diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,10563 @@ +{ + "best_metric": 2.0928452014923096, + "best_model_checkpoint": "/dodrio/scratch/projects/2023_005/llm-finetuning/results/falcon-7b-ft-mc4_nl_cleaned_tiny/tiny-3e-4lr+1152tbs+1ep+0.1wd/checkpoint-1700", + "epoch": 0.9995221255853961, + "global_step": 1743, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0, + "learning_rate": 0.0, + "loss": 4.2104, + "step": 1 + }, + { + "epoch": 0.0, + "learning_rate": 5.237502901441348e-05, + "loss": 4.1721, + "step": 2 + }, + { + "epoch": 0.0, + "learning_rate": 8.301245696202791e-05, + "loss": 4.1721, + "step": 3 + }, + { + "epoch": 0.0, + "learning_rate": 0.00010475005802882695, + "loss": 4.1338, + "step": 4 + }, + { + "epoch": 0.0, + "learning_rate": 0.0001216110513391074, + "loss": 3.9893, + "step": 5 + }, + { + "epoch": 0.0, + "learning_rate": 0.00013538748597644138, + "loss": 3.8484, + "step": 6 + }, + { + "epoch": 0.0, + "learning_rate": 0.0001470352954965235, + "loss": 3.8086, + "step": 7 + }, + { + "epoch": 0.0, + "learning_rate": 0.0001571250870432404, + "loss": 3.7847, + "step": 8 + }, + { + "epoch": 0.01, + "learning_rate": 0.00016602491392405581, + "loss": 3.7778, + "step": 9 + }, + { + "epoch": 0.01, + "learning_rate": 0.00017398608035352093, + "loss": 3.7383, + "step": 10 + }, + { + "epoch": 0.01, + "learning_rate": 0.00018118783139950783, + "loss": 3.6804, + "step": 11 + }, + { + "epoch": 0.01, + "learning_rate": 0.00018776251499085487, + "loss": 3.6416, + "step": 12 + }, + { + "epoch": 0.01, + "learning_rate": 0.00019381063760372773, + "loss": 3.6064, + "step": 13 + }, + { + "epoch": 0.01, + "learning_rate": 0.00019941032451093696, + "loss": 3.5791, + "step": 14 + }, + { + "epoch": 0.01, + "learning_rate": 0.0002046235083011353, + "loss": 3.5552, + "step": 15 + }, + { + "epoch": 0.01, + "learning_rate": 0.0002095001160576539, + "loss": 3.4915, + "step": 16 + }, + { + "epoch": 0.01, + "learning_rate": 0.0002140809849058235, + "loss": 3.467, + "step": 17 + }, + { + "epoch": 0.01, + "learning_rate": 0.00021839994293846928, + "loss": 3.4556, + "step": 18 + }, + { + "epoch": 0.01, + "learning_rate": 0.00022248532676773307, + "loss": 3.4517, + "step": 19 + }, + { + "epoch": 0.01, + "learning_rate": 0.00022636110936793434, + "loss": 3.4124, + "step": 20 + }, + { + "epoch": 0.01, + "learning_rate": 0.0002300477524585514, + "loss": 3.3765, + "step": 21 + }, + { + "epoch": 0.01, + "learning_rate": 0.00023356286041392133, + "loss": 3.3831, + "step": 22 + }, + { + "epoch": 0.01, + "learning_rate": 0.00023692168869698306, + "loss": 3.3608, + "step": 23 + }, + { + "epoch": 0.01, + "learning_rate": 0.00024013754400526832, + "loss": 3.3359, + "step": 24 + }, + { + "epoch": 0.01, + "learning_rate": 0.0002432221026782148, + "loss": 3.3135, + "step": 25 + }, + { + "epoch": 0.01, + "learning_rate": 0.0002461856666181412, + "loss": 3.3049, + "step": 26 + }, + { + "epoch": 0.02, + "learning_rate": 0.0002490373708860837, + "loss": 3.2999, + "step": 27 + }, + { + "epoch": 0.02, + "learning_rate": 0.00025178535352535046, + "loss": 3.2664, + "step": 28 + }, + { + "epoch": 0.02, + "learning_rate": 0.00025443689557127586, + "loss": 3.2631, + "step": 29 + }, + { + "epoch": 0.02, + "learning_rate": 0.0002569985373155488, + "loss": 3.2592, + "step": 30 + }, + { + "epoch": 0.02, + "learning_rate": 0.0002594761754996128, + "loss": 3.2306, + "step": 31 + }, + { + "epoch": 0.02, + "learning_rate": 0.0002618751450720674, + "loss": 3.2349, + "step": 32 + }, + { + "epoch": 0.02, + "learning_rate": 0.0002642002883615357, + "loss": 3.2136, + "step": 33 + }, + { + "epoch": 0.02, + "learning_rate": 0.00026645601392023696, + "loss": 3.2072, + "step": 34 + }, + { + "epoch": 0.02, + "learning_rate": 0.0002686463468356309, + "loss": 3.1746, + "step": 35 + }, + { + "epoch": 0.02, + "learning_rate": 0.00027077497195288275, + "loss": 3.1608, + "step": 36 + }, + { + "epoch": 0.02, + "learning_rate": 0.0002728452711740502, + "loss": 3.147, + "step": 37 + }, + { + "epoch": 0.02, + "learning_rate": 0.00027486035578214656, + "loss": 3.1412, + "step": 38 + }, + { + "epoch": 0.02, + "learning_rate": 0.0002768230945657556, + "loss": 3.1223, + "step": 39 + }, + { + "epoch": 0.02, + "learning_rate": 0.00027873613838234787, + "loss": 3.1102, + "step": 40 + }, + { + "epoch": 0.02, + "learning_rate": 0.0002806019416881012, + "loss": 3.1144, + "step": 41 + }, + { + "epoch": 0.02, + "learning_rate": 0.0002824227814729649, + "loss": 3.0933, + "step": 42 + }, + { + "epoch": 0.02, + "learning_rate": 0.0002842007739674116, + "loss": 3.1007, + "step": 43 + }, + { + "epoch": 0.03, + "learning_rate": 0.00028593788942833474, + "loss": 3.0845, + "step": 44 + }, + { + "epoch": 0.03, + "learning_rate": 0.0002876359652631632, + "loss": 3.0472, + "step": 45 + }, + { + "epoch": 0.03, + "learning_rate": 0.00028929671771139653, + "loss": 3.049, + "step": 46 + }, + { + "epoch": 0.03, + "learning_rate": 0.0002909217522697539, + "loss": 3.0353, + "step": 47 + }, + { + "epoch": 0.03, + "learning_rate": 0.0002925125730196818, + "loss": 3.022, + "step": 48 + }, + { + "epoch": 0.03, + "learning_rate": 0.000294070590993047, + "loss": 3.0054, + "step": 49 + }, + { + "epoch": 0.03, + "learning_rate": 0.0002955971316926283, + "loss": 2.9879, + "step": 50 + }, + { + "epoch": 0.03, + "learning_rate": 0.00029709344186785135, + "loss": 2.9656, + "step": 51 + }, + { + "epoch": 0.03, + "learning_rate": 0.00029856069563255467, + "loss": 2.9806, + "step": 52 + }, + { + "epoch": 0.03, + "learning_rate": 0.0003, + "loss": 2.9524, + "step": 53 + }, + { + "epoch": 0.03, + "learning_rate": 0.0003, + "loss": 2.9541, + "step": 54 + }, + { + "epoch": 0.03, + "learning_rate": 0.0003, + "loss": 2.9312, + "step": 55 + }, + { + "epoch": 0.03, + "learning_rate": 0.0003, + "loss": 2.9257, + "step": 56 + }, + { + "epoch": 0.03, + "learning_rate": 0.0003, + "loss": 2.9247, + "step": 57 + }, + { + "epoch": 0.03, + "learning_rate": 0.0003, + "loss": 2.911, + "step": 58 + }, + { + "epoch": 0.03, + "learning_rate": 0.0003, + "loss": 2.9258, + "step": 59 + }, + { + "epoch": 0.03, + "learning_rate": 0.0003, + "loss": 2.9089, + "step": 60 + }, + { + "epoch": 0.03, + "learning_rate": 0.0003, + "loss": 2.9037, + "step": 61 + }, + { + "epoch": 0.04, + "learning_rate": 0.0003, + "loss": 2.8817, + "step": 62 + }, + { + "epoch": 0.04, + "learning_rate": 0.0003, + "loss": 2.8732, + "step": 63 + }, + { + "epoch": 0.04, + "learning_rate": 0.0003, + "loss": 2.8701, + "step": 64 + }, + { + "epoch": 0.04, + "learning_rate": 0.0003, + "loss": 2.8821, + "step": 65 + }, + { + "epoch": 0.04, + "learning_rate": 0.0003, + "loss": 2.8696, + "step": 66 + }, + { + "epoch": 0.04, + "learning_rate": 0.0003, + "loss": 2.8462, + "step": 67 + }, + { + "epoch": 0.04, + "learning_rate": 0.0003, + "loss": 2.8441, + "step": 68 + }, + { + "epoch": 0.04, + "learning_rate": 0.0003, + "loss": 2.8455, + "step": 69 + }, + { + "epoch": 0.04, + "learning_rate": 0.0003, + "loss": 2.8361, + "step": 70 + }, + { + "epoch": 0.04, + "learning_rate": 0.0003, + "loss": 2.827, + "step": 71 + }, + { + "epoch": 0.04, + "learning_rate": 0.0003, + "loss": 2.8287, + "step": 72 + }, + { + "epoch": 0.04, + "learning_rate": 0.0003, + "loss": 2.8041, + "step": 73 + }, + { + "epoch": 0.04, + "learning_rate": 0.0003, + "loss": 2.8203, + "step": 74 + }, + { + "epoch": 0.04, + "learning_rate": 0.0003, + "loss": 2.7905, + "step": 75 + }, + { + "epoch": 0.04, + "learning_rate": 0.0003, + "loss": 2.8273, + "step": 76 + }, + { + "epoch": 0.04, + "learning_rate": 0.0003, + "loss": 2.7999, + "step": 77 + }, + { + "epoch": 0.04, + "learning_rate": 0.0003, + "loss": 2.7999, + "step": 78 + }, + { + "epoch": 0.05, + "learning_rate": 0.0003, + "loss": 2.7988, + "step": 79 + }, + { + "epoch": 0.05, + "learning_rate": 0.0003, + "loss": 2.7935, + "step": 80 + }, + { + "epoch": 0.05, + "learning_rate": 0.0003, + "loss": 2.8088, + "step": 81 + }, + { + "epoch": 0.05, + "learning_rate": 0.0003, + "loss": 2.7883, + "step": 82 + }, + { + "epoch": 0.05, + "learning_rate": 0.0003, + "loss": 2.7681, + "step": 83 + }, + { + "epoch": 0.05, + "learning_rate": 0.0003, + "loss": 2.7655, + "step": 84 + }, + { + "epoch": 0.05, + "learning_rate": 0.0003, + "loss": 2.7756, + "step": 85 + }, + { + "epoch": 0.05, + "learning_rate": 0.0003, + "loss": 2.7589, + "step": 86 + }, + { + "epoch": 0.05, + "learning_rate": 0.0003, + "loss": 2.7542, + "step": 87 + }, + { + "epoch": 0.05, + "learning_rate": 0.0003, + "loss": 2.7462, + "step": 88 + }, + { + "epoch": 0.05, + "learning_rate": 0.0003, + "loss": 2.7782, + "step": 89 + }, + { + "epoch": 0.05, + "learning_rate": 0.0003, + "loss": 2.7373, + "step": 90 + }, + { + "epoch": 0.05, + "learning_rate": 0.0003, + "loss": 2.748, + "step": 91 + }, + { + "epoch": 0.05, + "learning_rate": 0.0003, + "loss": 2.7476, + "step": 92 + }, + { + "epoch": 0.05, + "learning_rate": 0.0003, + "loss": 2.7506, + "step": 93 + }, + { + "epoch": 0.05, + "learning_rate": 0.0003, + "loss": 2.7318, + "step": 94 + }, + { + "epoch": 0.05, + "learning_rate": 0.0003, + "loss": 2.7477, + "step": 95 + }, + { + "epoch": 0.06, + "learning_rate": 0.0003, + "loss": 2.7357, + "step": 96 + }, + { + "epoch": 0.06, + "learning_rate": 0.0003, + "loss": 2.7202, + "step": 97 + }, + { + "epoch": 0.06, + "learning_rate": 0.0003, + "loss": 2.7266, + "step": 98 + }, + { + "epoch": 0.06, + "learning_rate": 0.0003, + "loss": 2.7301, + "step": 99 + }, + { + "epoch": 0.06, + "learning_rate": 0.0003, + "loss": 2.7336, + "step": 100 + }, + { + "epoch": 0.06, + "learning_rate": 0.0003, + "loss": 2.7251, + "step": 101 + }, + { + "epoch": 0.06, + "learning_rate": 0.0003, + "loss": 2.715, + "step": 102 + }, + { + "epoch": 0.06, + "learning_rate": 0.0003, + "loss": 2.7053, + "step": 103 + }, + { + "epoch": 0.06, + "learning_rate": 0.0003, + "loss": 2.7275, + "step": 104 + }, + { + "epoch": 0.06, + "learning_rate": 0.0003, + "loss": 2.6967, + "step": 105 + }, + { + "epoch": 0.06, + "learning_rate": 0.0003, + "loss": 2.6968, + "step": 106 + }, + { + "epoch": 0.06, + "learning_rate": 0.0003, + "loss": 2.7058, + "step": 107 + }, + { + "epoch": 0.06, + "learning_rate": 0.0003, + "loss": 2.705, + "step": 108 + }, + { + "epoch": 0.06, + "learning_rate": 0.0003, + "loss": 2.6992, + "step": 109 + }, + { + "epoch": 0.06, + "learning_rate": 0.0003, + "loss": 2.6975, + "step": 110 + }, + { + "epoch": 0.06, + "learning_rate": 0.0003, + "loss": 2.6932, + "step": 111 + }, + { + "epoch": 0.06, + "learning_rate": 0.0003, + "loss": 2.712, + "step": 112 + }, + { + "epoch": 0.06, + "learning_rate": 0.0003, + "loss": 2.6815, + "step": 113 + }, + { + "epoch": 0.07, + "learning_rate": 0.0003, + "loss": 2.6667, + "step": 114 + }, + { + "epoch": 0.07, + "learning_rate": 0.0003, + "loss": 2.7042, + "step": 115 + }, + { + "epoch": 0.07, + "learning_rate": 0.0003, + "loss": 2.696, + "step": 116 + }, + { + "epoch": 0.07, + "learning_rate": 0.0003, + "loss": 2.6992, + "step": 117 + }, + { + "epoch": 0.07, + "learning_rate": 0.0003, + "loss": 2.6871, + "step": 118 + }, + { + "epoch": 0.07, + "learning_rate": 0.0003, + "loss": 2.6862, + "step": 119 + }, + { + "epoch": 0.07, + "learning_rate": 0.0003, + "loss": 2.6887, + "step": 120 + }, + { + "epoch": 0.07, + "learning_rate": 0.0003, + "loss": 2.6815, + "step": 121 + }, + { + "epoch": 0.07, + "learning_rate": 0.0003, + "loss": 2.6837, + "step": 122 + }, + { + "epoch": 0.07, + "learning_rate": 0.0003, + "loss": 2.6664, + "step": 123 + }, + { + "epoch": 0.07, + "learning_rate": 0.0003, + "loss": 2.6797, + "step": 124 + }, + { + "epoch": 0.07, + "learning_rate": 0.0003, + "loss": 2.6735, + "step": 125 + }, + { + "epoch": 0.07, + "learning_rate": 0.0003, + "loss": 2.6709, + "step": 126 + }, + { + "epoch": 0.07, + "learning_rate": 0.0003, + "loss": 2.663, + "step": 127 + }, + { + "epoch": 0.07, + "learning_rate": 0.0003, + "loss": 2.6459, + "step": 128 + }, + { + "epoch": 0.07, + "learning_rate": 0.0003, + "loss": 2.6576, + "step": 129 + }, + { + "epoch": 0.07, + "learning_rate": 0.0003, + "loss": 2.6403, + "step": 130 + }, + { + "epoch": 0.08, + "learning_rate": 0.0003, + "loss": 2.653, + "step": 131 + }, + { + "epoch": 0.08, + "learning_rate": 0.0003, + "loss": 2.656, + "step": 132 + }, + { + "epoch": 0.08, + "learning_rate": 0.0003, + "loss": 2.6654, + "step": 133 + }, + { + "epoch": 0.08, + "learning_rate": 0.0003, + "loss": 2.6626, + "step": 134 + }, + { + "epoch": 0.08, + "learning_rate": 0.0003, + "loss": 2.6274, + "step": 135 + }, + { + "epoch": 0.08, + "learning_rate": 0.0003, + "loss": 2.6443, + "step": 136 + }, + { + "epoch": 0.08, + "learning_rate": 0.0003, + "loss": 2.649, + "step": 137 + }, + { + "epoch": 0.08, + "learning_rate": 0.0003, + "loss": 2.6444, + "step": 138 + }, + { + "epoch": 0.08, + "learning_rate": 0.0003, + "loss": 2.6447, + "step": 139 + }, + { + "epoch": 0.08, + "learning_rate": 0.0003, + "loss": 2.6469, + "step": 140 + }, + { + "epoch": 0.08, + "learning_rate": 0.0003, + "loss": 2.6604, + "step": 141 + }, + { + "epoch": 0.08, + "learning_rate": 0.0003, + "loss": 2.6233, + "step": 142 + }, + { + "epoch": 0.08, + "learning_rate": 0.0003, + "loss": 2.6315, + "step": 143 + }, + { + "epoch": 0.08, + "learning_rate": 0.0003, + "loss": 2.6233, + "step": 144 + }, + { + "epoch": 0.08, + "learning_rate": 0.0003, + "loss": 2.6442, + "step": 145 + }, + { + "epoch": 0.08, + "learning_rate": 0.0003, + "loss": 2.6294, + "step": 146 + }, + { + "epoch": 0.08, + "learning_rate": 0.0003, + "loss": 2.6359, + "step": 147 + }, + { + "epoch": 0.08, + "learning_rate": 0.0003, + "loss": 2.6283, + "step": 148 + }, + { + "epoch": 0.09, + "learning_rate": 0.0003, + "loss": 2.6069, + "step": 149 + }, + { + "epoch": 0.09, + "learning_rate": 0.0003, + "loss": 2.6334, + "step": 150 + }, + { + "epoch": 0.09, + "learning_rate": 0.0003, + "loss": 2.6166, + "step": 151 + }, + { + "epoch": 0.09, + "learning_rate": 0.0003, + "loss": 2.6276, + "step": 152 + }, + { + "epoch": 0.09, + "learning_rate": 0.0003, + "loss": 2.6279, + "step": 153 + }, + { + "epoch": 0.09, + "learning_rate": 0.0003, + "loss": 2.6163, + "step": 154 + }, + { + "epoch": 0.09, + "learning_rate": 0.0003, + "loss": 2.6223, + "step": 155 + }, + { + "epoch": 0.09, + "learning_rate": 0.0003, + "loss": 2.6271, + "step": 156 + }, + { + "epoch": 0.09, + "learning_rate": 0.0003, + "loss": 2.6188, + "step": 157 + }, + { + "epoch": 0.09, + "learning_rate": 0.0003, + "loss": 2.604, + "step": 158 + }, + { + "epoch": 0.09, + "learning_rate": 0.0003, + "loss": 2.6339, + "step": 159 + }, + { + "epoch": 0.09, + "learning_rate": 0.0003, + "loss": 2.6218, + "step": 160 + }, + { + "epoch": 0.09, + "learning_rate": 0.0003, + "loss": 2.6132, + "step": 161 + }, + { + "epoch": 0.09, + "learning_rate": 0.0003, + "loss": 2.6013, + "step": 162 + }, + { + "epoch": 0.09, + "learning_rate": 0.0003, + "loss": 2.6232, + "step": 163 + }, + { + "epoch": 0.09, + "learning_rate": 0.0003, + "loss": 2.6045, + "step": 164 + }, + { + "epoch": 0.09, + "learning_rate": 0.0003, + "loss": 2.6031, + "step": 165 + }, + { + "epoch": 0.1, + "learning_rate": 0.0003, + "loss": 2.6141, + "step": 166 + }, + { + "epoch": 0.1, + "learning_rate": 0.0003, + "loss": 2.6082, + "step": 167 + }, + { + "epoch": 0.1, + "learning_rate": 0.0003, + "loss": 2.5901, + "step": 168 + }, + { + "epoch": 0.1, + "learning_rate": 0.0003, + "loss": 2.6096, + "step": 169 + }, + { + "epoch": 0.1, + "learning_rate": 0.0003, + "loss": 2.6094, + "step": 170 + }, + { + "epoch": 0.1, + "eval_loss": 2.597968578338623, + "eval_runtime": 2023.7413, + "eval_samples_per_second": 52.123, + "eval_steps_per_second": 0.136, + "step": 170 + }, + { + "epoch": 0.1, + "learning_rate": 0.0003, + "loss": 2.5883, + "step": 171 + }, + { + "epoch": 0.1, + "learning_rate": 0.0003, + "loss": 2.5875, + "step": 172 + }, + { + "epoch": 0.1, + "learning_rate": 0.0003, + "loss": 2.5833, + "step": 173 + }, + { + "epoch": 0.1, + "learning_rate": 0.0003, + "loss": 2.5947, + "step": 174 + }, + { + "epoch": 0.1, + "learning_rate": 0.0003, + "loss": 2.5885, + "step": 175 + }, + { + "epoch": 0.1, + "learning_rate": 0.0003, + "loss": 2.587, + "step": 176 + }, + { + "epoch": 0.1, + "learning_rate": 0.0003, + "loss": 2.5936, + "step": 177 + }, + { + "epoch": 0.1, + "learning_rate": 0.0003, + "loss": 2.5754, + "step": 178 + }, + { + "epoch": 0.1, + "learning_rate": 0.0003, + "loss": 2.5938, + "step": 179 + }, + { + "epoch": 0.1, + "learning_rate": 0.0003, + "loss": 2.5767, + "step": 180 + }, + { + "epoch": 0.1, + "learning_rate": 0.0003, + "loss": 2.5706, + "step": 181 + }, + { + "epoch": 0.1, + "learning_rate": 0.0003, + "loss": 2.5713, + "step": 182 + }, + { + "epoch": 0.1, + "learning_rate": 0.0003, + "loss": 2.5765, + "step": 183 + }, + { + "epoch": 0.11, + "learning_rate": 0.0003, + "loss": 2.5885, + "step": 184 + }, + { + "epoch": 0.11, + "learning_rate": 0.0003, + "loss": 2.5938, + "step": 185 + }, + { + "epoch": 0.11, + "learning_rate": 0.0003, + "loss": 2.5797, + "step": 186 + }, + { + "epoch": 0.11, + "learning_rate": 0.0003, + "loss": 2.5862, + "step": 187 + }, + { + "epoch": 0.11, + "learning_rate": 0.0003, + "loss": 2.5769, + "step": 188 + }, + { + "epoch": 0.11, + "learning_rate": 0.0003, + "loss": 2.5646, + "step": 189 + }, + { + "epoch": 0.11, + "learning_rate": 0.0003, + "loss": 2.5612, + "step": 190 + }, + { + "epoch": 0.11, + "learning_rate": 0.0003, + "loss": 2.5712, + "step": 191 + }, + { + "epoch": 0.11, + "learning_rate": 0.0003, + "loss": 2.5712, + "step": 192 + }, + { + "epoch": 0.11, + "learning_rate": 0.0003, + "loss": 2.5662, + "step": 193 + }, + { + "epoch": 0.11, + "learning_rate": 0.0003, + "loss": 2.5802, + "step": 194 + }, + { + "epoch": 0.11, + "learning_rate": 0.0003, + "loss": 2.5674, + "step": 195 + }, + { + "epoch": 0.11, + "learning_rate": 0.0003, + "loss": 2.5715, + "step": 196 + }, + { + "epoch": 0.11, + "learning_rate": 0.0003, + "loss": 2.5675, + "step": 197 + }, + { + "epoch": 0.11, + "learning_rate": 0.0003, + "loss": 2.553, + "step": 198 + }, + { + "epoch": 0.11, + "learning_rate": 0.0003, + "loss": 2.5734, + "step": 199 + }, + { + "epoch": 0.11, + "learning_rate": 0.0003, + "loss": 2.5591, + "step": 200 + }, + { + "epoch": 0.12, + "learning_rate": 0.0003, + "loss": 2.5667, + "step": 201 + }, + { + "epoch": 0.12, + "learning_rate": 0.0003, + "loss": 2.5696, + "step": 202 + }, + { + "epoch": 0.12, + "learning_rate": 0.0003, + "loss": 2.5734, + "step": 203 + }, + { + "epoch": 0.12, + "learning_rate": 0.0003, + "loss": 2.5684, + "step": 204 + }, + { + "epoch": 0.12, + "learning_rate": 0.0003, + "loss": 2.5547, + "step": 205 + }, + { + "epoch": 0.12, + "learning_rate": 0.0003, + "loss": 2.5724, + "step": 206 + }, + { + "epoch": 0.12, + "learning_rate": 0.0003, + "loss": 2.5564, + "step": 207 + }, + { + "epoch": 0.12, + "learning_rate": 0.0003, + "loss": 2.5759, + "step": 208 + }, + { + "epoch": 0.12, + "learning_rate": 0.0003, + "loss": 2.5417, + "step": 209 + }, + { + "epoch": 0.12, + "learning_rate": 0.0003, + "loss": 2.5337, + "step": 210 + }, + { + "epoch": 0.12, + "learning_rate": 0.0003, + "loss": 2.5597, + "step": 211 + }, + { + "epoch": 0.12, + "learning_rate": 0.0003, + "loss": 2.5576, + "step": 212 + }, + { + "epoch": 0.12, + "learning_rate": 0.0003, + "loss": 2.5587, + "step": 213 + }, + { + "epoch": 0.12, + "learning_rate": 0.0003, + "loss": 2.5542, + "step": 214 + }, + { + "epoch": 0.12, + "learning_rate": 0.0003, + "loss": 2.5605, + "step": 215 + }, + { + "epoch": 0.12, + "learning_rate": 0.0003, + "loss": 2.5524, + "step": 216 + }, + { + "epoch": 0.12, + "learning_rate": 0.0003, + "loss": 2.5347, + "step": 217 + }, + { + "epoch": 0.13, + "learning_rate": 0.0003, + "loss": 2.5433, + "step": 218 + }, + { + "epoch": 0.13, + "learning_rate": 0.0003, + "loss": 2.5392, + "step": 219 + }, + { + "epoch": 0.13, + "learning_rate": 0.0003, + "loss": 2.5477, + "step": 220 + }, + { + "epoch": 0.13, + "learning_rate": 0.0003, + "loss": 2.5471, + "step": 221 + }, + { + "epoch": 0.13, + "learning_rate": 0.0003, + "loss": 2.5338, + "step": 222 + }, + { + "epoch": 0.13, + "learning_rate": 0.0003, + "loss": 2.5425, + "step": 223 + }, + { + "epoch": 0.13, + "learning_rate": 0.0003, + "loss": 2.536, + "step": 224 + }, + { + "epoch": 0.13, + "learning_rate": 0.0003, + "loss": 2.5502, + "step": 225 + }, + { + "epoch": 0.13, + "learning_rate": 0.0003, + "loss": 2.5173, + "step": 226 + }, + { + "epoch": 0.13, + "learning_rate": 0.0003, + "loss": 2.5114, + "step": 227 + }, + { + "epoch": 0.13, + "learning_rate": 0.0003, + "loss": 2.541, + "step": 228 + }, + { + "epoch": 0.13, + "learning_rate": 0.0003, + "loss": 2.5415, + "step": 229 + }, + { + "epoch": 0.13, + "learning_rate": 0.0003, + "loss": 2.5284, + "step": 230 + }, + { + "epoch": 0.13, + "learning_rate": 0.0003, + "loss": 2.5331, + "step": 231 + }, + { + "epoch": 0.13, + "learning_rate": 0.0003, + "loss": 2.5424, + "step": 232 + }, + { + "epoch": 0.13, + "learning_rate": 0.0003, + "loss": 2.5321, + "step": 233 + }, + { + "epoch": 0.13, + "learning_rate": 0.0003, + "loss": 2.5276, + "step": 234 + }, + { + "epoch": 0.13, + "learning_rate": 0.0003, + "loss": 2.5284, + "step": 235 + }, + { + "epoch": 0.14, + "learning_rate": 0.0003, + "loss": 2.533, + "step": 236 + }, + { + "epoch": 0.14, + "learning_rate": 0.0003, + "loss": 2.5334, + "step": 237 + }, + { + "epoch": 0.14, + "learning_rate": 0.0003, + "loss": 2.5155, + "step": 238 + }, + { + "epoch": 0.14, + "learning_rate": 0.0003, + "loss": 2.5365, + "step": 239 + }, + { + "epoch": 0.14, + "learning_rate": 0.0003, + "loss": 2.5221, + "step": 240 + }, + { + "epoch": 0.14, + "learning_rate": 0.0003, + "loss": 2.5116, + "step": 241 + }, + { + "epoch": 0.14, + "learning_rate": 0.0003, + "loss": 2.5258, + "step": 242 + }, + { + "epoch": 0.14, + "learning_rate": 0.0003, + "loss": 2.5239, + "step": 243 + }, + { + "epoch": 0.14, + "learning_rate": 0.0003, + "loss": 2.5172, + "step": 244 + }, + { + "epoch": 0.14, + "learning_rate": 0.0003, + "loss": 2.5134, + "step": 245 + }, + { + "epoch": 0.14, + "learning_rate": 0.0003, + "loss": 2.5103, + "step": 246 + }, + { + "epoch": 0.14, + "learning_rate": 0.0003, + "loss": 2.5175, + "step": 247 + }, + { + "epoch": 0.14, + "learning_rate": 0.0003, + "loss": 2.5155, + "step": 248 + }, + { + "epoch": 0.14, + "learning_rate": 0.0003, + "loss": 2.5157, + "step": 249 + }, + { + "epoch": 0.14, + "learning_rate": 0.0003, + "loss": 2.5088, + "step": 250 + }, + { + "epoch": 0.14, + "learning_rate": 0.0003, + "loss": 2.4944, + "step": 251 + }, + { + "epoch": 0.14, + "learning_rate": 0.0003, + "loss": 2.5162, + "step": 252 + }, + { + "epoch": 0.15, + "learning_rate": 0.0003, + "loss": 2.5199, + "step": 253 + }, + { + "epoch": 0.15, + "learning_rate": 0.0003, + "loss": 2.5033, + "step": 254 + }, + { + "epoch": 0.15, + "learning_rate": 0.0003, + "loss": 2.5067, + "step": 255 + }, + { + "epoch": 0.15, + "learning_rate": 0.0003, + "loss": 2.5111, + "step": 256 + }, + { + "epoch": 0.15, + "learning_rate": 0.0003, + "loss": 2.5153, + "step": 257 + }, + { + "epoch": 0.15, + "learning_rate": 0.0003, + "loss": 2.5074, + "step": 258 + }, + { + "epoch": 0.15, + "learning_rate": 0.0003, + "loss": 2.5093, + "step": 259 + }, + { + "epoch": 0.15, + "learning_rate": 0.0003, + "loss": 2.5101, + "step": 260 + }, + { + "epoch": 0.15, + "learning_rate": 0.0003, + "loss": 2.4918, + "step": 261 + }, + { + "epoch": 0.15, + "learning_rate": 0.0003, + "loss": 2.5114, + "step": 262 + }, + { + "epoch": 0.15, + "learning_rate": 0.0003, + "loss": 2.4883, + "step": 263 + }, + { + "epoch": 0.15, + "learning_rate": 0.0003, + "loss": 2.5061, + "step": 264 + }, + { + "epoch": 0.15, + "learning_rate": 0.0003, + "loss": 2.5, + "step": 265 + }, + { + "epoch": 0.15, + "learning_rate": 0.0003, + "loss": 2.4993, + "step": 266 + }, + { + "epoch": 0.15, + "learning_rate": 0.0003, + "loss": 2.5095, + "step": 267 + }, + { + "epoch": 0.15, + "learning_rate": 0.0003, + "loss": 2.5073, + "step": 268 + }, + { + "epoch": 0.15, + "learning_rate": 0.0003, + "loss": 2.4948, + "step": 269 + }, + { + "epoch": 0.15, + "learning_rate": 0.0003, + "loss": 2.4874, + "step": 270 + }, + { + "epoch": 0.16, + "learning_rate": 0.0003, + "loss": 2.5004, + "step": 271 + }, + { + "epoch": 0.16, + "learning_rate": 0.0003, + "loss": 2.4885, + "step": 272 + }, + { + "epoch": 0.16, + "learning_rate": 0.0003, + "loss": 2.4974, + "step": 273 + }, + { + "epoch": 0.16, + "learning_rate": 0.0003, + "loss": 2.4857, + "step": 274 + }, + { + "epoch": 0.16, + "learning_rate": 0.0003, + "loss": 2.499, + "step": 275 + }, + { + "epoch": 0.16, + "learning_rate": 0.0003, + "loss": 2.4764, + "step": 276 + }, + { + "epoch": 0.16, + "learning_rate": 0.0003, + "loss": 2.4943, + "step": 277 + }, + { + "epoch": 0.16, + "learning_rate": 0.0003, + "loss": 2.5027, + "step": 278 + }, + { + "epoch": 0.16, + "learning_rate": 0.0003, + "loss": 2.4895, + "step": 279 + }, + { + "epoch": 0.16, + "learning_rate": 0.0003, + "loss": 2.5068, + "step": 280 + }, + { + "epoch": 0.16, + "learning_rate": 0.0003, + "loss": 2.4602, + "step": 281 + }, + { + "epoch": 0.16, + "learning_rate": 0.0003, + "loss": 2.4696, + "step": 282 + }, + { + "epoch": 0.16, + "learning_rate": 0.0003, + "loss": 2.488, + "step": 283 + }, + { + "epoch": 0.16, + "learning_rate": 0.0003, + "loss": 2.4813, + "step": 284 + }, + { + "epoch": 0.16, + "learning_rate": 0.0003, + "loss": 2.483, + "step": 285 + }, + { + "epoch": 0.16, + "learning_rate": 0.0003, + "loss": 2.4839, + "step": 286 + }, + { + "epoch": 0.16, + "learning_rate": 0.0003, + "loss": 2.4772, + "step": 287 + }, + { + "epoch": 0.17, + "learning_rate": 0.0003, + "loss": 2.4855, + "step": 288 + }, + { + "epoch": 0.17, + "learning_rate": 0.0003, + "loss": 2.4927, + "step": 289 + }, + { + "epoch": 0.17, + "learning_rate": 0.0003, + "loss": 2.4691, + "step": 290 + }, + { + "epoch": 0.17, + "learning_rate": 0.0003, + "loss": 2.4749, + "step": 291 + }, + { + "epoch": 0.17, + "learning_rate": 0.0003, + "loss": 2.4841, + "step": 292 + }, + { + "epoch": 0.17, + "learning_rate": 0.0003, + "loss": 2.4791, + "step": 293 + }, + { + "epoch": 0.17, + "learning_rate": 0.0003, + "loss": 2.4744, + "step": 294 + }, + { + "epoch": 0.17, + "learning_rate": 0.0003, + "loss": 2.4619, + "step": 295 + }, + { + "epoch": 0.17, + "learning_rate": 0.0003, + "loss": 2.4788, + "step": 296 + }, + { + "epoch": 0.17, + "learning_rate": 0.0003, + "loss": 2.479, + "step": 297 + }, + { + "epoch": 0.17, + "learning_rate": 0.0003, + "loss": 2.4668, + "step": 298 + }, + { + "epoch": 0.17, + "learning_rate": 0.0003, + "loss": 2.4773, + "step": 299 + }, + { + "epoch": 0.17, + "learning_rate": 0.0003, + "loss": 2.4664, + "step": 300 + }, + { + "epoch": 0.17, + "learning_rate": 0.0003, + "loss": 2.467, + "step": 301 + }, + { + "epoch": 0.17, + "learning_rate": 0.0003, + "loss": 2.4764, + "step": 302 + }, + { + "epoch": 0.17, + "learning_rate": 0.0003, + "loss": 2.4626, + "step": 303 + }, + { + "epoch": 0.17, + "learning_rate": 0.0003, + "loss": 2.4656, + "step": 304 + }, + { + "epoch": 0.17, + "learning_rate": 0.0003, + "loss": 2.4617, + "step": 305 + }, + { + "epoch": 0.18, + "learning_rate": 0.0003, + "loss": 2.4637, + "step": 306 + }, + { + "epoch": 0.18, + "learning_rate": 0.0003, + "loss": 2.4652, + "step": 307 + }, + { + "epoch": 0.18, + "learning_rate": 0.0003, + "loss": 2.4617, + "step": 308 + }, + { + "epoch": 0.18, + "learning_rate": 0.0003, + "loss": 2.4493, + "step": 309 + }, + { + "epoch": 0.18, + "learning_rate": 0.0003, + "loss": 2.4572, + "step": 310 + }, + { + "epoch": 0.18, + "learning_rate": 0.0003, + "loss": 2.4662, + "step": 311 + }, + { + "epoch": 0.18, + "learning_rate": 0.0003, + "loss": 2.4504, + "step": 312 + }, + { + "epoch": 0.18, + "learning_rate": 0.0003, + "loss": 2.4644, + "step": 313 + }, + { + "epoch": 0.18, + "learning_rate": 0.0003, + "loss": 2.4673, + "step": 314 + }, + { + "epoch": 0.18, + "learning_rate": 0.0003, + "loss": 2.4557, + "step": 315 + }, + { + "epoch": 0.18, + "learning_rate": 0.0003, + "loss": 2.463, + "step": 316 + }, + { + "epoch": 0.18, + "learning_rate": 0.0003, + "loss": 2.4761, + "step": 317 + }, + { + "epoch": 0.18, + "learning_rate": 0.0003, + "loss": 2.4626, + "step": 318 + }, + { + "epoch": 0.18, + "learning_rate": 0.0003, + "loss": 2.4476, + "step": 319 + }, + { + "epoch": 0.18, + "learning_rate": 0.0003, + "loss": 2.4576, + "step": 320 + }, + { + "epoch": 0.18, + "learning_rate": 0.0003, + "loss": 2.4647, + "step": 321 + }, + { + "epoch": 0.18, + "learning_rate": 0.0003, + "loss": 2.4614, + "step": 322 + }, + { + "epoch": 0.19, + "learning_rate": 0.0003, + "loss": 2.4534, + "step": 323 + }, + { + "epoch": 0.19, + "learning_rate": 0.0003, + "loss": 2.453, + "step": 324 + }, + { + "epoch": 0.19, + "learning_rate": 0.0003, + "loss": 2.4573, + "step": 325 + }, + { + "epoch": 0.19, + "learning_rate": 0.0003, + "loss": 2.4558, + "step": 326 + }, + { + "epoch": 0.19, + "learning_rate": 0.0003, + "loss": 2.457, + "step": 327 + }, + { + "epoch": 0.19, + "learning_rate": 0.0003, + "loss": 2.4698, + "step": 328 + }, + { + "epoch": 0.19, + "learning_rate": 0.0003, + "loss": 2.4473, + "step": 329 + }, + { + "epoch": 0.19, + "learning_rate": 0.0003, + "loss": 2.4576, + "step": 330 + }, + { + "epoch": 0.19, + "learning_rate": 0.0003, + "loss": 2.4445, + "step": 331 + }, + { + "epoch": 0.19, + "learning_rate": 0.0003, + "loss": 2.4458, + "step": 332 + }, + { + "epoch": 0.19, + "learning_rate": 0.0003, + "loss": 2.4585, + "step": 333 + }, + { + "epoch": 0.19, + "learning_rate": 0.0003, + "loss": 2.4602, + "step": 334 + }, + { + "epoch": 0.19, + "learning_rate": 0.0003, + "loss": 2.4591, + "step": 335 + }, + { + "epoch": 0.19, + "learning_rate": 0.0003, + "loss": 2.4468, + "step": 336 + }, + { + "epoch": 0.19, + "learning_rate": 0.0003, + "loss": 2.4347, + "step": 337 + }, + { + "epoch": 0.19, + "learning_rate": 0.0003, + "loss": 2.4319, + "step": 338 + }, + { + "epoch": 0.19, + "learning_rate": 0.0003, + "loss": 2.4503, + "step": 339 + }, + { + "epoch": 0.19, + "learning_rate": 0.0003, + "loss": 2.4503, + "step": 340 + }, + { + "epoch": 0.19, + "eval_loss": 2.4404726028442383, + "eval_runtime": 2008.6664, + "eval_samples_per_second": 52.514, + "eval_steps_per_second": 0.137, + "step": 340 + }, + { + "epoch": 0.2, + "learning_rate": 0.0003, + "loss": 2.4263, + "step": 341 + }, + { + "epoch": 0.2, + "learning_rate": 0.0003, + "loss": 2.4492, + "step": 342 + }, + { + "epoch": 0.2, + "learning_rate": 0.0003, + "loss": 2.4585, + "step": 343 + }, + { + "epoch": 0.2, + "learning_rate": 0.0003, + "loss": 2.4452, + "step": 344 + }, + { + "epoch": 0.2, + "learning_rate": 0.0003, + "loss": 2.4491, + "step": 345 + }, + { + "epoch": 0.2, + "learning_rate": 0.0003, + "loss": 2.4307, + "step": 346 + }, + { + "epoch": 0.2, + "learning_rate": 0.0003, + "loss": 2.4489, + "step": 347 + }, + { + "epoch": 0.2, + "learning_rate": 0.0003, + "loss": 2.4377, + "step": 348 + }, + { + "epoch": 0.2, + "learning_rate": 0.0003, + "loss": 2.4376, + "step": 349 + }, + { + "epoch": 0.2, + "learning_rate": 0.0003, + "loss": 2.4412, + "step": 350 + }, + { + "epoch": 0.2, + "learning_rate": 0.0003, + "loss": 2.454, + "step": 351 + }, + { + "epoch": 0.2, + "learning_rate": 0.0003, + "loss": 2.4238, + "step": 352 + }, + { + "epoch": 0.2, + "learning_rate": 0.0003, + "loss": 2.4399, + "step": 353 + }, + { + "epoch": 0.2, + "learning_rate": 0.0003, + "loss": 2.421, + "step": 354 + }, + { + "epoch": 0.2, + "learning_rate": 0.0003, + "loss": 2.4409, + "step": 355 + }, + { + "epoch": 0.2, + "learning_rate": 0.0003, + "loss": 2.4294, + "step": 356 + }, + { + "epoch": 0.2, + "learning_rate": 0.0003, + "loss": 2.4354, + "step": 357 + }, + { + "epoch": 0.21, + "learning_rate": 0.0003, + "loss": 2.4299, + "step": 358 + }, + { + "epoch": 0.21, + "learning_rate": 0.0003, + "loss": 2.4137, + "step": 359 + }, + { + "epoch": 0.21, + "learning_rate": 0.0003, + "loss": 2.4293, + "step": 360 + }, + { + "epoch": 0.21, + "learning_rate": 0.0003, + "loss": 2.4191, + "step": 361 + }, + { + "epoch": 0.21, + "learning_rate": 0.0003, + "loss": 2.4316, + "step": 362 + }, + { + "epoch": 0.21, + "learning_rate": 0.0003, + "loss": 2.4276, + "step": 363 + }, + { + "epoch": 0.21, + "learning_rate": 0.0003, + "loss": 2.4379, + "step": 364 + }, + { + "epoch": 0.21, + "learning_rate": 0.0003, + "loss": 2.4098, + "step": 365 + }, + { + "epoch": 0.21, + "learning_rate": 0.0003, + "loss": 2.4307, + "step": 366 + }, + { + "epoch": 0.21, + "learning_rate": 0.0003, + "loss": 2.4102, + "step": 367 + }, + { + "epoch": 0.21, + "learning_rate": 0.0003, + "loss": 2.431, + "step": 368 + }, + { + "epoch": 0.21, + "learning_rate": 0.0003, + "loss": 2.4303, + "step": 369 + }, + { + "epoch": 0.21, + "learning_rate": 0.0003, + "loss": 2.4203, + "step": 370 + }, + { + "epoch": 0.21, + "learning_rate": 0.0003, + "loss": 2.417, + "step": 371 + }, + { + "epoch": 0.21, + "learning_rate": 0.0003, + "loss": 2.4255, + "step": 372 + }, + { + "epoch": 0.21, + "learning_rate": 0.0003, + "loss": 2.4141, + "step": 373 + }, + { + "epoch": 0.21, + "learning_rate": 0.0003, + "loss": 2.4302, + "step": 374 + }, + { + "epoch": 0.22, + "learning_rate": 0.0003, + "loss": 2.4165, + "step": 375 + }, + { + "epoch": 0.22, + "learning_rate": 0.0003, + "loss": 2.4274, + "step": 376 + }, + { + "epoch": 0.22, + "learning_rate": 0.0003, + "loss": 2.4143, + "step": 377 + }, + { + "epoch": 0.22, + "learning_rate": 0.0003, + "loss": 2.4293, + "step": 378 + }, + { + "epoch": 0.22, + "learning_rate": 0.0003, + "loss": 2.4025, + "step": 379 + }, + { + "epoch": 0.22, + "learning_rate": 0.0003, + "loss": 2.4155, + "step": 380 + }, + { + "epoch": 0.22, + "learning_rate": 0.0003, + "loss": 2.418, + "step": 381 + }, + { + "epoch": 0.22, + "learning_rate": 0.0003, + "loss": 2.4388, + "step": 382 + }, + { + "epoch": 0.22, + "learning_rate": 0.0003, + "loss": 2.4105, + "step": 383 + }, + { + "epoch": 0.22, + "learning_rate": 0.0003, + "loss": 2.427, + "step": 384 + }, + { + "epoch": 0.22, + "learning_rate": 0.0003, + "loss": 2.4067, + "step": 385 + }, + { + "epoch": 0.22, + "learning_rate": 0.0003, + "loss": 2.3934, + "step": 386 + }, + { + "epoch": 0.22, + "learning_rate": 0.0003, + "loss": 2.3995, + "step": 387 + }, + { + "epoch": 0.22, + "learning_rate": 0.0003, + "loss": 2.413, + "step": 388 + }, + { + "epoch": 0.22, + "learning_rate": 0.0003, + "loss": 2.4034, + "step": 389 + }, + { + "epoch": 0.22, + "learning_rate": 0.0003, + "loss": 2.4047, + "step": 390 + }, + { + "epoch": 0.22, + "learning_rate": 0.0003, + "loss": 2.4116, + "step": 391 + }, + { + "epoch": 0.22, + "learning_rate": 0.0003, + "loss": 2.4015, + "step": 392 + }, + { + "epoch": 0.23, + "learning_rate": 0.0003, + "loss": 2.3853, + "step": 393 + }, + { + "epoch": 0.23, + "learning_rate": 0.0003, + "loss": 2.4136, + "step": 394 + }, + { + "epoch": 0.23, + "learning_rate": 0.0003, + "loss": 2.4115, + "step": 395 + }, + { + "epoch": 0.23, + "learning_rate": 0.0003, + "loss": 2.4016, + "step": 396 + }, + { + "epoch": 0.23, + "learning_rate": 0.0003, + "loss": 2.3995, + "step": 397 + }, + { + "epoch": 0.23, + "learning_rate": 0.0003, + "loss": 2.3936, + "step": 398 + }, + { + "epoch": 0.23, + "learning_rate": 0.0003, + "loss": 2.3972, + "step": 399 + }, + { + "epoch": 0.23, + "learning_rate": 0.0003, + "loss": 2.4122, + "step": 400 + }, + { + "epoch": 0.23, + "learning_rate": 0.0003, + "loss": 2.405, + "step": 401 + }, + { + "epoch": 0.23, + "learning_rate": 0.0003, + "loss": 2.4066, + "step": 402 + }, + { + "epoch": 0.23, + "learning_rate": 0.0003, + "loss": 2.3806, + "step": 403 + }, + { + "epoch": 0.23, + "learning_rate": 0.0003, + "loss": 2.4055, + "step": 404 + }, + { + "epoch": 0.23, + "learning_rate": 0.0003, + "loss": 2.4219, + "step": 405 + }, + { + "epoch": 0.23, + "learning_rate": 0.0003, + "loss": 2.3945, + "step": 406 + }, + { + "epoch": 0.23, + "learning_rate": 0.0003, + "loss": 2.387, + "step": 407 + }, + { + "epoch": 0.23, + "learning_rate": 0.0003, + "loss": 2.3915, + "step": 408 + }, + { + "epoch": 0.23, + "learning_rate": 0.0003, + "loss": 2.3801, + "step": 409 + }, + { + "epoch": 0.24, + "learning_rate": 0.0003, + "loss": 2.3944, + "step": 410 + }, + { + "epoch": 0.24, + "learning_rate": 0.0003, + "loss": 2.381, + "step": 411 + }, + { + "epoch": 0.24, + "learning_rate": 0.0003, + "loss": 2.391, + "step": 412 + }, + { + "epoch": 0.24, + "learning_rate": 0.0003, + "loss": 2.421, + "step": 413 + }, + { + "epoch": 0.24, + "learning_rate": 0.0003, + "loss": 2.3994, + "step": 414 + }, + { + "epoch": 0.24, + "learning_rate": 0.0003, + "loss": 2.4015, + "step": 415 + }, + { + "epoch": 0.24, + "learning_rate": 0.0003, + "loss": 2.417, + "step": 416 + }, + { + "epoch": 0.24, + "learning_rate": 0.0003, + "loss": 2.396, + "step": 417 + }, + { + "epoch": 0.24, + "learning_rate": 0.0003, + "loss": 2.39, + "step": 418 + }, + { + "epoch": 0.24, + "learning_rate": 0.0003, + "loss": 2.3772, + "step": 419 + }, + { + "epoch": 0.24, + "learning_rate": 0.0003, + "loss": 2.3873, + "step": 420 + }, + { + "epoch": 0.24, + "learning_rate": 0.0003, + "loss": 2.3848, + "step": 421 + }, + { + "epoch": 0.24, + "learning_rate": 0.0003, + "loss": 2.3833, + "step": 422 + }, + { + "epoch": 0.24, + "learning_rate": 0.0003, + "loss": 2.3983, + "step": 423 + }, + { + "epoch": 0.24, + "learning_rate": 0.0003, + "loss": 2.3928, + "step": 424 + }, + { + "epoch": 0.24, + "learning_rate": 0.0003, + "loss": 2.3976, + "step": 425 + }, + { + "epoch": 0.24, + "learning_rate": 0.0003, + "loss": 2.3855, + "step": 426 + }, + { + "epoch": 0.24, + "learning_rate": 0.0003, + "loss": 2.3795, + "step": 427 + }, + { + "epoch": 0.25, + "learning_rate": 0.0003, + "loss": 2.3698, + "step": 428 + }, + { + "epoch": 0.25, + "learning_rate": 0.0003, + "loss": 2.3884, + "step": 429 + }, + { + "epoch": 0.25, + "learning_rate": 0.0003, + "loss": 2.3906, + "step": 430 + }, + { + "epoch": 0.25, + "learning_rate": 0.0003, + "loss": 2.3829, + "step": 431 + }, + { + "epoch": 0.25, + "learning_rate": 0.0003, + "loss": 2.3901, + "step": 432 + }, + { + "epoch": 0.25, + "learning_rate": 0.0003, + "loss": 2.3947, + "step": 433 + }, + { + "epoch": 0.25, + "learning_rate": 0.0003, + "loss": 2.4006, + "step": 434 + }, + { + "epoch": 0.25, + "learning_rate": 0.0003, + "loss": 2.3788, + "step": 435 + }, + { + "epoch": 0.25, + "learning_rate": 0.0003, + "loss": 2.4069, + "step": 436 + }, + { + "epoch": 0.25, + "learning_rate": 0.0003, + "loss": 2.3892, + "step": 437 + }, + { + "epoch": 0.25, + "learning_rate": 0.0003, + "loss": 2.3904, + "step": 438 + }, + { + "epoch": 0.25, + "learning_rate": 0.0003, + "loss": 2.3951, + "step": 439 + }, + { + "epoch": 0.25, + "learning_rate": 0.0003, + "loss": 2.3824, + "step": 440 + }, + { + "epoch": 0.25, + "learning_rate": 0.0003, + "loss": 2.3993, + "step": 441 + }, + { + "epoch": 0.25, + "learning_rate": 0.0003, + "loss": 2.3583, + "step": 442 + }, + { + "epoch": 0.25, + "learning_rate": 0.0003, + "loss": 2.3689, + "step": 443 + }, + { + "epoch": 0.25, + "learning_rate": 0.0003, + "loss": 2.371, + "step": 444 + }, + { + "epoch": 0.26, + "learning_rate": 0.0003, + "loss": 2.3771, + "step": 445 + }, + { + "epoch": 0.26, + "learning_rate": 0.0003, + "loss": 2.3654, + "step": 446 + }, + { + "epoch": 0.26, + "learning_rate": 0.0003, + "loss": 2.3801, + "step": 447 + }, + { + "epoch": 0.26, + "learning_rate": 0.0003, + "loss": 2.3784, + "step": 448 + }, + { + "epoch": 0.26, + "learning_rate": 0.0003, + "loss": 2.3579, + "step": 449 + }, + { + "epoch": 0.26, + "learning_rate": 0.0003, + "loss": 2.3735, + "step": 450 + }, + { + "epoch": 0.26, + "learning_rate": 0.0003, + "loss": 2.3788, + "step": 451 + }, + { + "epoch": 0.26, + "learning_rate": 0.0003, + "loss": 2.3546, + "step": 452 + }, + { + "epoch": 0.26, + "learning_rate": 0.0003, + "loss": 2.3691, + "step": 453 + }, + { + "epoch": 0.26, + "learning_rate": 0.0003, + "loss": 2.3632, + "step": 454 + }, + { + "epoch": 0.26, + "learning_rate": 0.0003, + "loss": 2.3422, + "step": 455 + }, + { + "epoch": 0.26, + "learning_rate": 0.0003, + "loss": 2.3749, + "step": 456 + }, + { + "epoch": 0.26, + "learning_rate": 0.0003, + "loss": 2.3722, + "step": 457 + }, + { + "epoch": 0.26, + "learning_rate": 0.0003, + "loss": 2.3669, + "step": 458 + }, + { + "epoch": 0.26, + "learning_rate": 0.0003, + "loss": 2.3696, + "step": 459 + }, + { + "epoch": 0.26, + "learning_rate": 0.0003, + "loss": 2.36, + "step": 460 + }, + { + "epoch": 0.26, + "learning_rate": 0.0003, + "loss": 2.3785, + "step": 461 + }, + { + "epoch": 0.26, + "learning_rate": 0.0003, + "loss": 2.3701, + "step": 462 + }, + { + "epoch": 0.27, + "learning_rate": 0.0003, + "loss": 2.3745, + "step": 463 + }, + { + "epoch": 0.27, + "learning_rate": 0.0003, + "loss": 2.3655, + "step": 464 + }, + { + "epoch": 0.27, + "learning_rate": 0.0003, + "loss": 2.3676, + "step": 465 + }, + { + "epoch": 0.27, + "learning_rate": 0.0003, + "loss": 2.3723, + "step": 466 + }, + { + "epoch": 0.27, + "learning_rate": 0.0003, + "loss": 2.3478, + "step": 467 + }, + { + "epoch": 0.27, + "learning_rate": 0.0003, + "loss": 2.3634, + "step": 468 + }, + { + "epoch": 0.27, + "learning_rate": 0.0003, + "loss": 2.3719, + "step": 469 + }, + { + "epoch": 0.27, + "learning_rate": 0.0003, + "loss": 2.3625, + "step": 470 + }, + { + "epoch": 0.27, + "learning_rate": 0.0003, + "loss": 2.3868, + "step": 471 + }, + { + "epoch": 0.27, + "learning_rate": 0.0003, + "loss": 2.3599, + "step": 472 + }, + { + "epoch": 0.27, + "learning_rate": 0.0003, + "loss": 2.3495, + "step": 473 + }, + { + "epoch": 0.27, + "learning_rate": 0.0003, + "loss": 2.3741, + "step": 474 + }, + { + "epoch": 0.27, + "learning_rate": 0.0003, + "loss": 2.3795, + "step": 475 + }, + { + "epoch": 0.27, + "learning_rate": 0.0003, + "loss": 2.3569, + "step": 476 + }, + { + "epoch": 0.27, + "learning_rate": 0.0003, + "loss": 2.3618, + "step": 477 + }, + { + "epoch": 0.27, + "learning_rate": 0.0003, + "loss": 2.3451, + "step": 478 + }, + { + "epoch": 0.27, + "learning_rate": 0.0003, + "loss": 2.3639, + "step": 479 + }, + { + "epoch": 0.28, + "learning_rate": 0.0003, + "loss": 2.3634, + "step": 480 + }, + { + "epoch": 0.28, + "learning_rate": 0.0003, + "loss": 2.373, + "step": 481 + }, + { + "epoch": 0.28, + "learning_rate": 0.0003, + "loss": 2.3545, + "step": 482 + }, + { + "epoch": 0.28, + "learning_rate": 0.0003, + "loss": 2.3627, + "step": 483 + }, + { + "epoch": 0.28, + "learning_rate": 0.0003, + "loss": 2.3563, + "step": 484 + }, + { + "epoch": 0.28, + "learning_rate": 0.0003, + "loss": 2.3549, + "step": 485 + }, + { + "epoch": 0.28, + "learning_rate": 0.0003, + "loss": 2.3586, + "step": 486 + }, + { + "epoch": 0.28, + "learning_rate": 0.0003, + "loss": 2.3488, + "step": 487 + }, + { + "epoch": 0.28, + "learning_rate": 0.0003, + "loss": 2.3784, + "step": 488 + }, + { + "epoch": 0.28, + "learning_rate": 0.0003, + "loss": 2.3632, + "step": 489 + }, + { + "epoch": 0.28, + "learning_rate": 0.0003, + "loss": 2.3544, + "step": 490 + }, + { + "epoch": 0.28, + "learning_rate": 0.0003, + "loss": 2.3571, + "step": 491 + }, + { + "epoch": 0.28, + "learning_rate": 0.0003, + "loss": 2.3568, + "step": 492 + }, + { + "epoch": 0.28, + "learning_rate": 0.0003, + "loss": 2.3539, + "step": 493 + }, + { + "epoch": 0.28, + "learning_rate": 0.0003, + "loss": 2.361, + "step": 494 + }, + { + "epoch": 0.28, + "learning_rate": 0.0003, + "loss": 2.3495, + "step": 495 + }, + { + "epoch": 0.28, + "learning_rate": 0.0003, + "loss": 2.3557, + "step": 496 + }, + { + "epoch": 0.29, + "learning_rate": 0.0003, + "loss": 2.353, + "step": 497 + }, + { + "epoch": 0.29, + "learning_rate": 0.0003, + "loss": 2.3533, + "step": 498 + }, + { + "epoch": 0.29, + "learning_rate": 0.0003, + "loss": 2.3606, + "step": 499 + }, + { + "epoch": 0.29, + "learning_rate": 0.0003, + "loss": 2.3453, + "step": 500 + }, + { + "epoch": 0.29, + "learning_rate": 0.0003, + "loss": 2.3494, + "step": 501 + }, + { + "epoch": 0.29, + "learning_rate": 0.0003, + "loss": 2.3386, + "step": 502 + }, + { + "epoch": 0.29, + "learning_rate": 0.0003, + "loss": 2.3395, + "step": 503 + }, + { + "epoch": 0.29, + "learning_rate": 0.0003, + "loss": 2.3625, + "step": 504 + }, + { + "epoch": 0.29, + "learning_rate": 0.0003, + "loss": 2.3501, + "step": 505 + }, + { + "epoch": 0.29, + "learning_rate": 0.0003, + "loss": 2.3547, + "step": 506 + }, + { + "epoch": 0.29, + "learning_rate": 0.0003, + "loss": 2.3389, + "step": 507 + }, + { + "epoch": 0.29, + "learning_rate": 0.0003, + "loss": 2.3661, + "step": 508 + }, + { + "epoch": 0.29, + "learning_rate": 0.0003, + "loss": 2.3278, + "step": 509 + }, + { + "epoch": 0.29, + "learning_rate": 0.0003, + "loss": 2.3243, + "step": 510 + }, + { + "epoch": 0.29, + "eval_loss": 2.3428287506103516, + "eval_runtime": 2007.442, + "eval_samples_per_second": 52.546, + "eval_steps_per_second": 0.137, + "step": 510 + }, + { + "epoch": 0.29, + "learning_rate": 0.0003, + "loss": 2.3339, + "step": 511 + }, + { + "epoch": 0.29, + "learning_rate": 0.0003, + "loss": 2.3323, + "step": 512 + }, + { + "epoch": 0.29, + "learning_rate": 0.0003, + "loss": 2.3518, + "step": 513 + }, + { + "epoch": 0.29, + "learning_rate": 0.0003, + "loss": 2.3601, + "step": 514 + }, + { + "epoch": 0.3, + "learning_rate": 0.0003, + "loss": 2.337, + "step": 515 + }, + { + "epoch": 0.3, + "learning_rate": 0.0003, + "loss": 2.3232, + "step": 516 + }, + { + "epoch": 0.3, + "learning_rate": 0.0003, + "loss": 2.3405, + "step": 517 + }, + { + "epoch": 0.3, + "learning_rate": 0.0003, + "loss": 2.3528, + "step": 518 + }, + { + "epoch": 0.3, + "learning_rate": 0.0003, + "loss": 2.3308, + "step": 519 + }, + { + "epoch": 0.3, + "learning_rate": 0.0003, + "loss": 2.3506, + "step": 520 + }, + { + "epoch": 0.3, + "learning_rate": 0.0003, + "loss": 2.3463, + "step": 521 + }, + { + "epoch": 0.3, + "learning_rate": 0.0003, + "loss": 2.348, + "step": 522 + }, + { + "epoch": 0.3, + "learning_rate": 0.0003, + "loss": 2.3351, + "step": 523 + }, + { + "epoch": 0.3, + "learning_rate": 0.0003, + "loss": 2.374, + "step": 524 + }, + { + "epoch": 0.3, + "learning_rate": 0.0003, + "loss": 2.3411, + "step": 525 + }, + { + "epoch": 0.3, + "learning_rate": 0.0003, + "loss": 2.3329, + "step": 526 + }, + { + "epoch": 0.3, + "learning_rate": 0.0003, + "loss": 2.3402, + "step": 527 + }, + { + "epoch": 0.3, + "learning_rate": 0.0003, + "loss": 2.3295, + "step": 528 + }, + { + "epoch": 0.3, + "learning_rate": 0.0003, + "loss": 2.3322, + "step": 529 + }, + { + "epoch": 0.3, + "learning_rate": 0.0003, + "loss": 2.3291, + "step": 530 + }, + { + "epoch": 0.3, + "learning_rate": 0.0003, + "loss": 2.3361, + "step": 531 + }, + { + "epoch": 0.31, + "learning_rate": 0.0003, + "loss": 2.328, + "step": 532 + }, + { + "epoch": 0.31, + "learning_rate": 0.0003, + "loss": 2.3279, + "step": 533 + }, + { + "epoch": 0.31, + "learning_rate": 0.0003, + "loss": 2.3265, + "step": 534 + }, + { + "epoch": 0.31, + "learning_rate": 0.0003, + "loss": 2.3455, + "step": 535 + }, + { + "epoch": 0.31, + "learning_rate": 0.0003, + "loss": 2.3464, + "step": 536 + }, + { + "epoch": 0.31, + "learning_rate": 0.0003, + "loss": 2.3759, + "step": 537 + }, + { + "epoch": 0.31, + "learning_rate": 0.0003, + "loss": 2.3312, + "step": 538 + }, + { + "epoch": 0.31, + "learning_rate": 0.0003, + "loss": 2.3274, + "step": 539 + }, + { + "epoch": 0.31, + "learning_rate": 0.0003, + "loss": 2.3552, + "step": 540 + }, + { + "epoch": 0.31, + "learning_rate": 0.0003, + "loss": 2.3341, + "step": 541 + }, + { + "epoch": 0.31, + "learning_rate": 0.0003, + "loss": 2.3223, + "step": 542 + }, + { + "epoch": 0.31, + "learning_rate": 0.0003, + "loss": 2.333, + "step": 543 + }, + { + "epoch": 0.31, + "learning_rate": 0.0003, + "loss": 2.3209, + "step": 544 + }, + { + "epoch": 0.31, + "learning_rate": 0.0003, + "loss": 2.3212, + "step": 545 + }, + { + "epoch": 0.31, + "learning_rate": 0.0003, + "loss": 2.3105, + "step": 546 + }, + { + "epoch": 0.31, + "learning_rate": 0.0003, + "loss": 2.3219, + "step": 547 + }, + { + "epoch": 0.31, + "learning_rate": 0.0003, + "loss": 2.328, + "step": 548 + }, + { + "epoch": 0.31, + "learning_rate": 0.0003, + "loss": 2.3247, + "step": 549 + }, + { + "epoch": 0.32, + "learning_rate": 0.0003, + "loss": 2.3143, + "step": 550 + }, + { + "epoch": 0.32, + "learning_rate": 0.0003, + "loss": 2.3258, + "step": 551 + }, + { + "epoch": 0.32, + "learning_rate": 0.0003, + "loss": 2.3251, + "step": 552 + }, + { + "epoch": 0.32, + "learning_rate": 0.0003, + "loss": 2.3265, + "step": 553 + }, + { + "epoch": 0.32, + "learning_rate": 0.0003, + "loss": 2.3239, + "step": 554 + }, + { + "epoch": 0.32, + "learning_rate": 0.0003, + "loss": 2.3326, + "step": 555 + }, + { + "epoch": 0.32, + "learning_rate": 0.0003, + "loss": 2.2976, + "step": 556 + }, + { + "epoch": 0.32, + "learning_rate": 0.0003, + "loss": 2.3228, + "step": 557 + }, + { + "epoch": 0.32, + "learning_rate": 0.0003, + "loss": 2.3163, + "step": 558 + }, + { + "epoch": 0.32, + "learning_rate": 0.0003, + "loss": 2.3151, + "step": 559 + }, + { + "epoch": 0.32, + "learning_rate": 0.0003, + "loss": 2.314, + "step": 560 + }, + { + "epoch": 0.32, + "learning_rate": 0.0003, + "loss": 2.3284, + "step": 561 + }, + { + "epoch": 0.32, + "learning_rate": 0.0003, + "loss": 2.3224, + "step": 562 + }, + { + "epoch": 0.32, + "learning_rate": 0.0003, + "loss": 2.3317, + "step": 563 + }, + { + "epoch": 0.32, + "learning_rate": 0.0003, + "loss": 2.326, + "step": 564 + }, + { + "epoch": 0.32, + "learning_rate": 0.0003, + "loss": 2.3112, + "step": 565 + }, + { + "epoch": 0.32, + "learning_rate": 0.0003, + "loss": 2.3278, + "step": 566 + }, + { + "epoch": 0.33, + "learning_rate": 0.0003, + "loss": 2.3289, + "step": 567 + }, + { + "epoch": 0.33, + "learning_rate": 0.0003, + "loss": 2.3132, + "step": 568 + }, + { + "epoch": 0.33, + "learning_rate": 0.0003, + "loss": 2.3284, + "step": 569 + }, + { + "epoch": 0.33, + "learning_rate": 0.0003, + "loss": 2.3151, + "step": 570 + }, + { + "epoch": 0.33, + "learning_rate": 0.0003, + "loss": 2.3246, + "step": 571 + }, + { + "epoch": 0.33, + "learning_rate": 0.0003, + "loss": 2.3131, + "step": 572 + }, + { + "epoch": 0.33, + "learning_rate": 0.0003, + "loss": 2.342, + "step": 573 + }, + { + "epoch": 0.33, + "learning_rate": 0.0003, + "loss": 2.3141, + "step": 574 + }, + { + "epoch": 0.33, + "learning_rate": 0.0003, + "loss": 2.3259, + "step": 575 + }, + { + "epoch": 0.33, + "learning_rate": 0.0003, + "loss": 2.3422, + "step": 576 + }, + { + "epoch": 0.33, + "learning_rate": 0.0003, + "loss": 2.3234, + "step": 577 + }, + { + "epoch": 0.33, + "learning_rate": 0.0003, + "loss": 2.3137, + "step": 578 + }, + { + "epoch": 0.33, + "learning_rate": 0.0003, + "loss": 2.3102, + "step": 579 + }, + { + "epoch": 0.33, + "learning_rate": 0.0003, + "loss": 2.3384, + "step": 580 + }, + { + "epoch": 0.33, + "learning_rate": 0.0003, + "loss": 2.2961, + "step": 581 + }, + { + "epoch": 0.33, + "learning_rate": 0.0003, + "loss": 2.319, + "step": 582 + }, + { + "epoch": 0.33, + "learning_rate": 0.0003, + "loss": 2.3129, + "step": 583 + }, + { + "epoch": 0.33, + "learning_rate": 0.0003, + "loss": 2.3274, + "step": 584 + }, + { + "epoch": 0.34, + "learning_rate": 0.0003, + "loss": 2.3319, + "step": 585 + }, + { + "epoch": 0.34, + "learning_rate": 0.0003, + "loss": 2.326, + "step": 586 + }, + { + "epoch": 0.34, + "learning_rate": 0.0003, + "loss": 2.3451, + "step": 587 + }, + { + "epoch": 0.34, + "learning_rate": 0.0003, + "loss": 2.3132, + "step": 588 + }, + { + "epoch": 0.34, + "learning_rate": 0.0003, + "loss": 2.2938, + "step": 589 + }, + { + "epoch": 0.34, + "learning_rate": 0.0003, + "loss": 2.3036, + "step": 590 + }, + { + "epoch": 0.34, + "learning_rate": 0.0003, + "loss": 2.313, + "step": 591 + }, + { + "epoch": 0.34, + "learning_rate": 0.0003, + "loss": 2.2843, + "step": 592 + }, + { + "epoch": 0.34, + "learning_rate": 0.0003, + "loss": 2.2959, + "step": 593 + }, + { + "epoch": 0.34, + "learning_rate": 0.0003, + "loss": 2.3075, + "step": 594 + }, + { + "epoch": 0.34, + "learning_rate": 0.0003, + "loss": 2.3159, + "step": 595 + }, + { + "epoch": 0.34, + "learning_rate": 0.0003, + "loss": 2.286, + "step": 596 + }, + { + "epoch": 0.34, + "learning_rate": 0.0003, + "loss": 2.3069, + "step": 597 + }, + { + "epoch": 0.34, + "learning_rate": 0.0003, + "loss": 2.2968, + "step": 598 + }, + { + "epoch": 0.34, + "learning_rate": 0.0003, + "loss": 2.3096, + "step": 599 + }, + { + "epoch": 0.34, + "learning_rate": 0.0003, + "loss": 2.2999, + "step": 600 + }, + { + "epoch": 0.34, + "learning_rate": 0.0003, + "loss": 2.2992, + "step": 601 + }, + { + "epoch": 0.35, + "learning_rate": 0.0003, + "loss": 2.3005, + "step": 602 + }, + { + "epoch": 0.35, + "learning_rate": 0.0003, + "loss": 2.2998, + "step": 603 + }, + { + "epoch": 0.35, + "learning_rate": 0.0003, + "loss": 2.2793, + "step": 604 + }, + { + "epoch": 0.35, + "learning_rate": 0.0003, + "loss": 2.307, + "step": 605 + }, + { + "epoch": 0.35, + "learning_rate": 0.0003, + "loss": 2.2954, + "step": 606 + }, + { + "epoch": 0.35, + "learning_rate": 0.0003, + "loss": 2.2975, + "step": 607 + }, + { + "epoch": 0.35, + "learning_rate": 0.0003, + "loss": 2.295, + "step": 608 + }, + { + "epoch": 0.35, + "learning_rate": 0.0003, + "loss": 2.3097, + "step": 609 + }, + { + "epoch": 0.35, + "learning_rate": 0.0003, + "loss": 2.2959, + "step": 610 + }, + { + "epoch": 0.35, + "learning_rate": 0.0003, + "loss": 2.3091, + "step": 611 + }, + { + "epoch": 0.35, + "learning_rate": 0.0003, + "loss": 2.3116, + "step": 612 + }, + { + "epoch": 0.35, + "learning_rate": 0.0003, + "loss": 2.309, + "step": 613 + }, + { + "epoch": 0.35, + "learning_rate": 0.0003, + "loss": 2.3115, + "step": 614 + }, + { + "epoch": 0.35, + "learning_rate": 0.0003, + "loss": 2.2852, + "step": 615 + }, + { + "epoch": 0.35, + "learning_rate": 0.0003, + "loss": 2.2999, + "step": 616 + }, + { + "epoch": 0.35, + "learning_rate": 0.0003, + "loss": 2.3208, + "step": 617 + }, + { + "epoch": 0.35, + "learning_rate": 0.0003, + "loss": 2.3148, + "step": 618 + }, + { + "epoch": 0.35, + "learning_rate": 0.0003, + "loss": 2.3085, + "step": 619 + }, + { + "epoch": 0.36, + "learning_rate": 0.0003, + "loss": 2.3037, + "step": 620 + }, + { + "epoch": 0.36, + "learning_rate": 0.0003, + "loss": 2.2906, + "step": 621 + }, + { + "epoch": 0.36, + "learning_rate": 0.0003, + "loss": 2.322, + "step": 622 + }, + { + "epoch": 0.36, + "learning_rate": 0.0003, + "loss": 2.2876, + "step": 623 + }, + { + "epoch": 0.36, + "learning_rate": 0.0003, + "loss": 2.304, + "step": 624 + }, + { + "epoch": 0.36, + "learning_rate": 0.0003, + "loss": 2.2941, + "step": 625 + }, + { + "epoch": 0.36, + "learning_rate": 0.0003, + "loss": 2.3184, + "step": 626 + }, + { + "epoch": 0.36, + "learning_rate": 0.0003, + "loss": 2.3018, + "step": 627 + }, + { + "epoch": 0.36, + "learning_rate": 0.0003, + "loss": 2.3101, + "step": 628 + }, + { + "epoch": 0.36, + "learning_rate": 0.0003, + "loss": 2.2815, + "step": 629 + }, + { + "epoch": 0.36, + "learning_rate": 0.0003, + "loss": 2.2909, + "step": 630 + }, + { + "epoch": 0.36, + "learning_rate": 0.0003, + "loss": 2.2916, + "step": 631 + }, + { + "epoch": 0.36, + "learning_rate": 0.0003, + "loss": 2.285, + "step": 632 + }, + { + "epoch": 0.36, + "learning_rate": 0.0003, + "loss": 2.2926, + "step": 633 + }, + { + "epoch": 0.36, + "learning_rate": 0.0003, + "loss": 2.278, + "step": 634 + }, + { + "epoch": 0.36, + "learning_rate": 0.0003, + "loss": 2.2863, + "step": 635 + }, + { + "epoch": 0.36, + "learning_rate": 0.0003, + "loss": 2.2821, + "step": 636 + }, + { + "epoch": 0.37, + "learning_rate": 0.0003, + "loss": 2.3015, + "step": 637 + }, + { + "epoch": 0.37, + "learning_rate": 0.0003, + "loss": 2.2793, + "step": 638 + }, + { + "epoch": 0.37, + "learning_rate": 0.0003, + "loss": 2.2968, + "step": 639 + }, + { + "epoch": 0.37, + "learning_rate": 0.0003, + "loss": 2.2869, + "step": 640 + }, + { + "epoch": 0.37, + "learning_rate": 0.0003, + "loss": 2.2681, + "step": 641 + }, + { + "epoch": 0.37, + "learning_rate": 0.0003, + "loss": 2.2764, + "step": 642 + }, + { + "epoch": 0.37, + "learning_rate": 0.0003, + "loss": 2.3, + "step": 643 + }, + { + "epoch": 0.37, + "learning_rate": 0.0003, + "loss": 2.299, + "step": 644 + }, + { + "epoch": 0.37, + "learning_rate": 0.0003, + "loss": 2.2882, + "step": 645 + }, + { + "epoch": 0.37, + "learning_rate": 0.0003, + "loss": 2.2856, + "step": 646 + }, + { + "epoch": 0.37, + "learning_rate": 0.0003, + "loss": 2.2948, + "step": 647 + }, + { + "epoch": 0.37, + "learning_rate": 0.0003, + "loss": 2.2754, + "step": 648 + }, + { + "epoch": 0.37, + "learning_rate": 0.0003, + "loss": 2.2843, + "step": 649 + }, + { + "epoch": 0.37, + "learning_rate": 0.0003, + "loss": 2.3031, + "step": 650 + }, + { + "epoch": 0.37, + "learning_rate": 0.0003, + "loss": 2.2955, + "step": 651 + }, + { + "epoch": 0.37, + "learning_rate": 0.0003, + "loss": 2.3, + "step": 652 + }, + { + "epoch": 0.37, + "learning_rate": 0.0003, + "loss": 2.3018, + "step": 653 + }, + { + "epoch": 0.38, + "learning_rate": 0.0003, + "loss": 2.2838, + "step": 654 + }, + { + "epoch": 0.38, + "learning_rate": 0.0003, + "loss": 2.2784, + "step": 655 + }, + { + "epoch": 0.38, + "learning_rate": 0.0003, + "loss": 2.2897, + "step": 656 + }, + { + "epoch": 0.38, + "learning_rate": 0.0003, + "loss": 2.2949, + "step": 657 + }, + { + "epoch": 0.38, + "learning_rate": 0.0003, + "loss": 2.2931, + "step": 658 + }, + { + "epoch": 0.38, + "learning_rate": 0.0003, + "loss": 2.2812, + "step": 659 + }, + { + "epoch": 0.38, + "learning_rate": 0.0003, + "loss": 2.2845, + "step": 660 + }, + { + "epoch": 0.38, + "learning_rate": 0.0003, + "loss": 2.2828, + "step": 661 + }, + { + "epoch": 0.38, + "learning_rate": 0.0003, + "loss": 2.2861, + "step": 662 + }, + { + "epoch": 0.38, + "learning_rate": 0.0003, + "loss": 2.2935, + "step": 663 + }, + { + "epoch": 0.38, + "learning_rate": 0.0003, + "loss": 2.2919, + "step": 664 + }, + { + "epoch": 0.38, + "learning_rate": 0.0003, + "loss": 2.3018, + "step": 665 + }, + { + "epoch": 0.38, + "learning_rate": 0.0003, + "loss": 2.2721, + "step": 666 + }, + { + "epoch": 0.38, + "learning_rate": 0.0003, + "loss": 2.286, + "step": 667 + }, + { + "epoch": 0.38, + "learning_rate": 0.0003, + "loss": 2.2947, + "step": 668 + }, + { + "epoch": 0.38, + "learning_rate": 0.0003, + "loss": 2.2906, + "step": 669 + }, + { + "epoch": 0.38, + "learning_rate": 0.0003, + "loss": 2.2777, + "step": 670 + }, + { + "epoch": 0.38, + "learning_rate": 0.0003, + "loss": 2.2858, + "step": 671 + }, + { + "epoch": 0.39, + "learning_rate": 0.0003, + "loss": 2.2809, + "step": 672 + }, + { + "epoch": 0.39, + "learning_rate": 0.0003, + "loss": 2.2786, + "step": 673 + }, + { + "epoch": 0.39, + "learning_rate": 0.0003, + "loss": 2.2618, + "step": 674 + }, + { + "epoch": 0.39, + "learning_rate": 0.0003, + "loss": 2.2787, + "step": 675 + }, + { + "epoch": 0.39, + "learning_rate": 0.0003, + "loss": 2.2711, + "step": 676 + }, + { + "epoch": 0.39, + "learning_rate": 0.0003, + "loss": 2.2666, + "step": 677 + }, + { + "epoch": 0.39, + "learning_rate": 0.0003, + "loss": 2.2889, + "step": 678 + }, + { + "epoch": 0.39, + "learning_rate": 0.0003, + "loss": 2.2828, + "step": 679 + }, + { + "epoch": 0.39, + "learning_rate": 0.0003, + "loss": 2.2822, + "step": 680 + }, + { + "epoch": 0.39, + "eval_loss": 2.2751543521881104, + "eval_runtime": 2014.3631, + "eval_samples_per_second": 52.366, + "eval_steps_per_second": 0.137, + "step": 680 + }, + { + "epoch": 0.39, + "learning_rate": 0.0003, + "loss": 2.2916, + "step": 681 + }, + { + "epoch": 0.39, + "learning_rate": 0.0003, + "loss": 2.2844, + "step": 682 + }, + { + "epoch": 0.39, + "learning_rate": 0.0003, + "loss": 2.2699, + "step": 683 + }, + { + "epoch": 0.39, + "learning_rate": 0.0003, + "loss": 2.278, + "step": 684 + }, + { + "epoch": 0.39, + "learning_rate": 0.0003, + "loss": 2.2573, + "step": 685 + }, + { + "epoch": 0.39, + "learning_rate": 0.0003, + "loss": 2.2841, + "step": 686 + }, + { + "epoch": 0.39, + "learning_rate": 0.0003, + "loss": 2.2734, + "step": 687 + }, + { + "epoch": 0.39, + "learning_rate": 0.0003, + "loss": 2.2708, + "step": 688 + }, + { + "epoch": 0.4, + "learning_rate": 0.0003, + "loss": 2.259, + "step": 689 + }, + { + "epoch": 0.4, + "learning_rate": 0.0003, + "loss": 2.2826, + "step": 690 + }, + { + "epoch": 0.4, + "learning_rate": 0.0003, + "loss": 2.2611, + "step": 691 + }, + { + "epoch": 0.4, + "learning_rate": 0.0003, + "loss": 2.2892, + "step": 692 + }, + { + "epoch": 0.4, + "learning_rate": 0.0003, + "loss": 2.2803, + "step": 693 + }, + { + "epoch": 0.4, + "learning_rate": 0.0003, + "loss": 2.2734, + "step": 694 + }, + { + "epoch": 0.4, + "learning_rate": 0.0003, + "loss": 2.2844, + "step": 695 + }, + { + "epoch": 0.4, + "learning_rate": 0.0003, + "loss": 2.2574, + "step": 696 + }, + { + "epoch": 0.4, + "learning_rate": 0.0003, + "loss": 2.2601, + "step": 697 + }, + { + "epoch": 0.4, + "learning_rate": 0.0003, + "loss": 2.2709, + "step": 698 + }, + { + "epoch": 0.4, + "learning_rate": 0.0003, + "loss": 2.2482, + "step": 699 + }, + { + "epoch": 0.4, + "learning_rate": 0.0003, + "loss": 2.2839, + "step": 700 + }, + { + "epoch": 0.4, + "learning_rate": 0.0003, + "loss": 2.2665, + "step": 701 + }, + { + "epoch": 0.4, + "learning_rate": 0.0003, + "loss": 2.2628, + "step": 702 + }, + { + "epoch": 0.4, + "learning_rate": 0.0003, + "loss": 2.2653, + "step": 703 + }, + { + "epoch": 0.4, + "learning_rate": 0.0003, + "loss": 2.277, + "step": 704 + }, + { + "epoch": 0.4, + "learning_rate": 0.0003, + "loss": 2.2539, + "step": 705 + }, + { + "epoch": 0.4, + "learning_rate": 0.0003, + "loss": 2.29, + "step": 706 + }, + { + "epoch": 0.41, + "learning_rate": 0.0003, + "loss": 2.2732, + "step": 707 + }, + { + "epoch": 0.41, + "learning_rate": 0.0003, + "loss": 2.2755, + "step": 708 + }, + { + "epoch": 0.41, + "learning_rate": 0.0003, + "loss": 2.278, + "step": 709 + }, + { + "epoch": 0.41, + "learning_rate": 0.0003, + "loss": 2.2611, + "step": 710 + }, + { + "epoch": 0.41, + "learning_rate": 0.0003, + "loss": 2.2623, + "step": 711 + }, + { + "epoch": 0.41, + "learning_rate": 0.0003, + "loss": 2.2654, + "step": 712 + }, + { + "epoch": 0.41, + "learning_rate": 0.0003, + "loss": 2.2521, + "step": 713 + }, + { + "epoch": 0.41, + "learning_rate": 0.0003, + "loss": 2.2662, + "step": 714 + }, + { + "epoch": 0.41, + "learning_rate": 0.0003, + "loss": 2.2488, + "step": 715 + }, + { + "epoch": 0.41, + "learning_rate": 0.0003, + "loss": 2.2719, + "step": 716 + }, + { + "epoch": 0.41, + "learning_rate": 0.0003, + "loss": 2.2655, + "step": 717 + }, + { + "epoch": 0.41, + "learning_rate": 0.0003, + "loss": 2.2649, + "step": 718 + }, + { + "epoch": 0.41, + "learning_rate": 0.0003, + "loss": 2.2939, + "step": 719 + }, + { + "epoch": 0.41, + "learning_rate": 0.0003, + "loss": 2.2506, + "step": 720 + }, + { + "epoch": 0.41, + "learning_rate": 0.0003, + "loss": 2.2518, + "step": 721 + }, + { + "epoch": 0.41, + "learning_rate": 0.0003, + "loss": 2.2599, + "step": 722 + }, + { + "epoch": 0.41, + "learning_rate": 0.0003, + "loss": 2.2833, + "step": 723 + }, + { + "epoch": 0.42, + "learning_rate": 0.0003, + "loss": 2.2548, + "step": 724 + }, + { + "epoch": 0.42, + "learning_rate": 0.0003, + "loss": 2.2739, + "step": 725 + }, + { + "epoch": 0.42, + "learning_rate": 0.0003, + "loss": 2.2672, + "step": 726 + }, + { + "epoch": 0.42, + "learning_rate": 0.0003, + "loss": 2.2714, + "step": 727 + }, + { + "epoch": 0.42, + "learning_rate": 0.0003, + "loss": 2.2512, + "step": 728 + }, + { + "epoch": 0.42, + "learning_rate": 0.0003, + "loss": 2.2618, + "step": 729 + }, + { + "epoch": 0.42, + "learning_rate": 0.0003, + "loss": 2.2798, + "step": 730 + }, + { + "epoch": 0.42, + "learning_rate": 0.0003, + "loss": 2.2573, + "step": 731 + }, + { + "epoch": 0.42, + "learning_rate": 0.0003, + "loss": 2.266, + "step": 732 + }, + { + "epoch": 0.42, + "learning_rate": 0.0003, + "loss": 2.2705, + "step": 733 + }, + { + "epoch": 0.42, + "learning_rate": 0.0003, + "loss": 2.2479, + "step": 734 + }, + { + "epoch": 0.42, + "learning_rate": 0.0003, + "loss": 2.2515, + "step": 735 + }, + { + "epoch": 0.42, + "learning_rate": 0.0003, + "loss": 2.2687, + "step": 736 + }, + { + "epoch": 0.42, + "learning_rate": 0.0003, + "loss": 2.2543, + "step": 737 + }, + { + "epoch": 0.42, + "learning_rate": 0.0003, + "loss": 2.2568, + "step": 738 + }, + { + "epoch": 0.42, + "learning_rate": 0.0003, + "loss": 2.2557, + "step": 739 + }, + { + "epoch": 0.42, + "learning_rate": 0.0003, + "loss": 2.2491, + "step": 740 + }, + { + "epoch": 0.42, + "learning_rate": 0.0003, + "loss": 2.2644, + "step": 741 + }, + { + "epoch": 0.43, + "learning_rate": 0.0003, + "loss": 2.2533, + "step": 742 + }, + { + "epoch": 0.43, + "learning_rate": 0.0003, + "loss": 2.2574, + "step": 743 + }, + { + "epoch": 0.43, + "learning_rate": 0.0003, + "loss": 2.26, + "step": 744 + }, + { + "epoch": 0.43, + "learning_rate": 0.0003, + "loss": 2.2488, + "step": 745 + }, + { + "epoch": 0.43, + "learning_rate": 0.0003, + "loss": 2.2637, + "step": 746 + }, + { + "epoch": 0.43, + "learning_rate": 0.0003, + "loss": 2.2734, + "step": 747 + }, + { + "epoch": 0.43, + "learning_rate": 0.0003, + "loss": 2.2402, + "step": 748 + }, + { + "epoch": 0.43, + "learning_rate": 0.0003, + "loss": 2.2523, + "step": 749 + }, + { + "epoch": 0.43, + "learning_rate": 0.0003, + "loss": 2.2538, + "step": 750 + }, + { + "epoch": 0.43, + "learning_rate": 0.0003, + "loss": 2.269, + "step": 751 + }, + { + "epoch": 0.43, + "learning_rate": 0.0003, + "loss": 2.2437, + "step": 752 + }, + { + "epoch": 0.43, + "learning_rate": 0.0003, + "loss": 2.2529, + "step": 753 + }, + { + "epoch": 0.43, + "learning_rate": 0.0003, + "loss": 2.2593, + "step": 754 + }, + { + "epoch": 0.43, + "learning_rate": 0.0003, + "loss": 2.2667, + "step": 755 + }, + { + "epoch": 0.43, + "learning_rate": 0.0003, + "loss": 2.2289, + "step": 756 + }, + { + "epoch": 0.43, + "learning_rate": 0.0003, + "loss": 2.2596, + "step": 757 + }, + { + "epoch": 0.43, + "learning_rate": 0.0003, + "loss": 2.2618, + "step": 758 + }, + { + "epoch": 0.44, + "learning_rate": 0.0003, + "loss": 2.2653, + "step": 759 + }, + { + "epoch": 0.44, + "learning_rate": 0.0003, + "loss": 2.2559, + "step": 760 + }, + { + "epoch": 0.44, + "learning_rate": 0.0003, + "loss": 2.2466, + "step": 761 + }, + { + "epoch": 0.44, + "learning_rate": 0.0003, + "loss": 2.2578, + "step": 762 + }, + { + "epoch": 0.44, + "learning_rate": 0.0003, + "loss": 2.2527, + "step": 763 + }, + { + "epoch": 0.44, + "learning_rate": 0.0003, + "loss": 2.251, + "step": 764 + }, + { + "epoch": 0.44, + "learning_rate": 0.0003, + "loss": 2.246, + "step": 765 + }, + { + "epoch": 0.44, + "learning_rate": 0.0003, + "loss": 2.252, + "step": 766 + }, + { + "epoch": 0.44, + "learning_rate": 0.0003, + "loss": 2.2423, + "step": 767 + }, + { + "epoch": 0.44, + "learning_rate": 0.0003, + "loss": 2.265, + "step": 768 + }, + { + "epoch": 0.44, + "learning_rate": 0.0003, + "loss": 2.2549, + "step": 769 + }, + { + "epoch": 0.44, + "learning_rate": 0.0003, + "loss": 2.262, + "step": 770 + }, + { + "epoch": 0.44, + "learning_rate": 0.0003, + "loss": 2.2422, + "step": 771 + }, + { + "epoch": 0.44, + "learning_rate": 0.0003, + "loss": 2.2504, + "step": 772 + }, + { + "epoch": 0.44, + "learning_rate": 0.0003, + "loss": 2.2375, + "step": 773 + }, + { + "epoch": 0.44, + "learning_rate": 0.0003, + "loss": 2.2389, + "step": 774 + }, + { + "epoch": 0.44, + "learning_rate": 0.0003, + "loss": 2.2621, + "step": 775 + }, + { + "epoch": 0.44, + "learning_rate": 0.0003, + "loss": 2.2451, + "step": 776 + }, + { + "epoch": 0.45, + "learning_rate": 0.0003, + "loss": 2.2588, + "step": 777 + }, + { + "epoch": 0.45, + "learning_rate": 0.0003, + "loss": 2.2657, + "step": 778 + }, + { + "epoch": 0.45, + "learning_rate": 0.0003, + "loss": 2.2357, + "step": 779 + }, + { + "epoch": 0.45, + "learning_rate": 0.0003, + "loss": 2.2344, + "step": 780 + }, + { + "epoch": 0.45, + "learning_rate": 0.0003, + "loss": 2.2294, + "step": 781 + }, + { + "epoch": 0.45, + "learning_rate": 0.0003, + "loss": 2.2535, + "step": 782 + }, + { + "epoch": 0.45, + "learning_rate": 0.0003, + "loss": 2.2629, + "step": 783 + }, + { + "epoch": 0.45, + "learning_rate": 0.0003, + "loss": 2.2329, + "step": 784 + }, + { + "epoch": 0.45, + "learning_rate": 0.0003, + "loss": 2.2561, + "step": 785 + }, + { + "epoch": 0.45, + "learning_rate": 0.0003, + "loss": 2.2646, + "step": 786 + }, + { + "epoch": 0.45, + "learning_rate": 0.0003, + "loss": 2.2241, + "step": 787 + }, + { + "epoch": 0.45, + "learning_rate": 0.0003, + "loss": 2.2383, + "step": 788 + }, + { + "epoch": 0.45, + "learning_rate": 0.0003, + "loss": 2.2494, + "step": 789 + }, + { + "epoch": 0.45, + "learning_rate": 0.0003, + "loss": 2.257, + "step": 790 + }, + { + "epoch": 0.45, + "learning_rate": 0.0003, + "loss": 2.2465, + "step": 791 + }, + { + "epoch": 0.45, + "learning_rate": 0.0003, + "loss": 2.246, + "step": 792 + }, + { + "epoch": 0.45, + "learning_rate": 0.0003, + "loss": 2.2407, + "step": 793 + }, + { + "epoch": 0.46, + "learning_rate": 0.0003, + "loss": 2.2216, + "step": 794 + }, + { + "epoch": 0.46, + "learning_rate": 0.0003, + "loss": 2.2418, + "step": 795 + }, + { + "epoch": 0.46, + "learning_rate": 0.0003, + "loss": 2.2448, + "step": 796 + }, + { + "epoch": 0.46, + "learning_rate": 0.0003, + "loss": 2.2449, + "step": 797 + }, + { + "epoch": 0.46, + "learning_rate": 0.0003, + "loss": 2.2349, + "step": 798 + }, + { + "epoch": 0.46, + "learning_rate": 0.0003, + "loss": 2.2565, + "step": 799 + }, + { + "epoch": 0.46, + "learning_rate": 0.0003, + "loss": 2.2384, + "step": 800 + }, + { + "epoch": 0.46, + "learning_rate": 0.0003, + "loss": 2.2604, + "step": 801 + }, + { + "epoch": 0.46, + "learning_rate": 0.0003, + "loss": 2.2479, + "step": 802 + }, + { + "epoch": 0.46, + "learning_rate": 0.0003, + "loss": 2.2495, + "step": 803 + }, + { + "epoch": 0.46, + "learning_rate": 0.0003, + "loss": 2.2511, + "step": 804 + }, + { + "epoch": 0.46, + "learning_rate": 0.0003, + "loss": 2.2458, + "step": 805 + }, + { + "epoch": 0.46, + "learning_rate": 0.0003, + "loss": 2.2339, + "step": 806 + }, + { + "epoch": 0.46, + "learning_rate": 0.0003, + "loss": 2.2513, + "step": 807 + }, + { + "epoch": 0.46, + "learning_rate": 0.0003, + "loss": 2.2242, + "step": 808 + }, + { + "epoch": 0.46, + "learning_rate": 0.0003, + "loss": 2.2437, + "step": 809 + }, + { + "epoch": 0.46, + "learning_rate": 0.0003, + "loss": 2.2465, + "step": 810 + }, + { + "epoch": 0.47, + "learning_rate": 0.0003, + "loss": 2.2437, + "step": 811 + }, + { + "epoch": 0.47, + "learning_rate": 0.0003, + "loss": 2.2095, + "step": 812 + }, + { + "epoch": 0.47, + "learning_rate": 0.0003, + "loss": 2.2333, + "step": 813 + }, + { + "epoch": 0.47, + "learning_rate": 0.0003, + "loss": 2.24, + "step": 814 + }, + { + "epoch": 0.47, + "learning_rate": 0.0003, + "loss": 2.2277, + "step": 815 + }, + { + "epoch": 0.47, + "learning_rate": 0.0003, + "loss": 2.2426, + "step": 816 + }, + { + "epoch": 0.47, + "learning_rate": 0.0003, + "loss": 2.2405, + "step": 817 + }, + { + "epoch": 0.47, + "learning_rate": 0.0003, + "loss": 2.229, + "step": 818 + }, + { + "epoch": 0.47, + "learning_rate": 0.0003, + "loss": 2.2419, + "step": 819 + }, + { + "epoch": 0.47, + "learning_rate": 0.0003, + "loss": 2.2358, + "step": 820 + }, + { + "epoch": 0.47, + "learning_rate": 0.0003, + "loss": 2.2295, + "step": 821 + }, + { + "epoch": 0.47, + "learning_rate": 0.0003, + "loss": 2.2151, + "step": 822 + }, + { + "epoch": 0.47, + "learning_rate": 0.0003, + "loss": 2.235, + "step": 823 + }, + { + "epoch": 0.47, + "learning_rate": 0.0003, + "loss": 2.2235, + "step": 824 + }, + { + "epoch": 0.47, + "learning_rate": 0.0003, + "loss": 2.2402, + "step": 825 + }, + { + "epoch": 0.47, + "learning_rate": 0.0003, + "loss": 2.2185, + "step": 826 + }, + { + "epoch": 0.47, + "learning_rate": 0.0003, + "loss": 2.2264, + "step": 827 + }, + { + "epoch": 0.47, + "learning_rate": 0.0003, + "loss": 2.2351, + "step": 828 + }, + { + "epoch": 0.48, + "learning_rate": 0.0003, + "loss": 2.2502, + "step": 829 + }, + { + "epoch": 0.48, + "learning_rate": 0.0003, + "loss": 2.2166, + "step": 830 + }, + { + "epoch": 0.48, + "learning_rate": 0.0003, + "loss": 2.2478, + "step": 831 + }, + { + "epoch": 0.48, + "learning_rate": 0.0003, + "loss": 2.2455, + "step": 832 + }, + { + "epoch": 0.48, + "learning_rate": 0.0003, + "loss": 2.241, + "step": 833 + }, + { + "epoch": 0.48, + "learning_rate": 0.0003, + "loss": 2.2488, + "step": 834 + }, + { + "epoch": 0.48, + "learning_rate": 0.0003, + "loss": 2.2396, + "step": 835 + }, + { + "epoch": 0.48, + "learning_rate": 0.0003, + "loss": 2.2479, + "step": 836 + }, + { + "epoch": 0.48, + "learning_rate": 0.0003, + "loss": 2.2037, + "step": 837 + }, + { + "epoch": 0.48, + "learning_rate": 0.0003, + "loss": 2.2253, + "step": 838 + }, + { + "epoch": 0.48, + "learning_rate": 0.0003, + "loss": 2.2191, + "step": 839 + }, + { + "epoch": 0.48, + "learning_rate": 0.0003, + "loss": 2.225, + "step": 840 + }, + { + "epoch": 0.48, + "learning_rate": 0.0003, + "loss": 2.254, + "step": 841 + }, + { + "epoch": 0.48, + "learning_rate": 0.0003, + "loss": 2.2063, + "step": 842 + }, + { + "epoch": 0.48, + "learning_rate": 0.0003, + "loss": 2.2079, + "step": 843 + }, + { + "epoch": 0.48, + "learning_rate": 0.0003, + "loss": 2.2384, + "step": 844 + }, + { + "epoch": 0.48, + "learning_rate": 0.0003, + "loss": 2.2308, + "step": 845 + }, + { + "epoch": 0.49, + "learning_rate": 0.0003, + "loss": 2.2521, + "step": 846 + }, + { + "epoch": 0.49, + "learning_rate": 0.0003, + "loss": 2.2316, + "step": 847 + }, + { + "epoch": 0.49, + "learning_rate": 0.0003, + "loss": 2.2097, + "step": 848 + }, + { + "epoch": 0.49, + "learning_rate": 0.0003, + "loss": 2.2209, + "step": 849 + }, + { + "epoch": 0.49, + "learning_rate": 0.0003, + "loss": 2.238, + "step": 850 + }, + { + "epoch": 0.49, + "eval_loss": 2.2248148918151855, + "eval_runtime": 2006.6082, + "eval_samples_per_second": 52.568, + "eval_steps_per_second": 0.137, + "step": 850 + }, + { + "epoch": 0.49, + "learning_rate": 0.0003, + "loss": 2.2324, + "step": 851 + }, + { + "epoch": 0.49, + "learning_rate": 0.0003, + "loss": 2.2268, + "step": 852 + }, + { + "epoch": 0.49, + "learning_rate": 0.0003, + "loss": 2.2094, + "step": 853 + }, + { + "epoch": 0.49, + "learning_rate": 0.0003, + "loss": 2.2263, + "step": 854 + }, + { + "epoch": 0.49, + "learning_rate": 0.0003, + "loss": 2.2231, + "step": 855 + }, + { + "epoch": 0.49, + "learning_rate": 0.0003, + "loss": 2.2177, + "step": 856 + }, + { + "epoch": 0.49, + "learning_rate": 0.0003, + "loss": 2.2247, + "step": 857 + }, + { + "epoch": 0.49, + "learning_rate": 0.0003, + "loss": 2.2296, + "step": 858 + }, + { + "epoch": 0.49, + "learning_rate": 0.0003, + "loss": 2.224, + "step": 859 + }, + { + "epoch": 0.49, + "learning_rate": 0.0003, + "loss": 2.2169, + "step": 860 + }, + { + "epoch": 0.49, + "learning_rate": 0.0003, + "loss": 2.231, + "step": 861 + }, + { + "epoch": 0.49, + "learning_rate": 0.0003, + "loss": 2.2363, + "step": 862 + }, + { + "epoch": 0.49, + "learning_rate": 0.0003, + "loss": 2.2305, + "step": 863 + }, + { + "epoch": 0.5, + "learning_rate": 0.0003, + "loss": 2.2185, + "step": 864 + }, + { + "epoch": 0.5, + "learning_rate": 0.0003, + "loss": 2.2297, + "step": 865 + }, + { + "epoch": 0.5, + "learning_rate": 0.0003, + "loss": 2.2218, + "step": 866 + }, + { + "epoch": 0.5, + "learning_rate": 0.0003, + "loss": 2.2245, + "step": 867 + }, + { + "epoch": 0.5, + "learning_rate": 0.0003, + "loss": 2.2233, + "step": 868 + }, + { + "epoch": 0.5, + "learning_rate": 0.0003, + "loss": 2.2303, + "step": 869 + }, + { + "epoch": 0.5, + "learning_rate": 0.0003, + "loss": 2.2325, + "step": 870 + }, + { + "epoch": 0.5, + "learning_rate": 0.0003, + "loss": 2.2305, + "step": 871 + }, + { + "epoch": 0.5, + "learning_rate": 0.0003, + "loss": 2.2245, + "step": 872 + }, + { + "epoch": 0.5, + "learning_rate": 0.0003, + "loss": 2.2334, + "step": 873 + }, + { + "epoch": 0.5, + "learning_rate": 0.0003, + "loss": 2.2307, + "step": 874 + }, + { + "epoch": 0.5, + "learning_rate": 0.0003, + "loss": 2.2168, + "step": 875 + }, + { + "epoch": 0.5, + "learning_rate": 0.0003, + "loss": 2.2106, + "step": 876 + }, + { + "epoch": 0.5, + "learning_rate": 0.0003, + "loss": 2.2188, + "step": 877 + }, + { + "epoch": 0.5, + "learning_rate": 0.0003, + "loss": 2.2355, + "step": 878 + }, + { + "epoch": 0.5, + "learning_rate": 0.0003, + "loss": 2.203, + "step": 879 + }, + { + "epoch": 0.5, + "learning_rate": 0.0003, + "loss": 2.2531, + "step": 880 + }, + { + "epoch": 0.51, + "learning_rate": 0.0003, + "loss": 2.2169, + "step": 881 + }, + { + "epoch": 0.51, + "learning_rate": 0.0003, + "loss": 2.2305, + "step": 882 + }, + { + "epoch": 0.51, + "learning_rate": 0.0003, + "loss": 2.2246, + "step": 883 + }, + { + "epoch": 0.51, + "learning_rate": 0.0003, + "loss": 2.2072, + "step": 884 + }, + { + "epoch": 0.51, + "learning_rate": 0.0003, + "loss": 2.2203, + "step": 885 + }, + { + "epoch": 0.51, + "learning_rate": 0.0003, + "loss": 2.2188, + "step": 886 + }, + { + "epoch": 0.51, + "learning_rate": 0.0003, + "loss": 2.2301, + "step": 887 + }, + { + "epoch": 0.51, + "learning_rate": 0.0003, + "loss": 2.2252, + "step": 888 + }, + { + "epoch": 0.51, + "learning_rate": 0.0003, + "loss": 2.2152, + "step": 889 + }, + { + "epoch": 0.51, + "learning_rate": 0.0003, + "loss": 2.2321, + "step": 890 + }, + { + "epoch": 0.51, + "learning_rate": 0.0003, + "loss": 2.2007, + "step": 891 + }, + { + "epoch": 0.51, + "learning_rate": 0.0003, + "loss": 2.2191, + "step": 892 + }, + { + "epoch": 0.51, + "learning_rate": 0.0003, + "loss": 2.2114, + "step": 893 + }, + { + "epoch": 0.51, + "learning_rate": 0.0003, + "loss": 2.2131, + "step": 894 + }, + { + "epoch": 0.51, + "learning_rate": 0.0003, + "loss": 2.213, + "step": 895 + }, + { + "epoch": 0.51, + "learning_rate": 0.0003, + "loss": 2.2162, + "step": 896 + }, + { + "epoch": 0.51, + "learning_rate": 0.0003, + "loss": 2.2294, + "step": 897 + }, + { + "epoch": 0.51, + "learning_rate": 0.0003, + "loss": 2.2172, + "step": 898 + }, + { + "epoch": 0.52, + "learning_rate": 0.0003, + "loss": 2.2122, + "step": 899 + }, + { + "epoch": 0.52, + "learning_rate": 0.0003, + "loss": 2.2213, + "step": 900 + }, + { + "epoch": 0.52, + "learning_rate": 0.0003, + "loss": 2.2261, + "step": 901 + }, + { + "epoch": 0.52, + "learning_rate": 0.0003, + "loss": 2.2168, + "step": 902 + }, + { + "epoch": 0.52, + "learning_rate": 0.0003, + "loss": 2.2196, + "step": 903 + }, + { + "epoch": 0.52, + "learning_rate": 0.0003, + "loss": 2.2249, + "step": 904 + }, + { + "epoch": 0.52, + "learning_rate": 0.0003, + "loss": 2.2163, + "step": 905 + }, + { + "epoch": 0.52, + "learning_rate": 0.0003, + "loss": 2.2091, + "step": 906 + }, + { + "epoch": 0.52, + "learning_rate": 0.0003, + "loss": 2.2128, + "step": 907 + }, + { + "epoch": 0.52, + "learning_rate": 0.0003, + "loss": 2.2001, + "step": 908 + }, + { + "epoch": 0.52, + "learning_rate": 0.0003, + "loss": 2.2313, + "step": 909 + }, + { + "epoch": 0.52, + "learning_rate": 0.0003, + "loss": 2.2147, + "step": 910 + }, + { + "epoch": 0.52, + "learning_rate": 0.0003, + "loss": 2.2106, + "step": 911 + }, + { + "epoch": 0.52, + "learning_rate": 0.0003, + "loss": 2.2274, + "step": 912 + }, + { + "epoch": 0.52, + "learning_rate": 0.0003, + "loss": 2.2057, + "step": 913 + }, + { + "epoch": 0.52, + "learning_rate": 0.0003, + "loss": 2.2076, + "step": 914 + }, + { + "epoch": 0.52, + "learning_rate": 0.0003, + "loss": 2.2257, + "step": 915 + }, + { + "epoch": 0.53, + "learning_rate": 0.0003, + "loss": 2.2213, + "step": 916 + }, + { + "epoch": 0.53, + "learning_rate": 0.0003, + "loss": 2.2119, + "step": 917 + }, + { + "epoch": 0.53, + "learning_rate": 0.0003, + "loss": 2.214, + "step": 918 + }, + { + "epoch": 0.53, + "learning_rate": 0.0003, + "loss": 2.2162, + "step": 919 + }, + { + "epoch": 0.53, + "learning_rate": 0.0003, + "loss": 2.2104, + "step": 920 + }, + { + "epoch": 0.53, + "learning_rate": 0.0003, + "loss": 2.2079, + "step": 921 + }, + { + "epoch": 0.53, + "learning_rate": 0.0003, + "loss": 2.2284, + "step": 922 + }, + { + "epoch": 0.53, + "learning_rate": 0.0003, + "loss": 2.2316, + "step": 923 + }, + { + "epoch": 0.53, + "learning_rate": 0.0003, + "loss": 2.2042, + "step": 924 + }, + { + "epoch": 0.53, + "learning_rate": 0.0003, + "loss": 2.2241, + "step": 925 + }, + { + "epoch": 0.53, + "learning_rate": 0.0003, + "loss": 2.2046, + "step": 926 + }, + { + "epoch": 0.53, + "learning_rate": 0.0003, + "loss": 2.2008, + "step": 927 + }, + { + "epoch": 0.53, + "learning_rate": 0.0003, + "loss": 2.2146, + "step": 928 + }, + { + "epoch": 0.53, + "learning_rate": 0.0003, + "loss": 2.2146, + "step": 929 + }, + { + "epoch": 0.53, + "learning_rate": 0.0003, + "loss": 2.204, + "step": 930 + }, + { + "epoch": 0.53, + "learning_rate": 0.0003, + "loss": 2.2316, + "step": 931 + }, + { + "epoch": 0.53, + "learning_rate": 0.0003, + "loss": 2.2041, + "step": 932 + }, + { + "epoch": 0.54, + "learning_rate": 0.0003, + "loss": 2.2141, + "step": 933 + }, + { + "epoch": 0.54, + "learning_rate": 0.0003, + "loss": 2.2178, + "step": 934 + }, + { + "epoch": 0.54, + "learning_rate": 0.0003, + "loss": 2.2058, + "step": 935 + }, + { + "epoch": 0.54, + "learning_rate": 0.0003, + "loss": 2.2068, + "step": 936 + }, + { + "epoch": 0.54, + "learning_rate": 0.0003, + "loss": 2.2159, + "step": 937 + }, + { + "epoch": 0.54, + "learning_rate": 0.0003, + "loss": 2.2108, + "step": 938 + }, + { + "epoch": 0.54, + "learning_rate": 0.0003, + "loss": 2.1992, + "step": 939 + }, + { + "epoch": 0.54, + "learning_rate": 0.0003, + "loss": 2.2147, + "step": 940 + }, + { + "epoch": 0.54, + "learning_rate": 0.0003, + "loss": 2.2058, + "step": 941 + }, + { + "epoch": 0.54, + "learning_rate": 0.0003, + "loss": 2.2124, + "step": 942 + }, + { + "epoch": 0.54, + "learning_rate": 0.0003, + "loss": 2.1974, + "step": 943 + }, + { + "epoch": 0.54, + "learning_rate": 0.0003, + "loss": 2.2003, + "step": 944 + }, + { + "epoch": 0.54, + "learning_rate": 0.0003, + "loss": 2.188, + "step": 945 + }, + { + "epoch": 0.54, + "learning_rate": 0.0003, + "loss": 2.2008, + "step": 946 + }, + { + "epoch": 0.54, + "learning_rate": 0.0003, + "loss": 2.1968, + "step": 947 + }, + { + "epoch": 0.54, + "learning_rate": 0.0003, + "loss": 2.2098, + "step": 948 + }, + { + "epoch": 0.54, + "learning_rate": 0.0003, + "loss": 2.1981, + "step": 949 + }, + { + "epoch": 0.54, + "learning_rate": 0.0003, + "loss": 2.2214, + "step": 950 + }, + { + "epoch": 0.55, + "learning_rate": 0.0003, + "loss": 2.2, + "step": 951 + }, + { + "epoch": 0.55, + "learning_rate": 0.0003, + "loss": 2.2104, + "step": 952 + }, + { + "epoch": 0.55, + "learning_rate": 0.0003, + "loss": 2.2129, + "step": 953 + }, + { + "epoch": 0.55, + "learning_rate": 0.0003, + "loss": 2.2043, + "step": 954 + }, + { + "epoch": 0.55, + "learning_rate": 0.0003, + "loss": 2.2051, + "step": 955 + }, + { + "epoch": 0.55, + "learning_rate": 0.0003, + "loss": 2.1974, + "step": 956 + }, + { + "epoch": 0.55, + "learning_rate": 0.0003, + "loss": 2.1943, + "step": 957 + }, + { + "epoch": 0.55, + "learning_rate": 0.0003, + "loss": 2.1964, + "step": 958 + }, + { + "epoch": 0.55, + "learning_rate": 0.0003, + "loss": 2.1936, + "step": 959 + }, + { + "epoch": 0.55, + "learning_rate": 0.0003, + "loss": 2.2017, + "step": 960 + }, + { + "epoch": 0.55, + "learning_rate": 0.0003, + "loss": 2.1963, + "step": 961 + }, + { + "epoch": 0.55, + "learning_rate": 0.0003, + "loss": 2.2084, + "step": 962 + }, + { + "epoch": 0.55, + "learning_rate": 0.0003, + "loss": 2.2195, + "step": 963 + }, + { + "epoch": 0.55, + "learning_rate": 0.0003, + "loss": 2.1926, + "step": 964 + }, + { + "epoch": 0.55, + "learning_rate": 0.0003, + "loss": 2.1805, + "step": 965 + }, + { + "epoch": 0.55, + "learning_rate": 0.0003, + "loss": 2.2206, + "step": 966 + }, + { + "epoch": 0.55, + "learning_rate": 0.0003, + "loss": 2.2021, + "step": 967 + }, + { + "epoch": 0.56, + "learning_rate": 0.0003, + "loss": 2.2085, + "step": 968 + }, + { + "epoch": 0.56, + "learning_rate": 0.0003, + "loss": 2.2001, + "step": 969 + }, + { + "epoch": 0.56, + "learning_rate": 0.0003, + "loss": 2.1749, + "step": 970 + }, + { + "epoch": 0.56, + "learning_rate": 0.0003, + "loss": 2.2198, + "step": 971 + }, + { + "epoch": 0.56, + "learning_rate": 0.0003, + "loss": 2.2029, + "step": 972 + }, + { + "epoch": 0.56, + "learning_rate": 0.0003, + "loss": 2.1986, + "step": 973 + }, + { + "epoch": 0.56, + "learning_rate": 0.0003, + "loss": 2.1909, + "step": 974 + }, + { + "epoch": 0.56, + "learning_rate": 0.0003, + "loss": 2.2095, + "step": 975 + }, + { + "epoch": 0.56, + "learning_rate": 0.0003, + "loss": 2.2047, + "step": 976 + }, + { + "epoch": 0.56, + "learning_rate": 0.0003, + "loss": 2.1923, + "step": 977 + }, + { + "epoch": 0.56, + "learning_rate": 0.0003, + "loss": 2.2053, + "step": 978 + }, + { + "epoch": 0.56, + "learning_rate": 0.0003, + "loss": 2.194, + "step": 979 + }, + { + "epoch": 0.56, + "learning_rate": 0.0003, + "loss": 2.1938, + "step": 980 + }, + { + "epoch": 0.56, + "learning_rate": 0.0003, + "loss": 2.1974, + "step": 981 + }, + { + "epoch": 0.56, + "learning_rate": 0.0003, + "loss": 2.198, + "step": 982 + }, + { + "epoch": 0.56, + "learning_rate": 0.0003, + "loss": 2.1976, + "step": 983 + }, + { + "epoch": 0.56, + "learning_rate": 0.0003, + "loss": 2.1987, + "step": 984 + }, + { + "epoch": 0.56, + "learning_rate": 0.0003, + "loss": 2.2135, + "step": 985 + }, + { + "epoch": 0.57, + "learning_rate": 0.0003, + "loss": 2.1895, + "step": 986 + }, + { + "epoch": 0.57, + "learning_rate": 0.0003, + "loss": 2.2025, + "step": 987 + }, + { + "epoch": 0.57, + "learning_rate": 0.0003, + "loss": 2.1804, + "step": 988 + }, + { + "epoch": 0.57, + "learning_rate": 0.0003, + "loss": 2.1938, + "step": 989 + }, + { + "epoch": 0.57, + "learning_rate": 0.0003, + "loss": 2.204, + "step": 990 + }, + { + "epoch": 0.57, + "learning_rate": 0.0003, + "loss": 2.2053, + "step": 991 + }, + { + "epoch": 0.57, + "learning_rate": 0.0003, + "loss": 2.2034, + "step": 992 + }, + { + "epoch": 0.57, + "learning_rate": 0.0003, + "loss": 2.1864, + "step": 993 + }, + { + "epoch": 0.57, + "learning_rate": 0.0003, + "loss": 2.1908, + "step": 994 + }, + { + "epoch": 0.57, + "learning_rate": 0.0003, + "loss": 2.2021, + "step": 995 + }, + { + "epoch": 0.57, + "learning_rate": 0.0003, + "loss": 2.189, + "step": 996 + }, + { + "epoch": 0.57, + "learning_rate": 0.0003, + "loss": 2.2018, + "step": 997 + }, + { + "epoch": 0.57, + "learning_rate": 0.0003, + "loss": 2.198, + "step": 998 + }, + { + "epoch": 0.57, + "learning_rate": 0.0003, + "loss": 2.1946, + "step": 999 + }, + { + "epoch": 0.57, + "learning_rate": 0.0003, + "loss": 2.1886, + "step": 1000 + }, + { + "epoch": 0.57, + "learning_rate": 0.0003, + "loss": 2.1919, + "step": 1001 + }, + { + "epoch": 0.57, + "learning_rate": 0.0003, + "loss": 2.205, + "step": 1002 + }, + { + "epoch": 0.58, + "learning_rate": 0.0003, + "loss": 2.2024, + "step": 1003 + }, + { + "epoch": 0.58, + "learning_rate": 0.0003, + "loss": 2.1835, + "step": 1004 + }, + { + "epoch": 0.58, + "learning_rate": 0.0003, + "loss": 2.2084, + "step": 1005 + }, + { + "epoch": 0.58, + "learning_rate": 0.0003, + "loss": 2.2009, + "step": 1006 + }, + { + "epoch": 0.58, + "learning_rate": 0.0003, + "loss": 2.1963, + "step": 1007 + }, + { + "epoch": 0.58, + "learning_rate": 0.0003, + "loss": 2.2008, + "step": 1008 + }, + { + "epoch": 0.58, + "learning_rate": 0.0003, + "loss": 2.2106, + "step": 1009 + }, + { + "epoch": 0.58, + "learning_rate": 0.0003, + "loss": 2.1888, + "step": 1010 + }, + { + "epoch": 0.58, + "learning_rate": 0.0003, + "loss": 2.1853, + "step": 1011 + }, + { + "epoch": 0.58, + "learning_rate": 0.0003, + "loss": 2.2042, + "step": 1012 + }, + { + "epoch": 0.58, + "learning_rate": 0.0003, + "loss": 2.1677, + "step": 1013 + }, + { + "epoch": 0.58, + "learning_rate": 0.0003, + "loss": 2.193, + "step": 1014 + }, + { + "epoch": 0.58, + "learning_rate": 0.0003, + "loss": 2.2048, + "step": 1015 + }, + { + "epoch": 0.58, + "learning_rate": 0.0003, + "loss": 2.1934, + "step": 1016 + }, + { + "epoch": 0.58, + "learning_rate": 0.0003, + "loss": 2.1863, + "step": 1017 + }, + { + "epoch": 0.58, + "learning_rate": 0.0003, + "loss": 2.2007, + "step": 1018 + }, + { + "epoch": 0.58, + "learning_rate": 0.0003, + "loss": 2.1763, + "step": 1019 + }, + { + "epoch": 0.58, + "learning_rate": 0.0003, + "loss": 2.2015, + "step": 1020 + }, + { + "epoch": 0.58, + "eval_loss": 2.1865060329437256, + "eval_runtime": 2005.632, + "eval_samples_per_second": 52.594, + "eval_steps_per_second": 0.137, + "step": 1020 + }, + { + "epoch": 0.59, + "learning_rate": 0.0003, + "loss": 2.2117, + "step": 1021 + }, + { + "epoch": 0.59, + "learning_rate": 0.0003, + "loss": 2.1819, + "step": 1022 + }, + { + "epoch": 0.59, + "learning_rate": 0.0003, + "loss": 2.1951, + "step": 1023 + }, + { + "epoch": 0.59, + "learning_rate": 0.0003, + "loss": 2.1814, + "step": 1024 + }, + { + "epoch": 0.59, + "learning_rate": 0.0003, + "loss": 2.174, + "step": 1025 + }, + { + "epoch": 0.59, + "learning_rate": 0.0003, + "loss": 2.179, + "step": 1026 + }, + { + "epoch": 0.59, + "learning_rate": 0.0003, + "loss": 2.1925, + "step": 1027 + }, + { + "epoch": 0.59, + "learning_rate": 0.0003, + "loss": 2.1818, + "step": 1028 + }, + { + "epoch": 0.59, + "learning_rate": 0.0003, + "loss": 2.1766, + "step": 1029 + }, + { + "epoch": 0.59, + "learning_rate": 0.0003, + "loss": 2.2008, + "step": 1030 + }, + { + "epoch": 0.59, + "learning_rate": 0.0003, + "loss": 2.1969, + "step": 1031 + }, + { + "epoch": 0.59, + "learning_rate": 0.0003, + "loss": 2.1815, + "step": 1032 + }, + { + "epoch": 0.59, + "learning_rate": 0.0003, + "loss": 2.1796, + "step": 1033 + }, + { + "epoch": 0.59, + "learning_rate": 0.0003, + "loss": 2.1777, + "step": 1034 + }, + { + "epoch": 0.59, + "learning_rate": 0.0003, + "loss": 2.1702, + "step": 1035 + }, + { + "epoch": 0.59, + "learning_rate": 0.0003, + "loss": 2.1946, + "step": 1036 + }, + { + "epoch": 0.59, + "learning_rate": 0.0003, + "loss": 2.1825, + "step": 1037 + }, + { + "epoch": 0.6, + "learning_rate": 0.0003, + "loss": 2.1925, + "step": 1038 + }, + { + "epoch": 0.6, + "learning_rate": 0.0003, + "loss": 2.191, + "step": 1039 + }, + { + "epoch": 0.6, + "learning_rate": 0.0003, + "loss": 2.1859, + "step": 1040 + }, + { + "epoch": 0.6, + "learning_rate": 0.0003, + "loss": 2.1863, + "step": 1041 + }, + { + "epoch": 0.6, + "learning_rate": 0.0003, + "loss": 2.1827, + "step": 1042 + }, + { + "epoch": 0.6, + "learning_rate": 0.0003, + "loss": 2.1848, + "step": 1043 + }, + { + "epoch": 0.6, + "learning_rate": 0.0003, + "loss": 2.1915, + "step": 1044 + }, + { + "epoch": 0.6, + "learning_rate": 0.0003, + "loss": 2.1932, + "step": 1045 + }, + { + "epoch": 0.6, + "learning_rate": 0.0003, + "loss": 2.1859, + "step": 1046 + }, + { + "epoch": 0.6, + "learning_rate": 0.0003, + "loss": 2.1722, + "step": 1047 + }, + { + "epoch": 0.6, + "learning_rate": 0.0003, + "loss": 2.187, + "step": 1048 + }, + { + "epoch": 0.6, + "learning_rate": 0.0003, + "loss": 2.1809, + "step": 1049 + }, + { + "epoch": 0.6, + "learning_rate": 0.0003, + "loss": 2.1846, + "step": 1050 + }, + { + "epoch": 0.6, + "learning_rate": 0.0003, + "loss": 2.2009, + "step": 1051 + }, + { + "epoch": 0.6, + "learning_rate": 0.0003, + "loss": 2.178, + "step": 1052 + }, + { + "epoch": 0.6, + "learning_rate": 0.0003, + "loss": 2.1692, + "step": 1053 + }, + { + "epoch": 0.6, + "learning_rate": 0.0003, + "loss": 2.1946, + "step": 1054 + }, + { + "epoch": 0.6, + "learning_rate": 0.0003, + "loss": 2.2009, + "step": 1055 + }, + { + "epoch": 0.61, + "learning_rate": 0.0003, + "loss": 2.1741, + "step": 1056 + }, + { + "epoch": 0.61, + "learning_rate": 0.0003, + "loss": 2.1827, + "step": 1057 + }, + { + "epoch": 0.61, + "learning_rate": 0.0003, + "loss": 2.1968, + "step": 1058 + }, + { + "epoch": 0.61, + "learning_rate": 0.0003, + "loss": 2.1744, + "step": 1059 + }, + { + "epoch": 0.61, + "learning_rate": 0.0003, + "loss": 2.1729, + "step": 1060 + }, + { + "epoch": 0.61, + "learning_rate": 0.0003, + "loss": 2.1755, + "step": 1061 + }, + { + "epoch": 0.61, + "learning_rate": 0.0003, + "loss": 2.1665, + "step": 1062 + }, + { + "epoch": 0.61, + "learning_rate": 0.0003, + "loss": 2.1881, + "step": 1063 + }, + { + "epoch": 0.61, + "learning_rate": 0.0003, + "loss": 2.1891, + "step": 1064 + }, + { + "epoch": 0.61, + "learning_rate": 0.0003, + "loss": 2.2041, + "step": 1065 + }, + { + "epoch": 0.61, + "learning_rate": 0.0003, + "loss": 2.1715, + "step": 1066 + }, + { + "epoch": 0.61, + "learning_rate": 0.0003, + "loss": 2.191, + "step": 1067 + }, + { + "epoch": 0.61, + "learning_rate": 0.0003, + "loss": 2.1815, + "step": 1068 + }, + { + "epoch": 0.61, + "learning_rate": 0.0003, + "loss": 2.1936, + "step": 1069 + }, + { + "epoch": 0.61, + "learning_rate": 0.0003, + "loss": 2.1846, + "step": 1070 + }, + { + "epoch": 0.61, + "learning_rate": 0.0003, + "loss": 2.1799, + "step": 1071 + }, + { + "epoch": 0.61, + "learning_rate": 0.0003, + "loss": 2.1956, + "step": 1072 + }, + { + "epoch": 0.62, + "learning_rate": 0.0003, + "loss": 2.198, + "step": 1073 + }, + { + "epoch": 0.62, + "learning_rate": 0.0003, + "loss": 2.1753, + "step": 1074 + }, + { + "epoch": 0.62, + "learning_rate": 0.0003, + "loss": 2.189, + "step": 1075 + }, + { + "epoch": 0.62, + "learning_rate": 0.0003, + "loss": 2.1754, + "step": 1076 + }, + { + "epoch": 0.62, + "learning_rate": 0.0003, + "loss": 2.1798, + "step": 1077 + }, + { + "epoch": 0.62, + "learning_rate": 0.0003, + "loss": 2.1838, + "step": 1078 + }, + { + "epoch": 0.62, + "learning_rate": 0.0003, + "loss": 2.1558, + "step": 1079 + }, + { + "epoch": 0.62, + "learning_rate": 0.0003, + "loss": 2.1653, + "step": 1080 + }, + { + "epoch": 0.62, + "learning_rate": 0.0003, + "loss": 2.1927, + "step": 1081 + }, + { + "epoch": 0.62, + "learning_rate": 0.0003, + "loss": 2.1757, + "step": 1082 + }, + { + "epoch": 0.62, + "learning_rate": 0.0003, + "loss": 2.1953, + "step": 1083 + }, + { + "epoch": 0.62, + "learning_rate": 0.0003, + "loss": 2.1826, + "step": 1084 + }, + { + "epoch": 0.62, + "learning_rate": 0.0003, + "loss": 2.1934, + "step": 1085 + }, + { + "epoch": 0.62, + "learning_rate": 0.0003, + "loss": 2.1661, + "step": 1086 + }, + { + "epoch": 0.62, + "learning_rate": 0.0003, + "loss": 2.1783, + "step": 1087 + }, + { + "epoch": 0.62, + "learning_rate": 0.0003, + "loss": 2.1718, + "step": 1088 + }, + { + "epoch": 0.62, + "learning_rate": 0.0003, + "loss": 2.1692, + "step": 1089 + }, + { + "epoch": 0.63, + "learning_rate": 0.0003, + "loss": 2.1614, + "step": 1090 + }, + { + "epoch": 0.63, + "learning_rate": 0.0003, + "loss": 2.1597, + "step": 1091 + }, + { + "epoch": 0.63, + "learning_rate": 0.0003, + "loss": 2.1613, + "step": 1092 + }, + { + "epoch": 0.63, + "learning_rate": 0.0003, + "loss": 2.1849, + "step": 1093 + }, + { + "epoch": 0.63, + "learning_rate": 0.0003, + "loss": 2.1896, + "step": 1094 + }, + { + "epoch": 0.63, + "learning_rate": 0.0003, + "loss": 2.1898, + "step": 1095 + }, + { + "epoch": 0.63, + "learning_rate": 0.0003, + "loss": 2.1736, + "step": 1096 + }, + { + "epoch": 0.63, + "learning_rate": 0.0003, + "loss": 2.1879, + "step": 1097 + }, + { + "epoch": 0.63, + "learning_rate": 0.0003, + "loss": 2.1854, + "step": 1098 + }, + { + "epoch": 0.63, + "learning_rate": 0.0003, + "loss": 2.1763, + "step": 1099 + }, + { + "epoch": 0.63, + "learning_rate": 0.0003, + "loss": 2.1794, + "step": 1100 + }, + { + "epoch": 0.63, + "learning_rate": 0.0003, + "loss": 2.1903, + "step": 1101 + }, + { + "epoch": 0.63, + "learning_rate": 0.0003, + "loss": 2.1826, + "step": 1102 + }, + { + "epoch": 0.63, + "learning_rate": 0.0003, + "loss": 2.1719, + "step": 1103 + }, + { + "epoch": 0.63, + "learning_rate": 0.0003, + "loss": 2.1735, + "step": 1104 + }, + { + "epoch": 0.63, + "learning_rate": 0.0003, + "loss": 2.184, + "step": 1105 + }, + { + "epoch": 0.63, + "learning_rate": 0.0003, + "loss": 2.1606, + "step": 1106 + }, + { + "epoch": 0.63, + "learning_rate": 0.0003, + "loss": 2.1924, + "step": 1107 + }, + { + "epoch": 0.64, + "learning_rate": 0.0003, + "loss": 2.1647, + "step": 1108 + }, + { + "epoch": 0.64, + "learning_rate": 0.0003, + "loss": 2.1847, + "step": 1109 + }, + { + "epoch": 0.64, + "learning_rate": 0.0003, + "loss": 2.1714, + "step": 1110 + }, + { + "epoch": 0.64, + "learning_rate": 0.0003, + "loss": 2.1577, + "step": 1111 + }, + { + "epoch": 0.64, + "learning_rate": 0.0003, + "loss": 2.1742, + "step": 1112 + }, + { + "epoch": 0.64, + "learning_rate": 0.0003, + "loss": 2.163, + "step": 1113 + }, + { + "epoch": 0.64, + "learning_rate": 0.0003, + "loss": 2.1742, + "step": 1114 + }, + { + "epoch": 0.64, + "learning_rate": 0.0003, + "loss": 2.1838, + "step": 1115 + }, + { + "epoch": 0.64, + "learning_rate": 0.0003, + "loss": 2.1832, + "step": 1116 + }, + { + "epoch": 0.64, + "learning_rate": 0.0003, + "loss": 2.1858, + "step": 1117 + }, + { + "epoch": 0.64, + "learning_rate": 0.0003, + "loss": 2.1744, + "step": 1118 + }, + { + "epoch": 0.64, + "learning_rate": 0.0003, + "loss": 2.1786, + "step": 1119 + }, + { + "epoch": 0.64, + "learning_rate": 0.0003, + "loss": 2.1747, + "step": 1120 + }, + { + "epoch": 0.64, + "learning_rate": 0.0003, + "loss": 2.1558, + "step": 1121 + }, + { + "epoch": 0.64, + "learning_rate": 0.0003, + "loss": 2.202, + "step": 1122 + }, + { + "epoch": 0.64, + "learning_rate": 0.0003, + "loss": 2.176, + "step": 1123 + }, + { + "epoch": 0.64, + "learning_rate": 0.0003, + "loss": 2.1589, + "step": 1124 + }, + { + "epoch": 0.65, + "learning_rate": 0.0003, + "loss": 2.1775, + "step": 1125 + }, + { + "epoch": 0.65, + "learning_rate": 0.0003, + "loss": 2.1996, + "step": 1126 + }, + { + "epoch": 0.65, + "learning_rate": 0.0003, + "loss": 2.1821, + "step": 1127 + }, + { + "epoch": 0.65, + "learning_rate": 0.0003, + "loss": 2.1848, + "step": 1128 + }, + { + "epoch": 0.65, + "learning_rate": 0.0003, + "loss": 2.1689, + "step": 1129 + }, + { + "epoch": 0.65, + "learning_rate": 0.0003, + "loss": 2.1938, + "step": 1130 + }, + { + "epoch": 0.65, + "learning_rate": 0.0003, + "loss": 2.1663, + "step": 1131 + }, + { + "epoch": 0.65, + "learning_rate": 0.0003, + "loss": 2.1805, + "step": 1132 + }, + { + "epoch": 0.65, + "learning_rate": 0.0003, + "loss": 2.1833, + "step": 1133 + }, + { + "epoch": 0.65, + "learning_rate": 0.0003, + "loss": 2.1665, + "step": 1134 + }, + { + "epoch": 0.65, + "learning_rate": 0.0003, + "loss": 2.1732, + "step": 1135 + }, + { + "epoch": 0.65, + "learning_rate": 0.0003, + "loss": 2.1483, + "step": 1136 + }, + { + "epoch": 0.65, + "learning_rate": 0.0003, + "loss": 2.163, + "step": 1137 + }, + { + "epoch": 0.65, + "learning_rate": 0.0003, + "loss": 2.194, + "step": 1138 + }, + { + "epoch": 0.65, + "learning_rate": 0.0003, + "loss": 2.1611, + "step": 1139 + }, + { + "epoch": 0.65, + "learning_rate": 0.0003, + "loss": 2.1672, + "step": 1140 + }, + { + "epoch": 0.65, + "learning_rate": 0.0003, + "loss": 2.1636, + "step": 1141 + }, + { + "epoch": 0.65, + "learning_rate": 0.0003, + "loss": 2.1647, + "step": 1142 + }, + { + "epoch": 0.66, + "learning_rate": 0.0003, + "loss": 2.1782, + "step": 1143 + }, + { + "epoch": 0.66, + "learning_rate": 0.0003, + "loss": 2.1761, + "step": 1144 + }, + { + "epoch": 0.66, + "learning_rate": 0.0003, + "loss": 2.1882, + "step": 1145 + }, + { + "epoch": 0.66, + "learning_rate": 0.0003, + "loss": 2.1599, + "step": 1146 + }, + { + "epoch": 0.66, + "learning_rate": 0.0003, + "loss": 2.1738, + "step": 1147 + }, + { + "epoch": 0.66, + "learning_rate": 0.0003, + "loss": 2.1533, + "step": 1148 + }, + { + "epoch": 0.66, + "learning_rate": 0.0003, + "loss": 2.1553, + "step": 1149 + }, + { + "epoch": 0.66, + "learning_rate": 0.0003, + "loss": 2.17, + "step": 1150 + }, + { + "epoch": 0.66, + "learning_rate": 0.0003, + "loss": 2.1764, + "step": 1151 + }, + { + "epoch": 0.66, + "learning_rate": 0.0003, + "loss": 2.1726, + "step": 1152 + }, + { + "epoch": 0.66, + "learning_rate": 0.0003, + "loss": 2.1757, + "step": 1153 + }, + { + "epoch": 0.66, + "learning_rate": 0.0003, + "loss": 2.155, + "step": 1154 + }, + { + "epoch": 0.66, + "learning_rate": 0.0003, + "loss": 2.1587, + "step": 1155 + }, + { + "epoch": 0.66, + "learning_rate": 0.0003, + "loss": 2.1772, + "step": 1156 + }, + { + "epoch": 0.66, + "learning_rate": 0.0003, + "loss": 2.1558, + "step": 1157 + }, + { + "epoch": 0.66, + "learning_rate": 0.0003, + "loss": 2.1725, + "step": 1158 + }, + { + "epoch": 0.66, + "learning_rate": 0.0003, + "loss": 2.1506, + "step": 1159 + }, + { + "epoch": 0.67, + "learning_rate": 0.0003, + "loss": 2.1674, + "step": 1160 + }, + { + "epoch": 0.67, + "learning_rate": 0.0003, + "loss": 2.1749, + "step": 1161 + }, + { + "epoch": 0.67, + "learning_rate": 0.0003, + "loss": 2.1644, + "step": 1162 + }, + { + "epoch": 0.67, + "learning_rate": 0.0003, + "loss": 2.1649, + "step": 1163 + }, + { + "epoch": 0.67, + "learning_rate": 0.0003, + "loss": 2.1707, + "step": 1164 + }, + { + "epoch": 0.67, + "learning_rate": 0.0003, + "loss": 2.1581, + "step": 1165 + }, + { + "epoch": 0.67, + "learning_rate": 0.0003, + "loss": 2.1544, + "step": 1166 + }, + { + "epoch": 0.67, + "learning_rate": 0.0003, + "loss": 2.1509, + "step": 1167 + }, + { + "epoch": 0.67, + "learning_rate": 0.0003, + "loss": 2.1718, + "step": 1168 + }, + { + "epoch": 0.67, + "learning_rate": 0.0003, + "loss": 2.1637, + "step": 1169 + }, + { + "epoch": 0.67, + "learning_rate": 0.0003, + "loss": 2.179, + "step": 1170 + }, + { + "epoch": 0.67, + "learning_rate": 0.0003, + "loss": 2.1801, + "step": 1171 + }, + { + "epoch": 0.67, + "learning_rate": 0.0003, + "loss": 2.1667, + "step": 1172 + }, + { + "epoch": 0.67, + "learning_rate": 0.0003, + "loss": 2.1514, + "step": 1173 + }, + { + "epoch": 0.67, + "learning_rate": 0.0003, + "loss": 2.1821, + "step": 1174 + }, + { + "epoch": 0.67, + "learning_rate": 0.0003, + "loss": 2.1545, + "step": 1175 + }, + { + "epoch": 0.67, + "learning_rate": 0.0003, + "loss": 2.1464, + "step": 1176 + }, + { + "epoch": 0.67, + "learning_rate": 0.0003, + "loss": 2.1646, + "step": 1177 + }, + { + "epoch": 0.68, + "learning_rate": 0.0003, + "loss": 2.1691, + "step": 1178 + }, + { + "epoch": 0.68, + "learning_rate": 0.0003, + "loss": 2.1578, + "step": 1179 + }, + { + "epoch": 0.68, + "learning_rate": 0.0003, + "loss": 2.1821, + "step": 1180 + }, + { + "epoch": 0.68, + "learning_rate": 0.0003, + "loss": 2.1575, + "step": 1181 + }, + { + "epoch": 0.68, + "learning_rate": 0.0003, + "loss": 2.1628, + "step": 1182 + }, + { + "epoch": 0.68, + "learning_rate": 0.0003, + "loss": 2.1488, + "step": 1183 + }, + { + "epoch": 0.68, + "learning_rate": 0.0003, + "loss": 2.1532, + "step": 1184 + }, + { + "epoch": 0.68, + "learning_rate": 0.0003, + "loss": 2.1605, + "step": 1185 + }, + { + "epoch": 0.68, + "learning_rate": 0.0003, + "loss": 2.1714, + "step": 1186 + }, + { + "epoch": 0.68, + "learning_rate": 0.0003, + "loss": 2.1584, + "step": 1187 + }, + { + "epoch": 0.68, + "learning_rate": 0.0003, + "loss": 2.1698, + "step": 1188 + }, + { + "epoch": 0.68, + "learning_rate": 0.0003, + "loss": 2.1731, + "step": 1189 + }, + { + "epoch": 0.68, + "learning_rate": 0.0003, + "loss": 2.1678, + "step": 1190 + }, + { + "epoch": 0.68, + "eval_loss": 2.1559829711914062, + "eval_runtime": 2009.3633, + "eval_samples_per_second": 52.496, + "eval_steps_per_second": 0.137, + "step": 1190 + }, + { + "epoch": 0.68, + "learning_rate": 0.0003, + "loss": 2.1783, + "step": 1191 + }, + { + "epoch": 0.68, + "learning_rate": 0.0003, + "loss": 2.1525, + "step": 1192 + }, + { + "epoch": 0.68, + "learning_rate": 0.0003, + "loss": 2.1621, + "step": 1193 + }, + { + "epoch": 0.68, + "learning_rate": 0.0003, + "loss": 2.1437, + "step": 1194 + }, + { + "epoch": 0.69, + "learning_rate": 0.0003, + "loss": 2.1654, + "step": 1195 + }, + { + "epoch": 0.69, + "learning_rate": 0.0003, + "loss": 2.1464, + "step": 1196 + }, + { + "epoch": 0.69, + "learning_rate": 0.0003, + "loss": 2.1483, + "step": 1197 + }, + { + "epoch": 0.69, + "learning_rate": 0.0003, + "loss": 2.1655, + "step": 1198 + }, + { + "epoch": 0.69, + "learning_rate": 0.0003, + "loss": 2.1608, + "step": 1199 + }, + { + "epoch": 0.69, + "learning_rate": 0.0003, + "loss": 2.1515, + "step": 1200 + }, + { + "epoch": 0.69, + "learning_rate": 0.0003, + "loss": 2.181, + "step": 1201 + }, + { + "epoch": 0.69, + "learning_rate": 0.0003, + "loss": 2.1703, + "step": 1202 + }, + { + "epoch": 0.69, + "learning_rate": 0.0003, + "loss": 2.1697, + "step": 1203 + }, + { + "epoch": 0.69, + "learning_rate": 0.0003, + "loss": 2.1631, + "step": 1204 + }, + { + "epoch": 0.69, + "learning_rate": 0.0003, + "loss": 2.1672, + "step": 1205 + }, + { + "epoch": 0.69, + "learning_rate": 0.0003, + "loss": 2.1416, + "step": 1206 + }, + { + "epoch": 0.69, + "learning_rate": 0.0003, + "loss": 2.1564, + "step": 1207 + }, + { + "epoch": 0.69, + "learning_rate": 0.0003, + "loss": 2.175, + "step": 1208 + }, + { + "epoch": 0.69, + "learning_rate": 0.0003, + "loss": 2.1569, + "step": 1209 + }, + { + "epoch": 0.69, + "learning_rate": 0.0003, + "loss": 2.1506, + "step": 1210 + }, + { + "epoch": 0.69, + "learning_rate": 0.0003, + "loss": 2.1403, + "step": 1211 + }, + { + "epoch": 0.7, + "learning_rate": 0.0003, + "loss": 2.1613, + "step": 1212 + }, + { + "epoch": 0.7, + "learning_rate": 0.0003, + "loss": 2.1483, + "step": 1213 + }, + { + "epoch": 0.7, + "learning_rate": 0.0003, + "loss": 2.1624, + "step": 1214 + }, + { + "epoch": 0.7, + "learning_rate": 0.0003, + "loss": 2.151, + "step": 1215 + }, + { + "epoch": 0.7, + "learning_rate": 0.0003, + "loss": 2.1646, + "step": 1216 + }, + { + "epoch": 0.7, + "learning_rate": 0.0003, + "loss": 2.1622, + "step": 1217 + }, + { + "epoch": 0.7, + "learning_rate": 0.0003, + "loss": 2.1548, + "step": 1218 + }, + { + "epoch": 0.7, + "learning_rate": 0.0003, + "loss": 2.1698, + "step": 1219 + }, + { + "epoch": 0.7, + "learning_rate": 0.0003, + "loss": 2.1448, + "step": 1220 + }, + { + "epoch": 0.7, + "learning_rate": 0.0003, + "loss": 2.1337, + "step": 1221 + }, + { + "epoch": 0.7, + "learning_rate": 0.0003, + "loss": 2.1609, + "step": 1222 + }, + { + "epoch": 0.7, + "learning_rate": 0.0003, + "loss": 2.1593, + "step": 1223 + }, + { + "epoch": 0.7, + "learning_rate": 0.0003, + "loss": 2.1545, + "step": 1224 + }, + { + "epoch": 0.7, + "learning_rate": 0.0003, + "loss": 2.146, + "step": 1225 + }, + { + "epoch": 0.7, + "learning_rate": 0.0003, + "loss": 2.1315, + "step": 1226 + }, + { + "epoch": 0.7, + "learning_rate": 0.0003, + "loss": 2.151, + "step": 1227 + }, + { + "epoch": 0.7, + "learning_rate": 0.0003, + "loss": 2.1606, + "step": 1228 + }, + { + "epoch": 0.7, + "learning_rate": 0.0003, + "loss": 2.1511, + "step": 1229 + }, + { + "epoch": 0.71, + "learning_rate": 0.0003, + "loss": 2.1576, + "step": 1230 + }, + { + "epoch": 0.71, + "learning_rate": 0.0003, + "loss": 2.1564, + "step": 1231 + }, + { + "epoch": 0.71, + "learning_rate": 0.0003, + "loss": 2.1395, + "step": 1232 + }, + { + "epoch": 0.71, + "learning_rate": 0.0003, + "loss": 2.1608, + "step": 1233 + }, + { + "epoch": 0.71, + "learning_rate": 0.0003, + "loss": 2.1406, + "step": 1234 + }, + { + "epoch": 0.71, + "learning_rate": 0.0003, + "loss": 2.1569, + "step": 1235 + }, + { + "epoch": 0.71, + "learning_rate": 0.0003, + "loss": 2.1611, + "step": 1236 + }, + { + "epoch": 0.71, + "learning_rate": 0.0003, + "loss": 2.1763, + "step": 1237 + }, + { + "epoch": 0.71, + "learning_rate": 0.0003, + "loss": 2.1635, + "step": 1238 + }, + { + "epoch": 0.71, + "learning_rate": 0.0003, + "loss": 2.1519, + "step": 1239 + }, + { + "epoch": 0.71, + "learning_rate": 0.0003, + "loss": 2.1575, + "step": 1240 + }, + { + "epoch": 0.71, + "learning_rate": 0.0003, + "loss": 2.1617, + "step": 1241 + }, + { + "epoch": 0.71, + "learning_rate": 0.0003, + "loss": 2.1616, + "step": 1242 + }, + { + "epoch": 0.71, + "learning_rate": 0.0003, + "loss": 2.1406, + "step": 1243 + }, + { + "epoch": 0.71, + "learning_rate": 0.0003, + "loss": 2.1656, + "step": 1244 + }, + { + "epoch": 0.71, + "learning_rate": 0.0003, + "loss": 2.1539, + "step": 1245 + }, + { + "epoch": 0.71, + "learning_rate": 0.0003, + "loss": 2.1564, + "step": 1246 + }, + { + "epoch": 0.72, + "learning_rate": 0.0003, + "loss": 2.153, + "step": 1247 + }, + { + "epoch": 0.72, + "learning_rate": 0.0003, + "loss": 2.1538, + "step": 1248 + }, + { + "epoch": 0.72, + "learning_rate": 0.0003, + "loss": 2.1541, + "step": 1249 + }, + { + "epoch": 0.72, + "learning_rate": 0.0003, + "loss": 2.1549, + "step": 1250 + }, + { + "epoch": 0.72, + "learning_rate": 0.0003, + "loss": 2.1589, + "step": 1251 + }, + { + "epoch": 0.72, + "learning_rate": 0.0003, + "loss": 2.1478, + "step": 1252 + }, + { + "epoch": 0.72, + "learning_rate": 0.0003, + "loss": 2.1542, + "step": 1253 + }, + { + "epoch": 0.72, + "learning_rate": 0.0003, + "loss": 2.1749, + "step": 1254 + }, + { + "epoch": 0.72, + "learning_rate": 0.0003, + "loss": 2.151, + "step": 1255 + }, + { + "epoch": 0.72, + "learning_rate": 0.0003, + "loss": 2.1349, + "step": 1256 + }, + { + "epoch": 0.72, + "learning_rate": 0.0003, + "loss": 2.1597, + "step": 1257 + }, + { + "epoch": 0.72, + "learning_rate": 0.0003, + "loss": 2.1589, + "step": 1258 + }, + { + "epoch": 0.72, + "learning_rate": 0.0003, + "loss": 2.1389, + "step": 1259 + }, + { + "epoch": 0.72, + "learning_rate": 0.0003, + "loss": 2.1392, + "step": 1260 + }, + { + "epoch": 0.72, + "learning_rate": 0.0003, + "loss": 2.1479, + "step": 1261 + }, + { + "epoch": 0.72, + "learning_rate": 0.0003, + "loss": 2.1667, + "step": 1262 + }, + { + "epoch": 0.72, + "learning_rate": 0.0003, + "loss": 2.1598, + "step": 1263 + }, + { + "epoch": 0.72, + "learning_rate": 0.0003, + "loss": 2.1628, + "step": 1264 + }, + { + "epoch": 0.73, + "learning_rate": 0.0003, + "loss": 2.1528, + "step": 1265 + }, + { + "epoch": 0.73, + "learning_rate": 0.0003, + "loss": 2.1527, + "step": 1266 + }, + { + "epoch": 0.73, + "learning_rate": 0.0003, + "loss": 2.144, + "step": 1267 + }, + { + "epoch": 0.73, + "learning_rate": 0.0003, + "loss": 2.1569, + "step": 1268 + }, + { + "epoch": 0.73, + "learning_rate": 0.0003, + "loss": 2.1156, + "step": 1269 + }, + { + "epoch": 0.73, + "learning_rate": 0.0003, + "loss": 2.1423, + "step": 1270 + }, + { + "epoch": 0.73, + "learning_rate": 0.0003, + "loss": 2.1471, + "step": 1271 + }, + { + "epoch": 0.73, + "learning_rate": 0.0003, + "loss": 2.1495, + "step": 1272 + }, + { + "epoch": 0.73, + "learning_rate": 0.0003, + "loss": 2.1437, + "step": 1273 + }, + { + "epoch": 0.73, + "learning_rate": 0.0003, + "loss": 2.149, + "step": 1274 + }, + { + "epoch": 0.73, + "learning_rate": 0.0003, + "loss": 2.1489, + "step": 1275 + }, + { + "epoch": 0.73, + "learning_rate": 0.0003, + "loss": 2.149, + "step": 1276 + }, + { + "epoch": 0.73, + "learning_rate": 0.0003, + "loss": 2.1426, + "step": 1277 + }, + { + "epoch": 0.73, + "learning_rate": 0.0003, + "loss": 2.1543, + "step": 1278 + }, + { + "epoch": 0.73, + "learning_rate": 0.0003, + "loss": 2.1416, + "step": 1279 + }, + { + "epoch": 0.73, + "learning_rate": 0.0003, + "loss": 2.1525, + "step": 1280 + }, + { + "epoch": 0.73, + "learning_rate": 0.0003, + "loss": 2.1548, + "step": 1281 + }, + { + "epoch": 0.74, + "learning_rate": 0.0003, + "loss": 2.1399, + "step": 1282 + }, + { + "epoch": 0.74, + "learning_rate": 0.0003, + "loss": 2.1525, + "step": 1283 + }, + { + "epoch": 0.74, + "learning_rate": 0.0003, + "loss": 2.1407, + "step": 1284 + }, + { + "epoch": 0.74, + "learning_rate": 0.0003, + "loss": 2.1669, + "step": 1285 + }, + { + "epoch": 0.74, + "learning_rate": 0.0003, + "loss": 2.1537, + "step": 1286 + }, + { + "epoch": 0.74, + "learning_rate": 0.0003, + "loss": 2.1522, + "step": 1287 + }, + { + "epoch": 0.74, + "learning_rate": 0.0003, + "loss": 2.1738, + "step": 1288 + }, + { + "epoch": 0.74, + "learning_rate": 0.0003, + "loss": 2.1462, + "step": 1289 + }, + { + "epoch": 0.74, + "learning_rate": 0.0003, + "loss": 2.1443, + "step": 1290 + }, + { + "epoch": 0.74, + "learning_rate": 0.0003, + "loss": 2.1343, + "step": 1291 + }, + { + "epoch": 0.74, + "learning_rate": 0.0003, + "loss": 2.135, + "step": 1292 + }, + { + "epoch": 0.74, + "learning_rate": 0.0003, + "loss": 2.1439, + "step": 1293 + }, + { + "epoch": 0.74, + "learning_rate": 0.0003, + "loss": 2.1539, + "step": 1294 + }, + { + "epoch": 0.74, + "learning_rate": 0.0003, + "loss": 2.1493, + "step": 1295 + }, + { + "epoch": 0.74, + "learning_rate": 0.0003, + "loss": 2.1467, + "step": 1296 + }, + { + "epoch": 0.74, + "learning_rate": 0.0003, + "loss": 2.144, + "step": 1297 + }, + { + "epoch": 0.74, + "learning_rate": 0.0003, + "loss": 2.1432, + "step": 1298 + }, + { + "epoch": 0.74, + "learning_rate": 0.0003, + "loss": 2.1614, + "step": 1299 + }, + { + "epoch": 0.75, + "learning_rate": 0.0003, + "loss": 2.149, + "step": 1300 + }, + { + "epoch": 0.75, + "learning_rate": 0.0003, + "loss": 2.1763, + "step": 1301 + }, + { + "epoch": 0.75, + "learning_rate": 0.0003, + "loss": 2.1514, + "step": 1302 + }, + { + "epoch": 0.75, + "learning_rate": 0.0003, + "loss": 2.1444, + "step": 1303 + }, + { + "epoch": 0.75, + "learning_rate": 0.0003, + "loss": 2.134, + "step": 1304 + }, + { + "epoch": 0.75, + "learning_rate": 0.0003, + "loss": 2.1331, + "step": 1305 + }, + { + "epoch": 0.75, + "learning_rate": 0.0003, + "loss": 2.1545, + "step": 1306 + }, + { + "epoch": 0.75, + "learning_rate": 0.0003, + "loss": 2.1355, + "step": 1307 + }, + { + "epoch": 0.75, + "learning_rate": 0.0003, + "loss": 2.141, + "step": 1308 + }, + { + "epoch": 0.75, + "learning_rate": 0.0003, + "loss": 2.1407, + "step": 1309 + }, + { + "epoch": 0.75, + "learning_rate": 0.0003, + "loss": 2.1429, + "step": 1310 + }, + { + "epoch": 0.75, + "learning_rate": 0.0003, + "loss": 2.132, + "step": 1311 + }, + { + "epoch": 0.75, + "learning_rate": 0.0003, + "loss": 2.1543, + "step": 1312 + }, + { + "epoch": 0.75, + "learning_rate": 0.0003, + "loss": 2.1682, + "step": 1313 + }, + { + "epoch": 0.75, + "learning_rate": 0.0003, + "loss": 2.146, + "step": 1314 + }, + { + "epoch": 0.75, + "learning_rate": 0.0003, + "loss": 2.1326, + "step": 1315 + }, + { + "epoch": 0.75, + "learning_rate": 0.0003, + "loss": 2.1421, + "step": 1316 + }, + { + "epoch": 0.76, + "learning_rate": 0.0003, + "loss": 2.1333, + "step": 1317 + }, + { + "epoch": 0.76, + "learning_rate": 0.0003, + "loss": 2.1342, + "step": 1318 + }, + { + "epoch": 0.76, + "learning_rate": 0.0003, + "loss": 2.1385, + "step": 1319 + }, + { + "epoch": 0.76, + "learning_rate": 0.0003, + "loss": 2.1301, + "step": 1320 + }, + { + "epoch": 0.76, + "learning_rate": 0.0003, + "loss": 2.1449, + "step": 1321 + }, + { + "epoch": 0.76, + "learning_rate": 0.0003, + "loss": 2.1255, + "step": 1322 + }, + { + "epoch": 0.76, + "learning_rate": 0.0003, + "loss": 2.1494, + "step": 1323 + }, + { + "epoch": 0.76, + "learning_rate": 0.0003, + "loss": 2.1385, + "step": 1324 + }, + { + "epoch": 0.76, + "learning_rate": 0.0003, + "loss": 2.1471, + "step": 1325 + }, + { + "epoch": 0.76, + "learning_rate": 0.0003, + "loss": 2.1257, + "step": 1326 + }, + { + "epoch": 0.76, + "learning_rate": 0.0003, + "loss": 2.1581, + "step": 1327 + }, + { + "epoch": 0.76, + "learning_rate": 0.0003, + "loss": 2.1405, + "step": 1328 + }, + { + "epoch": 0.76, + "learning_rate": 0.0003, + "loss": 2.1405, + "step": 1329 + }, + { + "epoch": 0.76, + "learning_rate": 0.0003, + "loss": 2.1469, + "step": 1330 + }, + { + "epoch": 0.76, + "learning_rate": 0.0003, + "loss": 2.1415, + "step": 1331 + }, + { + "epoch": 0.76, + "learning_rate": 0.0003, + "loss": 2.1316, + "step": 1332 + }, + { + "epoch": 0.76, + "learning_rate": 0.0003, + "loss": 2.1273, + "step": 1333 + }, + { + "epoch": 0.76, + "learning_rate": 0.0003, + "loss": 2.1392, + "step": 1334 + }, + { + "epoch": 0.77, + "learning_rate": 0.0003, + "loss": 2.1296, + "step": 1335 + }, + { + "epoch": 0.77, + "learning_rate": 0.0003, + "loss": 2.1478, + "step": 1336 + }, + { + "epoch": 0.77, + "learning_rate": 0.0003, + "loss": 2.132, + "step": 1337 + }, + { + "epoch": 0.77, + "learning_rate": 0.0003, + "loss": 2.1381, + "step": 1338 + }, + { + "epoch": 0.77, + "learning_rate": 0.0003, + "loss": 2.1521, + "step": 1339 + }, + { + "epoch": 0.77, + "learning_rate": 0.0003, + "loss": 2.1361, + "step": 1340 + }, + { + "epoch": 0.77, + "learning_rate": 0.0003, + "loss": 2.1309, + "step": 1341 + }, + { + "epoch": 0.77, + "learning_rate": 0.0003, + "loss": 2.1462, + "step": 1342 + }, + { + "epoch": 0.77, + "learning_rate": 0.0003, + "loss": 2.1193, + "step": 1343 + }, + { + "epoch": 0.77, + "learning_rate": 0.0003, + "loss": 2.1465, + "step": 1344 + }, + { + "epoch": 0.77, + "learning_rate": 0.0003, + "loss": 2.1433, + "step": 1345 + }, + { + "epoch": 0.77, + "learning_rate": 0.0003, + "loss": 2.1195, + "step": 1346 + }, + { + "epoch": 0.77, + "learning_rate": 0.0003, + "loss": 2.1229, + "step": 1347 + }, + { + "epoch": 0.77, + "learning_rate": 0.0003, + "loss": 2.1423, + "step": 1348 + }, + { + "epoch": 0.77, + "learning_rate": 0.0003, + "loss": 2.1418, + "step": 1349 + }, + { + "epoch": 0.77, + "learning_rate": 0.0003, + "loss": 2.1429, + "step": 1350 + }, + { + "epoch": 0.77, + "learning_rate": 0.0003, + "loss": 2.1191, + "step": 1351 + }, + { + "epoch": 0.78, + "learning_rate": 0.0003, + "loss": 2.1288, + "step": 1352 + }, + { + "epoch": 0.78, + "learning_rate": 0.0003, + "loss": 2.1392, + "step": 1353 + }, + { + "epoch": 0.78, + "learning_rate": 0.0003, + "loss": 2.1509, + "step": 1354 + }, + { + "epoch": 0.78, + "learning_rate": 0.0003, + "loss": 2.1221, + "step": 1355 + }, + { + "epoch": 0.78, + "learning_rate": 0.0003, + "loss": 2.1138, + "step": 1356 + }, + { + "epoch": 0.78, + "learning_rate": 0.0003, + "loss": 2.1461, + "step": 1357 + }, + { + "epoch": 0.78, + "learning_rate": 0.0003, + "loss": 2.1322, + "step": 1358 + }, + { + "epoch": 0.78, + "learning_rate": 0.0003, + "loss": 2.1362, + "step": 1359 + }, + { + "epoch": 0.78, + "learning_rate": 0.0003, + "loss": 2.1301, + "step": 1360 + }, + { + "epoch": 0.78, + "eval_loss": 2.1311893463134766, + "eval_runtime": 2006.6165, + "eval_samples_per_second": 52.568, + "eval_steps_per_second": 0.137, + "step": 1360 + }, + { + "epoch": 0.78, + "learning_rate": 0.0003, + "loss": 2.1521, + "step": 1361 + }, + { + "epoch": 0.78, + "learning_rate": 0.0003, + "loss": 2.1393, + "step": 1362 + }, + { + "epoch": 0.78, + "learning_rate": 0.0003, + "loss": 2.1355, + "step": 1363 + }, + { + "epoch": 0.78, + "learning_rate": 0.0003, + "loss": 2.1437, + "step": 1364 + }, + { + "epoch": 0.78, + "learning_rate": 0.0003, + "loss": 2.1311, + "step": 1365 + }, + { + "epoch": 0.78, + "learning_rate": 0.0003, + "loss": 2.1338, + "step": 1366 + }, + { + "epoch": 0.78, + "learning_rate": 0.0003, + "loss": 2.157, + "step": 1367 + }, + { + "epoch": 0.78, + "learning_rate": 0.0003, + "loss": 2.134, + "step": 1368 + }, + { + "epoch": 0.79, + "learning_rate": 0.0003, + "loss": 2.1368, + "step": 1369 + }, + { + "epoch": 0.79, + "learning_rate": 0.0003, + "loss": 2.1188, + "step": 1370 + }, + { + "epoch": 0.79, + "learning_rate": 0.0003, + "loss": 2.1256, + "step": 1371 + }, + { + "epoch": 0.79, + "learning_rate": 0.0003, + "loss": 2.1288, + "step": 1372 + }, + { + "epoch": 0.79, + "learning_rate": 0.0003, + "loss": 2.1335, + "step": 1373 + }, + { + "epoch": 0.79, + "learning_rate": 0.0003, + "loss": 2.1222, + "step": 1374 + }, + { + "epoch": 0.79, + "learning_rate": 0.0003, + "loss": 2.1277, + "step": 1375 + }, + { + "epoch": 0.79, + "learning_rate": 0.0003, + "loss": 2.1239, + "step": 1376 + }, + { + "epoch": 0.79, + "learning_rate": 0.0003, + "loss": 2.1376, + "step": 1377 + }, + { + "epoch": 0.79, + "learning_rate": 0.0003, + "loss": 2.1327, + "step": 1378 + }, + { + "epoch": 0.79, + "learning_rate": 0.0003, + "loss": 2.1462, + "step": 1379 + }, + { + "epoch": 0.79, + "learning_rate": 0.0003, + "loss": 2.1385, + "step": 1380 + }, + { + "epoch": 0.79, + "learning_rate": 0.0003, + "loss": 2.1445, + "step": 1381 + }, + { + "epoch": 0.79, + "learning_rate": 0.0003, + "loss": 2.1384, + "step": 1382 + }, + { + "epoch": 0.79, + "learning_rate": 0.0003, + "loss": 2.1294, + "step": 1383 + }, + { + "epoch": 0.79, + "learning_rate": 0.0003, + "loss": 2.1287, + "step": 1384 + }, + { + "epoch": 0.79, + "learning_rate": 0.0003, + "loss": 2.1412, + "step": 1385 + }, + { + "epoch": 0.79, + "learning_rate": 0.0003, + "loss": 2.1238, + "step": 1386 + }, + { + "epoch": 0.8, + "learning_rate": 0.0003, + "loss": 2.1406, + "step": 1387 + }, + { + "epoch": 0.8, + "learning_rate": 0.0003, + "loss": 2.1229, + "step": 1388 + }, + { + "epoch": 0.8, + "learning_rate": 0.0003, + "loss": 2.1375, + "step": 1389 + }, + { + "epoch": 0.8, + "learning_rate": 0.0003, + "loss": 2.1255, + "step": 1390 + }, + { + "epoch": 0.8, + "learning_rate": 0.0003, + "loss": 2.1176, + "step": 1391 + }, + { + "epoch": 0.8, + "learning_rate": 0.0003, + "loss": 2.1278, + "step": 1392 + }, + { + "epoch": 0.8, + "learning_rate": 0.0003, + "loss": 2.1375, + "step": 1393 + }, + { + "epoch": 0.8, + "learning_rate": 0.0003, + "loss": 2.1184, + "step": 1394 + }, + { + "epoch": 0.8, + "learning_rate": 0.0003, + "loss": 2.1355, + "step": 1395 + }, + { + "epoch": 0.8, + "learning_rate": 0.0003, + "loss": 2.1241, + "step": 1396 + }, + { + "epoch": 0.8, + "learning_rate": 0.0003, + "loss": 2.1177, + "step": 1397 + }, + { + "epoch": 0.8, + "learning_rate": 0.0003, + "loss": 2.1316, + "step": 1398 + }, + { + "epoch": 0.8, + "learning_rate": 0.0003, + "loss": 2.1111, + "step": 1399 + }, + { + "epoch": 0.8, + "learning_rate": 0.0003, + "loss": 2.1096, + "step": 1400 + }, + { + "epoch": 0.8, + "learning_rate": 0.0003, + "loss": 2.1371, + "step": 1401 + }, + { + "epoch": 0.8, + "learning_rate": 0.0003, + "loss": 2.1384, + "step": 1402 + }, + { + "epoch": 0.8, + "learning_rate": 0.0003, + "loss": 2.1385, + "step": 1403 + }, + { + "epoch": 0.81, + "learning_rate": 0.0003, + "loss": 2.1401, + "step": 1404 + }, + { + "epoch": 0.81, + "learning_rate": 0.0003, + "loss": 2.1277, + "step": 1405 + }, + { + "epoch": 0.81, + "learning_rate": 0.0003, + "loss": 2.1434, + "step": 1406 + }, + { + "epoch": 0.81, + "learning_rate": 0.0003, + "loss": 2.1194, + "step": 1407 + }, + { + "epoch": 0.81, + "learning_rate": 0.0003, + "loss": 2.1212, + "step": 1408 + }, + { + "epoch": 0.81, + "learning_rate": 0.0003, + "loss": 2.1198, + "step": 1409 + }, + { + "epoch": 0.81, + "learning_rate": 0.0003, + "loss": 2.1423, + "step": 1410 + }, + { + "epoch": 0.81, + "learning_rate": 0.0003, + "loss": 2.1262, + "step": 1411 + }, + { + "epoch": 0.81, + "learning_rate": 0.0003, + "loss": 2.1177, + "step": 1412 + }, + { + "epoch": 0.81, + "learning_rate": 0.0003, + "loss": 2.1361, + "step": 1413 + }, + { + "epoch": 0.81, + "learning_rate": 0.0003, + "loss": 2.1331, + "step": 1414 + }, + { + "epoch": 0.81, + "learning_rate": 0.0003, + "loss": 2.1248, + "step": 1415 + }, + { + "epoch": 0.81, + "learning_rate": 0.0003, + "loss": 2.123, + "step": 1416 + }, + { + "epoch": 0.81, + "learning_rate": 0.0003, + "loss": 2.1385, + "step": 1417 + }, + { + "epoch": 0.81, + "learning_rate": 0.0003, + "loss": 2.1285, + "step": 1418 + }, + { + "epoch": 0.81, + "learning_rate": 0.0003, + "loss": 2.1262, + "step": 1419 + }, + { + "epoch": 0.81, + "learning_rate": 0.0003, + "loss": 2.1261, + "step": 1420 + }, + { + "epoch": 0.81, + "learning_rate": 0.0003, + "loss": 2.1381, + "step": 1421 + }, + { + "epoch": 0.82, + "learning_rate": 0.0003, + "loss": 2.1351, + "step": 1422 + }, + { + "epoch": 0.82, + "learning_rate": 0.0003, + "loss": 2.125, + "step": 1423 + }, + { + "epoch": 0.82, + "learning_rate": 0.0003, + "loss": 2.1271, + "step": 1424 + }, + { + "epoch": 0.82, + "learning_rate": 0.0003, + "loss": 2.1274, + "step": 1425 + }, + { + "epoch": 0.82, + "learning_rate": 0.0003, + "loss": 2.115, + "step": 1426 + }, + { + "epoch": 0.82, + "learning_rate": 0.0003, + "loss": 2.1139, + "step": 1427 + }, + { + "epoch": 0.82, + "learning_rate": 0.0003, + "loss": 2.146, + "step": 1428 + }, + { + "epoch": 0.82, + "learning_rate": 0.0003, + "loss": 2.1222, + "step": 1429 + }, + { + "epoch": 0.82, + "learning_rate": 0.0003, + "loss": 2.1245, + "step": 1430 + }, + { + "epoch": 0.82, + "learning_rate": 0.0003, + "loss": 2.115, + "step": 1431 + }, + { + "epoch": 0.82, + "learning_rate": 0.0003, + "loss": 2.1276, + "step": 1432 + }, + { + "epoch": 0.82, + "learning_rate": 0.0003, + "loss": 2.1067, + "step": 1433 + }, + { + "epoch": 0.82, + "learning_rate": 0.0003, + "loss": 2.1189, + "step": 1434 + }, + { + "epoch": 0.82, + "learning_rate": 0.0003, + "loss": 2.1326, + "step": 1435 + }, + { + "epoch": 0.82, + "learning_rate": 0.0003, + "loss": 2.1146, + "step": 1436 + }, + { + "epoch": 0.82, + "learning_rate": 0.0003, + "loss": 2.1239, + "step": 1437 + }, + { + "epoch": 0.82, + "learning_rate": 0.0003, + "loss": 2.1195, + "step": 1438 + }, + { + "epoch": 0.83, + "learning_rate": 0.0003, + "loss": 2.1329, + "step": 1439 + }, + { + "epoch": 0.83, + "learning_rate": 0.0003, + "loss": 2.1171, + "step": 1440 + }, + { + "epoch": 0.83, + "learning_rate": 0.0003, + "loss": 2.1188, + "step": 1441 + }, + { + "epoch": 0.83, + "learning_rate": 0.0003, + "loss": 2.1532, + "step": 1442 + }, + { + "epoch": 0.83, + "learning_rate": 0.0003, + "loss": 2.1105, + "step": 1443 + }, + { + "epoch": 0.83, + "learning_rate": 0.0003, + "loss": 2.1357, + "step": 1444 + }, + { + "epoch": 0.83, + "learning_rate": 0.0003, + "loss": 2.147, + "step": 1445 + }, + { + "epoch": 0.83, + "learning_rate": 0.0003, + "loss": 2.1315, + "step": 1446 + }, + { + "epoch": 0.83, + "learning_rate": 0.0003, + "loss": 2.1443, + "step": 1447 + }, + { + "epoch": 0.83, + "learning_rate": 0.0003, + "loss": 2.1321, + "step": 1448 + }, + { + "epoch": 0.83, + "learning_rate": 0.0003, + "loss": 2.1123, + "step": 1449 + }, + { + "epoch": 0.83, + "learning_rate": 0.0003, + "loss": 2.1427, + "step": 1450 + }, + { + "epoch": 0.83, + "learning_rate": 0.0003, + "loss": 2.1219, + "step": 1451 + }, + { + "epoch": 0.83, + "learning_rate": 0.0003, + "loss": 2.1157, + "step": 1452 + }, + { + "epoch": 0.83, + "learning_rate": 0.0003, + "loss": 2.1307, + "step": 1453 + }, + { + "epoch": 0.83, + "learning_rate": 0.0003, + "loss": 2.1311, + "step": 1454 + }, + { + "epoch": 0.83, + "learning_rate": 0.0003, + "loss": 2.1089, + "step": 1455 + }, + { + "epoch": 0.83, + "learning_rate": 0.0003, + "loss": 2.1309, + "step": 1456 + }, + { + "epoch": 0.84, + "learning_rate": 0.0003, + "loss": 2.1111, + "step": 1457 + }, + { + "epoch": 0.84, + "learning_rate": 0.0003, + "loss": 2.1038, + "step": 1458 + }, + { + "epoch": 0.84, + "learning_rate": 0.0003, + "loss": 2.1064, + "step": 1459 + }, + { + "epoch": 0.84, + "learning_rate": 0.0003, + "loss": 2.1193, + "step": 1460 + }, + { + "epoch": 0.84, + "learning_rate": 0.0003, + "loss": 2.1028, + "step": 1461 + }, + { + "epoch": 0.84, + "learning_rate": 0.0003, + "loss": 2.1343, + "step": 1462 + }, + { + "epoch": 0.84, + "learning_rate": 0.0003, + "loss": 2.1226, + "step": 1463 + }, + { + "epoch": 0.84, + "learning_rate": 0.0003, + "loss": 2.1158, + "step": 1464 + }, + { + "epoch": 0.84, + "learning_rate": 0.0003, + "loss": 2.1023, + "step": 1465 + }, + { + "epoch": 0.84, + "learning_rate": 0.0003, + "loss": 2.1035, + "step": 1466 + }, + { + "epoch": 0.84, + "learning_rate": 0.0003, + "loss": 2.1, + "step": 1467 + }, + { + "epoch": 0.84, + "learning_rate": 0.0003, + "loss": 2.123, + "step": 1468 + }, + { + "epoch": 0.84, + "learning_rate": 0.0003, + "loss": 2.1239, + "step": 1469 + }, + { + "epoch": 0.84, + "learning_rate": 0.0003, + "loss": 2.1259, + "step": 1470 + }, + { + "epoch": 0.84, + "learning_rate": 0.0003, + "loss": 2.1345, + "step": 1471 + }, + { + "epoch": 0.84, + "learning_rate": 0.0003, + "loss": 2.1283, + "step": 1472 + }, + { + "epoch": 0.84, + "learning_rate": 0.0003, + "loss": 2.1331, + "step": 1473 + }, + { + "epoch": 0.85, + "learning_rate": 0.0003, + "loss": 2.1469, + "step": 1474 + }, + { + "epoch": 0.85, + "learning_rate": 0.0003, + "loss": 2.1234, + "step": 1475 + }, + { + "epoch": 0.85, + "learning_rate": 0.0003, + "loss": 2.1278, + "step": 1476 + }, + { + "epoch": 0.85, + "learning_rate": 0.0003, + "loss": 2.1248, + "step": 1477 + }, + { + "epoch": 0.85, + "learning_rate": 0.0003, + "loss": 2.1173, + "step": 1478 + }, + { + "epoch": 0.85, + "learning_rate": 0.0003, + "loss": 2.1208, + "step": 1479 + }, + { + "epoch": 0.85, + "learning_rate": 0.0003, + "loss": 2.1141, + "step": 1480 + }, + { + "epoch": 0.85, + "learning_rate": 0.0003, + "loss": 2.1334, + "step": 1481 + }, + { + "epoch": 0.85, + "learning_rate": 0.0003, + "loss": 2.1174, + "step": 1482 + }, + { + "epoch": 0.85, + "learning_rate": 0.0003, + "loss": 2.1263, + "step": 1483 + }, + { + "epoch": 0.85, + "learning_rate": 0.0003, + "loss": 2.1234, + "step": 1484 + }, + { + "epoch": 0.85, + "learning_rate": 0.0003, + "loss": 2.1093, + "step": 1485 + }, + { + "epoch": 0.85, + "learning_rate": 0.0003, + "loss": 2.1272, + "step": 1486 + }, + { + "epoch": 0.85, + "learning_rate": 0.0003, + "loss": 2.1221, + "step": 1487 + }, + { + "epoch": 0.85, + "learning_rate": 0.0003, + "loss": 2.1101, + "step": 1488 + }, + { + "epoch": 0.85, + "learning_rate": 0.0003, + "loss": 2.1221, + "step": 1489 + }, + { + "epoch": 0.85, + "learning_rate": 0.0003, + "loss": 2.1157, + "step": 1490 + }, + { + "epoch": 0.86, + "learning_rate": 0.0003, + "loss": 2.1143, + "step": 1491 + }, + { + "epoch": 0.86, + "learning_rate": 0.0003, + "loss": 2.1014, + "step": 1492 + }, + { + "epoch": 0.86, + "learning_rate": 0.0003, + "loss": 2.1176, + "step": 1493 + }, + { + "epoch": 0.86, + "learning_rate": 0.0003, + "loss": 2.1122, + "step": 1494 + }, + { + "epoch": 0.86, + "learning_rate": 0.0003, + "loss": 2.1272, + "step": 1495 + }, + { + "epoch": 0.86, + "learning_rate": 0.0003, + "loss": 2.1238, + "step": 1496 + }, + { + "epoch": 0.86, + "learning_rate": 0.0003, + "loss": 2.1307, + "step": 1497 + }, + { + "epoch": 0.86, + "learning_rate": 0.0003, + "loss": 2.1243, + "step": 1498 + }, + { + "epoch": 0.86, + "learning_rate": 0.0003, + "loss": 2.109, + "step": 1499 + }, + { + "epoch": 0.86, + "learning_rate": 0.0003, + "loss": 2.1406, + "step": 1500 + }, + { + "epoch": 0.86, + "learning_rate": 0.0003, + "loss": 2.1145, + "step": 1501 + }, + { + "epoch": 0.86, + "learning_rate": 0.0003, + "loss": 2.1289, + "step": 1502 + }, + { + "epoch": 0.86, + "learning_rate": 0.0003, + "loss": 2.1111, + "step": 1503 + }, + { + "epoch": 0.86, + "learning_rate": 0.0003, + "loss": 2.1185, + "step": 1504 + }, + { + "epoch": 0.86, + "learning_rate": 0.0003, + "loss": 2.1096, + "step": 1505 + }, + { + "epoch": 0.86, + "learning_rate": 0.0003, + "loss": 2.1165, + "step": 1506 + }, + { + "epoch": 0.86, + "learning_rate": 0.0003, + "loss": 2.1205, + "step": 1507 + }, + { + "epoch": 0.86, + "learning_rate": 0.0003, + "loss": 2.1051, + "step": 1508 + }, + { + "epoch": 0.87, + "learning_rate": 0.0003, + "loss": 2.1174, + "step": 1509 + }, + { + "epoch": 0.87, + "learning_rate": 0.0003, + "loss": 2.1244, + "step": 1510 + }, + { + "epoch": 0.87, + "learning_rate": 0.0003, + "loss": 2.1238, + "step": 1511 + }, + { + "epoch": 0.87, + "learning_rate": 0.0003, + "loss": 2.1461, + "step": 1512 + }, + { + "epoch": 0.87, + "learning_rate": 0.0003, + "loss": 2.1321, + "step": 1513 + }, + { + "epoch": 0.87, + "learning_rate": 0.0003, + "loss": 2.1167, + "step": 1514 + }, + { + "epoch": 0.87, + "learning_rate": 0.0003, + "loss": 2.1207, + "step": 1515 + }, + { + "epoch": 0.87, + "learning_rate": 0.0003, + "loss": 2.1178, + "step": 1516 + }, + { + "epoch": 0.87, + "learning_rate": 0.0003, + "loss": 2.1169, + "step": 1517 + }, + { + "epoch": 0.87, + "learning_rate": 0.0003, + "loss": 2.1224, + "step": 1518 + }, + { + "epoch": 0.87, + "learning_rate": 0.0003, + "loss": 2.126, + "step": 1519 + }, + { + "epoch": 0.87, + "learning_rate": 0.0003, + "loss": 2.1357, + "step": 1520 + }, + { + "epoch": 0.87, + "learning_rate": 0.0003, + "loss": 2.1021, + "step": 1521 + }, + { + "epoch": 0.87, + "learning_rate": 0.0003, + "loss": 2.1235, + "step": 1522 + }, + { + "epoch": 0.87, + "learning_rate": 0.0003, + "loss": 2.1014, + "step": 1523 + }, + { + "epoch": 0.87, + "learning_rate": 0.0003, + "loss": 2.1154, + "step": 1524 + }, + { + "epoch": 0.87, + "learning_rate": 0.0003, + "loss": 2.1017, + "step": 1525 + }, + { + "epoch": 0.88, + "learning_rate": 0.0003, + "loss": 2.1265, + "step": 1526 + }, + { + "epoch": 0.88, + "learning_rate": 0.0003, + "loss": 2.1073, + "step": 1527 + }, + { + "epoch": 0.88, + "learning_rate": 0.0003, + "loss": 2.1241, + "step": 1528 + }, + { + "epoch": 0.88, + "learning_rate": 0.0003, + "loss": 2.1, + "step": 1529 + }, + { + "epoch": 0.88, + "learning_rate": 0.0003, + "loss": 2.1161, + "step": 1530 + }, + { + "epoch": 0.88, + "eval_loss": 2.111154317855835, + "eval_runtime": 2007.8513, + "eval_samples_per_second": 52.536, + "eval_steps_per_second": 0.137, + "step": 1530 + }, + { + "epoch": 0.88, + "learning_rate": 0.0003, + "loss": 2.1178, + "step": 1531 + }, + { + "epoch": 0.88, + "learning_rate": 0.0003, + "loss": 2.1047, + "step": 1532 + }, + { + "epoch": 0.88, + "learning_rate": 0.0003, + "loss": 2.1157, + "step": 1533 + }, + { + "epoch": 0.88, + "learning_rate": 0.0003, + "loss": 2.1199, + "step": 1534 + }, + { + "epoch": 0.88, + "learning_rate": 0.0003, + "loss": 2.1061, + "step": 1535 + }, + { + "epoch": 0.88, + "learning_rate": 0.0003, + "loss": 2.113, + "step": 1536 + }, + { + "epoch": 0.88, + "learning_rate": 0.0003, + "loss": 2.1157, + "step": 1537 + }, + { + "epoch": 0.88, + "learning_rate": 0.0003, + "loss": 2.1158, + "step": 1538 + }, + { + "epoch": 0.88, + "learning_rate": 0.0003, + "loss": 2.1139, + "step": 1539 + }, + { + "epoch": 0.88, + "learning_rate": 0.0003, + "loss": 2.1134, + "step": 1540 + }, + { + "epoch": 0.88, + "learning_rate": 0.0003, + "loss": 2.1122, + "step": 1541 + }, + { + "epoch": 0.88, + "learning_rate": 0.0003, + "loss": 2.1133, + "step": 1542 + }, + { + "epoch": 0.88, + "learning_rate": 0.0003, + "loss": 2.0968, + "step": 1543 + }, + { + "epoch": 0.89, + "learning_rate": 0.0003, + "loss": 2.1158, + "step": 1544 + }, + { + "epoch": 0.89, + "learning_rate": 0.0003, + "loss": 2.1307, + "step": 1545 + }, + { + "epoch": 0.89, + "learning_rate": 0.0003, + "loss": 2.1241, + "step": 1546 + }, + { + "epoch": 0.89, + "learning_rate": 0.0003, + "loss": 2.1313, + "step": 1547 + }, + { + "epoch": 0.89, + "learning_rate": 0.0003, + "loss": 2.1053, + "step": 1548 + }, + { + "epoch": 0.89, + "learning_rate": 0.0003, + "loss": 2.1119, + "step": 1549 + }, + { + "epoch": 0.89, + "learning_rate": 0.0003, + "loss": 2.1158, + "step": 1550 + }, + { + "epoch": 0.89, + "learning_rate": 0.0003, + "loss": 2.0938, + "step": 1551 + }, + { + "epoch": 0.89, + "learning_rate": 0.0003, + "loss": 2.1003, + "step": 1552 + }, + { + "epoch": 0.89, + "learning_rate": 0.0003, + "loss": 2.1013, + "step": 1553 + }, + { + "epoch": 0.89, + "learning_rate": 0.0003, + "loss": 2.1123, + "step": 1554 + }, + { + "epoch": 0.89, + "learning_rate": 0.0003, + "loss": 2.1199, + "step": 1555 + }, + { + "epoch": 0.89, + "learning_rate": 0.0003, + "loss": 2.1085, + "step": 1556 + }, + { + "epoch": 0.89, + "learning_rate": 0.0003, + "loss": 2.1274, + "step": 1557 + }, + { + "epoch": 0.89, + "learning_rate": 0.0003, + "loss": 2.1309, + "step": 1558 + }, + { + "epoch": 0.89, + "learning_rate": 0.0003, + "loss": 2.1063, + "step": 1559 + }, + { + "epoch": 0.89, + "learning_rate": 0.0003, + "loss": 2.1166, + "step": 1560 + }, + { + "epoch": 0.9, + "learning_rate": 0.0003, + "loss": 2.118, + "step": 1561 + }, + { + "epoch": 0.9, + "learning_rate": 0.0003, + "loss": 2.1156, + "step": 1562 + }, + { + "epoch": 0.9, + "learning_rate": 0.0003, + "loss": 2.1041, + "step": 1563 + }, + { + "epoch": 0.9, + "learning_rate": 0.0003, + "loss": 2.1169, + "step": 1564 + }, + { + "epoch": 0.9, + "learning_rate": 0.0003, + "loss": 2.1094, + "step": 1565 + }, + { + "epoch": 0.9, + "learning_rate": 0.0003, + "loss": 2.1061, + "step": 1566 + }, + { + "epoch": 0.9, + "learning_rate": 0.0003, + "loss": 2.1051, + "step": 1567 + }, + { + "epoch": 0.9, + "learning_rate": 0.0003, + "loss": 2.1177, + "step": 1568 + }, + { + "epoch": 0.9, + "learning_rate": 0.0003, + "loss": 2.1097, + "step": 1569 + }, + { + "epoch": 0.9, + "learning_rate": 0.0003, + "loss": 2.1147, + "step": 1570 + }, + { + "epoch": 0.9, + "learning_rate": 0.0003, + "loss": 2.1149, + "step": 1571 + }, + { + "epoch": 0.9, + "learning_rate": 0.0003, + "loss": 2.1145, + "step": 1572 + }, + { + "epoch": 0.9, + "learning_rate": 0.0003, + "loss": 2.1145, + "step": 1573 + }, + { + "epoch": 0.9, + "learning_rate": 0.0003, + "loss": 2.0911, + "step": 1574 + }, + { + "epoch": 0.9, + "learning_rate": 0.0003, + "loss": 2.1107, + "step": 1575 + }, + { + "epoch": 0.9, + "learning_rate": 0.0003, + "loss": 2.1221, + "step": 1576 + }, + { + "epoch": 0.9, + "learning_rate": 0.0003, + "loss": 2.1101, + "step": 1577 + }, + { + "epoch": 0.9, + "learning_rate": 0.0003, + "loss": 2.1001, + "step": 1578 + }, + { + "epoch": 0.91, + "learning_rate": 0.0003, + "loss": 2.1279, + "step": 1579 + }, + { + "epoch": 0.91, + "learning_rate": 0.0003, + "loss": 2.1008, + "step": 1580 + }, + { + "epoch": 0.91, + "learning_rate": 0.0003, + "loss": 2.109, + "step": 1581 + }, + { + "epoch": 0.91, + "learning_rate": 0.0003, + "loss": 2.1299, + "step": 1582 + }, + { + "epoch": 0.91, + "learning_rate": 0.0003, + "loss": 2.1248, + "step": 1583 + }, + { + "epoch": 0.91, + "learning_rate": 0.0003, + "loss": 2.1016, + "step": 1584 + }, + { + "epoch": 0.91, + "learning_rate": 0.0003, + "loss": 2.1176, + "step": 1585 + }, + { + "epoch": 0.91, + "learning_rate": 0.0003, + "loss": 2.0852, + "step": 1586 + }, + { + "epoch": 0.91, + "learning_rate": 0.0003, + "loss": 2.1346, + "step": 1587 + }, + { + "epoch": 0.91, + "learning_rate": 0.0003, + "loss": 2.1239, + "step": 1588 + }, + { + "epoch": 0.91, + "learning_rate": 0.0003, + "loss": 2.1105, + "step": 1589 + }, + { + "epoch": 0.91, + "learning_rate": 0.0003, + "loss": 2.1187, + "step": 1590 + }, + { + "epoch": 0.91, + "learning_rate": 0.0003, + "loss": 2.1046, + "step": 1591 + }, + { + "epoch": 0.91, + "learning_rate": 0.0003, + "loss": 2.1144, + "step": 1592 + }, + { + "epoch": 0.91, + "learning_rate": 0.0003, + "loss": 2.1196, + "step": 1593 + }, + { + "epoch": 0.91, + "learning_rate": 0.0003, + "loss": 2.1185, + "step": 1594 + }, + { + "epoch": 0.91, + "learning_rate": 0.0003, + "loss": 2.1158, + "step": 1595 + }, + { + "epoch": 0.92, + "learning_rate": 0.0003, + "loss": 2.1172, + "step": 1596 + }, + { + "epoch": 0.92, + "learning_rate": 0.0003, + "loss": 2.1055, + "step": 1597 + }, + { + "epoch": 0.92, + "learning_rate": 0.0003, + "loss": 2.0725, + "step": 1598 + }, + { + "epoch": 0.92, + "learning_rate": 0.0003, + "loss": 2.0841, + "step": 1599 + }, + { + "epoch": 0.92, + "learning_rate": 0.0003, + "loss": 2.1122, + "step": 1600 + }, + { + "epoch": 0.92, + "learning_rate": 0.0003, + "loss": 2.0841, + "step": 1601 + }, + { + "epoch": 0.92, + "learning_rate": 0.0003, + "loss": 2.097, + "step": 1602 + }, + { + "epoch": 0.92, + "learning_rate": 0.0003, + "loss": 2.1278, + "step": 1603 + }, + { + "epoch": 0.92, + "learning_rate": 0.0003, + "loss": 2.0986, + "step": 1604 + }, + { + "epoch": 0.92, + "learning_rate": 0.0003, + "loss": 2.1073, + "step": 1605 + }, + { + "epoch": 0.92, + "learning_rate": 0.0003, + "loss": 2.1079, + "step": 1606 + }, + { + "epoch": 0.92, + "learning_rate": 0.0003, + "loss": 2.1238, + "step": 1607 + }, + { + "epoch": 0.92, + "learning_rate": 0.0003, + "loss": 2.1045, + "step": 1608 + }, + { + "epoch": 0.92, + "learning_rate": 0.0003, + "loss": 2.0995, + "step": 1609 + }, + { + "epoch": 0.92, + "learning_rate": 0.0003, + "loss": 2.104, + "step": 1610 + }, + { + "epoch": 0.92, + "learning_rate": 0.0003, + "loss": 2.1021, + "step": 1611 + }, + { + "epoch": 0.92, + "learning_rate": 0.0003, + "loss": 2.1108, + "step": 1612 + }, + { + "epoch": 0.92, + "learning_rate": 0.0003, + "loss": 2.1144, + "step": 1613 + }, + { + "epoch": 0.93, + "learning_rate": 0.0003, + "loss": 2.0884, + "step": 1614 + }, + { + "epoch": 0.93, + "learning_rate": 0.0003, + "loss": 2.1194, + "step": 1615 + }, + { + "epoch": 0.93, + "learning_rate": 0.0003, + "loss": 2.1144, + "step": 1616 + }, + { + "epoch": 0.93, + "learning_rate": 0.0003, + "loss": 2.1078, + "step": 1617 + }, + { + "epoch": 0.93, + "learning_rate": 0.0003, + "loss": 2.1052, + "step": 1618 + }, + { + "epoch": 0.93, + "learning_rate": 0.0003, + "loss": 2.0936, + "step": 1619 + }, + { + "epoch": 0.93, + "learning_rate": 0.0003, + "loss": 2.115, + "step": 1620 + }, + { + "epoch": 0.93, + "learning_rate": 0.0003, + "loss": 2.1047, + "step": 1621 + }, + { + "epoch": 0.93, + "learning_rate": 0.0003, + "loss": 2.0992, + "step": 1622 + }, + { + "epoch": 0.93, + "learning_rate": 0.0003, + "loss": 2.0912, + "step": 1623 + }, + { + "epoch": 0.93, + "learning_rate": 0.0003, + "loss": 2.1061, + "step": 1624 + }, + { + "epoch": 0.93, + "learning_rate": 0.0003, + "loss": 2.1101, + "step": 1625 + }, + { + "epoch": 0.93, + "learning_rate": 0.0003, + "loss": 2.087, + "step": 1626 + }, + { + "epoch": 0.93, + "learning_rate": 0.0003, + "loss": 2.0996, + "step": 1627 + }, + { + "epoch": 0.93, + "learning_rate": 0.0003, + "loss": 2.095, + "step": 1628 + }, + { + "epoch": 0.93, + "learning_rate": 0.0003, + "loss": 2.0939, + "step": 1629 + }, + { + "epoch": 0.93, + "learning_rate": 0.0003, + "loss": 2.1141, + "step": 1630 + }, + { + "epoch": 0.94, + "learning_rate": 0.0003, + "loss": 2.1141, + "step": 1631 + }, + { + "epoch": 0.94, + "learning_rate": 0.0003, + "loss": 2.1294, + "step": 1632 + }, + { + "epoch": 0.94, + "learning_rate": 0.0003, + "loss": 2.1119, + "step": 1633 + }, + { + "epoch": 0.94, + "learning_rate": 0.0003, + "loss": 2.1132, + "step": 1634 + }, + { + "epoch": 0.94, + "learning_rate": 0.0003, + "loss": 2.0918, + "step": 1635 + }, + { + "epoch": 0.94, + "learning_rate": 0.0003, + "loss": 2.1001, + "step": 1636 + }, + { + "epoch": 0.94, + "learning_rate": 0.0003, + "loss": 2.1078, + "step": 1637 + }, + { + "epoch": 0.94, + "learning_rate": 0.0003, + "loss": 2.0936, + "step": 1638 + }, + { + "epoch": 0.94, + "learning_rate": 0.0003, + "loss": 2.0997, + "step": 1639 + }, + { + "epoch": 0.94, + "learning_rate": 0.0003, + "loss": 2.1196, + "step": 1640 + }, + { + "epoch": 0.94, + "learning_rate": 0.0003, + "loss": 2.0979, + "step": 1641 + }, + { + "epoch": 0.94, + "learning_rate": 0.0003, + "loss": 2.1112, + "step": 1642 + }, + { + "epoch": 0.94, + "learning_rate": 0.0003, + "loss": 2.099, + "step": 1643 + }, + { + "epoch": 0.94, + "learning_rate": 0.0003, + "loss": 2.1064, + "step": 1644 + }, + { + "epoch": 0.94, + "learning_rate": 0.0003, + "loss": 2.1044, + "step": 1645 + }, + { + "epoch": 0.94, + "learning_rate": 0.0003, + "loss": 2.1001, + "step": 1646 + }, + { + "epoch": 0.94, + "learning_rate": 0.0003, + "loss": 2.1046, + "step": 1647 + }, + { + "epoch": 0.95, + "learning_rate": 0.0003, + "loss": 2.1, + "step": 1648 + }, + { + "epoch": 0.95, + "learning_rate": 0.0003, + "loss": 2.0863, + "step": 1649 + }, + { + "epoch": 0.95, + "learning_rate": 0.0003, + "loss": 2.1136, + "step": 1650 + }, + { + "epoch": 0.95, + "learning_rate": 0.0003, + "loss": 2.1117, + "step": 1651 + }, + { + "epoch": 0.95, + "learning_rate": 0.0003, + "loss": 2.1036, + "step": 1652 + }, + { + "epoch": 0.95, + "learning_rate": 0.0003, + "loss": 2.0984, + "step": 1653 + }, + { + "epoch": 0.95, + "learning_rate": 0.0003, + "loss": 2.1096, + "step": 1654 + }, + { + "epoch": 0.95, + "learning_rate": 0.0003, + "loss": 2.0905, + "step": 1655 + }, + { + "epoch": 0.95, + "learning_rate": 0.0003, + "loss": 2.1012, + "step": 1656 + }, + { + "epoch": 0.95, + "learning_rate": 0.0003, + "loss": 2.0815, + "step": 1657 + }, + { + "epoch": 0.95, + "learning_rate": 0.0003, + "loss": 2.0968, + "step": 1658 + }, + { + "epoch": 0.95, + "learning_rate": 0.0003, + "loss": 2.1017, + "step": 1659 + }, + { + "epoch": 0.95, + "learning_rate": 0.0003, + "loss": 2.0955, + "step": 1660 + }, + { + "epoch": 0.95, + "learning_rate": 0.0003, + "loss": 2.1256, + "step": 1661 + }, + { + "epoch": 0.95, + "learning_rate": 0.0003, + "loss": 2.088, + "step": 1662 + }, + { + "epoch": 0.95, + "learning_rate": 0.0003, + "loss": 2.1138, + "step": 1663 + }, + { + "epoch": 0.95, + "learning_rate": 0.0003, + "loss": 2.1038, + "step": 1664 + }, + { + "epoch": 0.95, + "learning_rate": 0.0003, + "loss": 2.0979, + "step": 1665 + }, + { + "epoch": 0.96, + "learning_rate": 0.0003, + "loss": 2.1158, + "step": 1666 + }, + { + "epoch": 0.96, + "learning_rate": 0.0003, + "loss": 2.1051, + "step": 1667 + }, + { + "epoch": 0.96, + "learning_rate": 0.0003, + "loss": 2.0946, + "step": 1668 + }, + { + "epoch": 0.96, + "learning_rate": 0.0003, + "loss": 2.1083, + "step": 1669 + }, + { + "epoch": 0.96, + "learning_rate": 0.0003, + "loss": 2.1234, + "step": 1670 + }, + { + "epoch": 0.96, + "learning_rate": 0.0003, + "loss": 2.1053, + "step": 1671 + }, + { + "epoch": 0.96, + "learning_rate": 0.0003, + "loss": 2.0784, + "step": 1672 + }, + { + "epoch": 0.96, + "learning_rate": 0.0003, + "loss": 2.1399, + "step": 1673 + }, + { + "epoch": 0.96, + "learning_rate": 0.0003, + "loss": 2.0868, + "step": 1674 + }, + { + "epoch": 0.96, + "learning_rate": 0.0003, + "loss": 2.0939, + "step": 1675 + }, + { + "epoch": 0.96, + "learning_rate": 0.0003, + "loss": 2.0894, + "step": 1676 + }, + { + "epoch": 0.96, + "learning_rate": 0.0003, + "loss": 2.0973, + "step": 1677 + }, + { + "epoch": 0.96, + "learning_rate": 0.0003, + "loss": 2.0952, + "step": 1678 + }, + { + "epoch": 0.96, + "learning_rate": 0.0003, + "loss": 2.0979, + "step": 1679 + }, + { + "epoch": 0.96, + "learning_rate": 0.0003, + "loss": 2.0836, + "step": 1680 + }, + { + "epoch": 0.96, + "learning_rate": 0.0003, + "loss": 2.1108, + "step": 1681 + }, + { + "epoch": 0.96, + "learning_rate": 0.0003, + "loss": 2.0991, + "step": 1682 + }, + { + "epoch": 0.97, + "learning_rate": 0.0003, + "loss": 2.113, + "step": 1683 + }, + { + "epoch": 0.97, + "learning_rate": 0.0003, + "loss": 2.1096, + "step": 1684 + }, + { + "epoch": 0.97, + "learning_rate": 0.0003, + "loss": 2.11, + "step": 1685 + }, + { + "epoch": 0.97, + "learning_rate": 0.0003, + "loss": 2.104, + "step": 1686 + }, + { + "epoch": 0.97, + "learning_rate": 0.0003, + "loss": 2.1066, + "step": 1687 + }, + { + "epoch": 0.97, + "learning_rate": 0.0003, + "loss": 2.1283, + "step": 1688 + }, + { + "epoch": 0.97, + "learning_rate": 0.0003, + "loss": 2.0972, + "step": 1689 + }, + { + "epoch": 0.97, + "learning_rate": 0.0003, + "loss": 2.1172, + "step": 1690 + }, + { + "epoch": 0.97, + "learning_rate": 0.0003, + "loss": 2.1033, + "step": 1691 + }, + { + "epoch": 0.97, + "learning_rate": 0.0003, + "loss": 2.0958, + "step": 1692 + }, + { + "epoch": 0.97, + "learning_rate": 0.0003, + "loss": 2.0938, + "step": 1693 + }, + { + "epoch": 0.97, + "learning_rate": 0.0003, + "loss": 2.1008, + "step": 1694 + }, + { + "epoch": 0.97, + "learning_rate": 0.0003, + "loss": 2.1055, + "step": 1695 + }, + { + "epoch": 0.97, + "learning_rate": 0.0003, + "loss": 2.1091, + "step": 1696 + }, + { + "epoch": 0.97, + "learning_rate": 0.0003, + "loss": 2.1006, + "step": 1697 + }, + { + "epoch": 0.97, + "learning_rate": 0.0003, + "loss": 2.1007, + "step": 1698 + }, + { + "epoch": 0.97, + "learning_rate": 0.0003, + "loss": 2.1, + "step": 1699 + }, + { + "epoch": 0.97, + "learning_rate": 0.0003, + "loss": 2.0997, + "step": 1700 + }, + { + "epoch": 0.97, + "eval_loss": 2.0928452014923096, + "eval_runtime": 2030.7734, + "eval_samples_per_second": 51.943, + "eval_steps_per_second": 0.135, + "step": 1700 + }, + { + "epoch": 0.98, + "learning_rate": 0.0003, + "loss": 2.0942, + "step": 1701 + }, + { + "epoch": 0.98, + "learning_rate": 0.0003, + "loss": 2.0994, + "step": 1702 + }, + { + "epoch": 0.98, + "learning_rate": 0.0003, + "loss": 2.0913, + "step": 1703 + }, + { + "epoch": 0.98, + "learning_rate": 0.0003, + "loss": 2.1155, + "step": 1704 + }, + { + "epoch": 0.98, + "learning_rate": 0.0003, + "loss": 2.0697, + "step": 1705 + }, + { + "epoch": 0.98, + "learning_rate": 0.0003, + "loss": 2.1021, + "step": 1706 + }, + { + "epoch": 0.98, + "learning_rate": 0.0003, + "loss": 2.1125, + "step": 1707 + }, + { + "epoch": 0.98, + "learning_rate": 0.0003, + "loss": 2.0771, + "step": 1708 + }, + { + "epoch": 0.98, + "learning_rate": 0.0003, + "loss": 2.1, + "step": 1709 + }, + { + "epoch": 0.98, + "learning_rate": 0.0003, + "loss": 2.1108, + "step": 1710 + }, + { + "epoch": 0.98, + "learning_rate": 0.0003, + "loss": 2.1063, + "step": 1711 + }, + { + "epoch": 0.98, + "learning_rate": 0.0003, + "loss": 2.087, + "step": 1712 + }, + { + "epoch": 0.98, + "learning_rate": 0.0003, + "loss": 2.0818, + "step": 1713 + }, + { + "epoch": 0.98, + "learning_rate": 0.0003, + "loss": 2.0886, + "step": 1714 + }, + { + "epoch": 0.98, + "learning_rate": 0.0003, + "loss": 2.09, + "step": 1715 + }, + { + "epoch": 0.98, + "learning_rate": 0.0003, + "loss": 2.0695, + "step": 1716 + }, + { + "epoch": 0.98, + "learning_rate": 0.0003, + "loss": 2.0957, + "step": 1717 + }, + { + "epoch": 0.99, + "learning_rate": 0.0003, + "loss": 2.1042, + "step": 1718 + }, + { + "epoch": 0.99, + "learning_rate": 0.0003, + "loss": 2.1017, + "step": 1719 + }, + { + "epoch": 0.99, + "learning_rate": 0.0003, + "loss": 2.0835, + "step": 1720 + }, + { + "epoch": 0.99, + "learning_rate": 0.0003, + "loss": 2.0986, + "step": 1721 + }, + { + "epoch": 0.99, + "learning_rate": 0.0003, + "loss": 2.0848, + "step": 1722 + }, + { + "epoch": 0.99, + "learning_rate": 0.0003, + "loss": 2.1111, + "step": 1723 + }, + { + "epoch": 0.99, + "learning_rate": 0.0003, + "loss": 2.0964, + "step": 1724 + }, + { + "epoch": 0.99, + "learning_rate": 0.0003, + "loss": 2.0967, + "step": 1725 + }, + { + "epoch": 0.99, + "learning_rate": 0.0003, + "loss": 2.0919, + "step": 1726 + }, + { + "epoch": 0.99, + "learning_rate": 0.0003, + "loss": 2.0785, + "step": 1727 + }, + { + "epoch": 0.99, + "learning_rate": 0.0003, + "loss": 2.0802, + "step": 1728 + }, + { + "epoch": 0.99, + "learning_rate": 0.0003, + "loss": 2.1207, + "step": 1729 + }, + { + "epoch": 0.99, + "learning_rate": 0.0003, + "loss": 2.0939, + "step": 1730 + }, + { + "epoch": 0.99, + "learning_rate": 0.0003, + "loss": 2.089, + "step": 1731 + }, + { + "epoch": 0.99, + "learning_rate": 0.0003, + "loss": 2.0968, + "step": 1732 + }, + { + "epoch": 0.99, + "learning_rate": 0.0003, + "loss": 2.0938, + "step": 1733 + }, + { + "epoch": 0.99, + "learning_rate": 0.0003, + "loss": 2.1082, + "step": 1734 + }, + { + "epoch": 0.99, + "learning_rate": 0.0003, + "loss": 2.1056, + "step": 1735 + }, + { + "epoch": 1.0, + "learning_rate": 0.0003, + "loss": 2.0906, + "step": 1736 + }, + { + "epoch": 1.0, + "learning_rate": 0.0003, + "loss": 2.084, + "step": 1737 + }, + { + "epoch": 1.0, + "learning_rate": 0.0003, + "loss": 2.0723, + "step": 1738 + }, + { + "epoch": 1.0, + "learning_rate": 0.0003, + "loss": 2.1074, + "step": 1739 + }, + { + "epoch": 1.0, + "learning_rate": 0.0003, + "loss": 2.0977, + "step": 1740 + }, + { + "epoch": 1.0, + "learning_rate": 0.0003, + "loss": 2.099, + "step": 1741 + }, + { + "epoch": 1.0, + "learning_rate": 0.0003, + "loss": 2.1049, + "step": 1742 + }, + { + "epoch": 1.0, + "learning_rate": 0.0003, + "loss": 2.0825, + "step": 1743 + }, + { + "epoch": 1.0, + "step": 1743, + "total_flos": 8.497123594010729e+19, + "train_loss": 0.05167265042312105, + "train_runtime": 3158.899, + "train_samples_per_second": 635.936, + "train_steps_per_second": 0.552 + } + ], + "max_steps": 1743, + "num_train_epochs": 1, + "total_flos": 8.497123594010729e+19, + "trial_name": null, + "trial_params": null +}