| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 1.5332869807057188, |
| "eval_steps": 150, |
| "global_step": 4410, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0006952894142186686, |
| "grad_norm": 5.413117383066536, |
| "learning_rate": 6.944444444444445e-07, |
| "loss": 2.3653, |
| "step": 2 |
| }, |
| { |
| "epoch": 0.0013905788284373371, |
| "grad_norm": 5.092870612337868, |
| "learning_rate": 1.388888888888889e-06, |
| "loss": 2.5001, |
| "step": 4 |
| }, |
| { |
| "epoch": 0.0020858682426560054, |
| "grad_norm": 3.6285625961711943, |
| "learning_rate": 2.0833333333333334e-06, |
| "loss": 2.3759, |
| "step": 6 |
| }, |
| { |
| "epoch": 0.0027811576568746743, |
| "grad_norm": 2.5025911526151075, |
| "learning_rate": 2.777777777777778e-06, |
| "loss": 1.944, |
| "step": 8 |
| }, |
| { |
| "epoch": 0.0034764470710933427, |
| "grad_norm": 3.077266049542496, |
| "learning_rate": 3.4722222222222224e-06, |
| "loss": 2.2504, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.004171736485312011, |
| "grad_norm": 4.068243939187174, |
| "learning_rate": 4.166666666666667e-06, |
| "loss": 2.0637, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.00486702589953068, |
| "grad_norm": 3.511444738830971, |
| "learning_rate": 4.861111111111111e-06, |
| "loss": 2.3604, |
| "step": 14 |
| }, |
| { |
| "epoch": 0.0055623153137493485, |
| "grad_norm": 5.925744892256934, |
| "learning_rate": 5.555555555555556e-06, |
| "loss": 2.6528, |
| "step": 16 |
| }, |
| { |
| "epoch": 0.0062576047279680165, |
| "grad_norm": 2.7950114871483405, |
| "learning_rate": 6.25e-06, |
| "loss": 2.3229, |
| "step": 18 |
| }, |
| { |
| "epoch": 0.006952894142186685, |
| "grad_norm": 6.586211843766182, |
| "learning_rate": 6.944444444444445e-06, |
| "loss": 2.342, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.0076481835564053535, |
| "grad_norm": 4.902839955269193, |
| "learning_rate": 7.63888888888889e-06, |
| "loss": 2.4188, |
| "step": 22 |
| }, |
| { |
| "epoch": 0.008343472970624021, |
| "grad_norm": 4.257062809771645, |
| "learning_rate": 8.333333333333334e-06, |
| "loss": 1.7957, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.009038762384842691, |
| "grad_norm": 4.460352004615699, |
| "learning_rate": 9.027777777777777e-06, |
| "loss": 2.0726, |
| "step": 26 |
| }, |
| { |
| "epoch": 0.00973405179906136, |
| "grad_norm": 5.858061506133739, |
| "learning_rate": 9.722222222222223e-06, |
| "loss": 2.0455, |
| "step": 28 |
| }, |
| { |
| "epoch": 0.010429341213280027, |
| "grad_norm": 4.331946668100709, |
| "learning_rate": 1.0416666666666668e-05, |
| "loss": 1.7645, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.011124630627498697, |
| "grad_norm": 5.428742204187391, |
| "learning_rate": 1.1111111111111112e-05, |
| "loss": 1.7446, |
| "step": 32 |
| }, |
| { |
| "epoch": 0.011819920041717365, |
| "grad_norm": 2.485909286541028, |
| "learning_rate": 1.1805555555555555e-05, |
| "loss": 1.5885, |
| "step": 34 |
| }, |
| { |
| "epoch": 0.012515209455936033, |
| "grad_norm": 3.2602949308730222, |
| "learning_rate": 1.25e-05, |
| "loss": 1.1014, |
| "step": 36 |
| }, |
| { |
| "epoch": 0.013210498870154701, |
| "grad_norm": 4.962187747415964, |
| "learning_rate": 1.3194444444444446e-05, |
| "loss": 1.3588, |
| "step": 38 |
| }, |
| { |
| "epoch": 0.01390578828437337, |
| "grad_norm": 7.627756894198461, |
| "learning_rate": 1.388888888888889e-05, |
| "loss": 1.4014, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.014601077698592039, |
| "grad_norm": 2.821848388410092, |
| "learning_rate": 1.4583333333333335e-05, |
| "loss": 1.0959, |
| "step": 42 |
| }, |
| { |
| "epoch": 0.015296367112810707, |
| "grad_norm": 10.989320944549025, |
| "learning_rate": 1.527777777777778e-05, |
| "loss": 1.739, |
| "step": 44 |
| }, |
| { |
| "epoch": 0.015991656527029375, |
| "grad_norm": 4.300936602591115, |
| "learning_rate": 1.597222222222222e-05, |
| "loss": 1.1078, |
| "step": 46 |
| }, |
| { |
| "epoch": 0.016686945941248043, |
| "grad_norm": 3.7539663261019856, |
| "learning_rate": 1.6666666666666667e-05, |
| "loss": 1.2701, |
| "step": 48 |
| }, |
| { |
| "epoch": 0.017382235355466714, |
| "grad_norm": 2.589248169352173, |
| "learning_rate": 1.736111111111111e-05, |
| "loss": 1.4452, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.018077524769685382, |
| "grad_norm": 3.6679301322156177, |
| "learning_rate": 1.8055555555555555e-05, |
| "loss": 1.4243, |
| "step": 52 |
| }, |
| { |
| "epoch": 0.01877281418390405, |
| "grad_norm": 2.0885660923860074, |
| "learning_rate": 1.8750000000000002e-05, |
| "loss": 1.4453, |
| "step": 54 |
| }, |
| { |
| "epoch": 0.01946810359812272, |
| "grad_norm": 2.955353237610474, |
| "learning_rate": 1.9444444444444445e-05, |
| "loss": 1.5388, |
| "step": 56 |
| }, |
| { |
| "epoch": 0.020163393012341387, |
| "grad_norm": 3.0527006398487018, |
| "learning_rate": 2.013888888888889e-05, |
| "loss": 1.4043, |
| "step": 58 |
| }, |
| { |
| "epoch": 0.020858682426560055, |
| "grad_norm": 1.9862208864154767, |
| "learning_rate": 2.0833333333333336e-05, |
| "loss": 1.1007, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.021553971840778723, |
| "grad_norm": 2.347361178472164, |
| "learning_rate": 2.152777777777778e-05, |
| "loss": 0.9291, |
| "step": 62 |
| }, |
| { |
| "epoch": 0.022249261254997394, |
| "grad_norm": 2.4812993223105995, |
| "learning_rate": 2.2222222222222223e-05, |
| "loss": 0.9592, |
| "step": 64 |
| }, |
| { |
| "epoch": 0.022944550669216062, |
| "grad_norm": 1.639333831845777, |
| "learning_rate": 2.2916666666666667e-05, |
| "loss": 1.0645, |
| "step": 66 |
| }, |
| { |
| "epoch": 0.02363984008343473, |
| "grad_norm": 2.92858351082494, |
| "learning_rate": 2.361111111111111e-05, |
| "loss": 1.4269, |
| "step": 68 |
| }, |
| { |
| "epoch": 0.024335129497653398, |
| "grad_norm": 2.8503434812871604, |
| "learning_rate": 2.4305555555555558e-05, |
| "loss": 0.7829, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.025030418911872066, |
| "grad_norm": 2.4875590459354107, |
| "learning_rate": 2.5e-05, |
| "loss": 0.9733, |
| "step": 72 |
| }, |
| { |
| "epoch": 0.025725708326090734, |
| "grad_norm": 5.2567545525905075, |
| "learning_rate": 2.5694444444444445e-05, |
| "loss": 1.3559, |
| "step": 74 |
| }, |
| { |
| "epoch": 0.026420997740309402, |
| "grad_norm": 4.68745753567611, |
| "learning_rate": 2.6388888888888892e-05, |
| "loss": 1.1499, |
| "step": 76 |
| }, |
| { |
| "epoch": 0.027116287154528074, |
| "grad_norm": 2.810345453706711, |
| "learning_rate": 2.7083333333333332e-05, |
| "loss": 0.8636, |
| "step": 78 |
| }, |
| { |
| "epoch": 0.02781157656874674, |
| "grad_norm": 3.0144493626195388, |
| "learning_rate": 2.777777777777778e-05, |
| "loss": 0.9274, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.02850686598296541, |
| "grad_norm": 2.3263036535508523, |
| "learning_rate": 2.8472222222222223e-05, |
| "loss": 1.3979, |
| "step": 82 |
| }, |
| { |
| "epoch": 0.029202155397184078, |
| "grad_norm": 1.6009019341419857, |
| "learning_rate": 2.916666666666667e-05, |
| "loss": 0.9039, |
| "step": 84 |
| }, |
| { |
| "epoch": 0.029897444811402746, |
| "grad_norm": 2.0637506891442294, |
| "learning_rate": 2.9861111111111113e-05, |
| "loss": 1.5187, |
| "step": 86 |
| }, |
| { |
| "epoch": 0.030592734225621414, |
| "grad_norm": 1.5533001268209932, |
| "learning_rate": 3.055555555555556e-05, |
| "loss": 0.8937, |
| "step": 88 |
| }, |
| { |
| "epoch": 0.03128802363984008, |
| "grad_norm": 1.9974857545819733, |
| "learning_rate": 3.125e-05, |
| "loss": 0.8218, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.03198331305405875, |
| "grad_norm": 1.871899331408953, |
| "learning_rate": 3.194444444444444e-05, |
| "loss": 1.2451, |
| "step": 92 |
| }, |
| { |
| "epoch": 0.03267860246827742, |
| "grad_norm": 1.591601830348497, |
| "learning_rate": 3.263888888888889e-05, |
| "loss": 0.7485, |
| "step": 94 |
| }, |
| { |
| "epoch": 0.033373891882496086, |
| "grad_norm": 2.404166647711005, |
| "learning_rate": 3.3333333333333335e-05, |
| "loss": 1.1587, |
| "step": 96 |
| }, |
| { |
| "epoch": 0.03406918129671476, |
| "grad_norm": 1.581080788392888, |
| "learning_rate": 3.402777777777778e-05, |
| "loss": 0.9578, |
| "step": 98 |
| }, |
| { |
| "epoch": 0.03476447071093343, |
| "grad_norm": 2.1551207338771547, |
| "learning_rate": 3.472222222222222e-05, |
| "loss": 1.1305, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.0354597601251521, |
| "grad_norm": 4.2246272812704, |
| "learning_rate": 3.541666666666667e-05, |
| "loss": 1.04, |
| "step": 102 |
| }, |
| { |
| "epoch": 0.036155049539370765, |
| "grad_norm": 2.5526898161765588, |
| "learning_rate": 3.611111111111111e-05, |
| "loss": 1.0743, |
| "step": 104 |
| }, |
| { |
| "epoch": 0.03685033895358943, |
| "grad_norm": 1.9784815137623597, |
| "learning_rate": 3.6805555555555556e-05, |
| "loss": 0.9667, |
| "step": 106 |
| }, |
| { |
| "epoch": 0.0375456283678081, |
| "grad_norm": 1.9131648200880944, |
| "learning_rate": 3.7500000000000003e-05, |
| "loss": 0.8951, |
| "step": 108 |
| }, |
| { |
| "epoch": 0.03824091778202677, |
| "grad_norm": 3.9405296351174575, |
| "learning_rate": 3.8194444444444444e-05, |
| "loss": 1.222, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.03893620719624544, |
| "grad_norm": 13.764861545850291, |
| "learning_rate": 3.888888888888889e-05, |
| "loss": 1.0497, |
| "step": 112 |
| }, |
| { |
| "epoch": 0.039631496610464105, |
| "grad_norm": 3.998123122175411, |
| "learning_rate": 3.958333333333333e-05, |
| "loss": 1.0901, |
| "step": 114 |
| }, |
| { |
| "epoch": 0.04032678602468277, |
| "grad_norm": 3.126193921423756, |
| "learning_rate": 4.027777777777778e-05, |
| "loss": 1.1143, |
| "step": 116 |
| }, |
| { |
| "epoch": 0.04102207543890144, |
| "grad_norm": 2.7839926692610613, |
| "learning_rate": 4.0972222222222225e-05, |
| "loss": 1.1637, |
| "step": 118 |
| }, |
| { |
| "epoch": 0.04171736485312011, |
| "grad_norm": 2.591162621161276, |
| "learning_rate": 4.166666666666667e-05, |
| "loss": 1.0624, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.04241265426733878, |
| "grad_norm": 1.4930703711545332, |
| "learning_rate": 4.236111111111111e-05, |
| "loss": 1.0239, |
| "step": 122 |
| }, |
| { |
| "epoch": 0.043107943681557445, |
| "grad_norm": 4.355150070532966, |
| "learning_rate": 4.305555555555556e-05, |
| "loss": 1.266, |
| "step": 124 |
| }, |
| { |
| "epoch": 0.04380323309577612, |
| "grad_norm": 1.8285608876277135, |
| "learning_rate": 4.375e-05, |
| "loss": 1.0667, |
| "step": 126 |
| }, |
| { |
| "epoch": 0.04449852250999479, |
| "grad_norm": 2.1848832742617055, |
| "learning_rate": 4.4444444444444447e-05, |
| "loss": 0.7549, |
| "step": 128 |
| }, |
| { |
| "epoch": 0.045193811924213456, |
| "grad_norm": 3.547857947451226, |
| "learning_rate": 4.5138888888888894e-05, |
| "loss": 0.7974, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.045889101338432124, |
| "grad_norm": 3.372457028128184, |
| "learning_rate": 4.5833333333333334e-05, |
| "loss": 1.0482, |
| "step": 132 |
| }, |
| { |
| "epoch": 0.04658439075265079, |
| "grad_norm": 2.2871138548091703, |
| "learning_rate": 4.652777777777778e-05, |
| "loss": 0.8246, |
| "step": 134 |
| }, |
| { |
| "epoch": 0.04727968016686946, |
| "grad_norm": 1.7489408054743605, |
| "learning_rate": 4.722222222222222e-05, |
| "loss": 1.0011, |
| "step": 136 |
| }, |
| { |
| "epoch": 0.04797496958108813, |
| "grad_norm": 1.944650349232646, |
| "learning_rate": 4.791666666666667e-05, |
| "loss": 0.7828, |
| "step": 138 |
| }, |
| { |
| "epoch": 0.048670258995306796, |
| "grad_norm": 2.321619459307342, |
| "learning_rate": 4.8611111111111115e-05, |
| "loss": 1.02, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.049365548409525464, |
| "grad_norm": 5.317732559595606, |
| "learning_rate": 4.930555555555556e-05, |
| "loss": 0.9539, |
| "step": 142 |
| }, |
| { |
| "epoch": 0.05006083782374413, |
| "grad_norm": 1.5887292172279854, |
| "learning_rate": 5e-05, |
| "loss": 0.7961, |
| "step": 144 |
| }, |
| { |
| "epoch": 0.0507561272379628, |
| "grad_norm": 2.9987019111854964, |
| "learning_rate": 5.069444444444444e-05, |
| "loss": 1.0018, |
| "step": 146 |
| }, |
| { |
| "epoch": 0.05145141665218147, |
| "grad_norm": 5.63878906132749, |
| "learning_rate": 5.138888888888889e-05, |
| "loss": 0.9278, |
| "step": 148 |
| }, |
| { |
| "epoch": 0.052146706066400136, |
| "grad_norm": 2.4849685204332834, |
| "learning_rate": 5.208333333333334e-05, |
| "loss": 0.7531, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.052146706066400136, |
| "eval_loss": 0.9286500215530396, |
| "eval_runtime": 711.2619, |
| "eval_samples_per_second": 6.808, |
| "eval_steps_per_second": 0.214, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.052841995480618804, |
| "grad_norm": 3.4927981580288776, |
| "learning_rate": 5.2777777777777784e-05, |
| "loss": 0.7422, |
| "step": 152 |
| }, |
| { |
| "epoch": 0.05353728489483748, |
| "grad_norm": 2.6593724943984682, |
| "learning_rate": 5.3472222222222224e-05, |
| "loss": 0.7762, |
| "step": 154 |
| }, |
| { |
| "epoch": 0.05423257430905615, |
| "grad_norm": 2.99709354088536, |
| "learning_rate": 5.4166666666666664e-05, |
| "loss": 0.8658, |
| "step": 156 |
| }, |
| { |
| "epoch": 0.054927863723274815, |
| "grad_norm": 3.9196361623391414, |
| "learning_rate": 5.486111111111112e-05, |
| "loss": 1.1783, |
| "step": 158 |
| }, |
| { |
| "epoch": 0.05562315313749348, |
| "grad_norm": 3.022562685608673, |
| "learning_rate": 5.555555555555556e-05, |
| "loss": 1.0602, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.05631844255171215, |
| "grad_norm": 3.292942684053579, |
| "learning_rate": 5.6250000000000005e-05, |
| "loss": 0.8561, |
| "step": 162 |
| }, |
| { |
| "epoch": 0.05701373196593082, |
| "grad_norm": 2.397775023338686, |
| "learning_rate": 5.6944444444444445e-05, |
| "loss": 0.9561, |
| "step": 164 |
| }, |
| { |
| "epoch": 0.05770902138014949, |
| "grad_norm": 3.0546714312119643, |
| "learning_rate": 5.7638888888888886e-05, |
| "loss": 0.8274, |
| "step": 166 |
| }, |
| { |
| "epoch": 0.058404310794368156, |
| "grad_norm": 2.4021883380894393, |
| "learning_rate": 5.833333333333334e-05, |
| "loss": 0.8492, |
| "step": 168 |
| }, |
| { |
| "epoch": 0.059099600208586824, |
| "grad_norm": 2.6052829770116293, |
| "learning_rate": 5.902777777777778e-05, |
| "loss": 1.1005, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.05979488962280549, |
| "grad_norm": 1.6838192892320467, |
| "learning_rate": 5.972222222222223e-05, |
| "loss": 1.0157, |
| "step": 172 |
| }, |
| { |
| "epoch": 0.06049017903702416, |
| "grad_norm": 3.98880214871885, |
| "learning_rate": 6.041666666666667e-05, |
| "loss": 0.8136, |
| "step": 174 |
| }, |
| { |
| "epoch": 0.06118546845124283, |
| "grad_norm": 4.81494345341073, |
| "learning_rate": 6.111111111111112e-05, |
| "loss": 0.9475, |
| "step": 176 |
| }, |
| { |
| "epoch": 0.061880757865461496, |
| "grad_norm": 1.9926278890091862, |
| "learning_rate": 6.180555555555556e-05, |
| "loss": 0.6549, |
| "step": 178 |
| }, |
| { |
| "epoch": 0.06257604727968016, |
| "grad_norm": 1.750353030728397, |
| "learning_rate": 6.25e-05, |
| "loss": 1.1442, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.06327133669389884, |
| "grad_norm": 4.77249782692129, |
| "learning_rate": 6.319444444444444e-05, |
| "loss": 1.1927, |
| "step": 182 |
| }, |
| { |
| "epoch": 0.0639666261081175, |
| "grad_norm": 1.866901788617278, |
| "learning_rate": 6.388888888888888e-05, |
| "loss": 1.0032, |
| "step": 184 |
| }, |
| { |
| "epoch": 0.06466191552233617, |
| "grad_norm": 1.7495859751833545, |
| "learning_rate": 6.458333333333334e-05, |
| "loss": 0.8439, |
| "step": 186 |
| }, |
| { |
| "epoch": 0.06535720493655484, |
| "grad_norm": 2.971674018184174, |
| "learning_rate": 6.527777777777778e-05, |
| "loss": 0.9817, |
| "step": 188 |
| }, |
| { |
| "epoch": 0.06605249435077351, |
| "grad_norm": 2.3753292673540165, |
| "learning_rate": 6.597222222222223e-05, |
| "loss": 0.9919, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.06674778376499217, |
| "grad_norm": 2.7019148741557744, |
| "learning_rate": 6.666666666666667e-05, |
| "loss": 1.1442, |
| "step": 192 |
| }, |
| { |
| "epoch": 0.06744307317921085, |
| "grad_norm": 1.8670113287712482, |
| "learning_rate": 6.736111111111112e-05, |
| "loss": 0.8905, |
| "step": 194 |
| }, |
| { |
| "epoch": 0.06813836259342952, |
| "grad_norm": 1.4199145680760579, |
| "learning_rate": 6.805555555555556e-05, |
| "loss": 0.7223, |
| "step": 196 |
| }, |
| { |
| "epoch": 0.06883365200764818, |
| "grad_norm": 3.894200902880186, |
| "learning_rate": 6.875e-05, |
| "loss": 0.9005, |
| "step": 198 |
| }, |
| { |
| "epoch": 0.06952894142186686, |
| "grad_norm": 3.2710376491241955, |
| "learning_rate": 6.944444444444444e-05, |
| "loss": 1.1605, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.07022423083608552, |
| "grad_norm": 1.6459754670035065, |
| "learning_rate": 7.013888888888888e-05, |
| "loss": 0.8551, |
| "step": 202 |
| }, |
| { |
| "epoch": 0.0709195202503042, |
| "grad_norm": 4.361031640374508, |
| "learning_rate": 7.083333333333334e-05, |
| "loss": 0.7007, |
| "step": 204 |
| }, |
| { |
| "epoch": 0.07161480966452285, |
| "grad_norm": 3.573741549123141, |
| "learning_rate": 7.152777777777778e-05, |
| "loss": 1.1396, |
| "step": 206 |
| }, |
| { |
| "epoch": 0.07231009907874153, |
| "grad_norm": 3.296990311359108, |
| "learning_rate": 7.222222222222222e-05, |
| "loss": 0.8695, |
| "step": 208 |
| }, |
| { |
| "epoch": 0.07300538849296019, |
| "grad_norm": 6.769659823038884, |
| "learning_rate": 7.291666666666667e-05, |
| "loss": 1.0511, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.07370067790717887, |
| "grad_norm": 1.6695233666860303, |
| "learning_rate": 7.361111111111111e-05, |
| "loss": 0.8174, |
| "step": 212 |
| }, |
| { |
| "epoch": 0.07439596732139753, |
| "grad_norm": 1.3543755216281146, |
| "learning_rate": 7.430555555555557e-05, |
| "loss": 0.7137, |
| "step": 214 |
| }, |
| { |
| "epoch": 0.0750912567356162, |
| "grad_norm": 4.861063813937456, |
| "learning_rate": 7.500000000000001e-05, |
| "loss": 1.0132, |
| "step": 216 |
| }, |
| { |
| "epoch": 0.07578654614983486, |
| "grad_norm": 5.715913176528681, |
| "learning_rate": 7.569444444444445e-05, |
| "loss": 1.1657, |
| "step": 218 |
| }, |
| { |
| "epoch": 0.07648183556405354, |
| "grad_norm": 4.0193839358302235, |
| "learning_rate": 7.638888888888889e-05, |
| "loss": 1.0547, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.0771771249782722, |
| "grad_norm": 2.26849134517291, |
| "learning_rate": 7.708333333333334e-05, |
| "loss": 0.9073, |
| "step": 222 |
| }, |
| { |
| "epoch": 0.07787241439249087, |
| "grad_norm": 4.570943697810998, |
| "learning_rate": 7.777777777777778e-05, |
| "loss": 1.3726, |
| "step": 224 |
| }, |
| { |
| "epoch": 0.07856770380670955, |
| "grad_norm": 1.1942436910880105, |
| "learning_rate": 7.847222222222222e-05, |
| "loss": 0.8146, |
| "step": 226 |
| }, |
| { |
| "epoch": 0.07926299322092821, |
| "grad_norm": 2.111849588751211, |
| "learning_rate": 7.916666666666666e-05, |
| "loss": 1.1157, |
| "step": 228 |
| }, |
| { |
| "epoch": 0.07995828263514689, |
| "grad_norm": 1.0283088880069582, |
| "learning_rate": 7.986111111111112e-05, |
| "loss": 0.767, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.08065357204936555, |
| "grad_norm": 1.2834055069208525, |
| "learning_rate": 8.055555555555556e-05, |
| "loss": 0.6625, |
| "step": 232 |
| }, |
| { |
| "epoch": 0.08134886146358422, |
| "grad_norm": 1.3772942873595098, |
| "learning_rate": 8.125000000000001e-05, |
| "loss": 0.8065, |
| "step": 234 |
| }, |
| { |
| "epoch": 0.08204415087780288, |
| "grad_norm": 1.5113456549735176, |
| "learning_rate": 8.194444444444445e-05, |
| "loss": 0.8606, |
| "step": 236 |
| }, |
| { |
| "epoch": 0.08273944029202156, |
| "grad_norm": 1.5765846352838255, |
| "learning_rate": 8.263888888888889e-05, |
| "loss": 0.8335, |
| "step": 238 |
| }, |
| { |
| "epoch": 0.08343472970624022, |
| "grad_norm": 2.4873462478329404, |
| "learning_rate": 8.333333333333334e-05, |
| "loss": 0.9705, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.0841300191204589, |
| "grad_norm": 1.2369219925635513, |
| "learning_rate": 8.402777777777778e-05, |
| "loss": 0.6061, |
| "step": 242 |
| }, |
| { |
| "epoch": 0.08482530853467755, |
| "grad_norm": 2.542132212473201, |
| "learning_rate": 8.472222222222222e-05, |
| "loss": 0.9142, |
| "step": 244 |
| }, |
| { |
| "epoch": 0.08552059794889623, |
| "grad_norm": 2.0301734217803022, |
| "learning_rate": 8.541666666666666e-05, |
| "loss": 0.8997, |
| "step": 246 |
| }, |
| { |
| "epoch": 0.08621588736311489, |
| "grad_norm": 1.8605316982945626, |
| "learning_rate": 8.611111111111112e-05, |
| "loss": 1.005, |
| "step": 248 |
| }, |
| { |
| "epoch": 0.08691117677733357, |
| "grad_norm": 1.193555257951713, |
| "learning_rate": 8.680555555555556e-05, |
| "loss": 0.8617, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.08760646619155224, |
| "grad_norm": 1.243815428863678, |
| "learning_rate": 8.75e-05, |
| "loss": 0.6261, |
| "step": 252 |
| }, |
| { |
| "epoch": 0.0883017556057709, |
| "grad_norm": 1.6487754861704442, |
| "learning_rate": 8.819444444444445e-05, |
| "loss": 0.9219, |
| "step": 254 |
| }, |
| { |
| "epoch": 0.08899704501998958, |
| "grad_norm": 1.1768410857322613, |
| "learning_rate": 8.888888888888889e-05, |
| "loss": 0.8563, |
| "step": 256 |
| }, |
| { |
| "epoch": 0.08969233443420824, |
| "grad_norm": 1.1155265191420587, |
| "learning_rate": 8.958333333333335e-05, |
| "loss": 0.998, |
| "step": 258 |
| }, |
| { |
| "epoch": 0.09038762384842691, |
| "grad_norm": 1.4432456616674065, |
| "learning_rate": 9.027777777777779e-05, |
| "loss": 0.8755, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.09108291326264557, |
| "grad_norm": 1.6013486668654413, |
| "learning_rate": 9.097222222222223e-05, |
| "loss": 0.6998, |
| "step": 262 |
| }, |
| { |
| "epoch": 0.09177820267686425, |
| "grad_norm": 0.7869968770186737, |
| "learning_rate": 9.166666666666667e-05, |
| "loss": 0.7574, |
| "step": 264 |
| }, |
| { |
| "epoch": 0.09247349209108291, |
| "grad_norm": 2.1117903903864566, |
| "learning_rate": 9.236111111111112e-05, |
| "loss": 0.8436, |
| "step": 266 |
| }, |
| { |
| "epoch": 0.09316878150530158, |
| "grad_norm": 1.3582999584721895, |
| "learning_rate": 9.305555555555556e-05, |
| "loss": 0.8503, |
| "step": 268 |
| }, |
| { |
| "epoch": 0.09386407091952025, |
| "grad_norm": 1.5674581009005415, |
| "learning_rate": 9.375e-05, |
| "loss": 0.9561, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.09455936033373892, |
| "grad_norm": 1.0274246934159952, |
| "learning_rate": 9.444444444444444e-05, |
| "loss": 0.7537, |
| "step": 272 |
| }, |
| { |
| "epoch": 0.09525464974795758, |
| "grad_norm": 0.9595343432519174, |
| "learning_rate": 9.513888888888888e-05, |
| "loss": 0.7855, |
| "step": 274 |
| }, |
| { |
| "epoch": 0.09594993916217626, |
| "grad_norm": 1.3313115114367815, |
| "learning_rate": 9.583333333333334e-05, |
| "loss": 0.7009, |
| "step": 276 |
| }, |
| { |
| "epoch": 0.09664522857639492, |
| "grad_norm": 1.4409463331317498, |
| "learning_rate": 9.652777777777779e-05, |
| "loss": 1.0332, |
| "step": 278 |
| }, |
| { |
| "epoch": 0.09734051799061359, |
| "grad_norm": 1.5445848213023137, |
| "learning_rate": 9.722222222222223e-05, |
| "loss": 0.7352, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.09803580740483227, |
| "grad_norm": 1.855518620927316, |
| "learning_rate": 9.791666666666667e-05, |
| "loss": 0.7191, |
| "step": 282 |
| }, |
| { |
| "epoch": 0.09873109681905093, |
| "grad_norm": 1.4033396985161997, |
| "learning_rate": 9.861111111111112e-05, |
| "loss": 0.7886, |
| "step": 284 |
| }, |
| { |
| "epoch": 0.0994263862332696, |
| "grad_norm": 1.9951521438049904, |
| "learning_rate": 9.930555555555556e-05, |
| "loss": 0.902, |
| "step": 286 |
| }, |
| { |
| "epoch": 0.10012167564748826, |
| "grad_norm": 1.1360934228431687, |
| "learning_rate": 0.0001, |
| "loss": 0.8655, |
| "step": 288 |
| }, |
| { |
| "epoch": 0.10081696506170694, |
| "grad_norm": 1.1225735433900375, |
| "learning_rate": 0.00010069444444444445, |
| "loss": 0.6851, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.1015122544759256, |
| "grad_norm": 0.770446891371583, |
| "learning_rate": 0.00010138888888888889, |
| "loss": 0.7775, |
| "step": 292 |
| }, |
| { |
| "epoch": 0.10220754389014428, |
| "grad_norm": 0.797064180835607, |
| "learning_rate": 0.00010208333333333333, |
| "loss": 0.6567, |
| "step": 294 |
| }, |
| { |
| "epoch": 0.10290283330436294, |
| "grad_norm": 0.9876188134326442, |
| "learning_rate": 0.00010277777777777778, |
| "loss": 0.7299, |
| "step": 296 |
| }, |
| { |
| "epoch": 0.10359812271858161, |
| "grad_norm": 0.799601623643537, |
| "learning_rate": 0.00010347222222222223, |
| "loss": 0.7891, |
| "step": 298 |
| }, |
| { |
| "epoch": 0.10429341213280027, |
| "grad_norm": 0.8791512281362982, |
| "learning_rate": 0.00010416666666666667, |
| "loss": 0.677, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.10429341213280027, |
| "eval_loss": 0.6939894556999207, |
| "eval_runtime": 709.5268, |
| "eval_samples_per_second": 6.824, |
| "eval_steps_per_second": 0.214, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.10498870154701895, |
| "grad_norm": 1.329564645811689, |
| "learning_rate": 0.00010486111111111113, |
| "loss": 0.7511, |
| "step": 302 |
| }, |
| { |
| "epoch": 0.10568399096123761, |
| "grad_norm": 0.7130415058241142, |
| "learning_rate": 0.00010555555555555557, |
| "loss": 0.709, |
| "step": 304 |
| }, |
| { |
| "epoch": 0.10637928037545628, |
| "grad_norm": 1.1294733411370705, |
| "learning_rate": 0.00010625000000000001, |
| "loss": 0.7781, |
| "step": 306 |
| }, |
| { |
| "epoch": 0.10707456978967496, |
| "grad_norm": 0.581727680778003, |
| "learning_rate": 0.00010694444444444445, |
| "loss": 0.6296, |
| "step": 308 |
| }, |
| { |
| "epoch": 0.10776985920389362, |
| "grad_norm": 0.9855391514311871, |
| "learning_rate": 0.00010763888888888889, |
| "loss": 0.7852, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.1084651486181123, |
| "grad_norm": 0.5708379609857769, |
| "learning_rate": 0.00010833333333333333, |
| "loss": 0.6759, |
| "step": 312 |
| }, |
| { |
| "epoch": 0.10916043803233096, |
| "grad_norm": 0.7243330859752051, |
| "learning_rate": 0.00010902777777777777, |
| "loss": 0.6326, |
| "step": 314 |
| }, |
| { |
| "epoch": 0.10985572744654963, |
| "grad_norm": 1.3922163003264714, |
| "learning_rate": 0.00010972222222222224, |
| "loss": 0.5406, |
| "step": 316 |
| }, |
| { |
| "epoch": 0.11055101686076829, |
| "grad_norm": 2.604851740418153, |
| "learning_rate": 0.00011041666666666668, |
| "loss": 0.8013, |
| "step": 318 |
| }, |
| { |
| "epoch": 0.11124630627498697, |
| "grad_norm": 1.83511037983023, |
| "learning_rate": 0.00011111111111111112, |
| "loss": 0.8048, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.11194159568920563, |
| "grad_norm": 1.6737764614655666, |
| "learning_rate": 0.00011180555555555556, |
| "loss": 0.6943, |
| "step": 322 |
| }, |
| { |
| "epoch": 0.1126368851034243, |
| "grad_norm": 1.077667781888673, |
| "learning_rate": 0.00011250000000000001, |
| "loss": 0.8054, |
| "step": 324 |
| }, |
| { |
| "epoch": 0.11333217451764296, |
| "grad_norm": 0.5597693838209001, |
| "learning_rate": 0.00011319444444444445, |
| "loss": 0.6251, |
| "step": 326 |
| }, |
| { |
| "epoch": 0.11402746393186164, |
| "grad_norm": 0.5636865078063477, |
| "learning_rate": 0.00011388888888888889, |
| "loss": 0.6832, |
| "step": 328 |
| }, |
| { |
| "epoch": 0.1147227533460803, |
| "grad_norm": 0.6445176566556912, |
| "learning_rate": 0.00011458333333333333, |
| "loss": 0.6661, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.11541804276029898, |
| "grad_norm": 0.607532412895966, |
| "learning_rate": 0.00011527777777777777, |
| "loss": 0.7026, |
| "step": 332 |
| }, |
| { |
| "epoch": 0.11611333217451764, |
| "grad_norm": 1.0496949694240345, |
| "learning_rate": 0.00011597222222222224, |
| "loss": 0.6451, |
| "step": 334 |
| }, |
| { |
| "epoch": 0.11680862158873631, |
| "grad_norm": 0.9469233272179363, |
| "learning_rate": 0.00011666666666666668, |
| "loss": 0.6694, |
| "step": 336 |
| }, |
| { |
| "epoch": 0.11750391100295499, |
| "grad_norm": 1.3789882614528595, |
| "learning_rate": 0.00011736111111111112, |
| "loss": 0.7031, |
| "step": 338 |
| }, |
| { |
| "epoch": 0.11819920041717365, |
| "grad_norm": 1.2016589928587922, |
| "learning_rate": 0.00011805555555555556, |
| "loss": 0.6676, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.11889448983139232, |
| "grad_norm": 1.3112845410865746, |
| "learning_rate": 0.00011875, |
| "loss": 0.7236, |
| "step": 342 |
| }, |
| { |
| "epoch": 0.11958977924561098, |
| "grad_norm": 0.6278789052805739, |
| "learning_rate": 0.00011944444444444445, |
| "loss": 0.6249, |
| "step": 344 |
| }, |
| { |
| "epoch": 0.12028506865982966, |
| "grad_norm": 0.5405148902553805, |
| "learning_rate": 0.0001201388888888889, |
| "loss": 0.6412, |
| "step": 346 |
| }, |
| { |
| "epoch": 0.12098035807404832, |
| "grad_norm": 1.6923616138961255, |
| "learning_rate": 0.00012083333333333333, |
| "loss": 0.7017, |
| "step": 348 |
| }, |
| { |
| "epoch": 0.121675647488267, |
| "grad_norm": 1.5994113095813072, |
| "learning_rate": 0.00012152777777777777, |
| "loss": 0.7075, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.12237093690248566, |
| "grad_norm": 2.3083516502461783, |
| "learning_rate": 0.00012222222222222224, |
| "loss": 0.7755, |
| "step": 352 |
| }, |
| { |
| "epoch": 0.12306622631670433, |
| "grad_norm": 0.8769167165306447, |
| "learning_rate": 0.00012291666666666668, |
| "loss": 0.7428, |
| "step": 354 |
| }, |
| { |
| "epoch": 0.12376151573092299, |
| "grad_norm": 0.4456469638797518, |
| "learning_rate": 0.00012361111111111112, |
| "loss": 0.7002, |
| "step": 356 |
| }, |
| { |
| "epoch": 0.12445680514514167, |
| "grad_norm": 0.8199993647792723, |
| "learning_rate": 0.00012430555555555556, |
| "loss": 0.7222, |
| "step": 358 |
| }, |
| { |
| "epoch": 0.12515209455936033, |
| "grad_norm": 0.3741084955970339, |
| "learning_rate": 0.000125, |
| "loss": 0.6821, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.125847383973579, |
| "grad_norm": 0.2972203241099783, |
| "learning_rate": 0.00012569444444444444, |
| "loss": 0.7083, |
| "step": 362 |
| }, |
| { |
| "epoch": 0.12654267338779768, |
| "grad_norm": 0.40651740632246575, |
| "learning_rate": 0.00012638888888888888, |
| "loss": 0.7144, |
| "step": 364 |
| }, |
| { |
| "epoch": 0.12723796280201635, |
| "grad_norm": 0.4388093199704831, |
| "learning_rate": 0.00012708333333333332, |
| "loss": 0.6863, |
| "step": 366 |
| }, |
| { |
| "epoch": 0.127933252216235, |
| "grad_norm": 0.45131267837810835, |
| "learning_rate": 0.00012777777777777776, |
| "loss": 0.6329, |
| "step": 368 |
| }, |
| { |
| "epoch": 0.12862854163045367, |
| "grad_norm": 0.3837830352665476, |
| "learning_rate": 0.00012847222222222223, |
| "loss": 0.6938, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.12932383104467235, |
| "grad_norm": 0.6378624076702905, |
| "learning_rate": 0.00012916666666666667, |
| "loss": 0.6433, |
| "step": 372 |
| }, |
| { |
| "epoch": 0.13001912045889102, |
| "grad_norm": 0.41344403509072675, |
| "learning_rate": 0.0001298611111111111, |
| "loss": 0.6892, |
| "step": 374 |
| }, |
| { |
| "epoch": 0.13071440987310967, |
| "grad_norm": 1.3428471718098582, |
| "learning_rate": 0.00013055555555555555, |
| "loss": 0.7793, |
| "step": 376 |
| }, |
| { |
| "epoch": 0.13140969928732835, |
| "grad_norm": 0.6202275876483008, |
| "learning_rate": 0.00013125000000000002, |
| "loss": 0.6836, |
| "step": 378 |
| }, |
| { |
| "epoch": 0.13210498870154702, |
| "grad_norm": 0.774014383396004, |
| "learning_rate": 0.00013194444444444446, |
| "loss": 0.701, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.1328002781157657, |
| "grad_norm": 1.0650744092292224, |
| "learning_rate": 0.0001326388888888889, |
| "loss": 0.6562, |
| "step": 382 |
| }, |
| { |
| "epoch": 0.13349556752998434, |
| "grad_norm": 0.5540168492927294, |
| "learning_rate": 0.00013333333333333334, |
| "loss": 0.6267, |
| "step": 384 |
| }, |
| { |
| "epoch": 0.13419085694420302, |
| "grad_norm": 0.6403086140920178, |
| "learning_rate": 0.00013402777777777778, |
| "loss": 0.6268, |
| "step": 386 |
| }, |
| { |
| "epoch": 0.1348861463584217, |
| "grad_norm": 0.9197218298114509, |
| "learning_rate": 0.00013472222222222225, |
| "loss": 0.7208, |
| "step": 388 |
| }, |
| { |
| "epoch": 0.13558143577264037, |
| "grad_norm": 0.4235387152144674, |
| "learning_rate": 0.0001354166666666667, |
| "loss": 0.6057, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.13627672518685904, |
| "grad_norm": 0.425648636635043, |
| "learning_rate": 0.00013611111111111113, |
| "loss": 0.6387, |
| "step": 392 |
| }, |
| { |
| "epoch": 0.1369720146010777, |
| "grad_norm": 1.034952396366508, |
| "learning_rate": 0.00013680555555555557, |
| "loss": 0.7307, |
| "step": 394 |
| }, |
| { |
| "epoch": 0.13766730401529637, |
| "grad_norm": 0.8865505732510678, |
| "learning_rate": 0.0001375, |
| "loss": 0.6547, |
| "step": 396 |
| }, |
| { |
| "epoch": 0.13836259342951504, |
| "grad_norm": 0.6490094573088415, |
| "learning_rate": 0.00013819444444444445, |
| "loss": 0.6909, |
| "step": 398 |
| }, |
| { |
| "epoch": 0.13905788284373372, |
| "grad_norm": 0.8592713330921486, |
| "learning_rate": 0.0001388888888888889, |
| "loss": 0.6442, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.13975317225795236, |
| "grad_norm": 1.0279375953862069, |
| "learning_rate": 0.00013958333333333333, |
| "loss": 0.6907, |
| "step": 402 |
| }, |
| { |
| "epoch": 0.14044846167217104, |
| "grad_norm": 0.6151057351983696, |
| "learning_rate": 0.00014027777777777777, |
| "loss": 0.697, |
| "step": 404 |
| }, |
| { |
| "epoch": 0.1411437510863897, |
| "grad_norm": 0.7417879717121144, |
| "learning_rate": 0.00014097222222222224, |
| "loss": 0.698, |
| "step": 406 |
| }, |
| { |
| "epoch": 0.1418390405006084, |
| "grad_norm": 0.5161453091859882, |
| "learning_rate": 0.00014166666666666668, |
| "loss": 0.6915, |
| "step": 408 |
| }, |
| { |
| "epoch": 0.14253432991482703, |
| "grad_norm": 0.5667521096080546, |
| "learning_rate": 0.00014236111111111112, |
| "loss": 0.6609, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.1432296193290457, |
| "grad_norm": 0.7560317583703429, |
| "learning_rate": 0.00014305555555555556, |
| "loss": 0.6392, |
| "step": 412 |
| }, |
| { |
| "epoch": 0.14392490874326438, |
| "grad_norm": 0.5456303981546313, |
| "learning_rate": 0.00014375, |
| "loss": 0.7561, |
| "step": 414 |
| }, |
| { |
| "epoch": 0.14462019815748306, |
| "grad_norm": 0.4218469018151631, |
| "learning_rate": 0.00014444444444444444, |
| "loss": 0.675, |
| "step": 416 |
| }, |
| { |
| "epoch": 0.14531548757170173, |
| "grad_norm": 0.3790517974518404, |
| "learning_rate": 0.00014513888888888888, |
| "loss": 0.6697, |
| "step": 418 |
| }, |
| { |
| "epoch": 0.14601077698592038, |
| "grad_norm": 0.36495171567914964, |
| "learning_rate": 0.00014583333333333335, |
| "loss": 0.6467, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.14670606640013906, |
| "grad_norm": 0.3793520501177419, |
| "learning_rate": 0.00014652777777777779, |
| "loss": 0.6234, |
| "step": 422 |
| }, |
| { |
| "epoch": 0.14740135581435773, |
| "grad_norm": 0.6214905371013544, |
| "learning_rate": 0.00014722222222222223, |
| "loss": 0.64, |
| "step": 424 |
| }, |
| { |
| "epoch": 0.1480966452285764, |
| "grad_norm": 0.4103438113660832, |
| "learning_rate": 0.0001479166666666667, |
| "loss": 0.6174, |
| "step": 426 |
| }, |
| { |
| "epoch": 0.14879193464279505, |
| "grad_norm": 0.5802311029669485, |
| "learning_rate": 0.00014861111111111113, |
| "loss": 0.6941, |
| "step": 428 |
| }, |
| { |
| "epoch": 0.14948722405701373, |
| "grad_norm": 0.4876572424583591, |
| "learning_rate": 0.00014930555555555557, |
| "loss": 0.6178, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.1501825134712324, |
| "grad_norm": 0.6252220373472688, |
| "learning_rate": 0.00015000000000000001, |
| "loss": 0.692, |
| "step": 432 |
| }, |
| { |
| "epoch": 0.15087780288545108, |
| "grad_norm": 1.516758455176553, |
| "learning_rate": 0.00015069444444444445, |
| "loss": 0.6812, |
| "step": 434 |
| }, |
| { |
| "epoch": 0.15157309229966973, |
| "grad_norm": 0.48294685446689867, |
| "learning_rate": 0.0001513888888888889, |
| "loss": 0.7085, |
| "step": 436 |
| }, |
| { |
| "epoch": 0.1522683817138884, |
| "grad_norm": 0.414196721150167, |
| "learning_rate": 0.00015208333333333333, |
| "loss": 0.673, |
| "step": 438 |
| }, |
| { |
| "epoch": 0.15296367112810708, |
| "grad_norm": 0.7294881062503469, |
| "learning_rate": 0.00015277777777777777, |
| "loss": 0.6415, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.15365896054232575, |
| "grad_norm": 0.4846932912054438, |
| "learning_rate": 0.00015347222222222224, |
| "loss": 0.6644, |
| "step": 442 |
| }, |
| { |
| "epoch": 0.1543542499565444, |
| "grad_norm": 0.6067400167426414, |
| "learning_rate": 0.00015416666666666668, |
| "loss": 0.6816, |
| "step": 444 |
| }, |
| { |
| "epoch": 0.15504953937076307, |
| "grad_norm": 0.5586278026791864, |
| "learning_rate": 0.00015486111111111112, |
| "loss": 0.6223, |
| "step": 446 |
| }, |
| { |
| "epoch": 0.15574482878498175, |
| "grad_norm": 0.3947236166856149, |
| "learning_rate": 0.00015555555555555556, |
| "loss": 0.7297, |
| "step": 448 |
| }, |
| { |
| "epoch": 0.15644011819920042, |
| "grad_norm": 0.371457756887947, |
| "learning_rate": 0.00015625, |
| "loss": 0.6099, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.15644011819920042, |
| "eval_loss": 0.6634477376937866, |
| "eval_runtime": 706.8027, |
| "eval_samples_per_second": 6.851, |
| "eval_steps_per_second": 0.215, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.1571354076134191, |
| "grad_norm": 0.4157928192338373, |
| "learning_rate": 0.00015694444444444444, |
| "loss": 0.7163, |
| "step": 452 |
| }, |
| { |
| "epoch": 0.15783069702763775, |
| "grad_norm": 0.6318513446857751, |
| "learning_rate": 0.00015763888888888888, |
| "loss": 0.658, |
| "step": 454 |
| }, |
| { |
| "epoch": 0.15852598644185642, |
| "grad_norm": 0.3188170086808347, |
| "learning_rate": 0.00015833333333333332, |
| "loss": 0.5966, |
| "step": 456 |
| }, |
| { |
| "epoch": 0.1592212758560751, |
| "grad_norm": 0.4873146052629784, |
| "learning_rate": 0.00015902777777777776, |
| "loss": 0.687, |
| "step": 458 |
| }, |
| { |
| "epoch": 0.15991656527029377, |
| "grad_norm": 0.3939618900754228, |
| "learning_rate": 0.00015972222222222223, |
| "loss": 0.6704, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.16061185468451242, |
| "grad_norm": 0.43442241310529234, |
| "learning_rate": 0.00016041666666666667, |
| "loss": 0.6941, |
| "step": 462 |
| }, |
| { |
| "epoch": 0.1613071440987311, |
| "grad_norm": 0.5944395304786771, |
| "learning_rate": 0.0001611111111111111, |
| "loss": 0.5934, |
| "step": 464 |
| }, |
| { |
| "epoch": 0.16200243351294977, |
| "grad_norm": 0.6186667715273749, |
| "learning_rate": 0.00016180555555555555, |
| "loss": 0.6604, |
| "step": 466 |
| }, |
| { |
| "epoch": 0.16269772292716844, |
| "grad_norm": 0.8945416923537124, |
| "learning_rate": 0.00016250000000000002, |
| "loss": 0.6224, |
| "step": 468 |
| }, |
| { |
| "epoch": 0.1633930123413871, |
| "grad_norm": 1.0526200270890014, |
| "learning_rate": 0.00016319444444444446, |
| "loss": 0.616, |
| "step": 470 |
| }, |
| { |
| "epoch": 0.16408830175560576, |
| "grad_norm": 0.4575985966727194, |
| "learning_rate": 0.0001638888888888889, |
| "loss": 0.6591, |
| "step": 472 |
| }, |
| { |
| "epoch": 0.16478359116982444, |
| "grad_norm": 0.4499931204645402, |
| "learning_rate": 0.00016458333333333334, |
| "loss": 0.6644, |
| "step": 474 |
| }, |
| { |
| "epoch": 0.16547888058404311, |
| "grad_norm": 0.8365574384741885, |
| "learning_rate": 0.00016527777777777778, |
| "loss": 0.6503, |
| "step": 476 |
| }, |
| { |
| "epoch": 0.1661741699982618, |
| "grad_norm": 0.5265284650577726, |
| "learning_rate": 0.00016597222222222225, |
| "loss": 0.557, |
| "step": 478 |
| }, |
| { |
| "epoch": 0.16686945941248044, |
| "grad_norm": 0.6004742828917694, |
| "learning_rate": 0.0001666666666666667, |
| "loss": 0.696, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.1675647488266991, |
| "grad_norm": 0.8653134209549845, |
| "learning_rate": 0.00016736111111111113, |
| "loss": 0.6544, |
| "step": 482 |
| }, |
| { |
| "epoch": 0.1682600382409178, |
| "grad_norm": 1.033230873418064, |
| "learning_rate": 0.00016805555555555557, |
| "loss": 0.641, |
| "step": 484 |
| }, |
| { |
| "epoch": 0.16895532765513646, |
| "grad_norm": 0.45457193405665913, |
| "learning_rate": 0.00016875, |
| "loss": 0.5952, |
| "step": 486 |
| }, |
| { |
| "epoch": 0.1696506170693551, |
| "grad_norm": 0.7413902971651539, |
| "learning_rate": 0.00016944444444444445, |
| "loss": 0.6306, |
| "step": 488 |
| }, |
| { |
| "epoch": 0.17034590648357378, |
| "grad_norm": 0.5808764628223875, |
| "learning_rate": 0.0001701388888888889, |
| "loss": 0.6714, |
| "step": 490 |
| }, |
| { |
| "epoch": 0.17104119589779246, |
| "grad_norm": 0.42505370383777924, |
| "learning_rate": 0.00017083333333333333, |
| "loss": 0.6432, |
| "step": 492 |
| }, |
| { |
| "epoch": 0.17173648531201113, |
| "grad_norm": 0.9403542369255028, |
| "learning_rate": 0.00017152777777777777, |
| "loss": 0.626, |
| "step": 494 |
| }, |
| { |
| "epoch": 0.17243177472622978, |
| "grad_norm": 0.41714707593148775, |
| "learning_rate": 0.00017222222222222224, |
| "loss": 0.6132, |
| "step": 496 |
| }, |
| { |
| "epoch": 0.17312706414044846, |
| "grad_norm": 0.6660896849726371, |
| "learning_rate": 0.00017291666666666668, |
| "loss": 0.7009, |
| "step": 498 |
| }, |
| { |
| "epoch": 0.17382235355466713, |
| "grad_norm": 0.5079993412341118, |
| "learning_rate": 0.00017361111111111112, |
| "loss": 0.6808, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.1745176429688858, |
| "grad_norm": 0.3426487947129772, |
| "learning_rate": 0.00017430555555555556, |
| "loss": 0.687, |
| "step": 502 |
| }, |
| { |
| "epoch": 0.17521293238310448, |
| "grad_norm": 0.7316051797581208, |
| "learning_rate": 0.000175, |
| "loss": 0.686, |
| "step": 504 |
| }, |
| { |
| "epoch": 0.17590822179732313, |
| "grad_norm": 0.4619785653282459, |
| "learning_rate": 0.00017569444444444444, |
| "loss": 0.7019, |
| "step": 506 |
| }, |
| { |
| "epoch": 0.1766035112115418, |
| "grad_norm": 0.47786977912472967, |
| "learning_rate": 0.0001763888888888889, |
| "loss": 0.7097, |
| "step": 508 |
| }, |
| { |
| "epoch": 0.17729880062576048, |
| "grad_norm": 0.4204398291864812, |
| "learning_rate": 0.00017708333333333335, |
| "loss": 0.6715, |
| "step": 510 |
| }, |
| { |
| "epoch": 0.17799409003997915, |
| "grad_norm": 0.5257761541709909, |
| "learning_rate": 0.00017777777777777779, |
| "loss": 0.6675, |
| "step": 512 |
| }, |
| { |
| "epoch": 0.1786893794541978, |
| "grad_norm": 0.5752239420884688, |
| "learning_rate": 0.00017847222222222225, |
| "loss": 0.5847, |
| "step": 514 |
| }, |
| { |
| "epoch": 0.17938466886841647, |
| "grad_norm": 0.5611432017880666, |
| "learning_rate": 0.0001791666666666667, |
| "loss": 0.5978, |
| "step": 516 |
| }, |
| { |
| "epoch": 0.18007995828263515, |
| "grad_norm": 0.5147097358785945, |
| "learning_rate": 0.00017986111111111113, |
| "loss": 0.641, |
| "step": 518 |
| }, |
| { |
| "epoch": 0.18077524769685382, |
| "grad_norm": 0.4387330962078066, |
| "learning_rate": 0.00018055555555555557, |
| "loss": 0.5771, |
| "step": 520 |
| }, |
| { |
| "epoch": 0.18147053711107247, |
| "grad_norm": 0.5712526440788663, |
| "learning_rate": 0.00018125000000000001, |
| "loss": 0.6212, |
| "step": 522 |
| }, |
| { |
| "epoch": 0.18216582652529115, |
| "grad_norm": 0.6357133907029916, |
| "learning_rate": 0.00018194444444444445, |
| "loss": 0.5981, |
| "step": 524 |
| }, |
| { |
| "epoch": 0.18286111593950982, |
| "grad_norm": 1.2129620643054042, |
| "learning_rate": 0.0001826388888888889, |
| "loss": 0.5848, |
| "step": 526 |
| }, |
| { |
| "epoch": 0.1835564053537285, |
| "grad_norm": 1.651599276688714, |
| "learning_rate": 0.00018333333333333334, |
| "loss": 0.7089, |
| "step": 528 |
| }, |
| { |
| "epoch": 0.18425169476794717, |
| "grad_norm": 0.9051688342434142, |
| "learning_rate": 0.00018402777777777778, |
| "loss": 0.7255, |
| "step": 530 |
| }, |
| { |
| "epoch": 0.18494698418216582, |
| "grad_norm": 0.6932441010863684, |
| "learning_rate": 0.00018472222222222224, |
| "loss": 0.6497, |
| "step": 532 |
| }, |
| { |
| "epoch": 0.1856422735963845, |
| "grad_norm": 1.2163006408955903, |
| "learning_rate": 0.00018541666666666668, |
| "loss": 0.7585, |
| "step": 534 |
| }, |
| { |
| "epoch": 0.18633756301060317, |
| "grad_norm": 0.4387074048377486, |
| "learning_rate": 0.00018611111111111112, |
| "loss": 0.6541, |
| "step": 536 |
| }, |
| { |
| "epoch": 0.18703285242482184, |
| "grad_norm": 0.49430581216479236, |
| "learning_rate": 0.00018680555555555556, |
| "loss": 0.6489, |
| "step": 538 |
| }, |
| { |
| "epoch": 0.1877281418390405, |
| "grad_norm": 0.47561442420926275, |
| "learning_rate": 0.0001875, |
| "loss": 0.6406, |
| "step": 540 |
| }, |
| { |
| "epoch": 0.18842343125325917, |
| "grad_norm": 0.7046092923664611, |
| "learning_rate": 0.00018819444444444444, |
| "loss": 0.6099, |
| "step": 542 |
| }, |
| { |
| "epoch": 0.18911872066747784, |
| "grad_norm": 0.46256136269201026, |
| "learning_rate": 0.00018888888888888888, |
| "loss": 0.6992, |
| "step": 544 |
| }, |
| { |
| "epoch": 0.18981401008169652, |
| "grad_norm": 0.5489877082982352, |
| "learning_rate": 0.00018958333333333332, |
| "loss": 0.5817, |
| "step": 546 |
| }, |
| { |
| "epoch": 0.19050929949591516, |
| "grad_norm": 0.46129441798908893, |
| "learning_rate": 0.00019027777777777776, |
| "loss": 0.645, |
| "step": 548 |
| }, |
| { |
| "epoch": 0.19120458891013384, |
| "grad_norm": 0.3724226655450749, |
| "learning_rate": 0.00019097222222222223, |
| "loss": 0.6177, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.1918998783243525, |
| "grad_norm": 0.4625796323902992, |
| "learning_rate": 0.00019166666666666667, |
| "loss": 0.6702, |
| "step": 552 |
| }, |
| { |
| "epoch": 0.1925951677385712, |
| "grad_norm": 1.0142124821047231, |
| "learning_rate": 0.0001923611111111111, |
| "loss": 0.7173, |
| "step": 554 |
| }, |
| { |
| "epoch": 0.19329045715278984, |
| "grad_norm": 0.5452830720753045, |
| "learning_rate": 0.00019305555555555558, |
| "loss": 0.6897, |
| "step": 556 |
| }, |
| { |
| "epoch": 0.1939857465670085, |
| "grad_norm": 0.5723214484723252, |
| "learning_rate": 0.00019375000000000002, |
| "loss": 0.6495, |
| "step": 558 |
| }, |
| { |
| "epoch": 0.19468103598122719, |
| "grad_norm": 0.39925844595054966, |
| "learning_rate": 0.00019444444444444446, |
| "loss": 0.6479, |
| "step": 560 |
| }, |
| { |
| "epoch": 0.19537632539544586, |
| "grad_norm": 0.3575565088751118, |
| "learning_rate": 0.0001951388888888889, |
| "loss": 0.6593, |
| "step": 562 |
| }, |
| { |
| "epoch": 0.19607161480966454, |
| "grad_norm": 0.6119892689536569, |
| "learning_rate": 0.00019583333333333334, |
| "loss": 0.6831, |
| "step": 564 |
| }, |
| { |
| "epoch": 0.19676690422388318, |
| "grad_norm": 0.5076569073121309, |
| "learning_rate": 0.00019652777777777778, |
| "loss": 0.6294, |
| "step": 566 |
| }, |
| { |
| "epoch": 0.19746219363810186, |
| "grad_norm": 0.622226505121079, |
| "learning_rate": 0.00019722222222222225, |
| "loss": 0.5551, |
| "step": 568 |
| }, |
| { |
| "epoch": 0.19815748305232053, |
| "grad_norm": 0.4106526912254415, |
| "learning_rate": 0.0001979166666666667, |
| "loss": 0.6503, |
| "step": 570 |
| }, |
| { |
| "epoch": 0.1988527724665392, |
| "grad_norm": 0.47164751383243125, |
| "learning_rate": 0.00019861111111111113, |
| "loss": 0.6992, |
| "step": 572 |
| }, |
| { |
| "epoch": 0.19954806188075785, |
| "grad_norm": 0.492328312395826, |
| "learning_rate": 0.00019930555555555557, |
| "loss": 0.7183, |
| "step": 574 |
| }, |
| { |
| "epoch": 0.20024335129497653, |
| "grad_norm": 0.5087240677439067, |
| "learning_rate": 0.0002, |
| "loss": 0.6429, |
| "step": 576 |
| }, |
| { |
| "epoch": 0.2009386407091952, |
| "grad_norm": 0.9311216098504759, |
| "learning_rate": 0.00019999992632143608, |
| "loss": 0.6586, |
| "step": 578 |
| }, |
| { |
| "epoch": 0.20163393012341388, |
| "grad_norm": 0.38840935170189844, |
| "learning_rate": 0.00019999970528585288, |
| "loss": 0.5905, |
| "step": 580 |
| }, |
| { |
| "epoch": 0.20232921953763253, |
| "grad_norm": 0.5966472765038273, |
| "learning_rate": 0.0001999993368935761, |
| "loss": 0.7112, |
| "step": 582 |
| }, |
| { |
| "epoch": 0.2030245089518512, |
| "grad_norm": 0.535232520705813, |
| "learning_rate": 0.00019999882114514863, |
| "loss": 0.7036, |
| "step": 584 |
| }, |
| { |
| "epoch": 0.20371979836606988, |
| "grad_norm": 0.5039916125175515, |
| "learning_rate": 0.0001999981580413304, |
| "loss": 0.7485, |
| "step": 586 |
| }, |
| { |
| "epoch": 0.20441508778028855, |
| "grad_norm": 0.6733262892097686, |
| "learning_rate": 0.0001999973475830986, |
| "loss": 0.7312, |
| "step": 588 |
| }, |
| { |
| "epoch": 0.20511037719450723, |
| "grad_norm": 0.42805740684614807, |
| "learning_rate": 0.00019999638977164747, |
| "loss": 0.6356, |
| "step": 590 |
| }, |
| { |
| "epoch": 0.20580566660872587, |
| "grad_norm": 0.976918546573707, |
| "learning_rate": 0.00019999528460838844, |
| "loss": 0.666, |
| "step": 592 |
| }, |
| { |
| "epoch": 0.20650095602294455, |
| "grad_norm": 0.674463735133944, |
| "learning_rate": 0.00019999403209495, |
| "loss": 0.6735, |
| "step": 594 |
| }, |
| { |
| "epoch": 0.20719624543716322, |
| "grad_norm": 0.412232119296174, |
| "learning_rate": 0.00019999263223317786, |
| "loss": 0.6364, |
| "step": 596 |
| }, |
| { |
| "epoch": 0.2078915348513819, |
| "grad_norm": 0.3611034413524545, |
| "learning_rate": 0.0001999910850251348, |
| "loss": 0.6368, |
| "step": 598 |
| }, |
| { |
| "epoch": 0.20858682426560055, |
| "grad_norm": 0.542506398091656, |
| "learning_rate": 0.0001999893904731007, |
| "loss": 0.7292, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.20858682426560055, |
| "eval_loss": 0.6517693400382996, |
| "eval_runtime": 707.4674, |
| "eval_samples_per_second": 6.844, |
| "eval_steps_per_second": 0.215, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.20928211367981922, |
| "grad_norm": 0.4799286290563759, |
| "learning_rate": 0.0001999875485795727, |
| "loss": 0.6609, |
| "step": 602 |
| }, |
| { |
| "epoch": 0.2099774030940379, |
| "grad_norm": 0.41830446575524177, |
| "learning_rate": 0.0001999855593472649, |
| "loss": 0.6084, |
| "step": 604 |
| }, |
| { |
| "epoch": 0.21067269250825657, |
| "grad_norm": 0.8865562086549428, |
| "learning_rate": 0.00019998342277910856, |
| "loss": 0.6204, |
| "step": 606 |
| }, |
| { |
| "epoch": 0.21136798192247522, |
| "grad_norm": 0.5567410910026902, |
| "learning_rate": 0.00019998113887825206, |
| "loss": 0.7087, |
| "step": 608 |
| }, |
| { |
| "epoch": 0.2120632713366939, |
| "grad_norm": 0.4229098163428897, |
| "learning_rate": 0.000199978707648061, |
| "loss": 0.6403, |
| "step": 610 |
| }, |
| { |
| "epoch": 0.21275856075091257, |
| "grad_norm": 0.7793555918286379, |
| "learning_rate": 0.00019997612909211784, |
| "loss": 0.6147, |
| "step": 612 |
| }, |
| { |
| "epoch": 0.21345385016513124, |
| "grad_norm": 0.49302383416102263, |
| "learning_rate": 0.00019997340321422228, |
| "loss": 0.5771, |
| "step": 614 |
| }, |
| { |
| "epoch": 0.21414913957934992, |
| "grad_norm": 1.0807551778698754, |
| "learning_rate": 0.00019997053001839115, |
| "loss": 0.6538, |
| "step": 616 |
| }, |
| { |
| "epoch": 0.21484442899356856, |
| "grad_norm": 0.5935042706034646, |
| "learning_rate": 0.0001999675095088583, |
| "loss": 0.6058, |
| "step": 618 |
| }, |
| { |
| "epoch": 0.21553971840778724, |
| "grad_norm": 0.5151203960605611, |
| "learning_rate": 0.00019996434169007468, |
| "loss": 0.6138, |
| "step": 620 |
| }, |
| { |
| "epoch": 0.21623500782200591, |
| "grad_norm": 0.7837579499467073, |
| "learning_rate": 0.00019996102656670824, |
| "loss": 0.657, |
| "step": 622 |
| }, |
| { |
| "epoch": 0.2169302972362246, |
| "grad_norm": 0.5438202413527335, |
| "learning_rate": 0.0001999575641436441, |
| "loss": 0.6642, |
| "step": 624 |
| }, |
| { |
| "epoch": 0.21762558665044324, |
| "grad_norm": 0.5566641749580342, |
| "learning_rate": 0.00019995395442598435, |
| "loss": 0.6813, |
| "step": 626 |
| }, |
| { |
| "epoch": 0.2183208760646619, |
| "grad_norm": 0.9345190838089675, |
| "learning_rate": 0.0001999501974190482, |
| "loss": 0.684, |
| "step": 628 |
| }, |
| { |
| "epoch": 0.2190161654788806, |
| "grad_norm": 0.48677420225080537, |
| "learning_rate": 0.00019994629312837186, |
| "loss": 0.6481, |
| "step": 630 |
| }, |
| { |
| "epoch": 0.21971145489309926, |
| "grad_norm": 1.3826373483976924, |
| "learning_rate": 0.00019994224155970856, |
| "loss": 0.7222, |
| "step": 632 |
| }, |
| { |
| "epoch": 0.2204067443073179, |
| "grad_norm": 0.9031189580678588, |
| "learning_rate": 0.00019993804271902857, |
| "loss": 0.6809, |
| "step": 634 |
| }, |
| { |
| "epoch": 0.22110203372153658, |
| "grad_norm": 0.5674519795217096, |
| "learning_rate": 0.00019993369661251923, |
| "loss": 0.6184, |
| "step": 636 |
| }, |
| { |
| "epoch": 0.22179732313575526, |
| "grad_norm": 2.5294344265719135, |
| "learning_rate": 0.0001999292032465848, |
| "loss": 0.7686, |
| "step": 638 |
| }, |
| { |
| "epoch": 0.22249261254997393, |
| "grad_norm": 2.8855656605632047, |
| "learning_rate": 0.00019992456262784658, |
| "loss": 0.8215, |
| "step": 640 |
| }, |
| { |
| "epoch": 0.2231879019641926, |
| "grad_norm": 1.5331524159142917, |
| "learning_rate": 0.00019991977476314286, |
| "loss": 0.7057, |
| "step": 642 |
| }, |
| { |
| "epoch": 0.22388319137841126, |
| "grad_norm": 0.8086145005161695, |
| "learning_rate": 0.0001999148396595289, |
| "loss": 0.6289, |
| "step": 644 |
| }, |
| { |
| "epoch": 0.22457848079262993, |
| "grad_norm": 0.767389193699688, |
| "learning_rate": 0.00019990975732427693, |
| "loss": 0.6117, |
| "step": 646 |
| }, |
| { |
| "epoch": 0.2252737702068486, |
| "grad_norm": 0.9275190697426144, |
| "learning_rate": 0.0001999045277648761, |
| "loss": 0.6582, |
| "step": 648 |
| }, |
| { |
| "epoch": 0.22596905962106728, |
| "grad_norm": 0.8004230202091956, |
| "learning_rate": 0.00019989915098903257, |
| "loss": 0.5387, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.22666434903528593, |
| "grad_norm": 0.8210833181388095, |
| "learning_rate": 0.0001998936270046694, |
| "loss": 0.5941, |
| "step": 652 |
| }, |
| { |
| "epoch": 0.2273596384495046, |
| "grad_norm": 2.1189675938584616, |
| "learning_rate": 0.00019988795581992656, |
| "loss": 0.7504, |
| "step": 654 |
| }, |
| { |
| "epoch": 0.22805492786372328, |
| "grad_norm": 0.5952032026226816, |
| "learning_rate": 0.000199882137443161, |
| "loss": 0.6643, |
| "step": 656 |
| }, |
| { |
| "epoch": 0.22875021727794195, |
| "grad_norm": 0.40481972159485846, |
| "learning_rate": 0.00019987617188294642, |
| "loss": 0.5225, |
| "step": 658 |
| }, |
| { |
| "epoch": 0.2294455066921606, |
| "grad_norm": 0.7478677085785272, |
| "learning_rate": 0.00019987005914807356, |
| "loss": 0.6561, |
| "step": 660 |
| }, |
| { |
| "epoch": 0.23014079610637928, |
| "grad_norm": 0.4468692254604696, |
| "learning_rate": 0.00019986379924754997, |
| "loss": 0.6262, |
| "step": 662 |
| }, |
| { |
| "epoch": 0.23083608552059795, |
| "grad_norm": 0.46697718493106893, |
| "learning_rate": 0.00019985739219060002, |
| "loss": 0.6178, |
| "step": 664 |
| }, |
| { |
| "epoch": 0.23153137493481663, |
| "grad_norm": 0.388728867903269, |
| "learning_rate": 0.000199850837986665, |
| "loss": 0.5544, |
| "step": 666 |
| }, |
| { |
| "epoch": 0.23222666434903527, |
| "grad_norm": 0.5731403139849517, |
| "learning_rate": 0.000199844136645403, |
| "loss": 0.5906, |
| "step": 668 |
| }, |
| { |
| "epoch": 0.23292195376325395, |
| "grad_norm": 0.7601215827849259, |
| "learning_rate": 0.0001998372881766889, |
| "loss": 0.6577, |
| "step": 670 |
| }, |
| { |
| "epoch": 0.23361724317747262, |
| "grad_norm": 0.8997786413980279, |
| "learning_rate": 0.00019983029259061446, |
| "loss": 0.6388, |
| "step": 672 |
| }, |
| { |
| "epoch": 0.2343125325916913, |
| "grad_norm": 0.519349430301096, |
| "learning_rate": 0.00019982314989748813, |
| "loss": 0.6561, |
| "step": 674 |
| }, |
| { |
| "epoch": 0.23500782200590997, |
| "grad_norm": 0.4918913590728881, |
| "learning_rate": 0.00019981586010783513, |
| "loss": 0.6176, |
| "step": 676 |
| }, |
| { |
| "epoch": 0.23570311142012862, |
| "grad_norm": 0.5324672662478441, |
| "learning_rate": 0.00019980842323239756, |
| "loss": 0.5801, |
| "step": 678 |
| }, |
| { |
| "epoch": 0.2363984008343473, |
| "grad_norm": 0.6375745879180794, |
| "learning_rate": 0.00019980083928213417, |
| "loss": 0.6885, |
| "step": 680 |
| }, |
| { |
| "epoch": 0.23709369024856597, |
| "grad_norm": 0.4317997282790841, |
| "learning_rate": 0.00019979310826822046, |
| "loss": 0.5727, |
| "step": 682 |
| }, |
| { |
| "epoch": 0.23778897966278464, |
| "grad_norm": 0.7933828103238656, |
| "learning_rate": 0.00019978523020204857, |
| "loss": 0.6703, |
| "step": 684 |
| }, |
| { |
| "epoch": 0.2384842690770033, |
| "grad_norm": 0.5465102255190097, |
| "learning_rate": 0.00019977720509522747, |
| "loss": 0.6675, |
| "step": 686 |
| }, |
| { |
| "epoch": 0.23917955849122197, |
| "grad_norm": 0.8356137363612883, |
| "learning_rate": 0.00019976903295958269, |
| "loss": 0.6558, |
| "step": 688 |
| }, |
| { |
| "epoch": 0.23987484790544064, |
| "grad_norm": 0.5190073187676806, |
| "learning_rate": 0.00019976071380715645, |
| "loss": 0.6779, |
| "step": 690 |
| }, |
| { |
| "epoch": 0.24057013731965932, |
| "grad_norm": 0.40561527338007225, |
| "learning_rate": 0.0001997522476502076, |
| "loss": 0.6687, |
| "step": 692 |
| }, |
| { |
| "epoch": 0.24126542673387796, |
| "grad_norm": 0.592821589706298, |
| "learning_rate": 0.0001997436345012117, |
| "loss": 0.6807, |
| "step": 694 |
| }, |
| { |
| "epoch": 0.24196071614809664, |
| "grad_norm": 0.686918345062074, |
| "learning_rate": 0.00019973487437286073, |
| "loss": 0.6831, |
| "step": 696 |
| }, |
| { |
| "epoch": 0.2426560055623153, |
| "grad_norm": 0.964075508243471, |
| "learning_rate": 0.00019972596727806346, |
| "loss": 0.665, |
| "step": 698 |
| }, |
| { |
| "epoch": 0.243351294976534, |
| "grad_norm": 0.739137434656244, |
| "learning_rate": 0.00019971691322994507, |
| "loss": 0.6387, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.24404658439075266, |
| "grad_norm": 0.3978099419540589, |
| "learning_rate": 0.00019970771224184737, |
| "loss": 0.6143, |
| "step": 702 |
| }, |
| { |
| "epoch": 0.2447418738049713, |
| "grad_norm": 0.4997978607153984, |
| "learning_rate": 0.0001996983643273287, |
| "loss": 0.6083, |
| "step": 704 |
| }, |
| { |
| "epoch": 0.24543716321918999, |
| "grad_norm": 1.3443518380439796, |
| "learning_rate": 0.0001996888695001638, |
| "loss": 0.6711, |
| "step": 706 |
| }, |
| { |
| "epoch": 0.24613245263340866, |
| "grad_norm": 0.5424379336505594, |
| "learning_rate": 0.000199679227774344, |
| "loss": 0.6321, |
| "step": 708 |
| }, |
| { |
| "epoch": 0.24682774204762734, |
| "grad_norm": 0.7689235498835658, |
| "learning_rate": 0.00019966943916407712, |
| "loss": 0.6918, |
| "step": 710 |
| }, |
| { |
| "epoch": 0.24752303146184598, |
| "grad_norm": 0.6467240394468301, |
| "learning_rate": 0.00019965950368378734, |
| "loss": 0.6368, |
| "step": 712 |
| }, |
| { |
| "epoch": 0.24821832087606466, |
| "grad_norm": 0.48819289451999526, |
| "learning_rate": 0.00019964942134811532, |
| "loss": 0.5875, |
| "step": 714 |
| }, |
| { |
| "epoch": 0.24891361029028333, |
| "grad_norm": 1.1929427463467126, |
| "learning_rate": 0.00019963919217191807, |
| "loss": 0.6795, |
| "step": 716 |
| }, |
| { |
| "epoch": 0.249608899704502, |
| "grad_norm": 0.4182713825442439, |
| "learning_rate": 0.00019962881617026902, |
| "loss": 0.5835, |
| "step": 718 |
| }, |
| { |
| "epoch": 0.25030418911872065, |
| "grad_norm": 0.5770262516296942, |
| "learning_rate": 0.00019961829335845795, |
| "loss": 0.6299, |
| "step": 720 |
| }, |
| { |
| "epoch": 0.25099947853293936, |
| "grad_norm": 0.9135250550485389, |
| "learning_rate": 0.00019960762375199095, |
| "loss": 0.6844, |
| "step": 722 |
| }, |
| { |
| "epoch": 0.251694767947158, |
| "grad_norm": 0.5511234805517417, |
| "learning_rate": 0.0001995968073665905, |
| "loss": 0.5745, |
| "step": 724 |
| }, |
| { |
| "epoch": 0.25239005736137665, |
| "grad_norm": 0.49137872896877094, |
| "learning_rate": 0.00019958584421819528, |
| "loss": 0.698, |
| "step": 726 |
| }, |
| { |
| "epoch": 0.25308534677559535, |
| "grad_norm": 0.4437253322782162, |
| "learning_rate": 0.00019957473432296026, |
| "loss": 0.6864, |
| "step": 728 |
| }, |
| { |
| "epoch": 0.253780636189814, |
| "grad_norm": 0.5064381489074585, |
| "learning_rate": 0.0001995634776972567, |
| "loss": 0.5758, |
| "step": 730 |
| }, |
| { |
| "epoch": 0.2544759256040327, |
| "grad_norm": 0.5464655664741616, |
| "learning_rate": 0.00019955207435767201, |
| "loss": 0.6351, |
| "step": 732 |
| }, |
| { |
| "epoch": 0.25517121501825135, |
| "grad_norm": 0.44060524677853835, |
| "learning_rate": 0.00019954052432100982, |
| "loss": 0.6837, |
| "step": 734 |
| }, |
| { |
| "epoch": 0.25586650443247, |
| "grad_norm": 0.942390696980061, |
| "learning_rate": 0.00019952882760428998, |
| "loss": 0.6378, |
| "step": 736 |
| }, |
| { |
| "epoch": 0.2565617938466887, |
| "grad_norm": 0.4710023299938762, |
| "learning_rate": 0.00019951698422474836, |
| "loss": 0.6968, |
| "step": 738 |
| }, |
| { |
| "epoch": 0.25725708326090735, |
| "grad_norm": 0.4179809222387651, |
| "learning_rate": 0.00019950499419983707, |
| "loss": 0.5643, |
| "step": 740 |
| }, |
| { |
| "epoch": 0.257952372675126, |
| "grad_norm": 0.4446278865199289, |
| "learning_rate": 0.00019949285754722426, |
| "loss": 0.5089, |
| "step": 742 |
| }, |
| { |
| "epoch": 0.2586476620893447, |
| "grad_norm": 0.914728851966275, |
| "learning_rate": 0.00019948057428479418, |
| "loss": 0.7933, |
| "step": 744 |
| }, |
| { |
| "epoch": 0.25934295150356335, |
| "grad_norm": 0.47142275185055055, |
| "learning_rate": 0.00019946814443064703, |
| "loss": 0.6384, |
| "step": 746 |
| }, |
| { |
| "epoch": 0.26003824091778205, |
| "grad_norm": 0.720916103725334, |
| "learning_rate": 0.00019945556800309917, |
| "loss": 0.5957, |
| "step": 748 |
| }, |
| { |
| "epoch": 0.2607335303320007, |
| "grad_norm": 0.8550360126036364, |
| "learning_rate": 0.00019944284502068275, |
| "loss": 0.6454, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.2607335303320007, |
| "eval_loss": 0.6373963952064514, |
| "eval_runtime": 728.7951, |
| "eval_samples_per_second": 6.644, |
| "eval_steps_per_second": 0.209, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.26142881974621934, |
| "grad_norm": 0.5467512313200382, |
| "learning_rate": 0.0001994299755021461, |
| "loss": 0.5781, |
| "step": 752 |
| }, |
| { |
| "epoch": 0.26212410916043805, |
| "grad_norm": 0.41230384162224254, |
| "learning_rate": 0.0001994169594664533, |
| "loss": 0.6467, |
| "step": 754 |
| }, |
| { |
| "epoch": 0.2628193985746567, |
| "grad_norm": 0.5635895964904281, |
| "learning_rate": 0.00019940379693278448, |
| "loss": 0.6182, |
| "step": 756 |
| }, |
| { |
| "epoch": 0.2635146879888754, |
| "grad_norm": 0.6456168984233462, |
| "learning_rate": 0.0001993904879205355, |
| "loss": 0.6086, |
| "step": 758 |
| }, |
| { |
| "epoch": 0.26420997740309404, |
| "grad_norm": 0.5060399335833471, |
| "learning_rate": 0.00019937703244931815, |
| "loss": 0.6407, |
| "step": 760 |
| }, |
| { |
| "epoch": 0.2649052668173127, |
| "grad_norm": 0.5033297464646784, |
| "learning_rate": 0.00019936343053896004, |
| "loss": 0.6157, |
| "step": 762 |
| }, |
| { |
| "epoch": 0.2656005562315314, |
| "grad_norm": 0.3911665523252255, |
| "learning_rate": 0.00019934968220950458, |
| "loss": 0.6227, |
| "step": 764 |
| }, |
| { |
| "epoch": 0.26629584564575004, |
| "grad_norm": 1.7130937466256138, |
| "learning_rate": 0.00019933578748121086, |
| "loss": 0.5994, |
| "step": 766 |
| }, |
| { |
| "epoch": 0.2669911350599687, |
| "grad_norm": 0.5747338886603724, |
| "learning_rate": 0.00019932174637455382, |
| "loss": 0.6238, |
| "step": 768 |
| }, |
| { |
| "epoch": 0.2676864244741874, |
| "grad_norm": 0.6226324357252033, |
| "learning_rate": 0.00019930755891022398, |
| "loss": 0.6255, |
| "step": 770 |
| }, |
| { |
| "epoch": 0.26838171388840604, |
| "grad_norm": 0.5282488794229044, |
| "learning_rate": 0.00019929322510912756, |
| "loss": 0.5808, |
| "step": 772 |
| }, |
| { |
| "epoch": 0.26907700330262474, |
| "grad_norm": 0.5596904577167024, |
| "learning_rate": 0.0001992787449923865, |
| "loss": 0.5361, |
| "step": 774 |
| }, |
| { |
| "epoch": 0.2697722927168434, |
| "grad_norm": 1.0477422618379753, |
| "learning_rate": 0.00019926411858133824, |
| "loss": 0.54, |
| "step": 776 |
| }, |
| { |
| "epoch": 0.27046758213106203, |
| "grad_norm": 1.1314704495635823, |
| "learning_rate": 0.00019924934589753582, |
| "loss": 0.5869, |
| "step": 778 |
| }, |
| { |
| "epoch": 0.27116287154528074, |
| "grad_norm": 0.5426548544381972, |
| "learning_rate": 0.00019923442696274794, |
| "loss": 0.5951, |
| "step": 780 |
| }, |
| { |
| "epoch": 0.2718581609594994, |
| "grad_norm": 0.8480025293040686, |
| "learning_rate": 0.00019921936179895862, |
| "loss": 0.6003, |
| "step": 782 |
| }, |
| { |
| "epoch": 0.2725534503737181, |
| "grad_norm": 0.5729452380906337, |
| "learning_rate": 0.0001992041504283675, |
| "loss": 0.6526, |
| "step": 784 |
| }, |
| { |
| "epoch": 0.27324873978793673, |
| "grad_norm": 0.7362085286496177, |
| "learning_rate": 0.00019918879287338957, |
| "loss": 0.6776, |
| "step": 786 |
| }, |
| { |
| "epoch": 0.2739440292021554, |
| "grad_norm": 0.7440059372391256, |
| "learning_rate": 0.00019917328915665535, |
| "loss": 0.6486, |
| "step": 788 |
| }, |
| { |
| "epoch": 0.2746393186163741, |
| "grad_norm": 0.6110842206790659, |
| "learning_rate": 0.0001991576393010106, |
| "loss": 0.6226, |
| "step": 790 |
| }, |
| { |
| "epoch": 0.27533460803059273, |
| "grad_norm": 0.8002336898560896, |
| "learning_rate": 0.00019914184332951655, |
| "loss": 0.6215, |
| "step": 792 |
| }, |
| { |
| "epoch": 0.2760298974448114, |
| "grad_norm": 0.6761736356192558, |
| "learning_rate": 0.00019912590126544964, |
| "loss": 0.5988, |
| "step": 794 |
| }, |
| { |
| "epoch": 0.2767251868590301, |
| "grad_norm": 1.1005016630123619, |
| "learning_rate": 0.00019910981313230172, |
| "loss": 0.6213, |
| "step": 796 |
| }, |
| { |
| "epoch": 0.27742047627324873, |
| "grad_norm": 0.6513452605734942, |
| "learning_rate": 0.00019909357895377973, |
| "loss": 0.682, |
| "step": 798 |
| }, |
| { |
| "epoch": 0.27811576568746743, |
| "grad_norm": 0.8536492677008404, |
| "learning_rate": 0.0001990771987538059, |
| "loss": 0.707, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.2788110551016861, |
| "grad_norm": 0.3362293750538688, |
| "learning_rate": 0.00019906067255651765, |
| "loss": 0.7108, |
| "step": 802 |
| }, |
| { |
| "epoch": 0.2795063445159047, |
| "grad_norm": 0.4792899916449288, |
| "learning_rate": 0.0001990440003862675, |
| "loss": 0.6055, |
| "step": 804 |
| }, |
| { |
| "epoch": 0.28020163393012343, |
| "grad_norm": 1.272937677307704, |
| "learning_rate": 0.00019902718226762304, |
| "loss": 0.6816, |
| "step": 806 |
| }, |
| { |
| "epoch": 0.2808969233443421, |
| "grad_norm": 0.30646692495778655, |
| "learning_rate": 0.00019901021822536704, |
| "loss": 0.688, |
| "step": 808 |
| }, |
| { |
| "epoch": 0.2815922127585608, |
| "grad_norm": 0.3035733218665055, |
| "learning_rate": 0.00019899310828449713, |
| "loss": 0.6746, |
| "step": 810 |
| }, |
| { |
| "epoch": 0.2822875021727794, |
| "grad_norm": 0.2763276026826459, |
| "learning_rate": 0.00019897585247022613, |
| "loss": 0.6632, |
| "step": 812 |
| }, |
| { |
| "epoch": 0.2829827915869981, |
| "grad_norm": 1.173118337900773, |
| "learning_rate": 0.00019895845080798166, |
| "loss": 0.6306, |
| "step": 814 |
| }, |
| { |
| "epoch": 0.2836780810012168, |
| "grad_norm": 0.4335278252167635, |
| "learning_rate": 0.0001989409033234063, |
| "loss": 0.6147, |
| "step": 816 |
| }, |
| { |
| "epoch": 0.2843733704154354, |
| "grad_norm": 0.3556694503062785, |
| "learning_rate": 0.00019892321004235755, |
| "loss": 0.5771, |
| "step": 818 |
| }, |
| { |
| "epoch": 0.28506865982965407, |
| "grad_norm": 0.9083240381897224, |
| "learning_rate": 0.00019890537099090768, |
| "loss": 0.5729, |
| "step": 820 |
| }, |
| { |
| "epoch": 0.2857639492438728, |
| "grad_norm": 0.5433346815141633, |
| "learning_rate": 0.00019888738619534385, |
| "loss": 0.5554, |
| "step": 822 |
| }, |
| { |
| "epoch": 0.2864592386580914, |
| "grad_norm": 0.5244705156138804, |
| "learning_rate": 0.0001988692556821679, |
| "loss": 0.6525, |
| "step": 824 |
| }, |
| { |
| "epoch": 0.2871545280723101, |
| "grad_norm": 0.7580866792170871, |
| "learning_rate": 0.00019885097947809648, |
| "loss": 0.6512, |
| "step": 826 |
| }, |
| { |
| "epoch": 0.28784981748652877, |
| "grad_norm": 0.7034554538404351, |
| "learning_rate": 0.00019883255761006082, |
| "loss": 0.5414, |
| "step": 828 |
| }, |
| { |
| "epoch": 0.2885451069007474, |
| "grad_norm": 0.6915575597289163, |
| "learning_rate": 0.00019881399010520688, |
| "loss": 0.6036, |
| "step": 830 |
| }, |
| { |
| "epoch": 0.2892403963149661, |
| "grad_norm": 0.6895372001781882, |
| "learning_rate": 0.00019879527699089524, |
| "loss": 0.5894, |
| "step": 832 |
| }, |
| { |
| "epoch": 0.28993568572918477, |
| "grad_norm": 0.7762412863407715, |
| "learning_rate": 0.00019877641829470094, |
| "loss": 0.7115, |
| "step": 834 |
| }, |
| { |
| "epoch": 0.29063097514340347, |
| "grad_norm": 0.5761220663639801, |
| "learning_rate": 0.00019875741404441367, |
| "loss": 0.6108, |
| "step": 836 |
| }, |
| { |
| "epoch": 0.2913262645576221, |
| "grad_norm": 0.47176943357070505, |
| "learning_rate": 0.00019873826426803755, |
| "loss": 0.634, |
| "step": 838 |
| }, |
| { |
| "epoch": 0.29202155397184076, |
| "grad_norm": 0.5985873148196751, |
| "learning_rate": 0.00019871896899379107, |
| "loss": 0.6528, |
| "step": 840 |
| }, |
| { |
| "epoch": 0.29271684338605947, |
| "grad_norm": 0.39514741111190665, |
| "learning_rate": 0.00019869952825010727, |
| "loss": 0.6034, |
| "step": 842 |
| }, |
| { |
| "epoch": 0.2934121328002781, |
| "grad_norm": 0.29787585594263405, |
| "learning_rate": 0.00019867994206563343, |
| "loss": 0.6063, |
| "step": 844 |
| }, |
| { |
| "epoch": 0.29410742221449676, |
| "grad_norm": 0.303320759302155, |
| "learning_rate": 0.00019866021046923118, |
| "loss": 0.6343, |
| "step": 846 |
| }, |
| { |
| "epoch": 0.29480271162871546, |
| "grad_norm": 0.33135450527244925, |
| "learning_rate": 0.00019864033348997645, |
| "loss": 0.6421, |
| "step": 848 |
| }, |
| { |
| "epoch": 0.2954980010429341, |
| "grad_norm": 0.553668190192523, |
| "learning_rate": 0.0001986203111571594, |
| "loss": 0.6503, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.2961932904571528, |
| "grad_norm": 0.31948016623126946, |
| "learning_rate": 0.00019860014350028438, |
| "loss": 0.6259, |
| "step": 852 |
| }, |
| { |
| "epoch": 0.29688857987137146, |
| "grad_norm": 0.5325237443938606, |
| "learning_rate": 0.0001985798305490698, |
| "loss": 0.6207, |
| "step": 854 |
| }, |
| { |
| "epoch": 0.2975838692855901, |
| "grad_norm": 0.5093186589927414, |
| "learning_rate": 0.00019855937233344831, |
| "loss": 0.5397, |
| "step": 856 |
| }, |
| { |
| "epoch": 0.2982791586998088, |
| "grad_norm": 0.5220573948537062, |
| "learning_rate": 0.00019853876888356652, |
| "loss": 0.6237, |
| "step": 858 |
| }, |
| { |
| "epoch": 0.29897444811402746, |
| "grad_norm": 0.9861332892020862, |
| "learning_rate": 0.00019851802022978506, |
| "loss": 0.689, |
| "step": 860 |
| }, |
| { |
| "epoch": 0.2996697375282461, |
| "grad_norm": 0.464669721879274, |
| "learning_rate": 0.00019849712640267861, |
| "loss": 0.522, |
| "step": 862 |
| }, |
| { |
| "epoch": 0.3003650269424648, |
| "grad_norm": 0.9223435358714303, |
| "learning_rate": 0.00019847608743303567, |
| "loss": 0.7491, |
| "step": 864 |
| }, |
| { |
| "epoch": 0.30106031635668346, |
| "grad_norm": 0.5058775377593727, |
| "learning_rate": 0.00019845490335185866, |
| "loss": 0.562, |
| "step": 866 |
| }, |
| { |
| "epoch": 0.30175560577090216, |
| "grad_norm": 0.5517767811356443, |
| "learning_rate": 0.00019843357419036382, |
| "loss": 0.6162, |
| "step": 868 |
| }, |
| { |
| "epoch": 0.3024508951851208, |
| "grad_norm": 0.49255497197537723, |
| "learning_rate": 0.00019841209997998127, |
| "loss": 0.6803, |
| "step": 870 |
| }, |
| { |
| "epoch": 0.30314618459933945, |
| "grad_norm": 0.41553745561512617, |
| "learning_rate": 0.0001983904807523547, |
| "loss": 0.6415, |
| "step": 872 |
| }, |
| { |
| "epoch": 0.30384147401355815, |
| "grad_norm": 0.49551628457734653, |
| "learning_rate": 0.00019836871653934162, |
| "loss": 0.6176, |
| "step": 874 |
| }, |
| { |
| "epoch": 0.3045367634277768, |
| "grad_norm": 0.7489091107060393, |
| "learning_rate": 0.00019834680737301313, |
| "loss": 0.6337, |
| "step": 876 |
| }, |
| { |
| "epoch": 0.3052320528419955, |
| "grad_norm": 0.32312869533576805, |
| "learning_rate": 0.00019832475328565398, |
| "loss": 0.6135, |
| "step": 878 |
| }, |
| { |
| "epoch": 0.30592734225621415, |
| "grad_norm": 0.304002075332943, |
| "learning_rate": 0.00019830255430976242, |
| "loss": 0.5533, |
| "step": 880 |
| }, |
| { |
| "epoch": 0.3066226316704328, |
| "grad_norm": 0.4137621036041215, |
| "learning_rate": 0.00019828021047805022, |
| "loss": 0.573, |
| "step": 882 |
| }, |
| { |
| "epoch": 0.3073179210846515, |
| "grad_norm": 0.7506870255042438, |
| "learning_rate": 0.00019825772182344262, |
| "loss": 0.6971, |
| "step": 884 |
| }, |
| { |
| "epoch": 0.30801321049887015, |
| "grad_norm": 0.7069489041589112, |
| "learning_rate": 0.00019823508837907828, |
| "loss": 0.5848, |
| "step": 886 |
| }, |
| { |
| "epoch": 0.3087084999130888, |
| "grad_norm": 0.49383355752727304, |
| "learning_rate": 0.00019821231017830914, |
| "loss": 0.6349, |
| "step": 888 |
| }, |
| { |
| "epoch": 0.3094037893273075, |
| "grad_norm": 0.7893505446859834, |
| "learning_rate": 0.0001981893872547005, |
| "loss": 0.6335, |
| "step": 890 |
| }, |
| { |
| "epoch": 0.31009907874152615, |
| "grad_norm": 1.0962653113728835, |
| "learning_rate": 0.00019816631964203097, |
| "loss": 0.6438, |
| "step": 892 |
| }, |
| { |
| "epoch": 0.31079436815574485, |
| "grad_norm": 0.40606329821748216, |
| "learning_rate": 0.0001981431073742923, |
| "loss": 0.557, |
| "step": 894 |
| }, |
| { |
| "epoch": 0.3114896575699635, |
| "grad_norm": 0.8061837126172193, |
| "learning_rate": 0.00019811975048568943, |
| "loss": 0.6334, |
| "step": 896 |
| }, |
| { |
| "epoch": 0.31218494698418214, |
| "grad_norm": 0.7808955990860935, |
| "learning_rate": 0.00019809624901064038, |
| "loss": 0.5775, |
| "step": 898 |
| }, |
| { |
| "epoch": 0.31288023639840085, |
| "grad_norm": 0.5527203146534614, |
| "learning_rate": 0.00019807260298377626, |
| "loss": 0.5934, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.31288023639840085, |
| "eval_loss": 0.666339099407196, |
| "eval_runtime": 759.5196, |
| "eval_samples_per_second": 6.375, |
| "eval_steps_per_second": 0.2, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.3135755258126195, |
| "grad_norm": 1.151650071753606, |
| "learning_rate": 0.00019804881243994118, |
| "loss": 0.6459, |
| "step": 902 |
| }, |
| { |
| "epoch": 0.3142708152268382, |
| "grad_norm": 0.37537177441864283, |
| "learning_rate": 0.00019802487741419218, |
| "loss": 0.5537, |
| "step": 904 |
| }, |
| { |
| "epoch": 0.31496610464105684, |
| "grad_norm": 0.39806583735978385, |
| "learning_rate": 0.00019800079794179927, |
| "loss": 0.5765, |
| "step": 906 |
| }, |
| { |
| "epoch": 0.3156613940552755, |
| "grad_norm": 0.9252532303995283, |
| "learning_rate": 0.00019797657405824524, |
| "loss": 0.6581, |
| "step": 908 |
| }, |
| { |
| "epoch": 0.3163566834694942, |
| "grad_norm": 0.4242008643262632, |
| "learning_rate": 0.00019795220579922572, |
| "loss": 0.663, |
| "step": 910 |
| }, |
| { |
| "epoch": 0.31705197288371284, |
| "grad_norm": 0.5557863138791925, |
| "learning_rate": 0.00019792769320064904, |
| "loss": 0.6492, |
| "step": 912 |
| }, |
| { |
| "epoch": 0.3177472622979315, |
| "grad_norm": 0.5743017982975046, |
| "learning_rate": 0.0001979030362986363, |
| "loss": 0.6425, |
| "step": 914 |
| }, |
| { |
| "epoch": 0.3184425517121502, |
| "grad_norm": 0.39667228882787314, |
| "learning_rate": 0.0001978782351295212, |
| "loss": 0.5658, |
| "step": 916 |
| }, |
| { |
| "epoch": 0.31913784112636884, |
| "grad_norm": 1.2742981139875873, |
| "learning_rate": 0.00019785328972985, |
| "loss": 0.6042, |
| "step": 918 |
| }, |
| { |
| "epoch": 0.31983313054058754, |
| "grad_norm": 0.7520790754771111, |
| "learning_rate": 0.00019782820013638158, |
| "loss": 0.6248, |
| "step": 920 |
| }, |
| { |
| "epoch": 0.3205284199548062, |
| "grad_norm": 1.1777266516894538, |
| "learning_rate": 0.0001978029663860872, |
| "loss": 0.6394, |
| "step": 922 |
| }, |
| { |
| "epoch": 0.32122370936902483, |
| "grad_norm": 0.5383416828808074, |
| "learning_rate": 0.00019777758851615058, |
| "loss": 0.6357, |
| "step": 924 |
| }, |
| { |
| "epoch": 0.32191899878324354, |
| "grad_norm": 0.5351088818608489, |
| "learning_rate": 0.00019775206656396787, |
| "loss": 0.6111, |
| "step": 926 |
| }, |
| { |
| "epoch": 0.3226142881974622, |
| "grad_norm": 0.7776255734128178, |
| "learning_rate": 0.00019772640056714744, |
| "loss": 0.5778, |
| "step": 928 |
| }, |
| { |
| "epoch": 0.3233095776116809, |
| "grad_norm": 0.5049904332607067, |
| "learning_rate": 0.00019770059056351, |
| "loss": 0.5978, |
| "step": 930 |
| }, |
| { |
| "epoch": 0.32400486702589953, |
| "grad_norm": 0.6894813643690206, |
| "learning_rate": 0.00019767463659108841, |
| "loss": 0.6727, |
| "step": 932 |
| }, |
| { |
| "epoch": 0.3247001564401182, |
| "grad_norm": 0.6230252249989028, |
| "learning_rate": 0.00019764853868812772, |
| "loss": 0.5911, |
| "step": 934 |
| }, |
| { |
| "epoch": 0.3253954458543369, |
| "grad_norm": 0.6699617199619087, |
| "learning_rate": 0.00019762229689308499, |
| "loss": 0.6694, |
| "step": 936 |
| }, |
| { |
| "epoch": 0.32609073526855553, |
| "grad_norm": 0.9762605521595761, |
| "learning_rate": 0.00019759591124462943, |
| "loss": 0.7053, |
| "step": 938 |
| }, |
| { |
| "epoch": 0.3267860246827742, |
| "grad_norm": 0.5216728233794251, |
| "learning_rate": 0.0001975693817816422, |
| "loss": 0.6958, |
| "step": 940 |
| }, |
| { |
| "epoch": 0.3274813140969929, |
| "grad_norm": 0.5943791708445256, |
| "learning_rate": 0.00019754270854321625, |
| "loss": 0.6342, |
| "step": 942 |
| }, |
| { |
| "epoch": 0.32817660351121153, |
| "grad_norm": 0.5341014737913188, |
| "learning_rate": 0.00019751589156865663, |
| "loss": 0.6272, |
| "step": 944 |
| }, |
| { |
| "epoch": 0.32887189292543023, |
| "grad_norm": 0.8411647140863245, |
| "learning_rate": 0.00019748893089747995, |
| "loss": 0.6041, |
| "step": 946 |
| }, |
| { |
| "epoch": 0.3295671823396489, |
| "grad_norm": 1.072323043427063, |
| "learning_rate": 0.00019746182656941473, |
| "loss": 0.7152, |
| "step": 948 |
| }, |
| { |
| "epoch": 0.3302624717538675, |
| "grad_norm": 0.6497829380326366, |
| "learning_rate": 0.00019743457862440115, |
| "loss": 0.6176, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.33095776116808623, |
| "grad_norm": 0.28736093186011447, |
| "learning_rate": 0.00019740718710259096, |
| "loss": 0.6453, |
| "step": 952 |
| }, |
| { |
| "epoch": 0.3316530505823049, |
| "grad_norm": 0.27868233108109625, |
| "learning_rate": 0.00019737965204434757, |
| "loss": 0.6051, |
| "step": 954 |
| }, |
| { |
| "epoch": 0.3323483399965236, |
| "grad_norm": 0.40709235855818693, |
| "learning_rate": 0.00019735197349024576, |
| "loss": 0.6255, |
| "step": 956 |
| }, |
| { |
| "epoch": 0.3330436294107422, |
| "grad_norm": 0.8385677925045294, |
| "learning_rate": 0.00019732415148107199, |
| "loss": 0.6455, |
| "step": 958 |
| }, |
| { |
| "epoch": 0.3337389188249609, |
| "grad_norm": 0.5642576200414804, |
| "learning_rate": 0.00019729618605782384, |
| "loss": 0.6971, |
| "step": 960 |
| }, |
| { |
| "epoch": 0.3344342082391796, |
| "grad_norm": 0.7034648545079693, |
| "learning_rate": 0.00019726807726171039, |
| "loss": 0.6177, |
| "step": 962 |
| }, |
| { |
| "epoch": 0.3351294976533982, |
| "grad_norm": 1.9840633930320113, |
| "learning_rate": 0.000197239825134152, |
| "loss": 0.6776, |
| "step": 964 |
| }, |
| { |
| "epoch": 0.33582478706761687, |
| "grad_norm": 1.0091982574836484, |
| "learning_rate": 0.00019721142971678015, |
| "loss": 0.6893, |
| "step": 966 |
| }, |
| { |
| "epoch": 0.3365200764818356, |
| "grad_norm": 0.9742560258590767, |
| "learning_rate": 0.00019718289105143753, |
| "loss": 0.744, |
| "step": 968 |
| }, |
| { |
| "epoch": 0.3372153658960542, |
| "grad_norm": 0.6897018399345455, |
| "learning_rate": 0.00019715420918017793, |
| "loss": 0.678, |
| "step": 970 |
| }, |
| { |
| "epoch": 0.3379106553102729, |
| "grad_norm": 0.29102959771453246, |
| "learning_rate": 0.00019712538414526606, |
| "loss": 0.6663, |
| "step": 972 |
| }, |
| { |
| "epoch": 0.33860594472449157, |
| "grad_norm": 0.7337107483377766, |
| "learning_rate": 0.0001970964159891777, |
| "loss": 0.663, |
| "step": 974 |
| }, |
| { |
| "epoch": 0.3393012341387102, |
| "grad_norm": 0.5817704647699353, |
| "learning_rate": 0.00019706730475459953, |
| "loss": 0.6398, |
| "step": 976 |
| }, |
| { |
| "epoch": 0.3399965235529289, |
| "grad_norm": 0.28703428796704483, |
| "learning_rate": 0.00019703805048442897, |
| "loss": 0.5906, |
| "step": 978 |
| }, |
| { |
| "epoch": 0.34069181296714757, |
| "grad_norm": 0.41383789019772477, |
| "learning_rate": 0.0001970086532217743, |
| "loss": 0.6709, |
| "step": 980 |
| }, |
| { |
| "epoch": 0.34138710238136627, |
| "grad_norm": 0.812487649001141, |
| "learning_rate": 0.00019697911300995443, |
| "loss": 0.6191, |
| "step": 982 |
| }, |
| { |
| "epoch": 0.3420823917955849, |
| "grad_norm": 0.798027200072012, |
| "learning_rate": 0.00019694942989249907, |
| "loss": 0.6608, |
| "step": 984 |
| }, |
| { |
| "epoch": 0.34277768120980356, |
| "grad_norm": 0.44029385955900757, |
| "learning_rate": 0.00019691960391314837, |
| "loss": 0.647, |
| "step": 986 |
| }, |
| { |
| "epoch": 0.34347297062402227, |
| "grad_norm": 0.3824484030698272, |
| "learning_rate": 0.00019688963511585295, |
| "loss": 0.6378, |
| "step": 988 |
| }, |
| { |
| "epoch": 0.3441682600382409, |
| "grad_norm": 0.4121768227084979, |
| "learning_rate": 0.0001968595235447741, |
| "loss": 0.5908, |
| "step": 990 |
| }, |
| { |
| "epoch": 0.34486354945245956, |
| "grad_norm": 0.5154693781246049, |
| "learning_rate": 0.0001968292692442833, |
| "loss": 0.5632, |
| "step": 992 |
| }, |
| { |
| "epoch": 0.34555883886667826, |
| "grad_norm": 0.3895510966829308, |
| "learning_rate": 0.0001967988722589624, |
| "loss": 0.4674, |
| "step": 994 |
| }, |
| { |
| "epoch": 0.3462541282808969, |
| "grad_norm": 0.507531875733667, |
| "learning_rate": 0.00019676833263360352, |
| "loss": 0.5581, |
| "step": 996 |
| }, |
| { |
| "epoch": 0.3469494176951156, |
| "grad_norm": 0.5476523355263471, |
| "learning_rate": 0.00019673765041320907, |
| "loss": 0.7421, |
| "step": 998 |
| }, |
| { |
| "epoch": 0.34764470710933426, |
| "grad_norm": 0.8417172933340035, |
| "learning_rate": 0.00019670682564299136, |
| "loss": 0.6774, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.3483399965235529, |
| "grad_norm": 0.6717112677412562, |
| "learning_rate": 0.00019667585836837299, |
| "loss": 0.6515, |
| "step": 1002 |
| }, |
| { |
| "epoch": 0.3490352859377716, |
| "grad_norm": 0.7599904388695796, |
| "learning_rate": 0.0001966447486349864, |
| "loss": 0.5679, |
| "step": 1004 |
| }, |
| { |
| "epoch": 0.34973057535199026, |
| "grad_norm": 0.44186748583335306, |
| "learning_rate": 0.000196613496488674, |
| "loss": 0.6067, |
| "step": 1006 |
| }, |
| { |
| "epoch": 0.35042586476620896, |
| "grad_norm": 0.4287165077907837, |
| "learning_rate": 0.00019658210197548805, |
| "loss": 0.5706, |
| "step": 1008 |
| }, |
| { |
| "epoch": 0.3511211541804276, |
| "grad_norm": 0.6051489125727973, |
| "learning_rate": 0.0001965505651416906, |
| "loss": 0.6178, |
| "step": 1010 |
| }, |
| { |
| "epoch": 0.35181644359464626, |
| "grad_norm": 0.5003034918118222, |
| "learning_rate": 0.00019651888603375346, |
| "loss": 0.6, |
| "step": 1012 |
| }, |
| { |
| "epoch": 0.35251173300886496, |
| "grad_norm": 0.6774513066433009, |
| "learning_rate": 0.00019648706469835804, |
| "loss": 0.6248, |
| "step": 1014 |
| }, |
| { |
| "epoch": 0.3532070224230836, |
| "grad_norm": 0.5666799893616385, |
| "learning_rate": 0.0001964551011823953, |
| "loss": 0.6256, |
| "step": 1016 |
| }, |
| { |
| "epoch": 0.35390231183730225, |
| "grad_norm": 0.6185519925235329, |
| "learning_rate": 0.00019642299553296582, |
| "loss": 0.5836, |
| "step": 1018 |
| }, |
| { |
| "epoch": 0.35459760125152096, |
| "grad_norm": 0.5020407783730059, |
| "learning_rate": 0.0001963907477973795, |
| "loss": 0.5148, |
| "step": 1020 |
| }, |
| { |
| "epoch": 0.3552928906657396, |
| "grad_norm": 0.966855697270511, |
| "learning_rate": 0.00019635835802315574, |
| "loss": 0.6335, |
| "step": 1022 |
| }, |
| { |
| "epoch": 0.3559881800799583, |
| "grad_norm": 0.8705091175729548, |
| "learning_rate": 0.00019632582625802317, |
| "loss": 0.6313, |
| "step": 1024 |
| }, |
| { |
| "epoch": 0.35668346949417695, |
| "grad_norm": 0.6028556619755229, |
| "learning_rate": 0.00019629315254991964, |
| "loss": 0.6483, |
| "step": 1026 |
| }, |
| { |
| "epoch": 0.3573787589083956, |
| "grad_norm": 0.6075038119620636, |
| "learning_rate": 0.00019626033694699214, |
| "loss": 0.6271, |
| "step": 1028 |
| }, |
| { |
| "epoch": 0.3580740483226143, |
| "grad_norm": 0.7923956541669288, |
| "learning_rate": 0.00019622737949759694, |
| "loss": 0.6338, |
| "step": 1030 |
| }, |
| { |
| "epoch": 0.35876933773683295, |
| "grad_norm": 1.2067992138100796, |
| "learning_rate": 0.00019619428025029905, |
| "loss": 0.6308, |
| "step": 1032 |
| }, |
| { |
| "epoch": 0.35946462715105165, |
| "grad_norm": 0.5446351671940789, |
| "learning_rate": 0.00019616103925387265, |
| "loss": 0.6475, |
| "step": 1034 |
| }, |
| { |
| "epoch": 0.3601599165652703, |
| "grad_norm": 0.8842474031361561, |
| "learning_rate": 0.0001961276565573007, |
| "loss": 0.6654, |
| "step": 1036 |
| }, |
| { |
| "epoch": 0.36085520597948895, |
| "grad_norm": 1.294693850012853, |
| "learning_rate": 0.00019609413220977496, |
| "loss": 0.6929, |
| "step": 1038 |
| }, |
| { |
| "epoch": 0.36155049539370765, |
| "grad_norm": 0.7435682846586636, |
| "learning_rate": 0.00019606046626069595, |
| "loss": 0.6791, |
| "step": 1040 |
| }, |
| { |
| "epoch": 0.3622457848079263, |
| "grad_norm": 0.45762946459115417, |
| "learning_rate": 0.0001960266587596729, |
| "loss": 0.5769, |
| "step": 1042 |
| }, |
| { |
| "epoch": 0.36294107422214494, |
| "grad_norm": 0.5614638042598611, |
| "learning_rate": 0.00019599270975652352, |
| "loss": 0.6047, |
| "step": 1044 |
| }, |
| { |
| "epoch": 0.36363636363636365, |
| "grad_norm": 1.6171161059961894, |
| "learning_rate": 0.0001959586193012741, |
| "loss": 0.6932, |
| "step": 1046 |
| }, |
| { |
| "epoch": 0.3643316530505823, |
| "grad_norm": 1.5847051714441287, |
| "learning_rate": 0.00019592438744415932, |
| "loss": 0.5908, |
| "step": 1048 |
| }, |
| { |
| "epoch": 0.365026942464801, |
| "grad_norm": 0.4282438415059217, |
| "learning_rate": 0.00019589001423562233, |
| "loss": 0.6749, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.365026942464801, |
| "eval_loss": 0.6080955266952515, |
| "eval_runtime": 710.9824, |
| "eval_samples_per_second": 6.81, |
| "eval_steps_per_second": 0.214, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.36572223187901964, |
| "grad_norm": 1.0612232842206784, |
| "learning_rate": 0.00019585549972631446, |
| "loss": 0.5669, |
| "step": 1052 |
| }, |
| { |
| "epoch": 0.3664175212932383, |
| "grad_norm": 2.6994673924740358, |
| "learning_rate": 0.0001958208439670953, |
| "loss": 0.6455, |
| "step": 1054 |
| }, |
| { |
| "epoch": 0.367112810707457, |
| "grad_norm": 2.022628249772274, |
| "learning_rate": 0.0001957860470090326, |
| "loss": 0.6395, |
| "step": 1056 |
| }, |
| { |
| "epoch": 0.36780810012167564, |
| "grad_norm": 0.607126211578616, |
| "learning_rate": 0.00019575110890340214, |
| "loss": 0.593, |
| "step": 1058 |
| }, |
| { |
| "epoch": 0.36850338953589434, |
| "grad_norm": 1.026539890410463, |
| "learning_rate": 0.00019571602970168775, |
| "loss": 0.5939, |
| "step": 1060 |
| }, |
| { |
| "epoch": 0.369198678950113, |
| "grad_norm": 0.6663599873173821, |
| "learning_rate": 0.00019568080945558104, |
| "loss": 0.6415, |
| "step": 1062 |
| }, |
| { |
| "epoch": 0.36989396836433164, |
| "grad_norm": 0.5967439670789174, |
| "learning_rate": 0.00019564544821698167, |
| "loss": 0.6348, |
| "step": 1064 |
| }, |
| { |
| "epoch": 0.37058925777855034, |
| "grad_norm": 0.6468802992284401, |
| "learning_rate": 0.00019560994603799682, |
| "loss": 0.5991, |
| "step": 1066 |
| }, |
| { |
| "epoch": 0.371284547192769, |
| "grad_norm": 1.2251498168873143, |
| "learning_rate": 0.00019557430297094158, |
| "loss": 0.6001, |
| "step": 1068 |
| }, |
| { |
| "epoch": 0.37197983660698763, |
| "grad_norm": 0.4681978143920913, |
| "learning_rate": 0.00019553851906833853, |
| "loss": 0.5664, |
| "step": 1070 |
| }, |
| { |
| "epoch": 0.37267512602120634, |
| "grad_norm": 0.7538053079694034, |
| "learning_rate": 0.00019550259438291782, |
| "loss": 0.5531, |
| "step": 1072 |
| }, |
| { |
| "epoch": 0.373370415435425, |
| "grad_norm": 0.4626868039226879, |
| "learning_rate": 0.00019546652896761696, |
| "loss": 0.6268, |
| "step": 1074 |
| }, |
| { |
| "epoch": 0.3740657048496437, |
| "grad_norm": 0.8012577314135656, |
| "learning_rate": 0.00019543032287558097, |
| "loss": 0.7819, |
| "step": 1076 |
| }, |
| { |
| "epoch": 0.37476099426386233, |
| "grad_norm": 0.7004564506452116, |
| "learning_rate": 0.0001953939761601621, |
| "loss": 0.6505, |
| "step": 1078 |
| }, |
| { |
| "epoch": 0.375456283678081, |
| "grad_norm": 1.352602706017517, |
| "learning_rate": 0.0001953574888749198, |
| "loss": 0.558, |
| "step": 1080 |
| }, |
| { |
| "epoch": 0.3761515730922997, |
| "grad_norm": 0.3012261239096098, |
| "learning_rate": 0.0001953208610736207, |
| "loss": 0.5746, |
| "step": 1082 |
| }, |
| { |
| "epoch": 0.37684686250651833, |
| "grad_norm": 0.49798092264749827, |
| "learning_rate": 0.0001952840928102385, |
| "loss": 0.5845, |
| "step": 1084 |
| }, |
| { |
| "epoch": 0.377542151920737, |
| "grad_norm": 0.4457559765569793, |
| "learning_rate": 0.00019524718413895382, |
| "loss": 0.5859, |
| "step": 1086 |
| }, |
| { |
| "epoch": 0.3782374413349557, |
| "grad_norm": 0.5943406854432696, |
| "learning_rate": 0.00019521013511415426, |
| "loss": 0.6581, |
| "step": 1088 |
| }, |
| { |
| "epoch": 0.37893273074917433, |
| "grad_norm": 1.8614189583072707, |
| "learning_rate": 0.00019517294579043414, |
| "loss": 0.7019, |
| "step": 1090 |
| }, |
| { |
| "epoch": 0.37962802016339303, |
| "grad_norm": 0.4212993231373999, |
| "learning_rate": 0.00019513561622259466, |
| "loss": 0.6106, |
| "step": 1092 |
| }, |
| { |
| "epoch": 0.3803233095776117, |
| "grad_norm": 0.5224177310348964, |
| "learning_rate": 0.00019509814646564354, |
| "loss": 0.6012, |
| "step": 1094 |
| }, |
| { |
| "epoch": 0.3810185989918303, |
| "grad_norm": 1.3980998842985117, |
| "learning_rate": 0.0001950605365747951, |
| "loss": 0.5875, |
| "step": 1096 |
| }, |
| { |
| "epoch": 0.38171388840604903, |
| "grad_norm": 0.7255638165052757, |
| "learning_rate": 0.0001950227866054703, |
| "loss": 0.6366, |
| "step": 1098 |
| }, |
| { |
| "epoch": 0.3824091778202677, |
| "grad_norm": 0.39401204779232035, |
| "learning_rate": 0.00019498489661329632, |
| "loss": 0.59, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.3831044672344864, |
| "grad_norm": 0.6249434294547875, |
| "learning_rate": 0.00019494686665410684, |
| "loss": 0.5555, |
| "step": 1102 |
| }, |
| { |
| "epoch": 0.383799756648705, |
| "grad_norm": 0.4970609140814689, |
| "learning_rate": 0.00019490869678394165, |
| "loss": 0.5872, |
| "step": 1104 |
| }, |
| { |
| "epoch": 0.3844950460629237, |
| "grad_norm": 0.5522933346523479, |
| "learning_rate": 0.0001948703870590468, |
| "loss": 0.6296, |
| "step": 1106 |
| }, |
| { |
| "epoch": 0.3851903354771424, |
| "grad_norm": 0.5358444643818808, |
| "learning_rate": 0.0001948319375358744, |
| "loss": 0.5677, |
| "step": 1108 |
| }, |
| { |
| "epoch": 0.385885624891361, |
| "grad_norm": 0.5839030994818908, |
| "learning_rate": 0.00019479334827108256, |
| "loss": 0.5443, |
| "step": 1110 |
| }, |
| { |
| "epoch": 0.38658091430557967, |
| "grad_norm": 0.9808370348899028, |
| "learning_rate": 0.00019475461932153533, |
| "loss": 0.6703, |
| "step": 1112 |
| }, |
| { |
| "epoch": 0.3872762037197984, |
| "grad_norm": 0.5567454433782846, |
| "learning_rate": 0.00019471575074430256, |
| "loss": 0.585, |
| "step": 1114 |
| }, |
| { |
| "epoch": 0.387971493134017, |
| "grad_norm": 0.8706120804701404, |
| "learning_rate": 0.00019467674259665985, |
| "loss": 0.5399, |
| "step": 1116 |
| }, |
| { |
| "epoch": 0.3886667825482357, |
| "grad_norm": 0.607175198823171, |
| "learning_rate": 0.00019463759493608854, |
| "loss": 0.6016, |
| "step": 1118 |
| }, |
| { |
| "epoch": 0.38936207196245437, |
| "grad_norm": 0.848040345554999, |
| "learning_rate": 0.0001945983078202754, |
| "loss": 0.5604, |
| "step": 1120 |
| }, |
| { |
| "epoch": 0.390057361376673, |
| "grad_norm": 0.39050505941263947, |
| "learning_rate": 0.00019455888130711296, |
| "loss": 0.5391, |
| "step": 1122 |
| }, |
| { |
| "epoch": 0.3907526507908917, |
| "grad_norm": 0.43688262723639604, |
| "learning_rate": 0.00019451931545469883, |
| "loss": 0.5911, |
| "step": 1124 |
| }, |
| { |
| "epoch": 0.39144794020511037, |
| "grad_norm": 0.5216486488482895, |
| "learning_rate": 0.00019447961032133623, |
| "loss": 0.7729, |
| "step": 1126 |
| }, |
| { |
| "epoch": 0.39214322961932907, |
| "grad_norm": 0.5238744592921366, |
| "learning_rate": 0.0001944397659655334, |
| "loss": 0.6561, |
| "step": 1128 |
| }, |
| { |
| "epoch": 0.3928385190335477, |
| "grad_norm": 0.5356265007770037, |
| "learning_rate": 0.00019439978244600392, |
| "loss": 0.6139, |
| "step": 1130 |
| }, |
| { |
| "epoch": 0.39353380844776636, |
| "grad_norm": 0.7845961219955971, |
| "learning_rate": 0.00019435965982166634, |
| "loss": 0.5682, |
| "step": 1132 |
| }, |
| { |
| "epoch": 0.39422909786198507, |
| "grad_norm": 0.5467699155830194, |
| "learning_rate": 0.0001943193981516442, |
| "loss": 0.5801, |
| "step": 1134 |
| }, |
| { |
| "epoch": 0.3949243872762037, |
| "grad_norm": 0.8265539141839208, |
| "learning_rate": 0.00019427899749526592, |
| "loss": 0.6112, |
| "step": 1136 |
| }, |
| { |
| "epoch": 0.39561967669042236, |
| "grad_norm": 0.6952363088057728, |
| "learning_rate": 0.0001942384579120648, |
| "loss": 0.5585, |
| "step": 1138 |
| }, |
| { |
| "epoch": 0.39631496610464106, |
| "grad_norm": 0.5390583855264917, |
| "learning_rate": 0.00019419777946177872, |
| "loss": 0.5437, |
| "step": 1140 |
| }, |
| { |
| "epoch": 0.3970102555188597, |
| "grad_norm": 0.7703174634754777, |
| "learning_rate": 0.0001941569622043504, |
| "loss": 0.6353, |
| "step": 1142 |
| }, |
| { |
| "epoch": 0.3977055449330784, |
| "grad_norm": 0.6599770887173079, |
| "learning_rate": 0.0001941160061999268, |
| "loss": 0.6984, |
| "step": 1144 |
| }, |
| { |
| "epoch": 0.39840083434729706, |
| "grad_norm": 0.6388540477012332, |
| "learning_rate": 0.0001940749115088597, |
| "loss": 0.6716, |
| "step": 1146 |
| }, |
| { |
| "epoch": 0.3990961237615157, |
| "grad_norm": 0.556740144667956, |
| "learning_rate": 0.00019403367819170495, |
| "loss": 0.5403, |
| "step": 1148 |
| }, |
| { |
| "epoch": 0.3997914131757344, |
| "grad_norm": 0.5811188003032377, |
| "learning_rate": 0.00019399230630922281, |
| "loss": 0.6038, |
| "step": 1150 |
| }, |
| { |
| "epoch": 0.40048670258995306, |
| "grad_norm": 0.5247246331826143, |
| "learning_rate": 0.00019395079592237767, |
| "loss": 0.5802, |
| "step": 1152 |
| }, |
| { |
| "epoch": 0.40118199200417176, |
| "grad_norm": 0.6218229787877086, |
| "learning_rate": 0.00019390914709233812, |
| "loss": 0.613, |
| "step": 1154 |
| }, |
| { |
| "epoch": 0.4018772814183904, |
| "grad_norm": 0.3768886648264559, |
| "learning_rate": 0.00019386735988047657, |
| "loss": 0.546, |
| "step": 1156 |
| }, |
| { |
| "epoch": 0.40257257083260906, |
| "grad_norm": 0.7192604288297088, |
| "learning_rate": 0.00019382543434836956, |
| "loss": 0.6414, |
| "step": 1158 |
| }, |
| { |
| "epoch": 0.40326786024682776, |
| "grad_norm": 0.6502767283266627, |
| "learning_rate": 0.00019378337055779725, |
| "loss": 0.6913, |
| "step": 1160 |
| }, |
| { |
| "epoch": 0.4039631496610464, |
| "grad_norm": 0.5821846511835889, |
| "learning_rate": 0.00019374116857074372, |
| "loss": 0.6135, |
| "step": 1162 |
| }, |
| { |
| "epoch": 0.40465843907526505, |
| "grad_norm": 0.4214614762979627, |
| "learning_rate": 0.00019369882844939656, |
| "loss": 0.5844, |
| "step": 1164 |
| }, |
| { |
| "epoch": 0.40535372848948376, |
| "grad_norm": 0.3995087581540831, |
| "learning_rate": 0.00019365635025614698, |
| "loss": 0.597, |
| "step": 1166 |
| }, |
| { |
| "epoch": 0.4060490179037024, |
| "grad_norm": 0.6448855951735888, |
| "learning_rate": 0.0001936137340535896, |
| "loss": 0.5964, |
| "step": 1168 |
| }, |
| { |
| "epoch": 0.4067443073179211, |
| "grad_norm": 0.4901354392109259, |
| "learning_rate": 0.00019357097990452244, |
| "loss": 0.6544, |
| "step": 1170 |
| }, |
| { |
| "epoch": 0.40743959673213975, |
| "grad_norm": 1.453302582941976, |
| "learning_rate": 0.0001935280878719468, |
| "loss": 0.7041, |
| "step": 1172 |
| }, |
| { |
| "epoch": 0.4081348861463584, |
| "grad_norm": 0.5029703758127726, |
| "learning_rate": 0.00019348505801906717, |
| "loss": 0.5963, |
| "step": 1174 |
| }, |
| { |
| "epoch": 0.4088301755605771, |
| "grad_norm": 0.42609969312576673, |
| "learning_rate": 0.00019344189040929104, |
| "loss": 0.5741, |
| "step": 1176 |
| }, |
| { |
| "epoch": 0.40952546497479575, |
| "grad_norm": 0.7354212284029352, |
| "learning_rate": 0.000193398585106229, |
| "loss": 0.5587, |
| "step": 1178 |
| }, |
| { |
| "epoch": 0.41022075438901445, |
| "grad_norm": 0.4221453871118396, |
| "learning_rate": 0.00019335514217369448, |
| "loss": 0.6008, |
| "step": 1180 |
| }, |
| { |
| "epoch": 0.4109160438032331, |
| "grad_norm": 0.4305199213179302, |
| "learning_rate": 0.00019331156167570377, |
| "loss": 0.6589, |
| "step": 1182 |
| }, |
| { |
| "epoch": 0.41161133321745175, |
| "grad_norm": 0.5908546860050797, |
| "learning_rate": 0.0001932678436764758, |
| "loss": 0.6791, |
| "step": 1184 |
| }, |
| { |
| "epoch": 0.41230662263167045, |
| "grad_norm": 0.592125562883329, |
| "learning_rate": 0.0001932239882404322, |
| "loss": 0.6163, |
| "step": 1186 |
| }, |
| { |
| "epoch": 0.4130019120458891, |
| "grad_norm": 0.3760868982177292, |
| "learning_rate": 0.00019317999543219707, |
| "loss": 0.6057, |
| "step": 1188 |
| }, |
| { |
| "epoch": 0.41369720146010774, |
| "grad_norm": 0.42271908071862413, |
| "learning_rate": 0.00019313586531659693, |
| "loss": 0.6055, |
| "step": 1190 |
| }, |
| { |
| "epoch": 0.41439249087432645, |
| "grad_norm": 0.7637973810401912, |
| "learning_rate": 0.00019309159795866067, |
| "loss": 0.6113, |
| "step": 1192 |
| }, |
| { |
| "epoch": 0.4150877802885451, |
| "grad_norm": 0.31348977991538335, |
| "learning_rate": 0.00019304719342361942, |
| "loss": 0.5906, |
| "step": 1194 |
| }, |
| { |
| "epoch": 0.4157830697027638, |
| "grad_norm": 0.34482048624358563, |
| "learning_rate": 0.00019300265177690635, |
| "loss": 0.5722, |
| "step": 1196 |
| }, |
| { |
| "epoch": 0.41647835911698244, |
| "grad_norm": 0.8513089082812038, |
| "learning_rate": 0.0001929579730841568, |
| "loss": 0.6293, |
| "step": 1198 |
| }, |
| { |
| "epoch": 0.4171736485312011, |
| "grad_norm": 0.4489163724605792, |
| "learning_rate": 0.00019291315741120802, |
| "loss": 0.5779, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.4171736485312011, |
| "eval_loss": 0.6036229133605957, |
| "eval_runtime": 710.4375, |
| "eval_samples_per_second": 6.816, |
| "eval_steps_per_second": 0.214, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.4178689379454198, |
| "grad_norm": 0.4187714520019529, |
| "learning_rate": 0.00019286820482409907, |
| "loss": 0.6429, |
| "step": 1202 |
| }, |
| { |
| "epoch": 0.41856422735963844, |
| "grad_norm": 0.6777550653965533, |
| "learning_rate": 0.0001928231153890708, |
| "loss": 0.7349, |
| "step": 1204 |
| }, |
| { |
| "epoch": 0.41925951677385714, |
| "grad_norm": 0.5320560870085244, |
| "learning_rate": 0.00019277788917256575, |
| "loss": 0.6171, |
| "step": 1206 |
| }, |
| { |
| "epoch": 0.4199548061880758, |
| "grad_norm": 0.4598764394317783, |
| "learning_rate": 0.0001927325262412279, |
| "loss": 0.568, |
| "step": 1208 |
| }, |
| { |
| "epoch": 0.42065009560229444, |
| "grad_norm": 0.6091029764488312, |
| "learning_rate": 0.0001926870266619028, |
| "loss": 0.6179, |
| "step": 1210 |
| }, |
| { |
| "epoch": 0.42134538501651314, |
| "grad_norm": 0.3831881756520618, |
| "learning_rate": 0.00019264139050163733, |
| "loss": 0.6277, |
| "step": 1212 |
| }, |
| { |
| "epoch": 0.4220406744307318, |
| "grad_norm": 0.46975774391187974, |
| "learning_rate": 0.00019259561782767964, |
| "loss": 0.6365, |
| "step": 1214 |
| }, |
| { |
| "epoch": 0.42273596384495044, |
| "grad_norm": 0.43958415828248065, |
| "learning_rate": 0.00019254970870747896, |
| "loss": 0.5599, |
| "step": 1216 |
| }, |
| { |
| "epoch": 0.42343125325916914, |
| "grad_norm": 0.3718315331231464, |
| "learning_rate": 0.00019250366320868573, |
| "loss": 0.5466, |
| "step": 1218 |
| }, |
| { |
| "epoch": 0.4241265426733878, |
| "grad_norm": 0.4132277343365207, |
| "learning_rate": 0.00019245748139915122, |
| "loss": 0.5505, |
| "step": 1220 |
| }, |
| { |
| "epoch": 0.4248218320876065, |
| "grad_norm": 0.699259138789713, |
| "learning_rate": 0.00019241116334692767, |
| "loss": 0.6155, |
| "step": 1222 |
| }, |
| { |
| "epoch": 0.42551712150182514, |
| "grad_norm": 0.8108211625089844, |
| "learning_rate": 0.00019236470912026795, |
| "loss": 0.6389, |
| "step": 1224 |
| }, |
| { |
| "epoch": 0.4262124109160438, |
| "grad_norm": 0.7991219996691611, |
| "learning_rate": 0.00019231811878762578, |
| "loss": 0.6218, |
| "step": 1226 |
| }, |
| { |
| "epoch": 0.4269077003302625, |
| "grad_norm": 0.6512869246829502, |
| "learning_rate": 0.00019227139241765527, |
| "loss": 0.6263, |
| "step": 1228 |
| }, |
| { |
| "epoch": 0.42760298974448113, |
| "grad_norm": 0.6616729861929296, |
| "learning_rate": 0.00019222453007921103, |
| "loss": 0.633, |
| "step": 1230 |
| }, |
| { |
| "epoch": 0.42829827915869984, |
| "grad_norm": 0.6441717114274472, |
| "learning_rate": 0.0001921775318413481, |
| "loss": 0.598, |
| "step": 1232 |
| }, |
| { |
| "epoch": 0.4289935685729185, |
| "grad_norm": 0.5398693786298329, |
| "learning_rate": 0.00019213039777332173, |
| "loss": 0.6128, |
| "step": 1234 |
| }, |
| { |
| "epoch": 0.42968885798713713, |
| "grad_norm": 0.7990572416423266, |
| "learning_rate": 0.00019208312794458734, |
| "loss": 0.6237, |
| "step": 1236 |
| }, |
| { |
| "epoch": 0.43038414740135583, |
| "grad_norm": 0.6792096783122197, |
| "learning_rate": 0.00019203572242480033, |
| "loss": 0.6167, |
| "step": 1238 |
| }, |
| { |
| "epoch": 0.4310794368155745, |
| "grad_norm": 0.569412739061848, |
| "learning_rate": 0.00019198818128381622, |
| "loss": 0.7147, |
| "step": 1240 |
| }, |
| { |
| "epoch": 0.4317747262297931, |
| "grad_norm": 0.3343242957299063, |
| "learning_rate": 0.00019194050459169016, |
| "loss": 0.5585, |
| "step": 1242 |
| }, |
| { |
| "epoch": 0.43247001564401183, |
| "grad_norm": 0.41389163470101814, |
| "learning_rate": 0.00019189269241867726, |
| "loss": 0.6206, |
| "step": 1244 |
| }, |
| { |
| "epoch": 0.4331653050582305, |
| "grad_norm": 0.6725681446750024, |
| "learning_rate": 0.00019184474483523208, |
| "loss": 0.6592, |
| "step": 1246 |
| }, |
| { |
| "epoch": 0.4338605944724492, |
| "grad_norm": 1.063023358388528, |
| "learning_rate": 0.0001917966619120088, |
| "loss": 0.6998, |
| "step": 1248 |
| }, |
| { |
| "epoch": 0.4345558838866678, |
| "grad_norm": 1.0481626997531774, |
| "learning_rate": 0.00019174844371986111, |
| "loss": 0.5649, |
| "step": 1250 |
| }, |
| { |
| "epoch": 0.4352511733008865, |
| "grad_norm": 0.6436744584316785, |
| "learning_rate": 0.00019170009032984188, |
| "loss": 0.6011, |
| "step": 1252 |
| }, |
| { |
| "epoch": 0.4359464627151052, |
| "grad_norm": 0.6100699160131396, |
| "learning_rate": 0.00019165160181320331, |
| "loss": 0.6229, |
| "step": 1254 |
| }, |
| { |
| "epoch": 0.4366417521293238, |
| "grad_norm": 0.43075456887586555, |
| "learning_rate": 0.00019160297824139671, |
| "loss": 0.5985, |
| "step": 1256 |
| }, |
| { |
| "epoch": 0.4373370415435425, |
| "grad_norm": 1.1133754307509498, |
| "learning_rate": 0.0001915542196860723, |
| "loss": 0.582, |
| "step": 1258 |
| }, |
| { |
| "epoch": 0.4380323309577612, |
| "grad_norm": 0.8115545368570724, |
| "learning_rate": 0.00019150532621907935, |
| "loss": 0.6281, |
| "step": 1260 |
| }, |
| { |
| "epoch": 0.4387276203719798, |
| "grad_norm": 0.6501357510133978, |
| "learning_rate": 0.00019145629791246586, |
| "loss": 0.7366, |
| "step": 1262 |
| }, |
| { |
| "epoch": 0.4394229097861985, |
| "grad_norm": 1.0301487119179336, |
| "learning_rate": 0.00019140713483847854, |
| "loss": 0.5878, |
| "step": 1264 |
| }, |
| { |
| "epoch": 0.44011819920041717, |
| "grad_norm": 0.9720409793058, |
| "learning_rate": 0.00019135783706956266, |
| "loss": 0.5096, |
| "step": 1266 |
| }, |
| { |
| "epoch": 0.4408134886146358, |
| "grad_norm": 0.5612642914060567, |
| "learning_rate": 0.000191308404678362, |
| "loss": 0.6624, |
| "step": 1268 |
| }, |
| { |
| "epoch": 0.4415087780288545, |
| "grad_norm": 0.46162422465509245, |
| "learning_rate": 0.00019125883773771874, |
| "loss": 0.5714, |
| "step": 1270 |
| }, |
| { |
| "epoch": 0.44220406744307317, |
| "grad_norm": 0.7669640725186674, |
| "learning_rate": 0.00019120913632067325, |
| "loss": 0.6786, |
| "step": 1272 |
| }, |
| { |
| "epoch": 0.44289935685729187, |
| "grad_norm": 0.4747976521609601, |
| "learning_rate": 0.00019115930050046416, |
| "loss": 0.5964, |
| "step": 1274 |
| }, |
| { |
| "epoch": 0.4435946462715105, |
| "grad_norm": 0.8528446125546961, |
| "learning_rate": 0.0001911093303505281, |
| "loss": 0.6584, |
| "step": 1276 |
| }, |
| { |
| "epoch": 0.44428993568572916, |
| "grad_norm": 0.444423482826933, |
| "learning_rate": 0.00019105922594449962, |
| "loss": 0.5931, |
| "step": 1278 |
| }, |
| { |
| "epoch": 0.44498522509994787, |
| "grad_norm": 0.4296567435369722, |
| "learning_rate": 0.00019100898735621114, |
| "loss": 0.5614, |
| "step": 1280 |
| }, |
| { |
| "epoch": 0.4456805145141665, |
| "grad_norm": 0.3735319415225358, |
| "learning_rate": 0.0001909586146596928, |
| "loss": 0.5585, |
| "step": 1282 |
| }, |
| { |
| "epoch": 0.4463758039283852, |
| "grad_norm": 0.3997395665103462, |
| "learning_rate": 0.0001909081079291724, |
| "loss": 0.4911, |
| "step": 1284 |
| }, |
| { |
| "epoch": 0.44707109334260386, |
| "grad_norm": 0.5067764089739433, |
| "learning_rate": 0.00019085746723907513, |
| "loss": 0.5912, |
| "step": 1286 |
| }, |
| { |
| "epoch": 0.4477663827568225, |
| "grad_norm": 1.5023247096958177, |
| "learning_rate": 0.00019080669266402373, |
| "loss": 0.7538, |
| "step": 1288 |
| }, |
| { |
| "epoch": 0.4484616721710412, |
| "grad_norm": 1.0536335728512713, |
| "learning_rate": 0.0001907557842788381, |
| "loss": 0.7347, |
| "step": 1290 |
| }, |
| { |
| "epoch": 0.44915696158525986, |
| "grad_norm": 1.3178937768983239, |
| "learning_rate": 0.00019070474215853543, |
| "loss": 0.6302, |
| "step": 1292 |
| }, |
| { |
| "epoch": 0.4498522509994785, |
| "grad_norm": 0.5893948676721078, |
| "learning_rate": 0.00019065356637832986, |
| "loss": 0.626, |
| "step": 1294 |
| }, |
| { |
| "epoch": 0.4505475404136972, |
| "grad_norm": 1.118265378089079, |
| "learning_rate": 0.0001906022570136326, |
| "loss": 0.5155, |
| "step": 1296 |
| }, |
| { |
| "epoch": 0.45124282982791586, |
| "grad_norm": 0.3585975576830803, |
| "learning_rate": 0.00019055081414005165, |
| "loss": 0.5558, |
| "step": 1298 |
| }, |
| { |
| "epoch": 0.45193811924213456, |
| "grad_norm": 1.3203699153765986, |
| "learning_rate": 0.00019049923783339171, |
| "loss": 0.6516, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.4526334086563532, |
| "grad_norm": 0.8836139230641887, |
| "learning_rate": 0.0001904475281696542, |
| "loss": 0.6532, |
| "step": 1302 |
| }, |
| { |
| "epoch": 0.45332869807057186, |
| "grad_norm": 0.7111007749248109, |
| "learning_rate": 0.00019039568522503694, |
| "loss": 0.6273, |
| "step": 1304 |
| }, |
| { |
| "epoch": 0.45402398748479056, |
| "grad_norm": 0.4303152450432435, |
| "learning_rate": 0.00019034370907593427, |
| "loss": 0.5596, |
| "step": 1306 |
| }, |
| { |
| "epoch": 0.4547192768990092, |
| "grad_norm": 0.46058879749054005, |
| "learning_rate": 0.00019029159979893669, |
| "loss": 0.6002, |
| "step": 1308 |
| }, |
| { |
| "epoch": 0.4554145663132279, |
| "grad_norm": 1.1757657004565762, |
| "learning_rate": 0.00019023935747083094, |
| "loss": 0.5425, |
| "step": 1310 |
| }, |
| { |
| "epoch": 0.45610985572744656, |
| "grad_norm": 0.6763815344574016, |
| "learning_rate": 0.00019018698216859985, |
| "loss": 0.7178, |
| "step": 1312 |
| }, |
| { |
| "epoch": 0.4568051451416652, |
| "grad_norm": 0.6557528693425259, |
| "learning_rate": 0.00019013447396942215, |
| "loss": 0.5482, |
| "step": 1314 |
| }, |
| { |
| "epoch": 0.4575004345558839, |
| "grad_norm": 0.40206624331116686, |
| "learning_rate": 0.0001900818329506724, |
| "loss": 0.5626, |
| "step": 1316 |
| }, |
| { |
| "epoch": 0.45819572397010255, |
| "grad_norm": 0.39959833266619904, |
| "learning_rate": 0.0001900290591899209, |
| "loss": 0.5805, |
| "step": 1318 |
| }, |
| { |
| "epoch": 0.4588910133843212, |
| "grad_norm": 1.157537080813341, |
| "learning_rate": 0.00018997615276493353, |
| "loss": 0.5895, |
| "step": 1320 |
| }, |
| { |
| "epoch": 0.4595863027985399, |
| "grad_norm": 1.9818188598958286, |
| "learning_rate": 0.0001899231137536717, |
| "loss": 0.6688, |
| "step": 1322 |
| }, |
| { |
| "epoch": 0.46028159221275855, |
| "grad_norm": 0.6448001142497345, |
| "learning_rate": 0.00018986994223429217, |
| "loss": 0.5878, |
| "step": 1324 |
| }, |
| { |
| "epoch": 0.46097688162697725, |
| "grad_norm": 0.7035140294155504, |
| "learning_rate": 0.000189816638285147, |
| "loss": 0.59, |
| "step": 1326 |
| }, |
| { |
| "epoch": 0.4616721710411959, |
| "grad_norm": 0.5547705850102905, |
| "learning_rate": 0.00018976320198478327, |
| "loss": 0.6194, |
| "step": 1328 |
| }, |
| { |
| "epoch": 0.46236746045541455, |
| "grad_norm": 0.8218894600210293, |
| "learning_rate": 0.00018970963341194327, |
| "loss": 0.5957, |
| "step": 1330 |
| }, |
| { |
| "epoch": 0.46306274986963325, |
| "grad_norm": 0.8582770909742338, |
| "learning_rate": 0.00018965593264556405, |
| "loss": 0.6027, |
| "step": 1332 |
| }, |
| { |
| "epoch": 0.4637580392838519, |
| "grad_norm": 0.5859159431078994, |
| "learning_rate": 0.00018960209976477755, |
| "loss": 0.5701, |
| "step": 1334 |
| }, |
| { |
| "epoch": 0.46445332869807054, |
| "grad_norm": 0.6055238281909472, |
| "learning_rate": 0.00018954813484891033, |
| "loss": 0.6069, |
| "step": 1336 |
| }, |
| { |
| "epoch": 0.46514861811228925, |
| "grad_norm": 0.7336223128037048, |
| "learning_rate": 0.00018949403797748356, |
| "loss": 0.639, |
| "step": 1338 |
| }, |
| { |
| "epoch": 0.4658439075265079, |
| "grad_norm": 0.4617499217742889, |
| "learning_rate": 0.0001894398092302128, |
| "loss": 0.5972, |
| "step": 1340 |
| }, |
| { |
| "epoch": 0.4665391969407266, |
| "grad_norm": 0.6626369850916174, |
| "learning_rate": 0.00018938544868700804, |
| "loss": 0.6254, |
| "step": 1342 |
| }, |
| { |
| "epoch": 0.46723448635494524, |
| "grad_norm": 0.6487571201649612, |
| "learning_rate": 0.00018933095642797336, |
| "loss": 0.6115, |
| "step": 1344 |
| }, |
| { |
| "epoch": 0.4679297757691639, |
| "grad_norm": 0.721231100905907, |
| "learning_rate": 0.00018927633253340703, |
| "loss": 0.5056, |
| "step": 1346 |
| }, |
| { |
| "epoch": 0.4686250651833826, |
| "grad_norm": 0.43640759465379947, |
| "learning_rate": 0.0001892215770838012, |
| "loss": 0.618, |
| "step": 1348 |
| }, |
| { |
| "epoch": 0.46932035459760124, |
| "grad_norm": 0.7472697329585503, |
| "learning_rate": 0.00018916669015984198, |
| "loss": 0.5145, |
| "step": 1350 |
| }, |
| { |
| "epoch": 0.46932035459760124, |
| "eval_loss": 0.6091039180755615, |
| "eval_runtime": 712.5909, |
| "eval_samples_per_second": 6.795, |
| "eval_steps_per_second": 0.213, |
| "step": 1350 |
| }, |
| { |
| "epoch": 0.47001564401181994, |
| "grad_norm": 0.8587847612083451, |
| "learning_rate": 0.00018911167184240915, |
| "loss": 0.6002, |
| "step": 1352 |
| }, |
| { |
| "epoch": 0.4707109334260386, |
| "grad_norm": 0.570124970949044, |
| "learning_rate": 0.0001890565222125761, |
| "loss": 0.6562, |
| "step": 1354 |
| }, |
| { |
| "epoch": 0.47140622284025724, |
| "grad_norm": 0.6984427176055181, |
| "learning_rate": 0.00018900124135160976, |
| "loss": 0.7262, |
| "step": 1356 |
| }, |
| { |
| "epoch": 0.47210151225447594, |
| "grad_norm": 0.765890528385208, |
| "learning_rate": 0.0001889458293409704, |
| "loss": 0.5891, |
| "step": 1358 |
| }, |
| { |
| "epoch": 0.4727968016686946, |
| "grad_norm": 0.609256703935707, |
| "learning_rate": 0.0001888902862623116, |
| "loss": 0.5919, |
| "step": 1360 |
| }, |
| { |
| "epoch": 0.47349209108291324, |
| "grad_norm": 0.4622539193764851, |
| "learning_rate": 0.00018883461219748, |
| "loss": 0.6179, |
| "step": 1362 |
| }, |
| { |
| "epoch": 0.47418738049713194, |
| "grad_norm": 0.46144204894965307, |
| "learning_rate": 0.00018877880722851536, |
| "loss": 0.6289, |
| "step": 1364 |
| }, |
| { |
| "epoch": 0.4748826699113506, |
| "grad_norm": 0.8200013455891142, |
| "learning_rate": 0.00018872287143765023, |
| "loss": 0.5644, |
| "step": 1366 |
| }, |
| { |
| "epoch": 0.4755779593255693, |
| "grad_norm": 0.5631892935159202, |
| "learning_rate": 0.00018866680490730998, |
| "loss": 0.5034, |
| "step": 1368 |
| }, |
| { |
| "epoch": 0.47627324873978794, |
| "grad_norm": 0.5739777810808598, |
| "learning_rate": 0.00018861060772011273, |
| "loss": 0.651, |
| "step": 1370 |
| }, |
| { |
| "epoch": 0.4769685381540066, |
| "grad_norm": 0.666694270081405, |
| "learning_rate": 0.00018855427995886892, |
| "loss": 0.58, |
| "step": 1372 |
| }, |
| { |
| "epoch": 0.4776638275682253, |
| "grad_norm": 0.4844198549303021, |
| "learning_rate": 0.00018849782170658158, |
| "loss": 0.5842, |
| "step": 1374 |
| }, |
| { |
| "epoch": 0.47835911698244393, |
| "grad_norm": 0.5289779719515055, |
| "learning_rate": 0.00018844123304644596, |
| "loss": 0.597, |
| "step": 1376 |
| }, |
| { |
| "epoch": 0.47905440639666264, |
| "grad_norm": 1.0867526542265191, |
| "learning_rate": 0.0001883845140618495, |
| "loss": 0.6741, |
| "step": 1378 |
| }, |
| { |
| "epoch": 0.4797496958108813, |
| "grad_norm": 0.6205435292114323, |
| "learning_rate": 0.00018832766483637165, |
| "loss": 0.6176, |
| "step": 1380 |
| }, |
| { |
| "epoch": 0.48044498522509993, |
| "grad_norm": 0.49806740313966835, |
| "learning_rate": 0.0001882706854537838, |
| "loss": 0.6003, |
| "step": 1382 |
| }, |
| { |
| "epoch": 0.48114027463931863, |
| "grad_norm": 0.713610401013068, |
| "learning_rate": 0.00018821357599804915, |
| "loss": 0.5746, |
| "step": 1384 |
| }, |
| { |
| "epoch": 0.4818355640535373, |
| "grad_norm": 0.7286362166201054, |
| "learning_rate": 0.00018815633655332252, |
| "loss": 0.5824, |
| "step": 1386 |
| }, |
| { |
| "epoch": 0.4825308534677559, |
| "grad_norm": 0.7609524076074863, |
| "learning_rate": 0.00018809896720395033, |
| "loss": 0.5422, |
| "step": 1388 |
| }, |
| { |
| "epoch": 0.48322614288197463, |
| "grad_norm": 0.8489750160154781, |
| "learning_rate": 0.0001880414680344704, |
| "loss": 0.5797, |
| "step": 1390 |
| }, |
| { |
| "epoch": 0.4839214322961933, |
| "grad_norm": 1.3137718080784093, |
| "learning_rate": 0.00018798383912961187, |
| "loss": 0.5397, |
| "step": 1392 |
| }, |
| { |
| "epoch": 0.484616721710412, |
| "grad_norm": 0.7192606649476782, |
| "learning_rate": 0.00018792608057429503, |
| "loss": 0.6948, |
| "step": 1394 |
| }, |
| { |
| "epoch": 0.4853120111246306, |
| "grad_norm": 0.8441204863561427, |
| "learning_rate": 0.00018786819245363118, |
| "loss": 0.5786, |
| "step": 1396 |
| }, |
| { |
| "epoch": 0.4860073005388493, |
| "grad_norm": 0.6806475957554041, |
| "learning_rate": 0.00018781017485292267, |
| "loss": 0.6033, |
| "step": 1398 |
| }, |
| { |
| "epoch": 0.486702589953068, |
| "grad_norm": 0.5168648519720452, |
| "learning_rate": 0.00018775202785766256, |
| "loss": 0.7098, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.4873978793672866, |
| "grad_norm": 0.465013130005678, |
| "learning_rate": 0.00018769375155353453, |
| "loss": 0.7186, |
| "step": 1402 |
| }, |
| { |
| "epoch": 0.4880931687815053, |
| "grad_norm": 0.453274392570444, |
| "learning_rate": 0.00018763534602641291, |
| "loss": 0.5782, |
| "step": 1404 |
| }, |
| { |
| "epoch": 0.488788458195724, |
| "grad_norm": 0.5972503276564156, |
| "learning_rate": 0.00018757681136236242, |
| "loss": 0.6221, |
| "step": 1406 |
| }, |
| { |
| "epoch": 0.4894837476099426, |
| "grad_norm": 0.36846650508391193, |
| "learning_rate": 0.00018751814764763806, |
| "loss": 0.5966, |
| "step": 1408 |
| }, |
| { |
| "epoch": 0.4901790370241613, |
| "grad_norm": 0.3451947155848497, |
| "learning_rate": 0.00018745935496868493, |
| "loss": 0.5508, |
| "step": 1410 |
| }, |
| { |
| "epoch": 0.49087432643837997, |
| "grad_norm": 0.5078133383734431, |
| "learning_rate": 0.00018740043341213832, |
| "loss": 0.5029, |
| "step": 1412 |
| }, |
| { |
| "epoch": 0.4915696158525986, |
| "grad_norm": 0.767602706979167, |
| "learning_rate": 0.00018734138306482332, |
| "loss": 0.6083, |
| "step": 1414 |
| }, |
| { |
| "epoch": 0.4922649052668173, |
| "grad_norm": 0.5648335118893248, |
| "learning_rate": 0.00018728220401375477, |
| "loss": 0.618, |
| "step": 1416 |
| }, |
| { |
| "epoch": 0.49296019468103597, |
| "grad_norm": 0.7659928375858898, |
| "learning_rate": 0.00018722289634613728, |
| "loss": 0.5907, |
| "step": 1418 |
| }, |
| { |
| "epoch": 0.49365548409525467, |
| "grad_norm": 0.9164796669061794, |
| "learning_rate": 0.0001871634601493649, |
| "loss": 0.4978, |
| "step": 1420 |
| }, |
| { |
| "epoch": 0.4943507735094733, |
| "grad_norm": 0.49130889452749815, |
| "learning_rate": 0.00018710389551102115, |
| "loss": 0.5277, |
| "step": 1422 |
| }, |
| { |
| "epoch": 0.49504606292369197, |
| "grad_norm": 0.5824962532822966, |
| "learning_rate": 0.00018704420251887868, |
| "loss": 0.5717, |
| "step": 1424 |
| }, |
| { |
| "epoch": 0.49574135233791067, |
| "grad_norm": 0.5964322775094694, |
| "learning_rate": 0.00018698438126089944, |
| "loss": 0.6249, |
| "step": 1426 |
| }, |
| { |
| "epoch": 0.4964366417521293, |
| "grad_norm": 0.42806592395765636, |
| "learning_rate": 0.0001869244318252343, |
| "loss": 0.5433, |
| "step": 1428 |
| }, |
| { |
| "epoch": 0.497131931166348, |
| "grad_norm": 0.8340709008133664, |
| "learning_rate": 0.000186864354300223, |
| "loss": 0.5714, |
| "step": 1430 |
| }, |
| { |
| "epoch": 0.49782722058056667, |
| "grad_norm": 0.6732191602218343, |
| "learning_rate": 0.00018680414877439406, |
| "loss": 0.6572, |
| "step": 1432 |
| }, |
| { |
| "epoch": 0.4985225099947853, |
| "grad_norm": 0.5785424946243081, |
| "learning_rate": 0.00018674381533646467, |
| "loss": 0.5831, |
| "step": 1434 |
| }, |
| { |
| "epoch": 0.499217799409004, |
| "grad_norm": 0.6435404053128269, |
| "learning_rate": 0.0001866833540753404, |
| "loss": 0.582, |
| "step": 1436 |
| }, |
| { |
| "epoch": 0.49991308882322266, |
| "grad_norm": 0.5713158720544551, |
| "learning_rate": 0.00018662276508011526, |
| "loss": 0.5276, |
| "step": 1438 |
| }, |
| { |
| "epoch": 0.5006083782374413, |
| "grad_norm": 0.43458150402178386, |
| "learning_rate": 0.00018656204844007143, |
| "loss": 0.5526, |
| "step": 1440 |
| }, |
| { |
| "epoch": 0.50130366765166, |
| "grad_norm": 0.5522953777612155, |
| "learning_rate": 0.00018650120424467922, |
| "loss": 0.6259, |
| "step": 1442 |
| }, |
| { |
| "epoch": 0.5019989570658787, |
| "grad_norm": 0.6700198262354904, |
| "learning_rate": 0.00018644023258359684, |
| "loss": 0.5976, |
| "step": 1444 |
| }, |
| { |
| "epoch": 0.5026942464800973, |
| "grad_norm": 0.653642473920571, |
| "learning_rate": 0.00018637913354667044, |
| "loss": 0.6187, |
| "step": 1446 |
| }, |
| { |
| "epoch": 0.503389535894316, |
| "grad_norm": 0.8572149566943323, |
| "learning_rate": 0.00018631790722393378, |
| "loss": 0.6685, |
| "step": 1448 |
| }, |
| { |
| "epoch": 0.5040848253085347, |
| "grad_norm": 0.4144221514106495, |
| "learning_rate": 0.00018625655370560823, |
| "loss": 0.514, |
| "step": 1450 |
| }, |
| { |
| "epoch": 0.5047801147227533, |
| "grad_norm": 0.8737004492907994, |
| "learning_rate": 0.00018619507308210255, |
| "loss": 0.6583, |
| "step": 1452 |
| }, |
| { |
| "epoch": 0.505475404136972, |
| "grad_norm": 0.3987400457590611, |
| "learning_rate": 0.00018613346544401281, |
| "loss": 0.5663, |
| "step": 1454 |
| }, |
| { |
| "epoch": 0.5061706935511907, |
| "grad_norm": 0.591401748694551, |
| "learning_rate": 0.0001860717308821223, |
| "loss": 0.6537, |
| "step": 1456 |
| }, |
| { |
| "epoch": 0.5068659829654093, |
| "grad_norm": 0.555392995714629, |
| "learning_rate": 0.00018600986948740124, |
| "loss": 0.5692, |
| "step": 1458 |
| }, |
| { |
| "epoch": 0.507561272379628, |
| "grad_norm": 0.39435725187187276, |
| "learning_rate": 0.00018594788135100688, |
| "loss": 0.6265, |
| "step": 1460 |
| }, |
| { |
| "epoch": 0.5082565617938467, |
| "grad_norm": 0.4077334073620459, |
| "learning_rate": 0.00018588576656428308, |
| "loss": 0.5554, |
| "step": 1462 |
| }, |
| { |
| "epoch": 0.5089518512080654, |
| "grad_norm": 0.38687853565198027, |
| "learning_rate": 0.00018582352521876046, |
| "loss": 0.6378, |
| "step": 1464 |
| }, |
| { |
| "epoch": 0.509647140622284, |
| "grad_norm": 0.31280184968676994, |
| "learning_rate": 0.00018576115740615606, |
| "loss": 0.5416, |
| "step": 1466 |
| }, |
| { |
| "epoch": 0.5103424300365027, |
| "grad_norm": 1.3853246114916373, |
| "learning_rate": 0.00018569866321837327, |
| "loss": 0.6763, |
| "step": 1468 |
| }, |
| { |
| "epoch": 0.5110377194507214, |
| "grad_norm": 0.5024331777389988, |
| "learning_rate": 0.00018563604274750178, |
| "loss": 0.6372, |
| "step": 1470 |
| }, |
| { |
| "epoch": 0.51173300886494, |
| "grad_norm": 0.32922333596717174, |
| "learning_rate": 0.0001855732960858173, |
| "loss": 0.5923, |
| "step": 1472 |
| }, |
| { |
| "epoch": 0.5124282982791587, |
| "grad_norm": 0.9059198527449924, |
| "learning_rate": 0.00018551042332578144, |
| "loss": 0.659, |
| "step": 1474 |
| }, |
| { |
| "epoch": 0.5131235876933774, |
| "grad_norm": 0.7045580743620987, |
| "learning_rate": 0.00018544742456004181, |
| "loss": 0.6416, |
| "step": 1476 |
| }, |
| { |
| "epoch": 0.513818877107596, |
| "grad_norm": 1.6530740850637509, |
| "learning_rate": 0.0001853842998814315, |
| "loss": 0.5863, |
| "step": 1478 |
| }, |
| { |
| "epoch": 0.5145141665218147, |
| "grad_norm": 0.742405889420684, |
| "learning_rate": 0.00018532104938296927, |
| "loss": 0.5942, |
| "step": 1480 |
| }, |
| { |
| "epoch": 0.5152094559360334, |
| "grad_norm": 0.43475036345006896, |
| "learning_rate": 0.0001852576731578592, |
| "loss": 0.54, |
| "step": 1482 |
| }, |
| { |
| "epoch": 0.515904745350252, |
| "grad_norm": 1.0939432717807218, |
| "learning_rate": 0.00018519417129949072, |
| "loss": 0.6034, |
| "step": 1484 |
| }, |
| { |
| "epoch": 0.5166000347644707, |
| "grad_norm": 1.1841858811933605, |
| "learning_rate": 0.0001851305439014383, |
| "loss": 0.6119, |
| "step": 1486 |
| }, |
| { |
| "epoch": 0.5172953241786894, |
| "grad_norm": 0.8950773429853176, |
| "learning_rate": 0.00018506679105746149, |
| "loss": 0.6224, |
| "step": 1488 |
| }, |
| { |
| "epoch": 0.5179906135929081, |
| "grad_norm": 0.9558807568134642, |
| "learning_rate": 0.0001850029128615046, |
| "loss": 0.5986, |
| "step": 1490 |
| }, |
| { |
| "epoch": 0.5186859030071267, |
| "grad_norm": 0.45382840521759343, |
| "learning_rate": 0.00018493890940769671, |
| "loss": 0.61, |
| "step": 1492 |
| }, |
| { |
| "epoch": 0.5193811924213454, |
| "grad_norm": 0.4697574815571038, |
| "learning_rate": 0.00018487478079035153, |
| "loss": 0.5299, |
| "step": 1494 |
| }, |
| { |
| "epoch": 0.5200764818355641, |
| "grad_norm": 0.699871643062895, |
| "learning_rate": 0.00018481052710396707, |
| "loss": 0.6021, |
| "step": 1496 |
| }, |
| { |
| "epoch": 0.5207717712497827, |
| "grad_norm": 0.9389496797384421, |
| "learning_rate": 0.0001847461484432258, |
| "loss": 0.6119, |
| "step": 1498 |
| }, |
| { |
| "epoch": 0.5214670606640014, |
| "grad_norm": 1.1160995208439726, |
| "learning_rate": 0.00018468164490299417, |
| "loss": 0.59, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.5214670606640014, |
| "eval_loss": 0.6035115718841553, |
| "eval_runtime": 374.7159, |
| "eval_samples_per_second": 12.922, |
| "eval_steps_per_second": 0.406, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.5221623500782201, |
| "grad_norm": 0.8572137642677378, |
| "learning_rate": 0.00018461701657832285, |
| "loss": 0.5865, |
| "step": 1502 |
| }, |
| { |
| "epoch": 0.5228576394924387, |
| "grad_norm": 0.9221217554185654, |
| "learning_rate": 0.0001845522635644462, |
| "loss": 0.5479, |
| "step": 1504 |
| }, |
| { |
| "epoch": 0.5235529289066574, |
| "grad_norm": 2.0492350557846577, |
| "learning_rate": 0.0001844873859567824, |
| "loss": 0.7292, |
| "step": 1506 |
| }, |
| { |
| "epoch": 0.5242482183208761, |
| "grad_norm": 1.1824788332559788, |
| "learning_rate": 0.0001844223838509333, |
| "loss": 0.6479, |
| "step": 1508 |
| }, |
| { |
| "epoch": 0.5249435077350947, |
| "grad_norm": 0.7039837396797187, |
| "learning_rate": 0.00018435725734268407, |
| "loss": 0.6149, |
| "step": 1510 |
| }, |
| { |
| "epoch": 0.5256387971493134, |
| "grad_norm": 0.4401391384170765, |
| "learning_rate": 0.0001842920065280033, |
| "loss": 0.5854, |
| "step": 1512 |
| }, |
| { |
| "epoch": 0.5263340865635321, |
| "grad_norm": 0.4100923191497827, |
| "learning_rate": 0.0001842266315030427, |
| "loss": 0.6091, |
| "step": 1514 |
| }, |
| { |
| "epoch": 0.5270293759777508, |
| "grad_norm": 0.5510661253393345, |
| "learning_rate": 0.00018416113236413705, |
| "loss": 0.5721, |
| "step": 1516 |
| }, |
| { |
| "epoch": 0.5277246653919694, |
| "grad_norm": 0.4386368092109241, |
| "learning_rate": 0.00018409550920780397, |
| "loss": 0.6232, |
| "step": 1518 |
| }, |
| { |
| "epoch": 0.5284199548061881, |
| "grad_norm": 0.5520600896880458, |
| "learning_rate": 0.00018402976213074385, |
| "loss": 0.6134, |
| "step": 1520 |
| }, |
| { |
| "epoch": 0.5291152442204068, |
| "grad_norm": 0.5101647982715836, |
| "learning_rate": 0.00018396389122983975, |
| "loss": 0.5918, |
| "step": 1522 |
| }, |
| { |
| "epoch": 0.5298105336346254, |
| "grad_norm": 0.7801958609328837, |
| "learning_rate": 0.00018389789660215708, |
| "loss": 0.6907, |
| "step": 1524 |
| }, |
| { |
| "epoch": 0.5305058230488441, |
| "grad_norm": 0.5012202349580023, |
| "learning_rate": 0.00018383177834494364, |
| "loss": 0.5818, |
| "step": 1526 |
| }, |
| { |
| "epoch": 0.5312011124630628, |
| "grad_norm": 0.9534096632353795, |
| "learning_rate": 0.00018376553655562945, |
| "loss": 0.661, |
| "step": 1528 |
| }, |
| { |
| "epoch": 0.5318964018772814, |
| "grad_norm": 0.7514828327361283, |
| "learning_rate": 0.00018369917133182644, |
| "loss": 0.5764, |
| "step": 1530 |
| }, |
| { |
| "epoch": 0.5325916912915001, |
| "grad_norm": 0.5336990715721825, |
| "learning_rate": 0.0001836326827713285, |
| "loss": 0.5652, |
| "step": 1532 |
| }, |
| { |
| "epoch": 0.5332869807057188, |
| "grad_norm": 0.5458405171879034, |
| "learning_rate": 0.0001835660709721113, |
| "loss": 0.577, |
| "step": 1534 |
| }, |
| { |
| "epoch": 0.5339822701199374, |
| "grad_norm": 0.649664481255766, |
| "learning_rate": 0.00018349933603233206, |
| "loss": 0.5599, |
| "step": 1536 |
| }, |
| { |
| "epoch": 0.5346775595341561, |
| "grad_norm": 0.4148065039425594, |
| "learning_rate": 0.0001834324780503295, |
| "loss": 0.6437, |
| "step": 1538 |
| }, |
| { |
| "epoch": 0.5353728489483748, |
| "grad_norm": 0.821550084995287, |
| "learning_rate": 0.00018336549712462353, |
| "loss": 0.6837, |
| "step": 1540 |
| }, |
| { |
| "epoch": 0.5360681383625935, |
| "grad_norm": 0.5564568145383851, |
| "learning_rate": 0.00018329839335391542, |
| "loss": 0.5715, |
| "step": 1542 |
| }, |
| { |
| "epoch": 0.5367634277768121, |
| "grad_norm": 0.6705822690313288, |
| "learning_rate": 0.0001832311668370873, |
| "loss": 0.5669, |
| "step": 1544 |
| }, |
| { |
| "epoch": 0.5374587171910308, |
| "grad_norm": 0.7821014824731976, |
| "learning_rate": 0.00018316381767320228, |
| "loss": 0.6117, |
| "step": 1546 |
| }, |
| { |
| "epoch": 0.5381540066052495, |
| "grad_norm": 0.416568249234258, |
| "learning_rate": 0.00018309634596150412, |
| "loss": 0.5604, |
| "step": 1548 |
| }, |
| { |
| "epoch": 0.5388492960194681, |
| "grad_norm": 0.6063645649473942, |
| "learning_rate": 0.0001830287518014172, |
| "loss": 0.6944, |
| "step": 1550 |
| }, |
| { |
| "epoch": 0.5395445854336868, |
| "grad_norm": 0.7965496657255345, |
| "learning_rate": 0.0001829610352925463, |
| "loss": 0.5101, |
| "step": 1552 |
| }, |
| { |
| "epoch": 0.5402398748479055, |
| "grad_norm": 0.6352930080090265, |
| "learning_rate": 0.0001828931965346766, |
| "loss": 0.5392, |
| "step": 1554 |
| }, |
| { |
| "epoch": 0.5409351642621241, |
| "grad_norm": 0.7018715777498272, |
| "learning_rate": 0.00018282523562777327, |
| "loss": 0.5115, |
| "step": 1556 |
| }, |
| { |
| "epoch": 0.5416304536763428, |
| "grad_norm": 0.7055081199927248, |
| "learning_rate": 0.0001827571526719816, |
| "loss": 0.6342, |
| "step": 1558 |
| }, |
| { |
| "epoch": 0.5423257430905615, |
| "grad_norm": 0.9176268923346018, |
| "learning_rate": 0.0001826889477676267, |
| "loss": 0.641, |
| "step": 1560 |
| }, |
| { |
| "epoch": 0.5430210325047801, |
| "grad_norm": 0.9614728125926221, |
| "learning_rate": 0.00018262062101521328, |
| "loss": 0.662, |
| "step": 1562 |
| }, |
| { |
| "epoch": 0.5437163219189988, |
| "grad_norm": 1.1496735145261052, |
| "learning_rate": 0.00018255217251542574, |
| "loss": 0.537, |
| "step": 1564 |
| }, |
| { |
| "epoch": 0.5444116113332175, |
| "grad_norm": 0.6789192030318737, |
| "learning_rate": 0.00018248360236912777, |
| "loss": 0.5446, |
| "step": 1566 |
| }, |
| { |
| "epoch": 0.5451069007474362, |
| "grad_norm": 0.8374235506929065, |
| "learning_rate": 0.00018241491067736242, |
| "loss": 0.6755, |
| "step": 1568 |
| }, |
| { |
| "epoch": 0.5458021901616548, |
| "grad_norm": 1.0198490239995306, |
| "learning_rate": 0.0001823460975413518, |
| "loss": 0.5292, |
| "step": 1570 |
| }, |
| { |
| "epoch": 0.5464974795758735, |
| "grad_norm": 0.5392136161037133, |
| "learning_rate": 0.0001822771630624969, |
| "loss": 0.6962, |
| "step": 1572 |
| }, |
| { |
| "epoch": 0.5471927689900922, |
| "grad_norm": 0.5006386802591807, |
| "learning_rate": 0.00018220810734237766, |
| "loss": 0.5946, |
| "step": 1574 |
| }, |
| { |
| "epoch": 0.5478880584043108, |
| "grad_norm": 0.5478881526480635, |
| "learning_rate": 0.0001821389304827526, |
| "loss": 0.5844, |
| "step": 1576 |
| }, |
| { |
| "epoch": 0.5485833478185295, |
| "grad_norm": 0.3440683975593449, |
| "learning_rate": 0.0001820696325855587, |
| "loss": 0.5598, |
| "step": 1578 |
| }, |
| { |
| "epoch": 0.5492786372327482, |
| "grad_norm": 0.36658750390429906, |
| "learning_rate": 0.0001820002137529114, |
| "loss": 0.6105, |
| "step": 1580 |
| }, |
| { |
| "epoch": 0.5499739266469668, |
| "grad_norm": 0.8580953766157888, |
| "learning_rate": 0.0001819306740871043, |
| "loss": 0.6685, |
| "step": 1582 |
| }, |
| { |
| "epoch": 0.5506692160611855, |
| "grad_norm": 0.7163438303185133, |
| "learning_rate": 0.00018186101369060902, |
| "loss": 0.593, |
| "step": 1584 |
| }, |
| { |
| "epoch": 0.5513645054754042, |
| "grad_norm": 0.34431820289313025, |
| "learning_rate": 0.00018179123266607515, |
| "loss": 0.6024, |
| "step": 1586 |
| }, |
| { |
| "epoch": 0.5520597948896228, |
| "grad_norm": 0.3504729261840606, |
| "learning_rate": 0.00018172133111633, |
| "loss": 0.6376, |
| "step": 1588 |
| }, |
| { |
| "epoch": 0.5527550843038415, |
| "grad_norm": 0.3806243378944845, |
| "learning_rate": 0.00018165130914437843, |
| "loss": 0.4432, |
| "step": 1590 |
| }, |
| { |
| "epoch": 0.5534503737180602, |
| "grad_norm": 0.44638264263646543, |
| "learning_rate": 0.00018158116685340286, |
| "loss": 0.584, |
| "step": 1592 |
| }, |
| { |
| "epoch": 0.5541456631322789, |
| "grad_norm": 0.8596816723413095, |
| "learning_rate": 0.00018151090434676296, |
| "loss": 0.6232, |
| "step": 1594 |
| }, |
| { |
| "epoch": 0.5548409525464975, |
| "grad_norm": 0.5102216453134625, |
| "learning_rate": 0.00018144052172799555, |
| "loss": 0.6171, |
| "step": 1596 |
| }, |
| { |
| "epoch": 0.5555362419607162, |
| "grad_norm": 0.728624434959544, |
| "learning_rate": 0.00018137001910081442, |
| "loss": 0.5828, |
| "step": 1598 |
| }, |
| { |
| "epoch": 0.5562315313749349, |
| "grad_norm": 0.6993841456831388, |
| "learning_rate": 0.00018129939656911018, |
| "loss": 0.6398, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.5569268207891535, |
| "grad_norm": 0.5081736994898716, |
| "learning_rate": 0.00018122865423695023, |
| "loss": 0.5907, |
| "step": 1602 |
| }, |
| { |
| "epoch": 0.5576221102033722, |
| "grad_norm": 0.9216105204641707, |
| "learning_rate": 0.00018115779220857835, |
| "loss": 0.475, |
| "step": 1604 |
| }, |
| { |
| "epoch": 0.5583173996175909, |
| "grad_norm": 0.5887668473902828, |
| "learning_rate": 0.0001810868105884149, |
| "loss": 0.6423, |
| "step": 1606 |
| }, |
| { |
| "epoch": 0.5590126890318095, |
| "grad_norm": 0.7052708958749251, |
| "learning_rate": 0.0001810157094810563, |
| "loss": 0.5979, |
| "step": 1608 |
| }, |
| { |
| "epoch": 0.5597079784460282, |
| "grad_norm": 0.6684687038470192, |
| "learning_rate": 0.00018094448899127506, |
| "loss": 0.5597, |
| "step": 1610 |
| }, |
| { |
| "epoch": 0.5604032678602469, |
| "grad_norm": 0.5472474373787231, |
| "learning_rate": 0.00018087314922401974, |
| "loss": 0.5604, |
| "step": 1612 |
| }, |
| { |
| "epoch": 0.5610985572744654, |
| "grad_norm": 1.046907488888251, |
| "learning_rate": 0.0001808016902844145, |
| "loss": 0.5588, |
| "step": 1614 |
| }, |
| { |
| "epoch": 0.5617938466886842, |
| "grad_norm": 0.599875325305421, |
| "learning_rate": 0.0001807301122777592, |
| "loss": 0.5333, |
| "step": 1616 |
| }, |
| { |
| "epoch": 0.5624891361029029, |
| "grad_norm": 0.7162383538759113, |
| "learning_rate": 0.0001806584153095291, |
| "loss": 0.5305, |
| "step": 1618 |
| }, |
| { |
| "epoch": 0.5631844255171216, |
| "grad_norm": 0.7631330239323099, |
| "learning_rate": 0.0001805865994853749, |
| "loss": 0.6774, |
| "step": 1620 |
| }, |
| { |
| "epoch": 0.5638797149313401, |
| "grad_norm": 0.5524440961094408, |
| "learning_rate": 0.00018051466491112223, |
| "loss": 0.5632, |
| "step": 1622 |
| }, |
| { |
| "epoch": 0.5645750043455589, |
| "grad_norm": 0.4987248734291855, |
| "learning_rate": 0.00018044261169277185, |
| "loss": 0.6215, |
| "step": 1624 |
| }, |
| { |
| "epoch": 0.5652702937597776, |
| "grad_norm": 0.7807389251113444, |
| "learning_rate": 0.00018037043993649935, |
| "loss": 0.6827, |
| "step": 1626 |
| }, |
| { |
| "epoch": 0.5659655831739961, |
| "grad_norm": 0.7200859056625541, |
| "learning_rate": 0.0001802981497486549, |
| "loss": 0.6462, |
| "step": 1628 |
| }, |
| { |
| "epoch": 0.5666608725882148, |
| "grad_norm": 0.5427861037081861, |
| "learning_rate": 0.0001802257412357632, |
| "loss": 0.5992, |
| "step": 1630 |
| }, |
| { |
| "epoch": 0.5673561620024336, |
| "grad_norm": 0.720385141439196, |
| "learning_rate": 0.0001801532145045235, |
| "loss": 0.5756, |
| "step": 1632 |
| }, |
| { |
| "epoch": 0.5680514514166521, |
| "grad_norm": 0.9552724965328024, |
| "learning_rate": 0.000180080569661809, |
| "loss": 0.6362, |
| "step": 1634 |
| }, |
| { |
| "epoch": 0.5687467408308708, |
| "grad_norm": 0.48995975515699663, |
| "learning_rate": 0.00018000780681466706, |
| "loss": 0.6285, |
| "step": 1636 |
| }, |
| { |
| "epoch": 0.5694420302450895, |
| "grad_norm": 0.3350821235074382, |
| "learning_rate": 0.00017993492607031896, |
| "loss": 0.6077, |
| "step": 1638 |
| }, |
| { |
| "epoch": 0.5701373196593081, |
| "grad_norm": 0.3998325347633491, |
| "learning_rate": 0.00017986192753615966, |
| "loss": 0.6461, |
| "step": 1640 |
| }, |
| { |
| "epoch": 0.5708326090735268, |
| "grad_norm": 0.4435600494500259, |
| "learning_rate": 0.0001797888113197577, |
| "loss": 0.6354, |
| "step": 1642 |
| }, |
| { |
| "epoch": 0.5715278984877455, |
| "grad_norm": 0.7021719076452477, |
| "learning_rate": 0.000179715577528855, |
| "loss": 0.609, |
| "step": 1644 |
| }, |
| { |
| "epoch": 0.5722231879019642, |
| "grad_norm": 0.3254773922584778, |
| "learning_rate": 0.00017964222627136684, |
| "loss": 0.5671, |
| "step": 1646 |
| }, |
| { |
| "epoch": 0.5729184773161828, |
| "grad_norm": 0.33864367997251066, |
| "learning_rate": 0.00017956875765538148, |
| "loss": 0.5037, |
| "step": 1648 |
| }, |
| { |
| "epoch": 0.5736137667304015, |
| "grad_norm": 0.7020501782188753, |
| "learning_rate": 0.00017949517178916018, |
| "loss": 0.6464, |
| "step": 1650 |
| }, |
| { |
| "epoch": 0.5736137667304015, |
| "eval_loss": 0.5962589979171753, |
| "eval_runtime": 371.2694, |
| "eval_samples_per_second": 13.042, |
| "eval_steps_per_second": 0.409, |
| "step": 1650 |
| }, |
| { |
| "epoch": 0.5743090561446202, |
| "grad_norm": 0.664359018460308, |
| "learning_rate": 0.00017942146878113696, |
| "loss": 0.54, |
| "step": 1652 |
| }, |
| { |
| "epoch": 0.5750043455588388, |
| "grad_norm": 0.4547846259763968, |
| "learning_rate": 0.00017934764873991842, |
| "loss": 0.5265, |
| "step": 1654 |
| }, |
| { |
| "epoch": 0.5756996349730575, |
| "grad_norm": 0.44499840681620506, |
| "learning_rate": 0.0001792737117742837, |
| "loss": 0.6455, |
| "step": 1656 |
| }, |
| { |
| "epoch": 0.5763949243872762, |
| "grad_norm": 1.1836252545555186, |
| "learning_rate": 0.00017919965799318414, |
| "loss": 0.5493, |
| "step": 1658 |
| }, |
| { |
| "epoch": 0.5770902138014948, |
| "grad_norm": 0.5250640289043013, |
| "learning_rate": 0.00017912548750574332, |
| "loss": 0.5912, |
| "step": 1660 |
| }, |
| { |
| "epoch": 0.5777855032157135, |
| "grad_norm": 0.5921000904212905, |
| "learning_rate": 0.00017905120042125674, |
| "loss": 0.6786, |
| "step": 1662 |
| }, |
| { |
| "epoch": 0.5784807926299322, |
| "grad_norm": 0.5069057332373537, |
| "learning_rate": 0.00017897679684919165, |
| "loss": 0.6578, |
| "step": 1664 |
| }, |
| { |
| "epoch": 0.5791760820441508, |
| "grad_norm": 0.8149193785886022, |
| "learning_rate": 0.00017890227689918708, |
| "loss": 0.5759, |
| "step": 1666 |
| }, |
| { |
| "epoch": 0.5798713714583695, |
| "grad_norm": 0.7674197048262045, |
| "learning_rate": 0.00017882764068105343, |
| "loss": 0.5992, |
| "step": 1668 |
| }, |
| { |
| "epoch": 0.5805666608725882, |
| "grad_norm": 0.7221408349004688, |
| "learning_rate": 0.00017875288830477255, |
| "loss": 0.6254, |
| "step": 1670 |
| }, |
| { |
| "epoch": 0.5812619502868069, |
| "grad_norm": 0.36822616889584014, |
| "learning_rate": 0.00017867801988049735, |
| "loss": 0.5479, |
| "step": 1672 |
| }, |
| { |
| "epoch": 0.5819572397010255, |
| "grad_norm": 0.6285282750057019, |
| "learning_rate": 0.00017860303551855183, |
| "loss": 0.6414, |
| "step": 1674 |
| }, |
| { |
| "epoch": 0.5826525291152442, |
| "grad_norm": 0.5000419603028949, |
| "learning_rate": 0.00017852793532943074, |
| "loss": 0.5929, |
| "step": 1676 |
| }, |
| { |
| "epoch": 0.5833478185294629, |
| "grad_norm": 0.46115086426294316, |
| "learning_rate": 0.0001784527194237996, |
| "loss": 0.5626, |
| "step": 1678 |
| }, |
| { |
| "epoch": 0.5840431079436815, |
| "grad_norm": 0.9767916828736493, |
| "learning_rate": 0.00017837738791249438, |
| "loss": 0.5637, |
| "step": 1680 |
| }, |
| { |
| "epoch": 0.5847383973579002, |
| "grad_norm": 0.4454990407423214, |
| "learning_rate": 0.00017830194090652146, |
| "loss": 0.5709, |
| "step": 1682 |
| }, |
| { |
| "epoch": 0.5854336867721189, |
| "grad_norm": 0.5297124460963825, |
| "learning_rate": 0.0001782263785170574, |
| "loss": 0.5378, |
| "step": 1684 |
| }, |
| { |
| "epoch": 0.5861289761863375, |
| "grad_norm": 0.5577701746789038, |
| "learning_rate": 0.0001781507008554487, |
| "loss": 0.5511, |
| "step": 1686 |
| }, |
| { |
| "epoch": 0.5868242656005562, |
| "grad_norm": 1.2633910061819977, |
| "learning_rate": 0.00017807490803321182, |
| "loss": 0.3968, |
| "step": 1688 |
| }, |
| { |
| "epoch": 0.5875195550147749, |
| "grad_norm": 0.6839443146946156, |
| "learning_rate": 0.00017799900016203293, |
| "loss": 0.5388, |
| "step": 1690 |
| }, |
| { |
| "epoch": 0.5882148444289935, |
| "grad_norm": 1.1726632523971559, |
| "learning_rate": 0.00017792297735376761, |
| "loss": 0.6796, |
| "step": 1692 |
| }, |
| { |
| "epoch": 0.5889101338432122, |
| "grad_norm": 1.1509568442319629, |
| "learning_rate": 0.00017784683972044095, |
| "loss": 0.6486, |
| "step": 1694 |
| }, |
| { |
| "epoch": 0.5896054232574309, |
| "grad_norm": 0.8030934293115218, |
| "learning_rate": 0.00017777058737424714, |
| "loss": 0.5735, |
| "step": 1696 |
| }, |
| { |
| "epoch": 0.5903007126716495, |
| "grad_norm": 0.9804610100362895, |
| "learning_rate": 0.00017769422042754947, |
| "loss": 0.6533, |
| "step": 1698 |
| }, |
| { |
| "epoch": 0.5909960020858682, |
| "grad_norm": 1.8668332123481661, |
| "learning_rate": 0.00017761773899288008, |
| "loss": 0.7231, |
| "step": 1700 |
| }, |
| { |
| "epoch": 0.5916912915000869, |
| "grad_norm": 0.7350290447809635, |
| "learning_rate": 0.00017754114318293982, |
| "loss": 0.7144, |
| "step": 1702 |
| }, |
| { |
| "epoch": 0.5923865809143056, |
| "grad_norm": 0.44898971630294976, |
| "learning_rate": 0.00017746443311059804, |
| "loss": 0.5719, |
| "step": 1704 |
| }, |
| { |
| "epoch": 0.5930818703285242, |
| "grad_norm": 0.5076851350638184, |
| "learning_rate": 0.00017738760888889253, |
| "loss": 0.6627, |
| "step": 1706 |
| }, |
| { |
| "epoch": 0.5937771597427429, |
| "grad_norm": 0.817639077876227, |
| "learning_rate": 0.00017731067063102925, |
| "loss": 0.5764, |
| "step": 1708 |
| }, |
| { |
| "epoch": 0.5944724491569616, |
| "grad_norm": 0.9017538169950253, |
| "learning_rate": 0.0001772336184503822, |
| "loss": 0.5917, |
| "step": 1710 |
| }, |
| { |
| "epoch": 0.5951677385711802, |
| "grad_norm": 0.40946041731193694, |
| "learning_rate": 0.00017715645246049326, |
| "loss": 0.6128, |
| "step": 1712 |
| }, |
| { |
| "epoch": 0.5958630279853989, |
| "grad_norm": 0.5611668605116188, |
| "learning_rate": 0.00017707917277507204, |
| "loss": 0.6213, |
| "step": 1714 |
| }, |
| { |
| "epoch": 0.5965583173996176, |
| "grad_norm": 0.530189165141287, |
| "learning_rate": 0.00017700177950799562, |
| "loss": 0.5389, |
| "step": 1716 |
| }, |
| { |
| "epoch": 0.5972536068138362, |
| "grad_norm": 1.2382772702026799, |
| "learning_rate": 0.00017692427277330857, |
| "loss": 0.5313, |
| "step": 1718 |
| }, |
| { |
| "epoch": 0.5979488962280549, |
| "grad_norm": 0.7009388588402383, |
| "learning_rate": 0.00017684665268522253, |
| "loss": 0.5763, |
| "step": 1720 |
| }, |
| { |
| "epoch": 0.5986441856422736, |
| "grad_norm": 0.4981948622376413, |
| "learning_rate": 0.00017676891935811622, |
| "loss": 0.6326, |
| "step": 1722 |
| }, |
| { |
| "epoch": 0.5993394750564922, |
| "grad_norm": 0.743480238762861, |
| "learning_rate": 0.00017669107290653528, |
| "loss": 0.5565, |
| "step": 1724 |
| }, |
| { |
| "epoch": 0.6000347644707109, |
| "grad_norm": 0.5198052077722481, |
| "learning_rate": 0.000176613113445192, |
| "loss": 0.4672, |
| "step": 1726 |
| }, |
| { |
| "epoch": 0.6007300538849296, |
| "grad_norm": 0.8298155463428107, |
| "learning_rate": 0.00017653504108896517, |
| "loss": 0.5494, |
| "step": 1728 |
| }, |
| { |
| "epoch": 0.6014253432991483, |
| "grad_norm": 0.819830510986772, |
| "learning_rate": 0.0001764568559529, |
| "loss": 0.7491, |
| "step": 1730 |
| }, |
| { |
| "epoch": 0.6021206327133669, |
| "grad_norm": 0.5441657313855216, |
| "learning_rate": 0.00017637855815220784, |
| "loss": 0.6218, |
| "step": 1732 |
| }, |
| { |
| "epoch": 0.6028159221275856, |
| "grad_norm": 0.6531514104491785, |
| "learning_rate": 0.00017630014780226612, |
| "loss": 0.5577, |
| "step": 1734 |
| }, |
| { |
| "epoch": 0.6035112115418043, |
| "grad_norm": 0.7295099022674164, |
| "learning_rate": 0.00017622162501861803, |
| "loss": 0.6255, |
| "step": 1736 |
| }, |
| { |
| "epoch": 0.6042065009560229, |
| "grad_norm": 0.6744185040628968, |
| "learning_rate": 0.0001761429899169725, |
| "loss": 0.6667, |
| "step": 1738 |
| }, |
| { |
| "epoch": 0.6049017903702416, |
| "grad_norm": 0.6781720439136543, |
| "learning_rate": 0.00017606424261320402, |
| "loss": 0.6276, |
| "step": 1740 |
| }, |
| { |
| "epoch": 0.6055970797844603, |
| "grad_norm": 0.5375368781216472, |
| "learning_rate": 0.00017598538322335228, |
| "loss": 0.6052, |
| "step": 1742 |
| }, |
| { |
| "epoch": 0.6062923691986789, |
| "grad_norm": 0.5740302297637102, |
| "learning_rate": 0.00017590641186362223, |
| "loss": 0.6069, |
| "step": 1744 |
| }, |
| { |
| "epoch": 0.6069876586128976, |
| "grad_norm": 0.7486920619809492, |
| "learning_rate": 0.0001758273286503838, |
| "loss": 0.6501, |
| "step": 1746 |
| }, |
| { |
| "epoch": 0.6076829480271163, |
| "grad_norm": 0.415243418739817, |
| "learning_rate": 0.00017574813370017177, |
| "loss": 0.6405, |
| "step": 1748 |
| }, |
| { |
| "epoch": 0.6083782374413349, |
| "grad_norm": 0.33508147043872166, |
| "learning_rate": 0.0001756688271296855, |
| "loss": 0.6736, |
| "step": 1750 |
| }, |
| { |
| "epoch": 0.6090735268555536, |
| "grad_norm": 0.6568565253926139, |
| "learning_rate": 0.0001755894090557889, |
| "loss": 0.5944, |
| "step": 1752 |
| }, |
| { |
| "epoch": 0.6097688162697723, |
| "grad_norm": 0.3600602933243024, |
| "learning_rate": 0.00017550987959551013, |
| "loss": 0.5912, |
| "step": 1754 |
| }, |
| { |
| "epoch": 0.610464105683991, |
| "grad_norm": 0.37379015662929643, |
| "learning_rate": 0.00017543023886604158, |
| "loss": 0.6873, |
| "step": 1756 |
| }, |
| { |
| "epoch": 0.6111593950982096, |
| "grad_norm": 0.4236250679062018, |
| "learning_rate": 0.00017535048698473951, |
| "loss": 0.5878, |
| "step": 1758 |
| }, |
| { |
| "epoch": 0.6118546845124283, |
| "grad_norm": 0.32077385067052405, |
| "learning_rate": 0.00017527062406912398, |
| "loss": 0.5699, |
| "step": 1760 |
| }, |
| { |
| "epoch": 0.612549973926647, |
| "grad_norm": 0.7516616152359099, |
| "learning_rate": 0.0001751906502368787, |
| "loss": 0.5472, |
| "step": 1762 |
| }, |
| { |
| "epoch": 0.6132452633408656, |
| "grad_norm": 0.40617593547014996, |
| "learning_rate": 0.00017511056560585083, |
| "loss": 0.6116, |
| "step": 1764 |
| }, |
| { |
| "epoch": 0.6139405527550843, |
| "grad_norm": 0.6589240069077935, |
| "learning_rate": 0.00017503037029405078, |
| "loss": 0.5217, |
| "step": 1766 |
| }, |
| { |
| "epoch": 0.614635842169303, |
| "grad_norm": 0.45796835520264856, |
| "learning_rate": 0.00017495006441965204, |
| "loss": 0.6046, |
| "step": 1768 |
| }, |
| { |
| "epoch": 0.6153311315835216, |
| "grad_norm": 0.7238836440483356, |
| "learning_rate": 0.00017486964810099106, |
| "loss": 0.5221, |
| "step": 1770 |
| }, |
| { |
| "epoch": 0.6160264209977403, |
| "grad_norm": 0.6979943995723972, |
| "learning_rate": 0.00017478912145656698, |
| "loss": 0.6296, |
| "step": 1772 |
| }, |
| { |
| "epoch": 0.616721710411959, |
| "grad_norm": 0.8707905832977048, |
| "learning_rate": 0.0001747084846050416, |
| "loss": 0.5417, |
| "step": 1774 |
| }, |
| { |
| "epoch": 0.6174169998261776, |
| "grad_norm": 0.7216519551841216, |
| "learning_rate": 0.00017462773766523904, |
| "loss": 0.641, |
| "step": 1776 |
| }, |
| { |
| "epoch": 0.6181122892403963, |
| "grad_norm": 0.661681773522186, |
| "learning_rate": 0.00017454688075614567, |
| "loss": 0.6498, |
| "step": 1778 |
| }, |
| { |
| "epoch": 0.618807578654615, |
| "grad_norm": 0.5732014964202184, |
| "learning_rate": 0.00017446591399690995, |
| "loss": 0.5302, |
| "step": 1780 |
| }, |
| { |
| "epoch": 0.6195028680688337, |
| "grad_norm": 0.8272395967155222, |
| "learning_rate": 0.00017438483750684212, |
| "loss": 0.6592, |
| "step": 1782 |
| }, |
| { |
| "epoch": 0.6201981574830523, |
| "grad_norm": 0.6210836312365995, |
| "learning_rate": 0.00017430365140541415, |
| "loss": 0.6877, |
| "step": 1784 |
| }, |
| { |
| "epoch": 0.620893446897271, |
| "grad_norm": 0.46170985659939656, |
| "learning_rate": 0.00017422235581225962, |
| "loss": 0.6255, |
| "step": 1786 |
| }, |
| { |
| "epoch": 0.6215887363114897, |
| "grad_norm": 0.3883099305940657, |
| "learning_rate": 0.00017414095084717333, |
| "loss": 0.5809, |
| "step": 1788 |
| }, |
| { |
| "epoch": 0.6222840257257083, |
| "grad_norm": 0.43704267209324255, |
| "learning_rate": 0.00017405943663011133, |
| "loss": 0.5123, |
| "step": 1790 |
| }, |
| { |
| "epoch": 0.622979315139927, |
| "grad_norm": 0.3930491987058269, |
| "learning_rate": 0.0001739778132811906, |
| "loss": 0.6012, |
| "step": 1792 |
| }, |
| { |
| "epoch": 0.6236746045541457, |
| "grad_norm": 0.412852657314726, |
| "learning_rate": 0.00017389608092068892, |
| "loss": 0.6531, |
| "step": 1794 |
| }, |
| { |
| "epoch": 0.6243698939683643, |
| "grad_norm": 0.36414563480361445, |
| "learning_rate": 0.00017381423966904486, |
| "loss": 0.5236, |
| "step": 1796 |
| }, |
| { |
| "epoch": 0.625065183382583, |
| "grad_norm": 0.5001274802739109, |
| "learning_rate": 0.00017373228964685726, |
| "loss": 0.579, |
| "step": 1798 |
| }, |
| { |
| "epoch": 0.6257604727968017, |
| "grad_norm": 0.4317988926228409, |
| "learning_rate": 0.00017365023097488534, |
| "loss": 0.5123, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.6257604727968017, |
| "eval_loss": 0.6052089929580688, |
| "eval_runtime": 370.9182, |
| "eval_samples_per_second": 13.054, |
| "eval_steps_per_second": 0.41, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.6264557622110203, |
| "grad_norm": 0.5204270522892355, |
| "learning_rate": 0.00017356806377404844, |
| "loss": 0.653, |
| "step": 1802 |
| }, |
| { |
| "epoch": 0.627151051625239, |
| "grad_norm": 0.7608529412962103, |
| "learning_rate": 0.00017348578816542572, |
| "loss": 0.6875, |
| "step": 1804 |
| }, |
| { |
| "epoch": 0.6278463410394577, |
| "grad_norm": 0.6530765061280092, |
| "learning_rate": 0.00017340340427025614, |
| "loss": 0.573, |
| "step": 1806 |
| }, |
| { |
| "epoch": 0.6285416304536764, |
| "grad_norm": 0.4895237421760521, |
| "learning_rate": 0.0001733209122099383, |
| "loss": 0.6624, |
| "step": 1808 |
| }, |
| { |
| "epoch": 0.629236919867895, |
| "grad_norm": 0.4397644961630881, |
| "learning_rate": 0.00017323831210603016, |
| "loss": 0.5753, |
| "step": 1810 |
| }, |
| { |
| "epoch": 0.6299322092821137, |
| "grad_norm": 0.8716885012986322, |
| "learning_rate": 0.0001731556040802488, |
| "loss": 0.634, |
| "step": 1812 |
| }, |
| { |
| "epoch": 0.6306274986963324, |
| "grad_norm": 0.53660029520689, |
| "learning_rate": 0.0001730727882544704, |
| "loss": 0.547, |
| "step": 1814 |
| }, |
| { |
| "epoch": 0.631322788110551, |
| "grad_norm": 0.6693656512741157, |
| "learning_rate": 0.00017298986475072997, |
| "loss": 0.542, |
| "step": 1816 |
| }, |
| { |
| "epoch": 0.6320180775247697, |
| "grad_norm": 0.6427881239550128, |
| "learning_rate": 0.00017290683369122122, |
| "loss": 0.5778, |
| "step": 1818 |
| }, |
| { |
| "epoch": 0.6327133669389884, |
| "grad_norm": 0.5700659225992518, |
| "learning_rate": 0.00017282369519829635, |
| "loss": 0.6321, |
| "step": 1820 |
| }, |
| { |
| "epoch": 0.633408656353207, |
| "grad_norm": 0.5947857078301825, |
| "learning_rate": 0.00017274044939446584, |
| "loss": 0.6398, |
| "step": 1822 |
| }, |
| { |
| "epoch": 0.6341039457674257, |
| "grad_norm": 0.8875363716928713, |
| "learning_rate": 0.00017265709640239835, |
| "loss": 0.6265, |
| "step": 1824 |
| }, |
| { |
| "epoch": 0.6347992351816444, |
| "grad_norm": 0.5083457538417298, |
| "learning_rate": 0.0001725736363449204, |
| "loss": 0.5039, |
| "step": 1826 |
| }, |
| { |
| "epoch": 0.635494524595863, |
| "grad_norm": 0.7572257750500727, |
| "learning_rate": 0.00017249006934501635, |
| "loss": 0.5638, |
| "step": 1828 |
| }, |
| { |
| "epoch": 0.6361898140100817, |
| "grad_norm": 0.4381329875262636, |
| "learning_rate": 0.0001724063955258281, |
| "loss": 0.5685, |
| "step": 1830 |
| }, |
| { |
| "epoch": 0.6368851034243004, |
| "grad_norm": 0.4769668851895218, |
| "learning_rate": 0.00017232261501065507, |
| "loss": 0.5773, |
| "step": 1832 |
| }, |
| { |
| "epoch": 0.6375803928385191, |
| "grad_norm": 0.6333517381851363, |
| "learning_rate": 0.00017223872792295378, |
| "loss": 0.5593, |
| "step": 1834 |
| }, |
| { |
| "epoch": 0.6382756822527377, |
| "grad_norm": 0.5103399981068355, |
| "learning_rate": 0.0001721547343863378, |
| "loss": 0.5575, |
| "step": 1836 |
| }, |
| { |
| "epoch": 0.6389709716669564, |
| "grad_norm": 0.6125228317995195, |
| "learning_rate": 0.00017207063452457763, |
| "loss": 0.565, |
| "step": 1838 |
| }, |
| { |
| "epoch": 0.6396662610811751, |
| "grad_norm": 0.7299577367692995, |
| "learning_rate": 0.0001719864284616004, |
| "loss": 0.7102, |
| "step": 1840 |
| }, |
| { |
| "epoch": 0.6403615504953937, |
| "grad_norm": 0.8462870412150455, |
| "learning_rate": 0.00017190211632148978, |
| "loss": 0.5988, |
| "step": 1842 |
| }, |
| { |
| "epoch": 0.6410568399096124, |
| "grad_norm": 0.6376476795268616, |
| "learning_rate": 0.00017181769822848563, |
| "loss": 0.6031, |
| "step": 1844 |
| }, |
| { |
| "epoch": 0.6417521293238311, |
| "grad_norm": 0.5498259619130056, |
| "learning_rate": 0.00017173317430698413, |
| "loss": 0.604, |
| "step": 1846 |
| }, |
| { |
| "epoch": 0.6424474187380497, |
| "grad_norm": 0.6402197493873277, |
| "learning_rate": 0.00017164854468153723, |
| "loss": 0.5663, |
| "step": 1848 |
| }, |
| { |
| "epoch": 0.6431427081522684, |
| "grad_norm": 0.49473627938254194, |
| "learning_rate": 0.00017156380947685275, |
| "loss": 0.4801, |
| "step": 1850 |
| }, |
| { |
| "epoch": 0.6438379975664871, |
| "grad_norm": 0.5864701801475597, |
| "learning_rate": 0.00017147896881779403, |
| "loss": 0.5787, |
| "step": 1852 |
| }, |
| { |
| "epoch": 0.6445332869807057, |
| "grad_norm": 0.9424086488351346, |
| "learning_rate": 0.00017139402282937986, |
| "loss": 0.6761, |
| "step": 1854 |
| }, |
| { |
| "epoch": 0.6452285763949244, |
| "grad_norm": 0.5754580414862842, |
| "learning_rate": 0.0001713089716367842, |
| "loss": 0.5768, |
| "step": 1856 |
| }, |
| { |
| "epoch": 0.6459238658091431, |
| "grad_norm": 0.6945657719528112, |
| "learning_rate": 0.000171223815365336, |
| "loss": 0.6376, |
| "step": 1858 |
| }, |
| { |
| "epoch": 0.6466191552233618, |
| "grad_norm": 0.5299829269045951, |
| "learning_rate": 0.00017113855414051918, |
| "loss": 0.6682, |
| "step": 1860 |
| }, |
| { |
| "epoch": 0.6473144446375804, |
| "grad_norm": 0.6178382762886556, |
| "learning_rate": 0.00017105318808797215, |
| "loss": 0.5745, |
| "step": 1862 |
| }, |
| { |
| "epoch": 0.6480097340517991, |
| "grad_norm": 0.5777288231745069, |
| "learning_rate": 0.00017096771733348794, |
| "loss": 0.4917, |
| "step": 1864 |
| }, |
| { |
| "epoch": 0.6487050234660178, |
| "grad_norm": 0.4060689941848452, |
| "learning_rate": 0.00017088214200301373, |
| "loss": 0.5554, |
| "step": 1866 |
| }, |
| { |
| "epoch": 0.6494003128802364, |
| "grad_norm": 0.35541750930109944, |
| "learning_rate": 0.0001707964622226509, |
| "loss": 0.6036, |
| "step": 1868 |
| }, |
| { |
| "epoch": 0.6500956022944551, |
| "grad_norm": 0.497757325255477, |
| "learning_rate": 0.00017071067811865476, |
| "loss": 0.5215, |
| "step": 1870 |
| }, |
| { |
| "epoch": 0.6507908917086738, |
| "grad_norm": 1.155534242094426, |
| "learning_rate": 0.00017062478981743424, |
| "loss": 0.6362, |
| "step": 1872 |
| }, |
| { |
| "epoch": 0.6514861811228924, |
| "grad_norm": 0.5222241972069097, |
| "learning_rate": 0.0001705387974455519, |
| "loss": 0.593, |
| "step": 1874 |
| }, |
| { |
| "epoch": 0.6521814705371111, |
| "grad_norm": 0.5041434419525581, |
| "learning_rate": 0.00017045270112972362, |
| "loss": 0.6426, |
| "step": 1876 |
| }, |
| { |
| "epoch": 0.6528767599513298, |
| "grad_norm": 0.7687464251703052, |
| "learning_rate": 0.00017036650099681847, |
| "loss": 0.5629, |
| "step": 1878 |
| }, |
| { |
| "epoch": 0.6535720493655484, |
| "grad_norm": 0.6523439997352839, |
| "learning_rate": 0.0001702801971738585, |
| "loss": 0.6066, |
| "step": 1880 |
| }, |
| { |
| "epoch": 0.6542673387797671, |
| "grad_norm": 0.7148868461231821, |
| "learning_rate": 0.0001701937897880185, |
| "loss": 0.6964, |
| "step": 1882 |
| }, |
| { |
| "epoch": 0.6549626281939858, |
| "grad_norm": 0.5711926468750704, |
| "learning_rate": 0.00017010727896662594, |
| "loss": 0.6143, |
| "step": 1884 |
| }, |
| { |
| "epoch": 0.6556579176082045, |
| "grad_norm": 0.5433509637373257, |
| "learning_rate": 0.00017002066483716067, |
| "loss": 0.5491, |
| "step": 1886 |
| }, |
| { |
| "epoch": 0.6563532070224231, |
| "grad_norm": 0.5034407658517631, |
| "learning_rate": 0.00016993394752725483, |
| "loss": 0.5197, |
| "step": 1888 |
| }, |
| { |
| "epoch": 0.6570484964366418, |
| "grad_norm": 0.8373667811859606, |
| "learning_rate": 0.0001698471271646925, |
| "loss": 0.5873, |
| "step": 1890 |
| }, |
| { |
| "epoch": 0.6577437858508605, |
| "grad_norm": 0.6227498724251839, |
| "learning_rate": 0.0001697602038774097, |
| "loss": 0.6962, |
| "step": 1892 |
| }, |
| { |
| "epoch": 0.658439075265079, |
| "grad_norm": 1.2351405074494481, |
| "learning_rate": 0.00016967317779349408, |
| "loss": 0.5342, |
| "step": 1894 |
| }, |
| { |
| "epoch": 0.6591343646792978, |
| "grad_norm": 0.5467600967193789, |
| "learning_rate": 0.00016958604904118477, |
| "loss": 0.5608, |
| "step": 1896 |
| }, |
| { |
| "epoch": 0.6598296540935165, |
| "grad_norm": 0.5963111593389001, |
| "learning_rate": 0.00016949881774887225, |
| "loss": 0.517, |
| "step": 1898 |
| }, |
| { |
| "epoch": 0.660524943507735, |
| "grad_norm": 0.7374594651213904, |
| "learning_rate": 0.00016941148404509796, |
| "loss": 0.4725, |
| "step": 1900 |
| }, |
| { |
| "epoch": 0.6612202329219538, |
| "grad_norm": 0.9472414514119026, |
| "learning_rate": 0.00016932404805855442, |
| "loss": 0.5568, |
| "step": 1902 |
| }, |
| { |
| "epoch": 0.6619155223361725, |
| "grad_norm": 1.0041060773211288, |
| "learning_rate": 0.00016923650991808472, |
| "loss": 0.603, |
| "step": 1904 |
| }, |
| { |
| "epoch": 0.662610811750391, |
| "grad_norm": 1.0580932026507945, |
| "learning_rate": 0.00016914886975268257, |
| "loss": 0.6036, |
| "step": 1906 |
| }, |
| { |
| "epoch": 0.6633061011646098, |
| "grad_norm": 0.8767482713317837, |
| "learning_rate": 0.00016906112769149204, |
| "loss": 0.5834, |
| "step": 1908 |
| }, |
| { |
| "epoch": 0.6640013905788285, |
| "grad_norm": 1.012539952053466, |
| "learning_rate": 0.00016897328386380732, |
| "loss": 0.6149, |
| "step": 1910 |
| }, |
| { |
| "epoch": 0.6646966799930472, |
| "grad_norm": 0.6627172772277559, |
| "learning_rate": 0.00016888533839907246, |
| "loss": 0.5652, |
| "step": 1912 |
| }, |
| { |
| "epoch": 0.6653919694072657, |
| "grad_norm": 0.5784691725846585, |
| "learning_rate": 0.00016879729142688142, |
| "loss": 0.5892, |
| "step": 1914 |
| }, |
| { |
| "epoch": 0.6660872588214845, |
| "grad_norm": 0.5351447888394464, |
| "learning_rate": 0.00016870914307697774, |
| "loss": 0.5569, |
| "step": 1916 |
| }, |
| { |
| "epoch": 0.6667825482357032, |
| "grad_norm": 0.6075480060783698, |
| "learning_rate": 0.00016862089347925423, |
| "loss": 0.5714, |
| "step": 1918 |
| }, |
| { |
| "epoch": 0.6674778376499217, |
| "grad_norm": 0.7061964040177015, |
| "learning_rate": 0.00016853254276375296, |
| "loss": 0.6222, |
| "step": 1920 |
| }, |
| { |
| "epoch": 0.6681731270641404, |
| "grad_norm": 0.7493781087201877, |
| "learning_rate": 0.00016844409106066505, |
| "loss": 0.5536, |
| "step": 1922 |
| }, |
| { |
| "epoch": 0.6688684164783592, |
| "grad_norm": 0.6889099042823535, |
| "learning_rate": 0.0001683555385003304, |
| "loss": 0.5381, |
| "step": 1924 |
| }, |
| { |
| "epoch": 0.6695637058925777, |
| "grad_norm": 0.686174384092237, |
| "learning_rate": 0.00016826688521323746, |
| "loss": 0.5206, |
| "step": 1926 |
| }, |
| { |
| "epoch": 0.6702589953067964, |
| "grad_norm": 0.6685983626439217, |
| "learning_rate": 0.0001681781313300232, |
| "loss": 0.619, |
| "step": 1928 |
| }, |
| { |
| "epoch": 0.6709542847210151, |
| "grad_norm": 0.6684510056843382, |
| "learning_rate": 0.0001680892769814728, |
| "loss": 0.5298, |
| "step": 1930 |
| }, |
| { |
| "epoch": 0.6716495741352337, |
| "grad_norm": 0.629367010591366, |
| "learning_rate": 0.00016800032229851943, |
| "loss": 0.556, |
| "step": 1932 |
| }, |
| { |
| "epoch": 0.6723448635494524, |
| "grad_norm": 0.6511793907808336, |
| "learning_rate": 0.0001679112674122442, |
| "loss": 0.5667, |
| "step": 1934 |
| }, |
| { |
| "epoch": 0.6730401529636711, |
| "grad_norm": 0.8544347958530318, |
| "learning_rate": 0.00016782211245387586, |
| "loss": 0.541, |
| "step": 1936 |
| }, |
| { |
| "epoch": 0.6737354423778898, |
| "grad_norm": 0.6862974328397455, |
| "learning_rate": 0.00016773285755479055, |
| "loss": 0.5718, |
| "step": 1938 |
| }, |
| { |
| "epoch": 0.6744307317921084, |
| "grad_norm": 0.9064411135544665, |
| "learning_rate": 0.00016764350284651172, |
| "loss": 0.5383, |
| "step": 1940 |
| }, |
| { |
| "epoch": 0.6751260212063271, |
| "grad_norm": 1.2798081471760363, |
| "learning_rate": 0.00016755404846070993, |
| "loss": 0.6409, |
| "step": 1942 |
| }, |
| { |
| "epoch": 0.6758213106205458, |
| "grad_norm": 0.7031747774944076, |
| "learning_rate": 0.0001674644945292026, |
| "loss": 0.6129, |
| "step": 1944 |
| }, |
| { |
| "epoch": 0.6765166000347644, |
| "grad_norm": 0.8672546418037201, |
| "learning_rate": 0.0001673748411839538, |
| "loss": 0.5397, |
| "step": 1946 |
| }, |
| { |
| "epoch": 0.6772118894489831, |
| "grad_norm": 1.3071689607309334, |
| "learning_rate": 0.00016728508855707412, |
| "loss": 0.6809, |
| "step": 1948 |
| }, |
| { |
| "epoch": 0.6779071788632018, |
| "grad_norm": 0.43492321965313424, |
| "learning_rate": 0.00016719523678082048, |
| "loss": 0.5308, |
| "step": 1950 |
| }, |
| { |
| "epoch": 0.6779071788632018, |
| "eval_loss": 0.6056139469146729, |
| "eval_runtime": 371.18, |
| "eval_samples_per_second": 13.045, |
| "eval_steps_per_second": 0.41, |
| "step": 1950 |
| }, |
| { |
| "epoch": 0.6786024682774204, |
| "grad_norm": 0.6569772769573995, |
| "learning_rate": 0.00016710528598759588, |
| "loss": 0.6151, |
| "step": 1952 |
| }, |
| { |
| "epoch": 0.6792977576916391, |
| "grad_norm": 0.6233123597755471, |
| "learning_rate": 0.00016701523630994921, |
| "loss": 0.6345, |
| "step": 1954 |
| }, |
| { |
| "epoch": 0.6799930471058578, |
| "grad_norm": 0.756483194755832, |
| "learning_rate": 0.00016692508788057508, |
| "loss": 0.6423, |
| "step": 1956 |
| }, |
| { |
| "epoch": 0.6806883365200764, |
| "grad_norm": 0.5012190330749898, |
| "learning_rate": 0.00016683484083231367, |
| "loss": 0.5771, |
| "step": 1958 |
| }, |
| { |
| "epoch": 0.6813836259342951, |
| "grad_norm": 0.45697084291469786, |
| "learning_rate": 0.00016674449529815037, |
| "loss": 0.6064, |
| "step": 1960 |
| }, |
| { |
| "epoch": 0.6820789153485138, |
| "grad_norm": 0.6839659288985819, |
| "learning_rate": 0.0001666540514112158, |
| "loss": 0.6019, |
| "step": 1962 |
| }, |
| { |
| "epoch": 0.6827742047627325, |
| "grad_norm": 0.6790048429974062, |
| "learning_rate": 0.0001665635093047855, |
| "loss": 0.5919, |
| "step": 1964 |
| }, |
| { |
| "epoch": 0.6834694941769511, |
| "grad_norm": 0.7067958138632173, |
| "learning_rate": 0.00016647286911227968, |
| "loss": 0.5737, |
| "step": 1966 |
| }, |
| { |
| "epoch": 0.6841647835911698, |
| "grad_norm": 0.6737064355710938, |
| "learning_rate": 0.00016638213096726312, |
| "loss": 0.6315, |
| "step": 1968 |
| }, |
| { |
| "epoch": 0.6848600730053885, |
| "grad_norm": 0.39284397968929874, |
| "learning_rate": 0.000166291295003445, |
| "loss": 0.5519, |
| "step": 1970 |
| }, |
| { |
| "epoch": 0.6855553624196071, |
| "grad_norm": 0.6110036966138074, |
| "learning_rate": 0.00016620036135467855, |
| "loss": 0.5699, |
| "step": 1972 |
| }, |
| { |
| "epoch": 0.6862506518338258, |
| "grad_norm": 0.674185155315815, |
| "learning_rate": 0.0001661093301549609, |
| "loss": 0.6632, |
| "step": 1974 |
| }, |
| { |
| "epoch": 0.6869459412480445, |
| "grad_norm": 1.34627131454609, |
| "learning_rate": 0.0001660182015384332, |
| "loss": 0.6096, |
| "step": 1976 |
| }, |
| { |
| "epoch": 0.6876412306622631, |
| "grad_norm": 0.644484450399573, |
| "learning_rate": 0.0001659269756393798, |
| "loss": 0.5991, |
| "step": 1978 |
| }, |
| { |
| "epoch": 0.6883365200764818, |
| "grad_norm": 0.4990172041361588, |
| "learning_rate": 0.00016583565259222863, |
| "loss": 0.5416, |
| "step": 1980 |
| }, |
| { |
| "epoch": 0.6890318094907005, |
| "grad_norm": 0.4505095876483144, |
| "learning_rate": 0.0001657442325315507, |
| "loss": 0.5676, |
| "step": 1982 |
| }, |
| { |
| "epoch": 0.6897270989049191, |
| "grad_norm": 0.46794666181349276, |
| "learning_rate": 0.00016565271559205997, |
| "loss": 0.5, |
| "step": 1984 |
| }, |
| { |
| "epoch": 0.6904223883191378, |
| "grad_norm": 0.4544548690857851, |
| "learning_rate": 0.0001655611019086132, |
| "loss": 0.4838, |
| "step": 1986 |
| }, |
| { |
| "epoch": 0.6911176777333565, |
| "grad_norm": 1.3379424680615957, |
| "learning_rate": 0.00016546939161620966, |
| "loss": 0.609, |
| "step": 1988 |
| }, |
| { |
| "epoch": 0.6918129671475752, |
| "grad_norm": 0.717144004391169, |
| "learning_rate": 0.00016537758484999105, |
| "loss": 0.587, |
| "step": 1990 |
| }, |
| { |
| "epoch": 0.6925082565617938, |
| "grad_norm": 0.8519131395212922, |
| "learning_rate": 0.0001652856817452411, |
| "loss": 0.6149, |
| "step": 1992 |
| }, |
| { |
| "epoch": 0.6932035459760125, |
| "grad_norm": 0.8247198494798238, |
| "learning_rate": 0.00016519368243738566, |
| "loss": 0.5559, |
| "step": 1994 |
| }, |
| { |
| "epoch": 0.6938988353902312, |
| "grad_norm": 0.9972243594284758, |
| "learning_rate": 0.00016510158706199222, |
| "loss": 0.7695, |
| "step": 1996 |
| }, |
| { |
| "epoch": 0.6945941248044498, |
| "grad_norm": 0.6775709255460454, |
| "learning_rate": 0.00016500939575476992, |
| "loss": 0.5546, |
| "step": 1998 |
| }, |
| { |
| "epoch": 0.6952894142186685, |
| "grad_norm": 0.6270536790838611, |
| "learning_rate": 0.0001649171086515692, |
| "loss": 0.6676, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.6959847036328872, |
| "grad_norm": 0.5247284479563707, |
| "learning_rate": 0.0001648247258883817, |
| "loss": 0.6573, |
| "step": 2002 |
| }, |
| { |
| "epoch": 0.6966799930471058, |
| "grad_norm": 1.3371640454637324, |
| "learning_rate": 0.00016473224760134, |
| "loss": 0.5688, |
| "step": 2004 |
| }, |
| { |
| "epoch": 0.6973752824613245, |
| "grad_norm": 0.537239749385301, |
| "learning_rate": 0.0001646396739267174, |
| "loss": 0.5238, |
| "step": 2006 |
| }, |
| { |
| "epoch": 0.6980705718755432, |
| "grad_norm": 0.683351825211432, |
| "learning_rate": 0.00016454700500092786, |
| "loss": 0.576, |
| "step": 2008 |
| }, |
| { |
| "epoch": 0.6987658612897618, |
| "grad_norm": 1.266816676228039, |
| "learning_rate": 0.00016445424096052567, |
| "loss": 0.592, |
| "step": 2010 |
| }, |
| { |
| "epoch": 0.6994611507039805, |
| "grad_norm": 0.8922765874170949, |
| "learning_rate": 0.00016436138194220518, |
| "loss": 0.5547, |
| "step": 2012 |
| }, |
| { |
| "epoch": 0.7001564401181992, |
| "grad_norm": 0.602053244835111, |
| "learning_rate": 0.00016426842808280086, |
| "loss": 0.6122, |
| "step": 2014 |
| }, |
| { |
| "epoch": 0.7008517295324179, |
| "grad_norm": 0.7130195810195582, |
| "learning_rate": 0.00016417537951928676, |
| "loss": 0.5297, |
| "step": 2016 |
| }, |
| { |
| "epoch": 0.7015470189466365, |
| "grad_norm": 0.8636165886047156, |
| "learning_rate": 0.00016408223638877665, |
| "loss": 0.5752, |
| "step": 2018 |
| }, |
| { |
| "epoch": 0.7022423083608552, |
| "grad_norm": 0.7876973687914621, |
| "learning_rate": 0.00016398899882852353, |
| "loss": 0.6652, |
| "step": 2020 |
| }, |
| { |
| "epoch": 0.7029375977750739, |
| "grad_norm": 0.6531287665955688, |
| "learning_rate": 0.0001638956669759196, |
| "loss": 0.6521, |
| "step": 2022 |
| }, |
| { |
| "epoch": 0.7036328871892925, |
| "grad_norm": 0.8086172007984764, |
| "learning_rate": 0.00016380224096849597, |
| "loss": 0.5881, |
| "step": 2024 |
| }, |
| { |
| "epoch": 0.7043281766035112, |
| "grad_norm": 0.47789458830184245, |
| "learning_rate": 0.00016370872094392255, |
| "loss": 0.5258, |
| "step": 2026 |
| }, |
| { |
| "epoch": 0.7050234660177299, |
| "grad_norm": 0.5867593662603546, |
| "learning_rate": 0.00016361510704000778, |
| "loss": 0.5487, |
| "step": 2028 |
| }, |
| { |
| "epoch": 0.7057187554319485, |
| "grad_norm": 0.743493145379613, |
| "learning_rate": 0.00016352139939469838, |
| "loss": 0.6646, |
| "step": 2030 |
| }, |
| { |
| "epoch": 0.7064140448461672, |
| "grad_norm": 0.5420751844584678, |
| "learning_rate": 0.00016342759814607928, |
| "loss": 0.6296, |
| "step": 2032 |
| }, |
| { |
| "epoch": 0.7071093342603859, |
| "grad_norm": 0.4767527091363965, |
| "learning_rate": 0.00016333370343237324, |
| "loss": 0.6135, |
| "step": 2034 |
| }, |
| { |
| "epoch": 0.7078046236746045, |
| "grad_norm": 0.7559831165564748, |
| "learning_rate": 0.0001632397153919409, |
| "loss": 0.6138, |
| "step": 2036 |
| }, |
| { |
| "epoch": 0.7084999130888232, |
| "grad_norm": 0.442317218000476, |
| "learning_rate": 0.0001631456341632803, |
| "loss": 0.5502, |
| "step": 2038 |
| }, |
| { |
| "epoch": 0.7091952025030419, |
| "grad_norm": 0.9012866177348007, |
| "learning_rate": 0.00016305145988502684, |
| "loss": 0.5459, |
| "step": 2040 |
| }, |
| { |
| "epoch": 0.7098904919172606, |
| "grad_norm": 0.7045775124035536, |
| "learning_rate": 0.000162957192695953, |
| "loss": 0.4986, |
| "step": 2042 |
| }, |
| { |
| "epoch": 0.7105857813314792, |
| "grad_norm": 0.5788828574335951, |
| "learning_rate": 0.00016286283273496825, |
| "loss": 0.5706, |
| "step": 2044 |
| }, |
| { |
| "epoch": 0.7112810707456979, |
| "grad_norm": 0.7709837646633183, |
| "learning_rate": 0.00016276838014111868, |
| "loss": 0.6708, |
| "step": 2046 |
| }, |
| { |
| "epoch": 0.7119763601599166, |
| "grad_norm": 0.9894566046216766, |
| "learning_rate": 0.00016267383505358694, |
| "loss": 0.4963, |
| "step": 2048 |
| }, |
| { |
| "epoch": 0.7126716495741352, |
| "grad_norm": 0.9904706283602799, |
| "learning_rate": 0.00016257919761169195, |
| "loss": 0.6312, |
| "step": 2050 |
| }, |
| { |
| "epoch": 0.7133669389883539, |
| "grad_norm": 1.2851758328250367, |
| "learning_rate": 0.00016248446795488874, |
| "loss": 0.55, |
| "step": 2052 |
| }, |
| { |
| "epoch": 0.7140622284025726, |
| "grad_norm": 1.280763519444461, |
| "learning_rate": 0.00016238964622276817, |
| "loss": 0.6931, |
| "step": 2054 |
| }, |
| { |
| "epoch": 0.7147575178167912, |
| "grad_norm": 1.4273656641407675, |
| "learning_rate": 0.00016229473255505693, |
| "loss": 0.5302, |
| "step": 2056 |
| }, |
| { |
| "epoch": 0.7154528072310099, |
| "grad_norm": 0.6347338138187268, |
| "learning_rate": 0.0001621997270916169, |
| "loss": 0.5815, |
| "step": 2058 |
| }, |
| { |
| "epoch": 0.7161480966452286, |
| "grad_norm": 0.6533886960783342, |
| "learning_rate": 0.00016210462997244554, |
| "loss": 0.6818, |
| "step": 2060 |
| }, |
| { |
| "epoch": 0.7168433860594472, |
| "grad_norm": 0.5939404538830146, |
| "learning_rate": 0.0001620094413376752, |
| "loss": 0.5961, |
| "step": 2062 |
| }, |
| { |
| "epoch": 0.7175386754736659, |
| "grad_norm": 1.0489562805261028, |
| "learning_rate": 0.00016191416132757305, |
| "loss": 0.5698, |
| "step": 2064 |
| }, |
| { |
| "epoch": 0.7182339648878846, |
| "grad_norm": 0.45667736793871694, |
| "learning_rate": 0.00016181879008254113, |
| "loss": 0.5632, |
| "step": 2066 |
| }, |
| { |
| "epoch": 0.7189292543021033, |
| "grad_norm": 0.6278641019713552, |
| "learning_rate": 0.00016172332774311562, |
| "loss": 0.6647, |
| "step": 2068 |
| }, |
| { |
| "epoch": 0.7196245437163219, |
| "grad_norm": 0.47629081990627226, |
| "learning_rate": 0.00016162777444996712, |
| "loss": 0.6373, |
| "step": 2070 |
| }, |
| { |
| "epoch": 0.7203198331305406, |
| "grad_norm": 0.6149498576605736, |
| "learning_rate": 0.00016153213034390026, |
| "loss": 0.5118, |
| "step": 2072 |
| }, |
| { |
| "epoch": 0.7210151225447593, |
| "grad_norm": 0.56362872937635, |
| "learning_rate": 0.00016143639556585342, |
| "loss": 0.5284, |
| "step": 2074 |
| }, |
| { |
| "epoch": 0.7217104119589779, |
| "grad_norm": 0.45928718600771984, |
| "learning_rate": 0.00016134057025689858, |
| "loss": 0.5591, |
| "step": 2076 |
| }, |
| { |
| "epoch": 0.7224057013731966, |
| "grad_norm": 0.9311134574972124, |
| "learning_rate": 0.00016124465455824127, |
| "loss": 0.6373, |
| "step": 2078 |
| }, |
| { |
| "epoch": 0.7231009907874153, |
| "grad_norm": 0.8117606338423109, |
| "learning_rate": 0.00016114864861122, |
| "loss": 0.6172, |
| "step": 2080 |
| }, |
| { |
| "epoch": 0.7237962802016339, |
| "grad_norm": 0.9415924380727096, |
| "learning_rate": 0.00016105255255730643, |
| "loss": 0.651, |
| "step": 2082 |
| }, |
| { |
| "epoch": 0.7244915696158526, |
| "grad_norm": 1.1201310609340696, |
| "learning_rate": 0.00016095636653810493, |
| "loss": 0.501, |
| "step": 2084 |
| }, |
| { |
| "epoch": 0.7251868590300713, |
| "grad_norm": 0.7532957479811233, |
| "learning_rate": 0.00016086009069535243, |
| "loss": 0.6211, |
| "step": 2086 |
| }, |
| { |
| "epoch": 0.7258821484442899, |
| "grad_norm": 0.4550840645762033, |
| "learning_rate": 0.00016076372517091831, |
| "loss": 0.5765, |
| "step": 2088 |
| }, |
| { |
| "epoch": 0.7265774378585086, |
| "grad_norm": 0.5721008820271746, |
| "learning_rate": 0.000160667270106804, |
| "loss": 0.5022, |
| "step": 2090 |
| }, |
| { |
| "epoch": 0.7272727272727273, |
| "grad_norm": 0.8683128387128771, |
| "learning_rate": 0.00016057072564514292, |
| "loss": 0.6232, |
| "step": 2092 |
| }, |
| { |
| "epoch": 0.727968016686946, |
| "grad_norm": 0.6494718738069951, |
| "learning_rate": 0.00016047409192820018, |
| "loss": 0.5538, |
| "step": 2094 |
| }, |
| { |
| "epoch": 0.7286633061011646, |
| "grad_norm": 1.1287879234892022, |
| "learning_rate": 0.00016037736909837254, |
| "loss": 0.6134, |
| "step": 2096 |
| }, |
| { |
| "epoch": 0.7293585955153833, |
| "grad_norm": 0.852132328455506, |
| "learning_rate": 0.00016028055729818785, |
| "loss": 0.599, |
| "step": 2098 |
| }, |
| { |
| "epoch": 0.730053884929602, |
| "grad_norm": 0.8301319013187483, |
| "learning_rate": 0.00016018365667030534, |
| "loss": 0.5511, |
| "step": 2100 |
| }, |
| { |
| "epoch": 0.730053884929602, |
| "eval_loss": 0.6051976680755615, |
| "eval_runtime": 370.8053, |
| "eval_samples_per_second": 13.058, |
| "eval_steps_per_second": 0.41, |
| "step": 2100 |
| }, |
| { |
| "epoch": 0.7307491743438206, |
| "grad_norm": 0.835223857619091, |
| "learning_rate": 0.00016008666735751488, |
| "loss": 0.5362, |
| "step": 2102 |
| }, |
| { |
| "epoch": 0.7314444637580393, |
| "grad_norm": 0.5859220724175553, |
| "learning_rate": 0.00015998958950273718, |
| "loss": 0.4207, |
| "step": 2104 |
| }, |
| { |
| "epoch": 0.732139753172258, |
| "grad_norm": 0.8195876868296976, |
| "learning_rate": 0.00015989242324902341, |
| "loss": 0.6456, |
| "step": 2106 |
| }, |
| { |
| "epoch": 0.7328350425864766, |
| "grad_norm": 0.5963426019757224, |
| "learning_rate": 0.00015979516873955495, |
| "loss": 0.5347, |
| "step": 2108 |
| }, |
| { |
| "epoch": 0.7335303320006953, |
| "grad_norm": 0.5598552464839238, |
| "learning_rate": 0.00015969782611764322, |
| "loss": 0.5609, |
| "step": 2110 |
| }, |
| { |
| "epoch": 0.734225621414914, |
| "grad_norm": 0.9190818791386364, |
| "learning_rate": 0.0001596003955267295, |
| "loss": 0.6444, |
| "step": 2112 |
| }, |
| { |
| "epoch": 0.7349209108291326, |
| "grad_norm": 0.7074763887519752, |
| "learning_rate": 0.00015950287711038482, |
| "loss": 0.5513, |
| "step": 2114 |
| }, |
| { |
| "epoch": 0.7356162002433513, |
| "grad_norm": 0.5303381980872908, |
| "learning_rate": 0.00015940527101230938, |
| "loss": 0.5649, |
| "step": 2116 |
| }, |
| { |
| "epoch": 0.73631148965757, |
| "grad_norm": 0.7961849958708207, |
| "learning_rate": 0.0001593075773763328, |
| "loss": 0.5032, |
| "step": 2118 |
| }, |
| { |
| "epoch": 0.7370067790717887, |
| "grad_norm": 0.6321864563119038, |
| "learning_rate": 0.00015920979634641364, |
| "loss": 0.5128, |
| "step": 2120 |
| }, |
| { |
| "epoch": 0.7377020684860073, |
| "grad_norm": 0.6402923888426749, |
| "learning_rate": 0.00015911192806663914, |
| "loss": 0.6083, |
| "step": 2122 |
| }, |
| { |
| "epoch": 0.738397357900226, |
| "grad_norm": 1.0036392568071666, |
| "learning_rate": 0.00015901397268122525, |
| "loss": 0.5654, |
| "step": 2124 |
| }, |
| { |
| "epoch": 0.7390926473144447, |
| "grad_norm": 0.639210671172982, |
| "learning_rate": 0.00015891593033451616, |
| "loss": 0.5165, |
| "step": 2126 |
| }, |
| { |
| "epoch": 0.7397879367286633, |
| "grad_norm": 0.5362341486809139, |
| "learning_rate": 0.00015881780117098427, |
| "loss": 0.6226, |
| "step": 2128 |
| }, |
| { |
| "epoch": 0.740483226142882, |
| "grad_norm": 0.5732167979609127, |
| "learning_rate": 0.00015871958533522995, |
| "loss": 0.5074, |
| "step": 2130 |
| }, |
| { |
| "epoch": 0.7411785155571007, |
| "grad_norm": 0.5175334271404235, |
| "learning_rate": 0.00015862128297198117, |
| "loss": 0.6992, |
| "step": 2132 |
| }, |
| { |
| "epoch": 0.7418738049713193, |
| "grad_norm": 0.6574502026701543, |
| "learning_rate": 0.0001585228942260935, |
| "loss": 0.5352, |
| "step": 2134 |
| }, |
| { |
| "epoch": 0.742569094385538, |
| "grad_norm": 0.6282329227137664, |
| "learning_rate": 0.00015842441924254975, |
| "loss": 0.5883, |
| "step": 2136 |
| }, |
| { |
| "epoch": 0.7432643837997567, |
| "grad_norm": 0.612347648551905, |
| "learning_rate": 0.00015832585816645986, |
| "loss": 0.6135, |
| "step": 2138 |
| }, |
| { |
| "epoch": 0.7439596732139753, |
| "grad_norm": 0.4562291436078377, |
| "learning_rate": 0.00015822721114306057, |
| "loss": 0.5953, |
| "step": 2140 |
| }, |
| { |
| "epoch": 0.744654962628194, |
| "grad_norm": 0.7370092367487316, |
| "learning_rate": 0.00015812847831771533, |
| "loss": 0.5933, |
| "step": 2142 |
| }, |
| { |
| "epoch": 0.7453502520424127, |
| "grad_norm": 0.734465754561315, |
| "learning_rate": 0.00015802965983591398, |
| "loss": 0.5499, |
| "step": 2144 |
| }, |
| { |
| "epoch": 0.7460455414566314, |
| "grad_norm": 0.5173035061957143, |
| "learning_rate": 0.0001579307558432726, |
| "loss": 0.5704, |
| "step": 2146 |
| }, |
| { |
| "epoch": 0.74674083087085, |
| "grad_norm": 0.42722584908793315, |
| "learning_rate": 0.00015783176648553323, |
| "loss": 0.5696, |
| "step": 2148 |
| }, |
| { |
| "epoch": 0.7474361202850687, |
| "grad_norm": 0.5700444049601131, |
| "learning_rate": 0.00015773269190856378, |
| "loss": 0.5371, |
| "step": 2150 |
| }, |
| { |
| "epoch": 0.7481314096992874, |
| "grad_norm": 0.5135665000835338, |
| "learning_rate": 0.00015763353225835776, |
| "loss": 0.5928, |
| "step": 2152 |
| }, |
| { |
| "epoch": 0.748826699113506, |
| "grad_norm": 0.9427562655572416, |
| "learning_rate": 0.00015753428768103386, |
| "loss": 0.5577, |
| "step": 2154 |
| }, |
| { |
| "epoch": 0.7495219885277247, |
| "grad_norm": 1.1255169215867615, |
| "learning_rate": 0.00015743495832283615, |
| "loss": 0.5936, |
| "step": 2156 |
| }, |
| { |
| "epoch": 0.7502172779419434, |
| "grad_norm": 0.9053073535451409, |
| "learning_rate": 0.00015733554433013347, |
| "loss": 0.5747, |
| "step": 2158 |
| }, |
| { |
| "epoch": 0.750912567356162, |
| "grad_norm": 0.5862209644479233, |
| "learning_rate": 0.0001572360458494194, |
| "loss": 0.5121, |
| "step": 2160 |
| }, |
| { |
| "epoch": 0.7516078567703807, |
| "grad_norm": 0.6727420322046515, |
| "learning_rate": 0.0001571364630273121, |
| "loss": 0.5634, |
| "step": 2162 |
| }, |
| { |
| "epoch": 0.7523031461845994, |
| "grad_norm": 1.1330927585243096, |
| "learning_rate": 0.00015703679601055393, |
| "loss": 0.5105, |
| "step": 2164 |
| }, |
| { |
| "epoch": 0.752998435598818, |
| "grad_norm": 0.9369102364760077, |
| "learning_rate": 0.0001569370449460113, |
| "loss": 0.61, |
| "step": 2166 |
| }, |
| { |
| "epoch": 0.7536937250130367, |
| "grad_norm": 0.7095814453472946, |
| "learning_rate": 0.00015683720998067462, |
| "loss": 0.4921, |
| "step": 2168 |
| }, |
| { |
| "epoch": 0.7543890144272554, |
| "grad_norm": 0.8737265359359387, |
| "learning_rate": 0.00015673729126165772, |
| "loss": 0.5291, |
| "step": 2170 |
| }, |
| { |
| "epoch": 0.755084303841474, |
| "grad_norm": 0.8444202602522437, |
| "learning_rate": 0.000156637288936198, |
| "loss": 0.5979, |
| "step": 2172 |
| }, |
| { |
| "epoch": 0.7557795932556927, |
| "grad_norm": 0.650071803598445, |
| "learning_rate": 0.000156537203151656, |
| "loss": 0.5416, |
| "step": 2174 |
| }, |
| { |
| "epoch": 0.7564748826699114, |
| "grad_norm": 0.8413408133328035, |
| "learning_rate": 0.0001564370340555153, |
| "loss": 0.6289, |
| "step": 2176 |
| }, |
| { |
| "epoch": 0.7571701720841301, |
| "grad_norm": 0.5687460625477972, |
| "learning_rate": 0.00015633678179538215, |
| "loss": 0.5903, |
| "step": 2178 |
| }, |
| { |
| "epoch": 0.7578654614983487, |
| "grad_norm": 0.7891119642696729, |
| "learning_rate": 0.00015623644651898544, |
| "loss": 0.6036, |
| "step": 2180 |
| }, |
| { |
| "epoch": 0.7585607509125674, |
| "grad_norm": 0.5465239219648831, |
| "learning_rate": 0.00015613602837417632, |
| "loss": 0.5858, |
| "step": 2182 |
| }, |
| { |
| "epoch": 0.7592560403267861, |
| "grad_norm": 0.6595658755524987, |
| "learning_rate": 0.00015603552750892812, |
| "loss": 0.507, |
| "step": 2184 |
| }, |
| { |
| "epoch": 0.7599513297410047, |
| "grad_norm": 0.4461549424157777, |
| "learning_rate": 0.000155934944071336, |
| "loss": 0.5526, |
| "step": 2186 |
| }, |
| { |
| "epoch": 0.7606466191552234, |
| "grad_norm": 0.6879482532243848, |
| "learning_rate": 0.00015583427820961682, |
| "loss": 0.587, |
| "step": 2188 |
| }, |
| { |
| "epoch": 0.7613419085694421, |
| "grad_norm": 0.990314646965209, |
| "learning_rate": 0.0001557335300721089, |
| "loss": 0.5649, |
| "step": 2190 |
| }, |
| { |
| "epoch": 0.7620371979836607, |
| "grad_norm": 1.101515439472337, |
| "learning_rate": 0.0001556326998072718, |
| "loss": 0.6442, |
| "step": 2192 |
| }, |
| { |
| "epoch": 0.7627324873978794, |
| "grad_norm": 0.5978051063323621, |
| "learning_rate": 0.00015553178756368618, |
| "loss": 0.5064, |
| "step": 2194 |
| }, |
| { |
| "epoch": 0.7634277768120981, |
| "grad_norm": 0.9387940256274154, |
| "learning_rate": 0.00015543079349005333, |
| "loss": 0.6245, |
| "step": 2196 |
| }, |
| { |
| "epoch": 0.7641230662263166, |
| "grad_norm": 0.7596273913172622, |
| "learning_rate": 0.00015532971773519528, |
| "loss": 0.6292, |
| "step": 2198 |
| }, |
| { |
| "epoch": 0.7648183556405354, |
| "grad_norm": 0.9249234619654654, |
| "learning_rate": 0.0001552285604480543, |
| "loss": 0.5532, |
| "step": 2200 |
| }, |
| { |
| "epoch": 0.765513645054754, |
| "grad_norm": 0.7260711670968186, |
| "learning_rate": 0.00015512732177769286, |
| "loss": 0.4786, |
| "step": 2202 |
| }, |
| { |
| "epoch": 0.7662089344689728, |
| "grad_norm": 0.7473521763806269, |
| "learning_rate": 0.00015502600187329346, |
| "loss": 0.5145, |
| "step": 2204 |
| }, |
| { |
| "epoch": 0.7669042238831913, |
| "grad_norm": 1.1602369586786794, |
| "learning_rate": 0.00015492460088415807, |
| "loss": 0.6282, |
| "step": 2206 |
| }, |
| { |
| "epoch": 0.76759951329741, |
| "grad_norm": 0.848038160748063, |
| "learning_rate": 0.00015482311895970838, |
| "loss": 0.5195, |
| "step": 2208 |
| }, |
| { |
| "epoch": 0.7682948027116288, |
| "grad_norm": 0.8249528004793321, |
| "learning_rate": 0.00015472155624948516, |
| "loss": 0.5051, |
| "step": 2210 |
| }, |
| { |
| "epoch": 0.7689900921258473, |
| "grad_norm": 1.034572812608102, |
| "learning_rate": 0.00015461991290314836, |
| "loss": 0.4521, |
| "step": 2212 |
| }, |
| { |
| "epoch": 0.769685381540066, |
| "grad_norm": 0.9105907481507995, |
| "learning_rate": 0.00015451818907047664, |
| "loss": 0.5475, |
| "step": 2214 |
| }, |
| { |
| "epoch": 0.7703806709542848, |
| "grad_norm": 1.0196726827311748, |
| "learning_rate": 0.00015441638490136736, |
| "loss": 0.5828, |
| "step": 2216 |
| }, |
| { |
| "epoch": 0.7710759603685033, |
| "grad_norm": 0.9621908020054395, |
| "learning_rate": 0.00015431450054583624, |
| "loss": 0.5734, |
| "step": 2218 |
| }, |
| { |
| "epoch": 0.771771249782722, |
| "grad_norm": 0.7586253979927021, |
| "learning_rate": 0.00015421253615401708, |
| "loss": 0.5281, |
| "step": 2220 |
| }, |
| { |
| "epoch": 0.7724665391969407, |
| "grad_norm": 0.7159186835240969, |
| "learning_rate": 0.00015411049187616168, |
| "loss": 0.5338, |
| "step": 2222 |
| }, |
| { |
| "epoch": 0.7731618286111593, |
| "grad_norm": 0.7654104439431867, |
| "learning_rate": 0.00015400836786263965, |
| "loss": 0.5465, |
| "step": 2224 |
| }, |
| { |
| "epoch": 0.773857118025378, |
| "grad_norm": 0.8844512066880119, |
| "learning_rate": 0.00015390616426393791, |
| "loss": 0.5622, |
| "step": 2226 |
| }, |
| { |
| "epoch": 0.7745524074395967, |
| "grad_norm": 0.7636977594059126, |
| "learning_rate": 0.00015380388123066075, |
| "loss": 0.5645, |
| "step": 2228 |
| }, |
| { |
| "epoch": 0.7752476968538154, |
| "grad_norm": 0.6363973262903718, |
| "learning_rate": 0.00015370151891352956, |
| "loss": 0.5933, |
| "step": 2230 |
| }, |
| { |
| "epoch": 0.775942986268034, |
| "grad_norm": 0.6190852796576922, |
| "learning_rate": 0.0001535990774633825, |
| "loss": 0.5674, |
| "step": 2232 |
| }, |
| { |
| "epoch": 0.7766382756822527, |
| "grad_norm": 0.6169591534781833, |
| "learning_rate": 0.00015349655703117434, |
| "loss": 0.491, |
| "step": 2234 |
| }, |
| { |
| "epoch": 0.7773335650964714, |
| "grad_norm": 0.901967565908769, |
| "learning_rate": 0.00015339395776797624, |
| "loss": 0.5222, |
| "step": 2236 |
| }, |
| { |
| "epoch": 0.77802885451069, |
| "grad_norm": 0.7375714747431651, |
| "learning_rate": 0.00015329127982497548, |
| "loss": 0.6043, |
| "step": 2238 |
| }, |
| { |
| "epoch": 0.7787241439249087, |
| "grad_norm": 0.5449413587750085, |
| "learning_rate": 0.00015318852335347545, |
| "loss": 0.4445, |
| "step": 2240 |
| }, |
| { |
| "epoch": 0.7794194333391274, |
| "grad_norm": 0.6555661162820788, |
| "learning_rate": 0.000153085688504895, |
| "loss": 0.5426, |
| "step": 2242 |
| }, |
| { |
| "epoch": 0.780114722753346, |
| "grad_norm": 0.9534465647450088, |
| "learning_rate": 0.00015298277543076872, |
| "loss": 0.4729, |
| "step": 2244 |
| }, |
| { |
| "epoch": 0.7808100121675647, |
| "grad_norm": 0.8698216407077228, |
| "learning_rate": 0.00015287978428274625, |
| "loss": 0.5977, |
| "step": 2246 |
| }, |
| { |
| "epoch": 0.7815053015817834, |
| "grad_norm": 0.900363520979286, |
| "learning_rate": 0.00015277671521259252, |
| "loss": 0.6219, |
| "step": 2248 |
| }, |
| { |
| "epoch": 0.782200590996002, |
| "grad_norm": 0.7530998286117955, |
| "learning_rate": 0.00015267356837218702, |
| "loss": 0.5554, |
| "step": 2250 |
| }, |
| { |
| "epoch": 0.782200590996002, |
| "eval_loss": 0.595500648021698, |
| "eval_runtime": 370.5074, |
| "eval_samples_per_second": 13.069, |
| "eval_steps_per_second": 0.41, |
| "step": 2250 |
| }, |
| { |
| "epoch": 0.7828958804102207, |
| "grad_norm": 1.1408998460850308, |
| "learning_rate": 0.00015257034391352403, |
| "loss": 0.6515, |
| "step": 2252 |
| }, |
| { |
| "epoch": 0.7835911698244394, |
| "grad_norm": 0.8969621990854273, |
| "learning_rate": 0.0001524670419887122, |
| "loss": 0.6501, |
| "step": 2254 |
| }, |
| { |
| "epoch": 0.7842864592386581, |
| "grad_norm": 0.6665465969738913, |
| "learning_rate": 0.00015236366274997418, |
| "loss": 0.6036, |
| "step": 2256 |
| }, |
| { |
| "epoch": 0.7849817486528767, |
| "grad_norm": 0.531024795387126, |
| "learning_rate": 0.0001522602063496467, |
| "loss": 0.5128, |
| "step": 2258 |
| }, |
| { |
| "epoch": 0.7856770380670954, |
| "grad_norm": 1.1619435853738966, |
| "learning_rate": 0.00015215667294018017, |
| "loss": 0.4987, |
| "step": 2260 |
| }, |
| { |
| "epoch": 0.7863723274813141, |
| "grad_norm": 0.55934494742226, |
| "learning_rate": 0.00015205306267413837, |
| "loss": 0.4729, |
| "step": 2262 |
| }, |
| { |
| "epoch": 0.7870676168955327, |
| "grad_norm": 0.7188169117405501, |
| "learning_rate": 0.00015194937570419848, |
| "loss": 0.5007, |
| "step": 2264 |
| }, |
| { |
| "epoch": 0.7877629063097514, |
| "grad_norm": 0.7195030544656635, |
| "learning_rate": 0.00015184561218315064, |
| "loss": 0.519, |
| "step": 2266 |
| }, |
| { |
| "epoch": 0.7884581957239701, |
| "grad_norm": 1.3598474735298196, |
| "learning_rate": 0.0001517417722638977, |
| "loss": 0.5392, |
| "step": 2268 |
| }, |
| { |
| "epoch": 0.7891534851381887, |
| "grad_norm": 1.0168579191876348, |
| "learning_rate": 0.0001516378560994553, |
| "loss": 0.5432, |
| "step": 2270 |
| }, |
| { |
| "epoch": 0.7898487745524074, |
| "grad_norm": 0.7947900880058245, |
| "learning_rate": 0.00015153386384295127, |
| "loss": 0.4383, |
| "step": 2272 |
| }, |
| { |
| "epoch": 0.7905440639666261, |
| "grad_norm": 1.347965316837954, |
| "learning_rate": 0.0001514297956476256, |
| "loss": 0.5153, |
| "step": 2274 |
| }, |
| { |
| "epoch": 0.7912393533808447, |
| "grad_norm": 1.0585653314732633, |
| "learning_rate": 0.0001513256516668302, |
| "loss": 0.5139, |
| "step": 2276 |
| }, |
| { |
| "epoch": 0.7919346427950634, |
| "grad_norm": 1.5897708606728695, |
| "learning_rate": 0.00015122143205402867, |
| "loss": 0.5554, |
| "step": 2278 |
| }, |
| { |
| "epoch": 0.7926299322092821, |
| "grad_norm": 0.9098511507133775, |
| "learning_rate": 0.00015111713696279601, |
| "loss": 0.5443, |
| "step": 2280 |
| }, |
| { |
| "epoch": 0.7933252216235008, |
| "grad_norm": 1.3752410406579205, |
| "learning_rate": 0.00015101276654681852, |
| "loss": 0.6228, |
| "step": 2282 |
| }, |
| { |
| "epoch": 0.7940205110377194, |
| "grad_norm": 0.8366795953408313, |
| "learning_rate": 0.00015090832095989335, |
| "loss": 0.6471, |
| "step": 2284 |
| }, |
| { |
| "epoch": 0.7947158004519381, |
| "grad_norm": 0.8632883922107255, |
| "learning_rate": 0.0001508038003559286, |
| "loss": 0.498, |
| "step": 2286 |
| }, |
| { |
| "epoch": 0.7954110898661568, |
| "grad_norm": 1.1032590650641334, |
| "learning_rate": 0.00015069920488894282, |
| "loss": 0.6356, |
| "step": 2288 |
| }, |
| { |
| "epoch": 0.7961063792803754, |
| "grad_norm": 1.2327371886597247, |
| "learning_rate": 0.00015059453471306488, |
| "loss": 0.5641, |
| "step": 2290 |
| }, |
| { |
| "epoch": 0.7968016686945941, |
| "grad_norm": 0.6849754809957169, |
| "learning_rate": 0.00015048978998253372, |
| "loss": 0.5461, |
| "step": 2292 |
| }, |
| { |
| "epoch": 0.7974969581088128, |
| "grad_norm": 1.1005494904216748, |
| "learning_rate": 0.00015038497085169818, |
| "loss": 0.6755, |
| "step": 2294 |
| }, |
| { |
| "epoch": 0.7981922475230314, |
| "grad_norm": 0.8037892664809793, |
| "learning_rate": 0.00015028007747501672, |
| "loss": 0.6166, |
| "step": 2296 |
| }, |
| { |
| "epoch": 0.7988875369372501, |
| "grad_norm": 0.515096865513587, |
| "learning_rate": 0.00015017511000705723, |
| "loss": 0.5248, |
| "step": 2298 |
| }, |
| { |
| "epoch": 0.7995828263514688, |
| "grad_norm": 0.526112593776873, |
| "learning_rate": 0.00015007006860249674, |
| "loss": 0.5081, |
| "step": 2300 |
| }, |
| { |
| "epoch": 0.8002781157656874, |
| "grad_norm": 0.6555698291475687, |
| "learning_rate": 0.00014996495341612121, |
| "loss": 0.5021, |
| "step": 2302 |
| }, |
| { |
| "epoch": 0.8009734051799061, |
| "grad_norm": 0.6702602021922074, |
| "learning_rate": 0.00014985976460282543, |
| "loss": 0.5523, |
| "step": 2304 |
| }, |
| { |
| "epoch": 0.8016686945941248, |
| "grad_norm": 0.7786218155389852, |
| "learning_rate": 0.0001497545023176126, |
| "loss": 0.494, |
| "step": 2306 |
| }, |
| { |
| "epoch": 0.8023639840083435, |
| "grad_norm": 0.7258281885786095, |
| "learning_rate": 0.00014964916671559412, |
| "loss": 0.4811, |
| "step": 2308 |
| }, |
| { |
| "epoch": 0.8030592734225621, |
| "grad_norm": 0.7470089527517397, |
| "learning_rate": 0.0001495437579519896, |
| "loss": 0.4933, |
| "step": 2310 |
| }, |
| { |
| "epoch": 0.8037545628367808, |
| "grad_norm": 0.6846825512730522, |
| "learning_rate": 0.00014943827618212632, |
| "loss": 0.4294, |
| "step": 2312 |
| }, |
| { |
| "epoch": 0.8044498522509995, |
| "grad_norm": 1.1043760342782873, |
| "learning_rate": 0.00014933272156143922, |
| "loss": 0.5303, |
| "step": 2314 |
| }, |
| { |
| "epoch": 0.8051451416652181, |
| "grad_norm": 1.2520441242526987, |
| "learning_rate": 0.00014922709424547052, |
| "loss": 0.6681, |
| "step": 2316 |
| }, |
| { |
| "epoch": 0.8058404310794368, |
| "grad_norm": 1.2907766696288898, |
| "learning_rate": 0.00014912139438986966, |
| "loss": 0.7507, |
| "step": 2318 |
| }, |
| { |
| "epoch": 0.8065357204936555, |
| "grad_norm": 1.140892292995435, |
| "learning_rate": 0.0001490156221503928, |
| "loss": 0.494, |
| "step": 2320 |
| }, |
| { |
| "epoch": 0.8072310099078741, |
| "grad_norm": 0.8752659425640287, |
| "learning_rate": 0.00014890977768290299, |
| "loss": 0.5153, |
| "step": 2322 |
| }, |
| { |
| "epoch": 0.8079262993220928, |
| "grad_norm": 1.217564740097299, |
| "learning_rate": 0.00014880386114336954, |
| "loss": 0.67, |
| "step": 2324 |
| }, |
| { |
| "epoch": 0.8086215887363115, |
| "grad_norm": 0.9665888600503119, |
| "learning_rate": 0.00014869787268786798, |
| "loss": 0.4828, |
| "step": 2326 |
| }, |
| { |
| "epoch": 0.8093168781505301, |
| "grad_norm": 0.8400739979551103, |
| "learning_rate": 0.00014859181247257996, |
| "loss": 0.574, |
| "step": 2328 |
| }, |
| { |
| "epoch": 0.8100121675647488, |
| "grad_norm": 0.7220947452102984, |
| "learning_rate": 0.0001484856806537927, |
| "loss": 0.6252, |
| "step": 2330 |
| }, |
| { |
| "epoch": 0.8107074569789675, |
| "grad_norm": 0.7663270835377898, |
| "learning_rate": 0.00014837947738789898, |
| "loss": 0.4473, |
| "step": 2332 |
| }, |
| { |
| "epoch": 0.8114027463931862, |
| "grad_norm": 0.6667034197533375, |
| "learning_rate": 0.00014827320283139695, |
| "loss": 0.4883, |
| "step": 2334 |
| }, |
| { |
| "epoch": 0.8120980358074048, |
| "grad_norm": 0.8633889989515121, |
| "learning_rate": 0.00014816685714088968, |
| "loss": 0.5967, |
| "step": 2336 |
| }, |
| { |
| "epoch": 0.8127933252216235, |
| "grad_norm": 0.9502089657040457, |
| "learning_rate": 0.00014806044047308516, |
| "loss": 0.5844, |
| "step": 2338 |
| }, |
| { |
| "epoch": 0.8134886146358422, |
| "grad_norm": 0.7241932558650107, |
| "learning_rate": 0.00014795395298479593, |
| "loss": 0.5417, |
| "step": 2340 |
| }, |
| { |
| "epoch": 0.8141839040500608, |
| "grad_norm": 0.6351527815368325, |
| "learning_rate": 0.0001478473948329389, |
| "loss": 0.5645, |
| "step": 2342 |
| }, |
| { |
| "epoch": 0.8148791934642795, |
| "grad_norm": 1.0981791843265052, |
| "learning_rate": 0.0001477407661745351, |
| "loss": 0.5939, |
| "step": 2344 |
| }, |
| { |
| "epoch": 0.8155744828784982, |
| "grad_norm": 0.7321642698624433, |
| "learning_rate": 0.00014763406716670943, |
| "loss": 0.5752, |
| "step": 2346 |
| }, |
| { |
| "epoch": 0.8162697722927168, |
| "grad_norm": 0.6239409142235018, |
| "learning_rate": 0.0001475272979666905, |
| "loss": 0.5851, |
| "step": 2348 |
| }, |
| { |
| "epoch": 0.8169650617069355, |
| "grad_norm": 0.8045965670853763, |
| "learning_rate": 0.00014742045873181037, |
| "loss": 0.5873, |
| "step": 2350 |
| }, |
| { |
| "epoch": 0.8176603511211542, |
| "grad_norm": 0.8459874011330281, |
| "learning_rate": 0.00014731354961950422, |
| "loss": 0.5444, |
| "step": 2352 |
| }, |
| { |
| "epoch": 0.8183556405353728, |
| "grad_norm": 0.7115510205306185, |
| "learning_rate": 0.00014720657078731025, |
| "loss": 0.5159, |
| "step": 2354 |
| }, |
| { |
| "epoch": 0.8190509299495915, |
| "grad_norm": 0.7218911584588392, |
| "learning_rate": 0.00014709952239286945, |
| "loss": 0.5151, |
| "step": 2356 |
| }, |
| { |
| "epoch": 0.8197462193638102, |
| "grad_norm": 0.6759319435463685, |
| "learning_rate": 0.00014699240459392518, |
| "loss": 0.4932, |
| "step": 2358 |
| }, |
| { |
| "epoch": 0.8204415087780289, |
| "grad_norm": 0.7842429800790871, |
| "learning_rate": 0.00014688521754832321, |
| "loss": 0.5117, |
| "step": 2360 |
| }, |
| { |
| "epoch": 0.8211367981922475, |
| "grad_norm": 1.187270675280872, |
| "learning_rate": 0.0001467779614140113, |
| "loss": 0.5774, |
| "step": 2362 |
| }, |
| { |
| "epoch": 0.8218320876064662, |
| "grad_norm": 1.3928493800154187, |
| "learning_rate": 0.00014667063634903895, |
| "loss": 0.5393, |
| "step": 2364 |
| }, |
| { |
| "epoch": 0.8225273770206849, |
| "grad_norm": 0.7432390567307604, |
| "learning_rate": 0.00014656324251155736, |
| "loss": 0.4451, |
| "step": 2366 |
| }, |
| { |
| "epoch": 0.8232226664349035, |
| "grad_norm": 0.745585544378624, |
| "learning_rate": 0.00014645578005981896, |
| "loss": 0.4071, |
| "step": 2368 |
| }, |
| { |
| "epoch": 0.8239179558491222, |
| "grad_norm": 1.1160515339209538, |
| "learning_rate": 0.00014634824915217733, |
| "loss": 0.6183, |
| "step": 2370 |
| }, |
| { |
| "epoch": 0.8246132452633409, |
| "grad_norm": 1.4887382057701017, |
| "learning_rate": 0.00014624064994708694, |
| "loss": 0.6139, |
| "step": 2372 |
| }, |
| { |
| "epoch": 0.8253085346775595, |
| "grad_norm": 0.8583174649910433, |
| "learning_rate": 0.0001461329826031029, |
| "loss": 0.6155, |
| "step": 2374 |
| }, |
| { |
| "epoch": 0.8260038240917782, |
| "grad_norm": 1.6239319050074577, |
| "learning_rate": 0.0001460252472788807, |
| "loss": 0.6581, |
| "step": 2376 |
| }, |
| { |
| "epoch": 0.8266991135059969, |
| "grad_norm": 1.0806676140550058, |
| "learning_rate": 0.00014591744413317603, |
| "loss": 0.5768, |
| "step": 2378 |
| }, |
| { |
| "epoch": 0.8273944029202155, |
| "grad_norm": 0.9532410697285137, |
| "learning_rate": 0.0001458095733248445, |
| "loss": 0.4727, |
| "step": 2380 |
| }, |
| { |
| "epoch": 0.8280896923344342, |
| "grad_norm": 1.013120781666776, |
| "learning_rate": 0.00014570163501284142, |
| "loss": 0.6277, |
| "step": 2382 |
| }, |
| { |
| "epoch": 0.8287849817486529, |
| "grad_norm": 0.9919742134433059, |
| "learning_rate": 0.00014559362935622164, |
| "loss": 0.5374, |
| "step": 2384 |
| }, |
| { |
| "epoch": 0.8294802711628716, |
| "grad_norm": 0.7478491420847245, |
| "learning_rate": 0.0001454855565141391, |
| "loss": 0.5655, |
| "step": 2386 |
| }, |
| { |
| "epoch": 0.8301755605770902, |
| "grad_norm": 1.0740936716392957, |
| "learning_rate": 0.00014537741664584695, |
| "loss": 0.5974, |
| "step": 2388 |
| }, |
| { |
| "epoch": 0.8308708499913089, |
| "grad_norm": 0.7493527064098355, |
| "learning_rate": 0.00014526920991069693, |
| "loss": 0.5789, |
| "step": 2390 |
| }, |
| { |
| "epoch": 0.8315661394055276, |
| "grad_norm": 0.7930930089550549, |
| "learning_rate": 0.00014516093646813936, |
| "loss": 0.5393, |
| "step": 2392 |
| }, |
| { |
| "epoch": 0.8322614288197462, |
| "grad_norm": 1.0607223540197912, |
| "learning_rate": 0.0001450525964777229, |
| "loss": 0.5463, |
| "step": 2394 |
| }, |
| { |
| "epoch": 0.8329567182339649, |
| "grad_norm": 0.7142937739779248, |
| "learning_rate": 0.0001449441900990943, |
| "loss": 0.5486, |
| "step": 2396 |
| }, |
| { |
| "epoch": 0.8336520076481836, |
| "grad_norm": 0.7204877069928376, |
| "learning_rate": 0.000144835717491998, |
| "loss": 0.4977, |
| "step": 2398 |
| }, |
| { |
| "epoch": 0.8343472970624022, |
| "grad_norm": 1.172477088190379, |
| "learning_rate": 0.00014472717881627614, |
| "loss": 0.5809, |
| "step": 2400 |
| }, |
| { |
| "epoch": 0.8343472970624022, |
| "eval_loss": 0.6061545014381409, |
| "eval_runtime": 370.4557, |
| "eval_samples_per_second": 13.07, |
| "eval_steps_per_second": 0.41, |
| "step": 2400 |
| }, |
| { |
| "epoch": 0.8350425864766209, |
| "grad_norm": 0.9883758967207753, |
| "learning_rate": 0.00014461857423186825, |
| "loss": 0.524, |
| "step": 2402 |
| }, |
| { |
| "epoch": 0.8357378758908396, |
| "grad_norm": 1.2220561945383148, |
| "learning_rate": 0.00014450990389881083, |
| "loss": 0.5457, |
| "step": 2404 |
| }, |
| { |
| "epoch": 0.8364331653050582, |
| "grad_norm": 1.463071490989689, |
| "learning_rate": 0.00014440116797723743, |
| "loss": 0.6799, |
| "step": 2406 |
| }, |
| { |
| "epoch": 0.8371284547192769, |
| "grad_norm": 0.9556553841303778, |
| "learning_rate": 0.0001442923666273782, |
| "loss": 0.5398, |
| "step": 2408 |
| }, |
| { |
| "epoch": 0.8378237441334956, |
| "grad_norm": 0.8609857491579432, |
| "learning_rate": 0.00014418350000955962, |
| "loss": 0.6049, |
| "step": 2410 |
| }, |
| { |
| "epoch": 0.8385190335477143, |
| "grad_norm": 0.9568357528475567, |
| "learning_rate": 0.00014407456828420445, |
| "loss": 0.5566, |
| "step": 2412 |
| }, |
| { |
| "epoch": 0.8392143229619329, |
| "grad_norm": 0.8570716316612914, |
| "learning_rate": 0.00014396557161183135, |
| "loss": 0.504, |
| "step": 2414 |
| }, |
| { |
| "epoch": 0.8399096123761516, |
| "grad_norm": 0.8647447256801674, |
| "learning_rate": 0.0001438565101530547, |
| "loss": 0.4693, |
| "step": 2416 |
| }, |
| { |
| "epoch": 0.8406049017903703, |
| "grad_norm": 1.0889945441110365, |
| "learning_rate": 0.00014374738406858429, |
| "loss": 0.553, |
| "step": 2418 |
| }, |
| { |
| "epoch": 0.8413001912045889, |
| "grad_norm": 0.9383001202479221, |
| "learning_rate": 0.00014363819351922522, |
| "loss": 0.5396, |
| "step": 2420 |
| }, |
| { |
| "epoch": 0.8419954806188076, |
| "grad_norm": 0.8092377755326117, |
| "learning_rate": 0.0001435289386658775, |
| "loss": 0.4548, |
| "step": 2422 |
| }, |
| { |
| "epoch": 0.8426907700330263, |
| "grad_norm": 1.310246150118249, |
| "learning_rate": 0.000143419619669536, |
| "loss": 0.5997, |
| "step": 2424 |
| }, |
| { |
| "epoch": 0.8433860594472449, |
| "grad_norm": 0.8511272016912201, |
| "learning_rate": 0.00014331023669129003, |
| "loss": 0.4683, |
| "step": 2426 |
| }, |
| { |
| "epoch": 0.8440813488614636, |
| "grad_norm": 0.861200809688603, |
| "learning_rate": 0.0001432007898923232, |
| "loss": 0.5187, |
| "step": 2428 |
| }, |
| { |
| "epoch": 0.8447766382756823, |
| "grad_norm": 0.9743662612070081, |
| "learning_rate": 0.0001430912794339132, |
| "loss": 0.4911, |
| "step": 2430 |
| }, |
| { |
| "epoch": 0.8454719276899009, |
| "grad_norm": 1.1305047733703086, |
| "learning_rate": 0.00014298170547743144, |
| "loss": 0.3942, |
| "step": 2432 |
| }, |
| { |
| "epoch": 0.8461672171041196, |
| "grad_norm": 1.3712238448240321, |
| "learning_rate": 0.000142872068184343, |
| "loss": 0.4884, |
| "step": 2434 |
| }, |
| { |
| "epoch": 0.8468625065183383, |
| "grad_norm": 1.1824893256721616, |
| "learning_rate": 0.00014276236771620625, |
| "loss": 0.5043, |
| "step": 2436 |
| }, |
| { |
| "epoch": 0.847557795932557, |
| "grad_norm": 1.0315998465733407, |
| "learning_rate": 0.0001426526042346726, |
| "loss": 0.4775, |
| "step": 2438 |
| }, |
| { |
| "epoch": 0.8482530853467756, |
| "grad_norm": 1.3235079091605297, |
| "learning_rate": 0.00014254277790148645, |
| "loss": 0.5928, |
| "step": 2440 |
| }, |
| { |
| "epoch": 0.8489483747609943, |
| "grad_norm": 1.1273370572626007, |
| "learning_rate": 0.00014243288887848463, |
| "loss": 0.4947, |
| "step": 2442 |
| }, |
| { |
| "epoch": 0.849643664175213, |
| "grad_norm": 0.8306982737650965, |
| "learning_rate": 0.00014232293732759652, |
| "loss": 0.5308, |
| "step": 2444 |
| }, |
| { |
| "epoch": 0.8503389535894316, |
| "grad_norm": 1.263229338137949, |
| "learning_rate": 0.00014221292341084354, |
| "loss": 0.5687, |
| "step": 2446 |
| }, |
| { |
| "epoch": 0.8510342430036503, |
| "grad_norm": 0.871724482752679, |
| "learning_rate": 0.00014210284729033903, |
| "loss": 0.444, |
| "step": 2448 |
| }, |
| { |
| "epoch": 0.851729532417869, |
| "grad_norm": 0.9273653240412545, |
| "learning_rate": 0.000141992709128288, |
| "loss": 0.4719, |
| "step": 2450 |
| }, |
| { |
| "epoch": 0.8524248218320876, |
| "grad_norm": 0.687896851633661, |
| "learning_rate": 0.00014188250908698692, |
| "loss": 0.4997, |
| "step": 2452 |
| }, |
| { |
| "epoch": 0.8531201112463063, |
| "grad_norm": 0.7506945949914205, |
| "learning_rate": 0.00014177224732882335, |
| "loss": 0.4686, |
| "step": 2454 |
| }, |
| { |
| "epoch": 0.853815400660525, |
| "grad_norm": 0.8162737115920249, |
| "learning_rate": 0.00014166192401627589, |
| "loss": 0.5283, |
| "step": 2456 |
| }, |
| { |
| "epoch": 0.8545106900747436, |
| "grad_norm": 1.1387603681759964, |
| "learning_rate": 0.0001415515393119138, |
| "loss": 0.5709, |
| "step": 2458 |
| }, |
| { |
| "epoch": 0.8552059794889623, |
| "grad_norm": 0.922996021627451, |
| "learning_rate": 0.0001414410933783968, |
| "loss": 0.5155, |
| "step": 2460 |
| }, |
| { |
| "epoch": 0.855901268903181, |
| "grad_norm": 1.076457839641558, |
| "learning_rate": 0.00014133058637847483, |
| "loss": 0.5211, |
| "step": 2462 |
| }, |
| { |
| "epoch": 0.8565965583173997, |
| "grad_norm": 1.0682472735103907, |
| "learning_rate": 0.0001412200184749879, |
| "loss": 0.6142, |
| "step": 2464 |
| }, |
| { |
| "epoch": 0.8572918477316183, |
| "grad_norm": 0.6935903391089329, |
| "learning_rate": 0.00014110938983086558, |
| "loss": 0.4511, |
| "step": 2466 |
| }, |
| { |
| "epoch": 0.857987137145837, |
| "grad_norm": 0.8437769362672389, |
| "learning_rate": 0.00014099870060912717, |
| "loss": 0.572, |
| "step": 2468 |
| }, |
| { |
| "epoch": 0.8586824265600557, |
| "grad_norm": 0.8789360074590062, |
| "learning_rate": 0.0001408879509728811, |
| "loss": 0.6124, |
| "step": 2470 |
| }, |
| { |
| "epoch": 0.8593777159742743, |
| "grad_norm": 0.8909899039496038, |
| "learning_rate": 0.00014077714108532482, |
| "loss": 0.4416, |
| "step": 2472 |
| }, |
| { |
| "epoch": 0.860073005388493, |
| "grad_norm": 1.1258001760188918, |
| "learning_rate": 0.00014066627110974458, |
| "loss": 0.4627, |
| "step": 2474 |
| }, |
| { |
| "epoch": 0.8607682948027117, |
| "grad_norm": 0.8686543206186234, |
| "learning_rate": 0.00014055534120951528, |
| "loss": 0.5104, |
| "step": 2476 |
| }, |
| { |
| "epoch": 0.8614635842169303, |
| "grad_norm": 1.0042374923348854, |
| "learning_rate": 0.00014044435154809997, |
| "loss": 0.6749, |
| "step": 2478 |
| }, |
| { |
| "epoch": 0.862158873631149, |
| "grad_norm": 1.4548986543137128, |
| "learning_rate": 0.00014033330228904983, |
| "loss": 0.5892, |
| "step": 2480 |
| }, |
| { |
| "epoch": 0.8628541630453677, |
| "grad_norm": 0.8709967374397177, |
| "learning_rate": 0.00014022219359600388, |
| "loss": 0.5093, |
| "step": 2482 |
| }, |
| { |
| "epoch": 0.8635494524595863, |
| "grad_norm": 0.896858736169805, |
| "learning_rate": 0.00014011102563268868, |
| "loss": 0.5164, |
| "step": 2484 |
| }, |
| { |
| "epoch": 0.864244741873805, |
| "grad_norm": 0.6741489014415273, |
| "learning_rate": 0.00013999979856291816, |
| "loss": 0.5496, |
| "step": 2486 |
| }, |
| { |
| "epoch": 0.8649400312880237, |
| "grad_norm": 0.7952403957972882, |
| "learning_rate": 0.00013988851255059332, |
| "loss": 0.5562, |
| "step": 2488 |
| }, |
| { |
| "epoch": 0.8656353207022424, |
| "grad_norm": 0.702264713556215, |
| "learning_rate": 0.00013977716775970207, |
| "loss": 0.5687, |
| "step": 2490 |
| }, |
| { |
| "epoch": 0.866330610116461, |
| "grad_norm": 1.5079773675274577, |
| "learning_rate": 0.00013966576435431885, |
| "loss": 0.6536, |
| "step": 2492 |
| }, |
| { |
| "epoch": 0.8670258995306797, |
| "grad_norm": 0.7598923711519189, |
| "learning_rate": 0.00013955430249860457, |
| "loss": 0.5378, |
| "step": 2494 |
| }, |
| { |
| "epoch": 0.8677211889448984, |
| "grad_norm": 0.8713369917712269, |
| "learning_rate": 0.00013944278235680614, |
| "loss": 0.5524, |
| "step": 2496 |
| }, |
| { |
| "epoch": 0.868416478359117, |
| "grad_norm": 0.6088275961040635, |
| "learning_rate": 0.00013933120409325653, |
| "loss": 0.4752, |
| "step": 2498 |
| }, |
| { |
| "epoch": 0.8691117677733357, |
| "grad_norm": 0.6210422474196712, |
| "learning_rate": 0.00013921956787237417, |
| "loss": 0.5281, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.8698070571875544, |
| "grad_norm": 0.8982266239311896, |
| "learning_rate": 0.00013910787385866308, |
| "loss": 0.5073, |
| "step": 2502 |
| }, |
| { |
| "epoch": 0.870502346601773, |
| "grad_norm": 0.8601407532546967, |
| "learning_rate": 0.0001389961222167123, |
| "loss": 0.4602, |
| "step": 2504 |
| }, |
| { |
| "epoch": 0.8711976360159916, |
| "grad_norm": 0.9709423553153457, |
| "learning_rate": 0.00013888431311119583, |
| "loss": 0.5811, |
| "step": 2506 |
| }, |
| { |
| "epoch": 0.8718929254302104, |
| "grad_norm": 1.113318453470144, |
| "learning_rate": 0.00013877244670687237, |
| "loss": 0.5, |
| "step": 2508 |
| }, |
| { |
| "epoch": 0.8725882148444289, |
| "grad_norm": 0.8887800367864689, |
| "learning_rate": 0.0001386605231685851, |
| "loss": 0.5353, |
| "step": 2510 |
| }, |
| { |
| "epoch": 0.8732835042586476, |
| "grad_norm": 0.8558201685685695, |
| "learning_rate": 0.00013854854266126123, |
| "loss": 0.4785, |
| "step": 2512 |
| }, |
| { |
| "epoch": 0.8739787936728663, |
| "grad_norm": 1.8897419247417928, |
| "learning_rate": 0.00013843650534991206, |
| "loss": 0.6461, |
| "step": 2514 |
| }, |
| { |
| "epoch": 0.874674083087085, |
| "grad_norm": 1.3579580716610924, |
| "learning_rate": 0.0001383244113996326, |
| "loss": 0.5925, |
| "step": 2516 |
| }, |
| { |
| "epoch": 0.8753693725013036, |
| "grad_norm": 1.3053323502456264, |
| "learning_rate": 0.00013821226097560123, |
| "loss": 0.5495, |
| "step": 2518 |
| }, |
| { |
| "epoch": 0.8760646619155223, |
| "grad_norm": 0.8765939036466984, |
| "learning_rate": 0.0001381000542430796, |
| "loss": 0.4576, |
| "step": 2520 |
| }, |
| { |
| "epoch": 0.876759951329741, |
| "grad_norm": 1.018205897173093, |
| "learning_rate": 0.00013798779136741228, |
| "loss": 0.6169, |
| "step": 2522 |
| }, |
| { |
| "epoch": 0.8774552407439596, |
| "grad_norm": 0.8145909570442347, |
| "learning_rate": 0.00013787547251402672, |
| "loss": 0.4436, |
| "step": 2524 |
| }, |
| { |
| "epoch": 0.8781505301581783, |
| "grad_norm": 1.0467029896962838, |
| "learning_rate": 0.0001377630978484327, |
| "loss": 0.5645, |
| "step": 2526 |
| }, |
| { |
| "epoch": 0.878845819572397, |
| "grad_norm": 1.37257571682647, |
| "learning_rate": 0.00013765066753622225, |
| "loss": 0.4969, |
| "step": 2528 |
| }, |
| { |
| "epoch": 0.8795411089866156, |
| "grad_norm": 1.1309679280831433, |
| "learning_rate": 0.00013753818174306955, |
| "loss": 0.6415, |
| "step": 2530 |
| }, |
| { |
| "epoch": 0.8802363984008343, |
| "grad_norm": 0.9783564147610989, |
| "learning_rate": 0.00013742564063473037, |
| "loss": 0.5622, |
| "step": 2532 |
| }, |
| { |
| "epoch": 0.880931687815053, |
| "grad_norm": 1.480288195795187, |
| "learning_rate": 0.00013731304437704207, |
| "loss": 0.5555, |
| "step": 2534 |
| }, |
| { |
| "epoch": 0.8816269772292716, |
| "grad_norm": 0.8056684648645962, |
| "learning_rate": 0.00013720039313592326, |
| "loss": 0.5015, |
| "step": 2536 |
| }, |
| { |
| "epoch": 0.8823222666434903, |
| "grad_norm": 0.836072520777469, |
| "learning_rate": 0.0001370876870773736, |
| "loss": 0.4655, |
| "step": 2538 |
| }, |
| { |
| "epoch": 0.883017556057709, |
| "grad_norm": 1.1940127467660082, |
| "learning_rate": 0.00013697492636747345, |
| "loss": 0.5142, |
| "step": 2540 |
| }, |
| { |
| "epoch": 0.8837128454719277, |
| "grad_norm": 0.7961826719577352, |
| "learning_rate": 0.00013686211117238382, |
| "loss": 0.4706, |
| "step": 2542 |
| }, |
| { |
| "epoch": 0.8844081348861463, |
| "grad_norm": 1.1956020907602014, |
| "learning_rate": 0.0001367492416583459, |
| "loss": 0.6053, |
| "step": 2544 |
| }, |
| { |
| "epoch": 0.885103424300365, |
| "grad_norm": 0.833906155396992, |
| "learning_rate": 0.0001366363179916809, |
| "loss": 0.517, |
| "step": 2546 |
| }, |
| { |
| "epoch": 0.8857987137145837, |
| "grad_norm": 1.2013026121334227, |
| "learning_rate": 0.00013652334033879, |
| "loss": 0.5974, |
| "step": 2548 |
| }, |
| { |
| "epoch": 0.8864940031288023, |
| "grad_norm": 1.057222497595888, |
| "learning_rate": 0.00013641030886615377, |
| "loss": 0.6188, |
| "step": 2550 |
| }, |
| { |
| "epoch": 0.8864940031288023, |
| "eval_loss": 0.5988921523094177, |
| "eval_runtime": 370.4051, |
| "eval_samples_per_second": 13.072, |
| "eval_steps_per_second": 0.41, |
| "step": 2550 |
| }, |
| { |
| "epoch": 0.887189292543021, |
| "grad_norm": 1.260946085888396, |
| "learning_rate": 0.00013629722374033216, |
| "loss": 0.5381, |
| "step": 2552 |
| }, |
| { |
| "epoch": 0.8878845819572397, |
| "grad_norm": 0.9825508876045091, |
| "learning_rate": 0.0001361840851279642, |
| "loss": 0.5306, |
| "step": 2554 |
| }, |
| { |
| "epoch": 0.8885798713714583, |
| "grad_norm": 1.110464958809028, |
| "learning_rate": 0.0001360708931957676, |
| "loss": 0.5052, |
| "step": 2556 |
| }, |
| { |
| "epoch": 0.889275160785677, |
| "grad_norm": 1.2573684773164022, |
| "learning_rate": 0.0001359576481105388, |
| "loss": 0.5999, |
| "step": 2558 |
| }, |
| { |
| "epoch": 0.8899704501998957, |
| "grad_norm": 1.0678263175308615, |
| "learning_rate": 0.0001358443500391525, |
| "loss": 0.4442, |
| "step": 2560 |
| }, |
| { |
| "epoch": 0.8906657396141143, |
| "grad_norm": 0.9900762732230567, |
| "learning_rate": 0.00013573099914856152, |
| "loss": 0.5135, |
| "step": 2562 |
| }, |
| { |
| "epoch": 0.891361029028333, |
| "grad_norm": 1.6166662252986856, |
| "learning_rate": 0.0001356175956057964, |
| "loss": 0.5902, |
| "step": 2564 |
| }, |
| { |
| "epoch": 0.8920563184425517, |
| "grad_norm": 1.3062109955219274, |
| "learning_rate": 0.00013550413957796545, |
| "loss": 0.5258, |
| "step": 2566 |
| }, |
| { |
| "epoch": 0.8927516078567704, |
| "grad_norm": 1.0624532752386364, |
| "learning_rate": 0.00013539063123225412, |
| "loss": 0.5258, |
| "step": 2568 |
| }, |
| { |
| "epoch": 0.893446897270989, |
| "grad_norm": 1.2569555959947833, |
| "learning_rate": 0.00013527707073592506, |
| "loss": 0.4077, |
| "step": 2570 |
| }, |
| { |
| "epoch": 0.8941421866852077, |
| "grad_norm": 1.957202257083521, |
| "learning_rate": 0.0001351634582563178, |
| "loss": 0.4637, |
| "step": 2572 |
| }, |
| { |
| "epoch": 0.8948374760994264, |
| "grad_norm": 1.0852820449574763, |
| "learning_rate": 0.0001350497939608484, |
| "loss": 0.4512, |
| "step": 2574 |
| }, |
| { |
| "epoch": 0.895532765513645, |
| "grad_norm": 1.1527392408428756, |
| "learning_rate": 0.00013493607801700925, |
| "loss": 0.6019, |
| "step": 2576 |
| }, |
| { |
| "epoch": 0.8962280549278637, |
| "grad_norm": 1.4110022221526575, |
| "learning_rate": 0.00013482231059236898, |
| "loss": 0.3984, |
| "step": 2578 |
| }, |
| { |
| "epoch": 0.8969233443420824, |
| "grad_norm": 1.1436468085493028, |
| "learning_rate": 0.00013470849185457198, |
| "loss": 0.6252, |
| "step": 2580 |
| }, |
| { |
| "epoch": 0.897618633756301, |
| "grad_norm": 1.1493865048392684, |
| "learning_rate": 0.00013459462197133826, |
| "loss": 0.5157, |
| "step": 2582 |
| }, |
| { |
| "epoch": 0.8983139231705197, |
| "grad_norm": 0.7337993535613259, |
| "learning_rate": 0.00013448070111046322, |
| "loss": 0.4864, |
| "step": 2584 |
| }, |
| { |
| "epoch": 0.8990092125847384, |
| "grad_norm": 1.5101702921506, |
| "learning_rate": 0.00013436672943981735, |
| "loss": 0.5036, |
| "step": 2586 |
| }, |
| { |
| "epoch": 0.899704501998957, |
| "grad_norm": 0.8657863919182651, |
| "learning_rate": 0.00013425270712734606, |
| "loss": 0.5167, |
| "step": 2588 |
| }, |
| { |
| "epoch": 0.9003997914131757, |
| "grad_norm": 0.8636648680520452, |
| "learning_rate": 0.0001341386343410693, |
| "loss": 0.4553, |
| "step": 2590 |
| }, |
| { |
| "epoch": 0.9010950808273944, |
| "grad_norm": 1.0243645157298624, |
| "learning_rate": 0.00013402451124908154, |
| "loss": 0.4713, |
| "step": 2592 |
| }, |
| { |
| "epoch": 0.9017903702416131, |
| "grad_norm": 1.0097975588278003, |
| "learning_rate": 0.0001339103380195512, |
| "loss": 0.4287, |
| "step": 2594 |
| }, |
| { |
| "epoch": 0.9024856596558317, |
| "grad_norm": 1.0308267158096918, |
| "learning_rate": 0.0001337961148207207, |
| "loss": 0.4882, |
| "step": 2596 |
| }, |
| { |
| "epoch": 0.9031809490700504, |
| "grad_norm": 1.5965054781850732, |
| "learning_rate": 0.0001336818418209061, |
| "loss": 0.4683, |
| "step": 2598 |
| }, |
| { |
| "epoch": 0.9038762384842691, |
| "grad_norm": 1.569464284225874, |
| "learning_rate": 0.0001335675191884968, |
| "loss": 0.5647, |
| "step": 2600 |
| }, |
| { |
| "epoch": 0.9045715278984877, |
| "grad_norm": 1.076629025329801, |
| "learning_rate": 0.0001334531470919553, |
| "loss": 0.5537, |
| "step": 2602 |
| }, |
| { |
| "epoch": 0.9052668173127064, |
| "grad_norm": 1.3725023912404772, |
| "learning_rate": 0.0001333387256998171, |
| "loss": 0.4526, |
| "step": 2604 |
| }, |
| { |
| "epoch": 0.9059621067269251, |
| "grad_norm": 1.532062124743962, |
| "learning_rate": 0.00013322425518069023, |
| "loss": 0.4768, |
| "step": 2606 |
| }, |
| { |
| "epoch": 0.9066573961411437, |
| "grad_norm": 0.8872105309349223, |
| "learning_rate": 0.00013310973570325517, |
| "loss": 0.4409, |
| "step": 2608 |
| }, |
| { |
| "epoch": 0.9073526855553624, |
| "grad_norm": 1.0623707057894844, |
| "learning_rate": 0.00013299516743626452, |
| "loss": 0.5586, |
| "step": 2610 |
| }, |
| { |
| "epoch": 0.9080479749695811, |
| "grad_norm": 1.5592936943462063, |
| "learning_rate": 0.00013288055054854286, |
| "loss": 0.5225, |
| "step": 2612 |
| }, |
| { |
| "epoch": 0.9087432643837997, |
| "grad_norm": 1.0882129513804535, |
| "learning_rate": 0.00013276588520898624, |
| "loss": 0.5074, |
| "step": 2614 |
| }, |
| { |
| "epoch": 0.9094385537980184, |
| "grad_norm": 0.8473823940307793, |
| "learning_rate": 0.00013265117158656227, |
| "loss": 0.474, |
| "step": 2616 |
| }, |
| { |
| "epoch": 0.9101338432122371, |
| "grad_norm": 1.43028993302617, |
| "learning_rate": 0.0001325364098503097, |
| "loss": 0.6691, |
| "step": 2618 |
| }, |
| { |
| "epoch": 0.9108291326264558, |
| "grad_norm": 0.8799124795333675, |
| "learning_rate": 0.00013242160016933803, |
| "loss": 0.4689, |
| "step": 2620 |
| }, |
| { |
| "epoch": 0.9115244220406744, |
| "grad_norm": 1.2478722240815094, |
| "learning_rate": 0.00013230674271282755, |
| "loss": 0.4807, |
| "step": 2622 |
| }, |
| { |
| "epoch": 0.9122197114548931, |
| "grad_norm": 1.3001366310730733, |
| "learning_rate": 0.00013219183765002888, |
| "loss": 0.468, |
| "step": 2624 |
| }, |
| { |
| "epoch": 0.9129150008691118, |
| "grad_norm": 0.8844623209103742, |
| "learning_rate": 0.00013207688515026286, |
| "loss": 0.5792, |
| "step": 2626 |
| }, |
| { |
| "epoch": 0.9136102902833304, |
| "grad_norm": 0.8809463539684392, |
| "learning_rate": 0.0001319618853829202, |
| "loss": 0.4704, |
| "step": 2628 |
| }, |
| { |
| "epoch": 0.9143055796975491, |
| "grad_norm": 0.891809393009814, |
| "learning_rate": 0.0001318468385174612, |
| "loss": 0.4186, |
| "step": 2630 |
| }, |
| { |
| "epoch": 0.9150008691117678, |
| "grad_norm": 1.1800222087320673, |
| "learning_rate": 0.00013173174472341567, |
| "loss": 0.5423, |
| "step": 2632 |
| }, |
| { |
| "epoch": 0.9156961585259864, |
| "grad_norm": 1.1334330539620814, |
| "learning_rate": 0.00013161660417038251, |
| "loss": 0.4116, |
| "step": 2634 |
| }, |
| { |
| "epoch": 0.9163914479402051, |
| "grad_norm": 1.2656185491277732, |
| "learning_rate": 0.0001315014170280295, |
| "loss": 0.4574, |
| "step": 2636 |
| }, |
| { |
| "epoch": 0.9170867373544238, |
| "grad_norm": 1.0727800107378958, |
| "learning_rate": 0.00013138618346609314, |
| "loss": 0.5088, |
| "step": 2638 |
| }, |
| { |
| "epoch": 0.9177820267686424, |
| "grad_norm": 1.6180690174111048, |
| "learning_rate": 0.00013127090365437828, |
| "loss": 0.5306, |
| "step": 2640 |
| }, |
| { |
| "epoch": 0.9184773161828611, |
| "grad_norm": 1.0058321963478047, |
| "learning_rate": 0.0001311555777627579, |
| "loss": 0.4775, |
| "step": 2642 |
| }, |
| { |
| "epoch": 0.9191726055970798, |
| "grad_norm": 1.169507937152265, |
| "learning_rate": 0.00013104020596117302, |
| "loss": 0.459, |
| "step": 2644 |
| }, |
| { |
| "epoch": 0.9198678950112984, |
| "grad_norm": 1.104893162104563, |
| "learning_rate": 0.00013092478841963216, |
| "loss": 0.3471, |
| "step": 2646 |
| }, |
| { |
| "epoch": 0.9205631844255171, |
| "grad_norm": 1.414806426255017, |
| "learning_rate": 0.00013080932530821126, |
| "loss": 0.5576, |
| "step": 2648 |
| }, |
| { |
| "epoch": 0.9212584738397358, |
| "grad_norm": 1.2158116380630863, |
| "learning_rate": 0.00013069381679705347, |
| "loss": 0.4344, |
| "step": 2650 |
| }, |
| { |
| "epoch": 0.9219537632539545, |
| "grad_norm": 1.3688353606618948, |
| "learning_rate": 0.00013057826305636888, |
| "loss": 0.514, |
| "step": 2652 |
| }, |
| { |
| "epoch": 0.9226490526681731, |
| "grad_norm": 1.1225231709200263, |
| "learning_rate": 0.00013046266425643404, |
| "loss": 0.5356, |
| "step": 2654 |
| }, |
| { |
| "epoch": 0.9233443420823918, |
| "grad_norm": 0.9635893691261951, |
| "learning_rate": 0.00013034702056759207, |
| "loss": 0.5756, |
| "step": 2656 |
| }, |
| { |
| "epoch": 0.9240396314966105, |
| "grad_norm": 0.8383991263110834, |
| "learning_rate": 0.00013023133216025229, |
| "loss": 0.3956, |
| "step": 2658 |
| }, |
| { |
| "epoch": 0.9247349209108291, |
| "grad_norm": 1.0343993568490624, |
| "learning_rate": 0.00013011559920488966, |
| "loss": 0.4683, |
| "step": 2660 |
| }, |
| { |
| "epoch": 0.9254302103250478, |
| "grad_norm": 0.9812375371477972, |
| "learning_rate": 0.00012999982187204503, |
| "loss": 0.4248, |
| "step": 2662 |
| }, |
| { |
| "epoch": 0.9261254997392665, |
| "grad_norm": 0.9770476760699544, |
| "learning_rate": 0.0001298840003323245, |
| "loss": 0.4754, |
| "step": 2664 |
| }, |
| { |
| "epoch": 0.9268207891534851, |
| "grad_norm": 1.309557007869239, |
| "learning_rate": 0.00012976813475639942, |
| "loss": 0.5502, |
| "step": 2666 |
| }, |
| { |
| "epoch": 0.9275160785677038, |
| "grad_norm": 1.2964006459833535, |
| "learning_rate": 0.00012965222531500595, |
| "loss": 0.4465, |
| "step": 2668 |
| }, |
| { |
| "epoch": 0.9282113679819225, |
| "grad_norm": 0.9977731540291016, |
| "learning_rate": 0.00012953627217894494, |
| "loss": 0.5301, |
| "step": 2670 |
| }, |
| { |
| "epoch": 0.9289066573961411, |
| "grad_norm": 1.5069957806353622, |
| "learning_rate": 0.00012942027551908153, |
| "loss": 0.6823, |
| "step": 2672 |
| }, |
| { |
| "epoch": 0.9296019468103598, |
| "grad_norm": 0.9910285199783668, |
| "learning_rate": 0.00012930423550634512, |
| "loss": 0.5514, |
| "step": 2674 |
| }, |
| { |
| "epoch": 0.9302972362245785, |
| "grad_norm": 0.9955906654151888, |
| "learning_rate": 0.00012918815231172893, |
| "loss": 0.4664, |
| "step": 2676 |
| }, |
| { |
| "epoch": 0.9309925256387972, |
| "grad_norm": 1.3738224815716196, |
| "learning_rate": 0.00012907202610628978, |
| "loss": 0.5161, |
| "step": 2678 |
| }, |
| { |
| "epoch": 0.9316878150530158, |
| "grad_norm": 1.1246636950555773, |
| "learning_rate": 0.00012895585706114803, |
| "loss": 0.5173, |
| "step": 2680 |
| }, |
| { |
| "epoch": 0.9323831044672345, |
| "grad_norm": 1.2094662027104317, |
| "learning_rate": 0.0001288396453474869, |
| "loss": 0.5526, |
| "step": 2682 |
| }, |
| { |
| "epoch": 0.9330783938814532, |
| "grad_norm": 1.0711854318426073, |
| "learning_rate": 0.00012872339113655277, |
| "loss": 0.6348, |
| "step": 2684 |
| }, |
| { |
| "epoch": 0.9337736832956718, |
| "grad_norm": 0.7121977366760148, |
| "learning_rate": 0.0001286070945996544, |
| "loss": 0.4679, |
| "step": 2686 |
| }, |
| { |
| "epoch": 0.9344689727098905, |
| "grad_norm": 1.2285647274291915, |
| "learning_rate": 0.00012849075590816306, |
| "loss": 0.5414, |
| "step": 2688 |
| }, |
| { |
| "epoch": 0.9351642621241092, |
| "grad_norm": 0.8500140633320541, |
| "learning_rate": 0.00012837437523351212, |
| "loss": 0.5336, |
| "step": 2690 |
| }, |
| { |
| "epoch": 0.9358595515383278, |
| "grad_norm": 0.9912987625423477, |
| "learning_rate": 0.0001282579527471968, |
| "loss": 0.5757, |
| "step": 2692 |
| }, |
| { |
| "epoch": 0.9365548409525465, |
| "grad_norm": 1.1294656511764527, |
| "learning_rate": 0.00012814148862077396, |
| "loss": 0.5435, |
| "step": 2694 |
| }, |
| { |
| "epoch": 0.9372501303667652, |
| "grad_norm": 0.9924499997227866, |
| "learning_rate": 0.00012802498302586169, |
| "loss": 0.5239, |
| "step": 2696 |
| }, |
| { |
| "epoch": 0.9379454197809838, |
| "grad_norm": 0.9931663516838377, |
| "learning_rate": 0.0001279084361341394, |
| "loss": 0.5379, |
| "step": 2698 |
| }, |
| { |
| "epoch": 0.9386407091952025, |
| "grad_norm": 0.8154595203564674, |
| "learning_rate": 0.00012779184811734714, |
| "loss": 0.5308, |
| "step": 2700 |
| }, |
| { |
| "epoch": 0.9386407091952025, |
| "eval_loss": 0.5993536114692688, |
| "eval_runtime": 370.6908, |
| "eval_samples_per_second": 13.062, |
| "eval_steps_per_second": 0.41, |
| "step": 2700 |
| }, |
| { |
| "epoch": 0.9393359986094212, |
| "grad_norm": 0.9207451002458883, |
| "learning_rate": 0.00012767521914728577, |
| "loss": 0.619, |
| "step": 2702 |
| }, |
| { |
| "epoch": 0.9400312880236399, |
| "grad_norm": 0.7709338052749476, |
| "learning_rate": 0.0001275585493958163, |
| "loss": 0.5353, |
| "step": 2704 |
| }, |
| { |
| "epoch": 0.9407265774378585, |
| "grad_norm": 0.9347819983326865, |
| "learning_rate": 0.00012744183903485998, |
| "loss": 0.5887, |
| "step": 2706 |
| }, |
| { |
| "epoch": 0.9414218668520772, |
| "grad_norm": 0.8400228462784727, |
| "learning_rate": 0.00012732508823639783, |
| "loss": 0.5349, |
| "step": 2708 |
| }, |
| { |
| "epoch": 0.9421171562662959, |
| "grad_norm": 0.8594099833585327, |
| "learning_rate": 0.00012720829717247046, |
| "loss": 0.5287, |
| "step": 2710 |
| }, |
| { |
| "epoch": 0.9428124456805145, |
| "grad_norm": 1.2418406505159845, |
| "learning_rate": 0.00012709146601517786, |
| "loss": 0.4746, |
| "step": 2712 |
| }, |
| { |
| "epoch": 0.9435077350947332, |
| "grad_norm": 1.3579625328098206, |
| "learning_rate": 0.00012697459493667904, |
| "loss": 0.5694, |
| "step": 2714 |
| }, |
| { |
| "epoch": 0.9442030245089519, |
| "grad_norm": 0.9671663851672909, |
| "learning_rate": 0.00012685768410919187, |
| "loss": 0.5845, |
| "step": 2716 |
| }, |
| { |
| "epoch": 0.9448983139231705, |
| "grad_norm": 0.8384748630048591, |
| "learning_rate": 0.0001267407337049928, |
| "loss": 0.5041, |
| "step": 2718 |
| }, |
| { |
| "epoch": 0.9455936033373892, |
| "grad_norm": 0.8023013976838196, |
| "learning_rate": 0.00012662374389641655, |
| "loss": 0.4349, |
| "step": 2720 |
| }, |
| { |
| "epoch": 0.9462888927516079, |
| "grad_norm": 0.9733910555938483, |
| "learning_rate": 0.000126506714855856, |
| "loss": 0.5265, |
| "step": 2722 |
| }, |
| { |
| "epoch": 0.9469841821658265, |
| "grad_norm": 0.8608103911393638, |
| "learning_rate": 0.00012638964675576174, |
| "loss": 0.5961, |
| "step": 2724 |
| }, |
| { |
| "epoch": 0.9476794715800452, |
| "grad_norm": 0.7587175703742405, |
| "learning_rate": 0.00012627253976864197, |
| "loss": 0.5209, |
| "step": 2726 |
| }, |
| { |
| "epoch": 0.9483747609942639, |
| "grad_norm": 0.8592614641060521, |
| "learning_rate": 0.00012615539406706219, |
| "loss": 0.4699, |
| "step": 2728 |
| }, |
| { |
| "epoch": 0.9490700504084826, |
| "grad_norm": 1.0954602475814819, |
| "learning_rate": 0.00012603820982364492, |
| "loss": 0.5598, |
| "step": 2730 |
| }, |
| { |
| "epoch": 0.9497653398227012, |
| "grad_norm": 0.9033493366456706, |
| "learning_rate": 0.0001259209872110695, |
| "loss": 0.4257, |
| "step": 2732 |
| }, |
| { |
| "epoch": 0.9504606292369199, |
| "grad_norm": 0.9647253788598223, |
| "learning_rate": 0.00012580372640207186, |
| "loss": 0.5263, |
| "step": 2734 |
| }, |
| { |
| "epoch": 0.9511559186511386, |
| "grad_norm": 1.1299299518105947, |
| "learning_rate": 0.00012568642756944409, |
| "loss": 0.6231, |
| "step": 2736 |
| }, |
| { |
| "epoch": 0.9518512080653572, |
| "grad_norm": 1.0864604863693257, |
| "learning_rate": 0.0001255690908860344, |
| "loss": 0.4388, |
| "step": 2738 |
| }, |
| { |
| "epoch": 0.9525464974795759, |
| "grad_norm": 1.2514655683613718, |
| "learning_rate": 0.00012545171652474674, |
| "loss": 0.6619, |
| "step": 2740 |
| }, |
| { |
| "epoch": 0.9532417868937946, |
| "grad_norm": 1.6269307483210904, |
| "learning_rate": 0.00012533430465854067, |
| "loss": 0.6272, |
| "step": 2742 |
| }, |
| { |
| "epoch": 0.9539370763080132, |
| "grad_norm": 0.9230510643625829, |
| "learning_rate": 0.00012521685546043086, |
| "loss": 0.4446, |
| "step": 2744 |
| }, |
| { |
| "epoch": 0.9546323657222319, |
| "grad_norm": 1.1798577172665998, |
| "learning_rate": 0.00012509936910348713, |
| "loss": 0.6538, |
| "step": 2746 |
| }, |
| { |
| "epoch": 0.9553276551364506, |
| "grad_norm": 0.9731594744172876, |
| "learning_rate": 0.000124981845760834, |
| "loss": 0.5092, |
| "step": 2748 |
| }, |
| { |
| "epoch": 0.9560229445506692, |
| "grad_norm": 0.9812795543703706, |
| "learning_rate": 0.00012486428560565043, |
| "loss": 0.5103, |
| "step": 2750 |
| }, |
| { |
| "epoch": 0.9567182339648879, |
| "grad_norm": 0.8144207800465867, |
| "learning_rate": 0.00012474668881116974, |
| "loss": 0.5147, |
| "step": 2752 |
| }, |
| { |
| "epoch": 0.9574135233791066, |
| "grad_norm": 1.0698715288092682, |
| "learning_rate": 0.0001246290555506792, |
| "loss": 0.4519, |
| "step": 2754 |
| }, |
| { |
| "epoch": 0.9581088127933253, |
| "grad_norm": 0.9593492727392562, |
| "learning_rate": 0.00012451138599751978, |
| "loss": 0.5212, |
| "step": 2756 |
| }, |
| { |
| "epoch": 0.9588041022075439, |
| "grad_norm": 0.9881767222663046, |
| "learning_rate": 0.0001243936803250859, |
| "loss": 0.5411, |
| "step": 2758 |
| }, |
| { |
| "epoch": 0.9594993916217626, |
| "grad_norm": 1.175314577812427, |
| "learning_rate": 0.00012427593870682538, |
| "loss": 0.5212, |
| "step": 2760 |
| }, |
| { |
| "epoch": 0.9601946810359813, |
| "grad_norm": 1.4700715481738074, |
| "learning_rate": 0.0001241581613162388, |
| "loss": 0.5457, |
| "step": 2762 |
| }, |
| { |
| "epoch": 0.9608899704501999, |
| "grad_norm": 0.8932068551574899, |
| "learning_rate": 0.0001240403483268796, |
| "loss": 0.484, |
| "step": 2764 |
| }, |
| { |
| "epoch": 0.9615852598644186, |
| "grad_norm": 1.3652074986078104, |
| "learning_rate": 0.00012392249991235353, |
| "loss": 0.4389, |
| "step": 2766 |
| }, |
| { |
| "epoch": 0.9622805492786373, |
| "grad_norm": 1.0421297234560185, |
| "learning_rate": 0.0001238046162463187, |
| "loss": 0.5435, |
| "step": 2768 |
| }, |
| { |
| "epoch": 0.9629758386928559, |
| "grad_norm": 1.3291690102104428, |
| "learning_rate": 0.0001236866975024851, |
| "loss": 0.4857, |
| "step": 2770 |
| }, |
| { |
| "epoch": 0.9636711281070746, |
| "grad_norm": 1.6521750685718746, |
| "learning_rate": 0.00012356874385461436, |
| "loss": 0.4814, |
| "step": 2772 |
| }, |
| { |
| "epoch": 0.9643664175212933, |
| "grad_norm": 1.3377291765621406, |
| "learning_rate": 0.00012345075547651958, |
| "loss": 0.5762, |
| "step": 2774 |
| }, |
| { |
| "epoch": 0.9650617069355119, |
| "grad_norm": 1.1200242362324335, |
| "learning_rate": 0.00012333273254206512, |
| "loss": 0.4229, |
| "step": 2776 |
| }, |
| { |
| "epoch": 0.9657569963497306, |
| "grad_norm": 1.3223390097770544, |
| "learning_rate": 0.0001232146752251661, |
| "loss": 0.3813, |
| "step": 2778 |
| }, |
| { |
| "epoch": 0.9664522857639493, |
| "grad_norm": 0.930329709935029, |
| "learning_rate": 0.00012309658369978843, |
| "loss": 0.3944, |
| "step": 2780 |
| }, |
| { |
| "epoch": 0.967147575178168, |
| "grad_norm": 1.4094106647839177, |
| "learning_rate": 0.00012297845813994838, |
| "loss": 0.5143, |
| "step": 2782 |
| }, |
| { |
| "epoch": 0.9678428645923866, |
| "grad_norm": 1.3679225707692713, |
| "learning_rate": 0.0001228602987197124, |
| "loss": 0.3728, |
| "step": 2784 |
| }, |
| { |
| "epoch": 0.9685381540066053, |
| "grad_norm": 1.4154349713192809, |
| "learning_rate": 0.0001227421056131968, |
| "loss": 0.446, |
| "step": 2786 |
| }, |
| { |
| "epoch": 0.969233443420824, |
| "grad_norm": 1.5183436364256648, |
| "learning_rate": 0.0001226238789945676, |
| "loss": 0.464, |
| "step": 2788 |
| }, |
| { |
| "epoch": 0.9699287328350426, |
| "grad_norm": 1.874321615574962, |
| "learning_rate": 0.00012250561903804006, |
| "loss": 0.6993, |
| "step": 2790 |
| }, |
| { |
| "epoch": 0.9706240222492613, |
| "grad_norm": 2.7438834256137095, |
| "learning_rate": 0.0001223873259178787, |
| "loss": 0.5092, |
| "step": 2792 |
| }, |
| { |
| "epoch": 0.97131931166348, |
| "grad_norm": 1.7396029948291611, |
| "learning_rate": 0.0001222689998083969, |
| "loss": 0.6264, |
| "step": 2794 |
| }, |
| { |
| "epoch": 0.9720146010776985, |
| "grad_norm": 1.0995732575416968, |
| "learning_rate": 0.00012215064088395657, |
| "loss": 0.3722, |
| "step": 2796 |
| }, |
| { |
| "epoch": 0.9727098904919173, |
| "grad_norm": 1.386292977935405, |
| "learning_rate": 0.00012203224931896802, |
| "loss": 0.433, |
| "step": 2798 |
| }, |
| { |
| "epoch": 0.973405179906136, |
| "grad_norm": 1.5401468207253657, |
| "learning_rate": 0.00012191382528788969, |
| "loss": 0.5386, |
| "step": 2800 |
| }, |
| { |
| "epoch": 0.9741004693203545, |
| "grad_norm": 1.2685559191897577, |
| "learning_rate": 0.00012179536896522779, |
| "loss": 0.3922, |
| "step": 2802 |
| }, |
| { |
| "epoch": 0.9747957587345732, |
| "grad_norm": 0.8703585645012714, |
| "learning_rate": 0.00012167688052553621, |
| "loss": 0.503, |
| "step": 2804 |
| }, |
| { |
| "epoch": 0.975491048148792, |
| "grad_norm": 1.3530989288275252, |
| "learning_rate": 0.00012155836014341604, |
| "loss": 0.543, |
| "step": 2806 |
| }, |
| { |
| "epoch": 0.9761863375630107, |
| "grad_norm": 1.026118209725581, |
| "learning_rate": 0.00012143980799351557, |
| "loss": 0.5235, |
| "step": 2808 |
| }, |
| { |
| "epoch": 0.9768816269772292, |
| "grad_norm": 1.5039102852667445, |
| "learning_rate": 0.00012132122425052983, |
| "loss": 0.493, |
| "step": 2810 |
| }, |
| { |
| "epoch": 0.977576916391448, |
| "grad_norm": 1.287806036517661, |
| "learning_rate": 0.0001212026090892004, |
| "loss": 0.5469, |
| "step": 2812 |
| }, |
| { |
| "epoch": 0.9782722058056667, |
| "grad_norm": 1.4035447243146326, |
| "learning_rate": 0.00012108396268431519, |
| "loss": 0.489, |
| "step": 2814 |
| }, |
| { |
| "epoch": 0.9789674952198852, |
| "grad_norm": 0.7604778680143648, |
| "learning_rate": 0.00012096528521070815, |
| "loss": 0.4487, |
| "step": 2816 |
| }, |
| { |
| "epoch": 0.979662784634104, |
| "grad_norm": 1.1794634261106687, |
| "learning_rate": 0.00012084657684325895, |
| "loss": 0.5343, |
| "step": 2818 |
| }, |
| { |
| "epoch": 0.9803580740483226, |
| "grad_norm": 0.9707197215548774, |
| "learning_rate": 0.0001207278377568929, |
| "loss": 0.5569, |
| "step": 2820 |
| }, |
| { |
| "epoch": 0.9810533634625412, |
| "grad_norm": 1.0346438602535457, |
| "learning_rate": 0.00012060906812658041, |
| "loss": 0.4768, |
| "step": 2822 |
| }, |
| { |
| "epoch": 0.9817486528767599, |
| "grad_norm": 1.462013422579743, |
| "learning_rate": 0.0001204902681273371, |
| "loss": 0.5664, |
| "step": 2824 |
| }, |
| { |
| "epoch": 0.9824439422909786, |
| "grad_norm": 1.1362693966197832, |
| "learning_rate": 0.00012037143793422316, |
| "loss": 0.4554, |
| "step": 2826 |
| }, |
| { |
| "epoch": 0.9831392317051972, |
| "grad_norm": 0.9765618194808099, |
| "learning_rate": 0.00012025257772234342, |
| "loss": 0.4801, |
| "step": 2828 |
| }, |
| { |
| "epoch": 0.9838345211194159, |
| "grad_norm": 0.8910777454759726, |
| "learning_rate": 0.00012013368766684679, |
| "loss": 0.4932, |
| "step": 2830 |
| }, |
| { |
| "epoch": 0.9845298105336346, |
| "grad_norm": 1.1134058361712267, |
| "learning_rate": 0.0001200147679429263, |
| "loss": 0.4397, |
| "step": 2832 |
| }, |
| { |
| "epoch": 0.9852250999478533, |
| "grad_norm": 1.4603822803174207, |
| "learning_rate": 0.0001198958187258186, |
| "loss": 0.5092, |
| "step": 2834 |
| }, |
| { |
| "epoch": 0.9859203893620719, |
| "grad_norm": 1.543670422145342, |
| "learning_rate": 0.00011977684019080388, |
| "loss": 0.4849, |
| "step": 2836 |
| }, |
| { |
| "epoch": 0.9866156787762906, |
| "grad_norm": 1.4552217324250358, |
| "learning_rate": 0.0001196578325132055, |
| "loss": 0.4884, |
| "step": 2838 |
| }, |
| { |
| "epoch": 0.9873109681905093, |
| "grad_norm": 1.383861083085062, |
| "learning_rate": 0.0001195387958683897, |
| "loss": 0.3629, |
| "step": 2840 |
| }, |
| { |
| "epoch": 0.9880062576047279, |
| "grad_norm": 1.0735309159208644, |
| "learning_rate": 0.00011941973043176549, |
| "loss": 0.5393, |
| "step": 2842 |
| }, |
| { |
| "epoch": 0.9887015470189466, |
| "grad_norm": 1.0426992770965624, |
| "learning_rate": 0.0001193006363787843, |
| "loss": 0.4881, |
| "step": 2844 |
| }, |
| { |
| "epoch": 0.9893968364331653, |
| "grad_norm": 1.3546987973664115, |
| "learning_rate": 0.00011918151388493966, |
| "loss": 0.6221, |
| "step": 2846 |
| }, |
| { |
| "epoch": 0.9900921258473839, |
| "grad_norm": 0.9148682244949863, |
| "learning_rate": 0.0001190623631257671, |
| "loss": 0.4166, |
| "step": 2848 |
| }, |
| { |
| "epoch": 0.9907874152616026, |
| "grad_norm": 1.286288275163198, |
| "learning_rate": 0.00011894318427684376, |
| "loss": 0.427, |
| "step": 2850 |
| }, |
| { |
| "epoch": 0.9907874152616026, |
| "eval_loss": 0.629009485244751, |
| "eval_runtime": 370.6874, |
| "eval_samples_per_second": 13.062, |
| "eval_steps_per_second": 0.41, |
| "step": 2850 |
| }, |
| { |
| "epoch": 0.9914827046758213, |
| "grad_norm": 1.3180773598651125, |
| "learning_rate": 0.0001188239775137881, |
| "loss": 0.4051, |
| "step": 2852 |
| }, |
| { |
| "epoch": 0.9921779940900399, |
| "grad_norm": 0.8747517343762433, |
| "learning_rate": 0.00011870474301225984, |
| "loss": 0.4323, |
| "step": 2854 |
| }, |
| { |
| "epoch": 0.9928732835042586, |
| "grad_norm": 1.5456791849636662, |
| "learning_rate": 0.00011858548094795954, |
| "loss": 0.4543, |
| "step": 2856 |
| }, |
| { |
| "epoch": 0.9935685729184773, |
| "grad_norm": 1.6784287649184517, |
| "learning_rate": 0.00011846619149662827, |
| "loss": 0.503, |
| "step": 2858 |
| }, |
| { |
| "epoch": 0.994263862332696, |
| "grad_norm": 1.4790633894661258, |
| "learning_rate": 0.00011834687483404763, |
| "loss": 0.503, |
| "step": 2860 |
| }, |
| { |
| "epoch": 0.9949591517469146, |
| "grad_norm": 1.780662023668694, |
| "learning_rate": 0.00011822753113603917, |
| "loss": 0.5344, |
| "step": 2862 |
| }, |
| { |
| "epoch": 0.9956544411611333, |
| "grad_norm": 1.8480048781379417, |
| "learning_rate": 0.00011810816057846432, |
| "loss": 0.6558, |
| "step": 2864 |
| }, |
| { |
| "epoch": 0.996349730575352, |
| "grad_norm": 1.3132278834941422, |
| "learning_rate": 0.00011798876333722416, |
| "loss": 0.5066, |
| "step": 2866 |
| }, |
| { |
| "epoch": 0.9970450199895706, |
| "grad_norm": 1.075594771509589, |
| "learning_rate": 0.00011786933958825904, |
| "loss": 0.4895, |
| "step": 2868 |
| }, |
| { |
| "epoch": 0.9977403094037893, |
| "grad_norm": 1.3769121244280504, |
| "learning_rate": 0.00011774988950754828, |
| "loss": 0.4602, |
| "step": 2870 |
| }, |
| { |
| "epoch": 0.998435598818008, |
| "grad_norm": 0.7606675238956178, |
| "learning_rate": 0.00011763041327111017, |
| "loss": 0.4272, |
| "step": 2872 |
| }, |
| { |
| "epoch": 0.9991308882322266, |
| "grad_norm": 1.0614427770230186, |
| "learning_rate": 0.00011751091105500148, |
| "loss": 0.4829, |
| "step": 2874 |
| }, |
| { |
| "epoch": 0.9998261776464453, |
| "grad_norm": 1.0472604077832082, |
| "learning_rate": 0.00011739138303531714, |
| "loss": 0.4973, |
| "step": 2876 |
| }, |
| { |
| "epoch": 1.0006952894142187, |
| "grad_norm": 0.8030004827307021, |
| "learning_rate": 0.0001172718293881903, |
| "loss": 0.5214, |
| "step": 2878 |
| }, |
| { |
| "epoch": 1.0013905788284374, |
| "grad_norm": 1.2035430089324888, |
| "learning_rate": 0.0001171522502897917, |
| "loss": 0.5692, |
| "step": 2880 |
| }, |
| { |
| "epoch": 1.002085868242656, |
| "grad_norm": 0.877246561662359, |
| "learning_rate": 0.00011703264591632974, |
| "loss": 0.3358, |
| "step": 2882 |
| }, |
| { |
| "epoch": 1.0027811576568746, |
| "grad_norm": 1.1099318959683315, |
| "learning_rate": 0.00011691301644404998, |
| "loss": 0.4113, |
| "step": 2884 |
| }, |
| { |
| "epoch": 1.0034764470710933, |
| "grad_norm": 1.0014213184547873, |
| "learning_rate": 0.00011679336204923492, |
| "loss": 0.3723, |
| "step": 2886 |
| }, |
| { |
| "epoch": 1.004171736485312, |
| "grad_norm": 0.9704642688079621, |
| "learning_rate": 0.00011667368290820389, |
| "loss": 0.2382, |
| "step": 2888 |
| }, |
| { |
| "epoch": 1.0048670258995307, |
| "grad_norm": 0.6311432514427833, |
| "learning_rate": 0.00011655397919731264, |
| "loss": 0.264, |
| "step": 2890 |
| }, |
| { |
| "epoch": 1.0055623153137494, |
| "grad_norm": 1.3297287791953407, |
| "learning_rate": 0.00011643425109295307, |
| "loss": 0.3517, |
| "step": 2892 |
| }, |
| { |
| "epoch": 1.006257604727968, |
| "grad_norm": 1.277582625895487, |
| "learning_rate": 0.00011631449877155311, |
| "loss": 0.3545, |
| "step": 2894 |
| }, |
| { |
| "epoch": 1.0069528941421866, |
| "grad_norm": 1.191996914639171, |
| "learning_rate": 0.00011619472240957634, |
| "loss": 0.3484, |
| "step": 2896 |
| }, |
| { |
| "epoch": 1.0076481835564053, |
| "grad_norm": 2.478742012406443, |
| "learning_rate": 0.00011607492218352178, |
| "loss": 0.3328, |
| "step": 2898 |
| }, |
| { |
| "epoch": 1.008343472970624, |
| "grad_norm": 3.154409064860738, |
| "learning_rate": 0.00011595509826992357, |
| "loss": 0.5825, |
| "step": 2900 |
| }, |
| { |
| "epoch": 1.0090387623848427, |
| "grad_norm": 2.1432175569882506, |
| "learning_rate": 0.00011583525084535083, |
| "loss": 0.3379, |
| "step": 2902 |
| }, |
| { |
| "epoch": 1.0097340517990614, |
| "grad_norm": 2.0514351597541025, |
| "learning_rate": 0.00011571538008640724, |
| "loss": 0.3369, |
| "step": 2904 |
| }, |
| { |
| "epoch": 1.01042934121328, |
| "grad_norm": 1.4488593406584533, |
| "learning_rate": 0.00011559548616973093, |
| "loss": 0.3111, |
| "step": 2906 |
| }, |
| { |
| "epoch": 1.0111246306274988, |
| "grad_norm": 1.9029926062831013, |
| "learning_rate": 0.0001154755692719941, |
| "loss": 0.3074, |
| "step": 2908 |
| }, |
| { |
| "epoch": 1.0118199200417173, |
| "grad_norm": 2.4361729668401915, |
| "learning_rate": 0.00011535562956990287, |
| "loss": 0.3903, |
| "step": 2910 |
| }, |
| { |
| "epoch": 1.012515209455936, |
| "grad_norm": 1.2486164104056807, |
| "learning_rate": 0.00011523566724019699, |
| "loss": 0.2979, |
| "step": 2912 |
| }, |
| { |
| "epoch": 1.0132104988701547, |
| "grad_norm": 2.671143152734771, |
| "learning_rate": 0.0001151156824596494, |
| "loss": 0.4503, |
| "step": 2914 |
| }, |
| { |
| "epoch": 1.0139057882843734, |
| "grad_norm": 1.7715037156920974, |
| "learning_rate": 0.0001149956754050663, |
| "loss": 0.4123, |
| "step": 2916 |
| }, |
| { |
| "epoch": 1.014601077698592, |
| "grad_norm": 2.3080827391994823, |
| "learning_rate": 0.00011487564625328664, |
| "loss": 0.4243, |
| "step": 2918 |
| }, |
| { |
| "epoch": 1.0152963671128108, |
| "grad_norm": 1.0738841028614674, |
| "learning_rate": 0.00011475559518118191, |
| "loss": 0.2884, |
| "step": 2920 |
| }, |
| { |
| "epoch": 1.0159916565270293, |
| "grad_norm": 1.7658509745092148, |
| "learning_rate": 0.00011463552236565591, |
| "loss": 0.3778, |
| "step": 2922 |
| }, |
| { |
| "epoch": 1.016686945941248, |
| "grad_norm": 1.5214276252745638, |
| "learning_rate": 0.00011451542798364452, |
| "loss": 0.3303, |
| "step": 2924 |
| }, |
| { |
| "epoch": 1.0173822353554667, |
| "grad_norm": 1.4572865433131137, |
| "learning_rate": 0.00011439531221211535, |
| "loss": 0.3729, |
| "step": 2926 |
| }, |
| { |
| "epoch": 1.0180775247696854, |
| "grad_norm": 1.6210350391088704, |
| "learning_rate": 0.00011427517522806757, |
| "loss": 0.3433, |
| "step": 2928 |
| }, |
| { |
| "epoch": 1.018772814183904, |
| "grad_norm": 1.0780851626803933, |
| "learning_rate": 0.00011415501720853156, |
| "loss": 0.2906, |
| "step": 2930 |
| }, |
| { |
| "epoch": 1.0194681035981228, |
| "grad_norm": 1.299775348765843, |
| "learning_rate": 0.00011403483833056876, |
| "loss": 0.3521, |
| "step": 2932 |
| }, |
| { |
| "epoch": 1.0201633930123415, |
| "grad_norm": 1.7570949365431192, |
| "learning_rate": 0.00011391463877127126, |
| "loss": 0.3818, |
| "step": 2934 |
| }, |
| { |
| "epoch": 1.02085868242656, |
| "grad_norm": 1.1493091777151956, |
| "learning_rate": 0.00011379441870776177, |
| "loss": 0.3145, |
| "step": 2936 |
| }, |
| { |
| "epoch": 1.0215539718407787, |
| "grad_norm": 0.6227735319725197, |
| "learning_rate": 0.00011367417831719305, |
| "loss": 0.2538, |
| "step": 2938 |
| }, |
| { |
| "epoch": 1.0222492612549974, |
| "grad_norm": 2.981206330497685, |
| "learning_rate": 0.00011355391777674791, |
| "loss": 0.4812, |
| "step": 2940 |
| }, |
| { |
| "epoch": 1.022944550669216, |
| "grad_norm": 2.046572491258154, |
| "learning_rate": 0.0001134336372636388, |
| "loss": 0.2542, |
| "step": 2942 |
| }, |
| { |
| "epoch": 1.0236398400834348, |
| "grad_norm": 1.795040362172509, |
| "learning_rate": 0.00011331333695510766, |
| "loss": 0.3658, |
| "step": 2944 |
| }, |
| { |
| "epoch": 1.0243351294976535, |
| "grad_norm": 1.4501160353545008, |
| "learning_rate": 0.00011319301702842557, |
| "loss": 0.3168, |
| "step": 2946 |
| }, |
| { |
| "epoch": 1.025030418911872, |
| "grad_norm": 1.5126318616897827, |
| "learning_rate": 0.0001130726776608925, |
| "loss": 0.3163, |
| "step": 2948 |
| }, |
| { |
| "epoch": 1.0257257083260907, |
| "grad_norm": 1.555522159696809, |
| "learning_rate": 0.00011295231902983712, |
| "loss": 0.2823, |
| "step": 2950 |
| }, |
| { |
| "epoch": 1.0264209977403094, |
| "grad_norm": 1.1751331166171803, |
| "learning_rate": 0.00011283194131261643, |
| "loss": 0.207, |
| "step": 2952 |
| }, |
| { |
| "epoch": 1.027116287154528, |
| "grad_norm": 2.0788516149416587, |
| "learning_rate": 0.00011271154468661555, |
| "loss": 0.4004, |
| "step": 2954 |
| }, |
| { |
| "epoch": 1.0278115765687468, |
| "grad_norm": 2.1492678206006754, |
| "learning_rate": 0.00011259112932924751, |
| "loss": 0.333, |
| "step": 2956 |
| }, |
| { |
| "epoch": 1.0285068659829655, |
| "grad_norm": 1.6372533276795482, |
| "learning_rate": 0.00011247069541795294, |
| "loss": 0.3118, |
| "step": 2958 |
| }, |
| { |
| "epoch": 1.0292021553971842, |
| "grad_norm": 2.2242948471766333, |
| "learning_rate": 0.00011235024313019978, |
| "loss": 0.2566, |
| "step": 2960 |
| }, |
| { |
| "epoch": 1.0298974448114027, |
| "grad_norm": 2.024230456754769, |
| "learning_rate": 0.00011222977264348305, |
| "loss": 0.3571, |
| "step": 2962 |
| }, |
| { |
| "epoch": 1.0305927342256214, |
| "grad_norm": 1.7637916725410678, |
| "learning_rate": 0.00011210928413532462, |
| "loss": 0.2523, |
| "step": 2964 |
| }, |
| { |
| "epoch": 1.03128802363984, |
| "grad_norm": 1.924942195361263, |
| "learning_rate": 0.00011198877778327289, |
| "loss": 0.2846, |
| "step": 2966 |
| }, |
| { |
| "epoch": 1.0319833130540588, |
| "grad_norm": 1.212829490834995, |
| "learning_rate": 0.00011186825376490253, |
| "loss": 0.2769, |
| "step": 2968 |
| }, |
| { |
| "epoch": 1.0326786024682775, |
| "grad_norm": 1.99251348753868, |
| "learning_rate": 0.0001117477122578143, |
| "loss": 0.2586, |
| "step": 2970 |
| }, |
| { |
| "epoch": 1.0333738918824962, |
| "grad_norm": 2.199610187348861, |
| "learning_rate": 0.00011162715343963473, |
| "loss": 0.219, |
| "step": 2972 |
| }, |
| { |
| "epoch": 1.0340691812967147, |
| "grad_norm": 2.1366147288309083, |
| "learning_rate": 0.00011150657748801578, |
| "loss": 0.4997, |
| "step": 2974 |
| }, |
| { |
| "epoch": 1.0347644707109334, |
| "grad_norm": 2.3719812564822362, |
| "learning_rate": 0.00011138598458063473, |
| "loss": 0.4039, |
| "step": 2976 |
| }, |
| { |
| "epoch": 1.035459760125152, |
| "grad_norm": 2.7532578468397526, |
| "learning_rate": 0.00011126537489519384, |
| "loss": 0.3589, |
| "step": 2978 |
| }, |
| { |
| "epoch": 1.0361550495393708, |
| "grad_norm": 1.8187149854764009, |
| "learning_rate": 0.00011114474860942005, |
| "loss": 0.3449, |
| "step": 2980 |
| }, |
| { |
| "epoch": 1.0368503389535895, |
| "grad_norm": 2.735044582952234, |
| "learning_rate": 0.0001110241059010648, |
| "loss": 0.4433, |
| "step": 2982 |
| }, |
| { |
| "epoch": 1.0375456283678082, |
| "grad_norm": 2.0532417560276146, |
| "learning_rate": 0.00011090344694790371, |
| "loss": 0.3922, |
| "step": 2984 |
| }, |
| { |
| "epoch": 1.0382409177820269, |
| "grad_norm": 1.6360035973449651, |
| "learning_rate": 0.00011078277192773641, |
| "loss": 0.264, |
| "step": 2986 |
| }, |
| { |
| "epoch": 1.0389362071962454, |
| "grad_norm": 1.0548243473635455, |
| "learning_rate": 0.00011066208101838604, |
| "loss": 0.3538, |
| "step": 2988 |
| }, |
| { |
| "epoch": 1.039631496610464, |
| "grad_norm": 1.0814694707236894, |
| "learning_rate": 0.00011054137439769934, |
| "loss": 0.3268, |
| "step": 2990 |
| }, |
| { |
| "epoch": 1.0403267860246828, |
| "grad_norm": 1.325877882349333, |
| "learning_rate": 0.00011042065224354607, |
| "loss": 0.329, |
| "step": 2992 |
| }, |
| { |
| "epoch": 1.0410220754389015, |
| "grad_norm": 1.2166101793088029, |
| "learning_rate": 0.00011029991473381897, |
| "loss": 0.3839, |
| "step": 2994 |
| }, |
| { |
| "epoch": 1.0417173648531202, |
| "grad_norm": 1.4941391669061694, |
| "learning_rate": 0.00011017916204643332, |
| "loss": 0.3518, |
| "step": 2996 |
| }, |
| { |
| "epoch": 1.0424126542673389, |
| "grad_norm": 1.7193756966994476, |
| "learning_rate": 0.0001100583943593269, |
| "loss": 0.3592, |
| "step": 2998 |
| }, |
| { |
| "epoch": 1.0431079436815573, |
| "grad_norm": 2.1541599899498345, |
| "learning_rate": 0.0001099376118504594, |
| "loss": 0.3107, |
| "step": 3000 |
| }, |
| { |
| "epoch": 1.0431079436815573, |
| "eval_loss": 0.6923864483833313, |
| "eval_runtime": 370.4448, |
| "eval_samples_per_second": 13.071, |
| "eval_steps_per_second": 0.41, |
| "step": 3000 |
| }, |
| { |
| "epoch": 1.043803233095776, |
| "grad_norm": 1.3154438707858545, |
| "learning_rate": 0.00010981681469781252, |
| "loss": 0.3689, |
| "step": 3002 |
| }, |
| { |
| "epoch": 1.0444985225099948, |
| "grad_norm": 1.3974965614228347, |
| "learning_rate": 0.00010969600307938945, |
| "loss": 0.273, |
| "step": 3004 |
| }, |
| { |
| "epoch": 1.0451938119242135, |
| "grad_norm": 2.130922469831274, |
| "learning_rate": 0.0001095751771732147, |
| "loss": 0.4886, |
| "step": 3006 |
| }, |
| { |
| "epoch": 1.0458891013384322, |
| "grad_norm": 1.7816159106806282, |
| "learning_rate": 0.00010945433715733391, |
| "loss": 0.3208, |
| "step": 3008 |
| }, |
| { |
| "epoch": 1.0465843907526509, |
| "grad_norm": 1.699901985683691, |
| "learning_rate": 0.00010933348320981341, |
| "loss": 0.433, |
| "step": 3010 |
| }, |
| { |
| "epoch": 1.0472796801668696, |
| "grad_norm": 2.251185242369719, |
| "learning_rate": 0.00010921261550874014, |
| "loss": 0.3986, |
| "step": 3012 |
| }, |
| { |
| "epoch": 1.047974969581088, |
| "grad_norm": 1.5744762113883837, |
| "learning_rate": 0.00010909173423222125, |
| "loss": 0.3524, |
| "step": 3014 |
| }, |
| { |
| "epoch": 1.0486702589953067, |
| "grad_norm": 1.0266582810737757, |
| "learning_rate": 0.0001089708395583839, |
| "loss": 0.2745, |
| "step": 3016 |
| }, |
| { |
| "epoch": 1.0493655484095255, |
| "grad_norm": 1.0487726667074528, |
| "learning_rate": 0.00010884993166537504, |
| "loss": 0.2744, |
| "step": 3018 |
| }, |
| { |
| "epoch": 1.0500608378237442, |
| "grad_norm": 1.6355652049596066, |
| "learning_rate": 0.00010872901073136111, |
| "loss": 0.4148, |
| "step": 3020 |
| }, |
| { |
| "epoch": 1.0507561272379629, |
| "grad_norm": 2.4928247247897692, |
| "learning_rate": 0.00010860807693452761, |
| "loss": 0.4917, |
| "step": 3022 |
| }, |
| { |
| "epoch": 1.0514514166521816, |
| "grad_norm": 1.6925879868016276, |
| "learning_rate": 0.00010848713045307923, |
| "loss": 0.3287, |
| "step": 3024 |
| }, |
| { |
| "epoch": 1.0521467060664, |
| "grad_norm": 1.2901087786275507, |
| "learning_rate": 0.00010836617146523915, |
| "loss": 0.3005, |
| "step": 3026 |
| }, |
| { |
| "epoch": 1.0528419954806187, |
| "grad_norm": 2.2323567950164995, |
| "learning_rate": 0.00010824520014924911, |
| "loss": 0.3863, |
| "step": 3028 |
| }, |
| { |
| "epoch": 1.0535372848948374, |
| "grad_norm": 1.2447828077577623, |
| "learning_rate": 0.00010812421668336892, |
| "loss": 0.2803, |
| "step": 3030 |
| }, |
| { |
| "epoch": 1.0542325743090561, |
| "grad_norm": 0.8464890827543972, |
| "learning_rate": 0.00010800322124587639, |
| "loss": 0.2337, |
| "step": 3032 |
| }, |
| { |
| "epoch": 1.0549278637232749, |
| "grad_norm": 0.9444985947269403, |
| "learning_rate": 0.00010788221401506687, |
| "loss": 0.1971, |
| "step": 3034 |
| }, |
| { |
| "epoch": 1.0556231531374936, |
| "grad_norm": 3.0444705523758193, |
| "learning_rate": 0.00010776119516925316, |
| "loss": 0.443, |
| "step": 3036 |
| }, |
| { |
| "epoch": 1.0563184425517123, |
| "grad_norm": 2.2652404241479727, |
| "learning_rate": 0.00010764016488676518, |
| "loss": 0.4478, |
| "step": 3038 |
| }, |
| { |
| "epoch": 1.0570137319659307, |
| "grad_norm": 1.2897111265446637, |
| "learning_rate": 0.00010751912334594965, |
| "loss": 0.2117, |
| "step": 3040 |
| }, |
| { |
| "epoch": 1.0577090213801494, |
| "grad_norm": 2.089718374050333, |
| "learning_rate": 0.00010739807072516995, |
| "loss": 0.3229, |
| "step": 3042 |
| }, |
| { |
| "epoch": 1.0584043107943681, |
| "grad_norm": 2.4601669056978435, |
| "learning_rate": 0.00010727700720280566, |
| "loss": 0.3973, |
| "step": 3044 |
| }, |
| { |
| "epoch": 1.0590996002085868, |
| "grad_norm": 1.4530054653221605, |
| "learning_rate": 0.00010715593295725259, |
| "loss": 0.3049, |
| "step": 3046 |
| }, |
| { |
| "epoch": 1.0597948896228055, |
| "grad_norm": 2.6285419130743564, |
| "learning_rate": 0.00010703484816692227, |
| "loss": 0.331, |
| "step": 3048 |
| }, |
| { |
| "epoch": 1.0604901790370242, |
| "grad_norm": 1.2936401916909233, |
| "learning_rate": 0.0001069137530102417, |
| "loss": 0.3232, |
| "step": 3050 |
| }, |
| { |
| "epoch": 1.0611854684512427, |
| "grad_norm": 1.4072117960980053, |
| "learning_rate": 0.00010679264766565329, |
| "loss": 0.2999, |
| "step": 3052 |
| }, |
| { |
| "epoch": 1.0618807578654614, |
| "grad_norm": 1.4724689492891376, |
| "learning_rate": 0.00010667153231161438, |
| "loss": 0.3069, |
| "step": 3054 |
| }, |
| { |
| "epoch": 1.0625760472796801, |
| "grad_norm": 1.523645819734336, |
| "learning_rate": 0.00010655040712659706, |
| "loss": 0.4126, |
| "step": 3056 |
| }, |
| { |
| "epoch": 1.0632713366938988, |
| "grad_norm": 1.116148852344772, |
| "learning_rate": 0.00010642927228908792, |
| "loss": 0.2886, |
| "step": 3058 |
| }, |
| { |
| "epoch": 1.0639666261081175, |
| "grad_norm": 0.6609181844137914, |
| "learning_rate": 0.00010630812797758782, |
| "loss": 0.1907, |
| "step": 3060 |
| }, |
| { |
| "epoch": 1.0646619155223362, |
| "grad_norm": 1.3746863098376871, |
| "learning_rate": 0.00010618697437061149, |
| "loss": 0.3442, |
| "step": 3062 |
| }, |
| { |
| "epoch": 1.065357204936555, |
| "grad_norm": 2.4423758266343394, |
| "learning_rate": 0.00010606581164668744, |
| "loss": 0.2808, |
| "step": 3064 |
| }, |
| { |
| "epoch": 1.0660524943507734, |
| "grad_norm": 1.7875712927098957, |
| "learning_rate": 0.00010594463998435757, |
| "loss": 0.2489, |
| "step": 3066 |
| }, |
| { |
| "epoch": 1.0667477837649921, |
| "grad_norm": 1.6433427674624645, |
| "learning_rate": 0.00010582345956217693, |
| "loss": 0.2301, |
| "step": 3068 |
| }, |
| { |
| "epoch": 1.0674430731792108, |
| "grad_norm": 1.4478686584577465, |
| "learning_rate": 0.00010570227055871357, |
| "loss": 0.3575, |
| "step": 3070 |
| }, |
| { |
| "epoch": 1.0681383625934295, |
| "grad_norm": 1.5346837447194575, |
| "learning_rate": 0.00010558107315254811, |
| "loss": 0.2515, |
| "step": 3072 |
| }, |
| { |
| "epoch": 1.0688336520076482, |
| "grad_norm": 2.7430648047331445, |
| "learning_rate": 0.00010545986752227351, |
| "loss": 0.3167, |
| "step": 3074 |
| }, |
| { |
| "epoch": 1.069528941421867, |
| "grad_norm": 1.2241028266722616, |
| "learning_rate": 0.00010533865384649495, |
| "loss": 0.2117, |
| "step": 3076 |
| }, |
| { |
| "epoch": 1.0702242308360854, |
| "grad_norm": 1.674653635413672, |
| "learning_rate": 0.00010521743230382946, |
| "loss": 0.3127, |
| "step": 3078 |
| }, |
| { |
| "epoch": 1.0709195202503041, |
| "grad_norm": 1.4735972752313786, |
| "learning_rate": 0.00010509620307290554, |
| "loss": 0.2728, |
| "step": 3080 |
| }, |
| { |
| "epoch": 1.0716148096645228, |
| "grad_norm": 2.1196399828192947, |
| "learning_rate": 0.00010497496633236314, |
| "loss": 0.2741, |
| "step": 3082 |
| }, |
| { |
| "epoch": 1.0723100990787415, |
| "grad_norm": 1.8274749623966444, |
| "learning_rate": 0.00010485372226085324, |
| "loss": 0.3354, |
| "step": 3084 |
| }, |
| { |
| "epoch": 1.0730053884929602, |
| "grad_norm": 1.4788924899318916, |
| "learning_rate": 0.00010473247103703761, |
| "loss": 0.3032, |
| "step": 3086 |
| }, |
| { |
| "epoch": 1.073700677907179, |
| "grad_norm": 1.96101750955112, |
| "learning_rate": 0.00010461121283958861, |
| "loss": 0.3423, |
| "step": 3088 |
| }, |
| { |
| "epoch": 1.0743959673213976, |
| "grad_norm": 1.710643053427844, |
| "learning_rate": 0.0001044899478471888, |
| "loss": 0.3627, |
| "step": 3090 |
| }, |
| { |
| "epoch": 1.0750912567356161, |
| "grad_norm": 0.7802744078492159, |
| "learning_rate": 0.00010436867623853079, |
| "loss": 0.2197, |
| "step": 3092 |
| }, |
| { |
| "epoch": 1.0757865461498348, |
| "grad_norm": 1.1912332482189396, |
| "learning_rate": 0.00010424739819231699, |
| "loss": 0.2935, |
| "step": 3094 |
| }, |
| { |
| "epoch": 1.0764818355640535, |
| "grad_norm": 1.2594281043358158, |
| "learning_rate": 0.00010412611388725917, |
| "loss": 0.2104, |
| "step": 3096 |
| }, |
| { |
| "epoch": 1.0771771249782722, |
| "grad_norm": 1.906616083048897, |
| "learning_rate": 0.00010400482350207843, |
| "loss": 0.3789, |
| "step": 3098 |
| }, |
| { |
| "epoch": 1.077872414392491, |
| "grad_norm": 0.9873117879124189, |
| "learning_rate": 0.00010388352721550482, |
| "loss": 0.3461, |
| "step": 3100 |
| }, |
| { |
| "epoch": 1.0785677038067096, |
| "grad_norm": 1.9556908244222917, |
| "learning_rate": 0.00010376222520627705, |
| "loss": 0.3783, |
| "step": 3102 |
| }, |
| { |
| "epoch": 1.0792629932209281, |
| "grad_norm": 0.9583766378982667, |
| "learning_rate": 0.00010364091765314225, |
| "loss": 0.3495, |
| "step": 3104 |
| }, |
| { |
| "epoch": 1.0799582826351468, |
| "grad_norm": 0.8968787755905534, |
| "learning_rate": 0.0001035196047348558, |
| "loss": 0.2365, |
| "step": 3106 |
| }, |
| { |
| "epoch": 1.0806535720493655, |
| "grad_norm": 2.0555123578910126, |
| "learning_rate": 0.00010339828663018087, |
| "loss": 0.3628, |
| "step": 3108 |
| }, |
| { |
| "epoch": 1.0813488614635842, |
| "grad_norm": 1.54992005628239, |
| "learning_rate": 0.00010327696351788838, |
| "loss": 0.2991, |
| "step": 3110 |
| }, |
| { |
| "epoch": 1.082044150877803, |
| "grad_norm": 2.158298643054877, |
| "learning_rate": 0.00010315563557675659, |
| "loss": 0.3089, |
| "step": 3112 |
| }, |
| { |
| "epoch": 1.0827394402920216, |
| "grad_norm": 0.8473973867847717, |
| "learning_rate": 0.0001030343029855708, |
| "loss": 0.2134, |
| "step": 3114 |
| }, |
| { |
| "epoch": 1.0834347297062403, |
| "grad_norm": 1.395215252675235, |
| "learning_rate": 0.00010291296592312331, |
| "loss": 0.2709, |
| "step": 3116 |
| }, |
| { |
| "epoch": 1.0841300191204588, |
| "grad_norm": 1.170273796430348, |
| "learning_rate": 0.0001027916245682129, |
| "loss": 0.3682, |
| "step": 3118 |
| }, |
| { |
| "epoch": 1.0848253085346775, |
| "grad_norm": 1.1300135663439852, |
| "learning_rate": 0.0001026702790996447, |
| "loss": 0.2398, |
| "step": 3120 |
| }, |
| { |
| "epoch": 1.0855205979488962, |
| "grad_norm": 1.4623244815140315, |
| "learning_rate": 0.00010254892969622993, |
| "loss": 0.3004, |
| "step": 3122 |
| }, |
| { |
| "epoch": 1.086215887363115, |
| "grad_norm": 1.863683822162162, |
| "learning_rate": 0.00010242757653678554, |
| "loss": 0.3608, |
| "step": 3124 |
| }, |
| { |
| "epoch": 1.0869111767773336, |
| "grad_norm": 1.6081535163252574, |
| "learning_rate": 0.00010230621980013413, |
| "loss": 0.2322, |
| "step": 3126 |
| }, |
| { |
| "epoch": 1.0876064661915523, |
| "grad_norm": 2.1399726299977826, |
| "learning_rate": 0.00010218485966510342, |
| "loss": 0.2012, |
| "step": 3128 |
| }, |
| { |
| "epoch": 1.0883017556057708, |
| "grad_norm": 1.9966195759490826, |
| "learning_rate": 0.00010206349631052628, |
| "loss": 0.3079, |
| "step": 3130 |
| }, |
| { |
| "epoch": 1.0889970450199895, |
| "grad_norm": 1.6601174124659477, |
| "learning_rate": 0.00010194212991524023, |
| "loss": 0.2804, |
| "step": 3132 |
| }, |
| { |
| "epoch": 1.0896923344342082, |
| "grad_norm": 1.641919684164699, |
| "learning_rate": 0.00010182076065808735, |
| "loss": 0.2881, |
| "step": 3134 |
| }, |
| { |
| "epoch": 1.090387623848427, |
| "grad_norm": 1.8908046042658808, |
| "learning_rate": 0.00010169938871791382, |
| "loss": 0.1913, |
| "step": 3136 |
| }, |
| { |
| "epoch": 1.0910829132626456, |
| "grad_norm": 1.305382644398651, |
| "learning_rate": 0.0001015780142735699, |
| "loss": 0.3521, |
| "step": 3138 |
| }, |
| { |
| "epoch": 1.0917782026768643, |
| "grad_norm": 1.8515758007120933, |
| "learning_rate": 0.00010145663750390948, |
| "loss": 0.339, |
| "step": 3140 |
| }, |
| { |
| "epoch": 1.092473492091083, |
| "grad_norm": 1.6956697847050364, |
| "learning_rate": 0.00010133525858778984, |
| "loss": 0.3091, |
| "step": 3142 |
| }, |
| { |
| "epoch": 1.0931687815053015, |
| "grad_norm": 2.0234087698860588, |
| "learning_rate": 0.00010121387770407151, |
| "loss": 0.2932, |
| "step": 3144 |
| }, |
| { |
| "epoch": 1.0938640709195202, |
| "grad_norm": 2.066956734331634, |
| "learning_rate": 0.00010109249503161788, |
| "loss": 0.1633, |
| "step": 3146 |
| }, |
| { |
| "epoch": 1.094559360333739, |
| "grad_norm": 1.3457014661271918, |
| "learning_rate": 0.00010097111074929493, |
| "loss": 0.3382, |
| "step": 3148 |
| }, |
| { |
| "epoch": 1.0952546497479576, |
| "grad_norm": 1.4339800505837978, |
| "learning_rate": 0.00010084972503597108, |
| "loss": 0.2467, |
| "step": 3150 |
| }, |
| { |
| "epoch": 1.0952546497479576, |
| "eval_loss": 0.762905478477478, |
| "eval_runtime": 370.5379, |
| "eval_samples_per_second": 13.067, |
| "eval_steps_per_second": 0.41, |
| "step": 3150 |
| }, |
| { |
| "epoch": 1.0959499391621763, |
| "grad_norm": 1.828765817436499, |
| "learning_rate": 0.00010072833807051683, |
| "loss": 0.3289, |
| "step": 3152 |
| }, |
| { |
| "epoch": 1.096645228576395, |
| "grad_norm": 3.714271831784244, |
| "learning_rate": 0.00010060695003180449, |
| "loss": 0.378, |
| "step": 3154 |
| }, |
| { |
| "epoch": 1.0973405179906135, |
| "grad_norm": 1.4551298556002794, |
| "learning_rate": 0.00010048556109870806, |
| "loss": 0.3927, |
| "step": 3156 |
| }, |
| { |
| "epoch": 1.0980358074048322, |
| "grad_norm": 2.0838238662365267, |
| "learning_rate": 0.00010036417145010271, |
| "loss": 0.332, |
| "step": 3158 |
| }, |
| { |
| "epoch": 1.098731096819051, |
| "grad_norm": 2.054490472770659, |
| "learning_rate": 0.00010024278126486475, |
| "loss": 0.4106, |
| "step": 3160 |
| }, |
| { |
| "epoch": 1.0994263862332696, |
| "grad_norm": 1.9642838650911554, |
| "learning_rate": 0.00010012139072187135, |
| "loss": 0.3058, |
| "step": 3162 |
| }, |
| { |
| "epoch": 1.1001216756474883, |
| "grad_norm": 2.1547535271957123, |
| "learning_rate": 0.0001, |
| "loss": 0.3644, |
| "step": 3164 |
| }, |
| { |
| "epoch": 1.100816965061707, |
| "grad_norm": 1.1845334778546894, |
| "learning_rate": 9.987860927812869e-05, |
| "loss": 0.3526, |
| "step": 3166 |
| }, |
| { |
| "epoch": 1.1015122544759257, |
| "grad_norm": 1.3796845491714145, |
| "learning_rate": 9.975721873513524e-05, |
| "loss": 0.302, |
| "step": 3168 |
| }, |
| { |
| "epoch": 1.1022075438901442, |
| "grad_norm": 0.8010153495082397, |
| "learning_rate": 9.963582854989734e-05, |
| "loss": 0.2645, |
| "step": 3170 |
| }, |
| { |
| "epoch": 1.102902833304363, |
| "grad_norm": 1.294960262323839, |
| "learning_rate": 9.951443890129199e-05, |
| "loss": 0.2792, |
| "step": 3172 |
| }, |
| { |
| "epoch": 1.1035981227185816, |
| "grad_norm": 2.1289097921537077, |
| "learning_rate": 9.939304996819552e-05, |
| "loss": 0.3352, |
| "step": 3174 |
| }, |
| { |
| "epoch": 1.1042934121328003, |
| "grad_norm": 1.6850360072439985, |
| "learning_rate": 9.927166192948318e-05, |
| "loss": 0.3774, |
| "step": 3176 |
| }, |
| { |
| "epoch": 1.104988701547019, |
| "grad_norm": 1.110074169304674, |
| "learning_rate": 9.915027496402895e-05, |
| "loss": 0.2093, |
| "step": 3178 |
| }, |
| { |
| "epoch": 1.1056839909612377, |
| "grad_norm": 1.385494626317771, |
| "learning_rate": 9.902888925070508e-05, |
| "loss": 0.3445, |
| "step": 3180 |
| }, |
| { |
| "epoch": 1.1063792803754562, |
| "grad_norm": 1.1831702789079874, |
| "learning_rate": 9.890750496838217e-05, |
| "loss": 0.3782, |
| "step": 3182 |
| }, |
| { |
| "epoch": 1.107074569789675, |
| "grad_norm": 2.5507185623302076, |
| "learning_rate": 9.878612229592852e-05, |
| "loss": 0.3288, |
| "step": 3184 |
| }, |
| { |
| "epoch": 1.1077698592038936, |
| "grad_norm": 2.0771530967413008, |
| "learning_rate": 9.866474141221018e-05, |
| "loss": 0.4368, |
| "step": 3186 |
| }, |
| { |
| "epoch": 1.1084651486181123, |
| "grad_norm": 1.8875898588427005, |
| "learning_rate": 9.854336249609055e-05, |
| "loss": 0.3462, |
| "step": 3188 |
| }, |
| { |
| "epoch": 1.109160438032331, |
| "grad_norm": 1.514246064837921, |
| "learning_rate": 9.842198572643011e-05, |
| "loss": 0.5697, |
| "step": 3190 |
| }, |
| { |
| "epoch": 1.1098557274465497, |
| "grad_norm": 1.64620495194456, |
| "learning_rate": 9.830061128208619e-05, |
| "loss": 0.3362, |
| "step": 3192 |
| }, |
| { |
| "epoch": 1.1105510168607684, |
| "grad_norm": 1.063195255674117, |
| "learning_rate": 9.81792393419127e-05, |
| "loss": 0.2596, |
| "step": 3194 |
| }, |
| { |
| "epoch": 1.1112463062749869, |
| "grad_norm": 1.1142788831420738, |
| "learning_rate": 9.805787008475979e-05, |
| "loss": 0.2299, |
| "step": 3196 |
| }, |
| { |
| "epoch": 1.1119415956892056, |
| "grad_norm": 1.3408028307136335, |
| "learning_rate": 9.793650368947373e-05, |
| "loss": 0.323, |
| "step": 3198 |
| }, |
| { |
| "epoch": 1.1126368851034243, |
| "grad_norm": 1.361703780523066, |
| "learning_rate": 9.781514033489659e-05, |
| "loss": 0.2893, |
| "step": 3200 |
| }, |
| { |
| "epoch": 1.113332174517643, |
| "grad_norm": 1.5338153901442921, |
| "learning_rate": 9.769378019986591e-05, |
| "loss": 0.3316, |
| "step": 3202 |
| }, |
| { |
| "epoch": 1.1140274639318617, |
| "grad_norm": 1.6733154143642206, |
| "learning_rate": 9.757242346321448e-05, |
| "loss": 0.248, |
| "step": 3204 |
| }, |
| { |
| "epoch": 1.1147227533460804, |
| "grad_norm": 1.3878503580173898, |
| "learning_rate": 9.745107030377012e-05, |
| "loss": 0.3286, |
| "step": 3206 |
| }, |
| { |
| "epoch": 1.1154180427602989, |
| "grad_norm": 2.206700603890188, |
| "learning_rate": 9.732972090035532e-05, |
| "loss": 0.2821, |
| "step": 3208 |
| }, |
| { |
| "epoch": 1.1161133321745176, |
| "grad_norm": 1.6511212826150234, |
| "learning_rate": 9.720837543178711e-05, |
| "loss": 0.3501, |
| "step": 3210 |
| }, |
| { |
| "epoch": 1.1168086215887363, |
| "grad_norm": 1.3187001488444274, |
| "learning_rate": 9.70870340768767e-05, |
| "loss": 0.2984, |
| "step": 3212 |
| }, |
| { |
| "epoch": 1.117503911002955, |
| "grad_norm": 2.8397254453881535, |
| "learning_rate": 9.69656970144292e-05, |
| "loss": 0.2971, |
| "step": 3214 |
| }, |
| { |
| "epoch": 1.1181992004171737, |
| "grad_norm": 2.34429898617825, |
| "learning_rate": 9.684436442324348e-05, |
| "loss": 0.2827, |
| "step": 3216 |
| }, |
| { |
| "epoch": 1.1188944898313924, |
| "grad_norm": 2.0722177368573327, |
| "learning_rate": 9.672303648211165e-05, |
| "loss": 0.3625, |
| "step": 3218 |
| }, |
| { |
| "epoch": 1.119589779245611, |
| "grad_norm": 1.5852262824713348, |
| "learning_rate": 9.660171336981914e-05, |
| "loss": 0.1897, |
| "step": 3220 |
| }, |
| { |
| "epoch": 1.1202850686598296, |
| "grad_norm": 2.0131207439149423, |
| "learning_rate": 9.648039526514422e-05, |
| "loss": 0.2734, |
| "step": 3222 |
| }, |
| { |
| "epoch": 1.1209803580740483, |
| "grad_norm": 2.9758355267326437, |
| "learning_rate": 9.635908234685774e-05, |
| "loss": 0.342, |
| "step": 3224 |
| }, |
| { |
| "epoch": 1.121675647488267, |
| "grad_norm": 1.8808493973435643, |
| "learning_rate": 9.623777479372296e-05, |
| "loss": 0.3068, |
| "step": 3226 |
| }, |
| { |
| "epoch": 1.1223709369024857, |
| "grad_norm": 1.6461222431744804, |
| "learning_rate": 9.61164727844952e-05, |
| "loss": 0.3143, |
| "step": 3228 |
| }, |
| { |
| "epoch": 1.1230662263167044, |
| "grad_norm": 0.8883976184476399, |
| "learning_rate": 9.59951764979216e-05, |
| "loss": 0.2162, |
| "step": 3230 |
| }, |
| { |
| "epoch": 1.123761515730923, |
| "grad_norm": 1.7372908597885963, |
| "learning_rate": 9.587388611274086e-05, |
| "loss": 0.3245, |
| "step": 3232 |
| }, |
| { |
| "epoch": 1.1244568051451416, |
| "grad_norm": 2.2525115429964897, |
| "learning_rate": 9.575260180768303e-05, |
| "loss": 0.3446, |
| "step": 3234 |
| }, |
| { |
| "epoch": 1.1251520945593603, |
| "grad_norm": 1.4178356051165701, |
| "learning_rate": 9.56313237614692e-05, |
| "loss": 0.2324, |
| "step": 3236 |
| }, |
| { |
| "epoch": 1.125847383973579, |
| "grad_norm": 1.5693811571076641, |
| "learning_rate": 9.551005215281123e-05, |
| "loss": 0.2195, |
| "step": 3238 |
| }, |
| { |
| "epoch": 1.1265426733877977, |
| "grad_norm": 2.0288466784320893, |
| "learning_rate": 9.538878716041141e-05, |
| "loss": 0.1936, |
| "step": 3240 |
| }, |
| { |
| "epoch": 1.1272379628020164, |
| "grad_norm": 2.8391123786621693, |
| "learning_rate": 9.52675289629624e-05, |
| "loss": 0.5663, |
| "step": 3242 |
| }, |
| { |
| "epoch": 1.127933252216235, |
| "grad_norm": 2.280040235856411, |
| "learning_rate": 9.514627773914679e-05, |
| "loss": 0.2267, |
| "step": 3244 |
| }, |
| { |
| "epoch": 1.1286285416304538, |
| "grad_norm": 2.362155084767197, |
| "learning_rate": 9.502503366763688e-05, |
| "loss": 0.3304, |
| "step": 3246 |
| }, |
| { |
| "epoch": 1.1293238310446723, |
| "grad_norm": 2.8247016482404623, |
| "learning_rate": 9.490379692709447e-05, |
| "loss": 0.4109, |
| "step": 3248 |
| }, |
| { |
| "epoch": 1.130019120458891, |
| "grad_norm": 1.6034607128714775, |
| "learning_rate": 9.47825676961706e-05, |
| "loss": 0.3215, |
| "step": 3250 |
| }, |
| { |
| "epoch": 1.1307144098731097, |
| "grad_norm": 1.5236567005284432, |
| "learning_rate": 9.466134615350506e-05, |
| "loss": 0.3018, |
| "step": 3252 |
| }, |
| { |
| "epoch": 1.1314096992873284, |
| "grad_norm": 1.8809233107902477, |
| "learning_rate": 9.454013247772652e-05, |
| "loss": 0.316, |
| "step": 3254 |
| }, |
| { |
| "epoch": 1.132104988701547, |
| "grad_norm": 1.9337629983851292, |
| "learning_rate": 9.441892684745193e-05, |
| "loss": 0.4087, |
| "step": 3256 |
| }, |
| { |
| "epoch": 1.1328002781157658, |
| "grad_norm": 1.2632097697654463, |
| "learning_rate": 9.429772944128643e-05, |
| "loss": 0.2446, |
| "step": 3258 |
| }, |
| { |
| "epoch": 1.1334955675299843, |
| "grad_norm": 1.7648187744502615, |
| "learning_rate": 9.417654043782305e-05, |
| "loss": 0.3, |
| "step": 3260 |
| }, |
| { |
| "epoch": 1.134190856944203, |
| "grad_norm": 1.99223862259603, |
| "learning_rate": 9.405536001564247e-05, |
| "loss": 0.1982, |
| "step": 3262 |
| }, |
| { |
| "epoch": 1.1348861463584217, |
| "grad_norm": 2.2338536845182033, |
| "learning_rate": 9.393418835331257e-05, |
| "loss": 0.2785, |
| "step": 3264 |
| }, |
| { |
| "epoch": 1.1355814357726404, |
| "grad_norm": 1.9607863958619112, |
| "learning_rate": 9.381302562938852e-05, |
| "loss": 0.2154, |
| "step": 3266 |
| }, |
| { |
| "epoch": 1.136276725186859, |
| "grad_norm": 1.382663969286361, |
| "learning_rate": 9.36918720224122e-05, |
| "loss": 0.2203, |
| "step": 3268 |
| }, |
| { |
| "epoch": 1.1369720146010778, |
| "grad_norm": 2.5893875856633755, |
| "learning_rate": 9.357072771091208e-05, |
| "loss": 0.3986, |
| "step": 3270 |
| }, |
| { |
| "epoch": 1.1376673040152965, |
| "grad_norm": 2.787065420105766, |
| "learning_rate": 9.344959287340299e-05, |
| "loss": 0.3962, |
| "step": 3272 |
| }, |
| { |
| "epoch": 1.138362593429515, |
| "grad_norm": 1.7086201727689936, |
| "learning_rate": 9.332846768838566e-05, |
| "loss": 0.4884, |
| "step": 3274 |
| }, |
| { |
| "epoch": 1.1390578828437337, |
| "grad_norm": 0.7328614452843403, |
| "learning_rate": 9.320735233434673e-05, |
| "loss": 0.1495, |
| "step": 3276 |
| }, |
| { |
| "epoch": 1.1397531722579524, |
| "grad_norm": 2.5733021655734367, |
| "learning_rate": 9.308624698975833e-05, |
| "loss": 0.4051, |
| "step": 3278 |
| }, |
| { |
| "epoch": 1.140448461672171, |
| "grad_norm": 1.9453547929901163, |
| "learning_rate": 9.296515183307777e-05, |
| "loss": 0.3171, |
| "step": 3280 |
| }, |
| { |
| "epoch": 1.1411437510863898, |
| "grad_norm": 1.8895993938794953, |
| "learning_rate": 9.28440670427474e-05, |
| "loss": 0.2833, |
| "step": 3282 |
| }, |
| { |
| "epoch": 1.1418390405006085, |
| "grad_norm": 1.1756800001253909, |
| "learning_rate": 9.272299279719437e-05, |
| "loss": 0.302, |
| "step": 3284 |
| }, |
| { |
| "epoch": 1.142534329914827, |
| "grad_norm": 1.790336645978303, |
| "learning_rate": 9.26019292748301e-05, |
| "loss": 0.3007, |
| "step": 3286 |
| }, |
| { |
| "epoch": 1.1432296193290457, |
| "grad_norm": 1.8980791991650203, |
| "learning_rate": 9.248087665405037e-05, |
| "loss": 0.2531, |
| "step": 3288 |
| }, |
| { |
| "epoch": 1.1439249087432644, |
| "grad_norm": 2.1129386514358064, |
| "learning_rate": 9.235983511323483e-05, |
| "loss": 0.2145, |
| "step": 3290 |
| }, |
| { |
| "epoch": 1.144620198157483, |
| "grad_norm": 2.968517428846931, |
| "learning_rate": 9.223880483074685e-05, |
| "loss": 0.3841, |
| "step": 3292 |
| }, |
| { |
| "epoch": 1.1453154875717018, |
| "grad_norm": 1.6296138128539726, |
| "learning_rate": 9.211778598493314e-05, |
| "loss": 0.2584, |
| "step": 3294 |
| }, |
| { |
| "epoch": 1.1460107769859205, |
| "grad_norm": 1.420028014094252, |
| "learning_rate": 9.199677875412366e-05, |
| "loss": 0.4057, |
| "step": 3296 |
| }, |
| { |
| "epoch": 1.1467060664001392, |
| "grad_norm": 0.9804698036411588, |
| "learning_rate": 9.187578331663112e-05, |
| "loss": 0.1644, |
| "step": 3298 |
| }, |
| { |
| "epoch": 1.1474013558143576, |
| "grad_norm": 1.5146888605529796, |
| "learning_rate": 9.175479985075092e-05, |
| "loss": 0.2841, |
| "step": 3300 |
| }, |
| { |
| "epoch": 1.1474013558143576, |
| "eval_loss": 0.7289093136787415, |
| "eval_runtime": 370.6055, |
| "eval_samples_per_second": 13.065, |
| "eval_steps_per_second": 0.41, |
| "step": 3300 |
| }, |
| { |
| "epoch": 1.1480966452285764, |
| "grad_norm": 1.8040435748834578, |
| "learning_rate": 9.163382853476088e-05, |
| "loss": 0.2262, |
| "step": 3302 |
| }, |
| { |
| "epoch": 1.148791934642795, |
| "grad_norm": 1.351570486907936, |
| "learning_rate": 9.151286954692078e-05, |
| "loss": 0.3039, |
| "step": 3304 |
| }, |
| { |
| "epoch": 1.1494872240570138, |
| "grad_norm": 1.1075958859715693, |
| "learning_rate": 9.13919230654724e-05, |
| "loss": 0.2476, |
| "step": 3306 |
| }, |
| { |
| "epoch": 1.1501825134712325, |
| "grad_norm": 2.3992449402046825, |
| "learning_rate": 9.127098926863896e-05, |
| "loss": 0.4374, |
| "step": 3308 |
| }, |
| { |
| "epoch": 1.1508778028854512, |
| "grad_norm": 1.842112886623376, |
| "learning_rate": 9.115006833462497e-05, |
| "loss": 0.2382, |
| "step": 3310 |
| }, |
| { |
| "epoch": 1.1515730922996696, |
| "grad_norm": 2.249242528990491, |
| "learning_rate": 9.102916044161611e-05, |
| "loss": 0.3044, |
| "step": 3312 |
| }, |
| { |
| "epoch": 1.1522683817138883, |
| "grad_norm": 1.8149109357611555, |
| "learning_rate": 9.090826576777877e-05, |
| "loss": 0.2641, |
| "step": 3314 |
| }, |
| { |
| "epoch": 1.152963671128107, |
| "grad_norm": 1.8204272433593236, |
| "learning_rate": 9.078738449125987e-05, |
| "loss": 0.2744, |
| "step": 3316 |
| }, |
| { |
| "epoch": 1.1536589605423258, |
| "grad_norm": 0.9815176816757393, |
| "learning_rate": 9.066651679018663e-05, |
| "loss": 0.1847, |
| "step": 3318 |
| }, |
| { |
| "epoch": 1.1543542499565445, |
| "grad_norm": 2.2097094746958033, |
| "learning_rate": 9.054566284266613e-05, |
| "loss": 0.3692, |
| "step": 3320 |
| }, |
| { |
| "epoch": 1.1550495393707632, |
| "grad_norm": 3.519634632687192, |
| "learning_rate": 9.042482282678532e-05, |
| "loss": 0.4586, |
| "step": 3322 |
| }, |
| { |
| "epoch": 1.1557448287849819, |
| "grad_norm": 2.9050001750636154, |
| "learning_rate": 9.030399692061058e-05, |
| "loss": 0.1915, |
| "step": 3324 |
| }, |
| { |
| "epoch": 1.1564401181992003, |
| "grad_norm": 2.025948172507856, |
| "learning_rate": 9.018318530218751e-05, |
| "loss": 0.3302, |
| "step": 3326 |
| }, |
| { |
| "epoch": 1.157135407613419, |
| "grad_norm": 2.579299334801871, |
| "learning_rate": 9.00623881495406e-05, |
| "loss": 0.282, |
| "step": 3328 |
| }, |
| { |
| "epoch": 1.1578306970276377, |
| "grad_norm": 2.1353228533630357, |
| "learning_rate": 8.994160564067314e-05, |
| "loss": 0.2505, |
| "step": 3330 |
| }, |
| { |
| "epoch": 1.1585259864418564, |
| "grad_norm": 2.2230423696690336, |
| "learning_rate": 8.982083795356669e-05, |
| "loss": 0.4342, |
| "step": 3332 |
| }, |
| { |
| "epoch": 1.1592212758560752, |
| "grad_norm": 1.2474718167160803, |
| "learning_rate": 8.970008526618105e-05, |
| "loss": 0.3003, |
| "step": 3334 |
| }, |
| { |
| "epoch": 1.1599165652702939, |
| "grad_norm": 1.4231173745700876, |
| "learning_rate": 8.957934775645394e-05, |
| "loss": 0.1642, |
| "step": 3336 |
| }, |
| { |
| "epoch": 1.1606118546845123, |
| "grad_norm": 1.9842382972182684, |
| "learning_rate": 8.945862560230067e-05, |
| "loss": 0.2948, |
| "step": 3338 |
| }, |
| { |
| "epoch": 1.161307144098731, |
| "grad_norm": 1.7614461854020373, |
| "learning_rate": 8.9337918981614e-05, |
| "loss": 0.2754, |
| "step": 3340 |
| }, |
| { |
| "epoch": 1.1620024335129497, |
| "grad_norm": 2.0268627780173913, |
| "learning_rate": 8.921722807226365e-05, |
| "loss": 0.2822, |
| "step": 3342 |
| }, |
| { |
| "epoch": 1.1626977229271684, |
| "grad_norm": 2.274054400948757, |
| "learning_rate": 8.90965530520963e-05, |
| "loss": 0.42, |
| "step": 3344 |
| }, |
| { |
| "epoch": 1.1633930123413871, |
| "grad_norm": 1.4215993359309038, |
| "learning_rate": 8.897589409893522e-05, |
| "loss": 0.2952, |
| "step": 3346 |
| }, |
| { |
| "epoch": 1.1640883017556058, |
| "grad_norm": 1.159726793123209, |
| "learning_rate": 8.885525139057996e-05, |
| "loss": 0.2538, |
| "step": 3348 |
| }, |
| { |
| "epoch": 1.1647835911698246, |
| "grad_norm": 1.8788294883309744, |
| "learning_rate": 8.873462510480617e-05, |
| "loss": 0.174, |
| "step": 3350 |
| }, |
| { |
| "epoch": 1.165478880584043, |
| "grad_norm": 0.6578510669281126, |
| "learning_rate": 8.861401541936529e-05, |
| "loss": 0.2229, |
| "step": 3352 |
| }, |
| { |
| "epoch": 1.1661741699982617, |
| "grad_norm": 1.9261772964063006, |
| "learning_rate": 8.849342251198424e-05, |
| "loss": 0.3012, |
| "step": 3354 |
| }, |
| { |
| "epoch": 1.1668694594124804, |
| "grad_norm": 1.5123656620750447, |
| "learning_rate": 8.83728465603653e-05, |
| "loss": 0.3131, |
| "step": 3356 |
| }, |
| { |
| "epoch": 1.1675647488266991, |
| "grad_norm": 2.5604371118380334, |
| "learning_rate": 8.82522877421857e-05, |
| "loss": 0.445, |
| "step": 3358 |
| }, |
| { |
| "epoch": 1.1682600382409178, |
| "grad_norm": 2.200591827083585, |
| "learning_rate": 8.813174623509748e-05, |
| "loss": 0.3465, |
| "step": 3360 |
| }, |
| { |
| "epoch": 1.1689553276551365, |
| "grad_norm": 2.1372889204884293, |
| "learning_rate": 8.801122221672712e-05, |
| "loss": 0.3285, |
| "step": 3362 |
| }, |
| { |
| "epoch": 1.169650617069355, |
| "grad_norm": 1.6482530697932805, |
| "learning_rate": 8.789071586467543e-05, |
| "loss": 0.237, |
| "step": 3364 |
| }, |
| { |
| "epoch": 1.1703459064835737, |
| "grad_norm": 1.1072594534543996, |
| "learning_rate": 8.777022735651699e-05, |
| "loss": 0.2383, |
| "step": 3366 |
| }, |
| { |
| "epoch": 1.1710411958977924, |
| "grad_norm": 1.7429314997642462, |
| "learning_rate": 8.764975686980025e-05, |
| "loss": 0.3579, |
| "step": 3368 |
| }, |
| { |
| "epoch": 1.1717364853120111, |
| "grad_norm": 2.248940591223371, |
| "learning_rate": 8.752930458204708e-05, |
| "loss": 0.3844, |
| "step": 3370 |
| }, |
| { |
| "epoch": 1.1724317747262298, |
| "grad_norm": 1.4375039607093951, |
| "learning_rate": 8.74088706707525e-05, |
| "loss": 0.2899, |
| "step": 3372 |
| }, |
| { |
| "epoch": 1.1731270641404485, |
| "grad_norm": 1.5637781545834029, |
| "learning_rate": 8.72884553133845e-05, |
| "loss": 0.2627, |
| "step": 3374 |
| }, |
| { |
| "epoch": 1.1738223535546672, |
| "grad_norm": 2.4113171543676883, |
| "learning_rate": 8.716805868738361e-05, |
| "loss": 0.2984, |
| "step": 3376 |
| }, |
| { |
| "epoch": 1.1745176429688857, |
| "grad_norm": 1.3528435543162218, |
| "learning_rate": 8.70476809701629e-05, |
| "loss": 0.3814, |
| "step": 3378 |
| }, |
| { |
| "epoch": 1.1752129323831044, |
| "grad_norm": 2.1489497368541928, |
| "learning_rate": 8.692732233910752e-05, |
| "loss": 0.2854, |
| "step": 3380 |
| }, |
| { |
| "epoch": 1.1759082217973231, |
| "grad_norm": 1.2792572425102118, |
| "learning_rate": 8.680698297157445e-05, |
| "loss": 0.1875, |
| "step": 3382 |
| }, |
| { |
| "epoch": 1.1766035112115418, |
| "grad_norm": 1.8648817931055814, |
| "learning_rate": 8.668666304489235e-05, |
| "loss": 0.208, |
| "step": 3384 |
| }, |
| { |
| "epoch": 1.1772988006257605, |
| "grad_norm": 1.2959520580439923, |
| "learning_rate": 8.656636273636125e-05, |
| "loss": 0.2676, |
| "step": 3386 |
| }, |
| { |
| "epoch": 1.1779940900399792, |
| "grad_norm": 1.4603962277925204, |
| "learning_rate": 8.644608222325215e-05, |
| "loss": 0.2655, |
| "step": 3388 |
| }, |
| { |
| "epoch": 1.1786893794541977, |
| "grad_norm": 1.1409367847137166, |
| "learning_rate": 8.632582168280698e-05, |
| "loss": 0.201, |
| "step": 3390 |
| }, |
| { |
| "epoch": 1.1793846688684164, |
| "grad_norm": 1.8319987390772146, |
| "learning_rate": 8.620558129223824e-05, |
| "loss": 0.3627, |
| "step": 3392 |
| }, |
| { |
| "epoch": 1.1800799582826351, |
| "grad_norm": 2.4420444984028533, |
| "learning_rate": 8.608536122872872e-05, |
| "loss": 0.2982, |
| "step": 3394 |
| }, |
| { |
| "epoch": 1.1807752476968538, |
| "grad_norm": 2.189521227134317, |
| "learning_rate": 8.596516166943125e-05, |
| "loss": 0.3137, |
| "step": 3396 |
| }, |
| { |
| "epoch": 1.1814705371110725, |
| "grad_norm": 1.2223819882056972, |
| "learning_rate": 8.584498279146847e-05, |
| "loss": 0.2512, |
| "step": 3398 |
| }, |
| { |
| "epoch": 1.1821658265252912, |
| "grad_norm": 2.4085762489031497, |
| "learning_rate": 8.572482477193247e-05, |
| "loss": 0.4008, |
| "step": 3400 |
| }, |
| { |
| "epoch": 1.18286111593951, |
| "grad_norm": 1.9788646282331432, |
| "learning_rate": 8.560468778788466e-05, |
| "loss": 0.382, |
| "step": 3402 |
| }, |
| { |
| "epoch": 1.1835564053537284, |
| "grad_norm": 1.3997615841673436, |
| "learning_rate": 8.548457201635552e-05, |
| "loss": 0.3067, |
| "step": 3404 |
| }, |
| { |
| "epoch": 1.1842516947679471, |
| "grad_norm": 1.332461336350933, |
| "learning_rate": 8.53644776343441e-05, |
| "loss": 0.2878, |
| "step": 3406 |
| }, |
| { |
| "epoch": 1.1849469841821658, |
| "grad_norm": 2.6380768106661945, |
| "learning_rate": 8.52444048188181e-05, |
| "loss": 0.3119, |
| "step": 3408 |
| }, |
| { |
| "epoch": 1.1856422735963845, |
| "grad_norm": 1.6210206136203409, |
| "learning_rate": 8.512435374671339e-05, |
| "loss": 0.4068, |
| "step": 3410 |
| }, |
| { |
| "epoch": 1.1863375630106032, |
| "grad_norm": 2.7299657529851955, |
| "learning_rate": 8.500432459493371e-05, |
| "loss": 0.2811, |
| "step": 3412 |
| }, |
| { |
| "epoch": 1.187032852424822, |
| "grad_norm": 1.769808826094503, |
| "learning_rate": 8.488431754035062e-05, |
| "loss": 0.2383, |
| "step": 3414 |
| }, |
| { |
| "epoch": 1.1877281418390404, |
| "grad_norm": 1.962430169310977, |
| "learning_rate": 8.476433275980305e-05, |
| "loss": 0.3461, |
| "step": 3416 |
| }, |
| { |
| "epoch": 1.188423431253259, |
| "grad_norm": 1.5705832259920427, |
| "learning_rate": 8.464437043009713e-05, |
| "loss": 0.2304, |
| "step": 3418 |
| }, |
| { |
| "epoch": 1.1891187206674778, |
| "grad_norm": 1.1565201875267868, |
| "learning_rate": 8.452443072800595e-05, |
| "loss": 0.2722, |
| "step": 3420 |
| }, |
| { |
| "epoch": 1.1898140100816965, |
| "grad_norm": 2.102978307271352, |
| "learning_rate": 8.440451383026913e-05, |
| "loss": 0.3181, |
| "step": 3422 |
| }, |
| { |
| "epoch": 1.1905092994959152, |
| "grad_norm": 1.513497747947106, |
| "learning_rate": 8.42846199135928e-05, |
| "loss": 0.362, |
| "step": 3424 |
| }, |
| { |
| "epoch": 1.191204588910134, |
| "grad_norm": 1.6867329113948406, |
| "learning_rate": 8.41647491546492e-05, |
| "loss": 0.1619, |
| "step": 3426 |
| }, |
| { |
| "epoch": 1.1918998783243526, |
| "grad_norm": 1.9695591001495825, |
| "learning_rate": 8.404490173007644e-05, |
| "loss": 0.3709, |
| "step": 3428 |
| }, |
| { |
| "epoch": 1.192595167738571, |
| "grad_norm": 0.7491722502459326, |
| "learning_rate": 8.392507781647823e-05, |
| "loss": 0.1392, |
| "step": 3430 |
| }, |
| { |
| "epoch": 1.1932904571527898, |
| "grad_norm": 1.774230078149517, |
| "learning_rate": 8.380527759042367e-05, |
| "loss": 0.2948, |
| "step": 3432 |
| }, |
| { |
| "epoch": 1.1939857465670085, |
| "grad_norm": 1.185904924173457, |
| "learning_rate": 8.368550122844691e-05, |
| "loss": 0.2258, |
| "step": 3434 |
| }, |
| { |
| "epoch": 1.1946810359812272, |
| "grad_norm": 2.3608753583375734, |
| "learning_rate": 8.356574890704696e-05, |
| "loss": 0.363, |
| "step": 3436 |
| }, |
| { |
| "epoch": 1.195376325395446, |
| "grad_norm": 1.8282541365525038, |
| "learning_rate": 8.344602080268739e-05, |
| "loss": 0.2482, |
| "step": 3438 |
| }, |
| { |
| "epoch": 1.1960716148096646, |
| "grad_norm": 2.4654473507150634, |
| "learning_rate": 8.33263170917961e-05, |
| "loss": 0.216, |
| "step": 3440 |
| }, |
| { |
| "epoch": 1.196766904223883, |
| "grad_norm": 2.4336610558248006, |
| "learning_rate": 8.320663795076508e-05, |
| "loss": 0.3436, |
| "step": 3442 |
| }, |
| { |
| "epoch": 1.1974621936381018, |
| "grad_norm": 1.864535910059934, |
| "learning_rate": 8.308698355595006e-05, |
| "loss": 0.3608, |
| "step": 3444 |
| }, |
| { |
| "epoch": 1.1981574830523205, |
| "grad_norm": 2.582709272409011, |
| "learning_rate": 8.296735408367028e-05, |
| "loss": 0.2636, |
| "step": 3446 |
| }, |
| { |
| "epoch": 1.1988527724665392, |
| "grad_norm": 2.5746350804379188, |
| "learning_rate": 8.284774971020832e-05, |
| "loss": 0.3171, |
| "step": 3448 |
| }, |
| { |
| "epoch": 1.199548061880758, |
| "grad_norm": 2.999111600960781, |
| "learning_rate": 8.272817061180973e-05, |
| "loss": 0.519, |
| "step": 3450 |
| }, |
| { |
| "epoch": 1.199548061880758, |
| "eval_loss": 0.8144862055778503, |
| "eval_runtime": 370.0535, |
| "eval_samples_per_second": 13.085, |
| "eval_steps_per_second": 0.411, |
| "step": 3450 |
| }, |
| { |
| "epoch": 1.2002433512949766, |
| "grad_norm": 1.8853883581896378, |
| "learning_rate": 8.260861696468288e-05, |
| "loss": 0.3958, |
| "step": 3452 |
| }, |
| { |
| "epoch": 1.2009386407091953, |
| "grad_norm": 3.4039218970105947, |
| "learning_rate": 8.248908894499859e-05, |
| "loss": 0.2991, |
| "step": 3454 |
| }, |
| { |
| "epoch": 1.2016339301234138, |
| "grad_norm": 1.321128890341765, |
| "learning_rate": 8.236958672888984e-05, |
| "loss": 0.2104, |
| "step": 3456 |
| }, |
| { |
| "epoch": 1.2023292195376325, |
| "grad_norm": 5.221525421158704, |
| "learning_rate": 8.225011049245174e-05, |
| "loss": 0.2686, |
| "step": 3458 |
| }, |
| { |
| "epoch": 1.2030245089518512, |
| "grad_norm": 1.5789179128197859, |
| "learning_rate": 8.2130660411741e-05, |
| "loss": 0.2212, |
| "step": 3460 |
| }, |
| { |
| "epoch": 1.20371979836607, |
| "grad_norm": 2.6043265958832196, |
| "learning_rate": 8.201123666277584e-05, |
| "loss": 0.3122, |
| "step": 3462 |
| }, |
| { |
| "epoch": 1.2044150877802886, |
| "grad_norm": 1.8647996647564873, |
| "learning_rate": 8.189183942153566e-05, |
| "loss": 0.2424, |
| "step": 3464 |
| }, |
| { |
| "epoch": 1.2051103771945073, |
| "grad_norm": 0.8884814097155843, |
| "learning_rate": 8.177246886396088e-05, |
| "loss": 0.1745, |
| "step": 3466 |
| }, |
| { |
| "epoch": 1.2058056666087258, |
| "grad_norm": 1.8627913890382377, |
| "learning_rate": 8.16531251659524e-05, |
| "loss": 0.2878, |
| "step": 3468 |
| }, |
| { |
| "epoch": 1.2065009560229445, |
| "grad_norm": 1.55297593480243, |
| "learning_rate": 8.153380850337174e-05, |
| "loss": 0.2906, |
| "step": 3470 |
| }, |
| { |
| "epoch": 1.2071962454371632, |
| "grad_norm": 1.3252322573417612, |
| "learning_rate": 8.141451905204049e-05, |
| "loss": 0.2295, |
| "step": 3472 |
| }, |
| { |
| "epoch": 1.207891534851382, |
| "grad_norm": 2.4004641916220804, |
| "learning_rate": 8.129525698774014e-05, |
| "loss": 0.3076, |
| "step": 3474 |
| }, |
| { |
| "epoch": 1.2085868242656006, |
| "grad_norm": 1.7033321689679786, |
| "learning_rate": 8.11760224862119e-05, |
| "loss": 0.1926, |
| "step": 3476 |
| }, |
| { |
| "epoch": 1.2092821136798193, |
| "grad_norm": 0.829590608355227, |
| "learning_rate": 8.105681572315629e-05, |
| "loss": 0.3428, |
| "step": 3478 |
| }, |
| { |
| "epoch": 1.209977403094038, |
| "grad_norm": 1.9496915772423573, |
| "learning_rate": 8.09376368742329e-05, |
| "loss": 0.2704, |
| "step": 3480 |
| }, |
| { |
| "epoch": 1.2106726925082565, |
| "grad_norm": 2.231280657442789, |
| "learning_rate": 8.081848611506035e-05, |
| "loss": 0.3256, |
| "step": 3482 |
| }, |
| { |
| "epoch": 1.2113679819224752, |
| "grad_norm": 1.9359732137866787, |
| "learning_rate": 8.069936362121573e-05, |
| "loss": 0.1879, |
| "step": 3484 |
| }, |
| { |
| "epoch": 1.212063271336694, |
| "grad_norm": 2.24974543336551, |
| "learning_rate": 8.05802695682345e-05, |
| "loss": 0.442, |
| "step": 3486 |
| }, |
| { |
| "epoch": 1.2127585607509126, |
| "grad_norm": 3.464604965717862, |
| "learning_rate": 8.046120413161035e-05, |
| "loss": 0.3197, |
| "step": 3488 |
| }, |
| { |
| "epoch": 1.2134538501651313, |
| "grad_norm": 1.6955404019862794, |
| "learning_rate": 8.034216748679455e-05, |
| "loss": 0.2491, |
| "step": 3490 |
| }, |
| { |
| "epoch": 1.21414913957935, |
| "grad_norm": 1.3734601258052566, |
| "learning_rate": 8.022315980919613e-05, |
| "loss": 0.2286, |
| "step": 3492 |
| }, |
| { |
| "epoch": 1.2148444289935685, |
| "grad_norm": 1.781453419827392, |
| "learning_rate": 8.01041812741814e-05, |
| "loss": 0.3284, |
| "step": 3494 |
| }, |
| { |
| "epoch": 1.2155397184077872, |
| "grad_norm": 1.268141200168223, |
| "learning_rate": 7.998523205707373e-05, |
| "loss": 0.1641, |
| "step": 3496 |
| }, |
| { |
| "epoch": 1.2162350078220059, |
| "grad_norm": 1.4444739618436466, |
| "learning_rate": 7.986631233315322e-05, |
| "loss": 0.1734, |
| "step": 3498 |
| }, |
| { |
| "epoch": 1.2169302972362246, |
| "grad_norm": 1.4062853458153892, |
| "learning_rate": 7.974742227765663e-05, |
| "loss": 0.1855, |
| "step": 3500 |
| }, |
| { |
| "epoch": 1.2176255866504433, |
| "grad_norm": 3.6497367613139247, |
| "learning_rate": 7.962856206577685e-05, |
| "loss": 0.5155, |
| "step": 3502 |
| }, |
| { |
| "epoch": 1.218320876064662, |
| "grad_norm": 2.3641021234312105, |
| "learning_rate": 7.950973187266291e-05, |
| "loss": 0.2977, |
| "step": 3504 |
| }, |
| { |
| "epoch": 1.2190161654788807, |
| "grad_norm": 0.8064333318001713, |
| "learning_rate": 7.93909318734196e-05, |
| "loss": 0.2148, |
| "step": 3506 |
| }, |
| { |
| "epoch": 1.2197114548930992, |
| "grad_norm": 2.7680896439429756, |
| "learning_rate": 7.927216224310713e-05, |
| "loss": 0.248, |
| "step": 3508 |
| }, |
| { |
| "epoch": 1.2204067443073179, |
| "grad_norm": 2.428621641475993, |
| "learning_rate": 7.915342315674103e-05, |
| "loss": 0.4271, |
| "step": 3510 |
| }, |
| { |
| "epoch": 1.2211020337215366, |
| "grad_norm": 1.719410561000375, |
| "learning_rate": 7.903471478929189e-05, |
| "loss": 0.3253, |
| "step": 3512 |
| }, |
| { |
| "epoch": 1.2217973231357553, |
| "grad_norm": 2.324008122067757, |
| "learning_rate": 7.891603731568482e-05, |
| "loss": 0.3414, |
| "step": 3514 |
| }, |
| { |
| "epoch": 1.222492612549974, |
| "grad_norm": 1.4608567083745039, |
| "learning_rate": 7.87973909107996e-05, |
| "loss": 0.295, |
| "step": 3516 |
| }, |
| { |
| "epoch": 1.2231879019641927, |
| "grad_norm": 1.6816758559841376, |
| "learning_rate": 7.867877574947018e-05, |
| "loss": 0.3005, |
| "step": 3518 |
| }, |
| { |
| "epoch": 1.2238831913784112, |
| "grad_norm": 1.04753563754324, |
| "learning_rate": 7.856019200648444e-05, |
| "loss": 0.1868, |
| "step": 3520 |
| }, |
| { |
| "epoch": 1.2245784807926299, |
| "grad_norm": 1.3160815552474603, |
| "learning_rate": 7.8441639856584e-05, |
| "loss": 0.361, |
| "step": 3522 |
| }, |
| { |
| "epoch": 1.2252737702068486, |
| "grad_norm": 1.2845412978218502, |
| "learning_rate": 7.832311947446384e-05, |
| "loss": 0.3377, |
| "step": 3524 |
| }, |
| { |
| "epoch": 1.2259690596210673, |
| "grad_norm": 1.495804403784908, |
| "learning_rate": 7.820463103477222e-05, |
| "loss": 0.4104, |
| "step": 3526 |
| }, |
| { |
| "epoch": 1.226664349035286, |
| "grad_norm": 2.0014377255648466, |
| "learning_rate": 7.808617471211032e-05, |
| "loss": 0.2124, |
| "step": 3528 |
| }, |
| { |
| "epoch": 1.2273596384495047, |
| "grad_norm": 1.4867899157314466, |
| "learning_rate": 7.7967750681032e-05, |
| "loss": 0.2661, |
| "step": 3530 |
| }, |
| { |
| "epoch": 1.2280549278637234, |
| "grad_norm": 2.4573652602685283, |
| "learning_rate": 7.784935911604345e-05, |
| "loss": 0.2678, |
| "step": 3532 |
| }, |
| { |
| "epoch": 1.2287502172779419, |
| "grad_norm": 1.2240289643877307, |
| "learning_rate": 7.773100019160313e-05, |
| "loss": 0.2595, |
| "step": 3534 |
| }, |
| { |
| "epoch": 1.2294455066921606, |
| "grad_norm": 0.849656976510099, |
| "learning_rate": 7.761267408212131e-05, |
| "loss": 0.1717, |
| "step": 3536 |
| }, |
| { |
| "epoch": 1.2301407961063793, |
| "grad_norm": 2.696718789588286, |
| "learning_rate": 7.749438096195995e-05, |
| "loss": 0.2405, |
| "step": 3538 |
| }, |
| { |
| "epoch": 1.230836085520598, |
| "grad_norm": 1.5470142516634993, |
| "learning_rate": 7.737612100543243e-05, |
| "loss": 0.3578, |
| "step": 3540 |
| }, |
| { |
| "epoch": 1.2315313749348167, |
| "grad_norm": 2.26065493202481, |
| "learning_rate": 7.725789438680318e-05, |
| "loss": 0.3515, |
| "step": 3542 |
| }, |
| { |
| "epoch": 1.2322266643490352, |
| "grad_norm": 1.9087525582302716, |
| "learning_rate": 7.713970128028759e-05, |
| "loss": 0.2461, |
| "step": 3544 |
| }, |
| { |
| "epoch": 1.2329219537632539, |
| "grad_norm": 2.955156031826718, |
| "learning_rate": 7.702154186005165e-05, |
| "loss": 0.3291, |
| "step": 3546 |
| }, |
| { |
| "epoch": 1.2336172431774726, |
| "grad_norm": 1.384553634504943, |
| "learning_rate": 7.690341630021161e-05, |
| "loss": 0.3115, |
| "step": 3548 |
| }, |
| { |
| "epoch": 1.2343125325916913, |
| "grad_norm": 1.6352563390713386, |
| "learning_rate": 7.678532477483393e-05, |
| "loss": 0.2185, |
| "step": 3550 |
| }, |
| { |
| "epoch": 1.23500782200591, |
| "grad_norm": 3.1027723668259437, |
| "learning_rate": 7.666726745793491e-05, |
| "loss": 0.5086, |
| "step": 3552 |
| }, |
| { |
| "epoch": 1.2357031114201287, |
| "grad_norm": 1.9977976367136538, |
| "learning_rate": 7.654924452348042e-05, |
| "loss": 0.2404, |
| "step": 3554 |
| }, |
| { |
| "epoch": 1.2363984008343474, |
| "grad_norm": 1.8019264636906462, |
| "learning_rate": 7.643125614538568e-05, |
| "loss": 0.229, |
| "step": 3556 |
| }, |
| { |
| "epoch": 1.237093690248566, |
| "grad_norm": 1.3606242454972741, |
| "learning_rate": 7.631330249751492e-05, |
| "loss": 0.234, |
| "step": 3558 |
| }, |
| { |
| "epoch": 1.2377889796627846, |
| "grad_norm": 2.205200701168964, |
| "learning_rate": 7.619538375368131e-05, |
| "loss": 0.1368, |
| "step": 3560 |
| }, |
| { |
| "epoch": 1.2384842690770033, |
| "grad_norm": 1.407813757325524, |
| "learning_rate": 7.607750008764647e-05, |
| "loss": 0.24, |
| "step": 3562 |
| }, |
| { |
| "epoch": 1.239179558491222, |
| "grad_norm": 1.7728463452972267, |
| "learning_rate": 7.595965167312041e-05, |
| "loss": 0.3096, |
| "step": 3564 |
| }, |
| { |
| "epoch": 1.2398748479054407, |
| "grad_norm": 1.0793002128199194, |
| "learning_rate": 7.584183868376117e-05, |
| "loss": 0.2486, |
| "step": 3566 |
| }, |
| { |
| "epoch": 1.2405701373196594, |
| "grad_norm": 1.966129412113088, |
| "learning_rate": 7.572406129317465e-05, |
| "loss": 0.282, |
| "step": 3568 |
| }, |
| { |
| "epoch": 1.2412654267338779, |
| "grad_norm": 2.0913122213839763, |
| "learning_rate": 7.56063196749141e-05, |
| "loss": 0.2789, |
| "step": 3570 |
| }, |
| { |
| "epoch": 1.2419607161480966, |
| "grad_norm": 1.5277956402626331, |
| "learning_rate": 7.548861400248027e-05, |
| "loss": 0.2982, |
| "step": 3572 |
| }, |
| { |
| "epoch": 1.2426560055623153, |
| "grad_norm": 1.3581976346862918, |
| "learning_rate": 7.537094444932082e-05, |
| "loss": 0.295, |
| "step": 3574 |
| }, |
| { |
| "epoch": 1.243351294976534, |
| "grad_norm": 1.8500974001237571, |
| "learning_rate": 7.525331118883026e-05, |
| "loss": 0.2187, |
| "step": 3576 |
| }, |
| { |
| "epoch": 1.2440465843907527, |
| "grad_norm": 1.861509679974732, |
| "learning_rate": 7.513571439434959e-05, |
| "loss": 0.2981, |
| "step": 3578 |
| }, |
| { |
| "epoch": 1.2447418738049714, |
| "grad_norm": 1.264025275050796, |
| "learning_rate": 7.501815423916605e-05, |
| "loss": 0.2714, |
| "step": 3580 |
| }, |
| { |
| "epoch": 1.24543716321919, |
| "grad_norm": 1.6807631386232256, |
| "learning_rate": 7.490063089651289e-05, |
| "loss": 0.3103, |
| "step": 3582 |
| }, |
| { |
| "epoch": 1.2461324526334088, |
| "grad_norm": 1.7621595082463557, |
| "learning_rate": 7.478314453956916e-05, |
| "loss": 0.2971, |
| "step": 3584 |
| }, |
| { |
| "epoch": 1.2468277420476273, |
| "grad_norm": 1.2860268659971938, |
| "learning_rate": 7.466569534145936e-05, |
| "loss": 0.1971, |
| "step": 3586 |
| }, |
| { |
| "epoch": 1.247523031461846, |
| "grad_norm": 2.622105467810232, |
| "learning_rate": 7.454828347525325e-05, |
| "loss": 0.3226, |
| "step": 3588 |
| }, |
| { |
| "epoch": 1.2482183208760647, |
| "grad_norm": 0.7993413824807574, |
| "learning_rate": 7.443090911396566e-05, |
| "loss": 0.2273, |
| "step": 3590 |
| }, |
| { |
| "epoch": 1.2489136102902834, |
| "grad_norm": 2.4653534356329816, |
| "learning_rate": 7.431357243055596e-05, |
| "loss": 0.3557, |
| "step": 3592 |
| }, |
| { |
| "epoch": 1.249608899704502, |
| "grad_norm": 2.1884871260152288, |
| "learning_rate": 7.419627359792817e-05, |
| "loss": 0.2919, |
| "step": 3594 |
| }, |
| { |
| "epoch": 1.2503041891187205, |
| "grad_norm": 1.9291328307496636, |
| "learning_rate": 7.40790127889305e-05, |
| "loss": 0.2519, |
| "step": 3596 |
| }, |
| { |
| "epoch": 1.2509994785329392, |
| "grad_norm": 2.7783950086565334, |
| "learning_rate": 7.39617901763551e-05, |
| "loss": 0.355, |
| "step": 3598 |
| }, |
| { |
| "epoch": 1.251694767947158, |
| "grad_norm": 1.3097549565134516, |
| "learning_rate": 7.384460593293782e-05, |
| "loss": 0.2001, |
| "step": 3600 |
| }, |
| { |
| "epoch": 1.251694767947158, |
| "eval_loss": 0.8054365515708923, |
| "eval_runtime": 370.2882, |
| "eval_samples_per_second": 13.076, |
| "eval_steps_per_second": 0.41, |
| "step": 3600 |
| }, |
| { |
| "epoch": 1.2523900573613767, |
| "grad_norm": 0.9923873707782574, |
| "learning_rate": 7.372746023135806e-05, |
| "loss": 0.2729, |
| "step": 3602 |
| }, |
| { |
| "epoch": 1.2530853467755954, |
| "grad_norm": 1.7475999673497644, |
| "learning_rate": 7.361035324423828e-05, |
| "loss": 0.2551, |
| "step": 3604 |
| }, |
| { |
| "epoch": 1.253780636189814, |
| "grad_norm": 2.548939456365811, |
| "learning_rate": 7.349328514414402e-05, |
| "loss": 0.3306, |
| "step": 3606 |
| }, |
| { |
| "epoch": 1.2544759256040328, |
| "grad_norm": 2.504108647851307, |
| "learning_rate": 7.337625610358346e-05, |
| "loss": 0.3652, |
| "step": 3608 |
| }, |
| { |
| "epoch": 1.2551712150182515, |
| "grad_norm": 1.5468257575274822, |
| "learning_rate": 7.325926629500722e-05, |
| "loss": 0.1624, |
| "step": 3610 |
| }, |
| { |
| "epoch": 1.25586650443247, |
| "grad_norm": 0.8958093611478746, |
| "learning_rate": 7.314231589080814e-05, |
| "loss": 0.0874, |
| "step": 3612 |
| }, |
| { |
| "epoch": 1.2565617938466886, |
| "grad_norm": 1.8259242408618002, |
| "learning_rate": 7.302540506332099e-05, |
| "loss": 0.3524, |
| "step": 3614 |
| }, |
| { |
| "epoch": 1.2572570832609073, |
| "grad_norm": 0.9286267489098781, |
| "learning_rate": 7.290853398482217e-05, |
| "loss": 0.1511, |
| "step": 3616 |
| }, |
| { |
| "epoch": 1.257952372675126, |
| "grad_norm": 1.6624360314411797, |
| "learning_rate": 7.279170282752955e-05, |
| "loss": 0.1911, |
| "step": 3618 |
| }, |
| { |
| "epoch": 1.2586476620893448, |
| "grad_norm": 0.9644429510957823, |
| "learning_rate": 7.267491176360219e-05, |
| "loss": 0.2792, |
| "step": 3620 |
| }, |
| { |
| "epoch": 1.2593429515035632, |
| "grad_norm": 3.603534829066384, |
| "learning_rate": 7.255816096514003e-05, |
| "loss": 0.3576, |
| "step": 3622 |
| }, |
| { |
| "epoch": 1.260038240917782, |
| "grad_norm": 1.5718715205715308, |
| "learning_rate": 7.244145060418373e-05, |
| "loss": 0.1743, |
| "step": 3624 |
| }, |
| { |
| "epoch": 1.2607335303320006, |
| "grad_norm": 2.4481162527836964, |
| "learning_rate": 7.232478085271428e-05, |
| "loss": 0.3542, |
| "step": 3626 |
| }, |
| { |
| "epoch": 1.2614288197462193, |
| "grad_norm": 2.080316425645612, |
| "learning_rate": 7.220815188265287e-05, |
| "loss": 0.3761, |
| "step": 3628 |
| }, |
| { |
| "epoch": 1.262124109160438, |
| "grad_norm": 1.6714144409696445, |
| "learning_rate": 7.209156386586064e-05, |
| "loss": 0.3901, |
| "step": 3630 |
| }, |
| { |
| "epoch": 1.2628193985746567, |
| "grad_norm": 2.3645735603968965, |
| "learning_rate": 7.197501697413834e-05, |
| "loss": 0.3567, |
| "step": 3632 |
| }, |
| { |
| "epoch": 1.2635146879888755, |
| "grad_norm": 1.0285755433235917, |
| "learning_rate": 7.185851137922608e-05, |
| "loss": 0.2822, |
| "step": 3634 |
| }, |
| { |
| "epoch": 1.2642099774030942, |
| "grad_norm": 1.019604725556347, |
| "learning_rate": 7.174204725280321e-05, |
| "loss": 0.1828, |
| "step": 3636 |
| }, |
| { |
| "epoch": 1.2649052668173126, |
| "grad_norm": 2.7058243966623907, |
| "learning_rate": 7.16256247664879e-05, |
| "loss": 0.3848, |
| "step": 3638 |
| }, |
| { |
| "epoch": 1.2656005562315313, |
| "grad_norm": 2.2242001170571495, |
| "learning_rate": 7.150924409183696e-05, |
| "loss": 0.4591, |
| "step": 3640 |
| }, |
| { |
| "epoch": 1.26629584564575, |
| "grad_norm": 2.155072688376517, |
| "learning_rate": 7.139290540034562e-05, |
| "loss": 0.2461, |
| "step": 3642 |
| }, |
| { |
| "epoch": 1.2669911350599687, |
| "grad_norm": 1.6476973655855016, |
| "learning_rate": 7.127660886344725e-05, |
| "loss": 0.1814, |
| "step": 3644 |
| }, |
| { |
| "epoch": 1.2676864244741874, |
| "grad_norm": 1.8352629064029073, |
| "learning_rate": 7.116035465251308e-05, |
| "loss": 0.3113, |
| "step": 3646 |
| }, |
| { |
| "epoch": 1.268381713888406, |
| "grad_norm": 1.483255711370553, |
| "learning_rate": 7.104414293885202e-05, |
| "loss": 0.2812, |
| "step": 3648 |
| }, |
| { |
| "epoch": 1.2690770033026246, |
| "grad_norm": 1.6962999261118368, |
| "learning_rate": 7.092797389371022e-05, |
| "loss": 0.1976, |
| "step": 3650 |
| }, |
| { |
| "epoch": 1.2697722927168433, |
| "grad_norm": 2.37929796474955, |
| "learning_rate": 7.081184768827111e-05, |
| "loss": 0.3611, |
| "step": 3652 |
| }, |
| { |
| "epoch": 1.270467582131062, |
| "grad_norm": 1.3062950416042882, |
| "learning_rate": 7.06957644936549e-05, |
| "loss": 0.3021, |
| "step": 3654 |
| }, |
| { |
| "epoch": 1.2711628715452807, |
| "grad_norm": 1.5706726442496965, |
| "learning_rate": 7.057972448091851e-05, |
| "loss": 0.2819, |
| "step": 3656 |
| }, |
| { |
| "epoch": 1.2718581609594994, |
| "grad_norm": 1.3335001105752498, |
| "learning_rate": 7.046372782105513e-05, |
| "loss": 0.3339, |
| "step": 3658 |
| }, |
| { |
| "epoch": 1.2725534503737181, |
| "grad_norm": 1.9239148861308164, |
| "learning_rate": 7.034777468499407e-05, |
| "loss": 0.2615, |
| "step": 3660 |
| }, |
| { |
| "epoch": 1.2732487397879368, |
| "grad_norm": 2.11200071751669, |
| "learning_rate": 7.023186524360062e-05, |
| "loss": 0.3613, |
| "step": 3662 |
| }, |
| { |
| "epoch": 1.2739440292021553, |
| "grad_norm": 1.3158474735048813, |
| "learning_rate": 7.011599966767552e-05, |
| "loss": 0.2882, |
| "step": 3664 |
| }, |
| { |
| "epoch": 1.274639318616374, |
| "grad_norm": 1.5842163962490339, |
| "learning_rate": 7.000017812795501e-05, |
| "loss": 0.3858, |
| "step": 3666 |
| }, |
| { |
| "epoch": 1.2753346080305927, |
| "grad_norm": 0.24870081173245925, |
| "learning_rate": 6.988440079511035e-05, |
| "loss": 0.2021, |
| "step": 3668 |
| }, |
| { |
| "epoch": 1.2760298974448114, |
| "grad_norm": 2.222630502074732, |
| "learning_rate": 6.976866783974778e-05, |
| "loss": 0.306, |
| "step": 3670 |
| }, |
| { |
| "epoch": 1.2767251868590301, |
| "grad_norm": 1.393284815680166, |
| "learning_rate": 6.965297943240794e-05, |
| "loss": 0.3569, |
| "step": 3672 |
| }, |
| { |
| "epoch": 1.2774204762732486, |
| "grad_norm": 1.275389156661581, |
| "learning_rate": 6.9537335743566e-05, |
| "loss": 0.2977, |
| "step": 3674 |
| }, |
| { |
| "epoch": 1.2781157656874673, |
| "grad_norm": 1.4112263984411733, |
| "learning_rate": 6.942173694363117e-05, |
| "loss": 0.2999, |
| "step": 3676 |
| }, |
| { |
| "epoch": 1.278811055101686, |
| "grad_norm": 1.6998237870179789, |
| "learning_rate": 6.930618320294652e-05, |
| "loss": 0.2992, |
| "step": 3678 |
| }, |
| { |
| "epoch": 1.2795063445159047, |
| "grad_norm": 1.2158429299805096, |
| "learning_rate": 6.919067469178876e-05, |
| "loss": 0.3488, |
| "step": 3680 |
| }, |
| { |
| "epoch": 1.2802016339301234, |
| "grad_norm": 1.370602961809805, |
| "learning_rate": 6.907521158036788e-05, |
| "loss": 0.4363, |
| "step": 3682 |
| }, |
| { |
| "epoch": 1.2808969233443421, |
| "grad_norm": 2.029481838134644, |
| "learning_rate": 6.8959794038827e-05, |
| "loss": 0.2906, |
| "step": 3684 |
| }, |
| { |
| "epoch": 1.2815922127585608, |
| "grad_norm": 1.8125218287231513, |
| "learning_rate": 6.88444222372421e-05, |
| "loss": 0.3106, |
| "step": 3686 |
| }, |
| { |
| "epoch": 1.2822875021727795, |
| "grad_norm": 1.9137800798111786, |
| "learning_rate": 6.872909634562176e-05, |
| "loss": 0.3052, |
| "step": 3688 |
| }, |
| { |
| "epoch": 1.282982791586998, |
| "grad_norm": 1.327035469536592, |
| "learning_rate": 6.861381653390688e-05, |
| "loss": 0.2589, |
| "step": 3690 |
| }, |
| { |
| "epoch": 1.2836780810012167, |
| "grad_norm": 1.978336934685258, |
| "learning_rate": 6.849858297197055e-05, |
| "loss": 0.266, |
| "step": 3692 |
| }, |
| { |
| "epoch": 1.2843733704154354, |
| "grad_norm": 0.9888583341888729, |
| "learning_rate": 6.838339582961754e-05, |
| "loss": 0.2079, |
| "step": 3694 |
| }, |
| { |
| "epoch": 1.2850686598296541, |
| "grad_norm": 2.1848537592557364, |
| "learning_rate": 6.826825527658435e-05, |
| "loss": 0.3686, |
| "step": 3696 |
| }, |
| { |
| "epoch": 1.2857639492438728, |
| "grad_norm": 2.2486970053744844, |
| "learning_rate": 6.81531614825388e-05, |
| "loss": 0.3209, |
| "step": 3698 |
| }, |
| { |
| "epoch": 1.2864592386580913, |
| "grad_norm": 2.043017242026317, |
| "learning_rate": 6.803811461707983e-05, |
| "loss": 0.2928, |
| "step": 3700 |
| }, |
| { |
| "epoch": 1.28715452807231, |
| "grad_norm": 1.8787348233874217, |
| "learning_rate": 6.792311484973715e-05, |
| "loss": 0.2521, |
| "step": 3702 |
| }, |
| { |
| "epoch": 1.2878498174865287, |
| "grad_norm": 1.4583799397954622, |
| "learning_rate": 6.780816234997117e-05, |
| "loss": 0.1117, |
| "step": 3704 |
| }, |
| { |
| "epoch": 1.2885451069007474, |
| "grad_norm": 1.216933373220705, |
| "learning_rate": 6.76932572871725e-05, |
| "loss": 0.2918, |
| "step": 3706 |
| }, |
| { |
| "epoch": 1.2892403963149661, |
| "grad_norm": 1.5062494871280234, |
| "learning_rate": 6.757839983066201e-05, |
| "loss": 0.1298, |
| "step": 3708 |
| }, |
| { |
| "epoch": 1.2899356857291848, |
| "grad_norm": 2.304816011902479, |
| "learning_rate": 6.746359014969033e-05, |
| "loss": 0.3133, |
| "step": 3710 |
| }, |
| { |
| "epoch": 1.2906309751434035, |
| "grad_norm": 1.2560733923473588, |
| "learning_rate": 6.734882841343771e-05, |
| "loss": 0.3067, |
| "step": 3712 |
| }, |
| { |
| "epoch": 1.2913262645576222, |
| "grad_norm": 1.4390911313124963, |
| "learning_rate": 6.723411479101376e-05, |
| "loss": 0.1879, |
| "step": 3714 |
| }, |
| { |
| "epoch": 1.2920215539718407, |
| "grad_norm": 2.310421126758864, |
| "learning_rate": 6.711944945145718e-05, |
| "loss": 0.332, |
| "step": 3716 |
| }, |
| { |
| "epoch": 1.2927168433860594, |
| "grad_norm": 1.624938499458673, |
| "learning_rate": 6.700483256373548e-05, |
| "loss": 0.45, |
| "step": 3718 |
| }, |
| { |
| "epoch": 1.2934121328002781, |
| "grad_norm": 2.148849146789743, |
| "learning_rate": 6.689026429674485e-05, |
| "loss": 0.3801, |
| "step": 3720 |
| }, |
| { |
| "epoch": 1.2941074222144968, |
| "grad_norm": 1.5460254554247057, |
| "learning_rate": 6.67757448193098e-05, |
| "loss": 0.3579, |
| "step": 3722 |
| }, |
| { |
| "epoch": 1.2948027116287155, |
| "grad_norm": 1.2002079993377932, |
| "learning_rate": 6.666127430018291e-05, |
| "loss": 0.2289, |
| "step": 3724 |
| }, |
| { |
| "epoch": 1.295498001042934, |
| "grad_norm": 1.1950493843872743, |
| "learning_rate": 6.654685290804474e-05, |
| "loss": 0.3682, |
| "step": 3726 |
| }, |
| { |
| "epoch": 1.2961932904571527, |
| "grad_norm": 1.8279025589413753, |
| "learning_rate": 6.643248081150325e-05, |
| "loss": 0.2497, |
| "step": 3728 |
| }, |
| { |
| "epoch": 1.2968885798713714, |
| "grad_norm": 1.2331515103180768, |
| "learning_rate": 6.631815817909391e-05, |
| "loss": 0.2277, |
| "step": 3730 |
| }, |
| { |
| "epoch": 1.29758386928559, |
| "grad_norm": 1.5456216326803007, |
| "learning_rate": 6.62038851792793e-05, |
| "loss": 0.2177, |
| "step": 3732 |
| }, |
| { |
| "epoch": 1.2982791586998088, |
| "grad_norm": 1.6658163950679008, |
| "learning_rate": 6.608966198044883e-05, |
| "loss": 0.2622, |
| "step": 3734 |
| }, |
| { |
| "epoch": 1.2989744481140275, |
| "grad_norm": 1.4391588043296901, |
| "learning_rate": 6.597548875091848e-05, |
| "loss": 0.2879, |
| "step": 3736 |
| }, |
| { |
| "epoch": 1.2996697375282462, |
| "grad_norm": 2.1244177276843867, |
| "learning_rate": 6.58613656589307e-05, |
| "loss": 0.2771, |
| "step": 3738 |
| }, |
| { |
| "epoch": 1.300365026942465, |
| "grad_norm": 1.469349548022793, |
| "learning_rate": 6.574729287265399e-05, |
| "loss": 0.1854, |
| "step": 3740 |
| }, |
| { |
| "epoch": 1.3010603163566834, |
| "grad_norm": 2.7817821148014366, |
| "learning_rate": 6.563327056018267e-05, |
| "loss": 0.4997, |
| "step": 3742 |
| }, |
| { |
| "epoch": 1.301755605770902, |
| "grad_norm": 1.7093126091054744, |
| "learning_rate": 6.551929888953681e-05, |
| "loss": 0.3019, |
| "step": 3744 |
| }, |
| { |
| "epoch": 1.3024508951851208, |
| "grad_norm": 1.444842766225459, |
| "learning_rate": 6.540537802866175e-05, |
| "loss": 0.19, |
| "step": 3746 |
| }, |
| { |
| "epoch": 1.3031461845993395, |
| "grad_norm": 1.5367589064913105, |
| "learning_rate": 6.529150814542802e-05, |
| "loss": 0.4442, |
| "step": 3748 |
| }, |
| { |
| "epoch": 1.3038414740135582, |
| "grad_norm": 1.424746458651605, |
| "learning_rate": 6.517768940763104e-05, |
| "loss": 0.3721, |
| "step": 3750 |
| }, |
| { |
| "epoch": 1.3038414740135582, |
| "eval_loss": 0.7538126707077026, |
| "eval_runtime": 369.9643, |
| "eval_samples_per_second": 13.088, |
| "eval_steps_per_second": 0.411, |
| "step": 3750 |
| }, |
| { |
| "epoch": 1.3045367634277767, |
| "grad_norm": 2.5440141972178067, |
| "learning_rate": 6.506392198299078e-05, |
| "loss": 0.2451, |
| "step": 3752 |
| }, |
| { |
| "epoch": 1.3052320528419954, |
| "grad_norm": 1.621992326913084, |
| "learning_rate": 6.495020603915165e-05, |
| "loss": 0.2254, |
| "step": 3754 |
| }, |
| { |
| "epoch": 1.305927342256214, |
| "grad_norm": 0.9766906856068927, |
| "learning_rate": 6.483654174368224e-05, |
| "loss": 0.2668, |
| "step": 3756 |
| }, |
| { |
| "epoch": 1.3066226316704328, |
| "grad_norm": 1.4322778535817728, |
| "learning_rate": 6.472292926407497e-05, |
| "loss": 0.2825, |
| "step": 3758 |
| }, |
| { |
| "epoch": 1.3073179210846515, |
| "grad_norm": 0.9448697060897363, |
| "learning_rate": 6.460936876774593e-05, |
| "loss": 0.3319, |
| "step": 3760 |
| }, |
| { |
| "epoch": 1.3080132104988702, |
| "grad_norm": 1.874401140914359, |
| "learning_rate": 6.449586042203458e-05, |
| "loss": 0.3274, |
| "step": 3762 |
| }, |
| { |
| "epoch": 1.308708499913089, |
| "grad_norm": 1.7515138182196655, |
| "learning_rate": 6.43824043942036e-05, |
| "loss": 0.2732, |
| "step": 3764 |
| }, |
| { |
| "epoch": 1.3094037893273076, |
| "grad_norm": 2.890804651252811, |
| "learning_rate": 6.42690008514385e-05, |
| "loss": 0.2776, |
| "step": 3766 |
| }, |
| { |
| "epoch": 1.310099078741526, |
| "grad_norm": 2.46389565031376, |
| "learning_rate": 6.415564996084748e-05, |
| "loss": 0.3617, |
| "step": 3768 |
| }, |
| { |
| "epoch": 1.3107943681557448, |
| "grad_norm": 1.6157924498265324, |
| "learning_rate": 6.40423518894612e-05, |
| "loss": 0.3028, |
| "step": 3770 |
| }, |
| { |
| "epoch": 1.3114896575699635, |
| "grad_norm": 0.6223193072150819, |
| "learning_rate": 6.392910680423245e-05, |
| "loss": 0.1145, |
| "step": 3772 |
| }, |
| { |
| "epoch": 1.3121849469841822, |
| "grad_norm": 1.271457081386872, |
| "learning_rate": 6.381591487203586e-05, |
| "loss": 0.265, |
| "step": 3774 |
| }, |
| { |
| "epoch": 1.312880236398401, |
| "grad_norm": 0.7944760935872012, |
| "learning_rate": 6.370277625966785e-05, |
| "loss": 0.3318, |
| "step": 3776 |
| }, |
| { |
| "epoch": 1.3135755258126194, |
| "grad_norm": 2.2526555804743427, |
| "learning_rate": 6.358969113384623e-05, |
| "loss": 0.2307, |
| "step": 3778 |
| }, |
| { |
| "epoch": 1.314270815226838, |
| "grad_norm": 2.906733400348192, |
| "learning_rate": 6.347665966121e-05, |
| "loss": 0.1961, |
| "step": 3780 |
| }, |
| { |
| "epoch": 1.3149661046410568, |
| "grad_norm": 1.2262030792921825, |
| "learning_rate": 6.336368200831912e-05, |
| "loss": 0.2548, |
| "step": 3782 |
| }, |
| { |
| "epoch": 1.3156613940552755, |
| "grad_norm": 1.2924828828899246, |
| "learning_rate": 6.325075834165417e-05, |
| "loss": 0.2532, |
| "step": 3784 |
| }, |
| { |
| "epoch": 1.3163566834694942, |
| "grad_norm": 2.560892079700494, |
| "learning_rate": 6.313788882761622e-05, |
| "loss": 0.3552, |
| "step": 3786 |
| }, |
| { |
| "epoch": 1.317051972883713, |
| "grad_norm": 1.7666862302668824, |
| "learning_rate": 6.302507363252656e-05, |
| "loss": 0.1494, |
| "step": 3788 |
| }, |
| { |
| "epoch": 1.3177472622979316, |
| "grad_norm": 2.022248178981181, |
| "learning_rate": 6.291231292262642e-05, |
| "loss": 0.1941, |
| "step": 3790 |
| }, |
| { |
| "epoch": 1.3184425517121503, |
| "grad_norm": 4.829869973416474, |
| "learning_rate": 6.279960686407674e-05, |
| "loss": 0.3437, |
| "step": 3792 |
| }, |
| { |
| "epoch": 1.3191378411263688, |
| "grad_norm": 2.4009038310356363, |
| "learning_rate": 6.268695562295797e-05, |
| "loss": 0.2214, |
| "step": 3794 |
| }, |
| { |
| "epoch": 1.3198331305405875, |
| "grad_norm": 1.8416804223103658, |
| "learning_rate": 6.257435936526968e-05, |
| "loss": 0.2688, |
| "step": 3796 |
| }, |
| { |
| "epoch": 1.3205284199548062, |
| "grad_norm": 2.5769021257497964, |
| "learning_rate": 6.24618182569305e-05, |
| "loss": 0.2782, |
| "step": 3798 |
| }, |
| { |
| "epoch": 1.321223709369025, |
| "grad_norm": 3.2879866025429685, |
| "learning_rate": 6.234933246377776e-05, |
| "loss": 0.3447, |
| "step": 3800 |
| }, |
| { |
| "epoch": 1.3219189987832436, |
| "grad_norm": 1.9730317910403659, |
| "learning_rate": 6.223690215156735e-05, |
| "loss": 0.2464, |
| "step": 3802 |
| }, |
| { |
| "epoch": 1.322614288197462, |
| "grad_norm": 1.3835776718177815, |
| "learning_rate": 6.212452748597331e-05, |
| "loss": 0.2452, |
| "step": 3804 |
| }, |
| { |
| "epoch": 1.3233095776116808, |
| "grad_norm": 5.009933452432127, |
| "learning_rate": 6.201220863258776e-05, |
| "loss": 0.2693, |
| "step": 3806 |
| }, |
| { |
| "epoch": 1.3240048670258995, |
| "grad_norm": 1.9258413436362602, |
| "learning_rate": 6.189994575692046e-05, |
| "loss": 0.3064, |
| "step": 3808 |
| }, |
| { |
| "epoch": 1.3247001564401182, |
| "grad_norm": 1.3573089654999126, |
| "learning_rate": 6.17877390243988e-05, |
| "loss": 0.3491, |
| "step": 3810 |
| }, |
| { |
| "epoch": 1.3253954458543369, |
| "grad_norm": 1.6703010672898377, |
| "learning_rate": 6.167558860036743e-05, |
| "loss": 0.2993, |
| "step": 3812 |
| }, |
| { |
| "epoch": 1.3260907352685556, |
| "grad_norm": 0.7142035982668372, |
| "learning_rate": 6.156349465008793e-05, |
| "loss": 0.2143, |
| "step": 3814 |
| }, |
| { |
| "epoch": 1.3267860246827743, |
| "grad_norm": 2.3840586614188175, |
| "learning_rate": 6.145145733873878e-05, |
| "loss": 0.3697, |
| "step": 3816 |
| }, |
| { |
| "epoch": 1.327481314096993, |
| "grad_norm": 1.7009132660775388, |
| "learning_rate": 6.133947683141495e-05, |
| "loss": 0.2492, |
| "step": 3818 |
| }, |
| { |
| "epoch": 1.3281766035112115, |
| "grad_norm": 1.9063403370321392, |
| "learning_rate": 6.122755329312764e-05, |
| "loss": 0.2232, |
| "step": 3820 |
| }, |
| { |
| "epoch": 1.3288718929254302, |
| "grad_norm": 1.741032082142445, |
| "learning_rate": 6.111568688880418e-05, |
| "loss": 0.4976, |
| "step": 3822 |
| }, |
| { |
| "epoch": 1.3295671823396489, |
| "grad_norm": 1.467048841507196, |
| "learning_rate": 6.100387778328771e-05, |
| "loss": 0.2605, |
| "step": 3824 |
| }, |
| { |
| "epoch": 1.3302624717538676, |
| "grad_norm": 1.0458033385629528, |
| "learning_rate": 6.0892126141336924e-05, |
| "loss": 0.313, |
| "step": 3826 |
| }, |
| { |
| "epoch": 1.3309577611680863, |
| "grad_norm": 1.4064954749922551, |
| "learning_rate": 6.078043212762581e-05, |
| "loss": 0.1169, |
| "step": 3828 |
| }, |
| { |
| "epoch": 1.3316530505823048, |
| "grad_norm": 1.6190851182554558, |
| "learning_rate": 6.06687959067435e-05, |
| "loss": 0.3406, |
| "step": 3830 |
| }, |
| { |
| "epoch": 1.3323483399965235, |
| "grad_norm": 2.0033922207158406, |
| "learning_rate": 6.055721764319386e-05, |
| "loss": 0.2865, |
| "step": 3832 |
| }, |
| { |
| "epoch": 1.3330436294107422, |
| "grad_norm": 1.186764888106539, |
| "learning_rate": 6.044569750139546e-05, |
| "loss": 0.1976, |
| "step": 3834 |
| }, |
| { |
| "epoch": 1.3337389188249609, |
| "grad_norm": 2.0392288891589256, |
| "learning_rate": 6.033423564568116e-05, |
| "loss": 0.3134, |
| "step": 3836 |
| }, |
| { |
| "epoch": 1.3344342082391796, |
| "grad_norm": 1.1223381552889025, |
| "learning_rate": 6.0222832240297944e-05, |
| "loss": 0.1529, |
| "step": 3838 |
| }, |
| { |
| "epoch": 1.3351294976533983, |
| "grad_norm": 1.8730009155305403, |
| "learning_rate": 6.0111487449406686e-05, |
| "loss": 0.343, |
| "step": 3840 |
| }, |
| { |
| "epoch": 1.335824787067617, |
| "grad_norm": 1.4964630737873887, |
| "learning_rate": 6.0000201437081874e-05, |
| "loss": 0.3388, |
| "step": 3842 |
| }, |
| { |
| "epoch": 1.3365200764818357, |
| "grad_norm": 1.1458928429879986, |
| "learning_rate": 5.9888974367311356e-05, |
| "loss": 0.1954, |
| "step": 3844 |
| }, |
| { |
| "epoch": 1.3372153658960542, |
| "grad_norm": 1.1289358402451455, |
| "learning_rate": 5.977780640399615e-05, |
| "loss": 0.221, |
| "step": 3846 |
| }, |
| { |
| "epoch": 1.3379106553102729, |
| "grad_norm": 2.4325291833992972, |
| "learning_rate": 5.9666697710950193e-05, |
| "loss": 0.4714, |
| "step": 3848 |
| }, |
| { |
| "epoch": 1.3386059447244916, |
| "grad_norm": 1.831392069750196, |
| "learning_rate": 5.955564845190005e-05, |
| "loss": 0.4101, |
| "step": 3850 |
| }, |
| { |
| "epoch": 1.3393012341387103, |
| "grad_norm": 1.7638383306287577, |
| "learning_rate": 5.9444658790484765e-05, |
| "loss": 0.4282, |
| "step": 3852 |
| }, |
| { |
| "epoch": 1.339996523552929, |
| "grad_norm": 1.6124114221140404, |
| "learning_rate": 5.933372889025543e-05, |
| "loss": 0.2827, |
| "step": 3854 |
| }, |
| { |
| "epoch": 1.3406918129671475, |
| "grad_norm": 3.0727072769375736, |
| "learning_rate": 5.922285891467523e-05, |
| "loss": 0.3148, |
| "step": 3856 |
| }, |
| { |
| "epoch": 1.3413871023813662, |
| "grad_norm": 2.54006055065279, |
| "learning_rate": 5.911204902711893e-05, |
| "loss": 0.2994, |
| "step": 3858 |
| }, |
| { |
| "epoch": 1.3420823917955849, |
| "grad_norm": 2.2646413953956825, |
| "learning_rate": 5.9001299390872845e-05, |
| "loss": 0.2999, |
| "step": 3860 |
| }, |
| { |
| "epoch": 1.3427776812098036, |
| "grad_norm": 1.8123438443119315, |
| "learning_rate": 5.8890610169134416e-05, |
| "loss": 0.2568, |
| "step": 3862 |
| }, |
| { |
| "epoch": 1.3434729706240223, |
| "grad_norm": 2.0388591613476854, |
| "learning_rate": 5.877998152501214e-05, |
| "loss": 0.4053, |
| "step": 3864 |
| }, |
| { |
| "epoch": 1.344168260038241, |
| "grad_norm": 0.6144592205294669, |
| "learning_rate": 5.8669413621525184e-05, |
| "loss": 0.1518, |
| "step": 3866 |
| }, |
| { |
| "epoch": 1.3448635494524597, |
| "grad_norm": 1.296717012165212, |
| "learning_rate": 5.8558906621603215e-05, |
| "loss": 0.2309, |
| "step": 3868 |
| }, |
| { |
| "epoch": 1.3455588388666784, |
| "grad_norm": 1.4203081029306004, |
| "learning_rate": 5.8448460688086203e-05, |
| "loss": 0.2535, |
| "step": 3870 |
| }, |
| { |
| "epoch": 1.3462541282808969, |
| "grad_norm": 2.1863435973935994, |
| "learning_rate": 5.8338075983724103e-05, |
| "loss": 0.4183, |
| "step": 3872 |
| }, |
| { |
| "epoch": 1.3469494176951156, |
| "grad_norm": 2.1740201042725267, |
| "learning_rate": 5.822775267117666e-05, |
| "loss": 0.2978, |
| "step": 3874 |
| }, |
| { |
| "epoch": 1.3476447071093343, |
| "grad_norm": 1.743329931575654, |
| "learning_rate": 5.811749091301312e-05, |
| "loss": 0.2085, |
| "step": 3876 |
| }, |
| { |
| "epoch": 1.348339996523553, |
| "grad_norm": 2.670160094279744, |
| "learning_rate": 5.800729087171204e-05, |
| "loss": 0.2404, |
| "step": 3878 |
| }, |
| { |
| "epoch": 1.3490352859377717, |
| "grad_norm": 1.025913600197322, |
| "learning_rate": 5.789715270966098e-05, |
| "loss": 0.1275, |
| "step": 3880 |
| }, |
| { |
| "epoch": 1.3497305753519901, |
| "grad_norm": 1.590095494001877, |
| "learning_rate": 5.77870765891565e-05, |
| "loss": 0.2631, |
| "step": 3882 |
| }, |
| { |
| "epoch": 1.3504258647662088, |
| "grad_norm": 2.7173746484868553, |
| "learning_rate": 5.767706267240347e-05, |
| "loss": 0.2186, |
| "step": 3884 |
| }, |
| { |
| "epoch": 1.3511211541804276, |
| "grad_norm": 1.924510083321827, |
| "learning_rate": 5.7567111121515404e-05, |
| "loss": 0.2128, |
| "step": 3886 |
| }, |
| { |
| "epoch": 1.3518164435946463, |
| "grad_norm": 1.749104710457094, |
| "learning_rate": 5.745722209851359e-05, |
| "loss": 0.2309, |
| "step": 3888 |
| }, |
| { |
| "epoch": 1.352511733008865, |
| "grad_norm": 1.8922831244281024, |
| "learning_rate": 5.7347395765327416e-05, |
| "loss": 0.3782, |
| "step": 3890 |
| }, |
| { |
| "epoch": 1.3532070224230837, |
| "grad_norm": 2.5121612481404516, |
| "learning_rate": 5.723763228379375e-05, |
| "loss": 0.2758, |
| "step": 3892 |
| }, |
| { |
| "epoch": 1.3539023118373024, |
| "grad_norm": 1.668245436257098, |
| "learning_rate": 5.712793181565701e-05, |
| "loss": 0.3209, |
| "step": 3894 |
| }, |
| { |
| "epoch": 1.354597601251521, |
| "grad_norm": 1.4921480516245784, |
| "learning_rate": 5.7018294522568574e-05, |
| "loss": 0.292, |
| "step": 3896 |
| }, |
| { |
| "epoch": 1.3552928906657395, |
| "grad_norm": 2.2331555392226443, |
| "learning_rate": 5.690872056608685e-05, |
| "loss": 0.3069, |
| "step": 3898 |
| }, |
| { |
| "epoch": 1.3559881800799582, |
| "grad_norm": 0.9857569332376791, |
| "learning_rate": 5.6799210107676804e-05, |
| "loss": 0.1369, |
| "step": 3900 |
| }, |
| { |
| "epoch": 1.3559881800799582, |
| "eval_loss": 0.8018413186073303, |
| "eval_runtime": 370.269, |
| "eval_samples_per_second": 13.077, |
| "eval_steps_per_second": 0.411, |
| "step": 3900 |
| }, |
| { |
| "epoch": 1.356683469494177, |
| "grad_norm": 1.7361986647664482, |
| "learning_rate": 5.668976330871e-05, |
| "loss": 0.2356, |
| "step": 3902 |
| }, |
| { |
| "epoch": 1.3573787589083957, |
| "grad_norm": 1.6018675256429025, |
| "learning_rate": 5.6580380330464046e-05, |
| "loss": 0.1639, |
| "step": 3904 |
| }, |
| { |
| "epoch": 1.3580740483226144, |
| "grad_norm": 2.1622207693853546, |
| "learning_rate": 5.647106133412251e-05, |
| "loss": 0.3555, |
| "step": 3906 |
| }, |
| { |
| "epoch": 1.3587693377368328, |
| "grad_norm": 1.7774697055684008, |
| "learning_rate": 5.636180648077487e-05, |
| "loss": 0.2626, |
| "step": 3908 |
| }, |
| { |
| "epoch": 1.3594646271510515, |
| "grad_norm": 2.154613781964978, |
| "learning_rate": 5.625261593141575e-05, |
| "loss": 0.2915, |
| "step": 3910 |
| }, |
| { |
| "epoch": 1.3601599165652702, |
| "grad_norm": 3.0267380775574346, |
| "learning_rate": 5.614348984694535e-05, |
| "loss": 0.3314, |
| "step": 3912 |
| }, |
| { |
| "epoch": 1.360855205979489, |
| "grad_norm": 2.39396844611769, |
| "learning_rate": 5.603442838816866e-05, |
| "loss": 0.51, |
| "step": 3914 |
| }, |
| { |
| "epoch": 1.3615504953937076, |
| "grad_norm": 2.6484623504559197, |
| "learning_rate": 5.592543171579556e-05, |
| "loss": 0.2956, |
| "step": 3916 |
| }, |
| { |
| "epoch": 1.3622457848079264, |
| "grad_norm": 1.9214228464042193, |
| "learning_rate": 5.581649999044036e-05, |
| "loss": 0.351, |
| "step": 3918 |
| }, |
| { |
| "epoch": 1.362941074222145, |
| "grad_norm": 1.525669381009347, |
| "learning_rate": 5.570763337262185e-05, |
| "loss": 0.2471, |
| "step": 3920 |
| }, |
| { |
| "epoch": 1.3636363636363638, |
| "grad_norm": 2.2559633273979345, |
| "learning_rate": 5.559883202276257e-05, |
| "loss": 0.3488, |
| "step": 3922 |
| }, |
| { |
| "epoch": 1.3643316530505822, |
| "grad_norm": 1.0945389454078283, |
| "learning_rate": 5.5490096101189193e-05, |
| "loss": 0.1633, |
| "step": 3924 |
| }, |
| { |
| "epoch": 1.365026942464801, |
| "grad_norm": 1.878375405998012, |
| "learning_rate": 5.5381425768131814e-05, |
| "loss": 0.2935, |
| "step": 3926 |
| }, |
| { |
| "epoch": 1.3657222318790196, |
| "grad_norm": 1.3652403041403325, |
| "learning_rate": 5.527282118372385e-05, |
| "loss": 0.2414, |
| "step": 3928 |
| }, |
| { |
| "epoch": 1.3664175212932383, |
| "grad_norm": 1.126776675687968, |
| "learning_rate": 5.516428250800203e-05, |
| "loss": 0.1693, |
| "step": 3930 |
| }, |
| { |
| "epoch": 1.367112810707457, |
| "grad_norm": 0.7104597421901008, |
| "learning_rate": 5.505580990090574e-05, |
| "loss": 0.1815, |
| "step": 3932 |
| }, |
| { |
| "epoch": 1.3678081001216755, |
| "grad_norm": 1.2846548486452432, |
| "learning_rate": 5.494740352227712e-05, |
| "loss": 0.3151, |
| "step": 3934 |
| }, |
| { |
| "epoch": 1.3685033895358942, |
| "grad_norm": 2.182215142213049, |
| "learning_rate": 5.483906353186065e-05, |
| "loss": 0.3221, |
| "step": 3936 |
| }, |
| { |
| "epoch": 1.369198678950113, |
| "grad_norm": 0.8096292350036893, |
| "learning_rate": 5.473079008930311e-05, |
| "loss": 0.2464, |
| "step": 3938 |
| }, |
| { |
| "epoch": 1.3698939683643316, |
| "grad_norm": 1.2542393748073035, |
| "learning_rate": 5.462258335415305e-05, |
| "loss": 0.208, |
| "step": 3940 |
| }, |
| { |
| "epoch": 1.3705892577785503, |
| "grad_norm": 1.1726858390729638, |
| "learning_rate": 5.451444348586092e-05, |
| "loss": 0.2301, |
| "step": 3942 |
| }, |
| { |
| "epoch": 1.371284547192769, |
| "grad_norm": 2.5437836827721125, |
| "learning_rate": 5.4406370643778406e-05, |
| "loss": 0.2712, |
| "step": 3944 |
| }, |
| { |
| "epoch": 1.3719798366069877, |
| "grad_norm": 2.216635898323048, |
| "learning_rate": 5.4298364987158614e-05, |
| "loss": 0.3255, |
| "step": 3946 |
| }, |
| { |
| "epoch": 1.3726751260212064, |
| "grad_norm": 2.3765688609704974, |
| "learning_rate": 5.419042667515553e-05, |
| "loss": 0.3263, |
| "step": 3948 |
| }, |
| { |
| "epoch": 1.373370415435425, |
| "grad_norm": 1.7807661008591051, |
| "learning_rate": 5.4082555866824e-05, |
| "loss": 0.2537, |
| "step": 3950 |
| }, |
| { |
| "epoch": 1.3740657048496436, |
| "grad_norm": 1.271737872520489, |
| "learning_rate": 5.397475272111934e-05, |
| "loss": 0.1095, |
| "step": 3952 |
| }, |
| { |
| "epoch": 1.3747609942638623, |
| "grad_norm": 1.5819730073696412, |
| "learning_rate": 5.3867017396897146e-05, |
| "loss": 0.2785, |
| "step": 3954 |
| }, |
| { |
| "epoch": 1.375456283678081, |
| "grad_norm": 1.8102440412726852, |
| "learning_rate": 5.375935005291311e-05, |
| "loss": 0.184, |
| "step": 3956 |
| }, |
| { |
| "epoch": 1.3761515730922997, |
| "grad_norm": 1.424570307586645, |
| "learning_rate": 5.3651750847822704e-05, |
| "loss": 0.3571, |
| "step": 3958 |
| }, |
| { |
| "epoch": 1.3768468625065182, |
| "grad_norm": 1.233782931234406, |
| "learning_rate": 5.3544219940181086e-05, |
| "loss": 0.1869, |
| "step": 3960 |
| }, |
| { |
| "epoch": 1.377542151920737, |
| "grad_norm": 2.2794558736336543, |
| "learning_rate": 5.3436757488442646e-05, |
| "loss": 0.2808, |
| "step": 3962 |
| }, |
| { |
| "epoch": 1.3782374413349556, |
| "grad_norm": 2.5185533118376227, |
| "learning_rate": 5.332936365096105e-05, |
| "loss": 0.228, |
| "step": 3964 |
| }, |
| { |
| "epoch": 1.3789327307491743, |
| "grad_norm": 1.8067537932516826, |
| "learning_rate": 5.322203858598873e-05, |
| "loss": 0.194, |
| "step": 3966 |
| }, |
| { |
| "epoch": 1.379628020163393, |
| "grad_norm": 1.871647013612042, |
| "learning_rate": 5.31147824516768e-05, |
| "loss": 0.3135, |
| "step": 3968 |
| }, |
| { |
| "epoch": 1.3803233095776117, |
| "grad_norm": 1.6386317816659686, |
| "learning_rate": 5.300759540607481e-05, |
| "loss": 0.2202, |
| "step": 3970 |
| }, |
| { |
| "epoch": 1.3810185989918304, |
| "grad_norm": 1.7620222274285318, |
| "learning_rate": 5.2900477607130574e-05, |
| "loss": 0.1665, |
| "step": 3972 |
| }, |
| { |
| "epoch": 1.3817138884060491, |
| "grad_norm": 2.132390398077624, |
| "learning_rate": 5.279342921268977e-05, |
| "loss": 0.2432, |
| "step": 3974 |
| }, |
| { |
| "epoch": 1.3824091778202676, |
| "grad_norm": 1.8773248003654746, |
| "learning_rate": 5.268645038049582e-05, |
| "loss": 0.2609, |
| "step": 3976 |
| }, |
| { |
| "epoch": 1.3831044672344863, |
| "grad_norm": 2.7272632898802036, |
| "learning_rate": 5.2579541268189644e-05, |
| "loss": 0.2431, |
| "step": 3978 |
| }, |
| { |
| "epoch": 1.383799756648705, |
| "grad_norm": 2.0535523305707524, |
| "learning_rate": 5.247270203330952e-05, |
| "loss": 0.4326, |
| "step": 3980 |
| }, |
| { |
| "epoch": 1.3844950460629237, |
| "grad_norm": 2.4360469454472446, |
| "learning_rate": 5.236593283329061e-05, |
| "loss": 0.2236, |
| "step": 3982 |
| }, |
| { |
| "epoch": 1.3851903354771424, |
| "grad_norm": 1.6268465852911962, |
| "learning_rate": 5.225923382546491e-05, |
| "loss": 0.2584, |
| "step": 3984 |
| }, |
| { |
| "epoch": 1.385885624891361, |
| "grad_norm": 2.2902631169054786, |
| "learning_rate": 5.215260516706111e-05, |
| "loss": 0.2328, |
| "step": 3986 |
| }, |
| { |
| "epoch": 1.3865809143055796, |
| "grad_norm": 1.4559178942653874, |
| "learning_rate": 5.2046047015204084e-05, |
| "loss": 0.1566, |
| "step": 3988 |
| }, |
| { |
| "epoch": 1.3872762037197983, |
| "grad_norm": 0.8568173758296125, |
| "learning_rate": 5.193955952691487e-05, |
| "loss": 0.2278, |
| "step": 3990 |
| }, |
| { |
| "epoch": 1.387971493134017, |
| "grad_norm": 2.851718287307149, |
| "learning_rate": 5.1833142859110314e-05, |
| "loss": 0.2524, |
| "step": 3992 |
| }, |
| { |
| "epoch": 1.3886667825482357, |
| "grad_norm": 1.6613041608064776, |
| "learning_rate": 5.1726797168603066e-05, |
| "loss": 0.4466, |
| "step": 3994 |
| }, |
| { |
| "epoch": 1.3893620719624544, |
| "grad_norm": 1.7949877673256964, |
| "learning_rate": 5.1620522612101e-05, |
| "loss": 0.3236, |
| "step": 3996 |
| }, |
| { |
| "epoch": 1.3900573613766731, |
| "grad_norm": 1.725496491320523, |
| "learning_rate": 5.151431934620732e-05, |
| "loss": 0.2484, |
| "step": 3998 |
| }, |
| { |
| "epoch": 1.3907526507908918, |
| "grad_norm": 2.232006563508218, |
| "learning_rate": 5.140818752742006e-05, |
| "loss": 0.2542, |
| "step": 4000 |
| }, |
| { |
| "epoch": 1.3914479402051103, |
| "grad_norm": 2.200202446282767, |
| "learning_rate": 5.130212731213203e-05, |
| "loss": 0.1905, |
| "step": 4002 |
| }, |
| { |
| "epoch": 1.392143229619329, |
| "grad_norm": 1.0765722717243909, |
| "learning_rate": 5.119613885663053e-05, |
| "loss": 0.083, |
| "step": 4004 |
| }, |
| { |
| "epoch": 1.3928385190335477, |
| "grad_norm": 1.9643568364973791, |
| "learning_rate": 5.109022231709704e-05, |
| "loss": 0.3048, |
| "step": 4006 |
| }, |
| { |
| "epoch": 1.3935338084477664, |
| "grad_norm": 2.200825711096068, |
| "learning_rate": 5.098437784960721e-05, |
| "loss": 0.3176, |
| "step": 4008 |
| }, |
| { |
| "epoch": 1.3942290978619851, |
| "grad_norm": 1.7528062866711556, |
| "learning_rate": 5.08786056101304e-05, |
| "loss": 0.2398, |
| "step": 4010 |
| }, |
| { |
| "epoch": 1.3949243872762036, |
| "grad_norm": 1.8567915577700007, |
| "learning_rate": 5.077290575452951e-05, |
| "loss": 0.204, |
| "step": 4012 |
| }, |
| { |
| "epoch": 1.3956196766904223, |
| "grad_norm": 1.053508655589504, |
| "learning_rate": 5.0667278438560783e-05, |
| "loss": 0.1996, |
| "step": 4014 |
| }, |
| { |
| "epoch": 1.396314966104641, |
| "grad_norm": 1.59190358848796, |
| "learning_rate": 5.0561723817873686e-05, |
| "loss": 0.1966, |
| "step": 4016 |
| }, |
| { |
| "epoch": 1.3970102555188597, |
| "grad_norm": 2.214401360672347, |
| "learning_rate": 5.0456242048010385e-05, |
| "loss": 0.3303, |
| "step": 4018 |
| }, |
| { |
| "epoch": 1.3977055449330784, |
| "grad_norm": 2.8646390206549617, |
| "learning_rate": 5.035083328440588e-05, |
| "loss": 0.3468, |
| "step": 4020 |
| }, |
| { |
| "epoch": 1.3984008343472971, |
| "grad_norm": 2.0445589179229313, |
| "learning_rate": 5.0245497682387446e-05, |
| "loss": 0.2855, |
| "step": 4022 |
| }, |
| { |
| "epoch": 1.3990961237615158, |
| "grad_norm": 1.9311580015983607, |
| "learning_rate": 5.014023539717458e-05, |
| "loss": 0.2142, |
| "step": 4024 |
| }, |
| { |
| "epoch": 1.3997914131757345, |
| "grad_norm": 1.2954578242074088, |
| "learning_rate": 5.003504658387876e-05, |
| "loss": 0.2394, |
| "step": 4026 |
| }, |
| { |
| "epoch": 1.400486702589953, |
| "grad_norm": 1.9646962338061735, |
| "learning_rate": 4.992993139750328e-05, |
| "loss": 0.3141, |
| "step": 4028 |
| }, |
| { |
| "epoch": 1.4011819920041717, |
| "grad_norm": 1.605707660220532, |
| "learning_rate": 4.982488999294279e-05, |
| "loss": 0.1654, |
| "step": 4030 |
| }, |
| { |
| "epoch": 1.4018772814183904, |
| "grad_norm": 1.3839678417234758, |
| "learning_rate": 4.971992252498327e-05, |
| "loss": 0.2028, |
| "step": 4032 |
| }, |
| { |
| "epoch": 1.4025725708326091, |
| "grad_norm": 2.0389449010271217, |
| "learning_rate": 4.9615029148301874e-05, |
| "loss": 0.2146, |
| "step": 4034 |
| }, |
| { |
| "epoch": 1.4032678602468278, |
| "grad_norm": 2.1085188012593488, |
| "learning_rate": 4.951021001746631e-05, |
| "loss": 0.3027, |
| "step": 4036 |
| }, |
| { |
| "epoch": 1.4039631496610463, |
| "grad_norm": 1.96976676512362, |
| "learning_rate": 4.940546528693516e-05, |
| "loss": 0.1932, |
| "step": 4038 |
| }, |
| { |
| "epoch": 1.404658439075265, |
| "grad_norm": 1.5987587366212639, |
| "learning_rate": 4.930079511105719e-05, |
| "loss": 0.1576, |
| "step": 4040 |
| }, |
| { |
| "epoch": 1.4053537284894837, |
| "grad_norm": 1.7788529668811892, |
| "learning_rate": 4.9196199644071404e-05, |
| "loss": 0.2504, |
| "step": 4042 |
| }, |
| { |
| "epoch": 1.4060490179037024, |
| "grad_norm": 1.9659292849812864, |
| "learning_rate": 4.909167904010666e-05, |
| "loss": 0.2339, |
| "step": 4044 |
| }, |
| { |
| "epoch": 1.406744307317921, |
| "grad_norm": 2.604990685917443, |
| "learning_rate": 4.8987233453181534e-05, |
| "loss": 0.3037, |
| "step": 4046 |
| }, |
| { |
| "epoch": 1.4074395967321398, |
| "grad_norm": 1.663670445381357, |
| "learning_rate": 4.888286303720398e-05, |
| "loss": 0.3984, |
| "step": 4048 |
| }, |
| { |
| "epoch": 1.4081348861463585, |
| "grad_norm": 1.993985314452789, |
| "learning_rate": 4.8778567945971346e-05, |
| "loss": 0.1987, |
| "step": 4050 |
| }, |
| { |
| "epoch": 1.4081348861463585, |
| "eval_loss": 0.8215903639793396, |
| "eval_runtime": 370.1142, |
| "eval_samples_per_second": 13.082, |
| "eval_steps_per_second": 0.411, |
| "step": 4050 |
| }, |
| { |
| "epoch": 1.4088301755605772, |
| "grad_norm": 1.2926112025819303, |
| "learning_rate": 4.867434833316982e-05, |
| "loss": 0.1994, |
| "step": 4052 |
| }, |
| { |
| "epoch": 1.4095254649747957, |
| "grad_norm": 2.0334536418120273, |
| "learning_rate": 4.8570204352374395e-05, |
| "loss": 0.3109, |
| "step": 4054 |
| }, |
| { |
| "epoch": 1.4102207543890144, |
| "grad_norm": 1.7071513805170115, |
| "learning_rate": 4.846613615704875e-05, |
| "loss": 0.2079, |
| "step": 4056 |
| }, |
| { |
| "epoch": 1.410916043803233, |
| "grad_norm": 0.8284654929989167, |
| "learning_rate": 4.836214390054472e-05, |
| "loss": 0.1428, |
| "step": 4058 |
| }, |
| { |
| "epoch": 1.4116113332174518, |
| "grad_norm": 2.1877463104140755, |
| "learning_rate": 4.8258227736102326e-05, |
| "loss": 0.3286, |
| "step": 4060 |
| }, |
| { |
| "epoch": 1.4123066226316705, |
| "grad_norm": 1.6394800348866228, |
| "learning_rate": 4.81543878168494e-05, |
| "loss": 0.2243, |
| "step": 4062 |
| }, |
| { |
| "epoch": 1.413001912045889, |
| "grad_norm": 2.29045701908158, |
| "learning_rate": 4.805062429580154e-05, |
| "loss": 0.2754, |
| "step": 4064 |
| }, |
| { |
| "epoch": 1.4136972014601077, |
| "grad_norm": 1.1593789840994158, |
| "learning_rate": 4.794693732586162e-05, |
| "loss": 0.2589, |
| "step": 4066 |
| }, |
| { |
| "epoch": 1.4143924908743264, |
| "grad_norm": 1.9076779586080495, |
| "learning_rate": 4.784332705981989e-05, |
| "loss": 0.3744, |
| "step": 4068 |
| }, |
| { |
| "epoch": 1.415087780288545, |
| "grad_norm": 1.0605351764371163, |
| "learning_rate": 4.773979365035331e-05, |
| "loss": 0.241, |
| "step": 4070 |
| }, |
| { |
| "epoch": 1.4157830697027638, |
| "grad_norm": 0.9400919890616967, |
| "learning_rate": 4.763633725002584e-05, |
| "loss": 0.1269, |
| "step": 4072 |
| }, |
| { |
| "epoch": 1.4164783591169825, |
| "grad_norm": 1.952528585538757, |
| "learning_rate": 4.753295801128782e-05, |
| "loss": 0.201, |
| "step": 4074 |
| }, |
| { |
| "epoch": 1.4171736485312012, |
| "grad_norm": 1.0134437003802954, |
| "learning_rate": 4.742965608647596e-05, |
| "loss": 0.179, |
| "step": 4076 |
| }, |
| { |
| "epoch": 1.41786893794542, |
| "grad_norm": 2.206549348385166, |
| "learning_rate": 4.732643162781301e-05, |
| "loss": 0.2293, |
| "step": 4078 |
| }, |
| { |
| "epoch": 1.4185642273596384, |
| "grad_norm": 1.9298045034198466, |
| "learning_rate": 4.722328478740754e-05, |
| "loss": 0.3132, |
| "step": 4080 |
| }, |
| { |
| "epoch": 1.419259516773857, |
| "grad_norm": 2.434300478677812, |
| "learning_rate": 4.712021571725378e-05, |
| "loss": 0.2811, |
| "step": 4082 |
| }, |
| { |
| "epoch": 1.4199548061880758, |
| "grad_norm": 1.1872316754317065, |
| "learning_rate": 4.701722456923131e-05, |
| "loss": 0.1675, |
| "step": 4084 |
| }, |
| { |
| "epoch": 1.4206500956022945, |
| "grad_norm": 2.0454164665381764, |
| "learning_rate": 4.691431149510501e-05, |
| "loss": 0.1872, |
| "step": 4086 |
| }, |
| { |
| "epoch": 1.4213453850165132, |
| "grad_norm": 2.6911117289279116, |
| "learning_rate": 4.681147664652455e-05, |
| "loss": 0.3075, |
| "step": 4088 |
| }, |
| { |
| "epoch": 1.4220406744307317, |
| "grad_norm": 2.341587620734492, |
| "learning_rate": 4.6708720175024536e-05, |
| "loss": 0.4096, |
| "step": 4090 |
| }, |
| { |
| "epoch": 1.4227359638449504, |
| "grad_norm": 2.7721201662792145, |
| "learning_rate": 4.660604223202378e-05, |
| "loss": 0.3148, |
| "step": 4092 |
| }, |
| { |
| "epoch": 1.423431253259169, |
| "grad_norm": 1.456250697767226, |
| "learning_rate": 4.650344296882568e-05, |
| "loss": 0.3385, |
| "step": 4094 |
| }, |
| { |
| "epoch": 1.4241265426733878, |
| "grad_norm": 1.3200715925518827, |
| "learning_rate": 4.640092253661749e-05, |
| "loss": 0.1961, |
| "step": 4096 |
| }, |
| { |
| "epoch": 1.4248218320876065, |
| "grad_norm": 2.3885713656146734, |
| "learning_rate": 4.629848108647043e-05, |
| "loss": 0.281, |
| "step": 4098 |
| }, |
| { |
| "epoch": 1.4255171215018252, |
| "grad_norm": 1.635923379123182, |
| "learning_rate": 4.619611876933925e-05, |
| "loss": 0.2244, |
| "step": 4100 |
| }, |
| { |
| "epoch": 1.426212410916044, |
| "grad_norm": 2.392775751436181, |
| "learning_rate": 4.609383573606213e-05, |
| "loss": 0.3849, |
| "step": 4102 |
| }, |
| { |
| "epoch": 1.4269077003302626, |
| "grad_norm": 1.3920312244467965, |
| "learning_rate": 4.599163213736036e-05, |
| "loss": 0.2543, |
| "step": 4104 |
| }, |
| { |
| "epoch": 1.427602989744481, |
| "grad_norm": 1.386923871502494, |
| "learning_rate": 4.588950812383832e-05, |
| "loss": 0.1941, |
| "step": 4106 |
| }, |
| { |
| "epoch": 1.4282982791586998, |
| "grad_norm": 2.3219780383747755, |
| "learning_rate": 4.578746384598296e-05, |
| "loss": 0.1857, |
| "step": 4108 |
| }, |
| { |
| "epoch": 1.4289935685729185, |
| "grad_norm": 2.3729132211845476, |
| "learning_rate": 4.5685499454163785e-05, |
| "loss": 0.4203, |
| "step": 4110 |
| }, |
| { |
| "epoch": 1.4296888579871372, |
| "grad_norm": 2.1324448188336067, |
| "learning_rate": 4.558361509863268e-05, |
| "loss": 0.3006, |
| "step": 4112 |
| }, |
| { |
| "epoch": 1.4303841474013559, |
| "grad_norm": 1.6516808941746237, |
| "learning_rate": 4.548181092952338e-05, |
| "loss": 0.2897, |
| "step": 4114 |
| }, |
| { |
| "epoch": 1.4310794368155744, |
| "grad_norm": 1.3205545315678033, |
| "learning_rate": 4.53800870968517e-05, |
| "loss": 0.2401, |
| "step": 4116 |
| }, |
| { |
| "epoch": 1.431774726229793, |
| "grad_norm": 1.7436544435374621, |
| "learning_rate": 4.527844375051485e-05, |
| "loss": 0.2569, |
| "step": 4118 |
| }, |
| { |
| "epoch": 1.4324700156440118, |
| "grad_norm": 1.9989349116349429, |
| "learning_rate": 4.517688104029165e-05, |
| "loss": 0.3491, |
| "step": 4120 |
| }, |
| { |
| "epoch": 1.4331653050582305, |
| "grad_norm": 1.9576009534010426, |
| "learning_rate": 4.507539911584191e-05, |
| "loss": 0.2261, |
| "step": 4122 |
| }, |
| { |
| "epoch": 1.4338605944724492, |
| "grad_norm": 1.6516212508066326, |
| "learning_rate": 4.497399812670658e-05, |
| "loss": 0.2482, |
| "step": 4124 |
| }, |
| { |
| "epoch": 1.4345558838866679, |
| "grad_norm": 2.272777724594952, |
| "learning_rate": 4.487267822230713e-05, |
| "loss": 0.2089, |
| "step": 4126 |
| }, |
| { |
| "epoch": 1.4352511733008866, |
| "grad_norm": 2.4736473517029456, |
| "learning_rate": 4.4771439551945726e-05, |
| "loss": 0.254, |
| "step": 4128 |
| }, |
| { |
| "epoch": 1.4359464627151053, |
| "grad_norm": 1.4673364182003128, |
| "learning_rate": 4.467028226480476e-05, |
| "loss": 0.2565, |
| "step": 4130 |
| }, |
| { |
| "epoch": 1.4366417521293238, |
| "grad_norm": 1.7889758290588789, |
| "learning_rate": 4.456920650994666e-05, |
| "loss": 0.1879, |
| "step": 4132 |
| }, |
| { |
| "epoch": 1.4373370415435425, |
| "grad_norm": 2.0477572135399713, |
| "learning_rate": 4.446821243631383e-05, |
| "loss": 0.204, |
| "step": 4134 |
| }, |
| { |
| "epoch": 1.4380323309577612, |
| "grad_norm": 2.1805910063720533, |
| "learning_rate": 4.436730019272819e-05, |
| "loss": 0.2547, |
| "step": 4136 |
| }, |
| { |
| "epoch": 1.4387276203719799, |
| "grad_norm": 1.9319389928527373, |
| "learning_rate": 4.426646992789114e-05, |
| "loss": 0.3026, |
| "step": 4138 |
| }, |
| { |
| "epoch": 1.4394229097861986, |
| "grad_norm": 1.0946954132339648, |
| "learning_rate": 4.416572179038322e-05, |
| "loss": 0.187, |
| "step": 4140 |
| }, |
| { |
| "epoch": 1.440118199200417, |
| "grad_norm": 2.1277556478497783, |
| "learning_rate": 4.406505592866404e-05, |
| "loss": 0.2602, |
| "step": 4142 |
| }, |
| { |
| "epoch": 1.4408134886146358, |
| "grad_norm": 2.53470768388119, |
| "learning_rate": 4.39644724910719e-05, |
| "loss": 0.3441, |
| "step": 4144 |
| }, |
| { |
| "epoch": 1.4415087780288545, |
| "grad_norm": 1.800880125581573, |
| "learning_rate": 4.386397162582373e-05, |
| "loss": 0.2163, |
| "step": 4146 |
| }, |
| { |
| "epoch": 1.4422040674430732, |
| "grad_norm": 1.2250300739306559, |
| "learning_rate": 4.376355348101459e-05, |
| "loss": 0.1879, |
| "step": 4148 |
| }, |
| { |
| "epoch": 1.4428993568572919, |
| "grad_norm": 2.91377651616678, |
| "learning_rate": 4.366321820461787e-05, |
| "loss": 0.266, |
| "step": 4150 |
| }, |
| { |
| "epoch": 1.4435946462715106, |
| "grad_norm": 0.8885674922158777, |
| "learning_rate": 4.3562965944484705e-05, |
| "loss": 0.1393, |
| "step": 4152 |
| }, |
| { |
| "epoch": 1.4442899356857293, |
| "grad_norm": 2.28353615381814, |
| "learning_rate": 4.3462796848344e-05, |
| "loss": 0.2155, |
| "step": 4154 |
| }, |
| { |
| "epoch": 1.444985225099948, |
| "grad_norm": 2.1440873985272684, |
| "learning_rate": 4.336271106380203e-05, |
| "loss": 0.1754, |
| "step": 4156 |
| }, |
| { |
| "epoch": 1.4456805145141665, |
| "grad_norm": 1.9129559236307636, |
| "learning_rate": 4.326270873834233e-05, |
| "loss": 0.3572, |
| "step": 4158 |
| }, |
| { |
| "epoch": 1.4463758039283852, |
| "grad_norm": 1.6962048087717518, |
| "learning_rate": 4.316279001932544e-05, |
| "loss": 0.2694, |
| "step": 4160 |
| }, |
| { |
| "epoch": 1.4470710933426039, |
| "grad_norm": 1.7138811037259571, |
| "learning_rate": 4.3062955053988695e-05, |
| "loss": 0.4043, |
| "step": 4162 |
| }, |
| { |
| "epoch": 1.4477663827568226, |
| "grad_norm": 1.2228126880390218, |
| "learning_rate": 4.296320398944611e-05, |
| "loss": 0.2559, |
| "step": 4164 |
| }, |
| { |
| "epoch": 1.4484616721710413, |
| "grad_norm": 1.6253628424429731, |
| "learning_rate": 4.28635369726879e-05, |
| "loss": 0.254, |
| "step": 4166 |
| }, |
| { |
| "epoch": 1.4491569615852598, |
| "grad_norm": 2.5329810986473813, |
| "learning_rate": 4.27639541505806e-05, |
| "loss": 0.3303, |
| "step": 4168 |
| }, |
| { |
| "epoch": 1.4498522509994785, |
| "grad_norm": 2.2952746855044794, |
| "learning_rate": 4.266445566986657e-05, |
| "loss": 0.3637, |
| "step": 4170 |
| }, |
| { |
| "epoch": 1.4505475404136972, |
| "grad_norm": 1.815951609291758, |
| "learning_rate": 4.256504167716389e-05, |
| "loss": 0.2179, |
| "step": 4172 |
| }, |
| { |
| "epoch": 1.4512428298279159, |
| "grad_norm": 1.3000244788207551, |
| "learning_rate": 4.2465712318966135e-05, |
| "loss": 0.1963, |
| "step": 4174 |
| }, |
| { |
| "epoch": 1.4519381192421346, |
| "grad_norm": 2.0274136674663996, |
| "learning_rate": 4.236646774164228e-05, |
| "loss": 0.2949, |
| "step": 4176 |
| }, |
| { |
| "epoch": 1.4526334086563533, |
| "grad_norm": 1.1360287047769313, |
| "learning_rate": 4.2267308091436234e-05, |
| "loss": 0.1784, |
| "step": 4178 |
| }, |
| { |
| "epoch": 1.453328698070572, |
| "grad_norm": 2.968136759801442, |
| "learning_rate": 4.216823351446682e-05, |
| "loss": 0.2602, |
| "step": 4180 |
| }, |
| { |
| "epoch": 1.4540239874847907, |
| "grad_norm": 1.7024994214096072, |
| "learning_rate": 4.206924415672744e-05, |
| "loss": 0.3289, |
| "step": 4182 |
| }, |
| { |
| "epoch": 1.4547192768990092, |
| "grad_norm": 2.3179695420629205, |
| "learning_rate": 4.197034016408605e-05, |
| "loss": 0.2539, |
| "step": 4184 |
| }, |
| { |
| "epoch": 1.4554145663132279, |
| "grad_norm": 1.7044194668309192, |
| "learning_rate": 4.18715216822847e-05, |
| "loss": 0.2625, |
| "step": 4186 |
| }, |
| { |
| "epoch": 1.4561098557274466, |
| "grad_norm": 3.3724003866954835, |
| "learning_rate": 4.1772788856939426e-05, |
| "loss": 0.2432, |
| "step": 4188 |
| }, |
| { |
| "epoch": 1.4568051451416653, |
| "grad_norm": 1.599492363871037, |
| "learning_rate": 4.167414183354016e-05, |
| "loss": 0.281, |
| "step": 4190 |
| }, |
| { |
| "epoch": 1.457500434555884, |
| "grad_norm": 0.9790835428522326, |
| "learning_rate": 4.157558075745027e-05, |
| "loss": 0.1965, |
| "step": 4192 |
| }, |
| { |
| "epoch": 1.4581957239701024, |
| "grad_norm": 1.1107576391833418, |
| "learning_rate": 4.147710577390654e-05, |
| "loss": 0.2249, |
| "step": 4194 |
| }, |
| { |
| "epoch": 1.4588910133843211, |
| "grad_norm": 1.9762973035181366, |
| "learning_rate": 4.137871702801885e-05, |
| "loss": 0.1317, |
| "step": 4196 |
| }, |
| { |
| "epoch": 1.4595863027985398, |
| "grad_norm": 1.6543005480458577, |
| "learning_rate": 4.128041466477008e-05, |
| "loss": 0.2481, |
| "step": 4198 |
| }, |
| { |
| "epoch": 1.4602815922127586, |
| "grad_norm": 1.0234088598666444, |
| "learning_rate": 4.118219882901572e-05, |
| "loss": 0.1397, |
| "step": 4200 |
| }, |
| { |
| "epoch": 1.4602815922127586, |
| "eval_loss": 0.8054793477058411, |
| "eval_runtime": 370.5501, |
| "eval_samples_per_second": 13.067, |
| "eval_steps_per_second": 0.41, |
| "step": 4200 |
| }, |
| { |
| "epoch": 1.4609768816269773, |
| "grad_norm": 1.205129289182479, |
| "learning_rate": 4.108406966548386e-05, |
| "loss": 0.2656, |
| "step": 4202 |
| }, |
| { |
| "epoch": 1.461672171041196, |
| "grad_norm": 1.1941309230087833, |
| "learning_rate": 4.098602731877479e-05, |
| "loss": 0.1182, |
| "step": 4204 |
| }, |
| { |
| "epoch": 1.4623674604554147, |
| "grad_norm": 1.2844012126052484, |
| "learning_rate": 4.088807193336089e-05, |
| "loss": 0.2754, |
| "step": 4206 |
| }, |
| { |
| "epoch": 1.4630627498696334, |
| "grad_norm": 1.698420166703449, |
| "learning_rate": 4.0790203653586414e-05, |
| "loss": 0.1675, |
| "step": 4208 |
| }, |
| { |
| "epoch": 1.4637580392838518, |
| "grad_norm": 2.9849246945771655, |
| "learning_rate": 4.069242262366719e-05, |
| "loss": 0.2109, |
| "step": 4210 |
| }, |
| { |
| "epoch": 1.4644533286980705, |
| "grad_norm": 1.874152841760584, |
| "learning_rate": 4.059472898769063e-05, |
| "loss": 0.2669, |
| "step": 4212 |
| }, |
| { |
| "epoch": 1.4651486181122892, |
| "grad_norm": 1.0555514585729977, |
| "learning_rate": 4.049712288961523e-05, |
| "loss": 0.1128, |
| "step": 4214 |
| }, |
| { |
| "epoch": 1.465843907526508, |
| "grad_norm": 1.8362636154896823, |
| "learning_rate": 4.0399604473270514e-05, |
| "loss": 0.1677, |
| "step": 4216 |
| }, |
| { |
| "epoch": 1.4665391969407267, |
| "grad_norm": 1.7738088257772673, |
| "learning_rate": 4.030217388235681e-05, |
| "loss": 0.1927, |
| "step": 4218 |
| }, |
| { |
| "epoch": 1.4672344863549451, |
| "grad_norm": 1.7097555028973528, |
| "learning_rate": 4.0204831260445084e-05, |
| "loss": 0.297, |
| "step": 4220 |
| }, |
| { |
| "epoch": 1.4679297757691638, |
| "grad_norm": 3.4118503905959536, |
| "learning_rate": 4.010757675097658e-05, |
| "loss": 0.3548, |
| "step": 4222 |
| }, |
| { |
| "epoch": 1.4686250651833825, |
| "grad_norm": 1.9412232483222223, |
| "learning_rate": 4.001041049726281e-05, |
| "loss": 0.3205, |
| "step": 4224 |
| }, |
| { |
| "epoch": 1.4693203545976012, |
| "grad_norm": 2.351944585629756, |
| "learning_rate": 3.991333264248513e-05, |
| "loss": 0.1789, |
| "step": 4226 |
| }, |
| { |
| "epoch": 1.47001564401182, |
| "grad_norm": 2.77981787149803, |
| "learning_rate": 3.9816343329694704e-05, |
| "loss": 0.4088, |
| "step": 4228 |
| }, |
| { |
| "epoch": 1.4707109334260386, |
| "grad_norm": 0.9224714463080386, |
| "learning_rate": 3.9719442701812136e-05, |
| "loss": 0.1771, |
| "step": 4230 |
| }, |
| { |
| "epoch": 1.4714062228402573, |
| "grad_norm": 3.0824480727396204, |
| "learning_rate": 3.96226309016275e-05, |
| "loss": 0.3542, |
| "step": 4232 |
| }, |
| { |
| "epoch": 1.472101512254476, |
| "grad_norm": 2.9165345999428376, |
| "learning_rate": 3.9525908071799834e-05, |
| "loss": 0.2431, |
| "step": 4234 |
| }, |
| { |
| "epoch": 1.4727968016686945, |
| "grad_norm": 1.8620585696734466, |
| "learning_rate": 3.942927435485708e-05, |
| "loss": 0.1597, |
| "step": 4236 |
| }, |
| { |
| "epoch": 1.4734920910829132, |
| "grad_norm": 1.2441583214781202, |
| "learning_rate": 3.933272989319604e-05, |
| "loss": 0.1882, |
| "step": 4238 |
| }, |
| { |
| "epoch": 1.474187380497132, |
| "grad_norm": 1.1066986528371985, |
| "learning_rate": 3.92362748290817e-05, |
| "loss": 0.1922, |
| "step": 4240 |
| }, |
| { |
| "epoch": 1.4748826699113506, |
| "grad_norm": 1.610275666207961, |
| "learning_rate": 3.913990930464758e-05, |
| "loss": 0.1412, |
| "step": 4242 |
| }, |
| { |
| "epoch": 1.4755779593255693, |
| "grad_norm": 1.8162225161291754, |
| "learning_rate": 3.904363346189508e-05, |
| "loss": 0.1995, |
| "step": 4244 |
| }, |
| { |
| "epoch": 1.4762732487397878, |
| "grad_norm": 1.982160632284015, |
| "learning_rate": 3.8947447442693585e-05, |
| "loss": 0.3077, |
| "step": 4246 |
| }, |
| { |
| "epoch": 1.4769685381540065, |
| "grad_norm": 2.3742426996369046, |
| "learning_rate": 3.885135138877999e-05, |
| "loss": 0.1937, |
| "step": 4248 |
| }, |
| { |
| "epoch": 1.4776638275682252, |
| "grad_norm": 3.75841446119527, |
| "learning_rate": 3.875534544175876e-05, |
| "loss": 0.4003, |
| "step": 4250 |
| }, |
| { |
| "epoch": 1.478359116982444, |
| "grad_norm": 2.1638375714521323, |
| "learning_rate": 3.865942974310139e-05, |
| "loss": 0.1966, |
| "step": 4252 |
| }, |
| { |
| "epoch": 1.4790544063966626, |
| "grad_norm": 2.974030616796864, |
| "learning_rate": 3.85636044341466e-05, |
| "loss": 0.2758, |
| "step": 4254 |
| }, |
| { |
| "epoch": 1.4797496958108813, |
| "grad_norm": 1.8926181608656496, |
| "learning_rate": 3.846786965609977e-05, |
| "loss": 0.1541, |
| "step": 4256 |
| }, |
| { |
| "epoch": 1.4804449852251, |
| "grad_norm": 0.9112054964455437, |
| "learning_rate": 3.837222555003287e-05, |
| "loss": 0.2522, |
| "step": 4258 |
| }, |
| { |
| "epoch": 1.4811402746393187, |
| "grad_norm": 0.9131730121835929, |
| "learning_rate": 3.827667225688444e-05, |
| "loss": 0.2207, |
| "step": 4260 |
| }, |
| { |
| "epoch": 1.4818355640535372, |
| "grad_norm": 1.8703351191767925, |
| "learning_rate": 3.81812099174589e-05, |
| "loss": 0.2774, |
| "step": 4262 |
| }, |
| { |
| "epoch": 1.482530853467756, |
| "grad_norm": 2.239704385434439, |
| "learning_rate": 3.808583867242694e-05, |
| "loss": 0.18, |
| "step": 4264 |
| }, |
| { |
| "epoch": 1.4832261428819746, |
| "grad_norm": 2.053956286354072, |
| "learning_rate": 3.7990558662324816e-05, |
| "loss": 0.1684, |
| "step": 4266 |
| }, |
| { |
| "epoch": 1.4839214322961933, |
| "grad_norm": 1.3838070739938082, |
| "learning_rate": 3.789537002755448e-05, |
| "loss": 0.2579, |
| "step": 4268 |
| }, |
| { |
| "epoch": 1.484616721710412, |
| "grad_norm": 2.655706752478529, |
| "learning_rate": 3.7800272908383084e-05, |
| "loss": 0.2681, |
| "step": 4270 |
| }, |
| { |
| "epoch": 1.4853120111246305, |
| "grad_norm": 1.1156606404182443, |
| "learning_rate": 3.770526744494314e-05, |
| "loss": 0.197, |
| "step": 4272 |
| }, |
| { |
| "epoch": 1.4860073005388492, |
| "grad_norm": 1.9260401094303623, |
| "learning_rate": 3.761035377723182e-05, |
| "loss": 0.1214, |
| "step": 4274 |
| }, |
| { |
| "epoch": 1.486702589953068, |
| "grad_norm": 2.1702994561727333, |
| "learning_rate": 3.751553204511128e-05, |
| "loss": 0.2046, |
| "step": 4276 |
| }, |
| { |
| "epoch": 1.4873978793672866, |
| "grad_norm": 3.141257123075563, |
| "learning_rate": 3.742080238830804e-05, |
| "loss": 0.112, |
| "step": 4278 |
| }, |
| { |
| "epoch": 1.4880931687815053, |
| "grad_norm": 2.170284970553849, |
| "learning_rate": 3.732616494641307e-05, |
| "loss": 0.1903, |
| "step": 4280 |
| }, |
| { |
| "epoch": 1.488788458195724, |
| "grad_norm": 2.021011549547257, |
| "learning_rate": 3.723161985888134e-05, |
| "loss": 0.2195, |
| "step": 4282 |
| }, |
| { |
| "epoch": 1.4894837476099427, |
| "grad_norm": 2.0088519149200246, |
| "learning_rate": 3.713716726503179e-05, |
| "loss": 0.1824, |
| "step": 4284 |
| }, |
| { |
| "epoch": 1.4901790370241614, |
| "grad_norm": 2.1469625632055376, |
| "learning_rate": 3.704280730404705e-05, |
| "loss": 0.2323, |
| "step": 4286 |
| }, |
| { |
| "epoch": 1.49087432643838, |
| "grad_norm": 2.497120708687944, |
| "learning_rate": 3.6948540114973186e-05, |
| "loss": 0.1967, |
| "step": 4288 |
| }, |
| { |
| "epoch": 1.4915696158525986, |
| "grad_norm": 2.2995235689237377, |
| "learning_rate": 3.6854365836719726e-05, |
| "loss": 0.3397, |
| "step": 4290 |
| }, |
| { |
| "epoch": 1.4922649052668173, |
| "grad_norm": 2.5645850593306854, |
| "learning_rate": 3.676028460805909e-05, |
| "loss": 0.2239, |
| "step": 4292 |
| }, |
| { |
| "epoch": 1.492960194681036, |
| "grad_norm": 2.1093766655679187, |
| "learning_rate": 3.666629656762679e-05, |
| "loss": 0.3691, |
| "step": 4294 |
| }, |
| { |
| "epoch": 1.4936554840952547, |
| "grad_norm": 2.1635076597384906, |
| "learning_rate": 3.657240185392076e-05, |
| "loss": 0.2085, |
| "step": 4296 |
| }, |
| { |
| "epoch": 1.4943507735094732, |
| "grad_norm": 3.572541006897479, |
| "learning_rate": 3.647860060530165e-05, |
| "loss": 0.2786, |
| "step": 4298 |
| }, |
| { |
| "epoch": 1.495046062923692, |
| "grad_norm": 1.1620393877776625, |
| "learning_rate": 3.6384892959992234e-05, |
| "loss": 0.1839, |
| "step": 4300 |
| }, |
| { |
| "epoch": 1.4957413523379106, |
| "grad_norm": 2.020166804194345, |
| "learning_rate": 3.6291279056077456e-05, |
| "loss": 0.1541, |
| "step": 4302 |
| }, |
| { |
| "epoch": 1.4964366417521293, |
| "grad_norm": 2.7755037338492934, |
| "learning_rate": 3.619775903150405e-05, |
| "loss": 0.2605, |
| "step": 4304 |
| }, |
| { |
| "epoch": 1.497131931166348, |
| "grad_norm": 1.301340684687448, |
| "learning_rate": 3.610433302408045e-05, |
| "loss": 0.1853, |
| "step": 4306 |
| }, |
| { |
| "epoch": 1.4978272205805667, |
| "grad_norm": 1.6835061945215053, |
| "learning_rate": 3.6011001171476474e-05, |
| "loss": 0.1665, |
| "step": 4308 |
| }, |
| { |
| "epoch": 1.4985225099947854, |
| "grad_norm": 1.0520177794795824, |
| "learning_rate": 3.591776361122337e-05, |
| "loss": 0.1487, |
| "step": 4310 |
| }, |
| { |
| "epoch": 1.4992177994090041, |
| "grad_norm": 0.859635070447827, |
| "learning_rate": 3.582462048071326e-05, |
| "loss": 0.1549, |
| "step": 4312 |
| }, |
| { |
| "epoch": 1.4999130888232226, |
| "grad_norm": 2.350772264108287, |
| "learning_rate": 3.573157191719915e-05, |
| "loss": 0.2473, |
| "step": 4314 |
| }, |
| { |
| "epoch": 1.5006083782374413, |
| "grad_norm": 3.2745568715369697, |
| "learning_rate": 3.563861805779482e-05, |
| "loss": 0.392, |
| "step": 4316 |
| }, |
| { |
| "epoch": 1.50130366765166, |
| "grad_norm": 3.03765330728192, |
| "learning_rate": 3.554575903947437e-05, |
| "loss": 0.1755, |
| "step": 4318 |
| }, |
| { |
| "epoch": 1.5019989570658787, |
| "grad_norm": 1.7919095698334324, |
| "learning_rate": 3.545299499907217e-05, |
| "loss": 0.3548, |
| "step": 4320 |
| }, |
| { |
| "epoch": 1.5026942464800972, |
| "grad_norm": 2.392033860296511, |
| "learning_rate": 3.536032607328261e-05, |
| "loss": 0.3514, |
| "step": 4322 |
| }, |
| { |
| "epoch": 1.503389535894316, |
| "grad_norm": 2.4590809095017097, |
| "learning_rate": 3.5267752398660045e-05, |
| "loss": 0.2609, |
| "step": 4324 |
| }, |
| { |
| "epoch": 1.5040848253085346, |
| "grad_norm": 1.997090793496329, |
| "learning_rate": 3.5175274111618306e-05, |
| "loss": 0.1827, |
| "step": 4326 |
| }, |
| { |
| "epoch": 1.5047801147227533, |
| "grad_norm": 2.2876339316317074, |
| "learning_rate": 3.5082891348430836e-05, |
| "loss": 0.2942, |
| "step": 4328 |
| }, |
| { |
| "epoch": 1.505475404136972, |
| "grad_norm": 2.5148757838486815, |
| "learning_rate": 3.499060424523009e-05, |
| "loss": 0.2426, |
| "step": 4330 |
| }, |
| { |
| "epoch": 1.5061706935511907, |
| "grad_norm": 1.5705902747689429, |
| "learning_rate": 3.489841293800779e-05, |
| "loss": 0.1518, |
| "step": 4332 |
| }, |
| { |
| "epoch": 1.5068659829654094, |
| "grad_norm": 3.3116294975461025, |
| "learning_rate": 3.480631756261438e-05, |
| "loss": 0.1892, |
| "step": 4334 |
| }, |
| { |
| "epoch": 1.5075612723796281, |
| "grad_norm": 1.077569115054968, |
| "learning_rate": 3.471431825475892e-05, |
| "loss": 0.1374, |
| "step": 4336 |
| }, |
| { |
| "epoch": 1.5082565617938468, |
| "grad_norm": 2.890287299390136, |
| "learning_rate": 3.462241515000899e-05, |
| "loss": 0.3915, |
| "step": 4338 |
| }, |
| { |
| "epoch": 1.5089518512080655, |
| "grad_norm": 2.1584605162140065, |
| "learning_rate": 3.4530608383790364e-05, |
| "loss": 0.2108, |
| "step": 4340 |
| }, |
| { |
| "epoch": 1.509647140622284, |
| "grad_norm": 2.89717751286833, |
| "learning_rate": 3.443889809138684e-05, |
| "loss": 0.2033, |
| "step": 4342 |
| }, |
| { |
| "epoch": 1.5103424300365027, |
| "grad_norm": 3.0897816198959442, |
| "learning_rate": 3.434728440794005e-05, |
| "loss": 0.256, |
| "step": 4344 |
| }, |
| { |
| "epoch": 1.5110377194507214, |
| "grad_norm": 2.287069750886783, |
| "learning_rate": 3.4255767468449343e-05, |
| "loss": 0.2385, |
| "step": 4346 |
| }, |
| { |
| "epoch": 1.5117330088649399, |
| "grad_norm": 2.2454468817458473, |
| "learning_rate": 3.416434740777138e-05, |
| "loss": 0.3107, |
| "step": 4348 |
| }, |
| { |
| "epoch": 1.5124282982791586, |
| "grad_norm": 2.776731800973786, |
| "learning_rate": 3.407302436062022e-05, |
| "loss": 0.2612, |
| "step": 4350 |
| }, |
| { |
| "epoch": 1.5124282982791586, |
| "eval_loss": 0.8387078642845154, |
| "eval_runtime": 370.5471, |
| "eval_samples_per_second": 13.067, |
| "eval_steps_per_second": 0.41, |
| "step": 4350 |
| }, |
| { |
| "epoch": 1.5131235876933773, |
| "grad_norm": 1.7010850031949087, |
| "learning_rate": 3.3981798461566835e-05, |
| "loss": 0.1537, |
| "step": 4352 |
| }, |
| { |
| "epoch": 1.513818877107596, |
| "grad_norm": 2.070406299336282, |
| "learning_rate": 3.389066984503909e-05, |
| "loss": 0.1882, |
| "step": 4354 |
| }, |
| { |
| "epoch": 1.5145141665218147, |
| "grad_norm": 2.671863349154566, |
| "learning_rate": 3.379963864532149e-05, |
| "loss": 0.3378, |
| "step": 4356 |
| }, |
| { |
| "epoch": 1.5152094559360334, |
| "grad_norm": 2.914185671524914, |
| "learning_rate": 3.370870499655502e-05, |
| "loss": 0.2734, |
| "step": 4358 |
| }, |
| { |
| "epoch": 1.515904745350252, |
| "grad_norm": 1.876554767248312, |
| "learning_rate": 3.361786903273689e-05, |
| "loss": 0.1835, |
| "step": 4360 |
| }, |
| { |
| "epoch": 1.5166000347644708, |
| "grad_norm": 0.8446347241252327, |
| "learning_rate": 3.352713088772036e-05, |
| "loss": 0.2531, |
| "step": 4362 |
| }, |
| { |
| "epoch": 1.5172953241786895, |
| "grad_norm": 2.0961074583438375, |
| "learning_rate": 3.343649069521455e-05, |
| "loss": 0.2916, |
| "step": 4364 |
| }, |
| { |
| "epoch": 1.5179906135929082, |
| "grad_norm": 1.8723369925355657, |
| "learning_rate": 3.3345948588784216e-05, |
| "loss": 0.209, |
| "step": 4366 |
| }, |
| { |
| "epoch": 1.5186859030071267, |
| "grad_norm": 1.8819556854019914, |
| "learning_rate": 3.325550470184966e-05, |
| "loss": 0.3215, |
| "step": 4368 |
| }, |
| { |
| "epoch": 1.5193811924213454, |
| "grad_norm": 2.0481068202906605, |
| "learning_rate": 3.3165159167686354e-05, |
| "loss": 0.2371, |
| "step": 4370 |
| }, |
| { |
| "epoch": 1.520076481835564, |
| "grad_norm": 1.7902777278997732, |
| "learning_rate": 3.307491211942493e-05, |
| "loss": 0.2751, |
| "step": 4372 |
| }, |
| { |
| "epoch": 1.5207717712497826, |
| "grad_norm": 1.4619929160492573, |
| "learning_rate": 3.2984763690050815e-05, |
| "loss": 0.19, |
| "step": 4374 |
| }, |
| { |
| "epoch": 1.5214670606640013, |
| "grad_norm": 1.3298285971628243, |
| "learning_rate": 3.289471401240415e-05, |
| "loss": 0.2477, |
| "step": 4376 |
| }, |
| { |
| "epoch": 1.52216235007822, |
| "grad_norm": 1.4560985599689389, |
| "learning_rate": 3.2804763219179526e-05, |
| "loss": 0.2034, |
| "step": 4378 |
| }, |
| { |
| "epoch": 1.5228576394924387, |
| "grad_norm": 1.311372444978652, |
| "learning_rate": 3.27149114429259e-05, |
| "loss": 0.2017, |
| "step": 4380 |
| }, |
| { |
| "epoch": 1.5235529289066574, |
| "grad_norm": 1.9290129350762313, |
| "learning_rate": 3.2625158816046245e-05, |
| "loss": 0.21, |
| "step": 4382 |
| }, |
| { |
| "epoch": 1.524248218320876, |
| "grad_norm": 1.5785080364435637, |
| "learning_rate": 3.253550547079741e-05, |
| "loss": 0.2844, |
| "step": 4384 |
| }, |
| { |
| "epoch": 1.5249435077350948, |
| "grad_norm": 2.785195126740178, |
| "learning_rate": 3.2445951539290096e-05, |
| "loss": 0.3006, |
| "step": 4386 |
| }, |
| { |
| "epoch": 1.5256387971493135, |
| "grad_norm": 2.719261081201672, |
| "learning_rate": 3.2356497153488285e-05, |
| "loss": 0.2907, |
| "step": 4388 |
| }, |
| { |
| "epoch": 1.5263340865635322, |
| "grad_norm": 2.553021678906123, |
| "learning_rate": 3.226714244520948e-05, |
| "loss": 0.3081, |
| "step": 4390 |
| }, |
| { |
| "epoch": 1.527029375977751, |
| "grad_norm": 1.3401671242636175, |
| "learning_rate": 3.217788754612414e-05, |
| "loss": 0.1174, |
| "step": 4392 |
| }, |
| { |
| "epoch": 1.5277246653919694, |
| "grad_norm": 2.7326813775310197, |
| "learning_rate": 3.20887325877558e-05, |
| "loss": 0.2397, |
| "step": 4394 |
| }, |
| { |
| "epoch": 1.528419954806188, |
| "grad_norm": 2.688550056559767, |
| "learning_rate": 3.199967770148059e-05, |
| "loss": 0.2804, |
| "step": 4396 |
| }, |
| { |
| "epoch": 1.5291152442204068, |
| "grad_norm": 2.159323400603022, |
| "learning_rate": 3.1910723018527256e-05, |
| "loss": 0.256, |
| "step": 4398 |
| }, |
| { |
| "epoch": 1.5298105336346253, |
| "grad_norm": 1.7034505082914522, |
| "learning_rate": 3.182186866997683e-05, |
| "loss": 0.2433, |
| "step": 4400 |
| }, |
| { |
| "epoch": 1.530505823048844, |
| "grad_norm": 2.1045193060188083, |
| "learning_rate": 3.1733114786762577e-05, |
| "loss": 0.1802, |
| "step": 4402 |
| }, |
| { |
| "epoch": 1.5312011124630627, |
| "grad_norm": 1.1348125292460949, |
| "learning_rate": 3.1644461499669606e-05, |
| "loss": 0.2146, |
| "step": 4404 |
| }, |
| { |
| "epoch": 1.5318964018772814, |
| "grad_norm": 2.930434624399605, |
| "learning_rate": 3.155590893933494e-05, |
| "loss": 0.2529, |
| "step": 4406 |
| }, |
| { |
| "epoch": 1.5325916912915, |
| "grad_norm": 2.187552627183412, |
| "learning_rate": 3.146745723624704e-05, |
| "loss": 0.3088, |
| "step": 4408 |
| }, |
| { |
| "epoch": 1.5332869807057188, |
| "grad_norm": 1.6924644719448223, |
| "learning_rate": 3.13791065207458e-05, |
| "loss": 0.2423, |
| "step": 4410 |
| } |
| ], |
| "logging_steps": 2, |
| "max_steps": 5752, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 2, |
| "save_steps": 90, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 1.6039462076678144e+16, |
| "train_batch_size": 4, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
