 0.0127, "step": 45630 }, { "epoch": 13.279022403258656, "grad_norm": 0.07909952849149704, "learning_rate": 4.458485842951998e-06, "loss": 0.0111, "step": 45640 }, { "epoch": 13.2819319173698, "grad_norm": 0.11888597160577774, "learning_rate": 4.452605891123999e-06, "loss": 0.0128, "step": 45650 }, { "epoch": 13.284841431480944, "grad_norm": 0.0935826227068901, "learning_rate": 4.446729143315832e-06, "loss": 0.0125, "step": 45660 }, { "epoch": 13.287750945592085, "grad_norm": 0.11208090931177139, "learning_rate": 4.440855601312713e-06, "loss": 0.0124, "step": 45670 }, { "epoch": 13.29066045970323, "grad_norm": 0.08275549113750458, "learning_rate": 4.434985266898866e-06, "loss": 0.0116, "step": 45680 }, { "epoch": 13.293569973814373, "grad_norm": 0.07178911566734314, "learning_rate": 4.4291181418575405e-06, "loss": 0.0128, "step": 45690 }, { "epoch": 13.296479487925517, "grad_norm": 0.08818624913692474, "learning_rate": 4.423254227971017e-06, "loss": 0.0119, "step": 45700 }, { "epoch": 13.29938900203666, "grad_norm": 0.08784140646457672, "learning_rate": 4.41739352702059e-06, "loss": 0.0133, "step": 45710 }, { "epoch": 13.302298516147804, "grad_norm": 0.08444219827651978, "learning_rate": 4.4115360407866026e-06, "loss": 0.0132, "step": 45720 }, { "epoch": 13.305208030258946, "grad_norm": 0.1247299537062645, "learning_rate": 4.405681771048384e-06, "loss": 0.0147, "step": 45730 }, { "epoch": 13.30811754437009, "grad_norm": 0.07903465628623962, "learning_rate": 4.3998307195843205e-06, "loss": 0.0124, "step": 45740 }, { "epoch": 13.311027058481233, "grad_norm": 0.08880617469549179, "learning_rate": 4.393982888171803e-06, "loss": 0.0134, "step": 45750 }, { "epoch": 13.313936572592377, "grad_norm": 0.16656915843486786, "learning_rate": 4.38813827858725e-06, "loss": 0.012, "step": 45760 }, { "epoch": 13.31684608670352, "grad_norm": 0.07365002483129501, "learning_rate": 4.382296892606097e-06, "loss": 0.0141, "step": 45770 }, { "epoch": 13.319755600814664, "grad_norm": 0.0918223038315773, "learning_rate": 4.376458732002803e-06, "loss": 0.0111, "step": 45780 }, { "epoch": 13.322665114925808, "grad_norm": 0.16756172478199005, "learning_rate": 4.370623798550846e-06, "loss": 0.0123, "step": 45790 }, { "epoch": 13.32557462903695, "grad_norm": 0.19294337928295135, "learning_rate": 4.364792094022733e-06, "loss": 0.0121, "step": 45800 }, { "epoch": 13.328484143148094, "grad_norm": 0.06664958596229553, "learning_rate": 4.3589636201899665e-06, "loss": 0.0115, "step": 45810 }, { "epoch": 13.331393657259238, "grad_norm": 0.12633995711803436, "learning_rate": 4.3531383788230975e-06, "loss": 0.0116, "step": 45820 }, { "epoch": 13.334303171370381, "grad_norm": 0.10141400992870331, "learning_rate": 4.347316371691675e-06, "loss": 0.0136, "step": 45830 }, { "epoch": 13.337212685481525, "grad_norm": 0.09784006327390671, "learning_rate": 4.34149760056427e-06, "loss": 0.0134, "step": 45840 }, { "epoch": 13.340122199592669, "grad_norm": 0.08761068433523178, "learning_rate": 4.3356820672084736e-06, "loss": 0.0124, "step": 45850 }, { "epoch": 13.343031713703812, "grad_norm": 0.07618606090545654, "learning_rate": 4.329869773390885e-06, "loss": 0.0119, "step": 45860 }, { "epoch": 13.345941227814954, "grad_norm": 0.14824408292770386, "learning_rate": 4.3240607208771395e-06, "loss": 0.013, "step": 45870 }, { "epoch": 13.348850741926098, "grad_norm": 0.0799684077501297, "learning_rate": 4.318254911431857e-06, "loss": 0.0134, "step": 45880 }, { "epoch": 13.351760256037242, "grad_norm": 0.09874878823757172, "learning_rate": 4.312452346818705e-06, "loss": 0.0127, "step": 45890 }, { "epoch": 13.354669770148385, "grad_norm": 0.08548915386199951, "learning_rate": 4.3066530288003335e-06, "loss": 0.0135, "step": 45900 }, { "epoch": 13.35757928425953, "grad_norm": 0.14171849191188812, "learning_rate": 4.300856959138436e-06, "loss": 0.0133, "step": 45910 }, { "epoch": 13.360488798370673, "grad_norm": 0.14575201272964478, "learning_rate": 4.295064139593698e-06, "loss": 0.0138, "step": 45920 }, { "epoch": 13.363398312481815, "grad_norm": 0.10319806635379791, "learning_rate": 4.289274571925828e-06, "loss": 0.0121, "step": 45930 }, { "epoch": 13.366307826592958, "grad_norm": 0.0829898789525032, "learning_rate": 4.2834882578935425e-06, "loss": 0.0121, "step": 45940 }, { "epoch": 13.369217340704102, "grad_norm": 0.08368202298879623, "learning_rate": 4.277705199254572e-06, "loss": 0.0126, "step": 45950 }, { "epoch": 13.372126854815246, "grad_norm": 0.09044221043586731, "learning_rate": 4.271925397765649e-06, "loss": 0.0111, "step": 45960 }, { "epoch": 13.37503636892639, "grad_norm": 0.10025473684072495, "learning_rate": 4.2661488551825386e-06, "loss": 0.0124, "step": 45970 }, { "epoch": 13.377945883037533, "grad_norm": 0.08373741060495377, "learning_rate": 4.260375573259994e-06, "loss": 0.0132, "step": 45980 }, { "epoch": 13.380855397148677, "grad_norm": 0.09074379503726959, "learning_rate": 4.254605553751787e-06, "loss": 0.0125, "step": 45990 }, { "epoch": 13.383764911259819, "grad_norm": 0.07325662672519684, "learning_rate": 4.248838798410698e-06, "loss": 0.0144, "step": 46000 }, { "epoch": 13.386674425370963, "grad_norm": 0.07641909271478653, "learning_rate": 4.243075308988509e-06, "loss": 0.0125, "step": 46010 }, { "epoch": 13.389583939482106, "grad_norm": 0.09666959196329117, "learning_rate": 4.23731508723603e-06, "loss": 0.0145, "step": 46020 }, { "epoch": 13.39249345359325, "grad_norm": 0.10226812213659286, "learning_rate": 4.231558134903045e-06, "loss": 0.0133, "step": 46030 }, { "epoch": 13.395402967704394, "grad_norm": 0.09789557009935379, "learning_rate": 4.225804453738386e-06, "loss": 0.0143, "step": 46040 }, { "epoch": 13.398312481815537, "grad_norm": 0.10462182015180588, "learning_rate": 4.2200540454898465e-06, "loss": 0.0128, "step": 46050 }, { "epoch": 13.40122199592668, "grad_norm": 0.09818199276924133, "learning_rate": 4.214306911904267e-06, "loss": 0.0133, "step": 46060 }, { "epoch": 13.404131510037823, "grad_norm": 0.08777053654193878, "learning_rate": 4.208563054727466e-06, "loss": 0.0117, "step": 46070 }, { "epoch": 13.407041024148967, "grad_norm": 0.10149826854467392, "learning_rate": 4.202822475704277e-06, "loss": 0.0145, "step": 46080 }, { "epoch": 13.40995053826011, "grad_norm": 0.09792733192443848, "learning_rate": 4.1970851765785375e-06, "loss": 0.0123, "step": 46090 }, { "epoch": 13.412860052371254, "grad_norm": 0.11213556677103043, "learning_rate": 4.191351159093087e-06, "loss": 0.0134, "step": 46100 }, { "epoch": 13.415769566482398, "grad_norm": 0.07949617505073547, "learning_rate": 4.185620424989764e-06, "loss": 0.0134, "step": 46110 }, { "epoch": 13.418679080593542, "grad_norm": 0.09539877623319626, "learning_rate": 4.1798929760094244e-06, "loss": 0.0129, "step": 46120 }, { "epoch": 13.421588594704684, "grad_norm": 0.07719457894563675, "learning_rate": 4.174168813891909e-06, "loss": 0.0119, "step": 46130 }, { "epoch": 13.424498108815827, "grad_norm": 0.09688746929168701, "learning_rate": 4.168447940376069e-06, "loss": 0.0118, "step": 46140 }, { "epoch": 13.427407622926971, "grad_norm": 0.16604654490947723, "learning_rate": 4.162730357199753e-06, "loss": 0.0118, "step": 46150 }, { "epoch": 13.430317137038115, "grad_norm": 0.09507827460765839, "learning_rate": 4.157016066099811e-06, "loss": 0.0136, "step": 46160 }, { "epoch": 13.433226651149258, "grad_norm": 0.08055183291435242, "learning_rate": 4.151305068812103e-06, "loss": 0.012, "step": 46170 }, { "epoch": 13.436136165260402, "grad_norm": 0.0951862707734108, "learning_rate": 4.145597367071462e-06, "loss": 0.0125, "step": 46180 }, { "epoch": 13.439045679371546, "grad_norm": 0.0842786505818367, "learning_rate": 4.139892962611757e-06, "loss": 0.0128, "step": 46190 }, { "epoch": 13.441955193482688, "grad_norm": 0.09230492264032364, "learning_rate": 4.1341918571658174e-06, "loss": 0.0131, "step": 46200 }, { "epoch": 13.444864707593831, "grad_norm": 0.0805114209651947, "learning_rate": 4.1284940524655e-06, "loss": 0.0125, "step": 46210 }, { "epoch": 13.447774221704975, "grad_norm": 0.09052389115095139, "learning_rate": 4.122799550241644e-06, "loss": 0.014, "step": 46220 }, { "epoch": 13.450683735816119, "grad_norm": 0.10231635719537735, "learning_rate": 4.117108352224089e-06, "loss": 0.0136, "step": 46230 }, { "epoch": 13.453593249927263, "grad_norm": 0.21150662004947662, "learning_rate": 4.111420460141671e-06, "loss": 0.0122, "step": 46240 }, { "epoch": 13.456502764038406, "grad_norm": 0.16571399569511414, "learning_rate": 4.10573587572222e-06, "loss": 0.0127, "step": 46250 }, { "epoch": 13.459412278149548, "grad_norm": 0.105082668364048, "learning_rate": 4.100054600692563e-06, "loss": 0.0118, "step": 46260 }, { "epoch": 13.462321792260692, "grad_norm": 0.09824279695749283, "learning_rate": 4.094376636778519e-06, "loss": 0.0109, "step": 46270 }, { "epoch": 13.465231306371836, "grad_norm": 0.07357224076986313, "learning_rate": 4.0887019857049095e-06, "loss": 0.0132, "step": 46280 }, { "epoch": 13.46814082048298, "grad_norm": 0.09299944341182709, "learning_rate": 4.083030649195541e-06, "loss": 0.0128, "step": 46290 }, { "epoch": 13.471050334594123, "grad_norm": 0.0914212241768837, "learning_rate": 4.077362628973217e-06, "loss": 0.0125, "step": 46300 }, { "epoch": 13.473959848705267, "grad_norm": 0.07639770209789276, "learning_rate": 4.071697926759728e-06, "loss": 0.0123, "step": 46310 }, { "epoch": 13.47686936281641, "grad_norm": 0.09768535941839218, "learning_rate": 4.066036544275866e-06, "loss": 0.0119, "step": 46320 }, { "epoch": 13.479778876927552, "grad_norm": 0.13114866614341736, "learning_rate": 4.060378483241402e-06, "loss": 0.0118, "step": 46330 }, { "epoch": 13.482688391038696, "grad_norm": 0.10899398475885391, "learning_rate": 4.054723745375121e-06, "loss": 0.0126, "step": 46340 }, { "epoch": 13.48559790514984, "grad_norm": 0.10191625356674194, "learning_rate": 4.049072332394763e-06, "loss": 0.013, "step": 46350 }, { "epoch": 13.488507419260984, "grad_norm": 0.09378261864185333, "learning_rate": 4.043424246017096e-06, "loss": 0.0137, "step": 46360 }, { "epoch": 13.491416933372127, "grad_norm": 0.09195581823587418, "learning_rate": 4.037779487957842e-06, "loss": 0.0135, "step": 46370 }, { "epoch": 13.494326447483271, "grad_norm": 0.14524990320205688, "learning_rate": 4.0321380599317456e-06, "loss": 0.013, "step": 46380 }, { "epoch": 13.497235961594413, "grad_norm": 0.10695957392454147, "learning_rate": 4.026499963652515e-06, "loss": 0.013, "step": 46390 }, { "epoch": 13.500145475705557, "grad_norm": 0.1427183896303177, "learning_rate": 4.0208652008328586e-06, "loss": 0.0132, "step": 46400 }, { "epoch": 13.5030549898167, "grad_norm": 0.10114685446023941, "learning_rate": 4.0152337731844665e-06, "loss": 0.013, "step": 46410 }, { "epoch": 13.505964503927844, "grad_norm": 0.09003158658742905, "learning_rate": 4.009605682418013e-06, "loss": 0.0124, "step": 46420 }, { "epoch": 13.508874018038988, "grad_norm": 0.1027403473854065, "learning_rate": 4.003980930243173e-06, "loss": 0.0122, "step": 46430 }, { "epoch": 13.511783532150131, "grad_norm": 0.09031980484724045, "learning_rate": 3.998359518368594e-06, "loss": 0.0127, "step": 46440 }, { "epoch": 13.514693046261275, "grad_norm": 0.10642839968204498, "learning_rate": 3.992741448501912e-06, "loss": 0.0123, "step": 46450 }, { "epoch": 13.517602560372417, "grad_norm": 0.08469890803098679, "learning_rate": 3.987126722349748e-06, "loss": 0.0117, "step": 46460 }, { "epoch": 13.52051207448356, "grad_norm": 0.07412269711494446, "learning_rate": 3.981515341617706e-06, "loss": 0.0147, "step": 46470 }, { "epoch": 13.523421588594704, "grad_norm": 0.08564141392707825, "learning_rate": 3.975907308010375e-06, "loss": 0.0123, "step": 46480 }, { "epoch": 13.526331102705848, "grad_norm": 0.12104244530200958, "learning_rate": 3.970302623231336e-06, "loss": 0.013, "step": 46490 }, { "epoch": 13.529240616816992, "grad_norm": 0.10458209365606308, "learning_rate": 3.964701288983132e-06, "loss": 0.0115, "step": 46500 }, { "epoch": 13.532150130928136, "grad_norm": 0.08920092135667801, "learning_rate": 3.959103306967312e-06, "loss": 0.0132, "step": 46510 }, { "epoch": 13.53505964503928, "grad_norm": 0.11685715615749359, "learning_rate": 3.953508678884382e-06, "loss": 0.0128, "step": 46520 }, { "epoch": 13.537969159150421, "grad_norm": 0.09056691825389862, "learning_rate": 3.947917406433853e-06, "loss": 0.0123, "step": 46530 }, { "epoch": 13.540878673261565, "grad_norm": 0.09083396196365356, "learning_rate": 3.942329491314205e-06, "loss": 0.0122, "step": 46540 }, { "epoch": 13.543788187372709, "grad_norm": 0.09560509026050568, "learning_rate": 3.936744935222895e-06, "loss": 0.0126, "step": 46550 }, { "epoch": 13.546697701483852, "grad_norm": 0.09954461455345154, "learning_rate": 3.931163739856365e-06, "loss": 0.0124, "step": 46560 }, { "epoch": 13.549607215594996, "grad_norm": 0.0800323560833931, "learning_rate": 3.9255859069100315e-06, "loss": 0.0135, "step": 46570 }, { "epoch": 13.55251672970614, "grad_norm": 0.1065642461180687, "learning_rate": 3.9200114380783005e-06, "loss": 0.0124, "step": 46580 }, { "epoch": 13.555426243817283, "grad_norm": 0.07882324606180191, "learning_rate": 3.914440335054544e-06, "loss": 0.011, "step": 46590 }, { "epoch": 13.558335757928425, "grad_norm": 0.09654084593057632, "learning_rate": 3.908872599531118e-06, "loss": 0.0129, "step": 46600 }, { "epoch": 13.561245272039569, "grad_norm": 0.09270605444908142, "learning_rate": 3.903308233199351e-06, "loss": 0.0121, "step": 46610 }, { "epoch": 13.564154786150713, "grad_norm": 0.0893683210015297, "learning_rate": 3.897747237749552e-06, "loss": 0.0115, "step": 46620 }, { "epoch": 13.567064300261857, "grad_norm": 0.13637372851371765, "learning_rate": 3.892189614870999e-06, "loss": 0.0126, "step": 46630 }, { "epoch": 13.569973814373, "grad_norm": 0.08941811323165894, "learning_rate": 3.886635366251966e-06, "loss": 0.0114, "step": 46640 }, { "epoch": 13.572883328484144, "grad_norm": 0.0834338515996933, "learning_rate": 3.8810844935796665e-06, "loss": 0.0126, "step": 46650 }, { "epoch": 13.575792842595286, "grad_norm": 0.07401090115308762, "learning_rate": 3.875536998540331e-06, "loss": 0.012, "step": 46660 }, { "epoch": 13.57870235670643, "grad_norm": 0.12284984439611435, "learning_rate": 3.86999288281912e-06, "loss": 0.015, "step": 46670 }, { "epoch": 13.581611870817573, "grad_norm": 0.08899541199207306, "learning_rate": 3.864452148100206e-06, "loss": 0.0132, "step": 46680 }, { "epoch": 13.584521384928717, "grad_norm": 0.08465959131717682, "learning_rate": 3.858914796066714e-06, "loss": 0.0143, "step": 46690 }, { "epoch": 13.58743089903986, "grad_norm": 0.13404329121112823, "learning_rate": 3.853380828400741e-06, "loss": 0.0111, "step": 46700 }, { "epoch": 13.590340413151004, "grad_norm": 0.09317620098590851, "learning_rate": 3.847850246783365e-06, "loss": 0.0118, "step": 46710 }, { "epoch": 13.593249927262146, "grad_norm": 0.07577964663505554, "learning_rate": 3.842323052894623e-06, "loss": 0.0122, "step": 46720 }, { "epoch": 13.59615944137329, "grad_norm": 0.11759639531373978, "learning_rate": 3.8367992484135444e-06, "loss": 0.015, "step": 46730 }, { "epoch": 13.599068955484434, "grad_norm": 0.09693257510662079, "learning_rate": 3.831278835018099e-06, "loss": 0.0141, "step": 46740 }, { "epoch": 13.601978469595577, "grad_norm": 0.1048366129398346, "learning_rate": 3.825761814385255e-06, "loss": 0.0137, "step": 46750 }, { "epoch": 13.604887983706721, "grad_norm": 0.08691791445016861, "learning_rate": 3.820248188190933e-06, "loss": 0.011, "step": 46760 }, { "epoch": 13.607797497817865, "grad_norm": 0.1072593405842781, "learning_rate": 3.814737958110026e-06, "loss": 0.0115, "step": 46770 }, { "epoch": 13.610707011929009, "grad_norm": 0.10652776807546616, "learning_rate": 3.809231125816399e-06, "loss": 0.0127, "step": 46780 }, { "epoch": 13.61361652604015, "grad_norm": 0.09538546204566956, "learning_rate": 3.8037276929828807e-06, "loss": 0.014, "step": 46790 }, { "epoch": 13.616526040151294, "grad_norm": 0.07726000249385834, "learning_rate": 3.798227661281265e-06, "loss": 0.012, "step": 46800 }, { "epoch": 13.619435554262438, "grad_norm": 0.07493803650140762, "learning_rate": 3.7927310323823274e-06, "loss": 0.0116, "step": 46810 }, { "epoch": 13.622345068373582, "grad_norm": 0.10798832774162292, "learning_rate": 3.7872378079557835e-06, "loss": 0.0121, "step": 46820 }, { "epoch": 13.625254582484725, "grad_norm": 0.10255969315767288, "learning_rate": 3.7817479896703424e-06, "loss": 0.0129, "step": 46830 }, { "epoch": 13.628164096595869, "grad_norm": 0.10472390055656433, "learning_rate": 3.7762615791936595e-06, "loss": 0.0114, "step": 46840 }, { "epoch": 13.631073610707013, "grad_norm": 0.08402520418167114, "learning_rate": 3.7707785781923654e-06, "loss": 0.0126, "step": 46850 }, { "epoch": 13.633983124818155, "grad_norm": 0.09076684713363647, "learning_rate": 3.765298988332048e-06, "loss": 0.0127, "step": 46860 }, { "epoch": 13.636892638929298, "grad_norm": 0.09210098534822464, "learning_rate": 3.7598228112772583e-06, "loss": 0.0131, "step": 46870 }, { "epoch": 13.639802153040442, "grad_norm": 0.08757917582988739, "learning_rate": 3.7543500486915262e-06, "loss": 0.0125, "step": 46880 }, { "epoch": 13.642711667151586, "grad_norm": 0.07861677557229996, "learning_rate": 3.748880702237318e-06, "loss": 0.013, "step": 46890 }, { "epoch": 13.64562118126273, "grad_norm": 0.10221253335475922, "learning_rate": 3.7434147735760876e-06, "loss": 0.0124, "step": 46900 }, { "epoch": 13.648530695373873, "grad_norm": 0.09148052334785461, "learning_rate": 3.7379522643682375e-06, "loss": 0.0127, "step": 46910 }, { "epoch": 13.651440209485017, "grad_norm": 0.08230800926685333, "learning_rate": 3.73249317627313e-06, "loss": 0.0114, "step": 46920 }, { "epoch": 13.654349723596159, "grad_norm": 0.0893329456448555, "learning_rate": 3.727037510949096e-06, "loss": 0.0133, "step": 46930 }, { "epoch": 13.657259237707303, "grad_norm": 0.08656566590070724, "learning_rate": 3.7215852700534213e-06, "loss": 0.0122, "step": 46940 }, { "epoch": 13.660168751818446, "grad_norm": 0.08813757449388504, "learning_rate": 3.716136455242346e-06, "loss": 0.0119, "step": 46950 }, { "epoch": 13.66307826592959, "grad_norm": 0.09934665262699127, "learning_rate": 3.7106910681710927e-06, "loss": 0.0132, "step": 46960 }, { "epoch": 13.665987780040734, "grad_norm": 0.09472576528787613, "learning_rate": 3.7052491104938064e-06, "loss": 0.0126, "step": 46970 }, { "epoch": 13.668897294151877, "grad_norm": 0.08598995208740234, "learning_rate": 3.6998105838636247e-06, "loss": 0.0127, "step": 46980 }, { "epoch": 13.67180680826302, "grad_norm": 0.08661438524723053, "learning_rate": 3.694375489932623e-06, "loss": 0.0122, "step": 46990 }, { "epoch": 13.674716322374163, "grad_norm": 0.08888706564903259, "learning_rate": 3.688943830351841e-06, "loss": 0.0124, "step": 47000 }, { "epoch": 13.677625836485307, "grad_norm": 0.08454279601573944, "learning_rate": 3.6835156067712693e-06, "loss": 0.0123, "step": 47010 }, { "epoch": 13.68053535059645, "grad_norm": 0.12606288492679596, "learning_rate": 3.67809082083986e-06, "loss": 0.0109, "step": 47020 }, { "epoch": 13.683444864707594, "grad_norm": 0.13668812811374664, "learning_rate": 3.6726694742055296e-06, "loss": 0.0125, "step": 47030 }, { "epoch": 13.686354378818738, "grad_norm": 0.12243428826332092, "learning_rate": 3.6672515685151236e-06, "loss": 0.0148, "step": 47040 }, { "epoch": 13.68926389292988, "grad_norm": 0.08057878911495209, "learning_rate": 3.661837105414471e-06, "loss": 0.0132, "step": 47050 }, { "epoch": 13.692173407041023, "grad_norm": 0.07550130784511566, "learning_rate": 3.6564260865483397e-06, "loss": 0.0132, "step": 47060 }, { "epoch": 13.695082921152167, "grad_norm": 0.09891857951879501, "learning_rate": 3.6510185135604525e-06, "loss": 0.0118, "step": 47070 }, { "epoch": 13.697992435263311, "grad_norm": 0.123092882335186, "learning_rate": 3.6456143880934887e-06, "loss": 0.0129, "step": 47080 }, { "epoch": 13.700901949374455, "grad_norm": 0.10836941748857498, "learning_rate": 3.6402137117890796e-06, "loss": 0.0115, "step": 47090 }, { "epoch": 13.703811463485598, "grad_norm": 0.09368189424276352, "learning_rate": 3.6348164862878063e-06, "loss": 0.0123, "step": 47100 }, { "epoch": 13.706720977596742, "grad_norm": 0.09559754282236099, "learning_rate": 3.6294227132292034e-06, "loss": 0.0133, "step": 47110 }, { "epoch": 13.709630491707884, "grad_norm": 0.07533422857522964, "learning_rate": 3.624032394251755e-06, "loss": 0.0117, "step": 47120 }, { "epoch": 13.712540005819028, "grad_norm": 0.13426394760608673, "learning_rate": 3.6186455309929028e-06, "loss": 0.0135, "step": 47130 }, { "epoch": 13.715449519930171, "grad_norm": 0.08486177027225494, "learning_rate": 3.613262125089033e-06, "loss": 0.0117, "step": 47140 }, { "epoch": 13.718359034041315, "grad_norm": 0.09037872403860092, "learning_rate": 3.607882178175479e-06, "loss": 0.013, "step": 47150 }, { "epoch": 13.721268548152459, "grad_norm": 0.07810931652784348, "learning_rate": 3.602505691886527e-06, "loss": 0.0128, "step": 47160 }, { "epoch": 13.724178062263602, "grad_norm": 0.12115877866744995, "learning_rate": 3.5971326678554083e-06, "loss": 0.0141, "step": 47170 }, { "epoch": 13.727087576374746, "grad_norm": 0.08036651462316513, "learning_rate": 3.59176310771432e-06, "loss": 0.012, "step": 47180 }, { "epoch": 13.729997090485888, "grad_norm": 0.11252550035715103, "learning_rate": 3.5863970130943728e-06, "loss": 0.0113, "step": 47190 }, { "epoch": 13.732906604597032, "grad_norm": 0.08598749339580536, "learning_rate": 3.5810343856256624e-06, "loss": 0.0126, "step": 47200 }, { "epoch": 13.735816118708176, "grad_norm": 0.14733457565307617, "learning_rate": 3.575675226937198e-06, "loss": 0.013, "step": 47210 }, { "epoch": 13.73872563281932, "grad_norm": 0.07156306505203247, "learning_rate": 3.5703195386569607e-06, "loss": 0.0123, "step": 47220 }, { "epoch": 13.741635146930463, "grad_norm": 0.07598625868558884, "learning_rate": 3.564967322411865e-06, "loss": 0.0126, "step": 47230 }, { "epoch": 13.744544661041607, "grad_norm": 0.11035879701375961, "learning_rate": 3.559618579827773e-06, "loss": 0.0121, "step": 47240 }, { "epoch": 13.74745417515275, "grad_norm": 0.10367942601442337, "learning_rate": 3.5542733125294902e-06, "loss": 0.0134, "step": 47250 }, { "epoch": 13.750363689263892, "grad_norm": 0.08922328054904938, "learning_rate": 3.5489315221407674e-06, "loss": 0.0146, "step": 47260 }, { "epoch": 13.753273203375036, "grad_norm": 0.08284555375576019, "learning_rate": 3.5435932102842973e-06, "loss": 0.0118, "step": 47270 }, { "epoch": 13.75618271748618, "grad_norm": 0.08350560814142227, "learning_rate": 3.5382583785817237e-06, "loss": 0.0125, "step": 47280 }, { "epoch": 13.759092231597323, "grad_norm": 0.0878366231918335, "learning_rate": 3.532927028653627e-06, "loss": 0.0116, "step": 47290 }, { "epoch": 13.762001745708467, "grad_norm": 0.06790715456008911, "learning_rate": 3.527599162119527e-06, "loss": 0.0133, "step": 47300 }, { "epoch": 13.76491125981961, "grad_norm": 0.0862140953540802, "learning_rate": 3.5222747805978906e-06, "loss": 0.0127, "step": 47310 }, { "epoch": 13.767820773930753, "grad_norm": 0.09038162231445312, "learning_rate": 3.516953885706119e-06, "loss": 0.0123, "step": 47320 }, { "epoch": 13.770730288041896, "grad_norm": 0.11455653607845306, "learning_rate": 3.511636479060574e-06, "loss": 0.0139, "step": 47330 }, { "epoch": 13.77363980215304, "grad_norm": 0.10324342548847198, "learning_rate": 3.506322562276526e-06, "loss": 0.0117, "step": 47340 }, { "epoch": 13.776549316264184, "grad_norm": 0.09814798831939697, "learning_rate": 3.501012136968218e-06, "loss": 0.0141, "step": 47350 }, { "epoch": 13.779458830375328, "grad_norm": 0.09621940553188324, "learning_rate": 3.4957052047488013e-06, "loss": 0.013, "step": 47360 }, { "epoch": 13.782368344486471, "grad_norm": 0.1099582239985466, "learning_rate": 3.490401767230395e-06, "loss": 0.0122, "step": 47370 }, { "epoch": 13.785277858597615, "grad_norm": 0.10023946315050125, "learning_rate": 3.4851018260240387e-06, "loss": 0.0122, "step": 47380 }, { "epoch": 13.788187372708757, "grad_norm": 0.22001248598098755, "learning_rate": 3.4798053827397142e-06, "loss": 0.0124, "step": 47390 }, { "epoch": 13.7910968868199, "grad_norm": 0.14375802874565125, "learning_rate": 3.4745124389863397e-06, "loss": 0.0128, "step": 47400 }, { "epoch": 13.794006400931044, "grad_norm": 0.09864316880702972, "learning_rate": 3.469222996371776e-06, "loss": 0.0129, "step": 47410 }, { "epoch": 13.796915915042188, "grad_norm": 0.06890972703695297, "learning_rate": 3.4639370565028074e-06, "loss": 0.0114, "step": 47420 }, { "epoch": 13.799825429153332, "grad_norm": 0.10436568409204483, "learning_rate": 3.4586546209851744e-06, "loss": 0.0113, "step": 47430 }, { "epoch": 13.802734943264475, "grad_norm": 0.11533033102750778, "learning_rate": 3.4533756914235337e-06, "loss": 0.0117, "step": 47440 }, { "epoch": 13.805644457375617, "grad_norm": 0.09739893674850464, "learning_rate": 3.448100269421489e-06, "loss": 0.0125, "step": 47450 }, { "epoch": 13.808553971486761, "grad_norm": 0.07328145205974579, "learning_rate": 3.4428283565815722e-06, "loss": 0.0134, "step": 47460 }, { "epoch": 13.811463485597905, "grad_norm": 0.11108795553445816, "learning_rate": 3.4375599545052498e-06, "loss": 0.0129, "step": 47470 }, { "epoch": 13.814372999709049, "grad_norm": 0.11606533825397491, "learning_rate": 3.432295064792926e-06, "loss": 0.0123, "step": 47480 }, { "epoch": 13.817282513820192, "grad_norm": 0.10358081012964249, "learning_rate": 3.42703368904393e-06, "loss": 0.0114, "step": 47490 }, { "epoch": 13.820192027931336, "grad_norm": 0.07268301397562027, "learning_rate": 3.42177582885654e-06, "loss": 0.0125, "step": 47500 }, { "epoch": 13.82310154204248, "grad_norm": 0.10789747536182404, "learning_rate": 3.41652148582794e-06, "loss": 0.0126, "step": 47510 }, { "epoch": 13.826011056153622, "grad_norm": 0.08018113672733307, "learning_rate": 3.4112706615542764e-06, "loss": 0.0112, "step": 47520 }, { "epoch": 13.828920570264765, "grad_norm": 0.09617459028959274, "learning_rate": 3.406023357630596e-06, "loss": 0.0129, "step": 47530 }, { "epoch": 13.831830084375909, "grad_norm": 0.07956665009260178, "learning_rate": 3.400779575650902e-06, "loss": 0.0118, "step": 47540 }, { "epoch": 13.834739598487053, "grad_norm": 0.08649753034114838, "learning_rate": 3.3955393172081136e-06, "loss": 0.0125, "step": 47550 }, { "epoch": 13.837649112598196, "grad_norm": 0.12008676677942276, "learning_rate": 3.3903025838940773e-06, "loss": 0.013, "step": 47560 }, { "epoch": 13.84055862670934, "grad_norm": 0.06987620890140533, "learning_rate": 3.385069377299589e-06, "loss": 0.0119, "step": 47570 }, { "epoch": 13.843468140820484, "grad_norm": 0.10017253458499908, "learning_rate": 3.3798396990143406e-06, "loss": 0.0113, "step": 47580 }, { "epoch": 13.846377654931626, "grad_norm": 0.07663611322641373, "learning_rate": 3.3746135506269845e-06, "loss": 0.0136, "step": 47590 }, { "epoch": 13.84928716904277, "grad_norm": 0.11780750006437302, "learning_rate": 3.3693909337250788e-06, "loss": 0.0124, "step": 47600 }, { "epoch": 13.852196683153913, "grad_norm": 0.0923454537987709, "learning_rate": 3.36417184989512e-06, "loss": 0.0126, "step": 47610 }, { "epoch": 13.855106197265057, "grad_norm": 0.14267699420452118, "learning_rate": 3.3589563007225282e-06, "loss": 0.0128, "step": 47620 }, { "epoch": 13.8580157113762, "grad_norm": 0.10692807286977768, "learning_rate": 3.3537442877916476e-06, "loss": 0.0125, "step": 47630 }, { "epoch": 13.860925225487344, "grad_norm": 0.11003539711236954, "learning_rate": 3.348535812685746e-06, "loss": 0.0134, "step": 47640 }, { "epoch": 13.863834739598486, "grad_norm": 0.08924530446529388, "learning_rate": 3.343330876987034e-06, "loss": 0.0116, "step": 47650 }, { "epoch": 13.86674425370963, "grad_norm": 0.11871614307165146, "learning_rate": 3.338129482276618e-06, "loss": 0.0141, "step": 47660 }, { "epoch": 13.869653767820774, "grad_norm": 0.09399480372667313, "learning_rate": 3.332931630134554e-06, "loss": 0.0114, "step": 47670 }, { "epoch": 13.872563281931917, "grad_norm": 0.09099527448415756, "learning_rate": 3.32773732213981e-06, "loss": 0.0125, "step": 47680 }, { "epoch": 13.875472796043061, "grad_norm": 0.10478828847408295, "learning_rate": 3.3225465598702797e-06, "loss": 0.0121, "step": 47690 }, { "epoch": 13.878382310154205, "grad_norm": 0.09195948392152786, "learning_rate": 3.3173593449027784e-06, "loss": 0.0127, "step": 47700 }, { "epoch": 13.881291824265348, "grad_norm": 0.1820259541273117, "learning_rate": 3.312175678813042e-06, "loss": 0.0114, "step": 47710 }, { "epoch": 13.88420133837649, "grad_norm": 0.07370330393314362, "learning_rate": 3.306995563175744e-06, "loss": 0.012, "step": 47720 }, { "epoch": 13.887110852487634, "grad_norm": 0.08692286163568497, "learning_rate": 3.30181899956445e-06, "loss": 0.0123, "step": 47730 }, { "epoch": 13.890020366598778, "grad_norm": 0.08680874854326248, "learning_rate": 3.2966459895516744e-06, "loss": 0.0118, "step": 47740 }, { "epoch": 13.892929880709922, "grad_norm": 0.08566546440124512, "learning_rate": 3.2914765347088386e-06, "loss": 0.0119, "step": 47750 }, { "epoch": 13.895839394821065, "grad_norm": 0.1030484065413475, "learning_rate": 3.2863106366062857e-06, "loss": 0.0121, "step": 47760 }, { "epoch": 13.898748908932209, "grad_norm": 0.09087258577346802, "learning_rate": 3.2811482968132796e-06, "loss": 0.0122, "step": 47770 }, { "epoch": 13.90165842304335, "grad_norm": 0.08237061649560928, "learning_rate": 3.2759895168980025e-06, "loss": 0.0138, "step": 47780 }, { "epoch": 13.904567937154495, "grad_norm": 0.09325297921895981, "learning_rate": 3.2708342984275515e-06, "loss": 0.0122, "step": 47790 }, { "epoch": 13.907477451265638, "grad_norm": 0.10287459939718246, "learning_rate": 3.265682642967958e-06, "loss": 0.0121, "step": 47800 }, { "epoch": 13.910386965376782, "grad_norm": 0.12751950323581696, "learning_rate": 3.260534552084143e-06, "loss": 0.0112, "step": 47810 }, { "epoch": 13.913296479487926, "grad_norm": 0.07207082211971283, "learning_rate": 3.255390027339973e-06, "loss": 0.0117, "step": 47820 }, { "epoch": 13.91620599359907, "grad_norm": 0.12366405129432678, "learning_rate": 3.250249070298213e-06, "loss": 0.013, "step": 47830 }, { "epoch": 13.919115507710213, "grad_norm": 0.08718981593847275, "learning_rate": 3.2451116825205524e-06, "loss": 0.0149, "step": 47840 }, { "epoch": 13.922025021821355, "grad_norm": 0.08421971648931503, "learning_rate": 3.2399778655675925e-06, "loss": 0.013, "step": 47850 }, { "epoch": 13.924934535932499, "grad_norm": 0.13246065378189087, "learning_rate": 3.2348476209988465e-06, "loss": 0.0123, "step": 47860 }, { "epoch": 13.927844050043642, "grad_norm": 0.11118330806493759, "learning_rate": 3.229720950372762e-06, "loss": 0.0132, "step": 47870 }, { "epoch": 13.930753564154786, "grad_norm": 0.09294462203979492, "learning_rate": 3.224597855246667e-06, "loss": 0.0113, "step": 47880 }, { "epoch": 13.93366307826593, "grad_norm": 0.08439967036247253, "learning_rate": 3.2194783371768418e-06, "loss": 0.0119, "step": 47890 }, { "epoch": 13.936572592377074, "grad_norm": 0.0913238599896431, "learning_rate": 3.2143623977184414e-06, "loss": 0.0117, "step": 47900 }, { "epoch": 13.939482106488217, "grad_norm": 0.11754840612411499, "learning_rate": 3.209250038425569e-06, "loss": 0.0125, "step": 47910 }, { "epoch": 13.94239162059936, "grad_norm": 0.09598129242658615, "learning_rate": 3.2041412608512166e-06, "loss": 0.011, "step": 47920 }, { "epoch": 13.945301134710503, "grad_norm": 0.08388040959835052, "learning_rate": 3.199036066547298e-06, "loss": 0.0137, "step": 47930 }, { "epoch": 13.948210648821647, "grad_norm": 0.1017238199710846, "learning_rate": 3.193934457064636e-06, "loss": 0.0126, "step": 47940 }, { "epoch": 13.95112016293279, "grad_norm": 0.11763064563274384, "learning_rate": 3.188836433952965e-06, "loss": 0.0119, "step": 47950 }, { "epoch": 13.954029677043934, "grad_norm": 0.08544391393661499, "learning_rate": 3.1837419987609246e-06, "loss": 0.0119, "step": 47960 }, { "epoch": 13.956939191155078, "grad_norm": 0.10982999205589294, "learning_rate": 3.178651153036079e-06, "loss": 0.0132, "step": 47970 }, { "epoch": 13.959848705266221, "grad_norm": 0.08350773900747299, "learning_rate": 3.1735638983248898e-06, "loss": 0.012, "step": 47980 }, { "epoch": 13.962758219377363, "grad_norm": 0.08319947123527527, "learning_rate": 3.168480236172728e-06, "loss": 0.0134, "step": 47990 }, { "epoch": 13.965667733488507, "grad_norm": 0.10572589188814163, "learning_rate": 3.163400168123877e-06, "loss": 0.0112, "step": 48000 }, { "epoch": 13.96857724759965, "grad_norm": 0.0807105228304863, "learning_rate": 3.1583236957215255e-06, "loss": 0.0118, "step": 48010 }, { "epoch": 13.971486761710795, "grad_norm": 0.09853789210319519, "learning_rate": 3.153250820507781e-06, "loss": 0.0127, "step": 48020 }, { "epoch": 13.974396275821938, "grad_norm": 0.09417697042226791, "learning_rate": 3.148181544023634e-06, "loss": 0.0105, "step": 48030 }, { "epoch": 13.977305789933082, "grad_norm": 0.09016052633523941, "learning_rate": 3.1431158678090127e-06, "loss": 0.012, "step": 48040 }, { "epoch": 13.980215304044224, "grad_norm": 0.08901971578598022, "learning_rate": 3.138053793402718e-06, "loss": 0.0114, "step": 48050 }, { "epoch": 13.983124818155368, "grad_norm": 0.10354455560445786, "learning_rate": 3.1329953223424902e-06, "loss": 0.0124, "step": 48060 }, { "epoch": 13.986034332266511, "grad_norm": 0.08666756004095078, "learning_rate": 3.1279404561649545e-06, "loss": 0.0109, "step": 48070 }, { "epoch": 13.988943846377655, "grad_norm": 0.08093476295471191, "learning_rate": 3.1228891964056428e-06, "loss": 0.0102, "step": 48080 }, { "epoch": 13.991853360488799, "grad_norm": 0.10681623220443726, "learning_rate": 3.1178415445989963e-06, "loss": 0.0114, "step": 48090 }, { "epoch": 13.994762874599942, "grad_norm": 0.12774765491485596, "learning_rate": 3.1127975022783593e-06, "loss": 0.0126, "step": 48100 }, { "epoch": 13.997672388711084, "grad_norm": 0.07565969228744507, "learning_rate": 3.107757070975972e-06, "loss": 0.0119, "step": 48110 }, { "epoch": 14.000581902822228, "grad_norm": 0.0736968144774437, "learning_rate": 3.1027202522229964e-06, "loss": 0.0119, "step": 48120 }, { "epoch": 14.003491416933372, "grad_norm": 0.11915725469589233, "learning_rate": 3.097687047549478e-06, "loss": 0.0125, "step": 48130 }, { "epoch": 14.006400931044515, "grad_norm": 0.10019274801015854, "learning_rate": 3.092657458484372e-06, "loss": 0.0122, "step": 48140 }, { "epoch": 14.00931044515566, "grad_norm": 0.09482284635305405, "learning_rate": 3.0876314865555333e-06, "loss": 0.0123, "step": 48150 }, { "epoch": 14.012219959266803, "grad_norm": 0.09677743166685104, "learning_rate": 3.0826091332897186e-06, "loss": 0.0115, "step": 48160 }, { "epoch": 14.015129473377947, "grad_norm": 0.11471506953239441, "learning_rate": 3.0775904002125967e-06, "loss": 0.0133, "step": 48170 }, { "epoch": 14.018038987489088, "grad_norm": 0.1066441684961319, "learning_rate": 3.0725752888487102e-06, "loss": 0.0117, "step": 48180 }, { "epoch": 14.020948501600232, "grad_norm": 0.12184887379407883, "learning_rate": 3.0675638007215345e-06, "loss": 0.0103, "step": 48190 }, { "epoch": 14.023858015711376, "grad_norm": 0.08194026350975037, "learning_rate": 3.0625559373534082e-06, "loss": 0.0122, "step": 48200 }, { "epoch": 14.02676752982252, "grad_norm": 0.10446789115667343, "learning_rate": 3.057551700265605e-06, "loss": 0.0119, "step": 48210 }, { "epoch": 14.029677043933663, "grad_norm": 0.09155169874429703, "learning_rate": 3.052551090978272e-06, "loss": 0.0129, "step": 48220 }, { "epoch": 14.032586558044807, "grad_norm": 0.07928315550088882, "learning_rate": 3.0475541110104616e-06, "loss": 0.0129, "step": 48230 }, { "epoch": 14.03549607215595, "grad_norm": 0.09387852251529694, "learning_rate": 3.042560761880129e-06, "loss": 0.0132, "step": 48240 }, { "epoch": 14.038405586267093, "grad_norm": 0.07945214956998825, "learning_rate": 3.0375710451041167e-06, "loss": 0.0128, "step": 48250 }, { "epoch": 14.041315100378236, "grad_norm": 0.08431250602006912, "learning_rate": 3.032584962198168e-06, "loss": 0.0109, "step": 48260 }, { "epoch": 14.04422461448938, "grad_norm": 0.11698916554450989, "learning_rate": 3.0276025146769306e-06, "loss": 0.0112, "step": 48270 }, { "epoch": 14.047134128600524, "grad_norm": 0.08655864745378494, "learning_rate": 3.0226237040539375e-06, "loss": 0.0124, "step": 48280 }, { "epoch": 14.050043642711668, "grad_norm": 0.08688220381736755, "learning_rate": 3.017648531841617e-06, "loss": 0.0121, "step": 48290 }, { "epoch": 14.052953156822811, "grad_norm": 0.08723986148834229, "learning_rate": 3.0126769995512987e-06, "loss": 0.0115, "step": 48300 }, { "epoch": 14.055862670933955, "grad_norm": 0.10134736448526382, "learning_rate": 3.0077091086931995e-06, "loss": 0.0125, "step": 48310 }, { "epoch": 14.058772185045097, "grad_norm": 0.08199354261159897, "learning_rate": 3.002744860776436e-06, "loss": 0.0134, "step": 48320 }, { "epoch": 14.06168169915624, "grad_norm": 0.09822769463062286, "learning_rate": 2.9977842573090093e-06, "loss": 0.0135, "step": 48330 }, { "epoch": 14.064591213267384, "grad_norm": 0.0981333777308464, "learning_rate": 2.9928272997978347e-06, "loss": 0.0114, "step": 48340 }, { "epoch": 14.067500727378528, "grad_norm": 0.08265111595392227, "learning_rate": 2.9878739897486854e-06, "loss": 0.012, "step": 48350 }, { "epoch": 14.070410241489672, "grad_norm": 0.09852782636880875, "learning_rate": 2.9829243286662655e-06, "loss": 0.0113, "step": 48360 }, { "epoch": 14.073319755600815, "grad_norm": 0.10420690476894379, "learning_rate": 2.9779783180541348e-06, "loss": 0.0145, "step": 48370 }, { "epoch": 14.076229269711957, "grad_norm": 0.08506511896848679, "learning_rate": 2.9730359594147716e-06, "loss": 0.0126, "step": 48380 }, { "epoch": 14.079138783823101, "grad_norm": 0.095717653632164, "learning_rate": 2.96809725424953e-06, "loss": 0.0126, "step": 48390 }, { "epoch": 14.082048297934245, "grad_norm": 0.08196638524532318, "learning_rate": 2.9631622040586583e-06, "loss": 0.0122, "step": 48400 }, { "epoch": 14.084957812045388, "grad_norm": 0.10126378387212753, "learning_rate": 2.9582308103412938e-06, "loss": 0.0133, "step": 48410 }, { "epoch": 14.087867326156532, "grad_norm": 0.08585000038146973, "learning_rate": 2.953303074595463e-06, "loss": 0.0132, "step": 48420 }, { "epoch": 14.090776840267676, "grad_norm": 0.07186255604028702, "learning_rate": 2.9483789983180857e-06, "loss": 0.0151, "step": 48430 }, { "epoch": 14.09368635437882, "grad_norm": 0.10063635557889938, "learning_rate": 2.9434585830049637e-06, "loss": 0.0116, "step": 48440 }, { "epoch": 14.096595868489961, "grad_norm": 0.08523969352245331, "learning_rate": 2.938541830150791e-06, "loss": 0.0127, "step": 48450 }, { "epoch": 14.099505382601105, "grad_norm": 0.09360211342573166, "learning_rate": 2.9336287412491464e-06, "loss": 0.0142, "step": 48460 }, { "epoch": 14.102414896712249, "grad_norm": 0.08601976931095123, "learning_rate": 2.9287193177924943e-06, "loss": 0.0126, "step": 48470 }, { "epoch": 14.105324410823393, "grad_norm": 0.10072314739227295, "learning_rate": 2.9238135612721867e-06, "loss": 0.0129, "step": 48480 }, { "epoch": 14.108233924934536, "grad_norm": 0.06382299959659576, "learning_rate": 2.918911473178475e-06, "loss": 0.0124, "step": 48490 }, { "epoch": 14.11114343904568, "grad_norm": 0.13382671773433685, "learning_rate": 2.9140130550004663e-06, "loss": 0.0132, "step": 48500 }, { "epoch": 14.114052953156822, "grad_norm": 0.09437806904315948, "learning_rate": 2.9091183082261868e-06, "loss": 0.0137, "step": 48510 }, { "epoch": 14.116962467267966, "grad_norm": 0.07821691036224365, "learning_rate": 2.904227234342516e-06, "loss": 0.0114, "step": 48520 }, { "epoch": 14.11987198137911, "grad_norm": 0.07441290467977524, "learning_rate": 2.8993398348352463e-06, "loss": 0.0135, "step": 48530 }, { "epoch": 14.122781495490253, "grad_norm": 0.10905373096466064, "learning_rate": 2.8944561111890337e-06, "loss": 0.013, "step": 48540 }, { "epoch": 14.125691009601397, "grad_norm": 0.09184795618057251, "learning_rate": 2.889576064887427e-06, "loss": 0.0116, "step": 48550 }, { "epoch": 14.12860052371254, "grad_norm": 0.14006227254867554, "learning_rate": 2.8846996974128546e-06, "loss": 0.0114, "step": 48560 }, { "epoch": 14.131510037823684, "grad_norm": 0.12471664696931839, "learning_rate": 2.879827010246624e-06, "loss": 0.0116, "step": 48570 }, { "epoch": 14.134419551934826, "grad_norm": 0.08797192573547363, "learning_rate": 2.8749580048689373e-06, "loss": 0.011, "step": 48580 }, { "epoch": 14.13732906604597, "grad_norm": 0.1171896904706955, "learning_rate": 2.8700926827588646e-06, "loss": 0.0143, "step": 48590 }, { "epoch": 14.140238580157114, "grad_norm": 0.10456521809101105, "learning_rate": 2.865231045394364e-06, "loss": 0.0127, "step": 48600 }, { "epoch": 14.143148094268257, "grad_norm": 0.07859615236520767, "learning_rate": 2.860373094252273e-06, "loss": 0.0119, "step": 48610 }, { "epoch": 14.146057608379401, "grad_norm": 0.15674468874931335, "learning_rate": 2.8555188308083065e-06, "loss": 0.0123, "step": 48620 }, { "epoch": 14.148967122490545, "grad_norm": 0.14042221009731293, "learning_rate": 2.850668256537059e-06, "loss": 0.0117, "step": 48630 }, { "epoch": 14.151876636601688, "grad_norm": 0.08551198244094849, "learning_rate": 2.8458213729120185e-06, "loss": 0.0129, "step": 48640 }, { "epoch": 14.15478615071283, "grad_norm": 0.09340638667345047, "learning_rate": 2.840978181405524e-06, "loss": 0.0128, "step": 48650 }, { "epoch": 14.157695664823974, "grad_norm": 0.08583881705999374, "learning_rate": 2.8361386834888253e-06, "loss": 0.0131, "step": 48660 }, { "epoch": 14.160605178935118, "grad_norm": 0.11229098588228226, "learning_rate": 2.8313028806320197e-06, "loss": 0.0127, "step": 48670 }, { "epoch": 14.163514693046261, "grad_norm": 0.11258826404809952, "learning_rate": 2.826470774304107e-06, "loss": 0.0114, "step": 48680 }, { "epoch": 14.166424207157405, "grad_norm": 0.11186370998620987, "learning_rate": 2.821642365972948e-06, "loss": 0.0145, "step": 48690 }, { "epoch": 14.169333721268549, "grad_norm": 0.09387464076280594, "learning_rate": 2.8168176571052894e-06, "loss": 0.0126, "step": 48700 }, { "epoch": 14.17224323537969, "grad_norm": 0.0917554497718811, "learning_rate": 2.811996649166746e-06, "loss": 0.0122, "step": 48710 }, { "epoch": 14.175152749490834, "grad_norm": 0.08259701728820801, "learning_rate": 2.8071793436218125e-06, "loss": 0.0121, "step": 48720 }, { "epoch": 14.178062263601978, "grad_norm": 0.10233868658542633, "learning_rate": 2.80236574193387e-06, "loss": 0.012, "step": 48730 }, { "epoch": 14.180971777713122, "grad_norm": 0.08348140865564346, "learning_rate": 2.7975558455651475e-06, "loss": 0.0119, "step": 48740 }, { "epoch": 14.183881291824266, "grad_norm": 0.08227545768022537, "learning_rate": 2.792749655976774e-06, "loss": 0.0121, "step": 48750 }, { "epoch": 14.18679080593541, "grad_norm": 0.08271995931863785, "learning_rate": 2.787947174628744e-06, "loss": 0.0116, "step": 48760 }, { "epoch": 14.189700320046553, "grad_norm": 0.09419432282447815, "learning_rate": 2.78314840297992e-06, "loss": 0.0124, "step": 48770 }, { "epoch": 14.192609834157695, "grad_norm": 0.10533536970615387, "learning_rate": 2.778353342488044e-06, "loss": 0.0123, "step": 48780 }, { "epoch": 14.195519348268839, "grad_norm": 0.08804851025342941, "learning_rate": 2.7735619946097286e-06, "loss": 0.0116, "step": 48790 }, { "epoch": 14.198428862379982, "grad_norm": 0.11863476037979126, "learning_rate": 2.7687743608004553e-06, "loss": 0.0123, "step": 48800 }, { "epoch": 14.201338376491126, "grad_norm": 0.09537691622972488, "learning_rate": 2.7639904425145924e-06, "loss": 0.0127, "step": 48810 }, { "epoch": 14.20424789060227, "grad_norm": 0.07826381176710129, "learning_rate": 2.759210241205351e-06, "loss": 0.0116, "step": 48820 }, { "epoch": 14.207157404713413, "grad_norm": 0.10188067704439163, "learning_rate": 2.754433758324845e-06, "loss": 0.011, "step": 48830 }, { "epoch": 14.210066918824555, "grad_norm": 0.08754003047943115, "learning_rate": 2.749660995324036e-06, "loss": 0.0137, "step": 48840 }, { "epoch": 14.2129764329357, "grad_norm": 0.09673837572336197, "learning_rate": 2.744891953652767e-06, "loss": 0.0118, "step": 48850 }, { "epoch": 14.215885947046843, "grad_norm": 0.12430671602487564, "learning_rate": 2.7401266347597433e-06, "loss": 0.0111, "step": 48860 }, { "epoch": 14.218795461157987, "grad_norm": 0.09951858222484589, "learning_rate": 2.735365040092541e-06, "loss": 0.0129, "step": 48870 }, { "epoch": 14.22170497526913, "grad_norm": 0.10381476581096649, "learning_rate": 2.730607171097619e-06, "loss": 0.0124, "step": 48880 }, { "epoch": 14.224614489380274, "grad_norm": 0.08838944137096405, "learning_rate": 2.725853029220275e-06, "loss": 0.0118, "step": 48890 }, { "epoch": 14.227524003491418, "grad_norm": 0.11592337489128113, "learning_rate": 2.7211026159047052e-06, "loss": 0.0128, "step": 48900 }, { "epoch": 14.23043351760256, "grad_norm": 0.0742204487323761, "learning_rate": 2.7163559325939558e-06, "loss": 0.0129, "step": 48910 }, { "epoch": 14.233343031713703, "grad_norm": 0.08461632579565048, "learning_rate": 2.7116129807299426e-06, "loss": 0.0123, "step": 48920 }, { "epoch": 14.236252545824847, "grad_norm": 0.10865478217601776, "learning_rate": 2.7068737617534505e-06, "loss": 0.013, "step": 48930 }, { "epoch": 14.23916205993599, "grad_norm": 0.08279480040073395, "learning_rate": 2.7021382771041276e-06, "loss": 0.0103, "step": 48940 }, { "epoch": 14.242071574047134, "grad_norm": 0.09591922163963318, "learning_rate": 2.6974065282204862e-06, "loss": 0.0123, "step": 48950 }, { "epoch": 14.244981088158278, "grad_norm": 0.12336662411689758, "learning_rate": 2.6926785165399188e-06, "loss": 0.0131, "step": 48960 }, { "epoch": 14.247890602269422, "grad_norm": 0.09806191921234131, "learning_rate": 2.6879542434986543e-06, "loss": 0.0116, "step": 48970 }, { "epoch": 14.250800116380564, "grad_norm": 0.11384034156799316, "learning_rate": 2.683233710531813e-06, "loss": 0.0109, "step": 48980 }, { "epoch": 14.253709630491707, "grad_norm": 0.12452972680330276, "learning_rate": 2.6785169190733657e-06, "loss": 0.0124, "step": 48990 }, { "epoch": 14.256619144602851, "grad_norm": 0.1314302533864975, "learning_rate": 2.673803870556148e-06, "loss": 0.0125, "step": 49000 }, { "epoch": 14.259528658713995, "grad_norm": 0.08687698096036911, "learning_rate": 2.6690945664118604e-06, "loss": 0.0122, "step": 49010 }, { "epoch": 14.262438172825139, "grad_norm": 0.09671022742986679, "learning_rate": 2.6643890080710603e-06, "loss": 0.0141, "step": 49020 }, { "epoch": 14.265347686936282, "grad_norm": 0.10208301246166229, "learning_rate": 2.6596871969631843e-06, "loss": 0.0128, "step": 49030 }, { "epoch": 14.268257201047424, "grad_norm": 0.08651352673768997, "learning_rate": 2.6549891345165043e-06, "loss": 0.0129, "step": 49040 }, { "epoch": 14.271166715158568, "grad_norm": 0.097753144800663, "learning_rate": 2.650294822158176e-06, "loss": 0.0107, "step": 49050 }, { "epoch": 14.274076229269712, "grad_norm": 0.08194651454687119, "learning_rate": 2.645604261314207e-06, "loss": 0.0115, "step": 49060 }, { "epoch": 14.276985743380855, "grad_norm": 0.11899197846651077, "learning_rate": 2.6409174534094643e-06, "loss": 0.0121, "step": 49070 }, { "epoch": 14.279895257491999, "grad_norm": 0.09666430950164795, "learning_rate": 2.636234399867677e-06, "loss": 0.0111, "step": 49080 }, { "epoch": 14.282804771603143, "grad_norm": 0.08858638256788254, "learning_rate": 2.631555102111432e-06, "loss": 0.0108, "step": 49090 }, { "epoch": 14.285714285714286, "grad_norm": 0.10053882747888565, "learning_rate": 2.6268795615621775e-06, "loss": 0.0124, "step": 49100 }, { "epoch": 14.288623799825428, "grad_norm": 0.09149634093046188, "learning_rate": 2.6222077796402193e-06, "loss": 0.0123, "step": 49110 }, { "epoch": 14.291533313936572, "grad_norm": 0.09241379052400589, "learning_rate": 2.617539757764717e-06, "loss": 0.0113, "step": 49120 }, { "epoch": 14.294442828047716, "grad_norm": 0.0909649208188057, "learning_rate": 2.6128754973536993e-06, "loss": 0.012, "step": 49130 }, { "epoch": 14.29735234215886, "grad_norm": 0.06978762894868851, "learning_rate": 2.6082149998240434e-06, "loss": 0.0117, "step": 49140 }, { "epoch": 14.300261856270003, "grad_norm": 0.14550654590129852, "learning_rate": 2.603558266591483e-06, "loss": 0.0119, "step": 49150 }, { "epoch": 14.303171370381147, "grad_norm": 0.09133557230234146, "learning_rate": 2.5989052990706134e-06, "loss": 0.0128, "step": 49160 }, { "epoch": 14.306080884492289, "grad_norm": 0.08237214386463165, "learning_rate": 2.5942560986748774e-06, "loss": 0.0115, "step": 49170 }, { "epoch": 14.308990398603433, "grad_norm": 0.09728901088237762, "learning_rate": 2.589610666816592e-06, "loss": 0.0122, "step": 49180 }, { "epoch": 14.311899912714576, "grad_norm": 0.061919622123241425, "learning_rate": 2.5849690049069013e-06, "loss": 0.0134, "step": 49190 }, { "epoch": 14.31480942682572, "grad_norm": 0.07585946470499039, "learning_rate": 2.5803311143558344e-06, "loss": 0.0113, "step": 49200 }, { "epoch": 14.317718940936864, "grad_norm": 0.10361964255571365, "learning_rate": 2.575696996572246e-06, "loss": 0.013, "step": 49210 }, { "epoch": 14.320628455048007, "grad_norm": 0.09773951023817062, "learning_rate": 2.5710666529638667e-06, "loss": 0.012, "step": 49220 }, { "epoch": 14.323537969159151, "grad_norm": 0.06791174411773682, "learning_rate": 2.5664400849372727e-06, "loss": 0.0124, "step": 49230 }, { "epoch": 14.326447483270293, "grad_norm": 0.07662210613489151, "learning_rate": 2.561817293897892e-06, "loss": 0.011, "step": 49240 }, { "epoch": 14.329356997381437, "grad_norm": 0.08843982219696045, "learning_rate": 2.5571982812500054e-06, "loss": 0.0118, "step": 49250 }, { "epoch": 14.33226651149258, "grad_norm": 0.09876159578561783, "learning_rate": 2.55258304839675e-06, "loss": 0.0125, "step": 49260 }, { "epoch": 14.335176025603724, "grad_norm": 0.08149085193872452, "learning_rate": 2.5479715967401034e-06, "loss": 0.012, "step": 49270 }, { "epoch": 14.338085539714868, "grad_norm": 0.09744831174612045, "learning_rate": 2.543363927680914e-06, "loss": 0.0132, "step": 49280 }, { "epoch": 14.340995053826012, "grad_norm": 0.12228932231664658, "learning_rate": 2.5387600426188633e-06, "loss": 0.0119, "step": 49290 }, { "epoch": 14.343904567937155, "grad_norm": 0.10269774496555328, "learning_rate": 2.534159942952493e-06, "loss": 0.0128, "step": 49300 }, { "epoch": 14.346814082048297, "grad_norm": 0.11014590412378311, "learning_rate": 2.5295636300791903e-06, "loss": 0.0115, "step": 49310 }, { "epoch": 14.349723596159441, "grad_norm": 0.11011907458305359, "learning_rate": 2.5249711053951913e-06, "loss": 0.0121, "step": 49320 }, { "epoch": 14.352633110270585, "grad_norm": 0.08734544366598129, "learning_rate": 2.520382370295592e-06, "loss": 0.011, "step": 49330 }, { "epoch": 14.355542624381728, "grad_norm": 0.13002635538578033, "learning_rate": 2.515797426174317e-06, "loss": 0.0139, "step": 49340 }, { "epoch": 14.358452138492872, "grad_norm": 0.09142989665269852, "learning_rate": 2.5112162744241646e-06, "loss": 0.0126, "step": 49350 }, { "epoch": 14.361361652604016, "grad_norm": 0.0983654335141182, "learning_rate": 2.5066389164367543e-06, "loss": 0.0134, "step": 49360 }, { "epoch": 14.36427116671516, "grad_norm": 0.08271544426679611, "learning_rate": 2.5020653536025756e-06, "loss": 0.0119, "step": 49370 }, { "epoch": 14.367180680826301, "grad_norm": 0.0866219699382782, "learning_rate": 2.497495587310954e-06, "loss": 0.0123, "step": 49380 }, { "epoch": 14.370090194937445, "grad_norm": 0.10837266594171524, "learning_rate": 2.492929618950064e-06, "loss": 0.0124, "step": 49390 }, { "epoch": 14.372999709048589, "grad_norm": 0.1158163920044899, "learning_rate": 2.488367449906926e-06, "loss": 0.011, "step": 49400 }, { "epoch": 14.375909223159733, "grad_norm": 0.09807369858026505, "learning_rate": 2.4838090815674046e-06, "loss": 0.0123, "step": 49410 }, { "epoch": 14.378818737270876, "grad_norm": 0.10945294052362442, "learning_rate": 2.479254515316211e-06, "loss": 0.0121, "step": 49420 }, { "epoch": 14.38172825138202, "grad_norm": 0.07243960350751877, "learning_rate": 2.474703752536908e-06, "loss": 0.0125, "step": 49430 }, { "epoch": 14.384637765493162, "grad_norm": 0.1025518923997879, "learning_rate": 2.4701567946118935e-06, "loss": 0.0115, "step": 49440 }, { "epoch": 14.387547279604306, "grad_norm": 0.09270524978637695, "learning_rate": 2.4656136429224124e-06, "loss": 0.0122, "step": 49450 }, { "epoch": 14.39045679371545, "grad_norm": 0.10404076427221298, "learning_rate": 2.4610742988485574e-06, "loss": 0.0129, "step": 49460 }, { "epoch": 14.393366307826593, "grad_norm": 0.07205504924058914, "learning_rate": 2.456538763769254e-06, "loss": 0.0123, "step": 49470 }, { "epoch": 14.396275821937737, "grad_norm": 0.09685557335615158, "learning_rate": 2.452007039062291e-06, "loss": 0.0134, "step": 49480 }, { "epoch": 14.39918533604888, "grad_norm": 0.0910172238945961, "learning_rate": 2.4474791261042716e-06, "loss": 0.0117, "step": 49490 }, { "epoch": 14.402094850160024, "grad_norm": 0.10105447471141815, "learning_rate": 2.4429550262706718e-06, "loss": 0.0116, "step": 49500 }, { "epoch": 14.405004364271166, "grad_norm": 0.09095427393913269, "learning_rate": 2.438434740935777e-06, "loss": 0.0116, "step": 49510 }, { "epoch": 14.40791387838231, "grad_norm": 0.14599470794200897, "learning_rate": 2.4339182714727477e-06, "loss": 0.0123, "step": 49520 }, { "epoch": 14.410823392493453, "grad_norm": 0.1117706447839737, "learning_rate": 2.4294056192535506e-06, "loss": 0.0131, "step": 49530 }, { "epoch": 14.413732906604597, "grad_norm": 0.08479428291320801, "learning_rate": 2.4248967856490233e-06, "loss": 0.011, "step": 49540 }, { "epoch": 14.41664242071574, "grad_norm": 0.07124438881874084, "learning_rate": 2.420391772028825e-06, "loss": 0.0119, "step": 49550 }, { "epoch": 14.419551934826885, "grad_norm": 0.06601312756538391, "learning_rate": 2.415890579761462e-06, "loss": 0.0107, "step": 49560 }, { "epoch": 14.422461448938027, "grad_norm": 0.08749895542860031, "learning_rate": 2.4113932102142765e-06, "loss": 0.0121, "step": 49570 }, { "epoch": 14.42537096304917, "grad_norm": 0.10380292683839798, "learning_rate": 2.4068996647534488e-06, "loss": 0.012, "step": 49580 }, { "epoch": 14.428280477160314, "grad_norm": 0.0899677649140358, "learning_rate": 2.402409944744003e-06, "loss": 0.013, "step": 49590 }, { "epoch": 14.431189991271458, "grad_norm": 0.09553395211696625, "learning_rate": 2.397924051549797e-06, "loss": 0.011, "step": 49600 }, { "epoch": 14.434099505382601, "grad_norm": 0.1230107918381691, "learning_rate": 2.3934419865335256e-06, "loss": 0.0122, "step": 49610 }, { "epoch": 14.437009019493745, "grad_norm": 0.08626090735197067, "learning_rate": 2.3889637510567197e-06, "loss": 0.013, "step": 49620 }, { "epoch": 14.439918533604889, "grad_norm": 0.11734754592180252, "learning_rate": 2.384489346479753e-06, "loss": 0.0139, "step": 49630 }, { "epoch": 14.44282804771603, "grad_norm": 0.09132812172174454, "learning_rate": 2.3800187741618263e-06, "loss": 0.0125, "step": 49640 }, { "epoch": 14.445737561827174, "grad_norm": 0.07973501831293106, "learning_rate": 2.3755520354609917e-06, "loss": 0.0113, "step": 49650 }, { "epoch": 14.448647075938318, "grad_norm": 0.10823278874158859, "learning_rate": 2.371089131734112e-06, "loss": 0.0125, "step": 49660 }, { "epoch": 14.451556590049462, "grad_norm": 0.12876322865486145, "learning_rate": 2.366630064336913e-06, "loss": 0.0133, "step": 49670 }, { "epoch": 14.454466104160606, "grad_norm": 0.12106049805879593, "learning_rate": 2.3621748346239295e-06, "loss": 0.0121, "step": 49680 }, { "epoch": 14.45737561827175, "grad_norm": 0.07543538510799408, "learning_rate": 2.357723443948554e-06, "loss": 0.0117, "step": 49690 }, { "epoch": 14.460285132382893, "grad_norm": 0.07478784769773483, "learning_rate": 2.3532758936629944e-06, "loss": 0.0128, "step": 49700 }, { "epoch": 14.463194646494035, "grad_norm": 0.09989750385284424, "learning_rate": 2.3488321851183005e-06, "loss": 0.0122, "step": 49710 }, { "epoch": 14.466104160605179, "grad_norm": 0.09905678778886795, "learning_rate": 2.344392319664355e-06, "loss": 0.0136, "step": 49720 }, { "epoch": 14.469013674716322, "grad_norm": 0.10889994353055954, "learning_rate": 2.3399562986498667e-06, "loss": 0.0131, "step": 49730 }, { "epoch": 14.471923188827466, "grad_norm": 0.10161155462265015, "learning_rate": 2.3355241234223906e-06, "loss": 0.0139, "step": 49740 }, { "epoch": 14.47483270293861, "grad_norm": 0.06844766438007355, "learning_rate": 2.331095795328298e-06, "loss": 0.0109, "step": 49750 }, { "epoch": 14.477742217049753, "grad_norm": 0.13794435560703278, "learning_rate": 2.326671315712801e-06, "loss": 0.0133, "step": 49760 }, { "epoch": 14.480651731160895, "grad_norm": 0.11344872415065765, "learning_rate": 2.3222506859199383e-06, "loss": 0.0121, "step": 49770 }, { "epoch": 14.483561245272039, "grad_norm": 0.09257906675338745, "learning_rate": 2.3178339072925804e-06, "loss": 0.0137, "step": 49780 }, { "epoch": 14.486470759383183, "grad_norm": 0.10086268186569214, "learning_rate": 2.3134209811724237e-06, "loss": 0.0133, "step": 49790 }, { "epoch": 14.489380273494326, "grad_norm": 0.1076267808675766, "learning_rate": 2.3090119089000117e-06, "loss": 0.0114, "step": 49800 }, { "epoch": 14.49228978760547, "grad_norm": 0.08745689690113068, "learning_rate": 2.3046066918146875e-06, "loss": 0.0111, "step": 49810 }, { "epoch": 14.495199301716614, "grad_norm": 0.13114485144615173, "learning_rate": 2.3002053312546558e-06, "loss": 0.0146, "step": 49820 }, { "epoch": 14.498108815827758, "grad_norm": 0.11664741486310959, "learning_rate": 2.2958078285569184e-06, "loss": 0.0121, "step": 49830 }, { "epoch": 14.5010183299389, "grad_norm": 0.09297125786542892, "learning_rate": 2.291414185057332e-06, "loss": 0.0119, "step": 49840 }, { "epoch": 14.503927844050043, "grad_norm": 0.09377152472734451, "learning_rate": 2.2870244020905644e-06, "loss": 0.0112, "step": 49850 }, { "epoch": 14.506837358161187, "grad_norm": 0.10018522292375565, "learning_rate": 2.282638480990117e-06, "loss": 0.0117, "step": 49860 }, { "epoch": 14.50974687227233, "grad_norm": 0.09682188928127289, "learning_rate": 2.278256423088317e-06, "loss": 0.0117, "step": 49870 }, { "epoch": 14.512656386383474, "grad_norm": 0.09718294441699982, "learning_rate": 2.2738782297163147e-06, "loss": 0.0115, "step": 49880 }, { "epoch": 14.515565900494618, "grad_norm": 0.07706218212842941, "learning_rate": 2.2695039022040947e-06, "loss": 0.0114, "step": 49890 }, { "epoch": 14.51847541460576, "grad_norm": 0.08035634458065033, "learning_rate": 2.265133441880461e-06, "loss": 0.0116, "step": 49900 }, { "epoch": 14.521384928716904, "grad_norm": 0.10340150445699692, "learning_rate": 2.260766850073041e-06, "loss": 0.0135, "step": 49910 }, { "epoch": 14.524294442828047, "grad_norm": 0.09734215587377548, "learning_rate": 2.2564041281082943e-06, "loss": 0.0119, "step": 49920 }, { "epoch": 14.527203956939191, "grad_norm": 0.071480393409729, "learning_rate": 2.252045277311497e-06, "loss": 0.0125, "step": 49930 }, { "epoch": 14.530113471050335, "grad_norm": 0.09711794555187225, "learning_rate": 2.247690299006755e-06, "loss": 0.0118, "step": 49940 }, { "epoch": 14.533022985161479, "grad_norm": 0.08545799553394318, "learning_rate": 2.2433391945169936e-06, "loss": 0.0108, "step": 49950 }, { "epoch": 14.535932499272622, "grad_norm": 0.114593505859375, "learning_rate": 2.238991965163962e-06, "loss": 0.0116, "step": 49960 }, { "epoch": 14.538842013383764, "grad_norm": 0.09044355899095535, "learning_rate": 2.2346486122682434e-06, "loss": 0.0117, "step": 49970 }, { "epoch": 14.541751527494908, "grad_norm": 0.12274841964244843, "learning_rate": 2.23030913714922e-06, "loss": 0.0112, "step": 49980 }, { "epoch": 14.544661041606052, "grad_norm": 0.12383049726486206, "learning_rate": 2.2259735411251203e-06, "loss": 0.0113, "step": 49990 }, { "epoch": 14.547570555717195, "grad_norm": 0.12053946405649185, "learning_rate": 2.2216418255129806e-06, "loss": 0.0129, "step": 50000 }, { "epoch": 14.550480069828339, "grad_norm": 0.08784061670303345, "learning_rate": 2.21731399162866e-06, "loss": 0.0113, "step": 50010 }, { "epoch": 14.553389583939483, "grad_norm": 0.10515328496694565, "learning_rate": 2.212990040786842e-06, "loss": 0.0111, "step": 50020 }, { "epoch": 14.556299098050626, "grad_norm": 0.0901910737156868, "learning_rate": 2.2086699743010256e-06, "loss": 0.0122, "step": 50030 }, { "epoch": 14.559208612161768, "grad_norm": 0.11406216770410538, "learning_rate": 2.204353793483542e-06, "loss": 0.0125, "step": 50040 }, { "epoch": 14.562118126272912, "grad_norm": 0.1236414834856987, "learning_rate": 2.200041499645519e-06, "loss": 0.0121, "step": 50050 }, { "epoch": 14.565027640384056, "grad_norm": 0.09615408629179001, "learning_rate": 2.195733094096929e-06, "loss": 0.0126, "step": 50060 }, { "epoch": 14.5679371544952, "grad_norm": 0.06520824879407883, "learning_rate": 2.191428578146547e-06, "loss": 0.0115, "step": 50070 }, { "epoch": 14.570846668606343, "grad_norm": 0.0812545195221901, "learning_rate": 2.187127953101973e-06, "loss": 0.0112, "step": 50080 }, { "epoch": 14.573756182717487, "grad_norm": 0.09782446920871735, "learning_rate": 2.1828312202696227e-06, "loss": 0.0124, "step": 50090 }, { "epoch": 14.57666569682863, "grad_norm": 0.07727734744548798, "learning_rate": 2.1785383809547292e-06, "loss": 0.0115, "step": 50100 }, { "epoch": 14.579575210939772, "grad_norm": 0.08983752131462097, "learning_rate": 2.174249436461343e-06, "loss": 0.0132, "step": 50110 }, { "epoch": 14.582484725050916, "grad_norm": 0.11380410194396973, "learning_rate": 2.169964388092338e-06, "loss": 0.0133, "step": 50120 }, { "epoch": 14.58539423916206, "grad_norm": 0.07385256141424179, "learning_rate": 2.1656832371493886e-06, "loss": 0.0118, "step": 50130 }, { "epoch": 14.588303753273204, "grad_norm": 0.07353557646274567, "learning_rate": 2.1614059849330054e-06, "loss": 0.0119, "step": 50140 }, { "epoch": 14.591213267384347, "grad_norm": 0.135919451713562, "learning_rate": 2.1571326327424996e-06, "loss": 0.0118, "step": 50150 }, { "epoch": 14.594122781495491, "grad_norm": 0.0830376148223877, "learning_rate": 2.1528631818760044e-06, "loss": 0.0115, "step": 50160 }, { "epoch": 14.597032295606633, "grad_norm": 0.1234852597117424, "learning_rate": 2.148597633630465e-06, "loss": 0.0121, "step": 50170 }, { "epoch": 14.599941809717777, "grad_norm": 0.09709334373474121, "learning_rate": 2.1443359893016394e-06, "loss": 0.011, "step": 50180 }, { "epoch": 14.60285132382892, "grad_norm": 0.06883984804153442, "learning_rate": 2.1400782501841133e-06, "loss": 0.0107, "step": 50190 }, { "epoch": 14.605760837940064, "grad_norm": 0.08006058633327484, "learning_rate": 2.1358244175712614e-06, "loss": 0.0099, "step": 50200 }, { "epoch": 14.608670352051208, "grad_norm": 0.11529383808374405, "learning_rate": 2.131574492755296e-06, "loss": 0.0123, "step": 50210 }, { "epoch": 14.611579866162351, "grad_norm": 0.07566758990287781, "learning_rate": 2.127328477027229e-06, "loss": 0.0117, "step": 50220 }, { "epoch": 14.614489380273493, "grad_norm": 0.10555518418550491, "learning_rate": 2.1230863716768867e-06, "loss": 0.013, "step": 50230 }, { "epoch": 14.617398894384637, "grad_norm": 0.09319521486759186, "learning_rate": 2.1188481779929103e-06, "loss": 0.0131, "step": 50240 }, { "epoch": 14.62030840849578, "grad_norm": 0.08198336511850357, "learning_rate": 2.1146138972627483e-06, "loss": 0.013, "step": 50250 }, { "epoch": 14.623217922606925, "grad_norm": 0.09506461024284363, "learning_rate": 2.110383530772665e-06, "loss": 0.0116, "step": 50260 }, { "epoch": 14.626127436718068, "grad_norm": 0.06813335418701172, "learning_rate": 2.1061570798077388e-06, "loss": 0.0105, "step": 50270 }, { "epoch": 14.629036950829212, "grad_norm": 0.08690328150987625, "learning_rate": 2.1019345456518435e-06, "loss": 0.0124, "step": 50280 }, { "epoch": 14.631946464940356, "grad_norm": 0.08759833872318268, "learning_rate": 2.0977159295876824e-06, "loss": 0.0126, "step": 50290 }, { "epoch": 14.634855979051498, "grad_norm": 0.08829271048307419, "learning_rate": 2.093501232896759e-06, "loss": 0.0129, "step": 50300 }, { "epoch": 14.637765493162641, "grad_norm": 0.09927614778280258, "learning_rate": 2.0892904568593824e-06, "loss": 0.0123, "step": 50310 }, { "epoch": 14.640675007273785, "grad_norm": 0.1050194576382637, "learning_rate": 2.0850836027546777e-06, "loss": 0.0113, "step": 50320 }, { "epoch": 14.643584521384929, "grad_norm": 0.0890030786395073, "learning_rate": 2.0808806718605746e-06, "loss": 0.0112, "step": 50330 }, { "epoch": 14.646494035496072, "grad_norm": 0.1035156324505806, "learning_rate": 2.076681665453818e-06, "loss": 0.0119, "step": 50340 }, { "epoch": 14.649403549607216, "grad_norm": 0.08892787992954254, "learning_rate": 2.0724865848099444e-06, "loss": 0.0106, "step": 50350 }, { "epoch": 14.65231306371836, "grad_norm": 0.1049601137638092, "learning_rate": 2.0682954312033233e-06, "loss": 0.0121, "step": 50360 }, { "epoch": 14.655222577829502, "grad_norm": 0.07682237029075623, "learning_rate": 2.064108205907102e-06, "loss": 0.0117, "step": 50370 }, { "epoch": 14.658132091940645, "grad_norm": 0.08870186656713486, "learning_rate": 2.0599249101932572e-06, "loss": 0.0117, "step": 50380 }, { "epoch": 14.66104160605179, "grad_norm": 0.11504615098237991, "learning_rate": 2.0557455453325617e-06, "loss": 0.0114, "step": 50390 }, { "epoch": 14.663951120162933, "grad_norm": 0.09578891098499298, "learning_rate": 2.051570112594597e-06, "loss": 0.0134, "step": 50400 }, { "epoch": 14.666860634274077, "grad_norm": 0.08325634151697159, "learning_rate": 2.047398613247749e-06, "loss": 0.0123, "step": 50410 }, { "epoch": 14.66977014838522, "grad_norm": 0.09023507684469223, "learning_rate": 2.0432310485592083e-06, "loss": 0.0115, "step": 50420 }, { "epoch": 14.672679662496364, "grad_norm": 0.09544577449560165, "learning_rate": 2.0390674197949673e-06, "loss": 0.0102, "step": 50430 }, { "epoch": 14.675589176607506, "grad_norm": 0.08957954496145248, "learning_rate": 2.034907728219834e-06, "loss": 0.0116, "step": 50440 }, { "epoch": 14.67849869071865, "grad_norm": 0.09770897775888443, "learning_rate": 2.03075197509741e-06, "loss": 0.0128, "step": 50450 }, { "epoch": 14.681408204829793, "grad_norm": 0.10899663716554642, "learning_rate": 2.026600161690104e-06, "loss": 0.0124, "step": 50460 }, { "epoch": 14.684317718940937, "grad_norm": 0.10722421854734421, "learning_rate": 2.022452289259124e-06, "loss": 0.0123, "step": 50470 }, { "epoch": 14.68722723305208, "grad_norm": 0.08227404206991196, "learning_rate": 2.0183083590644835e-06, "loss": 0.0128, "step": 50480 }, { "epoch": 14.690136747163224, "grad_norm": 0.0836542621254921, "learning_rate": 2.0141683723650083e-06, "loss": 0.0125, "step": 50490 }, { "epoch": 14.693046261274366, "grad_norm": 0.10419255495071411, "learning_rate": 2.0100323304183042e-06, "loss": 0.0114, "step": 50500 }, { "epoch": 14.69595577538551, "grad_norm": 0.08329235762357712, "learning_rate": 2.0059002344808046e-06, "loss": 0.012, "step": 50510 }, { "epoch": 14.698865289496654, "grad_norm": 0.09505216777324677, "learning_rate": 2.0017720858077187e-06, "loss": 0.0115, "step": 50520 }, { "epoch": 14.701774803607798, "grad_norm": 0.07242848724126816, "learning_rate": 1.9976478856530765e-06, "loss": 0.0106, "step": 50530 }, { "epoch": 14.704684317718941, "grad_norm": 0.11323084682226181, "learning_rate": 1.9935276352696996e-06, "loss": 0.0116, "step": 50540 }, { "epoch": 14.707593831830085, "grad_norm": 0.08029104024171829, "learning_rate": 1.9894113359092116e-06, "loss": 0.0123, "step": 50550 }, { "epoch": 14.710503345941227, "grad_norm": 0.08388765156269073, "learning_rate": 1.9852989888220353e-06, "loss": 0.0122, "step": 50560 }, { "epoch": 14.71341286005237, "grad_norm": 0.09141846001148224, "learning_rate": 1.9811905952573935e-06, "loss": 0.0125, "step": 50570 }, { "epoch": 14.716322374163514, "grad_norm": 0.09130623191595078, "learning_rate": 1.9770861564633023e-06, "loss": 0.0109, "step": 50580 }, { "epoch": 14.719231888274658, "grad_norm": 0.1058465987443924, "learning_rate": 1.9729856736865905e-06, "loss": 0.0123, "step": 50590 }, { "epoch": 14.722141402385802, "grad_norm": 0.09863259643316269, "learning_rate": 1.968889148172874e-06, "loss": 0.0123, "step": 50600 }, { "epoch": 14.725050916496945, "grad_norm": 0.08707144856452942, "learning_rate": 1.9647965811665656e-06, "loss": 0.0126, "step": 50610 }, { "epoch": 14.72796043060809, "grad_norm": 0.06995953619480133, "learning_rate": 1.960707973910882e-06, "loss": 0.0124, "step": 50620 }, { "epoch": 14.730869944719231, "grad_norm": 0.10442402958869934, "learning_rate": 1.9566233276478303e-06, "loss": 0.0126, "step": 50630 }, { "epoch": 14.733779458830375, "grad_norm": 0.08995931595563889, "learning_rate": 1.952542643618228e-06, "loss": 0.0114, "step": 50640 }, { "epoch": 14.736688972941518, "grad_norm": 0.08130946755409241, "learning_rate": 1.948465923061666e-06, "loss": 0.0114, "step": 50650 }, { "epoch": 14.739598487052662, "grad_norm": 0.08532612025737762, "learning_rate": 1.944393167216556e-06, "loss": 0.0127, "step": 50660 }, { "epoch": 14.742508001163806, "grad_norm": 0.08036969602108002, "learning_rate": 1.940324377320084e-06, "loss": 0.0116, "step": 50670 }, { "epoch": 14.74541751527495, "grad_norm": 0.11829972267150879, "learning_rate": 1.9362595546082463e-06, "loss": 0.0126, "step": 50680 }, { "epoch": 14.748327029386093, "grad_norm": 0.09298297762870789, "learning_rate": 1.9321987003158296e-06, "loss": 0.0103, "step": 50690 }, { "epoch": 14.751236543497235, "grad_norm": 0.0688321441411972, "learning_rate": 1.9281418156764113e-06, "loss": 0.0113, "step": 50700 }, { "epoch": 14.754146057608379, "grad_norm": 0.08477789163589478, "learning_rate": 1.9240889019223683e-06, "loss": 0.0112, "step": 50710 }, { "epoch": 14.757055571719523, "grad_norm": 0.08115758001804352, "learning_rate": 1.920039960284866e-06, "loss": 0.0109, "step": 50720 }, { "epoch": 14.759965085830666, "grad_norm": 0.08303546160459518, "learning_rate": 1.915994991993869e-06, "loss": 0.0113, "step": 50730 }, { "epoch": 14.76287459994181, "grad_norm": 0.09734106063842773, "learning_rate": 1.911953998278126e-06, "loss": 0.0126, "step": 50740 }, { "epoch": 14.765784114052954, "grad_norm": 0.08321495354175568, "learning_rate": 1.907916980365193e-06, "loss": 0.0118, "step": 50750 }, { "epoch": 14.768693628164097, "grad_norm": 0.10458218306303024, "learning_rate": 1.9038839394814029e-06, "loss": 0.0116, "step": 50760 }, { "epoch": 14.77160314227524, "grad_norm": 0.11044549942016602, "learning_rate": 1.8998548768518903e-06, "loss": 0.0116, "step": 50770 }, { "epoch": 14.774512656386383, "grad_norm": 0.08378085494041443, "learning_rate": 1.8958297937005764e-06, "loss": 0.0119, "step": 50780 }, { "epoch": 14.777422170497527, "grad_norm": 0.08562400937080383, "learning_rate": 1.8918086912501764e-06, "loss": 0.0114, "step": 50790 }, { "epoch": 14.78033168460867, "grad_norm": 0.07612711191177368, "learning_rate": 1.88779157072219e-06, "loss": 0.0127, "step": 50800 }, { "epoch": 14.783241198719814, "grad_norm": 0.12608659267425537, "learning_rate": 1.883778433336924e-06, "loss": 0.0118, "step": 50810 }, { "epoch": 14.786150712830958, "grad_norm": 0.09003332257270813, "learning_rate": 1.8797692803134487e-06, "loss": 0.0118, "step": 50820 }, { "epoch": 14.7890602269421, "grad_norm": 0.08862984925508499, "learning_rate": 1.875764112869653e-06, "loss": 0.0108, "step": 50830 }, { "epoch": 14.791969741053244, "grad_norm": 0.10809183120727539, "learning_rate": 1.8717629322221886e-06, "loss": 0.0139, "step": 50840 }, { "epoch": 14.794879255164387, "grad_norm": 0.08269622921943665, "learning_rate": 1.8677657395865183e-06, "loss": 0.0106, "step": 50850 }, { "epoch": 14.797788769275531, "grad_norm": 0.09544204920530319, "learning_rate": 1.8637725361768776e-06, "loss": 0.0132, "step": 50860 }, { "epoch": 14.800698283386675, "grad_norm": 0.08002714067697525, "learning_rate": 1.8597833232062995e-06, "loss": 0.0131, "step": 50870 }, { "epoch": 14.803607797497818, "grad_norm": 0.11024200916290283, "learning_rate": 1.8557981018866011e-06, "loss": 0.012, "step": 50880 }, { "epoch": 14.806517311608962, "grad_norm": 0.06721765547990799, "learning_rate": 1.851816873428384e-06, "loss": 0.0123, "step": 50890 }, { "epoch": 14.809426825720104, "grad_norm": 0.07877326011657715, "learning_rate": 1.8478396390410445e-06, "loss": 0.0122, "step": 50900 }, { "epoch": 14.812336339831248, "grad_norm": 0.09106653928756714, "learning_rate": 1.843866399932761e-06, "loss": 0.0127, "step": 50910 }, { "epoch": 14.815245853942391, "grad_norm": 0.0854325219988823, "learning_rate": 1.8398971573104978e-06, "loss": 0.0116, "step": 50920 }, { "epoch": 14.818155368053535, "grad_norm": 0.06935513764619827, "learning_rate": 1.8359319123800057e-06, "loss": 0.0123, "step": 50930 }, { "epoch": 14.821064882164679, "grad_norm": 0.09261160343885422, "learning_rate": 1.8319706663458208e-06, "loss": 0.0113, "step": 50940 }, { "epoch": 14.823974396275823, "grad_norm": 0.10323281586170197, "learning_rate": 1.8280134204112642e-06, "loss": 0.0132, "step": 50950 }, { "epoch": 14.826883910386965, "grad_norm": 0.08978305757045746, "learning_rate": 1.8240601757784492e-06, "loss": 0.0114, "step": 50960 }, { "epoch": 14.829793424498108, "grad_norm": 0.08454813063144684, "learning_rate": 1.8201109336482569e-06, "loss": 0.0113, "step": 50970 }, { "epoch": 14.832702938609252, "grad_norm": 0.13015563786029816, "learning_rate": 1.8161656952203726e-06, "loss": 0.0115, "step": 50980 }, { "epoch": 14.835612452720396, "grad_norm": 0.1167113184928894, "learning_rate": 1.812224461693246e-06, "loss": 0.0116, "step": 50990 }, { "epoch": 14.83852196683154, "grad_norm": 0.0787644162774086, "learning_rate": 1.8082872342641294e-06, "loss": 0.011, "step": 51000 }, { "epoch": 14.841431480942683, "grad_norm": 0.11761697381734848, "learning_rate": 1.8043540141290426e-06, "loss": 0.0123, "step": 51010 }, { "epoch": 14.844340995053827, "grad_norm": 0.10467718541622162, "learning_rate": 1.8004248024827951e-06, "loss": 0.0125, "step": 51020 }, { "epoch": 14.847250509164969, "grad_norm": 0.09248286485671997, "learning_rate": 1.7964996005189787e-06, "loss": 0.013, "step": 51030 }, { "epoch": 14.850160023276112, "grad_norm": 0.11098071187734604, "learning_rate": 1.792578409429963e-06, "loss": 0.0111, "step": 51040 }, { "epoch": 14.853069537387256, "grad_norm": 0.11222308874130249, "learning_rate": 1.7886612304069067e-06, "loss": 0.0113, "step": 51050 }, { "epoch": 14.8559790514984, "grad_norm": 0.10439327359199524, "learning_rate": 1.784748064639743e-06, "loss": 0.013, "step": 51060 }, { "epoch": 14.858888565609544, "grad_norm": 0.12254797667264938, "learning_rate": 1.7808389133171893e-06, "loss": 0.013, "step": 51070 }, { "epoch": 14.861798079720687, "grad_norm": 0.10894834250211716, "learning_rate": 1.7769337776267424e-06, "loss": 0.012, "step": 51080 }, { "epoch": 14.864707593831831, "grad_norm": 0.1279921680688858, "learning_rate": 1.7730326587546785e-06, "loss": 0.0114, "step": 51090 }, { "epoch": 14.867617107942973, "grad_norm": 0.10037247836589813, "learning_rate": 1.769135557886053e-06, "loss": 0.0119, "step": 51100 }, { "epoch": 14.870526622054117, "grad_norm": 0.08151151984930038, "learning_rate": 1.7652424762047103e-06, "loss": 0.0103, "step": 51110 }, { "epoch": 14.87343613616526, "grad_norm": 0.1605531871318817, "learning_rate": 1.7613534148932532e-06, "loss": 0.0113, "step": 51120 }, { "epoch": 14.876345650276404, "grad_norm": 0.10777812451124191, "learning_rate": 1.7574683751330878e-06, "loss": 0.0105, "step": 51130 }, { "epoch": 14.879255164387548, "grad_norm": 0.10002952814102173, "learning_rate": 1.7535873581043753e-06, "loss": 0.0105, "step": 51140 }, { "epoch": 14.882164678498691, "grad_norm": 0.10961098223924637, "learning_rate": 1.7497103649860762e-06, "loss": 0.0113, "step": 51150 }, { "epoch": 14.885074192609833, "grad_norm": 0.06095956638455391, "learning_rate": 1.745837396955915e-06, "loss": 0.0113, "step": 51160 }, { "epoch": 14.887983706720977, "grad_norm": 0.07068515568971634, "learning_rate": 1.7419684551903919e-06, "loss": 0.0119, "step": 51170 }, { "epoch": 14.89089322083212, "grad_norm": 0.09847506880760193, "learning_rate": 1.7381035408648006e-06, "loss": 0.0121, "step": 51180 }, { "epoch": 14.893802734943264, "grad_norm": 0.08099154382944107, "learning_rate": 1.7342426551531881e-06, "loss": 0.0113, "step": 51190 }, { "epoch": 14.896712249054408, "grad_norm": 0.11063745617866516, "learning_rate": 1.7303857992284016e-06, "loss": 0.0116, "step": 51200 }, { "epoch": 14.899621763165552, "grad_norm": 0.10280027985572815, "learning_rate": 1.7265329742620373e-06, "loss": 0.0112, "step": 51210 }, { "epoch": 14.902531277276696, "grad_norm": 0.06824269890785217, "learning_rate": 1.7226841814244937e-06, "loss": 0.0119, "step": 51220 }, { "epoch": 14.905440791387838, "grad_norm": 0.09263728559017181, "learning_rate": 1.7188394218849291e-06, "loss": 0.0122, "step": 51230 }, { "epoch": 14.908350305498981, "grad_norm": 0.06803730875253677, "learning_rate": 1.714998696811279e-06, "loss": 0.0121, "step": 51240 }, { "epoch": 14.911259819610125, "grad_norm": 0.0757465809583664, "learning_rate": 1.7111620073702554e-06, "loss": 0.0124, "step": 51250 }, { "epoch": 14.914169333721269, "grad_norm": 0.11182267963886261, "learning_rate": 1.7073293547273427e-06, "loss": 0.015, "step": 51260 }, { "epoch": 14.917078847832412, "grad_norm": 0.07855556905269623, "learning_rate": 1.7035007400467946e-06, "loss": 0.0125, "step": 51270 }, { "epoch": 14.919988361943556, "grad_norm": 0.09345870465040207, "learning_rate": 1.699676164491652e-06, "loss": 0.0122, "step": 51280 }, { "epoch": 14.922897876054698, "grad_norm": 0.08356611430644989, "learning_rate": 1.695855629223716e-06, "loss": 0.0122, "step": 51290 }, { "epoch": 14.925807390165842, "grad_norm": 0.08464175462722778, "learning_rate": 1.6920391354035647e-06, "loss": 0.0119, "step": 51300 }, { "epoch": 14.928716904276985, "grad_norm": 0.08778109401464462, "learning_rate": 1.6882266841905475e-06, "loss": 0.0125, "step": 51310 }, { "epoch": 14.931626418388129, "grad_norm": 0.07236059755086899, "learning_rate": 1.684418276742783e-06, "loss": 0.0133, "step": 51320 }, { "epoch": 14.934535932499273, "grad_norm": 0.09071695059537888, "learning_rate": 1.6806139142171756e-06, "loss": 0.013, "step": 51330 }, { "epoch": 14.937445446610417, "grad_norm": 0.07344535738229752, "learning_rate": 1.676813597769376e-06, "loss": 0.0122, "step": 51340 }, { "epoch": 14.94035496072156, "grad_norm": 0.07720153033733368, "learning_rate": 1.6730173285538326e-06, "loss": 0.0108, "step": 51350 }, { "epoch": 14.943264474832702, "grad_norm": 0.08965466171503067, "learning_rate": 1.669225107723741e-06, "loss": 0.0115, "step": 51360 }, { "epoch": 14.946173988943846, "grad_norm": 0.10031195729970932, "learning_rate": 1.6654369364310861e-06, "loss": 0.0105, "step": 51370 }, { "epoch": 14.94908350305499, "grad_norm": 0.14743420481681824, "learning_rate": 1.6616528158266114e-06, "loss": 0.0129, "step": 51380 }, { "epoch": 14.951993017166133, "grad_norm": 0.1043456494808197, "learning_rate": 1.657872747059832e-06, "loss": 0.013, "step": 51390 }, { "epoch": 14.954902531277277, "grad_norm": 0.12819896638393402, "learning_rate": 1.6540967312790328e-06, "loss": 0.013, "step": 51400 }, { "epoch": 14.95781204538842, "grad_norm": 0.0824633240699768, "learning_rate": 1.6503247696312696e-06, "loss": 0.0125, "step": 51410 }, { "epoch": 14.960721559499564, "grad_norm": 0.1471211314201355, "learning_rate": 1.6465568632623606e-06, "loss": 0.0109, "step": 51420 }, { "epoch": 14.963631073610706, "grad_norm": 0.07496079057455063, "learning_rate": 1.6427930133169023e-06, "loss": 0.0108, "step": 51430 }, { "epoch": 14.96654058772185, "grad_norm": 0.08624028414487839, "learning_rate": 1.6390332209382496e-06, "loss": 0.0116, "step": 51440 }, { "epoch": 14.969450101832994, "grad_norm": 0.0769573375582695, "learning_rate": 1.6352774872685279e-06, "loss": 0.0118, "step": 51450 }, { "epoch": 14.972359615944137, "grad_norm": 0.1577690839767456, "learning_rate": 1.6315258134486322e-06, "loss": 0.0108, "step": 51460 }, { "epoch": 14.975269130055281, "grad_norm": 0.14855101704597473, "learning_rate": 1.6277782006182163e-06, "loss": 0.0137, "step": 51470 }, { "epoch": 14.978178644166425, "grad_norm": 0.08167760819196701, "learning_rate": 1.6240346499157161e-06, "loss": 0.0114, "step": 51480 }, { "epoch": 14.981088158277569, "grad_norm": 0.07805924862623215, "learning_rate": 1.6202951624783102e-06, "loss": 0.0116, "step": 51490 }, { "epoch": 14.98399767238871, "grad_norm": 0.09089308977127075, "learning_rate": 1.61655973944197e-06, "loss": 0.0116, "step": 51500 }, { "epoch": 14.986907186499854, "grad_norm": 0.08414576202630997, "learning_rate": 1.6128283819414064e-06, "loss": 0.0108, "step": 51510 }, { "epoch": 14.989816700610998, "grad_norm": 0.08036577701568604, "learning_rate": 1.6091010911101156e-06, "loss": 0.0122, "step": 51520 }, { "epoch": 14.992726214722142, "grad_norm": 0.09639016538858414, "learning_rate": 1.605377868080347e-06, "loss": 0.0129, "step": 51530 }, { "epoch": 14.995635728833285, "grad_norm": 0.13987919688224792, "learning_rate": 1.6016587139831163e-06, "loss": 0.0142, "step": 51540 }, { "epoch": 14.998545242944429, "grad_norm": 0.08331600576639175, "learning_rate": 1.5979436299482076e-06, "loss": 0.0125, "step": 51550 }, { "epoch": 15.001454757055571, "grad_norm": 0.1080746129155159, "learning_rate": 1.5942326171041615e-06, "loss": 0.0122, "step": 51560 }, { "epoch": 15.004364271166715, "grad_norm": 0.09714958071708679, "learning_rate": 1.5905256765782883e-06, "loss": 0.0114, "step": 51570 }, { "epoch": 15.007273785277858, "grad_norm": 0.0938587486743927, "learning_rate": 1.5868228094966553e-06, "loss": 0.0109, "step": 51580 }, { "epoch": 15.010183299389002, "grad_norm": 0.08474763482809067, "learning_rate": 1.583124016984101e-06, "loss": 0.0123, "step": 51590 }, { "epoch": 15.013092813500146, "grad_norm": 0.0654349997639656, "learning_rate": 1.5794293001642186e-06, "loss": 0.0107, "step": 51600 }, { "epoch": 15.01600232761129, "grad_norm": 0.08787761628627777, "learning_rate": 1.5757386601593649e-06, "loss": 0.0131, "step": 51610 }, { "epoch": 15.018911841722433, "grad_norm": 0.08458449691534042, "learning_rate": 1.5720520980906584e-06, "loss": 0.0115, "step": 51620 }, { "epoch": 15.021821355833575, "grad_norm": 0.12371976673603058, "learning_rate": 1.5683696150779813e-06, "loss": 0.0111, "step": 51630 }, { "epoch": 15.024730869944719, "grad_norm": 0.07489611208438873, "learning_rate": 1.5646912122399692e-06, "loss": 0.0124, "step": 51640 }, { "epoch": 15.027640384055863, "grad_norm": 0.07979737222194672, "learning_rate": 1.5610168906940347e-06, "loss": 0.012, "step": 51650 }, { "epoch": 15.030549898167006, "grad_norm": 0.08209624886512756, "learning_rate": 1.5573466515563266e-06, "loss": 0.0112, "step": 51660 }, { "epoch": 15.03345941227815, "grad_norm": 0.132992222905159, "learning_rate": 1.55368049594178e-06, "loss": 0.011, "step": 51670 }, { "epoch": 15.036368926389294, "grad_norm": 0.09585091471672058, "learning_rate": 1.5500184249640619e-06, "loss": 0.0136, "step": 51680 }, { "epoch": 15.039278440500436, "grad_norm": 0.08746480196714401, "learning_rate": 1.5463604397356245e-06, "loss": 0.0114, "step": 51690 }, { "epoch": 15.04218795461158, "grad_norm": 0.08822964876890182, "learning_rate": 1.5427065413676622e-06, "loss": 0.0124, "step": 51700 }, { "epoch": 15.045097468722723, "grad_norm": 0.08234257996082306, "learning_rate": 1.5390567309701337e-06, "loss": 0.0115, "step": 51710 }, { "epoch": 15.048006982833867, "grad_norm": 0.08850196748971939, "learning_rate": 1.5354110096517537e-06, "loss": 0.0119, "step": 51720 }, { "epoch": 15.05091649694501, "grad_norm": 0.09200754016637802, "learning_rate": 1.5317693785199948e-06, "loss": 0.0114, "step": 51730 }, { "epoch": 15.053826011056154, "grad_norm": 0.11381151527166367, "learning_rate": 1.5281318386810938e-06, "loss": 0.0129, "step": 51740 }, { "epoch": 15.056735525167298, "grad_norm": 0.10452873259782791, "learning_rate": 1.5244983912400368e-06, "loss": 0.0123, "step": 51750 }, { "epoch": 15.05964503927844, "grad_norm": 0.09594248235225677, "learning_rate": 1.5208690373005675e-06, "loss": 0.0119, "step": 51760 }, { "epoch": 15.062554553389583, "grad_norm": 0.0601506307721138, "learning_rate": 1.5172437779651887e-06, "loss": 0.0102, "step": 51770 }, { "epoch": 15.065464067500727, "grad_norm": 0.09715622663497925, "learning_rate": 1.5136226143351583e-06, "loss": 0.0122, "step": 51780 }, { "epoch": 15.068373581611871, "grad_norm": 0.10888111591339111, "learning_rate": 1.5100055475104897e-06, "loss": 0.0121, "step": 51790 }, { "epoch": 15.071283095723015, "grad_norm": 0.10944156348705292, "learning_rate": 1.5063925785899568e-06, "loss": 0.0119, "step": 51800 }, { "epoch": 15.074192609834158, "grad_norm": 0.09870219230651855, "learning_rate": 1.5027837086710754e-06, "loss": 0.0108, "step": 51810 }, { "epoch": 15.077102123945302, "grad_norm": 0.10548843443393707, "learning_rate": 1.4991789388501354e-06, "loss": 0.0121, "step": 51820 }, { "epoch": 15.080011638056444, "grad_norm": 0.08653757721185684, "learning_rate": 1.4955782702221605e-06, "loss": 0.0125, "step": 51830 }, { "epoch": 15.082921152167588, "grad_norm": 0.09157120436429977, "learning_rate": 1.4919817038809442e-06, "loss": 0.0122, "step": 51840 }, { "epoch": 15.085830666278731, "grad_norm": 0.09803498536348343, "learning_rate": 1.4883892409190292e-06, "loss": 0.0138, "step": 51850 }, { "epoch": 15.088740180389875, "grad_norm": 0.08234868943691254, "learning_rate": 1.4848008824277093e-06, "loss": 0.0107, "step": 51860 }, { "epoch": 15.091649694501019, "grad_norm": 0.1219555139541626, "learning_rate": 1.4812166294970342e-06, "loss": 0.0125, "step": 51870 }, { "epoch": 15.094559208612162, "grad_norm": 0.07166316360235214, "learning_rate": 1.4776364832157996e-06, "loss": 0.0123, "step": 51880 }, { "epoch": 15.097468722723304, "grad_norm": 0.09275051206350327, "learning_rate": 1.4740604446715672e-06, "loss": 0.0109, "step": 51890 }, { "epoch": 15.100378236834448, "grad_norm": 0.1224413737654686, "learning_rate": 1.4704885149506414e-06, "loss": 0.011, "step": 51900 }, { "epoch": 15.103287750945592, "grad_norm": 0.10494129359722137, "learning_rate": 1.4669206951380777e-06, "loss": 0.0137, "step": 51910 }, { "epoch": 15.106197265056736, "grad_norm": 0.09569170325994492, "learning_rate": 1.4633569863176872e-06, "loss": 0.0115, "step": 51920 }, { "epoch": 15.10910677916788, "grad_norm": 0.09860784560441971, "learning_rate": 1.459797389572029e-06, "loss": 0.0118, "step": 51930 }, { "epoch": 15.112016293279023, "grad_norm": 0.10797607153654099, "learning_rate": 1.456241905982415e-06, "loss": 0.013, "step": 51940 }, { "epoch": 15.114925807390167, "grad_norm": 0.09140504151582718, "learning_rate": 1.4526905366289095e-06, "loss": 0.011, "step": 51950 }, { "epoch": 15.117835321501309, "grad_norm": 0.07743439823389053, "learning_rate": 1.4491432825903177e-06, "loss": 0.0108, "step": 51960 }, { "epoch": 15.120744835612452, "grad_norm": 0.13808098435401917, "learning_rate": 1.4456001449442146e-06, "loss": 0.0108, "step": 51970 }, { "epoch": 15.123654349723596, "grad_norm": 0.10954734683036804, "learning_rate": 1.4420611247668991e-06, "loss": 0.0123, "step": 51980 }, { "epoch": 15.12656386383474, "grad_norm": 0.09936072677373886, "learning_rate": 1.4385262231334395e-06, "loss": 0.0126, "step": 51990 }, { "epoch": 15.129473377945883, "grad_norm": 0.10557696968317032, "learning_rate": 1.434995441117642e-06, "loss": 0.0122, "step": 52000 }, { "epoch": 15.132382892057027, "grad_norm": 0.1231435090303421, "learning_rate": 1.4314687797920685e-06, "loss": 0.0105, "step": 52010 }, { "epoch": 15.135292406168169, "grad_norm": 0.0967821553349495, "learning_rate": 1.4279462402280218e-06, "loss": 0.0125, "step": 52020 }, { "epoch": 15.138201920279313, "grad_norm": 0.09026132524013519, "learning_rate": 1.4244278234955548e-06, "loss": 0.0121, "step": 52030 }, { "epoch": 15.141111434390456, "grad_norm": 0.07505800575017929, "learning_rate": 1.420913530663479e-06, "loss": 0.0108, "step": 52040 }, { "epoch": 15.1440209485016, "grad_norm": 0.09073499590158463, "learning_rate": 1.4174033627993322e-06, "loss": 0.0104, "step": 52050 }, { "epoch": 15.146930462612744, "grad_norm": 0.09797381609678268, "learning_rate": 1.4138973209694194e-06, "loss": 0.0112, "step": 52060 }, { "epoch": 15.149839976723888, "grad_norm": 0.06999389827251434, "learning_rate": 1.4103954062387792e-06, "loss": 0.0112, "step": 52070 }, { "epoch": 15.152749490835031, "grad_norm": 0.09867212176322937, "learning_rate": 1.4068976196712018e-06, "loss": 0.0117, "step": 52080 }, { "epoch": 15.155659004946173, "grad_norm": 0.07982252538204193, "learning_rate": 1.4034039623292245e-06, "loss": 0.0105, "step": 52090 }, { "epoch": 15.158568519057317, "grad_norm": 0.08566166460514069, "learning_rate": 1.3999144352741245e-06, "loss": 0.013, "step": 52100 }, { "epoch": 15.16147803316846, "grad_norm": 0.1170847937464714, "learning_rate": 1.3964290395659257e-06, "loss": 0.0114, "step": 52110 }, { "epoch": 15.164387547279604, "grad_norm": 0.10393167287111282, "learning_rate": 1.3929477762634113e-06, "loss": 0.0127, "step": 52120 }, { "epoch": 15.167297061390748, "grad_norm": 0.09684749692678452, "learning_rate": 1.389470646424082e-06, "loss": 0.0114, "step": 52130 }, { "epoch": 15.170206575501892, "grad_norm": 0.08791446685791016, "learning_rate": 1.3859976511042077e-06, "loss": 0.012, "step": 52140 }, { "epoch": 15.173116089613035, "grad_norm": 0.08819746971130371, "learning_rate": 1.3825287913587913e-06, "loss": 0.0114, "step": 52150 }, { "epoch": 15.176025603724177, "grad_norm": 0.09665551781654358, "learning_rate": 1.3790640682415784e-06, "loss": 0.0122, "step": 52160 }, { "epoch": 15.178935117835321, "grad_norm": 0.08780232071876526, "learning_rate": 1.3756034828050618e-06, "loss": 0.0118, "step": 52170 }, { "epoch": 15.181844631946465, "grad_norm": 0.09365908801555634, "learning_rate": 1.3721470361004724e-06, "loss": 0.0127, "step": 52180 }, { "epoch": 15.184754146057609, "grad_norm": 0.08497260510921478, "learning_rate": 1.3686947291777985e-06, "loss": 0.0114, "step": 52190 }, { "epoch": 15.187663660168752, "grad_norm": 0.09347938746213913, "learning_rate": 1.3652465630857446e-06, "loss": 0.0117, "step": 52200 }, { "epoch": 15.190573174279896, "grad_norm": 0.09563814103603363, "learning_rate": 1.3618025388717832e-06, "loss": 0.0124, "step": 52210 }, { "epoch": 15.193482688391038, "grad_norm": 0.07511107623577118, "learning_rate": 1.358362657582115e-06, "loss": 0.0116, "step": 52220 }, { "epoch": 15.196392202502182, "grad_norm": 0.13597145676612854, "learning_rate": 1.3549269202616849e-06, "loss": 0.0126, "step": 52230 }, { "epoch": 15.199301716613325, "grad_norm": 0.07785491645336151, "learning_rate": 1.3514953279541792e-06, "loss": 0.0118, "step": 52240 }, { "epoch": 15.202211230724469, "grad_norm": 0.09956119954586029, "learning_rate": 1.3480678817020242e-06, "loss": 0.0107, "step": 52250 }, { "epoch": 15.205120744835613, "grad_norm": 0.09504759311676025, "learning_rate": 1.344644582546387e-06, "loss": 0.0116, "step": 52260 }, { "epoch": 15.208030258946756, "grad_norm": 0.10946869850158691, "learning_rate": 1.341225431527181e-06, "loss": 0.0115, "step": 52270 }, { "epoch": 15.2109397730579, "grad_norm": 0.07868298143148422, "learning_rate": 1.3378104296830452e-06, "loss": 0.0107, "step": 52280 }, { "epoch": 15.213849287169042, "grad_norm": 0.10791709274053574, "learning_rate": 1.3343995780513745e-06, "loss": 0.0128, "step": 52290 }, { "epoch": 15.216758801280186, "grad_norm": 0.08933006227016449, "learning_rate": 1.3309928776682922e-06, "loss": 0.0112, "step": 52300 }, { "epoch": 15.21966831539133, "grad_norm": 0.07027390599250793, "learning_rate": 1.3275903295686636e-06, "loss": 0.0111, "step": 52310 }, { "epoch": 15.222577829502473, "grad_norm": 0.06704305112361908, "learning_rate": 1.3241919347860954e-06, "loss": 0.0092, "step": 52320 }, { "epoch": 15.225487343613617, "grad_norm": 0.092002272605896, "learning_rate": 1.3207976943529248e-06, "loss": 0.0114, "step": 52330 }, { "epoch": 15.22839685772476, "grad_norm": 0.09675654023885727, "learning_rate": 1.3174076093002423e-06, "loss": 0.0124, "step": 52340 }, { "epoch": 15.231306371835903, "grad_norm": 0.09977634996175766, "learning_rate": 1.3140216806578559e-06, "loss": 0.0125, "step": 52350 }, { "epoch": 15.234215885947046, "grad_norm": 0.08775138854980469, "learning_rate": 1.3106399094543292e-06, "loss": 0.012, "step": 52360 }, { "epoch": 15.23712540005819, "grad_norm": 0.09038499742746353, "learning_rate": 1.3072622967169456e-06, "loss": 0.0137, "step": 52370 }, { "epoch": 15.240034914169334, "grad_norm": 0.10809804499149323, "learning_rate": 1.3038888434717444e-06, "loss": 0.0133, "step": 52380 }, { "epoch": 15.242944428280477, "grad_norm": 0.11039973050355911, "learning_rate": 1.3005195507434875e-06, "loss": 0.012, "step": 52390 }, { "epoch": 15.245853942391621, "grad_norm": 0.08462492376565933, "learning_rate": 1.2971544195556777e-06, "loss": 0.0104, "step": 52400 }, { "epoch": 15.248763456502765, "grad_norm": 0.09547124803066254, "learning_rate": 1.293793450930551e-06, "loss": 0.0121, "step": 52410 }, { "epoch": 15.251672970613907, "grad_norm": 0.0683620348572731, "learning_rate": 1.2904366458890837e-06, "loss": 0.0116, "step": 52420 }, { "epoch": 15.25458248472505, "grad_norm": 0.09623546153306961, "learning_rate": 1.2870840054509803e-06, "loss": 0.0118, "step": 52430 }, { "epoch": 15.257491998836194, "grad_norm": 0.10722112655639648, "learning_rate": 1.283735530634691e-06, "loss": 0.0124, "step": 52440 }, { "epoch": 15.260401512947338, "grad_norm": 0.07861322909593582, "learning_rate": 1.2803912224573889e-06, "loss": 0.0125, "step": 52450 }, { "epoch": 15.263311027058482, "grad_norm": 0.07318713515996933, "learning_rate": 1.2770510819349896e-06, "loss": 0.0112, "step": 52460 }, { "epoch": 15.266220541169625, "grad_norm": 0.16173696517944336, "learning_rate": 1.2737151100821382e-06, "loss": 0.0115, "step": 52470 }, { "epoch": 15.269130055280769, "grad_norm": 0.0655955970287323, "learning_rate": 1.2703833079122107e-06, "loss": 0.012, "step": 52480 }, { "epoch": 15.27203956939191, "grad_norm": 0.10796403139829636, "learning_rate": 1.2670556764373308e-06, "loss": 0.0121, "step": 52490 }, { "epoch": 15.274949083503055, "grad_norm": 0.11076771467924118, "learning_rate": 1.2637322166683334e-06, "loss": 0.0124, "step": 52500 }, { "epoch": 15.277858597614198, "grad_norm": 0.05835793539881706, "learning_rate": 1.2604129296148092e-06, "loss": 0.0122, "step": 52510 }, { "epoch": 15.280768111725342, "grad_norm": 0.09086216241121292, "learning_rate": 1.2570978162850583e-06, "loss": 0.0113, "step": 52520 }, { "epoch": 15.283677625836486, "grad_norm": 0.08954429626464844, "learning_rate": 1.2537868776861332e-06, "loss": 0.0112, "step": 52530 }, { "epoch": 15.28658713994763, "grad_norm": 0.08914884924888611, "learning_rate": 1.2504801148238082e-06, "loss": 0.0108, "step": 52540 }, { "epoch": 15.289496654058771, "grad_norm": 0.09523703157901764, "learning_rate": 1.2471775287025877e-06, "loss": 0.0114, "step": 52550 }, { "epoch": 15.292406168169915, "grad_norm": 0.07657065987586975, "learning_rate": 1.2438791203257127e-06, "loss": 0.0122, "step": 52560 }, { "epoch": 15.295315682281059, "grad_norm": 0.07838627696037292, "learning_rate": 1.2405848906951517e-06, "loss": 0.0123, "step": 52570 }, { "epoch": 15.298225196392202, "grad_norm": 0.11201121658086777, "learning_rate": 1.2372948408116025e-06, "loss": 0.0119, "step": 52580 }, { "epoch": 15.301134710503346, "grad_norm": 0.08450622856616974, "learning_rate": 1.234008971674499e-06, "loss": 0.0105, "step": 52590 }, { "epoch": 15.30404422461449, "grad_norm": 0.08476007729768753, "learning_rate": 1.2307272842819995e-06, "loss": 0.0119, "step": 52600 }, { "epoch": 15.306953738725634, "grad_norm": 0.1016710177063942, "learning_rate": 1.2274497796309948e-06, "loss": 0.0137, "step": 52610 }, { "epoch": 15.309863252836776, "grad_norm": 0.0709109976887703, "learning_rate": 1.224176458717105e-06, "loss": 0.0113, "step": 52620 }, { "epoch": 15.31277276694792, "grad_norm": 0.08575864136219025, "learning_rate": 1.2209073225346734e-06, "loss": 0.0117, "step": 52630 }, { "epoch": 15.315682281059063, "grad_norm": 0.08431565016508102, "learning_rate": 1.217642372076787e-06, "loss": 0.0111, "step": 52640 }, { "epoch": 15.318591795170207, "grad_norm": 0.092427559196949, "learning_rate": 1.214381608335241e-06, "loss": 0.0121, "step": 52650 }, { "epoch": 15.32150130928135, "grad_norm": 0.1281408965587616, "learning_rate": 1.2111250323005795e-06, "loss": 0.0108, "step": 52660 }, { "epoch": 15.324410823392494, "grad_norm": 0.10501905530691147, "learning_rate": 1.2078726449620547e-06, "loss": 0.0127, "step": 52670 }, { "epoch": 15.327320337503636, "grad_norm": 0.0962478443980217, "learning_rate": 1.2046244473076611e-06, "loss": 0.0123, "step": 52680 }, { "epoch": 15.33022985161478, "grad_norm": 0.10184800624847412, "learning_rate": 1.2013804403241163e-06, "loss": 0.0109, "step": 52690 }, { "epoch": 15.333139365725923, "grad_norm": 0.10672658681869507, "learning_rate": 1.1981406249968618e-06, "loss": 0.0111, "step": 52700 }, { "epoch": 15.336048879837067, "grad_norm": 0.07536043971776962, "learning_rate": 1.1949050023100671e-06, "loss": 0.0134, "step": 52710 }, { "epoch": 15.33895839394821, "grad_norm": 0.10763435065746307, "learning_rate": 1.1916735732466328e-06, "loss": 0.012, "step": 52720 }, { "epoch": 15.341867908059355, "grad_norm": 0.07890904694795609, "learning_rate": 1.188446338788175e-06, "loss": 0.0114, "step": 52730 }, { "epoch": 15.344777422170498, "grad_norm": 0.08584202826023102, "learning_rate": 1.185223299915048e-06, "loss": 0.0117, "step": 52740 }, { "epoch": 15.34768693628164, "grad_norm": 0.08475328236818314, "learning_rate": 1.1820044576063255e-06, "loss": 0.0105, "step": 52750 }, { "epoch": 15.350596450392784, "grad_norm": 0.08449718356132507, "learning_rate": 1.1787898128398066e-06, "loss": 0.0128, "step": 52760 }, { "epoch": 15.353505964503928, "grad_norm": 0.09748369455337524, "learning_rate": 1.175579366592014e-06, "loss": 0.0118, "step": 52770 }, { "epoch": 15.356415478615071, "grad_norm": 0.08114702999591827, "learning_rate": 1.1723731198381971e-06, "loss": 0.0127, "step": 52780 }, { "epoch": 15.359324992726215, "grad_norm": 0.12519818544387817, "learning_rate": 1.1691710735523288e-06, "loss": 0.0109, "step": 52790 }, { "epoch": 15.362234506837359, "grad_norm": 0.14387091994285583, "learning_rate": 1.165973228707104e-06, "loss": 0.0124, "step": 52800 }, { "epoch": 15.365144020948502, "grad_norm": 0.07857068628072739, "learning_rate": 1.162779586273951e-06, "loss": 0.0122, "step": 52810 }, { "epoch": 15.368053535059644, "grad_norm": 0.09091871976852417, "learning_rate": 1.1595901472230052e-06, "loss": 0.0118, "step": 52820 }, { "epoch": 15.370963049170788, "grad_norm": 0.06758249551057816, "learning_rate": 1.1564049125231418e-06, "loss": 0.01, "step": 52830 }, { "epoch": 15.373872563281932, "grad_norm": 0.08940429985523224, "learning_rate": 1.1532238831419417e-06, "loss": 0.0115, "step": 52840 }, { "epoch": 15.376782077393075, "grad_norm": 0.09447848796844482, "learning_rate": 1.150047060045727e-06, "loss": 0.0105, "step": 52850 }, { "epoch": 15.37969159150422, "grad_norm": 0.08775518089532852, "learning_rate": 1.1468744441995271e-06, "loss": 0.0127, "step": 52860 }, { "epoch": 15.382601105615363, "grad_norm": 0.1161673441529274, "learning_rate": 1.1437060365670998e-06, "loss": 0.0109, "step": 52870 }, { "epoch": 15.385510619726507, "grad_norm": 0.09733232855796814, "learning_rate": 1.1405418381109267e-06, "loss": 0.0114, "step": 52880 }, { "epoch": 15.388420133837649, "grad_norm": 0.11452895402908325, "learning_rate": 1.1373818497922007e-06, "loss": 0.0113, "step": 52890 }, { "epoch": 15.391329647948792, "grad_norm": 0.1051020622253418, "learning_rate": 1.134226072570852e-06, "loss": 0.0121, "step": 52900 }, { "epoch": 15.394239162059936, "grad_norm": 0.09539856016635895, "learning_rate": 1.1310745074055156e-06, "loss": 0.0118, "step": 52910 }, { "epoch": 15.39714867617108, "grad_norm": 0.06958944350481033, "learning_rate": 1.1279271552535576e-06, "loss": 0.0106, "step": 52920 }, { "epoch": 15.400058190282223, "grad_norm": 0.09527347981929779, "learning_rate": 1.1247840170710595e-06, "loss": 0.0108, "step": 52930 }, { "epoch": 15.402967704393367, "grad_norm": 0.13866329193115234, "learning_rate": 1.1216450938128225e-06, "loss": 0.0119, "step": 52940 }, { "epoch": 15.405877218504509, "grad_norm": 0.09242328256368637, "learning_rate": 1.118510386432366e-06, "loss": 0.012, "step": 52950 }, { "epoch": 15.408786732615653, "grad_norm": 0.08777590095996857, "learning_rate": 1.1153798958819393e-06, "loss": 0.012, "step": 52960 }, { "epoch": 15.411696246726796, "grad_norm": 0.07698385417461395, "learning_rate": 1.1122536231124942e-06, "loss": 0.0119, "step": 52970 }, { "epoch": 15.41460576083794, "grad_norm": 0.09506160765886307, "learning_rate": 1.1091315690737185e-06, "loss": 0.0115, "step": 52980 }, { "epoch": 15.417515274949084, "grad_norm": 0.16122063994407654, "learning_rate": 1.1060137347140008e-06, "loss": 0.0112, "step": 52990 }, { "epoch": 15.420424789060228, "grad_norm": 0.09449933469295502, "learning_rate": 1.1029001209804617e-06, "loss": 0.0119, "step": 53000 }, { "epoch": 15.423334303171371, "grad_norm": 0.0861939787864685, "learning_rate": 1.0997907288189363e-06, "loss": 0.0118, "step": 53010 }, { "epoch": 15.426243817282513, "grad_norm": 0.13496477901935577, "learning_rate": 1.0966855591739721e-06, "loss": 0.0131, "step": 53020 }, { "epoch": 15.429153331393657, "grad_norm": 0.08828004449605942, "learning_rate": 1.09358461298884e-06, "loss": 0.0127, "step": 53030 }, { "epoch": 15.4320628455048, "grad_norm": 0.08898396044969559, "learning_rate": 1.0904878912055228e-06, "loss": 0.0116, "step": 53040 }, { "epoch": 15.434972359615944, "grad_norm": 0.08237753063440323, "learning_rate": 1.0873953947647263e-06, "loss": 0.013, "step": 53050 }, { "epoch": 15.437881873727088, "grad_norm": 0.08461853861808777, "learning_rate": 1.0843071246058678e-06, "loss": 0.0136, "step": 53060 }, { "epoch": 15.440791387838232, "grad_norm": 0.09665067493915558, "learning_rate": 1.0812230816670832e-06, "loss": 0.0122, "step": 53070 }, { "epoch": 15.443700901949374, "grad_norm": 0.13760730624198914, "learning_rate": 1.0781432668852216e-06, "loss": 0.0123, "step": 53080 }, { "epoch": 15.446610416060517, "grad_norm": 0.1086936816573143, "learning_rate": 1.0750676811958493e-06, "loss": 0.0123, "step": 53090 }, { "epoch": 15.449519930171661, "grad_norm": 0.11110381782054901, "learning_rate": 1.0719963255332477e-06, "loss": 0.0124, "step": 53100 }, { "epoch": 15.452429444282805, "grad_norm": 0.17378588020801544, "learning_rate": 1.0689292008304202e-06, "loss": 0.0116, "step": 53110 }, { "epoch": 15.455338958393948, "grad_norm": 0.11397083103656769, "learning_rate": 1.0658663080190667e-06, "loss": 0.0113, "step": 53120 }, { "epoch": 15.458248472505092, "grad_norm": 0.09255436062812805, "learning_rate": 1.062807648029626e-06, "loss": 0.0102, "step": 53130 }, { "epoch": 15.461157986616236, "grad_norm": 0.10815723985433578, "learning_rate": 1.0597532217912254e-06, "loss": 0.0113, "step": 53140 }, { "epoch": 15.464067500727378, "grad_norm": 0.08698753267526627, "learning_rate": 1.056703030231727e-06, "loss": 0.0106, "step": 53150 }, { "epoch": 15.466977014838521, "grad_norm": 0.08097495883703232, "learning_rate": 1.0536570742776969e-06, "loss": 0.0124, "step": 53160 }, { "epoch": 15.469886528949665, "grad_norm": 0.13826751708984375, "learning_rate": 1.0506153548544163e-06, "loss": 0.0113, "step": 53170 }, { "epoch": 15.472796043060809, "grad_norm": 0.07271800190210342, "learning_rate": 1.0475778728858793e-06, "loss": 0.0117, "step": 53180 }, { "epoch": 15.475705557171953, "grad_norm": 0.13451313972473145, "learning_rate": 1.0445446292947896e-06, "loss": 0.0105, "step": 53190 }, { "epoch": 15.478615071283096, "grad_norm": 0.08188934624195099, "learning_rate": 1.041515625002572e-06, "loss": 0.0127, "step": 53200 }, { "epoch": 15.48152458539424, "grad_norm": 0.08288110792636871, "learning_rate": 1.038490860929352e-06, "loss": 0.0105, "step": 53210 }, { "epoch": 15.484434099505382, "grad_norm": 0.08876632899045944, "learning_rate": 1.0354703379939779e-06, "loss": 0.0117, "step": 53220 }, { "epoch": 15.487343613616526, "grad_norm": 0.10227343440055847, "learning_rate": 1.032454057114004e-06, "loss": 0.0126, "step": 53230 }, { "epoch": 15.49025312772767, "grad_norm": 0.06516127288341522, "learning_rate": 1.0294420192056957e-06, "loss": 0.0111, "step": 53240 }, { "epoch": 15.493162641838813, "grad_norm": 0.17319194972515106, "learning_rate": 1.0264342251840298e-06, "loss": 0.0122, "step": 53250 }, { "epoch": 15.496072155949957, "grad_norm": 0.10818959772586823, "learning_rate": 1.0234306759626966e-06, "loss": 0.0123, "step": 53260 }, { "epoch": 15.4989816700611, "grad_norm": 0.10896884649991989, "learning_rate": 1.0204313724540898e-06, "loss": 0.013, "step": 53270 }, { "epoch": 15.501891184172242, "grad_norm": 0.08979371935129166, "learning_rate": 1.0174363155693306e-06, "loss": 0.012, "step": 53280 }, { "epoch": 15.504800698283386, "grad_norm": 0.11854945123195648, "learning_rate": 1.0144455062182229e-06, "loss": 0.0115, "step": 53290 }, { "epoch": 15.50771021239453, "grad_norm": 0.09441353380680084, "learning_rate": 1.011458945309308e-06, "loss": 0.0104, "step": 53300 }, { "epoch": 15.510619726505674, "grad_norm": 0.11519830673933029, "learning_rate": 1.0084766337498175e-06, "loss": 0.0105, "step": 53310 }, { "epoch": 15.513529240616817, "grad_norm": 0.15228819847106934, "learning_rate": 1.0054985724457027e-06, "loss": 0.0116, "step": 53320 }, { "epoch": 15.516438754727961, "grad_norm": 0.09938905388116837, "learning_rate": 1.002524762301617e-06, "loss": 0.0128, "step": 53330 }, { "epoch": 15.519348268839105, "grad_norm": 0.08621290326118469, "learning_rate": 9.995552042209245e-07, "loss": 0.0113, "step": 53340 }, { "epoch": 15.522257782950247, "grad_norm": 0.1864723265171051, "learning_rate": 9.965898991057065e-07, "loss": 0.0122, "step": 53350 }, { "epoch": 15.52516729706139, "grad_norm": 0.12843111157417297, "learning_rate": 9.936288478567313e-07, "loss": 0.0129, "step": 53360 }, { "epoch": 15.528076811172534, "grad_norm": 0.10364654660224915, "learning_rate": 9.906720513735007e-07, "loss": 0.0129, "step": 53370 }, { "epoch": 15.530986325283678, "grad_norm": 0.17381717264652252, "learning_rate": 9.87719510554203e-07, "loss": 0.012, "step": 53380 }, { "epoch": 15.533895839394821, "grad_norm": 0.08486466854810715, "learning_rate": 9.847712262957476e-07, "loss": 0.0115, "step": 53390 }, { "epoch": 15.536805353505965, "grad_norm": 0.07147325575351715, "learning_rate": 9.818271994937412e-07, "loss": 0.0132, "step": 53400 }, { "epoch": 15.539714867617107, "grad_norm": 0.10286661237478256, "learning_rate": 9.78887431042505e-07, "loss": 0.0113, "step": 53410 }, { "epoch": 15.54262438172825, "grad_norm": 0.11618088185787201, "learning_rate": 9.759519218350566e-07, "loss": 0.0108, "step": 53420 }, { "epoch": 15.545533895839394, "grad_norm": 0.08169975876808167, "learning_rate": 9.730206727631374e-07, "loss": 0.0109, "step": 53430 }, { "epoch": 15.548443409950538, "grad_norm": 0.08598600327968597, "learning_rate": 9.70093684717172e-07, "loss": 0.0122, "step": 53440 }, { "epoch": 15.551352924061682, "grad_norm": 0.07300785928964615, "learning_rate": 9.671709585863074e-07, "loss": 0.0122, "step": 53450 }, { "epoch": 15.554262438172826, "grad_norm": 0.10365014523267746, "learning_rate": 9.64252495258392e-07, "loss": 0.0135, "step": 53460 }, { "epoch": 15.55717195228397, "grad_norm": 0.11665376275777817, "learning_rate": 9.613382956199761e-07, "loss": 0.0126, "step": 53470 }, { "epoch": 15.560081466395111, "grad_norm": 0.08224164694547653, "learning_rate": 9.58428360556317e-07, "loss": 0.0121, "step": 53480 }, { "epoch": 15.562990980506255, "grad_norm": 0.10088920593261719, "learning_rate": 9.555226909513725e-07, "loss": 0.0119, "step": 53490 }, { "epoch": 15.565900494617399, "grad_norm": 0.08435119688510895, "learning_rate": 9.526212876878165e-07, "loss": 0.0111, "step": 53500 }, { "epoch": 15.568810008728542, "grad_norm": 0.08351238071918488, "learning_rate": 9.49724151647009e-07, "loss": 0.0112, "step": 53510 }, { "epoch": 15.571719522839686, "grad_norm": 0.09054926037788391, "learning_rate": 9.468312837090293e-07, "loss": 0.0128, "step": 53520 }, { "epoch": 15.57462903695083, "grad_norm": 0.09141785651445389, "learning_rate": 9.439426847526544e-07, "loss": 0.0125, "step": 53530 }, { "epoch": 15.577538551061973, "grad_norm": 0.06276761740446091, "learning_rate": 9.410583556553626e-07, "loss": 0.0106, "step": 53540 }, { "epoch": 15.580448065173115, "grad_norm": 0.09499858319759369, "learning_rate": 9.381782972933395e-07, "loss": 0.0119, "step": 53550 }, { "epoch": 15.58335757928426, "grad_norm": 0.12465433776378632, "learning_rate": 9.353025105414669e-07, "loss": 0.0122, "step": 53560 }, { "epoch": 15.586267093395403, "grad_norm": 0.09563599526882172, "learning_rate": 9.324309962733346e-07, "loss": 0.0102, "step": 53570 }, { "epoch": 15.589176607506547, "grad_norm": 0.10639164596796036, "learning_rate": 9.295637553612347e-07, "loss": 0.0118, "step": 53580 }, { "epoch": 15.59208612161769, "grad_norm": 0.12847422063350677, "learning_rate": 9.267007886761536e-07, "loss": 0.0105, "step": 53590 }, { "epoch": 15.594995635728834, "grad_norm": 0.07449863106012344, "learning_rate": 9.238420970877909e-07, "loss": 0.011, "step": 53600 }, { "epoch": 15.597905149839976, "grad_norm": 0.08976880460977554, "learning_rate": 9.209876814645418e-07, "loss": 0.0114, "step": 53610 }, { "epoch": 15.60081466395112, "grad_norm": 0.09617619216442108, "learning_rate": 9.181375426734978e-07, "loss": 0.0126, "step": 53620 }, { "epoch": 15.603724178062263, "grad_norm": 0.12408162653446198, "learning_rate": 9.152916815804597e-07, "loss": 0.011, "step": 53630 }, { "epoch": 15.606633692173407, "grad_norm": 0.1581140011548996, "learning_rate": 9.124500990499224e-07, "loss": 0.0118, "step": 53640 }, { "epoch": 15.60954320628455, "grad_norm": 0.0918363481760025, "learning_rate": 9.096127959450873e-07, "loss": 0.0111, "step": 53650 }, { "epoch": 15.612452720395694, "grad_norm": 0.09304439276456833, "learning_rate": 9.067797731278465e-07, "loss": 0.0119, "step": 53660 }, { "epoch": 15.615362234506838, "grad_norm": 0.07801972329616547, "learning_rate": 9.039510314588034e-07, "loss": 0.0135, "step": 53670 }, { "epoch": 15.61827174861798, "grad_norm": 0.08855285495519638, "learning_rate": 9.011265717972489e-07, "loss": 0.012, "step": 53680 }, { "epoch": 15.621181262729124, "grad_norm": 0.08053193986415863, "learning_rate": 8.983063950011866e-07, "loss": 0.012, "step": 53690 }, { "epoch": 15.624090776840267, "grad_norm": 0.0977129340171814, "learning_rate": 8.954905019273063e-07, "loss": 0.0107, "step": 53700 }, { "epoch": 15.627000290951411, "grad_norm": 0.09205608069896698, "learning_rate": 8.926788934310037e-07, "loss": 0.0111, "step": 53710 }, { "epoch": 15.629909805062555, "grad_norm": 0.11105834692716599, "learning_rate": 8.898715703663707e-07, "loss": 0.0125, "step": 53720 }, { "epoch": 15.632819319173699, "grad_norm": 0.08923038095235825, "learning_rate": 8.870685335861983e-07, "loss": 0.0108, "step": 53730 }, { "epoch": 15.63572883328484, "grad_norm": 0.12678050994873047, "learning_rate": 8.842697839419722e-07, "loss": 0.0108, "step": 53740 }, { "epoch": 15.638638347395984, "grad_norm": 0.0813073217868805, "learning_rate": 8.814753222838856e-07, "loss": 0.0122, "step": 53750 }, { "epoch": 15.641547861507128, "grad_norm": 0.12272245436906815, "learning_rate": 8.786851494608162e-07, "loss": 0.0123, "step": 53760 }, { "epoch": 15.644457375618272, "grad_norm": 0.0879233181476593, "learning_rate": 8.758992663203475e-07, "loss": 0.0105, "step": 53770 }, { "epoch": 15.647366889729415, "grad_norm": 0.07664027810096741, "learning_rate": 8.73117673708756e-07, "loss": 0.0111, "step": 53780 }, { "epoch": 15.650276403840559, "grad_norm": 0.09895162284374237, "learning_rate": 8.70340372471014e-07, "loss": 0.0117, "step": 53790 }, { "epoch": 15.653185917951703, "grad_norm": 0.09311854839324951, "learning_rate": 8.675673634508002e-07, "loss": 0.0122, "step": 53800 }, { "epoch": 15.656095432062845, "grad_norm": 0.11481723934412003, "learning_rate": 8.647986474904723e-07, "loss": 0.0121, "step": 53810 }, { "epoch": 15.659004946173988, "grad_norm": 0.10403174161911011, "learning_rate": 8.620342254310992e-07, "loss": 0.0121, "step": 53820 }, { "epoch": 15.661914460285132, "grad_norm": 0.08115637302398682, "learning_rate": 8.592740981124342e-07, "loss": 0.0118, "step": 53830 }, { "epoch": 15.664823974396276, "grad_norm": 0.08116407692432404, "learning_rate": 8.565182663729365e-07, "loss": 0.0111, "step": 53840 }, { "epoch": 15.66773348850742, "grad_norm": 0.30938491225242615, "learning_rate": 8.537667310497515e-07, "loss": 0.0115, "step": 53850 }, { "epoch": 15.670643002618563, "grad_norm": 0.08783246576786041, "learning_rate": 8.510194929787241e-07, "loss": 0.0102, "step": 53860 }, { "epoch": 15.673552516729707, "grad_norm": 0.09463668614625931, "learning_rate": 8.482765529943931e-07, "loss": 0.014, "step": 53870 }, { "epoch": 15.676462030840849, "grad_norm": 0.07552702724933624, "learning_rate": 8.455379119299889e-07, "loss": 0.0107, "step": 53880 }, { "epoch": 15.679371544951993, "grad_norm": 0.0783257856965065, "learning_rate": 8.428035706174392e-07, "loss": 0.0127, "step": 53890 }, { "epoch": 15.682281059063136, "grad_norm": 0.10752589255571365, "learning_rate": 8.400735298873679e-07, "loss": 0.0111, "step": 53900 }, { "epoch": 15.68519057317428, "grad_norm": 0.08911347389221191, "learning_rate": 8.373477905690869e-07, "loss": 0.0112, "step": 53910 }, { "epoch": 15.688100087285424, "grad_norm": 0.11137645691633224, "learning_rate": 8.346263534906051e-07, "loss": 0.01, "step": 53920 }, { "epoch": 15.691009601396567, "grad_norm": 0.09332557022571564, "learning_rate": 8.319092194786199e-07, "loss": 0.0118, "step": 53930 }, { "epoch": 15.693919115507711, "grad_norm": 0.10460235178470612, "learning_rate": 8.291963893585275e-07, "loss": 0.0122, "step": 53940 }, { "epoch": 15.696828629618853, "grad_norm": 0.07739679515361786, "learning_rate": 8.264878639544154e-07, "loss": 0.0128, "step": 53950 }, { "epoch": 15.699738143729997, "grad_norm": 0.09160152822732925, "learning_rate": 8.237836440890573e-07, "loss": 0.0116, "step": 53960 }, { "epoch": 15.70264765784114, "grad_norm": 0.09138782322406769, "learning_rate": 8.210837305839309e-07, "loss": 0.0128, "step": 53970 }, { "epoch": 15.705557171952284, "grad_norm": 0.08399567753076553, "learning_rate": 8.1838812425919e-07, "loss": 0.0109, "step": 53980 }, { "epoch": 15.708466686063428, "grad_norm": 0.094521164894104, "learning_rate": 8.156968259336983e-07, "loss": 0.0112, "step": 53990 }, { "epoch": 15.711376200174572, "grad_norm": 0.12120530009269714, "learning_rate": 8.130098364249916e-07, "loss": 0.0118, "step": 54000 }, { "epoch": 15.714285714285714, "grad_norm": 0.10022586584091187, "learning_rate": 8.10327156549312e-07, "loss": 0.0122, "step": 54010 }, { "epoch": 15.717195228396857, "grad_norm": 0.11220133304595947, "learning_rate": 8.076487871215859e-07, "loss": 0.0113, "step": 54020 }, { "epoch": 15.720104742508001, "grad_norm": 0.10182680189609528, "learning_rate": 8.04974728955431e-07, "loss": 0.0109, "step": 54030 }, { "epoch": 15.723014256619145, "grad_norm": 0.12482817471027374, "learning_rate": 8.023049828631557e-07, "loss": 0.0137, "step": 54040 }, { "epoch": 15.725923770730288, "grad_norm": 0.09306765347719193, "learning_rate": 7.996395496557562e-07, "loss": 0.0114, "step": 54050 }, { "epoch": 15.728833284841432, "grad_norm": 0.12656043469905853, "learning_rate": 7.96978430142925e-07, "loss": 0.0118, "step": 54060 }, { "epoch": 15.731742798952574, "grad_norm": 0.08442416787147522, "learning_rate": 7.943216251330371e-07, "loss": 0.0104, "step": 54070 }, { "epoch": 15.734652313063718, "grad_norm": 0.09374550729990005, "learning_rate": 7.916691354331618e-07, "loss": 0.0099, "step": 54080 }, { "epoch": 15.737561827174861, "grad_norm": 0.11977610737085342, "learning_rate": 7.890209618490512e-07, "loss": 0.0127, "step": 54090 }, { "epoch": 15.740471341286005, "grad_norm": 0.07829400151968002, "learning_rate": 7.863771051851553e-07, "loss": 0.0113, "step": 54100 }, { "epoch": 15.743380855397149, "grad_norm": 0.07913149893283844, "learning_rate": 7.837375662446033e-07, "loss": 0.0122, "step": 54110 }, { "epoch": 15.746290369508293, "grad_norm": 0.12901179492473602, "learning_rate": 7.811023458292238e-07, "loss": 0.0125, "step": 54120 }, { "epoch": 15.749199883619436, "grad_norm": 0.09208443760871887, "learning_rate": 7.784714447395197e-07, "loss": 0.0113, "step": 54130 }, { "epoch": 15.752109397730578, "grad_norm": 0.08025003969669342, "learning_rate": 7.758448637746968e-07, "loss": 0.012, "step": 54140 }, { "epoch": 15.755018911841722, "grad_norm": 0.0994059145450592, "learning_rate": 7.732226037326334e-07, "loss": 0.0106, "step": 54150 }, { "epoch": 15.757928425952866, "grad_norm": 0.12172173708677292, "learning_rate": 7.706046654099092e-07, "loss": 0.0122, "step": 54160 }, { "epoch": 15.76083794006401, "grad_norm": 0.09142499417066574, "learning_rate": 7.679910496017812e-07, "loss": 0.0123, "step": 54170 }, { "epoch": 15.763747454175153, "grad_norm": 0.09703979641199112, "learning_rate": 7.653817571021976e-07, "loss": 0.0116, "step": 54180 }, { "epoch": 15.766656968286297, "grad_norm": 0.09364034980535507, "learning_rate": 7.627767887037945e-07, "loss": 0.0109, "step": 54190 }, { "epoch": 15.76956648239744, "grad_norm": 0.07438793033361435, "learning_rate": 7.601761451978872e-07, "loss": 0.0112, "step": 54200 }, { "epoch": 15.772475996508582, "grad_norm": 0.07024288922548294, "learning_rate": 7.575798273744889e-07, "loss": 0.0116, "step": 54210 }, { "epoch": 15.775385510619726, "grad_norm": 0.09810782968997955, "learning_rate": 7.549878360222901e-07, "loss": 0.0119, "step": 54220 }, { "epoch": 15.77829502473087, "grad_norm": 0.12374924123287201, "learning_rate": 7.524001719286694e-07, "loss": 0.0114, "step": 54230 }, { "epoch": 15.781204538842013, "grad_norm": 0.1006663590669632, "learning_rate": 7.498168358796898e-07, "loss": 0.012, "step": 54240 }, { "epoch": 15.784114052953157, "grad_norm": 0.09439761191606522, "learning_rate": 7.472378286601e-07, "loss": 0.0109, "step": 54250 }, { "epoch": 15.7870235670643, "grad_norm": 0.06771901249885559, "learning_rate": 7.446631510533352e-07, "loss": 0.0127, "step": 54260 }, { "epoch": 15.789933081175445, "grad_norm": 0.15035393834114075, "learning_rate": 7.420928038415181e-07, "loss": 0.012, "step": 54270 }, { "epoch": 15.792842595286587, "grad_norm": 0.09238880127668381, "learning_rate": 7.39526787805444e-07, "loss": 0.0125, "step": 54280 }, { "epoch": 15.79575210939773, "grad_norm": 0.08263776451349258, "learning_rate": 7.369651037246111e-07, "loss": 0.0122, "step": 54290 }, { "epoch": 15.798661623508874, "grad_norm": 0.08483795821666718, "learning_rate": 7.344077523771819e-07, "loss": 0.0114, "step": 54300 }, { "epoch": 15.801571137620018, "grad_norm": 0.135190948843956, "learning_rate": 7.318547345400184e-07, "loss": 0.012, "step": 54310 }, { "epoch": 15.804480651731161, "grad_norm": 0.09024851024150848, "learning_rate": 7.293060509886567e-07, "loss": 0.0112, "step": 54320 }, { "epoch": 15.807390165842305, "grad_norm": 0.07647358626127243, "learning_rate": 7.267617024973228e-07, "loss": 0.0108, "step": 54330 }, { "epoch": 15.810299679953447, "grad_norm": 0.08566711097955704, "learning_rate": 7.242216898389197e-07, "loss": 0.0115, "step": 54340 }, { "epoch": 15.81320919406459, "grad_norm": 0.1164746955037117, "learning_rate": 7.216860137850356e-07, "loss": 0.0109, "step": 54350 }, { "epoch": 15.816118708175734, "grad_norm": 0.09196373075246811, "learning_rate": 7.191546751059442e-07, "loss": 0.0115, "step": 54360 }, { "epoch": 15.819028222286878, "grad_norm": 0.09709078073501587, "learning_rate": 7.166276745706002e-07, "loss": 0.0108, "step": 54370 }, { "epoch": 15.821937736398022, "grad_norm": 0.12500418722629547, "learning_rate": 7.141050129466364e-07, "loss": 0.0131, "step": 54380 }, { "epoch": 15.824847250509166, "grad_norm": 0.08838299661874771, "learning_rate": 7.115866910003732e-07, "loss": 0.011, "step": 54390 }, { "epoch": 15.82775676462031, "grad_norm": 0.09377413243055344, "learning_rate": 7.090727094968103e-07, "loss": 0.0114, "step": 54400 }, { "epoch": 15.830666278731451, "grad_norm": 0.1125236228108406, "learning_rate": 7.065630691996283e-07, "loss": 0.0117, "step": 54410 }, { "epoch": 15.833575792842595, "grad_norm": 0.1324186474084854, "learning_rate": 7.040577708711877e-07, "loss": 0.0114, "step": 54420 }, { "epoch": 15.836485306953739, "grad_norm": 0.09280654042959213, "learning_rate": 7.015568152725344e-07, "loss": 0.0122, "step": 54430 }, { "epoch": 15.839394821064882, "grad_norm": 0.10874421894550323, "learning_rate": 6.990602031633941e-07, "loss": 0.012, "step": 54440 }, { "epoch": 15.842304335176026, "grad_norm": 0.10084103792905807, "learning_rate": 6.965679353021653e-07, "loss": 0.0117, "step": 54450 }, { "epoch": 15.84521384928717, "grad_norm": 0.08665606379508972, "learning_rate": 6.940800124459407e-07, "loss": 0.012, "step": 54460 }, { "epoch": 15.848123363398312, "grad_norm": 0.07949475944042206, "learning_rate": 6.915964353504822e-07, "loss": 0.0107, "step": 54470 }, { "epoch": 15.851032877509455, "grad_norm": 0.09909848123788834, "learning_rate": 6.891172047702348e-07, "loss": 0.0129, "step": 54480 }, { "epoch": 15.853942391620599, "grad_norm": 0.0866534486413002, "learning_rate": 6.866423214583245e-07, "loss": 0.0104, "step": 54490 }, { "epoch": 15.856851905731743, "grad_norm": 0.0863005742430687, "learning_rate": 6.84171786166553e-07, "loss": 0.011, "step": 54500 }, { "epoch": 15.859761419842886, "grad_norm": 0.07949119806289673, "learning_rate": 6.817055996454069e-07, "loss": 0.0121, "step": 54510 }, { "epoch": 15.86267093395403, "grad_norm": 0.08563869446516037, "learning_rate": 6.792437626440451e-07, "loss": 0.0116, "step": 54520 }, { "epoch": 15.865580448065174, "grad_norm": 0.09091850370168686, "learning_rate": 6.767862759103128e-07, "loss": 0.0137, "step": 54530 }, { "epoch": 15.868489962176316, "grad_norm": 0.09345229715108871, "learning_rate": 6.74333140190726e-07, "loss": 0.0127, "step": 54540 }, { "epoch": 15.87139947628746, "grad_norm": 0.10930177569389343, "learning_rate": 6.718843562304838e-07, "loss": 0.0119, "step": 54550 }, { "epoch": 15.874308990398603, "grad_norm": 0.09053906798362732, "learning_rate": 6.694399247734628e-07, "loss": 0.0118, "step": 54560 }, { "epoch": 15.877218504509747, "grad_norm": 0.06080097332596779, "learning_rate": 6.669998465622157e-07, "loss": 0.0111, "step": 54570 }, { "epoch": 15.88012801862089, "grad_norm": 0.08511568605899811, "learning_rate": 6.645641223379729e-07, "loss": 0.0115, "step": 54580 }, { "epoch": 15.883037532732034, "grad_norm": 0.12849505245685577, "learning_rate": 6.621327528406474e-07, "loss": 0.0118, "step": 54590 }, { "epoch": 15.885947046843178, "grad_norm": 0.10449524223804474, "learning_rate": 6.597057388088201e-07, "loss": 0.0122, "step": 54600 }, { "epoch": 15.88885656095432, "grad_norm": 0.07742833346128464, "learning_rate": 6.572830809797564e-07, "loss": 0.012, "step": 54610 }, { "epoch": 15.891766075065464, "grad_norm": 0.08178432285785675, "learning_rate": 6.548647800893942e-07, "loss": 0.0121, "step": 54620 }, { "epoch": 15.894675589176607, "grad_norm": 0.09489351511001587, "learning_rate": 6.524508368723508e-07, "loss": 0.0115, "step": 54630 }, { "epoch": 15.897585103287751, "grad_norm": 0.085108682513237, "learning_rate": 6.500412520619198e-07, "loss": 0.0102, "step": 54640 }, { "epoch": 15.900494617398895, "grad_norm": 0.10231063514947891, "learning_rate": 6.476360263900654e-07, "loss": 0.0107, "step": 54650 }, { "epoch": 15.903404131510039, "grad_norm": 0.07310129702091217, "learning_rate": 6.452351605874368e-07, "loss": 0.01, "step": 54660 }, { "epoch": 15.90631364562118, "grad_norm": 0.08160512149333954, "learning_rate": 6.428386553833487e-07, "loss": 0.013, "step": 54670 }, { "epoch": 15.909223159732324, "grad_norm": 0.0760972872376442, "learning_rate": 6.404465115057989e-07, "loss": 0.0114, "step": 54680 }, { "epoch": 15.912132673843468, "grad_norm": 0.09755752235651016, "learning_rate": 6.380587296814561e-07, "loss": 0.0126, "step": 54690 }, { "epoch": 15.915042187954612, "grad_norm": 0.08947988599538803, "learning_rate": 6.356753106356666e-07, "loss": 0.0113, "step": 54700 }, { "epoch": 15.917951702065755, "grad_norm": 0.09579373151063919, "learning_rate": 6.332962550924498e-07, "loss": 0.0128, "step": 54710 }, { "epoch": 15.920861216176899, "grad_norm": 0.09021814912557602, "learning_rate": 6.309215637745008e-07, "loss": 0.0108, "step": 54720 }, { "epoch": 15.923770730288043, "grad_norm": 0.0946640744805336, "learning_rate": 6.285512374031843e-07, "loss": 0.0118, "step": 54730 }, { "epoch": 15.926680244399185, "grad_norm": 0.10411450266838074, "learning_rate": 6.261852766985476e-07, "loss": 0.0106, "step": 54740 }, { "epoch": 15.929589758510328, "grad_norm": 0.1537160873413086, "learning_rate": 6.238236823793059e-07, "loss": 0.0108, "step": 54750 }, { "epoch": 15.932499272621472, "grad_norm": 0.09869546443223953, "learning_rate": 6.214664551628485e-07, "loss": 0.0114, "step": 54760 }, { "epoch": 15.935408786732616, "grad_norm": 0.09556520730257034, "learning_rate": 6.191135957652377e-07, "loss": 0.0117, "step": 54770 }, { "epoch": 15.93831830084376, "grad_norm": 0.09685932099819183, "learning_rate": 6.167651049012101e-07, "loss": 0.0108, "step": 54780 }, { "epoch": 15.941227814954903, "grad_norm": 0.09615948051214218, "learning_rate": 6.144209832841802e-07, "loss": 0.0105, "step": 54790 }, { "epoch": 15.944137329066045, "grad_norm": 0.08786488324403763, "learning_rate": 6.120812316262231e-07, "loss": 0.0107, "step": 54800 }, { "epoch": 15.947046843177189, "grad_norm": 0.098539337515831, "learning_rate": 6.097458506381004e-07, "loss": 0.0115, "step": 54810 }, { "epoch": 15.949956357288332, "grad_norm": 0.14235273003578186, "learning_rate": 6.074148410292313e-07, "loss": 0.0126, "step": 54820 }, { "epoch": 15.952865871399476, "grad_norm": 0.10345710813999176, "learning_rate": 6.050882035077227e-07, "loss": 0.0114, "step": 54830 }, { "epoch": 15.95577538551062, "grad_norm": 0.08580337464809418, "learning_rate": 6.027659387803391e-07, "loss": 0.013, "step": 54840 }, { "epoch": 15.958684899621764, "grad_norm": 0.09380766749382019, "learning_rate": 6.004480475525282e-07, "loss": 0.0117, "step": 54850 }, { "epoch": 15.961594413732907, "grad_norm": 0.12029653787612915, "learning_rate": 5.98134530528403e-07, "loss": 0.012, "step": 54860 }, { "epoch": 15.96450392784405, "grad_norm": 0.08391543477773666, "learning_rate": 5.958253884107468e-07, "loss": 0.011, "step": 54870 }, { "epoch": 15.967413441955193, "grad_norm": 0.10825750231742859, "learning_rate": 5.935206219010181e-07, "loss": 0.011, "step": 54880 }, { "epoch": 15.970322956066337, "grad_norm": 0.07541627436876297, "learning_rate": 5.91220231699342e-07, "loss": 0.0117, "step": 54890 }, { "epoch": 15.97323247017748, "grad_norm": 0.1175304651260376, "learning_rate": 5.889242185045196e-07, "loss": 0.011, "step": 54900 }, { "epoch": 15.976141984288624, "grad_norm": 0.06821153312921524, "learning_rate": 5.866325830140162e-07, "loss": 0.0098, "step": 54910 }, { "epoch": 15.979051498399768, "grad_norm": 0.14285917580127716, "learning_rate": 5.843453259239718e-07, "loss": 0.01, "step": 54920 }, { "epoch": 15.981961012510912, "grad_norm": 0.12391611188650131, "learning_rate": 5.820624479291936e-07, "loss": 0.0123, "step": 54930 }, { "epoch": 15.984870526622053, "grad_norm": 0.09670617431402206, "learning_rate": 5.797839497231605e-07, "loss": 0.0122, "step": 54940 }, { "epoch": 15.987780040733197, "grad_norm": 0.11669552326202393, "learning_rate": 5.775098319980171e-07, "loss": 0.0128, "step": 54950 }, { "epoch": 15.99068955484434, "grad_norm": 0.1073010265827179, "learning_rate": 5.752400954445875e-07, "loss": 0.0116, "step": 54960 }, { "epoch": 15.993599068955485, "grad_norm": 0.10980924218893051, "learning_rate": 5.7297474075235e-07, "loss": 0.0121, "step": 54970 }, { "epoch": 15.996508583066628, "grad_norm": 0.07103278487920761, "learning_rate": 5.70713768609466e-07, "loss": 0.0115, "step": 54980 }, { "epoch": 15.999418097177772, "grad_norm": 0.12027391046285629, "learning_rate": 5.68457179702751e-07, "loss": 0.0117, "step": 54990 }, { "epoch": 16.002327611288916, "grad_norm": 0.08158803731203079, "learning_rate": 5.66204974717705e-07, "loss": 0.0112, "step": 55000 }, { "epoch": 16.005237125400058, "grad_norm": 0.10304597020149231, "learning_rate": 5.639571543384858e-07, "loss": 0.0112, "step": 55010 }, { "epoch": 16.008146639511203, "grad_norm": 0.09579392522573471, "learning_rate": 5.617137192479205e-07, "loss": 0.0117, "step": 55020 }, { "epoch": 16.011056153622345, "grad_norm": 0.11536318063735962, "learning_rate": 5.594746701275072e-07, "loss": 0.0117, "step": 55030 }, { "epoch": 16.013965667733487, "grad_norm": 0.0875900611281395, "learning_rate": 5.572400076574069e-07, "loss": 0.0105, "step": 55040 }, { "epoch": 16.016875181844632, "grad_norm": 0.06630432605743408, "learning_rate": 5.55009732516455e-07, "loss": 0.0117, "step": 55050 }, { "epoch": 16.019784695955774, "grad_norm": 0.1195671409368515, "learning_rate": 5.527838453821494e-07, "loss": 0.0109, "step": 55060 }, { "epoch": 16.02269421006692, "grad_norm": 0.09213083237409592, "learning_rate": 5.505623469306542e-07, "loss": 0.012, "step": 55070 }, { "epoch": 16.025603724178062, "grad_norm": 0.06708521395921707, "learning_rate": 5.483452378368014e-07, "loss": 0.012, "step": 55080 }, { "epoch": 16.028513238289207, "grad_norm": 0.0942012146115303, "learning_rate": 5.461325187740939e-07, "loss": 0.0122, "step": 55090 }, { "epoch": 16.03142275240035, "grad_norm": 0.14871074259281158, "learning_rate": 5.439241904146908e-07, "loss": 0.0122, "step": 55100 }, { "epoch": 16.03433226651149, "grad_norm": 0.08768892288208008, "learning_rate": 5.417202534294336e-07, "loss": 0.0122, "step": 55110 }, { "epoch": 16.037241780622637, "grad_norm": 0.0778961107134819, "learning_rate": 5.395207084878101e-07, "loss": 0.0111, "step": 55120 }, { "epoch": 16.04015129473378, "grad_norm": 0.11156190931797028, "learning_rate": 5.373255562579927e-07, "loss": 0.0117, "step": 55130 }, { "epoch": 16.043060808844924, "grad_norm": 0.10786201804876328, "learning_rate": 5.351347974068027e-07, "loss": 0.0121, "step": 55140 }, { "epoch": 16.045970322956066, "grad_norm": 0.08369269222021103, "learning_rate": 5.329484325997397e-07, "loss": 0.0113, "step": 55150 }, { "epoch": 16.04887983706721, "grad_norm": 0.06946083158254623, "learning_rate": 5.307664625009639e-07, "loss": 0.0107, "step": 55160 }, { "epoch": 16.051789351178353, "grad_norm": 0.08110413700342178, "learning_rate": 5.28588887773297e-07, "loss": 0.012, "step": 55170 }, { "epoch": 16.054698865289495, "grad_norm": 0.09665783494710922, "learning_rate": 5.264157090782312e-07, "loss": 0.0113, "step": 55180 }, { "epoch": 16.05760837940064, "grad_norm": 0.0840873122215271, "learning_rate": 5.242469270759187e-07, "loss": 0.0112, "step": 55190 }, { "epoch": 16.060517893511783, "grad_norm": 0.10549677163362503, "learning_rate": 5.22082542425184e-07, "loss": 0.0114, "step": 55200 }, { "epoch": 16.063427407622928, "grad_norm": 0.12666450440883636, "learning_rate": 5.199225557835013e-07, "loss": 0.0126, "step": 55210 }, { "epoch": 16.06633692173407, "grad_norm": 0.2014133632183075, "learning_rate": 5.177669678070241e-07, "loss": 0.0118, "step": 55220 }, { "epoch": 16.069246435845216, "grad_norm": 0.19373179972171783, "learning_rate": 5.1561577915056e-07, "loss": 0.0124, "step": 55230 }, { "epoch": 16.072155949956358, "grad_norm": 0.10018176585435867, "learning_rate": 5.134689904675848e-07, "loss": 0.0122, "step": 55240 }, { "epoch": 16.0750654640675, "grad_norm": 0.12577983736991882, "learning_rate": 5.113266024102353e-07, "loss": 0.0114, "step": 55250 }, { "epoch": 16.077974978178645, "grad_norm": 0.10376441478729248, "learning_rate": 5.091886156293124e-07, "loss": 0.0116, "step": 55260 }, { "epoch": 16.080884492289787, "grad_norm": 0.0789157822728157, "learning_rate": 5.070550307742783e-07, "loss": 0.0109, "step": 55270 }, { "epoch": 16.083794006400932, "grad_norm": 0.10672671347856522, "learning_rate": 5.049258484932661e-07, "loss": 0.0138, "step": 55280 }, { "epoch": 16.086703520512074, "grad_norm": 0.09349855035543442, "learning_rate": 5.028010694330571e-07, "loss": 0.0122, "step": 55290 }, { "epoch": 16.089613034623216, "grad_norm": 0.12575940787792206, "learning_rate": 5.006806942391095e-07, "loss": 0.011, "step": 55300 }, { "epoch": 16.09252254873436, "grad_norm": 0.08663033694028854, "learning_rate": 4.985647235555335e-07, "loss": 0.0117, "step": 55310 }, { "epoch": 16.095432062845504, "grad_norm": 0.060183241963386536, "learning_rate": 4.964531580251064e-07, "loss": 0.0112, "step": 55320 }, { "epoch": 16.09834157695665, "grad_norm": 0.0936867743730545, "learning_rate": 4.943459982892667e-07, "loss": 0.012, "step": 55330 }, { "epoch": 16.10125109106779, "grad_norm": 0.10368341952562332, "learning_rate": 4.922432449881109e-07, "loss": 0.0118, "step": 55340 }, { "epoch": 16.104160605178937, "grad_norm": 0.11026355624198914, "learning_rate": 4.901448987604046e-07, "loss": 0.0119, "step": 55350 }, { "epoch": 16.10707011929008, "grad_norm": 0.07193240523338318, "learning_rate": 4.880509602435663e-07, "loss": 0.0132, "step": 55360 }, { "epoch": 16.10997963340122, "grad_norm": 0.08986035734415054, "learning_rate": 4.859614300736804e-07, "loss": 0.0126, "step": 55370 }, { "epoch": 16.112889147512366, "grad_norm": 0.0855267345905304, "learning_rate": 4.838763088854909e-07, "loss": 0.0113, "step": 55380 }, { "epoch": 16.115798661623508, "grad_norm": 0.08404204994440079, "learning_rate": 4.817955973124044e-07, "loss": 0.0116, "step": 55390 }, { "epoch": 16.118708175734653, "grad_norm": 0.09786461293697357, "learning_rate": 4.797192959864821e-07, "loss": 0.0122, "step": 55400 }, { "epoch": 16.121617689845795, "grad_norm": 0.09236423671245575, "learning_rate": 4.776474055384544e-07, "loss": 0.0132, "step": 55410 }, { "epoch": 16.12452720395694, "grad_norm": 0.09012234210968018, "learning_rate": 4.755799265977012e-07, "loss": 0.0134, "step": 55420 }, { "epoch": 16.127436718068083, "grad_norm": 0.0816141664981842, "learning_rate": 4.735168597922751e-07, "loss": 0.0123, "step": 55430 }, { "epoch": 16.130346232179225, "grad_norm": 0.07969717681407928, "learning_rate": 4.714582057488731e-07, "loss": 0.0112, "step": 55440 }, { "epoch": 16.13325574629037, "grad_norm": 0.0864548534154892, "learning_rate": 4.6940396509286667e-07, "loss": 0.011, "step": 55450 }, { "epoch": 16.136165260401512, "grad_norm": 0.12646251916885376, "learning_rate": 4.6735413844827825e-07, "loss": 0.0135, "step": 55460 }, { "epoch": 16.139074774512657, "grad_norm": 0.1204310953617096, "learning_rate": 4.653087264377898e-07, "loss": 0.0103, "step": 55470 }, { "epoch": 16.1419842886238, "grad_norm": 0.08176077157258987, "learning_rate": 4.6326772968274256e-07, "loss": 0.0108, "step": 55480 }, { "epoch": 16.144893802734945, "grad_norm": 0.11314960569143295, "learning_rate": 4.61231148803139e-07, "loss": 0.012, "step": 55490 }, { "epoch": 16.147803316846087, "grad_norm": 0.09932362288236618, "learning_rate": 4.591989844176409e-07, "loss": 0.0107, "step": 55500 }, { "epoch": 16.15071283095723, "grad_norm": 0.06386063247919083, "learning_rate": 4.57171237143561e-07, "loss": 0.0127, "step": 55510 }, { "epoch": 16.153622345068374, "grad_norm": 0.08819176256656647, "learning_rate": 4.551479075968784e-07, "loss": 0.011, "step": 55520 }, { "epoch": 16.156531859179516, "grad_norm": 0.13220293819904327, "learning_rate": 4.53128996392228e-07, "loss": 0.0137, "step": 55530 }, { "epoch": 16.15944137329066, "grad_norm": 0.08561873435974121, "learning_rate": 4.51114504142901e-07, "loss": 0.0112, "step": 55540 }, { "epoch": 16.162350887401804, "grad_norm": 0.07988420873880386, "learning_rate": 4.491044314608461e-07, "loss": 0.0117, "step": 55550 }, { "epoch": 16.16526040151295, "grad_norm": 0.09746721386909485, "learning_rate": 4.470987789566716e-07, "loss": 0.0119, "step": 55560 }, { "epoch": 16.16816991562409, "grad_norm": 0.13910913467407227, "learning_rate": 4.4509754723963846e-07, "loss": 0.0116, "step": 55570 }, { "epoch": 16.171079429735233, "grad_norm": 0.06954044848680496, "learning_rate": 4.431007369176737e-07, "loss": 0.0125, "step": 55580 }, { "epoch": 16.17398894384638, "grad_norm": 0.09846465289592743, "learning_rate": 4.411083485973488e-07, "loss": 0.0105, "step": 55590 }, { "epoch": 16.17689845795752, "grad_norm": 0.08846480399370193, "learning_rate": 4.3912038288390465e-07, "loss": 0.011, "step": 55600 }, { "epoch": 16.179807972068666, "grad_norm": 0.08076992630958557, "learning_rate": 4.371368403812298e-07, "loss": 0.0106, "step": 55610 }, { "epoch": 16.182717486179808, "grad_norm": 0.09706950932741165, "learning_rate": 4.3515772169187396e-07, "loss": 0.0112, "step": 55620 }, { "epoch": 16.18562700029095, "grad_norm": 0.09650707989931107, "learning_rate": 4.331830274170379e-07, "loss": 0.011, "step": 55630 }, { "epoch": 16.188536514402095, "grad_norm": 0.10804105550050735, "learning_rate": 4.3121275815658336e-07, "loss": 0.012, "step": 55640 }, { "epoch": 16.191446028513237, "grad_norm": 0.08774913102388382, "learning_rate": 4.2924691450903e-07, "loss": 0.0117, "step": 55650 }, { "epoch": 16.194355542624383, "grad_norm": 0.08712165057659149, "learning_rate": 4.2728549707154176e-07, "loss": 0.0114, "step": 55660 }, { "epoch": 16.197265056735525, "grad_norm": 0.09876680374145508, "learning_rate": 4.2532850643995203e-07, "loss": 0.0122, "step": 55670 }, { "epoch": 16.20017457084667, "grad_norm": 0.0843614861369133, "learning_rate": 4.2337594320873695e-07, "loss": 0.012, "step": 55680 }, { "epoch": 16.203084084957812, "grad_norm": 0.13271620869636536, "learning_rate": 4.2142780797103875e-07, "loss": 0.0119, "step": 55690 }, { "epoch": 16.205993599068954, "grad_norm": 0.08565133810043335, "learning_rate": 4.1948410131864733e-07, "loss": 0.0105, "step": 55700 }, { "epoch": 16.2089031131801, "grad_norm": 0.086240753531456, "learning_rate": 4.175448238420104e-07, "loss": 0.0125, "step": 55710 }, { "epoch": 16.21181262729124, "grad_norm": 0.11087960749864578, "learning_rate": 4.156099761302284e-07, "loss": 0.0135, "step": 55720 }, { "epoch": 16.214722141402387, "grad_norm": 0.08802825957536697, "learning_rate": 4.1367955877105787e-07, "loss": 0.0115, "step": 55730 }, { "epoch": 16.21763165551353, "grad_norm": 0.08120393753051758, "learning_rate": 4.1175357235090474e-07, "loss": 0.011, "step": 55740 }, { "epoch": 16.220541169624674, "grad_norm": 0.1454407274723053, "learning_rate": 4.0983201745483755e-07, "loss": 0.0114, "step": 55750 }, { "epoch": 16.223450683735816, "grad_norm": 0.08867525309324265, "learning_rate": 4.079148946665728e-07, "loss": 0.0118, "step": 55760 }, { "epoch": 16.226360197846958, "grad_norm": 0.08609255403280258, "learning_rate": 4.0600220456848305e-07, "loss": 0.012, "step": 55770 }, { "epoch": 16.229269711958104, "grad_norm": 0.08209852129220963, "learning_rate": 4.0409394774159004e-07, "loss": 0.0117, "step": 55780 }, { "epoch": 16.232179226069245, "grad_norm": 0.08246788382530212, "learning_rate": 4.0219012476557014e-07, "loss": 0.0132, "step": 55790 }, { "epoch": 16.23508874018039, "grad_norm": 0.09061240404844284, "learning_rate": 4.002907362187624e-07, "loss": 0.0117, "step": 55800 }, { "epoch": 16.237998254291533, "grad_norm": 0.09512477368116379, "learning_rate": 3.98395782678142e-07, "loss": 0.0125, "step": 55810 }, { "epoch": 16.24090776840268, "grad_norm": 0.09118790924549103, "learning_rate": 3.9650526471935344e-07, "loss": 0.0105, "step": 55820 }, { "epoch": 16.24381728251382, "grad_norm": 0.08083838224411011, "learning_rate": 3.946191829166806e-07, "loss": 0.0125, "step": 55830 }, { "epoch": 16.246726796624962, "grad_norm": 0.1271066814661026, "learning_rate": 3.92737537843067e-07, "loss": 0.0106, "step": 55840 }, { "epoch": 16.249636310736108, "grad_norm": 0.1232009008526802, "learning_rate": 3.9086033007010693e-07, "loss": 0.0118, "step": 55850 }, { "epoch": 16.25254582484725, "grad_norm": 0.08730781823396683, "learning_rate": 3.889875601680476e-07, "loss": 0.0108, "step": 55860 }, { "epoch": 16.255455338958395, "grad_norm": 0.08006994426250458, "learning_rate": 3.8711922870578564e-07, "loss": 0.012, "step": 55870 }, { "epoch": 16.258364853069537, "grad_norm": 0.07581445574760437, "learning_rate": 3.8525533625087037e-07, "loss": 0.0108, "step": 55880 }, { "epoch": 16.261274367180683, "grad_norm": 0.09333062171936035, "learning_rate": 3.8339588336950394e-07, "loss": 0.011, "step": 55890 }, { "epoch": 16.264183881291824, "grad_norm": 0.11004074662923813, "learning_rate": 3.815408706265394e-07, "loss": 0.0109, "step": 55900 }, { "epoch": 16.267093395402966, "grad_norm": 0.09670096635818481, "learning_rate": 3.796902985854811e-07, "loss": 0.0108, "step": 55910 }, { "epoch": 16.270002909514112, "grad_norm": 0.11281813681125641, "learning_rate": 3.7784416780848265e-07, "loss": 0.012, "step": 55920 }, { "epoch": 16.272912423625254, "grad_norm": 0.09360385686159134, "learning_rate": 3.760024788563521e-07, "loss": 0.0116, "step": 55930 }, { "epoch": 16.2758219377364, "grad_norm": 0.1302497833967209, "learning_rate": 3.741652322885403e-07, "loss": 0.0109, "step": 55940 }, { "epoch": 16.27873145184754, "grad_norm": 0.11477503925561905, "learning_rate": 3.7233242866316244e-07, "loss": 0.0114, "step": 55950 }, { "epoch": 16.281640965958683, "grad_norm": 0.09588705003261566, "learning_rate": 3.7050406853696815e-07, "loss": 0.0115, "step": 55960 }, { "epoch": 16.28455048006983, "grad_norm": 0.1356741189956665, "learning_rate": 3.6868015246537156e-07, "loss": 0.0115, "step": 55970 }, { "epoch": 16.28745999418097, "grad_norm": 0.08885449171066284, "learning_rate": 3.6686068100242445e-07, "loss": 0.0127, "step": 55980 }, { "epoch": 16.290369508292116, "grad_norm": 0.08863024413585663, "learning_rate": 3.650456547008363e-07, "loss": 0.012, "step": 55990 }, { "epoch": 16.293279022403258, "grad_norm": 0.08707119524478912, "learning_rate": 3.632350741119661e-07, "loss": 0.0108, "step": 56000 }, { "epoch": 16.296188536514403, "grad_norm": 0.10438374429941177, "learning_rate": 3.614289397858189e-07, "loss": 0.0129, "step": 56010 }, { "epoch": 16.299098050625545, "grad_norm": 0.09186164289712906, "learning_rate": 3.5962725227105084e-07, "loss": 0.0118, "step": 56020 }, { "epoch": 16.302007564736687, "grad_norm": 0.07810229063034058, "learning_rate": 3.578300121149675e-07, "loss": 0.0122, "step": 56030 }, { "epoch": 16.304917078847833, "grad_norm": 0.07295363396406174, "learning_rate": 3.56037219863522e-07, "loss": 0.0115, "step": 56040 }, { "epoch": 16.307826592958975, "grad_norm": 0.09508568048477173, "learning_rate": 3.542488760613155e-07, "loss": 0.0117, "step": 56050 }, { "epoch": 16.31073610707012, "grad_norm": 0.13738133013248444, "learning_rate": 3.5246498125160344e-07, "loss": 0.0114, "step": 56060 }, { "epoch": 16.313645621181262, "grad_norm": 0.11625742167234421, "learning_rate": 3.506855359762856e-07, "loss": 0.0122, "step": 56070 }, { "epoch": 16.316555135292408, "grad_norm": 0.0824635773897171, "learning_rate": 3.48910540775908e-07, "loss": 0.0127, "step": 56080 }, { "epoch": 16.31946464940355, "grad_norm": 0.09761599451303482, "learning_rate": 3.471399961896676e-07, "loss": 0.011, "step": 56090 }, { "epoch": 16.32237416351469, "grad_norm": 0.10848204791545868, "learning_rate": 3.453739027554109e-07, "loss": 0.0104, "step": 56100 }, { "epoch": 16.325283677625837, "grad_norm": 0.07626063376665115, "learning_rate": 3.4361226100962874e-07, "loss": 0.0115, "step": 56110 }, { "epoch": 16.32819319173698, "grad_norm": 0.1097143366932869, "learning_rate": 3.4185507148746463e-07, "loss": 0.0109, "step": 56120 }, { "epoch": 16.331102705848124, "grad_norm": 0.07757284492254257, "learning_rate": 3.4010233472270167e-07, "loss": 0.0105, "step": 56130 }, { "epoch": 16.334012219959266, "grad_norm": 0.1242896318435669, "learning_rate": 3.38354051247779e-07, "loss": 0.0112, "step": 56140 }, { "epoch": 16.336921734070412, "grad_norm": 0.07779652625322342, "learning_rate": 3.3661022159377497e-07, "loss": 0.011, "step": 56150 }, { "epoch": 16.339831248181554, "grad_norm": 0.13563069701194763, "learning_rate": 3.3487084629042264e-07, "loss": 0.0115, "step": 56160 }, { "epoch": 16.342740762292696, "grad_norm": 0.09398581832647324, "learning_rate": 3.3313592586609597e-07, "loss": 0.0117, "step": 56170 }, { "epoch": 16.34565027640384, "grad_norm": 0.08059889078140259, "learning_rate": 3.314054608478201e-07, "loss": 0.0104, "step": 56180 }, { "epoch": 16.348559790514983, "grad_norm": 0.09805990010499954, "learning_rate": 3.296794517612628e-07, "loss": 0.0112, "step": 56190 }, { "epoch": 16.35146930462613, "grad_norm": 0.08008219301700592, "learning_rate": 3.279578991307397e-07, "loss": 0.0106, "step": 56200 }, { "epoch": 16.35437881873727, "grad_norm": 0.07641521096229553, "learning_rate": 3.262408034792158e-07, "loss": 0.0109, "step": 56210 }, { "epoch": 16.357288332848416, "grad_norm": 0.10242879390716553, "learning_rate": 3.2452816532829723e-07, "loss": 0.0106, "step": 56220 }, { "epoch": 16.360197846959558, "grad_norm": 0.09650111198425293, "learning_rate": 3.2281998519823776e-07, "loss": 0.0116, "step": 56230 }, { "epoch": 16.3631073610707, "grad_norm": 0.09004564583301544, "learning_rate": 3.21116263607939e-07, "loss": 0.012, "step": 56240 }, { "epoch": 16.366016875181845, "grad_norm": 0.10921170562505722, "learning_rate": 3.1941700107494533e-07, "loss": 0.0117, "step": 56250 }, { "epoch": 16.368926389292987, "grad_norm": 0.1584516167640686, "learning_rate": 3.177221981154488e-07, "loss": 0.011, "step": 56260 }, { "epoch": 16.371835903404133, "grad_norm": 0.11298254132270813, "learning_rate": 3.160318552442859e-07, "loss": 0.0104, "step": 56270 }, { "epoch": 16.374745417515275, "grad_norm": 0.07342516630887985, "learning_rate": 3.1434597297493593e-07, "loss": 0.0113, "step": 56280 }, { "epoch": 16.377654931626417, "grad_norm": 0.0923619195818901, "learning_rate": 3.126645518195309e-07, "loss": 0.0111, "step": 56290 }, { "epoch": 16.380564445737562, "grad_norm": 0.13370274007320404, "learning_rate": 3.109875922888339e-07, "loss": 0.0127, "step": 56300 }, { "epoch": 16.383473959848704, "grad_norm": 0.07956815510988235, "learning_rate": 3.093150948922674e-07, "loss": 0.0098, "step": 56310 }, { "epoch": 16.38638347395985, "grad_norm": 0.10282596200704575, "learning_rate": 3.0764706013789004e-07, "loss": 0.0118, "step": 56320 }, { "epoch": 16.38929298807099, "grad_norm": 0.09289344400167465, "learning_rate": 3.059834885324064e-07, "loss": 0.0094, "step": 56330 }, { "epoch": 16.392202502182137, "grad_norm": 0.15149438381195068, "learning_rate": 3.0432438058116743e-07, "loss": 0.0112, "step": 56340 }, { "epoch": 16.39511201629328, "grad_norm": 0.07160747796297073, "learning_rate": 3.0266973678816145e-07, "loss": 0.0123, "step": 56350 }, { "epoch": 16.39802153040442, "grad_norm": 0.07329075038433075, "learning_rate": 3.0101955765603153e-07, "loss": 0.011, "step": 56360 }, { "epoch": 16.400931044515566, "grad_norm": 0.09988530725240707, "learning_rate": 2.993738436860566e-07, "loss": 0.0122, "step": 56370 }, { "epoch": 16.403840558626708, "grad_norm": 0.09261975437402725, "learning_rate": 2.9773259537815844e-07, "loss": 0.012, "step": 56380 }, { "epoch": 16.406750072737854, "grad_norm": 0.0889090895652771, "learning_rate": 2.960958132309083e-07, "loss": 0.0123, "step": 56390 }, { "epoch": 16.409659586848996, "grad_norm": 0.09840505570173264, "learning_rate": 2.94463497741515e-07, "loss": 0.0104, "step": 56400 }, { "epoch": 16.41256910096014, "grad_norm": 0.09153786301612854, "learning_rate": 2.928356494058337e-07, "loss": 0.0115, "step": 56410 }, { "epoch": 16.415478615071283, "grad_norm": 0.13266287744045258, "learning_rate": 2.912122687183638e-07, "loss": 0.0104, "step": 56420 }, { "epoch": 16.418388129182425, "grad_norm": 0.0897618904709816, "learning_rate": 2.895933561722408e-07, "loss": 0.0112, "step": 56430 }, { "epoch": 16.42129764329357, "grad_norm": 0.08577492088079453, "learning_rate": 2.87978912259253e-07, "loss": 0.0112, "step": 56440 }, { "epoch": 16.424207157404712, "grad_norm": 0.11173397302627563, "learning_rate": 2.863689374698214e-07, "loss": 0.0126, "step": 56450 }, { "epoch": 16.427116671515858, "grad_norm": 0.12233150750398636, "learning_rate": 2.847634322930165e-07, "loss": 0.0106, "step": 56460 }, { "epoch": 16.430026185627, "grad_norm": 0.09293437749147415, "learning_rate": 2.831623972165481e-07, "loss": 0.0119, "step": 56470 }, { "epoch": 16.432935699738145, "grad_norm": 0.09507735818624496, "learning_rate": 2.8156583272676726e-07, "loss": 0.0118, "step": 56480 }, { "epoch": 16.435845213849287, "grad_norm": 0.0797976553440094, "learning_rate": 2.7997373930866934e-07, "loss": 0.0116, "step": 56490 }, { "epoch": 16.43875472796043, "grad_norm": 0.08001953363418579, "learning_rate": 2.783861174458874e-07, "loss": 0.0123, "step": 56500 }, { "epoch": 16.441664242071575, "grad_norm": 0.06813587993383408, "learning_rate": 2.76802967620704e-07, "loss": 0.0116, "step": 56510 }, { "epoch": 16.444573756182717, "grad_norm": 0.09411055594682693, "learning_rate": 2.7522429031403286e-07, "loss": 0.0116, "step": 56520 }, { "epoch": 16.447483270293862, "grad_norm": 0.1049087718129158, "learning_rate": 2.736500860054386e-07, "loss": 0.0114, "step": 56530 }, { "epoch": 16.450392784405004, "grad_norm": 0.0907863974571228, "learning_rate": 2.7208035517312215e-07, "loss": 0.0101, "step": 56540 }, { "epoch": 16.45330229851615, "grad_norm": 0.09736393392086029, "learning_rate": 2.7051509829392527e-07, "loss": 0.0109, "step": 56550 }, { "epoch": 16.45621181262729, "grad_norm": 0.18955232203006744, "learning_rate": 2.6895431584333276e-07, "loss": 0.0113, "step": 56560 }, { "epoch": 16.459121326738433, "grad_norm": 0.10023850202560425, "learning_rate": 2.673980082954669e-07, "loss": 0.0119, "step": 56570 }, { "epoch": 16.46203084084958, "grad_norm": 0.08242524415254593, "learning_rate": 2.658461761230929e-07, "loss": 0.0132, "step": 56580 }, { "epoch": 16.46494035496072, "grad_norm": 0.11304015666246414, "learning_rate": 2.6429881979762203e-07, "loss": 0.0135, "step": 56590 }, { "epoch": 16.467849869071866, "grad_norm": 0.08697564899921417, "learning_rate": 2.6275593978909165e-07, "loss": 0.0103, "step": 56600 }, { "epoch": 16.470759383183008, "grad_norm": 0.11782906204462051, "learning_rate": 2.6121753656619506e-07, "loss": 0.0132, "step": 56610 }, { "epoch": 16.473668897294154, "grad_norm": 0.07128351181745529, "learning_rate": 2.596836105962552e-07, "loss": 0.0116, "step": 56620 }, { "epoch": 16.476578411405296, "grad_norm": 0.08200784027576447, "learning_rate": 2.581541623452394e-07, "loss": 0.0105, "step": 56630 }, { "epoch": 16.479487925516437, "grad_norm": 0.06558416783809662, "learning_rate": 2.566291922777525e-07, "loss": 0.011, "step": 56640 }, { "epoch": 16.482397439627583, "grad_norm": 0.10195565223693848, "learning_rate": 2.5510870085704076e-07, "loss": 0.0115, "step": 56650 }, { "epoch": 16.485306953738725, "grad_norm": 0.09483636915683746, "learning_rate": 2.535926885449913e-07, "loss": 0.0113, "step": 56660 }, { "epoch": 16.48821646784987, "grad_norm": 0.07845690101385117, "learning_rate": 2.520811558021241e-07, "loss": 0.0104, "step": 56670 }, { "epoch": 16.491125981961012, "grad_norm": 0.09186077117919922, "learning_rate": 2.505741030876052e-07, "loss": 0.0116, "step": 56680 }, { "epoch": 16.494035496072154, "grad_norm": 0.10755203664302826, "learning_rate": 2.4907153085924005e-07, "loss": 0.0122, "step": 56690 }, { "epoch": 16.4969450101833, "grad_norm": 0.09758348017930984, "learning_rate": 2.475734395734669e-07, "loss": 0.0121, "step": 56700 }, { "epoch": 16.49985452429444, "grad_norm": 0.09991848468780518, "learning_rate": 2.4607982968536835e-07, "loss": 0.0119, "step": 56710 }, { "epoch": 16.502764038405587, "grad_norm": 0.08150311559438705, "learning_rate": 2.445907016486632e-07, "loss": 0.0117, "step": 56720 }, { "epoch": 16.50567355251673, "grad_norm": 0.09117865562438965, "learning_rate": 2.4310605591570614e-07, "loss": 0.0135, "step": 56730 }, { "epoch": 16.508583066627875, "grad_norm": 0.09194885939359665, "learning_rate": 2.416258929374998e-07, "loss": 0.0108, "step": 56740 }, { "epoch": 16.511492580739016, "grad_norm": 0.09209076315164566, "learning_rate": 2.4015021316367124e-07, "loss": 0.0107, "step": 56750 }, { "epoch": 16.51440209485016, "grad_norm": 0.0765521451830864, "learning_rate": 2.386790170424985e-07, "loss": 0.0121, "step": 56760 }, { "epoch": 16.517311608961304, "grad_norm": 0.07615144550800323, "learning_rate": 2.3721230502089076e-07, "loss": 0.0125, "step": 56770 }, { "epoch": 16.520221123072446, "grad_norm": 0.08096850663423538, "learning_rate": 2.3575007754439492e-07, "loss": 0.0114, "step": 56780 }, { "epoch": 16.52313063718359, "grad_norm": 0.09325650334358215, "learning_rate": 2.3429233505719904e-07, "loss": 0.0122, "step": 56790 }, { "epoch": 16.526040151294733, "grad_norm": 0.09609717130661011, "learning_rate": 2.3283907800212389e-07, "loss": 0.0108, "step": 56800 }, { "epoch": 16.52894966540588, "grad_norm": 0.07418093830347061, "learning_rate": 2.3139030682063634e-07, "loss": 0.0118, "step": 56810 }, { "epoch": 16.53185917951702, "grad_norm": 0.12749215960502625, "learning_rate": 2.2994602195282766e-07, "loss": 0.0123, "step": 56820 }, { "epoch": 16.534768693628163, "grad_norm": 0.08720911294221878, "learning_rate": 2.285062238374386e-07, "loss": 0.0116, "step": 56830 }, { "epoch": 16.537678207739308, "grad_norm": 0.08768227696418762, "learning_rate": 2.2707091291183923e-07, "loss": 0.0113, "step": 56840 }, { "epoch": 16.54058772185045, "grad_norm": 0.07740234583616257, "learning_rate": 2.256400896120392e-07, "loss": 0.0127, "step": 56850 }, { "epoch": 16.543497235961595, "grad_norm": 0.12898260354995728, "learning_rate": 2.2421375437268744e-07, "loss": 0.012, "step": 56860 }, { "epoch": 16.546406750072737, "grad_norm": 0.12538069486618042, "learning_rate": 2.2279190762706403e-07, "loss": 0.0105, "step": 56870 }, { "epoch": 16.549316264183883, "grad_norm": 0.11477803438901901, "learning_rate": 2.213745498070885e-07, "loss": 0.0118, "step": 56880 }, { "epoch": 16.552225778295025, "grad_norm": 0.08564262092113495, "learning_rate": 2.1996168134331974e-07, "loss": 0.0103, "step": 56890 }, { "epoch": 16.555135292406167, "grad_norm": 0.12774260342121124, "learning_rate": 2.185533026649461e-07, "loss": 0.0121, "step": 56900 }, { "epoch": 16.558044806517312, "grad_norm": 0.08655951172113419, "learning_rate": 2.1714941419979705e-07, "loss": 0.0126, "step": 56910 }, { "epoch": 16.560954320628454, "grad_norm": 0.09699159115552902, "learning_rate": 2.1575001637433977e-07, "loss": 0.0122, "step": 56920 }, { "epoch": 16.5638638347396, "grad_norm": 0.07440335303544998, "learning_rate": 2.1435510961367088e-07, "loss": 0.011, "step": 56930 }, { "epoch": 16.56677334885074, "grad_norm": 0.1009388193488121, "learning_rate": 2.129646943415281e-07, "loss": 0.0101, "step": 56940 }, { "epoch": 16.569682862961887, "grad_norm": 0.07652313262224197, "learning_rate": 2.1157877098028023e-07, "loss": 0.0113, "step": 56950 }, { "epoch": 16.57259237707303, "grad_norm": 0.07747745513916016, "learning_rate": 2.1019733995093882e-07, "loss": 0.0115, "step": 56960 }, { "epoch": 16.57550189118417, "grad_norm": 0.0871395543217659, "learning_rate": 2.0882040167313987e-07, "loss": 0.012, "step": 56970 }, { "epoch": 16.578411405295316, "grad_norm": 0.09566191583871841, "learning_rate": 2.074479565651688e-07, "loss": 0.0121, "step": 56980 }, { "epoch": 16.58132091940646, "grad_norm": 0.08162857592105865, "learning_rate": 2.0608000504393043e-07, "loss": 0.0121, "step": 56990 }, { "epoch": 16.584230433517604, "grad_norm": 0.09969542175531387, "learning_rate": 2.0471654752497738e-07, "loss": 0.0126, "step": 57000 }, { "epoch": 16.587139947628746, "grad_norm": 0.12227724492549896, "learning_rate": 2.033575844224883e-07, "loss": 0.0111, "step": 57010 }, { "epoch": 16.590049461739888, "grad_norm": 0.07086550444364548, "learning_rate": 2.020031161492847e-07, "loss": 0.0115, "step": 57020 }, { "epoch": 16.592958975851033, "grad_norm": 0.09420347958803177, "learning_rate": 2.006531431168157e-07, "loss": 0.0114, "step": 57030 }, { "epoch": 16.595868489962175, "grad_norm": 0.12626144289970398, "learning_rate": 1.9930766573516669e-07, "loss": 0.0136, "step": 57040 }, { "epoch": 16.59877800407332, "grad_norm": 0.12191484123468399, "learning_rate": 1.9796668441305899e-07, "loss": 0.0115, "step": 57050 }, { "epoch": 16.601687518184463, "grad_norm": 0.08378668129444122, "learning_rate": 1.9663019955784845e-07, "loss": 0.0116, "step": 57060 }, { "epoch": 16.604597032295608, "grad_norm": 0.09971773624420166, "learning_rate": 1.9529821157552197e-07, "loss": 0.0123, "step": 57070 }, { "epoch": 16.60750654640675, "grad_norm": 0.08736748993396759, "learning_rate": 1.939707208707042e-07, "loss": 0.0128, "step": 57080 }, { "epoch": 16.610416060517892, "grad_norm": 0.09277314692735672, "learning_rate": 1.9264772784665097e-07, "loss": 0.0114, "step": 57090 }, { "epoch": 16.613325574629037, "grad_norm": 0.09437324851751328, "learning_rate": 1.9132923290525074e-07, "loss": 0.0115, "step": 57100 }, { "epoch": 16.61623508874018, "grad_norm": 0.08818966895341873, "learning_rate": 1.9001523644703144e-07, "loss": 0.0114, "step": 57110 }, { "epoch": 16.619144602851325, "grad_norm": 0.12252809852361679, "learning_rate": 1.8870573887114718e-07, "loss": 0.0118, "step": 57120 }, { "epoch": 16.622054116962467, "grad_norm": 0.06247778236865997, "learning_rate": 1.8740074057539137e-07, "loss": 0.0117, "step": 57130 }, { "epoch": 16.624963631073612, "grad_norm": 0.07814272493124008, "learning_rate": 1.8610024195618358e-07, "loss": 0.0118, "step": 57140 }, { "epoch": 16.627873145184754, "grad_norm": 0.09633014351129532, "learning_rate": 1.8480424340858614e-07, "loss": 0.0119, "step": 57150 }, { "epoch": 16.630782659295896, "grad_norm": 0.08973978459835052, "learning_rate": 1.8351274532628748e-07, "loss": 0.011, "step": 57160 }, { "epoch": 16.63369217340704, "grad_norm": 0.09032611548900604, "learning_rate": 1.8222574810160874e-07, "loss": 0.0121, "step": 57170 }, { "epoch": 16.636601687518183, "grad_norm": 0.06631235033273697, "learning_rate": 1.8094325212550722e-07, "loss": 0.0108, "step": 57180 }, { "epoch": 16.63951120162933, "grad_norm": 0.08685899525880814, "learning_rate": 1.7966525778757292e-07, "loss": 0.0109, "step": 57190 }, { "epoch": 16.64242071574047, "grad_norm": 0.0871051475405693, "learning_rate": 1.7839176547602198e-07, "loss": 0.0114, "step": 57200 }, { "epoch": 16.645330229851616, "grad_norm": 0.09152296185493469, "learning_rate": 1.7712277557771328e-07, "loss": 0.0117, "step": 57210 }, { "epoch": 16.64823974396276, "grad_norm": 0.08220025897026062, "learning_rate": 1.7585828847812847e-07, "loss": 0.0129, "step": 57220 }, { "epoch": 16.6511492580739, "grad_norm": 0.10567856580018997, "learning_rate": 1.7459830456138692e-07, "loss": 0.0113, "step": 57230 }, { "epoch": 16.654058772185046, "grad_norm": 0.08110907673835754, "learning_rate": 1.7334282421023918e-07, "loss": 0.0129, "step": 57240 }, { "epoch": 16.656968286296188, "grad_norm": 0.08398941904306412, "learning_rate": 1.7209184780606679e-07, "loss": 0.0126, "step": 57250 }, { "epoch": 16.659877800407333, "grad_norm": 0.07635539770126343, "learning_rate": 1.7084537572888247e-07, "loss": 0.0123, "step": 57260 }, { "epoch": 16.662787314518475, "grad_norm": 0.0801951065659523, "learning_rate": 1.6960340835733002e-07, "loss": 0.0114, "step": 57270 }, { "epoch": 16.66569682862962, "grad_norm": 0.11388111114501953, "learning_rate": 1.6836594606869093e-07, "loss": 0.012, "step": 57280 }, { "epoch": 16.668606342740762, "grad_norm": 0.10472650080919266, "learning_rate": 1.6713298923886788e-07, "loss": 0.0118, "step": 57290 }, { "epoch": 16.671515856851904, "grad_norm": 0.0876389816403389, "learning_rate": 1.6590453824240627e-07, "loss": 0.0122, "step": 57300 }, { "epoch": 16.67442537096305, "grad_norm": 0.2044486403465271, "learning_rate": 1.6468059345247256e-07, "loss": 0.0125, "step": 57310 }, { "epoch": 16.677334885074192, "grad_norm": 0.07050540298223495, "learning_rate": 1.6346115524087101e-07, "loss": 0.0111, "step": 57320 }, { "epoch": 16.680244399185337, "grad_norm": 0.10893665254116058, "learning_rate": 1.6224622397803535e-07, "loss": 0.0124, "step": 57330 }, { "epoch": 16.68315391329648, "grad_norm": 0.08091922104358673, "learning_rate": 1.6103580003302698e-07, "loss": 0.0128, "step": 57340 }, { "epoch": 16.686063427407625, "grad_norm": 0.10488428920507431, "learning_rate": 1.5982988377354346e-07, "loss": 0.0125, "step": 57350 }, { "epoch": 16.688972941518767, "grad_norm": 0.08712555468082428, "learning_rate": 1.5862847556590841e-07, "loss": 0.0134, "step": 57360 }, { "epoch": 16.69188245562991, "grad_norm": 0.1030728667974472, "learning_rate": 1.5743157577507993e-07, "loss": 0.0104, "step": 57370 }, { "epoch": 16.694791969741054, "grad_norm": 0.1197739765048027, "learning_rate": 1.5623918476464383e-07, "loss": 0.0121, "step": 57380 }, { "epoch": 16.697701483852196, "grad_norm": 0.13041163980960846, "learning_rate": 1.5505130289681534e-07, "loss": 0.013, "step": 57390 }, { "epoch": 16.70061099796334, "grad_norm": 0.08400384336709976, "learning_rate": 1.538679305324442e-07, "loss": 0.0131, "step": 57400 }, { "epoch": 16.703520512074483, "grad_norm": 0.0811697393655777, "learning_rate": 1.5268906803100446e-07, "loss": 0.0122, "step": 57410 }, { "epoch": 16.706430026185625, "grad_norm": 0.07413394004106522, "learning_rate": 1.515147157506064e-07, "loss": 0.0112, "step": 57420 }, { "epoch": 16.70933954029677, "grad_norm": 0.09754657000303268, "learning_rate": 1.5034487404798625e-07, "loss": 0.0119, "step": 57430 }, { "epoch": 16.712249054407913, "grad_norm": 0.09012221544981003, "learning_rate": 1.491795432785098e-07, "loss": 0.0114, "step": 57440 }, { "epoch": 16.71515856851906, "grad_norm": 0.09117493778467178, "learning_rate": 1.480187237961772e-07, "loss": 0.012, "step": 57450 }, { "epoch": 16.7180680826302, "grad_norm": 0.12530571222305298, "learning_rate": 1.468624159536097e-07, "loss": 0.012, "step": 57460 }, { "epoch": 16.720977596741346, "grad_norm": 0.0690959244966507, "learning_rate": 1.4571062010206802e-07, "loss": 0.0102, "step": 57470 }, { "epoch": 16.723887110852488, "grad_norm": 0.10908710956573486, "learning_rate": 1.4456333659143395e-07, "loss": 0.0123, "step": 57480 }, { "epoch": 16.72679662496363, "grad_norm": 0.10805575549602509, "learning_rate": 1.4342056577022366e-07, "loss": 0.0112, "step": 57490 }, { "epoch": 16.729706139074775, "grad_norm": 0.09519830346107483, "learning_rate": 1.4228230798558118e-07, "loss": 0.0112, "step": 57500 }, { "epoch": 16.732615653185917, "grad_norm": 0.09441579133272171, "learning_rate": 1.4114856358327654e-07, "loss": 0.0121, "step": 57510 }, { "epoch": 16.735525167297062, "grad_norm": 0.07706833630800247, "learning_rate": 1.4001933290771263e-07, "loss": 0.0108, "step": 57520 }, { "epoch": 16.738434681408204, "grad_norm": 0.08815540373325348, "learning_rate": 1.3889461630192168e-07, "loss": 0.011, "step": 57530 }, { "epoch": 16.74134419551935, "grad_norm": 0.09980346262454987, "learning_rate": 1.3777441410756209e-07, "loss": 0.012, "step": 57540 }, { "epoch": 16.74425370963049, "grad_norm": 0.10831983387470245, "learning_rate": 1.3665872666491996e-07, "loss": 0.0113, "step": 57550 }, { "epoch": 16.747163223741634, "grad_norm": 0.10041162371635437, "learning_rate": 1.3554755431291254e-07, "loss": 0.013, "step": 57560 }, { "epoch": 16.75007273785278, "grad_norm": 0.07258038967847824, "learning_rate": 1.3444089738908483e-07, "loss": 0.0101, "step": 57570 }, { "epoch": 16.75298225196392, "grad_norm": 0.20413535833358765, "learning_rate": 1.3333875622961289e-07, "loss": 0.0124, "step": 57580 }, { "epoch": 16.755891766075067, "grad_norm": 0.09103474766016006, "learning_rate": 1.3224113116929226e-07, "loss": 0.0107, "step": 57590 }, { "epoch": 16.75880128018621, "grad_norm": 0.09720046818256378, "learning_rate": 1.3114802254155788e-07, "loss": 0.0109, "step": 57600 }, { "epoch": 16.761710794297354, "grad_norm": 0.09076264500617981, "learning_rate": 1.300594306784625e-07, "loss": 0.0118, "step": 57610 }, { "epoch": 16.764620308408496, "grad_norm": 0.09491293877363205, "learning_rate": 1.2897535591069654e-07, "loss": 0.0106, "step": 57620 }, { "epoch": 16.767529822519638, "grad_norm": 0.1404021680355072, "learning_rate": 1.2789579856756994e-07, "loss": 0.0109, "step": 57630 }, { "epoch": 16.770439336630783, "grad_norm": 0.09464658796787262, "learning_rate": 1.2682075897702372e-07, "loss": 0.0123, "step": 57640 }, { "epoch": 16.773348850741925, "grad_norm": 0.07822471112012863, "learning_rate": 1.2575023746562832e-07, "loss": 0.0123, "step": 57650 }, { "epoch": 16.77625836485307, "grad_norm": 0.07560540735721588, "learning_rate": 1.246842343585769e-07, "loss": 0.0109, "step": 57660 }, { "epoch": 16.779167878964213, "grad_norm": 0.07772955298423767, "learning_rate": 1.236227499796988e-07, "loss": 0.0107, "step": 57670 }, { "epoch": 16.782077393075355, "grad_norm": 0.11264344304800034, "learning_rate": 1.2256578465143774e-07, "loss": 0.0118, "step": 57680 }, { "epoch": 16.7849869071865, "grad_norm": 0.08591131865978241, "learning_rate": 1.2151333869487692e-07, "loss": 0.0131, "step": 57690 }, { "epoch": 16.787896421297642, "grad_norm": 0.09075089544057846, "learning_rate": 1.2046541242972054e-07, "loss": 0.0119, "step": 57700 }, { "epoch": 16.790805935408788, "grad_norm": 0.08231791108846664, "learning_rate": 1.1942200617430066e-07, "loss": 0.0121, "step": 57710 }, { "epoch": 16.79371544951993, "grad_norm": 0.11309477686882019, "learning_rate": 1.183831202455754e-07, "loss": 0.0127, "step": 57720 }, { "epoch": 16.796624963631075, "grad_norm": 0.08068186044692993, "learning_rate": 1.1734875495913233e-07, "loss": 0.0119, "step": 57730 }, { "epoch": 16.799534477742217, "grad_norm": 0.09901289641857147, "learning_rate": 1.1631891062918343e-07, "loss": 0.0122, "step": 57740 }, { "epoch": 16.80244399185336, "grad_norm": 0.08791325986385345, "learning_rate": 1.1529358756857011e-07, "loss": 0.0117, "step": 57750 }, { "epoch": 16.805353505964504, "grad_norm": 0.0685417577624321, "learning_rate": 1.1427278608875492e-07, "loss": 0.0112, "step": 57760 }, { "epoch": 16.808263020075646, "grad_norm": 0.1079944595694542, "learning_rate": 1.132565064998331e-07, "loss": 0.0135, "step": 57770 }, { "epoch": 16.81117253418679, "grad_norm": 0.0736834928393364, "learning_rate": 1.1224474911052274e-07, "loss": 0.0115, "step": 57780 }, { "epoch": 16.814082048297934, "grad_norm": 0.08989255130290985, "learning_rate": 1.112375142281713e-07, "loss": 0.0113, "step": 57790 }, { "epoch": 16.81699156240908, "grad_norm": 0.10893921554088593, "learning_rate": 1.102348021587457e-07, "loss": 0.0121, "step": 57800 }, { "epoch": 16.81990107652022, "grad_norm": 0.16236691176891327, "learning_rate": 1.0923661320684563e-07, "loss": 0.0118, "step": 57810 }, { "epoch": 16.822810590631363, "grad_norm": 0.08983712643384933, "learning_rate": 1.0824294767569687e-07, "loss": 0.0108, "step": 57820 }, { "epoch": 16.82572010474251, "grad_norm": 0.09115497022867203, "learning_rate": 1.0725380586714462e-07, "loss": 0.0113, "step": 57830 }, { "epoch": 16.82862961885365, "grad_norm": 0.07637634128332138, "learning_rate": 1.0626918808166686e-07, "loss": 0.0111, "step": 57840 }, { "epoch": 16.831539132964796, "grad_norm": 0.0886441171169281, "learning_rate": 1.0528909461836267e-07, "loss": 0.0115, "step": 57850 }, { "epoch": 16.834448647075938, "grad_norm": 0.09083319455385208, "learning_rate": 1.0431352577496056e-07, "loss": 0.0112, "step": 57860 }, { "epoch": 16.837358161187083, "grad_norm": 0.12638111412525177, "learning_rate": 1.0334248184781015e-07, "loss": 0.011, "step": 57870 }, { "epoch": 16.840267675298225, "grad_norm": 0.10746116191148758, "learning_rate": 1.0237596313189212e-07, "loss": 0.0124, "step": 57880 }, { "epoch": 16.843177189409367, "grad_norm": 0.10293295234441757, "learning_rate": 1.0141396992080498e-07, "loss": 0.0119, "step": 57890 }, { "epoch": 16.846086703520513, "grad_norm": 0.09717271476984024, "learning_rate": 1.0045650250678162e-07, "loss": 0.0108, "step": 57900 }, { "epoch": 16.848996217631655, "grad_norm": 0.0925435721874237, "learning_rate": 9.950356118067272e-08, "loss": 0.0136, "step": 57910 }, { "epoch": 16.8519057317428, "grad_norm": 0.11182243376970291, "learning_rate": 9.855514623195672e-08, "loss": 0.0113, "step": 57920 }, { "epoch": 16.854815245853942, "grad_norm": 0.08650000393390656, "learning_rate": 9.761125794873816e-08, "loss": 0.0106, "step": 57930 }, { "epoch": 16.857724759965087, "grad_norm": 0.09730862081050873, "learning_rate": 9.667189661774433e-08, "loss": 0.012, "step": 57940 }, { "epoch": 16.86063427407623, "grad_norm": 0.09848964214324951, "learning_rate": 9.573706252432867e-08, "loss": 0.0118, "step": 57950 }, { "epoch": 16.86354378818737, "grad_norm": 0.0617717020213604, "learning_rate": 9.480675595246735e-08, "loss": 0.0118, "step": 57960 }, { "epoch": 16.866453302298517, "grad_norm": 0.0936826840043068, "learning_rate": 9.388097718476763e-08, "loss": 0.012, "step": 57970 }, { "epoch": 16.86936281640966, "grad_norm": 0.07048794627189636, "learning_rate": 9.295972650245121e-08, "loss": 0.0101, "step": 57980 }, { "epoch": 16.872272330520804, "grad_norm": 0.07423155754804611, "learning_rate": 9.204300418537426e-08, "loss": 0.0104, "step": 57990 }, { "epoch": 16.875181844631946, "grad_norm": 0.1175612285733223, "learning_rate": 9.113081051201066e-08, "loss": 0.0116, "step": 58000 }, { "epoch": 16.87809135874309, "grad_norm": 0.09492985904216766, "learning_rate": 9.022314575946044e-08, "loss": 0.0115, "step": 58010 }, { "epoch": 16.881000872854234, "grad_norm": 0.07639354467391968, "learning_rate": 8.932001020344804e-08, "loss": 0.0123, "step": 58020 }, { "epoch": 16.883910386965375, "grad_norm": 0.11978814005851746, "learning_rate": 8.8421404118324e-08, "loss": 0.011, "step": 58030 }, { "epoch": 16.88681990107652, "grad_norm": 0.07644479721784592, "learning_rate": 8.75273277770583e-08, "loss": 0.0109, "step": 58040 }, { "epoch": 16.889729415187663, "grad_norm": 0.10150235146284103, "learning_rate": 8.663778145125201e-08, "loss": 0.0122, "step": 58050 }, { "epoch": 16.89263892929881, "grad_norm": 0.09054337441921234, "learning_rate": 8.575276541112064e-08, "loss": 0.012, "step": 58060 }, { "epoch": 16.89554844340995, "grad_norm": 0.08837748318910599, "learning_rate": 8.48722799255125e-08, "loss": 0.0118, "step": 58070 }, { "epoch": 16.898457957521096, "grad_norm": 0.1477501094341278, "learning_rate": 8.399632526189361e-08, "loss": 0.0123, "step": 58080 }, { "epoch": 16.901367471632238, "grad_norm": 0.12393353134393692, "learning_rate": 8.312490168635778e-08, "loss": 0.0121, "step": 58090 }, { "epoch": 16.90427698574338, "grad_norm": 0.08028803020715714, "learning_rate": 8.225800946361994e-08, "loss": 0.0124, "step": 58100 }, { "epoch": 16.907186499854525, "grad_norm": 0.08137380331754684, "learning_rate": 8.139564885701777e-08, "loss": 0.0128, "step": 58110 }, { "epoch": 16.910096013965667, "grad_norm": 0.10978769510984421, "learning_rate": 8.053782012851507e-08, "loss": 0.013, "step": 58120 }, { "epoch": 16.913005528076813, "grad_norm": 0.092011958360672, "learning_rate": 7.968452353869671e-08, "loss": 0.0109, "step": 58130 }, { "epoch": 16.915915042187954, "grad_norm": 0.14292454719543457, "learning_rate": 7.883575934677201e-08, "loss": 0.0096, "step": 58140 }, { "epoch": 16.918824556299096, "grad_norm": 0.06222616881132126, "learning_rate": 7.799152781057306e-08, "loss": 0.0102, "step": 58150 }, { "epoch": 16.921734070410242, "grad_norm": 0.10608738660812378, "learning_rate": 7.715182918655639e-08, "loss": 0.0105, "step": 58160 }, { "epoch": 16.924643584521384, "grad_norm": 0.11007139086723328, "learning_rate": 7.631666372979795e-08, "loss": 0.0114, "step": 58170 }, { "epoch": 16.92755309863253, "grad_norm": 0.0966365709900856, "learning_rate": 7.548603169399981e-08, "loss": 0.0112, "step": 58180 }, { "epoch": 16.93046261274367, "grad_norm": 0.09137297421693802, "learning_rate": 7.46599333314868e-08, "loss": 0.0124, "step": 58190 }, { "epoch": 16.933372126854817, "grad_norm": 0.1170419231057167, "learning_rate": 7.383836889320483e-08, "loss": 0.0128, "step": 58200 }, { "epoch": 16.93628164096596, "grad_norm": 0.1408759355545044, "learning_rate": 7.302133862872262e-08, "loss": 0.011, "step": 58210 }, { "epoch": 16.9391911550771, "grad_norm": 0.07299771904945374, "learning_rate": 7.220884278623496e-08, "loss": 0.0104, "step": 58220 }, { "epoch": 16.942100669188246, "grad_norm": 0.09400641173124313, "learning_rate": 7.140088161255609e-08, "loss": 0.0118, "step": 58230 }, { "epoch": 16.945010183299388, "grad_norm": 0.06580052524805069, "learning_rate": 7.059745535312301e-08, "loss": 0.0135, "step": 58240 }, { "epoch": 16.947919697410534, "grad_norm": 0.08021633327007294, "learning_rate": 6.979856425199382e-08, "loss": 0.0113, "step": 58250 }, { "epoch": 16.950829211521675, "grad_norm": 0.076602503657341, "learning_rate": 6.900420855185108e-08, "loss": 0.0119, "step": 58260 }, { "epoch": 16.95373872563282, "grad_norm": 0.08591817319393158, "learning_rate": 6.821438849400175e-08, "loss": 0.0113, "step": 58270 }, { "epoch": 16.956648239743963, "grad_norm": 0.07974372804164886, "learning_rate": 6.742910431836724e-08, "loss": 0.0114, "step": 58280 }, { "epoch": 16.959557753855105, "grad_norm": 0.08672792464494705, "learning_rate": 6.664835626350174e-08, "loss": 0.0109, "step": 58290 }, { "epoch": 16.96246726796625, "grad_norm": 0.10660488158464432, "learning_rate": 6.58721445665722e-08, "loss": 0.0127, "step": 58300 }, { "epoch": 16.965376782077392, "grad_norm": 0.083563432097435, "learning_rate": 6.510046946337166e-08, "loss": 0.0117, "step": 58310 }, { "epoch": 16.968286296188538, "grad_norm": 0.07314389944076538, "learning_rate": 6.433333118831764e-08, "loss": 0.0104, "step": 58320 }, { "epoch": 16.97119581029968, "grad_norm": 0.07172711938619614, "learning_rate": 6.357072997444202e-08, "loss": 0.01, "step": 58330 }, { "epoch": 16.974105324410825, "grad_norm": 0.06444241851568222, "learning_rate": 6.281266605340618e-08, "loss": 0.0109, "step": 58340 }, { "epoch": 16.977014838521967, "grad_norm": 0.10158644616603851, "learning_rate": 6.205913965548759e-08, "loss": 0.0118, "step": 58350 }, { "epoch": 16.97992435263311, "grad_norm": 0.11392220854759216, "learning_rate": 6.13101510095898e-08, "loss": 0.0138, "step": 58360 }, { "epoch": 16.982833866744254, "grad_norm": 0.0950474664568901, "learning_rate": 6.056570034323416e-08, "loss": 0.0099, "step": 58370 }, { "epoch": 16.985743380855396, "grad_norm": 0.07259244471788406, "learning_rate": 5.982578788256643e-08, "loss": 0.011, "step": 58380 }, { "epoch": 16.988652894966542, "grad_norm": 0.09210839867591858, "learning_rate": 5.9090413852353477e-08, "loss": 0.0115, "step": 58390 }, { "epoch": 16.991562409077684, "grad_norm": 0.11275658756494522, "learning_rate": 5.835957847597828e-08, "loss": 0.0102, "step": 58400 }, { "epoch": 16.994471923188826, "grad_norm": 0.0835161805152893, "learning_rate": 5.763328197545326e-08, "loss": 0.012, "step": 58410 }, { "epoch": 16.99738143729997, "grad_norm": 0.07419688999652863, "learning_rate": 5.691152457140691e-08, "loss": 0.0115, "step": 58420 }, { "epoch": 17.000290951411113, "grad_norm": 0.09703316539525986, "learning_rate": 5.6194306483088875e-08, "loss": 0.0101, "step": 58430 }, { "epoch": 17.00320046552226, "grad_norm": 0.08982633799314499, "learning_rate": 5.5481627928374855e-08, "loss": 0.0116, "step": 58440 }, { "epoch": 17.0061099796334, "grad_norm": 0.06105365604162216, "learning_rate": 5.477348912375168e-08, "loss": 0.0112, "step": 58450 }, { "epoch": 17.009019493744546, "grad_norm": 0.06042547523975372, "learning_rate": 5.406989028433895e-08, "loss": 0.0109, "step": 58460 }, { "epoch": 17.011929007855688, "grad_norm": 0.11863560974597931, "learning_rate": 5.337083162386902e-08, "loss": 0.0116, "step": 58470 }, { "epoch": 17.01483852196683, "grad_norm": 0.07765530049800873, "learning_rate": 5.2676313354697024e-08, "loss": 0.0105, "step": 58480 }, { "epoch": 17.017748036077975, "grad_norm": 0.1315426081418991, "learning_rate": 5.19863356877992e-08, "loss": 0.0114, "step": 58490 }, { "epoch": 17.020657550189117, "grad_norm": 0.06600496172904968, "learning_rate": 5.1300898832772894e-08, "loss": 0.0107, "step": 58500 }, { "epoch": 17.023567064300263, "grad_norm": 0.09444813430309296, "learning_rate": 5.06200029978382e-08, "loss": 0.0099, "step": 58510 }, { "epoch": 17.026476578411405, "grad_norm": 0.08919280767440796, "learning_rate": 4.9943648389829676e-08, "loss": 0.0124, "step": 58520 }, { "epoch": 17.02938609252255, "grad_norm": 0.10408365726470947, "learning_rate": 4.9271835214206304e-08, "loss": 0.0122, "step": 58530 }, { "epoch": 17.032295606633692, "grad_norm": 0.09090720117092133, "learning_rate": 4.860456367504984e-08, "loss": 0.0128, "step": 58540 }, { "epoch": 17.035205120744834, "grad_norm": 0.10328070819377899, "learning_rate": 4.7941833975058136e-08, "loss": 0.0104, "step": 58550 }, { "epoch": 17.03811463485598, "grad_norm": 0.09483779221773148, "learning_rate": 4.7283646315550155e-08, "loss": 0.0101, "step": 58560 }, { "epoch": 17.04102414896712, "grad_norm": 0.09738648682832718, "learning_rate": 4.663000089646763e-08, "loss": 0.0121, "step": 58570 }, { "epoch": 17.043933663078267, "grad_norm": 0.09409596025943756, "learning_rate": 4.5980897916370055e-08, "loss": 0.0128, "step": 58580 }, { "epoch": 17.04684317718941, "grad_norm": 0.12677986919879913, "learning_rate": 4.533633757243805e-08, "loss": 0.0107, "step": 58590 }, { "epoch": 17.049752691300554, "grad_norm": 0.1006055548787117, "learning_rate": 4.4696320060469975e-08, "loss": 0.0117, "step": 58600 }, { "epoch": 17.052662205411696, "grad_norm": 0.06669533252716064, "learning_rate": 4.406084557488865e-08, "loss": 0.0121, "step": 58610 }, { "epoch": 17.05557171952284, "grad_norm": 0.08248165994882584, "learning_rate": 4.3429914308734667e-08, "loss": 0.0104, "step": 58620 }, { "epoch": 17.058481233633984, "grad_norm": 0.07742199301719666, "learning_rate": 4.2803526453666365e-08, "loss": 0.0108, "step": 58630 }, { "epoch": 17.061390747745126, "grad_norm": 0.0900779440999031, "learning_rate": 4.2181682199964876e-08, "loss": 0.0108, "step": 58640 }, { "epoch": 17.06430026185627, "grad_norm": 0.09490861743688583, "learning_rate": 4.156438173653076e-08, "loss": 0.012, "step": 58650 }, { "epoch": 17.067209775967413, "grad_norm": 0.09158962219953537, "learning_rate": 4.095162525088236e-08, "loss": 0.0126, "step": 58660 }, { "epoch": 17.07011929007856, "grad_norm": 0.09107572585344315, "learning_rate": 4.034341292915911e-08, "loss": 0.0127, "step": 58670 }, { "epoch": 17.0730288041897, "grad_norm": 0.10330440104007721, "learning_rate": 3.9739744956121535e-08, "loss": 0.0117, "step": 58680 }, { "epoch": 17.075938318300842, "grad_norm": 0.08640255779027939, "learning_rate": 3.9140621515144634e-08, "loss": 0.0124, "step": 58690 }, { "epoch": 17.078847832411988, "grad_norm": 0.09853096306324005, "learning_rate": 3.8546042788229485e-08, "loss": 0.0113, "step": 58700 }, { "epoch": 17.08175734652313, "grad_norm": 0.06404776871204376, "learning_rate": 3.795600895599327e-08, "loss": 0.0117, "step": 58710 }, { "epoch": 17.084666860634275, "grad_norm": 0.08064322173595428, "learning_rate": 3.737052019766929e-08, "loss": 0.0106, "step": 58720 }, { "epoch": 17.087576374745417, "grad_norm": 0.09971122443675995, "learning_rate": 3.678957669111693e-08, "loss": 0.0117, "step": 58730 }, { "epoch": 17.090485888856563, "grad_norm": 0.09233132749795914, "learning_rate": 3.6213178612811706e-08, "loss": 0.0118, "step": 58740 }, { "epoch": 17.093395402967705, "grad_norm": 0.1134512796998024, "learning_rate": 3.564132613784521e-08, "loss": 0.0104, "step": 58750 }, { "epoch": 17.096304917078847, "grad_norm": 0.08406981825828552, "learning_rate": 3.507401943993349e-08, "loss": 0.0104, "step": 58760 }, { "epoch": 17.099214431189992, "grad_norm": 0.12983113527297974, "learning_rate": 3.4511258691408696e-08, "loss": 0.0111, "step": 58770 }, { "epoch": 17.102123945301134, "grad_norm": 0.13918152451515198, "learning_rate": 3.3953044063222414e-08, "loss": 0.0125, "step": 58780 }, { "epoch": 17.10503345941228, "grad_norm": 0.07980415225028992, "learning_rate": 3.339937572494733e-08, "loss": 0.0112, "step": 58790 }, { "epoch": 17.10794297352342, "grad_norm": 0.08414256572723389, "learning_rate": 3.285025384476892e-08, "loss": 0.0124, "step": 58800 }, { "epoch": 17.110852487634563, "grad_norm": 0.0879119485616684, "learning_rate": 3.2305678589502064e-08, "loss": 0.0123, "step": 58810 }, { "epoch": 17.11376200174571, "grad_norm": 0.08131054788827896, "learning_rate": 3.176565012456778e-08, "loss": 0.0106, "step": 58820 }, { "epoch": 17.11667151585685, "grad_norm": 0.08307932317256927, "learning_rate": 3.123016861401651e-08, "loss": 0.0121, "step": 58830 }, { "epoch": 17.119581029967996, "grad_norm": 0.079975925385952, "learning_rate": 3.069923422051313e-08, "loss": 0.0127, "step": 58840 }, { "epoch": 17.122490544079138, "grad_norm": 0.10004517436027527, "learning_rate": 3.017284710534196e-08, "loss": 0.0131, "step": 58850 }, { "epoch": 17.125400058190284, "grad_norm": 0.09118286520242691, "learning_rate": 2.965100742840343e-08, "loss": 0.0104, "step": 58860 }, { "epoch": 17.128309572301426, "grad_norm": 0.09919165819883347, "learning_rate": 2.913371534821907e-08, "loss": 0.0119, "step": 58870 }, { "epoch": 17.131219086412568, "grad_norm": 0.07629436254501343, "learning_rate": 2.862097102192984e-08, "loss": 0.0106, "step": 58880 }, { "epoch": 17.134128600523713, "grad_norm": 0.09898444265127182, "learning_rate": 2.8112774605292824e-08, "loss": 0.0114, "step": 58890 }, { "epoch": 17.137038114634855, "grad_norm": 0.09603488445281982, "learning_rate": 2.7609126252684526e-08, "loss": 0.0117, "step": 58900 }, { "epoch": 17.139947628746, "grad_norm": 0.08036711066961288, "learning_rate": 2.7110026117100893e-08, "loss": 0.012, "step": 58910 }, { "epoch": 17.142857142857142, "grad_norm": 0.33503100275993347, "learning_rate": 2.6615474350155654e-08, "loss": 0.0106, "step": 58920 }, { "epoch": 17.145766656968288, "grad_norm": 0.11598663032054901, "learning_rate": 2.61254711020803e-08, "loss": 0.0103, "step": 58930 }, { "epoch": 17.14867617107943, "grad_norm": 0.0843466967344284, "learning_rate": 2.5640016521724096e-08, "loss": 0.0114, "step": 58940 }, { "epoch": 17.15158568519057, "grad_norm": 0.14333687722682953, "learning_rate": 2.515911075655575e-08, "loss": 0.0127, "step": 58950 }, { "epoch": 17.154495199301717, "grad_norm": 0.06660350412130356, "learning_rate": 2.468275395266173e-08, "loss": 0.0116, "step": 58960 }, { "epoch": 17.15740471341286, "grad_norm": 0.07096102833747864, "learning_rate": 2.4210946254747955e-08, "loss": 0.0121, "step": 58970 }, { "epoch": 17.160314227524005, "grad_norm": 0.11717908084392548, "learning_rate": 2.3743687806134785e-08, "loss": 0.0119, "step": 58980 }, { "epoch": 17.163223741635147, "grad_norm": 0.15276968479156494, "learning_rate": 2.328097874876367e-08, "loss": 0.0111, "step": 58990 }, { "epoch": 17.166133255746292, "grad_norm": 0.1464248150587082, "learning_rate": 2.2822819223193847e-08, "loss": 0.0113, "step": 59000 }, { "epoch": 17.169042769857434, "grad_norm": 0.08280374854803085, "learning_rate": 2.2369209368602316e-08, "loss": 0.013, "step": 59010 }, { "epoch": 17.171952283968576, "grad_norm": 0.07994931191205978, "learning_rate": 2.1920149322783856e-08, "loss": 0.0126, "step": 59020 }, { "epoch": 17.17486179807972, "grad_norm": 0.12122359126806259, "learning_rate": 2.147563922215101e-08, "loss": 0.0119, "step": 59030 }, { "epoch": 17.177771312190863, "grad_norm": 0.0976220890879631, "learning_rate": 2.1035679201734102e-08, "loss": 0.0131, "step": 59040 }, { "epoch": 17.18068082630201, "grad_norm": 0.09036850184202194, "learning_rate": 2.0600269395181225e-08, "loss": 0.0107, "step": 59050 }, { "epoch": 17.18359034041315, "grad_norm": 0.10311731696128845, "learning_rate": 2.0169409934756578e-08, "loss": 0.0111, "step": 59060 }, { "epoch": 17.186499854524296, "grad_norm": 0.08036497980356216, "learning_rate": 1.9743100951347125e-08, "loss": 0.011, "step": 59070 }, { "epoch": 17.189409368635438, "grad_norm": 0.09244845807552338, "learning_rate": 1.9321342574452618e-08, "loss": 0.0112, "step": 59080 }, { "epoch": 17.19231888274658, "grad_norm": 0.09105968475341797, "learning_rate": 1.890413493219223e-08, "loss": 0.0106, "step": 59090 }, { "epoch": 17.195228396857726, "grad_norm": 0.13146720826625824, "learning_rate": 1.849147815130292e-08, "loss": 0.0123, "step": 59100 }, { "epoch": 17.198137910968867, "grad_norm": 0.07833882421255112, "learning_rate": 1.8083372357139415e-08, "loss": 0.0137, "step": 59110 }, { "epoch": 17.201047425080013, "grad_norm": 0.10520797222852707, "learning_rate": 1.767981767367255e-08, "loss": 0.0123, "step": 59120 }, { "epoch": 17.203956939191155, "grad_norm": 0.10047351568937302, "learning_rate": 1.7280814223490925e-08, "loss": 0.0109, "step": 59130 }, { "epoch": 17.206866453302297, "grad_norm": 0.08287650346755981, "learning_rate": 1.688636212780259e-08, "loss": 0.0126, "step": 59140 }, { "epoch": 17.209775967413442, "grad_norm": 0.09772749990224838, "learning_rate": 1.6496461506431694e-08, "loss": 0.0114, "step": 59150 }, { "epoch": 17.212685481524584, "grad_norm": 0.08734921365976334, "learning_rate": 1.611111247781849e-08, "loss": 0.0107, "step": 59160 }, { "epoch": 17.21559499563573, "grad_norm": 0.0858645886182785, "learning_rate": 1.573031515902268e-08, "loss": 0.0118, "step": 59170 }, { "epoch": 17.21850450974687, "grad_norm": 0.0770098939538002, "learning_rate": 1.5354069665721725e-08, "loss": 0.0103, "step": 59180 }, { "epoch": 17.221414023858017, "grad_norm": 0.1028493270277977, "learning_rate": 1.4982376112205874e-08, "loss": 0.0129, "step": 59190 }, { "epoch": 17.22432353796916, "grad_norm": 0.08643429726362228, "learning_rate": 1.461523461138814e-08, "loss": 0.0112, "step": 59200 }, { "epoch": 17.2272330520803, "grad_norm": 0.07401925325393677, "learning_rate": 1.4252645274797638e-08, "loss": 0.0116, "step": 59210 }, { "epoch": 17.230142566191446, "grad_norm": 0.07392755895853043, "learning_rate": 1.3894608212576266e-08, "loss": 0.0125, "step": 59220 }, { "epoch": 17.23305208030259, "grad_norm": 0.08685875684022903, "learning_rate": 1.354112353348702e-08, "loss": 0.0106, "step": 59230 }, { "epoch": 17.235961594413734, "grad_norm": 0.11242158710956573, "learning_rate": 1.3192191344910676e-08, "loss": 0.0117, "step": 59240 }, { "epoch": 17.238871108524876, "grad_norm": 0.09717456996440887, "learning_rate": 1.2847811752844107e-08, "loss": 0.0115, "step": 59250 }, { "epoch": 17.24178062263602, "grad_norm": 0.06377477943897247, "learning_rate": 1.2507984861898635e-08, "loss": 0.0119, "step": 59260 }, { "epoch": 17.244690136747163, "grad_norm": 0.08327606320381165, "learning_rate": 1.2172710775305019e-08, "loss": 0.0107, "step": 59270 }, { "epoch": 17.247599650858305, "grad_norm": 0.08109855651855469, "learning_rate": 1.1841989594911785e-08, "loss": 0.0108, "step": 59280 }, { "epoch": 17.25050916496945, "grad_norm": 0.07993289083242416, "learning_rate": 1.1515821421181905e-08, "loss": 0.0115, "step": 59290 }, { "epoch": 17.253418679080593, "grad_norm": 0.08290064334869385, "learning_rate": 1.1194206353199455e-08, "loss": 0.0122, "step": 59300 }, { "epoch": 17.256328193191738, "grad_norm": 0.07253984361886978, "learning_rate": 1.0877144488657953e-08, "loss": 0.0114, "step": 59310 }, { "epoch": 17.25923770730288, "grad_norm": 0.10793004930019379, "learning_rate": 1.056463592387702e-08, "loss": 0.0123, "step": 59320 }, { "epoch": 17.262147221414025, "grad_norm": 0.110496386885643, "learning_rate": 1.0256680753785718e-08, "loss": 0.0115, "step": 59330 }, { "epoch": 17.265056735525167, "grad_norm": 0.09616170078516006, "learning_rate": 9.953279071934218e-09, "loss": 0.0123, "step": 59340 }, { "epoch": 17.26796624963631, "grad_norm": 0.07594374567270279, "learning_rate": 9.654430970487127e-09, "loss": 0.0097, "step": 59350 }, { "epoch": 17.270875763747455, "grad_norm": 0.09048900753259659, "learning_rate": 9.360136540225161e-09, "loss": 0.0108, "step": 59360 }, { "epoch": 17.273785277858597, "grad_norm": 0.08759546279907227, "learning_rate": 9.070395870550142e-09, "loss": 0.0123, "step": 59370 }, { "epoch": 17.276694791969742, "grad_norm": 0.08201134949922562, "learning_rate": 8.785209049476661e-09, "loss": 0.0128, "step": 59380 }, { "epoch": 17.279604306080884, "grad_norm": 0.08236429840326309, "learning_rate": 8.50457616363709e-09, "loss": 0.0107, "step": 59390 }, { "epoch": 17.28251382019203, "grad_norm": 0.1033085361123085, "learning_rate": 8.228497298278237e-09, "loss": 0.011, "step": 59400 }, { "epoch": 17.28542333430317, "grad_norm": 0.10775876045227051, "learning_rate": 7.956972537268015e-09, "loss": 0.0102, "step": 59410 }, { "epoch": 17.288332848414314, "grad_norm": 0.10953635722398758, "learning_rate": 7.690001963087112e-09, "loss": 0.0131, "step": 59420 }, { "epoch": 17.29124236252546, "grad_norm": 0.07344184070825577, "learning_rate": 7.427585656835656e-09, "loss": 0.0113, "step": 59430 }, { "epoch": 17.2941518766366, "grad_norm": 0.08047362416982651, "learning_rate": 7.169723698226549e-09, "loss": 0.0112, "step": 59440 }, { "epoch": 17.297061390747746, "grad_norm": 0.07015752047300339, "learning_rate": 6.916416165592132e-09, "loss": 0.0109, "step": 59450 }, { "epoch": 17.29997090485889, "grad_norm": 0.07911106199026108, "learning_rate": 6.667663135882518e-09, "loss": 0.0106, "step": 59460 }, { "epoch": 17.30288041897003, "grad_norm": 0.07475234568119049, "learning_rate": 6.423464684660596e-09, "loss": 0.0119, "step": 59470 }, { "epoch": 17.305789933081176, "grad_norm": 0.08940697461366653, "learning_rate": 6.183820886107028e-09, "loss": 0.0111, "step": 59480 }, { "epoch": 17.308699447192318, "grad_norm": 0.09774718433618546, "learning_rate": 5.948731813020247e-09, "loss": 0.011, "step": 59490 }, { "epoch": 17.311608961303463, "grad_norm": 0.09831699728965759, "learning_rate": 5.71819753681313e-09, "loss": 0.0117, "step": 59500 }, { "epoch": 17.314518475414605, "grad_norm": 0.08863252401351929, "learning_rate": 5.492218127516324e-09, "loss": 0.0102, "step": 59510 }, { "epoch": 17.31742798952575, "grad_norm": 0.09480199217796326, "learning_rate": 5.27079365377825e-09, "loss": 0.0114, "step": 59520 }, { "epoch": 17.320337503636893, "grad_norm": 0.11940187960863113, "learning_rate": 5.053924182858438e-09, "loss": 0.0116, "step": 59530 }, { "epoch": 17.323247017748034, "grad_norm": 0.13658249378204346, "learning_rate": 4.841609780639189e-09, "loss": 0.0102, "step": 59540 }, { "epoch": 17.32615653185918, "grad_norm": 0.06514211744070053, "learning_rate": 4.6338505116139126e-09, "loss": 0.0111, "step": 59550 }, { "epoch": 17.329066045970322, "grad_norm": 0.09051632136106491, "learning_rate": 4.430646438895458e-09, "loss": 0.0122, "step": 59560 }, { "epoch": 17.331975560081467, "grad_norm": 0.08018273115158081, "learning_rate": 4.2319976242111145e-09, "loss": 0.011, "step": 59570 }, { "epoch": 17.33488507419261, "grad_norm": 0.12227515131235123, "learning_rate": 4.037904127905945e-09, "loss": 0.0141, "step": 59580 }, { "epoch": 17.337794588303755, "grad_norm": 0.07824793457984924, "learning_rate": 3.8483660089394525e-09, "loss": 0.0107, "step": 59590 }, { "epoch": 17.340704102414897, "grad_norm": 0.08994415402412415, "learning_rate": 3.663383324890579e-09, "loss": 0.0117, "step": 59600 }, { "epoch": 17.34361361652604, "grad_norm": 0.09650090336799622, "learning_rate": 3.4829561319493773e-09, "loss": 0.013, "step": 59610 }, { "epoch": 17.346523130637184, "grad_norm": 0.09670593589544296, "learning_rate": 3.307084484925338e-09, "loss": 0.0109, "step": 59620 }, { "epoch": 17.349432644748326, "grad_norm": 0.10990480333566666, "learning_rate": 3.135768437245723e-09, "loss": 0.0118, "step": 59630 }, { "epoch": 17.35234215885947, "grad_norm": 0.10179460048675537, "learning_rate": 2.9690080409489063e-09, "loss": 0.0109, "step": 59640 }, { "epoch": 17.355251672970613, "grad_norm": 0.11962249875068665, "learning_rate": 2.8068033466943643e-09, "loss": 0.0105, "step": 59650 }, { "epoch": 17.35816118708176, "grad_norm": 0.1304507851600647, "learning_rate": 2.6491544037560157e-09, "loss": 0.0131, "step": 59660 }, { "epoch": 17.3610707011929, "grad_norm": 0.10963866859674454, "learning_rate": 2.4960612600222197e-09, "loss": 0.012, "step": 59670 }, { "epoch": 17.363980215304043, "grad_norm": 0.09220848977565765, "learning_rate": 2.3475239619974443e-09, "loss": 0.0111, "step": 59680 }, { "epoch": 17.36688972941519, "grad_norm": 0.07579764723777771, "learning_rate": 2.203542554805593e-09, "loss": 0.0104, "step": 59690 }, { "epoch": 17.36979924352633, "grad_norm": 0.1322464793920517, "learning_rate": 2.0641170821833477e-09, "loss": 0.0143, "step": 59700 }, { "epoch": 17.372708757637476, "grad_norm": 0.0909576565027237, "learning_rate": 1.929247586485161e-09, "loss": 0.0119, "step": 59710 }, { "epoch": 17.375618271748618, "grad_norm": 0.10148561745882034, "learning_rate": 1.7989341086815936e-09, "loss": 0.0125, "step": 59720 }, { "epoch": 17.378527785859763, "grad_norm": 0.153689444065094, "learning_rate": 1.6731766883559818e-09, "loss": 0.0107, "step": 59730 }, { "epoch": 17.381437299970905, "grad_norm": 0.07036694139242172, "learning_rate": 1.5519753637111e-09, "loss": 0.0122, "step": 59740 }, { "epoch": 17.384346814082047, "grad_norm": 0.08846903592348099, "learning_rate": 1.4353301715658295e-09, "loss": 0.0128, "step": 59750 }, { "epoch": 17.387256328193192, "grad_norm": 0.10691361129283905, "learning_rate": 1.323241147351828e-09, "loss": 0.0107, "step": 59760 }, { "epoch": 17.390165842304334, "grad_norm": 0.11642947047948837, "learning_rate": 1.215708325121856e-09, "loss": 0.0117, "step": 59770 }, { "epoch": 17.39307535641548, "grad_norm": 0.08014802634716034, "learning_rate": 1.1127317375381197e-09, "loss": 0.0115, "step": 59780 }, { "epoch": 17.395984870526622, "grad_norm": 0.08156087249517441, "learning_rate": 1.0143114158839285e-09, "loss": 0.0115, "step": 59790 }, { "epoch": 17.398894384637764, "grad_norm": 0.09057051688432693, "learning_rate": 9.204473900586985e-10, "loss": 0.0117, "step": 59800 }, { "epoch": 17.40180389874891, "grad_norm": 0.08711527287960052, "learning_rate": 8.311396885712918e-10, "loss": 0.0105, "step": 59810 }, { "epoch": 17.40471341286005, "grad_norm": 0.09343179315328598, "learning_rate": 7.463883385550041e-10, "loss": 0.0134, "step": 59820 }, { "epoch": 17.407622926971197, "grad_norm": 0.0773305669426918, "learning_rate": 6.661933657525765e-10, "loss": 0.0109, "step": 59830 }, { "epoch": 17.41053244108234, "grad_norm": 0.08694391697645187, "learning_rate": 5.905547945278533e-10, "loss": 0.0109, "step": 59840 }, { "epoch": 17.413441955193484, "grad_norm": 0.0982549786567688, "learning_rate": 5.194726478541245e-10, "loss": 0.0116, "step": 59850 }, { "epoch": 17.416351469304626, "grad_norm": 0.0849807858467102, "learning_rate": 4.5294694732744834e-10, "loss": 0.0111, "step": 59860 }, { "epoch": 17.419260983415768, "grad_norm": 0.15580494701862335, "learning_rate": 3.909777131549941e-10, "loss": 0.0116, "step": 59870 }, { "epoch": 17.422170497526913, "grad_norm": 0.1474861353635788, "learning_rate": 3.3356496416170335e-10, "loss": 0.0126, "step": 59880 }, { "epoch": 17.425080011638055, "grad_norm": 0.08108297735452652, "learning_rate": 2.8070871778862475e-10, "loss": 0.0121, "step": 59890 }, { "epoch": 17.4279895257492, "grad_norm": 0.06739162653684616, "learning_rate": 2.324089900912485e-10, "loss": 0.011, "step": 59900 }, { "epoch": 17.430899039860343, "grad_norm": 0.08557594567537308, "learning_rate": 1.8866579574117194e-10, "loss": 0.0106, "step": 59910 }, { "epoch": 17.433808553971488, "grad_norm": 0.1009904071688652, "learning_rate": 1.4947914802942996e-10, "loss": 0.0122, "step": 59920 }, { "epoch": 17.43671806808263, "grad_norm": 0.091752789914608, "learning_rate": 1.148490588565032e-10, "loss": 0.0104, "step": 59930 }, { "epoch": 17.439627582193772, "grad_norm": 0.0807497501373291, "learning_rate": 8.477553874230992e-11, "loss": 0.0112, "step": 59940 }, { "epoch": 17.442537096304918, "grad_norm": 0.10733125358819962, "learning_rate": 5.925859682620604e-11, "loss": 0.011, "step": 59950 }, { "epoch": 17.44544661041606, "grad_norm": 0.10366243869066238, "learning_rate": 3.829824085532785e-11, "loss": 0.0124, "step": 59960 }, { "epoch": 17.448356124527205, "grad_norm": 0.07822789996862411, "learning_rate": 2.1894477197914598e-11, "loss": 0.0119, "step": 59970 }, { "epoch": 17.451265638638347, "grad_norm": 0.07172305136919022, "learning_rate": 1.0047310838312562e-11, "loss": 0.0115, "step": 59980 }, { "epoch": 17.454175152749492, "grad_norm": 0.1012803465127945, "learning_rate": 2.7567453736443428e-12, "loss": 0.0103, "step": 59990 }, { "epoch": 17.457084666860634, "grad_norm": 0.09286589175462723, "learning_rate": 2.2783020470207305e-14, "loss": 0.0108, "step": 60000 }, { "epoch": 17.457084666860634, "step": 60000, "total_flos": 0.0, "train_loss": 0.022227557626614967, "train_runtime": 127625.7119, "train_samples_per_second": 60.176, "train_steps_per_second": 0.47 } ], "logging_steps": 10, "max_steps": 60000, "num_input_tokens_seen": 0, "num_train_epochs": 18, "save_steps": 20000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 128, "trial_name": null, "trial_params": null }