diff --git "a/checkpoint-4378/trainer_state.json" "b/checkpoint-4378/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-4378/trainer_state.json"
@@ -0,0 +1,26464 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.4591023489932886,
+  "eval_steps": 199,
+  "global_step": 4378,
+  "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
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.928689129323009e-05,
+      "loss": 1.421,
+      "step": 1991
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.928553323396003e-05,
+      "loss": 1.5625,
+      "step": 1992
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.92841739306671e-05,
+      "loss": 1.5198,
+      "step": 1993
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9282813383533415e-05,
+      "loss": 1.5116,
+      "step": 1994
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9281451592741247e-05,
+      "loss": 1.4306,
+      "step": 1995
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9280088558473046e-05,
+      "loss": 1.5425,
+      "step": 1996
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9278724280911424e-05,
+      "loss": 1.4972,
+      "step": 1997
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9277358760239156e-05,
+      "loss": 1.5185,
+      "step": 1998
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9275991996639188e-05,
+      "loss": 1.5187,
+      "step": 1999
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.927462399029463e-05,
+      "loss": 1.4325,
+      "step": 2000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9273254741388763e-05,
+      "loss": 1.4397,
+      "step": 2001
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.927188425010503e-05,
+      "loss": 1.5319,
+      "step": 2002
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9270512516627042e-05,
+      "loss": 1.5664,
+      "step": 2003
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9269139541138574e-05,
+      "loss": 1.5227,
+      "step": 2004
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.926776532382357e-05,
+      "loss": 1.5187,
+      "step": 2005
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9266389864866134e-05,
+      "loss": 1.4029,
+      "step": 2006
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9265013164450557e-05,
+      "loss": 1.387,
+      "step": 2007
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9263635222761268e-05,
+      "loss": 1.5047,
+      "step": 2008
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9262256039982882e-05,
+      "loss": 1.4816,
+      "step": 2009
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9260875616300172e-05,
+      "loss": 1.4607,
+      "step": 2010
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9259493951898084e-05,
+      "loss": 1.628,
+      "step": 2011
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9258111046961714e-05,
+      "loss": 1.4149,
+      "step": 2012
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.925672690167635e-05,
+      "loss": 1.5303,
+      "step": 2013
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.925534151622742e-05,
+      "loss": 1.5198,
+      "step": 2014
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9253954890800535e-05,
+      "loss": 1.6422,
+      "step": 2015
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.925256702558147e-05,
+      "loss": 1.4738,
+      "step": 2016
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.925117792075616e-05,
+      "loss": 1.4752,
+      "step": 2017
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9249787576510713e-05,
+      "loss": 1.4547,
+      "step": 2018
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.924839599303139e-05,
+      "loss": 1.4864,
+      "step": 2019
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9247003170504638e-05,
+      "loss": 1.5037,
+      "step": 2020
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9245609109117053e-05,
+      "loss": 1.5089,
+      "step": 2021
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.924421380905541e-05,
+      "loss": 1.5616,
+      "step": 2022
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9242817270506634e-05,
+      "loss": 1.4592,
+      "step": 2023
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9241419493657832e-05,
+      "loss": 1.5592,
+      "step": 2024
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.924002047869627e-05,
+      "loss": 1.4897,
+      "step": 2025
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9238620225809375e-05,
+      "loss": 1.4743,
+      "step": 2026
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.923721873518475e-05,
+      "loss": 1.4686,
+      "step": 2027
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.923581600701016e-05,
+      "loss": 1.461,
+      "step": 2028
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9234412041473528e-05,
+      "loss": 1.6622,
+      "step": 2029
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9233006838762955e-05,
+      "loss": 1.4707,
+      "step": 2030
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.92316003990667e-05,
+      "loss": 1.5566,
+      "step": 2031
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9230192722573188e-05,
+      "loss": 1.4881,
+      "step": 2032
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9228783809471015e-05,
+      "loss": 1.6124,
+      "step": 2033
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9227373659948937e-05,
+      "loss": 1.3508,
+      "step": 2034
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9225962274195876e-05,
+      "loss": 1.4418,
+      "step": 2035
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9224549652400927e-05,
+      "loss": 1.3954,
+      "step": 2036
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.922313579475334e-05,
+      "loss": 1.4798,
+      "step": 2037
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.922172070144254e-05,
+      "loss": 1.4556,
+      "step": 2038
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.922030437265811e-05,
+      "loss": 1.4987,
+      "step": 2039
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9218886808589796e-05,
+      "loss": 1.4857,
+      "step": 2040
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9217468009427527e-05,
+      "loss": 1.5289,
+      "step": 2041
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.921604797536138e-05,
+      "loss": 1.5267,
+      "step": 2042
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9214626706581598e-05,
+      "loss": 1.4655,
+      "step": 2043
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9213204203278602e-05,
+      "loss": 1.4598,
+      "step": 2044
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9211780465642973e-05,
+      "loss": 1.4766,
+      "step": 2045
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.921035549386545e-05,
+      "loss": 1.5148,
+      "step": 2046
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9208929288136933e-05,
+      "loss": 1.5567,
+      "step": 2047
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.920750184864852e-05,
+      "loss": 1.4868,
+      "step": 2048
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.920607317559143e-05,
+      "loss": 1.4752,
+      "step": 2049
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.920464326915708e-05,
+      "loss": 1.4882,
+      "step": 2050
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9203212129537035e-05,
+      "loss": 1.5933,
+      "step": 2051
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9201779756923037e-05,
+      "loss": 1.4891,
+      "step": 2052
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9200346151506983e-05,
+      "loss": 1.4486,
+      "step": 2053
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9198911313480938e-05,
+      "loss": 1.5091,
+      "step": 2054
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.919747524303714e-05,
+      "loss": 1.4385,
+      "step": 2055
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9196037940367974e-05,
+      "loss": 1.4104,
+      "step": 2056
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9194599405666015e-05,
+      "loss": 1.4483,
+      "step": 2057
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9193159639123976e-05,
+      "loss": 1.4592,
+      "step": 2058
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.919171864093476e-05,
+      "loss": 1.4975,
+      "step": 2059
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9190276411291423e-05,
+      "loss": 1.4727,
+      "step": 2060
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9188832950387178e-05,
+      "loss": 1.4791,
+      "step": 2061
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9187388258415423e-05,
+      "loss": 1.3806,
+      "step": 2062
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9185942335569697e-05,
+      "loss": 1.4514,
+      "step": 2063
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9184495182043727e-05,
+      "loss": 1.4389,
+      "step": 2064
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.918304679803139e-05,
+      "loss": 1.5088,
+      "step": 2065
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9181597183726733e-05,
+      "loss": 1.4665,
+      "step": 2066
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9180146339323966e-05,
+      "loss": 1.489,
+      "step": 2067
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9178694265017466e-05,
+      "loss": 1.5215,
+      "step": 2068
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9177240961001775e-05,
+      "loss": 1.5152,
+      "step": 2069
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9175786427471594e-05,
+      "loss": 1.4418,
+      "step": 2070
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.91743306646218e-05,
+      "loss": 1.5001,
+      "step": 2071
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9172873672647417e-05,
+      "loss": 1.4879,
+      "step": 2072
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.917141545174366e-05,
+      "loss": 1.5299,
+      "step": 2073
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9169956002105877e-05,
+      "loss": 1.4602,
+      "step": 2074
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9168495323929605e-05,
+      "loss": 1.4921,
+      "step": 2075
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.916703341741054e-05,
+      "loss": 1.4659,
+      "step": 2076
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9165570282744534e-05,
+      "loss": 1.4707,
+      "step": 2077
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9164105920127613e-05,
+      "loss": 1.5354,
+      "step": 2078
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9162640329755962e-05,
+      "loss": 1.526,
+      "step": 2079
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.916117351182593e-05,
+      "loss": 1.6424,
+      "step": 2080
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.915970546653404e-05,
+      "loss": 1.5048,
+      "step": 2081
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.915823619407697e-05,
+      "loss": 1.5344,
+      "step": 2082
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.915676569465156e-05,
+      "loss": 1.4604,
+      "step": 2083
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9155293968454826e-05,
+      "loss": 1.4929,
+      "step": 2084
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9153821015683937e-05,
+      "loss": 1.4106,
+      "step": 2085
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.915234683653623e-05,
+      "loss": 1.4146,
+      "step": 2086
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9150871431209207e-05,
+      "loss": 1.4481,
+      "step": 2087
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.914939479990054e-05,
+      "loss": 1.5206,
+      "step": 2088
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9147916942808054e-05,
+      "loss": 1.5073,
+      "step": 2089
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9146437860129745e-05,
+      "loss": 1.5196,
+      "step": 2090
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9144957552063775e-05,
+      "loss": 1.5389,
+      "step": 2091
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9143476018808463e-05,
+      "loss": 1.5207,
+      "step": 2092
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9141993260562298e-05,
+      "loss": 1.3717,
+      "step": 2093
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9140509277523937e-05,
+      "loss": 1.471,
+      "step": 2094
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9139024069892185e-05,
+      "loss": 1.4553,
+      "step": 2095
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9137537637866027e-05,
+      "loss": 1.5311,
+      "step": 2096
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9136049981644614e-05,
+      "loss": 1.4369,
+      "step": 2097
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.913456110142724e-05,
+      "loss": 1.4528,
+      "step": 2098
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9133070997413384e-05,
+      "loss": 1.5294,
+      "step": 2099
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9131579669802685e-05,
+      "loss": 1.4588,
+      "step": 2100
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9130087118794934e-05,
+      "loss": 1.443,
+      "step": 2101
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9128593344590102e-05,
+      "loss": 1.4876,
+      "step": 2102
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.912709834738831e-05,
+      "loss": 1.4874,
+      "step": 2103
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9125602127389856e-05,
+      "loss": 1.4966,
+      "step": 2104
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.912410468479519e-05,
+      "loss": 1.5747,
+      "step": 2105
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9122606019804933e-05,
+      "loss": 1.4025,
+      "step": 2106
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.912110613261987e-05,
+      "loss": 1.4755,
+      "step": 2107
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.911960502344094e-05,
+      "loss": 1.4739,
+      "step": 2108
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.911810269246926e-05,
+      "loss": 1.4992,
+      "step": 2109
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.91165991399061e-05,
+      "loss": 1.4435,
+      "step": 2110
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.91150943659529e-05,
+      "loss": 1.4918,
+      "step": 2111
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9113588370811258e-05,
+      "loss": 1.3818,
+      "step": 2112
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.911208115468294e-05,
+      "loss": 1.5105,
+      "step": 2113
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9110572717769878e-05,
+      "loss": 1.6634,
+      "step": 2114
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9109063060274158e-05,
+      "loss": 1.5028,
+      "step": 2115
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9107552182398037e-05,
+      "loss": 1.6657,
+      "step": 2116
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9106040084343935e-05,
+      "loss": 1.5481,
+      "step": 2117
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.910452676631444e-05,
+      "loss": 1.4505,
+      "step": 2118
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9103012228512284e-05,
+      "loss": 1.5173,
+      "step": 2119
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9101496471140385e-05,
+      "loss": 1.3788,
+      "step": 2120
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9099979494401814e-05,
+      "loss": 1.4649,
+      "step": 2121
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9098461298499805e-05,
+      "loss": 1.4204,
+      "step": 2122
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9096941883637766e-05,
+      "loss": 1.4439,
+      "step": 2123
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.909542125001925e-05,
+      "loss": 1.6615,
+      "step": 2124
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9093899397847985e-05,
+      "loss": 1.4763,
+      "step": 2125
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.909237632732786e-05,
+      "loss": 1.5353,
+      "step": 2126
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.909085203866293e-05,
+      "loss": 1.442,
+      "step": 2127
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9089326532057408e-05,
+      "loss": 1.4782,
+      "step": 2128
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.908779980771567e-05,
+      "loss": 1.4933,
+      "step": 2129
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9086271865842264e-05,
+      "loss": 1.5285,
+      "step": 2130
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9084742706641894e-05,
+      "loss": 1.4736,
+      "step": 2131
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9083212330319427e-05,
+      "loss": 1.4496,
+      "step": 2132
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.908168073707989e-05,
+      "loss": 1.5099,
+      "step": 2133
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9080147927128483e-05,
+      "loss": 1.5235,
+      "step": 2134
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9078613900670556e-05,
+      "loss": 1.4769,
+      "step": 2135
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.907707865791164e-05,
+      "loss": 1.5771,
+      "step": 2136
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.907554219905741e-05,
+      "loss": 1.5116,
+      "step": 2137
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.907400452431371e-05,
+      "loss": 1.5161,
+      "step": 2138
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9072465633886557e-05,
+      "loss": 1.4356,
+      "step": 2139
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9070925527982117e-05,
+      "loss": 1.509,
+      "step": 2140
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9069384206806723e-05,
+      "loss": 1.4681,
+      "step": 2141
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9067841670566883e-05,
+      "loss": 1.5254,
+      "step": 2142
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9066297919469242e-05,
+      "loss": 1.4797,
+      "step": 2143
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9064752953720636e-05,
+      "loss": 1.3996,
+      "step": 2144
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9063206773528042e-05,
+      "loss": 1.4457,
+      "step": 2145
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9061659379098617e-05,
+      "loss": 1.5126,
+      "step": 2146
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9060110770639664e-05,
+      "loss": 1.5925,
+      "step": 2147
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.905856094835866e-05,
+      "loss": 1.5146,
+      "step": 2148
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.905700991246324e-05,
+      "loss": 1.7256,
+      "step": 2149
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.90554576631612e-05,
+      "loss": 1.57,
+      "step": 2150
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.905390420066051e-05,
+      "loss": 1.3722,
+      "step": 2151
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9052349525169292e-05,
+      "loss": 1.3999,
+      "step": 2152
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9050793636895828e-05,
+      "loss": 1.586,
+      "step": 2153
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9049236536048567e-05,
+      "loss": 1.4884,
+      "step": 2154
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9047678222836127e-05,
+      "loss": 1.5898,
+      "step": 2155
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9046118697467278e-05,
+      "loss": 1.569,
+      "step": 2156
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.904455796015095e-05,
+      "loss": 1.5592,
+      "step": 2157
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9042996011096253e-05,
+      "loss": 1.597,
+      "step": 2158
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.904143285051244e-05,
+      "loss": 1.422,
+      "step": 2159
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.903986847860894e-05,
+      "loss": 1.519,
+      "step": 2160
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9038302895595334e-05,
+      "loss": 1.5197,
+      "step": 2161
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.903673610168137e-05,
+      "loss": 1.4655,
+      "step": 2162
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9035168097076965e-05,
+      "loss": 1.3909,
+      "step": 2163
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.903359888199218e-05,
+      "loss": 1.4877,
+      "step": 2164
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.903202845663726e-05,
+      "loss": 1.5123,
+      "step": 2165
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9030456821222593e-05,
+      "loss": 1.4906,
+      "step": 2166
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9028883975958745e-05,
+      "loss": 1.4574,
+      "step": 2167
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9027309921056435e-05,
+      "loss": 1.4571,
+      "step": 2168
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.902573465672654e-05,
+      "loss": 1.4281,
+      "step": 2169
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9024158183180116e-05,
+      "loss": 1.5057,
+      "step": 2170
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.902258050062836e-05,
+      "loss": 1.4514,
+      "step": 2171
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9021001609282646e-05,
+      "loss": 1.5055,
+      "step": 2172
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9019421509354505e-05,
+      "loss": 1.4578,
+      "step": 2173
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9017840201055625e-05,
+      "loss": 1.459,
+      "step": 2174
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9016257684597864e-05,
+      "loss": 1.5059,
+      "step": 2175
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.901467396019324e-05,
+      "loss": 1.57,
+      "step": 2176
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.901308902805393e-05,
+      "loss": 1.4924,
+      "step": 2177
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9011502888392273e-05,
+      "loss": 1.4979,
+      "step": 2178
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.900991554142077e-05,
+      "loss": 1.5321,
+      "step": 2179
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9008326987352093e-05,
+      "loss": 1.493,
+      "step": 2180
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9006737226399056e-05,
+      "loss": 1.5111,
+      "step": 2181
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.900514625877465e-05,
+      "loss": 1.4154,
+      "step": 2182
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.900355408469203e-05,
+      "loss": 1.4934,
+      "step": 2183
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.90019607043645e-05,
+      "loss": 1.5975,
+      "step": 2184
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9000366118005532e-05,
+      "loss": 1.3537,
+      "step": 2185
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.899877032582876e-05,
+      "loss": 1.4583,
+      "step": 2186
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8997173328047984e-05,
+      "loss": 1.4586,
+      "step": 2187
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8995575124877156e-05,
+      "loss": 1.4774,
+      "step": 2188
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8993975716530394e-05,
+      "loss": 1.5025,
+      "step": 2189
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 1.5263344049453735,
+      "eval_runtime": 77.1492,
+      "eval_samples_per_second": 1.672,
+      "eval_steps_per_second": 0.117,
+      "step": 2189
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.899237510322198e-05,
+      "loss": 1.5438,
+      "step": 2190
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.899077328516635e-05,
+      "loss": 1.5181,
+      "step": 2191
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8989170262578112e-05,
+      "loss": 1.4454,
+      "step": 2192
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8987566035672028e-05,
+      "loss": 1.4882,
+      "step": 2193
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8985960604663026e-05,
+      "loss": 1.5073,
+      "step": 2194
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8984353969766187e-05,
+      "loss": 1.4478,
+      "step": 2195
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8982746131196763e-05,
+      "loss": 1.4768,
+      "step": 2196
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8981137089170162e-05,
+      "loss": 1.5116,
+      "step": 2197
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8979526843901952e-05,
+      "loss": 1.4202,
+      "step": 2198
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8977915395607867e-05,
+      "loss": 1.4736,
+      "step": 2199
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.89763027445038e-05,
+      "loss": 1.4489,
+      "step": 2200
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8974688890805804e-05,
+      "loss": 1.4512,
+      "step": 2201
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.897307383473009e-05,
+      "loss": 1.5091,
+      "step": 2202
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.897145757649304e-05,
+      "loss": 1.5512,
+      "step": 2203
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8969840116311192e-05,
+      "loss": 1.406,
+      "step": 2204
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.896822145440124e-05,
+      "loss": 1.4531,
+      "step": 2205
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.896660159098004e-05,
+      "loss": 1.4379,
+      "step": 2206
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.896498052626462e-05,
+      "loss": 1.4783,
+      "step": 2207
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.896335826047216e-05,
+      "loss": 1.4801,
+      "step": 2208
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.896173479381999e-05,
+      "loss": 1.4526,
+      "step": 2209
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8960110126525627e-05,
+      "loss": 1.4952,
+      "step": 2210
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.895848425880673e-05,
+      "loss": 1.5299,
+      "step": 2211
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8956857190881125e-05,
+      "loss": 1.5274,
+      "step": 2212
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8955228922966792e-05,
+      "loss": 1.4415,
+      "step": 2213
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8953599455281884e-05,
+      "loss": 1.5311,
+      "step": 2214
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8951968788044702e-05,
+      "loss": 1.4972,
+      "step": 2215
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8950336921473715e-05,
+      "loss": 1.4966,
+      "step": 2216
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8948703855787553e-05,
+      "loss": 1.4683,
+      "step": 2217
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8947069591205005e-05,
+      "loss": 1.503,
+      "step": 2218
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8945434127945014e-05,
+      "loss": 1.4902,
+      "step": 2219
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8943797466226702e-05,
+      "loss": 1.4835,
+      "step": 2220
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8942159606269332e-05,
+      "loss": 1.535,
+      "step": 2221
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8940520548292332e-05,
+      "loss": 1.491,
+      "step": 2222
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.89388802925153e-05,
+      "loss": 1.5157,
+      "step": 2223
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8937238839157987e-05,
+      "loss": 1.4779,
+      "step": 2224
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8935596188440302e-05,
+      "loss": 1.4902,
+      "step": 2225
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8933952340582323e-05,
+      "loss": 1.4988,
+      "step": 2226
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.893230729580428e-05,
+      "loss": 1.4568,
+      "step": 2227
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8930661054326567e-05,
+      "loss": 1.5163,
+      "step": 2228
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.892901361636974e-05,
+      "loss": 1.4472,
+      "step": 2229
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.892736498215451e-05,
+      "loss": 1.4658,
+      "step": 2230
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8925715151901754e-05,
+      "loss": 1.4281,
+      "step": 2231
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.892406412583251e-05,
+      "loss": 1.415,
+      "step": 2232
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8922411904167966e-05,
+      "loss": 1.464,
+      "step": 2233
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.892075848712948e-05,
+      "loss": 1.5014,
+      "step": 2234
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8919103874938568e-05,
+      "loss": 1.4397,
+      "step": 2235
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.891744806781691e-05,
+      "loss": 1.5508,
+      "step": 2236
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8915791065986335e-05,
+      "loss": 1.4749,
+      "step": 2237
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.891413286966884e-05,
+      "loss": 1.4538,
+      "step": 2238
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.891247347908658e-05,
+      "loss": 1.4405,
+      "step": 2239
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8910812894461878e-05,
+      "loss": 1.3562,
+      "step": 2240
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.89091511160172e-05,
+      "loss": 1.4811,
+      "step": 2241
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8907488143975186e-05,
+      "loss": 1.4286,
+      "step": 2242
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.890582397855863e-05,
+      "loss": 1.4493,
+      "step": 2243
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.890415861999049e-05,
+      "loss": 1.5639,
+      "step": 2244
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.890249206849388e-05,
+      "loss": 1.6375,
+      "step": 2245
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.890082432429208e-05,
+      "loss": 1.4982,
+      "step": 2246
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.889915538760851e-05,
+      "loss": 1.492,
+      "step": 2247
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8897485258666778e-05,
+      "loss": 1.4294,
+      "step": 2248
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8895813937690638e-05,
+      "loss": 1.4617,
+      "step": 2249
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8894141424904e-05,
+      "loss": 1.4763,
+      "step": 2250
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8892467720530937e-05,
+      "loss": 1.3989,
+      "step": 2251
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8890792824795687e-05,
+      "loss": 1.4803,
+      "step": 2252
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8889116737922638e-05,
+      "loss": 1.4734,
+      "step": 2253
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8887439460136342e-05,
+      "loss": 1.5069,
+      "step": 2254
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8885760991661514e-05,
+      "loss": 1.5565,
+      "step": 2255
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8884081332723028e-05,
+      "loss": 1.4707,
+      "step": 2256
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8882400483545913e-05,
+      "loss": 1.5077,
+      "step": 2257
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.888071844435536e-05,
+      "loss": 1.5075,
+      "step": 2258
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8879035215376717e-05,
+      "loss": 1.5185,
+      "step": 2259
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.887735079683549e-05,
+      "loss": 1.6266,
+      "step": 2260
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8875665188957354e-05,
+      "loss": 1.5013,
+      "step": 2261
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.887397839196814e-05,
+      "loss": 1.4362,
+      "step": 2262
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8872290406093827e-05,
+      "loss": 1.4718,
+      "step": 2263
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8870601231560567e-05,
+      "loss": 1.5202,
+      "step": 2264
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8868910868594664e-05,
+      "loss": 1.5168,
+      "step": 2265
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8867219317422588e-05,
+      "loss": 1.5142,
+      "step": 2266
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8865526578270954e-05,
+      "loss": 1.4885,
+      "step": 2267
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8863832651366552e-05,
+      "loss": 1.5441,
+      "step": 2268
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8862137536936328e-05,
+      "loss": 1.4647,
+      "step": 2269
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8860441235207374e-05,
+      "loss": 1.5352,
+      "step": 2270
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8858743746406958e-05,
+      "loss": 1.4568,
+      "step": 2271
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8857045070762497e-05,
+      "loss": 1.4865,
+      "step": 2272
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8855345208501573e-05,
+      "loss": 1.4637,
+      "step": 2273
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8853644159851918e-05,
+      "loss": 1.4348,
+      "step": 2274
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8851941925041436e-05,
+      "loss": 1.4413,
+      "step": 2275
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8850238504298177e-05,
+      "loss": 1.4434,
+      "step": 2276
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.884853389785036e-05,
+      "loss": 1.4822,
+      "step": 2277
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8846828105926356e-05,
+      "loss": 1.4632,
+      "step": 2278
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8845121128754698e-05,
+      "loss": 1.48,
+      "step": 2279
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8843412966564078e-05,
+      "loss": 1.4518,
+      "step": 2280
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.884170361958334e-05,
+      "loss": 1.4334,
+      "step": 2281
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8839993088041502e-05,
+      "loss": 1.5131,
+      "step": 2282
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8838281372167724e-05,
+      "loss": 1.508,
+      "step": 2283
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8836568472191335e-05,
+      "loss": 1.4086,
+      "step": 2284
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.883485438834182e-05,
+      "loss": 1.5021,
+      "step": 2285
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8833139120848817e-05,
+      "loss": 1.5083,
+      "step": 2286
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8831422669942137e-05,
+      "loss": 1.4162,
+      "step": 2287
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8829705035851734e-05,
+      "loss": 1.5494,
+      "step": 2288
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8827986218807724e-05,
+      "loss": 1.425,
+      "step": 2289
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8826266219040396e-05,
+      "loss": 1.5215,
+      "step": 2290
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8824545036780172e-05,
+      "loss": 1.5577,
+      "step": 2291
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8822822672257655e-05,
+      "loss": 1.4923,
+      "step": 2292
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8821099125703595e-05,
+      "loss": 1.4648,
+      "step": 2293
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8819374397348902e-05,
+      "loss": 1.471,
+      "step": 2294
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8817648487424648e-05,
+      "loss": 1.5658,
+      "step": 2295
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8815921396162058e-05,
+      "loss": 1.5348,
+      "step": 2296
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8814193123792518e-05,
+      "loss": 1.4081,
+      "step": 2297
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.881246367054757e-05,
+      "loss": 1.4762,
+      "step": 2298
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8810733036658925e-05,
+      "loss": 1.4399,
+      "step": 2299
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8809001222358432e-05,
+      "loss": 1.4763,
+      "step": 2300
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.880726822787812e-05,
+      "loss": 1.4874,
+      "step": 2301
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.880553405345016e-05,
+      "loss": 1.4951,
+      "step": 2302
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8803798699306887e-05,
+      "loss": 1.4697,
+      "step": 2303
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.880206216568079e-05,
+      "loss": 1.481,
+      "step": 2304
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.880032445280453e-05,
+      "loss": 1.4485,
+      "step": 2305
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.879858556091091e-05,
+      "loss": 1.4749,
+      "step": 2306
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8796845490232894e-05,
+      "loss": 1.5605,
+      "step": 2307
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.879510424100361e-05,
+      "loss": 1.4835,
+      "step": 2308
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8793361813456346e-05,
+      "loss": 1.4344,
+      "step": 2309
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8791618207824532e-05,
+      "loss": 1.4571,
+      "step": 2310
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8789873424341772e-05,
+      "loss": 1.482,
+      "step": 2311
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8788127463241822e-05,
+      "loss": 1.4736,
+      "step": 2312
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8786380324758597e-05,
+      "loss": 1.4807,
+      "step": 2313
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8784632009126167e-05,
+      "loss": 1.4418,
+      "step": 2314
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.878288251657876e-05,
+      "loss": 1.5248,
+      "step": 2315
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8781131847350768e-05,
+      "loss": 1.483,
+      "step": 2316
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.877938000167673e-05,
+      "loss": 1.5474,
+      "step": 2317
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8777626979791347e-05,
+      "loss": 1.4671,
+      "step": 2318
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.877587278192949e-05,
+      "loss": 1.3049,
+      "step": 2319
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.877411740832616e-05,
+      "loss": 1.483,
+      "step": 2320
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.877236085921655e-05,
+      "loss": 1.508,
+      "step": 2321
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8770603134835978e-05,
+      "loss": 1.4854,
+      "step": 2322
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8768844235419942e-05,
+      "loss": 1.4808,
+      "step": 2323
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8767084161204087e-05,
+      "loss": 1.5352,
+      "step": 2324
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8765322912424214e-05,
+      "loss": 1.5628,
+      "step": 2325
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8763560489316292e-05,
+      "loss": 1.494,
+      "step": 2326
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.876179689211643e-05,
+      "loss": 1.509,
+      "step": 2327
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8760032121060918e-05,
+      "loss": 1.5021,
+      "step": 2328
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.875826617638618e-05,
+      "loss": 1.4179,
+      "step": 2329
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8756499058328813e-05,
+      "loss": 1.4302,
+      "step": 2330
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8754730767125564e-05,
+      "loss": 1.4959,
+      "step": 2331
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8752961303013336e-05,
+      "loss": 1.4423,
+      "step": 2332
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8751190666229194e-05,
+      "loss": 1.4564,
+      "step": 2333
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8749418857010358e-05,
+      "loss": 1.4775,
+      "step": 2334
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8747645875594203e-05,
+      "loss": 1.4182,
+      "step": 2335
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.8745871722218265e-05,
+      "loss": 1.3957,
+      "step": 2336
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8744096397120235e-05,
+      "loss": 1.4633,
+      "step": 2337
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.874231990053796e-05,
+      "loss": 1.5512,
+      "step": 2338
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8740542232709444e-05,
+      "loss": 1.3461,
+      "step": 2339
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8738763393872848e-05,
+      "loss": 1.4754,
+      "step": 2340
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8736983384266495e-05,
+      "loss": 1.5448,
+      "step": 2341
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8735202204128854e-05,
+      "loss": 1.4491,
+      "step": 2342
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8733419853698566e-05,
+      "loss": 1.457,
+      "step": 2343
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8731636333214414e-05,
+      "loss": 1.4314,
+      "step": 2344
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8729851642915345e-05,
+      "loss": 1.3979,
+      "step": 2345
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8728065783040464e-05,
+      "loss": 1.4146,
+      "step": 2346
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8726278753829027e-05,
+      "loss": 1.4361,
+      "step": 2347
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.872449055552045e-05,
+      "loss": 1.4989,
+      "step": 2348
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.872270118835431e-05,
+      "loss": 1.5062,
+      "step": 2349
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.872091065257033e-05,
+      "loss": 1.4509,
+      "step": 2350
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8719118948408403e-05,
+      "loss": 1.4631,
+      "step": 2351
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8717326076108566e-05,
+      "loss": 1.5119,
+      "step": 2352
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8715532035911017e-05,
+      "loss": 1.6347,
+      "step": 2353
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8713736828056115e-05,
+      "loss": 1.5526,
+      "step": 2354
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8711940452784368e-05,
+      "loss": 1.6348,
+      "step": 2355
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.871014291033645e-05,
+      "loss": 1.5183,
+      "step": 2356
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8708344200953182e-05,
+      "loss": 1.458,
+      "step": 2357
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8706544324875545e-05,
+      "loss": 1.524,
+      "step": 2358
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8704743282344672e-05,
+      "loss": 1.4479,
+      "step": 2359
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.870294107360186e-05,
+      "loss": 1.5408,
+      "step": 2360
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8701137698888563e-05,
+      "loss": 1.5007,
+      "step": 2361
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8699333158446382e-05,
+      "loss": 1.5132,
+      "step": 2362
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.869752745251708e-05,
+      "loss": 1.4263,
+      "step": 2363
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8695720581342574e-05,
+      "loss": 1.422,
+      "step": 2364
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.869391254516494e-05,
+      "loss": 1.4411,
+      "step": 2365
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8692103344226406e-05,
+      "loss": 1.4594,
+      "step": 2366
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8690292978769364e-05,
+      "loss": 1.4992,
+      "step": 2367
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.868848144903635e-05,
+      "loss": 1.4822,
+      "step": 2368
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8686668755270064e-05,
+      "loss": 1.5402,
+      "step": 2369
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8684854897713363e-05,
+      "loss": 1.5268,
+      "step": 2370
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8683039876609258e-05,
+      "loss": 1.5587,
+      "step": 2371
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.868122369220091e-05,
+      "loss": 1.4214,
+      "step": 2372
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8679406344731645e-05,
+      "loss": 1.5089,
+      "step": 2373
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.867758783444494e-05,
+      "loss": 1.5067,
+      "step": 2374
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8675768161584427e-05,
+      "loss": 2.0616,
+      "step": 2375
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8673947326393903e-05,
+      "loss": 1.5617,
+      "step": 2376
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.86721253291173e-05,
+      "loss": 1.4422,
+      "step": 2377
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.867030216999873e-05,
+      "loss": 1.4727,
+      "step": 2378
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8668477849282443e-05,
+      "loss": 1.4152,
+      "step": 2379
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8666652367212856e-05,
+      "loss": 1.4376,
+      "step": 2380
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8664825724034534e-05,
+      "loss": 1.393,
+      "step": 2381
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8662997919992197e-05,
+      "loss": 1.4864,
+      "step": 2382
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.866116895533073e-05,
+      "loss": 1.4261,
+      "step": 2383
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8659338830295163e-05,
+      "loss": 1.4555,
+      "step": 2384
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8657507545130687e-05,
+      "loss": 1.4939,
+      "step": 2385
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.865567510008265e-05,
+      "loss": 1.4123,
+      "step": 2386
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8653841495396546e-05,
+      "loss": 1.4459,
+      "step": 2387
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8652006731318038e-05,
+      "loss": 1.396,
+      "step": 2388
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 1.5139706134796143,
+      "eval_runtime": 77.2904,
+      "eval_samples_per_second": 1.669,
+      "eval_steps_per_second": 0.116,
+      "step": 2388
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8650170808092936e-05,
+      "loss": 1.4666,
+      "step": 2389
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.86483337259672e-05,
+      "loss": 1.5154,
+      "step": 2390
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.864649548518696e-05,
+      "loss": 1.5163,
+      "step": 2391
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.864465608599849e-05,
+      "loss": 1.4734,
+      "step": 2392
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.864281552864822e-05,
+      "loss": 1.5217,
+      "step": 2393
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8640973813382737e-05,
+      "loss": 1.4694,
+      "step": 2394
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.863913094044879e-05,
+      "loss": 1.4761,
+      "step": 2395
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.863728691009327e-05,
+      "loss": 1.4766,
+      "step": 2396
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.863544172256323e-05,
+      "loss": 1.5695,
+      "step": 2397
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8633595378105882e-05,
+      "loss": 1.5079,
+      "step": 2398
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8631747876968584e-05,
+      "loss": 1.4154,
+      "step": 2399
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.862989921939886e-05,
+      "loss": 1.4488,
+      "step": 2400
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8628049405644376e-05,
+      "loss": 1.4279,
+      "step": 2401
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8626198435952966e-05,
+      "loss": 1.4896,
+      "step": 2402
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8624346310572602e-05,
+      "loss": 1.4315,
+      "step": 2403
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.862249302975143e-05,
+      "loss": 1.4919,
+      "step": 2404
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8620638593737743e-05,
+      "loss": 1.4561,
+      "step": 2405
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8618783002779985e-05,
+      "loss": 1.4622,
+      "step": 2406
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8616926257126754e-05,
+      "loss": 1.4917,
+      "step": 2407
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.861506835702681e-05,
+      "loss": 1.4962,
+      "step": 2408
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.861320930272906e-05,
+      "loss": 1.4164,
+      "step": 2409
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8611349094482578e-05,
+      "loss": 1.4889,
+      "step": 2410
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8609487732536576e-05,
+      "loss": 1.543,
+      "step": 2411
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8607625217140433e-05,
+      "loss": 1.5417,
+      "step": 2412
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8605761548543674e-05,
+      "loss": 1.5304,
+      "step": 2413
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.860389672699599e-05,
+      "loss": 1.4725,
+      "step": 2414
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.860203075274721e-05,
+      "loss": 1.4745,
+      "step": 2415
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8600163626047334e-05,
+      "loss": 1.4656,
+      "step": 2416
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8598295347146503e-05,
+      "loss": 1.5287,
+      "step": 2417
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8596425916295023e-05,
+      "loss": 1.5024,
+      "step": 2418
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8594555333743348e-05,
+      "loss": 1.4527,
+      "step": 2419
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.859268359974209e-05,
+      "loss": 1.4897,
+      "step": 2420
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.859081071454201e-05,
+      "loss": 1.3554,
+      "step": 2421
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.858893667839403e-05,
+      "loss": 1.4004,
+      "step": 2422
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.858706149154922e-05,
+      "loss": 1.5582,
+      "step": 2423
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8585185154258803e-05,
+      "loss": 1.505,
+      "step": 2424
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.858330766677417e-05,
+      "loss": 1.5043,
+      "step": 2425
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.858142902934685e-05,
+      "loss": 1.439,
+      "step": 2426
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.857954924222853e-05,
+      "loss": 1.4927,
+      "step": 2427
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8577668305671056e-05,
+      "loss": 1.5395,
+      "step": 2428
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8575786219926427e-05,
+      "loss": 1.4737,
+      "step": 2429
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.8573902985246794e-05,
+      "loss": 1.6151,
+      "step": 2430
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.857201860188446e-05,
+      "loss": 1.4313,
+      "step": 2431
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.857013307009188e-05,
+      "loss": 1.4509,
+      "step": 2432
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8568246390121673e-05,
+      "loss": 1.3998,
+      "step": 2433
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8566358562226603e-05,
+      "loss": 1.5268,
+      "step": 2434
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8564469586659593e-05,
+      "loss": 1.4369,
+      "step": 2435
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8562579463673713e-05,
+      "loss": 1.4759,
+      "step": 2436
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8560688193522197e-05,
+      "loss": 1.4092,
+      "step": 2437
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.855879577645842e-05,
+      "loss": 1.43,
+      "step": 2438
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.855690221273592e-05,
+      "loss": 1.5143,
+      "step": 2439
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8555007502608385e-05,
+      "loss": 1.5482,
+      "step": 2440
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.855311164632966e-05,
+      "loss": 1.4406,
+      "step": 2441
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8551214644153738e-05,
+      "loss": 1.4551,
+      "step": 2442
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8549316496334774e-05,
+      "loss": 1.4663,
+      "step": 2443
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8547417203127064e-05,
+      "loss": 1.4732,
+      "step": 2444
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8545516764785067e-05,
+      "loss": 1.5378,
+      "step": 2445
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8543615181563397e-05,
+      "loss": 1.4448,
+      "step": 2446
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.854171245371681e-05,
+      "loss": 1.2861,
+      "step": 2447
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8539808581500228e-05,
+      "loss": 1.4405,
+      "step": 2448
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8537903565168716e-05,
+      "loss": 1.4876,
+      "step": 2449
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8535997404977506e-05,
+      "loss": 1.4119,
+      "step": 2450
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8534090101181967e-05,
+      "loss": 1.4451,
+      "step": 2451
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8532181654037632e-05,
+      "loss": 1.4035,
+      "step": 2452
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.853027206380018e-05,
+      "loss": 1.4624,
+      "step": 2453
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.852836133072545e-05,
+      "loss": 1.4935,
+      "step": 2454
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8526449455069434e-05,
+      "loss": 1.4834,
+      "step": 2455
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8524536437088266e-05,
+      "loss": 1.498,
+      "step": 2456
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8522622277038247e-05,
+      "loss": 1.3417,
+      "step": 2457
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.852070697517583e-05,
+      "loss": 1.4455,
+      "step": 2458
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8518790531757607e-05,
+      "loss": 1.4119,
+      "step": 2459
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8516872947040337e-05,
+      "loss": 1.558,
+      "step": 2460
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8514954221280926e-05,
+      "loss": 1.4862,
+      "step": 2461
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8513034354736433e-05,
+      "loss": 1.6806,
+      "step": 2462
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8511113347664072e-05,
+      "loss": 1.438,
+      "step": 2463
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.850919120032121e-05,
+      "loss": 1.4975,
+      "step": 2464
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.850726791296536e-05,
+      "loss": 1.4409,
+      "step": 2465
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.85053434858542e-05,
+      "loss": 1.4769,
+      "step": 2466
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8503417919245548e-05,
+      "loss": 1.5823,
+      "step": 2467
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8501491213397386e-05,
+      "loss": 1.4045,
+      "step": 2468
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8499563368567837e-05,
+      "loss": 1.4828,
+      "step": 2469
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.849763438501519e-05,
+      "loss": 1.4684,
+      "step": 2470
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8495704262997873e-05,
+      "loss": 1.4577,
+      "step": 2471
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8493773002774475e-05,
+      "loss": 1.4759,
+      "step": 2472
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8491840604603736e-05,
+      "loss": 1.4259,
+      "step": 2473
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8489907068744543e-05,
+      "loss": 1.4717,
+      "step": 2474
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.848797239545595e-05,
+      "loss": 1.4247,
+      "step": 2475
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8486036584997143e-05,
+      "loss": 1.4954,
+      "step": 2476
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8484099637627472e-05,
+      "loss": 1.4926,
+      "step": 2477
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8482161553606447e-05,
+      "loss": 1.487,
+      "step": 2478
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8480222333193715e-05,
+      "loss": 1.4553,
+      "step": 2479
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.847828197664908e-05,
+      "loss": 1.4862,
+      "step": 2480
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8476340484232508e-05,
+      "loss": 1.4867,
+      "step": 2481
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8474397856204097e-05,
+      "loss": 1.4736,
+      "step": 2482
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8472454092824122e-05,
+      "loss": 1.4254,
+      "step": 2483
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8470509194352987e-05,
+      "loss": 1.4455,
+      "step": 2484
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8468563161051263e-05,
+      "loss": 1.4402,
+      "step": 2485
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.846661599317967e-05,
+      "loss": 1.4611,
+      "step": 2486
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8464667690999073e-05,
+      "loss": 1.4814,
+      "step": 2487
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8462718254770503e-05,
+      "loss": 1.4559,
+      "step": 2488
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8460767684755127e-05,
+      "loss": 1.4667,
+      "step": 2489
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8458815981214274e-05,
+      "loss": 1.4427,
+      "step": 2490
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.845686314440942e-05,
+      "loss": 1.4055,
+      "step": 2491
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8454909174602198e-05,
+      "loss": 1.5563,
+      "step": 2492
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.845295407205439e-05,
+      "loss": 1.4631,
+      "step": 2493
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8450997837027925e-05,
+      "loss": 1.4887,
+      "step": 2494
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8449040469784895e-05,
+      "loss": 1.5212,
+      "step": 2495
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.844708197058753e-05,
+      "loss": 1.5582,
+      "step": 2496
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.844512233969822e-05,
+      "loss": 1.4599,
+      "step": 2497
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.844316157737951e-05,
+      "loss": 1.4546,
+      "step": 2498
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.844119968389409e-05,
+      "loss": 1.4631,
+      "step": 2499
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.84392366595048e-05,
+      "loss": 1.4817,
+      "step": 2500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8437272504474636e-05,
+      "loss": 1.4214,
+      "step": 2501
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8435307219066744e-05,
+      "loss": 1.471,
+      "step": 2502
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8433340803544424e-05,
+      "loss": 1.4693,
+      "step": 2503
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8431373258171127e-05,
+      "loss": 1.6297,
+      "step": 2504
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.842940458321045e-05,
+      "loss": 1.5916,
+      "step": 2505
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.842743477892614e-05,
+      "loss": 1.4754,
+      "step": 2506
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.842546384558211e-05,
+      "loss": 1.4828,
+      "step": 2507
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.842349178344241e-05,
+      "loss": 1.4239,
+      "step": 2508
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8421518592771243e-05,
+      "loss": 1.5028,
+      "step": 2509
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.841954427383297e-05,
+      "loss": 1.4848,
+      "step": 2510
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8417568826892097e-05,
+      "loss": 1.4617,
+      "step": 2511
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8415592252213284e-05,
+      "loss": 1.4266,
+      "step": 2512
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8413614550061344e-05,
+      "loss": 1.4321,
+      "step": 2513
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.841163572070123e-05,
+      "loss": 1.4693,
+      "step": 2514
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.840965576439806e-05,
+      "loss": 1.4257,
+      "step": 2515
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.84076746814171e-05,
+      "loss": 1.4488,
+      "step": 2516
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.840569247202376e-05,
+      "loss": 1.5348,
+      "step": 2517
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8403709136483604e-05,
+      "loss": 1.477,
+      "step": 2518
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.840172467506235e-05,
+      "loss": 1.443,
+      "step": 2519
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.839973908802587e-05,
+      "loss": 1.4772,
+      "step": 2520
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.839775237564017e-05,
+      "loss": 1.4443,
+      "step": 2521
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.839576453817143e-05,
+      "loss": 1.4063,
+      "step": 2522
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8393775575885963e-05,
+      "loss": 1.5178,
+      "step": 2523
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8391785489050235e-05,
+      "loss": 1.4337,
+      "step": 2524
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8389794277930875e-05,
+      "loss": 1.4512,
+      "step": 2525
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.838780194279465e-05,
+      "loss": 1.5578,
+      "step": 2526
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.8385808483908484e-05,
+      "loss": 1.4606,
+      "step": 2527
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8383813901539444e-05,
+      "loss": 1.4957,
+      "step": 2528
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.838181819595476e-05,
+      "loss": 1.4914,
+      "step": 2529
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8379821367421802e-05,
+      "loss": 1.4785,
+      "step": 2530
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.837782341620809e-05,
+      "loss": 1.4052,
+      "step": 2531
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8375824342581303e-05,
+      "loss": 1.4807,
+      "step": 2532
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8373824146809264e-05,
+      "loss": 1.4313,
+      "step": 2533
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8371822829159945e-05,
+      "loss": 1.4374,
+      "step": 2534
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.836982038990148e-05,
+      "loss": 1.4482,
+      "step": 2535
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8367816829302133e-05,
+      "loss": 1.5231,
+      "step": 2536
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8365812147630337e-05,
+      "loss": 1.4101,
+      "step": 2537
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8363806345154666e-05,
+      "loss": 1.4372,
+      "step": 2538
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8361799422143845e-05,
+      "loss": 1.4808,
+      "step": 2539
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.835979137886675e-05,
+      "loss": 1.3515,
+      "step": 2540
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8357782215592412e-05,
+      "loss": 1.5284,
+      "step": 2541
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.835577193259e-05,
+      "loss": 1.4222,
+      "step": 2542
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8353760530128846e-05,
+      "loss": 1.4522,
+      "step": 2543
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.835174800847842e-05,
+      "loss": 1.4487,
+      "step": 2544
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8349734367908355e-05,
+      "loss": 1.3688,
+      "step": 2545
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8347719608688422e-05,
+      "loss": 1.483,
+      "step": 2546
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8345703731088554e-05,
+      "loss": 1.4733,
+      "step": 2547
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8343686735378816e-05,
+      "loss": 1.4594,
+      "step": 2548
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.834166862182944e-05,
+      "loss": 1.4648,
+      "step": 2549
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8339649390710807e-05,
+      "loss": 1.5249,
+      "step": 2550
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8337629042293434e-05,
+      "loss": 1.4744,
+      "step": 2551
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8335607576847994e-05,
+      "loss": 1.5094,
+      "step": 2552
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8333584994645318e-05,
+      "loss": 1.5122,
+      "step": 2553
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8331561295956374e-05,
+      "loss": 1.537,
+      "step": 2554
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.832953648105229e-05,
+      "loss": 1.4691,
+      "step": 2555
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8327510550204343e-05,
+      "loss": 1.4969,
+      "step": 2556
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8325483503683944e-05,
+      "loss": 1.4596,
+      "step": 2557
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8323455341762675e-05,
+      "loss": 1.435,
+      "step": 2558
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8321426064712255e-05,
+      "loss": 1.4883,
+      "step": 2559
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8319395672804554e-05,
+      "loss": 1.3758,
+      "step": 2560
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8317364166311592e-05,
+      "loss": 1.4352,
+      "step": 2561
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.831533154550554e-05,
+      "loss": 1.5072,
+      "step": 2562
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8313297810658718e-05,
+      "loss": 1.4476,
+      "step": 2563
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.831126296204359e-05,
+      "loss": 1.507,
+      "step": 2564
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8309226999932777e-05,
+      "loss": 1.4677,
+      "step": 2565
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8307189924599044e-05,
+      "loss": 1.4537,
+      "step": 2566
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8305151736315307e-05,
+      "loss": 1.4584,
+      "step": 2567
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8303112435354634e-05,
+      "loss": 1.4944,
+      "step": 2568
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8301072021990232e-05,
+      "loss": 1.4379,
+      "step": 2569
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8299030496495475e-05,
+      "loss": 1.539,
+      "step": 2570
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.829698785914386e-05,
+      "loss": 1.4421,
+      "step": 2571
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8294944110209063e-05,
+      "loss": 1.4693,
+      "step": 2572
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8292899249964887e-05,
+      "loss": 1.3239,
+      "step": 2573
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8290853278685284e-05,
+      "loss": 1.472,
+      "step": 2574
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8288806196644374e-05,
+      "loss": 1.536,
+      "step": 2575
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8286758004116405e-05,
+      "loss": 1.4644,
+      "step": 2576
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8284708701375783e-05,
+      "loss": 1.4537,
+      "step": 2577
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8282658288697067e-05,
+      "loss": 1.4359,
+      "step": 2578
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8280606766354955e-05,
+      "loss": 1.4842,
+      "step": 2579
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8278554134624302e-05,
+      "loss": 1.4664,
+      "step": 2580
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8276500393780105e-05,
+      "loss": 1.4632,
+      "step": 2581
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.827444554409751e-05,
+      "loss": 1.4996,
+      "step": 2582
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.827238958585182e-05,
+      "loss": 1.3399,
+      "step": 2583
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8270332519318476e-05,
+      "loss": 1.3928,
+      "step": 2584
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8268274344773074e-05,
+      "loss": 1.4993,
+      "step": 2585
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8266215062491356e-05,
+      "loss": 1.5398,
+      "step": 2586
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.826415467274921e-05,
+      "loss": 1.4353,
+      "step": 2587
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 1.5104212760925293,
+      "eval_runtime": 77.3448,
+      "eval_samples_per_second": 1.668,
+      "eval_steps_per_second": 0.116,
+      "step": 2587
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8262093175822685e-05,
+      "loss": 1.4889,
+      "step": 2588
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.826003057198796e-05,
+      "loss": 1.4912,
+      "step": 2589
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.825796686152137e-05,
+      "loss": 1.4884,
+      "step": 2590
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8255902044699404e-05,
+      "loss": 1.478,
+      "step": 2591
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8253836121798692e-05,
+      "loss": 1.6549,
+      "step": 2592
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8251769093096018e-05,
+      "loss": 1.5146,
+      "step": 2593
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8249700958868306e-05,
+      "loss": 1.4797,
+      "step": 2594
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8247631719392634e-05,
+      "loss": 1.4655,
+      "step": 2595
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.824556137494623e-05,
+      "loss": 1.4517,
+      "step": 2596
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8243489925806458e-05,
+      "loss": 1.4936,
+      "step": 2597
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8241417372250852e-05,
+      "loss": 1.4614,
+      "step": 2598
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8239343714557073e-05,
+      "loss": 1.4475,
+      "step": 2599
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8237268953002937e-05,
+      "loss": 1.5435,
+      "step": 2600
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8235193087866413e-05,
+      "loss": 1.4616,
+      "step": 2601
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8233116119425612e-05,
+      "loss": 1.4578,
+      "step": 2602
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8231038047958795e-05,
+      "loss": 1.434,
+      "step": 2603
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8228958873744368e-05,
+      "loss": 1.5335,
+      "step": 2604
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8226878597060887e-05,
+      "loss": 1.415,
+      "step": 2605
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8224797218187058e-05,
+      "loss": 1.6165,
+      "step": 2606
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.822271473740173e-05,
+      "loss": 1.372,
+      "step": 2607
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8220631154983907e-05,
+      "loss": 1.4855,
+      "step": 2608
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8218546471212727e-05,
+      "loss": 1.4093,
+      "step": 2609
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.821646068636749e-05,
+      "loss": 1.4852,
+      "step": 2610
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8214373800727637e-05,
+      "loss": 1.6959,
+      "step": 2611
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8212285814572755e-05,
+      "loss": 1.3874,
+      "step": 2612
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.821019672818258e-05,
+      "loss": 1.4425,
+      "step": 2613
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8208106541837e-05,
+      "loss": 1.6097,
+      "step": 2614
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8206015255816045e-05,
+      "loss": 1.5675,
+      "step": 2615
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.820392287039989e-05,
+      "loss": 1.4039,
+      "step": 2616
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8201829385868866e-05,
+      "loss": 1.4872,
+      "step": 2617
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.819973480250344e-05,
+      "loss": 1.4645,
+      "step": 2618
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8197639120584237e-05,
+      "loss": 1.4923,
+      "step": 2619
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8195542340392025e-05,
+      "loss": 1.4294,
+      "step": 2620
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8193444462207717e-05,
+      "loss": 1.4315,
+      "step": 2621
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.8191345486312372e-05,
+      "loss": 1.3972,
+      "step": 2622
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8189245412987205e-05,
+      "loss": 1.462,
+      "step": 2623
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8187144242513567e-05,
+      "loss": 1.4738,
+      "step": 2624
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.818504197517296e-05,
+      "loss": 1.4935,
+      "step": 2625
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8182938611247038e-05,
+      "loss": 1.4231,
+      "step": 2626
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.81808341510176e-05,
+      "loss": 1.4031,
+      "step": 2627
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.817872859476658e-05,
+      "loss": 1.5822,
+      "step": 2628
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8176621942776075e-05,
+      "loss": 1.432,
+      "step": 2629
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8174514195328324e-05,
+      "loss": 1.5565,
+      "step": 2630
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8172405352705704e-05,
+      "loss": 1.4895,
+      "step": 2631
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8170295415190754e-05,
+      "loss": 1.4032,
+      "step": 2632
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8168184383066144e-05,
+      "loss": 1.4253,
+      "step": 2633
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8166072256614702e-05,
+      "loss": 1.5162,
+      "step": 2634
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.81639590361194e-05,
+      "loss": 1.4609,
+      "step": 2635
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.816184472186335e-05,
+      "loss": 1.4351,
+      "step": 2636
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8159729314129823e-05,
+      "loss": 1.4738,
+      "step": 2637
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8157612813202224e-05,
+      "loss": 1.5986,
+      "step": 2638
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.815549521936411e-05,
+      "loss": 1.4633,
+      "step": 2639
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8153376532899184e-05,
+      "loss": 1.4404,
+      "step": 2640
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.81512567540913e-05,
+      "loss": 1.4256,
+      "step": 2641
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.814913588322445e-05,
+      "loss": 1.4054,
+      "step": 2642
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8147013920582773e-05,
+      "loss": 1.4134,
+      "step": 2643
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8144890866450567e-05,
+      "loss": 1.4098,
+      "step": 2644
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.814276672111226e-05,
+      "loss": 1.4593,
+      "step": 2645
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.814064148485243e-05,
+      "loss": 1.4679,
+      "step": 2646
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8138515157955812e-05,
+      "loss": 1.5234,
+      "step": 2647
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8136387740707272e-05,
+      "loss": 1.4568,
+      "step": 2648
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8134259233391836e-05,
+      "loss": 1.5057,
+      "step": 2649
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8132129636294665e-05,
+      "loss": 1.3643,
+      "step": 2650
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.812999894970107e-05,
+      "loss": 1.4766,
+      "step": 2651
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.812786717389651e-05,
+      "loss": 1.5701,
+      "step": 2652
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8125734309166587e-05,
+      "loss": 1.5341,
+      "step": 2653
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.812360035579705e-05,
+      "loss": 1.3596,
+      "step": 2654
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8121465314073793e-05,
+      "loss": 1.4847,
+      "step": 2655
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.811932918428286e-05,
+      "loss": 1.4664,
+      "step": 2656
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8117191966710438e-05,
+      "loss": 1.4673,
+      "step": 2657
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8115053661642856e-05,
+      "loss": 1.4712,
+      "step": 2658
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.811291426936659e-05,
+      "loss": 1.4458,
+      "step": 2659
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.811077379016827e-05,
+      "loss": 1.4667,
+      "step": 2660
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.810863222433466e-05,
+      "loss": 1.4695,
+      "step": 2661
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8106489572152677e-05,
+      "loss": 1.5263,
+      "step": 2662
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8104345833909376e-05,
+      "loss": 1.4148,
+      "step": 2663
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8102201009891976e-05,
+      "loss": 1.3955,
+      "step": 2664
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8100055100387814e-05,
+      "loss": 1.6056,
+      "step": 2665
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8097908105684396e-05,
+      "loss": 1.5394,
+      "step": 2666
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.809576002606936e-05,
+      "loss": 1.479,
+      "step": 2667
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8093610861830494e-05,
+      "loss": 1.4492,
+      "step": 2668
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8091460613255728e-05,
+      "loss": 1.4172,
+      "step": 2669
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.808930928063315e-05,
+      "loss": 1.4498,
+      "step": 2670
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.808715686425097e-05,
+      "loss": 1.6203,
+      "step": 2671
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8085003364397567e-05,
+      "loss": 1.435,
+      "step": 2672
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.808284878136145e-05,
+      "loss": 1.4489,
+      "step": 2673
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8080693115431277e-05,
+      "loss": 1.4416,
+      "step": 2674
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8078536366895853e-05,
+      "loss": 1.422,
+      "step": 2675
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8076378536044128e-05,
+      "loss": 1.4039,
+      "step": 2676
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8074219623165195e-05,
+      "loss": 1.4705,
+      "step": 2677
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.807205962854829e-05,
+      "loss": 1.5031,
+      "step": 2678
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8069898552482806e-05,
+      "loss": 1.4812,
+      "step": 2679
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.806773639525826e-05,
+      "loss": 1.4129,
+      "step": 2680
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8065573157164332e-05,
+      "loss": 1.4559,
+      "step": 2681
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.806340883849084e-05,
+      "loss": 1.3878,
+      "step": 2682
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8061243439527744e-05,
+      "loss": 1.4752,
+      "step": 2683
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8059076960565156e-05,
+      "loss": 1.4089,
+      "step": 2684
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8056909401893326e-05,
+      "loss": 1.4736,
+      "step": 2685
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8054740763802654e-05,
+      "loss": 1.3796,
+      "step": 2686
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.805257104658368e-05,
+      "loss": 1.4412,
+      "step": 2687
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8050400250527086e-05,
+      "loss": 1.4639,
+      "step": 2688
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.804822837592371e-05,
+      "loss": 1.4076,
+      "step": 2689
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.804605542306452e-05,
+      "loss": 1.4627,
+      "step": 2690
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8043881392240642e-05,
+      "loss": 1.5249,
+      "step": 2691
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.804170628374334e-05,
+      "loss": 1.4312,
+      "step": 2692
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.803953009786402e-05,
+      "loss": 1.5676,
+      "step": 2693
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8037352834894233e-05,
+      "loss": 1.5072,
+      "step": 2694
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.803517449512568e-05,
+      "loss": 1.5662,
+      "step": 2695
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8032995078850207e-05,
+      "loss": 1.4771,
+      "step": 2696
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.803081458635979e-05,
+      "loss": 1.412,
+      "step": 2697
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8028633017946564e-05,
+      "loss": 1.497,
+      "step": 2698
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8026450373902804e-05,
+      "loss": 1.4528,
+      "step": 2699
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.802426665452093e-05,
+      "loss": 1.4785,
+      "step": 2700
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8022081860093493e-05,
+      "loss": 1.6411,
+      "step": 2701
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8019895990913213e-05,
+      "loss": 1.4459,
+      "step": 2702
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8017709047272932e-05,
+      "loss": 1.4418,
+      "step": 2703
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.801552102946565e-05,
+      "loss": 1.4409,
+      "step": 2704
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8013331937784498e-05,
+      "loss": 1.4651,
+      "step": 2705
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8011141772522766e-05,
+      "loss": 1.4161,
+      "step": 2706
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8008950533973876e-05,
+      "loss": 1.458,
+      "step": 2707
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8006758222431398e-05,
+      "loss": 1.4477,
+      "step": 2708
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.800456483818904e-05,
+      "loss": 1.4433,
+      "step": 2709
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8002370381540672e-05,
+      "loss": 1.5005,
+      "step": 2710
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.8000174852780282e-05,
+      "loss": 1.4945,
+      "step": 2711
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.7997978252202023e-05,
+      "loss": 1.4298,
+      "step": 2712
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.7995780580100177e-05,
+      "loss": 1.3812,
+      "step": 2713
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.799358183676918e-05,
+      "loss": 1.5559,
+      "step": 2714
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.79913820225036e-05,
+      "loss": 1.4316,
+      "step": 2715
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.798918113759817e-05,
+      "loss": 1.419,
+      "step": 2716
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.7986979182347733e-05,
+      "loss": 1.43,
+      "step": 2717
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.798477615704731e-05,
+      "loss": 1.4964,
+      "step": 2718
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7982572061992044e-05,
+      "loss": 1.4417,
+      "step": 2719
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7980366897477227e-05,
+      "loss": 1.4562,
+      "step": 2720
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7978160663798295e-05,
+      "loss": 1.4386,
+      "step": 2721
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7975953361250824e-05,
+      "loss": 1.445,
+      "step": 2722
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.797374499013054e-05,
+      "loss": 1.4549,
+      "step": 2723
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7971535550733305e-05,
+      "loss": 1.5205,
+      "step": 2724
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.796932504335513e-05,
+      "loss": 1.4632,
+      "step": 2725
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.796711346829216e-05,
+      "loss": 1.3767,
+      "step": 2726
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.79649008258407e-05,
+      "loss": 1.4098,
+      "step": 2727
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7962687116297178e-05,
+      "loss": 1.4675,
+      "step": 2728
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7960472339958176e-05,
+      "loss": 1.6048,
+      "step": 2729
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7958256497120418e-05,
+      "loss": 1.5432,
+      "step": 2730
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7956039588080774e-05,
+      "loss": 1.5453,
+      "step": 2731
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7953821613136246e-05,
+      "loss": 1.4246,
+      "step": 2732
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7951602572583988e-05,
+      "loss": 1.3987,
+      "step": 2733
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7949382466721295e-05,
+      "loss": 1.4943,
+      "step": 2734
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7947161295845606e-05,
+      "loss": 1.4385,
+      "step": 2735
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7944939060254497e-05,
+      "loss": 1.4165,
+      "step": 2736
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7942715760245692e-05,
+      "loss": 1.4004,
+      "step": 2737
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7940491396117063e-05,
+      "loss": 1.4577,
+      "step": 2738
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7938265968166602e-05,
+      "loss": 1.4649,
+      "step": 2739
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7936039476692473e-05,
+      "loss": 1.5273,
+      "step": 2740
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7933811921992962e-05,
+      "loss": 1.4325,
+      "step": 2741
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.793158330436651e-05,
+      "loss": 1.4388,
+      "step": 2742
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7929353624111687e-05,
+      "loss": 1.4124,
+      "step": 2743
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7927122881527222e-05,
+      "loss": 1.5449,
+      "step": 2744
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7924891076911964e-05,
+      "loss": 1.4728,
+      "step": 2745
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7922658210564932e-05,
+      "loss": 1.5103,
+      "step": 2746
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.792042428278526e-05,
+      "loss": 1.5975,
+      "step": 2747
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7918189293872247e-05,
+      "loss": 1.567,
+      "step": 2748
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7915953244125318e-05,
+      "loss": 1.57,
+      "step": 2749
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.791371613384405e-05,
+      "loss": 1.4372,
+      "step": 2750
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7911477963328155e-05,
+      "loss": 1.4053,
+      "step": 2751
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7909238732877492e-05,
+      "loss": 1.3782,
+      "step": 2752
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.790699844279206e-05,
+      "loss": 1.4401,
+      "step": 2753
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7904757093372e-05,
+      "loss": 1.3749,
+      "step": 2754
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.79025146849176e-05,
+      "loss": 1.4909,
+      "step": 2755
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7900271217729273e-05,
+      "loss": 1.4103,
+      "step": 2756
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7898026692107597e-05,
+      "loss": 1.4986,
+      "step": 2757
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7895781108353278e-05,
+      "loss": 1.5449,
+      "step": 2758
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.789353446676717e-05,
+      "loss": 1.5583,
+      "step": 2759
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7891286767650254e-05,
+      "loss": 1.5342,
+      "step": 2760
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7889038011303673e-05,
+      "loss": 1.4233,
+      "step": 2761
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7886788198028707e-05,
+      "loss": 1.3433,
+      "step": 2762
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.788453732812676e-05,
+      "loss": 1.4451,
+      "step": 2763
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.78822854018994e-05,
+      "loss": 1.4506,
+      "step": 2764
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7880032419648324e-05,
+      "loss": 1.4386,
+      "step": 2765
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7877778381675374e-05,
+      "loss": 1.4512,
+      "step": 2766
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7875523288282535e-05,
+      "loss": 1.4319,
+      "step": 2767
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7873267139771932e-05,
+      "loss": 1.4753,
+      "step": 2768
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.787100993644583e-05,
+      "loss": 1.4044,
+      "step": 2769
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.786875167860663e-05,
+      "loss": 1.4382,
+      "step": 2770
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.786649236655689e-05,
+      "loss": 1.3977,
+      "step": 2771
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7864232000599298e-05,
+      "loss": 1.4473,
+      "step": 2772
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7861970581036678e-05,
+      "loss": 1.4666,
+      "step": 2773
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.785970810817201e-05,
+      "loss": 1.5299,
+      "step": 2774
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7857444582308402e-05,
+      "loss": 1.3974,
+      "step": 2775
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7855180003749114e-05,
+      "loss": 1.4603,
+      "step": 2776
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7852914372797535e-05,
+      "loss": 1.463,
+      "step": 2777
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7850647689757205e-05,
+      "loss": 1.5317,
+      "step": 2778
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.78483799549318e-05,
+      "loss": 1.4669,
+      "step": 2779
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.784611116862514e-05,
+      "loss": 1.5293,
+      "step": 2780
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7843841331141177e-05,
+      "loss": 1.5802,
+      "step": 2781
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7841570442784024e-05,
+      "loss": 1.4486,
+      "step": 2782
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7839298503857913e-05,
+      "loss": 1.4674,
+      "step": 2783
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7837025514667224e-05,
+      "loss": 1.4662,
+      "step": 2784
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7834751475516484e-05,
+      "loss": 1.3555,
+      "step": 2785
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.783247638671035e-05,
+      "loss": 1.4307,
+      "step": 2786
+    },
+    {
+      "epoch": 0.29,
+      "eval_loss": 1.500258445739746,
+      "eval_runtime": 77.406,
+      "eval_samples_per_second": 1.667,
+      "eval_steps_per_second": 0.116,
+      "step": 2786
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7830200248553633e-05,
+      "loss": 1.4596,
+      "step": 2787
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7827923061351273e-05,
+      "loss": 1.6118,
+      "step": 2788
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7825644825408353e-05,
+      "loss": 1.4782,
+      "step": 2789
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.78233655410301e-05,
+      "loss": 1.4681,
+      "step": 2790
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.782108520852188e-05,
+      "loss": 1.6281,
+      "step": 2791
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.78188038281892e-05,
+      "loss": 1.4446,
+      "step": 2792
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7816521400337704e-05,
+      "loss": 1.479,
+      "step": 2793
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7814237925273178e-05,
+      "loss": 1.5457,
+      "step": 2794
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7811953403301547e-05,
+      "loss": 1.3928,
+      "step": 2795
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.780966783472889e-05,
+      "loss": 1.4321,
+      "step": 2796
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.78073812198614e-05,
+      "loss": 1.5734,
+      "step": 2797
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7805093559005432e-05,
+      "loss": 1.4356,
+      "step": 2798
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7802804852467473e-05,
+      "loss": 1.4656,
+      "step": 2799
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.780051510055415e-05,
+      "loss": 1.4052,
+      "step": 2800
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7798224303572228e-05,
+      "loss": 1.4758,
+      "step": 2801
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7795932461828622e-05,
+      "loss": 1.4326,
+      "step": 2802
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7793639575630376e-05,
+      "loss": 1.3409,
+      "step": 2803
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7791345645284678e-05,
+      "loss": 1.4175,
+      "step": 2804
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7789050671098853e-05,
+      "loss": 1.383,
+      "step": 2805
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7786754653380373e-05,
+      "loss": 1.5162,
+      "step": 2806
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7784457592436847e-05,
+      "loss": 1.4899,
+      "step": 2807
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7782159488576016e-05,
+      "loss": 1.4874,
+      "step": 2808
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.777986034210577e-05,
+      "loss": 1.4363,
+      "step": 2809
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7777560153334134e-05,
+      "loss": 1.4223,
+      "step": 2810
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7775258922569273e-05,
+      "loss": 1.4792,
+      "step": 2811
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.77729566501195e-05,
+      "loss": 1.4475,
+      "step": 2812
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.7770653336293253e-05,
+      "loss": 1.5157,
+      "step": 2813
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.776834898139912e-05,
+      "loss": 1.389,
+      "step": 2814
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7766043585745824e-05,
+      "loss": 1.4358,
+      "step": 2815
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.776373714964223e-05,
+      "loss": 1.4297,
+      "step": 2816
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7761429673397344e-05,
+      "loss": 1.4102,
+      "step": 2817
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.77591211573203e-05,
+      "loss": 1.4572,
+      "step": 2818
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.775681160172039e-05,
+      "loss": 1.448,
+      "step": 2819
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7754501006907025e-05,
+      "loss": 1.4426,
+      "step": 2820
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7752189373189773e-05,
+      "loss": 1.4566,
+      "step": 2821
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7749876700878332e-05,
+      "loss": 1.4267,
+      "step": 2822
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7747562990282538e-05,
+      "loss": 1.3951,
+      "step": 2823
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7745248241712373e-05,
+      "loss": 1.4668,
+      "step": 2824
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.774293245547795e-05,
+      "loss": 1.4924,
+      "step": 2825
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7740615631889525e-05,
+      "loss": 1.5258,
+      "step": 2826
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7738297771257496e-05,
+      "loss": 1.4323,
+      "step": 2827
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7735978873892393e-05,
+      "loss": 1.439,
+      "step": 2828
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.773365894010489e-05,
+      "loss": 1.5658,
+      "step": 2829
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.77313379702058e-05,
+      "loss": 1.4607,
+      "step": 2830
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7729015964506073e-05,
+      "loss": 1.5544,
+      "step": 2831
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7726692923316798e-05,
+      "loss": 1.4232,
+      "step": 2832
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.77243688469492e-05,
+      "loss": 1.415,
+      "step": 2833
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.772204373571465e-05,
+      "loss": 1.4658,
+      "step": 2834
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7719717589924647e-05,
+      "loss": 1.4678,
+      "step": 2835
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.771739040989084e-05,
+      "loss": 1.3021,
+      "step": 2836
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.771506219592501e-05,
+      "loss": 1.4045,
+      "step": 2837
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7712732948339085e-05,
+      "loss": 1.4593,
+      "step": 2838
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7710402667445108e-05,
+      "loss": 1.4175,
+      "step": 2839
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.770807135355529e-05,
+      "loss": 1.3939,
+      "step": 2840
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7705739006981963e-05,
+      "loss": 1.6089,
+      "step": 2841
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.77034056280376e-05,
+      "loss": 1.3952,
+      "step": 2842
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7701071217034818e-05,
+      "loss": 1.5086,
+      "step": 2843
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7698735774286362e-05,
+      "loss": 1.4706,
+      "step": 2844
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7696399300105128e-05,
+      "loss": 1.4286,
+      "step": 2845
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7694061794804137e-05,
+      "loss": 1.4418,
+      "step": 2846
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7691723258696562e-05,
+      "loss": 1.516,
+      "step": 2847
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.76893836920957e-05,
+      "loss": 1.4354,
+      "step": 2848
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7687043095314998e-05,
+      "loss": 1.7503,
+      "step": 2849
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7684701468668032e-05,
+      "loss": 1.5985,
+      "step": 2850
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.768235881246852e-05,
+      "loss": 1.442,
+      "step": 2851
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7680015127030322e-05,
+      "loss": 1.4924,
+      "step": 2852
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7677670412667425e-05,
+      "loss": 1.4461,
+      "step": 2853
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7675324669693968e-05,
+      "loss": 1.3823,
+      "step": 2854
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.767297789842421e-05,
+      "loss": 1.452,
+      "step": 2855
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.767063009917257e-05,
+      "loss": 1.46,
+      "step": 2856
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7668281272253587e-05,
+      "loss": 1.4129,
+      "step": 2857
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.766593141798194e-05,
+      "loss": 1.4057,
+      "step": 2858
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7663580536672456e-05,
+      "loss": 1.5162,
+      "step": 2859
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7661228628640087e-05,
+      "loss": 1.5177,
+      "step": 2860
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7658875694199933e-05,
+      "loss": 1.4548,
+      "step": 2861
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7656521733667223e-05,
+      "loss": 1.4511,
+      "step": 2862
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.765416674735733e-05,
+      "loss": 1.4778,
+      "step": 2863
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.765181073558576e-05,
+      "loss": 1.3935,
+      "step": 2864
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7649453698668156e-05,
+      "loss": 1.4659,
+      "step": 2865
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7647095636920305e-05,
+      "loss": 1.7042,
+      "step": 2866
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7644736550658124e-05,
+      "loss": 1.7957,
+      "step": 2867
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.764237644019767e-05,
+      "loss": 1.5139,
+      "step": 2868
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.764001530585514e-05,
+      "loss": 1.4117,
+      "step": 2869
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7637653147946858e-05,
+      "loss": 1.4837,
+      "step": 2870
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7635289966789298e-05,
+      "loss": 1.4379,
+      "step": 2871
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7632925762699068e-05,
+      "loss": 1.4491,
+      "step": 2872
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7630560535992905e-05,
+      "loss": 1.4006,
+      "step": 2873
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7628194286987694e-05,
+      "loss": 1.5484,
+      "step": 2874
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7625827016000446e-05,
+      "loss": 1.4926,
+      "step": 2875
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.762345872334832e-05,
+      "loss": 1.3909,
+      "step": 2876
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.76210894093486e-05,
+      "loss": 1.405,
+      "step": 2877
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7618719074318718e-05,
+      "loss": 1.4736,
+      "step": 2878
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7616347718576238e-05,
+      "loss": 1.4445,
+      "step": 2879
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.761397534243886e-05,
+      "loss": 1.4878,
+      "step": 2880
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7611601946224417e-05,
+      "loss": 1.4414,
+      "step": 2881
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.760922753025089e-05,
+      "loss": 1.5657,
+      "step": 2882
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7606852094836385e-05,
+      "loss": 1.5008,
+      "step": 2883
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7604475640299153e-05,
+      "loss": 1.4922,
+      "step": 2884
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7602098166957577e-05,
+      "loss": 1.7451,
+      "step": 2885
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.759971967513017e-05,
+      "loss": 1.7077,
+      "step": 2886
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.75973401651356e-05,
+      "loss": 1.3986,
+      "step": 2887
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7594959637292656e-05,
+      "loss": 1.3966,
+      "step": 2888
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7592578091920264e-05,
+      "loss": 1.4086,
+      "step": 2889
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7590195529337495e-05,
+      "loss": 1.4481,
+      "step": 2890
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7587811949863548e-05,
+      "loss": 1.3974,
+      "step": 2891
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7585427353817758e-05,
+      "loss": 1.5013,
+      "step": 2892
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7583041741519608e-05,
+      "loss": 1.5454,
+      "step": 2893
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7580655113288708e-05,
+      "loss": 1.5033,
+      "step": 2894
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7578267469444793e-05,
+      "loss": 1.4332,
+      "step": 2895
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.757587881030776e-05,
+      "loss": 1.4667,
+      "step": 2896
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7573489136197623e-05,
+      "loss": 1.4314,
+      "step": 2897
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7571098447434532e-05,
+      "loss": 1.4514,
+      "step": 2898
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7568706744338784e-05,
+      "loss": 1.4872,
+      "step": 2899
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7566314027230808e-05,
+      "loss": 1.4841,
+      "step": 2900
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7563920296431158e-05,
+      "loss": 1.3964,
+      "step": 2901
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7561525552260538e-05,
+      "loss": 1.4932,
+      "step": 2902
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.755912979503978e-05,
+      "loss": 1.4153,
+      "step": 2903
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.755673302508986e-05,
+      "loss": 1.4939,
+      "step": 2904
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.755433524273187e-05,
+      "loss": 1.4437,
+      "step": 2905
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7551936448287067e-05,
+      "loss": 1.486,
+      "step": 2906
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.754953664207682e-05,
+      "loss": 1.4362,
+      "step": 2907
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.7547135824422635e-05,
+      "loss": 1.4954,
+      "step": 2908
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7544733995646174e-05,
+      "loss": 1.4552,
+      "step": 2909
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7542331156069213e-05,
+      "loss": 1.4654,
+      "step": 2910
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.753992730601367e-05,
+      "loss": 1.4531,
+      "step": 2911
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.75375224458016e-05,
+      "loss": 1.4949,
+      "step": 2912
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.753511657575519e-05,
+      "loss": 1.3998,
+      "step": 2913
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.753270969619677e-05,
+      "loss": 1.3904,
+      "step": 2914
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7530301807448794e-05,
+      "loss": 1.481,
+      "step": 2915
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7527892909833864e-05,
+      "loss": 1.4301,
+      "step": 2916
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7525483003674707e-05,
+      "loss": 1.4377,
+      "step": 2917
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7523072089294188e-05,
+      "loss": 1.5077,
+      "step": 2918
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.752066016701531e-05,
+      "loss": 1.3802,
+      "step": 2919
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7518247237161205e-05,
+      "loss": 1.3991,
+      "step": 2920
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7515833300055146e-05,
+      "loss": 1.4283,
+      "step": 2921
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7513418356020535e-05,
+      "loss": 1.3688,
+      "step": 2922
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.751100240538092e-05,
+      "loss": 1.4336,
+      "step": 2923
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7508585448459967e-05,
+      "loss": 1.4302,
+      "step": 2924
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7506167485581493e-05,
+      "loss": 1.4293,
+      "step": 2925
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.750374851706944e-05,
+      "loss": 1.4262,
+      "step": 2926
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7501328543247892e-05,
+      "loss": 1.4128,
+      "step": 2927
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7498907564441054e-05,
+      "loss": 1.4258,
+      "step": 2928
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.749648558097328e-05,
+      "loss": 1.4904,
+      "step": 2929
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.749406259316906e-05,
+      "loss": 1.4628,
+      "step": 2930
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7491638601353004e-05,
+      "loss": 1.4908,
+      "step": 2931
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7489213605849867e-05,
+      "loss": 1.4563,
+      "step": 2932
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7486787606984533e-05,
+      "loss": 1.5006,
+      "step": 2933
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.748436060508203e-05,
+      "loss": 1.424,
+      "step": 2934
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7481932600467508e-05,
+      "loss": 1.4136,
+      "step": 2935
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.747950359346626e-05,
+      "loss": 1.3835,
+      "step": 2936
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.747707358440371e-05,
+      "loss": 1.5284,
+      "step": 2937
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7474642573605418e-05,
+      "loss": 1.4592,
+      "step": 2938
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7472210561397075e-05,
+      "loss": 1.5483,
+      "step": 2939
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.746977754810451e-05,
+      "loss": 1.521,
+      "step": 2940
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7467343534053683e-05,
+      "loss": 1.4276,
+      "step": 2941
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.746490851957069e-05,
+      "loss": 1.4302,
+      "step": 2942
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.746247250498176e-05,
+      "loss": 1.4439,
+      "step": 2943
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7460035490613255e-05,
+      "loss": 1.4991,
+      "step": 2944
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7457597476791674e-05,
+      "loss": 1.4188,
+      "step": 2945
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7455158463843648e-05,
+      "loss": 1.4685,
+      "step": 2946
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.745271845209594e-05,
+      "loss": 1.3811,
+      "step": 2947
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7450277441875454e-05,
+      "loss": 1.4869,
+      "step": 2948
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7447835433509218e-05,
+      "loss": 1.5001,
+      "step": 2949
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.74453924273244e-05,
+      "loss": 1.4757,
+      "step": 2950
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.74429484236483e-05,
+      "loss": 1.5072,
+      "step": 2951
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.744050342280835e-05,
+      "loss": 1.4242,
+      "step": 2952
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.743805742513212e-05,
+      "loss": 1.4462,
+      "step": 2953
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.743561043094731e-05,
+      "loss": 1.6815,
+      "step": 2954
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7433162440581755e-05,
+      "loss": 1.427,
+      "step": 2955
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7430713454363423e-05,
+      "loss": 1.2933,
+      "step": 2956
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.742826347262041e-05,
+      "loss": 1.4247,
+      "step": 2957
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7425812495680955e-05,
+      "loss": 1.399,
+      "step": 2958
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.742336052387343e-05,
+      "loss": 1.4818,
+      "step": 2959
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.742090755752633e-05,
+      "loss": 1.3839,
+      "step": 2960
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.741845359696829e-05,
+      "loss": 1.4669,
+      "step": 2961
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.741599864252808e-05,
+      "loss": 1.468,
+      "step": 2962
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.74135426945346e-05,
+      "loss": 1.4135,
+      "step": 2963
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7411085753316882e-05,
+      "loss": 1.5899,
+      "step": 2964
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.74086278192041e-05,
+      "loss": 1.6115,
+      "step": 2965
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.740616889252554e-05,
+      "loss": 1.3988,
+      "step": 2966
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.740370897361065e-05,
+      "loss": 1.5001,
+      "step": 2967
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.740124806278899e-05,
+      "loss": 1.459,
+      "step": 2968
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7398786160390258e-05,
+      "loss": 1.4489,
+      "step": 2969
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7396323266744283e-05,
+      "loss": 1.4397,
+      "step": 2970
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7393859382181033e-05,
+      "loss": 1.4687,
+      "step": 2971
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7391394507030606e-05,
+      "loss": 1.4495,
+      "step": 2972
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.738892864162323e-05,
+      "loss": 1.4745,
+      "step": 2973
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7386461786289274e-05,
+      "loss": 1.3943,
+      "step": 2974
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.738399394135922e-05,
+      "loss": 1.4595,
+      "step": 2975
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.738152510716371e-05,
+      "loss": 1.4023,
+      "step": 2976
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7379055284033493e-05,
+      "loss": 1.3938,
+      "step": 2977
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7376584472299468e-05,
+      "loss": 1.4576,
+      "step": 2978
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.737411267229266e-05,
+      "loss": 1.424,
+      "step": 2979
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7371639884344225e-05,
+      "loss": 1.4224,
+      "step": 2980
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.736916610878545e-05,
+      "loss": 1.4228,
+      "step": 2981
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7366691345947768e-05,
+      "loss": 1.5021,
+      "step": 2982
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.736421559616272e-05,
+      "loss": 1.3589,
+      "step": 2983
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7361738859762007e-05,
+      "loss": 1.4748,
+      "step": 2984
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.735926113707743e-05,
+      "loss": 1.3974,
+      "step": 2985
+    },
+    {
+      "epoch": 0.31,
+      "eval_loss": 1.4849082231521606,
+      "eval_runtime": 77.3271,
+      "eval_samples_per_second": 1.668,
+      "eval_steps_per_second": 0.116,
+      "step": 2985
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.735678242844096e-05,
+      "loss": 1.392,
+      "step": 2986
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.735430273418467e-05,
+      "loss": 1.4277,
+      "step": 2987
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.735182205464077e-05,
+      "loss": 1.4216,
+      "step": 2988
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7349340390141617e-05,
+      "loss": 1.4546,
+      "step": 2989
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.734685774101969e-05,
+      "loss": 1.4264,
+      "step": 2990
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7344374107607593e-05,
+      "loss": 1.4307,
+      "step": 2991
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7341889490238076e-05,
+      "loss": 1.4266,
+      "step": 2992
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7339403889244008e-05,
+      "loss": 1.4486,
+      "step": 2993
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7336917304958403e-05,
+      "loss": 1.4345,
+      "step": 2994
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7334429737714395e-05,
+      "loss": 1.476,
+      "step": 2995
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7331941187845248e-05,
+      "loss": 1.5148,
+      "step": 2996
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7329451655684372e-05,
+      "loss": 1.4026,
+      "step": 2997
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7326961141565303e-05,
+      "loss": 1.3935,
+      "step": 2998
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.73244696458217e-05,
+      "loss": 1.4483,
+      "step": 2999
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7321977168787356e-05,
+      "loss": 1.5197,
+      "step": 3000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7319483710796208e-05,
+      "loss": 1.4648,
+      "step": 3001
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.7316989272182308e-05,
+      "loss": 1.4171,
+      "step": 3002
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.731449385327985e-05,
+      "loss": 1.4541,
+      "step": 3003
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.731199745442316e-05,
+      "loss": 1.4605,
+      "step": 3004
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7309500075946683e-05,
+      "loss": 1.5122,
+      "step": 3005
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7307001718185007e-05,
+      "loss": 1.5243,
+      "step": 3006
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7304502381472844e-05,
+      "loss": 1.4997,
+      "step": 3007
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.730200206614505e-05,
+      "loss": 1.4891,
+      "step": 3008
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.72995007725366e-05,
+      "loss": 1.4376,
+      "step": 3009
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7296998500982595e-05,
+      "loss": 1.4403,
+      "step": 3010
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7294495251818285e-05,
+      "loss": 1.5153,
+      "step": 3011
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.729199102537904e-05,
+      "loss": 1.4996,
+      "step": 3012
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7289485822000356e-05,
+      "loss": 1.5302,
+      "step": 3013
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7286979642017867e-05,
+      "loss": 1.5641,
+      "step": 3014
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.728447248576734e-05,
+      "loss": 1.438,
+      "step": 3015
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7281964353584674e-05,
+      "loss": 1.4796,
+      "step": 3016
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.727945524580588e-05,
+      "loss": 1.4413,
+      "step": 3017
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.727694516276713e-05,
+      "loss": 1.3966,
+      "step": 3018
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.72744341048047e-05,
+      "loss": 1.5267,
+      "step": 3019
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.727192207225501e-05,
+      "loss": 1.4284,
+      "step": 3020
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.726940906545461e-05,
+      "loss": 1.5137,
+      "step": 3021
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7266895084740173e-05,
+      "loss": 1.3966,
+      "step": 3022
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7264380130448514e-05,
+      "loss": 1.4156,
+      "step": 3023
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.726186420291657e-05,
+      "loss": 1.4644,
+      "step": 3024
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.725934730248141e-05,
+      "loss": 1.4768,
+      "step": 3025
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7256829429480233e-05,
+      "loss": 1.4428,
+      "step": 3026
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7254310584250373e-05,
+      "loss": 1.4495,
+      "step": 3027
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.725179076712929e-05,
+      "loss": 1.3867,
+      "step": 3028
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7249269978454567e-05,
+      "loss": 1.5011,
+      "step": 3029
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7246748218563937e-05,
+      "loss": 1.3916,
+      "step": 3030
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7244225487795238e-05,
+      "loss": 1.4984,
+      "step": 3031
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7241701786486464e-05,
+      "loss": 1.5776,
+      "step": 3032
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7239177114975717e-05,
+      "loss": 1.5018,
+      "step": 3033
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7236651473601244e-05,
+      "loss": 1.5042,
+      "step": 3034
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.723412486270141e-05,
+      "loss": 1.4636,
+      "step": 3035
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.723159728261472e-05,
+      "loss": 1.5612,
+      "step": 3036
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7229068733679803e-05,
+      "loss": 1.4399,
+      "step": 3037
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7226539216235424e-05,
+      "loss": 1.5635,
+      "step": 3038
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7224008730620468e-05,
+      "loss": 1.4329,
+      "step": 3039
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7221477277173956e-05,
+      "loss": 1.4886,
+      "step": 3040
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.721894485623504e-05,
+      "loss": 1.4741,
+      "step": 3041
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7216411468143e-05,
+      "loss": 1.4678,
+      "step": 3042
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.721387711323724e-05,
+      "loss": 1.4758,
+      "step": 3043
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.72113417918573e-05,
+      "loss": 1.441,
+      "step": 3044
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7208805504342855e-05,
+      "loss": 1.4328,
+      "step": 3045
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.720626825103369e-05,
+      "loss": 1.4254,
+      "step": 3046
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7203730032269744e-05,
+      "loss": 1.4462,
+      "step": 3047
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7201190848391066e-05,
+      "loss": 1.4826,
+      "step": 3048
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.719865069973784e-05,
+      "loss": 1.4215,
+      "step": 3049
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7196109586650385e-05,
+      "loss": 1.4291,
+      "step": 3050
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7193567509469146e-05,
+      "loss": 1.5005,
+      "step": 3051
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.719102446853469e-05,
+      "loss": 1.4216,
+      "step": 3052
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7188480464187725e-05,
+      "loss": 1.4349,
+      "step": 3053
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.718593549676908e-05,
+      "loss": 1.4784,
+      "step": 3054
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7183389566619714e-05,
+      "loss": 1.4325,
+      "step": 3055
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7180842674080717e-05,
+      "loss": 1.5387,
+      "step": 3056
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.717829481949331e-05,
+      "loss": 1.655,
+      "step": 3057
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.717574600319883e-05,
+      "loss": 1.4623,
+      "step": 3058
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7173196225538768e-05,
+      "loss": 1.4646,
+      "step": 3059
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.717064548685472e-05,
+      "loss": 1.4641,
+      "step": 3060
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7168093787488418e-05,
+      "loss": 1.4033,
+      "step": 3061
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7165541127781726e-05,
+      "loss": 1.4085,
+      "step": 3062
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.716298750807664e-05,
+      "loss": 1.387,
+      "step": 3063
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7160432928715272e-05,
+      "loss": 1.4197,
+      "step": 3064
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7157877390039872e-05,
+      "loss": 1.4883,
+      "step": 3065
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7155320892392818e-05,
+      "loss": 1.4865,
+      "step": 3066
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.715276343611662e-05,
+      "loss": 1.4532,
+      "step": 3067
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.71502050215539e-05,
+      "loss": 1.3991,
+      "step": 3068
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7147645649047424e-05,
+      "loss": 1.4318,
+      "step": 3069
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.714508531894009e-05,
+      "loss": 1.3616,
+      "step": 3070
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7142524031574904e-05,
+      "loss": 1.4481,
+      "step": 3071
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7139961787295022e-05,
+      "loss": 1.3897,
+      "step": 3072
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7137398586443716e-05,
+      "loss": 1.3128,
+      "step": 3073
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.713483442936439e-05,
+      "loss": 1.4481,
+      "step": 3074
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.713226931640057e-05,
+      "loss": 1.5374,
+      "step": 3075
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7129703247895923e-05,
+      "loss": 1.4247,
+      "step": 3076
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.712713622419423e-05,
+      "loss": 1.4098,
+      "step": 3077
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7124568245639414e-05,
+      "loss": 1.4558,
+      "step": 3078
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7121999312575507e-05,
+      "loss": 1.4484,
+      "step": 3079
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.711942942534669e-05,
+      "loss": 1.4236,
+      "step": 3080
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7116858584297253e-05,
+      "loss": 1.613,
+      "step": 3081
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.711428678977163e-05,
+      "loss": 1.4503,
+      "step": 3082
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7111714042114376e-05,
+      "loss": 1.4072,
+      "step": 3083
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7109140341670168e-05,
+      "loss": 1.5999,
+      "step": 3084
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7106565688783815e-05,
+      "loss": 1.4129,
+      "step": 3085
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.710399008380026e-05,
+      "loss": 1.3973,
+      "step": 3086
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.710141352706456e-05,
+      "loss": 1.4292,
+      "step": 3087
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.709883601892192e-05,
+      "loss": 1.4372,
+      "step": 3088
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7096257559717647e-05,
+      "loss": 1.4924,
+      "step": 3089
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7093678149797196e-05,
+      "loss": 1.4222,
+      "step": 3090
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7091097789506137e-05,
+      "loss": 1.4797,
+      "step": 3091
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7088516479190175e-05,
+      "loss": 1.4738,
+      "step": 3092
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.708593421919514e-05,
+      "loss": 1.5371,
+      "step": 3093
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7083351009866988e-05,
+      "loss": 1.3802,
+      "step": 3094
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7080766851551795e-05,
+      "loss": 1.4292,
+      "step": 3095
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.707818174459578e-05,
+      "loss": 1.4478,
+      "step": 3096
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7075595689345285e-05,
+      "loss": 1.4619,
+      "step": 3097
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7073008686146763e-05,
+      "loss": 1.4603,
+      "step": 3098
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.7070420735346815e-05,
+      "loss": 1.4548,
+      "step": 3099
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7067831837292157e-05,
+      "loss": 1.3638,
+      "step": 3100
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7065241992329634e-05,
+      "loss": 1.4284,
+      "step": 3101
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7062651200806217e-05,
+      "loss": 1.386,
+      "step": 3102
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7060059463069013e-05,
+      "loss": 1.5483,
+      "step": 3103
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.705746677946524e-05,
+      "loss": 1.4963,
+      "step": 3104
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7054873150342257e-05,
+      "loss": 1.4539,
+      "step": 3105
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.705227857604754e-05,
+      "loss": 1.4361,
+      "step": 3106
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7049683056928695e-05,
+      "loss": 1.3735,
+      "step": 3107
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7047086593333454e-05,
+      "loss": 1.4324,
+      "step": 3108
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7044489185609685e-05,
+      "loss": 1.5037,
+      "step": 3109
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.704189083410536e-05,
+      "loss": 1.3913,
+      "step": 3110
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7039291539168604e-05,
+      "loss": 1.3233,
+      "step": 3111
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.703669130114765e-05,
+      "loss": 1.4897,
+      "step": 3112
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7034090120390862e-05,
+      "loss": 1.4843,
+      "step": 3113
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7031487997246735e-05,
+      "loss": 1.3674,
+      "step": 3114
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7028884932063883e-05,
+      "loss": 1.4475,
+      "step": 3115
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.702628092519105e-05,
+      "loss": 1.4134,
+      "step": 3116
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7023675976977108e-05,
+      "loss": 1.4484,
+      "step": 3117
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.702107008777106e-05,
+      "loss": 1.407,
+      "step": 3118
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7018463257922005e-05,
+      "loss": 1.5443,
+      "step": 3119
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.701585548777922e-05,
+      "loss": 1.4548,
+      "step": 3120
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7013246777692063e-05,
+      "loss": 1.4252,
+      "step": 3121
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7010637128010037e-05,
+      "loss": 1.3748,
+      "step": 3122
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7008026539082767e-05,
+      "loss": 1.3804,
+      "step": 3123
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7005415011260008e-05,
+      "loss": 1.4203,
+      "step": 3124
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.7002802544891634e-05,
+      "loss": 1.4195,
+      "step": 3125
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.700018914032765e-05,
+      "loss": 1.4188,
+      "step": 3126
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.699757479791819e-05,
+      "loss": 1.3884,
+      "step": 3127
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6994959518013502e-05,
+      "loss": 1.4111,
+      "step": 3128
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6992343300963963e-05,
+      "loss": 1.5452,
+      "step": 3129
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6989726147120087e-05,
+      "loss": 1.4097,
+      "step": 3130
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6987108056832506e-05,
+      "loss": 1.481,
+      "step": 3131
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6984489030451973e-05,
+      "loss": 1.4505,
+      "step": 3132
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6981869068329368e-05,
+      "loss": 1.4281,
+      "step": 3133
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.69792481708157e-05,
+      "loss": 1.4555,
+      "step": 3134
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6976626338262103e-05,
+      "loss": 1.4525,
+      "step": 3135
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6974003571019837e-05,
+      "loss": 1.3765,
+      "step": 3136
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6971379869440286e-05,
+      "loss": 1.3998,
+      "step": 3137
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.696875523387495e-05,
+      "loss": 1.5103,
+      "step": 3138
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6966129664675474e-05,
+      "loss": 1.427,
+      "step": 3139
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6963503162193612e-05,
+      "loss": 1.4362,
+      "step": 3140
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6960875726781246e-05,
+      "loss": 1.4073,
+      "step": 3141
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6958247358790388e-05,
+      "loss": 1.4014,
+      "step": 3142
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6955618058573165e-05,
+      "loss": 1.4244,
+      "step": 3143
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6952987826481847e-05,
+      "loss": 1.47,
+      "step": 3144
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6950356662868814e-05,
+      "loss": 1.5112,
+      "step": 3145
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6947724568086562e-05,
+      "loss": 1.3398,
+      "step": 3146
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.694509154248774e-05,
+      "loss": 1.4584,
+      "step": 3147
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.69424575864251e-05,
+      "loss": 1.4383,
+      "step": 3148
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6939822700251523e-05,
+      "loss": 1.3869,
+      "step": 3149
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6937186884320015e-05,
+      "loss": 1.4453,
+      "step": 3150
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6934550138983713e-05,
+      "loss": 1.3355,
+      "step": 3151
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6931912464595866e-05,
+      "loss": 1.5573,
+      "step": 3152
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6929273861509865e-05,
+      "loss": 1.439,
+      "step": 3153
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6926634330079203e-05,
+      "loss": 1.6079,
+      "step": 3154
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.692399387065752e-05,
+      "loss": 1.6973,
+      "step": 3155
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.692135248359856e-05,
+      "loss": 1.3769,
+      "step": 3156
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6918710169256206e-05,
+      "loss": 1.4454,
+      "step": 3157
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.691606692798446e-05,
+      "loss": 1.4495,
+      "step": 3158
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.691342276013745e-05,
+      "loss": 1.4292,
+      "step": 3159
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.691077766606942e-05,
+      "loss": 1.4405,
+      "step": 3160
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6908131646134757e-05,
+      "loss": 1.3683,
+      "step": 3161
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.690548470068795e-05,
+      "loss": 1.4731,
+      "step": 3162
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.690283683008362e-05,
+      "loss": 1.4323,
+      "step": 3163
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6900188034676518e-05,
+      "loss": 1.5133,
+      "step": 3164
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6897538314821516e-05,
+      "loss": 1.5529,
+      "step": 3165
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6894887670873605e-05,
+      "loss": 1.4427,
+      "step": 3166
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6892236103187904e-05,
+      "loss": 1.4344,
+      "step": 3167
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6889583612119657e-05,
+      "loss": 1.4291,
+      "step": 3168
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6886930198024226e-05,
+      "loss": 1.4683,
+      "step": 3169
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6884275861257106e-05,
+      "loss": 1.4442,
+      "step": 3170
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6881620602173905e-05,
+      "loss": 1.4509,
+      "step": 3171
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6878964421130362e-05,
+      "loss": 1.5113,
+      "step": 3172
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6876307318482334e-05,
+      "loss": 1.4261,
+      "step": 3173
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6873649294585805e-05,
+      "loss": 1.445,
+      "step": 3174
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6870990349796886e-05,
+      "loss": 1.4603,
+      "step": 3175
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.68683304844718e-05,
+      "loss": 1.41,
+      "step": 3176
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.686566969896691e-05,
+      "loss": 1.4018,
+      "step": 3177
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6863007993638687e-05,
+      "loss": 1.422,
+      "step": 3178
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6860345368843727e-05,
+      "loss": 1.4412,
+      "step": 3179
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6857681824938768e-05,
+      "loss": 1.5411,
+      "step": 3180
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.685501736228064e-05,
+      "loss": 1.4416,
+      "step": 3181
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.685235198122632e-05,
+      "loss": 1.5113,
+      "step": 3182
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6849685682132903e-05,
+      "loss": 1.423,
+      "step": 3183
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.68470184653576e-05,
+      "loss": 1.404,
+      "step": 3184
+    },
+    {
+      "epoch": 0.33,
+      "eval_loss": 1.4771089553833008,
+      "eval_runtime": 77.4034,
+      "eval_samples_per_second": 1.667,
+      "eval_steps_per_second": 0.116,
+      "step": 3184
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6844350331257747e-05,
+      "loss": 1.4624,
+      "step": 3185
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6841681280190814e-05,
+      "loss": 1.4306,
+      "step": 3186
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.683901131251438e-05,
+      "loss": 1.42,
+      "step": 3187
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6836340428586155e-05,
+      "loss": 1.4787,
+      "step": 3188
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6833668628763965e-05,
+      "loss": 1.3485,
+      "step": 3189
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.683099591340576e-05,
+      "loss": 1.3674,
+      "step": 3190
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.6828322282869627e-05,
+      "loss": 1.5187,
+      "step": 3191
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.682564773751375e-05,
+      "loss": 1.4847,
+      "step": 3192
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.682297227769646e-05,
+      "loss": 1.4815,
+      "step": 3193
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.682029590377619e-05,
+      "loss": 1.4281,
+      "step": 3194
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6817618616111517e-05,
+      "loss": 1.4734,
+      "step": 3195
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.681494041506112e-05,
+      "loss": 1.4861,
+      "step": 3196
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6812261300983808e-05,
+      "loss": 1.4372,
+      "step": 3197
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.680958127423852e-05,
+      "loss": 1.4396,
+      "step": 3198
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.680690033518431e-05,
+      "loss": 1.4225,
+      "step": 3199
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6804218484180347e-05,
+      "loss": 1.4925,
+      "step": 3200
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6801535721585942e-05,
+      "loss": 1.4306,
+      "step": 3201
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6798852047760506e-05,
+      "loss": 1.4298,
+      "step": 3202
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6796167463063588e-05,
+      "loss": 1.4725,
+      "step": 3203
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.679348196785485e-05,
+      "loss": 1.4648,
+      "step": 3204
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6790795562494087e-05,
+      "loss": 1.4152,
+      "step": 3205
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6788108247341198e-05,
+      "loss": 1.4637,
+      "step": 3206
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6785420022756216e-05,
+      "loss": 1.4077,
+      "step": 3207
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.67827308890993e-05,
+      "loss": 1.4585,
+      "step": 3208
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.678004084673072e-05,
+      "loss": 1.3747,
+      "step": 3209
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6777349896010876e-05,
+      "loss": 1.4264,
+      "step": 3210
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.677465803730028e-05,
+      "loss": 1.4178,
+      "step": 3211
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6771965270959584e-05,
+      "loss": 1.3876,
+      "step": 3212
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6769271597349538e-05,
+      "loss": 1.4658,
+      "step": 3213
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6766577016831026e-05,
+      "loss": 1.3968,
+      "step": 3214
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.676388152976506e-05,
+      "loss": 1.5015,
+      "step": 3215
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6761185136512756e-05,
+      "loss": 1.4121,
+      "step": 3216
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.675848783743537e-05,
+      "loss": 1.4146,
+      "step": 3217
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.675578963289427e-05,
+      "loss": 1.4113,
+      "step": 3218
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6753090523250937e-05,
+      "loss": 1.4048,
+      "step": 3219
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6750390508866994e-05,
+      "loss": 1.4959,
+      "step": 3220
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6747689590104167e-05,
+      "loss": 1.4445,
+      "step": 3221
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6744987767324312e-05,
+      "loss": 1.4611,
+      "step": 3222
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.67422850408894e-05,
+      "loss": 1.4141,
+      "step": 3223
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6739581411161535e-05,
+      "loss": 1.4218,
+      "step": 3224
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6736876878502925e-05,
+      "loss": 1.4266,
+      "step": 3225
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6734171443275913e-05,
+      "loss": 1.4941,
+      "step": 3226
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6731465105842955e-05,
+      "loss": 1.4294,
+      "step": 3227
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6728757866566635e-05,
+      "loss": 1.4189,
+      "step": 3228
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6726049725809647e-05,
+      "loss": 1.4713,
+      "step": 3229
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6723340683934818e-05,
+      "loss": 1.4203,
+      "step": 3230
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6720630741305088e-05,
+      "loss": 1.4112,
+      "step": 3231
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6717919898283523e-05,
+      "loss": 1.3811,
+      "step": 3232
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6715208155233296e-05,
+      "loss": 1.4242,
+      "step": 3233
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6712495512517725e-05,
+      "loss": 1.4207,
+      "step": 3234
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6709781970500222e-05,
+      "loss": 1.4888,
+      "step": 3235
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.670706752954434e-05,
+      "loss": 1.4501,
+      "step": 3236
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.670435219001374e-05,
+      "loss": 1.4367,
+      "step": 3237
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6701635952272212e-05,
+      "loss": 1.3959,
+      "step": 3238
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6698918816683655e-05,
+      "loss": 1.4257,
+      "step": 3239
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6696200783612102e-05,
+      "loss": 1.4718,
+      "step": 3240
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6693481853421697e-05,
+      "loss": 1.5899,
+      "step": 3241
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6690762026476707e-05,
+      "loss": 1.418,
+      "step": 3242
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.668804130314152e-05,
+      "loss": 1.4076,
+      "step": 3243
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6685319683780645e-05,
+      "loss": 1.4571,
+      "step": 3244
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.66825971687587e-05,
+      "loss": 1.4066,
+      "step": 3245
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6679873758440442e-05,
+      "loss": 1.4639,
+      "step": 3246
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6677149453190732e-05,
+      "loss": 1.453,
+      "step": 3247
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6674424253374558e-05,
+      "loss": 1.4262,
+      "step": 3248
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.667169815935703e-05,
+      "loss": 1.3687,
+      "step": 3249
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6668971171503372e-05,
+      "loss": 1.5557,
+      "step": 3250
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6666243290178933e-05,
+      "loss": 1.4639,
+      "step": 3251
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6663514515749172e-05,
+      "loss": 1.4334,
+      "step": 3252
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6660784848579685e-05,
+      "loss": 1.3523,
+      "step": 3253
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6658054289036166e-05,
+      "loss": 1.3738,
+      "step": 3254
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.665532283748445e-05,
+      "loss": 1.4159,
+      "step": 3255
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6652590494290474e-05,
+      "loss": 1.3609,
+      "step": 3256
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6649857259820303e-05,
+      "loss": 1.4449,
+      "step": 3257
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6647123134440126e-05,
+      "loss": 1.424,
+      "step": 3258
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.664438811851624e-05,
+      "loss": 1.4737,
+      "step": 3259
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6641652212415066e-05,
+      "loss": 1.2717,
+      "step": 3260
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.663891541650315e-05,
+      "loss": 1.4845,
+      "step": 3261
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.663617773114715e-05,
+      "loss": 1.4115,
+      "step": 3262
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.663343915671384e-05,
+      "loss": 1.4237,
+      "step": 3263
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6630699693570127e-05,
+      "loss": 1.4209,
+      "step": 3264
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6627959342083026e-05,
+      "loss": 1.506,
+      "step": 3265
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6625218102619672e-05,
+      "loss": 1.4466,
+      "step": 3266
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6622475975547324e-05,
+      "loss": 1.4129,
+      "step": 3267
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.661973296123335e-05,
+      "loss": 1.4152,
+      "step": 3268
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.661698906004525e-05,
+      "loss": 1.43,
+      "step": 3269
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.661424427235064e-05,
+      "loss": 1.4701,
+      "step": 3270
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6611498598517238e-05,
+      "loss": 1.4598,
+      "step": 3271
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.66087520389129e-05,
+      "loss": 1.4795,
+      "step": 3272
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6606004593905595e-05,
+      "loss": 1.5264,
+      "step": 3273
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6603256263863415e-05,
+      "loss": 1.4412,
+      "step": 3274
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6600507049154556e-05,
+      "loss": 1.45,
+      "step": 3275
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.659775695014735e-05,
+      "loss": 1.4985,
+      "step": 3276
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.659500596721024e-05,
+      "loss": 1.496,
+      "step": 3277
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6592254100711775e-05,
+      "loss": 1.4386,
+      "step": 3278
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.658950135102065e-05,
+      "loss": 1.4073,
+      "step": 3279
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.658674771850565e-05,
+      "loss": 1.3912,
+      "step": 3280
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6583993203535702e-05,
+      "loss": 1.3628,
+      "step": 3281
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.658123780647983e-05,
+      "loss": 1.6089,
+      "step": 3282
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6578481527707196e-05,
+      "loss": 1.2367,
+      "step": 3283
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.657572436758706e-05,
+      "loss": 1.4058,
+      "step": 3284
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6572966326488825e-05,
+      "loss": 1.4067,
+      "step": 3285
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.657020740478198e-05,
+      "loss": 1.4187,
+      "step": 3286
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6567447602836166e-05,
+      "loss": 1.4599,
+      "step": 3287
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6564686921021115e-05,
+      "loss": 1.432,
+      "step": 3288
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.6561925359706693e-05,
+      "loss": 1.3941,
+      "step": 3289
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6559162919262874e-05,
+      "loss": 1.3644,
+      "step": 3290
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6556399600059756e-05,
+      "loss": 1.4628,
+      "step": 3291
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6553635402467555e-05,
+      "loss": 1.4355,
+      "step": 3292
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6550870326856596e-05,
+      "loss": 1.4627,
+      "step": 3293
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6548104373597335e-05,
+      "loss": 1.442,
+      "step": 3294
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6545337543060338e-05,
+      "loss": 1.3991,
+      "step": 3295
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6542569835616283e-05,
+      "loss": 1.4193,
+      "step": 3296
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6539801251635982e-05,
+      "loss": 1.4792,
+      "step": 3297
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.653703179149034e-05,
+      "loss": 1.4735,
+      "step": 3298
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6534261455550408e-05,
+      "loss": 1.451,
+      "step": 3299
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.653149024418733e-05,
+      "loss": 1.3975,
+      "step": 3300
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.652871815777238e-05,
+      "loss": 1.4843,
+      "step": 3301
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6525945196676948e-05,
+      "loss": 1.4782,
+      "step": 3302
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.652317136127254e-05,
+      "loss": 1.3865,
+      "step": 3303
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6520396651930777e-05,
+      "loss": 1.4359,
+      "step": 3304
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6517621069023397e-05,
+      "loss": 1.4289,
+      "step": 3305
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.651484461292226e-05,
+      "loss": 1.3526,
+      "step": 3306
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6512067283999336e-05,
+      "loss": 1.3671,
+      "step": 3307
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.650928908262672e-05,
+      "loss": 1.6041,
+      "step": 3308
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.650651000917662e-05,
+      "loss": 1.4322,
+      "step": 3309
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6503730064021354e-05,
+      "loss": 1.3992,
+      "step": 3310
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.650094924753337e-05,
+      "loss": 1.4928,
+      "step": 3311
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6498167560085222e-05,
+      "loss": 1.4075,
+      "step": 3312
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6495385002049583e-05,
+      "loss": 1.4535,
+      "step": 3313
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6492601573799253e-05,
+      "loss": 1.4221,
+      "step": 3314
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6489817275707133e-05,
+      "loss": 1.3688,
+      "step": 3315
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6487032108146243e-05,
+      "loss": 1.3889,
+      "step": 3316
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6484246071489733e-05,
+      "loss": 1.3942,
+      "step": 3317
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6481459166110858e-05,
+      "loss": 1.4004,
+      "step": 3318
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.647867139238299e-05,
+      "loss": 1.7308,
+      "step": 3319
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6475882750679622e-05,
+      "loss": 1.5001,
+      "step": 3320
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.647309324137435e-05,
+      "loss": 1.4325,
+      "step": 3321
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6470302864840913e-05,
+      "loss": 1.4494,
+      "step": 3322
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6467511621453136e-05,
+      "loss": 1.4418,
+      "step": 3323
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.646471951158498e-05,
+      "loss": 1.4046,
+      "step": 3324
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6461926535610517e-05,
+      "loss": 1.5031,
+      "step": 3325
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.645913269390393e-05,
+      "loss": 1.4813,
+      "step": 3326
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6456337986839525e-05,
+      "loss": 1.3842,
+      "step": 3327
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.645354241479172e-05,
+      "loss": 1.4539,
+      "step": 3328
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6450745978135048e-05,
+      "loss": 1.4503,
+      "step": 3329
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6447948677244163e-05,
+      "loss": 1.3741,
+      "step": 3330
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6445150512493832e-05,
+      "loss": 1.425,
+      "step": 3331
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.644235148425893e-05,
+      "loss": 1.4277,
+      "step": 3332
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6439551592914464e-05,
+      "loss": 1.4918,
+      "step": 3333
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6436750838835543e-05,
+      "loss": 1.3364,
+      "step": 3334
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6433949222397397e-05,
+      "loss": 1.4741,
+      "step": 3335
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.643114674397537e-05,
+      "loss": 1.3867,
+      "step": 3336
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.642834340394492e-05,
+      "loss": 1.3817,
+      "step": 3337
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6425539202681625e-05,
+      "loss": 1.4308,
+      "step": 3338
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6422734140561178e-05,
+      "loss": 1.4401,
+      "step": 3339
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.641992821795938e-05,
+      "loss": 1.4338,
+      "step": 3340
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.641712143525216e-05,
+      "loss": 1.3804,
+      "step": 3341
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6414313792815546e-05,
+      "loss": 1.371,
+      "step": 3342
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6411505291025695e-05,
+      "loss": 1.4102,
+      "step": 3343
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6408695930258875e-05,
+      "loss": 1.4433,
+      "step": 3344
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6405885710891467e-05,
+      "loss": 1.3895,
+      "step": 3345
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6403074633299966e-05,
+      "loss": 1.4321,
+      "step": 3346
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.640026269786098e-05,
+      "loss": 1.4207,
+      "step": 3347
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.639744990495125e-05,
+      "loss": 1.3758,
+      "step": 3348
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6394636254947612e-05,
+      "loss": 1.4203,
+      "step": 3349
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6391821748227013e-05,
+      "loss": 1.4596,
+      "step": 3350
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6389006385166537e-05,
+      "loss": 1.4344,
+      "step": 3351
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6386190166143366e-05,
+      "loss": 1.4048,
+      "step": 3352
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6383373091534798e-05,
+      "loss": 1.4348,
+      "step": 3353
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.638055516171825e-05,
+      "loss": 1.4972,
+      "step": 3354
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6377736377071258e-05,
+      "loss": 1.4685,
+      "step": 3355
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.637491673797146e-05,
+      "loss": 1.4093,
+      "step": 3356
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6372096244796616e-05,
+      "loss": 1.4452,
+      "step": 3357
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.63692748979246e-05,
+      "loss": 1.4458,
+      "step": 3358
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6366452697733404e-05,
+      "loss": 1.4187,
+      "step": 3359
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6363629644601126e-05,
+      "loss": 1.4052,
+      "step": 3360
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6360805738905984e-05,
+      "loss": 1.439,
+      "step": 3361
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.635798098102631e-05,
+      "loss": 1.4131,
+      "step": 3362
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6355155371340544e-05,
+      "loss": 1.4391,
+      "step": 3363
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6352328910227252e-05,
+      "loss": 1.4318,
+      "step": 3364
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.63495015980651e-05,
+      "loss": 1.5112,
+      "step": 3365
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6346673435232884e-05,
+      "loss": 1.4688,
+      "step": 3366
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6343844422109496e-05,
+      "loss": 1.3807,
+      "step": 3367
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.634101455907396e-05,
+      "loss": 1.4083,
+      "step": 3368
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6338183846505394e-05,
+      "loss": 1.3566,
+      "step": 3369
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.633535228478305e-05,
+      "loss": 1.4287,
+      "step": 3370
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6332519874286285e-05,
+      "loss": 1.3494,
+      "step": 3371
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6329686615394558e-05,
+      "loss": 1.3813,
+      "step": 3372
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6326852508487468e-05,
+      "loss": 1.4241,
+      "step": 3373
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.63240175539447e-05,
+      "loss": 1.424,
+      "step": 3374
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.632118175214607e-05,
+      "loss": 1.4212,
+      "step": 3375
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6318345103471504e-05,
+      "loss": 1.4671,
+      "step": 3376
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6315507608301044e-05,
+      "loss": 1.4593,
+      "step": 3377
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.631266926701483e-05,
+      "loss": 1.7816,
+      "step": 3378
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6309830079993134e-05,
+      "loss": 1.5134,
+      "step": 3379
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6306990047616334e-05,
+      "loss": 0.8933,
+      "step": 3380
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.630414917026492e-05,
+      "loss": 1.1033,
+      "step": 3381
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6301307448319497e-05,
+      "loss": 1.3494,
+      "step": 3382
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6298464882160784e-05,
+      "loss": 1.4299,
+      "step": 3383
+    },
+    {
+      "epoch": 0.35,
+      "eval_loss": 1.482450008392334,
+      "eval_runtime": 77.3737,
+      "eval_samples_per_second": 1.667,
+      "eval_steps_per_second": 0.116,
+      "step": 3383
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.629562147216961e-05,
+      "loss": 1.3971,
+      "step": 3384
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.6292777218726917e-05,
+      "loss": 1.3706,
+      "step": 3385
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6289932122213765e-05,
+      "loss": 1.4524,
+      "step": 3386
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6287086183011328e-05,
+      "loss": 1.4595,
+      "step": 3387
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.628423940150088e-05,
+      "loss": 1.492,
+      "step": 3388
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.628139177806382e-05,
+      "loss": 1.5576,
+      "step": 3389
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6278543313081657e-05,
+      "loss": 1.4615,
+      "step": 3390
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6275694006936015e-05,
+      "loss": 1.3456,
+      "step": 3391
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.627284386000862e-05,
+      "loss": 1.4254,
+      "step": 3392
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6269992872681325e-05,
+      "loss": 1.433,
+      "step": 3393
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.626714104533609e-05,
+      "loss": 1.5291,
+      "step": 3394
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.626428837835498e-05,
+      "loss": 1.4502,
+      "step": 3395
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.626143487212019e-05,
+      "loss": 1.3756,
+      "step": 3396
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6258580527014002e-05,
+      "loss": 1.414,
+      "step": 3397
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6255725343418835e-05,
+      "loss": 1.6289,
+      "step": 3398
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6252869321717208e-05,
+      "loss": 1.4862,
+      "step": 3399
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6250012462291756e-05,
+      "loss": 1.4213,
+      "step": 3400
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6247154765525217e-05,
+      "loss": 1.524,
+      "step": 3401
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.624429623180046e-05,
+      "loss": 1.6013,
+      "step": 3402
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.624143686150045e-05,
+      "loss": 1.4557,
+      "step": 3403
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6238576655008266e-05,
+      "loss": 1.3831,
+      "step": 3404
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6235715612707105e-05,
+      "loss": 1.4008,
+      "step": 3405
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.623285373498028e-05,
+      "loss": 1.4848,
+      "step": 3406
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6229991022211197e-05,
+      "loss": 1.4459,
+      "step": 3407
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6227127474783393e-05,
+      "loss": 1.4334,
+      "step": 3408
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6224263093080507e-05,
+      "loss": 1.4123,
+      "step": 3409
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6221397877486294e-05,
+      "loss": 1.3691,
+      "step": 3410
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6218531828384618e-05,
+      "loss": 1.4128,
+      "step": 3411
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6215664946159457e-05,
+      "loss": 1.3642,
+      "step": 3412
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6212797231194902e-05,
+      "loss": 1.4311,
+      "step": 3413
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6209928683875152e-05,
+      "loss": 1.4695,
+      "step": 3414
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6207059304584514e-05,
+      "loss": 1.3871,
+      "step": 3415
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6204189093707414e-05,
+      "loss": 1.4219,
+      "step": 3416
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.620131805162839e-05,
+      "loss": 1.6199,
+      "step": 3417
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.619844617873208e-05,
+      "loss": 1.4433,
+      "step": 3418
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.619557347540325e-05,
+      "loss": 1.4773,
+      "step": 3419
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6192699942026763e-05,
+      "loss": 1.4551,
+      "step": 3420
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6189825578987603e-05,
+      "loss": 1.4903,
+      "step": 3421
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6186950386670855e-05,
+      "loss": 1.473,
+      "step": 3422
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6184074365461722e-05,
+      "loss": 1.4082,
+      "step": 3423
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6181197515745523e-05,
+      "loss": 1.4746,
+      "step": 3424
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6178319837907674e-05,
+      "loss": 1.449,
+      "step": 3425
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6175441332333715e-05,
+      "loss": 1.377,
+      "step": 3426
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.617256199940929e-05,
+      "loss": 1.2364,
+      "step": 3427
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6169681839520155e-05,
+      "loss": 1.4437,
+      "step": 3428
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.616680085305218e-05,
+      "loss": 1.4726,
+      "step": 3429
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.616391904039134e-05,
+      "loss": 1.4824,
+      "step": 3430
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.616103640192373e-05,
+      "loss": 1.4052,
+      "step": 3431
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6158152938035537e-05,
+      "loss": 1.4131,
+      "step": 3432
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6155268649113084e-05,
+      "loss": 1.4267,
+      "step": 3433
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6152383535542786e-05,
+      "loss": 1.3735,
+      "step": 3434
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6149497597711175e-05,
+      "loss": 1.4249,
+      "step": 3435
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.614661083600489e-05,
+      "loss": 1.4312,
+      "step": 3436
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6143723250810686e-05,
+      "loss": 1.4831,
+      "step": 3437
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6140834842515425e-05,
+      "loss": 1.4183,
+      "step": 3438
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6137945611506078e-05,
+      "loss": 1.3908,
+      "step": 3439
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.613505555816973e-05,
+      "loss": 1.425,
+      "step": 3440
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.613216468289357e-05,
+      "loss": 1.3589,
+      "step": 3441
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6129272986064903e-05,
+      "loss": 1.4272,
+      "step": 3442
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6126380468071142e-05,
+      "loss": 1.4337,
+      "step": 3443
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6123487129299817e-05,
+      "loss": 1.3872,
+      "step": 3444
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.612059297013855e-05,
+      "loss": 1.374,
+      "step": 3445
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.611769799097509e-05,
+      "loss": 1.314,
+      "step": 3446
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6114802192197284e-05,
+      "loss": 1.3951,
+      "step": 3447
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6111905574193102e-05,
+      "loss": 1.4732,
+      "step": 3448
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6109008137350614e-05,
+      "loss": 1.3575,
+      "step": 3449
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6106109882058e-05,
+      "loss": 1.4606,
+      "step": 3450
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6103210808703554e-05,
+      "loss": 1.4357,
+      "step": 3451
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6100310917675678e-05,
+      "loss": 0.8456,
+      "step": 3452
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.609741020936288e-05,
+      "loss": 1.07,
+      "step": 3453
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6094508684153782e-05,
+      "loss": 1.4435,
+      "step": 3454
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.609160634243711e-05,
+      "loss": 1.3696,
+      "step": 3455
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6088703184601712e-05,
+      "loss": 1.4012,
+      "step": 3456
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.608579921103653e-05,
+      "loss": 1.5212,
+      "step": 3457
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.608289442213062e-05,
+      "loss": 1.4339,
+      "step": 3458
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.607998881827316e-05,
+      "loss": 1.4901,
+      "step": 3459
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6077082399853412e-05,
+      "loss": 1.4822,
+      "step": 3460
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.607417516726077e-05,
+      "loss": 1.3696,
+      "step": 3461
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6071267120884727e-05,
+      "loss": 1.4434,
+      "step": 3462
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6068358261114886e-05,
+      "loss": 1.3916,
+      "step": 3463
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.606544858834096e-05,
+      "loss": 1.422,
+      "step": 3464
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6062538102952763e-05,
+      "loss": 1.367,
+      "step": 3465
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.605962680534024e-05,
+      "loss": 1.3788,
+      "step": 3466
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.605671469589342e-05,
+      "loss": 1.4366,
+      "step": 3467
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6053801775002453e-05,
+      "loss": 1.4027,
+      "step": 3468
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6050888043057595e-05,
+      "loss": 1.4386,
+      "step": 3469
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6047973500449212e-05,
+      "loss": 1.4675,
+      "step": 3470
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6045058147567778e-05,
+      "loss": 1.6936,
+      "step": 3471
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6042141984803875e-05,
+      "loss": 1.4858,
+      "step": 3472
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6039225012548196e-05,
+      "loss": 1.5004,
+      "step": 3473
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6036307231191534e-05,
+      "loss": 1.4969,
+      "step": 3474
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6033388641124803e-05,
+      "loss": 1.4027,
+      "step": 3475
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.6030469242739016e-05,
+      "loss": 1.4741,
+      "step": 3476
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.60275490364253e-05,
+      "loss": 1.3887,
+      "step": 3477
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.602462802257488e-05,
+      "loss": 1.6735,
+      "step": 3478
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.602170620157911e-05,
+      "loss": 1.4997,
+      "step": 3479
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.601878357382943e-05,
+      "loss": 1.3973,
+      "step": 3480
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.6015860139717398e-05,
+      "loss": 1.416,
+      "step": 3481
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.6012935899634682e-05,
+      "loss": 1.4102,
+      "step": 3482
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.601001085397305e-05,
+      "loss": 1.4248,
+      "step": 3483
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.6007085003124386e-05,
+      "loss": 1.3909,
+      "step": 3484
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.600415834748068e-05,
+      "loss": 1.4029,
+      "step": 3485
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.6001230887434024e-05,
+      "loss": 1.4219,
+      "step": 3486
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5998302623376633e-05,
+      "loss": 1.4008,
+      "step": 3487
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5995373555700806e-05,
+      "loss": 1.5163,
+      "step": 3488
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.599244368479897e-05,
+      "loss": 1.4541,
+      "step": 3489
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.598951301106365e-05,
+      "loss": 1.4454,
+      "step": 3490
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.598658153488748e-05,
+      "loss": 1.1014,
+      "step": 3491
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.598364925666321e-05,
+      "loss": 1.407,
+      "step": 3492
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5980716176783682e-05,
+      "loss": 1.5046,
+      "step": 3493
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5977782295641857e-05,
+      "loss": 1.4265,
+      "step": 3494
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5974847613630796e-05,
+      "loss": 1.46,
+      "step": 3495
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5971912131143677e-05,
+      "loss": 1.3996,
+      "step": 3496
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5968975848573772e-05,
+      "loss": 1.4719,
+      "step": 3497
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.596603876631447e-05,
+      "loss": 1.43,
+      "step": 3498
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.596310088475927e-05,
+      "loss": 1.4109,
+      "step": 3499
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5960162204301768e-05,
+      "loss": 1.3612,
+      "step": 3500
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5957222725335672e-05,
+      "loss": 1.5045,
+      "step": 3501
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5954282448254794e-05,
+      "loss": 1.5323,
+      "step": 3502
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.595134137345306e-05,
+      "loss": 1.4518,
+      "step": 3503
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5948399501324495e-05,
+      "loss": 1.6131,
+      "step": 3504
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.594545683226324e-05,
+      "loss": 1.5076,
+      "step": 3505
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.594251336666353e-05,
+      "loss": 1.482,
+      "step": 3506
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5939569104919716e-05,
+      "loss": 1.4429,
+      "step": 3507
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5936624047426254e-05,
+      "loss": 1.4662,
+      "step": 3508
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5933678194577708e-05,
+      "loss": 1.425,
+      "step": 3509
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5930731546768744e-05,
+      "loss": 1.4312,
+      "step": 3510
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.592778410439414e-05,
+      "loss": 1.4336,
+      "step": 3511
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5924835867848775e-05,
+      "loss": 1.4197,
+      "step": 3512
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5921886837527638e-05,
+      "loss": 1.4456,
+      "step": 3513
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5918937013825823e-05,
+      "loss": 1.4394,
+      "step": 3514
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5915986397138526e-05,
+      "loss": 1.3606,
+      "step": 3515
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.591303498786106e-05,
+      "loss": 1.4382,
+      "step": 3516
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.591008278638884e-05,
+      "loss": 1.4185,
+      "step": 3517
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5907129793117374e-05,
+      "loss": 1.6692,
+      "step": 3518
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.59041760084423e-05,
+      "loss": 1.501,
+      "step": 3519
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5901221432759345e-05,
+      "loss": 1.5753,
+      "step": 3520
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.589826606646434e-05,
+      "loss": 1.4795,
+      "step": 3521
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5895309909953236e-05,
+      "loss": 1.4379,
+      "step": 3522
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5892352963622078e-05,
+      "loss": 1.3426,
+      "step": 3523
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5889395227867025e-05,
+      "loss": 1.4048,
+      "step": 3524
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5886436703084332e-05,
+      "loss": 1.4303,
+      "step": 3525
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5883477389670367e-05,
+      "loss": 1.3684,
+      "step": 3526
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5880517288021602e-05,
+      "loss": 1.3701,
+      "step": 3527
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.587755639853462e-05,
+      "loss": 1.4308,
+      "step": 3528
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5874594721606096e-05,
+      "loss": 1.3981,
+      "step": 3529
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5871632257632823e-05,
+      "loss": 1.5906,
+      "step": 3530
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5868669007011696e-05,
+      "loss": 1.4642,
+      "step": 3531
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.586570497013971e-05,
+      "loss": 1.411,
+      "step": 3532
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.586274014741397e-05,
+      "loss": 1.3942,
+      "step": 3533
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5859774539231693e-05,
+      "loss": 1.3952,
+      "step": 3534
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.585680814599019e-05,
+      "loss": 1.4009,
+      "step": 3535
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.585384096808688e-05,
+      "loss": 1.3238,
+      "step": 3536
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.585087300591929e-05,
+      "loss": 1.346,
+      "step": 3537
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.584790425988505e-05,
+      "loss": 1.3739,
+      "step": 3538
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5844934730381895e-05,
+      "loss": 1.4264,
+      "step": 3539
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5841964417807668e-05,
+      "loss": 1.4675,
+      "step": 3540
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5838993322560314e-05,
+      "loss": 1.3337,
+      "step": 3541
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5836021445037883e-05,
+      "loss": 1.4052,
+      "step": 3542
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.583304878563853e-05,
+      "loss": 1.4311,
+      "step": 3543
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5830075344760513e-05,
+      "loss": 1.386,
+      "step": 3544
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.58271011228022e-05,
+      "loss": 1.387,
+      "step": 3545
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.582412612016206e-05,
+      "loss": 1.3601,
+      "step": 3546
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5821150337238663e-05,
+      "loss": 1.5111,
+      "step": 3547
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5818173774430697e-05,
+      "loss": 1.471,
+      "step": 3548
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.581519643213693e-05,
+      "loss": 1.4442,
+      "step": 3549
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5812218310756263e-05,
+      "loss": 1.467,
+      "step": 3550
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5809239410687683e-05,
+      "loss": 1.3731,
+      "step": 3551
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5806259732330286e-05,
+      "loss": 1.4072,
+      "step": 3552
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5803279276083268e-05,
+      "loss": 1.3636,
+      "step": 3553
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5800298042345938e-05,
+      "loss": 1.3915,
+      "step": 3554
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5797316031517705e-05,
+      "loss": 1.5038,
+      "step": 3555
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5794333243998082e-05,
+      "loss": 1.4455,
+      "step": 3556
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5791349680186684e-05,
+      "loss": 1.5062,
+      "step": 3557
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5788365340483234e-05,
+      "loss": 1.3899,
+      "step": 3558
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5785380225287554e-05,
+      "loss": 1.4946,
+      "step": 3559
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5782394334999572e-05,
+      "loss": 1.3859,
+      "step": 3560
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.577940767001933e-05,
+      "loss": 1.541,
+      "step": 3561
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5776420230746952e-05,
+      "loss": 1.4952,
+      "step": 3562
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5773432017582684e-05,
+      "loss": 1.6055,
+      "step": 3563
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.577044303092687e-05,
+      "loss": 1.4661,
+      "step": 3564
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.576745327117996e-05,
+      "loss": 1.4082,
+      "step": 3565
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5764462738742493e-05,
+      "loss": 1.4005,
+      "step": 3566
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5761471434015137e-05,
+      "loss": 1.3959,
+      "step": 3567
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5758479357398645e-05,
+      "loss": 1.4567,
+      "step": 3568
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.575548650929388e-05,
+      "loss": 1.3791,
+      "step": 3569
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.57524928901018e-05,
+      "loss": 1.4428,
+      "step": 3570
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5749498500223486e-05,
+      "loss": 1.3864,
+      "step": 3571
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5746503340060097e-05,
+      "loss": 1.4019,
+      "step": 3572
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.574350741001291e-05,
+      "loss": 1.5436,
+      "step": 3573
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5740510710483308e-05,
+      "loss": 1.4637,
+      "step": 3574
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5737513241872767e-05,
+      "loss": 1.5097,
+      "step": 3575
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5734515004582873e-05,
+      "loss": 1.5143,
+      "step": 3576
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5731515999015315e-05,
+      "loss": 1.3926,
+      "step": 3577
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5728516225571874e-05,
+      "loss": 1.4242,
+      "step": 3578
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5725515684654453e-05,
+      "loss": 1.4429,
+      "step": 3579
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.572251437666504e-05,
+      "loss": 1.4638,
+      "step": 3580
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5719512302005735e-05,
+      "loss": 1.3381,
+      "step": 3581
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.571650946107874e-05,
+      "loss": 1.4342,
+      "step": 3582
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 1.4705015420913696,
+      "eval_runtime": 77.3664,
+      "eval_samples_per_second": 1.667,
+      "eval_steps_per_second": 0.116,
+      "step": 3582
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.571350585428636e-05,
+      "loss": 1.3991,
+      "step": 3583
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5710501482030998e-05,
+      "loss": 1.4274,
+      "step": 3584
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5707496344715166e-05,
+      "loss": 1.438,
+      "step": 3585
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.570449044274147e-05,
+      "loss": 1.4453,
+      "step": 3586
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.570148377651263e-05,
+      "loss": 1.4022,
+      "step": 3587
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5698476346431454e-05,
+      "loss": 1.4312,
+      "step": 3588
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.569546815290087e-05,
+      "loss": 1.4329,
+      "step": 3589
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.569245919632389e-05,
+      "loss": 1.4036,
+      "step": 3590
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5689449477103644e-05,
+      "loss": 1.412,
+      "step": 3591
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5686438995643353e-05,
+      "loss": 1.4492,
+      "step": 3592
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5683427752346343e-05,
+      "loss": 1.4555,
+      "step": 3593
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5680415747616047e-05,
+      "loss": 1.5594,
+      "step": 3594
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5677402981855993e-05,
+      "loss": 1.4813,
+      "step": 3595
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.567438945546982e-05,
+      "loss": 1.5919,
+      "step": 3596
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5671375168861255e-05,
+      "loss": 1.4235,
+      "step": 3597
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5668360122434135e-05,
+      "loss": 1.441,
+      "step": 3598
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.566534431659241e-05,
+      "loss": 1.5847,
+      "step": 3599
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.566232775174011e-05,
+      "loss": 1.6123,
+      "step": 3600
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5659310428281374e-05,
+      "loss": 1.4916,
+      "step": 3601
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5656292346620457e-05,
+      "loss": 1.4603,
+      "step": 3602
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.56532735071617e-05,
+      "loss": 1.4423,
+      "step": 3603
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5650253910309547e-05,
+      "loss": 1.4814,
+      "step": 3604
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5647233556468552e-05,
+      "loss": 1.4015,
+      "step": 3605
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.564421244604336e-05,
+      "loss": 1.3798,
+      "step": 3606
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5641190579438723e-05,
+      "loss": 1.4259,
+      "step": 3607
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5638167957059493e-05,
+      "loss": 1.3493,
+      "step": 3608
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5635144579310626e-05,
+      "loss": 1.3433,
+      "step": 3609
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5632120446597177e-05,
+      "loss": 1.4661,
+      "step": 3610
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5629095559324297e-05,
+      "loss": 1.4205,
+      "step": 3611
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.562606991789725e-05,
+      "loss": 1.3976,
+      "step": 3612
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5623043522721393e-05,
+      "loss": 1.4261,
+      "step": 3613
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.562001637420218e-05,
+      "loss": 1.4358,
+      "step": 3614
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5616988472745175e-05,
+      "loss": 1.437,
+      "step": 3615
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.561395981875604e-05,
+      "loss": 1.3735,
+      "step": 3616
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5610930412640532e-05,
+      "loss": 1.4357,
+      "step": 3617
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.560790025480452e-05,
+      "loss": 1.4003,
+      "step": 3618
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5604869345653965e-05,
+      "loss": 1.4127,
+      "step": 3619
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5601837685594927e-05,
+      "loss": 1.4666,
+      "step": 3620
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5598805275033574e-05,
+      "loss": 1.4106,
+      "step": 3621
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5595772114376174e-05,
+      "loss": 1.3857,
+      "step": 3622
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5592738204029086e-05,
+      "loss": 1.4704,
+      "step": 3623
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.558970354439878e-05,
+      "loss": 1.4649,
+      "step": 3624
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5586668135891824e-05,
+      "loss": 1.4812,
+      "step": 3625
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5583631978914882e-05,
+      "loss": 1.3507,
+      "step": 3626
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5580595073874718e-05,
+      "loss": 1.3762,
+      "step": 3627
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.557755742117821e-05,
+      "loss": 1.4169,
+      "step": 3628
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.557451902123231e-05,
+      "loss": 1.4098,
+      "step": 3629
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.55714798744441e-05,
+      "loss": 1.5526,
+      "step": 3630
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.556843998122074e-05,
+      "loss": 1.4414,
+      "step": 3631
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5565399341969497e-05,
+      "loss": 1.4134,
+      "step": 3632
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.556235795709774e-05,
+      "loss": 1.4718,
+      "step": 3633
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5559315827012937e-05,
+      "loss": 1.412,
+      "step": 3634
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5556272952122657e-05,
+      "loss": 1.4119,
+      "step": 3635
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5553229332834564e-05,
+      "loss": 1.3999,
+      "step": 3636
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5550184969556428e-05,
+      "loss": 1.4626,
+      "step": 3637
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5547139862696113e-05,
+      "loss": 1.3811,
+      "step": 3638
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5544094012661583e-05,
+      "loss": 1.4053,
+      "step": 3639
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5541047419860907e-05,
+      "loss": 1.4059,
+      "step": 3640
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.553800008470225e-05,
+      "loss": 1.4717,
+      "step": 3641
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.553495200759388e-05,
+      "loss": 1.399,
+      "step": 3642
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5531903188944154e-05,
+      "loss": 1.4143,
+      "step": 3643
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5528853629161536e-05,
+      "loss": 1.4414,
+      "step": 3644
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5525803328654592e-05,
+      "loss": 1.4206,
+      "step": 3645
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5522752287831982e-05,
+      "loss": 1.3869,
+      "step": 3646
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.551970050710247e-05,
+      "loss": 1.5434,
+      "step": 3647
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5516647986874914e-05,
+      "loss": 1.4574,
+      "step": 3648
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5513594727558273e-05,
+      "loss": 1.4353,
+      "step": 3649
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5510540729561606e-05,
+      "loss": 1.4122,
+      "step": 3650
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.550748599329407e-05,
+      "loss": 1.4077,
+      "step": 3651
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.550443051916492e-05,
+      "loss": 1.4433,
+      "step": 3652
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5501374307583514e-05,
+      "loss": 1.54,
+      "step": 3653
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5498317358959302e-05,
+      "loss": 1.4522,
+      "step": 3654
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5495259673701842e-05,
+      "loss": 1.4175,
+      "step": 3655
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5492201252220777e-05,
+      "loss": 1.4637,
+      "step": 3656
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5489142094925863e-05,
+      "loss": 1.4701,
+      "step": 3657
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.548608220222695e-05,
+      "loss": 1.4095,
+      "step": 3658
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.548302157453398e-05,
+      "loss": 1.3903,
+      "step": 3659
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5479960212257002e-05,
+      "loss": 1.4062,
+      "step": 3660
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5476898115806157e-05,
+      "loss": 1.4706,
+      "step": 3661
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.547383528559169e-05,
+      "loss": 1.4423,
+      "step": 3662
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.547077172202394e-05,
+      "loss": 1.4418,
+      "step": 3663
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5467707425513347e-05,
+      "loss": 1.4553,
+      "step": 3664
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5464642396470443e-05,
+      "loss": 1.435,
+      "step": 3665
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5461576635305873e-05,
+      "loss": 1.4105,
+      "step": 3666
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5458510142430364e-05,
+      "loss": 1.4118,
+      "step": 3667
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5455442918254746e-05,
+      "loss": 1.4096,
+      "step": 3668
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.545237496318995e-05,
+      "loss": 1.4248,
+      "step": 3669
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5449306277647003e-05,
+      "loss": 1.4591,
+      "step": 3670
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.544623686203703e-05,
+      "loss": 1.4764,
+      "step": 3671
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5443166716771255e-05,
+      "loss": 1.4329,
+      "step": 3672
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5440095842260995e-05,
+      "loss": 1.4023,
+      "step": 3673
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5437024238917674e-05,
+      "loss": 1.3802,
+      "step": 3674
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5433951907152806e-05,
+      "loss": 1.4627,
+      "step": 3675
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5430878847378e-05,
+      "loss": 1.5164,
+      "step": 3676
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.542780506000497e-05,
+      "loss": 1.3546,
+      "step": 3677
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.542473054544553e-05,
+      "loss": 1.4415,
+      "step": 3678
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5421655304111573e-05,
+      "loss": 1.4093,
+      "step": 3679
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5418579336415114e-05,
+      "loss": 1.3943,
+      "step": 3680
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5415502642768252e-05,
+      "loss": 1.3659,
+      "step": 3681
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.541242522358318e-05,
+      "loss": 1.4322,
+      "step": 3682
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5409347079272198e-05,
+      "loss": 1.4064,
+      "step": 3683
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5406268210247693e-05,
+      "loss": 1.4031,
+      "step": 3684
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.540318861692216e-05,
+      "loss": 1.3449,
+      "step": 3685
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5400108299708182e-05,
+      "loss": 1.5536,
+      "step": 3686
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5397027259018442e-05,
+      "loss": 1.4106,
+      "step": 3687
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5393945495265722e-05,
+      "loss": 1.4309,
+      "step": 3688
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.53908630088629e-05,
+      "loss": 1.3765,
+      "step": 3689
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5387779800222947e-05,
+      "loss": 1.3809,
+      "step": 3690
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5384695869758938e-05,
+      "loss": 1.4758,
+      "step": 3691
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5381611217884032e-05,
+      "loss": 1.4249,
+      "step": 3692
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5378525845011503e-05,
+      "loss": 1.4133,
+      "step": 3693
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5375439751554708e-05,
+      "loss": 1.418,
+      "step": 3694
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.53723529379271e-05,
+      "loss": 1.441,
+      "step": 3695
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.536926540454224e-05,
+      "loss": 1.3919,
+      "step": 3696
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5366177151813775e-05,
+      "loss": 1.3722,
+      "step": 3697
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.536308818015545e-05,
+      "loss": 1.4727,
+      "step": 3698
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.535999848998111e-05,
+      "loss": 1.442,
+      "step": 3699
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.535690808170469e-05,
+      "loss": 1.3884,
+      "step": 3700
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.535381695574023e-05,
+      "loss": 1.3226,
+      "step": 3701
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.535072511250186e-05,
+      "loss": 1.4403,
+      "step": 3702
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5347632552403808e-05,
+      "loss": 1.4144,
+      "step": 3703
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5344539275860395e-05,
+      "loss": 1.3514,
+      "step": 3704
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5341445283286044e-05,
+      "loss": 1.519,
+      "step": 3705
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5338350575095268e-05,
+      "loss": 1.4251,
+      "step": 3706
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5335255151702685e-05,
+      "loss": 1.392,
+      "step": 3707
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.533215901352299e-05,
+      "loss": 1.3838,
+      "step": 3708
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.532906216097099e-05,
+      "loss": 1.444,
+      "step": 3709
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5325964594461594e-05,
+      "loss": 1.3459,
+      "step": 3710
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5322866314409783e-05,
+      "loss": 1.4155,
+      "step": 3711
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5319767321230657e-05,
+      "loss": 1.4057,
+      "step": 3712
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5316667615339392e-05,
+      "loss": 1.382,
+      "step": 3713
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5313567197151277e-05,
+      "loss": 1.3634,
+      "step": 3714
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5310466067081688e-05,
+      "loss": 1.3434,
+      "step": 3715
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.530736422554609e-05,
+      "loss": 1.4553,
+      "step": 3716
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5304261672960055e-05,
+      "loss": 1.4193,
+      "step": 3717
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5301158409739242e-05,
+      "loss": 1.3955,
+      "step": 3718
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5298054436299413e-05,
+      "loss": 1.4592,
+      "step": 3719
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.529494975305642e-05,
+      "loss": 1.738,
+      "step": 3720
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.529184436042621e-05,
+      "loss": 1.438,
+      "step": 3721
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.528873825882482e-05,
+      "loss": 1.3445,
+      "step": 3722
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5285631448668395e-05,
+      "loss": 1.4168,
+      "step": 3723
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.528252393037317e-05,
+      "loss": 1.4088,
+      "step": 3724
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5279415704355468e-05,
+      "loss": 1.4484,
+      "step": 3725
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5276306771031708e-05,
+      "loss": 1.4706,
+      "step": 3726
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.527319713081841e-05,
+      "loss": 1.3466,
+      "step": 3727
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5270086784132196e-05,
+      "loss": 1.427,
+      "step": 3728
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5266975731389754e-05,
+      "loss": 1.4596,
+      "step": 3729
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.52638639730079e-05,
+      "loss": 1.405,
+      "step": 3730
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5260751509403518e-05,
+      "loss": 1.4572,
+      "step": 3731
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5257638340993608e-05,
+      "loss": 1.3827,
+      "step": 3732
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5254524468195251e-05,
+      "loss": 1.3712,
+      "step": 3733
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5251409891425626e-05,
+      "loss": 1.4541,
+      "step": 3734
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5248294611102006e-05,
+      "loss": 1.5279,
+      "step": 3735
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5245178627641759e-05,
+      "loss": 1.3722,
+      "step": 3736
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5242061941462348e-05,
+      "loss": 1.5158,
+      "step": 3737
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.523894455298132e-05,
+      "loss": 1.4274,
+      "step": 3738
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5235826462616338e-05,
+      "loss": 1.4543,
+      "step": 3739
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5232707670785144e-05,
+      "loss": 1.5011,
+      "step": 3740
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5229588177905568e-05,
+      "loss": 1.24,
+      "step": 3741
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5226467984395544e-05,
+      "loss": 1.4253,
+      "step": 3742
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5223347090673102e-05,
+      "loss": 1.3736,
+      "step": 3743
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.522022549715636e-05,
+      "loss": 1.4587,
+      "step": 3744
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.521710320426353e-05,
+      "loss": 1.5024,
+      "step": 3745
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.521398021241292e-05,
+      "loss": 1.427,
+      "step": 3746
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5210856522022929e-05,
+      "loss": 1.3387,
+      "step": 3747
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5207732133512055e-05,
+      "loss": 1.3819,
+      "step": 3748
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5204607047298883e-05,
+      "loss": 1.4571,
+      "step": 3749
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5201481263802094e-05,
+      "loss": 1.3592,
+      "step": 3750
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5198354783440464e-05,
+      "loss": 1.3912,
+      "step": 3751
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.519522760663286e-05,
+      "loss": 1.4986,
+      "step": 3752
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5192099733798243e-05,
+      "loss": 1.4106,
+      "step": 3753
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5188971165355667e-05,
+      "loss": 1.4145,
+      "step": 3754
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5185841901724282e-05,
+      "loss": 1.3915,
+      "step": 3755
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5182711943323329e-05,
+      "loss": 1.4014,
+      "step": 3756
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5179581290572141e-05,
+      "loss": 1.3651,
+      "step": 3757
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5176449943890147e-05,
+      "loss": 1.3653,
+      "step": 3758
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5173317903696862e-05,
+      "loss": 1.4486,
+      "step": 3759
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5170185170411906e-05,
+      "loss": 1.4459,
+      "step": 3760
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5167051744454977e-05,
+      "loss": 1.4215,
+      "step": 3761
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5163917626245877e-05,
+      "loss": 1.4379,
+      "step": 3762
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.51607828162045e-05,
+      "loss": 1.4147,
+      "step": 3763
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5157647314750827e-05,
+      "loss": 1.4011,
+      "step": 3764
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5154511122304936e-05,
+      "loss": 1.4309,
+      "step": 3765
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5151374239286993e-05,
+      "loss": 1.4382,
+      "step": 3766
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5148236666117266e-05,
+      "loss": 1.405,
+      "step": 3767
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5145098403216104e-05,
+      "loss": 1.4223,
+      "step": 3768
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5141959451003951e-05,
+      "loss": 1.4084,
+      "step": 3769
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5138819809901356e-05,
+      "loss": 1.4063,
+      "step": 3770
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.513567948032894e-05,
+      "loss": 1.3299,
+      "step": 3771
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5132538462707431e-05,
+      "loss": 1.3901,
+      "step": 3772
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5129396757457646e-05,
+      "loss": 1.4249,
+      "step": 3773
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.512625436500049e-05,
+      "loss": 1.4087,
+      "step": 3774
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5123111285756964e-05,
+      "loss": 1.3521,
+      "step": 3775
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5119967520148163e-05,
+      "loss": 1.4097,
+      "step": 3776
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5116823068595261e-05,
+      "loss": 1.4871,
+      "step": 3777
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5113677931519543e-05,
+      "loss": 1.44,
+      "step": 3778
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5110532109342377e-05,
+      "loss": 1.4518,
+      "step": 3779
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5107385602485214e-05,
+      "loss": 1.3961,
+      "step": 3780
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5104238411369614e-05,
+      "loss": 1.4341,
+      "step": 3781
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 1.4642821550369263,
+      "eval_runtime": 77.3595,
+      "eval_samples_per_second": 1.668,
+      "eval_steps_per_second": 0.116,
+      "step": 3781
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5101090536417218e-05,
+      "loss": 1.403,
+      "step": 3782
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5097941978049754e-05,
+      "loss": 1.4103,
+      "step": 3783
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5094792736689052e-05,
+      "loss": 1.3478,
+      "step": 3784
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5091642812757035e-05,
+      "loss": 1.4134,
+      "step": 3785
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5088492206675701e-05,
+      "loss": 1.4134,
+      "step": 3786
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5085340918867162e-05,
+      "loss": 1.1494,
+      "step": 3787
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5082188949753598e-05,
+      "loss": 1.2385,
+      "step": 3788
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5079036299757299e-05,
+      "loss": 1.4258,
+      "step": 3789
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5075882969300639e-05,
+      "loss": 1.3781,
+      "step": 3790
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.507272895880608e-05,
+      "loss": 1.5096,
+      "step": 3791
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5069574268696179e-05,
+      "loss": 1.366,
+      "step": 3792
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5066418899393584e-05,
+      "loss": 1.3792,
+      "step": 3793
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.506326285132104e-05,
+      "loss": 1.4987,
+      "step": 3794
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5060106124901364e-05,
+      "loss": 1.445,
+      "step": 3795
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5056948720557482e-05,
+      "loss": 1.5071,
+      "step": 3796
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.505379063871241e-05,
+      "loss": 1.4029,
+      "step": 3797
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5050631879789244e-05,
+      "loss": 1.5263,
+      "step": 3798
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5047472444211175e-05,
+      "loss": 1.3906,
+      "step": 3799
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5044312332401494e-05,
+      "loss": 1.7352,
+      "step": 3800
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5041151544783565e-05,
+      "loss": 1.3946,
+      "step": 3801
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5037990081780862e-05,
+      "loss": 1.3592,
+      "step": 3802
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.503482794381693e-05,
+      "loss": 1.3991,
+      "step": 3803
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5031665131315423e-05,
+      "loss": 1.4271,
+      "step": 3804
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.502850164470007e-05,
+      "loss": 1.4305,
+      "step": 3805
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5025337484394701e-05,
+      "loss": 1.4116,
+      "step": 3806
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5022172650823232e-05,
+      "loss": 1.4242,
+      "step": 3807
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5019007144409667e-05,
+      "loss": 1.4427,
+      "step": 3808
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5015840965578106e-05,
+      "loss": 1.3705,
+      "step": 3809
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5012674114752732e-05,
+      "loss": 1.4498,
+      "step": 3810
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5009506592357822e-05,
+      "loss": 1.489,
+      "step": 3811
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5006338398817746e-05,
+      "loss": 1.4801,
+      "step": 3812
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5003169534556956e-05,
+      "loss": 1.3468,
+      "step": 3813
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 1.4211,
+      "step": 3814
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4996829795571518e-05,
+      "loss": 1.4683,
+      "step": 3815
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.499365892169623e-05,
+      "loss": 1.4441,
+      "step": 3816
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4990487378798958e-05,
+      "loss": 1.4368,
+      "step": 3817
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4987315167304602e-05,
+      "loss": 1.4097,
+      "step": 3818
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4984142287638155e-05,
+      "loss": 1.4238,
+      "step": 3819
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4980968740224708e-05,
+      "loss": 1.497,
+      "step": 3820
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4977794525489433e-05,
+      "loss": 1.3989,
+      "step": 3821
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4974619643857588e-05,
+      "loss": 1.4415,
+      "step": 3822
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4971444095754533e-05,
+      "loss": 1.3931,
+      "step": 3823
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4968267881605705e-05,
+      "loss": 1.3426,
+      "step": 3824
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4965091001836636e-05,
+      "loss": 1.644,
+      "step": 3825
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4961913456872945e-05,
+      "loss": 1.3903,
+      "step": 3826
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4958735247140345e-05,
+      "loss": 1.3764,
+      "step": 3827
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4955556373064632e-05,
+      "loss": 1.5484,
+      "step": 3828
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4952376835071696e-05,
+      "loss": 1.5696,
+      "step": 3829
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4949196633587509e-05,
+      "loss": 1.3844,
+      "step": 3830
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.494601576903814e-05,
+      "loss": 1.3868,
+      "step": 3831
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4942834241849741e-05,
+      "loss": 1.4229,
+      "step": 3832
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4939652052448557e-05,
+      "loss": 1.4062,
+      "step": 3833
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4936469201260915e-05,
+      "loss": 1.3955,
+      "step": 3834
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4933285688713239e-05,
+      "loss": 1.3959,
+      "step": 3835
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4930101515232038e-05,
+      "loss": 1.4652,
+      "step": 3836
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4926916681243913e-05,
+      "loss": 1.5253,
+      "step": 3837
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4923731187175539e-05,
+      "loss": 1.4694,
+      "step": 3838
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4920545033453702e-05,
+      "loss": 1.3859,
+      "step": 3839
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4917358220505259e-05,
+      "loss": 1.4157,
+      "step": 3840
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4914170748757159e-05,
+      "loss": 1.4491,
+      "step": 3841
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4910982618636445e-05,
+      "loss": 1.4464,
+      "step": 3842
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4907793830570246e-05,
+      "loss": 1.3771,
+      "step": 3843
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4904604384985775e-05,
+      "loss": 1.3673,
+      "step": 3844
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4901414282310334e-05,
+      "loss": 1.4429,
+      "step": 3845
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4898223522971318e-05,
+      "loss": 1.4905,
+      "step": 3846
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4895032107396205e-05,
+      "loss": 1.4039,
+      "step": 3847
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4891840036012566e-05,
+      "loss": 1.4082,
+      "step": 3848
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4888647309248052e-05,
+      "loss": 1.3643,
+      "step": 3849
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4885453927530408e-05,
+      "loss": 1.3116,
+      "step": 3850
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4882259891287467e-05,
+      "loss": 1.4278,
+      "step": 3851
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4879065200947148e-05,
+      "loss": 1.3712,
+      "step": 3852
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4875869856937449e-05,
+      "loss": 1.387,
+      "step": 3853
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4872673859686475e-05,
+      "loss": 1.5053,
+      "step": 3854
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4869477209622404e-05,
+      "loss": 1.4427,
+      "step": 3855
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4866279907173502e-05,
+      "loss": 1.3404,
+      "step": 3856
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4863081952768126e-05,
+      "loss": 1.3126,
+      "step": 3857
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4859883346834725e-05,
+      "loss": 1.4588,
+      "step": 3858
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4856684089801822e-05,
+      "loss": 1.408,
+      "step": 3859
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4853484182098044e-05,
+      "loss": 1.4356,
+      "step": 3860
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4850283624152088e-05,
+      "loss": 1.3768,
+      "step": 3861
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.4847082416392753e-05,
+      "loss": 1.4563,
+      "step": 3862
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4843880559248915e-05,
+      "loss": 1.3648,
+      "step": 3863
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4840678053149542e-05,
+      "loss": 1.4146,
+      "step": 3864
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4837474898523686e-05,
+      "loss": 1.4423,
+      "step": 3865
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4834271095800487e-05,
+      "loss": 1.3351,
+      "step": 3866
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4831066645409178e-05,
+      "loss": 1.4534,
+      "step": 3867
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4827861547779066e-05,
+      "loss": 1.4763,
+      "step": 3868
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4824655803339553e-05,
+      "loss": 1.4019,
+      "step": 3869
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4821449412520132e-05,
+      "loss": 1.4223,
+      "step": 3870
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4818242375750371e-05,
+      "loss": 1.4065,
+      "step": 3871
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.481503469345993e-05,
+      "loss": 1.587,
+      "step": 3872
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4811826366078562e-05,
+      "loss": 1.3746,
+      "step": 3873
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4808617394036096e-05,
+      "loss": 1.3425,
+      "step": 3874
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.480540777776245e-05,
+      "loss": 0.6637,
+      "step": 3875
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4802197517687634e-05,
+      "loss": 1.3366,
+      "step": 3876
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.479898661424174e-05,
+      "loss": 1.4033,
+      "step": 3877
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4795775067854947e-05,
+      "loss": 1.4005,
+      "step": 3878
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4792562878957516e-05,
+      "loss": 1.3651,
+      "step": 3879
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.47893500479798e-05,
+      "loss": 1.4185,
+      "step": 3880
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4786136575352236e-05,
+      "loss": 1.4394,
+      "step": 3881
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.478292246150535e-05,
+      "loss": 1.7243,
+      "step": 3882
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4779707706869744e-05,
+      "loss": 1.3574,
+      "step": 3883
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4776492311876116e-05,
+      "loss": 1.4031,
+      "step": 3884
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4773276276955248e-05,
+      "loss": 1.4351,
+      "step": 3885
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4770059602538002e-05,
+      "loss": 1.3922,
+      "step": 3886
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4766842289055333e-05,
+      "loss": 1.3374,
+      "step": 3887
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4763624336938278e-05,
+      "loss": 1.428,
+      "step": 3888
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.476040574661796e-05,
+      "loss": 1.6087,
+      "step": 3889
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4757186518525586e-05,
+      "loss": 1.4054,
+      "step": 3890
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4753966653092447e-05,
+      "loss": 1.2595,
+      "step": 3891
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.475074615074993e-05,
+      "loss": 1.4052,
+      "step": 3892
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4747525011929492e-05,
+      "loss": 1.4296,
+      "step": 3893
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4744303237062689e-05,
+      "loss": 1.3718,
+      "step": 3894
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4741080826581151e-05,
+      "loss": 1.4498,
+      "step": 3895
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4737857780916602e-05,
+      "loss": 1.4029,
+      "step": 3896
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4734634100500846e-05,
+      "loss": 1.4136,
+      "step": 3897
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4731409785765774e-05,
+      "loss": 1.348,
+      "step": 3898
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4728184837143355e-05,
+      "loss": 1.3407,
+      "step": 3899
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.472495925506566e-05,
+      "loss": 1.3737,
+      "step": 3900
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4721733039964826e-05,
+      "loss": 2.2305,
+      "step": 3901
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4718506192273087e-05,
+      "loss": 1.4506,
+      "step": 3902
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4715278712422754e-05,
+      "loss": 1.3657,
+      "step": 3903
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.471205060084623e-05,
+      "loss": 1.4644,
+      "step": 3904
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4708821857976001e-05,
+      "loss": 1.4823,
+      "step": 3905
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4705592484244628e-05,
+      "loss": 1.3392,
+      "step": 3906
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4702362480084769e-05,
+      "loss": 1.3302,
+      "step": 3907
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4699131845929163e-05,
+      "loss": 1.8952,
+      "step": 3908
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.469590058221063e-05,
+      "loss": 1.4982,
+      "step": 3909
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4692668689362074e-05,
+      "loss": 1.407,
+      "step": 3910
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4689436167816489e-05,
+      "loss": 1.3911,
+      "step": 3911
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4686203018006948e-05,
+      "loss": 1.3421,
+      "step": 3912
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4682969240366612e-05,
+      "loss": 1.4627,
+      "step": 3913
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4679734835328718e-05,
+      "loss": 1.394,
+      "step": 3914
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4676499803326603e-05,
+      "loss": 1.4484,
+      "step": 3915
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4673264144793673e-05,
+      "loss": 1.4665,
+      "step": 3916
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4670027860163422e-05,
+      "loss": 1.3191,
+      "step": 3917
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.466679094986943e-05,
+      "loss": 1.4748,
+      "step": 3918
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4663553414345361e-05,
+      "loss": 1.2467,
+      "step": 3919
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.466031525402496e-05,
+      "loss": 1.4279,
+      "step": 3920
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4657076469342059e-05,
+      "loss": 1.3822,
+      "step": 3921
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4653837060730566e-05,
+      "loss": 1.3659,
+      "step": 3922
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.465059702862449e-05,
+      "loss": 1.3663,
+      "step": 3923
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4647356373457902e-05,
+      "loss": 1.4371,
+      "step": 3924
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4644115095664972e-05,
+      "loss": 1.3473,
+      "step": 3925
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4640873195679944e-05,
+      "loss": 1.4797,
+      "step": 3926
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.463763067393715e-05,
+      "loss": 1.3973,
+      "step": 3927
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4634387530871008e-05,
+      "loss": 1.3899,
+      "step": 3928
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4631143766916015e-05,
+      "loss": 1.3828,
+      "step": 3929
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4627899382506745e-05,
+      "loss": 1.4569,
+      "step": 3930
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4624654378077872e-05,
+      "loss": 1.4011,
+      "step": 3931
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.462140875406414e-05,
+      "loss": 1.3859,
+      "step": 3932
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4618162510900376e-05,
+      "loss": 1.4518,
+      "step": 3933
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4614915649021493e-05,
+      "loss": 1.4399,
+      "step": 3934
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4611668168862495e-05,
+      "loss": 1.4034,
+      "step": 3935
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4608420070858452e-05,
+      "loss": 1.3919,
+      "step": 3936
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4605171355444528e-05,
+      "loss": 1.391,
+      "step": 3937
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.460192202305597e-05,
+      "loss": 1.3717,
+      "step": 3938
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4598672074128103e-05,
+      "loss": 1.4351,
+      "step": 3939
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4595421509096339e-05,
+      "loss": 1.4202,
+      "step": 3940
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4592170328396165e-05,
+      "loss": 1.3559,
+      "step": 3941
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.458891853246316e-05,
+      "loss": 1.5878,
+      "step": 3942
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4585666121732982e-05,
+      "loss": 1.3818,
+      "step": 3943
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4582413096641369e-05,
+      "loss": 1.4009,
+      "step": 3944
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.457915945762414e-05,
+      "loss": 1.3937,
+      "step": 3945
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4575905205117203e-05,
+      "loss": 1.381,
+      "step": 3946
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.457265033955654e-05,
+      "loss": 1.3844,
+      "step": 3947
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4569394861378226e-05,
+      "loss": 1.4059,
+      "step": 3948
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4566138771018403e-05,
+      "loss": 1.3698,
+      "step": 3949
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4562882068913311e-05,
+      "loss": 1.4253,
+      "step": 3950
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.455962475549926e-05,
+      "loss": 1.4284,
+      "step": 3951
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4556366831212647e-05,
+      "loss": 1.4279,
+      "step": 3952
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4553108296489951e-05,
+      "loss": 1.3072,
+      "step": 3953
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4549849151767731e-05,
+      "loss": 1.4175,
+      "step": 3954
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4546589397482632e-05,
+      "loss": 1.5042,
+      "step": 3955
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.4543329034071371e-05,
+      "loss": 1.4363,
+      "step": 3956
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.454006806197076e-05,
+      "loss": 1.4111,
+      "step": 3957
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.453680648161768e-05,
+      "loss": 1.3817,
+      "step": 3958
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.45335442934491e-05,
+      "loss": 1.4689,
+      "step": 3959
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4530281497902072e-05,
+      "loss": 1.4164,
+      "step": 3960
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4527018095413726e-05,
+      "loss": 1.3943,
+      "step": 3961
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4523754086421271e-05,
+      "loss": 1.3851,
+      "step": 3962
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4520489471362003e-05,
+      "loss": 1.3907,
+      "step": 3963
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4517224250673297e-05,
+      "loss": 1.2877,
+      "step": 3964
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4513958424792605e-05,
+      "loss": 1.4696,
+      "step": 3965
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4510691994157471e-05,
+      "loss": 1.3785,
+      "step": 3966
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4507424959205505e-05,
+      "loss": 1.3652,
+      "step": 3967
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4504157320374408e-05,
+      "loss": 1.4198,
+      "step": 3968
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4500889078101962e-05,
+      "loss": 1.3476,
+      "step": 3969
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4497620232826028e-05,
+      "loss": 1.4356,
+      "step": 3970
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4494350784984543e-05,
+      "loss": 1.4993,
+      "step": 3971
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4491080735015533e-05,
+      "loss": 1.4022,
+      "step": 3972
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.44878100833571e-05,
+      "loss": 1.4074,
+      "step": 3973
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4484538830447423e-05,
+      "loss": 1.412,
+      "step": 3974
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4481266976724772e-05,
+      "loss": 1.3872,
+      "step": 3975
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4477994522627485e-05,
+      "loss": 1.4575,
+      "step": 3976
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4474721468593995e-05,
+      "loss": 1.3768,
+      "step": 3977
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.44714478150628e-05,
+      "loss": 1.4173,
+      "step": 3978
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4468173562472489e-05,
+      "loss": 1.3896,
+      "step": 3979
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4464898711261725e-05,
+      "loss": 1.4535,
+      "step": 3980
+    },
+    {
+      "epoch": 0.42,
+      "eval_loss": 1.4580057859420776,
+      "eval_runtime": 77.4537,
+      "eval_samples_per_second": 1.666,
+      "eval_steps_per_second": 0.116,
+      "step": 3980
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4461623261869258e-05,
+      "loss": 1.4046,
+      "step": 3981
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.445834721473391e-05,
+      "loss": 1.37,
+      "step": 3982
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4455070570294588e-05,
+      "loss": 1.4566,
+      "step": 3983
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.445179332899028e-05,
+      "loss": 1.4513,
+      "step": 3984
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4448515491260051e-05,
+      "loss": 1.3893,
+      "step": 3985
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4445237057543047e-05,
+      "loss": 1.3955,
+      "step": 3986
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4441958028278493e-05,
+      "loss": 1.4861,
+      "step": 3987
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4438678403905697e-05,
+      "loss": 1.4601,
+      "step": 3988
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4435398184864043e-05,
+      "loss": 1.398,
+      "step": 3989
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4432117371592995e-05,
+      "loss": 1.4392,
+      "step": 3990
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4428835964532098e-05,
+      "loss": 1.3772,
+      "step": 3991
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4425553964120979e-05,
+      "loss": 1.4241,
+      "step": 3992
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4422271370799338e-05,
+      "loss": 1.3237,
+      "step": 3993
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4418988185006964e-05,
+      "loss": 1.3813,
+      "step": 3994
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4415704407183709e-05,
+      "loss": 1.6139,
+      "step": 3995
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4412420037769525e-05,
+      "loss": 1.4431,
+      "step": 3996
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.440913507720443e-05,
+      "loss": 1.3862,
+      "step": 3997
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4405849525928523e-05,
+      "loss": 1.3895,
+      "step": 3998
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4402563384381981e-05,
+      "loss": 1.4068,
+      "step": 3999
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4399276653005072e-05,
+      "loss": 1.4501,
+      "step": 4000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4395989332238124e-05,
+      "loss": 1.3165,
+      "step": 4001
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4392701422521555e-05,
+      "loss": 1.3783,
+      "step": 4002
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4389412924295865e-05,
+      "loss": 1.343,
+      "step": 4003
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4386123838001625e-05,
+      "loss": 1.3456,
+      "step": 4004
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4382834164079486e-05,
+      "loss": 1.3689,
+      "step": 4005
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4379543902970183e-05,
+      "loss": 1.3789,
+      "step": 4006
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4376253055114528e-05,
+      "loss": 1.4912,
+      "step": 4007
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4372961620953407e-05,
+      "loss": 1.3692,
+      "step": 4008
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.436966960092779e-05,
+      "loss": 1.4901,
+      "step": 4009
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4366376995478717e-05,
+      "loss": 1.4431,
+      "step": 4010
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.436308380504732e-05,
+      "loss": 1.4626,
+      "step": 4011
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4359790030074798e-05,
+      "loss": 1.4241,
+      "step": 4012
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4356495671002435e-05,
+      "loss": 1.3834,
+      "step": 4013
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4353200728271586e-05,
+      "loss": 1.3626,
+      "step": 4014
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4349905202323692e-05,
+      "loss": 1.3557,
+      "step": 4015
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.434660909360027e-05,
+      "loss": 1.4077,
+      "step": 4016
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4343312402542911e-05,
+      "loss": 1.4337,
+      "step": 4017
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4340015129593286e-05,
+      "loss": 1.419,
+      "step": 4018
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4336717275193148e-05,
+      "loss": 1.3768,
+      "step": 4019
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4333418839784326e-05,
+      "loss": 1.383,
+      "step": 4020
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4330119823808719e-05,
+      "loss": 1.3951,
+      "step": 4021
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4326820227708316e-05,
+      "loss": 1.3985,
+      "step": 4022
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4323520051925177e-05,
+      "loss": 1.356,
+      "step": 4023
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4320219296901439e-05,
+      "loss": 1.5017,
+      "step": 4024
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4316917963079322e-05,
+      "loss": 1.3594,
+      "step": 4025
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4313616050901114e-05,
+      "loss": 1.6028,
+      "step": 4026
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4310313560809193e-05,
+      "loss": 1.4132,
+      "step": 4027
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4307010493246006e-05,
+      "loss": 1.2969,
+      "step": 4028
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4303706848654073e-05,
+      "loss": 1.5224,
+      "step": 4029
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4300402627476007e-05,
+      "loss": 1.3386,
+      "step": 4030
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.429709783015448e-05,
+      "loss": 1.3554,
+      "step": 4031
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4293792457132259e-05,
+      "loss": 1.3386,
+      "step": 4032
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.429048650885217e-05,
+      "loss": 1.4613,
+      "step": 4033
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4287179985757132e-05,
+      "loss": 1.3632,
+      "step": 4034
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.428387288829013e-05,
+      "loss": 1.3835,
+      "step": 4035
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4280565216894234e-05,
+      "loss": 1.3711,
+      "step": 4036
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4277256972012582e-05,
+      "loss": 1.3385,
+      "step": 4037
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4273948154088395e-05,
+      "loss": 1.4607,
+      "step": 4038
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4270638763564973e-05,
+      "loss": 1.4042,
+      "step": 4039
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4267328800885687e-05,
+      "loss": 1.4006,
+      "step": 4040
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4264018266493987e-05,
+      "loss": 1.386,
+      "step": 4041
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4260707160833401e-05,
+      "loss": 1.3487,
+      "step": 4042
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4257395484347529e-05,
+      "loss": 1.4199,
+      "step": 4043
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4254083237480054e-05,
+      "loss": 1.3236,
+      "step": 4044
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4250770420674726e-05,
+      "loss": 1.4056,
+      "step": 4045
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4247457034375387e-05,
+      "loss": 1.4391,
+      "step": 4046
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4244143079025938e-05,
+      "loss": 1.3935,
+      "step": 4047
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4240828555070368e-05,
+      "loss": 1.3623,
+      "step": 4048
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4237513462952733e-05,
+      "loss": 1.3254,
+      "step": 4049
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4234197803117174e-05,
+      "loss": 1.4217,
+      "step": 4050
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.4230881576007904e-05,
+      "loss": 1.435,
+      "step": 4051
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.422756478206921e-05,
+      "loss": 1.3687,
+      "step": 4052
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4224247421745464e-05,
+      "loss": 1.3965,
+      "step": 4053
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4220929495481097e-05,
+      "loss": 1.4159,
+      "step": 4054
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4217611003720632e-05,
+      "loss": 1.3632,
+      "step": 4055
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4214291946908661e-05,
+      "loss": 1.403,
+      "step": 4056
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4210972325489854e-05,
+      "loss": 1.4093,
+      "step": 4057
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4207652139908952e-05,
+      "loss": 1.4551,
+      "step": 4058
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4204331390610775e-05,
+      "loss": 1.4162,
+      "step": 4059
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4201010078040216e-05,
+      "loss": 1.4492,
+      "step": 4060
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.419768820264225e-05,
+      "loss": 1.4449,
+      "step": 4061
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4194365764861922e-05,
+      "loss": 1.3382,
+      "step": 4062
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.419104276514435e-05,
+      "loss": 1.3899,
+      "step": 4063
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4187719203934734e-05,
+      "loss": 1.3513,
+      "step": 4064
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4184395081678344e-05,
+      "loss": 1.3954,
+      "step": 4065
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4181070398820528e-05,
+      "loss": 1.442,
+      "step": 4066
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4177745155806708e-05,
+      "loss": 1.3907,
+      "step": 4067
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.417441935308238e-05,
+      "loss": 1.3757,
+      "step": 4068
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4171092991093118e-05,
+      "loss": 1.3507,
+      "step": 4069
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4167766070284568e-05,
+      "loss": 1.4136,
+      "step": 4070
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4164438591102452e-05,
+      "loss": 1.3835,
+      "step": 4071
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4161110553992565e-05,
+      "loss": 1.4981,
+      "step": 4072
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.415778195940078e-05,
+      "loss": 1.4387,
+      "step": 4073
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4154452807773047e-05,
+      "loss": 1.333,
+      "step": 4074
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4151123099555377e-05,
+      "loss": 1.3558,
+      "step": 4075
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4147792835193875e-05,
+      "loss": 1.4655,
+      "step": 4076
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4144462015134708e-05,
+      "loss": 1.4402,
+      "step": 4077
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4141130639824119e-05,
+      "loss": 1.4227,
+      "step": 4078
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4137798709708423e-05,
+      "loss": 1.4039,
+      "step": 4079
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4134466225234023e-05,
+      "loss": 1.3593,
+      "step": 4080
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4131133186847377e-05,
+      "loss": 1.372,
+      "step": 4081
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.412779959499503e-05,
+      "loss": 1.4487,
+      "step": 4082
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4124465450123599e-05,
+      "loss": 1.3972,
+      "step": 4083
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4121130752679772e-05,
+      "loss": 1.3768,
+      "step": 4084
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4117795503110313e-05,
+      "loss": 1.3827,
+      "step": 4085
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4114459701862062e-05,
+      "loss": 1.3922,
+      "step": 4086
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4111123349381925e-05,
+      "loss": 1.3736,
+      "step": 4087
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4107786446116894e-05,
+      "loss": 1.4338,
+      "step": 4088
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4104448992514023e-05,
+      "loss": 1.347,
+      "step": 4089
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.410111098902045e-05,
+      "loss": 1.3915,
+      "step": 4090
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4097772436083377e-05,
+      "loss": 1.3656,
+      "step": 4091
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4094433334150087e-05,
+      "loss": 1.3461,
+      "step": 4092
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4091093683667934e-05,
+      "loss": 1.7345,
+      "step": 4093
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4087753485084343e-05,
+      "loss": 1.6204,
+      "step": 4094
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4084412738846816e-05,
+      "loss": 1.4074,
+      "step": 4095
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.408107144540293e-05,
+      "loss": 1.3776,
+      "step": 4096
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.407772960520033e-05,
+      "loss": 1.4079,
+      "step": 4097
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4074387218686734e-05,
+      "loss": 1.4254,
+      "step": 4098
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.407104428630994e-05,
+      "loss": 1.38,
+      "step": 4099
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4067700808517815e-05,
+      "loss": 1.4175,
+      "step": 4100
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4064356785758298e-05,
+      "loss": 1.4085,
+      "step": 4101
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4061012218479397e-05,
+      "loss": 1.4529,
+      "step": 4102
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.405766710712921e-05,
+      "loss": 1.4836,
+      "step": 4103
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4054321452155885e-05,
+      "loss": 1.3915,
+      "step": 4104
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4050975254007659e-05,
+      "loss": 1.3044,
+      "step": 4105
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4047628513132835e-05,
+      "loss": 1.373,
+      "step": 4106
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4044281229979793e-05,
+      "loss": 1.4266,
+      "step": 4107
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4040933404996977e-05,
+      "loss": 1.3898,
+      "step": 4108
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4037585038632915e-05,
+      "loss": 0.9326,
+      "step": 4109
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.40342361313362e-05,
+      "loss": 1.0831,
+      "step": 4110
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.40308866835555e-05,
+      "loss": 1.3898,
+      "step": 4111
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4027536695739554e-05,
+      "loss": 1.5058,
+      "step": 4112
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4024186168337175e-05,
+      "loss": 1.3635,
+      "step": 4113
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4020835101797248e-05,
+      "loss": 1.3871,
+      "step": 4114
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.401748349656873e-05,
+      "loss": 1.3478,
+      "step": 4115
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4014131353100648e-05,
+      "loss": 1.3356,
+      "step": 4116
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4010778671842107e-05,
+      "loss": 1.3954,
+      "step": 4117
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4007425453242274e-05,
+      "loss": 1.3307,
+      "step": 4118
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4004071697750398e-05,
+      "loss": 1.4508,
+      "step": 4119
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.4000717405815799e-05,
+      "loss": 1.3752,
+      "step": 4120
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.399736257788786e-05,
+      "loss": 1.3654,
+      "step": 4121
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.3994007214416046e-05,
+      "loss": 1.4505,
+      "step": 4122
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.3990651315849886e-05,
+      "loss": 1.4307,
+      "step": 4123
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.398729488263899e-05,
+      "loss": 1.4481,
+      "step": 4124
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.3983937915233026e-05,
+      "loss": 1.4755,
+      "step": 4125
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.3980580414081746e-05,
+      "loss": 1.3402,
+      "step": 4126
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.3977222379634969e-05,
+      "loss": 1.3636,
+      "step": 4127
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.3973863812342585e-05,
+      "loss": 1.4165,
+      "step": 4128
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.3970504712654554e-05,
+      "loss": 1.3832,
+      "step": 4129
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.3967145081020912e-05,
+      "loss": 1.4236,
+      "step": 4130
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.396378491789176e-05,
+      "loss": 1.5022,
+      "step": 4131
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.3960424223717276e-05,
+      "loss": 1.3806,
+      "step": 4132
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.3957062998947707e-05,
+      "loss": 1.3649,
+      "step": 4133
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.3953701244033368e-05,
+      "loss": 1.3787,
+      "step": 4134
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.395033895942465e-05,
+      "loss": 1.3551,
+      "step": 4135
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.394697614557201e-05,
+      "loss": 1.3881,
+      "step": 4136
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.3943612802925983e-05,
+      "loss": 1.4848,
+      "step": 4137
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.3940248931937169e-05,
+      "loss": 1.425,
+      "step": 4138
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.3936884533056238e-05,
+      "loss": 1.4054,
+      "step": 4139
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.3933519606733934e-05,
+      "loss": 1.4028,
+      "step": 4140
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.3930154153421072e-05,
+      "loss": 1.4439,
+      "step": 4141
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.3926788173568536e-05,
+      "loss": 1.4482,
+      "step": 4142
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.392342166762728e-05,
+      "loss": 1.393,
+      "step": 4143
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.3920054636048328e-05,
+      "loss": 1.4097,
+      "step": 4144
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.391668707928278e-05,
+      "loss": 1.5142,
+      "step": 4145
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.3913318997781797e-05,
+      "loss": 1.4118,
+      "step": 4146
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.390995039199662e-05,
+      "loss": 1.4067,
+      "step": 4147
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.3906581262378552e-05,
+      "loss": 1.5214,
+      "step": 4148
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3903211609378971e-05,
+      "loss": 1.3618,
+      "step": 4149
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3899841433449327e-05,
+      "loss": 1.338,
+      "step": 4150
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3896470735041132e-05,
+      "loss": 1.3901,
+      "step": 4151
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3893099514605971e-05,
+      "loss": 1.4502,
+      "step": 4152
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.388972777259551e-05,
+      "loss": 1.4141,
+      "step": 4153
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3886355509461471e-05,
+      "loss": 1.3876,
+      "step": 4154
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.388298272565565e-05,
+      "loss": 1.4275,
+      "step": 4155
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3879609421629909e-05,
+      "loss": 1.3559,
+      "step": 4156
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3876235597836195e-05,
+      "loss": 1.416,
+      "step": 4157
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3872861254726506e-05,
+      "loss": 1.437,
+      "step": 4158
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3869486392752918e-05,
+      "loss": 1.3794,
+      "step": 4159
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3866111012367577e-05,
+      "loss": 1.4094,
+      "step": 4160
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3862735114022694e-05,
+      "loss": 1.3957,
+      "step": 4161
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3859358698170559e-05,
+      "loss": 1.3582,
+      "step": 4162
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3855981765263523e-05,
+      "loss": 1.4159,
+      "step": 4163
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3852604315754e-05,
+      "loss": 1.3498,
+      "step": 4164
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3849226350094491e-05,
+      "loss": 1.3951,
+      "step": 4165
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3845847868737559e-05,
+      "loss": 1.4087,
+      "step": 4166
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3842468872135821e-05,
+      "loss": 1.4318,
+      "step": 4167
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3839089360741989e-05,
+      "loss": 1.4174,
+      "step": 4168
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3835709335008822e-05,
+      "loss": 1.4499,
+      "step": 4169
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.383232879538916e-05,
+      "loss": 1.3554,
+      "step": 4170
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3828947742335905e-05,
+      "loss": 1.3873,
+      "step": 4171
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.382556617630204e-05,
+      "loss": 1.3804,
+      "step": 4172
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3822184097740598e-05,
+      "loss": 1.3856,
+      "step": 4173
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3818801507104697e-05,
+      "loss": 1.3901,
+      "step": 4174
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.381541840484751e-05,
+      "loss": 1.3714,
+      "step": 4175
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3812034791422297e-05,
+      "loss": 1.4945,
+      "step": 4176
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3808650667282368e-05,
+      "loss": 1.429,
+      "step": 4177
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3805266032881108e-05,
+      "loss": 1.3903,
+      "step": 4178
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3801880888671972e-05,
+      "loss": 1.4799,
+      "step": 4179
+    },
+    {
+      "epoch": 0.44,
+      "eval_loss": 1.4521424770355225,
+      "eval_runtime": 77.3596,
+      "eval_samples_per_second": 1.668,
+      "eval_steps_per_second": 0.116,
+      "step": 4179
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3798495235108484e-05,
+      "loss": 1.3912,
+      "step": 4180
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3795109072644235e-05,
+      "loss": 1.5837,
+      "step": 4181
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.379172240173288e-05,
+      "loss": 1.4255,
+      "step": 4182
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3788335222828148e-05,
+      "loss": 1.4666,
+      "step": 4183
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3784947536383837e-05,
+      "loss": 1.3964,
+      "step": 4184
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3781559342853807e-05,
+      "loss": 1.336,
+      "step": 4185
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3778170642691986e-05,
+      "loss": 1.4352,
+      "step": 4186
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.377478143635238e-05,
+      "loss": 1.3262,
+      "step": 4187
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3771391724289047e-05,
+      "loss": 1.3969,
+      "step": 4188
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.376800150695613e-05,
+      "loss": 1.4211,
+      "step": 4189
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3764610784807824e-05,
+      "loss": 1.4338,
+      "step": 4190
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3761219558298403e-05,
+      "loss": 1.3815,
+      "step": 4191
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3757827827882203e-05,
+      "loss": 1.3805,
+      "step": 4192
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3754435594013627e-05,
+      "loss": 1.4071,
+      "step": 4193
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3751042857147145e-05,
+      "loss": 1.343,
+      "step": 4194
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3747649617737304e-05,
+      "loss": 1.4129,
+      "step": 4195
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3744255876238708e-05,
+      "loss": 1.3939,
+      "step": 4196
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3740861633106029e-05,
+      "loss": 1.3963,
+      "step": 4197
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3737466888794008e-05,
+      "loss": 1.5181,
+      "step": 4198
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3734071643757456e-05,
+      "loss": 1.4397,
+      "step": 4199
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.373067589845125e-05,
+      "loss": 1.6293,
+      "step": 4200
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3727279653330328e-05,
+      "loss": 1.3498,
+      "step": 4201
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.37238829088497e-05,
+      "loss": 1.4509,
+      "step": 4202
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3720485665464446e-05,
+      "loss": 1.4477,
+      "step": 4203
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3717087923629709e-05,
+      "loss": 1.4399,
+      "step": 4204
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3713689683800695e-05,
+      "loss": 1.4417,
+      "step": 4205
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3710290946432685e-05,
+      "loss": 1.3957,
+      "step": 4206
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3706891711981021e-05,
+      "loss": 1.4476,
+      "step": 4207
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3703491980901113e-05,
+      "loss": 1.3894,
+      "step": 4208
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3700091753648437e-05,
+      "loss": 1.4745,
+      "step": 4209
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.369669103067854e-05,
+      "loss": 1.4038,
+      "step": 4210
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3693289812447023e-05,
+      "loss": 1.4151,
+      "step": 4211
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3689888099409572e-05,
+      "loss": 1.4214,
+      "step": 4212
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.368648589202192e-05,
+      "loss": 1.4287,
+      "step": 4213
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3683083190739883e-05,
+      "loss": 1.345,
+      "step": 4214
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.367967999601933e-05,
+      "loss": 1.3022,
+      "step": 4215
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3676276308316207e-05,
+      "loss": 1.3971,
+      "step": 4216
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3672872128086514e-05,
+      "loss": 1.3335,
+      "step": 4217
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.366946745578633e-05,
+      "loss": 1.3988,
+      "step": 4218
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.366606229187179e-05,
+      "loss": 1.4508,
+      "step": 4219
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3662656636799099e-05,
+      "loss": 1.3814,
+      "step": 4220
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3659250491024527e-05,
+      "loss": 1.4128,
+      "step": 4221
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3655843855004411e-05,
+      "loss": 1.3893,
+      "step": 4222
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3652436729195154e-05,
+      "loss": 1.2551,
+      "step": 4223
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3649029114053219e-05,
+      "loss": 1.4386,
+      "step": 4224
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.364562101003514e-05,
+      "loss": 1.4028,
+      "step": 4225
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.364221241759752e-05,
+      "loss": 1.4351,
+      "step": 4226
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.363880333719702e-05,
+      "loss": 1.3941,
+      "step": 4227
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3635393769290368e-05,
+      "loss": 1.3879,
+      "step": 4228
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3631983714334358e-05,
+      "loss": 1.3773,
+      "step": 4229
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3628573172785853e-05,
+      "loss": 1.3679,
+      "step": 4230
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3625162145101776e-05,
+      "loss": 1.3705,
+      "step": 4231
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3621750631739114e-05,
+      "loss": 1.8157,
+      "step": 4232
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3618338633154928e-05,
+      "loss": 1.4056,
+      "step": 4233
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3614926149806334e-05,
+      "loss": 1.3716,
+      "step": 4234
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3611513182150523e-05,
+      "loss": 1.3766,
+      "step": 4235
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3608099730644733e-05,
+      "loss": 1.41,
+      "step": 4236
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3604685795746291e-05,
+      "loss": 1.421,
+      "step": 4237
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.360127137791257e-05,
+      "loss": 1.3679,
+      "step": 4238
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3597856477601018e-05,
+      "loss": 1.4615,
+      "step": 4239
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.359444109526914e-05,
+      "loss": 1.3418,
+      "step": 4240
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3591025231374513e-05,
+      "loss": 1.3631,
+      "step": 4241
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3587608886374772e-05,
+      "loss": 1.3802,
+      "step": 4242
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.3584192060727622e-05,
+      "loss": 1.4156,
+      "step": 4243
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3580774754890827e-05,
+      "loss": 1.4161,
+      "step": 4244
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3577356969322222e-05,
+      "loss": 1.3535,
+      "step": 4245
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3573938704479699e-05,
+      "loss": 1.3859,
+      "step": 4246
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.357051996082122e-05,
+      "loss": 1.3595,
+      "step": 4247
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3567100738804806e-05,
+      "loss": 1.4191,
+      "step": 4248
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3563681038888548e-05,
+      "loss": 1.4186,
+      "step": 4249
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3560260861530596e-05,
+      "loss": 1.43,
+      "step": 4250
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3556840207189168e-05,
+      "loss": 1.4139,
+      "step": 4251
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3553419076322539e-05,
+      "loss": 1.4005,
+      "step": 4252
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3549997469389058e-05,
+      "loss": 1.3468,
+      "step": 4253
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3546575386847132e-05,
+      "loss": 1.3756,
+      "step": 4254
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3543152829155226e-05,
+      "loss": 1.4063,
+      "step": 4255
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3539729796771884e-05,
+      "loss": 1.4666,
+      "step": 4256
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3536306290155694e-05,
+      "loss": 1.4212,
+      "step": 4257
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3532882309765332e-05,
+      "loss": 1.4883,
+      "step": 4258
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3529457856059507e-05,
+      "loss": 1.6641,
+      "step": 4259
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3526032929497022e-05,
+      "loss": 1.3659,
+      "step": 4260
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3522607530536717e-05,
+      "loss": 1.5945,
+      "step": 4261
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.351918165963752e-05,
+      "loss": 1.3643,
+      "step": 4262
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3515755317258397e-05,
+      "loss": 1.3442,
+      "step": 4263
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3512328503858398e-05,
+      "loss": 1.3982,
+      "step": 4264
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3508901219896624e-05,
+      "loss": 1.4985,
+      "step": 4265
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.350547346583225e-05,
+      "loss": 1.3319,
+      "step": 4266
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3502045242124497e-05,
+      "loss": 1.3961,
+      "step": 4267
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3498616549232665e-05,
+      "loss": 1.4565,
+      "step": 4268
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.349518738761611e-05,
+      "loss": 1.375,
+      "step": 4269
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3491757757734254e-05,
+      "loss": 1.3005,
+      "step": 4270
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.348832766004657e-05,
+      "loss": 1.3185,
+      "step": 4271
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3484897095012613e-05,
+      "loss": 1.4449,
+      "step": 4272
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3481466063091988e-05,
+      "loss": 1.3587,
+      "step": 4273
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3478034564744362e-05,
+      "loss": 1.3844,
+      "step": 4274
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3474602600429468e-05,
+      "loss": 1.4471,
+      "step": 4275
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3471170170607105e-05,
+      "loss": 1.4686,
+      "step": 4276
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3467737275737127e-05,
+      "loss": 1.312,
+      "step": 4277
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3464303916279453e-05,
+      "loss": 1.407,
+      "step": 4278
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.346087009269407e-05,
+      "loss": 1.364,
+      "step": 4279
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3457435805441016e-05,
+      "loss": 1.4103,
+      "step": 4280
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3454001054980403e-05,
+      "loss": 1.4243,
+      "step": 4281
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3450565841772392e-05,
+      "loss": 1.3879,
+      "step": 4282
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.344713016627722e-05,
+      "loss": 1.4688,
+      "step": 4283
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3443694028955176e-05,
+      "loss": 1.3983,
+      "step": 4284
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.344025743026662e-05,
+      "loss": 1.4316,
+      "step": 4285
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3436820370671956e-05,
+      "loss": 1.3643,
+      "step": 4286
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3433382850631673e-05,
+      "loss": 1.4566,
+      "step": 4287
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3429944870606301e-05,
+      "loss": 1.1263,
+      "step": 4288
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3426506431056452e-05,
+      "loss": 1.2463,
+      "step": 4289
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.342306753244278e-05,
+      "loss": 1.4753,
+      "step": 4290
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3419628175226012e-05,
+      "loss": 1.3789,
+      "step": 4291
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3416188359866932e-05,
+      "loss": 1.3906,
+      "step": 4292
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3412748086826389e-05,
+      "loss": 1.3995,
+      "step": 4293
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3409307356565288e-05,
+      "loss": 1.4906,
+      "step": 4294
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3405866169544602e-05,
+      "loss": 1.3951,
+      "step": 4295
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3402424526225359e-05,
+      "loss": 1.3753,
+      "step": 4296
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3398982427068654e-05,
+      "loss": 1.3686,
+      "step": 4297
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3395539872535636e-05,
+      "loss": 1.3612,
+      "step": 4298
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3392096863087522e-05,
+      "loss": 1.4043,
+      "step": 4299
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3388653399185585e-05,
+      "loss": 1.4068,
+      "step": 4300
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.338520948129116e-05,
+      "loss": 1.4066,
+      "step": 4301
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3381765109865646e-05,
+      "loss": 1.4944,
+      "step": 4302
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3378320285370496e-05,
+      "loss": 1.3575,
+      "step": 4303
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3374875008267236e-05,
+      "loss": 1.3807,
+      "step": 4304
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3371429279017437e-05,
+      "loss": 1.4389,
+      "step": 4305
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3367983098082739e-05,
+      "loss": 1.4858,
+      "step": 4306
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3364536465924844e-05,
+      "loss": 1.3282,
+      "step": 4307
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3361089383005515e-05,
+      "loss": 1.4306,
+      "step": 4308
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3357641849786566e-05,
+      "loss": 1.3892,
+      "step": 4309
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3354193866729883e-05,
+      "loss": 1.3621,
+      "step": 4310
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3350745434297401e-05,
+      "loss": 1.3476,
+      "step": 4311
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.334729655295113e-05,
+      "loss": 1.4477,
+      "step": 4312
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3343847223153123e-05,
+      "loss": 1.3585,
+      "step": 4313
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.334039744536551e-05,
+      "loss": 1.3616,
+      "step": 4314
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3336947220050462e-05,
+      "loss": 1.3579,
+      "step": 4315
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3333496547670232e-05,
+      "loss": 1.3557,
+      "step": 4316
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3330045428687113e-05,
+      "loss": 1.4443,
+      "step": 4317
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.332659386356347e-05,
+      "loss": 1.3621,
+      "step": 4318
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3323141852761726e-05,
+      "loss": 1.3295,
+      "step": 4319
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3319689396744358e-05,
+      "loss": 1.4187,
+      "step": 4320
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3316236495973905e-05,
+      "loss": 1.335,
+      "step": 4321
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.331278315091297e-05,
+      "loss": 1.4382,
+      "step": 4322
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3309329362024212e-05,
+      "loss": 1.4244,
+      "step": 4323
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3305875129770351e-05,
+      "loss": 1.488,
+      "step": 4324
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3302420454614165e-05,
+      "loss": 1.3369,
+      "step": 4325
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3298965337018487e-05,
+      "loss": 1.3648,
+      "step": 4326
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3295509777446226e-05,
+      "loss": 1.3924,
+      "step": 4327
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3292053776360324e-05,
+      "loss": 1.3789,
+      "step": 4328
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3288597334223802e-05,
+      "loss": 1.3902,
+      "step": 4329
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3285140451499735e-05,
+      "loss": 1.4057,
+      "step": 4330
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.328168312865126e-05,
+      "loss": 1.4536,
+      "step": 4331
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3278225366141565e-05,
+      "loss": 1.3075,
+      "step": 4332
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3274767164433901e-05,
+      "loss": 1.4579,
+      "step": 4333
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3271308523991579e-05,
+      "loss": 1.3906,
+      "step": 4334
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.326784944527797e-05,
+      "loss": 1.3992,
+      "step": 4335
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3264389928756498e-05,
+      "loss": 1.4001,
+      "step": 4336
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3260929974890653e-05,
+      "loss": 1.2948,
+      "step": 4337
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.3257469584143976e-05,
+      "loss": 1.4041,
+      "step": 4338
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3254008756980076e-05,
+      "loss": 1.5035,
+      "step": 4339
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3250547493862609e-05,
+      "loss": 1.4867,
+      "step": 4340
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3247085795255302e-05,
+      "loss": 1.3612,
+      "step": 4341
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3243623661621925e-05,
+      "loss": 1.4193,
+      "step": 4342
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3240161093426322e-05,
+      "loss": 1.3596,
+      "step": 4343
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3236698091132383e-05,
+      "loss": 1.4448,
+      "step": 4344
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3233234655204066e-05,
+      "loss": 1.3924,
+      "step": 4345
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.322977078610538e-05,
+      "loss": 1.3552,
+      "step": 4346
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3226306484300395e-05,
+      "loss": 1.3949,
+      "step": 4347
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.322284175025324e-05,
+      "loss": 1.399,
+      "step": 4348
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3219376584428094e-05,
+      "loss": 1.4755,
+      "step": 4349
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3215910987289212e-05,
+      "loss": 1.3877,
+      "step": 4350
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3212444959300882e-05,
+      "loss": 1.3916,
+      "step": 4351
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3208978500927472e-05,
+      "loss": 1.3462,
+      "step": 4352
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3205511612633392e-05,
+      "loss": 1.3834,
+      "step": 4353
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3202044294883123e-05,
+      "loss": 1.3843,
+      "step": 4354
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3198576548141189e-05,
+      "loss": 1.2439,
+      "step": 4355
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3195108372872181e-05,
+      "loss": 1.4106,
+      "step": 4356
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3191639769540746e-05,
+      "loss": 1.4417,
+      "step": 4357
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3188170738611596e-05,
+      "loss": 1.4393,
+      "step": 4358
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3184701280549476e-05,
+      "loss": 1.4616,
+      "step": 4359
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3181231395819218e-05,
+      "loss": 1.3215,
+      "step": 4360
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3177761084885689e-05,
+      "loss": 1.368,
+      "step": 4361
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.317429034821383e-05,
+      "loss": 1.3247,
+      "step": 4362
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3170819186268618e-05,
+      "loss": 1.4512,
+      "step": 4363
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3167347599515111e-05,
+      "loss": 1.3129,
+      "step": 4364
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3163875588418404e-05,
+      "loss": 1.3934,
+      "step": 4365
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3160403153443666e-05,
+      "loss": 1.4518,
+      "step": 4366
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3156930295056105e-05,
+      "loss": 1.3959,
+      "step": 4367
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3153457013721e-05,
+      "loss": 1.3475,
+      "step": 4368
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3149983309903687e-05,
+      "loss": 1.4102,
+      "step": 4369
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3146509184069538e-05,
+      "loss": 1.3674,
+      "step": 4370
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.314303463668401e-05,
+      "loss": 1.3739,
+      "step": 4371
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3139559668212596e-05,
+      "loss": 1.4249,
+      "step": 4372
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3136084279120857e-05,
+      "loss": 1.4473,
+      "step": 4373
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3132608469874404e-05,
+      "loss": 1.4443,
+      "step": 4374
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3129132240938905e-05,
+      "loss": 1.3534,
+      "step": 4375
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3125655592780083e-05,
+      "loss": 1.5663,
+      "step": 4376
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3122178525863727e-05,
+      "loss": 1.3465,
+      "step": 4377
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.3118701040655669e-05,
+      "loss": 1.35,
+      "step": 4378
+    },
+    {
+      "epoch": 0.46,
+      "eval_loss": 1.447788953781128,
+      "eval_runtime": 77.4473,
+      "eval_samples_per_second": 1.666,
+      "eval_steps_per_second": 0.116,
+      "step": 4378
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 9536,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 398,
+  "total_flos": 1.9585598163698634e+20,
+  "trial_name": null,
+  "trial_params": null
+}