diff --git "a/checkpoint-1990/trainer_state.json" "b/checkpoint-1990/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1990/trainer_state.json"
@@ -0,0 +1,12040 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.20868288590604026,
+  "eval_steps": 199,
+  "global_step": 1990,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.0986358866736623e-08,
+      "loss": 2.1261,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1972717733473246e-08,
+      "loss": 2.0696,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.295907660020988e-08,
+      "loss": 2.0664,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.394543546694649e-08,
+      "loss": 2.0793,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.0493179433368311e-07,
+      "loss": 1.9493,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2591815320041975e-07,
+      "loss": 2.0603,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4690451206715635e-07,
+      "loss": 2.2201,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6789087093389298e-07,
+      "loss": 2.0475,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.888772298006296e-07,
+      "loss": 1.8393,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.0986358866736621e-07,
+      "loss": 2.1082,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.3084994753410284e-07,
+      "loss": 2.0846,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.518363064008395e-07,
+      "loss": 2.2172,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.7282266526757613e-07,
+      "loss": 2.0091,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.938090241343127e-07,
+      "loss": 2.1092,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1479538300104934e-07,
+      "loss": 2.0882,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3578174186778596e-07,
+      "loss": 2.0103,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.567681007345226e-07,
+      "loss": 2.119,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.777544596012592e-07,
+      "loss": 2.005,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.9874081846799585e-07,
+      "loss": 2.0261,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1972717733473243e-07,
+      "loss": 1.8976,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.407135362014691e-07,
+      "loss": 2.0584,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.616998950682057e-07,
+      "loss": 1.9565,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.826862539349423e-07,
+      "loss": 2.1067,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.03672612801679e-07,
+      "loss": 1.9939,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.246589716684156e-07,
+      "loss": 1.9337,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.456453305351523e-07,
+      "loss": 2.0869,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.666316894018888e-07,
+      "loss": 2.0134,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.876180482686254e-07,
+      "loss": 1.5282,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.086044071353621e-07,
+      "loss": 1.8605,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.295907660020987e-07,
+      "loss": 1.9561,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.505771248688352e-07,
+      "loss": 1.9562,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.715634837355719e-07,
+      "loss": 1.9124,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 6.925498426023086e-07,
+      "loss": 1.908,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.135362014690452e-07,
+      "loss": 1.9148,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.345225603357818e-07,
+      "loss": 1.9861,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.555089192025184e-07,
+      "loss": 1.8246,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.76495278069255e-07,
+      "loss": 1.9355,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.974816369359917e-07,
+      "loss": 1.9849,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.184679958027284e-07,
+      "loss": 1.8894,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.394543546694649e-07,
+      "loss": 1.9615,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.604407135362015e-07,
+      "loss": 1.8815,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.814270724029382e-07,
+      "loss": 1.9888,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.024134312696747e-07,
+      "loss": 1.8663,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.233997901364114e-07,
+      "loss": 1.8711,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.443861490031481e-07,
+      "loss": 1.8709,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.653725078698846e-07,
+      "loss": 1.8837,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.863588667366212e-07,
+      "loss": 1.9386,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.007345225603358e-06,
+      "loss": 1.8926,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0283315844700944e-06,
+      "loss": 1.8623,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0493179433368312e-06,
+      "loss": 1.9017,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0703043022035677e-06,
+      "loss": 1.883,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.0912906610703045e-06,
+      "loss": 2.0128,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1122770199370409e-06,
+      "loss": 1.9405,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1332633788037777e-06,
+      "loss": 2.1205,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1542497376705142e-06,
+      "loss": 1.8373,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1752360965372508e-06,
+      "loss": 1.9386,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1962224554039876e-06,
+      "loss": 1.831,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2172088142707242e-06,
+      "loss": 1.8523,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2381951731374608e-06,
+      "loss": 1.8641,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.2591815320041973e-06,
+      "loss": 1.8386,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.280167890870934e-06,
+      "loss": 1.8628,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3011542497376705e-06,
+      "loss": 1.8578,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3221406086044073e-06,
+      "loss": 1.868,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3431269674711439e-06,
+      "loss": 1.9356,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3641133263378806e-06,
+      "loss": 1.7525,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3850996852046172e-06,
+      "loss": 1.8815,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4060860440713536e-06,
+      "loss": 1.8752,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4270724029380904e-06,
+      "loss": 1.8017,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.448058761804827e-06,
+      "loss": 1.8209,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4690451206715635e-06,
+      "loss": 1.7979,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.4900314795383003e-06,
+      "loss": 1.8267,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.511017838405037e-06,
+      "loss": 1.862,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5320041972717733e-06,
+      "loss": 1.8278,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.55299055613851e-06,
+      "loss": 1.816,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5739769150052466e-06,
+      "loss": 1.8465,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5949632738719834e-06,
+      "loss": 1.7981,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.61594963273872e-06,
+      "loss": 1.9075,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6369359916054568e-06,
+      "loss": 1.7548,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6579223504721931e-06,
+      "loss": 1.8585,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6789087093389297e-06,
+      "loss": 1.7725,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6998950682056665e-06,
+      "loss": 1.8273,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.720881427072403e-06,
+      "loss": 1.777,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7418677859391397e-06,
+      "loss": 1.8077,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7628541448058764e-06,
+      "loss": 1.821,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.783840503672613e-06,
+      "loss": 1.8287,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8048268625393494e-06,
+      "loss": 1.7358,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8258132214060862e-06,
+      "loss": 1.7546,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8467995802728228e-06,
+      "loss": 1.8147,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8677859391395595e-06,
+      "loss": 1.8309,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8887722980062961e-06,
+      "loss": 1.6292,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.909758656873033e-06,
+      "loss": 1.7222,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9307450157397693e-06,
+      "loss": 1.809,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.951731374606506e-06,
+      "loss": 1.8691,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9727177334732424e-06,
+      "loss": 1.8777,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9937040923399792e-06,
+      "loss": 1.7701,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.014690451206716e-06,
+      "loss": 1.885,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0356768100734524e-06,
+      "loss": 1.7737,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0566631689401887e-06,
+      "loss": 1.7554,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0776495278069255e-06,
+      "loss": 1.8553,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.0986358866736623e-06,
+      "loss": 1.8386,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.119622245540399e-06,
+      "loss": 1.8325,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1406086044071355e-06,
+      "loss": 1.7812,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.1615949632738722e-06,
+      "loss": 1.7537,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.182581322140609e-06,
+      "loss": 1.7558,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2035676810073454e-06,
+      "loss": 1.7278,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2245540398740818e-06,
+      "loss": 1.761,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2455403987408186e-06,
+      "loss": 1.792,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2665267576075553e-06,
+      "loss": 1.7965,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.287513116474292e-06,
+      "loss": 1.7624,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3084994753410285e-06,
+      "loss": 1.7744,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.329485834207765e-06,
+      "loss": 1.7691,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3504721930745016e-06,
+      "loss": 1.8194,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3714585519412384e-06,
+      "loss": 1.691,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3924449108079752e-06,
+      "loss": 1.7505,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4134312696747116e-06,
+      "loss": 1.7492,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4344176285414484e-06,
+      "loss": 1.8426,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4554039874081847e-06,
+      "loss": 1.855,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.4763903462749215e-06,
+      "loss": 1.7778,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.497376705141658e-06,
+      "loss": 1.7715,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5183630640083947e-06,
+      "loss": 1.7382,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5393494228751315e-06,
+      "loss": 1.7027,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.560335781741868e-06,
+      "loss": 1.7508,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5813221406086046e-06,
+      "loss": 1.7119,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.602308499475341e-06,
+      "loss": 1.8291,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.623294858342078e-06,
+      "loss": 1.7568,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6442812172088146e-06,
+      "loss": 1.9176,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.665267576075551e-06,
+      "loss": 1.7036,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6862539349422877e-06,
+      "loss": 1.7842,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.707240293809024e-06,
+      "loss": 1.7883,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7282266526757613e-06,
+      "loss": 1.7662,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7492130115424977e-06,
+      "loss": 1.5587,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7701993704092344e-06,
+      "loss": 1.7486,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.791185729275971e-06,
+      "loss": 1.7515,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.812172088142707e-06,
+      "loss": 1.7668,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8331584470094444e-06,
+      "loss": 1.7895,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8541448058761808e-06,
+      "loss": 1.7961,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8751311647429175e-06,
+      "loss": 1.6693,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.896117523609654e-06,
+      "loss": 1.776,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9171038824763903e-06,
+      "loss": 1.7422,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.938090241343127e-06,
+      "loss": 1.7845,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.959076600209864e-06,
+      "loss": 1.8011,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9800629590766006e-06,
+      "loss": 1.6759,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.001049317943337e-06,
+      "loss": 1.7367,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.022035676810074e-06,
+      "loss": 1.7511,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.04302203567681e-06,
+      "loss": 1.779,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0640083945435465e-06,
+      "loss": 1.7542,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0849947534102837e-06,
+      "loss": 1.742,
+      "step": 147
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.10598111227702e-06,
+      "loss": 1.7408,
+      "step": 148
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.126967471143757e-06,
+      "loss": 1.6937,
+      "step": 149
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1479538300104932e-06,
+      "loss": 1.7198,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.1689401888772305e-06,
+      "loss": 1.674,
+      "step": 151
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.189926547743967e-06,
+      "loss": 1.7469,
+      "step": 152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.210912906610703e-06,
+      "loss": 1.7192,
+      "step": 153
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.23189926547744e-06,
+      "loss": 1.9507,
+      "step": 154
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2528856243441763e-06,
+      "loss": 1.8311,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2738719832109136e-06,
+      "loss": 1.7125,
+      "step": 156
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.29485834207765e-06,
+      "loss": 1.8235,
+      "step": 157
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3158447009443863e-06,
+      "loss": 1.7358,
+      "step": 158
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.336831059811123e-06,
+      "loss": 1.774,
+      "step": 159
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3578174186778594e-06,
+      "loss": 1.7717,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3788037775445962e-06,
+      "loss": 1.6815,
+      "step": 161
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.399790136411333e-06,
+      "loss": 1.6993,
+      "step": 162
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.42077649527807e-06,
+      "loss": 1.5911,
+      "step": 163
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.441762854144806e-06,
+      "loss": 1.7484,
+      "step": 164
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4627492130115425e-06,
+      "loss": 1.8047,
+      "step": 165
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4837355718782793e-06,
+      "loss": 1.7397,
+      "step": 166
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5047219307450157e-06,
+      "loss": 1.7384,
+      "step": 167
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.525708289611753e-06,
+      "loss": 1.7369,
+      "step": 168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5466946484784893e-06,
+      "loss": 1.7013,
+      "step": 169
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.567681007345226e-06,
+      "loss": 1.6896,
+      "step": 170
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5886673662119624e-06,
+      "loss": 1.6953,
+      "step": 171
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6096537250786988e-06,
+      "loss": 1.7638,
+      "step": 172
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.630640083945436e-06,
+      "loss": 1.6814,
+      "step": 173
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6516264428121724e-06,
+      "loss": 1.6938,
+      "step": 174
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.672612801678909e-06,
+      "loss": 1.8404,
+      "step": 175
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6935991605456455e-06,
+      "loss": 1.7429,
+      "step": 176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.714585519412382e-06,
+      "loss": 1.7408,
+      "step": 177
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.735571878279119e-06,
+      "loss": 1.6916,
+      "step": 178
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7565582371458554e-06,
+      "loss": 1.7668,
+      "step": 179
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7775445960125922e-06,
+      "loss": 1.7444,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7985309548793286e-06,
+      "loss": 1.7478,
+      "step": 181
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.819517313746066e-06,
+      "loss": 1.7812,
+      "step": 182
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.840503672612802e-06,
+      "loss": 1.7029,
+      "step": 183
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.8614900314795385e-06,
+      "loss": 1.66,
+      "step": 184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.882476390346275e-06,
+      "loss": 1.7375,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.903462749213012e-06,
+      "loss": 1.7702,
+      "step": 186
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.924449108079749e-06,
+      "loss": 1.7368,
+      "step": 187
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.945435466946485e-06,
+      "loss": 1.7121,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.966421825813222e-06,
+      "loss": 1.7639,
+      "step": 189
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.9874081846799584e-06,
+      "loss": 1.6665,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.008394543546694e-06,
+      "loss": 1.6536,
+      "step": 191
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.029380902413432e-06,
+      "loss": 1.7049,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.050367261280168e-06,
+      "loss": 1.8199,
+      "step": 193
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.071353620146905e-06,
+      "loss": 1.747,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0923399790136415e-06,
+      "loss": 1.7382,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1133263378803775e-06,
+      "loss": 1.6958,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134312696747115e-06,
+      "loss": 1.6974,
+      "step": 197
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.155299055613851e-06,
+      "loss": 1.6909,
+      "step": 198
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.176285414480588e-06,
+      "loss": 1.6995,
+      "step": 199
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 1.7673084735870361,
+      "eval_runtime": 77.3397,
+      "eval_samples_per_second": 1.668,
+      "eval_steps_per_second": 0.116,
+      "step": 199
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.197271773347325e-06,
+      "loss": 1.6732,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.218258132214061e-06,
+      "loss": 1.7269,
+      "step": 201
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.239244491080798e-06,
+      "loss": 1.6493,
+      "step": 202
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.260230849947534e-06,
+      "loss": 1.6544,
+      "step": 203
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.281217208814271e-06,
+      "loss": 1.8876,
+      "step": 204
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.302203567681008e-06,
+      "loss": 1.7196,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3231899265477445e-06,
+      "loss": 1.7659,
+      "step": 206
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.344176285414481e-06,
+      "loss": 1.7129,
+      "step": 207
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.365162644281218e-06,
+      "loss": 1.7269,
+      "step": 208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.386149003147954e-06,
+      "loss": 1.7957,
+      "step": 209
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.407135362014691e-06,
+      "loss": 1.681,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.428121720881428e-06,
+      "loss": 1.7112,
+      "step": 211
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4491080797481635e-06,
+      "loss": 1.7149,
+      "step": 212
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.470094438614901e-06,
+      "loss": 1.6928,
+      "step": 213
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.491080797481637e-06,
+      "loss": 1.6171,
+      "step": 214
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.512067156348374e-06,
+      "loss": 1.7275,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.533053515215111e-06,
+      "loss": 1.7502,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.554039874081847e-06,
+      "loss": 1.7948,
+      "step": 217
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.575026232948584e-06,
+      "loss": 1.8349,
+      "step": 218
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.59601259181532e-06,
+      "loss": 1.6717,
+      "step": 219
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.616998950682057e-06,
+      "loss": 1.7387,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637985309548794e-06,
+      "loss": 1.836,
+      "step": 221
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.65897166841553e-06,
+      "loss": 1.7285,
+      "step": 222
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.679958027282267e-06,
+      "loss": 1.6988,
+      "step": 223
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.700944386149003e-06,
+      "loss": 1.642,
+      "step": 224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.72193074501574e-06,
+      "loss": 1.7184,
+      "step": 225
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.742917103882477e-06,
+      "loss": 1.6935,
+      "step": 226
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.763903462749214e-06,
+      "loss": 1.7593,
+      "step": 227
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7848898216159504e-06,
+      "loss": 1.6854,
+      "step": 228
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.805876180482686e-06,
+      "loss": 1.7047,
+      "step": 229
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.826862539349423e-06,
+      "loss": 1.6543,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.84784889821616e-06,
+      "loss": 1.6321,
+      "step": 231
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.868835257082897e-06,
+      "loss": 1.6938,
+      "step": 232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889821615949633e-06,
+      "loss": 1.725,
+      "step": 233
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9108079748163695e-06,
+      "loss": 1.7119,
+      "step": 234
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.931794333683106e-06,
+      "loss": 1.7741,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.952780692549843e-06,
+      "loss": 1.702,
+      "step": 236
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.97376705141658e-06,
+      "loss": 1.6566,
+      "step": 237
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.994753410283316e-06,
+      "loss": 1.7506,
+      "step": 238
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.0157397691500534e-06,
+      "loss": 1.7185,
+      "step": 239
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.036726128016789e-06,
+      "loss": 1.8174,
+      "step": 240
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.057712486883526e-06,
+      "loss": 1.6781,
+      "step": 241
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.078698845750263e-06,
+      "loss": 1.7818,
+      "step": 242
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.099685204617e-06,
+      "loss": 1.6647,
+      "step": 243
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.120671563483736e-06,
+      "loss": 1.6918,
+      "step": 244
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.1416579223504725e-06,
+      "loss": 1.7119,
+      "step": 245
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.162644281217209e-06,
+      "loss": 1.612,
+      "step": 246
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.183630640083945e-06,
+      "loss": 1.5301,
+      "step": 247
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.204616998950682e-06,
+      "loss": 1.7407,
+      "step": 248
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.22560335781742e-06,
+      "loss": 1.7838,
+      "step": 249
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.246589716684156e-06,
+      "loss": 1.6825,
+      "step": 250
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.267576075550892e-06,
+      "loss": 1.6821,
+      "step": 251
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.288562434417629e-06,
+      "loss": 1.6712,
+      "step": 252
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.309548793284366e-06,
+      "loss": 1.6599,
+      "step": 253
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.330535152151102e-06,
+      "loss": 1.7009,
+      "step": 254
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.351521511017839e-06,
+      "loss": 1.6845,
+      "step": 255
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.3725078698845754e-06,
+      "loss": 1.6832,
+      "step": 256
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.393494228751312e-06,
+      "loss": 1.8023,
+      "step": 257
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.414480587618048e-06,
+      "loss": 1.718,
+      "step": 258
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.435466946484785e-06,
+      "loss": 1.602,
+      "step": 259
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.456453305351523e-06,
+      "loss": 1.6133,
+      "step": 260
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.4774396642182585e-06,
+      "loss": 1.704,
+      "step": 261
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.498426023084995e-06,
+      "loss": 1.6702,
+      "step": 262
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.519412381951732e-06,
+      "loss": 1.7256,
+      "step": 263
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.540398740818469e-06,
+      "loss": 1.7384,
+      "step": 264
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.561385099685205e-06,
+      "loss": 1.6504,
+      "step": 265
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.582371458551942e-06,
+      "loss": 1.6896,
+      "step": 266
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.603357817418678e-06,
+      "loss": 1.6566,
+      "step": 267
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.624344176285414e-06,
+      "loss": 1.7107,
+      "step": 268
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.645330535152151e-06,
+      "loss": 1.6883,
+      "step": 269
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.666316894018889e-06,
+      "loss": 1.74,
+      "step": 270
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.6873032528856256e-06,
+      "loss": 1.7273,
+      "step": 271
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.7082896117523615e-06,
+      "loss": 1.7397,
+      "step": 272
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.729275970619098e-06,
+      "loss": 1.723,
+      "step": 273
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.750262329485835e-06,
+      "loss": 1.6811,
+      "step": 274
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.771248688352571e-06,
+      "loss": 1.7044,
+      "step": 275
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.792235047219308e-06,
+      "loss": 1.6937,
+      "step": 276
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.813221406086045e-06,
+      "loss": 1.5984,
+      "step": 277
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.8342077649527805e-06,
+      "loss": 1.7853,
+      "step": 278
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.855194123819517e-06,
+      "loss": 1.6742,
+      "step": 279
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.876180482686254e-06,
+      "loss": 1.5612,
+      "step": 280
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.897166841552992e-06,
+      "loss": 1.6945,
+      "step": 281
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.918153200419728e-06,
+      "loss": 1.6618,
+      "step": 282
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.9391395592864645e-06,
+      "loss": 1.6928,
+      "step": 283
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.960125918153201e-06,
+      "loss": 1.7438,
+      "step": 284
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.981112277019937e-06,
+      "loss": 1.5858,
+      "step": 285
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.002098635886674e-06,
+      "loss": 1.6936,
+      "step": 286
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.023084994753411e-06,
+      "loss": 1.6522,
+      "step": 287
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.044071353620148e-06,
+      "loss": 1.7174,
+      "step": 288
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.0650577124868835e-06,
+      "loss": 1.6747,
+      "step": 289
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.08604407135362e-06,
+      "loss": 1.6525,
+      "step": 290
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.107030430220358e-06,
+      "loss": 1.6615,
+      "step": 291
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.128016789087093e-06,
+      "loss": 1.7633,
+      "step": 292
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.149003147953831e-06,
+      "loss": 1.6597,
+      "step": 293
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.1699895068205675e-06,
+      "loss": 1.6356,
+      "step": 294
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.190975865687304e-06,
+      "loss": 1.5906,
+      "step": 295
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.21196222455404e-06,
+      "loss": 1.7179,
+      "step": 296
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.232948583420777e-06,
+      "loss": 1.6571,
+      "step": 297
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.253934942287514e-06,
+      "loss": 1.655,
+      "step": 298
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.27492130115425e-06,
+      "loss": 1.6705,
+      "step": 299
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.2959076600209865e-06,
+      "loss": 1.6931,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.316894018887723e-06,
+      "loss": 1.6721,
+      "step": 301
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.337880377754461e-06,
+      "loss": 1.6573,
+      "step": 302
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.358866736621197e-06,
+      "loss": 1.7316,
+      "step": 303
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.379853095487934e-06,
+      "loss": 1.6635,
+      "step": 304
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.4008394543546704e-06,
+      "loss": 1.7145,
+      "step": 305
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.421825813221406e-06,
+      "loss": 1.6044,
+      "step": 306
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.442812172088143e-06,
+      "loss": 1.6659,
+      "step": 307
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.46379853095488e-06,
+      "loss": 1.6645,
+      "step": 308
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.484784889821617e-06,
+      "loss": 1.6803,
+      "step": 309
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.505771248688353e-06,
+      "loss": 1.6705,
+      "step": 310
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.5267576075550895e-06,
+      "loss": 1.6768,
+      "step": 311
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.547743966421827e-06,
+      "loss": 1.6768,
+      "step": 312
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.568730325288562e-06,
+      "loss": 1.6345,
+      "step": 313
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.5897166841553e-06,
+      "loss": 1.6363,
+      "step": 314
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.610703043022037e-06,
+      "loss": 1.6689,
+      "step": 315
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.6316894018887726e-06,
+      "loss": 1.6641,
+      "step": 316
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.652675760755509e-06,
+      "loss": 1.7221,
+      "step": 317
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.673662119622246e-06,
+      "loss": 1.6427,
+      "step": 318
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.694648478488983e-06,
+      "loss": 1.6451,
+      "step": 319
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.715634837355719e-06,
+      "loss": 1.8635,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.736621196222456e-06,
+      "loss": 1.6825,
+      "step": 321
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.7576075550891924e-06,
+      "loss": 1.7618,
+      "step": 322
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.778593913955928e-06,
+      "loss": 1.776,
+      "step": 323
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.799580272822666e-06,
+      "loss": 1.6075,
+      "step": 324
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.820566631689403e-06,
+      "loss": 1.5783,
+      "step": 325
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.84155299055614e-06,
+      "loss": 1.6573,
+      "step": 326
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.8625393494228755e-06,
+      "loss": 1.6736,
+      "step": 327
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.883525708289612e-06,
+      "loss": 1.7454,
+      "step": 328
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.904512067156349e-06,
+      "loss": 1.658,
+      "step": 329
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.925498426023085e-06,
+      "loss": 1.645,
+      "step": 330
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.946484784889822e-06,
+      "loss": 1.6697,
+      "step": 331
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.967471143756559e-06,
+      "loss": 1.633,
+      "step": 332
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.988457502623296e-06,
+      "loss": 1.879,
+      "step": 333
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.009443861490031e-06,
+      "loss": 1.6834,
+      "step": 334
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.030430220356769e-06,
+      "loss": 1.6658,
+      "step": 335
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.051416579223506e-06,
+      "loss": 1.6606,
+      "step": 336
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.072402938090242e-06,
+      "loss": 1.6488,
+      "step": 337
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.0933892969569785e-06,
+      "loss": 1.6356,
+      "step": 338
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.114375655823715e-06,
+      "loss": 1.6584,
+      "step": 339
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.135362014690452e-06,
+      "loss": 1.6851,
+      "step": 340
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.156348373557188e-06,
+      "loss": 1.5969,
+      "step": 341
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.177334732423925e-06,
+      "loss": 1.6009,
+      "step": 342
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.198321091290662e-06,
+      "loss": 1.6188,
+      "step": 343
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.2193074501573976e-06,
+      "loss": 1.6295,
+      "step": 344
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.240293809024135e-06,
+      "loss": 1.7956,
+      "step": 345
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.261280167890872e-06,
+      "loss": 1.7044,
+      "step": 346
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.282266526757609e-06,
+      "loss": 1.6785,
+      "step": 347
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.303252885624345e-06,
+      "loss": 1.6323,
+      "step": 348
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.3242392444910815e-06,
+      "loss": 1.5752,
+      "step": 349
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.345225603357818e-06,
+      "loss": 1.5829,
+      "step": 350
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.366211962224554e-06,
+      "loss": 1.6529,
+      "step": 351
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.387198321091291e-06,
+      "loss": 1.7484,
+      "step": 352
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.408184679958028e-06,
+      "loss": 1.6692,
+      "step": 353
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.429171038824764e-06,
+      "loss": 1.6153,
+      "step": 354
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.4501573976915005e-06,
+      "loss": 1.6408,
+      "step": 355
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.471143756558238e-06,
+      "loss": 1.5987,
+      "step": 356
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.492130115424975e-06,
+      "loss": 1.6204,
+      "step": 357
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.513116474291711e-06,
+      "loss": 1.6793,
+      "step": 358
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.534102833158448e-06,
+      "loss": 1.6736,
+      "step": 359
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.5550891920251845e-06,
+      "loss": 1.6899,
+      "step": 360
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.57607555089192e-06,
+      "loss": 1.668,
+      "step": 361
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.597061909758657e-06,
+      "loss": 1.6286,
+      "step": 362
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.618048268625394e-06,
+      "loss": 1.6475,
+      "step": 363
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.639034627492132e-06,
+      "loss": 1.6378,
+      "step": 364
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.660020986358867e-06,
+      "loss": 1.6705,
+      "step": 365
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.681007345225604e-06,
+      "loss": 1.5835,
+      "step": 366
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.70199370409234e-06,
+      "loss": 1.7316,
+      "step": 367
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.722980062959077e-06,
+      "loss": 1.6678,
+      "step": 368
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.743966421825814e-06,
+      "loss": 1.6155,
+      "step": 369
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.76495278069255e-06,
+      "loss": 1.6493,
+      "step": 370
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.785939139559287e-06,
+      "loss": 1.6038,
+      "step": 371
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.806925498426024e-06,
+      "loss": 1.6313,
+      "step": 372
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.827911857292761e-06,
+      "loss": 1.5703,
+      "step": 373
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.848898216159498e-06,
+      "loss": 1.7095,
+      "step": 374
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.869884575026233e-06,
+      "loss": 1.6805,
+      "step": 375
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.89087093389297e-06,
+      "loss": 1.6683,
+      "step": 376
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.911857292759706e-06,
+      "loss": 1.6528,
+      "step": 377
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.932843651626443e-06,
+      "loss": 1.6633,
+      "step": 378
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.95383001049318e-06,
+      "loss": 1.722,
+      "step": 379
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.974816369359917e-06,
+      "loss": 1.6455,
+      "step": 380
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.995802728226654e-06,
+      "loss": 1.7223,
+      "step": 381
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.016789087093389e-06,
+      "loss": 1.7372,
+      "step": 382
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.037775445960127e-06,
+      "loss": 1.7798,
+      "step": 383
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.058761804826864e-06,
+      "loss": 1.5855,
+      "step": 384
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.079748163693599e-06,
+      "loss": 1.651,
+      "step": 385
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.100734522560336e-06,
+      "loss": 1.6762,
+      "step": 386
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.121720881427073e-06,
+      "loss": 1.6231,
+      "step": 387
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.14270724029381e-06,
+      "loss": 1.6123,
+      "step": 388
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.163693599160546e-06,
+      "loss": 1.6126,
+      "step": 389
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.184679958027283e-06,
+      "loss": 1.6737,
+      "step": 390
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.20566631689402e-06,
+      "loss": 1.788,
+      "step": 391
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.226652675760755e-06,
+      "loss": 1.6406,
+      "step": 392
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.247639034627493e-06,
+      "loss": 1.6647,
+      "step": 393
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.26862539349423e-06,
+      "loss": 1.6685,
+      "step": 394
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.289611752360967e-06,
+      "loss": 1.6328,
+      "step": 395
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.310598111227702e-06,
+      "loss": 1.5943,
+      "step": 396
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.331584470094439e-06,
+      "loss": 1.8615,
+      "step": 397
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.352570828961176e-06,
+      "loss": 1.6949,
+      "step": 398
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 1.6879847049713135,
+      "eval_runtime": 77.2874,
+      "eval_samples_per_second": 1.669,
+      "eval_steps_per_second": 0.116,
+      "step": 398
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.373557187827912e-06,
+      "loss": 1.6748,
+      "step": 399
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.39454354669465e-06,
+      "loss": 1.6461,
+      "step": 400
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.415529905561386e-06,
+      "loss": 1.6726,
+      "step": 401
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.436516264428123e-06,
+      "loss": 1.7563,
+      "step": 402
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.457502623294858e-06,
+      "loss": 1.6398,
+      "step": 403
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.478488982161596e-06,
+      "loss": 1.6176,
+      "step": 404
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.499475341028333e-06,
+      "loss": 1.7346,
+      "step": 405
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.520461699895068e-06,
+      "loss": 1.5992,
+      "step": 406
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.541448058761805e-06,
+      "loss": 1.6699,
+      "step": 407
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.562434417628542e-06,
+      "loss": 1.6153,
+      "step": 408
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.583420776495279e-06,
+      "loss": 1.5813,
+      "step": 409
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.604407135362015e-06,
+      "loss": 1.6643,
+      "step": 410
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.625393494228752e-06,
+      "loss": 1.6114,
+      "step": 411
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.646379853095489e-06,
+      "loss": 1.6375,
+      "step": 412
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.667366211962224e-06,
+      "loss": 1.5735,
+      "step": 413
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.688352570828963e-06,
+      "loss": 1.6772,
+      "step": 414
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.7093389296957e-06,
+      "loss": 1.5998,
+      "step": 415
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.730325288562436e-06,
+      "loss": 1.6519,
+      "step": 416
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.751311647429171e-06,
+      "loss": 1.6153,
+      "step": 417
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.772298006295908e-06,
+      "loss": 1.696,
+      "step": 418
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.793284365162645e-06,
+      "loss": 1.8254,
+      "step": 419
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.814270724029382e-06,
+      "loss": 1.6961,
+      "step": 420
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.835257082896118e-06,
+      "loss": 1.7149,
+      "step": 421
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.856243441762855e-06,
+      "loss": 1.6383,
+      "step": 422
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.87722980062959e-06,
+      "loss": 1.5654,
+      "step": 423
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.898216159496327e-06,
+      "loss": 1.667,
+      "step": 424
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.919202518363066e-06,
+      "loss": 1.6383,
+      "step": 425
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.940188877229802e-06,
+      "loss": 1.5623,
+      "step": 426
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.961175236096537e-06,
+      "loss": 1.6241,
+      "step": 427
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.982161594963274e-06,
+      "loss": 1.6148,
+      "step": 428
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.003147953830011e-06,
+      "loss": 1.5868,
+      "step": 429
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.024134312696748e-06,
+      "loss": 1.567,
+      "step": 430
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.045120671563485e-06,
+      "loss": 1.6605,
+      "step": 431
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.066107030430221e-06,
+      "loss": 1.5792,
+      "step": 432
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.087093389296958e-06,
+      "loss": 1.6467,
+      "step": 433
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.108079748163693e-06,
+      "loss": 1.5783,
+      "step": 434
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.129066107030432e-06,
+      "loss": 1.6502,
+      "step": 435
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.150052465897169e-06,
+      "loss": 1.6256,
+      "step": 436
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.171038824763904e-06,
+      "loss": 1.6528,
+      "step": 437
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.19202518363064e-06,
+      "loss": 1.6235,
+      "step": 438
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.213011542497377e-06,
+      "loss": 1.5698,
+      "step": 439
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.233997901364114e-06,
+      "loss": 1.72,
+      "step": 440
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.25498426023085e-06,
+      "loss": 1.6564,
+      "step": 441
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.275970619097588e-06,
+      "loss": 1.6199,
+      "step": 442
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.296956977964324e-06,
+      "loss": 1.629,
+      "step": 443
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.31794333683106e-06,
+      "loss": 1.6373,
+      "step": 444
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.338929695697796e-06,
+      "loss": 1.5384,
+      "step": 445
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.359916054564535e-06,
+      "loss": 1.6797,
+      "step": 446
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.380902413431271e-06,
+      "loss": 1.7051,
+      "step": 447
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.401888772298007e-06,
+      "loss": 1.5762,
+      "step": 448
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.422875131164743e-06,
+      "loss": 1.644,
+      "step": 449
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.44386149003148e-06,
+      "loss": 1.6213,
+      "step": 450
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.464847848898217e-06,
+      "loss": 1.5922,
+      "step": 451
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.485834207764954e-06,
+      "loss": 1.6557,
+      "step": 452
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.50682056663169e-06,
+      "loss": 1.6089,
+      "step": 453
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.527806925498427e-06,
+      "loss": 1.5902,
+      "step": 454
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.548793284365162e-06,
+      "loss": 1.6236,
+      "step": 455
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.569779643231901e-06,
+      "loss": 1.5851,
+      "step": 456
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.590766002098638e-06,
+      "loss": 1.6097,
+      "step": 457
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.611752360965373e-06,
+      "loss": 1.6116,
+      "step": 458
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.63273871983211e-06,
+      "loss": 1.6326,
+      "step": 459
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.653725078698846e-06,
+      "loss": 1.6402,
+      "step": 460
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.674711437565583e-06,
+      "loss": 1.6108,
+      "step": 461
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.69569779643232e-06,
+      "loss": 1.6395,
+      "step": 462
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.716684155299057e-06,
+      "loss": 1.5837,
+      "step": 463
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.737670514165794e-06,
+      "loss": 1.6523,
+      "step": 464
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.758656873032529e-06,
+      "loss": 1.5901,
+      "step": 465
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.779643231899265e-06,
+      "loss": 1.5935,
+      "step": 466
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.800629590766004e-06,
+      "loss": 1.6661,
+      "step": 467
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.821615949632739e-06,
+      "loss": 1.6674,
+      "step": 468
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.842602308499476e-06,
+      "loss": 1.6202,
+      "step": 469
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.863588667366213e-06,
+      "loss": 2.3063,
+      "step": 470
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.88457502623295e-06,
+      "loss": 1.6931,
+      "step": 471
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.905561385099686e-06,
+      "loss": 1.6628,
+      "step": 472
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.926547743966423e-06,
+      "loss": 1.7019,
+      "step": 473
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.94753410283316e-06,
+      "loss": 1.6177,
+      "step": 474
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.968520461699895e-06,
+      "loss": 1.6433,
+      "step": 475
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 9.989506820566632e-06,
+      "loss": 1.5849,
+      "step": 476
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.001049317943337e-05,
+      "loss": 1.6873,
+      "step": 477
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0031479538300107e-05,
+      "loss": 1.5504,
+      "step": 478
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0052465897166844e-05,
+      "loss": 1.7235,
+      "step": 479
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0073452256033579e-05,
+      "loss": 1.5909,
+      "step": 480
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0094438614900316e-05,
+      "loss": 1.6982,
+      "step": 481
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0115424973767052e-05,
+      "loss": 1.6433,
+      "step": 482
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0136411332633789e-05,
+      "loss": 1.6281,
+      "step": 483
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0157397691500526e-05,
+      "loss": 1.6722,
+      "step": 484
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0178384050367263e-05,
+      "loss": 1.6298,
+      "step": 485
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0199370409234e-05,
+      "loss": 1.5791,
+      "step": 486
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0220356768100735e-05,
+      "loss": 1.8025,
+      "step": 487
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0241343126967471e-05,
+      "loss": 1.609,
+      "step": 488
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0262329485834208e-05,
+      "loss": 1.6356,
+      "step": 489
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0283315844700945e-05,
+      "loss": 1.5952,
+      "step": 490
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0304302203567682e-05,
+      "loss": 1.6137,
+      "step": 491
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0325288562434418e-05,
+      "loss": 1.5489,
+      "step": 492
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0346274921301155e-05,
+      "loss": 1.625,
+      "step": 493
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.036726128016789e-05,
+      "loss": 1.666,
+      "step": 494
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0388247639034627e-05,
+      "loss": 1.6412,
+      "step": 495
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0409233997901364e-05,
+      "loss": 1.5354,
+      "step": 496
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.04302203567681e-05,
+      "loss": 1.6083,
+      "step": 497
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.045120671563484e-05,
+      "loss": 1.6053,
+      "step": 498
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0472193074501576e-05,
+      "loss": 1.6284,
+      "step": 499
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0493179433368313e-05,
+      "loss": 1.62,
+      "step": 500
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0514165792235048e-05,
+      "loss": 1.6059,
+      "step": 501
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0535152151101785e-05,
+      "loss": 1.6375,
+      "step": 502
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0556138509968521e-05,
+      "loss": 1.5768,
+      "step": 503
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0577124868835258e-05,
+      "loss": 1.6851,
+      "step": 504
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0598111227701995e-05,
+      "loss": 1.5886,
+      "step": 505
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0619097586568732e-05,
+      "loss": 1.6843,
+      "step": 506
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0640083945435469e-05,
+      "loss": 1.6105,
+      "step": 507
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0661070304302204e-05,
+      "loss": 1.6059,
+      "step": 508
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.068205666316894e-05,
+      "loss": 1.6103,
+      "step": 509
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0703043022035677e-05,
+      "loss": 1.5691,
+      "step": 510
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0724029380902414e-05,
+      "loss": 1.7106,
+      "step": 511
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0745015739769151e-05,
+      "loss": 1.564,
+      "step": 512
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0766002098635888e-05,
+      "loss": 1.5815,
+      "step": 513
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0786988457502624e-05,
+      "loss": 1.6302,
+      "step": 514
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.080797481636936e-05,
+      "loss": 1.5964,
+      "step": 515
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0828961175236096e-05,
+      "loss": 1.5569,
+      "step": 516
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0849947534102833e-05,
+      "loss": 1.6172,
+      "step": 517
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.087093389296957e-05,
+      "loss": 1.6113,
+      "step": 518
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0891920251836308e-05,
+      "loss": 1.614,
+      "step": 519
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0912906610703045e-05,
+      "loss": 1.639,
+      "step": 520
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0933892969569782e-05,
+      "loss": 1.656,
+      "step": 521
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0954879328436517e-05,
+      "loss": 1.6146,
+      "step": 522
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0975865687303254e-05,
+      "loss": 1.6553,
+      "step": 523
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.099685204616999e-05,
+      "loss": 1.6367,
+      "step": 524
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1017838405036727e-05,
+      "loss": 1.6827,
+      "step": 525
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1038824763903464e-05,
+      "loss": 1.6049,
+      "step": 526
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1059811122770201e-05,
+      "loss": 1.5824,
+      "step": 527
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1080797481636938e-05,
+      "loss": 1.6282,
+      "step": 528
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1101783840503673e-05,
+      "loss": 1.6213,
+      "step": 529
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.112277019937041e-05,
+      "loss": 1.6201,
+      "step": 530
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1143756558237146e-05,
+      "loss": 1.6007,
+      "step": 531
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1164742917103883e-05,
+      "loss": 1.6385,
+      "step": 532
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.118572927597062e-05,
+      "loss": 1.614,
+      "step": 533
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1206715634837357e-05,
+      "loss": 1.6435,
+      "step": 534
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1227701993704094e-05,
+      "loss": 1.5764,
+      "step": 535
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1248688352570829e-05,
+      "loss": 1.67,
+      "step": 536
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1269674711437565e-05,
+      "loss": 1.5781,
+      "step": 537
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1290661070304302e-05,
+      "loss": 1.5912,
+      "step": 538
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1311647429171039e-05,
+      "loss": 1.5667,
+      "step": 539
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1332633788037778e-05,
+      "loss": 1.5808,
+      "step": 540
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1353620146904514e-05,
+      "loss": 1.6758,
+      "step": 541
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1374606505771251e-05,
+      "loss": 1.5756,
+      "step": 542
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1395592864637986e-05,
+      "loss": 1.6306,
+      "step": 543
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1416579223504723e-05,
+      "loss": 1.6251,
+      "step": 544
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.143756558237146e-05,
+      "loss": 1.5508,
+      "step": 545
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1458551941238197e-05,
+      "loss": 1.5342,
+      "step": 546
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1479538300104933e-05,
+      "loss": 1.6361,
+      "step": 547
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.150052465897167e-05,
+      "loss": 1.6582,
+      "step": 548
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1521511017838407e-05,
+      "loss": 1.6617,
+      "step": 549
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1542497376705142e-05,
+      "loss": 1.6189,
+      "step": 550
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1563483735571879e-05,
+      "loss": 1.5812,
+      "step": 551
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1584470094438616e-05,
+      "loss": 1.7048,
+      "step": 552
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1605456453305352e-05,
+      "loss": 1.6509,
+      "step": 553
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.162644281217209e-05,
+      "loss": 1.5905,
+      "step": 554
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1647429171038826e-05,
+      "loss": 1.5302,
+      "step": 555
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1668415529905561e-05,
+      "loss": 1.5373,
+      "step": 556
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1689401888772298e-05,
+      "loss": 1.551,
+      "step": 557
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1710388247639035e-05,
+      "loss": 1.5356,
+      "step": 558
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1731374606505771e-05,
+      "loss": 1.5445,
+      "step": 559
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1752360965372508e-05,
+      "loss": 1.949,
+      "step": 560
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1773347324239247e-05,
+      "loss": 1.7391,
+      "step": 561
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1794333683105984e-05,
+      "loss": 1.5847,
+      "step": 562
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1815320041972717e-05,
+      "loss": 1.6091,
+      "step": 563
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1836306400839455e-05,
+      "loss": 1.5864,
+      "step": 564
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1857292759706192e-05,
+      "loss": 1.6074,
+      "step": 565
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1878279118572929e-05,
+      "loss": 1.5872,
+      "step": 566
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1899265477439666e-05,
+      "loss": 1.579,
+      "step": 567
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1920251836306403e-05,
+      "loss": 1.6643,
+      "step": 568
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.194123819517314e-05,
+      "loss": 1.5755,
+      "step": 569
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1962224554039874e-05,
+      "loss": 1.5936,
+      "step": 570
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.1983210912906611e-05,
+      "loss": 1.5796,
+      "step": 571
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2004197271773348e-05,
+      "loss": 1.6583,
+      "step": 572
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2025183630640085e-05,
+      "loss": 1.5116,
+      "step": 573
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2046169989506822e-05,
+      "loss": 1.598,
+      "step": 574
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2067156348373558e-05,
+      "loss": 1.6624,
+      "step": 575
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2088142707240295e-05,
+      "loss": 1.5728,
+      "step": 576
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.210912906610703e-05,
+      "loss": 1.7116,
+      "step": 577
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2130115424973767e-05,
+      "loss": 1.8864,
+      "step": 578
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2151101783840504e-05,
+      "loss": 1.5962,
+      "step": 579
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.217208814270724e-05,
+      "loss": 1.5681,
+      "step": 580
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2193074501573977e-05,
+      "loss": 1.638,
+      "step": 581
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2214060860440716e-05,
+      "loss": 1.6252,
+      "step": 582
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2235047219307453e-05,
+      "loss": 1.5895,
+      "step": 583
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2256033578174186e-05,
+      "loss": 1.6143,
+      "step": 584
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2277019937040925e-05,
+      "loss": 1.6593,
+      "step": 585
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2298006295907661e-05,
+      "loss": 1.5382,
+      "step": 586
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2318992654774398e-05,
+      "loss": 1.5573,
+      "step": 587
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2339979013641135e-05,
+      "loss": 1.5502,
+      "step": 588
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2360965372507872e-05,
+      "loss": 1.6078,
+      "step": 589
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2381951731374608e-05,
+      "loss": 1.5852,
+      "step": 590
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2402938090241344e-05,
+      "loss": 1.5747,
+      "step": 591
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.242392444910808e-05,
+      "loss": 1.6304,
+      "step": 592
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2444910807974817e-05,
+      "loss": 1.5659,
+      "step": 593
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2465897166841554e-05,
+      "loss": 1.5801,
+      "step": 594
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.248688352570829e-05,
+      "loss": 1.576,
+      "step": 595
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2507869884575028e-05,
+      "loss": 1.6446,
+      "step": 596
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2528856243441764e-05,
+      "loss": 1.6377,
+      "step": 597
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 1.6429214477539062,
+      "eval_runtime": 77.2709,
+      "eval_samples_per_second": 1.669,
+      "eval_steps_per_second": 0.116,
+      "step": 597
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.25498426023085e-05,
+      "loss": 1.6151,
+      "step": 598
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2570828961175236e-05,
+      "loss": 2.2074,
+      "step": 599
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2591815320041973e-05,
+      "loss": 1.7116,
+      "step": 600
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.261280167890871e-05,
+      "loss": 1.5336,
+      "step": 601
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2633788037775447e-05,
+      "loss": 1.5948,
+      "step": 602
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2654774396642185e-05,
+      "loss": 1.6338,
+      "step": 603
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2675760755508922e-05,
+      "loss": 1.5316,
+      "step": 604
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2696747114375655e-05,
+      "loss": 1.5703,
+      "step": 605
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2717733473242394e-05,
+      "loss": 1.6207,
+      "step": 606
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.273871983210913e-05,
+      "loss": 1.625,
+      "step": 607
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2759706190975867e-05,
+      "loss": 1.5978,
+      "step": 608
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2780692549842604e-05,
+      "loss": 1.5619,
+      "step": 609
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2801678908709341e-05,
+      "loss": 1.6132,
+      "step": 610
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2822665267576078e-05,
+      "loss": 1.5934,
+      "step": 611
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2843651626442813e-05,
+      "loss": 1.5927,
+      "step": 612
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.286463798530955e-05,
+      "loss": 1.5585,
+      "step": 613
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2885624344176286e-05,
+      "loss": 1.5742,
+      "step": 614
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2906610703043023e-05,
+      "loss": 1.5541,
+      "step": 615
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.292759706190976e-05,
+      "loss": 1.7178,
+      "step": 616
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2948583420776497e-05,
+      "loss": 1.5816,
+      "step": 617
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2969569779643233e-05,
+      "loss": 1.5809,
+      "step": 618
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2990556138509969e-05,
+      "loss": 1.5444,
+      "step": 619
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3011542497376705e-05,
+      "loss": 1.6537,
+      "step": 620
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3032528856243442e-05,
+      "loss": 1.5833,
+      "step": 621
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3053515215110179e-05,
+      "loss": 1.6522,
+      "step": 622
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3074501573976916e-05,
+      "loss": 1.5983,
+      "step": 623
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3095487932843654e-05,
+      "loss": 1.4808,
+      "step": 624
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3116474291710388e-05,
+      "loss": 1.6065,
+      "step": 625
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3137460650577124e-05,
+      "loss": 1.4926,
+      "step": 626
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3158447009443863e-05,
+      "loss": 1.5984,
+      "step": 627
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.31794333683106e-05,
+      "loss": 1.5363,
+      "step": 628
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3200419727177336e-05,
+      "loss": 1.5889,
+      "step": 629
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3221406086044073e-05,
+      "loss": 1.5801,
+      "step": 630
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.324239244491081e-05,
+      "loss": 1.6418,
+      "step": 631
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3263378803777545e-05,
+      "loss": 1.6004,
+      "step": 632
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3284365162644282e-05,
+      "loss": 1.6094,
+      "step": 633
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3305351521511019e-05,
+      "loss": 1.5918,
+      "step": 634
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3326337880377755e-05,
+      "loss": 1.6469,
+      "step": 635
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3347324239244492e-05,
+      "loss": 1.5664,
+      "step": 636
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3368310598111229e-05,
+      "loss": 1.6178,
+      "step": 637
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3389296956977966e-05,
+      "loss": 1.6241,
+      "step": 638
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3410283315844701e-05,
+      "loss": 1.4875,
+      "step": 639
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3431269674711438e-05,
+      "loss": 1.5979,
+      "step": 640
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3452256033578175e-05,
+      "loss": 1.6157,
+      "step": 641
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3473242392444911e-05,
+      "loss": 1.5507,
+      "step": 642
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3494228751311648e-05,
+      "loss": 1.5974,
+      "step": 643
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3515215110178385e-05,
+      "loss": 1.5852,
+      "step": 644
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3536201469045123e-05,
+      "loss": 1.6494,
+      "step": 645
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3557187827911857e-05,
+      "loss": 1.6005,
+      "step": 646
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3578174186778594e-05,
+      "loss": 1.5783,
+      "step": 647
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3599160545645332e-05,
+      "loss": 1.5363,
+      "step": 648
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3620146904512069e-05,
+      "loss": 1.5401,
+      "step": 649
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3641133263378806e-05,
+      "loss": 1.611,
+      "step": 650
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3662119622245542e-05,
+      "loss": 1.6207,
+      "step": 651
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.368310598111228e-05,
+      "loss": 1.5563,
+      "step": 652
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3704092339979014e-05,
+      "loss": 1.583,
+      "step": 653
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3725078698845751e-05,
+      "loss": 1.6386,
+      "step": 654
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3746065057712488e-05,
+      "loss": 1.5913,
+      "step": 655
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3767051416579225e-05,
+      "loss": 1.6088,
+      "step": 656
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3788037775445961e-05,
+      "loss": 1.5821,
+      "step": 657
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3809024134312698e-05,
+      "loss": 1.4714,
+      "step": 658
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3830010493179435e-05,
+      "loss": 1.5414,
+      "step": 659
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.385099685204617e-05,
+      "loss": 1.5732,
+      "step": 660
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3871983210912907e-05,
+      "loss": 1.6069,
+      "step": 661
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3892969569779644e-05,
+      "loss": 1.6316,
+      "step": 662
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.391395592864638e-05,
+      "loss": 1.5834,
+      "step": 663
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3934942287513117e-05,
+      "loss": 1.5639,
+      "step": 664
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3955928646379854e-05,
+      "loss": 1.5151,
+      "step": 665
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3976915005246593e-05,
+      "loss": 1.5352,
+      "step": 666
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.3997901364113326e-05,
+      "loss": 1.5703,
+      "step": 667
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4018887722980063e-05,
+      "loss": 1.6609,
+      "step": 668
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4039874081846801e-05,
+      "loss": 1.5805,
+      "step": 669
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4060860440713538e-05,
+      "loss": 1.5583,
+      "step": 670
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4081846799580275e-05,
+      "loss": 1.5771,
+      "step": 671
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4102833158447012e-05,
+      "loss": 1.6637,
+      "step": 672
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4123819517313748e-05,
+      "loss": 1.5315,
+      "step": 673
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4144805876180483e-05,
+      "loss": 1.5379,
+      "step": 674
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.416579223504722e-05,
+      "loss": 1.5173,
+      "step": 675
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4186778593913957e-05,
+      "loss": 1.5092,
+      "step": 676
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4207764952780694e-05,
+      "loss": 1.5365,
+      "step": 677
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.422875131164743e-05,
+      "loss": 1.6046,
+      "step": 678
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4249737670514167e-05,
+      "loss": 1.6644,
+      "step": 679
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4270724029380904e-05,
+      "loss": 1.6196,
+      "step": 680
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.429171038824764e-05,
+      "loss": 1.5573,
+      "step": 681
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4312696747114376e-05,
+      "loss": 1.5819,
+      "step": 682
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4333683105981113e-05,
+      "loss": 1.5788,
+      "step": 683
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.435466946484785e-05,
+      "loss": 1.6133,
+      "step": 684
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4375655823714586e-05,
+      "loss": 1.5393,
+      "step": 685
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4396642182581323e-05,
+      "loss": 1.5798,
+      "step": 686
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4417628541448062e-05,
+      "loss": 1.5651,
+      "step": 687
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4438614900314795e-05,
+      "loss": 1.5316,
+      "step": 688
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4459601259181532e-05,
+      "loss": 1.604,
+      "step": 689
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.448058761804827e-05,
+      "loss": 1.5658,
+      "step": 690
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4501573976915007e-05,
+      "loss": 1.5505,
+      "step": 691
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4522560335781744e-05,
+      "loss": 1.5486,
+      "step": 692
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.454354669464848e-05,
+      "loss": 1.6992,
+      "step": 693
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4564533053515218e-05,
+      "loss": 1.6283,
+      "step": 694
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4585519412381953e-05,
+      "loss": 1.601,
+      "step": 695
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.460650577124869e-05,
+      "loss": 1.6118,
+      "step": 696
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4627492130115426e-05,
+      "loss": 1.5565,
+      "step": 697
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4648478488982163e-05,
+      "loss": 1.639,
+      "step": 698
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.46694648478489e-05,
+      "loss": 1.5883,
+      "step": 699
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4690451206715637e-05,
+      "loss": 1.5758,
+      "step": 700
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4711437565582372e-05,
+      "loss": 1.6378,
+      "step": 701
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4732423924449108e-05,
+      "loss": 1.6504,
+      "step": 702
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4753410283315845e-05,
+      "loss": 1.5978,
+      "step": 703
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4774396642182582e-05,
+      "loss": 1.5471,
+      "step": 704
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4795383001049319e-05,
+      "loss": 1.5618,
+      "step": 705
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4816369359916056e-05,
+      "loss": 1.6846,
+      "step": 706
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4837355718782794e-05,
+      "loss": 1.5838,
+      "step": 707
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4858342077649527e-05,
+      "loss": 1.5858,
+      "step": 708
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4879328436516264e-05,
+      "loss": 1.5917,
+      "step": 709
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4900314795383001e-05,
+      "loss": 1.5929,
+      "step": 710
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.492130115424974e-05,
+      "loss": 1.5476,
+      "step": 711
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4942287513116476e-05,
+      "loss": 1.587,
+      "step": 712
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4963273871983213e-05,
+      "loss": 1.6299,
+      "step": 713
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.498426023084995e-05,
+      "loss": 1.5708,
+      "step": 714
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.5005246589716685e-05,
+      "loss": 1.6735,
+      "step": 715
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5026232948583422e-05,
+      "loss": 1.5452,
+      "step": 716
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5047219307450159e-05,
+      "loss": 1.5681,
+      "step": 717
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5068205666316895e-05,
+      "loss": 1.7473,
+      "step": 718
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5089192025183632e-05,
+      "loss": 1.5579,
+      "step": 719
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5110178384050369e-05,
+      "loss": 1.6335,
+      "step": 720
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5131164742917106e-05,
+      "loss": 1.6215,
+      "step": 721
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.515215110178384e-05,
+      "loss": 1.5444,
+      "step": 722
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5173137460650578e-05,
+      "loss": 1.5838,
+      "step": 723
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5194123819517314e-05,
+      "loss": 1.5643,
+      "step": 724
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5215110178384051e-05,
+      "loss": 1.5059,
+      "step": 725
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5236096537250788e-05,
+      "loss": 1.5571,
+      "step": 726
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5257082896117525e-05,
+      "loss": 1.558,
+      "step": 727
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5278069254984263e-05,
+      "loss": 1.7263,
+      "step": 728
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5299055613851e-05,
+      "loss": 1.647,
+      "step": 729
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5320041972717733e-05,
+      "loss": 1.5449,
+      "step": 730
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5341028331584472e-05,
+      "loss": 1.5539,
+      "step": 731
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5362014690451207e-05,
+      "loss": 1.6365,
+      "step": 732
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5383001049317945e-05,
+      "loss": 1.6481,
+      "step": 733
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.540398740818468e-05,
+      "loss": 1.6083,
+      "step": 734
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.542497376705142e-05,
+      "loss": 1.5466,
+      "step": 735
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5445960125918154e-05,
+      "loss": 1.5735,
+      "step": 736
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.546694648478489e-05,
+      "loss": 1.5326,
+      "step": 737
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5487932843651628e-05,
+      "loss": 1.4715,
+      "step": 738
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5508919202518363e-05,
+      "loss": 1.5341,
+      "step": 739
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.55299055613851e-05,
+      "loss": 1.8178,
+      "step": 740
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.555089192025184e-05,
+      "loss": 1.597,
+      "step": 741
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5571878279118575e-05,
+      "loss": 1.5185,
+      "step": 742
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.559286463798531e-05,
+      "loss": 1.6017,
+      "step": 743
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.561385099685205e-05,
+      "loss": 1.535,
+      "step": 744
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5634837355718784e-05,
+      "loss": 1.5732,
+      "step": 745
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5655823714585522e-05,
+      "loss": 1.4951,
+      "step": 746
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5676810073452257e-05,
+      "loss": 1.6071,
+      "step": 747
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5697796432318996e-05,
+      "loss": 1.5926,
+      "step": 748
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.571878279118573e-05,
+      "loss": 1.5553,
+      "step": 749
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5739769150052466e-05,
+      "loss": 1.7759,
+      "step": 750
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5760755508919204e-05,
+      "loss": 1.6264,
+      "step": 751
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.578174186778594e-05,
+      "loss": 1.5323,
+      "step": 752
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5802728226652678e-05,
+      "loss": 1.5629,
+      "step": 753
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5823714585519413e-05,
+      "loss": 1.6094,
+      "step": 754
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.584470094438615e-05,
+      "loss": 1.6321,
+      "step": 755
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5865687303252887e-05,
+      "loss": 1.6866,
+      "step": 756
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.588667366211962e-05,
+      "loss": 1.6107,
+      "step": 757
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.590766002098636e-05,
+      "loss": 1.6612,
+      "step": 758
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5928646379853095e-05,
+      "loss": 1.8321,
+      "step": 759
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5949632738719834e-05,
+      "loss": 1.5567,
+      "step": 760
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.597061909758657e-05,
+      "loss": 1.5805,
+      "step": 761
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5991605456453307e-05,
+      "loss": 1.5654,
+      "step": 762
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6012591815320046e-05,
+      "loss": 1.5912,
+      "step": 763
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6033578174186777e-05,
+      "loss": 1.4732,
+      "step": 764
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6054564533053516e-05,
+      "loss": 1.5568,
+      "step": 765
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6075550891920254e-05,
+      "loss": 1.6122,
+      "step": 766
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.609653725078699e-05,
+      "loss": 1.5611,
+      "step": 767
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6117523609653728e-05,
+      "loss": 1.5996,
+      "step": 768
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6138509968520463e-05,
+      "loss": 1.5696,
+      "step": 769
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6159496327387198e-05,
+      "loss": 1.5436,
+      "step": 770
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6180482686253937e-05,
+      "loss": 1.5601,
+      "step": 771
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6201469045120672e-05,
+      "loss": 1.6172,
+      "step": 772
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.622245540398741e-05,
+      "loss": 1.5121,
+      "step": 773
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6243441762854145e-05,
+      "loss": 1.5513,
+      "step": 774
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6264428121720884e-05,
+      "loss": 1.5661,
+      "step": 775
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.628541448058762e-05,
+      "loss": 1.5684,
+      "step": 776
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6306400839454354e-05,
+      "loss": 1.5548,
+      "step": 777
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6327387198321092e-05,
+      "loss": 1.5235,
+      "step": 778
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6348373557187828e-05,
+      "loss": 1.5705,
+      "step": 779
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6369359916054566e-05,
+      "loss": 1.5811,
+      "step": 780
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.63903462749213e-05,
+      "loss": 1.6117,
+      "step": 781
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.641133263378804e-05,
+      "loss": 1.5395,
+      "step": 782
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6432318992654778e-05,
+      "loss": 1.543,
+      "step": 783
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.645330535152151e-05,
+      "loss": 1.5557,
+      "step": 784
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.647429171038825e-05,
+      "loss": 1.6305,
+      "step": 785
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6495278069254987e-05,
+      "loss": 1.5439,
+      "step": 786
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6516264428121722e-05,
+      "loss": 1.5047,
+      "step": 787
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.653725078698846e-05,
+      "loss": 1.6241,
+      "step": 788
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6558237145855195e-05,
+      "loss": 1.5123,
+      "step": 789
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6579223504721934e-05,
+      "loss": 1.5872,
+      "step": 790
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.660020986358867e-05,
+      "loss": 1.6289,
+      "step": 791
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6621196222455404e-05,
+      "loss": 1.5276,
+      "step": 792
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6642182581322143e-05,
+      "loss": 1.6422,
+      "step": 793
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6663168940188878e-05,
+      "loss": 1.634,
+      "step": 794
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6684155299055616e-05,
+      "loss": 1.5484,
+      "step": 795
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.670514165792235e-05,
+      "loss": 1.6011,
+      "step": 796
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 1.6383761167526245,
+      "eval_runtime": 77.154,
+      "eval_samples_per_second": 1.672,
+      "eval_steps_per_second": 0.117,
+      "step": 796
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.672612801678909e-05,
+      "loss": 1.5683,
+      "step": 797
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6747114375655825e-05,
+      "loss": 1.703,
+      "step": 798
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.676810073452256e-05,
+      "loss": 1.5666,
+      "step": 799
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.67890870933893e-05,
+      "loss": 1.586,
+      "step": 800
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6810073452256034e-05,
+      "loss": 1.5253,
+      "step": 801
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6831059811122772e-05,
+      "loss": 1.6577,
+      "step": 802
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.685204616998951e-05,
+      "loss": 1.6639,
+      "step": 803
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6873032528856246e-05,
+      "loss": 1.5758,
+      "step": 804
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.689401888772298e-05,
+      "loss": 1.624,
+      "step": 805
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6915005246589716e-05,
+      "loss": 1.5652,
+      "step": 806
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6935991605456454e-05,
+      "loss": 1.6917,
+      "step": 807
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6956977964323193e-05,
+      "loss": 1.5333,
+      "step": 808
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6977964323189928e-05,
+      "loss": 1.7285,
+      "step": 809
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6998950682056666e-05,
+      "loss": 1.7412,
+      "step": 810
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.70199370409234e-05,
+      "loss": 1.6371,
+      "step": 811
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7040923399790137e-05,
+      "loss": 1.6427,
+      "step": 812
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7061909758656875e-05,
+      "loss": 1.5976,
+      "step": 813
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.708289611752361e-05,
+      "loss": 1.5828,
+      "step": 814
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.710388247639035e-05,
+      "loss": 1.5149,
+      "step": 815
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7124868835257084e-05,
+      "loss": 1.5048,
+      "step": 816
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7145855194123822e-05,
+      "loss": 1.5827,
+      "step": 817
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7166841552990557e-05,
+      "loss": 1.5407,
+      "step": 818
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7187827911857292e-05,
+      "loss": 1.5344,
+      "step": 819
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.720881427072403e-05,
+      "loss": 1.6077,
+      "step": 820
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7229800629590766e-05,
+      "loss": 1.556,
+      "step": 821
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7250786988457504e-05,
+      "loss": 1.6861,
+      "step": 822
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.727177334732424e-05,
+      "loss": 1.5831,
+      "step": 823
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7292759706190978e-05,
+      "loss": 1.6158,
+      "step": 824
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7313746065057716e-05,
+      "loss": 1.4739,
+      "step": 825
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7334732423924448e-05,
+      "loss": 1.5502,
+      "step": 826
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7355718782791187e-05,
+      "loss": 1.5744,
+      "step": 827
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7376705141657925e-05,
+      "loss": 1.5399,
+      "step": 828
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.739769150052466e-05,
+      "loss": 1.606,
+      "step": 829
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.74186778593914e-05,
+      "loss": 1.5352,
+      "step": 830
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7439664218258134e-05,
+      "loss": 1.6509,
+      "step": 831
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7460650577124872e-05,
+      "loss": 1.5447,
+      "step": 832
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7481636935991607e-05,
+      "loss": 1.5429,
+      "step": 833
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7502623294858342e-05,
+      "loss": 1.6183,
+      "step": 834
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.752360965372508e-05,
+      "loss": 1.5557,
+      "step": 835
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7544596012591816e-05,
+      "loss": 1.573,
+      "step": 836
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7565582371458555e-05,
+      "loss": 1.5983,
+      "step": 837
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.758656873032529e-05,
+      "loss": 1.5386,
+      "step": 838
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7607555089192028e-05,
+      "loss": 1.5736,
+      "step": 839
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7628541448058763e-05,
+      "loss": 1.5197,
+      "step": 840
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7649527806925498e-05,
+      "loss": 1.5786,
+      "step": 841
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7670514165792237e-05,
+      "loss": 1.508,
+      "step": 842
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7691500524658972e-05,
+      "loss": 1.6392,
+      "step": 843
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.771248688352571e-05,
+      "loss": 1.1548,
+      "step": 844
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.773347324239245e-05,
+      "loss": 0.9182,
+      "step": 845
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.775445960125918e-05,
+      "loss": 0.9304,
+      "step": 846
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.777544596012592e-05,
+      "loss": 0.8864,
+      "step": 847
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7796432318992654e-05,
+      "loss": 1.0287,
+      "step": 848
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7817418677859393e-05,
+      "loss": 1.5457,
+      "step": 849
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.783840503672613e-05,
+      "loss": 1.6043,
+      "step": 850
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7859391395592866e-05,
+      "loss": 1.5983,
+      "step": 851
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7880377754459605e-05,
+      "loss": 1.5686,
+      "step": 852
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.790136411332634e-05,
+      "loss": 1.5884,
+      "step": 853
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7922350472193075e-05,
+      "loss": 1.6224,
+      "step": 854
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7943336831059813e-05,
+      "loss": 1.6756,
+      "step": 855
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.796432318992655e-05,
+      "loss": 1.6088,
+      "step": 856
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.7985309548793287e-05,
+      "loss": 1.5734,
+      "step": 857
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8006295907660022e-05,
+      "loss": 1.5858,
+      "step": 858
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.802728226652676e-05,
+      "loss": 1.5623,
+      "step": 859
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8048268625393496e-05,
+      "loss": 1.586,
+      "step": 860
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.806925498426023e-05,
+      "loss": 1.5358,
+      "step": 861
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.809024134312697e-05,
+      "loss": 1.5501,
+      "step": 862
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8111227701993704e-05,
+      "loss": 1.5555,
+      "step": 863
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8132214060860443e-05,
+      "loss": 1.5547,
+      "step": 864
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8153200419727178e-05,
+      "loss": 1.6531,
+      "step": 865
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8174186778593916e-05,
+      "loss": 1.5032,
+      "step": 866
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.819517313746065e-05,
+      "loss": 1.5672,
+      "step": 867
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8216159496327386e-05,
+      "loss": 1.588,
+      "step": 868
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8237145855194125e-05,
+      "loss": 1.5914,
+      "step": 869
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8258132214060863e-05,
+      "loss": 1.5484,
+      "step": 870
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.82791185729276e-05,
+      "loss": 1.5893,
+      "step": 871
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8300104931794337e-05,
+      "loss": 1.6191,
+      "step": 872
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8321091290661072e-05,
+      "loss": 1.55,
+      "step": 873
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8342077649527807e-05,
+      "loss": 1.5245,
+      "step": 874
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8363064008394546e-05,
+      "loss": 1.5176,
+      "step": 875
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.838405036726128e-05,
+      "loss": 1.64,
+      "step": 876
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.840503672612802e-05,
+      "loss": 1.5436,
+      "step": 877
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8426023084994754e-05,
+      "loss": 1.6874,
+      "step": 878
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8447009443861493e-05,
+      "loss": 1.788,
+      "step": 879
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8467995802728228e-05,
+      "loss": 1.5587,
+      "step": 880
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8488982161594963e-05,
+      "loss": 1.5652,
+      "step": 881
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.85099685204617e-05,
+      "loss": 1.5103,
+      "step": 882
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8530954879328437e-05,
+      "loss": 1.5965,
+      "step": 883
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8551941238195175e-05,
+      "loss": 1.548,
+      "step": 884
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.857292759706191e-05,
+      "loss": 1.6627,
+      "step": 885
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.859391395592865e-05,
+      "loss": 1.5303,
+      "step": 886
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8614900314795387e-05,
+      "loss": 1.6205,
+      "step": 887
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.863588667366212e-05,
+      "loss": 1.5995,
+      "step": 888
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8656873032528857e-05,
+      "loss": 1.6407,
+      "step": 889
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8677859391395592e-05,
+      "loss": 1.5884,
+      "step": 890
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.869884575026233e-05,
+      "loss": 1.535,
+      "step": 891
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.871983210912907e-05,
+      "loss": 1.5703,
+      "step": 892
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8740818467995805e-05,
+      "loss": 1.5606,
+      "step": 893
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8761804826862543e-05,
+      "loss": 1.6007,
+      "step": 894
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8782791185729278e-05,
+      "loss": 1.5868,
+      "step": 895
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8803777544596013e-05,
+      "loss": 1.5695,
+      "step": 896
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.882476390346275e-05,
+      "loss": 1.5453,
+      "step": 897
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8845750262329487e-05,
+      "loss": 1.5878,
+      "step": 898
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8866736621196225e-05,
+      "loss": 1.5869,
+      "step": 899
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.888772298006296e-05,
+      "loss": 1.5699,
+      "step": 900
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.89087093389297e-05,
+      "loss": 1.567,
+      "step": 901
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8929695697796434e-05,
+      "loss": 1.6158,
+      "step": 902
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.895068205666317e-05,
+      "loss": 1.5887,
+      "step": 903
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8971668415529907e-05,
+      "loss": 1.5829,
+      "step": 904
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8992654774396643e-05,
+      "loss": 1.5321,
+      "step": 905
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.901364113326338e-05,
+      "loss": 1.5625,
+      "step": 906
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9034627492130116e-05,
+      "loss": 1.5512,
+      "step": 907
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9055613850996855e-05,
+      "loss": 1.6162,
+      "step": 908
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.907660020986359e-05,
+      "loss": 1.5466,
+      "step": 909
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9097586568730325e-05,
+      "loss": 1.5268,
+      "step": 910
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9118572927597063e-05,
+      "loss": 1.5332,
+      "step": 911
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9139559286463802e-05,
+      "loss": 1.5468,
+      "step": 912
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9160545645330537e-05,
+      "loss": 1.6147,
+      "step": 913
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9181532004197275e-05,
+      "loss": 1.5977,
+      "step": 914
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.920251836306401e-05,
+      "loss": 1.5964,
+      "step": 915
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9223504721930746e-05,
+      "loss": 1.55,
+      "step": 916
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9244491080797484e-05,
+      "loss": 1.5841,
+      "step": 917
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.926547743966422e-05,
+      "loss": 1.6234,
+      "step": 918
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9286463798530958e-05,
+      "loss": 1.577,
+      "step": 919
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9307450157397693e-05,
+      "loss": 1.5822,
+      "step": 920
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.932843651626443e-05,
+      "loss": 1.5758,
+      "step": 921
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9349422875131166e-05,
+      "loss": 1.5701,
+      "step": 922
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.93704092339979e-05,
+      "loss": 1.5647,
+      "step": 923
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.939139559286464e-05,
+      "loss": 1.5702,
+      "step": 924
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9412381951731375e-05,
+      "loss": 1.6056,
+      "step": 925
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9433368310598113e-05,
+      "loss": 1.5622,
+      "step": 926
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.945435466946485e-05,
+      "loss": 1.639,
+      "step": 927
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9475341028331587e-05,
+      "loss": 1.5483,
+      "step": 928
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9496327387198322e-05,
+      "loss": 1.5727,
+      "step": 929
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9517313746065057e-05,
+      "loss": 1.5567,
+      "step": 930
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9538300104931796e-05,
+      "loss": 1.534,
+      "step": 931
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.955928646379853e-05,
+      "loss": 1.5235,
+      "step": 932
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.958027282266527e-05,
+      "loss": 1.5701,
+      "step": 933
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9601259181532008e-05,
+      "loss": 1.6203,
+      "step": 934
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9622245540398743e-05,
+      "loss": 1.532,
+      "step": 935
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9643231899265478e-05,
+      "loss": 1.6,
+      "step": 936
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9664218258132216e-05,
+      "loss": 1.5812,
+      "step": 937
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.968520461699895e-05,
+      "loss": 1.6079,
+      "step": 938
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.970619097586569e-05,
+      "loss": 1.5699,
+      "step": 939
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9727177334732425e-05,
+      "loss": 1.5731,
+      "step": 940
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9748163693599164e-05,
+      "loss": 1.6092,
+      "step": 941
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.97691500524659e-05,
+      "loss": 1.5324,
+      "step": 942
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9790136411332634e-05,
+      "loss": 1.5997,
+      "step": 943
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9811122770199372e-05,
+      "loss": 1.6267,
+      "step": 944
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9832109129066107e-05,
+      "loss": 1.5912,
+      "step": 945
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9853095487932846e-05,
+      "loss": 1.5486,
+      "step": 946
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.987408184679958e-05,
+      "loss": 1.5681,
+      "step": 947
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.989506820566632e-05,
+      "loss": 1.5839,
+      "step": 948
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9916054564533054e-05,
+      "loss": 1.5247,
+      "step": 949
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.993704092339979e-05,
+      "loss": 1.6515,
+      "step": 950
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9958027282266528e-05,
+      "loss": 1.5125,
+      "step": 951
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9979013641133263e-05,
+      "loss": 1.4875,
+      "step": 952
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 2e-05,
+      "loss": 1.5045,
+      "step": 953
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999999933012847e-05,
+      "loss": 1.5198,
+      "step": 954
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999997320513974e-05,
+      "loss": 1.6102,
+      "step": 955
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999999397115678e-05,
+      "loss": 1.5699,
+      "step": 956
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999998928205733e-05,
+      "loss": 1.5528,
+      "step": 957
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999998325321626e-05,
+      "loss": 1.5786,
+      "step": 958
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999975884634376e-05,
+      "loss": 1.6234,
+      "step": 959
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999967176312664e-05,
+      "loss": 1.524,
+      "step": 960
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999957128252288e-05,
+      "loss": 1.5765,
+      "step": 961
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999945740454604e-05,
+      "loss": 1.602,
+      "step": 962
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999933012921127e-05,
+      "loss": 1.5577,
+      "step": 963
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999918945653563e-05,
+      "loss": 1.5573,
+      "step": 964
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999903538653807e-05,
+      "loss": 1.5538,
+      "step": 965
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999988679192391e-05,
+      "loss": 1.578,
+      "step": 966
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999868705466126e-05,
+      "loss": 1.5514,
+      "step": 967
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999849279282872e-05,
+      "loss": 1.58,
+      "step": 968
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999828513376753e-05,
+      "loss": 1.5879,
+      "step": 969
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999980640775055e-05,
+      "loss": 1.6396,
+      "step": 970
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999782962407222e-05,
+      "loss": 1.5468,
+      "step": 971
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999975817734992e-05,
+      "loss": 1.6131,
+      "step": 972
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999732052581954e-05,
+      "loss": 1.7201,
+      "step": 973
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999704588106822e-05,
+      "loss": 1.567,
+      "step": 974
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999675783928216e-05,
+      "loss": 1.5601,
+      "step": 975
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999645640049988e-05,
+      "loss": 1.3966,
+      "step": 976
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999614156476177e-05,
+      "loss": 1.5359,
+      "step": 977
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999581333211e-05,
+      "loss": 1.5001,
+      "step": 978
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999547170258853e-05,
+      "loss": 1.6212,
+      "step": 979
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999951166762432e-05,
+      "loss": 1.5539,
+      "step": 980
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999947482531215e-05,
+      "loss": 1.5813,
+      "step": 981
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999943664332728e-05,
+      "loss": 1.6095,
+      "step": 982
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999397121674828e-05,
+      "loss": 1.4996,
+      "step": 983
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999356260360086e-05,
+      "loss": 1.5519,
+      "step": 984
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999314059388534e-05,
+      "loss": 1.5799,
+      "step": 985
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999927051876582e-05,
+      "loss": 1.5265,
+      "step": 986
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999922563849778e-05,
+      "loss": 1.6101,
+      "step": 987
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999179418590427e-05,
+      "loss": 1.5486,
+      "step": 988
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999131859049954e-05,
+      "loss": 1.5245,
+      "step": 989
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999082959882727e-05,
+      "loss": 1.6696,
+      "step": 990
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999032721095303e-05,
+      "loss": 1.5314,
+      "step": 991
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999898114269441e-05,
+      "loss": 1.564,
+      "step": 992
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9998928224686964e-05,
+      "loss": 1.5138,
+      "step": 993
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9998873967080046e-05,
+      "loss": 1.5107,
+      "step": 994
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999881836988093e-05,
+      "loss": 1.5196,
+      "step": 995
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 1.6060134172439575,
+      "eval_runtime": 77.2855,
+      "eval_samples_per_second": 1.669,
+      "eval_steps_per_second": 0.116,
+      "step": 995
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9998761433097066e-05,
+      "loss": 1.6001,
+      "step": 996
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9998703156736078e-05,
+      "loss": 1.6015,
+      "step": 997
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9998643540805776e-05,
+      "loss": 1.5665,
+      "step": 998
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9998582585314145e-05,
+      "loss": 1.6225,
+      "step": 999
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999852029026935e-05,
+      "loss": 1.5486,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9998456655679746e-05,
+      "loss": 1.5435,
+      "step": 1001
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999839168155385e-05,
+      "loss": 1.5444,
+      "step": 1002
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9998325367900368e-05,
+      "loss": 1.5587,
+      "step": 1003
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9998257714728187e-05,
+      "loss": 1.5964,
+      "step": 1004
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9998188722046366e-05,
+      "loss": 1.5703,
+      "step": 1005
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9998118389864153e-05,
+      "loss": 1.5835,
+      "step": 1006
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999804671819097e-05,
+      "loss": 1.6584,
+      "step": 1007
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999797370703642e-05,
+      "loss": 1.8835,
+      "step": 1008
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9997899356410276e-05,
+      "loss": 1.649,
+      "step": 1009
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999782366632251e-05,
+      "loss": 1.448,
+      "step": 1010
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9997746636783262e-05,
+      "loss": 1.5062,
+      "step": 1011
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9997668267802842e-05,
+      "loss": 1.533,
+      "step": 1012
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999758855939176e-05,
+      "loss": 1.6005,
+      "step": 1013
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9997507511560692e-05,
+      "loss": 1.561,
+      "step": 1014
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9997425124320494e-05,
+      "loss": 1.5232,
+      "step": 1015
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.99973413976822e-05,
+      "loss": 1.5334,
+      "step": 1016
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9997256331657037e-05,
+      "loss": 1.5691,
+      "step": 1017
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9997169926256397e-05,
+      "loss": 1.6222,
+      "step": 1018
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9997082181491855e-05,
+      "loss": 1.571,
+      "step": 1019
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9996993097375164e-05,
+      "loss": 1.5638,
+      "step": 1020
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9996902673918268e-05,
+      "loss": 1.5877,
+      "step": 1021
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9996810911133272e-05,
+      "loss": 1.5456,
+      "step": 1022
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9996717809032475e-05,
+      "loss": 1.5446,
+      "step": 1023
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999662336762835e-05,
+      "loss": 1.5521,
+      "step": 1024
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999652758693355e-05,
+      "loss": 1.5221,
+      "step": 1025
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9996430466960898e-05,
+      "loss": 1.5721,
+      "step": 1026
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999633200772342e-05,
+      "loss": 1.5508,
+      "step": 1027
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.99962322092343e-05,
+      "loss": 1.5228,
+      "step": 1028
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999613107150691e-05,
+      "loss": 1.523,
+      "step": 1029
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9996028594554794e-05,
+      "loss": 1.5308,
+      "step": 1030
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9995924778391686e-05,
+      "loss": 1.5751,
+      "step": 1031
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9995819623031498e-05,
+      "loss": 1.5671,
+      "step": 1032
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9995713128488313e-05,
+      "loss": 1.5968,
+      "step": 1033
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9995605294776403e-05,
+      "loss": 1.5225,
+      "step": 1034
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999549612191021e-05,
+      "loss": 1.5656,
+      "step": 1035
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9995385609904363e-05,
+      "loss": 1.6424,
+      "step": 1036
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9995273758773667e-05,
+      "loss": 1.4378,
+      "step": 1037
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999516056853311e-05,
+      "loss": 1.5453,
+      "step": 1038
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999504603919785e-05,
+      "loss": 1.5093,
+      "step": 1039
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9994930170783238e-05,
+      "loss": 1.5994,
+      "step": 1040
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9994812963304794e-05,
+      "loss": 1.5814,
+      "step": 1041
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999469441677822e-05,
+      "loss": 1.5619,
+      "step": 1042
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9994574531219402e-05,
+      "loss": 1.5861,
+      "step": 1043
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9994453306644397e-05,
+      "loss": 1.6174,
+      "step": 1044
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999433074306945e-05,
+      "loss": 1.6974,
+      "step": 1045
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9994206840510975e-05,
+      "loss": 1.7326,
+      "step": 1046
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9994081598985583e-05,
+      "loss": 1.5321,
+      "step": 1047
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9993955018510044e-05,
+      "loss": 1.5742,
+      "step": 1048
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9993827099101315e-05,
+      "loss": 1.5586,
+      "step": 1049
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999369784077654e-05,
+      "loss": 1.5359,
+      "step": 1050
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9993567243553036e-05,
+      "loss": 1.5896,
+      "step": 1051
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9993435307448297e-05,
+      "loss": 1.6072,
+      "step": 1052
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9993302032479996e-05,
+      "loss": 1.6707,
+      "step": 1053
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9993167418665997e-05,
+      "loss": 1.5265,
+      "step": 1054
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999303146602433e-05,
+      "loss": 1.6136,
+      "step": 1055
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9992894174573206e-05,
+      "loss": 1.6654,
+      "step": 1056
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9992755544331022e-05,
+      "loss": 1.5838,
+      "step": 1057
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9992615575316355e-05,
+      "loss": 1.5751,
+      "step": 1058
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9992474267547948e-05,
+      "loss": 1.451,
+      "step": 1059
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999233162104474e-05,
+      "loss": 1.4902,
+      "step": 1060
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9992187635825837e-05,
+      "loss": 1.5157,
+      "step": 1061
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9992042311910534e-05,
+      "loss": 1.6304,
+      "step": 1062
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9991895649318297e-05,
+      "loss": 1.6828,
+      "step": 1063
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9991747648068776e-05,
+      "loss": 1.5534,
+      "step": 1064
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9991598308181802e-05,
+      "loss": 1.5296,
+      "step": 1065
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999144762967738e-05,
+      "loss": 1.5465,
+      "step": 1066
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9991295612575697e-05,
+      "loss": 1.5044,
+      "step": 1067
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999114225689712e-05,
+      "loss": 1.5644,
+      "step": 1068
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9990987562662194e-05,
+      "loss": 1.5296,
+      "step": 1069
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9990831529891645e-05,
+      "loss": 1.5701,
+      "step": 1070
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9990674158606378e-05,
+      "loss": 1.5579,
+      "step": 1071
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9990515448827477e-05,
+      "loss": 1.5989,
+      "step": 1072
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.99903554005762e-05,
+      "loss": 1.5712,
+      "step": 1073
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9990194013873997e-05,
+      "loss": 1.5244,
+      "step": 1074
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9990031288742485e-05,
+      "loss": 1.5237,
+      "step": 1075
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9989867225203465e-05,
+      "loss": 1.5279,
+      "step": 1076
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9989701823278917e-05,
+      "loss": 1.5552,
+      "step": 1077
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9989535082991002e-05,
+      "loss": 1.5221,
+      "step": 1078
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.998936700436206e-05,
+      "loss": 1.5706,
+      "step": 1079
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9989197587414605e-05,
+      "loss": 1.469,
+      "step": 1080
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.998902683217134e-05,
+      "loss": 1.4973,
+      "step": 1081
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.998885473865514e-05,
+      "loss": 1.6985,
+      "step": 1082
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9988681306889058e-05,
+      "loss": 1.5801,
+      "step": 1083
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9988506536896336e-05,
+      "loss": 1.4841,
+      "step": 1084
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.998833042870038e-05,
+      "loss": 1.5433,
+      "step": 1085
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9988152982324788e-05,
+      "loss": 1.736,
+      "step": 1086
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9987974197793338e-05,
+      "loss": 1.5823,
+      "step": 1087
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9987794075129975e-05,
+      "loss": 1.528,
+      "step": 1088
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9987612614358835e-05,
+      "loss": 1.5557,
+      "step": 1089
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.998742981550423e-05,
+      "loss": 1.5393,
+      "step": 1090
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9987245678590644e-05,
+      "loss": 1.538,
+      "step": 1091
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9987060203642754e-05,
+      "loss": 1.5764,
+      "step": 1092
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9986873390685405e-05,
+      "loss": 1.4883,
+      "step": 1093
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9986685239743625e-05,
+      "loss": 1.5683,
+      "step": 1094
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9986495750842624e-05,
+      "loss": 1.5445,
+      "step": 1095
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9986304924007788e-05,
+      "loss": 1.5576,
+      "step": 1096
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.998611275926468e-05,
+      "loss": 1.5164,
+      "step": 1097
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9985919256639046e-05,
+      "loss": 1.5223,
+      "step": 1098
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9985724416156814e-05,
+      "loss": 1.5501,
+      "step": 1099
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9985528237844084e-05,
+      "loss": 1.5639,
+      "step": 1100
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9985330721727144e-05,
+      "loss": 1.561,
+      "step": 1101
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9985131867832445e-05,
+      "loss": 1.4891,
+      "step": 1102
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.998493167618664e-05,
+      "loss": 1.5038,
+      "step": 1103
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9984730146816543e-05,
+      "loss": 1.5025,
+      "step": 1104
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9984527279749157e-05,
+      "loss": 1.5344,
+      "step": 1105
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.998432307501166e-05,
+      "loss": 1.5562,
+      "step": 1106
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9984117532631407e-05,
+      "loss": 1.5406,
+      "step": 1107
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.998391065263594e-05,
+      "loss": 1.5669,
+      "step": 1108
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9983702435052976e-05,
+      "loss": 1.569,
+      "step": 1109
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9983492879910407e-05,
+      "loss": 1.6172,
+      "step": 1110
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.998328198723631e-05,
+      "loss": 1.5511,
+      "step": 1111
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9983069757058936e-05,
+      "loss": 1.585,
+      "step": 1112
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9982856189406722e-05,
+      "loss": 1.588,
+      "step": 1113
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9982641284308282e-05,
+      "loss": 1.5751,
+      "step": 1114
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9982425041792404e-05,
+      "loss": 1.5794,
+      "step": 1115
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9982207461888065e-05,
+      "loss": 1.5905,
+      "step": 1116
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9981988544624404e-05,
+      "loss": 1.5579,
+      "step": 1117
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.998176829003076e-05,
+      "loss": 1.5642,
+      "step": 1118
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9981546698136638e-05,
+      "loss": 1.5751,
+      "step": 1119
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9981323768971727e-05,
+      "loss": 1.506,
+      "step": 1120
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.998109950256589e-05,
+      "loss": 1.4748,
+      "step": 1121
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.998087389894918e-05,
+      "loss": 1.5659,
+      "step": 1122
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9980646958151814e-05,
+      "loss": 1.6106,
+      "step": 1123
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.99804186802042e-05,
+      "loss": 1.58,
+      "step": 1124
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9980189065136924e-05,
+      "loss": 1.5812,
+      "step": 1125
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.997995811298074e-05,
+      "loss": 1.6222,
+      "step": 1126
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9979725823766602e-05,
+      "loss": 1.485,
+      "step": 1127
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9979492197525622e-05,
+      "loss": 1.5596,
+      "step": 1128
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9979257234289106e-05,
+      "loss": 1.6098,
+      "step": 1129
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9979020934088525e-05,
+      "loss": 1.5064,
+      "step": 1130
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.997878329695554e-05,
+      "loss": 1.5365,
+      "step": 1131
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9978544322921993e-05,
+      "loss": 1.4867,
+      "step": 1132
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9978304012019898e-05,
+      "loss": 1.5235,
+      "step": 1133
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.997806236428145e-05,
+      "loss": 1.5239,
+      "step": 1134
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.997781937973902e-05,
+      "loss": 1.4978,
+      "step": 1135
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9977575058425166e-05,
+      "loss": 1.5116,
+      "step": 1136
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9977329400372622e-05,
+      "loss": 1.5382,
+      "step": 1137
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9977082405614293e-05,
+      "loss": 1.5398,
+      "step": 1138
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9976834074183276e-05,
+      "loss": 1.6178,
+      "step": 1139
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.997658440611284e-05,
+      "loss": 1.5242,
+      "step": 1140
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9976333401436433e-05,
+      "loss": 1.5686,
+      "step": 1141
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9976081060187686e-05,
+      "loss": 1.5536,
+      "step": 1142
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9975827382400403e-05,
+      "loss": 1.5566,
+      "step": 1143
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.997557236810857e-05,
+      "loss": 1.5506,
+      "step": 1144
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9975316017346356e-05,
+      "loss": 1.5996,
+      "step": 1145
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.99750583301481e-05,
+      "loss": 1.5492,
+      "step": 1146
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9974799306548333e-05,
+      "loss": 1.5857,
+      "step": 1147
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.997453894658175e-05,
+      "loss": 1.6161,
+      "step": 1148
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9974277250283237e-05,
+      "loss": 1.6501,
+      "step": 1149
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.997401421768785e-05,
+      "loss": 1.7257,
+      "step": 1150
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9973749848830835e-05,
+      "loss": 1.4917,
+      "step": 1151
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9973484143747604e-05,
+      "loss": 1.5374,
+      "step": 1152
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9973217102473765e-05,
+      "loss": 1.567,
+      "step": 1153
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9972948725045083e-05,
+      "loss": 1.5835,
+      "step": 1154
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.997267901149752e-05,
+      "loss": 1.5808,
+      "step": 1155
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9972407961867208e-05,
+      "loss": 1.6002,
+      "step": 1156
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.997213557619046e-05,
+      "loss": 1.534,
+      "step": 1157
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9971861854503772e-05,
+      "loss": 1.5806,
+      "step": 1158
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9971586796843815e-05,
+      "loss": 1.5429,
+      "step": 1159
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.997131040324744e-05,
+      "loss": 1.5448,
+      "step": 1160
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9971032673751673e-05,
+      "loss": 1.5598,
+      "step": 1161
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9970753608393727e-05,
+      "loss": 1.5355,
+      "step": 1162
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.997047320721099e-05,
+      "loss": 1.5599,
+      "step": 1163
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.997019147024102e-05,
+      "loss": 1.5372,
+      "step": 1164
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.996990839752157e-05,
+      "loss": 1.5287,
+      "step": 1165
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9969623989090566e-05,
+      "loss": 1.6099,
+      "step": 1166
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.996933824498611e-05,
+      "loss": 1.5386,
+      "step": 1167
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.996905116524648e-05,
+      "loss": 1.5641,
+      "step": 1168
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.996876274991014e-05,
+      "loss": 1.5728,
+      "step": 1169
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9968472999015734e-05,
+      "loss": 1.4114,
+      "step": 1170
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9968181912602074e-05,
+      "loss": 1.5279,
+      "step": 1171
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9967889490708164e-05,
+      "loss": 1.5374,
+      "step": 1172
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.996759573337318e-05,
+      "loss": 1.699,
+      "step": 1173
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9967300640636474e-05,
+      "loss": 1.6061,
+      "step": 1174
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9967004212537583e-05,
+      "loss": 1.4681,
+      "step": 1175
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9966706449116225e-05,
+      "loss": 1.5457,
+      "step": 1176
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9966407350412284e-05,
+      "loss": 1.5293,
+      "step": 1177
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9966106916465842e-05,
+      "loss": 1.5011,
+      "step": 1178
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.996580514731714e-05,
+      "loss": 1.534,
+      "step": 1179
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9965502043006613e-05,
+      "loss": 1.4692,
+      "step": 1180
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9965197603574867e-05,
+      "loss": 1.4711,
+      "step": 1181
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.996489182906269e-05,
+      "loss": 1.4407,
+      "step": 1182
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9964584719511048e-05,
+      "loss": 1.5076,
+      "step": 1183
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9964276274961086e-05,
+      "loss": 1.5695,
+      "step": 1184
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9963966495454126e-05,
+      "loss": 1.4995,
+      "step": 1185
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.996365538103167e-05,
+      "loss": 1.5609,
+      "step": 1186
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9963342931735402e-05,
+      "loss": 1.5052,
+      "step": 1187
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.996302914760718e-05,
+      "loss": 1.4803,
+      "step": 1188
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9962714028689046e-05,
+      "loss": 1.5778,
+      "step": 1189
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9962397575023216e-05,
+      "loss": 1.5188,
+      "step": 1190
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9962079786652084e-05,
+      "loss": 1.5627,
+      "step": 1191
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9961760663618232e-05,
+      "loss": 1.5204,
+      "step": 1192
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9961440205964404e-05,
+      "loss": 1.6796,
+      "step": 1193
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.996111841373354e-05,
+      "loss": 1.5158,
+      "step": 1194
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 1.5832440853118896,
+      "eval_runtime": 77.2765,
+      "eval_samples_per_second": 1.669,
+      "eval_steps_per_second": 0.116,
+      "step": 1194
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.996079528696876e-05,
+      "loss": 1.467,
+      "step": 1195
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.996047082571334e-05,
+      "loss": 1.5276,
+      "step": 1196
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9960145030010752e-05,
+      "loss": 1.4959,
+      "step": 1197
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9959817899904654e-05,
+      "loss": 1.5513,
+      "step": 1198
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9959489435438863e-05,
+      "loss": 1.6125,
+      "step": 1199
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9959159636657385e-05,
+      "loss": 1.5535,
+      "step": 1200
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9958828503604414e-05,
+      "loss": 1.5446,
+      "step": 1201
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9958496036324308e-05,
+      "loss": 1.4983,
+      "step": 1202
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9958162234861603e-05,
+      "loss": 1.5009,
+      "step": 1203
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9957827099261027e-05,
+      "loss": 1.5564,
+      "step": 1204
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.995749062956748e-05,
+      "loss": 1.4037,
+      "step": 1205
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9957152825826035e-05,
+      "loss": 1.607,
+      "step": 1206
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9956813688081954e-05,
+      "loss": 1.5404,
+      "step": 1207
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.995647321638067e-05,
+      "loss": 1.5485,
+      "step": 1208
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.99561314107678e-05,
+      "loss": 1.5897,
+      "step": 1209
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9955788271289132e-05,
+      "loss": 1.5297,
+      "step": 1210
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9955443797990647e-05,
+      "loss": 1.675,
+      "step": 1211
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9955097990918485e-05,
+      "loss": 1.5825,
+      "step": 1212
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9954750850118984e-05,
+      "loss": 1.553,
+      "step": 1213
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9954402375638643e-05,
+      "loss": 1.5298,
+      "step": 1214
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9954052567524158e-05,
+      "loss": 1.6591,
+      "step": 1215
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.995370142582239e-05,
+      "loss": 1.5252,
+      "step": 1216
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.995334895058038e-05,
+      "loss": 1.5124,
+      "step": 1217
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9952995141845357e-05,
+      "loss": 1.5119,
+      "step": 1218
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9952639999664716e-05,
+      "loss": 1.4908,
+      "step": 1219
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9952283524086043e-05,
+      "loss": 1.4798,
+      "step": 1220
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.995192571515709e-05,
+      "loss": 1.5116,
+      "step": 1221
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.99515665729258e-05,
+      "loss": 1.5127,
+      "step": 1222
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9951206097440284e-05,
+      "loss": 1.5158,
+      "step": 1223
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9950844288748842e-05,
+      "loss": 1.5894,
+      "step": 1224
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.995048114689994e-05,
+      "loss": 1.6717,
+      "step": 1225
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.995011667194224e-05,
+      "loss": 1.7595,
+      "step": 1226
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.994975086392456e-05,
+      "loss": 1.5495,
+      "step": 1227
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9949383722895918e-05,
+      "loss": 1.5844,
+      "step": 1228
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9949015248905493e-05,
+      "loss": 1.6022,
+      "step": 1229
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9948645442002663e-05,
+      "loss": 1.5495,
+      "step": 1230
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9948274302236964e-05,
+      "loss": 1.5465,
+      "step": 1231
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.994790182965812e-05,
+      "loss": 1.553,
+      "step": 1232
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9947528024316034e-05,
+      "loss": 1.5895,
+      "step": 1233
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9947152886260784e-05,
+      "loss": 1.4625,
+      "step": 1234
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9946776415542635e-05,
+      "loss": 1.5255,
+      "step": 1235
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.994639861221202e-05,
+      "loss": 1.5293,
+      "step": 1236
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9946019476319556e-05,
+      "loss": 1.6018,
+      "step": 1237
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9945639007916035e-05,
+      "loss": 1.53,
+      "step": 1238
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.994525720705243e-05,
+      "loss": 1.4906,
+      "step": 1239
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.99448740737799e-05,
+      "loss": 1.5203,
+      "step": 1240
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9944489608149763e-05,
+      "loss": 1.532,
+      "step": 1241
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.994410381021354e-05,
+      "loss": 1.5899,
+      "step": 1242
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9943716680022906e-05,
+      "loss": 1.5295,
+      "step": 1243
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9943328217629736e-05,
+      "loss": 1.5269,
+      "step": 1244
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9942938423086068e-05,
+      "loss": 1.5067,
+      "step": 1245
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.994254729644413e-05,
+      "loss": 1.557,
+      "step": 1246
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9942154837756318e-05,
+      "loss": 1.5343,
+      "step": 1247
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9941761047075214e-05,
+      "loss": 1.5815,
+      "step": 1248
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9941365924453574e-05,
+      "loss": 1.5704,
+      "step": 1249
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9940969469944334e-05,
+      "loss": 1.5795,
+      "step": 1250
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9940571683600614e-05,
+      "loss": 1.4563,
+      "step": 1251
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.99401725654757e-05,
+      "loss": 1.5214,
+      "step": 1252
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9939772115623065e-05,
+      "loss": 1.4633,
+      "step": 1253
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9939370334096363e-05,
+      "loss": 1.5729,
+      "step": 1254
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.993896722094942e-05,
+      "loss": 1.5483,
+      "step": 1255
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9938562776236245e-05,
+      "loss": 1.557,
+      "step": 1256
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.993815700001102e-05,
+      "loss": 1.5357,
+      "step": 1257
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9937749892328107e-05,
+      "loss": 1.5176,
+      "step": 1258
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9937341453242053e-05,
+      "loss": 1.4549,
+      "step": 1259
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9936931682807574e-05,
+      "loss": 1.5058,
+      "step": 1260
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9936520581079573e-05,
+      "loss": 1.5079,
+      "step": 1261
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9936108148113122e-05,
+      "loss": 1.4598,
+      "step": 1262
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.993569438396348e-05,
+      "loss": 1.5505,
+      "step": 1263
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9935279288686082e-05,
+      "loss": 1.5665,
+      "step": 1264
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9934862862336536e-05,
+      "loss": 1.4822,
+      "step": 1265
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9934445104970638e-05,
+      "loss": 1.4724,
+      "step": 1266
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.993402601664435e-05,
+      "loss": 1.4477,
+      "step": 1267
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9933605597413824e-05,
+      "loss": 1.4769,
+      "step": 1268
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9933183847335383e-05,
+      "loss": 1.5726,
+      "step": 1269
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9932760766465532e-05,
+      "loss": 1.4443,
+      "step": 1270
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.993233635486095e-05,
+      "loss": 1.6652,
+      "step": 1271
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.99319106125785e-05,
+      "loss": 1.7169,
+      "step": 1272
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9931483539675222e-05,
+      "loss": 1.4327,
+      "step": 1273
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.993105513620833e-05,
+      "loss": 1.5174,
+      "step": 1274
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.993062540223522e-05,
+      "loss": 1.5318,
+      "step": 1275
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9930194337813464e-05,
+      "loss": 1.4658,
+      "step": 1276
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9929761943000815e-05,
+      "loss": 1.5312,
+      "step": 1277
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9929328217855203e-05,
+      "loss": 1.5339,
+      "step": 1278
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9928893162434738e-05,
+      "loss": 1.5304,
+      "step": 1279
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.99284567767977e-05,
+      "loss": 1.5715,
+      "step": 1280
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.992801906100256e-05,
+      "loss": 1.5644,
+      "step": 1281
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9927580015107958e-05,
+      "loss": 1.5434,
+      "step": 1282
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9927139639172713e-05,
+      "loss": 1.7514,
+      "step": 1283
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9926697933255828e-05,
+      "loss": 1.5085,
+      "step": 1284
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9926254897416476e-05,
+      "loss": 1.5304,
+      "step": 1285
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9925810531714015e-05,
+      "loss": 1.557,
+      "step": 1286
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.992536483620798e-05,
+      "loss": 1.5159,
+      "step": 1287
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.992491781095808e-05,
+      "loss": 1.5138,
+      "step": 1288
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9924469456024206e-05,
+      "loss": 1.5264,
+      "step": 1289
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9924019771466427e-05,
+      "loss": 1.5213,
+      "step": 1290
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9923568757344985e-05,
+      "loss": 1.6664,
+      "step": 1291
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.992311641372031e-05,
+      "loss": 1.5334,
+      "step": 1292
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9922662740653003e-05,
+      "loss": 1.5541,
+      "step": 1293
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.992220773820384e-05,
+      "loss": 1.4788,
+      "step": 1294
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9921751406433783e-05,
+      "loss": 1.5413,
+      "step": 1295
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9921293745403972e-05,
+      "loss": 1.6117,
+      "step": 1296
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9920834755175714e-05,
+      "loss": 1.5822,
+      "step": 1297
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.992037443581051e-05,
+      "loss": 1.7076,
+      "step": 1298
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9919912787370025e-05,
+      "loss": 1.5648,
+      "step": 1299
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9919449809916114e-05,
+      "loss": 1.5338,
+      "step": 1300
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9918985503510796e-05,
+      "loss": 1.5155,
+      "step": 1301
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9918519868216284e-05,
+      "loss": 1.5157,
+      "step": 1302
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9918052904094955e-05,
+      "loss": 1.5318,
+      "step": 1303
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9917584611209376e-05,
+      "loss": 1.6008,
+      "step": 1304
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.991711498962228e-05,
+      "loss": 1.4786,
+      "step": 1305
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9916644039396592e-05,
+      "loss": 1.5199,
+      "step": 1306
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9916171760595398e-05,
+      "loss": 1.5193,
+      "step": 1307
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9915698153281977e-05,
+      "loss": 1.5439,
+      "step": 1308
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.991522321751978e-05,
+      "loss": 1.5282,
+      "step": 1309
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9914746953372435e-05,
+      "loss": 1.477,
+      "step": 1310
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.991426936090375e-05,
+      "loss": 1.5301,
+      "step": 1311
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9913790440177708e-05,
+      "loss": 1.4783,
+      "step": 1312
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9913310191258476e-05,
+      "loss": 1.5308,
+      "step": 1313
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.991282861421039e-05,
+      "loss": 1.5223,
+      "step": 1314
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9912345709097974e-05,
+      "loss": 1.5472,
+      "step": 1315
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9911861475985918e-05,
+      "loss": 1.5483,
+      "step": 1316
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9911375914939106e-05,
+      "loss": 1.5503,
+      "step": 1317
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9910889026022583e-05,
+      "loss": 1.684,
+      "step": 1318
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9910400809301585e-05,
+      "loss": 1.5614,
+      "step": 1319
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9909911264841514e-05,
+      "loss": 1.5388,
+      "step": 1320
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9909420392707965e-05,
+      "loss": 1.5537,
+      "step": 1321
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9908928192966697e-05,
+      "loss": 1.4993,
+      "step": 1322
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.990843466568365e-05,
+      "loss": 1.6789,
+      "step": 1323
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.990793981092495e-05,
+      "loss": 1.5477,
+      "step": 1324
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.990744362875689e-05,
+      "loss": 1.5408,
+      "step": 1325
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9906946119245945e-05,
+      "loss": 1.5897,
+      "step": 1326
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9906447282458777e-05,
+      "loss": 1.5161,
+      "step": 1327
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9905947118462207e-05,
+      "loss": 1.5298,
+      "step": 1328
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9905445627323247e-05,
+      "loss": 1.6533,
+      "step": 1329
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9904942809109088e-05,
+      "loss": 1.5175,
+      "step": 1330
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.990443866388709e-05,
+      "loss": 1.5637,
+      "step": 1331
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9903933191724798e-05,
+      "loss": 1.6731,
+      "step": 1332
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.990342639268993e-05,
+      "loss": 1.5462,
+      "step": 1333
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.990291826685039e-05,
+      "loss": 1.5438,
+      "step": 1334
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9902408814274246e-05,
+      "loss": 1.6058,
+      "step": 1335
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9901898035029753e-05,
+      "loss": 1.5122,
+      "step": 1336
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.990138592918535e-05,
+      "loss": 1.5462,
+      "step": 1337
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9900872496809636e-05,
+      "loss": 1.5612,
+      "step": 1338
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.99003577379714e-05,
+      "loss": 1.5255,
+      "step": 1339
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9899841652739616e-05,
+      "loss": 1.5259,
+      "step": 1340
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9899324241183413e-05,
+      "loss": 1.458,
+      "step": 1341
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9898805503372118e-05,
+      "loss": 1.4987,
+      "step": 1342
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.989828543937523e-05,
+      "loss": 1.5304,
+      "step": 1343
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9897764049262416e-05,
+      "loss": 1.5741,
+      "step": 1344
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.989724133310354e-05,
+      "loss": 1.5537,
+      "step": 1345
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.989671729096862e-05,
+      "loss": 1.5428,
+      "step": 1346
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9896191922927875e-05,
+      "loss": 1.4843,
+      "step": 1347
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9895665229051686e-05,
+      "loss": 1.5322,
+      "step": 1348
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9895137209410617e-05,
+      "loss": 1.5133,
+      "step": 1349
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9894607864075413e-05,
+      "loss": 1.5155,
+      "step": 1350
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9894077193116983e-05,
+      "loss": 1.5303,
+      "step": 1351
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9893545196606432e-05,
+      "loss": 1.549,
+      "step": 1352
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.989301187461503e-05,
+      "loss": 1.4972,
+      "step": 1353
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.989247722721423e-05,
+      "loss": 1.444,
+      "step": 1354
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.989194125447566e-05,
+      "loss": 1.6561,
+      "step": 1355
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.989140395647113e-05,
+      "loss": 1.7125,
+      "step": 1356
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.989086533327262e-05,
+      "loss": 1.5723,
+      "step": 1357
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9890325384952293e-05,
+      "loss": 1.4852,
+      "step": 1358
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.988978411158249e-05,
+      "loss": 1.591,
+      "step": 1359
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.988924151323572e-05,
+      "loss": 1.5385,
+      "step": 1360
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9888697589984685e-05,
+      "loss": 1.577,
+      "step": 1361
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.988815234190226e-05,
+      "loss": 1.4827,
+      "step": 1362
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9887605769061484e-05,
+      "loss": 1.5193,
+      "step": 1363
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9887057871535588e-05,
+      "loss": 1.5391,
+      "step": 1364
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.988650864939798e-05,
+      "loss": 1.7037,
+      "step": 1365
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9885958102722238e-05,
+      "loss": 1.4816,
+      "step": 1366
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.988540623158212e-05,
+      "loss": 1.4663,
+      "step": 1367
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9884853036051567e-05,
+      "loss": 1.5254,
+      "step": 1368
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9884298516204687e-05,
+      "loss": 1.5087,
+      "step": 1369
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9883742672115777e-05,
+      "loss": 1.5178,
+      "step": 1370
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9883185503859302e-05,
+      "loss": 1.473,
+      "step": 1371
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.988262701150991e-05,
+      "loss": 1.5245,
+      "step": 1372
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9882067195142426e-05,
+      "loss": 1.3492,
+      "step": 1373
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9881506054831847e-05,
+      "loss": 1.5338,
+      "step": 1374
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9880943590653356e-05,
+      "loss": 1.5214,
+      "step": 1375
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9880379802682305e-05,
+      "loss": 1.4902,
+      "step": 1376
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9879814690994226e-05,
+      "loss": 1.6882,
+      "step": 1377
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9879248255664836e-05,
+      "loss": 1.4952,
+      "step": 1378
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9878680496770016e-05,
+      "loss": 1.5286,
+      "step": 1379
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9878111414385837e-05,
+      "loss": 1.5065,
+      "step": 1380
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9877541008588537e-05,
+      "loss": 1.5064,
+      "step": 1381
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9876969279454538e-05,
+      "loss": 1.5136,
+      "step": 1382
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9876396227060437e-05,
+      "loss": 1.5103,
+      "step": 1383
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9875821851483002e-05,
+      "loss": 1.4873,
+      "step": 1384
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9875246152799192e-05,
+      "loss": 1.5025,
+      "step": 1385
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9874669131086138e-05,
+      "loss": 1.5064,
+      "step": 1386
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.987409078642114e-05,
+      "loss": 1.5356,
+      "step": 1387
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.987351111888168e-05,
+      "loss": 1.7061,
+      "step": 1388
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.987293012854542e-05,
+      "loss": 1.5003,
+      "step": 1389
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9872347815490207e-05,
+      "loss": 1.541,
+      "step": 1390
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9871764179794046e-05,
+      "loss": 1.5038,
+      "step": 1391
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9871179221535128e-05,
+      "loss": 1.502,
+      "step": 1392
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9870592940791828e-05,
+      "loss": 1.5181,
+      "step": 1393
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 1.5540661811828613,
+      "eval_runtime": 77.2268,
+      "eval_samples_per_second": 1.67,
+      "eval_steps_per_second": 0.117,
+      "step": 1393
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.987000533764269e-05,
+      "loss": 1.5556,
+      "step": 1394
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9869416412166443e-05,
+      "loss": 1.5699,
+      "step": 1395
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9868826164441977e-05,
+      "loss": 1.5438,
+      "step": 1396
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.986823459454838e-05,
+      "loss": 1.6003,
+      "step": 1397
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.98676417025649e-05,
+      "loss": 1.5009,
+      "step": 1398
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.986704748857098e-05,
+      "loss": 1.5066,
+      "step": 1399
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9866451952646212e-05,
+      "loss": 1.4497,
+      "step": 1400
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.98658550948704e-05,
+      "loss": 1.5807,
+      "step": 1401
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9865256915323495e-05,
+      "loss": 1.5399,
+      "step": 1402
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9864657414085648e-05,
+      "loss": 1.5325,
+      "step": 1403
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.986405659123717e-05,
+      "loss": 1.6155,
+      "step": 1404
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9863454446858558e-05,
+      "loss": 1.5468,
+      "step": 1405
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9862850981030482e-05,
+      "loss": 1.4848,
+      "step": 1406
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9862246193833793e-05,
+      "loss": 1.2032,
+      "step": 1407
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9861640085349516e-05,
+      "loss": 1.4645,
+      "step": 1408
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9861032655658852e-05,
+      "loss": 1.4402,
+      "step": 1409
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9860423904843187e-05,
+      "loss": 1.5101,
+      "step": 1410
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.985981383298407e-05,
+      "loss": 1.5337,
+      "step": 1411
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9859202440163243e-05,
+      "loss": 1.5493,
+      "step": 1412
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.985858972646261e-05,
+      "loss": 1.5393,
+      "step": 1413
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9857975691964264e-05,
+      "loss": 1.5077,
+      "step": 1414
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9857360336750465e-05,
+      "loss": 1.5027,
+      "step": 1415
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.985674366090366e-05,
+      "loss": 1.5995,
+      "step": 1416
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9856125664506468e-05,
+      "loss": 1.6216,
+      "step": 1417
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9855506347641678e-05,
+      "loss": 1.6285,
+      "step": 1418
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.985488571039227e-05,
+      "loss": 1.5688,
+      "step": 1419
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9854263752841385e-05,
+      "loss": 1.4989,
+      "step": 1420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.985364047507236e-05,
+      "loss": 1.4919,
+      "step": 1421
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9853015877168684e-05,
+      "loss": 1.5362,
+      "step": 1422
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.985238995921405e-05,
+      "loss": 1.5068,
+      "step": 1423
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9851762721292314e-05,
+      "loss": 1.5215,
+      "step": 1424
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.98511341634875e-05,
+      "loss": 1.6322,
+      "step": 1425
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9850504285883828e-05,
+      "loss": 1.589,
+      "step": 1426
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9849873088565684e-05,
+      "loss": 1.5862,
+      "step": 1427
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.984924057161763e-05,
+      "loss": 1.5645,
+      "step": 1428
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9848606735124402e-05,
+      "loss": 1.4309,
+      "step": 1429
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.984797157917093e-05,
+      "loss": 1.5717,
+      "step": 1430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9847335103842297e-05,
+      "loss": 1.5764,
+      "step": 1431
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.984669730922378e-05,
+      "loss": 1.4737,
+      "step": 1432
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.984605819540083e-05,
+      "loss": 1.5157,
+      "step": 1433
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9845417762459064e-05,
+      "loss": 1.5413,
+      "step": 1434
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.984477601048429e-05,
+      "loss": 1.5784,
+      "step": 1435
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.984413293956248e-05,
+      "loss": 1.4951,
+      "step": 1436
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.98434885497798e-05,
+      "loss": 1.4805,
+      "step": 1437
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9842842841222566e-05,
+      "loss": 1.5163,
+      "step": 1438
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.98421958139773e-05,
+      "loss": 1.5094,
+      "step": 1439
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9841547468130683e-05,
+      "loss": 1.511,
+      "step": 1440
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9840897803769574e-05,
+      "loss": 1.621,
+      "step": 1441
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9840246820981012e-05,
+      "loss": 1.6741,
+      "step": 1442
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9839594519852213e-05,
+      "loss": 1.5389,
+      "step": 1443
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9838940900470572e-05,
+      "loss": 1.6525,
+      "step": 1444
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9838285962923648e-05,
+      "loss": 1.5325,
+      "step": 1445
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9837629707299196e-05,
+      "loss": 1.4859,
+      "step": 1446
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9836972133685133e-05,
+      "loss": 1.4992,
+      "step": 1447
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9836313242169556e-05,
+      "loss": 1.4715,
+      "step": 1448
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.983565303284074e-05,
+      "loss": 1.7241,
+      "step": 1449
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9834991505787135e-05,
+      "loss": 1.6196,
+      "step": 1450
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.983432866109737e-05,
+      "loss": 1.5365,
+      "step": 1451
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9833664498860252e-05,
+      "loss": 1.4572,
+      "step": 1452
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.983299901916476e-05,
+      "loss": 1.5122,
+      "step": 1453
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9832332222100046e-05,
+      "loss": 1.5531,
+      "step": 1454
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.983166410775545e-05,
+      "loss": 1.5377,
+      "step": 1455
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.983099467622048e-05,
+      "loss": 1.5294,
+      "step": 1456
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9830323927584824e-05,
+      "loss": 1.5397,
+      "step": 1457
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9829651861938342e-05,
+      "loss": 1.5659,
+      "step": 1458
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9828978479371076e-05,
+      "loss": 1.5035,
+      "step": 1459
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.982830377997324e-05,
+      "loss": 1.3996,
+      "step": 1460
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9827627763835227e-05,
+      "loss": 1.4924,
+      "step": 1461
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9826950431047608e-05,
+      "loss": 1.5412,
+      "step": 1462
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9826271781701128e-05,
+      "loss": 1.5036,
+      "step": 1463
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9825591815886703e-05,
+      "loss": 1.4907,
+      "step": 1464
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9824910533695437e-05,
+      "loss": 1.4982,
+      "step": 1465
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9824227935218604e-05,
+      "loss": 1.466,
+      "step": 1466
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9823544020547652e-05,
+      "loss": 1.557,
+      "step": 1467
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.982285878977421e-05,
+      "loss": 1.5511,
+      "step": 1468
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9822172242990078e-05,
+      "loss": 1.5287,
+      "step": 1469
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.982148438028724e-05,
+      "loss": 1.532,
+      "step": 1470
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.982079520175785e-05,
+      "loss": 1.5558,
+      "step": 1471
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9820104707494238e-05,
+      "loss": 1.4756,
+      "step": 1472
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9819412897588913e-05,
+      "loss": 1.4943,
+      "step": 1473
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9818719772134566e-05,
+      "loss": 1.5284,
+      "step": 1474
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9818025331224052e-05,
+      "loss": 1.3625,
+      "step": 1475
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9817329574950406e-05,
+      "loss": 1.5042,
+      "step": 1476
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.981663250340685e-05,
+      "loss": 1.5595,
+      "step": 1477
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.981593411668677e-05,
+      "loss": 1.5317,
+      "step": 1478
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9815234414883723e-05,
+      "loss": 1.6197,
+      "step": 1479
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9814533398091463e-05,
+      "loss": 1.5507,
+      "step": 1480
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9813831066403905e-05,
+      "loss": 1.6891,
+      "step": 1481
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9813127419915143e-05,
+      "loss": 1.5163,
+      "step": 1482
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9812422458719448e-05,
+      "loss": 1.5262,
+      "step": 1483
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9811716182911262e-05,
+      "loss": 1.5741,
+      "step": 1484
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.981100859258521e-05,
+      "loss": 1.6361,
+      "step": 1485
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9810299687836093e-05,
+      "loss": 1.4907,
+      "step": 1486
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9809589468758886e-05,
+      "loss": 1.5192,
+      "step": 1487
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9808877935448747e-05,
+      "loss": 1.505,
+      "step": 1488
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9808165088000986e-05,
+      "loss": 1.4637,
+      "step": 1489
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.980745092651112e-05,
+      "loss": 1.4928,
+      "step": 1490
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.980673545107482e-05,
+      "loss": 1.7325,
+      "step": 1491
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.980601866178795e-05,
+      "loss": 1.5425,
+      "step": 1492
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9805300558746535e-05,
+      "loss": 1.5345,
+      "step": 1493
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9804581142046783e-05,
+      "loss": 1.5256,
+      "step": 1494
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.980386041178508e-05,
+      "loss": 1.4853,
+      "step": 1495
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9803138368057985e-05,
+      "loss": 1.4834,
+      "step": 1496
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.980241501096223e-05,
+      "loss": 1.504,
+      "step": 1497
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.980169034059473e-05,
+      "loss": 1.5379,
+      "step": 1498
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.980096435705257e-05,
+      "loss": 1.5264,
+      "step": 1499
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9800237060433014e-05,
+      "loss": 1.52,
+      "step": 1500
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.97995084508335e-05,
+      "loss": 1.5411,
+      "step": 1501
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9798778528351645e-05,
+      "loss": 1.509,
+      "step": 1502
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9798047293085236e-05,
+      "loss": 1.4869,
+      "step": 1503
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9797314745132247e-05,
+      "loss": 1.4347,
+      "step": 1504
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9796580884590813e-05,
+      "loss": 1.6105,
+      "step": 1505
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9795845711559254e-05,
+      "loss": 1.5335,
+      "step": 1506
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.979510922613607e-05,
+      "loss": 1.6023,
+      "step": 1507
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9794371428419922e-05,
+      "loss": 1.4844,
+      "step": 1508
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9793632318509664e-05,
+      "loss": 1.5876,
+      "step": 1509
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9792891896504313e-05,
+      "loss": 1.5566,
+      "step": 1510
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.979215016250307e-05,
+      "loss": 1.4877,
+      "step": 1511
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9791407116605304e-05,
+      "loss": 1.6041,
+      "step": 1512
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.979066275891057e-05,
+      "loss": 1.6285,
+      "step": 1513
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9789917089518584e-05,
+      "loss": 1.4881,
+      "step": 1514
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9789170108529256e-05,
+      "loss": 1.5249,
+      "step": 1515
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9788421816042657e-05,
+      "loss": 1.4931,
+      "step": 1516
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.978767221215904e-05,
+      "loss": 1.5066,
+      "step": 1517
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9786921296978834e-05,
+      "loss": 1.4409,
+      "step": 1518
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9786169070602638e-05,
+      "loss": 1.4639,
+      "step": 1519
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9785415533131237e-05,
+      "loss": 1.4504,
+      "step": 1520
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.978466068466558e-05,
+      "loss": 1.5162,
+      "step": 1521
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.97839045253068e-05,
+      "loss": 1.5009,
+      "step": 1522
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9783147055156203e-05,
+      "loss": 1.5544,
+      "step": 1523
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9782388274315272e-05,
+      "loss": 1.439,
+      "step": 1524
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9781628182885662e-05,
+      "loss": 1.5164,
+      "step": 1525
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9780866780969208e-05,
+      "loss": 1.5664,
+      "step": 1526
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9780104068667912e-05,
+      "loss": 1.5479,
+      "step": 1527
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9779340046083968e-05,
+      "loss": 1.588,
+      "step": 1528
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9778574713319726e-05,
+      "loss": 1.5177,
+      "step": 1529
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9777808070477726e-05,
+      "loss": 1.5231,
+      "step": 1530
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9777040117660676e-05,
+      "loss": 1.4482,
+      "step": 1531
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9776270854971466e-05,
+      "loss": 1.6153,
+      "step": 1532
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9775500282513153e-05,
+      "loss": 1.4942,
+      "step": 1533
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9774728400388973e-05,
+      "loss": 1.5073,
+      "step": 1534
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9773955208702346e-05,
+      "loss": 1.5183,
+      "step": 1535
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.977318070755685e-05,
+      "loss": 1.4683,
+      "step": 1536
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9772404897056255e-05,
+      "loss": 1.508,
+      "step": 1537
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.97716277773045e-05,
+      "loss": 1.4162,
+      "step": 1538
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9770849348405693e-05,
+      "loss": 1.7016,
+      "step": 1539
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9770069610464127e-05,
+      "loss": 1.563,
+      "step": 1540
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.976928856358427e-05,
+      "loss": 1.4684,
+      "step": 1541
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9768506207870754e-05,
+      "loss": 1.4793,
+      "step": 1542
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9767722543428408e-05,
+      "loss": 1.4873,
+      "step": 1543
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9766937570362205e-05,
+      "loss": 1.5214,
+      "step": 1544
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.976615128877733e-05,
+      "loss": 1.432,
+      "step": 1545
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.976536369877911e-05,
+      "loss": 1.5505,
+      "step": 1546
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.976457480047307e-05,
+      "loss": 1.4637,
+      "step": 1547
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9763784593964895e-05,
+      "loss": 1.5864,
+      "step": 1548
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9762993079360463e-05,
+      "loss": 1.4304,
+      "step": 1549
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9762200256765805e-05,
+      "loss": 1.5836,
+      "step": 1550
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9761406126287147e-05,
+      "loss": 1.5544,
+      "step": 1551
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9760610688030876e-05,
+      "loss": 1.4907,
+      "step": 1552
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.975981394210357e-05,
+      "loss": 1.5456,
+      "step": 1553
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9759015888611958e-05,
+      "loss": 1.5319,
+      "step": 1554
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.975821652766297e-05,
+      "loss": 1.5175,
+      "step": 1555
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.97574158593637e-05,
+      "loss": 1.5868,
+      "step": 1556
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.975661388382141e-05,
+      "loss": 1.4747,
+      "step": 1557
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9755810601143547e-05,
+      "loss": 1.4935,
+      "step": 1558
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9755006011437733e-05,
+      "loss": 1.5117,
+      "step": 1559
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.975420011481176e-05,
+      "loss": 1.5053,
+      "step": 1560
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9753392911373598e-05,
+      "loss": 1.5281,
+      "step": 1561
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.975258440123139e-05,
+      "loss": 1.6066,
+      "step": 1562
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9751774584493457e-05,
+      "loss": 1.47,
+      "step": 1563
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.975096346126829e-05,
+      "loss": 1.4469,
+      "step": 1564
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9750151031664567e-05,
+      "loss": 1.4246,
+      "step": 1565
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9749337295791127e-05,
+      "loss": 1.484,
+      "step": 1566
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9748522253756993e-05,
+      "loss": 1.5078,
+      "step": 1567
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9747705905671352e-05,
+      "loss": 1.4212,
+      "step": 1568
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9746888251643585e-05,
+      "loss": 1.5456,
+      "step": 1569
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9746069291783222e-05,
+      "loss": 1.4714,
+      "step": 1570
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.97452490262e-05,
+      "loss": 1.6635,
+      "step": 1571
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9744427455003796e-05,
+      "loss": 1.5673,
+      "step": 1572
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.974360457830469e-05,
+      "loss": 1.509,
+      "step": 1573
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.974278039621293e-05,
+      "loss": 1.5395,
+      "step": 1574
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9741954908838925e-05,
+      "loss": 1.5275,
+      "step": 1575
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9741128116293273e-05,
+      "loss": 1.5819,
+      "step": 1576
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9740300018686748e-05,
+      "loss": 1.4704,
+      "step": 1577
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9739470616130286e-05,
+      "loss": 1.4483,
+      "step": 1578
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.973863990873501e-05,
+      "loss": 1.4504,
+      "step": 1579
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.973780789661221e-05,
+      "loss": 1.4911,
+      "step": 1580
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9736974579873358e-05,
+      "loss": 1.5782,
+      "step": 1581
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.97361399586301e-05,
+      "loss": 1.5845,
+      "step": 1582
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9735304032994246e-05,
+      "loss": 1.4845,
+      "step": 1583
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.973446680307779e-05,
+      "loss": 1.4675,
+      "step": 1584
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9733628268992904e-05,
+      "loss": 1.4783,
+      "step": 1585
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9732788430851927e-05,
+      "loss": 1.4978,
+      "step": 1586
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9731947288767377e-05,
+      "loss": 1.4674,
+      "step": 1587
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.973110484285194e-05,
+      "loss": 1.4698,
+      "step": 1588
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9730261093218493e-05,
+      "loss": 1.5383,
+      "step": 1589
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.972941603998007e-05,
+      "loss": 1.5175,
+      "step": 1590
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9728569683249882e-05,
+      "loss": 1.4905,
+      "step": 1591
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9727722023141327e-05,
+      "loss": 1.4931,
+      "step": 1592
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 1.5493406057357788,
+      "eval_runtime": 77.1952,
+      "eval_samples_per_second": 1.671,
+      "eval_steps_per_second": 0.117,
+      "step": 1592
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9726873059767965e-05,
+      "loss": 1.4205,
+      "step": 1593
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.972602279324354e-05,
+      "loss": 1.454,
+      "step": 1594
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.972517122368196e-05,
+      "loss": 1.4603,
+      "step": 1595
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.972431835119732e-05,
+      "loss": 1.4607,
+      "step": 1596
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9723464175903874e-05,
+      "loss": 1.3517,
+      "step": 1597
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9722608697916068e-05,
+      "loss": 1.6674,
+      "step": 1598
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9721751917348507e-05,
+      "loss": 1.5046,
+      "step": 1599
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9720893834315985e-05,
+      "loss": 1.4661,
+      "step": 1600
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9720034448933463e-05,
+      "loss": 1.4868,
+      "step": 1601
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9719173761316068e-05,
+      "loss": 1.5264,
+      "step": 1602
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9718311771579114e-05,
+      "loss": 1.6156,
+      "step": 1603
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.971744847983809e-05,
+      "loss": 1.5134,
+      "step": 1604
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9716583886208652e-05,
+      "loss": 1.4823,
+      "step": 1605
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9715717990806627e-05,
+      "loss": 1.5776,
+      "step": 1606
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9714850793748032e-05,
+      "loss": 1.4972,
+      "step": 1607
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.971398229514905e-05,
+      "loss": 1.5078,
+      "step": 1608
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9713112495126027e-05,
+      "loss": 1.489,
+      "step": 1609
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.97122413937955e-05,
+      "loss": 1.5013,
+      "step": 1610
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.971136899127418e-05,
+      "loss": 1.4875,
+      "step": 1611
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9710495287678935e-05,
+      "loss": 1.489,
+      "step": 1612
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9709620283126827e-05,
+      "loss": 1.4884,
+      "step": 1613
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9708743977735082e-05,
+      "loss": 1.545,
+      "step": 1614
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.97078663716211e-05,
+      "loss": 1.5468,
+      "step": 1615
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9706987464902463e-05,
+      "loss": 1.5588,
+      "step": 1616
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9706107257696913e-05,
+      "loss": 1.5196,
+      "step": 1617
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9705225750122387e-05,
+      "loss": 1.4972,
+      "step": 1618
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9704342942296976e-05,
+      "loss": 1.5124,
+      "step": 1619
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9703458834338956e-05,
+      "loss": 1.5484,
+      "step": 1620
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.970257342636677e-05,
+      "loss": 1.6969,
+      "step": 1621
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.970168671849905e-05,
+      "loss": 1.5168,
+      "step": 1622
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9700798710854587e-05,
+      "loss": 1.5206,
+      "step": 1623
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.969990940355235e-05,
+      "loss": 1.5506,
+      "step": 1624
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.969901879671148e-05,
+      "loss": 1.4989,
+      "step": 1625
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9698126890451302e-05,
+      "loss": 1.497,
+      "step": 1626
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9697233684891307e-05,
+      "loss": 1.4643,
+      "step": 1627
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9696339180151158e-05,
+      "loss": 1.6567,
+      "step": 1628
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.96954433763507e-05,
+      "loss": 1.5363,
+      "step": 1629
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9694546273609946e-05,
+      "loss": 1.5112,
+      "step": 1630
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9693647872049085e-05,
+      "loss": 1.5175,
+      "step": 1631
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.969274817178848e-05,
+      "loss": 1.5139,
+      "step": 1632
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9691847172948663e-05,
+      "loss": 1.6995,
+      "step": 1633
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9690944875650354e-05,
+      "loss": 1.6399,
+      "step": 1634
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.969004128001443e-05,
+      "loss": 1.5386,
+      "step": 1635
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.968913638616195e-05,
+      "loss": 1.5735,
+      "step": 1636
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.968823019421415e-05,
+      "loss": 1.4948,
+      "step": 1637
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9687322704292436e-05,
+      "loss": 1.5309,
+      "step": 1638
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9686413916518388e-05,
+      "loss": 1.4924,
+      "step": 1639
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9685503831013757e-05,
+      "loss": 1.4575,
+      "step": 1640
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9684592447900478e-05,
+      "loss": 1.4604,
+      "step": 1641
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9683679767300643e-05,
+      "loss": 1.4625,
+      "step": 1642
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9682765789336534e-05,
+      "loss": 1.483,
+      "step": 1643
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9681850514130603e-05,
+      "loss": 1.4649,
+      "step": 1644
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.968093394180547e-05,
+      "loss": 1.5248,
+      "step": 1645
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.968001607248393e-05,
+      "loss": 1.5093,
+      "step": 1646
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9679096906288958e-05,
+      "loss": 1.5106,
+      "step": 1647
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9678176443343697e-05,
+      "loss": 1.6438,
+      "step": 1648
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9677254683771466e-05,
+      "loss": 1.5193,
+      "step": 1649
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9676331627695756e-05,
+      "loss": 1.5299,
+      "step": 1650
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.967540727524023e-05,
+      "loss": 1.6404,
+      "step": 1651
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9674481626528734e-05,
+      "loss": 1.5398,
+      "step": 1652
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9673554681685277e-05,
+      "loss": 1.5611,
+      "step": 1653
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9672626440834045e-05,
+      "loss": 1.4949,
+      "step": 1654
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9671696904099403e-05,
+      "loss": 1.5426,
+      "step": 1655
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9670766071605877e-05,
+      "loss": 1.5039,
+      "step": 1656
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9669833943478185e-05,
+      "loss": 1.418,
+      "step": 1657
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.96689005198412e-05,
+      "loss": 1.4904,
+      "step": 1658
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.966796580081998e-05,
+      "loss": 1.4831,
+      "step": 1659
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9667029786539754e-05,
+      "loss": 1.5287,
+      "step": 1660
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9666092477125923e-05,
+      "loss": 1.4392,
+      "step": 1661
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.966515387270406e-05,
+      "loss": 1.4647,
+      "step": 1662
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.966421397339992e-05,
+      "loss": 1.5023,
+      "step": 1663
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.966327277933942e-05,
+      "loss": 1.4992,
+      "step": 1664
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9662330290648658e-05,
+      "loss": 1.486,
+      "step": 1665
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9661386507453898e-05,
+      "loss": 1.8505,
+      "step": 1666
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9660441429881592e-05,
+      "loss": 1.4769,
+      "step": 1667
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.965949505805835e-05,
+      "loss": 1.4882,
+      "step": 1668
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9658547392110962e-05,
+      "loss": 1.5867,
+      "step": 1669
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9657598432166392e-05,
+      "loss": 1.5892,
+      "step": 1670
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9656648178351778e-05,
+      "loss": 1.9443,
+      "step": 1671
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9655696630794426e-05,
+      "loss": 1.4952,
+      "step": 1672
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.965474378962182e-05,
+      "loss": 1.461,
+      "step": 1673
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9653789654961618e-05,
+      "loss": 1.4979,
+      "step": 1674
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9652834226941647e-05,
+      "loss": 1.4456,
+      "step": 1675
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9651877505689912e-05,
+      "loss": 1.4105,
+      "step": 1676
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.965091949133459e-05,
+      "loss": 1.5391,
+      "step": 1677
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9649960184004027e-05,
+      "loss": 1.4479,
+      "step": 1678
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9648999583826747e-05,
+      "loss": 1.4842,
+      "step": 1679
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9648037690931446e-05,
+      "loss": 1.5078,
+      "step": 1680
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.964707450544699e-05,
+      "loss": 1.5522,
+      "step": 1681
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.964611002750243e-05,
+      "loss": 1.548,
+      "step": 1682
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9645144257226972e-05,
+      "loss": 1.4941,
+      "step": 1683
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9644177194750006e-05,
+      "loss": 1.4791,
+      "step": 1684
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9643208840201095e-05,
+      "loss": 1.5588,
+      "step": 1685
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9642239193709974e-05,
+      "loss": 1.5159,
+      "step": 1686
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9641268255406553e-05,
+      "loss": 1.5366,
+      "step": 1687
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9640296025420907e-05,
+      "loss": 1.501,
+      "step": 1688
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9639322503883292e-05,
+      "loss": 1.4766,
+      "step": 1689
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9638347690924138e-05,
+      "loss": 1.5004,
+      "step": 1690
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9637371586674044e-05,
+      "loss": 1.4943,
+      "step": 1691
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9636394191263777e-05,
+      "loss": 1.5243,
+      "step": 1692
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9635415504824294e-05,
+      "loss": 1.4976,
+      "step": 1693
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9634435527486704e-05,
+      "loss": 1.478,
+      "step": 1694
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.96334542593823e-05,
+      "loss": 1.4437,
+      "step": 1695
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9632471700642553e-05,
+      "loss": 1.4769,
+      "step": 1696
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9631487851399095e-05,
+      "loss": 1.4387,
+      "step": 1697
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9630502711783737e-05,
+      "loss": 1.5988,
+      "step": 1698
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.962951628192846e-05,
+      "loss": 1.5576,
+      "step": 1699
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.962852856196543e-05,
+      "loss": 1.5591,
+      "step": 1700
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.962753955202697e-05,
+      "loss": 1.5206,
+      "step": 1701
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9626549252245578e-05,
+      "loss": 1.5233,
+      "step": 1702
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9625557662753933e-05,
+      "loss": 1.5083,
+      "step": 1703
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.962456478368488e-05,
+      "loss": 1.5737,
+      "step": 1704
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9623570615171445e-05,
+      "loss": 1.5038,
+      "step": 1705
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9622575157346815e-05,
+      "loss": 1.3908,
+      "step": 1706
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9621578410344358e-05,
+      "loss": 1.5274,
+      "step": 1707
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.962058037429761e-05,
+      "loss": 1.508,
+      "step": 1708
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9619581049340288e-05,
+      "loss": 1.518,
+      "step": 1709
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9618580435606274e-05,
+      "loss": 1.5158,
+      "step": 1710
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9617578533229622e-05,
+      "loss": 1.4889,
+      "step": 1711
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.961657534234456e-05,
+      "loss": 1.5163,
+      "step": 1712
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9615570863085492e-05,
+      "loss": 1.5993,
+      "step": 1713
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9614565095586995e-05,
+      "loss": 1.4754,
+      "step": 1714
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.961355803998381e-05,
+      "loss": 1.5367,
+      "step": 1715
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9612549696410865e-05,
+      "loss": 1.5441,
+      "step": 1716
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9611540065003246e-05,
+      "loss": 1.5446,
+      "step": 1717
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9610529145896215e-05,
+      "loss": 1.5358,
+      "step": 1718
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9609516939225216e-05,
+      "loss": 1.4814,
+      "step": 1719
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9608503445125855e-05,
+      "loss": 1.4851,
+      "step": 1720
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9607488663733915e-05,
+      "loss": 1.4926,
+      "step": 1721
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.960647259518535e-05,
+      "loss": 1.5555,
+      "step": 1722
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9605455239616287e-05,
+      "loss": 1.4419,
+      "step": 1723
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9604436597163025e-05,
+      "loss": 1.4961,
+      "step": 1724
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9603416667962037e-05,
+      "loss": 1.5422,
+      "step": 1725
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9602395452149972e-05,
+      "loss": 1.5925,
+      "step": 1726
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.960137294986364e-05,
+      "loss": 1.4478,
+      "step": 1727
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.960034916124003e-05,
+      "loss": 1.5802,
+      "step": 1728
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9599324086416307e-05,
+      "loss": 1.5275,
+      "step": 1729
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.95982977255298e-05,
+      "loss": 1.4577,
+      "step": 1730
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9597270078718022e-05,
+      "loss": 1.5157,
+      "step": 1731
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9596241146118648e-05,
+      "loss": 1.5256,
+      "step": 1732
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9595210927869525e-05,
+      "loss": 1.5512,
+      "step": 1733
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.959417942410868e-05,
+      "loss": 1.5273,
+      "step": 1734
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9593146634974305e-05,
+      "loss": 1.494,
+      "step": 1735
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.959211256060477e-05,
+      "loss": 1.5764,
+      "step": 1736
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9591077201138612e-05,
+      "loss": 1.2567,
+      "step": 1737
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9590040556714546e-05,
+      "loss": 1.5334,
+      "step": 1738
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.958900262747145e-05,
+      "loss": 1.4827,
+      "step": 1739
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9587963413548388e-05,
+      "loss": 1.483,
+      "step": 1740
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.958692291508458e-05,
+      "loss": 1.5099,
+      "step": 1741
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.958588113221943e-05,
+      "loss": 1.5064,
+      "step": 1742
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.958483806509251e-05,
+      "loss": 1.5683,
+      "step": 1743
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9583793713843563e-05,
+      "loss": 1.4854,
+      "step": 1744
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9582748078612507e-05,
+      "loss": 1.461,
+      "step": 1745
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.958170115953943e-05,
+      "loss": 1.4945,
+      "step": 1746
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.958065295676459e-05,
+      "loss": 1.698,
+      "step": 1747
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.957960347042842e-05,
+      "loss": 1.4092,
+      "step": 1748
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9578552700671527e-05,
+      "loss": 1.4639,
+      "step": 1749
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9577500647634683e-05,
+      "loss": 1.4459,
+      "step": 1750
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.957644731145884e-05,
+      "loss": 1.4512,
+      "step": 1751
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9575392692285114e-05,
+      "loss": 1.4468,
+      "step": 1752
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.95743367902548e-05,
+      "loss": 1.5132,
+      "step": 1753
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.957327960550936e-05,
+      "loss": 1.5159,
+      "step": 1754
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9572221138190433e-05,
+      "loss": 1.498,
+      "step": 1755
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9571161388439823e-05,
+      "loss": 1.5547,
+      "step": 1756
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9570100356399505e-05,
+      "loss": 1.4997,
+      "step": 1757
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9569038042211642e-05,
+      "loss": 1.4873,
+      "step": 1758
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9567974446018546e-05,
+      "loss": 1.5333,
+      "step": 1759
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.956690956796272e-05,
+      "loss": 1.4821,
+      "step": 1760
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9565843408186816e-05,
+      "loss": 1.6178,
+      "step": 1761
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.956477596683369e-05,
+      "loss": 1.4491,
+      "step": 1762
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.956370724404634e-05,
+      "loss": 1.5331,
+      "step": 1763
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9562637239967953e-05,
+      "loss": 1.4756,
+      "step": 1764
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.956156595474188e-05,
+      "loss": 1.4727,
+      "step": 1765
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9560493388511642e-05,
+      "loss": 1.4892,
+      "step": 1766
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9559419541420942e-05,
+      "loss": 1.4899,
+      "step": 1767
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9558344413613645e-05,
+      "loss": 1.4929,
+      "step": 1768
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.955726800523379e-05,
+      "loss": 1.5537,
+      "step": 1769
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9556190316425583e-05,
+      "loss": 1.4552,
+      "step": 1770
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9555111347333414e-05,
+      "loss": 1.5802,
+      "step": 1771
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.955403109810184e-05,
+      "loss": 1.5085,
+      "step": 1772
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.955294956887558e-05,
+      "loss": 1.4562,
+      "step": 1773
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.955186675979953e-05,
+      "loss": 1.4304,
+      "step": 1774
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9550782671018766e-05,
+      "loss": 1.4944,
+      "step": 1775
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9549697302678518e-05,
+      "loss": 1.6818,
+      "step": 1776
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9548610654924206e-05,
+      "loss": 1.4304,
+      "step": 1777
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9547522727901408e-05,
+      "loss": 1.5663,
+      "step": 1778
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9546433521755883e-05,
+      "loss": 1.483,
+      "step": 1779
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9545343036633553e-05,
+      "loss": 1.4762,
+      "step": 1780
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9544251272680514e-05,
+      "loss": 1.4492,
+      "step": 1781
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.954315823004304e-05,
+      "loss": 1.5822,
+      "step": 1782
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9542063908867565e-05,
+      "loss": 1.4646,
+      "step": 1783
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.95409683093007e-05,
+      "loss": 1.5238,
+      "step": 1784
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.953987143148923e-05,
+      "loss": 1.5326,
+      "step": 1785
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9538773275580106e-05,
+      "loss": 1.4933,
+      "step": 1786
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9537673841720455e-05,
+      "loss": 1.5095,
+      "step": 1787
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.953657313005757e-05,
+      "loss": 1.5628,
+      "step": 1788
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9535471140738922e-05,
+      "loss": 1.4666,
+      "step": 1789
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9534367873912146e-05,
+      "loss": 1.5552,
+      "step": 1790
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9533263329725054e-05,
+      "loss": 1.4972,
+      "step": 1791
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 1.540722131729126,
+      "eval_runtime": 77.1546,
+      "eval_samples_per_second": 1.672,
+      "eval_steps_per_second": 0.117,
+      "step": 1791
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9532157508325622e-05,
+      "loss": 1.4485,
+      "step": 1792
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9531050409862008e-05,
+      "loss": 1.418,
+      "step": 1793
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.952994203448253e-05,
+      "loss": 1.4201,
+      "step": 1794
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9528832382335683e-05,
+      "loss": 1.4707,
+      "step": 1795
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.952772145357013e-05,
+      "loss": 1.4557,
+      "step": 1796
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9526609248334717e-05,
+      "loss": 1.5078,
+      "step": 1797
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9525495766778437e-05,
+      "loss": 1.5126,
+      "step": 1798
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9524381009050476e-05,
+      "loss": 1.565,
+      "step": 1799
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9523264975300175e-05,
+      "loss": 1.4532,
+      "step": 1800
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9522147665677067e-05,
+      "loss": 1.4712,
+      "step": 1801
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.952102908033083e-05,
+      "loss": 1.601,
+      "step": 1802
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9519909219411336e-05,
+      "loss": 1.5504,
+      "step": 1803
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.951878808306861e-05,
+      "loss": 1.4609,
+      "step": 1804
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9517665671452862e-05,
+      "loss": 1.5569,
+      "step": 1805
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9516541984714457e-05,
+      "loss": 1.6178,
+      "step": 1806
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9515417023003947e-05,
+      "loss": 1.5168,
+      "step": 1807
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.951429078647205e-05,
+      "loss": 1.6349,
+      "step": 1808
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9513163275269645e-05,
+      "loss": 1.4643,
+      "step": 1809
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9512034489547796e-05,
+      "loss": 1.4292,
+      "step": 1810
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.951090442945773e-05,
+      "loss": 1.4776,
+      "step": 1811
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9509773095150843e-05,
+      "loss": 1.5103,
+      "step": 1812
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.950864048677871e-05,
+      "loss": 1.3963,
+      "step": 1813
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9507506604493068e-05,
+      "loss": 1.5503,
+      "step": 1814
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9506371448445826e-05,
+      "loss": 1.5578,
+      "step": 1815
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.950523501878907e-05,
+      "loss": 1.5377,
+      "step": 1816
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.950409731567505e-05,
+      "loss": 1.5077,
+      "step": 1817
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9502958339256192e-05,
+      "loss": 1.4658,
+      "step": 1818
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.950181808968508e-05,
+      "loss": 1.5149,
+      "step": 1819
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9500676567114493e-05,
+      "loss": 1.4722,
+      "step": 1820
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9499533771697355e-05,
+      "loss": 1.5173,
+      "step": 1821
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9498389703586772e-05,
+      "loss": 1.4565,
+      "step": 1822
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9497244362936024e-05,
+      "loss": 1.5425,
+      "step": 1823
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9496097749898556e-05,
+      "loss": 1.5643,
+      "step": 1824
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9494949864627983e-05,
+      "loss": 1.6174,
+      "step": 1825
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9493800707278094e-05,
+      "loss": 1.4554,
+      "step": 1826
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9492650278002844e-05,
+      "loss": 1.4995,
+      "step": 1827
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9491498576956362e-05,
+      "loss": 1.3844,
+      "step": 1828
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.949034560429295e-05,
+      "loss": 1.4938,
+      "step": 1829
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.948919136016707e-05,
+      "loss": 1.4207,
+      "step": 1830
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9488035844733367e-05,
+      "loss": 1.5429,
+      "step": 1831
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9486879058146647e-05,
+      "loss": 1.4942,
+      "step": 1832
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.948572100056189e-05,
+      "loss": 1.461,
+      "step": 1833
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.948456167213425e-05,
+      "loss": 1.5331,
+      "step": 1834
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.948340107301904e-05,
+      "loss": 1.5576,
+      "step": 1835
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9482239203371755e-05,
+      "loss": 1.61,
+      "step": 1836
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.948107606334806e-05,
+      "loss": 1.5192,
+      "step": 1837
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9479911653103775e-05,
+      "loss": 1.4498,
+      "step": 1838
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9478745972794907e-05,
+      "loss": 1.5356,
+      "step": 1839
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.947757902257763e-05,
+      "loss": 1.574,
+      "step": 1840
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9476410802608278e-05,
+      "loss": 1.4712,
+      "step": 1841
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9475241313043372e-05,
+      "loss": 1.5113,
+      "step": 1842
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9474070554039587e-05,
+      "loss": 1.4674,
+      "step": 1843
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9472898525753776e-05,
+      "loss": 1.429,
+      "step": 1844
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.947172522834296e-05,
+      "loss": 1.5034,
+      "step": 1845
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9470550661964332e-05,
+      "loss": 1.4387,
+      "step": 1846
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9469374826775252e-05,
+      "loss": 1.4123,
+      "step": 1847
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9468197722933255e-05,
+      "loss": 1.5181,
+      "step": 1848
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9467019350596038e-05,
+      "loss": 1.5571,
+      "step": 1849
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9465839709921477e-05,
+      "loss": 1.527,
+      "step": 1850
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9464658801067612e-05,
+      "loss": 1.4841,
+      "step": 1851
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9463476624192655e-05,
+      "loss": 1.6462,
+      "step": 1852
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9462293179454985e-05,
+      "loss": 1.4558,
+      "step": 1853
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9461108467013155e-05,
+      "loss": 1.4901,
+      "step": 1854
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9459922487025883e-05,
+      "loss": 1.4548,
+      "step": 1855
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9458735239652068e-05,
+      "loss": 1.3878,
+      "step": 1856
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.945754672505076e-05,
+      "loss": 1.4575,
+      "step": 1857
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.94563569433812e-05,
+      "loss": 1.5251,
+      "step": 1858
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.945516589480278e-05,
+      "loss": 1.4804,
+      "step": 1859
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9453973579475073e-05,
+      "loss": 1.4914,
+      "step": 1860
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9452779997557817e-05,
+      "loss": 1.4812,
+      "step": 1861
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9451585149210923e-05,
+      "loss": 1.4823,
+      "step": 1862
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.945038903459447e-05,
+      "loss": 1.511,
+      "step": 1863
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9449191653868705e-05,
+      "loss": 1.5289,
+      "step": 1864
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9447993007194052e-05,
+      "loss": 1.4989,
+      "step": 1865
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.944679309473109e-05,
+      "loss": 1.5965,
+      "step": 1866
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.944559191664058e-05,
+      "loss": 1.4929,
+      "step": 1867
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9444389473083456e-05,
+      "loss": 1.5537,
+      "step": 1868
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9443185764220805e-05,
+      "loss": 1.5499,
+      "step": 1869
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9441980790213892e-05,
+      "loss": 1.4963,
+      "step": 1870
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.944077455122416e-05,
+      "loss": 1.5099,
+      "step": 1871
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9439567047413214e-05,
+      "loss": 1.5104,
+      "step": 1872
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9438358278942822e-05,
+      "loss": 1.548,
+      "step": 1873
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9437148245974934e-05,
+      "loss": 1.5168,
+      "step": 1874
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.943593694867166e-05,
+      "loss": 1.4347,
+      "step": 1875
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9434724387195282e-05,
+      "loss": 1.483,
+      "step": 1876
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9433510561708253e-05,
+      "loss": 1.5192,
+      "step": 1877
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.94322954723732e-05,
+      "loss": 1.4881,
+      "step": 1878
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.94310791193529e-05,
+      "loss": 1.4582,
+      "step": 1879
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9429861502810326e-05,
+      "loss": 1.4948,
+      "step": 1880
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9428642622908606e-05,
+      "loss": 1.3547,
+      "step": 1881
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9427422479811032e-05,
+      "loss": 1.5116,
+      "step": 1882
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.942620107368108e-05,
+      "loss": 1.6105,
+      "step": 1883
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.942497840468238e-05,
+      "loss": 1.5252,
+      "step": 1884
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.942375447297874e-05,
+      "loss": 1.583,
+      "step": 1885
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9422529278734133e-05,
+      "loss": 1.4526,
+      "step": 1886
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9421302822112708e-05,
+      "loss": 1.4635,
+      "step": 1887
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.942007510327878e-05,
+      "loss": 1.4575,
+      "step": 1888
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9418846122396827e-05,
+      "loss": 1.5382,
+      "step": 1889
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9417615879631508e-05,
+      "loss": 1.5819,
+      "step": 1890
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9416384375147635e-05,
+      "loss": 1.5065,
+      "step": 1891
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.94151516091102e-05,
+      "loss": 1.4488,
+      "step": 1892
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9413917581684366e-05,
+      "loss": 1.4391,
+      "step": 1893
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9412682293035458e-05,
+      "loss": 1.5256,
+      "step": 1894
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9411445743328974e-05,
+      "loss": 1.4471,
+      "step": 1895
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.941020793273058e-05,
+      "loss": 1.5442,
+      "step": 1896
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9408968861406113e-05,
+      "loss": 1.5141,
+      "step": 1897
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.940772852952157e-05,
+      "loss": 1.4379,
+      "step": 1898
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.940648693724313e-05,
+      "loss": 1.4989,
+      "step": 1899
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.940524408473713e-05,
+      "loss": 1.501,
+      "step": 1900
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9403999972170088e-05,
+      "loss": 1.5277,
+      "step": 1901
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9402754599708675e-05,
+      "loss": 1.5655,
+      "step": 1902
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9401507967519738e-05,
+      "loss": 1.5326,
+      "step": 1903
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9400260075770304e-05,
+      "loss": 1.508,
+      "step": 1904
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9399010924627548e-05,
+      "loss": 1.4971,
+      "step": 1905
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9397760514258825e-05,
+      "loss": 1.5248,
+      "step": 1906
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9396508844831663e-05,
+      "loss": 1.5119,
+      "step": 1907
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9395255916513753e-05,
+      "loss": 1.4492,
+      "step": 1908
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9394001729472954e-05,
+      "loss": 1.3187,
+      "step": 1909
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.939274628387729e-05,
+      "loss": 0.7899,
+      "step": 1910
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9391489579894967e-05,
+      "loss": 1.2956,
+      "step": 1911
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9390231617694343e-05,
+      "loss": 1.5321,
+      "step": 1912
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.938897239744396e-05,
+      "loss": 1.4484,
+      "step": 1913
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9387711919312516e-05,
+      "loss": 1.5293,
+      "step": 1914
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9386450183468883e-05,
+      "loss": 1.477,
+      "step": 1915
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.93851871900821e-05,
+      "loss": 1.4962,
+      "step": 1916
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9383922939321388e-05,
+      "loss": 1.4596,
+      "step": 1917
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9382657431356105e-05,
+      "loss": 1.6164,
+      "step": 1918
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.938139066635581e-05,
+      "loss": 1.4752,
+      "step": 1919
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.938012264449021e-05,
+      "loss": 1.5322,
+      "step": 1920
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.937885336592919e-05,
+      "loss": 1.4607,
+      "step": 1921
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9377582830842803e-05,
+      "loss": 1.6286,
+      "step": 1922
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9376311039401265e-05,
+      "loss": 1.553,
+      "step": 1923
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.937503799177497e-05,
+      "loss": 1.48,
+      "step": 1924
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.937376368813446e-05,
+      "loss": 1.5161,
+      "step": 1925
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.937248812865047e-05,
+      "loss": 1.5614,
+      "step": 1926
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.937121131349389e-05,
+      "loss": 1.503,
+      "step": 1927
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9369933242835773e-05,
+      "loss": 1.5059,
+      "step": 1928
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.936865391684736e-05,
+      "loss": 1.3895,
+      "step": 1929
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.936737333570004e-05,
+      "loss": 1.1535,
+      "step": 1930
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.936609149956538e-05,
+      "loss": 1.1182,
+      "step": 1931
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9364808408615118e-05,
+      "loss": 1.0568,
+      "step": 1932
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9363524063021143e-05,
+      "loss": 1.066,
+      "step": 1933
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9362238462955533e-05,
+      "loss": 0.9715,
+      "step": 1934
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.936095160859053e-05,
+      "loss": 0.9349,
+      "step": 1935
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9359663500098524e-05,
+      "loss": 0.9757,
+      "step": 1936
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9358374137652103e-05,
+      "loss": 0.9194,
+      "step": 1937
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9357083521424004e-05,
+      "loss": 1.4673,
+      "step": 1938
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9355791651587132e-05,
+      "loss": 1.5178,
+      "step": 1939
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9354498528314568e-05,
+      "loss": 1.5177,
+      "step": 1940
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.935320415177956e-05,
+      "loss": 1.4736,
+      "step": 1941
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.935190852215552e-05,
+      "loss": 1.4444,
+      "step": 1942
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9350611639616023e-05,
+      "loss": 1.5774,
+      "step": 1943
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9349313504334825e-05,
+      "loss": 1.4976,
+      "step": 1944
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.934801411648584e-05,
+      "loss": 1.4697,
+      "step": 1945
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9346713476243154e-05,
+      "loss": 1.5035,
+      "step": 1946
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.934541158378102e-05,
+      "loss": 1.4809,
+      "step": 1947
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9344108439273853e-05,
+      "loss": 1.5242,
+      "step": 1948
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9342804042896244e-05,
+      "loss": 1.4528,
+      "step": 1949
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9341498394822947e-05,
+      "loss": 1.473,
+      "step": 1950
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9340191495228892e-05,
+      "loss": 1.4217,
+      "step": 1951
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9338883344289165e-05,
+      "loss": 1.5586,
+      "step": 1952
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.933757394217902e-05,
+      "loss": 1.5473,
+      "step": 1953
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.933626328907389e-05,
+      "loss": 1.4718,
+      "step": 1954
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9334951385149367e-05,
+      "loss": 1.3981,
+      "step": 1955
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9333638230581214e-05,
+      "loss": 1.4659,
+      "step": 1956
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9332323825545355e-05,
+      "loss": 1.4542,
+      "step": 1957
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.933100817021789e-05,
+      "loss": 1.433,
+      "step": 1958
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9329691264775087e-05,
+      "loss": 1.5316,
+      "step": 1959
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.932837310939337e-05,
+      "loss": 1.443,
+      "step": 1960
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.932705370424934e-05,
+      "loss": 1.5022,
+      "step": 1961
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9325733049519763e-05,
+      "loss": 1.5283,
+      "step": 1962
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9324411145381575e-05,
+      "loss": 1.4691,
+      "step": 1963
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9323087992011878e-05,
+      "loss": 1.5221,
+      "step": 1964
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9321763589587943e-05,
+      "loss": 1.4747,
+      "step": 1965
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9320437938287198e-05,
+      "loss": 1.4269,
+      "step": 1966
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.931911103828725e-05,
+      "loss": 1.5038,
+      "step": 1967
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9317782889765872e-05,
+      "loss": 1.4501,
+      "step": 1968
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9316453492900996e-05,
+      "loss": 1.4949,
+      "step": 1969
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.931512284787073e-05,
+      "loss": 1.5188,
+      "step": 1970
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.931379095485335e-05,
+      "loss": 1.4313,
+      "step": 1971
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9312457814027293e-05,
+      "loss": 1.4908,
+      "step": 1972
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9311123425571164e-05,
+      "loss": 1.537,
+      "step": 1973
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9309787789663735e-05,
+      "loss": 1.4315,
+      "step": 1974
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.930845090648395e-05,
+      "loss": 1.4784,
+      "step": 1975
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.930711277621092e-05,
+      "loss": 1.6129,
+      "step": 1976
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9305773399023913e-05,
+      "loss": 1.5168,
+      "step": 1977
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9304432775102376e-05,
+      "loss": 1.479,
+      "step": 1978
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9303090904625917e-05,
+      "loss": 1.472,
+      "step": 1979
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.930174778777431e-05,
+      "loss": 1.4995,
+      "step": 1980
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.93004034247275e-05,
+      "loss": 1.5154,
+      "step": 1981
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9299057815665598e-05,
+      "loss": 1.5628,
+      "step": 1982
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.929771096076888e-05,
+      "loss": 1.5196,
+      "step": 1983
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.929636286021779e-05,
+      "loss": 1.4267,
+      "step": 1984
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9295013514192943e-05,
+      "loss": 1.5249,
+      "step": 1985
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.929366292287511e-05,
+      "loss": 1.4895,
+      "step": 1986
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9292311086445237e-05,
+      "loss": 1.4965,
+      "step": 1987
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9290958005084438e-05,
+      "loss": 1.5273,
+      "step": 1988
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9289603678973988e-05,
+      "loss": 1.4002,
+      "step": 1989
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9288248108295335e-05,
+      "loss": 1.5349,
+      "step": 1990
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 1.530476450920105,
+      "eval_runtime": 77.2115,
+      "eval_samples_per_second": 1.671,
+      "eval_steps_per_second": 0.117,
+      "step": 1990
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 9536,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 398,
+  "total_flos": 8.902544619863015e+19,
+  "trial_name": null,
+  "trial_params": null
+}