diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,54714 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.999936,
+  "eval_steps": 500,
+  "global_step": 7812,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 48.072471618652344,
+      "learning_rate": 2.1276595744680853e-06,
+      "loss": 9.3378,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 43.48775863647461,
+      "learning_rate": 4.255319148936171e-06,
+      "loss": 9.3274,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 43.6014518737793,
+      "learning_rate": 6.3829787234042555e-06,
+      "loss": 9.3945,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 44.38250732421875,
+      "learning_rate": 8.510638297872341e-06,
+      "loss": 9.3164,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 38.3822135925293,
+      "learning_rate": 1.0638297872340426e-05,
+      "loss": 9.0362,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 32.224544525146484,
+      "learning_rate": 1.2765957446808511e-05,
+      "loss": 9.0261,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 29.361282348632812,
+      "learning_rate": 1.4893617021276598e-05,
+      "loss": 8.3959,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 33.011817932128906,
+      "learning_rate": 1.7021276595744682e-05,
+      "loss": 8.6945,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 38.45726013183594,
+      "learning_rate": 1.9148936170212766e-05,
+      "loss": 8.4528,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 40.37697982788086,
+      "learning_rate": 2.1276595744680852e-05,
+      "loss": 7.6175,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 34.739898681640625,
+      "learning_rate": 2.3404255319148935e-05,
+      "loss": 7.2495,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 25.367467880249023,
+      "learning_rate": 2.5531914893617022e-05,
+      "loss": 7.1284,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.876768112182617,
+      "learning_rate": 2.7659574468085105e-05,
+      "loss": 6.8116,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 20.27662467956543,
+      "learning_rate": 2.9787234042553195e-05,
+      "loss": 6.7344,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 25.331371307373047,
+      "learning_rate": 3.1914893617021275e-05,
+      "loss": 6.4643,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 24.46682357788086,
+      "learning_rate": 3.4042553191489365e-05,
+      "loss": 6.318,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.2562837600708,
+      "learning_rate": 3.617021276595744e-05,
+      "loss": 5.8803,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.656604766845703,
+      "learning_rate": 3.829787234042553e-05,
+      "loss": 5.6743,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.968253135681152,
+      "learning_rate": 4.042553191489362e-05,
+      "loss": 5.4764,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.436161041259766,
+      "learning_rate": 4.2553191489361704e-05,
+      "loss": 5.4851,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.585102558135986,
+      "learning_rate": 4.468085106382979e-05,
+      "loss": 5.2845,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.892800807952881,
+      "learning_rate": 4.680851063829787e-05,
+      "loss": 5.1436,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.763721942901611,
+      "learning_rate": 4.893617021276596e-05,
+      "loss": 5.1792,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.559952735900879,
+      "learning_rate": 5.1063829787234044e-05,
+      "loss": 4.788,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.66323184967041,
+      "learning_rate": 5.319148936170213e-05,
+      "loss": 5.0929,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.006740093231201,
+      "learning_rate": 5.531914893617021e-05,
+      "loss": 5.1313,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.960521697998047,
+      "learning_rate": 5.74468085106383e-05,
+      "loss": 4.7744,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6684303283691406,
+      "learning_rate": 5.957446808510639e-05,
+      "loss": 4.7888,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.18186616897583,
+      "learning_rate": 6.170212765957447e-05,
+      "loss": 4.7324,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.048969268798828,
+      "learning_rate": 6.382978723404255e-05,
+      "loss": 4.7644,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6776039600372314,
+      "learning_rate": 6.595744680851063e-05,
+      "loss": 4.8169,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.373857021331787,
+      "learning_rate": 6.808510638297873e-05,
+      "loss": 4.7869,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.241915702819824,
+      "learning_rate": 7.021276595744681e-05,
+      "loss": 4.6614,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.0473105907440186,
+      "learning_rate": 7.234042553191488e-05,
+      "loss": 4.6913,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8055814504623413,
+      "learning_rate": 7.446808510638298e-05,
+      "loss": 4.4816,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.7576919794082642,
+      "learning_rate": 7.659574468085106e-05,
+      "loss": 4.4663,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6539785861968994,
+      "learning_rate": 7.872340425531916e-05,
+      "loss": 4.4183,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6569417715072632,
+      "learning_rate": 8.085106382978724e-05,
+      "loss": 4.5726,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.4927167892456055,
+      "learning_rate": 8.297872340425531e-05,
+      "loss": 4.5231,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.593092679977417,
+      "learning_rate": 8.510638297872341e-05,
+      "loss": 4.4761,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4695898294448853,
+      "learning_rate": 8.723404255319149e-05,
+      "loss": 4.309,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.461921215057373,
+      "learning_rate": 8.936170212765958e-05,
+      "loss": 4.591,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3230671882629395,
+      "learning_rate": 9.148936170212766e-05,
+      "loss": 4.5867,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3359841108322144,
+      "learning_rate": 9.361702127659574e-05,
+      "loss": 4.5312,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1943374872207642,
+      "learning_rate": 9.574468085106382e-05,
+      "loss": 4.444,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1472824811935425,
+      "learning_rate": 9.787234042553192e-05,
+      "loss": 4.5007,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.113065481185913,
+      "learning_rate": 0.0001,
+      "loss": 4.4621,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2492742538452148,
+      "learning_rate": 0.00010212765957446809,
+      "loss": 4.5417,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.063219428062439,
+      "learning_rate": 0.00010425531914893617,
+      "loss": 4.3506,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3006094694137573,
+      "learning_rate": 0.00010638297872340425,
+      "loss": 4.2705,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0823702812194824,
+      "learning_rate": 0.00010851063829787235,
+      "loss": 4.5639,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.171970248222351,
+      "learning_rate": 0.00011063829787234042,
+      "loss": 4.3682,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0258631706237793,
+      "learning_rate": 0.0001127659574468085,
+      "loss": 4.4626,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9669986963272095,
+      "learning_rate": 0.0001148936170212766,
+      "loss": 4.5103,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8417139649391174,
+      "learning_rate": 0.00011702127659574468,
+      "loss": 4.3414,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9252743721008301,
+      "learning_rate": 0.00011914893617021278,
+      "loss": 4.4555,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9143895506858826,
+      "learning_rate": 0.00012127659574468085,
+      "loss": 4.3624,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8701967597007751,
+      "learning_rate": 0.00012340425531914893,
+      "loss": 4.3983,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8114748001098633,
+      "learning_rate": 0.00012553191489361702,
+      "loss": 4.2508,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.831774115562439,
+      "learning_rate": 0.0001276595744680851,
+      "loss": 4.284,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8641481399536133,
+      "learning_rate": 0.00012978723404255318,
+      "loss": 4.3339,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7552957534790039,
+      "learning_rate": 0.00013191489361702127,
+      "loss": 4.3536,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7908809781074524,
+      "learning_rate": 0.00013404255319148938,
+      "loss": 4.4432,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7348148822784424,
+      "learning_rate": 0.00013617021276595746,
+      "loss": 4.369,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8431515693664551,
+      "learning_rate": 0.00013829787234042554,
+      "loss": 4.4683,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6354610919952393,
+      "learning_rate": 0.00014042553191489363,
+      "loss": 4.2249,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7959691286087036,
+      "learning_rate": 0.0001425531914893617,
+      "loss": 4.4541,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6812201738357544,
+      "learning_rate": 0.00014468085106382977,
+      "loss": 4.3104,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6995226740837097,
+      "learning_rate": 0.00014680851063829788,
+      "loss": 4.2944,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6939712166786194,
+      "learning_rate": 0.00014893617021276596,
+      "loss": 4.3758,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6890800595283508,
+      "learning_rate": 0.00015106382978723404,
+      "loss": 4.4136,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7452850937843323,
+      "learning_rate": 0.00015319148936170213,
+      "loss": 4.179,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7114067077636719,
+      "learning_rate": 0.0001553191489361702,
+      "loss": 4.4111,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6431289911270142,
+      "learning_rate": 0.00015744680851063832,
+      "loss": 4.3288,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7086479067802429,
+      "learning_rate": 0.0001595744680851064,
+      "loss": 4.321,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6875836849212646,
+      "learning_rate": 0.00016170212765957449,
+      "loss": 4.4143,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.697288990020752,
+      "learning_rate": 0.00016382978723404254,
+      "loss": 4.2854,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6955673694610596,
+      "learning_rate": 0.00016595744680851062,
+      "loss": 4.3061,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6468881368637085,
+      "learning_rate": 0.0001680851063829787,
+      "loss": 4.3326,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.678005039691925,
+      "learning_rate": 0.00017021276595744682,
+      "loss": 4.356,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6774478554725647,
+      "learning_rate": 0.0001723404255319149,
+      "loss": 4.2089,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7283062934875488,
+      "learning_rate": 0.00017446808510638298,
+      "loss": 4.3926,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7186111807823181,
+      "learning_rate": 0.00017659574468085107,
+      "loss": 4.3649,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6334117650985718,
+      "learning_rate": 0.00017872340425531915,
+      "loss": 4.3125,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6184477806091309,
+      "learning_rate": 0.00018085106382978726,
+      "loss": 4.3224,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6729727983474731,
+      "learning_rate": 0.00018297872340425532,
+      "loss": 4.2833,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.720475971698761,
+      "learning_rate": 0.0001851063829787234,
+      "loss": 4.3743,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.699727475643158,
+      "learning_rate": 0.00018723404255319148,
+      "loss": 4.3787,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7167297005653381,
+      "learning_rate": 0.00018936170212765957,
+      "loss": 4.4618,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7404178977012634,
+      "learning_rate": 0.00019148936170212765,
+      "loss": 4.4533,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7304544448852539,
+      "learning_rate": 0.00019361702127659576,
+      "loss": 4.4291,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7182232141494751,
+      "learning_rate": 0.00019574468085106384,
+      "loss": 4.3131,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7111774682998657,
+      "learning_rate": 0.00019787234042553193,
+      "loss": 4.2267,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7227694988250732,
+      "learning_rate": 0.0002,
+      "loss": 4.2252,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6536330580711365,
+      "learning_rate": 0.00020212765957446807,
+      "loss": 4.2824,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7753093242645264,
+      "learning_rate": 0.00020425531914893618,
+      "loss": 4.2619,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6675902009010315,
+      "learning_rate": 0.00020638297872340426,
+      "loss": 4.1745,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7651579976081848,
+      "learning_rate": 0.00020851063829787234,
+      "loss": 4.4243,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6589416265487671,
+      "learning_rate": 0.00021063829787234043,
+      "loss": 4.2501,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6965423822402954,
+      "learning_rate": 0.0002127659574468085,
+      "loss": 4.2279,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7227066159248352,
+      "learning_rate": 0.0002148936170212766,
+      "loss": 4.3669,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7057831287384033,
+      "learning_rate": 0.0002170212765957447,
+      "loss": 4.1575,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7887274622917175,
+      "learning_rate": 0.00021914893617021279,
+      "loss": 4.2733,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.601609468460083,
+      "learning_rate": 0.00022127659574468084,
+      "loss": 4.0538,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.783682107925415,
+      "learning_rate": 0.00022340425531914892,
+      "loss": 4.2188,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7353456616401672,
+      "learning_rate": 0.000225531914893617,
+      "loss": 4.318,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7265023589134216,
+      "learning_rate": 0.00022765957446808512,
+      "loss": 4.1894,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.710670530796051,
+      "learning_rate": 0.0002297872340425532,
+      "loss": 4.2075,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7732564806938171,
+      "learning_rate": 0.00023191489361702128,
+      "loss": 4.193,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8331496715545654,
+      "learning_rate": 0.00023404255319148937,
+      "loss": 4.313,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8698146939277649,
+      "learning_rate": 0.00023617021276595745,
+      "loss": 4.1215,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7106320261955261,
+      "learning_rate": 0.00023829787234042556,
+      "loss": 4.1094,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8148285150527954,
+      "learning_rate": 0.00024042553191489362,
+      "loss": 4.3917,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8150539994239807,
+      "learning_rate": 0.0002425531914893617,
+      "loss": 4.2951,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7719513773918152,
+      "learning_rate": 0.0002446808510638298,
+      "loss": 4.2279,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9067126512527466,
+      "learning_rate": 0.00024680851063829787,
+      "loss": 4.2159,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8653740286827087,
+      "learning_rate": 0.000248936170212766,
+      "loss": 4.3633,
+      "step": 117
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8690366744995117,
+      "learning_rate": 0.00025106382978723403,
+      "loss": 4.2861,
+      "step": 118
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8911768794059753,
+      "learning_rate": 0.0002531914893617021,
+      "loss": 4.2543,
+      "step": 119
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9171234965324402,
+      "learning_rate": 0.0002553191489361702,
+      "loss": 4.1359,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9806623458862305,
+      "learning_rate": 0.0002574468085106383,
+      "loss": 4.2147,
+      "step": 121
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9419201016426086,
+      "learning_rate": 0.00025957446808510637,
+      "loss": 4.1773,
+      "step": 122
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9396133422851562,
+      "learning_rate": 0.0002617021276595745,
+      "loss": 4.2412,
+      "step": 123
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9173053503036499,
+      "learning_rate": 0.00026382978723404253,
+      "loss": 4.3375,
+      "step": 124
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.922222912311554,
+      "learning_rate": 0.00026595744680851064,
+      "loss": 4.0584,
+      "step": 125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0579538345336914,
+      "learning_rate": 0.00026808510638297875,
+      "loss": 4.2104,
+      "step": 126
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.862234890460968,
+      "learning_rate": 0.0002702127659574468,
+      "loss": 4.2614,
+      "step": 127
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0590007305145264,
+      "learning_rate": 0.0002723404255319149,
+      "loss": 4.3827,
+      "step": 128
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0478700399398804,
+      "learning_rate": 0.000274468085106383,
+      "loss": 4.1992,
+      "step": 129
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0220136642456055,
+      "learning_rate": 0.0002765957446808511,
+      "loss": 4.2198,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0901641845703125,
+      "learning_rate": 0.00027872340425531914,
+      "loss": 4.2635,
+      "step": 131
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0935157537460327,
+      "learning_rate": 0.00028085106382978725,
+      "loss": 4.1894,
+      "step": 132
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.286113977432251,
+      "learning_rate": 0.00028297872340425536,
+      "loss": 4.3281,
+      "step": 133
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9515753388404846,
+      "learning_rate": 0.0002851063829787234,
+      "loss": 4.2958,
+      "step": 134
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2348365783691406,
+      "learning_rate": 0.00028723404255319153,
+      "loss": 4.0989,
+      "step": 135
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.276850700378418,
+      "learning_rate": 0.00028936170212765953,
+      "loss": 4.1455,
+      "step": 136
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.26995050907135,
+      "learning_rate": 0.00029148936170212764,
+      "loss": 4.1291,
+      "step": 137
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.523987889289856,
+      "learning_rate": 0.00029361702127659575,
+      "loss": 4.0667,
+      "step": 138
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3023549318313599,
+      "learning_rate": 0.0002957446808510638,
+      "loss": 4.1695,
+      "step": 139
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.365368127822876,
+      "learning_rate": 0.0002978723404255319,
+      "loss": 4.1791,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2599130868911743,
+      "learning_rate": 0.0003,
+      "loss": 4.2204,
+      "step": 141
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4464629888534546,
+      "learning_rate": 0.0003021276595744681,
+      "loss": 4.2318,
+      "step": 142
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1470637321472168,
+      "learning_rate": 0.0003042553191489362,
+      "loss": 4.1789,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.162804365158081,
+      "learning_rate": 0.00030638297872340425,
+      "loss": 4.1139,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.187334418296814,
+      "learning_rate": 0.00030851063829787236,
+      "loss": 3.9921,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4393917322158813,
+      "learning_rate": 0.0003106382978723404,
+      "loss": 4.1667,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2334777116775513,
+      "learning_rate": 0.0003127659574468085,
+      "loss": 3.9603,
+      "step": 147
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4265984296798706,
+      "learning_rate": 0.00031489361702127664,
+      "loss": 4.1494,
+      "step": 148
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3457194566726685,
+      "learning_rate": 0.0003170212765957447,
+      "loss": 4.035,
+      "step": 149
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.254904866218567,
+      "learning_rate": 0.0003191489361702128,
+      "loss": 4.1278,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3408393859863281,
+      "learning_rate": 0.00032127659574468086,
+      "loss": 4.0868,
+      "step": 151
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.306312918663025,
+      "learning_rate": 0.00032340425531914897,
+      "loss": 4.0424,
+      "step": 152
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2912030220031738,
+      "learning_rate": 0.0003255319148936171,
+      "loss": 3.9918,
+      "step": 153
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4844059944152832,
+      "learning_rate": 0.0003276595744680851,
+      "loss": 4.0833,
+      "step": 154
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3825430870056152,
+      "learning_rate": 0.0003297872340425532,
+      "loss": 4.0956,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3500663042068481,
+      "learning_rate": 0.00033191489361702125,
+      "loss": 4.1236,
+      "step": 156
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.284753441810608,
+      "learning_rate": 0.00033404255319148936,
+      "loss": 4.0041,
+      "step": 157
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.225710391998291,
+      "learning_rate": 0.0003361702127659574,
+      "loss": 4.0542,
+      "step": 158
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3104676008224487,
+      "learning_rate": 0.0003382978723404255,
+      "loss": 4.1674,
+      "step": 159
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2814124822616577,
+      "learning_rate": 0.00034042553191489364,
+      "loss": 4.1104,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.258671760559082,
+      "learning_rate": 0.0003425531914893617,
+      "loss": 3.9289,
+      "step": 161
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3215150833129883,
+      "learning_rate": 0.0003446808510638298,
+      "loss": 4.0127,
+      "step": 162
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.372815489768982,
+      "learning_rate": 0.00034680851063829786,
+      "loss": 4.1305,
+      "step": 163
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.322662115097046,
+      "learning_rate": 0.00034893617021276597,
+      "loss": 3.956,
+      "step": 164
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5933729410171509,
+      "learning_rate": 0.0003510638297872341,
+      "loss": 4.0883,
+      "step": 165
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2900516986846924,
+      "learning_rate": 0.00035319148936170213,
+      "loss": 4.0051,
+      "step": 166
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4232264757156372,
+      "learning_rate": 0.00035531914893617025,
+      "loss": 4.0694,
+      "step": 167
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.377969741821289,
+      "learning_rate": 0.0003574468085106383,
+      "loss": 4.0636,
+      "step": 168
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3479586839675903,
+      "learning_rate": 0.0003595744680851064,
+      "loss": 4.1845,
+      "step": 169
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3103374242782593,
+      "learning_rate": 0.0003617021276595745,
+      "loss": 3.9658,
+      "step": 170
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3195881843566895,
+      "learning_rate": 0.0003638297872340426,
+      "loss": 4.0497,
+      "step": 171
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.282640814781189,
+      "learning_rate": 0.00036595744680851063,
+      "loss": 4.024,
+      "step": 172
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4359363317489624,
+      "learning_rate": 0.0003680851063829787,
+      "loss": 3.9915,
+      "step": 173
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3655848503112793,
+      "learning_rate": 0.0003702127659574468,
+      "loss": 4.0806,
+      "step": 174
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3621926307678223,
+      "learning_rate": 0.0003723404255319149,
+      "loss": 4.1544,
+      "step": 175
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3997646570205688,
+      "learning_rate": 0.00037446808510638297,
+      "loss": 3.8485,
+      "step": 176
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.539406180381775,
+      "learning_rate": 0.0003765957446808511,
+      "loss": 4.1827,
+      "step": 177
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5509058237075806,
+      "learning_rate": 0.00037872340425531913,
+      "loss": 3.9315,
+      "step": 178
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.381636142730713,
+      "learning_rate": 0.00038085106382978724,
+      "loss": 3.9717,
+      "step": 179
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3316289186477661,
+      "learning_rate": 0.0003829787234042553,
+      "loss": 3.9606,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3810575008392334,
+      "learning_rate": 0.0003851063829787234,
+      "loss": 3.9959,
+      "step": 181
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6070252656936646,
+      "learning_rate": 0.0003872340425531915,
+      "loss": 4.0106,
+      "step": 182
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1589950323104858,
+      "learning_rate": 0.0003893617021276596,
+      "loss": 3.9195,
+      "step": 183
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1375452280044556,
+      "learning_rate": 0.0003914893617021277,
+      "loss": 3.9347,
+      "step": 184
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.451849102973938,
+      "learning_rate": 0.00039361702127659574,
+      "loss": 4.1053,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.171995997428894,
+      "learning_rate": 0.00039574468085106385,
+      "loss": 4.0659,
+      "step": 186
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3602004051208496,
+      "learning_rate": 0.00039787234042553196,
+      "loss": 3.9373,
+      "step": 187
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4252485036849976,
+      "learning_rate": 0.0004,
+      "loss": 3.9485,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3669577836990356,
+      "learning_rate": 0.00040212765957446813,
+      "loss": 3.9626,
+      "step": 189
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4510173797607422,
+      "learning_rate": 0.00040425531914893613,
+      "loss": 4.0763,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6044769287109375,
+      "learning_rate": 0.00040638297872340424,
+      "loss": 3.9054,
+      "step": 191
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3575087785720825,
+      "learning_rate": 0.00040851063829787235,
+      "loss": 3.9598,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4169284105300903,
+      "learning_rate": 0.0004106382978723404,
+      "loss": 4.0932,
+      "step": 193
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3452703952789307,
+      "learning_rate": 0.0004127659574468085,
+      "loss": 3.9784,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2284389734268188,
+      "learning_rate": 0.0004148936170212766,
+      "loss": 4.149,
+      "step": 195
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5200728178024292,
+      "learning_rate": 0.0004170212765957447,
+      "loss": 3.9903,
+      "step": 196
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7464290857315063,
+      "learning_rate": 0.0004191489361702128,
+      "loss": 3.9171,
+      "step": 197
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.219582438468933,
+      "learning_rate": 0.00042127659574468085,
+      "loss": 3.9455,
+      "step": 198
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3889364004135132,
+      "learning_rate": 0.00042340425531914896,
+      "loss": 4.0542,
+      "step": 199
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5458612442016602,
+      "learning_rate": 0.000425531914893617,
+      "loss": 3.9974,
+      "step": 200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3700460195541382,
+      "learning_rate": 0.00042765957446808513,
+      "loss": 3.919,
+      "step": 201
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6473002433776855,
+      "learning_rate": 0.0004297872340425532,
+      "loss": 3.9714,
+      "step": 202
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1797369718551636,
+      "learning_rate": 0.0004319148936170213,
+      "loss": 3.9766,
+      "step": 203
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1952027082443237,
+      "learning_rate": 0.0004340425531914894,
+      "loss": 3.9486,
+      "step": 204
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4129725694656372,
+      "learning_rate": 0.00043617021276595746,
+      "loss": 3.9886,
+      "step": 205
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2671421766281128,
+      "learning_rate": 0.00043829787234042557,
+      "loss": 3.7758,
+      "step": 206
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.422463059425354,
+      "learning_rate": 0.0004404255319148936,
+      "loss": 3.9077,
+      "step": 207
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4845503568649292,
+      "learning_rate": 0.0004425531914893617,
+      "loss": 3.8417,
+      "step": 208
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.202187418937683,
+      "learning_rate": 0.0004446808510638298,
+      "loss": 3.9389,
+      "step": 209
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.444098711013794,
+      "learning_rate": 0.00044680851063829785,
+      "loss": 3.951,
+      "step": 210
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4373562335968018,
+      "learning_rate": 0.00044893617021276596,
+      "loss": 3.9285,
+      "step": 211
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4413760900497437,
+      "learning_rate": 0.000451063829787234,
+      "loss": 3.9084,
+      "step": 212
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.457393765449524,
+      "learning_rate": 0.0004531914893617021,
+      "loss": 3.8667,
+      "step": 213
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4463461637496948,
+      "learning_rate": 0.00045531914893617024,
+      "loss": 3.8545,
+      "step": 214
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5317926406860352,
+      "learning_rate": 0.0004574468085106383,
+      "loss": 3.8489,
+      "step": 215
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3613853454589844,
+      "learning_rate": 0.0004595744680851064,
+      "loss": 3.8174,
+      "step": 216
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4517394304275513,
+      "learning_rate": 0.00046170212765957446,
+      "loss": 3.8263,
+      "step": 217
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1728559732437134,
+      "learning_rate": 0.00046382978723404257,
+      "loss": 3.8656,
+      "step": 218
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.242660641670227,
+      "learning_rate": 0.0004659574468085107,
+      "loss": 3.7996,
+      "step": 219
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4940431118011475,
+      "learning_rate": 0.00046808510638297874,
+      "loss": 4.0835,
+      "step": 220
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3037731647491455,
+      "learning_rate": 0.00047021276595744685,
+      "loss": 4.0274,
+      "step": 221
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3616893291473389,
+      "learning_rate": 0.0004723404255319149,
+      "loss": 4.0012,
+      "step": 222
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.081254482269287,
+      "learning_rate": 0.000474468085106383,
+      "loss": 3.7804,
+      "step": 223
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.778045654296875,
+      "learning_rate": 0.0004765957446808511,
+      "loss": 3.9001,
+      "step": 224
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1809135675430298,
+      "learning_rate": 0.0004787234042553192,
+      "loss": 3.8351,
+      "step": 225
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2070225477218628,
+      "learning_rate": 0.00048085106382978723,
+      "loss": 3.9147,
+      "step": 226
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4028456211090088,
+      "learning_rate": 0.0004829787234042553,
+      "loss": 3.8441,
+      "step": 227
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1965014934539795,
+      "learning_rate": 0.0004851063829787234,
+      "loss": 3.9585,
+      "step": 228
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2288951873779297,
+      "learning_rate": 0.00048723404255319146,
+      "loss": 3.9131,
+      "step": 229
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5023062229156494,
+      "learning_rate": 0.0004893617021276596,
+      "loss": 3.6967,
+      "step": 230
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2283579111099243,
+      "learning_rate": 0.0004914893617021277,
+      "loss": 3.949,
+      "step": 231
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3287562131881714,
+      "learning_rate": 0.0004936170212765957,
+      "loss": 3.9456,
+      "step": 232
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2116198539733887,
+      "learning_rate": 0.0004957446808510638,
+      "loss": 3.9013,
+      "step": 233
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.242817759513855,
+      "learning_rate": 0.000497872340425532,
+      "loss": 3.7293,
+      "step": 234
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2188082933425903,
+      "learning_rate": 0.0005,
+      "loss": 3.889,
+      "step": 235
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3149223327636719,
+      "learning_rate": 0.0004999999785110509,
+      "loss": 3.8368,
+      "step": 236
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1687288284301758,
+      "learning_rate": 0.0004999999140442072,
+      "loss": 3.7987,
+      "step": 237
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2290496826171875,
+      "learning_rate": 0.0004999998065994801,
+      "loss": 3.933,
+      "step": 238
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2678278684616089,
+      "learning_rate": 0.0004999996561768879,
+      "loss": 4.0096,
+      "step": 239
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1861952543258667,
+      "learning_rate": 0.0004999994627764566,
+      "loss": 3.8446,
+      "step": 240
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.275098443031311,
+      "learning_rate": 0.0004999992263982194,
+      "loss": 3.8766,
+      "step": 241
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.190368413925171,
+      "learning_rate": 0.000499998947042217,
+      "loss": 3.7794,
+      "step": 242
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6217900514602661,
+      "learning_rate": 0.0004999986247084974,
+      "loss": 3.8662,
+      "step": 243
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.160539150238037,
+      "learning_rate": 0.0004999982593971157,
+      "loss": 3.9881,
+      "step": 244
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2364122867584229,
+      "learning_rate": 0.0004999978511081353,
+      "loss": 3.8424,
+      "step": 245
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2432687282562256,
+      "learning_rate": 0.0004999973998416259,
+      "loss": 3.8258,
+      "step": 246
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4316729307174683,
+      "learning_rate": 0.0004999969055976653,
+      "loss": 3.8667,
+      "step": 247
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1562659740447998,
+      "learning_rate": 0.0004999963683763384,
+      "loss": 3.7427,
+      "step": 248
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1819325685501099,
+      "learning_rate": 0.0004999957881777376,
+      "loss": 3.7863,
+      "step": 249
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3457486629486084,
+      "learning_rate": 0.0004999951650019627,
+      "loss": 4.0465,
+      "step": 250
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.232448697090149,
+      "learning_rate": 0.0004999944988491207,
+      "loss": 3.7483,
+      "step": 251
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.208624243736267,
+      "learning_rate": 0.000499993789719326,
+      "loss": 3.9846,
+      "step": 252
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.267774224281311,
+      "learning_rate": 0.0004999930376127007,
+      "loss": 3.9279,
+      "step": 253
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2672343254089355,
+      "learning_rate": 0.0004999922425293743,
+      "loss": 3.7615,
+      "step": 254
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0653913021087646,
+      "learning_rate": 0.000499991404469483,
+      "loss": 3.9252,
+      "step": 255
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3523491621017456,
+      "learning_rate": 0.0004999905234331712,
+      "loss": 3.8587,
+      "step": 256
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3088959455490112,
+      "learning_rate": 0.0004999895994205903,
+      "loss": 3.9086,
+      "step": 257
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2633506059646606,
+      "learning_rate": 0.0004999886324318992,
+      "loss": 3.799,
+      "step": 258
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1152255535125732,
+      "learning_rate": 0.000499987622467264,
+      "loss": 3.8537,
+      "step": 259
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1998374462127686,
+      "learning_rate": 0.0004999865695268584,
+      "loss": 3.8254,
+      "step": 260
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.329124927520752,
+      "learning_rate": 0.0004999854736108633,
+      "loss": 3.8573,
+      "step": 261
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2538161277770996,
+      "learning_rate": 0.0004999843347194674,
+      "loss": 3.9046,
+      "step": 262
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1962872743606567,
+      "learning_rate": 0.0004999831528528662,
+      "loss": 3.8586,
+      "step": 263
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3084690570831299,
+      "learning_rate": 0.0004999819280112629,
+      "loss": 3.896,
+      "step": 264
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0731728076934814,
+      "learning_rate": 0.0004999806601948682,
+      "loss": 3.8193,
+      "step": 265
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3075398206710815,
+      "learning_rate": 0.0004999793494039,
+      "loss": 3.9347,
+      "step": 266
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2276506423950195,
+      "learning_rate": 0.0004999779956385836,
+      "loss": 3.8585,
+      "step": 267
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9807101488113403,
+      "learning_rate": 0.0004999765988991518,
+      "loss": 3.7606,
+      "step": 268
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0475084781646729,
+      "learning_rate": 0.0004999751591858447,
+      "loss": 3.7161,
+      "step": 269
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1833949089050293,
+      "learning_rate": 0.0004999736764989096,
+      "loss": 3.7091,
+      "step": 270
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1262840032577515,
+      "learning_rate": 0.0004999721508386018,
+      "loss": 3.8663,
+      "step": 271
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1360872983932495,
+      "learning_rate": 0.0004999705822051832,
+      "loss": 3.8956,
+      "step": 272
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.2529104948043823,
+      "learning_rate": 0.0004999689705989237,
+      "loss": 3.8955,
+      "step": 273
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1704559326171875,
+      "learning_rate": 0.0004999673160201001,
+      "loss": 3.683,
+      "step": 274
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3770536184310913,
+      "learning_rate": 0.0004999656184689972,
+      "loss": 3.8874,
+      "step": 275
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2912108898162842,
+      "learning_rate": 0.0004999638779459065,
+      "loss": 3.8122,
+      "step": 276
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2223007678985596,
+      "learning_rate": 0.0004999620944511274,
+      "loss": 3.8967,
+      "step": 277
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.205087661743164,
+      "learning_rate": 0.0004999602679849665,
+      "loss": 3.7982,
+      "step": 278
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1759825944900513,
+      "learning_rate": 0.0004999583985477377,
+      "loss": 3.7433,
+      "step": 279
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3122222423553467,
+      "learning_rate": 0.0004999564861397624,
+      "loss": 3.7569,
+      "step": 280
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.177197813987732,
+      "learning_rate": 0.0004999545307613695,
+      "loss": 3.9638,
+      "step": 281
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0554687976837158,
+      "learning_rate": 0.0004999525324128949,
+      "loss": 3.7023,
+      "step": 282
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1935759782791138,
+      "learning_rate": 0.0004999504910946824,
+      "loss": 3.8672,
+      "step": 283
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2850805521011353,
+      "learning_rate": 0.0004999484068070827,
+      "loss": 3.7865,
+      "step": 284
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1198339462280273,
+      "learning_rate": 0.0004999462795504542,
+      "loss": 3.829,
+      "step": 285
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.138165831565857,
+      "learning_rate": 0.0004999441093251627,
+      "loss": 3.854,
+      "step": 286
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0754280090332031,
+      "learning_rate": 0.0004999418961315812,
+      "loss": 3.832,
+      "step": 287
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0576859712600708,
+      "learning_rate": 0.0004999396399700902,
+      "loss": 3.8533,
+      "step": 288
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2031567096710205,
+      "learning_rate": 0.0004999373408410775,
+      "loss": 3.9285,
+      "step": 289
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1341794729232788,
+      "learning_rate": 0.0004999349987449384,
+      "loss": 3.9406,
+      "step": 290
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0620028972625732,
+      "learning_rate": 0.0004999326136820754,
+      "loss": 3.8794,
+      "step": 291
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2464852333068848,
+      "learning_rate": 0.0004999301856528989,
+      "loss": 3.9019,
+      "step": 292
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0631821155548096,
+      "learning_rate": 0.0004999277146578258,
+      "loss": 3.6962,
+      "step": 293
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0158509016036987,
+      "learning_rate": 0.0004999252006972813,
+      "loss": 3.7356,
+      "step": 294
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.038588047027588,
+      "learning_rate": 0.0004999226437716974,
+      "loss": 3.7918,
+      "step": 295
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1335421800613403,
+      "learning_rate": 0.0004999200438815136,
+      "loss": 3.7246,
+      "step": 296
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.126384973526001,
+      "learning_rate": 0.000499917401027177,
+      "loss": 3.7059,
+      "step": 297
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0297484397888184,
+      "learning_rate": 0.0004999147152091419,
+      "loss": 3.8476,
+      "step": 298
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1043933629989624,
+      "learning_rate": 0.0004999119864278699,
+      "loss": 3.7188,
+      "step": 299
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0596013069152832,
+      "learning_rate": 0.0004999092146838302,
+      "loss": 3.7761,
+      "step": 300
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2465026378631592,
+      "learning_rate": 0.0004999063999774994,
+      "loss": 3.8733,
+      "step": 301
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1994378566741943,
+      "learning_rate": 0.0004999035423093612,
+      "loss": 3.7724,
+      "step": 302
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.140863060951233,
+      "learning_rate": 0.000499900641679907,
+      "loss": 3.8162,
+      "step": 303
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0662174224853516,
+      "learning_rate": 0.0004998976980896354,
+      "loss": 3.831,
+      "step": 304
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.201188325881958,
+      "learning_rate": 0.0004998947115390524,
+      "loss": 3.9026,
+      "step": 305
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1321618556976318,
+      "learning_rate": 0.0004998916820286714,
+      "loss": 3.8513,
+      "step": 306
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9739273190498352,
+      "learning_rate": 0.0004998886095590134,
+      "loss": 3.9243,
+      "step": 307
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.112061858177185,
+      "learning_rate": 0.0004998854941306064,
+      "loss": 3.8003,
+      "step": 308
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1328226327896118,
+      "learning_rate": 0.000499882335743986,
+      "loss": 3.7843,
+      "step": 309
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1005468368530273,
+      "learning_rate": 0.0004998791343996952,
+      "loss": 3.7154,
+      "step": 310
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.102019190788269,
+      "learning_rate": 0.0004998758900982845,
+      "loss": 3.7542,
+      "step": 311
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2541944980621338,
+      "learning_rate": 0.0004998726028403114,
+      "loss": 3.7998,
+      "step": 312
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0166407823562622,
+      "learning_rate": 0.000499869272626341,
+      "loss": 3.7849,
+      "step": 313
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1922158002853394,
+      "learning_rate": 0.0004998658994569459,
+      "loss": 3.861,
+      "step": 314
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2430380582809448,
+      "learning_rate": 0.0004998624833327061,
+      "loss": 3.8154,
+      "step": 315
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2010189294815063,
+      "learning_rate": 0.0004998590242542087,
+      "loss": 3.8315,
+      "step": 316
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1250263452529907,
+      "learning_rate": 0.0004998555222220485,
+      "loss": 3.8601,
+      "step": 317
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1066341400146484,
+      "learning_rate": 0.0004998519772368273,
+      "loss": 3.8251,
+      "step": 318
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1382642984390259,
+      "learning_rate": 0.0004998483892991549,
+      "loss": 3.6828,
+      "step": 319
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.06110680103302,
+      "learning_rate": 0.0004998447584096477,
+      "loss": 3.8914,
+      "step": 320
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.044512391090393,
+      "learning_rate": 0.0004998410845689301,
+      "loss": 3.7508,
+      "step": 321
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.056077003479004,
+      "learning_rate": 0.0004998373677776337,
+      "loss": 3.7714,
+      "step": 322
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1969995498657227,
+      "learning_rate": 0.0004998336080363975,
+      "loss": 3.7438,
+      "step": 323
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0868432521820068,
+      "learning_rate": 0.0004998298053458676,
+      "loss": 3.7956,
+      "step": 324
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9396175146102905,
+      "learning_rate": 0.000499825959706698,
+      "loss": 3.6727,
+      "step": 325
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0107090473175049,
+      "learning_rate": 0.0004998220711195496,
+      "loss": 3.8141,
+      "step": 326
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1681958436965942,
+      "learning_rate": 0.0004998181395850911,
+      "loss": 3.6731,
+      "step": 327
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.969027042388916,
+      "learning_rate": 0.0004998141651039982,
+      "loss": 3.7648,
+      "step": 328
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.016277551651001,
+      "learning_rate": 0.0004998101476769542,
+      "loss": 3.8649,
+      "step": 329
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2155287265777588,
+      "learning_rate": 0.0004998060873046498,
+      "loss": 3.911,
+      "step": 330
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.138216495513916,
+      "learning_rate": 0.000499801983987783,
+      "loss": 3.7283,
+      "step": 331
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9919595718383789,
+      "learning_rate": 0.0004997978377270591,
+      "loss": 3.7001,
+      "step": 332
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0648598670959473,
+      "learning_rate": 0.0004997936485231911,
+      "loss": 3.8649,
+      "step": 333
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.069198727607727,
+      "learning_rate": 0.0004997894163768992,
+      "loss": 3.7799,
+      "step": 334
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1835734844207764,
+      "learning_rate": 0.0004997851412889106,
+      "loss": 3.7385,
+      "step": 335
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1583417654037476,
+      "learning_rate": 0.0004997808232599604,
+      "loss": 3.6885,
+      "step": 336
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1449414491653442,
+      "learning_rate": 0.0004997764622907911,
+      "loss": 3.6956,
+      "step": 337
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.168426513671875,
+      "learning_rate": 0.0004997720583821523,
+      "loss": 3.7565,
+      "step": 338
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0819423198699951,
+      "learning_rate": 0.000499767611534801,
+      "loss": 3.7627,
+      "step": 339
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2998234033584595,
+      "learning_rate": 0.0004997631217495018,
+      "loss": 3.7833,
+      "step": 340
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9576529860496521,
+      "learning_rate": 0.0004997585890270265,
+      "loss": 3.8186,
+      "step": 341
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0528411865234375,
+      "learning_rate": 0.0004997540133681541,
+      "loss": 3.706,
+      "step": 342
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0465948581695557,
+      "learning_rate": 0.0004997493947736715,
+      "loss": 3.7249,
+      "step": 343
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9764485955238342,
+      "learning_rate": 0.0004997447332443727,
+      "loss": 3.6947,
+      "step": 344
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9489080309867859,
+      "learning_rate": 0.0004997400287810587,
+      "loss": 3.7476,
+      "step": 345
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.995637059211731,
+      "learning_rate": 0.0004997352813845388,
+      "loss": 3.7786,
+      "step": 346
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9482467174530029,
+      "learning_rate": 0.0004997304910556288,
+      "loss": 3.74,
+      "step": 347
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1336299180984497,
+      "learning_rate": 0.0004997256577951521,
+      "loss": 3.7968,
+      "step": 348
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0206025838851929,
+      "learning_rate": 0.0004997207816039398,
+      "loss": 3.6832,
+      "step": 349
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0714061260223389,
+      "learning_rate": 0.0004997158624828303,
+      "loss": 3.6713,
+      "step": 350
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0384105443954468,
+      "learning_rate": 0.000499710900432669,
+      "loss": 3.7539,
+      "step": 351
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.263920783996582,
+      "learning_rate": 0.0004997058954543089,
+      "loss": 3.847,
+      "step": 352
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9722920656204224,
+      "learning_rate": 0.0004997008475486107,
+      "loss": 3.6932,
+      "step": 353
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9932896494865417,
+      "learning_rate": 0.000499695756716442,
+      "loss": 3.7424,
+      "step": 354
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5072218179702759,
+      "learning_rate": 0.0004996906229586778,
+      "loss": 3.7184,
+      "step": 355
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0924286842346191,
+      "learning_rate": 0.000499685446276201,
+      "loss": 3.741,
+      "step": 356
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9837846755981445,
+      "learning_rate": 0.0004996802266699014,
+      "loss": 3.6977,
+      "step": 357
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0560784339904785,
+      "learning_rate": 0.0004996749641406763,
+      "loss": 3.7476,
+      "step": 358
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.043904423713684,
+      "learning_rate": 0.0004996696586894304,
+      "loss": 3.8986,
+      "step": 359
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0850636959075928,
+      "learning_rate": 0.0004996643103170757,
+      "loss": 3.7739,
+      "step": 360
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0535517930984497,
+      "learning_rate": 0.0004996589190245318,
+      "loss": 3.9393,
+      "step": 361
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8589828610420227,
+      "learning_rate": 0.0004996534848127253,
+      "loss": 3.651,
+      "step": 362
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2381101846694946,
+      "learning_rate": 0.0004996480076825906,
+      "loss": 3.8707,
+      "step": 363
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9625450372695923,
+      "learning_rate": 0.0004996424876350692,
+      "loss": 3.7679,
+      "step": 364
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9322208762168884,
+      "learning_rate": 0.00049963692467111,
+      "loss": 3.6827,
+      "step": 365
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0279086828231812,
+      "learning_rate": 0.0004996313187916694,
+      "loss": 3.7496,
+      "step": 366
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1358555555343628,
+      "learning_rate": 0.0004996256699977112,
+      "loss": 3.7333,
+      "step": 367
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1529221534729004,
+      "learning_rate": 0.0004996199782902064,
+      "loss": 3.6571,
+      "step": 368
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.101511001586914,
+      "learning_rate": 0.0004996142436701336,
+      "loss": 3.692,
+      "step": 369
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1611484289169312,
+      "learning_rate": 0.0004996084661384783,
+      "loss": 3.8783,
+      "step": 370
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.027530312538147,
+      "learning_rate": 0.0004996026456962341,
+      "loss": 3.7984,
+      "step": 371
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1063411235809326,
+      "learning_rate": 0.0004995967823444015,
+      "loss": 3.6956,
+      "step": 372
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.212572693824768,
+      "learning_rate": 0.0004995908760839884,
+      "loss": 3.8137,
+      "step": 373
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.067589282989502,
+      "learning_rate": 0.0004995849269160102,
+      "loss": 3.8697,
+      "step": 374
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.170344352722168,
+      "learning_rate": 0.0004995789348414897,
+      "loss": 3.7037,
+      "step": 375
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0547147989273071,
+      "learning_rate": 0.0004995728998614567,
+      "loss": 3.8312,
+      "step": 376
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9910465478897095,
+      "learning_rate": 0.0004995668219769492,
+      "loss": 3.7954,
+      "step": 377
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.950949490070343,
+      "learning_rate": 0.0004995607011890115,
+      "loss": 3.6607,
+      "step": 378
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9628477692604065,
+      "learning_rate": 0.0004995545374986963,
+      "loss": 3.7922,
+      "step": 379
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.106607437133789,
+      "learning_rate": 0.000499548330907063,
+      "loss": 3.8673,
+      "step": 380
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1569526195526123,
+      "learning_rate": 0.0004995420814151786,
+      "loss": 3.7485,
+      "step": 381
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9462524056434631,
+      "learning_rate": 0.0004995357890241174,
+      "loss": 3.6742,
+      "step": 382
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.117493748664856,
+      "learning_rate": 0.0004995294537349612,
+      "loss": 3.7589,
+      "step": 383
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0731674432754517,
+      "learning_rate": 0.0004995230755487992,
+      "loss": 3.7896,
+      "step": 384
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9641139507293701,
+      "learning_rate": 0.0004995166544667278,
+      "loss": 3.6979,
+      "step": 385
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.982068657875061,
+      "learning_rate": 0.0004995101904898507,
+      "loss": 3.6966,
+      "step": 386
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9323788285255432,
+      "learning_rate": 0.0004995036836192793,
+      "loss": 3.7303,
+      "step": 387
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.083184838294983,
+      "learning_rate": 0.0004994971338561323,
+      "loss": 3.721,
+      "step": 388
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.071229100227356,
+      "learning_rate": 0.0004994905412015356,
+      "loss": 3.7621,
+      "step": 389
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0086432695388794,
+      "learning_rate": 0.0004994839056566225,
+      "loss": 3.8187,
+      "step": 390
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9265891313552856,
+      "learning_rate": 0.0004994772272225337,
+      "loss": 3.7917,
+      "step": 391
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0019060373306274,
+      "learning_rate": 0.0004994705059004174,
+      "loss": 3.648,
+      "step": 392
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1281073093414307,
+      "learning_rate": 0.000499463741691429,
+      "loss": 3.7661,
+      "step": 393
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1330609321594238,
+      "learning_rate": 0.0004994569345967314,
+      "loss": 3.6828,
+      "step": 394
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4806430339813232,
+      "learning_rate": 0.0004994500846174947,
+      "loss": 3.8634,
+      "step": 395
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1160086393356323,
+      "learning_rate": 0.0004994431917548966,
+      "loss": 3.6678,
+      "step": 396
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9635677933692932,
+      "learning_rate": 0.0004994362560101221,
+      "loss": 3.7078,
+      "step": 397
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0511142015457153,
+      "learning_rate": 0.0004994292773843635,
+      "loss": 3.7122,
+      "step": 398
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0360469818115234,
+      "learning_rate": 0.0004994222558788204,
+      "loss": 3.8101,
+      "step": 399
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9969998002052307,
+      "learning_rate": 0.0004994151914947001,
+      "loss": 3.6297,
+      "step": 400
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1051974296569824,
+      "learning_rate": 0.0004994080842332168,
+      "loss": 3.7338,
+      "step": 401
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9800593256950378,
+      "learning_rate": 0.0004994009340955924,
+      "loss": 3.8023,
+      "step": 402
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9644533395767212,
+      "learning_rate": 0.0004993937410830561,
+      "loss": 3.7269,
+      "step": 403
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1003131866455078,
+      "learning_rate": 0.0004993865051968447,
+      "loss": 3.7922,
+      "step": 404
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1602671146392822,
+      "learning_rate": 0.0004993792264382017,
+      "loss": 3.656,
+      "step": 405
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0589845180511475,
+      "learning_rate": 0.0004993719048083788,
+      "loss": 3.6774,
+      "step": 406
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1800709962844849,
+      "learning_rate": 0.0004993645403086344,
+      "loss": 3.8222,
+      "step": 407
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1202436685562134,
+      "learning_rate": 0.0004993571329402346,
+      "loss": 3.8316,
+      "step": 408
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.017310380935669,
+      "learning_rate": 0.0004993496827044529,
+      "loss": 3.6912,
+      "step": 409
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1052725315093994,
+      "learning_rate": 0.00049934218960257,
+      "loss": 3.8175,
+      "step": 410
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2691749334335327,
+      "learning_rate": 0.0004993346536358742,
+      "loss": 3.7636,
+      "step": 411
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0025804042816162,
+      "learning_rate": 0.0004993270748056607,
+      "loss": 3.7851,
+      "step": 412
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0031421184539795,
+      "learning_rate": 0.0004993194531132328,
+      "loss": 3.7159,
+      "step": 413
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0373765230178833,
+      "learning_rate": 0.0004993117885599004,
+      "loss": 3.6204,
+      "step": 414
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9760125875473022,
+      "learning_rate": 0.0004993040811469815,
+      "loss": 3.7047,
+      "step": 415
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1238311529159546,
+      "learning_rate": 0.0004992963308758006,
+      "loss": 3.8231,
+      "step": 416
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0206835269927979,
+      "learning_rate": 0.0004992885377476904,
+      "loss": 3.8061,
+      "step": 417
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9614518284797668,
+      "learning_rate": 0.0004992807017639906,
+      "loss": 3.6469,
+      "step": 418
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1135566234588623,
+      "learning_rate": 0.0004992728229260484,
+      "loss": 3.7795,
+      "step": 419
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.140252709388733,
+      "learning_rate": 0.000499264901235218,
+      "loss": 3.7001,
+      "step": 420
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1376891136169434,
+      "learning_rate": 0.0004992569366928612,
+      "loss": 3.6246,
+      "step": 421
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0964652299880981,
+      "learning_rate": 0.0004992489293003475,
+      "loss": 3.6815,
+      "step": 422
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9947781562805176,
+      "learning_rate": 0.0004992408790590534,
+      "loss": 3.7321,
+      "step": 423
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.073929786682129,
+      "learning_rate": 0.0004992327859703626,
+      "loss": 3.7496,
+      "step": 424
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0482063293457031,
+      "learning_rate": 0.0004992246500356665,
+      "loss": 3.6617,
+      "step": 425
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0412204265594482,
+      "learning_rate": 0.0004992164712563639,
+      "loss": 3.7565,
+      "step": 426
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1041027307510376,
+      "learning_rate": 0.0004992082496338607,
+      "loss": 3.7869,
+      "step": 427
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9422649145126343,
+      "learning_rate": 0.0004991999851695704,
+      "loss": 3.6871,
+      "step": 428
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9400534629821777,
+      "learning_rate": 0.0004991916778649135,
+      "loss": 3.7735,
+      "step": 429
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8632922172546387,
+      "learning_rate": 0.0004991833277213183,
+      "loss": 3.7803,
+      "step": 430
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9322618246078491,
+      "learning_rate": 0.0004991749347402205,
+      "loss": 3.7936,
+      "step": 431
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.067453384399414,
+      "learning_rate": 0.0004991664989230626,
+      "loss": 3.8727,
+      "step": 432
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9637219309806824,
+      "learning_rate": 0.0004991580202712949,
+      "loss": 3.6587,
+      "step": 433
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.019960880279541,
+      "learning_rate": 0.0004991494987863751,
+      "loss": 3.7486,
+      "step": 434
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9948220252990723,
+      "learning_rate": 0.0004991409344697681,
+      "loss": 3.7524,
+      "step": 435
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9766054749488831,
+      "learning_rate": 0.0004991323273229461,
+      "loss": 3.5528,
+      "step": 436
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9753291606903076,
+      "learning_rate": 0.0004991236773473889,
+      "loss": 3.7247,
+      "step": 437
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0976330041885376,
+      "learning_rate": 0.0004991149845445834,
+      "loss": 3.5883,
+      "step": 438
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1932387351989746,
+      "learning_rate": 0.0004991062489160241,
+      "loss": 3.754,
+      "step": 439
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9403764605522156,
+      "learning_rate": 0.0004990974704632128,
+      "loss": 3.6795,
+      "step": 440
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0581969022750854,
+      "learning_rate": 0.0004990886491876583,
+      "loss": 3.7334,
+      "step": 441
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9379636645317078,
+      "learning_rate": 0.0004990797850908775,
+      "loss": 3.7665,
+      "step": 442
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.97157883644104,
+      "learning_rate": 0.0004990708781743942,
+      "loss": 3.6782,
+      "step": 443
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9525388479232788,
+      "learning_rate": 0.0004990619284397393,
+      "loss": 3.8114,
+      "step": 444
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8939195275306702,
+      "learning_rate": 0.0004990529358884515,
+      "loss": 3.7224,
+      "step": 445
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0911587476730347,
+      "learning_rate": 0.0004990439005220768,
+      "loss": 3.7623,
+      "step": 446
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.012174367904663,
+      "learning_rate": 0.0004990348223421685,
+      "loss": 3.7234,
+      "step": 447
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0412369966506958,
+      "learning_rate": 0.0004990257013502871,
+      "loss": 3.8048,
+      "step": 448
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0075637102127075,
+      "learning_rate": 0.0004990165375480007,
+      "loss": 3.7575,
+      "step": 449
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.041245460510254,
+      "learning_rate": 0.0004990073309368847,
+      "loss": 3.6959,
+      "step": 450
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9612911343574524,
+      "learning_rate": 0.0004989980815185217,
+      "loss": 3.7159,
+      "step": 451
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.071912407875061,
+      "learning_rate": 0.0004989887892945018,
+      "loss": 3.6945,
+      "step": 452
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0854805707931519,
+      "learning_rate": 0.0004989794542664226,
+      "loss": 3.766,
+      "step": 453
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0065691471099854,
+      "learning_rate": 0.0004989700764358888,
+      "loss": 3.6882,
+      "step": 454
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0710570812225342,
+      "learning_rate": 0.0004989606558045126,
+      "loss": 3.6353,
+      "step": 455
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0000087022781372,
+      "learning_rate": 0.0004989511923739133,
+      "loss": 3.6365,
+      "step": 456
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9962633848190308,
+      "learning_rate": 0.0004989416861457181,
+      "loss": 3.7622,
+      "step": 457
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0738786458969116,
+      "learning_rate": 0.000498932137121561,
+      "loss": 3.7331,
+      "step": 458
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9874961972236633,
+      "learning_rate": 0.0004989225453030837,
+      "loss": 3.6447,
+      "step": 459
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9918415546417236,
+      "learning_rate": 0.000498912910691935,
+      "loss": 3.6844,
+      "step": 460
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9814335107803345,
+      "learning_rate": 0.0004989032332897714,
+      "loss": 3.7584,
+      "step": 461
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9892833828926086,
+      "learning_rate": 0.0004988935130982564,
+      "loss": 3.6989,
+      "step": 462
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.126273274421692,
+      "learning_rate": 0.000498883750119061,
+      "loss": 3.6964,
+      "step": 463
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0810917615890503,
+      "learning_rate": 0.0004988739443538638,
+      "loss": 3.6984,
+      "step": 464
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0964237451553345,
+      "learning_rate": 0.0004988640958043504,
+      "loss": 3.7274,
+      "step": 465
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0471069812774658,
+      "learning_rate": 0.0004988542044722138,
+      "loss": 3.735,
+      "step": 466
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0022653341293335,
+      "learning_rate": 0.0004988442703591545,
+      "loss": 3.7472,
+      "step": 467
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9772243499755859,
+      "learning_rate": 0.0004988342934668801,
+      "loss": 3.7419,
+      "step": 468
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0073323249816895,
+      "learning_rate": 0.0004988242737971061,
+      "loss": 3.7259,
+      "step": 469
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.921754777431488,
+      "learning_rate": 0.0004988142113515548,
+      "loss": 3.6942,
+      "step": 470
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8903483152389526,
+      "learning_rate": 0.000498804106131956,
+      "loss": 3.6989,
+      "step": 471
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.968635082244873,
+      "learning_rate": 0.000498793958140047,
+      "loss": 3.7787,
+      "step": 472
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9479289650917053,
+      "learning_rate": 0.0004987837673775723,
+      "loss": 3.6884,
+      "step": 473
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.1050664186477661,
+      "learning_rate": 0.000498773533846284,
+      "loss": 3.6962,
+      "step": 474
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9367997646331787,
+      "learning_rate": 0.000498763257547941,
+      "loss": 3.6899,
+      "step": 475
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.061821699142456,
+      "learning_rate": 0.0004987529384843102,
+      "loss": 3.5729,
+      "step": 476
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9343668818473816,
+      "learning_rate": 0.0004987425766571655,
+      "loss": 3.6983,
+      "step": 477
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.063883900642395,
+      "learning_rate": 0.0004987321720682882,
+      "loss": 3.7127,
+      "step": 478
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0057201385498047,
+      "learning_rate": 0.000498721724719467,
+      "loss": 3.8145,
+      "step": 479
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9163222908973694,
+      "learning_rate": 0.0004987112346124978,
+      "loss": 3.625,
+      "step": 480
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9235097169876099,
+      "learning_rate": 0.0004987007017491842,
+      "loss": 3.6672,
+      "step": 481
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2759689092636108,
+      "learning_rate": 0.0004986901261313366,
+      "loss": 3.8087,
+      "step": 482
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0272804498672485,
+      "learning_rate": 0.0004986795077607733,
+      "loss": 3.7867,
+      "step": 483
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9846420884132385,
+      "learning_rate": 0.0004986688466393198,
+      "loss": 3.6247,
+      "step": 484
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0052454471588135,
+      "learning_rate": 0.0004986581427688086,
+      "loss": 3.7265,
+      "step": 485
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9914617538452148,
+      "learning_rate": 0.0004986473961510801,
+      "loss": 3.6451,
+      "step": 486
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9145068526268005,
+      "learning_rate": 0.0004986366067879815,
+      "loss": 3.8237,
+      "step": 487
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.084983468055725,
+      "learning_rate": 0.0004986257746813678,
+      "loss": 3.6703,
+      "step": 488
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0903581380844116,
+      "learning_rate": 0.0004986148998331011,
+      "loss": 3.6616,
+      "step": 489
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.89190673828125,
+      "learning_rate": 0.0004986039822450509,
+      "loss": 3.7088,
+      "step": 490
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0566867589950562,
+      "learning_rate": 0.0004985930219190942,
+      "loss": 3.7702,
+      "step": 491
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9850021004676819,
+      "learning_rate": 0.0004985820188571151,
+      "loss": 3.6232,
+      "step": 492
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0485903024673462,
+      "learning_rate": 0.0004985709730610051,
+      "loss": 3.7174,
+      "step": 493
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9693613052368164,
+      "learning_rate": 0.0004985598845326631,
+      "loss": 3.6675,
+      "step": 494
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0052372217178345,
+      "learning_rate": 0.0004985487532739955,
+      "loss": 3.7537,
+      "step": 495
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.14641273021698,
+      "learning_rate": 0.0004985375792869157,
+      "loss": 3.6944,
+      "step": 496
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9626569747924805,
+      "learning_rate": 0.0004985263625733447,
+      "loss": 3.6664,
+      "step": 497
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0079176425933838,
+      "learning_rate": 0.0004985151031352108,
+      "loss": 3.7537,
+      "step": 498
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9485942721366882,
+      "learning_rate": 0.0004985038009744498,
+      "loss": 3.7139,
+      "step": 499
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0791044235229492,
+      "learning_rate": 0.0004984924560930044,
+      "loss": 3.7171,
+      "step": 500
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9445507526397705,
+      "learning_rate": 0.0004984810684928251,
+      "loss": 3.764,
+      "step": 501
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0104825496673584,
+      "learning_rate": 0.0004984696381758693,
+      "loss": 3.7245,
+      "step": 502
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9643205404281616,
+      "learning_rate": 0.0004984581651441024,
+      "loss": 3.7313,
+      "step": 503
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0203408002853394,
+      "learning_rate": 0.0004984466493994964,
+      "loss": 3.7902,
+      "step": 504
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9394329786300659,
+      "learning_rate": 0.0004984350909440311,
+      "loss": 3.5563,
+      "step": 505
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.996798038482666,
+      "learning_rate": 0.0004984234897796937,
+      "loss": 3.7845,
+      "step": 506
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9923459887504578,
+      "learning_rate": 0.0004984118459084783,
+      "loss": 3.7403,
+      "step": 507
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4790393114089966,
+      "learning_rate": 0.0004984001593323868,
+      "loss": 3.6578,
+      "step": 508
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0736733675003052,
+      "learning_rate": 0.0004983884300534283,
+      "loss": 3.7221,
+      "step": 509
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9900732636451721,
+      "learning_rate": 0.000498376658073619,
+      "loss": 3.5539,
+      "step": 510
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9970335364341736,
+      "learning_rate": 0.0004983648433949828,
+      "loss": 3.6776,
+      "step": 511
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0395703315734863,
+      "learning_rate": 0.0004983529860195507,
+      "loss": 3.712,
+      "step": 512
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.904090940952301,
+      "learning_rate": 0.0004983410859493611,
+      "loss": 3.6631,
+      "step": 513
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9814820289611816,
+      "learning_rate": 0.0004983291431864599,
+      "loss": 3.6864,
+      "step": 514
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0582082271575928,
+      "learning_rate": 0.0004983171577329001,
+      "loss": 3.6732,
+      "step": 515
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0297218561172485,
+      "learning_rate": 0.0004983051295907421,
+      "loss": 3.8229,
+      "step": 516
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9799910187721252,
+      "learning_rate": 0.0004982930587620537,
+      "loss": 3.6739,
+      "step": 517
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.02519690990448,
+      "learning_rate": 0.0004982809452489101,
+      "loss": 3.6983,
+      "step": 518
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9558513164520264,
+      "learning_rate": 0.0004982687890533936,
+      "loss": 3.5971,
+      "step": 519
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.03365957736969,
+      "learning_rate": 0.0004982565901775943,
+      "loss": 3.8095,
+      "step": 520
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0138533115386963,
+      "learning_rate": 0.0004982443486236089,
+      "loss": 3.7355,
+      "step": 521
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0666346549987793,
+      "learning_rate": 0.0004982320643935421,
+      "loss": 3.7923,
+      "step": 522
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0163183212280273,
+      "learning_rate": 0.0004982197374895058,
+      "loss": 3.7284,
+      "step": 523
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1301817893981934,
+      "learning_rate": 0.0004982073679136189,
+      "loss": 3.758,
+      "step": 524
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0546684265136719,
+      "learning_rate": 0.000498194955668008,
+      "loss": 3.7278,
+      "step": 525
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9236823916435242,
+      "learning_rate": 0.000498182500754807,
+      "loss": 3.6028,
+      "step": 526
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.940798282623291,
+      "learning_rate": 0.0004981700031761567,
+      "loss": 3.6304,
+      "step": 527
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0910943746566772,
+      "learning_rate": 0.0004981574629342061,
+      "loss": 3.7663,
+      "step": 528
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9817875027656555,
+      "learning_rate": 0.0004981448800311105,
+      "loss": 3.7661,
+      "step": 529
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9952785968780518,
+      "learning_rate": 0.0004981322544690335,
+      "loss": 3.7484,
+      "step": 530
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9223330020904541,
+      "learning_rate": 0.0004981195862501452,
+      "loss": 3.6418,
+      "step": 531
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0352973937988281,
+      "learning_rate": 0.0004981068753766237,
+      "loss": 3.7409,
+      "step": 532
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0839977264404297,
+      "learning_rate": 0.000498094121850654,
+      "loss": 3.729,
+      "step": 533
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0956603288650513,
+      "learning_rate": 0.0004980813256744286,
+      "loss": 3.6688,
+      "step": 534
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0349833965301514,
+      "learning_rate": 0.0004980684868501472,
+      "loss": 3.707,
+      "step": 535
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9875361323356628,
+      "learning_rate": 0.0004980556053800171,
+      "loss": 3.7233,
+      "step": 536
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0196729898452759,
+      "learning_rate": 0.0004980426812662527,
+      "loss": 3.5573,
+      "step": 537
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9404676556587219,
+      "learning_rate": 0.0004980297145110759,
+      "loss": 3.639,
+      "step": 538
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9518227577209473,
+      "learning_rate": 0.0004980167051167158,
+      "loss": 3.7485,
+      "step": 539
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0937633514404297,
+      "learning_rate": 0.0004980036530854088,
+      "loss": 3.679,
+      "step": 540
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9536413550376892,
+      "learning_rate": 0.0004979905584193986,
+      "loss": 3.7071,
+      "step": 541
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9441630840301514,
+      "learning_rate": 0.0004979774211209367,
+      "loss": 3.628,
+      "step": 542
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9995726943016052,
+      "learning_rate": 0.0004979642411922811,
+      "loss": 3.6734,
+      "step": 543
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9759696125984192,
+      "learning_rate": 0.0004979510186356979,
+      "loss": 3.7699,
+      "step": 544
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9783552289009094,
+      "learning_rate": 0.00049793775345346,
+      "loss": 3.7217,
+      "step": 545
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1201858520507812,
+      "learning_rate": 0.000497924445647848,
+      "loss": 3.7629,
+      "step": 546
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0142176151275635,
+      "learning_rate": 0.0004979110952211496,
+      "loss": 3.7308,
+      "step": 547
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9891877770423889,
+      "learning_rate": 0.00049789770217566,
+      "loss": 3.5363,
+      "step": 548
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.984727144241333,
+      "learning_rate": 0.0004978842665136814,
+      "loss": 3.7136,
+      "step": 549
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9455428123474121,
+      "learning_rate": 0.0004978707882375237,
+      "loss": 3.7227,
+      "step": 550
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9751055836677551,
+      "learning_rate": 0.000497857267349504,
+      "loss": 3.7307,
+      "step": 551
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9746074080467224,
+      "learning_rate": 0.0004978437038519465,
+      "loss": 3.6678,
+      "step": 552
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.972135066986084,
+      "learning_rate": 0.0004978300977471832,
+      "loss": 3.8389,
+      "step": 553
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.883399486541748,
+      "learning_rate": 0.0004978164490375529,
+      "loss": 3.7556,
+      "step": 554
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8993391990661621,
+      "learning_rate": 0.000497802757725402,
+      "loss": 3.6435,
+      "step": 555
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9784232378005981,
+      "learning_rate": 0.0004977890238130844,
+      "loss": 3.6716,
+      "step": 556
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0136913061141968,
+      "learning_rate": 0.000497775247302961,
+      "loss": 3.6346,
+      "step": 557
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1846816539764404,
+      "learning_rate": 0.0004977614281974002,
+      "loss": 3.7713,
+      "step": 558
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.601712703704834,
+      "learning_rate": 0.0004977475664987773,
+      "loss": 3.6996,
+      "step": 559
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9286705851554871,
+      "learning_rate": 0.0004977336622094759,
+      "loss": 3.6156,
+      "step": 560
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9827843308448792,
+      "learning_rate": 0.0004977197153318858,
+      "loss": 3.6352,
+      "step": 561
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.022339105606079,
+      "learning_rate": 0.000497705725868405,
+      "loss": 3.664,
+      "step": 562
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9422778487205505,
+      "learning_rate": 0.0004976916938214381,
+      "loss": 3.6822,
+      "step": 563
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8874156475067139,
+      "learning_rate": 0.0004976776191933976,
+      "loss": 3.651,
+      "step": 564
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.979041337966919,
+      "learning_rate": 0.000497663501986703,
+      "loss": 3.6669,
+      "step": 565
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0292575359344482,
+      "learning_rate": 0.0004976493422037813,
+      "loss": 3.727,
+      "step": 566
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9404982328414917,
+      "learning_rate": 0.0004976351398470667,
+      "loss": 3.7099,
+      "step": 567
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0229207277297974,
+      "learning_rate": 0.0004976208949190007,
+      "loss": 3.6508,
+      "step": 568
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.935306191444397,
+      "learning_rate": 0.0004976066074220322,
+      "loss": 3.8563,
+      "step": 569
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9937973618507385,
+      "learning_rate": 0.0004975922773586173,
+      "loss": 3.7509,
+      "step": 570
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9270636439323425,
+      "learning_rate": 0.0004975779047312198,
+      "loss": 3.6503,
+      "step": 571
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0206480026245117,
+      "learning_rate": 0.0004975634895423101,
+      "loss": 3.8005,
+      "step": 572
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0152143239974976,
+      "learning_rate": 0.0004975490317943666,
+      "loss": 3.7362,
+      "step": 573
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.946738600730896,
+      "learning_rate": 0.0004975345314898747,
+      "loss": 3.585,
+      "step": 574
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9271712303161621,
+      "learning_rate": 0.0004975199886313272,
+      "loss": 3.6641,
+      "step": 575
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9697306752204895,
+      "learning_rate": 0.0004975054032212241,
+      "loss": 3.7098,
+      "step": 576
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8851280808448792,
+      "learning_rate": 0.0004974907752620729,
+      "loss": 3.6736,
+      "step": 577
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9549417495727539,
+      "learning_rate": 0.0004974761047563883,
+      "loss": 3.536,
+      "step": 578
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9481258392333984,
+      "learning_rate": 0.0004974613917066923,
+      "loss": 3.6974,
+      "step": 579
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0045398473739624,
+      "learning_rate": 0.0004974466361155141,
+      "loss": 3.7196,
+      "step": 580
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9385789036750793,
+      "learning_rate": 0.0004974318379853907,
+      "loss": 3.724,
+      "step": 581
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0042544603347778,
+      "learning_rate": 0.0004974169973188658,
+      "loss": 3.6453,
+      "step": 582
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9435723423957825,
+      "learning_rate": 0.0004974021141184908,
+      "loss": 3.7277,
+      "step": 583
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9772533774375916,
+      "learning_rate": 0.0004973871883868242,
+      "loss": 3.6198,
+      "step": 584
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0190731287002563,
+      "learning_rate": 0.0004973722201264319,
+      "loss": 3.7273,
+      "step": 585
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0444413423538208,
+      "learning_rate": 0.0004973572093398873,
+      "loss": 3.7142,
+      "step": 586
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9854151606559753,
+      "learning_rate": 0.0004973421560297707,
+      "loss": 3.632,
+      "step": 587
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9795171618461609,
+      "learning_rate": 0.00049732706019867,
+      "loss": 3.5712,
+      "step": 588
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9471103549003601,
+      "learning_rate": 0.0004973119218491805,
+      "loss": 3.6515,
+      "step": 589
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0504335165023804,
+      "learning_rate": 0.0004972967409839045,
+      "loss": 3.7154,
+      "step": 590
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8781349062919617,
+      "learning_rate": 0.0004972815176054519,
+      "loss": 3.6802,
+      "step": 591
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8920720815658569,
+      "learning_rate": 0.0004972662517164396,
+      "loss": 3.5971,
+      "step": 592
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8624701499938965,
+      "learning_rate": 0.0004972509433194922,
+      "loss": 3.7251,
+      "step": 593
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8881915211677551,
+      "learning_rate": 0.0004972355924172411,
+      "loss": 3.6237,
+      "step": 594
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0436822175979614,
+      "learning_rate": 0.0004972201990123255,
+      "loss": 3.6556,
+      "step": 595
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9346268177032471,
+      "learning_rate": 0.0004972047631073917,
+      "loss": 3.7369,
+      "step": 596
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9199187159538269,
+      "learning_rate": 0.0004971892847050933,
+      "loss": 3.697,
+      "step": 597
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8995476961135864,
+      "learning_rate": 0.0004971737638080912,
+      "loss": 3.5557,
+      "step": 598
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8915233612060547,
+      "learning_rate": 0.0004971582004190536,
+      "loss": 3.6867,
+      "step": 599
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9910312294960022,
+      "learning_rate": 0.000497142594540656,
+      "loss": 3.7324,
+      "step": 600
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9255796074867249,
+      "learning_rate": 0.0004971269461755813,
+      "loss": 3.6542,
+      "step": 601
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9073544144630432,
+      "learning_rate": 0.0004971112553265195,
+      "loss": 3.6904,
+      "step": 602
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9548100233078003,
+      "learning_rate": 0.0004970955219961683,
+      "loss": 3.8423,
+      "step": 603
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9389156699180603,
+      "learning_rate": 0.0004970797461872323,
+      "loss": 3.8538,
+      "step": 604
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9951800107955933,
+      "learning_rate": 0.0004970639279024235,
+      "loss": 3.7453,
+      "step": 605
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9465222358703613,
+      "learning_rate": 0.0004970480671444613,
+      "loss": 3.727,
+      "step": 606
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.013688564300537,
+      "learning_rate": 0.0004970321639160723,
+      "loss": 3.7229,
+      "step": 607
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0122108459472656,
+      "learning_rate": 0.0004970162182199904,
+      "loss": 3.5292,
+      "step": 608
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.031880497932434,
+      "learning_rate": 0.0004970002300589571,
+      "loss": 3.718,
+      "step": 609
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9461362957954407,
+      "learning_rate": 0.0004969841994357207,
+      "loss": 3.7372,
+      "step": 610
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8112237453460693,
+      "learning_rate": 0.0004969681263530372,
+      "loss": 3.556,
+      "step": 611
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9459866881370544,
+      "learning_rate": 0.0004969520108136696,
+      "loss": 3.7211,
+      "step": 612
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.87845778465271,
+      "learning_rate": 0.0004969358528203884,
+      "loss": 3.6556,
+      "step": 613
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8740374445915222,
+      "learning_rate": 0.0004969196523759713,
+      "loss": 3.6772,
+      "step": 614
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9206240177154541,
+      "learning_rate": 0.0004969034094832036,
+      "loss": 3.6392,
+      "step": 615
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9973964691162109,
+      "learning_rate": 0.0004968871241448774,
+      "loss": 3.6835,
+      "step": 616
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8693141937255859,
+      "learning_rate": 0.0004968707963637924,
+      "loss": 3.6174,
+      "step": 617
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9450755715370178,
+      "learning_rate": 0.0004968544261427555,
+      "loss": 3.7216,
+      "step": 618
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9990179538726807,
+      "learning_rate": 0.0004968380134845811,
+      "loss": 3.8213,
+      "step": 619
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8987605571746826,
+      "learning_rate": 0.0004968215583920903,
+      "loss": 3.614,
+      "step": 620
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8715751767158508,
+      "learning_rate": 0.0004968050608681125,
+      "loss": 3.5939,
+      "step": 621
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9483709335327148,
+      "learning_rate": 0.0004967885209154833,
+      "loss": 3.6713,
+      "step": 622
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9490387439727783,
+      "learning_rate": 0.0004967719385370465,
+      "loss": 3.5721,
+      "step": 623
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9424468278884888,
+      "learning_rate": 0.0004967553137356527,
+      "loss": 3.7448,
+      "step": 624
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0020875930786133,
+      "learning_rate": 0.0004967386465141596,
+      "loss": 3.6202,
+      "step": 625
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9360184073448181,
+      "learning_rate": 0.0004967219368754329,
+      "loss": 3.5842,
+      "step": 626
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.961264431476593,
+      "learning_rate": 0.000496705184822345,
+      "loss": 3.6114,
+      "step": 627
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9263641238212585,
+      "learning_rate": 0.0004966883903577757,
+      "loss": 3.5792,
+      "step": 628
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8265263438224792,
+      "learning_rate": 0.0004966715534846123,
+      "loss": 3.6818,
+      "step": 629
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9687116742134094,
+      "learning_rate": 0.0004966546742057491,
+      "loss": 3.7072,
+      "step": 630
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9059762358665466,
+      "learning_rate": 0.000496637752524088,
+      "loss": 3.6933,
+      "step": 631
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9927087426185608,
+      "learning_rate": 0.000496620788442538,
+      "loss": 3.593,
+      "step": 632
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9516920447349548,
+      "learning_rate": 0.0004966037819640153,
+      "loss": 3.6701,
+      "step": 633
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9022125005722046,
+      "learning_rate": 0.0004965867330914437,
+      "loss": 3.6787,
+      "step": 634
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8861287832260132,
+      "learning_rate": 0.000496569641827754,
+      "loss": 3.7285,
+      "step": 635
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8411892056465149,
+      "learning_rate": 0.0004965525081758843,
+      "loss": 3.7808,
+      "step": 636
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8907370567321777,
+      "learning_rate": 0.0004965353321387803,
+      "loss": 3.6188,
+      "step": 637
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9059764742851257,
+      "learning_rate": 0.0004965181137193946,
+      "loss": 3.6357,
+      "step": 638
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9706633687019348,
+      "learning_rate": 0.0004965008529206872,
+      "loss": 3.4703,
+      "step": 639
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8777306079864502,
+      "learning_rate": 0.0004964835497456255,
+      "loss": 3.6519,
+      "step": 640
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9662359356880188,
+      "learning_rate": 0.0004964662041971841,
+      "loss": 3.6009,
+      "step": 641
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9643966555595398,
+      "learning_rate": 0.000496448816278345,
+      "loss": 3.5765,
+      "step": 642
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.965904951095581,
+      "learning_rate": 0.0004964313859920972,
+      "loss": 3.7324,
+      "step": 643
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9563618898391724,
+      "learning_rate": 0.0004964139133414373,
+      "loss": 3.7137,
+      "step": 644
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9152633547782898,
+      "learning_rate": 0.0004963963983293691,
+      "loss": 3.7103,
+      "step": 645
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.92173832654953,
+      "learning_rate": 0.0004963788409589035,
+      "loss": 3.6352,
+      "step": 646
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0018508434295654,
+      "learning_rate": 0.0004963612412330589,
+      "loss": 3.6581,
+      "step": 647
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9896060228347778,
+      "learning_rate": 0.0004963435991548608,
+      "loss": 3.7006,
+      "step": 648
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9749443531036377,
+      "learning_rate": 0.0004963259147273422,
+      "loss": 3.8044,
+      "step": 649
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8835206031799316,
+      "learning_rate": 0.0004963081879535431,
+      "loss": 3.6068,
+      "step": 650
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8468639850616455,
+      "learning_rate": 0.0004962904188365112,
+      "loss": 3.7224,
+      "step": 651
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8811071515083313,
+      "learning_rate": 0.000496272607379301,
+      "loss": 3.6173,
+      "step": 652
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8909751772880554,
+      "learning_rate": 0.0004962547535849745,
+      "loss": 3.6294,
+      "step": 653
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9179403185844421,
+      "learning_rate": 0.0004962368574566011,
+      "loss": 3.6692,
+      "step": 654
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.989331066608429,
+      "learning_rate": 0.0004962189189972573,
+      "loss": 3.6605,
+      "step": 655
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9141680598258972,
+      "learning_rate": 0.0004962009382100268,
+      "loss": 3.7447,
+      "step": 656
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0156253576278687,
+      "learning_rate": 0.0004961829150980009,
+      "loss": 3.8072,
+      "step": 657
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.904231071472168,
+      "learning_rate": 0.0004961648496642778,
+      "loss": 3.649,
+      "step": 658
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1194738149642944,
+      "learning_rate": 0.0004961467419119634,
+      "loss": 3.7544,
+      "step": 659
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9274888634681702,
+      "learning_rate": 0.0004961285918441704,
+      "loss": 3.6904,
+      "step": 660
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9246535897254944,
+      "learning_rate": 0.0004961103994640192,
+      "loss": 3.5769,
+      "step": 661
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.013225793838501,
+      "learning_rate": 0.000496092164774637,
+      "loss": 3.6117,
+      "step": 662
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9869426488876343,
+      "learning_rate": 0.0004960738877791589,
+      "loss": 3.601,
+      "step": 663
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9690237045288086,
+      "learning_rate": 0.0004960555684807266,
+      "loss": 3.6647,
+      "step": 664
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9183991551399231,
+      "learning_rate": 0.0004960372068824896,
+      "loss": 3.6451,
+      "step": 665
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9161506295204163,
+      "learning_rate": 0.0004960188029876044,
+      "loss": 3.6659,
+      "step": 666
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0477174520492554,
+      "learning_rate": 0.000496000356799235,
+      "loss": 3.7068,
+      "step": 667
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9970843195915222,
+      "learning_rate": 0.0004959818683205523,
+      "loss": 3.7798,
+      "step": 668
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9160288572311401,
+      "learning_rate": 0.0004959633375547348,
+      "loss": 3.6924,
+      "step": 669
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8371219635009766,
+      "learning_rate": 0.0004959447645049681,
+      "loss": 3.6469,
+      "step": 670
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2257660627365112,
+      "learning_rate": 0.0004959261491744452,
+      "loss": 3.6198,
+      "step": 671
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9227054715156555,
+      "learning_rate": 0.0004959074915663661,
+      "loss": 3.5663,
+      "step": 672
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9690736532211304,
+      "learning_rate": 0.0004958887916839386,
+      "loss": 3.7607,
+      "step": 673
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9061338305473328,
+      "learning_rate": 0.000495870049530377,
+      "loss": 3.7882,
+      "step": 674
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9429351091384888,
+      "learning_rate": 0.0004958512651089036,
+      "loss": 3.6591,
+      "step": 675
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8772246241569519,
+      "learning_rate": 0.0004958324384227477,
+      "loss": 3.5851,
+      "step": 676
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8267006874084473,
+      "learning_rate": 0.0004958135694751455,
+      "loss": 3.6412,
+      "step": 677
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8813744783401489,
+      "learning_rate": 0.0004957946582693412,
+      "loss": 3.7528,
+      "step": 678
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9163822531700134,
+      "learning_rate": 0.0004957757048085856,
+      "loss": 3.7047,
+      "step": 679
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8698851466178894,
+      "learning_rate": 0.0004957567090961369,
+      "loss": 3.7463,
+      "step": 680
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9418593049049377,
+      "learning_rate": 0.000495737671135261,
+      "loss": 3.7552,
+      "step": 681
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9326760768890381,
+      "learning_rate": 0.0004957185909292306,
+      "loss": 3.7517,
+      "step": 682
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9184606075286865,
+      "learning_rate": 0.0004956994684813257,
+      "loss": 3.6209,
+      "step": 683
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8675989508628845,
+      "learning_rate": 0.0004956803037948338,
+      "loss": 3.704,
+      "step": 684
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9268028140068054,
+      "learning_rate": 0.0004956610968730495,
+      "loss": 3.7821,
+      "step": 685
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9288507699966431,
+      "learning_rate": 0.0004956418477192748,
+      "loss": 3.6035,
+      "step": 686
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8930710554122925,
+      "learning_rate": 0.0004956225563368187,
+      "loss": 3.7,
+      "step": 687
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0005407333374023,
+      "learning_rate": 0.0004956032227289976,
+      "loss": 3.7182,
+      "step": 688
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0349626541137695,
+      "learning_rate": 0.0004955838468991353,
+      "loss": 3.6968,
+      "step": 689
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8456838130950928,
+      "learning_rate": 0.0004955644288505627,
+      "loss": 3.6307,
+      "step": 690
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9580467343330383,
+      "learning_rate": 0.0004955449685866179,
+      "loss": 3.6693,
+      "step": 691
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8633206486701965,
+      "learning_rate": 0.0004955254661106464,
+      "loss": 3.563,
+      "step": 692
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9008590579032898,
+      "learning_rate": 0.000495505921426001,
+      "loss": 3.5747,
+      "step": 693
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9075567722320557,
+      "learning_rate": 0.0004954863345360414,
+      "loss": 3.5997,
+      "step": 694
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8881756663322449,
+      "learning_rate": 0.000495466705444135,
+      "loss": 3.5613,
+      "step": 695
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8162988424301147,
+      "learning_rate": 0.0004954470341536563,
+      "loss": 3.623,
+      "step": 696
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9488250017166138,
+      "learning_rate": 0.000495427320667987,
+      "loss": 3.6673,
+      "step": 697
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8821372985839844,
+      "learning_rate": 0.000495407564990516,
+      "loss": 3.5725,
+      "step": 698
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8585495352745056,
+      "learning_rate": 0.0004953877671246395,
+      "loss": 3.5775,
+      "step": 699
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8403851985931396,
+      "learning_rate": 0.0004953679270737611,
+      "loss": 3.8019,
+      "step": 700
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9948201179504395,
+      "learning_rate": 0.0004953480448412914,
+      "loss": 3.6131,
+      "step": 701
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8229441046714783,
+      "learning_rate": 0.0004953281204306487,
+      "loss": 3.5401,
+      "step": 702
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8649178147315979,
+      "learning_rate": 0.0004953081538452579,
+      "loss": 3.4134,
+      "step": 703
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8966211080551147,
+      "learning_rate": 0.0004952881450885515,
+      "loss": 3.7468,
+      "step": 704
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0185855627059937,
+      "learning_rate": 0.0004952680941639694,
+      "loss": 3.6969,
+      "step": 705
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8968634605407715,
+      "learning_rate": 0.0004952480010749585,
+      "loss": 3.676,
+      "step": 706
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8802434206008911,
+      "learning_rate": 0.0004952278658249731,
+      "loss": 3.6423,
+      "step": 707
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9503376483917236,
+      "learning_rate": 0.0004952076884174746,
+      "loss": 3.6148,
+      "step": 708
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9169230461120605,
+      "learning_rate": 0.0004951874688559318,
+      "loss": 3.6117,
+      "step": 709
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9358011484146118,
+      "learning_rate": 0.0004951672071438207,
+      "loss": 3.7518,
+      "step": 710
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9428926110267639,
+      "learning_rate": 0.0004951469032846243,
+      "loss": 3.7773,
+      "step": 711
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9884240627288818,
+      "learning_rate": 0.0004951265572818334,
+      "loss": 3.6447,
+      "step": 712
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8522127866744995,
+      "learning_rate": 0.0004951061691389455,
+      "loss": 3.5563,
+      "step": 713
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8171650767326355,
+      "learning_rate": 0.0004950857388594656,
+      "loss": 3.6495,
+      "step": 714
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8851253390312195,
+      "learning_rate": 0.0004950652664469059,
+      "loss": 3.6131,
+      "step": 715
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8465543985366821,
+      "learning_rate": 0.000495044751904786,
+      "loss": 3.6781,
+      "step": 716
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9340558648109436,
+      "learning_rate": 0.0004950241952366322,
+      "loss": 3.7341,
+      "step": 717
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8248949646949768,
+      "learning_rate": 0.0004950035964459789,
+      "loss": 3.5921,
+      "step": 718
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9273523092269897,
+      "learning_rate": 0.0004949829555363669,
+      "loss": 3.6564,
+      "step": 719
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9097087979316711,
+      "learning_rate": 0.0004949622725113448,
+      "loss": 3.5832,
+      "step": 720
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.799820601940155,
+      "learning_rate": 0.0004949415473744683,
+      "loss": 3.6226,
+      "step": 721
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9359858632087708,
+      "learning_rate": 0.0004949207801293001,
+      "loss": 3.6449,
+      "step": 722
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8636605739593506,
+      "learning_rate": 0.0004948999707794105,
+      "loss": 3.6189,
+      "step": 723
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8730395436286926,
+      "learning_rate": 0.0004948791193283765,
+      "loss": 3.6658,
+      "step": 724
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9037566781044006,
+      "learning_rate": 0.0004948582257797834,
+      "loss": 3.6856,
+      "step": 725
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8765959143638611,
+      "learning_rate": 0.0004948372901372224,
+      "loss": 3.5039,
+      "step": 726
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9535045623779297,
+      "learning_rate": 0.000494816312404293,
+      "loss": 3.6947,
+      "step": 727
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9718337059020996,
+      "learning_rate": 0.0004947952925846013,
+      "loss": 3.5829,
+      "step": 728
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9038040041923523,
+      "learning_rate": 0.0004947742306817608,
+      "loss": 3.5878,
+      "step": 729
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9002828598022461,
+      "learning_rate": 0.0004947531266993924,
+      "loss": 3.6424,
+      "step": 730
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9017542600631714,
+      "learning_rate": 0.0004947319806411243,
+      "loss": 3.7354,
+      "step": 731
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8402687907218933,
+      "learning_rate": 0.0004947107925105912,
+      "loss": 3.6144,
+      "step": 732
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8400694727897644,
+      "learning_rate": 0.0004946895623114362,
+      "loss": 3.6566,
+      "step": 733
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.899815022945404,
+      "learning_rate": 0.0004946682900473087,
+      "loss": 3.6615,
+      "step": 734
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9010912179946899,
+      "learning_rate": 0.0004946469757218657,
+      "loss": 3.7581,
+      "step": 735
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8260379433631897,
+      "learning_rate": 0.0004946256193387714,
+      "loss": 3.6277,
+      "step": 736
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8734118342399597,
+      "learning_rate": 0.000494604220901697,
+      "loss": 3.5827,
+      "step": 737
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8689841032028198,
+      "learning_rate": 0.0004945827804143216,
+      "loss": 3.7219,
+      "step": 738
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8634633421897888,
+      "learning_rate": 0.0004945612978803307,
+      "loss": 3.4742,
+      "step": 739
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9236171841621399,
+      "learning_rate": 0.0004945397733034175,
+      "loss": 3.6558,
+      "step": 740
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9720011949539185,
+      "learning_rate": 0.0004945182066872823,
+      "loss": 3.6096,
+      "step": 741
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9223105311393738,
+      "learning_rate": 0.0004944965980356326,
+      "loss": 3.6497,
+      "step": 742
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8663374781608582,
+      "learning_rate": 0.0004944749473521834,
+      "loss": 3.695,
+      "step": 743
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7778543829917908,
+      "learning_rate": 0.0004944532546406565,
+      "loss": 3.6928,
+      "step": 744
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9464306831359863,
+      "learning_rate": 0.0004944315199047812,
+      "loss": 3.7198,
+      "step": 745
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.871847927570343,
+      "learning_rate": 0.0004944097431482939,
+      "loss": 3.6822,
+      "step": 746
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8734187483787537,
+      "learning_rate": 0.0004943879243749382,
+      "loss": 3.647,
+      "step": 747
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9731336832046509,
+      "learning_rate": 0.0004943660635884652,
+      "loss": 3.6557,
+      "step": 748
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9924262166023254,
+      "learning_rate": 0.000494344160792633,
+      "loss": 3.7084,
+      "step": 749
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9291201829910278,
+      "learning_rate": 0.0004943222159912069,
+      "loss": 3.6337,
+      "step": 750
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9033045172691345,
+      "learning_rate": 0.0004943002291879593,
+      "loss": 3.6388,
+      "step": 751
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0025124549865723,
+      "learning_rate": 0.0004942782003866703,
+      "loss": 3.7116,
+      "step": 752
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9180055260658264,
+      "learning_rate": 0.0004942561295911267,
+      "loss": 3.6105,
+      "step": 753
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8433054089546204,
+      "learning_rate": 0.0004942340168051226,
+      "loss": 3.6694,
+      "step": 754
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9406448602676392,
+      "learning_rate": 0.0004942118620324597,
+      "loss": 3.6136,
+      "step": 755
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8985976576805115,
+      "learning_rate": 0.0004941896652769466,
+      "loss": 3.6743,
+      "step": 756
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8895725011825562,
+      "learning_rate": 0.0004941674265423992,
+      "loss": 3.5532,
+      "step": 757
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8536813855171204,
+      "learning_rate": 0.0004941451458326404,
+      "loss": 3.7154,
+      "step": 758
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8130298256874084,
+      "learning_rate": 0.0004941228231515008,
+      "loss": 3.4941,
+      "step": 759
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.900328516960144,
+      "learning_rate": 0.0004941004585028177,
+      "loss": 3.6643,
+      "step": 760
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9134331345558167,
+      "learning_rate": 0.000494078051890436,
+      "loss": 3.6062,
+      "step": 761
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8630452156066895,
+      "learning_rate": 0.0004940556033182075,
+      "loss": 3.5736,
+      "step": 762
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.872016966342926,
+      "learning_rate": 0.0004940331127899915,
+      "loss": 3.7443,
+      "step": 763
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8469942808151245,
+      "learning_rate": 0.0004940105803096544,
+      "loss": 3.498,
+      "step": 764
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9625217914581299,
+      "learning_rate": 0.0004939880058810696,
+      "loss": 3.5907,
+      "step": 765
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0333412885665894,
+      "learning_rate": 0.0004939653895081181,
+      "loss": 3.5302,
+      "step": 766
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9309979677200317,
+      "learning_rate": 0.000493942731194688,
+      "loss": 3.5486,
+      "step": 767
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.873720109462738,
+      "learning_rate": 0.0004939200309446741,
+      "loss": 3.6667,
+      "step": 768
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.0009433031082153,
+      "learning_rate": 0.0004938972887619793,
+      "loss": 3.6041,
+      "step": 769
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8196940422058105,
+      "learning_rate": 0.0004938745046505129,
+      "loss": 3.6099,
+      "step": 770
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9370154738426208,
+      "learning_rate": 0.0004938516786141921,
+      "loss": 3.7256,
+      "step": 771
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9386774301528931,
+      "learning_rate": 0.0004938288106569407,
+      "loss": 3.7586,
+      "step": 772
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7597833275794983,
+      "learning_rate": 0.0004938059007826901,
+      "loss": 3.541,
+      "step": 773
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8484575748443604,
+      "learning_rate": 0.0004937829489953787,
+      "loss": 3.7377,
+      "step": 774
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.830502450466156,
+      "learning_rate": 0.0004937599552989521,
+      "loss": 3.6238,
+      "step": 775
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8506859540939331,
+      "learning_rate": 0.0004937369196973633,
+      "loss": 3.6765,
+      "step": 776
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8493182063102722,
+      "learning_rate": 0.0004937138421945724,
+      "loss": 3.6292,
+      "step": 777
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8898391723632812,
+      "learning_rate": 0.0004936907227945467,
+      "loss": 3.6496,
+      "step": 778
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9002225399017334,
+      "learning_rate": 0.0004936675615012606,
+      "loss": 3.673,
+      "step": 779
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8686527609825134,
+      "learning_rate": 0.0004936443583186958,
+      "loss": 3.663,
+      "step": 780
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8821991086006165,
+      "learning_rate": 0.0004936211132508413,
+      "loss": 3.6627,
+      "step": 781
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.852533757686615,
+      "learning_rate": 0.0004935978263016931,
+      "loss": 3.538,
+      "step": 782
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8961055278778076,
+      "learning_rate": 0.0004935744974752546,
+      "loss": 3.6461,
+      "step": 783
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9409698843955994,
+      "learning_rate": 0.0004935511267755361,
+      "loss": 3.545,
+      "step": 784
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8036273717880249,
+      "learning_rate": 0.0004935277142065556,
+      "loss": 3.5735,
+      "step": 785
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8235346674919128,
+      "learning_rate": 0.0004935042597723376,
+      "loss": 3.6634,
+      "step": 786
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8754570484161377,
+      "learning_rate": 0.0004934807634769145,
+      "loss": 3.7106,
+      "step": 787
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8720112442970276,
+      "learning_rate": 0.0004934572253243255,
+      "loss": 3.5811,
+      "step": 788
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8393718600273132,
+      "learning_rate": 0.0004934336453186171,
+      "loss": 3.6455,
+      "step": 789
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.795387327671051,
+      "learning_rate": 0.0004934100234638429,
+      "loss": 3.7253,
+      "step": 790
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9040440917015076,
+      "learning_rate": 0.0004933863597640638,
+      "loss": 3.623,
+      "step": 791
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8612719178199768,
+      "learning_rate": 0.0004933626542233478,
+      "loss": 3.7458,
+      "step": 792
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8143512606620789,
+      "learning_rate": 0.0004933389068457704,
+      "loss": 3.5961,
+      "step": 793
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7966370582580566,
+      "learning_rate": 0.0004933151176354138,
+      "loss": 3.6326,
+      "step": 794
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8990934491157532,
+      "learning_rate": 0.0004932912865963677,
+      "loss": 3.6588,
+      "step": 795
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9208589792251587,
+      "learning_rate": 0.000493267413732729,
+      "loss": 3.682,
+      "step": 796
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.840647280216217,
+      "learning_rate": 0.0004932434990486017,
+      "loss": 3.6069,
+      "step": 797
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.799932062625885,
+      "learning_rate": 0.000493219542548097,
+      "loss": 3.4894,
+      "step": 798
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9299008846282959,
+      "learning_rate": 0.0004931955442353333,
+      "loss": 3.7636,
+      "step": 799
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8207058906555176,
+      "learning_rate": 0.0004931715041144361,
+      "loss": 3.7038,
+      "step": 800
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9075020551681519,
+      "learning_rate": 0.0004931474221895383,
+      "loss": 3.58,
+      "step": 801
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8652316331863403,
+      "learning_rate": 0.0004931232984647798,
+      "loss": 3.5738,
+      "step": 802
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9437047243118286,
+      "learning_rate": 0.0004930991329443079,
+      "loss": 3.7435,
+      "step": 803
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8851674199104309,
+      "learning_rate": 0.0004930749256322766,
+      "loss": 3.7059,
+      "step": 804
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8456943035125732,
+      "learning_rate": 0.0004930506765328477,
+      "loss": 3.6934,
+      "step": 805
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8298068046569824,
+      "learning_rate": 0.0004930263856501899,
+      "loss": 3.51,
+      "step": 806
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8694912195205688,
+      "learning_rate": 0.0004930020529884789,
+      "loss": 3.6908,
+      "step": 807
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8628725409507751,
+      "learning_rate": 0.000492977678551898,
+      "loss": 3.7173,
+      "step": 808
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8303777575492859,
+      "learning_rate": 0.0004929532623446372,
+      "loss": 3.5715,
+      "step": 809
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7596468329429626,
+      "learning_rate": 0.000492928804370894,
+      "loss": 3.535,
+      "step": 810
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9455280900001526,
+      "learning_rate": 0.0004929043046348732,
+      "loss": 3.6283,
+      "step": 811
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8608766198158264,
+      "learning_rate": 0.0004928797631407863,
+      "loss": 3.6562,
+      "step": 812
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8241950869560242,
+      "learning_rate": 0.0004928551798928525,
+      "loss": 3.6279,
+      "step": 813
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.836773157119751,
+      "learning_rate": 0.0004928305548952978,
+      "loss": 3.5184,
+      "step": 814
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8365269303321838,
+      "learning_rate": 0.0004928058881523557,
+      "loss": 3.6565,
+      "step": 815
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8501466512680054,
+      "learning_rate": 0.0004927811796682666,
+      "loss": 3.7131,
+      "step": 816
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8528035879135132,
+      "learning_rate": 0.000492756429447278,
+      "loss": 3.5438,
+      "step": 817
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8384031057357788,
+      "learning_rate": 0.0004927316374936449,
+      "loss": 3.6685,
+      "step": 818
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.13548743724823,
+      "learning_rate": 0.0004927068038116293,
+      "loss": 3.6312,
+      "step": 819
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8726980090141296,
+      "learning_rate": 0.0004926819284055006,
+      "loss": 3.704,
+      "step": 820
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8911085724830627,
+      "learning_rate": 0.0004926570112795349,
+      "loss": 3.6911,
+      "step": 821
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8530990481376648,
+      "learning_rate": 0.0004926320524380159,
+      "loss": 3.6812,
+      "step": 822
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9349095821380615,
+      "learning_rate": 0.0004926070518852341,
+      "loss": 3.5695,
+      "step": 823
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8291351795196533,
+      "learning_rate": 0.0004925820096254877,
+      "loss": 3.5025,
+      "step": 824
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9228273630142212,
+      "learning_rate": 0.0004925569256630816,
+      "loss": 3.6815,
+      "step": 825
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8727307319641113,
+      "learning_rate": 0.0004925318000023279,
+      "loss": 3.5924,
+      "step": 826
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.853922426700592,
+      "learning_rate": 0.0004925066326475461,
+      "loss": 3.5766,
+      "step": 827
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9327370524406433,
+      "learning_rate": 0.0004924814236030629,
+      "loss": 3.6038,
+      "step": 828
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8882296681404114,
+      "learning_rate": 0.0004924561728732118,
+      "loss": 3.6118,
+      "step": 829
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8594300746917725,
+      "learning_rate": 0.0004924308804623339,
+      "loss": 3.6208,
+      "step": 830
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8235106468200684,
+      "learning_rate": 0.0004924055463747769,
+      "loss": 3.5596,
+      "step": 831
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8358860611915588,
+      "learning_rate": 0.0004923801706148965,
+      "loss": 3.5078,
+      "step": 832
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.863490104675293,
+      "learning_rate": 0.0004923547531870548,
+      "loss": 3.7563,
+      "step": 833
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.841188371181488,
+      "learning_rate": 0.0004923292940956215,
+      "loss": 3.6264,
+      "step": 834
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9689059257507324,
+      "learning_rate": 0.0004923037933449731,
+      "loss": 3.6982,
+      "step": 835
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8078739643096924,
+      "learning_rate": 0.0004922782509394937,
+      "loss": 3.627,
+      "step": 836
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8360235095024109,
+      "learning_rate": 0.0004922526668835741,
+      "loss": 3.5822,
+      "step": 837
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8151078820228577,
+      "learning_rate": 0.0004922270411816126,
+      "loss": 3.6288,
+      "step": 838
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8234032392501831,
+      "learning_rate": 0.0004922013738380147,
+      "loss": 3.6923,
+      "step": 839
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8429092168807983,
+      "learning_rate": 0.0004921756648571928,
+      "loss": 3.6477,
+      "step": 840
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7991187572479248,
+      "learning_rate": 0.0004921499142435666,
+      "loss": 3.6118,
+      "step": 841
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7951987981796265,
+      "learning_rate": 0.0004921241220015627,
+      "loss": 3.6724,
+      "step": 842
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8163877129554749,
+      "learning_rate": 0.0004920982881356156,
+      "loss": 3.6376,
+      "step": 843
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8272448778152466,
+      "learning_rate": 0.0004920724126501659,
+      "loss": 3.622,
+      "step": 844
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9418625831604004,
+      "learning_rate": 0.0004920464955496622,
+      "loss": 3.6689,
+      "step": 845
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8513910174369812,
+      "learning_rate": 0.0004920205368385598,
+      "loss": 3.4536,
+      "step": 846
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7715315222740173,
+      "learning_rate": 0.0004919945365213214,
+      "loss": 3.5744,
+      "step": 847
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8094183802604675,
+      "learning_rate": 0.0004919684946024168,
+      "loss": 3.7615,
+      "step": 848
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8636196851730347,
+      "learning_rate": 0.0004919424110863227,
+      "loss": 3.6143,
+      "step": 849
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8138225674629211,
+      "learning_rate": 0.0004919162859775235,
+      "loss": 3.6715,
+      "step": 850
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8703018426895142,
+      "learning_rate": 0.00049189011928051,
+      "loss": 3.5849,
+      "step": 851
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7516616582870483,
+      "learning_rate": 0.0004918639109997809,
+      "loss": 3.6899,
+      "step": 852
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8473501205444336,
+      "learning_rate": 0.0004918376611398415,
+      "loss": 3.6175,
+      "step": 853
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9071518778800964,
+      "learning_rate": 0.0004918113697052046,
+      "loss": 3.6823,
+      "step": 854
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7907838225364685,
+      "learning_rate": 0.0004917850367003898,
+      "loss": 3.612,
+      "step": 855
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8364929556846619,
+      "learning_rate": 0.0004917586621299243,
+      "loss": 3.6682,
+      "step": 856
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8281691670417786,
+      "learning_rate": 0.0004917322459983421,
+      "loss": 3.5506,
+      "step": 857
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7997773289680481,
+      "learning_rate": 0.0004917057883101843,
+      "loss": 3.5342,
+      "step": 858
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9194799065589905,
+      "learning_rate": 0.0004916792890699995,
+      "loss": 3.5594,
+      "step": 859
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8962545394897461,
+      "learning_rate": 0.000491652748282343,
+      "loss": 3.7193,
+      "step": 860
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8621538281440735,
+      "learning_rate": 0.0004916261659517777,
+      "loss": 3.6989,
+      "step": 861
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7588165402412415,
+      "learning_rate": 0.0004915995420828732,
+      "loss": 3.4604,
+      "step": 862
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9234063029289246,
+      "learning_rate": 0.0004915728766802066,
+      "loss": 3.6228,
+      "step": 863
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8838184475898743,
+      "learning_rate": 0.000491546169748362,
+      "loss": 3.6242,
+      "step": 864
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8651776313781738,
+      "learning_rate": 0.0004915194212919305,
+      "loss": 3.7815,
+      "step": 865
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0226529836654663,
+      "learning_rate": 0.0004914926313155106,
+      "loss": 3.4453,
+      "step": 866
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8438092470169067,
+      "learning_rate": 0.0004914657998237078,
+      "loss": 3.6444,
+      "step": 867
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9107071161270142,
+      "learning_rate": 0.0004914389268211346,
+      "loss": 3.7138,
+      "step": 868
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7964327335357666,
+      "learning_rate": 0.0004914120123124108,
+      "loss": 3.6776,
+      "step": 869
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8652677536010742,
+      "learning_rate": 0.0004913850563021636,
+      "loss": 3.78,
+      "step": 870
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8645572662353516,
+      "learning_rate": 0.0004913580587950267,
+      "loss": 3.5867,
+      "step": 871
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.805211067199707,
+      "learning_rate": 0.0004913310197956416,
+      "loss": 3.6309,
+      "step": 872
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8409008979797363,
+      "learning_rate": 0.0004913039393086563,
+      "loss": 3.5873,
+      "step": 873
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8912443518638611,
+      "learning_rate": 0.0004912768173387264,
+      "loss": 3.6151,
+      "step": 874
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8684778809547424,
+      "learning_rate": 0.0004912496538905145,
+      "loss": 3.6755,
+      "step": 875
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8110811114311218,
+      "learning_rate": 0.0004912224489686903,
+      "loss": 3.4404,
+      "step": 876
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9000563025474548,
+      "learning_rate": 0.0004911952025779306,
+      "loss": 3.7488,
+      "step": 877
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8925985097885132,
+      "learning_rate": 0.0004911679147229194,
+      "loss": 3.7425,
+      "step": 878
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7726958394050598,
+      "learning_rate": 0.0004911405854083479,
+      "loss": 3.6524,
+      "step": 879
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.786937415599823,
+      "learning_rate": 0.0004911132146389141,
+      "loss": 3.6178,
+      "step": 880
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8084043264389038,
+      "learning_rate": 0.0004910858024193236,
+      "loss": 3.531,
+      "step": 881
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8019929528236389,
+      "learning_rate": 0.0004910583487542886,
+      "loss": 3.7352,
+      "step": 882
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7468410730361938,
+      "learning_rate": 0.0004910308536485291,
+      "loss": 3.484,
+      "step": 883
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8613249659538269,
+      "learning_rate": 0.0004910033171067713,
+      "loss": 3.5572,
+      "step": 884
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8318970203399658,
+      "learning_rate": 0.0004909757391337496,
+      "loss": 3.5578,
+      "step": 885
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8411234617233276,
+      "learning_rate": 0.0004909481197342046,
+      "loss": 3.5871,
+      "step": 886
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7730182409286499,
+      "learning_rate": 0.0004909204589128845,
+      "loss": 3.6533,
+      "step": 887
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8421241641044617,
+      "learning_rate": 0.0004908927566745446,
+      "loss": 3.6285,
+      "step": 888
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8909102082252502,
+      "learning_rate": 0.000490865013023947,
+      "loss": 3.6253,
+      "step": 889
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8533161878585815,
+      "learning_rate": 0.0004908372279658614,
+      "loss": 3.5395,
+      "step": 890
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8164993524551392,
+      "learning_rate": 0.0004908094015050643,
+      "loss": 3.5685,
+      "step": 891
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7901835441589355,
+      "learning_rate": 0.0004907815336463394,
+      "loss": 3.6154,
+      "step": 892
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9597400426864624,
+      "learning_rate": 0.0004907536243944774,
+      "loss": 3.7743,
+      "step": 893
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8602568507194519,
+      "learning_rate": 0.0004907256737542765,
+      "loss": 3.5457,
+      "step": 894
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8954022526741028,
+      "learning_rate": 0.0004906976817305413,
+      "loss": 3.5667,
+      "step": 895
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9862939119338989,
+      "learning_rate": 0.0004906696483280844,
+      "loss": 3.664,
+      "step": 896
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8378450870513916,
+      "learning_rate": 0.0004906415735517248,
+      "loss": 3.68,
+      "step": 897
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7724777460098267,
+      "learning_rate": 0.0004906134574062888,
+      "loss": 3.5164,
+      "step": 898
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8759022951126099,
+      "learning_rate": 0.0004905852998966102,
+      "loss": 3.5678,
+      "step": 899
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7621853947639465,
+      "learning_rate": 0.0004905571010275295,
+      "loss": 3.466,
+      "step": 900
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8283673524856567,
+      "learning_rate": 0.0004905288608038943,
+      "loss": 3.5888,
+      "step": 901
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8321616053581238,
+      "learning_rate": 0.0004905005792305595,
+      "loss": 3.6799,
+      "step": 902
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7852991223335266,
+      "learning_rate": 0.0004904722563123869,
+      "loss": 3.566,
+      "step": 903
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8530229330062866,
+      "learning_rate": 0.0004904438920542457,
+      "loss": 3.58,
+      "step": 904
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8264917135238647,
+      "learning_rate": 0.000490415486461012,
+      "loss": 3.6062,
+      "step": 905
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7997876405715942,
+      "learning_rate": 0.0004903870395375691,
+      "loss": 3.4823,
+      "step": 906
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8135886192321777,
+      "learning_rate": 0.0004903585512888072,
+      "loss": 3.5646,
+      "step": 907
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8352776765823364,
+      "learning_rate": 0.0004903300217196239,
+      "loss": 3.6718,
+      "step": 908
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8962761163711548,
+      "learning_rate": 0.0004903014508349237,
+      "loss": 3.5917,
+      "step": 909
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8091096878051758,
+      "learning_rate": 0.0004902728386396184,
+      "loss": 3.6274,
+      "step": 910
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7973507642745972,
+      "learning_rate": 0.0004902441851386266,
+      "loss": 3.6302,
+      "step": 911
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8099647164344788,
+      "learning_rate": 0.0004902154903368741,
+      "loss": 3.544,
+      "step": 912
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7872130870819092,
+      "learning_rate": 0.0004901867542392942,
+      "loss": 3.6173,
+      "step": 913
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8407408595085144,
+      "learning_rate": 0.0004901579768508265,
+      "loss": 3.6839,
+      "step": 914
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8333362936973572,
+      "learning_rate": 0.0004901291581764186,
+      "loss": 3.6181,
+      "step": 915
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8242466449737549,
+      "learning_rate": 0.0004901002982210246,
+      "loss": 3.6812,
+      "step": 916
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7317993640899658,
+      "learning_rate": 0.0004900713969896059,
+      "loss": 3.598,
+      "step": 917
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8498689532279968,
+      "learning_rate": 0.0004900424544871308,
+      "loss": 3.6426,
+      "step": 918
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8187620043754578,
+      "learning_rate": 0.0004900134707185751,
+      "loss": 3.6288,
+      "step": 919
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7988176941871643,
+      "learning_rate": 0.000489984445688921,
+      "loss": 3.6693,
+      "step": 920
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9066382646560669,
+      "learning_rate": 0.0004899553794031589,
+      "loss": 3.5619,
+      "step": 921
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7891650199890137,
+      "learning_rate": 0.0004899262718662851,
+      "loss": 3.8446,
+      "step": 922
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8057774305343628,
+      "learning_rate": 0.0004898971230833037,
+      "loss": 3.6976,
+      "step": 923
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8969120979309082,
+      "learning_rate": 0.0004898679330592259,
+      "loss": 3.5516,
+      "step": 924
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8318272829055786,
+      "learning_rate": 0.0004898387017990694,
+      "loss": 3.525,
+      "step": 925
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8637136220932007,
+      "learning_rate": 0.0004898094293078597,
+      "loss": 3.6406,
+      "step": 926
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7998780608177185,
+      "learning_rate": 0.000489780115590629,
+      "loss": 3.4636,
+      "step": 927
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8367539048194885,
+      "learning_rate": 0.0004897507606524167,
+      "loss": 3.6949,
+      "step": 928
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8102987408638,
+      "learning_rate": 0.0004897213644982691,
+      "loss": 3.5146,
+      "step": 929
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7719710469245911,
+      "learning_rate": 0.0004896919271332399,
+      "loss": 3.492,
+      "step": 930
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7758731245994568,
+      "learning_rate": 0.0004896624485623896,
+      "loss": 3.5188,
+      "step": 931
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7850921154022217,
+      "learning_rate": 0.0004896329287907862,
+      "loss": 3.5319,
+      "step": 932
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7877423167228699,
+      "learning_rate": 0.0004896033678235041,
+      "loss": 3.6709,
+      "step": 933
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.736785352230072,
+      "learning_rate": 0.0004895737656656254,
+      "loss": 3.555,
+      "step": 934
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.770460307598114,
+      "learning_rate": 0.000489544122322239,
+      "loss": 3.6151,
+      "step": 935
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8316922783851624,
+      "learning_rate": 0.000489514437798441,
+      "loss": 3.648,
+      "step": 936
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8192751407623291,
+      "learning_rate": 0.0004894847120993344,
+      "loss": 3.528,
+      "step": 937
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8254303932189941,
+      "learning_rate": 0.0004894549452300294,
+      "loss": 3.6145,
+      "step": 938
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8344351053237915,
+      "learning_rate": 0.0004894251371956433,
+      "loss": 3.6076,
+      "step": 939
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8560154438018799,
+      "learning_rate": 0.0004893952880013005,
+      "loss": 3.5497,
+      "step": 940
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8088158369064331,
+      "learning_rate": 0.0004893653976521325,
+      "loss": 3.5911,
+      "step": 941
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8597379922866821,
+      "learning_rate": 0.0004893354661532776,
+      "loss": 3.6256,
+      "step": 942
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7602822780609131,
+      "learning_rate": 0.0004893054935098814,
+      "loss": 3.607,
+      "step": 943
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8641725778579712,
+      "learning_rate": 0.0004892754797270967,
+      "loss": 3.5751,
+      "step": 944
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8781532049179077,
+      "learning_rate": 0.000489245424810083,
+      "loss": 3.6763,
+      "step": 945
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8362348675727844,
+      "learning_rate": 0.0004892153287640072,
+      "loss": 3.6466,
+      "step": 946
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8070446252822876,
+      "learning_rate": 0.0004891851915940433,
+      "loss": 3.5171,
+      "step": 947
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8270018696784973,
+      "learning_rate": 0.000489155013305372,
+      "loss": 3.5635,
+      "step": 948
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8150656819343567,
+      "learning_rate": 0.0004891247939031814,
+      "loss": 3.5941,
+      "step": 949
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.830181360244751,
+      "learning_rate": 0.0004890945333926665,
+      "loss": 3.6124,
+      "step": 950
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.808355450630188,
+      "learning_rate": 0.0004890642317790296,
+      "loss": 3.5877,
+      "step": 951
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7882115840911865,
+      "learning_rate": 0.0004890338890674797,
+      "loss": 3.6326,
+      "step": 952
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.828498125076294,
+      "learning_rate": 0.0004890035052632333,
+      "loss": 3.5982,
+      "step": 953
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7736223936080933,
+      "learning_rate": 0.0004889730803715133,
+      "loss": 3.6843,
+      "step": 954
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8319690227508545,
+      "learning_rate": 0.0004889426143975507,
+      "loss": 3.621,
+      "step": 955
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8388370275497437,
+      "learning_rate": 0.0004889121073465825,
+      "loss": 3.6456,
+      "step": 956
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8199196457862854,
+      "learning_rate": 0.0004888815592238532,
+      "loss": 3.5886,
+      "step": 957
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7664119005203247,
+      "learning_rate": 0.0004888509700346146,
+      "loss": 3.5408,
+      "step": 958
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7739691734313965,
+      "learning_rate": 0.0004888203397841253,
+      "loss": 3.5643,
+      "step": 959
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7797976136207581,
+      "learning_rate": 0.0004887896684776508,
+      "loss": 3.5595,
+      "step": 960
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8414448499679565,
+      "learning_rate": 0.0004887589561204642,
+      "loss": 3.5498,
+      "step": 961
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8154247999191284,
+      "learning_rate": 0.000488728202717845,
+      "loss": 3.6078,
+      "step": 962
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8003179430961609,
+      "learning_rate": 0.0004886974082750803,
+      "loss": 3.4887,
+      "step": 963
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8092113137245178,
+      "learning_rate": 0.0004886665727974638,
+      "loss": 3.5388,
+      "step": 964
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7909213304519653,
+      "learning_rate": 0.0004886356962902965,
+      "loss": 3.6211,
+      "step": 965
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7772814035415649,
+      "learning_rate": 0.0004886047787588866,
+      "loss": 3.594,
+      "step": 966
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.801804780960083,
+      "learning_rate": 0.0004885738202085491,
+      "loss": 3.6257,
+      "step": 967
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7865604758262634,
+      "learning_rate": 0.0004885428206446061,
+      "loss": 3.5625,
+      "step": 968
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8398476839065552,
+      "learning_rate": 0.0004885117800723869,
+      "loss": 3.6431,
+      "step": 969
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.853718638420105,
+      "learning_rate": 0.0004884806984972274,
+      "loss": 3.6405,
+      "step": 970
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8108567595481873,
+      "learning_rate": 0.0004884495759244713,
+      "loss": 3.5849,
+      "step": 971
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8177028298377991,
+      "learning_rate": 0.0004884184123594687,
+      "loss": 3.6298,
+      "step": 972
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8156768083572388,
+      "learning_rate": 0.0004883872078075771,
+      "loss": 3.5049,
+      "step": 973
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8004074692726135,
+      "learning_rate": 0.0004883559622741607,
+      "loss": 3.6292,
+      "step": 974
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8555276989936829,
+      "learning_rate": 0.0004883246757645914,
+      "loss": 3.6365,
+      "step": 975
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8024552464485168,
+      "learning_rate": 0.0004882933482842472,
+      "loss": 3.4914,
+      "step": 976
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7471779584884644,
+      "learning_rate": 0.00048826197983851396,
+      "loss": 3.5452,
+      "step": 977
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8516356945037842,
+      "learning_rate": 0.00048823057043278427,
+      "loss": 3.6427,
+      "step": 978
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8830351829528809,
+      "learning_rate": 0.00048819912007245765,
+      "loss": 3.6421,
+      "step": 979
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8727092742919922,
+      "learning_rate": 0.0004881676287629409,
+      "loss": 3.5592,
+      "step": 980
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8191733956336975,
+      "learning_rate": 0.0004881360965096476,
+      "loss": 3.5201,
+      "step": 981
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6910346150398254,
+      "learning_rate": 0.0004881045233179987,
+      "loss": 3.4483,
+      "step": 982
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8177911639213562,
+      "learning_rate": 0.00048807290919342187,
+      "loss": 3.6568,
+      "step": 983
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8026474714279175,
+      "learning_rate": 0.00048804125414135194,
+      "loss": 3.6078,
+      "step": 984
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7814860343933105,
+      "learning_rate": 0.00048800955816723083,
+      "loss": 3.4803,
+      "step": 985
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.772703230381012,
+      "learning_rate": 0.0004879778212765075,
+      "loss": 3.5197,
+      "step": 986
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8092082142829895,
+      "learning_rate": 0.0004879460434746377,
+      "loss": 3.5248,
+      "step": 987
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8448889255523682,
+      "learning_rate": 0.0004879142247670846,
+      "loss": 3.6247,
+      "step": 988
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.814842939376831,
+      "learning_rate": 0.00048788236515931807,
+      "loss": 3.6079,
+      "step": 989
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.789014995098114,
+      "learning_rate": 0.00048785046465681515,
+      "loss": 3.5347,
+      "step": 990
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7849494814872742,
+      "learning_rate": 0.00048781852326506,
+      "loss": 3.5534,
+      "step": 991
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8483198881149292,
+      "learning_rate": 0.0004877865409895438,
+      "loss": 3.734,
+      "step": 992
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7456423044204712,
+      "learning_rate": 0.0004877545178357644,
+      "loss": 3.5662,
+      "step": 993
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8502069711685181,
+      "learning_rate": 0.0004877224538092271,
+      "loss": 3.6267,
+      "step": 994
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.794623851776123,
+      "learning_rate": 0.00048769034891544416,
+      "loss": 3.6479,
+      "step": 995
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8556898832321167,
+      "learning_rate": 0.00048765820315993465,
+      "loss": 3.5804,
+      "step": 996
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8531051874160767,
+      "learning_rate": 0.0004876260165482248,
+      "loss": 3.5223,
+      "step": 997
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8192583322525024,
+      "learning_rate": 0.000487593789085848,
+      "loss": 3.5352,
+      "step": 998
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8201632499694824,
+      "learning_rate": 0.00048756152077834435,
+      "loss": 3.7469,
+      "step": 999
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.767375648021698,
+      "learning_rate": 0.0004875292116312613,
+      "loss": 3.66,
+      "step": 1000
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7790427803993225,
+      "learning_rate": 0.00048749686165015305,
+      "loss": 3.3572,
+      "step": 1001
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7836480140686035,
+      "learning_rate": 0.00048746447084058105,
+      "loss": 3.5477,
+      "step": 1002
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7594608068466187,
+      "learning_rate": 0.0004874320392081135,
+      "loss": 3.6529,
+      "step": 1003
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8746945261955261,
+      "learning_rate": 0.00048739956675832594,
+      "loss": 3.5856,
+      "step": 1004
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8683459758758545,
+      "learning_rate": 0.00048736705349680063,
+      "loss": 3.6031,
+      "step": 1005
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7671228051185608,
+      "learning_rate": 0.000487334499429127,
+      "loss": 3.5888,
+      "step": 1006
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7802176475524902,
+      "learning_rate": 0.00048730190456090154,
+      "loss": 3.5723,
+      "step": 1007
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7627329230308533,
+      "learning_rate": 0.00048726926889772773,
+      "loss": 3.5489,
+      "step": 1008
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8381568193435669,
+      "learning_rate": 0.00048723659244521576,
+      "loss": 3.5454,
+      "step": 1009
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7972750663757324,
+      "learning_rate": 0.00048720387520898343,
+      "loss": 3.5405,
+      "step": 1010
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8035085797309875,
+      "learning_rate": 0.0004871711171946549,
+      "loss": 3.5673,
+      "step": 1011
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8251034021377563,
+      "learning_rate": 0.00048713831840786195,
+      "loss": 3.6157,
+      "step": 1012
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.785926342010498,
+      "learning_rate": 0.0004871054788542428,
+      "loss": 3.5963,
+      "step": 1013
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7499585747718811,
+      "learning_rate": 0.0004870725985394431,
+      "loss": 3.4179,
+      "step": 1014
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8161090612411499,
+      "learning_rate": 0.00048703967746911535,
+      "loss": 3.6996,
+      "step": 1015
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8196116089820862,
+      "learning_rate": 0.0004870067156489191,
+      "loss": 3.6429,
+      "step": 1016
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7670063972473145,
+      "learning_rate": 0.00048697371308452077,
+      "loss": 3.6078,
+      "step": 1017
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.857550859451294,
+      "learning_rate": 0.0004869406697815939,
+      "loss": 3.6364,
+      "step": 1018
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8311168551445007,
+      "learning_rate": 0.00048690758574581905,
+      "loss": 3.3805,
+      "step": 1019
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8697710037231445,
+      "learning_rate": 0.00048687446098288374,
+      "loss": 3.497,
+      "step": 1020
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8013821244239807,
+      "learning_rate": 0.0004868412954984825,
+      "loss": 3.5215,
+      "step": 1021
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8414774537086487,
+      "learning_rate": 0.0004868080892983169,
+      "loss": 3.7113,
+      "step": 1022
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7996682524681091,
+      "learning_rate": 0.0004867748423880954,
+      "loss": 3.5724,
+      "step": 1023
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7667548060417175,
+      "learning_rate": 0.00048674155477353354,
+      "loss": 3.4878,
+      "step": 1024
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7549991011619568,
+      "learning_rate": 0.00048670822646035397,
+      "loss": 3.4915,
+      "step": 1025
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.809350311756134,
+      "learning_rate": 0.00048667485745428594,
+      "loss": 3.5255,
+      "step": 1026
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8527643084526062,
+      "learning_rate": 0.00048664144776106635,
+      "loss": 3.5635,
+      "step": 1027
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8219663500785828,
+      "learning_rate": 0.00048660799738643837,
+      "loss": 3.5104,
+      "step": 1028
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.910542905330658,
+      "learning_rate": 0.0004865745063361526,
+      "loss": 3.4771,
+      "step": 1029
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7761480212211609,
+      "learning_rate": 0.00048654097461596675,
+      "loss": 3.6563,
+      "step": 1030
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7041064500808716,
+      "learning_rate": 0.00048650740223164503,
+      "loss": 3.4496,
+      "step": 1031
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7807188630104065,
+      "learning_rate": 0.00048647378918895904,
+      "loss": 3.6611,
+      "step": 1032
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7950218319892883,
+      "learning_rate": 0.0004864401354936873,
+      "loss": 3.4923,
+      "step": 1033
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8378319144248962,
+      "learning_rate": 0.00048640644115161523,
+      "loss": 3.622,
+      "step": 1034
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7991623878479004,
+      "learning_rate": 0.00048637270616853527,
+      "loss": 3.5162,
+      "step": 1035
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8185338973999023,
+      "learning_rate": 0.0004863389305502467,
+      "loss": 3.6818,
+      "step": 1036
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8307229280471802,
+      "learning_rate": 0.0004863051143025563,
+      "loss": 3.6152,
+      "step": 1037
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7207797169685364,
+      "learning_rate": 0.0004862712574312771,
+      "loss": 3.5715,
+      "step": 1038
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8925043344497681,
+      "learning_rate": 0.00048623735994222976,
+      "loss": 3.7058,
+      "step": 1039
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8738099336624146,
+      "learning_rate": 0.00048620342184124156,
+      "loss": 3.6992,
+      "step": 1040
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7224860191345215,
+      "learning_rate": 0.0004861694431341468,
+      "loss": 3.4439,
+      "step": 1041
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8354774713516235,
+      "learning_rate": 0.0004861354238267868,
+      "loss": 3.6577,
+      "step": 1042
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9021112322807312,
+      "learning_rate": 0.00048610136392501004,
+      "loss": 3.6562,
+      "step": 1043
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9086030721664429,
+      "learning_rate": 0.0004860672634346716,
+      "loss": 3.6625,
+      "step": 1044
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7535606622695923,
+      "learning_rate": 0.0004860331223616339,
+      "loss": 3.6021,
+      "step": 1045
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8022335171699524,
+      "learning_rate": 0.0004859989407117661,
+      "loss": 3.68,
+      "step": 1046
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7648836374282837,
+      "learning_rate": 0.0004859647184909445,
+      "loss": 3.6551,
+      "step": 1047
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7472444176673889,
+      "learning_rate": 0.00048593045570505223,
+      "loss": 3.5951,
+      "step": 1048
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7668543457984924,
+      "learning_rate": 0.00048589615235997954,
+      "loss": 3.4882,
+      "step": 1049
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7949318289756775,
+      "learning_rate": 0.0004858618084616234,
+      "loss": 3.6551,
+      "step": 1050
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7682514190673828,
+      "learning_rate": 0.00048582742401588814,
+      "loss": 3.5208,
+      "step": 1051
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7590826749801636,
+      "learning_rate": 0.0004857929990286847,
+      "loss": 3.632,
+      "step": 1052
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7686331868171692,
+      "learning_rate": 0.0004857585335059312,
+      "loss": 3.5244,
+      "step": 1053
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8039591312408447,
+      "learning_rate": 0.0004857240274535526,
+      "loss": 3.5638,
+      "step": 1054
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7751696109771729,
+      "learning_rate": 0.000485689480877481,
+      "loss": 3.5261,
+      "step": 1055
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7615006566047668,
+      "learning_rate": 0.00048565489378365524,
+      "loss": 3.5094,
+      "step": 1056
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7346290946006775,
+      "learning_rate": 0.00048562026617802133,
+      "loss": 3.4812,
+      "step": 1057
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8013098239898682,
+      "learning_rate": 0.00048558559806653214,
+      "loss": 3.5405,
+      "step": 1058
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8227730393409729,
+      "learning_rate": 0.0004855508894551474,
+      "loss": 3.7083,
+      "step": 1059
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8623093962669373,
+      "learning_rate": 0.0004855161403498341,
+      "loss": 3.7428,
+      "step": 1060
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7669157981872559,
+      "learning_rate": 0.00048548135075656595,
+      "loss": 3.6555,
+      "step": 1061
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7940523028373718,
+      "learning_rate": 0.00048544652068132363,
+      "loss": 3.6781,
+      "step": 1062
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7166727781295776,
+      "learning_rate": 0.0004854116501300949,
+      "loss": 3.4161,
+      "step": 1063
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.809562623500824,
+      "learning_rate": 0.00048537673910887435,
+      "loss": 3.5386,
+      "step": 1064
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7145100235939026,
+      "learning_rate": 0.00048534178762366364,
+      "loss": 3.6118,
+      "step": 1065
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7584477066993713,
+      "learning_rate": 0.0004853067956804713,
+      "loss": 3.5688,
+      "step": 1066
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8329288959503174,
+      "learning_rate": 0.0004852717632853129,
+      "loss": 3.5135,
+      "step": 1067
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8227055668830872,
+      "learning_rate": 0.0004852366904442109,
+      "loss": 3.6212,
+      "step": 1068
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7810276746749878,
+      "learning_rate": 0.00048520157716319467,
+      "loss": 3.5009,
+      "step": 1069
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7705519199371338,
+      "learning_rate": 0.00048516642344830077,
+      "loss": 3.6416,
+      "step": 1070
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8475935459136963,
+      "learning_rate": 0.0004851312293055722,
+      "loss": 3.5342,
+      "step": 1071
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9521417617797852,
+      "learning_rate": 0.0004850959947410596,
+      "loss": 3.6362,
+      "step": 1072
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7932273745536804,
+      "learning_rate": 0.00048506071976081995,
+      "loss": 3.5728,
+      "step": 1073
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7444732189178467,
+      "learning_rate": 0.00048502540437091754,
+      "loss": 3.5696,
+      "step": 1074
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7930514812469482,
+      "learning_rate": 0.00048499004857742347,
+      "loss": 3.5353,
+      "step": 1075
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8234135508537292,
+      "learning_rate": 0.0004849546523864158,
+      "loss": 3.6051,
+      "step": 1076
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.818554162979126,
+      "learning_rate": 0.00048491921580397956,
+      "loss": 3.6196,
+      "step": 1077
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7329668998718262,
+      "learning_rate": 0.00048488373883620676,
+      "loss": 3.5506,
+      "step": 1078
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8094332218170166,
+      "learning_rate": 0.00048484822148919614,
+      "loss": 3.5721,
+      "step": 1079
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8114806413650513,
+      "learning_rate": 0.0004848126637690537,
+      "loss": 3.5818,
+      "step": 1080
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7351101040840149,
+      "learning_rate": 0.00048477706568189215,
+      "loss": 3.5156,
+      "step": 1081
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7758623957633972,
+      "learning_rate": 0.0004847414272338313,
+      "loss": 3.6086,
+      "step": 1082
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7500880360603333,
+      "learning_rate": 0.0004847057484309977,
+      "loss": 3.5158,
+      "step": 1083
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7690058350563049,
+      "learning_rate": 0.00048467002927952507,
+      "loss": 3.4918,
+      "step": 1084
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7274742126464844,
+      "learning_rate": 0.0004846342697855538,
+      "loss": 3.4803,
+      "step": 1085
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.736820638179779,
+      "learning_rate": 0.0004845984699552315,
+      "loss": 3.5871,
+      "step": 1086
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7568606734275818,
+      "learning_rate": 0.0004845626297947125,
+      "loss": 3.4236,
+      "step": 1087
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7281172871589661,
+      "learning_rate": 0.0004845267493101582,
+      "loss": 3.5822,
+      "step": 1088
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7004794478416443,
+      "learning_rate": 0.0004844908285077367,
+      "loss": 3.5296,
+      "step": 1089
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.698142409324646,
+      "learning_rate": 0.00048445486739362345,
+      "loss": 3.5203,
+      "step": 1090
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7830263376235962,
+      "learning_rate": 0.00048441886597400043,
+      "loss": 3.5399,
+      "step": 1091
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.735805094242096,
+      "learning_rate": 0.0004843828242550566,
+      "loss": 3.5375,
+      "step": 1092
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7252518534660339,
+      "learning_rate": 0.00048434674224298824,
+      "loss": 3.5794,
+      "step": 1093
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8261983394622803,
+      "learning_rate": 0.0004843106199439981,
+      "loss": 3.7589,
+      "step": 1094
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8342580199241638,
+      "learning_rate": 0.000484274457364296,
+      "loss": 3.6699,
+      "step": 1095
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7839307188987732,
+      "learning_rate": 0.0004842382545100987,
+      "loss": 3.5323,
+      "step": 1096
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8153157234191895,
+      "learning_rate": 0.00048420201138762997,
+      "loss": 3.6985,
+      "step": 1097
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7271867394447327,
+      "learning_rate": 0.0004841657280031204,
+      "loss": 3.3959,
+      "step": 1098
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7656581401824951,
+      "learning_rate": 0.0004841294043628074,
+      "loss": 3.6082,
+      "step": 1099
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8586941957473755,
+      "learning_rate": 0.00048409304047293566,
+      "loss": 3.4874,
+      "step": 1100
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7712840437889099,
+      "learning_rate": 0.0004840566363397564,
+      "loss": 3.6099,
+      "step": 1101
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7870004773139954,
+      "learning_rate": 0.00048402019196952783,
+      "loss": 3.5546,
+      "step": 1102
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7695828676223755,
+      "learning_rate": 0.0004839837073685154,
+      "loss": 3.6071,
+      "step": 1103
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7353809475898743,
+      "learning_rate": 0.00048394718254299096,
+      "loss": 3.5236,
+      "step": 1104
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6917027831077576,
+      "learning_rate": 0.0004839106174992338,
+      "loss": 3.3997,
+      "step": 1105
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7143733501434326,
+      "learning_rate": 0.0004838740122435298,
+      "loss": 3.4964,
+      "step": 1106
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7869515419006348,
+      "learning_rate": 0.00048383736678217173,
+      "loss": 3.5131,
+      "step": 1107
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7774674296379089,
+      "learning_rate": 0.00048380068112145936,
+      "loss": 3.5641,
+      "step": 1108
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7332097291946411,
+      "learning_rate": 0.0004837639552676996,
+      "loss": 3.4813,
+      "step": 1109
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.737568736076355,
+      "learning_rate": 0.00048372718922720583,
+      "loss": 3.5873,
+      "step": 1110
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8007314801216125,
+      "learning_rate": 0.00048369038300629866,
+      "loss": 3.5338,
+      "step": 1111
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8464657068252563,
+      "learning_rate": 0.0004836535366113054,
+      "loss": 3.5136,
+      "step": 1112
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7695111632347107,
+      "learning_rate": 0.0004836166500485605,
+      "loss": 3.5132,
+      "step": 1113
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7972545623779297,
+      "learning_rate": 0.0004835797233244052,
+      "loss": 3.4944,
+      "step": 1114
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8033908009529114,
+      "learning_rate": 0.0004835427564451875,
+      "loss": 3.5781,
+      "step": 1115
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7441543340682983,
+      "learning_rate": 0.0004835057494172625,
+      "loss": 3.4414,
+      "step": 1116
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8694853782653809,
+      "learning_rate": 0.00048346870224699225,
+      "loss": 3.5197,
+      "step": 1117
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.833680272102356,
+      "learning_rate": 0.0004834316149407454,
+      "loss": 3.4152,
+      "step": 1118
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9092240929603577,
+      "learning_rate": 0.0004833944875048978,
+      "loss": 3.5894,
+      "step": 1119
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7969741225242615,
+      "learning_rate": 0.000483357319945832,
+      "loss": 3.6246,
+      "step": 1120
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8331993222236633,
+      "learning_rate": 0.0004833201122699377,
+      "loss": 3.7095,
+      "step": 1121
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7968263030052185,
+      "learning_rate": 0.0004832828644836111,
+      "loss": 3.5234,
+      "step": 1122
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7297738790512085,
+      "learning_rate": 0.0004832455765932557,
+      "loss": 3.6672,
+      "step": 1123
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7510380148887634,
+      "learning_rate": 0.00048320824860528165,
+      "loss": 3.5526,
+      "step": 1124
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6990682482719421,
+      "learning_rate": 0.0004831708805261061,
+      "loss": 3.6454,
+      "step": 1125
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7424012422561646,
+      "learning_rate": 0.00048313347236215304,
+      "loss": 3.5578,
+      "step": 1126
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7626960277557373,
+      "learning_rate": 0.0004830960241198534,
+      "loss": 3.5351,
+      "step": 1127
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7487967610359192,
+      "learning_rate": 0.00048305853580564483,
+      "loss": 3.5848,
+      "step": 1128
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7351952195167542,
+      "learning_rate": 0.0004830210074259722,
+      "loss": 3.57,
+      "step": 1129
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7571800351142883,
+      "learning_rate": 0.000482983438987287,
+      "loss": 3.5361,
+      "step": 1130
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7057940363883972,
+      "learning_rate": 0.00048294583049604756,
+      "loss": 3.5936,
+      "step": 1131
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.713702917098999,
+      "learning_rate": 0.00048290818195871946,
+      "loss": 3.5384,
+      "step": 1132
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7351755499839783,
+      "learning_rate": 0.0004828704933817746,
+      "loss": 3.5273,
+      "step": 1133
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7459946274757385,
+      "learning_rate": 0.0004828327647716924,
+      "loss": 3.6704,
+      "step": 1134
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7656434178352356,
+      "learning_rate": 0.0004827949961349587,
+      "loss": 3.6132,
+      "step": 1135
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7966132164001465,
+      "learning_rate": 0.0004827571874780663,
+      "loss": 3.53,
+      "step": 1136
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7394230365753174,
+      "learning_rate": 0.0004827193388075151,
+      "loss": 3.5357,
+      "step": 1137
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7922983765602112,
+      "learning_rate": 0.00048268145012981157,
+      "loss": 3.5202,
+      "step": 1138
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8919665813446045,
+      "learning_rate": 0.0004826435214514693,
+      "loss": 3.4918,
+      "step": 1139
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7942013740539551,
+      "learning_rate": 0.0004826055527790087,
+      "loss": 3.6465,
+      "step": 1140
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8015908598899841,
+      "learning_rate": 0.00048256754411895697,
+      "loss": 3.4992,
+      "step": 1141
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8327279686927795,
+      "learning_rate": 0.0004825294954778482,
+      "loss": 3.7522,
+      "step": 1142
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7686452865600586,
+      "learning_rate": 0.0004824914068622235,
+      "loss": 3.5357,
+      "step": 1143
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8076480627059937,
+      "learning_rate": 0.0004824532782786307,
+      "loss": 3.7146,
+      "step": 1144
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7591144442558289,
+      "learning_rate": 0.00048241510973362446,
+      "loss": 3.6272,
+      "step": 1145
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.764695942401886,
+      "learning_rate": 0.00048237690123376656,
+      "loss": 3.5952,
+      "step": 1146
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.737902045249939,
+      "learning_rate": 0.0004823386527856253,
+      "loss": 3.5014,
+      "step": 1147
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7147963643074036,
+      "learning_rate": 0.00048230036439577615,
+      "loss": 3.4634,
+      "step": 1148
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7402358055114746,
+      "learning_rate": 0.0004822620360708014,
+      "loss": 3.5073,
+      "step": 1149
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8155220150947571,
+      "learning_rate": 0.00048222366781729,
+      "loss": 3.7277,
+      "step": 1150
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8458031415939331,
+      "learning_rate": 0.0004821852596418379,
+      "loss": 3.5606,
+      "step": 1151
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.730534017086029,
+      "learning_rate": 0.0004821468115510479,
+      "loss": 3.6716,
+      "step": 1152
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7740826606750488,
+      "learning_rate": 0.0004821083235515298,
+      "loss": 3.6137,
+      "step": 1153
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7634129524230957,
+      "learning_rate": 0.0004820697956499,
+      "loss": 3.6249,
+      "step": 1154
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7905582189559937,
+      "learning_rate": 0.00048203122785278195,
+      "loss": 3.5613,
+      "step": 1155
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7820989489555359,
+      "learning_rate": 0.0004819926201668059,
+      "loss": 3.5696,
+      "step": 1156
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7696579098701477,
+      "learning_rate": 0.00048195397259860894,
+      "loss": 3.5467,
+      "step": 1157
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8046624064445496,
+      "learning_rate": 0.00048191528515483504,
+      "loss": 3.6504,
+      "step": 1158
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7997883558273315,
+      "learning_rate": 0.0004818765578421351,
+      "loss": 3.5677,
+      "step": 1159
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7544205784797668,
+      "learning_rate": 0.0004818377906671667,
+      "loss": 3.6235,
+      "step": 1160
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7897548079490662,
+      "learning_rate": 0.00048179898363659435,
+      "loss": 3.6026,
+      "step": 1161
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7426437735557556,
+      "learning_rate": 0.0004817601367570895,
+      "loss": 3.5164,
+      "step": 1162
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7978121638298035,
+      "learning_rate": 0.00048172125003533034,
+      "loss": 3.6887,
+      "step": 1163
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7501871585845947,
+      "learning_rate": 0.00048168232347800197,
+      "loss": 3.4855,
+      "step": 1164
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7330640554428101,
+      "learning_rate": 0.0004816433570917963,
+      "loss": 3.4732,
+      "step": 1165
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7803331017494202,
+      "learning_rate": 0.000481604350883412,
+      "loss": 3.5258,
+      "step": 1166
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7541994452476501,
+      "learning_rate": 0.00048156530485955496,
+      "loss": 3.4746,
+      "step": 1167
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.779971182346344,
+      "learning_rate": 0.0004815262190269374,
+      "loss": 3.4933,
+      "step": 1168
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7977837920188904,
+      "learning_rate": 0.00048148709339227867,
+      "loss": 3.5619,
+      "step": 1169
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8055017590522766,
+      "learning_rate": 0.000481447927962305,
+      "loss": 3.4803,
+      "step": 1170
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.854934811592102,
+      "learning_rate": 0.00048140872274374935,
+      "loss": 3.5664,
+      "step": 1171
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7964548468589783,
+      "learning_rate": 0.00048136947774335154,
+      "loss": 3.6014,
+      "step": 1172
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.76549232006073,
+      "learning_rate": 0.00048133019296785825,
+      "loss": 3.6647,
+      "step": 1173
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7582719922065735,
+      "learning_rate": 0.0004812908684240229,
+      "loss": 3.4162,
+      "step": 1174
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7677952647209167,
+      "learning_rate": 0.000481251504118606,
+      "loss": 3.6258,
+      "step": 1175
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.820792019367218,
+      "learning_rate": 0.00048121210005837463,
+      "loss": 3.6514,
+      "step": 1176
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8015343546867371,
+      "learning_rate": 0.0004811726562501028,
+      "loss": 3.499,
+      "step": 1177
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0623503923416138,
+      "learning_rate": 0.00048113317270057136,
+      "loss": 3.6353,
+      "step": 1178
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7586822509765625,
+      "learning_rate": 0.0004810936494165681,
+      "loss": 3.5914,
+      "step": 1179
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.744174063205719,
+      "learning_rate": 0.00048105408640488733,
+      "loss": 3.4081,
+      "step": 1180
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7595331072807312,
+      "learning_rate": 0.0004810144836723305,
+      "loss": 3.5483,
+      "step": 1181
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7022777199745178,
+      "learning_rate": 0.00048097484122570585,
+      "loss": 3.4242,
+      "step": 1182
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7804768085479736,
+      "learning_rate": 0.0004809351590718283,
+      "loss": 3.5572,
+      "step": 1183
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7230439782142639,
+      "learning_rate": 0.00048089543721751963,
+      "loss": 3.538,
+      "step": 1184
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8133188486099243,
+      "learning_rate": 0.0004808556756696085,
+      "loss": 3.7011,
+      "step": 1185
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7499143481254578,
+      "learning_rate": 0.00048081587443493044,
+      "loss": 3.5564,
+      "step": 1186
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7320449948310852,
+      "learning_rate": 0.0004807760335203277,
+      "loss": 3.5104,
+      "step": 1187
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6878542304039001,
+      "learning_rate": 0.0004807361529326495,
+      "loss": 3.4946,
+      "step": 1188
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7565658688545227,
+      "learning_rate": 0.0004806962326787516,
+      "loss": 3.4685,
+      "step": 1189
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.302788496017456,
+      "learning_rate": 0.00048065627276549697,
+      "loss": 3.4976,
+      "step": 1190
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8028513789176941,
+      "learning_rate": 0.00048061627319975493,
+      "loss": 3.5922,
+      "step": 1191
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7514439225196838,
+      "learning_rate": 0.000480576233988402,
+      "loss": 3.5921,
+      "step": 1192
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8432983160018921,
+      "learning_rate": 0.0004805361551383214,
+      "loss": 3.5551,
+      "step": 1193
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7517931461334229,
+      "learning_rate": 0.0004804960366564032,
+      "loss": 3.5159,
+      "step": 1194
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7453200817108154,
+      "learning_rate": 0.000480455878549544,
+      "loss": 3.4694,
+      "step": 1195
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7308317422866821,
+      "learning_rate": 0.00048041568082464773,
+      "loss": 3.4796,
+      "step": 1196
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7596452236175537,
+      "learning_rate": 0.00048037544348862474,
+      "loss": 3.4077,
+      "step": 1197
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.703974187374115,
+      "learning_rate": 0.00048033516654839213,
+      "loss": 3.607,
+      "step": 1198
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.666873574256897,
+      "learning_rate": 0.0004802948500108743,
+      "loss": 3.5187,
+      "step": 1199
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7371575832366943,
+      "learning_rate": 0.0004802544938830018,
+      "loss": 3.6909,
+      "step": 1200
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7255220413208008,
+      "learning_rate": 0.0004802140981717125,
+      "loss": 3.6693,
+      "step": 1201
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.757788896560669,
+      "learning_rate": 0.0004801736628839509,
+      "loss": 3.5358,
+      "step": 1202
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7462393045425415,
+      "learning_rate": 0.0004801331880266682,
+      "loss": 3.5705,
+      "step": 1203
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.827492892742157,
+      "learning_rate": 0.0004800926736068225,
+      "loss": 3.6334,
+      "step": 1204
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.803577721118927,
+      "learning_rate": 0.0004800521196313788,
+      "loss": 3.4963,
+      "step": 1205
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8225682377815247,
+      "learning_rate": 0.0004800115261073088,
+      "loss": 3.4747,
+      "step": 1206
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.806121826171875,
+      "learning_rate": 0.00047997089304159085,
+      "loss": 3.544,
+      "step": 1207
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8418048024177551,
+      "learning_rate": 0.00047993022044121036,
+      "loss": 3.6103,
+      "step": 1208
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7457411289215088,
+      "learning_rate": 0.00047988950831315947,
+      "loss": 3.5944,
+      "step": 1209
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7606494426727295,
+      "learning_rate": 0.0004798487566644369,
+      "loss": 3.6472,
+      "step": 1210
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7392157912254333,
+      "learning_rate": 0.00047980796550204844,
+      "loss": 3.4584,
+      "step": 1211
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7845761179924011,
+      "learning_rate": 0.0004797671348330066,
+      "loss": 3.6173,
+      "step": 1212
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7397451996803284,
+      "learning_rate": 0.0004797262646643305,
+      "loss": 3.5771,
+      "step": 1213
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7049579620361328,
+      "learning_rate": 0.0004796853550030463,
+      "loss": 3.4425,
+      "step": 1214
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7008839249610901,
+      "learning_rate": 0.000479644405856187,
+      "loss": 3.5106,
+      "step": 1215
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7417147755622864,
+      "learning_rate": 0.00047960341723079185,
+      "loss": 3.4835,
+      "step": 1216
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8001255989074707,
+      "learning_rate": 0.0004795623891339076,
+      "loss": 3.5156,
+      "step": 1217
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7629938125610352,
+      "learning_rate": 0.0004795213215725873,
+      "loss": 3.5797,
+      "step": 1218
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7776039242744446,
+      "learning_rate": 0.000479480214553891,
+      "loss": 3.6313,
+      "step": 1219
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7509599924087524,
+      "learning_rate": 0.00047943906808488546,
+      "loss": 3.4204,
+      "step": 1220
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.837563157081604,
+      "learning_rate": 0.00047939788217264424,
+      "loss": 3.5478,
+      "step": 1221
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7498551607131958,
+      "learning_rate": 0.00047935665682424767,
+      "loss": 3.5485,
+      "step": 1222
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7325605154037476,
+      "learning_rate": 0.00047931539204678286,
+      "loss": 3.4524,
+      "step": 1223
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.691773533821106,
+      "learning_rate": 0.0004792740878473437,
+      "loss": 3.5594,
+      "step": 1224
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7024340033531189,
+      "learning_rate": 0.0004792327442330309,
+      "loss": 3.4423,
+      "step": 1225
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7005559802055359,
+      "learning_rate": 0.00047919136121095173,
+      "loss": 3.4901,
+      "step": 1226
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7247700691223145,
+      "learning_rate": 0.00047914993878822067,
+      "loss": 3.6014,
+      "step": 1227
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6847344636917114,
+      "learning_rate": 0.0004791084769719586,
+      "loss": 3.4629,
+      "step": 1228
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7518942356109619,
+      "learning_rate": 0.00047906697576929327,
+      "loss": 3.6171,
+      "step": 1229
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7332186102867126,
+      "learning_rate": 0.0004790254351873592,
+      "loss": 3.5396,
+      "step": 1230
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7920958995819092,
+      "learning_rate": 0.0004789838552332978,
+      "loss": 3.5926,
+      "step": 1231
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7580140829086304,
+      "learning_rate": 0.000478942235914257,
+      "loss": 3.4796,
+      "step": 1232
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.728755533695221,
+      "learning_rate": 0.00047890057723739177,
+      "loss": 3.5943,
+      "step": 1233
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7519229054450989,
+      "learning_rate": 0.0004788588792098637,
+      "loss": 3.5283,
+      "step": 1234
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7821226119995117,
+      "learning_rate": 0.00047881714183884105,
+      "loss": 3.6506,
+      "step": 1235
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8124991059303284,
+      "learning_rate": 0.0004787753651314991,
+      "loss": 3.6492,
+      "step": 1236
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7577442526817322,
+      "learning_rate": 0.00047873354909501963,
+      "loss": 3.4795,
+      "step": 1237
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7887035012245178,
+      "learning_rate": 0.00047869169373659145,
+      "loss": 3.4658,
+      "step": 1238
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7463725805282593,
+      "learning_rate": 0.0004786497990634099,
+      "loss": 3.5866,
+      "step": 1239
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8219547867774963,
+      "learning_rate": 0.0004786078650826771,
+      "loss": 3.6231,
+      "step": 1240
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6816055774688721,
+      "learning_rate": 0.00047856589180160204,
+      "loss": 3.5007,
+      "step": 1241
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7251628637313843,
+      "learning_rate": 0.00047852387922740037,
+      "loss": 3.4619,
+      "step": 1242
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7888408303260803,
+      "learning_rate": 0.0004784818273672947,
+      "loss": 3.5239,
+      "step": 1243
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7807585000991821,
+      "learning_rate": 0.00047843973622851407,
+      "loss": 3.5853,
+      "step": 1244
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7896850109100342,
+      "learning_rate": 0.00047839760581829444,
+      "loss": 3.5603,
+      "step": 1245
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.706608235836029,
+      "learning_rate": 0.0004783554361438786,
+      "loss": 3.6437,
+      "step": 1246
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7114312052726746,
+      "learning_rate": 0.0004783132272125159,
+      "loss": 3.5691,
+      "step": 1247
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6859767436981201,
+      "learning_rate": 0.00047827097903146266,
+      "loss": 3.446,
+      "step": 1248
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8105244636535645,
+      "learning_rate": 0.00047822869160798185,
+      "loss": 3.5147,
+      "step": 1249
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.761033296585083,
+      "learning_rate": 0.00047818636494934295,
+      "loss": 3.5753,
+      "step": 1250
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7481651306152344,
+      "learning_rate": 0.0004781439990628227,
+      "loss": 3.6124,
+      "step": 1251
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7689448595046997,
+      "learning_rate": 0.00047810159395570405,
+      "loss": 3.5605,
+      "step": 1252
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.758472740650177,
+      "learning_rate": 0.00047805914963527707,
+      "loss": 3.518,
+      "step": 1253
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7787196040153503,
+      "learning_rate": 0.00047801666610883833,
+      "loss": 3.3512,
+      "step": 1254
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8059257864952087,
+      "learning_rate": 0.00047797414338369125,
+      "loss": 3.5148,
+      "step": 1255
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8021636009216309,
+      "learning_rate": 0.0004779315814671461,
+      "loss": 3.6353,
+      "step": 1256
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.817902684211731,
+      "learning_rate": 0.00047788898036651967,
+      "loss": 3.6066,
+      "step": 1257
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8657376170158386,
+      "learning_rate": 0.00047784634008913565,
+      "loss": 3.636,
+      "step": 1258
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7575492262840271,
+      "learning_rate": 0.00047780366064232427,
+      "loss": 3.4298,
+      "step": 1259
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7412144541740417,
+      "learning_rate": 0.0004777609420334227,
+      "loss": 3.5585,
+      "step": 1260
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7263116836547852,
+      "learning_rate": 0.00047771818426977475,
+      "loss": 3.5636,
+      "step": 1261
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7642979621887207,
+      "learning_rate": 0.0004776753873587311,
+      "loss": 3.6134,
+      "step": 1262
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7336152195930481,
+      "learning_rate": 0.0004776325513076488,
+      "loss": 3.5417,
+      "step": 1263
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7328927516937256,
+      "learning_rate": 0.00047758967612389206,
+      "loss": 3.4612,
+      "step": 1264
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7464151382446289,
+      "learning_rate": 0.00047754676181483146,
+      "loss": 3.6533,
+      "step": 1265
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7545056343078613,
+      "learning_rate": 0.0004775038083878446,
+      "loss": 3.5238,
+      "step": 1266
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7579200863838196,
+      "learning_rate": 0.00047746081585031566,
+      "loss": 3.4982,
+      "step": 1267
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7628843188285828,
+      "learning_rate": 0.00047741778420963554,
+      "loss": 3.5309,
+      "step": 1268
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7624720931053162,
+      "learning_rate": 0.0004773747134732018,
+      "loss": 3.689,
+      "step": 1269
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6972590684890747,
+      "learning_rate": 0.0004773316036484189,
+      "loss": 3.5652,
+      "step": 1270
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7528889775276184,
+      "learning_rate": 0.0004772884547426979,
+      "loss": 3.599,
+      "step": 1271
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6812743544578552,
+      "learning_rate": 0.0004772452667634565,
+      "loss": 3.4173,
+      "step": 1272
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7377366423606873,
+      "learning_rate": 0.0004772020397181194,
+      "loss": 3.5766,
+      "step": 1273
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7267171740531921,
+      "learning_rate": 0.0004771587736141176,
+      "loss": 3.5296,
+      "step": 1274
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7521837949752808,
+      "learning_rate": 0.0004771154684588892,
+      "loss": 3.6125,
+      "step": 1275
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6818501949310303,
+      "learning_rate": 0.00047707212425987885,
+      "loss": 3.5312,
+      "step": 1276
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7633997797966003,
+      "learning_rate": 0.00047702874102453785,
+      "loss": 3.4141,
+      "step": 1277
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7707627415657043,
+      "learning_rate": 0.00047698531876032437,
+      "loss": 3.5135,
+      "step": 1278
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7294227480888367,
+      "learning_rate": 0.0004769418574747032,
+      "loss": 3.6082,
+      "step": 1279
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7334704995155334,
+      "learning_rate": 0.00047689835717514576,
+      "loss": 3.4137,
+      "step": 1280
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7663036584854126,
+      "learning_rate": 0.0004768548178691302,
+      "loss": 3.605,
+      "step": 1281
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7646681070327759,
+      "learning_rate": 0.00047681123956414164,
+      "loss": 3.571,
+      "step": 1282
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7949337959289551,
+      "learning_rate": 0.0004767676222676715,
+      "loss": 3.7506,
+      "step": 1283
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.725398063659668,
+      "learning_rate": 0.00047672396598721824,
+      "loss": 3.5381,
+      "step": 1284
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7394816279411316,
+      "learning_rate": 0.00047668027073028676,
+      "loss": 3.555,
+      "step": 1285
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7550774216651917,
+      "learning_rate": 0.0004766365365043889,
+      "loss": 3.4987,
+      "step": 1286
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.736352801322937,
+      "learning_rate": 0.00047659276331704295,
+      "loss": 3.3987,
+      "step": 1287
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7474049925804138,
+      "learning_rate": 0.00047654895117577414,
+      "loss": 3.4919,
+      "step": 1288
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8658040761947632,
+      "learning_rate": 0.0004765051000881142,
+      "loss": 3.7013,
+      "step": 1289
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7833415269851685,
+      "learning_rate": 0.0004764612100616017,
+      "loss": 3.6412,
+      "step": 1290
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7740939855575562,
+      "learning_rate": 0.0004764172811037818,
+      "loss": 3.5839,
+      "step": 1291
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8426244258880615,
+      "learning_rate": 0.0004763733132222065,
+      "loss": 3.3914,
+      "step": 1292
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.733086884021759,
+      "learning_rate": 0.0004763293064244343,
+      "loss": 3.6397,
+      "step": 1293
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8235949873924255,
+      "learning_rate": 0.00047628526071803046,
+      "loss": 3.5314,
+      "step": 1294
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7923353314399719,
+      "learning_rate": 0.00047624117611056696,
+      "loss": 3.4461,
+      "step": 1295
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7275211811065674,
+      "learning_rate": 0.0004761970526096225,
+      "loss": 3.6302,
+      "step": 1296
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7095808386802673,
+      "learning_rate": 0.0004761528902227824,
+      "loss": 3.5677,
+      "step": 1297
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7407586574554443,
+      "learning_rate": 0.00047610868895763865,
+      "loss": 3.4968,
+      "step": 1298
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7782431244850159,
+      "learning_rate": 0.00047606444882179,
+      "loss": 3.717,
+      "step": 1299
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6857940554618835,
+      "learning_rate": 0.0004760201698228419,
+      "loss": 3.4115,
+      "step": 1300
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8255395293235779,
+      "learning_rate": 0.0004759758519684063,
+      "loss": 3.6371,
+      "step": 1301
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7767965197563171,
+      "learning_rate": 0.000475931495266102,
+      "loss": 3.6189,
+      "step": 1302
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7226409316062927,
+      "learning_rate": 0.0004758870997235545,
+      "loss": 3.5922,
+      "step": 1303
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7387113571166992,
+      "learning_rate": 0.00047584266534839577,
+      "loss": 3.5281,
+      "step": 1304
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7528977394104004,
+      "learning_rate": 0.0004757981921482647,
+      "loss": 3.5226,
+      "step": 1305
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8862900137901306,
+      "learning_rate": 0.00047575368013080675,
+      "loss": 3.4604,
+      "step": 1306
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8055890202522278,
+      "learning_rate": 0.000475709129303674,
+      "loss": 3.5837,
+      "step": 1307
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7440634965896606,
+      "learning_rate": 0.0004756645396745253,
+      "loss": 3.5179,
+      "step": 1308
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7975066304206848,
+      "learning_rate": 0.0004756199112510261,
+      "loss": 3.4997,
+      "step": 1309
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7775300145149231,
+      "learning_rate": 0.0004755752440408485,
+      "loss": 3.5861,
+      "step": 1310
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7195447683334351,
+      "learning_rate": 0.00047553053805167136,
+      "loss": 3.4673,
+      "step": 1311
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7289524078369141,
+      "learning_rate": 0.0004754857932911802,
+      "loss": 3.4836,
+      "step": 1312
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7409530282020569,
+      "learning_rate": 0.0004754410097670671,
+      "loss": 3.4789,
+      "step": 1313
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7885916829109192,
+      "learning_rate": 0.0004753961874870309,
+      "loss": 3.4292,
+      "step": 1314
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7096027731895447,
+      "learning_rate": 0.000475351326458777,
+      "loss": 3.4578,
+      "step": 1315
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.802874743938446,
+      "learning_rate": 0.0004753064266900177,
+      "loss": 3.5491,
+      "step": 1316
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7888942360877991,
+      "learning_rate": 0.00047526148818847157,
+      "loss": 3.451,
+      "step": 1317
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7798328995704651,
+      "learning_rate": 0.00047521651096186417,
+      "loss": 3.6149,
+      "step": 1318
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7312204241752625,
+      "learning_rate": 0.00047517149501792767,
+      "loss": 3.6303,
+      "step": 1319
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7797731757164001,
+      "learning_rate": 0.0004751264403644007,
+      "loss": 3.4529,
+      "step": 1320
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7609673142433167,
+      "learning_rate": 0.00047508134700902874,
+      "loss": 3.5962,
+      "step": 1321
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6947163939476013,
+      "learning_rate": 0.0004750362149595639,
+      "loss": 3.4685,
+      "step": 1322
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7074885964393616,
+      "learning_rate": 0.0004749910442237648,
+      "loss": 3.4133,
+      "step": 1323
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7470065355300903,
+      "learning_rate": 0.000474945834809397,
+      "loss": 3.5763,
+      "step": 1324
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.752415120601654,
+      "learning_rate": 0.0004749005867242323,
+      "loss": 3.5299,
+      "step": 1325
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7444427609443665,
+      "learning_rate": 0.00047485529997604947,
+      "loss": 3.5643,
+      "step": 1326
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7178149819374084,
+      "learning_rate": 0.00047480997457263394,
+      "loss": 3.5211,
+      "step": 1327
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7430973052978516,
+      "learning_rate": 0.0004747646105217774,
+      "loss": 3.5017,
+      "step": 1328
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7771696448326111,
+      "learning_rate": 0.00047471920783127874,
+      "loss": 3.5547,
+      "step": 1329
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8592125773429871,
+      "learning_rate": 0.000474673766508943,
+      "loss": 3.6266,
+      "step": 1330
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7617010474205017,
+      "learning_rate": 0.0004746282865625822,
+      "loss": 3.7154,
+      "step": 1331
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7075968980789185,
+      "learning_rate": 0.00047458276800001485,
+      "loss": 3.5496,
+      "step": 1332
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8044525384902954,
+      "learning_rate": 0.0004745372108290661,
+      "loss": 3.5343,
+      "step": 1333
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7208424806594849,
+      "learning_rate": 0.00047449161505756766,
+      "loss": 3.6053,
+      "step": 1334
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6809619069099426,
+      "learning_rate": 0.0004744459806933581,
+      "loss": 3.4405,
+      "step": 1335
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7048726081848145,
+      "learning_rate": 0.0004744003077442825,
+      "loss": 3.541,
+      "step": 1336
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7939316630363464,
+      "learning_rate": 0.00047435459621819257,
+      "loss": 3.6154,
+      "step": 1337
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7555193305015564,
+      "learning_rate": 0.00047430884612294645,
+      "loss": 3.4798,
+      "step": 1338
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7547550201416016,
+      "learning_rate": 0.0004742630574664094,
+      "loss": 3.4525,
+      "step": 1339
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7709076404571533,
+      "learning_rate": 0.0004742172302564528,
+      "loss": 3.4592,
+      "step": 1340
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7275447845458984,
+      "learning_rate": 0.00047417136450095504,
+      "loss": 3.6308,
+      "step": 1341
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.773053765296936,
+      "learning_rate": 0.0004741254602078009,
+      "loss": 3.5169,
+      "step": 1342
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7104327082633972,
+      "learning_rate": 0.00047407951738488187,
+      "loss": 3.5841,
+      "step": 1343
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7337343692779541,
+      "learning_rate": 0.00047403353604009605,
+      "loss": 3.4557,
+      "step": 1344
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8200117945671082,
+      "learning_rate": 0.0004739875161813481,
+      "loss": 3.6261,
+      "step": 1345
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7903836369514465,
+      "learning_rate": 0.0004739414578165495,
+      "loss": 3.5732,
+      "step": 1346
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8006786108016968,
+      "learning_rate": 0.00047389536095361807,
+      "loss": 3.708,
+      "step": 1347
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7467408776283264,
+      "learning_rate": 0.00047384922560047855,
+      "loss": 3.5375,
+      "step": 1348
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7394277453422546,
+      "learning_rate": 0.00047380305176506203,
+      "loss": 3.4562,
+      "step": 1349
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7740619778633118,
+      "learning_rate": 0.0004737568394553064,
+      "loss": 3.6413,
+      "step": 1350
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7343472242355347,
+      "learning_rate": 0.00047371058867915606,
+      "loss": 3.4088,
+      "step": 1351
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7520578503608704,
+      "learning_rate": 0.000473664299444562,
+      "loss": 3.6681,
+      "step": 1352
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8136751651763916,
+      "learning_rate": 0.000473617971759482,
+      "loss": 3.5463,
+      "step": 1353
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6983349323272705,
+      "learning_rate": 0.0004735716056318802,
+      "loss": 3.4922,
+      "step": 1354
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7741418480873108,
+      "learning_rate": 0.0004735252010697275,
+      "loss": 3.6106,
+      "step": 1355
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.692803144454956,
+      "learning_rate": 0.00047347875808100145,
+      "loss": 3.4923,
+      "step": 1356
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7374471426010132,
+      "learning_rate": 0.00047343227667368605,
+      "loss": 3.5939,
+      "step": 1357
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7494756579399109,
+      "learning_rate": 0.0004733857568557721,
+      "loss": 3.5447,
+      "step": 1358
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7510193586349487,
+      "learning_rate": 0.0004733391986352568,
+      "loss": 3.5397,
+      "step": 1359
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7476963996887207,
+      "learning_rate": 0.0004732926020201441,
+      "loss": 3.5822,
+      "step": 1360
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7323873043060303,
+      "learning_rate": 0.0004732459670184445,
+      "loss": 3.5792,
+      "step": 1361
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7584927678108215,
+      "learning_rate": 0.000473199293638175,
+      "loss": 3.5797,
+      "step": 1362
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7634500861167908,
+      "learning_rate": 0.00047315258188735954,
+      "loss": 3.4843,
+      "step": 1363
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.750939667224884,
+      "learning_rate": 0.0004731058317740281,
+      "loss": 3.3346,
+      "step": 1364
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7266926765441895,
+      "learning_rate": 0.0004730590433062178,
+      "loss": 3.416,
+      "step": 1365
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.722404420375824,
+      "learning_rate": 0.000473012216491972,
+      "loss": 3.6912,
+      "step": 1366
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6876066327095032,
+      "learning_rate": 0.0004729653513393408,
+      "loss": 3.5736,
+      "step": 1367
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7364122271537781,
+      "learning_rate": 0.00047291844785638085,
+      "loss": 3.6518,
+      "step": 1368
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7460653781890869,
+      "learning_rate": 0.0004728715060511555,
+      "loss": 3.5476,
+      "step": 1369
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6969454288482666,
+      "learning_rate": 0.0004728245259317344,
+      "loss": 3.4781,
+      "step": 1370
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7020079493522644,
+      "learning_rate": 0.00047277750750619415,
+      "loss": 3.5167,
+      "step": 1371
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7419220805168152,
+      "learning_rate": 0.00047273045078261765,
+      "loss": 3.5804,
+      "step": 1372
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7115669846534729,
+      "learning_rate": 0.0004726833557690946,
+      "loss": 3.5391,
+      "step": 1373
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7343298196792603,
+      "learning_rate": 0.0004726362224737211,
+      "loss": 3.5557,
+      "step": 1374
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7240042090415955,
+      "learning_rate": 0.00047258905090459996,
+      "loss": 3.5377,
+      "step": 1375
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7479480504989624,
+      "learning_rate": 0.0004725418410698405,
+      "loss": 3.4328,
+      "step": 1376
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7856737971305847,
+      "learning_rate": 0.0004724945929775586,
+      "loss": 3.5831,
+      "step": 1377
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7368362545967102,
+      "learning_rate": 0.0004724473066358768,
+      "loss": 3.4749,
+      "step": 1378
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7227674126625061,
+      "learning_rate": 0.00047239998205292425,
+      "loss": 3.5272,
+      "step": 1379
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7797960042953491,
+      "learning_rate": 0.0004723526192368364,
+      "loss": 3.4572,
+      "step": 1380
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7110636234283447,
+      "learning_rate": 0.0004723052181957556,
+      "loss": 3.5097,
+      "step": 1381
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7274758219718933,
+      "learning_rate": 0.00047225777893783054,
+      "loss": 3.5489,
+      "step": 1382
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7741771340370178,
+      "learning_rate": 0.0004722103014712167,
+      "loss": 3.4818,
+      "step": 1383
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7286615371704102,
+      "learning_rate": 0.00047216278580407603,
+      "loss": 3.517,
+      "step": 1384
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7044836282730103,
+      "learning_rate": 0.00047211523194457683,
+      "loss": 3.5782,
+      "step": 1385
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7229937314987183,
+      "learning_rate": 0.0004720676399008943,
+      "loss": 3.5155,
+      "step": 1386
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7034634351730347,
+      "learning_rate": 0.00047202000968121007,
+      "loss": 3.5449,
+      "step": 1387
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6847106218338013,
+      "learning_rate": 0.00047197234129371234,
+      "loss": 3.5222,
+      "step": 1388
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7309967279434204,
+      "learning_rate": 0.00047192463474659576,
+      "loss": 3.5243,
+      "step": 1389
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7710961103439331,
+      "learning_rate": 0.0004718768900480617,
+      "loss": 3.5449,
+      "step": 1390
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7054787278175354,
+      "learning_rate": 0.00047182910720631804,
+      "loss": 3.5429,
+      "step": 1391
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.732120931148529,
+      "learning_rate": 0.00047178128622957916,
+      "loss": 3.5658,
+      "step": 1392
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7226571440696716,
+      "learning_rate": 0.00047173342712606613,
+      "loss": 3.5384,
+      "step": 1393
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.753541111946106,
+      "learning_rate": 0.00047168552990400637,
+      "loss": 3.5268,
+      "step": 1394
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7406083941459656,
+      "learning_rate": 0.000471637594571634,
+      "loss": 3.5432,
+      "step": 1395
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7489895224571228,
+      "learning_rate": 0.0004715896211371897,
+      "loss": 3.3925,
+      "step": 1396
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.742703378200531,
+      "learning_rate": 0.00047154160960892065,
+      "loss": 3.4482,
+      "step": 1397
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7691521048545837,
+      "learning_rate": 0.00047149355999508066,
+      "loss": 3.6035,
+      "step": 1398
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.847844660282135,
+      "learning_rate": 0.0004714454723039299,
+      "loss": 3.6594,
+      "step": 1399
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6853086948394775,
+      "learning_rate": 0.00047139734654373514,
+      "loss": 3.4047,
+      "step": 1400
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6940163373947144,
+      "learning_rate": 0.0004713491827227699,
+      "loss": 3.552,
+      "step": 1401
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7334827780723572,
+      "learning_rate": 0.00047130098084931406,
+      "loss": 3.5911,
+      "step": 1402
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7077183723449707,
+      "learning_rate": 0.00047125274093165405,
+      "loss": 3.6431,
+      "step": 1403
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7522363066673279,
+      "learning_rate": 0.00047120446297808287,
+      "loss": 3.4579,
+      "step": 1404
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6966984868049622,
+      "learning_rate": 0.00047115614699690014,
+      "loss": 3.5049,
+      "step": 1405
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7402704358100891,
+      "learning_rate": 0.0004711077929964118,
+      "loss": 3.606,
+      "step": 1406
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7326617240905762,
+      "learning_rate": 0.0004710594009849306,
+      "loss": 3.5428,
+      "step": 1407
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7809644937515259,
+      "learning_rate": 0.0004710109709707757,
+      "loss": 3.6287,
+      "step": 1408
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7565213441848755,
+      "learning_rate": 0.0004709625029622726,
+      "loss": 3.3696,
+      "step": 1409
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7164292931556702,
+      "learning_rate": 0.0004709139969677537,
+      "loss": 3.3233,
+      "step": 1410
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7206988334655762,
+      "learning_rate": 0.0004708654529955576,
+      "loss": 3.6306,
+      "step": 1411
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.713064968585968,
+      "learning_rate": 0.00047081687105402967,
+      "loss": 3.4532,
+      "step": 1412
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6608602404594421,
+      "learning_rate": 0.00047076825115152166,
+      "loss": 3.5389,
+      "step": 1413
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7010859251022339,
+      "learning_rate": 0.0004707195932963919,
+      "loss": 3.4688,
+      "step": 1414
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.700348973274231,
+      "learning_rate": 0.00047067089749700534,
+      "loss": 3.4571,
+      "step": 1415
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6964651346206665,
+      "learning_rate": 0.00047062216376173315,
+      "loss": 3.5361,
+      "step": 1416
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7558305859565735,
+      "learning_rate": 0.0004705733920989534,
+      "loss": 3.4434,
+      "step": 1417
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.784369945526123,
+      "learning_rate": 0.00047052458251705043,
+      "loss": 3.6063,
+      "step": 1418
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7479367256164551,
+      "learning_rate": 0.0004704757350244152,
+      "loss": 3.5864,
+      "step": 1419
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7332322001457214,
+      "learning_rate": 0.0004704268496294451,
+      "loss": 3.5199,
+      "step": 1420
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7303996086120605,
+      "learning_rate": 0.00047037792634054416,
+      "loss": 3.5091,
+      "step": 1421
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7379922270774841,
+      "learning_rate": 0.0004703289651661229,
+      "loss": 3.5665,
+      "step": 1422
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7322970628738403,
+      "learning_rate": 0.0004702799661145981,
+      "loss": 3.5029,
+      "step": 1423
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7706109285354614,
+      "learning_rate": 0.0004702309291943936,
+      "loss": 3.632,
+      "step": 1424
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6745412349700928,
+      "learning_rate": 0.00047018185441393914,
+      "loss": 3.5001,
+      "step": 1425
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7590019106864929,
+      "learning_rate": 0.00047013274178167136,
+      "loss": 3.6288,
+      "step": 1426
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.681084156036377,
+      "learning_rate": 0.00047008359130603326,
+      "loss": 3.6016,
+      "step": 1427
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7189148664474487,
+      "learning_rate": 0.00047003440299547437,
+      "loss": 3.7531,
+      "step": 1428
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6640321016311646,
+      "learning_rate": 0.0004699851768584508,
+      "loss": 3.4807,
+      "step": 1429
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7376559376716614,
+      "learning_rate": 0.0004699359129034251,
+      "loss": 3.483,
+      "step": 1430
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6923270225524902,
+      "learning_rate": 0.0004698866111388661,
+      "loss": 3.5111,
+      "step": 1431
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6715227961540222,
+      "learning_rate": 0.00046983727157324964,
+      "loss": 3.5518,
+      "step": 1432
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6784403920173645,
+      "learning_rate": 0.0004697878942150575,
+      "loss": 3.4985,
+      "step": 1433
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7522211670875549,
+      "learning_rate": 0.00046973847907277844,
+      "loss": 3.4804,
+      "step": 1434
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8737341165542603,
+      "learning_rate": 0.0004696890261549073,
+      "loss": 3.4973,
+      "step": 1435
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7290940880775452,
+      "learning_rate": 0.00046963953546994583,
+      "loss": 3.4597,
+      "step": 1436
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3438891172409058,
+      "learning_rate": 0.00046959000702640185,
+      "loss": 3.5925,
+      "step": 1437
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8091414570808411,
+      "learning_rate": 0.00046954044083279004,
+      "loss": 3.4549,
+      "step": 1438
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8499096035957336,
+      "learning_rate": 0.00046949083689763114,
+      "loss": 3.6355,
+      "step": 1439
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7301666736602783,
+      "learning_rate": 0.00046944119522945307,
+      "loss": 3.4349,
+      "step": 1440
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7843241095542908,
+      "learning_rate": 0.0004693915158367894,
+      "loss": 3.4676,
+      "step": 1441
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7359977960586548,
+      "learning_rate": 0.00046934179872818073,
+      "loss": 3.5503,
+      "step": 1442
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7981695532798767,
+      "learning_rate": 0.00046929204391217414,
+      "loss": 3.4857,
+      "step": 1443
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8104868531227112,
+      "learning_rate": 0.0004692422513973229,
+      "loss": 3.5549,
+      "step": 1444
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7411574721336365,
+      "learning_rate": 0.000469192421192187,
+      "loss": 3.3795,
+      "step": 1445
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.715009868144989,
+      "learning_rate": 0.00046914255330533273,
+      "loss": 3.613,
+      "step": 1446
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7242358326911926,
+      "learning_rate": 0.00046909264774533307,
+      "loss": 3.537,
+      "step": 1447
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8142811059951782,
+      "learning_rate": 0.0004690427045207673,
+      "loss": 3.4606,
+      "step": 1448
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7344624996185303,
+      "learning_rate": 0.00046899272364022126,
+      "loss": 3.433,
+      "step": 1449
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7056677937507629,
+      "learning_rate": 0.0004689427051122873,
+      "loss": 3.5493,
+      "step": 1450
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6946106553077698,
+      "learning_rate": 0.00046889264894556406,
+      "loss": 3.5235,
+      "step": 1451
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7007893323898315,
+      "learning_rate": 0.00046884255514865694,
+      "loss": 3.4545,
+      "step": 1452
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6814204454421997,
+      "learning_rate": 0.00046879242373017746,
+      "loss": 3.5959,
+      "step": 1453
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7359189987182617,
+      "learning_rate": 0.0004687422546987439,
+      "loss": 3.7141,
+      "step": 1454
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7151730060577393,
+      "learning_rate": 0.00046869204806298094,
+      "loss": 3.5196,
+      "step": 1455
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7060539126396179,
+      "learning_rate": 0.0004686418038315196,
+      "loss": 3.5085,
+      "step": 1456
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6896477341651917,
+      "learning_rate": 0.00046859152201299736,
+      "loss": 3.5511,
+      "step": 1457
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.731194019317627,
+      "learning_rate": 0.0004685412026160584,
+      "loss": 3.437,
+      "step": 1458
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8108443021774292,
+      "learning_rate": 0.00046849084564935323,
+      "loss": 3.5826,
+      "step": 1459
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.751708447933197,
+      "learning_rate": 0.00046844045112153865,
+      "loss": 3.5659,
+      "step": 1460
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7761220932006836,
+      "learning_rate": 0.0004683900190412782,
+      "loss": 3.5145,
+      "step": 1461
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8092109560966492,
+      "learning_rate": 0.0004683395494172417,
+      "loss": 3.5084,
+      "step": 1462
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8220993876457214,
+      "learning_rate": 0.0004682890422581054,
+      "loss": 3.5945,
+      "step": 1463
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7602437734603882,
+      "learning_rate": 0.0004682384975725522,
+      "loss": 3.4458,
+      "step": 1464
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6738101840019226,
+      "learning_rate": 0.0004681879153692711,
+      "loss": 3.486,
+      "step": 1465
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7281554341316223,
+      "learning_rate": 0.00046813729565695793,
+      "loss": 3.7172,
+      "step": 1466
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7399675846099854,
+      "learning_rate": 0.0004680866384443149,
+      "loss": 3.6074,
+      "step": 1467
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6513975262641907,
+      "learning_rate": 0.0004680359437400503,
+      "loss": 3.3342,
+      "step": 1468
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6898742914199829,
+      "learning_rate": 0.0004679852115528793,
+      "loss": 3.4383,
+      "step": 1469
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7081462740898132,
+      "learning_rate": 0.0004679344418915234,
+      "loss": 3.5978,
+      "step": 1470
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7182134985923767,
+      "learning_rate": 0.0004678836347647104,
+      "loss": 3.5795,
+      "step": 1471
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7415251135826111,
+      "learning_rate": 0.0004678327901811746,
+      "loss": 3.5322,
+      "step": 1472
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7213219404220581,
+      "learning_rate": 0.00046778190814965694,
+      "loss": 3.5142,
+      "step": 1473
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9378483295440674,
+      "learning_rate": 0.0004677309886789044,
+      "loss": 3.5648,
+      "step": 1474
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7091275453567505,
+      "learning_rate": 0.0004676800317776708,
+      "loss": 3.5034,
+      "step": 1475
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7190051078796387,
+      "learning_rate": 0.0004676290374547162,
+      "loss": 3.4522,
+      "step": 1476
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7434428930282593,
+      "learning_rate": 0.0004675780057188071,
+      "loss": 3.4756,
+      "step": 1477
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.822962760925293,
+      "learning_rate": 0.00046752693657871645,
+      "loss": 3.5878,
+      "step": 1478
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7522849440574646,
+      "learning_rate": 0.00046747583004322357,
+      "loss": 3.6562,
+      "step": 1479
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.769645631313324,
+      "learning_rate": 0.0004674246861211143,
+      "loss": 3.5938,
+      "step": 1480
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8104962706565857,
+      "learning_rate": 0.0004673735048211809,
+      "loss": 3.6472,
+      "step": 1481
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7202017903327942,
+      "learning_rate": 0.00046732228615222203,
+      "loss": 3.4748,
+      "step": 1482
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7173405289649963,
+      "learning_rate": 0.00046727103012304274,
+      "loss": 3.4189,
+      "step": 1483
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7136857509613037,
+      "learning_rate": 0.00046721973674245453,
+      "loss": 3.4613,
+      "step": 1484
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7353072762489319,
+      "learning_rate": 0.00046716840601927534,
+      "loss": 3.5693,
+      "step": 1485
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7930543422698975,
+      "learning_rate": 0.00046711703796232954,
+      "loss": 3.5676,
+      "step": 1486
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6626320481300354,
+      "learning_rate": 0.0004670656325804479,
+      "loss": 3.4426,
+      "step": 1487
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7427566647529602,
+      "learning_rate": 0.0004670141898824676,
+      "loss": 3.4905,
+      "step": 1488
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7173172831535339,
+      "learning_rate": 0.0004669627098772321,
+      "loss": 3.5909,
+      "step": 1489
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.768047034740448,
+      "learning_rate": 0.0004669111925735916,
+      "loss": 3.4904,
+      "step": 1490
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6736437082290649,
+      "learning_rate": 0.00046685963798040247,
+      "loss": 3.4495,
+      "step": 1491
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7259313464164734,
+      "learning_rate": 0.0004668080461065275,
+      "loss": 3.5522,
+      "step": 1492
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7390080690383911,
+      "learning_rate": 0.00046675641696083595,
+      "loss": 3.5692,
+      "step": 1493
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7455681562423706,
+      "learning_rate": 0.00046670475055220347,
+      "loss": 3.5092,
+      "step": 1494
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7226245999336243,
+      "learning_rate": 0.0004666530468895121,
+      "loss": 3.4348,
+      "step": 1495
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8105984926223755,
+      "learning_rate": 0.0004666013059816503,
+      "loss": 3.5211,
+      "step": 1496
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6822407841682434,
+      "learning_rate": 0.000466549527837513,
+      "loss": 3.5263,
+      "step": 1497
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7036212086677551,
+      "learning_rate": 0.00046649771246600136,
+      "loss": 3.4583,
+      "step": 1498
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7508665323257446,
+      "learning_rate": 0.00046644585987602304,
+      "loss": 3.4535,
+      "step": 1499
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6954589486122131,
+      "learning_rate": 0.0004663939700764923,
+      "loss": 3.5332,
+      "step": 1500
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6855521202087402,
+      "learning_rate": 0.0004663420430763293,
+      "loss": 3.3302,
+      "step": 1501
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7412664890289307,
+      "learning_rate": 0.00046629007888446115,
+      "loss": 3.4485,
+      "step": 1502
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7577595114707947,
+      "learning_rate": 0.00046623807750982094,
+      "loss": 3.4485,
+      "step": 1503
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7610722184181213,
+      "learning_rate": 0.00046618603896134836,
+      "loss": 3.6569,
+      "step": 1504
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7613515257835388,
+      "learning_rate": 0.00046613396324798943,
+      "loss": 3.4656,
+      "step": 1505
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7554030418395996,
+      "learning_rate": 0.0004660818503786965,
+      "loss": 3.4672,
+      "step": 1506
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7763833999633789,
+      "learning_rate": 0.00046602970036242866,
+      "loss": 3.6245,
+      "step": 1507
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8957259058952332,
+      "learning_rate": 0.00046597751320815074,
+      "loss": 3.4625,
+      "step": 1508
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7575466632843018,
+      "learning_rate": 0.00046592528892483453,
+      "loss": 3.5558,
+      "step": 1509
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7638049125671387,
+      "learning_rate": 0.00046587302752145793,
+      "loss": 3.5342,
+      "step": 1510
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7103486657142639,
+      "learning_rate": 0.00046582072900700524,
+      "loss": 3.4839,
+      "step": 1511
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7354874610900879,
+      "learning_rate": 0.00046576839339046726,
+      "loss": 3.547,
+      "step": 1512
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6934744715690613,
+      "learning_rate": 0.00046571602068084107,
+      "loss": 3.4264,
+      "step": 1513
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7283271551132202,
+      "learning_rate": 0.00046566361088713016,
+      "loss": 3.5117,
+      "step": 1514
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.738408088684082,
+      "learning_rate": 0.00046561116401834426,
+      "loss": 3.4262,
+      "step": 1515
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6994336843490601,
+      "learning_rate": 0.0004655586800834998,
+      "loss": 3.3924,
+      "step": 1516
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.723712682723999,
+      "learning_rate": 0.0004655061590916191,
+      "loss": 3.4571,
+      "step": 1517
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7158104777336121,
+      "learning_rate": 0.0004654536010517314,
+      "loss": 3.5852,
+      "step": 1518
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7097963690757751,
+      "learning_rate": 0.00046540100597287193,
+      "loss": 3.5034,
+      "step": 1519
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6684868335723877,
+      "learning_rate": 0.00046534837386408236,
+      "loss": 3.3736,
+      "step": 1520
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7114806175231934,
+      "learning_rate": 0.0004652957047344108,
+      "loss": 3.4636,
+      "step": 1521
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7204274535179138,
+      "learning_rate": 0.00046524299859291164,
+      "loss": 3.5365,
+      "step": 1522
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6938864588737488,
+      "learning_rate": 0.0004651902554486458,
+      "loss": 3.4385,
+      "step": 1523
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6994484663009644,
+      "learning_rate": 0.0004651374753106803,
+      "loss": 3.5652,
+      "step": 1524
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6984156966209412,
+      "learning_rate": 0.00046508465818808866,
+      "loss": 3.5164,
+      "step": 1525
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7477048635482788,
+      "learning_rate": 0.00046503180408995085,
+      "loss": 3.6011,
+      "step": 1526
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7297093868255615,
+      "learning_rate": 0.00046497891302535303,
+      "loss": 3.6558,
+      "step": 1527
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7676709890365601,
+      "learning_rate": 0.00046492598500338787,
+      "loss": 3.5184,
+      "step": 1528
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7769712209701538,
+      "learning_rate": 0.0004648730200331542,
+      "loss": 3.5287,
+      "step": 1529
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7823746204376221,
+      "learning_rate": 0.0004648200181237574,
+      "loss": 3.4073,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7949132323265076,
+      "learning_rate": 0.000464766979284309,
+      "loss": 3.6725,
+      "step": 1531
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7303690910339355,
+      "learning_rate": 0.0004647139035239272,
+      "loss": 3.6023,
+      "step": 1532
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7428348064422607,
+      "learning_rate": 0.0004646607908517361,
+      "loss": 3.582,
+      "step": 1533
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7381561994552612,
+      "learning_rate": 0.0004646076412768665,
+      "loss": 3.6566,
+      "step": 1534
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7015333771705627,
+      "learning_rate": 0.00046455445480845543,
+      "loss": 3.5161,
+      "step": 1535
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7148547172546387,
+      "learning_rate": 0.0004645012314556463,
+      "loss": 3.5613,
+      "step": 1536
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6725534796714783,
+      "learning_rate": 0.0004644479712275887,
+      "loss": 3.4803,
+      "step": 1537
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7428320050239563,
+      "learning_rate": 0.0004643946741334387,
+      "loss": 3.53,
+      "step": 1538
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6852494478225708,
+      "learning_rate": 0.00046434134018235885,
+      "loss": 3.584,
+      "step": 1539
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.717242419719696,
+      "learning_rate": 0.0004642879693835178,
+      "loss": 3.6089,
+      "step": 1540
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7618151903152466,
+      "learning_rate": 0.00046423456174609045,
+      "loss": 3.5611,
+      "step": 1541
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7496387362480164,
+      "learning_rate": 0.0004641811172792584,
+      "loss": 3.5066,
+      "step": 1542
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6788463592529297,
+      "learning_rate": 0.00046412763599220925,
+      "loss": 3.6358,
+      "step": 1543
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.735870361328125,
+      "learning_rate": 0.00046407411789413714,
+      "loss": 3.6166,
+      "step": 1544
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.766202986240387,
+      "learning_rate": 0.0004640205629942423,
+      "loss": 3.4675,
+      "step": 1545
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7808355093002319,
+      "learning_rate": 0.00046396697130173165,
+      "loss": 3.4217,
+      "step": 1546
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7486604452133179,
+      "learning_rate": 0.0004639133428258181,
+      "loss": 3.571,
+      "step": 1547
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.693970263004303,
+      "learning_rate": 0.000463859677575721,
+      "loss": 3.4341,
+      "step": 1548
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7594770193099976,
+      "learning_rate": 0.00046380597556066607,
+      "loss": 3.459,
+      "step": 1549
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.740746259689331,
+      "learning_rate": 0.0004637522367898852,
+      "loss": 3.5711,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7217738032341003,
+      "learning_rate": 0.00046369846127261696,
+      "loss": 3.5348,
+      "step": 1551
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7293241620063782,
+      "learning_rate": 0.0004636446490181057,
+      "loss": 3.5425,
+      "step": 1552
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7295405268669128,
+      "learning_rate": 0.0004635908000356025,
+      "loss": 3.6256,
+      "step": 1553
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7225673198699951,
+      "learning_rate": 0.00046353691433436464,
+      "loss": 3.5693,
+      "step": 1554
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7160700559616089,
+      "learning_rate": 0.00046348299192365566,
+      "loss": 3.6359,
+      "step": 1555
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7208033204078674,
+      "learning_rate": 0.00046342903281274553,
+      "loss": 3.4644,
+      "step": 1556
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.725329577922821,
+      "learning_rate": 0.00046337503701091026,
+      "loss": 3.5131,
+      "step": 1557
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.66863614320755,
+      "learning_rate": 0.0004633210045274325,
+      "loss": 3.5729,
+      "step": 1558
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6695690751075745,
+      "learning_rate": 0.00046326693537160113,
+      "loss": 3.5065,
+      "step": 1559
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6696643829345703,
+      "learning_rate": 0.0004632128295527111,
+      "loss": 3.5872,
+      "step": 1560
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6737921833992004,
+      "learning_rate": 0.00046315868708006393,
+      "loss": 3.3002,
+      "step": 1561
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6874504089355469,
+      "learning_rate": 0.0004631045079629672,
+      "loss": 3.4913,
+      "step": 1562
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7577842473983765,
+      "learning_rate": 0.00046305029221073516,
+      "loss": 3.5991,
+      "step": 1563
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7580335736274719,
+      "learning_rate": 0.0004629960398326879,
+      "loss": 3.541,
+      "step": 1564
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7388274669647217,
+      "learning_rate": 0.00046294175083815215,
+      "loss": 3.5442,
+      "step": 1565
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7765889763832092,
+      "learning_rate": 0.0004628874252364609,
+      "loss": 3.557,
+      "step": 1566
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.758765459060669,
+      "learning_rate": 0.00046283306303695316,
+      "loss": 3.5705,
+      "step": 1567
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7454854846000671,
+      "learning_rate": 0.0004627786642489745,
+      "loss": 3.5865,
+      "step": 1568
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7236392498016357,
+      "learning_rate": 0.0004627242288818767,
+      "loss": 3.4642,
+      "step": 1569
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7050071954727173,
+      "learning_rate": 0.0004626697569450179,
+      "loss": 3.3845,
+      "step": 1570
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6823355555534363,
+      "learning_rate": 0.00046261524844776235,
+      "loss": 3.4824,
+      "step": 1571
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7080915570259094,
+      "learning_rate": 0.0004625607033994808,
+      "loss": 3.6313,
+      "step": 1572
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7229819893836975,
+      "learning_rate": 0.0004625061218095501,
+      "loss": 3.4139,
+      "step": 1573
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7221214175224304,
+      "learning_rate": 0.00046245150368735344,
+      "loss": 3.5608,
+      "step": 1574
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7289392352104187,
+      "learning_rate": 0.0004623968490422804,
+      "loss": 3.5833,
+      "step": 1575
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7396621704101562,
+      "learning_rate": 0.0004623421578837267,
+      "loss": 3.4815,
+      "step": 1576
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.753093957901001,
+      "learning_rate": 0.0004622874302210943,
+      "loss": 3.5491,
+      "step": 1577
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7183561325073242,
+      "learning_rate": 0.00046223266606379166,
+      "loss": 3.5169,
+      "step": 1578
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7659851312637329,
+      "learning_rate": 0.0004621778654212333,
+      "loss": 3.6561,
+      "step": 1579
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.74198317527771,
+      "learning_rate": 0.00046212302830284015,
+      "loss": 3.5021,
+      "step": 1580
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7400766015052795,
+      "learning_rate": 0.0004620681547180392,
+      "loss": 3.5211,
+      "step": 1581
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7454343438148499,
+      "learning_rate": 0.00046201324467626405,
+      "loss": 3.6262,
+      "step": 1582
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6770966649055481,
+      "learning_rate": 0.0004619582981869542,
+      "loss": 3.4959,
+      "step": 1583
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6676375269889832,
+      "learning_rate": 0.00046190331525955566,
+      "loss": 3.4918,
+      "step": 1584
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7769284248352051,
+      "learning_rate": 0.0004618482959035206,
+      "loss": 3.5092,
+      "step": 1585
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7751719951629639,
+      "learning_rate": 0.0004617932401283076,
+      "loss": 3.4876,
+      "step": 1586
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6945412755012512,
+      "learning_rate": 0.0004617381479433813,
+      "loss": 3.3976,
+      "step": 1587
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.803017795085907,
+      "learning_rate": 0.0004616830193582127,
+      "loss": 3.4917,
+      "step": 1588
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7180715799331665,
+      "learning_rate": 0.00046162785438227895,
+      "loss": 3.4079,
+      "step": 1589
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7560957074165344,
+      "learning_rate": 0.0004615726530250637,
+      "loss": 3.5088,
+      "step": 1590
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8058301210403442,
+      "learning_rate": 0.00046151741529605654,
+      "loss": 3.5657,
+      "step": 1591
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7559605240821838,
+      "learning_rate": 0.00046146214120475367,
+      "loss": 3.635,
+      "step": 1592
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7491931319236755,
+      "learning_rate": 0.0004614068307606572,
+      "loss": 3.4993,
+      "step": 1593
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7718519568443298,
+      "learning_rate": 0.0004613514839732757,
+      "loss": 3.5777,
+      "step": 1594
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7517576813697815,
+      "learning_rate": 0.00046129610085212394,
+      "loss": 3.4151,
+      "step": 1595
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7024547457695007,
+      "learning_rate": 0.00046124068140672284,
+      "loss": 3.4908,
+      "step": 1596
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7161441445350647,
+      "learning_rate": 0.0004611852256465997,
+      "loss": 3.5818,
+      "step": 1597
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6977832317352295,
+      "learning_rate": 0.00046112973358128796,
+      "loss": 3.4198,
+      "step": 1598
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6729937195777893,
+      "learning_rate": 0.0004610742052203275,
+      "loss": 3.6251,
+      "step": 1599
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7269434332847595,
+      "learning_rate": 0.0004610186405732641,
+      "loss": 3.5543,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7198590636253357,
+      "learning_rate": 0.00046096303964965004,
+      "loss": 3.5553,
+      "step": 1601
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7438438534736633,
+      "learning_rate": 0.00046090740245904383,
+      "loss": 3.4807,
+      "step": 1602
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7416555285453796,
+      "learning_rate": 0.00046085172901101006,
+      "loss": 3.4353,
+      "step": 1603
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7552011013031006,
+      "learning_rate": 0.0004607960193151197,
+      "loss": 3.5533,
+      "step": 1604
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6969509124755859,
+      "learning_rate": 0.00046074027338094983,
+      "loss": 3.5417,
+      "step": 1605
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6895467042922974,
+      "learning_rate": 0.0004606844912180839,
+      "loss": 3.5665,
+      "step": 1606
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6941189765930176,
+      "learning_rate": 0.0004606286728361113,
+      "loss": 3.5748,
+      "step": 1607
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6663969159126282,
+      "learning_rate": 0.0004605728182446282,
+      "loss": 3.3967,
+      "step": 1608
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7020400166511536,
+      "learning_rate": 0.0004605169274532364,
+      "loss": 3.4237,
+      "step": 1609
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7018489837646484,
+      "learning_rate": 0.00046046100047154425,
+      "loss": 3.6572,
+      "step": 1610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7054951190948486,
+      "learning_rate": 0.0004604050373091663,
+      "loss": 3.3447,
+      "step": 1611
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8908802270889282,
+      "learning_rate": 0.0004603490379757232,
+      "loss": 3.5154,
+      "step": 1612
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6973745226860046,
+      "learning_rate": 0.00046029300248084183,
+      "loss": 3.4249,
+      "step": 1613
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6773630976676941,
+      "learning_rate": 0.0004602369308341555,
+      "loss": 3.4243,
+      "step": 1614
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7102717757225037,
+      "learning_rate": 0.0004601808230453034,
+      "loss": 3.6138,
+      "step": 1615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6981760859489441,
+      "learning_rate": 0.00046012467912393126,
+      "loss": 3.3881,
+      "step": 1616
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7004696726799011,
+      "learning_rate": 0.0004600684990796907,
+      "loss": 3.5152,
+      "step": 1617
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.77801513671875,
+      "learning_rate": 0.00046001228292223993,
+      "loss": 3.5409,
+      "step": 1618
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7654480338096619,
+      "learning_rate": 0.00045995603066124305,
+      "loss": 3.375,
+      "step": 1619
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7013776302337646,
+      "learning_rate": 0.00045989974230637045,
+      "loss": 3.4237,
+      "step": 1620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7696542739868164,
+      "learning_rate": 0.0004598434178672988,
+      "loss": 3.6602,
+      "step": 1621
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.912253201007843,
+      "learning_rate": 0.00045978705735371083,
+      "loss": 3.5991,
+      "step": 1622
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7334491610527039,
+      "learning_rate": 0.00045973066077529574,
+      "loss": 3.3882,
+      "step": 1623
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7463875412940979,
+      "learning_rate": 0.00045967422814174863,
+      "loss": 3.4675,
+      "step": 1624
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.730761706829071,
+      "learning_rate": 0.0004596177594627709,
+      "loss": 3.6244,
+      "step": 1625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6837449669837952,
+      "learning_rate": 0.00045956125474807034,
+      "loss": 3.3481,
+      "step": 1626
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.722831130027771,
+      "learning_rate": 0.0004595047140073605,
+      "loss": 3.546,
+      "step": 1627
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7261539101600647,
+      "learning_rate": 0.0004594481372503616,
+      "loss": 3.5072,
+      "step": 1628
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7304078936576843,
+      "learning_rate": 0.00045939152448679977,
+      "loss": 3.5156,
+      "step": 1629
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6343042254447937,
+      "learning_rate": 0.0004593348757264074,
+      "loss": 3.567,
+      "step": 1630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7293105125427246,
+      "learning_rate": 0.0004592781909789231,
+      "loss": 3.536,
+      "step": 1631
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7170096635818481,
+      "learning_rate": 0.0004592214702540916,
+      "loss": 3.5915,
+      "step": 1632
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7097491025924683,
+      "learning_rate": 0.00045916471356166383,
+      "loss": 3.5545,
+      "step": 1633
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6664694547653198,
+      "learning_rate": 0.000459107920911397,
+      "loss": 3.5604,
+      "step": 1634
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7300134301185608,
+      "learning_rate": 0.00045905109231305437,
+      "loss": 3.4362,
+      "step": 1635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7318281531333923,
+      "learning_rate": 0.00045899422777640543,
+      "loss": 3.5118,
+      "step": 1636
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7411096692085266,
+      "learning_rate": 0.00045893732731122584,
+      "loss": 3.472,
+      "step": 1637
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6925008893013,
+      "learning_rate": 0.0004588803909272975,
+      "loss": 3.4881,
+      "step": 1638
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.735426664352417,
+      "learning_rate": 0.0004588234186344084,
+      "loss": 3.5375,
+      "step": 1639
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7070428729057312,
+      "learning_rate": 0.0004587664104423528,
+      "loss": 3.5516,
+      "step": 1640
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6997687816619873,
+      "learning_rate": 0.000458709366360931,
+      "loss": 3.5889,
+      "step": 1641
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.709517240524292,
+      "learning_rate": 0.0004586522863999495,
+      "loss": 3.6729,
+      "step": 1642
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7886548638343811,
+      "learning_rate": 0.0004585951705692211,
+      "loss": 3.6147,
+      "step": 1643
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7466222643852234,
+      "learning_rate": 0.0004585380188785646,
+      "loss": 3.4541,
+      "step": 1644
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6836808323860168,
+      "learning_rate": 0.0004584808313378051,
+      "loss": 3.4846,
+      "step": 1645
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6720550060272217,
+      "learning_rate": 0.0004584236079567738,
+      "loss": 3.6527,
+      "step": 1646
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6817904114723206,
+      "learning_rate": 0.000458366348745308,
+      "loss": 3.4742,
+      "step": 1647
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7797605991363525,
+      "learning_rate": 0.00045830905371325125,
+      "loss": 3.5033,
+      "step": 1648
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7567112445831299,
+      "learning_rate": 0.0004582517228704533,
+      "loss": 3.4196,
+      "step": 1649
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7362318634986877,
+      "learning_rate": 0.00045819435622676985,
+      "loss": 3.5975,
+      "step": 1650
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7481383085250854,
+      "learning_rate": 0.0004581369537920631,
+      "loss": 3.5421,
+      "step": 1651
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.741283118724823,
+      "learning_rate": 0.000458079515576201,
+      "loss": 3.6068,
+      "step": 1652
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7184082865715027,
+      "learning_rate": 0.00045802204158905787,
+      "loss": 3.4343,
+      "step": 1653
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7001619935035706,
+      "learning_rate": 0.00045796453184051417,
+      "loss": 3.5601,
+      "step": 1654
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7675745487213135,
+      "learning_rate": 0.0004579069863404566,
+      "loss": 3.5921,
+      "step": 1655
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6656587719917297,
+      "learning_rate": 0.0004578494050987777,
+      "loss": 3.5763,
+      "step": 1656
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7041658759117126,
+      "learning_rate": 0.0004577917881253766,
+      "loss": 3.5689,
+      "step": 1657
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6746188998222351,
+      "learning_rate": 0.0004577341354301581,
+      "loss": 3.4682,
+      "step": 1658
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7306371927261353,
+      "learning_rate": 0.00045767644702303346,
+      "loss": 3.5342,
+      "step": 1659
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6800777912139893,
+      "learning_rate": 0.00045761872291392005,
+      "loss": 3.3995,
+      "step": 1660
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7372608780860901,
+      "learning_rate": 0.00045756096311274123,
+      "loss": 3.4871,
+      "step": 1661
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7147559523582458,
+      "learning_rate": 0.00045750316762942656,
+      "loss": 3.4816,
+      "step": 1662
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6870706677436829,
+      "learning_rate": 0.0004574453364739118,
+      "loss": 3.3661,
+      "step": 1663
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7589773535728455,
+      "learning_rate": 0.00045738746965613876,
+      "loss": 3.5928,
+      "step": 1664
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7578521966934204,
+      "learning_rate": 0.0004573295671860555,
+      "loss": 3.5298,
+      "step": 1665
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7245187163352966,
+      "learning_rate": 0.0004572716290736161,
+      "loss": 3.4699,
+      "step": 1666
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.807602047920227,
+      "learning_rate": 0.00045721365532878065,
+      "loss": 3.4792,
+      "step": 1667
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.735205352306366,
+      "learning_rate": 0.0004571556459615157,
+      "loss": 3.5042,
+      "step": 1668
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7455587983131409,
+      "learning_rate": 0.00045709760098179363,
+      "loss": 3.6033,
+      "step": 1669
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7397889494895935,
+      "learning_rate": 0.00045703952039959306,
+      "loss": 3.6053,
+      "step": 1670
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6866623163223267,
+      "learning_rate": 0.00045698140422489875,
+      "loss": 3.5595,
+      "step": 1671
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6840174198150635,
+      "learning_rate": 0.00045692325246770156,
+      "loss": 3.4672,
+      "step": 1672
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7191321849822998,
+      "learning_rate": 0.0004568650651379984,
+      "loss": 3.5568,
+      "step": 1673
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6546381115913391,
+      "learning_rate": 0.0004568068422457923,
+      "loss": 3.5321,
+      "step": 1674
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6783313155174255,
+      "learning_rate": 0.0004567485838010925,
+      "loss": 3.5561,
+      "step": 1675
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7149572968482971,
+      "learning_rate": 0.00045669028981391434,
+      "loss": 3.4968,
+      "step": 1676
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.754173755645752,
+      "learning_rate": 0.00045663196029427925,
+      "loss": 3.4835,
+      "step": 1677
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7751779556274414,
+      "learning_rate": 0.00045657359525221465,
+      "loss": 3.4883,
+      "step": 1678
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6891859769821167,
+      "learning_rate": 0.0004565151946977542,
+      "loss": 3.507,
+      "step": 1679
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7515348196029663,
+      "learning_rate": 0.00045645675864093766,
+      "loss": 3.5647,
+      "step": 1680
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7117422223091125,
+      "learning_rate": 0.0004563982870918109,
+      "loss": 3.4871,
+      "step": 1681
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7142488956451416,
+      "learning_rate": 0.00045633978006042575,
+      "loss": 3.5224,
+      "step": 1682
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7000136375427246,
+      "learning_rate": 0.00045628123755684036,
+      "loss": 3.4883,
+      "step": 1683
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.766757071018219,
+      "learning_rate": 0.0004562226595911188,
+      "loss": 3.5763,
+      "step": 1684
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7023899555206299,
+      "learning_rate": 0.0004561640461733313,
+      "loss": 3.4384,
+      "step": 1685
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7456826567649841,
+      "learning_rate": 0.0004561053973135543,
+      "loss": 3.5366,
+      "step": 1686
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7246689200401306,
+      "learning_rate": 0.00045604671302187,
+      "loss": 3.6063,
+      "step": 1687
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7071058750152588,
+      "learning_rate": 0.000455987993308367,
+      "loss": 3.5176,
+      "step": 1688
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7235510349273682,
+      "learning_rate": 0.00045592923818314014,
+      "loss": 3.4324,
+      "step": 1689
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6815816760063171,
+      "learning_rate": 0.00045587044765628973,
+      "loss": 3.405,
+      "step": 1690
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7223538756370544,
+      "learning_rate": 0.0004558116217379228,
+      "loss": 3.5051,
+      "step": 1691
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.698800265789032,
+      "learning_rate": 0.00045575276043815203,
+      "loss": 3.5592,
+      "step": 1692
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7396284937858582,
+      "learning_rate": 0.00045569386376709655,
+      "loss": 3.4582,
+      "step": 1693
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6998549103736877,
+      "learning_rate": 0.0004556349317348812,
+      "loss": 3.4187,
+      "step": 1694
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6666313409805298,
+      "learning_rate": 0.0004555759643516372,
+      "loss": 3.546,
+      "step": 1695
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7003994584083557,
+      "learning_rate": 0.0004555169616275017,
+      "loss": 3.5365,
+      "step": 1696
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7141129374504089,
+      "learning_rate": 0.00045545792357261784,
+      "loss": 3.5452,
+      "step": 1697
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6956498622894287,
+      "learning_rate": 0.0004553988501971351,
+      "loss": 3.4388,
+      "step": 1698
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6885257363319397,
+      "learning_rate": 0.00045533974151120896,
+      "loss": 3.4795,
+      "step": 1699
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8513539433479309,
+      "learning_rate": 0.0004552805975250006,
+      "loss": 3.5629,
+      "step": 1700
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7191634178161621,
+      "learning_rate": 0.0004552214182486777,
+      "loss": 3.4614,
+      "step": 1701
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6562285423278809,
+      "learning_rate": 0.0004551622036924139,
+      "loss": 3.6592,
+      "step": 1702
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.673965334892273,
+      "learning_rate": 0.0004551029538663889,
+      "loss": 3.6226,
+      "step": 1703
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7001772522926331,
+      "learning_rate": 0.00045504366878078826,
+      "loss": 3.4116,
+      "step": 1704
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6701346635818481,
+      "learning_rate": 0.0004549843484458041,
+      "loss": 3.4371,
+      "step": 1705
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.659070611000061,
+      "learning_rate": 0.0004549249928716338,
+      "loss": 3.4649,
+      "step": 1706
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7089521884918213,
+      "learning_rate": 0.0004548656020684817,
+      "loss": 3.4124,
+      "step": 1707
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6980241537094116,
+      "learning_rate": 0.0004548061760465575,
+      "loss": 3.4206,
+      "step": 1708
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7338328957557678,
+      "learning_rate": 0.00045474671481607744,
+      "loss": 3.4859,
+      "step": 1709
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7270656824111938,
+      "learning_rate": 0.00045468721838726336,
+      "loss": 3.5368,
+      "step": 1710
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6947497129440308,
+      "learning_rate": 0.00045462768677034354,
+      "loss": 3.5811,
+      "step": 1711
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7697454690933228,
+      "learning_rate": 0.00045456811997555215,
+      "loss": 3.47,
+      "step": 1712
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7647615671157837,
+      "learning_rate": 0.0004545085180131293,
+      "loss": 3.523,
+      "step": 1713
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7090777158737183,
+      "learning_rate": 0.0004544488808933214,
+      "loss": 3.4826,
+      "step": 1714
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6906019449234009,
+      "learning_rate": 0.0004543892086263807,
+      "loss": 3.4338,
+      "step": 1715
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.709435760974884,
+      "learning_rate": 0.0004543295012225656,
+      "loss": 3.3788,
+      "step": 1716
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6874732971191406,
+      "learning_rate": 0.00045426975869214035,
+      "loss": 3.5806,
+      "step": 1717
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7035239338874817,
+      "learning_rate": 0.0004542099810453755,
+      "loss": 3.5524,
+      "step": 1718
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.682473361492157,
+      "learning_rate": 0.0004541501682925475,
+      "loss": 3.4603,
+      "step": 1719
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.756766676902771,
+      "learning_rate": 0.0004540903204439389,
+      "loss": 3.5824,
+      "step": 1720
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.716594934463501,
+      "learning_rate": 0.0004540304375098382,
+      "loss": 3.4851,
+      "step": 1721
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6676660180091858,
+      "learning_rate": 0.0004539705195005399,
+      "loss": 3.4948,
+      "step": 1722
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7688947319984436,
+      "learning_rate": 0.00045391056642634476,
+      "loss": 3.3265,
+      "step": 1723
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7121802568435669,
+      "learning_rate": 0.00045385057829755925,
+      "loss": 3.4246,
+      "step": 1724
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7320714592933655,
+      "learning_rate": 0.00045379055512449615,
+      "loss": 3.5915,
+      "step": 1725
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7020367980003357,
+      "learning_rate": 0.00045373049691747403,
+      "loss": 3.585,
+      "step": 1726
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7828790545463562,
+      "learning_rate": 0.0004536704036868177,
+      "loss": 3.4396,
+      "step": 1727
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7378258109092712,
+      "learning_rate": 0.0004536102754428577,
+      "loss": 3.5462,
+      "step": 1728
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6563218235969543,
+      "learning_rate": 0.00045355011219593103,
+      "loss": 3.5116,
+      "step": 1729
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7141712307929993,
+      "learning_rate": 0.00045348991395638026,
+      "loss": 3.6652,
+      "step": 1730
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6334406733512878,
+      "learning_rate": 0.00045342968073455427,
+      "loss": 3.5444,
+      "step": 1731
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6695120334625244,
+      "learning_rate": 0.0004533694125408078,
+      "loss": 3.5061,
+      "step": 1732
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6740787029266357,
+      "learning_rate": 0.00045330910938550157,
+      "loss": 3.6269,
+      "step": 1733
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6897600889205933,
+      "learning_rate": 0.00045324877127900253,
+      "loss": 3.3207,
+      "step": 1734
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6967800259590149,
+      "learning_rate": 0.00045318839823168345,
+      "loss": 3.4868,
+      "step": 1735
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8088453412055969,
+      "learning_rate": 0.00045312799025392313,
+      "loss": 3.5164,
+      "step": 1736
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7367091774940491,
+      "learning_rate": 0.00045306754735610643,
+      "loss": 3.4248,
+      "step": 1737
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7140554189682007,
+      "learning_rate": 0.00045300706954862425,
+      "loss": 3.3806,
+      "step": 1738
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7674543857574463,
+      "learning_rate": 0.00045294655684187325,
+      "loss": 3.5559,
+      "step": 1739
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.745649516582489,
+      "learning_rate": 0.00045288600924625643,
+      "loss": 3.5134,
+      "step": 1740
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7173015475273132,
+      "learning_rate": 0.00045282542677218255,
+      "loss": 3.4533,
+      "step": 1741
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6862587928771973,
+      "learning_rate": 0.00045276480943006646,
+      "loss": 3.6471,
+      "step": 1742
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5865806341171265,
+      "learning_rate": 0.00045270415723032897,
+      "loss": 3.3734,
+      "step": 1743
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.682898223400116,
+      "learning_rate": 0.0004526434701833969,
+      "loss": 3.6128,
+      "step": 1744
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6636016964912415,
+      "learning_rate": 0.000452582748299703,
+      "loss": 3.5103,
+      "step": 1745
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6695670485496521,
+      "learning_rate": 0.0004525219915896863,
+      "loss": 3.4878,
+      "step": 1746
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7170305848121643,
+      "learning_rate": 0.00045246120006379125,
+      "loss": 3.5729,
+      "step": 1747
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6496245265007019,
+      "learning_rate": 0.00045240037373246885,
+      "loss": 3.5039,
+      "step": 1748
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.727599561214447,
+      "learning_rate": 0.0004523395126061757,
+      "loss": 3.4796,
+      "step": 1749
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6873193383216858,
+      "learning_rate": 0.00045227861669537475,
+      "loss": 3.5528,
+      "step": 1750
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7259441614151001,
+      "learning_rate": 0.0004522176860105345,
+      "loss": 3.5181,
+      "step": 1751
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7488285303115845,
+      "learning_rate": 0.00045215672056212966,
+      "loss": 3.4781,
+      "step": 1752
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7147641777992249,
+      "learning_rate": 0.00045209572036064105,
+      "loss": 3.6021,
+      "step": 1753
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.760727047920227,
+      "learning_rate": 0.00045203468541655514,
+      "loss": 3.4344,
+      "step": 1754
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7726491689682007,
+      "learning_rate": 0.00045197361574036466,
+      "loss": 3.4809,
+      "step": 1755
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7099810838699341,
+      "learning_rate": 0.0004519125113425681,
+      "loss": 3.5666,
+      "step": 1756
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6434507966041565,
+      "learning_rate": 0.0004518513722336701,
+      "loss": 3.4913,
+      "step": 1757
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6790302395820618,
+      "learning_rate": 0.00045179019842418126,
+      "loss": 3.5466,
+      "step": 1758
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7132038474082947,
+      "learning_rate": 0.0004517289899246178,
+      "loss": 3.4942,
+      "step": 1759
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6917662620544434,
+      "learning_rate": 0.0004516677467455024,
+      "loss": 3.5228,
+      "step": 1760
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7276488542556763,
+      "learning_rate": 0.00045160646889736333,
+      "loss": 3.491,
+      "step": 1761
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.671357274055481,
+      "learning_rate": 0.00045154515639073513,
+      "loss": 3.476,
+      "step": 1762
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6840657591819763,
+      "learning_rate": 0.00045148380923615804,
+      "loss": 3.4813,
+      "step": 1763
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6774691343307495,
+      "learning_rate": 0.0004514224274441783,
+      "loss": 3.46,
+      "step": 1764
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7096502184867859,
+      "learning_rate": 0.00045136101102534823,
+      "loss": 3.3175,
+      "step": 1765
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7626500129699707,
+      "learning_rate": 0.00045129955999022585,
+      "loss": 3.4774,
+      "step": 1766
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7550767064094543,
+      "learning_rate": 0.0004512380743493756,
+      "loss": 3.5936,
+      "step": 1767
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.718117892742157,
+      "learning_rate": 0.00045117655411336735,
+      "loss": 3.4214,
+      "step": 1768
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7564663887023926,
+      "learning_rate": 0.00045111499929277723,
+      "loss": 3.5129,
+      "step": 1769
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7053793668746948,
+      "learning_rate": 0.0004510534098981872,
+      "loss": 3.4276,
+      "step": 1770
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7066434621810913,
+      "learning_rate": 0.0004509917859401852,
+      "loss": 3.4055,
+      "step": 1771
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6779627799987793,
+      "learning_rate": 0.0004509301274293651,
+      "loss": 3.3915,
+      "step": 1772
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6719902157783508,
+      "learning_rate": 0.00045086843437632673,
+      "loss": 3.3999,
+      "step": 1773
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6965342164039612,
+      "learning_rate": 0.0004508067067916758,
+      "loss": 3.3755,
+      "step": 1774
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7080169320106506,
+      "learning_rate": 0.000450744944686024,
+      "loss": 3.602,
+      "step": 1775
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8627835512161255,
+      "learning_rate": 0.000450683148069989,
+      "loss": 3.4469,
+      "step": 1776
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7523432374000549,
+      "learning_rate": 0.00045062131695419434,
+      "loss": 3.5508,
+      "step": 1777
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7530139088630676,
+      "learning_rate": 0.00045055945134926944,
+      "loss": 3.4976,
+      "step": 1778
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7500725388526917,
+      "learning_rate": 0.00045049755126584987,
+      "loss": 3.4723,
+      "step": 1779
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7203992009162903,
+      "learning_rate": 0.00045043561671457677,
+      "loss": 3.5489,
+      "step": 1780
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7485403418540955,
+      "learning_rate": 0.0004503736477060976,
+      "loss": 3.6029,
+      "step": 1781
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.65134197473526,
+      "learning_rate": 0.0004503116442510654,
+      "loss": 3.5845,
+      "step": 1782
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.671821653842926,
+      "learning_rate": 0.00045024960636013935,
+      "loss": 3.4996,
+      "step": 1783
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7114441394805908,
+      "learning_rate": 0.00045018753404398444,
+      "loss": 3.5538,
+      "step": 1784
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6860888600349426,
+      "learning_rate": 0.0004501254273132717,
+      "loss": 3.5501,
+      "step": 1785
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6691362261772156,
+      "learning_rate": 0.0004500632861786779,
+      "loss": 3.495,
+      "step": 1786
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6683976054191589,
+      "learning_rate": 0.00045000111065088597,
+      "loss": 3.4391,
+      "step": 1787
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6653891801834106,
+      "learning_rate": 0.00044993890074058443,
+      "loss": 3.5648,
+      "step": 1788
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6728489995002747,
+      "learning_rate": 0.000449876656458468,
+      "loss": 3.4839,
+      "step": 1789
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6840108036994934,
+      "learning_rate": 0.00044981437781523714,
+      "loss": 3.4492,
+      "step": 1790
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7206718921661377,
+      "learning_rate": 0.00044975206482159827,
+      "loss": 3.4534,
+      "step": 1791
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6660496592521667,
+      "learning_rate": 0.0004496897174882637,
+      "loss": 3.4322,
+      "step": 1792
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6936440467834473,
+      "learning_rate": 0.0004496273358259517,
+      "loss": 3.5069,
+      "step": 1793
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7100087404251099,
+      "learning_rate": 0.0004495649198453865,
+      "loss": 3.5615,
+      "step": 1794
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7237424850463867,
+      "learning_rate": 0.00044950246955729794,
+      "loss": 3.6016,
+      "step": 1795
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7479493618011475,
+      "learning_rate": 0.00044943998497242203,
+      "loss": 3.5855,
+      "step": 1796
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6842243075370789,
+      "learning_rate": 0.00044937746610150065,
+      "loss": 3.3482,
+      "step": 1797
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7530412077903748,
+      "learning_rate": 0.00044931491295528144,
+      "loss": 3.5059,
+      "step": 1798
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6601586937904358,
+      "learning_rate": 0.000449252325544518,
+      "loss": 3.3957,
+      "step": 1799
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7175249457359314,
+      "learning_rate": 0.0004491897038799699,
+      "loss": 3.5889,
+      "step": 1800
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7499895095825195,
+      "learning_rate": 0.00044912704797240243,
+      "loss": 3.3806,
+      "step": 1801
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7383548617362976,
+      "learning_rate": 0.000449064357832587,
+      "loss": 3.4735,
+      "step": 1802
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6868014335632324,
+      "learning_rate": 0.00044900163347130073,
+      "loss": 3.5739,
+      "step": 1803
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7332748770713806,
+      "learning_rate": 0.0004489388748993266,
+      "loss": 3.3961,
+      "step": 1804
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.695375382900238,
+      "learning_rate": 0.0004488760821274536,
+      "loss": 3.4391,
+      "step": 1805
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6780656576156616,
+      "learning_rate": 0.00044881325516647654,
+      "loss": 3.5523,
+      "step": 1806
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6870729923248291,
+      "learning_rate": 0.00044875039402719606,
+      "loss": 3.4855,
+      "step": 1807
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7325615286827087,
+      "learning_rate": 0.0004486874987204187,
+      "loss": 3.5555,
+      "step": 1808
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7024620175361633,
+      "learning_rate": 0.0004486245692569569,
+      "loss": 3.4482,
+      "step": 1809
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7054506540298462,
+      "learning_rate": 0.00044856160564762904,
+      "loss": 3.4569,
+      "step": 1810
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7186829447746277,
+      "learning_rate": 0.0004484986079032592,
+      "loss": 3.4807,
+      "step": 1811
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6961916089057922,
+      "learning_rate": 0.00044843557603467754,
+      "loss": 3.4053,
+      "step": 1812
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.797313928604126,
+      "learning_rate": 0.00044837251005271984,
+      "loss": 3.4249,
+      "step": 1813
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7082807421684265,
+      "learning_rate": 0.000448309409968228,
+      "loss": 3.6033,
+      "step": 1814
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7119776606559753,
+      "learning_rate": 0.00044824627579204953,
+      "loss": 3.4228,
+      "step": 1815
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.724306583404541,
+      "learning_rate": 0.000448183107535038,
+      "loss": 3.42,
+      "step": 1816
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7299683690071106,
+      "learning_rate": 0.0004481199052080527,
+      "loss": 3.4756,
+      "step": 1817
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7112978100776672,
+      "learning_rate": 0.0004480566688219589,
+      "loss": 3.4947,
+      "step": 1818
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6957636475563049,
+      "learning_rate": 0.0004479933983876278,
+      "loss": 3.5313,
+      "step": 1819
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6685547232627869,
+      "learning_rate": 0.000447930093915936,
+      "loss": 3.4602,
+      "step": 1820
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6561869382858276,
+      "learning_rate": 0.00044786675541776653,
+      "loss": 3.4492,
+      "step": 1821
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6853393912315369,
+      "learning_rate": 0.0004478033829040079,
+      "loss": 3.5309,
+      "step": 1822
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8650455474853516,
+      "learning_rate": 0.0004477399763855547,
+      "loss": 3.4467,
+      "step": 1823
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6917409896850586,
+      "learning_rate": 0.0004476765358733071,
+      "loss": 3.3812,
+      "step": 1824
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6957911252975464,
+      "learning_rate": 0.0004476130613781713,
+      "loss": 3.5238,
+      "step": 1825
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7351962327957153,
+      "learning_rate": 0.0004475495529110594,
+      "loss": 3.6196,
+      "step": 1826
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7193084955215454,
+      "learning_rate": 0.00044748601048288906,
+      "loss": 3.4937,
+      "step": 1827
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6810345649719238,
+      "learning_rate": 0.00044742243410458415,
+      "loss": 3.5354,
+      "step": 1828
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7194337844848633,
+      "learning_rate": 0.00044735882378707406,
+      "loss": 3.4368,
+      "step": 1829
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6895034909248352,
+      "learning_rate": 0.00044729517954129416,
+      "loss": 3.4814,
+      "step": 1830
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7407213449478149,
+      "learning_rate": 0.0004472315013781857,
+      "loss": 3.5947,
+      "step": 1831
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6990834474563599,
+      "learning_rate": 0.0004471677893086956,
+      "loss": 3.5198,
+      "step": 1832
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6894617080688477,
+      "learning_rate": 0.00044710404334377684,
+      "loss": 3.3916,
+      "step": 1833
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7195038795471191,
+      "learning_rate": 0.000447040263494388,
+      "loss": 3.5726,
+      "step": 1834
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6901001334190369,
+      "learning_rate": 0.00044697644977149346,
+      "loss": 3.5009,
+      "step": 1835
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7657378911972046,
+      "learning_rate": 0.0004469126021860637,
+      "loss": 3.483,
+      "step": 1836
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6822826862335205,
+      "learning_rate": 0.0004468487207490749,
+      "loss": 3.5181,
+      "step": 1837
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7345016598701477,
+      "learning_rate": 0.00044678480547150886,
+      "loss": 3.4361,
+      "step": 1838
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7520294189453125,
+      "learning_rate": 0.00044672085636435346,
+      "loss": 3.5237,
+      "step": 1839
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7586507797241211,
+      "learning_rate": 0.0004466568734386023,
+      "loss": 3.4596,
+      "step": 1840
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6959397196769714,
+      "learning_rate": 0.00044659285670525464,
+      "loss": 3.4634,
+      "step": 1841
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.708108127117157,
+      "learning_rate": 0.00044652880617531587,
+      "loss": 3.4588,
+      "step": 1842
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7102747559547424,
+      "learning_rate": 0.000446464721859797,
+      "loss": 3.6206,
+      "step": 1843
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6688615083694458,
+      "learning_rate": 0.0004464006037697148,
+      "loss": 3.3673,
+      "step": 1844
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7083567976951599,
+      "learning_rate": 0.000446336451916092,
+      "loss": 3.4632,
+      "step": 1845
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.655749499797821,
+      "learning_rate": 0.0004462722663099569,
+      "loss": 3.4694,
+      "step": 1846
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7156842947006226,
+      "learning_rate": 0.00044620804696234387,
+      "loss": 3.4285,
+      "step": 1847
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6834205985069275,
+      "learning_rate": 0.0004461437938842929,
+      "loss": 3.5335,
+      "step": 1848
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6689387559890747,
+      "learning_rate": 0.00044607950708685,
+      "loss": 3.4063,
+      "step": 1849
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7429822683334351,
+      "learning_rate": 0.00044601518658106653,
+      "loss": 3.4659,
+      "step": 1850
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6987894177436829,
+      "learning_rate": 0.0004459508323780001,
+      "loss": 3.636,
+      "step": 1851
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6989704370498657,
+      "learning_rate": 0.00044588644448871395,
+      "loss": 3.4733,
+      "step": 1852
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.69915372133255,
+      "learning_rate": 0.0004458220229242771,
+      "loss": 3.5979,
+      "step": 1853
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7444015145301819,
+      "learning_rate": 0.0004457575676957644,
+      "loss": 3.5744,
+      "step": 1854
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6756484508514404,
+      "learning_rate": 0.0004456930788142563,
+      "loss": 3.4098,
+      "step": 1855
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6732588410377502,
+      "learning_rate": 0.0004456285562908393,
+      "loss": 3.4806,
+      "step": 1856
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6970331072807312,
+      "learning_rate": 0.0004455640001366056,
+      "loss": 3.6507,
+      "step": 1857
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7336878180503845,
+      "learning_rate": 0.00044549941036265306,
+      "loss": 3.6066,
+      "step": 1858
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6298251152038574,
+      "learning_rate": 0.00044543478698008546,
+      "loss": 3.4287,
+      "step": 1859
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6941573023796082,
+      "learning_rate": 0.0004453701300000124,
+      "loss": 3.5047,
+      "step": 1860
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6953024864196777,
+      "learning_rate": 0.00044530543943354896,
+      "loss": 3.521,
+      "step": 1861
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6954164505004883,
+      "learning_rate": 0.0004452407152918163,
+      "loss": 3.4325,
+      "step": 1862
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7267298102378845,
+      "learning_rate": 0.0004451759575859413,
+      "loss": 3.4498,
+      "step": 1863
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6570194959640503,
+      "learning_rate": 0.0004451111663270565,
+      "loss": 3.4265,
+      "step": 1864
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7331037521362305,
+      "learning_rate": 0.00044504634152630024,
+      "loss": 3.5847,
+      "step": 1865
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7002983689308167,
+      "learning_rate": 0.0004449814831948168,
+      "loss": 3.5015,
+      "step": 1866
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7224550247192383,
+      "learning_rate": 0.00044491659134375587,
+      "loss": 3.6215,
+      "step": 1867
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6379685997962952,
+      "learning_rate": 0.0004448516659842733,
+      "loss": 3.5124,
+      "step": 1868
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6856171488761902,
+      "learning_rate": 0.00044478670712753043,
+      "loss": 3.3486,
+      "step": 1869
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.671360969543457,
+      "learning_rate": 0.0004447217147846944,
+      "loss": 3.4822,
+      "step": 1870
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7054650783538818,
+      "learning_rate": 0.0004446566889669382,
+      "loss": 3.5302,
+      "step": 1871
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7077500820159912,
+      "learning_rate": 0.00044459162968544055,
+      "loss": 3.5209,
+      "step": 1872
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7756421566009521,
+      "learning_rate": 0.00044452653695138585,
+      "loss": 3.5391,
+      "step": 1873
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7363382577896118,
+      "learning_rate": 0.00044446141077596424,
+      "loss": 3.6557,
+      "step": 1874
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6910688877105713,
+      "learning_rate": 0.00044439625117037183,
+      "loss": 3.3939,
+      "step": 1875
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6563981771469116,
+      "learning_rate": 0.0004443310581458102,
+      "loss": 3.4331,
+      "step": 1876
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6942573189735413,
+      "learning_rate": 0.00044426583171348666,
+      "loss": 3.463,
+      "step": 1877
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7408201098442078,
+      "learning_rate": 0.0004442005718846147,
+      "loss": 3.5645,
+      "step": 1878
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7020135521888733,
+      "learning_rate": 0.00044413527867041293,
+      "loss": 3.5329,
+      "step": 1879
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7310847640037537,
+      "learning_rate": 0.0004440699520821062,
+      "loss": 3.4169,
+      "step": 1880
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7448086142539978,
+      "learning_rate": 0.00044400459213092487,
+      "loss": 3.4812,
+      "step": 1881
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6619753837585449,
+      "learning_rate": 0.0004439391988281051,
+      "loss": 3.5827,
+      "step": 1882
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6850958466529846,
+      "learning_rate": 0.00044387377218488863,
+      "loss": 3.4599,
+      "step": 1883
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7675847411155701,
+      "learning_rate": 0.00044380831221252316,
+      "loss": 3.4061,
+      "step": 1884
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7381026148796082,
+      "learning_rate": 0.00044374281892226204,
+      "loss": 3.4934,
+      "step": 1885
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7035375833511353,
+      "learning_rate": 0.00044367729232536423,
+      "loss": 3.7558,
+      "step": 1886
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6353806853294373,
+      "learning_rate": 0.0004436117324330946,
+      "loss": 3.4913,
+      "step": 1887
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.687185525894165,
+      "learning_rate": 0.0004435461392567236,
+      "loss": 3.6783,
+      "step": 1888
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6932180523872375,
+      "learning_rate": 0.00044348051280752756,
+      "loss": 3.4155,
+      "step": 1889
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6853898763656616,
+      "learning_rate": 0.0004434148530967883,
+      "loss": 3.3687,
+      "step": 1890
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6757237911224365,
+      "learning_rate": 0.0004433491601357935,
+      "loss": 3.5026,
+      "step": 1891
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.653641402721405,
+      "learning_rate": 0.0004432834339358367,
+      "loss": 3.3618,
+      "step": 1892
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6783159971237183,
+      "learning_rate": 0.00044321767450821673,
+      "loss": 3.5522,
+      "step": 1893
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6962612867355347,
+      "learning_rate": 0.0004431518818642386,
+      "loss": 3.3733,
+      "step": 1894
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7038624286651611,
+      "learning_rate": 0.0004430860560152128,
+      "loss": 3.4771,
+      "step": 1895
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6692087650299072,
+      "learning_rate": 0.00044302019697245546,
+      "loss": 3.5181,
+      "step": 1896
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6483880877494812,
+      "learning_rate": 0.0004429543047472886,
+      "loss": 3.5177,
+      "step": 1897
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.687965452671051,
+      "learning_rate": 0.0004428883793510399,
+      "loss": 3.4136,
+      "step": 1898
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7006435394287109,
+      "learning_rate": 0.0004428224207950425,
+      "loss": 3.4867,
+      "step": 1899
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7762485146522522,
+      "learning_rate": 0.0004427564290906357,
+      "loss": 3.4821,
+      "step": 1900
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.672616183757782,
+      "learning_rate": 0.00044269040424916407,
+      "loss": 3.5393,
+      "step": 1901
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.75141841173172,
+      "learning_rate": 0.00044262434628197805,
+      "loss": 3.6058,
+      "step": 1902
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6640050411224365,
+      "learning_rate": 0.00044255825520043393,
+      "loss": 3.4241,
+      "step": 1903
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6671122908592224,
+      "learning_rate": 0.00044249213101589323,
+      "loss": 3.3735,
+      "step": 1904
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6726486086845398,
+      "learning_rate": 0.0004424259737397238,
+      "loss": 3.4756,
+      "step": 1905
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6656337380409241,
+      "learning_rate": 0.00044235978338329863,
+      "loss": 3.4923,
+      "step": 1906
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6552636623382568,
+      "learning_rate": 0.0004422935599579967,
+      "loss": 3.4121,
+      "step": 1907
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7019140720367432,
+      "learning_rate": 0.00044222730347520255,
+      "loss": 3.4499,
+      "step": 1908
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6775514483451843,
+      "learning_rate": 0.0004421610139463064,
+      "loss": 3.3372,
+      "step": 1909
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6777592301368713,
+      "learning_rate": 0.0004420946913827043,
+      "loss": 3.4424,
+      "step": 1910
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7406557202339172,
+      "learning_rate": 0.00044202833579579776,
+      "loss": 3.3874,
+      "step": 1911
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6490644216537476,
+      "learning_rate": 0.0004419619471969941,
+      "loss": 3.3859,
+      "step": 1912
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7046037316322327,
+      "learning_rate": 0.00044189552559770635,
+      "loss": 3.5542,
+      "step": 1913
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.681759774684906,
+      "learning_rate": 0.0004418290710093531,
+      "loss": 3.4768,
+      "step": 1914
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7443521618843079,
+      "learning_rate": 0.0004417625834433586,
+      "loss": 3.6244,
+      "step": 1915
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6745380163192749,
+      "learning_rate": 0.00044169606291115295,
+      "loss": 3.3206,
+      "step": 1916
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7358022332191467,
+      "learning_rate": 0.0004416295094241718,
+      "loss": 3.3504,
+      "step": 1917
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7005290985107422,
+      "learning_rate": 0.00044156292299385636,
+      "loss": 3.5177,
+      "step": 1918
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6522749066352844,
+      "learning_rate": 0.0004414963036316537,
+      "loss": 3.504,
+      "step": 1919
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7436257600784302,
+      "learning_rate": 0.00044142965134901635,
+      "loss": 3.5254,
+      "step": 1920
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.682768702507019,
+      "learning_rate": 0.0004413629661574028,
+      "loss": 3.1955,
+      "step": 1921
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.715193510055542,
+      "learning_rate": 0.00044129624806827684,
+      "loss": 3.4625,
+      "step": 1922
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7119811773300171,
+      "learning_rate": 0.00044122949709310817,
+      "loss": 3.5076,
+      "step": 1923
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6985951066017151,
+      "learning_rate": 0.00044116271324337196,
+      "loss": 3.4379,
+      "step": 1924
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.644829273223877,
+      "learning_rate": 0.00044109589653054925,
+      "loss": 3.5178,
+      "step": 1925
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7521076798439026,
+      "learning_rate": 0.0004410290469661266,
+      "loss": 3.5462,
+      "step": 1926
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7019145488739014,
+      "learning_rate": 0.0004409621645615961,
+      "loss": 3.4109,
+      "step": 1927
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.731626033782959,
+      "learning_rate": 0.0004408952493284557,
+      "loss": 3.548,
+      "step": 1928
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6920309066772461,
+      "learning_rate": 0.00044082830127820897,
+      "loss": 3.5405,
+      "step": 1929
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7317379713058472,
+      "learning_rate": 0.0004407613204223648,
+      "loss": 3.4448,
+      "step": 1930
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7225735187530518,
+      "learning_rate": 0.00044069430677243834,
+      "loss": 3.4579,
+      "step": 1931
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6989052891731262,
+      "learning_rate": 0.0004406272603399497,
+      "loss": 3.4951,
+      "step": 1932
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.706935703754425,
+      "learning_rate": 0.00044056018113642514,
+      "loss": 3.4935,
+      "step": 1933
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6633912920951843,
+      "learning_rate": 0.00044049306917339626,
+      "loss": 3.5158,
+      "step": 1934
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6701186895370483,
+      "learning_rate": 0.00044042592446240044,
+      "loss": 3.4929,
+      "step": 1935
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6783208250999451,
+      "learning_rate": 0.0004403587470149806,
+      "loss": 3.4407,
+      "step": 1936
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8055753707885742,
+      "learning_rate": 0.00044029153684268526,
+      "loss": 3.3577,
+      "step": 1937
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6336120963096619,
+      "learning_rate": 0.0004402242939570687,
+      "loss": 3.3685,
+      "step": 1938
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6730412840843201,
+      "learning_rate": 0.00044015701836969075,
+      "loss": 3.5368,
+      "step": 1939
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7138683795928955,
+      "learning_rate": 0.00044008971009211684,
+      "loss": 3.5457,
+      "step": 1940
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.759942889213562,
+      "learning_rate": 0.0004400223691359181,
+      "loss": 3.4003,
+      "step": 1941
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.752402126789093,
+      "learning_rate": 0.00043995499551267115,
+      "loss": 3.4526,
+      "step": 1942
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6543063521385193,
+      "learning_rate": 0.0004398875892339583,
+      "loss": 3.5303,
+      "step": 1943
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7147708535194397,
+      "learning_rate": 0.0004398201503113675,
+      "loss": 3.4889,
+      "step": 1944
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6864874958992004,
+      "learning_rate": 0.0004397526787564923,
+      "loss": 3.5839,
+      "step": 1945
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7960362434387207,
+      "learning_rate": 0.00043968517458093184,
+      "loss": 3.5179,
+      "step": 1946
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7662225961685181,
+      "learning_rate": 0.0004396176377962908,
+      "loss": 3.5034,
+      "step": 1947
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6254611015319824,
+      "learning_rate": 0.0004395500684141797,
+      "loss": 3.3934,
+      "step": 1948
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6561282873153687,
+      "learning_rate": 0.00043948246644621427,
+      "loss": 3.5359,
+      "step": 1949
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6546863317489624,
+      "learning_rate": 0.0004394148319040162,
+      "loss": 3.464,
+      "step": 1950
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6724408864974976,
+      "learning_rate": 0.00043934716479921267,
+      "loss": 3.4396,
+      "step": 1951
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6835522055625916,
+      "learning_rate": 0.00043927946514343635,
+      "loss": 3.4584,
+      "step": 1952
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7021266222000122,
+      "learning_rate": 0.00043921173294832574,
+      "loss": 3.3387,
+      "step": 1953
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7060502767562866,
+      "learning_rate": 0.0004391439682255247,
+      "loss": 3.4091,
+      "step": 1954
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7331516742706299,
+      "learning_rate": 0.0004390761709866827,
+      "loss": 3.5226,
+      "step": 1955
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7018919587135315,
+      "learning_rate": 0.000439008341243455,
+      "loss": 3.5414,
+      "step": 1956
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8398563861846924,
+      "learning_rate": 0.00043894047900750225,
+      "loss": 3.4953,
+      "step": 1957
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6954603791236877,
+      "learning_rate": 0.0004388725842904908,
+      "loss": 3.5214,
+      "step": 1958
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6625902056694031,
+      "learning_rate": 0.00043880465710409243,
+      "loss": 3.4665,
+      "step": 1959
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6996656060218811,
+      "learning_rate": 0.0004387366974599848,
+      "loss": 3.4562,
+      "step": 1960
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7215577363967896,
+      "learning_rate": 0.0004386687053698508,
+      "loss": 3.4082,
+      "step": 1961
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7131018042564392,
+      "learning_rate": 0.0004386006808453792,
+      "loss": 3.436,
+      "step": 1962
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7280438542366028,
+      "learning_rate": 0.000438532623898264,
+      "loss": 3.3886,
+      "step": 1963
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7057799696922302,
+      "learning_rate": 0.00043846453454020513,
+      "loss": 3.4871,
+      "step": 1964
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7401341795921326,
+      "learning_rate": 0.00043839641278290787,
+      "loss": 3.5752,
+      "step": 1965
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7227087616920471,
+      "learning_rate": 0.0004383282586380832,
+      "loss": 3.5417,
+      "step": 1966
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.778991162776947,
+      "learning_rate": 0.0004382600721174477,
+      "loss": 3.493,
+      "step": 1967
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6657295227050781,
+      "learning_rate": 0.00043819185323272313,
+      "loss": 3.4669,
+      "step": 1968
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6682909727096558,
+      "learning_rate": 0.0004381236019956374,
+      "loss": 3.3524,
+      "step": 1969
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6612628102302551,
+      "learning_rate": 0.00043805531841792345,
+      "loss": 3.5043,
+      "step": 1970
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7476475238800049,
+      "learning_rate": 0.0004379870025113203,
+      "loss": 3.5374,
+      "step": 1971
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6901301741600037,
+      "learning_rate": 0.00043791865428757196,
+      "loss": 3.3478,
+      "step": 1972
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6781008243560791,
+      "learning_rate": 0.00043785027375842846,
+      "loss": 3.4154,
+      "step": 1973
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7976479530334473,
+      "learning_rate": 0.0004377818609356451,
+      "loss": 3.4864,
+      "step": 1974
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6746860146522522,
+      "learning_rate": 0.00043771341583098293,
+      "loss": 3.3971,
+      "step": 1975
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6861461400985718,
+      "learning_rate": 0.00043764493845620847,
+      "loss": 3.5625,
+      "step": 1976
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6744572520256042,
+      "learning_rate": 0.00043757642882309364,
+      "loss": 3.4734,
+      "step": 1977
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.684846818447113,
+      "learning_rate": 0.00043750788694341613,
+      "loss": 3.388,
+      "step": 1978
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6695400476455688,
+      "learning_rate": 0.0004374393128289591,
+      "loss": 3.5149,
+      "step": 1979
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6745089292526245,
+      "learning_rate": 0.0004373707064915112,
+      "loss": 3.376,
+      "step": 1980
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7593763470649719,
+      "learning_rate": 0.0004373020679428667,
+      "loss": 3.5023,
+      "step": 1981
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6825028657913208,
+      "learning_rate": 0.0004372333971948253,
+      "loss": 3.5532,
+      "step": 1982
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7344127893447876,
+      "learning_rate": 0.0004371646942591924,
+      "loss": 3.3933,
+      "step": 1983
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7311540246009827,
+      "learning_rate": 0.00043709595914777865,
+      "loss": 3.3001,
+      "step": 1984
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7440100312232971,
+      "learning_rate": 0.0004370271918724006,
+      "loss": 3.5478,
+      "step": 1985
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7262304425239563,
+      "learning_rate": 0.00043695839244488,
+      "loss": 3.432,
+      "step": 1986
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7508566379547119,
+      "learning_rate": 0.00043688956087704434,
+      "loss": 3.451,
+      "step": 1987
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.701371431350708,
+      "learning_rate": 0.0004368206971807266,
+      "loss": 3.4946,
+      "step": 1988
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6847150921821594,
+      "learning_rate": 0.00043675180136776515,
+      "loss": 3.3828,
+      "step": 1989
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7214049100875854,
+      "learning_rate": 0.00043668287345000403,
+      "loss": 3.5557,
+      "step": 1990
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6651790142059326,
+      "learning_rate": 0.0004366139134392928,
+      "loss": 3.3798,
+      "step": 1991
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7046433091163635,
+      "learning_rate": 0.00043654492134748634,
+      "loss": 3.4396,
+      "step": 1992
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7146850824356079,
+      "learning_rate": 0.00043647589718644544,
+      "loss": 3.4823,
+      "step": 1993
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6687988042831421,
+      "learning_rate": 0.00043640684096803585,
+      "loss": 3.4075,
+      "step": 1994
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6961211562156677,
+      "learning_rate": 0.0004363377527041294,
+      "loss": 3.408,
+      "step": 1995
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7626001238822937,
+      "learning_rate": 0.00043626863240660296,
+      "loss": 3.6133,
+      "step": 1996
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6749371886253357,
+      "learning_rate": 0.0004361994800873392,
+      "loss": 3.366,
+      "step": 1997
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.74397212266922,
+      "learning_rate": 0.00043613029575822625,
+      "loss": 3.4656,
+      "step": 1998
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6702881455421448,
+      "learning_rate": 0.0004360610794311577,
+      "loss": 3.5005,
+      "step": 1999
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6869195699691772,
+      "learning_rate": 0.0004359918311180325,
+      "loss": 3.6151,
+      "step": 2000
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7465648055076599,
+      "learning_rate": 0.0004359225508307554,
+      "loss": 3.5636,
+      "step": 2001
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6688586473464966,
+      "learning_rate": 0.00043585323858123635,
+      "loss": 3.5731,
+      "step": 2002
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6723427176475525,
+      "learning_rate": 0.000435783894381391,
+      "loss": 3.5229,
+      "step": 2003
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6933841109275818,
+      "learning_rate": 0.0004357145182431405,
+      "loss": 3.504,
+      "step": 2004
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6604121327400208,
+      "learning_rate": 0.00043564511017841123,
+      "loss": 3.3467,
+      "step": 2005
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6695713400840759,
+      "learning_rate": 0.00043557567019913534,
+      "loss": 3.5138,
+      "step": 2006
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7155663371086121,
+      "learning_rate": 0.00043550619831725037,
+      "loss": 3.4378,
+      "step": 2007
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7241122722625732,
+      "learning_rate": 0.00043543669454469935,
+      "loss": 3.391,
+      "step": 2008
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6676546335220337,
+      "learning_rate": 0.00043536715889343084,
+      "loss": 3.4446,
+      "step": 2009
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7022714614868164,
+      "learning_rate": 0.0004352975913753987,
+      "loss": 3.4875,
+      "step": 2010
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6732816696166992,
+      "learning_rate": 0.0004352279920025624,
+      "loss": 3.3945,
+      "step": 2011
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7686198949813843,
+      "learning_rate": 0.00043515836078688696,
+      "loss": 3.3351,
+      "step": 2012
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7455644011497498,
+      "learning_rate": 0.00043508869774034275,
+      "loss": 3.5383,
+      "step": 2013
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7511581778526306,
+      "learning_rate": 0.0004350190028749057,
+      "loss": 3.4616,
+      "step": 2014
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6484218835830688,
+      "learning_rate": 0.00043494927620255715,
+      "loss": 3.4581,
+      "step": 2015
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7067961096763611,
+      "learning_rate": 0.0004348795177352839,
+      "loss": 3.4788,
+      "step": 2016
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6927537322044373,
+      "learning_rate": 0.0004348097274850782,
+      "loss": 3.4899,
+      "step": 2017
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6653082370758057,
+      "learning_rate": 0.00043473990546393784,
+      "loss": 3.5604,
+      "step": 2018
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7367866039276123,
+      "learning_rate": 0.0004346700516838661,
+      "loss": 3.4888,
+      "step": 2019
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7152118682861328,
+      "learning_rate": 0.0004346001661568715,
+      "loss": 3.5531,
+      "step": 2020
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6760360598564148,
+      "learning_rate": 0.0004345302488949684,
+      "loss": 3.4032,
+      "step": 2021
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6933254599571228,
+      "learning_rate": 0.0004344602999101761,
+      "loss": 3.5824,
+      "step": 2022
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6791170239448547,
+      "learning_rate": 0.00043439031921451994,
+      "loss": 3.299,
+      "step": 2023
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6987645030021667,
+      "learning_rate": 0.0004343203068200302,
+      "loss": 3.4105,
+      "step": 2024
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6751917004585266,
+      "learning_rate": 0.0004342502627387429,
+      "loss": 3.4956,
+      "step": 2025
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7118316888809204,
+      "learning_rate": 0.00043418018698269945,
+      "loss": 3.4484,
+      "step": 2026
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7071229815483093,
+      "learning_rate": 0.00043411007956394666,
+      "loss": 3.3642,
+      "step": 2027
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6810958981513977,
+      "learning_rate": 0.0004340399404945368,
+      "loss": 3.5119,
+      "step": 2028
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7358177900314331,
+      "learning_rate": 0.0004339697697865276,
+      "loss": 3.4273,
+      "step": 2029
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7156998515129089,
+      "learning_rate": 0.0004338995674519822,
+      "loss": 3.5882,
+      "step": 2030
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7202603220939636,
+      "learning_rate": 0.00043382933350296916,
+      "loss": 3.5176,
+      "step": 2031
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7155426740646362,
+      "learning_rate": 0.0004337590679515626,
+      "loss": 3.4288,
+      "step": 2032
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7273238301277161,
+      "learning_rate": 0.00043368877080984195,
+      "loss": 3.5389,
+      "step": 2033
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7046866416931152,
+      "learning_rate": 0.0004336184420898921,
+      "loss": 3.4743,
+      "step": 2034
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.724859356880188,
+      "learning_rate": 0.00043354808180380333,
+      "loss": 3.4958,
+      "step": 2035
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7297008633613586,
+      "learning_rate": 0.0004334776899636714,
+      "loss": 3.5521,
+      "step": 2036
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7214579582214355,
+      "learning_rate": 0.00043340726658159764,
+      "loss": 3.3943,
+      "step": 2037
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.682220458984375,
+      "learning_rate": 0.0004333368116696884,
+      "loss": 3.4364,
+      "step": 2038
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6749826073646545,
+      "learning_rate": 0.00043326632524005583,
+      "loss": 3.5132,
+      "step": 2039
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6864546537399292,
+      "learning_rate": 0.0004331958073048174,
+      "loss": 3.568,
+      "step": 2040
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7034383416175842,
+      "learning_rate": 0.0004331252578760959,
+      "loss": 3.5975,
+      "step": 2041
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7162829637527466,
+      "learning_rate": 0.00043305467696601953,
+      "loss": 3.4938,
+      "step": 2042
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6995472311973572,
+      "learning_rate": 0.00043298406458672207,
+      "loss": 3.5466,
+      "step": 2043
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.672080397605896,
+      "learning_rate": 0.00043291342075034255,
+      "loss": 3.3307,
+      "step": 2044
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6504019498825073,
+      "learning_rate": 0.00043284274546902555,
+      "loss": 3.497,
+      "step": 2045
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6735064387321472,
+      "learning_rate": 0.00043277203875492087,
+      "loss": 3.4493,
+      "step": 2046
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7060967087745667,
+      "learning_rate": 0.0004327013006201839,
+      "loss": 3.554,
+      "step": 2047
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6433796882629395,
+      "learning_rate": 0.00043263053107697524,
+      "loss": 3.5399,
+      "step": 2048
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6759264469146729,
+      "learning_rate": 0.0004325597301374611,
+      "loss": 3.5371,
+      "step": 2049
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6785297393798828,
+      "learning_rate": 0.00043248889781381285,
+      "loss": 3.4896,
+      "step": 2050
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6828736662864685,
+      "learning_rate": 0.00043241803411820756,
+      "loss": 3.3398,
+      "step": 2051
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.699984610080719,
+      "learning_rate": 0.0004323471390628274,
+      "loss": 3.3404,
+      "step": 2052
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6868505477905273,
+      "learning_rate": 0.00043227621265986004,
+      "loss": 3.4911,
+      "step": 2053
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6925286054611206,
+      "learning_rate": 0.0004322052549214987,
+      "loss": 3.3701,
+      "step": 2054
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7424176931381226,
+      "learning_rate": 0.0004321342658599416,
+      "loss": 3.4955,
+      "step": 2055
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6669638156890869,
+      "learning_rate": 0.0004320632454873929,
+      "loss": 3.3172,
+      "step": 2056
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6610919237136841,
+      "learning_rate": 0.00043199219381606153,
+      "loss": 3.5237,
+      "step": 2057
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6433104276657104,
+      "learning_rate": 0.0004319211108581622,
+      "loss": 3.4581,
+      "step": 2058
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6824292540550232,
+      "learning_rate": 0.000431849996625915,
+      "loss": 3.5064,
+      "step": 2059
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6911894679069519,
+      "learning_rate": 0.00043177885113154503,
+      "loss": 3.4415,
+      "step": 2060
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7084999084472656,
+      "learning_rate": 0.0004317076743872833,
+      "loss": 3.4936,
+      "step": 2061
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7088779211044312,
+      "learning_rate": 0.0004316364664053658,
+      "loss": 3.4208,
+      "step": 2062
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6912571787834167,
+      "learning_rate": 0.000431565227198034,
+      "loss": 3.5643,
+      "step": 2063
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6518218517303467,
+      "learning_rate": 0.0004314939567775347,
+      "loss": 3.4999,
+      "step": 2064
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.758113384246826,
+      "learning_rate": 0.0004314226551561202,
+      "loss": 3.4841,
+      "step": 2065
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6662098169326782,
+      "learning_rate": 0.00043135132234604814,
+      "loss": 3.4227,
+      "step": 2066
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7417008876800537,
+      "learning_rate": 0.0004312799583595813,
+      "loss": 3.5523,
+      "step": 2067
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7159901857376099,
+      "learning_rate": 0.00043120856320898806,
+      "loss": 3.4275,
+      "step": 2068
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7324928045272827,
+      "learning_rate": 0.000431137136906542,
+      "loss": 3.5198,
+      "step": 2069
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7330513000488281,
+      "learning_rate": 0.00043106567946452225,
+      "loss": 3.6361,
+      "step": 2070
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7240403294563293,
+      "learning_rate": 0.0004309941908952131,
+      "loss": 3.4916,
+      "step": 2071
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7005897760391235,
+      "learning_rate": 0.0004309226712109042,
+      "loss": 3.4402,
+      "step": 2072
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7218866348266602,
+      "learning_rate": 0.00043085112042389075,
+      "loss": 3.5906,
+      "step": 2073
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7103626132011414,
+      "learning_rate": 0.000430779538546473,
+      "loss": 3.5515,
+      "step": 2074
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7686103582382202,
+      "learning_rate": 0.0004307079255909569,
+      "loss": 3.4956,
+      "step": 2075
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7192513942718506,
+      "learning_rate": 0.0004306362815696534,
+      "loss": 3.5604,
+      "step": 2076
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6870089173316956,
+      "learning_rate": 0.00043056460649487904,
+      "loss": 3.4542,
+      "step": 2077
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7580736875534058,
+      "learning_rate": 0.0004304929003789555,
+      "loss": 3.4443,
+      "step": 2078
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6816487908363342,
+      "learning_rate": 0.0004304211632342099,
+      "loss": 3.3623,
+      "step": 2079
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.724460780620575,
+      "learning_rate": 0.0004303493950729748,
+      "loss": 3.3841,
+      "step": 2080
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.719296395778656,
+      "learning_rate": 0.0004302775959075878,
+      "loss": 3.3878,
+      "step": 2081
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6661654114723206,
+      "learning_rate": 0.00043020576575039215,
+      "loss": 3.4579,
+      "step": 2082
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6661612391471863,
+      "learning_rate": 0.00043013390461373626,
+      "loss": 3.4368,
+      "step": 2083
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.683428168296814,
+      "learning_rate": 0.0004300620125099738,
+      "loss": 3.4621,
+      "step": 2084
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7038209438323975,
+      "learning_rate": 0.000429990089451464,
+      "loss": 3.5481,
+      "step": 2085
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6688196659088135,
+      "learning_rate": 0.0004299181354505712,
+      "loss": 3.4501,
+      "step": 2086
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6909682154655457,
+      "learning_rate": 0.00042984615051966515,
+      "loss": 3.4476,
+      "step": 2087
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.70419842004776,
+      "learning_rate": 0.00042977413467112084,
+      "loss": 3.6017,
+      "step": 2088
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8000498414039612,
+      "learning_rate": 0.00042970208791731857,
+      "loss": 3.413,
+      "step": 2089
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6451414227485657,
+      "learning_rate": 0.00042963001027064416,
+      "loss": 3.5047,
+      "step": 2090
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7211129069328308,
+      "learning_rate": 0.00042955790174348864,
+      "loss": 3.5806,
+      "step": 2091
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7346986532211304,
+      "learning_rate": 0.0004294857623482481,
+      "loss": 3.502,
+      "step": 2092
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7390275597572327,
+      "learning_rate": 0.0004294135920973242,
+      "loss": 3.4617,
+      "step": 2093
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7164575457572937,
+      "learning_rate": 0.000429341391003124,
+      "loss": 3.5179,
+      "step": 2094
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7850090861320496,
+      "learning_rate": 0.0004292691590780595,
+      "loss": 3.5023,
+      "step": 2095
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7475020289421082,
+      "learning_rate": 0.00042919689633454827,
+      "loss": 3.4549,
+      "step": 2096
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7047721147537231,
+      "learning_rate": 0.0004291246027850132,
+      "loss": 3.4779,
+      "step": 2097
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.656097948551178,
+      "learning_rate": 0.00042905227844188226,
+      "loss": 3.3747,
+      "step": 2098
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6438138484954834,
+      "learning_rate": 0.00042897992331758896,
+      "loss": 3.5208,
+      "step": 2099
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6644891500473022,
+      "learning_rate": 0.0004289075374245719,
+      "loss": 3.4167,
+      "step": 2100
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6867865324020386,
+      "learning_rate": 0.00042883512077527506,
+      "loss": 3.615,
+      "step": 2101
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7456568479537964,
+      "learning_rate": 0.0004287626733821477,
+      "loss": 3.4706,
+      "step": 2102
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7014803290367126,
+      "learning_rate": 0.0004286901952576445,
+      "loss": 3.5279,
+      "step": 2103
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6818939447402954,
+      "learning_rate": 0.00042861768641422505,
+      "loss": 3.4957,
+      "step": 2104
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6658956408500671,
+      "learning_rate": 0.0004285451468643546,
+      "loss": 3.5563,
+      "step": 2105
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6703715920448303,
+      "learning_rate": 0.0004284725766205037,
+      "loss": 3.45,
+      "step": 2106
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.697333574295044,
+      "learning_rate": 0.00042839997569514766,
+      "loss": 3.4967,
+      "step": 2107
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6819553375244141,
+      "learning_rate": 0.0004283273441007677,
+      "loss": 3.2826,
+      "step": 2108
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7194439768791199,
+      "learning_rate": 0.00042825468184984995,
+      "loss": 3.4084,
+      "step": 2109
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7378134727478027,
+      "learning_rate": 0.0004281819889548858,
+      "loss": 3.4858,
+      "step": 2110
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7612019777297974,
+      "learning_rate": 0.00042810926542837213,
+      "loss": 3.56,
+      "step": 2111
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6910973191261292,
+      "learning_rate": 0.000428036511282811,
+      "loss": 3.4338,
+      "step": 2112
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7197093963623047,
+      "learning_rate": 0.00042796372653070946,
+      "loss": 3.5613,
+      "step": 2113
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.673733651638031,
+      "learning_rate": 0.00042789091118458037,
+      "loss": 3.4107,
+      "step": 2114
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6490559577941895,
+      "learning_rate": 0.00042781806525694124,
+      "loss": 3.5269,
+      "step": 2115
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6586740612983704,
+      "learning_rate": 0.0004277451887603152,
+      "loss": 3.4372,
+      "step": 2116
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6464582085609436,
+      "learning_rate": 0.0004276722817072307,
+      "loss": 3.3763,
+      "step": 2117
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7137161493301392,
+      "learning_rate": 0.0004275993441102212,
+      "loss": 3.4224,
+      "step": 2118
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6621758341789246,
+      "learning_rate": 0.00042752637598182555,
+      "loss": 3.4509,
+      "step": 2119
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6240204572677612,
+      "learning_rate": 0.0004274533773345878,
+      "loss": 3.4812,
+      "step": 2120
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6391758918762207,
+      "learning_rate": 0.0004273803481810573,
+      "loss": 3.4246,
+      "step": 2121
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.702078104019165,
+      "learning_rate": 0.0004273072885337885,
+      "loss": 3.5051,
+      "step": 2122
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.767562985420227,
+      "learning_rate": 0.0004272341984053413,
+      "loss": 3.4407,
+      "step": 2123
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6934359073638916,
+      "learning_rate": 0.0004271610778082807,
+      "loss": 3.5191,
+      "step": 2124
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7259646058082581,
+      "learning_rate": 0.00042708792675517703,
+      "loss": 3.5403,
+      "step": 2125
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7107850909233093,
+      "learning_rate": 0.00042701474525860583,
+      "loss": 3.5209,
+      "step": 2126
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7186352014541626,
+      "learning_rate": 0.0004269415333311477,
+      "loss": 3.402,
+      "step": 2127
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6743739247322083,
+      "learning_rate": 0.0004268682909853888,
+      "loss": 3.4583,
+      "step": 2128
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7082521319389343,
+      "learning_rate": 0.0004267950182339201,
+      "loss": 3.4462,
+      "step": 2129
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7342261672019958,
+      "learning_rate": 0.0004267217150893383,
+      "loss": 3.4504,
+      "step": 2130
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.70721834897995,
+      "learning_rate": 0.0004266483815642449,
+      "loss": 3.4127,
+      "step": 2131
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6457509994506836,
+      "learning_rate": 0.00042657501767124685,
+      "loss": 3.3845,
+      "step": 2132
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7977797389030457,
+      "learning_rate": 0.0004265016234229563,
+      "loss": 3.4275,
+      "step": 2133
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7149450182914734,
+      "learning_rate": 0.00042642819883199033,
+      "loss": 3.5007,
+      "step": 2134
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6619107723236084,
+      "learning_rate": 0.0004263547439109717,
+      "loss": 3.3743,
+      "step": 2135
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6747405529022217,
+      "learning_rate": 0.0004262812586725282,
+      "loss": 3.5574,
+      "step": 2136
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7770102620124817,
+      "learning_rate": 0.00042620774312929265,
+      "loss": 3.5815,
+      "step": 2137
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.711304247379303,
+      "learning_rate": 0.0004261341972939033,
+      "loss": 3.4309,
+      "step": 2138
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6804766058921814,
+      "learning_rate": 0.0004260606211790035,
+      "loss": 3.3478,
+      "step": 2139
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.700182318687439,
+      "learning_rate": 0.0004259870147972419,
+      "loss": 3.4619,
+      "step": 2140
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6286213994026184,
+      "learning_rate": 0.0004259133781612722,
+      "loss": 3.5719,
+      "step": 2141
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7141399383544922,
+      "learning_rate": 0.0004258397112837534,
+      "loss": 3.5161,
+      "step": 2142
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6732556819915771,
+      "learning_rate": 0.0004257660141773497,
+      "loss": 3.4989,
+      "step": 2143
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6593499183654785,
+      "learning_rate": 0.0004256922868547306,
+      "loss": 3.3962,
+      "step": 2144
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6658518314361572,
+      "learning_rate": 0.00042561852932857045,
+      "loss": 3.4766,
+      "step": 2145
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6705014705657959,
+      "learning_rate": 0.00042554474161154933,
+      "loss": 3.4349,
+      "step": 2146
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6827144622802734,
+      "learning_rate": 0.000425470923716352,
+      "loss": 3.484,
+      "step": 2147
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7057637572288513,
+      "learning_rate": 0.0004253970756556685,
+      "loss": 3.3785,
+      "step": 2148
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6769725680351257,
+      "learning_rate": 0.0004253231974421945,
+      "loss": 3.4597,
+      "step": 2149
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7392288446426392,
+      "learning_rate": 0.00042524928908863025,
+      "loss": 3.4506,
+      "step": 2150
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6754869222640991,
+      "learning_rate": 0.0004251753506076816,
+      "loss": 3.4491,
+      "step": 2151
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.688669741153717,
+      "learning_rate": 0.00042510138201205935,
+      "loss": 3.622,
+      "step": 2152
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7528844475746155,
+      "learning_rate": 0.0004250273833144795,
+      "loss": 3.5109,
+      "step": 2153
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7025592923164368,
+      "learning_rate": 0.00042495335452766346,
+      "loss": 3.5784,
+      "step": 2154
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.678760290145874,
+      "learning_rate": 0.0004248792956643376,
+      "loss": 3.5936,
+      "step": 2155
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7542569041252136,
+      "learning_rate": 0.00042480520673723334,
+      "loss": 3.6332,
+      "step": 2156
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7015278935432434,
+      "learning_rate": 0.0004247310877590875,
+      "loss": 3.3464,
+      "step": 2157
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7442424893379211,
+      "learning_rate": 0.00042465693874264203,
+      "loss": 3.4356,
+      "step": 2158
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7039641737937927,
+      "learning_rate": 0.00042458275970064404,
+      "loss": 3.3722,
+      "step": 2159
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6853299736976624,
+      "learning_rate": 0.0004245085506458457,
+      "loss": 3.4757,
+      "step": 2160
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6797823309898376,
+      "learning_rate": 0.0004244343115910044,
+      "loss": 3.4409,
+      "step": 2161
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7026341557502747,
+      "learning_rate": 0.00042436004254888286,
+      "loss": 3.4307,
+      "step": 2162
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7010685205459595,
+      "learning_rate": 0.00042428574353224846,
+      "loss": 3.3267,
+      "step": 2163
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6813393831253052,
+      "learning_rate": 0.0004242114145538744,
+      "loss": 3.3881,
+      "step": 2164
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7339474558830261,
+      "learning_rate": 0.00042413705562653847,
+      "loss": 3.4078,
+      "step": 2165
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6485927104949951,
+      "learning_rate": 0.0004240626667630239,
+      "loss": 3.4319,
+      "step": 2166
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6844644546508789,
+      "learning_rate": 0.000423988247976119,
+      "loss": 3.412,
+      "step": 2167
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6630472540855408,
+      "learning_rate": 0.0004239137992786173,
+      "loss": 3.4041,
+      "step": 2168
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.715288519859314,
+      "learning_rate": 0.00042383932068331727,
+      "loss": 3.4277,
+      "step": 2169
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7145828604698181,
+      "learning_rate": 0.0004237648122030227,
+      "loss": 3.407,
+      "step": 2170
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7224670648574829,
+      "learning_rate": 0.00042369027385054245,
+      "loss": 3.501,
+      "step": 2171
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7074441909790039,
+      "learning_rate": 0.00042361570563869056,
+      "loss": 3.5399,
+      "step": 2172
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6857562065124512,
+      "learning_rate": 0.00042354110758028614,
+      "loss": 3.4285,
+      "step": 2173
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6324169039726257,
+      "learning_rate": 0.00042346647968815346,
+      "loss": 3.311,
+      "step": 2174
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7110226154327393,
+      "learning_rate": 0.00042339182197512193,
+      "loss": 3.5438,
+      "step": 2175
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6994052529335022,
+      "learning_rate": 0.000423317134454026,
+      "loss": 3.5169,
+      "step": 2176
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6708390712738037,
+      "learning_rate": 0.0004232424171377055,
+      "loss": 3.3347,
+      "step": 2177
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.681258499622345,
+      "learning_rate": 0.00042316767003900503,
+      "loss": 3.5251,
+      "step": 2178
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7101171612739563,
+      "learning_rate": 0.0004230928931707746,
+      "loss": 3.453,
+      "step": 2179
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.66106778383255,
+      "learning_rate": 0.00042301808654586915,
+      "loss": 3.4855,
+      "step": 2180
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7216284871101379,
+      "learning_rate": 0.00042294325017714875,
+      "loss": 3.5795,
+      "step": 2181
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6846936345100403,
+      "learning_rate": 0.00042286838407747877,
+      "loss": 3.603,
+      "step": 2182
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6562172174453735,
+      "learning_rate": 0.00042279348825972955,
+      "loss": 3.3686,
+      "step": 2183
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6876118779182434,
+      "learning_rate": 0.0004227185627367765,
+      "loss": 3.3676,
+      "step": 2184
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6920157670974731,
+      "learning_rate": 0.0004226436075215001,
+      "loss": 3.4078,
+      "step": 2185
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6572172045707703,
+      "learning_rate": 0.0004225686226267862,
+      "loss": 3.4962,
+      "step": 2186
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6494218707084656,
+      "learning_rate": 0.0004224936080655255,
+      "loss": 3.4851,
+      "step": 2187
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7011899948120117,
+      "learning_rate": 0.0004224185638506138,
+      "loss": 3.4603,
+      "step": 2188
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6653608083724976,
+      "learning_rate": 0.0004223434899949521,
+      "loss": 3.4034,
+      "step": 2189
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6927925944328308,
+      "learning_rate": 0.0004222683865114465,
+      "loss": 3.4491,
+      "step": 2190
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.650065004825592,
+      "learning_rate": 0.0004221932534130082,
+      "loss": 3.4603,
+      "step": 2191
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6736721992492676,
+      "learning_rate": 0.00042211809071255344,
+      "loss": 3.3901,
+      "step": 2192
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6661523580551147,
+      "learning_rate": 0.00042204289842300344,
+      "loss": 3.4231,
+      "step": 2193
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7264108061790466,
+      "learning_rate": 0.00042196767655728473,
+      "loss": 3.62,
+      "step": 2194
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6918498277664185,
+      "learning_rate": 0.00042189242512832875,
+      "loss": 3.5098,
+      "step": 2195
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7076089978218079,
+      "learning_rate": 0.00042181714414907224,
+      "loss": 3.4579,
+      "step": 2196
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6871483325958252,
+      "learning_rate": 0.00042174183363245674,
+      "loss": 3.4601,
+      "step": 2197
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6901012063026428,
+      "learning_rate": 0.0004216664935914291,
+      "loss": 3.5506,
+      "step": 2198
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6661196947097778,
+      "learning_rate": 0.0004215911240389409,
+      "loss": 3.3883,
+      "step": 2199
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6737020611763,
+      "learning_rate": 0.0004215157249879493,
+      "loss": 3.4412,
+      "step": 2200
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6551852822303772,
+      "learning_rate": 0.0004214402964514163,
+      "loss": 3.3396,
+      "step": 2201
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.682401180267334,
+      "learning_rate": 0.00042136483844230877,
+      "loss": 3.4852,
+      "step": 2202
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6794586777687073,
+      "learning_rate": 0.0004212893509735989,
+      "loss": 3.3587,
+      "step": 2203
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6415001153945923,
+      "learning_rate": 0.00042121383405826373,
+      "loss": 3.4704,
+      "step": 2204
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6676735877990723,
+      "learning_rate": 0.00042113828770928575,
+      "loss": 3.4323,
+      "step": 2205
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7048613429069519,
+      "learning_rate": 0.000421062711939652,
+      "loss": 3.3844,
+      "step": 2206
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6825292706489563,
+      "learning_rate": 0.000420987106762355,
+      "loss": 3.5218,
+      "step": 2207
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6771185398101807,
+      "learning_rate": 0.00042091147219039203,
+      "loss": 3.3899,
+      "step": 2208
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6721455454826355,
+      "learning_rate": 0.0004208358082367657,
+      "loss": 3.4341,
+      "step": 2209
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7785545587539673,
+      "learning_rate": 0.0004207601149144834,
+      "loss": 3.516,
+      "step": 2210
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7200364470481873,
+      "learning_rate": 0.00042068439223655774,
+      "loss": 3.4922,
+      "step": 2211
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6667042374610901,
+      "learning_rate": 0.0004206086402160063,
+      "loss": 3.3768,
+      "step": 2212
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6721999645233154,
+      "learning_rate": 0.0004205328588658517,
+      "loss": 3.4743,
+      "step": 2213
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6536881923675537,
+      "learning_rate": 0.0004204570481991217,
+      "loss": 3.3868,
+      "step": 2214
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6732337474822998,
+      "learning_rate": 0.00042038120822884904,
+      "loss": 3.3522,
+      "step": 2215
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6934797167778015,
+      "learning_rate": 0.00042030533896807143,
+      "loss": 3.5299,
+      "step": 2216
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6694208383560181,
+      "learning_rate": 0.0004202294404298317,
+      "loss": 3.4935,
+      "step": 2217
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6776825189590454,
+      "learning_rate": 0.0004201535126271777,
+      "loss": 3.4775,
+      "step": 2218
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6322728395462036,
+      "learning_rate": 0.0004200775555731623,
+      "loss": 3.496,
+      "step": 2219
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6704550981521606,
+      "learning_rate": 0.00042000156928084336,
+      "loss": 3.5436,
+      "step": 2220
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6826850175857544,
+      "learning_rate": 0.00041992555376328385,
+      "loss": 3.577,
+      "step": 2221
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6952018737792969,
+      "learning_rate": 0.00041984950903355166,
+      "loss": 3.4282,
+      "step": 2222
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6880483031272888,
+      "learning_rate": 0.0004197734351047199,
+      "loss": 3.4562,
+      "step": 2223
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7053331732749939,
+      "learning_rate": 0.0004196973319898664,
+      "loss": 3.4515,
+      "step": 2224
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6839045882225037,
+      "learning_rate": 0.0004196211997020742,
+      "loss": 3.5339,
+      "step": 2225
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6083318591117859,
+      "learning_rate": 0.0004195450382544315,
+      "loss": 3.4035,
+      "step": 2226
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6685528755187988,
+      "learning_rate": 0.000419468847660031,
+      "loss": 3.447,
+      "step": 2227
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6794456243515015,
+      "learning_rate": 0.00041939262793197105,
+      "loss": 3.3722,
+      "step": 2228
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6989598274230957,
+      "learning_rate": 0.00041931637908335453,
+      "loss": 3.4522,
+      "step": 2229
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7004005312919617,
+      "learning_rate": 0.0004192401011272896,
+      "loss": 3.2647,
+      "step": 2230
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6671735048294067,
+      "learning_rate": 0.0004191637940768893,
+      "loss": 3.4905,
+      "step": 2231
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7051229476928711,
+      "learning_rate": 0.0004190874579452716,
+      "loss": 3.3168,
+      "step": 2232
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6781712174415588,
+      "learning_rate": 0.0004190110927455597,
+      "loss": 3.4034,
+      "step": 2233
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7480825185775757,
+      "learning_rate": 0.0004189346984908816,
+      "loss": 3.3314,
+      "step": 2234
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7438943386077881,
+      "learning_rate": 0.00041885827519437047,
+      "loss": 3.4312,
+      "step": 2235
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6712598204612732,
+      "learning_rate": 0.0004187818228691641,
+      "loss": 3.5298,
+      "step": 2236
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6709827184677124,
+      "learning_rate": 0.0004187053415284058,
+      "loss": 3.3879,
+      "step": 2237
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.65529465675354,
+      "learning_rate": 0.0004186288311852435,
+      "loss": 3.3781,
+      "step": 2238
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6793820858001709,
+      "learning_rate": 0.0004185522918528302,
+      "loss": 3.51,
+      "step": 2239
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6451491713523865,
+      "learning_rate": 0.0004184757235443238,
+      "loss": 3.366,
+      "step": 2240
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7233899235725403,
+      "learning_rate": 0.0004183991262728875,
+      "loss": 3.4044,
+      "step": 2241
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6855872273445129,
+      "learning_rate": 0.0004183225000516891,
+      "loss": 3.3616,
+      "step": 2242
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7209010720252991,
+      "learning_rate": 0.0004182458448939016,
+      "loss": 3.4268,
+      "step": 2243
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6535835862159729,
+      "learning_rate": 0.00041816916081270286,
+      "loss": 3.5758,
+      "step": 2244
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6555159091949463,
+      "learning_rate": 0.00041809244782127573,
+      "loss": 3.4047,
+      "step": 2245
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6891555190086365,
+      "learning_rate": 0.00041801570593280824,
+      "loss": 3.4448,
+      "step": 2246
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6463344693183899,
+      "learning_rate": 0.00041793893516049307,
+      "loss": 3.4945,
+      "step": 2247
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6066601276397705,
+      "learning_rate": 0.00041786213551752804,
+      "loss": 3.3468,
+      "step": 2248
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6507159471511841,
+      "learning_rate": 0.0004177853070171159,
+      "loss": 3.355,
+      "step": 2249
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6900511980056763,
+      "learning_rate": 0.00041770844967246423,
+      "loss": 3.4342,
+      "step": 2250
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7094230055809021,
+      "learning_rate": 0.0004176315634967859,
+      "loss": 3.5087,
+      "step": 2251
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6796755194664001,
+      "learning_rate": 0.00041755464850329847,
+      "loss": 3.4755,
+      "step": 2252
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7061808109283447,
+      "learning_rate": 0.0004174777047052245,
+      "loss": 3.449,
+      "step": 2253
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6908069252967834,
+      "learning_rate": 0.00041740073211579156,
+      "loss": 3.4706,
+      "step": 2254
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6602137684822083,
+      "learning_rate": 0.00041732373074823204,
+      "loss": 3.521,
+      "step": 2255
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.670946478843689,
+      "learning_rate": 0.0004172467006157834,
+      "loss": 3.5037,
+      "step": 2256
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6751448512077332,
+      "learning_rate": 0.00041716964173168803,
+      "loss": 3.3769,
+      "step": 2257
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.651528000831604,
+      "learning_rate": 0.00041709255410919335,
+      "loss": 3.4461,
+      "step": 2258
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6197723150253296,
+      "learning_rate": 0.00041701543776155136,
+      "loss": 3.4872,
+      "step": 2259
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6545820832252502,
+      "learning_rate": 0.0004169382927020196,
+      "loss": 3.3789,
+      "step": 2260
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6743528246879578,
+      "learning_rate": 0.0004168611189438598,
+      "loss": 3.442,
+      "step": 2261
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7139307856559753,
+      "learning_rate": 0.00041678391650033943,
+      "loss": 3.4165,
+      "step": 2262
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.694196879863739,
+      "learning_rate": 0.0004167066853847302,
+      "loss": 3.5189,
+      "step": 2263
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6422727704048157,
+      "learning_rate": 0.00041662942561030905,
+      "loss": 3.5793,
+      "step": 2264
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6588732004165649,
+      "learning_rate": 0.000416552137190358,
+      "loss": 3.5192,
+      "step": 2265
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6799389123916626,
+      "learning_rate": 0.00041647482013816366,
+      "loss": 3.4602,
+      "step": 2266
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6588906645774841,
+      "learning_rate": 0.0004163974744670179,
+      "loss": 3.388,
+      "step": 2267
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7076383233070374,
+      "learning_rate": 0.00041632010019021715,
+      "loss": 3.3662,
+      "step": 2268
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6735315918922424,
+      "learning_rate": 0.000416242697321063,
+      "loss": 3.3095,
+      "step": 2269
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6909252405166626,
+      "learning_rate": 0.000416165265872862,
+      "loss": 3.486,
+      "step": 2270
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7013283371925354,
+      "learning_rate": 0.0004160878058589255,
+      "loss": 3.4681,
+      "step": 2271
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6996357440948486,
+      "learning_rate": 0.00041601031729256963,
+      "loss": 3.2933,
+      "step": 2272
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7182360887527466,
+      "learning_rate": 0.00041593280018711567,
+      "loss": 3.4998,
+      "step": 2273
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7021991610527039,
+      "learning_rate": 0.0004158552545558897,
+      "loss": 3.4124,
+      "step": 2274
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6800808906555176,
+      "learning_rate": 0.00041577768041222265,
+      "loss": 3.3624,
+      "step": 2275
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6459739208221436,
+      "learning_rate": 0.0004157000777694506,
+      "loss": 3.3325,
+      "step": 2276
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.621931791305542,
+      "learning_rate": 0.0004156224466409141,
+      "loss": 3.4246,
+      "step": 2277
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7608678936958313,
+      "learning_rate": 0.00041554478703995903,
+      "loss": 3.6148,
+      "step": 2278
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6395514011383057,
+      "learning_rate": 0.00041546709897993594,
+      "loss": 3.3696,
+      "step": 2279
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.671046257019043,
+      "learning_rate": 0.0004153893824742002,
+      "loss": 3.552,
+      "step": 2280
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6708778142929077,
+      "learning_rate": 0.00041531163753611236,
+      "loss": 3.4801,
+      "step": 2281
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6961725950241089,
+      "learning_rate": 0.00041523386417903745,
+      "loss": 3.4896,
+      "step": 2282
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7047480940818787,
+      "learning_rate": 0.00041515606241634577,
+      "loss": 3.4363,
+      "step": 2283
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6927662491798401,
+      "learning_rate": 0.0004150782322614124,
+      "loss": 3.454,
+      "step": 2284
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6934167146682739,
+      "learning_rate": 0.00041500037372761705,
+      "loss": 3.4441,
+      "step": 2285
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6866016983985901,
+      "learning_rate": 0.00041492248682834464,
+      "loss": 3.4035,
+      "step": 2286
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7255225777626038,
+      "learning_rate": 0.00041484457157698473,
+      "loss": 3.6204,
+      "step": 2287
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.657615602016449,
+      "learning_rate": 0.00041476662798693205,
+      "loss": 3.4804,
+      "step": 2288
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7463390231132507,
+      "learning_rate": 0.0004146886560715858,
+      "loss": 3.5594,
+      "step": 2289
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7018217444419861,
+      "learning_rate": 0.0004146106558443504,
+      "loss": 3.4931,
+      "step": 2290
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.686687171459198,
+      "learning_rate": 0.0004145326273186348,
+      "loss": 3.4155,
+      "step": 2291
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7059077620506287,
+      "learning_rate": 0.0004144545705078533,
+      "loss": 3.5494,
+      "step": 2292
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6666622757911682,
+      "learning_rate": 0.0004143764854254245,
+      "loss": 3.3401,
+      "step": 2293
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6963295936584473,
+      "learning_rate": 0.0004142983720847723,
+      "loss": 3.3885,
+      "step": 2294
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6787527203559875,
+      "learning_rate": 0.0004142202304993252,
+      "loss": 3.479,
+      "step": 2295
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.658642053604126,
+      "learning_rate": 0.0004141420606825166,
+      "loss": 3.4086,
+      "step": 2296
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.625916600227356,
+      "learning_rate": 0.00041406386264778496,
+      "loss": 3.3912,
+      "step": 2297
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6570123434066772,
+      "learning_rate": 0.0004139856364085733,
+      "loss": 3.5255,
+      "step": 2298
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6459956765174866,
+      "learning_rate": 0.00041390738197832975,
+      "loss": 3.5721,
+      "step": 2299
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.667670488357544,
+      "learning_rate": 0.00041382909937050694,
+      "loss": 3.5063,
+      "step": 2300
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7464162707328796,
+      "learning_rate": 0.00041375078859856275,
+      "loss": 3.3505,
+      "step": 2301
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6404237747192383,
+      "learning_rate": 0.00041367244967595963,
+      "loss": 3.4405,
+      "step": 2302
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6771771907806396,
+      "learning_rate": 0.0004135940826161649,
+      "loss": 3.4206,
+      "step": 2303
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6354660391807556,
+      "learning_rate": 0.0004135156874326509,
+      "loss": 3.4234,
+      "step": 2304
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6784554123878479,
+      "learning_rate": 0.0004134372641388946,
+      "loss": 3.5672,
+      "step": 2305
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6180399060249329,
+      "learning_rate": 0.0004133588127483778,
+      "loss": 3.5065,
+      "step": 2306
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7070502638816833,
+      "learning_rate": 0.00041328033327458726,
+      "loss": 3.3293,
+      "step": 2307
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6411788463592529,
+      "learning_rate": 0.00041320182573101463,
+      "loss": 3.2946,
+      "step": 2308
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6687417030334473,
+      "learning_rate": 0.0004131232901311561,
+      "loss": 3.4136,
+      "step": 2309
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.618337094783783,
+      "learning_rate": 0.00041304472648851285,
+      "loss": 3.2932,
+      "step": 2310
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7408819794654846,
+      "learning_rate": 0.00041296613481659096,
+      "loss": 3.3637,
+      "step": 2311
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6739766597747803,
+      "learning_rate": 0.0004128875151289013,
+      "loss": 3.508,
+      "step": 2312
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7727325558662415,
+      "learning_rate": 0.0004128088674389594,
+      "loss": 3.4848,
+      "step": 2313
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6911073923110962,
+      "learning_rate": 0.0004127301917602857,
+      "loss": 3.4613,
+      "step": 2314
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7270300984382629,
+      "learning_rate": 0.0004126514881064055,
+      "loss": 3.4827,
+      "step": 2315
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6577227711677551,
+      "learning_rate": 0.00041257275649084896,
+      "loss": 3.4467,
+      "step": 2316
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7177078127861023,
+      "learning_rate": 0.00041249399692715085,
+      "loss": 3.2912,
+      "step": 2317
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7087587714195251,
+      "learning_rate": 0.00041241520942885085,
+      "loss": 3.4813,
+      "step": 2318
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6729384660720825,
+      "learning_rate": 0.00041233639400949345,
+      "loss": 3.5241,
+      "step": 2319
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6875993013381958,
+      "learning_rate": 0.00041225755068262804,
+      "loss": 3.5853,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7308191061019897,
+      "learning_rate": 0.00041217867946180856,
+      "loss": 3.4682,
+      "step": 2321
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.697512149810791,
+      "learning_rate": 0.00041209978036059393,
+      "loss": 3.3957,
+      "step": 2322
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7060604691505432,
+      "learning_rate": 0.00041202085339254785,
+      "loss": 3.5362,
+      "step": 2323
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.697181224822998,
+      "learning_rate": 0.00041194189857123877,
+      "loss": 3.4826,
+      "step": 2324
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7139546871185303,
+      "learning_rate": 0.00041186291591023994,
+      "loss": 3.5293,
+      "step": 2325
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7130452990531921,
+      "learning_rate": 0.00041178390542312936,
+      "loss": 3.3592,
+      "step": 2326
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.680930495262146,
+      "learning_rate": 0.0004117048671234899,
+      "loss": 3.408,
+      "step": 2327
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7060506343841553,
+      "learning_rate": 0.00041162580102490915,
+      "loss": 3.3227,
+      "step": 2328
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6682820916175842,
+      "learning_rate": 0.0004115467071409794,
+      "loss": 3.4366,
+      "step": 2329
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6864296197891235,
+      "learning_rate": 0.000411467585485298,
+      "loss": 3.3235,
+      "step": 2330
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6731339693069458,
+      "learning_rate": 0.0004113884360714667,
+      "loss": 3.4194,
+      "step": 2331
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6872637271881104,
+      "learning_rate": 0.0004113092589130923,
+      "loss": 3.4454,
+      "step": 2332
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6677448153495789,
+      "learning_rate": 0.00041123005402378615,
+      "loss": 3.3855,
+      "step": 2333
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7034282088279724,
+      "learning_rate": 0.00041115082141716464,
+      "loss": 3.4623,
+      "step": 2334
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6950244307518005,
+      "learning_rate": 0.00041107156110684875,
+      "loss": 3.3473,
+      "step": 2335
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6312594413757324,
+      "learning_rate": 0.00041099227310646415,
+      "loss": 3.5203,
+      "step": 2336
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6454058289527893,
+      "learning_rate": 0.0004109129574296414,
+      "loss": 3.3099,
+      "step": 2337
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6628250479698181,
+      "learning_rate": 0.0004108336140900158,
+      "loss": 3.3756,
+      "step": 2338
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6997645497322083,
+      "learning_rate": 0.00041075424310122745,
+      "loss": 3.3865,
+      "step": 2339
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6497191190719604,
+      "learning_rate": 0.00041067484447692104,
+      "loss": 3.5008,
+      "step": 2340
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7443307042121887,
+      "learning_rate": 0.0004105954182307462,
+      "loss": 3.5211,
+      "step": 2341
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7580959796905518,
+      "learning_rate": 0.00041051596437635717,
+      "loss": 3.377,
+      "step": 2342
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6966285109519958,
+      "learning_rate": 0.000410436482927413,
+      "loss": 3.4408,
+      "step": 2343
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6279584765434265,
+      "learning_rate": 0.00041035697389757745,
+      "loss": 3.405,
+      "step": 2344
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6910143494606018,
+      "learning_rate": 0.0004102774373005191,
+      "loss": 3.3697,
+      "step": 2345
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7200734615325928,
+      "learning_rate": 0.0004101978731499112,
+      "loss": 3.4753,
+      "step": 2346
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7063745856285095,
+      "learning_rate": 0.00041011828145943167,
+      "loss": 3.4143,
+      "step": 2347
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6820845603942871,
+      "learning_rate": 0.00041003866224276333,
+      "loss": 3.49,
+      "step": 2348
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6298126578330994,
+      "learning_rate": 0.00040995901551359366,
+      "loss": 3.4667,
+      "step": 2349
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6085259914398193,
+      "learning_rate": 0.00040987934128561477,
+      "loss": 3.4444,
+      "step": 2350
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7015394568443298,
+      "learning_rate": 0.0004097996395725237,
+      "loss": 3.3604,
+      "step": 2351
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6495139002799988,
+      "learning_rate": 0.0004097199103880219,
+      "loss": 3.3688,
+      "step": 2352
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.729513943195343,
+      "learning_rate": 0.00040964015374581606,
+      "loss": 3.5373,
+      "step": 2353
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6324521899223328,
+      "learning_rate": 0.00040956036965961694,
+      "loss": 3.5387,
+      "step": 2354
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6699215173721313,
+      "learning_rate": 0.0004094805581431406,
+      "loss": 3.521,
+      "step": 2355
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6890226602554321,
+      "learning_rate": 0.00040940071921010735,
+      "loss": 3.3704,
+      "step": 2356
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7144984602928162,
+      "learning_rate": 0.0004093208528742426,
+      "loss": 3.3831,
+      "step": 2357
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6853955984115601,
+      "learning_rate": 0.0004092409591492762,
+      "loss": 3.4309,
+      "step": 2358
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7000523209571838,
+      "learning_rate": 0.0004091610380489429,
+      "loss": 3.3446,
+      "step": 2359
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6360917091369629,
+      "learning_rate": 0.000409081089586982,
+      "loss": 3.5088,
+      "step": 2360
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6459645628929138,
+      "learning_rate": 0.00040900111377713743,
+      "loss": 3.5955,
+      "step": 2361
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7057053446769714,
+      "learning_rate": 0.0004089211106331583,
+      "loss": 3.4484,
+      "step": 2362
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6538532376289368,
+      "learning_rate": 0.0004088410801687977,
+      "loss": 3.4237,
+      "step": 2363
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.727017343044281,
+      "learning_rate": 0.00040876102239781407,
+      "loss": 3.4437,
+      "step": 2364
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7080045342445374,
+      "learning_rate": 0.0004086809373339703,
+      "loss": 3.4506,
+      "step": 2365
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6796673536300659,
+      "learning_rate": 0.00040860082499103356,
+      "loss": 3.5654,
+      "step": 2366
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.739126443862915,
+      "learning_rate": 0.0004085206853827765,
+      "loss": 3.4897,
+      "step": 2367
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6407649517059326,
+      "learning_rate": 0.0004084405185229759,
+      "loss": 3.3694,
+      "step": 2368
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.063560962677002,
+      "learning_rate": 0.0004083603244254134,
+      "loss": 3.5015,
+      "step": 2369
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6587662696838379,
+      "learning_rate": 0.0004082801031038751,
+      "loss": 3.4524,
+      "step": 2370
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6542607545852661,
+      "learning_rate": 0.00040819985457215223,
+      "loss": 3.3431,
+      "step": 2371
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7463224530220032,
+      "learning_rate": 0.0004081195788440404,
+      "loss": 3.4656,
+      "step": 2372
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7438334822654724,
+      "learning_rate": 0.0004080392759333398,
+      "loss": 3.3541,
+      "step": 2373
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6712779402732849,
+      "learning_rate": 0.00040795894585385557,
+      "loss": 3.475,
+      "step": 2374
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6458290219306946,
+      "learning_rate": 0.0004078785886193973,
+      "loss": 3.3563,
+      "step": 2375
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6500002145767212,
+      "learning_rate": 0.00040779820424377935,
+      "loss": 3.4716,
+      "step": 2376
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6747543811798096,
+      "learning_rate": 0.0004077177927408208,
+      "loss": 3.5333,
+      "step": 2377
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7234296798706055,
+      "learning_rate": 0.0004076373541243452,
+      "loss": 3.5203,
+      "step": 2378
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6669906973838806,
+      "learning_rate": 0.00040755688840818095,
+      "loss": 3.425,
+      "step": 2379
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6469231843948364,
+      "learning_rate": 0.000407476395606161,
+      "loss": 3.4563,
+      "step": 2380
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8278177976608276,
+      "learning_rate": 0.0004073958757321231,
+      "loss": 3.4265,
+      "step": 2381
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6605141758918762,
+      "learning_rate": 0.0004073153287999094,
+      "loss": 3.4845,
+      "step": 2382
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6517539620399475,
+      "learning_rate": 0.00040723475482336703,
+      "loss": 3.5435,
+      "step": 2383
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6293243169784546,
+      "learning_rate": 0.00040715415381634737,
+      "loss": 3.4723,
+      "step": 2384
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6572651267051697,
+      "learning_rate": 0.0004070735257927068,
+      "loss": 3.3577,
+      "step": 2385
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6305806636810303,
+      "learning_rate": 0.0004069928707663062,
+      "loss": 3.4141,
+      "step": 2386
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6208084225654602,
+      "learning_rate": 0.00040691218875101113,
+      "loss": 3.4277,
+      "step": 2387
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7063469290733337,
+      "learning_rate": 0.0004068314797606917,
+      "loss": 3.3206,
+      "step": 2388
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6245595812797546,
+      "learning_rate": 0.0004067507438092227,
+      "loss": 3.4599,
+      "step": 2389
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7018944025039673,
+      "learning_rate": 0.0004066699809104837,
+      "loss": 3.4735,
+      "step": 2390
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6688552498817444,
+      "learning_rate": 0.0004065891910783587,
+      "loss": 3.4035,
+      "step": 2391
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.699343740940094,
+      "learning_rate": 0.00040650837432673647,
+      "loss": 3.4211,
+      "step": 2392
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6850763559341431,
+      "learning_rate": 0.00040642753066951024,
+      "loss": 3.5136,
+      "step": 2393
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.714257538318634,
+      "learning_rate": 0.00040634666012057797,
+      "loss": 3.3978,
+      "step": 2394
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6932154893875122,
+      "learning_rate": 0.00040626576269384244,
+      "loss": 3.4032,
+      "step": 2395
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6759898066520691,
+      "learning_rate": 0.0004061848384032106,
+      "loss": 3.3437,
+      "step": 2396
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6919507384300232,
+      "learning_rate": 0.0004061038872625944,
+      "loss": 3.4002,
+      "step": 2397
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6822661757469177,
+      "learning_rate": 0.00040602290928591026,
+      "loss": 3.4483,
+      "step": 2398
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7360864281654358,
+      "learning_rate": 0.00040594190448707925,
+      "loss": 3.2807,
+      "step": 2399
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6921037435531616,
+      "learning_rate": 0.00040586087288002707,
+      "loss": 3.3602,
+      "step": 2400
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7205747365951538,
+      "learning_rate": 0.0004057798144786839,
+      "loss": 3.4747,
+      "step": 2401
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6903738379478455,
+      "learning_rate": 0.0004056987292969847,
+      "loss": 3.4369,
+      "step": 2402
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7312140464782715,
+      "learning_rate": 0.00040561761734886885,
+      "loss": 3.4649,
+      "step": 2403
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6566363573074341,
+      "learning_rate": 0.0004055364786482806,
+      "loss": 3.4276,
+      "step": 2404
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6498070359230042,
+      "learning_rate": 0.00040545531320916844,
+      "loss": 3.4,
+      "step": 2405
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6509925723075867,
+      "learning_rate": 0.0004053741210454859,
+      "loss": 3.3766,
+      "step": 2406
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6313583254814148,
+      "learning_rate": 0.00040529290217119053,
+      "loss": 3.4926,
+      "step": 2407
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6571773290634155,
+      "learning_rate": 0.00040521165660024503,
+      "loss": 3.5759,
+      "step": 2408
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6563805341720581,
+      "learning_rate": 0.0004051303843466164,
+      "loss": 3.4646,
+      "step": 2409
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6175151467323303,
+      "learning_rate": 0.0004050490854242763,
+      "loss": 3.3364,
+      "step": 2410
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7010701894760132,
+      "learning_rate": 0.0004049677598472009,
+      "loss": 3.4465,
+      "step": 2411
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7390778660774231,
+      "learning_rate": 0.000404886407629371,
+      "loss": 3.5124,
+      "step": 2412
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6527614593505859,
+      "learning_rate": 0.0004048050287847721,
+      "loss": 3.4107,
+      "step": 2413
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6538462042808533,
+      "learning_rate": 0.0004047236233273941,
+      "loss": 3.5326,
+      "step": 2414
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6972126364707947,
+      "learning_rate": 0.00040464219127123147,
+      "loss": 3.4837,
+      "step": 2415
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7578407526016235,
+      "learning_rate": 0.00040456073263028347,
+      "loss": 3.5219,
+      "step": 2416
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6868325471878052,
+      "learning_rate": 0.0004044792474185537,
+      "loss": 3.5467,
+      "step": 2417
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6207484006881714,
+      "learning_rate": 0.00040439773565005034,
+      "loss": 3.3662,
+      "step": 2418
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6857748031616211,
+      "learning_rate": 0.00040431619733878637,
+      "loss": 3.3901,
+      "step": 2419
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6813026070594788,
+      "learning_rate": 0.0004042346324987791,
+      "loss": 3.2639,
+      "step": 2420
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6619722843170166,
+      "learning_rate": 0.0004041530411440504,
+      "loss": 3.5003,
+      "step": 2421
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6328409314155579,
+      "learning_rate": 0.0004040714232886269,
+      "loss": 3.5067,
+      "step": 2422
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.68046635389328,
+      "learning_rate": 0.00040398977894653954,
+      "loss": 3.4216,
+      "step": 2423
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6497738361358643,
+      "learning_rate": 0.000403908108131824,
+      "loss": 3.4894,
+      "step": 2424
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6499060988426208,
+      "learning_rate": 0.0004038264108585204,
+      "loss": 3.4544,
+      "step": 2425
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6749792695045471,
+      "learning_rate": 0.00040374468714067345,
+      "loss": 3.4488,
+      "step": 2426
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7037678360939026,
+      "learning_rate": 0.00040366293699233246,
+      "loss": 3.4747,
+      "step": 2427
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6845188736915588,
+      "learning_rate": 0.00040358116042755115,
+      "loss": 3.4011,
+      "step": 2428
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.651651918888092,
+      "learning_rate": 0.00040349935746038793,
+      "loss": 3.4295,
+      "step": 2429
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6486951112747192,
+      "learning_rate": 0.0004034175281049056,
+      "loss": 3.3702,
+      "step": 2430
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7313594818115234,
+      "learning_rate": 0.00040333567237517165,
+      "loss": 3.4048,
+      "step": 2431
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6645215153694153,
+      "learning_rate": 0.000403253790285258,
+      "loss": 3.4562,
+      "step": 2432
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6896566152572632,
+      "learning_rate": 0.00040317188184924116,
+      "loss": 3.3233,
+      "step": 2433
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6973954439163208,
+      "learning_rate": 0.00040308994708120207,
+      "loss": 3.4185,
+      "step": 2434
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6570093631744385,
+      "learning_rate": 0.00040300798599522626,
+      "loss": 3.3588,
+      "step": 2435
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6232661604881287,
+      "learning_rate": 0.00040292599860540393,
+      "loss": 3.4517,
+      "step": 2436
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6920954585075378,
+      "learning_rate": 0.00040284398492582954,
+      "loss": 3.4176,
+      "step": 2437
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7019113898277283,
+      "learning_rate": 0.00040276194497060227,
+      "loss": 3.4126,
+      "step": 2438
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6729170680046082,
+      "learning_rate": 0.0004026798787538256,
+      "loss": 3.4647,
+      "step": 2439
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6775332689285278,
+      "learning_rate": 0.00040259778628960773,
+      "loss": 3.4055,
+      "step": 2440
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7059245109558105,
+      "learning_rate": 0.0004025156675920614,
+      "loss": 3.4201,
+      "step": 2441
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6790223717689514,
+      "learning_rate": 0.0004024335226753037,
+      "loss": 3.4517,
+      "step": 2442
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6908589005470276,
+      "learning_rate": 0.0004023513515534563,
+      "loss": 3.3657,
+      "step": 2443
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6590731143951416,
+      "learning_rate": 0.0004022691542406453,
+      "loss": 3.4982,
+      "step": 2444
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6416826248168945,
+      "learning_rate": 0.00040218693075100146,
+      "loss": 3.3781,
+      "step": 2445
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6760276556015015,
+      "learning_rate": 0.0004021046810986599,
+      "loss": 3.4182,
+      "step": 2446
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6423786282539368,
+      "learning_rate": 0.0004020224052977604,
+      "loss": 3.5088,
+      "step": 2447
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7082242369651794,
+      "learning_rate": 0.00040194010336244696,
+      "loss": 3.3883,
+      "step": 2448
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6391067504882812,
+      "learning_rate": 0.0004018577753068683,
+      "loss": 3.2714,
+      "step": 2449
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7196114659309387,
+      "learning_rate": 0.00040177542114517765,
+      "loss": 3.5096,
+      "step": 2450
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6951587200164795,
+      "learning_rate": 0.0004016930408915325,
+      "loss": 3.3887,
+      "step": 2451
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.622295081615448,
+      "learning_rate": 0.00040161063456009506,
+      "loss": 3.3613,
+      "step": 2452
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6327260732650757,
+      "learning_rate": 0.00040152820216503196,
+      "loss": 3.4621,
+      "step": 2453
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6431770920753479,
+      "learning_rate": 0.00040144574372051416,
+      "loss": 3.4108,
+      "step": 2454
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6560462117195129,
+      "learning_rate": 0.0004013632592407174,
+      "loss": 3.4254,
+      "step": 2455
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.672125518321991,
+      "learning_rate": 0.00040128074873982156,
+      "loss": 3.45,
+      "step": 2456
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6163976788520813,
+      "learning_rate": 0.0004011982122320112,
+      "loss": 3.3067,
+      "step": 2457
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.667873740196228,
+      "learning_rate": 0.0004011156497314754,
+      "loss": 3.3488,
+      "step": 2458
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6594126224517822,
+      "learning_rate": 0.00040103306125240746,
+      "loss": 3.5699,
+      "step": 2459
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6237178444862366,
+      "learning_rate": 0.0004009504468090054,
+      "loss": 3.5031,
+      "step": 2460
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6896465420722961,
+      "learning_rate": 0.00040086780641547157,
+      "loss": 3.4244,
+      "step": 2461
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6883692145347595,
+      "learning_rate": 0.00040078514008601274,
+      "loss": 3.4431,
+      "step": 2462
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6940760016441345,
+      "learning_rate": 0.00040070244783484035,
+      "loss": 3.3533,
+      "step": 2463
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6569497585296631,
+      "learning_rate": 0.00040061972967617013,
+      "loss": 3.4175,
+      "step": 2464
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6395750641822815,
+      "learning_rate": 0.00040053698562422216,
+      "loss": 3.5232,
+      "step": 2465
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6653283834457397,
+      "learning_rate": 0.00040045421569322116,
+      "loss": 3.3437,
+      "step": 2466
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6142536997795105,
+      "learning_rate": 0.00040037141989739626,
+      "loss": 3.3303,
+      "step": 2467
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6975305080413818,
+      "learning_rate": 0.0004002885982509811,
+      "loss": 3.3792,
+      "step": 2468
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6583127975463867,
+      "learning_rate": 0.0004002057507682136,
+      "loss": 3.5056,
+      "step": 2469
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6988340616226196,
+      "learning_rate": 0.0004001228774633362,
+      "loss": 3.4311,
+      "step": 2470
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7292492985725403,
+      "learning_rate": 0.0004000399783505957,
+      "loss": 3.3629,
+      "step": 2471
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6581091284751892,
+      "learning_rate": 0.0003999570534442436,
+      "loss": 3.3712,
+      "step": 2472
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6563900113105774,
+      "learning_rate": 0.0003998741027585356,
+      "loss": 3.4227,
+      "step": 2473
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7054997086524963,
+      "learning_rate": 0.0003997911263077318,
+      "loss": 3.4234,
+      "step": 2474
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7008715271949768,
+      "learning_rate": 0.0003997081241060968,
+      "loss": 3.505,
+      "step": 2475
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6388863921165466,
+      "learning_rate": 0.00039962509616789984,
+      "loss": 3.2693,
+      "step": 2476
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8109036087989807,
+      "learning_rate": 0.0003995420425074141,
+      "loss": 3.3847,
+      "step": 2477
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6357604265213013,
+      "learning_rate": 0.00039945896313891774,
+      "loss": 3.3092,
+      "step": 2478
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6333456039428711,
+      "learning_rate": 0.00039937585807669286,
+      "loss": 3.29,
+      "step": 2479
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6567932963371277,
+      "learning_rate": 0.00039929272733502623,
+      "loss": 3.5207,
+      "step": 2480
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6872777938842773,
+      "learning_rate": 0.0003992095709282091,
+      "loss": 3.3851,
+      "step": 2481
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6259168982505798,
+      "learning_rate": 0.0003991263888705369,
+      "loss": 3.4827,
+      "step": 2482
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.637649655342102,
+      "learning_rate": 0.00039904318117630965,
+      "loss": 3.6457,
+      "step": 2483
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6806047558784485,
+      "learning_rate": 0.0003989599478598317,
+      "loss": 3.3285,
+      "step": 2484
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6640322208404541,
+      "learning_rate": 0.0003988766889354118,
+      "loss": 3.4148,
+      "step": 2485
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6474546790122986,
+      "learning_rate": 0.00039879340441736314,
+      "loss": 3.3632,
+      "step": 2486
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6708735823631287,
+      "learning_rate": 0.0003987100943200033,
+      "loss": 3.4945,
+      "step": 2487
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7857782244682312,
+      "learning_rate": 0.0003986267586576543,
+      "loss": 3.45,
+      "step": 2488
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.733543872833252,
+      "learning_rate": 0.0003985433974446424,
+      "loss": 3.4585,
+      "step": 2489
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.687157392501831,
+      "learning_rate": 0.0003984600106952985,
+      "loss": 3.2763,
+      "step": 2490
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6997659206390381,
+      "learning_rate": 0.00039837659842395756,
+      "loss": 3.529,
+      "step": 2491
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6506561636924744,
+      "learning_rate": 0.0003982931606449592,
+      "loss": 3.5072,
+      "step": 2492
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6445373296737671,
+      "learning_rate": 0.00039820969737264743,
+      "loss": 3.3944,
+      "step": 2493
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6242894530296326,
+      "learning_rate": 0.00039812620862137056,
+      "loss": 3.4263,
+      "step": 2494
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5803747773170471,
+      "learning_rate": 0.0003980426944054811,
+      "loss": 3.3099,
+      "step": 2495
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6158910989761353,
+      "learning_rate": 0.00039795915473933623,
+      "loss": 3.5117,
+      "step": 2496
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6407814621925354,
+      "learning_rate": 0.00039787558963729744,
+      "loss": 3.3517,
+      "step": 2497
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6843712329864502,
+      "learning_rate": 0.0003977919991137304,
+      "loss": 3.4314,
+      "step": 2498
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7148441076278687,
+      "learning_rate": 0.0003977083831830054,
+      "loss": 3.3656,
+      "step": 2499
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6620389819145203,
+      "learning_rate": 0.00039762474185949686,
+      "loss": 3.4349,
+      "step": 2500
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6523193120956421,
+      "learning_rate": 0.0003975410751575839,
+      "loss": 3.4915,
+      "step": 2501
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6660597920417786,
+      "learning_rate": 0.0003974573830916496,
+      "loss": 3.4874,
+      "step": 2502
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7087952494621277,
+      "learning_rate": 0.0003973736656760817,
+      "loss": 3.5746,
+      "step": 2503
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7073147296905518,
+      "learning_rate": 0.0003972899229252721,
+      "loss": 3.3991,
+      "step": 2504
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6184029579162598,
+      "learning_rate": 0.00039720615485361733,
+      "loss": 3.4825,
+      "step": 2505
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6575698256492615,
+      "learning_rate": 0.00039712236147551795,
+      "loss": 3.4063,
+      "step": 2506
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.648165762424469,
+      "learning_rate": 0.000397038542805379,
+      "loss": 3.5048,
+      "step": 2507
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6533029079437256,
+      "learning_rate": 0.00039695469885761,
+      "loss": 3.3698,
+      "step": 2508
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6076450943946838,
+      "learning_rate": 0.0003968708296466245,
+      "loss": 3.3721,
+      "step": 2509
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.631220281124115,
+      "learning_rate": 0.00039678693518684083,
+      "loss": 3.3633,
+      "step": 2510
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6237044334411621,
+      "learning_rate": 0.0003967030154926813,
+      "loss": 3.4118,
+      "step": 2511
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.712161660194397,
+      "learning_rate": 0.00039661907057857263,
+      "loss": 3.386,
+      "step": 2512
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6828798651695251,
+      "learning_rate": 0.0003965351004589459,
+      "loss": 3.4011,
+      "step": 2513
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6511789560317993,
+      "learning_rate": 0.0003964511051482367,
+      "loss": 3.3554,
+      "step": 2514
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7005689740180969,
+      "learning_rate": 0.00039636708466088476,
+      "loss": 3.4527,
+      "step": 2515
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6651748418807983,
+      "learning_rate": 0.00039628303901133413,
+      "loss": 3.4961,
+      "step": 2516
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6886200308799744,
+      "learning_rate": 0.00039619896821403315,
+      "loss": 3.3853,
+      "step": 2517
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6898598670959473,
+      "learning_rate": 0.0003961148722834347,
+      "loss": 3.3678,
+      "step": 2518
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6316580772399902,
+      "learning_rate": 0.0003960307512339958,
+      "loss": 3.3588,
+      "step": 2519
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6656391620635986,
+      "learning_rate": 0.00039594660508017774,
+      "loss": 3.4436,
+      "step": 2520
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7290733456611633,
+      "learning_rate": 0.00039586243383644645,
+      "loss": 3.4326,
+      "step": 2521
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.684203565120697,
+      "learning_rate": 0.00039577823751727175,
+      "loss": 3.4813,
+      "step": 2522
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6514954566955566,
+      "learning_rate": 0.00039569401613712797,
+      "loss": 3.4506,
+      "step": 2523
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6980857849121094,
+      "learning_rate": 0.00039560976971049386,
+      "loss": 3.3717,
+      "step": 2524
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6839146614074707,
+      "learning_rate": 0.00039552549825185224,
+      "loss": 3.3792,
+      "step": 2525
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6302343010902405,
+      "learning_rate": 0.0003954412017756904,
+      "loss": 3.4761,
+      "step": 2526
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6123268604278564,
+      "learning_rate": 0.00039535688029649984,
+      "loss": 3.3494,
+      "step": 2527
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6835804581642151,
+      "learning_rate": 0.0003952725338287765,
+      "loss": 3.4457,
+      "step": 2528
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6334277391433716,
+      "learning_rate": 0.0003951881623870204,
+      "loss": 3.3608,
+      "step": 2529
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6541914939880371,
+      "learning_rate": 0.00039510376598573605,
+      "loss": 3.4332,
+      "step": 2530
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6691238880157471,
+      "learning_rate": 0.0003950193446394321,
+      "loss": 3.553,
+      "step": 2531
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5947278738021851,
+      "learning_rate": 0.00039493489836262165,
+      "loss": 3.4345,
+      "step": 2532
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6068588495254517,
+      "learning_rate": 0.0003948504271698219,
+      "loss": 3.4011,
+      "step": 2533
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6819738745689392,
+      "learning_rate": 0.0003947659310755545,
+      "loss": 3.3812,
+      "step": 2534
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6502918004989624,
+      "learning_rate": 0.00039468141009434533,
+      "loss": 3.5099,
+      "step": 2535
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6570164561271667,
+      "learning_rate": 0.0003945968642407244,
+      "loss": 3.4287,
+      "step": 2536
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7019553184509277,
+      "learning_rate": 0.0003945122935292261,
+      "loss": 3.3873,
+      "step": 2537
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6633791923522949,
+      "learning_rate": 0.00039442769797438937,
+      "loss": 3.4676,
+      "step": 2538
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6608754396438599,
+      "learning_rate": 0.0003943430775907569,
+      "loss": 3.3399,
+      "step": 2539
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6182772517204285,
+      "learning_rate": 0.0003942584323928762,
+      "loss": 3.4271,
+      "step": 2540
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7042610049247742,
+      "learning_rate": 0.00039417376239529847,
+      "loss": 3.4318,
+      "step": 2541
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.691856861114502,
+      "learning_rate": 0.00039408906761257957,
+      "loss": 3.435,
+      "step": 2542
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.692318856716156,
+      "learning_rate": 0.00039400434805927947,
+      "loss": 3.5448,
+      "step": 2543
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6041357517242432,
+      "learning_rate": 0.00039391960374996263,
+      "loss": 3.3525,
+      "step": 2544
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7062011361122131,
+      "learning_rate": 0.0003938348346991973,
+      "loss": 3.5398,
+      "step": 2545
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6786819100379944,
+      "learning_rate": 0.0003937500409215565,
+      "loss": 3.3596,
+      "step": 2546
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6160567998886108,
+      "learning_rate": 0.0003936652224316172,
+      "loss": 3.3977,
+      "step": 2547
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6385580897331238,
+      "learning_rate": 0.00039358037924396055,
+      "loss": 3.4078,
+      "step": 2548
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6399730443954468,
+      "learning_rate": 0.0003934955113731723,
+      "loss": 3.329,
+      "step": 2549
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6369842290878296,
+      "learning_rate": 0.000393410618833842,
+      "loss": 3.5006,
+      "step": 2550
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6642215251922607,
+      "learning_rate": 0.0003933257016405638,
+      "loss": 3.3534,
+      "step": 2551
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7062375545501709,
+      "learning_rate": 0.0003932407598079359,
+      "loss": 3.4078,
+      "step": 2552
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6758290529251099,
+      "learning_rate": 0.0003931557933505608,
+      "loss": 3.3559,
+      "step": 2553
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6561087965965271,
+      "learning_rate": 0.00039307080228304526,
+      "loss": 3.4428,
+      "step": 2554
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6365450620651245,
+      "learning_rate": 0.0003929857866200002,
+      "loss": 3.3149,
+      "step": 2555
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6529794335365295,
+      "learning_rate": 0.0003929007463760407,
+      "loss": 3.4746,
+      "step": 2556
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6441689133644104,
+      "learning_rate": 0.0003928156815657863,
+      "loss": 3.3952,
+      "step": 2557
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6540311574935913,
+      "learning_rate": 0.00039273059220386065,
+      "loss": 3.522,
+      "step": 2558
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6645547747612,
+      "learning_rate": 0.0003926454783048914,
+      "loss": 3.4708,
+      "step": 2559
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6746283173561096,
+      "learning_rate": 0.0003925603398835108,
+      "loss": 3.3483,
+      "step": 2560
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6782620549201965,
+      "learning_rate": 0.00039247517695435507,
+      "loss": 3.3578,
+      "step": 2561
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6735712289810181,
+      "learning_rate": 0.00039238998953206473,
+      "loss": 3.3935,
+      "step": 2562
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6714909672737122,
+      "learning_rate": 0.0003923047776312844,
+      "loss": 3.4678,
+      "step": 2563
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7226071953773499,
+      "learning_rate": 0.00039221954126666313,
+      "loss": 3.5213,
+      "step": 2564
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6498965620994568,
+      "learning_rate": 0.00039213428045285385,
+      "loss": 3.4121,
+      "step": 2565
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6417312622070312,
+      "learning_rate": 0.0003920489952045141,
+      "loss": 3.3316,
+      "step": 2566
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6538100838661194,
+      "learning_rate": 0.00039196368553630535,
+      "loss": 3.3282,
+      "step": 2567
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6624993085861206,
+      "learning_rate": 0.0003918783514628932,
+      "loss": 3.3879,
+      "step": 2568
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.662112295627594,
+      "learning_rate": 0.0003917929929989476,
+      "loss": 3.3907,
+      "step": 2569
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6703125238418579,
+      "learning_rate": 0.0003917076101591427,
+      "loss": 3.5137,
+      "step": 2570
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6566250920295715,
+      "learning_rate": 0.00039162220295815684,
+      "loss": 3.4737,
+      "step": 2571
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6487689018249512,
+      "learning_rate": 0.00039153677141067254,
+      "loss": 3.3527,
+      "step": 2572
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.704270601272583,
+      "learning_rate": 0.0003914513155313763,
+      "loss": 3.5182,
+      "step": 2573
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.793128252029419,
+      "learning_rate": 0.00039136583533495905,
+      "loss": 3.3832,
+      "step": 2574
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6730992197990417,
+      "learning_rate": 0.0003912803308361159,
+      "loss": 3.6286,
+      "step": 2575
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6524163484573364,
+      "learning_rate": 0.00039119480204954606,
+      "loss": 3.4998,
+      "step": 2576
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7632148265838623,
+      "learning_rate": 0.00039110924898995284,
+      "loss": 3.4395,
+      "step": 2577
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6825524568557739,
+      "learning_rate": 0.0003910236716720438,
+      "loss": 3.4272,
+      "step": 2578
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6332438588142395,
+      "learning_rate": 0.00039093807011053073,
+      "loss": 3.4783,
+      "step": 2579
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6003892421722412,
+      "learning_rate": 0.0003908524443201296,
+      "loss": 3.3101,
+      "step": 2580
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6262263655662537,
+      "learning_rate": 0.0003907667943155603,
+      "loss": 3.4023,
+      "step": 2581
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6281822919845581,
+      "learning_rate": 0.00039068112011154715,
+      "loss": 3.4619,
+      "step": 2582
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6222302317619324,
+      "learning_rate": 0.0003905954217228186,
+      "loss": 3.2979,
+      "step": 2583
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6705178022384644,
+      "learning_rate": 0.0003905096991641071,
+      "loss": 3.4174,
+      "step": 2584
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6498988270759583,
+      "learning_rate": 0.00039042395245014933,
+      "loss": 3.4733,
+      "step": 2585
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6690146327018738,
+      "learning_rate": 0.00039033818159568624,
+      "loss": 3.3617,
+      "step": 2586
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7530152201652527,
+      "learning_rate": 0.0003902523866154628,
+      "loss": 3.3473,
+      "step": 2587
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.647878110408783,
+      "learning_rate": 0.00039016656752422814,
+      "loss": 3.3918,
+      "step": 2588
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6253050565719604,
+      "learning_rate": 0.00039008072433673556,
+      "loss": 3.3227,
+      "step": 2589
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6417919993400574,
+      "learning_rate": 0.00038999485706774263,
+      "loss": 3.3752,
+      "step": 2590
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7062673568725586,
+      "learning_rate": 0.0003899089657320107,
+      "loss": 3.4478,
+      "step": 2591
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6611924767494202,
+      "learning_rate": 0.0003898230503443055,
+      "loss": 3.4042,
+      "step": 2592
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6270459294319153,
+      "learning_rate": 0.0003897371109193972,
+      "loss": 3.4388,
+      "step": 2593
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6408137083053589,
+      "learning_rate": 0.00038965114747205944,
+      "loss": 3.3497,
+      "step": 2594
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7065438032150269,
+      "learning_rate": 0.0003895651600170705,
+      "loss": 3.5494,
+      "step": 2595
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6970211863517761,
+      "learning_rate": 0.0003894791485692125,
+      "loss": 3.4482,
+      "step": 2596
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6874566078186035,
+      "learning_rate": 0.0003893931131432719,
+      "loss": 3.3486,
+      "step": 2597
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6408682465553284,
+      "learning_rate": 0.00038930705375403923,
+      "loss": 3.4459,
+      "step": 2598
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6832528114318848,
+      "learning_rate": 0.000389220970416309,
+      "loss": 3.479,
+      "step": 2599
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7190806865692139,
+      "learning_rate": 0.00038913486314488,
+      "loss": 3.3861,
+      "step": 2600
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6788740754127502,
+      "learning_rate": 0.000389048731954555,
+      "loss": 3.5252,
+      "step": 2601
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7985811233520508,
+      "learning_rate": 0.0003889625768601409,
+      "loss": 3.3963,
+      "step": 2602
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.654829740524292,
+      "learning_rate": 0.00038887639787644904,
+      "loss": 3.4419,
+      "step": 2603
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6960733532905579,
+      "learning_rate": 0.00038879019501829423,
+      "loss": 3.3564,
+      "step": 2604
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6409716010093689,
+      "learning_rate": 0.00038870396830049595,
+      "loss": 3.4309,
+      "step": 2605
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6938721537590027,
+      "learning_rate": 0.00038861771773787744,
+      "loss": 3.407,
+      "step": 2606
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6538400053977966,
+      "learning_rate": 0.00038853144334526643,
+      "loss": 3.4347,
+      "step": 2607
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6794121265411377,
+      "learning_rate": 0.0003884451451374942,
+      "loss": 3.3708,
+      "step": 2608
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6384621858596802,
+      "learning_rate": 0.00038835882312939656,
+      "loss": 3.4581,
+      "step": 2609
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6200606226921082,
+      "learning_rate": 0.0003882724773358133,
+      "loss": 3.2224,
+      "step": 2610
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7154632210731506,
+      "learning_rate": 0.00038818610777158804,
+      "loss": 3.3628,
+      "step": 2611
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7068455219268799,
+      "learning_rate": 0.000388099714451569,
+      "loss": 3.4331,
+      "step": 2612
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6456942558288574,
+      "learning_rate": 0.000388013297390608,
+      "loss": 3.3843,
+      "step": 2613
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7880919575691223,
+      "learning_rate": 0.0003879268566035612,
+      "loss": 3.3801,
+      "step": 2614
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7418599128723145,
+      "learning_rate": 0.0003878403921052887,
+      "loss": 3.436,
+      "step": 2615
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7047538161277771,
+      "learning_rate": 0.0003877539039106549,
+      "loss": 3.3842,
+      "step": 2616
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6789987683296204,
+      "learning_rate": 0.000387667392034528,
+      "loss": 3.4646,
+      "step": 2617
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6848494410514832,
+      "learning_rate": 0.00038758085649178044,
+      "loss": 3.4293,
+      "step": 2618
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6500349044799805,
+      "learning_rate": 0.0003874942972972887,
+      "loss": 3.4177,
+      "step": 2619
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7213976979255676,
+      "learning_rate": 0.00038740771446593323,
+      "loss": 3.5621,
+      "step": 2620
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6804102659225464,
+      "learning_rate": 0.00038732110801259874,
+      "loss": 3.4393,
+      "step": 2621
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6893405318260193,
+      "learning_rate": 0.00038723447795217374,
+      "loss": 3.482,
+      "step": 2622
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6176344752311707,
+      "learning_rate": 0.0003871478242995511,
+      "loss": 3.4335,
+      "step": 2623
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6017330288887024,
+      "learning_rate": 0.0003870611470696275,
+      "loss": 3.3067,
+      "step": 2624
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5874341726303101,
+      "learning_rate": 0.00038697444627730374,
+      "loss": 3.4075,
+      "step": 2625
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6801062226295471,
+      "learning_rate": 0.00038688772193748466,
+      "loss": 3.4583,
+      "step": 2626
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6636302471160889,
+      "learning_rate": 0.0003868009740650794,
+      "loss": 3.464,
+      "step": 2627
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6342689990997314,
+      "learning_rate": 0.00038671420267500067,
+      "loss": 3.4211,
+      "step": 2628
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.679341733455658,
+      "learning_rate": 0.0003866274077821655,
+      "loss": 3.5502,
+      "step": 2629
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6680846810340881,
+      "learning_rate": 0.0003865405894014951,
+      "loss": 3.4471,
+      "step": 2630
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6836916208267212,
+      "learning_rate": 0.00038645374754791445,
+      "loss": 3.499,
+      "step": 2631
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.698597252368927,
+      "learning_rate": 0.00038636688223635273,
+      "loss": 3.5174,
+      "step": 2632
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6660541296005249,
+      "learning_rate": 0.000386279993481743,
+      "loss": 3.4333,
+      "step": 2633
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6841675043106079,
+      "learning_rate": 0.0003861930812990225,
+      "loss": 3.511,
+      "step": 2634
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6387922167778015,
+      "learning_rate": 0.00038610614570313244,
+      "loss": 3.3794,
+      "step": 2635
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6555662751197815,
+      "learning_rate": 0.00038601918670901807,
+      "loss": 3.3818,
+      "step": 2636
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6639174222946167,
+      "learning_rate": 0.0003859322043316287,
+      "loss": 3.5306,
+      "step": 2637
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6673606038093567,
+      "learning_rate": 0.0003858451985859175,
+      "loss": 3.489,
+      "step": 2638
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6728341579437256,
+      "learning_rate": 0.0003857581694868417,
+      "loss": 3.3419,
+      "step": 2639
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6590186953544617,
+      "learning_rate": 0.00038567111704936285,
+      "loss": 3.3493,
+      "step": 2640
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6790418028831482,
+      "learning_rate": 0.00038558404128844604,
+      "loss": 3.4853,
+      "step": 2641
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.660210371017456,
+      "learning_rate": 0.00038549694221906084,
+      "loss": 3.4395,
+      "step": 2642
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6540093421936035,
+      "learning_rate": 0.00038540981985618036,
+      "loss": 3.4581,
+      "step": 2643
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.647119402885437,
+      "learning_rate": 0.00038532267421478205,
+      "loss": 3.5108,
+      "step": 2644
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6743528842926025,
+      "learning_rate": 0.0003852355053098473,
+      "loss": 3.3326,
+      "step": 2645
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6162041425704956,
+      "learning_rate": 0.00038514831315636134,
+      "loss": 3.4508,
+      "step": 2646
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6437183022499084,
+      "learning_rate": 0.00038506109776931366,
+      "loss": 3.3653,
+      "step": 2647
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6612119674682617,
+      "learning_rate": 0.0003849738591636974,
+      "loss": 3.4162,
+      "step": 2648
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6219061613082886,
+      "learning_rate": 0.00038488659735451,
+      "loss": 3.4372,
+      "step": 2649
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6129502654075623,
+      "learning_rate": 0.00038479931235675294,
+      "loss": 3.4622,
+      "step": 2650
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6514201164245605,
+      "learning_rate": 0.0003847120041854312,
+      "loss": 3.4952,
+      "step": 2651
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6656510233879089,
+      "learning_rate": 0.00038462467285555423,
+      "loss": 3.4955,
+      "step": 2652
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6361547708511353,
+      "learning_rate": 0.00038453731838213533,
+      "loss": 3.4209,
+      "step": 2653
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6556544303894043,
+      "learning_rate": 0.0003844499407801918,
+      "loss": 3.4026,
+      "step": 2654
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6475023627281189,
+      "learning_rate": 0.0003843625400647446,
+      "loss": 3.5694,
+      "step": 2655
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6740972995758057,
+      "learning_rate": 0.00038427511625081925,
+      "loss": 3.457,
+      "step": 2656
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7039819955825806,
+      "learning_rate": 0.00038418766935344466,
+      "loss": 3.5753,
+      "step": 2657
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6474445462226868,
+      "learning_rate": 0.00038410019938765413,
+      "loss": 3.4807,
+      "step": 2658
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6333473920822144,
+      "learning_rate": 0.00038401270636848474,
+      "loss": 3.423,
+      "step": 2659
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6756442785263062,
+      "learning_rate": 0.00038392519031097745,
+      "loss": 3.3655,
+      "step": 2660
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6421991586685181,
+      "learning_rate": 0.0003838376512301773,
+      "loss": 3.3833,
+      "step": 2661
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.650489330291748,
+      "learning_rate": 0.0003837500891411334,
+      "loss": 3.4429,
+      "step": 2662
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6732882857322693,
+      "learning_rate": 0.0003836625040588986,
+      "loss": 3.3899,
+      "step": 2663
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6878411173820496,
+      "learning_rate": 0.00038357489599852984,
+      "loss": 3.4397,
+      "step": 2664
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6262323260307312,
+      "learning_rate": 0.0003834872649750879,
+      "loss": 3.2506,
+      "step": 2665
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6440932750701904,
+      "learning_rate": 0.00038339961100363753,
+      "loss": 3.4439,
+      "step": 2666
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6685276031494141,
+      "learning_rate": 0.0003833119340992476,
+      "loss": 3.5265,
+      "step": 2667
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6493540406227112,
+      "learning_rate": 0.0003832242342769907,
+      "loss": 3.3814,
+      "step": 2668
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6593091487884521,
+      "learning_rate": 0.00038313651155194345,
+      "loss": 3.3719,
+      "step": 2669
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.647038459777832,
+      "learning_rate": 0.00038304876593918646,
+      "loss": 3.359,
+      "step": 2670
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.697995126247406,
+      "learning_rate": 0.0003829609974538041,
+      "loss": 3.4759,
+      "step": 2671
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6449376344680786,
+      "learning_rate": 0.00038287320611088486,
+      "loss": 3.4125,
+      "step": 2672
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.674048900604248,
+      "learning_rate": 0.00038278539192552113,
+      "loss": 3.4293,
+      "step": 2673
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6361321806907654,
+      "learning_rate": 0.00038269755491280916,
+      "loss": 3.4286,
+      "step": 2674
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.673592746257782,
+      "learning_rate": 0.00038260969508784916,
+      "loss": 3.4306,
+      "step": 2675
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.597294807434082,
+      "learning_rate": 0.00038252181246574516,
+      "loss": 3.3668,
+      "step": 2676
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6507534384727478,
+      "learning_rate": 0.0003824339070616053,
+      "loss": 3.4632,
+      "step": 2677
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6554521322250366,
+      "learning_rate": 0.0003823459788905415,
+      "loss": 3.5535,
+      "step": 2678
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.668842077255249,
+      "learning_rate": 0.00038225802796766964,
+      "loss": 3.6099,
+      "step": 2679
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6702171564102173,
+      "learning_rate": 0.0003821700543081095,
+      "loss": 3.4481,
+      "step": 2680
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7185189127922058,
+      "learning_rate": 0.00038208205792698483,
+      "loss": 3.3188,
+      "step": 2681
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6936577558517456,
+      "learning_rate": 0.0003819940388394232,
+      "loss": 3.4154,
+      "step": 2682
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6211004257202148,
+      "learning_rate": 0.00038190599706055595,
+      "loss": 3.3764,
+      "step": 2683
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7152443528175354,
+      "learning_rate": 0.00038181793260551875,
+      "loss": 3.499,
+      "step": 2684
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5948370695114136,
+      "learning_rate": 0.00038172984548945067,
+      "loss": 3.3368,
+      "step": 2685
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6092764139175415,
+      "learning_rate": 0.00038164173572749504,
+      "loss": 3.3187,
+      "step": 2686
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6423379182815552,
+      "learning_rate": 0.0003815536033347989,
+      "loss": 3.5358,
+      "step": 2687
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6405006051063538,
+      "learning_rate": 0.0003814654483265134,
+      "loss": 3.3378,
+      "step": 2688
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7445230484008789,
+      "learning_rate": 0.00038137727071779304,
+      "loss": 3.4888,
+      "step": 2689
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6603292226791382,
+      "learning_rate": 0.0003812890705237969,
+      "loss": 3.4612,
+      "step": 2690
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6725325584411621,
+      "learning_rate": 0.0003812008477596875,
+      "loss": 3.4146,
+      "step": 2691
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.647088885307312,
+      "learning_rate": 0.0003811126024406314,
+      "loss": 3.3068,
+      "step": 2692
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6597961187362671,
+      "learning_rate": 0.00038102433458179883,
+      "loss": 3.4599,
+      "step": 2693
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6334750652313232,
+      "learning_rate": 0.0003809360441983643,
+      "loss": 3.2942,
+      "step": 2694
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6588044166564941,
+      "learning_rate": 0.00038084773130550576,
+      "loss": 3.2866,
+      "step": 2695
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5897568464279175,
+      "learning_rate": 0.0003807593959184053,
+      "loss": 3.4746,
+      "step": 2696
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6646286845207214,
+      "learning_rate": 0.0003806710380522488,
+      "loss": 3.3076,
+      "step": 2697
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6470707654953003,
+      "learning_rate": 0.000380582657722226,
+      "loss": 3.504,
+      "step": 2698
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6288967728614807,
+      "learning_rate": 0.00038049425494353047,
+      "loss": 3.4962,
+      "step": 2699
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6686561703681946,
+      "learning_rate": 0.00038040582973135974,
+      "loss": 3.4595,
+      "step": 2700
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6200351119041443,
+      "learning_rate": 0.00038031738210091506,
+      "loss": 3.3326,
+      "step": 2701
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.634735643863678,
+      "learning_rate": 0.0003802289120674016,
+      "loss": 3.3659,
+      "step": 2702
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6304633021354675,
+      "learning_rate": 0.00038014041964602843,
+      "loss": 3.4768,
+      "step": 2703
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6305013298988342,
+      "learning_rate": 0.00038005190485200844,
+      "loss": 3.5768,
+      "step": 2704
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6871489882469177,
+      "learning_rate": 0.0003799633677005583,
+      "loss": 3.4581,
+      "step": 2705
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6479337215423584,
+      "learning_rate": 0.0003798748082068986,
+      "loss": 3.4541,
+      "step": 2706
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6679111123085022,
+      "learning_rate": 0.0003797862263862537,
+      "loss": 3.5238,
+      "step": 2707
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6705268621444702,
+      "learning_rate": 0.00037969762225385176,
+      "loss": 3.5165,
+      "step": 2708
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6185088157653809,
+      "learning_rate": 0.00037960899582492515,
+      "loss": 3.3261,
+      "step": 2709
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.641598105430603,
+      "learning_rate": 0.00037952034711470953,
+      "loss": 3.3865,
+      "step": 2710
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6470401883125305,
+      "learning_rate": 0.00037943167613844477,
+      "loss": 3.4629,
+      "step": 2711
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6144194602966309,
+      "learning_rate": 0.0003793429829113743,
+      "loss": 3.2882,
+      "step": 2712
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6533951163291931,
+      "learning_rate": 0.0003792542674487456,
+      "loss": 3.4835,
+      "step": 2713
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6079481244087219,
+      "learning_rate": 0.00037916552976580993,
+      "loss": 3.408,
+      "step": 2714
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6616194248199463,
+      "learning_rate": 0.00037907676987782225,
+      "loss": 3.354,
+      "step": 2715
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6493124961853027,
+      "learning_rate": 0.0003789879878000415,
+      "loss": 3.3074,
+      "step": 2716
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6650919318199158,
+      "learning_rate": 0.00037889918354773016,
+      "loss": 3.2749,
+      "step": 2717
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6138140559196472,
+      "learning_rate": 0.00037881035713615495,
+      "loss": 3.3002,
+      "step": 2718
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6505001783370972,
+      "learning_rate": 0.000378721508580586,
+      "loss": 3.2198,
+      "step": 2719
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6526826024055481,
+      "learning_rate": 0.00037863263789629753,
+      "loss": 3.4036,
+      "step": 2720
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.711990475654602,
+      "learning_rate": 0.00037854374509856733,
+      "loss": 3.4434,
+      "step": 2721
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6838361620903015,
+      "learning_rate": 0.00037845483020267713,
+      "loss": 3.4861,
+      "step": 2722
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7093937993049622,
+      "learning_rate": 0.00037836589322391234,
+      "loss": 3.3011,
+      "step": 2723
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6648284196853638,
+      "learning_rate": 0.00037827693417756254,
+      "loss": 3.3851,
+      "step": 2724
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6639228463172913,
+      "learning_rate": 0.00037818795307892057,
+      "loss": 3.4538,
+      "step": 2725
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6106069087982178,
+      "learning_rate": 0.0003780989499432833,
+      "loss": 3.3033,
+      "step": 2726
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6387519240379333,
+      "learning_rate": 0.0003780099247859516,
+      "loss": 3.371,
+      "step": 2727
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6642487049102783,
+      "learning_rate": 0.0003779208776222298,
+      "loss": 3.4189,
+      "step": 2728
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6124227046966553,
+      "learning_rate": 0.00037783180846742617,
+      "loss": 3.3224,
+      "step": 2729
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6687057614326477,
+      "learning_rate": 0.0003777427173368526,
+      "loss": 3.4333,
+      "step": 2730
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6118618249893188,
+      "learning_rate": 0.000377653604245825,
+      "loss": 3.4947,
+      "step": 2731
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6521013379096985,
+      "learning_rate": 0.00037756446920966303,
+      "loss": 3.3169,
+      "step": 2732
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6459051370620728,
+      "learning_rate": 0.00037747531224368994,
+      "loss": 3.4129,
+      "step": 2733
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6482908725738525,
+      "learning_rate": 0.0003773861333632328,
+      "loss": 3.4848,
+      "step": 2734
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6932248473167419,
+      "learning_rate": 0.00037729693258362263,
+      "loss": 3.3471,
+      "step": 2735
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7700514197349548,
+      "learning_rate": 0.00037720770992019393,
+      "loss": 3.4745,
+      "step": 2736
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6803534030914307,
+      "learning_rate": 0.0003771184653882852,
+      "loss": 3.4634,
+      "step": 2737
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.655329704284668,
+      "learning_rate": 0.00037702919900323856,
+      "loss": 3.4858,
+      "step": 2738
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.666143536567688,
+      "learning_rate": 0.0003769399107804,
+      "loss": 3.3936,
+      "step": 2739
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6388490200042725,
+      "learning_rate": 0.0003768506007351191,
+      "loss": 3.3838,
+      "step": 2740
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6686370372772217,
+      "learning_rate": 0.00037676126888274943,
+      "loss": 3.4085,
+      "step": 2741
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7050430178642273,
+      "learning_rate": 0.0003766719152386481,
+      "loss": 3.3971,
+      "step": 2742
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.687915563583374,
+      "learning_rate": 0.00037658253981817603,
+      "loss": 3.4354,
+      "step": 2743
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6310634016990662,
+      "learning_rate": 0.00037649314263669785,
+      "loss": 3.4568,
+      "step": 2744
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6493231654167175,
+      "learning_rate": 0.000376403723709582,
+      "loss": 3.3632,
+      "step": 2745
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.669558048248291,
+      "learning_rate": 0.0003763142830522007,
+      "loss": 3.3468,
+      "step": 2746
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.721985399723053,
+      "learning_rate": 0.00037622482067992976,
+      "loss": 3.4897,
+      "step": 2747
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6772878766059875,
+      "learning_rate": 0.0003761353366081488,
+      "loss": 3.5065,
+      "step": 2748
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7323068380355835,
+      "learning_rate": 0.00037604583085224126,
+      "loss": 3.5024,
+      "step": 2749
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6881877779960632,
+      "learning_rate": 0.0003759563034275941,
+      "loss": 3.4767,
+      "step": 2750
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7242856621742249,
+      "learning_rate": 0.0003758667543495982,
+      "loss": 3.5577,
+      "step": 2751
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6376597285270691,
+      "learning_rate": 0.000375777183633648,
+      "loss": 3.5718,
+      "step": 2752
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6114027500152588,
+      "learning_rate": 0.00037568759129514185,
+      "loss": 3.324,
+      "step": 2753
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6938390731811523,
+      "learning_rate": 0.0003755979773494816,
+      "loss": 3.3563,
+      "step": 2754
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6962039470672607,
+      "learning_rate": 0.000375508341812073,
+      "loss": 3.4616,
+      "step": 2755
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6147626042366028,
+      "learning_rate": 0.00037541868469832547,
+      "loss": 3.3883,
+      "step": 2756
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.646985650062561,
+      "learning_rate": 0.00037532900602365205,
+      "loss": 3.428,
+      "step": 2757
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.669619083404541,
+      "learning_rate": 0.00037523930580346955,
+      "loss": 3.43,
+      "step": 2758
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.720965564250946,
+      "learning_rate": 0.00037514958405319846,
+      "loss": 3.5539,
+      "step": 2759
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7012642025947571,
+      "learning_rate": 0.00037505984078826303,
+      "loss": 3.4185,
+      "step": 2760
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6531978845596313,
+      "learning_rate": 0.0003749700760240912,
+      "loss": 3.355,
+      "step": 2761
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6693536043167114,
+      "learning_rate": 0.0003748802897761144,
+      "loss": 3.4314,
+      "step": 2762
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6955552697181702,
+      "learning_rate": 0.0003747904820597682,
+      "loss": 3.308,
+      "step": 2763
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6971192955970764,
+      "learning_rate": 0.00037470065289049135,
+      "loss": 3.4387,
+      "step": 2764
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6421964764595032,
+      "learning_rate": 0.0003746108022837266,
+      "loss": 3.4253,
+      "step": 2765
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6529431939125061,
+      "learning_rate": 0.0003745209302549204,
+      "loss": 3.3858,
+      "step": 2766
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6435075998306274,
+      "learning_rate": 0.0003744310368195227,
+      "loss": 3.3976,
+      "step": 2767
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5953412055969238,
+      "learning_rate": 0.0003743411219929872,
+      "loss": 3.3261,
+      "step": 2768
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6816799640655518,
+      "learning_rate": 0.0003742511857907713,
+      "loss": 3.4159,
+      "step": 2769
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6983208656311035,
+      "learning_rate": 0.0003741612282283362,
+      "loss": 3.5298,
+      "step": 2770
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6705322861671448,
+      "learning_rate": 0.0003740712493211466,
+      "loss": 3.3556,
+      "step": 2771
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.640531599521637,
+      "learning_rate": 0.0003739812490846708,
+      "loss": 3.3856,
+      "step": 2772
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6185344457626343,
+      "learning_rate": 0.000373891227534381,
+      "loss": 3.2589,
+      "step": 2773
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6326441168785095,
+      "learning_rate": 0.0003738011846857529,
+      "loss": 3.309,
+      "step": 2774
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6673996448516846,
+      "learning_rate": 0.000373711120554266,
+      "loss": 3.4861,
+      "step": 2775
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6781924962997437,
+      "learning_rate": 0.0003736210351554032,
+      "loss": 3.5098,
+      "step": 2776
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6595224142074585,
+      "learning_rate": 0.0003735309285046513,
+      "loss": 3.5276,
+      "step": 2777
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6162325143814087,
+      "learning_rate": 0.0003734408006175008,
+      "loss": 3.2696,
+      "step": 2778
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6548352241516113,
+      "learning_rate": 0.00037335065150944556,
+      "loss": 3.4189,
+      "step": 2779
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6298694610595703,
+      "learning_rate": 0.0003732604811959834,
+      "loss": 3.5247,
+      "step": 2780
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.632470428943634,
+      "learning_rate": 0.00037317028969261547,
+      "loss": 3.3577,
+      "step": 2781
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6624454855918884,
+      "learning_rate": 0.00037308007701484684,
+      "loss": 3.3773,
+      "step": 2782
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.641908586025238,
+      "learning_rate": 0.00037298984317818613,
+      "loss": 3.4905,
+      "step": 2783
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6339795589447021,
+      "learning_rate": 0.00037289958819814557,
+      "loss": 3.4135,
+      "step": 2784
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6595509052276611,
+      "learning_rate": 0.00037280931209024106,
+      "loss": 3.5151,
+      "step": 2785
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6234700679779053,
+      "learning_rate": 0.000372719014869992,
+      "loss": 3.3336,
+      "step": 2786
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6775194406509399,
+      "learning_rate": 0.0003726286965529216,
+      "loss": 3.3802,
+      "step": 2787
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6999110579490662,
+      "learning_rate": 0.00037253835715455664,
+      "loss": 3.4496,
+      "step": 2788
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6700520515441895,
+      "learning_rate": 0.00037244799669042754,
+      "loss": 3.4214,
+      "step": 2789
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.659300684928894,
+      "learning_rate": 0.00037235761517606826,
+      "loss": 3.3645,
+      "step": 2790
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6916134357452393,
+      "learning_rate": 0.00037226721262701633,
+      "loss": 3.4435,
+      "step": 2791
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6676251292228699,
+      "learning_rate": 0.00037217678905881324,
+      "loss": 3.4753,
+      "step": 2792
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6123244762420654,
+      "learning_rate": 0.00037208634448700374,
+      "loss": 3.2478,
+      "step": 2793
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6385574340820312,
+      "learning_rate": 0.00037199587892713617,
+      "loss": 3.3875,
+      "step": 2794
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6606059074401855,
+      "learning_rate": 0.0003719053923947628,
+      "loss": 3.3749,
+      "step": 2795
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6060505509376526,
+      "learning_rate": 0.0003718148849054391,
+      "loss": 3.3241,
+      "step": 2796
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6498494148254395,
+      "learning_rate": 0.00037172435647472466,
+      "loss": 3.398,
+      "step": 2797
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6244494318962097,
+      "learning_rate": 0.0003716338071181821,
+      "loss": 3.531,
+      "step": 2798
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6035483479499817,
+      "learning_rate": 0.00037154323685137803,
+      "loss": 3.4189,
+      "step": 2799
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6718020439147949,
+      "learning_rate": 0.0003714526456898824,
+      "loss": 3.4351,
+      "step": 2800
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6160071492195129,
+      "learning_rate": 0.000371362033649269,
+      "loss": 3.396,
+      "step": 2801
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.664903461933136,
+      "learning_rate": 0.00037127140074511516,
+      "loss": 3.4186,
+      "step": 2802
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6706874966621399,
+      "learning_rate": 0.0003711807469930016,
+      "loss": 3.4037,
+      "step": 2803
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6155506372451782,
+      "learning_rate": 0.0003710900724085128,
+      "loss": 3.3301,
+      "step": 2804
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.616083025932312,
+      "learning_rate": 0.00037099937700723663,
+      "loss": 3.2487,
+      "step": 2805
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6222215890884399,
+      "learning_rate": 0.00037090866080476495,
+      "loss": 3.3808,
+      "step": 2806
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5854479670524597,
+      "learning_rate": 0.00037081792381669275,
+      "loss": 3.4023,
+      "step": 2807
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6443006992340088,
+      "learning_rate": 0.0003707271660586188,
+      "loss": 3.3804,
+      "step": 2808
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6444540619850159,
+      "learning_rate": 0.0003706363875461454,
+      "loss": 3.4384,
+      "step": 2809
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6260697841644287,
+      "learning_rate": 0.00037054558829487837,
+      "loss": 3.4442,
+      "step": 2810
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6451106667518616,
+      "learning_rate": 0.00037045476832042734,
+      "loss": 3.282,
+      "step": 2811
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6747676134109497,
+      "learning_rate": 0.00037036392763840513,
+      "loss": 3.4037,
+      "step": 2812
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6337499618530273,
+      "learning_rate": 0.00037027306626442847,
+      "loss": 3.3404,
+      "step": 2813
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6735157370567322,
+      "learning_rate": 0.0003701821842141173,
+      "loss": 3.5064,
+      "step": 2814
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6192700266838074,
+      "learning_rate": 0.0003700912815030955,
+      "loss": 3.3892,
+      "step": 2815
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6197276711463928,
+      "learning_rate": 0.0003700003581469901,
+      "loss": 3.4002,
+      "step": 2816
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6331234574317932,
+      "learning_rate": 0.000369909414161432,
+      "loss": 3.434,
+      "step": 2817
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6590570211410522,
+      "learning_rate": 0.0003698184495620555,
+      "loss": 3.5184,
+      "step": 2818
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6396684050559998,
+      "learning_rate": 0.00036972746436449844,
+      "loss": 3.4706,
+      "step": 2819
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6504268050193787,
+      "learning_rate": 0.0003696364585844023,
+      "loss": 3.4953,
+      "step": 2820
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6527693867683411,
+      "learning_rate": 0.000369545432237412,
+      "loss": 3.487,
+      "step": 2821
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6094881892204285,
+      "learning_rate": 0.0003694543853391759,
+      "loss": 3.2907,
+      "step": 2822
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6652140617370605,
+      "learning_rate": 0.0003693633179053462,
+      "loss": 3.3554,
+      "step": 2823
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6317107677459717,
+      "learning_rate": 0.00036927222995157837,
+      "loss": 3.4327,
+      "step": 2824
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6838727593421936,
+      "learning_rate": 0.0003691811214935315,
+      "loss": 3.3449,
+      "step": 2825
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6608859300613403,
+      "learning_rate": 0.00036908999254686817,
+      "loss": 3.4886,
+      "step": 2826
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7537527680397034,
+      "learning_rate": 0.00036899884312725453,
+      "loss": 3.4995,
+      "step": 2827
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6652011871337891,
+      "learning_rate": 0.0003689076732503601,
+      "loss": 3.5182,
+      "step": 2828
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6610903739929199,
+      "learning_rate": 0.0003688164829318583,
+      "loss": 3.4118,
+      "step": 2829
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6542768478393555,
+      "learning_rate": 0.00036872527218742557,
+      "loss": 3.2537,
+      "step": 2830
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6550228595733643,
+      "learning_rate": 0.00036863404103274215,
+      "loss": 3.4192,
+      "step": 2831
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6263365745544434,
+      "learning_rate": 0.00036854278948349184,
+      "loss": 3.3837,
+      "step": 2832
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6227504014968872,
+      "learning_rate": 0.0003684515175553616,
+      "loss": 3.3923,
+      "step": 2833
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7218218445777893,
+      "learning_rate": 0.00036836022526404237,
+      "loss": 3.4124,
+      "step": 2834
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6374344229698181,
+      "learning_rate": 0.0003682689126252283,
+      "loss": 3.3984,
+      "step": 2835
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6587662696838379,
+      "learning_rate": 0.0003681775796546171,
+      "loss": 3.4672,
+      "step": 2836
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6785812377929688,
+      "learning_rate": 0.0003680862263679097,
+      "loss": 3.3096,
+      "step": 2837
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6638329029083252,
+      "learning_rate": 0.00036799485278081126,
+      "loss": 3.4854,
+      "step": 2838
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6179649829864502,
+      "learning_rate": 0.0003679034589090296,
+      "loss": 3.4328,
+      "step": 2839
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5974080562591553,
+      "learning_rate": 0.0003678120447682765,
+      "loss": 3.2263,
+      "step": 2840
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6987083554267883,
+      "learning_rate": 0.0003677206103742671,
+      "loss": 3.486,
+      "step": 2841
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6530375480651855,
+      "learning_rate": 0.0003676291557427201,
+      "loss": 3.3044,
+      "step": 2842
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6578952670097351,
+      "learning_rate": 0.0003675376808893575,
+      "loss": 3.5347,
+      "step": 2843
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6601744890213013,
+      "learning_rate": 0.00036744618582990496,
+      "loss": 3.2729,
+      "step": 2844
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6604732275009155,
+      "learning_rate": 0.00036735467058009153,
+      "loss": 3.5031,
+      "step": 2845
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6677465438842773,
+      "learning_rate": 0.0003672631351556498,
+      "loss": 3.3901,
+      "step": 2846
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6700396537780762,
+      "learning_rate": 0.0003671715795723156,
+      "loss": 3.4247,
+      "step": 2847
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6824732422828674,
+      "learning_rate": 0.00036708000384582854,
+      "loss": 3.2445,
+      "step": 2848
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6635594964027405,
+      "learning_rate": 0.00036698840799193153,
+      "loss": 3.5453,
+      "step": 2849
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6748292446136475,
+      "learning_rate": 0.00036689679202637094,
+      "loss": 3.4804,
+      "step": 2850
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.584503173828125,
+      "learning_rate": 0.0003668051559648965,
+      "loss": 3.4471,
+      "step": 2851
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6716008186340332,
+      "learning_rate": 0.00036671349982326173,
+      "loss": 3.3442,
+      "step": 2852
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6274146437644958,
+      "learning_rate": 0.00036662182361722333,
+      "loss": 3.3139,
+      "step": 2853
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6152539849281311,
+      "learning_rate": 0.00036653012736254136,
+      "loss": 3.4598,
+      "step": 2854
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6842306852340698,
+      "learning_rate": 0.0003664384110749797,
+      "loss": 3.4182,
+      "step": 2855
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.705602765083313,
+      "learning_rate": 0.00036634667477030515,
+      "loss": 3.3613,
+      "step": 2856
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6812083125114441,
+      "learning_rate": 0.0003662549184642885,
+      "loss": 3.3681,
+      "step": 2857
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.615364670753479,
+      "learning_rate": 0.00036616314217270354,
+      "loss": 3.3965,
+      "step": 2858
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5788252353668213,
+      "learning_rate": 0.00036607134591132783,
+      "loss": 3.3278,
+      "step": 2859
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6545475125312805,
+      "learning_rate": 0.00036597952969594217,
+      "loss": 3.4891,
+      "step": 2860
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6398218274116516,
+      "learning_rate": 0.0003658876935423307,
+      "loss": 3.429,
+      "step": 2861
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6287596821784973,
+      "learning_rate": 0.00036579583746628125,
+      "loss": 3.4181,
+      "step": 2862
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6481361985206604,
+      "learning_rate": 0.00036570396148358496,
+      "loss": 3.3941,
+      "step": 2863
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6730766892433167,
+      "learning_rate": 0.00036561206561003636,
+      "loss": 3.5306,
+      "step": 2864
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6656659841537476,
+      "learning_rate": 0.00036552014986143336,
+      "loss": 3.3658,
+      "step": 2865
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6573489904403687,
+      "learning_rate": 0.0003654282142535773,
+      "loss": 3.4707,
+      "step": 2866
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6266399621963501,
+      "learning_rate": 0.0003653362588022732,
+      "loss": 3.4187,
+      "step": 2867
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6363852620124817,
+      "learning_rate": 0.0003652442835233291,
+      "loss": 3.314,
+      "step": 2868
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6544065475463867,
+      "learning_rate": 0.0003651522884325565,
+      "loss": 3.3298,
+      "step": 2869
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6237480640411377,
+      "learning_rate": 0.0003650602735457706,
+      "loss": 3.2989,
+      "step": 2870
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.627240777015686,
+      "learning_rate": 0.0003649682388787898,
+      "loss": 3.3804,
+      "step": 2871
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7018293142318726,
+      "learning_rate": 0.000364876184447436,
+      "loss": 3.3385,
+      "step": 2872
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6876696348190308,
+      "learning_rate": 0.0003647841102675342,
+      "loss": 3.3663,
+      "step": 2873
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6464250683784485,
+      "learning_rate": 0.0003646920163549132,
+      "loss": 3.4385,
+      "step": 2874
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.703446626663208,
+      "learning_rate": 0.00036459990272540507,
+      "loss": 3.4518,
+      "step": 2875
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6278776526451111,
+      "learning_rate": 0.00036450776939484495,
+      "loss": 3.4672,
+      "step": 2876
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.622167706489563,
+      "learning_rate": 0.0003644156163790719,
+      "loss": 3.4943,
+      "step": 2877
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.643734335899353,
+      "learning_rate": 0.00036432344369392795,
+      "loss": 3.3365,
+      "step": 2878
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6084978580474854,
+      "learning_rate": 0.0003642312513552586,
+      "loss": 3.4268,
+      "step": 2879
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6124347448348999,
+      "learning_rate": 0.0003641390393789129,
+      "loss": 3.2882,
+      "step": 2880
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6220024824142456,
+      "learning_rate": 0.00036404680778074315,
+      "loss": 3.3692,
+      "step": 2881
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6398417949676514,
+      "learning_rate": 0.00036395455657660505,
+      "loss": 3.3702,
+      "step": 2882
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6671603918075562,
+      "learning_rate": 0.0003638622857823575,
+      "loss": 3.4744,
+      "step": 2883
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.663680374622345,
+      "learning_rate": 0.000363769995413863,
+      "loss": 3.4448,
+      "step": 2884
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7110529541969299,
+      "learning_rate": 0.00036367768548698745,
+      "loss": 3.4378,
+      "step": 2885
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6788253784179688,
+      "learning_rate": 0.0003635853560175998,
+      "loss": 3.4987,
+      "step": 2886
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6238468289375305,
+      "learning_rate": 0.0003634930070215727,
+      "loss": 3.5116,
+      "step": 2887
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6631554365158081,
+      "learning_rate": 0.000363400638514782,
+      "loss": 3.2929,
+      "step": 2888
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6805015206336975,
+      "learning_rate": 0.0003633082505131069,
+      "loss": 3.4705,
+      "step": 2889
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6465250849723816,
+      "learning_rate": 0.0003632158430324299,
+      "loss": 3.3477,
+      "step": 2890
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5989782810211182,
+      "learning_rate": 0.000363123416088637,
+      "loss": 3.2646,
+      "step": 2891
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6498996019363403,
+      "learning_rate": 0.0003630309696976175,
+      "loss": 3.5004,
+      "step": 2892
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6347329020500183,
+      "learning_rate": 0.00036293850387526377,
+      "loss": 3.4345,
+      "step": 2893
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6744681000709534,
+      "learning_rate": 0.000362846018637472,
+      "loss": 3.3837,
+      "step": 2894
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6721587181091309,
+      "learning_rate": 0.00036275351400014144,
+      "loss": 3.3357,
+      "step": 2895
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6546547412872314,
+      "learning_rate": 0.00036266098997917476,
+      "loss": 3.4083,
+      "step": 2896
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6251435875892639,
+      "learning_rate": 0.00036256844659047767,
+      "loss": 3.4083,
+      "step": 2897
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6282581686973572,
+      "learning_rate": 0.0003624758838499596,
+      "loss": 3.4596,
+      "step": 2898
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6379485130310059,
+      "learning_rate": 0.0003623833017735333,
+      "loss": 3.3715,
+      "step": 2899
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6379944682121277,
+      "learning_rate": 0.00036229070037711446,
+      "loss": 3.2941,
+      "step": 2900
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7039225697517395,
+      "learning_rate": 0.0003621980796766225,
+      "loss": 3.4274,
+      "step": 2901
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5860920548439026,
+      "learning_rate": 0.0003621054396879798,
+      "loss": 3.333,
+      "step": 2902
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.688922107219696,
+      "learning_rate": 0.00036201278042711257,
+      "loss": 3.4805,
+      "step": 2903
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6412578821182251,
+      "learning_rate": 0.0003619201019099497,
+      "loss": 3.4894,
+      "step": 2904
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7155341506004333,
+      "learning_rate": 0.0003618274041524239,
+      "loss": 3.4138,
+      "step": 2905
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7104753255844116,
+      "learning_rate": 0.00036173468717047086,
+      "loss": 3.4911,
+      "step": 2906
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6984987854957581,
+      "learning_rate": 0.0003616419509800297,
+      "loss": 3.4344,
+      "step": 2907
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6836965084075928,
+      "learning_rate": 0.00036154919559704303,
+      "loss": 3.2565,
+      "step": 2908
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6347119808197021,
+      "learning_rate": 0.0003614564210374563,
+      "loss": 3.3677,
+      "step": 2909
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.577217698097229,
+      "learning_rate": 0.00036136362731721886,
+      "loss": 3.3257,
+      "step": 2910
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6259645819664001,
+      "learning_rate": 0.0003612708144522827,
+      "loss": 3.5003,
+      "step": 2911
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6633533239364624,
+      "learning_rate": 0.00036117798245860364,
+      "loss": 3.3711,
+      "step": 2912
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6357201337814331,
+      "learning_rate": 0.0003610851313521405,
+      "loss": 3.4742,
+      "step": 2913
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6130713820457458,
+      "learning_rate": 0.00036099226114885545,
+      "loss": 3.3955,
+      "step": 2914
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6467316150665283,
+      "learning_rate": 0.000360899371864714,
+      "loss": 3.5528,
+      "step": 2915
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6171731948852539,
+      "learning_rate": 0.00036080646351568485,
+      "loss": 3.3667,
+      "step": 2916
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6676499843597412,
+      "learning_rate": 0.0003607135361177401,
+      "loss": 3.3411,
+      "step": 2917
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6495273113250732,
+      "learning_rate": 0.000360620589686855,
+      "loss": 3.3802,
+      "step": 2918
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6387211680412292,
+      "learning_rate": 0.000360527624239008,
+      "loss": 3.3713,
+      "step": 2919
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6411932110786438,
+      "learning_rate": 0.0003604346397901811,
+      "loss": 3.3835,
+      "step": 2920
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.614315927028656,
+      "learning_rate": 0.0003603416363563593,
+      "loss": 3.3809,
+      "step": 2921
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6927322745323181,
+      "learning_rate": 0.0003602486139535311,
+      "loss": 3.4335,
+      "step": 2922
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6915454864501953,
+      "learning_rate": 0.000360155572597688,
+      "loss": 3.3863,
+      "step": 2923
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6321678161621094,
+      "learning_rate": 0.000360062512304825,
+      "loss": 3.4565,
+      "step": 2924
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6756112575531006,
+      "learning_rate": 0.0003599694330909401,
+      "loss": 3.4476,
+      "step": 2925
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6556795239448547,
+      "learning_rate": 0.00035987633497203483,
+      "loss": 3.1846,
+      "step": 2926
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6876122355461121,
+      "learning_rate": 0.00035978321796411374,
+      "loss": 3.3614,
+      "step": 2927
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.68753981590271,
+      "learning_rate": 0.00035969008208318483,
+      "loss": 3.3603,
+      "step": 2928
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6918085813522339,
+      "learning_rate": 0.0003595969273452591,
+      "loss": 3.2825,
+      "step": 2929
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6255301833152771,
+      "learning_rate": 0.00035950375376635104,
+      "loss": 3.4746,
+      "step": 2930
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6332477331161499,
+      "learning_rate": 0.00035941056136247826,
+      "loss": 3.3609,
+      "step": 2931
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6548829674720764,
+      "learning_rate": 0.0003593173501496616,
+      "loss": 3.4232,
+      "step": 2932
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6732133030891418,
+      "learning_rate": 0.00035922412014392503,
+      "loss": 3.4736,
+      "step": 2933
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6473472118377686,
+      "learning_rate": 0.0003591308713612961,
+      "loss": 3.4885,
+      "step": 2934
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7150130271911621,
+      "learning_rate": 0.0003590376038178051,
+      "loss": 3.352,
+      "step": 2935
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6881107687950134,
+      "learning_rate": 0.00035894431752948603,
+      "loss": 3.4317,
+      "step": 2936
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6364063024520874,
+      "learning_rate": 0.0003588510125123757,
+      "loss": 3.4329,
+      "step": 2937
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.616174042224884,
+      "learning_rate": 0.00035875768878251444,
+      "loss": 3.4787,
+      "step": 2938
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7194497585296631,
+      "learning_rate": 0.0003586643463559457,
+      "loss": 3.5125,
+      "step": 2939
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6634018421173096,
+      "learning_rate": 0.000358570985248716,
+      "loss": 3.5313,
+      "step": 2940
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6058551073074341,
+      "learning_rate": 0.00035847760547687535,
+      "loss": 3.3203,
+      "step": 2941
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6337698698043823,
+      "learning_rate": 0.0003583842070564768,
+      "loss": 3.4544,
+      "step": 2942
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6180306077003479,
+      "learning_rate": 0.00035829079000357643,
+      "loss": 3.4748,
+      "step": 2943
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6458591222763062,
+      "learning_rate": 0.0003581973543342338,
+      "loss": 3.3677,
+      "step": 2944
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.669301450252533,
+      "learning_rate": 0.00035810390006451177,
+      "loss": 3.3496,
+      "step": 2945
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6883298754692078,
+      "learning_rate": 0.00035801042721047606,
+      "loss": 3.4491,
+      "step": 2946
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.565108597278595,
+      "learning_rate": 0.00035791693578819564,
+      "loss": 3.2918,
+      "step": 2947
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5855687856674194,
+      "learning_rate": 0.00035782342581374294,
+      "loss": 3.4187,
+      "step": 2948
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6642069220542908,
+      "learning_rate": 0.0003577298973031933,
+      "loss": 3.2816,
+      "step": 2949
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6544449925422668,
+      "learning_rate": 0.0003576363502726255,
+      "loss": 3.4615,
+      "step": 2950
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6845769882202148,
+      "learning_rate": 0.0003575427847381212,
+      "loss": 3.4364,
+      "step": 2951
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6053401827812195,
+      "learning_rate": 0.0003574492007157655,
+      "loss": 3.4218,
+      "step": 2952
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6466869711875916,
+      "learning_rate": 0.0003573555982216465,
+      "loss": 3.4526,
+      "step": 2953
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6494457721710205,
+      "learning_rate": 0.0003572619772718556,
+      "loss": 3.5021,
+      "step": 2954
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6246287226676941,
+      "learning_rate": 0.0003571683378824873,
+      "loss": 3.2659,
+      "step": 2955
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6110139489173889,
+      "learning_rate": 0.00035707468006963936,
+      "loss": 3.3118,
+      "step": 2956
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7253236174583435,
+      "learning_rate": 0.00035698100384941257,
+      "loss": 3.3502,
+      "step": 2957
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6442704796791077,
+      "learning_rate": 0.0003568873092379109,
+      "loss": 3.4506,
+      "step": 2958
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6649218201637268,
+      "learning_rate": 0.00035679359625124174,
+      "loss": 3.3949,
+      "step": 2959
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6611589789390564,
+      "learning_rate": 0.00035669986490551535,
+      "loss": 3.4551,
+      "step": 2960
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7117242217063904,
+      "learning_rate": 0.00035660611521684516,
+      "loss": 3.5832,
+      "step": 2961
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6103567481040955,
+      "learning_rate": 0.0003565123472013478,
+      "loss": 3.3479,
+      "step": 2962
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.630219042301178,
+      "learning_rate": 0.00035641856087514337,
+      "loss": 3.3387,
+      "step": 2963
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5978261828422546,
+      "learning_rate": 0.0003563247562543545,
+      "loss": 3.3176,
+      "step": 2964
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6091091632843018,
+      "learning_rate": 0.00035623093335510736,
+      "loss": 3.2688,
+      "step": 2965
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6181175708770752,
+      "learning_rate": 0.0003561370921935313,
+      "loss": 3.2935,
+      "step": 2966
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6397256851196289,
+      "learning_rate": 0.00035604323278575856,
+      "loss": 3.3793,
+      "step": 2967
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6449773907661438,
+      "learning_rate": 0.0003559493551479249,
+      "loss": 3.4294,
+      "step": 2968
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6925517320632935,
+      "learning_rate": 0.00035585545929616867,
+      "loss": 3.4042,
+      "step": 2969
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6502744555473328,
+      "learning_rate": 0.0003557615452466319,
+      "loss": 3.5209,
+      "step": 2970
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6259154677391052,
+      "learning_rate": 0.0003556676130154594,
+      "loss": 3.3345,
+      "step": 2971
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6316712498664856,
+      "learning_rate": 0.00035557366261879916,
+      "loss": 3.304,
+      "step": 2972
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5849267840385437,
+      "learning_rate": 0.0003554796940728024,
+      "loss": 3.2746,
+      "step": 2973
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6785985231399536,
+      "learning_rate": 0.0003553857073936235,
+      "loss": 3.5004,
+      "step": 2974
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6512728333473206,
+      "learning_rate": 0.00035529170259741973,
+      "loss": 3.354,
+      "step": 2975
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6976585388183594,
+      "learning_rate": 0.00035519767970035147,
+      "loss": 3.3957,
+      "step": 2976
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6594891548156738,
+      "learning_rate": 0.0003551036387185827,
+      "loss": 3.2895,
+      "step": 2977
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7098269462585449,
+      "learning_rate": 0.00035500957966827994,
+      "loss": 3.5641,
+      "step": 2978
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6764531135559082,
+      "learning_rate": 0.000354915502565613,
+      "loss": 3.4566,
+      "step": 2979
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6904491186141968,
+      "learning_rate": 0.000354821407426755,
+      "loss": 3.3431,
+      "step": 2980
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6289408206939697,
+      "learning_rate": 0.00035472729426788176,
+      "loss": 3.4602,
+      "step": 2981
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5534930229187012,
+      "learning_rate": 0.0003546331631051726,
+      "loss": 3.2183,
+      "step": 2982
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6324417591094971,
+      "learning_rate": 0.0003545390139548096,
+      "loss": 3.3314,
+      "step": 2983
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6006765961647034,
+      "learning_rate": 0.0003544448468329783,
+      "loss": 3.4489,
+      "step": 2984
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3233200311660767,
+      "learning_rate": 0.0003543506617558669,
+      "loss": 3.4516,
+      "step": 2985
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6815298199653625,
+      "learning_rate": 0.0003542564587396671,
+      "loss": 3.3749,
+      "step": 2986
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5787482857704163,
+      "learning_rate": 0.0003541622378005733,
+      "loss": 3.3712,
+      "step": 2987
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6376410126686096,
+      "learning_rate": 0.0003540679989547833,
+      "loss": 3.2685,
+      "step": 2988
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6883410811424255,
+      "learning_rate": 0.00035397374221849786,
+      "loss": 3.5234,
+      "step": 2989
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7224739789962769,
+      "learning_rate": 0.00035387946760792073,
+      "loss": 3.576,
+      "step": 2990
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6121779084205627,
+      "learning_rate": 0.00035378517513925885,
+      "loss": 3.2946,
+      "step": 2991
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.710949718952179,
+      "learning_rate": 0.0003536908648287222,
+      "loss": 3.4816,
+      "step": 2992
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5981950163841248,
+      "learning_rate": 0.0003535965366925238,
+      "loss": 3.4682,
+      "step": 2993
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6085212826728821,
+      "learning_rate": 0.0003535021907468797,
+      "loss": 3.2202,
+      "step": 2994
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6419751644134521,
+      "learning_rate": 0.0003534078270080091,
+      "loss": 3.344,
+      "step": 2995
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6410698890686035,
+      "learning_rate": 0.00035331344549213435,
+      "loss": 3.3171,
+      "step": 2996
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.66268390417099,
+      "learning_rate": 0.0003532190462154805,
+      "loss": 3.3896,
+      "step": 2997
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6045586466789246,
+      "learning_rate": 0.000353124629194276,
+      "loss": 3.3563,
+      "step": 2998
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6830160021781921,
+      "learning_rate": 0.00035303019444475224,
+      "loss": 3.4542,
+      "step": 2999
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6896335482597351,
+      "learning_rate": 0.0003529357419831437,
+      "loss": 3.4167,
+      "step": 3000
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6890286803245544,
+      "learning_rate": 0.00035284127182568767,
+      "loss": 3.3208,
+      "step": 3001
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6445521712303162,
+      "learning_rate": 0.0003527467839886248,
+      "loss": 3.3125,
+      "step": 3002
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6485596299171448,
+      "learning_rate": 0.00035265227848819866,
+      "loss": 3.4856,
+      "step": 3003
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5812082886695862,
+      "learning_rate": 0.0003525577553406557,
+      "loss": 3.4143,
+      "step": 3004
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6749499440193176,
+      "learning_rate": 0.0003524632145622457,
+      "loss": 3.3787,
+      "step": 3005
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6414878964424133,
+      "learning_rate": 0.0003523686561692213,
+      "loss": 3.4175,
+      "step": 3006
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6318088173866272,
+      "learning_rate": 0.00035227408017783813,
+      "loss": 3.337,
+      "step": 3007
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6485508680343628,
+      "learning_rate": 0.0003521794866043549,
+      "loss": 3.4666,
+      "step": 3008
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6505287289619446,
+      "learning_rate": 0.0003520848754650333,
+      "loss": 3.4531,
+      "step": 3009
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6688392758369446,
+      "learning_rate": 0.00035199024677613826,
+      "loss": 3.5422,
+      "step": 3010
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6040188074111938,
+      "learning_rate": 0.00035189560055393744,
+      "loss": 3.2983,
+      "step": 3011
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6814796924591064,
+      "learning_rate": 0.0003518009368147015,
+      "loss": 3.4625,
+      "step": 3012
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6356769800186157,
+      "learning_rate": 0.00035170625557470433,
+      "loss": 3.4569,
+      "step": 3013
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6647804975509644,
+      "learning_rate": 0.0003516115568502228,
+      "loss": 3.3751,
+      "step": 3014
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.651704728603363,
+      "learning_rate": 0.00035151684065753665,
+      "loss": 3.3811,
+      "step": 3015
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6553705930709839,
+      "learning_rate": 0.00035142210701292873,
+      "loss": 3.522,
+      "step": 3016
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6140713095664978,
+      "learning_rate": 0.0003513273559326848,
+      "loss": 3.371,
+      "step": 3017
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.636650562286377,
+      "learning_rate": 0.00035123258743309375,
+      "loss": 3.3064,
+      "step": 3018
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6213472485542297,
+      "learning_rate": 0.0003511378015304473,
+      "loss": 3.5391,
+      "step": 3019
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6385437846183777,
+      "learning_rate": 0.00035104299824104026,
+      "loss": 3.3907,
+      "step": 3020
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6367380023002625,
+      "learning_rate": 0.0003509481775811705,
+      "loss": 3.4114,
+      "step": 3021
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6261208057403564,
+      "learning_rate": 0.0003508533395671386,
+      "loss": 3.3958,
+      "step": 3022
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6035501956939697,
+      "learning_rate": 0.00035075848421524847,
+      "loss": 3.3486,
+      "step": 3023
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6485708951950073,
+      "learning_rate": 0.0003506636115418069,
+      "loss": 3.4355,
+      "step": 3024
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6530906558036804,
+      "learning_rate": 0.0003505687215631235,
+      "loss": 3.4487,
+      "step": 3025
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6586419939994812,
+      "learning_rate": 0.0003504738142955109,
+      "loss": 3.4389,
+      "step": 3026
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6540868282318115,
+      "learning_rate": 0.0003503788897552848,
+      "loss": 3.3689,
+      "step": 3027
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6517311930656433,
+      "learning_rate": 0.00035028394795876396,
+      "loss": 3.5082,
+      "step": 3028
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5996099710464478,
+      "learning_rate": 0.00035018898892226984,
+      "loss": 3.3514,
+      "step": 3029
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6310666799545288,
+      "learning_rate": 0.00035009401266212706,
+      "loss": 3.3497,
+      "step": 3030
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.623947262763977,
+      "learning_rate": 0.00034999901919466304,
+      "loss": 3.2739,
+      "step": 3031
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6056064963340759,
+      "learning_rate": 0.00034990400853620836,
+      "loss": 3.4216,
+      "step": 3032
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6672126054763794,
+      "learning_rate": 0.00034980898070309643,
+      "loss": 3.4154,
+      "step": 3033
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6156476736068726,
+      "learning_rate": 0.0003497139357116637,
+      "loss": 3.3667,
+      "step": 3034
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6445411443710327,
+      "learning_rate": 0.00034961887357824934,
+      "loss": 3.345,
+      "step": 3035
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6492227911949158,
+      "learning_rate": 0.0003495237943191957,
+      "loss": 3.333,
+      "step": 3036
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6277163624763489,
+      "learning_rate": 0.00034942869795084815,
+      "loss": 3.4001,
+      "step": 3037
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7017390131950378,
+      "learning_rate": 0.0003493335844895547,
+      "loss": 3.3455,
+      "step": 3038
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6569957733154297,
+      "learning_rate": 0.00034923845395166655,
+      "loss": 3.4971,
+      "step": 3039
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6246014833450317,
+      "learning_rate": 0.00034914330635353765,
+      "loss": 3.3908,
+      "step": 3040
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6579762697219849,
+      "learning_rate": 0.00034904814171152497,
+      "loss": 3.4869,
+      "step": 3041
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6611732840538025,
+      "learning_rate": 0.00034895296004198854,
+      "loss": 3.3098,
+      "step": 3042
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6445024609565735,
+      "learning_rate": 0.0003488577613612911,
+      "loss": 3.4028,
+      "step": 3043
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7002721428871155,
+      "learning_rate": 0.0003487625456857984,
+      "loss": 3.3255,
+      "step": 3044
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6145609617233276,
+      "learning_rate": 0.00034866731303187916,
+      "loss": 3.4165,
+      "step": 3045
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6688539385795593,
+      "learning_rate": 0.00034857206341590497,
+      "loss": 3.3142,
+      "step": 3046
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7072897553443909,
+      "learning_rate": 0.00034847679685425024,
+      "loss": 3.4557,
+      "step": 3047
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.664849579334259,
+      "learning_rate": 0.00034838151336329267,
+      "loss": 3.5348,
+      "step": 3048
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6490936875343323,
+      "learning_rate": 0.0003482862129594123,
+      "loss": 3.3588,
+      "step": 3049
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5882840156555176,
+      "learning_rate": 0.00034819089565899246,
+      "loss": 3.2787,
+      "step": 3050
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.622391402721405,
+      "learning_rate": 0.00034809556147841935,
+      "loss": 3.4638,
+      "step": 3051
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6462031602859497,
+      "learning_rate": 0.0003480002104340821,
+      "loss": 3.4412,
+      "step": 3052
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7033052444458008,
+      "learning_rate": 0.0003479048425423725,
+      "loss": 3.4586,
+      "step": 3053
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6103463768959045,
+      "learning_rate": 0.0003478094578196854,
+      "loss": 3.2599,
+      "step": 3054
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6081366539001465,
+      "learning_rate": 0.00034771405628241867,
+      "loss": 3.4528,
+      "step": 3055
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.612507164478302,
+      "learning_rate": 0.00034761863794697287,
+      "loss": 3.4103,
+      "step": 3056
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6493777632713318,
+      "learning_rate": 0.00034752320282975145,
+      "loss": 3.3473,
+      "step": 3057
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7184935212135315,
+      "learning_rate": 0.00034742775094716093,
+      "loss": 3.4663,
+      "step": 3058
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6439271569252014,
+      "learning_rate": 0.00034733228231561056,
+      "loss": 3.2053,
+      "step": 3059
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6064172983169556,
+      "learning_rate": 0.00034723679695151244,
+      "loss": 3.4793,
+      "step": 3060
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6584929823875427,
+      "learning_rate": 0.00034714129487128166,
+      "loss": 3.402,
+      "step": 3061
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6590911149978638,
+      "learning_rate": 0.00034704577609133614,
+      "loss": 3.4054,
+      "step": 3062
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6318978667259216,
+      "learning_rate": 0.00034695024062809666,
+      "loss": 3.3923,
+      "step": 3063
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6369098424911499,
+      "learning_rate": 0.0003468546884979868,
+      "loss": 3.3737,
+      "step": 3064
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9716199040412903,
+      "learning_rate": 0.0003467591197174332,
+      "loss": 3.2404,
+      "step": 3065
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.676282525062561,
+      "learning_rate": 0.00034666353430286534,
+      "loss": 3.3358,
+      "step": 3066
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.655470609664917,
+      "learning_rate": 0.0003465679322707152,
+      "loss": 3.5153,
+      "step": 3067
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6594122648239136,
+      "learning_rate": 0.00034647231363741797,
+      "loss": 3.282,
+      "step": 3068
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6830422878265381,
+      "learning_rate": 0.00034637667841941154,
+      "loss": 3.4893,
+      "step": 3069
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9497554302215576,
+      "learning_rate": 0.000346281026633137,
+      "loss": 3.3481,
+      "step": 3070
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6353447437286377,
+      "learning_rate": 0.0003461853582950377,
+      "loss": 3.3376,
+      "step": 3071
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6244276165962219,
+      "learning_rate": 0.00034608967342156016,
+      "loss": 3.3032,
+      "step": 3072
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6454404592514038,
+      "learning_rate": 0.00034599397202915384,
+      "loss": 3.3049,
+      "step": 3073
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6199542880058289,
+      "learning_rate": 0.0003458982541342709,
+      "loss": 3.3146,
+      "step": 3074
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7060653567314148,
+      "learning_rate": 0.00034580251975336636,
+      "loss": 3.4034,
+      "step": 3075
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6460721492767334,
+      "learning_rate": 0.00034570676890289794,
+      "loss": 3.3228,
+      "step": 3076
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6966212391853333,
+      "learning_rate": 0.00034561100159932653,
+      "loss": 3.5031,
+      "step": 3077
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6779978275299072,
+      "learning_rate": 0.00034551521785911537,
+      "loss": 3.466,
+      "step": 3078
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6837643980979919,
+      "learning_rate": 0.0003454194176987311,
+      "loss": 3.3513,
+      "step": 3079
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6901535391807556,
+      "learning_rate": 0.0003453236011346427,
+      "loss": 3.3933,
+      "step": 3080
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6862813830375671,
+      "learning_rate": 0.0003452277681833221,
+      "loss": 3.4137,
+      "step": 3081
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.686568558216095,
+      "learning_rate": 0.0003451319188612443,
+      "loss": 3.5128,
+      "step": 3082
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6437336802482605,
+      "learning_rate": 0.0003450360531848866,
+      "loss": 3.4203,
+      "step": 3083
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6524853706359863,
+      "learning_rate": 0.00034494017117072973,
+      "loss": 3.3342,
+      "step": 3084
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6673775911331177,
+      "learning_rate": 0.00034484427283525685,
+      "loss": 3.4508,
+      "step": 3085
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6682735681533813,
+      "learning_rate": 0.0003447483581949538,
+      "loss": 3.4717,
+      "step": 3086
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6896383762359619,
+      "learning_rate": 0.0003446524272663096,
+      "loss": 3.41,
+      "step": 3087
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6797161102294922,
+      "learning_rate": 0.0003445564800658159,
+      "loss": 3.4678,
+      "step": 3088
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7360920906066895,
+      "learning_rate": 0.00034446051660996714,
+      "loss": 3.5791,
+      "step": 3089
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6560612320899963,
+      "learning_rate": 0.00034436453691526035,
+      "loss": 3.4878,
+      "step": 3090
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7072687745094299,
+      "learning_rate": 0.00034426854099819584,
+      "loss": 3.4051,
+      "step": 3091
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1792899370193481,
+      "learning_rate": 0.00034417252887527616,
+      "loss": 3.3725,
+      "step": 3092
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6721067428588867,
+      "learning_rate": 0.0003440765005630071,
+      "loss": 3.4718,
+      "step": 3093
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6501315832138062,
+      "learning_rate": 0.0003439804560778968,
+      "loss": 3.3161,
+      "step": 3094
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6631578207015991,
+      "learning_rate": 0.00034388439543645683,
+      "loss": 3.5299,
+      "step": 3095
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6838598251342773,
+      "learning_rate": 0.0003437883186552008,
+      "loss": 3.3965,
+      "step": 3096
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6257017850875854,
+      "learning_rate": 0.0003436922257506454,
+      "loss": 3.2741,
+      "step": 3097
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6229751706123352,
+      "learning_rate": 0.00034359611673931027,
+      "loss": 3.3837,
+      "step": 3098
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6309128999710083,
+      "learning_rate": 0.0003434999916377177,
+      "loss": 3.4685,
+      "step": 3099
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6558688282966614,
+      "learning_rate": 0.0003434038504623926,
+      "loss": 3.4456,
+      "step": 3100
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6286501288414001,
+      "learning_rate": 0.00034330769322986267,
+      "loss": 3.4159,
+      "step": 3101
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6474177837371826,
+      "learning_rate": 0.0003432115199566586,
+      "loss": 3.4199,
+      "step": 3102
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5949915647506714,
+      "learning_rate": 0.00034311533065931374,
+      "loss": 3.5586,
+      "step": 3103
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.584155797958374,
+      "learning_rate": 0.00034301912535436395,
+      "loss": 3.4332,
+      "step": 3104
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6542870402336121,
+      "learning_rate": 0.0003429229040583482,
+      "loss": 3.4701,
+      "step": 3105
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.634269654750824,
+      "learning_rate": 0.0003428266667878079,
+      "loss": 3.3685,
+      "step": 3106
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7071416974067688,
+      "learning_rate": 0.0003427304135592876,
+      "loss": 3.4869,
+      "step": 3107
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6527315378189087,
+      "learning_rate": 0.00034263414438933407,
+      "loss": 3.5248,
+      "step": 3108
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.634596586227417,
+      "learning_rate": 0.00034253785929449723,
+      "loss": 3.5002,
+      "step": 3109
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6589738130569458,
+      "learning_rate": 0.00034244155829132953,
+      "loss": 3.4708,
+      "step": 3110
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6533238887786865,
+      "learning_rate": 0.0003423452413963864,
+      "loss": 3.3354,
+      "step": 3111
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6337494254112244,
+      "learning_rate": 0.00034224890862622566,
+      "loss": 3.4836,
+      "step": 3112
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6837114691734314,
+      "learning_rate": 0.00034215255999740806,
+      "loss": 3.4551,
+      "step": 3113
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6452608108520508,
+      "learning_rate": 0.00034205619552649715,
+      "loss": 3.5575,
+      "step": 3114
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6340389847755432,
+      "learning_rate": 0.00034195981523005894,
+      "loss": 3.3972,
+      "step": 3115
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6513810157775879,
+      "learning_rate": 0.0003418634191246624,
+      "loss": 3.3774,
+      "step": 3116
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.632177472114563,
+      "learning_rate": 0.0003417670072268792,
+      "loss": 3.4479,
+      "step": 3117
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6260270476341248,
+      "learning_rate": 0.00034167057955328365,
+      "loss": 3.499,
+      "step": 3118
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.66219162940979,
+      "learning_rate": 0.0003415741361204526,
+      "loss": 3.4481,
+      "step": 3119
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6389588117599487,
+      "learning_rate": 0.00034147767694496604,
+      "loss": 3.3789,
+      "step": 3120
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6469547152519226,
+      "learning_rate": 0.00034138120204340623,
+      "loss": 3.4273,
+      "step": 3121
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6160049438476562,
+      "learning_rate": 0.00034128471143235847,
+      "loss": 3.5509,
+      "step": 3122
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6240808367729187,
+      "learning_rate": 0.00034118820512841054,
+      "loss": 3.4135,
+      "step": 3123
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6506920456886292,
+      "learning_rate": 0.000341091683148153,
+      "loss": 3.4338,
+      "step": 3124
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6893854737281799,
+      "learning_rate": 0.00034099514550817904,
+      "loss": 3.3949,
+      "step": 3125
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6406022310256958,
+      "learning_rate": 0.00034089859222508474,
+      "loss": 3.5067,
+      "step": 3126
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5929657220840454,
+      "learning_rate": 0.0003408020233154686,
+      "loss": 3.3074,
+      "step": 3127
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6202057600021362,
+      "learning_rate": 0.00034070543879593206,
+      "loss": 3.3579,
+      "step": 3128
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6425173878669739,
+      "learning_rate": 0.00034060883868307893,
+      "loss": 3.3264,
+      "step": 3129
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6303429007530212,
+      "learning_rate": 0.0003405122229935161,
+      "loss": 3.4209,
+      "step": 3130
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7062411308288574,
+      "learning_rate": 0.0003404155917438528,
+      "loss": 3.4781,
+      "step": 3131
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5809078812599182,
+      "learning_rate": 0.0003403189449507012,
+      "loss": 3.3356,
+      "step": 3132
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6062414050102234,
+      "learning_rate": 0.0003402222826306757,
+      "loss": 3.3905,
+      "step": 3133
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.651741623878479,
+      "learning_rate": 0.0003401256048003939,
+      "loss": 3.3978,
+      "step": 3134
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6697308421134949,
+      "learning_rate": 0.0003400289114764759,
+      "loss": 3.4333,
+      "step": 3135
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.668836236000061,
+      "learning_rate": 0.0003399322026755443,
+      "loss": 3.3994,
+      "step": 3136
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6523700952529907,
+      "learning_rate": 0.0003398354784142244,
+      "loss": 3.4765,
+      "step": 3137
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6296630501747131,
+      "learning_rate": 0.00033973873870914433,
+      "loss": 3.4468,
+      "step": 3138
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6403985023498535,
+      "learning_rate": 0.00033964198357693475,
+      "loss": 3.4852,
+      "step": 3139
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6643931269645691,
+      "learning_rate": 0.00033954521303422905,
+      "loss": 3.3481,
+      "step": 3140
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6210636496543884,
+      "learning_rate": 0.00033944842709766307,
+      "loss": 3.4064,
+      "step": 3141
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6406473517417908,
+      "learning_rate": 0.00033935162578387545,
+      "loss": 3.356,
+      "step": 3142
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6525415778160095,
+      "learning_rate": 0.00033925480910950757,
+      "loss": 3.4282,
+      "step": 3143
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.637851893901825,
+      "learning_rate": 0.0003391579770912032,
+      "loss": 3.4665,
+      "step": 3144
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6637407541275024,
+      "learning_rate": 0.00033906112974560904,
+      "loss": 3.3821,
+      "step": 3145
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6572132110595703,
+      "learning_rate": 0.0003389642670893742,
+      "loss": 3.385,
+      "step": 3146
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6763609647750854,
+      "learning_rate": 0.0003388673891391504,
+      "loss": 3.2326,
+      "step": 3147
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.630200982093811,
+      "learning_rate": 0.00033877049591159213,
+      "loss": 3.3535,
+      "step": 3148
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.671553373336792,
+      "learning_rate": 0.00033867358742335665,
+      "loss": 3.4566,
+      "step": 3149
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6508437395095825,
+      "learning_rate": 0.0003385766636911035,
+      "loss": 3.3464,
+      "step": 3150
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6745772957801819,
+      "learning_rate": 0.00033847972473149485,
+      "loss": 3.4444,
+      "step": 3151
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5989139676094055,
+      "learning_rate": 0.0003383827705611958,
+      "loss": 3.2801,
+      "step": 3152
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6041255593299866,
+      "learning_rate": 0.00033828580119687395,
+      "loss": 3.412,
+      "step": 3153
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6439777612686157,
+      "learning_rate": 0.0003381888166551994,
+      "loss": 3.4712,
+      "step": 3154
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6362516283988953,
+      "learning_rate": 0.0003380918169528448,
+      "loss": 3.3767,
+      "step": 3155
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6359475255012512,
+      "learning_rate": 0.00033799480210648566,
+      "loss": 3.3803,
+      "step": 3156
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6420576572418213,
+      "learning_rate": 0.0003378977721327999,
+      "loss": 3.3852,
+      "step": 3157
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6671146750450134,
+      "learning_rate": 0.0003378007270484681,
+      "loss": 3.5088,
+      "step": 3158
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6650534868240356,
+      "learning_rate": 0.00033770366687017346,
+      "loss": 3.4395,
+      "step": 3159
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6656182408332825,
+      "learning_rate": 0.00033760659161460177,
+      "loss": 3.4362,
+      "step": 3160
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6491706967353821,
+      "learning_rate": 0.00033750950129844134,
+      "loss": 3.3303,
+      "step": 3161
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6670464277267456,
+      "learning_rate": 0.0003374123959383831,
+      "loss": 3.3414,
+      "step": 3162
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6329524517059326,
+      "learning_rate": 0.0003373152755511207,
+      "loss": 3.4744,
+      "step": 3163
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6125639081001282,
+      "learning_rate": 0.0003372181401533501,
+      "loss": 3.3642,
+      "step": 3164
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6461024880409241,
+      "learning_rate": 0.00033712098976177015,
+      "loss": 3.4479,
+      "step": 3165
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6762595772743225,
+      "learning_rate": 0.000337023824393082,
+      "loss": 3.3881,
+      "step": 3166
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.632960319519043,
+      "learning_rate": 0.0003369266440639897,
+      "loss": 3.476,
+      "step": 3167
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6530134677886963,
+      "learning_rate": 0.0003368294487911995,
+      "loss": 3.3649,
+      "step": 3168
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5805623531341553,
+      "learning_rate": 0.00033673223859142034,
+      "loss": 3.3062,
+      "step": 3169
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6365443468093872,
+      "learning_rate": 0.0003366350134813639,
+      "loss": 3.4325,
+      "step": 3170
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6298810839653015,
+      "learning_rate": 0.0003365377734777444,
+      "loss": 3.3691,
+      "step": 3171
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.67560875415802,
+      "learning_rate": 0.0003364405185972783,
+      "loss": 3.444,
+      "step": 3172
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6174789071083069,
+      "learning_rate": 0.000336343248856685,
+      "loss": 3.3915,
+      "step": 3173
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6995825171470642,
+      "learning_rate": 0.00033624596427268627,
+      "loss": 3.4796,
+      "step": 3174
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.676274836063385,
+      "learning_rate": 0.00033614866486200636,
+      "loss": 3.4281,
+      "step": 3175
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6361021399497986,
+      "learning_rate": 0.0003360513506413722,
+      "loss": 3.3182,
+      "step": 3176
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6266982555389404,
+      "learning_rate": 0.00033595402162751344,
+      "loss": 3.3538,
+      "step": 3177
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6403414011001587,
+      "learning_rate": 0.00033585667783716177,
+      "loss": 3.5016,
+      "step": 3178
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6274116039276123,
+      "learning_rate": 0.00033575931928705186,
+      "loss": 3.4155,
+      "step": 3179
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6344843506813049,
+      "learning_rate": 0.00033566194599392074,
+      "loss": 3.3734,
+      "step": 3180
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6367878913879395,
+      "learning_rate": 0.00033556455797450807,
+      "loss": 3.4927,
+      "step": 3181
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6092458963394165,
+      "learning_rate": 0.00033546715524555593,
+      "loss": 3.3209,
+      "step": 3182
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.64115971326828,
+      "learning_rate": 0.00033536973782380896,
+      "loss": 3.3135,
+      "step": 3183
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6055205464363098,
+      "learning_rate": 0.0003352723057260144,
+      "loss": 3.4147,
+      "step": 3184
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.65650874376297,
+      "learning_rate": 0.0003351748589689219,
+      "loss": 3.3556,
+      "step": 3185
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6385210752487183,
+      "learning_rate": 0.00033507739756928373,
+      "loss": 3.3497,
+      "step": 3186
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6704872846603394,
+      "learning_rate": 0.0003349799215438546,
+      "loss": 3.4498,
+      "step": 3187
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6331678628921509,
+      "learning_rate": 0.0003348824309093918,
+      "loss": 3.3611,
+      "step": 3188
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6341878175735474,
+      "learning_rate": 0.00033478492568265507,
+      "loss": 3.4496,
+      "step": 3189
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6627287268638611,
+      "learning_rate": 0.00033468740588040667,
+      "loss": 3.4181,
+      "step": 3190
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6297523975372314,
+      "learning_rate": 0.0003345898715194115,
+      "loss": 3.3115,
+      "step": 3191
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6300864815711975,
+      "learning_rate": 0.00033449232261643675,
+      "loss": 3.2662,
+      "step": 3192
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6536930203437805,
+      "learning_rate": 0.00033439475918825215,
+      "loss": 3.2435,
+      "step": 3193
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6698246002197266,
+      "learning_rate": 0.00033429718125163013,
+      "loss": 3.4452,
+      "step": 3194
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6215016841888428,
+      "learning_rate": 0.0003341995888233453,
+      "loss": 3.4275,
+      "step": 3195
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.623037576675415,
+      "learning_rate": 0.00033410198192017517,
+      "loss": 3.3506,
+      "step": 3196
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6066840887069702,
+      "learning_rate": 0.00033400436055889925,
+      "loss": 3.329,
+      "step": 3197
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6748824715614319,
+      "learning_rate": 0.0003339067247562999,
+      "loss": 3.4759,
+      "step": 3198
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6000560522079468,
+      "learning_rate": 0.0003338090745291619,
+      "loss": 3.256,
+      "step": 3199
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6228510141372681,
+      "learning_rate": 0.0003337114098942723,
+      "loss": 3.4539,
+      "step": 3200
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6298632025718689,
+      "learning_rate": 0.0003336137308684209,
+      "loss": 3.2606,
+      "step": 3201
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6771034598350525,
+      "learning_rate": 0.00033351603746839987,
+      "loss": 3.4276,
+      "step": 3202
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6480305790901184,
+      "learning_rate": 0.00033341832971100376,
+      "loss": 3.3418,
+      "step": 3203
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6708489656448364,
+      "learning_rate": 0.00033332060761302973,
+      "loss": 3.4634,
+      "step": 3204
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6806980967521667,
+      "learning_rate": 0.0003332228711912773,
+      "loss": 3.4818,
+      "step": 3205
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6376710534095764,
+      "learning_rate": 0.0003331251204625485,
+      "loss": 3.4056,
+      "step": 3206
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6477569937705994,
+      "learning_rate": 0.00033302735544364794,
+      "loss": 3.2535,
+      "step": 3207
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6241592764854431,
+      "learning_rate": 0.0003329295761513822,
+      "loss": 3.4414,
+      "step": 3208
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6071279048919678,
+      "learning_rate": 0.0003328317826025611,
+      "loss": 3.2777,
+      "step": 3209
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.585581362247467,
+      "learning_rate": 0.0003327339748139963,
+      "loss": 3.3901,
+      "step": 3210
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6255743503570557,
+      "learning_rate": 0.000332636152802502,
+      "loss": 3.351,
+      "step": 3211
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6320606470108032,
+      "learning_rate": 0.0003325383165848951,
+      "loss": 3.4299,
+      "step": 3212
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6342218518257141,
+      "learning_rate": 0.00033244046617799473,
+      "loss": 3.3082,
+      "step": 3213
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6162002682685852,
+      "learning_rate": 0.00033234260159862244,
+      "loss": 3.3157,
+      "step": 3214
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5912756323814392,
+      "learning_rate": 0.00033224472286360235,
+      "loss": 3.2381,
+      "step": 3215
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6357508897781372,
+      "learning_rate": 0.00033214682998976096,
+      "loss": 3.4788,
+      "step": 3216
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6640059351921082,
+      "learning_rate": 0.00033204892299392715,
+      "loss": 3.3802,
+      "step": 3217
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6069798469543457,
+      "learning_rate": 0.0003319510018929323,
+      "loss": 3.34,
+      "step": 3218
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6166238188743591,
+      "learning_rate": 0.00033185306670361015,
+      "loss": 3.4072,
+      "step": 3219
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5803248882293701,
+      "learning_rate": 0.00033175511744279695,
+      "loss": 3.3099,
+      "step": 3220
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6132253408432007,
+      "learning_rate": 0.0003316571541273312,
+      "loss": 3.3057,
+      "step": 3221
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5996150374412537,
+      "learning_rate": 0.00033155917677405404,
+      "loss": 3.4043,
+      "step": 3222
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.684146523475647,
+      "learning_rate": 0.00033146118539980886,
+      "loss": 3.3739,
+      "step": 3223
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5961861610412598,
+      "learning_rate": 0.0003313631800214416,
+      "loss": 3.371,
+      "step": 3224
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6282367706298828,
+      "learning_rate": 0.0003312651606558004,
+      "loss": 3.3869,
+      "step": 3225
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5975234508514404,
+      "learning_rate": 0.00033116712731973597,
+      "loss": 3.3008,
+      "step": 3226
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6370994448661804,
+      "learning_rate": 0.0003310690800301014,
+      "loss": 3.4603,
+      "step": 3227
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6064805388450623,
+      "learning_rate": 0.0003309710188037521,
+      "loss": 3.2699,
+      "step": 3228
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.620232105255127,
+      "learning_rate": 0.00033087294365754605,
+      "loss": 3.5315,
+      "step": 3229
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.635590136051178,
+      "learning_rate": 0.0003307748546083434,
+      "loss": 3.5964,
+      "step": 3230
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6045178174972534,
+      "learning_rate": 0.00033067675167300673,
+      "loss": 3.2993,
+      "step": 3231
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6263872385025024,
+      "learning_rate": 0.0003305786348684013,
+      "loss": 3.5154,
+      "step": 3232
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6414194703102112,
+      "learning_rate": 0.0003304805042113943,
+      "loss": 3.3522,
+      "step": 3233
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5951778292655945,
+      "learning_rate": 0.00033038235971885566,
+      "loss": 3.4194,
+      "step": 3234
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.700951337814331,
+      "learning_rate": 0.00033028420140765745,
+      "loss": 3.406,
+      "step": 3235
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6830416321754456,
+      "learning_rate": 0.0003301860292946743,
+      "loss": 3.4461,
+      "step": 3236
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.657326340675354,
+      "learning_rate": 0.0003300878433967832,
+      "loss": 3.37,
+      "step": 3237
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.665967583656311,
+      "learning_rate": 0.0003299896437308633,
+      "loss": 3.3558,
+      "step": 3238
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6418531537055969,
+      "learning_rate": 0.0003298914303137964,
+      "loss": 3.4379,
+      "step": 3239
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6427636742591858,
+      "learning_rate": 0.0003297932031624663,
+      "loss": 3.3113,
+      "step": 3240
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6846147179603577,
+      "learning_rate": 0.0003296949622937596,
+      "loss": 3.3432,
+      "step": 3241
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5913913249969482,
+      "learning_rate": 0.000329596707724565,
+      "loss": 3.3784,
+      "step": 3242
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6321174502372742,
+      "learning_rate": 0.0003294984394717736,
+      "loss": 3.4374,
+      "step": 3243
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7750719785690308,
+      "learning_rate": 0.0003294001575522788,
+      "loss": 3.3214,
+      "step": 3244
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6348200440406799,
+      "learning_rate": 0.0003293018619829764,
+      "loss": 3.3458,
+      "step": 3245
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6247170567512512,
+      "learning_rate": 0.00032920355278076466,
+      "loss": 3.3382,
+      "step": 3246
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6537448167800903,
+      "learning_rate": 0.0003291052299625439,
+      "loss": 3.5199,
+      "step": 3247
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6451489329338074,
+      "learning_rate": 0.00032900689354521705,
+      "loss": 3.3319,
+      "step": 3248
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6447813510894775,
+      "learning_rate": 0.00032890854354568935,
+      "loss": 3.3783,
+      "step": 3249
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6620329022407532,
+      "learning_rate": 0.00032881017998086817,
+      "loss": 3.439,
+      "step": 3250
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6444206237792969,
+      "learning_rate": 0.0003287118028676634,
+      "loss": 3.4197,
+      "step": 3251
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6440566778182983,
+      "learning_rate": 0.00032861341222298725,
+      "loss": 3.29,
+      "step": 3252
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.64480060338974,
+      "learning_rate": 0.0003285150080637542,
+      "loss": 3.4399,
+      "step": 3253
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6301221251487732,
+      "learning_rate": 0.00032841659040688093,
+      "loss": 3.3271,
+      "step": 3254
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6687061786651611,
+      "learning_rate": 0.00032831815926928675,
+      "loss": 3.4727,
+      "step": 3255
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7571885585784912,
+      "learning_rate": 0.00032821971466789304,
+      "loss": 3.3199,
+      "step": 3256
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6162340641021729,
+      "learning_rate": 0.0003281212566196236,
+      "loss": 3.53,
+      "step": 3257
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7048136591911316,
+      "learning_rate": 0.00032802278514140444,
+      "loss": 3.3968,
+      "step": 3258
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7401905655860901,
+      "learning_rate": 0.0003279243002501641,
+      "loss": 3.5071,
+      "step": 3259
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.66029953956604,
+      "learning_rate": 0.00032782580196283307,
+      "loss": 3.3956,
+      "step": 3260
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6099365949630737,
+      "learning_rate": 0.0003277272902963445,
+      "loss": 3.4846,
+      "step": 3261
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.636657178401947,
+      "learning_rate": 0.00032762876526763355,
+      "loss": 3.4264,
+      "step": 3262
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6442105770111084,
+      "learning_rate": 0.00032753022689363797,
+      "loss": 3.2815,
+      "step": 3263
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6299678087234497,
+      "learning_rate": 0.0003274316751912976,
+      "loss": 3.4017,
+      "step": 3264
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6098466515541077,
+      "learning_rate": 0.0003273331101775546,
+      "loss": 3.2246,
+      "step": 3265
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5735189914703369,
+      "learning_rate": 0.0003272345318693534,
+      "loss": 3.4318,
+      "step": 3266
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6352166533470154,
+      "learning_rate": 0.0003271359402836408,
+      "loss": 3.2772,
+      "step": 3267
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6368380784988403,
+      "learning_rate": 0.00032703733543736584,
+      "loss": 3.4721,
+      "step": 3268
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6893098950386047,
+      "learning_rate": 0.00032693871734747984,
+      "loss": 3.3131,
+      "step": 3269
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.660503089427948,
+      "learning_rate": 0.00032684008603093634,
+      "loss": 3.375,
+      "step": 3270
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6506989598274231,
+      "learning_rate": 0.00032674144150469133,
+      "loss": 3.3489,
+      "step": 3271
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.612584114074707,
+      "learning_rate": 0.0003266427837857027,
+      "loss": 3.2255,
+      "step": 3272
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6004265546798706,
+      "learning_rate": 0.0003265441128909311,
+      "loss": 3.4389,
+      "step": 3273
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6289479732513428,
+      "learning_rate": 0.0003264454288373392,
+      "loss": 3.5457,
+      "step": 3274
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6494922041893005,
+      "learning_rate": 0.00032634673164189175,
+      "loss": 3.4552,
+      "step": 3275
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6543040871620178,
+      "learning_rate": 0.0003262480213215561,
+      "loss": 3.3199,
+      "step": 3276
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6678844690322876,
+      "learning_rate": 0.0003261492978933016,
+      "loss": 3.3874,
+      "step": 3277
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.614801287651062,
+      "learning_rate": 0.0003260505613741,
+      "loss": 3.4826,
+      "step": 3278
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6474515199661255,
+      "learning_rate": 0.00032595181178092525,
+      "loss": 3.4237,
+      "step": 3279
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6905706524848938,
+      "learning_rate": 0.00032585304913075353,
+      "loss": 3.5723,
+      "step": 3280
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6413430571556091,
+      "learning_rate": 0.0003257542734405634,
+      "loss": 3.4314,
+      "step": 3281
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5918505787849426,
+      "learning_rate": 0.0003256554847273353,
+      "loss": 3.2961,
+      "step": 3282
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.658825695514679,
+      "learning_rate": 0.00032555668300805236,
+      "loss": 3.4234,
+      "step": 3283
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7836254835128784,
+      "learning_rate": 0.0003254578682996997,
+      "loss": 3.2429,
+      "step": 3284
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6375472545623779,
+      "learning_rate": 0.00032535904061926473,
+      "loss": 3.4131,
+      "step": 3285
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6533145308494568,
+      "learning_rate": 0.000325260199983737,
+      "loss": 3.4104,
+      "step": 3286
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6835505962371826,
+      "learning_rate": 0.00032516134641010836,
+      "loss": 3.3225,
+      "step": 3287
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6734188199043274,
+      "learning_rate": 0.000325062479915373,
+      "loss": 3.4174,
+      "step": 3288
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6586959362030029,
+      "learning_rate": 0.00032496360051652713,
+      "loss": 3.4702,
+      "step": 3289
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5997111201286316,
+      "learning_rate": 0.00032486470823056925,
+      "loss": 3.2781,
+      "step": 3290
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6650357842445374,
+      "learning_rate": 0.0003247658030745001,
+      "loss": 3.3645,
+      "step": 3291
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6417433619499207,
+      "learning_rate": 0.0003246668850653227,
+      "loss": 3.4462,
+      "step": 3292
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6094464063644409,
+      "learning_rate": 0.00032456795422004217,
+      "loss": 3.4812,
+      "step": 3293
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6310552954673767,
+      "learning_rate": 0.00032446901055566574,
+      "loss": 3.4605,
+      "step": 3294
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6238374710083008,
+      "learning_rate": 0.0003243700540892031,
+      "loss": 3.3442,
+      "step": 3295
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6162290573120117,
+      "learning_rate": 0.000324271084837666,
+      "loss": 3.4078,
+      "step": 3296
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6043792366981506,
+      "learning_rate": 0.0003241721028180684,
+      "loss": 3.3685,
+      "step": 3297
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6471230983734131,
+      "learning_rate": 0.0003240731080474265,
+      "loss": 3.26,
+      "step": 3298
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6456151008605957,
+      "learning_rate": 0.0003239741005427586,
+      "loss": 3.4192,
+      "step": 3299
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6801007986068726,
+      "learning_rate": 0.0003238750803210851,
+      "loss": 3.4669,
+      "step": 3300
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6284809708595276,
+      "learning_rate": 0.000323776047399429,
+      "loss": 3.3718,
+      "step": 3301
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5813315510749817,
+      "learning_rate": 0.000323677001794815,
+      "loss": 3.4051,
+      "step": 3302
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6104584336280823,
+      "learning_rate": 0.00032357794352427035,
+      "loss": 3.2907,
+      "step": 3303
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6234532594680786,
+      "learning_rate": 0.0003234788726048241,
+      "loss": 3.3436,
+      "step": 3304
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.633009135723114,
+      "learning_rate": 0.00032337978905350786,
+      "loss": 3.3842,
+      "step": 3305
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5902324318885803,
+      "learning_rate": 0.0003232806928873552,
+      "loss": 3.1955,
+      "step": 3306
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.636461079120636,
+      "learning_rate": 0.00032318158412340185,
+      "loss": 3.3598,
+      "step": 3307
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6180666089057922,
+      "learning_rate": 0.0003230824627786858,
+      "loss": 3.5227,
+      "step": 3308
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5958223938941956,
+      "learning_rate": 0.00032298332887024716,
+      "loss": 3.396,
+      "step": 3309
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6031109094619751,
+      "learning_rate": 0.00032288418241512826,
+      "loss": 3.3235,
+      "step": 3310
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.583393394947052,
+      "learning_rate": 0.00032278502343037344,
+      "loss": 3.379,
+      "step": 3311
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.626627504825592,
+      "learning_rate": 0.0003226858519330292,
+      "loss": 3.4544,
+      "step": 3312
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6014848351478577,
+      "learning_rate": 0.0003225866679401444,
+      "loss": 3.3026,
+      "step": 3313
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6087637543678284,
+      "learning_rate": 0.0003224874714687699,
+      "loss": 3.3001,
+      "step": 3314
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6320148706436157,
+      "learning_rate": 0.0003223882625359587,
+      "loss": 3.3306,
+      "step": 3315
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6071418523788452,
+      "learning_rate": 0.00032228904115876603,
+      "loss": 3.418,
+      "step": 3316
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6349049806594849,
+      "learning_rate": 0.00032218980735424906,
+      "loss": 3.4945,
+      "step": 3317
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6723031997680664,
+      "learning_rate": 0.00032209056113946733,
+      "loss": 3.5165,
+      "step": 3318
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5983545780181885,
+      "learning_rate": 0.0003219913025314824,
+      "loss": 3.2624,
+      "step": 3319
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5908249020576477,
+      "learning_rate": 0.00032189203154735804,
+      "loss": 3.3144,
+      "step": 3320
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6061389446258545,
+      "learning_rate": 0.0003217927482041599,
+      "loss": 3.2718,
+      "step": 3321
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6404465436935425,
+      "learning_rate": 0.0003216934525189561,
+      "loss": 3.3912,
+      "step": 3322
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6172356605529785,
+      "learning_rate": 0.00032159414450881667,
+      "loss": 3.4738,
+      "step": 3323
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6124842762947083,
+      "learning_rate": 0.0003214948241908138,
+      "loss": 3.5237,
+      "step": 3324
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.628217875957489,
+      "learning_rate": 0.00032139549158202187,
+      "loss": 3.5643,
+      "step": 3325
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6124613881111145,
+      "learning_rate": 0.00032129614669951717,
+      "loss": 3.3889,
+      "step": 3326
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6479032039642334,
+      "learning_rate": 0.0003211967895603784,
+      "loss": 3.5663,
+      "step": 3327
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6235885620117188,
+      "learning_rate": 0.00032109742018168605,
+      "loss": 3.3537,
+      "step": 3328
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6726940870285034,
+      "learning_rate": 0.000320998038580523,
+      "loss": 3.3948,
+      "step": 3329
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6240055561065674,
+      "learning_rate": 0.000320898644773974,
+      "loss": 3.2737,
+      "step": 3330
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6884067058563232,
+      "learning_rate": 0.00032079923877912606,
+      "loss": 3.3856,
+      "step": 3331
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6275650858879089,
+      "learning_rate": 0.00032069982061306814,
+      "loss": 3.4113,
+      "step": 3332
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6332144141197205,
+      "learning_rate": 0.0003206003902928914,
+      "loss": 3.2665,
+      "step": 3333
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6318677663803101,
+      "learning_rate": 0.00032050094783568915,
+      "loss": 3.2838,
+      "step": 3334
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6149763464927673,
+      "learning_rate": 0.0003204014932585567,
+      "loss": 3.2776,
+      "step": 3335
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6720367074012756,
+      "learning_rate": 0.0003203020265785914,
+      "loss": 3.4391,
+      "step": 3336
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6117024421691895,
+      "learning_rate": 0.0003202025478128926,
+      "loss": 3.2989,
+      "step": 3337
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6262743473052979,
+      "learning_rate": 0.00032010305697856214,
+      "loss": 3.4251,
+      "step": 3338
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.635115921497345,
+      "learning_rate": 0.0003200035540927034,
+      "loss": 3.5293,
+      "step": 3339
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6411104798316956,
+      "learning_rate": 0.0003199040391724221,
+      "loss": 3.3625,
+      "step": 3340
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6015012860298157,
+      "learning_rate": 0.00031980451223482615,
+      "loss": 3.288,
+      "step": 3341
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5813014507293701,
+      "learning_rate": 0.00031970497329702525,
+      "loss": 3.316,
+      "step": 3342
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.619752824306488,
+      "learning_rate": 0.00031960542237613145,
+      "loss": 3.4123,
+      "step": 3343
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6338851451873779,
+      "learning_rate": 0.0003195058594892585,
+      "loss": 3.3534,
+      "step": 3344
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.661378026008606,
+      "learning_rate": 0.00031940628465352254,
+      "loss": 3.4172,
+      "step": 3345
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5985472202301025,
+      "learning_rate": 0.0003193066978860416,
+      "loss": 3.353,
+      "step": 3346
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5869202017784119,
+      "learning_rate": 0.00031920709920393593,
+      "loss": 3.3536,
+      "step": 3347
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5923748016357422,
+      "learning_rate": 0.0003191074886243275,
+      "loss": 3.3209,
+      "step": 3348
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5979978442192078,
+      "learning_rate": 0.0003190078661643406,
+      "loss": 3.3844,
+      "step": 3349
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6152983903884888,
+      "learning_rate": 0.0003189082318411016,
+      "loss": 3.3938,
+      "step": 3350
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6220348477363586,
+      "learning_rate": 0.00031880858567173855,
+      "loss": 3.4024,
+      "step": 3351
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.650018572807312,
+      "learning_rate": 0.00031870892767338197,
+      "loss": 3.5015,
+      "step": 3352
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6519098281860352,
+      "learning_rate": 0.0003186092578631643,
+      "loss": 3.3158,
+      "step": 3353
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6117034554481506,
+      "learning_rate": 0.0003185095762582197,
+      "loss": 3.4774,
+      "step": 3354
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6011911034584045,
+      "learning_rate": 0.00031840988287568474,
+      "loss": 3.3814,
+      "step": 3355
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6428784728050232,
+      "learning_rate": 0.00031831017773269784,
+      "loss": 3.3486,
+      "step": 3356
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6304425001144409,
+      "learning_rate": 0.0003182104608463995,
+      "loss": 3.4399,
+      "step": 3357
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6032813787460327,
+      "learning_rate": 0.00031811073223393205,
+      "loss": 3.3493,
+      "step": 3358
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6222589015960693,
+      "learning_rate": 0.0003180109919124403,
+      "loss": 3.4479,
+      "step": 3359
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6171223521232605,
+      "learning_rate": 0.00031791123989907043,
+      "loss": 3.2138,
+      "step": 3360
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6898640394210815,
+      "learning_rate": 0.00031781147621097125,
+      "loss": 3.4864,
+      "step": 3361
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6144060492515564,
+      "learning_rate": 0.0003177117008652931,
+      "loss": 3.4161,
+      "step": 3362
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6295124292373657,
+      "learning_rate": 0.00031761191387918867,
+      "loss": 3.3,
+      "step": 3363
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6242650747299194,
+      "learning_rate": 0.00031751211526981235,
+      "loss": 3.3961,
+      "step": 3364
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6680017113685608,
+      "learning_rate": 0.00031741230505432066,
+      "loss": 3.4074,
+      "step": 3365
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6949132680892944,
+      "learning_rate": 0.0003173124832498724,
+      "loss": 3.3573,
+      "step": 3366
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8361984491348267,
+      "learning_rate": 0.0003172126498736278,
+      "loss": 3.396,
+      "step": 3367
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6838335990905762,
+      "learning_rate": 0.0003171128049427496,
+      "loss": 3.2927,
+      "step": 3368
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6476494073867798,
+      "learning_rate": 0.00031701294847440197,
+      "loss": 3.4097,
+      "step": 3369
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6326252222061157,
+      "learning_rate": 0.00031691308048575185,
+      "loss": 3.4594,
+      "step": 3370
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.655439019203186,
+      "learning_rate": 0.0003168132009939674,
+      "loss": 3.3764,
+      "step": 3371
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6298957467079163,
+      "learning_rate": 0.00031671331001621906,
+      "loss": 3.3612,
+      "step": 3372
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6700654625892639,
+      "learning_rate": 0.00031661340756967943,
+      "loss": 3.4241,
+      "step": 3373
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6177738308906555,
+      "learning_rate": 0.0003165134936715227,
+      "loss": 3.3666,
+      "step": 3374
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6501065492630005,
+      "learning_rate": 0.00031641356833892544,
+      "loss": 3.4278,
+      "step": 3375
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.655293345451355,
+      "learning_rate": 0.0003163136315890657,
+      "loss": 3.3245,
+      "step": 3376
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.659319281578064,
+      "learning_rate": 0.000316213683439124,
+      "loss": 3.3409,
+      "step": 3377
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.615034818649292,
+      "learning_rate": 0.0003161137239062825,
+      "loss": 3.3011,
+      "step": 3378
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.645544707775116,
+      "learning_rate": 0.00031601375300772537,
+      "loss": 3.4239,
+      "step": 3379
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.595169723033905,
+      "learning_rate": 0.0003159137707606388,
+      "loss": 3.2856,
+      "step": 3380
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.607345700263977,
+      "learning_rate": 0.0003158137771822109,
+      "loss": 3.4077,
+      "step": 3381
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5927847027778625,
+      "learning_rate": 0.00031571377228963184,
+      "loss": 3.3385,
+      "step": 3382
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6054478883743286,
+      "learning_rate": 0.0003156137561000933,
+      "loss": 3.4818,
+      "step": 3383
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6634422540664673,
+      "learning_rate": 0.00031551372863078956,
+      "loss": 3.2794,
+      "step": 3384
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.591606080532074,
+      "learning_rate": 0.0003154136898989164,
+      "loss": 3.443,
+      "step": 3385
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.672604501247406,
+      "learning_rate": 0.0003153136399216715,
+      "loss": 3.3172,
+      "step": 3386
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5803119540214539,
+      "learning_rate": 0.0003152135787162548,
+      "loss": 3.2582,
+      "step": 3387
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6184684634208679,
+      "learning_rate": 0.0003151135062998678,
+      "loss": 3.3465,
+      "step": 3388
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6289113163948059,
+      "learning_rate": 0.00031501342268971434,
+      "loss": 3.3795,
+      "step": 3389
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6137139201164246,
+      "learning_rate": 0.00031491332790299973,
+      "loss": 3.3531,
+      "step": 3390
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6423360705375671,
+      "learning_rate": 0.0003148132219569315,
+      "loss": 3.4468,
+      "step": 3391
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6035122871398926,
+      "learning_rate": 0.0003147131048687191,
+      "loss": 3.3405,
+      "step": 3392
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.598874032497406,
+      "learning_rate": 0.0003146129766555737,
+      "loss": 3.3547,
+      "step": 3393
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6974937319755554,
+      "learning_rate": 0.0003145128373347086,
+      "loss": 3.4845,
+      "step": 3394
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6463556289672852,
+      "learning_rate": 0.0003144126869233389,
+      "loss": 3.2956,
+      "step": 3395
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.69242262840271,
+      "learning_rate": 0.0003143125254386815,
+      "loss": 3.5167,
+      "step": 3396
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6210297346115112,
+      "learning_rate": 0.00031421235289795534,
+      "loss": 3.5848,
+      "step": 3397
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0055627822875977,
+      "learning_rate": 0.0003141121693183814,
+      "loss": 3.3848,
+      "step": 3398
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7247527241706848,
+      "learning_rate": 0.00031401197471718235,
+      "loss": 3.3554,
+      "step": 3399
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5828782916069031,
+      "learning_rate": 0.0003139117691115827,
+      "loss": 3.4218,
+      "step": 3400
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6309610605239868,
+      "learning_rate": 0.000313811552518809,
+      "loss": 3.433,
+      "step": 3401
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6716001033782959,
+      "learning_rate": 0.0003137113249560896,
+      "loss": 3.3232,
+      "step": 3402
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6095533967018127,
+      "learning_rate": 0.00031361108644065497,
+      "loss": 3.3684,
+      "step": 3403
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6483668088912964,
+      "learning_rate": 0.00031351083698973704,
+      "loss": 3.2485,
+      "step": 3404
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6276049613952637,
+      "learning_rate": 0.00031341057662057,
+      "loss": 3.3453,
+      "step": 3405
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6165592074394226,
+      "learning_rate": 0.00031331030535038963,
+      "loss": 3.4065,
+      "step": 3406
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5991979837417603,
+      "learning_rate": 0.00031321002319643394,
+      "loss": 3.2915,
+      "step": 3407
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6335907578468323,
+      "learning_rate": 0.00031310973017594233,
+      "loss": 3.4469,
+      "step": 3408
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6385035514831543,
+      "learning_rate": 0.00031300942630615666,
+      "loss": 3.3794,
+      "step": 3409
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6390557885169983,
+      "learning_rate": 0.00031290911160432,
+      "loss": 3.4735,
+      "step": 3410
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5964511632919312,
+      "learning_rate": 0.0003128087860876778,
+      "loss": 3.2477,
+      "step": 3411
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6422802209854126,
+      "learning_rate": 0.00031270844977347706,
+      "loss": 3.464,
+      "step": 3412
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6574854850769043,
+      "learning_rate": 0.00031260810267896686,
+      "loss": 3.3135,
+      "step": 3413
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6653745770454407,
+      "learning_rate": 0.0003125077448213981,
+      "loss": 3.4306,
+      "step": 3414
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6036469340324402,
+      "learning_rate": 0.0003124073762180232,
+      "loss": 3.4437,
+      "step": 3415
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6745374798774719,
+      "learning_rate": 0.00031230699688609693,
+      "loss": 3.2753,
+      "step": 3416
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6684151291847229,
+      "learning_rate": 0.0003122066068428756,
+      "loss": 3.3819,
+      "step": 3417
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6250439882278442,
+      "learning_rate": 0.00031210620610561736,
+      "loss": 3.3884,
+      "step": 3418
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6410674452781677,
+      "learning_rate": 0.0003120057946915822,
+      "loss": 3.3788,
+      "step": 3419
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6384410262107849,
+      "learning_rate": 0.00031190537261803215,
+      "loss": 3.4571,
+      "step": 3420
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6605309844017029,
+      "learning_rate": 0.0003118049399022309,
+      "loss": 3.3624,
+      "step": 3421
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6254794001579285,
+      "learning_rate": 0.000311704496561444,
+      "loss": 3.3778,
+      "step": 3422
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6189180016517639,
+      "learning_rate": 0.0003116040426129388,
+      "loss": 3.3505,
+      "step": 3423
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6452104449272156,
+      "learning_rate": 0.0003115035780739845,
+      "loss": 3.3617,
+      "step": 3424
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6219657063484192,
+      "learning_rate": 0.000311403102961852,
+      "loss": 3.3035,
+      "step": 3425
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6689308285713196,
+      "learning_rate": 0.0003113026172938144,
+      "loss": 3.4442,
+      "step": 3426
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6046881079673767,
+      "learning_rate": 0.0003112021210871462,
+      "loss": 3.2917,
+      "step": 3427
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6043495535850525,
+      "learning_rate": 0.00031110161435912384,
+      "loss": 3.4112,
+      "step": 3428
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6852877140045166,
+      "learning_rate": 0.0003110010971270256,
+      "loss": 3.3943,
+      "step": 3429
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6157641410827637,
+      "learning_rate": 0.0003109005694081315,
+      "loss": 3.2675,
+      "step": 3430
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5903221368789673,
+      "learning_rate": 0.0003108000312197236,
+      "loss": 3.3839,
+      "step": 3431
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6359903812408447,
+      "learning_rate": 0.0003106994825790855,
+      "loss": 3.4311,
+      "step": 3432
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6401050090789795,
+      "learning_rate": 0.0003105989235035026,
+      "loss": 3.3574,
+      "step": 3433
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5832918286323547,
+      "learning_rate": 0.0003104983540102622,
+      "loss": 3.3243,
+      "step": 3434
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6364819407463074,
+      "learning_rate": 0.0003103977741166535,
+      "loss": 3.3136,
+      "step": 3435
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5973476767539978,
+      "learning_rate": 0.0003102971838399672,
+      "loss": 3.4813,
+      "step": 3436
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6046728491783142,
+      "learning_rate": 0.0003101965831974959,
+      "loss": 3.2598,
+      "step": 3437
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6494289636611938,
+      "learning_rate": 0.00031009597220653417,
+      "loss": 3.3655,
+      "step": 3438
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6511953473091125,
+      "learning_rate": 0.00030999535088437806,
+      "loss": 3.4139,
+      "step": 3439
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6205424070358276,
+      "learning_rate": 0.0003098947192483256,
+      "loss": 3.32,
+      "step": 3440
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6335713863372803,
+      "learning_rate": 0.0003097940773156766,
+      "loss": 3.4043,
+      "step": 3441
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6370178461074829,
+      "learning_rate": 0.00030969342510373253,
+      "loss": 3.3922,
+      "step": 3442
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6742094159126282,
+      "learning_rate": 0.00030959276262979663,
+      "loss": 3.4295,
+      "step": 3443
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5739489197731018,
+      "learning_rate": 0.00030949208991117396,
+      "loss": 3.3301,
+      "step": 3444
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6312046051025391,
+      "learning_rate": 0.00030939140696517137,
+      "loss": 3.2065,
+      "step": 3445
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6611394882202148,
+      "learning_rate": 0.0003092907138090974,
+      "loss": 3.3841,
+      "step": 3446
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6275805234909058,
+      "learning_rate": 0.0003091900104602623,
+      "loss": 3.3972,
+      "step": 3447
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6278124451637268,
+      "learning_rate": 0.00030908929693597827,
+      "loss": 3.3839,
+      "step": 3448
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6052056550979614,
+      "learning_rate": 0.00030898857325355905,
+      "loss": 3.4597,
+      "step": 3449
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6291190385818481,
+      "learning_rate": 0.00030888783943032027,
+      "loss": 3.2394,
+      "step": 3450
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6044225692749023,
+      "learning_rate": 0.0003087870954835792,
+      "loss": 3.4053,
+      "step": 3451
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6691856980323792,
+      "learning_rate": 0.00030868634143065487,
+      "loss": 3.3968,
+      "step": 3452
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6347593069076538,
+      "learning_rate": 0.00030858557728886813,
+      "loss": 3.3502,
+      "step": 3453
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6214762330055237,
+      "learning_rate": 0.0003084848030755415,
+      "loss": 3.3677,
+      "step": 3454
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6205227375030518,
+      "learning_rate": 0.0003083840188079991,
+      "loss": 3.4138,
+      "step": 3455
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6083345413208008,
+      "learning_rate": 0.0003082832245035671,
+      "loss": 3.4667,
+      "step": 3456
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6635501980781555,
+      "learning_rate": 0.0003081824201795731,
+      "loss": 3.3298,
+      "step": 3457
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6177193522453308,
+      "learning_rate": 0.00030808160585334653,
+      "loss": 3.3713,
+      "step": 3458
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6598303914070129,
+      "learning_rate": 0.0003079807815422187,
+      "loss": 3.4621,
+      "step": 3459
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6189092397689819,
+      "learning_rate": 0.0003078799472635223,
+      "loss": 3.3052,
+      "step": 3460
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5945866107940674,
+      "learning_rate": 0.00030777910303459187,
+      "loss": 3.3532,
+      "step": 3461
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6236907839775085,
+      "learning_rate": 0.00030767824887276385,
+      "loss": 3.2658,
+      "step": 3462
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6429352164268494,
+      "learning_rate": 0.00030757738479537624,
+      "loss": 3.5409,
+      "step": 3463
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6596398949623108,
+      "learning_rate": 0.0003074765108197687,
+      "loss": 3.3922,
+      "step": 3464
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6131845116615295,
+      "learning_rate": 0.00030737562696328254,
+      "loss": 3.5003,
+      "step": 3465
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6321629881858826,
+      "learning_rate": 0.000307274733243261,
+      "loss": 3.3875,
+      "step": 3466
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5874630212783813,
+      "learning_rate": 0.0003071738296770488,
+      "loss": 3.3837,
+      "step": 3467
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6010831594467163,
+      "learning_rate": 0.0003070729162819925,
+      "loss": 3.3813,
+      "step": 3468
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6139161586761475,
+      "learning_rate": 0.00030697199307544014,
+      "loss": 3.3483,
+      "step": 3469
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6040095090866089,
+      "learning_rate": 0.0003068710600747418,
+      "loss": 3.4057,
+      "step": 3470
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6296358704566956,
+      "learning_rate": 0.0003067701172972489,
+      "loss": 3.3732,
+      "step": 3471
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6295837163925171,
+      "learning_rate": 0.0003066691647603147,
+      "loss": 3.3701,
+      "step": 3472
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6771396398544312,
+      "learning_rate": 0.0003065682024812941,
+      "loss": 3.31,
+      "step": 3473
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6971045732498169,
+      "learning_rate": 0.0003064672304775438,
+      "loss": 3.4291,
+      "step": 3474
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5894261598587036,
+      "learning_rate": 0.0003063662487664218,
+      "loss": 3.4009,
+      "step": 3475
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6298110485076904,
+      "learning_rate": 0.0003062652573652882,
+      "loss": 3.3064,
+      "step": 3476
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6593163013458252,
+      "learning_rate": 0.00030616425629150467,
+      "loss": 3.3307,
+      "step": 3477
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6416672468185425,
+      "learning_rate": 0.00030606324556243435,
+      "loss": 3.3134,
+      "step": 3478
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6436814665794373,
+      "learning_rate": 0.00030596222519544216,
+      "loss": 3.4234,
+      "step": 3479
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6197558641433716,
+      "learning_rate": 0.00030586119520789465,
+      "loss": 3.3476,
+      "step": 3480
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6113980412483215,
+      "learning_rate": 0.00030576015561716014,
+      "loss": 3.2472,
+      "step": 3481
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.64369136095047,
+      "learning_rate": 0.00030565910644060845,
+      "loss": 3.4599,
+      "step": 3482
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6595032811164856,
+      "learning_rate": 0.00030555804769561107,
+      "loss": 3.307,
+      "step": 3483
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6180384159088135,
+      "learning_rate": 0.00030545697939954124,
+      "loss": 3.2461,
+      "step": 3484
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6257269382476807,
+      "learning_rate": 0.00030535590156977375,
+      "loss": 3.3398,
+      "step": 3485
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6344363689422607,
+      "learning_rate": 0.00030525481422368503,
+      "loss": 3.4293,
+      "step": 3486
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6301414370536804,
+      "learning_rate": 0.00030515371737865313,
+      "loss": 3.3741,
+      "step": 3487
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6213796138763428,
+      "learning_rate": 0.00030505261105205785,
+      "loss": 3.4556,
+      "step": 3488
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.0104610919952393,
+      "learning_rate": 0.00030495149526128053,
+      "loss": 3.3195,
+      "step": 3489
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6484731435775757,
+      "learning_rate": 0.000304850370023704,
+      "loss": 3.2919,
+      "step": 3490
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6554868221282959,
+      "learning_rate": 0.0003047492353567131,
+      "loss": 3.3719,
+      "step": 3491
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5816743969917297,
+      "learning_rate": 0.000304648091277694,
+      "loss": 3.1774,
+      "step": 3492
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6390587091445923,
+      "learning_rate": 0.0003045469378040344,
+      "loss": 3.3829,
+      "step": 3493
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6130589842796326,
+      "learning_rate": 0.00030444577495312373,
+      "loss": 3.3452,
+      "step": 3494
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6152821183204651,
+      "learning_rate": 0.0003043446027423533,
+      "loss": 3.3599,
+      "step": 3495
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6135820746421814,
+      "learning_rate": 0.0003042434211891156,
+      "loss": 3.4351,
+      "step": 3496
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5926715731620789,
+      "learning_rate": 0.0003041422303108049,
+      "loss": 3.5196,
+      "step": 3497
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6294316649436951,
+      "learning_rate": 0.0003040410301248172,
+      "loss": 3.4092,
+      "step": 3498
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6111506819725037,
+      "learning_rate": 0.00030393982064854985,
+      "loss": 3.2415,
+      "step": 3499
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6075042486190796,
+      "learning_rate": 0.0003038386018994021,
+      "loss": 3.3598,
+      "step": 3500
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.586727499961853,
+      "learning_rate": 0.00030373737389477446,
+      "loss": 3.4594,
+      "step": 3501
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7585539221763611,
+      "learning_rate": 0.00030363613665206936,
+      "loss": 3.4881,
+      "step": 3502
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5955037474632263,
+      "learning_rate": 0.0003035348901886905,
+      "loss": 3.4064,
+      "step": 3503
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5939720869064331,
+      "learning_rate": 0.00030343363452204334,
+      "loss": 3.3177,
+      "step": 3504
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6512664556503296,
+      "learning_rate": 0.000303332369669535,
+      "loss": 3.3653,
+      "step": 3505
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6145567297935486,
+      "learning_rate": 0.0003032310956485741,
+      "loss": 3.3844,
+      "step": 3506
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6215113997459412,
+      "learning_rate": 0.0003031298124765707,
+      "loss": 3.3526,
+      "step": 3507
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.619853675365448,
+      "learning_rate": 0.00030302852017093654,
+      "loss": 3.4662,
+      "step": 3508
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6302598118782043,
+      "learning_rate": 0.000302927218749085,
+      "loss": 3.3894,
+      "step": 3509
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6114816069602966,
+      "learning_rate": 0.0003028259082284311,
+      "loss": 3.3717,
+      "step": 3510
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6622714400291443,
+      "learning_rate": 0.00030272458862639105,
+      "loss": 3.3851,
+      "step": 3511
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6394220590591431,
+      "learning_rate": 0.00030262325996038305,
+      "loss": 3.3535,
+      "step": 3512
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6466477513313293,
+      "learning_rate": 0.00030252192224782654,
+      "loss": 3.2252,
+      "step": 3513
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.582334578037262,
+      "learning_rate": 0.00030242057550614276,
+      "loss": 3.3672,
+      "step": 3514
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6474951505661011,
+      "learning_rate": 0.0003023192197527543,
+      "loss": 3.3184,
+      "step": 3515
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6074023246765137,
+      "learning_rate": 0.0003022178550050855,
+      "loss": 3.2978,
+      "step": 3516
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6092503070831299,
+      "learning_rate": 0.00030211648128056196,
+      "loss": 3.4776,
+      "step": 3517
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6604886054992676,
+      "learning_rate": 0.0003020150985966111,
+      "loss": 3.4106,
+      "step": 3518
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.599673330783844,
+      "learning_rate": 0.00030191370697066187,
+      "loss": 3.4092,
+      "step": 3519
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6192481517791748,
+      "learning_rate": 0.00030181230642014456,
+      "loss": 3.3152,
+      "step": 3520
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6263299584388733,
+      "learning_rate": 0.0003017108969624911,
+      "loss": 3.4797,
+      "step": 3521
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6188412308692932,
+      "learning_rate": 0.000301609478615135,
+      "loss": 3.329,
+      "step": 3522
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6249926090240479,
+      "learning_rate": 0.00030150805139551114,
+      "loss": 3.2953,
+      "step": 3523
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6381181478500366,
+      "learning_rate": 0.0003014066153210562,
+      "loss": 3.4152,
+      "step": 3524
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6533681750297546,
+      "learning_rate": 0.0003013051704092081,
+      "loss": 3.3498,
+      "step": 3525
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6145551800727844,
+      "learning_rate": 0.0003012037166774063,
+      "loss": 3.4357,
+      "step": 3526
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6011974215507507,
+      "learning_rate": 0.00030110225414309213,
+      "loss": 3.365,
+      "step": 3527
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6385951042175293,
+      "learning_rate": 0.00030100078282370804,
+      "loss": 3.3649,
+      "step": 3528
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5937651991844177,
+      "learning_rate": 0.0003008993027366981,
+      "loss": 3.271,
+      "step": 3529
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6630327701568604,
+      "learning_rate": 0.000300797813899508,
+      "loss": 3.2964,
+      "step": 3530
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7359730005264282,
+      "learning_rate": 0.00030069631632958474,
+      "loss": 3.3561,
+      "step": 3531
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.631149172782898,
+      "learning_rate": 0.00030059481004437696,
+      "loss": 3.3569,
+      "step": 3532
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6512892842292786,
+      "learning_rate": 0.00030049329506133483,
+      "loss": 3.3519,
+      "step": 3533
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6602842807769775,
+      "learning_rate": 0.0003003917713979099,
+      "loss": 3.5396,
+      "step": 3534
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6186116337776184,
+      "learning_rate": 0.0003002902390715553,
+      "loss": 3.4058,
+      "step": 3535
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6194803714752197,
+      "learning_rate": 0.00030018869809972555,
+      "loss": 3.33,
+      "step": 3536
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6519399285316467,
+      "learning_rate": 0.0003000871484998767,
+      "loss": 3.3879,
+      "step": 3537
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6017206311225891,
+      "learning_rate": 0.00029998559028946646,
+      "loss": 3.3329,
+      "step": 3538
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6431175470352173,
+      "learning_rate": 0.0002998840234859538,
+      "loss": 3.3358,
+      "step": 3539
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6184296011924744,
+      "learning_rate": 0.00029978244810679913,
+      "loss": 3.4958,
+      "step": 3540
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.625480592250824,
+      "learning_rate": 0.0002996808641694644,
+      "loss": 3.3546,
+      "step": 3541
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6198711395263672,
+      "learning_rate": 0.00029957927169141336,
+      "loss": 3.2546,
+      "step": 3542
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6898179650306702,
+      "learning_rate": 0.0002994776706901107,
+      "loss": 3.3301,
+      "step": 3543
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6591171622276306,
+      "learning_rate": 0.0002993760611830228,
+      "loss": 3.2971,
+      "step": 3544
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6145486831665039,
+      "learning_rate": 0.00029927444318761755,
+      "loss": 3.2786,
+      "step": 3545
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6447986364364624,
+      "learning_rate": 0.0002991728167213643,
+      "loss": 3.2749,
+      "step": 3546
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6319825649261475,
+      "learning_rate": 0.0002990711818017338,
+      "loss": 3.2753,
+      "step": 3547
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6972938776016235,
+      "learning_rate": 0.00029896953844619833,
+      "loss": 3.5112,
+      "step": 3548
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.618682861328125,
+      "learning_rate": 0.0002988678866722315,
+      "loss": 3.2993,
+      "step": 3549
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5898873209953308,
+      "learning_rate": 0.0002987662264973083,
+      "loss": 3.4458,
+      "step": 3550
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6917723417282104,
+      "learning_rate": 0.00029866455793890544,
+      "loss": 3.3236,
+      "step": 3551
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6008872389793396,
+      "learning_rate": 0.000298562881014501,
+      "loss": 3.4311,
+      "step": 3552
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.674096941947937,
+      "learning_rate": 0.0002984611957415743,
+      "loss": 3.3805,
+      "step": 3553
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6469094157218933,
+      "learning_rate": 0.0002983595021376062,
+      "loss": 3.3483,
+      "step": 3554
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6237218379974365,
+      "learning_rate": 0.000298257800220079,
+      "loss": 3.3999,
+      "step": 3555
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.1701252460479736,
+      "learning_rate": 0.0002981560900064766,
+      "loss": 3.4764,
+      "step": 3556
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6512659788131714,
+      "learning_rate": 0.00029805437151428403,
+      "loss": 3.3742,
+      "step": 3557
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6231496334075928,
+      "learning_rate": 0.00029795264476098786,
+      "loss": 3.2932,
+      "step": 3558
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5999680757522583,
+      "learning_rate": 0.00029785090976407614,
+      "loss": 3.3411,
+      "step": 3559
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6761013269424438,
+      "learning_rate": 0.0002977491665410383,
+      "loss": 3.472,
+      "step": 3560
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6520906090736389,
+      "learning_rate": 0.0002976474151093653,
+      "loss": 3.3184,
+      "step": 3561
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6317535638809204,
+      "learning_rate": 0.0002975456554865491,
+      "loss": 3.3547,
+      "step": 3562
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.635465681552887,
+      "learning_rate": 0.00029744388769008365,
+      "loss": 3.453,
+      "step": 3563
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6405307650566101,
+      "learning_rate": 0.0002973421117374637,
+      "loss": 3.2819,
+      "step": 3564
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.622593343257904,
+      "learning_rate": 0.00029724032764618605,
+      "loss": 3.4364,
+      "step": 3565
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6837420463562012,
+      "learning_rate": 0.00029713853543374846,
+      "loss": 3.4766,
+      "step": 3566
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6647900938987732,
+      "learning_rate": 0.0002970367351176501,
+      "loss": 3.4147,
+      "step": 3567
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6509084105491638,
+      "learning_rate": 0.00029693492671539165,
+      "loss": 3.3641,
+      "step": 3568
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5970133543014526,
+      "learning_rate": 0.0002968331102444752,
+      "loss": 3.3489,
+      "step": 3569
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7497248649597168,
+      "learning_rate": 0.0002967312857224042,
+      "loss": 3.4326,
+      "step": 3570
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6541271805763245,
+      "learning_rate": 0.0002966294531666834,
+      "loss": 3.4407,
+      "step": 3571
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6058679819107056,
+      "learning_rate": 0.000296527612594819,
+      "loss": 3.433,
+      "step": 3572
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5970894694328308,
+      "learning_rate": 0.0002964257640243186,
+      "loss": 3.381,
+      "step": 3573
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6197634339332581,
+      "learning_rate": 0.00029632390747269115,
+      "loss": 3.3566,
+      "step": 3574
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5964658260345459,
+      "learning_rate": 0.000296222042957447,
+      "loss": 3.2884,
+      "step": 3575
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5752972960472107,
+      "learning_rate": 0.0002961201704960978,
+      "loss": 3.4334,
+      "step": 3576
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6440456509590149,
+      "learning_rate": 0.00029601829010615665,
+      "loss": 3.4766,
+      "step": 3577
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6194326877593994,
+      "learning_rate": 0.0002959164018051379,
+      "loss": 3.3685,
+      "step": 3578
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6755337119102478,
+      "learning_rate": 0.0002958145056105575,
+      "loss": 3.4672,
+      "step": 3579
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6031014919281006,
+      "learning_rate": 0.00029571260153993224,
+      "loss": 3.2198,
+      "step": 3580
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.601763904094696,
+      "learning_rate": 0.000295610689610781,
+      "loss": 3.25,
+      "step": 3581
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5831199884414673,
+      "learning_rate": 0.00029550876984062337,
+      "loss": 3.315,
+      "step": 3582
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6840198040008545,
+      "learning_rate": 0.00029540684224698056,
+      "loss": 3.4738,
+      "step": 3583
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6487699151039124,
+      "learning_rate": 0.0002953049068473753,
+      "loss": 3.3388,
+      "step": 3584
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6016435027122498,
+      "learning_rate": 0.0002952029636593313,
+      "loss": 3.2886,
+      "step": 3585
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6112717390060425,
+      "learning_rate": 0.00029510101270037377,
+      "loss": 3.4304,
+      "step": 3586
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6386420130729675,
+      "learning_rate": 0.00029499905398802927,
+      "loss": 3.3713,
+      "step": 3587
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6484072804450989,
+      "learning_rate": 0.0002948970875398258,
+      "loss": 3.4334,
+      "step": 3588
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6308383345603943,
+      "learning_rate": 0.0002947951133732925,
+      "loss": 3.2712,
+      "step": 3589
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6079784035682678,
+      "learning_rate": 0.00029469313150595983,
+      "loss": 3.4595,
+      "step": 3590
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6424318552017212,
+      "learning_rate": 0.00029459114195535977,
+      "loss": 3.408,
+      "step": 3591
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5923727750778198,
+      "learning_rate": 0.00029448914473902544,
+      "loss": 3.3762,
+      "step": 3592
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5885934829711914,
+      "learning_rate": 0.0002943871398744914,
+      "loss": 3.4226,
+      "step": 3593
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6437953114509583,
+      "learning_rate": 0.00029428512737929334,
+      "loss": 3.3637,
+      "step": 3594
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6215007901191711,
+      "learning_rate": 0.00029418310727096864,
+      "loss": 3.3717,
+      "step": 3595
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6558961868286133,
+      "learning_rate": 0.00029408107956705544,
+      "loss": 3.1917,
+      "step": 3596
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.606677770614624,
+      "learning_rate": 0.0002939790442850937,
+      "loss": 3.3311,
+      "step": 3597
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5694128274917603,
+      "learning_rate": 0.0002938770014426244,
+      "loss": 3.2667,
+      "step": 3598
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6147801876068115,
+      "learning_rate": 0.00029377495105718993,
+      "loss": 3.5328,
+      "step": 3599
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5825375318527222,
+      "learning_rate": 0.00029367289314633383,
+      "loss": 3.3569,
+      "step": 3600
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6331030130386353,
+      "learning_rate": 0.00029357082772760103,
+      "loss": 3.3965,
+      "step": 3601
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6068410873413086,
+      "learning_rate": 0.000293468754818538,
+      "loss": 3.2792,
+      "step": 3602
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6427342891693115,
+      "learning_rate": 0.00029336667443669197,
+      "loss": 3.3827,
+      "step": 3603
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6383602619171143,
+      "learning_rate": 0.0002932645865996119,
+      "loss": 3.3444,
+      "step": 3604
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.671267569065094,
+      "learning_rate": 0.00029316249132484775,
+      "loss": 3.4324,
+      "step": 3605
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6190444231033325,
+      "learning_rate": 0.000293060388629951,
+      "loss": 3.3451,
+      "step": 3606
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6308755278587341,
+      "learning_rate": 0.0002929582785324743,
+      "loss": 3.4204,
+      "step": 3607
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6411634087562561,
+      "learning_rate": 0.0002928561610499714,
+      "loss": 3.434,
+      "step": 3608
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6541515588760376,
+      "learning_rate": 0.0002927540361999977,
+      "loss": 3.3911,
+      "step": 3609
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6651307940483093,
+      "learning_rate": 0.0002926519040001094,
+      "loss": 3.2249,
+      "step": 3610
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6221703290939331,
+      "learning_rate": 0.00029254976446786447,
+      "loss": 3.2465,
+      "step": 3611
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6201170086860657,
+      "learning_rate": 0.00029244761762082166,
+      "loss": 3.3435,
+      "step": 3612
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7041198015213013,
+      "learning_rate": 0.00029234546347654133,
+      "loss": 3.3102,
+      "step": 3613
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6195172667503357,
+      "learning_rate": 0.0002922433020525849,
+      "loss": 3.4033,
+      "step": 3614
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5839033722877502,
+      "learning_rate": 0.0002921411333665151,
+      "loss": 3.4524,
+      "step": 3615
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5902976393699646,
+      "learning_rate": 0.00029203895743589597,
+      "loss": 3.3454,
+      "step": 3616
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.583528459072113,
+      "learning_rate": 0.00029193677427829276,
+      "loss": 3.365,
+      "step": 3617
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6564446091651917,
+      "learning_rate": 0.0002918345839112717,
+      "loss": 3.3221,
+      "step": 3618
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6363897919654846,
+      "learning_rate": 0.0002917323863524008,
+      "loss": 3.349,
+      "step": 3619
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6170995831489563,
+      "learning_rate": 0.00029163018161924885,
+      "loss": 3.3675,
+      "step": 3620
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5862497687339783,
+      "learning_rate": 0.0002915279697293862,
+      "loss": 3.2222,
+      "step": 3621
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6244729161262512,
+      "learning_rate": 0.0002914257507003839,
+      "loss": 3.343,
+      "step": 3622
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6157928109169006,
+      "learning_rate": 0.0002913235245498149,
+      "loss": 3.3368,
+      "step": 3623
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5982532501220703,
+      "learning_rate": 0.000291221291295253,
+      "loss": 3.3366,
+      "step": 3624
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6363582015037537,
+      "learning_rate": 0.00029111905095427325,
+      "loss": 3.3574,
+      "step": 3625
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6117314100265503,
+      "learning_rate": 0.00029101680354445185,
+      "loss": 3.3052,
+      "step": 3626
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6107704043388367,
+      "learning_rate": 0.0002909145490833666,
+      "loss": 3.3321,
+      "step": 3627
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6260195970535278,
+      "learning_rate": 0.00029081228758859583,
+      "loss": 3.3971,
+      "step": 3628
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7223565578460693,
+      "learning_rate": 0.00029071001907771983,
+      "loss": 3.4466,
+      "step": 3629
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6113673448562622,
+      "learning_rate": 0.00029060774356831946,
+      "loss": 3.3126,
+      "step": 3630
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6077743768692017,
+      "learning_rate": 0.0002905054610779774,
+      "loss": 3.3025,
+      "step": 3631
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6696673631668091,
+      "learning_rate": 0.00029040317162427686,
+      "loss": 3.3966,
+      "step": 3632
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5980106592178345,
+      "learning_rate": 0.0002903008752248027,
+      "loss": 3.3802,
+      "step": 3633
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.666287362575531,
+      "learning_rate": 0.00029019857189714095,
+      "loss": 3.3208,
+      "step": 3634
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6521688103675842,
+      "learning_rate": 0.0002900962616588787,
+      "loss": 3.3544,
+      "step": 3635
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5930438041687012,
+      "learning_rate": 0.00028999394452760413,
+      "loss": 3.3636,
+      "step": 3636
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6384075880050659,
+      "learning_rate": 0.000289891620520907,
+      "loss": 3.299,
+      "step": 3637
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6537858247756958,
+      "learning_rate": 0.0002897892896563777,
+      "loss": 3.4557,
+      "step": 3638
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5836814641952515,
+      "learning_rate": 0.00028968695195160824,
+      "loss": 3.4368,
+      "step": 3639
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6711363792419434,
+      "learning_rate": 0.0002895846074241917,
+      "loss": 3.4193,
+      "step": 3640
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5976700782775879,
+      "learning_rate": 0.00028948225609172224,
+      "loss": 3.2866,
+      "step": 3641
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6466230750083923,
+      "learning_rate": 0.0002893798979717952,
+      "loss": 3.4144,
+      "step": 3642
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6126176714897156,
+      "learning_rate": 0.0002892775330820072,
+      "loss": 3.3428,
+      "step": 3643
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6454432606697083,
+      "learning_rate": 0.00028917516143995594,
+      "loss": 3.4041,
+      "step": 3644
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5797054171562195,
+      "learning_rate": 0.00028907278306324025,
+      "loss": 3.3597,
+      "step": 3645
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6614933013916016,
+      "learning_rate": 0.0002889703979694602,
+      "loss": 3.3663,
+      "step": 3646
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6284923553466797,
+      "learning_rate": 0.0002888680061762169,
+      "loss": 3.3435,
+      "step": 3647
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.633361279964447,
+      "learning_rate": 0.00028876560770111276,
+      "loss": 3.3691,
+      "step": 3648
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6137580871582031,
+      "learning_rate": 0.00028866320256175126,
+      "loss": 3.4468,
+      "step": 3649
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6484012603759766,
+      "learning_rate": 0.00028856079077573704,
+      "loss": 3.3441,
+      "step": 3650
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6236936450004578,
+      "learning_rate": 0.00028845837236067583,
+      "loss": 3.2894,
+      "step": 3651
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6233891248703003,
+      "learning_rate": 0.00028835594733417455,
+      "loss": 3.4351,
+      "step": 3652
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6302180886268616,
+      "learning_rate": 0.0002882535157138413,
+      "loss": 3.3731,
+      "step": 3653
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6548006534576416,
+      "learning_rate": 0.0002881510775172851,
+      "loss": 3.4157,
+      "step": 3654
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5937785506248474,
+      "learning_rate": 0.0002880486327621166,
+      "loss": 3.4101,
+      "step": 3655
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5982134938240051,
+      "learning_rate": 0.00028794618146594684,
+      "loss": 3.4519,
+      "step": 3656
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6334812045097351,
+      "learning_rate": 0.0002878437236463887,
+      "loss": 3.3228,
+      "step": 3657
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5932255983352661,
+      "learning_rate": 0.00028774125932105567,
+      "loss": 3.4541,
+      "step": 3658
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6145234704017639,
+      "learning_rate": 0.0002876387885075627,
+      "loss": 3.3212,
+      "step": 3659
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6053762435913086,
+      "learning_rate": 0.00028753631122352565,
+      "loss": 3.3865,
+      "step": 3660
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6523486375808716,
+      "learning_rate": 0.00028743382748656145,
+      "loss": 3.4043,
+      "step": 3661
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6297605037689209,
+      "learning_rate": 0.00028733133731428836,
+      "loss": 3.3975,
+      "step": 3662
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6141480803489685,
+      "learning_rate": 0.0002872288407243256,
+      "loss": 3.4173,
+      "step": 3663
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5871593952178955,
+      "learning_rate": 0.00028712633773429356,
+      "loss": 3.2913,
+      "step": 3664
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6223506331443787,
+      "learning_rate": 0.0002870238283618136,
+      "loss": 3.4494,
+      "step": 3665
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.591840386390686,
+      "learning_rate": 0.0002869213126245084,
+      "loss": 3.3138,
+      "step": 3666
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5490025877952576,
+      "learning_rate": 0.0002868187905400015,
+      "loss": 3.2517,
+      "step": 3667
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5901401042938232,
+      "learning_rate": 0.0002867162621259177,
+      "loss": 3.3777,
+      "step": 3668
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5850833058357239,
+      "learning_rate": 0.00028661372739988275,
+      "loss": 3.4323,
+      "step": 3669
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6089956164360046,
+      "learning_rate": 0.00028651118637952356,
+      "loss": 3.3793,
+      "step": 3670
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6123780608177185,
+      "learning_rate": 0.00028640863908246825,
+      "loss": 3.3569,
+      "step": 3671
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6722397208213806,
+      "learning_rate": 0.00028630608552634567,
+      "loss": 3.4235,
+      "step": 3672
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6255530714988708,
+      "learning_rate": 0.0002862035257287862,
+      "loss": 3.3247,
+      "step": 3673
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6261695623397827,
+      "learning_rate": 0.00028610095970742097,
+      "loss": 3.3805,
+      "step": 3674
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6323671936988831,
+      "learning_rate": 0.00028599838747988215,
+      "loss": 3.3112,
+      "step": 3675
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6552383303642273,
+      "learning_rate": 0.00028589580906380324,
+      "loss": 3.4024,
+      "step": 3676
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6854276657104492,
+      "learning_rate": 0.0002857932244768186,
+      "loss": 3.5205,
+      "step": 3677
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6248034834861755,
+      "learning_rate": 0.0002856906337365638,
+      "loss": 3.3893,
+      "step": 3678
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6380196809768677,
+      "learning_rate": 0.0002855880368606752,
+      "loss": 3.2897,
+      "step": 3679
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6448647379875183,
+      "learning_rate": 0.0002854854338667905,
+      "loss": 3.4351,
+      "step": 3680
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.668107271194458,
+      "learning_rate": 0.0002853828247725484,
+      "loss": 3.3766,
+      "step": 3681
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6216070055961609,
+      "learning_rate": 0.0002852802095955885,
+      "loss": 3.4826,
+      "step": 3682
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6218451857566833,
+      "learning_rate": 0.0002851775883535515,
+      "loss": 3.3688,
+      "step": 3683
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6319736242294312,
+      "learning_rate": 0.0002850749610640793,
+      "loss": 3.3736,
+      "step": 3684
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6772602200508118,
+      "learning_rate": 0.00028497232774481467,
+      "loss": 3.4899,
+      "step": 3685
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6682306528091431,
+      "learning_rate": 0.0002848696884134016,
+      "loss": 3.4507,
+      "step": 3686
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6152041554450989,
+      "learning_rate": 0.00028476704308748465,
+      "loss": 3.2639,
+      "step": 3687
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6305086612701416,
+      "learning_rate": 0.00028466439178471,
+      "loss": 3.2784,
+      "step": 3688
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6088946461677551,
+      "learning_rate": 0.00028456173452272456,
+      "loss": 3.3655,
+      "step": 3689
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6327491402626038,
+      "learning_rate": 0.0002844590713191762,
+      "loss": 3.3668,
+      "step": 3690
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5848463177680969,
+      "learning_rate": 0.0002843564021917141,
+      "loss": 3.3243,
+      "step": 3691
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.664267361164093,
+      "learning_rate": 0.00028425372715798816,
+      "loss": 3.5031,
+      "step": 3692
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6153045296669006,
+      "learning_rate": 0.0002841510462356494,
+      "loss": 3.3878,
+      "step": 3693
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6261101365089417,
+      "learning_rate": 0.0002840483594423498,
+      "loss": 3.4501,
+      "step": 3694
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6328792572021484,
+      "learning_rate": 0.00028394566679574256,
+      "loss": 3.4624,
+      "step": 3695
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6128214001655579,
+      "learning_rate": 0.0002838429683134817,
+      "loss": 3.2798,
+      "step": 3696
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6675438284873962,
+      "learning_rate": 0.00028374026401322224,
+      "loss": 3.3174,
+      "step": 3697
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6769039630889893,
+      "learning_rate": 0.00028363755391262014,
+      "loss": 3.2651,
+      "step": 3698
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6832557916641235,
+      "learning_rate": 0.00028353483802933267,
+      "loss": 3.4916,
+      "step": 3699
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6211268901824951,
+      "learning_rate": 0.0002834321163810178,
+      "loss": 3.2836,
+      "step": 3700
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6350568532943726,
+      "learning_rate": 0.0002833293889853345,
+      "loss": 3.2902,
+      "step": 3701
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.631552517414093,
+      "learning_rate": 0.0002832266558599428,
+      "loss": 3.293,
+      "step": 3702
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6681779026985168,
+      "learning_rate": 0.0002831239170225038,
+      "loss": 3.1951,
+      "step": 3703
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6160247325897217,
+      "learning_rate": 0.0002830211724906794,
+      "loss": 3.4099,
+      "step": 3704
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5667205452919006,
+      "learning_rate": 0.0002829184222821327,
+      "loss": 3.2883,
+      "step": 3705
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6098401546478271,
+      "learning_rate": 0.00028281566641452763,
+      "loss": 3.3686,
+      "step": 3706
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6125772595405579,
+      "learning_rate": 0.000282712904905529,
+      "loss": 3.3227,
+      "step": 3707
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6355741024017334,
+      "learning_rate": 0.00028261013777280275,
+      "loss": 3.4768,
+      "step": 3708
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5942621231079102,
+      "learning_rate": 0.00028250736503401586,
+      "loss": 3.3359,
+      "step": 3709
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6091006994247437,
+      "learning_rate": 0.000282404586706836,
+      "loss": 3.1985,
+      "step": 3710
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.60381019115448,
+      "learning_rate": 0.000282301802808932,
+      "loss": 3.4062,
+      "step": 3711
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5802066922187805,
+      "learning_rate": 0.00028219901335797365,
+      "loss": 3.4317,
+      "step": 3712
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6418296098709106,
+      "learning_rate": 0.00028209621837163167,
+      "loss": 3.4678,
+      "step": 3713
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6086550354957581,
+      "learning_rate": 0.00028199341786757764,
+      "loss": 3.2207,
+      "step": 3714
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8619450926780701,
+      "learning_rate": 0.0002818906118634841,
+      "loss": 3.5654,
+      "step": 3715
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6018812656402588,
+      "learning_rate": 0.0002817878003770247,
+      "loss": 3.4463,
+      "step": 3716
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6410269737243652,
+      "learning_rate": 0.000281684983425874,
+      "loss": 3.3419,
+      "step": 3717
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5836138129234314,
+      "learning_rate": 0.0002815821610277073,
+      "loss": 3.3654,
+      "step": 3718
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5505456328392029,
+      "learning_rate": 0.00028147933320020084,
+      "loss": 3.3156,
+      "step": 3719
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6228007674217224,
+      "learning_rate": 0.0002813764999610322,
+      "loss": 3.3534,
+      "step": 3720
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6110643744468689,
+      "learning_rate": 0.0002812736613278794,
+      "loss": 3.3265,
+      "step": 3721
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6205087304115295,
+      "learning_rate": 0.0002811708173184217,
+      "loss": 3.3997,
+      "step": 3722
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.605631411075592,
+      "learning_rate": 0.00028106796795033916,
+      "loss": 3.3011,
+      "step": 3723
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6018489599227905,
+      "learning_rate": 0.00028096511324131275,
+      "loss": 3.3225,
+      "step": 3724
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6378231048583984,
+      "learning_rate": 0.0002808622532090244,
+      "loss": 3.4259,
+      "step": 3725
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6305454969406128,
+      "learning_rate": 0.00028075938787115685,
+      "loss": 3.2959,
+      "step": 3726
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6639338135719299,
+      "learning_rate": 0.00028065651724539404,
+      "loss": 3.3189,
+      "step": 3727
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6513733863830566,
+      "learning_rate": 0.00028055364134942047,
+      "loss": 3.3276,
+      "step": 3728
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5964550971984863,
+      "learning_rate": 0.00028045076020092176,
+      "loss": 3.3355,
+      "step": 3729
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6000546216964722,
+      "learning_rate": 0.00028034787381758433,
+      "loss": 3.3898,
+      "step": 3730
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.601989209651947,
+      "learning_rate": 0.00028024498221709555,
+      "loss": 3.2769,
+      "step": 3731
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6208813786506653,
+      "learning_rate": 0.0002801420854171438,
+      "loss": 3.3688,
+      "step": 3732
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5917370915412903,
+      "learning_rate": 0.000280039183435418,
+      "loss": 3.4519,
+      "step": 3733
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6015068888664246,
+      "learning_rate": 0.0002799362762896084,
+      "loss": 3.3931,
+      "step": 3734
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5841699242591858,
+      "learning_rate": 0.00027983336399740584,
+      "loss": 3.4089,
+      "step": 3735
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6215444803237915,
+      "learning_rate": 0.00027973044657650213,
+      "loss": 3.3379,
+      "step": 3736
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6177813410758972,
+      "learning_rate": 0.00027962752404459,
+      "loss": 3.3657,
+      "step": 3737
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6230924129486084,
+      "learning_rate": 0.0002795245964193631,
+      "loss": 3.3429,
+      "step": 3738
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6166138052940369,
+      "learning_rate": 0.00027942166371851566,
+      "loss": 3.4291,
+      "step": 3739
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5767776370048523,
+      "learning_rate": 0.0002793187259597432,
+      "loss": 3.3826,
+      "step": 3740
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6381042003631592,
+      "learning_rate": 0.0002792157831607418,
+      "loss": 3.3958,
+      "step": 3741
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6046695113182068,
+      "learning_rate": 0.00027911283533920865,
+      "loss": 3.358,
+      "step": 3742
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6162915825843811,
+      "learning_rate": 0.0002790098825128416,
+      "loss": 3.3353,
+      "step": 3743
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5994752645492554,
+      "learning_rate": 0.0002789069246993394,
+      "loss": 3.2982,
+      "step": 3744
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6190649271011353,
+      "learning_rate": 0.0002788039619164017,
+      "loss": 3.4497,
+      "step": 3745
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6186133027076721,
+      "learning_rate": 0.0002787009941817291,
+      "loss": 3.464,
+      "step": 3746
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6915348172187805,
+      "learning_rate": 0.0002785980215130228,
+      "loss": 3.3655,
+      "step": 3747
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.613624095916748,
+      "learning_rate": 0.0002784950439279852,
+      "loss": 3.4417,
+      "step": 3748
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6338350772857666,
+      "learning_rate": 0.00027839206144431906,
+      "loss": 3.3845,
+      "step": 3749
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6367514729499817,
+      "learning_rate": 0.0002782890740797285,
+      "loss": 3.4026,
+      "step": 3750
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6197525262832642,
+      "learning_rate": 0.0002781860818519181,
+      "loss": 3.4527,
+      "step": 3751
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6463971734046936,
+      "learning_rate": 0.0002780830847785935,
+      "loss": 3.4799,
+      "step": 3752
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6383652687072754,
+      "learning_rate": 0.00027798008287746096,
+      "loss": 3.3539,
+      "step": 3753
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6330506205558777,
+      "learning_rate": 0.0002778770761662279,
+      "loss": 3.4025,
+      "step": 3754
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6345783472061157,
+      "learning_rate": 0.0002777740646626022,
+      "loss": 3.3139,
+      "step": 3755
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6289472579956055,
+      "learning_rate": 0.00027767104838429283,
+      "loss": 3.4035,
+      "step": 3756
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6420667767524719,
+      "learning_rate": 0.00027756802734900957,
+      "loss": 3.3938,
+      "step": 3757
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7068736553192139,
+      "learning_rate": 0.0002774650015744626,
+      "loss": 3.3751,
+      "step": 3758
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6125309467315674,
+      "learning_rate": 0.00027736197107836356,
+      "loss": 3.3893,
+      "step": 3759
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6480317115783691,
+      "learning_rate": 0.00027725893587842457,
+      "loss": 3.3795,
+      "step": 3760
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6217477917671204,
+      "learning_rate": 0.00027715589599235834,
+      "loss": 3.404,
+      "step": 3761
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6079472303390503,
+      "learning_rate": 0.0002770528514378788,
+      "loss": 3.4794,
+      "step": 3762
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6414199471473694,
+      "learning_rate": 0.00027694980223270054,
+      "loss": 3.3759,
+      "step": 3763
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6436086297035217,
+      "learning_rate": 0.00027684674839453886,
+      "loss": 3.5266,
+      "step": 3764
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.594540536403656,
+      "learning_rate": 0.0002767436899411098,
+      "loss": 3.3593,
+      "step": 3765
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6494881510734558,
+      "learning_rate": 0.00027664062689013055,
+      "loss": 3.3226,
+      "step": 3766
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6552349925041199,
+      "learning_rate": 0.0002765375592593186,
+      "loss": 3.3331,
+      "step": 3767
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6475743651390076,
+      "learning_rate": 0.0002764344870663925,
+      "loss": 3.3533,
+      "step": 3768
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6364050507545471,
+      "learning_rate": 0.00027633141032907166,
+      "loss": 3.497,
+      "step": 3769
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5534500479698181,
+      "learning_rate": 0.00027622832906507624,
+      "loss": 3.2724,
+      "step": 3770
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6304947137832642,
+      "learning_rate": 0.00027612524329212685,
+      "loss": 3.3515,
+      "step": 3771
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6687121987342834,
+      "learning_rate": 0.0002760221530279453,
+      "loss": 3.4175,
+      "step": 3772
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.618787407875061,
+      "learning_rate": 0.000275919058290254,
+      "loss": 3.5104,
+      "step": 3773
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6386235356330872,
+      "learning_rate": 0.00027581595909677607,
+      "loss": 3.3924,
+      "step": 3774
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6412214040756226,
+      "learning_rate": 0.00027571285546523555,
+      "loss": 3.2804,
+      "step": 3775
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6224966049194336,
+      "learning_rate": 0.00027560974741335696,
+      "loss": 3.3536,
+      "step": 3776
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6877733469009399,
+      "learning_rate": 0.000275506634958866,
+      "loss": 3.4325,
+      "step": 3777
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6142483949661255,
+      "learning_rate": 0.0002754035181194888,
+      "loss": 3.4013,
+      "step": 3778
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6085824966430664,
+      "learning_rate": 0.00027530039691295224,
+      "loss": 3.3554,
+      "step": 3779
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5958852171897888,
+      "learning_rate": 0.0002751972713569842,
+      "loss": 3.3051,
+      "step": 3780
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5997690558433533,
+      "learning_rate": 0.00027509414146931313,
+      "loss": 3.3604,
+      "step": 3781
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6032097339630127,
+      "learning_rate": 0.00027499100726766817,
+      "loss": 3.3303,
+      "step": 3782
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5886170268058777,
+      "learning_rate": 0.0002748878687697794,
+      "loss": 3.324,
+      "step": 3783
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5861546397209167,
+      "learning_rate": 0.0002747847259933774,
+      "loss": 3.4258,
+      "step": 3784
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6559439301490784,
+      "learning_rate": 0.0002746815789561938,
+      "loss": 3.4873,
+      "step": 3785
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6742950081825256,
+      "learning_rate": 0.0002745784276759605,
+      "loss": 3.3851,
+      "step": 3786
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6020086407661438,
+      "learning_rate": 0.00027447527217041055,
+      "loss": 3.4026,
+      "step": 3787
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.624190092086792,
+      "learning_rate": 0.00027437211245727764,
+      "loss": 3.4239,
+      "step": 3788
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5968867540359497,
+      "learning_rate": 0.0002742689485542961,
+      "loss": 3.4466,
+      "step": 3789
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6325806975364685,
+      "learning_rate": 0.0002741657804792008,
+      "loss": 3.2536,
+      "step": 3790
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.595104992389679,
+      "learning_rate": 0.0002740626082497277,
+      "loss": 3.2624,
+      "step": 3791
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6479557752609253,
+      "learning_rate": 0.00027395943188361345,
+      "loss": 3.4156,
+      "step": 3792
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6428397297859192,
+      "learning_rate": 0.000273856251398595,
+      "loss": 3.4348,
+      "step": 3793
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.571712851524353,
+      "learning_rate": 0.0002737530668124104,
+      "loss": 3.407,
+      "step": 3794
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5848454236984253,
+      "learning_rate": 0.0002736498781427981,
+      "loss": 3.2116,
+      "step": 3795
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6058680415153503,
+      "learning_rate": 0.00027354668540749774,
+      "loss": 3.3176,
+      "step": 3796
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6119828820228577,
+      "learning_rate": 0.0002734434886242491,
+      "loss": 3.3076,
+      "step": 3797
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6115705370903015,
+      "learning_rate": 0.000273340287810793,
+      "loss": 3.2607,
+      "step": 3798
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6166749596595764,
+      "learning_rate": 0.00027323708298487094,
+      "loss": 3.3529,
+      "step": 3799
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6365509629249573,
+      "learning_rate": 0.0002731338741642248,
+      "loss": 3.3274,
+      "step": 3800
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6337140798568726,
+      "learning_rate": 0.0002730306613665975,
+      "loss": 3.3724,
+      "step": 3801
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6156590580940247,
+      "learning_rate": 0.00027292744460973256,
+      "loss": 3.2998,
+      "step": 3802
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6068218350410461,
+      "learning_rate": 0.0002728242239113741,
+      "loss": 3.3315,
+      "step": 3803
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5837210416793823,
+      "learning_rate": 0.00027272099928926695,
+      "loss": 3.3249,
+      "step": 3804
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6302131414413452,
+      "learning_rate": 0.00027261777076115657,
+      "loss": 3.3445,
+      "step": 3805
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5965090990066528,
+      "learning_rate": 0.0002725145383447893,
+      "loss": 3.4302,
+      "step": 3806
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5544455647468567,
+      "learning_rate": 0.0002724113020579118,
+      "loss": 3.2505,
+      "step": 3807
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6283453106880188,
+      "learning_rate": 0.00027230806191827164,
+      "loss": 3.3896,
+      "step": 3808
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6319368481636047,
+      "learning_rate": 0.00027220481794361704,
+      "loss": 3.372,
+      "step": 3809
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5720791220664978,
+      "learning_rate": 0.00027210157015169687,
+      "loss": 3.2416,
+      "step": 3810
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6046373844146729,
+      "learning_rate": 0.00027199831856026044,
+      "loss": 3.1444,
+      "step": 3811
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6764358282089233,
+      "learning_rate": 0.00027189506318705815,
+      "loss": 3.3285,
+      "step": 3812
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6244410872459412,
+      "learning_rate": 0.0002717918040498406,
+      "loss": 3.3421,
+      "step": 3813
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6250739693641663,
+      "learning_rate": 0.0002716885411663593,
+      "loss": 3.4526,
+      "step": 3814
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.600410521030426,
+      "learning_rate": 0.0002715852745543663,
+      "loss": 3.4254,
+      "step": 3815
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6418642401695251,
+      "learning_rate": 0.0002714820042316145,
+      "loss": 3.1948,
+      "step": 3816
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.639212429523468,
+      "learning_rate": 0.0002713787302158571,
+      "loss": 3.2942,
+      "step": 3817
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6001421809196472,
+      "learning_rate": 0.000271275452524848,
+      "loss": 3.3185,
+      "step": 3818
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6354403495788574,
+      "learning_rate": 0.00027117217117634206,
+      "loss": 3.3556,
+      "step": 3819
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6731590032577515,
+      "learning_rate": 0.0002710688861880945,
+      "loss": 3.4374,
+      "step": 3820
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6121606230735779,
+      "learning_rate": 0.0002709655975778611,
+      "loss": 3.3958,
+      "step": 3821
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6243732571601868,
+      "learning_rate": 0.0002708623053633984,
+      "loss": 3.3648,
+      "step": 3822
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6269816756248474,
+      "learning_rate": 0.00027075900956246353,
+      "loss": 3.3031,
+      "step": 3823
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6408427953720093,
+      "learning_rate": 0.0002706557101928143,
+      "loss": 3.2254,
+      "step": 3824
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.629331648349762,
+      "learning_rate": 0.0002705524072722091,
+      "loss": 3.4062,
+      "step": 3825
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6464104652404785,
+      "learning_rate": 0.00027044910081840665,
+      "loss": 3.3696,
+      "step": 3826
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6373887658119202,
+      "learning_rate": 0.0002703457908491668,
+      "loss": 3.3367,
+      "step": 3827
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.704435408115387,
+      "learning_rate": 0.0002702424773822497,
+      "loss": 3.3642,
+      "step": 3828
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.634634792804718,
+      "learning_rate": 0.00027013916043541595,
+      "loss": 3.2776,
+      "step": 3829
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6082105040550232,
+      "learning_rate": 0.0002700358400264271,
+      "loss": 3.3809,
+      "step": 3830
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5796101689338684,
+      "learning_rate": 0.00026993251617304513,
+      "loss": 3.3239,
+      "step": 3831
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6139970421791077,
+      "learning_rate": 0.0002698291888930325,
+      "loss": 3.4051,
+      "step": 3832
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6833216547966003,
+      "learning_rate": 0.00026972585820415234,
+      "loss": 3.419,
+      "step": 3833
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6220802068710327,
+      "learning_rate": 0.0002696225241241686,
+      "loss": 3.358,
+      "step": 3834
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6287161707878113,
+      "learning_rate": 0.00026951918667084547,
+      "loss": 3.2347,
+      "step": 3835
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6108514070510864,
+      "learning_rate": 0.00026941584586194777,
+      "loss": 3.3711,
+      "step": 3836
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5812754034996033,
+      "learning_rate": 0.00026931250171524114,
+      "loss": 3.4014,
+      "step": 3837
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6192893981933594,
+      "learning_rate": 0.00026920915424849155,
+      "loss": 3.3689,
+      "step": 3838
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6789095401763916,
+      "learning_rate": 0.00026910580347946575,
+      "loss": 3.4586,
+      "step": 3839
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6124496459960938,
+      "learning_rate": 0.00026900244942593067,
+      "loss": 3.3106,
+      "step": 3840
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.663774847984314,
+      "learning_rate": 0.00026889909210565434,
+      "loss": 3.4751,
+      "step": 3841
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6115865111351013,
+      "learning_rate": 0.00026879573153640497,
+      "loss": 3.4109,
+      "step": 3842
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5876237750053406,
+      "learning_rate": 0.00026869236773595145,
+      "loss": 3.3461,
+      "step": 3843
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6677536368370056,
+      "learning_rate": 0.0002685890007220632,
+      "loss": 3.3114,
+      "step": 3844
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6193956136703491,
+      "learning_rate": 0.0002684856305125103,
+      "loss": 3.3138,
+      "step": 3845
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5813229084014893,
+      "learning_rate": 0.0002683822571250631,
+      "loss": 3.3238,
+      "step": 3846
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7698580622673035,
+      "learning_rate": 0.00026827888057749275,
+      "loss": 3.2871,
+      "step": 3847
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6021171808242798,
+      "learning_rate": 0.0002681755008875711,
+      "loss": 3.2632,
+      "step": 3848
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.590849757194519,
+      "learning_rate": 0.0002680721180730701,
+      "loss": 3.3846,
+      "step": 3849
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6216306686401367,
+      "learning_rate": 0.0002679687321517624,
+      "loss": 3.4654,
+      "step": 3850
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6439468860626221,
+      "learning_rate": 0.00026786534314142136,
+      "loss": 3.4023,
+      "step": 3851
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6653583645820618,
+      "learning_rate": 0.0002677619510598208,
+      "loss": 3.3591,
+      "step": 3852
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.625581681728363,
+      "learning_rate": 0.0002676585559247349,
+      "loss": 3.2551,
+      "step": 3853
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6035146117210388,
+      "learning_rate": 0.0002675551577539384,
+      "loss": 3.222,
+      "step": 3854
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6973259449005127,
+      "learning_rate": 0.0002674517565652069,
+      "loss": 3.2016,
+      "step": 3855
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6433961391448975,
+      "learning_rate": 0.00026734835237631605,
+      "loss": 3.3871,
+      "step": 3856
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6088781356811523,
+      "learning_rate": 0.0002672449452050424,
+      "loss": 3.373,
+      "step": 3857
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6201007962226868,
+      "learning_rate": 0.0002671415350691627,
+      "loss": 3.3112,
+      "step": 3858
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5991825461387634,
+      "learning_rate": 0.0002670381219864544,
+      "loss": 3.3877,
+      "step": 3859
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5942227244377136,
+      "learning_rate": 0.0002669347059746954,
+      "loss": 3.3144,
+      "step": 3860
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6242623329162598,
+      "learning_rate": 0.00026683128705166416,
+      "loss": 3.2541,
+      "step": 3861
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6556063890457153,
+      "learning_rate": 0.00026672786523513947,
+      "loss": 3.3878,
+      "step": 3862
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6340212821960449,
+      "learning_rate": 0.00026662444054290085,
+      "loss": 3.3619,
+      "step": 3863
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6830992698669434,
+      "learning_rate": 0.0002665210129927282,
+      "loss": 3.4276,
+      "step": 3864
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.578670084476471,
+      "learning_rate": 0.00026641758260240184,
+      "loss": 3.439,
+      "step": 3865
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6785685420036316,
+      "learning_rate": 0.0002663141493897028,
+      "loss": 3.4199,
+      "step": 3866
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6079252362251282,
+      "learning_rate": 0.00026621071337241226,
+      "loss": 3.3678,
+      "step": 3867
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6454574465751648,
+      "learning_rate": 0.00026610727456831217,
+      "loss": 3.3315,
+      "step": 3868
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.614564836025238,
+      "learning_rate": 0.0002660038329951849,
+      "loss": 3.3012,
+      "step": 3869
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6483057141304016,
+      "learning_rate": 0.00026590038867081316,
+      "loss": 3.2884,
+      "step": 3870
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6164308190345764,
+      "learning_rate": 0.0002657969416129803,
+      "loss": 3.2987,
+      "step": 3871
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6032328605651855,
+      "learning_rate": 0.00026569349183947,
+      "loss": 3.299,
+      "step": 3872
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5902796387672424,
+      "learning_rate": 0.0002655900393680666,
+      "loss": 3.4066,
+      "step": 3873
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.59278804063797,
+      "learning_rate": 0.00026548658421655464,
+      "loss": 3.414,
+      "step": 3874
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5973041653633118,
+      "learning_rate": 0.00026538312640271935,
+      "loss": 3.437,
+      "step": 3875
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.608325183391571,
+      "learning_rate": 0.0002652796659443462,
+      "loss": 3.4709,
+      "step": 3876
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6026968359947205,
+      "learning_rate": 0.00026517620285922144,
+      "loss": 3.3073,
+      "step": 3877
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6208993196487427,
+      "learning_rate": 0.00026507273716513144,
+      "loss": 3.3494,
+      "step": 3878
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5892390012741089,
+      "learning_rate": 0.00026496926887986324,
+      "loss": 3.357,
+      "step": 3879
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6406593918800354,
+      "learning_rate": 0.00026486579802120406,
+      "loss": 3.3358,
+      "step": 3880
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6420329809188843,
+      "learning_rate": 0.00026476232460694195,
+      "loss": 3.3091,
+      "step": 3881
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5696841478347778,
+      "learning_rate": 0.0002646588486548651,
+      "loss": 3.3782,
+      "step": 3882
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6150279641151428,
+      "learning_rate": 0.0002645553701827621,
+      "loss": 3.3737,
+      "step": 3883
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6361640691757202,
+      "learning_rate": 0.0002644518892084224,
+      "loss": 3.4473,
+      "step": 3884
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6132729053497314,
+      "learning_rate": 0.0002643484057496353,
+      "loss": 3.324,
+      "step": 3885
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6346362829208374,
+      "learning_rate": 0.00026424491982419095,
+      "loss": 3.3577,
+      "step": 3886
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6215100884437561,
+      "learning_rate": 0.00026414143144987975,
+      "loss": 3.4401,
+      "step": 3887
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6370041370391846,
+      "learning_rate": 0.0002640379406444925,
+      "loss": 3.3201,
+      "step": 3888
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5833559036254883,
+      "learning_rate": 0.0002639344474258206,
+      "loss": 3.2599,
+      "step": 3889
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6063408255577087,
+      "learning_rate": 0.00026383095181165553,
+      "loss": 3.251,
+      "step": 3890
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6340535879135132,
+      "learning_rate": 0.0002637274538197896,
+      "loss": 3.3845,
+      "step": 3891
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6207410097122192,
+      "learning_rate": 0.0002636239534680151,
+      "loss": 3.2976,
+      "step": 3892
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6404011249542236,
+      "learning_rate": 0.00026352045077412516,
+      "loss": 3.4276,
+      "step": 3893
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.614709198474884,
+      "learning_rate": 0.0002634169457559129,
+      "loss": 3.4475,
+      "step": 3894
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6316673159599304,
+      "learning_rate": 0.0002633134384311722,
+      "loss": 3.4259,
+      "step": 3895
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5817269086837769,
+      "learning_rate": 0.0002632099288176971,
+      "loss": 3.345,
+      "step": 3896
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.624113142490387,
+      "learning_rate": 0.00026310641693328205,
+      "loss": 3.5048,
+      "step": 3897
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6005644798278809,
+      "learning_rate": 0.000263002902795722,
+      "loss": 3.3095,
+      "step": 3898
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6340106129646301,
+      "learning_rate": 0.0002628993864228122,
+      "loss": 3.404,
+      "step": 3899
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5678104162216187,
+      "learning_rate": 0.0002627958678323484,
+      "loss": 3.2185,
+      "step": 3900
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6310855746269226,
+      "learning_rate": 0.0002626923470421264,
+      "loss": 3.4734,
+      "step": 3901
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6362099647521973,
+      "learning_rate": 0.000262588824069943,
+      "loss": 3.4269,
+      "step": 3902
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.63554447889328,
+      "learning_rate": 0.00026248529893359475,
+      "loss": 3.3321,
+      "step": 3903
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.619417130947113,
+      "learning_rate": 0.0002623817716508788,
+      "loss": 3.4176,
+      "step": 3904
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6085559725761414,
+      "learning_rate": 0.00026227824223959287,
+      "loss": 3.362,
+      "step": 3905
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.610848605632782,
+      "learning_rate": 0.00026217471071753463,
+      "loss": 3.2489,
+      "step": 3906
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.572178840637207,
+      "learning_rate": 0.00026207117710250254,
+      "loss": 3.357,
+      "step": 3907
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6343911290168762,
+      "learning_rate": 0.0002619676414122952,
+      "loss": 3.4127,
+      "step": 3908
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6293900609016418,
+      "learning_rate": 0.0002618641036647115,
+      "loss": 3.3451,
+      "step": 3909
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6710907816886902,
+      "learning_rate": 0.0002617605638775509,
+      "loss": 3.3697,
+      "step": 3910
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5918707251548767,
+      "learning_rate": 0.000261657022068613,
+      "loss": 3.4208,
+      "step": 3911
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6613056063652039,
+      "learning_rate": 0.00026155347825569783,
+      "loss": 3.2565,
+      "step": 3912
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6077964305877686,
+      "learning_rate": 0.0002614499324566059,
+      "loss": 3.4051,
+      "step": 3913
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6586365699768066,
+      "learning_rate": 0.0002613463846891377,
+      "loss": 3.4608,
+      "step": 3914
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6008127927780151,
+      "learning_rate": 0.00026124283497109445,
+      "loss": 3.5188,
+      "step": 3915
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6196286678314209,
+      "learning_rate": 0.00026113928332027763,
+      "loss": 3.2877,
+      "step": 3916
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5656877756118774,
+      "learning_rate": 0.00026103572975448884,
+      "loss": 3.2677,
+      "step": 3917
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5755074620246887,
+      "learning_rate": 0.0002609321742915302,
+      "loss": 3.3962,
+      "step": 3918
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6107060313224792,
+      "learning_rate": 0.000260828616949204,
+      "loss": 3.2938,
+      "step": 3919
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6584084630012512,
+      "learning_rate": 0.00026072505774531304,
+      "loss": 3.3053,
+      "step": 3920
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5812655091285706,
+      "learning_rate": 0.0002606214966976603,
+      "loss": 3.4265,
+      "step": 3921
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6057932376861572,
+      "learning_rate": 0.00026051793382404916,
+      "loss": 3.442,
+      "step": 3922
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.617680549621582,
+      "learning_rate": 0.0002604143691422833,
+      "loss": 3.2939,
+      "step": 3923
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.616668164730072,
+      "learning_rate": 0.00026031080267016664,
+      "loss": 3.3975,
+      "step": 3924
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.571333110332489,
+      "learning_rate": 0.0002602072344255034,
+      "loss": 3.289,
+      "step": 3925
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6522971391677856,
+      "learning_rate": 0.0002601036644260982,
+      "loss": 3.3841,
+      "step": 3926
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5657179951667786,
+      "learning_rate": 0.00026000009268975613,
+      "loss": 3.4224,
+      "step": 3927
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5910373330116272,
+      "learning_rate": 0.0002598965192342821,
+      "loss": 3.242,
+      "step": 3928
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6280388832092285,
+      "learning_rate": 0.0002597929440774816,
+      "loss": 3.5236,
+      "step": 3929
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6481741666793823,
+      "learning_rate": 0.0002596893672371605,
+      "loss": 3.2832,
+      "step": 3930
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6153259873390198,
+      "learning_rate": 0.00025958578873112496,
+      "loss": 3.459,
+      "step": 3931
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.59522944688797,
+      "learning_rate": 0.0002594822085771812,
+      "loss": 3.249,
+      "step": 3932
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5696998834609985,
+      "learning_rate": 0.00025937862679313573,
+      "loss": 3.3491,
+      "step": 3933
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6377519369125366,
+      "learning_rate": 0.00025927504339679565,
+      "loss": 3.2992,
+      "step": 3934
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5743221044540405,
+      "learning_rate": 0.0002591714584059681,
+      "loss": 3.3483,
+      "step": 3935
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6047176122665405,
+      "learning_rate": 0.0002590678718384605,
+      "loss": 3.3561,
+      "step": 3936
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5904937982559204,
+      "learning_rate": 0.0002589642837120807,
+      "loss": 3.4039,
+      "step": 3937
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6137279868125916,
+      "learning_rate": 0.00025886069404463646,
+      "loss": 3.3784,
+      "step": 3938
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6516735553741455,
+      "learning_rate": 0.00025875710285393634,
+      "loss": 3.4415,
+      "step": 3939
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5683122277259827,
+      "learning_rate": 0.0002586535101577886,
+      "loss": 3.3377,
+      "step": 3940
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6334203481674194,
+      "learning_rate": 0.00025854991597400217,
+      "loss": 3.2492,
+      "step": 3941
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6036867499351501,
+      "learning_rate": 0.0002584463203203861,
+      "loss": 3.3008,
+      "step": 3942
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.636132001876831,
+      "learning_rate": 0.00025834272321474964,
+      "loss": 3.3408,
+      "step": 3943
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6284753680229187,
+      "learning_rate": 0.0002582391246749023,
+      "loss": 3.4375,
+      "step": 3944
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6076215505599976,
+      "learning_rate": 0.000258135524718654,
+      "loss": 3.3355,
+      "step": 3945
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6466711163520813,
+      "learning_rate": 0.00025803192336381465,
+      "loss": 3.268,
+      "step": 3946
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6400593519210815,
+      "learning_rate": 0.0002579283206281945,
+      "loss": 3.4288,
+      "step": 3947
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6525185704231262,
+      "learning_rate": 0.0002578247165296041,
+      "loss": 3.3486,
+      "step": 3948
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6444108486175537,
+      "learning_rate": 0.0002577211110858543,
+      "loss": 3.518,
+      "step": 3949
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5654917359352112,
+      "learning_rate": 0.00025761750431475605,
+      "loss": 3.2278,
+      "step": 3950
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5556711554527283,
+      "learning_rate": 0.0002575138962341204,
+      "loss": 3.3468,
+      "step": 3951
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5785199999809265,
+      "learning_rate": 0.00025741028686175887,
+      "loss": 3.2947,
+      "step": 3952
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.605494499206543,
+      "learning_rate": 0.0002573066762154832,
+      "loss": 3.382,
+      "step": 3953
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6059899926185608,
+      "learning_rate": 0.0002572030643131051,
+      "loss": 3.3576,
+      "step": 3954
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.632280170917511,
+      "learning_rate": 0.00025709945117243676,
+      "loss": 3.2547,
+      "step": 3955
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5962053537368774,
+      "learning_rate": 0.0002569958368112905,
+      "loss": 3.3773,
+      "step": 3956
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.604576826095581,
+      "learning_rate": 0.00025689222124747877,
+      "loss": 3.39,
+      "step": 3957
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5862025618553162,
+      "learning_rate": 0.00025678860449881427,
+      "loss": 3.2868,
+      "step": 3958
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6356115341186523,
+      "learning_rate": 0.00025668498658311,
+      "loss": 3.3923,
+      "step": 3959
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6119865775108337,
+      "learning_rate": 0.00025658136751817906,
+      "loss": 3.3857,
+      "step": 3960
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6909923553466797,
+      "learning_rate": 0.0002564777473218347,
+      "loss": 3.1973,
+      "step": 3961
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6727779507637024,
+      "learning_rate": 0.0002563741260118904,
+      "loss": 3.4471,
+      "step": 3962
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6961916089057922,
+      "learning_rate": 0.0002562705036061601,
+      "loss": 3.5349,
+      "step": 3963
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6283586621284485,
+      "learning_rate": 0.0002561668801224575,
+      "loss": 3.5474,
+      "step": 3964
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6693708896636963,
+      "learning_rate": 0.00025606325557859665,
+      "loss": 3.4094,
+      "step": 3965
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6046238541603088,
+      "learning_rate": 0.00025595962999239185,
+      "loss": 3.3631,
+      "step": 3966
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6080331206321716,
+      "learning_rate": 0.00025585600338165764,
+      "loss": 3.5612,
+      "step": 3967
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6154356598854065,
+      "learning_rate": 0.00025575237576420864,
+      "loss": 3.3989,
+      "step": 3968
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5944554209709167,
+      "learning_rate": 0.00025564874715785943,
+      "loss": 3.1934,
+      "step": 3969
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.627717912197113,
+      "learning_rate": 0.00025554511758042514,
+      "loss": 3.154,
+      "step": 3970
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6271909475326538,
+      "learning_rate": 0.00025544148704972084,
+      "loss": 3.3213,
+      "step": 3971
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.612829327583313,
+      "learning_rate": 0.00025533785558356175,
+      "loss": 3.2844,
+      "step": 3972
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5844332575798035,
+      "learning_rate": 0.00025523422319976354,
+      "loss": 3.4009,
+      "step": 3973
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6311678886413574,
+      "learning_rate": 0.00025513058991614166,
+      "loss": 3.4696,
+      "step": 3974
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5532718300819397,
+      "learning_rate": 0.0002550269557505118,
+      "loss": 3.2258,
+      "step": 3975
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6182968020439148,
+      "learning_rate": 0.00025492332072069,
+      "loss": 3.3138,
+      "step": 3976
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6046236753463745,
+      "learning_rate": 0.00025481968484449236,
+      "loss": 3.2735,
+      "step": 3977
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6428443789482117,
+      "learning_rate": 0.00025471604813973503,
+      "loss": 3.3787,
+      "step": 3978
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6348656415939331,
+      "learning_rate": 0.0002546124106242343,
+      "loss": 3.4603,
+      "step": 3979
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5930755734443665,
+      "learning_rate": 0.00025450877231580665,
+      "loss": 3.4622,
+      "step": 3980
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.636692464351654,
+      "learning_rate": 0.0002544051332322689,
+      "loss": 3.3818,
+      "step": 3981
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5703644752502441,
+      "learning_rate": 0.0002543014933914376,
+      "loss": 3.3697,
+      "step": 3982
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6083183884620667,
+      "learning_rate": 0.0002541978528111297,
+      "loss": 3.3912,
+      "step": 3983
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6166694760322571,
+      "learning_rate": 0.00025409421150916235,
+      "loss": 3.2652,
+      "step": 3984
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6031867265701294,
+      "learning_rate": 0.0002539905695033524,
+      "loss": 3.3129,
+      "step": 3985
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6122192740440369,
+      "learning_rate": 0.0002538869268115175,
+      "loss": 3.3169,
+      "step": 3986
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6112638711929321,
+      "learning_rate": 0.0002537832834514747,
+      "loss": 3.2996,
+      "step": 3987
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6045041084289551,
+      "learning_rate": 0.0002536796394410416,
+      "loss": 3.3229,
+      "step": 3988
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6106022596359253,
+      "learning_rate": 0.00025357599479803586,
+      "loss": 3.364,
+      "step": 3989
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6055294275283813,
+      "learning_rate": 0.00025347234954027506,
+      "loss": 3.275,
+      "step": 3990
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5865367650985718,
+      "learning_rate": 0.0002533687036855772,
+      "loss": 3.3175,
+      "step": 3991
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6457139849662781,
+      "learning_rate": 0.0002532650572517602,
+      "loss": 3.3128,
+      "step": 3992
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6387559771537781,
+      "learning_rate": 0.00025316141025664193,
+      "loss": 3.2759,
+      "step": 3993
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6478233337402344,
+      "learning_rate": 0.0002530577627180405,
+      "loss": 3.4826,
+      "step": 3994
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6275167465209961,
+      "learning_rate": 0.0002529541146537743,
+      "loss": 3.3614,
+      "step": 3995
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5962046980857849,
+      "learning_rate": 0.00025285046608166165,
+      "loss": 3.4151,
+      "step": 3996
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6149052977561951,
+      "learning_rate": 0.0002527468170195207,
+      "loss": 3.2768,
+      "step": 3997
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.607078492641449,
+      "learning_rate": 0.0002526431674851701,
+      "loss": 3.3428,
+      "step": 3998
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6066473126411438,
+      "learning_rate": 0.0002525395174964283,
+      "loss": 3.4584,
+      "step": 3999
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5871192812919617,
+      "learning_rate": 0.0002524358670711141,
+      "loss": 3.5256,
+      "step": 4000
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5904553532600403,
+      "learning_rate": 0.00025233221622704606,
+      "loss": 3.2749,
+      "step": 4001
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5975815057754517,
+      "learning_rate": 0.000252228564982043,
+      "loss": 3.4215,
+      "step": 4002
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6611238121986389,
+      "learning_rate": 0.0002521249133539238,
+      "loss": 3.3904,
+      "step": 4003
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6311246156692505,
+      "learning_rate": 0.0002520212613605074,
+      "loss": 3.2204,
+      "step": 4004
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5951711535453796,
+      "learning_rate": 0.0002519176090196127,
+      "loss": 3.3291,
+      "step": 4005
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6938546895980835,
+      "learning_rate": 0.0002518139563490588,
+      "loss": 3.3354,
+      "step": 4006
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5967162847518921,
+      "learning_rate": 0.0002517103033666648,
+      "loss": 3.2652,
+      "step": 4007
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6288487911224365,
+      "learning_rate": 0.0002516066500902497,
+      "loss": 3.1915,
+      "step": 4008
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6464089155197144,
+      "learning_rate": 0.0002515029965376329,
+      "loss": 3.2344,
+      "step": 4009
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6463294625282288,
+      "learning_rate": 0.0002513993427266336,
+      "loss": 3.4581,
+      "step": 4010
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.617191731929779,
+      "learning_rate": 0.00025129568867507103,
+      "loss": 3.4698,
+      "step": 4011
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6267497539520264,
+      "learning_rate": 0.00025119203440076453,
+      "loss": 3.3611,
+      "step": 4012
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6050522327423096,
+      "learning_rate": 0.00025108837992153353,
+      "loss": 3.3997,
+      "step": 4013
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6314012408256531,
+      "learning_rate": 0.0002509847252551974,
+      "loss": 3.3887,
+      "step": 4014
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7093877792358398,
+      "learning_rate": 0.00025088107041957566,
+      "loss": 3.3945,
+      "step": 4015
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5881847739219666,
+      "learning_rate": 0.0002507774154324876,
+      "loss": 3.3512,
+      "step": 4016
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6320837736129761,
+      "learning_rate": 0.0002506737603117528,
+      "loss": 3.3946,
+      "step": 4017
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.617286205291748,
+      "learning_rate": 0.0002505701050751909,
+      "loss": 3.2815,
+      "step": 4018
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6056340336799622,
+      "learning_rate": 0.00025046644974062123,
+      "loss": 3.3245,
+      "step": 4019
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6097458600997925,
+      "learning_rate": 0.0002503627943258635,
+      "loss": 3.2694,
+      "step": 4020
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5899346470832825,
+      "learning_rate": 0.00025025913884873727,
+      "loss": 3.3591,
+      "step": 4021
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5989256501197815,
+      "learning_rate": 0.0002501554833270619,
+      "loss": 3.4002,
+      "step": 4022
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6305152773857117,
+      "learning_rate": 0.00025005182777865725,
+      "loss": 3.2025,
+      "step": 4023
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6572516560554504,
+      "learning_rate": 0.00024994817222134276,
+      "loss": 3.3328,
+      "step": 4024
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6383114457130432,
+      "learning_rate": 0.0002498445166729381,
+      "loss": 3.2991,
+      "step": 4025
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6050015091896057,
+      "learning_rate": 0.0002497408611512628,
+      "loss": 3.3712,
+      "step": 4026
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6256352066993713,
+      "learning_rate": 0.0002496372056741365,
+      "loss": 3.4663,
+      "step": 4027
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6097555756568909,
+      "learning_rate": 0.0002495335502593788,
+      "loss": 3.2967,
+      "step": 4028
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6122956275939941,
+      "learning_rate": 0.00024942989492480913,
+      "loss": 3.3454,
+      "step": 4029
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6701467633247375,
+      "learning_rate": 0.00024932623968824724,
+      "loss": 3.2917,
+      "step": 4030
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5594620704650879,
+      "learning_rate": 0.00024922258456751246,
+      "loss": 3.3435,
+      "step": 4031
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5902441143989563,
+      "learning_rate": 0.0002491189295804244,
+      "loss": 3.3574,
+      "step": 4032
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6207271814346313,
+      "learning_rate": 0.0002490152747448026,
+      "loss": 3.4042,
+      "step": 4033
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5825251936912537,
+      "learning_rate": 0.0002489116200784664,
+      "loss": 3.2744,
+      "step": 4034
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5591505169868469,
+      "learning_rate": 0.0002488079655992355,
+      "loss": 3.3777,
+      "step": 4035
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6506365537643433,
+      "learning_rate": 0.00024870431132492904,
+      "loss": 3.3424,
+      "step": 4036
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5781224370002747,
+      "learning_rate": 0.00024860065727336646,
+      "loss": 3.4152,
+      "step": 4037
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6120777130126953,
+      "learning_rate": 0.00024849700346236714,
+      "loss": 3.3249,
+      "step": 4038
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6162705421447754,
+      "learning_rate": 0.0002483933499097504,
+      "loss": 3.3129,
+      "step": 4039
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6134864091873169,
+      "learning_rate": 0.00024828969663333533,
+      "loss": 3.3345,
+      "step": 4040
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6703715920448303,
+      "learning_rate": 0.0002481860436509413,
+      "loss": 3.5297,
+      "step": 4041
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7207913398742676,
+      "learning_rate": 0.00024808239098038744,
+      "loss": 3.4621,
+      "step": 4042
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9156608581542969,
+      "learning_rate": 0.00024797873863949266,
+      "loss": 3.3278,
+      "step": 4043
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6291712522506714,
+      "learning_rate": 0.0002478750866460762,
+      "loss": 3.2997,
+      "step": 4044
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5725845098495483,
+      "learning_rate": 0.000247771435017957,
+      "loss": 3.4005,
+      "step": 4045
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6406933069229126,
+      "learning_rate": 0.000247667783772954,
+      "loss": 3.3842,
+      "step": 4046
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.633264422416687,
+      "learning_rate": 0.0002475641329288859,
+      "loss": 3.3048,
+      "step": 4047
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5785310864448547,
+      "learning_rate": 0.00024746048250357173,
+      "loss": 3.2927,
+      "step": 4048
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6772027015686035,
+      "learning_rate": 0.00024735683251483003,
+      "loss": 3.3779,
+      "step": 4049
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6316635608673096,
+      "learning_rate": 0.00024725318298047936,
+      "loss": 3.3251,
+      "step": 4050
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5939462780952454,
+      "learning_rate": 0.00024714953391833847,
+      "loss": 3.3048,
+      "step": 4051
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5867147445678711,
+      "learning_rate": 0.00024704588534622563,
+      "loss": 3.3868,
+      "step": 4052
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6310352087020874,
+      "learning_rate": 0.0002469422372819595,
+      "loss": 3.4479,
+      "step": 4053
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6316264867782593,
+      "learning_rate": 0.00024683858974335814,
+      "loss": 3.2812,
+      "step": 4054
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6058422327041626,
+      "learning_rate": 0.00024673494274823985,
+      "loss": 3.3781,
+      "step": 4055
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6459770202636719,
+      "learning_rate": 0.0002466312963144228,
+      "loss": 3.5122,
+      "step": 4056
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6532690525054932,
+      "learning_rate": 0.00024652765045972495,
+      "loss": 3.3902,
+      "step": 4057
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5904327034950256,
+      "learning_rate": 0.0002464240052019642,
+      "loss": 3.3848,
+      "step": 4058
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6357998847961426,
+      "learning_rate": 0.0002463203605589585,
+      "loss": 3.268,
+      "step": 4059
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5934481024742126,
+      "learning_rate": 0.00024621671654852543,
+      "loss": 3.2678,
+      "step": 4060
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6189697980880737,
+      "learning_rate": 0.00024611307318848257,
+      "loss": 3.5115,
+      "step": 4061
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6146605610847473,
+      "learning_rate": 0.00024600943049664755,
+      "loss": 3.1718,
+      "step": 4062
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.667597234249115,
+      "learning_rate": 0.0002459057884908377,
+      "loss": 3.3191,
+      "step": 4063
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6192696690559387,
+      "learning_rate": 0.0002458021471888703,
+      "loss": 3.3508,
+      "step": 4064
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6276788711547852,
+      "learning_rate": 0.0002456985066085624,
+      "loss": 3.373,
+      "step": 4065
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6441596150398254,
+      "learning_rate": 0.0002455948667677312,
+      "loss": 3.3959,
+      "step": 4066
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.541420042514801,
+      "learning_rate": 0.00024549122768419336,
+      "loss": 3.3434,
+      "step": 4067
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6219912171363831,
+      "learning_rate": 0.00024538758937576576,
+      "loss": 3.2958,
+      "step": 4068
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6137280464172363,
+      "learning_rate": 0.0002452839518602651,
+      "loss": 3.3196,
+      "step": 4069
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6087313890457153,
+      "learning_rate": 0.0002451803151555076,
+      "loss": 3.3415,
+      "step": 4070
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5690910220146179,
+      "learning_rate": 0.00024507667927930993,
+      "loss": 3.2481,
+      "step": 4071
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6003113985061646,
+      "learning_rate": 0.0002449730442494882,
+      "loss": 3.3616,
+      "step": 4072
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6427178978919983,
+      "learning_rate": 0.0002448694100838584,
+      "loss": 3.3604,
+      "step": 4073
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.648370623588562,
+      "learning_rate": 0.0002447657768002365,
+      "loss": 3.3164,
+      "step": 4074
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6166640520095825,
+      "learning_rate": 0.00024466214441643826,
+      "loss": 3.3474,
+      "step": 4075
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5990052819252014,
+      "learning_rate": 0.00024455851295027923,
+      "loss": 3.4471,
+      "step": 4076
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.61072838306427,
+      "learning_rate": 0.000244454882419575,
+      "loss": 3.2597,
+      "step": 4077
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5881423950195312,
+      "learning_rate": 0.0002443512528421407,
+      "loss": 3.3991,
+      "step": 4078
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6317446231842041,
+      "learning_rate": 0.0002442476242357915,
+      "loss": 3.3498,
+      "step": 4079
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6786907911300659,
+      "learning_rate": 0.0002441439966183423,
+      "loss": 3.3058,
+      "step": 4080
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.652067244052887,
+      "learning_rate": 0.0002440403700076081,
+      "loss": 3.4007,
+      "step": 4081
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6863986253738403,
+      "learning_rate": 0.0002439367444214034,
+      "loss": 3.3557,
+      "step": 4082
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6483095288276672,
+      "learning_rate": 0.00024383311987754254,
+      "loss": 3.3917,
+      "step": 4083
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6491280794143677,
+      "learning_rate": 0.00024372949639383992,
+      "loss": 3.4275,
+      "step": 4084
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6307315230369568,
+      "learning_rate": 0.00024362587398810959,
+      "loss": 3.3075,
+      "step": 4085
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6512874364852905,
+      "learning_rate": 0.00024352225267816537,
+      "loss": 3.3041,
+      "step": 4086
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6538076400756836,
+      "learning_rate": 0.00024341863248182103,
+      "loss": 3.306,
+      "step": 4087
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6164134740829468,
+      "learning_rate": 0.00024331501341689007,
+      "loss": 3.3771,
+      "step": 4088
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6030963063240051,
+      "learning_rate": 0.00024321139550118572,
+      "loss": 3.2981,
+      "step": 4089
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5784734487533569,
+      "learning_rate": 0.00024310777875252127,
+      "loss": 3.2589,
+      "step": 4090
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6351543664932251,
+      "learning_rate": 0.0002430041631887095,
+      "loss": 3.2737,
+      "step": 4091
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6241032481193542,
+      "learning_rate": 0.00024290054882756325,
+      "loss": 3.2979,
+      "step": 4092
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6360937356948853,
+      "learning_rate": 0.00024279693568689495,
+      "loss": 3.4559,
+      "step": 4093
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.571132481098175,
+      "learning_rate": 0.00024269332378451687,
+      "loss": 3.3975,
+      "step": 4094
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6465764045715332,
+      "learning_rate": 0.00024258971313824115,
+      "loss": 3.3546,
+      "step": 4095
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6149590611457825,
+      "learning_rate": 0.00024248610376587971,
+      "loss": 3.2858,
+      "step": 4096
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6058986186981201,
+      "learning_rate": 0.00024238249568524404,
+      "loss": 3.3587,
+      "step": 4097
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6025921106338501,
+      "learning_rate": 0.00024227888891414562,
+      "loss": 3.1882,
+      "step": 4098
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5954164266586304,
+      "learning_rate": 0.00024217528347039583,
+      "loss": 3.4369,
+      "step": 4099
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6313084363937378,
+      "learning_rate": 0.0002420716793718055,
+      "loss": 3.3698,
+      "step": 4100
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6320266127586365,
+      "learning_rate": 0.0002419680766361854,
+      "loss": 3.3298,
+      "step": 4101
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6496672034263611,
+      "learning_rate": 0.00024186447528134605,
+      "loss": 3.405,
+      "step": 4102
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6242969632148743,
+      "learning_rate": 0.0002417608753250977,
+      "loss": 3.4386,
+      "step": 4103
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6158754229545593,
+      "learning_rate": 0.0002416572767852504,
+      "loss": 3.3669,
+      "step": 4104
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6036068797111511,
+      "learning_rate": 0.00024155367967961395,
+      "loss": 3.3808,
+      "step": 4105
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6248922944068909,
+      "learning_rate": 0.0002414500840259979,
+      "loss": 3.3149,
+      "step": 4106
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6088889241218567,
+      "learning_rate": 0.0002413464898422114,
+      "loss": 3.1844,
+      "step": 4107
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6099678874015808,
+      "learning_rate": 0.00024124289714606368,
+      "loss": 3.3684,
+      "step": 4108
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6642912030220032,
+      "learning_rate": 0.00024113930595536353,
+      "loss": 3.4227,
+      "step": 4109
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6614413261413574,
+      "learning_rate": 0.00024103571628791937,
+      "loss": 3.3632,
+      "step": 4110
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5714897513389587,
+      "learning_rate": 0.00024093212816153953,
+      "loss": 3.445,
+      "step": 4111
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7094261050224304,
+      "learning_rate": 0.00024082854159403192,
+      "loss": 3.4828,
+      "step": 4112
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6622090935707092,
+      "learning_rate": 0.00024072495660320439,
+      "loss": 3.4051,
+      "step": 4113
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5947360396385193,
+      "learning_rate": 0.00024062137320686436,
+      "loss": 3.31,
+      "step": 4114
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5824505090713501,
+      "learning_rate": 0.00024051779142281892,
+      "loss": 3.348,
+      "step": 4115
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6096441149711609,
+      "learning_rate": 0.00024041421126887514,
+      "loss": 3.2762,
+      "step": 4116
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5826987624168396,
+      "learning_rate": 0.00024031063276283945,
+      "loss": 3.292,
+      "step": 4117
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6296156644821167,
+      "learning_rate": 0.00024020705592251842,
+      "loss": 3.3162,
+      "step": 4118
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6289445757865906,
+      "learning_rate": 0.00024010348076571798,
+      "loss": 3.2726,
+      "step": 4119
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6153998374938965,
+      "learning_rate": 0.00023999990731024396,
+      "loss": 3.3769,
+      "step": 4120
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6376085877418518,
+      "learning_rate": 0.0002398963355739018,
+      "loss": 3.2786,
+      "step": 4121
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5849770903587341,
+      "learning_rate": 0.00023979276557449663,
+      "loss": 3.2887,
+      "step": 4122
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.638300895690918,
+      "learning_rate": 0.00023968919732983345,
+      "loss": 3.3049,
+      "step": 4123
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6314319968223572,
+      "learning_rate": 0.0002395856308577168,
+      "loss": 3.2331,
+      "step": 4124
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6285073757171631,
+      "learning_rate": 0.00023948206617595088,
+      "loss": 3.137,
+      "step": 4125
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6688522696495056,
+      "learning_rate": 0.00023937850330233966,
+      "loss": 3.3979,
+      "step": 4126
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.649043619632721,
+      "learning_rate": 0.00023927494225468694,
+      "loss": 3.3499,
+      "step": 4127
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6987082958221436,
+      "learning_rate": 0.000239171383050796,
+      "loss": 3.4631,
+      "step": 4128
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6388396620750427,
+      "learning_rate": 0.0002390678257084698,
+      "loss": 3.2583,
+      "step": 4129
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6269456744194031,
+      "learning_rate": 0.00023896427024551115,
+      "loss": 3.422,
+      "step": 4130
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6069709062576294,
+      "learning_rate": 0.0002388607166797224,
+      "loss": 3.4043,
+      "step": 4131
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6211990118026733,
+      "learning_rate": 0.00023875716502890556,
+      "loss": 3.3723,
+      "step": 4132
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.596342146396637,
+      "learning_rate": 0.00023865361531086234,
+      "loss": 3.4274,
+      "step": 4133
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5964327454566956,
+      "learning_rate": 0.00023855006754339424,
+      "loss": 3.3164,
+      "step": 4134
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5660306215286255,
+      "learning_rate": 0.00023844652174430218,
+      "loss": 3.4161,
+      "step": 4135
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6334840059280396,
+      "learning_rate": 0.00023834297793138708,
+      "loss": 3.3481,
+      "step": 4136
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5846516489982605,
+      "learning_rate": 0.00023823943612244914,
+      "loss": 3.4382,
+      "step": 4137
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6413622498512268,
+      "learning_rate": 0.00023813589633528854,
+      "loss": 3.3584,
+      "step": 4138
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6269566416740417,
+      "learning_rate": 0.00023803235858770489,
+      "loss": 3.3366,
+      "step": 4139
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.609159529209137,
+      "learning_rate": 0.00023792882289749747,
+      "loss": 3.2685,
+      "step": 4140
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5730066895484924,
+      "learning_rate": 0.0002378252892824654,
+      "loss": 3.2286,
+      "step": 4141
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6422752737998962,
+      "learning_rate": 0.00023772175776040727,
+      "loss": 3.3926,
+      "step": 4142
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6044661402702332,
+      "learning_rate": 0.00023761822834912126,
+      "loss": 3.1972,
+      "step": 4143
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6210265159606934,
+      "learning_rate": 0.00023751470106640526,
+      "loss": 3.1652,
+      "step": 4144
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6290984153747559,
+      "learning_rate": 0.00023741117593005702,
+      "loss": 3.4232,
+      "step": 4145
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6228407621383667,
+      "learning_rate": 0.00023730765295787356,
+      "loss": 3.2965,
+      "step": 4146
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.628711462020874,
+      "learning_rate": 0.00023720413216765166,
+      "loss": 3.4672,
+      "step": 4147
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6025664806365967,
+      "learning_rate": 0.00023710061357718783,
+      "loss": 3.3333,
+      "step": 4148
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.613601803779602,
+      "learning_rate": 0.00023699709720427807,
+      "loss": 3.3516,
+      "step": 4149
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5990720987319946,
+      "learning_rate": 0.000236893583066718,
+      "loss": 3.3035,
+      "step": 4150
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.589634358882904,
+      "learning_rate": 0.000236790071182303,
+      "loss": 3.4067,
+      "step": 4151
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5927738547325134,
+      "learning_rate": 0.00023668656156882787,
+      "loss": 3.4319,
+      "step": 4152
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6159043908119202,
+      "learning_rate": 0.00023658305424408718,
+      "loss": 3.4393,
+      "step": 4153
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6087044477462769,
+      "learning_rate": 0.0002364795492258749,
+      "loss": 3.3349,
+      "step": 4154
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6531714200973511,
+      "learning_rate": 0.0002363760465319849,
+      "loss": 3.3839,
+      "step": 4155
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6283435821533203,
+      "learning_rate": 0.00023627254618021048,
+      "loss": 3.337,
+      "step": 4156
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6097371578216553,
+      "learning_rate": 0.00023616904818834453,
+      "loss": 3.4198,
+      "step": 4157
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6434276700019836,
+      "learning_rate": 0.0002360655525741795,
+      "loss": 3.3535,
+      "step": 4158
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6356515288352966,
+      "learning_rate": 0.00023596205935550756,
+      "loss": 3.4611,
+      "step": 4159
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.632684051990509,
+      "learning_rate": 0.00023585856855012037,
+      "loss": 3.412,
+      "step": 4160
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6770361065864563,
+      "learning_rate": 0.0002357550801758091,
+      "loss": 3.377,
+      "step": 4161
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5955840349197388,
+      "learning_rate": 0.0002356515942503648,
+      "loss": 3.2534,
+      "step": 4162
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6225008368492126,
+      "learning_rate": 0.00023554811079157763,
+      "loss": 3.375,
+      "step": 4163
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6350579261779785,
+      "learning_rate": 0.00023544462981723788,
+      "loss": 3.4148,
+      "step": 4164
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.622137188911438,
+      "learning_rate": 0.00023534115134513496,
+      "loss": 3.3173,
+      "step": 4165
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6317146420478821,
+      "learning_rate": 0.0002352376753930581,
+      "loss": 3.2954,
+      "step": 4166
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6163128018379211,
+      "learning_rate": 0.00023513420197879598,
+      "loss": 3.2012,
+      "step": 4167
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6241321563720703,
+      "learning_rate": 0.00023503073112013685,
+      "loss": 3.3465,
+      "step": 4168
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7008969187736511,
+      "learning_rate": 0.00023492726283486862,
+      "loss": 3.5505,
+      "step": 4169
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6041575074195862,
+      "learning_rate": 0.00023482379714077865,
+      "loss": 3.4479,
+      "step": 4170
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6186978220939636,
+      "learning_rate": 0.00023472033405565388,
+      "loss": 3.275,
+      "step": 4171
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6767882108688354,
+      "learning_rate": 0.00023461687359728071,
+      "loss": 3.3668,
+      "step": 4172
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6426897048950195,
+      "learning_rate": 0.00023451341578344538,
+      "loss": 3.2808,
+      "step": 4173
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5685768723487854,
+      "learning_rate": 0.00023440996063193344,
+      "loss": 3.33,
+      "step": 4174
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6065342426300049,
+      "learning_rate": 0.00023430650816053,
+      "loss": 3.378,
+      "step": 4175
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6212944984436035,
+      "learning_rate": 0.00023420305838701971,
+      "loss": 3.496,
+      "step": 4176
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6290854811668396,
+      "learning_rate": 0.00023409961132918688,
+      "loss": 3.343,
+      "step": 4177
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6344352960586548,
+      "learning_rate": 0.00023399616700481518,
+      "loss": 3.4487,
+      "step": 4178
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6236302256584167,
+      "learning_rate": 0.00023389272543168784,
+      "loss": 3.4238,
+      "step": 4179
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6058031320571899,
+      "learning_rate": 0.0002337892866275878,
+      "loss": 3.3207,
+      "step": 4180
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5936098694801331,
+      "learning_rate": 0.00023368585061029723,
+      "loss": 3.3671,
+      "step": 4181
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5796845555305481,
+      "learning_rate": 0.00023358241739759815,
+      "loss": 3.2806,
+      "step": 4182
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6267536878585815,
+      "learning_rate": 0.0002334789870072718,
+      "loss": 3.3728,
+      "step": 4183
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6423101425170898,
+      "learning_rate": 0.00023337555945709916,
+      "loss": 3.281,
+      "step": 4184
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6725943684577942,
+      "learning_rate": 0.0002332721347648606,
+      "loss": 3.4129,
+      "step": 4185
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6199285984039307,
+      "learning_rate": 0.0002331687129483359,
+      "loss": 3.2585,
+      "step": 4186
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6867667436599731,
+      "learning_rate": 0.00023306529402530467,
+      "loss": 3.3677,
+      "step": 4187
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6624594926834106,
+      "learning_rate": 0.0002329618780135457,
+      "loss": 3.411,
+      "step": 4188
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6970866918563843,
+      "learning_rate": 0.00023285846493083736,
+      "loss": 3.3584,
+      "step": 4189
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6615217924118042,
+      "learning_rate": 0.00023275505479495768,
+      "loss": 3.4853,
+      "step": 4190
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6189938187599182,
+      "learning_rate": 0.0002326516476236839,
+      "loss": 3.2827,
+      "step": 4191
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6213204264640808,
+      "learning_rate": 0.00023254824343479314,
+      "loss": 3.4017,
+      "step": 4192
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6040611267089844,
+      "learning_rate": 0.0002324448422460616,
+      "loss": 3.4199,
+      "step": 4193
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6316700577735901,
+      "learning_rate": 0.0002323414440752652,
+      "loss": 3.3436,
+      "step": 4194
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5814554691314697,
+      "learning_rate": 0.0002322380489401793,
+      "loss": 3.2686,
+      "step": 4195
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6496289968490601,
+      "learning_rate": 0.0002321346568585787,
+      "loss": 3.3125,
+      "step": 4196
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5613424777984619,
+      "learning_rate": 0.00023203126784823765,
+      "loss": 3.2389,
+      "step": 4197
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6151560544967651,
+      "learning_rate": 0.00023192788192693002,
+      "loss": 3.3445,
+      "step": 4198
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6019925475120544,
+      "learning_rate": 0.000231824499112429,
+      "loss": 3.1763,
+      "step": 4199
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6324076056480408,
+      "learning_rate": 0.00023172111942250717,
+      "loss": 3.2282,
+      "step": 4200
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6262516379356384,
+      "learning_rate": 0.0002316177428749369,
+      "loss": 3.4438,
+      "step": 4201
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6653327345848083,
+      "learning_rate": 0.00023151436948748978,
+      "loss": 3.3371,
+      "step": 4202
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.64223712682724,
+      "learning_rate": 0.00023141099927793683,
+      "loss": 3.4776,
+      "step": 4203
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6516994833946228,
+      "learning_rate": 0.00023130763226404858,
+      "loss": 3.4086,
+      "step": 4204
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6069992184638977,
+      "learning_rate": 0.00023120426846359507,
+      "loss": 3.3633,
+      "step": 4205
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6335489153862,
+      "learning_rate": 0.00023110090789434573,
+      "loss": 3.3145,
+      "step": 4206
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.636854350566864,
+      "learning_rate": 0.00023099755057406934,
+      "loss": 3.3186,
+      "step": 4207
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7066942453384399,
+      "learning_rate": 0.0002308941965205344,
+      "loss": 3.4711,
+      "step": 4208
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6394109129905701,
+      "learning_rate": 0.00023079084575150844,
+      "loss": 3.2384,
+      "step": 4209
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6655170917510986,
+      "learning_rate": 0.00023068749828475887,
+      "loss": 3.3054,
+      "step": 4210
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.603547215461731,
+      "learning_rate": 0.00023058415413805227,
+      "loss": 3.4559,
+      "step": 4211
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6678892970085144,
+      "learning_rate": 0.0002304808133291546,
+      "loss": 3.4117,
+      "step": 4212
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5886514186859131,
+      "learning_rate": 0.00023037747587583145,
+      "loss": 3.373,
+      "step": 4213
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.615600049495697,
+      "learning_rate": 0.0002302741417958477,
+      "loss": 3.3118,
+      "step": 4214
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6184048652648926,
+      "learning_rate": 0.0002301708111069676,
+      "loss": 3.3858,
+      "step": 4215
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6221560835838318,
+      "learning_rate": 0.00023006748382695496,
+      "loss": 3.4005,
+      "step": 4216
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6305278539657593,
+      "learning_rate": 0.00022996415997357297,
+      "loss": 3.31,
+      "step": 4217
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6164804697036743,
+      "learning_rate": 0.00022986083956458403,
+      "loss": 3.2852,
+      "step": 4218
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6002761125564575,
+      "learning_rate": 0.0002297575226177503,
+      "loss": 3.3573,
+      "step": 4219
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6530587077140808,
+      "learning_rate": 0.0002296542091508332,
+      "loss": 3.4403,
+      "step": 4220
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6755840182304382,
+      "learning_rate": 0.00022955089918159333,
+      "loss": 3.3594,
+      "step": 4221
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6349380016326904,
+      "learning_rate": 0.00022944759272779098,
+      "loss": 3.3848,
+      "step": 4222
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5848196148872375,
+      "learning_rate": 0.00022934428980718571,
+      "loss": 3.2475,
+      "step": 4223
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6374369859695435,
+      "learning_rate": 0.00022924099043753648,
+      "loss": 3.5159,
+      "step": 4224
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6684466600418091,
+      "learning_rate": 0.00022913769463660163,
+      "loss": 3.3954,
+      "step": 4225
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6474562883377075,
+      "learning_rate": 0.000229034402422139,
+      "loss": 3.3546,
+      "step": 4226
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5929523706436157,
+      "learning_rate": 0.00022893111381190563,
+      "loss": 3.2398,
+      "step": 4227
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.586125910282135,
+      "learning_rate": 0.0002288278288236579,
+      "loss": 3.2269,
+      "step": 4228
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5903507471084595,
+      "learning_rate": 0.00022872454747515197,
+      "loss": 3.2724,
+      "step": 4229
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6329900026321411,
+      "learning_rate": 0.00022862126978414295,
+      "loss": 3.3877,
+      "step": 4230
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6692850589752197,
+      "learning_rate": 0.00022851799576838552,
+      "loss": 3.3185,
+      "step": 4231
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5712085962295532,
+      "learning_rate": 0.00022841472544563368,
+      "loss": 3.2557,
+      "step": 4232
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6429996490478516,
+      "learning_rate": 0.0002283114588336407,
+      "loss": 3.2549,
+      "step": 4233
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.621189296245575,
+      "learning_rate": 0.00022820819595015945,
+      "loss": 3.3142,
+      "step": 4234
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6361286044120789,
+      "learning_rate": 0.00022810493681294197,
+      "loss": 3.372,
+      "step": 4235
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6397047638893127,
+      "learning_rate": 0.0002280016814397396,
+      "loss": 3.3356,
+      "step": 4236
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5639482736587524,
+      "learning_rate": 0.00022789842984830317,
+      "loss": 3.3375,
+      "step": 4237
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5667389631271362,
+      "learning_rate": 0.00022779518205638297,
+      "loss": 3.2657,
+      "step": 4238
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6697838306427002,
+      "learning_rate": 0.00022769193808172843,
+      "loss": 3.4399,
+      "step": 4239
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.634571373462677,
+      "learning_rate": 0.00022758869794208824,
+      "loss": 3.2491,
+      "step": 4240
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6573125123977661,
+      "learning_rate": 0.00022748546165521079,
+      "loss": 3.4672,
+      "step": 4241
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6508284211158752,
+      "learning_rate": 0.00022738222923884347,
+      "loss": 3.2154,
+      "step": 4242
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6569792628288269,
+      "learning_rate": 0.00022727900071073308,
+      "loss": 3.375,
+      "step": 4243
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6264458894729614,
+      "learning_rate": 0.00022717577608862596,
+      "loss": 3.4025,
+      "step": 4244
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6183469891548157,
+      "learning_rate": 0.00022707255539026753,
+      "loss": 3.4382,
+      "step": 4245
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6360318660736084,
+      "learning_rate": 0.0002269693386334025,
+      "loss": 3.3828,
+      "step": 4246
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5953824520111084,
+      "learning_rate": 0.00022686612583577525,
+      "loss": 3.3895,
+      "step": 4247
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6206636428833008,
+      "learning_rate": 0.00022676291701512912,
+      "loss": 3.3062,
+      "step": 4248
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6012082695960999,
+      "learning_rate": 0.000226659712189207,
+      "loss": 3.2807,
+      "step": 4249
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.56572026014328,
+      "learning_rate": 0.00022655651137575095,
+      "loss": 3.3347,
+      "step": 4250
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6169419884681702,
+      "learning_rate": 0.00022645331459250233,
+      "loss": 3.4329,
+      "step": 4251
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6697790622711182,
+      "learning_rate": 0.00022635012185720193,
+      "loss": 3.3228,
+      "step": 4252
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5939697623252869,
+      "learning_rate": 0.00022624693318758977,
+      "loss": 3.4995,
+      "step": 4253
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6752704381942749,
+      "learning_rate": 0.00022614374860140511,
+      "loss": 3.3671,
+      "step": 4254
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5698899030685425,
+      "learning_rate": 0.00022604056811638656,
+      "loss": 3.3342,
+      "step": 4255
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6277320981025696,
+      "learning_rate": 0.00022593739175027222,
+      "loss": 3.4509,
+      "step": 4256
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6074933409690857,
+      "learning_rate": 0.00022583421952079925,
+      "loss": 3.3508,
+      "step": 4257
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6025525331497192,
+      "learning_rate": 0.000225731051445704,
+      "loss": 3.3106,
+      "step": 4258
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6493105292320251,
+      "learning_rate": 0.0002256278875427224,
+      "loss": 3.4314,
+      "step": 4259
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6410335302352905,
+      "learning_rate": 0.0002255247278295895,
+      "loss": 3.4247,
+      "step": 4260
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.604914665222168,
+      "learning_rate": 0.00022542157232403957,
+      "loss": 3.3722,
+      "step": 4261
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6296116709709167,
+      "learning_rate": 0.00022531842104380633,
+      "loss": 3.316,
+      "step": 4262
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.660125732421875,
+      "learning_rate": 0.00022521527400662267,
+      "loss": 3.3388,
+      "step": 4263
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5822790861129761,
+      "learning_rate": 0.00022511213123022067,
+      "loss": 3.2431,
+      "step": 4264
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6079729199409485,
+      "learning_rate": 0.00022500899273233184,
+      "loss": 3.4239,
+      "step": 4265
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6291740536689758,
+      "learning_rate": 0.00022490585853068688,
+      "loss": 3.3707,
+      "step": 4266
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6137354969978333,
+      "learning_rate": 0.00022480272864301582,
+      "loss": 3.3738,
+      "step": 4267
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6171749234199524,
+      "learning_rate": 0.0002246996030870478,
+      "loss": 3.5086,
+      "step": 4268
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6187096238136292,
+      "learning_rate": 0.00022459648188051127,
+      "loss": 3.3847,
+      "step": 4269
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5911290049552917,
+      "learning_rate": 0.00022449336504113405,
+      "loss": 3.3809,
+      "step": 4270
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6221075057983398,
+      "learning_rate": 0.0002243902525866431,
+      "loss": 3.3216,
+      "step": 4271
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6293851137161255,
+      "learning_rate": 0.00022428714453476457,
+      "loss": 3.3067,
+      "step": 4272
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.603039562702179,
+      "learning_rate": 0.00022418404090322403,
+      "loss": 3.3561,
+      "step": 4273
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5857105255126953,
+      "learning_rate": 0.00022408094170974603,
+      "loss": 3.4122,
+      "step": 4274
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5771323442459106,
+      "learning_rate": 0.00022397784697205473,
+      "loss": 3.3669,
+      "step": 4275
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.61953204870224,
+      "learning_rate": 0.00022387475670787317,
+      "loss": 3.1603,
+      "step": 4276
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6458093523979187,
+      "learning_rate": 0.00022377167093492385,
+      "loss": 3.4174,
+      "step": 4277
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6441712379455566,
+      "learning_rate": 0.00022366858967092835,
+      "loss": 3.3808,
+      "step": 4278
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6296495795249939,
+      "learning_rate": 0.0002235655129336075,
+      "loss": 3.3364,
+      "step": 4279
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6566582918167114,
+      "learning_rate": 0.0002234624407406815,
+      "loss": 3.3708,
+      "step": 4280
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5854193568229675,
+      "learning_rate": 0.0002233593731098696,
+      "loss": 3.1429,
+      "step": 4281
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.9514264464378357,
+      "learning_rate": 0.00022325631005889023,
+      "loss": 3.3458,
+      "step": 4282
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.584960401058197,
+      "learning_rate": 0.00022315325160546118,
+      "loss": 3.2647,
+      "step": 4283
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.636075496673584,
+      "learning_rate": 0.00022305019776729942,
+      "loss": 3.2481,
+      "step": 4284
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6747125387191772,
+      "learning_rate": 0.00022294714856212116,
+      "loss": 3.3448,
+      "step": 4285
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6393190026283264,
+      "learning_rate": 0.0002228441040076417,
+      "loss": 3.3644,
+      "step": 4286
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6362060904502869,
+      "learning_rate": 0.00022274106412157552,
+      "loss": 3.389,
+      "step": 4287
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6304023265838623,
+      "learning_rate": 0.00022263802892163645,
+      "loss": 3.3914,
+      "step": 4288
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6571694016456604,
+      "learning_rate": 0.00022253499842553746,
+      "loss": 3.4138,
+      "step": 4289
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6255675554275513,
+      "learning_rate": 0.00022243197265099058,
+      "loss": 3.3106,
+      "step": 4290
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6484401822090149,
+      "learning_rate": 0.00022232895161570723,
+      "loss": 3.4181,
+      "step": 4291
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.626513659954071,
+      "learning_rate": 0.0002222259353373978,
+      "loss": 3.3272,
+      "step": 4292
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6421791911125183,
+      "learning_rate": 0.00022212292383377215,
+      "loss": 3.2945,
+      "step": 4293
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.64449143409729,
+      "learning_rate": 0.00022201991712253905,
+      "loss": 3.2933,
+      "step": 4294
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6221674084663391,
+      "learning_rate": 0.0002219169152214066,
+      "loss": 3.3385,
+      "step": 4295
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6388044953346252,
+      "learning_rate": 0.000221813918148082,
+      "loss": 3.3818,
+      "step": 4296
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6350899934768677,
+      "learning_rate": 0.00022171092592027157,
+      "loss": 3.3079,
+      "step": 4297
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6287100315093994,
+      "learning_rate": 0.00022160793855568098,
+      "loss": 3.4988,
+      "step": 4298
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6106211543083191,
+      "learning_rate": 0.00022150495607201493,
+      "loss": 3.3329,
+      "step": 4299
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5845887660980225,
+      "learning_rate": 0.00022140197848697718,
+      "loss": 3.2069,
+      "step": 4300
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6291304230690002,
+      "learning_rate": 0.00022129900581827094,
+      "loss": 3.3692,
+      "step": 4301
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5868272185325623,
+      "learning_rate": 0.00022119603808359823,
+      "loss": 3.3427,
+      "step": 4302
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6013039350509644,
+      "learning_rate": 0.00022109307530066062,
+      "loss": 3.444,
+      "step": 4303
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6117579340934753,
+      "learning_rate": 0.00022099011748715844,
+      "loss": 3.3554,
+      "step": 4304
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6590527296066284,
+      "learning_rate": 0.00022088716466079134,
+      "loss": 3.4342,
+      "step": 4305
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5997858643531799,
+      "learning_rate": 0.0002207842168392582,
+      "loss": 3.3164,
+      "step": 4306
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5858597755432129,
+      "learning_rate": 0.0002206812740402569,
+      "loss": 3.2571,
+      "step": 4307
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.644085168838501,
+      "learning_rate": 0.0002205783362814844,
+      "loss": 3.2653,
+      "step": 4308
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6230582594871521,
+      "learning_rate": 0.00022047540358063707,
+      "loss": 3.3099,
+      "step": 4309
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6548083424568176,
+      "learning_rate": 0.0002203724759554101,
+      "loss": 3.2571,
+      "step": 4310
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6090371012687683,
+      "learning_rate": 0.00022026955342349788,
+      "loss": 3.314,
+      "step": 4311
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.572574257850647,
+      "learning_rate": 0.00022016663600259417,
+      "loss": 3.4233,
+      "step": 4312
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6050710082054138,
+      "learning_rate": 0.00022006372371039163,
+      "loss": 3.3334,
+      "step": 4313
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6001432538032532,
+      "learning_rate": 0.00021996081656458204,
+      "loss": 3.3587,
+      "step": 4314
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5992408990859985,
+      "learning_rate": 0.00021985791458285626,
+      "loss": 3.4552,
+      "step": 4315
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6127637624740601,
+      "learning_rate": 0.00021975501778290446,
+      "loss": 3.3865,
+      "step": 4316
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6167628169059753,
+      "learning_rate": 0.00021965212618241576,
+      "loss": 3.3978,
+      "step": 4317
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6233075261116028,
+      "learning_rate": 0.0002195492397990783,
+      "loss": 3.4385,
+      "step": 4318
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5850648880004883,
+      "learning_rate": 0.0002194463586505796,
+      "loss": 3.2513,
+      "step": 4319
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6017858386039734,
+      "learning_rate": 0.00021934348275460597,
+      "loss": 3.3278,
+      "step": 4320
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6295514702796936,
+      "learning_rate": 0.00021924061212884313,
+      "loss": 3.3648,
+      "step": 4321
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6282049417495728,
+      "learning_rate": 0.00021913774679097568,
+      "loss": 3.3844,
+      "step": 4322
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5880584716796875,
+      "learning_rate": 0.00021903488675868726,
+      "loss": 3.2539,
+      "step": 4323
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6382170915603638,
+      "learning_rate": 0.00021893203204966088,
+      "loss": 3.4829,
+      "step": 4324
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6000595092773438,
+      "learning_rate": 0.00021882918268157834,
+      "loss": 3.3531,
+      "step": 4325
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6396212577819824,
+      "learning_rate": 0.0002187263386721206,
+      "loss": 3.3567,
+      "step": 4326
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6351012587547302,
+      "learning_rate": 0.00021862350003896787,
+      "loss": 3.2683,
+      "step": 4327
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6076213717460632,
+      "learning_rate": 0.00021852066679979923,
+      "loss": 3.3865,
+      "step": 4328
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5674775838851929,
+      "learning_rate": 0.00021841783897229278,
+      "loss": 3.2943,
+      "step": 4329
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6063193082809448,
+      "learning_rate": 0.000218315016574126,
+      "loss": 3.3139,
+      "step": 4330
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5930963754653931,
+      "learning_rate": 0.00021821219962297529,
+      "loss": 3.2054,
+      "step": 4331
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6369614601135254,
+      "learning_rate": 0.00021810938813651592,
+      "loss": 3.3401,
+      "step": 4332
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.638029932975769,
+      "learning_rate": 0.00021800658213242243,
+      "loss": 3.3599,
+      "step": 4333
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6605929732322693,
+      "learning_rate": 0.00021790378162836837,
+      "loss": 3.3288,
+      "step": 4334
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6818827986717224,
+      "learning_rate": 0.0002178009866420264,
+      "loss": 3.3761,
+      "step": 4335
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.60908442735672,
+      "learning_rate": 0.000217698197191068,
+      "loss": 3.3528,
+      "step": 4336
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.594711422920227,
+      "learning_rate": 0.00021759541329316408,
+      "loss": 3.3716,
+      "step": 4337
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6263545155525208,
+      "learning_rate": 0.00021749263496598426,
+      "loss": 3.433,
+      "step": 4338
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6035946011543274,
+      "learning_rate": 0.00021738986222719723,
+      "loss": 3.3038,
+      "step": 4339
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6282978057861328,
+      "learning_rate": 0.00021728709509447102,
+      "loss": 3.3346,
+      "step": 4340
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5939093232154846,
+      "learning_rate": 0.0002171843335854724,
+      "loss": 3.2278,
+      "step": 4341
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6777181029319763,
+      "learning_rate": 0.00021708157771786732,
+      "loss": 3.4534,
+      "step": 4342
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6568247675895691,
+      "learning_rate": 0.00021697882750932064,
+      "loss": 3.4082,
+      "step": 4343
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5942575335502625,
+      "learning_rate": 0.00021687608297749625,
+      "loss": 3.2439,
+      "step": 4344
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6373928189277649,
+      "learning_rate": 0.0002167733441400573,
+      "loss": 3.2647,
+      "step": 4345
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6113847494125366,
+      "learning_rate": 0.00021667061101466565,
+      "loss": 3.3904,
+      "step": 4346
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6256155967712402,
+      "learning_rate": 0.0002165678836189823,
+      "loss": 3.2756,
+      "step": 4347
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6141061782836914,
+      "learning_rate": 0.0002164651619706673,
+      "loss": 3.3737,
+      "step": 4348
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5906838774681091,
+      "learning_rate": 0.00021636244608737982,
+      "loss": 3.2511,
+      "step": 4349
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6241053938865662,
+      "learning_rate": 0.00021625973598677785,
+      "loss": 3.3067,
+      "step": 4350
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6333268880844116,
+      "learning_rate": 0.00021615703168651832,
+      "loss": 3.4568,
+      "step": 4351
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6115931868553162,
+      "learning_rate": 0.00021605433320425743,
+      "loss": 3.321,
+      "step": 4352
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.589478611946106,
+      "learning_rate": 0.00021595164055765022,
+      "loss": 3.2818,
+      "step": 4353
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6029682159423828,
+      "learning_rate": 0.00021584895376435068,
+      "loss": 3.2876,
+      "step": 4354
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5721759796142578,
+      "learning_rate": 0.00021574627284201193,
+      "loss": 3.3056,
+      "step": 4355
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6214892864227295,
+      "learning_rate": 0.00021564359780828598,
+      "loss": 3.2567,
+      "step": 4356
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6280972957611084,
+      "learning_rate": 0.0002155409286808238,
+      "loss": 3.446,
+      "step": 4357
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6329858303070068,
+      "learning_rate": 0.00021543826547727545,
+      "loss": 3.2714,
+      "step": 4358
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5748723745346069,
+      "learning_rate": 0.00021533560821529002,
+      "loss": 3.2944,
+      "step": 4359
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6321572065353394,
+      "learning_rate": 0.00021523295691251544,
+      "loss": 3.3506,
+      "step": 4360
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5913103818893433,
+      "learning_rate": 0.00021513031158659852,
+      "loss": 3.4111,
+      "step": 4361
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7028201818466187,
+      "learning_rate": 0.00021502767225518532,
+      "loss": 3.3399,
+      "step": 4362
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5699731707572937,
+      "learning_rate": 0.00021492503893592074,
+      "loss": 3.3424,
+      "step": 4363
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6355714201927185,
+      "learning_rate": 0.00021482241164644855,
+      "loss": 3.3267,
+      "step": 4364
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6434081196784973,
+      "learning_rate": 0.0002147197904044116,
+      "loss": 3.2562,
+      "step": 4365
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6165071725845337,
+      "learning_rate": 0.00021461717522745162,
+      "loss": 3.4498,
+      "step": 4366
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6426162123680115,
+      "learning_rate": 0.0002145145661332095,
+      "loss": 3.3448,
+      "step": 4367
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6269713044166565,
+      "learning_rate": 0.00021441196313932485,
+      "loss": 3.352,
+      "step": 4368
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5989927649497986,
+      "learning_rate": 0.00021430936626343626,
+      "loss": 3.2152,
+      "step": 4369
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.586698055267334,
+      "learning_rate": 0.00021420677552318141,
+      "loss": 3.4072,
+      "step": 4370
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5867027640342712,
+      "learning_rate": 0.00021410419093619682,
+      "loss": 3.2013,
+      "step": 4371
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6716873049736023,
+      "learning_rate": 0.00021400161252011786,
+      "loss": 3.325,
+      "step": 4372
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6525076031684875,
+      "learning_rate": 0.00021389904029257912,
+      "loss": 3.4317,
+      "step": 4373
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6117497682571411,
+      "learning_rate": 0.00021379647427121387,
+      "loss": 3.3695,
+      "step": 4374
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6429139971733093,
+      "learning_rate": 0.00021369391447365437,
+      "loss": 3.3836,
+      "step": 4375
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5918415784835815,
+      "learning_rate": 0.00021359136091753176,
+      "loss": 3.2832,
+      "step": 4376
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6165708899497986,
+      "learning_rate": 0.00021348881362047643,
+      "loss": 3.2793,
+      "step": 4377
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6108337044715881,
+      "learning_rate": 0.00021338627260011732,
+      "loss": 3.297,
+      "step": 4378
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6316553354263306,
+      "learning_rate": 0.00021328373787408235,
+      "loss": 3.4442,
+      "step": 4379
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.594231903553009,
+      "learning_rate": 0.00021318120945999853,
+      "loss": 3.3557,
+      "step": 4380
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6255188584327698,
+      "learning_rate": 0.00021307868737549166,
+      "loss": 3.2208,
+      "step": 4381
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5767228007316589,
+      "learning_rate": 0.00021297617163818639,
+      "loss": 3.2629,
+      "step": 4382
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6460601687431335,
+      "learning_rate": 0.0002128736622657065,
+      "loss": 3.2394,
+      "step": 4383
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6205249428749084,
+      "learning_rate": 0.00021277115927567446,
+      "loss": 3.2985,
+      "step": 4384
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6857031583786011,
+      "learning_rate": 0.00021266866268571168,
+      "loss": 3.3526,
+      "step": 4385
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6057483553886414,
+      "learning_rate": 0.00021256617251343862,
+      "loss": 3.361,
+      "step": 4386
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6175010800361633,
+      "learning_rate": 0.00021246368877647442,
+      "loss": 3.2765,
+      "step": 4387
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5943877696990967,
+      "learning_rate": 0.00021236121149243733,
+      "loss": 3.3107,
+      "step": 4388
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6185916066169739,
+      "learning_rate": 0.00021225874067894434,
+      "loss": 3.1669,
+      "step": 4389
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6013955473899841,
+      "learning_rate": 0.00021215627635361135,
+      "loss": 3.374,
+      "step": 4390
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5989227890968323,
+      "learning_rate": 0.00021205381853405317,
+      "loss": 3.4375,
+      "step": 4391
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6267585754394531,
+      "learning_rate": 0.0002119513672378835,
+      "loss": 3.2624,
+      "step": 4392
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.635187029838562,
+      "learning_rate": 0.00021184892248271489,
+      "loss": 3.2997,
+      "step": 4393
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6231619715690613,
+      "learning_rate": 0.0002117464842861587,
+      "loss": 3.4194,
+      "step": 4394
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5973530411720276,
+      "learning_rate": 0.00021164405266582546,
+      "loss": 3.4433,
+      "step": 4395
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5918546319007874,
+      "learning_rate": 0.0002115416276393242,
+      "loss": 3.2967,
+      "step": 4396
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6279117465019226,
+      "learning_rate": 0.00021143920922426298,
+      "loss": 3.3549,
+      "step": 4397
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6497898101806641,
+      "learning_rate": 0.00021133679743824877,
+      "loss": 3.517,
+      "step": 4398
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6653177738189697,
+      "learning_rate": 0.0002112343922988873,
+      "loss": 3.369,
+      "step": 4399
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6002407670021057,
+      "learning_rate": 0.00021113199382378312,
+      "loss": 3.3906,
+      "step": 4400
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6072937846183777,
+      "learning_rate": 0.0002110296020305399,
+      "loss": 3.2641,
+      "step": 4401
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.621131420135498,
+      "learning_rate": 0.00021092721693675984,
+      "loss": 3.3461,
+      "step": 4402
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6232835054397583,
+      "learning_rate": 0.00021082483856004405,
+      "loss": 3.4475,
+      "step": 4403
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6901021003723145,
+      "learning_rate": 0.0002107224669179928,
+      "loss": 3.3947,
+      "step": 4404
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5754798650741577,
+      "learning_rate": 0.00021062010202820477,
+      "loss": 3.2069,
+      "step": 4405
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6407782435417175,
+      "learning_rate": 0.00021051774390827777,
+      "loss": 3.26,
+      "step": 4406
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6157494187355042,
+      "learning_rate": 0.00021041539257580832,
+      "loss": 3.3776,
+      "step": 4407
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6090334057807922,
+      "learning_rate": 0.00021031304804839177,
+      "loss": 3.2972,
+      "step": 4408
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6406842470169067,
+      "learning_rate": 0.0002102107103436224,
+      "loss": 3.3999,
+      "step": 4409
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6397127509117126,
+      "learning_rate": 0.00021010837947909314,
+      "loss": 3.2272,
+      "step": 4410
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5834441184997559,
+      "learning_rate": 0.0002100060554723959,
+      "loss": 3.3716,
+      "step": 4411
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6270722150802612,
+      "learning_rate": 0.00020990373834112142,
+      "loss": 3.3023,
+      "step": 4412
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6306389570236206,
+      "learning_rate": 0.00020980142810285904,
+      "loss": 3.3696,
+      "step": 4413
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5938227772712708,
+      "learning_rate": 0.00020969912477519732,
+      "loss": 3.3412,
+      "step": 4414
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6692408323287964,
+      "learning_rate": 0.00020959682837572318,
+      "loss": 3.4345,
+      "step": 4415
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6993062496185303,
+      "learning_rate": 0.0002094945389220227,
+      "loss": 3.2873,
+      "step": 4416
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.62058424949646,
+      "learning_rate": 0.00020939225643168055,
+      "loss": 3.2635,
+      "step": 4417
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6588623523712158,
+      "learning_rate": 0.00020928998092228023,
+      "loss": 3.2024,
+      "step": 4418
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6288304328918457,
+      "learning_rate": 0.00020918771241140423,
+      "loss": 3.3743,
+      "step": 4419
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6205514669418335,
+      "learning_rate": 0.00020908545091663356,
+      "loss": 3.3437,
+      "step": 4420
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5885340571403503,
+      "learning_rate": 0.00020898319645554816,
+      "loss": 3.2602,
+      "step": 4421
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5855981707572937,
+      "learning_rate": 0.0002088809490457268,
+      "loss": 3.2926,
+      "step": 4422
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.623570442199707,
+      "learning_rate": 0.000208778708704747,
+      "loss": 3.2314,
+      "step": 4423
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6036929488182068,
+      "learning_rate": 0.0002086764754501851,
+      "loss": 3.4226,
+      "step": 4424
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6300530433654785,
+      "learning_rate": 0.00020857424929961613,
+      "loss": 3.3704,
+      "step": 4425
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5847938060760498,
+      "learning_rate": 0.0002084720302706139,
+      "loss": 3.3486,
+      "step": 4426
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6898181438446045,
+      "learning_rate": 0.00020836981838075113,
+      "loss": 3.4942,
+      "step": 4427
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5924224257469177,
+      "learning_rate": 0.00020826761364759925,
+      "loss": 3.3576,
+      "step": 4428
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6309162974357605,
+      "learning_rate": 0.0002081654160887283,
+      "loss": 3.3284,
+      "step": 4429
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5869278907775879,
+      "learning_rate": 0.0002080632257217074,
+      "loss": 3.3286,
+      "step": 4430
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6758648157119751,
+      "learning_rate": 0.000207961042564104,
+      "loss": 3.4192,
+      "step": 4431
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6503282189369202,
+      "learning_rate": 0.00020785886663348492,
+      "loss": 3.4772,
+      "step": 4432
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6114601492881775,
+      "learning_rate": 0.00020775669794741508,
+      "loss": 3.3737,
+      "step": 4433
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6376871466636658,
+      "learning_rate": 0.0002076545365234587,
+      "loss": 3.2028,
+      "step": 4434
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6289983987808228,
+      "learning_rate": 0.0002075523823791784,
+      "loss": 3.4294,
+      "step": 4435
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6352116465568542,
+      "learning_rate": 0.00020745023553213557,
+      "loss": 3.2161,
+      "step": 4436
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5930891633033752,
+      "learning_rate": 0.00020734809599989062,
+      "loss": 3.2802,
+      "step": 4437
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6394661664962769,
+      "learning_rate": 0.0002072459638000024,
+      "loss": 3.3941,
+      "step": 4438
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6554189920425415,
+      "learning_rate": 0.00020714383895002863,
+      "loss": 3.4242,
+      "step": 4439
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6214420795440674,
+      "learning_rate": 0.00020704172146752576,
+      "loss": 3.3585,
+      "step": 4440
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5942106246948242,
+      "learning_rate": 0.000206939611370049,
+      "loss": 3.3558,
+      "step": 4441
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5977739691734314,
+      "learning_rate": 0.00020683750867515226,
+      "loss": 3.343,
+      "step": 4442
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6262596249580383,
+      "learning_rate": 0.0002067354134003882,
+      "loss": 3.5039,
+      "step": 4443
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6008438467979431,
+      "learning_rate": 0.00020663332556330807,
+      "loss": 3.2965,
+      "step": 4444
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6184266209602356,
+      "learning_rate": 0.0002065312451814621,
+      "loss": 3.3596,
+      "step": 4445
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6176638603210449,
+      "learning_rate": 0.00020642917227239898,
+      "loss": 3.2312,
+      "step": 4446
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.599486768245697,
+      "learning_rate": 0.00020632710685366623,
+      "loss": 3.3167,
+      "step": 4447
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5841536521911621,
+      "learning_rate": 0.00020622504894281018,
+      "loss": 3.4116,
+      "step": 4448
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.62180495262146,
+      "learning_rate": 0.0002061229985573757,
+      "loss": 3.3403,
+      "step": 4449
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5866765975952148,
+      "learning_rate": 0.0002060209557149063,
+      "loss": 3.3545,
+      "step": 4450
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5751224160194397,
+      "learning_rate": 0.00020591892043294452,
+      "loss": 3.3138,
+      "step": 4451
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.621759295463562,
+      "learning_rate": 0.00020581689272903143,
+      "loss": 3.4496,
+      "step": 4452
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6080735921859741,
+      "learning_rate": 0.00020571487262070664,
+      "loss": 3.2598,
+      "step": 4453
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6722729802131653,
+      "learning_rate": 0.00020561286012550864,
+      "loss": 3.3551,
+      "step": 4454
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6033018827438354,
+      "learning_rate": 0.0002055108552609746,
+      "loss": 3.2812,
+      "step": 4455
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6231434345245361,
+      "learning_rate": 0.00020540885804464033,
+      "loss": 3.3449,
+      "step": 4456
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5670124292373657,
+      "learning_rate": 0.0002053068684940402,
+      "loss": 3.1999,
+      "step": 4457
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6172345280647278,
+      "learning_rate": 0.0002052048866267076,
+      "loss": 3.2949,
+      "step": 4458
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6621881127357483,
+      "learning_rate": 0.00020510291246017415,
+      "loss": 3.4038,
+      "step": 4459
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5925076007843018,
+      "learning_rate": 0.0002050009460119707,
+      "loss": 3.3718,
+      "step": 4460
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6103377342224121,
+      "learning_rate": 0.00020489898729962627,
+      "loss": 3.455,
+      "step": 4461
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6944873332977295,
+      "learning_rate": 0.00020479703634066873,
+      "loss": 3.2554,
+      "step": 4462
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6063147783279419,
+      "learning_rate": 0.00020469509315262474,
+      "loss": 3.3391,
+      "step": 4463
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5826002359390259,
+      "learning_rate": 0.00020459315775301945,
+      "loss": 3.357,
+      "step": 4464
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5819864273071289,
+      "learning_rate": 0.0002044912301593767,
+      "loss": 3.3449,
+      "step": 4465
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6028755307197571,
+      "learning_rate": 0.00020438931038921913,
+      "loss": 3.4394,
+      "step": 4466
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.597665548324585,
+      "learning_rate": 0.00020428739846006783,
+      "loss": 3.3657,
+      "step": 4467
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6453766822814941,
+      "learning_rate": 0.00020418549438944262,
+      "loss": 3.2768,
+      "step": 4468
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6111141443252563,
+      "learning_rate": 0.00020408359819486206,
+      "loss": 3.3182,
+      "step": 4469
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6410272717475891,
+      "learning_rate": 0.00020398170989384336,
+      "loss": 3.2668,
+      "step": 4470
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6027970314025879,
+      "learning_rate": 0.00020387982950390222,
+      "loss": 3.2543,
+      "step": 4471
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5735467076301575,
+      "learning_rate": 0.000203777957042553,
+      "loss": 3.2246,
+      "step": 4472
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6231405735015869,
+      "learning_rate": 0.00020367609252730886,
+      "loss": 3.5919,
+      "step": 4473
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6022270917892456,
+      "learning_rate": 0.00020357423597568147,
+      "loss": 3.3906,
+      "step": 4474
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6529914736747742,
+      "learning_rate": 0.00020347238740518107,
+      "loss": 3.2762,
+      "step": 4475
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5402228236198425,
+      "learning_rate": 0.00020337054683331672,
+      "loss": 3.2418,
+      "step": 4476
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6032045483589172,
+      "learning_rate": 0.00020326871427759583,
+      "loss": 3.2748,
+      "step": 4477
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6211556196212769,
+      "learning_rate": 0.00020316688975552483,
+      "loss": 3.2738,
+      "step": 4478
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.587247908115387,
+      "learning_rate": 0.00020306507328460838,
+      "loss": 3.3226,
+      "step": 4479
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6140707731246948,
+      "learning_rate": 0.00020296326488234996,
+      "loss": 3.3092,
+      "step": 4480
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6291319131851196,
+      "learning_rate": 0.0002028614645662516,
+      "loss": 3.403,
+      "step": 4481
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6498053669929504,
+      "learning_rate": 0.00020275967235381398,
+      "loss": 3.3862,
+      "step": 4482
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5758363604545593,
+      "learning_rate": 0.00020265788826253627,
+      "loss": 3.3229,
+      "step": 4483
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6611259579658508,
+      "learning_rate": 0.0002025561123099165,
+      "loss": 3.2656,
+      "step": 4484
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6483119130134583,
+      "learning_rate": 0.00020245434451345102,
+      "loss": 3.274,
+      "step": 4485
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.580224871635437,
+      "learning_rate": 0.00020235258489063486,
+      "loss": 3.3803,
+      "step": 4486
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6108076572418213,
+      "learning_rate": 0.00020225083345896163,
+      "loss": 3.33,
+      "step": 4487
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.599166214466095,
+      "learning_rate": 0.00020214909023592387,
+      "loss": 3.2939,
+      "step": 4488
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6520143747329712,
+      "learning_rate": 0.00020204735523901218,
+      "loss": 3.3922,
+      "step": 4489
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6151868104934692,
+      "learning_rate": 0.000201945628485716,
+      "loss": 3.3768,
+      "step": 4490
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6231768131256104,
+      "learning_rate": 0.00020184390999352346,
+      "loss": 3.2957,
+      "step": 4491
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5855582356452942,
+      "learning_rate": 0.00020174219977992102,
+      "loss": 3.3735,
+      "step": 4492
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6440154314041138,
+      "learning_rate": 0.00020164049786239386,
+      "loss": 3.3244,
+      "step": 4493
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6473847031593323,
+      "learning_rate": 0.0002015388042584258,
+      "loss": 3.261,
+      "step": 4494
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6261402368545532,
+      "learning_rate": 0.0002014371189854991,
+      "loss": 3.3561,
+      "step": 4495
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6150282025337219,
+      "learning_rate": 0.0002013354420610945,
+      "loss": 3.2941,
+      "step": 4496
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6966652870178223,
+      "learning_rate": 0.00020123377350269176,
+      "loss": 3.4082,
+      "step": 4497
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6378828883171082,
+      "learning_rate": 0.0002011321133277686,
+      "loss": 3.465,
+      "step": 4498
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6646527051925659,
+      "learning_rate": 0.00020103046155380173,
+      "loss": 3.4109,
+      "step": 4499
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5820347666740417,
+      "learning_rate": 0.00020092881819826623,
+      "loss": 3.2323,
+      "step": 4500
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6171141266822815,
+      "learning_rate": 0.0002008271832786357,
+      "loss": 3.3112,
+      "step": 4501
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5929138660430908,
+      "learning_rate": 0.0002007255568123825,
+      "loss": 3.3975,
+      "step": 4502
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6141952276229858,
+      "learning_rate": 0.00020062393881697732,
+      "loss": 3.281,
+      "step": 4503
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5851192474365234,
+      "learning_rate": 0.0002005223293098894,
+      "loss": 3.4033,
+      "step": 4504
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6415019631385803,
+      "learning_rate": 0.00020042072830858663,
+      "loss": 3.3742,
+      "step": 4505
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5388271808624268,
+      "learning_rate": 0.0002003191358305355,
+      "loss": 3.4478,
+      "step": 4506
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6231642961502075,
+      "learning_rate": 0.00020021755189320096,
+      "loss": 3.3979,
+      "step": 4507
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.726058304309845,
+      "learning_rate": 0.00020011597651404625,
+      "loss": 3.2798,
+      "step": 4508
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5680627822875977,
+      "learning_rate": 0.00020001440971053355,
+      "loss": 3.2948,
+      "step": 4509
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.618209183216095,
+      "learning_rate": 0.00019991285150012332,
+      "loss": 3.3348,
+      "step": 4510
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5917927026748657,
+      "learning_rate": 0.00019981130190027452,
+      "loss": 3.3388,
+      "step": 4511
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6505729556083679,
+      "learning_rate": 0.0001997097609284448,
+      "loss": 3.263,
+      "step": 4512
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6598019003868103,
+      "learning_rate": 0.0001996082286020902,
+      "loss": 3.3134,
+      "step": 4513
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6504053473472595,
+      "learning_rate": 0.00019950670493866518,
+      "loss": 3.4686,
+      "step": 4514
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6098240613937378,
+      "learning_rate": 0.00019940518995562306,
+      "loss": 3.4063,
+      "step": 4515
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.677740216255188,
+      "learning_rate": 0.0001993036836704153,
+      "loss": 3.3569,
+      "step": 4516
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5991215705871582,
+      "learning_rate": 0.00019920218610049205,
+      "loss": 3.277,
+      "step": 4517
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6191042065620422,
+      "learning_rate": 0.00019910069726330196,
+      "loss": 3.5172,
+      "step": 4518
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5997809171676636,
+      "learning_rate": 0.000198999217176292,
+      "loss": 3.3544,
+      "step": 4519
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6340522170066833,
+      "learning_rate": 0.00019889774585690794,
+      "loss": 3.2665,
+      "step": 4520
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5887564420700073,
+      "learning_rate": 0.00019879628332259376,
+      "loss": 3.2665,
+      "step": 4521
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6379936337471008,
+      "learning_rate": 0.00019869482959079205,
+      "loss": 3.3426,
+      "step": 4522
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6129528880119324,
+      "learning_rate": 0.00019859338467894395,
+      "loss": 3.3432,
+      "step": 4523
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6111044883728027,
+      "learning_rate": 0.00019849194860448887,
+      "loss": 3.3171,
+      "step": 4524
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6117994785308838,
+      "learning_rate": 0.00019839052138486508,
+      "loss": 3.3349,
+      "step": 4525
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6637835502624512,
+      "learning_rate": 0.0001982891030375089,
+      "loss": 3.2723,
+      "step": 4526
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6433278322219849,
+      "learning_rate": 0.00019818769357985547,
+      "loss": 3.2436,
+      "step": 4527
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6748256087303162,
+      "learning_rate": 0.00019808629302933817,
+      "loss": 3.4467,
+      "step": 4528
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6236575245857239,
+      "learning_rate": 0.00019798490140338887,
+      "loss": 3.3715,
+      "step": 4529
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6512519121170044,
+      "learning_rate": 0.0001978835187194381,
+      "loss": 3.4483,
+      "step": 4530
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5926584601402283,
+      "learning_rate": 0.00019778214499491462,
+      "loss": 3.319,
+      "step": 4531
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7325878739356995,
+      "learning_rate": 0.00019768078024724576,
+      "loss": 3.3067,
+      "step": 4532
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8272165060043335,
+      "learning_rate": 0.00019757942449385725,
+      "loss": 3.4562,
+      "step": 4533
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6049686670303345,
+      "learning_rate": 0.00019747807775217344,
+      "loss": 3.4144,
+      "step": 4534
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5436964631080627,
+      "learning_rate": 0.000197376740039617,
+      "loss": 3.347,
+      "step": 4535
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5583860874176025,
+      "learning_rate": 0.000197275411373609,
+      "loss": 3.3251,
+      "step": 4536
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5891790986061096,
+      "learning_rate": 0.00019717409177156893,
+      "loss": 3.262,
+      "step": 4537
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6078890562057495,
+      "learning_rate": 0.000197072781250915,
+      "loss": 3.2676,
+      "step": 4538
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5872478485107422,
+      "learning_rate": 0.00019697147982906355,
+      "loss": 3.332,
+      "step": 4539
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6866099238395691,
+      "learning_rate": 0.00019687018752342938,
+      "loss": 3.1723,
+      "step": 4540
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5853214263916016,
+      "learning_rate": 0.00019676890435142602,
+      "loss": 3.3885,
+      "step": 4541
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6210541129112244,
+      "learning_rate": 0.00019666763033046497,
+      "loss": 3.3431,
+      "step": 4542
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6124565601348877,
+      "learning_rate": 0.00019656636547795664,
+      "loss": 3.3899,
+      "step": 4543
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6130770444869995,
+      "learning_rate": 0.0001964651098113095,
+      "loss": 3.293,
+      "step": 4544
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6430736184120178,
+      "learning_rate": 0.0001963638633479307,
+      "loss": 3.3275,
+      "step": 4545
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5826516151428223,
+      "learning_rate": 0.00019626262610522558,
+      "loss": 3.3288,
+      "step": 4546
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6000692248344421,
+      "learning_rate": 0.00019616139810059793,
+      "loss": 3.4573,
+      "step": 4547
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6481806635856628,
+      "learning_rate": 0.00019606017935145018,
+      "loss": 3.2728,
+      "step": 4548
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6128011345863342,
+      "learning_rate": 0.00019595896987518292,
+      "loss": 3.4553,
+      "step": 4549
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6444360613822937,
+      "learning_rate": 0.00019585776968919516,
+      "loss": 3.284,
+      "step": 4550
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5891768336296082,
+      "learning_rate": 0.0001957565788108844,
+      "loss": 3.224,
+      "step": 4551
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6095583438873291,
+      "learning_rate": 0.0001956553972576467,
+      "loss": 3.3322,
+      "step": 4552
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5875852704048157,
+      "learning_rate": 0.00019555422504687625,
+      "loss": 3.3107,
+      "step": 4553
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6559893488883972,
+      "learning_rate": 0.00019545306219596564,
+      "loss": 3.391,
+      "step": 4554
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6290282011032104,
+      "learning_rate": 0.00019535190872230608,
+      "loss": 3.339,
+      "step": 4555
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5857229232788086,
+      "learning_rate": 0.0001952507646432869,
+      "loss": 3.4297,
+      "step": 4556
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6085899472236633,
+      "learning_rate": 0.00019514962997629602,
+      "loss": 3.2916,
+      "step": 4557
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6261692643165588,
+      "learning_rate": 0.00019504850473871954,
+      "loss": 3.2903,
+      "step": 4558
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.8973466753959656,
+      "learning_rate": 0.00019494738894794222,
+      "loss": 3.3536,
+      "step": 4559
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6741258502006531,
+      "learning_rate": 0.00019484628262134696,
+      "loss": 3.4425,
+      "step": 4560
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6239340305328369,
+      "learning_rate": 0.00019474518577631503,
+      "loss": 3.2413,
+      "step": 4561
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6356580257415771,
+      "learning_rate": 0.00019464409843022627,
+      "loss": 3.366,
+      "step": 4562
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5412483811378479,
+      "learning_rate": 0.00019454302060045877,
+      "loss": 3.2558,
+      "step": 4563
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6215941309928894,
+      "learning_rate": 0.00019444195230438897,
+      "loss": 3.3711,
+      "step": 4564
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6361460089683533,
+      "learning_rate": 0.00019434089355939159,
+      "loss": 3.257,
+      "step": 4565
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6052807569503784,
+      "learning_rate": 0.0001942398443828399,
+      "loss": 3.3148,
+      "step": 4566
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5979077219963074,
+      "learning_rate": 0.00019413880479210538,
+      "loss": 3.3381,
+      "step": 4567
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.631401777267456,
+      "learning_rate": 0.0001940377748045579,
+      "loss": 3.3188,
+      "step": 4568
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6828151345252991,
+      "learning_rate": 0.00019393675443756572,
+      "loss": 3.3572,
+      "step": 4569
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6513455510139465,
+      "learning_rate": 0.00019383574370849532,
+      "loss": 3.4064,
+      "step": 4570
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6232062578201294,
+      "learning_rate": 0.00019373474263471177,
+      "loss": 3.3425,
+      "step": 4571
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5913916230201721,
+      "learning_rate": 0.0001936337512335782,
+      "loss": 3.2361,
+      "step": 4572
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6242142915725708,
+      "learning_rate": 0.00019353276952245628,
+      "loss": 3.3271,
+      "step": 4573
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5948387980461121,
+      "learning_rate": 0.00019343179751870593,
+      "loss": 3.2691,
+      "step": 4574
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6828065514564514,
+      "learning_rate": 0.0001933308352396853,
+      "loss": 3.258,
+      "step": 4575
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5905175805091858,
+      "learning_rate": 0.00019322988270275115,
+      "loss": 3.2377,
+      "step": 4576
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6073306202888489,
+      "learning_rate": 0.00019312893992525827,
+      "loss": 3.3904,
+      "step": 4577
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6556875705718994,
+      "learning_rate": 0.00019302800692455995,
+      "loss": 3.3669,
+      "step": 4578
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.609798014163971,
+      "learning_rate": 0.00019292708371800756,
+      "loss": 3.3693,
+      "step": 4579
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6277830600738525,
+      "learning_rate": 0.00019282617032295122,
+      "loss": 3.2681,
+      "step": 4580
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6359354853630066,
+      "learning_rate": 0.00019272526675673908,
+      "loss": 3.2995,
+      "step": 4581
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6259021759033203,
+      "learning_rate": 0.00019262437303671753,
+      "loss": 3.2703,
+      "step": 4582
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6437623500823975,
+      "learning_rate": 0.00019252348918023137,
+      "loss": 3.3741,
+      "step": 4583
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5862172245979309,
+      "learning_rate": 0.00019242261520462382,
+      "loss": 3.2279,
+      "step": 4584
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5978466868400574,
+      "learning_rate": 0.0001923217511272362,
+      "loss": 3.2782,
+      "step": 4585
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6125479936599731,
+      "learning_rate": 0.00019222089696540814,
+      "loss": 3.3482,
+      "step": 4586
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6221272945404053,
+      "learning_rate": 0.0001921200527364778,
+      "loss": 3.4018,
+      "step": 4587
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5960904359817505,
+      "learning_rate": 0.0001920192184577813,
+      "loss": 3.2547,
+      "step": 4588
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6198006272315979,
+      "learning_rate": 0.0001919183941466534,
+      "loss": 3.4055,
+      "step": 4589
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6370771527290344,
+      "learning_rate": 0.00019181757982042686,
+      "loss": 3.3125,
+      "step": 4590
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6158603429794312,
+      "learning_rate": 0.00019171677549643292,
+      "loss": 3.2231,
+      "step": 4591
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6988177299499512,
+      "learning_rate": 0.0001916159811920009,
+      "loss": 3.305,
+      "step": 4592
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5879431962966919,
+      "learning_rate": 0.00019151519692445858,
+      "loss": 3.3562,
+      "step": 4593
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6362794637680054,
+      "learning_rate": 0.0001914144227111319,
+      "loss": 3.2479,
+      "step": 4594
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6738634705543518,
+      "learning_rate": 0.0001913136585693452,
+      "loss": 3.2917,
+      "step": 4595
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6603161096572876,
+      "learning_rate": 0.0001912129045164209,
+      "loss": 3.3998,
+      "step": 4596
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6091164946556091,
+      "learning_rate": 0.0001911121605696798,
+      "loss": 3.3063,
+      "step": 4597
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6261993050575256,
+      "learning_rate": 0.0001910114267464409,
+      "loss": 3.3927,
+      "step": 4598
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6107027530670166,
+      "learning_rate": 0.00019091070306402174,
+      "loss": 3.2642,
+      "step": 4599
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6385813355445862,
+      "learning_rate": 0.0001908099895397377,
+      "loss": 3.3738,
+      "step": 4600
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6593822240829468,
+      "learning_rate": 0.00019070928619090266,
+      "loss": 3.3662,
+      "step": 4601
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6259279251098633,
+      "learning_rate": 0.0001906085930348287,
+      "loss": 3.3263,
+      "step": 4602
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6008145809173584,
+      "learning_rate": 0.00019050791008882613,
+      "loss": 3.1958,
+      "step": 4603
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6144819855690002,
+      "learning_rate": 0.00019040723737020343,
+      "loss": 3.2219,
+      "step": 4604
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6299683451652527,
+      "learning_rate": 0.00019030657489626753,
+      "loss": 3.2164,
+      "step": 4605
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5852930545806885,
+      "learning_rate": 0.00019020592268432346,
+      "loss": 3.2383,
+      "step": 4606
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6271327137947083,
+      "learning_rate": 0.0001901052807516744,
+      "loss": 3.5819,
+      "step": 4607
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6168107986450195,
+      "learning_rate": 0.00019000464911562192,
+      "loss": 3.2793,
+      "step": 4608
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6508430242538452,
+      "learning_rate": 0.0001899040277934659,
+      "loss": 3.2548,
+      "step": 4609
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5964395999908447,
+      "learning_rate": 0.00018980341680250413,
+      "loss": 3.335,
+      "step": 4610
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5980967879295349,
+      "learning_rate": 0.00018970281616003285,
+      "loss": 3.2761,
+      "step": 4611
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6565948724746704,
+      "learning_rate": 0.00018960222588334656,
+      "loss": 3.3041,
+      "step": 4612
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.569970965385437,
+      "learning_rate": 0.00018950164598973782,
+      "loss": 3.2828,
+      "step": 4613
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6044802069664001,
+      "learning_rate": 0.00018940107649649743,
+      "loss": 3.3015,
+      "step": 4614
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6506989598274231,
+      "learning_rate": 0.0001893005174209146,
+      "loss": 3.3676,
+      "step": 4615
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5685774087905884,
+      "learning_rate": 0.00018919996878027635,
+      "loss": 3.3147,
+      "step": 4616
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6195859313011169,
+      "learning_rate": 0.00018909943059186845,
+      "loss": 3.4735,
+      "step": 4617
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6199584007263184,
+      "learning_rate": 0.0001889989028729745,
+      "loss": 3.3806,
+      "step": 4618
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6059188842773438,
+      "learning_rate": 0.00018889838564087623,
+      "loss": 3.3009,
+      "step": 4619
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6111743450164795,
+      "learning_rate": 0.0001887978789128539,
+      "loss": 3.2618,
+      "step": 4620
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6045678853988647,
+      "learning_rate": 0.00018869738270618566,
+      "loss": 3.3107,
+      "step": 4621
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6162798404693604,
+      "learning_rate": 0.00018859689703814797,
+      "loss": 3.3628,
+      "step": 4622
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5983816981315613,
+      "learning_rate": 0.0001884964219260156,
+      "loss": 3.3064,
+      "step": 4623
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6370041966438293,
+      "learning_rate": 0.0001883959573870613,
+      "loss": 3.491,
+      "step": 4624
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6961913704872131,
+      "learning_rate": 0.000188295503438556,
+      "loss": 3.3276,
+      "step": 4625
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6662081480026245,
+      "learning_rate": 0.00018819506009776904,
+      "loss": 3.3432,
+      "step": 4626
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.589362621307373,
+      "learning_rate": 0.00018809462738196783,
+      "loss": 3.3602,
+      "step": 4627
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8106807470321655,
+      "learning_rate": 0.0001879942053084178,
+      "loss": 3.4204,
+      "step": 4628
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6647022366523743,
+      "learning_rate": 0.00018789379389438273,
+      "loss": 3.3072,
+      "step": 4629
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6003039479255676,
+      "learning_rate": 0.0001877933931571245,
+      "loss": 3.3546,
+      "step": 4630
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6438256502151489,
+      "learning_rate": 0.00018769300311390313,
+      "loss": 3.3693,
+      "step": 4631
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6219691038131714,
+      "learning_rate": 0.0001875926237819768,
+      "loss": 3.3282,
+      "step": 4632
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5815635323524475,
+      "learning_rate": 0.000187492255178602,
+      "loss": 3.4142,
+      "step": 4633
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6395319700241089,
+      "learning_rate": 0.00018739189732103317,
+      "loss": 3.2867,
+      "step": 4634
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6617757081985474,
+      "learning_rate": 0.00018729155022652292,
+      "loss": 3.3896,
+      "step": 4635
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5858076810836792,
+      "learning_rate": 0.00018719121391232225,
+      "loss": 3.3204,
+      "step": 4636
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.660744845867157,
+      "learning_rate": 0.00018709088839568,
+      "loss": 3.2615,
+      "step": 4637
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6023468375205994,
+      "learning_rate": 0.00018699057369384343,
+      "loss": 3.2328,
+      "step": 4638
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.593385636806488,
+      "learning_rate": 0.00018689026982405765,
+      "loss": 3.3062,
+      "step": 4639
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6836429238319397,
+      "learning_rate": 0.0001867899768035661,
+      "loss": 3.3262,
+      "step": 4640
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5973769426345825,
+      "learning_rate": 0.00018668969464961038,
+      "loss": 3.2149,
+      "step": 4641
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5919034481048584,
+      "learning_rate": 0.00018658942337943012,
+      "loss": 3.2584,
+      "step": 4642
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5940399169921875,
+      "learning_rate": 0.00018648916301026303,
+      "loss": 3.2865,
+      "step": 4643
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5559197664260864,
+      "learning_rate": 0.00018638891355934504,
+      "loss": 3.3067,
+      "step": 4644
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5843505263328552,
+      "learning_rate": 0.00018628867504391037,
+      "loss": 3.4005,
+      "step": 4645
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5966205596923828,
+      "learning_rate": 0.00018618844748119107,
+      "loss": 3.3431,
+      "step": 4646
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6731165647506714,
+      "learning_rate": 0.00018608823088841732,
+      "loss": 3.3499,
+      "step": 4647
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6583471298217773,
+      "learning_rate": 0.00018598802528281772,
+      "loss": 3.4808,
+      "step": 4648
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6014115214347839,
+      "learning_rate": 0.00018588783068161863,
+      "loss": 3.3479,
+      "step": 4649
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.621559202671051,
+      "learning_rate": 0.00018578764710204467,
+      "loss": 3.451,
+      "step": 4650
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6271108984947205,
+      "learning_rate": 0.0001856874745613186,
+      "loss": 3.392,
+      "step": 4651
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6189903020858765,
+      "learning_rate": 0.00018558731307666127,
+      "loss": 3.3413,
+      "step": 4652
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6160193681716919,
+      "learning_rate": 0.00018548716266529141,
+      "loss": 3.171,
+      "step": 4653
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5830660462379456,
+      "learning_rate": 0.00018538702334442632,
+      "loss": 3.2925,
+      "step": 4654
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6012463569641113,
+      "learning_rate": 0.00018528689513128092,
+      "loss": 3.2814,
+      "step": 4655
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6324112415313721,
+      "learning_rate": 0.0001851867780430685,
+      "loss": 3.392,
+      "step": 4656
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5771598219871521,
+      "learning_rate": 0.00018508667209700034,
+      "loss": 3.368,
+      "step": 4657
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6254698634147644,
+      "learning_rate": 0.00018498657731028572,
+      "loss": 3.2715,
+      "step": 4658
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5659250617027283,
+      "learning_rate": 0.00018488649370013223,
+      "loss": 3.3505,
+      "step": 4659
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6227028369903564,
+      "learning_rate": 0.0001847864212837453,
+      "loss": 3.1114,
+      "step": 4660
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6223110556602478,
+      "learning_rate": 0.00018468636007832852,
+      "loss": 3.3056,
+      "step": 4661
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5822505354881287,
+      "learning_rate": 0.00018458631010108363,
+      "loss": 3.3205,
+      "step": 4662
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6166062951087952,
+      "learning_rate": 0.00018448627136921042,
+      "loss": 3.3337,
+      "step": 4663
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6271905303001404,
+      "learning_rate": 0.00018438624389990665,
+      "loss": 3.3123,
+      "step": 4664
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6071668863296509,
+      "learning_rate": 0.00018428622771036822,
+      "loss": 3.3749,
+      "step": 4665
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5966046452522278,
+      "learning_rate": 0.00018418622281778907,
+      "loss": 3.3569,
+      "step": 4666
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5757268667221069,
+      "learning_rate": 0.00018408622923936124,
+      "loss": 3.1887,
+      "step": 4667
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.652225136756897,
+      "learning_rate": 0.00018398624699227467,
+      "loss": 3.3437,
+      "step": 4668
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6238378882408142,
+      "learning_rate": 0.00018388627609371757,
+      "loss": 3.4043,
+      "step": 4669
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6278160810470581,
+      "learning_rate": 0.00018378631656087608,
+      "loss": 3.3641,
+      "step": 4670
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6459201574325562,
+      "learning_rate": 0.00018368636841093434,
+      "loss": 3.4107,
+      "step": 4671
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6520346999168396,
+      "learning_rate": 0.00018358643166107463,
+      "loss": 3.3059,
+      "step": 4672
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5968659520149231,
+      "learning_rate": 0.00018348650632847726,
+      "loss": 3.2756,
+      "step": 4673
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6091008186340332,
+      "learning_rate": 0.00018338659243032063,
+      "loss": 3.2954,
+      "step": 4674
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6272988319396973,
+      "learning_rate": 0.00018328668998378095,
+      "loss": 3.2951,
+      "step": 4675
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6257034540176392,
+      "learning_rate": 0.00018318679900603265,
+      "loss": 3.4839,
+      "step": 4676
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6726770997047424,
+      "learning_rate": 0.00018308691951424822,
+      "loss": 3.3251,
+      "step": 4677
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6617605090141296,
+      "learning_rate": 0.000182987051525598,
+      "loss": 3.2709,
+      "step": 4678
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6106722354888916,
+      "learning_rate": 0.00018288719505725056,
+      "loss": 3.4031,
+      "step": 4679
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6694716811180115,
+      "learning_rate": 0.00018278735012637227,
+      "loss": 3.2723,
+      "step": 4680
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6016599535942078,
+      "learning_rate": 0.0001826875167501276,
+      "loss": 3.5272,
+      "step": 4681
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6514267921447754,
+      "learning_rate": 0.00018258769494567932,
+      "loss": 3.3313,
+      "step": 4682
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6131722331047058,
+      "learning_rate": 0.0001824878847301877,
+      "loss": 3.3894,
+      "step": 4683
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6339254379272461,
+      "learning_rate": 0.00018238808612081143,
+      "loss": 3.4352,
+      "step": 4684
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6149921417236328,
+      "learning_rate": 0.00018228829913470695,
+      "loss": 3.3429,
+      "step": 4685
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6236720085144043,
+      "learning_rate": 0.0001821885237890288,
+      "loss": 3.3032,
+      "step": 4686
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5988636612892151,
+      "learning_rate": 0.00018208876010092958,
+      "loss": 3.2897,
+      "step": 4687
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5963378548622131,
+      "learning_rate": 0.0001819890080875598,
+      "loss": 3.3148,
+      "step": 4688
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6665579676628113,
+      "learning_rate": 0.00018188926776606796,
+      "loss": 3.3134,
+      "step": 4689
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6786206960678101,
+      "learning_rate": 0.00018178953915360058,
+      "loss": 3.3366,
+      "step": 4690
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6171886324882507,
+      "learning_rate": 0.0001816898222673022,
+      "loss": 3.243,
+      "step": 4691
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6409314274787903,
+      "learning_rate": 0.00018159011712431527,
+      "loss": 3.3624,
+      "step": 4692
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.60154789686203,
+      "learning_rate": 0.00018149042374178038,
+      "loss": 3.3147,
+      "step": 4693
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6252798438072205,
+      "learning_rate": 0.00018139074213683576,
+      "loss": 3.2986,
+      "step": 4694
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5952434539794922,
+      "learning_rate": 0.000181291072326618,
+      "loss": 3.2963,
+      "step": 4695
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6407507658004761,
+      "learning_rate": 0.0001811914143282615,
+      "loss": 3.5287,
+      "step": 4696
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5918746590614319,
+      "learning_rate": 0.0001810917681588985,
+      "loss": 3.4047,
+      "step": 4697
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.658079206943512,
+      "learning_rate": 0.00018099213383565945,
+      "loss": 3.4384,
+      "step": 4698
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5900827646255493,
+      "learning_rate": 0.00018089251137567254,
+      "loss": 3.2725,
+      "step": 4699
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6569340825080872,
+      "learning_rate": 0.00018079290079606414,
+      "loss": 3.2884,
+      "step": 4700
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5840054750442505,
+      "learning_rate": 0.00018069330211395833,
+      "loss": 3.3072,
+      "step": 4701
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7141736149787903,
+      "learning_rate": 0.0001805937153464775,
+      "loss": 3.3235,
+      "step": 4702
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6071069240570068,
+      "learning_rate": 0.0001804941405107416,
+      "loss": 3.3522,
+      "step": 4703
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6229115128517151,
+      "learning_rate": 0.00018039457762386864,
+      "loss": 3.3637,
+      "step": 4704
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.618442177772522,
+      "learning_rate": 0.00018029502670297479,
+      "loss": 3.3666,
+      "step": 4705
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.606401801109314,
+      "learning_rate": 0.00018019548776517395,
+      "loss": 3.2345,
+      "step": 4706
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6161041855812073,
+      "learning_rate": 0.00018009596082757794,
+      "loss": 3.2574,
+      "step": 4707
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6366614699363708,
+      "learning_rate": 0.0001799964459072967,
+      "loss": 3.2993,
+      "step": 4708
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.636702835559845,
+      "learning_rate": 0.00017989694302143788,
+      "loss": 3.4084,
+      "step": 4709
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6429193615913391,
+      "learning_rate": 0.00017979745218710735,
+      "loss": 3.3006,
+      "step": 4710
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6209059357643127,
+      "learning_rate": 0.00017969797342140868,
+      "loss": 3.2954,
+      "step": 4711
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.634597897529602,
+      "learning_rate": 0.0001795985067414433,
+      "loss": 3.2651,
+      "step": 4712
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5833476781845093,
+      "learning_rate": 0.00017949905216431083,
+      "loss": 3.2958,
+      "step": 4713
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6205071210861206,
+      "learning_rate": 0.00017939960970710865,
+      "loss": 3.3498,
+      "step": 4714
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6410182118415833,
+      "learning_rate": 0.00017930017938693193,
+      "loss": 3.3563,
+      "step": 4715
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6970769166946411,
+      "learning_rate": 0.0001792007612208741,
+      "loss": 3.4069,
+      "step": 4716
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6904531717300415,
+      "learning_rate": 0.00017910135522602614,
+      "loss": 3.4065,
+      "step": 4717
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6469513773918152,
+      "learning_rate": 0.00017900196141947705,
+      "loss": 3.3064,
+      "step": 4718
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6153637766838074,
+      "learning_rate": 0.00017890257981831393,
+      "loss": 3.1413,
+      "step": 4719
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6310357451438904,
+      "learning_rate": 0.00017880321043962165,
+      "loss": 3.351,
+      "step": 4720
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6128477454185486,
+      "learning_rate": 0.00017870385330048284,
+      "loss": 3.3062,
+      "step": 4721
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6027767062187195,
+      "learning_rate": 0.00017860450841797814,
+      "loss": 3.2342,
+      "step": 4722
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6310712695121765,
+      "learning_rate": 0.0001785051758091862,
+      "loss": 3.3841,
+      "step": 4723
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6277130842208862,
+      "learning_rate": 0.00017840585549118337,
+      "loss": 3.3743,
+      "step": 4724
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6063037514686584,
+      "learning_rate": 0.0001783065474810439,
+      "loss": 3.1617,
+      "step": 4725
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6380765438079834,
+      "learning_rate": 0.00017820725179584014,
+      "loss": 3.4209,
+      "step": 4726
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6219593286514282,
+      "learning_rate": 0.000178107968452642,
+      "loss": 3.3852,
+      "step": 4727
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5906947255134583,
+      "learning_rate": 0.00017800869746851757,
+      "loss": 3.3816,
+      "step": 4728
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6144208312034607,
+      "learning_rate": 0.00017790943886053268,
+      "loss": 3.3553,
+      "step": 4729
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6082693934440613,
+      "learning_rate": 0.00017781019264575092,
+      "loss": 3.2825,
+      "step": 4730
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6370079517364502,
+      "learning_rate": 0.00017771095884123404,
+      "loss": 3.2945,
+      "step": 4731
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6162868738174438,
+      "learning_rate": 0.00017761173746404135,
+      "loss": 3.1309,
+      "step": 4732
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6330640912055969,
+      "learning_rate": 0.00017751252853123012,
+      "loss": 3.2824,
+      "step": 4733
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6064519882202148,
+      "learning_rate": 0.00017741333205985565,
+      "loss": 3.2224,
+      "step": 4734
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6307936310768127,
+      "learning_rate": 0.0001773141480669709,
+      "loss": 3.2816,
+      "step": 4735
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.631202220916748,
+      "learning_rate": 0.00017721497656962665,
+      "loss": 3.338,
+      "step": 4736
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6187067031860352,
+      "learning_rate": 0.00017711581758487178,
+      "loss": 3.2692,
+      "step": 4737
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6404892206192017,
+      "learning_rate": 0.00017701667112975285,
+      "loss": 3.2894,
+      "step": 4738
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6452632546424866,
+      "learning_rate": 0.00017691753722131424,
+      "loss": 3.3684,
+      "step": 4739
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6127825379371643,
+      "learning_rate": 0.00017681841587659816,
+      "loss": 3.3527,
+      "step": 4740
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6171119809150696,
+      "learning_rate": 0.00017671930711264487,
+      "loss": 3.2473,
+      "step": 4741
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6060885190963745,
+      "learning_rate": 0.00017662021094649223,
+      "loss": 3.3261,
+      "step": 4742
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6263689398765564,
+      "learning_rate": 0.00017652112739517596,
+      "loss": 3.2627,
+      "step": 4743
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5872975587844849,
+      "learning_rate": 0.00017642205647572975,
+      "loss": 3.3252,
+      "step": 4744
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6522191762924194,
+      "learning_rate": 0.00017632299820518504,
+      "loss": 3.327,
+      "step": 4745
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6609612703323364,
+      "learning_rate": 0.000176223952600571,
+      "loss": 3.2529,
+      "step": 4746
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6605116128921509,
+      "learning_rate": 0.00017612491967891486,
+      "loss": 3.4559,
+      "step": 4747
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6365745663642883,
+      "learning_rate": 0.00017602589945724144,
+      "loss": 3.3159,
+      "step": 4748
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6035125255584717,
+      "learning_rate": 0.00017592689195257354,
+      "loss": 3.2803,
+      "step": 4749
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6476607918739319,
+      "learning_rate": 0.0001758278971819316,
+      "loss": 3.3746,
+      "step": 4750
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6447315812110901,
+      "learning_rate": 0.00017572891516233398,
+      "loss": 3.2185,
+      "step": 4751
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6167468428611755,
+      "learning_rate": 0.00017562994591079696,
+      "loss": 3.4329,
+      "step": 4752
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6049765944480896,
+      "learning_rate": 0.00017553098944433435,
+      "loss": 3.2662,
+      "step": 4753
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5891309380531311,
+      "learning_rate": 0.00017543204577995795,
+      "loss": 3.3656,
+      "step": 4754
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7036178708076477,
+      "learning_rate": 0.0001753331149346773,
+      "loss": 3.2953,
+      "step": 4755
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6219146847724915,
+      "learning_rate": 0.00017523419692549988,
+      "loss": 3.2968,
+      "step": 4756
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6210200190544128,
+      "learning_rate": 0.0001751352917694308,
+      "loss": 3.3618,
+      "step": 4757
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6193158626556396,
+      "learning_rate": 0.00017503639948347288,
+      "loss": 3.3273,
+      "step": 4758
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6548976302146912,
+      "learning_rate": 0.00017493752008462704,
+      "loss": 3.4085,
+      "step": 4759
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6289055943489075,
+      "learning_rate": 0.00017483865358989168,
+      "loss": 3.3661,
+      "step": 4760
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5951789021492004,
+      "learning_rate": 0.00017473980001626304,
+      "loss": 3.3544,
+      "step": 4761
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5678247213363647,
+      "learning_rate": 0.00017464095938073536,
+      "loss": 3.3182,
+      "step": 4762
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6432581543922424,
+      "learning_rate": 0.00017454213170030037,
+      "loss": 3.2518,
+      "step": 4763
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6164613366127014,
+      "learning_rate": 0.00017444331699194762,
+      "loss": 3.359,
+      "step": 4764
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6441646218299866,
+      "learning_rate": 0.00017434451527266465,
+      "loss": 3.3631,
+      "step": 4765
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5996881127357483,
+      "learning_rate": 0.00017424572655943665,
+      "loss": 3.2875,
+      "step": 4766
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6040083765983582,
+      "learning_rate": 0.00017414695086924648,
+      "loss": 3.3622,
+      "step": 4767
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.630914568901062,
+      "learning_rate": 0.00017404818821907482,
+      "loss": 3.2048,
+      "step": 4768
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6699578762054443,
+      "learning_rate": 0.00017394943862590004,
+      "loss": 3.5253,
+      "step": 4769
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.665324866771698,
+      "learning_rate": 0.0001738507021066985,
+      "loss": 3.2589,
+      "step": 4770
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6398959755897522,
+      "learning_rate": 0.00017375197867844401,
+      "loss": 3.2727,
+      "step": 4771
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.697909951210022,
+      "learning_rate": 0.00017365326835810832,
+      "loss": 3.4117,
+      "step": 4772
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6068934798240662,
+      "learning_rate": 0.0001735545711626608,
+      "loss": 3.2946,
+      "step": 4773
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6205888390541077,
+      "learning_rate": 0.0001734558871090689,
+      "loss": 3.3062,
+      "step": 4774
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6135295629501343,
+      "learning_rate": 0.00017335721621429732,
+      "loss": 3.4012,
+      "step": 4775
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.663394570350647,
+      "learning_rate": 0.00017325855849530876,
+      "loss": 3.2613,
+      "step": 4776
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6539154052734375,
+      "learning_rate": 0.00017315991396906372,
+      "loss": 3.3191,
+      "step": 4777
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6040697693824768,
+      "learning_rate": 0.00017306128265252025,
+      "loss": 3.2666,
+      "step": 4778
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6077384948730469,
+      "learning_rate": 0.0001729626645626342,
+      "loss": 3.3127,
+      "step": 4779
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6014003753662109,
+      "learning_rate": 0.00017286405971635928,
+      "loss": 3.3564,
+      "step": 4780
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5933483839035034,
+      "learning_rate": 0.0001727654681306467,
+      "loss": 3.2653,
+      "step": 4781
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6310820579528809,
+      "learning_rate": 0.0001726668898224455,
+      "loss": 3.3651,
+      "step": 4782
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6220866441726685,
+      "learning_rate": 0.00017256832480870237,
+      "loss": 3.3343,
+      "step": 4783
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6071862578392029,
+      "learning_rate": 0.00017246977310636201,
+      "loss": 3.3427,
+      "step": 4784
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5943565964698792,
+      "learning_rate": 0.00017237123473236643,
+      "loss": 3.2754,
+      "step": 4785
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6503081917762756,
+      "learning_rate": 0.00017227270970365555,
+      "loss": 3.4084,
+      "step": 4786
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6690330505371094,
+      "learning_rate": 0.00017217419803716697,
+      "loss": 3.3706,
+      "step": 4787
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6268274784088135,
+      "learning_rate": 0.000172075699749836,
+      "loss": 3.4523,
+      "step": 4788
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6357588768005371,
+      "learning_rate": 0.00017197721485859562,
+      "loss": 3.5234,
+      "step": 4789
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6383059620857239,
+      "learning_rate": 0.00017187874338037644,
+      "loss": 3.3371,
+      "step": 4790
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.646907389163971,
+      "learning_rate": 0.00017178028533210705,
+      "loss": 3.3089,
+      "step": 4791
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5765965580940247,
+      "learning_rate": 0.00017168184073071324,
+      "loss": 3.3526,
+      "step": 4792
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5881819725036621,
+      "learning_rate": 0.0001715834095931191,
+      "loss": 3.286,
+      "step": 4793
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6026200652122498,
+      "learning_rate": 0.00017148499193624586,
+      "loss": 3.2993,
+      "step": 4794
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7471684813499451,
+      "learning_rate": 0.0001713865877770128,
+      "loss": 3.3015,
+      "step": 4795
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6499928832054138,
+      "learning_rate": 0.00017128819713233665,
+      "loss": 3.2953,
+      "step": 4796
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6148751378059387,
+      "learning_rate": 0.00017118982001913187,
+      "loss": 3.2791,
+      "step": 4797
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.681559145450592,
+      "learning_rate": 0.00017109145645431074,
+      "loss": 3.3396,
+      "step": 4798
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6669842600822449,
+      "learning_rate": 0.000170993106454783,
+      "loss": 3.3768,
+      "step": 4799
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6817641258239746,
+      "learning_rate": 0.00017089477003745618,
+      "loss": 3.461,
+      "step": 4800
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6329324841499329,
+      "learning_rate": 0.00017079644721923538,
+      "loss": 3.3652,
+      "step": 4801
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6417033076286316,
+      "learning_rate": 0.00017069813801702362,
+      "loss": 3.4907,
+      "step": 4802
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7128257155418396,
+      "learning_rate": 0.00017059984244772124,
+      "loss": 3.3174,
+      "step": 4803
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6172081232070923,
+      "learning_rate": 0.0001705015605282264,
+      "loss": 3.3234,
+      "step": 4804
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6061226725578308,
+      "learning_rate": 0.000170403292275435,
+      "loss": 3.3656,
+      "step": 4805
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5999137163162231,
+      "learning_rate": 0.00017030503770624046,
+      "loss": 3.3599,
+      "step": 4806
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5816587805747986,
+      "learning_rate": 0.0001702067968375337,
+      "loss": 3.2664,
+      "step": 4807
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.63521808385849,
+      "learning_rate": 0.00017010856968620373,
+      "loss": 3.4519,
+      "step": 4808
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6459574699401855,
+      "learning_rate": 0.00017001035626913678,
+      "loss": 3.258,
+      "step": 4809
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5902713537216187,
+      "learning_rate": 0.00016991215660321679,
+      "loss": 3.4486,
+      "step": 4810
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6274678707122803,
+      "learning_rate": 0.00016981397070532566,
+      "loss": 3.4292,
+      "step": 4811
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5549883246421814,
+      "learning_rate": 0.0001697157985923425,
+      "loss": 3.1933,
+      "step": 4812
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6459517478942871,
+      "learning_rate": 0.00016961764028114437,
+      "loss": 3.3454,
+      "step": 4813
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6086466908454895,
+      "learning_rate": 0.00016951949578860575,
+      "loss": 3.2549,
+      "step": 4814
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6149626970291138,
+      "learning_rate": 0.00016942136513159873,
+      "loss": 3.166,
+      "step": 4815
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6007905006408691,
+      "learning_rate": 0.00016932324832699325,
+      "loss": 3.3945,
+      "step": 4816
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6186566948890686,
+      "learning_rate": 0.0001692251453916567,
+      "loss": 3.376,
+      "step": 4817
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6166824102401733,
+      "learning_rate": 0.000169127056342454,
+      "loss": 3.3302,
+      "step": 4818
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6184083223342896,
+      "learning_rate": 0.00016902898119624794,
+      "loss": 3.251,
+      "step": 4819
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6125866770744324,
+      "learning_rate": 0.0001689309199698986,
+      "loss": 3.1848,
+      "step": 4820
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.604458212852478,
+      "learning_rate": 0.00016883287268026404,
+      "loss": 3.3634,
+      "step": 4821
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6307944655418396,
+      "learning_rate": 0.00016873483934419959,
+      "loss": 3.3647,
+      "step": 4822
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5935248136520386,
+      "learning_rate": 0.00016863681997855844,
+      "loss": 3.3112,
+      "step": 4823
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6155415773391724,
+      "learning_rate": 0.00016853881460019115,
+      "loss": 3.3834,
+      "step": 4824
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6031137704849243,
+      "learning_rate": 0.00016844082322594597,
+      "loss": 3.2919,
+      "step": 4825
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.713043212890625,
+      "learning_rate": 0.00016834284587266885,
+      "loss": 3.4577,
+      "step": 4826
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6130597591400146,
+      "learning_rate": 0.00016824488255720317,
+      "loss": 3.3312,
+      "step": 4827
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.620797336101532,
+      "learning_rate": 0.00016814693329638992,
+      "loss": 3.1559,
+      "step": 4828
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6594494581222534,
+      "learning_rate": 0.00016804899810706774,
+      "loss": 3.3446,
+      "step": 4829
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6507055163383484,
+      "learning_rate": 0.00016795107700607286,
+      "loss": 3.3051,
+      "step": 4830
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6062747836112976,
+      "learning_rate": 0.00016785317001023905,
+      "loss": 3.2461,
+      "step": 4831
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6378657221794128,
+      "learning_rate": 0.0001677552771363977,
+      "loss": 3.4018,
+      "step": 4832
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6071016192436218,
+      "learning_rate": 0.00016765739840137757,
+      "loss": 3.1889,
+      "step": 4833
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5815123319625854,
+      "learning_rate": 0.00016755953382200533,
+      "loss": 3.3255,
+      "step": 4834
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5857453942298889,
+      "learning_rate": 0.00016746168341510499,
+      "loss": 3.3624,
+      "step": 4835
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6075772643089294,
+      "learning_rate": 0.00016736384719749804,
+      "loss": 3.2764,
+      "step": 4836
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6338539719581604,
+      "learning_rate": 0.00016726602518600382,
+      "loss": 3.2596,
+      "step": 4837
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6194092035293579,
+      "learning_rate": 0.00016716821739743887,
+      "loss": 3.4251,
+      "step": 4838
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6281957030296326,
+      "learning_rate": 0.00016707042384861775,
+      "loss": 3.2663,
+      "step": 4839
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.583798348903656,
+      "learning_rate": 0.00016697264455635213,
+      "loss": 3.2086,
+      "step": 4840
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6021174788475037,
+      "learning_rate": 0.0001668748795374515,
+      "loss": 3.2542,
+      "step": 4841
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6187352538108826,
+      "learning_rate": 0.00016677712880872275,
+      "loss": 3.2407,
+      "step": 4842
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5813688039779663,
+      "learning_rate": 0.00016667939238697028,
+      "loss": 3.2851,
+      "step": 4843
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6295503377914429,
+      "learning_rate": 0.0001665816702889963,
+      "loss": 3.3658,
+      "step": 4844
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5846412777900696,
+      "learning_rate": 0.0001664839625316002,
+      "loss": 3.2269,
+      "step": 4845
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6589176058769226,
+      "learning_rate": 0.0001663862691315791,
+      "loss": 3.2868,
+      "step": 4846
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6424933671951294,
+      "learning_rate": 0.00016628859010572771,
+      "loss": 3.3598,
+      "step": 4847
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6255198121070862,
+      "learning_rate": 0.00016619092547083813,
+      "loss": 3.4012,
+      "step": 4848
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.656611442565918,
+      "learning_rate": 0.00016609327524370012,
+      "loss": 3.2743,
+      "step": 4849
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6469491720199585,
+      "learning_rate": 0.0001659956394411008,
+      "loss": 3.2577,
+      "step": 4850
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6087647676467896,
+      "learning_rate": 0.00016589801807982487,
+      "loss": 3.3577,
+      "step": 4851
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6154561638832092,
+      "learning_rate": 0.00016580041117665467,
+      "loss": 3.3821,
+      "step": 4852
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6215868592262268,
+      "learning_rate": 0.00016570281874836996,
+      "loss": 3.2785,
+      "step": 4853
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6483801007270813,
+      "learning_rate": 0.0001656052408117479,
+      "loss": 3.3564,
+      "step": 4854
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6019678115844727,
+      "learning_rate": 0.00016550767738356337,
+      "loss": 3.2738,
+      "step": 4855
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.626493513584137,
+      "learning_rate": 0.0001654101284805886,
+      "loss": 3.182,
+      "step": 4856
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6123741269111633,
+      "learning_rate": 0.00016531259411959332,
+      "loss": 3.3446,
+      "step": 4857
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6224625706672668,
+      "learning_rate": 0.00016521507431734492,
+      "loss": 3.3272,
+      "step": 4858
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6573577523231506,
+      "learning_rate": 0.0001651175690906082,
+      "loss": 3.325,
+      "step": 4859
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6397596001625061,
+      "learning_rate": 0.0001650200784561454,
+      "loss": 3.2862,
+      "step": 4860
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6268782019615173,
+      "learning_rate": 0.00016492260243071628,
+      "loss": 3.2584,
+      "step": 4861
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6043176054954529,
+      "learning_rate": 0.00016482514103107813,
+      "loss": 3.2901,
+      "step": 4862
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6334238648414612,
+      "learning_rate": 0.00016472769427398564,
+      "loss": 3.507,
+      "step": 4863
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6372676491737366,
+      "learning_rate": 0.00016463026217619105,
+      "loss": 3.2654,
+      "step": 4864
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6464550495147705,
+      "learning_rate": 0.00016453284475444413,
+      "loss": 3.2341,
+      "step": 4865
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6223431825637817,
+      "learning_rate": 0.00016443544202549189,
+      "loss": 3.3866,
+      "step": 4866
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5811198949813843,
+      "learning_rate": 0.00016433805400607924,
+      "loss": 3.2517,
+      "step": 4867
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6073024868965149,
+      "learning_rate": 0.00016424068071294818,
+      "loss": 3.3954,
+      "step": 4868
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6416819095611572,
+      "learning_rate": 0.00016414332216283827,
+      "loss": 3.3483,
+      "step": 4869
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6186531782150269,
+      "learning_rate": 0.00016404597837248663,
+      "loss": 3.3467,
+      "step": 4870
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.622018039226532,
+      "learning_rate": 0.0001639486493586278,
+      "loss": 3.359,
+      "step": 4871
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6513142585754395,
+      "learning_rate": 0.00016385133513799368,
+      "loss": 3.2877,
+      "step": 4872
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6026149988174438,
+      "learning_rate": 0.00016375403572731385,
+      "loss": 3.3425,
+      "step": 4873
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5751915574073792,
+      "learning_rate": 0.00016365675114331512,
+      "loss": 3.4418,
+      "step": 4874
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6088400483131409,
+      "learning_rate": 0.0001635594814027217,
+      "loss": 3.2938,
+      "step": 4875
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6468400955200195,
+      "learning_rate": 0.0001634622265222556,
+      "loss": 3.3112,
+      "step": 4876
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6035825610160828,
+      "learning_rate": 0.00016336498651863609,
+      "loss": 3.3464,
+      "step": 4877
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5807467699050903,
+      "learning_rate": 0.0001632677614085797,
+      "loss": 3.3286,
+      "step": 4878
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6074662804603577,
+      "learning_rate": 0.00016317055120880058,
+      "loss": 3.36,
+      "step": 4879
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.646324872970581,
+      "learning_rate": 0.00016307335593601036,
+      "loss": 3.491,
+      "step": 4880
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6242514848709106,
+      "learning_rate": 0.000162976175606918,
+      "loss": 3.5013,
+      "step": 4881
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6329924464225769,
+      "learning_rate": 0.00016287901023822988,
+      "loss": 3.3805,
+      "step": 4882
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6353751420974731,
+      "learning_rate": 0.00016278185984664996,
+      "loss": 3.3469,
+      "step": 4883
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6465364694595337,
+      "learning_rate": 0.00016268472444887932,
+      "loss": 3.384,
+      "step": 4884
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5814542770385742,
+      "learning_rate": 0.0001625876040616169,
+      "loss": 3.4397,
+      "step": 4885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5991275310516357,
+      "learning_rate": 0.00016249049870155875,
+      "loss": 3.3694,
+      "step": 4886
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6458463072776794,
+      "learning_rate": 0.00016239340838539827,
+      "loss": 3.2171,
+      "step": 4887
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5924831628799438,
+      "learning_rate": 0.00016229633312982655,
+      "loss": 3.3777,
+      "step": 4888
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6339801549911499,
+      "learning_rate": 0.00016219927295153198,
+      "loss": 3.3329,
+      "step": 4889
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5737445950508118,
+      "learning_rate": 0.00016210222786720014,
+      "loss": 3.2879,
+      "step": 4890
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6337724328041077,
+      "learning_rate": 0.00016200519789351443,
+      "loss": 3.3842,
+      "step": 4891
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5654064416885376,
+      "learning_rate": 0.0001619081830471553,
+      "loss": 3.2704,
+      "step": 4892
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6272947788238525,
+      "learning_rate": 0.00016181118334480073,
+      "loss": 3.4049,
+      "step": 4893
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7220709919929504,
+      "learning_rate": 0.00016171419880312604,
+      "loss": 3.4469,
+      "step": 4894
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6143007278442383,
+      "learning_rate": 0.00016161722943880416,
+      "loss": 3.3387,
+      "step": 4895
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6261044144630432,
+      "learning_rate": 0.00016152027526850519,
+      "loss": 3.3093,
+      "step": 4896
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.61616051197052,
+      "learning_rate": 0.0001614233363088966,
+      "loss": 3.3521,
+      "step": 4897
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6149140000343323,
+      "learning_rate": 0.0001613264125766434,
+      "loss": 3.2737,
+      "step": 4898
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.616252601146698,
+      "learning_rate": 0.00016122950408840785,
+      "loss": 3.279,
+      "step": 4899
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6254578828811646,
+      "learning_rate": 0.00016113261086084962,
+      "loss": 3.4044,
+      "step": 4900
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6441753506660461,
+      "learning_rate": 0.0001610357329106259,
+      "loss": 3.4349,
+      "step": 4901
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6455859541893005,
+      "learning_rate": 0.00016093887025439105,
+      "loss": 3.3825,
+      "step": 4902
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.632778525352478,
+      "learning_rate": 0.00016084202290879678,
+      "loss": 3.4236,
+      "step": 4903
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6407384872436523,
+      "learning_rate": 0.0001607451908904925,
+      "loss": 3.1993,
+      "step": 4904
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6580832004547119,
+      "learning_rate": 0.00016064837421612456,
+      "loss": 3.1839,
+      "step": 4905
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6400294899940491,
+      "learning_rate": 0.00016055157290233702,
+      "loss": 3.4126,
+      "step": 4906
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.60333251953125,
+      "learning_rate": 0.00016045478696577104,
+      "loss": 3.3403,
+      "step": 4907
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6334813833236694,
+      "learning_rate": 0.00016035801642306526,
+      "loss": 3.3563,
+      "step": 4908
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6013305187225342,
+      "learning_rate": 0.00016026126129085568,
+      "loss": 3.2698,
+      "step": 4909
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5717803239822388,
+      "learning_rate": 0.00016016452158577565,
+      "loss": 3.4212,
+      "step": 4910
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6447166204452515,
+      "learning_rate": 0.00016006779732445582,
+      "loss": 3.2419,
+      "step": 4911
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6601414084434509,
+      "learning_rate": 0.0001599710885235241,
+      "loss": 3.4306,
+      "step": 4912
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6667841672897339,
+      "learning_rate": 0.00015987439519960607,
+      "loss": 3.2909,
+      "step": 4913
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6420342922210693,
+      "learning_rate": 0.00015977771736932434,
+      "loss": 3.1509,
+      "step": 4914
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6115997433662415,
+      "learning_rate": 0.00015968105504929892,
+      "loss": 3.3316,
+      "step": 4915
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5998908281326294,
+      "learning_rate": 0.00015958440825614723,
+      "loss": 3.2116,
+      "step": 4916
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.654397189617157,
+      "learning_rate": 0.00015948777700648396,
+      "loss": 3.2198,
+      "step": 4917
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6434248685836792,
+      "learning_rate": 0.00015939116131692105,
+      "loss": 3.2118,
+      "step": 4918
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6479857563972473,
+      "learning_rate": 0.000159294561204068,
+      "loss": 3.384,
+      "step": 4919
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6829742193222046,
+      "learning_rate": 0.00015919797668453146,
+      "loss": 3.3759,
+      "step": 4920
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.643349289894104,
+      "learning_rate": 0.00015910140777491527,
+      "loss": 3.3625,
+      "step": 4921
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6263314485549927,
+      "learning_rate": 0.00015900485449182094,
+      "loss": 3.3384,
+      "step": 4922
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6248134970664978,
+      "learning_rate": 0.00015890831685184704,
+      "loss": 3.4037,
+      "step": 4923
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6030375361442566,
+      "learning_rate": 0.00015881179487158952,
+      "loss": 3.1696,
+      "step": 4924
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6380594372749329,
+      "learning_rate": 0.00015871528856764163,
+      "loss": 3.3414,
+      "step": 4925
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6514424085617065,
+      "learning_rate": 0.00015861879795659378,
+      "loss": 3.2199,
+      "step": 4926
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6601086854934692,
+      "learning_rate": 0.00015852232305503406,
+      "loss": 3.3459,
+      "step": 4927
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.663547933101654,
+      "learning_rate": 0.0001584258638795475,
+      "loss": 3.293,
+      "step": 4928
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6009930372238159,
+      "learning_rate": 0.00015832942044671647,
+      "loss": 3.2446,
+      "step": 4929
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6301782727241516,
+      "learning_rate": 0.00015823299277312086,
+      "loss": 3.2444,
+      "step": 4930
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6136788129806519,
+      "learning_rate": 0.00015813658087533757,
+      "loss": 3.4593,
+      "step": 4931
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5674062371253967,
+      "learning_rate": 0.0001580401847699411,
+      "loss": 3.2897,
+      "step": 4932
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5971000790596008,
+      "learning_rate": 0.00015794380447350288,
+      "loss": 3.3371,
+      "step": 4933
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6050028800964355,
+      "learning_rate": 0.00015784744000259195,
+      "loss": 3.3715,
+      "step": 4934
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6050082445144653,
+      "learning_rate": 0.0001577510913737744,
+      "loss": 3.2101,
+      "step": 4935
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6067320704460144,
+      "learning_rate": 0.00015765475860361365,
+      "loss": 3.3618,
+      "step": 4936
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.685673713684082,
+      "learning_rate": 0.00015755844170867048,
+      "loss": 3.2557,
+      "step": 4937
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6283639669418335,
+      "learning_rate": 0.00015746214070550286,
+      "loss": 3.4148,
+      "step": 4938
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5921222567558289,
+      "learning_rate": 0.00015736585561066603,
+      "loss": 3.2689,
+      "step": 4939
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6454620361328125,
+      "learning_rate": 0.00015726958644071248,
+      "loss": 3.2757,
+      "step": 4940
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6581416726112366,
+      "learning_rate": 0.00015717333321219206,
+      "loss": 3.393,
+      "step": 4941
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.608899712562561,
+      "learning_rate": 0.00015707709594165185,
+      "loss": 3.3873,
+      "step": 4942
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6669918298721313,
+      "learning_rate": 0.0001569808746456361,
+      "loss": 3.2282,
+      "step": 4943
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6613045334815979,
+      "learning_rate": 0.00015688466934068632,
+      "loss": 3.2838,
+      "step": 4944
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6391863822937012,
+      "learning_rate": 0.0001567884800433414,
+      "loss": 3.2624,
+      "step": 4945
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5971353650093079,
+      "learning_rate": 0.0001566923067701374,
+      "loss": 3.253,
+      "step": 4946
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.608292818069458,
+      "learning_rate": 0.0001565961495376075,
+      "loss": 3.3042,
+      "step": 4947
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6255372762680054,
+      "learning_rate": 0.0001565000083622824,
+      "loss": 3.2865,
+      "step": 4948
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6835498213768005,
+      "learning_rate": 0.00015640388326068969,
+      "loss": 3.3385,
+      "step": 4949
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6317881941795349,
+      "learning_rate": 0.0001563077742493546,
+      "loss": 3.3869,
+      "step": 4950
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6487517952919006,
+      "learning_rate": 0.00015621168134479925,
+      "loss": 3.3362,
+      "step": 4951
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6169304847717285,
+      "learning_rate": 0.00015611560456354323,
+      "loss": 3.2885,
+      "step": 4952
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6524413228034973,
+      "learning_rate": 0.00015601954392210316,
+      "loss": 3.2434,
+      "step": 4953
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6151243448257446,
+      "learning_rate": 0.00015592349943699296,
+      "loss": 3.2181,
+      "step": 4954
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.61757493019104,
+      "learning_rate": 0.0001558274711247239,
+      "loss": 3.374,
+      "step": 4955
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7145484089851379,
+      "learning_rate": 0.0001557314590018043,
+      "loss": 3.2289,
+      "step": 4956
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6981588006019592,
+      "learning_rate": 0.00015563546308473966,
+      "loss": 3.3561,
+      "step": 4957
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5915506482124329,
+      "learning_rate": 0.00015553948339003287,
+      "loss": 3.281,
+      "step": 4958
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6296604871749878,
+      "learning_rate": 0.00015544351993418404,
+      "loss": 3.404,
+      "step": 4959
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6014420986175537,
+      "learning_rate": 0.00015534757273369038,
+      "loss": 3.3068,
+      "step": 4960
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.621677815914154,
+      "learning_rate": 0.00015525164180504622,
+      "loss": 3.3712,
+      "step": 4961
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6184865832328796,
+      "learning_rate": 0.00015515572716474322,
+      "loss": 3.3306,
+      "step": 4962
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6415227651596069,
+      "learning_rate": 0.00015505982882927028,
+      "loss": 3.2403,
+      "step": 4963
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6465979218482971,
+      "learning_rate": 0.00015496394681511344,
+      "loss": 3.1784,
+      "step": 4964
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6245262026786804,
+      "learning_rate": 0.00015486808113875582,
+      "loss": 3.3508,
+      "step": 4965
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6168780326843262,
+      "learning_rate": 0.00015477223181667796,
+      "loss": 3.2795,
+      "step": 4966
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6115142107009888,
+      "learning_rate": 0.00015467639886535746,
+      "loss": 3.2947,
+      "step": 4967
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5774526000022888,
+      "learning_rate": 0.00015458058230126892,
+      "loss": 3.4082,
+      "step": 4968
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5983077883720398,
+      "learning_rate": 0.00015448478214088456,
+      "loss": 3.4332,
+      "step": 4969
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5960877537727356,
+      "learning_rate": 0.00015438899840067356,
+      "loss": 3.2596,
+      "step": 4970
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.601218581199646,
+      "learning_rate": 0.00015429323109710207,
+      "loss": 3.2533,
+      "step": 4971
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.587386429309845,
+      "learning_rate": 0.0001541974802466337,
+      "loss": 3.3526,
+      "step": 4972
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6104817986488342,
+      "learning_rate": 0.00015410174586572912,
+      "loss": 3.2944,
+      "step": 4973
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6067850589752197,
+      "learning_rate": 0.0001540060279708462,
+      "loss": 3.361,
+      "step": 4974
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6144149303436279,
+      "learning_rate": 0.00015391032657843988,
+      "loss": 3.2056,
+      "step": 4975
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6430124640464783,
+      "learning_rate": 0.0001538146417049624,
+      "loss": 3.322,
+      "step": 4976
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6340519785881042,
+      "learning_rate": 0.00015371897336686303,
+      "loss": 3.2947,
+      "step": 4977
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6688194870948792,
+      "learning_rate": 0.0001536233215805884,
+      "loss": 3.2453,
+      "step": 4978
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7047507166862488,
+      "learning_rate": 0.000153527686362582,
+      "loss": 3.3035,
+      "step": 4979
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6342984437942505,
+      "learning_rate": 0.00015343206772928486,
+      "loss": 3.4161,
+      "step": 4980
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6949705481529236,
+      "learning_rate": 0.00015333646569713473,
+      "loss": 3.2528,
+      "step": 4981
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6137412786483765,
+      "learning_rate": 0.00015324088028256677,
+      "loss": 3.3913,
+      "step": 4982
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6763321757316589,
+      "learning_rate": 0.00015314531150201316,
+      "loss": 3.3005,
+      "step": 4983
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6877267956733704,
+      "learning_rate": 0.0001530497593719034,
+      "loss": 3.3418,
+      "step": 4984
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.633033812046051,
+      "learning_rate": 0.00015295422390866398,
+      "loss": 3.2597,
+      "step": 4985
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6207406520843506,
+      "learning_rate": 0.00015285870512871835,
+      "loss": 3.3557,
+      "step": 4986
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6227355599403381,
+      "learning_rate": 0.00015276320304848757,
+      "loss": 3.2997,
+      "step": 4987
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6410094499588013,
+      "learning_rate": 0.00015266771768438948,
+      "loss": 3.3759,
+      "step": 4988
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6022902131080627,
+      "learning_rate": 0.00015257224905283913,
+      "loss": 3.2474,
+      "step": 4989
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6035560965538025,
+      "learning_rate": 0.00015247679717024854,
+      "loss": 3.3183,
+      "step": 4990
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6439880132675171,
+      "learning_rate": 0.0001523813620530272,
+      "loss": 3.3303,
+      "step": 4991
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6099724173545837,
+      "learning_rate": 0.00015228594371758137,
+      "loss": 3.223,
+      "step": 4992
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6340009570121765,
+      "learning_rate": 0.00015219054218031458,
+      "loss": 3.2488,
+      "step": 4993
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6087921261787415,
+      "learning_rate": 0.0001520951574576276,
+      "loss": 3.2731,
+      "step": 4994
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6423810124397278,
+      "learning_rate": 0.0001519997895659179,
+      "loss": 3.1672,
+      "step": 4995
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5887874960899353,
+      "learning_rate": 0.00015190443852158058,
+      "loss": 3.2284,
+      "step": 4996
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6974184513092041,
+      "learning_rate": 0.00015180910434100747,
+      "loss": 3.3678,
+      "step": 4997
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5889269113540649,
+      "learning_rate": 0.00015171378704058772,
+      "loss": 3.2248,
+      "step": 4998
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6352993249893188,
+      "learning_rate": 0.0001516184866367074,
+      "loss": 3.2472,
+      "step": 4999
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5974810719490051,
+      "learning_rate": 0.00015152320314574974,
+      "loss": 3.2709,
+      "step": 5000
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6414270997047424,
+      "learning_rate": 0.0001514279365840951,
+      "loss": 3.2805,
+      "step": 5001
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6457729339599609,
+      "learning_rate": 0.0001513326869681209,
+      "loss": 3.2465,
+      "step": 5002
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5834159851074219,
+      "learning_rate": 0.00015123745431420169,
+      "loss": 3.3469,
+      "step": 5003
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6347243189811707,
+      "learning_rate": 0.000151142238638709,
+      "loss": 3.2968,
+      "step": 5004
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6431541442871094,
+      "learning_rate": 0.00015104703995801145,
+      "loss": 3.2673,
+      "step": 5005
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6189345717430115,
+      "learning_rate": 0.000150951858288475,
+      "loss": 3.3624,
+      "step": 5006
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6066557168960571,
+      "learning_rate": 0.00015085669364646242,
+      "loss": 3.2837,
+      "step": 5007
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6095768809318542,
+      "learning_rate": 0.0001507615460483335,
+      "loss": 3.2273,
+      "step": 5008
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6279232501983643,
+      "learning_rate": 0.00015066641551044532,
+      "loss": 3.3132,
+      "step": 5009
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6557057499885559,
+      "learning_rate": 0.00015057130204915192,
+      "loss": 3.4269,
+      "step": 5010
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6350106596946716,
+      "learning_rate": 0.00015047620568080428,
+      "loss": 3.2787,
+      "step": 5011
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6168920993804932,
+      "learning_rate": 0.00015038112642175072,
+      "loss": 3.2072,
+      "step": 5012
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6138224005699158,
+      "learning_rate": 0.00015028606428833645,
+      "loss": 3.3762,
+      "step": 5013
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6427010893821716,
+      "learning_rate": 0.00015019101929690358,
+      "loss": 3.4177,
+      "step": 5014
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6091998815536499,
+      "learning_rate": 0.00015009599146379162,
+      "loss": 3.3719,
+      "step": 5015
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6215112209320068,
+      "learning_rate": 0.00015000098080533697,
+      "loss": 3.3247,
+      "step": 5016
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6604453325271606,
+      "learning_rate": 0.00014990598733787304,
+      "loss": 3.2659,
+      "step": 5017
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.643018364906311,
+      "learning_rate": 0.0001498110110777302,
+      "loss": 3.1471,
+      "step": 5018
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6082643270492554,
+      "learning_rate": 0.00014971605204123608,
+      "loss": 3.2765,
+      "step": 5019
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6341992020606995,
+      "learning_rate": 0.00014962111024471522,
+      "loss": 3.4197,
+      "step": 5020
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6241186261177063,
+      "learning_rate": 0.00014952618570448923,
+      "loss": 3.3389,
+      "step": 5021
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6271200180053711,
+      "learning_rate": 0.00014943127843687658,
+      "loss": 3.2994,
+      "step": 5022
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6608973741531372,
+      "learning_rate": 0.0001493363884581931,
+      "loss": 3.4353,
+      "step": 5023
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6565616726875305,
+      "learning_rate": 0.0001492415157847515,
+      "loss": 3.3183,
+      "step": 5024
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5876268744468689,
+      "learning_rate": 0.0001491466604328614,
+      "loss": 3.3067,
+      "step": 5025
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5943182706832886,
+      "learning_rate": 0.00014905182241882955,
+      "loss": 3.276,
+      "step": 5026
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5989049673080444,
+      "learning_rate": 0.00014895700175895978,
+      "loss": 3.3227,
+      "step": 5027
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6210485696792603,
+      "learning_rate": 0.00014886219846955276,
+      "loss": 3.2309,
+      "step": 5028
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6408283114433289,
+      "learning_rate": 0.0001487674125669063,
+      "loss": 3.3571,
+      "step": 5029
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6775360107421875,
+      "learning_rate": 0.00014867264406731524,
+      "loss": 3.3817,
+      "step": 5030
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7827131748199463,
+      "learning_rate": 0.00014857789298707133,
+      "loss": 3.2939,
+      "step": 5031
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6316560506820679,
+      "learning_rate": 0.0001484831593424633,
+      "loss": 3.0717,
+      "step": 5032
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5866991877555847,
+      "learning_rate": 0.00014838844314977719,
+      "loss": 3.4829,
+      "step": 5033
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.652138888835907,
+      "learning_rate": 0.00014829374442529563,
+      "loss": 3.3932,
+      "step": 5034
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5911225080490112,
+      "learning_rate": 0.00014819906318529858,
+      "loss": 3.3848,
+      "step": 5035
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6234878301620483,
+      "learning_rate": 0.00014810439944606263,
+      "loss": 3.2459,
+      "step": 5036
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6251317262649536,
+      "learning_rate": 0.00014800975322386175,
+      "loss": 3.2907,
+      "step": 5037
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6247319579124451,
+      "learning_rate": 0.00014791512453496669,
+      "loss": 3.3405,
+      "step": 5038
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6416252255439758,
+      "learning_rate": 0.00014782051339564512,
+      "loss": 3.3899,
+      "step": 5039
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6290828585624695,
+      "learning_rate": 0.00014772591982216193,
+      "loss": 3.3738,
+      "step": 5040
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5640152096748352,
+      "learning_rate": 0.00014763134383077875,
+      "loss": 3.3445,
+      "step": 5041
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6488463878631592,
+      "learning_rate": 0.00014753678543775428,
+      "loss": 3.3709,
+      "step": 5042
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6063884496688843,
+      "learning_rate": 0.0001474422446593443,
+      "loss": 3.4388,
+      "step": 5043
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6267374157905579,
+      "learning_rate": 0.0001473477215118014,
+      "loss": 3.2639,
+      "step": 5044
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6429750919342041,
+      "learning_rate": 0.00014725321601137526,
+      "loss": 3.3561,
+      "step": 5045
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6482391357421875,
+      "learning_rate": 0.00014715872817431242,
+      "loss": 3.2559,
+      "step": 5046
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6009674072265625,
+      "learning_rate": 0.0001470642580168564,
+      "loss": 3.3531,
+      "step": 5047
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6284033060073853,
+      "learning_rate": 0.0001469698055552478,
+      "loss": 3.323,
+      "step": 5048
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6527823209762573,
+      "learning_rate": 0.00014687537080572405,
+      "loss": 3.2579,
+      "step": 5049
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.630158007144928,
+      "learning_rate": 0.00014678095378451955,
+      "loss": 3.4153,
+      "step": 5050
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.631565272808075,
+      "learning_rate": 0.00014668655450786566,
+      "loss": 3.2825,
+      "step": 5051
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6248773336410522,
+      "learning_rate": 0.00014659217299199084,
+      "loss": 3.2931,
+      "step": 5052
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.608366847038269,
+      "learning_rate": 0.00014649780925312034,
+      "loss": 3.3213,
+      "step": 5053
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6143530607223511,
+      "learning_rate": 0.00014640346330747623,
+      "loss": 3.3564,
+      "step": 5054
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6777532696723938,
+      "learning_rate": 0.00014630913517127786,
+      "loss": 3.3336,
+      "step": 5055
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5899236798286438,
+      "learning_rate": 0.00014621482486074121,
+      "loss": 3.1634,
+      "step": 5056
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6264376640319824,
+      "learning_rate": 0.00014612053239207928,
+      "loss": 3.2912,
+      "step": 5057
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5884184241294861,
+      "learning_rate": 0.0001460262577815022,
+      "loss": 3.1504,
+      "step": 5058
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6290861368179321,
+      "learning_rate": 0.0001459320010452167,
+      "loss": 3.2264,
+      "step": 5059
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.648082435131073,
+      "learning_rate": 0.00014583776219942665,
+      "loss": 3.3968,
+      "step": 5060
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6225313544273376,
+      "learning_rate": 0.00014574354126033293,
+      "loss": 3.2704,
+      "step": 5061
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6109769344329834,
+      "learning_rate": 0.0001456493382441331,
+      "loss": 3.1933,
+      "step": 5062
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6114202737808228,
+      "learning_rate": 0.00014555515316702175,
+      "loss": 3.2708,
+      "step": 5063
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6244300603866577,
+      "learning_rate": 0.0001454609860451904,
+      "loss": 3.3708,
+      "step": 5064
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6992241740226746,
+      "learning_rate": 0.0001453668368948275,
+      "loss": 3.1895,
+      "step": 5065
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6371669769287109,
+      "learning_rate": 0.00014527270573211823,
+      "loss": 3.3398,
+      "step": 5066
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6179408431053162,
+      "learning_rate": 0.00014517859257324507,
+      "loss": 3.2487,
+      "step": 5067
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6645209193229675,
+      "learning_rate": 0.000145084497434387,
+      "loss": 3.313,
+      "step": 5068
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5824407339096069,
+      "learning_rate": 0.00014499042033172,
+      "loss": 3.2539,
+      "step": 5069
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6037798523902893,
+      "learning_rate": 0.00014489636128141726,
+      "loss": 3.2942,
+      "step": 5070
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6334280967712402,
+      "learning_rate": 0.00014480232029964849,
+      "loss": 3.3655,
+      "step": 5071
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5741010308265686,
+      "learning_rate": 0.0001447082974025804,
+      "loss": 3.2918,
+      "step": 5072
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6136914491653442,
+      "learning_rate": 0.0001446142926063766,
+      "loss": 3.2696,
+      "step": 5073
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6093493103981018,
+      "learning_rate": 0.00014452030592719756,
+      "loss": 3.1996,
+      "step": 5074
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6550522446632385,
+      "learning_rate": 0.00014442633738120087,
+      "loss": 3.3622,
+      "step": 5075
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6372076869010925,
+      "learning_rate": 0.0001443323869845407,
+      "loss": 3.3896,
+      "step": 5076
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6629884839057922,
+      "learning_rate": 0.0001442384547533682,
+      "loss": 3.348,
+      "step": 5077
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6217200756072998,
+      "learning_rate": 0.00014414454070383142,
+      "loss": 3.2738,
+      "step": 5078
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6206686496734619,
+      "learning_rate": 0.00014405064485207517,
+      "loss": 3.2477,
+      "step": 5079
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6034120321273804,
+      "learning_rate": 0.00014395676721424145,
+      "loss": 3.3615,
+      "step": 5080
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5921332836151123,
+      "learning_rate": 0.00014386290780646872,
+      "loss": 3.3844,
+      "step": 5081
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6570034027099609,
+      "learning_rate": 0.00014376906664489265,
+      "loss": 3.4145,
+      "step": 5082
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5640652179718018,
+      "learning_rate": 0.00014367524374564556,
+      "loss": 3.2098,
+      "step": 5083
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6425613760948181,
+      "learning_rate": 0.00014358143912485672,
+      "loss": 3.3405,
+      "step": 5084
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6288067698478699,
+      "learning_rate": 0.0001434876527986522,
+      "loss": 3.3572,
+      "step": 5085
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5896071195602417,
+      "learning_rate": 0.00014339388478315496,
+      "loss": 3.2689,
+      "step": 5086
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6366032361984253,
+      "learning_rate": 0.00014330013509448468,
+      "loss": 3.4776,
+      "step": 5087
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6185347437858582,
+      "learning_rate": 0.00014320640374875827,
+      "loss": 3.2119,
+      "step": 5088
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6984763145446777,
+      "learning_rate": 0.00014311269076208903,
+      "loss": 3.3938,
+      "step": 5089
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6636654734611511,
+      "learning_rate": 0.00014301899615058747,
+      "loss": 3.2879,
+      "step": 5090
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.576370894908905,
+      "learning_rate": 0.0001429253199303607,
+      "loss": 3.3888,
+      "step": 5091
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6354531645774841,
+      "learning_rate": 0.00014283166211751276,
+      "loss": 3.2763,
+      "step": 5092
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6325899362564087,
+      "learning_rate": 0.0001427380227281445,
+      "loss": 3.0955,
+      "step": 5093
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6241953372955322,
+      "learning_rate": 0.00014264440177835363,
+      "loss": 3.2864,
+      "step": 5094
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6474228501319885,
+      "learning_rate": 0.00014255079928423455,
+      "loss": 3.2663,
+      "step": 5095
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6273337006568909,
+      "learning_rate": 0.00014245721526187882,
+      "loss": 3.4285,
+      "step": 5096
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5948605537414551,
+      "learning_rate": 0.00014236364972737447,
+      "loss": 3.2559,
+      "step": 5097
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6316484808921814,
+      "learning_rate": 0.00014227010269680663,
+      "loss": 3.297,
+      "step": 5098
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6476810574531555,
+      "learning_rate": 0.00014217657418625707,
+      "loss": 3.3244,
+      "step": 5099
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6473725438117981,
+      "learning_rate": 0.0001420830642118044,
+      "loss": 3.2591,
+      "step": 5100
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7240094542503357,
+      "learning_rate": 0.00014198957278952406,
+      "loss": 3.3884,
+      "step": 5101
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6182361245155334,
+      "learning_rate": 0.00014189609993548824,
+      "loss": 3.293,
+      "step": 5102
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6440767049789429,
+      "learning_rate": 0.00014180264566576617,
+      "loss": 3.3392,
+      "step": 5103
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6826469898223877,
+      "learning_rate": 0.0001417092099964236,
+      "loss": 3.2833,
+      "step": 5104
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6628642082214355,
+      "learning_rate": 0.00014161579294352333,
+      "loss": 3.3416,
+      "step": 5105
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6261721849441528,
+      "learning_rate": 0.0001415223945231246,
+      "loss": 3.3862,
+      "step": 5106
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5831178426742554,
+      "learning_rate": 0.00014142901475128395,
+      "loss": 3.272,
+      "step": 5107
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6141115427017212,
+      "learning_rate": 0.00014133565364405433,
+      "loss": 3.3945,
+      "step": 5108
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.58644038438797,
+      "learning_rate": 0.0001412423112174856,
+      "loss": 3.2059,
+      "step": 5109
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5988340377807617,
+      "learning_rate": 0.0001411489874876243,
+      "loss": 3.3375,
+      "step": 5110
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6062052249908447,
+      "learning_rate": 0.00014105568247051403,
+      "loss": 3.3096,
+      "step": 5111
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6299193501472473,
+      "learning_rate": 0.00014096239618219492,
+      "loss": 3.3522,
+      "step": 5112
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6424022316932678,
+      "learning_rate": 0.00014086912863870403,
+      "loss": 3.1548,
+      "step": 5113
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6243067979812622,
+      "learning_rate": 0.00014077587985607504,
+      "loss": 3.3714,
+      "step": 5114
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6548874378204346,
+      "learning_rate": 0.00014068264985033857,
+      "loss": 3.354,
+      "step": 5115
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6031063199043274,
+      "learning_rate": 0.00014058943863752178,
+      "loss": 3.2291,
+      "step": 5116
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6349620223045349,
+      "learning_rate": 0.0001404962462336489,
+      "loss": 3.2758,
+      "step": 5117
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7188941836357117,
+      "learning_rate": 0.00014040307265474086,
+      "loss": 3.208,
+      "step": 5118
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6274533867835999,
+      "learning_rate": 0.00014030991791681518,
+      "loss": 3.3663,
+      "step": 5119
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6238290071487427,
+      "learning_rate": 0.00014021678203588627,
+      "loss": 3.3302,
+      "step": 5120
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6041349172592163,
+      "learning_rate": 0.00014012366502796526,
+      "loss": 3.2348,
+      "step": 5121
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6019901037216187,
+      "learning_rate": 0.00014003056690906,
+      "loss": 3.2653,
+      "step": 5122
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6081122756004333,
+      "learning_rate": 0.00013993748769517507,
+      "loss": 3.4407,
+      "step": 5123
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6346868872642517,
+      "learning_rate": 0.00013984442740231203,
+      "loss": 3.287,
+      "step": 5124
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.587018609046936,
+      "learning_rate": 0.00013975138604646888,
+      "loss": 3.4186,
+      "step": 5125
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5837512612342834,
+      "learning_rate": 0.00013965836364364067,
+      "loss": 3.249,
+      "step": 5126
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.608168363571167,
+      "learning_rate": 0.00013956536020981897,
+      "loss": 3.2367,
+      "step": 5127
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6023334264755249,
+      "learning_rate": 0.0001394723757609921,
+      "loss": 3.3814,
+      "step": 5128
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6250126361846924,
+      "learning_rate": 0.00013937941031314516,
+      "loss": 3.2889,
+      "step": 5129
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6232753992080688,
+      "learning_rate": 0.00013928646388226002,
+      "loss": 3.329,
+      "step": 5130
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6164801716804504,
+      "learning_rate": 0.00013919353648431516,
+      "loss": 3.297,
+      "step": 5131
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6190966963768005,
+      "learning_rate": 0.00013910062813528605,
+      "loss": 3.3271,
+      "step": 5132
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6317085027694702,
+      "learning_rate": 0.0001390077388511446,
+      "loss": 3.4416,
+      "step": 5133
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5752818584442139,
+      "learning_rate": 0.0001389148686478595,
+      "loss": 3.2217,
+      "step": 5134
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6402003169059753,
+      "learning_rate": 0.00013882201754139638,
+      "loss": 3.3543,
+      "step": 5135
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6417957544326782,
+      "learning_rate": 0.0001387291855477173,
+      "loss": 3.2738,
+      "step": 5136
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6363396644592285,
+      "learning_rate": 0.00013863637268278123,
+      "loss": 3.2036,
+      "step": 5137
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6913270950317383,
+      "learning_rate": 0.0001385435789625436,
+      "loss": 3.3316,
+      "step": 5138
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6306250691413879,
+      "learning_rate": 0.00013845080440295698,
+      "loss": 3.3967,
+      "step": 5139
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6392128467559814,
+      "learning_rate": 0.00013835804901997029,
+      "loss": 3.4899,
+      "step": 5140
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6294986605644226,
+      "learning_rate": 0.0001382653128295292,
+      "loss": 3.2473,
+      "step": 5141
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6198727488517761,
+      "learning_rate": 0.00013817259584757619,
+      "loss": 3.3429,
+      "step": 5142
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6190699338912964,
+      "learning_rate": 0.0001380798980900503,
+      "loss": 3.3584,
+      "step": 5143
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5895346999168396,
+      "learning_rate": 0.00013798721957288747,
+      "loss": 3.4016,
+      "step": 5144
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6533825993537903,
+      "learning_rate": 0.0001378945603120202,
+      "loss": 3.3099,
+      "step": 5145
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6693593263626099,
+      "learning_rate": 0.00013780192032337752,
+      "loss": 3.23,
+      "step": 5146
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6089479327201843,
+      "learning_rate": 0.00013770929962288552,
+      "loss": 3.3238,
+      "step": 5147
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.644583523273468,
+      "learning_rate": 0.00013761669822646676,
+      "loss": 3.2007,
+      "step": 5148
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6629884243011475,
+      "learning_rate": 0.0001375241161500404,
+      "loss": 3.2607,
+      "step": 5149
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6261684894561768,
+      "learning_rate": 0.00013743155340952242,
+      "loss": 3.3117,
+      "step": 5150
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6273629665374756,
+      "learning_rate": 0.00013733901002082544,
+      "loss": 3.2865,
+      "step": 5151
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6482716202735901,
+      "learning_rate": 0.00013724648599985857,
+      "loss": 3.328,
+      "step": 5152
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6402632594108582,
+      "learning_rate": 0.00013715398136252794,
+      "loss": 3.3984,
+      "step": 5153
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6050584316253662,
+      "learning_rate": 0.0001370614961247362,
+      "loss": 3.4794,
+      "step": 5154
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6375922560691833,
+      "learning_rate": 0.00013696903030238262,
+      "loss": 3.2708,
+      "step": 5155
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6117022037506104,
+      "learning_rate": 0.00013687658391136305,
+      "loss": 3.4379,
+      "step": 5156
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6070494651794434,
+      "learning_rate": 0.00013678415696757016,
+      "loss": 3.2773,
+      "step": 5157
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5802207589149475,
+      "learning_rate": 0.00013669174948689318,
+      "loss": 3.2079,
+      "step": 5158
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5739595293998718,
+      "learning_rate": 0.000136599361485218,
+      "loss": 3.3533,
+      "step": 5159
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6401107907295227,
+      "learning_rate": 0.0001365069929784273,
+      "loss": 3.31,
+      "step": 5160
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6143526434898376,
+      "learning_rate": 0.00013641464398240021,
+      "loss": 3.3205,
+      "step": 5161
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6221523284912109,
+      "learning_rate": 0.00013632231451301256,
+      "loss": 3.3169,
+      "step": 5162
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6104118824005127,
+      "learning_rate": 0.000136230004586137,
+      "loss": 3.3427,
+      "step": 5163
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6160389184951782,
+      "learning_rate": 0.00013613771421764254,
+      "loss": 3.3304,
+      "step": 5164
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.607818603515625,
+      "learning_rate": 0.00013604544342339506,
+      "loss": 3.2499,
+      "step": 5165
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.64719557762146,
+      "learning_rate": 0.0001359531922192569,
+      "loss": 3.2432,
+      "step": 5166
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.597808837890625,
+      "learning_rate": 0.0001358609606210871,
+      "loss": 3.3301,
+      "step": 5167
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6585919260978699,
+      "learning_rate": 0.00013576874864474142,
+      "loss": 3.3205,
+      "step": 5168
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6387555599212646,
+      "learning_rate": 0.0001356765563060721,
+      "loss": 3.3725,
+      "step": 5169
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6279529929161072,
+      "learning_rate": 0.00013558438362092816,
+      "loss": 3.3865,
+      "step": 5170
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6626907587051392,
+      "learning_rate": 0.00013549223060515503,
+      "loss": 3.2946,
+      "step": 5171
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6042270064353943,
+      "learning_rate": 0.000135400097274595,
+      "loss": 3.3273,
+      "step": 5172
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5902952551841736,
+      "learning_rate": 0.00013530798364508678,
+      "loss": 3.4654,
+      "step": 5173
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6035054326057434,
+      "learning_rate": 0.00013521588973246573,
+      "loss": 3.1904,
+      "step": 5174
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6701990962028503,
+      "learning_rate": 0.00013512381555256403,
+      "loss": 3.2674,
+      "step": 5175
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6053285598754883,
+      "learning_rate": 0.0001350317611212102,
+      "loss": 3.2791,
+      "step": 5176
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6039866805076599,
+      "learning_rate": 0.00013493972645422942,
+      "loss": 3.2819,
+      "step": 5177
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6498487591743469,
+      "learning_rate": 0.00013484771156744356,
+      "loss": 3.4138,
+      "step": 5178
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5941669940948486,
+      "learning_rate": 0.0001347557164766711,
+      "loss": 3.3125,
+      "step": 5179
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5859676599502563,
+      "learning_rate": 0.00013466374119772685,
+      "loss": 3.2442,
+      "step": 5180
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6273823976516724,
+      "learning_rate": 0.0001345717857464226,
+      "loss": 3.3161,
+      "step": 5181
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.641112744808197,
+      "learning_rate": 0.00013447985013856665,
+      "loss": 3.3408,
+      "step": 5182
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6055769324302673,
+      "learning_rate": 0.00013438793438996365,
+      "loss": 3.2671,
+      "step": 5183
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6235153675079346,
+      "learning_rate": 0.00013429603851641505,
+      "loss": 3.2637,
+      "step": 5184
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6751651763916016,
+      "learning_rate": 0.0001342041625337188,
+      "loss": 3.1511,
+      "step": 5185
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6396401524543762,
+      "learning_rate": 0.00013411230645766936,
+      "loss": 3.2449,
+      "step": 5186
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.623742938041687,
+      "learning_rate": 0.000134020470304058,
+      "loss": 3.3573,
+      "step": 5187
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6788823008537292,
+      "learning_rate": 0.00013392865408867223,
+      "loss": 3.3257,
+      "step": 5188
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6515249013900757,
+      "learning_rate": 0.0001338368578272965,
+      "loss": 3.3549,
+      "step": 5189
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.624242901802063,
+      "learning_rate": 0.00013374508153571153,
+      "loss": 3.2207,
+      "step": 5190
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7381793260574341,
+      "learning_rate": 0.00013365332522969486,
+      "loss": 3.3485,
+      "step": 5191
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5954594016075134,
+      "learning_rate": 0.00013356158892502038,
+      "loss": 3.2959,
+      "step": 5192
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6428266763687134,
+      "learning_rate": 0.00013346987263745862,
+      "loss": 3.2255,
+      "step": 5193
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6748354434967041,
+      "learning_rate": 0.00013337817638277673,
+      "loss": 3.3791,
+      "step": 5194
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.612354040145874,
+      "learning_rate": 0.0001332865001767382,
+      "loss": 3.3375,
+      "step": 5195
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6095655560493469,
+      "learning_rate": 0.00013319484403510345,
+      "loss": 3.2663,
+      "step": 5196
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5803064107894897,
+      "learning_rate": 0.00013310320797362915,
+      "loss": 3.3649,
+      "step": 5197
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6813079118728638,
+      "learning_rate": 0.00013301159200806856,
+      "loss": 3.3442,
+      "step": 5198
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6259683966636658,
+      "learning_rate": 0.00013291999615417147,
+      "loss": 3.4004,
+      "step": 5199
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5949110388755798,
+      "learning_rate": 0.00013282842042768446,
+      "loss": 3.2591,
+      "step": 5200
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.641460120677948,
+      "learning_rate": 0.0001327368648443503,
+      "loss": 3.3612,
+      "step": 5201
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5888059735298157,
+      "learning_rate": 0.00013264532941990853,
+      "loss": 3.3093,
+      "step": 5202
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.629054605960846,
+      "learning_rate": 0.00013255381417009502,
+      "loss": 3.2571,
+      "step": 5203
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6009963154792786,
+      "learning_rate": 0.0001324623191106425,
+      "loss": 3.2093,
+      "step": 5204
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6035163998603821,
+      "learning_rate": 0.00013237084425727995,
+      "loss": 3.346,
+      "step": 5205
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6143518090248108,
+      "learning_rate": 0.00013227938962573295,
+      "loss": 3.3584,
+      "step": 5206
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6170941591262817,
+      "learning_rate": 0.0001321879552317236,
+      "loss": 3.2979,
+      "step": 5207
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5544931888580322,
+      "learning_rate": 0.00013209654109097043,
+      "loss": 3.3117,
+      "step": 5208
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5998968482017517,
+      "learning_rate": 0.00013200514721918883,
+      "loss": 3.1606,
+      "step": 5209
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6752246618270874,
+      "learning_rate": 0.00013191377363209022,
+      "loss": 3.2952,
+      "step": 5210
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.656449019908905,
+      "learning_rate": 0.000131822420345383,
+      "loss": 3.2877,
+      "step": 5211
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6037731170654297,
+      "learning_rate": 0.00013173108737477173,
+      "loss": 3.3158,
+      "step": 5212
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6379036903381348,
+      "learning_rate": 0.00013163977473595767,
+      "loss": 3.3287,
+      "step": 5213
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.569927990436554,
+      "learning_rate": 0.00013154848244463846,
+      "loss": 3.297,
+      "step": 5214
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6360076665878296,
+      "learning_rate": 0.00013145721051650833,
+      "loss": 3.3574,
+      "step": 5215
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5665431618690491,
+      "learning_rate": 0.00013136595896725786,
+      "loss": 3.2449,
+      "step": 5216
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.583529531955719,
+      "learning_rate": 0.00013127472781257439,
+      "loss": 3.2924,
+      "step": 5217
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5858355164527893,
+      "learning_rate": 0.0001311835170681417,
+      "loss": 3.2698,
+      "step": 5218
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6250311732292175,
+      "learning_rate": 0.00013109232674963982,
+      "loss": 3.3428,
+      "step": 5219
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6676336526870728,
+      "learning_rate": 0.0001310011568727455,
+      "loss": 3.27,
+      "step": 5220
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6124276518821716,
+      "learning_rate": 0.00013091000745313187,
+      "loss": 3.2371,
+      "step": 5221
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6961532235145569,
+      "learning_rate": 0.00013081887850646857,
+      "loss": 3.2868,
+      "step": 5222
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6264037489891052,
+      "learning_rate": 0.0001307277700484217,
+      "loss": 3.2977,
+      "step": 5223
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6278262138366699,
+      "learning_rate": 0.0001306366820946538,
+      "loss": 3.4374,
+      "step": 5224
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.598902702331543,
+      "learning_rate": 0.00013054561466082411,
+      "loss": 3.2982,
+      "step": 5225
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6131187081336975,
+      "learning_rate": 0.00013045456776258812,
+      "loss": 3.502,
+      "step": 5226
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6919788718223572,
+      "learning_rate": 0.0001303635414155977,
+      "loss": 3.3583,
+      "step": 5227
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6344329714775085,
+      "learning_rate": 0.00013027253563550157,
+      "loss": 3.3588,
+      "step": 5228
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6490606069564819,
+      "learning_rate": 0.00013018155043794454,
+      "loss": 3.2091,
+      "step": 5229
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6615791320800781,
+      "learning_rate": 0.00013009058583856808,
+      "loss": 3.3525,
+      "step": 5230
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5891029834747314,
+      "learning_rate": 0.0001299996418530099,
+      "loss": 3.3269,
+      "step": 5231
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.602630078792572,
+      "learning_rate": 0.00012990871849690455,
+      "loss": 3.3488,
+      "step": 5232
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6360301375389099,
+      "learning_rate": 0.0001298178157858827,
+      "loss": 3.3218,
+      "step": 5233
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5955541729927063,
+      "learning_rate": 0.0001297269337355716,
+      "loss": 3.2836,
+      "step": 5234
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5877676606178284,
+      "learning_rate": 0.0001296360723615949,
+      "loss": 3.3302,
+      "step": 5235
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5851530432701111,
+      "learning_rate": 0.00012954523167957267,
+      "loss": 3.4018,
+      "step": 5236
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6754768490791321,
+      "learning_rate": 0.0001294544117051216,
+      "loss": 3.3248,
+      "step": 5237
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6499289870262146,
+      "learning_rate": 0.00012936361245385457,
+      "loss": 3.271,
+      "step": 5238
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5828032493591309,
+      "learning_rate": 0.00012927283394138122,
+      "loss": 3.3106,
+      "step": 5239
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.622321367263794,
+      "learning_rate": 0.0001291820761833073,
+      "loss": 3.2305,
+      "step": 5240
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.635196328163147,
+      "learning_rate": 0.0001290913391952351,
+      "loss": 3.2854,
+      "step": 5241
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6155919432640076,
+      "learning_rate": 0.00012900062299276338,
+      "loss": 3.3467,
+      "step": 5242
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6434239149093628,
+      "learning_rate": 0.00012890992759148734,
+      "loss": 3.409,
+      "step": 5243
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6506776213645935,
+      "learning_rate": 0.00012881925300699853,
+      "loss": 3.3755,
+      "step": 5244
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6123660802841187,
+      "learning_rate": 0.00012872859925488488,
+      "loss": 3.2831,
+      "step": 5245
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5690043568611145,
+      "learning_rate": 0.0001286379663507309,
+      "loss": 3.1536,
+      "step": 5246
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6750291585922241,
+      "learning_rate": 0.00012854735431011758,
+      "loss": 3.3318,
+      "step": 5247
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.647638738155365,
+      "learning_rate": 0.00012845676314862203,
+      "loss": 3.4185,
+      "step": 5248
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6368696093559265,
+      "learning_rate": 0.00012836619288181795,
+      "loss": 3.2747,
+      "step": 5249
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6043297052383423,
+      "learning_rate": 0.00012827564352527543,
+      "loss": 3.2943,
+      "step": 5250
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6385737061500549,
+      "learning_rate": 0.00012818511509456092,
+      "loss": 3.3061,
+      "step": 5251
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6400846838951111,
+      "learning_rate": 0.0001280946076052372,
+      "loss": 3.3456,
+      "step": 5252
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5763105154037476,
+      "learning_rate": 0.00012800412107286384,
+      "loss": 3.3069,
+      "step": 5253
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6130650639533997,
+      "learning_rate": 0.00012791365551299624,
+      "loss": 3.3633,
+      "step": 5254
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6135804057121277,
+      "learning_rate": 0.00012782321094118672,
+      "loss": 3.2839,
+      "step": 5255
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5776688456535339,
+      "learning_rate": 0.0001277327873729836,
+      "loss": 3.242,
+      "step": 5256
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6195380091667175,
+      "learning_rate": 0.0001276423848239318,
+      "loss": 3.2915,
+      "step": 5257
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5916157364845276,
+      "learning_rate": 0.0001275520033095725,
+      "loss": 3.2218,
+      "step": 5258
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6159975528717041,
+      "learning_rate": 0.00012746164284544332,
+      "loss": 3.2407,
+      "step": 5259
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6072033643722534,
+      "learning_rate": 0.00012737130344707843,
+      "loss": 3.4078,
+      "step": 5260
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6531447768211365,
+      "learning_rate": 0.00012728098513000805,
+      "loss": 3.145,
+      "step": 5261
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5923706889152527,
+      "learning_rate": 0.00012719068790975906,
+      "loss": 3.366,
+      "step": 5262
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6577826738357544,
+      "learning_rate": 0.0001271004118018545,
+      "loss": 3.3171,
+      "step": 5263
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6210270524024963,
+      "learning_rate": 0.00012701015682181385,
+      "loss": 3.3118,
+      "step": 5264
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.758233904838562,
+      "learning_rate": 0.00012691992298515317,
+      "loss": 3.3728,
+      "step": 5265
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6270756125450134,
+      "learning_rate": 0.0001268297103073846,
+      "loss": 3.3101,
+      "step": 5266
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6202570199966431,
+      "learning_rate": 0.00012673951880401663,
+      "loss": 3.3395,
+      "step": 5267
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5952986478805542,
+      "learning_rate": 0.00012664934849055442,
+      "loss": 3.36,
+      "step": 5268
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6283809542655945,
+      "learning_rate": 0.00012655919938249922,
+      "loss": 3.1854,
+      "step": 5269
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6460493803024292,
+      "learning_rate": 0.0001264690714953487,
+      "loss": 3.3881,
+      "step": 5270
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6002459526062012,
+      "learning_rate": 0.00012637896484459687,
+      "loss": 3.276,
+      "step": 5271
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6768071055412292,
+      "learning_rate": 0.00012628887944573413,
+      "loss": 3.3012,
+      "step": 5272
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6223396062850952,
+      "learning_rate": 0.00012619881531424713,
+      "loss": 3.2877,
+      "step": 5273
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.73650723695755,
+      "learning_rate": 0.00012610877246561897,
+      "loss": 3.3105,
+      "step": 5274
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6108227968215942,
+      "learning_rate": 0.0001260187509153292,
+      "loss": 3.166,
+      "step": 5275
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6116386651992798,
+      "learning_rate": 0.00012592875067885345,
+      "loss": 3.2562,
+      "step": 5276
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6081870198249817,
+      "learning_rate": 0.0001258387717716638,
+      "loss": 3.2611,
+      "step": 5277
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6102370023727417,
+      "learning_rate": 0.00012574881420922873,
+      "loss": 3.2183,
+      "step": 5278
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5910782217979431,
+      "learning_rate": 0.00012565887800701291,
+      "loss": 3.2597,
+      "step": 5279
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5789437294006348,
+      "learning_rate": 0.00012556896318047733,
+      "loss": 3.3493,
+      "step": 5280
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5884090065956116,
+      "learning_rate": 0.00012547906974507968,
+      "loss": 3.36,
+      "step": 5281
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6605008840560913,
+      "learning_rate": 0.00012538919771627334,
+      "loss": 3.3206,
+      "step": 5282
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6063432097434998,
+      "learning_rate": 0.00012529934710950864,
+      "loss": 3.3791,
+      "step": 5283
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6339197754859924,
+      "learning_rate": 0.00012520951794023184,
+      "loss": 3.3382,
+      "step": 5284
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6664906740188599,
+      "learning_rate": 0.00012511971022388557,
+      "loss": 3.2971,
+      "step": 5285
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6363011002540588,
+      "learning_rate": 0.0001250299239759089,
+      "loss": 3.4032,
+      "step": 5286
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6400793194770813,
+      "learning_rate": 0.00012494015921173704,
+      "loss": 3.2862,
+      "step": 5287
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5983986258506775,
+      "learning_rate": 0.00012485041594680155,
+      "loss": 3.2912,
+      "step": 5288
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5790104866027832,
+      "learning_rate": 0.0001247606941965305,
+      "loss": 3.3903,
+      "step": 5289
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6162522435188293,
+      "learning_rate": 0.00012467099397634802,
+      "loss": 3.2884,
+      "step": 5290
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5953441262245178,
+      "learning_rate": 0.00012458131530167452,
+      "loss": 3.3475,
+      "step": 5291
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6053186058998108,
+      "learning_rate": 0.000124491658187927,
+      "loss": 3.3778,
+      "step": 5292
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6237776875495911,
+      "learning_rate": 0.00012440202265051844,
+      "loss": 3.3701,
+      "step": 5293
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.640582799911499,
+      "learning_rate": 0.00012431240870485824,
+      "loss": 3.3292,
+      "step": 5294
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6169292330741882,
+      "learning_rate": 0.00012422281636635202,
+      "loss": 3.3628,
+      "step": 5295
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6268018484115601,
+      "learning_rate": 0.00012413324565040186,
+      "loss": 3.3225,
+      "step": 5296
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5968641638755798,
+      "learning_rate": 0.00012404369657240596,
+      "loss": 3.3803,
+      "step": 5297
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6453279256820679,
+      "learning_rate": 0.0001239541691477588,
+      "loss": 3.3369,
+      "step": 5298
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6305952668190002,
+      "learning_rate": 0.00012386466339185125,
+      "loss": 3.5307,
+      "step": 5299
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6009957194328308,
+      "learning_rate": 0.00012377517932007033,
+      "loss": 3.3499,
+      "step": 5300
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5801147222518921,
+      "learning_rate": 0.00012368571694779934,
+      "loss": 3.3182,
+      "step": 5301
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6128845810890198,
+      "learning_rate": 0.00012359627629041805,
+      "loss": 3.2508,
+      "step": 5302
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6393544673919678,
+      "learning_rate": 0.00012350685736330216,
+      "loss": 3.3248,
+      "step": 5303
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.632507860660553,
+      "learning_rate": 0.000123417460181824,
+      "loss": 3.2552,
+      "step": 5304
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.64713054895401,
+      "learning_rate": 0.00012332808476135193,
+      "loss": 3.3147,
+      "step": 5305
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6159345507621765,
+      "learning_rate": 0.00012323873111725063,
+      "loss": 3.3826,
+      "step": 5306
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6228779554367065,
+      "learning_rate": 0.00012314939926488095,
+      "loss": 3.3276,
+      "step": 5307
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5892593860626221,
+      "learning_rate": 0.0001230600892196001,
+      "loss": 3.307,
+      "step": 5308
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6291082501411438,
+      "learning_rate": 0.00012297080099676146,
+      "loss": 3.3732,
+      "step": 5309
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5819165110588074,
+      "learning_rate": 0.0001228815346117148,
+      "loss": 3.2115,
+      "step": 5310
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6499049067497253,
+      "learning_rate": 0.00012279229007980605,
+      "loss": 3.416,
+      "step": 5311
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6292195320129395,
+      "learning_rate": 0.0001227030674163774,
+      "loss": 3.3248,
+      "step": 5312
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5752406120300293,
+      "learning_rate": 0.00012261386663676722,
+      "loss": 3.29,
+      "step": 5313
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5899255871772766,
+      "learning_rate": 0.00012252468775631012,
+      "loss": 3.264,
+      "step": 5314
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6782875657081604,
+      "learning_rate": 0.00012243553079033703,
+      "loss": 3.3667,
+      "step": 5315
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6177646517753601,
+      "learning_rate": 0.00012234639575417497,
+      "loss": 3.3603,
+      "step": 5316
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6094318628311157,
+      "learning_rate": 0.00012225728266314746,
+      "loss": 3.3114,
+      "step": 5317
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6286597847938538,
+      "learning_rate": 0.00012216819153257398,
+      "loss": 3.2324,
+      "step": 5318
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6156710982322693,
+      "learning_rate": 0.00012207912237777022,
+      "loss": 3.4317,
+      "step": 5319
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6111778020858765,
+      "learning_rate": 0.0001219900752140484,
+      "loss": 3.205,
+      "step": 5320
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6007053256034851,
+      "learning_rate": 0.0001219010500567167,
+      "loss": 3.3469,
+      "step": 5321
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5899903178215027,
+      "learning_rate": 0.00012181204692107952,
+      "loss": 3.3366,
+      "step": 5322
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.636367678642273,
+      "learning_rate": 0.00012172306582243756,
+      "loss": 3.3244,
+      "step": 5323
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6870309710502625,
+      "learning_rate": 0.0001216341067760876,
+      "loss": 3.2732,
+      "step": 5324
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6357422471046448,
+      "learning_rate": 0.00012154516979732295,
+      "loss": 3.1816,
+      "step": 5325
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6212683916091919,
+      "learning_rate": 0.00012145625490143275,
+      "loss": 3.314,
+      "step": 5326
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5981928110122681,
+      "learning_rate": 0.00012136736210370255,
+      "loss": 3.1853,
+      "step": 5327
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.636627733707428,
+      "learning_rate": 0.00012127849141941396,
+      "loss": 3.3097,
+      "step": 5328
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6140633821487427,
+      "learning_rate": 0.00012118964286384506,
+      "loss": 3.3549,
+      "step": 5329
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5917346477508545,
+      "learning_rate": 0.00012110081645226986,
+      "loss": 3.2242,
+      "step": 5330
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6547762155532837,
+      "learning_rate": 0.00012101201219995853,
+      "loss": 3.232,
+      "step": 5331
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6183657050132751,
+      "learning_rate": 0.00012092323012217774,
+      "loss": 3.3344,
+      "step": 5332
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6329478621482849,
+      "learning_rate": 0.00012083447023419009,
+      "loss": 3.3021,
+      "step": 5333
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6323261260986328,
+      "learning_rate": 0.00012074573255125442,
+      "loss": 3.2491,
+      "step": 5334
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6243185997009277,
+      "learning_rate": 0.00012065701708862578,
+      "loss": 3.3159,
+      "step": 5335
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6621860861778259,
+      "learning_rate": 0.00012056832386155536,
+      "loss": 3.2703,
+      "step": 5336
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6325708031654358,
+      "learning_rate": 0.0001204796528852905,
+      "loss": 3.2953,
+      "step": 5337
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6371250152587891,
+      "learning_rate": 0.0001203910041750749,
+      "loss": 3.4275,
+      "step": 5338
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6301656365394592,
+      "learning_rate": 0.00012030237774614816,
+      "loss": 3.4774,
+      "step": 5339
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6328499913215637,
+      "learning_rate": 0.00012021377361374636,
+      "loss": 3.2954,
+      "step": 5340
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6077415347099304,
+      "learning_rate": 0.0001201251917931015,
+      "loss": 3.3703,
+      "step": 5341
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6206401586532593,
+      "learning_rate": 0.00012003663229944178,
+      "loss": 3.3644,
+      "step": 5342
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6526326537132263,
+      "learning_rate": 0.00011994809514799166,
+      "loss": 3.3512,
+      "step": 5343
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6655504107475281,
+      "learning_rate": 0.00011985958035397169,
+      "loss": 3.2669,
+      "step": 5344
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5940502882003784,
+      "learning_rate": 0.00011977108793259845,
+      "loss": 3.2079,
+      "step": 5345
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5870606899261475,
+      "learning_rate": 0.00011968261789908502,
+      "loss": 3.2211,
+      "step": 5346
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6249194145202637,
+      "learning_rate": 0.00011959417026864025,
+      "loss": 3.3497,
+      "step": 5347
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6449898481369019,
+      "learning_rate": 0.00011950574505646952,
+      "loss": 3.4259,
+      "step": 5348
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.659328281879425,
+      "learning_rate": 0.00011941734227777403,
+      "loss": 3.3543,
+      "step": 5349
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6789844632148743,
+      "learning_rate": 0.00011932896194775125,
+      "loss": 3.2992,
+      "step": 5350
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6676364541053772,
+      "learning_rate": 0.00011924060408159477,
+      "loss": 3.448,
+      "step": 5351
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6708555221557617,
+      "learning_rate": 0.00011915226869449425,
+      "loss": 3.3006,
+      "step": 5352
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6383585929870605,
+      "learning_rate": 0.00011906395580163576,
+      "loss": 3.2522,
+      "step": 5353
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6652035713195801,
+      "learning_rate": 0.0001189756654182012,
+      "loss": 3.3669,
+      "step": 5354
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6032835841178894,
+      "learning_rate": 0.00011888739755936873,
+      "loss": 3.4457,
+      "step": 5355
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6228095293045044,
+      "learning_rate": 0.00011879915224031249,
+      "loss": 3.4268,
+      "step": 5356
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6080935001373291,
+      "learning_rate": 0.0001187109294762031,
+      "loss": 3.2699,
+      "step": 5357
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6331431865692139,
+      "learning_rate": 0.00011862272928220696,
+      "loss": 3.3719,
+      "step": 5358
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6054500341415405,
+      "learning_rate": 0.00011853455167348673,
+      "loss": 3.1855,
+      "step": 5359
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6649967432022095,
+      "learning_rate": 0.00011844639666520105,
+      "loss": 3.3242,
+      "step": 5360
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6024637818336487,
+      "learning_rate": 0.00011835826427250496,
+      "loss": 3.3709,
+      "step": 5361
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.654136061668396,
+      "learning_rate": 0.00011827015451054937,
+      "loss": 3.2647,
+      "step": 5362
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6690921783447266,
+      "learning_rate": 0.00011818206739448137,
+      "loss": 3.3242,
+      "step": 5363
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6012855768203735,
+      "learning_rate": 0.00011809400293944414,
+      "loss": 3.237,
+      "step": 5364
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6119986772537231,
+      "learning_rate": 0.00011800596116057688,
+      "loss": 3.2263,
+      "step": 5365
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6060097217559814,
+      "learning_rate": 0.00011791794207301524,
+      "loss": 3.375,
+      "step": 5366
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6539217233657837,
+      "learning_rate": 0.00011782994569189045,
+      "loss": 3.2729,
+      "step": 5367
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6135516166687012,
+      "learning_rate": 0.00011774197203233037,
+      "loss": 3.3101,
+      "step": 5368
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6137378811836243,
+      "learning_rate": 0.00011765402110945852,
+      "loss": 3.2379,
+      "step": 5369
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6105403900146484,
+      "learning_rate": 0.00011756609293839477,
+      "loss": 3.2976,
+      "step": 5370
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6659379005432129,
+      "learning_rate": 0.00011747818753425493,
+      "loss": 3.2012,
+      "step": 5371
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6903918385505676,
+      "learning_rate": 0.00011739030491215097,
+      "loss": 3.294,
+      "step": 5372
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6265830397605896,
+      "learning_rate": 0.00011730244508719087,
+      "loss": 3.3366,
+      "step": 5373
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5908456444740295,
+      "learning_rate": 0.00011721460807447889,
+      "loss": 3.2259,
+      "step": 5374
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6120139956474304,
+      "learning_rate": 0.0001171267938891151,
+      "loss": 3.2589,
+      "step": 5375
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5990294218063354,
+      "learning_rate": 0.0001170390025461959,
+      "loss": 3.299,
+      "step": 5376
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6606252193450928,
+      "learning_rate": 0.00011695123406081359,
+      "loss": 3.3959,
+      "step": 5377
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.689970850944519,
+      "learning_rate": 0.00011686348844805659,
+      "loss": 3.3365,
+      "step": 5378
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7213213443756104,
+      "learning_rate": 0.00011677576572300935,
+      "loss": 3.3814,
+      "step": 5379
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6567277908325195,
+      "learning_rate": 0.00011668806590075248,
+      "loss": 3.3028,
+      "step": 5380
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.636063277721405,
+      "learning_rate": 0.00011660038899636247,
+      "loss": 3.2835,
+      "step": 5381
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6282566785812378,
+      "learning_rate": 0.00011651273502491216,
+      "loss": 3.3486,
+      "step": 5382
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6272147297859192,
+      "learning_rate": 0.00011642510400147025,
+      "loss": 3.257,
+      "step": 5383
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5836976766586304,
+      "learning_rate": 0.00011633749594110139,
+      "loss": 3.2378,
+      "step": 5384
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6571261286735535,
+      "learning_rate": 0.00011624991085886661,
+      "loss": 3.288,
+      "step": 5385
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6557903289794922,
+      "learning_rate": 0.0001161623487698227,
+      "loss": 3.2989,
+      "step": 5386
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6257616877555847,
+      "learning_rate": 0.00011607480968902264,
+      "loss": 3.2491,
+      "step": 5387
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6553791761398315,
+      "learning_rate": 0.0001159872936315153,
+      "loss": 3.3515,
+      "step": 5388
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.631804883480072,
+      "learning_rate": 0.00011589980061234587,
+      "loss": 3.1635,
+      "step": 5389
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6316229701042175,
+      "learning_rate": 0.00011581233064655536,
+      "loss": 3.347,
+      "step": 5390
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6077477335929871,
+      "learning_rate": 0.00011572488374918083,
+      "loss": 3.2972,
+      "step": 5391
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6585051417350769,
+      "learning_rate": 0.00011563745993525543,
+      "loss": 3.3662,
+      "step": 5392
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6453659534454346,
+      "learning_rate": 0.00011555005921980825,
+      "loss": 3.3,
+      "step": 5393
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6500974297523499,
+      "learning_rate": 0.00011546268161786466,
+      "loss": 3.2243,
+      "step": 5394
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6436439156532288,
+      "learning_rate": 0.0001153753271444458,
+      "loss": 3.3304,
+      "step": 5395
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6569267511367798,
+      "learning_rate": 0.00011528799581456878,
+      "loss": 3.2496,
+      "step": 5396
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6070961356163025,
+      "learning_rate": 0.00011520068764324712,
+      "loss": 3.4124,
+      "step": 5397
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.597518801689148,
+      "learning_rate": 0.00011511340264548997,
+      "loss": 3.2643,
+      "step": 5398
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6514595150947571,
+      "learning_rate": 0.00011502614083630264,
+      "loss": 3.4758,
+      "step": 5399
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.615273654460907,
+      "learning_rate": 0.00011493890223068646,
+      "loss": 3.4179,
+      "step": 5400
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6327627301216125,
+      "learning_rate": 0.00011485168684363876,
+      "loss": 3.3016,
+      "step": 5401
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6123901009559631,
+      "learning_rate": 0.00011476449469015276,
+      "loss": 3.3365,
+      "step": 5402
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6408461332321167,
+      "learning_rate": 0.0001146773257852179,
+      "loss": 3.2455,
+      "step": 5403
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.606148362159729,
+      "learning_rate": 0.00011459018014381963,
+      "loss": 3.3178,
+      "step": 5404
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6506169438362122,
+      "learning_rate": 0.0001145030577809392,
+      "loss": 3.4223,
+      "step": 5405
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6217191219329834,
+      "learning_rate": 0.00011441595871155397,
+      "loss": 3.3234,
+      "step": 5406
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6534956693649292,
+      "learning_rate": 0.00011432888295063723,
+      "loss": 3.2476,
+      "step": 5407
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6531282663345337,
+      "learning_rate": 0.00011424183051315837,
+      "loss": 3.2446,
+      "step": 5408
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6264949440956116,
+      "learning_rate": 0.00011415480141408258,
+      "loss": 3.3105,
+      "step": 5409
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6574244499206543,
+      "learning_rate": 0.00011406779566837139,
+      "loss": 3.3587,
+      "step": 5410
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6147937178611755,
+      "learning_rate": 0.00011398081329098198,
+      "loss": 3.4396,
+      "step": 5411
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6084353923797607,
+      "learning_rate": 0.00011389385429686752,
+      "loss": 3.2026,
+      "step": 5412
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.617586612701416,
+      "learning_rate": 0.0001138069187009775,
+      "loss": 3.341,
+      "step": 5413
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6375697255134583,
+      "learning_rate": 0.00011372000651825703,
+      "loss": 3.3281,
+      "step": 5414
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6307726502418518,
+      "learning_rate": 0.00011363311776364735,
+      "loss": 3.3064,
+      "step": 5415
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5901523232460022,
+      "learning_rate": 0.0001135462524520856,
+      "loss": 3.2324,
+      "step": 5416
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6218355298042297,
+      "learning_rate": 0.0001134594105985049,
+      "loss": 3.2884,
+      "step": 5417
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6189782023429871,
+      "learning_rate": 0.00011337259221783453,
+      "loss": 3.3595,
+      "step": 5418
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6642634272575378,
+      "learning_rate": 0.00011328579732499944,
+      "loss": 3.3093,
+      "step": 5419
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6610681414604187,
+      "learning_rate": 0.00011319902593492074,
+      "loss": 3.3262,
+      "step": 5420
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6154919266700745,
+      "learning_rate": 0.00011311227806251531,
+      "loss": 3.2144,
+      "step": 5421
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6145983338356018,
+      "learning_rate": 0.00011302555372269633,
+      "loss": 3.3124,
+      "step": 5422
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6004980206489563,
+      "learning_rate": 0.00011293885293037259,
+      "loss": 3.4122,
+      "step": 5423
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6144428253173828,
+      "learning_rate": 0.0001128521757004489,
+      "loss": 3.1873,
+      "step": 5424
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.652301549911499,
+      "learning_rate": 0.00011276552204782625,
+      "loss": 3.4181,
+      "step": 5425
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5871480107307434,
+      "learning_rate": 0.00011267889198740131,
+      "loss": 3.2486,
+      "step": 5426
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5704430341720581,
+      "learning_rate": 0.0001125922855340668,
+      "loss": 3.1964,
+      "step": 5427
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6235558390617371,
+      "learning_rate": 0.0001125057027027114,
+      "loss": 3.0768,
+      "step": 5428
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6705940365791321,
+      "learning_rate": 0.00011241914350821967,
+      "loss": 3.4425,
+      "step": 5429
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6448222994804382,
+      "learning_rate": 0.00011233260796547201,
+      "loss": 3.3121,
+      "step": 5430
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6028103232383728,
+      "learning_rate": 0.00011224609608934505,
+      "loss": 3.3336,
+      "step": 5431
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6391410231590271,
+      "learning_rate": 0.00011215960789471125,
+      "loss": 3.2327,
+      "step": 5432
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.673503577709198,
+      "learning_rate": 0.00011207314339643884,
+      "loss": 3.4889,
+      "step": 5433
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5765765309333801,
+      "learning_rate": 0.00011198670260939206,
+      "loss": 3.2223,
+      "step": 5434
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6331235766410828,
+      "learning_rate": 0.00011190028554843107,
+      "loss": 3.2458,
+      "step": 5435
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6514334082603455,
+      "learning_rate": 0.00011181389222841201,
+      "loss": 3.3787,
+      "step": 5436
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6331505179405212,
+      "learning_rate": 0.00011172752266418684,
+      "loss": 3.3041,
+      "step": 5437
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6335639953613281,
+      "learning_rate": 0.00011164117687060346,
+      "loss": 3.3845,
+      "step": 5438
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6040472388267517,
+      "learning_rate": 0.00011155485486250574,
+      "loss": 3.2573,
+      "step": 5439
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.625055193901062,
+      "learning_rate": 0.00011146855665473355,
+      "loss": 3.3227,
+      "step": 5440
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6047588586807251,
+      "learning_rate": 0.00011138228226212249,
+      "loss": 3.3813,
+      "step": 5441
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6100949645042419,
+      "learning_rate": 0.00011129603169950409,
+      "loss": 3.4167,
+      "step": 5442
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6181287169456482,
+      "learning_rate": 0.00011120980498170583,
+      "loss": 3.344,
+      "step": 5443
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6023228764533997,
+      "learning_rate": 0.0001111236021235511,
+      "loss": 3.3657,
+      "step": 5444
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6248236894607544,
+      "learning_rate": 0.00011103742313985906,
+      "loss": 3.295,
+      "step": 5445
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5963073372840881,
+      "learning_rate": 0.00011095126804544505,
+      "loss": 3.1363,
+      "step": 5446
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6147260665893555,
+      "learning_rate": 0.0001108651368551201,
+      "loss": 3.4925,
+      "step": 5447
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.630146861076355,
+      "learning_rate": 0.00011077902958369106,
+      "loss": 3.2746,
+      "step": 5448
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6169753074645996,
+      "learning_rate": 0.00011069294624596077,
+      "loss": 3.1804,
+      "step": 5449
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6223546266555786,
+      "learning_rate": 0.0001106068868567281,
+      "loss": 3.2929,
+      "step": 5450
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6127028465270996,
+      "learning_rate": 0.00011052085143078752,
+      "loss": 3.2841,
+      "step": 5451
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6518468260765076,
+      "learning_rate": 0.00011043483998292949,
+      "loss": 3.2391,
+      "step": 5452
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.623802900314331,
+      "learning_rate": 0.00011034885252794056,
+      "loss": 3.2241,
+      "step": 5453
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5940724611282349,
+      "learning_rate": 0.00011026288908060284,
+      "loss": 3.1555,
+      "step": 5454
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6354788541793823,
+      "learning_rate": 0.00011017694965569447,
+      "loss": 3.3109,
+      "step": 5455
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5967780351638794,
+      "learning_rate": 0.00011009103426798939,
+      "loss": 3.3411,
+      "step": 5456
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6164358258247375,
+      "learning_rate": 0.0001100051429322575,
+      "loss": 3.33,
+      "step": 5457
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6327795386314392,
+      "learning_rate": 0.00010991927566326443,
+      "loss": 3.2257,
+      "step": 5458
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.637921154499054,
+      "learning_rate": 0.00010983343247577187,
+      "loss": 3.2734,
+      "step": 5459
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.61724853515625,
+      "learning_rate": 0.00010974761338453718,
+      "loss": 3.1948,
+      "step": 5460
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6523029208183289,
+      "learning_rate": 0.00010966181840431375,
+      "loss": 3.3381,
+      "step": 5461
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6485067009925842,
+      "learning_rate": 0.0001095760475498507,
+      "loss": 3.356,
+      "step": 5462
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6275986433029175,
+      "learning_rate": 0.000109490300835893,
+      "loss": 3.297,
+      "step": 5463
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6262179017066956,
+      "learning_rate": 0.00010940457827718151,
+      "loss": 3.2337,
+      "step": 5464
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6072761416435242,
+      "learning_rate": 0.00010931887988845294,
+      "loss": 3.3163,
+      "step": 5465
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6281195878982544,
+      "learning_rate": 0.00010923320568443972,
+      "loss": 3.3204,
+      "step": 5466
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6109669208526611,
+      "learning_rate": 0.0001091475556798704,
+      "loss": 3.3305,
+      "step": 5467
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.660270094871521,
+      "learning_rate": 0.00010906192988946922,
+      "loss": 3.2994,
+      "step": 5468
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6059504151344299,
+      "learning_rate": 0.00010897632832795618,
+      "loss": 3.2588,
+      "step": 5469
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6001191735267639,
+      "learning_rate": 0.0001088907510100472,
+      "loss": 3.3192,
+      "step": 5470
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5829064846038818,
+      "learning_rate": 0.00010880519795045399,
+      "loss": 3.2155,
+      "step": 5471
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5633309483528137,
+      "learning_rate": 0.00010871966916388415,
+      "loss": 3.308,
+      "step": 5472
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5973705649375916,
+      "learning_rate": 0.00010863416466504092,
+      "loss": 3.3308,
+      "step": 5473
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6159566640853882,
+      "learning_rate": 0.00010854868446862373,
+      "loss": 3.4214,
+      "step": 5474
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.654606282711029,
+      "learning_rate": 0.00010846322858932756,
+      "loss": 3.2694,
+      "step": 5475
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6203795671463013,
+      "learning_rate": 0.00010837779704184311,
+      "loss": 3.3132,
+      "step": 5476
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6621377468109131,
+      "learning_rate": 0.00010829238984085727,
+      "loss": 3.3296,
+      "step": 5477
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6005878448486328,
+      "learning_rate": 0.00010820700700105244,
+      "loss": 3.2034,
+      "step": 5478
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6048322916030884,
+      "learning_rate": 0.00010812164853710687,
+      "loss": 3.2214,
+      "step": 5479
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6573078632354736,
+      "learning_rate": 0.00010803631446369477,
+      "loss": 3.3038,
+      "step": 5480
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6411568522453308,
+      "learning_rate": 0.00010795100479548586,
+      "loss": 3.3197,
+      "step": 5481
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6481874585151672,
+      "learning_rate": 0.00010786571954714613,
+      "loss": 3.2614,
+      "step": 5482
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6276066303253174,
+      "learning_rate": 0.00010778045873333695,
+      "loss": 3.2463,
+      "step": 5483
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6247035264968872,
+      "learning_rate": 0.00010769522236871568,
+      "loss": 3.2256,
+      "step": 5484
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6492263078689575,
+      "learning_rate": 0.00010761001046793539,
+      "loss": 3.3523,
+      "step": 5485
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6784762144088745,
+      "learning_rate": 0.00010752482304564495,
+      "loss": 3.3337,
+      "step": 5486
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6118037104606628,
+      "learning_rate": 0.00010743966011648926,
+      "loss": 3.2237,
+      "step": 5487
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.637303352355957,
+      "learning_rate": 0.0001073545216951086,
+      "loss": 3.2963,
+      "step": 5488
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6309385299682617,
+      "learning_rate": 0.00010726940779613942,
+      "loss": 3.2413,
+      "step": 5489
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5993750691413879,
+      "learning_rate": 0.0001071843184342137,
+      "loss": 3.281,
+      "step": 5490
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6420213580131531,
+      "learning_rate": 0.00010709925362395933,
+      "loss": 3.1684,
+      "step": 5491
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6415120959281921,
+      "learning_rate": 0.0001070142133799999,
+      "loss": 3.3344,
+      "step": 5492
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6192628145217896,
+      "learning_rate": 0.00010692919771695483,
+      "loss": 3.2524,
+      "step": 5493
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6713070273399353,
+      "learning_rate": 0.0001068442066494392,
+      "loss": 3.3718,
+      "step": 5494
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6320802569389343,
+      "learning_rate": 0.00010675924019206415,
+      "loss": 3.3318,
+      "step": 5495
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6208996772766113,
+      "learning_rate": 0.00010667429835943617,
+      "loss": 3.4287,
+      "step": 5496
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.601486086845398,
+      "learning_rate": 0.00010658938116615802,
+      "loss": 3.3564,
+      "step": 5497
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6085739135742188,
+      "learning_rate": 0.00010650448862682777,
+      "loss": 3.4097,
+      "step": 5498
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6288726329803467,
+      "learning_rate": 0.00010641962075603948,
+      "loss": 3.2182,
+      "step": 5499
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6331077814102173,
+      "learning_rate": 0.00010633477756838292,
+      "loss": 3.2166,
+      "step": 5500
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6663286089897156,
+      "learning_rate": 0.0001062499590784436,
+      "loss": 3.3958,
+      "step": 5501
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6179254651069641,
+      "learning_rate": 0.00010616516530080269,
+      "loss": 3.3268,
+      "step": 5502
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5864897966384888,
+      "learning_rate": 0.00010608039625003746,
+      "loss": 3.2097,
+      "step": 5503
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.670877993106842,
+      "learning_rate": 0.00010599565194072047,
+      "loss": 3.3851,
+      "step": 5504
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6081317663192749,
+      "learning_rate": 0.00010591093238742047,
+      "loss": 3.2122,
+      "step": 5505
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.627242922782898,
+      "learning_rate": 0.00010582623760470159,
+      "loss": 3.3746,
+      "step": 5506
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6064150333404541,
+      "learning_rate": 0.00010574156760712389,
+      "loss": 3.2696,
+      "step": 5507
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6662350296974182,
+      "learning_rate": 0.00010565692240924307,
+      "loss": 3.3341,
+      "step": 5508
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.606359601020813,
+      "learning_rate": 0.0001055723020256106,
+      "loss": 3.2804,
+      "step": 5509
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5908784866333008,
+      "learning_rate": 0.00010548770647077385,
+      "loss": 3.2003,
+      "step": 5510
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5890214443206787,
+      "learning_rate": 0.00010540313575927568,
+      "loss": 3.2922,
+      "step": 5511
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6119219660758972,
+      "learning_rate": 0.00010531858990565477,
+      "loss": 3.2116,
+      "step": 5512
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6429023146629333,
+      "learning_rate": 0.00010523406892444549,
+      "loss": 3.3435,
+      "step": 5513
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6105777621269226,
+      "learning_rate": 0.00010514957283017809,
+      "loss": 3.1645,
+      "step": 5514
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5838789939880371,
+      "learning_rate": 0.0001050651016373784,
+      "loss": 3.2575,
+      "step": 5515
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5827857255935669,
+      "learning_rate": 0.00010498065536056794,
+      "loss": 3.4108,
+      "step": 5516
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6028188467025757,
+      "learning_rate": 0.00010489623401426396,
+      "loss": 3.2953,
+      "step": 5517
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.616176426410675,
+      "learning_rate": 0.00010481183761297961,
+      "loss": 3.3607,
+      "step": 5518
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.652996301651001,
+      "learning_rate": 0.00010472746617122356,
+      "loss": 3.2498,
+      "step": 5519
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6454377174377441,
+      "learning_rate": 0.00010464311970350021,
+      "loss": 3.087,
+      "step": 5520
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6470353603363037,
+      "learning_rate": 0.00010455879822430969,
+      "loss": 3.2994,
+      "step": 5521
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6124841570854187,
+      "learning_rate": 0.00010447450174814787,
+      "loss": 3.3696,
+      "step": 5522
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6404425501823425,
+      "learning_rate": 0.0001043902302895062,
+      "loss": 3.2451,
+      "step": 5523
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6828598976135254,
+      "learning_rate": 0.00010430598386287199,
+      "loss": 3.1555,
+      "step": 5524
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6182655692100525,
+      "learning_rate": 0.00010422176248272825,
+      "loss": 3.2486,
+      "step": 5525
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6098328828811646,
+      "learning_rate": 0.00010413756616355358,
+      "loss": 3.3218,
+      "step": 5526
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6310797333717346,
+      "learning_rate": 0.00010405339491982224,
+      "loss": 3.2505,
+      "step": 5527
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6132204532623291,
+      "learning_rate": 0.00010396924876600428,
+      "loss": 3.44,
+      "step": 5528
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6618490219116211,
+      "learning_rate": 0.00010388512771656539,
+      "loss": 3.2299,
+      "step": 5529
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6412935853004456,
+      "learning_rate": 0.00010380103178596686,
+      "loss": 3.3538,
+      "step": 5530
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6475511193275452,
+      "learning_rate": 0.00010371696098866596,
+      "loss": 3.3645,
+      "step": 5531
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6022229790687561,
+      "learning_rate": 0.00010363291533911523,
+      "loss": 3.2734,
+      "step": 5532
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6337286829948425,
+      "learning_rate": 0.00010354889485176328,
+      "loss": 3.2877,
+      "step": 5533
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6246282458305359,
+      "learning_rate": 0.0001034648995410541,
+      "loss": 3.2926,
+      "step": 5534
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6323986053466797,
+      "learning_rate": 0.00010338092942142746,
+      "loss": 3.4426,
+      "step": 5535
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6178189516067505,
+      "learning_rate": 0.0001032969845073188,
+      "loss": 3.2257,
+      "step": 5536
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.653200626373291,
+      "learning_rate": 0.00010321306481315926,
+      "loss": 3.3105,
+      "step": 5537
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6622167229652405,
+      "learning_rate": 0.00010312917035337546,
+      "loss": 3.2909,
+      "step": 5538
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5952175259590149,
+      "learning_rate": 0.00010304530114239008,
+      "loss": 3.2572,
+      "step": 5539
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5932444334030151,
+      "learning_rate": 0.00010296145719462105,
+      "loss": 3.2586,
+      "step": 5540
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6411198973655701,
+      "learning_rate": 0.00010287763852448207,
+      "loss": 3.2888,
+      "step": 5541
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6064850687980652,
+      "learning_rate": 0.00010279384514638268,
+      "loss": 3.1833,
+      "step": 5542
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6550402641296387,
+      "learning_rate": 0.00010271007707472788,
+      "loss": 3.3338,
+      "step": 5543
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6096448302268982,
+      "learning_rate": 0.00010262633432391838,
+      "loss": 3.2992,
+      "step": 5544
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6269118189811707,
+      "learning_rate": 0.0001025426169083504,
+      "loss": 3.2611,
+      "step": 5545
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6356010437011719,
+      "learning_rate": 0.00010245892484241615,
+      "loss": 3.2572,
+      "step": 5546
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6577345132827759,
+      "learning_rate": 0.00010237525814050316,
+      "loss": 3.2427,
+      "step": 5547
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6274149417877197,
+      "learning_rate": 0.0001022916168169947,
+      "loss": 3.2137,
+      "step": 5548
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6737575531005859,
+      "learning_rate": 0.00010220800088626969,
+      "loss": 3.2899,
+      "step": 5549
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6428675651550293,
+      "learning_rate": 0.00010212441036270271,
+      "loss": 3.3025,
+      "step": 5550
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6240596175193787,
+      "learning_rate": 0.0001020408452606638,
+      "loss": 3.2887,
+      "step": 5551
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6059290766716003,
+      "learning_rate": 0.00010195730559451893,
+      "loss": 3.2722,
+      "step": 5552
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5900526642799377,
+      "learning_rate": 0.00010187379137862945,
+      "loss": 3.2072,
+      "step": 5553
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5984786152839661,
+      "learning_rate": 0.00010179030262735254,
+      "loss": 3.284,
+      "step": 5554
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6868179440498352,
+      "learning_rate": 0.00010170683935504077,
+      "loss": 3.3633,
+      "step": 5555
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6394917964935303,
+      "learning_rate": 0.00010162340157604252,
+      "loss": 3.2281,
+      "step": 5556
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6540482640266418,
+      "learning_rate": 0.00010153998930470165,
+      "loss": 3.1823,
+      "step": 5557
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6355370879173279,
+      "learning_rate": 0.00010145660255535771,
+      "loss": 3.4014,
+      "step": 5558
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6422974467277527,
+      "learning_rate": 0.00010137324134234577,
+      "loss": 3.3509,
+      "step": 5559
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6201578974723816,
+      "learning_rate": 0.00010128990567999666,
+      "loss": 3.3104,
+      "step": 5560
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6185092926025391,
+      "learning_rate": 0.00010120659558263687,
+      "loss": 3.3656,
+      "step": 5561
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6379492878913879,
+      "learning_rate": 0.00010112331106458825,
+      "loss": 3.3114,
+      "step": 5562
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6754159331321716,
+      "learning_rate": 0.00010104005214016837,
+      "loss": 3.3098,
+      "step": 5563
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6179805994033813,
+      "learning_rate": 0.00010095681882369042,
+      "loss": 3.4452,
+      "step": 5564
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6396198272705078,
+      "learning_rate": 0.00010087361112946319,
+      "loss": 3.3147,
+      "step": 5565
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6521846652030945,
+      "learning_rate": 0.00010079042907179092,
+      "loss": 3.2994,
+      "step": 5566
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6344322562217712,
+      "learning_rate": 0.0001007072726649738,
+      "loss": 3.3356,
+      "step": 5567
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6148968935012817,
+      "learning_rate": 0.00010062414192330724,
+      "loss": 3.3081,
+      "step": 5568
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6300048828125,
+      "learning_rate": 0.00010054103686108229,
+      "loss": 3.3595,
+      "step": 5569
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6113799214363098,
+      "learning_rate": 0.0001004579574925859,
+      "loss": 3.2506,
+      "step": 5570
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6313521265983582,
+      "learning_rate": 0.00010037490383210024,
+      "loss": 3.3954,
+      "step": 5571
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6265667080879211,
+      "learning_rate": 0.0001002918758939032,
+      "loss": 3.4398,
+      "step": 5572
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.67408686876297,
+      "learning_rate": 0.0001002088736922683,
+      "loss": 3.3008,
+      "step": 5573
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6352943778038025,
+      "learning_rate": 0.00010012589724146443,
+      "loss": 3.3582,
+      "step": 5574
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5675966739654541,
+      "learning_rate": 0.00010004294655575639,
+      "loss": 3.2766,
+      "step": 5575
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.695729672908783,
+      "learning_rate": 9.996002164940429e-05,
+      "loss": 3.3943,
+      "step": 5576
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5746981501579285,
+      "learning_rate": 9.98771225366639e-05,
+      "loss": 3.3425,
+      "step": 5577
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5896180272102356,
+      "learning_rate": 9.979424923178643e-05,
+      "loss": 3.2756,
+      "step": 5578
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6412181854248047,
+      "learning_rate": 9.971140174901891e-05,
+      "loss": 3.267,
+      "step": 5579
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5925517082214355,
+      "learning_rate": 9.962858010260376e-05,
+      "loss": 3.2467,
+      "step": 5580
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6303708553314209,
+      "learning_rate": 9.954578430677882e-05,
+      "loss": 3.3277,
+      "step": 5581
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6667519807815552,
+      "learning_rate": 9.94630143757779e-05,
+      "loss": 3.4194,
+      "step": 5582
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6062800288200378,
+      "learning_rate": 9.938027032382996e-05,
+      "loss": 3.3892,
+      "step": 5583
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6656795144081116,
+      "learning_rate": 9.92975521651597e-05,
+      "loss": 3.3456,
+      "step": 5584
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6574938297271729,
+      "learning_rate": 9.92148599139873e-05,
+      "loss": 3.3215,
+      "step": 5585
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6665852069854736,
+      "learning_rate": 9.913219358452855e-05,
+      "loss": 3.3117,
+      "step": 5586
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.640589714050293,
+      "learning_rate": 9.904955319099462e-05,
+      "loss": 3.1973,
+      "step": 5587
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.615400493144989,
+      "learning_rate": 9.896693874759257e-05,
+      "loss": 3.3768,
+      "step": 5588
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6466597318649292,
+      "learning_rate": 9.888435026852458e-05,
+      "loss": 3.2447,
+      "step": 5589
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6233015060424805,
+      "learning_rate": 9.880178776798876e-05,
+      "loss": 3.305,
+      "step": 5590
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.658562183380127,
+      "learning_rate": 9.871925126017845e-05,
+      "loss": 3.1554,
+      "step": 5591
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6282123327255249,
+      "learning_rate": 9.863674075928267e-05,
+      "loss": 3.1391,
+      "step": 5592
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6806384921073914,
+      "learning_rate": 9.855425627948587e-05,
+      "loss": 3.422,
+      "step": 5593
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6451425552368164,
+      "learning_rate": 9.847179783496815e-05,
+      "loss": 3.2525,
+      "step": 5594
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6309500932693481,
+      "learning_rate": 9.838936543990495e-05,
+      "loss": 3.3627,
+      "step": 5595
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.641991138458252,
+      "learning_rate": 9.830695910846754e-05,
+      "loss": 3.2728,
+      "step": 5596
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6529327630996704,
+      "learning_rate": 9.822457885482237e-05,
+      "loss": 3.306,
+      "step": 5597
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6511504650115967,
+      "learning_rate": 9.814222469313166e-05,
+      "loss": 3.2857,
+      "step": 5598
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5973052978515625,
+      "learning_rate": 9.805989663755308e-05,
+      "loss": 3.3806,
+      "step": 5599
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6324095726013184,
+      "learning_rate": 9.797759470223966e-05,
+      "loss": 3.3219,
+      "step": 5600
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6935181021690369,
+      "learning_rate": 9.789531890134012e-05,
+      "loss": 3.3186,
+      "step": 5601
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6352891325950623,
+      "learning_rate": 9.781306924899852e-05,
+      "loss": 3.3261,
+      "step": 5602
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6081901788711548,
+      "learning_rate": 9.773084575935471e-05,
+      "loss": 3.2191,
+      "step": 5603
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6001847386360168,
+      "learning_rate": 9.764864844654379e-05,
+      "loss": 3.4207,
+      "step": 5604
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5756590962409973,
+      "learning_rate": 9.756647732469636e-05,
+      "loss": 3.1644,
+      "step": 5605
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6494952440261841,
+      "learning_rate": 9.748433240793858e-05,
+      "loss": 3.3798,
+      "step": 5606
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.597460150718689,
+      "learning_rate": 9.740221371039226e-05,
+      "loss": 3.2834,
+      "step": 5607
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.64914470911026,
+      "learning_rate": 9.732012124617449e-05,
+      "loss": 3.3143,
+      "step": 5608
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6491999626159668,
+      "learning_rate": 9.723805502939786e-05,
+      "loss": 3.3448,
+      "step": 5609
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6296149492263794,
+      "learning_rate": 9.715601507417046e-05,
+      "loss": 3.3611,
+      "step": 5610
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5787907838821411,
+      "learning_rate": 9.70740013945961e-05,
+      "loss": 3.154,
+      "step": 5611
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6179164052009583,
+      "learning_rate": 9.699201400477372e-05,
+      "loss": 3.2012,
+      "step": 5612
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6448367834091187,
+      "learning_rate": 9.691005291879801e-05,
+      "loss": 3.2615,
+      "step": 5613
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6444175839424133,
+      "learning_rate": 9.682811815075895e-05,
+      "loss": 3.1871,
+      "step": 5614
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6346524357795715,
+      "learning_rate": 9.674620971474202e-05,
+      "loss": 3.1654,
+      "step": 5615
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6192304491996765,
+      "learning_rate": 9.666432762482838e-05,
+      "loss": 3.2435,
+      "step": 5616
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6740929484367371,
+      "learning_rate": 9.658247189509436e-05,
+      "loss": 3.2574,
+      "step": 5617
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6283291578292847,
+      "learning_rate": 9.650064253961208e-05,
+      "loss": 3.4429,
+      "step": 5618
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6806941032409668,
+      "learning_rate": 9.641883957244887e-05,
+      "loss": 3.1937,
+      "step": 5619
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6300070881843567,
+      "learning_rate": 9.633706300766759e-05,
+      "loss": 3.4148,
+      "step": 5620
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.68531334400177,
+      "learning_rate": 9.62553128593266e-05,
+      "loss": 3.3577,
+      "step": 5621
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6283920407295227,
+      "learning_rate": 9.617358914147969e-05,
+      "loss": 3.3075,
+      "step": 5622
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6550459265708923,
+      "learning_rate": 9.609189186817604e-05,
+      "loss": 3.2821,
+      "step": 5623
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.596026599407196,
+      "learning_rate": 9.60102210534605e-05,
+      "loss": 3.4144,
+      "step": 5624
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6399941444396973,
+      "learning_rate": 9.59285767113731e-05,
+      "loss": 3.4119,
+      "step": 5625
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6629415154457092,
+      "learning_rate": 9.584695885594957e-05,
+      "loss": 3.3527,
+      "step": 5626
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5877668261528015,
+      "learning_rate": 9.576536750122094e-05,
+      "loss": 3.2068,
+      "step": 5627
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6166104674339294,
+      "learning_rate": 9.568380266121366e-05,
+      "loss": 3.196,
+      "step": 5628
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6430399417877197,
+      "learning_rate": 9.56022643499497e-05,
+      "loss": 3.2587,
+      "step": 5629
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6009620428085327,
+      "learning_rate": 9.55207525814464e-05,
+      "loss": 3.3185,
+      "step": 5630
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5799506902694702,
+      "learning_rate": 9.543926736971656e-05,
+      "loss": 3.2342,
+      "step": 5631
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6568311452865601,
+      "learning_rate": 9.535780872876857e-05,
+      "loss": 3.3809,
+      "step": 5632
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5879119038581848,
+      "learning_rate": 9.5276376672606e-05,
+      "loss": 3.2301,
+      "step": 5633
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6337886452674866,
+      "learning_rate": 9.519497121522791e-05,
+      "loss": 3.3523,
+      "step": 5634
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6023512482643127,
+      "learning_rate": 9.5113592370629e-05,
+      "loss": 3.3258,
+      "step": 5635
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6364913582801819,
+      "learning_rate": 9.503224015279916e-05,
+      "loss": 3.2814,
+      "step": 5636
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6315597295761108,
+      "learning_rate": 9.49509145757238e-05,
+      "loss": 3.2524,
+      "step": 5637
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6659747958183289,
+      "learning_rate": 9.48696156533836e-05,
+      "loss": 3.3089,
+      "step": 5638
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6412807703018188,
+      "learning_rate": 9.478834339975498e-05,
+      "loss": 3.3407,
+      "step": 5639
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.654059112071991,
+      "learning_rate": 9.470709782880952e-05,
+      "loss": 3.2988,
+      "step": 5640
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6097717881202698,
+      "learning_rate": 9.462587895451424e-05,
+      "loss": 3.269,
+      "step": 5641
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6349428296089172,
+      "learning_rate": 9.454468679083161e-05,
+      "loss": 3.3455,
+      "step": 5642
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6521540880203247,
+      "learning_rate": 9.446352135171943e-05,
+      "loss": 3.1802,
+      "step": 5643
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.596436083316803,
+      "learning_rate": 9.438238265113116e-05,
+      "loss": 3.2043,
+      "step": 5644
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6569159030914307,
+      "learning_rate": 9.43012707030153e-05,
+      "loss": 3.3805,
+      "step": 5645
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6213423013687134,
+      "learning_rate": 9.422018552131611e-05,
+      "loss": 3.3731,
+      "step": 5646
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6433305740356445,
+      "learning_rate": 9.413912711997297e-05,
+      "loss": 3.391,
+      "step": 5647
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6438114643096924,
+      "learning_rate": 9.405809551292077e-05,
+      "loss": 3.1735,
+      "step": 5648
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6398864984512329,
+      "learning_rate": 9.39770907140898e-05,
+      "loss": 3.4009,
+      "step": 5649
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6665001511573792,
+      "learning_rate": 9.38961127374057e-05,
+      "loss": 3.3683,
+      "step": 5650
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6613116264343262,
+      "learning_rate": 9.381516159678955e-05,
+      "loss": 3.2784,
+      "step": 5651
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6008492708206177,
+      "learning_rate": 9.373423730615766e-05,
+      "loss": 3.4032,
+      "step": 5652
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6395289301872253,
+      "learning_rate": 9.365333987942199e-05,
+      "loss": 3.2685,
+      "step": 5653
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6400224566459656,
+      "learning_rate": 9.357246933048977e-05,
+      "loss": 3.5009,
+      "step": 5654
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6652866005897522,
+      "learning_rate": 9.349162567326355e-05,
+      "loss": 3.2542,
+      "step": 5655
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6197553873062134,
+      "learning_rate": 9.34108089216413e-05,
+      "loss": 3.3471,
+      "step": 5656
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5985637903213501,
+      "learning_rate": 9.333001908951633e-05,
+      "loss": 3.2991,
+      "step": 5657
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5842049717903137,
+      "learning_rate": 9.324925619077732e-05,
+      "loss": 3.4584,
+      "step": 5658
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.59958815574646,
+      "learning_rate": 9.316852023930832e-05,
+      "loss": 3.1319,
+      "step": 5659
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5971797108650208,
+      "learning_rate": 9.308781124898894e-05,
+      "loss": 3.2662,
+      "step": 5660
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5735284686088562,
+      "learning_rate": 9.300712923369387e-05,
+      "loss": 3.2252,
+      "step": 5661
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6035478711128235,
+      "learning_rate": 9.29264742072932e-05,
+      "loss": 3.1127,
+      "step": 5662
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6161136627197266,
+      "learning_rate": 9.284584618365266e-05,
+      "loss": 3.3258,
+      "step": 5663
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.684057354927063,
+      "learning_rate": 9.276524517663306e-05,
+      "loss": 3.2352,
+      "step": 5664
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.620516300201416,
+      "learning_rate": 9.268467120009063e-05,
+      "loss": 3.2965,
+      "step": 5665
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6349815130233765,
+      "learning_rate": 9.26041242678769e-05,
+      "loss": 3.2535,
+      "step": 5666
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6179966926574707,
+      "learning_rate": 9.252360439383897e-05,
+      "loss": 3.2433,
+      "step": 5667
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6889963150024414,
+      "learning_rate": 9.244311159181906e-05,
+      "loss": 3.3033,
+      "step": 5668
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6488670110702515,
+      "learning_rate": 9.236264587565485e-05,
+      "loss": 3.3602,
+      "step": 5669
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6509321331977844,
+      "learning_rate": 9.22822072591793e-05,
+      "loss": 3.2645,
+      "step": 5670
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6226929426193237,
+      "learning_rate": 9.220179575622065e-05,
+      "loss": 3.4073,
+      "step": 5671
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6354256868362427,
+      "learning_rate": 9.212141138060273e-05,
+      "loss": 3.3473,
+      "step": 5672
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6060647368431091,
+      "learning_rate": 9.20410541461445e-05,
+      "loss": 3.3694,
+      "step": 5673
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6140630841255188,
+      "learning_rate": 9.19607240666602e-05,
+      "loss": 3.2368,
+      "step": 5674
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6258276104927063,
+      "learning_rate": 9.188042115595966e-05,
+      "loss": 3.3254,
+      "step": 5675
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.633005678653717,
+      "learning_rate": 9.180014542784779e-05,
+      "loss": 3.3143,
+      "step": 5676
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6406413316726685,
+      "learning_rate": 9.171989689612495e-05,
+      "loss": 3.3556,
+      "step": 5677
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6403318047523499,
+      "learning_rate": 9.163967557458675e-05,
+      "loss": 3.2525,
+      "step": 5678
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5798841714859009,
+      "learning_rate": 9.155948147702419e-05,
+      "loss": 3.1693,
+      "step": 5679
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5711493492126465,
+      "learning_rate": 9.14793146172235e-05,
+      "loss": 3.3399,
+      "step": 5680
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6775694489479065,
+      "learning_rate": 9.139917500896635e-05,
+      "loss": 3.3542,
+      "step": 5681
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6326046586036682,
+      "learning_rate": 9.131906266602977e-05,
+      "loss": 3.3323,
+      "step": 5682
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6392051577568054,
+      "learning_rate": 9.123897760218589e-05,
+      "loss": 3.3526,
+      "step": 5683
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6865871548652649,
+      "learning_rate": 9.115891983120228e-05,
+      "loss": 3.3018,
+      "step": 5684
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6464802622795105,
+      "learning_rate": 9.107888936684181e-05,
+      "loss": 3.2994,
+      "step": 5685
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6245539784431458,
+      "learning_rate": 9.099888622286262e-05,
+      "loss": 3.2927,
+      "step": 5686
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6925250291824341,
+      "learning_rate": 9.091891041301808e-05,
+      "loss": 3.3068,
+      "step": 5687
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.62113356590271,
+      "learning_rate": 9.083896195105718e-05,
+      "loss": 3.2647,
+      "step": 5688
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6427499055862427,
+      "learning_rate": 9.075904085072375e-05,
+      "loss": 3.3155,
+      "step": 5689
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6446628570556641,
+      "learning_rate": 9.06791471257574e-05,
+      "loss": 3.35,
+      "step": 5690
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.596994161605835,
+      "learning_rate": 9.059928078989266e-05,
+      "loss": 3.3549,
+      "step": 5691
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6257395148277283,
+      "learning_rate": 9.051944185685948e-05,
+      "loss": 3.3217,
+      "step": 5692
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6585277915000916,
+      "learning_rate": 9.04396303403831e-05,
+      "loss": 3.3329,
+      "step": 5693
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6199955344200134,
+      "learning_rate": 9.035984625418406e-05,
+      "loss": 3.3027,
+      "step": 5694
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6597419381141663,
+      "learning_rate": 9.028008961197803e-05,
+      "loss": 3.284,
+      "step": 5695
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6863325834274292,
+      "learning_rate": 9.020036042747637e-05,
+      "loss": 3.3692,
+      "step": 5696
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6006118059158325,
+      "learning_rate": 9.012065871438527e-05,
+      "loss": 3.374,
+      "step": 5697
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.643882691860199,
+      "learning_rate": 9.004098448640643e-05,
+      "loss": 3.2162,
+      "step": 5698
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6381259560585022,
+      "learning_rate": 8.996133775723666e-05,
+      "loss": 3.3138,
+      "step": 5699
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7250969409942627,
+      "learning_rate": 8.988171854056837e-05,
+      "loss": 3.3412,
+      "step": 5700
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5688900351524353,
+      "learning_rate": 8.98021268500889e-05,
+      "loss": 3.2126,
+      "step": 5701
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6205479502677917,
+      "learning_rate": 8.97225626994809e-05,
+      "loss": 3.376,
+      "step": 5702
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6234270334243774,
+      "learning_rate": 8.964302610242256e-05,
+      "loss": 3.2247,
+      "step": 5703
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6274899244308472,
+      "learning_rate": 8.956351707258705e-05,
+      "loss": 3.2646,
+      "step": 5704
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6678943037986755,
+      "learning_rate": 8.948403562364291e-05,
+      "loss": 3.3395,
+      "step": 5705
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6532748341560364,
+      "learning_rate": 8.940458176925389e-05,
+      "loss": 3.3363,
+      "step": 5706
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6314519643783569,
+      "learning_rate": 8.932515552307904e-05,
+      "loss": 3.2931,
+      "step": 5707
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6630345582962036,
+      "learning_rate": 8.92457568987726e-05,
+      "loss": 3.1622,
+      "step": 5708
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6650614738464355,
+      "learning_rate": 8.916638590998425e-05,
+      "loss": 3.3558,
+      "step": 5709
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6153234243392944,
+      "learning_rate": 8.90870425703586e-05,
+      "loss": 3.2526,
+      "step": 5710
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.650212824344635,
+      "learning_rate": 8.900772689353589e-05,
+      "loss": 3.3059,
+      "step": 5711
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6286440491676331,
+      "learning_rate": 8.892843889315133e-05,
+      "loss": 3.3715,
+      "step": 5712
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6217995285987854,
+      "learning_rate": 8.88491785828354e-05,
+      "loss": 3.3575,
+      "step": 5713
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6089176535606384,
+      "learning_rate": 8.876994597621391e-05,
+      "loss": 3.3028,
+      "step": 5714
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.631025493144989,
+      "learning_rate": 8.869074108690783e-05,
+      "loss": 3.3126,
+      "step": 5715
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6296160817146301,
+      "learning_rate": 8.861156392853334e-05,
+      "loss": 3.344,
+      "step": 5716
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6157452464103699,
+      "learning_rate": 8.853241451470198e-05,
+      "loss": 3.1725,
+      "step": 5717
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6230807304382324,
+      "learning_rate": 8.845329285902054e-05,
+      "loss": 3.2362,
+      "step": 5718
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7034242749214172,
+      "learning_rate": 8.837419897509086e-05,
+      "loss": 3.3586,
+      "step": 5719
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6778315305709839,
+      "learning_rate": 8.829513287651011e-05,
+      "loss": 3.3762,
+      "step": 5720
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6450397372245789,
+      "learning_rate": 8.821609457687069e-05,
+      "loss": 3.2485,
+      "step": 5721
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6391881108283997,
+      "learning_rate": 8.813708408976015e-05,
+      "loss": 3.484,
+      "step": 5722
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6348711848258972,
+      "learning_rate": 8.805810142876123e-05,
+      "loss": 3.2173,
+      "step": 5723
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6497623920440674,
+      "learning_rate": 8.797914660745218e-05,
+      "loss": 3.29,
+      "step": 5724
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.581919252872467,
+      "learning_rate": 8.790021963940612e-05,
+      "loss": 3.3053,
+      "step": 5725
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6217665672302246,
+      "learning_rate": 8.782132053819145e-05,
+      "loss": 3.2744,
+      "step": 5726
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.645945131778717,
+      "learning_rate": 8.774244931737197e-05,
+      "loss": 3.2167,
+      "step": 5727
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6536757349967957,
+      "learning_rate": 8.766360599050654e-05,
+      "loss": 3.4259,
+      "step": 5728
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6132533550262451,
+      "learning_rate": 8.758479057114917e-05,
+      "loss": 3.2033,
+      "step": 5729
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6727334260940552,
+      "learning_rate": 8.750600307284922e-05,
+      "loss": 3.3885,
+      "step": 5730
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.676634669303894,
+      "learning_rate": 8.742724350915102e-05,
+      "loss": 3.4119,
+      "step": 5731
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6278587579727173,
+      "learning_rate": 8.734851189359447e-05,
+      "loss": 3.3256,
+      "step": 5732
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.623261034488678,
+      "learning_rate": 8.726980823971434e-05,
+      "loss": 3.3331,
+      "step": 5733
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6241022944450378,
+      "learning_rate": 8.719113256104069e-05,
+      "loss": 3.3075,
+      "step": 5734
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.593539297580719,
+      "learning_rate": 8.71124848710988e-05,
+      "loss": 3.2942,
+      "step": 5735
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6482180953025818,
+      "learning_rate": 8.703386518340902e-05,
+      "loss": 3.3901,
+      "step": 5736
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6159878969192505,
+      "learning_rate": 8.695527351148719e-05,
+      "loss": 3.431,
+      "step": 5737
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6078242659568787,
+      "learning_rate": 8.687670986884394e-05,
+      "loss": 3.3015,
+      "step": 5738
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.604847252368927,
+      "learning_rate": 8.67981742689854e-05,
+      "loss": 3.3483,
+      "step": 5739
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6486718058586121,
+      "learning_rate": 8.671966672541273e-05,
+      "loss": 3.378,
+      "step": 5740
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6555278897285461,
+      "learning_rate": 8.664118725162226e-05,
+      "loss": 3.2885,
+      "step": 5741
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6854482293128967,
+      "learning_rate": 8.656273586110549e-05,
+      "loss": 3.4235,
+      "step": 5742
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6202027797698975,
+      "learning_rate": 8.648431256734918e-05,
+      "loss": 3.2464,
+      "step": 5743
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6301712393760681,
+      "learning_rate": 8.640591738383507e-05,
+      "loss": 3.3117,
+      "step": 5744
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5973697304725647,
+      "learning_rate": 8.632755032404041e-05,
+      "loss": 3.2841,
+      "step": 5745
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.601158082485199,
+      "learning_rate": 8.624921140143722e-05,
+      "loss": 3.1326,
+      "step": 5746
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5854448676109314,
+      "learning_rate": 8.617090062949303e-05,
+      "loss": 3.3146,
+      "step": 5747
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6340086460113525,
+      "learning_rate": 8.609261802167029e-05,
+      "loss": 3.3485,
+      "step": 5748
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.634315013885498,
+      "learning_rate": 8.60143635914267e-05,
+      "loss": 3.2879,
+      "step": 5749
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.639906644821167,
+      "learning_rate": 8.593613735221506e-05,
+      "loss": 3.3795,
+      "step": 5750
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6317019462585449,
+      "learning_rate": 8.585793931748343e-05,
+      "loss": 3.2666,
+      "step": 5751
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6535657048225403,
+      "learning_rate": 8.577976950067484e-05,
+      "loss": 3.3267,
+      "step": 5752
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6275309324264526,
+      "learning_rate": 8.570162791522776e-05,
+      "loss": 3.5098,
+      "step": 5753
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6426060199737549,
+      "learning_rate": 8.562351457457549e-05,
+      "loss": 3.4356,
+      "step": 5754
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6221879124641418,
+      "learning_rate": 8.554542949214672e-05,
+      "loss": 3.2748,
+      "step": 5755
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6630797982215881,
+      "learning_rate": 8.546737268136518e-05,
+      "loss": 3.4619,
+      "step": 5756
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6118805408477783,
+      "learning_rate": 8.538934415564966e-05,
+      "loss": 3.3449,
+      "step": 5757
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6531286239624023,
+      "learning_rate": 8.531134392841424e-05,
+      "loss": 3.3845,
+      "step": 5758
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6513779759407043,
+      "learning_rate": 8.523337201306796e-05,
+      "loss": 3.3936,
+      "step": 5759
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.653174638748169,
+      "learning_rate": 8.515542842301524e-05,
+      "loss": 3.1839,
+      "step": 5760
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6509855389595032,
+      "learning_rate": 8.507751317165541e-05,
+      "loss": 3.3573,
+      "step": 5761
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.640022873878479,
+      "learning_rate": 8.499962627238302e-05,
+      "loss": 3.35,
+      "step": 5762
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6861554384231567,
+      "learning_rate": 8.492176773858765e-05,
+      "loss": 3.3539,
+      "step": 5763
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6354894042015076,
+      "learning_rate": 8.484393758365422e-05,
+      "loss": 3.3855,
+      "step": 5764
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.679643988609314,
+      "learning_rate": 8.476613582096257e-05,
+      "loss": 3.2839,
+      "step": 5765
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6623861789703369,
+      "learning_rate": 8.468836246388772e-05,
+      "loss": 3.3924,
+      "step": 5766
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6140841841697693,
+      "learning_rate": 8.461061752579976e-05,
+      "loss": 3.3562,
+      "step": 5767
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.613014280796051,
+      "learning_rate": 8.453290102006408e-05,
+      "loss": 3.2021,
+      "step": 5768
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6353635191917419,
+      "learning_rate": 8.445521296004099e-05,
+      "loss": 3.2412,
+      "step": 5769
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6085248589515686,
+      "learning_rate": 8.437755335908592e-05,
+      "loss": 3.2289,
+      "step": 5770
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6465585231781006,
+      "learning_rate": 8.429992223054952e-05,
+      "loss": 3.4146,
+      "step": 5771
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6053239703178406,
+      "learning_rate": 8.422231958777743e-05,
+      "loss": 3.3973,
+      "step": 5772
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6017369627952576,
+      "learning_rate": 8.414474544411038e-05,
+      "loss": 3.3329,
+      "step": 5773
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6545090675354004,
+      "learning_rate": 8.406719981288436e-05,
+      "loss": 3.3531,
+      "step": 5774
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6039037704467773,
+      "learning_rate": 8.398968270743041e-05,
+      "loss": 3.3622,
+      "step": 5775
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6185946464538574,
+      "learning_rate": 8.391219414107456e-05,
+      "loss": 3.3219,
+      "step": 5776
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6226698756217957,
+      "learning_rate": 8.383473412713802e-05,
+      "loss": 3.292,
+      "step": 5777
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6575097441673279,
+      "learning_rate": 8.375730267893703e-05,
+      "loss": 3.1514,
+      "step": 5778
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6401361227035522,
+      "learning_rate": 8.367989980978294e-05,
+      "loss": 3.3897,
+      "step": 5779
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6216908097267151,
+      "learning_rate": 8.360252553298214e-05,
+      "loss": 3.2323,
+      "step": 5780
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5935060381889343,
+      "learning_rate": 8.352517986183636e-05,
+      "loss": 3.3945,
+      "step": 5781
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6383286118507385,
+      "learning_rate": 8.344786280964197e-05,
+      "loss": 3.3122,
+      "step": 5782
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6526837348937988,
+      "learning_rate": 8.337057438969092e-05,
+      "loss": 3.2942,
+      "step": 5783
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6335547566413879,
+      "learning_rate": 8.329331461526988e-05,
+      "loss": 3.3683,
+      "step": 5784
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.619326114654541,
+      "learning_rate": 8.321608349966065e-05,
+      "loss": 3.3295,
+      "step": 5785
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6175655722618103,
+      "learning_rate": 8.31388810561402e-05,
+      "loss": 3.1881,
+      "step": 5786
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6244990825653076,
+      "learning_rate": 8.306170729798054e-05,
+      "loss": 3.3252,
+      "step": 5787
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6253812313079834,
+      "learning_rate": 8.29845622384486e-05,
+      "loss": 3.2042,
+      "step": 5788
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6498683094978333,
+      "learning_rate": 8.290744589080674e-05,
+      "loss": 3.3758,
+      "step": 5789
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6260440945625305,
+      "learning_rate": 8.283035826831201e-05,
+      "loss": 3.2997,
+      "step": 5790
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6759835481643677,
+      "learning_rate": 8.275329938421658e-05,
+      "loss": 3.3167,
+      "step": 5791
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6252330541610718,
+      "learning_rate": 8.2676269251768e-05,
+      "loss": 3.2903,
+      "step": 5792
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6032154560089111,
+      "learning_rate": 8.25992678842085e-05,
+      "loss": 3.189,
+      "step": 5793
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6188944578170776,
+      "learning_rate": 8.252229529477554e-05,
+      "loss": 3.165,
+      "step": 5794
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.630266547203064,
+      "learning_rate": 8.244535149670148e-05,
+      "loss": 3.1759,
+      "step": 5795
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6536327600479126,
+      "learning_rate": 8.236843650321408e-05,
+      "loss": 3.2775,
+      "step": 5796
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6232385635375977,
+      "learning_rate": 8.229155032753577e-05,
+      "loss": 3.2553,
+      "step": 5797
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6598358750343323,
+      "learning_rate": 8.221469298288419e-05,
+      "loss": 3.392,
+      "step": 5798
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6670900583267212,
+      "learning_rate": 8.213786448247205e-05,
+      "loss": 3.3819,
+      "step": 5799
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6062140464782715,
+      "learning_rate": 8.206106483950693e-05,
+      "loss": 3.3811,
+      "step": 5800
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6420729160308838,
+      "learning_rate": 8.198429406719177e-05,
+      "loss": 3.196,
+      "step": 5801
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6605176329612732,
+      "learning_rate": 8.190755217872425e-05,
+      "loss": 3.4285,
+      "step": 5802
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6510549783706665,
+      "learning_rate": 8.183083918729713e-05,
+      "loss": 3.4052,
+      "step": 5803
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6223390698432922,
+      "learning_rate": 8.175415510609843e-05,
+      "loss": 3.257,
+      "step": 5804
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6070486903190613,
+      "learning_rate": 8.167749994831092e-05,
+      "loss": 3.2376,
+      "step": 5805
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6169131994247437,
+      "learning_rate": 8.160087372711256e-05,
+      "loss": 3.2959,
+      "step": 5806
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6251973509788513,
+      "learning_rate": 8.152427645567622e-05,
+      "loss": 3.3106,
+      "step": 5807
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.629302442073822,
+      "learning_rate": 8.144770814716993e-05,
+      "loss": 3.1936,
+      "step": 5808
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6064233183860779,
+      "learning_rate": 8.137116881475653e-05,
+      "loss": 3.3217,
+      "step": 5809
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5733712315559387,
+      "learning_rate": 8.129465847159414e-05,
+      "loss": 3.2617,
+      "step": 5810
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6017459630966187,
+      "learning_rate": 8.121817713083584e-05,
+      "loss": 3.2549,
+      "step": 5811
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5999737977981567,
+      "learning_rate": 8.114172480562957e-05,
+      "loss": 3.3636,
+      "step": 5812
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6227502226829529,
+      "learning_rate": 8.106530150911837e-05,
+      "loss": 3.2725,
+      "step": 5813
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.620937705039978,
+      "learning_rate": 8.098890725444033e-05,
+      "loss": 3.2942,
+      "step": 5814
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5800301432609558,
+      "learning_rate": 8.091254205472846e-05,
+      "loss": 3.2182,
+      "step": 5815
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6282032132148743,
+      "learning_rate": 8.083620592311075e-05,
+      "loss": 3.2537,
+      "step": 5816
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6425186991691589,
+      "learning_rate": 8.075989887271043e-05,
+      "loss": 3.3287,
+      "step": 5817
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6459394097328186,
+      "learning_rate": 8.068362091664552e-05,
+      "loss": 3.2264,
+      "step": 5818
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6475614905357361,
+      "learning_rate": 8.060737206802896e-05,
+      "loss": 3.2835,
+      "step": 5819
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6001794934272766,
+      "learning_rate": 8.0531152339969e-05,
+      "loss": 3.341,
+      "step": 5820
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6368500590324402,
+      "learning_rate": 8.04549617455686e-05,
+      "loss": 3.2757,
+      "step": 5821
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6508361101150513,
+      "learning_rate": 8.037880029792582e-05,
+      "loss": 3.3815,
+      "step": 5822
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5668452978134155,
+      "learning_rate": 8.030266801013366e-05,
+      "loss": 3.3402,
+      "step": 5823
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6491958498954773,
+      "learning_rate": 8.022656489528013e-05,
+      "loss": 3.4139,
+      "step": 5824
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6883427500724792,
+      "learning_rate": 8.015049096644833e-05,
+      "loss": 3.3844,
+      "step": 5825
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6037605404853821,
+      "learning_rate": 8.007444623671619e-05,
+      "loss": 3.3302,
+      "step": 5826
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6611215472221375,
+      "learning_rate": 7.999843071915671e-05,
+      "loss": 3.2731,
+      "step": 5827
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6562877893447876,
+      "learning_rate": 7.992244442683771e-05,
+      "loss": 3.3768,
+      "step": 5828
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6370260119438171,
+      "learning_rate": 7.984648737282232e-05,
+      "loss": 3.3764,
+      "step": 5829
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6698316931724548,
+      "learning_rate": 7.977055957016835e-05,
+      "loss": 3.2406,
+      "step": 5830
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6381048560142517,
+      "learning_rate": 7.969466103192858e-05,
+      "loss": 3.245,
+      "step": 5831
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6150162816047668,
+      "learning_rate": 7.961879177115097e-05,
+      "loss": 3.1812,
+      "step": 5832
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5887827277183533,
+      "learning_rate": 7.954295180087831e-05,
+      "loss": 3.2838,
+      "step": 5833
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5896087884902954,
+      "learning_rate": 7.946714113414836e-05,
+      "loss": 3.3149,
+      "step": 5834
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6396685838699341,
+      "learning_rate": 7.939135978399382e-05,
+      "loss": 3.3465,
+      "step": 5835
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5983510613441467,
+      "learning_rate": 7.931560776344238e-05,
+      "loss": 3.3145,
+      "step": 5836
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6477048397064209,
+      "learning_rate": 7.923988508551663e-05,
+      "loss": 3.3467,
+      "step": 5837
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6282901763916016,
+      "learning_rate": 7.916419176323428e-05,
+      "loss": 3.3294,
+      "step": 5838
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5895499587059021,
+      "learning_rate": 7.908852780960794e-05,
+      "loss": 3.3558,
+      "step": 5839
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6594005823135376,
+      "learning_rate": 7.901289323764502e-05,
+      "loss": 3.2958,
+      "step": 5840
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6194105744361877,
+      "learning_rate": 7.893728806034803e-05,
+      "loss": 3.08,
+      "step": 5841
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6449106931686401,
+      "learning_rate": 7.886171229071434e-05,
+      "loss": 3.3395,
+      "step": 5842
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6128101944923401,
+      "learning_rate": 7.878616594173632e-05,
+      "loss": 3.2356,
+      "step": 5843
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6564944386482239,
+      "learning_rate": 7.871064902640124e-05,
+      "loss": 3.2198,
+      "step": 5844
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.650698721408844,
+      "learning_rate": 7.863516155769129e-05,
+      "loss": 3.2793,
+      "step": 5845
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6543959379196167,
+      "learning_rate": 7.855970354858378e-05,
+      "loss": 3.246,
+      "step": 5846
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6274881958961487,
+      "learning_rate": 7.848427501205064e-05,
+      "loss": 3.2797,
+      "step": 5847
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5793599486351013,
+      "learning_rate": 7.840887596105909e-05,
+      "loss": 3.2347,
+      "step": 5848
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6473877429962158,
+      "learning_rate": 7.833350640857101e-05,
+      "loss": 3.391,
+      "step": 5849
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6663014888763428,
+      "learning_rate": 7.825816636754333e-05,
+      "loss": 3.3715,
+      "step": 5850
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6312239170074463,
+      "learning_rate": 7.818285585092783e-05,
+      "loss": 3.2068,
+      "step": 5851
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.622334361076355,
+      "learning_rate": 7.810757487167122e-05,
+      "loss": 3.1758,
+      "step": 5852
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6082350015640259,
+      "learning_rate": 7.803232344271532e-05,
+      "loss": 3.3347,
+      "step": 5853
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6351848244667053,
+      "learning_rate": 7.795710157699662e-05,
+      "loss": 3.1847,
+      "step": 5854
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6521065831184387,
+      "learning_rate": 7.788190928744668e-05,
+      "loss": 3.331,
+      "step": 5855
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6365451216697693,
+      "learning_rate": 7.78067465869918e-05,
+      "loss": 3.4912,
+      "step": 5856
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6078124046325684,
+      "learning_rate": 7.773161348855349e-05,
+      "loss": 3.2794,
+      "step": 5857
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6526720523834229,
+      "learning_rate": 7.765651000504795e-05,
+      "loss": 3.2953,
+      "step": 5858
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6162887811660767,
+      "learning_rate": 7.758143614938621e-05,
+      "loss": 3.3665,
+      "step": 5859
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6105762720108032,
+      "learning_rate": 7.750639193447454e-05,
+      "loss": 3.4606,
+      "step": 5860
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5999521017074585,
+      "learning_rate": 7.743137737321381e-05,
+      "loss": 3.2395,
+      "step": 5861
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6496285200119019,
+      "learning_rate": 7.73563924784999e-05,
+      "loss": 3.2687,
+      "step": 5862
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6636385321617126,
+      "learning_rate": 7.728143726322359e-05,
+      "loss": 3.3295,
+      "step": 5863
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.648780882358551,
+      "learning_rate": 7.720651174027051e-05,
+      "loss": 3.2538,
+      "step": 5864
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6651748418807983,
+      "learning_rate": 7.713161592252121e-05,
+      "loss": 3.2258,
+      "step": 5865
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6317049264907837,
+      "learning_rate": 7.705674982285127e-05,
+      "loss": 3.2107,
+      "step": 5866
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6234962344169617,
+      "learning_rate": 7.698191345413086e-05,
+      "loss": 3.2145,
+      "step": 5867
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6090578436851501,
+      "learning_rate": 7.690710682922541e-05,
+      "loss": 3.3738,
+      "step": 5868
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6581776738166809,
+      "learning_rate": 7.683232996099498e-05,
+      "loss": 3.2474,
+      "step": 5869
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6431095004081726,
+      "learning_rate": 7.675758286229456e-05,
+      "loss": 3.2965,
+      "step": 5870
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.648125171661377,
+      "learning_rate": 7.668286554597404e-05,
+      "loss": 3.2822,
+      "step": 5871
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6426742672920227,
+      "learning_rate": 7.660817802487819e-05,
+      "loss": 3.3757,
+      "step": 5872
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6470495462417603,
+      "learning_rate": 7.65335203118466e-05,
+      "loss": 3.3586,
+      "step": 5873
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6357805132865906,
+      "learning_rate": 7.645889241971384e-05,
+      "loss": 3.1924,
+      "step": 5874
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5984390377998352,
+      "learning_rate": 7.638429436130945e-05,
+      "loss": 3.2409,
+      "step": 5875
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6378467679023743,
+      "learning_rate": 7.630972614945756e-05,
+      "loss": 3.3597,
+      "step": 5876
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7053393721580505,
+      "learning_rate": 7.623518779697733e-05,
+      "loss": 3.4355,
+      "step": 5877
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.673662006855011,
+      "learning_rate": 7.616067931668277e-05,
+      "loss": 3.3187,
+      "step": 5878
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6976150870323181,
+      "learning_rate": 7.608620072138278e-05,
+      "loss": 3.3573,
+      "step": 5879
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6496651768684387,
+      "learning_rate": 7.601175202388097e-05,
+      "loss": 3.2845,
+      "step": 5880
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6530563831329346,
+      "learning_rate": 7.593733323697613e-05,
+      "loss": 3.317,
+      "step": 5881
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6424029469490051,
+      "learning_rate": 7.586294437346158e-05,
+      "loss": 3.2376,
+      "step": 5882
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6783373355865479,
+      "learning_rate": 7.578858544612571e-05,
+      "loss": 3.2549,
+      "step": 5883
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6218166947364807,
+      "learning_rate": 7.571425646775151e-05,
+      "loss": 3.3064,
+      "step": 5884
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6839401125907898,
+      "learning_rate": 7.563995745111724e-05,
+      "loss": 3.295,
+      "step": 5885
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6496739983558655,
+      "learning_rate": 7.55656884089956e-05,
+      "loss": 3.3063,
+      "step": 5886
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6135286688804626,
+      "learning_rate": 7.549144935415434e-05,
+      "loss": 3.3386,
+      "step": 5887
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6319910883903503,
+      "learning_rate": 7.541724029935596e-05,
+      "loss": 3.269,
+      "step": 5888
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.691398561000824,
+      "learning_rate": 7.534306125735796e-05,
+      "loss": 3.2814,
+      "step": 5889
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6186633706092834,
+      "learning_rate": 7.526891224091254e-05,
+      "loss": 3.2561,
+      "step": 5890
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6427898406982422,
+      "learning_rate": 7.519479326276677e-05,
+      "loss": 3.4017,
+      "step": 5891
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6385214328765869,
+      "learning_rate": 7.512070433566253e-05,
+      "loss": 3.404,
+      "step": 5892
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6215147972106934,
+      "learning_rate": 7.504664547233655e-05,
+      "loss": 3.3073,
+      "step": 5893
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6701691746711731,
+      "learning_rate": 7.497261668552049e-05,
+      "loss": 3.2871,
+      "step": 5894
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5981207489967346,
+      "learning_rate": 7.489861798794065e-05,
+      "loss": 3.2929,
+      "step": 5895
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6607575416564941,
+      "learning_rate": 7.482464939231842e-05,
+      "loss": 3.4491,
+      "step": 5896
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5619738101959229,
+      "learning_rate": 7.475071091136973e-05,
+      "loss": 3.2607,
+      "step": 5897
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6099634766578674,
+      "learning_rate": 7.467680255780555e-05,
+      "loss": 3.2743,
+      "step": 5898
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6769587397575378,
+      "learning_rate": 7.460292434433147e-05,
+      "loss": 3.3378,
+      "step": 5899
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6428564786911011,
+      "learning_rate": 7.45290762836481e-05,
+      "loss": 3.3223,
+      "step": 5900
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7414649128913879,
+      "learning_rate": 7.445525838845076e-05,
+      "loss": 3.2627,
+      "step": 5901
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6124725341796875,
+      "learning_rate": 7.43814706714295e-05,
+      "loss": 3.2105,
+      "step": 5902
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6249970197677612,
+      "learning_rate": 7.43077131452694e-05,
+      "loss": 3.4095,
+      "step": 5903
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7355794310569763,
+      "learning_rate": 7.423398582265026e-05,
+      "loss": 3.2868,
+      "step": 5904
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6775880455970764,
+      "learning_rate": 7.416028871624664e-05,
+      "loss": 3.2442,
+      "step": 5905
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6577247381210327,
+      "learning_rate": 7.408662183872786e-05,
+      "loss": 3.3793,
+      "step": 5906
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7472736239433289,
+      "learning_rate": 7.401298520275817e-05,
+      "loss": 3.2458,
+      "step": 5907
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6332255005836487,
+      "learning_rate": 7.393937882099656e-05,
+      "loss": 3.3351,
+      "step": 5908
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6670913696289062,
+      "learning_rate": 7.386580270609669e-05,
+      "loss": 3.267,
+      "step": 5909
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6957175731658936,
+      "learning_rate": 7.379225687070734e-05,
+      "loss": 3.381,
+      "step": 5910
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6249982118606567,
+      "learning_rate": 7.371874132747175e-05,
+      "loss": 3.3498,
+      "step": 5911
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6238232851028442,
+      "learning_rate": 7.364525608902823e-05,
+      "loss": 3.2145,
+      "step": 5912
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6316736936569214,
+      "learning_rate": 7.357180116800965e-05,
+      "loss": 3.2978,
+      "step": 5913
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.559715211391449,
+      "learning_rate": 7.349837657704378e-05,
+      "loss": 3.328,
+      "step": 5914
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6490960717201233,
+      "learning_rate": 7.342498232875319e-05,
+      "loss": 3.3539,
+      "step": 5915
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6523640751838684,
+      "learning_rate": 7.335161843575505e-05,
+      "loss": 3.2819,
+      "step": 5916
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6588881015777588,
+      "learning_rate": 7.327828491066169e-05,
+      "loss": 3.3489,
+      "step": 5917
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6438858509063721,
+      "learning_rate": 7.32049817660799e-05,
+      "loss": 3.2901,
+      "step": 5918
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6138406991958618,
+      "learning_rate": 7.31317090146113e-05,
+      "loss": 3.1689,
+      "step": 5919
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5893608927726746,
+      "learning_rate": 7.305846666885236e-05,
+      "loss": 3.3019,
+      "step": 5920
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6357761025428772,
+      "learning_rate": 7.298525474139419e-05,
+      "loss": 3.3627,
+      "step": 5921
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6008619666099548,
+      "learning_rate": 7.291207324482296e-05,
+      "loss": 3.3415,
+      "step": 5922
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6560606956481934,
+      "learning_rate": 7.283892219171933e-05,
+      "loss": 3.2799,
+      "step": 5923
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5996524691581726,
+      "learning_rate": 7.27658015946587e-05,
+      "loss": 3.2154,
+      "step": 5924
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5937591195106506,
+      "learning_rate": 7.269271146621153e-05,
+      "loss": 3.2677,
+      "step": 5925
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6687285900115967,
+      "learning_rate": 7.26196518189428e-05,
+      "loss": 3.3694,
+      "step": 5926
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6421836614608765,
+      "learning_rate": 7.254662266541229e-05,
+      "loss": 3.2305,
+      "step": 5927
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6282448172569275,
+      "learning_rate": 7.247362401817456e-05,
+      "loss": 3.4225,
+      "step": 5928
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6458451747894287,
+      "learning_rate": 7.24006558897789e-05,
+      "loss": 3.3958,
+      "step": 5929
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6398035287857056,
+      "learning_rate": 7.232771829276935e-05,
+      "loss": 3.2902,
+      "step": 5930
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6559479832649231,
+      "learning_rate": 7.225481123968475e-05,
+      "loss": 3.3943,
+      "step": 5931
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6435176134109497,
+      "learning_rate": 7.218193474305881e-05,
+      "loss": 3.3508,
+      "step": 5932
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7084299921989441,
+      "learning_rate": 7.21090888154197e-05,
+      "loss": 3.3744,
+      "step": 5933
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6554321050643921,
+      "learning_rate": 7.203627346929053e-05,
+      "loss": 3.3302,
+      "step": 5934
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.624515950679779,
+      "learning_rate": 7.196348871718907e-05,
+      "loss": 3.3803,
+      "step": 5935
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6426070928573608,
+      "learning_rate": 7.18907345716279e-05,
+      "loss": 3.2487,
+      "step": 5936
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6676350235939026,
+      "learning_rate": 7.181801104511417e-05,
+      "loss": 3.3991,
+      "step": 5937
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5947633981704712,
+      "learning_rate": 7.174531815015009e-05,
+      "loss": 3.2309,
+      "step": 5938
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6089289784431458,
+      "learning_rate": 7.167265589923228e-05,
+      "loss": 3.2029,
+      "step": 5939
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6404508352279663,
+      "learning_rate": 7.16000243048523e-05,
+      "loss": 3.3253,
+      "step": 5940
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6133842468261719,
+      "learning_rate": 7.152742337949636e-05,
+      "loss": 3.2687,
+      "step": 5941
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6085859537124634,
+      "learning_rate": 7.145485313564536e-05,
+      "loss": 3.327,
+      "step": 5942
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5930094122886658,
+      "learning_rate": 7.138231358577498e-05,
+      "loss": 3.1893,
+      "step": 5943
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6218498349189758,
+      "learning_rate": 7.13098047423556e-05,
+      "loss": 3.2663,
+      "step": 5944
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6165961623191833,
+      "learning_rate": 7.12373266178523e-05,
+      "loss": 3.3039,
+      "step": 5945
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5737981796264648,
+      "learning_rate": 7.116487922472498e-05,
+      "loss": 3.1787,
+      "step": 5946
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6428831815719604,
+      "learning_rate": 7.109246257542817e-05,
+      "loss": 3.3723,
+      "step": 5947
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6483456492424011,
+      "learning_rate": 7.102007668241107e-05,
+      "loss": 3.3862,
+      "step": 5948
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6161538362503052,
+      "learning_rate": 7.094772155811774e-05,
+      "loss": 3.356,
+      "step": 5949
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.618384063243866,
+      "learning_rate": 7.087539721498684e-05,
+      "loss": 3.1761,
+      "step": 5950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6284879446029663,
+      "learning_rate": 7.080310366545176e-05,
+      "loss": 3.2568,
+      "step": 5951
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6222849488258362,
+      "learning_rate": 7.073084092194049e-05,
+      "loss": 3.326,
+      "step": 5952
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7094422578811646,
+      "learning_rate": 7.065860899687604e-05,
+      "loss": 3.3456,
+      "step": 5953
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5881149768829346,
+      "learning_rate": 7.05864079026758e-05,
+      "loss": 3.2518,
+      "step": 5954
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5845149755477905,
+      "learning_rate": 7.051423765175194e-05,
+      "loss": 3.2955,
+      "step": 5955
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6444967985153198,
+      "learning_rate": 7.044209825651143e-05,
+      "loss": 3.3345,
+      "step": 5956
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5796284675598145,
+      "learning_rate": 7.036998972935585e-05,
+      "loss": 3.2418,
+      "step": 5957
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5962539911270142,
+      "learning_rate": 7.02979120826814e-05,
+      "loss": 3.3014,
+      "step": 5958
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6504797339439392,
+      "learning_rate": 7.022586532887926e-05,
+      "loss": 3.443,
+      "step": 5959
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6190051436424255,
+      "learning_rate": 7.015384948033487e-05,
+      "loss": 3.2198,
+      "step": 5960
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6398943662643433,
+      "learning_rate": 7.00818645494288e-05,
+      "loss": 3.3484,
+      "step": 5961
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6127356290817261,
+      "learning_rate": 7.0009910548536e-05,
+      "loss": 3.077,
+      "step": 5962
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6366421580314636,
+      "learning_rate": 6.993798749002622e-05,
+      "loss": 3.395,
+      "step": 5963
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.606545090675354,
+      "learning_rate": 6.986609538626384e-05,
+      "loss": 3.325,
+      "step": 5964
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5849049091339111,
+      "learning_rate": 6.979423424960795e-05,
+      "loss": 3.2585,
+      "step": 5965
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6326369047164917,
+      "learning_rate": 6.972240409241224e-05,
+      "loss": 3.2424,
+      "step": 5966
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.642951488494873,
+      "learning_rate": 6.965060492702525e-05,
+      "loss": 3.2929,
+      "step": 5967
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6404616236686707,
+      "learning_rate": 6.95788367657901e-05,
+      "loss": 3.3485,
+      "step": 5968
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5683895945549011,
+      "learning_rate": 6.950709962104454e-05,
+      "loss": 3.2801,
+      "step": 5969
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.614601731300354,
+      "learning_rate": 6.943539350512101e-05,
+      "loss": 3.2453,
+      "step": 5970
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6009972095489502,
+      "learning_rate": 6.936371843034663e-05,
+      "loss": 3.3278,
+      "step": 5971
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6426170468330383,
+      "learning_rate": 6.929207440904318e-05,
+      "loss": 3.2078,
+      "step": 5972
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6109356880187988,
+      "learning_rate": 6.922046145352698e-05,
+      "loss": 3.2783,
+      "step": 5973
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5790854096412659,
+      "learning_rate": 6.91488795761093e-05,
+      "loss": 3.3016,
+      "step": 5974
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6267958283424377,
+      "learning_rate": 6.907732878909587e-05,
+      "loss": 3.3314,
+      "step": 5975
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6615608334541321,
+      "learning_rate": 6.900580910478693e-05,
+      "loss": 3.2845,
+      "step": 5976
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6061287522315979,
+      "learning_rate": 6.89343205354778e-05,
+      "loss": 3.2213,
+      "step": 5977
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6183081865310669,
+      "learning_rate": 6.886286309345801e-05,
+      "loss": 3.212,
+      "step": 5978
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.615390419960022,
+      "learning_rate": 6.879143679101202e-05,
+      "loss": 3.3775,
+      "step": 5979
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6467695832252502,
+      "learning_rate": 6.872004164041878e-05,
+      "loss": 3.2776,
+      "step": 5980
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.635640561580658,
+      "learning_rate": 6.864867765395188e-05,
+      "loss": 3.3567,
+      "step": 5981
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6547449231147766,
+      "learning_rate": 6.857734484387976e-05,
+      "loss": 3.2996,
+      "step": 5982
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6103988289833069,
+      "learning_rate": 6.850604322246532e-05,
+      "loss": 3.1834,
+      "step": 5983
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6807669401168823,
+      "learning_rate": 6.843477280196609e-05,
+      "loss": 3.3163,
+      "step": 5984
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5808712244033813,
+      "learning_rate": 6.836353359463424e-05,
+      "loss": 3.3389,
+      "step": 5985
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6186822056770325,
+      "learning_rate": 6.829232561271672e-05,
+      "loss": 3.3118,
+      "step": 5986
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6658785343170166,
+      "learning_rate": 6.822114886845498e-05,
+      "loss": 3.1867,
+      "step": 5987
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6463857889175415,
+      "learning_rate": 6.815000337408506e-05,
+      "loss": 3.2857,
+      "step": 5988
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6342170238494873,
+      "learning_rate": 6.80788891418378e-05,
+      "loss": 3.4287,
+      "step": 5989
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6567581295967102,
+      "learning_rate": 6.800780618393851e-05,
+      "loss": 3.3444,
+      "step": 5990
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6194071173667908,
+      "learning_rate": 6.793675451260717e-05,
+      "loss": 3.2548,
+      "step": 5991
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6007659435272217,
+      "learning_rate": 6.78657341400584e-05,
+      "loss": 3.2942,
+      "step": 5992
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6221792101860046,
+      "learning_rate": 6.779474507850139e-05,
+      "loss": 3.3124,
+      "step": 5993
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6629132032394409,
+      "learning_rate": 6.772378734013992e-05,
+      "loss": 3.2519,
+      "step": 5994
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6474263668060303,
+      "learning_rate": 6.765286093717265e-05,
+      "loss": 3.3981,
+      "step": 5995
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6070880889892578,
+      "learning_rate": 6.758196588179244e-05,
+      "loss": 3.3196,
+      "step": 5996
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5742640495300293,
+      "learning_rate": 6.751110218618714e-05,
+      "loss": 3.2057,
+      "step": 5997
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6586142778396606,
+      "learning_rate": 6.744026986253895e-05,
+      "loss": 3.3675,
+      "step": 5998
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5742660760879517,
+      "learning_rate": 6.736946892302481e-05,
+      "loss": 3.2621,
+      "step": 5999
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6078349351882935,
+      "learning_rate": 6.729869937981619e-05,
+      "loss": 3.2636,
+      "step": 6000
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5926090478897095,
+      "learning_rate": 6.72279612450792e-05,
+      "loss": 3.2122,
+      "step": 6001
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6195435523986816,
+      "learning_rate": 6.715725453097446e-05,
+      "loss": 3.3282,
+      "step": 6002
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.699246883392334,
+      "learning_rate": 6.708657924965746e-05,
+      "loss": 3.3197,
+      "step": 6003
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6332224011421204,
+      "learning_rate": 6.701593541327792e-05,
+      "loss": 3.2629,
+      "step": 6004
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6227474212646484,
+      "learning_rate": 6.694532303398048e-05,
+      "loss": 3.3366,
+      "step": 6005
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6562210917472839,
+      "learning_rate": 6.687474212390418e-05,
+      "loss": 3.3181,
+      "step": 6006
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6172775030136108,
+      "learning_rate": 6.680419269518265e-05,
+      "loss": 3.3731,
+      "step": 6007
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6271465420722961,
+      "learning_rate": 6.673367475994421e-05,
+      "loss": 3.3334,
+      "step": 6008
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.626900851726532,
+      "learning_rate": 6.66631883303116e-05,
+      "loss": 3.4734,
+      "step": 6009
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7117727398872375,
+      "learning_rate": 6.659273341840241e-05,
+      "loss": 3.3387,
+      "step": 6010
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6394992470741272,
+      "learning_rate": 6.652231003632858e-05,
+      "loss": 3.4417,
+      "step": 6011
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6169131398200989,
+      "learning_rate": 6.645191819619672e-05,
+      "loss": 3.3076,
+      "step": 6012
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6018803715705872,
+      "learning_rate": 6.638155791010791e-05,
+      "loss": 3.2318,
+      "step": 6013
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6905596256256104,
+      "learning_rate": 6.631122919015809e-05,
+      "loss": 3.2227,
+      "step": 6014
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5747807025909424,
+      "learning_rate": 6.624093204843746e-05,
+      "loss": 3.1994,
+      "step": 6015
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6372610926628113,
+      "learning_rate": 6.61706664970309e-05,
+      "loss": 3.337,
+      "step": 6016
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6507368087768555,
+      "learning_rate": 6.610043254801784e-05,
+      "loss": 3.3928,
+      "step": 6017
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6256568431854248,
+      "learning_rate": 6.603023021347246e-05,
+      "loss": 3.192,
+      "step": 6018
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6386218667030334,
+      "learning_rate": 6.596005950546327e-05,
+      "loss": 3.3401,
+      "step": 6019
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6473775506019592,
+      "learning_rate": 6.58899204360534e-05,
+      "loss": 3.1698,
+      "step": 6020
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6212315559387207,
+      "learning_rate": 6.58198130173006e-05,
+      "loss": 3.3463,
+      "step": 6021
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6566527485847473,
+      "learning_rate": 6.57497372612571e-05,
+      "loss": 3.3118,
+      "step": 6022
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5994009375572205,
+      "learning_rate": 6.567969317996982e-05,
+      "loss": 3.169,
+      "step": 6023
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6485239267349243,
+      "learning_rate": 6.560968078548005e-05,
+      "loss": 3.288,
+      "step": 6024
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.606691837310791,
+      "learning_rate": 6.553970008982385e-05,
+      "loss": 3.1417,
+      "step": 6025
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6456628441810608,
+      "learning_rate": 6.546975110503164e-05,
+      "loss": 3.2006,
+      "step": 6026
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6619232296943665,
+      "learning_rate": 6.53998338431285e-05,
+      "loss": 3.3092,
+      "step": 6027
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6379346251487732,
+      "learning_rate": 6.532994831613398e-05,
+      "loss": 3.422,
+      "step": 6028
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6716301441192627,
+      "learning_rate": 6.526009453606224e-05,
+      "loss": 3.3259,
+      "step": 6029
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6789576411247253,
+      "learning_rate": 6.519027251492185e-05,
+      "loss": 3.3459,
+      "step": 6030
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6274928450584412,
+      "learning_rate": 6.512048226471617e-05,
+      "loss": 3.2663,
+      "step": 6031
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6633818745613098,
+      "learning_rate": 6.505072379744283e-05,
+      "loss": 3.3433,
+      "step": 6032
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6291759014129639,
+      "learning_rate": 6.498099712509428e-05,
+      "loss": 3.2835,
+      "step": 6033
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.624902606010437,
+      "learning_rate": 6.491130225965722e-05,
+      "loss": 3.3141,
+      "step": 6034
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6179114580154419,
+      "learning_rate": 6.484163921311306e-05,
+      "loss": 3.2325,
+      "step": 6035
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6924793720245361,
+      "learning_rate": 6.477200799743766e-05,
+      "loss": 3.2907,
+      "step": 6036
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5994296073913574,
+      "learning_rate": 6.470240862460142e-05,
+      "loss": 3.1898,
+      "step": 6037
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6019535660743713,
+      "learning_rate": 6.46328411065692e-05,
+      "loss": 3.3504,
+      "step": 6038
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6485642790794373,
+      "learning_rate": 6.456330545530065e-05,
+      "loss": 3.4536,
+      "step": 6039
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6324289441108704,
+      "learning_rate": 6.449380168274965e-05,
+      "loss": 3.2148,
+      "step": 6040
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6326262354850769,
+      "learning_rate": 6.442432980086466e-05,
+      "loss": 3.317,
+      "step": 6041
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8014912009239197,
+      "learning_rate": 6.43548898215888e-05,
+      "loss": 3.2295,
+      "step": 6042
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7135244607925415,
+      "learning_rate": 6.428548175685958e-05,
+      "loss": 3.3388,
+      "step": 6043
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6073774099349976,
+      "learning_rate": 6.421610561860902e-05,
+      "loss": 3.2621,
+      "step": 6044
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6199702024459839,
+      "learning_rate": 6.414676141876363e-05,
+      "loss": 3.3099,
+      "step": 6045
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.683590292930603,
+      "learning_rate": 6.407744916924463e-05,
+      "loss": 3.3776,
+      "step": 6046
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6720425486564636,
+      "learning_rate": 6.400816888196751e-05,
+      "loss": 3.1343,
+      "step": 6047
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6250442266464233,
+      "learning_rate": 6.393892056884234e-05,
+      "loss": 3.2088,
+      "step": 6048
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6554825305938721,
+      "learning_rate": 6.386970424177376e-05,
+      "loss": 3.3075,
+      "step": 6049
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6803299188613892,
+      "learning_rate": 6.380051991266075e-05,
+      "loss": 3.3917,
+      "step": 6050
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6328518390655518,
+      "learning_rate": 6.373136759339704e-05,
+      "loss": 3.3815,
+      "step": 6051
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6592798829078674,
+      "learning_rate": 6.366224729587067e-05,
+      "loss": 3.1769,
+      "step": 6052
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.627619743347168,
+      "learning_rate": 6.359315903196411e-05,
+      "loss": 3.2093,
+      "step": 6053
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6786208152770996,
+      "learning_rate": 6.352410281355461e-05,
+      "loss": 3.2703,
+      "step": 6054
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6732574701309204,
+      "learning_rate": 6.345507865251366e-05,
+      "loss": 3.3439,
+      "step": 6055
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5812044143676758,
+      "learning_rate": 6.338608656070727e-05,
+      "loss": 3.2492,
+      "step": 6056
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6608644723892212,
+      "learning_rate": 6.331712654999602e-05,
+      "loss": 3.2478,
+      "step": 6057
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.623185932636261,
+      "learning_rate": 6.324819863223497e-05,
+      "loss": 3.2782,
+      "step": 6058
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6139093637466431,
+      "learning_rate": 6.317930281927348e-05,
+      "loss": 3.3213,
+      "step": 6059
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6272095441818237,
+      "learning_rate": 6.311043912295563e-05,
+      "loss": 3.1978,
+      "step": 6060
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6622560620307922,
+      "learning_rate": 6.304160755512003e-05,
+      "loss": 3.3403,
+      "step": 6061
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6419727206230164,
+      "learning_rate": 6.297280812759945e-05,
+      "loss": 3.3154,
+      "step": 6062
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6320186853408813,
+      "learning_rate": 6.29040408522214e-05,
+      "loss": 3.2966,
+      "step": 6063
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6026878952980042,
+      "learning_rate": 6.28353057408077e-05,
+      "loss": 3.3277,
+      "step": 6064
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6492076516151428,
+      "learning_rate": 6.276660280517477e-05,
+      "loss": 3.2501,
+      "step": 6065
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6005073189735413,
+      "learning_rate": 6.269793205713331e-05,
+      "loss": 3.2953,
+      "step": 6066
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5965752005577087,
+      "learning_rate": 6.262929350848881e-05,
+      "loss": 3.2117,
+      "step": 6067
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6626747846603394,
+      "learning_rate": 6.256068717104093e-05,
+      "loss": 3.3177,
+      "step": 6068
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6364305019378662,
+      "learning_rate": 6.249211305658384e-05,
+      "loss": 3.1901,
+      "step": 6069
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6597917675971985,
+      "learning_rate": 6.242357117690639e-05,
+      "loss": 3.2494,
+      "step": 6070
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6223085522651672,
+      "learning_rate": 6.235506154379161e-05,
+      "loss": 3.2459,
+      "step": 6071
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5980460047721863,
+      "learning_rate": 6.228658416901711e-05,
+      "loss": 3.1854,
+      "step": 6072
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6632031798362732,
+      "learning_rate": 6.221813906435494e-05,
+      "loss": 3.2937,
+      "step": 6073
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6205509305000305,
+      "learning_rate": 6.214972624157159e-05,
+      "loss": 3.3785,
+      "step": 6074
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6477987766265869,
+      "learning_rate": 6.208134571242808e-05,
+      "loss": 3.1485,
+      "step": 6075
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6303578019142151,
+      "learning_rate": 6.20129974886798e-05,
+      "loss": 3.1722,
+      "step": 6076
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6248361468315125,
+      "learning_rate": 6.194468158207659e-05,
+      "loss": 3.263,
+      "step": 6077
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8012433648109436,
+      "learning_rate": 6.187639800436265e-05,
+      "loss": 3.2556,
+      "step": 6078
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6323918104171753,
+      "learning_rate": 6.18081467672769e-05,
+      "loss": 3.2504,
+      "step": 6079
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.626590371131897,
+      "learning_rate": 6.17399278825524e-05,
+      "loss": 3.3973,
+      "step": 6080
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6283361911773682,
+      "learning_rate": 6.167174136191675e-05,
+      "loss": 3.2398,
+      "step": 6081
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.636721134185791,
+      "learning_rate": 6.160358721709212e-05,
+      "loss": 3.2098,
+      "step": 6082
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6429245471954346,
+      "learning_rate": 6.15354654597949e-05,
+      "loss": 3.3347,
+      "step": 6083
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6272968649864197,
+      "learning_rate": 6.146737610173606e-05,
+      "loss": 3.3246,
+      "step": 6084
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6349840760231018,
+      "learning_rate": 6.13993191546209e-05,
+      "loss": 3.252,
+      "step": 6085
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6135838627815247,
+      "learning_rate": 6.133129463014924e-05,
+      "loss": 3.3089,
+      "step": 6086
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6382942199707031,
+      "learning_rate": 6.126330254001522e-05,
+      "loss": 3.2533,
+      "step": 6087
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.690299391746521,
+      "learning_rate": 6.119534289590747e-05,
+      "loss": 3.091,
+      "step": 6088
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6160559058189392,
+      "learning_rate": 6.112741570950919e-05,
+      "loss": 3.2888,
+      "step": 6089
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6733001470565796,
+      "learning_rate": 6.105952099249776e-05,
+      "loss": 3.2529,
+      "step": 6090
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5812699794769287,
+      "learning_rate": 6.099165875654503e-05,
+      "loss": 3.3015,
+      "step": 6091
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6092056035995483,
+      "learning_rate": 6.092382901331733e-05,
+      "loss": 3.1526,
+      "step": 6092
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6755577921867371,
+      "learning_rate": 6.0856031774475407e-05,
+      "loss": 3.3499,
+      "step": 6093
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.596678614616394,
+      "learning_rate": 6.078826705167437e-05,
+      "loss": 3.2249,
+      "step": 6094
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6615753173828125,
+      "learning_rate": 6.072053485656365e-05,
+      "loss": 3.3419,
+      "step": 6095
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6795618534088135,
+      "learning_rate": 6.065283520078732e-05,
+      "loss": 3.289,
+      "step": 6096
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6887994408607483,
+      "learning_rate": 6.05851680959838e-05,
+      "loss": 3.3519,
+      "step": 6097
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7297321557998657,
+      "learning_rate": 6.051753355378578e-05,
+      "loss": 3.1386,
+      "step": 6098
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6433356404304504,
+      "learning_rate": 6.044993158582038e-05,
+      "loss": 3.2025,
+      "step": 6099
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6871981024742126,
+      "learning_rate": 6.038236220370921e-05,
+      "loss": 3.3401,
+      "step": 6100
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6435525417327881,
+      "learning_rate": 6.031482541906821e-05,
+      "loss": 3.4581,
+      "step": 6101
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6278454661369324,
+      "learning_rate": 6.024732124350768e-05,
+      "loss": 3.3444,
+      "step": 6102
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6780098676681519,
+      "learning_rate": 6.017984968863249e-05,
+      "loss": 3.3399,
+      "step": 6103
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5871052742004395,
+      "learning_rate": 6.0112410766041735e-05,
+      "loss": 3.1699,
+      "step": 6104
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6621236205101013,
+      "learning_rate": 6.0045004487328917e-05,
+      "loss": 3.2539,
+      "step": 6105
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6626326441764832,
+      "learning_rate": 5.997763086408192e-05,
+      "loss": 3.273,
+      "step": 6106
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5612792372703552,
+      "learning_rate": 5.991028990788316e-05,
+      "loss": 3.2697,
+      "step": 6107
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6139649748802185,
+      "learning_rate": 5.984298163030929e-05,
+      "loss": 3.2308,
+      "step": 6108
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6269494891166687,
+      "learning_rate": 5.977570604293128e-05,
+      "loss": 3.3231,
+      "step": 6109
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6203356981277466,
+      "learning_rate": 5.9708463157314765e-05,
+      "loss": 3.3443,
+      "step": 6110
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6309711337089539,
+      "learning_rate": 5.964125298501946e-05,
+      "loss": 3.2053,
+      "step": 6111
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7104734182357788,
+      "learning_rate": 5.957407553759961e-05,
+      "loss": 3.2847,
+      "step": 6112
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6571914553642273,
+      "learning_rate": 5.950693082660377e-05,
+      "loss": 3.3378,
+      "step": 6113
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6446319818496704,
+      "learning_rate": 5.9439818863574913e-05,
+      "loss": 3.1806,
+      "step": 6114
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.632908284664154,
+      "learning_rate": 5.937273966005028e-05,
+      "loss": 3.284,
+      "step": 6115
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6450886726379395,
+      "learning_rate": 5.9305693227561715e-05,
+      "loss": 3.2289,
+      "step": 6116
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6438875198364258,
+      "learning_rate": 5.923867957763512e-05,
+      "loss": 3.2964,
+      "step": 6117
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.632624089717865,
+      "learning_rate": 5.917169872179109e-05,
+      "loss": 3.3564,
+      "step": 6118
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6316910982131958,
+      "learning_rate": 5.91047506715443e-05,
+      "loss": 3.3003,
+      "step": 6119
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6209931373596191,
+      "learning_rate": 5.903783543840393e-05,
+      "loss": 3.2094,
+      "step": 6120
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5965805053710938,
+      "learning_rate": 5.8970953033873484e-05,
+      "loss": 3.1804,
+      "step": 6121
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6775824427604675,
+      "learning_rate": 5.89041034694508e-05,
+      "loss": 3.3426,
+      "step": 6122
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6145557165145874,
+      "learning_rate": 5.8837286756628025e-05,
+      "loss": 3.2335,
+      "step": 6123
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6069118976593018,
+      "learning_rate": 5.877050290689182e-05,
+      "loss": 3.1969,
+      "step": 6124
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6316295266151428,
+      "learning_rate": 5.870375193172314e-05,
+      "loss": 3.3389,
+      "step": 6125
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.654758870601654,
+      "learning_rate": 5.8637033842597224e-05,
+      "loss": 3.2345,
+      "step": 6126
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6531403064727783,
+      "learning_rate": 5.857034865098365e-05,
+      "loss": 3.2278,
+      "step": 6127
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6327181458473206,
+      "learning_rate": 5.8503696368346374e-05,
+      "loss": 3.1833,
+      "step": 6128
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6120930314064026,
+      "learning_rate": 5.84370770061437e-05,
+      "loss": 3.3167,
+      "step": 6129
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6629424095153809,
+      "learning_rate": 5.8370490575828216e-05,
+      "loss": 3.2911,
+      "step": 6130
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6397058367729187,
+      "learning_rate": 5.8303937088847045e-05,
+      "loss": 3.2963,
+      "step": 6131
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6212075352668762,
+      "learning_rate": 5.823741655664141e-05,
+      "loss": 3.4139,
+      "step": 6132
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.635893702507019,
+      "learning_rate": 5.817092899064691e-05,
+      "loss": 3.398,
+      "step": 6133
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6343691945075989,
+      "learning_rate": 5.810447440229366e-05,
+      "loss": 3.2888,
+      "step": 6134
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6382713913917542,
+      "learning_rate": 5.8038052803005894e-05,
+      "loss": 3.263,
+      "step": 6135
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6321362257003784,
+      "learning_rate": 5.797166420420228e-05,
+      "loss": 3.327,
+      "step": 6136
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.631368100643158,
+      "learning_rate": 5.790530861729576e-05,
+      "loss": 3.2032,
+      "step": 6137
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5903132557868958,
+      "learning_rate": 5.783898605369356e-05,
+      "loss": 3.2422,
+      "step": 6138
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6581616401672363,
+      "learning_rate": 5.777269652479747e-05,
+      "loss": 3.2814,
+      "step": 6139
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.638910710811615,
+      "learning_rate": 5.770644004200332e-05,
+      "loss": 3.2615,
+      "step": 6140
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6400853991508484,
+      "learning_rate": 5.764021661670141e-05,
+      "loss": 3.2304,
+      "step": 6141
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6094925999641418,
+      "learning_rate": 5.757402626027625e-05,
+      "loss": 3.2817,
+      "step": 6142
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7051050066947937,
+      "learning_rate": 5.750786898410673e-05,
+      "loss": 3.2862,
+      "step": 6143
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6410810351371765,
+      "learning_rate": 5.744174479956615e-05,
+      "loss": 3.3413,
+      "step": 6144
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6460927724838257,
+      "learning_rate": 5.737565371802189e-05,
+      "loss": 3.2164,
+      "step": 6145
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6438145041465759,
+      "learning_rate": 5.730959575083594e-05,
+      "loss": 3.2533,
+      "step": 6146
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6696481108665466,
+      "learning_rate": 5.7243570909364326e-05,
+      "loss": 3.2115,
+      "step": 6147
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6192052960395813,
+      "learning_rate": 5.71775792049575e-05,
+      "loss": 3.2049,
+      "step": 6148
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5886045694351196,
+      "learning_rate": 5.711162064896019e-05,
+      "loss": 3.2451,
+      "step": 6149
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6182309985160828,
+      "learning_rate": 5.7045695252711475e-05,
+      "loss": 3.1899,
+      "step": 6150
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6360798478126526,
+      "learning_rate": 5.697980302754458e-05,
+      "loss": 3.3021,
+      "step": 6151
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6290764808654785,
+      "learning_rate": 5.691394398478727e-05,
+      "loss": 3.2406,
+      "step": 6152
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6354265213012695,
+      "learning_rate": 5.684811813576138e-05,
+      "loss": 3.3239,
+      "step": 6153
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.659576952457428,
+      "learning_rate": 5.678232549178328e-05,
+      "loss": 3.4077,
+      "step": 6154
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6638350486755371,
+      "learning_rate": 5.671656606416337e-05,
+      "loss": 3.2351,
+      "step": 6155
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.657824695110321,
+      "learning_rate": 5.6650839864206496e-05,
+      "loss": 3.371,
+      "step": 6156
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.648253858089447,
+      "learning_rate": 5.658514690321176e-05,
+      "loss": 3.2927,
+      "step": 6157
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6625542044639587,
+      "learning_rate": 5.651948719247252e-05,
+      "loss": 3.1641,
+      "step": 6158
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6146804094314575,
+      "learning_rate": 5.645386074327638e-05,
+      "loss": 3.2208,
+      "step": 6159
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7196055054664612,
+      "learning_rate": 5.63882675669054e-05,
+      "loss": 3.2925,
+      "step": 6160
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6016031503677368,
+      "learning_rate": 5.6322707674635736e-05,
+      "loss": 3.3658,
+      "step": 6161
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6775092482566833,
+      "learning_rate": 5.6257181077737975e-05,
+      "loss": 3.2381,
+      "step": 6162
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6624186635017395,
+      "learning_rate": 5.619168778747685e-05,
+      "loss": 3.2365,
+      "step": 6163
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6462388634681702,
+      "learning_rate": 5.6126227815111425e-05,
+      "loss": 3.3843,
+      "step": 6164
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6163438558578491,
+      "learning_rate": 5.6060801171894996e-05,
+      "loss": 3.2424,
+      "step": 6165
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5882861018180847,
+      "learning_rate": 5.599540786907512e-05,
+      "loss": 3.446,
+      "step": 6166
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6447424292564392,
+      "learning_rate": 5.59300479178938e-05,
+      "loss": 3.2844,
+      "step": 6167
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5792222023010254,
+      "learning_rate": 5.5864721329587084e-05,
+      "loss": 3.1562,
+      "step": 6168
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6093016266822815,
+      "learning_rate": 5.579942811538538e-05,
+      "loss": 3.2774,
+      "step": 6169
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6209796071052551,
+      "learning_rate": 5.573416828651329e-05,
+      "loss": 3.4412,
+      "step": 6170
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.62843257188797,
+      "learning_rate": 5.566894185418986e-05,
+      "loss": 3.3524,
+      "step": 6171
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6866121292114258,
+      "learning_rate": 5.560374882962821e-05,
+      "loss": 3.3448,
+      "step": 6172
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6563045382499695,
+      "learning_rate": 5.553858922403576e-05,
+      "loss": 3.2293,
+      "step": 6173
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6468909382820129,
+      "learning_rate": 5.5473463048614144e-05,
+      "loss": 3.3161,
+      "step": 6174
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6102748513221741,
+      "learning_rate": 5.540837031455945e-05,
+      "loss": 3.3122,
+      "step": 6175
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6245465278625488,
+      "learning_rate": 5.534331103306181e-05,
+      "loss": 3.2799,
+      "step": 6176
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6340338587760925,
+      "learning_rate": 5.527828521530562e-05,
+      "loss": 3.3966,
+      "step": 6177
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6553022265434265,
+      "learning_rate": 5.521329287246965e-05,
+      "loss": 3.1975,
+      "step": 6178
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6260964274406433,
+      "learning_rate": 5.5148334015726774e-05,
+      "loss": 3.181,
+      "step": 6179
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6341874003410339,
+      "learning_rate": 5.5083408656244146e-05,
+      "loss": 3.1921,
+      "step": 6180
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6693719029426575,
+      "learning_rate": 5.501851680518322e-05,
+      "loss": 3.3772,
+      "step": 6181
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6416365504264832,
+      "learning_rate": 5.4953658473699734e-05,
+      "loss": 3.3323,
+      "step": 6182
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6330123543739319,
+      "learning_rate": 5.488883367294353e-05,
+      "loss": 3.3311,
+      "step": 6183
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6906682252883911,
+      "learning_rate": 5.482404241405875e-05,
+      "loss": 3.2435,
+      "step": 6184
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6447123289108276,
+      "learning_rate": 5.4759284708183755e-05,
+      "loss": 3.2492,
+      "step": 6185
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6190230250358582,
+      "learning_rate": 5.469456056645114e-05,
+      "loss": 3.276,
+      "step": 6186
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.60328209400177,
+      "learning_rate": 5.4629869999987675e-05,
+      "loss": 3.2909,
+      "step": 6187
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6737046837806702,
+      "learning_rate": 5.4565213019914544e-05,
+      "loss": 3.3374,
+      "step": 6188
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6252620816230774,
+      "learning_rate": 5.450058963734691e-05,
+      "loss": 3.3156,
+      "step": 6189
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6689559817314148,
+      "learning_rate": 5.44359998633944e-05,
+      "loss": 3.4685,
+      "step": 6190
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6267184019088745,
+      "learning_rate": 5.437144370916069e-05,
+      "loss": 3.3537,
+      "step": 6191
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6087378859519958,
+      "learning_rate": 5.430692118574374e-05,
+      "loss": 3.3843,
+      "step": 6192
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6335881352424622,
+      "learning_rate": 5.424243230423567e-05,
+      "loss": 3.2503,
+      "step": 6193
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6103877425193787,
+      "learning_rate": 5.417797707572294e-05,
+      "loss": 3.1548,
+      "step": 6194
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6789931058883667,
+      "learning_rate": 5.411355551128602e-05,
+      "loss": 3.2277,
+      "step": 6195
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6586472392082214,
+      "learning_rate": 5.404916762199991e-05,
+      "loss": 3.3497,
+      "step": 6196
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6357256174087524,
+      "learning_rate": 5.3984813418933545e-05,
+      "loss": 3.3711,
+      "step": 6197
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6573581695556641,
+      "learning_rate": 5.3920492913150074e-05,
+      "loss": 3.2539,
+      "step": 6198
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6482775211334229,
+      "learning_rate": 5.3856206115707077e-05,
+      "loss": 3.1406,
+      "step": 6199
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6598178148269653,
+      "learning_rate": 5.379195303765616e-05,
+      "loss": 3.2314,
+      "step": 6200
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6228412985801697,
+      "learning_rate": 5.372773369004314e-05,
+      "loss": 3.1364,
+      "step": 6201
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7382637858390808,
+      "learning_rate": 5.366354808390803e-05,
+      "loss": 3.236,
+      "step": 6202
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6506236791610718,
+      "learning_rate": 5.35993962302852e-05,
+      "loss": 3.1999,
+      "step": 6203
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6304946541786194,
+      "learning_rate": 5.353527814020301e-05,
+      "loss": 3.3298,
+      "step": 6204
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6180459260940552,
+      "learning_rate": 5.347119382468413e-05,
+      "loss": 3.2057,
+      "step": 6205
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6518542766571045,
+      "learning_rate": 5.340714329474541e-05,
+      "loss": 3.1527,
+      "step": 6206
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6927798390388489,
+      "learning_rate": 5.3343126561397756e-05,
+      "loss": 3.2682,
+      "step": 6207
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6122478246688843,
+      "learning_rate": 5.327914363564659e-05,
+      "loss": 3.3978,
+      "step": 6208
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.607306182384491,
+      "learning_rate": 5.3215194528491186e-05,
+      "loss": 3.2244,
+      "step": 6209
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7095955610275269,
+      "learning_rate": 5.315127925092511e-05,
+      "loss": 3.3604,
+      "step": 6210
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6393181681632996,
+      "learning_rate": 5.308739781393629e-05,
+      "loss": 3.142,
+      "step": 6211
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7591974139213562,
+      "learning_rate": 5.302355022850655e-05,
+      "loss": 3.3069,
+      "step": 6212
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6667575836181641,
+      "learning_rate": 5.295973650561212e-05,
+      "loss": 3.3487,
+      "step": 6213
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6396272778511047,
+      "learning_rate": 5.289595665622324e-05,
+      "loss": 3.2997,
+      "step": 6214
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6105965375900269,
+      "learning_rate": 5.283221069130442e-05,
+      "loss": 3.4331,
+      "step": 6215
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6773115992546082,
+      "learning_rate": 5.276849862181432e-05,
+      "loss": 3.3996,
+      "step": 6216
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6334084868431091,
+      "learning_rate": 5.2704820458705785e-05,
+      "loss": 3.1817,
+      "step": 6217
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6564429402351379,
+      "learning_rate": 5.264117621292594e-05,
+      "loss": 3.1336,
+      "step": 6218
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.631753146648407,
+      "learning_rate": 5.2577565895415876e-05,
+      "loss": 3.2437,
+      "step": 6219
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6693889498710632,
+      "learning_rate": 5.251398951711095e-05,
+      "loss": 3.3941,
+      "step": 6220
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6456241607666016,
+      "learning_rate": 5.245044708894067e-05,
+      "loss": 3.2437,
+      "step": 6221
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6354429125785828,
+      "learning_rate": 5.238693862182875e-05,
+      "loss": 3.1984,
+      "step": 6222
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6462532877922058,
+      "learning_rate": 5.232346412669292e-05,
+      "loss": 3.4046,
+      "step": 6223
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6371812224388123,
+      "learning_rate": 5.2260023614445344e-05,
+      "loss": 3.3915,
+      "step": 6224
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6222655773162842,
+      "learning_rate": 5.2196617095992103e-05,
+      "loss": 3.273,
+      "step": 6225
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.618720531463623,
+      "learning_rate": 5.213324458223345e-05,
+      "loss": 3.3342,
+      "step": 6226
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6369234919548035,
+      "learning_rate": 5.2069906084064006e-05,
+      "loss": 3.3978,
+      "step": 6227
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6682491302490234,
+      "learning_rate": 5.200660161237228e-05,
+      "loss": 3.1945,
+      "step": 6228
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6570479273796082,
+      "learning_rate": 5.1943331178041094e-05,
+      "loss": 3.2701,
+      "step": 6229
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6652297973632812,
+      "learning_rate": 5.188009479194736e-05,
+      "loss": 3.3443,
+      "step": 6230
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6708425283432007,
+      "learning_rate": 5.1816892464962046e-05,
+      "loss": 3.3297,
+      "step": 6231
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.653319776058197,
+      "learning_rate": 5.1753724207950517e-05,
+      "loss": 3.2837,
+      "step": 6232
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6871150732040405,
+      "learning_rate": 5.1690590031772066e-05,
+      "loss": 3.3362,
+      "step": 6233
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6416165828704834,
+      "learning_rate": 5.16274899472802e-05,
+      "loss": 3.2679,
+      "step": 6234
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6724268198013306,
+      "learning_rate": 5.156442396532246e-05,
+      "loss": 3.3087,
+      "step": 6235
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5732466578483582,
+      "learning_rate": 5.150139209674079e-05,
+      "loss": 3.2837,
+      "step": 6236
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6079490780830383,
+      "learning_rate": 5.1438394352370994e-05,
+      "loss": 3.3867,
+      "step": 6237
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6082061529159546,
+      "learning_rate": 5.137543074304307e-05,
+      "loss": 3.4159,
+      "step": 6238
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6031527519226074,
+      "learning_rate": 5.131250127958134e-05,
+      "loss": 3.2258,
+      "step": 6239
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.68482905626297,
+      "learning_rate": 5.124960597280401e-05,
+      "loss": 3.3032,
+      "step": 6240
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6390005946159363,
+      "learning_rate": 5.1186744833523533e-05,
+      "loss": 3.3787,
+      "step": 6241
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6423435807228088,
+      "learning_rate": 5.112391787254645e-05,
+      "loss": 3.3118,
+      "step": 6242
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6239182353019714,
+      "learning_rate": 5.106112510067346e-05,
+      "loss": 3.25,
+      "step": 6243
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6259147524833679,
+      "learning_rate": 5.0998366528699266e-05,
+      "loss": 3.2261,
+      "step": 6244
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6671559810638428,
+      "learning_rate": 5.0935642167413e-05,
+      "loss": 3.5002,
+      "step": 6245
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6126053333282471,
+      "learning_rate": 5.087295202759751e-05,
+      "loss": 3.4132,
+      "step": 6246
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6345648765563965,
+      "learning_rate": 5.081029612003013e-05,
+      "loss": 3.3136,
+      "step": 6247
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6129223704338074,
+      "learning_rate": 5.074767445548204e-05,
+      "loss": 3.3363,
+      "step": 6248
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6308622360229492,
+      "learning_rate": 5.0685087044718635e-05,
+      "loss": 3.3067,
+      "step": 6249
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5842102766036987,
+      "learning_rate": 5.062253389849941e-05,
+      "loss": 3.1752,
+      "step": 6250
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.62709641456604,
+      "learning_rate": 5.056001502757801e-05,
+      "loss": 3.3369,
+      "step": 6251
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6009126901626587,
+      "learning_rate": 5.0497530442702086e-05,
+      "loss": 3.2501,
+      "step": 6252
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6700008511543274,
+      "learning_rate": 5.043508015461354e-05,
+      "loss": 3.3443,
+      "step": 6253
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.66363126039505,
+      "learning_rate": 5.037266417404823e-05,
+      "loss": 3.211,
+      "step": 6254
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.669608473777771,
+      "learning_rate": 5.031028251173628e-05,
+      "loss": 3.1384,
+      "step": 6255
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6509547233581543,
+      "learning_rate": 5.024793517840176e-05,
+      "loss": 3.2467,
+      "step": 6256
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5959102511405945,
+      "learning_rate": 5.018562218476294e-05,
+      "loss": 3.2967,
+      "step": 6257
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.650132954120636,
+      "learning_rate": 5.012334354153208e-05,
+      "loss": 3.2285,
+      "step": 6258
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5863680243492126,
+      "learning_rate": 5.00610992594156e-05,
+      "loss": 3.3313,
+      "step": 6259
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6746791005134583,
+      "learning_rate": 4.999888934911409e-05,
+      "loss": 3.2599,
+      "step": 6260
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6146058440208435,
+      "learning_rate": 4.993671382132212e-05,
+      "loss": 3.2519,
+      "step": 6261
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6945891976356506,
+      "learning_rate": 4.987457268672837e-05,
+      "loss": 3.3743,
+      "step": 6262
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6706831455230713,
+      "learning_rate": 4.981246595601555e-05,
+      "loss": 3.2523,
+      "step": 6263
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6371322870254517,
+      "learning_rate": 4.9750393639860694e-05,
+      "loss": 3.2625,
+      "step": 6264
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6765096187591553,
+      "learning_rate": 4.9688355748934654e-05,
+      "loss": 3.2962,
+      "step": 6265
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6514853835105896,
+      "learning_rate": 4.9626352293902474e-05,
+      "loss": 3.4608,
+      "step": 6266
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6448090076446533,
+      "learning_rate": 4.9564383285423204e-05,
+      "loss": 3.3251,
+      "step": 6267
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6782955527305603,
+      "learning_rate": 4.950244873415016e-05,
+      "loss": 3.2665,
+      "step": 6268
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6292037963867188,
+      "learning_rate": 4.9440548650730555e-05,
+      "loss": 3.3043,
+      "step": 6269
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6162392497062683,
+      "learning_rate": 4.93786830458057e-05,
+      "loss": 3.2856,
+      "step": 6270
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6886681914329529,
+      "learning_rate": 4.931685193001106e-05,
+      "loss": 3.2479,
+      "step": 6271
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6889098882675171,
+      "learning_rate": 4.9255055313976003e-05,
+      "loss": 3.4246,
+      "step": 6272
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6550235152244568,
+      "learning_rate": 4.9193293208324246e-05,
+      "loss": 3.3797,
+      "step": 6273
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6581763625144958,
+      "learning_rate": 4.9131565623673266e-05,
+      "loss": 3.3268,
+      "step": 6274
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6184753775596619,
+      "learning_rate": 4.906987257063489e-05,
+      "loss": 3.3674,
+      "step": 6275
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6205751299858093,
+      "learning_rate": 4.900821405981482e-05,
+      "loss": 3.3093,
+      "step": 6276
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6369295120239258,
+      "learning_rate": 4.894659010181282e-05,
+      "loss": 3.1849,
+      "step": 6277
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.691066324710846,
+      "learning_rate": 4.888500070722282e-05,
+      "loss": 3.3285,
+      "step": 6278
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6090084314346313,
+      "learning_rate": 4.882344588663271e-05,
+      "loss": 3.2637,
+      "step": 6279
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6285091638565063,
+      "learning_rate": 4.87619256506244e-05,
+      "loss": 3.2199,
+      "step": 6280
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6004552245140076,
+      "learning_rate": 4.870044000977406e-05,
+      "loss": 3.2288,
+      "step": 6281
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6280658841133118,
+      "learning_rate": 4.86389889746518e-05,
+      "loss": 3.1535,
+      "step": 6282
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6477638483047485,
+      "learning_rate": 4.857757255582171e-05,
+      "loss": 3.2006,
+      "step": 6283
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6137582063674927,
+      "learning_rate": 4.851619076384201e-05,
+      "loss": 3.3103,
+      "step": 6284
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6008321642875671,
+      "learning_rate": 4.845484360926489e-05,
+      "loss": 3.2552,
+      "step": 6285
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6434144973754883,
+      "learning_rate": 4.8393531102636664e-05,
+      "loss": 3.3374,
+      "step": 6286
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6633194088935852,
+      "learning_rate": 4.8332253254497665e-05,
+      "loss": 3.3185,
+      "step": 6287
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7374275326728821,
+      "learning_rate": 4.8271010075382205e-05,
+      "loss": 3.1933,
+      "step": 6288
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6177479028701782,
+      "learning_rate": 4.8209801575818835e-05,
+      "loss": 3.1641,
+      "step": 6289
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6840469837188721,
+      "learning_rate": 4.8148627766329914e-05,
+      "loss": 3.4058,
+      "step": 6290
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6524780988693237,
+      "learning_rate": 4.808748865743187e-05,
+      "loss": 3.3234,
+      "step": 6291
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6769046783447266,
+      "learning_rate": 4.802638425963537e-05,
+      "loss": 3.3129,
+      "step": 6292
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6700226068496704,
+      "learning_rate": 4.796531458344491e-05,
+      "loss": 3.2718,
+      "step": 6293
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6332312226295471,
+      "learning_rate": 4.790427963935903e-05,
+      "loss": 3.314,
+      "step": 6294
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6040202379226685,
+      "learning_rate": 4.784327943787034e-05,
+      "loss": 3.2749,
+      "step": 6295
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6835915446281433,
+      "learning_rate": 4.7782313989465556e-05,
+      "loss": 3.3133,
+      "step": 6296
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6584433317184448,
+      "learning_rate": 4.772138330462533e-05,
+      "loss": 3.1339,
+      "step": 6297
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6218266487121582,
+      "learning_rate": 4.766048739382431e-05,
+      "loss": 3.2933,
+      "step": 6298
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6843163371086121,
+      "learning_rate": 4.7599626267531225e-05,
+      "loss": 3.3426,
+      "step": 6299
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6411788463592529,
+      "learning_rate": 4.753879993620877e-05,
+      "loss": 3.2508,
+      "step": 6300
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6439712643623352,
+      "learning_rate": 4.7478008410313775e-05,
+      "loss": 3.1786,
+      "step": 6301
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6473103761672974,
+      "learning_rate": 4.741725170029693e-05,
+      "loss": 3.3076,
+      "step": 6302
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6452195644378662,
+      "learning_rate": 4.73565298166031e-05,
+      "loss": 3.3539,
+      "step": 6303
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6400468945503235,
+      "learning_rate": 4.7295842769671053e-05,
+      "loss": 3.2893,
+      "step": 6304
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6331076622009277,
+      "learning_rate": 4.723519056993358e-05,
+      "loss": 3.2634,
+      "step": 6305
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6293200850486755,
+      "learning_rate": 4.717457322781749e-05,
+      "loss": 3.2414,
+      "step": 6306
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7003605365753174,
+      "learning_rate": 4.711399075374362e-05,
+      "loss": 3.203,
+      "step": 6307
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6010873317718506,
+      "learning_rate": 4.705344315812682e-05,
+      "loss": 3.2742,
+      "step": 6308
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6363465189933777,
+      "learning_rate": 4.699293045137582e-05,
+      "loss": 3.3849,
+      "step": 6309
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6699336767196655,
+      "learning_rate": 4.693245264389351e-05,
+      "loss": 3.36,
+      "step": 6310
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6721777319908142,
+      "learning_rate": 4.687200974607683e-05,
+      "loss": 3.4133,
+      "step": 6311
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6349652409553528,
+      "learning_rate": 4.6811601768316556e-05,
+      "loss": 3.4149,
+      "step": 6312
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6694733500480652,
+      "learning_rate": 4.675122872099749e-05,
+      "loss": 3.2205,
+      "step": 6313
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5913556814193726,
+      "learning_rate": 4.6690890614498466e-05,
+      "loss": 3.2009,
+      "step": 6314
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6164645552635193,
+      "learning_rate": 4.66305874591923e-05,
+      "loss": 3.3531,
+      "step": 6315
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6450039148330688,
+      "learning_rate": 4.657031926544575e-05,
+      "loss": 3.3254,
+      "step": 6316
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6537266373634338,
+      "learning_rate": 4.651008604361975e-05,
+      "loss": 3.2913,
+      "step": 6317
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.65716952085495,
+      "learning_rate": 4.6449887804068944e-05,
+      "loss": 3.3379,
+      "step": 6318
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6729187369346619,
+      "learning_rate": 4.638972455714224e-05,
+      "loss": 3.2672,
+      "step": 6319
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6832636594772339,
+      "learning_rate": 4.632959631318234e-05,
+      "loss": 3.2275,
+      "step": 6320
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6566662192344666,
+      "learning_rate": 4.6269503082526e-05,
+      "loss": 3.2217,
+      "step": 6321
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.671359658241272,
+      "learning_rate": 4.620944487550391e-05,
+      "loss": 3.4067,
+      "step": 6322
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.677776038646698,
+      "learning_rate": 4.6149421702440743e-05,
+      "loss": 3.2207,
+      "step": 6323
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6239826679229736,
+      "learning_rate": 4.6089433573655276e-05,
+      "loss": 3.231,
+      "step": 6324
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6734199523925781,
+      "learning_rate": 4.6029480499460095e-05,
+      "loss": 3.268,
+      "step": 6325
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6146829128265381,
+      "learning_rate": 4.596956249016188e-05,
+      "loss": 3.1548,
+      "step": 6326
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6150223612785339,
+      "learning_rate": 4.590967955606115e-05,
+      "loss": 3.1679,
+      "step": 6327
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.690636396408081,
+      "learning_rate": 4.5849831707452496e-05,
+      "loss": 3.2218,
+      "step": 6328
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6366089582443237,
+      "learning_rate": 4.579001895462453e-05,
+      "loss": 3.2026,
+      "step": 6329
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6616598963737488,
+      "learning_rate": 4.573024130785972e-05,
+      "loss": 3.2072,
+      "step": 6330
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6787472367286682,
+      "learning_rate": 4.5670498777434456e-05,
+      "loss": 3.2212,
+      "step": 6331
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6337720155715942,
+      "learning_rate": 4.561079137361932e-05,
+      "loss": 3.2873,
+      "step": 6332
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.680767834186554,
+      "learning_rate": 4.555111910667861e-05,
+      "loss": 3.4622,
+      "step": 6333
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5902546048164368,
+      "learning_rate": 4.5491481986870726e-05,
+      "loss": 3.0807,
+      "step": 6334
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6405333280563354,
+      "learning_rate": 4.543188002444795e-05,
+      "loss": 3.3239,
+      "step": 6335
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6532682180404663,
+      "learning_rate": 4.537231322965654e-05,
+      "loss": 3.3386,
+      "step": 6336
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6565823554992676,
+      "learning_rate": 4.531278161273667e-05,
+      "loss": 3.2401,
+      "step": 6337
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6898584961891174,
+      "learning_rate": 4.5253285183922574e-05,
+      "loss": 3.4491,
+      "step": 6338
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6294739246368408,
+      "learning_rate": 4.519382395344246e-05,
+      "loss": 3.2714,
+      "step": 6339
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.673034131526947,
+      "learning_rate": 4.513439793151833e-05,
+      "loss": 3.372,
+      "step": 6340
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6405984163284302,
+      "learning_rate": 4.507500712836621e-05,
+      "loss": 3.3182,
+      "step": 6341
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.619519054889679,
+      "learning_rate": 4.5015651554196015e-05,
+      "loss": 3.3016,
+      "step": 6342
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5913485288619995,
+      "learning_rate": 4.495633121921175e-05,
+      "loss": 3.3034,
+      "step": 6343
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6833899021148682,
+      "learning_rate": 4.489704613361112e-05,
+      "loss": 3.401,
+      "step": 6344
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7165760397911072,
+      "learning_rate": 4.4837796307586085e-05,
+      "loss": 3.3631,
+      "step": 6345
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6703233122825623,
+      "learning_rate": 4.477858175132227e-05,
+      "loss": 3.1297,
+      "step": 6346
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6346496939659119,
+      "learning_rate": 4.4719402474999424e-05,
+      "loss": 3.2481,
+      "step": 6347
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6317062973976135,
+      "learning_rate": 4.4660258488791125e-05,
+      "loss": 3.2427,
+      "step": 6348
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6202000379562378,
+      "learning_rate": 4.4601149802864864e-05,
+      "loss": 3.3742,
+      "step": 6349
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.655680775642395,
+      "learning_rate": 4.454207642738217e-05,
+      "loss": 3.212,
+      "step": 6350
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6488626599311829,
+      "learning_rate": 4.4483038372498397e-05,
+      "loss": 3.3244,
+      "step": 6351
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6419368386268616,
+      "learning_rate": 4.4424035648362836e-05,
+      "loss": 3.3423,
+      "step": 6352
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6373838186264038,
+      "learning_rate": 4.4365068265118825e-05,
+      "loss": 3.2937,
+      "step": 6353
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6217588186264038,
+      "learning_rate": 4.4306136232903524e-05,
+      "loss": 3.2487,
+      "step": 6354
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6756614446640015,
+      "learning_rate": 4.424723956184795e-05,
+      "loss": 3.2515,
+      "step": 6355
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.641849935054779,
+      "learning_rate": 4.418837826207725e-05,
+      "loss": 3.3007,
+      "step": 6356
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6056436896324158,
+      "learning_rate": 4.41295523437103e-05,
+      "loss": 3.2913,
+      "step": 6357
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6423781514167786,
+      "learning_rate": 4.407076181685996e-05,
+      "loss": 3.2333,
+      "step": 6358
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6782287955284119,
+      "learning_rate": 4.401200669163291e-05,
+      "loss": 3.4017,
+      "step": 6359
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5897096395492554,
+      "learning_rate": 4.3953286978130005e-05,
+      "loss": 3.1897,
+      "step": 6360
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6132126450538635,
+      "learning_rate": 4.389460268644577e-05,
+      "loss": 3.3616,
+      "step": 6361
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6592351198196411,
+      "learning_rate": 4.383595382666872e-05,
+      "loss": 3.2439,
+      "step": 6362
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6139822602272034,
+      "learning_rate": 4.3777340408881263e-05,
+      "loss": 3.2914,
+      "step": 6363
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6376251578330994,
+      "learning_rate": 4.3718762443159725e-05,
+      "loss": 3.3195,
+      "step": 6364
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6536754965782166,
+      "learning_rate": 4.366021993957428e-05,
+      "loss": 3.314,
+      "step": 6365
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6610123515129089,
+      "learning_rate": 4.360171290818918e-05,
+      "loss": 3.3501,
+      "step": 6366
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6074190735816956,
+      "learning_rate": 4.354324135906235e-05,
+      "loss": 3.2911,
+      "step": 6367
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6405759453773499,
+      "learning_rate": 4.3484805302245826e-05,
+      "loss": 3.2542,
+      "step": 6368
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6312056183815002,
+      "learning_rate": 4.342640474778542e-05,
+      "loss": 3.2063,
+      "step": 6369
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6049206256866455,
+      "learning_rate": 4.3368039705720844e-05,
+      "loss": 3.3249,
+      "step": 6370
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6044682264328003,
+      "learning_rate": 4.33097101860857e-05,
+      "loss": 3.306,
+      "step": 6371
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6150915622711182,
+      "learning_rate": 4.3251416198907576e-05,
+      "loss": 3.3062,
+      "step": 6372
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6390911936759949,
+      "learning_rate": 4.319315775420776e-05,
+      "loss": 3.2433,
+      "step": 6373
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5915288329124451,
+      "learning_rate": 4.3134934862001624e-05,
+      "loss": 3.2095,
+      "step": 6374
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6328829526901245,
+      "learning_rate": 4.307674753229846e-05,
+      "loss": 3.4263,
+      "step": 6375
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6330004930496216,
+      "learning_rate": 4.301859577510123e-05,
+      "loss": 3.1894,
+      "step": 6376
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6698534488677979,
+      "learning_rate": 4.2960479600406917e-05,
+      "loss": 3.3107,
+      "step": 6377
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6564880609512329,
+      "learning_rate": 4.290239901820639e-05,
+      "loss": 3.346,
+      "step": 6378
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5670497417449951,
+      "learning_rate": 4.284435403848436e-05,
+      "loss": 3.2161,
+      "step": 6379
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5879086852073669,
+      "learning_rate": 4.2786344671219334e-05,
+      "loss": 3.2026,
+      "step": 6380
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6338858008384705,
+      "learning_rate": 4.2728370926383956e-05,
+      "loss": 3.22,
+      "step": 6381
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6421663761138916,
+      "learning_rate": 4.267043281394453e-05,
+      "loss": 3.3388,
+      "step": 6382
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6467295289039612,
+      "learning_rate": 4.26125303438612e-05,
+      "loss": 3.3544,
+      "step": 6383
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6295877695083618,
+      "learning_rate": 4.255466352608822e-05,
+      "loss": 3.2793,
+      "step": 6384
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6589932441711426,
+      "learning_rate": 4.2496832370573476e-05,
+      "loss": 3.2592,
+      "step": 6385
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6921527981758118,
+      "learning_rate": 4.2439036887258837e-05,
+      "loss": 3.1894,
+      "step": 6386
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6357716917991638,
+      "learning_rate": 4.2381277086080026e-05,
+      "loss": 3.2358,
+      "step": 6387
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6349096298217773,
+      "learning_rate": 4.2323552976966525e-05,
+      "loss": 3.3883,
+      "step": 6388
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6470059156417847,
+      "learning_rate": 4.226586456984191e-05,
+      "loss": 3.1993,
+      "step": 6389
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6028220653533936,
+      "learning_rate": 4.220821187462345e-05,
+      "loss": 3.1472,
+      "step": 6390
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6410771608352661,
+      "learning_rate": 4.2150594901222306e-05,
+      "loss": 3.2464,
+      "step": 6391
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6089470386505127,
+      "learning_rate": 4.209301365954343e-05,
+      "loss": 3.2082,
+      "step": 6392
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6100988984107971,
+      "learning_rate": 4.2035468159485835e-05,
+      "loss": 3.3006,
+      "step": 6393
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6550276875495911,
+      "learning_rate": 4.197795841094221e-05,
+      "loss": 3.2269,
+      "step": 6394
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6482765078544617,
+      "learning_rate": 4.192048442379903e-05,
+      "loss": 3.4126,
+      "step": 6395
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6182270646095276,
+      "learning_rate": 4.1863046207936934e-05,
+      "loss": 3.3065,
+      "step": 6396
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6417246460914612,
+      "learning_rate": 4.180564377323012e-05,
+      "loss": 3.2214,
+      "step": 6397
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6832561492919922,
+      "learning_rate": 4.1748277129546735e-05,
+      "loss": 3.441,
+      "step": 6398
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6649682521820068,
+      "learning_rate": 4.169094628674877e-05,
+      "loss": 3.3143,
+      "step": 6399
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6525314450263977,
+      "learning_rate": 4.163365125469207e-05,
+      "loss": 3.1577,
+      "step": 6400
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7238406538963318,
+      "learning_rate": 4.157639204322625e-05,
+      "loss": 3.3571,
+      "step": 6401
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6186272501945496,
+      "learning_rate": 4.1519168662194935e-05,
+      "loss": 3.277,
+      "step": 6402
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.605095624923706,
+      "learning_rate": 4.14619811214354e-05,
+      "loss": 3.3446,
+      "step": 6403
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.643415093421936,
+      "learning_rate": 4.140482943077895e-05,
+      "loss": 3.3538,
+      "step": 6404
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6669523119926453,
+      "learning_rate": 4.1347713600050547e-05,
+      "loss": 3.2147,
+      "step": 6405
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6525630354881287,
+      "learning_rate": 4.12906336390691e-05,
+      "loss": 3.2834,
+      "step": 6406
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6735588908195496,
+      "learning_rate": 4.123358955764728e-05,
+      "loss": 3.4095,
+      "step": 6407
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6865549087524414,
+      "learning_rate": 4.117658136559166e-05,
+      "loss": 3.3061,
+      "step": 6408
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6206440925598145,
+      "learning_rate": 4.1119609072702515e-05,
+      "loss": 3.2481,
+      "step": 6409
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6011683940887451,
+      "learning_rate": 4.10626726887742e-05,
+      "loss": 3.1444,
+      "step": 6410
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6887205243110657,
+      "learning_rate": 4.1005772223594583e-05,
+      "loss": 3.3052,
+      "step": 6411
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.643696129322052,
+      "learning_rate": 4.0948907686945675e-05,
+      "loss": 3.326,
+      "step": 6412
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.647750198841095,
+      "learning_rate": 4.089207908860304e-05,
+      "loss": 3.2997,
+      "step": 6413
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6211200952529907,
+      "learning_rate": 4.08352864383362e-05,
+      "loss": 3.3226,
+      "step": 6414
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6293831467628479,
+      "learning_rate": 4.077852974590846e-05,
+      "loss": 3.2841,
+      "step": 6415
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6573365330696106,
+      "learning_rate": 4.0721809021076916e-05,
+      "loss": 3.3569,
+      "step": 6416
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6426891088485718,
+      "learning_rate": 4.066512427359262e-05,
+      "loss": 3.2425,
+      "step": 6417
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6337445974349976,
+      "learning_rate": 4.060847551320027e-05,
+      "loss": 3.2817,
+      "step": 6418
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6924945116043091,
+      "learning_rate": 4.055186274963846e-05,
+      "loss": 3.263,
+      "step": 6419
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6352530717849731,
+      "learning_rate": 4.0495285992639494e-05,
+      "loss": 3.3007,
+      "step": 6420
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6801749467849731,
+      "learning_rate": 4.0438745251929756e-05,
+      "loss": 3.2665,
+      "step": 6421
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6209421753883362,
+      "learning_rate": 4.038224053722911e-05,
+      "loss": 3.2686,
+      "step": 6422
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6174101829528809,
+      "learning_rate": 4.0325771858251424e-05,
+      "loss": 3.1971,
+      "step": 6423
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6516974568367004,
+      "learning_rate": 4.026933922470427e-05,
+      "loss": 3.2522,
+      "step": 6424
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6284974813461304,
+      "learning_rate": 4.021294264628914e-05,
+      "loss": 3.3534,
+      "step": 6425
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6337777972221375,
+      "learning_rate": 4.015658213270126e-05,
+      "loss": 3.3124,
+      "step": 6426
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6753039956092834,
+      "learning_rate": 4.01002576936296e-05,
+      "loss": 3.2471,
+      "step": 6427
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6429099440574646,
+      "learning_rate": 4.004396933875701e-05,
+      "loss": 3.1807,
+      "step": 6428
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6312372088432312,
+      "learning_rate": 3.9987717077760054e-05,
+      "loss": 3.326,
+      "step": 6429
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6393996477127075,
+      "learning_rate": 3.993150092030928e-05,
+      "loss": 3.3049,
+      "step": 6430
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6306151151657104,
+      "learning_rate": 3.987532087606874e-05,
+      "loss": 3.3042,
+      "step": 6431
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6168709993362427,
+      "learning_rate": 3.981917695469658e-05,
+      "loss": 3.2459,
+      "step": 6432
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6085794568061829,
+      "learning_rate": 3.9763069165844546e-05,
+      "loss": 3.4072,
+      "step": 6433
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6100152730941772,
+      "learning_rate": 3.9706997519158185e-05,
+      "loss": 3.2662,
+      "step": 6434
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7263420820236206,
+      "learning_rate": 3.965096202427687e-05,
+      "loss": 3.1309,
+      "step": 6435
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6868110299110413,
+      "learning_rate": 3.959496269083376e-05,
+      "loss": 3.4455,
+      "step": 6436
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6217072606086731,
+      "learning_rate": 3.953899952845572e-05,
+      "loss": 3.2807,
+      "step": 6437
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6387251019477844,
+      "learning_rate": 3.9483072546763615e-05,
+      "loss": 3.2012,
+      "step": 6438
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6551047563552856,
+      "learning_rate": 3.9427181755371774e-05,
+      "loss": 3.2802,
+      "step": 6439
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.626193642616272,
+      "learning_rate": 3.937132716388864e-05,
+      "loss": 3.2716,
+      "step": 6440
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6018974781036377,
+      "learning_rate": 3.931550878191617e-05,
+      "loss": 3.286,
+      "step": 6441
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6675443649291992,
+      "learning_rate": 3.9259726619050225e-05,
+      "loss": 3.2788,
+      "step": 6442
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6622194051742554,
+      "learning_rate": 3.920398068488037e-05,
+      "loss": 3.244,
+      "step": 6443
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6131778359413147,
+      "learning_rate": 3.914827098898999e-05,
+      "loss": 3.3079,
+      "step": 6444
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6365820169448853,
+      "learning_rate": 3.909259754095617e-05,
+      "loss": 3.3133,
+      "step": 6445
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.618618369102478,
+      "learning_rate": 3.9036960350349954e-05,
+      "loss": 3.3683,
+      "step": 6446
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6009793877601624,
+      "learning_rate": 3.8981359426735945e-05,
+      "loss": 3.2897,
+      "step": 6447
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6588253378868103,
+      "learning_rate": 3.892579477967253e-05,
+      "loss": 3.3201,
+      "step": 6448
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6340826749801636,
+      "learning_rate": 3.887026641871203e-05,
+      "loss": 3.2024,
+      "step": 6449
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6288912892341614,
+      "learning_rate": 3.8814774353400335e-05,
+      "loss": 3.3785,
+      "step": 6450
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6233227252960205,
+      "learning_rate": 3.875931859327722e-05,
+      "loss": 3.3332,
+      "step": 6451
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6542572975158691,
+      "learning_rate": 3.870389914787609e-05,
+      "loss": 3.3217,
+      "step": 6452
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6089885234832764,
+      "learning_rate": 3.8648516026724314e-05,
+      "loss": 3.3098,
+      "step": 6453
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6347541213035583,
+      "learning_rate": 3.859316923934284e-05,
+      "loss": 3.2931,
+      "step": 6454
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6597635746002197,
+      "learning_rate": 3.8537858795246404e-05,
+      "loss": 3.2143,
+      "step": 6455
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6348395943641663,
+      "learning_rate": 3.84825847039435e-05,
+      "loss": 3.2582,
+      "step": 6456
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6304396390914917,
+      "learning_rate": 3.8427346974936355e-05,
+      "loss": 3.2717,
+      "step": 6457
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6483598351478577,
+      "learning_rate": 3.837214561772109e-05,
+      "loss": 3.3369,
+      "step": 6458
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6771075129508972,
+      "learning_rate": 3.83169806417874e-05,
+      "loss": 3.3749,
+      "step": 6459
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6419274806976318,
+      "learning_rate": 3.826185205661872e-05,
+      "loss": 3.2732,
+      "step": 6460
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6026401519775391,
+      "learning_rate": 3.820675987169239e-05,
+      "loss": 3.2216,
+      "step": 6461
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6403732299804688,
+      "learning_rate": 3.815170409647939e-05,
+      "loss": 3.3156,
+      "step": 6462
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6257762908935547,
+      "learning_rate": 3.8096684740444395e-05,
+      "loss": 3.2642,
+      "step": 6463
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6160749197006226,
+      "learning_rate": 3.804170181304587e-05,
+      "loss": 3.26,
+      "step": 6464
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6488245725631714,
+      "learning_rate": 3.798675532373605e-05,
+      "loss": 3.3118,
+      "step": 6465
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6134005188941956,
+      "learning_rate": 3.7931845281960807e-05,
+      "loss": 3.2695,
+      "step": 6466
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6293632388114929,
+      "learning_rate": 3.787697169715984e-05,
+      "loss": 3.3878,
+      "step": 6467
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6290237903594971,
+      "learning_rate": 3.7822134578766654e-05,
+      "loss": 3.3034,
+      "step": 6468
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.642416775226593,
+      "learning_rate": 3.776733393620832e-05,
+      "loss": 3.2255,
+      "step": 6469
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6392046213150024,
+      "learning_rate": 3.771256977890569e-05,
+      "loss": 3.2484,
+      "step": 6470
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6874974370002747,
+      "learning_rate": 3.765784211627335e-05,
+      "loss": 3.3158,
+      "step": 6471
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.61274653673172,
+      "learning_rate": 3.760315095771966e-05,
+      "loss": 3.2667,
+      "step": 6472
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6175403594970703,
+      "learning_rate": 3.754849631264653e-05,
+      "loss": 3.2756,
+      "step": 6473
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6124956011772156,
+      "learning_rate": 3.749387819044994e-05,
+      "loss": 3.3829,
+      "step": 6474
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6410250067710876,
+      "learning_rate": 3.743929660051923e-05,
+      "loss": 3.27,
+      "step": 6475
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6674247980117798,
+      "learning_rate": 3.738475155223761e-05,
+      "loss": 3.2707,
+      "step": 6476
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6086504459381104,
+      "learning_rate": 3.7330243054982116e-05,
+      "loss": 3.3489,
+      "step": 6477
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5963006615638733,
+      "learning_rate": 3.7275771118123305e-05,
+      "loss": 3.1831,
+      "step": 6478
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6756417155265808,
+      "learning_rate": 3.722133575102554e-05,
+      "loss": 3.3405,
+      "step": 6479
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5943170189857483,
+      "learning_rate": 3.7166936963046933e-05,
+      "loss": 3.2267,
+      "step": 6480
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5919336676597595,
+      "learning_rate": 3.711257476353916e-05,
+      "loss": 3.3203,
+      "step": 6481
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6363521218299866,
+      "learning_rate": 3.705824916184783e-05,
+      "loss": 3.2974,
+      "step": 6482
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5848808288574219,
+      "learning_rate": 3.700396016731214e-05,
+      "loss": 3.1555,
+      "step": 6483
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.658295214176178,
+      "learning_rate": 3.694970778926493e-05,
+      "loss": 3.2425,
+      "step": 6484
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5989205241203308,
+      "learning_rate": 3.6895492037032766e-05,
+      "loss": 3.2431,
+      "step": 6485
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6400282382965088,
+      "learning_rate": 3.684131291993614e-05,
+      "loss": 3.3204,
+      "step": 6486
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6289013028144836,
+      "learning_rate": 3.678717044728894e-05,
+      "loss": 3.297,
+      "step": 6487
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6446638703346252,
+      "learning_rate": 3.6733064628398874e-05,
+      "loss": 3.2334,
+      "step": 6488
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6189967393875122,
+      "learning_rate": 3.667899547256745e-05,
+      "loss": 3.3623,
+      "step": 6489
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6518852710723877,
+      "learning_rate": 3.6624962989089736e-05,
+      "loss": 3.4634,
+      "step": 6490
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6169833540916443,
+      "learning_rate": 3.657096718725456e-05,
+      "loss": 3.3121,
+      "step": 6491
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6083911061286926,
+      "learning_rate": 3.651700807634439e-05,
+      "loss": 3.3423,
+      "step": 6492
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6496937870979309,
+      "learning_rate": 3.646308566563541e-05,
+      "loss": 3.3,
+      "step": 6493
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6268054842948914,
+      "learning_rate": 3.640919996439751e-05,
+      "loss": 3.4026,
+      "step": 6494
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6139602661132812,
+      "learning_rate": 3.63553509818943e-05,
+      "loss": 3.1487,
+      "step": 6495
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6239408254623413,
+      "learning_rate": 3.6301538727383094e-05,
+      "loss": 3.3932,
+      "step": 6496
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6659420728683472,
+      "learning_rate": 3.624776321011478e-05,
+      "loss": 3.2684,
+      "step": 6497
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6213231682777405,
+      "learning_rate": 3.619402443933398e-05,
+      "loss": 3.1686,
+      "step": 6498
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6232683658599854,
+      "learning_rate": 3.614032242427903e-05,
+      "loss": 3.3488,
+      "step": 6499
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.604383111000061,
+      "learning_rate": 3.608665717418197e-05,
+      "loss": 3.2447,
+      "step": 6500
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5992885231971741,
+      "learning_rate": 3.603302869826841e-05,
+      "loss": 3.2639,
+      "step": 6501
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6617333292961121,
+      "learning_rate": 3.5979437005757674e-05,
+      "loss": 3.2442,
+      "step": 6502
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6387820243835449,
+      "learning_rate": 3.592588210586287e-05,
+      "loss": 3.2685,
+      "step": 6503
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6541626453399658,
+      "learning_rate": 3.587236400779073e-05,
+      "loss": 3.1938,
+      "step": 6504
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6782626509666443,
+      "learning_rate": 3.58188827207416e-05,
+      "loss": 3.3333,
+      "step": 6505
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6350370049476624,
+      "learning_rate": 3.576543825390954e-05,
+      "loss": 3.3704,
+      "step": 6506
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6781575083732605,
+      "learning_rate": 3.571203061648226e-05,
+      "loss": 3.3183,
+      "step": 6507
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6166051030158997,
+      "learning_rate": 3.565865981764116e-05,
+      "loss": 3.178,
+      "step": 6508
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6483914256095886,
+      "learning_rate": 3.560532586656126e-05,
+      "loss": 3.3736,
+      "step": 6509
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6345580220222473,
+      "learning_rate": 3.555202877241134e-05,
+      "loss": 3.1971,
+      "step": 6510
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6691543459892273,
+      "learning_rate": 3.54987685443538e-05,
+      "loss": 3.2858,
+      "step": 6511
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6558595895767212,
+      "learning_rate": 3.544554519154464e-05,
+      "loss": 3.2686,
+      "step": 6512
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6313934922218323,
+      "learning_rate": 3.539235872313354e-05,
+      "loss": 3.379,
+      "step": 6513
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6743571758270264,
+      "learning_rate": 3.533920914826397e-05,
+      "loss": 3.2874,
+      "step": 6514
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6316173672676086,
+      "learning_rate": 3.528609647607289e-05,
+      "loss": 3.2425,
+      "step": 6515
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6446629762649536,
+      "learning_rate": 3.523302071569098e-05,
+      "loss": 3.2588,
+      "step": 6516
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6505565047264099,
+      "learning_rate": 3.5179981876242634e-05,
+      "loss": 3.3203,
+      "step": 6517
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6769336462020874,
+      "learning_rate": 3.5126979966845826e-05,
+      "loss": 3.3582,
+      "step": 6518
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6479410529136658,
+      "learning_rate": 3.50740149966122e-05,
+      "loss": 3.353,
+      "step": 6519
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6441712975502014,
+      "learning_rate": 3.502108697464701e-05,
+      "loss": 3.2151,
+      "step": 6520
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6619863510131836,
+      "learning_rate": 3.496819591004921e-05,
+      "loss": 3.2887,
+      "step": 6521
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6329262256622314,
+      "learning_rate": 3.491534181191136e-05,
+      "loss": 3.2962,
+      "step": 6522
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6719884872436523,
+      "learning_rate": 3.4862524689319776e-05,
+      "loss": 3.3552,
+      "step": 6523
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6617701649665833,
+      "learning_rate": 3.480974455135422e-05,
+      "loss": 3.2037,
+      "step": 6524
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6485932469367981,
+      "learning_rate": 3.4757001407088346e-05,
+      "loss": 3.2314,
+      "step": 6525
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6022593975067139,
+      "learning_rate": 3.470429526558921e-05,
+      "loss": 3.1837,
+      "step": 6526
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6153969168663025,
+      "learning_rate": 3.465162613591769e-05,
+      "loss": 3.3104,
+      "step": 6527
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6110339164733887,
+      "learning_rate": 3.459899402712813e-05,
+      "loss": 3.2397,
+      "step": 6528
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.620966911315918,
+      "learning_rate": 3.4546398948268665e-05,
+      "loss": 3.318,
+      "step": 6529
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6802229881286621,
+      "learning_rate": 3.4493840908380895e-05,
+      "loss": 3.3426,
+      "step": 6530
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6216890215873718,
+      "learning_rate": 3.444131991650024e-05,
+      "loss": 3.2002,
+      "step": 6531
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6214625239372253,
+      "learning_rate": 3.4388835981655737e-05,
+      "loss": 3.2711,
+      "step": 6532
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6334563493728638,
+      "learning_rate": 3.433638911286987e-05,
+      "loss": 3.2746,
+      "step": 6533
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6457828283309937,
+      "learning_rate": 3.428397931915894e-05,
+      "loss": 3.2577,
+      "step": 6534
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6323023438453674,
+      "learning_rate": 3.423160660953276e-05,
+      "loss": 3.1932,
+      "step": 6535
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6734870672225952,
+      "learning_rate": 3.417927099299478e-05,
+      "loss": 3.2799,
+      "step": 6536
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.692999541759491,
+      "learning_rate": 3.4126972478542076e-05,
+      "loss": 3.3364,
+      "step": 6537
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6212725639343262,
+      "learning_rate": 3.407471107516549e-05,
+      "loss": 3.2772,
+      "step": 6538
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6048080325126648,
+      "learning_rate": 3.402248679184927e-05,
+      "loss": 3.2096,
+      "step": 6539
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6433602571487427,
+      "learning_rate": 3.397029963757134e-05,
+      "loss": 3.306,
+      "step": 6540
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6071273684501648,
+      "learning_rate": 3.391814962130341e-05,
+      "loss": 3.1467,
+      "step": 6541
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6584873795509338,
+      "learning_rate": 3.3866036752010585e-05,
+      "loss": 3.3005,
+      "step": 6542
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6271588206291199,
+      "learning_rate": 3.381396103865167e-05,
+      "loss": 3.1721,
+      "step": 6543
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6568840742111206,
+      "learning_rate": 3.376192249017912e-05,
+      "loss": 3.4128,
+      "step": 6544
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6104627847671509,
+      "learning_rate": 3.370992111553886e-05,
+      "loss": 3.1912,
+      "step": 6545
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6521439552307129,
+      "learning_rate": 3.365795692367069e-05,
+      "loss": 3.1236,
+      "step": 6546
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6480180621147156,
+      "learning_rate": 3.360602992350775e-05,
+      "loss": 3.3413,
+      "step": 6547
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7322046160697937,
+      "learning_rate": 3.3554140123976954e-05,
+      "loss": 3.3761,
+      "step": 6548
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6070480942726135,
+      "learning_rate": 3.35022875339987e-05,
+      "loss": 3.2662,
+      "step": 6549
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6356043219566345,
+      "learning_rate": 3.345047216248703e-05,
+      "loss": 3.2459,
+      "step": 6550
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6720722317695618,
+      "learning_rate": 3.3398694018349715e-05,
+      "loss": 3.2603,
+      "step": 6551
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6010866761207581,
+      "learning_rate": 3.334695311048788e-05,
+      "loss": 3.3378,
+      "step": 6552
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6258826851844788,
+      "learning_rate": 3.329524944779655e-05,
+      "loss": 3.2521,
+      "step": 6553
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6346193552017212,
+      "learning_rate": 3.3243583039164054e-05,
+      "loss": 3.3683,
+      "step": 6554
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6157397627830505,
+      "learning_rate": 3.319195389347251e-05,
+      "loss": 3.2757,
+      "step": 6555
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6648643612861633,
+      "learning_rate": 3.314036201959755e-05,
+      "loss": 3.3403,
+      "step": 6556
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6080839037895203,
+      "learning_rate": 3.3088807426408434e-05,
+      "loss": 3.1593,
+      "step": 6557
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6104198694229126,
+      "learning_rate": 3.3037290122767873e-05,
+      "loss": 3.2418,
+      "step": 6558
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.614980161190033,
+      "learning_rate": 3.298581011753246e-05,
+      "loss": 3.321,
+      "step": 6559
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6636888980865479,
+      "learning_rate": 3.293436741955208e-05,
+      "loss": 3.2396,
+      "step": 6560
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6287356615066528,
+      "learning_rate": 3.288296203767044e-05,
+      "loss": 3.2945,
+      "step": 6561
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6043485999107361,
+      "learning_rate": 3.2831593980724664e-05,
+      "loss": 3.4645,
+      "step": 6562
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6181927919387817,
+      "learning_rate": 3.278026325754552e-05,
+      "loss": 3.2439,
+      "step": 6563
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6592283248901367,
+      "learning_rate": 3.272896987695734e-05,
+      "loss": 3.38,
+      "step": 6564
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6501290798187256,
+      "learning_rate": 3.267771384777804e-05,
+      "loss": 3.3721,
+      "step": 6565
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6183682680130005,
+      "learning_rate": 3.2626495178819134e-05,
+      "loss": 3.2261,
+      "step": 6566
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6056748032569885,
+      "learning_rate": 3.257531387888574e-05,
+      "loss": 3.3504,
+      "step": 6567
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6474144458770752,
+      "learning_rate": 3.252416995677646e-05,
+      "loss": 3.3172,
+      "step": 6568
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8973803520202637,
+      "learning_rate": 3.247306342128359e-05,
+      "loss": 3.1872,
+      "step": 6569
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6066573858261108,
+      "learning_rate": 3.2421994281192915e-05,
+      "loss": 3.3767,
+      "step": 6570
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.628512442111969,
+      "learning_rate": 3.23709625452838e-05,
+      "loss": 3.2634,
+      "step": 6571
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6511738300323486,
+      "learning_rate": 3.2319968222329216e-05,
+      "loss": 3.2386,
+      "step": 6572
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6188576221466064,
+      "learning_rate": 3.226901132109558e-05,
+      "loss": 3.294,
+      "step": 6573
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6552718281745911,
+      "learning_rate": 3.221809185034311e-05,
+      "loss": 3.2761,
+      "step": 6574
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.638786792755127,
+      "learning_rate": 3.21672098188254e-05,
+      "loss": 3.3197,
+      "step": 6575
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.654994010925293,
+      "learning_rate": 3.211636523528966e-05,
+      "loss": 3.1878,
+      "step": 6576
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6370866298675537,
+      "learning_rate": 3.2065558108476615e-05,
+      "loss": 3.3911,
+      "step": 6577
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6311317086219788,
+      "learning_rate": 3.201478844712069e-05,
+      "loss": 3.3334,
+      "step": 6578
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6666600108146667,
+      "learning_rate": 3.196405625994972e-05,
+      "loss": 3.2714,
+      "step": 6579
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.638438880443573,
+      "learning_rate": 3.1913361555685196e-05,
+      "loss": 3.2415,
+      "step": 6580
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6487619280815125,
+      "learning_rate": 3.1862704343042e-05,
+      "loss": 3.2731,
+      "step": 6581
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6603454947471619,
+      "learning_rate": 3.181208463072888e-05,
+      "loss": 3.355,
+      "step": 6582
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6582320332527161,
+      "learning_rate": 3.1761502427447855e-05,
+      "loss": 3.3843,
+      "step": 6583
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.682377278804779,
+      "learning_rate": 3.1710957741894614e-05,
+      "loss": 3.2483,
+      "step": 6584
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6121364235877991,
+      "learning_rate": 3.166045058275835e-05,
+      "loss": 3.2538,
+      "step": 6585
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6380255222320557,
+      "learning_rate": 3.160998095872183e-05,
+      "loss": 3.319,
+      "step": 6586
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6388247013092041,
+      "learning_rate": 3.1559548878461325e-05,
+      "loss": 3.1455,
+      "step": 6587
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.614047110080719,
+      "learning_rate": 3.1509154350646745e-05,
+      "loss": 3.2948,
+      "step": 6588
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6629258394241333,
+      "learning_rate": 3.145879738394156e-05,
+      "loss": 3.3795,
+      "step": 6589
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6300737857818604,
+      "learning_rate": 3.140847798700267e-05,
+      "loss": 3.1161,
+      "step": 6590
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6252375841140747,
+      "learning_rate": 3.1358196168480515e-05,
+      "loss": 3.2803,
+      "step": 6591
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6455172300338745,
+      "learning_rate": 3.130795193701916e-05,
+      "loss": 3.2143,
+      "step": 6592
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6555576920509338,
+      "learning_rate": 3.1257745301256165e-05,
+      "loss": 3.3123,
+      "step": 6593
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6197571158409119,
+      "learning_rate": 3.1207576269822566e-05,
+      "loss": 3.2767,
+      "step": 6594
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6083783507347107,
+      "learning_rate": 3.115744485134314e-05,
+      "loss": 3.1134,
+      "step": 6595
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6117764711380005,
+      "learning_rate": 3.1107351054435906e-05,
+      "loss": 3.3644,
+      "step": 6596
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6059695482254028,
+      "learning_rate": 3.105729488771272e-05,
+      "loss": 3.4464,
+      "step": 6597
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.628901481628418,
+      "learning_rate": 3.100727635977873e-05,
+      "loss": 3.391,
+      "step": 6598
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6364163756370544,
+      "learning_rate": 3.095729547923273e-05,
+      "loss": 3.2195,
+      "step": 6599
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6437022089958191,
+      "learning_rate": 3.0907352254666985e-05,
+      "loss": 3.3625,
+      "step": 6600
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6143481135368347,
+      "learning_rate": 3.085744669466733e-05,
+      "loss": 3.1303,
+      "step": 6601
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7401031255722046,
+      "learning_rate": 3.080757880781307e-05,
+      "loss": 3.3805,
+      "step": 6602
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5818541049957275,
+      "learning_rate": 3.075774860267716e-05,
+      "loss": 3.276,
+      "step": 6603
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6768269538879395,
+      "learning_rate": 3.0707956087825923e-05,
+      "loss": 3.3174,
+      "step": 6604
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6687585115432739,
+      "learning_rate": 3.065820127181923e-05,
+      "loss": 3.2422,
+      "step": 6605
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6788862943649292,
+      "learning_rate": 3.060848416321063e-05,
+      "loss": 3.3239,
+      "step": 6606
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6286535263061523,
+      "learning_rate": 3.055880477054701e-05,
+      "loss": 3.2169,
+      "step": 6607
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6973392367362976,
+      "learning_rate": 3.0509163102368815e-05,
+      "loss": 3.3658,
+      "step": 6608
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6668422818183899,
+      "learning_rate": 3.0459559167209993e-05,
+      "loss": 3.4155,
+      "step": 6609
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6291588544845581,
+      "learning_rate": 3.0409992973598145e-05,
+      "loss": 3.2352,
+      "step": 6610
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6531401872634888,
+      "learning_rate": 3.03604645300542e-05,
+      "loss": 3.2694,
+      "step": 6611
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6205223202705383,
+      "learning_rate": 3.031097384509271e-05,
+      "loss": 3.3903,
+      "step": 6612
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.662426769733429,
+      "learning_rate": 3.0261520927221647e-05,
+      "loss": 3.3381,
+      "step": 6613
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5979740023612976,
+      "learning_rate": 3.021210578494249e-05,
+      "loss": 3.3004,
+      "step": 6614
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6564514636993408,
+      "learning_rate": 3.016272842675044e-05,
+      "loss": 3.27,
+      "step": 6615
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.683449923992157,
+      "learning_rate": 3.0113388861133907e-05,
+      "loss": 3.3458,
+      "step": 6616
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6408655047416687,
+      "learning_rate": 3.006408709657496e-05,
+      "loss": 3.0898,
+      "step": 6617
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6579421162605286,
+      "learning_rate": 3.0014823141549186e-05,
+      "loss": 3.2406,
+      "step": 6618
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6416338682174683,
+      "learning_rate": 2.9965597004525614e-05,
+      "loss": 3.2816,
+      "step": 6619
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.375583171844482,
+      "learning_rate": 2.991640869396675e-05,
+      "loss": 3.33,
+      "step": 6620
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6254746913909912,
+      "learning_rate": 2.9867258218328668e-05,
+      "loss": 3.2705,
+      "step": 6621
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6137012839317322,
+      "learning_rate": 2.9818145586060912e-05,
+      "loss": 3.3699,
+      "step": 6622
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6229578852653503,
+      "learning_rate": 2.976907080560645e-05,
+      "loss": 3.0818,
+      "step": 6623
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6454364061355591,
+      "learning_rate": 2.9720033885401816e-05,
+      "loss": 3.3186,
+      "step": 6624
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6500315070152283,
+      "learning_rate": 2.9671034833877147e-05,
+      "loss": 3.1809,
+      "step": 6625
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6044383645057678,
+      "learning_rate": 2.962207365945585e-05,
+      "loss": 3.2998,
+      "step": 6626
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6520947813987732,
+      "learning_rate": 2.9573150370554942e-05,
+      "loss": 3.2195,
+      "step": 6627
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6258668303489685,
+      "learning_rate": 2.9524264975584887e-05,
+      "loss": 3.245,
+      "step": 6628
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6491183042526245,
+      "learning_rate": 2.9475417482949657e-05,
+      "loss": 3.282,
+      "step": 6629
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6811879277229309,
+      "learning_rate": 2.9426607901046622e-05,
+      "loss": 3.407,
+      "step": 6630
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5877832770347595,
+      "learning_rate": 2.937783623826687e-05,
+      "loss": 3.2315,
+      "step": 6631
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6555650234222412,
+      "learning_rate": 2.9329102502994753e-05,
+      "loss": 3.3984,
+      "step": 6632
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6136705875396729,
+      "learning_rate": 2.92804067036081e-05,
+      "loss": 3.356,
+      "step": 6633
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6061446666717529,
+      "learning_rate": 2.9231748848478373e-05,
+      "loss": 3.1808,
+      "step": 6634
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6542891263961792,
+      "learning_rate": 2.918312894597039e-05,
+      "loss": 3.2174,
+      "step": 6635
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6651108860969543,
+      "learning_rate": 2.9134547004442456e-05,
+      "loss": 3.301,
+      "step": 6636
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7036979794502258,
+      "learning_rate": 2.90860030322464e-05,
+      "loss": 3.3462,
+      "step": 6637
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6571182012557983,
+      "learning_rate": 2.903749703772743e-05,
+      "loss": 3.2802,
+      "step": 6638
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6431148648262024,
+      "learning_rate": 2.8989029029224374e-05,
+      "loss": 3.3584,
+      "step": 6639
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6809457540512085,
+      "learning_rate": 2.8940599015069403e-05,
+      "loss": 3.2448,
+      "step": 6640
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6790762543678284,
+      "learning_rate": 2.8892207003588218e-05,
+      "loss": 3.1304,
+      "step": 6641
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7012802958488464,
+      "learning_rate": 2.8843853003099885e-05,
+      "loss": 3.4279,
+      "step": 6642
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6387409567832947,
+      "learning_rate": 2.8795537021917144e-05,
+      "loss": 3.2683,
+      "step": 6643
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6141805052757263,
+      "learning_rate": 2.874725906834602e-05,
+      "loss": 3.2577,
+      "step": 6644
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6355705261230469,
+      "learning_rate": 2.8699019150685958e-05,
+      "loss": 3.2237,
+      "step": 6645
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6300574541091919,
+      "learning_rate": 2.8650817277230123e-05,
+      "loss": 3.2852,
+      "step": 6646
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6067410707473755,
+      "learning_rate": 2.8602653456264893e-05,
+      "loss": 3.1998,
+      "step": 6647
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5872777104377747,
+      "learning_rate": 2.85545276960702e-05,
+      "loss": 3.2829,
+      "step": 6648
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6420096158981323,
+      "learning_rate": 2.850644000491942e-05,
+      "loss": 3.2486,
+      "step": 6649
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6760454773902893,
+      "learning_rate": 2.8458390391079365e-05,
+      "loss": 3.3123,
+      "step": 6650
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6170526742935181,
+      "learning_rate": 2.8410378862810255e-05,
+      "loss": 3.2622,
+      "step": 6651
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8137573599815369,
+      "learning_rate": 2.836240542836599e-05,
+      "loss": 3.3784,
+      "step": 6652
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6231951117515564,
+      "learning_rate": 2.831447009599361e-05,
+      "loss": 3.1708,
+      "step": 6653
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.642604410648346,
+      "learning_rate": 2.826657287393389e-05,
+      "loss": 3.2427,
+      "step": 6654
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6338666677474976,
+      "learning_rate": 2.8218713770420816e-05,
+      "loss": 3.3477,
+      "step": 6655
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.656091570854187,
+      "learning_rate": 2.8170892793681984e-05,
+      "loss": 3.2722,
+      "step": 6656
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6734408736228943,
+      "learning_rate": 2.812310995193834e-05,
+      "loss": 3.3415,
+      "step": 6657
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.636033833026886,
+      "learning_rate": 2.807536525340429e-05,
+      "loss": 3.2382,
+      "step": 6658
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.633607029914856,
+      "learning_rate": 2.8027658706287677e-05,
+      "loss": 3.3441,
+      "step": 6659
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6181923151016235,
+      "learning_rate": 2.7979990318789922e-05,
+      "loss": 3.1268,
+      "step": 6660
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6916999220848083,
+      "learning_rate": 2.7932360099105658e-05,
+      "loss": 3.2991,
+      "step": 6661
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6611192226409912,
+      "learning_rate": 2.7884768055423172e-05,
+      "loss": 3.3341,
+      "step": 6662
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6485355496406555,
+      "learning_rate": 2.7837214195924027e-05,
+      "loss": 3.1795,
+      "step": 6663
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6241832375526428,
+      "learning_rate": 2.7789698528783323e-05,
+      "loss": 3.2311,
+      "step": 6664
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6207740902900696,
+      "learning_rate": 2.7742221062169505e-05,
+      "loss": 3.3232,
+      "step": 6665
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6482567191123962,
+      "learning_rate": 2.769478180424445e-05,
+      "loss": 3.3199,
+      "step": 6666
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6029701828956604,
+      "learning_rate": 2.7647380763163673e-05,
+      "loss": 3.2725,
+      "step": 6667
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6569728851318359,
+      "learning_rate": 2.760001794707584e-05,
+      "loss": 3.149,
+      "step": 6668
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7240564823150635,
+      "learning_rate": 2.755269336412322e-05,
+      "loss": 3.2414,
+      "step": 6669
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6327884793281555,
+      "learning_rate": 2.750540702244139e-05,
+      "loss": 3.2251,
+      "step": 6670
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6695579886436462,
+      "learning_rate": 2.7458158930159516e-05,
+      "loss": 3.2286,
+      "step": 6671
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6714457869529724,
+      "learning_rate": 2.7410949095400067e-05,
+      "loss": 3.1332,
+      "step": 6672
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6200273036956787,
+      "learning_rate": 2.7363777526278915e-05,
+      "loss": 3.2911,
+      "step": 6673
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6575629115104675,
+      "learning_rate": 2.731664423090541e-05,
+      "loss": 3.2926,
+      "step": 6674
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6633983850479126,
+      "learning_rate": 2.726954921738234e-05,
+      "loss": 3.3532,
+      "step": 6675
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6128534078598022,
+      "learning_rate": 2.72224924938059e-05,
+      "loss": 3.2628,
+      "step": 6676
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6728388071060181,
+      "learning_rate": 2.7175474068265648e-05,
+      "loss": 3.3687,
+      "step": 6677
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6939319968223572,
+      "learning_rate": 2.7128493948844617e-05,
+      "loss": 3.477,
+      "step": 6678
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7096392512321472,
+      "learning_rate": 2.708155214361918e-05,
+      "loss": 3.104,
+      "step": 6679
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6583896279335022,
+      "learning_rate": 2.7034648660659246e-05,
+      "loss": 3.3376,
+      "step": 6680
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6055124402046204,
+      "learning_rate": 2.6987783508028023e-05,
+      "loss": 3.2004,
+      "step": 6681
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6449880003929138,
+      "learning_rate": 2.6940956693782215e-05,
+      "loss": 3.2938,
+      "step": 6682
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5981688499450684,
+      "learning_rate": 2.6894168225971876e-05,
+      "loss": 3.267,
+      "step": 6683
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.611633837223053,
+      "learning_rate": 2.684741811264052e-05,
+      "loss": 3.2029,
+      "step": 6684
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6699308156967163,
+      "learning_rate": 2.6800706361824966e-05,
+      "loss": 3.2592,
+      "step": 6685
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6041182279586792,
+      "learning_rate": 2.6754032981555544e-05,
+      "loss": 3.2229,
+      "step": 6686
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5983774662017822,
+      "learning_rate": 2.6707397979855885e-05,
+      "loss": 3.2502,
+      "step": 6687
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6428108811378479,
+      "learning_rate": 2.6660801364743138e-05,
+      "loss": 3.3587,
+      "step": 6688
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6479592323303223,
+      "learning_rate": 2.661424314422789e-05,
+      "loss": 3.2364,
+      "step": 6689
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6503172516822815,
+      "learning_rate": 2.656772332631391e-05,
+      "loss": 3.4001,
+      "step": 6690
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6411533951759338,
+      "learning_rate": 2.6521241918998572e-05,
+      "loss": 3.2245,
+      "step": 6691
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7209354043006897,
+      "learning_rate": 2.647479893027252e-05,
+      "loss": 3.1744,
+      "step": 6692
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6692107915878296,
+      "learning_rate": 2.642839436811986e-05,
+      "loss": 3.1625,
+      "step": 6693
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6229325532913208,
+      "learning_rate": 2.638202824051808e-05,
+      "loss": 3.3834,
+      "step": 6694
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6708949208259583,
+      "learning_rate": 2.633570055543802e-05,
+      "loss": 3.2889,
+      "step": 6695
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6829221844673157,
+      "learning_rate": 2.6289411320843974e-05,
+      "loss": 3.2398,
+      "step": 6696
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6486997008323669,
+      "learning_rate": 2.624316054469364e-05,
+      "loss": 3.2258,
+      "step": 6697
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6232972741127014,
+      "learning_rate": 2.6196948234937955e-05,
+      "loss": 3.2935,
+      "step": 6698
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6385177969932556,
+      "learning_rate": 2.615077439952146e-05,
+      "loss": 3.23,
+      "step": 6699
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6205930113792419,
+      "learning_rate": 2.6104639046381912e-05,
+      "loss": 3.3083,
+      "step": 6700
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7248706817626953,
+      "learning_rate": 2.605854218345055e-05,
+      "loss": 3.2182,
+      "step": 6701
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6649443507194519,
+      "learning_rate": 2.60124838186519e-05,
+      "loss": 3.2294,
+      "step": 6702
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7588315010070801,
+      "learning_rate": 2.5966463959904013e-05,
+      "loss": 3.3313,
+      "step": 6703
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6877144575119019,
+      "learning_rate": 2.5920482615118173e-05,
+      "loss": 3.365,
+      "step": 6704
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6987777948379517,
+      "learning_rate": 2.5874539792199142e-05,
+      "loss": 3.3193,
+      "step": 6705
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.636402428150177,
+      "learning_rate": 2.582863549904499e-05,
+      "loss": 3.1899,
+      "step": 6706
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.668968677520752,
+      "learning_rate": 2.578276974354718e-05,
+      "loss": 3.2208,
+      "step": 6707
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6446101069450378,
+      "learning_rate": 2.573694253359063e-05,
+      "loss": 3.1072,
+      "step": 6708
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6740022301673889,
+      "learning_rate": 2.5691153877053564e-05,
+      "loss": 3.2844,
+      "step": 6709
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6222270131111145,
+      "learning_rate": 2.5645403781807492e-05,
+      "loss": 3.3166,
+      "step": 6710
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.688671350479126,
+      "learning_rate": 2.5599692255717512e-05,
+      "loss": 3.2215,
+      "step": 6711
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6682297587394714,
+      "learning_rate": 2.555401930664189e-05,
+      "loss": 3.3589,
+      "step": 6712
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6356056928634644,
+      "learning_rate": 2.550838494243235e-05,
+      "loss": 3.2856,
+      "step": 6713
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6925631165504456,
+      "learning_rate": 2.5462789170933976e-05,
+      "loss": 3.2134,
+      "step": 6714
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6714974641799927,
+      "learning_rate": 2.541723199998522e-05,
+      "loss": 3.4193,
+      "step": 6715
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6370111107826233,
+      "learning_rate": 2.5371713437417787e-05,
+      "loss": 3.1771,
+      "step": 6716
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6870541572570801,
+      "learning_rate": 2.5326233491056948e-05,
+      "loss": 3.3673,
+      "step": 6717
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6816743612289429,
+      "learning_rate": 2.5280792168721257e-05,
+      "loss": 3.3273,
+      "step": 6718
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6466469168663025,
+      "learning_rate": 2.523538947822257e-05,
+      "loss": 3.3984,
+      "step": 6719
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6508055925369263,
+      "learning_rate": 2.519002542736612e-05,
+      "loss": 3.1394,
+      "step": 6720
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6685908436775208,
+      "learning_rate": 2.5144700023950527e-05,
+      "loss": 3.2152,
+      "step": 6721
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6493694186210632,
+      "learning_rate": 2.509941327576773e-05,
+      "loss": 3.3361,
+      "step": 6722
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6266916990280151,
+      "learning_rate": 2.5054165190603022e-05,
+      "loss": 3.1545,
+      "step": 6723
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5824097394943237,
+      "learning_rate": 2.500895577623516e-05,
+      "loss": 3.1569,
+      "step": 6724
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6279321312904358,
+      "learning_rate": 2.4963785040436088e-05,
+      "loss": 3.4809,
+      "step": 6725
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6300782561302185,
+      "learning_rate": 2.4918652990971236e-05,
+      "loss": 3.2169,
+      "step": 6726
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5826421976089478,
+      "learning_rate": 2.4873559635599313e-05,
+      "loss": 3.2049,
+      "step": 6727
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6628600358963013,
+      "learning_rate": 2.4828504982072397e-05,
+      "loss": 3.2544,
+      "step": 6728
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5862376093864441,
+      "learning_rate": 2.4783489038135847e-05,
+      "loss": 3.2769,
+      "step": 6729
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6180055737495422,
+      "learning_rate": 2.47385118115285e-05,
+      "loss": 3.2885,
+      "step": 6730
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6752247214317322,
+      "learning_rate": 2.4693573309982342e-05,
+      "loss": 3.3776,
+      "step": 6731
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6708443760871887,
+      "learning_rate": 2.4648673541222994e-05,
+      "loss": 3.3513,
+      "step": 6732
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7268015742301941,
+      "learning_rate": 2.4603812512969142e-05,
+      "loss": 3.3318,
+      "step": 6733
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6242460012435913,
+      "learning_rate": 2.4558990232932955e-05,
+      "loss": 3.28,
+      "step": 6734
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6382979154586792,
+      "learning_rate": 2.451420670881982e-05,
+      "loss": 3.3254,
+      "step": 6735
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.629286527633667,
+      "learning_rate": 2.4469461948328635e-05,
+      "loss": 3.1642,
+      "step": 6736
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6485285758972168,
+      "learning_rate": 2.4424755959151558e-05,
+      "loss": 3.2271,
+      "step": 6737
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6550916433334351,
+      "learning_rate": 2.4380088748973933e-05,
+      "loss": 3.2394,
+      "step": 6738
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6315405368804932,
+      "learning_rate": 2.4335460325474736e-05,
+      "loss": 3.2613,
+      "step": 6739
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6707183122634888,
+      "learning_rate": 2.4290870696326046e-05,
+      "loss": 3.4281,
+      "step": 6740
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6594843864440918,
+      "learning_rate": 2.424631986919332e-05,
+      "loss": 3.4797,
+      "step": 6741
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6843511462211609,
+      "learning_rate": 2.420180785173534e-05,
+      "loss": 3.3491,
+      "step": 6742
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6677063703536987,
+      "learning_rate": 2.4157334651604302e-05,
+      "loss": 3.2779,
+      "step": 6743
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6253190636634827,
+      "learning_rate": 2.411290027644558e-05,
+      "loss": 3.2955,
+      "step": 6744
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6013156175613403,
+      "learning_rate": 2.4068504733897988e-05,
+      "loss": 3.2331,
+      "step": 6745
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6724177598953247,
+      "learning_rate": 2.4024148031593723e-05,
+      "loss": 3.2156,
+      "step": 6746
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6198339462280273,
+      "learning_rate": 2.397983017715813e-05,
+      "loss": 3.2331,
+      "step": 6747
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6026395559310913,
+      "learning_rate": 2.3935551178210004e-05,
+      "loss": 3.1947,
+      "step": 6748
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6195902228355408,
+      "learning_rate": 2.3891311042361364e-05,
+      "loss": 3.2482,
+      "step": 6749
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.635100245475769,
+      "learning_rate": 2.3847109777217658e-05,
+      "loss": 3.3386,
+      "step": 6750
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6043824553489685,
+      "learning_rate": 2.3802947390377554e-05,
+      "loss": 3.118,
+      "step": 6751
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6124756932258606,
+      "learning_rate": 2.375882388943307e-05,
+      "loss": 3.1976,
+      "step": 6752
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6314394474029541,
+      "learning_rate": 2.3714739281969545e-05,
+      "loss": 3.2971,
+      "step": 6753
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6392366886138916,
+      "learning_rate": 2.3670693575565726e-05,
+      "loss": 3.2638,
+      "step": 6754
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6057491898536682,
+      "learning_rate": 2.3626686777793503e-05,
+      "loss": 3.2018,
+      "step": 6755
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6680116057395935,
+      "learning_rate": 2.3582718896218185e-05,
+      "loss": 3.3387,
+      "step": 6756
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6639420986175537,
+      "learning_rate": 2.3538789938398335e-05,
+      "loss": 3.4412,
+      "step": 6757
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6201452016830444,
+      "learning_rate": 2.3494899911885857e-05,
+      "loss": 3.2545,
+      "step": 6758
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6092161536216736,
+      "learning_rate": 2.3451048824225912e-05,
+      "loss": 3.2522,
+      "step": 6759
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6514724493026733,
+      "learning_rate": 2.3407236682957106e-05,
+      "loss": 3.2481,
+      "step": 6760
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5797368884086609,
+      "learning_rate": 2.336346349561119e-05,
+      "loss": 3.1768,
+      "step": 6761
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.661661684513092,
+      "learning_rate": 2.3319729269713263e-05,
+      "loss": 3.1709,
+      "step": 6762
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6890133619308472,
+      "learning_rate": 2.3276034012781803e-05,
+      "loss": 3.2111,
+      "step": 6763
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6165077090263367,
+      "learning_rate": 2.323237773232853e-05,
+      "loss": 3.2561,
+      "step": 6764
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6231985092163086,
+      "learning_rate": 2.3188760435858436e-05,
+      "loss": 3.4661,
+      "step": 6765
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6593947410583496,
+      "learning_rate": 2.3145182130869772e-05,
+      "loss": 3.3824,
+      "step": 6766
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6102452278137207,
+      "learning_rate": 2.3101642824854302e-05,
+      "loss": 3.3386,
+      "step": 6767
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6081363558769226,
+      "learning_rate": 2.3058142525296864e-05,
+      "loss": 3.1532,
+      "step": 6768
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6321461200714111,
+      "learning_rate": 2.301468123967565e-05,
+      "loss": 3.243,
+      "step": 6769
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6125012636184692,
+      "learning_rate": 2.297125897546215e-05,
+      "loss": 3.2151,
+      "step": 6770
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6407002806663513,
+      "learning_rate": 2.29278757401212e-05,
+      "loss": 3.1811,
+      "step": 6771
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6125441789627075,
+      "learning_rate": 2.288453154111081e-05,
+      "loss": 3.2838,
+      "step": 6772
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6470776796340942,
+      "learning_rate": 2.284122638588243e-05,
+      "loss": 3.485,
+      "step": 6773
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6441331505775452,
+      "learning_rate": 2.2797960281880664e-05,
+      "loss": 3.3061,
+      "step": 6774
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.65076744556427,
+      "learning_rate": 2.2754733236543506e-05,
+      "loss": 3.283,
+      "step": 6775
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6396564841270447,
+      "learning_rate": 2.2711545257302152e-05,
+      "loss": 3.177,
+      "step": 6776
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6858227849006653,
+      "learning_rate": 2.2668396351581134e-05,
+      "loss": 3.3496,
+      "step": 6777
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5985760688781738,
+      "learning_rate": 2.262528652679824e-05,
+      "loss": 3.29,
+      "step": 6778
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6282758712768555,
+      "learning_rate": 2.2582215790364542e-05,
+      "loss": 3.4646,
+      "step": 6779
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6286015510559082,
+      "learning_rate": 2.2539184149684338e-05,
+      "loss": 3.2247,
+      "step": 6780
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6121549010276794,
+      "learning_rate": 2.2496191612155355e-05,
+      "loss": 3.3774,
+      "step": 6781
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6485132575035095,
+      "learning_rate": 2.2453238185168505e-05,
+      "loss": 3.3468,
+      "step": 6782
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6902052164077759,
+      "learning_rate": 2.2410323876107974e-05,
+      "loss": 3.3393,
+      "step": 6783
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5603453516960144,
+      "learning_rate": 2.2367448692351216e-05,
+      "loss": 3.1907,
+      "step": 6784
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6382895708084106,
+      "learning_rate": 2.2324612641268975e-05,
+      "loss": 3.2211,
+      "step": 6785
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6377125978469849,
+      "learning_rate": 2.2281815730225252e-05,
+      "loss": 3.335,
+      "step": 6786
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.638985812664032,
+      "learning_rate": 2.2239057966577297e-05,
+      "loss": 3.2314,
+      "step": 6787
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6332682371139526,
+      "learning_rate": 2.219633935767576e-05,
+      "loss": 3.2651,
+      "step": 6788
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6259728074073792,
+      "learning_rate": 2.215365991086443e-05,
+      "loss": 3.211,
+      "step": 6789
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6506862640380859,
+      "learning_rate": 2.2111019633480306e-05,
+      "loss": 3.3028,
+      "step": 6790
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6298990845680237,
+      "learning_rate": 2.2068418532853878e-05,
+      "loss": 3.1886,
+      "step": 6791
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7053328156471252,
+      "learning_rate": 2.202585661630871e-05,
+      "loss": 3.1508,
+      "step": 6792
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6330047845840454,
+      "learning_rate": 2.198333389116172e-05,
+      "loss": 3.2599,
+      "step": 6793
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7028726935386658,
+      "learning_rate": 2.1940850364723014e-05,
+      "loss": 3.3423,
+      "step": 6794
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.631902277469635,
+      "learning_rate": 2.1898406044295964e-05,
+      "loss": 3.2537,
+      "step": 6795
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5942831039428711,
+      "learning_rate": 2.185600093717735e-05,
+      "loss": 3.3153,
+      "step": 6796
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6552851796150208,
+      "learning_rate": 2.1813635050657032e-05,
+      "loss": 3.2042,
+      "step": 6797
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6387913823127747,
+      "learning_rate": 2.1771308392018213e-05,
+      "loss": 3.2958,
+      "step": 6798
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7131662368774414,
+      "learning_rate": 2.1729020968537296e-05,
+      "loss": 3.2817,
+      "step": 6799
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6315956711769104,
+      "learning_rate": 2.1686772787484072e-05,
+      "loss": 3.2904,
+      "step": 6800
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6526632905006409,
+      "learning_rate": 2.164456385612143e-05,
+      "loss": 3.2094,
+      "step": 6801
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6810610294342041,
+      "learning_rate": 2.1602394181705564e-05,
+      "loss": 3.3241,
+      "step": 6802
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6127844452857971,
+      "learning_rate": 2.1560263771485983e-05,
+      "loss": 3.3316,
+      "step": 6803
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7066926956176758,
+      "learning_rate": 2.1518172632705334e-05,
+      "loss": 3.3411,
+      "step": 6804
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6281033158302307,
+      "learning_rate": 2.1476120772599613e-05,
+      "loss": 3.2523,
+      "step": 6805
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6492352485656738,
+      "learning_rate": 2.1434108198398027e-05,
+      "loss": 3.2758,
+      "step": 6806
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6568070650100708,
+      "learning_rate": 2.1392134917322975e-05,
+      "loss": 3.2168,
+      "step": 6807
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6375867128372192,
+      "learning_rate": 2.1350200936590152e-05,
+      "loss": 3.2557,
+      "step": 6808
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5970162153244019,
+      "learning_rate": 2.130830626340857e-05,
+      "loss": 3.2336,
+      "step": 6809
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6471285820007324,
+      "learning_rate": 2.1266450904980335e-05,
+      "loss": 3.2764,
+      "step": 6810
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6247797608375549,
+      "learning_rate": 2.1224634868500902e-05,
+      "loss": 3.31,
+      "step": 6811
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6753485202789307,
+      "learning_rate": 2.1182858161158947e-05,
+      "loss": 3.3195,
+      "step": 6812
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6527683734893799,
+      "learning_rate": 2.114112079013636e-05,
+      "loss": 3.3438,
+      "step": 6813
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6041711568832397,
+      "learning_rate": 2.1099422762608262e-05,
+      "loss": 3.2407,
+      "step": 6814
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.670981764793396,
+      "learning_rate": 2.105776408574303e-05,
+      "loss": 3.2142,
+      "step": 6815
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6561753153800964,
+      "learning_rate": 2.1016144766702217e-05,
+      "loss": 3.2738,
+      "step": 6816
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6334289312362671,
+      "learning_rate": 2.097456481264079e-05,
+      "loss": 3.2525,
+      "step": 6817
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6357218623161316,
+      "learning_rate": 2.0933024230706726e-05,
+      "loss": 3.2613,
+      "step": 6818
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.639295220375061,
+      "learning_rate": 2.08915230280414e-05,
+      "loss": 3.3472,
+      "step": 6819
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6403231024742126,
+      "learning_rate": 2.085006121177932e-05,
+      "loss": 3.335,
+      "step": 6820
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6622771620750427,
+      "learning_rate": 2.0808638789048263e-05,
+      "loss": 3.1739,
+      "step": 6821
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.708372175693512,
+      "learning_rate": 2.0767255766969195e-05,
+      "loss": 3.3605,
+      "step": 6822
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6363511681556702,
+      "learning_rate": 2.0725912152656317e-05,
+      "loss": 3.1817,
+      "step": 6823
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6383823752403259,
+      "learning_rate": 2.0684607953217164e-05,
+      "loss": 3.395,
+      "step": 6824
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6685783863067627,
+      "learning_rate": 2.0643343175752367e-05,
+      "loss": 3.2711,
+      "step": 6825
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6340213418006897,
+      "learning_rate": 2.060211782735577e-05,
+      "loss": 3.3396,
+      "step": 6826
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6279425621032715,
+      "learning_rate": 2.0560931915114518e-05,
+      "loss": 3.4247,
+      "step": 6827
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6689784526824951,
+      "learning_rate": 2.0519785446108996e-05,
+      "loss": 3.3333,
+      "step": 6828
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6482875347137451,
+      "learning_rate": 2.0478678427412718e-05,
+      "loss": 3.322,
+      "step": 6829
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6636989712715149,
+      "learning_rate": 2.0437610866092442e-05,
+      "loss": 3.2851,
+      "step": 6830
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6471637487411499,
+      "learning_rate": 2.0396582769208137e-05,
+      "loss": 3.3048,
+      "step": 6831
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6450260281562805,
+      "learning_rate": 2.0355594143813078e-05,
+      "loss": 3.2115,
+      "step": 6832
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6283779144287109,
+      "learning_rate": 2.031464499695368e-05,
+      "loss": 3.1822,
+      "step": 6833
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6070658564567566,
+      "learning_rate": 2.0273735335669536e-05,
+      "loss": 3.3539,
+      "step": 6834
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6264026165008545,
+      "learning_rate": 2.0232865166993492e-05,
+      "loss": 3.2446,
+      "step": 6835
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5890588760375977,
+      "learning_rate": 2.0192034497951566e-05,
+      "loss": 3.2071,
+      "step": 6836
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6800095438957214,
+      "learning_rate": 2.0151243335563146e-05,
+      "loss": 3.298,
+      "step": 6837
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6465356945991516,
+      "learning_rate": 2.0110491686840564e-05,
+      "loss": 3.3531,
+      "step": 6838
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6601016521453857,
+      "learning_rate": 2.006977955878961e-05,
+      "loss": 3.1851,
+      "step": 6839
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6479170322418213,
+      "learning_rate": 2.0029106958409148e-05,
+      "loss": 3.2531,
+      "step": 6840
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6518640518188477,
+      "learning_rate": 1.9988473892691235e-05,
+      "loss": 3.258,
+      "step": 6841
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6924657225608826,
+      "learning_rate": 1.994788036862119e-05,
+      "loss": 3.2473,
+      "step": 6842
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6231380701065063,
+      "learning_rate": 1.990732639317752e-05,
+      "loss": 3.2671,
+      "step": 6843
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6526206135749817,
+      "learning_rate": 1.9866811973331846e-05,
+      "loss": 3.3348,
+      "step": 6844
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6139849424362183,
+      "learning_rate": 1.982633711604917e-05,
+      "loss": 3.3864,
+      "step": 6845
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6477641463279724,
+      "learning_rate": 1.978590182828749e-05,
+      "loss": 3.3935,
+      "step": 6846
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6017580032348633,
+      "learning_rate": 1.9745506116998214e-05,
+      "loss": 3.1474,
+      "step": 6847
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5790743231773376,
+      "learning_rate": 1.970514998912576e-05,
+      "loss": 3.2047,
+      "step": 6848
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6539261937141418,
+      "learning_rate": 1.9664833451607856e-05,
+      "loss": 3.1933,
+      "step": 6849
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6438460946083069,
+      "learning_rate": 1.9624556511375347e-05,
+      "loss": 3.2391,
+      "step": 6850
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.621233344078064,
+      "learning_rate": 1.9584319175352307e-05,
+      "loss": 3.1789,
+      "step": 6851
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.676813542842865,
+      "learning_rate": 1.9544121450455976e-05,
+      "loss": 3.1908,
+      "step": 6852
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5853865146636963,
+      "learning_rate": 1.950396334359686e-05,
+      "loss": 3.3179,
+      "step": 6853
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6410489082336426,
+      "learning_rate": 1.9463844861678627e-05,
+      "loss": 3.2646,
+      "step": 6854
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6273015737533569,
+      "learning_rate": 1.942376601159798e-05,
+      "loss": 3.1283,
+      "step": 6855
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6359155178070068,
+      "learning_rate": 1.9383726800245083e-05,
+      "loss": 3.2482,
+      "step": 6856
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6588960886001587,
+      "learning_rate": 1.934372723450309e-05,
+      "loss": 3.2952,
+      "step": 6857
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6521344780921936,
+      "learning_rate": 1.9303767321248388e-05,
+      "loss": 3.1928,
+      "step": 6858
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6000865697860718,
+      "learning_rate": 1.926384706735049e-05,
+      "loss": 3.2937,
+      "step": 6859
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6228570938110352,
+      "learning_rate": 1.9223966479672255e-05,
+      "loss": 3.3858,
+      "step": 6860
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.643111526966095,
+      "learning_rate": 1.9184125565069543e-05,
+      "loss": 3.2005,
+      "step": 6861
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6268308162689209,
+      "learning_rate": 1.914432433039151e-05,
+      "loss": 3.2329,
+      "step": 6862
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6840165257453918,
+      "learning_rate": 1.9104562782480434e-05,
+      "loss": 3.2345,
+      "step": 6863
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6919675469398499,
+      "learning_rate": 1.906484092817176e-05,
+      "loss": 3.2749,
+      "step": 6864
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6603330969810486,
+      "learning_rate": 1.902515877429417e-05,
+      "loss": 3.2708,
+      "step": 6865
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6835293173789978,
+      "learning_rate": 1.8985516327669512e-05,
+      "loss": 3.3131,
+      "step": 6866
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6213630437850952,
+      "learning_rate": 1.894591359511269e-05,
+      "loss": 3.2018,
+      "step": 6867
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6311686635017395,
+      "learning_rate": 1.890635058343196e-05,
+      "loss": 3.1391,
+      "step": 6868
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.603550136089325,
+      "learning_rate": 1.886682729942865e-05,
+      "loss": 3.2515,
+      "step": 6869
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.640389084815979,
+      "learning_rate": 1.8827343749897224e-05,
+      "loss": 3.2602,
+      "step": 6870
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6399551033973694,
+      "learning_rate": 1.8787899941625413e-05,
+      "loss": 3.2608,
+      "step": 6871
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6673776507377625,
+      "learning_rate": 1.8748495881394046e-05,
+      "loss": 3.2441,
+      "step": 6872
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5939326882362366,
+      "learning_rate": 1.870913157597709e-05,
+      "loss": 3.1888,
+      "step": 6873
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.64223712682724,
+      "learning_rate": 1.866980703214177e-05,
+      "loss": 3.2668,
+      "step": 6874
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5804606080055237,
+      "learning_rate": 1.8630522256648463e-05,
+      "loss": 3.1411,
+      "step": 6875
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6221579909324646,
+      "learning_rate": 1.8591277256250648e-05,
+      "loss": 3.1402,
+      "step": 6876
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6562621593475342,
+      "learning_rate": 1.855207203769499e-05,
+      "loss": 3.3352,
+      "step": 6877
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.654252827167511,
+      "learning_rate": 1.8512906607721342e-05,
+      "loss": 3.2961,
+      "step": 6878
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6618490219116211,
+      "learning_rate": 1.8473780973062655e-05,
+      "loss": 3.2589,
+      "step": 6879
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6368651390075684,
+      "learning_rate": 1.8434695140445074e-05,
+      "loss": 3.2181,
+      "step": 6880
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6331031322479248,
+      "learning_rate": 1.8395649116587974e-05,
+      "loss": 3.3327,
+      "step": 6881
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6702656149864197,
+      "learning_rate": 1.8356642908203767e-05,
+      "loss": 3.2164,
+      "step": 6882
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6424792408943176,
+      "learning_rate": 1.8317676521998033e-05,
+      "loss": 3.2294,
+      "step": 6883
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6082201600074768,
+      "learning_rate": 1.827874996466966e-05,
+      "loss": 3.3042,
+      "step": 6884
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6452752947807312,
+      "learning_rate": 1.823986324291052e-05,
+      "loss": 3.3743,
+      "step": 6885
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6559005379676819,
+      "learning_rate": 1.8201016363405653e-05,
+      "loss": 3.3308,
+      "step": 6886
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6568729877471924,
+      "learning_rate": 1.816220933283336e-05,
+      "loss": 3.2974,
+      "step": 6887
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7080766558647156,
+      "learning_rate": 1.8123442157864907e-05,
+      "loss": 3.1731,
+      "step": 6888
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6404166221618652,
+      "learning_rate": 1.8084714845164912e-05,
+      "loss": 3.3248,
+      "step": 6889
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.664342999458313,
+      "learning_rate": 1.804602740139105e-05,
+      "loss": 3.3336,
+      "step": 6890
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6884958744049072,
+      "learning_rate": 1.8007379833194142e-05,
+      "loss": 3.1601,
+      "step": 6891
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7024304270744324,
+      "learning_rate": 1.7968772147218067e-05,
+      "loss": 3.2835,
+      "step": 6892
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.670708954334259,
+      "learning_rate": 1.793020435010004e-05,
+      "loss": 3.3666,
+      "step": 6893
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6311020255088806,
+      "learning_rate": 1.7891676448470255e-05,
+      "loss": 3.2645,
+      "step": 6894
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7195715308189392,
+      "learning_rate": 1.785318844895209e-05,
+      "loss": 3.2093,
+      "step": 6895
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6251908540725708,
+      "learning_rate": 1.7814740358162136e-05,
+      "loss": 3.3117,
+      "step": 6896
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6668151617050171,
+      "learning_rate": 1.7776332182710047e-05,
+      "loss": 3.3754,
+      "step": 6897
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.655120313167572,
+      "learning_rate": 1.773796392919863e-05,
+      "loss": 3.246,
+      "step": 6898
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6486814618110657,
+      "learning_rate": 1.769963560422383e-05,
+      "loss": 3.3588,
+      "step": 6899
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6619285941123962,
+      "learning_rate": 1.7661347214374706e-05,
+      "loss": 3.2986,
+      "step": 6900
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.677575945854187,
+      "learning_rate": 1.762309876623347e-05,
+      "loss": 3.1905,
+      "step": 6901
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6310055255889893,
+      "learning_rate": 1.7584890266375552e-05,
+      "loss": 3.3665,
+      "step": 6902
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6316937804222107,
+      "learning_rate": 1.7546721721369314e-05,
+      "loss": 3.477,
+      "step": 6903
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6531652808189392,
+      "learning_rate": 1.7508593137776503e-05,
+      "loss": 3.3253,
+      "step": 6904
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6354236602783203,
+      "learning_rate": 1.7470504522151792e-05,
+      "loss": 3.3458,
+      "step": 6905
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6475649476051331,
+      "learning_rate": 1.7432455881043085e-05,
+      "loss": 3.3281,
+      "step": 6906
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6476089954376221,
+      "learning_rate": 1.7394447220991342e-05,
+      "loss": 3.426,
+      "step": 6907
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5931687951087952,
+      "learning_rate": 1.735647854853073e-05,
+      "loss": 3.1769,
+      "step": 6908
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6188100576400757,
+      "learning_rate": 1.7318549870188468e-05,
+      "loss": 3.2636,
+      "step": 6909
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6311696767807007,
+      "learning_rate": 1.728066119248492e-05,
+      "loss": 3.2662,
+      "step": 6910
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6498907804489136,
+      "learning_rate": 1.724281252193369e-05,
+      "loss": 3.2602,
+      "step": 6911
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6863357424736023,
+      "learning_rate": 1.7205003865041342e-05,
+      "loss": 3.2302,
+      "step": 6912
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6403094530105591,
+      "learning_rate": 1.7167235228307627e-05,
+      "loss": 3.2029,
+      "step": 6913
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6653391122817993,
+      "learning_rate": 1.7129506618225376e-05,
+      "loss": 3.3177,
+      "step": 6914
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.616364598274231,
+      "learning_rate": 1.7091818041280626e-05,
+      "loss": 3.3601,
+      "step": 6915
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6449361443519592,
+      "learning_rate": 1.7054169503952415e-05,
+      "loss": 3.3162,
+      "step": 6916
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6562541127204895,
+      "learning_rate": 1.701656101271304e-05,
+      "loss": 3.3603,
+      "step": 6917
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6591519117355347,
+      "learning_rate": 1.6978992574027824e-05,
+      "loss": 3.3095,
+      "step": 6918
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.627376914024353,
+      "learning_rate": 1.6941464194355188e-05,
+      "loss": 3.334,
+      "step": 6919
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6193311810493469,
+      "learning_rate": 1.6903975880146638e-05,
+      "loss": 3.2708,
+      "step": 6920
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6567133069038391,
+      "learning_rate": 1.686652763784699e-05,
+      "loss": 3.2111,
+      "step": 6921
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6930579543113708,
+      "learning_rate": 1.6829119473893927e-05,
+      "loss": 3.3558,
+      "step": 6922
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6540127992630005,
+      "learning_rate": 1.679175139471839e-05,
+      "loss": 3.2656,
+      "step": 6923
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5813849568367004,
+      "learning_rate": 1.6754423406744323e-05,
+      "loss": 3.3181,
+      "step": 6924
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6777455806732178,
+      "learning_rate": 1.6717135516388925e-05,
+      "loss": 3.3117,
+      "step": 6925
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.657409131526947,
+      "learning_rate": 1.6679887730062404e-05,
+      "loss": 3.2066,
+      "step": 6926
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6671031713485718,
+      "learning_rate": 1.6642680054168026e-05,
+      "loss": 3.3455,
+      "step": 6927
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5967366695404053,
+      "learning_rate": 1.6605512495102282e-05,
+      "loss": 3.2231,
+      "step": 6928
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6444880366325378,
+      "learning_rate": 1.656838505925462e-05,
+      "loss": 3.2665,
+      "step": 6929
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6808575391769409,
+      "learning_rate": 1.6531297753007795e-05,
+      "loss": 3.2324,
+      "step": 6930
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6104139089584351,
+      "learning_rate": 1.649425058273743e-05,
+      "loss": 3.2293,
+      "step": 6931
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6585542559623718,
+      "learning_rate": 1.6457243554812486e-05,
+      "loss": 3.1955,
+      "step": 6932
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6599720120429993,
+      "learning_rate": 1.6420276675594814e-05,
+      "loss": 3.4477,
+      "step": 6933
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6509204506874084,
+      "learning_rate": 1.6383349951439475e-05,
+      "loss": 3.2605,
+      "step": 6934
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7017077207565308,
+      "learning_rate": 1.634646338869461e-05,
+      "loss": 3.2774,
+      "step": 6935
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6533510684967041,
+      "learning_rate": 1.6309616993701426e-05,
+      "loss": 3.3077,
+      "step": 6936
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6311295032501221,
+      "learning_rate": 1.6272810772794218e-05,
+      "loss": 3.2622,
+      "step": 6937
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6376855969429016,
+      "learning_rate": 1.623604473230042e-05,
+      "loss": 3.287,
+      "step": 6938
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.683590829372406,
+      "learning_rate": 1.6199318878540593e-05,
+      "loss": 3.3705,
+      "step": 6939
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6251614689826965,
+      "learning_rate": 1.616263321782832e-05,
+      "loss": 3.295,
+      "step": 6940
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.608283281326294,
+      "learning_rate": 1.6125987756470257e-05,
+      "loss": 3.1217,
+      "step": 6941
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6173918843269348,
+      "learning_rate": 1.6089382500766193e-05,
+      "loss": 3.299,
+      "step": 6942
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6270238161087036,
+      "learning_rate": 1.605281745700904e-05,
+      "loss": 3.2012,
+      "step": 6943
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.670943021774292,
+      "learning_rate": 1.6016292631484684e-05,
+      "loss": 3.2473,
+      "step": 6944
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6275277137756348,
+      "learning_rate": 1.5979808030472164e-05,
+      "loss": 3.0922,
+      "step": 6945
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6165827512741089,
+      "learning_rate": 1.5943363660243655e-05,
+      "loss": 3.2648,
+      "step": 6946
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6089856028556824,
+      "learning_rate": 1.5906959527064334e-05,
+      "loss": 3.2855,
+      "step": 6947
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6459159255027771,
+      "learning_rate": 1.5870595637192535e-05,
+      "loss": 3.268,
+      "step": 6948
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.669853150844574,
+      "learning_rate": 1.5834271996879644e-05,
+      "loss": 3.1253,
+      "step": 6949
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6710006594657898,
+      "learning_rate": 1.579798861237003e-05,
+      "loss": 3.1732,
+      "step": 6950
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6368153691291809,
+      "learning_rate": 1.5761745489901307e-05,
+      "loss": 3.28,
+      "step": 6951
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6223729848861694,
+      "learning_rate": 1.5725542635704026e-05,
+      "loss": 3.1827,
+      "step": 6952
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6653406023979187,
+      "learning_rate": 1.5689380056001927e-05,
+      "loss": 3.3634,
+      "step": 6953
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7627063393592834,
+      "learning_rate": 1.5653257757011763e-05,
+      "loss": 3.3921,
+      "step": 6954
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6704054474830627,
+      "learning_rate": 1.5617175744943368e-05,
+      "loss": 3.3215,
+      "step": 6955
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6059884428977966,
+      "learning_rate": 1.5581134025999644e-05,
+      "loss": 3.3057,
+      "step": 6956
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6775001287460327,
+      "learning_rate": 1.5545132606376605e-05,
+      "loss": 3.3801,
+      "step": 6957
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6409911513328552,
+      "learning_rate": 1.5509171492263302e-05,
+      "loss": 3.2189,
+      "step": 6958
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6561943888664246,
+      "learning_rate": 1.5473250689841843e-05,
+      "loss": 3.3577,
+      "step": 6959
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6522544026374817,
+      "learning_rate": 1.5437370205287515e-05,
+      "loss": 3.4143,
+      "step": 6960
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6523046493530273,
+      "learning_rate": 1.540153004476852e-05,
+      "loss": 3.1973,
+      "step": 6961
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6480640769004822,
+      "learning_rate": 1.5365730214446204e-05,
+      "loss": 3.1805,
+      "step": 6962
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6580466628074646,
+      "learning_rate": 1.5329970720474985e-05,
+      "loss": 3.3363,
+      "step": 6963
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6183872818946838,
+      "learning_rate": 1.52942515690023e-05,
+      "loss": 3.3039,
+      "step": 6964
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6598934531211853,
+      "learning_rate": 1.5258572766168738e-05,
+      "loss": 3.3041,
+      "step": 6965
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7518994808197021,
+      "learning_rate": 1.5222934318107839e-05,
+      "loss": 3.4231,
+      "step": 6966
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6579861640930176,
+      "learning_rate": 1.5187336230946285e-05,
+      "loss": 3.3033,
+      "step": 6967
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6535706520080566,
+      "learning_rate": 1.5151778510803877e-05,
+      "loss": 3.3467,
+      "step": 6968
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6427934765815735,
+      "learning_rate": 1.5116261163793332e-05,
+      "loss": 3.1088,
+      "step": 6969
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6205170750617981,
+      "learning_rate": 1.5080784196020491e-05,
+      "loss": 3.2518,
+      "step": 6970
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6465139389038086,
+      "learning_rate": 1.5045347613584253e-05,
+      "loss": 3.3603,
+      "step": 6971
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6643823385238647,
+      "learning_rate": 1.5009951422576607e-05,
+      "loss": 3.2904,
+      "step": 6972
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6664116382598877,
+      "learning_rate": 1.4974595629082488e-05,
+      "loss": 3.3235,
+      "step": 6973
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6498314142227173,
+      "learning_rate": 1.4939280239180091e-05,
+      "loss": 3.29,
+      "step": 6974
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6376214027404785,
+      "learning_rate": 1.4904005258940424e-05,
+      "loss": 3.3026,
+      "step": 6975
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6093677878379822,
+      "learning_rate": 1.4868770694427768e-05,
+      "loss": 3.2407,
+      "step": 6976
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6138312220573425,
+      "learning_rate": 1.4833576551699285e-05,
+      "loss": 3.258,
+      "step": 6977
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6545621752738953,
+      "learning_rate": 1.4798422836805298e-05,
+      "loss": 3.3727,
+      "step": 6978
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6456170678138733,
+      "learning_rate": 1.4763309555789111e-05,
+      "loss": 3.2684,
+      "step": 6979
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6681527495384216,
+      "learning_rate": 1.4728236714687066e-05,
+      "loss": 3.498,
+      "step": 6980
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.644694447517395,
+      "learning_rate": 1.4693204319528696e-05,
+      "loss": 3.1514,
+      "step": 6981
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6642745137214661,
+      "learning_rate": 1.4658212376336384e-05,
+      "loss": 3.3157,
+      "step": 6982
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6468070149421692,
+      "learning_rate": 1.46232608911257e-05,
+      "loss": 3.3904,
+      "step": 6983
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6560948491096497,
+      "learning_rate": 1.4588349869905149e-05,
+      "loss": 3.3689,
+      "step": 6984
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6244742274284363,
+      "learning_rate": 1.4553479318676398e-05,
+      "loss": 3.3091,
+      "step": 6985
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6304239630699158,
+      "learning_rate": 1.45186492434341e-05,
+      "loss": 3.306,
+      "step": 6986
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6679442524909973,
+      "learning_rate": 1.4483859650165937e-05,
+      "loss": 3.3438,
+      "step": 6987
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6513766050338745,
+      "learning_rate": 1.4449110544852596e-05,
+      "loss": 3.266,
+      "step": 6988
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6204814910888672,
+      "learning_rate": 1.4414401933467907e-05,
+      "loss": 3.3811,
+      "step": 6989
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6290676593780518,
+      "learning_rate": 1.4379733821978686e-05,
+      "loss": 3.3356,
+      "step": 6990
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6321442127227783,
+      "learning_rate": 1.4345106216344772e-05,
+      "loss": 3.2128,
+      "step": 6991
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6475712656974792,
+      "learning_rate": 1.4310519122519045e-05,
+      "loss": 3.4042,
+      "step": 6992
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6464070677757263,
+      "learning_rate": 1.4275972546447412e-05,
+      "loss": 3.333,
+      "step": 6993
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6454011797904968,
+      "learning_rate": 1.4241466494068822e-05,
+      "loss": 3.1697,
+      "step": 6994
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6559916734695435,
+      "learning_rate": 1.4207000971315276e-05,
+      "loss": 3.2271,
+      "step": 6995
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7605160474777222,
+      "learning_rate": 1.4172575984111869e-05,
+      "loss": 3.2859,
+      "step": 6996
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6258525252342224,
+      "learning_rate": 1.4138191538376587e-05,
+      "loss": 3.3107,
+      "step": 6997
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6340823769569397,
+      "learning_rate": 1.4103847640020511e-05,
+      "loss": 3.2286,
+      "step": 6998
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6625699996948242,
+      "learning_rate": 1.4069544294947779e-05,
+      "loss": 3.1828,
+      "step": 6999
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.640891432762146,
+      "learning_rate": 1.4035281509055531e-05,
+      "loss": 3.2585,
+      "step": 7000
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6654657125473022,
+      "learning_rate": 1.4001059288233892e-05,
+      "loss": 3.3142,
+      "step": 7001
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7290911674499512,
+      "learning_rate": 1.3966877638366127e-05,
+      "loss": 3.4767,
+      "step": 7002
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6988948583602905,
+      "learning_rate": 1.3932736565328396e-05,
+      "loss": 3.2548,
+      "step": 7003
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6619759202003479,
+      "learning_rate": 1.389863607498998e-05,
+      "loss": 3.3499,
+      "step": 7004
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6396152377128601,
+      "learning_rate": 1.3864576173213183e-05,
+      "loss": 3.2954,
+      "step": 7005
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6701951026916504,
+      "learning_rate": 1.3830556865853244e-05,
+      "loss": 3.2475,
+      "step": 7006
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6315203309059143,
+      "learning_rate": 1.3796578158758483e-05,
+      "loss": 3.243,
+      "step": 7007
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6182517409324646,
+      "learning_rate": 1.3762640057770253e-05,
+      "loss": 3.2879,
+      "step": 7008
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6174279451370239,
+      "learning_rate": 1.3728742568722864e-05,
+      "loss": 3.1597,
+      "step": 7009
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6286531686782837,
+      "learning_rate": 1.369488569744376e-05,
+      "loss": 3.2261,
+      "step": 7010
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6284041404724121,
+      "learning_rate": 1.366106944975326e-05,
+      "loss": 3.2638,
+      "step": 7011
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.711616575717926,
+      "learning_rate": 1.3627293831464771e-05,
+      "loss": 3.3522,
+      "step": 7012
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.649397075176239,
+      "learning_rate": 1.3593558848384785e-05,
+      "loss": 3.3519,
+      "step": 7013
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6184853315353394,
+      "learning_rate": 1.3559864506312691e-05,
+      "loss": 3.159,
+      "step": 7014
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6492748856544495,
+      "learning_rate": 1.352621081104094e-05,
+      "loss": 3.2205,
+      "step": 7015
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6151764988899231,
+      "learning_rate": 1.3492597768354959e-05,
+      "loss": 3.3327,
+      "step": 7016
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6303196549415588,
+      "learning_rate": 1.3459025384033264e-05,
+      "loss": 3.1337,
+      "step": 7017
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6831106543540955,
+      "learning_rate": 1.3425493663847349e-05,
+      "loss": 3.1622,
+      "step": 7018
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6606957912445068,
+      "learning_rate": 1.339200261356166e-05,
+      "loss": 3.1899,
+      "step": 7019
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6598381400108337,
+      "learning_rate": 1.335855223893373e-05,
+      "loss": 3.2937,
+      "step": 7020
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6386822462081909,
+      "learning_rate": 1.3325142545714014e-05,
+      "loss": 3.294,
+      "step": 7021
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6143596768379211,
+      "learning_rate": 1.3291773539646112e-05,
+      "loss": 3.1769,
+      "step": 7022
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6505967378616333,
+      "learning_rate": 1.3258445226466464e-05,
+      "loss": 3.2047,
+      "step": 7023
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6997924447059631,
+      "learning_rate": 1.3225157611904625e-05,
+      "loss": 3.4018,
+      "step": 7024
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6924464106559753,
+      "learning_rate": 1.3191910701683129e-05,
+      "loss": 3.3137,
+      "step": 7025
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6976944804191589,
+      "learning_rate": 1.3158704501517516e-05,
+      "loss": 3.2224,
+      "step": 7026
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6997489333152771,
+      "learning_rate": 1.31255390171163e-05,
+      "loss": 3.3012,
+      "step": 7027
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.641872227191925,
+      "learning_rate": 1.3092414254181006e-05,
+      "loss": 3.2138,
+      "step": 7028
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6068212389945984,
+      "learning_rate": 1.3059330218406162e-05,
+      "loss": 3.2716,
+      "step": 7029
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6634566187858582,
+      "learning_rate": 1.3026286915479273e-05,
+      "loss": 3.3622,
+      "step": 7030
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6330064535140991,
+      "learning_rate": 1.2993284351080909e-05,
+      "loss": 3.2639,
+      "step": 7031
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6173626780509949,
+      "learning_rate": 1.296032253088461e-05,
+      "loss": 3.2587,
+      "step": 7032
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7010987401008606,
+      "learning_rate": 1.2927401460556876e-05,
+      "loss": 3.2308,
+      "step": 7033
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6823770999908447,
+      "learning_rate": 1.2894521145757205e-05,
+      "loss": 3.3011,
+      "step": 7034
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6275352239608765,
+      "learning_rate": 1.2861681592138103e-05,
+      "loss": 3.1642,
+      "step": 7035
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6606463193893433,
+      "learning_rate": 1.282888280534511e-05,
+      "loss": 3.3432,
+      "step": 7036
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5987416505813599,
+      "learning_rate": 1.2796124791016605e-05,
+      "loss": 3.1984,
+      "step": 7037
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6557232737541199,
+      "learning_rate": 1.2763407554784223e-05,
+      "loss": 3.2765,
+      "step": 7038
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7637114524841309,
+      "learning_rate": 1.2730731102272352e-05,
+      "loss": 3.3681,
+      "step": 7039
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6272114515304565,
+      "learning_rate": 1.2698095439098445e-05,
+      "loss": 3.1892,
+      "step": 7040
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6716676950454712,
+      "learning_rate": 1.2665500570872984e-05,
+      "loss": 3.3333,
+      "step": 7041
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7103142142295837,
+      "learning_rate": 1.2632946503199406e-05,
+      "loss": 3.2958,
+      "step": 7042
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6846051216125488,
+      "learning_rate": 1.260043324167412e-05,
+      "loss": 3.3182,
+      "step": 7043
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6335909366607666,
+      "learning_rate": 1.2567960791886518e-05,
+      "loss": 3.3162,
+      "step": 7044
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6781107783317566,
+      "learning_rate": 1.2535529159418968e-05,
+      "loss": 3.2821,
+      "step": 7045
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5994791984558105,
+      "learning_rate": 1.2503138349846926e-05,
+      "loss": 3.3098,
+      "step": 7046
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6859517693519592,
+      "learning_rate": 1.2470788368738717e-05,
+      "loss": 3.1433,
+      "step": 7047
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6935350298881531,
+      "learning_rate": 1.2438479221655641e-05,
+      "loss": 3.3322,
+      "step": 7048
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6661397218704224,
+      "learning_rate": 1.2406210914152005e-05,
+      "loss": 3.213,
+      "step": 7049
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6652267575263977,
+      "learning_rate": 1.2373983451775179e-05,
+      "loss": 3.1398,
+      "step": 7050
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.652113139629364,
+      "learning_rate": 1.2341796840065366e-05,
+      "loss": 3.3114,
+      "step": 7051
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6672348976135254,
+      "learning_rate": 1.230965108455584e-05,
+      "loss": 3.364,
+      "step": 7052
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6418783664703369,
+      "learning_rate": 1.227754619077287e-05,
+      "loss": 3.3023,
+      "step": 7053
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.637625515460968,
+      "learning_rate": 1.2245482164235627e-05,
+      "loss": 3.2907,
+      "step": 7054
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6356735825538635,
+      "learning_rate": 1.2213459010456285e-05,
+      "loss": 3.1777,
+      "step": 7055
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6418339014053345,
+      "learning_rate": 1.2181476734939968e-05,
+      "loss": 3.2971,
+      "step": 7056
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7009538412094116,
+      "learning_rate": 1.2149535343184858e-05,
+      "loss": 3.3479,
+      "step": 7057
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6452652812004089,
+      "learning_rate": 1.2117634840681984e-05,
+      "loss": 3.1921,
+      "step": 7058
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6060606837272644,
+      "learning_rate": 1.2085775232915485e-05,
+      "loss": 3.3612,
+      "step": 7059
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6534036993980408,
+      "learning_rate": 1.2053956525362314e-05,
+      "loss": 3.3251,
+      "step": 7060
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6830912828445435,
+      "learning_rate": 1.2022178723492566e-05,
+      "loss": 3.3881,
+      "step": 7061
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6198063492774963,
+      "learning_rate": 1.1990441832769178e-05,
+      "loss": 3.2956,
+      "step": 7062
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6613783240318298,
+      "learning_rate": 1.195874585864809e-05,
+      "loss": 3.2565,
+      "step": 7063
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6410684585571289,
+      "learning_rate": 1.1927090806578195e-05,
+      "loss": 3.2802,
+      "step": 7064
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6888789534568787,
+      "learning_rate": 1.189547668200136e-05,
+      "loss": 3.2388,
+      "step": 7065
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6376722455024719,
+      "learning_rate": 1.1863903490352379e-05,
+      "loss": 3.3282,
+      "step": 7066
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6477070450782776,
+      "learning_rate": 1.1832371237059158e-05,
+      "loss": 3.2316,
+      "step": 7067
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6830076575279236,
+      "learning_rate": 1.1800879927542335e-05,
+      "loss": 3.3049,
+      "step": 7068
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6432408690452576,
+      "learning_rate": 1.1769429567215773e-05,
+      "loss": 3.1772,
+      "step": 7069
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6343344449996948,
+      "learning_rate": 1.1738020161486035e-05,
+      "loss": 3.153,
+      "step": 7070
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6284183263778687,
+      "learning_rate": 1.1706651715752803e-05,
+      "loss": 3.2556,
+      "step": 7071
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6398439407348633,
+      "learning_rate": 1.1675324235408707e-05,
+      "loss": 3.3268,
+      "step": 7072
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6415259838104248,
+      "learning_rate": 1.164403772583919e-05,
+      "loss": 3.3415,
+      "step": 7073
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6458741426467896,
+      "learning_rate": 1.1612792192422922e-05,
+      "loss": 3.2342,
+      "step": 7074
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6145205497741699,
+      "learning_rate": 1.1581587640531272e-05,
+      "loss": 3.2521,
+      "step": 7075
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6816922426223755,
+      "learning_rate": 1.1550424075528697e-05,
+      "loss": 3.2076,
+      "step": 7076
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7199867367744446,
+      "learning_rate": 1.1519301502772556e-05,
+      "loss": 3.3487,
+      "step": 7077
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6625239253044128,
+      "learning_rate": 1.1488219927613202e-05,
+      "loss": 3.2905,
+      "step": 7078
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.615695059299469,
+      "learning_rate": 1.145717935539392e-05,
+      "loss": 3.3327,
+      "step": 7079
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.669804573059082,
+      "learning_rate": 1.1426179791450913e-05,
+      "loss": 3.3124,
+      "step": 7080
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6174100041389465,
+      "learning_rate": 1.1395221241113363e-05,
+      "loss": 3.2424,
+      "step": 7081
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6850603818893433,
+      "learning_rate": 1.1364303709703482e-05,
+      "loss": 3.3598,
+      "step": 7082
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6397016048431396,
+      "learning_rate": 1.1333427202536273e-05,
+      "loss": 3.2598,
+      "step": 7083
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6480134129524231,
+      "learning_rate": 1.1302591724919791e-05,
+      "loss": 3.3201,
+      "step": 7084
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6504730582237244,
+      "learning_rate": 1.1271797282154994e-05,
+      "loss": 3.28,
+      "step": 7085
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.646886944770813,
+      "learning_rate": 1.1241043879535811e-05,
+      "loss": 3.1932,
+      "step": 7086
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6623911261558533,
+      "learning_rate": 1.1210331522349126e-05,
+      "loss": 3.2725,
+      "step": 7087
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6475020051002502,
+      "learning_rate": 1.1179660215874715e-05,
+      "loss": 3.2217,
+      "step": 7088
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6451123952865601,
+      "learning_rate": 1.1149029965385416e-05,
+      "loss": 3.1674,
+      "step": 7089
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7544106245040894,
+      "learning_rate": 1.1118440776146821e-05,
+      "loss": 3.206,
+      "step": 7090
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6707612872123718,
+      "learning_rate": 1.1087892653417642e-05,
+      "loss": 3.2419,
+      "step": 7091
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6471689343452454,
+      "learning_rate": 1.10573856024494e-05,
+      "loss": 3.1596,
+      "step": 7092
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7012263536453247,
+      "learning_rate": 1.1026919628486647e-05,
+      "loss": 3.2828,
+      "step": 7093
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6403650641441345,
+      "learning_rate": 1.0996494736766782e-05,
+      "loss": 3.2227,
+      "step": 7094
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6718841791152954,
+      "learning_rate": 1.0966110932520285e-05,
+      "loss": 3.1116,
+      "step": 7095
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5933142900466919,
+      "learning_rate": 1.0935768220970393e-05,
+      "loss": 3.1487,
+      "step": 7096
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6481373906135559,
+      "learning_rate": 1.0905466607333465e-05,
+      "loss": 3.2511,
+      "step": 7097
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6603955030441284,
+      "learning_rate": 1.0875206096818607e-05,
+      "loss": 3.1641,
+      "step": 7098
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6217958331108093,
+      "learning_rate": 1.0844986694628022e-05,
+      "loss": 3.2658,
+      "step": 7099
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6261694431304932,
+      "learning_rate": 1.081480840595675e-05,
+      "loss": 3.2886,
+      "step": 7100
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6567366123199463,
+      "learning_rate": 1.0784671235992777e-05,
+      "loss": 3.2645,
+      "step": 7101
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6423852443695068,
+      "learning_rate": 1.0754575189917015e-05,
+      "loss": 3.2489,
+      "step": 7102
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6532287001609802,
+      "learning_rate": 1.0724520272903382e-05,
+      "loss": 3.3499,
+      "step": 7103
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.67435222864151,
+      "learning_rate": 1.0694506490118632e-05,
+      "loss": 3.1951,
+      "step": 7104
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6489763855934143,
+      "learning_rate": 1.0664533846722447e-05,
+      "loss": 3.2103,
+      "step": 7105
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6705418825149536,
+      "learning_rate": 1.0634602347867533e-05,
+      "loss": 3.2122,
+      "step": 7106
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6227537393569946,
+      "learning_rate": 1.0604711998699445e-05,
+      "loss": 3.133,
+      "step": 7107
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6649758219718933,
+      "learning_rate": 1.0574862804356683e-05,
+      "loss": 3.3684,
+      "step": 7108
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7490100860595703,
+      "learning_rate": 1.0545054769970614e-05,
+      "loss": 3.4781,
+      "step": 7109
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6477136611938477,
+      "learning_rate": 1.0515287900665666e-05,
+      "loss": 3.1671,
+      "step": 7110
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6387073397636414,
+      "learning_rate": 1.0485562201559079e-05,
+      "loss": 3.3198,
+      "step": 7111
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.648479700088501,
+      "learning_rate": 1.0455877677761044e-05,
+      "loss": 3.2667,
+      "step": 7112
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6808549761772156,
+      "learning_rate": 1.0426234334374647e-05,
+      "loss": 3.2661,
+      "step": 7113
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6373841166496277,
+      "learning_rate": 1.0396632176495946e-05,
+      "loss": 3.2111,
+      "step": 7114
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6293783187866211,
+      "learning_rate": 1.0367071209213902e-05,
+      "loss": 3.3596,
+      "step": 7115
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6526897549629211,
+      "learning_rate": 1.0337551437610365e-05,
+      "loss": 3.3849,
+      "step": 7116
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6583970785140991,
+      "learning_rate": 1.0308072866760137e-05,
+      "loss": 3.2599,
+      "step": 7117
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6063961982727051,
+      "learning_rate": 1.027863550173097e-05,
+      "loss": 3.3776,
+      "step": 7118
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6646319627761841,
+      "learning_rate": 1.0249239347583428e-05,
+      "loss": 3.2485,
+      "step": 7119
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7042518854141235,
+      "learning_rate": 1.0219884409371077e-05,
+      "loss": 3.2221,
+      "step": 7120
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.685055673122406,
+      "learning_rate": 1.0190570692140355e-05,
+      "loss": 3.2521,
+      "step": 7121
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6594273447990417,
+      "learning_rate": 1.0161298200930647e-05,
+      "loss": 3.3004,
+      "step": 7122
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.61384117603302,
+      "learning_rate": 1.0132066940774203e-05,
+      "loss": 3.2558,
+      "step": 7123
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6446554660797119,
+      "learning_rate": 1.010287691669623e-05,
+      "loss": 3.265,
+      "step": 7124
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6155528426170349,
+      "learning_rate": 1.0073728133714877e-05,
+      "loss": 3.23,
+      "step": 7125
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5826809406280518,
+      "learning_rate": 1.0044620596841136e-05,
+      "loss": 3.171,
+      "step": 7126
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7251071333885193,
+      "learning_rate": 1.0015554311078895e-05,
+      "loss": 3.3523,
+      "step": 7127
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6796685457229614,
+      "learning_rate": 9.986529281425016e-06,
+      "loss": 3.3612,
+      "step": 7128
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6957389712333679,
+      "learning_rate": 9.957545512869231e-06,
+      "loss": 3.2444,
+      "step": 7129
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6471418738365173,
+      "learning_rate": 9.928603010394138e-06,
+      "loss": 3.2354,
+      "step": 7130
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6198047995567322,
+      "learning_rate": 9.899701778975395e-06,
+      "loss": 3.1523,
+      "step": 7131
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6490558981895447,
+      "learning_rate": 9.870841823581362e-06,
+      "loss": 3.191,
+      "step": 7132
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6650616526603699,
+      "learning_rate": 9.842023149173428e-06,
+      "loss": 3.3283,
+      "step": 7133
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6575379371643066,
+      "learning_rate": 9.813245760705886e-06,
+      "loss": 3.3364,
+      "step": 7134
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6252859830856323,
+      "learning_rate": 9.784509663125884e-06,
+      "loss": 3.3314,
+      "step": 7135
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6415067911148071,
+      "learning_rate": 9.755814861373502e-06,
+      "loss": 3.0821,
+      "step": 7136
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6143196225166321,
+      "learning_rate": 9.727161360381681e-06,
+      "loss": 3.3058,
+      "step": 7137
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6523025035858154,
+      "learning_rate": 9.69854916507626e-06,
+      "loss": 3.2572,
+      "step": 7138
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6369534134864807,
+      "learning_rate": 9.669978280376107e-06,
+      "loss": 3.1735,
+      "step": 7139
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6777133345603943,
+      "learning_rate": 9.641448711192796e-06,
+      "loss": 3.1831,
+      "step": 7140
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6214920878410339,
+      "learning_rate": 9.61296046243096e-06,
+      "loss": 3.211,
+      "step": 7141
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6660655736923218,
+      "learning_rate": 9.584513538987983e-06,
+      "loss": 3.2759,
+      "step": 7142
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6243155002593994,
+      "learning_rate": 9.556107945754316e-06,
+      "loss": 3.3113,
+      "step": 7143
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6599859595298767,
+      "learning_rate": 9.527743687613138e-06,
+      "loss": 3.2715,
+      "step": 7144
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6295204758644104,
+      "learning_rate": 9.499420769440576e-06,
+      "loss": 3.2068,
+      "step": 7145
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6949333548545837,
+      "learning_rate": 9.471139196105732e-06,
+      "loss": 3.3363,
+      "step": 7146
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6422164440155029,
+      "learning_rate": 9.442898972470526e-06,
+      "loss": 3.2773,
+      "step": 7147
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6916665434837341,
+      "learning_rate": 9.414700103389768e-06,
+      "loss": 3.2584,
+      "step": 7148
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.722175121307373,
+      "learning_rate": 9.386542593711162e-06,
+      "loss": 3.3599,
+      "step": 7149
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6117916107177734,
+      "learning_rate": 9.358426448275309e-06,
+      "loss": 3.1653,
+      "step": 7150
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6152571439743042,
+      "learning_rate": 9.330351671915676e-06,
+      "loss": 3.3316,
+      "step": 7151
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6863966584205627,
+      "learning_rate": 9.302318269458682e-06,
+      "loss": 3.1733,
+      "step": 7152
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6250137686729431,
+      "learning_rate": 9.274326245723607e-06,
+      "loss": 3.1781,
+      "step": 7153
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6183449625968933,
+      "learning_rate": 9.246375605522578e-06,
+      "loss": 3.4088,
+      "step": 7154
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6490113735198975,
+      "learning_rate": 9.218466353660637e-06,
+      "loss": 3.2649,
+      "step": 7155
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6815605759620667,
+      "learning_rate": 9.190598494935726e-06,
+      "loss": 3.1625,
+      "step": 7156
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.623877227306366,
+      "learning_rate": 9.162772034138623e-06,
+      "loss": 3.3315,
+      "step": 7157
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.654930591583252,
+      "learning_rate": 9.134986976053028e-06,
+      "loss": 3.1119,
+      "step": 7158
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6280533075332642,
+      "learning_rate": 9.10724332545551e-06,
+      "loss": 3.3371,
+      "step": 7159
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6446752548217773,
+      "learning_rate": 9.079541087115506e-06,
+      "loss": 3.271,
+      "step": 7160
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6848363876342773,
+      "learning_rate": 9.051880265795426e-06,
+      "loss": 3.2908,
+      "step": 7161
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6229819655418396,
+      "learning_rate": 9.024260866250439e-06,
+      "loss": 3.2014,
+      "step": 7162
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6721832156181335,
+      "learning_rate": 8.996682893228609e-06,
+      "loss": 3.1534,
+      "step": 7163
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6590028405189514,
+      "learning_rate": 8.969146351470974e-06,
+      "loss": 3.3421,
+      "step": 7164
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6596424579620361,
+      "learning_rate": 8.941651245711336e-06,
+      "loss": 3.1913,
+      "step": 7165
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6490575075149536,
+      "learning_rate": 8.914197580676409e-06,
+      "loss": 3.2886,
+      "step": 7166
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.643104612827301,
+      "learning_rate": 8.886785361085865e-06,
+      "loss": 3.2904,
+      "step": 7167
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.633247971534729,
+      "learning_rate": 8.859414591652126e-06,
+      "loss": 3.1506,
+      "step": 7168
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6761441826820374,
+      "learning_rate": 8.832085277080571e-06,
+      "loss": 3.234,
+      "step": 7169
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6587713360786438,
+      "learning_rate": 8.804797422069383e-06,
+      "loss": 3.2586,
+      "step": 7170
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6706406474113464,
+      "learning_rate": 8.777551031309727e-06,
+      "loss": 3.2162,
+      "step": 7171
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6373216509819031,
+      "learning_rate": 8.750346109485525e-06,
+      "loss": 3.2112,
+      "step": 7172
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6625058650970459,
+      "learning_rate": 8.723182661273615e-06,
+      "loss": 3.3164,
+      "step": 7173
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6026141047477722,
+      "learning_rate": 8.696060691343765e-06,
+      "loss": 3.1936,
+      "step": 7174
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6548628211021423,
+      "learning_rate": 8.668980204358496e-06,
+      "loss": 3.3851,
+      "step": 7175
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6361516714096069,
+      "learning_rate": 8.641941204973274e-06,
+      "loss": 3.3051,
+      "step": 7176
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6361994743347168,
+      "learning_rate": 8.61494369783644e-06,
+      "loss": 3.3744,
+      "step": 7177
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6931582093238831,
+      "learning_rate": 8.587987687589172e-06,
+      "loss": 3.3163,
+      "step": 7178
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6600695252418518,
+      "learning_rate": 8.561073178865453e-06,
+      "loss": 3.3885,
+      "step": 7179
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6554650068283081,
+      "learning_rate": 8.534200176292305e-06,
+      "loss": 3.2736,
+      "step": 7180
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6693041920661926,
+      "learning_rate": 8.507368684489397e-06,
+      "loss": 3.1823,
+      "step": 7181
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6274213790893555,
+      "learning_rate": 8.48057870806951e-06,
+      "loss": 3.292,
+      "step": 7182
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6852020621299744,
+      "learning_rate": 8.45383025163804e-06,
+      "loss": 3.306,
+      "step": 7183
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6665197610855103,
+      "learning_rate": 8.427123319793395e-06,
+      "loss": 3.3709,
+      "step": 7184
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6755871772766113,
+      "learning_rate": 8.400457917126819e-06,
+      "loss": 3.1938,
+      "step": 7185
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6387326717376709,
+      "learning_rate": 8.373834048222394e-06,
+      "loss": 3.3682,
+      "step": 7186
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6742923259735107,
+      "learning_rate": 8.347251717657018e-06,
+      "loss": 3.2947,
+      "step": 7187
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6418575644493103,
+      "learning_rate": 8.320710930000586e-06,
+      "loss": 3.2402,
+      "step": 7188
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6079849600791931,
+      "learning_rate": 8.294211689815729e-06,
+      "loss": 3.3532,
+      "step": 7189
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6370933055877686,
+      "learning_rate": 8.267754001657969e-06,
+      "loss": 3.2438,
+      "step": 7190
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6640612483024597,
+      "learning_rate": 8.241337870075721e-06,
+      "loss": 3.308,
+      "step": 7191
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6398223042488098,
+      "learning_rate": 8.214963299610189e-06,
+      "loss": 3.4203,
+      "step": 7192
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6995474696159363,
+      "learning_rate": 8.188630294795469e-06,
+      "loss": 3.2194,
+      "step": 7193
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5878394842147827,
+      "learning_rate": 8.16233886015852e-06,
+      "loss": 3.23,
+      "step": 7194
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.66357421875,
+      "learning_rate": 8.136089000219144e-06,
+      "loss": 3.2357,
+      "step": 7195
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6442654132843018,
+      "learning_rate": 8.10988071949001e-06,
+      "loss": 3.1958,
+      "step": 7196
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6443668007850647,
+      "learning_rate": 8.083714022476568e-06,
+      "loss": 3.2293,
+      "step": 7197
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6623659729957581,
+      "learning_rate": 8.057588913677277e-06,
+      "loss": 3.0882,
+      "step": 7198
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6151493787765503,
+      "learning_rate": 8.031505397583267e-06,
+      "loss": 3.2242,
+      "step": 7199
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5729960799217224,
+      "learning_rate": 8.005463478678615e-06,
+      "loss": 3.2981,
+      "step": 7200
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.695530354976654,
+      "learning_rate": 7.979463161440242e-06,
+      "loss": 3.0923,
+      "step": 7201
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6356866359710693,
+      "learning_rate": 7.953504450337879e-06,
+      "loss": 3.2798,
+      "step": 7202
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7238433361053467,
+      "learning_rate": 7.927587349834148e-06,
+      "loss": 3.2346,
+      "step": 7203
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5977983474731445,
+      "learning_rate": 7.901711864384515e-06,
+      "loss": 3.2292,
+      "step": 7204
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6451910138130188,
+      "learning_rate": 7.875877998437226e-06,
+      "loss": 3.2024,
+      "step": 7205
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6787285208702087,
+      "learning_rate": 7.850085756433478e-06,
+      "loss": 3.2673,
+      "step": 7206
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6958125233650208,
+      "learning_rate": 7.824335142807198e-06,
+      "loss": 3.2676,
+      "step": 7207
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6704130172729492,
+      "learning_rate": 7.79862616198529e-06,
+      "loss": 3.2651,
+      "step": 7208
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6035682559013367,
+      "learning_rate": 7.772958818387326e-06,
+      "loss": 3.4288,
+      "step": 7209
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6599581241607666,
+      "learning_rate": 7.747333116425947e-06,
+      "loss": 3.3845,
+      "step": 7210
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6396064758300781,
+      "learning_rate": 7.721749060506406e-06,
+      "loss": 3.1495,
+      "step": 7211
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6582738757133484,
+      "learning_rate": 7.696206655026933e-06,
+      "loss": 3.3308,
+      "step": 7212
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6420777440071106,
+      "learning_rate": 7.670705904378572e-06,
+      "loss": 3.2123,
+      "step": 7213
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6579254865646362,
+      "learning_rate": 7.645246812945206e-06,
+      "loss": 3.3212,
+      "step": 7214
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6444306373596191,
+      "learning_rate": 7.6198293851034715e-06,
+      "loss": 3.121,
+      "step": 7215
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7798821926116943,
+      "learning_rate": 7.594453625223013e-06,
+      "loss": 3.1915,
+      "step": 7216
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6160184144973755,
+      "learning_rate": 7.569119537666175e-06,
+      "loss": 3.268,
+      "step": 7217
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7422967553138733,
+      "learning_rate": 7.543827126788194e-06,
+      "loss": 3.2423,
+      "step": 7218
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6377444863319397,
+      "learning_rate": 7.5185763969371215e-06,
+      "loss": 3.1848,
+      "step": 7219
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6461215615272522,
+      "learning_rate": 7.493367352453873e-06,
+      "loss": 3.2925,
+      "step": 7220
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6127544045448303,
+      "learning_rate": 7.468199997672148e-06,
+      "loss": 3.1988,
+      "step": 7221
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6574669480323792,
+      "learning_rate": 7.443074336918487e-06,
+      "loss": 3.2854,
+      "step": 7222
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6241658329963684,
+      "learning_rate": 7.417990374512296e-06,
+      "loss": 3.3113,
+      "step": 7223
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6702455282211304,
+      "learning_rate": 7.392948114765846e-06,
+      "loss": 3.2695,
+      "step": 7224
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6530889272689819,
+      "learning_rate": 7.36794756198414e-06,
+      "loss": 3.2646,
+      "step": 7225
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.681586503982544,
+      "learning_rate": 7.3429887204650994e-06,
+      "loss": 3.2925,
+      "step": 7226
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6429810523986816,
+      "learning_rate": 7.318071594499403e-06,
+      "loss": 3.2615,
+      "step": 7227
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6187373399734497,
+      "learning_rate": 7.293196188370627e-06,
+      "loss": 3.2572,
+      "step": 7228
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6678377389907837,
+      "learning_rate": 7.268362506355125e-06,
+      "loss": 3.3227,
+      "step": 7229
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6457262635231018,
+      "learning_rate": 7.243570552722067e-06,
+      "loss": 3.1622,
+      "step": 7230
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6468750834465027,
+      "learning_rate": 7.2188203317335165e-06,
+      "loss": 3.2397,
+      "step": 7231
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6340529918670654,
+      "learning_rate": 7.194111847644347e-06,
+      "loss": 3.1485,
+      "step": 7232
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6241766214370728,
+      "learning_rate": 7.16944510470216e-06,
+      "loss": 3.2466,
+      "step": 7233
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6593299508094788,
+      "learning_rate": 7.144820107147482e-06,
+      "loss": 3.3613,
+      "step": 7234
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6926560997962952,
+      "learning_rate": 7.120236859213674e-06,
+      "loss": 3.3678,
+      "step": 7235
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6539257168769836,
+      "learning_rate": 7.095695365126858e-06,
+      "loss": 3.217,
+      "step": 7236
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6468613147735596,
+      "learning_rate": 7.071195629105992e-06,
+      "loss": 3.1489,
+      "step": 7237
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.662559986114502,
+      "learning_rate": 7.046737655362845e-06,
+      "loss": 3.3925,
+      "step": 7238
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6389094591140747,
+      "learning_rate": 7.0223214481020535e-06,
+      "loss": 3.2455,
+      "step": 7239
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6499673128128052,
+      "learning_rate": 6.997947011521067e-06,
+      "loss": 3.3585,
+      "step": 7240
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6148930191993713,
+      "learning_rate": 6.973614349810115e-06,
+      "loss": 3.2433,
+      "step": 7241
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6275553703308105,
+      "learning_rate": 6.949323467152269e-06,
+      "loss": 3.2461,
+      "step": 7242
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7042184472084045,
+      "learning_rate": 6.925074367723383e-06,
+      "loss": 3.3104,
+      "step": 7243
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6173494458198547,
+      "learning_rate": 6.90086705569215e-06,
+      "loss": 3.2534,
+      "step": 7244
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6344504952430725,
+      "learning_rate": 6.876701535220131e-06,
+      "loss": 3.1419,
+      "step": 7245
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6360830664634705,
+      "learning_rate": 6.8525778104616685e-06,
+      "loss": 3.4543,
+      "step": 7246
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6810656189918518,
+      "learning_rate": 6.82849588556389e-06,
+      "loss": 3.1706,
+      "step": 7247
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6513180732727051,
+      "learning_rate": 6.804455764666733e-06,
+      "loss": 3.4029,
+      "step": 7248
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6299558877944946,
+      "learning_rate": 6.7804574519030325e-06,
+      "loss": 3.2123,
+      "step": 7249
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6327093839645386,
+      "learning_rate": 6.75650095139832e-06,
+      "loss": 3.322,
+      "step": 7250
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.667273223400116,
+      "learning_rate": 6.732586267270968e-06,
+      "loss": 3.3847,
+      "step": 7251
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6162838935852051,
+      "learning_rate": 6.708713403632299e-06,
+      "loss": 3.2832,
+      "step": 7252
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6861327886581421,
+      "learning_rate": 6.684882364586226e-06,
+      "loss": 3.2907,
+      "step": 7253
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6113225221633911,
+      "learning_rate": 6.661093154229636e-06,
+      "loss": 3.2132,
+      "step": 7254
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6668545603752136,
+      "learning_rate": 6.637345776652176e-06,
+      "loss": 3.3351,
+      "step": 7255
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7097353339195251,
+      "learning_rate": 6.613640235936275e-06,
+      "loss": 3.4091,
+      "step": 7256
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6316132545471191,
+      "learning_rate": 6.589976536157199e-06,
+      "loss": 3.2332,
+      "step": 7257
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.637789785861969,
+      "learning_rate": 6.566354681383002e-06,
+      "loss": 3.2688,
+      "step": 7258
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6592211723327637,
+      "learning_rate": 6.542774675674546e-06,
+      "loss": 3.2023,
+      "step": 7259
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6290187835693359,
+      "learning_rate": 6.519236523085531e-06,
+      "loss": 3.2602,
+      "step": 7260
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.651838481426239,
+      "learning_rate": 6.495740227662445e-06,
+      "loss": 3.3902,
+      "step": 7261
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6233133673667908,
+      "learning_rate": 6.472285793444499e-06,
+      "loss": 3.1979,
+      "step": 7262
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6357033848762512,
+      "learning_rate": 6.448873224463914e-06,
+      "loss": 3.1953,
+      "step": 7263
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.704164981842041,
+      "learning_rate": 6.4255025247454694e-06,
+      "loss": 3.1846,
+      "step": 7264
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6579105854034424,
+      "learning_rate": 6.402173698306924e-06,
+      "loss": 3.3871,
+      "step": 7265
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6748107075691223,
+      "learning_rate": 6.378886749158708e-06,
+      "loss": 3.2592,
+      "step": 7266
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6929618716239929,
+      "learning_rate": 6.355641681304175e-06,
+      "loss": 3.3654,
+      "step": 7267
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.651310920715332,
+      "learning_rate": 6.332438498739435e-06,
+      "loss": 3.28,
+      "step": 7268
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.639167845249176,
+      "learning_rate": 6.309277205453323e-06,
+      "loss": 3.1737,
+      "step": 7269
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6047765612602234,
+      "learning_rate": 6.2861578054276e-06,
+      "loss": 3.2351,
+      "step": 7270
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6707802414894104,
+      "learning_rate": 6.26308030263667e-06,
+      "loss": 3.3016,
+      "step": 7271
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6374141573905945,
+      "learning_rate": 6.240044701047915e-06,
+      "loss": 3.3161,
+      "step": 7272
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6345334649085999,
+      "learning_rate": 6.21705100462136e-06,
+      "loss": 3.356,
+      "step": 7273
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6492778658866882,
+      "learning_rate": 6.194099217309901e-06,
+      "loss": 3.2443,
+      "step": 7274
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.642340898513794,
+      "learning_rate": 6.171189343059269e-06,
+      "loss": 3.255,
+      "step": 7275
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6277442574501038,
+      "learning_rate": 6.14832138580787e-06,
+      "loss": 3.3729,
+      "step": 7276
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6362481713294983,
+      "learning_rate": 6.125495349487003e-06,
+      "loss": 3.3552,
+      "step": 7277
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6513189673423767,
+      "learning_rate": 6.102711238020725e-06,
+      "loss": 3.1394,
+      "step": 7278
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6888836622238159,
+      "learning_rate": 6.079969055325901e-06,
+      "loss": 3.2644,
+      "step": 7279
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6480313539505005,
+      "learning_rate": 6.057268805312127e-06,
+      "loss": 3.1271,
+      "step": 7280
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6510853171348572,
+      "learning_rate": 6.0346104918818645e-06,
+      "loss": 3.3147,
+      "step": 7281
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6157244443893433,
+      "learning_rate": 6.011994118930358e-06,
+      "loss": 3.2939,
+      "step": 7282
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6814156770706177,
+      "learning_rate": 5.9894196903456376e-06,
+      "loss": 3.2736,
+      "step": 7283
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6630598306655884,
+      "learning_rate": 5.966887210008487e-06,
+      "loss": 3.2568,
+      "step": 7284
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6240792274475098,
+      "learning_rate": 5.944396681792474e-06,
+      "loss": 3.3374,
+      "step": 7285
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6431431174278259,
+      "learning_rate": 5.921948109564035e-06,
+      "loss": 3.3313,
+      "step": 7286
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5855844020843506,
+      "learning_rate": 5.899541497182276e-06,
+      "loss": 3.2376,
+      "step": 7287
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6474109888076782,
+      "learning_rate": 5.8771768484992e-06,
+      "loss": 3.2878,
+      "step": 7288
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7335505485534668,
+      "learning_rate": 5.854854167359564e-06,
+      "loss": 3.3169,
+      "step": 7289
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6480274200439453,
+      "learning_rate": 5.832573457600826e-06,
+      "loss": 3.2585,
+      "step": 7290
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.640687882900238,
+      "learning_rate": 5.810334723053367e-06,
+      "loss": 3.3847,
+      "step": 7291
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.64142245054245,
+      "learning_rate": 5.788137967540269e-06,
+      "loss": 3.1795,
+      "step": 7292
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6028220653533936,
+      "learning_rate": 5.765983194877394e-06,
+      "loss": 3.1765,
+      "step": 7293
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.672534704208374,
+      "learning_rate": 5.743870408873419e-06,
+      "loss": 3.2826,
+      "step": 7294
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6606788039207458,
+      "learning_rate": 5.7217996133297476e-06,
+      "loss": 3.1391,
+      "step": 7295
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6740570664405823,
+      "learning_rate": 5.6997708120406775e-06,
+      "loss": 3.2925,
+      "step": 7296
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.668215811252594,
+      "learning_rate": 5.677784008793152e-06,
+      "loss": 3.2491,
+      "step": 7297
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6743031144142151,
+      "learning_rate": 5.655839207367008e-06,
+      "loss": 3.3067,
+      "step": 7298
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6616876125335693,
+      "learning_rate": 5.633936411534729e-06,
+      "loss": 3.2424,
+      "step": 7299
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6089078783988953,
+      "learning_rate": 5.612075625061774e-06,
+      "loss": 3.1401,
+      "step": 7300
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6700180768966675,
+      "learning_rate": 5.590256851706193e-06,
+      "loss": 3.3681,
+      "step": 7301
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6170392632484436,
+      "learning_rate": 5.5684800952188755e-06,
+      "loss": 3.2534,
+      "step": 7302
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6896737813949585,
+      "learning_rate": 5.546745359343547e-06,
+      "loss": 3.1538,
+      "step": 7303
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6510875821113586,
+      "learning_rate": 5.525052647816636e-06,
+      "loss": 3.4175,
+      "step": 7304
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6070569157600403,
+      "learning_rate": 5.503401964367355e-06,
+      "loss": 3.3609,
+      "step": 7305
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6779491901397705,
+      "learning_rate": 5.481793312717753e-06,
+      "loss": 3.288,
+      "step": 7306
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6362120509147644,
+      "learning_rate": 5.460226696582555e-06,
+      "loss": 3.2862,
+      "step": 7307
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.9496941566467285,
+      "learning_rate": 5.43870211966932e-06,
+      "loss": 3.2342,
+      "step": 7308
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7293429970741272,
+      "learning_rate": 5.417219585678423e-06,
+      "loss": 3.3709,
+      "step": 7309
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6373938322067261,
+      "learning_rate": 5.395779098302911e-06,
+      "loss": 3.3523,
+      "step": 7310
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6489766836166382,
+      "learning_rate": 5.374380661228667e-06,
+      "loss": 3.1637,
+      "step": 7311
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6648989915847778,
+      "learning_rate": 5.353024278134333e-06,
+      "loss": 3.2934,
+      "step": 7312
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6867431998252869,
+      "learning_rate": 5.331709952691333e-06,
+      "loss": 3.1268,
+      "step": 7313
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6261318922042847,
+      "learning_rate": 5.3104376885638185e-06,
+      "loss": 3.3558,
+      "step": 7314
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.635741114616394,
+      "learning_rate": 5.289207489408754e-06,
+      "loss": 3.2383,
+      "step": 7315
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6128219962120056,
+      "learning_rate": 5.268019358875803e-06,
+      "loss": 3.3608,
+      "step": 7316
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6505506038665771,
+      "learning_rate": 5.246873300607552e-06,
+      "loss": 3.2795,
+      "step": 7317
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6969970464706421,
+      "learning_rate": 5.225769318239177e-06,
+      "loss": 3.3382,
+      "step": 7318
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6488733887672424,
+      "learning_rate": 5.2047074153987465e-06,
+      "loss": 3.1926,
+      "step": 7319
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6368139982223511,
+      "learning_rate": 5.183687595707032e-06,
+      "loss": 3.2677,
+      "step": 7320
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6244758367538452,
+      "learning_rate": 5.1627098627775594e-06,
+      "loss": 3.2698,
+      "step": 7321
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6960594654083252,
+      "learning_rate": 5.1417742202166665e-06,
+      "loss": 3.4216,
+      "step": 7322
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6352282762527466,
+      "learning_rate": 5.12088067162339e-06,
+      "loss": 3.2368,
+      "step": 7323
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6699858903884888,
+      "learning_rate": 5.100029220589636e-06,
+      "loss": 3.4182,
+      "step": 7324
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.628677487373352,
+      "learning_rate": 5.07921987069998e-06,
+      "loss": 3.2176,
+      "step": 7325
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7023717164993286,
+      "learning_rate": 5.058452625531812e-06,
+      "loss": 3.3189,
+      "step": 7326
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6396088600158691,
+      "learning_rate": 5.037727488655192e-06,
+      "loss": 3.4466,
+      "step": 7327
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6315101385116577,
+      "learning_rate": 5.0170444636331025e-06,
+      "loss": 3.2949,
+      "step": 7328
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6428784728050232,
+      "learning_rate": 4.996403554021145e-06,
+      "loss": 3.2748,
+      "step": 7329
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6924943327903748,
+      "learning_rate": 4.975804763367758e-06,
+      "loss": 3.3769,
+      "step": 7330
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6694445610046387,
+      "learning_rate": 4.955248095214082e-06,
+      "loss": 3.1891,
+      "step": 7331
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6170761585235596,
+      "learning_rate": 4.934733553094068e-06,
+      "loss": 3.3613,
+      "step": 7332
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.675621747970581,
+      "learning_rate": 4.914261140534393e-06,
+      "loss": 3.3318,
+      "step": 7333
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.675338625907898,
+      "learning_rate": 4.8938308610544935e-06,
+      "loss": 3.2877,
+      "step": 7334
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6300745010375977,
+      "learning_rate": 4.8734427181666156e-06,
+      "loss": 3.1573,
+      "step": 7335
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6434711217880249,
+      "learning_rate": 4.853096715375649e-06,
+      "loss": 3.2582,
+      "step": 7336
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6582839488983154,
+      "learning_rate": 4.8327928561793525e-06,
+      "loss": 3.4553,
+      "step": 7337
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6042643785476685,
+      "learning_rate": 4.812531144068183e-06,
+      "loss": 3.1254,
+      "step": 7338
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6512244939804077,
+      "learning_rate": 4.792311582525383e-06,
+      "loss": 3.2403,
+      "step": 7339
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6681080460548401,
+      "learning_rate": 4.772134175026921e-06,
+      "loss": 3.2801,
+      "step": 7340
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6504518985748291,
+      "learning_rate": 4.751998925041495e-06,
+      "loss": 3.2723,
+      "step": 7341
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6400660872459412,
+      "learning_rate": 4.731905836030642e-06,
+      "loss": 3.3012,
+      "step": 7342
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6214766502380371,
+      "learning_rate": 4.711854911448543e-06,
+      "loss": 3.284,
+      "step": 7343
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6600997447967529,
+      "learning_rate": 4.691846154742191e-06,
+      "loss": 3.3809,
+      "step": 7344
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7025464773178101,
+      "learning_rate": 4.671879569351362e-06,
+      "loss": 3.2058,
+      "step": 7345
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6420849561691284,
+      "learning_rate": 4.651955158708532e-06,
+      "loss": 3.1747,
+      "step": 7346
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5835869908332825,
+      "learning_rate": 4.632072926238934e-06,
+      "loss": 3.2178,
+      "step": 7347
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6520231366157532,
+      "learning_rate": 4.612232875360528e-06,
+      "loss": 3.2645,
+      "step": 7348
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6406843066215515,
+      "learning_rate": 4.592435009484086e-06,
+      "loss": 3.439,
+      "step": 7349
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6760379076004028,
+      "learning_rate": 4.57267933201308e-06,
+      "loss": 3.4426,
+      "step": 7350
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5957873463630676,
+      "learning_rate": 4.552965846343709e-06,
+      "loss": 3.3527,
+      "step": 7351
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6975240111351013,
+      "learning_rate": 4.533294555864986e-06,
+      "loss": 3.3895,
+      "step": 7352
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.699790894985199,
+      "learning_rate": 4.513665463958621e-06,
+      "loss": 3.2091,
+      "step": 7353
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6839869022369385,
+      "learning_rate": 4.494078573999111e-06,
+      "loss": 3.2557,
+      "step": 7354
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6462808847427368,
+      "learning_rate": 4.474533889353594e-06,
+      "loss": 3.23,
+      "step": 7355
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6232364773750305,
+      "learning_rate": 4.455031413382104e-06,
+      "loss": 3.2319,
+      "step": 7356
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8861619234085083,
+      "learning_rate": 4.435571149437323e-06,
+      "loss": 3.3775,
+      "step": 7357
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6382137537002563,
+      "learning_rate": 4.416153100864684e-06,
+      "loss": 3.2468,
+      "step": 7358
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6784188747406006,
+      "learning_rate": 4.396777271002378e-06,
+      "loss": 3.2836,
+      "step": 7359
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6564819812774658,
+      "learning_rate": 4.377443663181324e-06,
+      "loss": 3.3778,
+      "step": 7360
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6596285104751587,
+      "learning_rate": 4.358152280725225e-06,
+      "loss": 3.2579,
+      "step": 7361
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6444254517555237,
+      "learning_rate": 4.338903126950483e-06,
+      "loss": 3.2217,
+      "step": 7362
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6608470678329468,
+      "learning_rate": 4.3196962051662014e-06,
+      "loss": 3.0983,
+      "step": 7363
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6407880783081055,
+      "learning_rate": 4.300531518674322e-06,
+      "loss": 3.2262,
+      "step": 7364
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6535629034042358,
+      "learning_rate": 4.281409070769487e-06,
+      "loss": 3.3884,
+      "step": 7365
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.063302755355835,
+      "learning_rate": 4.26232886473904e-06,
+      "loss": 3.2577,
+      "step": 7366
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6154136657714844,
+      "learning_rate": 4.2432909038630785e-06,
+      "loss": 3.1855,
+      "step": 7367
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5912913084030151,
+      "learning_rate": 4.224295191414512e-06,
+      "loss": 3.2211,
+      "step": 7368
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6943714618682861,
+      "learning_rate": 4.205341730658841e-06,
+      "loss": 3.3071,
+      "step": 7369
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6367833614349365,
+      "learning_rate": 4.186430524854429e-06,
+      "loss": 3.2245,
+      "step": 7370
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.658357560634613,
+      "learning_rate": 4.167561577252343e-06,
+      "loss": 3.3038,
+      "step": 7371
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6754440665245056,
+      "learning_rate": 4.148734891096351e-06,
+      "loss": 3.3077,
+      "step": 7372
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6419435143470764,
+      "learning_rate": 4.129950469622945e-06,
+      "loss": 3.3789,
+      "step": 7373
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6624208092689514,
+      "learning_rate": 4.1112083160614326e-06,
+      "loss": 3.1705,
+      "step": 7374
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6808865666389465,
+      "learning_rate": 4.0925084336338455e-06,
+      "loss": 3.1439,
+      "step": 7375
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6722561120986938,
+      "learning_rate": 4.073850825554837e-06,
+      "loss": 3.3548,
+      "step": 7376
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6091050505638123,
+      "learning_rate": 4.0552354950318944e-06,
+      "loss": 3.1668,
+      "step": 7377
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6220724582672119,
+      "learning_rate": 4.03666244526521e-06,
+      "loss": 3.2689,
+      "step": 7378
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6284779906272888,
+      "learning_rate": 4.018131679447701e-06,
+      "loss": 3.2768,
+      "step": 7379
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6638051867485046,
+      "learning_rate": 3.999643200764985e-06,
+      "loss": 3.3569,
+      "step": 7380
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6347754001617432,
+      "learning_rate": 3.981197012395521e-06,
+      "loss": 3.2362,
+      "step": 7381
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7009361982345581,
+      "learning_rate": 3.962793117510383e-06,
+      "loss": 3.2804,
+      "step": 7382
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6505569219589233,
+      "learning_rate": 3.944431519273401e-06,
+      "loss": 3.3916,
+      "step": 7383
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6859057545661926,
+      "learning_rate": 3.926112220841188e-06,
+      "loss": 3.2671,
+      "step": 7384
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6320242285728455,
+      "learning_rate": 3.907835225363005e-06,
+      "loss": 3.1011,
+      "step": 7385
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6930665969848633,
+      "learning_rate": 3.889600535980892e-06,
+      "loss": 3.3667,
+      "step": 7386
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6389195919036865,
+      "learning_rate": 3.8714081558295925e-06,
+      "loss": 3.2719,
+      "step": 7387
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6377859711647034,
+      "learning_rate": 3.853258088036604e-06,
+      "loss": 3.2008,
+      "step": 7388
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6412763595581055,
+      "learning_rate": 3.835150335722154e-06,
+      "loss": 3.2153,
+      "step": 7389
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7105813026428223,
+      "learning_rate": 3.817084901999113e-06,
+      "loss": 3.299,
+      "step": 7390
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6779466867446899,
+      "learning_rate": 3.7990617899731904e-06,
+      "loss": 3.2306,
+      "step": 7391
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6576768159866333,
+      "learning_rate": 3.7810810027427424e-06,
+      "loss": 3.4647,
+      "step": 7392
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6411910653114319,
+      "learning_rate": 3.7631425433989062e-06,
+      "loss": 3.2822,
+      "step": 7393
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6772045493125916,
+      "learning_rate": 3.745246415025466e-06,
+      "loss": 3.4465,
+      "step": 7394
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6431482434272766,
+      "learning_rate": 3.727392620699016e-06,
+      "loss": 3.2325,
+      "step": 7395
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6533840298652649,
+      "learning_rate": 3.7095811634887956e-06,
+      "loss": 3.1974,
+      "step": 7396
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6330516338348389,
+      "learning_rate": 3.691812046456827e-06,
+      "loss": 3.2032,
+      "step": 7397
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.617141604423523,
+      "learning_rate": 3.6740852726578067e-06,
+      "loss": 3.2139,
+      "step": 7398
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7009109258651733,
+      "learning_rate": 3.6564008451392127e-06,
+      "loss": 3.2776,
+      "step": 7399
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6232280731201172,
+      "learning_rate": 3.6387587669411416e-06,
+      "loss": 3.2027,
+      "step": 7400
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7022527456283569,
+      "learning_rate": 3.6211590410965e-06,
+      "loss": 3.3376,
+      "step": 7401
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6588423848152161,
+      "learning_rate": 3.603601670630896e-06,
+      "loss": 3.3283,
+      "step": 7402
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6571477055549622,
+      "learning_rate": 3.5860866585626373e-06,
+      "loss": 3.2783,
+      "step": 7403
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6520159244537354,
+      "learning_rate": 3.5686140079027598e-06,
+      "loss": 3.2508,
+      "step": 7404
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6331124901771545,
+      "learning_rate": 3.551183721655027e-06,
+      "loss": 3.2829,
+      "step": 7405
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6849514245986938,
+      "learning_rate": 3.5337958028158754e-06,
+      "loss": 3.188,
+      "step": 7406
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5790781378746033,
+      "learning_rate": 3.5164502543745257e-06,
+      "loss": 3.1526,
+      "step": 7407
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.634568452835083,
+      "learning_rate": 3.499147079312842e-06,
+      "loss": 3.3117,
+      "step": 7408
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6393299102783203,
+      "learning_rate": 3.481886280605445e-06,
+      "loss": 3.3075,
+      "step": 7409
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6796958446502686,
+      "learning_rate": 3.4646678612196837e-06,
+      "loss": 3.2816,
+      "step": 7410
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6363698840141296,
+      "learning_rate": 3.447491824115606e-06,
+      "loss": 3.3547,
+      "step": 7411
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.611214280128479,
+      "learning_rate": 3.43035817224599e-06,
+      "loss": 3.2755,
+      "step": 7412
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6416151523590088,
+      "learning_rate": 3.413266908556256e-06,
+      "loss": 3.3442,
+      "step": 7413
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6357517838478088,
+      "learning_rate": 3.3962180359846386e-06,
+      "loss": 3.308,
+      "step": 7414
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6159369945526123,
+      "learning_rate": 3.379211557462014e-06,
+      "loss": 3.2884,
+      "step": 7415
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6278047561645508,
+      "learning_rate": 3.362247475911989e-06,
+      "loss": 3.1813,
+      "step": 7416
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6572278141975403,
+      "learning_rate": 3.3453257942508972e-06,
+      "loss": 3.3234,
+      "step": 7417
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6664751172065735,
+      "learning_rate": 3.328446515387773e-06,
+      "loss": 3.3132,
+      "step": 7418
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6599653363227844,
+      "learning_rate": 3.3116096422243225e-06,
+      "loss": 3.2526,
+      "step": 7419
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.654568612575531,
+      "learning_rate": 3.294815177655064e-06,
+      "loss": 3.207,
+      "step": 7420
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6287360191345215,
+      "learning_rate": 3.2780631245671043e-06,
+      "loss": 3.209,
+      "step": 7421
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6323820948600769,
+      "learning_rate": 3.2613534858403625e-06,
+      "loss": 3.2948,
+      "step": 7422
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6450842022895813,
+      "learning_rate": 3.244686264347374e-06,
+      "loss": 3.3435,
+      "step": 7423
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7002039551734924,
+      "learning_rate": 3.2280614629534853e-06,
+      "loss": 3.4343,
+      "step": 7424
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6468720436096191,
+      "learning_rate": 3.211479084516633e-06,
+      "loss": 3.3368,
+      "step": 7425
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7393363118171692,
+      "learning_rate": 3.194939131887564e-06,
+      "loss": 3.2048,
+      "step": 7426
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6139890551567078,
+      "learning_rate": 3.1784416079096433e-06,
+      "loss": 3.3042,
+      "step": 7427
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6313186883926392,
+      "learning_rate": 3.1619865154190186e-06,
+      "loss": 3.2822,
+      "step": 7428
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6245015263557434,
+      "learning_rate": 3.1455738572444824e-06,
+      "loss": 3.3669,
+      "step": 7429
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6937124729156494,
+      "learning_rate": 3.1292036362076117e-06,
+      "loss": 3.2656,
+      "step": 7430
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6102727055549622,
+      "learning_rate": 3.1128758551225988e-06,
+      "loss": 3.1886,
+      "step": 7431
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6378781795501709,
+      "learning_rate": 3.0965905167963935e-06,
+      "loss": 3.2937,
+      "step": 7432
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6180261969566345,
+      "learning_rate": 3.0803476240286178e-06,
+      "loss": 3.1587,
+      "step": 7433
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.66991126537323,
+      "learning_rate": 3.064147179611648e-06,
+      "loss": 3.2151,
+      "step": 7434
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6227942705154419,
+      "learning_rate": 3.04798918633048e-06,
+      "loss": 3.2695,
+      "step": 7435
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.670200765132904,
+      "learning_rate": 3.0318736469628906e-06,
+      "loss": 3.1475,
+      "step": 7436
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5999524593353271,
+      "learning_rate": 3.015800564279303e-06,
+      "loss": 3.2645,
+      "step": 7437
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6560450196266174,
+      "learning_rate": 2.9997699410428956e-06,
+      "loss": 3.2083,
+      "step": 7438
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6428449153900146,
+      "learning_rate": 2.9837817800095203e-06,
+      "loss": 3.4943,
+      "step": 7439
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6920236945152283,
+      "learning_rate": 2.9678360839277276e-06,
+      "loss": 3.2001,
+      "step": 7440
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.606275200843811,
+      "learning_rate": 2.9519328555387417e-06,
+      "loss": 3.3291,
+      "step": 7441
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.665324330329895,
+      "learning_rate": 2.9360720975765144e-06,
+      "loss": 3.2925,
+      "step": 7442
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6416666507720947,
+      "learning_rate": 2.920253812767698e-06,
+      "loss": 3.4125,
+      "step": 7443
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.621583104133606,
+      "learning_rate": 2.9044780038316456e-06,
+      "loss": 3.2132,
+      "step": 7444
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6322230696678162,
+      "learning_rate": 2.888744673480437e-06,
+      "loss": 3.2776,
+      "step": 7445
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6622276306152344,
+      "learning_rate": 2.873053824418742e-06,
+      "loss": 3.3062,
+      "step": 7446
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5983544588088989,
+      "learning_rate": 2.8574054593440423e-06,
+      "loss": 3.2411,
+      "step": 7447
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6524794697761536,
+      "learning_rate": 2.841799580946464e-06,
+      "loss": 3.2322,
+      "step": 7448
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5979216694831848,
+      "learning_rate": 2.82623619190886e-06,
+      "loss": 3.1827,
+      "step": 7449
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6573725938796997,
+      "learning_rate": 2.8107152949067295e-06,
+      "loss": 3.4223,
+      "step": 7450
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6488075256347656,
+      "learning_rate": 2.795236892608327e-06,
+      "loss": 3.1364,
+      "step": 7451
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6476569771766663,
+      "learning_rate": 2.779800987674497e-06,
+      "loss": 3.2683,
+      "step": 7452
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5997454524040222,
+      "learning_rate": 2.764407582758921e-06,
+      "loss": 3.245,
+      "step": 7453
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6412538290023804,
+      "learning_rate": 2.7490566805078996e-06,
+      "loss": 3.2369,
+      "step": 7454
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6506205201148987,
+      "learning_rate": 2.7337482835604056e-06,
+      "loss": 3.3683,
+      "step": 7455
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6723852157592773,
+      "learning_rate": 2.7184823945481105e-06,
+      "loss": 3.2889,
+      "step": 7456
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6422088742256165,
+      "learning_rate": 2.7032590160954716e-06,
+      "loss": 3.2476,
+      "step": 7457
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6394736170768738,
+      "learning_rate": 2.6880781508194784e-06,
+      "loss": 3.2505,
+      "step": 7458
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6404505372047424,
+      "learning_rate": 2.67293980132996e-06,
+      "loss": 3.2505,
+      "step": 7459
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6839563846588135,
+      "learning_rate": 2.6578439702293344e-06,
+      "loss": 3.2388,
+      "step": 7460
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6375935673713684,
+      "learning_rate": 2.642790660112776e-06,
+      "loss": 3.1646,
+      "step": 7461
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6834440231323242,
+      "learning_rate": 2.6277798735681315e-06,
+      "loss": 3.1828,
+      "step": 7462
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6728748083114624,
+      "learning_rate": 2.6128116131758918e-06,
+      "loss": 3.2923,
+      "step": 7463
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.619761049747467,
+      "learning_rate": 2.5978858815093045e-06,
+      "loss": 3.2659,
+      "step": 7464
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.683583676815033,
+      "learning_rate": 2.5830026811342335e-06,
+      "loss": 3.2574,
+      "step": 7465
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6158591508865356,
+      "learning_rate": 2.5681620146093266e-06,
+      "loss": 3.3406,
+      "step": 7466
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6639404296875,
+      "learning_rate": 2.5533638844858486e-06,
+      "loss": 3.1972,
+      "step": 7467
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6213057637214661,
+      "learning_rate": 2.538608293307765e-06,
+      "loss": 3.2496,
+      "step": 7468
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6411742568016052,
+      "learning_rate": 2.52389524361174e-06,
+      "loss": 3.357,
+      "step": 7469
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.664076566696167,
+      "learning_rate": 2.5092247379271126e-06,
+      "loss": 3.2863,
+      "step": 7470
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.586796224117279,
+      "learning_rate": 2.494596778775893e-06,
+      "loss": 3.2527,
+      "step": 7471
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.661478579044342,
+      "learning_rate": 2.4800113686728467e-06,
+      "loss": 3.345,
+      "step": 7472
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6605984568595886,
+      "learning_rate": 2.4654685101253014e-06,
+      "loss": 3.2981,
+      "step": 7473
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6903208494186401,
+      "learning_rate": 2.450968205633425e-06,
+      "loss": 3.3093,
+      "step": 7474
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7049779295921326,
+      "learning_rate": 2.436510457689917e-06,
+      "loss": 3.2576,
+      "step": 7475
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6223276853561401,
+      "learning_rate": 2.4220952687802887e-06,
+      "loss": 3.3374,
+      "step": 7476
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6552082896232605,
+      "learning_rate": 2.4077226413826426e-06,
+      "loss": 3.2579,
+      "step": 7477
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6227458119392395,
+      "learning_rate": 2.3933925779678346e-06,
+      "loss": 3.2856,
+      "step": 7478
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6535359025001526,
+      "learning_rate": 2.379105080999311e-06,
+      "loss": 3.2242,
+      "step": 7479
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6697153449058533,
+      "learning_rate": 2.3648601529333014e-06,
+      "loss": 3.2368,
+      "step": 7480
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7097249627113342,
+      "learning_rate": 2.35065779621868e-06,
+      "loss": 3.3781,
+      "step": 7481
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6668369770050049,
+      "learning_rate": 2.336498013296967e-06,
+      "loss": 3.272,
+      "step": 7482
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6335403323173523,
+      "learning_rate": 2.3223808066024086e-06,
+      "loss": 3.1968,
+      "step": 7483
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6395009756088257,
+      "learning_rate": 2.3083061785618977e-06,
+      "loss": 3.3057,
+      "step": 7484
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6572073101997375,
+      "learning_rate": 2.2942741315950534e-06,
+      "loss": 3.3194,
+      "step": 7485
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.64683997631073,
+      "learning_rate": 2.2802846681141685e-06,
+      "loss": 3.1422,
+      "step": 7486
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6475039720535278,
+      "learning_rate": 2.2663377905241257e-06,
+      "loss": 3.2925,
+      "step": 7487
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6652488708496094,
+      "learning_rate": 2.2524335012225894e-06,
+      "loss": 3.1857,
+      "step": 7488
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6927341222763062,
+      "learning_rate": 2.238571802599898e-06,
+      "loss": 3.3621,
+      "step": 7489
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6448782086372375,
+      "learning_rate": 2.2247526970390064e-06,
+      "loss": 3.2789,
+      "step": 7490
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6733150482177734,
+      "learning_rate": 2.2109761869155697e-06,
+      "loss": 3.2261,
+      "step": 7491
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6071258187294006,
+      "learning_rate": 2.1972422745979436e-06,
+      "loss": 3.2484,
+      "step": 7492
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6346178650856018,
+      "learning_rate": 2.1835509624471562e-06,
+      "loss": 3.4007,
+      "step": 7493
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6324289441108704,
+      "learning_rate": 2.1699022528168797e-06,
+      "loss": 3.1415,
+      "step": 7494
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6148743033409119,
+      "learning_rate": 2.156296148053488e-06,
+      "loss": 3.3537,
+      "step": 7495
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6335322856903076,
+      "learning_rate": 2.142732650496082e-06,
+      "loss": 3.1739,
+      "step": 7496
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7053571343421936,
+      "learning_rate": 2.1292117624763243e-06,
+      "loss": 3.274,
+      "step": 7497
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6378585696220398,
+      "learning_rate": 2.115733486318605e-06,
+      "loss": 3.1187,
+      "step": 7498
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6480412483215332,
+      "learning_rate": 2.102297824340044e-06,
+      "loss": 3.2563,
+      "step": 7499
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6488012075424194,
+      "learning_rate": 2.0889047788503755e-06,
+      "loss": 3.2886,
+      "step": 7500
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6246366500854492,
+      "learning_rate": 2.0755543521519815e-06,
+      "loss": 3.2876,
+      "step": 7501
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.702854573726654,
+      "learning_rate": 2.06224654653997e-06,
+      "loss": 3.3309,
+      "step": 7502
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.630124568939209,
+      "learning_rate": 2.0489813643021493e-06,
+      "loss": 3.2226,
+      "step": 7503
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6372682452201843,
+      "learning_rate": 2.035758807718918e-06,
+      "loss": 3.2873,
+      "step": 7504
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6328387260437012,
+      "learning_rate": 2.0225788790633746e-06,
+      "loss": 3.227,
+      "step": 7505
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6260595917701721,
+      "learning_rate": 2.0094415806013454e-06,
+      "loss": 3.4956,
+      "step": 7506
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6554728746414185,
+      "learning_rate": 1.996346914591274e-06,
+      "loss": 3.2805,
+      "step": 7507
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6232674717903137,
+      "learning_rate": 1.983294883284248e-06,
+      "loss": 3.2885,
+      "step": 7508
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.624148964881897,
+      "learning_rate": 1.970285488924084e-06,
+      "loss": 3.4587,
+      "step": 7509
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6816232800483704,
+      "learning_rate": 1.957318733747271e-06,
+      "loss": 3.3016,
+      "step": 7510
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7068300247192383,
+      "learning_rate": 1.944394619982942e-06,
+      "loss": 3.2474,
+      "step": 7511
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6904548406600952,
+      "learning_rate": 1.931513149852848e-06,
+      "loss": 3.1786,
+      "step": 7512
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6330033540725708,
+      "learning_rate": 1.9186743255714955e-06,
+      "loss": 3.2149,
+      "step": 7513
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6272470355033875,
+      "learning_rate": 1.905878149346063e-06,
+      "loss": 3.286,
+      "step": 7514
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6389195919036865,
+      "learning_rate": 1.893124623376319e-06,
+      "loss": 3.269,
+      "step": 7515
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6015783548355103,
+      "learning_rate": 1.8804137498547592e-06,
+      "loss": 3.2182,
+      "step": 7516
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6698102951049805,
+      "learning_rate": 1.8677455309664971e-06,
+      "loss": 3.2994,
+      "step": 7517
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.654212474822998,
+      "learning_rate": 1.8551199688894016e-06,
+      "loss": 3.2943,
+      "step": 7518
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6759641766548157,
+      "learning_rate": 1.842537065793931e-06,
+      "loss": 3.2493,
+      "step": 7519
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6238897442817688,
+      "learning_rate": 1.8299968238432163e-06,
+      "loss": 3.2418,
+      "step": 7520
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6189209222793579,
+      "learning_rate": 1.8174992451930605e-06,
+      "loss": 3.2654,
+      "step": 7521
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6399158835411072,
+      "learning_rate": 1.805044331991995e-06,
+      "loss": 3.3872,
+      "step": 7522
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6976350545883179,
+      "learning_rate": 1.7926320863811129e-06,
+      "loss": 3.2379,
+      "step": 7523
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6221190094947815,
+      "learning_rate": 1.7802625104942627e-06,
+      "loss": 3.1995,
+      "step": 7524
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6344885230064392,
+      "learning_rate": 1.7679356064578821e-06,
+      "loss": 3.2165,
+      "step": 7525
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6899000406265259,
+      "learning_rate": 1.7556513763911096e-06,
+      "loss": 3.259,
+      "step": 7526
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6491447687149048,
+      "learning_rate": 1.7434098224057838e-06,
+      "loss": 3.2962,
+      "step": 7527
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6144093871116638,
+      "learning_rate": 1.7312109466063597e-06,
+      "loss": 3.1326,
+      "step": 7528
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6447715163230896,
+      "learning_rate": 1.71905475108991e-06,
+      "loss": 3.3746,
+      "step": 7529
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.596442461013794,
+      "learning_rate": 1.7069412379462911e-06,
+      "loss": 3.1959,
+      "step": 7530
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6180348992347717,
+      "learning_rate": 1.6948704092579205e-06,
+      "loss": 3.2475,
+      "step": 7531
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6351748704910278,
+      "learning_rate": 1.6828422670999432e-06,
+      "loss": 3.2058,
+      "step": 7532
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.642833411693573,
+      "learning_rate": 1.6708568135401225e-06,
+      "loss": 3.1984,
+      "step": 7533
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.638550341129303,
+      "learning_rate": 1.6589140506388933e-06,
+      "loss": 3.2736,
+      "step": 7534
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6374490261077881,
+      "learning_rate": 1.6470139804493357e-06,
+      "loss": 3.3581,
+      "step": 7535
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6551685333251953,
+      "learning_rate": 1.6351566050172573e-06,
+      "loss": 3.2042,
+      "step": 7536
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6481794118881226,
+      "learning_rate": 1.6233419263810278e-06,
+      "loss": 3.2001,
+      "step": 7537
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6699073314666748,
+      "learning_rate": 1.611569946571745e-06,
+      "loss": 3.3134,
+      "step": 7538
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.680874764919281,
+      "learning_rate": 1.5998406676131783e-06,
+      "loss": 3.3234,
+      "step": 7539
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6119950413703918,
+      "learning_rate": 1.5881540915216875e-06,
+      "loss": 3.078,
+      "step": 7540
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6424537897109985,
+      "learning_rate": 1.5765102203063596e-06,
+      "loss": 3.254,
+      "step": 7541
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6495968103408813,
+      "learning_rate": 1.5649090559688716e-06,
+      "loss": 3.2139,
+      "step": 7542
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6209322214126587,
+      "learning_rate": 1.5533506005036557e-06,
+      "loss": 3.3226,
+      "step": 7543
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6560537219047546,
+      "learning_rate": 1.5418348558977058e-06,
+      "loss": 3.2831,
+      "step": 7544
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6782567501068115,
+      "learning_rate": 1.5303618241306883e-06,
+      "loss": 3.2048,
+      "step": 7545
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.648374617099762,
+      "learning_rate": 1.5189315071749977e-06,
+      "loss": 3.2015,
+      "step": 7546
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6472437381744385,
+      "learning_rate": 1.5075439069956453e-06,
+      "loss": 3.1896,
+      "step": 7547
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6487230062484741,
+      "learning_rate": 1.4961990255502323e-06,
+      "loss": 3.2939,
+      "step": 7548
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6969477534294128,
+      "learning_rate": 1.4848968647891148e-06,
+      "loss": 3.3404,
+      "step": 7549
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6641603708267212,
+      "learning_rate": 1.4736374266552943e-06,
+      "loss": 3.2283,
+      "step": 7550
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6257874965667725,
+      "learning_rate": 1.4624207130843336e-06,
+      "loss": 3.2594,
+      "step": 7551
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6400957703590393,
+      "learning_rate": 1.4512467260045514e-06,
+      "loss": 3.3074,
+      "step": 7552
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6433885097503662,
+      "learning_rate": 1.4401154673368833e-06,
+      "loss": 3.3894,
+      "step": 7553
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6520471572875977,
+      "learning_rate": 1.4290269389949095e-06,
+      "loss": 3.2277,
+      "step": 7554
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6392160058021545,
+      "learning_rate": 1.417981142884911e-06,
+      "loss": 3.2823,
+      "step": 7555
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.637761116027832,
+      "learning_rate": 1.4069780809057575e-06,
+      "loss": 3.2336,
+      "step": 7556
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6488739252090454,
+      "learning_rate": 1.3960177549489917e-06,
+      "loss": 3.38,
+      "step": 7557
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6583778858184814,
+      "learning_rate": 1.3851001668988562e-06,
+      "loss": 3.2938,
+      "step": 7558
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6335779428482056,
+      "learning_rate": 1.3742253186321829e-06,
+      "loss": 3.2275,
+      "step": 7559
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6512758135795593,
+      "learning_rate": 1.3633932120184766e-06,
+      "loss": 3.3365,
+      "step": 7560
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6726590991020203,
+      "learning_rate": 1.3526038489199421e-06,
+      "loss": 3.3447,
+      "step": 7561
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6761517524719238,
+      "learning_rate": 1.3418572311913735e-06,
+      "loss": 3.3358,
+      "step": 7562
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5900736451148987,
+      "learning_rate": 1.3311533606802651e-06,
+      "loss": 3.2874,
+      "step": 7563
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6390765905380249,
+      "learning_rate": 1.3204922392266728e-06,
+      "loss": 3.211,
+      "step": 7564
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6341561079025269,
+      "learning_rate": 1.309873868663436e-06,
+      "loss": 3.3539,
+      "step": 7565
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6070687174797058,
+      "learning_rate": 1.2992982508159e-06,
+      "loss": 3.2892,
+      "step": 7566
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6941978931427002,
+      "learning_rate": 1.2887653875021944e-06,
+      "loss": 3.3742,
+      "step": 7567
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6048576235771179,
+      "learning_rate": 1.2782752805330366e-06,
+      "loss": 3.1947,
+      "step": 7568
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6762664318084717,
+      "learning_rate": 1.2678279317117903e-06,
+      "loss": 3.2131,
+      "step": 7569
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6474283337593079,
+      "learning_rate": 1.2574233428344905e-06,
+      "loss": 3.3567,
+      "step": 7570
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6327533721923828,
+      "learning_rate": 1.2470615156897624e-06,
+      "loss": 3.2799,
+      "step": 7571
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6646479368209839,
+      "learning_rate": 1.2367424520589588e-06,
+      "loss": 3.3822,
+      "step": 7572
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6395760178565979,
+      "learning_rate": 1.2264661537160492e-06,
+      "loss": 3.3372,
+      "step": 7573
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7232096195220947,
+      "learning_rate": 1.216232622427621e-06,
+      "loss": 3.3805,
+      "step": 7574
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6458327174186707,
+      "learning_rate": 1.206041859952961e-06,
+      "loss": 3.2589,
+      "step": 7575
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6875505447387695,
+      "learning_rate": 1.1958938680439736e-06,
+      "loss": 3.1891,
+      "step": 7576
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6274546384811401,
+      "learning_rate": 1.1857886484452073e-06,
+      "loss": 3.2288,
+      "step": 7577
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6373465657234192,
+      "learning_rate": 1.1757262028938842e-06,
+      "loss": 3.312,
+      "step": 7578
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.624261736869812,
+      "learning_rate": 1.1657065331198425e-06,
+      "loss": 3.1332,
+      "step": 7579
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6104755997657776,
+      "learning_rate": 1.1557296408455932e-06,
+      "loss": 3.2463,
+      "step": 7580
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6304614543914795,
+      "learning_rate": 1.1457955277862641e-06,
+      "loss": 3.3576,
+      "step": 7581
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6486703157424927,
+      "learning_rate": 1.1359041956496286e-06,
+      "loss": 3.2021,
+      "step": 7582
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6310344934463501,
+      "learning_rate": 1.1260556461361592e-06,
+      "loss": 3.2776,
+      "step": 7583
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6773595213890076,
+      "learning_rate": 1.1162498809389188e-06,
+      "loss": 3.2133,
+      "step": 7584
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6367495059967041,
+      "learning_rate": 1.106486901743642e-06,
+      "loss": 3.2301,
+      "step": 7585
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6119555830955505,
+      "learning_rate": 1.0967667102286527e-06,
+      "loss": 3.2219,
+      "step": 7586
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6391817927360535,
+      "learning_rate": 1.0870893080650313e-06,
+      "loss": 3.1323,
+      "step": 7587
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6428825855255127,
+      "learning_rate": 1.0774546969163912e-06,
+      "loss": 3.2592,
+      "step": 7588
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6393899917602539,
+      "learning_rate": 1.0678628784390466e-06,
+      "loss": 3.3378,
+      "step": 7589
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6011338829994202,
+      "learning_rate": 1.0583138542819558e-06,
+      "loss": 3.3527,
+      "step": 7590
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6448006629943848,
+      "learning_rate": 1.0488076260866952e-06,
+      "loss": 3.2675,
+      "step": 7591
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6437724828720093,
+      "learning_rate": 1.0393441954874849e-06,
+      "loss": 3.3337,
+      "step": 7592
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6157968640327454,
+      "learning_rate": 1.0299235641111904e-06,
+      "loss": 3.2734,
+      "step": 7593
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6341884136199951,
+      "learning_rate": 1.0205457335773493e-06,
+      "loss": 3.2529,
+      "step": 7594
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.8305909037590027,
+      "learning_rate": 1.0112107054981167e-06,
+      "loss": 3.2854,
+      "step": 7595
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.61409991979599,
+      "learning_rate": 1.0019184814782923e-06,
+      "loss": 3.2995,
+      "step": 7596
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6470026969909668,
+      "learning_rate": 9.9266906311532e-07,
+      "loss": 3.3683,
+      "step": 7597
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6454616785049438,
+      "learning_rate": 9.834624519992897e-07,
+      "loss": 3.2503,
+      "step": 7598
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.614826500415802,
+      "learning_rate": 9.742986497128792e-07,
+      "loss": 3.265,
+      "step": 7599
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.647479772567749,
+      "learning_rate": 9.65177657831523e-07,
+      "loss": 3.1108,
+      "step": 7600
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.689254641532898,
+      "learning_rate": 9.560994779231613e-07,
+      "loss": 3.2805,
+      "step": 7601
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6509751081466675,
+      "learning_rate": 9.470641115484624e-07,
+      "loss": 3.2505,
+      "step": 7602
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6674864292144775,
+      "learning_rate": 9.380715602607115e-07,
+      "loss": 3.2593,
+      "step": 7603
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6391346454620361,
+      "learning_rate": 9.291218256058387e-07,
+      "loss": 3.4132,
+      "step": 7604
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6764358878135681,
+      "learning_rate": 9.20214909122391e-07,
+      "loss": 3.0903,
+      "step": 7605
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6738482117652893,
+      "learning_rate": 9.113508123415881e-07,
+      "loss": 3.2611,
+      "step": 7606
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6696187853813171,
+      "learning_rate": 9.025295367872665e-07,
+      "loss": 3.3648,
+      "step": 7607
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6360906362533569,
+      "learning_rate": 8.937510839759078e-07,
+      "loss": 3.2088,
+      "step": 7608
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6381098031997681,
+      "learning_rate": 8.850154554166101e-07,
+      "loss": 3.326,
+      "step": 7609
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7050420045852661,
+      "learning_rate": 8.763226526111723e-07,
+      "loss": 3.2243,
+      "step": 7610
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.664466381072998,
+      "learning_rate": 8.676726770539267e-07,
+      "loss": 3.3126,
+      "step": 7611
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6208410263061523,
+      "learning_rate": 8.590655302319616e-07,
+      "loss": 3.256,
+      "step": 7612
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6272687315940857,
+      "learning_rate": 8.505012136249268e-07,
+      "loss": 3.272,
+      "step": 7613
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6727380752563477,
+      "learning_rate": 8.41979728705089e-07,
+      "loss": 3.2288,
+      "step": 7614
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7296229600906372,
+      "learning_rate": 8.335010769374429e-07,
+      "loss": 3.3521,
+      "step": 7615
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6695315837860107,
+      "learning_rate": 8.25065259779545e-07,
+      "loss": 3.2912,
+      "step": 7616
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6159558892250061,
+      "learning_rate": 8.166722786816239e-07,
+      "loss": 3.2625,
+      "step": 7617
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.639655590057373,
+      "learning_rate": 8.083221350865256e-07,
+      "loss": 3.2781,
+      "step": 7618
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6372049450874329,
+      "learning_rate": 8.000148304297128e-07,
+      "loss": 3.2502,
+      "step": 7619
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6868239641189575,
+      "learning_rate": 7.917503661393211e-07,
+      "loss": 3.2418,
+      "step": 7620
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6578603982925415,
+      "learning_rate": 7.835287436361305e-07,
+      "loss": 3.3421,
+      "step": 7621
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6327383518218994,
+      "learning_rate": 7.753499643334827e-07,
+      "loss": 3.2392,
+      "step": 7622
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6470942497253418,
+      "learning_rate": 7.672140296374475e-07,
+      "loss": 3.1826,
+      "step": 7623
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6480328440666199,
+      "learning_rate": 7.591209409466837e-07,
+      "loss": 3.34,
+      "step": 7624
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6624470949172974,
+      "learning_rate": 7.510706996524675e-07,
+      "loss": 3.2978,
+      "step": 7625
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6373721361160278,
+      "learning_rate": 7.430633071387749e-07,
+      "loss": 3.3639,
+      "step": 7626
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6560044288635254,
+      "learning_rate": 7.350987647820884e-07,
+      "loss": 3.3718,
+      "step": 7627
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6382562518119812,
+      "learning_rate": 7.271770739516737e-07,
+      "loss": 3.254,
+      "step": 7628
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6614659428596497,
+      "learning_rate": 7.19298236009358e-07,
+      "loss": 3.2579,
+      "step": 7629
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6337676048278809,
+      "learning_rate": 7.114622523095305e-07,
+      "loss": 3.099,
+      "step": 7630
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6617096066474915,
+      "learning_rate": 7.036691241993909e-07,
+      "loss": 3.2443,
+      "step": 7631
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6335844397544861,
+      "learning_rate": 6.9591885301859e-07,
+      "loss": 3.3422,
+      "step": 7632
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6168643236160278,
+      "learning_rate": 6.882114400995343e-07,
+      "loss": 3.253,
+      "step": 7633
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6648966073989868,
+      "learning_rate": 6.805468867672193e-07,
+      "loss": 3.162,
+      "step": 7634
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.680281400680542,
+      "learning_rate": 6.729251943392301e-07,
+      "loss": 3.223,
+      "step": 7635
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6999646425247192,
+      "learning_rate": 6.653463641258517e-07,
+      "loss": 3.2631,
+      "step": 7636
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6539798974990845,
+      "learning_rate": 6.578103974299588e-07,
+      "loss": 3.28,
+      "step": 7637
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.676738440990448,
+      "learning_rate": 6.503172955470982e-07,
+      "loss": 3.3316,
+      "step": 7638
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6027320623397827,
+      "learning_rate": 6.428670597654062e-07,
+      "loss": 3.2476,
+      "step": 7639
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5785861611366272,
+      "learning_rate": 6.354596913656363e-07,
+      "loss": 3.1649,
+      "step": 7640
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6677677631378174,
+      "learning_rate": 6.280951916212418e-07,
+      "loss": 3.2489,
+      "step": 7641
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.611767590045929,
+      "learning_rate": 6.207735617982657e-07,
+      "loss": 3.3958,
+      "step": 7642
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7346791625022888,
+      "learning_rate": 6.134948031553678e-07,
+      "loss": 3.2172,
+      "step": 7643
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6609097719192505,
+      "learning_rate": 6.062589169438248e-07,
+      "loss": 3.3464,
+      "step": 7644
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6335502862930298,
+      "learning_rate": 5.990659044076141e-07,
+      "loss": 3.178,
+      "step": 7645
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6232840418815613,
+      "learning_rate": 5.919157667832464e-07,
+      "loss": 3.2284,
+      "step": 7646
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5935794115066528,
+      "learning_rate": 5.848085052999885e-07,
+      "loss": 3.2545,
+      "step": 7647
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6684829592704773,
+      "learning_rate": 5.777441211795853e-07,
+      "loss": 3.3676,
+      "step": 7648
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6543155908584595,
+      "learning_rate": 5.707226156365375e-07,
+      "loss": 3.219,
+      "step": 7649
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5975430607795715,
+      "learning_rate": 5.637439898779073e-07,
+      "loss": 3.2246,
+      "step": 7650
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6876344084739685,
+      "learning_rate": 5.56808245103374e-07,
+      "loss": 3.3489,
+      "step": 7651
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6838445067405701,
+      "learning_rate": 5.499153825053171e-07,
+      "loss": 3.3739,
+      "step": 7652
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6268404722213745,
+      "learning_rate": 5.430654032686777e-07,
+      "loss": 3.1471,
+      "step": 7653
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.679879903793335,
+      "learning_rate": 5.362583085710416e-07,
+      "loss": 3.3478,
+      "step": 7654
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6650727391242981,
+      "learning_rate": 5.294940995826392e-07,
+      "loss": 3.3497,
+      "step": 7655
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6430428624153137,
+      "learning_rate": 5.227727774663182e-07,
+      "loss": 3.1272,
+      "step": 7656
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7259106636047363,
+      "learning_rate": 5.160943433775434e-07,
+      "loss": 3.2782,
+      "step": 7657
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6815977692604065,
+      "learning_rate": 5.094587984643962e-07,
+      "loss": 3.2856,
+      "step": 7658
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6110831499099731,
+      "learning_rate": 5.028661438676308e-07,
+      "loss": 3.354,
+      "step": 7659
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6547110080718994,
+      "learning_rate": 4.963163807205906e-07,
+      "loss": 3.2981,
+      "step": 7660
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6048975586891174,
+      "learning_rate": 4.898095101492916e-07,
+      "loss": 3.2398,
+      "step": 7661
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6170310378074646,
+      "learning_rate": 4.833455332722836e-07,
+      "loss": 3.306,
+      "step": 7662
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6575043797492981,
+      "learning_rate": 4.769244512008164e-07,
+      "loss": 3.2691,
+      "step": 7663
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6412174701690674,
+      "learning_rate": 4.7054626503878484e-07,
+      "loss": 3.4241,
+      "step": 7664
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6265059113502502,
+      "learning_rate": 4.642109758826174e-07,
+      "loss": 3.2871,
+      "step": 7665
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6097034215927124,
+      "learning_rate": 4.579185848214429e-07,
+      "loss": 3.2071,
+      "step": 7666
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6445879340171814,
+      "learning_rate": 4.5166909293703487e-07,
+      "loss": 3.285,
+      "step": 7667
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6280282735824585,
+      "learning_rate": 4.454625013037006e-07,
+      "loss": 3.1142,
+      "step": 7668
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.644930362701416,
+      "learning_rate": 4.392988109884477e-07,
+      "loss": 3.415,
+      "step": 7669
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7137476205825806,
+      "learning_rate": 4.331780230509008e-07,
+      "loss": 3.3423,
+      "step": 7670
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6461427211761475,
+      "learning_rate": 4.271001385432738e-07,
+      "loss": 3.3923,
+      "step": 7671
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6740446090698242,
+      "learning_rate": 4.2106515851042524e-07,
+      "loss": 3.1089,
+      "step": 7672
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6676669120788574,
+      "learning_rate": 4.150730839898309e-07,
+      "loss": 3.315,
+      "step": 7673
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6839563846588135,
+      "learning_rate": 4.0912391601161115e-07,
+      "loss": 3.3375,
+      "step": 7674
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7062957286834717,
+      "learning_rate": 4.032176555985034e-07,
+      "loss": 3.2208,
+      "step": 7675
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6304193139076233,
+      "learning_rate": 3.9735430376586224e-07,
+      "loss": 3.2194,
+      "step": 7676
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6265396475791931,
+      "learning_rate": 3.9153386152165905e-07,
+      "loss": 3.1607,
+      "step": 7677
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6499339938163757,
+      "learning_rate": 3.8575632986648236e-07,
+      "loss": 3.1646,
+      "step": 7678
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6155187487602234,
+      "learning_rate": 3.800217097935932e-07,
+      "loss": 3.2666,
+      "step": 7679
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6204526424407959,
+      "learning_rate": 3.7433000228878635e-07,
+      "loss": 3.255,
+      "step": 7680
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5999842882156372,
+      "learning_rate": 3.6868120833055686e-07,
+      "loss": 3.1958,
+      "step": 7681
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6637298464775085,
+      "learning_rate": 3.630753288900446e-07,
+      "loss": 3.2996,
+      "step": 7682
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6675429344177246,
+      "learning_rate": 3.575123649308953e-07,
+      "loss": 3.387,
+      "step": 7683
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6250631809234619,
+      "learning_rate": 3.5199231740945525e-07,
+      "loss": 3.2661,
+      "step": 7684
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6430096626281738,
+      "learning_rate": 3.4651518727474294e-07,
+      "loss": 3.2979,
+      "step": 7685
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6707577109336853,
+      "learning_rate": 3.410809754682831e-07,
+      "loss": 3.3047,
+      "step": 7686
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6935476064682007,
+      "learning_rate": 3.356896829243006e-07,
+      "loss": 3.1893,
+      "step": 7687
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.661466658115387,
+      "learning_rate": 3.303413105696096e-07,
+      "loss": 3.2446,
+      "step": 7688
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6921482682228088,
+      "learning_rate": 3.250358593236691e-07,
+      "loss": 3.3028,
+      "step": 7689
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6858127117156982,
+      "learning_rate": 3.19773330098555e-07,
+      "loss": 3.2903,
+      "step": 7690
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.602769136428833,
+      "learning_rate": 3.1455372379893267e-07,
+      "loss": 3.0643,
+      "step": 7691
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6282801628112793,
+      "learning_rate": 3.0937704132213975e-07,
+      "loss": 3.2068,
+      "step": 7692
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6082367300987244,
+      "learning_rate": 3.0424328355810326e-07,
+      "loss": 3.3177,
+      "step": 7693
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6054982542991638,
+      "learning_rate": 2.9915245138933957e-07,
+      "loss": 3.3331,
+      "step": 7694
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6351670026779175,
+      "learning_rate": 2.9410454569106516e-07,
+      "loss": 3.3554,
+      "step": 7695
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6396512985229492,
+      "learning_rate": 2.8909956733105815e-07,
+      "loss": 3.2822,
+      "step": 7696
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6445683836936951,
+      "learning_rate": 2.841375171697413e-07,
+      "loss": 3.2697,
+      "step": 7697
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6352219581604004,
+      "learning_rate": 2.792183960601269e-07,
+      "loss": 3.1338,
+      "step": 7698
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6588396430015564,
+      "learning_rate": 2.743422048478994e-07,
+      "loss": 3.3444,
+      "step": 7699
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6465908288955688,
+      "learning_rate": 2.69508944371305e-07,
+      "loss": 3.2591,
+      "step": 7700
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6593931913375854,
+      "learning_rate": 2.6471861546123465e-07,
+      "loss": 3.1718,
+      "step": 7701
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6716919541358948,
+      "learning_rate": 2.59971218941224e-07,
+      "loss": 3.3324,
+      "step": 7702
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6339880228042603,
+      "learning_rate": 2.55266755627398e-07,
+      "loss": 3.2649,
+      "step": 7703
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.641206681728363,
+      "learning_rate": 2.506052263284986e-07,
+      "loss": 3.3873,
+      "step": 7704
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6437599658966064,
+      "learning_rate": 2.4598663184591254e-07,
+      "loss": 3.3047,
+      "step": 7705
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6551904082298279,
+      "learning_rate": 2.414109729736158e-07,
+      "loss": 3.512,
+      "step": 7706
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6614031791687012,
+      "learning_rate": 2.368782504982292e-07,
+      "loss": 3.3369,
+      "step": 7707
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6629523038864136,
+      "learning_rate": 2.3238846519896274e-07,
+      "loss": 3.2633,
+      "step": 7708
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6999644041061401,
+      "learning_rate": 2.2794161784769896e-07,
+      "loss": 3.2459,
+      "step": 7709
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6301155686378479,
+      "learning_rate": 2.235377092088542e-07,
+      "loss": 3.3455,
+      "step": 7710
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6277669668197632,
+      "learning_rate": 2.1917674003954502e-07,
+      "loss": 3.355,
+      "step": 7711
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6536439061164856,
+      "learning_rate": 2.1485871108944955e-07,
+      "loss": 3.2062,
+      "step": 7712
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.622828483581543,
+      "learning_rate": 2.1058362310091837e-07,
+      "loss": 3.2024,
+      "step": 7713
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6759048700332642,
+      "learning_rate": 2.0635147680886368e-07,
+      "loss": 3.2899,
+      "step": 7714
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6670531034469604,
+      "learning_rate": 2.0216227294084234e-07,
+      "loss": 3.3518,
+      "step": 7715
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6462798714637756,
+      "learning_rate": 1.9801601221702825e-07,
+      "loss": 3.2263,
+      "step": 7716
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6871739625930786,
+      "learning_rate": 1.939126953502124e-07,
+      "loss": 3.2723,
+      "step": 7717
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6767607927322388,
+      "learning_rate": 1.8985232304580268e-07,
+      "loss": 3.2441,
+      "step": 7718
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6675797700881958,
+      "learning_rate": 1.8583489600182413e-07,
+      "loss": 3.2221,
+      "step": 7719
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6533353328704834,
+      "learning_rate": 1.8186041490894646e-07,
+      "loss": 3.2238,
+      "step": 7720
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6005282402038574,
+      "learning_rate": 1.7792888045037315e-07,
+      "loss": 3.364,
+      "step": 7721
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6536868810653687,
+      "learning_rate": 1.7404029330203574e-07,
+      "loss": 3.2367,
+      "step": 7722
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6557430624961853,
+      "learning_rate": 1.7019465413239955e-07,
+      "loss": 3.3412,
+      "step": 7723
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6770214438438416,
+      "learning_rate": 1.6639196360257458e-07,
+      "loss": 3.2828,
+      "step": 7724
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6555509567260742,
+      "learning_rate": 1.6263222236628794e-07,
+      "loss": 3.2589,
+      "step": 7725
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6666334271430969,
+      "learning_rate": 1.589154310698837e-07,
+      "loss": 3.3113,
+      "step": 7726
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5893369317054749,
+      "learning_rate": 1.55241590352323e-07,
+      "loss": 3.2455,
+      "step": 7727
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7154968976974487,
+      "learning_rate": 1.5161070084518392e-07,
+      "loss": 3.325,
+      "step": 7728
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6864626407623291,
+      "learning_rate": 1.4802276317266162e-07,
+      "loss": 3.3101,
+      "step": 7729
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6598794460296631,
+      "learning_rate": 1.444777779515405e-07,
+      "loss": 3.2696,
+      "step": 7730
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6975576877593994,
+      "learning_rate": 1.4097574579127749e-07,
+      "loss": 3.2838,
+      "step": 7731
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6082096695899963,
+      "learning_rate": 1.37516667293891e-07,
+      "loss": 3.1761,
+      "step": 7732
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5750147104263306,
+      "learning_rate": 1.3410054305404428e-07,
+      "loss": 3.3239,
+      "step": 7733
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6296765208244324,
+      "learning_rate": 1.3072737365901755e-07,
+      "loss": 3.3048,
+      "step": 7734
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6614196300506592,
+      "learning_rate": 1.2739715968868028e-07,
+      "loss": 3.2068,
+      "step": 7735
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6254644393920898,
+      "learning_rate": 1.241099017155467e-07,
+      "loss": 3.2677,
+      "step": 7736
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6854985952377319,
+      "learning_rate": 1.2086560030474813e-07,
+      "loss": 3.3053,
+      "step": 7737
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6591229438781738,
+      "learning_rate": 1.1766425601397734e-07,
+      "loss": 3.2759,
+      "step": 7738
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6330887079238892,
+      "learning_rate": 1.1450586939362739e-07,
+      "loss": 3.3254,
+      "step": 7739
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.645741879940033,
+      "learning_rate": 1.1139044098662509e-07,
+      "loss": 3.2391,
+      "step": 7740
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.674430251121521,
+      "learning_rate": 1.0831797132854204e-07,
+      "loss": 3.3575,
+      "step": 7741
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6620994210243225,
+      "learning_rate": 1.0528846094762234e-07,
+      "loss": 3.3184,
+      "step": 7742
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6199248433113098,
+      "learning_rate": 1.0230191036464388e-07,
+      "loss": 3.2341,
+      "step": 7743
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6632201671600342,
+      "learning_rate": 9.93583200930015e-08,
+      "loss": 3.2446,
+      "step": 7744
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6452274918556213,
+      "learning_rate": 9.645769063879039e-08,
+      "loss": 3.3059,
+      "step": 7745
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6447122693061829,
+      "learning_rate": 9.360002250061172e-08,
+      "loss": 3.3207,
+      "step": 7746
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6201050281524658,
+      "learning_rate": 9.078531616976693e-08,
+      "loss": 3.3251,
+      "step": 7747
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.663913905620575,
+      "learning_rate": 8.801357213011896e-08,
+      "loss": 3.2393,
+      "step": 7748
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6402385234832764,
+      "learning_rate": 8.528479085817554e-08,
+      "loss": 3.1977,
+      "step": 7749
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.680420458316803,
+      "learning_rate": 8.259897282303363e-08,
+      "loss": 3.3444,
+      "step": 7750
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6284847855567932,
+      "learning_rate": 7.995611848640728e-08,
+      "loss": 3.2691,
+      "step": 7751
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6648733019828796,
+      "learning_rate": 7.735622830265521e-08,
+      "loss": 3.1771,
+      "step": 7752
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6077075600624084,
+      "learning_rate": 7.479930271869773e-08,
+      "loss": 3.2351,
+      "step": 7753
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6494194269180298,
+      "learning_rate": 7.228534217415539e-08,
+      "loss": 3.2726,
+      "step": 7754
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6784036755561829,
+      "learning_rate": 6.981434710115475e-08,
+      "loss": 3.2135,
+      "step": 7755
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6575074195861816,
+      "learning_rate": 6.738631792452266e-08,
+      "loss": 3.2102,
+      "step": 7756
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6784305572509766,
+      "learning_rate": 6.500125506161969e-08,
+      "loss": 3.3353,
+      "step": 7757
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6398900151252747,
+      "learning_rate": 6.265915892253448e-08,
+      "loss": 3.1519,
+      "step": 7758
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6406898498535156,
+      "learning_rate": 6.036002990983391e-08,
+      "loss": 3.236,
+      "step": 7759
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6392099857330322,
+      "learning_rate": 5.810386841878512e-08,
+      "loss": 3.2477,
+      "step": 7760
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6563682556152344,
+      "learning_rate": 5.5890674837272285e-08,
+      "loss": 3.3468,
+      "step": 7761
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6428896188735962,
+      "learning_rate": 5.3720449545768826e-08,
+      "loss": 3.3645,
+      "step": 7762
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6636173725128174,
+      "learning_rate": 5.159319291733744e-08,
+      "loss": 3.2402,
+      "step": 7763
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6179137229919434,
+      "learning_rate": 4.950890531765784e-08,
+      "loss": 3.2196,
+      "step": 7764
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6699796319007874,
+      "learning_rate": 4.746758710511001e-08,
+      "loss": 3.3711,
+      "step": 7765
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6111055016517639,
+      "learning_rate": 4.546923863055219e-08,
+      "loss": 3.271,
+      "step": 7766
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6459084153175354,
+      "learning_rate": 4.351386023757064e-08,
+      "loss": 3.3405,
+      "step": 7767
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6312695741653442,
+      "learning_rate": 4.1601452262313155e-08,
+      "loss": 3.3122,
+      "step": 7768
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5960135459899902,
+      "learning_rate": 3.973201503351675e-08,
+      "loss": 3.381,
+      "step": 7769
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6069164276123047,
+      "learning_rate": 3.790554887256326e-08,
+      "loss": 3.2818,
+      "step": 7770
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6660847663879395,
+      "learning_rate": 3.612205409347924e-08,
+      "loss": 3.2356,
+      "step": 7771
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7008808255195618,
+      "learning_rate": 3.438153100282504e-08,
+      "loss": 3.3235,
+      "step": 7772
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6775698065757751,
+      "learning_rate": 3.2683979899833514e-08,
+      "loss": 3.2186,
+      "step": 7773
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6351761817932129,
+      "learning_rate": 3.1029401076354546e-08,
+      "loss": 3.2799,
+      "step": 7774
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6113251447677612,
+      "learning_rate": 2.9417794816799515e-08,
+      "loss": 3.2026,
+      "step": 7775
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6366956233978271,
+      "learning_rate": 2.7849161398224575e-08,
+      "loss": 3.1522,
+      "step": 7776
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.66960209608078,
+      "learning_rate": 2.632350109033066e-08,
+      "loss": 3.2107,
+      "step": 7777
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6164228916168213,
+      "learning_rate": 2.484081415535244e-08,
+      "loss": 3.1453,
+      "step": 7778
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6470010280609131,
+      "learning_rate": 2.3401100848197132e-08,
+      "loss": 3.2593,
+      "step": 7779
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5902237892150879,
+      "learning_rate": 2.2004361416361195e-08,
+      "loss": 3.2198,
+      "step": 7780
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6584739089012146,
+      "learning_rate": 2.0650596099985874e-08,
+      "loss": 3.3246,
+      "step": 7781
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6267509460449219,
+      "learning_rate": 1.9339805131773912e-08,
+      "loss": 3.3515,
+      "step": 7782
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6580419540405273,
+      "learning_rate": 1.8071988737100585e-08,
+      "loss": 3.2328,
+      "step": 7783
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6391124129295349,
+      "learning_rate": 1.6847147133847163e-08,
+      "loss": 3.2001,
+      "step": 7784
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6607205271720886,
+      "learning_rate": 1.5665280532650705e-08,
+      "loss": 3.2384,
+      "step": 7785
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6454439163208008,
+      "learning_rate": 1.4526389136654273e-08,
+      "loss": 3.2399,
+      "step": 7786
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6675859093666077,
+      "learning_rate": 1.3430473141645694e-08,
+      "loss": 3.1598,
+      "step": 7787
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6172164678573608,
+      "learning_rate": 1.2377532736057572e-08,
+      "loss": 3.3129,
+      "step": 7788
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6432853937149048,
+      "learning_rate": 1.1367568100856262e-08,
+      "loss": 3.2288,
+      "step": 7789
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6783219575881958,
+      "learning_rate": 1.0400579409680643e-08,
+      "loss": 3.2685,
+      "step": 7790
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6008654832839966,
+      "learning_rate": 9.476566828786615e-09,
+      "loss": 3.2806,
+      "step": 7791
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6340771317481995,
+      "learning_rate": 8.595530516991579e-09,
+      "loss": 3.2155,
+      "step": 7792
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.636785626411438,
+      "learning_rate": 7.757470625785467e-09,
+      "loss": 3.2274,
+      "step": 7793
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6771584749221802,
+      "learning_rate": 6.962387299219719e-09,
+      "loss": 3.2748,
+      "step": 7794
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6428625583648682,
+      "learning_rate": 6.210280674018298e-09,
+      "loss": 3.2257,
+      "step": 7795
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6587328910827637,
+      "learning_rate": 5.501150879411165e-09,
+      "loss": 3.3533,
+      "step": 7796
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6639074683189392,
+      "learning_rate": 4.8349980373563195e-09,
+      "loss": 3.2496,
+      "step": 7797
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6632134318351746,
+      "learning_rate": 4.2118222623455105e-09,
+      "loss": 3.2102,
+      "step": 7798
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6787715554237366,
+      "learning_rate": 3.6316236615430154e-09,
+      "loss": 3.2833,
+      "step": 7799
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6268180012702942,
+      "learning_rate": 3.0944023346746175e-09,
+      "loss": 3.1667,
+      "step": 7800
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6189209222793579,
+      "learning_rate": 2.6001583740553615e-09,
+      "loss": 3.2903,
+      "step": 7801
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6033608913421631,
+      "learning_rate": 2.1488918647283307e-09,
+      "loss": 3.2652,
+      "step": 7802
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6378599405288696,
+      "learning_rate": 1.7406028842148481e-09,
+      "loss": 3.3589,
+      "step": 7803
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6148380637168884,
+      "learning_rate": 1.3752915027087642e-09,
+      "loss": 3.2446,
+      "step": 7804
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.720740020275116,
+      "learning_rate": 1.0529577830209468e-09,
+      "loss": 3.3218,
+      "step": 7805
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6575115919113159,
+      "learning_rate": 7.736017805792805e-10,
+      "loss": 3.2913,
+      "step": 7806
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6172797679901123,
+      "learning_rate": 5.37223543400911e-10,
+      "loss": 3.1487,
+      "step": 7807
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6265156865119934,
+      "learning_rate": 3.4382311209224526e-10,
+      "loss": 3.3373,
+      "step": 7808
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.676074743270874,
+      "learning_rate": 1.9340051995997422e-10,
+      "loss": 3.1338,
+      "step": 7809
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6446760296821594,
+      "learning_rate": 8.595579281678311e-11,
+      "loss": 3.2552,
+      "step": 7810
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6062079071998596,
+      "learning_rate": 2.1488949120129775e-11,
+      "loss": 3.1867,
+      "step": 7811
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6387574076652527,
+      "learning_rate": 0.0,
+      "loss": 3.367,
+      "step": 7812
+    },
+    {
+      "epoch": 1.0,
+      "step": 7812,
+      "total_flos": 9.9855816916258e+17,
+      "train_loss": 3.4287229562440533,
+      "train_runtime": 19633.4504,
+      "train_samples_per_second": 101.867,
+      "train_steps_per_second": 0.398
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 7812,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 24000,
+  "total_flos": 9.9855816916258e+17,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}