File size: 6,084 Bytes
5c7eb73 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 |
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.292181069958848,
"eval_steps": 500,
"global_step": 100,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.3292181069958848,
"grad_norm": 0.5561477541923523,
"learning_rate": 8.333333333333333e-07,
"logits/chosen": 1.7485754489898682,
"logits/rejected": 1.8832639455795288,
"logps/chosen": -70.18267059326172,
"logps/rejected": -77.9986343383789,
"loss": 0.6938,
"rewards/accuracies": 0.4749999940395355,
"rewards/chosen": 0.005453853867948055,
"rewards/margins": 0.013218576088547707,
"rewards/rejected": -0.007764720823615789,
"step": 10
},
{
"epoch": 0.6584362139917695,
"grad_norm": 0.48141908645629883,
"learning_rate": 1.6666666666666667e-06,
"logits/chosen": 1.9016907215118408,
"logits/rejected": 1.9251121282577515,
"logps/chosen": -96.5027847290039,
"logps/rejected": -87.00735473632812,
"loss": 0.6937,
"rewards/accuracies": 0.5249999761581421,
"rewards/chosen": -0.005797500256448984,
"rewards/margins": -0.0004409264656715095,
"rewards/rejected": -0.005356573965400457,
"step": 20
},
{
"epoch": 0.9876543209876543,
"grad_norm": 0.4639015197753906,
"learning_rate": 2.5e-06,
"logits/chosen": 1.7938541173934937,
"logits/rejected": 1.6962993144989014,
"logps/chosen": -71.47590637207031,
"logps/rejected": -66.45989227294922,
"loss": 0.6942,
"rewards/accuracies": 0.4124999940395355,
"rewards/chosen": -0.004830303601920605,
"rewards/margins": -0.00935445912182331,
"rewards/rejected": 0.004524155054241419,
"step": 30
},
{
"epoch": 1.316872427983539,
"grad_norm": 0.44931092858314514,
"learning_rate": 3.3333333333333333e-06,
"logits/chosen": 1.8256213665008545,
"logits/rejected": 1.8677200078964233,
"logps/chosen": -75.90711975097656,
"logps/rejected": -76.26548767089844,
"loss": 0.6935,
"rewards/accuracies": 0.5375000238418579,
"rewards/chosen": -0.004859285429120064,
"rewards/margins": -0.007447429001331329,
"rewards/rejected": 0.0025881435722112656,
"step": 40
},
{
"epoch": 1.646090534979424,
"grad_norm": 0.512350857257843,
"learning_rate": 4.166666666666667e-06,
"logits/chosen": 1.7572576999664307,
"logits/rejected": 1.7408854961395264,
"logps/chosen": -80.90664672851562,
"logps/rejected": -85.82096862792969,
"loss": 0.6937,
"rewards/accuracies": 0.4749999940395355,
"rewards/chosen": -0.00120059447363019,
"rewards/margins": -0.0018891148502007127,
"rewards/rejected": 0.0006885197362862527,
"step": 50
},
{
"epoch": 1.9753086419753085,
"grad_norm": 0.5260242819786072,
"learning_rate": 5e-06,
"logits/chosen": 1.827275037765503,
"logits/rejected": 1.8168131113052368,
"logps/chosen": -86.74467468261719,
"logps/rejected": -79.20576477050781,
"loss": 0.6937,
"rewards/accuracies": 0.4124999940395355,
"rewards/chosen": 0.0014693590346723795,
"rewards/margins": -0.0025326632894575596,
"rewards/rejected": 0.00400202302262187,
"step": 60
},
{
"epoch": 2.3045267489711936,
"grad_norm": 0.5231289267539978,
"learning_rate": 4.995770395678171e-06,
"logits/chosen": 1.7851394414901733,
"logits/rejected": 1.8952877521514893,
"logps/chosen": -81.03253173828125,
"logps/rejected": -88.5263442993164,
"loss": 0.6932,
"rewards/accuracies": 0.44999998807907104,
"rewards/chosen": -0.003017458599060774,
"rewards/margins": -0.002731734188273549,
"rewards/rejected": -0.0002857256622519344,
"step": 70
},
{
"epoch": 2.633744855967078,
"grad_norm": 0.5552361011505127,
"learning_rate": 4.983095894354858e-06,
"logits/chosen": 1.8227930068969727,
"logits/rejected": 1.7752052545547485,
"logps/chosen": -89.98479461669922,
"logps/rejected": -72.01054382324219,
"loss": 0.6928,
"rewards/accuracies": 0.48750001192092896,
"rewards/chosen": 0.0013810636010020971,
"rewards/margins": -0.0008344938978552818,
"rewards/rejected": 0.0022155570331960917,
"step": 80
},
{
"epoch": 2.962962962962963,
"grad_norm": 0.603434681892395,
"learning_rate": 4.962019382530521e-06,
"logits/chosen": 1.829049825668335,
"logits/rejected": 1.78665030002594,
"logps/chosen": -81.95549011230469,
"logps/rejected": -76.07003021240234,
"loss": 0.6909,
"rewards/accuracies": 0.5625,
"rewards/chosen": 0.0062006814405322075,
"rewards/margins": 0.003528360743075609,
"rewards/rejected": 0.00267231953330338,
"step": 90
},
{
"epoch": 3.292181069958848,
"grad_norm": 0.6200206279754639,
"learning_rate": 4.93261217644956e-06,
"logits/chosen": 1.7920262813568115,
"logits/rejected": 1.7817165851593018,
"logps/chosen": -84.39167022705078,
"logps/rejected": -85.04205322265625,
"loss": 0.6907,
"rewards/accuracies": 0.6000000238418579,
"rewards/chosen": 0.0063859038054943085,
"rewards/margins": 0.007152262143790722,
"rewards/rejected": -0.0007663581636734307,
"step": 100
}
],
"logging_steps": 10,
"max_steps": 600,
"num_input_tokens_seen": 0,
"num_train_epochs": 20,
"save_steps": 50,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 3.407626107753595e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}
|