| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 87, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.034482758620689655, | |
| "grad_norm": 1.8968472863706825, | |
| "learning_rate": 1.111111111111111e-06, | |
| "loss": 0.435, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.06896551724137931, | |
| "grad_norm": 1.3787873388390994, | |
| "learning_rate": 2.222222222222222e-06, | |
| "loss": 0.3556, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.10344827586206896, | |
| "grad_norm": 1.500745719809916, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 0.3895, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.13793103448275862, | |
| "grad_norm": 1.3758212628963946, | |
| "learning_rate": 4.444444444444444e-06, | |
| "loss": 0.3798, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.1724137931034483, | |
| "grad_norm": 1.0725134036871664, | |
| "learning_rate": 5.555555555555557e-06, | |
| "loss": 0.3282, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.20689655172413793, | |
| "grad_norm": 0.7977790045197081, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.2932, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.2413793103448276, | |
| "grad_norm": 0.8837691300884692, | |
| "learning_rate": 7.77777777777778e-06, | |
| "loss": 0.3141, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.27586206896551724, | |
| "grad_norm": 0.705940070600187, | |
| "learning_rate": 8.888888888888888e-06, | |
| "loss": 0.2514, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.3103448275862069, | |
| "grad_norm": 0.5506871024791313, | |
| "learning_rate": 1e-05, | |
| "loss": 0.2285, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.3448275862068966, | |
| "grad_norm": 0.9069273874218927, | |
| "learning_rate": 9.995944990857848e-06, | |
| "loss": 0.3054, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3793103448275862, | |
| "grad_norm": 0.7378234840973698, | |
| "learning_rate": 9.983786540671052e-06, | |
| "loss": 0.2962, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.41379310344827586, | |
| "grad_norm": 0.6054420618930146, | |
| "learning_rate": 9.96354437049027e-06, | |
| "loss": 0.2805, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.4482758620689655, | |
| "grad_norm": 0.5093901922135685, | |
| "learning_rate": 9.935251313189564e-06, | |
| "loss": 0.2664, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.4827586206896552, | |
| "grad_norm": 0.5788307564961488, | |
| "learning_rate": 9.89895326021134e-06, | |
| "loss": 0.2748, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.5172413793103449, | |
| "grad_norm": 0.4070144278077285, | |
| "learning_rate": 9.854709087130261e-06, | |
| "loss": 0.1965, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.5517241379310345, | |
| "grad_norm": 0.44452387037855084, | |
| "learning_rate": 9.802590558156863e-06, | |
| "loss": 0.2413, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.5862068965517241, | |
| "grad_norm": 0.46435536348228057, | |
| "learning_rate": 9.742682209735727e-06, | |
| "loss": 0.216, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.6206896551724138, | |
| "grad_norm": 0.3764221335514055, | |
| "learning_rate": 9.675081213427076e-06, | |
| "loss": 0.2056, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.6551724137931034, | |
| "grad_norm": 0.4301176989354492, | |
| "learning_rate": 9.599897218294122e-06, | |
| "loss": 0.2318, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.6896551724137931, | |
| "grad_norm": 0.47375751198098554, | |
| "learning_rate": 9.517252173051912e-06, | |
| "loss": 0.2306, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7241379310344828, | |
| "grad_norm": 0.36713668347068407, | |
| "learning_rate": 9.427280128266049e-06, | |
| "loss": 0.1925, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.7586206896551724, | |
| "grad_norm": 0.3605603008015653, | |
| "learning_rate": 9.330127018922195e-06, | |
| "loss": 0.2035, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.7931034482758621, | |
| "grad_norm": 0.4148028274133272, | |
| "learning_rate": 9.225950427718974e-06, | |
| "loss": 0.2447, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.8275862068965517, | |
| "grad_norm": 0.37584064602396255, | |
| "learning_rate": 9.114919329468283e-06, | |
| "loss": 0.1961, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.8620689655172413, | |
| "grad_norm": 0.37162969905754883, | |
| "learning_rate": 8.997213817017508e-06, | |
| "loss": 0.2143, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.896551724137931, | |
| "grad_norm": 0.3768066216380791, | |
| "learning_rate": 8.873024809138272e-06, | |
| "loss": 0.2177, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.9310344827586207, | |
| "grad_norm": 0.3632688725752434, | |
| "learning_rate": 8.742553740855507e-06, | |
| "loss": 0.2078, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.9655172413793104, | |
| "grad_norm": 0.3933393816831956, | |
| "learning_rate": 8.606012236719073e-06, | |
| "loss": 0.2053, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.37560278938859387, | |
| "learning_rate": 8.463621767547998e-06, | |
| "loss": 0.2078, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.0344827586206897, | |
| "grad_norm": 0.31034148392401967, | |
| "learning_rate": 8.315613291203977e-06, | |
| "loss": 0.1503, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.0689655172413792, | |
| "grad_norm": 0.3472551839526283, | |
| "learning_rate": 8.162226877976886e-06, | |
| "loss": 0.1624, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.103448275862069, | |
| "grad_norm": 0.3119751636900065, | |
| "learning_rate": 8.003711321189895e-06, | |
| "loss": 0.1525, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.1379310344827587, | |
| "grad_norm": 0.36473590972460807, | |
| "learning_rate": 7.84032373365578e-06, | |
| "loss": 0.1743, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.1724137931034484, | |
| "grad_norm": 0.3754222460318716, | |
| "learning_rate": 7.672329130639007e-06, | |
| "loss": 0.1839, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.206896551724138, | |
| "grad_norm": 0.3440125880926874, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.1691, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.2413793103448276, | |
| "grad_norm": 0.32158510549152597, | |
| "learning_rate": 7.323615860218844e-06, | |
| "loss": 0.1535, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.2758620689655173, | |
| "grad_norm": 0.3339665959815591, | |
| "learning_rate": 7.143462807015271e-06, | |
| "loss": 0.1556, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.3103448275862069, | |
| "grad_norm": 0.34776270320077474, | |
| "learning_rate": 6.959833049300376e-06, | |
| "loss": 0.1635, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.3448275862068966, | |
| "grad_norm": 0.3664731009524177, | |
| "learning_rate": 6.773024435212678e-06, | |
| "loss": 0.1629, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.3793103448275863, | |
| "grad_norm": 0.35319022882448053, | |
| "learning_rate": 6.583339969007364e-06, | |
| "loss": 0.1548, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.4137931034482758, | |
| "grad_norm": 0.322074992178244, | |
| "learning_rate": 6.391087319582264e-06, | |
| "loss": 0.1492, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.4482758620689655, | |
| "grad_norm": 0.3745066020307014, | |
| "learning_rate": 6.1965783214377895e-06, | |
| "loss": 0.1678, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.4827586206896552, | |
| "grad_norm": 0.3349892745642585, | |
| "learning_rate": 6.000128468880223e-06, | |
| "loss": 0.1352, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.5172413793103448, | |
| "grad_norm": 0.3621146034278814, | |
| "learning_rate": 5.8020564042888015e-06, | |
| "loss": 0.1535, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.5517241379310345, | |
| "grad_norm": 0.3356997566620184, | |
| "learning_rate": 5.6026834012766155e-06, | |
| "loss": 0.1389, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.5862068965517242, | |
| "grad_norm": 0.3393649872853136, | |
| "learning_rate": 5.402332843583631e-06, | |
| "loss": 0.1418, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.6206896551724137, | |
| "grad_norm": 0.3288903512386936, | |
| "learning_rate": 5.201329700547077e-06, | |
| "loss": 0.1318, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.6551724137931034, | |
| "grad_norm": 0.35291058020727184, | |
| "learning_rate": 5e-06, | |
| "loss": 0.139, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.6896551724137931, | |
| "grad_norm": 0.31859428188018823, | |
| "learning_rate": 4.798670299452926e-06, | |
| "loss": 0.132, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.7241379310344827, | |
| "grad_norm": 0.36948885648092306, | |
| "learning_rate": 4.597667156416371e-06, | |
| "loss": 0.1459, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.7586206896551724, | |
| "grad_norm": 0.4140907666360901, | |
| "learning_rate": 4.397316598723385e-06, | |
| "loss": 0.1898, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.793103448275862, | |
| "grad_norm": 0.37454232433271367, | |
| "learning_rate": 4.1979435957111984e-06, | |
| "loss": 0.1519, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.8275862068965516, | |
| "grad_norm": 0.36698948317992974, | |
| "learning_rate": 3.999871531119779e-06, | |
| "loss": 0.1563, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.8620689655172413, | |
| "grad_norm": 0.3987168791517662, | |
| "learning_rate": 3.803421678562213e-06, | |
| "loss": 0.1538, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.896551724137931, | |
| "grad_norm": 0.35639343871652807, | |
| "learning_rate": 3.6089126804177373e-06, | |
| "loss": 0.1312, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.9310344827586206, | |
| "grad_norm": 0.3051946271043259, | |
| "learning_rate": 3.416660030992639e-06, | |
| "loss": 0.1248, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.9655172413793105, | |
| "grad_norm": 0.32352874901380435, | |
| "learning_rate": 3.226975564787322e-06, | |
| "loss": 0.1217, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.3834889635507212, | |
| "learning_rate": 3.040166950699626e-06, | |
| "loss": 0.1601, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.0344827586206895, | |
| "grad_norm": 0.311091599579286, | |
| "learning_rate": 2.8565371929847286e-06, | |
| "loss": 0.1127, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.0689655172413794, | |
| "grad_norm": 0.35138772971857607, | |
| "learning_rate": 2.6763841397811576e-06, | |
| "loss": 0.1301, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.103448275862069, | |
| "grad_norm": 0.3412886489094977, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.1243, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.1379310344827585, | |
| "grad_norm": 0.33728563398551564, | |
| "learning_rate": 2.3276708693609947e-06, | |
| "loss": 0.1144, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.1724137931034484, | |
| "grad_norm": 0.3115082544602725, | |
| "learning_rate": 2.159676266344222e-06, | |
| "loss": 0.1063, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.206896551724138, | |
| "grad_norm": 0.3525243286834508, | |
| "learning_rate": 1.996288678810105e-06, | |
| "loss": 0.1259, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.2413793103448274, | |
| "grad_norm": 0.32371434370720314, | |
| "learning_rate": 1.8377731220231144e-06, | |
| "loss": 0.1019, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.2758620689655173, | |
| "grad_norm": 0.30148326256239505, | |
| "learning_rate": 1.6843867087960252e-06, | |
| "loss": 0.1049, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.310344827586207, | |
| "grad_norm": 0.3311910822062836, | |
| "learning_rate": 1.5363782324520033e-06, | |
| "loss": 0.1106, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.344827586206897, | |
| "grad_norm": 0.3506409615806702, | |
| "learning_rate": 1.3939877632809279e-06, | |
| "loss": 0.1123, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.3793103448275863, | |
| "grad_norm": 0.305444065384095, | |
| "learning_rate": 1.257446259144494e-06, | |
| "loss": 0.102, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.413793103448276, | |
| "grad_norm": 0.2776673923244958, | |
| "learning_rate": 1.1269751908617277e-06, | |
| "loss": 0.0889, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.4482758620689653, | |
| "grad_norm": 0.2853344145059976, | |
| "learning_rate": 1.0027861829824953e-06, | |
| "loss": 0.0905, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.4827586206896552, | |
| "grad_norm": 0.3297908783701314, | |
| "learning_rate": 8.850806705317183e-07, | |
| "loss": 0.1079, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.5172413793103448, | |
| "grad_norm": 0.32560370624682045, | |
| "learning_rate": 7.740495722810271e-07, | |
| "loss": 0.1079, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.5517241379310347, | |
| "grad_norm": 0.35586337791264894, | |
| "learning_rate": 6.698729810778065e-07, | |
| "loss": 0.1037, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.586206896551724, | |
| "grad_norm": 0.3315175811831225, | |
| "learning_rate": 5.727198717339511e-07, | |
| "loss": 0.1075, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.6206896551724137, | |
| "grad_norm": 0.35277939107078204, | |
| "learning_rate": 4.827478269480895e-07, | |
| "loss": 0.1132, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.655172413793103, | |
| "grad_norm": 0.3970709020417667, | |
| "learning_rate": 4.001027817058789e-07, | |
| "loss": 0.1272, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 2.689655172413793, | |
| "grad_norm": 0.31343582687046845, | |
| "learning_rate": 3.2491878657292643e-07, | |
| "loss": 0.1086, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.7241379310344827, | |
| "grad_norm": 0.36817444381442266, | |
| "learning_rate": 2.573177902642726e-07, | |
| "loss": 0.1214, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 2.7586206896551726, | |
| "grad_norm": 0.3201480217673011, | |
| "learning_rate": 1.9740944184313882e-07, | |
| "loss": 0.0993, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.793103448275862, | |
| "grad_norm": 0.3588278131749075, | |
| "learning_rate": 1.4529091286973994e-07, | |
| "loss": 0.1071, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 2.8275862068965516, | |
| "grad_norm": 0.3616467526970753, | |
| "learning_rate": 1.0104673978866164e-07, | |
| "loss": 0.1112, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 2.862068965517241, | |
| "grad_norm": 0.3722334553870228, | |
| "learning_rate": 6.474868681043578e-08, | |
| "loss": 0.126, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 2.896551724137931, | |
| "grad_norm": 0.312011836899645, | |
| "learning_rate": 3.645562950973014e-08, | |
| "loss": 0.0979, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 2.9310344827586206, | |
| "grad_norm": 0.3293502901375024, | |
| "learning_rate": 1.6213459328950355e-08, | |
| "loss": 0.1006, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 2.9655172413793105, | |
| "grad_norm": 0.33847790149852947, | |
| "learning_rate": 4.055009142152066e-09, | |
| "loss": 0.1012, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.3498051357440655, | |
| "learning_rate": 0.0, | |
| "loss": 0.1129, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 87, | |
| "total_flos": 3.952374498459648e+16, | |
| "train_loss": 0.17466499442341685, | |
| "train_runtime": 4967.3247, | |
| "train_samples_per_second": 0.84, | |
| "train_steps_per_second": 0.018 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 87, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.952374498459648e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |