| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 1.0, |
| "eval_steps": 500, |
| "global_step": 518, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0019305019305019305, |
| "grad_norm": 9.087823867797852, |
| "learning_rate": 9.980694980694981e-06, |
| "loss": 763.5779, |
| "step": 1 |
| }, |
| { |
| "epoch": 0.003861003861003861, |
| "grad_norm": 8.02014446258545, |
| "learning_rate": 9.961389961389962e-06, |
| "loss": 742.0983, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.005791505791505791, |
| "grad_norm": 6.846795558929443, |
| "learning_rate": 9.942084942084944e-06, |
| "loss": 754.556, |
| "step": 3 |
| }, |
| { |
| "epoch": 0.007722007722007722, |
| "grad_norm": 5.642313003540039, |
| "learning_rate": 9.922779922779924e-06, |
| "loss": 760.8958, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.009652509652509652, |
| "grad_norm": 4.954174995422363, |
| "learning_rate": 9.903474903474905e-06, |
| "loss": 773.4886, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.011583011583011582, |
| "grad_norm": 4.209826946258545, |
| "learning_rate": 9.884169884169885e-06, |
| "loss": 752.0172, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.013513513513513514, |
| "grad_norm": 3.8238916397094727, |
| "learning_rate": 9.864864864864865e-06, |
| "loss": 765.1586, |
| "step": 7 |
| }, |
| { |
| "epoch": 0.015444015444015444, |
| "grad_norm": 3.472975730895996, |
| "learning_rate": 9.845559845559847e-06, |
| "loss": 758.0544, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.017374517374517374, |
| "grad_norm": 3.3365626335144043, |
| "learning_rate": 9.826254826254828e-06, |
| "loss": 760.6899, |
| "step": 9 |
| }, |
| { |
| "epoch": 0.019305019305019305, |
| "grad_norm": 3.205853223800659, |
| "learning_rate": 9.806949806949808e-06, |
| "loss": 763.1088, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.021235521235521235, |
| "grad_norm": 2.9912545680999756, |
| "learning_rate": 9.787644787644789e-06, |
| "loss": 751.3445, |
| "step": 11 |
| }, |
| { |
| "epoch": 0.023166023166023165, |
| "grad_norm": 2.9894907474517822, |
| "learning_rate": 9.768339768339769e-06, |
| "loss": 762.5296, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.025096525096525095, |
| "grad_norm": 2.973778009414673, |
| "learning_rate": 9.749034749034751e-06, |
| "loss": 766.5692, |
| "step": 13 |
| }, |
| { |
| "epoch": 0.02702702702702703, |
| "grad_norm": 2.866206645965576, |
| "learning_rate": 9.729729729729732e-06, |
| "loss": 759.659, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.02895752895752896, |
| "grad_norm": 2.7454729080200195, |
| "learning_rate": 9.71042471042471e-06, |
| "loss": 749.6011, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.03088803088803089, |
| "grad_norm": 2.7870497703552246, |
| "learning_rate": 9.69111969111969e-06, |
| "loss": 757.2128, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.032818532818532815, |
| "grad_norm": 2.727757453918457, |
| "learning_rate": 9.671814671814673e-06, |
| "loss": 752.8752, |
| "step": 17 |
| }, |
| { |
| "epoch": 0.03474903474903475, |
| "grad_norm": 2.761842966079712, |
| "learning_rate": 9.652509652509653e-06, |
| "loss": 763.5405, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.03667953667953668, |
| "grad_norm": 2.64555025100708, |
| "learning_rate": 9.633204633204634e-06, |
| "loss": 752.1604, |
| "step": 19 |
| }, |
| { |
| "epoch": 0.03861003861003861, |
| "grad_norm": 2.604011297225952, |
| "learning_rate": 9.613899613899614e-06, |
| "loss": 745.1989, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.04054054054054054, |
| "grad_norm": 2.5989909172058105, |
| "learning_rate": 9.594594594594594e-06, |
| "loss": 744.575, |
| "step": 21 |
| }, |
| { |
| "epoch": 0.04247104247104247, |
| "grad_norm": 2.646167278289795, |
| "learning_rate": 9.575289575289576e-06, |
| "loss": 753.7839, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.0444015444015444, |
| "grad_norm": 2.6674346923828125, |
| "learning_rate": 9.555984555984557e-06, |
| "loss": 764.2125, |
| "step": 23 |
| }, |
| { |
| "epoch": 0.04633204633204633, |
| "grad_norm": 2.6234052181243896, |
| "learning_rate": 9.536679536679537e-06, |
| "loss": 753.5481, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.04826254826254826, |
| "grad_norm": 2.550206422805786, |
| "learning_rate": 9.517374517374518e-06, |
| "loss": 744.3314, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.05019305019305019, |
| "grad_norm": 2.634002208709717, |
| "learning_rate": 9.498069498069498e-06, |
| "loss": 756.7723, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.052123552123552123, |
| "grad_norm": 2.632336378097534, |
| "learning_rate": 9.47876447876448e-06, |
| "loss": 756.9872, |
| "step": 27 |
| }, |
| { |
| "epoch": 0.05405405405405406, |
| "grad_norm": 2.628897190093994, |
| "learning_rate": 9.45945945945946e-06, |
| "loss": 756.0316, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.055984555984555984, |
| "grad_norm": 2.6032018661499023, |
| "learning_rate": 9.440154440154441e-06, |
| "loss": 754.0662, |
| "step": 29 |
| }, |
| { |
| "epoch": 0.05791505791505792, |
| "grad_norm": 2.5998055934906006, |
| "learning_rate": 9.420849420849421e-06, |
| "loss": 749.3468, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.059845559845559844, |
| "grad_norm": 2.623134136199951, |
| "learning_rate": 9.401544401544402e-06, |
| "loss": 753.1194, |
| "step": 31 |
| }, |
| { |
| "epoch": 0.06177606177606178, |
| "grad_norm": 2.636357069015503, |
| "learning_rate": 9.382239382239384e-06, |
| "loss": 755.1291, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.0637065637065637, |
| "grad_norm": 2.709622621536255, |
| "learning_rate": 9.362934362934364e-06, |
| "loss": 774.9333, |
| "step": 33 |
| }, |
| { |
| "epoch": 0.06563706563706563, |
| "grad_norm": 2.6019175052642822, |
| "learning_rate": 9.343629343629345e-06, |
| "loss": 758.8455, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.06756756756756757, |
| "grad_norm": 2.620014190673828, |
| "learning_rate": 9.324324324324325e-06, |
| "loss": 751.6978, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.0694980694980695, |
| "grad_norm": 2.6363117694854736, |
| "learning_rate": 9.305019305019305e-06, |
| "loss": 755.4987, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.07142857142857142, |
| "grad_norm": 2.697289228439331, |
| "learning_rate": 9.285714285714288e-06, |
| "loss": 769.0797, |
| "step": 37 |
| }, |
| { |
| "epoch": 0.07335907335907337, |
| "grad_norm": 2.684602975845337, |
| "learning_rate": 9.266409266409268e-06, |
| "loss": 771.785, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.07528957528957529, |
| "grad_norm": 2.6869068145751953, |
| "learning_rate": 9.247104247104248e-06, |
| "loss": 771.0751, |
| "step": 39 |
| }, |
| { |
| "epoch": 0.07722007722007722, |
| "grad_norm": 2.65386962890625, |
| "learning_rate": 9.227799227799229e-06, |
| "loss": 766.6141, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.07915057915057915, |
| "grad_norm": 2.5492303371429443, |
| "learning_rate": 9.20849420849421e-06, |
| "loss": 741.012, |
| "step": 41 |
| }, |
| { |
| "epoch": 0.08108108108108109, |
| "grad_norm": 2.659393548965454, |
| "learning_rate": 9.189189189189191e-06, |
| "loss": 761.9696, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.08301158301158301, |
| "grad_norm": 2.6672518253326416, |
| "learning_rate": 9.169884169884172e-06, |
| "loss": 762.8466, |
| "step": 43 |
| }, |
| { |
| "epoch": 0.08494208494208494, |
| "grad_norm": 2.60377836227417, |
| "learning_rate": 9.15057915057915e-06, |
| "loss": 758.5039, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.08687258687258688, |
| "grad_norm": 2.622608184814453, |
| "learning_rate": 9.13127413127413e-06, |
| "loss": 750.5551, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.0888030888030888, |
| "grad_norm": 2.6671438217163086, |
| "learning_rate": 9.111969111969113e-06, |
| "loss": 766.5024, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.09073359073359073, |
| "grad_norm": 2.5831432342529297, |
| "learning_rate": 9.092664092664093e-06, |
| "loss": 747.624, |
| "step": 47 |
| }, |
| { |
| "epoch": 0.09266409266409266, |
| "grad_norm": 2.61625075340271, |
| "learning_rate": 9.073359073359074e-06, |
| "loss": 759.9783, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.0945945945945946, |
| "grad_norm": 2.6503899097442627, |
| "learning_rate": 9.054054054054054e-06, |
| "loss": 766.5831, |
| "step": 49 |
| }, |
| { |
| "epoch": 0.09652509652509653, |
| "grad_norm": 2.713566303253174, |
| "learning_rate": 9.034749034749034e-06, |
| "loss": 764.6612, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.09845559845559845, |
| "grad_norm": 2.6607675552368164, |
| "learning_rate": 9.015444015444017e-06, |
| "loss": 759.2146, |
| "step": 51 |
| }, |
| { |
| "epoch": 0.10038610038610038, |
| "grad_norm": 2.703293800354004, |
| "learning_rate": 8.996138996138997e-06, |
| "loss": 771.3636, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.10231660231660232, |
| "grad_norm": 2.6380341053009033, |
| "learning_rate": 8.976833976833977e-06, |
| "loss": 762.6519, |
| "step": 53 |
| }, |
| { |
| "epoch": 0.10424710424710425, |
| "grad_norm": 2.6804656982421875, |
| "learning_rate": 8.957528957528958e-06, |
| "loss": 764.4567, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.10617760617760617, |
| "grad_norm": 2.6722824573516846, |
| "learning_rate": 8.938223938223938e-06, |
| "loss": 762.7928, |
| "step": 55 |
| }, |
| { |
| "epoch": 0.10810810810810811, |
| "grad_norm": 2.5956473350524902, |
| "learning_rate": 8.91891891891892e-06, |
| "loss": 757.1573, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.11003861003861004, |
| "grad_norm": 2.67158579826355, |
| "learning_rate": 8.8996138996139e-06, |
| "loss": 762.9989, |
| "step": 57 |
| }, |
| { |
| "epoch": 0.11196911196911197, |
| "grad_norm": 2.6719319820404053, |
| "learning_rate": 8.880308880308881e-06, |
| "loss": 765.219, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.1138996138996139, |
| "grad_norm": 2.690995931625366, |
| "learning_rate": 8.861003861003861e-06, |
| "loss": 755.7704, |
| "step": 59 |
| }, |
| { |
| "epoch": 0.11583011583011583, |
| "grad_norm": 2.672006845474243, |
| "learning_rate": 8.841698841698842e-06, |
| "loss": 755.6411, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.11776061776061776, |
| "grad_norm": 2.6625969409942627, |
| "learning_rate": 8.822393822393824e-06, |
| "loss": 755.6705, |
| "step": 61 |
| }, |
| { |
| "epoch": 0.11969111969111969, |
| "grad_norm": 2.6644089221954346, |
| "learning_rate": 8.803088803088804e-06, |
| "loss": 755.7343, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.12162162162162163, |
| "grad_norm": 2.6913063526153564, |
| "learning_rate": 8.783783783783785e-06, |
| "loss": 760.8112, |
| "step": 63 |
| }, |
| { |
| "epoch": 0.12355212355212356, |
| "grad_norm": 2.640939474105835, |
| "learning_rate": 8.764478764478765e-06, |
| "loss": 752.9957, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.12548262548262548, |
| "grad_norm": 2.706070899963379, |
| "learning_rate": 8.745173745173746e-06, |
| "loss": 751.4651, |
| "step": 65 |
| }, |
| { |
| "epoch": 0.1274131274131274, |
| "grad_norm": 2.6751809120178223, |
| "learning_rate": 8.725868725868728e-06, |
| "loss": 760.4512, |
| "step": 66 |
| }, |
| { |
| "epoch": 0.12934362934362933, |
| "grad_norm": 2.731750965118408, |
| "learning_rate": 8.706563706563708e-06, |
| "loss": 766.3732, |
| "step": 67 |
| }, |
| { |
| "epoch": 0.13127413127413126, |
| "grad_norm": 2.688737392425537, |
| "learning_rate": 8.687258687258689e-06, |
| "loss": 749.8923, |
| "step": 68 |
| }, |
| { |
| "epoch": 0.13320463320463322, |
| "grad_norm": 2.67354679107666, |
| "learning_rate": 8.667953667953669e-06, |
| "loss": 751.1083, |
| "step": 69 |
| }, |
| { |
| "epoch": 0.13513513513513514, |
| "grad_norm": 2.6414129734039307, |
| "learning_rate": 8.64864864864865e-06, |
| "loss": 733.0116, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.13706563706563707, |
| "grad_norm": 3.0136189460754395, |
| "learning_rate": 8.629343629343631e-06, |
| "loss": 764.5856, |
| "step": 71 |
| }, |
| { |
| "epoch": 0.138996138996139, |
| "grad_norm": 2.7174079418182373, |
| "learning_rate": 8.61003861003861e-06, |
| "loss": 753.4861, |
| "step": 72 |
| }, |
| { |
| "epoch": 0.14092664092664092, |
| "grad_norm": 2.714317560195923, |
| "learning_rate": 8.59073359073359e-06, |
| "loss": 754.7225, |
| "step": 73 |
| }, |
| { |
| "epoch": 0.14285714285714285, |
| "grad_norm": 2.7697298526763916, |
| "learning_rate": 8.571428571428571e-06, |
| "loss": 765.4945, |
| "step": 74 |
| }, |
| { |
| "epoch": 0.14478764478764478, |
| "grad_norm": 2.7908132076263428, |
| "learning_rate": 8.552123552123553e-06, |
| "loss": 779.0895, |
| "step": 75 |
| }, |
| { |
| "epoch": 0.14671814671814673, |
| "grad_norm": 2.7896623611450195, |
| "learning_rate": 8.532818532818533e-06, |
| "loss": 767.8585, |
| "step": 76 |
| }, |
| { |
| "epoch": 0.14864864864864866, |
| "grad_norm": 2.718611240386963, |
| "learning_rate": 8.513513513513514e-06, |
| "loss": 760.3116, |
| "step": 77 |
| }, |
| { |
| "epoch": 0.15057915057915058, |
| "grad_norm": 2.78110671043396, |
| "learning_rate": 8.494208494208494e-06, |
| "loss": 761.649, |
| "step": 78 |
| }, |
| { |
| "epoch": 0.1525096525096525, |
| "grad_norm": 2.822859048843384, |
| "learning_rate": 8.474903474903475e-06, |
| "loss": 766.2858, |
| "step": 79 |
| }, |
| { |
| "epoch": 0.15444015444015444, |
| "grad_norm": 2.719345808029175, |
| "learning_rate": 8.455598455598457e-06, |
| "loss": 755.3741, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.15637065637065636, |
| "grad_norm": 2.710378408432007, |
| "learning_rate": 8.436293436293437e-06, |
| "loss": 758.8147, |
| "step": 81 |
| }, |
| { |
| "epoch": 0.1583011583011583, |
| "grad_norm": 2.7957656383514404, |
| "learning_rate": 8.416988416988418e-06, |
| "loss": 761.3373, |
| "step": 82 |
| }, |
| { |
| "epoch": 0.16023166023166024, |
| "grad_norm": 2.8298914432525635, |
| "learning_rate": 8.397683397683398e-06, |
| "loss": 769.8754, |
| "step": 83 |
| }, |
| { |
| "epoch": 0.16216216216216217, |
| "grad_norm": 2.762465715408325, |
| "learning_rate": 8.378378378378378e-06, |
| "loss": 769.7825, |
| "step": 84 |
| }, |
| { |
| "epoch": 0.1640926640926641, |
| "grad_norm": 2.773038625717163, |
| "learning_rate": 8.35907335907336e-06, |
| "loss": 749.3358, |
| "step": 85 |
| }, |
| { |
| "epoch": 0.16602316602316602, |
| "grad_norm": 2.7703516483306885, |
| "learning_rate": 8.33976833976834e-06, |
| "loss": 757.9563, |
| "step": 86 |
| }, |
| { |
| "epoch": 0.16795366795366795, |
| "grad_norm": 2.930047035217285, |
| "learning_rate": 8.320463320463321e-06, |
| "loss": 764.3655, |
| "step": 87 |
| }, |
| { |
| "epoch": 0.16988416988416988, |
| "grad_norm": 2.7261102199554443, |
| "learning_rate": 8.301158301158302e-06, |
| "loss": 748.2144, |
| "step": 88 |
| }, |
| { |
| "epoch": 0.1718146718146718, |
| "grad_norm": 2.793009042739868, |
| "learning_rate": 8.281853281853282e-06, |
| "loss": 754.8672, |
| "step": 89 |
| }, |
| { |
| "epoch": 0.17374517374517376, |
| "grad_norm": 2.8894758224487305, |
| "learning_rate": 8.262548262548264e-06, |
| "loss": 764.7795, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.17567567567567569, |
| "grad_norm": 2.8510689735412598, |
| "learning_rate": 8.243243243243245e-06, |
| "loss": 753.4908, |
| "step": 91 |
| }, |
| { |
| "epoch": 0.1776061776061776, |
| "grad_norm": 2.757417917251587, |
| "learning_rate": 8.223938223938225e-06, |
| "loss": 760.0729, |
| "step": 92 |
| }, |
| { |
| "epoch": 0.17953667953667954, |
| "grad_norm": 2.8439671993255615, |
| "learning_rate": 8.204633204633205e-06, |
| "loss": 771.863, |
| "step": 93 |
| }, |
| { |
| "epoch": 0.18146718146718147, |
| "grad_norm": 2.8313705921173096, |
| "learning_rate": 8.185328185328186e-06, |
| "loss": 765.174, |
| "step": 94 |
| }, |
| { |
| "epoch": 0.1833976833976834, |
| "grad_norm": 2.7856626510620117, |
| "learning_rate": 8.166023166023168e-06, |
| "loss": 758.1021, |
| "step": 95 |
| }, |
| { |
| "epoch": 0.18532818532818532, |
| "grad_norm": 2.8734123706817627, |
| "learning_rate": 8.146718146718148e-06, |
| "loss": 768.5622, |
| "step": 96 |
| }, |
| { |
| "epoch": 0.18725868725868725, |
| "grad_norm": 2.833714723587036, |
| "learning_rate": 8.127413127413129e-06, |
| "loss": 753.1698, |
| "step": 97 |
| }, |
| { |
| "epoch": 0.1891891891891892, |
| "grad_norm": 2.8014986515045166, |
| "learning_rate": 8.108108108108109e-06, |
| "loss": 752.7419, |
| "step": 98 |
| }, |
| { |
| "epoch": 0.19111969111969113, |
| "grad_norm": 2.7595949172973633, |
| "learning_rate": 8.08880308880309e-06, |
| "loss": 755.9596, |
| "step": 99 |
| }, |
| { |
| "epoch": 0.19305019305019305, |
| "grad_norm": 2.7833123207092285, |
| "learning_rate": 8.06949806949807e-06, |
| "loss": 765.9607, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.19498069498069498, |
| "grad_norm": 2.777548313140869, |
| "learning_rate": 8.05019305019305e-06, |
| "loss": 766.981, |
| "step": 101 |
| }, |
| { |
| "epoch": 0.1969111969111969, |
| "grad_norm": 2.7974212169647217, |
| "learning_rate": 8.03088803088803e-06, |
| "loss": 753.3157, |
| "step": 102 |
| }, |
| { |
| "epoch": 0.19884169884169883, |
| "grad_norm": 3.194911003112793, |
| "learning_rate": 8.011583011583011e-06, |
| "loss": 773.3227, |
| "step": 103 |
| }, |
| { |
| "epoch": 0.20077220077220076, |
| "grad_norm": 2.7325661182403564, |
| "learning_rate": 7.992277992277993e-06, |
| "loss": 746.1143, |
| "step": 104 |
| }, |
| { |
| "epoch": 0.20270270270270271, |
| "grad_norm": 2.740203619003296, |
| "learning_rate": 7.972972972972974e-06, |
| "loss": 750.4532, |
| "step": 105 |
| }, |
| { |
| "epoch": 0.20463320463320464, |
| "grad_norm": 2.764615297317505, |
| "learning_rate": 7.953667953667954e-06, |
| "loss": 757.8738, |
| "step": 106 |
| }, |
| { |
| "epoch": 0.20656370656370657, |
| "grad_norm": 2.783639430999756, |
| "learning_rate": 7.934362934362934e-06, |
| "loss": 759.7949, |
| "step": 107 |
| }, |
| { |
| "epoch": 0.2084942084942085, |
| "grad_norm": 2.933793783187866, |
| "learning_rate": 7.915057915057915e-06, |
| "loss": 771.2281, |
| "step": 108 |
| }, |
| { |
| "epoch": 0.21042471042471042, |
| "grad_norm": 2.795241117477417, |
| "learning_rate": 7.895752895752897e-06, |
| "loss": 747.9264, |
| "step": 109 |
| }, |
| { |
| "epoch": 0.21235521235521235, |
| "grad_norm": 2.755584239959717, |
| "learning_rate": 7.876447876447877e-06, |
| "loss": 746.8754, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.21428571428571427, |
| "grad_norm": 2.802136182785034, |
| "learning_rate": 7.857142857142858e-06, |
| "loss": 745.1733, |
| "step": 111 |
| }, |
| { |
| "epoch": 0.21621621621621623, |
| "grad_norm": 3.5105533599853516, |
| "learning_rate": 7.837837837837838e-06, |
| "loss": 745.6346, |
| "step": 112 |
| }, |
| { |
| "epoch": 0.21814671814671815, |
| "grad_norm": 2.8462111949920654, |
| "learning_rate": 7.818532818532818e-06, |
| "loss": 757.0023, |
| "step": 113 |
| }, |
| { |
| "epoch": 0.22007722007722008, |
| "grad_norm": 3.079543113708496, |
| "learning_rate": 7.7992277992278e-06, |
| "loss": 757.2842, |
| "step": 114 |
| }, |
| { |
| "epoch": 0.222007722007722, |
| "grad_norm": 2.8306119441986084, |
| "learning_rate": 7.779922779922781e-06, |
| "loss": 752.3796, |
| "step": 115 |
| }, |
| { |
| "epoch": 0.22393822393822393, |
| "grad_norm": 3.1957056522369385, |
| "learning_rate": 7.760617760617761e-06, |
| "loss": 753.2894, |
| "step": 116 |
| }, |
| { |
| "epoch": 0.22586872586872586, |
| "grad_norm": 2.90783953666687, |
| "learning_rate": 7.741312741312742e-06, |
| "loss": 753.8042, |
| "step": 117 |
| }, |
| { |
| "epoch": 0.2277992277992278, |
| "grad_norm": 2.8529882431030273, |
| "learning_rate": 7.722007722007722e-06, |
| "loss": 758.0192, |
| "step": 118 |
| }, |
| { |
| "epoch": 0.22972972972972974, |
| "grad_norm": 2.8292617797851562, |
| "learning_rate": 7.702702702702704e-06, |
| "loss": 756.563, |
| "step": 119 |
| }, |
| { |
| "epoch": 0.23166023166023167, |
| "grad_norm": 2.8797755241394043, |
| "learning_rate": 7.683397683397685e-06, |
| "loss": 761.1346, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.2335907335907336, |
| "grad_norm": 2.7390382289886475, |
| "learning_rate": 7.664092664092665e-06, |
| "loss": 750.3627, |
| "step": 121 |
| }, |
| { |
| "epoch": 0.23552123552123552, |
| "grad_norm": 2.7686545848846436, |
| "learning_rate": 7.644787644787645e-06, |
| "loss": 750.1412, |
| "step": 122 |
| }, |
| { |
| "epoch": 0.23745173745173745, |
| "grad_norm": 2.798008680343628, |
| "learning_rate": 7.625482625482627e-06, |
| "loss": 742.588, |
| "step": 123 |
| }, |
| { |
| "epoch": 0.23938223938223938, |
| "grad_norm": 4.088344097137451, |
| "learning_rate": 7.606177606177607e-06, |
| "loss": 758.9719, |
| "step": 124 |
| }, |
| { |
| "epoch": 0.2413127413127413, |
| "grad_norm": 2.9229252338409424, |
| "learning_rate": 7.5868725868725875e-06, |
| "loss": 756.5762, |
| "step": 125 |
| }, |
| { |
| "epoch": 0.24324324324324326, |
| "grad_norm": 3.1966676712036133, |
| "learning_rate": 7.567567567567569e-06, |
| "loss": 768.2112, |
| "step": 126 |
| }, |
| { |
| "epoch": 0.24517374517374518, |
| "grad_norm": 3.255308151245117, |
| "learning_rate": 7.548262548262549e-06, |
| "loss": 768.8184, |
| "step": 127 |
| }, |
| { |
| "epoch": 0.2471042471042471, |
| "grad_norm": 2.803837537765503, |
| "learning_rate": 7.5289575289575304e-06, |
| "loss": 758.1424, |
| "step": 128 |
| }, |
| { |
| "epoch": 0.24903474903474904, |
| "grad_norm": 2.8075811862945557, |
| "learning_rate": 7.509652509652511e-06, |
| "loss": 748.3246, |
| "step": 129 |
| }, |
| { |
| "epoch": 0.25096525096525096, |
| "grad_norm": 3.1468892097473145, |
| "learning_rate": 7.49034749034749e-06, |
| "loss": 759.9876, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.2528957528957529, |
| "grad_norm": 3.324525833129883, |
| "learning_rate": 7.471042471042471e-06, |
| "loss": 756.2051, |
| "step": 131 |
| }, |
| { |
| "epoch": 0.2548262548262548, |
| "grad_norm": 2.8439035415649414, |
| "learning_rate": 7.451737451737452e-06, |
| "loss": 755.392, |
| "step": 132 |
| }, |
| { |
| "epoch": 0.25675675675675674, |
| "grad_norm": 2.882779598236084, |
| "learning_rate": 7.4324324324324324e-06, |
| "loss": 771.8354, |
| "step": 133 |
| }, |
| { |
| "epoch": 0.25868725868725867, |
| "grad_norm": 3.0841407775878906, |
| "learning_rate": 7.413127413127414e-06, |
| "loss": 754.2871, |
| "step": 134 |
| }, |
| { |
| "epoch": 0.2606177606177606, |
| "grad_norm": 3.2702057361602783, |
| "learning_rate": 7.393822393822394e-06, |
| "loss": 742.1606, |
| "step": 135 |
| }, |
| { |
| "epoch": 0.2625482625482625, |
| "grad_norm": 2.7956244945526123, |
| "learning_rate": 7.3745173745173745e-06, |
| "loss": 750.0869, |
| "step": 136 |
| }, |
| { |
| "epoch": 0.2644787644787645, |
| "grad_norm": 3.004718542098999, |
| "learning_rate": 7.355212355212356e-06, |
| "loss": 781.2869, |
| "step": 137 |
| }, |
| { |
| "epoch": 0.26640926640926643, |
| "grad_norm": 3.8438968658447266, |
| "learning_rate": 7.335907335907336e-06, |
| "loss": 755.309, |
| "step": 138 |
| }, |
| { |
| "epoch": 0.26833976833976836, |
| "grad_norm": 3.0307114124298096, |
| "learning_rate": 7.316602316602317e-06, |
| "loss": 757.5778, |
| "step": 139 |
| }, |
| { |
| "epoch": 0.2702702702702703, |
| "grad_norm": 2.9648211002349854, |
| "learning_rate": 7.297297297297298e-06, |
| "loss": 753.4247, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.2722007722007722, |
| "grad_norm": 2.9691784381866455, |
| "learning_rate": 7.277992277992278e-06, |
| "loss": 757.2734, |
| "step": 141 |
| }, |
| { |
| "epoch": 0.27413127413127414, |
| "grad_norm": 2.8389272689819336, |
| "learning_rate": 7.2586872586872595e-06, |
| "loss": 758.1973, |
| "step": 142 |
| }, |
| { |
| "epoch": 0.27606177606177607, |
| "grad_norm": 2.8457741737365723, |
| "learning_rate": 7.23938223938224e-06, |
| "loss": 759.3056, |
| "step": 143 |
| }, |
| { |
| "epoch": 0.277992277992278, |
| "grad_norm": 2.9947755336761475, |
| "learning_rate": 7.22007722007722e-06, |
| "loss": 756.6718, |
| "step": 144 |
| }, |
| { |
| "epoch": 0.2799227799227799, |
| "grad_norm": 3.285741090774536, |
| "learning_rate": 7.2007722007722015e-06, |
| "loss": 741.2257, |
| "step": 145 |
| }, |
| { |
| "epoch": 0.28185328185328185, |
| "grad_norm": 3.118593692779541, |
| "learning_rate": 7.181467181467182e-06, |
| "loss": 757.1205, |
| "step": 146 |
| }, |
| { |
| "epoch": 0.28378378378378377, |
| "grad_norm": 3.1158804893493652, |
| "learning_rate": 7.162162162162163e-06, |
| "loss": 747.1119, |
| "step": 147 |
| }, |
| { |
| "epoch": 0.2857142857142857, |
| "grad_norm": 2.9939019680023193, |
| "learning_rate": 7.1428571428571436e-06, |
| "loss": 738.4834, |
| "step": 148 |
| }, |
| { |
| "epoch": 0.2876447876447876, |
| "grad_norm": 3.792679786682129, |
| "learning_rate": 7.123552123552124e-06, |
| "loss": 758.1683, |
| "step": 149 |
| }, |
| { |
| "epoch": 0.28957528957528955, |
| "grad_norm": 3.3988595008850098, |
| "learning_rate": 7.104247104247105e-06, |
| "loss": 757.7793, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.2915057915057915, |
| "grad_norm": 3.4164488315582275, |
| "learning_rate": 7.084942084942086e-06, |
| "loss": 753.686, |
| "step": 151 |
| }, |
| { |
| "epoch": 0.29343629343629346, |
| "grad_norm": 4.295523166656494, |
| "learning_rate": 7.065637065637067e-06, |
| "loss": 749.4221, |
| "step": 152 |
| }, |
| { |
| "epoch": 0.2953667953667954, |
| "grad_norm": 3.1985435485839844, |
| "learning_rate": 7.046332046332047e-06, |
| "loss": 766.7484, |
| "step": 153 |
| }, |
| { |
| "epoch": 0.2972972972972973, |
| "grad_norm": 2.811443567276001, |
| "learning_rate": 7.027027027027028e-06, |
| "loss": 744.4705, |
| "step": 154 |
| }, |
| { |
| "epoch": 0.29922779922779924, |
| "grad_norm": 2.8985869884490967, |
| "learning_rate": 7.007722007722009e-06, |
| "loss": 771.2823, |
| "step": 155 |
| }, |
| { |
| "epoch": 0.30115830115830117, |
| "grad_norm": 2.913670301437378, |
| "learning_rate": 6.988416988416989e-06, |
| "loss": 767.1945, |
| "step": 156 |
| }, |
| { |
| "epoch": 0.3030888030888031, |
| "grad_norm": 3.091149091720581, |
| "learning_rate": 6.9691119691119706e-06, |
| "loss": 770.168, |
| "step": 157 |
| }, |
| { |
| "epoch": 0.305019305019305, |
| "grad_norm": 2.89846134185791, |
| "learning_rate": 6.949806949806951e-06, |
| "loss": 746.1279, |
| "step": 158 |
| }, |
| { |
| "epoch": 0.30694980694980695, |
| "grad_norm": 2.981823444366455, |
| "learning_rate": 6.9305019305019305e-06, |
| "loss": 749.2059, |
| "step": 159 |
| }, |
| { |
| "epoch": 0.3088803088803089, |
| "grad_norm": 3.6476705074310303, |
| "learning_rate": 6.911196911196911e-06, |
| "loss": 757.4434, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.3108108108108108, |
| "grad_norm": 2.946174383163452, |
| "learning_rate": 6.891891891891892e-06, |
| "loss": 759.8816, |
| "step": 161 |
| }, |
| { |
| "epoch": 0.3127413127413127, |
| "grad_norm": 3.030103921890259, |
| "learning_rate": 6.872586872586873e-06, |
| "loss": 757.3468, |
| "step": 162 |
| }, |
| { |
| "epoch": 0.31467181467181465, |
| "grad_norm": 3.3172245025634766, |
| "learning_rate": 6.853281853281854e-06, |
| "loss": 753.3054, |
| "step": 163 |
| }, |
| { |
| "epoch": 0.3166023166023166, |
| "grad_norm": 3.128265857696533, |
| "learning_rate": 6.833976833976834e-06, |
| "loss": 764.0436, |
| "step": 164 |
| }, |
| { |
| "epoch": 0.3185328185328185, |
| "grad_norm": 2.8332438468933105, |
| "learning_rate": 6.814671814671815e-06, |
| "loss": 747.4676, |
| "step": 165 |
| }, |
| { |
| "epoch": 0.3204633204633205, |
| "grad_norm": 2.9192118644714355, |
| "learning_rate": 6.795366795366796e-06, |
| "loss": 759.479, |
| "step": 166 |
| }, |
| { |
| "epoch": 0.3223938223938224, |
| "grad_norm": 2.926225423812866, |
| "learning_rate": 6.776061776061776e-06, |
| "loss": 749.5502, |
| "step": 167 |
| }, |
| { |
| "epoch": 0.32432432432432434, |
| "grad_norm": 3.739107608795166, |
| "learning_rate": 6.7567567567567575e-06, |
| "loss": 761.9246, |
| "step": 168 |
| }, |
| { |
| "epoch": 0.32625482625482627, |
| "grad_norm": 3.1689376831054688, |
| "learning_rate": 6.737451737451738e-06, |
| "loss": 759.9313, |
| "step": 169 |
| }, |
| { |
| "epoch": 0.3281853281853282, |
| "grad_norm": 4.102996349334717, |
| "learning_rate": 6.718146718146718e-06, |
| "loss": 755.9019, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.3301158301158301, |
| "grad_norm": 4.32927131652832, |
| "learning_rate": 6.6988416988417e-06, |
| "loss": 748.2408, |
| "step": 171 |
| }, |
| { |
| "epoch": 0.33204633204633205, |
| "grad_norm": 2.9341273307800293, |
| "learning_rate": 6.67953667953668e-06, |
| "loss": 767.1997, |
| "step": 172 |
| }, |
| { |
| "epoch": 0.333976833976834, |
| "grad_norm": 2.9747681617736816, |
| "learning_rate": 6.66023166023166e-06, |
| "loss": 749.2424, |
| "step": 173 |
| }, |
| { |
| "epoch": 0.3359073359073359, |
| "grad_norm": 3.03954815864563, |
| "learning_rate": 6.640926640926642e-06, |
| "loss": 753.1185, |
| "step": 174 |
| }, |
| { |
| "epoch": 0.33783783783783783, |
| "grad_norm": 2.896865129470825, |
| "learning_rate": 6.621621621621622e-06, |
| "loss": 760.0892, |
| "step": 175 |
| }, |
| { |
| "epoch": 0.33976833976833976, |
| "grad_norm": 4.517345905303955, |
| "learning_rate": 6.602316602316603e-06, |
| "loss": 761.2517, |
| "step": 176 |
| }, |
| { |
| "epoch": 0.3416988416988417, |
| "grad_norm": 3.4095852375030518, |
| "learning_rate": 6.583011583011584e-06, |
| "loss": 760.7, |
| "step": 177 |
| }, |
| { |
| "epoch": 0.3436293436293436, |
| "grad_norm": 3.5481455326080322, |
| "learning_rate": 6.563706563706564e-06, |
| "loss": 761.6207, |
| "step": 178 |
| }, |
| { |
| "epoch": 0.34555984555984554, |
| "grad_norm": 4.07427978515625, |
| "learning_rate": 6.544401544401545e-06, |
| "loss": 762.6121, |
| "step": 179 |
| }, |
| { |
| "epoch": 0.3474903474903475, |
| "grad_norm": 3.08604097366333, |
| "learning_rate": 6.525096525096526e-06, |
| "loss": 743.1675, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.34942084942084944, |
| "grad_norm": 4.791824817657471, |
| "learning_rate": 6.505791505791507e-06, |
| "loss": 773.7855, |
| "step": 181 |
| }, |
| { |
| "epoch": 0.35135135135135137, |
| "grad_norm": 5.513062953948975, |
| "learning_rate": 6.486486486486487e-06, |
| "loss": 755.7749, |
| "step": 182 |
| }, |
| { |
| "epoch": 0.3532818532818533, |
| "grad_norm": 4.703281879425049, |
| "learning_rate": 6.467181467181468e-06, |
| "loss": 762.9774, |
| "step": 183 |
| }, |
| { |
| "epoch": 0.3552123552123552, |
| "grad_norm": 4.766940593719482, |
| "learning_rate": 6.447876447876449e-06, |
| "loss": 757.7225, |
| "step": 184 |
| }, |
| { |
| "epoch": 0.35714285714285715, |
| "grad_norm": 2.995068073272705, |
| "learning_rate": 6.4285714285714295e-06, |
| "loss": 757.5352, |
| "step": 185 |
| }, |
| { |
| "epoch": 0.3590733590733591, |
| "grad_norm": 4.172158241271973, |
| "learning_rate": 6.409266409266411e-06, |
| "loss": 758.1395, |
| "step": 186 |
| }, |
| { |
| "epoch": 0.361003861003861, |
| "grad_norm": 3.057084798812866, |
| "learning_rate": 6.389961389961391e-06, |
| "loss": 763.8903, |
| "step": 187 |
| }, |
| { |
| "epoch": 0.36293436293436293, |
| "grad_norm": 3.4208884239196777, |
| "learning_rate": 6.370656370656371e-06, |
| "loss": 755.9663, |
| "step": 188 |
| }, |
| { |
| "epoch": 0.36486486486486486, |
| "grad_norm": 3.5881717205047607, |
| "learning_rate": 6.351351351351351e-06, |
| "loss": 764.6107, |
| "step": 189 |
| }, |
| { |
| "epoch": 0.3667953667953668, |
| "grad_norm": 4.713850498199463, |
| "learning_rate": 6.332046332046332e-06, |
| "loss": 748.8102, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.3687258687258687, |
| "grad_norm": 3.244720935821533, |
| "learning_rate": 6.312741312741313e-06, |
| "loss": 764.7598, |
| "step": 191 |
| }, |
| { |
| "epoch": 0.37065637065637064, |
| "grad_norm": 4.447683334350586, |
| "learning_rate": 6.293436293436294e-06, |
| "loss": 752.6754, |
| "step": 192 |
| }, |
| { |
| "epoch": 0.37258687258687256, |
| "grad_norm": 3.9772603511810303, |
| "learning_rate": 6.274131274131274e-06, |
| "loss": 756.0471, |
| "step": 193 |
| }, |
| { |
| "epoch": 0.3745173745173745, |
| "grad_norm": 4.803740978240967, |
| "learning_rate": 6.254826254826255e-06, |
| "loss": 769.636, |
| "step": 194 |
| }, |
| { |
| "epoch": 0.3764478764478765, |
| "grad_norm": 4.907277584075928, |
| "learning_rate": 6.235521235521236e-06, |
| "loss": 765.9659, |
| "step": 195 |
| }, |
| { |
| "epoch": 0.3783783783783784, |
| "grad_norm": 3.6414902210235596, |
| "learning_rate": 6.2162162162162164e-06, |
| "loss": 760.1011, |
| "step": 196 |
| }, |
| { |
| "epoch": 0.3803088803088803, |
| "grad_norm": 3.6037609577178955, |
| "learning_rate": 6.196911196911197e-06, |
| "loss": 750.7067, |
| "step": 197 |
| }, |
| { |
| "epoch": 0.38223938223938225, |
| "grad_norm": 2.8772242069244385, |
| "learning_rate": 6.177606177606178e-06, |
| "loss": 756.6414, |
| "step": 198 |
| }, |
| { |
| "epoch": 0.3841698841698842, |
| "grad_norm": 5.2583184242248535, |
| "learning_rate": 6.1583011583011585e-06, |
| "loss": 751.6912, |
| "step": 199 |
| }, |
| { |
| "epoch": 0.3861003861003861, |
| "grad_norm": 2.8050296306610107, |
| "learning_rate": 6.13899613899614e-06, |
| "loss": 743.5164, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.38803088803088803, |
| "grad_norm": 3.5677103996276855, |
| "learning_rate": 6.11969111969112e-06, |
| "loss": 764.8164, |
| "step": 201 |
| }, |
| { |
| "epoch": 0.38996138996138996, |
| "grad_norm": 3.1550533771514893, |
| "learning_rate": 6.1003861003861005e-06, |
| "loss": 762.9319, |
| "step": 202 |
| }, |
| { |
| "epoch": 0.3918918918918919, |
| "grad_norm": 2.991600751876831, |
| "learning_rate": 6.081081081081082e-06, |
| "loss": 770.1368, |
| "step": 203 |
| }, |
| { |
| "epoch": 0.3938223938223938, |
| "grad_norm": 2.9870667457580566, |
| "learning_rate": 6.061776061776062e-06, |
| "loss": 751.1348, |
| "step": 204 |
| }, |
| { |
| "epoch": 0.39575289575289574, |
| "grad_norm": 2.980593204498291, |
| "learning_rate": 6.0424710424710434e-06, |
| "loss": 769.3608, |
| "step": 205 |
| }, |
| { |
| "epoch": 0.39768339768339767, |
| "grad_norm": 3.3218741416931152, |
| "learning_rate": 6.023166023166024e-06, |
| "loss": 763.5135, |
| "step": 206 |
| }, |
| { |
| "epoch": 0.3996138996138996, |
| "grad_norm": 3.1396753787994385, |
| "learning_rate": 6.003861003861004e-06, |
| "loss": 757.1105, |
| "step": 207 |
| }, |
| { |
| "epoch": 0.4015444015444015, |
| "grad_norm": 3.2232918739318848, |
| "learning_rate": 5.9845559845559855e-06, |
| "loss": 743.5112, |
| "step": 208 |
| }, |
| { |
| "epoch": 0.4034749034749035, |
| "grad_norm": 3.3327605724334717, |
| "learning_rate": 5.965250965250966e-06, |
| "loss": 766.4379, |
| "step": 209 |
| }, |
| { |
| "epoch": 0.40540540540540543, |
| "grad_norm": 3.4697072505950928, |
| "learning_rate": 5.945945945945947e-06, |
| "loss": 767.7292, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.40733590733590735, |
| "grad_norm": 4.161364555358887, |
| "learning_rate": 5.9266409266409275e-06, |
| "loss": 746.888, |
| "step": 211 |
| }, |
| { |
| "epoch": 0.4092664092664093, |
| "grad_norm": 3.0299108028411865, |
| "learning_rate": 5.907335907335908e-06, |
| "loss": 773.0205, |
| "step": 212 |
| }, |
| { |
| "epoch": 0.4111969111969112, |
| "grad_norm": 3.702420949935913, |
| "learning_rate": 5.888030888030889e-06, |
| "loss": 762.8275, |
| "step": 213 |
| }, |
| { |
| "epoch": 0.41312741312741313, |
| "grad_norm": 4.044224262237549, |
| "learning_rate": 5.86872586872587e-06, |
| "loss": 762.0264, |
| "step": 214 |
| }, |
| { |
| "epoch": 0.41505791505791506, |
| "grad_norm": 3.09975266456604, |
| "learning_rate": 5.84942084942085e-06, |
| "loss": 738.4187, |
| "step": 215 |
| }, |
| { |
| "epoch": 0.416988416988417, |
| "grad_norm": 3.779367208480835, |
| "learning_rate": 5.83011583011583e-06, |
| "loss": 764.1588, |
| "step": 216 |
| }, |
| { |
| "epoch": 0.4189189189189189, |
| "grad_norm": 3.9344935417175293, |
| "learning_rate": 5.810810810810811e-06, |
| "loss": 751.987, |
| "step": 217 |
| }, |
| { |
| "epoch": 0.42084942084942084, |
| "grad_norm": 3.073119878768921, |
| "learning_rate": 5.791505791505791e-06, |
| "loss": 763.5579, |
| "step": 218 |
| }, |
| { |
| "epoch": 0.42277992277992277, |
| "grad_norm": 5.221838474273682, |
| "learning_rate": 5.7722007722007725e-06, |
| "loss": 736.679, |
| "step": 219 |
| }, |
| { |
| "epoch": 0.4247104247104247, |
| "grad_norm": 3.210583209991455, |
| "learning_rate": 5.752895752895753e-06, |
| "loss": 743.9155, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.4266409266409266, |
| "grad_norm": 3.032778739929199, |
| "learning_rate": 5.733590733590734e-06, |
| "loss": 756.9208, |
| "step": 221 |
| }, |
| { |
| "epoch": 0.42857142857142855, |
| "grad_norm": 3.052091598510742, |
| "learning_rate": 5.7142857142857145e-06, |
| "loss": 753.4189, |
| "step": 222 |
| }, |
| { |
| "epoch": 0.4305019305019305, |
| "grad_norm": 2.934615135192871, |
| "learning_rate": 5.694980694980695e-06, |
| "loss": 744.2302, |
| "step": 223 |
| }, |
| { |
| "epoch": 0.43243243243243246, |
| "grad_norm": 3.453640937805176, |
| "learning_rate": 5.675675675675676e-06, |
| "loss": 753.5289, |
| "step": 224 |
| }, |
| { |
| "epoch": 0.4343629343629344, |
| "grad_norm": 4.493204116821289, |
| "learning_rate": 5.6563706563706566e-06, |
| "loss": 773.2119, |
| "step": 225 |
| }, |
| { |
| "epoch": 0.4362934362934363, |
| "grad_norm": 3.1990344524383545, |
| "learning_rate": 5.637065637065637e-06, |
| "loss": 771.2296, |
| "step": 226 |
| }, |
| { |
| "epoch": 0.43822393822393824, |
| "grad_norm": 4.990328788757324, |
| "learning_rate": 5.617760617760618e-06, |
| "loss": 766.2954, |
| "step": 227 |
| }, |
| { |
| "epoch": 0.44015444015444016, |
| "grad_norm": 6.218008518218994, |
| "learning_rate": 5.598455598455599e-06, |
| "loss": 754.2188, |
| "step": 228 |
| }, |
| { |
| "epoch": 0.4420849420849421, |
| "grad_norm": 4.668118476867676, |
| "learning_rate": 5.57915057915058e-06, |
| "loss": 764.908, |
| "step": 229 |
| }, |
| { |
| "epoch": 0.444015444015444, |
| "grad_norm": 3.144949436187744, |
| "learning_rate": 5.55984555984556e-06, |
| "loss": 769.907, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.44594594594594594, |
| "grad_norm": 3.761617422103882, |
| "learning_rate": 5.540540540540541e-06, |
| "loss": 762.5698, |
| "step": 231 |
| }, |
| { |
| "epoch": 0.44787644787644787, |
| "grad_norm": 3.3352720737457275, |
| "learning_rate": 5.521235521235522e-06, |
| "loss": 760.8379, |
| "step": 232 |
| }, |
| { |
| "epoch": 0.4498069498069498, |
| "grad_norm": 3.646754741668701, |
| "learning_rate": 5.501930501930502e-06, |
| "loss": 762.5335, |
| "step": 233 |
| }, |
| { |
| "epoch": 0.4517374517374517, |
| "grad_norm": 5.615658283233643, |
| "learning_rate": 5.4826254826254836e-06, |
| "loss": 763.7914, |
| "step": 234 |
| }, |
| { |
| "epoch": 0.45366795366795365, |
| "grad_norm": 3.1293540000915527, |
| "learning_rate": 5.463320463320464e-06, |
| "loss": 765.2655, |
| "step": 235 |
| }, |
| { |
| "epoch": 0.4555984555984556, |
| "grad_norm": 3.9679007530212402, |
| "learning_rate": 5.444015444015444e-06, |
| "loss": 776.4496, |
| "step": 236 |
| }, |
| { |
| "epoch": 0.4575289575289575, |
| "grad_norm": 4.110316753387451, |
| "learning_rate": 5.424710424710426e-06, |
| "loss": 769.2201, |
| "step": 237 |
| }, |
| { |
| "epoch": 0.4594594594594595, |
| "grad_norm": 3.5175106525421143, |
| "learning_rate": 5.405405405405406e-06, |
| "loss": 759.53, |
| "step": 238 |
| }, |
| { |
| "epoch": 0.4613899613899614, |
| "grad_norm": 3.206360101699829, |
| "learning_rate": 5.386100386100387e-06, |
| "loss": 753.7003, |
| "step": 239 |
| }, |
| { |
| "epoch": 0.46332046332046334, |
| "grad_norm": 2.994739294052124, |
| "learning_rate": 5.366795366795368e-06, |
| "loss": 750.8613, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.46525096525096526, |
| "grad_norm": 2.966420888900757, |
| "learning_rate": 5.347490347490348e-06, |
| "loss": 755.6821, |
| "step": 241 |
| }, |
| { |
| "epoch": 0.4671814671814672, |
| "grad_norm": 5.035081386566162, |
| "learning_rate": 5.328185328185329e-06, |
| "loss": 753.4547, |
| "step": 242 |
| }, |
| { |
| "epoch": 0.4691119691119691, |
| "grad_norm": 3.399747371673584, |
| "learning_rate": 5.30888030888031e-06, |
| "loss": 752.4637, |
| "step": 243 |
| }, |
| { |
| "epoch": 0.47104247104247104, |
| "grad_norm": 2.9818832874298096, |
| "learning_rate": 5.28957528957529e-06, |
| "loss": 761.8242, |
| "step": 244 |
| }, |
| { |
| "epoch": 0.47297297297297297, |
| "grad_norm": 3.317462921142578, |
| "learning_rate": 5.2702702702702705e-06, |
| "loss": 741.8611, |
| "step": 245 |
| }, |
| { |
| "epoch": 0.4749034749034749, |
| "grad_norm": 3.9936470985412598, |
| "learning_rate": 5.250965250965251e-06, |
| "loss": 755.0592, |
| "step": 246 |
| }, |
| { |
| "epoch": 0.4768339768339768, |
| "grad_norm": 3.263794422149658, |
| "learning_rate": 5.231660231660231e-06, |
| "loss": 748.0864, |
| "step": 247 |
| }, |
| { |
| "epoch": 0.47876447876447875, |
| "grad_norm": 3.1874217987060547, |
| "learning_rate": 5.212355212355213e-06, |
| "loss": 751.8619, |
| "step": 248 |
| }, |
| { |
| "epoch": 0.4806949806949807, |
| "grad_norm": 4.754833698272705, |
| "learning_rate": 5.193050193050193e-06, |
| "loss": 763.7399, |
| "step": 249 |
| }, |
| { |
| "epoch": 0.4826254826254826, |
| "grad_norm": 3.21230149269104, |
| "learning_rate": 5.173745173745173e-06, |
| "loss": 757.1412, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.48455598455598453, |
| "grad_norm": 3.2287521362304688, |
| "learning_rate": 5.154440154440155e-06, |
| "loss": 749.6309, |
| "step": 251 |
| }, |
| { |
| "epoch": 0.4864864864864865, |
| "grad_norm": 3.137535572052002, |
| "learning_rate": 5.135135135135135e-06, |
| "loss": 759.4156, |
| "step": 252 |
| }, |
| { |
| "epoch": 0.48841698841698844, |
| "grad_norm": 3.6146438121795654, |
| "learning_rate": 5.115830115830116e-06, |
| "loss": 741.8262, |
| "step": 253 |
| }, |
| { |
| "epoch": 0.49034749034749037, |
| "grad_norm": 2.9531517028808594, |
| "learning_rate": 5.096525096525097e-06, |
| "loss": 763.7745, |
| "step": 254 |
| }, |
| { |
| "epoch": 0.4922779922779923, |
| "grad_norm": 3.0513949394226074, |
| "learning_rate": 5.077220077220077e-06, |
| "loss": 755.271, |
| "step": 255 |
| }, |
| { |
| "epoch": 0.4942084942084942, |
| "grad_norm": 3.2384207248687744, |
| "learning_rate": 5.057915057915058e-06, |
| "loss": 765.3377, |
| "step": 256 |
| }, |
| { |
| "epoch": 0.49613899613899615, |
| "grad_norm": 3.8821310997009277, |
| "learning_rate": 5.038610038610039e-06, |
| "loss": 763.2088, |
| "step": 257 |
| }, |
| { |
| "epoch": 0.4980694980694981, |
| "grad_norm": 2.926225423812866, |
| "learning_rate": 5.01930501930502e-06, |
| "loss": 758.5596, |
| "step": 258 |
| }, |
| { |
| "epoch": 0.5, |
| "grad_norm": 2.9751245975494385, |
| "learning_rate": 5e-06, |
| "loss": 757.9719, |
| "step": 259 |
| }, |
| { |
| "epoch": 0.5019305019305019, |
| "grad_norm": 4.772404193878174, |
| "learning_rate": 4.980694980694981e-06, |
| "loss": 757.3872, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.5038610038610039, |
| "grad_norm": 3.096656322479248, |
| "learning_rate": 4.961389961389962e-06, |
| "loss": 760.4723, |
| "step": 261 |
| }, |
| { |
| "epoch": 0.5057915057915058, |
| "grad_norm": 3.504852056503296, |
| "learning_rate": 4.9420849420849425e-06, |
| "loss": 747.486, |
| "step": 262 |
| }, |
| { |
| "epoch": 0.5077220077220077, |
| "grad_norm": 3.33747935295105, |
| "learning_rate": 4.922779922779924e-06, |
| "loss": 758.9821, |
| "step": 263 |
| }, |
| { |
| "epoch": 0.5096525096525096, |
| "grad_norm": 3.4812850952148438, |
| "learning_rate": 4.903474903474904e-06, |
| "loss": 765.7698, |
| "step": 264 |
| }, |
| { |
| "epoch": 0.5115830115830116, |
| "grad_norm": 3.174971580505371, |
| "learning_rate": 4.8841698841698845e-06, |
| "loss": 756.7213, |
| "step": 265 |
| }, |
| { |
| "epoch": 0.5135135135135135, |
| "grad_norm": 3.601083755493164, |
| "learning_rate": 4.864864864864866e-06, |
| "loss": 762.7404, |
| "step": 266 |
| }, |
| { |
| "epoch": 0.5154440154440154, |
| "grad_norm": 3.3433072566986084, |
| "learning_rate": 4.845559845559845e-06, |
| "loss": 747.8812, |
| "step": 267 |
| }, |
| { |
| "epoch": 0.5173745173745173, |
| "grad_norm": 3.7494962215423584, |
| "learning_rate": 4.8262548262548266e-06, |
| "loss": 756.0776, |
| "step": 268 |
| }, |
| { |
| "epoch": 0.5193050193050193, |
| "grad_norm": 4.609169960021973, |
| "learning_rate": 4.806949806949807e-06, |
| "loss": 750.7801, |
| "step": 269 |
| }, |
| { |
| "epoch": 0.5212355212355212, |
| "grad_norm": 5.737427711486816, |
| "learning_rate": 4.787644787644788e-06, |
| "loss": 751.1343, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.5231660231660231, |
| "grad_norm": 3.3391857147216797, |
| "learning_rate": 4.768339768339769e-06, |
| "loss": 760.2618, |
| "step": 271 |
| }, |
| { |
| "epoch": 0.525096525096525, |
| "grad_norm": 3.1758346557617188, |
| "learning_rate": 4.749034749034749e-06, |
| "loss": 756.4235, |
| "step": 272 |
| }, |
| { |
| "epoch": 0.527027027027027, |
| "grad_norm": 4.515329360961914, |
| "learning_rate": 4.72972972972973e-06, |
| "loss": 770.3212, |
| "step": 273 |
| }, |
| { |
| "epoch": 0.528957528957529, |
| "grad_norm": 3.2932803630828857, |
| "learning_rate": 4.710424710424711e-06, |
| "loss": 751.3934, |
| "step": 274 |
| }, |
| { |
| "epoch": 0.5308880308880309, |
| "grad_norm": 3.047992467880249, |
| "learning_rate": 4.691119691119692e-06, |
| "loss": 749.6434, |
| "step": 275 |
| }, |
| { |
| "epoch": 0.5328185328185329, |
| "grad_norm": 3.2720017433166504, |
| "learning_rate": 4.671814671814672e-06, |
| "loss": 771.3862, |
| "step": 276 |
| }, |
| { |
| "epoch": 0.5347490347490348, |
| "grad_norm": 3.6635942459106445, |
| "learning_rate": 4.652509652509653e-06, |
| "loss": 758.3706, |
| "step": 277 |
| }, |
| { |
| "epoch": 0.5366795366795367, |
| "grad_norm": 3.4579808712005615, |
| "learning_rate": 4.633204633204634e-06, |
| "loss": 747.9475, |
| "step": 278 |
| }, |
| { |
| "epoch": 0.5386100386100386, |
| "grad_norm": 3.525111675262451, |
| "learning_rate": 4.613899613899614e-06, |
| "loss": 756.5895, |
| "step": 279 |
| }, |
| { |
| "epoch": 0.5405405405405406, |
| "grad_norm": 3.336205005645752, |
| "learning_rate": 4.594594594594596e-06, |
| "loss": 772.9059, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.5424710424710425, |
| "grad_norm": 3.758382558822632, |
| "learning_rate": 4.575289575289575e-06, |
| "loss": 769.8763, |
| "step": 281 |
| }, |
| { |
| "epoch": 0.5444015444015444, |
| "grad_norm": 3.054797649383545, |
| "learning_rate": 4.5559845559845564e-06, |
| "loss": 768.0063, |
| "step": 282 |
| }, |
| { |
| "epoch": 0.5463320463320464, |
| "grad_norm": 4.434202671051025, |
| "learning_rate": 4.536679536679537e-06, |
| "loss": 760.1692, |
| "step": 283 |
| }, |
| { |
| "epoch": 0.5482625482625483, |
| "grad_norm": 4.3827667236328125, |
| "learning_rate": 4.517374517374517e-06, |
| "loss": 744.6565, |
| "step": 284 |
| }, |
| { |
| "epoch": 0.5501930501930502, |
| "grad_norm": 6.237070560455322, |
| "learning_rate": 4.4980694980694985e-06, |
| "loss": 767.5054, |
| "step": 285 |
| }, |
| { |
| "epoch": 0.5521235521235521, |
| "grad_norm": 3.162343740463257, |
| "learning_rate": 4.478764478764479e-06, |
| "loss": 762.6232, |
| "step": 286 |
| }, |
| { |
| "epoch": 0.5540540540540541, |
| "grad_norm": 3.9073193073272705, |
| "learning_rate": 4.45945945945946e-06, |
| "loss": 778.0848, |
| "step": 287 |
| }, |
| { |
| "epoch": 0.555984555984556, |
| "grad_norm": 6.03090763092041, |
| "learning_rate": 4.4401544401544405e-06, |
| "loss": 759.6763, |
| "step": 288 |
| }, |
| { |
| "epoch": 0.5579150579150579, |
| "grad_norm": 4.3667311668396, |
| "learning_rate": 4.420849420849421e-06, |
| "loss": 762.0499, |
| "step": 289 |
| }, |
| { |
| "epoch": 0.5598455598455598, |
| "grad_norm": 3.5320377349853516, |
| "learning_rate": 4.401544401544402e-06, |
| "loss": 769.951, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.5617760617760618, |
| "grad_norm": 3.7424259185791016, |
| "learning_rate": 4.382239382239383e-06, |
| "loss": 767.4916, |
| "step": 291 |
| }, |
| { |
| "epoch": 0.5637065637065637, |
| "grad_norm": 3.3271844387054443, |
| "learning_rate": 4.362934362934364e-06, |
| "loss": 761.4987, |
| "step": 292 |
| }, |
| { |
| "epoch": 0.5656370656370656, |
| "grad_norm": 2.957545757293701, |
| "learning_rate": 4.343629343629344e-06, |
| "loss": 762.5405, |
| "step": 293 |
| }, |
| { |
| "epoch": 0.5675675675675675, |
| "grad_norm": 3.087904453277588, |
| "learning_rate": 4.324324324324325e-06, |
| "loss": 752.0869, |
| "step": 294 |
| }, |
| { |
| "epoch": 0.5694980694980695, |
| "grad_norm": 3.4047775268554688, |
| "learning_rate": 4.305019305019305e-06, |
| "loss": 764.2693, |
| "step": 295 |
| }, |
| { |
| "epoch": 0.5714285714285714, |
| "grad_norm": 3.244701385498047, |
| "learning_rate": 4.2857142857142855e-06, |
| "loss": 754.4578, |
| "step": 296 |
| }, |
| { |
| "epoch": 0.5733590733590733, |
| "grad_norm": 3.3166253566741943, |
| "learning_rate": 4.266409266409267e-06, |
| "loss": 773.7292, |
| "step": 297 |
| }, |
| { |
| "epoch": 0.5752895752895753, |
| "grad_norm": 3.0471668243408203, |
| "learning_rate": 4.247104247104247e-06, |
| "loss": 778.6743, |
| "step": 298 |
| }, |
| { |
| "epoch": 0.5772200772200772, |
| "grad_norm": 3.6165595054626465, |
| "learning_rate": 4.227799227799228e-06, |
| "loss": 755.6964, |
| "step": 299 |
| }, |
| { |
| "epoch": 0.5791505791505791, |
| "grad_norm": 3.426254987716675, |
| "learning_rate": 4.208494208494209e-06, |
| "loss": 768.4984, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.581081081081081, |
| "grad_norm": 4.2183518409729, |
| "learning_rate": 4.189189189189189e-06, |
| "loss": 756.4444, |
| "step": 301 |
| }, |
| { |
| "epoch": 0.583011583011583, |
| "grad_norm": 4.252375602722168, |
| "learning_rate": 4.16988416988417e-06, |
| "loss": 760.7186, |
| "step": 302 |
| }, |
| { |
| "epoch": 0.584942084942085, |
| "grad_norm": 3.898979425430298, |
| "learning_rate": 4.150579150579151e-06, |
| "loss": 755.8693, |
| "step": 303 |
| }, |
| { |
| "epoch": 0.5868725868725869, |
| "grad_norm": 4.586599349975586, |
| "learning_rate": 4.131274131274132e-06, |
| "loss": 767.7788, |
| "step": 304 |
| }, |
| { |
| "epoch": 0.5888030888030888, |
| "grad_norm": 3.4545576572418213, |
| "learning_rate": 4.1119691119691125e-06, |
| "loss": 755.6498, |
| "step": 305 |
| }, |
| { |
| "epoch": 0.5907335907335908, |
| "grad_norm": 3.056624174118042, |
| "learning_rate": 4.092664092664093e-06, |
| "loss": 779.4406, |
| "step": 306 |
| }, |
| { |
| "epoch": 0.5926640926640927, |
| "grad_norm": 3.1570804119110107, |
| "learning_rate": 4.073359073359074e-06, |
| "loss": 756.1407, |
| "step": 307 |
| }, |
| { |
| "epoch": 0.5945945945945946, |
| "grad_norm": 5.404672622680664, |
| "learning_rate": 4.0540540540540545e-06, |
| "loss": 770.681, |
| "step": 308 |
| }, |
| { |
| "epoch": 0.5965250965250966, |
| "grad_norm": 3.0374257564544678, |
| "learning_rate": 4.034749034749035e-06, |
| "loss": 759.7466, |
| "step": 309 |
| }, |
| { |
| "epoch": 0.5984555984555985, |
| "grad_norm": 3.2685699462890625, |
| "learning_rate": 4.015444015444015e-06, |
| "loss": 762.7603, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.6003861003861004, |
| "grad_norm": 7.29334831237793, |
| "learning_rate": 3.996138996138997e-06, |
| "loss": 742.7489, |
| "step": 311 |
| }, |
| { |
| "epoch": 0.6023166023166023, |
| "grad_norm": 3.852389335632324, |
| "learning_rate": 3.976833976833977e-06, |
| "loss": 767.3596, |
| "step": 312 |
| }, |
| { |
| "epoch": 0.6042471042471043, |
| "grad_norm": 3.200193166732788, |
| "learning_rate": 3.957528957528957e-06, |
| "loss": 766.8765, |
| "step": 313 |
| }, |
| { |
| "epoch": 0.6061776061776062, |
| "grad_norm": 3.5071451663970947, |
| "learning_rate": 3.938223938223939e-06, |
| "loss": 776.7089, |
| "step": 314 |
| }, |
| { |
| "epoch": 0.6081081081081081, |
| "grad_norm": 4.16587495803833, |
| "learning_rate": 3.918918918918919e-06, |
| "loss": 741.2016, |
| "step": 315 |
| }, |
| { |
| "epoch": 0.61003861003861, |
| "grad_norm": 5.703463554382324, |
| "learning_rate": 3.8996138996139e-06, |
| "loss": 756.0547, |
| "step": 316 |
| }, |
| { |
| "epoch": 0.611969111969112, |
| "grad_norm": 5.430912017822266, |
| "learning_rate": 3.880308880308881e-06, |
| "loss": 756.722, |
| "step": 317 |
| }, |
| { |
| "epoch": 0.6138996138996139, |
| "grad_norm": 8.650154113769531, |
| "learning_rate": 3.861003861003861e-06, |
| "loss": 745.7667, |
| "step": 318 |
| }, |
| { |
| "epoch": 0.6158301158301158, |
| "grad_norm": 13.80824089050293, |
| "learning_rate": 3.841698841698842e-06, |
| "loss": 755.2433, |
| "step": 319 |
| }, |
| { |
| "epoch": 0.6177606177606177, |
| "grad_norm": 7.08932638168335, |
| "learning_rate": 3.822393822393823e-06, |
| "loss": 753.7036, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.6196911196911197, |
| "grad_norm": 3.6433534622192383, |
| "learning_rate": 3.8030888030888036e-06, |
| "loss": 773.3873, |
| "step": 321 |
| }, |
| { |
| "epoch": 0.6216216216216216, |
| "grad_norm": 3.2302098274230957, |
| "learning_rate": 3.7837837837837844e-06, |
| "loss": 758.14, |
| "step": 322 |
| }, |
| { |
| "epoch": 0.6235521235521235, |
| "grad_norm": 3.731865644454956, |
| "learning_rate": 3.7644787644787652e-06, |
| "loss": 766.5978, |
| "step": 323 |
| }, |
| { |
| "epoch": 0.6254826254826255, |
| "grad_norm": 3.244114875793457, |
| "learning_rate": 3.745173745173745e-06, |
| "loss": 748.5593, |
| "step": 324 |
| }, |
| { |
| "epoch": 0.6274131274131274, |
| "grad_norm": 6.71386194229126, |
| "learning_rate": 3.725868725868726e-06, |
| "loss": 750.8797, |
| "step": 325 |
| }, |
| { |
| "epoch": 0.6293436293436293, |
| "grad_norm": 6.857434272766113, |
| "learning_rate": 3.706563706563707e-06, |
| "loss": 751.0732, |
| "step": 326 |
| }, |
| { |
| "epoch": 0.6312741312741312, |
| "grad_norm": 3.1057608127593994, |
| "learning_rate": 3.6872586872586872e-06, |
| "loss": 751.9279, |
| "step": 327 |
| }, |
| { |
| "epoch": 0.6332046332046332, |
| "grad_norm": 4.462062835693359, |
| "learning_rate": 3.667953667953668e-06, |
| "loss": 761.2999, |
| "step": 328 |
| }, |
| { |
| "epoch": 0.6351351351351351, |
| "grad_norm": 5.394342422485352, |
| "learning_rate": 3.648648648648649e-06, |
| "loss": 765.5782, |
| "step": 329 |
| }, |
| { |
| "epoch": 0.637065637065637, |
| "grad_norm": 4.586513996124268, |
| "learning_rate": 3.6293436293436297e-06, |
| "loss": 759.7944, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.638996138996139, |
| "grad_norm": 6.204144477844238, |
| "learning_rate": 3.61003861003861e-06, |
| "loss": 749.0181, |
| "step": 331 |
| }, |
| { |
| "epoch": 0.640926640926641, |
| "grad_norm": 3.065887451171875, |
| "learning_rate": 3.590733590733591e-06, |
| "loss": 764.4531, |
| "step": 332 |
| }, |
| { |
| "epoch": 0.6428571428571429, |
| "grad_norm": 3.5210766792297363, |
| "learning_rate": 3.5714285714285718e-06, |
| "loss": 757.1459, |
| "step": 333 |
| }, |
| { |
| "epoch": 0.6447876447876448, |
| "grad_norm": 4.145437240600586, |
| "learning_rate": 3.5521235521235526e-06, |
| "loss": 772.7172, |
| "step": 334 |
| }, |
| { |
| "epoch": 0.6467181467181468, |
| "grad_norm": 3.4393627643585205, |
| "learning_rate": 3.5328185328185334e-06, |
| "loss": 748.3174, |
| "step": 335 |
| }, |
| { |
| "epoch": 0.6486486486486487, |
| "grad_norm": 6.470264911651611, |
| "learning_rate": 3.513513513513514e-06, |
| "loss": 761.2096, |
| "step": 336 |
| }, |
| { |
| "epoch": 0.6505791505791506, |
| "grad_norm": 3.4703409671783447, |
| "learning_rate": 3.4942084942084947e-06, |
| "loss": 752.6991, |
| "step": 337 |
| }, |
| { |
| "epoch": 0.6525096525096525, |
| "grad_norm": 9.870447158813477, |
| "learning_rate": 3.4749034749034755e-06, |
| "loss": 771.034, |
| "step": 338 |
| }, |
| { |
| "epoch": 0.6544401544401545, |
| "grad_norm": 9.164502143859863, |
| "learning_rate": 3.4555984555984555e-06, |
| "loss": 759.8434, |
| "step": 339 |
| }, |
| { |
| "epoch": 0.6563706563706564, |
| "grad_norm": 8.678630828857422, |
| "learning_rate": 3.4362934362934363e-06, |
| "loss": 760.7397, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.6583011583011583, |
| "grad_norm": 4.10188102722168, |
| "learning_rate": 3.416988416988417e-06, |
| "loss": 752.9614, |
| "step": 341 |
| }, |
| { |
| "epoch": 0.6602316602316602, |
| "grad_norm": 5.000187873840332, |
| "learning_rate": 3.397683397683398e-06, |
| "loss": 741.5472, |
| "step": 342 |
| }, |
| { |
| "epoch": 0.6621621621621622, |
| "grad_norm": 5.491066932678223, |
| "learning_rate": 3.3783783783783788e-06, |
| "loss": 752.3487, |
| "step": 343 |
| }, |
| { |
| "epoch": 0.6640926640926641, |
| "grad_norm": 4.743298530578613, |
| "learning_rate": 3.359073359073359e-06, |
| "loss": 740.1313, |
| "step": 344 |
| }, |
| { |
| "epoch": 0.666023166023166, |
| "grad_norm": 3.8254621028900146, |
| "learning_rate": 3.33976833976834e-06, |
| "loss": 751.6027, |
| "step": 345 |
| }, |
| { |
| "epoch": 0.667953667953668, |
| "grad_norm": 3.4094951152801514, |
| "learning_rate": 3.320463320463321e-06, |
| "loss": 747.6785, |
| "step": 346 |
| }, |
| { |
| "epoch": 0.6698841698841699, |
| "grad_norm": 4.3971757888793945, |
| "learning_rate": 3.3011583011583016e-06, |
| "loss": 765.18, |
| "step": 347 |
| }, |
| { |
| "epoch": 0.6718146718146718, |
| "grad_norm": 5.712260723114014, |
| "learning_rate": 3.281853281853282e-06, |
| "loss": 754.3199, |
| "step": 348 |
| }, |
| { |
| "epoch": 0.6737451737451737, |
| "grad_norm": 2.944014310836792, |
| "learning_rate": 3.262548262548263e-06, |
| "loss": 759.125, |
| "step": 349 |
| }, |
| { |
| "epoch": 0.6756756756756757, |
| "grad_norm": 5.154186725616455, |
| "learning_rate": 3.2432432432432437e-06, |
| "loss": 773.4568, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.6776061776061776, |
| "grad_norm": 4.158810138702393, |
| "learning_rate": 3.2239382239382245e-06, |
| "loss": 773.0182, |
| "step": 351 |
| }, |
| { |
| "epoch": 0.6795366795366795, |
| "grad_norm": 2.9692142009735107, |
| "learning_rate": 3.2046332046332054e-06, |
| "loss": 759.3475, |
| "step": 352 |
| }, |
| { |
| "epoch": 0.6814671814671814, |
| "grad_norm": 5.837861061096191, |
| "learning_rate": 3.1853281853281853e-06, |
| "loss": 758.8844, |
| "step": 353 |
| }, |
| { |
| "epoch": 0.6833976833976834, |
| "grad_norm": 4.23380708694458, |
| "learning_rate": 3.166023166023166e-06, |
| "loss": 762.8468, |
| "step": 354 |
| }, |
| { |
| "epoch": 0.6853281853281853, |
| "grad_norm": 5.026414394378662, |
| "learning_rate": 3.146718146718147e-06, |
| "loss": 738.6548, |
| "step": 355 |
| }, |
| { |
| "epoch": 0.6872586872586872, |
| "grad_norm": 4.041588306427002, |
| "learning_rate": 3.1274131274131274e-06, |
| "loss": 766.1196, |
| "step": 356 |
| }, |
| { |
| "epoch": 0.6891891891891891, |
| "grad_norm": 4.39390230178833, |
| "learning_rate": 3.1081081081081082e-06, |
| "loss": 749.96, |
| "step": 357 |
| }, |
| { |
| "epoch": 0.6911196911196911, |
| "grad_norm": 4.870336055755615, |
| "learning_rate": 3.088803088803089e-06, |
| "loss": 750.3604, |
| "step": 358 |
| }, |
| { |
| "epoch": 0.693050193050193, |
| "grad_norm": 6.5269036293029785, |
| "learning_rate": 3.06949806949807e-06, |
| "loss": 765.9295, |
| "step": 359 |
| }, |
| { |
| "epoch": 0.694980694980695, |
| "grad_norm": 3.1647534370422363, |
| "learning_rate": 3.0501930501930503e-06, |
| "loss": 747.9913, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.696911196911197, |
| "grad_norm": 3.435527801513672, |
| "learning_rate": 3.030888030888031e-06, |
| "loss": 770.7856, |
| "step": 361 |
| }, |
| { |
| "epoch": 0.6988416988416989, |
| "grad_norm": 5.05522346496582, |
| "learning_rate": 3.011583011583012e-06, |
| "loss": 773.0812, |
| "step": 362 |
| }, |
| { |
| "epoch": 0.7007722007722008, |
| "grad_norm": 4.161330223083496, |
| "learning_rate": 2.9922779922779927e-06, |
| "loss": 771.5272, |
| "step": 363 |
| }, |
| { |
| "epoch": 0.7027027027027027, |
| "grad_norm": 4.4345526695251465, |
| "learning_rate": 2.9729729729729736e-06, |
| "loss": 754.7959, |
| "step": 364 |
| }, |
| { |
| "epoch": 0.7046332046332047, |
| "grad_norm": 2.9777281284332275, |
| "learning_rate": 2.953667953667954e-06, |
| "loss": 749.3711, |
| "step": 365 |
| }, |
| { |
| "epoch": 0.7065637065637066, |
| "grad_norm": 2.855987310409546, |
| "learning_rate": 2.934362934362935e-06, |
| "loss": 749.5988, |
| "step": 366 |
| }, |
| { |
| "epoch": 0.7084942084942085, |
| "grad_norm": 3.307720184326172, |
| "learning_rate": 2.915057915057915e-06, |
| "loss": 761.0093, |
| "step": 367 |
| }, |
| { |
| "epoch": 0.7104247104247104, |
| "grad_norm": 3.787365198135376, |
| "learning_rate": 2.8957528957528956e-06, |
| "loss": 751.6588, |
| "step": 368 |
| }, |
| { |
| "epoch": 0.7123552123552124, |
| "grad_norm": 3.029486656188965, |
| "learning_rate": 2.8764478764478764e-06, |
| "loss": 755.2015, |
| "step": 369 |
| }, |
| { |
| "epoch": 0.7142857142857143, |
| "grad_norm": 3.0125091075897217, |
| "learning_rate": 2.8571428571428573e-06, |
| "loss": 766.543, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.7162162162162162, |
| "grad_norm": 4.560346603393555, |
| "learning_rate": 2.837837837837838e-06, |
| "loss": 768.441, |
| "step": 371 |
| }, |
| { |
| "epoch": 0.7181467181467182, |
| "grad_norm": 3.190322160720825, |
| "learning_rate": 2.8185328185328185e-06, |
| "loss": 753.3053, |
| "step": 372 |
| }, |
| { |
| "epoch": 0.7200772200772201, |
| "grad_norm": 3.4040372371673584, |
| "learning_rate": 2.7992277992277993e-06, |
| "loss": 769.1228, |
| "step": 373 |
| }, |
| { |
| "epoch": 0.722007722007722, |
| "grad_norm": 5.434439182281494, |
| "learning_rate": 2.77992277992278e-06, |
| "loss": 754.5753, |
| "step": 374 |
| }, |
| { |
| "epoch": 0.7239382239382239, |
| "grad_norm": 3.3807380199432373, |
| "learning_rate": 2.760617760617761e-06, |
| "loss": 768.5261, |
| "step": 375 |
| }, |
| { |
| "epoch": 0.7258687258687259, |
| "grad_norm": 3.2653305530548096, |
| "learning_rate": 2.7413127413127418e-06, |
| "loss": 768.1964, |
| "step": 376 |
| }, |
| { |
| "epoch": 0.7277992277992278, |
| "grad_norm": 3.041628360748291, |
| "learning_rate": 2.722007722007722e-06, |
| "loss": 771.9323, |
| "step": 377 |
| }, |
| { |
| "epoch": 0.7297297297297297, |
| "grad_norm": 3.092245101928711, |
| "learning_rate": 2.702702702702703e-06, |
| "loss": 753.942, |
| "step": 378 |
| }, |
| { |
| "epoch": 0.7316602316602316, |
| "grad_norm": 4.467896938323975, |
| "learning_rate": 2.683397683397684e-06, |
| "loss": 749.8477, |
| "step": 379 |
| }, |
| { |
| "epoch": 0.7335907335907336, |
| "grad_norm": 2.989520788192749, |
| "learning_rate": 2.6640926640926647e-06, |
| "loss": 759.7245, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.7355212355212355, |
| "grad_norm": 2.9288792610168457, |
| "learning_rate": 2.644787644787645e-06, |
| "loss": 761.0784, |
| "step": 381 |
| }, |
| { |
| "epoch": 0.7374517374517374, |
| "grad_norm": 3.411576986312866, |
| "learning_rate": 2.6254826254826255e-06, |
| "loss": 769.0735, |
| "step": 382 |
| }, |
| { |
| "epoch": 0.7393822393822393, |
| "grad_norm": 7.500954627990723, |
| "learning_rate": 2.6061776061776063e-06, |
| "loss": 752.552, |
| "step": 383 |
| }, |
| { |
| "epoch": 0.7413127413127413, |
| "grad_norm": 4.436896324157715, |
| "learning_rate": 2.5868725868725867e-06, |
| "loss": 772.7829, |
| "step": 384 |
| }, |
| { |
| "epoch": 0.7432432432432432, |
| "grad_norm": 3.8016252517700195, |
| "learning_rate": 2.5675675675675675e-06, |
| "loss": 756.5428, |
| "step": 385 |
| }, |
| { |
| "epoch": 0.7451737451737451, |
| "grad_norm": 3.1768035888671875, |
| "learning_rate": 2.5482625482625484e-06, |
| "loss": 761.3637, |
| "step": 386 |
| }, |
| { |
| "epoch": 0.747104247104247, |
| "grad_norm": 3.24792218208313, |
| "learning_rate": 2.528957528957529e-06, |
| "loss": 764.2406, |
| "step": 387 |
| }, |
| { |
| "epoch": 0.749034749034749, |
| "grad_norm": 3.6789443492889404, |
| "learning_rate": 2.50965250965251e-06, |
| "loss": 764.6515, |
| "step": 388 |
| }, |
| { |
| "epoch": 0.750965250965251, |
| "grad_norm": 2.998337507247925, |
| "learning_rate": 2.4903474903474904e-06, |
| "loss": 745.9545, |
| "step": 389 |
| }, |
| { |
| "epoch": 0.752895752895753, |
| "grad_norm": 5.5688796043396, |
| "learning_rate": 2.4710424710424712e-06, |
| "loss": 757.1937, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.7548262548262549, |
| "grad_norm": 3.491425037384033, |
| "learning_rate": 2.451737451737452e-06, |
| "loss": 747.8929, |
| "step": 391 |
| }, |
| { |
| "epoch": 0.7567567567567568, |
| "grad_norm": 5.370115280151367, |
| "learning_rate": 2.432432432432433e-06, |
| "loss": 760.0563, |
| "step": 392 |
| }, |
| { |
| "epoch": 0.7586872586872587, |
| "grad_norm": 3.907552480697632, |
| "learning_rate": 2.4131274131274133e-06, |
| "loss": 752.7516, |
| "step": 393 |
| }, |
| { |
| "epoch": 0.7606177606177607, |
| "grad_norm": 3.052731513977051, |
| "learning_rate": 2.393822393822394e-06, |
| "loss": 752.1655, |
| "step": 394 |
| }, |
| { |
| "epoch": 0.7625482625482626, |
| "grad_norm": 5.308294296264648, |
| "learning_rate": 2.3745173745173745e-06, |
| "loss": 767.8257, |
| "step": 395 |
| }, |
| { |
| "epoch": 0.7644787644787645, |
| "grad_norm": 3.72119140625, |
| "learning_rate": 2.3552123552123553e-06, |
| "loss": 764.8139, |
| "step": 396 |
| }, |
| { |
| "epoch": 0.7664092664092664, |
| "grad_norm": 3.117110252380371, |
| "learning_rate": 2.335907335907336e-06, |
| "loss": 766.5566, |
| "step": 397 |
| }, |
| { |
| "epoch": 0.7683397683397684, |
| "grad_norm": 3.8100202083587646, |
| "learning_rate": 2.316602316602317e-06, |
| "loss": 767.1085, |
| "step": 398 |
| }, |
| { |
| "epoch": 0.7702702702702703, |
| "grad_norm": 3.1219418048858643, |
| "learning_rate": 2.297297297297298e-06, |
| "loss": 750.5438, |
| "step": 399 |
| }, |
| { |
| "epoch": 0.7722007722007722, |
| "grad_norm": 3.212282180786133, |
| "learning_rate": 2.2779922779922782e-06, |
| "loss": 748.5366, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.7741312741312741, |
| "grad_norm": 3.0505459308624268, |
| "learning_rate": 2.2586872586872586e-06, |
| "loss": 743.7197, |
| "step": 401 |
| }, |
| { |
| "epoch": 0.7760617760617761, |
| "grad_norm": 3.131042718887329, |
| "learning_rate": 2.2393822393822394e-06, |
| "loss": 756.7755, |
| "step": 402 |
| }, |
| { |
| "epoch": 0.777992277992278, |
| "grad_norm": 4.418597221374512, |
| "learning_rate": 2.2200772200772203e-06, |
| "loss": 754.0249, |
| "step": 403 |
| }, |
| { |
| "epoch": 0.7799227799227799, |
| "grad_norm": 4.271742820739746, |
| "learning_rate": 2.200772200772201e-06, |
| "loss": 768.3597, |
| "step": 404 |
| }, |
| { |
| "epoch": 0.7818532818532818, |
| "grad_norm": 3.788788080215454, |
| "learning_rate": 2.181467181467182e-06, |
| "loss": 758.9531, |
| "step": 405 |
| }, |
| { |
| "epoch": 0.7837837837837838, |
| "grad_norm": 3.5197062492370605, |
| "learning_rate": 2.1621621621621623e-06, |
| "loss": 754.1716, |
| "step": 406 |
| }, |
| { |
| "epoch": 0.7857142857142857, |
| "grad_norm": 3.414795398712158, |
| "learning_rate": 2.1428571428571427e-06, |
| "loss": 751.0524, |
| "step": 407 |
| }, |
| { |
| "epoch": 0.7876447876447876, |
| "grad_norm": 3.1258320808410645, |
| "learning_rate": 2.1235521235521236e-06, |
| "loss": 761.6645, |
| "step": 408 |
| }, |
| { |
| "epoch": 0.7895752895752896, |
| "grad_norm": 6.547064781188965, |
| "learning_rate": 2.1042471042471044e-06, |
| "loss": 765.9225, |
| "step": 409 |
| }, |
| { |
| "epoch": 0.7915057915057915, |
| "grad_norm": 4.772284030914307, |
| "learning_rate": 2.084942084942085e-06, |
| "loss": 764.9897, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.7934362934362934, |
| "grad_norm": 4.287495136260986, |
| "learning_rate": 2.065637065637066e-06, |
| "loss": 756.7947, |
| "step": 411 |
| }, |
| { |
| "epoch": 0.7953667953667953, |
| "grad_norm": 5.286579132080078, |
| "learning_rate": 2.0463320463320464e-06, |
| "loss": 756.2988, |
| "step": 412 |
| }, |
| { |
| "epoch": 0.7972972972972973, |
| "grad_norm": 3.132582902908325, |
| "learning_rate": 2.0270270270270273e-06, |
| "loss": 760.6235, |
| "step": 413 |
| }, |
| { |
| "epoch": 0.7992277992277992, |
| "grad_norm": 4.088491439819336, |
| "learning_rate": 2.0077220077220077e-06, |
| "loss": 751.6579, |
| "step": 414 |
| }, |
| { |
| "epoch": 0.8011583011583011, |
| "grad_norm": 3.853825807571411, |
| "learning_rate": 1.9884169884169885e-06, |
| "loss": 751.7198, |
| "step": 415 |
| }, |
| { |
| "epoch": 0.803088803088803, |
| "grad_norm": 3.8805510997772217, |
| "learning_rate": 1.9691119691119693e-06, |
| "loss": 754.175, |
| "step": 416 |
| }, |
| { |
| "epoch": 0.805019305019305, |
| "grad_norm": 3.465831756591797, |
| "learning_rate": 1.94980694980695e-06, |
| "loss": 763.5966, |
| "step": 417 |
| }, |
| { |
| "epoch": 0.806949806949807, |
| "grad_norm": 3.4799764156341553, |
| "learning_rate": 1.9305019305019305e-06, |
| "loss": 757.6487, |
| "step": 418 |
| }, |
| { |
| "epoch": 0.8088803088803089, |
| "grad_norm": 4.306357383728027, |
| "learning_rate": 1.9111969111969114e-06, |
| "loss": 770.5115, |
| "step": 419 |
| }, |
| { |
| "epoch": 0.8108108108108109, |
| "grad_norm": 5.999144554138184, |
| "learning_rate": 1.8918918918918922e-06, |
| "loss": 750.0549, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.8127413127413128, |
| "grad_norm": 3.353970527648926, |
| "learning_rate": 1.8725868725868726e-06, |
| "loss": 770.1407, |
| "step": 421 |
| }, |
| { |
| "epoch": 0.8146718146718147, |
| "grad_norm": 3.1123435497283936, |
| "learning_rate": 1.8532818532818534e-06, |
| "loss": 755.7714, |
| "step": 422 |
| }, |
| { |
| "epoch": 0.8166023166023166, |
| "grad_norm": 3.5919506549835205, |
| "learning_rate": 1.833976833976834e-06, |
| "loss": 768.7866, |
| "step": 423 |
| }, |
| { |
| "epoch": 0.8185328185328186, |
| "grad_norm": 5.815296649932861, |
| "learning_rate": 1.8146718146718149e-06, |
| "loss": 769.049, |
| "step": 424 |
| }, |
| { |
| "epoch": 0.8204633204633205, |
| "grad_norm": 6.508057117462158, |
| "learning_rate": 1.7953667953667955e-06, |
| "loss": 744.9335, |
| "step": 425 |
| }, |
| { |
| "epoch": 0.8223938223938224, |
| "grad_norm": 3.8051586151123047, |
| "learning_rate": 1.7760617760617763e-06, |
| "loss": 755.9141, |
| "step": 426 |
| }, |
| { |
| "epoch": 0.8243243243243243, |
| "grad_norm": 3.628269910812378, |
| "learning_rate": 1.756756756756757e-06, |
| "loss": 758.3513, |
| "step": 427 |
| }, |
| { |
| "epoch": 0.8262548262548263, |
| "grad_norm": 5.430158615112305, |
| "learning_rate": 1.7374517374517377e-06, |
| "loss": 754.9781, |
| "step": 428 |
| }, |
| { |
| "epoch": 0.8281853281853282, |
| "grad_norm": 3.8903748989105225, |
| "learning_rate": 1.7181467181467181e-06, |
| "loss": 762.0396, |
| "step": 429 |
| }, |
| { |
| "epoch": 0.8301158301158301, |
| "grad_norm": 3.2768187522888184, |
| "learning_rate": 1.698841698841699e-06, |
| "loss": 748.2692, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.832046332046332, |
| "grad_norm": 4.247284889221191, |
| "learning_rate": 1.6795366795366796e-06, |
| "loss": 759.6968, |
| "step": 431 |
| }, |
| { |
| "epoch": 0.833976833976834, |
| "grad_norm": 3.205296039581299, |
| "learning_rate": 1.6602316602316604e-06, |
| "loss": 745.4493, |
| "step": 432 |
| }, |
| { |
| "epoch": 0.8359073359073359, |
| "grad_norm": 3.3943800926208496, |
| "learning_rate": 1.640926640926641e-06, |
| "loss": 759.4459, |
| "step": 433 |
| }, |
| { |
| "epoch": 0.8378378378378378, |
| "grad_norm": 3.8933515548706055, |
| "learning_rate": 1.6216216216216219e-06, |
| "loss": 768.5825, |
| "step": 434 |
| }, |
| { |
| "epoch": 0.8397683397683398, |
| "grad_norm": 3.123276710510254, |
| "learning_rate": 1.6023166023166027e-06, |
| "loss": 749.004, |
| "step": 435 |
| }, |
| { |
| "epoch": 0.8416988416988417, |
| "grad_norm": 3.028196334838867, |
| "learning_rate": 1.583011583011583e-06, |
| "loss": 754.6147, |
| "step": 436 |
| }, |
| { |
| "epoch": 0.8436293436293436, |
| "grad_norm": 5.627590179443359, |
| "learning_rate": 1.5637065637065637e-06, |
| "loss": 764.9931, |
| "step": 437 |
| }, |
| { |
| "epoch": 0.8455598455598455, |
| "grad_norm": 4.444796562194824, |
| "learning_rate": 1.5444015444015445e-06, |
| "loss": 756.2324, |
| "step": 438 |
| }, |
| { |
| "epoch": 0.8474903474903475, |
| "grad_norm": 4.994812965393066, |
| "learning_rate": 1.5250965250965251e-06, |
| "loss": 764.1788, |
| "step": 439 |
| }, |
| { |
| "epoch": 0.8494208494208494, |
| "grad_norm": 3.8100221157073975, |
| "learning_rate": 1.505791505791506e-06, |
| "loss": 748.0216, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.8513513513513513, |
| "grad_norm": 3.284438371658325, |
| "learning_rate": 1.4864864864864868e-06, |
| "loss": 744.6481, |
| "step": 441 |
| }, |
| { |
| "epoch": 0.8532818532818532, |
| "grad_norm": 3.754772901535034, |
| "learning_rate": 1.4671814671814674e-06, |
| "loss": 759.6836, |
| "step": 442 |
| }, |
| { |
| "epoch": 0.8552123552123552, |
| "grad_norm": 3.4503631591796875, |
| "learning_rate": 1.4478764478764478e-06, |
| "loss": 757.1558, |
| "step": 443 |
| }, |
| { |
| "epoch": 0.8571428571428571, |
| "grad_norm": 3.9186127185821533, |
| "learning_rate": 1.4285714285714286e-06, |
| "loss": 749.7752, |
| "step": 444 |
| }, |
| { |
| "epoch": 0.859073359073359, |
| "grad_norm": 5.067336082458496, |
| "learning_rate": 1.4092664092664092e-06, |
| "loss": 768.12, |
| "step": 445 |
| }, |
| { |
| "epoch": 0.861003861003861, |
| "grad_norm": 3.5622620582580566, |
| "learning_rate": 1.38996138996139e-06, |
| "loss": 749.7698, |
| "step": 446 |
| }, |
| { |
| "epoch": 0.862934362934363, |
| "grad_norm": 3.2070393562316895, |
| "learning_rate": 1.3706563706563709e-06, |
| "loss": 765.8137, |
| "step": 447 |
| }, |
| { |
| "epoch": 0.8648648648648649, |
| "grad_norm": 3.920966386795044, |
| "learning_rate": 1.3513513513513515e-06, |
| "loss": 763.2885, |
| "step": 448 |
| }, |
| { |
| "epoch": 0.8667953667953668, |
| "grad_norm": 4.828410625457764, |
| "learning_rate": 1.3320463320463323e-06, |
| "loss": 744.1356, |
| "step": 449 |
| }, |
| { |
| "epoch": 0.8687258687258688, |
| "grad_norm": 2.9684271812438965, |
| "learning_rate": 1.3127413127413127e-06, |
| "loss": 748.5911, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.8706563706563707, |
| "grad_norm": 3.4309277534484863, |
| "learning_rate": 1.2934362934362933e-06, |
| "loss": 754.4804, |
| "step": 451 |
| }, |
| { |
| "epoch": 0.8725868725868726, |
| "grad_norm": 3.2176759243011475, |
| "learning_rate": 1.2741312741312742e-06, |
| "loss": 763.7363, |
| "step": 452 |
| }, |
| { |
| "epoch": 0.8745173745173745, |
| "grad_norm": 4.412746429443359, |
| "learning_rate": 1.254826254826255e-06, |
| "loss": 762.2845, |
| "step": 453 |
| }, |
| { |
| "epoch": 0.8764478764478765, |
| "grad_norm": 3.3320093154907227, |
| "learning_rate": 1.2355212355212356e-06, |
| "loss": 737.4437, |
| "step": 454 |
| }, |
| { |
| "epoch": 0.8783783783783784, |
| "grad_norm": 3.1412718296051025, |
| "learning_rate": 1.2162162162162164e-06, |
| "loss": 752.1414, |
| "step": 455 |
| }, |
| { |
| "epoch": 0.8803088803088803, |
| "grad_norm": 2.9298524856567383, |
| "learning_rate": 1.196911196911197e-06, |
| "loss": 749.2881, |
| "step": 456 |
| }, |
| { |
| "epoch": 0.8822393822393823, |
| "grad_norm": 3.4683337211608887, |
| "learning_rate": 1.1776061776061777e-06, |
| "loss": 750.4893, |
| "step": 457 |
| }, |
| { |
| "epoch": 0.8841698841698842, |
| "grad_norm": 8.293436050415039, |
| "learning_rate": 1.1583011583011585e-06, |
| "loss": 759.465, |
| "step": 458 |
| }, |
| { |
| "epoch": 0.8861003861003861, |
| "grad_norm": 3.0078611373901367, |
| "learning_rate": 1.1389961389961391e-06, |
| "loss": 765.3209, |
| "step": 459 |
| }, |
| { |
| "epoch": 0.888030888030888, |
| "grad_norm": 3.5911474227905273, |
| "learning_rate": 1.1196911196911197e-06, |
| "loss": 748.1154, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.88996138996139, |
| "grad_norm": 3.626723527908325, |
| "learning_rate": 1.1003861003861005e-06, |
| "loss": 763.6529, |
| "step": 461 |
| }, |
| { |
| "epoch": 0.8918918918918919, |
| "grad_norm": 3.322850465774536, |
| "learning_rate": 1.0810810810810812e-06, |
| "loss": 757.5947, |
| "step": 462 |
| }, |
| { |
| "epoch": 0.8938223938223938, |
| "grad_norm": 3.866856575012207, |
| "learning_rate": 1.0617760617760618e-06, |
| "loss": 761.439, |
| "step": 463 |
| }, |
| { |
| "epoch": 0.8957528957528957, |
| "grad_norm": 3.293731927871704, |
| "learning_rate": 1.0424710424710426e-06, |
| "loss": 763.1173, |
| "step": 464 |
| }, |
| { |
| "epoch": 0.8976833976833977, |
| "grad_norm": 3.4967000484466553, |
| "learning_rate": 1.0231660231660232e-06, |
| "loss": 771.8659, |
| "step": 465 |
| }, |
| { |
| "epoch": 0.8996138996138996, |
| "grad_norm": 2.9513790607452393, |
| "learning_rate": 1.0038610038610038e-06, |
| "loss": 759.2766, |
| "step": 466 |
| }, |
| { |
| "epoch": 0.9015444015444015, |
| "grad_norm": 3.051725387573242, |
| "learning_rate": 9.845559845559847e-07, |
| "loss": 761.7484, |
| "step": 467 |
| }, |
| { |
| "epoch": 0.9034749034749034, |
| "grad_norm": 3.222289562225342, |
| "learning_rate": 9.652509652509653e-07, |
| "loss": 753.144, |
| "step": 468 |
| }, |
| { |
| "epoch": 0.9054054054054054, |
| "grad_norm": 4.781485557556152, |
| "learning_rate": 9.459459459459461e-07, |
| "loss": 766.5778, |
| "step": 469 |
| }, |
| { |
| "epoch": 0.9073359073359073, |
| "grad_norm": 9.611944198608398, |
| "learning_rate": 9.266409266409267e-07, |
| "loss": 757.17, |
| "step": 470 |
| }, |
| { |
| "epoch": 0.9092664092664092, |
| "grad_norm": 5.166945457458496, |
| "learning_rate": 9.073359073359074e-07, |
| "loss": 758.192, |
| "step": 471 |
| }, |
| { |
| "epoch": 0.9111969111969112, |
| "grad_norm": 3.2668616771698, |
| "learning_rate": 8.880308880308882e-07, |
| "loss": 772.9589, |
| "step": 472 |
| }, |
| { |
| "epoch": 0.9131274131274131, |
| "grad_norm": 3.808668851852417, |
| "learning_rate": 8.687258687258689e-07, |
| "loss": 743.6304, |
| "step": 473 |
| }, |
| { |
| "epoch": 0.915057915057915, |
| "grad_norm": 3.6966071128845215, |
| "learning_rate": 8.494208494208495e-07, |
| "loss": 739.0352, |
| "step": 474 |
| }, |
| { |
| "epoch": 0.916988416988417, |
| "grad_norm": 5.23868465423584, |
| "learning_rate": 8.301158301158302e-07, |
| "loss": 758.2206, |
| "step": 475 |
| }, |
| { |
| "epoch": 0.918918918918919, |
| "grad_norm": 5.4658732414245605, |
| "learning_rate": 8.108108108108109e-07, |
| "loss": 745.1991, |
| "step": 476 |
| }, |
| { |
| "epoch": 0.9208494208494209, |
| "grad_norm": 4.291003227233887, |
| "learning_rate": 7.915057915057915e-07, |
| "loss": 765.403, |
| "step": 477 |
| }, |
| { |
| "epoch": 0.9227799227799228, |
| "grad_norm": 3.528947353363037, |
| "learning_rate": 7.722007722007723e-07, |
| "loss": 764.4042, |
| "step": 478 |
| }, |
| { |
| "epoch": 0.9247104247104247, |
| "grad_norm": 3.179927349090576, |
| "learning_rate": 7.52895752895753e-07, |
| "loss": 757.1093, |
| "step": 479 |
| }, |
| { |
| "epoch": 0.9266409266409267, |
| "grad_norm": 4.567697048187256, |
| "learning_rate": 7.335907335907337e-07, |
| "loss": 759.746, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.9285714285714286, |
| "grad_norm": 3.929335355758667, |
| "learning_rate": 7.142857142857143e-07, |
| "loss": 758.0447, |
| "step": 481 |
| }, |
| { |
| "epoch": 0.9305019305019305, |
| "grad_norm": 5.094730377197266, |
| "learning_rate": 6.94980694980695e-07, |
| "loss": 757.1946, |
| "step": 482 |
| }, |
| { |
| "epoch": 0.9324324324324325, |
| "grad_norm": 3.7125465869903564, |
| "learning_rate": 6.756756756756758e-07, |
| "loss": 770.7656, |
| "step": 483 |
| }, |
| { |
| "epoch": 0.9343629343629344, |
| "grad_norm": 3.3825149536132812, |
| "learning_rate": 6.563706563706564e-07, |
| "loss": 747.5059, |
| "step": 484 |
| }, |
| { |
| "epoch": 0.9362934362934363, |
| "grad_norm": 3.3807835578918457, |
| "learning_rate": 6.370656370656371e-07, |
| "loss": 767.5899, |
| "step": 485 |
| }, |
| { |
| "epoch": 0.9382239382239382, |
| "grad_norm": 2.997955322265625, |
| "learning_rate": 6.177606177606178e-07, |
| "loss": 758.338, |
| "step": 486 |
| }, |
| { |
| "epoch": 0.9401544401544402, |
| "grad_norm": 3.478146553039551, |
| "learning_rate": 5.984555984555985e-07, |
| "loss": 758.3116, |
| "step": 487 |
| }, |
| { |
| "epoch": 0.9420849420849421, |
| "grad_norm": 3.0303502082824707, |
| "learning_rate": 5.791505791505792e-07, |
| "loss": 752.7023, |
| "step": 488 |
| }, |
| { |
| "epoch": 0.944015444015444, |
| "grad_norm": 4.182223796844482, |
| "learning_rate": 5.598455598455599e-07, |
| "loss": 767.7703, |
| "step": 489 |
| }, |
| { |
| "epoch": 0.9459459459459459, |
| "grad_norm": 3.0645620822906494, |
| "learning_rate": 5.405405405405406e-07, |
| "loss": 756.7206, |
| "step": 490 |
| }, |
| { |
| "epoch": 0.9478764478764479, |
| "grad_norm": 3.5093557834625244, |
| "learning_rate": 5.212355212355213e-07, |
| "loss": 759.0178, |
| "step": 491 |
| }, |
| { |
| "epoch": 0.9498069498069498, |
| "grad_norm": 5.231660842895508, |
| "learning_rate": 5.019305019305019e-07, |
| "loss": 758.1909, |
| "step": 492 |
| }, |
| { |
| "epoch": 0.9517374517374517, |
| "grad_norm": 4.316704273223877, |
| "learning_rate": 4.826254826254826e-07, |
| "loss": 756.6906, |
| "step": 493 |
| }, |
| { |
| "epoch": 0.9536679536679536, |
| "grad_norm": 3.842200994491577, |
| "learning_rate": 4.6332046332046336e-07, |
| "loss": 757.8163, |
| "step": 494 |
| }, |
| { |
| "epoch": 0.9555984555984556, |
| "grad_norm": 3.351325511932373, |
| "learning_rate": 4.440154440154441e-07, |
| "loss": 762.0304, |
| "step": 495 |
| }, |
| { |
| "epoch": 0.9575289575289575, |
| "grad_norm": 3.0159342288970947, |
| "learning_rate": 4.2471042471042474e-07, |
| "loss": 761.4176, |
| "step": 496 |
| }, |
| { |
| "epoch": 0.9594594594594594, |
| "grad_norm": 3.074551582336426, |
| "learning_rate": 4.0540540540540546e-07, |
| "loss": 751.5521, |
| "step": 497 |
| }, |
| { |
| "epoch": 0.9613899613899614, |
| "grad_norm": 3.896904945373535, |
| "learning_rate": 3.8610038610038613e-07, |
| "loss": 763.3478, |
| "step": 498 |
| }, |
| { |
| "epoch": 0.9633204633204633, |
| "grad_norm": 3.919020175933838, |
| "learning_rate": 3.6679536679536685e-07, |
| "loss": 768.0314, |
| "step": 499 |
| }, |
| { |
| "epoch": 0.9652509652509652, |
| "grad_norm": 2.949843645095825, |
| "learning_rate": 3.474903474903475e-07, |
| "loss": 755.5469, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.9671814671814671, |
| "grad_norm": 3.6389143466949463, |
| "learning_rate": 3.281853281853282e-07, |
| "loss": 744.184, |
| "step": 501 |
| }, |
| { |
| "epoch": 0.9691119691119691, |
| "grad_norm": 4.229357719421387, |
| "learning_rate": 3.088803088803089e-07, |
| "loss": 761.7263, |
| "step": 502 |
| }, |
| { |
| "epoch": 0.971042471042471, |
| "grad_norm": 3.1734883785247803, |
| "learning_rate": 2.895752895752896e-07, |
| "loss": 758.8358, |
| "step": 503 |
| }, |
| { |
| "epoch": 0.972972972972973, |
| "grad_norm": 3.089228868484497, |
| "learning_rate": 2.702702702702703e-07, |
| "loss": 760.397, |
| "step": 504 |
| }, |
| { |
| "epoch": 0.974903474903475, |
| "grad_norm": 3.926948070526123, |
| "learning_rate": 2.5096525096525096e-07, |
| "loss": 755.6573, |
| "step": 505 |
| }, |
| { |
| "epoch": 0.9768339768339769, |
| "grad_norm": 2.9857828617095947, |
| "learning_rate": 2.3166023166023168e-07, |
| "loss": 767.6666, |
| "step": 506 |
| }, |
| { |
| "epoch": 0.9787644787644788, |
| "grad_norm": 4.68814754486084, |
| "learning_rate": 2.1235521235521237e-07, |
| "loss": 767.6501, |
| "step": 507 |
| }, |
| { |
| "epoch": 0.9806949806949807, |
| "grad_norm": 3.457016706466675, |
| "learning_rate": 1.9305019305019306e-07, |
| "loss": 760.5441, |
| "step": 508 |
| }, |
| { |
| "epoch": 0.9826254826254827, |
| "grad_norm": 4.365758419036865, |
| "learning_rate": 1.7374517374517376e-07, |
| "loss": 760.8904, |
| "step": 509 |
| }, |
| { |
| "epoch": 0.9845559845559846, |
| "grad_norm": 3.6141843795776367, |
| "learning_rate": 1.5444015444015445e-07, |
| "loss": 754.2842, |
| "step": 510 |
| }, |
| { |
| "epoch": 0.9864864864864865, |
| "grad_norm": 4.281587600708008, |
| "learning_rate": 1.3513513513513515e-07, |
| "loss": 754.181, |
| "step": 511 |
| }, |
| { |
| "epoch": 0.9884169884169884, |
| "grad_norm": 3.3735556602478027, |
| "learning_rate": 1.1583011583011584e-07, |
| "loss": 755.9177, |
| "step": 512 |
| }, |
| { |
| "epoch": 0.9903474903474904, |
| "grad_norm": 3.126232624053955, |
| "learning_rate": 9.652509652509653e-08, |
| "loss": 750.3485, |
| "step": 513 |
| }, |
| { |
| "epoch": 0.9922779922779923, |
| "grad_norm": 5.9563727378845215, |
| "learning_rate": 7.722007722007723e-08, |
| "loss": 766.379, |
| "step": 514 |
| }, |
| { |
| "epoch": 0.9942084942084942, |
| "grad_norm": 3.0199837684631348, |
| "learning_rate": 5.791505791505792e-08, |
| "loss": 753.6818, |
| "step": 515 |
| }, |
| { |
| "epoch": 0.9961389961389961, |
| "grad_norm": 3.0909945964813232, |
| "learning_rate": 3.861003861003861e-08, |
| "loss": 749.71, |
| "step": 516 |
| }, |
| { |
| "epoch": 0.9980694980694981, |
| "grad_norm": 3.416806221008301, |
| "learning_rate": 1.9305019305019306e-08, |
| "loss": 772.7269, |
| "step": 517 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 5.65108585357666, |
| "learning_rate": 0.0, |
| "loss": 744.4813, |
| "step": 518 |
| }, |
| { |
| "epoch": 1.0, |
| "step": 518, |
| "total_flos": 1.1089078720895386e+18, |
| "train_loss": 758.2892525573495, |
| "train_runtime": 2215.2409, |
| "train_samples_per_second": 239.051, |
| "train_steps_per_second": 0.234 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 518, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 1, |
| "save_steps": 500, |
| "total_flos": 1.1089078720895386e+18, |
| "train_batch_size": 1024, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
