diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,13159 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 1875,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0016,
+      "grad_norm": 5.62226299207291,
+      "learning_rate": 0.0,
+      "loss": 1.8733,
+      "step": 1
+    },
+    {
+      "epoch": 0.0032,
+      "grad_norm": 5.746994099322661,
+      "learning_rate": 5.319148936170213e-08,
+      "loss": 1.8807,
+      "step": 2
+    },
+    {
+      "epoch": 0.0048,
+      "grad_norm": 5.401377984537998,
+      "learning_rate": 1.0638297872340426e-07,
+      "loss": 1.8942,
+      "step": 3
+    },
+    {
+      "epoch": 0.0064,
+      "grad_norm": 5.716062893888617,
+      "learning_rate": 1.5957446808510638e-07,
+      "loss": 1.9057,
+      "step": 4
+    },
+    {
+      "epoch": 0.008,
+      "grad_norm": 5.8346254356312155,
+      "learning_rate": 2.1276595744680852e-07,
+      "loss": 1.8542,
+      "step": 5
+    },
+    {
+      "epoch": 0.0096,
+      "grad_norm": 5.571054857366781,
+      "learning_rate": 2.6595744680851066e-07,
+      "loss": 1.9659,
+      "step": 6
+    },
+    {
+      "epoch": 0.0112,
+      "grad_norm": 5.207531427391658,
+      "learning_rate": 3.1914893617021275e-07,
+      "loss": 1.8469,
+      "step": 7
+    },
+    {
+      "epoch": 0.0128,
+      "grad_norm": 5.198538330592293,
+      "learning_rate": 3.723404255319149e-07,
+      "loss": 1.8721,
+      "step": 8
+    },
+    {
+      "epoch": 0.0144,
+      "grad_norm": 5.352146679154644,
+      "learning_rate": 4.2553191489361704e-07,
+      "loss": 1.9468,
+      "step": 9
+    },
+    {
+      "epoch": 0.016,
+      "grad_norm": 4.84542012091441,
+      "learning_rate": 4.787234042553192e-07,
+      "loss": 1.87,
+      "step": 10
+    },
+    {
+      "epoch": 0.0176,
+      "grad_norm": 4.507714803947995,
+      "learning_rate": 5.319148936170213e-07,
+      "loss": 1.8445,
+      "step": 11
+    },
+    {
+      "epoch": 0.0192,
+      "grad_norm": 4.25682123781333,
+      "learning_rate": 5.851063829787235e-07,
+      "loss": 1.8657,
+      "step": 12
+    },
+    {
+      "epoch": 0.0208,
+      "grad_norm": 4.2512762482562,
+      "learning_rate": 6.382978723404255e-07,
+      "loss": 1.8149,
+      "step": 13
+    },
+    {
+      "epoch": 0.0224,
+      "grad_norm": 3.0991414644056996,
+      "learning_rate": 6.914893617021278e-07,
+      "loss": 1.8229,
+      "step": 14
+    },
+    {
+      "epoch": 0.024,
+      "grad_norm": 3.094485558434669,
+      "learning_rate": 7.446808510638298e-07,
+      "loss": 1.8517,
+      "step": 15
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 2.894798261378148,
+      "learning_rate": 7.97872340425532e-07,
+      "loss": 1.8629,
+      "step": 16
+    },
+    {
+      "epoch": 0.0272,
+      "grad_norm": 2.6099171919639654,
+      "learning_rate": 8.510638297872341e-07,
+      "loss": 1.7829,
+      "step": 17
+    },
+    {
+      "epoch": 0.0288,
+      "grad_norm": 2.6694918386782667,
+      "learning_rate": 9.042553191489363e-07,
+      "loss": 1.8358,
+      "step": 18
+    },
+    {
+      "epoch": 0.0304,
+      "grad_norm": 3.8778092642079147,
+      "learning_rate": 9.574468085106384e-07,
+      "loss": 1.755,
+      "step": 19
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 2.9150215078892208,
+      "learning_rate": 1.0106382978723404e-06,
+      "loss": 1.7525,
+      "step": 20
+    },
+    {
+      "epoch": 0.0336,
+      "grad_norm": 3.2310864302875046,
+      "learning_rate": 1.0638297872340427e-06,
+      "loss": 1.7894,
+      "step": 21
+    },
+    {
+      "epoch": 0.0352,
+      "grad_norm": 3.0167825463628803,
+      "learning_rate": 1.1170212765957447e-06,
+      "loss": 1.7192,
+      "step": 22
+    },
+    {
+      "epoch": 0.0368,
+      "grad_norm": 2.5759199108690978,
+      "learning_rate": 1.170212765957447e-06,
+      "loss": 1.7481,
+      "step": 23
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 2.217729037173822,
+      "learning_rate": 1.223404255319149e-06,
+      "loss": 1.7528,
+      "step": 24
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8104091510434737,
+      "learning_rate": 1.276595744680851e-06,
+      "loss": 1.7437,
+      "step": 25
+    },
+    {
+      "epoch": 0.0416,
+      "grad_norm": 2.2270747460022537,
+      "learning_rate": 1.3297872340425533e-06,
+      "loss": 1.748,
+      "step": 26
+    },
+    {
+      "epoch": 0.0432,
+      "grad_norm": 2.2619494752002494,
+      "learning_rate": 1.3829787234042555e-06,
+      "loss": 1.69,
+      "step": 27
+    },
+    {
+      "epoch": 0.0448,
+      "grad_norm": 1.8529655827289973,
+      "learning_rate": 1.4361702127659578e-06,
+      "loss": 1.7036,
+      "step": 28
+    },
+    {
+      "epoch": 0.0464,
+      "grad_norm": 2.0219180144477957,
+      "learning_rate": 1.4893617021276596e-06,
+      "loss": 1.687,
+      "step": 29
+    },
+    {
+      "epoch": 0.048,
+      "grad_norm": 1.8077499955144491,
+      "learning_rate": 1.5425531914893618e-06,
+      "loss": 1.6709,
+      "step": 30
+    },
+    {
+      "epoch": 0.0496,
+      "grad_norm": 1.6923861830308737,
+      "learning_rate": 1.595744680851064e-06,
+      "loss": 1.7229,
+      "step": 31
+    },
+    {
+      "epoch": 0.0512,
+      "grad_norm": 1.3348578380210163,
+      "learning_rate": 1.648936170212766e-06,
+      "loss": 1.6978,
+      "step": 32
+    },
+    {
+      "epoch": 0.0528,
+      "grad_norm": 1.4293758734983528,
+      "learning_rate": 1.7021276595744682e-06,
+      "loss": 1.7038,
+      "step": 33
+    },
+    {
+      "epoch": 0.0544,
+      "grad_norm": 1.2527084708900378,
+      "learning_rate": 1.7553191489361704e-06,
+      "loss": 1.6235,
+      "step": 34
+    },
+    {
+      "epoch": 0.056,
+      "grad_norm": 1.0856416737106502,
+      "learning_rate": 1.8085106382978727e-06,
+      "loss": 1.6138,
+      "step": 35
+    },
+    {
+      "epoch": 0.0576,
+      "grad_norm": 1.216031054870184,
+      "learning_rate": 1.8617021276595745e-06,
+      "loss": 1.6944,
+      "step": 36
+    },
+    {
+      "epoch": 0.0592,
+      "grad_norm": 1.2934177985215147,
+      "learning_rate": 1.9148936170212767e-06,
+      "loss": 1.6465,
+      "step": 37
+    },
+    {
+      "epoch": 0.0608,
+      "grad_norm": 1.0529150612329066,
+      "learning_rate": 1.968085106382979e-06,
+      "loss": 1.5554,
+      "step": 38
+    },
+    {
+      "epoch": 0.0624,
+      "grad_norm": 1.2736584368658097,
+      "learning_rate": 2.021276595744681e-06,
+      "loss": 1.6868,
+      "step": 39
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 1.1546039150026979,
+      "learning_rate": 2.074468085106383e-06,
+      "loss": 1.6295,
+      "step": 40
+    },
+    {
+      "epoch": 0.0656,
+      "grad_norm": 1.041523047929129,
+      "learning_rate": 2.1276595744680853e-06,
+      "loss": 1.6934,
+      "step": 41
+    },
+    {
+      "epoch": 0.0672,
+      "grad_norm": 1.124607570234757,
+      "learning_rate": 2.1808510638297876e-06,
+      "loss": 1.6203,
+      "step": 42
+    },
+    {
+      "epoch": 0.0688,
+      "grad_norm": 1.028113746803128,
+      "learning_rate": 2.2340425531914894e-06,
+      "loss": 1.6165,
+      "step": 43
+    },
+    {
+      "epoch": 0.0704,
+      "grad_norm": 0.9444985774904561,
+      "learning_rate": 2.2872340425531916e-06,
+      "loss": 1.5539,
+      "step": 44
+    },
+    {
+      "epoch": 0.072,
+      "grad_norm": 1.098218215448554,
+      "learning_rate": 2.340425531914894e-06,
+      "loss": 1.5242,
+      "step": 45
+    },
+    {
+      "epoch": 0.0736,
+      "grad_norm": 1.017113084727177,
+      "learning_rate": 2.393617021276596e-06,
+      "loss": 1.6258,
+      "step": 46
+    },
+    {
+      "epoch": 0.0752,
+      "grad_norm": 0.9139246169810218,
+      "learning_rate": 2.446808510638298e-06,
+      "loss": 1.6024,
+      "step": 47
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 0.9740165382581546,
+      "learning_rate": 2.5e-06,
+      "loss": 1.5723,
+      "step": 48
+    },
+    {
+      "epoch": 0.0784,
+      "grad_norm": 0.9897585433312512,
+      "learning_rate": 2.553191489361702e-06,
+      "loss": 1.605,
+      "step": 49
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9377756033250716,
+      "learning_rate": 2.6063829787234047e-06,
+      "loss": 1.6074,
+      "step": 50
+    },
+    {
+      "epoch": 0.0816,
+      "grad_norm": 0.888361442088947,
+      "learning_rate": 2.6595744680851065e-06,
+      "loss": 1.551,
+      "step": 51
+    },
+    {
+      "epoch": 0.0832,
+      "grad_norm": 0.9010425723038035,
+      "learning_rate": 2.7127659574468084e-06,
+      "loss": 1.5885,
+      "step": 52
+    },
+    {
+      "epoch": 0.0848,
+      "grad_norm": 0.8782680735159686,
+      "learning_rate": 2.765957446808511e-06,
+      "loss": 1.627,
+      "step": 53
+    },
+    {
+      "epoch": 0.0864,
+      "grad_norm": 0.8205194269781229,
+      "learning_rate": 2.819148936170213e-06,
+      "loss": 1.6255,
+      "step": 54
+    },
+    {
+      "epoch": 0.088,
+      "grad_norm": 0.8203697326474236,
+      "learning_rate": 2.8723404255319155e-06,
+      "loss": 1.4998,
+      "step": 55
+    },
+    {
+      "epoch": 0.0896,
+      "grad_norm": 0.91633549445533,
+      "learning_rate": 2.9255319148936174e-06,
+      "loss": 1.5603,
+      "step": 56
+    },
+    {
+      "epoch": 0.0912,
+      "grad_norm": 0.8067047612808607,
+      "learning_rate": 2.978723404255319e-06,
+      "loss": 1.6116,
+      "step": 57
+    },
+    {
+      "epoch": 0.0928,
+      "grad_norm": 0.8551505034793769,
+      "learning_rate": 3.031914893617022e-06,
+      "loss": 1.5906,
+      "step": 58
+    },
+    {
+      "epoch": 0.0944,
+      "grad_norm": 0.8193957244504834,
+      "learning_rate": 3.0851063829787237e-06,
+      "loss": 1.5703,
+      "step": 59
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 0.81504827175592,
+      "learning_rate": 3.1382978723404255e-06,
+      "loss": 1.5237,
+      "step": 60
+    },
+    {
+      "epoch": 0.0976,
+      "grad_norm": 0.8013808408175694,
+      "learning_rate": 3.191489361702128e-06,
+      "loss": 1.479,
+      "step": 61
+    },
+    {
+      "epoch": 0.0992,
+      "grad_norm": 0.8583688010765025,
+      "learning_rate": 3.24468085106383e-06,
+      "loss": 1.545,
+      "step": 62
+    },
+    {
+      "epoch": 0.1008,
+      "grad_norm": 0.9422040438358964,
+      "learning_rate": 3.297872340425532e-06,
+      "loss": 1.5166,
+      "step": 63
+    },
+    {
+      "epoch": 0.1024,
+      "grad_norm": 0.8406334698414787,
+      "learning_rate": 3.3510638297872345e-06,
+      "loss": 1.6014,
+      "step": 64
+    },
+    {
+      "epoch": 0.104,
+      "grad_norm": 0.901094194243374,
+      "learning_rate": 3.4042553191489363e-06,
+      "loss": 1.5144,
+      "step": 65
+    },
+    {
+      "epoch": 0.1056,
+      "grad_norm": 0.8626716726835978,
+      "learning_rate": 3.457446808510639e-06,
+      "loss": 1.4725,
+      "step": 66
+    },
+    {
+      "epoch": 0.1072,
+      "grad_norm": 0.8752420983503114,
+      "learning_rate": 3.510638297872341e-06,
+      "loss": 1.5298,
+      "step": 67
+    },
+    {
+      "epoch": 0.1088,
+      "grad_norm": 0.8626445551053299,
+      "learning_rate": 3.5638297872340426e-06,
+      "loss": 1.5273,
+      "step": 68
+    },
+    {
+      "epoch": 0.1104,
+      "grad_norm": 0.8192875362726145,
+      "learning_rate": 3.6170212765957453e-06,
+      "loss": 1.5124,
+      "step": 69
+    },
+    {
+      "epoch": 0.112,
+      "grad_norm": 0.9009606657014252,
+      "learning_rate": 3.670212765957447e-06,
+      "loss": 1.5266,
+      "step": 70
+    },
+    {
+      "epoch": 0.1136,
+      "grad_norm": 0.8276903097298349,
+      "learning_rate": 3.723404255319149e-06,
+      "loss": 1.493,
+      "step": 71
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 0.9171447342972514,
+      "learning_rate": 3.7765957446808516e-06,
+      "loss": 1.4786,
+      "step": 72
+    },
+    {
+      "epoch": 0.1168,
+      "grad_norm": 0.8322023288307312,
+      "learning_rate": 3.8297872340425535e-06,
+      "loss": 1.5629,
+      "step": 73
+    },
+    {
+      "epoch": 0.1184,
+      "grad_norm": 0.944192968157738,
+      "learning_rate": 3.882978723404256e-06,
+      "loss": 1.5307,
+      "step": 74
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9580241987746934,
+      "learning_rate": 3.936170212765958e-06,
+      "loss": 1.5363,
+      "step": 75
+    },
+    {
+      "epoch": 0.1216,
+      "grad_norm": 0.8991111767051639,
+      "learning_rate": 3.98936170212766e-06,
+      "loss": 1.4944,
+      "step": 76
+    },
+    {
+      "epoch": 0.1232,
+      "grad_norm": 0.9889663553382402,
+      "learning_rate": 4.042553191489362e-06,
+      "loss": 1.5405,
+      "step": 77
+    },
+    {
+      "epoch": 0.1248,
+      "grad_norm": 0.8479913644893209,
+      "learning_rate": 4.095744680851064e-06,
+      "loss": 1.463,
+      "step": 78
+    },
+    {
+      "epoch": 0.1264,
+      "grad_norm": 0.9176082023606945,
+      "learning_rate": 4.148936170212766e-06,
+      "loss": 1.4824,
+      "step": 79
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 0.8909642982060679,
+      "learning_rate": 4.202127659574468e-06,
+      "loss": 1.4836,
+      "step": 80
+    },
+    {
+      "epoch": 0.1296,
+      "grad_norm": 0.8653628680533969,
+      "learning_rate": 4.255319148936171e-06,
+      "loss": 1.4718,
+      "step": 81
+    },
+    {
+      "epoch": 0.1312,
+      "grad_norm": 0.9395830752243142,
+      "learning_rate": 4.308510638297873e-06,
+      "loss": 1.4712,
+      "step": 82
+    },
+    {
+      "epoch": 0.1328,
+      "grad_norm": 0.8354475901357226,
+      "learning_rate": 4.361702127659575e-06,
+      "loss": 1.5099,
+      "step": 83
+    },
+    {
+      "epoch": 0.1344,
+      "grad_norm": 0.8829667006626285,
+      "learning_rate": 4.414893617021277e-06,
+      "loss": 1.5071,
+      "step": 84
+    },
+    {
+      "epoch": 0.136,
+      "grad_norm": 0.8418008245421322,
+      "learning_rate": 4.468085106382979e-06,
+      "loss": 1.4995,
+      "step": 85
+    },
+    {
+      "epoch": 0.1376,
+      "grad_norm": 0.929375614410642,
+      "learning_rate": 4.521276595744681e-06,
+      "loss": 1.5093,
+      "step": 86
+    },
+    {
+      "epoch": 0.1392,
+      "grad_norm": 0.9799697790983709,
+      "learning_rate": 4.574468085106383e-06,
+      "loss": 1.5582,
+      "step": 87
+    },
+    {
+      "epoch": 0.1408,
+      "grad_norm": 0.9376049799622733,
+      "learning_rate": 4.6276595744680855e-06,
+      "loss": 1.4829,
+      "step": 88
+    },
+    {
+      "epoch": 0.1424,
+      "grad_norm": 0.9715674585160516,
+      "learning_rate": 4.680851063829788e-06,
+      "loss": 1.4719,
+      "step": 89
+    },
+    {
+      "epoch": 0.144,
+      "grad_norm": 0.8783987698225258,
+      "learning_rate": 4.73404255319149e-06,
+      "loss": 1.5009,
+      "step": 90
+    },
+    {
+      "epoch": 0.1456,
+      "grad_norm": 0.9602386073117885,
+      "learning_rate": 4.787234042553192e-06,
+      "loss": 1.4912,
+      "step": 91
+    },
+    {
+      "epoch": 0.1472,
+      "grad_norm": 0.9387133124172151,
+      "learning_rate": 4.840425531914894e-06,
+      "loss": 1.5279,
+      "step": 92
+    },
+    {
+      "epoch": 0.1488,
+      "grad_norm": 0.9146748043740972,
+      "learning_rate": 4.893617021276596e-06,
+      "loss": 1.5179,
+      "step": 93
+    },
+    {
+      "epoch": 0.1504,
+      "grad_norm": 0.9982307186397839,
+      "learning_rate": 4.946808510638298e-06,
+      "loss": 1.4905,
+      "step": 94
+    },
+    {
+      "epoch": 0.152,
+      "grad_norm": 0.9337613489901015,
+      "learning_rate": 5e-06,
+      "loss": 1.5128,
+      "step": 95
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 0.9172748445523281,
+      "learning_rate": 5.053191489361703e-06,
+      "loss": 1.4938,
+      "step": 96
+    },
+    {
+      "epoch": 0.1552,
+      "grad_norm": 0.9274752534831828,
+      "learning_rate": 5.106382978723404e-06,
+      "loss": 1.4861,
+      "step": 97
+    },
+    {
+      "epoch": 0.1568,
+      "grad_norm": 0.9801282245424324,
+      "learning_rate": 5.159574468085107e-06,
+      "loss": 1.555,
+      "step": 98
+    },
+    {
+      "epoch": 0.1584,
+      "grad_norm": 1.0745582620969478,
+      "learning_rate": 5.212765957446809e-06,
+      "loss": 1.4869,
+      "step": 99
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9165855156089596,
+      "learning_rate": 5.265957446808511e-06,
+      "loss": 1.5017,
+      "step": 100
+    },
+    {
+      "epoch": 0.1616,
+      "grad_norm": 0.9647801798617938,
+      "learning_rate": 5.319148936170213e-06,
+      "loss": 1.5078,
+      "step": 101
+    },
+    {
+      "epoch": 0.1632,
+      "grad_norm": 0.9914374293552304,
+      "learning_rate": 5.372340425531915e-06,
+      "loss": 1.4292,
+      "step": 102
+    },
+    {
+      "epoch": 0.1648,
+      "grad_norm": 0.9617555827535809,
+      "learning_rate": 5.425531914893617e-06,
+      "loss": 1.5904,
+      "step": 103
+    },
+    {
+      "epoch": 0.1664,
+      "grad_norm": 1.0969039952352677,
+      "learning_rate": 5.47872340425532e-06,
+      "loss": 1.4839,
+      "step": 104
+    },
+    {
+      "epoch": 0.168,
+      "grad_norm": 1.0212040140567873,
+      "learning_rate": 5.531914893617022e-06,
+      "loss": 1.4989,
+      "step": 105
+    },
+    {
+      "epoch": 0.1696,
+      "grad_norm": 0.9798596488851384,
+      "learning_rate": 5.5851063829787235e-06,
+      "loss": 1.4484,
+      "step": 106
+    },
+    {
+      "epoch": 0.1712,
+      "grad_norm": 1.250809769676185,
+      "learning_rate": 5.638297872340426e-06,
+      "loss": 1.4566,
+      "step": 107
+    },
+    {
+      "epoch": 0.1728,
+      "grad_norm": 0.8965018194000115,
+      "learning_rate": 5.691489361702128e-06,
+      "loss": 1.4419,
+      "step": 108
+    },
+    {
+      "epoch": 0.1744,
+      "grad_norm": 1.3529774065671574,
+      "learning_rate": 5.744680851063831e-06,
+      "loss": 1.4853,
+      "step": 109
+    },
+    {
+      "epoch": 0.176,
+      "grad_norm": 1.0112519966485178,
+      "learning_rate": 5.7978723404255325e-06,
+      "loss": 1.4689,
+      "step": 110
+    },
+    {
+      "epoch": 0.1776,
+      "grad_norm": 1.3101785904851901,
+      "learning_rate": 5.851063829787235e-06,
+      "loss": 1.4557,
+      "step": 111
+    },
+    {
+      "epoch": 0.1792,
+      "grad_norm": 0.9973601848151146,
+      "learning_rate": 5.904255319148937e-06,
+      "loss": 1.4434,
+      "step": 112
+    },
+    {
+      "epoch": 0.1808,
+      "grad_norm": 1.24469233265297,
+      "learning_rate": 5.957446808510638e-06,
+      "loss": 1.4413,
+      "step": 113
+    },
+    {
+      "epoch": 0.1824,
+      "grad_norm": 1.04461117119082,
+      "learning_rate": 6.010638297872341e-06,
+      "loss": 1.449,
+      "step": 114
+    },
+    {
+      "epoch": 0.184,
+      "grad_norm": 1.178991500026275,
+      "learning_rate": 6.063829787234044e-06,
+      "loss": 1.4077,
+      "step": 115
+    },
+    {
+      "epoch": 0.1856,
+      "grad_norm": 1.7963536228796226,
+      "learning_rate": 6.117021276595745e-06,
+      "loss": 1.4772,
+      "step": 116
+    },
+    {
+      "epoch": 0.1872,
+      "grad_norm": 1.20367371677769,
+      "learning_rate": 6.170212765957447e-06,
+      "loss": 1.4249,
+      "step": 117
+    },
+    {
+      "epoch": 0.1888,
+      "grad_norm": 1.3257492405901747,
+      "learning_rate": 6.22340425531915e-06,
+      "loss": 1.4985,
+      "step": 118
+    },
+    {
+      "epoch": 0.1904,
+      "grad_norm": 1.1457033681970583,
+      "learning_rate": 6.276595744680851e-06,
+      "loss": 1.3747,
+      "step": 119
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 1.5856534710463634,
+      "learning_rate": 6.329787234042554e-06,
+      "loss": 1.4692,
+      "step": 120
+    },
+    {
+      "epoch": 0.1936,
+      "grad_norm": 1.183885873709775,
+      "learning_rate": 6.382978723404256e-06,
+      "loss": 1.4711,
+      "step": 121
+    },
+    {
+      "epoch": 0.1952,
+      "grad_norm": 1.6239310664454123,
+      "learning_rate": 6.436170212765958e-06,
+      "loss": 1.5127,
+      "step": 122
+    },
+    {
+      "epoch": 0.1968,
+      "grad_norm": 1.1138912453951073,
+      "learning_rate": 6.48936170212766e-06,
+      "loss": 1.4626,
+      "step": 123
+    },
+    {
+      "epoch": 0.1984,
+      "grad_norm": 1.5411328325184284,
+      "learning_rate": 6.542553191489362e-06,
+      "loss": 1.452,
+      "step": 124
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.1320386769924315,
+      "learning_rate": 6.595744680851064e-06,
+      "loss": 1.4587,
+      "step": 125
+    },
+    {
+      "epoch": 0.2016,
+      "grad_norm": 1.6143584376503772,
+      "learning_rate": 6.648936170212767e-06,
+      "loss": 1.4417,
+      "step": 126
+    },
+    {
+      "epoch": 0.2032,
+      "grad_norm": 1.3320857270304862,
+      "learning_rate": 6.702127659574469e-06,
+      "loss": 1.4724,
+      "step": 127
+    },
+    {
+      "epoch": 0.2048,
+      "grad_norm": 2.063909527247485,
+      "learning_rate": 6.75531914893617e-06,
+      "loss": 1.4487,
+      "step": 128
+    },
+    {
+      "epoch": 0.2064,
+      "grad_norm": 1.8724041182580602,
+      "learning_rate": 6.808510638297873e-06,
+      "loss": 1.4893,
+      "step": 129
+    },
+    {
+      "epoch": 0.208,
+      "grad_norm": 1.3869931885645097,
+      "learning_rate": 6.861702127659575e-06,
+      "loss": 1.4138,
+      "step": 130
+    },
+    {
+      "epoch": 0.2096,
+      "grad_norm": 1.475602605911623,
+      "learning_rate": 6.914893617021278e-06,
+      "loss": 1.4615,
+      "step": 131
+    },
+    {
+      "epoch": 0.2112,
+      "grad_norm": 1.2407192665538644,
+      "learning_rate": 6.968085106382979e-06,
+      "loss": 1.4192,
+      "step": 132
+    },
+    {
+      "epoch": 0.2128,
+      "grad_norm": 1.1825286219646471,
+      "learning_rate": 7.021276595744682e-06,
+      "loss": 1.4969,
+      "step": 133
+    },
+    {
+      "epoch": 0.2144,
+      "grad_norm": 1.2689242172202735,
+      "learning_rate": 7.074468085106384e-06,
+      "loss": 1.4761,
+      "step": 134
+    },
+    {
+      "epoch": 0.216,
+      "grad_norm": 1.0401960112997706,
+      "learning_rate": 7.127659574468085e-06,
+      "loss": 1.5053,
+      "step": 135
+    },
+    {
+      "epoch": 0.2176,
+      "grad_norm": 1.6449075435346958,
+      "learning_rate": 7.1808510638297875e-06,
+      "loss": 1.3993,
+      "step": 136
+    },
+    {
+      "epoch": 0.2192,
+      "grad_norm": 1.1939112115849293,
+      "learning_rate": 7.234042553191491e-06,
+      "loss": 1.4561,
+      "step": 137
+    },
+    {
+      "epoch": 0.2208,
+      "grad_norm": 1.4624227727389152,
+      "learning_rate": 7.287234042553192e-06,
+      "loss": 1.3991,
+      "step": 138
+    },
+    {
+      "epoch": 0.2224,
+      "grad_norm": 1.3246662525246184,
+      "learning_rate": 7.340425531914894e-06,
+      "loss": 1.4272,
+      "step": 139
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 1.3618272673689185,
+      "learning_rate": 7.3936170212765965e-06,
+      "loss": 1.4507,
+      "step": 140
+    },
+    {
+      "epoch": 0.2256,
+      "grad_norm": 1.167006686401738,
+      "learning_rate": 7.446808510638298e-06,
+      "loss": 1.4401,
+      "step": 141
+    },
+    {
+      "epoch": 0.2272,
+      "grad_norm": 1.444460747495841,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 1.4669,
+      "step": 142
+    },
+    {
+      "epoch": 0.2288,
+      "grad_norm": 1.2261285031056548,
+      "learning_rate": 7.553191489361703e-06,
+      "loss": 1.4123,
+      "step": 143
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 1.4475067097815897,
+      "learning_rate": 7.606382978723405e-06,
+      "loss": 1.4954,
+      "step": 144
+    },
+    {
+      "epoch": 0.232,
+      "grad_norm": 1.0828745077995214,
+      "learning_rate": 7.659574468085107e-06,
+      "loss": 1.4552,
+      "step": 145
+    },
+    {
+      "epoch": 0.2336,
+      "grad_norm": 1.502983008936816,
+      "learning_rate": 7.71276595744681e-06,
+      "loss": 1.4215,
+      "step": 146
+    },
+    {
+      "epoch": 0.2352,
+      "grad_norm": 1.3382100648101942,
+      "learning_rate": 7.765957446808511e-06,
+      "loss": 1.3715,
+      "step": 147
+    },
+    {
+      "epoch": 0.2368,
+      "grad_norm": 1.4785464803871713,
+      "learning_rate": 7.819148936170213e-06,
+      "loss": 1.4703,
+      "step": 148
+    },
+    {
+      "epoch": 0.2384,
+      "grad_norm": 1.2835665398984137,
+      "learning_rate": 7.872340425531916e-06,
+      "loss": 1.4455,
+      "step": 149
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.297870043729999,
+      "learning_rate": 7.925531914893617e-06,
+      "loss": 1.4935,
+      "step": 150
+    },
+    {
+      "epoch": 0.2416,
+      "grad_norm": 1.206693651474362,
+      "learning_rate": 7.97872340425532e-06,
+      "loss": 1.4043,
+      "step": 151
+    },
+    {
+      "epoch": 0.2432,
+      "grad_norm": 1.4117751668208294,
+      "learning_rate": 8.031914893617022e-06,
+      "loss": 1.453,
+      "step": 152
+    },
+    {
+      "epoch": 0.2448,
+      "grad_norm": 1.265548101816677,
+      "learning_rate": 8.085106382978723e-06,
+      "loss": 1.3904,
+      "step": 153
+    },
+    {
+      "epoch": 0.2464,
+      "grad_norm": 1.3865944561901362,
+      "learning_rate": 8.138297872340426e-06,
+      "loss": 1.4095,
+      "step": 154
+    },
+    {
+      "epoch": 0.248,
+      "grad_norm": 1.335359896063033,
+      "learning_rate": 8.191489361702128e-06,
+      "loss": 1.4466,
+      "step": 155
+    },
+    {
+      "epoch": 0.2496,
+      "grad_norm": 2.5651247238176746,
+      "learning_rate": 8.24468085106383e-06,
+      "loss": 1.4254,
+      "step": 156
+    },
+    {
+      "epoch": 0.2512,
+      "grad_norm": 1.8837443282352924,
+      "learning_rate": 8.297872340425532e-06,
+      "loss": 1.43,
+      "step": 157
+    },
+    {
+      "epoch": 0.2528,
+      "grad_norm": 1.1778299527602927,
+      "learning_rate": 8.351063829787235e-06,
+      "loss": 1.4527,
+      "step": 158
+    },
+    {
+      "epoch": 0.2544,
+      "grad_norm": 1.8402397363273713,
+      "learning_rate": 8.404255319148937e-06,
+      "loss": 1.4884,
+      "step": 159
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 1.7496428286114352,
+      "learning_rate": 8.457446808510638e-06,
+      "loss": 1.4535,
+      "step": 160
+    },
+    {
+      "epoch": 0.2576,
+      "grad_norm": 1.3027535838000748,
+      "learning_rate": 8.510638297872341e-06,
+      "loss": 1.5046,
+      "step": 161
+    },
+    {
+      "epoch": 0.2592,
+      "grad_norm": 1.4319570316900039,
+      "learning_rate": 8.563829787234044e-06,
+      "loss": 1.4992,
+      "step": 162
+    },
+    {
+      "epoch": 0.2608,
+      "grad_norm": 1.2499891064835174,
+      "learning_rate": 8.617021276595746e-06,
+      "loss": 1.4791,
+      "step": 163
+    },
+    {
+      "epoch": 0.2624,
+      "grad_norm": 1.2451658812747937,
+      "learning_rate": 8.670212765957447e-06,
+      "loss": 1.4937,
+      "step": 164
+    },
+    {
+      "epoch": 0.264,
+      "grad_norm": 1.262154781498061,
+      "learning_rate": 8.72340425531915e-06,
+      "loss": 1.4753,
+      "step": 165
+    },
+    {
+      "epoch": 0.2656,
+      "grad_norm": 1.2742566156930526,
+      "learning_rate": 8.776595744680852e-06,
+      "loss": 1.3944,
+      "step": 166
+    },
+    {
+      "epoch": 0.2672,
+      "grad_norm": 6.712772229056138,
+      "learning_rate": 8.829787234042555e-06,
+      "loss": 1.4525,
+      "step": 167
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 1.4052153262235774,
+      "learning_rate": 8.882978723404256e-06,
+      "loss": 1.4236,
+      "step": 168
+    },
+    {
+      "epoch": 0.2704,
+      "grad_norm": 1.0939282687297194,
+      "learning_rate": 8.936170212765958e-06,
+      "loss": 1.3998,
+      "step": 169
+    },
+    {
+      "epoch": 0.272,
+      "grad_norm": 1.2348863514466097,
+      "learning_rate": 8.98936170212766e-06,
+      "loss": 1.4361,
+      "step": 170
+    },
+    {
+      "epoch": 0.2736,
+      "grad_norm": 1.105220954815482,
+      "learning_rate": 9.042553191489362e-06,
+      "loss": 1.3938,
+      "step": 171
+    },
+    {
+      "epoch": 0.2752,
+      "grad_norm": 1.1628857331602198,
+      "learning_rate": 9.095744680851063e-06,
+      "loss": 1.4662,
+      "step": 172
+    },
+    {
+      "epoch": 0.2768,
+      "grad_norm": 1.1328843365557233,
+      "learning_rate": 9.148936170212767e-06,
+      "loss": 1.4611,
+      "step": 173
+    },
+    {
+      "epoch": 0.2784,
+      "grad_norm": 0.996845201983398,
+      "learning_rate": 9.20212765957447e-06,
+      "loss": 1.5037,
+      "step": 174
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1573232730095613,
+      "learning_rate": 9.255319148936171e-06,
+      "loss": 1.5602,
+      "step": 175
+    },
+    {
+      "epoch": 0.2816,
+      "grad_norm": 1.0826966959072166,
+      "learning_rate": 9.308510638297872e-06,
+      "loss": 1.4871,
+      "step": 176
+    },
+    {
+      "epoch": 0.2832,
+      "grad_norm": 1.111725169811083,
+      "learning_rate": 9.361702127659576e-06,
+      "loss": 1.4528,
+      "step": 177
+    },
+    {
+      "epoch": 0.2848,
+      "grad_norm": 1.196889850030109,
+      "learning_rate": 9.414893617021279e-06,
+      "loss": 1.4784,
+      "step": 178
+    },
+    {
+      "epoch": 0.2864,
+      "grad_norm": 0.9516395615222921,
+      "learning_rate": 9.46808510638298e-06,
+      "loss": 1.4072,
+      "step": 179
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 1.2239862023448242,
+      "learning_rate": 9.521276595744681e-06,
+      "loss": 1.4296,
+      "step": 180
+    },
+    {
+      "epoch": 0.2896,
+      "grad_norm": 1.0372029954992623,
+      "learning_rate": 9.574468085106385e-06,
+      "loss": 1.5251,
+      "step": 181
+    },
+    {
+      "epoch": 0.2912,
+      "grad_norm": 1.1658585976692784,
+      "learning_rate": 9.627659574468086e-06,
+      "loss": 1.4614,
+      "step": 182
+    },
+    {
+      "epoch": 0.2928,
+      "grad_norm": 1.0304634292489934,
+      "learning_rate": 9.680851063829787e-06,
+      "loss": 1.4147,
+      "step": 183
+    },
+    {
+      "epoch": 0.2944,
+      "grad_norm": 0.9942046926192526,
+      "learning_rate": 9.73404255319149e-06,
+      "loss": 1.3759,
+      "step": 184
+    },
+    {
+      "epoch": 0.296,
+      "grad_norm": 1.2079897841849294,
+      "learning_rate": 9.787234042553192e-06,
+      "loss": 1.4828,
+      "step": 185
+    },
+    {
+      "epoch": 0.2976,
+      "grad_norm": 1.1462202233764602,
+      "learning_rate": 9.840425531914895e-06,
+      "loss": 1.4472,
+      "step": 186
+    },
+    {
+      "epoch": 0.2992,
+      "grad_norm": 1.3854519633915403,
+      "learning_rate": 9.893617021276596e-06,
+      "loss": 1.5271,
+      "step": 187
+    },
+    {
+      "epoch": 0.3008,
+      "grad_norm": 1.2515539841569043,
+      "learning_rate": 9.946808510638298e-06,
+      "loss": 1.5307,
+      "step": 188
+    },
+    {
+      "epoch": 0.3024,
+      "grad_norm": 0.9715507772528107,
+      "learning_rate": 1e-05,
+      "loss": 1.4166,
+      "step": 189
+    },
+    {
+      "epoch": 0.304,
+      "grad_norm": 1.167833273215218,
+      "learning_rate": 9.99999133019233e-06,
+      "loss": 1.4122,
+      "step": 190
+    },
+    {
+      "epoch": 0.3056,
+      "grad_norm": 1.1389425903102333,
+      "learning_rate": 9.999965320799377e-06,
+      "loss": 1.4011,
+      "step": 191
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 0.9586589114852498,
+      "learning_rate": 9.999921971911345e-06,
+      "loss": 1.4236,
+      "step": 192
+    },
+    {
+      "epoch": 0.3088,
+      "grad_norm": 1.0067545033042806,
+      "learning_rate": 9.999861283678563e-06,
+      "loss": 1.4316,
+      "step": 193
+    },
+    {
+      "epoch": 0.3104,
+      "grad_norm": 1.2179943948791738,
+      "learning_rate": 9.999783256311494e-06,
+      "loss": 1.4598,
+      "step": 194
+    },
+    {
+      "epoch": 0.312,
+      "grad_norm": 1.1146117708963228,
+      "learning_rate": 9.99968789008073e-06,
+      "loss": 1.3802,
+      "step": 195
+    },
+    {
+      "epoch": 0.3136,
+      "grad_norm": 0.9071324281566625,
+      "learning_rate": 9.999575185316994e-06,
+      "loss": 1.448,
+      "step": 196
+    },
+    {
+      "epoch": 0.3152,
+      "grad_norm": 1.1544163315457692,
+      "learning_rate": 9.999445142411139e-06,
+      "loss": 1.4373,
+      "step": 197
+    },
+    {
+      "epoch": 0.3168,
+      "grad_norm": 0.8908523254707694,
+      "learning_rate": 9.99929776181414e-06,
+      "loss": 1.5117,
+      "step": 198
+    },
+    {
+      "epoch": 0.3184,
+      "grad_norm": 1.034899371455761,
+      "learning_rate": 9.999133044037107e-06,
+      "loss": 1.4313,
+      "step": 199
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1070751163669361,
+      "learning_rate": 9.998950989651261e-06,
+      "loss": 1.4658,
+      "step": 200
+    },
+    {
+      "epoch": 0.3216,
+      "grad_norm": 1.2156642025562623,
+      "learning_rate": 9.99875159928796e-06,
+      "loss": 1.3833,
+      "step": 201
+    },
+    {
+      "epoch": 0.3232,
+      "grad_norm": 1.2348049349170562,
+      "learning_rate": 9.99853487363867e-06,
+      "loss": 1.3945,
+      "step": 202
+    },
+    {
+      "epoch": 0.3248,
+      "grad_norm": 1.0826327117852885,
+      "learning_rate": 9.998300813454981e-06,
+      "loss": 1.5233,
+      "step": 203
+    },
+    {
+      "epoch": 0.3264,
+      "grad_norm": 1.1945292691325835,
+      "learning_rate": 9.998049419548597e-06,
+      "loss": 1.3781,
+      "step": 204
+    },
+    {
+      "epoch": 0.328,
+      "grad_norm": 1.1305105140898009,
+      "learning_rate": 9.997780692791329e-06,
+      "loss": 1.3962,
+      "step": 205
+    },
+    {
+      "epoch": 0.3296,
+      "grad_norm": 1.276031196266578,
+      "learning_rate": 9.997494634115101e-06,
+      "loss": 1.4053,
+      "step": 206
+    },
+    {
+      "epoch": 0.3312,
+      "grad_norm": 1.3501251619486347,
+      "learning_rate": 9.997191244511947e-06,
+      "loss": 1.4453,
+      "step": 207
+    },
+    {
+      "epoch": 0.3328,
+      "grad_norm": 1.0837280639245943,
+      "learning_rate": 9.996870525033994e-06,
+      "loss": 1.4219,
+      "step": 208
+    },
+    {
+      "epoch": 0.3344,
+      "grad_norm": 1.9512682751102763,
+      "learning_rate": 9.996532476793475e-06,
+      "loss": 1.4437,
+      "step": 209
+    },
+    {
+      "epoch": 0.336,
+      "grad_norm": 1.2103554988663232,
+      "learning_rate": 9.996177100962714e-06,
+      "loss": 1.4403,
+      "step": 210
+    },
+    {
+      "epoch": 0.3376,
+      "grad_norm": 1.440552463037835,
+      "learning_rate": 9.995804398774129e-06,
+      "loss": 1.4348,
+      "step": 211
+    },
+    {
+      "epoch": 0.3392,
+      "grad_norm": 1.291707104794343,
+      "learning_rate": 9.99541437152022e-06,
+      "loss": 1.4104,
+      "step": 212
+    },
+    {
+      "epoch": 0.3408,
+      "grad_norm": 1.5316596171467,
+      "learning_rate": 9.995007020553572e-06,
+      "loss": 1.429,
+      "step": 213
+    },
+    {
+      "epoch": 0.3424,
+      "grad_norm": 1.5800926294658175,
+      "learning_rate": 9.994582347286849e-06,
+      "loss": 1.5191,
+      "step": 214
+    },
+    {
+      "epoch": 0.344,
+      "grad_norm": 1.1760392407548164,
+      "learning_rate": 9.994140353192782e-06,
+      "loss": 1.4868,
+      "step": 215
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 1.0725653121971608,
+      "learning_rate": 9.993681039804176e-06,
+      "loss": 1.4456,
+      "step": 216
+    },
+    {
+      "epoch": 0.3472,
+      "grad_norm": 1.188218443216402,
+      "learning_rate": 9.99320440871389e-06,
+      "loss": 1.4378,
+      "step": 217
+    },
+    {
+      "epoch": 0.3488,
+      "grad_norm": 1.073633475587723,
+      "learning_rate": 9.99271046157485e-06,
+      "loss": 1.423,
+      "step": 218
+    },
+    {
+      "epoch": 0.3504,
+      "grad_norm": 1.0746171362988375,
+      "learning_rate": 9.99219920010002e-06,
+      "loss": 1.479,
+      "step": 219
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 1.0809945648377464,
+      "learning_rate": 9.991670626062422e-06,
+      "loss": 1.5114,
+      "step": 220
+    },
+    {
+      "epoch": 0.3536,
+      "grad_norm": 0.9638054305266779,
+      "learning_rate": 9.991124741295106e-06,
+      "loss": 1.4407,
+      "step": 221
+    },
+    {
+      "epoch": 0.3552,
+      "grad_norm": 1.02003743210777,
+      "learning_rate": 9.990561547691159e-06,
+      "loss": 1.4495,
+      "step": 222
+    },
+    {
+      "epoch": 0.3568,
+      "grad_norm": 1.0040443438802942,
+      "learning_rate": 9.989981047203693e-06,
+      "loss": 1.4419,
+      "step": 223
+    },
+    {
+      "epoch": 0.3584,
+      "grad_norm": 0.9327882260954362,
+      "learning_rate": 9.98938324184584e-06,
+      "loss": 1.3952,
+      "step": 224
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1614141809516667,
+      "learning_rate": 9.988768133690741e-06,
+      "loss": 1.4317,
+      "step": 225
+    },
+    {
+      "epoch": 0.3616,
+      "grad_norm": 1.368954414267439,
+      "learning_rate": 9.988135724871546e-06,
+      "loss": 1.4043,
+      "step": 226
+    },
+    {
+      "epoch": 0.3632,
+      "grad_norm": 0.9304802015773106,
+      "learning_rate": 9.987486017581401e-06,
+      "loss": 1.4326,
+      "step": 227
+    },
+    {
+      "epoch": 0.3648,
+      "grad_norm": 1.21113260171531,
+      "learning_rate": 9.986819014073436e-06,
+      "loss": 1.4791,
+      "step": 228
+    },
+    {
+      "epoch": 0.3664,
+      "grad_norm": 0.9472509409960091,
+      "learning_rate": 9.986134716660774e-06,
+      "loss": 1.5087,
+      "step": 229
+    },
+    {
+      "epoch": 0.368,
+      "grad_norm": 1.0509014522366695,
+      "learning_rate": 9.9854331277165e-06,
+      "loss": 1.4668,
+      "step": 230
+    },
+    {
+      "epoch": 0.3696,
+      "grad_norm": 0.9718800937304604,
+      "learning_rate": 9.984714249673676e-06,
+      "loss": 1.4619,
+      "step": 231
+    },
+    {
+      "epoch": 0.3712,
+      "grad_norm": 1.0686473845644762,
+      "learning_rate": 9.98397808502531e-06,
+      "loss": 1.4557,
+      "step": 232
+    },
+    {
+      "epoch": 0.3728,
+      "grad_norm": 0.9725472399146279,
+      "learning_rate": 9.983224636324369e-06,
+      "loss": 1.4443,
+      "step": 233
+    },
+    {
+      "epoch": 0.3744,
+      "grad_norm": 1.0944140475753608,
+      "learning_rate": 9.982453906183754e-06,
+      "loss": 1.4337,
+      "step": 234
+    },
+    {
+      "epoch": 0.376,
+      "grad_norm": 0.9431022694515178,
+      "learning_rate": 9.981665897276298e-06,
+      "loss": 1.4555,
+      "step": 235
+    },
+    {
+      "epoch": 0.3776,
+      "grad_norm": 1.0100013648537385,
+      "learning_rate": 9.980860612334753e-06,
+      "loss": 1.4637,
+      "step": 236
+    },
+    {
+      "epoch": 0.3792,
+      "grad_norm": 1.0058202516707364,
+      "learning_rate": 9.980038054151789e-06,
+      "loss": 1.4628,
+      "step": 237
+    },
+    {
+      "epoch": 0.3808,
+      "grad_norm": 0.9913153380720154,
+      "learning_rate": 9.979198225579968e-06,
+      "loss": 1.4358,
+      "step": 238
+    },
+    {
+      "epoch": 0.3824,
+      "grad_norm": 0.8988623338067386,
+      "learning_rate": 9.97834112953176e-06,
+      "loss": 1.4629,
+      "step": 239
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 0.9740147279656486,
+      "learning_rate": 9.9774667689795e-06,
+      "loss": 1.4357,
+      "step": 240
+    },
+    {
+      "epoch": 0.3856,
+      "grad_norm": 0.9951294105411214,
+      "learning_rate": 9.976575146955409e-06,
+      "loss": 1.4434,
+      "step": 241
+    },
+    {
+      "epoch": 0.3872,
+      "grad_norm": 0.9951718301507533,
+      "learning_rate": 9.97566626655156e-06,
+      "loss": 1.4113,
+      "step": 242
+    },
+    {
+      "epoch": 0.3888,
+      "grad_norm": 0.8802415880068204,
+      "learning_rate": 9.974740130919883e-06,
+      "loss": 1.3559,
+      "step": 243
+    },
+    {
+      "epoch": 0.3904,
+      "grad_norm": 1.0325126883354452,
+      "learning_rate": 9.973796743272141e-06,
+      "loss": 1.438,
+      "step": 244
+    },
+    {
+      "epoch": 0.392,
+      "grad_norm": 0.9666747686408942,
+      "learning_rate": 9.972836106879936e-06,
+      "loss": 1.4049,
+      "step": 245
+    },
+    {
+      "epoch": 0.3936,
+      "grad_norm": 0.9944733014612315,
+      "learning_rate": 9.971858225074672e-06,
+      "loss": 1.4266,
+      "step": 246
+    },
+    {
+      "epoch": 0.3952,
+      "grad_norm": 0.9587433451633552,
+      "learning_rate": 9.970863101247578e-06,
+      "loss": 1.3906,
+      "step": 247
+    },
+    {
+      "epoch": 0.3968,
+      "grad_norm": 1.016922781470055,
+      "learning_rate": 9.96985073884966e-06,
+      "loss": 1.5062,
+      "step": 248
+    },
+    {
+      "epoch": 0.3984,
+      "grad_norm": 0.9358973589751296,
+      "learning_rate": 9.968821141391716e-06,
+      "loss": 1.4976,
+      "step": 249
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0741883076450827,
+      "learning_rate": 9.96777431244431e-06,
+      "loss": 1.4495,
+      "step": 250
+    },
+    {
+      "epoch": 0.4016,
+      "grad_norm": 1.125348613337464,
+      "learning_rate": 9.966710255637764e-06,
+      "loss": 1.4026,
+      "step": 251
+    },
+    {
+      "epoch": 0.4032,
+      "grad_norm": 0.9686939955694517,
+      "learning_rate": 9.965628974662145e-06,
+      "loss": 1.4014,
+      "step": 252
+    },
+    {
+      "epoch": 0.4048,
+      "grad_norm": 1.08227919396946,
+      "learning_rate": 9.964530473267253e-06,
+      "loss": 1.4719,
+      "step": 253
+    },
+    {
+      "epoch": 0.4064,
+      "grad_norm": 0.9262434213987709,
+      "learning_rate": 9.963414755262606e-06,
+      "loss": 1.4696,
+      "step": 254
+    },
+    {
+      "epoch": 0.408,
+      "grad_norm": 0.9577795270680981,
+      "learning_rate": 9.962281824517427e-06,
+      "loss": 1.5227,
+      "step": 255
+    },
+    {
+      "epoch": 0.4096,
+      "grad_norm": 1.1494604740495509,
+      "learning_rate": 9.961131684960635e-06,
+      "loss": 1.496,
+      "step": 256
+    },
+    {
+      "epoch": 0.4112,
+      "grad_norm": 0.9050495086472392,
+      "learning_rate": 9.959964340580823e-06,
+      "loss": 1.4223,
+      "step": 257
+    },
+    {
+      "epoch": 0.4128,
+      "grad_norm": 1.31748177655284,
+      "learning_rate": 9.958779795426253e-06,
+      "loss": 1.4907,
+      "step": 258
+    },
+    {
+      "epoch": 0.4144,
+      "grad_norm": 0.9509393936828862,
+      "learning_rate": 9.957578053604837e-06,
+      "loss": 1.4263,
+      "step": 259
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 0.9514843035483065,
+      "learning_rate": 9.956359119284123e-06,
+      "loss": 1.4102,
+      "step": 260
+    },
+    {
+      "epoch": 0.4176,
+      "grad_norm": 0.8620538134245733,
+      "learning_rate": 9.955122996691278e-06,
+      "loss": 1.4225,
+      "step": 261
+    },
+    {
+      "epoch": 0.4192,
+      "grad_norm": 0.9362753717603248,
+      "learning_rate": 9.953869690113085e-06,
+      "loss": 1.5078,
+      "step": 262
+    },
+    {
+      "epoch": 0.4208,
+      "grad_norm": 1.0111889456194498,
+      "learning_rate": 9.952599203895912e-06,
+      "loss": 1.5283,
+      "step": 263
+    },
+    {
+      "epoch": 0.4224,
+      "grad_norm": 5.588721884545647,
+      "learning_rate": 9.95131154244571e-06,
+      "loss": 1.4652,
+      "step": 264
+    },
+    {
+      "epoch": 0.424,
+      "grad_norm": 1.3343427006397177,
+      "learning_rate": 9.950006710227986e-06,
+      "loss": 1.4454,
+      "step": 265
+    },
+    {
+      "epoch": 0.4256,
+      "grad_norm": 0.9461480376428831,
+      "learning_rate": 9.9486847117678e-06,
+      "loss": 1.46,
+      "step": 266
+    },
+    {
+      "epoch": 0.4272,
+      "grad_norm": 1.1160970651843485,
+      "learning_rate": 9.947345551649741e-06,
+      "loss": 1.4167,
+      "step": 267
+    },
+    {
+      "epoch": 0.4288,
+      "grad_norm": 0.9466702965216292,
+      "learning_rate": 9.945989234517913e-06,
+      "loss": 1.4291,
+      "step": 268
+    },
+    {
+      "epoch": 0.4304,
+      "grad_norm": 1.1318514717267618,
+      "learning_rate": 9.94461576507592e-06,
+      "loss": 1.4974,
+      "step": 269
+    },
+    {
+      "epoch": 0.432,
+      "grad_norm": 1.014737325917967,
+      "learning_rate": 9.943225148086846e-06,
+      "loss": 1.4468,
+      "step": 270
+    },
+    {
+      "epoch": 0.4336,
+      "grad_norm": 1.004892905335674,
+      "learning_rate": 9.941817388373248e-06,
+      "loss": 1.4054,
+      "step": 271
+    },
+    {
+      "epoch": 0.4352,
+      "grad_norm": 1.1145542428042499,
+      "learning_rate": 9.940392490817124e-06,
+      "loss": 1.4433,
+      "step": 272
+    },
+    {
+      "epoch": 0.4368,
+      "grad_norm": 0.9262280452051491,
+      "learning_rate": 9.938950460359912e-06,
+      "loss": 1.4592,
+      "step": 273
+    },
+    {
+      "epoch": 0.4384,
+      "grad_norm": 1.0872574594228337,
+      "learning_rate": 9.937491302002462e-06,
+      "loss": 1.397,
+      "step": 274
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0158478894607412,
+      "learning_rate": 9.936015020805022e-06,
+      "loss": 1.4159,
+      "step": 275
+    },
+    {
+      "epoch": 0.4416,
+      "grad_norm": 0.957580286646448,
+      "learning_rate": 9.934521621887223e-06,
+      "loss": 1.3818,
+      "step": 276
+    },
+    {
+      "epoch": 0.4432,
+      "grad_norm": 0.9141063972864393,
+      "learning_rate": 9.933011110428058e-06,
+      "loss": 1.4233,
+      "step": 277
+    },
+    {
+      "epoch": 0.4448,
+      "grad_norm": 0.9648086674817686,
+      "learning_rate": 9.93148349166586e-06,
+      "loss": 1.5066,
+      "step": 278
+    },
+    {
+      "epoch": 0.4464,
+      "grad_norm": 0.8897509056946189,
+      "learning_rate": 9.929938770898299e-06,
+      "loss": 1.4374,
+      "step": 279
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 0.9674740981855364,
+      "learning_rate": 9.928376953482343e-06,
+      "loss": 1.3939,
+      "step": 280
+    },
+    {
+      "epoch": 0.4496,
+      "grad_norm": 0.8344484419936378,
+      "learning_rate": 9.926798044834261e-06,
+      "loss": 1.4545,
+      "step": 281
+    },
+    {
+      "epoch": 0.4512,
+      "grad_norm": 1.0393346090471922,
+      "learning_rate": 9.92520205042958e-06,
+      "loss": 1.4894,
+      "step": 282
+    },
+    {
+      "epoch": 0.4528,
+      "grad_norm": 1.2717326998929173,
+      "learning_rate": 9.92358897580309e-06,
+      "loss": 1.3688,
+      "step": 283
+    },
+    {
+      "epoch": 0.4544,
+      "grad_norm": 0.8455840803687033,
+      "learning_rate": 9.921958826548808e-06,
+      "loss": 1.3876,
+      "step": 284
+    },
+    {
+      "epoch": 0.456,
+      "grad_norm": 1.2015086720989145,
+      "learning_rate": 9.920311608319968e-06,
+      "loss": 1.3888,
+      "step": 285
+    },
+    {
+      "epoch": 0.4576,
+      "grad_norm": 1.017465345182967,
+      "learning_rate": 9.918647326828993e-06,
+      "loss": 1.3333,
+      "step": 286
+    },
+    {
+      "epoch": 0.4592,
+      "grad_norm": 0.9351732766128413,
+      "learning_rate": 9.916965987847485e-06,
+      "loss": 1.3827,
+      "step": 287
+    },
+    {
+      "epoch": 0.4608,
+      "grad_norm": 1.0140632309246405,
+      "learning_rate": 9.915267597206198e-06,
+      "loss": 1.4095,
+      "step": 288
+    },
+    {
+      "epoch": 0.4624,
+      "grad_norm": 0.8347111036164164,
+      "learning_rate": 9.913552160795022e-06,
+      "loss": 1.3896,
+      "step": 289
+    },
+    {
+      "epoch": 0.464,
+      "grad_norm": 0.9092047051822805,
+      "learning_rate": 9.911819684562954e-06,
+      "loss": 1.4032,
+      "step": 290
+    },
+    {
+      "epoch": 0.4656,
+      "grad_norm": 1.0444567149642918,
+      "learning_rate": 9.910070174518093e-06,
+      "loss": 1.441,
+      "step": 291
+    },
+    {
+      "epoch": 0.4672,
+      "grad_norm": 0.861149684657332,
+      "learning_rate": 9.908303636727604e-06,
+      "loss": 1.4313,
+      "step": 292
+    },
+    {
+      "epoch": 0.4688,
+      "grad_norm": 0.8545363541845673,
+      "learning_rate": 9.9065200773177e-06,
+      "loss": 1.4126,
+      "step": 293
+    },
+    {
+      "epoch": 0.4704,
+      "grad_norm": 0.794720729089051,
+      "learning_rate": 9.904719502473635e-06,
+      "loss": 1.4051,
+      "step": 294
+    },
+    {
+      "epoch": 0.472,
+      "grad_norm": 0.772032657664973,
+      "learning_rate": 9.902901918439658e-06,
+      "loss": 1.3828,
+      "step": 295
+    },
+    {
+      "epoch": 0.4736,
+      "grad_norm": 0.8731064748154745,
+      "learning_rate": 9.901067331519013e-06,
+      "loss": 1.4737,
+      "step": 296
+    },
+    {
+      "epoch": 0.4752,
+      "grad_norm": 0.8682557725072058,
+      "learning_rate": 9.899215748073906e-06,
+      "loss": 1.4793,
+      "step": 297
+    },
+    {
+      "epoch": 0.4768,
+      "grad_norm": 0.9378808411525816,
+      "learning_rate": 9.897347174525487e-06,
+      "loss": 1.4984,
+      "step": 298
+    },
+    {
+      "epoch": 0.4784,
+      "grad_norm": 0.8988523412951636,
+      "learning_rate": 9.895461617353823e-06,
+      "loss": 1.435,
+      "step": 299
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8486005562873248,
+      "learning_rate": 9.893559083097885e-06,
+      "loss": 1.4625,
+      "step": 300
+    },
+    {
+      "epoch": 0.4816,
+      "grad_norm": 0.9003055077064087,
+      "learning_rate": 9.891639578355511e-06,
+      "loss": 1.3725,
+      "step": 301
+    },
+    {
+      "epoch": 0.4832,
+      "grad_norm": 0.8452670592813603,
+      "learning_rate": 9.8897031097834e-06,
+      "loss": 1.4051,
+      "step": 302
+    },
+    {
+      "epoch": 0.4848,
+      "grad_norm": 0.9032763695692412,
+      "learning_rate": 9.887749684097072e-06,
+      "loss": 1.3984,
+      "step": 303
+    },
+    {
+      "epoch": 0.4864,
+      "grad_norm": 0.9217880570166753,
+      "learning_rate": 9.88577930807086e-06,
+      "loss": 1.4255,
+      "step": 304
+    },
+    {
+      "epoch": 0.488,
+      "grad_norm": 0.7900676526757542,
+      "learning_rate": 9.883791988537874e-06,
+      "loss": 1.4211,
+      "step": 305
+    },
+    {
+      "epoch": 0.4896,
+      "grad_norm": 0.8569444599747695,
+      "learning_rate": 9.881787732389987e-06,
+      "loss": 1.4147,
+      "step": 306
+    },
+    {
+      "epoch": 0.4912,
+      "grad_norm": 0.8740378248774224,
+      "learning_rate": 9.879766546577805e-06,
+      "loss": 1.3612,
+      "step": 307
+    },
+    {
+      "epoch": 0.4928,
+      "grad_norm": 0.980903430839926,
+      "learning_rate": 9.877728438110645e-06,
+      "loss": 1.4075,
+      "step": 308
+    },
+    {
+      "epoch": 0.4944,
+      "grad_norm": 1.029244252794369,
+      "learning_rate": 9.87567341405651e-06,
+      "loss": 1.4236,
+      "step": 309
+    },
+    {
+      "epoch": 0.496,
+      "grad_norm": 0.9957378500968583,
+      "learning_rate": 9.873601481542065e-06,
+      "loss": 1.4852,
+      "step": 310
+    },
+    {
+      "epoch": 0.4976,
+      "grad_norm": 1.108990944925085,
+      "learning_rate": 9.871512647752612e-06,
+      "loss": 1.458,
+      "step": 311
+    },
+    {
+      "epoch": 0.4992,
+      "grad_norm": 0.9150661274668994,
+      "learning_rate": 9.86940691993207e-06,
+      "loss": 1.4226,
+      "step": 312
+    },
+    {
+      "epoch": 0.5008,
+      "grad_norm": 0.9593681804173295,
+      "learning_rate": 9.867284305382936e-06,
+      "loss": 1.4019,
+      "step": 313
+    },
+    {
+      "epoch": 0.5024,
+      "grad_norm": 0.8956176658563437,
+      "learning_rate": 9.865144811466275e-06,
+      "loss": 1.3989,
+      "step": 314
+    },
+    {
+      "epoch": 0.504,
+      "grad_norm": 1.026222428206968,
+      "learning_rate": 9.86298844560169e-06,
+      "loss": 1.4142,
+      "step": 315
+    },
+    {
+      "epoch": 0.5056,
+      "grad_norm": 0.9583316360660304,
+      "learning_rate": 9.860815215267288e-06,
+      "loss": 1.4104,
+      "step": 316
+    },
+    {
+      "epoch": 0.5072,
+      "grad_norm": 0.8667240559496524,
+      "learning_rate": 9.858625127999668e-06,
+      "loss": 1.4324,
+      "step": 317
+    },
+    {
+      "epoch": 0.5088,
+      "grad_norm": 0.8927831395239118,
+      "learning_rate": 9.856418191393881e-06,
+      "loss": 1.3604,
+      "step": 318
+    },
+    {
+      "epoch": 0.5104,
+      "grad_norm": 0.8303006133699132,
+      "learning_rate": 9.854194413103418e-06,
+      "loss": 1.5048,
+      "step": 319
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 0.9968602083561718,
+      "learning_rate": 9.851953800840166e-06,
+      "loss": 1.3992,
+      "step": 320
+    },
+    {
+      "epoch": 0.5136,
+      "grad_norm": 0.7804241501889839,
+      "learning_rate": 9.849696362374399e-06,
+      "loss": 1.3579,
+      "step": 321
+    },
+    {
+      "epoch": 0.5152,
+      "grad_norm": 1.0762037181795934,
+      "learning_rate": 9.847422105534739e-06,
+      "loss": 1.3892,
+      "step": 322
+    },
+    {
+      "epoch": 0.5168,
+      "grad_norm": 0.9024512096292459,
+      "learning_rate": 9.845131038208135e-06,
+      "loss": 1.4391,
+      "step": 323
+    },
+    {
+      "epoch": 0.5184,
+      "grad_norm": 0.950375240891984,
+      "learning_rate": 9.84282316833983e-06,
+      "loss": 1.389,
+      "step": 324
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.9217313418343899,
+      "learning_rate": 9.84049850393334e-06,
+      "loss": 1.3515,
+      "step": 325
+    },
+    {
+      "epoch": 0.5216,
+      "grad_norm": 0.8903073470541886,
+      "learning_rate": 9.838157053050423e-06,
+      "loss": 1.3974,
+      "step": 326
+    },
+    {
+      "epoch": 0.5232,
+      "grad_norm": 0.9404268131102451,
+      "learning_rate": 9.83579882381105e-06,
+      "loss": 1.446,
+      "step": 327
+    },
+    {
+      "epoch": 0.5248,
+      "grad_norm": 0.9459049756654574,
+      "learning_rate": 9.83342382439338e-06,
+      "loss": 1.4236,
+      "step": 328
+    },
+    {
+      "epoch": 0.5264,
+      "grad_norm": 1.0106062698333564,
+      "learning_rate": 9.831032063033726e-06,
+      "loss": 1.403,
+      "step": 329
+    },
+    {
+      "epoch": 0.528,
+      "grad_norm": 1.095802072613435,
+      "learning_rate": 9.828623548026533e-06,
+      "loss": 1.3984,
+      "step": 330
+    },
+    {
+      "epoch": 0.5296,
+      "grad_norm": 0.9794496689594635,
+      "learning_rate": 9.826198287724346e-06,
+      "loss": 1.4209,
+      "step": 331
+    },
+    {
+      "epoch": 0.5312,
+      "grad_norm": 1.0329217358111102,
+      "learning_rate": 9.823756290537783e-06,
+      "loss": 1.5127,
+      "step": 332
+    },
+    {
+      "epoch": 0.5328,
+      "grad_norm": 1.244494504884671,
+      "learning_rate": 9.821297564935499e-06,
+      "loss": 1.5077,
+      "step": 333
+    },
+    {
+      "epoch": 0.5344,
+      "grad_norm": 0.8650595110892331,
+      "learning_rate": 9.81882211944417e-06,
+      "loss": 1.401,
+      "step": 334
+    },
+    {
+      "epoch": 0.536,
+      "grad_norm": 1.3822246676846863,
+      "learning_rate": 9.816329962648444e-06,
+      "loss": 1.4377,
+      "step": 335
+    },
+    {
+      "epoch": 0.5376,
+      "grad_norm": 0.8812946489401868,
+      "learning_rate": 9.813821103190932e-06,
+      "loss": 1.3933,
+      "step": 336
+    },
+    {
+      "epoch": 0.5392,
+      "grad_norm": 1.1414874527523327,
+      "learning_rate": 9.811295549772169e-06,
+      "loss": 1.4089,
+      "step": 337
+    },
+    {
+      "epoch": 0.5408,
+      "grad_norm": 1.060913973993358,
+      "learning_rate": 9.808753311150575e-06,
+      "loss": 1.4874,
+      "step": 338
+    },
+    {
+      "epoch": 0.5424,
+      "grad_norm": 0.9957337198496105,
+      "learning_rate": 9.80619439614244e-06,
+      "loss": 1.3987,
+      "step": 339
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 1.0409320307872303,
+      "learning_rate": 9.803618813621885e-06,
+      "loss": 1.3526,
+      "step": 340
+    },
+    {
+      "epoch": 0.5456,
+      "grad_norm": 0.9734407627489675,
+      "learning_rate": 9.801026572520832e-06,
+      "loss": 1.4344,
+      "step": 341
+    },
+    {
+      "epoch": 0.5472,
+      "grad_norm": 0.8453680652659735,
+      "learning_rate": 9.798417681828972e-06,
+      "loss": 1.3923,
+      "step": 342
+    },
+    {
+      "epoch": 0.5488,
+      "grad_norm": 0.8786393793484176,
+      "learning_rate": 9.795792150593739e-06,
+      "loss": 1.4324,
+      "step": 343
+    },
+    {
+      "epoch": 0.5504,
+      "grad_norm": 0.8268750774072728,
+      "learning_rate": 9.793149987920273e-06,
+      "loss": 1.3761,
+      "step": 344
+    },
+    {
+      "epoch": 0.552,
+      "grad_norm": 0.9114027195922206,
+      "learning_rate": 9.79049120297139e-06,
+      "loss": 1.481,
+      "step": 345
+    },
+    {
+      "epoch": 0.5536,
+      "grad_norm": 0.7958391757750285,
+      "learning_rate": 9.787815804967554e-06,
+      "loss": 1.3837,
+      "step": 346
+    },
+    {
+      "epoch": 0.5552,
+      "grad_norm": 0.9299753980373837,
+      "learning_rate": 9.785123803186834e-06,
+      "loss": 1.407,
+      "step": 347
+    },
+    {
+      "epoch": 0.5568,
+      "grad_norm": 0.8489252489922379,
+      "learning_rate": 9.782415206964892e-06,
+      "loss": 1.4575,
+      "step": 348
+    },
+    {
+      "epoch": 0.5584,
+      "grad_norm": 0.8216508916025077,
+      "learning_rate": 9.779690025694926e-06,
+      "loss": 1.4167,
+      "step": 349
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7896132782582262,
+      "learning_rate": 9.776948268827658e-06,
+      "loss": 1.4131,
+      "step": 350
+    },
+    {
+      "epoch": 0.5616,
+      "grad_norm": 0.9281112339792017,
+      "learning_rate": 9.77418994587129e-06,
+      "loss": 1.4109,
+      "step": 351
+    },
+    {
+      "epoch": 0.5632,
+      "grad_norm": 0.7734560307524083,
+      "learning_rate": 9.771415066391473e-06,
+      "loss": 1.4144,
+      "step": 352
+    },
+    {
+      "epoch": 0.5648,
+      "grad_norm": 0.840688180659481,
+      "learning_rate": 9.768623640011272e-06,
+      "loss": 1.3986,
+      "step": 353
+    },
+    {
+      "epoch": 0.5664,
+      "grad_norm": 0.8566314664588407,
+      "learning_rate": 9.765815676411145e-06,
+      "loss": 1.4045,
+      "step": 354
+    },
+    {
+      "epoch": 0.568,
+      "grad_norm": 0.7800596217794905,
+      "learning_rate": 9.762991185328891e-06,
+      "loss": 1.462,
+      "step": 355
+    },
+    {
+      "epoch": 0.5696,
+      "grad_norm": 0.7928638656219332,
+      "learning_rate": 9.760150176559627e-06,
+      "loss": 1.3529,
+      "step": 356
+    },
+    {
+      "epoch": 0.5712,
+      "grad_norm": 0.8335381645209639,
+      "learning_rate": 9.757292659955755e-06,
+      "loss": 1.4744,
+      "step": 357
+    },
+    {
+      "epoch": 0.5728,
+      "grad_norm": 0.8332870438639828,
+      "learning_rate": 9.754418645426919e-06,
+      "loss": 1.3784,
+      "step": 358
+    },
+    {
+      "epoch": 0.5744,
+      "grad_norm": 0.8464098664199892,
+      "learning_rate": 9.751528142939986e-06,
+      "loss": 1.4763,
+      "step": 359
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 0.9106041994798749,
+      "learning_rate": 9.74862116251899e-06,
+      "loss": 1.4475,
+      "step": 360
+    },
+    {
+      "epoch": 0.5776,
+      "grad_norm": 0.8210258623150267,
+      "learning_rate": 9.74569771424512e-06,
+      "loss": 1.4287,
+      "step": 361
+    },
+    {
+      "epoch": 0.5792,
+      "grad_norm": 0.834120073367641,
+      "learning_rate": 9.742757808256667e-06,
+      "loss": 1.481,
+      "step": 362
+    },
+    {
+      "epoch": 0.5808,
+      "grad_norm": 0.7609703999551664,
+      "learning_rate": 9.739801454749e-06,
+      "loss": 1.4107,
+      "step": 363
+    },
+    {
+      "epoch": 0.5824,
+      "grad_norm": 0.8527009973131312,
+      "learning_rate": 9.736828663974527e-06,
+      "loss": 1.4673,
+      "step": 364
+    },
+    {
+      "epoch": 0.584,
+      "grad_norm": 0.8164119126594951,
+      "learning_rate": 9.733839446242655e-06,
+      "loss": 1.4568,
+      "step": 365
+    },
+    {
+      "epoch": 0.5856,
+      "grad_norm": 0.8530109563089022,
+      "learning_rate": 9.730833811919763e-06,
+      "loss": 1.4691,
+      "step": 366
+    },
+    {
+      "epoch": 0.5872,
+      "grad_norm": 0.8574974137029094,
+      "learning_rate": 9.727811771429158e-06,
+      "loss": 1.4202,
+      "step": 367
+    },
+    {
+      "epoch": 0.5888,
+      "grad_norm": 0.7757474313569022,
+      "learning_rate": 9.724773335251046e-06,
+      "loss": 1.4442,
+      "step": 368
+    },
+    {
+      "epoch": 0.5904,
+      "grad_norm": 0.8268622128771634,
+      "learning_rate": 9.721718513922488e-06,
+      "loss": 1.3717,
+      "step": 369
+    },
+    {
+      "epoch": 0.592,
+      "grad_norm": 0.8647588386447276,
+      "learning_rate": 9.71864731803737e-06,
+      "loss": 1.4694,
+      "step": 370
+    },
+    {
+      "epoch": 0.5936,
+      "grad_norm": 0.8191953738475698,
+      "learning_rate": 9.715559758246363e-06,
+      "loss": 1.4853,
+      "step": 371
+    },
+    {
+      "epoch": 0.5952,
+      "grad_norm": 0.861549421079886,
+      "learning_rate": 9.712455845256888e-06,
+      "loss": 1.4471,
+      "step": 372
+    },
+    {
+      "epoch": 0.5968,
+      "grad_norm": 0.918802411641663,
+      "learning_rate": 9.709335589833076e-06,
+      "loss": 1.4074,
+      "step": 373
+    },
+    {
+      "epoch": 0.5984,
+      "grad_norm": 0.942735416823841,
+      "learning_rate": 9.70619900279573e-06,
+      "loss": 1.4472,
+      "step": 374
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8711434350714989,
+      "learning_rate": 9.703046095022297e-06,
+      "loss": 1.4279,
+      "step": 375
+    },
+    {
+      "epoch": 0.6016,
+      "grad_norm": 1.0622725466024596,
+      "learning_rate": 9.699876877446815e-06,
+      "loss": 1.3826,
+      "step": 376
+    },
+    {
+      "epoch": 0.6032,
+      "grad_norm": 0.8418366821072758,
+      "learning_rate": 9.696691361059886e-06,
+      "loss": 1.416,
+      "step": 377
+    },
+    {
+      "epoch": 0.6048,
+      "grad_norm": 0.9910826628515811,
+      "learning_rate": 9.693489556908641e-06,
+      "loss": 1.3915,
+      "step": 378
+    },
+    {
+      "epoch": 0.6064,
+      "grad_norm": 1.0888710053129054,
+      "learning_rate": 9.690271476096686e-06,
+      "loss": 1.4433,
+      "step": 379
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 0.8466935648343848,
+      "learning_rate": 9.68703712978408e-06,
+      "loss": 1.398,
+      "step": 380
+    },
+    {
+      "epoch": 0.6096,
+      "grad_norm": 1.2307760645198296,
+      "learning_rate": 9.683786529187287e-06,
+      "loss": 1.457,
+      "step": 381
+    },
+    {
+      "epoch": 0.6112,
+      "grad_norm": 0.8666906197338995,
+      "learning_rate": 9.680519685579137e-06,
+      "loss": 1.4237,
+      "step": 382
+    },
+    {
+      "epoch": 0.6128,
+      "grad_norm": 1.0486659449704756,
+      "learning_rate": 9.677236610288797e-06,
+      "loss": 1.424,
+      "step": 383
+    },
+    {
+      "epoch": 0.6144,
+      "grad_norm": 1.0255123414407383,
+      "learning_rate": 9.673937314701714e-06,
+      "loss": 1.4487,
+      "step": 384
+    },
+    {
+      "epoch": 0.616,
+      "grad_norm": 0.8693116691689832,
+      "learning_rate": 9.670621810259596e-06,
+      "loss": 1.3721,
+      "step": 385
+    },
+    {
+      "epoch": 0.6176,
+      "grad_norm": 0.9153400028137791,
+      "learning_rate": 9.667290108460354e-06,
+      "loss": 1.4081,
+      "step": 386
+    },
+    {
+      "epoch": 0.6192,
+      "grad_norm": 0.916585652034894,
+      "learning_rate": 9.663942220858075e-06,
+      "loss": 1.4121,
+      "step": 387
+    },
+    {
+      "epoch": 0.6208,
+      "grad_norm": 0.8343830479165492,
+      "learning_rate": 9.660578159062977e-06,
+      "loss": 1.4126,
+      "step": 388
+    },
+    {
+      "epoch": 0.6224,
+      "grad_norm": 0.9571680157207823,
+      "learning_rate": 9.657197934741366e-06,
+      "loss": 1.4707,
+      "step": 389
+    },
+    {
+      "epoch": 0.624,
+      "grad_norm": 2.8719854455225433,
+      "learning_rate": 9.6538015596156e-06,
+      "loss": 1.3906,
+      "step": 390
+    },
+    {
+      "epoch": 0.6256,
+      "grad_norm": 1.1767372971747092,
+      "learning_rate": 9.650389045464046e-06,
+      "loss": 1.424,
+      "step": 391
+    },
+    {
+      "epoch": 0.6272,
+      "grad_norm": 0.8156334997101575,
+      "learning_rate": 9.646960404121042e-06,
+      "loss": 1.3929,
+      "step": 392
+    },
+    {
+      "epoch": 0.6288,
+      "grad_norm": 0.9423342294113503,
+      "learning_rate": 9.643515647476851e-06,
+      "loss": 1.4163,
+      "step": 393
+    },
+    {
+      "epoch": 0.6304,
+      "grad_norm": 0.8902190034125246,
+      "learning_rate": 9.640054787477626e-06,
+      "loss": 1.3622,
+      "step": 394
+    },
+    {
+      "epoch": 0.632,
+      "grad_norm": 0.9097314147125211,
+      "learning_rate": 9.63657783612536e-06,
+      "loss": 1.3513,
+      "step": 395
+    },
+    {
+      "epoch": 0.6336,
+      "grad_norm": 0.9483183125118382,
+      "learning_rate": 9.633084805477857e-06,
+      "loss": 1.4232,
+      "step": 396
+    },
+    {
+      "epoch": 0.6352,
+      "grad_norm": 0.8924129889108191,
+      "learning_rate": 9.629575707648675e-06,
+      "loss": 1.4627,
+      "step": 397
+    },
+    {
+      "epoch": 0.6368,
+      "grad_norm": 0.8609545147886143,
+      "learning_rate": 9.626050554807096e-06,
+      "loss": 1.4283,
+      "step": 398
+    },
+    {
+      "epoch": 0.6384,
+      "grad_norm": 0.9228557892274488,
+      "learning_rate": 9.62250935917808e-06,
+      "loss": 1.4203,
+      "step": 399
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8455061116567486,
+      "learning_rate": 9.618952133042223e-06,
+      "loss": 1.3984,
+      "step": 400
+    },
+    {
+      "epoch": 0.6416,
+      "grad_norm": 0.8962131053637206,
+      "learning_rate": 9.615378888735706e-06,
+      "loss": 1.4502,
+      "step": 401
+    },
+    {
+      "epoch": 0.6432,
+      "grad_norm": 0.9137751301128598,
+      "learning_rate": 9.611789638650269e-06,
+      "loss": 1.4307,
+      "step": 402
+    },
+    {
+      "epoch": 0.6448,
+      "grad_norm": 0.7808560850363229,
+      "learning_rate": 9.608184395233156e-06,
+      "loss": 1.4245,
+      "step": 403
+    },
+    {
+      "epoch": 0.6464,
+      "grad_norm": 0.8362321415841635,
+      "learning_rate": 9.604563170987072e-06,
+      "loss": 1.3611,
+      "step": 404
+    },
+    {
+      "epoch": 0.648,
+      "grad_norm": 0.9205598422546907,
+      "learning_rate": 9.600925978470143e-06,
+      "loss": 1.4304,
+      "step": 405
+    },
+    {
+      "epoch": 0.6496,
+      "grad_norm": 0.7759206091046329,
+      "learning_rate": 9.597272830295877e-06,
+      "loss": 1.3979,
+      "step": 406
+    },
+    {
+      "epoch": 0.6512,
+      "grad_norm": 0.947730026645376,
+      "learning_rate": 9.593603739133105e-06,
+      "loss": 1.4049,
+      "step": 407
+    },
+    {
+      "epoch": 0.6528,
+      "grad_norm": 1.002390593941524,
+      "learning_rate": 9.589918717705957e-06,
+      "loss": 1.347,
+      "step": 408
+    },
+    {
+      "epoch": 0.6544,
+      "grad_norm": 0.9170494063455187,
+      "learning_rate": 9.586217778793804e-06,
+      "loss": 1.4224,
+      "step": 409
+    },
+    {
+      "epoch": 0.656,
+      "grad_norm": 0.8478380983626956,
+      "learning_rate": 9.582500935231215e-06,
+      "loss": 1.3836,
+      "step": 410
+    },
+    {
+      "epoch": 0.6576,
+      "grad_norm": 1.0238360505298216,
+      "learning_rate": 9.57876819990792e-06,
+      "loss": 1.3797,
+      "step": 411
+    },
+    {
+      "epoch": 0.6592,
+      "grad_norm": 0.8781470308651935,
+      "learning_rate": 9.575019585768758e-06,
+      "loss": 1.5014,
+      "step": 412
+    },
+    {
+      "epoch": 0.6608,
+      "grad_norm": 0.9215480847724832,
+      "learning_rate": 9.571255105813632e-06,
+      "loss": 1.442,
+      "step": 413
+    },
+    {
+      "epoch": 0.6624,
+      "grad_norm": 0.9366334306450057,
+      "learning_rate": 9.567474773097469e-06,
+      "loss": 1.4312,
+      "step": 414
+    },
+    {
+      "epoch": 0.664,
+      "grad_norm": 0.8085361491556748,
+      "learning_rate": 9.563678600730175e-06,
+      "loss": 1.3952,
+      "step": 415
+    },
+    {
+      "epoch": 0.6656,
+      "grad_norm": 0.9797467918073007,
+      "learning_rate": 9.559866601876581e-06,
+      "loss": 1.4006,
+      "step": 416
+    },
+    {
+      "epoch": 0.6672,
+      "grad_norm": 0.7875247346611903,
+      "learning_rate": 9.556038789756407e-06,
+      "loss": 1.4434,
+      "step": 417
+    },
+    {
+      "epoch": 0.6688,
+      "grad_norm": 0.8646379134587265,
+      "learning_rate": 9.55219517764421e-06,
+      "loss": 1.459,
+      "step": 418
+    },
+    {
+      "epoch": 0.6704,
+      "grad_norm": 0.758799569096621,
+      "learning_rate": 9.548335778869342e-06,
+      "loss": 1.384,
+      "step": 419
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 0.7746275652822688,
+      "learning_rate": 9.544460606815901e-06,
+      "loss": 1.4165,
+      "step": 420
+    },
+    {
+      "epoch": 0.6736,
+      "grad_norm": 0.7158654192176451,
+      "learning_rate": 9.540569674922685e-06,
+      "loss": 1.3996,
+      "step": 421
+    },
+    {
+      "epoch": 0.6752,
+      "grad_norm": 0.7027455721712468,
+      "learning_rate": 9.536662996683146e-06,
+      "loss": 1.3867,
+      "step": 422
+    },
+    {
+      "epoch": 0.6768,
+      "grad_norm": 0.7372716549607049,
+      "learning_rate": 9.532740585645346e-06,
+      "loss": 1.3957,
+      "step": 423
+    },
+    {
+      "epoch": 0.6784,
+      "grad_norm": 0.7438353862313923,
+      "learning_rate": 9.528802455411902e-06,
+      "loss": 1.3963,
+      "step": 424
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8078197965954008,
+      "learning_rate": 9.52484861963995e-06,
+      "loss": 1.425,
+      "step": 425
+    },
+    {
+      "epoch": 0.6816,
+      "grad_norm": 0.7440732373898441,
+      "learning_rate": 9.520879092041085e-06,
+      "loss": 1.3679,
+      "step": 426
+    },
+    {
+      "epoch": 0.6832,
+      "grad_norm": 0.8255186215076498,
+      "learning_rate": 9.516893886381324e-06,
+      "loss": 1.4301,
+      "step": 427
+    },
+    {
+      "epoch": 0.6848,
+      "grad_norm": 0.7578884617843161,
+      "learning_rate": 9.512893016481053e-06,
+      "loss": 1.3571,
+      "step": 428
+    },
+    {
+      "epoch": 0.6864,
+      "grad_norm": 0.7828225273097063,
+      "learning_rate": 9.508876496214983e-06,
+      "loss": 1.3333,
+      "step": 429
+    },
+    {
+      "epoch": 0.688,
+      "grad_norm": 0.8612791870905079,
+      "learning_rate": 9.504844339512096e-06,
+      "loss": 1.4613,
+      "step": 430
+    },
+    {
+      "epoch": 0.6896,
+      "grad_norm": 0.7693367492174544,
+      "learning_rate": 9.500796560355603e-06,
+      "loss": 1.3582,
+      "step": 431
+    },
+    {
+      "epoch": 0.6912,
+      "grad_norm": 0.8089036040160911,
+      "learning_rate": 9.496733172782889e-06,
+      "loss": 1.4509,
+      "step": 432
+    },
+    {
+      "epoch": 0.6928,
+      "grad_norm": 0.8607231738322668,
+      "learning_rate": 9.492654190885469e-06,
+      "loss": 1.3868,
+      "step": 433
+    },
+    {
+      "epoch": 0.6944,
+      "grad_norm": 0.772545086994621,
+      "learning_rate": 9.488559628808939e-06,
+      "loss": 1.442,
+      "step": 434
+    },
+    {
+      "epoch": 0.696,
+      "grad_norm": 0.8756033149939567,
+      "learning_rate": 9.484449500752927e-06,
+      "loss": 1.3518,
+      "step": 435
+    },
+    {
+      "epoch": 0.6976,
+      "grad_norm": 0.852866352429728,
+      "learning_rate": 9.480323820971039e-06,
+      "loss": 1.4421,
+      "step": 436
+    },
+    {
+      "epoch": 0.6992,
+      "grad_norm": 0.8462963818147305,
+      "learning_rate": 9.476182603770814e-06,
+      "loss": 1.421,
+      "step": 437
+    },
+    {
+      "epoch": 0.7008,
+      "grad_norm": 0.8178560240891339,
+      "learning_rate": 9.472025863513676e-06,
+      "loss": 1.441,
+      "step": 438
+    },
+    {
+      "epoch": 0.7024,
+      "grad_norm": 0.8488745798784146,
+      "learning_rate": 9.467853614614883e-06,
+      "loss": 1.4234,
+      "step": 439
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 0.8506852720553284,
+      "learning_rate": 9.46366587154347e-06,
+      "loss": 1.3921,
+      "step": 440
+    },
+    {
+      "epoch": 0.7056,
+      "grad_norm": 0.8817876113520201,
+      "learning_rate": 9.459462648822209e-06,
+      "loss": 1.409,
+      "step": 441
+    },
+    {
+      "epoch": 0.7072,
+      "grad_norm": 0.7830891410440286,
+      "learning_rate": 9.45524396102755e-06,
+      "loss": 1.3541,
+      "step": 442
+    },
+    {
+      "epoch": 0.7088,
+      "grad_norm": 0.816291855175921,
+      "learning_rate": 9.451009822789583e-06,
+      "loss": 1.476,
+      "step": 443
+    },
+    {
+      "epoch": 0.7104,
+      "grad_norm": 0.8051727540979733,
+      "learning_rate": 9.44676024879197e-06,
+      "loss": 1.4324,
+      "step": 444
+    },
+    {
+      "epoch": 0.712,
+      "grad_norm": 0.8459867104324923,
+      "learning_rate": 9.442495253771909e-06,
+      "loss": 1.4406,
+      "step": 445
+    },
+    {
+      "epoch": 0.7136,
+      "grad_norm": 0.7802166814206262,
+      "learning_rate": 9.438214852520073e-06,
+      "loss": 1.4112,
+      "step": 446
+    },
+    {
+      "epoch": 0.7152,
+      "grad_norm": 0.8504004892247203,
+      "learning_rate": 9.433919059880564e-06,
+      "loss": 1.4365,
+      "step": 447
+    },
+    {
+      "epoch": 0.7168,
+      "grad_norm": 0.7608614188240382,
+      "learning_rate": 9.429607890750863e-06,
+      "loss": 1.4301,
+      "step": 448
+    },
+    {
+      "epoch": 0.7184,
+      "grad_norm": 0.8126317547872607,
+      "learning_rate": 9.425281360081769e-06,
+      "loss": 1.4133,
+      "step": 449
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8107857163379556,
+      "learning_rate": 9.420939482877359e-06,
+      "loss": 1.3136,
+      "step": 450
+    },
+    {
+      "epoch": 0.7216,
+      "grad_norm": 0.8374339463916678,
+      "learning_rate": 9.416582274194929e-06,
+      "loss": 1.3531,
+      "step": 451
+    },
+    {
+      "epoch": 0.7232,
+      "grad_norm": 0.8793221787536417,
+      "learning_rate": 9.412209749144947e-06,
+      "loss": 1.3929,
+      "step": 452
+    },
+    {
+      "epoch": 0.7248,
+      "grad_norm": 0.9401120457058638,
+      "learning_rate": 9.40782192289099e-06,
+      "loss": 1.395,
+      "step": 453
+    },
+    {
+      "epoch": 0.7264,
+      "grad_norm": 1.1135379391350015,
+      "learning_rate": 9.4034188106497e-06,
+      "loss": 1.3624,
+      "step": 454
+    },
+    {
+      "epoch": 0.728,
+      "grad_norm": 0.8965033564578084,
+      "learning_rate": 9.399000427690736e-06,
+      "loss": 1.4325,
+      "step": 455
+    },
+    {
+      "epoch": 0.7296,
+      "grad_norm": 0.9169662353659955,
+      "learning_rate": 9.394566789336707e-06,
+      "loss": 1.4518,
+      "step": 456
+    },
+    {
+      "epoch": 0.7312,
+      "grad_norm": 0.9297823728738536,
+      "learning_rate": 9.390117910963132e-06,
+      "loss": 1.4357,
+      "step": 457
+    },
+    {
+      "epoch": 0.7328,
+      "grad_norm": 0.950514791864914,
+      "learning_rate": 9.385653807998376e-06,
+      "loss": 1.4477,
+      "step": 458
+    },
+    {
+      "epoch": 0.7344,
+      "grad_norm": 0.8838875602963611,
+      "learning_rate": 9.381174495923608e-06,
+      "loss": 1.3355,
+      "step": 459
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 0.9312747737980149,
+      "learning_rate": 9.376679990272736e-06,
+      "loss": 1.3921,
+      "step": 460
+    },
+    {
+      "epoch": 0.7376,
+      "grad_norm": 1.259042078655765,
+      "learning_rate": 9.37217030663236e-06,
+      "loss": 1.397,
+      "step": 461
+    },
+    {
+      "epoch": 0.7392,
+      "grad_norm": 0.7890045513528599,
+      "learning_rate": 9.367645460641716e-06,
+      "loss": 1.3674,
+      "step": 462
+    },
+    {
+      "epoch": 0.7408,
+      "grad_norm": 1.076291710104549,
+      "learning_rate": 9.36310546799262e-06,
+      "loss": 1.4191,
+      "step": 463
+    },
+    {
+      "epoch": 0.7424,
+      "grad_norm": 0.9007048846658345,
+      "learning_rate": 9.358550344429421e-06,
+      "loss": 1.3892,
+      "step": 464
+    },
+    {
+      "epoch": 0.744,
+      "grad_norm": 1.0977024251297094,
+      "learning_rate": 9.353980105748934e-06,
+      "loss": 1.441,
+      "step": 465
+    },
+    {
+      "epoch": 0.7456,
+      "grad_norm": 1.2093055668478268,
+      "learning_rate": 9.349394767800397e-06,
+      "loss": 1.5066,
+      "step": 466
+    },
+    {
+      "epoch": 0.7472,
+      "grad_norm": 0.8480683427051406,
+      "learning_rate": 9.344794346485408e-06,
+      "loss": 1.3336,
+      "step": 467
+    },
+    {
+      "epoch": 0.7488,
+      "grad_norm": 1.1756880847210887,
+      "learning_rate": 9.340178857757876e-06,
+      "loss": 1.3751,
+      "step": 468
+    },
+    {
+      "epoch": 0.7504,
+      "grad_norm": 0.8613181944812055,
+      "learning_rate": 9.335548317623957e-06,
+      "loss": 1.3893,
+      "step": 469
+    },
+    {
+      "epoch": 0.752,
+      "grad_norm": 1.0624454785191673,
+      "learning_rate": 9.330902742142013e-06,
+      "loss": 1.3824,
+      "step": 470
+    },
+    {
+      "epoch": 0.7536,
+      "grad_norm": 1.1272607765843166,
+      "learning_rate": 9.326242147422538e-06,
+      "loss": 1.4453,
+      "step": 471
+    },
+    {
+      "epoch": 0.7552,
+      "grad_norm": 1.0837953328133374,
+      "learning_rate": 9.321566549628118e-06,
+      "loss": 1.3709,
+      "step": 472
+    },
+    {
+      "epoch": 0.7568,
+      "grad_norm": 1.268678650702657,
+      "learning_rate": 9.316875964973366e-06,
+      "loss": 1.4476,
+      "step": 473
+    },
+    {
+      "epoch": 0.7584,
+      "grad_norm": 0.9599581255680933,
+      "learning_rate": 9.31217040972487e-06,
+      "loss": 1.3527,
+      "step": 474
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2114088700139298,
+      "learning_rate": 9.307449900201132e-06,
+      "loss": 1.4508,
+      "step": 475
+    },
+    {
+      "epoch": 0.7616,
+      "grad_norm": 1.0613734813415028,
+      "learning_rate": 9.302714452772515e-06,
+      "loss": 1.3664,
+      "step": 476
+    },
+    {
+      "epoch": 0.7632,
+      "grad_norm": 1.343113595509049,
+      "learning_rate": 9.29796408386119e-06,
+      "loss": 1.3716,
+      "step": 477
+    },
+    {
+      "epoch": 0.7648,
+      "grad_norm": 1.1004543943467968,
+      "learning_rate": 9.293198809941067e-06,
+      "loss": 1.3809,
+      "step": 478
+    },
+    {
+      "epoch": 0.7664,
+      "grad_norm": 1.3061097093993579,
+      "learning_rate": 9.288418647537752e-06,
+      "loss": 1.4026,
+      "step": 479
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 0.9948898156259072,
+      "learning_rate": 9.283623613228479e-06,
+      "loss": 1.4657,
+      "step": 480
+    },
+    {
+      "epoch": 0.7696,
+      "grad_norm": 1.2847207174268944,
+      "learning_rate": 9.27881372364206e-06,
+      "loss": 1.4015,
+      "step": 481
+    },
+    {
+      "epoch": 0.7712,
+      "grad_norm": 0.9305699856616605,
+      "learning_rate": 9.27398899545882e-06,
+      "loss": 1.3939,
+      "step": 482
+    },
+    {
+      "epoch": 0.7728,
+      "grad_norm": 1.5116072506582783,
+      "learning_rate": 9.269149445410545e-06,
+      "loss": 1.4132,
+      "step": 483
+    },
+    {
+      "epoch": 0.7744,
+      "grad_norm": 1.0086293795751071,
+      "learning_rate": 9.264295090280424e-06,
+      "loss": 1.4387,
+      "step": 484
+    },
+    {
+      "epoch": 0.776,
+      "grad_norm": 1.847408228066701,
+      "learning_rate": 9.259425946902987e-06,
+      "loss": 1.4244,
+      "step": 485
+    },
+    {
+      "epoch": 0.7776,
+      "grad_norm": 1.4078695634234875,
+      "learning_rate": 9.254542032164047e-06,
+      "loss": 1.3942,
+      "step": 486
+    },
+    {
+      "epoch": 0.7792,
+      "grad_norm": 1.4517182219499245,
+      "learning_rate": 9.249643363000645e-06,
+      "loss": 1.4315,
+      "step": 487
+    },
+    {
+      "epoch": 0.7808,
+      "grad_norm": 1.451406669904316,
+      "learning_rate": 9.24472995640099e-06,
+      "loss": 1.3913,
+      "step": 488
+    },
+    {
+      "epoch": 0.7824,
+      "grad_norm": 1.0938409865761924,
+      "learning_rate": 9.239801829404396e-06,
+      "loss": 1.4313,
+      "step": 489
+    },
+    {
+      "epoch": 0.784,
+      "grad_norm": 1.2506758979933086,
+      "learning_rate": 9.234858999101232e-06,
+      "loss": 1.4715,
+      "step": 490
+    },
+    {
+      "epoch": 0.7856,
+      "grad_norm": 1.1351321562564738,
+      "learning_rate": 9.22990148263285e-06,
+      "loss": 1.4522,
+      "step": 491
+    },
+    {
+      "epoch": 0.7872,
+      "grad_norm": 1.1943926446564825,
+      "learning_rate": 9.224929297191536e-06,
+      "loss": 1.3648,
+      "step": 492
+    },
+    {
+      "epoch": 0.7888,
+      "grad_norm": 1.3695921540675913,
+      "learning_rate": 9.219942460020447e-06,
+      "loss": 1.3939,
+      "step": 493
+    },
+    {
+      "epoch": 0.7904,
+      "grad_norm": 1.1374940277890149,
+      "learning_rate": 9.214940988413552e-06,
+      "loss": 1.3704,
+      "step": 494
+    },
+    {
+      "epoch": 0.792,
+      "grad_norm": 1.207913655341118,
+      "learning_rate": 9.20992489971557e-06,
+      "loss": 1.3793,
+      "step": 495
+    },
+    {
+      "epoch": 0.7936,
+      "grad_norm": 1.1705673770849725,
+      "learning_rate": 9.204894211321906e-06,
+      "loss": 1.4118,
+      "step": 496
+    },
+    {
+      "epoch": 0.7952,
+      "grad_norm": 1.3128400573537733,
+      "learning_rate": 9.199848940678607e-06,
+      "loss": 1.3585,
+      "step": 497
+    },
+    {
+      "epoch": 0.7968,
+      "grad_norm": 1.3091607424430376,
+      "learning_rate": 9.194789105282277e-06,
+      "loss": 1.3669,
+      "step": 498
+    },
+    {
+      "epoch": 0.7984,
+      "grad_norm": 1.477658102699647,
+      "learning_rate": 9.189714722680041e-06,
+      "loss": 1.4156,
+      "step": 499
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3115376817623345,
+      "learning_rate": 9.184625810469468e-06,
+      "loss": 1.3535,
+      "step": 500
+    },
+    {
+      "epoch": 0.8016,
+      "grad_norm": 1.0977364496603061,
+      "learning_rate": 9.179522386298508e-06,
+      "loss": 1.4219,
+      "step": 501
+    },
+    {
+      "epoch": 0.8032,
+      "grad_norm": 1.381776506444585,
+      "learning_rate": 9.174404467865447e-06,
+      "loss": 1.4198,
+      "step": 502
+    },
+    {
+      "epoch": 0.8048,
+      "grad_norm": 0.9817475942284672,
+      "learning_rate": 9.169272072918834e-06,
+      "loss": 1.416,
+      "step": 503
+    },
+    {
+      "epoch": 0.8064,
+      "grad_norm": 1.0728224880994846,
+      "learning_rate": 9.164125219257419e-06,
+      "loss": 1.4459,
+      "step": 504
+    },
+    {
+      "epoch": 0.808,
+      "grad_norm": 1.0181572151654306,
+      "learning_rate": 9.158963924730092e-06,
+      "loss": 1.4039,
+      "step": 505
+    },
+    {
+      "epoch": 0.8096,
+      "grad_norm": 0.9065528635153179,
+      "learning_rate": 9.153788207235827e-06,
+      "loss": 1.3904,
+      "step": 506
+    },
+    {
+      "epoch": 0.8112,
+      "grad_norm": 0.9346490813023509,
+      "learning_rate": 9.148598084723615e-06,
+      "loss": 1.3786,
+      "step": 507
+    },
+    {
+      "epoch": 0.8128,
+      "grad_norm": 0.8659246501890807,
+      "learning_rate": 9.143393575192402e-06,
+      "loss": 1.3721,
+      "step": 508
+    },
+    {
+      "epoch": 0.8144,
+      "grad_norm": 0.9814840082344324,
+      "learning_rate": 9.138174696691025e-06,
+      "loss": 1.3909,
+      "step": 509
+    },
+    {
+      "epoch": 0.816,
+      "grad_norm": 0.8647337721059558,
+      "learning_rate": 9.132941467318152e-06,
+      "loss": 1.3737,
+      "step": 510
+    },
+    {
+      "epoch": 0.8176,
+      "grad_norm": 1.083778198151699,
+      "learning_rate": 9.127693905222223e-06,
+      "loss": 1.3622,
+      "step": 511
+    },
+    {
+      "epoch": 0.8192,
+      "grad_norm": 0.8729519901988788,
+      "learning_rate": 9.122432028601377e-06,
+      "loss": 1.3898,
+      "step": 512
+    },
+    {
+      "epoch": 0.8208,
+      "grad_norm": 1.1217148807433088,
+      "learning_rate": 9.1171558557034e-06,
+      "loss": 1.4515,
+      "step": 513
+    },
+    {
+      "epoch": 0.8224,
+      "grad_norm": 0.9022165699341304,
+      "learning_rate": 9.111865404825652e-06,
+      "loss": 1.4496,
+      "step": 514
+    },
+    {
+      "epoch": 0.824,
+      "grad_norm": 0.9888214552365132,
+      "learning_rate": 9.10656069431501e-06,
+      "loss": 1.3738,
+      "step": 515
+    },
+    {
+      "epoch": 0.8256,
+      "grad_norm": 0.9468409789818583,
+      "learning_rate": 9.101241742567802e-06,
+      "loss": 1.418,
+      "step": 516
+    },
+    {
+      "epoch": 0.8272,
+      "grad_norm": 0.8972310461247193,
+      "learning_rate": 9.095908568029741e-06,
+      "loss": 1.4316,
+      "step": 517
+    },
+    {
+      "epoch": 0.8288,
+      "grad_norm": 0.8669980947883519,
+      "learning_rate": 9.09056118919587e-06,
+      "loss": 1.4441,
+      "step": 518
+    },
+    {
+      "epoch": 0.8304,
+      "grad_norm": 0.8175057891786413,
+      "learning_rate": 9.085199624610486e-06,
+      "loss": 1.3556,
+      "step": 519
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 0.9367182742699341,
+      "learning_rate": 9.079823892867083e-06,
+      "loss": 1.4464,
+      "step": 520
+    },
+    {
+      "epoch": 0.8336,
+      "grad_norm": 0.7796305912637057,
+      "learning_rate": 9.074434012608282e-06,
+      "loss": 1.4291,
+      "step": 521
+    },
+    {
+      "epoch": 0.8352,
+      "grad_norm": 0.9069526169820784,
+      "learning_rate": 9.069030002525777e-06,
+      "loss": 1.3164,
+      "step": 522
+    },
+    {
+      "epoch": 0.8368,
+      "grad_norm": 0.837307689038858,
+      "learning_rate": 9.063611881360258e-06,
+      "loss": 1.3571,
+      "step": 523
+    },
+    {
+      "epoch": 0.8384,
+      "grad_norm": 0.8941222777198047,
+      "learning_rate": 9.05817966790135e-06,
+      "loss": 1.361,
+      "step": 524
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.843789303204077,
+      "learning_rate": 9.052733380987555e-06,
+      "loss": 1.4937,
+      "step": 525
+    },
+    {
+      "epoch": 0.8416,
+      "grad_norm": 0.8309049745967986,
+      "learning_rate": 9.047273039506174e-06,
+      "loss": 1.3751,
+      "step": 526
+    },
+    {
+      "epoch": 0.8432,
+      "grad_norm": 0.8444741514042486,
+      "learning_rate": 9.041798662393255e-06,
+      "loss": 1.4154,
+      "step": 527
+    },
+    {
+      "epoch": 0.8448,
+      "grad_norm": 0.7443057914206279,
+      "learning_rate": 9.036310268633515e-06,
+      "loss": 1.43,
+      "step": 528
+    },
+    {
+      "epoch": 0.8464,
+      "grad_norm": 0.880401296762967,
+      "learning_rate": 9.030807877260278e-06,
+      "loss": 1.4005,
+      "step": 529
+    },
+    {
+      "epoch": 0.848,
+      "grad_norm": 0.6716991446528822,
+      "learning_rate": 9.025291507355419e-06,
+      "loss": 1.2757,
+      "step": 530
+    },
+    {
+      "epoch": 0.8496,
+      "grad_norm": 0.8260344337985983,
+      "learning_rate": 9.01976117804928e-06,
+      "loss": 1.3531,
+      "step": 531
+    },
+    {
+      "epoch": 0.8512,
+      "grad_norm": 0.732475965261935,
+      "learning_rate": 9.014216908520619e-06,
+      "loss": 1.3923,
+      "step": 532
+    },
+    {
+      "epoch": 0.8528,
+      "grad_norm": 0.7781784793257497,
+      "learning_rate": 9.008658717996538e-06,
+      "loss": 1.4252,
+      "step": 533
+    },
+    {
+      "epoch": 0.8544,
+      "grad_norm": 0.7470827467959966,
+      "learning_rate": 9.003086625752414e-06,
+      "loss": 1.5211,
+      "step": 534
+    },
+    {
+      "epoch": 0.856,
+      "grad_norm": 0.893401171942822,
+      "learning_rate": 8.997500651111833e-06,
+      "loss": 1.4675,
+      "step": 535
+    },
+    {
+      "epoch": 0.8576,
+      "grad_norm": 0.7496425990746173,
+      "learning_rate": 8.991900813446523e-06,
+      "loss": 1.3747,
+      "step": 536
+    },
+    {
+      "epoch": 0.8592,
+      "grad_norm": 0.8941121755968526,
+      "learning_rate": 8.986287132176295e-06,
+      "loss": 1.4127,
+      "step": 537
+    },
+    {
+      "epoch": 0.8608,
+      "grad_norm": 0.852246198988937,
+      "learning_rate": 8.980659626768961e-06,
+      "loss": 1.4069,
+      "step": 538
+    },
+    {
+      "epoch": 0.8624,
+      "grad_norm": 0.7696950633368015,
+      "learning_rate": 8.975018316740278e-06,
+      "loss": 1.3873,
+      "step": 539
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 0.8082984195611446,
+      "learning_rate": 8.969363221653875e-06,
+      "loss": 1.4684,
+      "step": 540
+    },
+    {
+      "epoch": 0.8656,
+      "grad_norm": 0.8128028102014774,
+      "learning_rate": 8.963694361121186e-06,
+      "loss": 1.4019,
+      "step": 541
+    },
+    {
+      "epoch": 0.8672,
+      "grad_norm": 0.6720640233339992,
+      "learning_rate": 8.958011754801383e-06,
+      "loss": 1.3883,
+      "step": 542
+    },
+    {
+      "epoch": 0.8688,
+      "grad_norm": 0.8439128125139631,
+      "learning_rate": 8.952315422401307e-06,
+      "loss": 1.4326,
+      "step": 543
+    },
+    {
+      "epoch": 0.8704,
+      "grad_norm": 0.7083982477117675,
+      "learning_rate": 8.946605383675403e-06,
+      "loss": 1.3882,
+      "step": 544
+    },
+    {
+      "epoch": 0.872,
+      "grad_norm": 0.8274456057681323,
+      "learning_rate": 8.940881658425645e-06,
+      "loss": 1.3796,
+      "step": 545
+    },
+    {
+      "epoch": 0.8736,
+      "grad_norm": 0.7144147412305394,
+      "learning_rate": 8.93514426650147e-06,
+      "loss": 1.4396,
+      "step": 546
+    },
+    {
+      "epoch": 0.8752,
+      "grad_norm": 0.7638988341896414,
+      "learning_rate": 8.929393227799715e-06,
+      "loss": 1.4192,
+      "step": 547
+    },
+    {
+      "epoch": 0.8768,
+      "grad_norm": 0.7657318679800025,
+      "learning_rate": 8.923628562264536e-06,
+      "loss": 1.3994,
+      "step": 548
+    },
+    {
+      "epoch": 0.8784,
+      "grad_norm": 0.7793813938401812,
+      "learning_rate": 8.917850289887353e-06,
+      "loss": 1.3896,
+      "step": 549
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8770291138872169,
+      "learning_rate": 8.91205843070677e-06,
+      "loss": 1.4351,
+      "step": 550
+    },
+    {
+      "epoch": 0.8816,
+      "grad_norm": 0.8206781095731741,
+      "learning_rate": 8.906253004808506e-06,
+      "loss": 1.4389,
+      "step": 551
+    },
+    {
+      "epoch": 0.8832,
+      "grad_norm": 0.7810396801244909,
+      "learning_rate": 8.900434032325332e-06,
+      "loss": 1.3316,
+      "step": 552
+    },
+    {
+      "epoch": 0.8848,
+      "grad_norm": 0.8485630183902447,
+      "learning_rate": 8.894601533437e-06,
+      "loss": 1.451,
+      "step": 553
+    },
+    {
+      "epoch": 0.8864,
+      "grad_norm": 0.7671493124236161,
+      "learning_rate": 8.888755528370163e-06,
+      "loss": 1.3241,
+      "step": 554
+    },
+    {
+      "epoch": 0.888,
+      "grad_norm": 0.8507523598271359,
+      "learning_rate": 8.882896037398322e-06,
+      "loss": 1.3401,
+      "step": 555
+    },
+    {
+      "epoch": 0.8896,
+      "grad_norm": 0.7826825855082984,
+      "learning_rate": 8.877023080841739e-06,
+      "loss": 1.4857,
+      "step": 556
+    },
+    {
+      "epoch": 0.8912,
+      "grad_norm": 0.9309896728064471,
+      "learning_rate": 8.871136679067372e-06,
+      "loss": 1.4469,
+      "step": 557
+    },
+    {
+      "epoch": 0.8928,
+      "grad_norm": 0.7865685397938487,
+      "learning_rate": 8.865236852488813e-06,
+      "loss": 1.4324,
+      "step": 558
+    },
+    {
+      "epoch": 0.8944,
+      "grad_norm": 0.9169877802542052,
+      "learning_rate": 8.859323621566207e-06,
+      "loss": 1.3443,
+      "step": 559
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 0.8312864547987815,
+      "learning_rate": 8.853397006806183e-06,
+      "loss": 1.3738,
+      "step": 560
+    },
+    {
+      "epoch": 0.8976,
+      "grad_norm": 0.8654767265603543,
+      "learning_rate": 8.847457028761783e-06,
+      "loss": 1.4199,
+      "step": 561
+    },
+    {
+      "epoch": 0.8992,
+      "grad_norm": 0.8164510514060842,
+      "learning_rate": 8.841503708032398e-06,
+      "loss": 1.369,
+      "step": 562
+    },
+    {
+      "epoch": 0.9008,
+      "grad_norm": 0.8440095293724006,
+      "learning_rate": 8.835537065263684e-06,
+      "loss": 1.3878,
+      "step": 563
+    },
+    {
+      "epoch": 0.9024,
+      "grad_norm": 0.8714469266860524,
+      "learning_rate": 8.829557121147499e-06,
+      "loss": 1.3663,
+      "step": 564
+    },
+    {
+      "epoch": 0.904,
+      "grad_norm": 0.8343895256971336,
+      "learning_rate": 8.82356389642183e-06,
+      "loss": 1.361,
+      "step": 565
+    },
+    {
+      "epoch": 0.9056,
+      "grad_norm": 1.0196547909632545,
+      "learning_rate": 8.817557411870717e-06,
+      "loss": 1.3728,
+      "step": 566
+    },
+    {
+      "epoch": 0.9072,
+      "grad_norm": 0.7577032213218948,
+      "learning_rate": 8.811537688324187e-06,
+      "loss": 1.3841,
+      "step": 567
+    },
+    {
+      "epoch": 0.9088,
+      "grad_norm": 0.911560919949406,
+      "learning_rate": 8.805504746658183e-06,
+      "loss": 1.4591,
+      "step": 568
+    },
+    {
+      "epoch": 0.9104,
+      "grad_norm": 0.8156307131074223,
+      "learning_rate": 8.799458607794476e-06,
+      "loss": 1.4198,
+      "step": 569
+    },
+    {
+      "epoch": 0.912,
+      "grad_norm": 0.8011154864034381,
+      "learning_rate": 8.793399292700616e-06,
+      "loss": 1.4119,
+      "step": 570
+    },
+    {
+      "epoch": 0.9136,
+      "grad_norm": 0.8526685441242386,
+      "learning_rate": 8.787326822389836e-06,
+      "loss": 1.4242,
+      "step": 571
+    },
+    {
+      "epoch": 0.9152,
+      "grad_norm": 0.7400921918737187,
+      "learning_rate": 8.781241217921e-06,
+      "loss": 1.3382,
+      "step": 572
+    },
+    {
+      "epoch": 0.9168,
+      "grad_norm": 0.9312947506236492,
+      "learning_rate": 8.775142500398513e-06,
+      "loss": 1.3812,
+      "step": 573
+    },
+    {
+      "epoch": 0.9184,
+      "grad_norm": 0.7336472742040459,
+      "learning_rate": 8.769030690972262e-06,
+      "loss": 1.3789,
+      "step": 574
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8493178435884886,
+      "learning_rate": 8.76290581083753e-06,
+      "loss": 1.3671,
+      "step": 575
+    },
+    {
+      "epoch": 0.9216,
+      "grad_norm": 0.8368764087136009,
+      "learning_rate": 8.756767881234928e-06,
+      "loss": 1.377,
+      "step": 576
+    },
+    {
+      "epoch": 0.9232,
+      "grad_norm": 0.7698225633480201,
+      "learning_rate": 8.750616923450328e-06,
+      "loss": 1.4029,
+      "step": 577
+    },
+    {
+      "epoch": 0.9248,
+      "grad_norm": 0.8361698981543135,
+      "learning_rate": 8.744452958814775e-06,
+      "loss": 1.4173,
+      "step": 578
+    },
+    {
+      "epoch": 0.9264,
+      "grad_norm": 0.7587508604626323,
+      "learning_rate": 8.738276008704426e-06,
+      "loss": 1.3932,
+      "step": 579
+    },
+    {
+      "epoch": 0.928,
+      "grad_norm": 0.8546549871905071,
+      "learning_rate": 8.732086094540467e-06,
+      "loss": 1.3361,
+      "step": 580
+    },
+    {
+      "epoch": 0.9296,
+      "grad_norm": 0.7421912173920685,
+      "learning_rate": 8.725883237789046e-06,
+      "loss": 1.3456,
+      "step": 581
+    },
+    {
+      "epoch": 0.9312,
+      "grad_norm": 0.825615878564762,
+      "learning_rate": 8.719667459961191e-06,
+      "loss": 1.3463,
+      "step": 582
+    },
+    {
+      "epoch": 0.9328,
+      "grad_norm": 0.7698272298025547,
+      "learning_rate": 8.713438782612743e-06,
+      "loss": 1.3966,
+      "step": 583
+    },
+    {
+      "epoch": 0.9344,
+      "grad_norm": 0.7218931975678564,
+      "learning_rate": 8.707197227344275e-06,
+      "loss": 1.4218,
+      "step": 584
+    },
+    {
+      "epoch": 0.936,
+      "grad_norm": 0.7955230482449089,
+      "learning_rate": 8.700942815801023e-06,
+      "loss": 1.3754,
+      "step": 585
+    },
+    {
+      "epoch": 0.9376,
+      "grad_norm": 0.791857871190122,
+      "learning_rate": 8.6946755696728e-06,
+      "loss": 1.3982,
+      "step": 586
+    },
+    {
+      "epoch": 0.9392,
+      "grad_norm": 0.7999920197140337,
+      "learning_rate": 8.688395510693939e-06,
+      "loss": 1.3688,
+      "step": 587
+    },
+    {
+      "epoch": 0.9408,
+      "grad_norm": 0.7866559631960182,
+      "learning_rate": 8.682102660643196e-06,
+      "loss": 1.3761,
+      "step": 588
+    },
+    {
+      "epoch": 0.9424,
+      "grad_norm": 0.9296039514954354,
+      "learning_rate": 8.675797041343696e-06,
+      "loss": 1.3435,
+      "step": 589
+    },
+    {
+      "epoch": 0.944,
+      "grad_norm": 0.7357362550391597,
+      "learning_rate": 8.669478674662839e-06,
+      "loss": 1.3931,
+      "step": 590
+    },
+    {
+      "epoch": 0.9456,
+      "grad_norm": 0.8334261274846441,
+      "learning_rate": 8.663147582512232e-06,
+      "loss": 1.4285,
+      "step": 591
+    },
+    {
+      "epoch": 0.9472,
+      "grad_norm": 0.7846577474630272,
+      "learning_rate": 8.65680378684762e-06,
+      "loss": 1.4735,
+      "step": 592
+    },
+    {
+      "epoch": 0.9488,
+      "grad_norm": 0.8000689076804095,
+      "learning_rate": 8.6504473096688e-06,
+      "loss": 1.396,
+      "step": 593
+    },
+    {
+      "epoch": 0.9504,
+      "grad_norm": 0.8644457289912572,
+      "learning_rate": 8.64407817301954e-06,
+      "loss": 1.3883,
+      "step": 594
+    },
+    {
+      "epoch": 0.952,
+      "grad_norm": 0.8170562231694108,
+      "learning_rate": 8.637696398987517e-06,
+      "loss": 1.3259,
+      "step": 595
+    },
+    {
+      "epoch": 0.9536,
+      "grad_norm": 0.8371277076085974,
+      "learning_rate": 8.631302009704235e-06,
+      "loss": 1.3252,
+      "step": 596
+    },
+    {
+      "epoch": 0.9552,
+      "grad_norm": 0.8071669179884052,
+      "learning_rate": 8.624895027344943e-06,
+      "loss": 1.4278,
+      "step": 597
+    },
+    {
+      "epoch": 0.9568,
+      "grad_norm": 0.7606065945932591,
+      "learning_rate": 8.618475474128563e-06,
+      "loss": 1.3994,
+      "step": 598
+    },
+    {
+      "epoch": 0.9584,
+      "grad_norm": 0.7911564860445692,
+      "learning_rate": 8.61204337231761e-06,
+      "loss": 1.4521,
+      "step": 599
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7585068470599076,
+      "learning_rate": 8.605598744218122e-06,
+      "loss": 1.3503,
+      "step": 600
+    },
+    {
+      "epoch": 0.9616,
+      "grad_norm": 0.7265778403448,
+      "learning_rate": 8.599141612179572e-06,
+      "loss": 1.3914,
+      "step": 601
+    },
+    {
+      "epoch": 0.9632,
+      "grad_norm": 0.7470831601666178,
+      "learning_rate": 8.592671998594794e-06,
+      "loss": 1.3741,
+      "step": 602
+    },
+    {
+      "epoch": 0.9648,
+      "grad_norm": 0.81581183512277,
+      "learning_rate": 8.586189925899913e-06,
+      "loss": 1.4238,
+      "step": 603
+    },
+    {
+      "epoch": 0.9664,
+      "grad_norm": 0.7993638957190318,
+      "learning_rate": 8.57969541657426e-06,
+      "loss": 1.3444,
+      "step": 604
+    },
+    {
+      "epoch": 0.968,
+      "grad_norm": 0.9408590183214677,
+      "learning_rate": 8.57318849314029e-06,
+      "loss": 1.3237,
+      "step": 605
+    },
+    {
+      "epoch": 0.9696,
+      "grad_norm": 0.7631902378064993,
+      "learning_rate": 8.566669178163513e-06,
+      "loss": 1.3711,
+      "step": 606
+    },
+    {
+      "epoch": 0.9712,
+      "grad_norm": 0.7635574738613359,
+      "learning_rate": 8.560137494252416e-06,
+      "loss": 1.3937,
+      "step": 607
+    },
+    {
+      "epoch": 0.9728,
+      "grad_norm": 0.7659536219288624,
+      "learning_rate": 8.553593464058374e-06,
+      "loss": 1.4676,
+      "step": 608
+    },
+    {
+      "epoch": 0.9744,
+      "grad_norm": 0.7311299523187105,
+      "learning_rate": 8.54703711027558e-06,
+      "loss": 1.3734,
+      "step": 609
+    },
+    {
+      "epoch": 0.976,
+      "grad_norm": 0.8365969393256748,
+      "learning_rate": 8.540468455640964e-06,
+      "loss": 1.3712,
+      "step": 610
+    },
+    {
+      "epoch": 0.9776,
+      "grad_norm": 0.728495021132979,
+      "learning_rate": 8.533887522934114e-06,
+      "loss": 1.4531,
+      "step": 611
+    },
+    {
+      "epoch": 0.9792,
+      "grad_norm": 0.8344628082616679,
+      "learning_rate": 8.527294334977201e-06,
+      "loss": 1.4244,
+      "step": 612
+    },
+    {
+      "epoch": 0.9808,
+      "grad_norm": 0.8072727935508021,
+      "learning_rate": 8.520688914634894e-06,
+      "loss": 1.5216,
+      "step": 613
+    },
+    {
+      "epoch": 0.9824,
+      "grad_norm": 0.8300186765698186,
+      "learning_rate": 8.51407128481428e-06,
+      "loss": 1.4609,
+      "step": 614
+    },
+    {
+      "epoch": 0.984,
+      "grad_norm": 0.9000138499069039,
+      "learning_rate": 8.507441468464792e-06,
+      "loss": 1.4012,
+      "step": 615
+    },
+    {
+      "epoch": 0.9856,
+      "grad_norm": 0.7761668117516761,
+      "learning_rate": 8.50079948857812e-06,
+      "loss": 1.4176,
+      "step": 616
+    },
+    {
+      "epoch": 0.9872,
+      "grad_norm": 0.9530505538051826,
+      "learning_rate": 8.494145368188143e-06,
+      "loss": 1.3909,
+      "step": 617
+    },
+    {
+      "epoch": 0.9888,
+      "grad_norm": 0.8035311888047029,
+      "learning_rate": 8.487479130370838e-06,
+      "loss": 1.4116,
+      "step": 618
+    },
+    {
+      "epoch": 0.9904,
+      "grad_norm": 0.8862605817145259,
+      "learning_rate": 8.480800798244202e-06,
+      "loss": 1.358,
+      "step": 619
+    },
+    {
+      "epoch": 0.992,
+      "grad_norm": 0.7318635814501507,
+      "learning_rate": 8.47411039496818e-06,
+      "loss": 1.4554,
+      "step": 620
+    },
+    {
+      "epoch": 0.9936,
+      "grad_norm": 0.9500697250471559,
+      "learning_rate": 8.467407943744574e-06,
+      "loss": 1.3743,
+      "step": 621
+    },
+    {
+      "epoch": 0.9952,
+      "grad_norm": 0.7676402074781837,
+      "learning_rate": 8.460693467816972e-06,
+      "loss": 1.3213,
+      "step": 622
+    },
+    {
+      "epoch": 0.9968,
+      "grad_norm": 0.810130778656904,
+      "learning_rate": 8.453966990470656e-06,
+      "loss": 1.4763,
+      "step": 623
+    },
+    {
+      "epoch": 0.9984,
+      "grad_norm": 0.7776259967378215,
+      "learning_rate": 8.447228535032536e-06,
+      "loss": 1.3362,
+      "step": 624
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7880775004505816,
+      "learning_rate": 8.440478124871054e-06,
+      "loss": 1.4009,
+      "step": 625
+    },
+    {
+      "epoch": 1.0016,
+      "grad_norm": 0.8867802400151525,
+      "learning_rate": 8.433715783396115e-06,
+      "loss": 1.3205,
+      "step": 626
+    },
+    {
+      "epoch": 1.0032,
+      "grad_norm": 0.7704672946258602,
+      "learning_rate": 8.426941534058999e-06,
+      "loss": 1.321,
+      "step": 627
+    },
+    {
+      "epoch": 1.0048,
+      "grad_norm": 0.8242884968004639,
+      "learning_rate": 8.420155400352279e-06,
+      "loss": 1.298,
+      "step": 628
+    },
+    {
+      "epoch": 1.0064,
+      "grad_norm": 0.7976797442221665,
+      "learning_rate": 8.413357405809748e-06,
+      "loss": 1.2982,
+      "step": 629
+    },
+    {
+      "epoch": 1.008,
+      "grad_norm": 0.8331209000285394,
+      "learning_rate": 8.406547574006326e-06,
+      "loss": 1.362,
+      "step": 630
+    },
+    {
+      "epoch": 1.0096,
+      "grad_norm": 0.9079042062408942,
+      "learning_rate": 8.399725928557985e-06,
+      "loss": 1.3712,
+      "step": 631
+    },
+    {
+      "epoch": 1.0112,
+      "grad_norm": 0.8056151779345211,
+      "learning_rate": 8.39289249312167e-06,
+      "loss": 1.3904,
+      "step": 632
+    },
+    {
+      "epoch": 1.0128,
+      "grad_norm": 0.7981574156500866,
+      "learning_rate": 8.386047291395208e-06,
+      "loss": 1.2818,
+      "step": 633
+    },
+    {
+      "epoch": 1.0144,
+      "grad_norm": 0.8472205588779337,
+      "learning_rate": 8.37919034711723e-06,
+      "loss": 1.3237,
+      "step": 634
+    },
+    {
+      "epoch": 1.016,
+      "grad_norm": 0.8251340809986435,
+      "learning_rate": 8.372321684067092e-06,
+      "loss": 1.4003,
+      "step": 635
+    },
+    {
+      "epoch": 1.0176,
+      "grad_norm": 0.9766998483870328,
+      "learning_rate": 8.36544132606479e-06,
+      "loss": 1.3457,
+      "step": 636
+    },
+    {
+      "epoch": 1.0192,
+      "grad_norm": 0.7990690452078489,
+      "learning_rate": 8.358549296970877e-06,
+      "loss": 1.3664,
+      "step": 637
+    },
+    {
+      "epoch": 1.0208,
+      "grad_norm": 0.7931453037773057,
+      "learning_rate": 8.351645620686377e-06,
+      "loss": 1.2956,
+      "step": 638
+    },
+    {
+      "epoch": 1.0224,
+      "grad_norm": 0.796625151037376,
+      "learning_rate": 8.34473032115271e-06,
+      "loss": 1.3428,
+      "step": 639
+    },
+    {
+      "epoch": 1.024,
+      "grad_norm": 0.7914312126729364,
+      "learning_rate": 8.337803422351602e-06,
+      "loss": 1.3165,
+      "step": 640
+    },
+    {
+      "epoch": 1.0256,
+      "grad_norm": 0.8508455746721232,
+      "learning_rate": 8.33086494830501e-06,
+      "loss": 1.3266,
+      "step": 641
+    },
+    {
+      "epoch": 1.0272,
+      "grad_norm": 0.8351115999256439,
+      "learning_rate": 8.323914923075018e-06,
+      "loss": 1.3695,
+      "step": 642
+    },
+    {
+      "epoch": 1.0288,
+      "grad_norm": 0.8365236361304882,
+      "learning_rate": 8.316953370763788e-06,
+      "loss": 1.3296,
+      "step": 643
+    },
+    {
+      "epoch": 1.0304,
+      "grad_norm": 0.8655039393465203,
+      "learning_rate": 8.309980315513444e-06,
+      "loss": 1.348,
+      "step": 644
+    },
+    {
+      "epoch": 1.032,
+      "grad_norm": 0.9412337563817211,
+      "learning_rate": 8.302995781506007e-06,
+      "loss": 1.3371,
+      "step": 645
+    },
+    {
+      "epoch": 1.0336,
+      "grad_norm": 0.9819048382543066,
+      "learning_rate": 8.295999792963301e-06,
+      "loss": 1.338,
+      "step": 646
+    },
+    {
+      "epoch": 1.0352,
+      "grad_norm": 0.8803888815417968,
+      "learning_rate": 8.288992374146878e-06,
+      "loss": 1.3841,
+      "step": 647
+    },
+    {
+      "epoch": 1.0368,
+      "grad_norm": 0.8873381083249229,
+      "learning_rate": 8.281973549357927e-06,
+      "loss": 1.3489,
+      "step": 648
+    },
+    {
+      "epoch": 1.0384,
+      "grad_norm": 0.8110257851682673,
+      "learning_rate": 8.274943342937191e-06,
+      "loss": 1.3977,
+      "step": 649
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.7568949087069258,
+      "learning_rate": 8.267901779264889e-06,
+      "loss": 1.3049,
+      "step": 650
+    },
+    {
+      "epoch": 1.0416,
+      "grad_norm": 0.8695655997977444,
+      "learning_rate": 8.260848882760616e-06,
+      "loss": 1.3195,
+      "step": 651
+    },
+    {
+      "epoch": 1.0432,
+      "grad_norm": 0.8456850505104982,
+      "learning_rate": 8.25378467788328e-06,
+      "loss": 1.3353,
+      "step": 652
+    },
+    {
+      "epoch": 1.0448,
+      "grad_norm": 0.7701724352760845,
+      "learning_rate": 8.246709189130997e-06,
+      "loss": 1.2889,
+      "step": 653
+    },
+    {
+      "epoch": 1.0464,
+      "grad_norm": 0.9339573362592145,
+      "learning_rate": 8.23962244104102e-06,
+      "loss": 1.3934,
+      "step": 654
+    },
+    {
+      "epoch": 1.048,
+      "grad_norm": 0.7715767204608774,
+      "learning_rate": 8.232524458189644e-06,
+      "loss": 1.3182,
+      "step": 655
+    },
+    {
+      "epoch": 1.0496,
+      "grad_norm": 0.8263393275238416,
+      "learning_rate": 8.225415265192126e-06,
+      "loss": 1.3516,
+      "step": 656
+    },
+    {
+      "epoch": 1.0512,
+      "grad_norm": 0.8074184002839383,
+      "learning_rate": 8.218294886702606e-06,
+      "loss": 1.356,
+      "step": 657
+    },
+    {
+      "epoch": 1.0528,
+      "grad_norm": 0.8222603075760709,
+      "learning_rate": 8.211163347414005e-06,
+      "loss": 1.3942,
+      "step": 658
+    },
+    {
+      "epoch": 1.0544,
+      "grad_norm": 0.8648103668843804,
+      "learning_rate": 8.20402067205795e-06,
+      "loss": 1.3699,
+      "step": 659
+    },
+    {
+      "epoch": 1.056,
+      "grad_norm": 0.7815582606112759,
+      "learning_rate": 8.196866885404697e-06,
+      "loss": 1.2801,
+      "step": 660
+    },
+    {
+      "epoch": 1.0576,
+      "grad_norm": 0.8495229318095386,
+      "learning_rate": 8.18970201226302e-06,
+      "loss": 1.3495,
+      "step": 661
+    },
+    {
+      "epoch": 1.0592,
+      "grad_norm": 0.7463706798831494,
+      "learning_rate": 8.182526077480153e-06,
+      "loss": 1.3119,
+      "step": 662
+    },
+    {
+      "epoch": 1.0608,
+      "grad_norm": 0.8479455595082831,
+      "learning_rate": 8.175339105941685e-06,
+      "loss": 1.3228,
+      "step": 663
+    },
+    {
+      "epoch": 1.0624,
+      "grad_norm": 0.7206075398678762,
+      "learning_rate": 8.168141122571478e-06,
+      "loss": 1.3533,
+      "step": 664
+    },
+    {
+      "epoch": 1.064,
+      "grad_norm": 0.7136084973860434,
+      "learning_rate": 8.160932152331587e-06,
+      "loss": 1.3376,
+      "step": 665
+    },
+    {
+      "epoch": 1.0656,
+      "grad_norm": 0.7554190988596579,
+      "learning_rate": 8.153712220222163e-06,
+      "loss": 1.3233,
+      "step": 666
+    },
+    {
+      "epoch": 1.0672,
+      "grad_norm": 0.8231062501131365,
+      "learning_rate": 8.14648135128138e-06,
+      "loss": 1.2972,
+      "step": 667
+    },
+    {
+      "epoch": 1.0688,
+      "grad_norm": 0.7337004363624503,
+      "learning_rate": 8.139239570585334e-06,
+      "loss": 1.3526,
+      "step": 668
+    },
+    {
+      "epoch": 1.0704,
+      "grad_norm": 0.6888414813060557,
+      "learning_rate": 8.131986903247959e-06,
+      "loss": 1.319,
+      "step": 669
+    },
+    {
+      "epoch": 1.072,
+      "grad_norm": 0.8010656753693667,
+      "learning_rate": 8.124723374420951e-06,
+      "loss": 1.2795,
+      "step": 670
+    },
+    {
+      "epoch": 1.0735999999999999,
+      "grad_norm": 0.6786098646282538,
+      "learning_rate": 8.117449009293668e-06,
+      "loss": 1.3291,
+      "step": 671
+    },
+    {
+      "epoch": 1.0752,
+      "grad_norm": 0.7514142318880065,
+      "learning_rate": 8.11016383309305e-06,
+      "loss": 1.2829,
+      "step": 672
+    },
+    {
+      "epoch": 1.0768,
+      "grad_norm": 0.7185909282143551,
+      "learning_rate": 8.102867871083528e-06,
+      "loss": 1.3258,
+      "step": 673
+    },
+    {
+      "epoch": 1.0784,
+      "grad_norm": 0.7466318983684804,
+      "learning_rate": 8.095561148566932e-06,
+      "loss": 1.3476,
+      "step": 674
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.806660372632054,
+      "learning_rate": 8.088243690882421e-06,
+      "loss": 1.3246,
+      "step": 675
+    },
+    {
+      "epoch": 1.0816,
+      "grad_norm": 0.6971522813323874,
+      "learning_rate": 8.080915523406371e-06,
+      "loss": 1.306,
+      "step": 676
+    },
+    {
+      "epoch": 1.0832,
+      "grad_norm": 0.744537846025582,
+      "learning_rate": 8.073576671552303e-06,
+      "loss": 1.2986,
+      "step": 677
+    },
+    {
+      "epoch": 1.0848,
+      "grad_norm": 0.7770302333753993,
+      "learning_rate": 8.06622716077079e-06,
+      "loss": 1.3673,
+      "step": 678
+    },
+    {
+      "epoch": 1.0864,
+      "grad_norm": 0.7430935263291278,
+      "learning_rate": 8.058867016549372e-06,
+      "loss": 1.314,
+      "step": 679
+    },
+    {
+      "epoch": 1.088,
+      "grad_norm": 0.7329047709128129,
+      "learning_rate": 8.051496264412464e-06,
+      "loss": 1.3525,
+      "step": 680
+    },
+    {
+      "epoch": 1.0896,
+      "grad_norm": 0.7027439776396316,
+      "learning_rate": 8.044114929921264e-06,
+      "loss": 1.3224,
+      "step": 681
+    },
+    {
+      "epoch": 1.0912,
+      "grad_norm": 0.6796328435447017,
+      "learning_rate": 8.036723038673675e-06,
+      "loss": 1.3182,
+      "step": 682
+    },
+    {
+      "epoch": 1.0928,
+      "grad_norm": 0.7535723193130166,
+      "learning_rate": 8.029320616304204e-06,
+      "loss": 1.2934,
+      "step": 683
+    },
+    {
+      "epoch": 1.0944,
+      "grad_norm": 0.68817751196644,
+      "learning_rate": 8.021907688483885e-06,
+      "loss": 1.3373,
+      "step": 684
+    },
+    {
+      "epoch": 1.096,
+      "grad_norm": 0.751933090820658,
+      "learning_rate": 8.01448428092018e-06,
+      "loss": 1.3275,
+      "step": 685
+    },
+    {
+      "epoch": 1.0976,
+      "grad_norm": 0.7480069902232998,
+      "learning_rate": 8.007050419356898e-06,
+      "loss": 1.3301,
+      "step": 686
+    },
+    {
+      "epoch": 1.0992,
+      "grad_norm": 0.7516952235670269,
+      "learning_rate": 7.999606129574096e-06,
+      "loss": 1.3889,
+      "step": 687
+    },
+    {
+      "epoch": 1.1008,
+      "grad_norm": 0.7518171526695501,
+      "learning_rate": 7.992151437387999e-06,
+      "loss": 1.4044,
+      "step": 688
+    },
+    {
+      "epoch": 1.1024,
+      "grad_norm": 0.7624230806802528,
+      "learning_rate": 7.984686368650907e-06,
+      "loss": 1.3371,
+      "step": 689
+    },
+    {
+      "epoch": 1.104,
+      "grad_norm": 0.7236521573251403,
+      "learning_rate": 7.977210949251102e-06,
+      "loss": 1.3154,
+      "step": 690
+    },
+    {
+      "epoch": 1.1056,
+      "grad_norm": 0.7357364142064718,
+      "learning_rate": 7.969725205112766e-06,
+      "loss": 1.3821,
+      "step": 691
+    },
+    {
+      "epoch": 1.1072,
+      "grad_norm": 0.7603695919029521,
+      "learning_rate": 7.962229162195882e-06,
+      "loss": 1.3634,
+      "step": 692
+    },
+    {
+      "epoch": 1.1088,
+      "grad_norm": 0.7120471373762784,
+      "learning_rate": 7.95472284649615e-06,
+      "loss": 1.3448,
+      "step": 693
+    },
+    {
+      "epoch": 1.1104,
+      "grad_norm": 0.9010198182606531,
+      "learning_rate": 7.947206284044896e-06,
+      "loss": 1.3321,
+      "step": 694
+    },
+    {
+      "epoch": 1.112,
+      "grad_norm": 0.7094998599511237,
+      "learning_rate": 7.939679500908982e-06,
+      "loss": 1.3535,
+      "step": 695
+    },
+    {
+      "epoch": 1.1136,
+      "grad_norm": 0.7861916286185477,
+      "learning_rate": 7.932142523190711e-06,
+      "loss": 1.3264,
+      "step": 696
+    },
+    {
+      "epoch": 1.1152,
+      "grad_norm": 0.8088077238492102,
+      "learning_rate": 7.924595377027741e-06,
+      "loss": 1.3195,
+      "step": 697
+    },
+    {
+      "epoch": 1.1168,
+      "grad_norm": 0.722503169380539,
+      "learning_rate": 7.917038088592997e-06,
+      "loss": 1.3653,
+      "step": 698
+    },
+    {
+      "epoch": 1.1184,
+      "grad_norm": 0.8996799292971065,
+      "learning_rate": 7.90947068409457e-06,
+      "loss": 1.4148,
+      "step": 699
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.7255935049433854,
+      "learning_rate": 7.90189318977564e-06,
+      "loss": 1.3631,
+      "step": 700
+    },
+    {
+      "epoch": 1.1216,
+      "grad_norm": 0.8670647859300297,
+      "learning_rate": 7.894305631914373e-06,
+      "loss": 1.3297,
+      "step": 701
+    },
+    {
+      "epoch": 1.1232,
+      "grad_norm": 0.8051339707672398,
+      "learning_rate": 7.886708036823838e-06,
+      "loss": 1.3732,
+      "step": 702
+    },
+    {
+      "epoch": 1.1248,
+      "grad_norm": 0.7837140887874579,
+      "learning_rate": 7.879100430851907e-06,
+      "loss": 1.3202,
+      "step": 703
+    },
+    {
+      "epoch": 1.1264,
+      "grad_norm": 0.8645754913641339,
+      "learning_rate": 7.871482840381174e-06,
+      "loss": 1.3222,
+      "step": 704
+    },
+    {
+      "epoch": 1.1280000000000001,
+      "grad_norm": 0.7053511027412408,
+      "learning_rate": 7.863855291828857e-06,
+      "loss": 1.3488,
+      "step": 705
+    },
+    {
+      "epoch": 1.1296,
+      "grad_norm": 0.8539612906227079,
+      "learning_rate": 7.856217811646707e-06,
+      "loss": 1.3923,
+      "step": 706
+    },
+    {
+      "epoch": 1.1312,
+      "grad_norm": 0.7701566229930726,
+      "learning_rate": 7.848570426320918e-06,
+      "loss": 1.3155,
+      "step": 707
+    },
+    {
+      "epoch": 1.1328,
+      "grad_norm": 0.764173939470731,
+      "learning_rate": 7.840913162372032e-06,
+      "loss": 1.3502,
+      "step": 708
+    },
+    {
+      "epoch": 1.1344,
+      "grad_norm": 3.839058204496018,
+      "learning_rate": 7.833246046354856e-06,
+      "loss": 1.3617,
+      "step": 709
+    },
+    {
+      "epoch": 1.1360000000000001,
+      "grad_norm": 1.060708335537188,
+      "learning_rate": 7.825569104858353e-06,
+      "loss": 1.3275,
+      "step": 710
+    },
+    {
+      "epoch": 1.1376,
+      "grad_norm": 0.704270204332644,
+      "learning_rate": 7.81788236450557e-06,
+      "loss": 1.384,
+      "step": 711
+    },
+    {
+      "epoch": 1.1392,
+      "grad_norm": 0.8332556759446079,
+      "learning_rate": 7.810185851953529e-06,
+      "loss": 1.394,
+      "step": 712
+    },
+    {
+      "epoch": 1.1408,
+      "grad_norm": 0.8564429803496905,
+      "learning_rate": 7.802479593893142e-06,
+      "loss": 1.436,
+      "step": 713
+    },
+    {
+      "epoch": 1.1424,
+      "grad_norm": 0.7259708402969878,
+      "learning_rate": 7.794763617049124e-06,
+      "loss": 1.3286,
+      "step": 714
+    },
+    {
+      "epoch": 1.144,
+      "grad_norm": 0.8212173748174225,
+      "learning_rate": 7.787037948179884e-06,
+      "loss": 1.3311,
+      "step": 715
+    },
+    {
+      "epoch": 1.1456,
+      "grad_norm": 0.8627077751259059,
+      "learning_rate": 7.779302614077449e-06,
+      "loss": 1.3776,
+      "step": 716
+    },
+    {
+      "epoch": 1.1472,
+      "grad_norm": 0.7296982871098114,
+      "learning_rate": 7.771557641567363e-06,
+      "loss": 1.3552,
+      "step": 717
+    },
+    {
+      "epoch": 1.1488,
+      "grad_norm": 0.896911970689186,
+      "learning_rate": 7.763803057508594e-06,
+      "loss": 1.2956,
+      "step": 718
+    },
+    {
+      "epoch": 1.1504,
+      "grad_norm": 0.7297346310776905,
+      "learning_rate": 7.756038888793446e-06,
+      "loss": 1.4314,
+      "step": 719
+    },
+    {
+      "epoch": 1.152,
+      "grad_norm": 0.806699697495796,
+      "learning_rate": 7.748265162347455e-06,
+      "loss": 1.4013,
+      "step": 720
+    },
+    {
+      "epoch": 1.1536,
+      "grad_norm": 0.8168290700509154,
+      "learning_rate": 7.740481905129307e-06,
+      "loss": 1.3599,
+      "step": 721
+    },
+    {
+      "epoch": 1.1552,
+      "grad_norm": 0.7672552378369193,
+      "learning_rate": 7.732689144130741e-06,
+      "loss": 1.3612,
+      "step": 722
+    },
+    {
+      "epoch": 1.1568,
+      "grad_norm": 0.7905631075794937,
+      "learning_rate": 7.724886906376451e-06,
+      "loss": 1.3347,
+      "step": 723
+    },
+    {
+      "epoch": 1.1584,
+      "grad_norm": 4.433399430205663,
+      "learning_rate": 7.717075218923998e-06,
+      "loss": 1.3354,
+      "step": 724
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 1.0790501227639715,
+      "learning_rate": 7.709254108863714e-06,
+      "loss": 1.4111,
+      "step": 725
+    },
+    {
+      "epoch": 1.1616,
+      "grad_norm": 0.8572386554446184,
+      "learning_rate": 7.701423603318605e-06,
+      "loss": 1.3575,
+      "step": 726
+    },
+    {
+      "epoch": 1.1632,
+      "grad_norm": 0.816499204623045,
+      "learning_rate": 7.693583729444263e-06,
+      "loss": 1.3145,
+      "step": 727
+    },
+    {
+      "epoch": 1.1648,
+      "grad_norm": 0.94430977579828,
+      "learning_rate": 7.685734514428767e-06,
+      "loss": 1.3583,
+      "step": 728
+    },
+    {
+      "epoch": 1.1663999999999999,
+      "grad_norm": 0.7818703885546086,
+      "learning_rate": 7.677875985492591e-06,
+      "loss": 1.2702,
+      "step": 729
+    },
+    {
+      "epoch": 1.168,
+      "grad_norm": 0.8962891180451786,
+      "learning_rate": 7.67000816988851e-06,
+      "loss": 1.334,
+      "step": 730
+    },
+    {
+      "epoch": 1.1696,
+      "grad_norm": 0.8003963489050018,
+      "learning_rate": 7.662131094901499e-06,
+      "loss": 1.3333,
+      "step": 731
+    },
+    {
+      "epoch": 1.1712,
+      "grad_norm": 0.7274294763545434,
+      "learning_rate": 7.654244787848655e-06,
+      "loss": 1.3103,
+      "step": 732
+    },
+    {
+      "epoch": 1.1728,
+      "grad_norm": 0.7839565338890555,
+      "learning_rate": 7.646349276079079e-06,
+      "loss": 1.2974,
+      "step": 733
+    },
+    {
+      "epoch": 1.1743999999999999,
+      "grad_norm": 0.7278577525399996,
+      "learning_rate": 7.6384445869738e-06,
+      "loss": 1.2949,
+      "step": 734
+    },
+    {
+      "epoch": 1.176,
+      "grad_norm": 0.7777315701890748,
+      "learning_rate": 7.630530747945672e-06,
+      "loss": 1.3727,
+      "step": 735
+    },
+    {
+      "epoch": 1.1776,
+      "grad_norm": 0.751017376541754,
+      "learning_rate": 7.622607786439279e-06,
+      "loss": 1.3629,
+      "step": 736
+    },
+    {
+      "epoch": 1.1792,
+      "grad_norm": 0.7116622877129397,
+      "learning_rate": 7.6146757299308406e-06,
+      "loss": 1.355,
+      "step": 737
+    },
+    {
+      "epoch": 1.1808,
+      "grad_norm": 0.7124041729642461,
+      "learning_rate": 7.606734605928123e-06,
+      "loss": 1.3624,
+      "step": 738
+    },
+    {
+      "epoch": 1.1824,
+      "grad_norm": 0.7319762609507647,
+      "learning_rate": 7.598784441970329e-06,
+      "loss": 1.3583,
+      "step": 739
+    },
+    {
+      "epoch": 1.184,
+      "grad_norm": 0.7541702544780959,
+      "learning_rate": 7.590825265628019e-06,
+      "loss": 1.3499,
+      "step": 740
+    },
+    {
+      "epoch": 1.1856,
+      "grad_norm": 0.7448038157300539,
+      "learning_rate": 7.5828571045030005e-06,
+      "loss": 1.2992,
+      "step": 741
+    },
+    {
+      "epoch": 1.1872,
+      "grad_norm": 0.7853348571167494,
+      "learning_rate": 7.574879986228245e-06,
+      "loss": 1.2577,
+      "step": 742
+    },
+    {
+      "epoch": 1.1888,
+      "grad_norm": 0.7165061526251901,
+      "learning_rate": 7.566893938467788e-06,
+      "loss": 1.359,
+      "step": 743
+    },
+    {
+      "epoch": 1.1904,
+      "grad_norm": 0.820829901467796,
+      "learning_rate": 7.558898988916624e-06,
+      "loss": 1.2942,
+      "step": 744
+    },
+    {
+      "epoch": 1.192,
+      "grad_norm": 0.7185009909981739,
+      "learning_rate": 7.550895165300626e-06,
+      "loss": 1.3718,
+      "step": 745
+    },
+    {
+      "epoch": 1.1936,
+      "grad_norm": 0.8231342500310124,
+      "learning_rate": 7.542882495376437e-06,
+      "loss": 1.3428,
+      "step": 746
+    },
+    {
+      "epoch": 1.1952,
+      "grad_norm": 0.7575261900009206,
+      "learning_rate": 7.5348610069313795e-06,
+      "loss": 1.3012,
+      "step": 747
+    },
+    {
+      "epoch": 1.1968,
+      "grad_norm": 0.7522949635337716,
+      "learning_rate": 7.5268307277833605e-06,
+      "loss": 1.3694,
+      "step": 748
+    },
+    {
+      "epoch": 1.1984,
+      "grad_norm": 0.7208573828735061,
+      "learning_rate": 7.518791685780769e-06,
+      "loss": 1.3222,
+      "step": 749
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.707584389682235,
+      "learning_rate": 7.5107439088023845e-06,
+      "loss": 1.3656,
+      "step": 750
+    },
+    {
+      "epoch": 1.2016,
+      "grad_norm": 0.7072800497016226,
+      "learning_rate": 7.502687424757278e-06,
+      "loss": 1.3168,
+      "step": 751
+    },
+    {
+      "epoch": 1.2032,
+      "grad_norm": 0.7459474061447671,
+      "learning_rate": 7.4946222615847165e-06,
+      "loss": 1.3393,
+      "step": 752
+    },
+    {
+      "epoch": 1.2048,
+      "grad_norm": 0.6791104234917247,
+      "learning_rate": 7.486548447254065e-06,
+      "loss": 1.2976,
+      "step": 753
+    },
+    {
+      "epoch": 1.2064,
+      "grad_norm": 0.732751228713868,
+      "learning_rate": 7.478466009764692e-06,
+      "loss": 1.332,
+      "step": 754
+    },
+    {
+      "epoch": 1.208,
+      "grad_norm": 0.7157404411753864,
+      "learning_rate": 7.470374977145867e-06,
+      "loss": 1.3068,
+      "step": 755
+    },
+    {
+      "epoch": 1.2096,
+      "grad_norm": 0.6904779851769556,
+      "learning_rate": 7.462275377456671e-06,
+      "loss": 1.3709,
+      "step": 756
+    },
+    {
+      "epoch": 1.2112,
+      "grad_norm": 0.7319404631729015,
+      "learning_rate": 7.4541672387858895e-06,
+      "loss": 1.3605,
+      "step": 757
+    },
+    {
+      "epoch": 1.2128,
+      "grad_norm": 0.7276335779355101,
+      "learning_rate": 7.446050589251928e-06,
+      "loss": 1.2614,
+      "step": 758
+    },
+    {
+      "epoch": 1.2144,
+      "grad_norm": 0.7915409202872087,
+      "learning_rate": 7.437925457002697e-06,
+      "loss": 1.4232,
+      "step": 759
+    },
+    {
+      "epoch": 1.216,
+      "grad_norm": 0.7308729767975439,
+      "learning_rate": 7.429791870215535e-06,
+      "loss": 1.3211,
+      "step": 760
+    },
+    {
+      "epoch": 1.2176,
+      "grad_norm": 0.7735854230802189,
+      "learning_rate": 7.421649857097092e-06,
+      "loss": 1.4013,
+      "step": 761
+    },
+    {
+      "epoch": 1.2192,
+      "grad_norm": 0.7941891833380622,
+      "learning_rate": 7.413499445883245e-06,
+      "loss": 1.3376,
+      "step": 762
+    },
+    {
+      "epoch": 1.2208,
+      "grad_norm": 0.7669491957124617,
+      "learning_rate": 7.405340664838994e-06,
+      "loss": 1.4,
+      "step": 763
+    },
+    {
+      "epoch": 1.2224,
+      "grad_norm": 0.7359597501151763,
+      "learning_rate": 7.39717354225836e-06,
+      "loss": 1.3353,
+      "step": 764
+    },
+    {
+      "epoch": 1.224,
+      "grad_norm": 0.7647539179585882,
+      "learning_rate": 7.3889981064643e-06,
+      "loss": 1.3668,
+      "step": 765
+    },
+    {
+      "epoch": 1.2256,
+      "grad_norm": 0.7291962021503118,
+      "learning_rate": 7.380814385808594e-06,
+      "loss": 1.3257,
+      "step": 766
+    },
+    {
+      "epoch": 1.2272,
+      "grad_norm": 0.7551922713552083,
+      "learning_rate": 7.372622408671757e-06,
+      "loss": 1.369,
+      "step": 767
+    },
+    {
+      "epoch": 1.2288000000000001,
+      "grad_norm": 0.7740321863297959,
+      "learning_rate": 7.364422203462935e-06,
+      "loss": 1.3597,
+      "step": 768
+    },
+    {
+      "epoch": 1.2304,
+      "grad_norm": 0.718683566292131,
+      "learning_rate": 7.3562137986198065e-06,
+      "loss": 1.3399,
+      "step": 769
+    },
+    {
+      "epoch": 1.232,
+      "grad_norm": 0.8101489574389433,
+      "learning_rate": 7.3479972226084925e-06,
+      "loss": 1.3435,
+      "step": 770
+    },
+    {
+      "epoch": 1.2336,
+      "grad_norm": 0.7029635544100846,
+      "learning_rate": 7.339772503923445e-06,
+      "loss": 1.3445,
+      "step": 771
+    },
+    {
+      "epoch": 1.2352,
+      "grad_norm": 0.753075618355313,
+      "learning_rate": 7.331539671087353e-06,
+      "loss": 1.3281,
+      "step": 772
+    },
+    {
+      "epoch": 1.2368000000000001,
+      "grad_norm": 0.7766893298591357,
+      "learning_rate": 7.32329875265105e-06,
+      "loss": 1.3149,
+      "step": 773
+    },
+    {
+      "epoch": 1.2384,
+      "grad_norm": 0.7199746311881295,
+      "learning_rate": 7.315049777193407e-06,
+      "loss": 1.3279,
+      "step": 774
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.7200377742411487,
+      "learning_rate": 7.306792773321234e-06,
+      "loss": 1.3575,
+      "step": 775
+    },
+    {
+      "epoch": 1.2416,
+      "grad_norm": 0.7224104347665683,
+      "learning_rate": 7.298527769669188e-06,
+      "loss": 1.3792,
+      "step": 776
+    },
+    {
+      "epoch": 1.2432,
+      "grad_norm": 0.7587408608350531,
+      "learning_rate": 7.290254794899665e-06,
+      "loss": 1.322,
+      "step": 777
+    },
+    {
+      "epoch": 1.2448,
+      "grad_norm": 0.7347356641366618,
+      "learning_rate": 7.281973877702705e-06,
+      "loss": 1.3391,
+      "step": 778
+    },
+    {
+      "epoch": 1.2464,
+      "grad_norm": 0.7655500308754767,
+      "learning_rate": 7.2736850467958905e-06,
+      "loss": 1.355,
+      "step": 779
+    },
+    {
+      "epoch": 1.248,
+      "grad_norm": 0.7447908103157936,
+      "learning_rate": 7.26538833092425e-06,
+      "loss": 1.3332,
+      "step": 780
+    },
+    {
+      "epoch": 1.2496,
+      "grad_norm": 0.7345591983659606,
+      "learning_rate": 7.257083758860159e-06,
+      "loss": 1.3204,
+      "step": 781
+    },
+    {
+      "epoch": 1.2511999999999999,
+      "grad_norm": 0.7995800623880757,
+      "learning_rate": 7.248771359403231e-06,
+      "loss": 1.4102,
+      "step": 782
+    },
+    {
+      "epoch": 1.2528000000000001,
+      "grad_norm": 0.8020542837835895,
+      "learning_rate": 7.240451161380226e-06,
+      "loss": 1.254,
+      "step": 783
+    },
+    {
+      "epoch": 1.2544,
+      "grad_norm": 0.7270111444552266,
+      "learning_rate": 7.232123193644957e-06,
+      "loss": 1.3354,
+      "step": 784
+    },
+    {
+      "epoch": 1.256,
+      "grad_norm": 0.7533596074354689,
+      "learning_rate": 7.22378748507817e-06,
+      "loss": 1.3813,
+      "step": 785
+    },
+    {
+      "epoch": 1.2576,
+      "grad_norm": 0.7330121616511528,
+      "learning_rate": 7.215444064587462e-06,
+      "loss": 1.382,
+      "step": 786
+    },
+    {
+      "epoch": 1.2591999999999999,
+      "grad_norm": 0.8032146664515214,
+      "learning_rate": 7.207092961107176e-06,
+      "loss": 1.2979,
+      "step": 787
+    },
+    {
+      "epoch": 1.2608,
+      "grad_norm": 0.7561254917030892,
+      "learning_rate": 7.198734203598294e-06,
+      "loss": 1.3789,
+      "step": 788
+    },
+    {
+      "epoch": 1.2624,
+      "grad_norm": 0.7959324722696368,
+      "learning_rate": 7.190367821048346e-06,
+      "loss": 1.384,
+      "step": 789
+    },
+    {
+      "epoch": 1.264,
+      "grad_norm": 0.7020868905071856,
+      "learning_rate": 7.181993842471301e-06,
+      "loss": 1.3199,
+      "step": 790
+    },
+    {
+      "epoch": 1.2656,
+      "grad_norm": 0.772575737188667,
+      "learning_rate": 7.173612296907473e-06,
+      "loss": 1.3386,
+      "step": 791
+    },
+    {
+      "epoch": 1.2671999999999999,
+      "grad_norm": 0.6636642847616839,
+      "learning_rate": 7.165223213423416e-06,
+      "loss": 1.361,
+      "step": 792
+    },
+    {
+      "epoch": 1.2688,
+      "grad_norm": 0.7951682928716968,
+      "learning_rate": 7.15682662111183e-06,
+      "loss": 1.3785,
+      "step": 793
+    },
+    {
+      "epoch": 1.2704,
+      "grad_norm": 0.6935703595646292,
+      "learning_rate": 7.148422549091447e-06,
+      "loss": 1.4312,
+      "step": 794
+    },
+    {
+      "epoch": 1.272,
+      "grad_norm": 0.7733330936529429,
+      "learning_rate": 7.140011026506945e-06,
+      "loss": 1.2827,
+      "step": 795
+    },
+    {
+      "epoch": 1.2736,
+      "grad_norm": 0.7319860039091094,
+      "learning_rate": 7.131592082528837e-06,
+      "loss": 1.3149,
+      "step": 796
+    },
+    {
+      "epoch": 1.2752,
+      "grad_norm": 0.7470810172238302,
+      "learning_rate": 7.12316574635337e-06,
+      "loss": 1.3698,
+      "step": 797
+    },
+    {
+      "epoch": 1.2768,
+      "grad_norm": 0.6984732630891479,
+      "learning_rate": 7.114732047202433e-06,
+      "loss": 1.2664,
+      "step": 798
+    },
+    {
+      "epoch": 1.2784,
+      "grad_norm": 0.6912226189894565,
+      "learning_rate": 7.106291014323445e-06,
+      "loss": 1.2684,
+      "step": 799
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.7152386668298167,
+      "learning_rate": 7.0978426769892585e-06,
+      "loss": 1.2924,
+      "step": 800
+    },
+    {
+      "epoch": 1.2816,
+      "grad_norm": 0.67245219413985,
+      "learning_rate": 7.089387064498057e-06,
+      "loss": 1.3255,
+      "step": 801
+    },
+    {
+      "epoch": 1.2832,
+      "grad_norm": 0.6981130588897566,
+      "learning_rate": 7.080924206173253e-06,
+      "loss": 1.3211,
+      "step": 802
+    },
+    {
+      "epoch": 1.2848,
+      "grad_norm": 0.6996847117802274,
+      "learning_rate": 7.072454131363391e-06,
+      "loss": 1.3502,
+      "step": 803
+    },
+    {
+      "epoch": 1.2864,
+      "grad_norm": 0.6787885738821787,
+      "learning_rate": 7.063976869442037e-06,
+      "loss": 1.3418,
+      "step": 804
+    },
+    {
+      "epoch": 1.288,
+      "grad_norm": 0.6965533922980114,
+      "learning_rate": 7.055492449807684e-06,
+      "loss": 1.2715,
+      "step": 805
+    },
+    {
+      "epoch": 1.2896,
+      "grad_norm": 0.6683423793972584,
+      "learning_rate": 7.047000901883646e-06,
+      "loss": 1.412,
+      "step": 806
+    },
+    {
+      "epoch": 1.2912,
+      "grad_norm": 0.7104685344288328,
+      "learning_rate": 7.038502255117957e-06,
+      "loss": 1.408,
+      "step": 807
+    },
+    {
+      "epoch": 1.2928,
+      "grad_norm": 0.6662217061996702,
+      "learning_rate": 7.029996538983273e-06,
+      "loss": 1.2499,
+      "step": 808
+    },
+    {
+      "epoch": 1.2944,
+      "grad_norm": 0.7327212754942412,
+      "learning_rate": 7.021483782976759e-06,
+      "loss": 1.4001,
+      "step": 809
+    },
+    {
+      "epoch": 1.296,
+      "grad_norm": 0.7136037541795051,
+      "learning_rate": 7.012964016620002e-06,
+      "loss": 1.3419,
+      "step": 810
+    },
+    {
+      "epoch": 1.2976,
+      "grad_norm": 0.7366117239499376,
+      "learning_rate": 7.004437269458894e-06,
+      "loss": 1.4046,
+      "step": 811
+    },
+    {
+      "epoch": 1.2992,
+      "grad_norm": 0.7338053105626233,
+      "learning_rate": 6.995903571063541e-06,
+      "loss": 1.35,
+      "step": 812
+    },
+    {
+      "epoch": 1.3008,
+      "grad_norm": 0.706065395114804,
+      "learning_rate": 6.987362951028147e-06,
+      "loss": 1.3304,
+      "step": 813
+    },
+    {
+      "epoch": 1.3024,
+      "grad_norm": 0.6670624704813073,
+      "learning_rate": 6.97881543897093e-06,
+      "loss": 1.3338,
+      "step": 814
+    },
+    {
+      "epoch": 1.304,
+      "grad_norm": 0.6680614225263798,
+      "learning_rate": 6.970261064534003e-06,
+      "loss": 1.3065,
+      "step": 815
+    },
+    {
+      "epoch": 1.3056,
+      "grad_norm": 0.6826823454622892,
+      "learning_rate": 6.961699857383279e-06,
+      "loss": 1.3179,
+      "step": 816
+    },
+    {
+      "epoch": 1.3072,
+      "grad_norm": 0.7520654915531269,
+      "learning_rate": 6.953131847208365e-06,
+      "loss": 1.356,
+      "step": 817
+    },
+    {
+      "epoch": 1.3088,
+      "grad_norm": 0.7386405066565618,
+      "learning_rate": 6.944557063722459e-06,
+      "loss": 1.3672,
+      "step": 818
+    },
+    {
+      "epoch": 1.3104,
+      "grad_norm": 0.6377238841149768,
+      "learning_rate": 6.935975536662254e-06,
+      "loss": 1.2824,
+      "step": 819
+    },
+    {
+      "epoch": 1.312,
+      "grad_norm": 0.7171487278365545,
+      "learning_rate": 6.9273872957878255e-06,
+      "loss": 1.3003,
+      "step": 820
+    },
+    {
+      "epoch": 1.3136,
+      "grad_norm": 0.6667054066480461,
+      "learning_rate": 6.91879237088253e-06,
+      "loss": 1.3022,
+      "step": 821
+    },
+    {
+      "epoch": 1.3152,
+      "grad_norm": 0.6822790945605449,
+      "learning_rate": 6.910190791752907e-06,
+      "loss": 1.2339,
+      "step": 822
+    },
+    {
+      "epoch": 1.3168,
+      "grad_norm": 0.7241564127095226,
+      "learning_rate": 6.90158258822857e-06,
+      "loss": 1.3195,
+      "step": 823
+    },
+    {
+      "epoch": 1.3184,
+      "grad_norm": 0.6942814014365648,
+      "learning_rate": 6.892967790162109e-06,
+      "loss": 1.3482,
+      "step": 824
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.6623929950986044,
+      "learning_rate": 6.884346427428978e-06,
+      "loss": 1.339,
+      "step": 825
+    },
+    {
+      "epoch": 1.3216,
+      "grad_norm": 0.6994192083583599,
+      "learning_rate": 6.875718529927404e-06,
+      "loss": 1.3578,
+      "step": 826
+    },
+    {
+      "epoch": 1.3232,
+      "grad_norm": 0.6924050892291411,
+      "learning_rate": 6.867084127578267e-06,
+      "loss": 1.3195,
+      "step": 827
+    },
+    {
+      "epoch": 1.3248,
+      "grad_norm": 0.7205500382558482,
+      "learning_rate": 6.858443250325013e-06,
+      "loss": 1.3336,
+      "step": 828
+    },
+    {
+      "epoch": 1.3264,
+      "grad_norm": 0.6717992009416129,
+      "learning_rate": 6.849795928133538e-06,
+      "loss": 1.3797,
+      "step": 829
+    },
+    {
+      "epoch": 1.328,
+      "grad_norm": 0.7238080633946886,
+      "learning_rate": 6.841142190992092e-06,
+      "loss": 1.3527,
+      "step": 830
+    },
+    {
+      "epoch": 1.3296000000000001,
+      "grad_norm": 0.6954200960558923,
+      "learning_rate": 6.832482068911167e-06,
+      "loss": 1.259,
+      "step": 831
+    },
+    {
+      "epoch": 1.3312,
+      "grad_norm": 0.6882879989973665,
+      "learning_rate": 6.823815591923402e-06,
+      "loss": 1.3183,
+      "step": 832
+    },
+    {
+      "epoch": 1.3328,
+      "grad_norm": 0.7300188016561705,
+      "learning_rate": 6.815142790083473e-06,
+      "loss": 1.3112,
+      "step": 833
+    },
+    {
+      "epoch": 1.3344,
+      "grad_norm": 0.7275904240471627,
+      "learning_rate": 6.8064636934679885e-06,
+      "loss": 1.32,
+      "step": 834
+    },
+    {
+      "epoch": 1.336,
+      "grad_norm": 0.6357231171846449,
+      "learning_rate": 6.797778332175387e-06,
+      "loss": 1.2817,
+      "step": 835
+    },
+    {
+      "epoch": 1.3376000000000001,
+      "grad_norm": 0.7238522629666839,
+      "learning_rate": 6.789086736325834e-06,
+      "loss": 1.2725,
+      "step": 836
+    },
+    {
+      "epoch": 1.3392,
+      "grad_norm": 0.6891608861323208,
+      "learning_rate": 6.780388936061118e-06,
+      "loss": 1.284,
+      "step": 837
+    },
+    {
+      "epoch": 1.3408,
+      "grad_norm": 0.7171161016867675,
+      "learning_rate": 6.771684961544537e-06,
+      "loss": 1.2949,
+      "step": 838
+    },
+    {
+      "epoch": 1.3424,
+      "grad_norm": 0.7408998698740566,
+      "learning_rate": 6.7629748429608076e-06,
+      "loss": 1.3239,
+      "step": 839
+    },
+    {
+      "epoch": 1.3439999999999999,
+      "grad_norm": 0.737720815896351,
+      "learning_rate": 6.754258610515949e-06,
+      "loss": 1.2856,
+      "step": 840
+    },
+    {
+      "epoch": 1.3456000000000001,
+      "grad_norm": 0.715134172718235,
+      "learning_rate": 6.745536294437187e-06,
+      "loss": 1.386,
+      "step": 841
+    },
+    {
+      "epoch": 1.3472,
+      "grad_norm": 0.7268565656956834,
+      "learning_rate": 6.736807924972841e-06,
+      "loss": 1.2559,
+      "step": 842
+    },
+    {
+      "epoch": 1.3488,
+      "grad_norm": 0.6823895436049205,
+      "learning_rate": 6.728073532392226e-06,
+      "loss": 1.3547,
+      "step": 843
+    },
+    {
+      "epoch": 1.3504,
+      "grad_norm": 0.7613129547031245,
+      "learning_rate": 6.719333146985544e-06,
+      "loss": 1.3278,
+      "step": 844
+    },
+    {
+      "epoch": 1.3519999999999999,
+      "grad_norm": 0.7356898395807632,
+      "learning_rate": 6.710586799063777e-06,
+      "loss": 1.3172,
+      "step": 845
+    },
+    {
+      "epoch": 1.3536000000000001,
+      "grad_norm": 0.7317108859931903,
+      "learning_rate": 6.701834518958587e-06,
+      "loss": 1.3382,
+      "step": 846
+    },
+    {
+      "epoch": 1.3552,
+      "grad_norm": 0.7519561439867526,
+      "learning_rate": 6.6930763370222104e-06,
+      "loss": 1.3007,
+      "step": 847
+    },
+    {
+      "epoch": 1.3568,
+      "grad_norm": 0.7007561159805317,
+      "learning_rate": 6.684312283627348e-06,
+      "loss": 1.3871,
+      "step": 848
+    },
+    {
+      "epoch": 1.3584,
+      "grad_norm": 0.7121592446390924,
+      "learning_rate": 6.6755423891670605e-06,
+      "loss": 1.2893,
+      "step": 849
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 0.6678880922506275,
+      "learning_rate": 6.6667666840546685e-06,
+      "loss": 1.3611,
+      "step": 850
+    },
+    {
+      "epoch": 1.3616,
+      "grad_norm": 0.7046350986963248,
+      "learning_rate": 6.6579851987236435e-06,
+      "loss": 1.3383,
+      "step": 851
+    },
+    {
+      "epoch": 1.3632,
+      "grad_norm": 0.7427388955580942,
+      "learning_rate": 6.649197963627497e-06,
+      "loss": 1.3111,
+      "step": 852
+    },
+    {
+      "epoch": 1.3648,
+      "grad_norm": 0.6860328314354844,
+      "learning_rate": 6.640405009239689e-06,
+      "loss": 1.3946,
+      "step": 853
+    },
+    {
+      "epoch": 1.3664,
+      "grad_norm": 0.6907593638088485,
+      "learning_rate": 6.631606366053507e-06,
+      "loss": 1.3099,
+      "step": 854
+    },
+    {
+      "epoch": 1.3679999999999999,
+      "grad_norm": 0.6922101701270538,
+      "learning_rate": 6.622802064581968e-06,
+      "loss": 1.3789,
+      "step": 855
+    },
+    {
+      "epoch": 1.3696,
+      "grad_norm": 0.7256423674408752,
+      "learning_rate": 6.613992135357713e-06,
+      "loss": 1.3528,
+      "step": 856
+    },
+    {
+      "epoch": 1.3712,
+      "grad_norm": 0.708063312974822,
+      "learning_rate": 6.605176608932897e-06,
+      "loss": 1.2881,
+      "step": 857
+    },
+    {
+      "epoch": 1.3728,
+      "grad_norm": 0.7475519018999066,
+      "learning_rate": 6.596355515879091e-06,
+      "loss": 1.361,
+      "step": 858
+    },
+    {
+      "epoch": 1.3744,
+      "grad_norm": 0.7019365339748909,
+      "learning_rate": 6.587528886787165e-06,
+      "loss": 1.3212,
+      "step": 859
+    },
+    {
+      "epoch": 1.376,
+      "grad_norm": 0.6947358410261755,
+      "learning_rate": 6.578696752267189e-06,
+      "loss": 1.3285,
+      "step": 860
+    },
+    {
+      "epoch": 1.3776,
+      "grad_norm": 0.661458016108248,
+      "learning_rate": 6.5698591429483286e-06,
+      "loss": 1.3216,
+      "step": 861
+    },
+    {
+      "epoch": 1.3792,
+      "grad_norm": 0.7392992803092605,
+      "learning_rate": 6.5610160894787275e-06,
+      "loss": 1.2877,
+      "step": 862
+    },
+    {
+      "epoch": 1.3808,
+      "grad_norm": 0.6627027997732061,
+      "learning_rate": 6.552167622525421e-06,
+      "loss": 1.3402,
+      "step": 863
+    },
+    {
+      "epoch": 1.3824,
+      "grad_norm": 0.8065479994972355,
+      "learning_rate": 6.543313772774209e-06,
+      "loss": 1.3118,
+      "step": 864
+    },
+    {
+      "epoch": 1.384,
+      "grad_norm": 0.6511836583027392,
+      "learning_rate": 6.534454570929563e-06,
+      "loss": 1.3449,
+      "step": 865
+    },
+    {
+      "epoch": 1.3856,
+      "grad_norm": 0.7796970526049094,
+      "learning_rate": 6.52559004771451e-06,
+      "loss": 1.3459,
+      "step": 866
+    },
+    {
+      "epoch": 1.3872,
+      "grad_norm": 0.669997525854838,
+      "learning_rate": 6.516720233870538e-06,
+      "loss": 1.3407,
+      "step": 867
+    },
+    {
+      "epoch": 1.3888,
+      "grad_norm": 0.701341522774775,
+      "learning_rate": 6.507845160157476e-06,
+      "loss": 1.3904,
+      "step": 868
+    },
+    {
+      "epoch": 1.3904,
+      "grad_norm": 0.6822376037518737,
+      "learning_rate": 6.498964857353401e-06,
+      "loss": 1.3525,
+      "step": 869
+    },
+    {
+      "epoch": 1.392,
+      "grad_norm": 0.729026029452657,
+      "learning_rate": 6.4900793562545165e-06,
+      "loss": 1.3612,
+      "step": 870
+    },
+    {
+      "epoch": 1.3936,
+      "grad_norm": 0.638522137710986,
+      "learning_rate": 6.481188687675057e-06,
+      "loss": 1.2948,
+      "step": 871
+    },
+    {
+      "epoch": 1.3952,
+      "grad_norm": 0.7504118361733407,
+      "learning_rate": 6.47229288244718e-06,
+      "loss": 1.3215,
+      "step": 872
+    },
+    {
+      "epoch": 1.3968,
+      "grad_norm": 0.9036845021808154,
+      "learning_rate": 6.46339197142085e-06,
+      "loss": 1.3641,
+      "step": 873
+    },
+    {
+      "epoch": 1.3984,
+      "grad_norm": 0.755371680086072,
+      "learning_rate": 6.454485985463742e-06,
+      "loss": 1.3639,
+      "step": 874
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.691478776478371,
+      "learning_rate": 6.445574955461134e-06,
+      "loss": 1.3441,
+      "step": 875
+    },
+    {
+      "epoch": 1.4016,
+      "grad_norm": 0.7553839121451498,
+      "learning_rate": 6.436658912315789e-06,
+      "loss": 1.3754,
+      "step": 876
+    },
+    {
+      "epoch": 1.4032,
+      "grad_norm": 0.6780617107245691,
+      "learning_rate": 6.427737886947859e-06,
+      "loss": 1.3483,
+      "step": 877
+    },
+    {
+      "epoch": 1.4048,
+      "grad_norm": 0.660948745290708,
+      "learning_rate": 6.418811910294776e-06,
+      "loss": 1.3509,
+      "step": 878
+    },
+    {
+      "epoch": 1.4064,
+      "grad_norm": 0.697410370070993,
+      "learning_rate": 6.409881013311136e-06,
+      "loss": 1.3109,
+      "step": 879
+    },
+    {
+      "epoch": 1.408,
+      "grad_norm": 0.707316793645158,
+      "learning_rate": 6.400945226968607e-06,
+      "loss": 1.2951,
+      "step": 880
+    },
+    {
+      "epoch": 1.4096,
+      "grad_norm": 0.7543456734916378,
+      "learning_rate": 6.392004582255807e-06,
+      "loss": 1.3578,
+      "step": 881
+    },
+    {
+      "epoch": 1.4112,
+      "grad_norm": 0.7294553251723411,
+      "learning_rate": 6.383059110178205e-06,
+      "loss": 1.3265,
+      "step": 882
+    },
+    {
+      "epoch": 1.4128,
+      "grad_norm": 0.729887887883092,
+      "learning_rate": 6.374108841758006e-06,
+      "loss": 1.3039,
+      "step": 883
+    },
+    {
+      "epoch": 1.4144,
+      "grad_norm": 0.7789797651078585,
+      "learning_rate": 6.365153808034057e-06,
+      "loss": 1.339,
+      "step": 884
+    },
+    {
+      "epoch": 1.416,
+      "grad_norm": 0.705598929830418,
+      "learning_rate": 6.356194040061725e-06,
+      "loss": 1.3371,
+      "step": 885
+    },
+    {
+      "epoch": 1.4176,
+      "grad_norm": 0.7162746384863019,
+      "learning_rate": 6.3472295689127946e-06,
+      "loss": 1.2921,
+      "step": 886
+    },
+    {
+      "epoch": 1.4192,
+      "grad_norm": 0.6933822503950068,
+      "learning_rate": 6.338260425675365e-06,
+      "loss": 1.2447,
+      "step": 887
+    },
+    {
+      "epoch": 1.4208,
+      "grad_norm": 0.774886428622685,
+      "learning_rate": 6.329286641453729e-06,
+      "loss": 1.3525,
+      "step": 888
+    },
+    {
+      "epoch": 1.4224,
+      "grad_norm": 0.63522193386074,
+      "learning_rate": 6.320308247368285e-06,
+      "loss": 1.3336,
+      "step": 889
+    },
+    {
+      "epoch": 1.424,
+      "grad_norm": 0.8641940133565976,
+      "learning_rate": 6.311325274555413e-06,
+      "loss": 1.315,
+      "step": 890
+    },
+    {
+      "epoch": 1.4256,
+      "grad_norm": 0.6417054903128508,
+      "learning_rate": 6.302337754167369e-06,
+      "loss": 1.3683,
+      "step": 891
+    },
+    {
+      "epoch": 1.4272,
+      "grad_norm": 0.8516849338065954,
+      "learning_rate": 6.2933457173721855e-06,
+      "loss": 1.3744,
+      "step": 892
+    },
+    {
+      "epoch": 1.4288,
+      "grad_norm": 0.6731539728880818,
+      "learning_rate": 6.2843491953535515e-06,
+      "loss": 1.3468,
+      "step": 893
+    },
+    {
+      "epoch": 1.4304000000000001,
+      "grad_norm": 0.7062275161950584,
+      "learning_rate": 6.275348219310715e-06,
+      "loss": 1.3054,
+      "step": 894
+    },
+    {
+      "epoch": 1.432,
+      "grad_norm": 0.6747441769261444,
+      "learning_rate": 6.266342820458366e-06,
+      "loss": 1.3598,
+      "step": 895
+    },
+    {
+      "epoch": 1.4336,
+      "grad_norm": 0.7610102484222968,
+      "learning_rate": 6.2573330300265375e-06,
+      "loss": 1.3484,
+      "step": 896
+    },
+    {
+      "epoch": 1.4352,
+      "grad_norm": 0.74686931623709,
+      "learning_rate": 6.248318879260488e-06,
+      "loss": 1.3909,
+      "step": 897
+    },
+    {
+      "epoch": 1.4368,
+      "grad_norm": 0.6746810294799923,
+      "learning_rate": 6.239300399420601e-06,
+      "loss": 1.3109,
+      "step": 898
+    },
+    {
+      "epoch": 1.4384000000000001,
+      "grad_norm": 0.7383829925879524,
+      "learning_rate": 6.230277621782269e-06,
+      "loss": 1.3378,
+      "step": 899
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.7123873724460076,
+      "learning_rate": 6.221250577635791e-06,
+      "loss": 1.319,
+      "step": 900
+    },
+    {
+      "epoch": 1.4416,
+      "grad_norm": 0.697366729845929,
+      "learning_rate": 6.2122192982862615e-06,
+      "loss": 1.41,
+      "step": 901
+    },
+    {
+      "epoch": 1.4432,
+      "grad_norm": 0.7305412961533443,
+      "learning_rate": 6.203183815053463e-06,
+      "loss": 1.3076,
+      "step": 902
+    },
+    {
+      "epoch": 1.4447999999999999,
+      "grad_norm": 0.6855558803474534,
+      "learning_rate": 6.1941441592717564e-06,
+      "loss": 1.3279,
+      "step": 903
+    },
+    {
+      "epoch": 1.4464000000000001,
+      "grad_norm": 0.7284037988244817,
+      "learning_rate": 6.185100362289972e-06,
+      "loss": 1.3987,
+      "step": 904
+    },
+    {
+      "epoch": 1.448,
+      "grad_norm": 0.7042660983622786,
+      "learning_rate": 6.176052455471302e-06,
+      "loss": 1.3048,
+      "step": 905
+    },
+    {
+      "epoch": 1.4496,
+      "grad_norm": 0.6953725229148289,
+      "learning_rate": 6.167000470193189e-06,
+      "loss": 1.3047,
+      "step": 906
+    },
+    {
+      "epoch": 1.4512,
+      "grad_norm": 0.8079661227760134,
+      "learning_rate": 6.157944437847226e-06,
+      "loss": 1.2888,
+      "step": 907
+    },
+    {
+      "epoch": 1.4527999999999999,
+      "grad_norm": 0.7518799267794672,
+      "learning_rate": 6.148884389839035e-06,
+      "loss": 1.3394,
+      "step": 908
+    },
+    {
+      "epoch": 1.4544000000000001,
+      "grad_norm": 0.710486654915963,
+      "learning_rate": 6.1398203575881645e-06,
+      "loss": 1.2938,
+      "step": 909
+    },
+    {
+      "epoch": 1.456,
+      "grad_norm": 0.7626175839309549,
+      "learning_rate": 6.130752372527981e-06,
+      "loss": 1.3373,
+      "step": 910
+    },
+    {
+      "epoch": 1.4576,
+      "grad_norm": 0.6844134428830544,
+      "learning_rate": 6.121680466105559e-06,
+      "loss": 1.319,
+      "step": 911
+    },
+    {
+      "epoch": 1.4592,
+      "grad_norm": 0.771399391857818,
+      "learning_rate": 6.112604669781572e-06,
+      "loss": 1.3298,
+      "step": 912
+    },
+    {
+      "epoch": 1.4607999999999999,
+      "grad_norm": 0.7434081159590018,
+      "learning_rate": 6.1035250150301864e-06,
+      "loss": 1.3031,
+      "step": 913
+    },
+    {
+      "epoch": 1.4624,
+      "grad_norm": 0.7803970705829358,
+      "learning_rate": 6.0944415333389405e-06,
+      "loss": 1.4003,
+      "step": 914
+    },
+    {
+      "epoch": 1.464,
+      "grad_norm": 0.7137997206430831,
+      "learning_rate": 6.085354256208655e-06,
+      "loss": 1.3051,
+      "step": 915
+    },
+    {
+      "epoch": 1.4656,
+      "grad_norm": 0.7793612255344321,
+      "learning_rate": 6.076263215153308e-06,
+      "loss": 1.3458,
+      "step": 916
+    },
+    {
+      "epoch": 1.4672,
+      "grad_norm": 0.6919866659570898,
+      "learning_rate": 6.067168441699927e-06,
+      "loss": 1.2904,
+      "step": 917
+    },
+    {
+      "epoch": 1.4687999999999999,
+      "grad_norm": 0.7392059059890401,
+      "learning_rate": 6.058069967388489e-06,
+      "loss": 1.2607,
+      "step": 918
+    },
+    {
+      "epoch": 1.4704,
+      "grad_norm": 0.7545942547246565,
+      "learning_rate": 6.048967823771802e-06,
+      "loss": 1.2837,
+      "step": 919
+    },
+    {
+      "epoch": 1.472,
+      "grad_norm": 0.6679767943543515,
+      "learning_rate": 6.039862042415401e-06,
+      "loss": 1.3596,
+      "step": 920
+    },
+    {
+      "epoch": 1.4736,
+      "grad_norm": 0.8164664716288693,
+      "learning_rate": 6.030752654897435e-06,
+      "loss": 1.3923,
+      "step": 921
+    },
+    {
+      "epoch": 1.4752,
+      "grad_norm": 0.738079629002926,
+      "learning_rate": 6.021639692808558e-06,
+      "loss": 1.2482,
+      "step": 922
+    },
+    {
+      "epoch": 1.4768,
+      "grad_norm": 0.6848545061130017,
+      "learning_rate": 6.0125231877518205e-06,
+      "loss": 1.2998,
+      "step": 923
+    },
+    {
+      "epoch": 1.4784,
+      "grad_norm": 0.7909017607726269,
+      "learning_rate": 6.0034031713425636e-06,
+      "loss": 1.3453,
+      "step": 924
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.688639589177177,
+      "learning_rate": 5.994279675208302e-06,
+      "loss": 1.3245,
+      "step": 925
+    },
+    {
+      "epoch": 1.4816,
+      "grad_norm": 0.7420015940666387,
+      "learning_rate": 5.985152730988617e-06,
+      "loss": 1.3447,
+      "step": 926
+    },
+    {
+      "epoch": 1.4832,
+      "grad_norm": 0.7871372626059635,
+      "learning_rate": 5.9760223703350495e-06,
+      "loss": 1.3368,
+      "step": 927
+    },
+    {
+      "epoch": 1.4848,
+      "grad_norm": 0.7107145720911654,
+      "learning_rate": 5.966888624910989e-06,
+      "loss": 1.2863,
+      "step": 928
+    },
+    {
+      "epoch": 1.4864,
+      "grad_norm": 0.674410963118814,
+      "learning_rate": 5.957751526391558e-06,
+      "loss": 1.2441,
+      "step": 929
+    },
+    {
+      "epoch": 1.488,
+      "grad_norm": 0.7018708093370339,
+      "learning_rate": 5.948611106463518e-06,
+      "loss": 1.2568,
+      "step": 930
+    },
+    {
+      "epoch": 1.4896,
+      "grad_norm": 0.7029381592947972,
+      "learning_rate": 5.939467396825137e-06,
+      "loss": 1.3326,
+      "step": 931
+    },
+    {
+      "epoch": 1.4912,
+      "grad_norm": 0.6799277071218487,
+      "learning_rate": 5.9303204291860975e-06,
+      "loss": 1.3048,
+      "step": 932
+    },
+    {
+      "epoch": 1.4928,
+      "grad_norm": 0.7990568152343718,
+      "learning_rate": 5.92117023526738e-06,
+      "loss": 1.3246,
+      "step": 933
+    },
+    {
+      "epoch": 1.4944,
+      "grad_norm": 0.6804338171743042,
+      "learning_rate": 5.912016846801153e-06,
+      "loss": 1.3694,
+      "step": 934
+    },
+    {
+      "epoch": 1.496,
+      "grad_norm": 0.7688680357014016,
+      "learning_rate": 5.902860295530665e-06,
+      "loss": 1.2898,
+      "step": 935
+    },
+    {
+      "epoch": 1.4976,
+      "grad_norm": 0.6575074273703849,
+      "learning_rate": 5.893700613210128e-06,
+      "loss": 1.283,
+      "step": 936
+    },
+    {
+      "epoch": 1.4992,
+      "grad_norm": 0.6632876180495973,
+      "learning_rate": 5.88453783160462e-06,
+      "loss": 1.326,
+      "step": 937
+    },
+    {
+      "epoch": 1.5008,
+      "grad_norm": 0.7261376069660928,
+      "learning_rate": 5.875371982489959e-06,
+      "loss": 1.3355,
+      "step": 938
+    },
+    {
+      "epoch": 1.5024,
+      "grad_norm": 0.698717536628437,
+      "learning_rate": 5.866203097652605e-06,
+      "loss": 1.3141,
+      "step": 939
+    },
+    {
+      "epoch": 1.504,
+      "grad_norm": 0.6974231071817996,
+      "learning_rate": 5.857031208889548e-06,
+      "loss": 1.4058,
+      "step": 940
+    },
+    {
+      "epoch": 1.5056,
+      "grad_norm": 0.8083595450566096,
+      "learning_rate": 5.847856348008188e-06,
+      "loss": 1.3671,
+      "step": 941
+    },
+    {
+      "epoch": 1.5072,
+      "grad_norm": 0.6858637886005077,
+      "learning_rate": 5.838678546826242e-06,
+      "loss": 1.3266,
+      "step": 942
+    },
+    {
+      "epoch": 1.5088,
+      "grad_norm": 0.7934686705320552,
+      "learning_rate": 5.829497837171616e-06,
+      "loss": 1.3656,
+      "step": 943
+    },
+    {
+      "epoch": 1.5104,
+      "grad_norm": 0.6567558487680852,
+      "learning_rate": 5.820314250882304e-06,
+      "loss": 1.3436,
+      "step": 944
+    },
+    {
+      "epoch": 1.512,
+      "grad_norm": 0.7183964565031455,
+      "learning_rate": 5.811127819806277e-06,
+      "loss": 1.2538,
+      "step": 945
+    },
+    {
+      "epoch": 1.5135999999999998,
+      "grad_norm": 0.7543147918920082,
+      "learning_rate": 5.801938575801372e-06,
+      "loss": 1.3949,
+      "step": 946
+    },
+    {
+      "epoch": 1.5152,
+      "grad_norm": 0.6487634395018614,
+      "learning_rate": 5.792746550735182e-06,
+      "loss": 1.3852,
+      "step": 947
+    },
+    {
+      "epoch": 1.5168,
+      "grad_norm": 0.7404786888249957,
+      "learning_rate": 5.7835517764849395e-06,
+      "loss": 1.4052,
+      "step": 948
+    },
+    {
+      "epoch": 1.5184,
+      "grad_norm": 0.6371333320810849,
+      "learning_rate": 5.7743542849374155e-06,
+      "loss": 1.3781,
+      "step": 949
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6732358154653978,
+      "learning_rate": 5.765154107988803e-06,
+      "loss": 1.3888,
+      "step": 950
+    },
+    {
+      "epoch": 1.5215999999999998,
+      "grad_norm": 0.6510570111010161,
+      "learning_rate": 5.755951277544607e-06,
+      "loss": 1.351,
+      "step": 951
+    },
+    {
+      "epoch": 1.5232,
+      "grad_norm": 0.7259778160561365,
+      "learning_rate": 5.746745825519539e-06,
+      "loss": 1.4158,
+      "step": 952
+    },
+    {
+      "epoch": 1.5248,
+      "grad_norm": 0.6679406598144906,
+      "learning_rate": 5.737537783837395e-06,
+      "loss": 1.3864,
+      "step": 953
+    },
+    {
+      "epoch": 1.5264,
+      "grad_norm": 0.7287634426223066,
+      "learning_rate": 5.728327184430955e-06,
+      "loss": 1.3694,
+      "step": 954
+    },
+    {
+      "epoch": 1.528,
+      "grad_norm": 0.6677194166194163,
+      "learning_rate": 5.719114059241871e-06,
+      "loss": 1.3506,
+      "step": 955
+    },
+    {
+      "epoch": 1.5295999999999998,
+      "grad_norm": 0.6812695285815242,
+      "learning_rate": 5.709898440220552e-06,
+      "loss": 1.2856,
+      "step": 956
+    },
+    {
+      "epoch": 1.5312000000000001,
+      "grad_norm": 0.7781535793474241,
+      "learning_rate": 5.700680359326055e-06,
+      "loss": 1.3696,
+      "step": 957
+    },
+    {
+      "epoch": 1.5328,
+      "grad_norm": 0.7239499668939477,
+      "learning_rate": 5.691459848525977e-06,
+      "loss": 1.3204,
+      "step": 958
+    },
+    {
+      "epoch": 1.5344,
+      "grad_norm": 0.7964550990945548,
+      "learning_rate": 5.682236939796337e-06,
+      "loss": 1.3393,
+      "step": 959
+    },
+    {
+      "epoch": 1.536,
+      "grad_norm": 0.7096940218933419,
+      "learning_rate": 5.673011665121477e-06,
+      "loss": 1.3161,
+      "step": 960
+    },
+    {
+      "epoch": 1.5375999999999999,
+      "grad_norm": 0.6831335235145152,
+      "learning_rate": 5.663784056493936e-06,
+      "loss": 1.3491,
+      "step": 961
+    },
+    {
+      "epoch": 1.5392000000000001,
+      "grad_norm": 0.7330090809524039,
+      "learning_rate": 5.6545541459143535e-06,
+      "loss": 1.3349,
+      "step": 962
+    },
+    {
+      "epoch": 1.5408,
+      "grad_norm": 0.7173808101736036,
+      "learning_rate": 5.6453219653913495e-06,
+      "loss": 1.3516,
+      "step": 963
+    },
+    {
+      "epoch": 1.5424,
+      "grad_norm": 0.7213252944658283,
+      "learning_rate": 5.636087546941413e-06,
+      "loss": 1.3733,
+      "step": 964
+    },
+    {
+      "epoch": 1.544,
+      "grad_norm": 0.721930917723235,
+      "learning_rate": 5.6268509225888005e-06,
+      "loss": 1.2949,
+      "step": 965
+    },
+    {
+      "epoch": 1.5455999999999999,
+      "grad_norm": 0.6869128020361388,
+      "learning_rate": 5.617612124365411e-06,
+      "loss": 1.3222,
+      "step": 966
+    },
+    {
+      "epoch": 1.5472000000000001,
+      "grad_norm": 0.6939928448454886,
+      "learning_rate": 5.608371184310688e-06,
+      "loss": 1.2913,
+      "step": 967
+    },
+    {
+      "epoch": 1.5488,
+      "grad_norm": 0.6501944579455962,
+      "learning_rate": 5.5991281344714984e-06,
+      "loss": 1.3419,
+      "step": 968
+    },
+    {
+      "epoch": 1.5504,
+      "grad_norm": 0.6042956861581,
+      "learning_rate": 5.5898830069020325e-06,
+      "loss": 1.3375,
+      "step": 969
+    },
+    {
+      "epoch": 1.552,
+      "grad_norm": 0.6853101882091047,
+      "learning_rate": 5.580635833663679e-06,
+      "loss": 1.3094,
+      "step": 970
+    },
+    {
+      "epoch": 1.5535999999999999,
+      "grad_norm": 0.6462711645109748,
+      "learning_rate": 5.5713866468249235e-06,
+      "loss": 1.3128,
+      "step": 971
+    },
+    {
+      "epoch": 1.5552000000000001,
+      "grad_norm": 0.6447497665323855,
+      "learning_rate": 5.562135478461234e-06,
+      "loss": 1.3297,
+      "step": 972
+    },
+    {
+      "epoch": 1.5568,
+      "grad_norm": 0.6627991099714495,
+      "learning_rate": 5.55288236065495e-06,
+      "loss": 1.2975,
+      "step": 973
+    },
+    {
+      "epoch": 1.5584,
+      "grad_norm": 0.6433048937471955,
+      "learning_rate": 5.5436273254951734e-06,
+      "loss": 1.2784,
+      "step": 974
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6441960845452053,
+      "learning_rate": 5.5343704050776535e-06,
+      "loss": 1.3104,
+      "step": 975
+    },
+    {
+      "epoch": 1.5615999999999999,
+      "grad_norm": 0.6299789892271472,
+      "learning_rate": 5.5251116315046785e-06,
+      "loss": 1.3837,
+      "step": 976
+    },
+    {
+      "epoch": 1.5632000000000001,
+      "grad_norm": 0.7048590361849487,
+      "learning_rate": 5.515851036884964e-06,
+      "loss": 1.3434,
+      "step": 977
+    },
+    {
+      "epoch": 1.5648,
+      "grad_norm": 0.6727046690022734,
+      "learning_rate": 5.5065886533335355e-06,
+      "loss": 1.3333,
+      "step": 978
+    },
+    {
+      "epoch": 1.5664,
+      "grad_norm": 0.6723499858438524,
+      "learning_rate": 5.497324512971632e-06,
+      "loss": 1.3089,
+      "step": 979
+    },
+    {
+      "epoch": 1.568,
+      "grad_norm": 0.6580827069148176,
+      "learning_rate": 5.4880586479265774e-06,
+      "loss": 1.2507,
+      "step": 980
+    },
+    {
+      "epoch": 1.5695999999999999,
+      "grad_norm": 0.6578048389928468,
+      "learning_rate": 5.478791090331677e-06,
+      "loss": 1.3331,
+      "step": 981
+    },
+    {
+      "epoch": 1.5712000000000002,
+      "grad_norm": 0.6787903086607003,
+      "learning_rate": 5.4695218723261115e-06,
+      "loss": 1.3638,
+      "step": 982
+    },
+    {
+      "epoch": 1.5728,
+      "grad_norm": 0.6771294871181338,
+      "learning_rate": 5.46025102605481e-06,
+      "loss": 1.3806,
+      "step": 983
+    },
+    {
+      "epoch": 1.5744,
+      "grad_norm": 0.658051303187384,
+      "learning_rate": 5.4509785836683606e-06,
+      "loss": 1.312,
+      "step": 984
+    },
+    {
+      "epoch": 1.576,
+      "grad_norm": 0.6396941923519958,
+      "learning_rate": 5.441704577322877e-06,
+      "loss": 1.3203,
+      "step": 985
+    },
+    {
+      "epoch": 1.5776,
+      "grad_norm": 0.6709259033261713,
+      "learning_rate": 5.4324290391798995e-06,
+      "loss": 1.2796,
+      "step": 986
+    },
+    {
+      "epoch": 1.5792000000000002,
+      "grad_norm": 0.6796742067662461,
+      "learning_rate": 5.423152001406282e-06,
+      "loss": 1.336,
+      "step": 987
+    },
+    {
+      "epoch": 1.5808,
+      "grad_norm": 0.6445502466525074,
+      "learning_rate": 5.413873496174077e-06,
+      "loss": 1.2609,
+      "step": 988
+    },
+    {
+      "epoch": 1.5824,
+      "grad_norm": 0.6880008170930073,
+      "learning_rate": 5.404593555660424e-06,
+      "loss": 1.305,
+      "step": 989
+    },
+    {
+      "epoch": 1.584,
+      "grad_norm": 0.6923697756557811,
+      "learning_rate": 5.39531221204745e-06,
+      "loss": 1.325,
+      "step": 990
+    },
+    {
+      "epoch": 1.5856,
+      "grad_norm": 0.6861246117770287,
+      "learning_rate": 5.3860294975221335e-06,
+      "loss": 1.3013,
+      "step": 991
+    },
+    {
+      "epoch": 1.5872000000000002,
+      "grad_norm": 0.7255374735711748,
+      "learning_rate": 5.376745444276219e-06,
+      "loss": 1.3928,
+      "step": 992
+    },
+    {
+      "epoch": 1.5888,
+      "grad_norm": 0.7167817340663303,
+      "learning_rate": 5.3674600845060856e-06,
+      "loss": 1.3081,
+      "step": 993
+    },
+    {
+      "epoch": 1.5904,
+      "grad_norm": 0.6637194016304895,
+      "learning_rate": 5.358173450412649e-06,
+      "loss": 1.3382,
+      "step": 994
+    },
+    {
+      "epoch": 1.592,
+      "grad_norm": 0.7196584223339284,
+      "learning_rate": 5.34888557420124e-06,
+      "loss": 1.3776,
+      "step": 995
+    },
+    {
+      "epoch": 1.5936,
+      "grad_norm": 0.6546003746621862,
+      "learning_rate": 5.339596488081501e-06,
+      "loss": 1.356,
+      "step": 996
+    },
+    {
+      "epoch": 1.5952,
+      "grad_norm": 0.6793006723135461,
+      "learning_rate": 5.330306224267268e-06,
+      "loss": 1.3708,
+      "step": 997
+    },
+    {
+      "epoch": 1.5968,
+      "grad_norm": 0.7039683107083351,
+      "learning_rate": 5.321014814976459e-06,
+      "loss": 1.3763,
+      "step": 998
+    },
+    {
+      "epoch": 1.5984,
+      "grad_norm": 0.6422007759823517,
+      "learning_rate": 5.311722292430966e-06,
+      "loss": 1.2448,
+      "step": 999
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.7046446722704103,
+      "learning_rate": 5.302428688856544e-06,
+      "loss": 1.2619,
+      "step": 1000
+    },
+    {
+      "epoch": 1.6016,
+      "grad_norm": 0.6776242294882786,
+      "learning_rate": 5.293134036482697e-06,
+      "loss": 1.3077,
+      "step": 1001
+    },
+    {
+      "epoch": 1.6032,
+      "grad_norm": 0.7121529946843654,
+      "learning_rate": 5.283838367542562e-06,
+      "loss": 1.3077,
+      "step": 1002
+    },
+    {
+      "epoch": 1.6048,
+      "grad_norm": 0.6767845793214963,
+      "learning_rate": 5.274541714272805e-06,
+      "loss": 1.3303,
+      "step": 1003
+    },
+    {
+      "epoch": 1.6064,
+      "grad_norm": 0.7058053408872256,
+      "learning_rate": 5.265244108913503e-06,
+      "loss": 1.3549,
+      "step": 1004
+    },
+    {
+      "epoch": 1.608,
+      "grad_norm": 0.6856724343424353,
+      "learning_rate": 5.255945583708037e-06,
+      "loss": 1.3881,
+      "step": 1005
+    },
+    {
+      "epoch": 1.6096,
+      "grad_norm": 0.7136599283136974,
+      "learning_rate": 5.2466461709029755e-06,
+      "loss": 1.3534,
+      "step": 1006
+    },
+    {
+      "epoch": 1.6112,
+      "grad_norm": 0.6426659402549703,
+      "learning_rate": 5.237345902747969e-06,
+      "loss": 1.273,
+      "step": 1007
+    },
+    {
+      "epoch": 1.6128,
+      "grad_norm": 0.7044391754353025,
+      "learning_rate": 5.228044811495632e-06,
+      "loss": 1.3952,
+      "step": 1008
+    },
+    {
+      "epoch": 1.6143999999999998,
+      "grad_norm": 0.6916348817848671,
+      "learning_rate": 5.218742929401432e-06,
+      "loss": 1.3021,
+      "step": 1009
+    },
+    {
+      "epoch": 1.616,
+      "grad_norm": 0.6674033908599477,
+      "learning_rate": 5.2094402887235805e-06,
+      "loss": 1.3186,
+      "step": 1010
+    },
+    {
+      "epoch": 1.6176,
+      "grad_norm": 0.703218580548564,
+      "learning_rate": 5.200136921722919e-06,
+      "loss": 1.2899,
+      "step": 1011
+    },
+    {
+      "epoch": 1.6192,
+      "grad_norm": 0.6931388436072926,
+      "learning_rate": 5.1908328606628114e-06,
+      "loss": 1.3392,
+      "step": 1012
+    },
+    {
+      "epoch": 1.6208,
+      "grad_norm": 0.6381832447084699,
+      "learning_rate": 5.181528137809023e-06,
+      "loss": 1.336,
+      "step": 1013
+    },
+    {
+      "epoch": 1.6223999999999998,
+      "grad_norm": 0.6563297836181966,
+      "learning_rate": 5.1722227854296195e-06,
+      "loss": 1.3443,
+      "step": 1014
+    },
+    {
+      "epoch": 1.624,
+      "grad_norm": 0.6386683444856855,
+      "learning_rate": 5.162916835794843e-06,
+      "loss": 1.3577,
+      "step": 1015
+    },
+    {
+      "epoch": 1.6256,
+      "grad_norm": 0.609801975510135,
+      "learning_rate": 5.1536103211770135e-06,
+      "loss": 1.2981,
+      "step": 1016
+    },
+    {
+      "epoch": 1.6272,
+      "grad_norm": 0.6616774090463595,
+      "learning_rate": 5.14430327385041e-06,
+      "loss": 1.3331,
+      "step": 1017
+    },
+    {
+      "epoch": 1.6288,
+      "grad_norm": 0.6710494915322406,
+      "learning_rate": 5.134995726091152e-06,
+      "loss": 1.3624,
+      "step": 1018
+    },
+    {
+      "epoch": 1.6303999999999998,
+      "grad_norm": 0.6427499798655534,
+      "learning_rate": 5.1256877101771015e-06,
+      "loss": 1.2942,
+      "step": 1019
+    },
+    {
+      "epoch": 1.6320000000000001,
+      "grad_norm": 0.6118288062570725,
+      "learning_rate": 5.116379258387742e-06,
+      "loss": 1.3039,
+      "step": 1020
+    },
+    {
+      "epoch": 1.6336,
+      "grad_norm": 0.6535235926894215,
+      "learning_rate": 5.1070704030040675e-06,
+      "loss": 1.3087,
+      "step": 1021
+    },
+    {
+      "epoch": 1.6352,
+      "grad_norm": 0.6268822402582575,
+      "learning_rate": 5.097761176308471e-06,
+      "loss": 1.3048,
+      "step": 1022
+    },
+    {
+      "epoch": 1.6368,
+      "grad_norm": 0.6398785242744189,
+      "learning_rate": 5.088451610584638e-06,
+      "loss": 1.3122,
+      "step": 1023
+    },
+    {
+      "epoch": 1.6383999999999999,
+      "grad_norm": 0.6118275584123517,
+      "learning_rate": 5.079141738117423e-06,
+      "loss": 1.2609,
+      "step": 1024
+    },
+    {
+      "epoch": 1.6400000000000001,
+      "grad_norm": 0.6599367598082485,
+      "learning_rate": 5.06983159119275e-06,
+      "loss": 1.3613,
+      "step": 1025
+    },
+    {
+      "epoch": 1.6416,
+      "grad_norm": 0.6296132637799025,
+      "learning_rate": 5.060521202097491e-06,
+      "loss": 1.2854,
+      "step": 1026
+    },
+    {
+      "epoch": 1.6432,
+      "grad_norm": 0.6495770358552447,
+      "learning_rate": 5.051210603119358e-06,
+      "loss": 1.3002,
+      "step": 1027
+    },
+    {
+      "epoch": 1.6448,
+      "grad_norm": 0.6442511703517048,
+      "learning_rate": 5.041899826546791e-06,
+      "loss": 1.3711,
+      "step": 1028
+    },
+    {
+      "epoch": 1.6463999999999999,
+      "grad_norm": 0.6472700904472487,
+      "learning_rate": 5.032588904668851e-06,
+      "loss": 1.3201,
+      "step": 1029
+    },
+    {
+      "epoch": 1.6480000000000001,
+      "grad_norm": 0.7150550673953523,
+      "learning_rate": 5.023277869775097e-06,
+      "loss": 1.3633,
+      "step": 1030
+    },
+    {
+      "epoch": 1.6496,
+      "grad_norm": 0.6208136907324723,
+      "learning_rate": 5.013966754155482e-06,
+      "loss": 1.3185,
+      "step": 1031
+    },
+    {
+      "epoch": 1.6512,
+      "grad_norm": 0.6754478207390778,
+      "learning_rate": 5.004655590100238e-06,
+      "loss": 1.3804,
+      "step": 1032
+    },
+    {
+      "epoch": 1.6528,
+      "grad_norm": 0.6731237842271147,
+      "learning_rate": 4.995344409899764e-06,
+      "loss": 1.3049,
+      "step": 1033
+    },
+    {
+      "epoch": 1.6543999999999999,
+      "grad_norm": 0.6146268515269073,
+      "learning_rate": 4.986033245844519e-06,
+      "loss": 1.284,
+      "step": 1034
+    },
+    {
+      "epoch": 1.6560000000000001,
+      "grad_norm": 0.6602348229313003,
+      "learning_rate": 4.976722130224904e-06,
+      "loss": 1.3512,
+      "step": 1035
+    },
+    {
+      "epoch": 1.6576,
+      "grad_norm": 0.6442164432659385,
+      "learning_rate": 4.967411095331149e-06,
+      "loss": 1.3583,
+      "step": 1036
+    },
+    {
+      "epoch": 1.6592,
+      "grad_norm": 0.7117332566740452,
+      "learning_rate": 4.95810017345321e-06,
+      "loss": 1.3426,
+      "step": 1037
+    },
+    {
+      "epoch": 1.6608,
+      "grad_norm": 0.6532902324728062,
+      "learning_rate": 4.948789396880644e-06,
+      "loss": 1.3054,
+      "step": 1038
+    },
+    {
+      "epoch": 1.6623999999999999,
+      "grad_norm": 0.6833306056359661,
+      "learning_rate": 4.939478797902512e-06,
+      "loss": 1.3001,
+      "step": 1039
+    },
+    {
+      "epoch": 1.6640000000000001,
+      "grad_norm": 0.6927267498021275,
+      "learning_rate": 4.930168408807252e-06,
+      "loss": 1.4079,
+      "step": 1040
+    },
+    {
+      "epoch": 1.6656,
+      "grad_norm": 0.647064891159953,
+      "learning_rate": 4.920858261882578e-06,
+      "loss": 1.2676,
+      "step": 1041
+    },
+    {
+      "epoch": 1.6672,
+      "grad_norm": 0.6420745499601633,
+      "learning_rate": 4.911548389415363e-06,
+      "loss": 1.3133,
+      "step": 1042
+    },
+    {
+      "epoch": 1.6688,
+      "grad_norm": 0.6779247746803815,
+      "learning_rate": 4.9022388236915306e-06,
+      "loss": 1.2579,
+      "step": 1043
+    },
+    {
+      "epoch": 1.6703999999999999,
+      "grad_norm": 0.658650307415508,
+      "learning_rate": 4.892929596995934e-06,
+      "loss": 1.3957,
+      "step": 1044
+    },
+    {
+      "epoch": 1.6720000000000002,
+      "grad_norm": 0.6741505575557593,
+      "learning_rate": 4.883620741612259e-06,
+      "loss": 1.2919,
+      "step": 1045
+    },
+    {
+      "epoch": 1.6736,
+      "grad_norm": 0.6233207770911169,
+      "learning_rate": 4.8743122898229e-06,
+      "loss": 1.2986,
+      "step": 1046
+    },
+    {
+      "epoch": 1.6752,
+      "grad_norm": 0.7007853430104943,
+      "learning_rate": 4.865004273908851e-06,
+      "loss": 1.2789,
+      "step": 1047
+    },
+    {
+      "epoch": 1.6768,
+      "grad_norm": 0.6859956860952263,
+      "learning_rate": 4.855696726149593e-06,
+      "loss": 1.3421,
+      "step": 1048
+    },
+    {
+      "epoch": 1.6784,
+      "grad_norm": 0.6284667111907238,
+      "learning_rate": 4.846389678822987e-06,
+      "loss": 1.3721,
+      "step": 1049
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 0.6120395964519181,
+      "learning_rate": 4.837083164205159e-06,
+      "loss": 1.3256,
+      "step": 1050
+    },
+    {
+      "epoch": 1.6816,
+      "grad_norm": 9.477627744514695,
+      "learning_rate": 4.827777214570384e-06,
+      "loss": 1.4344,
+      "step": 1051
+    },
+    {
+      "epoch": 1.6832,
+      "grad_norm": 0.7630133975015444,
+      "learning_rate": 4.818471862190979e-06,
+      "loss": 1.2813,
+      "step": 1052
+    },
+    {
+      "epoch": 1.6848,
+      "grad_norm": 0.6602127460281478,
+      "learning_rate": 4.809167139337191e-06,
+      "loss": 1.2883,
+      "step": 1053
+    },
+    {
+      "epoch": 1.6864,
+      "grad_norm": 0.674046897952832,
+      "learning_rate": 4.799863078277082e-06,
+      "loss": 1.3264,
+      "step": 1054
+    },
+    {
+      "epoch": 1.688,
+      "grad_norm": 0.6875303254174575,
+      "learning_rate": 4.790559711276422e-06,
+      "loss": 1.2771,
+      "step": 1055
+    },
+    {
+      "epoch": 1.6896,
+      "grad_norm": 0.6231529017870409,
+      "learning_rate": 4.781257070598571e-06,
+      "loss": 1.3577,
+      "step": 1056
+    },
+    {
+      "epoch": 1.6912,
+      "grad_norm": 0.6737834660651841,
+      "learning_rate": 4.771955188504371e-06,
+      "loss": 1.3013,
+      "step": 1057
+    },
+    {
+      "epoch": 1.6928,
+      "grad_norm": 0.6549813584473074,
+      "learning_rate": 4.762654097252033e-06,
+      "loss": 1.3499,
+      "step": 1058
+    },
+    {
+      "epoch": 1.6944,
+      "grad_norm": 0.6335210467281328,
+      "learning_rate": 4.753353829097025e-06,
+      "loss": 1.3617,
+      "step": 1059
+    },
+    {
+      "epoch": 1.696,
+      "grad_norm": 0.6589745427538588,
+      "learning_rate": 4.7440544162919645e-06,
+      "loss": 1.3597,
+      "step": 1060
+    },
+    {
+      "epoch": 1.6976,
+      "grad_norm": 0.6546037989920358,
+      "learning_rate": 4.734755891086498e-06,
+      "loss": 1.3387,
+      "step": 1061
+    },
+    {
+      "epoch": 1.6992,
+      "grad_norm": 0.6376552606514404,
+      "learning_rate": 4.725458285727195e-06,
+      "loss": 1.2862,
+      "step": 1062
+    },
+    {
+      "epoch": 1.7008,
+      "grad_norm": 0.6505276539357651,
+      "learning_rate": 4.716161632457438e-06,
+      "loss": 1.3391,
+      "step": 1063
+    },
+    {
+      "epoch": 1.7024,
+      "grad_norm": 0.6150080362086289,
+      "learning_rate": 4.7068659635173034e-06,
+      "loss": 1.3168,
+      "step": 1064
+    },
+    {
+      "epoch": 1.704,
+      "grad_norm": 0.6676577825157071,
+      "learning_rate": 4.6975713111434556e-06,
+      "loss": 1.2968,
+      "step": 1065
+    },
+    {
+      "epoch": 1.7056,
+      "grad_norm": 0.6623871024780047,
+      "learning_rate": 4.688277707569035e-06,
+      "loss": 1.3935,
+      "step": 1066
+    },
+    {
+      "epoch": 1.7072,
+      "grad_norm": 0.6356498428266403,
+      "learning_rate": 4.678985185023542e-06,
+      "loss": 1.2918,
+      "step": 1067
+    },
+    {
+      "epoch": 1.7088,
+      "grad_norm": 0.6589239326644672,
+      "learning_rate": 4.669693775732733e-06,
+      "loss": 1.2978,
+      "step": 1068
+    },
+    {
+      "epoch": 1.7104,
+      "grad_norm": 0.6617616871815194,
+      "learning_rate": 4.660403511918499e-06,
+      "loss": 1.3574,
+      "step": 1069
+    },
+    {
+      "epoch": 1.712,
+      "grad_norm": 0.673461102751327,
+      "learning_rate": 4.65111442579876e-06,
+      "loss": 1.2326,
+      "step": 1070
+    },
+    {
+      "epoch": 1.7136,
+      "grad_norm": 0.6465104098753137,
+      "learning_rate": 4.641826549587352e-06,
+      "loss": 1.3584,
+      "step": 1071
+    },
+    {
+      "epoch": 1.7151999999999998,
+      "grad_norm": 0.6837177250746823,
+      "learning_rate": 4.632539915493915e-06,
+      "loss": 1.3199,
+      "step": 1072
+    },
+    {
+      "epoch": 1.7168,
+      "grad_norm": 0.6378927101361298,
+      "learning_rate": 4.623254555723783e-06,
+      "loss": 1.3117,
+      "step": 1073
+    },
+    {
+      "epoch": 1.7184,
+      "grad_norm": 0.6001502744569444,
+      "learning_rate": 4.613970502477867e-06,
+      "loss": 1.2717,
+      "step": 1074
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.7347941195132434,
+      "learning_rate": 4.604687787952552e-06,
+      "loss": 1.3065,
+      "step": 1075
+    },
+    {
+      "epoch": 1.7216,
+      "grad_norm": 0.6569992799984579,
+      "learning_rate": 4.5954064443395765e-06,
+      "loss": 1.3382,
+      "step": 1076
+    },
+    {
+      "epoch": 1.7231999999999998,
+      "grad_norm": 0.646642968361254,
+      "learning_rate": 4.586126503825925e-06,
+      "loss": 1.2715,
+      "step": 1077
+    },
+    {
+      "epoch": 1.7248,
+      "grad_norm": 0.6447465277176916,
+      "learning_rate": 4.57684799859372e-06,
+      "loss": 1.2508,
+      "step": 1078
+    },
+    {
+      "epoch": 1.7264,
+      "grad_norm": 0.6246677764173614,
+      "learning_rate": 4.567570960820101e-06,
+      "loss": 1.3005,
+      "step": 1079
+    },
+    {
+      "epoch": 1.728,
+      "grad_norm": 0.718619859214415,
+      "learning_rate": 4.558295422677124e-06,
+      "loss": 1.3835,
+      "step": 1080
+    },
+    {
+      "epoch": 1.7296,
+      "grad_norm": 0.666726460395986,
+      "learning_rate": 4.54902141633164e-06,
+      "loss": 1.2965,
+      "step": 1081
+    },
+    {
+      "epoch": 1.7311999999999999,
+      "grad_norm": 0.6653200201084793,
+      "learning_rate": 4.539748973945191e-06,
+      "loss": 1.3058,
+      "step": 1082
+    },
+    {
+      "epoch": 1.7328000000000001,
+      "grad_norm": 0.7681018644486702,
+      "learning_rate": 4.53047812767389e-06,
+      "loss": 1.2987,
+      "step": 1083
+    },
+    {
+      "epoch": 1.7344,
+      "grad_norm": 0.6398663640816541,
+      "learning_rate": 4.5212089096683234e-06,
+      "loss": 1.32,
+      "step": 1084
+    },
+    {
+      "epoch": 1.736,
+      "grad_norm": 0.6800151750926361,
+      "learning_rate": 4.511941352073424e-06,
+      "loss": 1.35,
+      "step": 1085
+    },
+    {
+      "epoch": 1.7376,
+      "grad_norm": 0.6718688083849886,
+      "learning_rate": 4.5026754870283695e-06,
+      "loss": 1.3506,
+      "step": 1086
+    },
+    {
+      "epoch": 1.7391999999999999,
+      "grad_norm": 0.6663763372973037,
+      "learning_rate": 4.493411346666465e-06,
+      "loss": 1.3781,
+      "step": 1087
+    },
+    {
+      "epoch": 1.7408000000000001,
+      "grad_norm": 0.6526940986450056,
+      "learning_rate": 4.484148963115038e-06,
+      "loss": 1.2862,
+      "step": 1088
+    },
+    {
+      "epoch": 1.7424,
+      "grad_norm": 0.6390643824071945,
+      "learning_rate": 4.474888368495322e-06,
+      "loss": 1.2854,
+      "step": 1089
+    },
+    {
+      "epoch": 1.744,
+      "grad_norm": 0.6300013294518952,
+      "learning_rate": 4.465629594922348e-06,
+      "loss": 1.3778,
+      "step": 1090
+    },
+    {
+      "epoch": 1.7456,
+      "grad_norm": 0.6734594226919762,
+      "learning_rate": 4.456372674504828e-06,
+      "loss": 1.3236,
+      "step": 1091
+    },
+    {
+      "epoch": 1.7471999999999999,
+      "grad_norm": 0.6474947506469003,
+      "learning_rate": 4.447117639345052e-06,
+      "loss": 1.3486,
+      "step": 1092
+    },
+    {
+      "epoch": 1.7488000000000001,
+      "grad_norm": 0.6246843007099236,
+      "learning_rate": 4.437864521538768e-06,
+      "loss": 1.3212,
+      "step": 1093
+    },
+    {
+      "epoch": 1.7504,
+      "grad_norm": 0.6905194790145287,
+      "learning_rate": 4.428613353175078e-06,
+      "loss": 1.3938,
+      "step": 1094
+    },
+    {
+      "epoch": 1.752,
+      "grad_norm": 0.6126709874156032,
+      "learning_rate": 4.4193641663363214e-06,
+      "loss": 1.3891,
+      "step": 1095
+    },
+    {
+      "epoch": 1.7536,
+      "grad_norm": 0.9554152874649552,
+      "learning_rate": 4.410116993097968e-06,
+      "loss": 1.3403,
+      "step": 1096
+    },
+    {
+      "epoch": 1.7551999999999999,
+      "grad_norm": 0.6394553805815464,
+      "learning_rate": 4.400871865528502e-06,
+      "loss": 1.3615,
+      "step": 1097
+    },
+    {
+      "epoch": 1.7568000000000001,
+      "grad_norm": 0.6692972471235337,
+      "learning_rate": 4.391628815689314e-06,
+      "loss": 1.33,
+      "step": 1098
+    },
+    {
+      "epoch": 1.7584,
+      "grad_norm": 0.6319306005857699,
+      "learning_rate": 4.382387875634592e-06,
+      "loss": 1.3584,
+      "step": 1099
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6327249949086355,
+      "learning_rate": 4.373149077411203e-06,
+      "loss": 1.2973,
+      "step": 1100
+    },
+    {
+      "epoch": 1.7616,
+      "grad_norm": 0.6354375106302664,
+      "learning_rate": 4.363912453058589e-06,
+      "loss": 1.3054,
+      "step": 1101
+    },
+    {
+      "epoch": 1.7631999999999999,
+      "grad_norm": 0.6296910321672626,
+      "learning_rate": 4.354678034608654e-06,
+      "loss": 1.3193,
+      "step": 1102
+    },
+    {
+      "epoch": 1.7648000000000001,
+      "grad_norm": 0.6163418293015167,
+      "learning_rate": 4.345445854085649e-06,
+      "loss": 1.3092,
+      "step": 1103
+    },
+    {
+      "epoch": 1.7664,
+      "grad_norm": 0.6249755666929132,
+      "learning_rate": 4.336215943506066e-06,
+      "loss": 1.342,
+      "step": 1104
+    },
+    {
+      "epoch": 1.768,
+      "grad_norm": 0.6553674995128351,
+      "learning_rate": 4.326988334878526e-06,
+      "loss": 1.3148,
+      "step": 1105
+    },
+    {
+      "epoch": 1.7696,
+      "grad_norm": 0.6706602137256806,
+      "learning_rate": 4.317763060203665e-06,
+      "loss": 1.3003,
+      "step": 1106
+    },
+    {
+      "epoch": 1.7711999999999999,
+      "grad_norm": 0.6124841462934393,
+      "learning_rate": 4.308540151474027e-06,
+      "loss": 1.2895,
+      "step": 1107
+    },
+    {
+      "epoch": 1.7728000000000002,
+      "grad_norm": 0.6360543916442886,
+      "learning_rate": 4.299319640673948e-06,
+      "loss": 1.432,
+      "step": 1108
+    },
+    {
+      "epoch": 1.7744,
+      "grad_norm": 0.693064249327211,
+      "learning_rate": 4.290101559779451e-06,
+      "loss": 1.344,
+      "step": 1109
+    },
+    {
+      "epoch": 1.776,
+      "grad_norm": 0.6191542641131123,
+      "learning_rate": 4.280885940758131e-06,
+      "loss": 1.3562,
+      "step": 1110
+    },
+    {
+      "epoch": 1.7776,
+      "grad_norm": 0.644719947360838,
+      "learning_rate": 4.271672815569047e-06,
+      "loss": 1.3486,
+      "step": 1111
+    },
+    {
+      "epoch": 1.7792,
+      "grad_norm": 0.6861381977735842,
+      "learning_rate": 4.262462216162606e-06,
+      "loss": 1.3153,
+      "step": 1112
+    },
+    {
+      "epoch": 1.7808000000000002,
+      "grad_norm": 0.5927922956595179,
+      "learning_rate": 4.253254174480462e-06,
+      "loss": 1.3468,
+      "step": 1113
+    },
+    {
+      "epoch": 1.7824,
+      "grad_norm": 0.6512078491954844,
+      "learning_rate": 4.244048722455393e-06,
+      "loss": 1.3225,
+      "step": 1114
+    },
+    {
+      "epoch": 1.784,
+      "grad_norm": 0.6643115798059935,
+      "learning_rate": 4.234845892011198e-06,
+      "loss": 1.3283,
+      "step": 1115
+    },
+    {
+      "epoch": 1.7856,
+      "grad_norm": 0.6590872003967726,
+      "learning_rate": 4.225645715062585e-06,
+      "loss": 1.3268,
+      "step": 1116
+    },
+    {
+      "epoch": 1.7872,
+      "grad_norm": 0.6093150767645139,
+      "learning_rate": 4.216448223515061e-06,
+      "loss": 1.3442,
+      "step": 1117
+    },
+    {
+      "epoch": 1.7888,
+      "grad_norm": 0.6374009687576389,
+      "learning_rate": 4.2072534492648184e-06,
+      "loss": 1.2968,
+      "step": 1118
+    },
+    {
+      "epoch": 1.7904,
+      "grad_norm": 0.6632602593723123,
+      "learning_rate": 4.198061424198627e-06,
+      "loss": 1.3119,
+      "step": 1119
+    },
+    {
+      "epoch": 1.792,
+      "grad_norm": 0.6154913941198491,
+      "learning_rate": 4.188872180193723e-06,
+      "loss": 1.3554,
+      "step": 1120
+    },
+    {
+      "epoch": 1.7936,
+      "grad_norm": 0.6667738684866328,
+      "learning_rate": 4.179685749117698e-06,
+      "loss": 1.3608,
+      "step": 1121
+    },
+    {
+      "epoch": 1.7952,
+      "grad_norm": 0.6542083607913017,
+      "learning_rate": 4.170502162828385e-06,
+      "loss": 1.249,
+      "step": 1122
+    },
+    {
+      "epoch": 1.7968,
+      "grad_norm": 0.6577532681015594,
+      "learning_rate": 4.161321453173759e-06,
+      "loss": 1.3565,
+      "step": 1123
+    },
+    {
+      "epoch": 1.7984,
+      "grad_norm": 0.7028418988884404,
+      "learning_rate": 4.152143651991812e-06,
+      "loss": 1.2523,
+      "step": 1124
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.6264576552195866,
+      "learning_rate": 4.142968791110455e-06,
+      "loss": 1.3218,
+      "step": 1125
+    },
+    {
+      "epoch": 1.8016,
+      "grad_norm": 0.7150455943315028,
+      "learning_rate": 4.133796902347397e-06,
+      "loss": 1.3483,
+      "step": 1126
+    },
+    {
+      "epoch": 1.8032,
+      "grad_norm": 0.680846661361651,
+      "learning_rate": 4.124628017510043e-06,
+      "loss": 1.3012,
+      "step": 1127
+    },
+    {
+      "epoch": 1.8048,
+      "grad_norm": 0.6674439174878533,
+      "learning_rate": 4.115462168395382e-06,
+      "loss": 1.3076,
+      "step": 1128
+    },
+    {
+      "epoch": 1.8064,
+      "grad_norm": 0.7410498613010643,
+      "learning_rate": 4.106299386789873e-06,
+      "loss": 1.3402,
+      "step": 1129
+    },
+    {
+      "epoch": 1.808,
+      "grad_norm": 0.6043783458729426,
+      "learning_rate": 4.097139704469337e-06,
+      "loss": 1.313,
+      "step": 1130
+    },
+    {
+      "epoch": 1.8096,
+      "grad_norm": 0.6418409247208597,
+      "learning_rate": 4.0879831531988485e-06,
+      "loss": 1.3127,
+      "step": 1131
+    },
+    {
+      "epoch": 1.8112,
+      "grad_norm": 0.7070947740415923,
+      "learning_rate": 4.078829764732621e-06,
+      "loss": 1.4422,
+      "step": 1132
+    },
+    {
+      "epoch": 1.8128,
+      "grad_norm": 0.6079565278967628,
+      "learning_rate": 4.069679570813903e-06,
+      "loss": 1.3842,
+      "step": 1133
+    },
+    {
+      "epoch": 1.8144,
+      "grad_norm": 0.6961923923309354,
+      "learning_rate": 4.060532603174865e-06,
+      "loss": 1.2826,
+      "step": 1134
+    },
+    {
+      "epoch": 1.8159999999999998,
+      "grad_norm": 0.6741760201006882,
+      "learning_rate": 4.051388893536484e-06,
+      "loss": 1.317,
+      "step": 1135
+    },
+    {
+      "epoch": 1.8176,
+      "grad_norm": 0.6320597729702291,
+      "learning_rate": 4.042248473608442e-06,
+      "loss": 1.3303,
+      "step": 1136
+    },
+    {
+      "epoch": 1.8192,
+      "grad_norm": 0.6872058122292877,
+      "learning_rate": 4.033111375089013e-06,
+      "loss": 1.2303,
+      "step": 1137
+    },
+    {
+      "epoch": 1.8208,
+      "grad_norm": 0.640554774472593,
+      "learning_rate": 4.023977629664951e-06,
+      "loss": 1.3168,
+      "step": 1138
+    },
+    {
+      "epoch": 1.8224,
+      "grad_norm": 0.6666998532099543,
+      "learning_rate": 4.0148472690113845e-06,
+      "loss": 1.3457,
+      "step": 1139
+    },
+    {
+      "epoch": 1.8239999999999998,
+      "grad_norm": 8.882663559326115,
+      "learning_rate": 4.0057203247917e-06,
+      "loss": 1.3282,
+      "step": 1140
+    },
+    {
+      "epoch": 1.8256000000000001,
+      "grad_norm": 0.7719246604410158,
+      "learning_rate": 3.996596828657437e-06,
+      "loss": 1.3496,
+      "step": 1141
+    },
+    {
+      "epoch": 1.8272,
+      "grad_norm": 0.6779289351162358,
+      "learning_rate": 3.987476812248181e-06,
+      "loss": 1.3235,
+      "step": 1142
+    },
+    {
+      "epoch": 1.8288,
+      "grad_norm": 0.6661528545784979,
+      "learning_rate": 3.978360307191444e-06,
+      "loss": 1.2895,
+      "step": 1143
+    },
+    {
+      "epoch": 1.8304,
+      "grad_norm": 0.6437488234826078,
+      "learning_rate": 3.969247345102567e-06,
+      "loss": 1.4024,
+      "step": 1144
+    },
+    {
+      "epoch": 1.8319999999999999,
+      "grad_norm": 0.6433176751665047,
+      "learning_rate": 3.960137957584601e-06,
+      "loss": 1.3374,
+      "step": 1145
+    },
+    {
+      "epoch": 1.8336000000000001,
+      "grad_norm": 0.6463711322148512,
+      "learning_rate": 3.9510321762282e-06,
+      "loss": 1.3456,
+      "step": 1146
+    },
+    {
+      "epoch": 1.8352,
+      "grad_norm": 0.6849911460711375,
+      "learning_rate": 3.941930032611513e-06,
+      "loss": 1.2735,
+      "step": 1147
+    },
+    {
+      "epoch": 1.8368,
+      "grad_norm": 0.675903044323276,
+      "learning_rate": 3.932831558300074e-06,
+      "loss": 1.3319,
+      "step": 1148
+    },
+    {
+      "epoch": 1.8384,
+      "grad_norm": 0.5965111552137392,
+      "learning_rate": 3.923736784846693e-06,
+      "loss": 1.3358,
+      "step": 1149
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 0.7019611467511079,
+      "learning_rate": 3.914645743791346e-06,
+      "loss": 1.3691,
+      "step": 1150
+    },
+    {
+      "epoch": 1.8416000000000001,
+      "grad_norm": 0.6315833648027265,
+      "learning_rate": 3.90555846666106e-06,
+      "loss": 1.3187,
+      "step": 1151
+    },
+    {
+      "epoch": 1.8432,
+      "grad_norm": 0.6735962976408618,
+      "learning_rate": 3.896474984969817e-06,
+      "loss": 1.3441,
+      "step": 1152
+    },
+    {
+      "epoch": 1.8448,
+      "grad_norm": 0.6957894766068956,
+      "learning_rate": 3.887395330218429e-06,
+      "loss": 1.2923,
+      "step": 1153
+    },
+    {
+      "epoch": 1.8464,
+      "grad_norm": 0.620429070821768,
+      "learning_rate": 3.878319533894443e-06,
+      "loss": 1.3123,
+      "step": 1154
+    },
+    {
+      "epoch": 1.8479999999999999,
+      "grad_norm": 0.7264378050956252,
+      "learning_rate": 3.869247627472021e-06,
+      "loss": 1.3117,
+      "step": 1155
+    },
+    {
+      "epoch": 1.8496000000000001,
+      "grad_norm": 0.63239511778322,
+      "learning_rate": 3.860179642411838e-06,
+      "loss": 1.3048,
+      "step": 1156
+    },
+    {
+      "epoch": 1.8512,
+      "grad_norm": 0.6312975175784915,
+      "learning_rate": 3.851115610160967e-06,
+      "loss": 1.2683,
+      "step": 1157
+    },
+    {
+      "epoch": 1.8528,
+      "grad_norm": 0.6512203166918432,
+      "learning_rate": 3.842055562152775e-06,
+      "loss": 1.286,
+      "step": 1158
+    },
+    {
+      "epoch": 1.8544,
+      "grad_norm": 0.671195552146292,
+      "learning_rate": 3.8329995298068114e-06,
+      "loss": 1.328,
+      "step": 1159
+    },
+    {
+      "epoch": 1.8559999999999999,
+      "grad_norm": 0.6845751953551978,
+      "learning_rate": 3.8239475445287015e-06,
+      "loss": 1.4028,
+      "step": 1160
+    },
+    {
+      "epoch": 1.8576000000000001,
+      "grad_norm": 0.9772474164224244,
+      "learning_rate": 3.814899637710031e-06,
+      "loss": 1.3311,
+      "step": 1161
+    },
+    {
+      "epoch": 1.8592,
+      "grad_norm": 0.7218057172813025,
+      "learning_rate": 3.8058558407282465e-06,
+      "loss": 1.3017,
+      "step": 1162
+    },
+    {
+      "epoch": 1.8608,
+      "grad_norm": 0.6761286129178691,
+      "learning_rate": 3.7968161849465395e-06,
+      "loss": 1.3104,
+      "step": 1163
+    },
+    {
+      "epoch": 1.8624,
+      "grad_norm": 0.6478451127663627,
+      "learning_rate": 3.78778070171374e-06,
+      "loss": 1.3093,
+      "step": 1164
+    },
+    {
+      "epoch": 1.8639999999999999,
+      "grad_norm": 0.7121302352065212,
+      "learning_rate": 3.7787494223642096e-06,
+      "loss": 1.285,
+      "step": 1165
+    },
+    {
+      "epoch": 1.8656000000000001,
+      "grad_norm": 0.6647705098807707,
+      "learning_rate": 3.7697223782177304e-06,
+      "loss": 1.3803,
+      "step": 1166
+    },
+    {
+      "epoch": 1.8672,
+      "grad_norm": 0.6680092679641494,
+      "learning_rate": 3.760699600579399e-06,
+      "loss": 1.3269,
+      "step": 1167
+    },
+    {
+      "epoch": 1.8688,
+      "grad_norm": 0.6992904245094892,
+      "learning_rate": 3.7516811207395116e-06,
+      "loss": 1.2732,
+      "step": 1168
+    },
+    {
+      "epoch": 1.8704,
+      "grad_norm": 0.6598163110885773,
+      "learning_rate": 3.742666969973463e-06,
+      "loss": 1.343,
+      "step": 1169
+    },
+    {
+      "epoch": 1.8719999999999999,
+      "grad_norm": 0.6264796856692565,
+      "learning_rate": 3.733657179541635e-06,
+      "loss": 1.3174,
+      "step": 1170
+    },
+    {
+      "epoch": 1.8736000000000002,
+      "grad_norm": 0.6330848168521771,
+      "learning_rate": 3.724651780689286e-06,
+      "loss": 1.3019,
+      "step": 1171
+    },
+    {
+      "epoch": 1.8752,
+      "grad_norm": 0.665171636940861,
+      "learning_rate": 3.715650804646449e-06,
+      "loss": 1.2677,
+      "step": 1172
+    },
+    {
+      "epoch": 1.8768,
+      "grad_norm": 0.629132876699752,
+      "learning_rate": 3.7066542826278153e-06,
+      "loss": 1.2814,
+      "step": 1173
+    },
+    {
+      "epoch": 1.8784,
+      "grad_norm": 0.6325547949599222,
+      "learning_rate": 3.6976622458326308e-06,
+      "loss": 1.3375,
+      "step": 1174
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.6781881190715783,
+      "learning_rate": 3.6886747254445877e-06,
+      "loss": 1.3058,
+      "step": 1175
+    },
+    {
+      "epoch": 1.8816000000000002,
+      "grad_norm": 0.5990488091310996,
+      "learning_rate": 3.6796917526317153e-06,
+      "loss": 1.29,
+      "step": 1176
+    },
+    {
+      "epoch": 1.8832,
+      "grad_norm": 0.6408821012219653,
+      "learning_rate": 3.6707133585462713e-06,
+      "loss": 1.3526,
+      "step": 1177
+    },
+    {
+      "epoch": 1.8848,
+      "grad_norm": 0.6617626586072831,
+      "learning_rate": 3.6617395743246375e-06,
+      "loss": 1.3327,
+      "step": 1178
+    },
+    {
+      "epoch": 1.8864,
+      "grad_norm": 0.6088346406101846,
+      "learning_rate": 3.652770431087206e-06,
+      "loss": 1.3086,
+      "step": 1179
+    },
+    {
+      "epoch": 1.888,
+      "grad_norm": 0.6372787326141296,
+      "learning_rate": 3.6438059599382765e-06,
+      "loss": 1.3141,
+      "step": 1180
+    },
+    {
+      "epoch": 1.8896,
+      "grad_norm": 0.6636094535641841,
+      "learning_rate": 3.634846191965944e-06,
+      "loss": 1.2622,
+      "step": 1181
+    },
+    {
+      "epoch": 1.8912,
+      "grad_norm": 0.6281506508011241,
+      "learning_rate": 3.625891158241994e-06,
+      "loss": 1.299,
+      "step": 1182
+    },
+    {
+      "epoch": 1.8928,
+      "grad_norm": 0.6306823952130184,
+      "learning_rate": 3.6169408898217973e-06,
+      "loss": 1.3001,
+      "step": 1183
+    },
+    {
+      "epoch": 1.8944,
+      "grad_norm": 0.6183015040722184,
+      "learning_rate": 3.6079954177441945e-06,
+      "loss": 1.3292,
+      "step": 1184
+    },
+    {
+      "epoch": 1.896,
+      "grad_norm": 0.6341225964752806,
+      "learning_rate": 3.599054773031394e-06,
+      "loss": 1.3602,
+      "step": 1185
+    },
+    {
+      "epoch": 1.8976,
+      "grad_norm": 0.6582568320297879,
+      "learning_rate": 3.5901189866888654e-06,
+      "loss": 1.2527,
+      "step": 1186
+    },
+    {
+      "epoch": 1.8992,
+      "grad_norm": 0.6286638810669035,
+      "learning_rate": 3.581188089705226e-06,
+      "loss": 1.268,
+      "step": 1187
+    },
+    {
+      "epoch": 1.9008,
+      "grad_norm": 0.5909670525821452,
+      "learning_rate": 3.572262113052142e-06,
+      "loss": 1.2769,
+      "step": 1188
+    },
+    {
+      "epoch": 1.9024,
+      "grad_norm": 0.6126787771025278,
+      "learning_rate": 3.563341087684213e-06,
+      "loss": 1.3205,
+      "step": 1189
+    },
+    {
+      "epoch": 1.904,
+      "grad_norm": 0.6265727237392489,
+      "learning_rate": 3.554425044538868e-06,
+      "loss": 1.2862,
+      "step": 1190
+    },
+    {
+      "epoch": 1.9056,
+      "grad_norm": 0.6068139853803607,
+      "learning_rate": 3.5455140145362587e-06,
+      "loss": 1.3321,
+      "step": 1191
+    },
+    {
+      "epoch": 1.9072,
+      "grad_norm": 0.6199201485393414,
+      "learning_rate": 3.5366080285791516e-06,
+      "loss": 1.3064,
+      "step": 1192
+    },
+    {
+      "epoch": 1.9088,
+      "grad_norm": 0.6661489496137575,
+      "learning_rate": 3.527707117552822e-06,
+      "loss": 1.3098,
+      "step": 1193
+    },
+    {
+      "epoch": 1.9104,
+      "grad_norm": 0.569322294451224,
+      "learning_rate": 3.5188113123249435e-06,
+      "loss": 1.3588,
+      "step": 1194
+    },
+    {
+      "epoch": 1.912,
+      "grad_norm": 0.6130321107811649,
+      "learning_rate": 3.5099206437454852e-06,
+      "loss": 1.3304,
+      "step": 1195
+    },
+    {
+      "epoch": 1.9136,
+      "grad_norm": 0.6133318497885671,
+      "learning_rate": 3.5010351426466006e-06,
+      "loss": 1.3307,
+      "step": 1196
+    },
+    {
+      "epoch": 1.9152,
+      "grad_norm": 0.597826912533194,
+      "learning_rate": 3.4921548398425246e-06,
+      "loss": 1.2867,
+      "step": 1197
+    },
+    {
+      "epoch": 1.9167999999999998,
+      "grad_norm": 0.6300356866252554,
+      "learning_rate": 3.4832797661294633e-06,
+      "loss": 1.3089,
+      "step": 1198
+    },
+    {
+      "epoch": 1.9184,
+      "grad_norm": 0.5962886820182318,
+      "learning_rate": 3.4744099522854914e-06,
+      "loss": 1.3179,
+      "step": 1199
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.6247700898715137,
+      "learning_rate": 3.4655454290704393e-06,
+      "loss": 1.3379,
+      "step": 1200
+    },
+    {
+      "epoch": 1.9216,
+      "grad_norm": 0.6598079556346362,
+      "learning_rate": 3.4566862272257923e-06,
+      "loss": 1.3425,
+      "step": 1201
+    },
+    {
+      "epoch": 1.9232,
+      "grad_norm": 0.5972977510327708,
+      "learning_rate": 3.44783237747458e-06,
+      "loss": 1.2955,
+      "step": 1202
+    },
+    {
+      "epoch": 1.9247999999999998,
+      "grad_norm": 0.7769678378965986,
+      "learning_rate": 3.438983910521273e-06,
+      "loss": 1.3699,
+      "step": 1203
+    },
+    {
+      "epoch": 1.9264000000000001,
+      "grad_norm": 0.6311067556397424,
+      "learning_rate": 3.430140857051675e-06,
+      "loss": 1.3163,
+      "step": 1204
+    },
+    {
+      "epoch": 1.928,
+      "grad_norm": 0.671858101923529,
+      "learning_rate": 3.421303247732813e-06,
+      "loss": 1.3491,
+      "step": 1205
+    },
+    {
+      "epoch": 1.9296,
+      "grad_norm": 0.6077265226224458,
+      "learning_rate": 3.4124711132128374e-06,
+      "loss": 1.3046,
+      "step": 1206
+    },
+    {
+      "epoch": 1.9312,
+      "grad_norm": 0.6179109747692213,
+      "learning_rate": 3.4036444841209113e-06,
+      "loss": 1.2342,
+      "step": 1207
+    },
+    {
+      "epoch": 1.9327999999999999,
+      "grad_norm": 0.6149170950043792,
+      "learning_rate": 3.3948233910671036e-06,
+      "loss": 1.3291,
+      "step": 1208
+    },
+    {
+      "epoch": 1.9344000000000001,
+      "grad_norm": 0.6013239927648985,
+      "learning_rate": 3.3860078646422894e-06,
+      "loss": 1.3942,
+      "step": 1209
+    },
+    {
+      "epoch": 1.936,
+      "grad_norm": 0.6406331507491955,
+      "learning_rate": 3.3771979354180343e-06,
+      "loss": 1.345,
+      "step": 1210
+    },
+    {
+      "epoch": 1.9376,
+      "grad_norm": 0.6104568967472018,
+      "learning_rate": 3.3683936339464957e-06,
+      "loss": 1.2755,
+      "step": 1211
+    },
+    {
+      "epoch": 1.9392,
+      "grad_norm": 0.602448049771855,
+      "learning_rate": 3.359594990760313e-06,
+      "loss": 1.2745,
+      "step": 1212
+    },
+    {
+      "epoch": 1.9407999999999999,
+      "grad_norm": 0.6334534932438354,
+      "learning_rate": 3.3508020363725043e-06,
+      "loss": 1.3345,
+      "step": 1213
+    },
+    {
+      "epoch": 1.9424000000000001,
+      "grad_norm": 0.6641236747179002,
+      "learning_rate": 3.34201480127636e-06,
+      "loss": 1.293,
+      "step": 1214
+    },
+    {
+      "epoch": 1.944,
+      "grad_norm": 0.6277382807867748,
+      "learning_rate": 3.333233315945333e-06,
+      "loss": 1.3393,
+      "step": 1215
+    },
+    {
+      "epoch": 1.9456,
+      "grad_norm": 0.6039443052733019,
+      "learning_rate": 3.324457610832942e-06,
+      "loss": 1.2631,
+      "step": 1216
+    },
+    {
+      "epoch": 1.9472,
+      "grad_norm": 0.6208690924269173,
+      "learning_rate": 3.315687716372655e-06,
+      "loss": 1.3419,
+      "step": 1217
+    },
+    {
+      "epoch": 1.9487999999999999,
+      "grad_norm": 0.6277171670054564,
+      "learning_rate": 3.306923662977789e-06,
+      "loss": 1.3382,
+      "step": 1218
+    },
+    {
+      "epoch": 1.9504000000000001,
+      "grad_norm": 0.6277751798708293,
+      "learning_rate": 3.2981654810414128e-06,
+      "loss": 1.3479,
+      "step": 1219
+    },
+    {
+      "epoch": 1.952,
+      "grad_norm": 0.5872626665004217,
+      "learning_rate": 3.2894132009362245e-06,
+      "loss": 1.289,
+      "step": 1220
+    },
+    {
+      "epoch": 1.9536,
+      "grad_norm": 0.6264329770165874,
+      "learning_rate": 3.280666853014457e-06,
+      "loss": 1.3507,
+      "step": 1221
+    },
+    {
+      "epoch": 1.9552,
+      "grad_norm": 0.6782825784320365,
+      "learning_rate": 3.271926467607774e-06,
+      "loss": 1.3606,
+      "step": 1222
+    },
+    {
+      "epoch": 1.9567999999999999,
+      "grad_norm": 0.6143724830135365,
+      "learning_rate": 3.2631920750271594e-06,
+      "loss": 1.3134,
+      "step": 1223
+    },
+    {
+      "epoch": 1.9584000000000001,
+      "grad_norm": 0.6675338173354348,
+      "learning_rate": 3.2544637055628135e-06,
+      "loss": 1.3491,
+      "step": 1224
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.6517148056698626,
+      "learning_rate": 3.2457413894840516e-06,
+      "loss": 1.2581,
+      "step": 1225
+    },
+    {
+      "epoch": 1.9616,
+      "grad_norm": 0.6024569095993069,
+      "learning_rate": 3.2370251570391933e-06,
+      "loss": 1.3304,
+      "step": 1226
+    },
+    {
+      "epoch": 1.9632,
+      "grad_norm": 1.9890498546517974,
+      "learning_rate": 3.2283150384554642e-06,
+      "loss": 1.3441,
+      "step": 1227
+    },
+    {
+      "epoch": 1.9647999999999999,
+      "grad_norm": 0.7296146055528164,
+      "learning_rate": 3.219611063938883e-06,
+      "loss": 1.331,
+      "step": 1228
+    },
+    {
+      "epoch": 1.9664000000000001,
+      "grad_norm": 0.6299668809386478,
+      "learning_rate": 3.210913263674166e-06,
+      "loss": 1.2941,
+      "step": 1229
+    },
+    {
+      "epoch": 1.968,
+      "grad_norm": 0.5876769661849001,
+      "learning_rate": 3.2022216678246145e-06,
+      "loss": 1.3098,
+      "step": 1230
+    },
+    {
+      "epoch": 1.9696,
+      "grad_norm": 0.6939557318747765,
+      "learning_rate": 3.193536306532013e-06,
+      "loss": 1.3074,
+      "step": 1231
+    },
+    {
+      "epoch": 1.9712,
+      "grad_norm": 0.6338426865763995,
+      "learning_rate": 3.184857209916528e-06,
+      "loss": 1.3246,
+      "step": 1232
+    },
+    {
+      "epoch": 1.9727999999999999,
+      "grad_norm": 0.63814151517842,
+      "learning_rate": 3.1761844080765993e-06,
+      "loss": 1.2651,
+      "step": 1233
+    },
+    {
+      "epoch": 1.9744000000000002,
+      "grad_norm": 0.6391071664482869,
+      "learning_rate": 3.1675179310888344e-06,
+      "loss": 1.3197,
+      "step": 1234
+    },
+    {
+      "epoch": 1.976,
+      "grad_norm": 0.6202076209313251,
+      "learning_rate": 3.15885780900791e-06,
+      "loss": 1.3045,
+      "step": 1235
+    },
+    {
+      "epoch": 1.9776,
+      "grad_norm": 0.6975384073164714,
+      "learning_rate": 3.150204071866464e-06,
+      "loss": 1.3012,
+      "step": 1236
+    },
+    {
+      "epoch": 1.9792,
+      "grad_norm": 0.6486041044203757,
+      "learning_rate": 3.141556749674988e-06,
+      "loss": 1.2619,
+      "step": 1237
+    },
+    {
+      "epoch": 1.9808,
+      "grad_norm": 0.5954374562460272,
+      "learning_rate": 3.132915872421734e-06,
+      "loss": 1.2655,
+      "step": 1238
+    },
+    {
+      "epoch": 1.9824000000000002,
+      "grad_norm": 0.6826062255146707,
+      "learning_rate": 3.1242814700725977e-06,
+      "loss": 1.2742,
+      "step": 1239
+    },
+    {
+      "epoch": 1.984,
+      "grad_norm": 0.6473177416655974,
+      "learning_rate": 3.1156535725710224e-06,
+      "loss": 1.2725,
+      "step": 1240
+    },
+    {
+      "epoch": 1.9856,
+      "grad_norm": 0.5923720983735592,
+      "learning_rate": 3.1070322098378925e-06,
+      "loss": 1.3051,
+      "step": 1241
+    },
+    {
+      "epoch": 1.9872,
+      "grad_norm": 0.6818359815911758,
+      "learning_rate": 3.0984174117714306e-06,
+      "loss": 1.2747,
+      "step": 1242
+    },
+    {
+      "epoch": 1.9888,
+      "grad_norm": 0.5970764083240043,
+      "learning_rate": 3.0898092082470943e-06,
+      "loss": 1.3197,
+      "step": 1243
+    },
+    {
+      "epoch": 1.9904,
+      "grad_norm": 0.5961514209423736,
+      "learning_rate": 3.081207629117472e-06,
+      "loss": 1.299,
+      "step": 1244
+    },
+    {
+      "epoch": 1.992,
+      "grad_norm": 0.5957933590992069,
+      "learning_rate": 3.0726127042121766e-06,
+      "loss": 1.4162,
+      "step": 1245
+    },
+    {
+      "epoch": 1.9936,
+      "grad_norm": 0.6251599591120667,
+      "learning_rate": 3.064024463337747e-06,
+      "loss": 1.3509,
+      "step": 1246
+    },
+    {
+      "epoch": 1.9952,
+      "grad_norm": 0.5934767166739614,
+      "learning_rate": 3.0554429362775417e-06,
+      "loss": 1.3145,
+      "step": 1247
+    },
+    {
+      "epoch": 1.9968,
+      "grad_norm": 0.580805796744601,
+      "learning_rate": 3.046868152791638e-06,
+      "loss": 1.2639,
+      "step": 1248
+    },
+    {
+      "epoch": 1.9984,
+      "grad_norm": 0.6366974087302847,
+      "learning_rate": 3.038300142616723e-06,
+      "loss": 1.3354,
+      "step": 1249
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.6200678574564001,
+      "learning_rate": 3.0297389354659984e-06,
+      "loss": 1.2865,
+      "step": 1250
+    },
+    {
+      "epoch": 2.0016,
+      "grad_norm": 0.8471924318972989,
+      "learning_rate": 3.021184561029071e-06,
+      "loss": 1.3162,
+      "step": 1251
+    },
+    {
+      "epoch": 2.0032,
+      "grad_norm": 0.6948521908658133,
+      "learning_rate": 3.0126370489718537e-06,
+      "loss": 1.3137,
+      "step": 1252
+    },
+    {
+      "epoch": 2.0048,
+      "grad_norm": 0.7208464794765871,
+      "learning_rate": 3.0040964289364618e-06,
+      "loss": 1.2607,
+      "step": 1253
+    },
+    {
+      "epoch": 2.0064,
+      "grad_norm": 0.7151298772316619,
+      "learning_rate": 2.9955627305411074e-06,
+      "loss": 1.2194,
+      "step": 1254
+    },
+    {
+      "epoch": 2.008,
+      "grad_norm": 0.6732834157081544,
+      "learning_rate": 2.9870359833799994e-06,
+      "loss": 1.1986,
+      "step": 1255
+    },
+    {
+      "epoch": 2.0096,
+      "grad_norm": 0.6984493979312703,
+      "learning_rate": 2.978516217023243e-06,
+      "loss": 1.2305,
+      "step": 1256
+    },
+    {
+      "epoch": 2.0112,
+      "grad_norm": 0.6811589914097944,
+      "learning_rate": 2.97000346101673e-06,
+      "loss": 1.287,
+      "step": 1257
+    },
+    {
+      "epoch": 2.0128,
+      "grad_norm": 0.7024885217768756,
+      "learning_rate": 2.9614977448820444e-06,
+      "loss": 1.306,
+      "step": 1258
+    },
+    {
+      "epoch": 2.0144,
+      "grad_norm": 0.6718615264499539,
+      "learning_rate": 2.952999098116356e-06,
+      "loss": 1.2859,
+      "step": 1259
+    },
+    {
+      "epoch": 2.016,
+      "grad_norm": 0.7194957467922413,
+      "learning_rate": 2.944507550192318e-06,
+      "loss": 1.2938,
+      "step": 1260
+    },
+    {
+      "epoch": 2.0176,
+      "grad_norm": 0.6848788365939565,
+      "learning_rate": 2.9360231305579645e-06,
+      "loss": 1.2218,
+      "step": 1261
+    },
+    {
+      "epoch": 2.0192,
+      "grad_norm": 0.6783116051322475,
+      "learning_rate": 2.9275458686366108e-06,
+      "loss": 1.3068,
+      "step": 1262
+    },
+    {
+      "epoch": 2.0208,
+      "grad_norm": 0.6724432880909419,
+      "learning_rate": 2.9190757938267477e-06,
+      "loss": 1.2761,
+      "step": 1263
+    },
+    {
+      "epoch": 2.0224,
+      "grad_norm": 0.6552752317545698,
+      "learning_rate": 2.9106129355019464e-06,
+      "loss": 1.2999,
+      "step": 1264
+    },
+    {
+      "epoch": 2.024,
+      "grad_norm": 0.6358125456675352,
+      "learning_rate": 2.9021573230107436e-06,
+      "loss": 1.2787,
+      "step": 1265
+    },
+    {
+      "epoch": 2.0256,
+      "grad_norm": 0.6597714528250314,
+      "learning_rate": 2.8937089856765564e-06,
+      "loss": 1.2061,
+      "step": 1266
+    },
+    {
+      "epoch": 2.0272,
+      "grad_norm": 0.6667260791825,
+      "learning_rate": 2.885267952797569e-06,
+      "loss": 1.3379,
+      "step": 1267
+    },
+    {
+      "epoch": 2.0288,
+      "grad_norm": 0.6117661152924256,
+      "learning_rate": 2.876834253646631e-06,
+      "loss": 1.279,
+      "step": 1268
+    },
+    {
+      "epoch": 2.0304,
+      "grad_norm": 0.638256006361779,
+      "learning_rate": 2.8684079174711665e-06,
+      "loss": 1.2215,
+      "step": 1269
+    },
+    {
+      "epoch": 2.032,
+      "grad_norm": 0.6469849102603812,
+      "learning_rate": 2.8599889734930548e-06,
+      "loss": 1.253,
+      "step": 1270
+    },
+    {
+      "epoch": 2.0336,
+      "grad_norm": 0.5971907983795129,
+      "learning_rate": 2.8515774509085535e-06,
+      "loss": 1.2993,
+      "step": 1271
+    },
+    {
+      "epoch": 2.0352,
+      "grad_norm": 0.6034704603848413,
+      "learning_rate": 2.8431733788881703e-06,
+      "loss": 1.2702,
+      "step": 1272
+    },
+    {
+      "epoch": 2.0368,
+      "grad_norm": 0.6500646128406273,
+      "learning_rate": 2.8347767865765828e-06,
+      "loss": 1.2904,
+      "step": 1273
+    },
+    {
+      "epoch": 2.0384,
+      "grad_norm": 0.6091402764691797,
+      "learning_rate": 2.826387703092528e-06,
+      "loss": 1.256,
+      "step": 1274
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.592715808924595,
+      "learning_rate": 2.8180061575286995e-06,
+      "loss": 1.2259,
+      "step": 1275
+    },
+    {
+      "epoch": 2.0416,
+      "grad_norm": 0.6355460542753687,
+      "learning_rate": 2.8096321789516557e-06,
+      "loss": 1.2535,
+      "step": 1276
+    },
+    {
+      "epoch": 2.0432,
+      "grad_norm": 0.6582124852548594,
+      "learning_rate": 2.8012657964017056e-06,
+      "loss": 1.2373,
+      "step": 1277
+    },
+    {
+      "epoch": 2.0448,
+      "grad_norm": 0.6476498054875318,
+      "learning_rate": 2.792907038892823e-06,
+      "loss": 1.3432,
+      "step": 1278
+    },
+    {
+      "epoch": 2.0464,
+      "grad_norm": 0.6191157695469318,
+      "learning_rate": 2.784555935412538e-06,
+      "loss": 1.2975,
+      "step": 1279
+    },
+    {
+      "epoch": 2.048,
+      "grad_norm": 0.6206643191234933,
+      "learning_rate": 2.77621251492183e-06,
+      "loss": 1.2722,
+      "step": 1280
+    },
+    {
+      "epoch": 2.0496,
+      "grad_norm": 0.6587991587931553,
+      "learning_rate": 2.7678768063550454e-06,
+      "loss": 1.1774,
+      "step": 1281
+    },
+    {
+      "epoch": 2.0512,
+      "grad_norm": 0.6303379084338204,
+      "learning_rate": 2.759548838619774e-06,
+      "loss": 1.2523,
+      "step": 1282
+    },
+    {
+      "epoch": 2.0528,
+      "grad_norm": 0.6373058408759629,
+      "learning_rate": 2.7512286405967726e-06,
+      "loss": 1.2462,
+      "step": 1283
+    },
+    {
+      "epoch": 2.0544,
+      "grad_norm": 0.6355650890762419,
+      "learning_rate": 2.742916241139843e-06,
+      "loss": 1.2794,
+      "step": 1284
+    },
+    {
+      "epoch": 2.056,
+      "grad_norm": 0.5840615741333903,
+      "learning_rate": 2.7346116690757496e-06,
+      "loss": 1.3187,
+      "step": 1285
+    },
+    {
+      "epoch": 2.0576,
+      "grad_norm": 0.6160009448707741,
+      "learning_rate": 2.726314953204111e-06,
+      "loss": 1.2831,
+      "step": 1286
+    },
+    {
+      "epoch": 2.0592,
+      "grad_norm": 0.9517704072527321,
+      "learning_rate": 2.718026122297297e-06,
+      "loss": 1.2971,
+      "step": 1287
+    },
+    {
+      "epoch": 2.0608,
+      "grad_norm": 0.6518604113787492,
+      "learning_rate": 2.7097452051003375e-06,
+      "loss": 1.286,
+      "step": 1288
+    },
+    {
+      "epoch": 2.0624,
+      "grad_norm": 0.6203468812864008,
+      "learning_rate": 2.701472230330813e-06,
+      "loss": 1.1959,
+      "step": 1289
+    },
+    {
+      "epoch": 2.064,
+      "grad_norm": 0.6063328763871871,
+      "learning_rate": 2.693207226678767e-06,
+      "loss": 1.2523,
+      "step": 1290
+    },
+    {
+      "epoch": 2.0656,
+      "grad_norm": 0.6702920781609866,
+      "learning_rate": 2.684950222806596e-06,
+      "loss": 1.2597,
+      "step": 1291
+    },
+    {
+      "epoch": 2.0672,
+      "grad_norm": 0.6321325134963872,
+      "learning_rate": 2.676701247348951e-06,
+      "loss": 1.2571,
+      "step": 1292
+    },
+    {
+      "epoch": 2.0688,
+      "grad_norm": 0.6141593722542322,
+      "learning_rate": 2.6684603289126492e-06,
+      "loss": 1.2764,
+      "step": 1293
+    },
+    {
+      "epoch": 2.0704,
+      "grad_norm": 0.6271923290250505,
+      "learning_rate": 2.660227496076557e-06,
+      "loss": 1.3008,
+      "step": 1294
+    },
+    {
+      "epoch": 2.072,
+      "grad_norm": 0.6048507470598293,
+      "learning_rate": 2.6520027773915075e-06,
+      "loss": 1.2447,
+      "step": 1295
+    },
+    {
+      "epoch": 2.0736,
+      "grad_norm": 0.628375453906248,
+      "learning_rate": 2.643786201380194e-06,
+      "loss": 1.2975,
+      "step": 1296
+    },
+    {
+      "epoch": 2.0752,
+      "grad_norm": 0.5940511382044983,
+      "learning_rate": 2.6355777965370665e-06,
+      "loss": 1.2589,
+      "step": 1297
+    },
+    {
+      "epoch": 2.0768,
+      "grad_norm": 0.6148512071405243,
+      "learning_rate": 2.627377591328245e-06,
+      "loss": 1.2679,
+      "step": 1298
+    },
+    {
+      "epoch": 2.0784,
+      "grad_norm": 0.6387766214803621,
+      "learning_rate": 2.6191856141914074e-06,
+      "loss": 1.333,
+      "step": 1299
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.6078005795789057,
+      "learning_rate": 2.6110018935357005e-06,
+      "loss": 1.2587,
+      "step": 1300
+    },
+    {
+      "epoch": 2.0816,
+      "grad_norm": 0.6146422917970676,
+      "learning_rate": 2.6028264577416418e-06,
+      "loss": 1.2336,
+      "step": 1301
+    },
+    {
+      "epoch": 2.0832,
+      "grad_norm": 0.6593830319570091,
+      "learning_rate": 2.594659335161008e-06,
+      "loss": 1.3113,
+      "step": 1302
+    },
+    {
+      "epoch": 2.0848,
+      "grad_norm": 0.629700901457414,
+      "learning_rate": 2.586500554116757e-06,
+      "loss": 1.2391,
+      "step": 1303
+    },
+    {
+      "epoch": 2.0864,
+      "grad_norm": 0.6260907554240207,
+      "learning_rate": 2.578350142902909e-06,
+      "loss": 1.2923,
+      "step": 1304
+    },
+    {
+      "epoch": 2.088,
+      "grad_norm": 0.6179157451842253,
+      "learning_rate": 2.570208129784466e-06,
+      "loss": 1.2092,
+      "step": 1305
+    },
+    {
+      "epoch": 2.0896,
+      "grad_norm": 0.6381573626173542,
+      "learning_rate": 2.562074542997305e-06,
+      "loss": 1.3037,
+      "step": 1306
+    },
+    {
+      "epoch": 2.0912,
+      "grad_norm": 0.6252980941478117,
+      "learning_rate": 2.5539494107480746e-06,
+      "loss": 1.3038,
+      "step": 1307
+    },
+    {
+      "epoch": 2.0928,
+      "grad_norm": 0.7089224458053554,
+      "learning_rate": 2.545832761214112e-06,
+      "loss": 1.2786,
+      "step": 1308
+    },
+    {
+      "epoch": 2.0944,
+      "grad_norm": 0.6700882391000433,
+      "learning_rate": 2.5377246225433306e-06,
+      "loss": 1.2955,
+      "step": 1309
+    },
+    {
+      "epoch": 2.096,
+      "grad_norm": 0.6784643629444751,
+      "learning_rate": 2.529625022854133e-06,
+      "loss": 1.3162,
+      "step": 1310
+    },
+    {
+      "epoch": 2.0976,
+      "grad_norm": 0.6335661573719359,
+      "learning_rate": 2.5215339902353097e-06,
+      "loss": 1.2849,
+      "step": 1311
+    },
+    {
+      "epoch": 2.0992,
+      "grad_norm": 0.6017064135213758,
+      "learning_rate": 2.513451552745936e-06,
+      "loss": 1.3012,
+      "step": 1312
+    },
+    {
+      "epoch": 2.1008,
+      "grad_norm": 0.6267702045243329,
+      "learning_rate": 2.505377738415286e-06,
+      "loss": 1.3337,
+      "step": 1313
+    },
+    {
+      "epoch": 2.1024,
+      "grad_norm": 0.6665121939834827,
+      "learning_rate": 2.4973125752427243e-06,
+      "loss": 1.2709,
+      "step": 1314
+    },
+    {
+      "epoch": 2.104,
+      "grad_norm": 0.6335816881321071,
+      "learning_rate": 2.4892560911976167e-06,
+      "loss": 1.3126,
+      "step": 1315
+    },
+    {
+      "epoch": 2.1056,
+      "grad_norm": 0.6387538941121612,
+      "learning_rate": 2.481208314219233e-06,
+      "loss": 1.2574,
+      "step": 1316
+    },
+    {
+      "epoch": 2.1072,
+      "grad_norm": 0.6392725702012532,
+      "learning_rate": 2.4731692722166408e-06,
+      "loss": 1.2694,
+      "step": 1317
+    },
+    {
+      "epoch": 2.1088,
+      "grad_norm": 0.6160417309262448,
+      "learning_rate": 2.4651389930686226e-06,
+      "loss": 1.2125,
+      "step": 1318
+    },
+    {
+      "epoch": 2.1104,
+      "grad_norm": 0.6390648314935018,
+      "learning_rate": 2.457117504623565e-06,
+      "loss": 1.2192,
+      "step": 1319
+    },
+    {
+      "epoch": 2.112,
+      "grad_norm": 0.6713378876149517,
+      "learning_rate": 2.4491048346993756e-06,
+      "loss": 1.2783,
+      "step": 1320
+    },
+    {
+      "epoch": 2.1136,
+      "grad_norm": 0.6631473322035947,
+      "learning_rate": 2.4411010110833783e-06,
+      "loss": 1.2292,
+      "step": 1321
+    },
+    {
+      "epoch": 2.1152,
+      "grad_norm": 0.6195981377481148,
+      "learning_rate": 2.433106061532214e-06,
+      "loss": 1.2175,
+      "step": 1322
+    },
+    {
+      "epoch": 2.1168,
+      "grad_norm": 0.611258999812973,
+      "learning_rate": 2.4251200137717545e-06,
+      "loss": 1.2292,
+      "step": 1323
+    },
+    {
+      "epoch": 2.1184,
+      "grad_norm": 0.6588792970158625,
+      "learning_rate": 2.4171428954969995e-06,
+      "loss": 1.2486,
+      "step": 1324
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 0.6237596524260902,
+      "learning_rate": 2.4091747343719828e-06,
+      "loss": 1.2486,
+      "step": 1325
+    },
+    {
+      "epoch": 2.1216,
+      "grad_norm": 0.6732447610480806,
+      "learning_rate": 2.401215558029671e-06,
+      "loss": 1.2489,
+      "step": 1326
+    },
+    {
+      "epoch": 2.1232,
+      "grad_norm": 0.6552954967870803,
+      "learning_rate": 2.3932653940718784e-06,
+      "loss": 1.3237,
+      "step": 1327
+    },
+    {
+      "epoch": 2.1248,
+      "grad_norm": 0.6119614806773075,
+      "learning_rate": 2.3853242700691594e-06,
+      "loss": 1.2824,
+      "step": 1328
+    },
+    {
+      "epoch": 2.1264,
+      "grad_norm": 0.655698809881722,
+      "learning_rate": 2.3773922135607217e-06,
+      "loss": 1.2519,
+      "step": 1329
+    },
+    {
+      "epoch": 2.128,
+      "grad_norm": 0.6715881157007074,
+      "learning_rate": 2.3694692520543293e-06,
+      "loss": 1.2736,
+      "step": 1330
+    },
+    {
+      "epoch": 2.1296,
+      "grad_norm": 0.6409744997268509,
+      "learning_rate": 2.3615554130262003e-06,
+      "loss": 1.3013,
+      "step": 1331
+    },
+    {
+      "epoch": 2.1312,
+      "grad_norm": 0.6061600712739071,
+      "learning_rate": 2.3536507239209223e-06,
+      "loss": 1.1987,
+      "step": 1332
+    },
+    {
+      "epoch": 2.1328,
+      "grad_norm": 0.623286003731786,
+      "learning_rate": 2.3457552121513455e-06,
+      "loss": 1.3172,
+      "step": 1333
+    },
+    {
+      "epoch": 2.1344,
+      "grad_norm": 0.6486170953071254,
+      "learning_rate": 2.337868905098499e-06,
+      "loss": 1.2703,
+      "step": 1334
+    },
+    {
+      "epoch": 2.136,
+      "grad_norm": 0.6079486935638616,
+      "learning_rate": 2.329991830111492e-06,
+      "loss": 1.2912,
+      "step": 1335
+    },
+    {
+      "epoch": 2.1376,
+      "grad_norm": 0.6468973733252784,
+      "learning_rate": 2.32212401450741e-06,
+      "loss": 1.216,
+      "step": 1336
+    },
+    {
+      "epoch": 2.1391999999999998,
+      "grad_norm": 0.6187124789738179,
+      "learning_rate": 2.3142654855712353e-06,
+      "loss": 1.233,
+      "step": 1337
+    },
+    {
+      "epoch": 2.1408,
+      "grad_norm": 0.6245264169413748,
+      "learning_rate": 2.3064162705557387e-06,
+      "loss": 1.215,
+      "step": 1338
+    },
+    {
+      "epoch": 2.1424,
+      "grad_norm": 0.6191651190575526,
+      "learning_rate": 2.2985763966813963e-06,
+      "loss": 1.312,
+      "step": 1339
+    },
+    {
+      "epoch": 2.144,
+      "grad_norm": 0.6376444008020343,
+      "learning_rate": 2.2907458911362885e-06,
+      "loss": 1.2358,
+      "step": 1340
+    },
+    {
+      "epoch": 2.1456,
+      "grad_norm": 0.628562575803481,
+      "learning_rate": 2.2829247810760023e-06,
+      "loss": 1.2225,
+      "step": 1341
+    },
+    {
+      "epoch": 2.1471999999999998,
+      "grad_norm": 0.6056235372374407,
+      "learning_rate": 2.275113093623551e-06,
+      "loss": 1.268,
+      "step": 1342
+    },
+    {
+      "epoch": 2.1488,
+      "grad_norm": 0.6650855789900607,
+      "learning_rate": 2.2673108558692603e-06,
+      "loss": 1.2434,
+      "step": 1343
+    },
+    {
+      "epoch": 2.1504,
+      "grad_norm": 0.6294483366948181,
+      "learning_rate": 2.259518094870693e-06,
+      "loss": 1.2269,
+      "step": 1344
+    },
+    {
+      "epoch": 2.152,
+      "grad_norm": 0.6302774586279425,
+      "learning_rate": 2.251734837652547e-06,
+      "loss": 1.2865,
+      "step": 1345
+    },
+    {
+      "epoch": 2.1536,
+      "grad_norm": 0.6189939263773941,
+      "learning_rate": 2.243961111206555e-06,
+      "loss": 1.2507,
+      "step": 1346
+    },
+    {
+      "epoch": 2.1552,
+      "grad_norm": 0.6199396732557294,
+      "learning_rate": 2.236196942491407e-06,
+      "loss": 1.2672,
+      "step": 1347
+    },
+    {
+      "epoch": 2.1568,
+      "grad_norm": 0.6394446835705898,
+      "learning_rate": 2.228442358432638e-06,
+      "loss": 1.2511,
+      "step": 1348
+    },
+    {
+      "epoch": 2.1584,
+      "grad_norm": 0.6097874579647334,
+      "learning_rate": 2.2206973859225518e-06,
+      "loss": 1.2576,
+      "step": 1349
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.6569624782340872,
+      "learning_rate": 2.2129620518201184e-06,
+      "loss": 1.3024,
+      "step": 1350
+    },
+    {
+      "epoch": 2.1616,
+      "grad_norm": 0.6631037017074078,
+      "learning_rate": 2.2052363829508776e-06,
+      "loss": 1.2231,
+      "step": 1351
+    },
+    {
+      "epoch": 2.1632,
+      "grad_norm": 0.6427703446038863,
+      "learning_rate": 2.1975204061068594e-06,
+      "loss": 1.3604,
+      "step": 1352
+    },
+    {
+      "epoch": 2.1648,
+      "grad_norm": 0.600378747064922,
+      "learning_rate": 2.189814148046473e-06,
+      "loss": 1.276,
+      "step": 1353
+    },
+    {
+      "epoch": 2.1664,
+      "grad_norm": 0.6194499024663324,
+      "learning_rate": 2.182117635494431e-06,
+      "loss": 1.2368,
+      "step": 1354
+    },
+    {
+      "epoch": 2.168,
+      "grad_norm": 0.6487146482041286,
+      "learning_rate": 2.1744308951416483e-06,
+      "loss": 1.3458,
+      "step": 1355
+    },
+    {
+      "epoch": 2.1696,
+      "grad_norm": 0.6386371651048787,
+      "learning_rate": 2.1667539536451455e-06,
+      "loss": 1.229,
+      "step": 1356
+    },
+    {
+      "epoch": 2.1712,
+      "grad_norm": 0.6045124072648208,
+      "learning_rate": 2.1590868376279693e-06,
+      "loss": 1.2396,
+      "step": 1357
+    },
+    {
+      "epoch": 2.1728,
+      "grad_norm": 0.6342298636285454,
+      "learning_rate": 2.151429573679084e-06,
+      "loss": 1.2275,
+      "step": 1358
+    },
+    {
+      "epoch": 2.1744,
+      "grad_norm": 0.6281026400810641,
+      "learning_rate": 2.1437821883532956e-06,
+      "loss": 1.2426,
+      "step": 1359
+    },
+    {
+      "epoch": 2.176,
+      "grad_norm": 0.623915337248547,
+      "learning_rate": 2.136144708171145e-06,
+      "loss": 1.3044,
+      "step": 1360
+    },
+    {
+      "epoch": 2.1776,
+      "grad_norm": 0.5850801342450666,
+      "learning_rate": 2.128517159618827e-06,
+      "loss": 1.2984,
+      "step": 1361
+    },
+    {
+      "epoch": 2.1792,
+      "grad_norm": 0.6326969928348949,
+      "learning_rate": 2.1208995691480947e-06,
+      "loss": 1.2921,
+      "step": 1362
+    },
+    {
+      "epoch": 2.1808,
+      "grad_norm": 0.5995144024942897,
+      "learning_rate": 2.1132919631761637e-06,
+      "loss": 1.2627,
+      "step": 1363
+    },
+    {
+      "epoch": 2.1824,
+      "grad_norm": 0.5733693880869144,
+      "learning_rate": 2.1056943680856286e-06,
+      "loss": 1.2461,
+      "step": 1364
+    },
+    {
+      "epoch": 2.184,
+      "grad_norm": 0.5897946623538239,
+      "learning_rate": 2.098106810224362e-06,
+      "loss": 1.2605,
+      "step": 1365
+    },
+    {
+      "epoch": 2.1856,
+      "grad_norm": 0.6272913941408451,
+      "learning_rate": 2.0905293159054315e-06,
+      "loss": 1.2149,
+      "step": 1366
+    },
+    {
+      "epoch": 2.1872,
+      "grad_norm": 0.6229467008741375,
+      "learning_rate": 2.0829619114070068e-06,
+      "loss": 1.2766,
+      "step": 1367
+    },
+    {
+      "epoch": 2.1888,
+      "grad_norm": 0.5870504981716799,
+      "learning_rate": 2.075404622972261e-06,
+      "loss": 1.3097,
+      "step": 1368
+    },
+    {
+      "epoch": 2.1904,
+      "grad_norm": 0.6237020563591745,
+      "learning_rate": 2.0678574768092926e-06,
+      "loss": 1.336,
+      "step": 1369
+    },
+    {
+      "epoch": 2.192,
+      "grad_norm": 0.6315261452955289,
+      "learning_rate": 2.0603204990910195e-06,
+      "loss": 1.2936,
+      "step": 1370
+    },
+    {
+      "epoch": 2.1936,
+      "grad_norm": 0.5896687937856695,
+      "learning_rate": 2.0527937159551044e-06,
+      "loss": 1.2411,
+      "step": 1371
+    },
+    {
+      "epoch": 2.1952,
+      "grad_norm": 0.6072975790629751,
+      "learning_rate": 2.0452771535038518e-06,
+      "loss": 1.2628,
+      "step": 1372
+    },
+    {
+      "epoch": 2.1968,
+      "grad_norm": 0.5864472780851144,
+      "learning_rate": 2.03777083780412e-06,
+      "loss": 1.242,
+      "step": 1373
+    },
+    {
+      "epoch": 2.1984,
+      "grad_norm": 0.6046350705090998,
+      "learning_rate": 2.030274794887237e-06,
+      "loss": 1.3375,
+      "step": 1374
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 0.5906835440590078,
+      "learning_rate": 2.0227890507488993e-06,
+      "loss": 1.2809,
+      "step": 1375
+    },
+    {
+      "epoch": 2.2016,
+      "grad_norm": 0.6052316630660587,
+      "learning_rate": 2.0153136313490945e-06,
+      "loss": 1.3129,
+      "step": 1376
+    },
+    {
+      "epoch": 2.2032,
+      "grad_norm": 0.617854728052535,
+      "learning_rate": 2.0078485626120015e-06,
+      "loss": 1.2911,
+      "step": 1377
+    },
+    {
+      "epoch": 2.2048,
+      "grad_norm": 0.6039036623553512,
+      "learning_rate": 2.000393870425904e-06,
+      "loss": 1.2849,
+      "step": 1378
+    },
+    {
+      "epoch": 2.2064,
+      "grad_norm": 0.6079474478070289,
+      "learning_rate": 1.9929495806431024e-06,
+      "loss": 1.27,
+      "step": 1379
+    },
+    {
+      "epoch": 2.208,
+      "grad_norm": 0.5933819921603732,
+      "learning_rate": 1.985515719079819e-06,
+      "loss": 1.2254,
+      "step": 1380
+    },
+    {
+      "epoch": 2.2096,
+      "grad_norm": 0.5951715222552574,
+      "learning_rate": 1.978092311516116e-06,
+      "loss": 1.2811,
+      "step": 1381
+    },
+    {
+      "epoch": 2.2112,
+      "grad_norm": 0.6003055736312115,
+      "learning_rate": 1.9706793836957964e-06,
+      "loss": 1.2351,
+      "step": 1382
+    },
+    {
+      "epoch": 2.2128,
+      "grad_norm": 0.6041285863567887,
+      "learning_rate": 1.963276961326326e-06,
+      "loss": 1.2365,
+      "step": 1383
+    },
+    {
+      "epoch": 2.2144,
+      "grad_norm": 0.6157237623250096,
+      "learning_rate": 1.955885070078737e-06,
+      "loss": 1.3074,
+      "step": 1384
+    },
+    {
+      "epoch": 2.216,
+      "grad_norm": 0.5976543680886747,
+      "learning_rate": 1.948503735587537e-06,
+      "loss": 1.3292,
+      "step": 1385
+    },
+    {
+      "epoch": 2.2176,
+      "grad_norm": 0.596770325416233,
+      "learning_rate": 1.9411329834506286e-06,
+      "loss": 1.3345,
+      "step": 1386
+    },
+    {
+      "epoch": 2.2192,
+      "grad_norm": 0.5938058985522667,
+      "learning_rate": 1.9337728392292104e-06,
+      "loss": 1.2959,
+      "step": 1387
+    },
+    {
+      "epoch": 2.2208,
+      "grad_norm": 0.6085575623447018,
+      "learning_rate": 1.926423328447698e-06,
+      "loss": 1.2381,
+      "step": 1388
+    },
+    {
+      "epoch": 2.2224,
+      "grad_norm": 0.5973141296672015,
+      "learning_rate": 1.919084476593631e-06,
+      "loss": 1.3219,
+      "step": 1389
+    },
+    {
+      "epoch": 2.224,
+      "grad_norm": 0.6010869415434419,
+      "learning_rate": 1.9117563091175795e-06,
+      "loss": 1.3334,
+      "step": 1390
+    },
+    {
+      "epoch": 2.2256,
+      "grad_norm": 0.5835608850864904,
+      "learning_rate": 1.904438851433068e-06,
+      "loss": 1.2581,
+      "step": 1391
+    },
+    {
+      "epoch": 2.2272,
+      "grad_norm": 0.6065481924895109,
+      "learning_rate": 1.897132128916474e-06,
+      "loss": 1.3181,
+      "step": 1392
+    },
+    {
+      "epoch": 2.2288,
+      "grad_norm": 0.6033643663870863,
+      "learning_rate": 1.8898361669069497e-06,
+      "loss": 1.2974,
+      "step": 1393
+    },
+    {
+      "epoch": 2.2304,
+      "grad_norm": 0.626407761860978,
+      "learning_rate": 1.8825509907063328e-06,
+      "loss": 1.248,
+      "step": 1394
+    },
+    {
+      "epoch": 2.232,
+      "grad_norm": 0.5919306562261474,
+      "learning_rate": 1.87527662557905e-06,
+      "loss": 1.2869,
+      "step": 1395
+    },
+    {
+      "epoch": 2.2336,
+      "grad_norm": 0.5933156214757883,
+      "learning_rate": 1.8680130967520433e-06,
+      "loss": 1.2965,
+      "step": 1396
+    },
+    {
+      "epoch": 2.2352,
+      "grad_norm": 0.656781514607513,
+      "learning_rate": 1.8607604294146685e-06,
+      "loss": 1.253,
+      "step": 1397
+    },
+    {
+      "epoch": 2.2368,
+      "grad_norm": 0.9332680543767776,
+      "learning_rate": 1.8535186487186213e-06,
+      "loss": 1.2838,
+      "step": 1398
+    },
+    {
+      "epoch": 2.2384,
+      "grad_norm": 0.6007551376736266,
+      "learning_rate": 1.8462877797778367e-06,
+      "loss": 1.31,
+      "step": 1399
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 0.576561305779861,
+      "learning_rate": 1.8390678476684143e-06,
+      "loss": 1.2817,
+      "step": 1400
+    },
+    {
+      "epoch": 2.2416,
+      "grad_norm": 0.5842420492249342,
+      "learning_rate": 1.831858877428524e-06,
+      "loss": 1.2414,
+      "step": 1401
+    },
+    {
+      "epoch": 2.2432,
+      "grad_norm": 0.6020069907784787,
+      "learning_rate": 1.8246608940583166e-06,
+      "loss": 1.2606,
+      "step": 1402
+    },
+    {
+      "epoch": 2.2448,
+      "grad_norm": 0.6041132399543367,
+      "learning_rate": 1.8174739225198485e-06,
+      "loss": 1.2978,
+      "step": 1403
+    },
+    {
+      "epoch": 2.2464,
+      "grad_norm": 0.5769808559569037,
+      "learning_rate": 1.8102979877369808e-06,
+      "loss": 1.2755,
+      "step": 1404
+    },
+    {
+      "epoch": 2.248,
+      "grad_norm": 0.5919838165556751,
+      "learning_rate": 1.8031331145953047e-06,
+      "loss": 1.3299,
+      "step": 1405
+    },
+    {
+      "epoch": 2.2496,
+      "grad_norm": 0.613177858879072,
+      "learning_rate": 1.7959793279420507e-06,
+      "loss": 1.3072,
+      "step": 1406
+    },
+    {
+      "epoch": 2.2512,
+      "grad_norm": 0.5895948976118959,
+      "learning_rate": 1.7888366525859968e-06,
+      "loss": 1.2485,
+      "step": 1407
+    },
+    {
+      "epoch": 2.2528,
+      "grad_norm": 0.6149703112939133,
+      "learning_rate": 1.781705113297396e-06,
+      "loss": 1.2635,
+      "step": 1408
+    },
+    {
+      "epoch": 2.2544,
+      "grad_norm": 0.6307208055912222,
+      "learning_rate": 1.7745847348078742e-06,
+      "loss": 1.2773,
+      "step": 1409
+    },
+    {
+      "epoch": 2.2560000000000002,
+      "grad_norm": 0.5935025136071642,
+      "learning_rate": 1.7674755418103578e-06,
+      "loss": 1.2537,
+      "step": 1410
+    },
+    {
+      "epoch": 2.2576,
+      "grad_norm": 0.5849373850726203,
+      "learning_rate": 1.7603775589589821e-06,
+      "loss": 1.2386,
+      "step": 1411
+    },
+    {
+      "epoch": 2.2592,
+      "grad_norm": 0.5831650766294275,
+      "learning_rate": 1.7532908108690038e-06,
+      "loss": 1.2717,
+      "step": 1412
+    },
+    {
+      "epoch": 2.2608,
+      "grad_norm": 0.5922304359943565,
+      "learning_rate": 1.7462153221167222e-06,
+      "loss": 1.312,
+      "step": 1413
+    },
+    {
+      "epoch": 2.2624,
+      "grad_norm": 0.6047253864922058,
+      "learning_rate": 1.7391511172393849e-06,
+      "loss": 1.2678,
+      "step": 1414
+    },
+    {
+      "epoch": 2.2640000000000002,
+      "grad_norm": 0.5859189847084657,
+      "learning_rate": 1.7320982207351128e-06,
+      "loss": 1.2544,
+      "step": 1415
+    },
+    {
+      "epoch": 2.2656,
+      "grad_norm": 0.5858171686917456,
+      "learning_rate": 1.7250566570628103e-06,
+      "loss": 1.2743,
+      "step": 1416
+    },
+    {
+      "epoch": 2.2672,
+      "grad_norm": 0.5957103928439663,
+      "learning_rate": 1.7180264506420746e-06,
+      "loss": 1.2714,
+      "step": 1417
+    },
+    {
+      "epoch": 2.2688,
+      "grad_norm": 0.6060514081115188,
+      "learning_rate": 1.7110076258531244e-06,
+      "loss": 1.2767,
+      "step": 1418
+    },
+    {
+      "epoch": 2.2704,
+      "grad_norm": 0.5900840599437479,
+      "learning_rate": 1.7040002070367006e-06,
+      "loss": 1.2342,
+      "step": 1419
+    },
+    {
+      "epoch": 2.2720000000000002,
+      "grad_norm": 0.5964456927892774,
+      "learning_rate": 1.6970042184939943e-06,
+      "loss": 1.225,
+      "step": 1420
+    },
+    {
+      "epoch": 2.2736,
+      "grad_norm": 0.6064086324974391,
+      "learning_rate": 1.6900196844865575e-06,
+      "loss": 1.3169,
+      "step": 1421
+    },
+    {
+      "epoch": 2.2752,
+      "grad_norm": 0.611372321023836,
+      "learning_rate": 1.683046629236213e-06,
+      "loss": 1.2067,
+      "step": 1422
+    },
+    {
+      "epoch": 2.2768,
+      "grad_norm": 0.6607542915282189,
+      "learning_rate": 1.6760850769249837e-06,
+      "loss": 1.3387,
+      "step": 1423
+    },
+    {
+      "epoch": 2.2784,
+      "grad_norm": 0.597216513932068,
+      "learning_rate": 1.669135051694994e-06,
+      "loss": 1.2507,
+      "step": 1424
+    },
+    {
+      "epoch": 2.2800000000000002,
+      "grad_norm": 0.6073226221249906,
+      "learning_rate": 1.662196577648398e-06,
+      "loss": 1.3253,
+      "step": 1425
+    },
+    {
+      "epoch": 2.2816,
+      "grad_norm": 0.5991564460662722,
+      "learning_rate": 1.6552696788472921e-06,
+      "loss": 1.2721,
+      "step": 1426
+    },
+    {
+      "epoch": 2.2832,
+      "grad_norm": 0.5916265111713015,
+      "learning_rate": 1.6483543793136247e-06,
+      "loss": 1.2082,
+      "step": 1427
+    },
+    {
+      "epoch": 2.2848,
+      "grad_norm": 0.5947762564301382,
+      "learning_rate": 1.6414507030291249e-06,
+      "loss": 1.3363,
+      "step": 1428
+    },
+    {
+      "epoch": 2.2864,
+      "grad_norm": 0.5807301396037421,
+      "learning_rate": 1.6345586739352105e-06,
+      "loss": 1.2359,
+      "step": 1429
+    },
+    {
+      "epoch": 2.288,
+      "grad_norm": 0.6366752560521612,
+      "learning_rate": 1.6276783159329095e-06,
+      "loss": 1.2772,
+      "step": 1430
+    },
+    {
+      "epoch": 2.2896,
+      "grad_norm": 0.5896791902794976,
+      "learning_rate": 1.6208096528827717e-06,
+      "loss": 1.2979,
+      "step": 1431
+    },
+    {
+      "epoch": 2.2912,
+      "grad_norm": 0.5671592861021616,
+      "learning_rate": 1.6139527086047929e-06,
+      "loss": 1.2086,
+      "step": 1432
+    },
+    {
+      "epoch": 2.2928,
+      "grad_norm": 0.585184215316201,
+      "learning_rate": 1.6071075068783303e-06,
+      "loss": 1.257,
+      "step": 1433
+    },
+    {
+      "epoch": 2.2944,
+      "grad_norm": 0.5805256925092802,
+      "learning_rate": 1.600274071442014e-06,
+      "loss": 1.2457,
+      "step": 1434
+    },
+    {
+      "epoch": 2.296,
+      "grad_norm": 0.586120372735018,
+      "learning_rate": 1.5934524259936757e-06,
+      "loss": 1.1949,
+      "step": 1435
+    },
+    {
+      "epoch": 2.2976,
+      "grad_norm": 0.5834966102642732,
+      "learning_rate": 1.5866425941902524e-06,
+      "loss": 1.2765,
+      "step": 1436
+    },
+    {
+      "epoch": 2.2992,
+      "grad_norm": 0.5802077624018527,
+      "learning_rate": 1.5798445996477219e-06,
+      "loss": 1.2015,
+      "step": 1437
+    },
+    {
+      "epoch": 2.3008,
+      "grad_norm": 0.5739329907227468,
+      "learning_rate": 1.573058465941002e-06,
+      "loss": 1.2297,
+      "step": 1438
+    },
+    {
+      "epoch": 2.3024,
+      "grad_norm": 0.5930254892448077,
+      "learning_rate": 1.5662842166038844e-06,
+      "loss": 1.2491,
+      "step": 1439
+    },
+    {
+      "epoch": 2.304,
+      "grad_norm": 0.5801445584936561,
+      "learning_rate": 1.5595218751289465e-06,
+      "loss": 1.2512,
+      "step": 1440
+    },
+    {
+      "epoch": 2.3056,
+      "grad_norm": 0.5814098810789083,
+      "learning_rate": 1.5527714649674641e-06,
+      "loss": 1.2807,
+      "step": 1441
+    },
+    {
+      "epoch": 2.3072,
+      "grad_norm": 0.5902236544112721,
+      "learning_rate": 1.5460330095293447e-06,
+      "loss": 1.2481,
+      "step": 1442
+    },
+    {
+      "epoch": 2.3088,
+      "grad_norm": 0.5968048258151835,
+      "learning_rate": 1.5393065321830292e-06,
+      "loss": 1.2421,
+      "step": 1443
+    },
+    {
+      "epoch": 2.3104,
+      "grad_norm": 0.5804530862180891,
+      "learning_rate": 1.5325920562554259e-06,
+      "loss": 1.2288,
+      "step": 1444
+    },
+    {
+      "epoch": 2.312,
+      "grad_norm": 0.6153118581827133,
+      "learning_rate": 1.5258896050318217e-06,
+      "loss": 1.2374,
+      "step": 1445
+    },
+    {
+      "epoch": 2.3136,
+      "grad_norm": 0.5963993336230258,
+      "learning_rate": 1.5191992017557994e-06,
+      "loss": 1.2702,
+      "step": 1446
+    },
+    {
+      "epoch": 2.3152,
+      "grad_norm": 0.6281557815178028,
+      "learning_rate": 1.512520869629165e-06,
+      "loss": 1.2416,
+      "step": 1447
+    },
+    {
+      "epoch": 2.3168,
+      "grad_norm": 0.5824688278780765,
+      "learning_rate": 1.5058546318118583e-06,
+      "loss": 1.2354,
+      "step": 1448
+    },
+    {
+      "epoch": 2.3184,
+      "grad_norm": 0.5921319594259772,
+      "learning_rate": 1.4992005114218805e-06,
+      "loss": 1.2358,
+      "step": 1449
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 0.5799783412842386,
+      "learning_rate": 1.4925585315352108e-06,
+      "loss": 1.2628,
+      "step": 1450
+    },
+    {
+      "epoch": 2.3216,
+      "grad_norm": 0.5945867471050821,
+      "learning_rate": 1.485928715185721e-06,
+      "loss": 1.2378,
+      "step": 1451
+    },
+    {
+      "epoch": 2.3232,
+      "grad_norm": 0.5805492572296385,
+      "learning_rate": 1.4793110853651077e-06,
+      "loss": 1.2124,
+      "step": 1452
+    },
+    {
+      "epoch": 2.3247999999999998,
+      "grad_norm": 0.5817250656001614,
+      "learning_rate": 1.472705665022799e-06,
+      "loss": 1.273,
+      "step": 1453
+    },
+    {
+      "epoch": 2.3264,
+      "grad_norm": 0.5939855546222484,
+      "learning_rate": 1.4661124770658857e-06,
+      "loss": 1.3103,
+      "step": 1454
+    },
+    {
+      "epoch": 2.328,
+      "grad_norm": 0.5999908254792997,
+      "learning_rate": 1.459531544359038e-06,
+      "loss": 1.2864,
+      "step": 1455
+    },
+    {
+      "epoch": 2.3296,
+      "grad_norm": 0.6369816795001034,
+      "learning_rate": 1.4529628897244214e-06,
+      "loss": 1.2257,
+      "step": 1456
+    },
+    {
+      "epoch": 2.3312,
+      "grad_norm": 0.5891544164345975,
+      "learning_rate": 1.4464065359416274e-06,
+      "loss": 1.2269,
+      "step": 1457
+    },
+    {
+      "epoch": 2.3327999999999998,
+      "grad_norm": 0.6104621744526847,
+      "learning_rate": 1.4398625057475845e-06,
+      "loss": 1.3267,
+      "step": 1458
+    },
+    {
+      "epoch": 2.3344,
+      "grad_norm": 0.5848199984520441,
+      "learning_rate": 1.4333308218364861e-06,
+      "loss": 1.2429,
+      "step": 1459
+    },
+    {
+      "epoch": 2.336,
+      "grad_norm": 0.5876181838044481,
+      "learning_rate": 1.4268115068597122e-06,
+      "loss": 1.2707,
+      "step": 1460
+    },
+    {
+      "epoch": 2.3376,
+      "grad_norm": 0.5827759615842357,
+      "learning_rate": 1.4203045834257418e-06,
+      "loss": 1.306,
+      "step": 1461
+    },
+    {
+      "epoch": 2.3392,
+      "grad_norm": 0.6003049743114269,
+      "learning_rate": 1.4138100741000888e-06,
+      "loss": 1.2856,
+      "step": 1462
+    },
+    {
+      "epoch": 2.3407999999999998,
+      "grad_norm": 0.6206564355597641,
+      "learning_rate": 1.4073280014052077e-06,
+      "loss": 1.3405,
+      "step": 1463
+    },
+    {
+      "epoch": 2.3424,
+      "grad_norm": 2.0242812600403397,
+      "learning_rate": 1.4008583878204297e-06,
+      "loss": 1.2564,
+      "step": 1464
+    },
+    {
+      "epoch": 2.344,
+      "grad_norm": 0.5887198234505773,
+      "learning_rate": 1.3944012557818793e-06,
+      "loss": 1.2654,
+      "step": 1465
+    },
+    {
+      "epoch": 2.3456,
+      "grad_norm": 0.5809147456062125,
+      "learning_rate": 1.3879566276823896e-06,
+      "loss": 1.2919,
+      "step": 1466
+    },
+    {
+      "epoch": 2.3472,
+      "grad_norm": 0.6048064684804156,
+      "learning_rate": 1.3815245258714393e-06,
+      "loss": 1.2145,
+      "step": 1467
+    },
+    {
+      "epoch": 2.3487999999999998,
+      "grad_norm": 0.5907655163704777,
+      "learning_rate": 1.3751049726550587e-06,
+      "loss": 1.2583,
+      "step": 1468
+    },
+    {
+      "epoch": 2.3504,
+      "grad_norm": 0.599906529538663,
+      "learning_rate": 1.368697990295766e-06,
+      "loss": 1.2412,
+      "step": 1469
+    },
+    {
+      "epoch": 2.352,
+      "grad_norm": 0.6192543108809913,
+      "learning_rate": 1.3623036010124845e-06,
+      "loss": 1.274,
+      "step": 1470
+    },
+    {
+      "epoch": 2.3536,
+      "grad_norm": 0.6041162236463318,
+      "learning_rate": 1.3559218269804624e-06,
+      "loss": 1.2107,
+      "step": 1471
+    },
+    {
+      "epoch": 2.3552,
+      "grad_norm": 0.6295795087128394,
+      "learning_rate": 1.3495526903312029e-06,
+      "loss": 1.3017,
+      "step": 1472
+    },
+    {
+      "epoch": 2.3568,
+      "grad_norm": 0.6130685107401797,
+      "learning_rate": 1.3431962131523796e-06,
+      "loss": 1.2958,
+      "step": 1473
+    },
+    {
+      "epoch": 2.3584,
+      "grad_norm": 0.5981815021099564,
+      "learning_rate": 1.3368524174877679e-06,
+      "loss": 1.257,
+      "step": 1474
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 0.573145009398753,
+      "learning_rate": 1.330521325337164e-06,
+      "loss": 1.3095,
+      "step": 1475
+    },
+    {
+      "epoch": 2.3616,
+      "grad_norm": 0.5960233966329017,
+      "learning_rate": 1.3242029586563054e-06,
+      "loss": 1.2925,
+      "step": 1476
+    },
+    {
+      "epoch": 2.3632,
+      "grad_norm": 0.5890912794633566,
+      "learning_rate": 1.3178973393568055e-06,
+      "loss": 1.3218,
+      "step": 1477
+    },
+    {
+      "epoch": 2.3648,
+      "grad_norm": 0.6159890204978258,
+      "learning_rate": 1.3116044893060637e-06,
+      "loss": 1.3089,
+      "step": 1478
+    },
+    {
+      "epoch": 2.3664,
+      "grad_norm": 0.5924389723923869,
+      "learning_rate": 1.3053244303272022e-06,
+      "loss": 1.2549,
+      "step": 1479
+    },
+    {
+      "epoch": 2.368,
+      "grad_norm": 0.5801453251709701,
+      "learning_rate": 1.2990571841989796e-06,
+      "loss": 1.219,
+      "step": 1480
+    },
+    {
+      "epoch": 2.3696,
+      "grad_norm": 0.6044729675826989,
+      "learning_rate": 1.2928027726557257e-06,
+      "loss": 1.29,
+      "step": 1481
+    },
+    {
+      "epoch": 2.3712,
+      "grad_norm": 0.5945345497476346,
+      "learning_rate": 1.2865612173872577e-06,
+      "loss": 1.277,
+      "step": 1482
+    },
+    {
+      "epoch": 2.3728,
+      "grad_norm": 0.5755250724719723,
+      "learning_rate": 1.2803325400388095e-06,
+      "loss": 1.2759,
+      "step": 1483
+    },
+    {
+      "epoch": 2.3744,
+      "grad_norm": 0.5829640662425055,
+      "learning_rate": 1.2741167622109557e-06,
+      "loss": 1.2721,
+      "step": 1484
+    },
+    {
+      "epoch": 2.376,
+      "grad_norm": 0.5831858767096839,
+      "learning_rate": 1.2679139054595335e-06,
+      "loss": 1.2944,
+      "step": 1485
+    },
+    {
+      "epoch": 2.3776,
+      "grad_norm": 0.5744519630728254,
+      "learning_rate": 1.261723991295576e-06,
+      "loss": 1.2532,
+      "step": 1486
+    },
+    {
+      "epoch": 2.3792,
+      "grad_norm": 0.6315002760686891,
+      "learning_rate": 1.2555470411852262e-06,
+      "loss": 1.2712,
+      "step": 1487
+    },
+    {
+      "epoch": 2.3808,
+      "grad_norm": 0.5987416912920919,
+      "learning_rate": 1.2493830765496724e-06,
+      "loss": 1.2883,
+      "step": 1488
+    },
+    {
+      "epoch": 2.3824,
+      "grad_norm": 0.5999875484870067,
+      "learning_rate": 1.2432321187650726e-06,
+      "loss": 1.24,
+      "step": 1489
+    },
+    {
+      "epoch": 2.384,
+      "grad_norm": 0.5701645084323782,
+      "learning_rate": 1.237094189162471e-06,
+      "loss": 1.2444,
+      "step": 1490
+    },
+    {
+      "epoch": 2.3856,
+      "grad_norm": 0.5920904701945038,
+      "learning_rate": 1.2309693090277392e-06,
+      "loss": 1.2285,
+      "step": 1491
+    },
+    {
+      "epoch": 2.3872,
+      "grad_norm": 0.5614829108339787,
+      "learning_rate": 1.2248574996014872e-06,
+      "loss": 1.3082,
+      "step": 1492
+    },
+    {
+      "epoch": 2.3888,
+      "grad_norm": 0.5952290550168348,
+      "learning_rate": 1.218758782079001e-06,
+      "loss": 1.2869,
+      "step": 1493
+    },
+    {
+      "epoch": 2.3904,
+      "grad_norm": 0.572936584565802,
+      "learning_rate": 1.2126731776101657e-06,
+      "loss": 1.2515,
+      "step": 1494
+    },
+    {
+      "epoch": 2.392,
+      "grad_norm": 0.5994452771041506,
+      "learning_rate": 1.2066007072993856e-06,
+      "loss": 1.2951,
+      "step": 1495
+    },
+    {
+      "epoch": 2.3936,
+      "grad_norm": 0.5752311288846104,
+      "learning_rate": 1.2005413922055247e-06,
+      "loss": 1.2379,
+      "step": 1496
+    },
+    {
+      "epoch": 2.3952,
+      "grad_norm": 0.5847348180009007,
+      "learning_rate": 1.194495253341818e-06,
+      "loss": 1.2464,
+      "step": 1497
+    },
+    {
+      "epoch": 2.3968,
+      "grad_norm": 0.5701172501071853,
+      "learning_rate": 1.1884623116758121e-06,
+      "loss": 1.3174,
+      "step": 1498
+    },
+    {
+      "epoch": 2.3984,
+      "grad_norm": 0.5877183041821828,
+      "learning_rate": 1.1824425881292846e-06,
+      "loss": 1.2704,
+      "step": 1499
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.6015316386367004,
+      "learning_rate": 1.1764361035781718e-06,
+      "loss": 1.2449,
+      "step": 1500
+    },
+    {
+      "epoch": 2.4016,
+      "grad_norm": 0.5816822662033109,
+      "learning_rate": 1.170442878852503e-06,
+      "loss": 1.2644,
+      "step": 1501
+    },
+    {
+      "epoch": 2.4032,
+      "grad_norm": 0.5844200422828078,
+      "learning_rate": 1.1644629347363173e-06,
+      "loss": 1.2324,
+      "step": 1502
+    },
+    {
+      "epoch": 2.4048,
+      "grad_norm": 0.5836941529145264,
+      "learning_rate": 1.1584962919676024e-06,
+      "loss": 1.2513,
+      "step": 1503
+    },
+    {
+      "epoch": 2.4064,
+      "grad_norm": 0.5685914320901377,
+      "learning_rate": 1.1525429712382175e-06,
+      "loss": 1.1775,
+      "step": 1504
+    },
+    {
+      "epoch": 2.408,
+      "grad_norm": 0.5739668319772913,
+      "learning_rate": 1.1466029931938182e-06,
+      "loss": 1.3201,
+      "step": 1505
+    },
+    {
+      "epoch": 2.4096,
+      "grad_norm": 0.5858432478082448,
+      "learning_rate": 1.1406763784337948e-06,
+      "loss": 1.2391,
+      "step": 1506
+    },
+    {
+      "epoch": 2.4112,
+      "grad_norm": 0.5798756292994779,
+      "learning_rate": 1.1347631475111882e-06,
+      "loss": 1.2441,
+      "step": 1507
+    },
+    {
+      "epoch": 2.4128,
+      "grad_norm": 0.5826366664379028,
+      "learning_rate": 1.1288633209326288e-06,
+      "loss": 1.2708,
+      "step": 1508
+    },
+    {
+      "epoch": 2.4144,
+      "grad_norm": 0.6150382821598519,
+      "learning_rate": 1.122976919158264e-06,
+      "loss": 1.2574,
+      "step": 1509
+    },
+    {
+      "epoch": 2.416,
+      "grad_norm": 0.5811502329434302,
+      "learning_rate": 1.1171039626016789e-06,
+      "loss": 1.262,
+      "step": 1510
+    },
+    {
+      "epoch": 2.4176,
+      "grad_norm": 0.6007506611007667,
+      "learning_rate": 1.1112444716298381e-06,
+      "loss": 1.305,
+      "step": 1511
+    },
+    {
+      "epoch": 2.4192,
+      "grad_norm": 1.902619016618781,
+      "learning_rate": 1.1053984665630025e-06,
+      "loss": 1.3083,
+      "step": 1512
+    },
+    {
+      "epoch": 2.4208,
+      "grad_norm": 0.586907916326899,
+      "learning_rate": 1.0995659676746706e-06,
+      "loss": 1.2233,
+      "step": 1513
+    },
+    {
+      "epoch": 2.4224,
+      "grad_norm": 0.5886511547422887,
+      "learning_rate": 1.093746995191497e-06,
+      "loss": 1.2424,
+      "step": 1514
+    },
+    {
+      "epoch": 2.424,
+      "grad_norm": 0.5723103308376278,
+      "learning_rate": 1.0879415692932328e-06,
+      "loss": 1.2698,
+      "step": 1515
+    },
+    {
+      "epoch": 2.4256,
+      "grad_norm": 0.5923006482157545,
+      "learning_rate": 1.0821497101126487e-06,
+      "loss": 1.3247,
+      "step": 1516
+    },
+    {
+      "epoch": 2.4272,
+      "grad_norm": 0.5776924978411566,
+      "learning_rate": 1.076371437735465e-06,
+      "loss": 1.2133,
+      "step": 1517
+    },
+    {
+      "epoch": 2.4288,
+      "grad_norm": 0.6026993233141199,
+      "learning_rate": 1.0706067722002877e-06,
+      "loss": 1.2246,
+      "step": 1518
+    },
+    {
+      "epoch": 2.4304,
+      "grad_norm": 0.5854287427238086,
+      "learning_rate": 1.064855733498531e-06,
+      "loss": 1.2955,
+      "step": 1519
+    },
+    {
+      "epoch": 2.432,
+      "grad_norm": 0.578216485438422,
+      "learning_rate": 1.0591183415743562e-06,
+      "loss": 1.2821,
+      "step": 1520
+    },
+    {
+      "epoch": 2.4336,
+      "grad_norm": 0.572314298173481,
+      "learning_rate": 1.0533946163245984e-06,
+      "loss": 1.2815,
+      "step": 1521
+    },
+    {
+      "epoch": 2.4352,
+      "grad_norm": 0.5670262740534067,
+      "learning_rate": 1.047684577598694e-06,
+      "loss": 1.2919,
+      "step": 1522
+    },
+    {
+      "epoch": 2.4368,
+      "grad_norm": 0.5925335275343767,
+      "learning_rate": 1.0419882451986197e-06,
+      "loss": 1.249,
+      "step": 1523
+    },
+    {
+      "epoch": 2.4384,
+      "grad_norm": 0.5730834614549865,
+      "learning_rate": 1.0363056388788162e-06,
+      "loss": 1.2228,
+      "step": 1524
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 0.5585212924469906,
+      "learning_rate": 1.0306367783461258e-06,
+      "loss": 1.1778,
+      "step": 1525
+    },
+    {
+      "epoch": 2.4416,
+      "grad_norm": 0.5785630139760682,
+      "learning_rate": 1.024981683259723e-06,
+      "loss": 1.233,
+      "step": 1526
+    },
+    {
+      "epoch": 2.4432,
+      "grad_norm": 0.5783008177381842,
+      "learning_rate": 1.0193403732310392e-06,
+      "loss": 1.2179,
+      "step": 1527
+    },
+    {
+      "epoch": 2.4448,
+      "grad_norm": 0.5661583763283703,
+      "learning_rate": 1.0137128678237062e-06,
+      "loss": 1.2878,
+      "step": 1528
+    },
+    {
+      "epoch": 2.4464,
+      "grad_norm": 0.5694034025323197,
+      "learning_rate": 1.0080991865534773e-06,
+      "loss": 1.2816,
+      "step": 1529
+    },
+    {
+      "epoch": 2.448,
+      "grad_norm": 0.5684583530504835,
+      "learning_rate": 1.002499348888169e-06,
+      "loss": 1.2339,
+      "step": 1530
+    },
+    {
+      "epoch": 2.4496,
+      "grad_norm": 0.5628935413983448,
+      "learning_rate": 9.969133742475883e-07,
+      "loss": 1.2958,
+      "step": 1531
+    },
+    {
+      "epoch": 2.4512,
+      "grad_norm": 0.5715197638142439,
+      "learning_rate": 9.913412820034629e-07,
+      "loss": 1.2529,
+      "step": 1532
+    },
+    {
+      "epoch": 2.4528,
+      "grad_norm": 0.5860068906764546,
+      "learning_rate": 9.857830914793827e-07,
+      "loss": 1.1818,
+      "step": 1533
+    },
+    {
+      "epoch": 2.4544,
+      "grad_norm": 0.5778860400642872,
+      "learning_rate": 9.802388219507215e-07,
+      "loss": 1.2612,
+      "step": 1534
+    },
+    {
+      "epoch": 2.456,
+      "grad_norm": 0.5717055303138866,
+      "learning_rate": 9.747084926445839e-07,
+      "loss": 1.2262,
+      "step": 1535
+    },
+    {
+      "epoch": 2.4576000000000002,
+      "grad_norm": 0.5909897074193656,
+      "learning_rate": 9.691921227397227e-07,
+      "loss": 1.2453,
+      "step": 1536
+    },
+    {
+      "epoch": 2.4592,
+      "grad_norm": 0.590600262668449,
+      "learning_rate": 9.63689731366486e-07,
+      "loss": 1.2448,
+      "step": 1537
+    },
+    {
+      "epoch": 2.4608,
+      "grad_norm": 0.5645667052878954,
+      "learning_rate": 9.58201337606745e-07,
+      "loss": 1.3064,
+      "step": 1538
+    },
+    {
+      "epoch": 2.4624,
+      "grad_norm": 0.5659495125994496,
+      "learning_rate": 9.527269604938249e-07,
+      "loss": 1.25,
+      "step": 1539
+    },
+    {
+      "epoch": 2.464,
+      "grad_norm": 0.6087128882606626,
+      "learning_rate": 9.472666190124457e-07,
+      "loss": 1.2944,
+      "step": 1540
+    },
+    {
+      "epoch": 2.4656000000000002,
+      "grad_norm": 0.570389057833244,
+      "learning_rate": 9.418203320986502e-07,
+      "loss": 1.2382,
+      "step": 1541
+    },
+    {
+      "epoch": 2.4672,
+      "grad_norm": 0.5839826517390007,
+      "learning_rate": 9.363881186397434e-07,
+      "loss": 1.2362,
+      "step": 1542
+    },
+    {
+      "epoch": 2.4688,
+      "grad_norm": 0.575303612521702,
+      "learning_rate": 9.309699974742243e-07,
+      "loss": 1.2901,
+      "step": 1543
+    },
+    {
+      "epoch": 2.4704,
+      "grad_norm": 0.5709857933259976,
+      "learning_rate": 9.255659873917183e-07,
+      "loss": 1.2059,
+      "step": 1544
+    },
+    {
+      "epoch": 2.472,
+      "grad_norm": 0.5771372031547738,
+      "learning_rate": 9.201761071329196e-07,
+      "loss": 1.3263,
+      "step": 1545
+    },
+    {
+      "epoch": 2.4736000000000002,
+      "grad_norm": 0.5748892855287724,
+      "learning_rate": 9.148003753895146e-07,
+      "loss": 1.2935,
+      "step": 1546
+    },
+    {
+      "epoch": 2.4752,
+      "grad_norm": 0.5652460945609818,
+      "learning_rate": 9.094388108041302e-07,
+      "loss": 1.3097,
+      "step": 1547
+    },
+    {
+      "epoch": 2.4768,
+      "grad_norm": 0.5781293040811305,
+      "learning_rate": 9.040914319702598e-07,
+      "loss": 1.3302,
+      "step": 1548
+    },
+    {
+      "epoch": 2.4784,
+      "grad_norm": 0.5616769866106647,
+      "learning_rate": 8.987582574321996e-07,
+      "loss": 1.266,
+      "step": 1549
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 0.5869247211752077,
+      "learning_rate": 8.934393056849921e-07,
+      "loss": 1.3079,
+      "step": 1550
+    },
+    {
+      "epoch": 2.4816,
+      "grad_norm": 0.5785022212454402,
+      "learning_rate": 8.881345951743486e-07,
+      "loss": 1.2961,
+      "step": 1551
+    },
+    {
+      "epoch": 2.4832,
+      "grad_norm": 0.5745864558375826,
+      "learning_rate": 8.828441442966013e-07,
+      "loss": 1.2107,
+      "step": 1552
+    },
+    {
+      "epoch": 2.4848,
+      "grad_norm": 0.5922037163674431,
+      "learning_rate": 8.775679713986235e-07,
+      "loss": 1.2894,
+      "step": 1553
+    },
+    {
+      "epoch": 2.4864,
+      "grad_norm": 0.5817592775435674,
+      "learning_rate": 8.723060947777778e-07,
+      "loss": 1.2685,
+      "step": 1554
+    },
+    {
+      "epoch": 2.488,
+      "grad_norm": 0.5704843672700182,
+      "learning_rate": 8.670585326818493e-07,
+      "loss": 1.205,
+      "step": 1555
+    },
+    {
+      "epoch": 2.4896,
+      "grad_norm": 0.5711539764665575,
+      "learning_rate": 8.618253033089768e-07,
+      "loss": 1.2432,
+      "step": 1556
+    },
+    {
+      "epoch": 2.4912,
+      "grad_norm": 0.5701562009129715,
+      "learning_rate": 8.566064248076001e-07,
+      "loss": 1.2549,
+      "step": 1557
+    },
+    {
+      "epoch": 2.4928,
+      "grad_norm": 0.590838583540458,
+      "learning_rate": 8.514019152763852e-07,
+      "loss": 1.2831,
+      "step": 1558
+    },
+    {
+      "epoch": 2.4944,
+      "grad_norm": 0.5696754189422609,
+      "learning_rate": 8.462117927641733e-07,
+      "loss": 1.3182,
+      "step": 1559
+    },
+    {
+      "epoch": 2.496,
+      "grad_norm": 0.5664268253735034,
+      "learning_rate": 8.410360752699099e-07,
+      "loss": 1.205,
+      "step": 1560
+    },
+    {
+      "epoch": 2.4976,
+      "grad_norm": 0.5683295705996511,
+      "learning_rate": 8.358747807425827e-07,
+      "loss": 1.309,
+      "step": 1561
+    },
+    {
+      "epoch": 2.4992,
+      "grad_norm": 0.5719144930185068,
+      "learning_rate": 8.307279270811675e-07,
+      "loss": 1.2645,
+      "step": 1562
+    },
+    {
+      "epoch": 2.5008,
+      "grad_norm": 0.5681254683021824,
+      "learning_rate": 8.255955321345533e-07,
+      "loss": 1.2708,
+      "step": 1563
+    },
+    {
+      "epoch": 2.5023999999999997,
+      "grad_norm": 0.5589196005341329,
+      "learning_rate": 8.20477613701493e-07,
+      "loss": 1.3096,
+      "step": 1564
+    },
+    {
+      "epoch": 2.504,
+      "grad_norm": 0.5704898081890937,
+      "learning_rate": 8.153741895305351e-07,
+      "loss": 1.2553,
+      "step": 1565
+    },
+    {
+      "epoch": 2.5056000000000003,
+      "grad_norm": 0.5685419762619763,
+      "learning_rate": 8.102852773199588e-07,
+      "loss": 1.2469,
+      "step": 1566
+    },
+    {
+      "epoch": 2.5072,
+      "grad_norm": 0.582518408202899,
+      "learning_rate": 8.052108947177234e-07,
+      "loss": 1.2622,
+      "step": 1567
+    },
+    {
+      "epoch": 2.5088,
+      "grad_norm": 0.5820841718584535,
+      "learning_rate": 8.001510593213946e-07,
+      "loss": 1.2409,
+      "step": 1568
+    },
+    {
+      "epoch": 2.5103999999999997,
+      "grad_norm": 0.5812347477026885,
+      "learning_rate": 7.951057886780939e-07,
+      "loss": 1.325,
+      "step": 1569
+    },
+    {
+      "epoch": 2.512,
+      "grad_norm": 1.1571657119489638,
+      "learning_rate": 7.900751002844326e-07,
+      "loss": 1.2915,
+      "step": 1570
+    },
+    {
+      "epoch": 2.5136,
+      "grad_norm": 0.5754598758131384,
+      "learning_rate": 7.850590115864481e-07,
+      "loss": 1.2747,
+      "step": 1571
+    },
+    {
+      "epoch": 2.5152,
+      "grad_norm": 0.609767129530317,
+      "learning_rate": 7.80057539979554e-07,
+      "loss": 1.207,
+      "step": 1572
+    },
+    {
+      "epoch": 2.5168,
+      "grad_norm": 0.5761909168117831,
+      "learning_rate": 7.750707028084653e-07,
+      "loss": 1.3059,
+      "step": 1573
+    },
+    {
+      "epoch": 2.5183999999999997,
+      "grad_norm": 0.5552897640880203,
+      "learning_rate": 7.70098517367151e-07,
+      "loss": 1.2954,
+      "step": 1574
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.5781155573535764,
+      "learning_rate": 7.651410008987698e-07,
+      "loss": 1.2937,
+      "step": 1575
+    },
+    {
+      "epoch": 2.5216,
+      "grad_norm": 0.5717300016964076,
+      "learning_rate": 7.601981705956041e-07,
+      "loss": 1.2341,
+      "step": 1576
+    },
+    {
+      "epoch": 2.5232,
+      "grad_norm": 0.5462791663828772,
+      "learning_rate": 7.552700435990123e-07,
+      "loss": 1.2259,
+      "step": 1577
+    },
+    {
+      "epoch": 2.5248,
+      "grad_norm": 0.565140213659957,
+      "learning_rate": 7.503566369993564e-07,
+      "loss": 1.2036,
+      "step": 1578
+    },
+    {
+      "epoch": 2.5263999999999998,
+      "grad_norm": 0.5831706191745931,
+      "learning_rate": 7.454579678359547e-07,
+      "loss": 1.2887,
+      "step": 1579
+    },
+    {
+      "epoch": 2.528,
+      "grad_norm": 0.5800338051763073,
+      "learning_rate": 7.405740530970157e-07,
+      "loss": 1.288,
+      "step": 1580
+    },
+    {
+      "epoch": 2.5296,
+      "grad_norm": 0.5727144206726773,
+      "learning_rate": 7.357049097195773e-07,
+      "loss": 1.2095,
+      "step": 1581
+    },
+    {
+      "epoch": 2.5312,
+      "grad_norm": 0.5662300093904085,
+      "learning_rate": 7.308505545894567e-07,
+      "loss": 1.3077,
+      "step": 1582
+    },
+    {
+      "epoch": 2.5328,
+      "grad_norm": 0.5835910606913989,
+      "learning_rate": 7.260110045411816e-07,
+      "loss": 1.3038,
+      "step": 1583
+    },
+    {
+      "epoch": 2.5343999999999998,
+      "grad_norm": 0.5783539382953341,
+      "learning_rate": 7.211862763579414e-07,
+      "loss": 1.2314,
+      "step": 1584
+    },
+    {
+      "epoch": 2.536,
+      "grad_norm": 0.5649488588645777,
+      "learning_rate": 7.163763867715218e-07,
+      "loss": 1.2337,
+      "step": 1585
+    },
+    {
+      "epoch": 2.5376,
+      "grad_norm": 0.5631438545947426,
+      "learning_rate": 7.115813524622489e-07,
+      "loss": 1.2193,
+      "step": 1586
+    },
+    {
+      "epoch": 2.5392,
+      "grad_norm": 0.5590679011644212,
+      "learning_rate": 7.068011900589333e-07,
+      "loss": 1.2583,
+      "step": 1587
+    },
+    {
+      "epoch": 2.5408,
+      "grad_norm": 0.5906082364132665,
+      "learning_rate": 7.020359161388108e-07,
+      "loss": 1.2517,
+      "step": 1588
+    },
+    {
+      "epoch": 2.5423999999999998,
+      "grad_norm": 0.5602540879445478,
+      "learning_rate": 6.972855472274853e-07,
+      "loss": 1.2893,
+      "step": 1589
+    },
+    {
+      "epoch": 2.544,
+      "grad_norm": 0.557700285161506,
+      "learning_rate": 6.925500997988694e-07,
+      "loss": 1.27,
+      "step": 1590
+    },
+    {
+      "epoch": 2.5456,
+      "grad_norm": 0.5788373350120426,
+      "learning_rate": 6.87829590275132e-07,
+      "loss": 1.2858,
+      "step": 1591
+    },
+    {
+      "epoch": 2.5472,
+      "grad_norm": 0.5733265326434638,
+      "learning_rate": 6.83124035026635e-07,
+      "loss": 1.2599,
+      "step": 1592
+    },
+    {
+      "epoch": 2.5488,
+      "grad_norm": 0.570809715983484,
+      "learning_rate": 6.784334503718826e-07,
+      "loss": 1.2714,
+      "step": 1593
+    },
+    {
+      "epoch": 2.5504,
+      "grad_norm": 0.558688175063394,
+      "learning_rate": 6.737578525774636e-07,
+      "loss": 1.1827,
+      "step": 1594
+    },
+    {
+      "epoch": 2.552,
+      "grad_norm": 0.6014327429819659,
+      "learning_rate": 6.690972578579886e-07,
+      "loss": 1.2695,
+      "step": 1595
+    },
+    {
+      "epoch": 2.5536,
+      "grad_norm": 0.588142635157579,
+      "learning_rate": 6.644516823760439e-07,
+      "loss": 1.2952,
+      "step": 1596
+    },
+    {
+      "epoch": 2.5552,
+      "grad_norm": 0.589449782874084,
+      "learning_rate": 6.598211422421258e-07,
+      "loss": 1.294,
+      "step": 1597
+    },
+    {
+      "epoch": 2.5568,
+      "grad_norm": 0.5826488842546474,
+      "learning_rate": 6.552056535145917e-07,
+      "loss": 1.2711,
+      "step": 1598
+    },
+    {
+      "epoch": 2.5584,
+      "grad_norm": 0.5891532908893267,
+      "learning_rate": 6.506052321996037e-07,
+      "loss": 1.2783,
+      "step": 1599
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 0.5829053241592642,
+      "learning_rate": 6.46019894251066e-07,
+      "loss": 1.2215,
+      "step": 1600
+    },
+    {
+      "epoch": 2.5616,
+      "grad_norm": 0.572411848479696,
+      "learning_rate": 6.414496555705802e-07,
+      "loss": 1.2194,
+      "step": 1601
+    },
+    {
+      "epoch": 2.5632,
+      "grad_norm": 0.5936269474742342,
+      "learning_rate": 6.368945320073799e-07,
+      "loss": 1.3096,
+      "step": 1602
+    },
+    {
+      "epoch": 2.5648,
+      "grad_norm": 0.5642427040020616,
+      "learning_rate": 6.323545393582847e-07,
+      "loss": 1.2418,
+      "step": 1603
+    },
+    {
+      "epoch": 2.5664,
+      "grad_norm": 0.5772145575247725,
+      "learning_rate": 6.278296933676414e-07,
+      "loss": 1.3006,
+      "step": 1604
+    },
+    {
+      "epoch": 2.568,
+      "grad_norm": 0.5730307355954474,
+      "learning_rate": 6.233200097272646e-07,
+      "loss": 1.2475,
+      "step": 1605
+    },
+    {
+      "epoch": 2.5696,
+      "grad_norm": 0.5722699631010523,
+      "learning_rate": 6.188255040763929e-07,
+      "loss": 1.2069,
+      "step": 1606
+    },
+    {
+      "epoch": 2.5712,
+      "grad_norm": 0.5585877936693405,
+      "learning_rate": 6.143461920016247e-07,
+      "loss": 1.2408,
+      "step": 1607
+    },
+    {
+      "epoch": 2.5728,
+      "grad_norm": 0.5700083672750869,
+      "learning_rate": 6.098820890368696e-07,
+      "loss": 1.2833,
+      "step": 1608
+    },
+    {
+      "epoch": 2.5744,
+      "grad_norm": 0.6649938017946225,
+      "learning_rate": 6.054332106632943e-07,
+      "loss": 1.2172,
+      "step": 1609
+    },
+    {
+      "epoch": 2.576,
+      "grad_norm": 0.6586014480346398,
+      "learning_rate": 6.009995723092655e-07,
+      "loss": 1.2689,
+      "step": 1610
+    },
+    {
+      "epoch": 2.5776,
+      "grad_norm": 0.5488014461458274,
+      "learning_rate": 5.965811893503015e-07,
+      "loss": 1.221,
+      "step": 1611
+    },
+    {
+      "epoch": 2.5792,
+      "grad_norm": 0.59457168014389,
+      "learning_rate": 5.921780771090124e-07,
+      "loss": 1.3015,
+      "step": 1612
+    },
+    {
+      "epoch": 2.5808,
+      "grad_norm": 0.5618409387025245,
+      "learning_rate": 5.877902508550542e-07,
+      "loss": 1.262,
+      "step": 1613
+    },
+    {
+      "epoch": 2.5824,
+      "grad_norm": 0.5503473209107296,
+      "learning_rate": 5.834177258050711e-07,
+      "loss": 1.2357,
+      "step": 1614
+    },
+    {
+      "epoch": 2.584,
+      "grad_norm": 0.5531748896652627,
+      "learning_rate": 5.790605171226421e-07,
+      "loss": 1.2289,
+      "step": 1615
+    },
+    {
+      "epoch": 2.5856,
+      "grad_norm": 0.5518135432478328,
+      "learning_rate": 5.747186399182336e-07,
+      "loss": 1.2295,
+      "step": 1616
+    },
+    {
+      "epoch": 2.5872,
+      "grad_norm": 2.8333265848608726,
+      "learning_rate": 5.703921092491393e-07,
+      "loss": 1.2502,
+      "step": 1617
+    },
+    {
+      "epoch": 2.5888,
+      "grad_norm": 0.5587630452524199,
+      "learning_rate": 5.660809401194362e-07,
+      "loss": 1.2481,
+      "step": 1618
+    },
+    {
+      "epoch": 2.5904,
+      "grad_norm": 0.5902357730535409,
+      "learning_rate": 5.617851474799285e-07,
+      "loss": 1.2508,
+      "step": 1619
+    },
+    {
+      "epoch": 2.592,
+      "grad_norm": 0.5613550895419033,
+      "learning_rate": 5.575047462280919e-07,
+      "loss": 1.3005,
+      "step": 1620
+    },
+    {
+      "epoch": 2.5936,
+      "grad_norm": 0.5807095119813283,
+      "learning_rate": 5.532397512080306e-07,
+      "loss": 1.272,
+      "step": 1621
+    },
+    {
+      "epoch": 2.5952,
+      "grad_norm": 0.5654542500132802,
+      "learning_rate": 5.489901772104178e-07,
+      "loss": 1.2812,
+      "step": 1622
+    },
+    {
+      "epoch": 2.5968,
+      "grad_norm": 0.5789325935060607,
+      "learning_rate": 5.447560389724499e-07,
+      "loss": 1.2463,
+      "step": 1623
+    },
+    {
+      "epoch": 2.5984,
+      "grad_norm": 0.5943273235818389,
+      "learning_rate": 5.405373511777939e-07,
+      "loss": 1.2819,
+      "step": 1624
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 0.5818311518648028,
+      "learning_rate": 5.363341284565316e-07,
+      "loss": 1.2916,
+      "step": 1625
+    },
+    {
+      "epoch": 2.6016,
+      "grad_norm": 0.580923057262521,
+      "learning_rate": 5.321463853851189e-07,
+      "loss": 1.2444,
+      "step": 1626
+    },
+    {
+      "epoch": 2.6032,
+      "grad_norm": 0.5914721139192901,
+      "learning_rate": 5.279741364863244e-07,
+      "loss": 1.2497,
+      "step": 1627
+    },
+    {
+      "epoch": 2.6048,
+      "grad_norm": 0.5906420084796773,
+      "learning_rate": 5.238173962291881e-07,
+      "loss": 1.2676,
+      "step": 1628
+    },
+    {
+      "epoch": 2.6064,
+      "grad_norm": 0.5869300975019845,
+      "learning_rate": 5.196761790289639e-07,
+      "loss": 1.2499,
+      "step": 1629
+    },
+    {
+      "epoch": 2.608,
+      "grad_norm": 0.5875676224774843,
+      "learning_rate": 5.155504992470751e-07,
+      "loss": 1.2348,
+      "step": 1630
+    },
+    {
+      "epoch": 2.6096,
+      "grad_norm": 0.5524115159467096,
+      "learning_rate": 5.114403711910631e-07,
+      "loss": 1.2592,
+      "step": 1631
+    },
+    {
+      "epoch": 2.6112,
+      "grad_norm": 0.562584714521394,
+      "learning_rate": 5.073458091145328e-07,
+      "loss": 1.2348,
+      "step": 1632
+    },
+    {
+      "epoch": 2.6128,
+      "grad_norm": 0.5569097850727853,
+      "learning_rate": 5.032668272171138e-07,
+      "loss": 1.2474,
+      "step": 1633
+    },
+    {
+      "epoch": 2.6144,
+      "grad_norm": 0.570361354765726,
+      "learning_rate": 4.99203439644399e-07,
+      "loss": 1.2716,
+      "step": 1634
+    },
+    {
+      "epoch": 2.616,
+      "grad_norm": 0.5749622043446121,
+      "learning_rate": 4.951556604879049e-07,
+      "loss": 1.2865,
+      "step": 1635
+    },
+    {
+      "epoch": 2.6176,
+      "grad_norm": 0.5707576815839768,
+      "learning_rate": 4.911235037850187e-07,
+      "loss": 1.2677,
+      "step": 1636
+    },
+    {
+      "epoch": 2.6192,
+      "grad_norm": 0.573407999306702,
+      "learning_rate": 4.871069835189485e-07,
+      "loss": 1.2922,
+      "step": 1637
+    },
+    {
+      "epoch": 2.6208,
+      "grad_norm": 0.5658168803831255,
+      "learning_rate": 4.831061136186787e-07,
+      "loss": 1.1682,
+      "step": 1638
+    },
+    {
+      "epoch": 2.6224,
+      "grad_norm": 0.5528314541354952,
+      "learning_rate": 4.791209079589165e-07,
+      "loss": 1.276,
+      "step": 1639
+    },
+    {
+      "epoch": 2.624,
+      "grad_norm": 0.5713026290929196,
+      "learning_rate": 4.7515138036005157e-07,
+      "loss": 1.2663,
+      "step": 1640
+    },
+    {
+      "epoch": 2.6256,
+      "grad_norm": 0.5964816023059739,
+      "learning_rate": 4.7119754458809727e-07,
+      "loss": 1.3364,
+      "step": 1641
+    },
+    {
+      "epoch": 2.6272,
+      "grad_norm": 0.5714947544409424,
+      "learning_rate": 4.672594143546538e-07,
+      "loss": 1.2228,
+      "step": 1642
+    },
+    {
+      "epoch": 2.6288,
+      "grad_norm": 0.6000635979915935,
+      "learning_rate": 4.6333700331685385e-07,
+      "loss": 1.2671,
+      "step": 1643
+    },
+    {
+      "epoch": 2.6304,
+      "grad_norm": 0.5666267344589055,
+      "learning_rate": 4.594303250773152e-07,
+      "loss": 1.3141,
+      "step": 1644
+    },
+    {
+      "epoch": 2.632,
+      "grad_norm": 0.5904866556179479,
+      "learning_rate": 4.555393931841001e-07,
+      "loss": 1.2814,
+      "step": 1645
+    },
+    {
+      "epoch": 2.6336,
+      "grad_norm": 0.5859995037469117,
+      "learning_rate": 4.5166422113065877e-07,
+      "loss": 1.2737,
+      "step": 1646
+    },
+    {
+      "epoch": 2.6352,
+      "grad_norm": 0.5564513005508059,
+      "learning_rate": 4.478048223557907e-07,
+      "loss": 1.2868,
+      "step": 1647
+    },
+    {
+      "epoch": 2.6368,
+      "grad_norm": 0.5445357022320175,
+      "learning_rate": 4.439612102435942e-07,
+      "loss": 1.2289,
+      "step": 1648
+    },
+    {
+      "epoch": 2.6384,
+      "grad_norm": 0.5499592338581805,
+      "learning_rate": 4.401333981234196e-07,
+      "loss": 1.2791,
+      "step": 1649
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.5603603043337102,
+      "learning_rate": 4.3632139926982676e-07,
+      "loss": 1.2538,
+      "step": 1650
+    },
+    {
+      "epoch": 2.6416,
+      "grad_norm": 0.5618191597170836,
+      "learning_rate": 4.325252269025315e-07,
+      "loss": 1.2579,
+      "step": 1651
+    },
+    {
+      "epoch": 2.6432,
+      "grad_norm": 0.6039041372985126,
+      "learning_rate": 4.287448941863692e-07,
+      "loss": 1.2692,
+      "step": 1652
+    },
+    {
+      "epoch": 2.6448,
+      "grad_norm": 0.5822566221862587,
+      "learning_rate": 4.249804142312436e-07,
+      "loss": 1.3088,
+      "step": 1653
+    },
+    {
+      "epoch": 2.6464,
+      "grad_norm": 0.5419950064159947,
+      "learning_rate": 4.2123180009207956e-07,
+      "loss": 1.2666,
+      "step": 1654
+    },
+    {
+      "epoch": 2.648,
+      "grad_norm": 0.5524210190765074,
+      "learning_rate": 4.1749906476878486e-07,
+      "loss": 1.2281,
+      "step": 1655
+    },
+    {
+      "epoch": 2.6496,
+      "grad_norm": 0.5800791260964875,
+      "learning_rate": 4.137822212061965e-07,
+      "loss": 1.2946,
+      "step": 1656
+    },
+    {
+      "epoch": 2.6512000000000002,
+      "grad_norm": 0.5574660768384481,
+      "learning_rate": 4.100812822940431e-07,
+      "loss": 1.2466,
+      "step": 1657
+    },
+    {
+      "epoch": 2.6528,
+      "grad_norm": 0.5503172374110917,
+      "learning_rate": 4.063962608668959e-07,
+      "loss": 1.2739,
+      "step": 1658
+    },
+    {
+      "epoch": 2.6544,
+      "grad_norm": 0.5846803189329707,
+      "learning_rate": 4.0272716970412516e-07,
+      "loss": 1.3413,
+      "step": 1659
+    },
+    {
+      "epoch": 2.656,
+      "grad_norm": 0.5784809766784924,
+      "learning_rate": 3.990740215298583e-07,
+      "loss": 1.2555,
+      "step": 1660
+    },
+    {
+      "epoch": 2.6576,
+      "grad_norm": 0.5557005060529346,
+      "learning_rate": 3.954368290129301e-07,
+      "loss": 1.2071,
+      "step": 1661
+    },
+    {
+      "epoch": 2.6592000000000002,
+      "grad_norm": 0.5724384940244169,
+      "learning_rate": 3.918156047668453e-07,
+      "loss": 1.2725,
+      "step": 1662
+    },
+    {
+      "epoch": 2.6608,
+      "grad_norm": 0.5622555319217774,
+      "learning_rate": 3.882103613497318e-07,
+      "loss": 1.3313,
+      "step": 1663
+    },
+    {
+      "epoch": 2.6624,
+      "grad_norm": 0.5832435420214503,
+      "learning_rate": 3.84621111264295e-07,
+      "loss": 1.2667,
+      "step": 1664
+    },
+    {
+      "epoch": 2.664,
+      "grad_norm": 0.5781335567299438,
+      "learning_rate": 3.810478669577794e-07,
+      "loss": 1.282,
+      "step": 1665
+    },
+    {
+      "epoch": 2.6656,
+      "grad_norm": 0.5625982602679049,
+      "learning_rate": 3.7749064082191976e-07,
+      "loss": 1.2638,
+      "step": 1666
+    },
+    {
+      "epoch": 2.6672000000000002,
+      "grad_norm": 0.5690130791455427,
+      "learning_rate": 3.739494451929049e-07,
+      "loss": 1.2084,
+      "step": 1667
+    },
+    {
+      "epoch": 2.6688,
+      "grad_norm": 0.557342713903056,
+      "learning_rate": 3.7042429235132625e-07,
+      "loss": 1.2311,
+      "step": 1668
+    },
+    {
+      "epoch": 2.6704,
+      "grad_norm": 0.5380613819239237,
+      "learning_rate": 3.6691519452214387e-07,
+      "loss": 1.249,
+      "step": 1669
+    },
+    {
+      "epoch": 2.672,
+      "grad_norm": 0.5658002507891863,
+      "learning_rate": 3.6342216387464047e-07,
+      "loss": 1.3192,
+      "step": 1670
+    },
+    {
+      "epoch": 2.6736,
+      "grad_norm": 0.5841584365577271,
+      "learning_rate": 3.5994521252237516e-07,
+      "loss": 1.2655,
+      "step": 1671
+    },
+    {
+      "epoch": 2.6752000000000002,
+      "grad_norm": 0.5510555787121992,
+      "learning_rate": 3.564843525231498e-07,
+      "loss": 1.2021,
+      "step": 1672
+    },
+    {
+      "epoch": 2.6768,
+      "grad_norm": 0.5692505251039209,
+      "learning_rate": 3.53039595878959e-07,
+      "loss": 1.3248,
+      "step": 1673
+    },
+    {
+      "epoch": 2.6784,
+      "grad_norm": 0.5675899017560704,
+      "learning_rate": 3.496109545359544e-07,
+      "loss": 1.2154,
+      "step": 1674
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 0.5577041775528249,
+      "learning_rate": 3.461984403844015e-07,
+      "loss": 1.2284,
+      "step": 1675
+    },
+    {
+      "epoch": 2.6816,
+      "grad_norm": 0.5393962375581134,
+      "learning_rate": 3.42802065258635e-07,
+      "loss": 1.2347,
+      "step": 1676
+    },
+    {
+      "epoch": 2.6832000000000003,
+      "grad_norm": 0.5767340361586724,
+      "learning_rate": 3.394218409370242e-07,
+      "loss": 1.2703,
+      "step": 1677
+    },
+    {
+      "epoch": 2.6848,
+      "grad_norm": 0.5547540763753328,
+      "learning_rate": 3.360577791419256e-07,
+      "loss": 1.2455,
+      "step": 1678
+    },
+    {
+      "epoch": 2.6864,
+      "grad_norm": 0.5567347375194274,
+      "learning_rate": 3.3270989153964707e-07,
+      "loss": 1.2984,
+      "step": 1679
+    },
+    {
+      "epoch": 2.6879999999999997,
+      "grad_norm": 0.5713455471384397,
+      "learning_rate": 3.2937818974040637e-07,
+      "loss": 1.2344,
+      "step": 1680
+    },
+    {
+      "epoch": 2.6896,
+      "grad_norm": 0.833227236334107,
+      "learning_rate": 3.260626852982873e-07,
+      "loss": 1.2557,
+      "step": 1681
+    },
+    {
+      "epoch": 2.6912000000000003,
+      "grad_norm": 0.5573478132148704,
+      "learning_rate": 3.227633897112059e-07,
+      "loss": 1.2566,
+      "step": 1682
+    },
+    {
+      "epoch": 2.6928,
+      "grad_norm": 0.5538435370719593,
+      "learning_rate": 3.194803144208636e-07,
+      "loss": 1.2258,
+      "step": 1683
+    },
+    {
+      "epoch": 2.6944,
+      "grad_norm": 0.5354974249210992,
+      "learning_rate": 3.16213470812714e-07,
+      "loss": 1.1948,
+      "step": 1684
+    },
+    {
+      "epoch": 2.6959999999999997,
+      "grad_norm": 0.5573775830561629,
+      "learning_rate": 3.129628702159204e-07,
+      "loss": 1.2373,
+      "step": 1685
+    },
+    {
+      "epoch": 2.6976,
+      "grad_norm": 0.5465879056572422,
+      "learning_rate": 3.097285239033138e-07,
+      "loss": 1.2731,
+      "step": 1686
+    },
+    {
+      "epoch": 2.6992000000000003,
+      "grad_norm": 0.5769130656162111,
+      "learning_rate": 3.0651044309136016e-07,
+      "loss": 1.266,
+      "step": 1687
+    },
+    {
+      "epoch": 2.7008,
+      "grad_norm": 0.5671117315453904,
+      "learning_rate": 3.033086389401141e-07,
+      "loss": 1.2677,
+      "step": 1688
+    },
+    {
+      "epoch": 2.7024,
+      "grad_norm": 0.5741045868380483,
+      "learning_rate": 3.0012312255318696e-07,
+      "loss": 1.3348,
+      "step": 1689
+    },
+    {
+      "epoch": 2.7039999999999997,
+      "grad_norm": 0.5462603168888583,
+      "learning_rate": 2.9695390497770535e-07,
+      "loss": 1.2156,
+      "step": 1690
+    },
+    {
+      "epoch": 2.7056,
+      "grad_norm": 0.5518773877014073,
+      "learning_rate": 2.93800997204271e-07,
+      "loss": 1.31,
+      "step": 1691
+    },
+    {
+      "epoch": 2.7072000000000003,
+      "grad_norm": 0.5757408741523948,
+      "learning_rate": 2.9066441016692594e-07,
+      "loss": 1.2562,
+      "step": 1692
+    },
+    {
+      "epoch": 2.7088,
+      "grad_norm": 0.5476620828478335,
+      "learning_rate": 2.8754415474311235e-07,
+      "loss": 1.2363,
+      "step": 1693
+    },
+    {
+      "epoch": 2.7104,
+      "grad_norm": 0.5706612414035004,
+      "learning_rate": 2.844402417536374e-07,
+      "loss": 1.2772,
+      "step": 1694
+    },
+    {
+      "epoch": 2.7119999999999997,
+      "grad_norm": 0.541510251087581,
+      "learning_rate": 2.8135268196263055e-07,
+      "loss": 1.2803,
+      "step": 1695
+    },
+    {
+      "epoch": 2.7136,
+      "grad_norm": 0.5747067535258159,
+      "learning_rate": 2.782814860775124e-07,
+      "loss": 1.3146,
+      "step": 1696
+    },
+    {
+      "epoch": 2.7152,
+      "grad_norm": 0.5467734425967701,
+      "learning_rate": 2.752266647489549e-07,
+      "loss": 1.2326,
+      "step": 1697
+    },
+    {
+      "epoch": 2.7168,
+      "grad_norm": 0.5758681613346702,
+      "learning_rate": 2.7218822857084217e-07,
+      "loss": 1.2714,
+      "step": 1698
+    },
+    {
+      "epoch": 2.7184,
+      "grad_norm": 0.5585531005547033,
+      "learning_rate": 2.691661880802382e-07,
+      "loss": 1.288,
+      "step": 1699
+    },
+    {
+      "epoch": 2.7199999999999998,
+      "grad_norm": 0.5391193255301477,
+      "learning_rate": 2.661605537573453e-07,
+      "loss": 1.2488,
+      "step": 1700
+    },
+    {
+      "epoch": 2.7216,
+      "grad_norm": 0.5438447673124908,
+      "learning_rate": 2.631713360254734e-07,
+      "loss": 1.2208,
+      "step": 1701
+    },
+    {
+      "epoch": 2.7232,
+      "grad_norm": 0.55104906611935,
+      "learning_rate": 2.6019854525099977e-07,
+      "loss": 1.2954,
+      "step": 1702
+    },
+    {
+      "epoch": 2.7248,
+      "grad_norm": 0.5676994108159752,
+      "learning_rate": 2.572421917433332e-07,
+      "loss": 1.2659,
+      "step": 1703
+    },
+    {
+      "epoch": 2.7264,
+      "grad_norm": 0.5523169149970274,
+      "learning_rate": 2.5430228575488156e-07,
+      "loss": 1.2715,
+      "step": 1704
+    },
+    {
+      "epoch": 2.7279999999999998,
+      "grad_norm": 0.552232995728815,
+      "learning_rate": 2.513788374810111e-07,
+      "loss": 1.299,
+      "step": 1705
+    },
+    {
+      "epoch": 2.7296,
+      "grad_norm": 0.5601590729174357,
+      "learning_rate": 2.4847185706001643e-07,
+      "loss": 1.2746,
+      "step": 1706
+    },
+    {
+      "epoch": 2.7312,
+      "grad_norm": 0.562292238700845,
+      "learning_rate": 2.455813545730812e-07,
+      "loss": 1.2457,
+      "step": 1707
+    },
+    {
+      "epoch": 2.7328,
+      "grad_norm": 0.5635210925734802,
+      "learning_rate": 2.4270734004424643e-07,
+      "loss": 1.2781,
+      "step": 1708
+    },
+    {
+      "epoch": 2.7344,
+      "grad_norm": 0.5719231077574123,
+      "learning_rate": 2.39849823440374e-07,
+      "loss": 1.2366,
+      "step": 1709
+    },
+    {
+      "epoch": 2.7359999999999998,
+      "grad_norm": 0.5754100752092469,
+      "learning_rate": 2.3700881467111025e-07,
+      "loss": 1.2629,
+      "step": 1710
+    },
+    {
+      "epoch": 2.7376,
+      "grad_norm": 0.5750180165381704,
+      "learning_rate": 2.3418432358885633e-07,
+      "loss": 1.2524,
+      "step": 1711
+    },
+    {
+      "epoch": 2.7392,
+      "grad_norm": 0.5581236003728374,
+      "learning_rate": 2.3137635998872808e-07,
+      "loss": 1.3031,
+      "step": 1712
+    },
+    {
+      "epoch": 2.7408,
+      "grad_norm": 0.5566408946071852,
+      "learning_rate": 2.285849336085294e-07,
+      "loss": 1.2969,
+      "step": 1713
+    },
+    {
+      "epoch": 2.7424,
+      "grad_norm": 0.5650277215685173,
+      "learning_rate": 2.258100541287117e-07,
+      "loss": 1.2712,
+      "step": 1714
+    },
+    {
+      "epoch": 2.7439999999999998,
+      "grad_norm": 0.579829670140934,
+      "learning_rate": 2.2305173117234236e-07,
+      "loss": 1.2698,
+      "step": 1715
+    },
+    {
+      "epoch": 2.7456,
+      "grad_norm": 0.5614650736355169,
+      "learning_rate": 2.2030997430507462e-07,
+      "loss": 1.2545,
+      "step": 1716
+    },
+    {
+      "epoch": 2.7472,
+      "grad_norm": 0.5465402367379577,
+      "learning_rate": 2.1758479303510937e-07,
+      "loss": 1.261,
+      "step": 1717
+    },
+    {
+      "epoch": 2.7488,
+      "grad_norm": 0.5457246643954163,
+      "learning_rate": 2.148761968131663e-07,
+      "loss": 1.3527,
+      "step": 1718
+    },
+    {
+      "epoch": 2.7504,
+      "grad_norm": 0.5489080933995129,
+      "learning_rate": 2.121841950324488e-07,
+      "loss": 1.2098,
+      "step": 1719
+    },
+    {
+      "epoch": 2.752,
+      "grad_norm": 0.5473277548830675,
+      "learning_rate": 2.0950879702861082e-07,
+      "loss": 1.24,
+      "step": 1720
+    },
+    {
+      "epoch": 2.7536,
+      "grad_norm": 0.5702853745132767,
+      "learning_rate": 2.0685001207972843e-07,
+      "loss": 1.2671,
+      "step": 1721
+    },
+    {
+      "epoch": 2.7552,
+      "grad_norm": 0.5392017075083879,
+      "learning_rate": 2.042078494062616e-07,
+      "loss": 1.1973,
+      "step": 1722
+    },
+    {
+      "epoch": 2.7568,
+      "grad_norm": 0.5590205903367657,
+      "learning_rate": 2.0158231817102858e-07,
+      "loss": 1.2833,
+      "step": 1723
+    },
+    {
+      "epoch": 2.7584,
+      "grad_norm": 0.5672607874010511,
+      "learning_rate": 1.9897342747916938e-07,
+      "loss": 1.3125,
+      "step": 1724
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.5675816421102972,
+      "learning_rate": 1.9638118637811564e-07,
+      "loss": 1.1871,
+      "step": 1725
+    },
+    {
+      "epoch": 2.7616,
+      "grad_norm": 0.5546409402688645,
+      "learning_rate": 1.9380560385756088e-07,
+      "loss": 1.2313,
+      "step": 1726
+    },
+    {
+      "epoch": 2.7632,
+      "grad_norm": 0.5439962608523701,
+      "learning_rate": 1.9124668884942632e-07,
+      "loss": 1.2708,
+      "step": 1727
+    },
+    {
+      "epoch": 2.7648,
+      "grad_norm": 0.5599662172118572,
+      "learning_rate": 1.8870445022783234e-07,
+      "loss": 1.2838,
+      "step": 1728
+    },
+    {
+      "epoch": 2.7664,
+      "grad_norm": 0.5780013010181216,
+      "learning_rate": 1.861788968090683e-07,
+      "loss": 1.2817,
+      "step": 1729
+    },
+    {
+      "epoch": 2.768,
+      "grad_norm": 0.5403680119488328,
+      "learning_rate": 1.8367003735155764e-07,
+      "loss": 1.2328,
+      "step": 1730
+    },
+    {
+      "epoch": 2.7696,
+      "grad_norm": 0.560561298966086,
+      "learning_rate": 1.8117788055583286e-07,
+      "loss": 1.2274,
+      "step": 1731
+    },
+    {
+      "epoch": 2.7712,
+      "grad_norm": 0.5543554854029232,
+      "learning_rate": 1.7870243506450113e-07,
+      "loss": 1.2786,
+      "step": 1732
+    },
+    {
+      "epoch": 2.7728,
+      "grad_norm": 0.5483654707203095,
+      "learning_rate": 1.762437094622177e-07,
+      "loss": 1.2496,
+      "step": 1733
+    },
+    {
+      "epoch": 2.7744,
+      "grad_norm": 0.5541440242010299,
+      "learning_rate": 1.738017122756541e-07,
+      "loss": 1.2715,
+      "step": 1734
+    },
+    {
+      "epoch": 2.776,
+      "grad_norm": 0.5493365087188109,
+      "learning_rate": 1.713764519734673e-07,
+      "loss": 1.2624,
+      "step": 1735
+    },
+    {
+      "epoch": 2.7776,
+      "grad_norm": 0.5506895581941508,
+      "learning_rate": 1.68967936966275e-07,
+      "loss": 1.2515,
+      "step": 1736
+    },
+    {
+      "epoch": 2.7792,
+      "grad_norm": 0.5743299004646728,
+      "learning_rate": 1.6657617560662088e-07,
+      "loss": 1.2849,
+      "step": 1737
+    },
+    {
+      "epoch": 2.7808,
+      "grad_norm": 0.5637493127397919,
+      "learning_rate": 1.6420117618895003e-07,
+      "loss": 1.2973,
+      "step": 1738
+    },
+    {
+      "epoch": 2.7824,
+      "grad_norm": 0.5482123218560021,
+      "learning_rate": 1.6184294694957747e-07,
+      "loss": 1.1998,
+      "step": 1739
+    },
+    {
+      "epoch": 2.784,
+      "grad_norm": 0.5572929724255375,
+      "learning_rate": 1.5950149606666077e-07,
+      "loss": 1.2955,
+      "step": 1740
+    },
+    {
+      "epoch": 2.7856,
+      "grad_norm": 0.5610585292284709,
+      "learning_rate": 1.5717683166017184e-07,
+      "loss": 1.3234,
+      "step": 1741
+    },
+    {
+      "epoch": 2.7872,
+      "grad_norm": 0.5590657574820085,
+      "learning_rate": 1.5486896179186693e-07,
+      "loss": 1.2522,
+      "step": 1742
+    },
+    {
+      "epoch": 2.7888,
+      "grad_norm": 0.5635316592445546,
+      "learning_rate": 1.5257789446526172e-07,
+      "loss": 1.2136,
+      "step": 1743
+    },
+    {
+      "epoch": 2.7904,
+      "grad_norm": 0.5563256733490789,
+      "learning_rate": 1.5030363762560228e-07,
+      "loss": 1.248,
+      "step": 1744
+    },
+    {
+      "epoch": 2.792,
+      "grad_norm": 0.5972603512819085,
+      "learning_rate": 1.480461991598353e-07,
+      "loss": 1.2648,
+      "step": 1745
+    },
+    {
+      "epoch": 2.7936,
+      "grad_norm": 0.5621460572695758,
+      "learning_rate": 1.458055868965841e-07,
+      "loss": 1.2374,
+      "step": 1746
+    },
+    {
+      "epoch": 2.7952,
+      "grad_norm": 0.5670447144333843,
+      "learning_rate": 1.4358180860611913e-07,
+      "loss": 1.2585,
+      "step": 1747
+    },
+    {
+      "epoch": 2.7968,
+      "grad_norm": 0.5819116642427915,
+      "learning_rate": 1.4137487200033383e-07,
+      "loss": 1.2046,
+      "step": 1748
+    },
+    {
+      "epoch": 2.7984,
+      "grad_norm": 0.5568896030472557,
+      "learning_rate": 1.3918478473271325e-07,
+      "loss": 1.3358,
+      "step": 1749
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 0.5804439020989762,
+      "learning_rate": 1.3701155439831249e-07,
+      "loss": 1.2715,
+      "step": 1750
+    },
+    {
+      "epoch": 2.8016,
+      "grad_norm": 0.5847295390943986,
+      "learning_rate": 1.3485518853372625e-07,
+      "loss": 1.2942,
+      "step": 1751
+    },
+    {
+      "epoch": 2.8032,
+      "grad_norm": 0.5564863967564779,
+      "learning_rate": 1.3271569461706547e-07,
+      "loss": 1.254,
+      "step": 1752
+    },
+    {
+      "epoch": 2.8048,
+      "grad_norm": 0.5821597829500411,
+      "learning_rate": 1.305930800679317e-07,
+      "loss": 1.2641,
+      "step": 1753
+    },
+    {
+      "epoch": 2.8064,
+      "grad_norm": 0.5614263976427496,
+      "learning_rate": 1.2848735224738729e-07,
+      "loss": 1.2516,
+      "step": 1754
+    },
+    {
+      "epoch": 2.808,
+      "grad_norm": 0.5402974900152122,
+      "learning_rate": 1.2639851845793583e-07,
+      "loss": 1.2775,
+      "step": 1755
+    },
+    {
+      "epoch": 2.8096,
+      "grad_norm": 0.5473757639682008,
+      "learning_rate": 1.2432658594349113e-07,
+      "loss": 1.2763,
+      "step": 1756
+    },
+    {
+      "epoch": 2.8112,
+      "grad_norm": 0.5534558211430033,
+      "learning_rate": 1.2227156188935552e-07,
+      "loss": 1.2625,
+      "step": 1757
+    },
+    {
+      "epoch": 2.8128,
+      "grad_norm": 0.5462880068144754,
+      "learning_rate": 1.202334534221955e-07,
+      "loss": 1.2435,
+      "step": 1758
+    },
+    {
+      "epoch": 2.8144,
+      "grad_norm": 0.5532480692782625,
+      "learning_rate": 1.1821226761001391e-07,
+      "loss": 1.3015,
+      "step": 1759
+    },
+    {
+      "epoch": 2.816,
+      "grad_norm": 0.5565049166907481,
+      "learning_rate": 1.1620801146212723e-07,
+      "loss": 1.2515,
+      "step": 1760
+    },
+    {
+      "epoch": 2.8176,
+      "grad_norm": 0.5424741306268211,
+      "learning_rate": 1.1422069192914221e-07,
+      "loss": 1.3148,
+      "step": 1761
+    },
+    {
+      "epoch": 2.8192,
+      "grad_norm": 0.5510168420123431,
+      "learning_rate": 1.1225031590292923e-07,
+      "loss": 1.2496,
+      "step": 1762
+    },
+    {
+      "epoch": 2.8208,
+      "grad_norm": 0.5608846863291166,
+      "learning_rate": 1.1029689021660183e-07,
+      "loss": 1.366,
+      "step": 1763
+    },
+    {
+      "epoch": 2.8224,
+      "grad_norm": 0.557422367604704,
+      "learning_rate": 1.0836042164448945e-07,
+      "loss": 1.2138,
+      "step": 1764
+    },
+    {
+      "epoch": 2.824,
+      "grad_norm": 0.5823443873149928,
+      "learning_rate": 1.0644091690211633e-07,
+      "loss": 1.2424,
+      "step": 1765
+    },
+    {
+      "epoch": 2.8256,
+      "grad_norm": 0.5421481298388121,
+      "learning_rate": 1.0453838264617711e-07,
+      "loss": 1.3096,
+      "step": 1766
+    },
+    {
+      "epoch": 2.8272,
+      "grad_norm": 0.5612996449980537,
+      "learning_rate": 1.0265282547451405e-07,
+      "loss": 1.2782,
+      "step": 1767
+    },
+    {
+      "epoch": 2.8288,
+      "grad_norm": 0.5685227995684492,
+      "learning_rate": 1.0078425192609487e-07,
+      "loss": 1.3294,
+      "step": 1768
+    },
+    {
+      "epoch": 2.8304,
+      "grad_norm": 0.5692148398087089,
+      "learning_rate": 9.893266848098826e-08,
+      "loss": 1.2928,
+      "step": 1769
+    },
+    {
+      "epoch": 2.832,
+      "grad_norm": 0.5606204136497016,
+      "learning_rate": 9.709808156034394e-08,
+      "loss": 1.2158,
+      "step": 1770
+    },
+    {
+      "epoch": 2.8336,
+      "grad_norm": 0.5461779131372307,
+      "learning_rate": 9.528049752636714e-08,
+      "loss": 1.2363,
+      "step": 1771
+    },
+    {
+      "epoch": 2.8352,
+      "grad_norm": 0.5717931937607929,
+      "learning_rate": 9.347992268230022e-08,
+      "loss": 1.322,
+      "step": 1772
+    },
+    {
+      "epoch": 2.8368,
+      "grad_norm": 0.6124021503719553,
+      "learning_rate": 9.169636327239883e-08,
+      "loss": 1.2628,
+      "step": 1773
+    },
+    {
+      "epoch": 2.8384,
+      "grad_norm": 0.5527650207668391,
+      "learning_rate": 8.992982548190809e-08,
+      "loss": 1.2769,
+      "step": 1774
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 0.5587460601150641,
+      "learning_rate": 8.818031543704641e-08,
+      "loss": 1.1826,
+      "step": 1775
+    },
+    {
+      "epoch": 2.8416,
+      "grad_norm": 0.5547570670036566,
+      "learning_rate": 8.644783920498001e-08,
+      "loss": 1.2579,
+      "step": 1776
+    },
+    {
+      "epoch": 2.8432,
+      "grad_norm": 0.5603062999480989,
+      "learning_rate": 8.473240279380235e-08,
+      "loss": 1.2464,
+      "step": 1777
+    },
+    {
+      "epoch": 2.8448,
+      "grad_norm": 0.5645188579405525,
+      "learning_rate": 8.303401215251583e-08,
+      "loss": 1.2125,
+      "step": 1778
+    },
+    {
+      "epoch": 2.8464,
+      "grad_norm": 0.5647207872092516,
+      "learning_rate": 8.135267317100792e-08,
+      "loss": 1.3015,
+      "step": 1779
+    },
+    {
+      "epoch": 2.848,
+      "grad_norm": 0.5571408831702612,
+      "learning_rate": 7.968839168003395e-08,
+      "loss": 1.2539,
+      "step": 1780
+    },
+    {
+      "epoch": 2.8496,
+      "grad_norm": 2.5911876971039427,
+      "learning_rate": 7.804117345119266e-08,
+      "loss": 1.3049,
+      "step": 1781
+    },
+    {
+      "epoch": 2.8512,
+      "grad_norm": 0.5396169314276171,
+      "learning_rate": 7.64110241969107e-08,
+      "loss": 1.2626,
+      "step": 1782
+    },
+    {
+      "epoch": 2.8528000000000002,
+      "grad_norm": 0.5548772559064917,
+      "learning_rate": 7.479794957042041e-08,
+      "loss": 1.2352,
+      "step": 1783
+    },
+    {
+      "epoch": 2.8544,
+      "grad_norm": 0.5517301416184537,
+      "learning_rate": 7.320195516574036e-08,
+      "loss": 1.2743,
+      "step": 1784
+    },
+    {
+      "epoch": 2.856,
+      "grad_norm": 0.5455374081059908,
+      "learning_rate": 7.16230465176565e-08,
+      "loss": 1.2207,
+      "step": 1785
+    },
+    {
+      "epoch": 2.8576,
+      "grad_norm": 0.5538076256095801,
+      "learning_rate": 7.00612291017022e-08,
+      "loss": 1.2845,
+      "step": 1786
+    },
+    {
+      "epoch": 2.8592,
+      "grad_norm": 0.5460088941095047,
+      "learning_rate": 6.851650833414103e-08,
+      "loss": 1.2955,
+      "step": 1787
+    },
+    {
+      "epoch": 2.8608000000000002,
+      "grad_norm": 0.5795766443005916,
+      "learning_rate": 6.698888957194505e-08,
+      "loss": 1.2718,
+      "step": 1788
+    },
+    {
+      "epoch": 2.8624,
+      "grad_norm": 0.5534242953773871,
+      "learning_rate": 6.547837811277824e-08,
+      "loss": 1.2371,
+      "step": 1789
+    },
+    {
+      "epoch": 2.864,
+      "grad_norm": 0.557595341030343,
+      "learning_rate": 6.39849791949787e-08,
+      "loss": 1.2438,
+      "step": 1790
+    },
+    {
+      "epoch": 2.8656,
+      "grad_norm": 0.5575363105183772,
+      "learning_rate": 6.250869799753866e-08,
+      "loss": 1.2304,
+      "step": 1791
+    },
+    {
+      "epoch": 2.8672,
+      "grad_norm": 0.5511957581682512,
+      "learning_rate": 6.104953964008897e-08,
+      "loss": 1.3133,
+      "step": 1792
+    },
+    {
+      "epoch": 2.8688000000000002,
+      "grad_norm": 0.5211492542882514,
+      "learning_rate": 5.960750918287627e-08,
+      "loss": 1.226,
+      "step": 1793
+    },
+    {
+      "epoch": 2.8704,
+      "grad_norm": 0.5624150656631208,
+      "learning_rate": 5.818261162675309e-08,
+      "loss": 1.3049,
+      "step": 1794
+    },
+    {
+      "epoch": 2.872,
+      "grad_norm": 0.5583884284219154,
+      "learning_rate": 5.677485191315391e-08,
+      "loss": 1.2099,
+      "step": 1795
+    },
+    {
+      "epoch": 2.8736,
+      "grad_norm": 0.563326361572727,
+      "learning_rate": 5.538423492408129e-08,
+      "loss": 1.2277,
+      "step": 1796
+    },
+    {
+      "epoch": 2.8752,
+      "grad_norm": 0.5503718663661139,
+      "learning_rate": 5.401076548208761e-08,
+      "loss": 1.2771,
+      "step": 1797
+    },
+    {
+      "epoch": 2.8768000000000002,
+      "grad_norm": 0.5768693124576363,
+      "learning_rate": 5.265444835025946e-08,
+      "loss": 1.3491,
+      "step": 1798
+    },
+    {
+      "epoch": 2.8784,
+      "grad_norm": 0.5662430315367616,
+      "learning_rate": 5.1315288232201e-08,
+      "loss": 1.3278,
+      "step": 1799
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.5532212264664487,
+      "learning_rate": 4.9993289772015116e-08,
+      "loss": 1.2929,
+      "step": 1800
+    },
+    {
+      "epoch": 2.8816,
+      "grad_norm": 0.5448398243846065,
+      "learning_rate": 4.8688457554291746e-08,
+      "loss": 1.2226,
+      "step": 1801
+    },
+    {
+      "epoch": 2.8832,
+      "grad_norm": 0.5836962978476233,
+      "learning_rate": 4.7400796104088434e-08,
+      "loss": 1.3033,
+      "step": 1802
+    },
+    {
+      "epoch": 2.8848000000000003,
+      "grad_norm": 0.5569168645000776,
+      "learning_rate": 4.613030988691536e-08,
+      "loss": 1.3226,
+      "step": 1803
+    },
+    {
+      "epoch": 2.8864,
+      "grad_norm": 0.5437619322877473,
+      "learning_rate": 4.4877003308722575e-08,
+      "loss": 1.1876,
+      "step": 1804
+    },
+    {
+      "epoch": 2.888,
+      "grad_norm": 0.5518534849018494,
+      "learning_rate": 4.364088071587891e-08,
+      "loss": 1.2105,
+      "step": 1805
+    },
+    {
+      "epoch": 2.8895999999999997,
+      "grad_norm": 0.5491187383658536,
+      "learning_rate": 4.2421946395164174e-08,
+      "loss": 1.2772,
+      "step": 1806
+    },
+    {
+      "epoch": 2.8912,
+      "grad_norm": 0.5624016663599182,
+      "learning_rate": 4.1220204573747534e-08,
+      "loss": 1.2901,
+      "step": 1807
+    },
+    {
+      "epoch": 2.8928000000000003,
+      "grad_norm": 0.5447675078444026,
+      "learning_rate": 4.0035659419178086e-08,
+      "loss": 1.2685,
+      "step": 1808
+    },
+    {
+      "epoch": 2.8944,
+      "grad_norm": 0.5606247308583606,
+      "learning_rate": 3.88683150393665e-08,
+      "loss": 1.2502,
+      "step": 1809
+    },
+    {
+      "epoch": 2.896,
+      "grad_norm": 0.5300843675573687,
+      "learning_rate": 3.771817548257395e-08,
+      "loss": 1.2475,
+      "step": 1810
+    },
+    {
+      "epoch": 2.8975999999999997,
+      "grad_norm": 0.5343342293102542,
+      "learning_rate": 3.658524473739544e-08,
+      "loss": 1.2384,
+      "step": 1811
+    },
+    {
+      "epoch": 2.8992,
+      "grad_norm": 0.5617525854075,
+      "learning_rate": 3.546952673274817e-08,
+      "loss": 1.1979,
+      "step": 1812
+    },
+    {
+      "epoch": 2.9008000000000003,
+      "grad_norm": 0.6389617017119005,
+      "learning_rate": 3.437102533785541e-08,
+      "loss": 1.2359,
+      "step": 1813
+    },
+    {
+      "epoch": 2.9024,
+      "grad_norm": 0.5685069692244814,
+      "learning_rate": 3.328974436223709e-08,
+      "loss": 1.3189,
+      "step": 1814
+    },
+    {
+      "epoch": 2.904,
+      "grad_norm": 0.5543820403552354,
+      "learning_rate": 3.2225687555690886e-08,
+      "loss": 1.2379,
+      "step": 1815
+    },
+    {
+      "epoch": 2.9055999999999997,
+      "grad_norm": 0.5656407809510097,
+      "learning_rate": 3.117885860828396e-08,
+      "loss": 1.2821,
+      "step": 1816
+    },
+    {
+      "epoch": 2.9072,
+      "grad_norm": 0.5355099980915239,
+      "learning_rate": 3.014926115034012e-08,
+      "loss": 1.2309,
+      "step": 1817
+    },
+    {
+      "epoch": 2.9088000000000003,
+      "grad_norm": 0.5904342153734854,
+      "learning_rate": 2.9136898752422648e-08,
+      "loss": 1.24,
+      "step": 1818
+    },
+    {
+      "epoch": 2.9104,
+      "grad_norm": 0.5536692444451957,
+      "learning_rate": 2.8141774925327103e-08,
+      "loss": 1.264,
+      "step": 1819
+    },
+    {
+      "epoch": 2.912,
+      "grad_norm": 0.5562705346310348,
+      "learning_rate": 2.7163893120066288e-08,
+      "loss": 1.2602,
+      "step": 1820
+    },
+    {
+      "epoch": 2.9135999999999997,
+      "grad_norm": 0.574848336584587,
+      "learning_rate": 2.6203256727859172e-08,
+      "loss": 1.2632,
+      "step": 1821
+    },
+    {
+      "epoch": 2.9152,
+      "grad_norm": 0.5886039324136505,
+      "learning_rate": 2.5259869080118127e-08,
+      "loss": 1.2774,
+      "step": 1822
+    },
+    {
+      "epoch": 2.9168,
+      "grad_norm": 0.5410660015106693,
+      "learning_rate": 2.4333733448440033e-08,
+      "loss": 1.1799,
+      "step": 1823
+    },
+    {
+      "epoch": 2.9184,
+      "grad_norm": 0.5597997165370372,
+      "learning_rate": 2.34248530445913e-08,
+      "loss": 1.2227,
+      "step": 1824
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 0.5635645821862755,
+      "learning_rate": 2.2533231020499536e-08,
+      "loss": 1.2383,
+      "step": 1825
+    },
+    {
+      "epoch": 2.9215999999999998,
+      "grad_norm": 0.5389487278505842,
+      "learning_rate": 2.1658870468241332e-08,
+      "loss": 1.2519,
+      "step": 1826
+    },
+    {
+      "epoch": 2.9232,
+      "grad_norm": 0.5674970718856741,
+      "learning_rate": 2.0801774420031172e-08,
+      "loss": 1.2647,
+      "step": 1827
+    },
+    {
+      "epoch": 2.9248,
+      "grad_norm": 0.5780746858570248,
+      "learning_rate": 1.9961945848213092e-08,
+      "loss": 1.293,
+      "step": 1828
+    },
+    {
+      "epoch": 2.9264,
+      "grad_norm": 0.5417375634162136,
+      "learning_rate": 1.9139387665247922e-08,
+      "loss": 1.3201,
+      "step": 1829
+    },
+    {
+      "epoch": 2.928,
+      "grad_norm": 0.5523903432136753,
+      "learning_rate": 1.8334102723703286e-08,
+      "loss": 1.3458,
+      "step": 1830
+    },
+    {
+      "epoch": 2.9295999999999998,
+      "grad_norm": 0.5443868077260675,
+      "learning_rate": 1.754609381624639e-08,
+      "loss": 1.2336,
+      "step": 1831
+    },
+    {
+      "epoch": 2.9312,
+      "grad_norm": 0.5457778893362586,
+      "learning_rate": 1.677536367563126e-08,
+      "loss": 1.328,
+      "step": 1832
+    },
+    {
+      "epoch": 2.9328,
+      "grad_norm": 0.5454037603298351,
+      "learning_rate": 1.6021914974690413e-08,
+      "loss": 1.2408,
+      "step": 1833
+    },
+    {
+      "epoch": 2.9344,
+      "grad_norm": 0.5583448097371418,
+      "learning_rate": 1.5285750326325953e-08,
+      "loss": 1.2674,
+      "step": 1834
+    },
+    {
+      "epoch": 2.936,
+      "grad_norm": 0.5668230258107845,
+      "learning_rate": 1.4566872283500733e-08,
+      "loss": 1.271,
+      "step": 1835
+    },
+    {
+      "epoch": 2.9375999999999998,
+      "grad_norm": 0.5482869178771883,
+      "learning_rate": 1.3865283339228319e-08,
+      "loss": 1.2562,
+      "step": 1836
+    },
+    {
+      "epoch": 2.9392,
+      "grad_norm": 0.5507018654403958,
+      "learning_rate": 1.3180985926564693e-08,
+      "loss": 1.2774,
+      "step": 1837
+    },
+    {
+      "epoch": 2.9408,
+      "grad_norm": 0.5473348063102211,
+      "learning_rate": 1.2513982418601024e-08,
+      "loss": 1.2395,
+      "step": 1838
+    },
+    {
+      "epoch": 2.9424,
+      "grad_norm": 0.5668009078025288,
+      "learning_rate": 1.1864275128454783e-08,
+      "loss": 1.2864,
+      "step": 1839
+    },
+    {
+      "epoch": 2.944,
+      "grad_norm": 0.5391472255038026,
+      "learning_rate": 1.1231866309259764e-08,
+      "loss": 1.2724,
+      "step": 1840
+    },
+    {
+      "epoch": 2.9455999999999998,
+      "grad_norm": 0.5421694153454758,
+      "learning_rate": 1.0616758154161633e-08,
+      "loss": 1.237,
+      "step": 1841
+    },
+    {
+      "epoch": 2.9472,
+      "grad_norm": 0.5657024622516142,
+      "learning_rate": 1.0018952796307934e-08,
+      "loss": 1.2832,
+      "step": 1842
+    },
+    {
+      "epoch": 2.9488,
+      "grad_norm": 0.5496610601217906,
+      "learning_rate": 9.438452308841995e-09,
+      "loss": 1.3006,
+      "step": 1843
+    },
+    {
+      "epoch": 2.9504,
+      "grad_norm": 0.5327307639429272,
+      "learning_rate": 8.87525870489514e-09,
+      "loss": 1.2173,
+      "step": 1844
+    },
+    {
+      "epoch": 2.952,
+      "grad_norm": 0.546440924950516,
+      "learning_rate": 8.329373937578378e-09,
+      "loss": 1.2915,
+      "step": 1845
+    },
+    {
+      "epoch": 2.9536,
+      "grad_norm": 0.5515127668393153,
+      "learning_rate": 7.800799899979061e-09,
+      "loss": 1.3057,
+      "step": 1846
+    },
+    {
+      "epoch": 2.9552,
+      "grad_norm": 0.5558634581954316,
+      "learning_rate": 7.289538425150899e-09,
+      "loss": 1.3002,
+      "step": 1847
+    },
+    {
+      "epoch": 2.9568,
+      "grad_norm": 0.5711742561472705,
+      "learning_rate": 6.7955912861095155e-09,
+      "loss": 1.2389,
+      "step": 1848
+    },
+    {
+      "epoch": 2.9584,
+      "grad_norm": 0.5432936387646515,
+      "learning_rate": 6.31896019582523e-09,
+      "loss": 1.2736,
+      "step": 1849
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 0.5616986791759175,
+      "learning_rate": 5.8596468072180665e-09,
+      "loss": 1.1747,
+      "step": 1850
+    },
+    {
+      "epoch": 2.9616,
+      "grad_norm": 0.5677288266095453,
+      "learning_rate": 5.417652713152199e-09,
+      "loss": 1.2398,
+      "step": 1851
+    },
+    {
+      "epoch": 2.9632,
+      "grad_norm": 0.5712537199373301,
+      "learning_rate": 4.992979446428736e-09,
+      "loss": 1.2518,
+      "step": 1852
+    },
+    {
+      "epoch": 2.9648,
+      "grad_norm": 0.5471915227156728,
+      "learning_rate": 4.585628479781279e-09,
+      "loss": 1.2514,
+      "step": 1853
+    },
+    {
+      "epoch": 2.9664,
+      "grad_norm": 0.546884922246981,
+      "learning_rate": 4.195601225872592e-09,
+      "loss": 1.2436,
+      "step": 1854
+    },
+    {
+      "epoch": 2.968,
+      "grad_norm": 0.5618097409745267,
+      "learning_rate": 3.822899037286276e-09,
+      "loss": 1.2949,
+      "step": 1855
+    },
+    {
+      "epoch": 2.9696,
+      "grad_norm": 0.5886138892603576,
+      "learning_rate": 3.4675232065256583e-09,
+      "loss": 1.2038,
+      "step": 1856
+    },
+    {
+      "epoch": 2.9712,
+      "grad_norm": 0.5377709529757217,
+      "learning_rate": 3.129474966006574e-09,
+      "loss": 1.2401,
+      "step": 1857
+    },
+    {
+      "epoch": 2.9728,
+      "grad_norm": 0.5439473347454598,
+      "learning_rate": 2.808755488054038e-09,
+      "loss": 1.336,
+      "step": 1858
+    },
+    {
+      "epoch": 2.9744,
+      "grad_norm": 0.5723116937874024,
+      "learning_rate": 2.5053658848989137e-09,
+      "loss": 1.3087,
+      "step": 1859
+    },
+    {
+      "epoch": 2.976,
+      "grad_norm": 0.5785610708085795,
+      "learning_rate": 2.219307208672361e-09,
+      "loss": 1.2539,
+      "step": 1860
+    },
+    {
+      "epoch": 2.9776,
+      "grad_norm": 0.5434356087541273,
+      "learning_rate": 1.9505804514047266e-09,
+      "loss": 1.1983,
+      "step": 1861
+    },
+    {
+      "epoch": 2.9792,
+      "grad_norm": 0.5624167534914531,
+      "learning_rate": 1.6991865450188827e-09,
+      "loss": 1.2534,
+      "step": 1862
+    },
+    {
+      "epoch": 2.9808,
+      "grad_norm": 0.5433474098577605,
+      "learning_rate": 1.465126361330227e-09,
+      "loss": 1.2617,
+      "step": 1863
+    },
+    {
+      "epoch": 2.9824,
+      "grad_norm": 0.5508551148093853,
+      "learning_rate": 1.2484007120411312e-09,
+      "loss": 1.235,
+      "step": 1864
+    },
+    {
+      "epoch": 2.984,
+      "grad_norm": 0.5479685723139754,
+      "learning_rate": 1.0490103487392766e-09,
+      "loss": 1.2808,
+      "step": 1865
+    },
+    {
+      "epoch": 2.9856,
+      "grad_norm": 0.577054439506447,
+      "learning_rate": 8.669559628954327e-10,
+      "loss": 1.2614,
+      "step": 1866
+    },
+    {
+      "epoch": 2.9872,
+      "grad_norm": 0.6032830455473198,
+      "learning_rate": 7.02238185860682e-10,
+      "loss": 1.3016,
+      "step": 1867
+    },
+    {
+      "epoch": 2.9888,
+      "grad_norm": 0.5484454257718967,
+      "learning_rate": 5.54857588862534e-10,
+      "loss": 1.3108,
+      "step": 1868
+    },
+    {
+      "epoch": 2.9904,
+      "grad_norm": 0.5550007614202449,
+      "learning_rate": 4.2481468300603625e-10,
+      "loss": 1.2465,
+      "step": 1869
+    },
+    {
+      "epoch": 2.992,
+      "grad_norm": 0.562128424684606,
+      "learning_rate": 3.1210991927044244e-10,
+      "loss": 1.2641,
+      "step": 1870
+    },
+    {
+      "epoch": 2.9936,
+      "grad_norm": 0.5449954023145134,
+      "learning_rate": 2.167436885064378e-10,
+      "loss": 1.2595,
+      "step": 1871
+    },
+    {
+      "epoch": 2.9952,
+      "grad_norm": 0.5412493384642634,
+      "learning_rate": 1.387163214372489e-10,
+      "loss": 1.3001,
+      "step": 1872
+    },
+    {
+      "epoch": 2.9968,
+      "grad_norm": 0.5461775541428863,
+      "learning_rate": 7.80280886558682e-11,
+      "loss": 1.3092,
+      "step": 1873
+    },
+    {
+      "epoch": 2.9984,
+      "grad_norm": 0.5825810868073731,
+      "learning_rate": 3.467920062394381e-11,
+      "loss": 1.2788,
+      "step": 1874
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.546220196616214,
+      "learning_rate": 8.669807672334606e-12,
+      "loss": 1.2903,
+      "step": 1875
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1875,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 2000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1439107152936960.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}