| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.2810248198558847, | |
| "eval_steps": 500, | |
| "global_step": 100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.012810248198558846, | |
| "grad_norm": 14.422553569632921, | |
| "learning_rate": 1.25e-07, | |
| "loss": 0.3727, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.025620496397117692, | |
| "grad_norm": 15.27369825740346, | |
| "learning_rate": 2.5e-07, | |
| "loss": 0.3497, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03843074459567654, | |
| "grad_norm": 14.934103011301683, | |
| "learning_rate": 3.75e-07, | |
| "loss": 0.3837, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.051240992794235385, | |
| "grad_norm": 14.736814811452579, | |
| "learning_rate": 5e-07, | |
| "loss": 0.3396, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.06405124099279423, | |
| "grad_norm": 14.197297787283581, | |
| "learning_rate": 6.249999999999999e-07, | |
| "loss": 0.3499, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.07686148919135308, | |
| "grad_norm": 11.495057131004048, | |
| "learning_rate": 7.5e-07, | |
| "loss": 0.3044, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.08967173738991192, | |
| "grad_norm": 12.10719572952739, | |
| "learning_rate": 8.75e-07, | |
| "loss": 0.3489, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.10248198558847077, | |
| "grad_norm": 17.82877298105584, | |
| "learning_rate": 1e-06, | |
| "loss": 0.3319, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.11529223378702963, | |
| "grad_norm": 20.757666788469532, | |
| "learning_rate": 1.125e-06, | |
| "loss": 0.299, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.12810248198558846, | |
| "grad_norm": 19.06167752186461, | |
| "learning_rate": 1.2499999999999999e-06, | |
| "loss": 0.3083, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.14091273018414732, | |
| "grad_norm": 17.862857075250577, | |
| "learning_rate": 1.375e-06, | |
| "loss": 0.289, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.15372297838270615, | |
| "grad_norm": 24.872907286961055, | |
| "learning_rate": 1.5e-06, | |
| "loss": 0.2736, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.16653322658126501, | |
| "grad_norm": 26.286295562663373, | |
| "learning_rate": 1.625e-06, | |
| "loss": 0.2917, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.17934347477982385, | |
| "grad_norm": 27.853856612497783, | |
| "learning_rate": 1.75e-06, | |
| "loss": 0.3264, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.1921537229783827, | |
| "grad_norm": 29.926337896366878, | |
| "learning_rate": 1.8749999999999998e-06, | |
| "loss": 0.4142, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.20496397117694154, | |
| "grad_norm": 32.72677695738821, | |
| "learning_rate": 2e-06, | |
| "loss": 0.3857, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2177742193755004, | |
| "grad_norm": 25.15556631940085, | |
| "learning_rate": 1.9997482349425066e-06, | |
| "loss": 0.4154, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.23058446757405926, | |
| "grad_norm": 25.6375788037228, | |
| "learning_rate": 1.9989930665413145e-06, | |
| "loss": 0.393, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2433947157726181, | |
| "grad_norm": 25.369663386464367, | |
| "learning_rate": 1.997734875046456e-06, | |
| "loss": 0.3194, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.2562049639711769, | |
| "grad_norm": 22.33431342233663, | |
| "learning_rate": 1.995974293995239e-06, | |
| "loss": 0.3499, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2690152121697358, | |
| "grad_norm": 17.384680330316037, | |
| "learning_rate": 1.9937122098932426e-06, | |
| "loss": 0.3306, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.28182546036829464, | |
| "grad_norm": 19.864128436150512, | |
| "learning_rate": 1.9909497617679347e-06, | |
| "loss": 0.3596, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2946357085668535, | |
| "grad_norm": 20.439398805573077, | |
| "learning_rate": 1.9876883405951377e-06, | |
| "loss": 0.328, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3074459567654123, | |
| "grad_norm": 16.880587852141087, | |
| "learning_rate": 1.9839295885986295e-06, | |
| "loss": 0.2997, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32025620496397117, | |
| "grad_norm": 14.039028426941659, | |
| "learning_rate": 1.9796753984232355e-06, | |
| "loss": 0.2995, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.33306645316253003, | |
| "grad_norm": 16.079146586971746, | |
| "learning_rate": 1.9749279121818236e-06, | |
| "loss": 0.3277, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3458767013610889, | |
| "grad_norm": 16.540864706951407, | |
| "learning_rate": 1.9696895203766866e-06, | |
| "loss": 0.3492, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3586869495596477, | |
| "grad_norm": 12.581391856103888, | |
| "learning_rate": 1.9639628606958534e-06, | |
| "loss": 0.2907, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.37149719775820655, | |
| "grad_norm": 11.140674253319334, | |
| "learning_rate": 1.9577508166849303e-06, | |
| "loss": 0.2977, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.3843074459567654, | |
| "grad_norm": 14.824558664988295, | |
| "learning_rate": 1.9510565162951534e-06, | |
| "loss": 0.3211, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3971176941553243, | |
| "grad_norm": 13.545165184398623, | |
| "learning_rate": 1.9438833303083674e-06, | |
| "loss": 0.3418, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4099279423538831, | |
| "grad_norm": 13.298757689081272, | |
| "learning_rate": 1.936234870639737e-06, | |
| "loss": 0.3298, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.42273819055244194, | |
| "grad_norm": 13.834770648936974, | |
| "learning_rate": 1.928114988519039e-06, | |
| "loss": 0.3406, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4355484387510008, | |
| "grad_norm": 14.126211131938067, | |
| "learning_rate": 1.9195277725514506e-06, | |
| "loss": 0.3268, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.44835868694955966, | |
| "grad_norm": 12.61217701807801, | |
| "learning_rate": 1.9104775466588157e-06, | |
| "loss": 0.3421, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4611689351481185, | |
| "grad_norm": 12.638382733573478, | |
| "learning_rate": 1.9009688679024189e-06, | |
| "loss": 0.3311, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4739791833466773, | |
| "grad_norm": 13.067396281150065, | |
| "learning_rate": 1.8910065241883678e-06, | |
| "loss": 0.3278, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4867894315452362, | |
| "grad_norm": 12.323205100444895, | |
| "learning_rate": 1.8805955318567379e-06, | |
| "loss": 0.3021, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.49959967974379504, | |
| "grad_norm": 15.67799328226583, | |
| "learning_rate": 1.8697411331556953e-06, | |
| "loss": 0.356, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.5124099279423538, | |
| "grad_norm": 11.446755693352566, | |
| "learning_rate": 1.858448793601866e-06, | |
| "loss": 0.2987, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5252201761409128, | |
| "grad_norm": 13.823237028823755, | |
| "learning_rate": 1.8467241992282841e-06, | |
| "loss": 0.3212, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5380304243394716, | |
| "grad_norm": 14.097283611242158, | |
| "learning_rate": 1.8345732537213026e-06, | |
| "loss": 0.2847, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5508406725380304, | |
| "grad_norm": 13.339593293425303, | |
| "learning_rate": 1.82200207544791e-06, | |
| "loss": 0.3023, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5636509207365893, | |
| "grad_norm": 11.710132196735142, | |
| "learning_rate": 1.8090169943749474e-06, | |
| "loss": 0.288, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.5764611689351481, | |
| "grad_norm": 15.421270027221475, | |
| "learning_rate": 1.795624548881781e-06, | |
| "loss": 0.3191, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.589271417133707, | |
| "grad_norm": 15.196185783638896, | |
| "learning_rate": 1.7818314824680298e-06, | |
| "loss": 0.2917, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6020816653322658, | |
| "grad_norm": 14.417769465666634, | |
| "learning_rate": 1.767644740358011e-06, | |
| "loss": 0.3065, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6148919135308246, | |
| "grad_norm": 15.806685481243651, | |
| "learning_rate": 1.753071466003611e-06, | |
| "loss": 0.3133, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6277021617293835, | |
| "grad_norm": 12.804839878435857, | |
| "learning_rate": 1.7381189974873407e-06, | |
| "loss": 0.2924, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.6405124099279423, | |
| "grad_norm": 12.41861829102855, | |
| "learning_rate": 1.7227948638273915e-06, | |
| "loss": 0.2884, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6533226581265013, | |
| "grad_norm": 14.521843299259071, | |
| "learning_rate": 1.7071067811865474e-06, | |
| "loss": 0.3172, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6661329063250601, | |
| "grad_norm": 12.369940546930579, | |
| "learning_rate": 1.6910626489868648e-06, | |
| "loss": 0.3064, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6789431545236189, | |
| "grad_norm": 13.075907929491837, | |
| "learning_rate": 1.6746705459320744e-06, | |
| "loss": 0.3077, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6917534027221778, | |
| "grad_norm": 15.91307811791754, | |
| "learning_rate": 1.6579387259397126e-06, | |
| "loss": 0.3202, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.7045636509207366, | |
| "grad_norm": 16.794702833290078, | |
| "learning_rate": 1.640875613985024e-06, | |
| "loss": 0.3509, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7173738991192954, | |
| "grad_norm": 11.645931539680932, | |
| "learning_rate": 1.6234898018587336e-06, | |
| "loss": 0.2737, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7301841473178543, | |
| "grad_norm": 12.164042324452254, | |
| "learning_rate": 1.6057900438408199e-06, | |
| "loss": 0.2805, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7429943955164131, | |
| "grad_norm": 11.003522632810185, | |
| "learning_rate": 1.587785252292473e-06, | |
| "loss": 0.2935, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.755804643714972, | |
| "grad_norm": 14.747618364345833, | |
| "learning_rate": 1.569484493168452e-06, | |
| "loss": 0.3185, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.7686148919135308, | |
| "grad_norm": 13.515142475366316, | |
| "learning_rate": 1.5508969814521024e-06, | |
| "loss": 0.2973, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7814251401120896, | |
| "grad_norm": 12.718973924439272, | |
| "learning_rate": 1.5320320765153365e-06, | |
| "loss": 0.2956, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7942353883106485, | |
| "grad_norm": 11.338497570766252, | |
| "learning_rate": 1.5128992774059062e-06, | |
| "loss": 0.3127, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8070456365092074, | |
| "grad_norm": 11.44549619969602, | |
| "learning_rate": 1.4935082180643467e-06, | |
| "loss": 0.277, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8198558847077662, | |
| "grad_norm": 12.057954732711101, | |
| "learning_rate": 1.4738686624729987e-06, | |
| "loss": 0.2957, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.8326661329063251, | |
| "grad_norm": 12.801694035766888, | |
| "learning_rate": 1.4539904997395467e-06, | |
| "loss": 0.2907, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8454763811048839, | |
| "grad_norm": 12.990091142294833, | |
| "learning_rate": 1.433883739117558e-06, | |
| "loss": 0.2973, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8582866293034428, | |
| "grad_norm": 12.952175792300656, | |
| "learning_rate": 1.4135585049665206e-06, | |
| "loss": 0.3178, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8710968775020016, | |
| "grad_norm": 12.728890379024975, | |
| "learning_rate": 1.3930250316539235e-06, | |
| "loss": 0.2974, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8839071257005604, | |
| "grad_norm": 14.929834721351973, | |
| "learning_rate": 1.3722936584019451e-06, | |
| "loss": 0.2682, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.8967173738991193, | |
| "grad_norm": 13.241782488686244, | |
| "learning_rate": 1.3513748240813427e-06, | |
| "loss": 0.3008, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9095276220976781, | |
| "grad_norm": 11.993846811913796, | |
| "learning_rate": 1.3302790619551672e-06, | |
| "loss": 0.2954, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.922337870296237, | |
| "grad_norm": 11.51482248851063, | |
| "learning_rate": 1.3090169943749473e-06, | |
| "loss": 0.2839, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9351481184947958, | |
| "grad_norm": 16.30894659038507, | |
| "learning_rate": 1.2875993274320173e-06, | |
| "loss": 0.2697, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9479583666933546, | |
| "grad_norm": 14.18702210008118, | |
| "learning_rate": 1.266036845566675e-06, | |
| "loss": 0.3257, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.9607686148919136, | |
| "grad_norm": 12.829612982865, | |
| "learning_rate": 1.244340406137894e-06, | |
| "loss": 0.2843, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9735788630904724, | |
| "grad_norm": 11.280572000591482, | |
| "learning_rate": 1.2225209339563143e-06, | |
| "loss": 0.2877, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9863891112890312, | |
| "grad_norm": 11.890947903730776, | |
| "learning_rate": 1.2005894157832728e-06, | |
| "loss": 0.3159, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9991993594875901, | |
| "grad_norm": 11.43791196956486, | |
| "learning_rate": 1.1785568947986366e-06, | |
| "loss": 0.2783, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.012009607686149, | |
| "grad_norm": 6.708403841137654, | |
| "learning_rate": 1.156434465040231e-06, | |
| "loss": 0.1292, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.0248198558847077, | |
| "grad_norm": 7.089840959480307, | |
| "learning_rate": 1.1342332658176555e-06, | |
| "loss": 0.1441, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0376301040832665, | |
| "grad_norm": 7.0309806199688705, | |
| "learning_rate": 1.1119644761033077e-06, | |
| "loss": 0.1405, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0504403522818255, | |
| "grad_norm": 8.710657806703308, | |
| "learning_rate": 1.0896393089034335e-06, | |
| "loss": 0.1265, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0632506004803843, | |
| "grad_norm": 8.365738125681913, | |
| "learning_rate": 1.0672690056120398e-06, | |
| "loss": 0.121, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0760608486789431, | |
| "grad_norm": 6.240558508799005, | |
| "learning_rate": 1.044864830350515e-06, | |
| "loss": 0.1072, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088871096877502, | |
| "grad_norm": 8.786219507218826, | |
| "learning_rate": 1.022438064295805e-06, | |
| "loss": 0.1256, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1016813450760607, | |
| "grad_norm": 7.367696121948762, | |
| "learning_rate": 1e-06, | |
| "loss": 0.1169, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1144915932746198, | |
| "grad_norm": 8.286075087364262, | |
| "learning_rate": 9.77561935704195e-07, | |
| "loss": 0.1302, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1273018414731786, | |
| "grad_norm": 6.69695777215013, | |
| "learning_rate": 9.551351696494853e-07, | |
| "loss": 0.1191, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1401120896717374, | |
| "grad_norm": 12.22193531590102, | |
| "learning_rate": 9.327309943879603e-07, | |
| "loss": 0.1089, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.1529223378702962, | |
| "grad_norm": 9.221889701081018, | |
| "learning_rate": 9.103606910965665e-07, | |
| "loss": 0.1224, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.165732586068855, | |
| "grad_norm": 19.718305211304475, | |
| "learning_rate": 8.880355238966921e-07, | |
| "loss": 0.1211, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.178542834267414, | |
| "grad_norm": 8.925338055742081, | |
| "learning_rate": 8.657667341823448e-07, | |
| "loss": 0.1178, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1913530824659728, | |
| "grad_norm": 8.25397448128133, | |
| "learning_rate": 8.435655349597689e-07, | |
| "loss": 0.127, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2041633306645316, | |
| "grad_norm": 10.64375599591567, | |
| "learning_rate": 8.214431052013634e-07, | |
| "loss": 0.1162, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.2169735788630904, | |
| "grad_norm": 9.047517084249554, | |
| "learning_rate": 7.994105842167272e-07, | |
| "loss": 0.1386, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2297838270616492, | |
| "grad_norm": 14.178667620453645, | |
| "learning_rate": 7.774790660436857e-07, | |
| "loss": 0.1137, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2425940752602083, | |
| "grad_norm": 7.4649396301693836, | |
| "learning_rate": 7.556595938621058e-07, | |
| "loss": 0.1189, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.255404323458767, | |
| "grad_norm": 9.120043718229002, | |
| "learning_rate": 7.33963154433325e-07, | |
| "loss": 0.1218, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2682145716573259, | |
| "grad_norm": 9.320538317849792, | |
| "learning_rate": 7.124006725679828e-07, | |
| "loss": 0.1215, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.2810248198558847, | |
| "grad_norm": 8.775423616734841, | |
| "learning_rate": 6.909830056250526e-07, | |
| "loss": 0.1106, | |
| "step": 100 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 156, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 135343586148352.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |