diff --git "a/checkpoint-796/trainer_state.json" "b/checkpoint-796/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-796/trainer_state.json"
@@ -0,0 +1,4828 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.0834731543624161,
+  "eval_steps": 199,
+  "global_step": 796,
+  "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
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 9536,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 398,
+  "total_flos": 3.561017847945206e+19,
+  "trial_name": null,
+  "trial_params": null
+}