| { | |
| "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": 31525.89453125, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 108.938, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.022121446742616967, | |
| "grad_norm": 276.1221008300781, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 71.4733, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.03318217011392545, | |
| "grad_norm": 3096.194580078125, | |
| "learning_rate": 3e-06, | |
| "loss": 119.8372, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.044242893485233935, | |
| "grad_norm": 715.627197265625, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 131.9085, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.05530361685654242, | |
| "grad_norm": 8888.615234375, | |
| "learning_rate": 5e-06, | |
| "loss": 120.4494, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.0663643402278509, | |
| "grad_norm": 4452.72119140625, | |
| "learning_rate": 6e-06, | |
| "loss": 100.8797, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.07742506359915939, | |
| "grad_norm": 66307.34375, | |
| "learning_rate": 7e-06, | |
| "loss": 145.084, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.08848578697046787, | |
| "grad_norm": 602.1884765625, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 68.911, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.09954651034177635, | |
| "grad_norm": 1317.8251953125, | |
| "learning_rate": 9e-06, | |
| "loss": 150.1405, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.11060723371308484, | |
| "grad_norm": 716.1102905273438, | |
| "learning_rate": 1e-05, | |
| "loss": 54.1946, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.12166795708439332, | |
| "grad_norm": 1669.779052734375, | |
| "learning_rate": 9.88888888888889e-06, | |
| "loss": 119.7225, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.1327286804557018, | |
| "grad_norm": 1085.806884765625, | |
| "learning_rate": 9.777777777777779e-06, | |
| "loss": 104.1395, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.14378940382701028, | |
| "grad_norm": 317.991943359375, | |
| "learning_rate": 9.666666666666667e-06, | |
| "loss": 167.9486, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.15485012719831878, | |
| "grad_norm": 225.4032745361328, | |
| "learning_rate": 9.555555555555556e-06, | |
| "loss": 65.5434, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.16591085056962726, | |
| "grad_norm": 443.49407958984375, | |
| "learning_rate": 9.444444444444445e-06, | |
| "loss": 68.5829, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.17697157394093574, | |
| "grad_norm": 492.5722961425781, | |
| "learning_rate": 9.333333333333334e-06, | |
| "loss": 97.6923, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.18803229731224422, | |
| "grad_norm": 1269.8448486328125, | |
| "learning_rate": 9.222222222222224e-06, | |
| "loss": 95.6809, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.1990930206835527, | |
| "grad_norm": 271.0899963378906, | |
| "learning_rate": 9.111111111111112e-06, | |
| "loss": 68.7351, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.21015374405486117, | |
| "grad_norm": 300.30364990234375, | |
| "learning_rate": 9e-06, | |
| "loss": 113.9254, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.22121446742616968, | |
| "grad_norm": 2832.990234375, | |
| "learning_rate": 8.888888888888888e-06, | |
| "loss": 176.4239, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.23227519079747816, | |
| "grad_norm": 1707.886962890625, | |
| "learning_rate": 8.777777777777778e-06, | |
| "loss": 121.4123, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.24333591416878664, | |
| "grad_norm": 432.3907470703125, | |
| "learning_rate": 8.666666666666668e-06, | |
| "loss": 90.216, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.2543966375400951, | |
| "grad_norm": 759.7860717773438, | |
| "learning_rate": 8.555555555555556e-06, | |
| "loss": 80.3294, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.2654573609114036, | |
| "grad_norm": 1113.482666015625, | |
| "learning_rate": 8.444444444444446e-06, | |
| "loss": 100.8702, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.2765180842827121, | |
| "grad_norm": 242.53280639648438, | |
| "learning_rate": 8.333333333333334e-06, | |
| "loss": 200.5789, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.28757880765402055, | |
| "grad_norm": 14315.33203125, | |
| "learning_rate": 8.222222222222222e-06, | |
| "loss": 131.5917, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.29863953102532903, | |
| "grad_norm": 1412.022705078125, | |
| "learning_rate": 8.111111111111112e-06, | |
| "loss": 114.6078, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.30970025439663756, | |
| "grad_norm": 481.417724609375, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 85.9446, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.32076097776794604, | |
| "grad_norm": 1741.899169921875, | |
| "learning_rate": 7.88888888888889e-06, | |
| "loss": 76.5052, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.3318217011392545, | |
| "grad_norm": 2144.94189453125, | |
| "learning_rate": 7.77777777777778e-06, | |
| "loss": 84.6808, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.342882424510563, | |
| "grad_norm": 641.1427001953125, | |
| "learning_rate": 7.666666666666667e-06, | |
| "loss": 85.521, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.3539431478818715, | |
| "grad_norm": 305.78216552734375, | |
| "learning_rate": 7.555555555555556e-06, | |
| "loss": 64.7245, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.36500387125317996, | |
| "grad_norm": 591.6135864257812, | |
| "learning_rate": 7.444444444444445e-06, | |
| "loss": 77.5208, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.37606459462448844, | |
| "grad_norm": 391.4192810058594, | |
| "learning_rate": 7.333333333333333e-06, | |
| "loss": 62.7838, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.3871253179957969, | |
| "grad_norm": 534.8387451171875, | |
| "learning_rate": 7.222222222222223e-06, | |
| "loss": 75.2527, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.3981860413671054, | |
| "grad_norm": 202.5685577392578, | |
| "learning_rate": 7.111111111111112e-06, | |
| "loss": 89.4137, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.40924676473841387, | |
| "grad_norm": 714.9501953125, | |
| "learning_rate": 7e-06, | |
| "loss": 71.5902, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.42030748810972235, | |
| "grad_norm": 127.53585052490234, | |
| "learning_rate": 6.88888888888889e-06, | |
| "loss": 121.4861, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.4313682114810309, | |
| "grad_norm": 195.28065490722656, | |
| "learning_rate": 6.777777777777779e-06, | |
| "loss": 103.2692, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.44242893485233936, | |
| "grad_norm": 452.83935546875, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 173.6033, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.45348965822364784, | |
| "grad_norm": 2889.515869140625, | |
| "learning_rate": 6.555555555555556e-06, | |
| "loss": 52.3384, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.4645503815949563, | |
| "grad_norm": 3502.135986328125, | |
| "learning_rate": 6.444444444444445e-06, | |
| "loss": 89.3744, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.4756111049662648, | |
| "grad_norm": 224.1280517578125, | |
| "learning_rate": 6.333333333333333e-06, | |
| "loss": 84.7305, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.4866718283375733, | |
| "grad_norm": 78.53791046142578, | |
| "learning_rate": 6.222222222222223e-06, | |
| "loss": 115.2134, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.49773255170888175, | |
| "grad_norm": 146.20570373535156, | |
| "learning_rate": 6.111111111111112e-06, | |
| "loss": 84.5645, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.5087932750801902, | |
| "grad_norm": 535.9326171875, | |
| "learning_rate": 6e-06, | |
| "loss": 95.7671, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.5198539984514987, | |
| "grad_norm": 356.42462158203125, | |
| "learning_rate": 5.88888888888889e-06, | |
| "loss": 78.2208, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.5309147218228072, | |
| "grad_norm": 744.9382934570312, | |
| "learning_rate": 5.777777777777778e-06, | |
| "loss": 67.7703, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.5419754451941157, | |
| "grad_norm": 585.5819091796875, | |
| "learning_rate": 5.666666666666667e-06, | |
| "loss": 115.4916, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.5530361685654241, | |
| "grad_norm": 955.3712768554688, | |
| "learning_rate": 5.555555555555557e-06, | |
| "loss": 46.3012, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.5640968919367326, | |
| "grad_norm": 675.9095458984375, | |
| "learning_rate": 5.444444444444445e-06, | |
| "loss": 191.4264, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.5751576153080411, | |
| "grad_norm": 525.0625, | |
| "learning_rate": 5.333333333333334e-06, | |
| "loss": 73.2462, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.5862183386793496, | |
| "grad_norm": 649.7520141601562, | |
| "learning_rate": 5.2222222222222226e-06, | |
| "loss": 57.1971, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.5972790620506581, | |
| "grad_norm": 167.98040771484375, | |
| "learning_rate": 5.1111111111111115e-06, | |
| "loss": 128.9888, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.6083397854219667, | |
| "grad_norm": 450.4929504394531, | |
| "learning_rate": 5e-06, | |
| "loss": 97.4584, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.6194005087932751, | |
| "grad_norm": 994.5988159179688, | |
| "learning_rate": 4.888888888888889e-06, | |
| "loss": 148.1406, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 0.6304612321645836, | |
| "grad_norm": 977.2152099609375, | |
| "learning_rate": 4.777777777777778e-06, | |
| "loss": 59.5398, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.6415219555358921, | |
| "grad_norm": 692.0723876953125, | |
| "learning_rate": 4.666666666666667e-06, | |
| "loss": 151.1167, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 0.6525826789072006, | |
| "grad_norm": 906.3406982421875, | |
| "learning_rate": 4.555555555555556e-06, | |
| "loss": 80.2065, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 0.663643402278509, | |
| "grad_norm": 654.7086181640625, | |
| "learning_rate": 4.444444444444444e-06, | |
| "loss": 125.2885, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.6747041256498175, | |
| "grad_norm": 470.31787109375, | |
| "learning_rate": 4.333333333333334e-06, | |
| "loss": 48.1181, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 0.685764849021126, | |
| "grad_norm": 462.6188049316406, | |
| "learning_rate": 4.222222222222223e-06, | |
| "loss": 72.0648, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 0.6968255723924345, | |
| "grad_norm": 669.2234497070312, | |
| "learning_rate": 4.111111111111111e-06, | |
| "loss": 92.0509, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.707886295763743, | |
| "grad_norm": 428.6507263183594, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 64.739, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 0.7189470191350514, | |
| "grad_norm": 893.17138671875, | |
| "learning_rate": 3.88888888888889e-06, | |
| "loss": 79.4529, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.7300077425063599, | |
| "grad_norm": 1005.3417358398438, | |
| "learning_rate": 3.777777777777778e-06, | |
| "loss": 45.9996, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.7410684658776684, | |
| "grad_norm": 1213.5457763671875, | |
| "learning_rate": 3.6666666666666666e-06, | |
| "loss": 84.1884, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 0.7521291892489769, | |
| "grad_norm": 706.6162109375, | |
| "learning_rate": 3.555555555555556e-06, | |
| "loss": 80.2486, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 0.7631899126202853, | |
| "grad_norm": 311.6005859375, | |
| "learning_rate": 3.444444444444445e-06, | |
| "loss": 86.258, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.7742506359915938, | |
| "grad_norm": 1002.1148071289062, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 48.6228, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.7853113593629023, | |
| "grad_norm": 238.88232421875, | |
| "learning_rate": 3.2222222222222227e-06, | |
| "loss": 58.1885, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 0.7963720827342108, | |
| "grad_norm": 332.437255859375, | |
| "learning_rate": 3.1111111111111116e-06, | |
| "loss": 53.2679, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.8074328061055193, | |
| "grad_norm": 365.4303283691406, | |
| "learning_rate": 3e-06, | |
| "loss": 68.5481, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 0.8184935294768277, | |
| "grad_norm": 2730.581787109375, | |
| "learning_rate": 2.888888888888889e-06, | |
| "loss": 87.514, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 0.8295542528481362, | |
| "grad_norm": 276.9474792480469, | |
| "learning_rate": 2.7777777777777783e-06, | |
| "loss": 50.4965, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.8406149762194447, | |
| "grad_norm": 2384.289306640625, | |
| "learning_rate": 2.666666666666667e-06, | |
| "loss": 93.6163, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 0.8516756995907533, | |
| "grad_norm": 2390.2421875, | |
| "learning_rate": 2.5555555555555557e-06, | |
| "loss": 56.3228, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 0.8627364229620618, | |
| "grad_norm": 818.0576782226562, | |
| "learning_rate": 2.4444444444444447e-06, | |
| "loss": 68.8869, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 0.8737971463333702, | |
| "grad_norm": 792.1302490234375, | |
| "learning_rate": 2.3333333333333336e-06, | |
| "loss": 67.4239, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 0.8848578697046787, | |
| "grad_norm": 995.8483276367188, | |
| "learning_rate": 2.222222222222222e-06, | |
| "loss": 63.2453, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.8959185930759872, | |
| "grad_norm": 788.3203125, | |
| "learning_rate": 2.1111111111111114e-06, | |
| "loss": 131.5255, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 0.9069793164472957, | |
| "grad_norm": 708.2405395507812, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 57.5729, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 0.9180400398186042, | |
| "grad_norm": 1030.069580078125, | |
| "learning_rate": 1.888888888888889e-06, | |
| "loss": 100.5658, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 0.9291007631899126, | |
| "grad_norm": 927.4437866210938, | |
| "learning_rate": 1.777777777777778e-06, | |
| "loss": 73.2141, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 0.9401614865612211, | |
| "grad_norm": 204.16241455078125, | |
| "learning_rate": 1.6666666666666667e-06, | |
| "loss": 59.6098, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.9512222099325296, | |
| "grad_norm": 242.94900512695312, | |
| "learning_rate": 1.5555555555555558e-06, | |
| "loss": 84.2695, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 0.9622829333038381, | |
| "grad_norm": 318.7410888671875, | |
| "learning_rate": 1.4444444444444445e-06, | |
| "loss": 43.2253, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 0.9733436566751466, | |
| "grad_norm": 791.7606811523438, | |
| "learning_rate": 1.3333333333333334e-06, | |
| "loss": 138.4619, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 0.984404380046455, | |
| "grad_norm": 389.98565673828125, | |
| "learning_rate": 1.2222222222222223e-06, | |
| "loss": 83.4004, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 0.9954651034177635, | |
| "grad_norm": 158.41607666015625, | |
| "learning_rate": 1.111111111111111e-06, | |
| "loss": 84.5722, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 78.85076141357422, | |
| "eval_runtime": 53.3119, | |
| "eval_samples_per_second": 37.703, | |
| "eval_steps_per_second": 18.851, | |
| "step": 9041 | |
| }, | |
| { | |
| "epoch": 1.006525826789072, | |
| "grad_norm": 783.9674072265625, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 65.1962, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 1.0175865501603805, | |
| "grad_norm": 1137.7186279296875, | |
| "learning_rate": 8.88888888888889e-07, | |
| "loss": 58.0089, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 1.028647273531689, | |
| "grad_norm": 88.58700561523438, | |
| "learning_rate": 7.777777777777779e-07, | |
| "loss": 49.665, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 1.0397079969029974, | |
| "grad_norm": 218.33413696289062, | |
| "learning_rate": 6.666666666666667e-07, | |
| "loss": 64.1321, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 1.050768720274306, | |
| "grad_norm": 222.10977172851562, | |
| "learning_rate": 5.555555555555555e-07, | |
| "loss": 117.5688, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.0618294436456144, | |
| "grad_norm": 1043.607177734375, | |
| "learning_rate": 4.444444444444445e-07, | |
| "loss": 72.3682, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 1.0728901670169229, | |
| "grad_norm": 369.0968017578125, | |
| "learning_rate": 3.3333333333333335e-07, | |
| "loss": 128.3121, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 1.0839508903882313, | |
| "grad_norm": 939.8422241210938, | |
| "learning_rate": 2.2222222222222224e-07, | |
| "loss": 112.3189, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 1.0950116137595398, | |
| "grad_norm": 643.8301391601562, | |
| "learning_rate": 1.1111111111111112e-07, | |
| "loss": 55.195, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 1.1060723371308483, | |
| "grad_norm": 335.5890808105469, | |
| "learning_rate": 0.0, | |
| "loss": 146.6059, | |
| "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 | |
| } | |