| { |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.4067443073179211, |
| "eval_steps": 150, |
| "global_step": 1170, |
| "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 |
| } |
| ], |
| "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": 4661946054868992.0, |
| "train_batch_size": 4, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
