| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.16704, | |
| "eval_steps": 500, | |
| "global_step": 261, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00064, | |
| "grad_norm": 2.3657337561838188, | |
| "learning_rate": 0.0, | |
| "loss": 0.5764, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00128, | |
| "grad_norm": 2.577238188562681, | |
| "learning_rate": 6.329113924050633e-07, | |
| "loss": 0.6154, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.00192, | |
| "grad_norm": 2.486400760684678, | |
| "learning_rate": 1.2658227848101265e-06, | |
| "loss": 0.5765, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.00256, | |
| "grad_norm": 2.427317708214583, | |
| "learning_rate": 1.8987341772151901e-06, | |
| "loss": 0.5898, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0032, | |
| "grad_norm": 2.3689778807783104, | |
| "learning_rate": 2.531645569620253e-06, | |
| "loss": 0.5846, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.00384, | |
| "grad_norm": 2.533864042007203, | |
| "learning_rate": 3.1645569620253167e-06, | |
| "loss": 0.6002, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.00448, | |
| "grad_norm": 1.879595075845785, | |
| "learning_rate": 3.7974683544303802e-06, | |
| "loss": 0.5581, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.00512, | |
| "grad_norm": 2.056563531274739, | |
| "learning_rate": 4.430379746835443e-06, | |
| "loss": 0.5859, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.00576, | |
| "grad_norm": 1.4703684433180533, | |
| "learning_rate": 5.063291139240506e-06, | |
| "loss": 0.5479, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.0064, | |
| "grad_norm": 1.2614967629103893, | |
| "learning_rate": 5.69620253164557e-06, | |
| "loss": 0.5303, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.00704, | |
| "grad_norm": 1.1460599905324345, | |
| "learning_rate": 6.329113924050633e-06, | |
| "loss": 0.5568, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.00768, | |
| "grad_norm": 1.6204644872566638, | |
| "learning_rate": 6.9620253164556965e-06, | |
| "loss": 0.5111, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.00832, | |
| "grad_norm": 1.6361395875225335, | |
| "learning_rate": 7.5949367088607605e-06, | |
| "loss": 0.4678, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.00896, | |
| "grad_norm": 1.6402122162855226, | |
| "learning_rate": 8.227848101265822e-06, | |
| "loss": 0.4811, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.0096, | |
| "grad_norm": 1.2490639825476322, | |
| "learning_rate": 8.860759493670886e-06, | |
| "loss": 0.442, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.01024, | |
| "grad_norm": 1.0588915477376701, | |
| "learning_rate": 9.49367088607595e-06, | |
| "loss": 0.4263, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.01088, | |
| "grad_norm": 1.40604786962106, | |
| "learning_rate": 1.0126582278481012e-05, | |
| "loss": 0.4549, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.01152, | |
| "grad_norm": 1.242365760774633, | |
| "learning_rate": 1.0759493670886076e-05, | |
| "loss": 0.4531, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.01216, | |
| "grad_norm": 0.9396101260126887, | |
| "learning_rate": 1.139240506329114e-05, | |
| "loss": 0.415, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.0128, | |
| "grad_norm": 0.7578271337493023, | |
| "learning_rate": 1.2025316455696203e-05, | |
| "loss": 0.4268, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.01344, | |
| "grad_norm": 0.7671817679988738, | |
| "learning_rate": 1.2658227848101267e-05, | |
| "loss": 0.4314, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.01408, | |
| "grad_norm": 0.6494114557426621, | |
| "learning_rate": 1.3291139240506329e-05, | |
| "loss": 0.3666, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.01472, | |
| "grad_norm": 0.6806456497077759, | |
| "learning_rate": 1.3924050632911393e-05, | |
| "loss": 0.3823, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.01536, | |
| "grad_norm": 0.6595430154008608, | |
| "learning_rate": 1.4556962025316457e-05, | |
| "loss": 0.3529, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.016, | |
| "grad_norm": 0.7689071362735744, | |
| "learning_rate": 1.5189873417721521e-05, | |
| "loss": 0.3932, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.01664, | |
| "grad_norm": 0.6854685982052705, | |
| "learning_rate": 1.5822784810126583e-05, | |
| "loss": 0.3738, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.01728, | |
| "grad_norm": 0.558852717411362, | |
| "learning_rate": 1.6455696202531644e-05, | |
| "loss": 0.3708, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.01792, | |
| "grad_norm": 0.49876042518765007, | |
| "learning_rate": 1.7088607594936708e-05, | |
| "loss": 0.3696, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.01856, | |
| "grad_norm": 0.5271567613623678, | |
| "learning_rate": 1.7721518987341772e-05, | |
| "loss": 0.3558, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.0192, | |
| "grad_norm": 0.5978920045933569, | |
| "learning_rate": 1.8354430379746836e-05, | |
| "loss": 0.3827, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.01984, | |
| "grad_norm": 0.4975313988318536, | |
| "learning_rate": 1.89873417721519e-05, | |
| "loss": 0.3772, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.02048, | |
| "grad_norm": 0.5766330443378189, | |
| "learning_rate": 1.962025316455696e-05, | |
| "loss": 0.3832, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.02112, | |
| "grad_norm": 0.5567055560538855, | |
| "learning_rate": 2.0253164556962025e-05, | |
| "loss": 0.3477, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.02176, | |
| "grad_norm": 0.46871971492473, | |
| "learning_rate": 2.088607594936709e-05, | |
| "loss": 0.3604, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.0224, | |
| "grad_norm": 0.43316810660778116, | |
| "learning_rate": 2.1518987341772153e-05, | |
| "loss": 0.3592, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.02304, | |
| "grad_norm": 0.4312186195231089, | |
| "learning_rate": 2.2151898734177217e-05, | |
| "loss": 0.3711, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.02368, | |
| "grad_norm": 0.46951838290882675, | |
| "learning_rate": 2.278481012658228e-05, | |
| "loss": 0.3618, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.02432, | |
| "grad_norm": 0.468516818608629, | |
| "learning_rate": 2.341772151898734e-05, | |
| "loss": 0.3578, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.02496, | |
| "grad_norm": 0.45909911923620816, | |
| "learning_rate": 2.4050632911392405e-05, | |
| "loss": 0.3521, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 0.36223572387755915, | |
| "learning_rate": 2.468354430379747e-05, | |
| "loss": 0.3439, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.02624, | |
| "grad_norm": 0.41920026035277985, | |
| "learning_rate": 2.5316455696202533e-05, | |
| "loss": 0.3052, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.02688, | |
| "grad_norm": 0.46410873807902825, | |
| "learning_rate": 2.5949367088607597e-05, | |
| "loss": 0.3324, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.02752, | |
| "grad_norm": 0.44804958203514283, | |
| "learning_rate": 2.6582278481012658e-05, | |
| "loss": 0.322, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.02816, | |
| "grad_norm": 0.4408934955100543, | |
| "learning_rate": 2.7215189873417722e-05, | |
| "loss": 0.3399, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.0288, | |
| "grad_norm": 0.4536362310706478, | |
| "learning_rate": 2.7848101265822786e-05, | |
| "loss": 0.3508, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.02944, | |
| "grad_norm": 0.4735438789842479, | |
| "learning_rate": 2.848101265822785e-05, | |
| "loss": 0.3431, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.03008, | |
| "grad_norm": 0.3960360490220664, | |
| "learning_rate": 2.9113924050632914e-05, | |
| "loss": 0.3358, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.03072, | |
| "grad_norm": 0.4449245975582831, | |
| "learning_rate": 2.9746835443037974e-05, | |
| "loss": 0.3472, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.03136, | |
| "grad_norm": 0.41598379019590886, | |
| "learning_rate": 3.0379746835443042e-05, | |
| "loss": 0.3295, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.032, | |
| "grad_norm": 0.39770652562313213, | |
| "learning_rate": 3.10126582278481e-05, | |
| "loss": 0.339, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03264, | |
| "grad_norm": 0.4036450937079526, | |
| "learning_rate": 3.1645569620253167e-05, | |
| "loss": 0.326, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.03328, | |
| "grad_norm": 0.43082778666471144, | |
| "learning_rate": 3.227848101265823e-05, | |
| "loss": 0.3258, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.03392, | |
| "grad_norm": 0.402275292978643, | |
| "learning_rate": 3.291139240506329e-05, | |
| "loss": 0.2905, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.03456, | |
| "grad_norm": 0.43612618786182106, | |
| "learning_rate": 3.354430379746836e-05, | |
| "loss": 0.334, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.0352, | |
| "grad_norm": 0.45984837714015897, | |
| "learning_rate": 3.4177215189873416e-05, | |
| "loss": 0.3283, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.03584, | |
| "grad_norm": 0.4484461315187549, | |
| "learning_rate": 3.4810126582278487e-05, | |
| "loss": 0.3277, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.03648, | |
| "grad_norm": 0.4052074034352622, | |
| "learning_rate": 3.5443037974683544e-05, | |
| "loss": 0.3271, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.03712, | |
| "grad_norm": 0.46620186767863914, | |
| "learning_rate": 3.607594936708861e-05, | |
| "loss": 0.3464, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.03776, | |
| "grad_norm": 0.42593292833096424, | |
| "learning_rate": 3.670886075949367e-05, | |
| "loss": 0.3331, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 0.4815687673225032, | |
| "learning_rate": 3.7341772151898736e-05, | |
| "loss": 0.3269, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.03904, | |
| "grad_norm": 0.455342196474994, | |
| "learning_rate": 3.79746835443038e-05, | |
| "loss": 0.3343, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.03968, | |
| "grad_norm": 0.4316078425366754, | |
| "learning_rate": 3.8607594936708864e-05, | |
| "loss": 0.3145, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.04032, | |
| "grad_norm": 0.4604468007808017, | |
| "learning_rate": 3.924050632911392e-05, | |
| "loss": 0.3206, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.04096, | |
| "grad_norm": 0.516259514156323, | |
| "learning_rate": 3.987341772151899e-05, | |
| "loss": 0.3538, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.0416, | |
| "grad_norm": 0.3911206186305256, | |
| "learning_rate": 4.050632911392405e-05, | |
| "loss": 0.3116, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.04224, | |
| "grad_norm": 0.5059680673369054, | |
| "learning_rate": 4.113924050632912e-05, | |
| "loss": 0.3461, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.04288, | |
| "grad_norm": 0.4477077725372915, | |
| "learning_rate": 4.177215189873418e-05, | |
| "loss": 0.3142, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.04352, | |
| "grad_norm": 0.4568181894979607, | |
| "learning_rate": 4.240506329113924e-05, | |
| "loss": 0.3258, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.04416, | |
| "grad_norm": 0.436426112682463, | |
| "learning_rate": 4.3037974683544305e-05, | |
| "loss": 0.2942, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.0448, | |
| "grad_norm": 0.4130845172287253, | |
| "learning_rate": 4.367088607594937e-05, | |
| "loss": 0.3293, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.04544, | |
| "grad_norm": 0.4907470169969436, | |
| "learning_rate": 4.430379746835443e-05, | |
| "loss": 0.3152, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.04608, | |
| "grad_norm": 0.44557023947548996, | |
| "learning_rate": 4.49367088607595e-05, | |
| "loss": 0.323, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.04672, | |
| "grad_norm": 0.43347032964452825, | |
| "learning_rate": 4.556962025316456e-05, | |
| "loss": 0.3132, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.04736, | |
| "grad_norm": 0.41315778348357357, | |
| "learning_rate": 4.6202531645569625e-05, | |
| "loss": 0.3076, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.048, | |
| "grad_norm": 0.38689299616228223, | |
| "learning_rate": 4.683544303797468e-05, | |
| "loss": 0.2971, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.04864, | |
| "grad_norm": 0.6649393785343345, | |
| "learning_rate": 4.7468354430379746e-05, | |
| "loss": 0.3196, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.04928, | |
| "grad_norm": 0.4741860571119842, | |
| "learning_rate": 4.810126582278481e-05, | |
| "loss": 0.3261, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.04992, | |
| "grad_norm": 0.39850926668653464, | |
| "learning_rate": 4.8734177215189874e-05, | |
| "loss": 0.2909, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.05056, | |
| "grad_norm": 0.45993342960946926, | |
| "learning_rate": 4.936708860759494e-05, | |
| "loss": 0.3232, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 0.41758744126936076, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3138, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.05184, | |
| "grad_norm": 0.4190137054884969, | |
| "learning_rate": 4.999994390459423e-05, | |
| "loss": 0.324, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.05248, | |
| "grad_norm": 0.38007532443802017, | |
| "learning_rate": 4.999977561862863e-05, | |
| "loss": 0.3167, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.05312, | |
| "grad_norm": 0.3894813419979866, | |
| "learning_rate": 4.999949514285842e-05, | |
| "loss": 0.3169, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.05376, | |
| "grad_norm": 0.4646671073219221, | |
| "learning_rate": 4.999910247854227e-05, | |
| "loss": 0.3295, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.0544, | |
| "grad_norm": 0.46440352293136283, | |
| "learning_rate": 4.99985976274423e-05, | |
| "loss": 0.3259, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.05504, | |
| "grad_norm": 0.4754742191954697, | |
| "learning_rate": 4.9997980591824126e-05, | |
| "loss": 0.3073, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.05568, | |
| "grad_norm": 0.4102331568111904, | |
| "learning_rate": 4.999725137445674e-05, | |
| "loss": 0.3179, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.05632, | |
| "grad_norm": 0.49865407563428993, | |
| "learning_rate": 4.999640997861264e-05, | |
| "loss": 0.3198, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.05696, | |
| "grad_norm": 0.450995150648752, | |
| "learning_rate": 4.999545640806766e-05, | |
| "loss": 0.3215, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.0576, | |
| "grad_norm": 0.42715082345561, | |
| "learning_rate": 4.999439066710111e-05, | |
| "loss": 0.3406, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.05824, | |
| "grad_norm": 0.47007665924261693, | |
| "learning_rate": 4.999321276049563e-05, | |
| "loss": 0.3169, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.05888, | |
| "grad_norm": 0.37692477648671757, | |
| "learning_rate": 4.9991922693537226e-05, | |
| "loss": 0.3052, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.05952, | |
| "grad_norm": 0.5254380682221013, | |
| "learning_rate": 4.9990520472015246e-05, | |
| "loss": 0.2928, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.06016, | |
| "grad_norm": 0.4524991571079602, | |
| "learning_rate": 4.998900610222236e-05, | |
| "loss": 0.3035, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.0608, | |
| "grad_norm": 0.39419071585883453, | |
| "learning_rate": 4.998737959095449e-05, | |
| "loss": 0.2764, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.06144, | |
| "grad_norm": 0.44021580096048935, | |
| "learning_rate": 4.998564094551081e-05, | |
| "loss": 0.3137, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.06208, | |
| "grad_norm": 0.4397078772822271, | |
| "learning_rate": 4.998379017369375e-05, | |
| "loss": 0.3092, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.06272, | |
| "grad_norm": 0.4612518337175716, | |
| "learning_rate": 4.9981827283808855e-05, | |
| "loss": 0.3065, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.06336, | |
| "grad_norm": 0.4491229684653282, | |
| "learning_rate": 4.997975228466489e-05, | |
| "loss": 0.3327, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 0.40909998580765855, | |
| "learning_rate": 4.997756518557368e-05, | |
| "loss": 0.2787, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.06464, | |
| "grad_norm": 0.3785712105730604, | |
| "learning_rate": 4.997526599635011e-05, | |
| "loss": 0.2837, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.06528, | |
| "grad_norm": 0.5009245567404541, | |
| "learning_rate": 4.997285472731211e-05, | |
| "loss": 0.3325, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.06592, | |
| "grad_norm": 0.41940515716785753, | |
| "learning_rate": 4.997033138928056e-05, | |
| "loss": 0.3126, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.06656, | |
| "grad_norm": 0.4356254674524969, | |
| "learning_rate": 4.996769599357927e-05, | |
| "loss": 0.3079, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.0672, | |
| "grad_norm": 0.4485794513852225, | |
| "learning_rate": 4.996494855203493e-05, | |
| "loss": 0.3227, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.06784, | |
| "grad_norm": 0.5339271359603054, | |
| "learning_rate": 4.996208907697706e-05, | |
| "loss": 0.3197, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.06848, | |
| "grad_norm": 0.47913200680485823, | |
| "learning_rate": 4.995911758123791e-05, | |
| "loss": 0.2829, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.06912, | |
| "grad_norm": 0.48619705540043057, | |
| "learning_rate": 4.995603407815249e-05, | |
| "loss": 0.295, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.06976, | |
| "grad_norm": 0.44791381540090275, | |
| "learning_rate": 4.9952838581558406e-05, | |
| "loss": 0.3009, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.0704, | |
| "grad_norm": 0.4217427510368746, | |
| "learning_rate": 4.994953110579588e-05, | |
| "loss": 0.3124, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.07104, | |
| "grad_norm": 0.44605995933104997, | |
| "learning_rate": 4.9946111665707645e-05, | |
| "loss": 0.3232, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.07168, | |
| "grad_norm": 0.3580173680896881, | |
| "learning_rate": 4.99425802766389e-05, | |
| "loss": 0.288, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.07232, | |
| "grad_norm": 0.4281198427243427, | |
| "learning_rate": 4.993893695443721e-05, | |
| "loss": 0.3008, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.07296, | |
| "grad_norm": 0.3748012845273175, | |
| "learning_rate": 4.993518171545247e-05, | |
| "loss": 0.2898, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.0736, | |
| "grad_norm": 0.41133247941906176, | |
| "learning_rate": 4.993131457653682e-05, | |
| "loss": 0.3007, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.07424, | |
| "grad_norm": 0.4049279526177947, | |
| "learning_rate": 4.992733555504454e-05, | |
| "loss": 0.3156, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.07488, | |
| "grad_norm": 0.3907854290589919, | |
| "learning_rate": 4.9923244668832025e-05, | |
| "loss": 0.305, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.07552, | |
| "grad_norm": 3.3672297158831523, | |
| "learning_rate": 4.991904193625768e-05, | |
| "loss": 0.3099, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.07616, | |
| "grad_norm": 0.45896342700641674, | |
| "learning_rate": 4.991472737618181e-05, | |
| "loss": 0.2766, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 0.41249483633287526, | |
| "learning_rate": 4.991030100796659e-05, | |
| "loss": 0.3026, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.07744, | |
| "grad_norm": 0.39435147330028136, | |
| "learning_rate": 4.9905762851475915e-05, | |
| "loss": 0.3151, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.07808, | |
| "grad_norm": 0.38174217860012233, | |
| "learning_rate": 4.990111292707537e-05, | |
| "loss": 0.3132, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.07872, | |
| "grad_norm": 0.4106762400890441, | |
| "learning_rate": 4.989635125563211e-05, | |
| "loss": 0.2883, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.07936, | |
| "grad_norm": 0.3777494709551966, | |
| "learning_rate": 4.989147785851477e-05, | |
| "loss": 0.3075, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.3812281839176885, | |
| "learning_rate": 4.988649275759335e-05, | |
| "loss": 0.3173, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.08064, | |
| "grad_norm": 0.3800442807387135, | |
| "learning_rate": 4.988139597523916e-05, | |
| "loss": 0.2897, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.08128, | |
| "grad_norm": 0.4101363073522896, | |
| "learning_rate": 4.98761875343247e-05, | |
| "loss": 0.2822, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.08192, | |
| "grad_norm": 0.3515414786001439, | |
| "learning_rate": 4.987086745822352e-05, | |
| "loss": 0.2831, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.08256, | |
| "grad_norm": 0.38143533758711873, | |
| "learning_rate": 4.986543577081017e-05, | |
| "loss": 0.3193, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.0832, | |
| "grad_norm": 0.44066201212578954, | |
| "learning_rate": 4.985989249646007e-05, | |
| "loss": 0.3006, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.08384, | |
| "grad_norm": 0.3699946189372801, | |
| "learning_rate": 4.985423766004939e-05, | |
| "loss": 0.309, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.08448, | |
| "grad_norm": 0.4732003041900607, | |
| "learning_rate": 4.9848471286954965e-05, | |
| "loss": 0.3253, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.08512, | |
| "grad_norm": 0.35724915771132804, | |
| "learning_rate": 4.984259340305416e-05, | |
| "loss": 0.2876, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.08576, | |
| "grad_norm": 0.39925687072469707, | |
| "learning_rate": 4.9836604034724753e-05, | |
| "loss": 0.2906, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.0864, | |
| "grad_norm": 0.43582991089095635, | |
| "learning_rate": 4.9830503208844835e-05, | |
| "loss": 0.2928, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.08704, | |
| "grad_norm": 0.3638802742085539, | |
| "learning_rate": 4.982429095279266e-05, | |
| "loss": 0.2924, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.08768, | |
| "grad_norm": 0.44414884826796713, | |
| "learning_rate": 4.981796729444655e-05, | |
| "loss": 0.3225, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.08832, | |
| "grad_norm": 0.4179539121719692, | |
| "learning_rate": 4.981153226218477e-05, | |
| "loss": 0.3067, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.08896, | |
| "grad_norm": 0.3518602463362222, | |
| "learning_rate": 4.9804985884885366e-05, | |
| "loss": 0.2854, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 0.4013030661247676, | |
| "learning_rate": 4.979832819192608e-05, | |
| "loss": 0.3113, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.09024, | |
| "grad_norm": 0.3700912839771966, | |
| "learning_rate": 4.97915592131842e-05, | |
| "loss": 0.2873, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.09088, | |
| "grad_norm": 0.41429834446700914, | |
| "learning_rate": 4.97846789790364e-05, | |
| "loss": 0.3064, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.09152, | |
| "grad_norm": 0.3743350290044281, | |
| "learning_rate": 4.977768752035865e-05, | |
| "loss": 0.311, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.09216, | |
| "grad_norm": 0.4436631708270473, | |
| "learning_rate": 4.9770584868526046e-05, | |
| "loss": 0.3105, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.0928, | |
| "grad_norm": 0.4341903366251428, | |
| "learning_rate": 4.9763371055412685e-05, | |
| "loss": 0.307, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.09344, | |
| "grad_norm": 0.4147022361169306, | |
| "learning_rate": 4.9756046113391496e-05, | |
| "loss": 0.3288, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.09408, | |
| "grad_norm": 0.3884970059580215, | |
| "learning_rate": 4.974861007533413e-05, | |
| "loss": 0.2847, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.09472, | |
| "grad_norm": 0.4258271028958422, | |
| "learning_rate": 4.97410629746108e-05, | |
| "loss": 0.2908, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.09536, | |
| "grad_norm": 0.41527352386736543, | |
| "learning_rate": 4.9733404845090116e-05, | |
| "loss": 0.2828, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 0.41352693469232554, | |
| "learning_rate": 4.972563572113894e-05, | |
| "loss": 0.2895, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.09664, | |
| "grad_norm": 0.4391732737160089, | |
| "learning_rate": 4.971775563762226e-05, | |
| "loss": 0.2908, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.09728, | |
| "grad_norm": 0.42637768991672836, | |
| "learning_rate": 4.970976462990299e-05, | |
| "loss": 0.3103, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.09792, | |
| "grad_norm": 0.38123935717217666, | |
| "learning_rate": 4.970166273384183e-05, | |
| "loss": 0.2859, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.09856, | |
| "grad_norm": 0.3432072306652636, | |
| "learning_rate": 4.969344998579712e-05, | |
| "loss": 0.2653, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.0992, | |
| "grad_norm": 0.38378318854854493, | |
| "learning_rate": 4.9685126422624644e-05, | |
| "loss": 0.3135, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.09984, | |
| "grad_norm": 0.3871767364493639, | |
| "learning_rate": 4.96766920816775e-05, | |
| "loss": 0.3129, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.10048, | |
| "grad_norm": 0.361665333999729, | |
| "learning_rate": 4.966814700080592e-05, | |
| "loss": 0.3175, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.10112, | |
| "grad_norm": 0.394466639141269, | |
| "learning_rate": 4.965949121835707e-05, | |
| "loss": 0.333, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.10176, | |
| "grad_norm": 0.396292343881353, | |
| "learning_rate": 4.965072477317493e-05, | |
| "loss": 0.3084, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 0.3468981265806776, | |
| "learning_rate": 4.964184770460007e-05, | |
| "loss": 0.2927, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.10304, | |
| "grad_norm": 0.39865168519337385, | |
| "learning_rate": 4.963286005246953e-05, | |
| "loss": 0.2815, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.10368, | |
| "grad_norm": 0.37451640530098157, | |
| "learning_rate": 4.9623761857116594e-05, | |
| "loss": 0.3188, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.10432, | |
| "grad_norm": 0.37483826123841657, | |
| "learning_rate": 4.961455315937059e-05, | |
| "loss": 0.2813, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.10496, | |
| "grad_norm": 0.33130533329924106, | |
| "learning_rate": 4.960523400055679e-05, | |
| "loss": 0.2924, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.1056, | |
| "grad_norm": 0.3364997696428208, | |
| "learning_rate": 4.959580442249615e-05, | |
| "loss": 0.3001, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.10624, | |
| "grad_norm": 0.3183485250122132, | |
| "learning_rate": 4.958626446750515e-05, | |
| "loss": 0.3195, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.10688, | |
| "grad_norm": 0.3417933668875517, | |
| "learning_rate": 4.9576614178395595e-05, | |
| "loss": 0.2991, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.10752, | |
| "grad_norm": 0.3487109852934047, | |
| "learning_rate": 4.956685359847445e-05, | |
| "loss": 0.2959, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.10816, | |
| "grad_norm": 0.31315058374358634, | |
| "learning_rate": 4.95569827715436e-05, | |
| "loss": 0.2799, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.1088, | |
| "grad_norm": 0.35913320358975465, | |
| "learning_rate": 4.9547001741899693e-05, | |
| "loss": 0.297, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.10944, | |
| "grad_norm": 0.3331478821334656, | |
| "learning_rate": 4.953691055433391e-05, | |
| "loss": 0.2901, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.11008, | |
| "grad_norm": 0.39224813006559756, | |
| "learning_rate": 4.952670925413182e-05, | |
| "loss": 0.3026, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.11072, | |
| "grad_norm": 0.3480414792318722, | |
| "learning_rate": 4.951639788707307e-05, | |
| "loss": 0.2845, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.11136, | |
| "grad_norm": 0.34970614984057585, | |
| "learning_rate": 4.950597649943132e-05, | |
| "loss": 0.2775, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.112, | |
| "grad_norm": 0.347207358588287, | |
| "learning_rate": 4.9495445137973905e-05, | |
| "loss": 0.2923, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.11264, | |
| "grad_norm": 0.34884090954360897, | |
| "learning_rate": 4.948480384996172e-05, | |
| "loss": 0.2862, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.11328, | |
| "grad_norm": 0.32979237586574495, | |
| "learning_rate": 4.947405268314894e-05, | |
| "loss": 0.2983, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.11392, | |
| "grad_norm": 0.43991275496703847, | |
| "learning_rate": 4.9463191685782865e-05, | |
| "loss": 0.3068, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.11456, | |
| "grad_norm": 0.3039302314073942, | |
| "learning_rate": 4.945222090660364e-05, | |
| "loss": 0.2841, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 0.38589337384034356, | |
| "learning_rate": 4.944114039484411e-05, | |
| "loss": 0.3091, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.11584, | |
| "grad_norm": 0.34105198100381573, | |
| "learning_rate": 4.942995020022954e-05, | |
| "loss": 0.3228, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.11648, | |
| "grad_norm": 0.34780611263986955, | |
| "learning_rate": 4.9418650372977384e-05, | |
| "loss": 0.2874, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.11712, | |
| "grad_norm": 0.32541445621708004, | |
| "learning_rate": 4.940724096379714e-05, | |
| "loss": 0.282, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.11776, | |
| "grad_norm": 0.3082894269244735, | |
| "learning_rate": 4.9395722023890035e-05, | |
| "loss": 0.2799, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.1184, | |
| "grad_norm": 0.32812745331187065, | |
| "learning_rate": 4.938409360494883e-05, | |
| "loss": 0.2827, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.11904, | |
| "grad_norm": 0.34755898651365463, | |
| "learning_rate": 4.937235575915761e-05, | |
| "loss": 0.2917, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.11968, | |
| "grad_norm": 0.3413504764617461, | |
| "learning_rate": 4.93605085391915e-05, | |
| "loss": 0.2918, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.12032, | |
| "grad_norm": 0.3705985820501455, | |
| "learning_rate": 4.934855199821647e-05, | |
| "loss": 0.2826, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.12096, | |
| "grad_norm": 0.3414000814514884, | |
| "learning_rate": 4.93364861898891e-05, | |
| "loss": 0.276, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.1216, | |
| "grad_norm": 0.3589734824395955, | |
| "learning_rate": 4.932431116835628e-05, | |
| "loss": 0.2959, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.12224, | |
| "grad_norm": 0.345670488984043, | |
| "learning_rate": 4.931202698825503e-05, | |
| "loss": 0.281, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.12288, | |
| "grad_norm": 0.3235692312545686, | |
| "learning_rate": 4.929963370471226e-05, | |
| "loss": 0.2886, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.12352, | |
| "grad_norm": 0.3034427958798936, | |
| "learning_rate": 4.928713137334445e-05, | |
| "loss": 0.2812, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.12416, | |
| "grad_norm": 0.34712496586477004, | |
| "learning_rate": 4.927452005025748e-05, | |
| "loss": 0.3018, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.1248, | |
| "grad_norm": 0.3222467927855389, | |
| "learning_rate": 4.926179979204633e-05, | |
| "loss": 0.3023, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.12544, | |
| "grad_norm": 0.3243348951725544, | |
| "learning_rate": 4.9248970655794836e-05, | |
| "loss": 0.2917, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.12608, | |
| "grad_norm": 0.3096795363251828, | |
| "learning_rate": 4.923603269907545e-05, | |
| "loss": 0.2737, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.12672, | |
| "grad_norm": 0.31152060487499617, | |
| "learning_rate": 4.922298597994897e-05, | |
| "loss": 0.2828, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.12736, | |
| "grad_norm": 0.3405465489598549, | |
| "learning_rate": 4.920983055696428e-05, | |
| "loss": 0.2927, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 0.3423946912166355, | |
| "learning_rate": 4.919656648915807e-05, | |
| "loss": 0.2835, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.12864, | |
| "grad_norm": 0.3464727040800733, | |
| "learning_rate": 4.918319383605462e-05, | |
| "loss": 0.2575, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.12928, | |
| "grad_norm": 0.3620779326329196, | |
| "learning_rate": 4.916971265766547e-05, | |
| "loss": 0.2944, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.12992, | |
| "grad_norm": 0.3699616777622434, | |
| "learning_rate": 4.915612301448919e-05, | |
| "loss": 0.2942, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.13056, | |
| "grad_norm": 0.38798124559835534, | |
| "learning_rate": 4.914242496751111e-05, | |
| "loss": 0.2882, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.1312, | |
| "grad_norm": 0.4306348775923099, | |
| "learning_rate": 4.912861857820303e-05, | |
| "loss": 0.3026, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.13184, | |
| "grad_norm": 0.37275998073469474, | |
| "learning_rate": 4.911470390852295e-05, | |
| "loss": 0.3008, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.13248, | |
| "grad_norm": 0.33076859723493407, | |
| "learning_rate": 4.91006810209148e-05, | |
| "loss": 0.2881, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.13312, | |
| "grad_norm": 0.38285708110841127, | |
| "learning_rate": 4.9086549978308125e-05, | |
| "loss": 0.299, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.13376, | |
| "grad_norm": 0.3130925296009032, | |
| "learning_rate": 4.9072310844117875e-05, | |
| "loss": 0.2714, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.1344, | |
| "grad_norm": 0.3556404363171034, | |
| "learning_rate": 4.9057963682244045e-05, | |
| "loss": 0.291, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.13504, | |
| "grad_norm": 0.33759319834147206, | |
| "learning_rate": 4.904350855707142e-05, | |
| "loss": 0.2868, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.13568, | |
| "grad_norm": 0.3171722280640699, | |
| "learning_rate": 4.9028945533469276e-05, | |
| "loss": 0.2646, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.13632, | |
| "grad_norm": 0.39240914087101075, | |
| "learning_rate": 4.9014274676791134e-05, | |
| "loss": 0.302, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.13696, | |
| "grad_norm": 0.34184628116877047, | |
| "learning_rate": 4.899949605287439e-05, | |
| "loss": 0.2792, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.1376, | |
| "grad_norm": 0.3700664187254164, | |
| "learning_rate": 4.898460972804009e-05, | |
| "loss": 0.3088, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.13824, | |
| "grad_norm": 0.36421921549881453, | |
| "learning_rate": 4.896961576909258e-05, | |
| "loss": 0.2808, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.13888, | |
| "grad_norm": 0.3565240796039332, | |
| "learning_rate": 4.895451424331924e-05, | |
| "loss": 0.2714, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.13952, | |
| "grad_norm": 0.3396297883795516, | |
| "learning_rate": 4.893930521849016e-05, | |
| "loss": 0.267, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.14016, | |
| "grad_norm": 0.39278607376772506, | |
| "learning_rate": 4.892398876285785e-05, | |
| "loss": 0.2896, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 0.38437399542076783, | |
| "learning_rate": 4.890856494515696e-05, | |
| "loss": 0.2875, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.14144, | |
| "grad_norm": 0.37920192384172824, | |
| "learning_rate": 4.889303383460389e-05, | |
| "loss": 0.2806, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.14208, | |
| "grad_norm": 0.35504283495161315, | |
| "learning_rate": 4.887739550089656e-05, | |
| "loss": 0.2781, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.14272, | |
| "grad_norm": 0.3866107761723795, | |
| "learning_rate": 4.886165001421408e-05, | |
| "loss": 0.3091, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.14336, | |
| "grad_norm": 0.30439690273605985, | |
| "learning_rate": 4.884579744521638e-05, | |
| "loss": 0.2842, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.144, | |
| "grad_norm": 0.3551417352689114, | |
| "learning_rate": 4.8829837865043984e-05, | |
| "loss": 0.2909, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.14464, | |
| "grad_norm": 0.309111612445454, | |
| "learning_rate": 4.881377134531762e-05, | |
| "loss": 0.268, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.14528, | |
| "grad_norm": 0.3282993602648903, | |
| "learning_rate": 4.8797597958137914e-05, | |
| "loss": 0.28, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.14592, | |
| "grad_norm": 0.3581890513495606, | |
| "learning_rate": 4.878131777608509e-05, | |
| "loss": 0.2842, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.14656, | |
| "grad_norm": 0.32315603560305584, | |
| "learning_rate": 4.876493087221862e-05, | |
| "loss": 0.2961, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.1472, | |
| "grad_norm": 0.33067996529729515, | |
| "learning_rate": 4.87484373200769e-05, | |
| "loss": 0.2891, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.14784, | |
| "grad_norm": 0.3178546016599421, | |
| "learning_rate": 4.8731837193676944e-05, | |
| "loss": 0.2781, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.14848, | |
| "grad_norm": 0.3414289060571217, | |
| "learning_rate": 4.871513056751401e-05, | |
| "loss": 0.2983, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.14912, | |
| "grad_norm": 0.3420355616767074, | |
| "learning_rate": 4.8698317516561295e-05, | |
| "loss": 0.2987, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.14976, | |
| "grad_norm": 0.347117902288678, | |
| "learning_rate": 4.8681398116269596e-05, | |
| "loss": 0.2857, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.1504, | |
| "grad_norm": 0.31648517023684664, | |
| "learning_rate": 4.8664372442566954e-05, | |
| "loss": 0.2921, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.15104, | |
| "grad_norm": 0.3308519821164363, | |
| "learning_rate": 4.864724057185834e-05, | |
| "loss": 0.2672, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.15168, | |
| "grad_norm": 0.31208238851833453, | |
| "learning_rate": 4.86300025810253e-05, | |
| "loss": 0.2798, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.15232, | |
| "grad_norm": 0.2945629591271816, | |
| "learning_rate": 4.861265854742559e-05, | |
| "loss": 0.2732, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.15296, | |
| "grad_norm": 0.3383764371143183, | |
| "learning_rate": 4.859520854889287e-05, | |
| "loss": 0.2923, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 0.2935231514068793, | |
| "learning_rate": 4.857765266373631e-05, | |
| "loss": 0.2945, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.15424, | |
| "grad_norm": 0.3265200869108488, | |
| "learning_rate": 4.855999097074027e-05, | |
| "loss": 0.2925, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.15488, | |
| "grad_norm": 0.301565323998506, | |
| "learning_rate": 4.8542223549163946e-05, | |
| "loss": 0.271, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.15552, | |
| "grad_norm": 0.3121729061732836, | |
| "learning_rate": 4.8524350478740985e-05, | |
| "loss": 0.266, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.15616, | |
| "grad_norm": 0.3612945591439241, | |
| "learning_rate": 4.850637183967917e-05, | |
| "loss": 0.2825, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.1568, | |
| "grad_norm": 0.32526112140366076, | |
| "learning_rate": 4.848828771266002e-05, | |
| "loss": 0.2875, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.15744, | |
| "grad_norm": 0.3724034080949958, | |
| "learning_rate": 4.847009817883844e-05, | |
| "loss": 0.2829, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.15808, | |
| "grad_norm": 0.31233652215083263, | |
| "learning_rate": 4.845180331984239e-05, | |
| "loss": 0.2713, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.15872, | |
| "grad_norm": 0.3480570732169317, | |
| "learning_rate": 4.843340321777246e-05, | |
| "loss": 0.2641, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.15936, | |
| "grad_norm": 0.320447396199795, | |
| "learning_rate": 4.8414897955201556e-05, | |
| "loss": 0.261, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.30011124568595027, | |
| "learning_rate": 4.8396287615174484e-05, | |
| "loss": 0.2566, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.16064, | |
| "grad_norm": 0.3416591207698499, | |
| "learning_rate": 4.837757228120762e-05, | |
| "loss": 0.2886, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.16128, | |
| "grad_norm": 0.3347844484474534, | |
| "learning_rate": 4.835875203728849e-05, | |
| "loss": 0.2546, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.16192, | |
| "grad_norm": 0.3805038941147735, | |
| "learning_rate": 4.8339826967875443e-05, | |
| "loss": 0.2867, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.16256, | |
| "grad_norm": 0.36899369410322225, | |
| "learning_rate": 4.832079715789724e-05, | |
| "loss": 0.2969, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.1632, | |
| "grad_norm": 0.3671046083988311, | |
| "learning_rate": 4.830166269275267e-05, | |
| "loss": 0.2903, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.16384, | |
| "grad_norm": 0.3268166037896424, | |
| "learning_rate": 4.828242365831016e-05, | |
| "loss": 0.2744, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.16448, | |
| "grad_norm": 0.38930768380996184, | |
| "learning_rate": 4.8263080140907455e-05, | |
| "loss": 0.2898, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.16512, | |
| "grad_norm": 0.3527164726419632, | |
| "learning_rate": 4.824363222735114e-05, | |
| "loss": 0.2751, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.16576, | |
| "grad_norm": 0.3831694490194351, | |
| "learning_rate": 4.82240800049163e-05, | |
| "loss": 0.2797, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 0.39066664017067054, | |
| "learning_rate": 4.820442356134612e-05, | |
| "loss": 0.2716, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.16704, | |
| "grad_norm": 0.36528936400377576, | |
| "learning_rate": 4.8184662984851505e-05, | |
| "loss": 0.3001, | |
| "step": 261 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1562, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 261, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 119007652085760.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |