diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6053 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 6.998144712430427,
+  "eval_steps": 20,
+  "global_step": 943,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 1.2998,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 1.2882,
+      "step": 2
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 9e-07,
+      "loss": 1.3159,
+      "step": 3
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 1.2737,
+      "step": 4
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.5e-06,
+      "loss": 1.2977,
+      "step": 5
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.8e-06,
+      "loss": 1.2353,
+      "step": 6
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 2.1e-06,
+      "loss": 1.2186,
+      "step": 7
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 1.1946,
+      "step": 8
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.7e-06,
+      "loss": 1.1438,
+      "step": 9
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 3e-06,
+      "loss": 1.1702,
+      "step": 10
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.999995815365561e-06,
+      "loss": 1.155,
+      "step": 11
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.9999832614855924e-06,
+      "loss": 1.1791,
+      "step": 12
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.9999623384301386e-06,
+      "loss": 1.0963,
+      "step": 13
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2.99993304631594e-06,
+      "loss": 1.1439,
+      "step": 14
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 2.9998953853064323e-06,
+      "loss": 1.1887,
+      "step": 15
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 2.9998493556117455e-06,
+      "loss": 1.1146,
+      "step": 16
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.9997949574887036e-06,
+      "loss": 1.125,
+      "step": 17
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 2.9997321912408206e-06,
+      "loss": 1.1295,
+      "step": 18
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 2.999661057218302e-06,
+      "loss": 1.1022,
+      "step": 19
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 2.999581555818041e-06,
+      "loss": 1.1109,
+      "step": 20
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 1.2603144645690918,
+      "eval_runtime": 1.8751,
+      "eval_samples_per_second": 35.199,
+      "eval_steps_per_second": 3.2,
+      "step": 20
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.9994936874836168e-06,
+      "loss": 1.1012,
+      "step": 21
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 2.999397452705292e-06,
+      "loss": 1.0692,
+      "step": 22
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 2.9992928520200088e-06,
+      "loss": 1.0678,
+      "step": 23
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 2.9991798860113893e-06,
+      "loss": 1.0603,
+      "step": 24
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.9990585553097277e-06,
+      "loss": 1.0493,
+      "step": 25
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 2.998928860591991e-06,
+      "loss": 1.0547,
+      "step": 26
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 2.998790802581812e-06,
+      "loss": 1.077,
+      "step": 27
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 2.9986443820494873e-06,
+      "loss": 1.0809,
+      "step": 28
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.998489599811972e-06,
+      "loss": 1.0791,
+      "step": 29
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 2.9983264567328758e-06,
+      "loss": 1.0607,
+      "step": 30
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 2.998154953722457e-06,
+      "loss": 1.0367,
+      "step": 31
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.9979750917376195e-06,
+      "loss": 1.0573,
+      "step": 32
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 2.9977868717819053e-06,
+      "loss": 1.0617,
+      "step": 33
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 2.99759029490549e-06,
+      "loss": 1.0861,
+      "step": 34
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 2.9973853622051763e-06,
+      "loss": 1.0268,
+      "step": 35
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.9971720748243892e-06,
+      "loss": 1.0708,
+      "step": 36
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 2.9969504339531687e-06,
+      "loss": 1.0817,
+      "step": 37
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 2.996720440828162e-06,
+      "loss": 1.0567,
+      "step": 38
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 2.996482096732619e-06,
+      "loss": 1.0048,
+      "step": 39
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.9962354029963836e-06,
+      "loss": 1.0765,
+      "step": 40
+    },
+    {
+      "epoch": 0.3,
+      "eval_loss": 1.204389214515686,
+      "eval_runtime": 1.8834,
+      "eval_samples_per_second": 35.043,
+      "eval_steps_per_second": 3.186,
+      "step": 40
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 2.995980360995887e-06,
+      "loss": 1.0945,
+      "step": 41
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 2.9957169721541385e-06,
+      "loss": 1.0433,
+      "step": 42
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 2.99544523794072e-06,
+      "loss": 1.0581,
+      "step": 43
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.9951651598717757e-06,
+      "loss": 1.0783,
+      "step": 44
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 2.994876739510005e-06,
+      "loss": 1.014,
+      "step": 45
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 2.9945799784646523e-06,
+      "loss": 1.045,
+      "step": 46
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 2.9942748783915007e-06,
+      "loss": 1.052,
+      "step": 47
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.993961440992859e-06,
+      "loss": 1.0564,
+      "step": 48
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 2.9936396680175547e-06,
+      "loss": 1.0365,
+      "step": 49
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 2.9933095612609256e-06,
+      "loss": 1.0159,
+      "step": 50
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 2.992971122564805e-06,
+      "loss": 1.0114,
+      "step": 51
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.9926243538175173e-06,
+      "loss": 1.0261,
+      "step": 52
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 2.992269256953862e-06,
+      "loss": 1.039,
+      "step": 53
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 2.9919058339551067e-06,
+      "loss": 1.0318,
+      "step": 54
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 2.991534086848975e-06,
+      "loss": 1.0412,
+      "step": 55
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.9911540177096343e-06,
+      "loss": 1.0141,
+      "step": 56
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 2.9907656286576856e-06,
+      "loss": 1.0176,
+      "step": 57
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 2.9903689218601497e-06,
+      "loss": 1.0551,
+      "step": 58
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 2.9899638995304575e-06,
+      "loss": 1.0776,
+      "step": 59
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.9895505639284363e-06,
+      "loss": 0.9753,
+      "step": 60
+    },
+    {
+      "epoch": 0.45,
+      "eval_loss": 1.188205361366272,
+      "eval_runtime": 1.8807,
+      "eval_samples_per_second": 35.092,
+      "eval_steps_per_second": 3.19,
+      "step": 60
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 2.989128917360297e-06,
+      "loss": 1.0406,
+      "step": 61
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 2.9886989621786224e-06,
+      "loss": 1.0383,
+      "step": 62
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.9882607007823527e-06,
+      "loss": 1.0095,
+      "step": 63
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 2.9878141356167727e-06,
+      "loss": 1.0179,
+      "step": 64
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 2.9873592691734985e-06,
+      "loss": 1.0344,
+      "step": 65
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 2.986896103990463e-06,
+      "loss": 1.0537,
+      "step": 66
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.986424642651902e-06,
+      "loss": 1.0404,
+      "step": 67
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 2.9859448877883407e-06,
+      "loss": 1.0181,
+      "step": 68
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 2.985456842076577e-06,
+      "loss": 1.0378,
+      "step": 69
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 2.9849605082396677e-06,
+      "loss": 1.0473,
+      "step": 70
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.984455889046914e-06,
+      "loss": 1.0301,
+      "step": 71
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 2.983942987313845e-06,
+      "loss": 1.001,
+      "step": 72
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 2.9834218059022026e-06,
+      "loss": 1.0133,
+      "step": 73
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 2.9828923477199252e-06,
+      "loss": 1.0371,
+      "step": 74
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.9823546157211302e-06,
+      "loss": 1.0751,
+      "step": 75
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 2.981808612906102e-06,
+      "loss": 1.0052,
+      "step": 76
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 2.9812543423212685e-06,
+      "loss": 1.0069,
+      "step": 77
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 2.98069180705919e-06,
+      "loss": 0.9995,
+      "step": 78
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.9801210102585393e-06,
+      "loss": 1.0232,
+      "step": 79
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 2.9795419551040837e-06,
+      "loss": 1.0347,
+      "step": 80
+    },
+    {
+      "epoch": 0.59,
+      "eval_loss": 1.1748546361923218,
+      "eval_runtime": 1.8792,
+      "eval_samples_per_second": 35.122,
+      "eval_steps_per_second": 3.193,
+      "step": 80
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 2.9789546448266694e-06,
+      "loss": 1.0095,
+      "step": 81
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 2.978359082703201e-06,
+      "loss": 1.0476,
+      "step": 82
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.9777552720566255e-06,
+      "loss": 1.0336,
+      "step": 83
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 2.977143216255911e-06,
+      "loss": 1.0061,
+      "step": 84
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 2.976522918716032e-06,
+      "loss": 0.9965,
+      "step": 85
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 2.9758943828979446e-06,
+      "loss": 1.0092,
+      "step": 86
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.9752576123085738e-06,
+      "loss": 1.0331,
+      "step": 87
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 2.9746126105007884e-06,
+      "loss": 0.9892,
+      "step": 88
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 2.973959381073384e-06,
+      "loss": 1.0255,
+      "step": 89
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 2.9732979276710633e-06,
+      "loss": 1.0191,
+      "step": 90
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.972628253984413e-06,
+      "loss": 1.0334,
+      "step": 91
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 2.9719503637498856e-06,
+      "loss": 1.0312,
+      "step": 92
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 2.9712642607497798e-06,
+      "loss": 1.0269,
+      "step": 93
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 2.970569948812214e-06,
+      "loss": 1.0559,
+      "step": 94
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.969867431811112e-06,
+      "loss": 1.0286,
+      "step": 95
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 2.9691567136661748e-06,
+      "loss": 0.9996,
+      "step": 96
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 2.968437798342864e-06,
+      "loss": 1.0735,
+      "step": 97
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.967710689852377e-06,
+      "loss": 1.0005,
+      "step": 98
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 2.966975392251624e-06,
+      "loss": 1.0126,
+      "step": 99
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 2.9662319096432077e-06,
+      "loss": 1.0253,
+      "step": 100
+    },
+    {
+      "epoch": 0.74,
+      "eval_loss": 1.1693907976150513,
+      "eval_runtime": 1.8764,
+      "eval_samples_per_second": 35.175,
+      "eval_steps_per_second": 3.198,
+      "step": 100
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.9654802461753993e-06,
+      "loss": 1.0553,
+      "step": 101
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.964720406042113e-06,
+      "loss": 1.0182,
+      "step": 102
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.9639523934828875e-06,
+      "loss": 1.0239,
+      "step": 103
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.9631762127828583e-06,
+      "loss": 1.0104,
+      "step": 104
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.9623918682727352e-06,
+      "loss": 0.9997,
+      "step": 105
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.9615993643287788e-06,
+      "loss": 1.0302,
+      "step": 106
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.9607987053727745e-06,
+      "loss": 1.05,
+      "step": 107
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.959989895872009e-06,
+      "loss": 0.9937,
+      "step": 108
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 2.959172940339245e-06,
+      "loss": 1.0005,
+      "step": 109
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.958347843332696e-06,
+      "loss": 0.9855,
+      "step": 110
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 2.9575146094560015e-06,
+      "loss": 1.0268,
+      "step": 111
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.9566732433582005e-06,
+      "loss": 0.9781,
+      "step": 112
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.9558237497337052e-06,
+      "loss": 0.9862,
+      "step": 113
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.9549661333222765e-06,
+      "loss": 1.0124,
+      "step": 114
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.9541003989089957e-06,
+      "loss": 1.0266,
+      "step": 115
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.953226551324239e-06,
+      "loss": 1.0128,
+      "step": 116
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 2.9523445954436498e-06,
+      "loss": 1.0174,
+      "step": 117
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.9514545361881123e-06,
+      "loss": 0.9973,
+      "step": 118
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 2.950556378523723e-06,
+      "loss": 1.0416,
+      "step": 119
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 2.949650127461764e-06,
+      "loss": 0.996,
+      "step": 120
+    },
+    {
+      "epoch": 0.89,
+      "eval_loss": 1.16244637966156,
+      "eval_runtime": 1.8733,
+      "eval_samples_per_second": 35.231,
+      "eval_steps_per_second": 3.203,
+      "step": 120
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 2.9487357880586744e-06,
+      "loss": 1.049,
+      "step": 121
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.9478133654160233e-06,
+      "loss": 1.059,
+      "step": 122
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 2.946882864680478e-06,
+      "loss": 1.0366,
+      "step": 123
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.94594429104378e-06,
+      "loss": 1.0233,
+      "step": 124
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.944997649742713e-06,
+      "loss": 1.0102,
+      "step": 125
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.9440429460590727e-06,
+      "loss": 1.0632,
+      "step": 126
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.9430801853196413e-06,
+      "loss": 1.0004,
+      "step": 127
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.9421093728961544e-06,
+      "loss": 0.961,
+      "step": 128
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.9411305142052725e-06,
+      "loss": 0.9959,
+      "step": 129
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.940143614708549e-06,
+      "loss": 1.0057,
+      "step": 130
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 2.9391486799124027e-06,
+      "loss": 1.0646,
+      "step": 131
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.938145715368085e-06,
+      "loss": 0.9998,
+      "step": 132
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.9371347266716486e-06,
+      "loss": 1.0092,
+      "step": 133
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.9361157194639183e-06,
+      "loss": 1.0177,
+      "step": 134
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.935088699430457e-06,
+      "loss": 0.9934,
+      "step": 135
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9340536723015366e-06,
+      "loss": 0.969,
+      "step": 136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.933010643852103e-06,
+      "loss": 0.9459,
+      "step": 137
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9319596199017477e-06,
+      "loss": 0.883,
+      "step": 138
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 2.9309006063146715e-06,
+      "loss": 0.9093,
+      "step": 139
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 2.9298336089996537e-06,
+      "loss": 0.9029,
+      "step": 140
+    },
+    {
+      "epoch": 1.04,
+      "eval_loss": 1.1602718830108643,
+      "eval_runtime": 1.887,
+      "eval_samples_per_second": 34.977,
+      "eval_steps_per_second": 3.18,
+      "step": 140
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.92875863391002e-06,
+      "loss": 0.9516,
+      "step": 141
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 2.9276756870436073e-06,
+      "loss": 0.9498,
+      "step": 142
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 2.9265847744427307e-06,
+      "loss": 0.9322,
+      "step": 143
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 2.925485902194151e-06,
+      "loss": 0.9398,
+      "step": 144
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.9243790764290398e-06,
+      "loss": 0.9348,
+      "step": 145
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 2.9232643033229455e-06,
+      "loss": 0.9251,
+      "step": 146
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 2.9221415890957587e-06,
+      "loss": 0.9558,
+      "step": 147
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 2.921010940011677e-06,
+      "loss": 0.9426,
+      "step": 148
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.9198723623791723e-06,
+      "loss": 0.9465,
+      "step": 149
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 2.918725862550952e-06,
+      "loss": 0.9072,
+      "step": 150
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 2.917571446923927e-06,
+      "loss": 0.9049,
+      "step": 151
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 2.9164091219391724e-06,
+      "loss": 0.9459,
+      "step": 152
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.9152388940818966e-06,
+      "loss": 0.9159,
+      "step": 153
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 2.9140607698814e-06,
+      "loss": 0.919,
+      "step": 154
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 2.9128747559110412e-06,
+      "loss": 0.948,
+      "step": 155
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 2.9116808587882004e-06,
+      "loss": 0.882,
+      "step": 156
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.910479085174242e-06,
+      "loss": 0.9349,
+      "step": 157
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 2.9092694417744764e-06,
+      "loss": 0.9317,
+      "step": 158
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 2.908051935338124e-06,
+      "loss": 0.9657,
+      "step": 159
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.906826572658278e-06,
+      "loss": 0.9011,
+      "step": 160
+    },
+    {
+      "epoch": 1.19,
+      "eval_loss": 1.156259298324585,
+      "eval_runtime": 1.8951,
+      "eval_samples_per_second": 34.826,
+      "eval_steps_per_second": 3.166,
+      "step": 160
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 2.9055933605718644e-06,
+      "loss": 0.921,
+      "step": 161
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 2.904352305959606e-06,
+      "loss": 0.9666,
+      "step": 162
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 2.9031034157459823e-06,
+      "loss": 0.8779,
+      "step": 163
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.9018466968991914e-06,
+      "loss": 0.9605,
+      "step": 164
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 2.9005821564311125e-06,
+      "loss": 0.9313,
+      "step": 165
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 2.8993098013972654e-06,
+      "loss": 0.9219,
+      "step": 166
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 2.898029638896771e-06,
+      "loss": 0.9332,
+      "step": 167
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.8967416760723113e-06,
+      "loss": 0.9591,
+      "step": 168
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 2.895445920110092e-06,
+      "loss": 0.8924,
+      "step": 169
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 2.8941423782397986e-06,
+      "loss": 0.9655,
+      "step": 170
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 2.8928310577345605e-06,
+      "loss": 0.9281,
+      "step": 171
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.8915119659109067e-06,
+      "loss": 0.9103,
+      "step": 172
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 2.890185110128727e-06,
+      "loss": 0.88,
+      "step": 173
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 2.8888504977912286e-06,
+      "loss": 0.9418,
+      "step": 174
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 2.887508136344899e-06,
+      "loss": 0.9366,
+      "step": 175
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.8861580332794604e-06,
+      "loss": 0.9156,
+      "step": 176
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 2.88480019612783e-06,
+      "loss": 0.9382,
+      "step": 177
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 2.8834346324660774e-06,
+      "loss": 0.8794,
+      "step": 178
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 2.8820613499133817e-06,
+      "loss": 0.9314,
+      "step": 179
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.8806803561319903e-06,
+      "loss": 0.9196,
+      "step": 180
+    },
+    {
+      "epoch": 1.34,
+      "eval_loss": 1.15269935131073,
+      "eval_runtime": 1.8922,
+      "eval_samples_per_second": 34.88,
+      "eval_steps_per_second": 3.171,
+      "step": 180
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 2.879291658827176e-06,
+      "loss": 0.8976,
+      "step": 181
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 2.8778952657471924e-06,
+      "loss": 0.9203,
+      "step": 182
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 2.8764911846832326e-06,
+      "loss": 0.9648,
+      "step": 183
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.875079423469384e-06,
+      "loss": 0.9633,
+      "step": 184
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 2.8736599899825863e-06,
+      "loss": 0.9179,
+      "step": 185
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 2.8722328921425862e-06,
+      "loss": 0.887,
+      "step": 186
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 2.8707981379118954e-06,
+      "loss": 0.9141,
+      "step": 187
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.8693557352957416e-06,
+      "loss": 0.9819,
+      "step": 188
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 2.8679056923420294e-06,
+      "loss": 0.9378,
+      "step": 189
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 2.866448017141291e-06,
+      "loss": 0.9547,
+      "step": 190
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.864982717826644e-06,
+      "loss": 0.9451,
+      "step": 191
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 2.8635098025737437e-06,
+      "loss": 0.9197,
+      "step": 192
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 2.8620292796007402e-06,
+      "loss": 0.9473,
+      "step": 193
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 2.8605411571682297e-06,
+      "loss": 0.8969,
+      "step": 194
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.8590454435792094e-06,
+      "loss": 0.9269,
+      "step": 195
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 2.8575421471790334e-06,
+      "loss": 0.9612,
+      "step": 196
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 2.8560312763553615e-06,
+      "loss": 0.9374,
+      "step": 197
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 2.8545128395381175e-06,
+      "loss": 0.9385,
+      "step": 198
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.8529868451994386e-06,
+      "loss": 0.9177,
+      "step": 199
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 2.851453301853629e-06,
+      "loss": 0.9683,
+      "step": 200
+    },
+    {
+      "epoch": 1.48,
+      "eval_loss": 1.1511304378509521,
+      "eval_runtime": 1.8899,
+      "eval_samples_per_second": 34.922,
+      "eval_steps_per_second": 3.175,
+      "step": 200
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 2.849912218057113e-06,
+      "loss": 0.9105,
+      "step": 201
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 2.8483636024083876e-06,
+      "loss": 0.9264,
+      "step": 202
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.846807463547973e-06,
+      "loss": 0.8927,
+      "step": 203
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 2.845243810158365e-06,
+      "loss": 0.9575,
+      "step": 204
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 2.8436726509639884e-06,
+      "loss": 1.0133,
+      "step": 205
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 2.842093994731145e-06,
+      "loss": 0.9193,
+      "step": 206
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.840507850267968e-06,
+      "loss": 0.8911,
+      "step": 207
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 2.838914226424369e-06,
+      "loss": 0.9449,
+      "step": 208
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 2.8373131320919934e-06,
+      "loss": 0.9166,
+      "step": 209
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 2.835704576204167e-06,
+      "loss": 0.9177,
+      "step": 210
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.834088567735847e-06,
+      "loss": 0.8971,
+      "step": 211
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 2.8324651157035737e-06,
+      "loss": 0.9655,
+      "step": 212
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 2.830834229165418e-06,
+      "loss": 0.9564,
+      "step": 213
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 2.8291959172209315e-06,
+      "loss": 0.939,
+      "step": 214
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.8275501890110964e-06,
+      "loss": 0.9139,
+      "step": 215
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 2.8258970537182743e-06,
+      "loss": 0.9264,
+      "step": 216
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 2.824236520566154e-06,
+      "loss": 0.9819,
+      "step": 217
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 2.822568598819702e-06,
+      "loss": 0.9442,
+      "step": 218
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.820893297785107e-06,
+      "loss": 0.9502,
+      "step": 219
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 2.8192106268097337e-06,
+      "loss": 0.9787,
+      "step": 220
+    },
+    {
+      "epoch": 1.63,
+      "eval_loss": 1.1459424495697021,
+      "eval_runtime": 1.8901,
+      "eval_samples_per_second": 34.919,
+      "eval_steps_per_second": 3.174,
+      "step": 220
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 2.8175205952820653e-06,
+      "loss": 0.9031,
+      "step": 221
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.8158232126316546e-06,
+      "loss": 0.9335,
+      "step": 222
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 2.814118488329068e-06,
+      "loss": 0.9318,
+      "step": 223
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 2.812406431885838e-06,
+      "loss": 0.9363,
+      "step": 224
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 2.8106870528544043e-06,
+      "loss": 0.8961,
+      "step": 225
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.8089603608280635e-06,
+      "loss": 0.9538,
+      "step": 226
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 2.8072263654409156e-06,
+      "loss": 0.9078,
+      "step": 227
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 2.8054850763678107e-06,
+      "loss": 0.9695,
+      "step": 228
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 2.8037365033242915e-06,
+      "loss": 0.9252,
+      "step": 229
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.8019806560665448e-06,
+      "loss": 0.8447,
+      "step": 230
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 2.800217544391341e-06,
+      "loss": 0.9675,
+      "step": 231
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 2.7984471781359856e-06,
+      "loss": 0.89,
+      "step": 232
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 2.7966695671782577e-06,
+      "loss": 0.8937,
+      "step": 233
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.794884721436361e-06,
+      "loss": 0.8984,
+      "step": 234
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 2.7930926508688646e-06,
+      "loss": 0.9553,
+      "step": 235
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 2.7912933654746483e-06,
+      "loss": 0.9257,
+      "step": 236
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 2.7894868752928486e-06,
+      "loss": 0.9379,
+      "step": 237
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.7876731904027993e-06,
+      "loss": 0.9437,
+      "step": 238
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 2.7858523209239784e-06,
+      "loss": 0.9327,
+      "step": 239
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 2.7840242770159503e-06,
+      "loss": 0.9435,
+      "step": 240
+    },
+    {
+      "epoch": 1.78,
+      "eval_loss": 1.1447596549987793,
+      "eval_runtime": 1.8875,
+      "eval_samples_per_second": 34.967,
+      "eval_steps_per_second": 3.179,
+      "step": 240
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 2.7821890688783088e-06,
+      "loss": 0.8707,
+      "step": 241
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.7803467067506204e-06,
+      "loss": 0.9592,
+      "step": 242
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 2.778497200912368e-06,
+      "loss": 0.8997,
+      "step": 243
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 2.776640561682894e-06,
+      "loss": 0.9318,
+      "step": 244
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 2.774776799421338e-06,
+      "loss": 0.9361,
+      "step": 245
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.7729059245265865e-06,
+      "loss": 0.9341,
+      "step": 246
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 2.7710279474372093e-06,
+      "loss": 0.9429,
+      "step": 247
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 2.769142878631403e-06,
+      "loss": 0.9573,
+      "step": 248
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 2.7672507286269333e-06,
+      "loss": 0.9454,
+      "step": 249
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.7653515079810744e-06,
+      "loss": 0.9435,
+      "step": 250
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 2.7634452272905513e-06,
+      "loss": 0.9099,
+      "step": 251
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 2.761531897191482e-06,
+      "loss": 0.9693,
+      "step": 252
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 2.7596115283593157e-06,
+      "loss": 0.9181,
+      "step": 253
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 2.7576841315087746e-06,
+      "loss": 0.9396,
+      "step": 254
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 2.7557497173937928e-06,
+      "loss": 0.9125,
+      "step": 255
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 2.7538082968074594e-06,
+      "loss": 0.9439,
+      "step": 256
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.751859880581954e-06,
+      "loss": 0.8935,
+      "step": 257
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 2.74990447958849e-06,
+      "loss": 0.9319,
+      "step": 258
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 2.747942104737252e-06,
+      "loss": 0.9317,
+      "step": 259
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 2.7459727669773343e-06,
+      "loss": 0.9165,
+      "step": 260
+    },
+    {
+      "epoch": 1.93,
+      "eval_loss": 1.1428970098495483,
+      "eval_runtime": 1.8802,
+      "eval_samples_per_second": 35.102,
+      "eval_steps_per_second": 3.191,
+      "step": 260
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 2.7439964772966828e-06,
+      "loss": 0.9671,
+      "step": 261
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 2.74201324672203e-06,
+      "loss": 0.9457,
+      "step": 262
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 2.740023086318836e-06,
+      "loss": 0.9174,
+      "step": 263
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 2.738026007191226e-06,
+      "loss": 0.9464,
+      "step": 264
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.736022020481928e-06,
+      "loss": 0.9291,
+      "step": 265
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 2.7340111373722105e-06,
+      "loss": 0.9212,
+      "step": 266
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 2.7319933690818224e-06,
+      "loss": 0.9734,
+      "step": 267
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 2.729968726868926e-06,
+      "loss": 0.9395,
+      "step": 268
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.727937222030039e-06,
+      "loss": 0.9272,
+      "step": 269
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 2.725898865899967e-06,
+      "loss": 0.8969,
+      "step": 270
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 2.723853669851745e-06,
+      "loss": 0.8236,
+      "step": 271
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 2.721801645296569e-06,
+      "loss": 0.8658,
+      "step": 272
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.719742803683737e-06,
+      "loss": 0.8593,
+      "step": 273
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 2.7176771565005804e-06,
+      "loss": 0.8553,
+      "step": 274
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 2.7156047152724047e-06,
+      "loss": 0.776,
+      "step": 275
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 2.713525491562421e-06,
+      "loss": 0.8605,
+      "step": 276
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.711439496971685e-06,
+      "loss": 0.8427,
+      "step": 277
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 2.7093467431390293e-06,
+      "loss": 0.8234,
+      "step": 278
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 2.707247241741e-06,
+      "loss": 0.8284,
+      "step": 279
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 2.705141004491792e-06,
+      "loss": 0.8777,
+      "step": 280
+    },
+    {
+      "epoch": 2.08,
+      "eval_loss": 1.1463375091552734,
+      "eval_runtime": 1.8701,
+      "eval_samples_per_second": 35.293,
+      "eval_steps_per_second": 3.208,
+      "step": 280
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.7030280431431824e-06,
+      "loss": 0.8527,
+      "step": 281
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 2.7009083694844656e-06,
+      "loss": 0.8092,
+      "step": 282
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 2.698781995342387e-06,
+      "loss": 0.8572,
+      "step": 283
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 2.696648932581079e-06,
+      "loss": 0.8329,
+      "step": 284
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.6945091931019912e-06,
+      "loss": 0.8427,
+      "step": 285
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 2.6923627888438277e-06,
+      "loss": 0.8342,
+      "step": 286
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 2.690209731782477e-06,
+      "loss": 0.8691,
+      "step": 287
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.688050033930949e-06,
+      "loss": 0.8164,
+      "step": 288
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 2.6858837073393047e-06,
+      "loss": 0.8888,
+      "step": 289
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 2.6837107640945906e-06,
+      "loss": 0.8631,
+      "step": 290
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 2.6815312163207708e-06,
+      "loss": 0.8379,
+      "step": 291
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.67934507617866e-06,
+      "loss": 0.8861,
+      "step": 292
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 2.6771523558658536e-06,
+      "loss": 0.8741,
+      "step": 293
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 2.674953067616663e-06,
+      "loss": 0.8474,
+      "step": 294
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 2.6727472237020448e-06,
+      "loss": 0.8773,
+      "step": 295
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.6705348364295324e-06,
+      "loss": 0.8139,
+      "step": 296
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 2.6683159181431688e-06,
+      "loss": 0.8672,
+      "step": 297
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 2.666090481223436e-06,
+      "loss": 0.7782,
+      "step": 298
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 2.6638585380871882e-06,
+      "loss": 0.8846,
+      "step": 299
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.6616201011875793e-06,
+      "loss": 0.8255,
+      "step": 300
+    },
+    {
+      "epoch": 2.23,
+      "eval_loss": 1.1532946825027466,
+      "eval_runtime": 1.8972,
+      "eval_samples_per_second": 34.789,
+      "eval_steps_per_second": 3.163,
+      "step": 300
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 2.659375183013996e-06,
+      "loss": 0.8399,
+      "step": 301
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 2.657123796091989e-06,
+      "loss": 0.8808,
+      "step": 302
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 2.654865952983198e-06,
+      "loss": 0.8478,
+      "step": 303
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.652601666285289e-06,
+      "loss": 0.8781,
+      "step": 304
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 2.6503309486318763e-06,
+      "loss": 0.8565,
+      "step": 305
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.6480538126924586e-06,
+      "loss": 0.8995,
+      "step": 306
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.6457702711723445e-06,
+      "loss": 0.9137,
+      "step": 307
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.6434803368125824e-06,
+      "loss": 0.8168,
+      "step": 308
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.64118402238989e-06,
+      "loss": 0.9208,
+      "step": 309
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.638881340716583e-06,
+      "loss": 0.874,
+      "step": 310
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.6365723046405024e-06,
+      "loss": 0.839,
+      "step": 311
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.634256927044943e-06,
+      "loss": 0.8603,
+      "step": 312
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.631935220848585e-06,
+      "loss": 0.8274,
+      "step": 313
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.6296071990054165e-06,
+      "loss": 0.8907,
+      "step": 314
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.6272728745046647e-06,
+      "loss": 0.8703,
+      "step": 315
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.6249322603707224e-06,
+      "loss": 0.8357,
+      "step": 316
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.622585369663076e-06,
+      "loss": 0.8152,
+      "step": 317
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.620232215476231e-06,
+      "loss": 0.8823,
+      "step": 318
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.6178728109396414e-06,
+      "loss": 0.8447,
+      "step": 319
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.615507169217635e-06,
+      "loss": 0.8334,
+      "step": 320
+    },
+    {
+      "epoch": 2.37,
+      "eval_loss": 1.1515859365463257,
+      "eval_runtime": 1.8978,
+      "eval_samples_per_second": 34.777,
+      "eval_steps_per_second": 3.162,
+      "step": 320
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.613135303509338e-06,
+      "loss": 0.8715,
+      "step": 321
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.6107572270486066e-06,
+      "loss": 0.868,
+      "step": 322
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.608372953103948e-06,
+      "loss": 0.8163,
+      "step": 323
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.6059824949784475e-06,
+      "loss": 0.8287,
+      "step": 324
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 2.603585866009698e-06,
+      "loss": 0.8716,
+      "step": 325
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 2.601183079569719e-06,
+      "loss": 0.8878,
+      "step": 326
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.5987741490648896e-06,
+      "loss": 0.8602,
+      "step": 327
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 2.5963590879358674e-06,
+      "loss": 0.8889,
+      "step": 328
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 2.5939379096575157e-06,
+      "loss": 0.8558,
+      "step": 329
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 2.5915106277388294e-06,
+      "loss": 0.8329,
+      "step": 330
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 2.589077255722859e-06,
+      "loss": 0.8948,
+      "step": 331
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 2.5866378071866335e-06,
+      "loss": 0.891,
+      "step": 332
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 2.584192295741087e-06,
+      "loss": 0.821,
+      "step": 333
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 2.5817407350309826e-06,
+      "loss": 0.8006,
+      "step": 334
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 2.5792831387348325e-06,
+      "loss": 0.8367,
+      "step": 335
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 2.5768195205648263e-06,
+      "loss": 0.8606,
+      "step": 336
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 2.574349894266753e-06,
+      "loss": 0.8561,
+      "step": 337
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 2.5718742736199244e-06,
+      "loss": 0.8584,
+      "step": 338
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 2.5693926724370957e-06,
+      "loss": 0.821,
+      "step": 339
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 2.566905104564393e-06,
+      "loss": 0.8762,
+      "step": 340
+    },
+    {
+      "epoch": 2.52,
+      "eval_loss": 1.1476701498031616,
+      "eval_runtime": 1.8933,
+      "eval_samples_per_second": 34.86,
+      "eval_steps_per_second": 3.169,
+      "step": 340
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 2.5644115838812316e-06,
+      "loss": 0.8535,
+      "step": 341
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 2.561912124300242e-06,
+      "loss": 0.8704,
+      "step": 342
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2.5594067397671903e-06,
+      "loss": 0.8927,
+      "step": 343
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 2.5568954442609014e-06,
+      "loss": 0.8688,
+      "step": 344
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 2.55437825179318e-06,
+      "loss": 0.8235,
+      "step": 345
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 2.551855176408733e-06,
+      "loss": 0.7796,
+      "step": 346
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.5493262321850913e-06,
+      "loss": 0.9045,
+      "step": 347
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 2.5467914332325314e-06,
+      "loss": 0.8974,
+      "step": 348
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 2.544250793693995e-06,
+      "loss": 0.8263,
+      "step": 349
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.5417043277450128e-06,
+      "loss": 0.8562,
+      "step": 350
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 2.5391520495936246e-06,
+      "loss": 0.8782,
+      "step": 351
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 2.5365939734802974e-06,
+      "loss": 0.8627,
+      "step": 352
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 2.5340301136778488e-06,
+      "loss": 0.82,
+      "step": 353
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.5314604844913677e-06,
+      "loss": 0.9275,
+      "step": 354
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 2.5288851002581322e-06,
+      "loss": 0.8527,
+      "step": 355
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 2.5263039753475312e-06,
+      "loss": 0.8548,
+      "step": 356
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 2.523717124160983e-06,
+      "loss": 0.8639,
+      "step": 357
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.521124561131857e-06,
+      "loss": 0.8676,
+      "step": 358
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 2.518526300725391e-06,
+      "loss": 0.8644,
+      "step": 359
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 2.515922357438612e-06,
+      "loss": 0.8883,
+      "step": 360
+    },
+    {
+      "epoch": 2.67,
+      "eval_loss": 1.148936152458191,
+      "eval_runtime": 1.8897,
+      "eval_samples_per_second": 34.927,
+      "eval_steps_per_second": 3.175,
+      "step": 360
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 2.5133127458002535e-06,
+      "loss": 0.856,
+      "step": 361
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.510697480370677e-06,
+      "loss": 0.8909,
+      "step": 362
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 2.5080765757417887e-06,
+      "loss": 0.8923,
+      "step": 363
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 2.5054500465369594e-06,
+      "loss": 0.8675,
+      "step": 364
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 2.5028179074109417e-06,
+      "loss": 0.8263,
+      "step": 365
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.5001801730497888e-06,
+      "loss": 0.8297,
+      "step": 366
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 2.4975368581707723e-06,
+      "loss": 0.8592,
+      "step": 367
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 2.4948879775223016e-06,
+      "loss": 0.838,
+      "step": 368
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 2.49223354588384e-06,
+      "loss": 0.8562,
+      "step": 369
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 2.489573578065821e-06,
+      "loss": 0.8598,
+      "step": 370
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 2.486908088909569e-06,
+      "loss": 0.7962,
+      "step": 371
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 2.4842370932872145e-06,
+      "loss": 0.9059,
+      "step": 372
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.481560606101611e-06,
+      "loss": 0.9074,
+      "step": 373
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.478878642286253e-06,
+      "loss": 0.8676,
+      "step": 374
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.4761912168051904e-06,
+      "loss": 0.8292,
+      "step": 375
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.4734983446529485e-06,
+      "loss": 0.8717,
+      "step": 376
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.47080004085444e-06,
+      "loss": 0.8997,
+      "step": 377
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.4680963204648868e-06,
+      "loss": 0.8365,
+      "step": 378
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.4653871985697293e-06,
+      "loss": 0.848,
+      "step": 379
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 2.4626726902845476e-06,
+      "loss": 0.8689,
+      "step": 380
+    },
+    {
+      "epoch": 2.82,
+      "eval_loss": 1.1483380794525146,
+      "eval_runtime": 1.8888,
+      "eval_samples_per_second": 34.943,
+      "eval_steps_per_second": 3.177,
+      "step": 380
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.459952810754975e-06,
+      "loss": 0.811,
+      "step": 381
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 2.457227575156613e-06,
+      "loss": 0.8707,
+      "step": 382
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 2.454496998694949e-06,
+      "loss": 0.9079,
+      "step": 383
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 2.451761096605268e-06,
+      "loss": 0.9,
+      "step": 384
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.4490198841525704e-06,
+      "loss": 0.876,
+      "step": 385
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 2.446273376631486e-06,
+      "loss": 0.8777,
+      "step": 386
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 2.443521589366187e-06,
+      "loss": 0.7991,
+      "step": 387
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 2.4407645377103057e-06,
+      "loss": 0.8337,
+      "step": 388
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 2.4380022370468466e-06,
+      "loss": 0.8483,
+      "step": 389
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 2.4352347027881005e-06,
+      "loss": 0.8779,
+      "step": 390
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 2.4324619503755603e-06,
+      "loss": 0.873,
+      "step": 391
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 2.4296839952798324e-06,
+      "loss": 0.8904,
+      "step": 392
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 2.4269008530005523e-06,
+      "loss": 0.8555,
+      "step": 393
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 2.424112539066298e-06,
+      "loss": 0.8698,
+      "step": 394
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.421319069034502e-06,
+      "loss": 0.8633,
+      "step": 395
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.4185204584913648e-06,
+      "loss": 0.8637,
+      "step": 396
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 2.41571672305177e-06,
+      "loss": 0.8717,
+      "step": 397
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 2.412907878359194e-06,
+      "loss": 0.8667,
+      "step": 398
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 2.4100939400856215e-06,
+      "loss": 0.8401,
+      "step": 399
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 2.4072749239314563e-06,
+      "loss": 0.8465,
+      "step": 400
+    },
+    {
+      "epoch": 2.97,
+      "eval_loss": 1.148036241531372,
+      "eval_runtime": 1.8799,
+      "eval_samples_per_second": 35.109,
+      "eval_steps_per_second": 3.192,
+      "step": 400
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.404450845625435e-06,
+      "loss": 0.8723,
+      "step": 401
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.4016217209245378e-06,
+      "loss": 0.8306,
+      "step": 402
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 2.3987875656139014e-06,
+      "loss": 0.8828,
+      "step": 403
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 2.3959483955067312e-06,
+      "loss": 0.8913,
+      "step": 404
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 2.3931042264442123e-06,
+      "loss": 0.7621,
+      "step": 405
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 2.3902550742954223e-06,
+      "loss": 0.7881,
+      "step": 406
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 2.3874009549572413e-06,
+      "loss": 0.7749,
+      "step": 407
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 2.3845418843542636e-06,
+      "loss": 0.8033,
+      "step": 408
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 2.3816778784387097e-06,
+      "loss": 0.7568,
+      "step": 409
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 2.3788089531903374e-06,
+      "loss": 0.7794,
+      "step": 410
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 2.3759351246163502e-06,
+      "loss": 0.7215,
+      "step": 411
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 2.3730564087513115e-06,
+      "loss": 0.8205,
+      "step": 412
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 2.3701728216570525e-06,
+      "loss": 0.7681,
+      "step": 413
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 2.367284379422584e-06,
+      "loss": 0.7828,
+      "step": 414
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 2.364391098164005e-06,
+      "loss": 0.7704,
+      "step": 415
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 2.3614929940244153e-06,
+      "loss": 0.7685,
+      "step": 416
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 2.3585900831738237e-06,
+      "loss": 0.7795,
+      "step": 417
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 2.355682381809057e-06,
+      "loss": 0.7783,
+      "step": 418
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 2.3527699061536726e-06,
+      "loss": 0.7422,
+      "step": 419
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 2.3498526724578637e-06,
+      "loss": 0.8175,
+      "step": 420
+    },
+    {
+      "epoch": 3.12,
+      "eval_loss": 1.1665019989013672,
+      "eval_runtime": 1.8849,
+      "eval_samples_per_second": 35.015,
+      "eval_steps_per_second": 3.183,
+      "step": 420
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 2.346930696998373e-06,
+      "loss": 0.7785,
+      "step": 421
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 2.3440039960784e-06,
+      "loss": 0.8415,
+      "step": 422
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 2.3410725860275093e-06,
+      "loss": 0.8219,
+      "step": 423
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 2.3381364832015388e-06,
+      "loss": 0.7653,
+      "step": 424
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 2.3351957039825128e-06,
+      "loss": 0.7541,
+      "step": 425
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 2.332250264778545e-06,
+      "loss": 0.7887,
+      "step": 426
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 2.3293001820237507e-06,
+      "loss": 0.7835,
+      "step": 427
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 2.3263454721781537e-06,
+      "loss": 0.8103,
+      "step": 428
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 2.3233861517275953e-06,
+      "loss": 0.7349,
+      "step": 429
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 2.320422237183641e-06,
+      "loss": 0.7939,
+      "step": 430
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 2.3174537450834883e-06,
+      "loss": 0.7693,
+      "step": 431
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 2.314480691989877e-06,
+      "loss": 0.7564,
+      "step": 432
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 2.3115030944909947e-06,
+      "loss": 0.7882,
+      "step": 433
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 2.3085209692003836e-06,
+      "loss": 0.7792,
+      "step": 434
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 2.305534332756849e-06,
+      "loss": 0.8427,
+      "step": 435
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 2.302543201824367e-06,
+      "loss": 0.7481,
+      "step": 436
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 2.2995475930919907e-06,
+      "loss": 0.7896,
+      "step": 437
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 2.296547523273756e-06,
+      "loss": 0.8052,
+      "step": 438
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 2.2935430091085907e-06,
+      "loss": 0.7615,
+      "step": 439
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 2.2905340673602185e-06,
+      "loss": 0.7487,
+      "step": 440
+    },
+    {
+      "epoch": 3.27,
+      "eval_loss": 1.1718708276748657,
+      "eval_runtime": 1.8991,
+      "eval_samples_per_second": 34.753,
+      "eval_steps_per_second": 3.159,
+      "step": 440
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 2.2875207148170684e-06,
+      "loss": 0.8262,
+      "step": 441
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 2.284502968292179e-06,
+      "loss": 0.7856,
+      "step": 442
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 2.2814808446231046e-06,
+      "loss": 0.7435,
+      "step": 443
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 2.278454360671823e-06,
+      "loss": 0.8163,
+      "step": 444
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 2.275423533324638e-06,
+      "loss": 0.7505,
+      "step": 445
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 2.272388379492091e-06,
+      "loss": 0.7919,
+      "step": 446
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 2.2693489161088594e-06,
+      "loss": 0.7943,
+      "step": 447
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 2.266305160133668e-06,
+      "loss": 0.7781,
+      "step": 448
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 2.263257128549191e-06,
+      "loss": 0.794,
+      "step": 449
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 2.26020483836196e-06,
+      "loss": 0.7782,
+      "step": 450
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 2.257148306602266e-06,
+      "loss": 0.798,
+      "step": 451
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 2.2540875503240656e-06,
+      "loss": 0.8059,
+      "step": 452
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 2.2510225866048884e-06,
+      "loss": 0.8247,
+      "step": 453
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 2.2479534325457374e-06,
+      "loss": 0.7228,
+      "step": 454
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 2.244880105270996e-06,
+      "loss": 0.7524,
+      "step": 455
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 2.2418026219283332e-06,
+      "loss": 0.8087,
+      "step": 456
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 2.238720999688605e-06,
+      "loss": 0.808,
+      "step": 457
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 2.2356352557457624e-06,
+      "loss": 0.77,
+      "step": 458
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 2.232545407316752e-06,
+      "loss": 0.8192,
+      "step": 459
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 2.229451471641422e-06,
+      "loss": 0.7782,
+      "step": 460
+    },
+    {
+      "epoch": 3.41,
+      "eval_loss": 1.1698108911514282,
+      "eval_runtime": 1.8889,
+      "eval_samples_per_second": 34.941,
+      "eval_steps_per_second": 3.176,
+      "step": 460
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 2.226353465982426e-06,
+      "loss": 0.82,
+      "step": 461
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 2.2232514076251247e-06,
+      "loss": 0.7974,
+      "step": 462
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 2.220145313877493e-06,
+      "loss": 0.794,
+      "step": 463
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 2.217035202070019e-06,
+      "loss": 0.7961,
+      "step": 464
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 2.2139210895556107e-06,
+      "loss": 0.7925,
+      "step": 465
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 2.210802993709498e-06,
+      "loss": 0.7905,
+      "step": 466
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 2.2076809319291363e-06,
+      "loss": 0.7778,
+      "step": 467
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 2.2045549216341095e-06,
+      "loss": 0.7611,
+      "step": 468
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 2.20142498026603e-06,
+      "loss": 0.8309,
+      "step": 469
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 2.1982911252884453e-06,
+      "loss": 0.7868,
+      "step": 470
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 2.1951533741867384e-06,
+      "loss": 0.7749,
+      "step": 471
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 2.192011744468032e-06,
+      "loss": 0.7628,
+      "step": 472
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 2.188866253661088e-06,
+      "loss": 0.7945,
+      "step": 473
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 2.185716919316212e-06,
+      "loss": 0.7941,
+      "step": 474
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 2.182563759005155e-06,
+      "loss": 0.8122,
+      "step": 475
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 2.179406790321014e-06,
+      "loss": 0.7727,
+      "step": 476
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 2.1762460308781368e-06,
+      "loss": 0.7917,
+      "step": 477
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 2.1730814983120186e-06,
+      "loss": 0.8089,
+      "step": 478
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 2.16991321027921e-06,
+      "loss": 0.7781,
+      "step": 479
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 2.166741184457214e-06,
+      "loss": 0.8008,
+      "step": 480
+    },
+    {
+      "epoch": 3.56,
+      "eval_loss": 1.1715506315231323,
+      "eval_runtime": 1.894,
+      "eval_samples_per_second": 34.846,
+      "eval_steps_per_second": 3.168,
+      "step": 480
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 2.1635654385443873e-06,
+      "loss": 0.8241,
+      "step": 481
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 2.1603859902598455e-06,
+      "loss": 0.7675,
+      "step": 482
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 2.1572028573433596e-06,
+      "loss": 0.8259,
+      "step": 483
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 2.1540160575552603e-06,
+      "loss": 0.8168,
+      "step": 484
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 2.150825608676337e-06,
+      "loss": 0.7856,
+      "step": 485
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 2.147631528507739e-06,
+      "loss": 0.7947,
+      "step": 486
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 2.144433834870878e-06,
+      "loss": 0.8247,
+      "step": 487
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 2.141232545607324e-06,
+      "loss": 0.8087,
+      "step": 488
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 2.138027678578712e-06,
+      "loss": 0.7621,
+      "step": 489
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 2.1348192516666377e-06,
+      "loss": 0.8085,
+      "step": 490
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 2.131607282772559e-06,
+      "loss": 0.7395,
+      "step": 491
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 2.1283917898176972e-06,
+      "loss": 0.7809,
+      "step": 492
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 2.1251727907429357e-06,
+      "loss": 0.8311,
+      "step": 493
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 2.1219503035087204e-06,
+      "loss": 0.7925,
+      "step": 494
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 2.118724346094959e-06,
+      "loss": 0.7815,
+      "step": 495
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 2.1154949365009215e-06,
+      "loss": 0.788,
+      "step": 496
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 2.11226209274514e-06,
+      "loss": 0.7893,
+      "step": 497
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 2.109025832865307e-06,
+      "loss": 0.7982,
+      "step": 498
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 2.1057861749181742e-06,
+      "loss": 0.7929,
+      "step": 499
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 2.1025431369794542e-06,
+      "loss": 0.8505,
+      "step": 500
+    },
+    {
+      "epoch": 3.71,
+      "eval_loss": 1.1672272682189941,
+      "eval_runtime": 1.8949,
+      "eval_samples_per_second": 34.83,
+      "eval_steps_per_second": 3.166,
+      "step": 500
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 2.0992967371437185e-06,
+      "loss": 0.8078,
+      "step": 501
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 2.0960469935242956e-06,
+      "loss": 0.8023,
+      "step": 502
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 2.092793924253171e-06,
+      "loss": 0.8141,
+      "step": 503
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 2.0895375474808856e-06,
+      "loss": 0.7886,
+      "step": 504
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 2.086277881376434e-06,
+      "loss": 0.7546,
+      "step": 505
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 2.0830149441271657e-06,
+      "loss": 0.7664,
+      "step": 506
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 2.0797487539386784e-06,
+      "loss": 0.8069,
+      "step": 507
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 2.0764793290347222e-06,
+      "loss": 0.8065,
+      "step": 508
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 2.0732066876570953e-06,
+      "loss": 0.8278,
+      "step": 509
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 2.06993084806554e-06,
+      "loss": 0.7892,
+      "step": 510
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 2.0666518285376455e-06,
+      "loss": 0.7862,
+      "step": 511
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 2.0633696473687415e-06,
+      "loss": 0.7902,
+      "step": 512
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 2.0600843228718007e-06,
+      "loss": 0.7739,
+      "step": 513
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 2.0567958733773313e-06,
+      "loss": 0.7619,
+      "step": 514
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 2.0535043172332786e-06,
+      "loss": 0.8042,
+      "step": 515
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 2.0502096728049225e-06,
+      "loss": 0.7272,
+      "step": 516
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 2.0469119584747724e-06,
+      "loss": 0.8391,
+      "step": 517
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 2.043611192642466e-06,
+      "loss": 0.8057,
+      "step": 518
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 2.0403073937246695e-06,
+      "loss": 0.8066,
+      "step": 519
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 2.03700058015497e-06,
+      "loss": 0.7578,
+      "step": 520
+    },
+    {
+      "epoch": 3.86,
+      "eval_loss": 1.1677265167236328,
+      "eval_runtime": 1.8886,
+      "eval_samples_per_second": 34.946,
+      "eval_steps_per_second": 3.177,
+      "step": 520
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 2.033690770383775e-06,
+      "loss": 0.7881,
+      "step": 521
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 2.0303779828782097e-06,
+      "loss": 0.7894,
+      "step": 522
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 2.027062236122014e-06,
+      "loss": 0.804,
+      "step": 523
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 2.0237435486154387e-06,
+      "loss": 0.7958,
+      "step": 524
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 2.020421938875143e-06,
+      "loss": 0.7689,
+      "step": 525
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 2.0170974254340887e-06,
+      "loss": 0.8254,
+      "step": 526
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 2.013770026841442e-06,
+      "loss": 0.8067,
+      "step": 527
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 2.010439761662465e-06,
+      "loss": 0.8033,
+      "step": 528
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 2.007106648478413e-06,
+      "loss": 0.7673,
+      "step": 529
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 2.0037707058864345e-06,
+      "loss": 0.7588,
+      "step": 530
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 2.0004319524994617e-06,
+      "loss": 0.7752,
+      "step": 531
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 1.9970904069461117e-06,
+      "loss": 0.7761,
+      "step": 532
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 1.99374608787058e-06,
+      "loss": 0.8285,
+      "step": 533
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 1.9903990139325377e-06,
+      "loss": 0.7835,
+      "step": 534
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 1.9870492038070255e-06,
+      "loss": 0.7891,
+      "step": 535
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 1.9836966761843508e-06,
+      "loss": 0.7667,
+      "step": 536
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 1.9803414497699833e-06,
+      "loss": 0.76,
+      "step": 537
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 1.9769835432844523e-06,
+      "loss": 0.8053,
+      "step": 538
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 1.973622975463238e-06,
+      "loss": 0.7687,
+      "step": 539
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 1.970259765056672e-06,
+      "loss": 0.7235,
+      "step": 540
+    },
+    {
+      "epoch": 4.01,
+      "eval_loss": 1.1656824350357056,
+      "eval_runtime": 1.8763,
+      "eval_samples_per_second": 35.176,
+      "eval_steps_per_second": 3.198,
+      "step": 540
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 1.966893930829828e-06,
+      "loss": 0.7476,
+      "step": 541
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 1.963525491562421e-06,
+      "loss": 0.6848,
+      "step": 542
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 1.9601544660487007e-06,
+      "loss": 0.7517,
+      "step": 543
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 1.9567808730973453e-06,
+      "loss": 0.7553,
+      "step": 544
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 1.95340473153136e-06,
+      "loss": 0.7807,
+      "step": 545
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 1.950026060187969e-06,
+      "loss": 0.7643,
+      "step": 546
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 1.9466448779185117e-06,
+      "loss": 0.6956,
+      "step": 547
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 1.9432612035883365e-06,
+      "loss": 0.7316,
+      "step": 548
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 1.9398750560766976e-06,
+      "loss": 0.7282,
+      "step": 549
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 1.936486454276647e-06,
+      "loss": 0.7159,
+      "step": 550
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 1.933095417094931e-06,
+      "loss": 0.6873,
+      "step": 551
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 1.9297019634518857e-06,
+      "loss": 0.7381,
+      "step": 552
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 1.926306112281327e-06,
+      "loss": 0.7012,
+      "step": 553
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 1.92290788253045e-06,
+      "loss": 0.72,
+      "step": 554
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 1.919507293159719e-06,
+      "loss": 0.7262,
+      "step": 555
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 1.916104363142767e-06,
+      "loss": 0.7034,
+      "step": 556
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 1.9126991114662838e-06,
+      "loss": 0.6763,
+      "step": 557
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 1.9092915571299144e-06,
+      "loss": 0.7297,
+      "step": 558
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 1.905881719146151e-06,
+      "loss": 0.7242,
+      "step": 559
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1.9024696165402271e-06,
+      "loss": 0.7537,
+      "step": 560
+    },
+    {
+      "epoch": 4.16,
+      "eval_loss": 1.2012686729431152,
+      "eval_runtime": 1.8849,
+      "eval_samples_per_second": 35.014,
+      "eval_steps_per_second": 3.183,
+      "step": 560
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 1.8990552683500125e-06,
+      "loss": 0.7345,
+      "step": 561
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 1.8956386936259065e-06,
+      "loss": 0.7603,
+      "step": 562
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 1.8922199114307298e-06,
+      "loss": 0.6958,
+      "step": 563
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 1.8887989408396212e-06,
+      "loss": 0.6731,
+      "step": 564
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 1.8853758009399288e-06,
+      "loss": 0.7308,
+      "step": 565
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 1.8819505108311053e-06,
+      "loss": 0.7179,
+      "step": 566
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 1.8785230896245992e-06,
+      "loss": 0.7405,
+      "step": 567
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 1.8750935564437512e-06,
+      "loss": 0.7034,
+      "step": 568
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 1.8716619304236844e-06,
+      "loss": 0.7059,
+      "step": 569
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 1.8682282307111988e-06,
+      "loss": 0.7262,
+      "step": 570
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 1.8647924764646661e-06,
+      "loss": 0.6885,
+      "step": 571
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 1.8613546868539196e-06,
+      "loss": 0.7533,
+      "step": 572
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 1.8579148810601505e-06,
+      "loss": 0.6855,
+      "step": 573
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 1.8544730782757977e-06,
+      "loss": 0.6793,
+      "step": 574
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 1.8510292977044434e-06,
+      "loss": 0.7536,
+      "step": 575
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 1.8475835585607053e-06,
+      "loss": 0.7172,
+      "step": 576
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 1.8441358800701275e-06,
+      "loss": 0.7391,
+      "step": 577
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 1.840686281469076e-06,
+      "loss": 0.6517,
+      "step": 578
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 1.837234782004629e-06,
+      "loss": 0.7022,
+      "step": 579
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 1.8337814009344715e-06,
+      "loss": 0.6614,
+      "step": 580
+    },
+    {
+      "epoch": 4.3,
+      "eval_loss": 1.202240228652954,
+      "eval_runtime": 1.8878,
+      "eval_samples_per_second": 34.96,
+      "eval_steps_per_second": 3.178,
+      "step": 580
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 1.8303261575267871e-06,
+      "loss": 0.7196,
+      "step": 581
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 1.8268690710601495e-06,
+      "loss": 0.7177,
+      "step": 582
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 1.8234101608234164e-06,
+      "loss": 0.6744,
+      "step": 583
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 1.8199494461156204e-06,
+      "loss": 0.735,
+      "step": 584
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 1.8164869462458636e-06,
+      "loss": 0.7484,
+      "step": 585
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 1.8130226805332063e-06,
+      "loss": 0.6986,
+      "step": 586
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 1.8095566683065638e-06,
+      "loss": 0.743,
+      "step": 587
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 1.8060889289045952e-06,
+      "loss": 0.7289,
+      "step": 588
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 1.8026194816755951e-06,
+      "loss": 0.7529,
+      "step": 589
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 1.7991483459773888e-06,
+      "loss": 0.7316,
+      "step": 590
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 1.7956755411772204e-06,
+      "loss": 0.7085,
+      "step": 591
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 1.7922010866516499e-06,
+      "loss": 0.6792,
+      "step": 592
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 1.788725001786438e-06,
+      "loss": 0.7389,
+      "step": 593
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 1.7852473059764454e-06,
+      "loss": 0.7344,
+      "step": 594
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 1.7817680186255197e-06,
+      "loss": 0.7414,
+      "step": 595
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 1.7782871591463869e-06,
+      "loss": 0.7644,
+      "step": 596
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 1.774804746960548e-06,
+      "loss": 0.711,
+      "step": 597
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 1.7713208014981651e-06,
+      "loss": 0.7147,
+      "step": 598
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 1.7678353421979551e-06,
+      "loss": 0.7274,
+      "step": 599
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 1.7643483885070828e-06,
+      "loss": 0.7596,
+      "step": 600
+    },
+    {
+      "epoch": 4.45,
+      "eval_loss": 1.200880765914917,
+      "eval_runtime": 1.891,
+      "eval_samples_per_second": 34.902,
+      "eval_steps_per_second": 3.173,
+      "step": 600
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 1.7608599598810494e-06,
+      "loss": 0.7393,
+      "step": 601
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 1.7573700757835875e-06,
+      "loss": 0.7719,
+      "step": 602
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 1.7538787556865488e-06,
+      "loss": 0.7093,
+      "step": 603
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 1.7503860190697968e-06,
+      "loss": 0.7511,
+      "step": 604
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 1.746891885421101e-06,
+      "loss": 0.7397,
+      "step": 605
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 1.7433963742360229e-06,
+      "loss": 0.7208,
+      "step": 606
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 1.7398995050178117e-06,
+      "loss": 0.7158,
+      "step": 607
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 1.736401297277293e-06,
+      "loss": 0.7395,
+      "step": 608
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 1.7329017705327609e-06,
+      "loss": 0.7271,
+      "step": 609
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 1.7294009443098692e-06,
+      "loss": 0.7256,
+      "step": 610
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 1.725898838141521e-06,
+      "loss": 0.6933,
+      "step": 611
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 1.722395471567763e-06,
+      "loss": 0.6941,
+      "step": 612
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 1.7188908641356723e-06,
+      "loss": 0.7977,
+      "step": 613
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 1.7153850353992504e-06,
+      "loss": 0.7242,
+      "step": 614
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 1.711878004919313e-06,
+      "loss": 0.7427,
+      "step": 615
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 1.70836979226338e-06,
+      "loss": 0.7344,
+      "step": 616
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 1.7048604170055693e-06,
+      "loss": 0.7447,
+      "step": 617
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 1.7013498987264833e-06,
+      "loss": 0.7589,
+      "step": 618
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 1.6978382570131036e-06,
+      "loss": 0.7398,
+      "step": 619
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 1.6943255114586788e-06,
+      "loss": 0.7049,
+      "step": 620
+    },
+    {
+      "epoch": 4.6,
+      "eval_loss": 1.2018176317214966,
+      "eval_runtime": 1.8792,
+      "eval_samples_per_second": 35.121,
+      "eval_steps_per_second": 3.193,
+      "step": 620
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 1.6908116816626172e-06,
+      "loss": 0.6686,
+      "step": 621
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 1.6872967872303765e-06,
+      "loss": 0.7073,
+      "step": 622
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 1.6837808477733537e-06,
+      "loss": 0.6985,
+      "step": 623
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 1.6802638829087788e-06,
+      "loss": 0.6903,
+      "step": 624
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 1.6767459122595998e-06,
+      "loss": 0.7469,
+      "step": 625
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 1.6732269554543796e-06,
+      "loss": 0.7233,
+      "step": 626
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 1.6697070321271805e-06,
+      "loss": 0.7559,
+      "step": 627
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 1.6661861619174603e-06,
+      "loss": 0.6888,
+      "step": 628
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 1.6626643644699583e-06,
+      "loss": 0.754,
+      "step": 629
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 1.6591416594345871e-06,
+      "loss": 0.7619,
+      "step": 630
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 1.6556180664663244e-06,
+      "loss": 0.7401,
+      "step": 631
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 1.6520936052251014e-06,
+      "loss": 0.7013,
+      "step": 632
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 1.6485682953756943e-06,
+      "loss": 0.7098,
+      "step": 633
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 1.6450421565876128e-06,
+      "loss": 0.7579,
+      "step": 634
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 1.6415152085349934e-06,
+      "loss": 0.711,
+      "step": 635
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 1.6379874708964875e-06,
+      "loss": 0.7365,
+      "step": 636
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 1.6344589633551504e-06,
+      "loss": 0.6762,
+      "step": 637
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 1.6309297055983354e-06,
+      "loss": 0.7176,
+      "step": 638
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 1.6273997173175802e-06,
+      "loss": 0.7319,
+      "step": 639
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 1.6238690182084987e-06,
+      "loss": 0.7377,
+      "step": 640
+    },
+    {
+      "epoch": 4.75,
+      "eval_loss": 1.2016925811767578,
+      "eval_runtime": 1.8787,
+      "eval_samples_per_second": 35.131,
+      "eval_steps_per_second": 3.194,
+      "step": 640
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 1.6203376279706712e-06,
+      "loss": 0.6862,
+      "step": 641
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 1.6168055663075337e-06,
+      "loss": 0.7598,
+      "step": 642
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 1.6132728529262696e-06,
+      "loss": 0.7638,
+      "step": 643
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 1.6097395075376966e-06,
+      "loss": 0.7765,
+      "step": 644
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 1.6062055498561607e-06,
+      "loss": 0.7284,
+      "step": 645
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 1.602670999599422e-06,
+      "loss": 0.7261,
+      "step": 646
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 1.599135876488549e-06,
+      "loss": 0.7467,
+      "step": 647
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 1.5956002002478062e-06,
+      "loss": 0.7068,
+      "step": 648
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 1.5920639906045424e-06,
+      "loss": 0.736,
+      "step": 649
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 1.5885272672890843e-06,
+      "loss": 0.6772,
+      "step": 650
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 1.5849900500346244e-06,
+      "loss": 0.7179,
+      "step": 651
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 1.5814523585771104e-06,
+      "loss": 0.734,
+      "step": 652
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 1.5779142126551373e-06,
+      "loss": 0.7077,
+      "step": 653
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 1.5743756320098334e-06,
+      "loss": 0.7114,
+      "step": 654
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 1.5708366363847552e-06,
+      "loss": 0.7567,
+      "step": 655
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 1.5672972455257725e-06,
+      "loss": 0.7148,
+      "step": 656
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 1.5637574791809616e-06,
+      "loss": 0.7239,
+      "step": 657
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 1.5602173571004933e-06,
+      "loss": 0.7832,
+      "step": 658
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 1.5566768990365226e-06,
+      "loss": 0.7385,
+      "step": 659
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 1.553136124743081e-06,
+      "loss": 0.6787,
+      "step": 660
+    },
+    {
+      "epoch": 4.9,
+      "eval_loss": 1.2029359340667725,
+      "eval_runtime": 1.8784,
+      "eval_samples_per_second": 35.137,
+      "eval_steps_per_second": 3.194,
+      "step": 660
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 1.5495950539759623e-06,
+      "loss": 0.7358,
+      "step": 661
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 1.546053706492616e-06,
+      "loss": 0.7815,
+      "step": 662
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 1.5425121020520348e-06,
+      "loss": 0.7123,
+      "step": 663
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 1.538970260414645e-06,
+      "loss": 0.7439,
+      "step": 664
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 1.535428201342197e-06,
+      "loss": 0.7348,
+      "step": 665
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 1.5318859445976535e-06,
+      "loss": 0.7079,
+      "step": 666
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 1.528343509945081e-06,
+      "loss": 0.7244,
+      "step": 667
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 1.524800917149538e-06,
+      "loss": 0.7538,
+      "step": 668
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 1.5212581859769652e-06,
+      "loss": 0.7628,
+      "step": 669
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 1.517715336194077e-06,
+      "loss": 0.6712,
+      "step": 670
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 1.5141723875682468e-06,
+      "loss": 0.7134,
+      "step": 671
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 1.510629359867402e-06,
+      "loss": 0.7365,
+      "step": 672
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 1.5070862728599103e-06,
+      "loss": 0.7095,
+      "step": 673
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 1.5035431463144702e-06,
+      "loss": 0.6718,
+      "step": 674
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 1.5e-06,
+      "loss": 0.6433,
+      "step": 675
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 1.4964568536855303e-06,
+      "loss": 0.6893,
+      "step": 676
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 1.49291372714009e-06,
+      "loss": 0.6974,
+      "step": 677
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 1.4893706401325978e-06,
+      "loss": 0.6199,
+      "step": 678
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 1.4858276124317535e-06,
+      "loss": 0.6859,
+      "step": 679
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 1.4822846638059234e-06,
+      "loss": 0.6532,
+      "step": 680
+    },
+    {
+      "epoch": 5.05,
+      "eval_loss": 1.2448418140411377,
+      "eval_runtime": 1.8757,
+      "eval_samples_per_second": 35.187,
+      "eval_steps_per_second": 3.199,
+      "step": 680
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 1.4787418140230348e-06,
+      "loss": 0.6454,
+      "step": 681
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 1.4751990828504623e-06,
+      "loss": 0.6857,
+      "step": 682
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 1.4716564900549195e-06,
+      "loss": 0.6432,
+      "step": 683
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 1.4681140554023468e-06,
+      "loss": 0.6536,
+      "step": 684
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 1.464571798657803e-06,
+      "loss": 0.6288,
+      "step": 685
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 1.4610297395853553e-06,
+      "loss": 0.7039,
+      "step": 686
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 1.4574878979479652e-06,
+      "loss": 0.7148,
+      "step": 687
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 1.453946293507384e-06,
+      "loss": 0.6258,
+      "step": 688
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 1.4504049460240376e-06,
+      "loss": 0.6293,
+      "step": 689
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 1.4468638752569193e-06,
+      "loss": 0.6808,
+      "step": 690
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 1.4433231009634775e-06,
+      "loss": 0.6593,
+      "step": 691
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 1.4397826428995074e-06,
+      "loss": 0.691,
+      "step": 692
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 1.436242520819039e-06,
+      "loss": 0.6125,
+      "step": 693
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 1.4327027544742282e-06,
+      "loss": 0.6734,
+      "step": 694
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 1.4291633636152451e-06,
+      "loss": 0.7129,
+      "step": 695
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 1.4256243679901664e-06,
+      "loss": 0.6828,
+      "step": 696
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 1.4220857873448632e-06,
+      "loss": 0.6765,
+      "step": 697
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 1.4185476414228896e-06,
+      "loss": 0.6531,
+      "step": 698
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 1.415009949965376e-06,
+      "loss": 0.6887,
+      "step": 699
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 1.411472732710916e-06,
+      "loss": 0.6774,
+      "step": 700
+    },
+    {
+      "epoch": 5.19,
+      "eval_loss": 1.2385097742080688,
+      "eval_runtime": 1.8997,
+      "eval_samples_per_second": 34.743,
+      "eval_steps_per_second": 3.158,
+      "step": 700
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 1.4079360093954583e-06,
+      "loss": 0.6887,
+      "step": 701
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 1.4043997997521941e-06,
+      "loss": 0.6968,
+      "step": 702
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 1.400864123511451e-06,
+      "loss": 0.6861,
+      "step": 703
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 1.397329000400578e-06,
+      "loss": 0.6834,
+      "step": 704
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 1.3937944501438398e-06,
+      "loss": 0.6621,
+      "step": 705
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 1.3902604924623037e-06,
+      "loss": 0.6826,
+      "step": 706
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 1.3867271470737303e-06,
+      "loss": 0.6758,
+      "step": 707
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 1.3831944336924663e-06,
+      "loss": 0.6089,
+      "step": 708
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 1.379662372029329e-06,
+      "loss": 0.6735,
+      "step": 709
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 1.3761309817915016e-06,
+      "loss": 0.5866,
+      "step": 710
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 1.37260028268242e-06,
+      "loss": 0.6271,
+      "step": 711
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 1.3690702944016651e-06,
+      "loss": 0.6588,
+      "step": 712
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 1.3655410366448499e-06,
+      "loss": 0.6722,
+      "step": 713
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 1.3620125291035132e-06,
+      "loss": 0.6661,
+      "step": 714
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 1.3584847914650067e-06,
+      "loss": 0.6138,
+      "step": 715
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 1.354957843412387e-06,
+      "loss": 0.6911,
+      "step": 716
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 1.351431704624306e-06,
+      "loss": 0.6661,
+      "step": 717
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 1.3479063947748985e-06,
+      "loss": 0.638,
+      "step": 718
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 1.3443819335336756e-06,
+      "loss": 0.7198,
+      "step": 719
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 1.3408583405654131e-06,
+      "loss": 0.6523,
+      "step": 720
+    },
+    {
+      "epoch": 5.34,
+      "eval_loss": 1.238602876663208,
+      "eval_runtime": 1.896,
+      "eval_samples_per_second": 34.81,
+      "eval_steps_per_second": 3.165,
+      "step": 720
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 1.3373356355300424e-06,
+      "loss": 0.6303,
+      "step": 721
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 1.33381383808254e-06,
+      "loss": 0.6506,
+      "step": 722
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 1.33029296787282e-06,
+      "loss": 0.6491,
+      "step": 723
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 1.326773044545621e-06,
+      "loss": 0.6619,
+      "step": 724
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 1.3232540877404e-06,
+      "loss": 0.6354,
+      "step": 725
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 1.3197361170912215e-06,
+      "loss": 0.687,
+      "step": 726
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 1.3162191522266461e-06,
+      "loss": 0.6967,
+      "step": 727
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 1.3127032127696238e-06,
+      "loss": 0.6881,
+      "step": 728
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 1.309188318337383e-06,
+      "loss": 0.6658,
+      "step": 729
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 1.3056744885413216e-06,
+      "loss": 0.6705,
+      "step": 730
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 1.3021617429868966e-06,
+      "loss": 0.6619,
+      "step": 731
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 1.2986501012735172e-06,
+      "loss": 0.7074,
+      "step": 732
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 1.295139582994431e-06,
+      "loss": 0.6854,
+      "step": 733
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 1.2916302077366203e-06,
+      "loss": 0.6469,
+      "step": 734
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 1.2881219950806876e-06,
+      "loss": 0.6799,
+      "step": 735
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 1.2846149646007497e-06,
+      "loss": 0.6892,
+      "step": 736
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 1.281109135864328e-06,
+      "loss": 0.6788,
+      "step": 737
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 1.2776045284322369e-06,
+      "loss": 0.6626,
+      "step": 738
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 1.2741011618584794e-06,
+      "loss": 0.6507,
+      "step": 739
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 1.270599055690131e-06,
+      "loss": 0.7073,
+      "step": 740
+    },
+    {
+      "epoch": 5.49,
+      "eval_loss": 1.240828037261963,
+      "eval_runtime": 1.8983,
+      "eval_samples_per_second": 34.767,
+      "eval_steps_per_second": 3.161,
+      "step": 740
+    },
+    {
+      "epoch": 5.5,
+      "learning_rate": 1.2670982294672394e-06,
+      "loss": 0.6756,
+      "step": 741
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 1.2635987027227074e-06,
+      "loss": 0.6911,
+      "step": 742
+    },
+    {
+      "epoch": 5.51,
+      "learning_rate": 1.2601004949821888e-06,
+      "loss": 0.6465,
+      "step": 743
+    },
+    {
+      "epoch": 5.52,
+      "learning_rate": 1.2566036257639774e-06,
+      "loss": 0.6711,
+      "step": 744
+    },
+    {
+      "epoch": 5.53,
+      "learning_rate": 1.2531081145788989e-06,
+      "loss": 0.6543,
+      "step": 745
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 1.2496139809302033e-06,
+      "loss": 0.6941,
+      "step": 746
+    },
+    {
+      "epoch": 5.54,
+      "learning_rate": 1.2461212443134513e-06,
+      "loss": 0.681,
+      "step": 747
+    },
+    {
+      "epoch": 5.55,
+      "learning_rate": 1.2426299242164126e-06,
+      "loss": 0.7029,
+      "step": 748
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 1.2391400401189504e-06,
+      "loss": 0.6122,
+      "step": 749
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 1.2356516114929177e-06,
+      "loss": 0.6976,
+      "step": 750
+    },
+    {
+      "epoch": 5.57,
+      "learning_rate": 1.2321646578020452e-06,
+      "loss": 0.7029,
+      "step": 751
+    },
+    {
+      "epoch": 5.58,
+      "learning_rate": 1.2286791985018356e-06,
+      "loss": 0.6697,
+      "step": 752
+    },
+    {
+      "epoch": 5.59,
+      "learning_rate": 1.2251952530394524e-06,
+      "loss": 0.6747,
+      "step": 753
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 1.2217128408536134e-06,
+      "loss": 0.6873,
+      "step": 754
+    },
+    {
+      "epoch": 5.6,
+      "learning_rate": 1.2182319813744808e-06,
+      "loss": 0.6643,
+      "step": 755
+    },
+    {
+      "epoch": 5.61,
+      "learning_rate": 1.2147526940235546e-06,
+      "loss": 0.662,
+      "step": 756
+    },
+    {
+      "epoch": 5.62,
+      "learning_rate": 1.211274998213562e-06,
+      "loss": 0.6604,
+      "step": 757
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 1.2077989133483506e-06,
+      "loss": 0.6699,
+      "step": 758
+    },
+    {
+      "epoch": 5.63,
+      "learning_rate": 1.2043244588227797e-06,
+      "loss": 0.6715,
+      "step": 759
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 1.2008516540226115e-06,
+      "loss": 0.641,
+      "step": 760
+    },
+    {
+      "epoch": 5.64,
+      "eval_loss": 1.2447619438171387,
+      "eval_runtime": 1.8934,
+      "eval_samples_per_second": 34.859,
+      "eval_steps_per_second": 3.169,
+      "step": 760
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 1.1973805183244054e-06,
+      "loss": 0.6393,
+      "step": 761
+    },
+    {
+      "epoch": 5.65,
+      "learning_rate": 1.193911071095405e-06,
+      "loss": 0.6917,
+      "step": 762
+    },
+    {
+      "epoch": 5.66,
+      "learning_rate": 1.1904433316934363e-06,
+      "loss": 0.6576,
+      "step": 763
+    },
+    {
+      "epoch": 5.67,
+      "learning_rate": 1.1869773194667937e-06,
+      "loss": 0.666,
+      "step": 764
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 1.183513053754137e-06,
+      "loss": 0.6874,
+      "step": 765
+    },
+    {
+      "epoch": 5.68,
+      "learning_rate": 1.18005055388438e-06,
+      "loss": 0.67,
+      "step": 766
+    },
+    {
+      "epoch": 5.69,
+      "learning_rate": 1.1765898391765837e-06,
+      "loss": 0.6728,
+      "step": 767
+    },
+    {
+      "epoch": 5.7,
+      "learning_rate": 1.1731309289398508e-06,
+      "loss": 0.6532,
+      "step": 768
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 1.169673842473213e-06,
+      "loss": 0.6842,
+      "step": 769
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 1.1662185990655286e-06,
+      "loss": 0.6749,
+      "step": 770
+    },
+    {
+      "epoch": 5.72,
+      "learning_rate": 1.1627652179953711e-06,
+      "loss": 0.6752,
+      "step": 771
+    },
+    {
+      "epoch": 5.73,
+      "learning_rate": 1.1593137185309248e-06,
+      "loss": 0.6702,
+      "step": 772
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 1.1558641199298728e-06,
+      "loss": 0.6698,
+      "step": 773
+    },
+    {
+      "epoch": 5.74,
+      "learning_rate": 1.1524164414392952e-06,
+      "loss": 0.6577,
+      "step": 774
+    },
+    {
+      "epoch": 5.75,
+      "learning_rate": 1.1489707022955566e-06,
+      "loss": 0.7082,
+      "step": 775
+    },
+    {
+      "epoch": 5.76,
+      "learning_rate": 1.1455269217242022e-06,
+      "loss": 0.6982,
+      "step": 776
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 1.1420851189398496e-06,
+      "loss": 0.6843,
+      "step": 777
+    },
+    {
+      "epoch": 5.77,
+      "learning_rate": 1.1386453131460803e-06,
+      "loss": 0.6929,
+      "step": 778
+    },
+    {
+      "epoch": 5.78,
+      "learning_rate": 1.1352075235353344e-06,
+      "loss": 0.6854,
+      "step": 779
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 1.1317717692888015e-06,
+      "loss": 0.7031,
+      "step": 780
+    },
+    {
+      "epoch": 5.79,
+      "eval_loss": 1.241090178489685,
+      "eval_runtime": 1.8839,
+      "eval_samples_per_second": 35.034,
+      "eval_steps_per_second": 3.185,
+      "step": 780
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 1.1283380695763163e-06,
+      "loss": 0.6876,
+      "step": 781
+    },
+    {
+      "epoch": 5.8,
+      "learning_rate": 1.124906443556249e-06,
+      "loss": 0.6889,
+      "step": 782
+    },
+    {
+      "epoch": 5.81,
+      "learning_rate": 1.121476910375401e-06,
+      "loss": 0.6821,
+      "step": 783
+    },
+    {
+      "epoch": 5.82,
+      "learning_rate": 1.118049489168895e-06,
+      "loss": 0.7038,
+      "step": 784
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 1.1146241990600713e-06,
+      "loss": 0.6854,
+      "step": 785
+    },
+    {
+      "epoch": 5.83,
+      "learning_rate": 1.111201059160379e-06,
+      "loss": 0.675,
+      "step": 786
+    },
+    {
+      "epoch": 5.84,
+      "learning_rate": 1.1077800885692703e-06,
+      "loss": 0.6396,
+      "step": 787
+    },
+    {
+      "epoch": 5.85,
+      "learning_rate": 1.104361306374094e-06,
+      "loss": 0.6649,
+      "step": 788
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 1.1009447316499875e-06,
+      "loss": 0.6661,
+      "step": 789
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 1.0975303834597734e-06,
+      "loss": 0.6559,
+      "step": 790
+    },
+    {
+      "epoch": 5.87,
+      "learning_rate": 1.094118280853849e-06,
+      "loss": 0.6685,
+      "step": 791
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 1.0907084428700861e-06,
+      "loss": 0.6742,
+      "step": 792
+    },
+    {
+      "epoch": 5.88,
+      "learning_rate": 1.0873008885337163e-06,
+      "loss": 0.6427,
+      "step": 793
+    },
+    {
+      "epoch": 5.89,
+      "learning_rate": 1.0838956368572335e-06,
+      "loss": 0.6489,
+      "step": 794
+    },
+    {
+      "epoch": 5.9,
+      "learning_rate": 1.0804927068402811e-06,
+      "loss": 0.6667,
+      "step": 795
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 1.0770921174695505e-06,
+      "loss": 0.6603,
+      "step": 796
+    },
+    {
+      "epoch": 5.91,
+      "learning_rate": 1.0736938877186735e-06,
+      "loss": 0.6544,
+      "step": 797
+    },
+    {
+      "epoch": 5.92,
+      "learning_rate": 1.0702980365481144e-06,
+      "loss": 0.6675,
+      "step": 798
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 1.0669045829050691e-06,
+      "loss": 0.6519,
+      "step": 799
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 1.0635135457233534e-06,
+      "loss": 0.6764,
+      "step": 800
+    },
+    {
+      "epoch": 5.94,
+      "eval_loss": 1.2441211938858032,
+      "eval_runtime": 1.8821,
+      "eval_samples_per_second": 35.068,
+      "eval_steps_per_second": 3.188,
+      "step": 800
+    },
+    {
+      "epoch": 5.94,
+      "learning_rate": 1.0601249439233032e-06,
+      "loss": 0.6552,
+      "step": 801
+    },
+    {
+      "epoch": 5.95,
+      "learning_rate": 1.0567387964116638e-06,
+      "loss": 0.6838,
+      "step": 802
+    },
+    {
+      "epoch": 5.96,
+      "learning_rate": 1.0533551220814888e-06,
+      "loss": 0.6826,
+      "step": 803
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 1.0499739398120311e-06,
+      "loss": 0.6504,
+      "step": 804
+    },
+    {
+      "epoch": 5.97,
+      "learning_rate": 1.0465952684686397e-06,
+      "loss": 0.6699,
+      "step": 805
+    },
+    {
+      "epoch": 5.98,
+      "learning_rate": 1.043219126902655e-06,
+      "loss": 0.6717,
+      "step": 806
+    },
+    {
+      "epoch": 5.99,
+      "learning_rate": 1.0398455339512994e-06,
+      "loss": 0.6655,
+      "step": 807
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 1.036474508437579e-06,
+      "loss": 0.6905,
+      "step": 808
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 1.033106069170172e-06,
+      "loss": 0.5937,
+      "step": 809
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 1.0297402349433286e-06,
+      "loss": 0.6226,
+      "step": 810
+    },
+    {
+      "epoch": 6.02,
+      "learning_rate": 1.026377024536762e-06,
+      "loss": 0.6643,
+      "step": 811
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 1.0230164567155482e-06,
+      "loss": 0.6261,
+      "step": 812
+    },
+    {
+      "epoch": 6.03,
+      "learning_rate": 1.019658550230017e-06,
+      "loss": 0.6455,
+      "step": 813
+    },
+    {
+      "epoch": 6.04,
+      "learning_rate": 1.0163033238156493e-06,
+      "loss": 0.6651,
+      "step": 814
+    },
+    {
+      "epoch": 6.05,
+      "learning_rate": 1.012950796192975e-06,
+      "loss": 0.6287,
+      "step": 815
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 1.0096009860674622e-06,
+      "loss": 0.6414,
+      "step": 816
+    },
+    {
+      "epoch": 6.06,
+      "learning_rate": 1.0062539121294198e-06,
+      "loss": 0.6287,
+      "step": 817
+    },
+    {
+      "epoch": 6.07,
+      "learning_rate": 1.0029095930538886e-06,
+      "loss": 0.5848,
+      "step": 818
+    },
+    {
+      "epoch": 6.08,
+      "learning_rate": 9.995680475005393e-07,
+      "loss": 0.6597,
+      "step": 819
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 9.962292941135662e-07,
+      "loss": 0.6135,
+      "step": 820
+    },
+    {
+      "epoch": 6.09,
+      "eval_loss": 1.2782245874404907,
+      "eval_runtime": 1.8831,
+      "eval_samples_per_second": 35.048,
+      "eval_steps_per_second": 3.186,
+      "step": 820
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 9.928933515215876e-07,
+      "loss": 0.6064,
+      "step": 821
+    },
+    {
+      "epoch": 6.1,
+      "learning_rate": 9.895602383375354e-07,
+      "loss": 0.6353,
+      "step": 822
+    },
+    {
+      "epoch": 6.11,
+      "learning_rate": 9.862299731585582e-07,
+      "loss": 0.6341,
+      "step": 823
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 9.829025745659111e-07,
+      "loss": 0.5561,
+      "step": 824
+    },
+    {
+      "epoch": 6.12,
+      "learning_rate": 9.795780611248574e-07,
+      "loss": 0.6033,
+      "step": 825
+    },
+    {
+      "epoch": 6.13,
+      "learning_rate": 9.762564513845614e-07,
+      "loss": 0.6155,
+      "step": 826
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 9.729377638779858e-07,
+      "loss": 0.601,
+      "step": 827
+    },
+    {
+      "epoch": 6.14,
+      "learning_rate": 9.696220171217908e-07,
+      "loss": 0.6182,
+      "step": 828
+    },
+    {
+      "epoch": 6.15,
+      "learning_rate": 9.663092296162252e-07,
+      "loss": 0.64,
+      "step": 829
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 9.629994198450304e-07,
+      "loss": 0.6277,
+      "step": 830
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 9.596926062753304e-07,
+      "loss": 0.629,
+      "step": 831
+    },
+    {
+      "epoch": 6.17,
+      "learning_rate": 9.56388807357534e-07,
+      "loss": 0.6547,
+      "step": 832
+    },
+    {
+      "epoch": 6.18,
+      "learning_rate": 9.530880415252283e-07,
+      "loss": 0.5858,
+      "step": 833
+    },
+    {
+      "epoch": 6.19,
+      "learning_rate": 9.497903271950775e-07,
+      "loss": 0.569,
+      "step": 834
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 9.464956827667215e-07,
+      "loss": 0.6242,
+      "step": 835
+    },
+    {
+      "epoch": 6.2,
+      "learning_rate": 9.432041266226686e-07,
+      "loss": 0.615,
+      "step": 836
+    },
+    {
+      "epoch": 6.21,
+      "learning_rate": 9.399156771281995e-07,
+      "loss": 0.6756,
+      "step": 837
+    },
+    {
+      "epoch": 6.22,
+      "learning_rate": 9.366303526312581e-07,
+      "loss": 0.6179,
+      "step": 838
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 9.333481714623548e-07,
+      "loss": 0.6275,
+      "step": 839
+    },
+    {
+      "epoch": 6.23,
+      "learning_rate": 9.300691519344602e-07,
+      "loss": 0.5778,
+      "step": 840
+    },
+    {
+      "epoch": 6.23,
+      "eval_loss": 1.2758677005767822,
+      "eval_runtime": 1.8866,
+      "eval_samples_per_second": 34.984,
+      "eval_steps_per_second": 3.18,
+      "step": 840
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 9.267933123429051e-07,
+      "loss": 0.6309,
+      "step": 841
+    },
+    {
+      "epoch": 6.25,
+      "learning_rate": 9.235206709652777e-07,
+      "loss": 0.6355,
+      "step": 842
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 9.202512460613221e-07,
+      "loss": 0.6009,
+      "step": 843
+    },
+    {
+      "epoch": 6.26,
+      "learning_rate": 9.169850558728346e-07,
+      "loss": 0.672,
+      "step": 844
+    },
+    {
+      "epoch": 6.27,
+      "learning_rate": 9.137221186235657e-07,
+      "loss": 0.6201,
+      "step": 845
+    },
+    {
+      "epoch": 6.28,
+      "learning_rate": 9.104624525191147e-07,
+      "loss": 0.6036,
+      "step": 846
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 9.07206075746829e-07,
+      "loss": 0.6741,
+      "step": 847
+    },
+    {
+      "epoch": 6.29,
+      "learning_rate": 9.039530064757047e-07,
+      "loss": 0.6133,
+      "step": 848
+    },
+    {
+      "epoch": 6.3,
+      "learning_rate": 9.007032628562816e-07,
+      "loss": 0.6179,
+      "step": 849
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 8.974568630205462e-07,
+      "loss": 0.6222,
+      "step": 850
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 8.942138250818262e-07,
+      "loss": 0.6739,
+      "step": 851
+    },
+    {
+      "epoch": 6.32,
+      "learning_rate": 8.909741671346939e-07,
+      "loss": 0.6565,
+      "step": 852
+    },
+    {
+      "epoch": 6.33,
+      "learning_rate": 8.877379072548599e-07,
+      "loss": 0.6461,
+      "step": 853
+    },
+    {
+      "epoch": 6.34,
+      "learning_rate": 8.845050634990782e-07,
+      "loss": 0.6424,
+      "step": 854
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 8.812756539050414e-07,
+      "loss": 0.6272,
+      "step": 855
+    },
+    {
+      "epoch": 6.35,
+      "learning_rate": 8.780496964912799e-07,
+      "loss": 0.6136,
+      "step": 856
+    },
+    {
+      "epoch": 6.36,
+      "learning_rate": 8.748272092570648e-07,
+      "loss": 0.6076,
+      "step": 857
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 8.71608210182303e-07,
+      "loss": 0.6568,
+      "step": 858
+    },
+    {
+      "epoch": 6.37,
+      "learning_rate": 8.683927172274413e-07,
+      "loss": 0.6333,
+      "step": 859
+    },
+    {
+      "epoch": 6.38,
+      "learning_rate": 8.651807483333627e-07,
+      "loss": 0.6685,
+      "step": 860
+    },
+    {
+      "epoch": 6.38,
+      "eval_loss": 1.2791975736618042,
+      "eval_runtime": 1.8881,
+      "eval_samples_per_second": 34.956,
+      "eval_steps_per_second": 3.178,
+      "step": 860
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 8.619723214212887e-07,
+      "loss": 0.6996,
+      "step": 861
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 8.587674543926759e-07,
+      "loss": 0.6577,
+      "step": 862
+    },
+    {
+      "epoch": 6.4,
+      "learning_rate": 8.555661651291226e-07,
+      "loss": 0.6222,
+      "step": 863
+    },
+    {
+      "epoch": 6.41,
+      "learning_rate": 8.523684714922608e-07,
+      "loss": 0.6179,
+      "step": 864
+    },
+    {
+      "epoch": 6.42,
+      "learning_rate": 8.49174391323663e-07,
+      "loss": 0.5978,
+      "step": 865
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 8.459839424447394e-07,
+      "loss": 0.6716,
+      "step": 866
+    },
+    {
+      "epoch": 6.43,
+      "learning_rate": 8.427971426566402e-07,
+      "loss": 0.5939,
+      "step": 867
+    },
+    {
+      "epoch": 6.44,
+      "learning_rate": 8.396140097401551e-07,
+      "loss": 0.5966,
+      "step": 868
+    },
+    {
+      "epoch": 6.45,
+      "learning_rate": 8.364345614556126e-07,
+      "loss": 0.6231,
+      "step": 869
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 8.332588155427869e-07,
+      "loss": 0.6093,
+      "step": 870
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 8.300867897207904e-07,
+      "loss": 0.6676,
+      "step": 871
+    },
+    {
+      "epoch": 6.47,
+      "learning_rate": 8.269185016879815e-07,
+      "loss": 0.6241,
+      "step": 872
+    },
+    {
+      "epoch": 6.48,
+      "learning_rate": 8.237539691218637e-07,
+      "loss": 0.6165,
+      "step": 873
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 8.205932096789858e-07,
+      "loss": 0.6517,
+      "step": 874
+    },
+    {
+      "epoch": 6.49,
+      "learning_rate": 8.174362409948456e-07,
+      "loss": 0.6405,
+      "step": 875
+    },
+    {
+      "epoch": 6.5,
+      "learning_rate": 8.142830806837877e-07,
+      "loss": 0.6362,
+      "step": 876
+    },
+    {
+      "epoch": 6.51,
+      "learning_rate": 8.111337463389124e-07,
+      "loss": 0.6627,
+      "step": 877
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 8.079882555319686e-07,
+      "loss": 0.6467,
+      "step": 878
+    },
+    {
+      "epoch": 6.52,
+      "learning_rate": 8.048466258132618e-07,
+      "loss": 0.5837,
+      "step": 879
+    },
+    {
+      "epoch": 6.53,
+      "learning_rate": 8.017088747115553e-07,
+      "loss": 0.6443,
+      "step": 880
+    },
+    {
+      "epoch": 6.53,
+      "eval_loss": 1.2795201539993286,
+      "eval_runtime": 1.8894,
+      "eval_samples_per_second": 34.932,
+      "eval_steps_per_second": 3.176,
+      "step": 880
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 7.98575019733971e-07,
+      "loss": 0.6019,
+      "step": 881
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 7.954450783658911e-07,
+      "loss": 0.5963,
+      "step": 882
+    },
+    {
+      "epoch": 6.55,
+      "learning_rate": 7.923190680708637e-07,
+      "loss": 0.6128,
+      "step": 883
+    },
+    {
+      "epoch": 6.56,
+      "learning_rate": 7.891970062905021e-07,
+      "loss": 0.5946,
+      "step": 884
+    },
+    {
+      "epoch": 6.57,
+      "learning_rate": 7.860789104443897e-07,
+      "loss": 0.6165,
+      "step": 885
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 7.829647979299815e-07,
+      "loss": 0.6339,
+      "step": 886
+    },
+    {
+      "epoch": 6.58,
+      "learning_rate": 7.798546861225071e-07,
+      "loss": 0.6352,
+      "step": 887
+    },
+    {
+      "epoch": 6.59,
+      "learning_rate": 7.767485923748755e-07,
+      "loss": 0.654,
+      "step": 888
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 7.736465340175744e-07,
+      "loss": 0.612,
+      "step": 889
+    },
+    {
+      "epoch": 6.6,
+      "learning_rate": 7.70548528358578e-07,
+      "loss": 0.6352,
+      "step": 890
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 7.674545926832481e-07,
+      "loss": 0.6613,
+      "step": 891
+    },
+    {
+      "epoch": 6.62,
+      "learning_rate": 7.643647442542383e-07,
+      "loss": 0.6241,
+      "step": 892
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 7.612790003113949e-07,
+      "loss": 0.6404,
+      "step": 893
+    },
+    {
+      "epoch": 6.63,
+      "learning_rate": 7.581973780716668e-07,
+      "loss": 0.61,
+      "step": 894
+    },
+    {
+      "epoch": 6.64,
+      "learning_rate": 7.551198947290043e-07,
+      "loss": 0.6391,
+      "step": 895
+    },
+    {
+      "epoch": 6.65,
+      "learning_rate": 7.520465674542628e-07,
+      "loss": 0.6046,
+      "step": 896
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 7.489774133951117e-07,
+      "loss": 0.5885,
+      "step": 897
+    },
+    {
+      "epoch": 6.66,
+      "learning_rate": 7.459124496759343e-07,
+      "loss": 0.6582,
+      "step": 898
+    },
+    {
+      "epoch": 6.67,
+      "learning_rate": 7.428516933977348e-07,
+      "loss": 0.6459,
+      "step": 899
+    },
+    {
+      "epoch": 6.68,
+      "learning_rate": 7.397951616380401e-07,
+      "loss": 0.5658,
+      "step": 900
+    },
+    {
+      "epoch": 6.68,
+      "eval_loss": 1.2807255983352661,
+      "eval_runtime": 1.8871,
+      "eval_samples_per_second": 34.974,
+      "eval_steps_per_second": 3.179,
+      "step": 900
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 7.367428714508092e-07,
+      "loss": 0.6467,
+      "step": 901
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 7.336948398663325e-07,
+      "loss": 0.6538,
+      "step": 902
+    },
+    {
+      "epoch": 6.7,
+      "learning_rate": 7.30651083891141e-07,
+      "loss": 0.6567,
+      "step": 903
+    },
+    {
+      "epoch": 6.71,
+      "learning_rate": 7.276116205079093e-07,
+      "loss": 0.6455,
+      "step": 904
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 7.245764666753617e-07,
+      "loss": 0.6346,
+      "step": 905
+    },
+    {
+      "epoch": 6.72,
+      "learning_rate": 7.215456393281777e-07,
+      "loss": 0.618,
+      "step": 906
+    },
+    {
+      "epoch": 6.73,
+      "learning_rate": 7.185191553768952e-07,
+      "loss": 0.6005,
+      "step": 907
+    },
+    {
+      "epoch": 6.74,
+      "learning_rate": 7.154970317078214e-07,
+      "loss": 0.6174,
+      "step": 908
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 7.124792851829318e-07,
+      "loss": 0.6054,
+      "step": 909
+    },
+    {
+      "epoch": 6.75,
+      "learning_rate": 7.094659326397818e-07,
+      "loss": 0.6477,
+      "step": 910
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 7.064569908914097e-07,
+      "loss": 0.6712,
+      "step": 911
+    },
+    {
+      "epoch": 6.77,
+      "learning_rate": 7.034524767262446e-07,
+      "loss": 0.6384,
+      "step": 912
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 7.004524069080097e-07,
+      "loss": 0.6302,
+      "step": 913
+    },
+    {
+      "epoch": 6.78,
+      "learning_rate": 6.974567981756324e-07,
+      "loss": 0.6326,
+      "step": 914
+    },
+    {
+      "epoch": 6.79,
+      "learning_rate": 6.944656672431511e-07,
+      "loss": 0.6447,
+      "step": 915
+    },
+    {
+      "epoch": 6.8,
+      "learning_rate": 6.914790307996167e-07,
+      "loss": 0.623,
+      "step": 916
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 6.884969055090056e-07,
+      "loss": 0.6078,
+      "step": 917
+    },
+    {
+      "epoch": 6.81,
+      "learning_rate": 6.85519308010123e-07,
+      "loss": 0.6442,
+      "step": 918
+    },
+    {
+      "epoch": 6.82,
+      "learning_rate": 6.825462549165124e-07,
+      "loss": 0.6032,
+      "step": 919
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 6.7957776281636e-07,
+      "loss": 0.6825,
+      "step": 920
+    },
+    {
+      "epoch": 6.83,
+      "eval_loss": 1.2800445556640625,
+      "eval_runtime": 1.8853,
+      "eval_samples_per_second": 35.007,
+      "eval_steps_per_second": 3.182,
+      "step": 920
+    },
+    {
+      "epoch": 6.83,
+      "learning_rate": 6.766138482724051e-07,
+      "loss": 0.589,
+      "step": 921
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 6.736545278218464e-07,
+      "loss": 0.5906,
+      "step": 922
+    },
+    {
+      "epoch": 6.85,
+      "learning_rate": 6.7069981797625e-07,
+      "loss": 0.5974,
+      "step": 923
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 6.67749735221455e-07,
+      "loss": 0.6646,
+      "step": 924
+    },
+    {
+      "epoch": 6.86,
+      "learning_rate": 6.648042960174872e-07,
+      "loss": 0.6045,
+      "step": 925
+    },
+    {
+      "epoch": 6.87,
+      "learning_rate": 6.618635167984613e-07,
+      "loss": 0.6193,
+      "step": 926
+    },
+    {
+      "epoch": 6.88,
+      "learning_rate": 6.589274139724911e-07,
+      "loss": 0.6409,
+      "step": 927
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 6.559960039215999e-07,
+      "loss": 0.6383,
+      "step": 928
+    },
+    {
+      "epoch": 6.89,
+      "learning_rate": 6.530693030016268e-07,
+      "loss": 0.6366,
+      "step": 929
+    },
+    {
+      "epoch": 6.9,
+      "learning_rate": 6.501473275421371e-07,
+      "loss": 0.6119,
+      "step": 930
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 6.472300938463277e-07,
+      "loss": 0.6433,
+      "step": 931
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 6.443176181909433e-07,
+      "loss": 0.6474,
+      "step": 932
+    },
+    {
+      "epoch": 6.92,
+      "learning_rate": 6.414099168261768e-07,
+      "loss": 0.5944,
+      "step": 933
+    },
+    {
+      "epoch": 6.93,
+      "learning_rate": 6.385070059755846e-07,
+      "loss": 0.649,
+      "step": 934
+    },
+    {
+      "epoch": 6.94,
+      "learning_rate": 6.356089018359949e-07,
+      "loss": 0.6469,
+      "step": 935
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 6.327156205774161e-07,
+      "loss": 0.6081,
+      "step": 936
+    },
+    {
+      "epoch": 6.95,
+      "learning_rate": 6.298271783429478e-07,
+      "loss": 0.6489,
+      "step": 937
+    },
+    {
+      "epoch": 6.96,
+      "learning_rate": 6.269435912486883e-07,
+      "loss": 0.6154,
+      "step": 938
+    },
+    {
+      "epoch": 6.97,
+      "learning_rate": 6.240648753836501e-07,
+      "loss": 0.6206,
+      "step": 939
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 6.211910468096631e-07,
+      "loss": 0.6234,
+      "step": 940
+    },
+    {
+      "epoch": 6.98,
+      "eval_loss": 1.2810416221618652,
+      "eval_runtime": 1.8893,
+      "eval_samples_per_second": 34.934,
+      "eval_steps_per_second": 3.176,
+      "step": 940
+    },
+    {
+      "epoch": 6.98,
+      "learning_rate": 6.183221215612905e-07,
+      "loss": 0.6427,
+      "step": 941
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 6.154581156457368e-07,
+      "loss": 0.5942,
+      "step": 942
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 6.125990450427591e-07,
+      "loss": 0.659,
+      "step": 943
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1340,
+  "num_train_epochs": 10,
+  "save_steps": 500,
+  "total_flos": 1.6678837104158966e+18,
+  "trial_name": null,
+  "trial_params": null
+}