| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 50000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 2.37485408782959, | |
| "learning_rate": 0.0002, | |
| "loss": 0.7725, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.3556101322174072, | |
| "learning_rate": 0.0004, | |
| "loss": 0.73, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 4.192414283752441, | |
| "learning_rate": 0.0006, | |
| "loss": 0.7414, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 3.5939464569091797, | |
| "learning_rate": 0.0008, | |
| "loss": 0.7544, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 15.336837768554688, | |
| "learning_rate": 0.001, | |
| "loss": 0.7723, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 3.893497943878174, | |
| "learning_rate": 0.0009894736842105264, | |
| "loss": 0.7893, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 11.713598251342773, | |
| "learning_rate": 0.0009789473684210528, | |
| "loss": 0.7843, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 13.350496292114258, | |
| "learning_rate": 0.000968421052631579, | |
| "loss": 0.7894, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 2.8563292026519775, | |
| "learning_rate": 0.0009578947368421053, | |
| "loss": 0.7843, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 9.680468559265137, | |
| "learning_rate": 0.0009473684210526315, | |
| "loss": 0.8511, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 4.680222034454346, | |
| "learning_rate": 0.0009368421052631579, | |
| "loss": 0.8686, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 12.037607192993164, | |
| "learning_rate": 0.0009263157894736843, | |
| "loss": 0.8393, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 8.154277801513672, | |
| "learning_rate": 0.0009157894736842105, | |
| "loss": 0.8252, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 3.256535530090332, | |
| "learning_rate": 0.0009052631578947369, | |
| "loss": 0.7854, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 4.168020725250244, | |
| "learning_rate": 0.0008947368421052632, | |
| "loss": 0.8059, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 2.775940418243408, | |
| "learning_rate": 0.0008842105263157894, | |
| "loss": 0.8272, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 14.476820945739746, | |
| "learning_rate": 0.0008736842105263159, | |
| "loss": 0.8006, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 14.232259750366211, | |
| "learning_rate": 0.0008631578947368422, | |
| "loss": 0.7537, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 5.338253021240234, | |
| "learning_rate": 0.0008526315789473684, | |
| "loss": 0.7831, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 5.454962730407715, | |
| "learning_rate": 0.0008421052631578947, | |
| "loss": 0.7633, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 8.250731468200684, | |
| "learning_rate": 0.0008315789473684212, | |
| "loss": 0.7565, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 8.233278274536133, | |
| "learning_rate": 0.0008210526315789474, | |
| "loss": 0.7724, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 10.603631019592285, | |
| "learning_rate": 0.0008105263157894737, | |
| "loss": 0.782, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 3.75233793258667, | |
| "learning_rate": 0.0008, | |
| "loss": 0.7519, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 4.677459239959717, | |
| "learning_rate": 0.0007894736842105263, | |
| "loss": 0.7556, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 4.583637714385986, | |
| "learning_rate": 0.0007789473684210527, | |
| "loss": 0.7459, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 9.343656539916992, | |
| "learning_rate": 0.0007684210526315789, | |
| "loss": 0.7802, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 1.4162571430206299, | |
| "learning_rate": 0.0007578947368421053, | |
| "loss": 0.7384, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 6.586658477783203, | |
| "learning_rate": 0.0007473684210526316, | |
| "loss": 0.7795, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 5.946737289428711, | |
| "learning_rate": 0.0007368421052631579, | |
| "loss": 0.7381, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 5.825309753417969, | |
| "learning_rate": 0.0007263157894736843, | |
| "loss": 0.7482, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 2.1794614791870117, | |
| "learning_rate": 0.0007157894736842105, | |
| "loss": 0.7381, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 7.837281227111816, | |
| "learning_rate": 0.0007052631578947368, | |
| "loss": 0.7314, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 4.612753868103027, | |
| "learning_rate": 0.0006947368421052632, | |
| "loss": 0.7384, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 6.038923263549805, | |
| "learning_rate": 0.0006842105263157895, | |
| "loss": 0.752, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 5.17457389831543, | |
| "learning_rate": 0.0006736842105263158, | |
| "loss": 0.7421, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 4.990045547485352, | |
| "learning_rate": 0.0006631578947368421, | |
| "loss": 0.7253, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 3.1730449199676514, | |
| "learning_rate": 0.0006526315789473684, | |
| "loss": 0.7202, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 1.5063399076461792, | |
| "learning_rate": 0.0006421052631578948, | |
| "loss": 0.7488, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 3.6247060298919678, | |
| "learning_rate": 0.0006315789473684211, | |
| "loss": 0.7263, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 1.4821553230285645, | |
| "learning_rate": 0.0006210526315789474, | |
| "loss": 0.7176, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 2.641312837600708, | |
| "learning_rate": 0.0006105263157894737, | |
| "loss": 0.7182, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "grad_norm": 3.3185555934906006, | |
| "learning_rate": 0.0006, | |
| "loss": 0.7257, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 2.4133715629577637, | |
| "learning_rate": 0.0005894736842105263, | |
| "loss": 0.7202, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "grad_norm": 1.8694628477096558, | |
| "learning_rate": 0.0005789473684210527, | |
| "loss": 0.7172, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 5.248977184295654, | |
| "learning_rate": 0.0005684210526315789, | |
| "loss": 0.7148, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 6.179994106292725, | |
| "learning_rate": 0.0005578947368421052, | |
| "loss": 0.7308, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.9322715997695923, | |
| "learning_rate": 0.0005473684210526317, | |
| "loss": 0.7163, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 1.6366279125213623, | |
| "learning_rate": 0.0005368421052631579, | |
| "loss": 0.7262, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 5.179625034332275, | |
| "learning_rate": 0.0005263157894736842, | |
| "loss": 0.7054, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "grad_norm": 1.7737842798233032, | |
| "learning_rate": 0.0005157894736842106, | |
| "loss": 0.7104, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 1.6575433015823364, | |
| "learning_rate": 0.0005052631578947368, | |
| "loss": 0.7114, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "grad_norm": 3.6518235206604004, | |
| "learning_rate": 0.0004947368421052632, | |
| "loss": 0.7087, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 1.3658417463302612, | |
| "learning_rate": 0.0004842105263157895, | |
| "loss": 0.7106, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 1.312829852104187, | |
| "learning_rate": 0.00047368421052631577, | |
| "loss": 0.7118, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 2.94315767288208, | |
| "learning_rate": 0.00046315789473684214, | |
| "loss": 0.7102, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "grad_norm": 1.349211573600769, | |
| "learning_rate": 0.00045263157894736845, | |
| "loss": 0.7058, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 1.3551281690597534, | |
| "learning_rate": 0.0004421052631578947, | |
| "loss": 0.708, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "grad_norm": 1.3995367288589478, | |
| "learning_rate": 0.0004315789473684211, | |
| "loss": 0.7106, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.9616995453834534, | |
| "learning_rate": 0.00042105263157894734, | |
| "loss": 0.7021, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 3.05, | |
| "grad_norm": 2.390662908554077, | |
| "learning_rate": 0.0004105263157894737, | |
| "loss": 0.7061, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 1.2397950887680054, | |
| "learning_rate": 0.0004, | |
| "loss": 0.705, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 3.15, | |
| "grad_norm": 3.110950231552124, | |
| "learning_rate": 0.00038947368421052633, | |
| "loss": 0.7031, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 1.1811543703079224, | |
| "learning_rate": 0.00037894736842105265, | |
| "loss": 0.7013, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "grad_norm": 1.1221048831939697, | |
| "learning_rate": 0.00036842105263157896, | |
| "loss": 0.7026, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 0.9882568717002869, | |
| "learning_rate": 0.0003578947368421053, | |
| "loss": 0.7012, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 3.35, | |
| "grad_norm": 0.8155710697174072, | |
| "learning_rate": 0.0003473684210526316, | |
| "loss": 0.7033, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 1.1295320987701416, | |
| "learning_rate": 0.0003368421052631579, | |
| "loss": 0.701, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 3.45, | |
| "grad_norm": 2.0776407718658447, | |
| "learning_rate": 0.0003263157894736842, | |
| "loss": 0.6986, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 1.7907339334487915, | |
| "learning_rate": 0.00031578947368421053, | |
| "loss": 0.698, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "grad_norm": 1.978267788887024, | |
| "learning_rate": 0.00030526315789473684, | |
| "loss": 0.7031, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 3.065232515335083, | |
| "learning_rate": 0.00029473684210526316, | |
| "loss": 0.697, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "grad_norm": 0.9516103267669678, | |
| "learning_rate": 0.00028421052631578947, | |
| "loss": 0.7006, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "grad_norm": 1.0045952796936035, | |
| "learning_rate": 0.00027368421052631584, | |
| "loss": 0.6986, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 1.6901030540466309, | |
| "learning_rate": 0.0002631578947368421, | |
| "loss": 0.6995, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 1.8813843727111816, | |
| "learning_rate": 0.0002526315789473684, | |
| "loss": 0.6987, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "grad_norm": 1.8174165487289429, | |
| "learning_rate": 0.00024210526315789475, | |
| "loss": 0.6999, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 1.7451245784759521, | |
| "learning_rate": 0.00023157894736842107, | |
| "loss": 0.6981, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 3.95, | |
| "grad_norm": 0.9422154426574707, | |
| "learning_rate": 0.00022105263157894735, | |
| "loss": 0.6962, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 2.196399688720703, | |
| "learning_rate": 0.00021052631578947367, | |
| "loss": 0.6984, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 4.05, | |
| "grad_norm": 1.940382719039917, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6945, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "grad_norm": 1.2243621349334717, | |
| "learning_rate": 0.00018947368421052632, | |
| "loss": 0.6965, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 4.15, | |
| "grad_norm": 2.004392623901367, | |
| "learning_rate": 0.00017894736842105264, | |
| "loss": 0.6992, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 1.8801628351211548, | |
| "learning_rate": 0.00016842105263157895, | |
| "loss": 0.6944, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 4.25, | |
| "grad_norm": 0.9306873679161072, | |
| "learning_rate": 0.00015789473684210527, | |
| "loss": 0.693, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "grad_norm": 1.9730675220489502, | |
| "learning_rate": 0.00014736842105263158, | |
| "loss": 0.6957, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 4.35, | |
| "grad_norm": 1.0207428932189941, | |
| "learning_rate": 0.00013684210526315792, | |
| "loss": 0.7004, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 0.9524587988853455, | |
| "learning_rate": 0.0001263157894736842, | |
| "loss": 0.6927, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 4.45, | |
| "grad_norm": 0.9829258322715759, | |
| "learning_rate": 0.00011578947368421053, | |
| "loss": 0.6962, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 1.317120909690857, | |
| "learning_rate": 0.00010526315789473683, | |
| "loss": 0.6943, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 4.55, | |
| "grad_norm": 1.8932749032974243, | |
| "learning_rate": 9.473684210526316e-05, | |
| "loss": 0.6933, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "grad_norm": 0.8536391854286194, | |
| "learning_rate": 8.421052631578948e-05, | |
| "loss": 0.693, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 4.65, | |
| "grad_norm": 1.9397382736206055, | |
| "learning_rate": 7.368421052631579e-05, | |
| "loss": 0.6942, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "grad_norm": 1.846257209777832, | |
| "learning_rate": 6.31578947368421e-05, | |
| "loss": 0.6941, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 4.75, | |
| "grad_norm": 0.937384307384491, | |
| "learning_rate": 5.263157894736842e-05, | |
| "loss": 0.6956, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.9019428491592407, | |
| "learning_rate": 4.210526315789474e-05, | |
| "loss": 0.6935, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 4.85, | |
| "grad_norm": 1.892066478729248, | |
| "learning_rate": 3.157894736842105e-05, | |
| "loss": 0.6945, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "grad_norm": 1.764987826347351, | |
| "learning_rate": 2.105263157894737e-05, | |
| "loss": 0.695, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 4.95, | |
| "grad_norm": 1.626028060913086, | |
| "learning_rate": 1.0526315789473684e-05, | |
| "loss": 0.6941, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.6325174570083618, | |
| "learning_rate": 0.0, | |
| "loss": 0.6935, | |
| "step": 50000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 50000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 2877777168000000.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |