{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 1329, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0007524454477050414, "grad_norm": 187.81736755371094, "learning_rate": 0.0, "loss": 10.1133, "step": 1 }, { "epoch": 0.0015048908954100827, "grad_norm": 151.3889923095703, "learning_rate": 1.4925373134328358e-07, "loss": 9.2148, "step": 2 }, { "epoch": 0.002257336343115124, "grad_norm": 195.20321655273438, "learning_rate": 2.9850746268656716e-07, "loss": 11.0938, "step": 3 }, { "epoch": 0.0030097817908201654, "grad_norm": 348.9756774902344, "learning_rate": 4.4776119402985074e-07, "loss": 9.832, "step": 4 }, { "epoch": 0.003762227238525207, "grad_norm": 163.2272491455078, "learning_rate": 5.970149253731343e-07, "loss": 11.2422, "step": 5 }, { "epoch": 0.004514672686230248, "grad_norm": 176.78646850585938, "learning_rate": 7.462686567164179e-07, "loss": 11.4922, "step": 6 }, { "epoch": 0.005267118133935289, "grad_norm": 171.82962036132812, "learning_rate": 8.955223880597015e-07, "loss": 10.2266, "step": 7 }, { "epoch": 0.006019563581640331, "grad_norm": 193.08480834960938, "learning_rate": 1.044776119402985e-06, "loss": 10.5117, "step": 8 }, { "epoch": 0.006772009029345372, "grad_norm": 162.14559936523438, "learning_rate": 1.1940298507462686e-06, "loss": 9.6094, "step": 9 }, { "epoch": 0.007524454477050414, "grad_norm": 150.88475036621094, "learning_rate": 1.3432835820895524e-06, "loss": 8.7812, "step": 10 }, { "epoch": 0.008276899924755455, "grad_norm": 173.12478637695312, "learning_rate": 1.4925373134328358e-06, "loss": 10.4766, "step": 11 }, { "epoch": 0.009029345372460496, "grad_norm": 172.7010498046875, "learning_rate": 1.6417910447761196e-06, "loss": 9.1289, "step": 12 }, { "epoch": 0.009781790820165538, "grad_norm": 140.74220275878906, "learning_rate": 1.791044776119403e-06, "loss": 7.6797, "step": 13 }, { "epoch": 0.010534236267870579, "grad_norm": 163.679443359375, "learning_rate": 1.9402985074626867e-06, "loss": 9.1445, "step": 14 }, { "epoch": 0.011286681715575621, "grad_norm": 167.57139587402344, "learning_rate": 2.08955223880597e-06, "loss": 9.4844, "step": 15 }, { "epoch": 0.012039127163280662, "grad_norm": 181.3193359375, "learning_rate": 2.238805970149254e-06, "loss": 8.9805, "step": 16 }, { "epoch": 0.012791572610985704, "grad_norm": 121.30020904541016, "learning_rate": 2.3880597014925373e-06, "loss": 7.5312, "step": 17 }, { "epoch": 0.013544018058690745, "grad_norm": 128.07334899902344, "learning_rate": 2.537313432835821e-06, "loss": 7.7539, "step": 18 }, { "epoch": 0.014296463506395787, "grad_norm": 135.87106323242188, "learning_rate": 2.686567164179105e-06, "loss": 8.1602, "step": 19 }, { "epoch": 0.015048908954100828, "grad_norm": 128.00323486328125, "learning_rate": 2.835820895522388e-06, "loss": 6.6758, "step": 20 }, { "epoch": 0.01580135440180587, "grad_norm": 137.32179260253906, "learning_rate": 2.9850746268656716e-06, "loss": 7.0859, "step": 21 }, { "epoch": 0.01655379984951091, "grad_norm": 165.83033752441406, "learning_rate": 3.1343283582089558e-06, "loss": 6.5977, "step": 22 }, { "epoch": 0.01730624529721595, "grad_norm": 147.80575561523438, "learning_rate": 3.283582089552239e-06, "loss": 7.1211, "step": 23 }, { "epoch": 0.01805869074492099, "grad_norm": 178.7413787841797, "learning_rate": 3.4328358208955225e-06, "loss": 6.6016, "step": 24 }, { "epoch": 0.018811136192626036, "grad_norm": 87.64071655273438, "learning_rate": 3.582089552238806e-06, "loss": 6.8906, "step": 25 }, { "epoch": 0.019563581640331076, "grad_norm": 206.30386352539062, "learning_rate": 3.73134328358209e-06, "loss": 5.9648, "step": 26 }, { "epoch": 0.020316027088036117, "grad_norm": 319.09381103515625, "learning_rate": 3.8805970149253735e-06, "loss": 6.6172, "step": 27 }, { "epoch": 0.021068472535741158, "grad_norm": 70.70848846435547, "learning_rate": 4.029850746268657e-06, "loss": 6.2734, "step": 28 }, { "epoch": 0.0218209179834462, "grad_norm": 156.22230529785156, "learning_rate": 4.17910447761194e-06, "loss": 5.5195, "step": 29 }, { "epoch": 0.022573363431151242, "grad_norm": 56.316650390625, "learning_rate": 4.3283582089552236e-06, "loss": 5.3242, "step": 30 }, { "epoch": 0.023325808878856283, "grad_norm": 77.25025177001953, "learning_rate": 4.477611940298508e-06, "loss": 5.0781, "step": 31 }, { "epoch": 0.024078254326561323, "grad_norm": 48.7443962097168, "learning_rate": 4.626865671641791e-06, "loss": 5.1797, "step": 32 }, { "epoch": 0.024830699774266364, "grad_norm": 58.58309555053711, "learning_rate": 4.7761194029850745e-06, "loss": 5.3711, "step": 33 }, { "epoch": 0.025583145221971408, "grad_norm": 54.83428955078125, "learning_rate": 4.925373134328359e-06, "loss": 4.5898, "step": 34 }, { "epoch": 0.02633559066967645, "grad_norm": 153.2790069580078, "learning_rate": 5.074626865671642e-06, "loss": 4.7715, "step": 35 }, { "epoch": 0.02708803611738149, "grad_norm": 126.50374603271484, "learning_rate": 5.2238805970149255e-06, "loss": 4.9414, "step": 36 }, { "epoch": 0.02784048156508653, "grad_norm": 283.3973388671875, "learning_rate": 5.37313432835821e-06, "loss": 5.0352, "step": 37 }, { "epoch": 0.028592927012791574, "grad_norm": 84.64131164550781, "learning_rate": 5.522388059701493e-06, "loss": 4.5098, "step": 38 }, { "epoch": 0.029345372460496615, "grad_norm": 117.75650024414062, "learning_rate": 5.671641791044776e-06, "loss": 4.5391, "step": 39 }, { "epoch": 0.030097817908201655, "grad_norm": 292.9526672363281, "learning_rate": 5.820895522388061e-06, "loss": 5.1367, "step": 40 }, { "epoch": 0.030850263355906696, "grad_norm": 82.79414367675781, "learning_rate": 5.970149253731343e-06, "loss": 4.5117, "step": 41 }, { "epoch": 0.03160270880361174, "grad_norm": 389.5240783691406, "learning_rate": 6.119402985074627e-06, "loss": 5.6406, "step": 42 }, { "epoch": 0.03235515425131678, "grad_norm": 400.56854248046875, "learning_rate": 6.2686567164179116e-06, "loss": 5.3984, "step": 43 }, { "epoch": 0.03310759969902182, "grad_norm": 110.48643493652344, "learning_rate": 6.417910447761194e-06, "loss": 3.9082, "step": 44 }, { "epoch": 0.033860045146726865, "grad_norm": 102.55659484863281, "learning_rate": 6.567164179104478e-06, "loss": 4.4297, "step": 45 }, { "epoch": 0.0346124905944319, "grad_norm": 192.07241821289062, "learning_rate": 6.7164179104477625e-06, "loss": 4.1855, "step": 46 }, { "epoch": 0.035364936042136946, "grad_norm": 148.1069793701172, "learning_rate": 6.865671641791045e-06, "loss": 4.291, "step": 47 }, { "epoch": 0.03611738148984198, "grad_norm": 238.80014038085938, "learning_rate": 7.014925373134329e-06, "loss": 4.5215, "step": 48 }, { "epoch": 0.03686982693754703, "grad_norm": 220.73306274414062, "learning_rate": 7.164179104477612e-06, "loss": 4.5684, "step": 49 }, { "epoch": 0.03762227238525207, "grad_norm": 35.45397186279297, "learning_rate": 7.313432835820896e-06, "loss": 3.9785, "step": 50 }, { "epoch": 0.03837471783295711, "grad_norm": 257.580810546875, "learning_rate": 7.46268656716418e-06, "loss": 4.3066, "step": 51 }, { "epoch": 0.03912716328066215, "grad_norm": 131.84400939941406, "learning_rate": 7.611940298507463e-06, "loss": 4.2578, "step": 52 }, { "epoch": 0.0398796087283672, "grad_norm": 121.55342102050781, "learning_rate": 7.761194029850747e-06, "loss": 4.0273, "step": 53 }, { "epoch": 0.040632054176072234, "grad_norm": 157.86026000976562, "learning_rate": 7.91044776119403e-06, "loss": 4.3867, "step": 54 }, { "epoch": 0.04138449962377728, "grad_norm": 66.99703216552734, "learning_rate": 8.059701492537314e-06, "loss": 3.707, "step": 55 }, { "epoch": 0.042136945071482315, "grad_norm": 35.16154861450195, "learning_rate": 8.208955223880599e-06, "loss": 3.2441, "step": 56 }, { "epoch": 0.04288939051918736, "grad_norm": 44.11848449707031, "learning_rate": 8.35820895522388e-06, "loss": 4.0781, "step": 57 }, { "epoch": 0.0436418359668924, "grad_norm": 67.31927490234375, "learning_rate": 8.507462686567165e-06, "loss": 3.3984, "step": 58 }, { "epoch": 0.04439428141459744, "grad_norm": 66.82479858398438, "learning_rate": 8.656716417910447e-06, "loss": 4.0352, "step": 59 }, { "epoch": 0.045146726862302484, "grad_norm": 56.568931579589844, "learning_rate": 8.805970149253732e-06, "loss": 3.6328, "step": 60 }, { "epoch": 0.04589917231000752, "grad_norm": 124.09092712402344, "learning_rate": 8.955223880597016e-06, "loss": 4.0215, "step": 61 }, { "epoch": 0.046651617757712566, "grad_norm": 99.17076873779297, "learning_rate": 9.104477611940299e-06, "loss": 3.8984, "step": 62 }, { "epoch": 0.04740406320541761, "grad_norm": 232.51084899902344, "learning_rate": 9.253731343283582e-06, "loss": 4.1426, "step": 63 }, { "epoch": 0.04815650865312265, "grad_norm": 213.96304321289062, "learning_rate": 9.402985074626867e-06, "loss": 4.5879, "step": 64 }, { "epoch": 0.04890895410082769, "grad_norm": 47.709999084472656, "learning_rate": 9.552238805970149e-06, "loss": 3.8027, "step": 65 }, { "epoch": 0.04966139954853273, "grad_norm": 233.9221954345703, "learning_rate": 9.701492537313434e-06, "loss": 4.9902, "step": 66 }, { "epoch": 0.05041384499623777, "grad_norm": 239.6158905029297, "learning_rate": 9.850746268656717e-06, "loss": 5.2031, "step": 67 }, { "epoch": 0.051166290443942816, "grad_norm": 157.87442016601562, "learning_rate": 1e-05, "loss": 4.0195, "step": 68 }, { "epoch": 0.05191873589164785, "grad_norm": 63.87594223022461, "learning_rate": 9.992076069730588e-06, "loss": 3.9609, "step": 69 }, { "epoch": 0.0526711813393529, "grad_norm": 148.4019317626953, "learning_rate": 9.984152139461173e-06, "loss": 3.498, "step": 70 }, { "epoch": 0.05342362678705794, "grad_norm": 83.16004180908203, "learning_rate": 9.97622820919176e-06, "loss": 3.6074, "step": 71 }, { "epoch": 0.05417607223476298, "grad_norm": 115.74369049072266, "learning_rate": 9.968304278922346e-06, "loss": 4.2363, "step": 72 }, { "epoch": 0.05492851768246802, "grad_norm": 193.71302795410156, "learning_rate": 9.960380348652933e-06, "loss": 4.709, "step": 73 }, { "epoch": 0.05568096313017306, "grad_norm": 53.86237716674805, "learning_rate": 9.95245641838352e-06, "loss": 3.0889, "step": 74 }, { "epoch": 0.056433408577878104, "grad_norm": 190.28834533691406, "learning_rate": 9.944532488114107e-06, "loss": 3.9102, "step": 75 }, { "epoch": 0.05718585402558315, "grad_norm": 217.49977111816406, "learning_rate": 9.936608557844692e-06, "loss": 4.3711, "step": 76 }, { "epoch": 0.057938299473288185, "grad_norm": 113.78182220458984, "learning_rate": 9.928684627575277e-06, "loss": 3.9707, "step": 77 }, { "epoch": 0.05869074492099323, "grad_norm": 44.046791076660156, "learning_rate": 9.920760697305864e-06, "loss": 3.3438, "step": 78 }, { "epoch": 0.059443190368698266, "grad_norm": 118.92435455322266, "learning_rate": 9.912836767036451e-06, "loss": 3.291, "step": 79 }, { "epoch": 0.06019563581640331, "grad_norm": 32.814632415771484, "learning_rate": 9.904912836767039e-06, "loss": 2.9922, "step": 80 }, { "epoch": 0.060948081264108354, "grad_norm": 43.7454833984375, "learning_rate": 9.896988906497624e-06, "loss": 3.2129, "step": 81 }, { "epoch": 0.06170052671181339, "grad_norm": 70.15777587890625, "learning_rate": 9.88906497622821e-06, "loss": 3.8359, "step": 82 }, { "epoch": 0.062452972159518436, "grad_norm": 74.7402114868164, "learning_rate": 9.881141045958796e-06, "loss": 3.3516, "step": 83 }, { "epoch": 0.06320541760722348, "grad_norm": 28.77251625061035, "learning_rate": 9.873217115689383e-06, "loss": 3.2402, "step": 84 }, { "epoch": 0.06395786305492852, "grad_norm": 52.40445327758789, "learning_rate": 9.86529318541997e-06, "loss": 3.3154, "step": 85 }, { "epoch": 0.06471030850263355, "grad_norm": 56.26734924316406, "learning_rate": 9.857369255150556e-06, "loss": 3.5293, "step": 86 }, { "epoch": 0.0654627539503386, "grad_norm": 55.244136810302734, "learning_rate": 9.849445324881141e-06, "loss": 3.3105, "step": 87 }, { "epoch": 0.06621519939804364, "grad_norm": 39.8226432800293, "learning_rate": 9.841521394611728e-06, "loss": 3.3848, "step": 88 }, { "epoch": 0.06696764484574869, "grad_norm": 24.03954315185547, "learning_rate": 9.833597464342315e-06, "loss": 3.2285, "step": 89 }, { "epoch": 0.06772009029345373, "grad_norm": 90.6539535522461, "learning_rate": 9.825673534072902e-06, "loss": 3.1191, "step": 90 }, { "epoch": 0.06847253574115876, "grad_norm": 100.41248321533203, "learning_rate": 9.817749603803487e-06, "loss": 3.1475, "step": 91 }, { "epoch": 0.0692249811888638, "grad_norm": 119.3066177368164, "learning_rate": 9.809825673534073e-06, "loss": 3.5273, "step": 92 }, { "epoch": 0.06997742663656885, "grad_norm": 132.11619567871094, "learning_rate": 9.80190174326466e-06, "loss": 3.5645, "step": 93 }, { "epoch": 0.07072987208427389, "grad_norm": 100.85049438476562, "learning_rate": 9.793977812995247e-06, "loss": 3.2793, "step": 94 }, { "epoch": 0.07148231753197894, "grad_norm": 100.26934051513672, "learning_rate": 9.786053882725834e-06, "loss": 4.3945, "step": 95 }, { "epoch": 0.07223476297968397, "grad_norm": 118.85063934326172, "learning_rate": 9.77812995245642e-06, "loss": 3.6211, "step": 96 }, { "epoch": 0.07298720842738901, "grad_norm": 103.21385955810547, "learning_rate": 9.770206022187005e-06, "loss": 3.7793, "step": 97 }, { "epoch": 0.07373965387509406, "grad_norm": 126.40459442138672, "learning_rate": 9.762282091917592e-06, "loss": 3.2441, "step": 98 }, { "epoch": 0.0744920993227991, "grad_norm": 167.4326171875, "learning_rate": 9.754358161648179e-06, "loss": 4.0898, "step": 99 }, { "epoch": 0.07524454477050414, "grad_norm": 112.69780731201172, "learning_rate": 9.746434231378766e-06, "loss": 3.5723, "step": 100 }, { "epoch": 0.07599699021820917, "grad_norm": 30.508968353271484, "learning_rate": 9.738510301109351e-06, "loss": 3.5879, "step": 101 }, { "epoch": 0.07674943566591422, "grad_norm": 76.39116668701172, "learning_rate": 9.730586370839936e-06, "loss": 2.8037, "step": 102 }, { "epoch": 0.07750188111361926, "grad_norm": 38.17439651489258, "learning_rate": 9.722662440570524e-06, "loss": 3.2109, "step": 103 }, { "epoch": 0.0782543265613243, "grad_norm": 30.669612884521484, "learning_rate": 9.71473851030111e-06, "loss": 3.1318, "step": 104 }, { "epoch": 0.07900677200902935, "grad_norm": 120.1872787475586, "learning_rate": 9.706814580031696e-06, "loss": 3.4766, "step": 105 }, { "epoch": 0.0797592174567344, "grad_norm": 128.8622589111328, "learning_rate": 9.698890649762283e-06, "loss": 3.4961, "step": 106 }, { "epoch": 0.08051166290443942, "grad_norm": 59.050209045410156, "learning_rate": 9.69096671949287e-06, "loss": 3.1543, "step": 107 }, { "epoch": 0.08126410835214447, "grad_norm": 85.37413024902344, "learning_rate": 9.683042789223455e-06, "loss": 3.3301, "step": 108 }, { "epoch": 0.08201655379984951, "grad_norm": 105.35246276855469, "learning_rate": 9.675118858954042e-06, "loss": 3.3145, "step": 109 }, { "epoch": 0.08276899924755456, "grad_norm": 73.28331756591797, "learning_rate": 9.667194928684628e-06, "loss": 2.8535, "step": 110 }, { "epoch": 0.0835214446952596, "grad_norm": 34.09514617919922, "learning_rate": 9.659270998415215e-06, "loss": 3.668, "step": 111 }, { "epoch": 0.08427389014296463, "grad_norm": 196.14305114746094, "learning_rate": 9.651347068145802e-06, "loss": 4.4473, "step": 112 }, { "epoch": 0.08502633559066967, "grad_norm": 226.3737030029297, "learning_rate": 9.643423137876387e-06, "loss": 4.4688, "step": 113 }, { "epoch": 0.08577878103837472, "grad_norm": 148.45594787597656, "learning_rate": 9.635499207606974e-06, "loss": 3.9121, "step": 114 }, { "epoch": 0.08653122648607976, "grad_norm": 22.061586380004883, "learning_rate": 9.62757527733756e-06, "loss": 3.1465, "step": 115 }, { "epoch": 0.0872836719337848, "grad_norm": 115.12189483642578, "learning_rate": 9.619651347068147e-06, "loss": 3.4834, "step": 116 }, { "epoch": 0.08803611738148984, "grad_norm": 142.36178588867188, "learning_rate": 9.611727416798734e-06, "loss": 3.7148, "step": 117 }, { "epoch": 0.08878856282919488, "grad_norm": 171.8651123046875, "learning_rate": 9.603803486529319e-06, "loss": 4.502, "step": 118 }, { "epoch": 0.08954100827689992, "grad_norm": 133.11654663085938, "learning_rate": 9.595879556259906e-06, "loss": 4.0439, "step": 119 }, { "epoch": 0.09029345372460497, "grad_norm": 26.849279403686523, "learning_rate": 9.587955625990491e-06, "loss": 2.6719, "step": 120 }, { "epoch": 0.09104589917231001, "grad_norm": 114.27269744873047, "learning_rate": 9.580031695721078e-06, "loss": 4.2188, "step": 121 }, { "epoch": 0.09179834462001504, "grad_norm": 203.3996124267578, "learning_rate": 9.572107765451665e-06, "loss": 4.4941, "step": 122 }, { "epoch": 0.09255079006772009, "grad_norm": 165.19558715820312, "learning_rate": 9.56418383518225e-06, "loss": 4.5957, "step": 123 }, { "epoch": 0.09330323551542513, "grad_norm": 125.5126724243164, "learning_rate": 9.556259904912838e-06, "loss": 3.4189, "step": 124 }, { "epoch": 0.09405568096313018, "grad_norm": 79.48650360107422, "learning_rate": 9.548335974643423e-06, "loss": 3.2871, "step": 125 }, { "epoch": 0.09480812641083522, "grad_norm": 66.11795806884766, "learning_rate": 9.54041204437401e-06, "loss": 3.1514, "step": 126 }, { "epoch": 0.09556057185854025, "grad_norm": 74.2788314819336, "learning_rate": 9.532488114104597e-06, "loss": 3.3936, "step": 127 }, { "epoch": 0.0963130173062453, "grad_norm": 35.67583465576172, "learning_rate": 9.524564183835183e-06, "loss": 3.3691, "step": 128 }, { "epoch": 0.09706546275395034, "grad_norm": 52.17413330078125, "learning_rate": 9.51664025356577e-06, "loss": 2.8535, "step": 129 }, { "epoch": 0.09781790820165538, "grad_norm": 119.83101654052734, "learning_rate": 9.508716323296355e-06, "loss": 3.5293, "step": 130 }, { "epoch": 0.09857035364936043, "grad_norm": 146.635986328125, "learning_rate": 9.500792393026942e-06, "loss": 3.2285, "step": 131 }, { "epoch": 0.09932279909706546, "grad_norm": 34.28512191772461, "learning_rate": 9.492868462757529e-06, "loss": 3.127, "step": 132 }, { "epoch": 0.1000752445447705, "grad_norm": 84.61096954345703, "learning_rate": 9.484944532488114e-06, "loss": 3.1426, "step": 133 }, { "epoch": 0.10082768999247554, "grad_norm": 88.358642578125, "learning_rate": 9.477020602218701e-06, "loss": 2.7129, "step": 134 }, { "epoch": 0.10158013544018059, "grad_norm": 53.102962493896484, "learning_rate": 9.469096671949287e-06, "loss": 3.1367, "step": 135 }, { "epoch": 0.10233258088788563, "grad_norm": 55.46281814575195, "learning_rate": 9.461172741679874e-06, "loss": 3.3574, "step": 136 }, { "epoch": 0.10308502633559068, "grad_norm": 72.24687957763672, "learning_rate": 9.45324881141046e-06, "loss": 3.1113, "step": 137 }, { "epoch": 0.1038374717832957, "grad_norm": 84.3861083984375, "learning_rate": 9.445324881141046e-06, "loss": 3.3047, "step": 138 }, { "epoch": 0.10458991723100075, "grad_norm": 88.76127624511719, "learning_rate": 9.437400950871633e-06, "loss": 3.6875, "step": 139 }, { "epoch": 0.1053423626787058, "grad_norm": 55.702484130859375, "learning_rate": 9.429477020602219e-06, "loss": 3.3535, "step": 140 }, { "epoch": 0.10609480812641084, "grad_norm": 68.59583282470703, "learning_rate": 9.421553090332806e-06, "loss": 2.8721, "step": 141 }, { "epoch": 0.10684725357411588, "grad_norm": 89.981689453125, "learning_rate": 9.413629160063393e-06, "loss": 3.2344, "step": 142 }, { "epoch": 0.10759969902182091, "grad_norm": 92.87030792236328, "learning_rate": 9.405705229793978e-06, "loss": 3.5117, "step": 143 }, { "epoch": 0.10835214446952596, "grad_norm": 27.346115112304688, "learning_rate": 9.397781299524565e-06, "loss": 2.7305, "step": 144 }, { "epoch": 0.109104589917231, "grad_norm": 43.55765151977539, "learning_rate": 9.38985736925515e-06, "loss": 3.2832, "step": 145 }, { "epoch": 0.10985703536493605, "grad_norm": 91.49488067626953, "learning_rate": 9.381933438985737e-06, "loss": 3.1992, "step": 146 }, { "epoch": 0.11060948081264109, "grad_norm": 67.19734954833984, "learning_rate": 9.374009508716324e-06, "loss": 3.3184, "step": 147 }, { "epoch": 0.11136192626034612, "grad_norm": 47.810272216796875, "learning_rate": 9.366085578446912e-06, "loss": 3.3203, "step": 148 }, { "epoch": 0.11211437170805116, "grad_norm": 70.10330200195312, "learning_rate": 9.358161648177497e-06, "loss": 4.127, "step": 149 }, { "epoch": 0.11286681715575621, "grad_norm": 43.671142578125, "learning_rate": 9.350237717908082e-06, "loss": 2.5596, "step": 150 }, { "epoch": 0.11361926260346125, "grad_norm": 22.097169876098633, "learning_rate": 9.34231378763867e-06, "loss": 2.8418, "step": 151 }, { "epoch": 0.1143717080511663, "grad_norm": 26.284515380859375, "learning_rate": 9.334389857369256e-06, "loss": 3.3594, "step": 152 }, { "epoch": 0.11512415349887133, "grad_norm": 39.73884582519531, "learning_rate": 9.326465927099843e-06, "loss": 3.6699, "step": 153 }, { "epoch": 0.11587659894657637, "grad_norm": 27.474496841430664, "learning_rate": 9.318541996830429e-06, "loss": 2.9297, "step": 154 }, { "epoch": 0.11662904439428141, "grad_norm": 33.951560974121094, "learning_rate": 9.310618066561014e-06, "loss": 3.5488, "step": 155 }, { "epoch": 0.11738148984198646, "grad_norm": 26.033336639404297, "learning_rate": 9.302694136291601e-06, "loss": 2.9043, "step": 156 }, { "epoch": 0.1181339352896915, "grad_norm": 22.7733211517334, "learning_rate": 9.294770206022188e-06, "loss": 3.4277, "step": 157 }, { "epoch": 0.11888638073739653, "grad_norm": 78.39336395263672, "learning_rate": 9.286846275752775e-06, "loss": 3.4902, "step": 158 }, { "epoch": 0.11963882618510158, "grad_norm": 26.790979385375977, "learning_rate": 9.27892234548336e-06, "loss": 3.002, "step": 159 }, { "epoch": 0.12039127163280662, "grad_norm": 21.88675308227539, "learning_rate": 9.270998415213946e-06, "loss": 3.2793, "step": 160 }, { "epoch": 0.12114371708051166, "grad_norm": 36.12630844116211, "learning_rate": 9.263074484944533e-06, "loss": 3.1387, "step": 161 }, { "epoch": 0.12189616252821671, "grad_norm": 61.90777587890625, "learning_rate": 9.25515055467512e-06, "loss": 3.1348, "step": 162 }, { "epoch": 0.12264860797592174, "grad_norm": 53.83844757080078, "learning_rate": 9.247226624405707e-06, "loss": 3.1816, "step": 163 }, { "epoch": 0.12340105342362678, "grad_norm": 48.54594039916992, "learning_rate": 9.239302694136292e-06, "loss": 2.8086, "step": 164 }, { "epoch": 0.12415349887133183, "grad_norm": 36.44153594970703, "learning_rate": 9.231378763866878e-06, "loss": 3.2012, "step": 165 }, { "epoch": 0.12490594431903687, "grad_norm": 69.61957550048828, "learning_rate": 9.223454833597465e-06, "loss": 3.1953, "step": 166 }, { "epoch": 0.1256583897667419, "grad_norm": 103.36559295654297, "learning_rate": 9.215530903328052e-06, "loss": 3.1553, "step": 167 }, { "epoch": 0.12641083521444696, "grad_norm": 66.82914733886719, "learning_rate": 9.207606973058639e-06, "loss": 3.2227, "step": 168 }, { "epoch": 0.127163280662152, "grad_norm": 51.18250274658203, "learning_rate": 9.199683042789224e-06, "loss": 3.082, "step": 169 }, { "epoch": 0.12791572610985705, "grad_norm": 72.96109771728516, "learning_rate": 9.19175911251981e-06, "loss": 3.6699, "step": 170 }, { "epoch": 0.12866817155756208, "grad_norm": 75.40308380126953, "learning_rate": 9.183835182250396e-06, "loss": 3.2969, "step": 171 }, { "epoch": 0.1294206170052671, "grad_norm": 28.419187545776367, "learning_rate": 9.175911251980984e-06, "loss": 2.9824, "step": 172 }, { "epoch": 0.13017306245297217, "grad_norm": 53.255279541015625, "learning_rate": 9.16798732171157e-06, "loss": 3.2578, "step": 173 }, { "epoch": 0.1309255079006772, "grad_norm": 71.64155578613281, "learning_rate": 9.160063391442156e-06, "loss": 3.0547, "step": 174 }, { "epoch": 0.13167795334838225, "grad_norm": 28.97214126586914, "learning_rate": 9.152139461172741e-06, "loss": 3.0166, "step": 175 }, { "epoch": 0.13243039879608728, "grad_norm": 50.072044372558594, "learning_rate": 9.144215530903328e-06, "loss": 3.4219, "step": 176 }, { "epoch": 0.13318284424379231, "grad_norm": 21.372264862060547, "learning_rate": 9.136291600633915e-06, "loss": 2.9746, "step": 177 }, { "epoch": 0.13393528969149737, "grad_norm": 24.634939193725586, "learning_rate": 9.128367670364502e-06, "loss": 3.1484, "step": 178 }, { "epoch": 0.1346877351392024, "grad_norm": 62.78826904296875, "learning_rate": 9.120443740095088e-06, "loss": 3.3242, "step": 179 }, { "epoch": 0.13544018058690746, "grad_norm": 31.50160789489746, "learning_rate": 9.112519809825675e-06, "loss": 2.6777, "step": 180 }, { "epoch": 0.1361926260346125, "grad_norm": 39.34284591674805, "learning_rate": 9.10459587955626e-06, "loss": 2.5, "step": 181 }, { "epoch": 0.13694507148231752, "grad_norm": 32.132713317871094, "learning_rate": 9.096671949286847e-06, "loss": 2.5635, "step": 182 }, { "epoch": 0.13769751693002258, "grad_norm": 41.65974044799805, "learning_rate": 9.088748019017434e-06, "loss": 3.3496, "step": 183 }, { "epoch": 0.1384499623777276, "grad_norm": 38.184654235839844, "learning_rate": 9.08082408874802e-06, "loss": 3.9766, "step": 184 }, { "epoch": 0.13920240782543267, "grad_norm": 75.4859848022461, "learning_rate": 9.072900158478607e-06, "loss": 3.1914, "step": 185 }, { "epoch": 0.1399548532731377, "grad_norm": 34.991886138916016, "learning_rate": 9.064976228209192e-06, "loss": 3.0166, "step": 186 }, { "epoch": 0.14070729872084273, "grad_norm": 69.22996520996094, "learning_rate": 9.057052297939779e-06, "loss": 3.0352, "step": 187 }, { "epoch": 0.14145974416854779, "grad_norm": 45.45055389404297, "learning_rate": 9.049128367670366e-06, "loss": 3.2051, "step": 188 }, { "epoch": 0.14221218961625282, "grad_norm": 37.568206787109375, "learning_rate": 9.041204437400951e-06, "loss": 3.2168, "step": 189 }, { "epoch": 0.14296463506395787, "grad_norm": 30.181642532348633, "learning_rate": 9.033280507131538e-06, "loss": 3.2676, "step": 190 }, { "epoch": 0.1437170805116629, "grad_norm": 23.631832122802734, "learning_rate": 9.025356576862124e-06, "loss": 3.0547, "step": 191 }, { "epoch": 0.14446952595936793, "grad_norm": 40.3087043762207, "learning_rate": 9.01743264659271e-06, "loss": 2.9238, "step": 192 }, { "epoch": 0.145221971407073, "grad_norm": 28.530550003051758, "learning_rate": 9.009508716323298e-06, "loss": 2.5859, "step": 193 }, { "epoch": 0.14597441685477802, "grad_norm": 74.48800659179688, "learning_rate": 9.001584786053883e-06, "loss": 3.0322, "step": 194 }, { "epoch": 0.14672686230248308, "grad_norm": 106.76486206054688, "learning_rate": 8.99366085578447e-06, "loss": 3.7148, "step": 195 }, { "epoch": 0.1474793077501881, "grad_norm": 67.58019256591797, "learning_rate": 8.985736925515056e-06, "loss": 2.9502, "step": 196 }, { "epoch": 0.14823175319789314, "grad_norm": 54.92237091064453, "learning_rate": 8.977812995245643e-06, "loss": 2.8799, "step": 197 }, { "epoch": 0.1489841986455982, "grad_norm": 59.95341873168945, "learning_rate": 8.96988906497623e-06, "loss": 3.7773, "step": 198 }, { "epoch": 0.14973664409330323, "grad_norm": 38.49965286254883, "learning_rate": 8.961965134706815e-06, "loss": 3.5742, "step": 199 }, { "epoch": 0.1504890895410083, "grad_norm": 32.147010803222656, "learning_rate": 8.954041204437402e-06, "loss": 2.9717, "step": 200 }, { "epoch": 0.15124153498871332, "grad_norm": 33.17499542236328, "learning_rate": 8.946117274167987e-06, "loss": 2.9531, "step": 201 }, { "epoch": 0.15199398043641835, "grad_norm": 26.919301986694336, "learning_rate": 8.938193343898574e-06, "loss": 2.9121, "step": 202 }, { "epoch": 0.1527464258841234, "grad_norm": 21.65958595275879, "learning_rate": 8.930269413629161e-06, "loss": 2.9121, "step": 203 }, { "epoch": 0.15349887133182843, "grad_norm": 20.233850479125977, "learning_rate": 8.922345483359747e-06, "loss": 3.0332, "step": 204 }, { "epoch": 0.1542513167795335, "grad_norm": 31.14151382446289, "learning_rate": 8.914421553090334e-06, "loss": 3.0938, "step": 205 }, { "epoch": 0.15500376222723852, "grad_norm": 29.060792922973633, "learning_rate": 8.90649762282092e-06, "loss": 3.4043, "step": 206 }, { "epoch": 0.15575620767494355, "grad_norm": 43.115238189697266, "learning_rate": 8.898573692551506e-06, "loss": 3.208, "step": 207 }, { "epoch": 0.1565086531226486, "grad_norm": 25.682483673095703, "learning_rate": 8.890649762282093e-06, "loss": 3.2988, "step": 208 }, { "epoch": 0.15726109857035364, "grad_norm": 23.325185775756836, "learning_rate": 8.882725832012679e-06, "loss": 3.0186, "step": 209 }, { "epoch": 0.1580135440180587, "grad_norm": 19.003551483154297, "learning_rate": 8.874801901743266e-06, "loss": 2.5508, "step": 210 }, { "epoch": 0.15876598946576373, "grad_norm": 34.20718765258789, "learning_rate": 8.866877971473851e-06, "loss": 3.1582, "step": 211 }, { "epoch": 0.1595184349134688, "grad_norm": 25.554487228393555, "learning_rate": 8.858954041204438e-06, "loss": 3.0859, "step": 212 }, { "epoch": 0.16027088036117382, "grad_norm": 25.85458755493164, "learning_rate": 8.851030110935025e-06, "loss": 3.0215, "step": 213 }, { "epoch": 0.16102332580887885, "grad_norm": 89.72293853759766, "learning_rate": 8.84310618066561e-06, "loss": 3.2988, "step": 214 }, { "epoch": 0.1617757712565839, "grad_norm": 61.3826789855957, "learning_rate": 8.835182250396197e-06, "loss": 3.2188, "step": 215 }, { "epoch": 0.16252821670428894, "grad_norm": 35.555458068847656, "learning_rate": 8.827258320126783e-06, "loss": 3.2764, "step": 216 }, { "epoch": 0.163280662151994, "grad_norm": 88.35000610351562, "learning_rate": 8.81933438985737e-06, "loss": 3.1133, "step": 217 }, { "epoch": 0.16403310759969902, "grad_norm": 114.05996704101562, "learning_rate": 8.811410459587957e-06, "loss": 3.9531, "step": 218 }, { "epoch": 0.16478555304740405, "grad_norm": 78.45552062988281, "learning_rate": 8.803486529318542e-06, "loss": 2.957, "step": 219 }, { "epoch": 0.1655379984951091, "grad_norm": 31.660160064697266, "learning_rate": 8.79556259904913e-06, "loss": 2.8115, "step": 220 }, { "epoch": 0.16629044394281414, "grad_norm": 31.968887329101562, "learning_rate": 8.787638668779716e-06, "loss": 2.6855, "step": 221 }, { "epoch": 0.1670428893905192, "grad_norm": 49.5087890625, "learning_rate": 8.779714738510302e-06, "loss": 3.3604, "step": 222 }, { "epoch": 0.16779533483822423, "grad_norm": 22.69972038269043, "learning_rate": 8.771790808240889e-06, "loss": 2.6787, "step": 223 }, { "epoch": 0.16854778028592926, "grad_norm": 47.977970123291016, "learning_rate": 8.763866877971474e-06, "loss": 3.0547, "step": 224 }, { "epoch": 0.16930022573363432, "grad_norm": 65.0685043334961, "learning_rate": 8.755942947702061e-06, "loss": 3.3545, "step": 225 }, { "epoch": 0.17005267118133935, "grad_norm": 43.86599349975586, "learning_rate": 8.748019017432648e-06, "loss": 2.8574, "step": 226 }, { "epoch": 0.1708051166290444, "grad_norm": 72.90522003173828, "learning_rate": 8.740095087163233e-06, "loss": 3.7168, "step": 227 }, { "epoch": 0.17155756207674944, "grad_norm": 41.62009048461914, "learning_rate": 8.73217115689382e-06, "loss": 2.8691, "step": 228 }, { "epoch": 0.17231000752445447, "grad_norm": 53.0240478515625, "learning_rate": 8.724247226624406e-06, "loss": 2.8555, "step": 229 }, { "epoch": 0.17306245297215953, "grad_norm": 24.85753059387207, "learning_rate": 8.716323296354993e-06, "loss": 3.0605, "step": 230 }, { "epoch": 0.17381489841986456, "grad_norm": 21.22967529296875, "learning_rate": 8.70839936608558e-06, "loss": 3.2324, "step": 231 }, { "epoch": 0.1745673438675696, "grad_norm": 104.60713958740234, "learning_rate": 8.700475435816165e-06, "loss": 3.541, "step": 232 }, { "epoch": 0.17531978931527464, "grad_norm": 102.33487701416016, "learning_rate": 8.692551505546752e-06, "loss": 3.2305, "step": 233 }, { "epoch": 0.17607223476297967, "grad_norm": 35.243202209472656, "learning_rate": 8.684627575277338e-06, "loss": 3.1211, "step": 234 }, { "epoch": 0.17682468021068473, "grad_norm": 25.380643844604492, "learning_rate": 8.676703645007925e-06, "loss": 2.9746, "step": 235 }, { "epoch": 0.17757712565838976, "grad_norm": 94.3141860961914, "learning_rate": 8.668779714738512e-06, "loss": 3.5654, "step": 236 }, { "epoch": 0.17832957110609482, "grad_norm": 130.24545288085938, "learning_rate": 8.660855784469097e-06, "loss": 4.0391, "step": 237 }, { "epoch": 0.17908201655379985, "grad_norm": 76.28020477294922, "learning_rate": 8.652931854199684e-06, "loss": 2.9375, "step": 238 }, { "epoch": 0.17983446200150488, "grad_norm": 84.25048065185547, "learning_rate": 8.64500792393027e-06, "loss": 3.4512, "step": 239 }, { "epoch": 0.18058690744920994, "grad_norm": 25.853254318237305, "learning_rate": 8.637083993660857e-06, "loss": 2.8232, "step": 240 }, { "epoch": 0.18133935289691497, "grad_norm": 51.54713439941406, "learning_rate": 8.629160063391444e-06, "loss": 3.04, "step": 241 }, { "epoch": 0.18209179834462003, "grad_norm": 81.95057678222656, "learning_rate": 8.621236133122029e-06, "loss": 3.2344, "step": 242 }, { "epoch": 0.18284424379232506, "grad_norm": 41.63261032104492, "learning_rate": 8.613312202852616e-06, "loss": 3.0566, "step": 243 }, { "epoch": 0.1835966892400301, "grad_norm": 41.84830856323242, "learning_rate": 8.605388272583201e-06, "loss": 3.1377, "step": 244 }, { "epoch": 0.18434913468773514, "grad_norm": 37.13138961791992, "learning_rate": 8.597464342313788e-06, "loss": 2.3369, "step": 245 }, { "epoch": 0.18510158013544017, "grad_norm": 39.32001876831055, "learning_rate": 8.589540412044375e-06, "loss": 2.8877, "step": 246 }, { "epoch": 0.18585402558314523, "grad_norm": 32.617576599121094, "learning_rate": 8.58161648177496e-06, "loss": 2.7949, "step": 247 }, { "epoch": 0.18660647103085026, "grad_norm": 24.77182388305664, "learning_rate": 8.573692551505548e-06, "loss": 2.7861, "step": 248 }, { "epoch": 0.1873589164785553, "grad_norm": 26.878522872924805, "learning_rate": 8.565768621236133e-06, "loss": 2.8887, "step": 249 }, { "epoch": 0.18811136192626035, "grad_norm": 46.47650909423828, "learning_rate": 8.55784469096672e-06, "loss": 2.4863, "step": 250 }, { "epoch": 0.18886380737396538, "grad_norm": 33.06129455566406, "learning_rate": 8.549920760697307e-06, "loss": 2.7051, "step": 251 }, { "epoch": 0.18961625282167044, "grad_norm": 43.33230209350586, "learning_rate": 8.541996830427893e-06, "loss": 4.084, "step": 252 }, { "epoch": 0.19036869826937547, "grad_norm": 27.43387222290039, "learning_rate": 8.53407290015848e-06, "loss": 2.6455, "step": 253 }, { "epoch": 0.1911211437170805, "grad_norm": 34.015403747558594, "learning_rate": 8.526148969889065e-06, "loss": 3.3154, "step": 254 }, { "epoch": 0.19187358916478556, "grad_norm": 22.817705154418945, "learning_rate": 8.518225039619652e-06, "loss": 2.9551, "step": 255 }, { "epoch": 0.1926260346124906, "grad_norm": 81.8350601196289, "learning_rate": 8.510301109350239e-06, "loss": 2.792, "step": 256 }, { "epoch": 0.19337848006019565, "grad_norm": 84.2726058959961, "learning_rate": 8.502377179080824e-06, "loss": 2.9844, "step": 257 }, { "epoch": 0.19413092550790068, "grad_norm": 67.91531372070312, "learning_rate": 8.494453248811411e-06, "loss": 2.6729, "step": 258 }, { "epoch": 0.1948833709556057, "grad_norm": 20.062740325927734, "learning_rate": 8.486529318541997e-06, "loss": 2.79, "step": 259 }, { "epoch": 0.19563581640331076, "grad_norm": 59.18655014038086, "learning_rate": 8.478605388272584e-06, "loss": 3.4775, "step": 260 }, { "epoch": 0.1963882618510158, "grad_norm": 59.75342559814453, "learning_rate": 8.47068145800317e-06, "loss": 3.0488, "step": 261 }, { "epoch": 0.19714070729872085, "grad_norm": 70.16788482666016, "learning_rate": 8.462757527733758e-06, "loss": 3.7871, "step": 262 }, { "epoch": 0.19789315274642588, "grad_norm": 39.20167541503906, "learning_rate": 8.454833597464343e-06, "loss": 2.959, "step": 263 }, { "epoch": 0.1986455981941309, "grad_norm": 55.009517669677734, "learning_rate": 8.446909667194929e-06, "loss": 3.0439, "step": 264 }, { "epoch": 0.19939804364183597, "grad_norm": 113.7911148071289, "learning_rate": 8.438985736925516e-06, "loss": 3.6328, "step": 265 }, { "epoch": 0.200150489089541, "grad_norm": 70.44924926757812, "learning_rate": 8.431061806656103e-06, "loss": 3.0176, "step": 266 }, { "epoch": 0.20090293453724606, "grad_norm": 45.5489616394043, "learning_rate": 8.42313787638669e-06, "loss": 2.9434, "step": 267 }, { "epoch": 0.2016553799849511, "grad_norm": 62.986175537109375, "learning_rate": 8.415213946117275e-06, "loss": 3.1152, "step": 268 }, { "epoch": 0.20240782543265612, "grad_norm": 37.91203689575195, "learning_rate": 8.40729001584786e-06, "loss": 3.6875, "step": 269 }, { "epoch": 0.20316027088036118, "grad_norm": 75.06108856201172, "learning_rate": 8.399366085578447e-06, "loss": 3.127, "step": 270 }, { "epoch": 0.2039127163280662, "grad_norm": 45.03618240356445, "learning_rate": 8.391442155309034e-06, "loss": 2.7373, "step": 271 }, { "epoch": 0.20466516177577126, "grad_norm": 20.670963287353516, "learning_rate": 8.383518225039621e-06, "loss": 2.6094, "step": 272 }, { "epoch": 0.2054176072234763, "grad_norm": 84.76927947998047, "learning_rate": 8.375594294770207e-06, "loss": 3.1133, "step": 273 }, { "epoch": 0.20617005267118135, "grad_norm": 36.80317306518555, "learning_rate": 8.367670364500792e-06, "loss": 2.7793, "step": 274 }, { "epoch": 0.20692249811888638, "grad_norm": 38.66841506958008, "learning_rate": 8.35974643423138e-06, "loss": 3.0479, "step": 275 }, { "epoch": 0.2076749435665914, "grad_norm": 47.253578186035156, "learning_rate": 8.351822503961966e-06, "loss": 3.1748, "step": 276 }, { "epoch": 0.20842738901429647, "grad_norm": 21.542232513427734, "learning_rate": 8.343898573692553e-06, "loss": 2.8145, "step": 277 }, { "epoch": 0.2091798344620015, "grad_norm": 55.47714614868164, "learning_rate": 8.335974643423139e-06, "loss": 2.498, "step": 278 }, { "epoch": 0.20993227990970656, "grad_norm": 54.076080322265625, "learning_rate": 8.328050713153724e-06, "loss": 2.8848, "step": 279 }, { "epoch": 0.2106847253574116, "grad_norm": 45.00600051879883, "learning_rate": 8.320126782884311e-06, "loss": 3.2305, "step": 280 }, { "epoch": 0.21143717080511662, "grad_norm": 27.784671783447266, "learning_rate": 8.312202852614898e-06, "loss": 2.8555, "step": 281 }, { "epoch": 0.21218961625282168, "grad_norm": 21.135204315185547, "learning_rate": 8.304278922345485e-06, "loss": 2.8125, "step": 282 }, { "epoch": 0.2129420617005267, "grad_norm": 27.72955894470215, "learning_rate": 8.29635499207607e-06, "loss": 3.0762, "step": 283 }, { "epoch": 0.21369450714823177, "grad_norm": 23.81853485107422, "learning_rate": 8.288431061806656e-06, "loss": 2.5938, "step": 284 }, { "epoch": 0.2144469525959368, "grad_norm": 41.66539764404297, "learning_rate": 8.280507131537243e-06, "loss": 2.8809, "step": 285 }, { "epoch": 0.21519939804364183, "grad_norm": 38.57289123535156, "learning_rate": 8.27258320126783e-06, "loss": 2.6543, "step": 286 }, { "epoch": 0.21595184349134688, "grad_norm": 46.27495193481445, "learning_rate": 8.264659270998417e-06, "loss": 2.6543, "step": 287 }, { "epoch": 0.21670428893905191, "grad_norm": 88.909912109375, "learning_rate": 8.256735340729002e-06, "loss": 3.042, "step": 288 }, { "epoch": 0.21745673438675697, "grad_norm": 56.075382232666016, "learning_rate": 8.24881141045959e-06, "loss": 2.9414, "step": 289 }, { "epoch": 0.218209179834462, "grad_norm": 26.580392837524414, "learning_rate": 8.240887480190175e-06, "loss": 3.0, "step": 290 }, { "epoch": 0.21896162528216703, "grad_norm": 61.84196853637695, "learning_rate": 8.232963549920762e-06, "loss": 2.8662, "step": 291 }, { "epoch": 0.2197140707298721, "grad_norm": 43.39738082885742, "learning_rate": 8.225039619651349e-06, "loss": 3.292, "step": 292 }, { "epoch": 0.22046651617757712, "grad_norm": 41.86874008178711, "learning_rate": 8.217115689381934e-06, "loss": 3.2148, "step": 293 }, { "epoch": 0.22121896162528218, "grad_norm": 20.22987937927246, "learning_rate": 8.209191759112521e-06, "loss": 3.0645, "step": 294 }, { "epoch": 0.2219714070729872, "grad_norm": 99.28943634033203, "learning_rate": 8.201267828843106e-06, "loss": 3.373, "step": 295 }, { "epoch": 0.22272385252069224, "grad_norm": 121.67118835449219, "learning_rate": 8.193343898573693e-06, "loss": 3.5176, "step": 296 }, { "epoch": 0.2234762979683973, "grad_norm": 105.28216552734375, "learning_rate": 8.18541996830428e-06, "loss": 3.1875, "step": 297 }, { "epoch": 0.22422874341610233, "grad_norm": 70.92816925048828, "learning_rate": 8.177496038034866e-06, "loss": 3.2969, "step": 298 }, { "epoch": 0.22498118886380739, "grad_norm": 22.64484977722168, "learning_rate": 8.169572107765453e-06, "loss": 3.3965, "step": 299 }, { "epoch": 0.22573363431151242, "grad_norm": 39.87211608886719, "learning_rate": 8.161648177496038e-06, "loss": 2.9863, "step": 300 }, { "epoch": 0.22648607975921745, "grad_norm": 62.50984573364258, "learning_rate": 8.153724247226625e-06, "loss": 2.3682, "step": 301 }, { "epoch": 0.2272385252069225, "grad_norm": 32.058197021484375, "learning_rate": 8.145800316957212e-06, "loss": 2.9736, "step": 302 }, { "epoch": 0.22799097065462753, "grad_norm": 39.14604568481445, "learning_rate": 8.137876386687798e-06, "loss": 3.2285, "step": 303 }, { "epoch": 0.2287434161023326, "grad_norm": 27.052776336669922, "learning_rate": 8.129952456418385e-06, "loss": 2.9492, "step": 304 }, { "epoch": 0.22949586155003762, "grad_norm": 83.68045806884766, "learning_rate": 8.12202852614897e-06, "loss": 3.6328, "step": 305 }, { "epoch": 0.23024830699774265, "grad_norm": 117.3252944946289, "learning_rate": 8.114104595879557e-06, "loss": 3.5703, "step": 306 }, { "epoch": 0.2310007524454477, "grad_norm": 65.46855926513672, "learning_rate": 8.106180665610144e-06, "loss": 2.8477, "step": 307 }, { "epoch": 0.23175319789315274, "grad_norm": 70.63121795654297, "learning_rate": 8.09825673534073e-06, "loss": 2.9668, "step": 308 }, { "epoch": 0.2325056433408578, "grad_norm": 37.77394104003906, "learning_rate": 8.090332805071317e-06, "loss": 3.0215, "step": 309 }, { "epoch": 0.23325808878856283, "grad_norm": 52.81776809692383, "learning_rate": 8.082408874801902e-06, "loss": 2.7529, "step": 310 }, { "epoch": 0.23401053423626786, "grad_norm": 97.96956634521484, "learning_rate": 8.074484944532489e-06, "loss": 3.8711, "step": 311 }, { "epoch": 0.23476297968397292, "grad_norm": 94.45693969726562, "learning_rate": 8.066561014263076e-06, "loss": 3.2363, "step": 312 }, { "epoch": 0.23551542513167795, "grad_norm": 113.64404296875, "learning_rate": 8.058637083993661e-06, "loss": 3.5781, "step": 313 }, { "epoch": 0.236267870579383, "grad_norm": 44.172882080078125, "learning_rate": 8.050713153724248e-06, "loss": 3.1006, "step": 314 }, { "epoch": 0.23702031602708803, "grad_norm": 51.807559967041016, "learning_rate": 8.042789223454834e-06, "loss": 2.8213, "step": 315 }, { "epoch": 0.23777276147479307, "grad_norm": 54.96208190917969, "learning_rate": 8.03486529318542e-06, "loss": 2.9482, "step": 316 }, { "epoch": 0.23852520692249812, "grad_norm": 34.15311813354492, "learning_rate": 8.026941362916006e-06, "loss": 2.998, "step": 317 }, { "epoch": 0.23927765237020315, "grad_norm": 57.66706085205078, "learning_rate": 8.019017432646593e-06, "loss": 2.8281, "step": 318 }, { "epoch": 0.2400300978179082, "grad_norm": 31.129295349121094, "learning_rate": 8.01109350237718e-06, "loss": 3.3594, "step": 319 }, { "epoch": 0.24078254326561324, "grad_norm": 67.8250732421875, "learning_rate": 8.003169572107765e-06, "loss": 3.6387, "step": 320 }, { "epoch": 0.24153498871331827, "grad_norm": 78.51448822021484, "learning_rate": 7.995245641838353e-06, "loss": 3.2305, "step": 321 }, { "epoch": 0.24228743416102333, "grad_norm": 82.17552185058594, "learning_rate": 7.987321711568938e-06, "loss": 2.8574, "step": 322 }, { "epoch": 0.24303987960872836, "grad_norm": 76.5505599975586, "learning_rate": 7.979397781299525e-06, "loss": 3.1113, "step": 323 }, { "epoch": 0.24379232505643342, "grad_norm": 19.232769012451172, "learning_rate": 7.971473851030112e-06, "loss": 3.1807, "step": 324 }, { "epoch": 0.24454477050413845, "grad_norm": 65.15279388427734, "learning_rate": 7.963549920760697e-06, "loss": 2.8574, "step": 325 }, { "epoch": 0.24529721595184348, "grad_norm": 83.49951171875, "learning_rate": 7.955625990491284e-06, "loss": 3.1113, "step": 326 }, { "epoch": 0.24604966139954854, "grad_norm": 32.631622314453125, "learning_rate": 7.94770206022187e-06, "loss": 2.7305, "step": 327 }, { "epoch": 0.24680210684725357, "grad_norm": 45.0832405090332, "learning_rate": 7.939778129952457e-06, "loss": 3.2754, "step": 328 }, { "epoch": 0.24755455229495862, "grad_norm": 20.239614486694336, "learning_rate": 7.931854199683044e-06, "loss": 2.8242, "step": 329 }, { "epoch": 0.24830699774266365, "grad_norm": 62.87268829345703, "learning_rate": 7.92393026941363e-06, "loss": 3.1045, "step": 330 }, { "epoch": 0.24905944319036868, "grad_norm": 74.40349578857422, "learning_rate": 7.916006339144216e-06, "loss": 3.1553, "step": 331 }, { "epoch": 0.24981188863807374, "grad_norm": 47.66246795654297, "learning_rate": 7.908082408874802e-06, "loss": 3.1406, "step": 332 }, { "epoch": 0.2505643340857788, "grad_norm": 52.15084457397461, "learning_rate": 7.900158478605389e-06, "loss": 3.4297, "step": 333 }, { "epoch": 0.2513167795334838, "grad_norm": 24.649290084838867, "learning_rate": 7.892234548335976e-06, "loss": 3.1738, "step": 334 }, { "epoch": 0.2520692249811889, "grad_norm": 94.12933349609375, "learning_rate": 7.884310618066563e-06, "loss": 3.291, "step": 335 }, { "epoch": 0.2528216704288939, "grad_norm": 80.10469818115234, "learning_rate": 7.876386687797148e-06, "loss": 3.1777, "step": 336 }, { "epoch": 0.25357411587659895, "grad_norm": 103.66165161132812, "learning_rate": 7.868462757527733e-06, "loss": 3.0996, "step": 337 }, { "epoch": 0.254326561324304, "grad_norm": 41.690093994140625, "learning_rate": 7.86053882725832e-06, "loss": 2.875, "step": 338 }, { "epoch": 0.255079006772009, "grad_norm": 29.56729507446289, "learning_rate": 7.852614896988907e-06, "loss": 3.1309, "step": 339 }, { "epoch": 0.2558314522197141, "grad_norm": 50.44151306152344, "learning_rate": 7.844690966719494e-06, "loss": 3.0566, "step": 340 }, { "epoch": 0.2565838976674191, "grad_norm": 64.72400665283203, "learning_rate": 7.83676703645008e-06, "loss": 3.165, "step": 341 }, { "epoch": 0.25733634311512416, "grad_norm": 20.706647872924805, "learning_rate": 7.828843106180665e-06, "loss": 2.7725, "step": 342 }, { "epoch": 0.2580887885628292, "grad_norm": 31.02082633972168, "learning_rate": 7.820919175911252e-06, "loss": 2.7695, "step": 343 }, { "epoch": 0.2588412340105342, "grad_norm": 21.273944854736328, "learning_rate": 7.81299524564184e-06, "loss": 2.6152, "step": 344 }, { "epoch": 0.2595936794582393, "grad_norm": 21.953739166259766, "learning_rate": 7.805071315372426e-06, "loss": 2.7334, "step": 345 }, { "epoch": 0.26034612490594433, "grad_norm": 27.68943977355957, "learning_rate": 7.797147385103012e-06, "loss": 3.1729, "step": 346 }, { "epoch": 0.26109857035364936, "grad_norm": 21.063129425048828, "learning_rate": 7.789223454833597e-06, "loss": 2.4844, "step": 347 }, { "epoch": 0.2618510158013544, "grad_norm": 27.426313400268555, "learning_rate": 7.781299524564184e-06, "loss": 2.748, "step": 348 }, { "epoch": 0.2626034612490594, "grad_norm": 21.69107437133789, "learning_rate": 7.773375594294771e-06, "loss": 2.7773, "step": 349 }, { "epoch": 0.2633559066967645, "grad_norm": 32.103153228759766, "learning_rate": 7.765451664025358e-06, "loss": 2.709, "step": 350 }, { "epoch": 0.26410835214446954, "grad_norm": 63.66896438598633, "learning_rate": 7.757527733755943e-06, "loss": 3.2031, "step": 351 }, { "epoch": 0.26486079759217457, "grad_norm": 29.75193214416504, "learning_rate": 7.749603803486529e-06, "loss": 3.168, "step": 352 }, { "epoch": 0.2656132430398796, "grad_norm": 25.284969329833984, "learning_rate": 7.741679873217116e-06, "loss": 2.5859, "step": 353 }, { "epoch": 0.26636568848758463, "grad_norm": 25.517702102661133, "learning_rate": 7.733755942947703e-06, "loss": 2.791, "step": 354 }, { "epoch": 0.2671181339352897, "grad_norm": 67.4950942993164, "learning_rate": 7.72583201267829e-06, "loss": 2.9072, "step": 355 }, { "epoch": 0.26787057938299474, "grad_norm": 31.88636016845703, "learning_rate": 7.717908082408875e-06, "loss": 2.8047, "step": 356 }, { "epoch": 0.2686230248306998, "grad_norm": 19.853824615478516, "learning_rate": 7.70998415213946e-06, "loss": 2.5537, "step": 357 }, { "epoch": 0.2693754702784048, "grad_norm": 68.49687957763672, "learning_rate": 7.702060221870048e-06, "loss": 3.1318, "step": 358 }, { "epoch": 0.27012791572610984, "grad_norm": 44.60654830932617, "learning_rate": 7.694136291600635e-06, "loss": 2.7461, "step": 359 }, { "epoch": 0.2708803611738149, "grad_norm": 54.63933181762695, "learning_rate": 7.686212361331222e-06, "loss": 2.6074, "step": 360 }, { "epoch": 0.27163280662151995, "grad_norm": 32.24433517456055, "learning_rate": 7.678288431061807e-06, "loss": 3.6348, "step": 361 }, { "epoch": 0.272385252069225, "grad_norm": 23.62687873840332, "learning_rate": 7.670364500792394e-06, "loss": 2.8691, "step": 362 }, { "epoch": 0.27313769751693, "grad_norm": 24.92971420288086, "learning_rate": 7.66244057052298e-06, "loss": 2.2783, "step": 363 }, { "epoch": 0.27389014296463504, "grad_norm": 72.9671859741211, "learning_rate": 7.654516640253566e-06, "loss": 3.1533, "step": 364 }, { "epoch": 0.2746425884123401, "grad_norm": 23.203428268432617, "learning_rate": 7.646592709984154e-06, "loss": 2.4209, "step": 365 }, { "epoch": 0.27539503386004516, "grad_norm": 26.601865768432617, "learning_rate": 7.638668779714739e-06, "loss": 2.5693, "step": 366 }, { "epoch": 0.2761474793077502, "grad_norm": 22.43524932861328, "learning_rate": 7.630744849445326e-06, "loss": 3.1426, "step": 367 }, { "epoch": 0.2768999247554552, "grad_norm": 25.846656799316406, "learning_rate": 7.622820919175912e-06, "loss": 2.8037, "step": 368 }, { "epoch": 0.27765237020316025, "grad_norm": 22.966304779052734, "learning_rate": 7.614896988906498e-06, "loss": 3.2334, "step": 369 }, { "epoch": 0.27840481565086533, "grad_norm": 31.72697639465332, "learning_rate": 7.606973058637085e-06, "loss": 2.6318, "step": 370 }, { "epoch": 0.27915726109857036, "grad_norm": 26.754465103149414, "learning_rate": 7.5990491283676715e-06, "loss": 2.3926, "step": 371 }, { "epoch": 0.2799097065462754, "grad_norm": 29.274229049682617, "learning_rate": 7.591125198098257e-06, "loss": 2.9092, "step": 372 }, { "epoch": 0.2806621519939804, "grad_norm": 19.97159194946289, "learning_rate": 7.583201267828844e-06, "loss": 2.5391, "step": 373 }, { "epoch": 0.28141459744168545, "grad_norm": 23.58123779296875, "learning_rate": 7.57527733755943e-06, "loss": 2.9385, "step": 374 }, { "epoch": 0.28216704288939054, "grad_norm": 22.98595428466797, "learning_rate": 7.567353407290017e-06, "loss": 2.6777, "step": 375 }, { "epoch": 0.28291948833709557, "grad_norm": 65.53314208984375, "learning_rate": 7.559429477020603e-06, "loss": 3.1846, "step": 376 }, { "epoch": 0.2836719337848006, "grad_norm": 63.19049835205078, "learning_rate": 7.551505546751189e-06, "loss": 2.5977, "step": 377 }, { "epoch": 0.28442437923250563, "grad_norm": 28.387653350830078, "learning_rate": 7.543581616481776e-06, "loss": 2.8359, "step": 378 }, { "epoch": 0.28517682468021066, "grad_norm": 26.09599494934082, "learning_rate": 7.535657686212362e-06, "loss": 2.6807, "step": 379 }, { "epoch": 0.28592927012791575, "grad_norm": 73.98117065429688, "learning_rate": 7.527733755942949e-06, "loss": 3.3809, "step": 380 }, { "epoch": 0.2866817155756208, "grad_norm": 53.673030853271484, "learning_rate": 7.519809825673535e-06, "loss": 3.1035, "step": 381 }, { "epoch": 0.2874341610233258, "grad_norm": 49.4766845703125, "learning_rate": 7.5118858954041205e-06, "loss": 2.8252, "step": 382 }, { "epoch": 0.28818660647103084, "grad_norm": 27.39165496826172, "learning_rate": 7.5039619651347075e-06, "loss": 2.5625, "step": 383 }, { "epoch": 0.28893905191873587, "grad_norm": 26.559101104736328, "learning_rate": 7.496038034865294e-06, "loss": 2.7559, "step": 384 }, { "epoch": 0.28969149736644095, "grad_norm": 34.693641662597656, "learning_rate": 7.488114104595881e-06, "loss": 2.5527, "step": 385 }, { "epoch": 0.290443942814146, "grad_norm": 31.282804489135742, "learning_rate": 7.480190174326466e-06, "loss": 3.1592, "step": 386 }, { "epoch": 0.291196388261851, "grad_norm": 23.59662628173828, "learning_rate": 7.472266244057052e-06, "loss": 2.3418, "step": 387 }, { "epoch": 0.29194883370955604, "grad_norm": 37.472572326660156, "learning_rate": 7.464342313787639e-06, "loss": 2.3193, "step": 388 }, { "epoch": 0.2927012791572611, "grad_norm": 54.54274368286133, "learning_rate": 7.4564183835182255e-06, "loss": 3.084, "step": 389 }, { "epoch": 0.29345372460496616, "grad_norm": 52.258155822753906, "learning_rate": 7.4484944532488126e-06, "loss": 3.1748, "step": 390 }, { "epoch": 0.2942061700526712, "grad_norm": 29.557064056396484, "learning_rate": 7.440570522979398e-06, "loss": 3.1416, "step": 391 }, { "epoch": 0.2949586155003762, "grad_norm": 21.459806442260742, "learning_rate": 7.432646592709984e-06, "loss": 2.6309, "step": 392 }, { "epoch": 0.29571106094808125, "grad_norm": 65.10810089111328, "learning_rate": 7.424722662440571e-06, "loss": 2.875, "step": 393 }, { "epoch": 0.2964635063957863, "grad_norm": 69.97591400146484, "learning_rate": 7.416798732171157e-06, "loss": 2.9814, "step": 394 }, { "epoch": 0.29721595184349137, "grad_norm": 58.1292839050293, "learning_rate": 7.408874801901744e-06, "loss": 3.0332, "step": 395 }, { "epoch": 0.2979683972911964, "grad_norm": 24.810606002807617, "learning_rate": 7.40095087163233e-06, "loss": 2.7246, "step": 396 }, { "epoch": 0.2987208427389014, "grad_norm": 35.730674743652344, "learning_rate": 7.393026941362916e-06, "loss": 3.0996, "step": 397 }, { "epoch": 0.29947328818660646, "grad_norm": 89.5046615600586, "learning_rate": 7.385103011093503e-06, "loss": 3.4258, "step": 398 }, { "epoch": 0.3002257336343115, "grad_norm": 32.95557403564453, "learning_rate": 7.377179080824089e-06, "loss": 2.3613, "step": 399 }, { "epoch": 0.3009781790820166, "grad_norm": 17.055770874023438, "learning_rate": 7.369255150554676e-06, "loss": 2.6543, "step": 400 }, { "epoch": 0.3017306245297216, "grad_norm": 37.67470932006836, "learning_rate": 7.3613312202852615e-06, "loss": 3.0732, "step": 401 }, { "epoch": 0.30248306997742663, "grad_norm": 28.4992618560791, "learning_rate": 7.3534072900158486e-06, "loss": 2.7207, "step": 402 }, { "epoch": 0.30323551542513166, "grad_norm": 35.26691818237305, "learning_rate": 7.345483359746435e-06, "loss": 2.8926, "step": 403 }, { "epoch": 0.3039879608728367, "grad_norm": 40.095428466796875, "learning_rate": 7.337559429477021e-06, "loss": 3.1211, "step": 404 }, { "epoch": 0.3047404063205418, "grad_norm": 18.42124366760254, "learning_rate": 7.329635499207608e-06, "loss": 2.501, "step": 405 }, { "epoch": 0.3054928517682468, "grad_norm": 21.575937271118164, "learning_rate": 7.321711568938193e-06, "loss": 2.5635, "step": 406 }, { "epoch": 0.30624529721595184, "grad_norm": 35.04372787475586, "learning_rate": 7.31378763866878e-06, "loss": 2.5801, "step": 407 }, { "epoch": 0.30699774266365687, "grad_norm": 38.118961334228516, "learning_rate": 7.305863708399367e-06, "loss": 3.166, "step": 408 }, { "epoch": 0.3077501881113619, "grad_norm": 25.873291015625, "learning_rate": 7.297939778129954e-06, "loss": 2.8125, "step": 409 }, { "epoch": 0.308502633559067, "grad_norm": 22.422454833984375, "learning_rate": 7.29001584786054e-06, "loss": 2.8438, "step": 410 }, { "epoch": 0.309255079006772, "grad_norm": 30.403430938720703, "learning_rate": 7.282091917591125e-06, "loss": 2.7852, "step": 411 }, { "epoch": 0.31000752445447705, "grad_norm": 47.75285339355469, "learning_rate": 7.274167987321712e-06, "loss": 3.0742, "step": 412 }, { "epoch": 0.3107599699021821, "grad_norm": 51.34233474731445, "learning_rate": 7.266244057052298e-06, "loss": 3.1562, "step": 413 }, { "epoch": 0.3115124153498871, "grad_norm": 24.690597534179688, "learning_rate": 7.2583201267828854e-06, "loss": 2.7275, "step": 414 }, { "epoch": 0.3122648607975922, "grad_norm": 26.2873592376709, "learning_rate": 7.250396196513472e-06, "loss": 2.9141, "step": 415 }, { "epoch": 0.3130173062452972, "grad_norm": 18.63983726501465, "learning_rate": 7.242472266244057e-06, "loss": 3.1152, "step": 416 }, { "epoch": 0.31376975169300225, "grad_norm": 55.87972640991211, "learning_rate": 7.234548335974644e-06, "loss": 3.2871, "step": 417 }, { "epoch": 0.3145221971407073, "grad_norm": 41.152286529541016, "learning_rate": 7.22662440570523e-06, "loss": 2.6465, "step": 418 }, { "epoch": 0.3152746425884123, "grad_norm": 47.55572509765625, "learning_rate": 7.218700475435817e-06, "loss": 3.0742, "step": 419 }, { "epoch": 0.3160270880361174, "grad_norm": 24.060579299926758, "learning_rate": 7.2107765451664034e-06, "loss": 3.0566, "step": 420 }, { "epoch": 0.31677953348382243, "grad_norm": 20.185791015625, "learning_rate": 7.202852614896989e-06, "loss": 3.1035, "step": 421 }, { "epoch": 0.31753197893152746, "grad_norm": 36.20608139038086, "learning_rate": 7.194928684627576e-06, "loss": 3.0059, "step": 422 }, { "epoch": 0.3182844243792325, "grad_norm": 34.9112434387207, "learning_rate": 7.187004754358162e-06, "loss": 2.8789, "step": 423 }, { "epoch": 0.3190368698269376, "grad_norm": 16.00592803955078, "learning_rate": 7.179080824088749e-06, "loss": 2.5615, "step": 424 }, { "epoch": 0.3197893152746426, "grad_norm": 39.905330657958984, "learning_rate": 7.171156893819335e-06, "loss": 2.6543, "step": 425 }, { "epoch": 0.32054176072234764, "grad_norm": 39.642520904541016, "learning_rate": 7.163232963549921e-06, "loss": 2.8311, "step": 426 }, { "epoch": 0.32129420617005267, "grad_norm": 34.66411209106445, "learning_rate": 7.155309033280508e-06, "loss": 3.3242, "step": 427 }, { "epoch": 0.3220466516177577, "grad_norm": 25.36420249938965, "learning_rate": 7.147385103011094e-06, "loss": 2.7295, "step": 428 }, { "epoch": 0.3227990970654628, "grad_norm": 26.767236709594727, "learning_rate": 7.139461172741681e-06, "loss": 2.9414, "step": 429 }, { "epoch": 0.3235515425131678, "grad_norm": 29.003219604492188, "learning_rate": 7.131537242472267e-06, "loss": 2.8301, "step": 430 }, { "epoch": 0.32430398796087284, "grad_norm": 17.377384185791016, "learning_rate": 7.123613312202852e-06, "loss": 2.8506, "step": 431 }, { "epoch": 0.32505643340857787, "grad_norm": 28.27768898010254, "learning_rate": 7.1156893819334394e-06, "loss": 2.8789, "step": 432 }, { "epoch": 0.3258088788562829, "grad_norm": 25.299936294555664, "learning_rate": 7.107765451664026e-06, "loss": 2.6748, "step": 433 }, { "epoch": 0.326561324303988, "grad_norm": 20.38831901550293, "learning_rate": 7.099841521394613e-06, "loss": 2.498, "step": 434 }, { "epoch": 0.327313769751693, "grad_norm": 33.19630432128906, "learning_rate": 7.091917591125199e-06, "loss": 3.2666, "step": 435 }, { "epoch": 0.32806621519939805, "grad_norm": 26.92388153076172, "learning_rate": 7.083993660855785e-06, "loss": 2.6367, "step": 436 }, { "epoch": 0.3288186606471031, "grad_norm": 33.628597259521484, "learning_rate": 7.076069730586371e-06, "loss": 2.7441, "step": 437 }, { "epoch": 0.3295711060948081, "grad_norm": 47.53849411010742, "learning_rate": 7.0681458003169574e-06, "loss": 2.251, "step": 438 }, { "epoch": 0.3303235515425132, "grad_norm": 90.70731353759766, "learning_rate": 7.0602218700475445e-06, "loss": 3.3291, "step": 439 }, { "epoch": 0.3310759969902182, "grad_norm": 90.42861938476562, "learning_rate": 7.052297939778131e-06, "loss": 3.8516, "step": 440 }, { "epoch": 0.33182844243792325, "grad_norm": 27.976673126220703, "learning_rate": 7.044374009508717e-06, "loss": 2.7227, "step": 441 }, { "epoch": 0.3325808878856283, "grad_norm": 32.398536682128906, "learning_rate": 7.036450079239303e-06, "loss": 3.0078, "step": 442 }, { "epoch": 0.3333333333333333, "grad_norm": 16.50065803527832, "learning_rate": 7.02852614896989e-06, "loss": 2.5225, "step": 443 }, { "epoch": 0.3340857787810384, "grad_norm": 17.448528289794922, "learning_rate": 7.020602218700476e-06, "loss": 3.0, "step": 444 }, { "epoch": 0.33483822422874343, "grad_norm": 42.91080093383789, "learning_rate": 7.0126782884310625e-06, "loss": 2.8887, "step": 445 }, { "epoch": 0.33559066967644846, "grad_norm": 43.43963623046875, "learning_rate": 7.004754358161649e-06, "loss": 2.7139, "step": 446 }, { "epoch": 0.3363431151241535, "grad_norm": 21.494733810424805, "learning_rate": 6.996830427892235e-06, "loss": 2.8984, "step": 447 }, { "epoch": 0.3370955605718585, "grad_norm": 19.013137817382812, "learning_rate": 6.988906497622822e-06, "loss": 2.6895, "step": 448 }, { "epoch": 0.3378480060195636, "grad_norm": 59.25289535522461, "learning_rate": 6.980982567353408e-06, "loss": 2.7393, "step": 449 }, { "epoch": 0.33860045146726864, "grad_norm": 35.03165817260742, "learning_rate": 6.973058637083995e-06, "loss": 2.5195, "step": 450 }, { "epoch": 0.33935289691497367, "grad_norm": 39.91790008544922, "learning_rate": 6.9651347068145805e-06, "loss": 3.0195, "step": 451 }, { "epoch": 0.3401053423626787, "grad_norm": 31.990583419799805, "learning_rate": 6.957210776545167e-06, "loss": 3.2422, "step": 452 }, { "epoch": 0.34085778781038373, "grad_norm": 20.681529998779297, "learning_rate": 6.949286846275754e-06, "loss": 2.4619, "step": 453 }, { "epoch": 0.3416102332580888, "grad_norm": 20.630233764648438, "learning_rate": 6.94136291600634e-06, "loss": 2.6602, "step": 454 }, { "epoch": 0.34236267870579384, "grad_norm": 45.344696044921875, "learning_rate": 6.933438985736925e-06, "loss": 2.5986, "step": 455 }, { "epoch": 0.3431151241534989, "grad_norm": 32.19722366333008, "learning_rate": 6.925515055467512e-06, "loss": 2.791, "step": 456 }, { "epoch": 0.3438675696012039, "grad_norm": 26.99687385559082, "learning_rate": 6.9175911251980985e-06, "loss": 2.3369, "step": 457 }, { "epoch": 0.34462001504890893, "grad_norm": 18.11928367614746, "learning_rate": 6.9096671949286855e-06, "loss": 2.8008, "step": 458 }, { "epoch": 0.345372460496614, "grad_norm": 26.664819717407227, "learning_rate": 6.901743264659272e-06, "loss": 3.1514, "step": 459 }, { "epoch": 0.34612490594431905, "grad_norm": 19.563344955444336, "learning_rate": 6.893819334389857e-06, "loss": 2.3389, "step": 460 }, { "epoch": 0.3468773513920241, "grad_norm": 23.58795928955078, "learning_rate": 6.885895404120444e-06, "loss": 2.8252, "step": 461 }, { "epoch": 0.3476297968397291, "grad_norm": 30.469026565551758, "learning_rate": 6.87797147385103e-06, "loss": 2.6855, "step": 462 }, { "epoch": 0.34838224228743414, "grad_norm": 28.226003646850586, "learning_rate": 6.870047543581617e-06, "loss": 2.8213, "step": 463 }, { "epoch": 0.3491346877351392, "grad_norm": 25.442476272583008, "learning_rate": 6.8621236133122035e-06, "loss": 2.2754, "step": 464 }, { "epoch": 0.34988713318284426, "grad_norm": 37.09988021850586, "learning_rate": 6.854199683042789e-06, "loss": 2.9346, "step": 465 }, { "epoch": 0.3506395786305493, "grad_norm": 22.494609832763672, "learning_rate": 6.846275752773376e-06, "loss": 2.9346, "step": 466 }, { "epoch": 0.3513920240782543, "grad_norm": 24.106929779052734, "learning_rate": 6.838351822503962e-06, "loss": 2.8359, "step": 467 }, { "epoch": 0.35214446952595935, "grad_norm": 40.518829345703125, "learning_rate": 6.830427892234549e-06, "loss": 2.3438, "step": 468 }, { "epoch": 0.35289691497366443, "grad_norm": 17.524858474731445, "learning_rate": 6.822503961965135e-06, "loss": 2.2739, "step": 469 }, { "epoch": 0.35364936042136946, "grad_norm": 27.322208404541016, "learning_rate": 6.8145800316957216e-06, "loss": 3.1016, "step": 470 }, { "epoch": 0.3544018058690745, "grad_norm": 24.024763107299805, "learning_rate": 6.806656101426308e-06, "loss": 3.291, "step": 471 }, { "epoch": 0.3551542513167795, "grad_norm": 20.18027114868164, "learning_rate": 6.798732171156894e-06, "loss": 2.4229, "step": 472 }, { "epoch": 0.35590669676448455, "grad_norm": 26.4113826751709, "learning_rate": 6.790808240887481e-06, "loss": 2.5786, "step": 473 }, { "epoch": 0.35665914221218964, "grad_norm": 21.726591110229492, "learning_rate": 6.782884310618067e-06, "loss": 2.4438, "step": 474 }, { "epoch": 0.35741158765989467, "grad_norm": 28.190261840820312, "learning_rate": 6.774960380348653e-06, "loss": 3.1172, "step": 475 }, { "epoch": 0.3581640331075997, "grad_norm": 16.553571701049805, "learning_rate": 6.7670364500792396e-06, "loss": 2.7676, "step": 476 }, { "epoch": 0.35891647855530473, "grad_norm": 31.165143966674805, "learning_rate": 6.759112519809827e-06, "loss": 2.6465, "step": 477 }, { "epoch": 0.35966892400300976, "grad_norm": 33.22994613647461, "learning_rate": 6.751188589540413e-06, "loss": 2.7314, "step": 478 }, { "epoch": 0.36042136945071485, "grad_norm": 22.18958854675293, "learning_rate": 6.743264659270999e-06, "loss": 2.7051, "step": 479 }, { "epoch": 0.3611738148984199, "grad_norm": 35.11039733886719, "learning_rate": 6.735340729001585e-06, "loss": 2.9121, "step": 480 }, { "epoch": 0.3619262603461249, "grad_norm": 52.86737060546875, "learning_rate": 6.727416798732171e-06, "loss": 3.2871, "step": 481 }, { "epoch": 0.36267870579382994, "grad_norm": 31.692039489746094, "learning_rate": 6.719492868462758e-06, "loss": 2.6289, "step": 482 }, { "epoch": 0.36343115124153497, "grad_norm": 28.485288619995117, "learning_rate": 6.711568938193345e-06, "loss": 2.6816, "step": 483 }, { "epoch": 0.36418359668924005, "grad_norm": 45.19233703613281, "learning_rate": 6.703645007923932e-06, "loss": 2.7295, "step": 484 }, { "epoch": 0.3649360421369451, "grad_norm": 21.647083282470703, "learning_rate": 6.695721077654517e-06, "loss": 2.9443, "step": 485 }, { "epoch": 0.3656884875846501, "grad_norm": 45.4495735168457, "learning_rate": 6.687797147385103e-06, "loss": 3.082, "step": 486 }, { "epoch": 0.36644093303235514, "grad_norm": 28.85840606689453, "learning_rate": 6.67987321711569e-06, "loss": 2.8994, "step": 487 }, { "epoch": 0.3671933784800602, "grad_norm": 34.27227020263672, "learning_rate": 6.671949286846276e-06, "loss": 2.832, "step": 488 }, { "epoch": 0.36794582392776526, "grad_norm": 63.276527404785156, "learning_rate": 6.6640253565768635e-06, "loss": 3.7012, "step": 489 }, { "epoch": 0.3686982693754703, "grad_norm": 56.69678497314453, "learning_rate": 6.656101426307449e-06, "loss": 3.2578, "step": 490 }, { "epoch": 0.3694507148231753, "grad_norm": 24.20551109313965, "learning_rate": 6.648177496038035e-06, "loss": 2.3926, "step": 491 }, { "epoch": 0.37020316027088035, "grad_norm": 22.95747947692871, "learning_rate": 6.640253565768622e-06, "loss": 2.4639, "step": 492 }, { "epoch": 0.3709556057185854, "grad_norm": 69.18743896484375, "learning_rate": 6.632329635499208e-06, "loss": 2.7012, "step": 493 }, { "epoch": 0.37170805116629047, "grad_norm": 70.50599670410156, "learning_rate": 6.624405705229795e-06, "loss": 3.1016, "step": 494 }, { "epoch": 0.3724604966139955, "grad_norm": 49.86539077758789, "learning_rate": 6.616481774960381e-06, "loss": 2.7959, "step": 495 }, { "epoch": 0.3732129420617005, "grad_norm": 24.0435848236084, "learning_rate": 6.608557844690967e-06, "loss": 2.502, "step": 496 }, { "epoch": 0.37396538750940556, "grad_norm": 21.732086181640625, "learning_rate": 6.600633914421554e-06, "loss": 3.2559, "step": 497 }, { "epoch": 0.3747178329571106, "grad_norm": 36.107078552246094, "learning_rate": 6.59270998415214e-06, "loss": 2.8994, "step": 498 }, { "epoch": 0.37547027840481567, "grad_norm": 56.2667236328125, "learning_rate": 6.584786053882727e-06, "loss": 2.8262, "step": 499 }, { "epoch": 0.3762227238525207, "grad_norm": 39.14205551147461, "learning_rate": 6.5768621236133124e-06, "loss": 2.4551, "step": 500 }, { "epoch": 0.37697516930022573, "grad_norm": 19.946632385253906, "learning_rate": 6.568938193343899e-06, "loss": 2.6494, "step": 501 }, { "epoch": 0.37772761474793076, "grad_norm": 22.237207412719727, "learning_rate": 6.561014263074486e-06, "loss": 2.1465, "step": 502 }, { "epoch": 0.3784800601956358, "grad_norm": 32.88526916503906, "learning_rate": 6.553090332805072e-06, "loss": 2.4648, "step": 503 }, { "epoch": 0.3792325056433409, "grad_norm": 17.94049644470215, "learning_rate": 6.545166402535659e-06, "loss": 2.2949, "step": 504 }, { "epoch": 0.3799849510910459, "grad_norm": 53.39393615722656, "learning_rate": 6.537242472266244e-06, "loss": 3.2441, "step": 505 }, { "epoch": 0.38073739653875094, "grad_norm": 28.436603546142578, "learning_rate": 6.5293185419968304e-06, "loss": 2.377, "step": 506 }, { "epoch": 0.38148984198645597, "grad_norm": 37.433231353759766, "learning_rate": 6.5213946117274175e-06, "loss": 2.5381, "step": 507 }, { "epoch": 0.382242287434161, "grad_norm": 25.794477462768555, "learning_rate": 6.513470681458004e-06, "loss": 2.3105, "step": 508 }, { "epoch": 0.3829947328818661, "grad_norm": 49.73271560668945, "learning_rate": 6.505546751188591e-06, "loss": 3.2764, "step": 509 }, { "epoch": 0.3837471783295711, "grad_norm": 41.859580993652344, "learning_rate": 6.497622820919176e-06, "loss": 2.9141, "step": 510 }, { "epoch": 0.38449962377727614, "grad_norm": 47.98792266845703, "learning_rate": 6.489698890649762e-06, "loss": 2.7812, "step": 511 }, { "epoch": 0.3852520692249812, "grad_norm": 23.1630802154541, "learning_rate": 6.481774960380349e-06, "loss": 2.377, "step": 512 }, { "epoch": 0.3860045146726862, "grad_norm": 41.17527770996094, "learning_rate": 6.4738510301109355e-06, "loss": 2.959, "step": 513 }, { "epoch": 0.3867569601203913, "grad_norm": 35.60078048706055, "learning_rate": 6.4659270998415225e-06, "loss": 2.6143, "step": 514 }, { "epoch": 0.3875094055680963, "grad_norm": 26.28261375427246, "learning_rate": 6.458003169572108e-06, "loss": 2.8418, "step": 515 }, { "epoch": 0.38826185101580135, "grad_norm": 23.71259307861328, "learning_rate": 6.450079239302695e-06, "loss": 2.8457, "step": 516 }, { "epoch": 0.3890142964635064, "grad_norm": 26.641910552978516, "learning_rate": 6.442155309033281e-06, "loss": 2.6387, "step": 517 }, { "epoch": 0.3897667419112114, "grad_norm": 49.22462844848633, "learning_rate": 6.434231378763868e-06, "loss": 2.8408, "step": 518 }, { "epoch": 0.3905191873589165, "grad_norm": 27.4183406829834, "learning_rate": 6.426307448494454e-06, "loss": 2.168, "step": 519 }, { "epoch": 0.3912716328066215, "grad_norm": 21.35692596435547, "learning_rate": 6.41838351822504e-06, "loss": 2.6094, "step": 520 }, { "epoch": 0.39202407825432656, "grad_norm": 23.8715763092041, "learning_rate": 6.410459587955627e-06, "loss": 2.5928, "step": 521 }, { "epoch": 0.3927765237020316, "grad_norm": 76.4928970336914, "learning_rate": 6.402535657686213e-06, "loss": 3.166, "step": 522 }, { "epoch": 0.3935289691497366, "grad_norm": 44.17985916137695, "learning_rate": 6.3946117274168e-06, "loss": 2.8154, "step": 523 }, { "epoch": 0.3942814145974417, "grad_norm": 21.460453033447266, "learning_rate": 6.386687797147385e-06, "loss": 2.8242, "step": 524 }, { "epoch": 0.39503386004514673, "grad_norm": 28.290786743164062, "learning_rate": 6.3787638668779715e-06, "loss": 2.4619, "step": 525 }, { "epoch": 0.39578630549285176, "grad_norm": 21.564462661743164, "learning_rate": 6.3708399366085585e-06, "loss": 2.665, "step": 526 }, { "epoch": 0.3965387509405568, "grad_norm": 37.88861083984375, "learning_rate": 6.362916006339145e-06, "loss": 2.8418, "step": 527 }, { "epoch": 0.3972911963882618, "grad_norm": 24.181049346923828, "learning_rate": 6.354992076069732e-06, "loss": 2.3916, "step": 528 }, { "epoch": 0.3980436418359669, "grad_norm": 27.649242401123047, "learning_rate": 6.347068145800317e-06, "loss": 2.7217, "step": 529 }, { "epoch": 0.39879608728367194, "grad_norm": 26.146526336669922, "learning_rate": 6.339144215530903e-06, "loss": 3.3066, "step": 530 }, { "epoch": 0.39954853273137697, "grad_norm": 24.51007843017578, "learning_rate": 6.33122028526149e-06, "loss": 2.6338, "step": 531 }, { "epoch": 0.400300978179082, "grad_norm": 44.78768539428711, "learning_rate": 6.3232963549920765e-06, "loss": 2.9102, "step": 532 }, { "epoch": 0.40105342362678703, "grad_norm": 31.947906494140625, "learning_rate": 6.3153724247226636e-06, "loss": 2.7119, "step": 533 }, { "epoch": 0.4018058690744921, "grad_norm": 20.330984115600586, "learning_rate": 6.307448494453249e-06, "loss": 2.332, "step": 534 }, { "epoch": 0.40255831452219715, "grad_norm": 18.617891311645508, "learning_rate": 6.299524564183835e-06, "loss": 2.416, "step": 535 }, { "epoch": 0.4033107599699022, "grad_norm": 31.218690872192383, "learning_rate": 6.291600633914422e-06, "loss": 2.6289, "step": 536 }, { "epoch": 0.4040632054176072, "grad_norm": 40.97818374633789, "learning_rate": 6.283676703645008e-06, "loss": 2.3281, "step": 537 }, { "epoch": 0.40481565086531224, "grad_norm": 44.885047912597656, "learning_rate": 6.275752773375595e-06, "loss": 3.1621, "step": 538 }, { "epoch": 0.4055680963130173, "grad_norm": 28.18929100036621, "learning_rate": 6.267828843106181e-06, "loss": 2.3223, "step": 539 }, { "epoch": 0.40632054176072235, "grad_norm": 24.03481674194336, "learning_rate": 6.259904912836767e-06, "loss": 2.5801, "step": 540 }, { "epoch": 0.4070729872084274, "grad_norm": 21.83283805847168, "learning_rate": 6.251980982567354e-06, "loss": 2.6436, "step": 541 }, { "epoch": 0.4078254326561324, "grad_norm": 19.02834129333496, "learning_rate": 6.24405705229794e-06, "loss": 2.5078, "step": 542 }, { "epoch": 0.40857787810383744, "grad_norm": 27.406829833984375, "learning_rate": 6.236133122028527e-06, "loss": 2.6074, "step": 543 }, { "epoch": 0.40933032355154253, "grad_norm": 51.72572708129883, "learning_rate": 6.2282091917591125e-06, "loss": 2.7842, "step": 544 }, { "epoch": 0.41008276899924756, "grad_norm": 51.11612319946289, "learning_rate": 6.220285261489699e-06, "loss": 2.6377, "step": 545 }, { "epoch": 0.4108352144469526, "grad_norm": 27.588716506958008, "learning_rate": 6.212361331220286e-06, "loss": 2.5391, "step": 546 }, { "epoch": 0.4115876598946576, "grad_norm": 39.39955139160156, "learning_rate": 6.204437400950872e-06, "loss": 3.208, "step": 547 }, { "epoch": 0.4123401053423627, "grad_norm": 15.496928215026855, "learning_rate": 6.196513470681459e-06, "loss": 2.146, "step": 548 }, { "epoch": 0.41309255079006774, "grad_norm": 18.776878356933594, "learning_rate": 6.188589540412044e-06, "loss": 2.5137, "step": 549 }, { "epoch": 0.41384499623777277, "grad_norm": 29.392667770385742, "learning_rate": 6.180665610142631e-06, "loss": 2.5322, "step": 550 }, { "epoch": 0.4145974416854778, "grad_norm": 21.60759925842285, "learning_rate": 6.172741679873218e-06, "loss": 2.6914, "step": 551 }, { "epoch": 0.4153498871331828, "grad_norm": 27.371747970581055, "learning_rate": 6.164817749603804e-06, "loss": 2.6299, "step": 552 }, { "epoch": 0.4161023325808879, "grad_norm": 18.83663558959961, "learning_rate": 6.156893819334391e-06, "loss": 2.1299, "step": 553 }, { "epoch": 0.41685477802859294, "grad_norm": 20.581167221069336, "learning_rate": 6.148969889064976e-06, "loss": 2.3896, "step": 554 }, { "epoch": 0.417607223476298, "grad_norm": 28.915456771850586, "learning_rate": 6.141045958795563e-06, "loss": 2.8584, "step": 555 }, { "epoch": 0.418359668924003, "grad_norm": 27.82697105407715, "learning_rate": 6.133122028526149e-06, "loss": 2.2695, "step": 556 }, { "epoch": 0.41911211437170803, "grad_norm": 35.71852111816406, "learning_rate": 6.1251980982567364e-06, "loss": 2.6367, "step": 557 }, { "epoch": 0.4198645598194131, "grad_norm": 30.599428176879883, "learning_rate": 6.117274167987323e-06, "loss": 3.1572, "step": 558 }, { "epoch": 0.42061700526711815, "grad_norm": 33.88134002685547, "learning_rate": 6.109350237717908e-06, "loss": 2.3701, "step": 559 }, { "epoch": 0.4213694507148232, "grad_norm": 35.95499038696289, "learning_rate": 6.101426307448495e-06, "loss": 2.7178, "step": 560 }, { "epoch": 0.4221218961625282, "grad_norm": 25.86326026916504, "learning_rate": 6.093502377179081e-06, "loss": 2.1943, "step": 561 }, { "epoch": 0.42287434161023324, "grad_norm": 51.53483200073242, "learning_rate": 6.085578446909668e-06, "loss": 3.207, "step": 562 }, { "epoch": 0.4236267870579383, "grad_norm": 27.713850021362305, "learning_rate": 6.0776545166402544e-06, "loss": 2.7949, "step": 563 }, { "epoch": 0.42437923250564336, "grad_norm": 31.49678611755371, "learning_rate": 6.06973058637084e-06, "loss": 2.6533, "step": 564 }, { "epoch": 0.4251316779533484, "grad_norm": 32.17995071411133, "learning_rate": 6.061806656101427e-06, "loss": 2.8281, "step": 565 }, { "epoch": 0.4258841234010534, "grad_norm": 27.868188858032227, "learning_rate": 6.053882725832013e-06, "loss": 2.1934, "step": 566 }, { "epoch": 0.42663656884875845, "grad_norm": 20.03648567199707, "learning_rate": 6.0459587955626e-06, "loss": 2.6455, "step": 567 }, { "epoch": 0.42738901429646353, "grad_norm": 32.74396896362305, "learning_rate": 6.038034865293186e-06, "loss": 2.5557, "step": 568 }, { "epoch": 0.42814145974416856, "grad_norm": 21.329883575439453, "learning_rate": 6.030110935023772e-06, "loss": 2.6055, "step": 569 }, { "epoch": 0.4288939051918736, "grad_norm": 28.882789611816406, "learning_rate": 6.022187004754359e-06, "loss": 2.8438, "step": 570 }, { "epoch": 0.4296463506395786, "grad_norm": 31.306201934814453, "learning_rate": 6.014263074484945e-06, "loss": 3.2129, "step": 571 }, { "epoch": 0.43039879608728365, "grad_norm": 45.59975814819336, "learning_rate": 6.006339144215532e-06, "loss": 2.5244, "step": 572 }, { "epoch": 0.43115124153498874, "grad_norm": 34.31132507324219, "learning_rate": 5.998415213946118e-06, "loss": 2.8154, "step": 573 }, { "epoch": 0.43190368698269377, "grad_norm": 34.59330749511719, "learning_rate": 5.990491283676703e-06, "loss": 2.6465, "step": 574 }, { "epoch": 0.4326561324303988, "grad_norm": 27.188886642456055, "learning_rate": 5.9825673534072905e-06, "loss": 2.9131, "step": 575 }, { "epoch": 0.43340857787810383, "grad_norm": 25.449323654174805, "learning_rate": 5.974643423137877e-06, "loss": 2.7393, "step": 576 }, { "epoch": 0.43416102332580886, "grad_norm": 21.703617095947266, "learning_rate": 5.966719492868464e-06, "loss": 2.7178, "step": 577 }, { "epoch": 0.43491346877351394, "grad_norm": 16.706140518188477, "learning_rate": 5.95879556259905e-06, "loss": 2.0869, "step": 578 }, { "epoch": 0.435665914221219, "grad_norm": 16.854990005493164, "learning_rate": 5.950871632329635e-06, "loss": 2.4287, "step": 579 }, { "epoch": 0.436418359668924, "grad_norm": 36.00377655029297, "learning_rate": 5.942947702060222e-06, "loss": 2.8838, "step": 580 }, { "epoch": 0.43717080511662904, "grad_norm": 44.43525695800781, "learning_rate": 5.9350237717908085e-06, "loss": 2.4482, "step": 581 }, { "epoch": 0.43792325056433407, "grad_norm": 24.75007438659668, "learning_rate": 5.9270998415213955e-06, "loss": 2.5791, "step": 582 }, { "epoch": 0.43867569601203915, "grad_norm": 29.451013565063477, "learning_rate": 5.919175911251982e-06, "loss": 2.5898, "step": 583 }, { "epoch": 0.4394281414597442, "grad_norm": 26.943174362182617, "learning_rate": 5.911251980982568e-06, "loss": 2.542, "step": 584 }, { "epoch": 0.4401805869074492, "grad_norm": 23.72393226623535, "learning_rate": 5.903328050713154e-06, "loss": 2.2354, "step": 585 }, { "epoch": 0.44093303235515424, "grad_norm": 31.409198760986328, "learning_rate": 5.89540412044374e-06, "loss": 2.5586, "step": 586 }, { "epoch": 0.44168547780285927, "grad_norm": 44.14702224731445, "learning_rate": 5.887480190174327e-06, "loss": 2.4023, "step": 587 }, { "epoch": 0.44243792325056436, "grad_norm": 33.159690856933594, "learning_rate": 5.8795562599049135e-06, "loss": 2.4902, "step": 588 }, { "epoch": 0.4431903686982694, "grad_norm": 33.96577072143555, "learning_rate": 5.8716323296355e-06, "loss": 2.9766, "step": 589 }, { "epoch": 0.4439428141459744, "grad_norm": 26.576927185058594, "learning_rate": 5.863708399366086e-06, "loss": 2.541, "step": 590 }, { "epoch": 0.44469525959367945, "grad_norm": 20.750225067138672, "learning_rate": 5.855784469096673e-06, "loss": 2.2998, "step": 591 }, { "epoch": 0.4454477050413845, "grad_norm": 32.86729431152344, "learning_rate": 5.847860538827259e-06, "loss": 2.498, "step": 592 }, { "epoch": 0.44620015048908956, "grad_norm": 26.496009826660156, "learning_rate": 5.839936608557845e-06, "loss": 2.1748, "step": 593 }, { "epoch": 0.4469525959367946, "grad_norm": 28.07709312438965, "learning_rate": 5.8320126782884315e-06, "loss": 2.8682, "step": 594 }, { "epoch": 0.4477050413844996, "grad_norm": 21.558242797851562, "learning_rate": 5.824088748019018e-06, "loss": 2.6094, "step": 595 }, { "epoch": 0.44845748683220465, "grad_norm": 46.38016128540039, "learning_rate": 5.816164817749605e-06, "loss": 2.9346, "step": 596 }, { "epoch": 0.4492099322799097, "grad_norm": 61.013710021972656, "learning_rate": 5.808240887480191e-06, "loss": 2.7021, "step": 597 }, { "epoch": 0.44996237772761477, "grad_norm": 46.63782501220703, "learning_rate": 5.800316957210776e-06, "loss": 3.0039, "step": 598 }, { "epoch": 0.4507148231753198, "grad_norm": 29.919158935546875, "learning_rate": 5.792393026941363e-06, "loss": 2.8818, "step": 599 }, { "epoch": 0.45146726862302483, "grad_norm": 28.168922424316406, "learning_rate": 5.7844690966719495e-06, "loss": 3.1172, "step": 600 }, { "epoch": 0.45221971407072986, "grad_norm": 39.816036224365234, "learning_rate": 5.7765451664025366e-06, "loss": 2.4238, "step": 601 }, { "epoch": 0.4529721595184349, "grad_norm": 21.569623947143555, "learning_rate": 5.768621236133123e-06, "loss": 2.3555, "step": 602 }, { "epoch": 0.45372460496614, "grad_norm": 19.12834358215332, "learning_rate": 5.760697305863708e-06, "loss": 2.7578, "step": 603 }, { "epoch": 0.454477050413845, "grad_norm": 35.17719268798828, "learning_rate": 5.752773375594295e-06, "loss": 2.8389, "step": 604 }, { "epoch": 0.45522949586155004, "grad_norm": 21.667118072509766, "learning_rate": 5.744849445324881e-06, "loss": 2.6777, "step": 605 }, { "epoch": 0.45598194130925507, "grad_norm": 43.21159362792969, "learning_rate": 5.736925515055468e-06, "loss": 2.5498, "step": 606 }, { "epoch": 0.4567343867569601, "grad_norm": 27.8854923248291, "learning_rate": 5.7290015847860546e-06, "loss": 2.793, "step": 607 }, { "epoch": 0.4574868322046652, "grad_norm": 24.826038360595703, "learning_rate": 5.72107765451664e-06, "loss": 2.4551, "step": 608 }, { "epoch": 0.4582392776523702, "grad_norm": 39.36992263793945, "learning_rate": 5.713153724247227e-06, "loss": 2.7695, "step": 609 }, { "epoch": 0.45899172310007524, "grad_norm": 23.196136474609375, "learning_rate": 5.705229793977813e-06, "loss": 2.9688, "step": 610 }, { "epoch": 0.4597441685477803, "grad_norm": 46.75358581542969, "learning_rate": 5.6973058637084e-06, "loss": 2.4863, "step": 611 }, { "epoch": 0.4604966139954853, "grad_norm": 26.586414337158203, "learning_rate": 5.689381933438986e-06, "loss": 2.6416, "step": 612 }, { "epoch": 0.4612490594431904, "grad_norm": 21.940500259399414, "learning_rate": 5.681458003169572e-06, "loss": 2.6592, "step": 613 }, { "epoch": 0.4620015048908954, "grad_norm": 20.193708419799805, "learning_rate": 5.673534072900159e-06, "loss": 3.4102, "step": 614 }, { "epoch": 0.46275395033860045, "grad_norm": 20.76471519470215, "learning_rate": 5.665610142630745e-06, "loss": 2.6309, "step": 615 }, { "epoch": 0.4635063957863055, "grad_norm": 38.27462387084961, "learning_rate": 5.657686212361332e-06, "loss": 2.4277, "step": 616 }, { "epoch": 0.4642588412340105, "grad_norm": 14.277875900268555, "learning_rate": 5.649762282091918e-06, "loss": 2.7598, "step": 617 }, { "epoch": 0.4650112866817156, "grad_norm": 23.343847274780273, "learning_rate": 5.6418383518225035e-06, "loss": 2.6494, "step": 618 }, { "epoch": 0.4657637321294206, "grad_norm": 21.005477905273438, "learning_rate": 5.6339144215530906e-06, "loss": 2.8164, "step": 619 }, { "epoch": 0.46651617757712566, "grad_norm": 43.00020217895508, "learning_rate": 5.625990491283677e-06, "loss": 2.5244, "step": 620 }, { "epoch": 0.4672686230248307, "grad_norm": 35.95077896118164, "learning_rate": 5.618066561014264e-06, "loss": 2.3779, "step": 621 }, { "epoch": 0.4680210684725357, "grad_norm": 38.55207443237305, "learning_rate": 5.61014263074485e-06, "loss": 2.9541, "step": 622 }, { "epoch": 0.4687735139202408, "grad_norm": 28.089881896972656, "learning_rate": 5.602218700475436e-06, "loss": 2.4502, "step": 623 }, { "epoch": 0.46952595936794583, "grad_norm": 22.415802001953125, "learning_rate": 5.594294770206022e-06, "loss": 2.5527, "step": 624 }, { "epoch": 0.47027840481565086, "grad_norm": 20.582796096801758, "learning_rate": 5.586370839936609e-06, "loss": 2.707, "step": 625 }, { "epoch": 0.4710308502633559, "grad_norm": 21.10422134399414, "learning_rate": 5.578446909667196e-06, "loss": 2.6953, "step": 626 }, { "epoch": 0.4717832957110609, "grad_norm": 41.894432067871094, "learning_rate": 5.570522979397782e-06, "loss": 2.8154, "step": 627 }, { "epoch": 0.472535741158766, "grad_norm": 24.479263305664062, "learning_rate": 5.562599049128368e-06, "loss": 2.292, "step": 628 }, { "epoch": 0.47328818660647104, "grad_norm": 16.830801010131836, "learning_rate": 5.554675118858954e-06, "loss": 2.0449, "step": 629 }, { "epoch": 0.47404063205417607, "grad_norm": 17.007564544677734, "learning_rate": 5.546751188589541e-06, "loss": 1.9434, "step": 630 }, { "epoch": 0.4747930775018811, "grad_norm": 16.903120040893555, "learning_rate": 5.5388272583201274e-06, "loss": 1.9873, "step": 631 }, { "epoch": 0.47554552294958613, "grad_norm": 22.137493133544922, "learning_rate": 5.5309033280507145e-06, "loss": 2.7607, "step": 632 }, { "epoch": 0.4762979683972912, "grad_norm": 21.33124542236328, "learning_rate": 5.5229793977813e-06, "loss": 2.1914, "step": 633 }, { "epoch": 0.47705041384499625, "grad_norm": 55.744205474853516, "learning_rate": 5.515055467511886e-06, "loss": 3.4844, "step": 634 }, { "epoch": 0.4778028592927013, "grad_norm": 41.97389602661133, "learning_rate": 5.507131537242473e-06, "loss": 2.457, "step": 635 }, { "epoch": 0.4785553047404063, "grad_norm": 28.83306884765625, "learning_rate": 5.499207606973059e-06, "loss": 2.375, "step": 636 }, { "epoch": 0.47930775018811134, "grad_norm": 31.097944259643555, "learning_rate": 5.491283676703646e-06, "loss": 3.2861, "step": 637 }, { "epoch": 0.4800601956358164, "grad_norm": 20.576139450073242, "learning_rate": 5.483359746434232e-06, "loss": 2.5293, "step": 638 }, { "epoch": 0.48081264108352145, "grad_norm": 21.134475708007812, "learning_rate": 5.475435816164818e-06, "loss": 2.4531, "step": 639 }, { "epoch": 0.4815650865312265, "grad_norm": 29.152189254760742, "learning_rate": 5.467511885895405e-06, "loss": 2.6094, "step": 640 }, { "epoch": 0.4823175319789315, "grad_norm": 19.158748626708984, "learning_rate": 5.459587955625991e-06, "loss": 2.5801, "step": 641 }, { "epoch": 0.48306997742663654, "grad_norm": 50.753875732421875, "learning_rate": 5.451664025356578e-06, "loss": 2.7539, "step": 642 }, { "epoch": 0.48382242287434163, "grad_norm": 44.789730072021484, "learning_rate": 5.4437400950871634e-06, "loss": 2.6309, "step": 643 }, { "epoch": 0.48457486832204666, "grad_norm": 14.936765670776367, "learning_rate": 5.43581616481775e-06, "loss": 2.7637, "step": 644 }, { "epoch": 0.4853273137697517, "grad_norm": 19.86391830444336, "learning_rate": 5.427892234548337e-06, "loss": 2.5547, "step": 645 }, { "epoch": 0.4860797592174567, "grad_norm": 19.42682456970215, "learning_rate": 5.419968304278923e-06, "loss": 2.1025, "step": 646 }, { "epoch": 0.48683220466516175, "grad_norm": 51.899208068847656, "learning_rate": 5.41204437400951e-06, "loss": 2.3691, "step": 647 }, { "epoch": 0.48758465011286684, "grad_norm": 40.9661750793457, "learning_rate": 5.404120443740095e-06, "loss": 2.5078, "step": 648 }, { "epoch": 0.48833709556057187, "grad_norm": 22.135168075561523, "learning_rate": 5.3961965134706814e-06, "loss": 2.019, "step": 649 }, { "epoch": 0.4890895410082769, "grad_norm": 27.344635009765625, "learning_rate": 5.3882725832012685e-06, "loss": 2.835, "step": 650 }, { "epoch": 0.4898419864559819, "grad_norm": 42.776100158691406, "learning_rate": 5.380348652931855e-06, "loss": 2.9199, "step": 651 }, { "epoch": 0.49059443190368696, "grad_norm": 34.481712341308594, "learning_rate": 5.372424722662442e-06, "loss": 3.0693, "step": 652 }, { "epoch": 0.49134687735139204, "grad_norm": 57.45832061767578, "learning_rate": 5.364500792393027e-06, "loss": 2.6128, "step": 653 }, { "epoch": 0.49209932279909707, "grad_norm": 37.97975540161133, "learning_rate": 5.356576862123613e-06, "loss": 2.4619, "step": 654 }, { "epoch": 0.4928517682468021, "grad_norm": 34.988643646240234, "learning_rate": 5.3486529318542e-06, "loss": 2.1592, "step": 655 }, { "epoch": 0.49360421369450713, "grad_norm": 21.3216552734375, "learning_rate": 5.3407290015847865e-06, "loss": 2.5303, "step": 656 }, { "epoch": 0.49435665914221216, "grad_norm": 21.58885955810547, "learning_rate": 5.3328050713153735e-06, "loss": 2.0869, "step": 657 }, { "epoch": 0.49510910458991725, "grad_norm": 41.567867279052734, "learning_rate": 5.324881141045959e-06, "loss": 2.9033, "step": 658 }, { "epoch": 0.4958615500376223, "grad_norm": 52.82680892944336, "learning_rate": 5.316957210776545e-06, "loss": 2.5273, "step": 659 }, { "epoch": 0.4966139954853273, "grad_norm": 35.885414123535156, "learning_rate": 5.309033280507132e-06, "loss": 2.915, "step": 660 }, { "epoch": 0.49736644093303234, "grad_norm": 22.456951141357422, "learning_rate": 5.301109350237718e-06, "loss": 2.0615, "step": 661 }, { "epoch": 0.49811888638073737, "grad_norm": 18.177186965942383, "learning_rate": 5.293185419968305e-06, "loss": 2.1211, "step": 662 }, { "epoch": 0.49887133182844245, "grad_norm": 41.87432098388672, "learning_rate": 5.285261489698891e-06, "loss": 2.4551, "step": 663 }, { "epoch": 0.4996237772761475, "grad_norm": 33.8329963684082, "learning_rate": 5.277337559429478e-06, "loss": 2.9609, "step": 664 }, { "epoch": 0.5003762227238525, "grad_norm": 32.781150817871094, "learning_rate": 5.269413629160064e-06, "loss": 2.5273, "step": 665 }, { "epoch": 0.5011286681715575, "grad_norm": 19.160852432250977, "learning_rate": 5.26148969889065e-06, "loss": 2.7012, "step": 666 }, { "epoch": 0.5018811136192626, "grad_norm": 23.525550842285156, "learning_rate": 5.253565768621236e-06, "loss": 2.6514, "step": 667 }, { "epoch": 0.5026335590669676, "grad_norm": 23.408510208129883, "learning_rate": 5.2456418383518225e-06, "loss": 2.8096, "step": 668 }, { "epoch": 0.5033860045146726, "grad_norm": 34.85614776611328, "learning_rate": 5.2377179080824095e-06, "loss": 2.1074, "step": 669 }, { "epoch": 0.5041384499623778, "grad_norm": 22.156509399414062, "learning_rate": 5.229793977812996e-06, "loss": 2.4199, "step": 670 }, { "epoch": 0.5048908954100828, "grad_norm": 24.354738235473633, "learning_rate": 5.221870047543583e-06, "loss": 2.5723, "step": 671 }, { "epoch": 0.5056433408577878, "grad_norm": 18.33073616027832, "learning_rate": 5.213946117274168e-06, "loss": 2.585, "step": 672 }, { "epoch": 0.5063957863054929, "grad_norm": 34.77322769165039, "learning_rate": 5.206022187004754e-06, "loss": 2.8555, "step": 673 }, { "epoch": 0.5071482317531979, "grad_norm": 15.46507453918457, "learning_rate": 5.198098256735341e-06, "loss": 1.9648, "step": 674 }, { "epoch": 0.5079006772009029, "grad_norm": 20.744842529296875, "learning_rate": 5.1901743264659275e-06, "loss": 2.2227, "step": 675 }, { "epoch": 0.508653122648608, "grad_norm": 26.760475158691406, "learning_rate": 5.182250396196515e-06, "loss": 2.6406, "step": 676 }, { "epoch": 0.509405568096313, "grad_norm": 47.131629943847656, "learning_rate": 5.1743264659271e-06, "loss": 2.5049, "step": 677 }, { "epoch": 0.510158013544018, "grad_norm": 40.88092041015625, "learning_rate": 5.166402535657686e-06, "loss": 2.7178, "step": 678 }, { "epoch": 0.510910458991723, "grad_norm": 44.75162887573242, "learning_rate": 5.158478605388273e-06, "loss": 2.5703, "step": 679 }, { "epoch": 0.5116629044394282, "grad_norm": 24.004650115966797, "learning_rate": 5.150554675118859e-06, "loss": 2.5869, "step": 680 }, { "epoch": 0.5124153498871332, "grad_norm": 29.9807071685791, "learning_rate": 5.142630744849446e-06, "loss": 2.7354, "step": 681 }, { "epoch": 0.5131677953348383, "grad_norm": 26.726299285888672, "learning_rate": 5.134706814580032e-06, "loss": 2.4414, "step": 682 }, { "epoch": 0.5139202407825433, "grad_norm": 23.798734664916992, "learning_rate": 5.126782884310618e-06, "loss": 2.8223, "step": 683 }, { "epoch": 0.5146726862302483, "grad_norm": 45.92054748535156, "learning_rate": 5.118858954041205e-06, "loss": 2.998, "step": 684 }, { "epoch": 0.5154251316779533, "grad_norm": 40.293460845947266, "learning_rate": 5.110935023771791e-06, "loss": 2.1846, "step": 685 }, { "epoch": 0.5161775771256584, "grad_norm": 26.087142944335938, "learning_rate": 5.103011093502378e-06, "loss": 2.582, "step": 686 }, { "epoch": 0.5169300225733634, "grad_norm": 33.69111251831055, "learning_rate": 5.0950871632329636e-06, "loss": 2.2061, "step": 687 }, { "epoch": 0.5176824680210684, "grad_norm": 20.17823600769043, "learning_rate": 5.08716323296355e-06, "loss": 2.2803, "step": 688 }, { "epoch": 0.5184349134687735, "grad_norm": 26.224407196044922, "learning_rate": 5.079239302694137e-06, "loss": 2.6611, "step": 689 }, { "epoch": 0.5191873589164786, "grad_norm": 21.59946632385254, "learning_rate": 5.071315372424723e-06, "loss": 2.8125, "step": 690 }, { "epoch": 0.5199398043641836, "grad_norm": 16.415241241455078, "learning_rate": 5.06339144215531e-06, "loss": 2.0098, "step": 691 }, { "epoch": 0.5206922498118887, "grad_norm": 69.1198501586914, "learning_rate": 5.055467511885895e-06, "loss": 2.7305, "step": 692 }, { "epoch": 0.5214446952595937, "grad_norm": 52.50381851196289, "learning_rate": 5.0475435816164816e-06, "loss": 2.7988, "step": 693 }, { "epoch": 0.5221971407072987, "grad_norm": 43.8442268371582, "learning_rate": 5.039619651347069e-06, "loss": 2.7412, "step": 694 }, { "epoch": 0.5229495861550038, "grad_norm": 23.745641708374023, "learning_rate": 5.031695721077655e-06, "loss": 2.2217, "step": 695 }, { "epoch": 0.5237020316027088, "grad_norm": 21.59581184387207, "learning_rate": 5.023771790808242e-06, "loss": 2.2441, "step": 696 }, { "epoch": 0.5244544770504138, "grad_norm": 41.607452392578125, "learning_rate": 5.015847860538827e-06, "loss": 2.2539, "step": 697 }, { "epoch": 0.5252069224981188, "grad_norm": 44.614837646484375, "learning_rate": 5.007923930269414e-06, "loss": 2.8076, "step": 698 }, { "epoch": 0.5259593679458239, "grad_norm": 31.962142944335938, "learning_rate": 5e-06, "loss": 2.1489, "step": 699 }, { "epoch": 0.526711813393529, "grad_norm": 43.057865142822266, "learning_rate": 4.992076069730587e-06, "loss": 3.002, "step": 700 }, { "epoch": 0.527464258841234, "grad_norm": 31.688989639282227, "learning_rate": 4.984152139461173e-06, "loss": 2.0718, "step": 701 }, { "epoch": 0.5282167042889391, "grad_norm": 35.44694900512695, "learning_rate": 4.97622820919176e-06, "loss": 2.2939, "step": 702 }, { "epoch": 0.5289691497366441, "grad_norm": 54.46326446533203, "learning_rate": 4.968304278922346e-06, "loss": 2.6084, "step": 703 }, { "epoch": 0.5297215951843491, "grad_norm": 22.62406349182129, "learning_rate": 4.960380348652932e-06, "loss": 2.6279, "step": 704 }, { "epoch": 0.5304740406320542, "grad_norm": 37.78304672241211, "learning_rate": 4.952456418383519e-06, "loss": 2.7197, "step": 705 }, { "epoch": 0.5312264860797592, "grad_norm": 29.49541473388672, "learning_rate": 4.944532488114105e-06, "loss": 2.5576, "step": 706 }, { "epoch": 0.5319789315274642, "grad_norm": 19.669097900390625, "learning_rate": 4.936608557844692e-06, "loss": 2.8604, "step": 707 }, { "epoch": 0.5327313769751693, "grad_norm": 38.62272262573242, "learning_rate": 4.928684627575278e-06, "loss": 2.4404, "step": 708 }, { "epoch": 0.5334838224228743, "grad_norm": 21.684125900268555, "learning_rate": 4.920760697305864e-06, "loss": 2.8936, "step": 709 }, { "epoch": 0.5342362678705794, "grad_norm": 17.511150360107422, "learning_rate": 4.912836767036451e-06, "loss": 2.5947, "step": 710 }, { "epoch": 0.5349887133182845, "grad_norm": 33.95096206665039, "learning_rate": 4.904912836767036e-06, "loss": 2.5986, "step": 711 }, { "epoch": 0.5357411587659895, "grad_norm": 16.671566009521484, "learning_rate": 4.8969889064976235e-06, "loss": 2.8828, "step": 712 }, { "epoch": 0.5364936042136945, "grad_norm": 21.556991577148438, "learning_rate": 4.88906497622821e-06, "loss": 2.5596, "step": 713 }, { "epoch": 0.5372460496613995, "grad_norm": 19.081218719482422, "learning_rate": 4.881141045958796e-06, "loss": 2.6641, "step": 714 }, { "epoch": 0.5379984951091046, "grad_norm": 19.19892120361328, "learning_rate": 4.873217115689383e-06, "loss": 2.957, "step": 715 }, { "epoch": 0.5387509405568096, "grad_norm": 20.48495101928711, "learning_rate": 4.865293185419968e-06, "loss": 2.915, "step": 716 }, { "epoch": 0.5395033860045146, "grad_norm": 16.29701805114746, "learning_rate": 4.857369255150555e-06, "loss": 2.5693, "step": 717 }, { "epoch": 0.5402558314522197, "grad_norm": 20.534658432006836, "learning_rate": 4.8494453248811415e-06, "loss": 2.3193, "step": 718 }, { "epoch": 0.5410082768999247, "grad_norm": 20.344722747802734, "learning_rate": 4.841521394611728e-06, "loss": 2.7607, "step": 719 }, { "epoch": 0.5417607223476298, "grad_norm": 50.48695755004883, "learning_rate": 4.833597464342314e-06, "loss": 2.8174, "step": 720 }, { "epoch": 0.5425131677953349, "grad_norm": 21.673259735107422, "learning_rate": 4.825673534072901e-06, "loss": 2.1982, "step": 721 }, { "epoch": 0.5432656132430399, "grad_norm": 20.62663459777832, "learning_rate": 4.817749603803487e-06, "loss": 2.2461, "step": 722 }, { "epoch": 0.5440180586907449, "grad_norm": 29.592178344726562, "learning_rate": 4.809825673534073e-06, "loss": 3.083, "step": 723 }, { "epoch": 0.54477050413845, "grad_norm": 28.092294692993164, "learning_rate": 4.8019017432646595e-06, "loss": 2.4023, "step": 724 }, { "epoch": 0.545522949586155, "grad_norm": 26.026023864746094, "learning_rate": 4.793977812995246e-06, "loss": 2.79, "step": 725 }, { "epoch": 0.54627539503386, "grad_norm": 34.060489654541016, "learning_rate": 4.786053882725833e-06, "loss": 2.8916, "step": 726 }, { "epoch": 0.547027840481565, "grad_norm": 23.83104705810547, "learning_rate": 4.778129952456419e-06, "loss": 2.4453, "step": 727 }, { "epoch": 0.5477802859292701, "grad_norm": 27.824453353881836, "learning_rate": 4.770206022187005e-06, "loss": 3.0391, "step": 728 }, { "epoch": 0.5485327313769752, "grad_norm": 46.81410598754883, "learning_rate": 4.762282091917591e-06, "loss": 2.4746, "step": 729 }, { "epoch": 0.5492851768246803, "grad_norm": 30.927135467529297, "learning_rate": 4.7543581616481775e-06, "loss": 2.6143, "step": 730 }, { "epoch": 0.5500376222723853, "grad_norm": 23.535202026367188, "learning_rate": 4.7464342313787645e-06, "loss": 2.3213, "step": 731 }, { "epoch": 0.5507900677200903, "grad_norm": 21.322975158691406, "learning_rate": 4.738510301109351e-06, "loss": 2.4536, "step": 732 }, { "epoch": 0.5515425131677953, "grad_norm": 22.54486656188965, "learning_rate": 4.730586370839937e-06, "loss": 2.623, "step": 733 }, { "epoch": 0.5522949586155004, "grad_norm": 26.132186889648438, "learning_rate": 4.722662440570523e-06, "loss": 1.9863, "step": 734 }, { "epoch": 0.5530474040632054, "grad_norm": 36.999229431152344, "learning_rate": 4.714738510301109e-06, "loss": 2.8018, "step": 735 }, { "epoch": 0.5537998495109104, "grad_norm": 55.02450180053711, "learning_rate": 4.706814580031696e-06, "loss": 2.1533, "step": 736 }, { "epoch": 0.5545522949586155, "grad_norm": 27.328872680664062, "learning_rate": 4.6988906497622825e-06, "loss": 2.6514, "step": 737 }, { "epoch": 0.5553047404063205, "grad_norm": 29.182422637939453, "learning_rate": 4.690966719492869e-06, "loss": 2.9434, "step": 738 }, { "epoch": 0.5560571858540256, "grad_norm": 23.462890625, "learning_rate": 4.683042789223456e-06, "loss": 2.5322, "step": 739 }, { "epoch": 0.5568096313017307, "grad_norm": 25.650156021118164, "learning_rate": 4.675118858954041e-06, "loss": 3.1553, "step": 740 }, { "epoch": 0.5575620767494357, "grad_norm": 36.816558837890625, "learning_rate": 4.667194928684628e-06, "loss": 2.9004, "step": 741 }, { "epoch": 0.5583145221971407, "grad_norm": 40.148284912109375, "learning_rate": 4.659270998415214e-06, "loss": 2.457, "step": 742 }, { "epoch": 0.5590669676448458, "grad_norm": 20.173446655273438, "learning_rate": 4.6513470681458005e-06, "loss": 2.6436, "step": 743 }, { "epoch": 0.5598194130925508, "grad_norm": 26.01722526550293, "learning_rate": 4.6434231378763876e-06, "loss": 2.5381, "step": 744 }, { "epoch": 0.5605718585402558, "grad_norm": 16.02857208251953, "learning_rate": 4.635499207606973e-06, "loss": 2.0049, "step": 745 }, { "epoch": 0.5613243039879608, "grad_norm": 17.349332809448242, "learning_rate": 4.62757527733756e-06, "loss": 2.4648, "step": 746 }, { "epoch": 0.5620767494356659, "grad_norm": 39.4999885559082, "learning_rate": 4.619651347068146e-06, "loss": 2.707, "step": 747 }, { "epoch": 0.5628291948833709, "grad_norm": 38.11347198486328, "learning_rate": 4.611727416798732e-06, "loss": 2.666, "step": 748 }, { "epoch": 0.563581640331076, "grad_norm": 34.45002365112305, "learning_rate": 4.603803486529319e-06, "loss": 2.8965, "step": 749 }, { "epoch": 0.5643340857787811, "grad_norm": 24.278921127319336, "learning_rate": 4.595879556259905e-06, "loss": 2.9795, "step": 750 }, { "epoch": 0.5650865312264861, "grad_norm": 36.90974426269531, "learning_rate": 4.587955625990492e-06, "loss": 2.7666, "step": 751 }, { "epoch": 0.5658389766741911, "grad_norm": 21.165529251098633, "learning_rate": 4.580031695721078e-06, "loss": 2.478, "step": 752 }, { "epoch": 0.5665914221218962, "grad_norm": 23.80830955505371, "learning_rate": 4.572107765451664e-06, "loss": 2.4609, "step": 753 }, { "epoch": 0.5673438675696012, "grad_norm": 30.803369522094727, "learning_rate": 4.564183835182251e-06, "loss": 2.8076, "step": 754 }, { "epoch": 0.5680963130173062, "grad_norm": 22.102771759033203, "learning_rate": 4.556259904912837e-06, "loss": 2.5107, "step": 755 }, { "epoch": 0.5688487584650113, "grad_norm": 16.512863159179688, "learning_rate": 4.5483359746434236e-06, "loss": 2.1992, "step": 756 }, { "epoch": 0.5696012039127163, "grad_norm": 43.92239761352539, "learning_rate": 4.54041204437401e-06, "loss": 3.0449, "step": 757 }, { "epoch": 0.5703536493604213, "grad_norm": 31.005207061767578, "learning_rate": 4.532488114104596e-06, "loss": 2.6104, "step": 758 }, { "epoch": 0.5711060948081265, "grad_norm": 25.206518173217773, "learning_rate": 4.524564183835183e-06, "loss": 2.3311, "step": 759 }, { "epoch": 0.5718585402558315, "grad_norm": 25.389394760131836, "learning_rate": 4.516640253565769e-06, "loss": 1.9697, "step": 760 }, { "epoch": 0.5726109857035365, "grad_norm": 28.341934204101562, "learning_rate": 4.508716323296355e-06, "loss": 2.7266, "step": 761 }, { "epoch": 0.5733634311512416, "grad_norm": 55.47808074951172, "learning_rate": 4.500792393026942e-06, "loss": 2.7305, "step": 762 }, { "epoch": 0.5741158765989466, "grad_norm": 28.395244598388672, "learning_rate": 4.492868462757528e-06, "loss": 1.937, "step": 763 }, { "epoch": 0.5748683220466516, "grad_norm": 15.90774154663086, "learning_rate": 4.484944532488115e-06, "loss": 2.2666, "step": 764 }, { "epoch": 0.5756207674943566, "grad_norm": 23.520118713378906, "learning_rate": 4.477020602218701e-06, "loss": 2.3945, "step": 765 }, { "epoch": 0.5763732129420617, "grad_norm": 24.288490295410156, "learning_rate": 4.469096671949287e-06, "loss": 2.8779, "step": 766 }, { "epoch": 0.5771256583897667, "grad_norm": 20.855445861816406, "learning_rate": 4.461172741679873e-06, "loss": 2.1201, "step": 767 }, { "epoch": 0.5778781038374717, "grad_norm": 23.80067253112793, "learning_rate": 4.45324881141046e-06, "loss": 2.3896, "step": 768 }, { "epoch": 0.5786305492851769, "grad_norm": 27.069873809814453, "learning_rate": 4.445324881141047e-06, "loss": 3.1094, "step": 769 }, { "epoch": 0.5793829947328819, "grad_norm": 33.97768783569336, "learning_rate": 4.437400950871633e-06, "loss": 2.501, "step": 770 }, { "epoch": 0.5801354401805869, "grad_norm": 24.86118507385254, "learning_rate": 4.429477020602219e-06, "loss": 2.8809, "step": 771 }, { "epoch": 0.580887885628292, "grad_norm": 30.0194091796875, "learning_rate": 4.421553090332805e-06, "loss": 2.0205, "step": 772 }, { "epoch": 0.581640331075997, "grad_norm": 18.77094841003418, "learning_rate": 4.413629160063391e-06, "loss": 2.5732, "step": 773 }, { "epoch": 0.582392776523702, "grad_norm": 28.159244537353516, "learning_rate": 4.4057052297939784e-06, "loss": 2.7314, "step": 774 }, { "epoch": 0.5831452219714071, "grad_norm": 30.579219818115234, "learning_rate": 4.397781299524565e-06, "loss": 2.0508, "step": 775 }, { "epoch": 0.5838976674191121, "grad_norm": 24.77485466003418, "learning_rate": 4.389857369255151e-06, "loss": 2.2637, "step": 776 }, { "epoch": 0.5846501128668171, "grad_norm": 17.51915740966797, "learning_rate": 4.381933438985737e-06, "loss": 2.1367, "step": 777 }, { "epoch": 0.5854025583145221, "grad_norm": 18.201080322265625, "learning_rate": 4.374009508716324e-06, "loss": 2.3232, "step": 778 }, { "epoch": 0.5861550037622273, "grad_norm": 19.63605499267578, "learning_rate": 4.36608557844691e-06, "loss": 2.875, "step": 779 }, { "epoch": 0.5869074492099323, "grad_norm": 27.237627029418945, "learning_rate": 4.3581616481774964e-06, "loss": 2.6475, "step": 780 }, { "epoch": 0.5876598946576373, "grad_norm": 17.824848175048828, "learning_rate": 4.350237717908083e-06, "loss": 1.8301, "step": 781 }, { "epoch": 0.5884123401053424, "grad_norm": 21.104175567626953, "learning_rate": 4.342313787638669e-06, "loss": 1.7656, "step": 782 }, { "epoch": 0.5891647855530474, "grad_norm": 23.795610427856445, "learning_rate": 4.334389857369256e-06, "loss": 3.1006, "step": 783 }, { "epoch": 0.5899172310007524, "grad_norm": 23.139272689819336, "learning_rate": 4.326465927099842e-06, "loss": 1.9956, "step": 784 }, { "epoch": 0.5906696764484575, "grad_norm": 25.720857620239258, "learning_rate": 4.318541996830428e-06, "loss": 1.9121, "step": 785 }, { "epoch": 0.5914221218961625, "grad_norm": 21.973257064819336, "learning_rate": 4.3106180665610144e-06, "loss": 2.7388, "step": 786 }, { "epoch": 0.5921745673438675, "grad_norm": 30.504762649536133, "learning_rate": 4.302694136291601e-06, "loss": 2.4004, "step": 787 }, { "epoch": 0.5929270127915726, "grad_norm": 21.440141677856445, "learning_rate": 4.294770206022188e-06, "loss": 2.5205, "step": 788 }, { "epoch": 0.5936794582392777, "grad_norm": 19.103830337524414, "learning_rate": 4.286846275752774e-06, "loss": 2.1895, "step": 789 }, { "epoch": 0.5944319036869827, "grad_norm": 25.186803817749023, "learning_rate": 4.27892234548336e-06, "loss": 2.9531, "step": 790 }, { "epoch": 0.5951843491346878, "grad_norm": 23.751972198486328, "learning_rate": 4.270998415213946e-06, "loss": 2.5225, "step": 791 }, { "epoch": 0.5959367945823928, "grad_norm": 30.363698959350586, "learning_rate": 4.2630744849445325e-06, "loss": 3.0273, "step": 792 }, { "epoch": 0.5966892400300978, "grad_norm": 28.2047061920166, "learning_rate": 4.2551505546751195e-06, "loss": 2.4619, "step": 793 }, { "epoch": 0.5974416854778029, "grad_norm": 18.83759880065918, "learning_rate": 4.247226624405706e-06, "loss": 2.3926, "step": 794 }, { "epoch": 0.5981941309255079, "grad_norm": 16.413726806640625, "learning_rate": 4.239302694136292e-06, "loss": 2.1045, "step": 795 }, { "epoch": 0.5989465763732129, "grad_norm": 24.916812896728516, "learning_rate": 4.231378763866879e-06, "loss": 2.0381, "step": 796 }, { "epoch": 0.5996990218209179, "grad_norm": 22.680103302001953, "learning_rate": 4.223454833597464e-06, "loss": 2.5254, "step": 797 }, { "epoch": 0.600451467268623, "grad_norm": 17.02437400817871, "learning_rate": 4.215530903328051e-06, "loss": 2.4658, "step": 798 }, { "epoch": 0.6012039127163281, "grad_norm": 40.846778869628906, "learning_rate": 4.2076069730586375e-06, "loss": 2.2783, "step": 799 }, { "epoch": 0.6019563581640331, "grad_norm": 32.31572723388672, "learning_rate": 4.199683042789224e-06, "loss": 2.0996, "step": 800 }, { "epoch": 0.6027088036117382, "grad_norm": 28.528690338134766, "learning_rate": 4.191759112519811e-06, "loss": 2.1387, "step": 801 }, { "epoch": 0.6034612490594432, "grad_norm": 19.70213508605957, "learning_rate": 4.183835182250396e-06, "loss": 2.2217, "step": 802 }, { "epoch": 0.6042136945071482, "grad_norm": 33.513206481933594, "learning_rate": 4.175911251980983e-06, "loss": 2.9395, "step": 803 }, { "epoch": 0.6049661399548533, "grad_norm": 25.72749900817871, "learning_rate": 4.167987321711569e-06, "loss": 2.7637, "step": 804 }, { "epoch": 0.6057185854025583, "grad_norm": 51.15900421142578, "learning_rate": 4.1600633914421555e-06, "loss": 2.708, "step": 805 }, { "epoch": 0.6064710308502633, "grad_norm": 36.24783706665039, "learning_rate": 4.1521394611727425e-06, "loss": 2.4258, "step": 806 }, { "epoch": 0.6072234762979684, "grad_norm": 22.41239356994629, "learning_rate": 4.144215530903328e-06, "loss": 2.4014, "step": 807 }, { "epoch": 0.6079759217456734, "grad_norm": 27.632205963134766, "learning_rate": 4.136291600633915e-06, "loss": 2.1709, "step": 808 }, { "epoch": 0.6087283671933785, "grad_norm": 24.77882194519043, "learning_rate": 4.128367670364501e-06, "loss": 2.5947, "step": 809 }, { "epoch": 0.6094808126410836, "grad_norm": 18.057119369506836, "learning_rate": 4.120443740095087e-06, "loss": 2.4512, "step": 810 }, { "epoch": 0.6102332580887886, "grad_norm": 23.234073638916016, "learning_rate": 4.112519809825674e-06, "loss": 2.5381, "step": 811 }, { "epoch": 0.6109857035364936, "grad_norm": 31.53257179260254, "learning_rate": 4.1045958795562605e-06, "loss": 2.2202, "step": 812 }, { "epoch": 0.6117381489841986, "grad_norm": 23.6964168548584, "learning_rate": 4.096671949286847e-06, "loss": 2.2139, "step": 813 }, { "epoch": 0.6124905944319037, "grad_norm": 31.621091842651367, "learning_rate": 4.088748019017433e-06, "loss": 2.8975, "step": 814 }, { "epoch": 0.6132430398796087, "grad_norm": 28.83985137939453, "learning_rate": 4.080824088748019e-06, "loss": 2.8174, "step": 815 }, { "epoch": 0.6139954853273137, "grad_norm": 19.182252883911133, "learning_rate": 4.072900158478606e-06, "loss": 2.4541, "step": 816 }, { "epoch": 0.6147479307750188, "grad_norm": 16.74087142944336, "learning_rate": 4.064976228209192e-06, "loss": 2.0703, "step": 817 }, { "epoch": 0.6155003762227238, "grad_norm": 22.0015869140625, "learning_rate": 4.0570522979397786e-06, "loss": 2.2656, "step": 818 }, { "epoch": 0.6162528216704289, "grad_norm": 40.94550323486328, "learning_rate": 4.049128367670365e-06, "loss": 2.499, "step": 819 }, { "epoch": 0.617005267118134, "grad_norm": 18.632471084594727, "learning_rate": 4.041204437400951e-06, "loss": 2.1719, "step": 820 }, { "epoch": 0.617757712565839, "grad_norm": 30.165760040283203, "learning_rate": 4.033280507131538e-06, "loss": 2.5557, "step": 821 }, { "epoch": 0.618510158013544, "grad_norm": 25.652528762817383, "learning_rate": 4.025356576862124e-06, "loss": 2.7559, "step": 822 }, { "epoch": 0.6192626034612491, "grad_norm": 27.220035552978516, "learning_rate": 4.01743264659271e-06, "loss": 2.9453, "step": 823 }, { "epoch": 0.6200150489089541, "grad_norm": 22.394641876220703, "learning_rate": 4.0095087163232966e-06, "loss": 2.2749, "step": 824 }, { "epoch": 0.6207674943566591, "grad_norm": 22.86411476135254, "learning_rate": 4.001584786053883e-06, "loss": 2.6201, "step": 825 }, { "epoch": 0.6215199398043642, "grad_norm": 20.468843460083008, "learning_rate": 3.993660855784469e-06, "loss": 1.8496, "step": 826 }, { "epoch": 0.6222723852520692, "grad_norm": 21.8277530670166, "learning_rate": 3.985736925515056e-06, "loss": 1.79, "step": 827 }, { "epoch": 0.6230248306997742, "grad_norm": 20.242324829101562, "learning_rate": 3.977812995245642e-06, "loss": 2.3403, "step": 828 }, { "epoch": 0.6237772761474794, "grad_norm": 25.844030380249023, "learning_rate": 3.969889064976228e-06, "loss": 2.2485, "step": 829 }, { "epoch": 0.6245297215951844, "grad_norm": 42.171226501464844, "learning_rate": 3.961965134706815e-06, "loss": 2.3018, "step": 830 }, { "epoch": 0.6252821670428894, "grad_norm": 37.05338668823242, "learning_rate": 3.954041204437401e-06, "loss": 2.584, "step": 831 }, { "epoch": 0.6260346124905944, "grad_norm": 17.200483322143555, "learning_rate": 3.946117274167988e-06, "loss": 2.2383, "step": 832 }, { "epoch": 0.6267870579382995, "grad_norm": 39.11799240112305, "learning_rate": 3.938193343898574e-06, "loss": 2.7314, "step": 833 }, { "epoch": 0.6275395033860045, "grad_norm": 25.06464958190918, "learning_rate": 3.93026941362916e-06, "loss": 2.7432, "step": 834 }, { "epoch": 0.6282919488337095, "grad_norm": 16.09369468688965, "learning_rate": 3.922345483359747e-06, "loss": 1.9019, "step": 835 }, { "epoch": 0.6290443942814146, "grad_norm": 38.14901351928711, "learning_rate": 3.9144215530903326e-06, "loss": 2.4102, "step": 836 }, { "epoch": 0.6297968397291196, "grad_norm": 18.259464263916016, "learning_rate": 3.90649762282092e-06, "loss": 2.2725, "step": 837 }, { "epoch": 0.6305492851768246, "grad_norm": 48.563934326171875, "learning_rate": 3.898573692551506e-06, "loss": 2.4902, "step": 838 }, { "epoch": 0.6313017306245298, "grad_norm": 35.83158874511719, "learning_rate": 3.890649762282092e-06, "loss": 2.2686, "step": 839 }, { "epoch": 0.6320541760722348, "grad_norm": 25.336288452148438, "learning_rate": 3.882725832012679e-06, "loss": 2.1768, "step": 840 }, { "epoch": 0.6328066215199398, "grad_norm": 29.03192901611328, "learning_rate": 3.874801901743264e-06, "loss": 2.5166, "step": 841 }, { "epoch": 0.6335590669676449, "grad_norm": 32.02183151245117, "learning_rate": 3.866877971473851e-06, "loss": 2.5801, "step": 842 }, { "epoch": 0.6343115124153499, "grad_norm": 40.6337776184082, "learning_rate": 3.858954041204438e-06, "loss": 2.1455, "step": 843 }, { "epoch": 0.6350639578630549, "grad_norm": 40.04218292236328, "learning_rate": 3.851030110935024e-06, "loss": 2.0576, "step": 844 }, { "epoch": 0.63581640331076, "grad_norm": 21.415939331054688, "learning_rate": 3.843106180665611e-06, "loss": 2.1914, "step": 845 }, { "epoch": 0.636568848758465, "grad_norm": 33.20722579956055, "learning_rate": 3.835182250396197e-06, "loss": 2.9111, "step": 846 }, { "epoch": 0.63732129420617, "grad_norm": 47.37984085083008, "learning_rate": 3.827258320126783e-06, "loss": 2.5576, "step": 847 }, { "epoch": 0.6380737396538751, "grad_norm": 51.218971252441406, "learning_rate": 3.8193343898573694e-06, "loss": 3.1221, "step": 848 }, { "epoch": 0.6388261851015802, "grad_norm": 22.271236419677734, "learning_rate": 3.811410459587956e-06, "loss": 2.5859, "step": 849 }, { "epoch": 0.6395786305492852, "grad_norm": 20.053430557250977, "learning_rate": 3.8034865293185427e-06, "loss": 2.4189, "step": 850 }, { "epoch": 0.6403310759969902, "grad_norm": 25.01032257080078, "learning_rate": 3.7955625990491284e-06, "loss": 3.1533, "step": 851 }, { "epoch": 0.6410835214446953, "grad_norm": 18.8804988861084, "learning_rate": 3.787638668779715e-06, "loss": 2.3438, "step": 852 }, { "epoch": 0.6418359668924003, "grad_norm": 20.691516876220703, "learning_rate": 3.7797147385103017e-06, "loss": 2.4365, "step": 853 }, { "epoch": 0.6425884123401053, "grad_norm": 24.400440216064453, "learning_rate": 3.771790808240888e-06, "loss": 2.2988, "step": 854 }, { "epoch": 0.6433408577878104, "grad_norm": 23.2838077545166, "learning_rate": 3.7638668779714745e-06, "loss": 2.0371, "step": 855 }, { "epoch": 0.6440933032355154, "grad_norm": 28.432334899902344, "learning_rate": 3.7559429477020602e-06, "loss": 1.9199, "step": 856 }, { "epoch": 0.6448457486832204, "grad_norm": 18.364774703979492, "learning_rate": 3.748019017432647e-06, "loss": 2.3975, "step": 857 }, { "epoch": 0.6455981941309256, "grad_norm": 26.896156311035156, "learning_rate": 3.740095087163233e-06, "loss": 2.3828, "step": 858 }, { "epoch": 0.6463506395786306, "grad_norm": 20.20989227294922, "learning_rate": 3.7321711568938197e-06, "loss": 2.9043, "step": 859 }, { "epoch": 0.6471030850263356, "grad_norm": 51.84599685668945, "learning_rate": 3.7242472266244063e-06, "loss": 2.377, "step": 860 }, { "epoch": 0.6478555304740407, "grad_norm": 26.037067413330078, "learning_rate": 3.716323296354992e-06, "loss": 2.5391, "step": 861 }, { "epoch": 0.6486079759217457, "grad_norm": 36.53365707397461, "learning_rate": 3.7083993660855787e-06, "loss": 2.3242, "step": 862 }, { "epoch": 0.6493604213694507, "grad_norm": 27.17704963684082, "learning_rate": 3.700475435816165e-06, "loss": 2.3184, "step": 863 }, { "epoch": 0.6501128668171557, "grad_norm": 46.70039367675781, "learning_rate": 3.6925515055467515e-06, "loss": 2.2803, "step": 864 }, { "epoch": 0.6508653122648608, "grad_norm": 20.36431121826172, "learning_rate": 3.684627575277338e-06, "loss": 2.4951, "step": 865 }, { "epoch": 0.6516177577125658, "grad_norm": 35.15061950683594, "learning_rate": 3.6767036450079243e-06, "loss": 2.1846, "step": 866 }, { "epoch": 0.6523702031602708, "grad_norm": 25.147287368774414, "learning_rate": 3.6687797147385105e-06, "loss": 2.3579, "step": 867 }, { "epoch": 0.653122648607976, "grad_norm": 23.692909240722656, "learning_rate": 3.6608557844690967e-06, "loss": 2.8662, "step": 868 }, { "epoch": 0.653875094055681, "grad_norm": 28.69312858581543, "learning_rate": 3.6529318541996833e-06, "loss": 2.2891, "step": 869 }, { "epoch": 0.654627539503386, "grad_norm": 18.955467224121094, "learning_rate": 3.64500792393027e-06, "loss": 2.5234, "step": 870 }, { "epoch": 0.6553799849510911, "grad_norm": 16.43120574951172, "learning_rate": 3.637083993660856e-06, "loss": 1.9131, "step": 871 }, { "epoch": 0.6561324303987961, "grad_norm": 25.750568389892578, "learning_rate": 3.6291600633914427e-06, "loss": 2.6816, "step": 872 }, { "epoch": 0.6568848758465011, "grad_norm": 23.512271881103516, "learning_rate": 3.6212361331220285e-06, "loss": 2.207, "step": 873 }, { "epoch": 0.6576373212942062, "grad_norm": 25.66358757019043, "learning_rate": 3.613312202852615e-06, "loss": 2.1592, "step": 874 }, { "epoch": 0.6583897667419112, "grad_norm": 18.43921661376953, "learning_rate": 3.6053882725832017e-06, "loss": 1.6904, "step": 875 }, { "epoch": 0.6591422121896162, "grad_norm": 22.820770263671875, "learning_rate": 3.597464342313788e-06, "loss": 2.502, "step": 876 }, { "epoch": 0.6598946576373212, "grad_norm": 27.273193359375, "learning_rate": 3.5895404120443745e-06, "loss": 2.0859, "step": 877 }, { "epoch": 0.6606471030850264, "grad_norm": 25.108341217041016, "learning_rate": 3.5816164817749603e-06, "loss": 2.7012, "step": 878 }, { "epoch": 0.6613995485327314, "grad_norm": 31.935596466064453, "learning_rate": 3.573692551505547e-06, "loss": 2.5947, "step": 879 }, { "epoch": 0.6621519939804364, "grad_norm": 29.96102523803711, "learning_rate": 3.5657686212361335e-06, "loss": 2.4814, "step": 880 }, { "epoch": 0.6629044394281415, "grad_norm": 30.193965911865234, "learning_rate": 3.5578446909667197e-06, "loss": 2.2949, "step": 881 }, { "epoch": 0.6636568848758465, "grad_norm": 21.395172119140625, "learning_rate": 3.5499207606973063e-06, "loss": 2.4214, "step": 882 }, { "epoch": 0.6644093303235515, "grad_norm": 39.11041259765625, "learning_rate": 3.5419968304278925e-06, "loss": 2.4277, "step": 883 }, { "epoch": 0.6651617757712566, "grad_norm": 39.84598922729492, "learning_rate": 3.5340729001584787e-06, "loss": 2.5557, "step": 884 }, { "epoch": 0.6659142212189616, "grad_norm": 22.32029914855957, "learning_rate": 3.5261489698890653e-06, "loss": 2.6318, "step": 885 }, { "epoch": 0.6666666666666666, "grad_norm": 47.957611083984375, "learning_rate": 3.5182250396196515e-06, "loss": 2.3809, "step": 886 }, { "epoch": 0.6674191121143717, "grad_norm": 18.636598587036133, "learning_rate": 3.510301109350238e-06, "loss": 2.2236, "step": 887 }, { "epoch": 0.6681715575620768, "grad_norm": 26.87550163269043, "learning_rate": 3.5023771790808243e-06, "loss": 1.9717, "step": 888 }, { "epoch": 0.6689240030097818, "grad_norm": 29.201322555541992, "learning_rate": 3.494453248811411e-06, "loss": 2.0947, "step": 889 }, { "epoch": 0.6696764484574869, "grad_norm": 24.85993194580078, "learning_rate": 3.4865293185419976e-06, "loss": 2.4033, "step": 890 }, { "epoch": 0.6704288939051919, "grad_norm": 20.475067138671875, "learning_rate": 3.4786053882725833e-06, "loss": 2.0635, "step": 891 }, { "epoch": 0.6711813393528969, "grad_norm": 27.593063354492188, "learning_rate": 3.47068145800317e-06, "loss": 2.4746, "step": 892 }, { "epoch": 0.671933784800602, "grad_norm": 19.604068756103516, "learning_rate": 3.462757527733756e-06, "loss": 2.3154, "step": 893 }, { "epoch": 0.672686230248307, "grad_norm": 37.860382080078125, "learning_rate": 3.4548335974643428e-06, "loss": 2.1055, "step": 894 }, { "epoch": 0.673438675696012, "grad_norm": 20.184446334838867, "learning_rate": 3.4469096671949285e-06, "loss": 2.1973, "step": 895 }, { "epoch": 0.674191121143717, "grad_norm": 27.288856506347656, "learning_rate": 3.438985736925515e-06, "loss": 2.4014, "step": 896 }, { "epoch": 0.6749435665914221, "grad_norm": 26.36166000366211, "learning_rate": 3.4310618066561018e-06, "loss": 2.4404, "step": 897 }, { "epoch": 0.6756960120391272, "grad_norm": 22.578481674194336, "learning_rate": 3.423137876386688e-06, "loss": 2.3037, "step": 898 }, { "epoch": 0.6764484574868322, "grad_norm": 22.866018295288086, "learning_rate": 3.4152139461172746e-06, "loss": 2.585, "step": 899 }, { "epoch": 0.6772009029345373, "grad_norm": 23.510807037353516, "learning_rate": 3.4072900158478608e-06, "loss": 1.9209, "step": 900 }, { "epoch": 0.6779533483822423, "grad_norm": 41.13738250732422, "learning_rate": 3.399366085578447e-06, "loss": 2.1924, "step": 901 }, { "epoch": 0.6787057938299473, "grad_norm": 22.456920623779297, "learning_rate": 3.3914421553090336e-06, "loss": 2.7236, "step": 902 }, { "epoch": 0.6794582392776524, "grad_norm": 42.51992416381836, "learning_rate": 3.3835182250396198e-06, "loss": 2.707, "step": 903 }, { "epoch": 0.6802106847253574, "grad_norm": 34.998878479003906, "learning_rate": 3.3755942947702064e-06, "loss": 2.7812, "step": 904 }, { "epoch": 0.6809631301730624, "grad_norm": 21.747095108032227, "learning_rate": 3.3676703645007926e-06, "loss": 2.3213, "step": 905 }, { "epoch": 0.6817155756207675, "grad_norm": 18.266437530517578, "learning_rate": 3.359746434231379e-06, "loss": 1.957, "step": 906 }, { "epoch": 0.6824680210684725, "grad_norm": 44.764652252197266, "learning_rate": 3.351822503961966e-06, "loss": 3.2598, "step": 907 }, { "epoch": 0.6832204665161776, "grad_norm": 21.49046516418457, "learning_rate": 3.3438985736925516e-06, "loss": 2.2383, "step": 908 }, { "epoch": 0.6839729119638827, "grad_norm": 20.83407211303711, "learning_rate": 3.335974643423138e-06, "loss": 2.4727, "step": 909 }, { "epoch": 0.6847253574115877, "grad_norm": 52.79618453979492, "learning_rate": 3.3280507131537244e-06, "loss": 2.8594, "step": 910 }, { "epoch": 0.6854778028592927, "grad_norm": 33.63954162597656, "learning_rate": 3.320126782884311e-06, "loss": 2.2773, "step": 911 }, { "epoch": 0.6862302483069977, "grad_norm": 36.19979476928711, "learning_rate": 3.3122028526148976e-06, "loss": 2.6445, "step": 912 }, { "epoch": 0.6869826937547028, "grad_norm": 21.096250534057617, "learning_rate": 3.3042789223454834e-06, "loss": 1.8545, "step": 913 }, { "epoch": 0.6877351392024078, "grad_norm": 22.03937530517578, "learning_rate": 3.29635499207607e-06, "loss": 2.6279, "step": 914 }, { "epoch": 0.6884875846501128, "grad_norm": 31.599443435668945, "learning_rate": 3.2884310618066562e-06, "loss": 2.6084, "step": 915 }, { "epoch": 0.6892400300978179, "grad_norm": 50.962554931640625, "learning_rate": 3.280507131537243e-06, "loss": 3.5176, "step": 916 }, { "epoch": 0.6899924755455229, "grad_norm": 35.13251876831055, "learning_rate": 3.2725832012678294e-06, "loss": 2.2129, "step": 917 }, { "epoch": 0.690744920993228, "grad_norm": 21.519636154174805, "learning_rate": 3.2646592709984152e-06, "loss": 2.6162, "step": 918 }, { "epoch": 0.6914973664409331, "grad_norm": 22.46910285949707, "learning_rate": 3.256735340729002e-06, "loss": 2.4355, "step": 919 }, { "epoch": 0.6922498118886381, "grad_norm": 24.34375, "learning_rate": 3.248811410459588e-06, "loss": 2.3936, "step": 920 }, { "epoch": 0.6930022573363431, "grad_norm": 19.537216186523438, "learning_rate": 3.2408874801901746e-06, "loss": 2.2109, "step": 921 }, { "epoch": 0.6937547027840482, "grad_norm": 25.251739501953125, "learning_rate": 3.2329635499207613e-06, "loss": 2.5303, "step": 922 }, { "epoch": 0.6945071482317532, "grad_norm": 19.39518928527832, "learning_rate": 3.2250396196513475e-06, "loss": 2.082, "step": 923 }, { "epoch": 0.6952595936794582, "grad_norm": 23.059263229370117, "learning_rate": 3.217115689381934e-06, "loss": 2.6348, "step": 924 }, { "epoch": 0.6960120391271633, "grad_norm": 24.483816146850586, "learning_rate": 3.20919175911252e-06, "loss": 2.0283, "step": 925 }, { "epoch": 0.6967644845748683, "grad_norm": 27.044761657714844, "learning_rate": 3.2012678288431065e-06, "loss": 1.9946, "step": 926 }, { "epoch": 0.6975169300225733, "grad_norm": 24.321277618408203, "learning_rate": 3.1933438985736926e-06, "loss": 2.2832, "step": 927 }, { "epoch": 0.6982693754702785, "grad_norm": 24.639503479003906, "learning_rate": 3.1854199683042793e-06, "loss": 2.165, "step": 928 }, { "epoch": 0.6990218209179835, "grad_norm": 21.22592544555664, "learning_rate": 3.177496038034866e-06, "loss": 2.1494, "step": 929 }, { "epoch": 0.6997742663656885, "grad_norm": 22.190229415893555, "learning_rate": 3.1695721077654516e-06, "loss": 2.1587, "step": 930 }, { "epoch": 0.7005267118133935, "grad_norm": 20.478050231933594, "learning_rate": 3.1616481774960383e-06, "loss": 1.9883, "step": 931 }, { "epoch": 0.7012791572610986, "grad_norm": 29.108064651489258, "learning_rate": 3.1537242472266245e-06, "loss": 2.8047, "step": 932 }, { "epoch": 0.7020316027088036, "grad_norm": 21.049943923950195, "learning_rate": 3.145800316957211e-06, "loss": 2.1621, "step": 933 }, { "epoch": 0.7027840481565086, "grad_norm": 35.474388122558594, "learning_rate": 3.1378763866877977e-06, "loss": 2.6064, "step": 934 }, { "epoch": 0.7035364936042137, "grad_norm": 22.91497039794922, "learning_rate": 3.1299524564183835e-06, "loss": 2.1855, "step": 935 }, { "epoch": 0.7042889390519187, "grad_norm": 24.449195861816406, "learning_rate": 3.12202852614897e-06, "loss": 2.0146, "step": 936 }, { "epoch": 0.7050413844996237, "grad_norm": 20.208375930786133, "learning_rate": 3.1141045958795563e-06, "loss": 1.6709, "step": 937 }, { "epoch": 0.7057938299473289, "grad_norm": 37.04838943481445, "learning_rate": 3.106180665610143e-06, "loss": 2.0508, "step": 938 }, { "epoch": 0.7065462753950339, "grad_norm": 26.501577377319336, "learning_rate": 3.0982567353407295e-06, "loss": 2.418, "step": 939 }, { "epoch": 0.7072987208427389, "grad_norm": 27.098785400390625, "learning_rate": 3.0903328050713157e-06, "loss": 2.5312, "step": 940 }, { "epoch": 0.708051166290444, "grad_norm": 33.405418395996094, "learning_rate": 3.082408874801902e-06, "loss": 2.3438, "step": 941 }, { "epoch": 0.708803611738149, "grad_norm": 28.807294845581055, "learning_rate": 3.074484944532488e-06, "loss": 2.6094, "step": 942 }, { "epoch": 0.709556057185854, "grad_norm": 22.320127487182617, "learning_rate": 3.0665610142630747e-06, "loss": 1.645, "step": 943 }, { "epoch": 0.710308502633559, "grad_norm": 35.30049514770508, "learning_rate": 3.0586370839936613e-06, "loss": 2.5449, "step": 944 }, { "epoch": 0.7110609480812641, "grad_norm": 26.332334518432617, "learning_rate": 3.0507131537242475e-06, "loss": 2.6299, "step": 945 }, { "epoch": 0.7118133935289691, "grad_norm": 21.510595321655273, "learning_rate": 3.042789223454834e-06, "loss": 2.1787, "step": 946 }, { "epoch": 0.7125658389766741, "grad_norm": 30.63580322265625, "learning_rate": 3.03486529318542e-06, "loss": 2.3877, "step": 947 }, { "epoch": 0.7133182844243793, "grad_norm": 24.134626388549805, "learning_rate": 3.0269413629160065e-06, "loss": 3.1738, "step": 948 }, { "epoch": 0.7140707298720843, "grad_norm": 25.873851776123047, "learning_rate": 3.019017432646593e-06, "loss": 2.1338, "step": 949 }, { "epoch": 0.7148231753197893, "grad_norm": 20.822885513305664, "learning_rate": 3.0110935023771793e-06, "loss": 2.5898, "step": 950 }, { "epoch": 0.7155756207674944, "grad_norm": 21.63072395324707, "learning_rate": 3.003169572107766e-06, "loss": 2.1758, "step": 951 }, { "epoch": 0.7163280662151994, "grad_norm": 21.866065979003906, "learning_rate": 2.9952456418383517e-06, "loss": 2.1553, "step": 952 }, { "epoch": 0.7170805116629044, "grad_norm": 16.720848083496094, "learning_rate": 2.9873217115689383e-06, "loss": 2.7178, "step": 953 }, { "epoch": 0.7178329571106095, "grad_norm": 17.399744033813477, "learning_rate": 2.979397781299525e-06, "loss": 1.9292, "step": 954 }, { "epoch": 0.7185854025583145, "grad_norm": 18.03335952758789, "learning_rate": 2.971473851030111e-06, "loss": 2.499, "step": 955 }, { "epoch": 0.7193378480060195, "grad_norm": 18.41415023803711, "learning_rate": 2.9635499207606977e-06, "loss": 1.9775, "step": 956 }, { "epoch": 0.7200902934537246, "grad_norm": 21.423734664916992, "learning_rate": 2.955625990491284e-06, "loss": 2.2734, "step": 957 }, { "epoch": 0.7208427389014297, "grad_norm": 20.504459381103516, "learning_rate": 2.94770206022187e-06, "loss": 2.478, "step": 958 }, { "epoch": 0.7215951843491347, "grad_norm": 26.129322052001953, "learning_rate": 2.9397781299524568e-06, "loss": 2.8418, "step": 959 }, { "epoch": 0.7223476297968398, "grad_norm": 23.1035213470459, "learning_rate": 2.931854199683043e-06, "loss": 2.1973, "step": 960 }, { "epoch": 0.7231000752445448, "grad_norm": 35.233028411865234, "learning_rate": 2.9239302694136296e-06, "loss": 2.3066, "step": 961 }, { "epoch": 0.7238525206922498, "grad_norm": 30.950870513916016, "learning_rate": 2.9160063391442158e-06, "loss": 2.5923, "step": 962 }, { "epoch": 0.7246049661399548, "grad_norm": 20.875741958618164, "learning_rate": 2.9080824088748024e-06, "loss": 2.1748, "step": 963 }, { "epoch": 0.7253574115876599, "grad_norm": 33.3520622253418, "learning_rate": 2.900158478605388e-06, "loss": 2.707, "step": 964 }, { "epoch": 0.7261098570353649, "grad_norm": 17.897817611694336, "learning_rate": 2.8922345483359748e-06, "loss": 1.9912, "step": 965 }, { "epoch": 0.7268623024830699, "grad_norm": 23.978641510009766, "learning_rate": 2.8843106180665614e-06, "loss": 2.3223, "step": 966 }, { "epoch": 0.7276147479307751, "grad_norm": 39.2357177734375, "learning_rate": 2.8763866877971476e-06, "loss": 3.0039, "step": 967 }, { "epoch": 0.7283671933784801, "grad_norm": 21.9942569732666, "learning_rate": 2.868462757527734e-06, "loss": 2.4873, "step": 968 }, { "epoch": 0.7291196388261851, "grad_norm": 35.64141845703125, "learning_rate": 2.86053882725832e-06, "loss": 2.7051, "step": 969 }, { "epoch": 0.7298720842738902, "grad_norm": 45.55552673339844, "learning_rate": 2.8526148969889066e-06, "loss": 2.7578, "step": 970 }, { "epoch": 0.7306245297215952, "grad_norm": 19.321287155151367, "learning_rate": 2.844690966719493e-06, "loss": 2.1943, "step": 971 }, { "epoch": 0.7313769751693002, "grad_norm": 24.38201332092285, "learning_rate": 2.8367670364500794e-06, "loss": 2.4873, "step": 972 }, { "epoch": 0.7321294206170053, "grad_norm": 22.028440475463867, "learning_rate": 2.828843106180666e-06, "loss": 2.5156, "step": 973 }, { "epoch": 0.7328818660647103, "grad_norm": 26.298290252685547, "learning_rate": 2.8209191759112518e-06, "loss": 2.7314, "step": 974 }, { "epoch": 0.7336343115124153, "grad_norm": 30.888376235961914, "learning_rate": 2.8129952456418384e-06, "loss": 2.71, "step": 975 }, { "epoch": 0.7343867569601203, "grad_norm": 23.976533889770508, "learning_rate": 2.805071315372425e-06, "loss": 2.3789, "step": 976 }, { "epoch": 0.7351392024078255, "grad_norm": 31.292388916015625, "learning_rate": 2.797147385103011e-06, "loss": 2.5811, "step": 977 }, { "epoch": 0.7358916478555305, "grad_norm": 19.439210891723633, "learning_rate": 2.789223454833598e-06, "loss": 2.6377, "step": 978 }, { "epoch": 0.7366440933032355, "grad_norm": 19.47076988220215, "learning_rate": 2.781299524564184e-06, "loss": 2.5439, "step": 979 }, { "epoch": 0.7373965387509406, "grad_norm": 20.66202735900879, "learning_rate": 2.7733755942947706e-06, "loss": 2.0352, "step": 980 }, { "epoch": 0.7381489841986456, "grad_norm": 24.99919891357422, "learning_rate": 2.7654516640253572e-06, "loss": 2.1938, "step": 981 }, { "epoch": 0.7389014296463506, "grad_norm": 32.048587799072266, "learning_rate": 2.757527733755943e-06, "loss": 2.3311, "step": 982 }, { "epoch": 0.7396538750940557, "grad_norm": 24.181419372558594, "learning_rate": 2.7496038034865296e-06, "loss": 2.415, "step": 983 }, { "epoch": 0.7404063205417607, "grad_norm": 21.071016311645508, "learning_rate": 2.741679873217116e-06, "loss": 2.7002, "step": 984 }, { "epoch": 0.7411587659894657, "grad_norm": 23.08110237121582, "learning_rate": 2.7337559429477024e-06, "loss": 2.6738, "step": 985 }, { "epoch": 0.7419112114371708, "grad_norm": 19.222745895385742, "learning_rate": 2.725832012678289e-06, "loss": 2.168, "step": 986 }, { "epoch": 0.7426636568848759, "grad_norm": 36.517940521240234, "learning_rate": 2.717908082408875e-06, "loss": 1.9395, "step": 987 }, { "epoch": 0.7434161023325809, "grad_norm": 21.429569244384766, "learning_rate": 2.7099841521394614e-06, "loss": 2.1455, "step": 988 }, { "epoch": 0.744168547780286, "grad_norm": 25.69976043701172, "learning_rate": 2.7020602218700476e-06, "loss": 2.5479, "step": 989 }, { "epoch": 0.744920993227991, "grad_norm": 23.321632385253906, "learning_rate": 2.6941362916006342e-06, "loss": 2.5186, "step": 990 }, { "epoch": 0.745673438675696, "grad_norm": 18.7263126373291, "learning_rate": 2.686212361331221e-06, "loss": 2.4336, "step": 991 }, { "epoch": 0.746425884123401, "grad_norm": 23.556350708007812, "learning_rate": 2.6782884310618066e-06, "loss": 2.4292, "step": 992 }, { "epoch": 0.7471783295711061, "grad_norm": 22.186206817626953, "learning_rate": 2.6703645007923932e-06, "loss": 2.4961, "step": 993 }, { "epoch": 0.7479307750188111, "grad_norm": 23.775405883789062, "learning_rate": 2.6624405705229794e-06, "loss": 2.4878, "step": 994 }, { "epoch": 0.7486832204665161, "grad_norm": 33.59183883666992, "learning_rate": 2.654516640253566e-06, "loss": 2.7451, "step": 995 }, { "epoch": 0.7494356659142212, "grad_norm": 18.139806747436523, "learning_rate": 2.6465927099841527e-06, "loss": 2.0146, "step": 996 }, { "epoch": 0.7501881113619263, "grad_norm": 17.442407608032227, "learning_rate": 2.638668779714739e-06, "loss": 1.8535, "step": 997 }, { "epoch": 0.7509405568096313, "grad_norm": 30.57038116455078, "learning_rate": 2.630744849445325e-06, "loss": 2.3447, "step": 998 }, { "epoch": 0.7516930022573364, "grad_norm": 18.132648468017578, "learning_rate": 2.6228209191759112e-06, "loss": 2.1621, "step": 999 }, { "epoch": 0.7524454477050414, "grad_norm": 27.932897567749023, "learning_rate": 2.614896988906498e-06, "loss": 2.4824, "step": 1000 }, { "epoch": 0.7531978931527464, "grad_norm": 33.71088790893555, "learning_rate": 2.606973058637084e-06, "loss": 2.457, "step": 1001 }, { "epoch": 0.7539503386004515, "grad_norm": 20.68466567993164, "learning_rate": 2.5990491283676707e-06, "loss": 2.0396, "step": 1002 }, { "epoch": 0.7547027840481565, "grad_norm": 21.711130142211914, "learning_rate": 2.5911251980982573e-06, "loss": 2.6387, "step": 1003 }, { "epoch": 0.7554552294958615, "grad_norm": 28.704111099243164, "learning_rate": 2.583201267828843e-06, "loss": 2.3721, "step": 1004 }, { "epoch": 0.7562076749435666, "grad_norm": 33.77923583984375, "learning_rate": 2.5752773375594297e-06, "loss": 2.7725, "step": 1005 }, { "epoch": 0.7569601203912716, "grad_norm": 18.9777774810791, "learning_rate": 2.567353407290016e-06, "loss": 2.3325, "step": 1006 }, { "epoch": 0.7577125658389767, "grad_norm": 34.291168212890625, "learning_rate": 2.5594294770206025e-06, "loss": 2.2109, "step": 1007 }, { "epoch": 0.7584650112866818, "grad_norm": 30.99751853942871, "learning_rate": 2.551505546751189e-06, "loss": 2.2041, "step": 1008 }, { "epoch": 0.7592174567343868, "grad_norm": 28.4600830078125, "learning_rate": 2.543581616481775e-06, "loss": 3.1016, "step": 1009 }, { "epoch": 0.7599699021820918, "grad_norm": 32.46207046508789, "learning_rate": 2.5356576862123615e-06, "loss": 2.2295, "step": 1010 }, { "epoch": 0.7607223476297968, "grad_norm": 30.938323974609375, "learning_rate": 2.5277337559429477e-06, "loss": 2.6611, "step": 1011 }, { "epoch": 0.7614747930775019, "grad_norm": 33.81102752685547, "learning_rate": 2.5198098256735343e-06, "loss": 1.8555, "step": 1012 }, { "epoch": 0.7622272385252069, "grad_norm": 33.322418212890625, "learning_rate": 2.511885895404121e-06, "loss": 2.1885, "step": 1013 }, { "epoch": 0.7629796839729119, "grad_norm": 26.505693435668945, "learning_rate": 2.503961965134707e-06, "loss": 2.9521, "step": 1014 }, { "epoch": 0.763732129420617, "grad_norm": 25.791608810424805, "learning_rate": 2.4960380348652933e-06, "loss": 2.3506, "step": 1015 }, { "epoch": 0.764484574868322, "grad_norm": 34.77909851074219, "learning_rate": 2.48811410459588e-06, "loss": 2.5376, "step": 1016 }, { "epoch": 0.7652370203160271, "grad_norm": 22.43770408630371, "learning_rate": 2.480190174326466e-06, "loss": 1.6523, "step": 1017 }, { "epoch": 0.7659894657637322, "grad_norm": 48.59208297729492, "learning_rate": 2.4722662440570523e-06, "loss": 2.3481, "step": 1018 }, { "epoch": 0.7667419112114372, "grad_norm": 19.5473575592041, "learning_rate": 2.464342313787639e-06, "loss": 2.248, "step": 1019 }, { "epoch": 0.7674943566591422, "grad_norm": 17.761960983276367, "learning_rate": 2.4564183835182255e-06, "loss": 1.9951, "step": 1020 }, { "epoch": 0.7682468021068473, "grad_norm": 39.91436004638672, "learning_rate": 2.4484944532488117e-06, "loss": 2.8223, "step": 1021 }, { "epoch": 0.7689992475545523, "grad_norm": 16.507017135620117, "learning_rate": 2.440570522979398e-06, "loss": 2.2949, "step": 1022 }, { "epoch": 0.7697516930022573, "grad_norm": 19.836973190307617, "learning_rate": 2.432646592709984e-06, "loss": 2.1084, "step": 1023 }, { "epoch": 0.7705041384499624, "grad_norm": 24.554914474487305, "learning_rate": 2.4247226624405707e-06, "loss": 2.0625, "step": 1024 }, { "epoch": 0.7712565838976674, "grad_norm": 27.172744750976562, "learning_rate": 2.416798732171157e-06, "loss": 2.2188, "step": 1025 }, { "epoch": 0.7720090293453724, "grad_norm": 29.99898910522461, "learning_rate": 2.4088748019017435e-06, "loss": 2.3184, "step": 1026 }, { "epoch": 0.7727614747930776, "grad_norm": 29.691516876220703, "learning_rate": 2.4009508716323297e-06, "loss": 2.2871, "step": 1027 }, { "epoch": 0.7735139202407826, "grad_norm": 25.524301528930664, "learning_rate": 2.3930269413629164e-06, "loss": 2.543, "step": 1028 }, { "epoch": 0.7742663656884876, "grad_norm": 23.30611801147461, "learning_rate": 2.3851030110935025e-06, "loss": 2.9375, "step": 1029 }, { "epoch": 0.7750188111361926, "grad_norm": 18.776622772216797, "learning_rate": 2.3771790808240887e-06, "loss": 1.8032, "step": 1030 }, { "epoch": 0.7757712565838977, "grad_norm": 24.228599548339844, "learning_rate": 2.3692551505546754e-06, "loss": 2.1113, "step": 1031 }, { "epoch": 0.7765237020316027, "grad_norm": 37.123287200927734, "learning_rate": 2.3613312202852615e-06, "loss": 2.1606, "step": 1032 }, { "epoch": 0.7772761474793077, "grad_norm": 16.47333335876465, "learning_rate": 2.353407290015848e-06, "loss": 1.8906, "step": 1033 }, { "epoch": 0.7780285929270128, "grad_norm": 17.088388442993164, "learning_rate": 2.3454833597464344e-06, "loss": 2.2246, "step": 1034 }, { "epoch": 0.7787810383747178, "grad_norm": 22.337467193603516, "learning_rate": 2.3375594294770205e-06, "loss": 2.2988, "step": 1035 }, { "epoch": 0.7795334838224228, "grad_norm": 20.124755859375, "learning_rate": 2.329635499207607e-06, "loss": 2.3594, "step": 1036 }, { "epoch": 0.780285929270128, "grad_norm": 23.33896255493164, "learning_rate": 2.3217115689381938e-06, "loss": 2.1382, "step": 1037 }, { "epoch": 0.781038374717833, "grad_norm": 23.720806121826172, "learning_rate": 2.31378763866878e-06, "loss": 2.0571, "step": 1038 }, { "epoch": 0.781790820165538, "grad_norm": 17.930727005004883, "learning_rate": 2.305863708399366e-06, "loss": 2.0557, "step": 1039 }, { "epoch": 0.782543265613243, "grad_norm": 27.225309371948242, "learning_rate": 2.2979397781299524e-06, "loss": 2.1133, "step": 1040 }, { "epoch": 0.7832957110609481, "grad_norm": 21.77535057067871, "learning_rate": 2.290015847860539e-06, "loss": 2.0059, "step": 1041 }, { "epoch": 0.7840481565086531, "grad_norm": 31.862289428710938, "learning_rate": 2.2820919175911256e-06, "loss": 2.6641, "step": 1042 }, { "epoch": 0.7848006019563581, "grad_norm": 21.0166072845459, "learning_rate": 2.2741679873217118e-06, "loss": 2.4512, "step": 1043 }, { "epoch": 0.7855530474040632, "grad_norm": 18.94448471069336, "learning_rate": 2.266244057052298e-06, "loss": 2.0381, "step": 1044 }, { "epoch": 0.7863054928517682, "grad_norm": 22.67876625061035, "learning_rate": 2.2583201267828846e-06, "loss": 2.6689, "step": 1045 }, { "epoch": 0.7870579382994732, "grad_norm": 18.799575805664062, "learning_rate": 2.250396196513471e-06, "loss": 2.0874, "step": 1046 }, { "epoch": 0.7878103837471784, "grad_norm": 21.82719612121582, "learning_rate": 2.2424722662440574e-06, "loss": 2.4414, "step": 1047 }, { "epoch": 0.7885628291948834, "grad_norm": 26.506649017333984, "learning_rate": 2.2345483359746436e-06, "loss": 2.5059, "step": 1048 }, { "epoch": 0.7893152746425884, "grad_norm": 19.323801040649414, "learning_rate": 2.22662440570523e-06, "loss": 2.3105, "step": 1049 }, { "epoch": 0.7900677200902935, "grad_norm": 37.666603088378906, "learning_rate": 2.2187004754358164e-06, "loss": 1.9395, "step": 1050 }, { "epoch": 0.7908201655379985, "grad_norm": 19.168920516967773, "learning_rate": 2.2107765451664026e-06, "loss": 2.2466, "step": 1051 }, { "epoch": 0.7915726109857035, "grad_norm": 38.56500244140625, "learning_rate": 2.2028526148969892e-06, "loss": 2.4648, "step": 1052 }, { "epoch": 0.7923250564334086, "grad_norm": 23.99535369873047, "learning_rate": 2.1949286846275754e-06, "loss": 2.876, "step": 1053 }, { "epoch": 0.7930775018811136, "grad_norm": 40.79738998413086, "learning_rate": 2.187004754358162e-06, "loss": 2.6338, "step": 1054 }, { "epoch": 0.7938299473288186, "grad_norm": 42.04122543334961, "learning_rate": 2.1790808240887482e-06, "loss": 2.3599, "step": 1055 }, { "epoch": 0.7945823927765236, "grad_norm": 22.063844680786133, "learning_rate": 2.1711568938193344e-06, "loss": 2.4443, "step": 1056 }, { "epoch": 0.7953348382242288, "grad_norm": 20.444236755371094, "learning_rate": 2.163232963549921e-06, "loss": 2.2334, "step": 1057 }, { "epoch": 0.7960872836719338, "grad_norm": 31.267562866210938, "learning_rate": 2.1553090332805072e-06, "loss": 2.167, "step": 1058 }, { "epoch": 0.7968397291196389, "grad_norm": 18.25832176208496, "learning_rate": 2.147385103011094e-06, "loss": 2.7217, "step": 1059 }, { "epoch": 0.7975921745673439, "grad_norm": 21.88541030883789, "learning_rate": 2.13946117274168e-06, "loss": 2.3711, "step": 1060 }, { "epoch": 0.7983446200150489, "grad_norm": 21.097015380859375, "learning_rate": 2.1315372424722662e-06, "loss": 2.1494, "step": 1061 }, { "epoch": 0.7990970654627539, "grad_norm": 27.271303176879883, "learning_rate": 2.123613312202853e-06, "loss": 2.0898, "step": 1062 }, { "epoch": 0.799849510910459, "grad_norm": 23.769197463989258, "learning_rate": 2.1156893819334395e-06, "loss": 2.5215, "step": 1063 }, { "epoch": 0.800601956358164, "grad_norm": 27.01453399658203, "learning_rate": 2.1077654516640257e-06, "loss": 1.9517, "step": 1064 }, { "epoch": 0.801354401805869, "grad_norm": 21.828453063964844, "learning_rate": 2.099841521394612e-06, "loss": 2.6025, "step": 1065 }, { "epoch": 0.8021068472535741, "grad_norm": 21.228717803955078, "learning_rate": 2.091917591125198e-06, "loss": 2.0664, "step": 1066 }, { "epoch": 0.8028592927012792, "grad_norm": 32.78303909301758, "learning_rate": 2.0839936608557847e-06, "loss": 2.1641, "step": 1067 }, { "epoch": 0.8036117381489842, "grad_norm": 26.406593322753906, "learning_rate": 2.0760697305863713e-06, "loss": 2.2568, "step": 1068 }, { "epoch": 0.8043641835966893, "grad_norm": 26.602384567260742, "learning_rate": 2.0681458003169575e-06, "loss": 2.9897, "step": 1069 }, { "epoch": 0.8051166290443943, "grad_norm": 25.720447540283203, "learning_rate": 2.0602218700475437e-06, "loss": 2.3076, "step": 1070 }, { "epoch": 0.8058690744920993, "grad_norm": 28.323110580444336, "learning_rate": 2.0522979397781303e-06, "loss": 2.4307, "step": 1071 }, { "epoch": 0.8066215199398044, "grad_norm": 21.25274085998535, "learning_rate": 2.0443740095087165e-06, "loss": 2.4961, "step": 1072 }, { "epoch": 0.8073739653875094, "grad_norm": 20.181909561157227, "learning_rate": 2.036450079239303e-06, "loss": 2.2793, "step": 1073 }, { "epoch": 0.8081264108352144, "grad_norm": 19.529109954833984, "learning_rate": 2.0285261489698893e-06, "loss": 2.3984, "step": 1074 }, { "epoch": 0.8088788562829194, "grad_norm": 27.370973587036133, "learning_rate": 2.0206022187004755e-06, "loss": 2.3125, "step": 1075 }, { "epoch": 0.8096313017306245, "grad_norm": 22.67449378967285, "learning_rate": 2.012678288431062e-06, "loss": 2.1079, "step": 1076 }, { "epoch": 0.8103837471783296, "grad_norm": 30.263660430908203, "learning_rate": 2.0047543581616483e-06, "loss": 2.3652, "step": 1077 }, { "epoch": 0.8111361926260346, "grad_norm": 21.763072967529297, "learning_rate": 1.9968304278922345e-06, "loss": 2.3447, "step": 1078 }, { "epoch": 0.8118886380737397, "grad_norm": 26.992006301879883, "learning_rate": 1.988906497622821e-06, "loss": 2.4863, "step": 1079 }, { "epoch": 0.8126410835214447, "grad_norm": 22.929861068725586, "learning_rate": 1.9809825673534077e-06, "loss": 2.2822, "step": 1080 }, { "epoch": 0.8133935289691497, "grad_norm": 22.8551082611084, "learning_rate": 1.973058637083994e-06, "loss": 2.377, "step": 1081 }, { "epoch": 0.8141459744168548, "grad_norm": 26.721590042114258, "learning_rate": 1.96513470681458e-06, "loss": 2.7217, "step": 1082 }, { "epoch": 0.8148984198645598, "grad_norm": 17.116928100585938, "learning_rate": 1.9572107765451663e-06, "loss": 1.8262, "step": 1083 }, { "epoch": 0.8156508653122648, "grad_norm": 28.113954544067383, "learning_rate": 1.949286846275753e-06, "loss": 2.9307, "step": 1084 }, { "epoch": 0.8164033107599699, "grad_norm": 15.134292602539062, "learning_rate": 1.9413629160063395e-06, "loss": 1.7354, "step": 1085 }, { "epoch": 0.8171557562076749, "grad_norm": 25.39183807373047, "learning_rate": 1.9334389857369257e-06, "loss": 1.7539, "step": 1086 }, { "epoch": 0.81790820165538, "grad_norm": 29.560977935791016, "learning_rate": 1.925515055467512e-06, "loss": 2.2822, "step": 1087 }, { "epoch": 0.8186606471030851, "grad_norm": 33.22944259643555, "learning_rate": 1.9175911251980985e-06, "loss": 2.252, "step": 1088 }, { "epoch": 0.8194130925507901, "grad_norm": 19.814882278442383, "learning_rate": 1.9096671949286847e-06, "loss": 2.2861, "step": 1089 }, { "epoch": 0.8201655379984951, "grad_norm": 31.048839569091797, "learning_rate": 1.9017432646592713e-06, "loss": 2.1475, "step": 1090 }, { "epoch": 0.8209179834462002, "grad_norm": 16.46122932434082, "learning_rate": 1.8938193343898575e-06, "loss": 2.1201, "step": 1091 }, { "epoch": 0.8216704288939052, "grad_norm": 24.316068649291992, "learning_rate": 1.885895404120444e-06, "loss": 1.7188, "step": 1092 }, { "epoch": 0.8224228743416102, "grad_norm": 22.87055778503418, "learning_rate": 1.8779714738510301e-06, "loss": 2.0498, "step": 1093 }, { "epoch": 0.8231753197893152, "grad_norm": 21.308670043945312, "learning_rate": 1.8700475435816165e-06, "loss": 2.3086, "step": 1094 }, { "epoch": 0.8239277652370203, "grad_norm": 24.183080673217773, "learning_rate": 1.8621236133122031e-06, "loss": 2.4404, "step": 1095 }, { "epoch": 0.8246802106847254, "grad_norm": 26.43325424194336, "learning_rate": 1.8541996830427893e-06, "loss": 2.7451, "step": 1096 }, { "epoch": 0.8254326561324304, "grad_norm": 20.95301628112793, "learning_rate": 1.8462757527733757e-06, "loss": 2.1162, "step": 1097 }, { "epoch": 0.8261851015801355, "grad_norm": 22.1946964263916, "learning_rate": 1.8383518225039621e-06, "loss": 2.2295, "step": 1098 }, { "epoch": 0.8269375470278405, "grad_norm": 23.528072357177734, "learning_rate": 1.8304278922345483e-06, "loss": 2.2217, "step": 1099 }, { "epoch": 0.8276899924755455, "grad_norm": 26.633913040161133, "learning_rate": 1.822503961965135e-06, "loss": 2.5986, "step": 1100 }, { "epoch": 0.8284424379232506, "grad_norm": 22.813688278198242, "learning_rate": 1.8145800316957214e-06, "loss": 2.4839, "step": 1101 }, { "epoch": 0.8291948833709556, "grad_norm": 19.97185707092285, "learning_rate": 1.8066561014263076e-06, "loss": 2.1719, "step": 1102 }, { "epoch": 0.8299473288186606, "grad_norm": 17.617671966552734, "learning_rate": 1.798732171156894e-06, "loss": 1.6826, "step": 1103 }, { "epoch": 0.8306997742663657, "grad_norm": 25.46587371826172, "learning_rate": 1.7908082408874801e-06, "loss": 2.4238, "step": 1104 }, { "epoch": 0.8314522197140707, "grad_norm": 27.127437591552734, "learning_rate": 1.7828843106180668e-06, "loss": 2.2109, "step": 1105 }, { "epoch": 0.8322046651617758, "grad_norm": 39.776519775390625, "learning_rate": 1.7749603803486532e-06, "loss": 2.2764, "step": 1106 }, { "epoch": 0.8329571106094809, "grad_norm": 22.824893951416016, "learning_rate": 1.7670364500792394e-06, "loss": 2.8037, "step": 1107 }, { "epoch": 0.8337095560571859, "grad_norm": 28.222532272338867, "learning_rate": 1.7591125198098258e-06, "loss": 2.5269, "step": 1108 }, { "epoch": 0.8344620015048909, "grad_norm": 17.80474281311035, "learning_rate": 1.7511885895404122e-06, "loss": 1.8408, "step": 1109 }, { "epoch": 0.835214446952596, "grad_norm": 36.78727340698242, "learning_rate": 1.7432646592709988e-06, "loss": 2.3691, "step": 1110 }, { "epoch": 0.835966892400301, "grad_norm": 18.49757957458496, "learning_rate": 1.735340729001585e-06, "loss": 2.2607, "step": 1111 }, { "epoch": 0.836719337848006, "grad_norm": 22.9407958984375, "learning_rate": 1.7274167987321714e-06, "loss": 2.75, "step": 1112 }, { "epoch": 0.837471783295711, "grad_norm": 19.55245018005371, "learning_rate": 1.7194928684627576e-06, "loss": 2.248, "step": 1113 }, { "epoch": 0.8382242287434161, "grad_norm": 21.383373260498047, "learning_rate": 1.711568938193344e-06, "loss": 2.6201, "step": 1114 }, { "epoch": 0.8389766741911211, "grad_norm": 19.142234802246094, "learning_rate": 1.7036450079239304e-06, "loss": 1.8564, "step": 1115 }, { "epoch": 0.8397291196388262, "grad_norm": 17.51129913330078, "learning_rate": 1.6957210776545168e-06, "loss": 1.571, "step": 1116 }, { "epoch": 0.8404815650865313, "grad_norm": 21.427541732788086, "learning_rate": 1.6877971473851032e-06, "loss": 2.2041, "step": 1117 }, { "epoch": 0.8412340105342363, "grad_norm": 22.243871688842773, "learning_rate": 1.6798732171156896e-06, "loss": 2.1689, "step": 1118 }, { "epoch": 0.8419864559819413, "grad_norm": 29.453140258789062, "learning_rate": 1.6719492868462758e-06, "loss": 2.4053, "step": 1119 }, { "epoch": 0.8427389014296464, "grad_norm": 27.817018508911133, "learning_rate": 1.6640253565768622e-06, "loss": 2.54, "step": 1120 }, { "epoch": 0.8434913468773514, "grad_norm": 20.197452545166016, "learning_rate": 1.6561014263074488e-06, "loss": 2.2637, "step": 1121 }, { "epoch": 0.8442437923250564, "grad_norm": 22.869123458862305, "learning_rate": 1.648177496038035e-06, "loss": 2.3154, "step": 1122 }, { "epoch": 0.8449962377727614, "grad_norm": 29.69762420654297, "learning_rate": 1.6402535657686214e-06, "loss": 2.209, "step": 1123 }, { "epoch": 0.8457486832204665, "grad_norm": 21.819568634033203, "learning_rate": 1.6323296354992076e-06, "loss": 2.1567, "step": 1124 }, { "epoch": 0.8465011286681715, "grad_norm": 20.180278778076172, "learning_rate": 1.624405705229794e-06, "loss": 2.3389, "step": 1125 }, { "epoch": 0.8472535741158767, "grad_norm": 18.55019187927246, "learning_rate": 1.6164817749603806e-06, "loss": 1.8306, "step": 1126 }, { "epoch": 0.8480060195635817, "grad_norm": 33.824039459228516, "learning_rate": 1.608557844690967e-06, "loss": 2.9336, "step": 1127 }, { "epoch": 0.8487584650112867, "grad_norm": 28.131019592285156, "learning_rate": 1.6006339144215532e-06, "loss": 2.2119, "step": 1128 }, { "epoch": 0.8495109104589917, "grad_norm": 29.770904541015625, "learning_rate": 1.5927099841521396e-06, "loss": 1.7197, "step": 1129 }, { "epoch": 0.8502633559066968, "grad_norm": 21.487455368041992, "learning_rate": 1.5847860538827258e-06, "loss": 2.4395, "step": 1130 }, { "epoch": 0.8510158013544018, "grad_norm": 36.66151809692383, "learning_rate": 1.5768621236133122e-06, "loss": 1.9824, "step": 1131 }, { "epoch": 0.8517682468021068, "grad_norm": 38.3604850769043, "learning_rate": 1.5689381933438988e-06, "loss": 2.7559, "step": 1132 }, { "epoch": 0.8525206922498119, "grad_norm": 30.512598037719727, "learning_rate": 1.561014263074485e-06, "loss": 2.1714, "step": 1133 }, { "epoch": 0.8532731376975169, "grad_norm": 38.16960525512695, "learning_rate": 1.5530903328050714e-06, "loss": 2.3359, "step": 1134 }, { "epoch": 0.8540255831452219, "grad_norm": 20.379392623901367, "learning_rate": 1.5451664025356578e-06, "loss": 2.3291, "step": 1135 }, { "epoch": 0.8547780285929271, "grad_norm": 22.714763641357422, "learning_rate": 1.537242472266244e-06, "loss": 2.3301, "step": 1136 }, { "epoch": 0.8555304740406321, "grad_norm": 18.340160369873047, "learning_rate": 1.5293185419968307e-06, "loss": 1.7939, "step": 1137 }, { "epoch": 0.8562829194883371, "grad_norm": 51.088600158691406, "learning_rate": 1.521394611727417e-06, "loss": 2.3945, "step": 1138 }, { "epoch": 0.8570353649360422, "grad_norm": 30.266088485717773, "learning_rate": 1.5134706814580033e-06, "loss": 2.0225, "step": 1139 }, { "epoch": 0.8577878103837472, "grad_norm": 32.32817459106445, "learning_rate": 1.5055467511885897e-06, "loss": 2.5049, "step": 1140 }, { "epoch": 0.8585402558314522, "grad_norm": 16.877460479736328, "learning_rate": 1.4976228209191759e-06, "loss": 1.998, "step": 1141 }, { "epoch": 0.8592927012791572, "grad_norm": 27.261184692382812, "learning_rate": 1.4896988906497625e-06, "loss": 2.1865, "step": 1142 }, { "epoch": 0.8600451467268623, "grad_norm": 20.137563705444336, "learning_rate": 1.4817749603803489e-06, "loss": 2.1787, "step": 1143 }, { "epoch": 0.8607975921745673, "grad_norm": 18.921850204467773, "learning_rate": 1.473851030110935e-06, "loss": 1.6484, "step": 1144 }, { "epoch": 0.8615500376222723, "grad_norm": 22.27260971069336, "learning_rate": 1.4659270998415215e-06, "loss": 2.6924, "step": 1145 }, { "epoch": 0.8623024830699775, "grad_norm": 23.848426818847656, "learning_rate": 1.4580031695721079e-06, "loss": 2.5029, "step": 1146 }, { "epoch": 0.8630549285176825, "grad_norm": 38.04443359375, "learning_rate": 1.450079239302694e-06, "loss": 1.7031, "step": 1147 }, { "epoch": 0.8638073739653875, "grad_norm": 33.86138153076172, "learning_rate": 1.4421553090332807e-06, "loss": 2.2695, "step": 1148 }, { "epoch": 0.8645598194130926, "grad_norm": 22.020042419433594, "learning_rate": 1.434231378763867e-06, "loss": 2.0498, "step": 1149 }, { "epoch": 0.8653122648607976, "grad_norm": 25.139564514160156, "learning_rate": 1.4263074484944533e-06, "loss": 1.9087, "step": 1150 }, { "epoch": 0.8660647103085026, "grad_norm": 21.85015106201172, "learning_rate": 1.4183835182250397e-06, "loss": 2.2334, "step": 1151 }, { "epoch": 0.8668171557562077, "grad_norm": 29.274450302124023, "learning_rate": 1.4104595879556259e-06, "loss": 2.4326, "step": 1152 }, { "epoch": 0.8675696012039127, "grad_norm": 27.56648063659668, "learning_rate": 1.4025356576862125e-06, "loss": 1.917, "step": 1153 }, { "epoch": 0.8683220466516177, "grad_norm": 23.9324951171875, "learning_rate": 1.394611727416799e-06, "loss": 2.3818, "step": 1154 }, { "epoch": 0.8690744920993227, "grad_norm": 25.811965942382812, "learning_rate": 1.3866877971473853e-06, "loss": 2.2402, "step": 1155 }, { "epoch": 0.8698269375470279, "grad_norm": 20.791032791137695, "learning_rate": 1.3787638668779715e-06, "loss": 1.9756, "step": 1156 }, { "epoch": 0.8705793829947329, "grad_norm": 24.097496032714844, "learning_rate": 1.370839936608558e-06, "loss": 2.2168, "step": 1157 }, { "epoch": 0.871331828442438, "grad_norm": 22.489957809448242, "learning_rate": 1.3629160063391445e-06, "loss": 2.1572, "step": 1158 }, { "epoch": 0.872084273890143, "grad_norm": 37.01300811767578, "learning_rate": 1.3549920760697307e-06, "loss": 2.4814, "step": 1159 }, { "epoch": 0.872836719337848, "grad_norm": 18.329721450805664, "learning_rate": 1.3470681458003171e-06, "loss": 2.2793, "step": 1160 }, { "epoch": 0.873589164785553, "grad_norm": 27.86050033569336, "learning_rate": 1.3391442155309033e-06, "loss": 2.1816, "step": 1161 }, { "epoch": 0.8743416102332581, "grad_norm": 23.993194580078125, "learning_rate": 1.3312202852614897e-06, "loss": 2.7588, "step": 1162 }, { "epoch": 0.8750940556809631, "grad_norm": 22.006431579589844, "learning_rate": 1.3232963549920763e-06, "loss": 1.7539, "step": 1163 }, { "epoch": 0.8758465011286681, "grad_norm": 32.45913314819336, "learning_rate": 1.3153724247226625e-06, "loss": 2.4727, "step": 1164 }, { "epoch": 0.8765989465763732, "grad_norm": 22.738069534301758, "learning_rate": 1.307448494453249e-06, "loss": 2.1963, "step": 1165 }, { "epoch": 0.8773513920240783, "grad_norm": 19.741193771362305, "learning_rate": 1.2995245641838353e-06, "loss": 1.9995, "step": 1166 }, { "epoch": 0.8781038374717833, "grad_norm": 22.3425235748291, "learning_rate": 1.2916006339144215e-06, "loss": 2.584, "step": 1167 }, { "epoch": 0.8788562829194884, "grad_norm": 18.124391555786133, "learning_rate": 1.283676703645008e-06, "loss": 1.9648, "step": 1168 }, { "epoch": 0.8796087283671934, "grad_norm": 20.027179718017578, "learning_rate": 1.2757527733755946e-06, "loss": 2.0439, "step": 1169 }, { "epoch": 0.8803611738148984, "grad_norm": 20.139663696289062, "learning_rate": 1.2678288431061807e-06, "loss": 2.1653, "step": 1170 }, { "epoch": 0.8811136192626035, "grad_norm": 18.437063217163086, "learning_rate": 1.2599049128367671e-06, "loss": 2.1299, "step": 1171 }, { "epoch": 0.8818660647103085, "grad_norm": 19.62114715576172, "learning_rate": 1.2519809825673536e-06, "loss": 1.9883, "step": 1172 }, { "epoch": 0.8826185101580135, "grad_norm": 32.16853713989258, "learning_rate": 1.24405705229794e-06, "loss": 1.9561, "step": 1173 }, { "epoch": 0.8833709556057185, "grad_norm": 31.71169090270996, "learning_rate": 1.2361331220285262e-06, "loss": 2.4297, "step": 1174 }, { "epoch": 0.8841234010534236, "grad_norm": 27.213788986206055, "learning_rate": 1.2282091917591128e-06, "loss": 1.7173, "step": 1175 }, { "epoch": 0.8848758465011287, "grad_norm": 17.10274314880371, "learning_rate": 1.220285261489699e-06, "loss": 2.0039, "step": 1176 }, { "epoch": 0.8856282919488337, "grad_norm": 20.940834045410156, "learning_rate": 1.2123613312202854e-06, "loss": 2.0146, "step": 1177 }, { "epoch": 0.8863807373965388, "grad_norm": 21.25338363647461, "learning_rate": 1.2044374009508718e-06, "loss": 2.4961, "step": 1178 }, { "epoch": 0.8871331828442438, "grad_norm": 18.00691032409668, "learning_rate": 1.1965134706814582e-06, "loss": 1.9626, "step": 1179 }, { "epoch": 0.8878856282919488, "grad_norm": 28.76074981689453, "learning_rate": 1.1885895404120444e-06, "loss": 2.0713, "step": 1180 }, { "epoch": 0.8886380737396539, "grad_norm": 34.1113395690918, "learning_rate": 1.1806656101426308e-06, "loss": 1.8286, "step": 1181 }, { "epoch": 0.8893905191873589, "grad_norm": 19.671710968017578, "learning_rate": 1.1727416798732172e-06, "loss": 2.1851, "step": 1182 }, { "epoch": 0.8901429646350639, "grad_norm": 27.864065170288086, "learning_rate": 1.1648177496038036e-06, "loss": 2.6172, "step": 1183 }, { "epoch": 0.890895410082769, "grad_norm": 23.696453094482422, "learning_rate": 1.15689381933439e-06, "loss": 2.0938, "step": 1184 }, { "epoch": 0.891647855530474, "grad_norm": 25.530902862548828, "learning_rate": 1.1489698890649762e-06, "loss": 2.0488, "step": 1185 }, { "epoch": 0.8924003009781791, "grad_norm": 18.939353942871094, "learning_rate": 1.1410459587955628e-06, "loss": 1.9146, "step": 1186 }, { "epoch": 0.8931527464258842, "grad_norm": 21.1546630859375, "learning_rate": 1.133122028526149e-06, "loss": 1.918, "step": 1187 }, { "epoch": 0.8939051918735892, "grad_norm": 16.869848251342773, "learning_rate": 1.1251980982567354e-06, "loss": 1.4795, "step": 1188 }, { "epoch": 0.8946576373212942, "grad_norm": 23.08144760131836, "learning_rate": 1.1172741679873218e-06, "loss": 2.001, "step": 1189 }, { "epoch": 0.8954100827689992, "grad_norm": 17.821651458740234, "learning_rate": 1.1093502377179082e-06, "loss": 1.7251, "step": 1190 }, { "epoch": 0.8961625282167043, "grad_norm": 27.713359832763672, "learning_rate": 1.1014263074484946e-06, "loss": 2.6104, "step": 1191 }, { "epoch": 0.8969149736644093, "grad_norm": 25.884157180786133, "learning_rate": 1.093502377179081e-06, "loss": 1.751, "step": 1192 }, { "epoch": 0.8976674191121143, "grad_norm": 28.492116928100586, "learning_rate": 1.0855784469096672e-06, "loss": 2.0498, "step": 1193 }, { "epoch": 0.8984198645598194, "grad_norm": 33.259117126464844, "learning_rate": 1.0776545166402536e-06, "loss": 2.5342, "step": 1194 }, { "epoch": 0.8991723100075244, "grad_norm": 31.423341751098633, "learning_rate": 1.06973058637084e-06, "loss": 1.9385, "step": 1195 }, { "epoch": 0.8999247554552295, "grad_norm": 21.758747100830078, "learning_rate": 1.0618066561014264e-06, "loss": 2.3877, "step": 1196 }, { "epoch": 0.9006772009029346, "grad_norm": 32.80032730102539, "learning_rate": 1.0538827258320128e-06, "loss": 1.9854, "step": 1197 }, { "epoch": 0.9014296463506396, "grad_norm": 24.677797317504883, "learning_rate": 1.045958795562599e-06, "loss": 2.1455, "step": 1198 }, { "epoch": 0.9021820917983446, "grad_norm": 28.141752243041992, "learning_rate": 1.0380348652931856e-06, "loss": 2.4795, "step": 1199 }, { "epoch": 0.9029345372460497, "grad_norm": 23.31185531616211, "learning_rate": 1.0301109350237718e-06, "loss": 2.3716, "step": 1200 }, { "epoch": 0.9036869826937547, "grad_norm": 41.367584228515625, "learning_rate": 1.0221870047543582e-06, "loss": 2.3271, "step": 1201 }, { "epoch": 0.9044394281414597, "grad_norm": 21.5883731842041, "learning_rate": 1.0142630744849446e-06, "loss": 2.2163, "step": 1202 }, { "epoch": 0.9051918735891648, "grad_norm": 27.00819969177246, "learning_rate": 1.006339144215531e-06, "loss": 2.7285, "step": 1203 }, { "epoch": 0.9059443190368698, "grad_norm": 18.80669593811035, "learning_rate": 9.984152139461172e-07, "loss": 1.9658, "step": 1204 }, { "epoch": 0.9066967644845748, "grad_norm": 24.32224464416504, "learning_rate": 9.904912836767039e-07, "loss": 2.1133, "step": 1205 }, { "epoch": 0.90744920993228, "grad_norm": 21.93855094909668, "learning_rate": 9.8256735340729e-07, "loss": 2.0918, "step": 1206 }, { "epoch": 0.908201655379985, "grad_norm": 21.887161254882812, "learning_rate": 9.746434231378764e-07, "loss": 1.9634, "step": 1207 }, { "epoch": 0.90895410082769, "grad_norm": 32.432167053222656, "learning_rate": 9.667194928684629e-07, "loss": 2.4023, "step": 1208 }, { "epoch": 0.909706546275395, "grad_norm": 22.609710693359375, "learning_rate": 9.587955625990493e-07, "loss": 2.417, "step": 1209 }, { "epoch": 0.9104589917231001, "grad_norm": 36.706790924072266, "learning_rate": 9.508716323296357e-07, "loss": 1.7422, "step": 1210 }, { "epoch": 0.9112114371708051, "grad_norm": 46.86161804199219, "learning_rate": 9.42947702060222e-07, "loss": 2.5244, "step": 1211 }, { "epoch": 0.9119638826185101, "grad_norm": 31.959516525268555, "learning_rate": 9.350237717908083e-07, "loss": 2.3408, "step": 1212 }, { "epoch": 0.9127163280662152, "grad_norm": 21.65765380859375, "learning_rate": 9.270998415213947e-07, "loss": 1.7529, "step": 1213 }, { "epoch": 0.9134687735139202, "grad_norm": 18.55181884765625, "learning_rate": 9.191759112519811e-07, "loss": 1.8828, "step": 1214 }, { "epoch": 0.9142212189616253, "grad_norm": 21.481040954589844, "learning_rate": 9.112519809825675e-07, "loss": 2.3438, "step": 1215 }, { "epoch": 0.9149736644093304, "grad_norm": 29.45285987854004, "learning_rate": 9.033280507131538e-07, "loss": 2.1631, "step": 1216 }, { "epoch": 0.9157261098570354, "grad_norm": 29.11750030517578, "learning_rate": 8.954041204437401e-07, "loss": 2.2168, "step": 1217 }, { "epoch": 0.9164785553047404, "grad_norm": 24.79958152770996, "learning_rate": 8.874801901743266e-07, "loss": 2.1729, "step": 1218 }, { "epoch": 0.9172310007524455, "grad_norm": 23.481903076171875, "learning_rate": 8.795562599049129e-07, "loss": 1.9126, "step": 1219 }, { "epoch": 0.9179834462001505, "grad_norm": 21.63692855834961, "learning_rate": 8.716323296354994e-07, "loss": 2.04, "step": 1220 }, { "epoch": 0.9187358916478555, "grad_norm": 30.64655876159668, "learning_rate": 8.637083993660857e-07, "loss": 2.207, "step": 1221 }, { "epoch": 0.9194883370955605, "grad_norm": 37.03800582885742, "learning_rate": 8.55784469096672e-07, "loss": 2.0947, "step": 1222 }, { "epoch": 0.9202407825432656, "grad_norm": 29.37847328186035, "learning_rate": 8.478605388272584e-07, "loss": 1.6558, "step": 1223 }, { "epoch": 0.9209932279909706, "grad_norm": 27.9404296875, "learning_rate": 8.399366085578448e-07, "loss": 2.4248, "step": 1224 }, { "epoch": 0.9217456734386757, "grad_norm": 25.742717742919922, "learning_rate": 8.320126782884311e-07, "loss": 2.4385, "step": 1225 }, { "epoch": 0.9224981188863808, "grad_norm": 17.55823516845703, "learning_rate": 8.240887480190175e-07, "loss": 2.1953, "step": 1226 }, { "epoch": 0.9232505643340858, "grad_norm": 21.086668014526367, "learning_rate": 8.161648177496038e-07, "loss": 2.3545, "step": 1227 }, { "epoch": 0.9240030097817908, "grad_norm": 18.935644149780273, "learning_rate": 8.082408874801903e-07, "loss": 1.7041, "step": 1228 }, { "epoch": 0.9247554552294959, "grad_norm": 19.29863739013672, "learning_rate": 8.003169572107766e-07, "loss": 1.8579, "step": 1229 }, { "epoch": 0.9255079006772009, "grad_norm": 28.539777755737305, "learning_rate": 7.923930269413629e-07, "loss": 2.9229, "step": 1230 }, { "epoch": 0.9262603461249059, "grad_norm": 23.44830894470215, "learning_rate": 7.844690966719494e-07, "loss": 2.0674, "step": 1231 }, { "epoch": 0.927012791572611, "grad_norm": 24.71790313720703, "learning_rate": 7.765451664025357e-07, "loss": 2.1211, "step": 1232 }, { "epoch": 0.927765237020316, "grad_norm": 21.996749877929688, "learning_rate": 7.68621236133122e-07, "loss": 2.1787, "step": 1233 }, { "epoch": 0.928517682468021, "grad_norm": 30.374027252197266, "learning_rate": 7.606973058637085e-07, "loss": 2.1143, "step": 1234 }, { "epoch": 0.9292701279157262, "grad_norm": 27.452085494995117, "learning_rate": 7.527733755942948e-07, "loss": 2.0303, "step": 1235 }, { "epoch": 0.9300225733634312, "grad_norm": 22.854434967041016, "learning_rate": 7.448494453248812e-07, "loss": 2.0107, "step": 1236 }, { "epoch": 0.9307750188111362, "grad_norm": 41.897850036621094, "learning_rate": 7.369255150554675e-07, "loss": 1.9414, "step": 1237 }, { "epoch": 0.9315274642588413, "grad_norm": 17.855621337890625, "learning_rate": 7.290015847860539e-07, "loss": 2.1172, "step": 1238 }, { "epoch": 0.9322799097065463, "grad_norm": 23.190418243408203, "learning_rate": 7.210776545166403e-07, "loss": 2.3647, "step": 1239 }, { "epoch": 0.9330323551542513, "grad_norm": 22.126853942871094, "learning_rate": 7.131537242472266e-07, "loss": 2.7139, "step": 1240 }, { "epoch": 0.9337848006019563, "grad_norm": 23.941364288330078, "learning_rate": 7.052297939778129e-07, "loss": 2.2729, "step": 1241 }, { "epoch": 0.9345372460496614, "grad_norm": 30.85197639465332, "learning_rate": 6.973058637083995e-07, "loss": 2.5273, "step": 1242 }, { "epoch": 0.9352896914973664, "grad_norm": 42.83109664916992, "learning_rate": 6.893819334389858e-07, "loss": 2.3301, "step": 1243 }, { "epoch": 0.9360421369450714, "grad_norm": 27.508867263793945, "learning_rate": 6.814580031695723e-07, "loss": 2.6396, "step": 1244 }, { "epoch": 0.9367945823927766, "grad_norm": 26.918298721313477, "learning_rate": 6.735340729001586e-07, "loss": 2.4189, "step": 1245 }, { "epoch": 0.9375470278404816, "grad_norm": 19.541494369506836, "learning_rate": 6.656101426307449e-07, "loss": 2.501, "step": 1246 }, { "epoch": 0.9382994732881866, "grad_norm": 21.202077865600586, "learning_rate": 6.576862123613313e-07, "loss": 2.0938, "step": 1247 }, { "epoch": 0.9390519187358917, "grad_norm": 20.202423095703125, "learning_rate": 6.497622820919177e-07, "loss": 2.1953, "step": 1248 }, { "epoch": 0.9398043641835967, "grad_norm": 24.55869483947754, "learning_rate": 6.41838351822504e-07, "loss": 2.1191, "step": 1249 }, { "epoch": 0.9405568096313017, "grad_norm": 35.99618911743164, "learning_rate": 6.339144215530904e-07, "loss": 2.4551, "step": 1250 }, { "epoch": 0.9413092550790068, "grad_norm": 27.567293167114258, "learning_rate": 6.259904912836768e-07, "loss": 1.7285, "step": 1251 }, { "epoch": 0.9420617005267118, "grad_norm": 24.59737205505371, "learning_rate": 6.180665610142631e-07, "loss": 2.9912, "step": 1252 }, { "epoch": 0.9428141459744168, "grad_norm": 20.435455322265625, "learning_rate": 6.101426307448495e-07, "loss": 2.1045, "step": 1253 }, { "epoch": 0.9435665914221218, "grad_norm": 25.387649536132812, "learning_rate": 6.022187004754359e-07, "loss": 1.8057, "step": 1254 }, { "epoch": 0.944319036869827, "grad_norm": 19.371496200561523, "learning_rate": 5.942947702060222e-07, "loss": 2.0029, "step": 1255 }, { "epoch": 0.945071482317532, "grad_norm": 19.404163360595703, "learning_rate": 5.863708399366086e-07, "loss": 1.791, "step": 1256 }, { "epoch": 0.945823927765237, "grad_norm": 36.32655715942383, "learning_rate": 5.78446909667195e-07, "loss": 1.9404, "step": 1257 }, { "epoch": 0.9465763732129421, "grad_norm": 26.90781021118164, "learning_rate": 5.705229793977814e-07, "loss": 2.1074, "step": 1258 }, { "epoch": 0.9473288186606471, "grad_norm": 25.104326248168945, "learning_rate": 5.625990491283677e-07, "loss": 2.0557, "step": 1259 }, { "epoch": 0.9480812641083521, "grad_norm": 20.98943519592285, "learning_rate": 5.546751188589541e-07, "loss": 1.9453, "step": 1260 }, { "epoch": 0.9488337095560572, "grad_norm": 18.50478172302246, "learning_rate": 5.467511885895405e-07, "loss": 2.0991, "step": 1261 }, { "epoch": 0.9495861550037622, "grad_norm": 32.75830078125, "learning_rate": 5.388272583201268e-07, "loss": 2.3687, "step": 1262 }, { "epoch": 0.9503386004514672, "grad_norm": 21.3521785736084, "learning_rate": 5.309033280507132e-07, "loss": 2.4414, "step": 1263 }, { "epoch": 0.9510910458991723, "grad_norm": 20.554296493530273, "learning_rate": 5.229793977812995e-07, "loss": 1.7754, "step": 1264 }, { "epoch": 0.9518434913468774, "grad_norm": 22.69301414489746, "learning_rate": 5.150554675118859e-07, "loss": 2.3613, "step": 1265 }, { "epoch": 0.9525959367945824, "grad_norm": 26.686033248901367, "learning_rate": 5.071315372424723e-07, "loss": 2.6074, "step": 1266 }, { "epoch": 0.9533483822422875, "grad_norm": 19.40122413635254, "learning_rate": 4.992076069730586e-07, "loss": 2.0215, "step": 1267 }, { "epoch": 0.9541008276899925, "grad_norm": 30.532285690307617, "learning_rate": 4.91283676703645e-07, "loss": 1.71, "step": 1268 }, { "epoch": 0.9548532731376975, "grad_norm": 21.527864456176758, "learning_rate": 4.833597464342314e-07, "loss": 2.6123, "step": 1269 }, { "epoch": 0.9556057185854026, "grad_norm": 23.632919311523438, "learning_rate": 4.7543581616481783e-07, "loss": 2.293, "step": 1270 }, { "epoch": 0.9563581640331076, "grad_norm": 20.96917724609375, "learning_rate": 4.6751188589540413e-07, "loss": 2.4497, "step": 1271 }, { "epoch": 0.9571106094808126, "grad_norm": 36.56161880493164, "learning_rate": 4.5958795562599054e-07, "loss": 2.1211, "step": 1272 }, { "epoch": 0.9578630549285176, "grad_norm": 30.125322341918945, "learning_rate": 4.516640253565769e-07, "loss": 2.375, "step": 1273 }, { "epoch": 0.9586155003762227, "grad_norm": 28.567535400390625, "learning_rate": 4.437400950871633e-07, "loss": 2.1699, "step": 1274 }, { "epoch": 0.9593679458239278, "grad_norm": 26.635305404663086, "learning_rate": 4.358161648177497e-07, "loss": 2.0947, "step": 1275 }, { "epoch": 0.9601203912716328, "grad_norm": 25.772676467895508, "learning_rate": 4.27892234548336e-07, "loss": 2.0137, "step": 1276 }, { "epoch": 0.9608728367193379, "grad_norm": 22.599668502807617, "learning_rate": 4.199683042789224e-07, "loss": 2.7881, "step": 1277 }, { "epoch": 0.9616252821670429, "grad_norm": 24.592472076416016, "learning_rate": 4.1204437400950875e-07, "loss": 2.1631, "step": 1278 }, { "epoch": 0.9623777276147479, "grad_norm": 23.07600212097168, "learning_rate": 4.0412044374009516e-07, "loss": 2.127, "step": 1279 }, { "epoch": 0.963130173062453, "grad_norm": 24.716087341308594, "learning_rate": 3.9619651347068146e-07, "loss": 1.8828, "step": 1280 }, { "epoch": 0.963882618510158, "grad_norm": 21.794273376464844, "learning_rate": 3.8827258320126786e-07, "loss": 2.3472, "step": 1281 }, { "epoch": 0.964635063957863, "grad_norm": 28.48988914489746, "learning_rate": 3.8034865293185427e-07, "loss": 2.3457, "step": 1282 }, { "epoch": 0.9653875094055681, "grad_norm": 21.161699295043945, "learning_rate": 3.724247226624406e-07, "loss": 2.4102, "step": 1283 }, { "epoch": 0.9661399548532731, "grad_norm": 26.318927764892578, "learning_rate": 3.6450079239302697e-07, "loss": 2.4238, "step": 1284 }, { "epoch": 0.9668924003009782, "grad_norm": 33.44709014892578, "learning_rate": 3.565768621236133e-07, "loss": 2.5547, "step": 1285 }, { "epoch": 0.9676448457486833, "grad_norm": 27.735422134399414, "learning_rate": 3.486529318541997e-07, "loss": 2.8076, "step": 1286 }, { "epoch": 0.9683972911963883, "grad_norm": 20.180845260620117, "learning_rate": 3.4072900158478613e-07, "loss": 2.3311, "step": 1287 }, { "epoch": 0.9691497366440933, "grad_norm": 26.456632614135742, "learning_rate": 3.3280507131537243e-07, "loss": 2.4238, "step": 1288 }, { "epoch": 0.9699021820917983, "grad_norm": 24.67302131652832, "learning_rate": 3.2488114104595883e-07, "loss": 2.251, "step": 1289 }, { "epoch": 0.9706546275395034, "grad_norm": 25.135757446289062, "learning_rate": 3.169572107765452e-07, "loss": 2.0205, "step": 1290 }, { "epoch": 0.9714070729872084, "grad_norm": 46.5058708190918, "learning_rate": 3.0903328050713154e-07, "loss": 2.8018, "step": 1291 }, { "epoch": 0.9721595184349134, "grad_norm": 29.164596557617188, "learning_rate": 3.0110935023771794e-07, "loss": 2.541, "step": 1292 }, { "epoch": 0.9729119638826185, "grad_norm": 19.991748809814453, "learning_rate": 2.931854199683043e-07, "loss": 2.126, "step": 1293 }, { "epoch": 0.9736644093303235, "grad_norm": 19.993240356445312, "learning_rate": 2.852614896988907e-07, "loss": 2.1357, "step": 1294 }, { "epoch": 0.9744168547780286, "grad_norm": 30.28998565673828, "learning_rate": 2.7733755942947705e-07, "loss": 2.7969, "step": 1295 }, { "epoch": 0.9751693002257337, "grad_norm": 32.665958404541016, "learning_rate": 2.694136291600634e-07, "loss": 2.0776, "step": 1296 }, { "epoch": 0.9759217456734387, "grad_norm": 32.19696044921875, "learning_rate": 2.6148969889064975e-07, "loss": 2.1387, "step": 1297 }, { "epoch": 0.9766741911211437, "grad_norm": 24.746389389038086, "learning_rate": 2.5356576862123616e-07, "loss": 1.9941, "step": 1298 }, { "epoch": 0.9774266365688488, "grad_norm": 30.807029724121094, "learning_rate": 2.456418383518225e-07, "loss": 2.4258, "step": 1299 }, { "epoch": 0.9781790820165538, "grad_norm": 19.208616256713867, "learning_rate": 2.3771790808240892e-07, "loss": 2.1309, "step": 1300 }, { "epoch": 0.9789315274642588, "grad_norm": 24.626890182495117, "learning_rate": 2.2979397781299527e-07, "loss": 2.1152, "step": 1301 }, { "epoch": 0.9796839729119639, "grad_norm": 18.25760269165039, "learning_rate": 2.2187004754358165e-07, "loss": 2.1079, "step": 1302 }, { "epoch": 0.9804364183596689, "grad_norm": 19.399744033813477, "learning_rate": 2.13946117274168e-07, "loss": 1.9756, "step": 1303 }, { "epoch": 0.9811888638073739, "grad_norm": 32.49583053588867, "learning_rate": 2.0602218700475438e-07, "loss": 1.9805, "step": 1304 }, { "epoch": 0.981941309255079, "grad_norm": 21.51534652709961, "learning_rate": 1.9809825673534073e-07, "loss": 1.9082, "step": 1305 }, { "epoch": 0.9826937547027841, "grad_norm": 29.631881713867188, "learning_rate": 1.9017432646592713e-07, "loss": 2.188, "step": 1306 }, { "epoch": 0.9834462001504891, "grad_norm": 23.09160804748535, "learning_rate": 1.8225039619651348e-07, "loss": 2.3164, "step": 1307 }, { "epoch": 0.9841986455981941, "grad_norm": 22.58462905883789, "learning_rate": 1.7432646592709986e-07, "loss": 2.4209, "step": 1308 }, { "epoch": 0.9849510910458992, "grad_norm": 35.65068435668945, "learning_rate": 1.6640253565768621e-07, "loss": 2.5005, "step": 1309 }, { "epoch": 0.9857035364936042, "grad_norm": 19.236482620239258, "learning_rate": 1.584786053882726e-07, "loss": 1.9756, "step": 1310 }, { "epoch": 0.9864559819413092, "grad_norm": 22.260229110717773, "learning_rate": 1.5055467511885897e-07, "loss": 2.3398, "step": 1311 }, { "epoch": 0.9872084273890143, "grad_norm": 23.633285522460938, "learning_rate": 1.4263074484944535e-07, "loss": 2.1797, "step": 1312 }, { "epoch": 0.9879608728367193, "grad_norm": 25.102895736694336, "learning_rate": 1.347068145800317e-07, "loss": 2.6221, "step": 1313 }, { "epoch": 0.9887133182844243, "grad_norm": 22.20816993713379, "learning_rate": 1.2678288431061808e-07, "loss": 2.418, "step": 1314 }, { "epoch": 0.9894657637321295, "grad_norm": 19.845365524291992, "learning_rate": 1.1885895404120446e-07, "loss": 1.8213, "step": 1315 }, { "epoch": 0.9902182091798345, "grad_norm": 31.5147647857666, "learning_rate": 1.1093502377179082e-07, "loss": 2.2051, "step": 1316 }, { "epoch": 0.9909706546275395, "grad_norm": 21.427370071411133, "learning_rate": 1.0301109350237719e-07, "loss": 2.3633, "step": 1317 }, { "epoch": 0.9917231000752446, "grad_norm": 21.501970291137695, "learning_rate": 9.508716323296357e-08, "loss": 2.2607, "step": 1318 }, { "epoch": 0.9924755455229496, "grad_norm": 24.47955322265625, "learning_rate": 8.716323296354993e-08, "loss": 1.9883, "step": 1319 }, { "epoch": 0.9932279909706546, "grad_norm": 33.163780212402344, "learning_rate": 7.92393026941363e-08, "loss": 1.9531, "step": 1320 }, { "epoch": 0.9939804364183596, "grad_norm": 25.771236419677734, "learning_rate": 7.131537242472267e-08, "loss": 2.1113, "step": 1321 }, { "epoch": 0.9947328818660647, "grad_norm": 30.363033294677734, "learning_rate": 6.339144215530904e-08, "loss": 2.3027, "step": 1322 }, { "epoch": 0.9954853273137697, "grad_norm": 22.448755264282227, "learning_rate": 5.546751188589541e-08, "loss": 2.293, "step": 1323 }, { "epoch": 0.9962377727614747, "grad_norm": 23.290424346923828, "learning_rate": 4.754358161648178e-08, "loss": 2.6187, "step": 1324 }, { "epoch": 0.9969902182091799, "grad_norm": 24.690128326416016, "learning_rate": 3.961965134706815e-08, "loss": 2.208, "step": 1325 }, { "epoch": 0.9977426636568849, "grad_norm": 22.215335845947266, "learning_rate": 3.169572107765452e-08, "loss": 2.2271, "step": 1326 }, { "epoch": 0.9984951091045899, "grad_norm": 37.365562438964844, "learning_rate": 2.377179080824089e-08, "loss": 2.3867, "step": 1327 }, { "epoch": 0.999247554552295, "grad_norm": 21.778207778930664, "learning_rate": 1.584786053882726e-08, "loss": 2.248, "step": 1328 }, { "epoch": 1.0, "grad_norm": 24.791046142578125, "learning_rate": 7.92393026941363e-09, "loss": 2.4365, "step": 1329 }, { "epoch": 1.0, "step": 1329, "total_flos": 2.392704719865774e+18, "train_loss": 2.812455360292043, "train_runtime": 1486.3359, "train_samples_per_second": 228.75, "train_steps_per_second": 0.894 } ], "logging_steps": 1, "max_steps": 1329, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": false, "should_training_stop": false }, "attributes": {} } }, "total_flos": 2.392704719865774e+18, "train_batch_size": 32, "trial_name": null, "trial_params": null }