| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.1060723371308483, | |
| "eval_steps": 500, | |
| "global_step": 10000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.011060723371308484, | |
| "grad_norm": 56.4532356262207, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 46.4173, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.022121446742616967, | |
| "grad_norm": 1428.7451171875, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 59.5443, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.03318217011392545, | |
| "grad_norm": 93.24847412109375, | |
| "learning_rate": 3e-06, | |
| "loss": 47.2897, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.044242893485233935, | |
| "grad_norm": 205.97618103027344, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 78.6319, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.05530361685654242, | |
| "grad_norm": 553.5042724609375, | |
| "learning_rate": 5e-06, | |
| "loss": 19.4572, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.0663643402278509, | |
| "grad_norm": 135.06678771972656, | |
| "learning_rate": 6e-06, | |
| "loss": 36.1564, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.07742506359915939, | |
| "grad_norm": 856.98779296875, | |
| "learning_rate": 7e-06, | |
| "loss": 63.0676, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.08848578697046787, | |
| "grad_norm": 1598.5091552734375, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 47.5267, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.09954651034177635, | |
| "grad_norm": 247.8220977783203, | |
| "learning_rate": 9e-06, | |
| "loss": 40.4608, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.11060723371308484, | |
| "grad_norm": 1963.9495849609375, | |
| "learning_rate": 1e-05, | |
| "loss": 41.4591, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.12166795708439332, | |
| "grad_norm": 50.89236068725586, | |
| "learning_rate": 9.88888888888889e-06, | |
| "loss": 54.2236, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.1327286804557018, | |
| "grad_norm": 984.9762573242188, | |
| "learning_rate": 9.777777777777779e-06, | |
| "loss": 41.6753, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.14378940382701028, | |
| "grad_norm": 148.56124877929688, | |
| "learning_rate": 9.666666666666667e-06, | |
| "loss": 35.156, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.15485012719831878, | |
| "grad_norm": 714.2054443359375, | |
| "learning_rate": 9.555555555555556e-06, | |
| "loss": 29.9971, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.16591085056962726, | |
| "grad_norm": 563.89111328125, | |
| "learning_rate": 9.444444444444445e-06, | |
| "loss": 29.874, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.17697157394093574, | |
| "grad_norm": 1477.522705078125, | |
| "learning_rate": 9.333333333333334e-06, | |
| "loss": 22.5016, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.18803229731224422, | |
| "grad_norm": 332.4978942871094, | |
| "learning_rate": 9.222222222222224e-06, | |
| "loss": 68.0949, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.1990930206835527, | |
| "grad_norm": 498.601318359375, | |
| "learning_rate": 9.111111111111112e-06, | |
| "loss": 27.2102, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.21015374405486117, | |
| "grad_norm": 34.91910934448242, | |
| "learning_rate": 9e-06, | |
| "loss": 41.4839, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.22121446742616968, | |
| "grad_norm": 298.38714599609375, | |
| "learning_rate": 8.888888888888888e-06, | |
| "loss": 51.3326, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.23227519079747816, | |
| "grad_norm": 117.34264373779297, | |
| "learning_rate": 8.777777777777778e-06, | |
| "loss": 46.8127, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.24333591416878664, | |
| "grad_norm": 55.05023956298828, | |
| "learning_rate": 8.666666666666668e-06, | |
| "loss": 54.3789, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.2543966375400951, | |
| "grad_norm": 317.3476257324219, | |
| "learning_rate": 8.555555555555556e-06, | |
| "loss": 53.2983, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.2654573609114036, | |
| "grad_norm": 12889.2529296875, | |
| "learning_rate": 8.444444444444446e-06, | |
| "loss": 45.7081, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.2765180842827121, | |
| "grad_norm": 305.0364990234375, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 36.1243, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.28757880765402055, | |
| "grad_norm": 381.5853576660156, | |
| "learning_rate": 8.222222222222222e-06, | |
| "loss": 25.3867, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.29863953102532903, | |
| "grad_norm": 79.50729370117188, | |
| "learning_rate": 8.111111111111112e-06, | |
| "loss": 37.5862, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.30970025439663756, | |
| "grad_norm": 620.679931640625, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 55.266, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.32076097776794604, | |
| "grad_norm": 402.6932678222656, | |
| "learning_rate": 7.88888888888889e-06, | |
| "loss": 41.631, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.3318217011392545, | |
| "grad_norm": 192.93832397460938, | |
| "learning_rate": 7.77777777777778e-06, | |
| "loss": 44.7148, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.342882424510563, | |
| "grad_norm": 859.4608154296875, | |
| "learning_rate": 7.666666666666667e-06, | |
| "loss": 17.0516, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.3539431478818715, | |
| "grad_norm": 1979.3770751953125, | |
| "learning_rate": 7.555555555555556e-06, | |
| "loss": 28.492, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.36500387125317996, | |
| "grad_norm": 417.1896057128906, | |
| "learning_rate": 7.444444444444445e-06, | |
| "loss": 44.7208, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.37606459462448844, | |
| "grad_norm": 510.34063720703125, | |
| "learning_rate": 7.333333333333333e-06, | |
| "loss": 48.0172, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.3871253179957969, | |
| "grad_norm": 841.9534912109375, | |
| "learning_rate": 7.222222222222223e-06, | |
| "loss": 36.351, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.3981860413671054, | |
| "grad_norm": 876.4202880859375, | |
| "learning_rate": 7.111111111111112e-06, | |
| "loss": 22.4034, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.40924676473841387, | |
| "grad_norm": 25.35300064086914, | |
| "learning_rate": 7e-06, | |
| "loss": 28.6452, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.42030748810972235, | |
| "grad_norm": 384.4833679199219, | |
| "learning_rate": 6.88888888888889e-06, | |
| "loss": 36.1268, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.4313682114810309, | |
| "grad_norm": 45193.96875, | |
| "learning_rate": 6.777777777777779e-06, | |
| "loss": 46.502, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.44242893485233936, | |
| "grad_norm": 31.551109313964844, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 51.0344, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.45348965822364784, | |
| "grad_norm": 314.2434387207031, | |
| "learning_rate": 6.555555555555556e-06, | |
| "loss": 17.7956, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.4645503815949563, | |
| "grad_norm": 141.0003204345703, | |
| "learning_rate": 6.444444444444445e-06, | |
| "loss": 46.316, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.4756111049662648, | |
| "grad_norm": 228.07168579101562, | |
| "learning_rate": 6.333333333333333e-06, | |
| "loss": 33.4187, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.4866718283375733, | |
| "grad_norm": 111.94170379638672, | |
| "learning_rate": 6.222222222222223e-06, | |
| "loss": 55.4457, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.49773255170888175, | |
| "grad_norm": 623.2265625, | |
| "learning_rate": 6.111111111111112e-06, | |
| "loss": 43.8869, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.5087932750801902, | |
| "grad_norm": 727.6326293945312, | |
| "learning_rate": 6e-06, | |
| "loss": 20.9696, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.5198539984514987, | |
| "grad_norm": 82.59809875488281, | |
| "learning_rate": 5.88888888888889e-06, | |
| "loss": 27.3194, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.5309147218228072, | |
| "grad_norm": 5570.04736328125, | |
| "learning_rate": 5.777777777777778e-06, | |
| "loss": 34.3857, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.5419754451941157, | |
| "grad_norm": 170.74560546875, | |
| "learning_rate": 5.666666666666667e-06, | |
| "loss": 30.1674, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.5530361685654241, | |
| "grad_norm": 143.09498596191406, | |
| "learning_rate": 5.555555555555557e-06, | |
| "loss": 68.6396, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.5640968919367326, | |
| "grad_norm": 573.5573120117188, | |
| "learning_rate": 5.444444444444445e-06, | |
| "loss": 38.1837, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.5751576153080411, | |
| "grad_norm": 420.1831970214844, | |
| "learning_rate": 5.333333333333334e-06, | |
| "loss": 18.1636, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.5862183386793496, | |
| "grad_norm": 372.7723388671875, | |
| "learning_rate": 5.2222222222222226e-06, | |
| "loss": 22.4455, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.5972790620506581, | |
| "grad_norm": 58.206329345703125, | |
| "learning_rate": 5.1111111111111115e-06, | |
| "loss": 41.3742, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.6083397854219667, | |
| "grad_norm": 439.6692199707031, | |
| "learning_rate": 5e-06, | |
| "loss": 43.9376, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.6194005087932751, | |
| "grad_norm": 482.8997497558594, | |
| "learning_rate": 4.888888888888889e-06, | |
| "loss": 24.5571, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.6304612321645836, | |
| "grad_norm": 102.2362289428711, | |
| "learning_rate": 4.777777777777778e-06, | |
| "loss": 30.6074, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.6415219555358921, | |
| "grad_norm": 297.0746765136719, | |
| "learning_rate": 4.666666666666667e-06, | |
| "loss": 27.421, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.6525826789072006, | |
| "grad_norm": 753.6423950195312, | |
| "learning_rate": 4.555555555555556e-06, | |
| "loss": 31.0388, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.663643402278509, | |
| "grad_norm": 12171.5869140625, | |
| "learning_rate": 4.444444444444444e-06, | |
| "loss": 56.092, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.6747041256498175, | |
| "grad_norm": 512.933349609375, | |
| "learning_rate": 4.333333333333334e-06, | |
| "loss": 16.8703, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.685764849021126, | |
| "grad_norm": 1244.9775390625, | |
| "learning_rate": 4.222222222222223e-06, | |
| "loss": 25.927, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.6968255723924345, | |
| "grad_norm": 235.6149139404297, | |
| "learning_rate": 4.111111111111111e-06, | |
| "loss": 33.8786, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.707886295763743, | |
| "grad_norm": 233.86119079589844, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 58.0722, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.7189470191350514, | |
| "grad_norm": 136.34292602539062, | |
| "learning_rate": 3.88888888888889e-06, | |
| "loss": 35.814, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.7300077425063599, | |
| "grad_norm": 32.31587600708008, | |
| "learning_rate": 3.777777777777778e-06, | |
| "loss": 40.7047, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.7410684658776684, | |
| "grad_norm": 22.009552001953125, | |
| "learning_rate": 3.6666666666666666e-06, | |
| "loss": 32.949, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 0.7521291892489769, | |
| "grad_norm": 59.19048309326172, | |
| "learning_rate": 3.555555555555556e-06, | |
| "loss": 35.1464, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.7631899126202853, | |
| "grad_norm": 974.19140625, | |
| "learning_rate": 3.444444444444445e-06, | |
| "loss": 41.7756, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.7742506359915938, | |
| "grad_norm": 228.2376708984375, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 18.9142, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.7853113593629023, | |
| "grad_norm": 366.40478515625, | |
| "learning_rate": 3.2222222222222227e-06, | |
| "loss": 37.3152, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 0.7963720827342108, | |
| "grad_norm": 417.91851806640625, | |
| "learning_rate": 3.1111111111111116e-06, | |
| "loss": 19.5427, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.8074328061055193, | |
| "grad_norm": 45.679840087890625, | |
| "learning_rate": 3e-06, | |
| "loss": 18.8838, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 0.8184935294768277, | |
| "grad_norm": 291.4388732910156, | |
| "learning_rate": 2.888888888888889e-06, | |
| "loss": 28.37, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 0.8295542528481362, | |
| "grad_norm": 779.9126586914062, | |
| "learning_rate": 2.7777777777777783e-06, | |
| "loss": 31.1144, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.8406149762194447, | |
| "grad_norm": 653.708740234375, | |
| "learning_rate": 2.666666666666667e-06, | |
| "loss": 21.9558, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 0.8516756995907533, | |
| "grad_norm": 534.4269409179688, | |
| "learning_rate": 2.5555555555555557e-06, | |
| "loss": 17.3903, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 0.8627364229620618, | |
| "grad_norm": 425.115478515625, | |
| "learning_rate": 2.4444444444444447e-06, | |
| "loss": 16.8976, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 0.8737971463333702, | |
| "grad_norm": 1288.64306640625, | |
| "learning_rate": 2.3333333333333336e-06, | |
| "loss": 23.9973, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 0.8848578697046787, | |
| "grad_norm": 22.14337921142578, | |
| "learning_rate": 2.222222222222222e-06, | |
| "loss": 17.8739, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.8959185930759872, | |
| "grad_norm": 100.16365051269531, | |
| "learning_rate": 2.1111111111111114e-06, | |
| "loss": 43.2343, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 0.9069793164472957, | |
| "grad_norm": 462.9579772949219, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 27.7818, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 0.9180400398186042, | |
| "grad_norm": 67.15435791015625, | |
| "learning_rate": 1.888888888888889e-06, | |
| "loss": 29.433, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 0.9291007631899126, | |
| "grad_norm": 289.3615417480469, | |
| "learning_rate": 1.777777777777778e-06, | |
| "loss": 66.6806, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 0.9401614865612211, | |
| "grad_norm": 730.722412109375, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 19.1097, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.9512222099325296, | |
| "grad_norm": 495.9251708984375, | |
| "learning_rate": 1.5555555555555558e-06, | |
| "loss": 24.1341, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 0.9622829333038381, | |
| "grad_norm": 663.7501220703125, | |
| "learning_rate": 1.4444444444444445e-06, | |
| "loss": 35.4828, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 0.9733436566751466, | |
| "grad_norm": 178.17031860351562, | |
| "learning_rate": 1.3333333333333334e-06, | |
| "loss": 26.9061, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 0.984404380046455, | |
| "grad_norm": 392.52386474609375, | |
| "learning_rate": 1.2222222222222223e-06, | |
| "loss": 34.1691, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 0.9954651034177635, | |
| "grad_norm": 668.2710571289062, | |
| "learning_rate": 1.111111111111111e-06, | |
| "loss": 16.6653, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 76.04228973388672, | |
| "eval_runtime": 85.9957, | |
| "eval_samples_per_second": 23.373, | |
| "eval_steps_per_second": 11.687, | |
| "step": 9041 | |
| }, | |
| { | |
| "epoch": 1.006525826789072, | |
| "grad_norm": 156.24392700195312, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 24.6885, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 1.0175865501603805, | |
| "grad_norm": 851.2970581054688, | |
| "learning_rate": 8.88888888888889e-07, | |
| "loss": 21.6874, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 1.028647273531689, | |
| "grad_norm": 558.8850708007812, | |
| "learning_rate": 7.777777777777779e-07, | |
| "loss": 11.7699, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 1.0397079969029974, | |
| "grad_norm": 235.15342712402344, | |
| "learning_rate": 6.666666666666667e-07, | |
| "loss": 51.2119, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 1.050768720274306, | |
| "grad_norm": 595.2941284179688, | |
| "learning_rate": 5.555555555555555e-07, | |
| "loss": 21.2429, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.0618294436456144, | |
| "grad_norm": 719.56396484375, | |
| "learning_rate": 4.444444444444445e-07, | |
| "loss": 18.2018, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 1.0728901670169229, | |
| "grad_norm": 226.12979125976562, | |
| "learning_rate": 3.3333333333333335e-07, | |
| "loss": 18.8277, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 1.0839508903882313, | |
| "grad_norm": 57.15268325805664, | |
| "learning_rate": 2.2222222222222224e-07, | |
| "loss": 17.7595, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 1.0950116137595398, | |
| "grad_norm": 129.44561767578125, | |
| "learning_rate": 1.1111111111111112e-07, | |
| "loss": 32.7305, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 1.1060723371308483, | |
| "grad_norm": 354.8187255859375, | |
| "learning_rate": 0.0, | |
| "loss": 21.932, | |
| "step": 10000 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 10000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 5000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |