diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,17530 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 2500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.929184960782006,
+      "learning_rate": 2.666666666666667e-07,
+      "loss": 12.7783,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.874932389885064,
+      "learning_rate": 5.333333333333335e-07,
+      "loss": 12.7829,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.638474194681965,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 12.7492,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.691613562458222,
+      "learning_rate": 1.066666666666667e-06,
+      "loss": 12.7644,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.183637154093034,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 12.7361,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.394137566345053,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 12.7412,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 16.41810330919611,
+      "learning_rate": 1.8666666666666669e-06,
+      "loss": 12.7497,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.703173654632728,
+      "learning_rate": 2.133333333333334e-06,
+      "loss": 12.6923,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.28258963587572,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 12.633,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.979699480878217,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 12.4999,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.901151346983722,
+      "learning_rate": 2.9333333333333338e-06,
+      "loss": 12.4535,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.29923785648717,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 12.3986,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.471178506450912,
+      "learning_rate": 3.4666666666666672e-06,
+      "loss": 11.9871,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.838514649534746,
+      "learning_rate": 3.7333333333333337e-06,
+      "loss": 11.9267,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.978160035469458,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 11.8394,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.141794873591074,
+      "learning_rate": 4.266666666666668e-06,
+      "loss": 11.7715,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.298617438327794,
+      "learning_rate": 4.533333333333334e-06,
+      "loss": 11.3915,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.25498056051837,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 11.17,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.283997955198105,
+      "learning_rate": 5.0666666666666676e-06,
+      "loss": 11.0347,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.780727278149014,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 10.8716,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.854956693311563,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 10.8734,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.375861574845129,
+      "learning_rate": 5.8666666666666675e-06,
+      "loss": 10.7295,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.478804987639091,
+      "learning_rate": 6.133333333333334e-06,
+      "loss": 10.5077,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.061475152313659,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 10.4395,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.26564612763676,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 10.2625,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.538214638535141,
+      "learning_rate": 6.9333333333333344e-06,
+      "loss": 10.1915,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.337218899132541,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 10.1794,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.415165410790723,
+      "learning_rate": 7.4666666666666675e-06,
+      "loss": 10.2069,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.187915970859848,
+      "learning_rate": 7.733333333333334e-06,
+      "loss": 9.895,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.631030212687751,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 9.9977,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.631218434317244,
+      "learning_rate": 8.266666666666667e-06,
+      "loss": 9.8149,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.8158165880323365,
+      "learning_rate": 8.533333333333335e-06,
+      "loss": 9.7459,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.841399667684455,
+      "learning_rate": 8.8e-06,
+      "loss": 9.7156,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.610263893321706,
+      "learning_rate": 9.066666666666667e-06,
+      "loss": 9.6304,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.235066753616497,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 9.41,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.452655010828091,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 9.498,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.938743934202386,
+      "learning_rate": 9.866666666666668e-06,
+      "loss": 9.3042,
+      "step": 37
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9544713684597945,
+      "learning_rate": 1.0133333333333335e-05,
+      "loss": 9.4645,
+      "step": 38
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6572227241816835,
+      "learning_rate": 1.04e-05,
+      "loss": 9.2902,
+      "step": 39
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.466505871627913,
+      "learning_rate": 1.0666666666666667e-05,
+      "loss": 9.3163,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.778564965485367,
+      "learning_rate": 1.0933333333333334e-05,
+      "loss": 9.1139,
+      "step": 41
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.410075586798703,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 9.0775,
+      "step": 42
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.122133392766589,
+      "learning_rate": 1.1466666666666668e-05,
+      "loss": 9.146,
+      "step": 43
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6810042264298595,
+      "learning_rate": 1.1733333333333335e-05,
+      "loss": 9.0396,
+      "step": 44
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.227042286632877,
+      "learning_rate": 1.2e-05,
+      "loss": 9.0946,
+      "step": 45
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.629133823957898,
+      "learning_rate": 1.2266666666666667e-05,
+      "loss": 8.9797,
+      "step": 46
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7271542497913503,
+      "learning_rate": 1.2533333333333336e-05,
+      "loss": 8.9485,
+      "step": 47
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.242646810059511,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 8.8412,
+      "step": 48
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.076863398153946,
+      "learning_rate": 1.3066666666666668e-05,
+      "loss": 8.9591,
+      "step": 49
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.744253371158508,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 8.849,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.4165606869895475,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 8.8491,
+      "step": 51
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.857593202175524,
+      "learning_rate": 1.3866666666666669e-05,
+      "loss": 8.8226,
+      "step": 52
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.475861448921801,
+      "learning_rate": 1.4133333333333334e-05,
+      "loss": 8.808,
+      "step": 53
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.8910608136482905,
+      "learning_rate": 1.4400000000000001e-05,
+      "loss": 8.6174,
+      "step": 54
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.759581203767945,
+      "learning_rate": 1.4666666666666666e-05,
+      "loss": 8.7544,
+      "step": 55
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.430056602810506,
+      "learning_rate": 1.4933333333333335e-05,
+      "loss": 8.7811,
+      "step": 56
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6362047197084126,
+      "learning_rate": 1.5200000000000002e-05,
+      "loss": 8.6588,
+      "step": 57
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.307397557457014,
+      "learning_rate": 1.546666666666667e-05,
+      "loss": 8.6764,
+      "step": 58
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.163868735280828,
+      "learning_rate": 1.5733333333333334e-05,
+      "loss": 8.5804,
+      "step": 59
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.926174882030974,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 8.4866,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.622342586255422,
+      "learning_rate": 1.6266666666666668e-05,
+      "loss": 8.4828,
+      "step": 61
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.589183178299784,
+      "learning_rate": 1.6533333333333333e-05,
+      "loss": 8.5785,
+      "step": 62
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.736115426468919,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 8.3527,
+      "step": 63
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6673697167117676,
+      "learning_rate": 1.706666666666667e-05,
+      "loss": 8.4629,
+      "step": 64
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.672070519476944,
+      "learning_rate": 1.7333333333333336e-05,
+      "loss": 8.2062,
+      "step": 65
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.187720892130302,
+      "learning_rate": 1.76e-05,
+      "loss": 8.3856,
+      "step": 66
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.15197703033149,
+      "learning_rate": 1.7866666666666666e-05,
+      "loss": 8.2769,
+      "step": 67
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.912600166722876,
+      "learning_rate": 1.8133333333333335e-05,
+      "loss": 8.3477,
+      "step": 68
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.244802476034463,
+      "learning_rate": 1.8400000000000003e-05,
+      "loss": 8.1459,
+      "step": 69
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.722980011319047,
+      "learning_rate": 1.866666666666667e-05,
+      "loss": 8.2767,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.174831807901733,
+      "learning_rate": 1.8933333333333334e-05,
+      "loss": 8.1762,
+      "step": 71
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.528943545214611,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 8.2168,
+      "step": 72
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.639645536558215,
+      "learning_rate": 1.9466666666666668e-05,
+      "loss": 8.1628,
+      "step": 73
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.52886572716252,
+      "learning_rate": 1.9733333333333336e-05,
+      "loss": 8.1529,
+      "step": 74
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.075612772806507,
+      "learning_rate": 2e-05,
+      "loss": 7.8176,
+      "step": 75
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.783051863921666,
+      "learning_rate": 1.9999991608372392e-05,
+      "loss": 8.028,
+      "step": 76
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4072414013421133,
+      "learning_rate": 1.999996643350365e-05,
+      "loss": 8.018,
+      "step": 77
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.2518888973752365,
+      "learning_rate": 1.999992447543603e-05,
+      "loss": 8.0281,
+      "step": 78
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.342242469543838,
+      "learning_rate": 1.999986573423995e-05,
+      "loss": 7.7778,
+      "step": 79
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.492869195214892,
+      "learning_rate": 1.999979021001399e-05,
+      "loss": 7.9055,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.969553752122128,
+      "learning_rate": 1.999969790288491e-05,
+      "loss": 7.8983,
+      "step": 81
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.683302910481509,
+      "learning_rate": 1.999958881300763e-05,
+      "loss": 7.7587,
+      "step": 82
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.970945676308067,
+      "learning_rate": 1.9999462940565242e-05,
+      "loss": 7.7445,
+      "step": 83
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.102972572272466,
+      "learning_rate": 1.9999320285769e-05,
+      "loss": 7.73,
+      "step": 84
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.501542226169796,
+      "learning_rate": 1.999916084885832e-05,
+      "loss": 7.7589,
+      "step": 85
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.2240889084532633,
+      "learning_rate": 1.999898463010079e-05,
+      "loss": 7.7714,
+      "step": 86
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.2224615585033245,
+      "learning_rate": 1.9998791629792172e-05,
+      "loss": 7.863,
+      "step": 87
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.918041691814854,
+      "learning_rate": 1.999858184825637e-05,
+      "loss": 7.6648,
+      "step": 88
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.96906876524204,
+      "learning_rate": 1.9998355285845473e-05,
+      "loss": 7.6013,
+      "step": 89
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.462241702890193,
+      "learning_rate": 1.9998111942939727e-05,
+      "loss": 7.6095,
+      "step": 90
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.324181915141825,
+      "learning_rate": 1.9997851819947537e-05,
+      "loss": 7.5812,
+      "step": 91
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.031985651604528,
+      "learning_rate": 1.999757491730548e-05,
+      "loss": 7.6367,
+      "step": 92
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.638146112260251,
+      "learning_rate": 1.999728123547828e-05,
+      "loss": 7.5123,
+      "step": 93
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.290147893102006,
+      "learning_rate": 1.9996970774958836e-05,
+      "loss": 7.3667,
+      "step": 94
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6217639310893253,
+      "learning_rate": 1.9996643536268202e-05,
+      "loss": 7.3957,
+      "step": 95
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.565797170343041,
+      "learning_rate": 1.999629951995559e-05,
+      "loss": 7.4662,
+      "step": 96
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3508669824426702,
+      "learning_rate": 1.9995938726598374e-05,
+      "loss": 7.378,
+      "step": 97
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.424192916253537,
+      "learning_rate": 1.999556115680208e-05,
+      "loss": 7.3639,
+      "step": 98
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.232652541715236,
+      "learning_rate": 1.999516681120039e-05,
+      "loss": 7.2653,
+      "step": 99
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.129496340056508,
+      "learning_rate": 1.9994755690455154e-05,
+      "loss": 7.4364,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.233080193988463,
+      "learning_rate": 1.999432779525635e-05,
+      "loss": 7.2178,
+      "step": 101
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.176827501719459,
+      "learning_rate": 1.9993883126322142e-05,
+      "loss": 7.254,
+      "step": 102
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.758141329766479,
+      "learning_rate": 1.9993421684398825e-05,
+      "loss": 7.1993,
+      "step": 103
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.405265927651899,
+      "learning_rate": 1.9992943470260845e-05,
+      "loss": 7.3139,
+      "step": 104
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.152161371033381,
+      "learning_rate": 1.99924484847108e-05,
+      "loss": 7.0201,
+      "step": 105
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7267363140868968,
+      "learning_rate": 1.9991936728579438e-05,
+      "loss": 7.0296,
+      "step": 106
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.3497962003683925,
+      "learning_rate": 1.999140820272566e-05,
+      "loss": 7.2447,
+      "step": 107
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.298332761550472,
+      "learning_rate": 1.9990862908036492e-05,
+      "loss": 7.0496,
+      "step": 108
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.846041734386933,
+      "learning_rate": 1.9990300845427123e-05,
+      "loss": 7.015,
+      "step": 109
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.155174975783645,
+      "learning_rate": 1.998972201584088e-05,
+      "loss": 6.952,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.420517963314204,
+      "learning_rate": 1.998912642024922e-05,
+      "loss": 7.0024,
+      "step": 111
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.174524593326725,
+      "learning_rate": 1.998851405965175e-05,
+      "loss": 6.9717,
+      "step": 112
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.633064351625967,
+      "learning_rate": 1.9987884935076213e-05,
+      "loss": 7.0441,
+      "step": 113
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.841624516136794,
+      "learning_rate": 1.9987239047578482e-05,
+      "loss": 6.9816,
+      "step": 114
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.012980808319145,
+      "learning_rate": 1.9986576398242566e-05,
+      "loss": 6.8687,
+      "step": 115
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.412369426332821,
+      "learning_rate": 1.9985896988180607e-05,
+      "loss": 6.9009,
+      "step": 116
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.597355208644285,
+      "learning_rate": 1.9985200818532873e-05,
+      "loss": 6.9828,
+      "step": 117
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7175005237062635,
+      "learning_rate": 1.9984487890467773e-05,
+      "loss": 7.0573,
+      "step": 118
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.03043168020476,
+      "learning_rate": 1.9983758205181824e-05,
+      "loss": 6.8331,
+      "step": 119
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.88783774497073,
+      "learning_rate": 1.9983011763899674e-05,
+      "loss": 6.6711,
+      "step": 120
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.280680262363997,
+      "learning_rate": 1.9982248567874098e-05,
+      "loss": 6.8023,
+      "step": 121
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.577251349440423,
+      "learning_rate": 1.998146861838599e-05,
+      "loss": 6.8457,
+      "step": 122
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5891117859872472,
+      "learning_rate": 1.9980671916744356e-05,
+      "loss": 6.8008,
+      "step": 123
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.099854105360618,
+      "learning_rate": 1.9979858464286317e-05,
+      "loss": 6.726,
+      "step": 124
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9821886115445677,
+      "learning_rate": 1.997902826237712e-05,
+      "loss": 6.6006,
+      "step": 125
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.681735516522654,
+      "learning_rate": 1.9978181312410104e-05,
+      "loss": 6.7371,
+      "step": 126
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4290275825448147,
+      "learning_rate": 1.9977317615806738e-05,
+      "loss": 6.7282,
+      "step": 127
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.803532374626138,
+      "learning_rate": 1.9976437174016575e-05,
+      "loss": 6.6815,
+      "step": 128
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.32888767020378,
+      "learning_rate": 1.997553998851729e-05,
+      "loss": 6.6643,
+      "step": 129
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.066789919091347,
+      "learning_rate": 1.997462606081465e-05,
+      "loss": 6.6853,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.708099567003482,
+      "learning_rate": 1.997369539244252e-05,
+      "loss": 6.7651,
+      "step": 131
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.245589097217936,
+      "learning_rate": 1.997274798496287e-05,
+      "loss": 6.6351,
+      "step": 132
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.240512001177884,
+      "learning_rate": 1.9971783839965756e-05,
+      "loss": 6.509,
+      "step": 133
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9243143342593974,
+      "learning_rate": 1.997080295906933e-05,
+      "loss": 6.6422,
+      "step": 134
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9373753907930666,
+      "learning_rate": 1.9969805343919822e-05,
+      "loss": 6.6452,
+      "step": 135
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.29054775888337,
+      "learning_rate": 1.996879099619156e-05,
+      "loss": 6.5095,
+      "step": 136
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.39805253052477,
+      "learning_rate": 1.9967759917586953e-05,
+      "loss": 6.4897,
+      "step": 137
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7120504701706247,
+      "learning_rate": 1.9966712109836476e-05,
+      "loss": 6.4993,
+      "step": 138
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.288830803394844,
+      "learning_rate": 1.9965647574698705e-05,
+      "loss": 6.5248,
+      "step": 139
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.908954938672954,
+      "learning_rate": 1.9964566313960265e-05,
+      "loss": 6.382,
+      "step": 140
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.400649085165072,
+      "learning_rate": 1.9963468329435872e-05,
+      "loss": 6.6161,
+      "step": 141
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6352934691721006,
+      "learning_rate": 1.9962353622968296e-05,
+      "loss": 6.2676,
+      "step": 142
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.869261630058483,
+      "learning_rate": 1.996122219642838e-05,
+      "loss": 6.6562,
+      "step": 143
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.386275014923958,
+      "learning_rate": 1.9960074051715022e-05,
+      "loss": 6.4506,
+      "step": 144
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.209204783609756,
+      "learning_rate": 1.995890919075519e-05,
+      "loss": 6.3668,
+      "step": 145
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.247052362466022,
+      "learning_rate": 1.995772761550389e-05,
+      "loss": 6.3787,
+      "step": 146
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.27883778192868,
+      "learning_rate": 1.9956529327944198e-05,
+      "loss": 6.3488,
+      "step": 147
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.635114569082134,
+      "learning_rate": 1.9955314330087225e-05,
+      "loss": 6.343,
+      "step": 148
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8840203593424962,
+      "learning_rate": 1.9954082623972143e-05,
+      "loss": 6.3235,
+      "step": 149
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5382476230277056,
+      "learning_rate": 1.995283421166614e-05,
+      "loss": 6.3011,
+      "step": 150
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.234026714115879,
+      "learning_rate": 1.9951569095264473e-05,
+      "loss": 6.2582,
+      "step": 151
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4587502541675947,
+      "learning_rate": 1.995028727689041e-05,
+      "loss": 6.3841,
+      "step": 152
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.061472086983153,
+      "learning_rate": 1.9948988758695263e-05,
+      "loss": 6.1556,
+      "step": 153
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.147311138116654,
+      "learning_rate": 1.994767354285837e-05,
+      "loss": 6.1378,
+      "step": 154
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.231891466089776,
+      "learning_rate": 1.9946341631587086e-05,
+      "loss": 6.2871,
+      "step": 155
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.434419768405808,
+      "learning_rate": 1.9944993027116798e-05,
+      "loss": 6.332,
+      "step": 156
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.531188775167551,
+      "learning_rate": 1.9943627731710896e-05,
+      "loss": 6.0622,
+      "step": 157
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.843702023801363,
+      "learning_rate": 1.9942245747660797e-05,
+      "loss": 6.2306,
+      "step": 158
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.766495021175679,
+      "learning_rate": 1.9940847077285918e-05,
+      "loss": 6.1657,
+      "step": 159
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.226745630752804,
+      "learning_rate": 1.9939431722933678e-05,
+      "loss": 6.2713,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.213643944441643,
+      "learning_rate": 1.993799968697951e-05,
+      "loss": 6.2101,
+      "step": 161
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.320563316825058,
+      "learning_rate": 1.9936550971826835e-05,
+      "loss": 6.1438,
+      "step": 162
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.816127954551905,
+      "learning_rate": 1.9935085579907064e-05,
+      "loss": 6.2491,
+      "step": 163
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.6329522483416845,
+      "learning_rate": 1.9933603513679604e-05,
+      "loss": 6.1279,
+      "step": 164
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.027487870638996,
+      "learning_rate": 1.9932104775631847e-05,
+      "loss": 6.0602,
+      "step": 165
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.98578715574418,
+      "learning_rate": 1.993058936827916e-05,
+      "loss": 6.0202,
+      "step": 166
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.604110505866004,
+      "learning_rate": 1.9929057294164894e-05,
+      "loss": 6.0824,
+      "step": 167
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.246568749376813,
+      "learning_rate": 1.992750855586036e-05,
+      "loss": 5.9831,
+      "step": 168
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7019710360614204,
+      "learning_rate": 1.9925943155964857e-05,
+      "loss": 6.0617,
+      "step": 169
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.276255882489964,
+      "learning_rate": 1.9924361097105624e-05,
+      "loss": 5.8458,
+      "step": 170
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.92369718317991,
+      "learning_rate": 1.992276238193788e-05,
+      "loss": 5.9862,
+      "step": 171
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.178133635023751,
+      "learning_rate": 1.9921147013144782e-05,
+      "loss": 5.8947,
+      "step": 172
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.894193449640341,
+      "learning_rate": 1.9919514993437445e-05,
+      "loss": 5.864,
+      "step": 173
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.701018088179099,
+      "learning_rate": 1.9917866325554936e-05,
+      "loss": 6.0865,
+      "step": 174
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.066248382914103,
+      "learning_rate": 1.9916201012264255e-05,
+      "loss": 6.032,
+      "step": 175
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.103277009648441,
+      "learning_rate": 1.991451905636033e-05,
+      "loss": 6.1079,
+      "step": 176
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.025146053384431,
+      "learning_rate": 1.9912820460666046e-05,
+      "loss": 5.9556,
+      "step": 177
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.297603739475917,
+      "learning_rate": 1.9911105228032186e-05,
+      "loss": 5.9033,
+      "step": 178
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.973721423418755,
+      "learning_rate": 1.9909373361337475e-05,
+      "loss": 5.9913,
+      "step": 179
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7597853707129434,
+      "learning_rate": 1.990762486348855e-05,
+      "loss": 5.91,
+      "step": 180
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.429544778679046,
+      "learning_rate": 1.990585973741996e-05,
+      "loss": 5.9034,
+      "step": 181
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.746428539982332,
+      "learning_rate": 1.9904077986094153e-05,
+      "loss": 5.8531,
+      "step": 182
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.325501317331493,
+      "learning_rate": 1.9902279612501494e-05,
+      "loss": 5.8167,
+      "step": 183
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.410086017324782,
+      "learning_rate": 1.9900464619660243e-05,
+      "loss": 5.7331,
+      "step": 184
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.99166656509949,
+      "learning_rate": 1.989863301061654e-05,
+      "loss": 5.9225,
+      "step": 185
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.535795454860607,
+      "learning_rate": 1.989678478844443e-05,
+      "loss": 5.8929,
+      "step": 186
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.547069866912832,
+      "learning_rate": 1.9894919956245825e-05,
+      "loss": 5.9094,
+      "step": 187
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4283860612781245,
+      "learning_rate": 1.9893038517150526e-05,
+      "loss": 5.9043,
+      "step": 188
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.162844234054413,
+      "learning_rate": 1.9891140474316197e-05,
+      "loss": 5.7997,
+      "step": 189
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.305168114018011,
+      "learning_rate": 1.9889225830928365e-05,
+      "loss": 5.7349,
+      "step": 190
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.055746785153127,
+      "learning_rate": 1.9887294590200437e-05,
+      "loss": 5.8111,
+      "step": 191
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7328825091394324,
+      "learning_rate": 1.988534675537366e-05,
+      "loss": 5.8384,
+      "step": 192
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1019661386474735,
+      "learning_rate": 1.988338232971713e-05,
+      "loss": 5.8723,
+      "step": 193
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.497989489932904,
+      "learning_rate": 1.9881401316527795e-05,
+      "loss": 5.9128,
+      "step": 194
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.8831242401657087,
+      "learning_rate": 1.987940371913044e-05,
+      "loss": 5.8428,
+      "step": 195
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.174688702002974,
+      "learning_rate": 1.9877389540877686e-05,
+      "loss": 5.7467,
+      "step": 196
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.055915757767483,
+      "learning_rate": 1.9875358785149982e-05,
+      "loss": 5.7186,
+      "step": 197
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.782137439007435,
+      "learning_rate": 1.987331145535559e-05,
+      "loss": 5.8389,
+      "step": 198
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2379181329566777,
+      "learning_rate": 1.98712475549306e-05,
+      "loss": 5.7215,
+      "step": 199
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6456344440328845,
+      "learning_rate": 1.9869167087338908e-05,
+      "loss": 5.618,
+      "step": 200
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7619772961608255,
+      "learning_rate": 1.9867070056072215e-05,
+      "loss": 5.8516,
+      "step": 201
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5740327111798322,
+      "learning_rate": 1.9864956464650027e-05,
+      "loss": 5.7066,
+      "step": 202
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.033968317687985,
+      "learning_rate": 1.986282631661963e-05,
+      "loss": 5.6463,
+      "step": 203
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.644873695688273,
+      "learning_rate": 1.9860679615556112e-05,
+      "loss": 5.5472,
+      "step": 204
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2695708458307617,
+      "learning_rate": 1.9858516365062334e-05,
+      "loss": 5.7738,
+      "step": 205
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.43428089723817,
+      "learning_rate": 1.9856336568768936e-05,
+      "loss": 5.6885,
+      "step": 206
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.203310963319522,
+      "learning_rate": 1.9854140230334323e-05,
+      "loss": 5.6778,
+      "step": 207
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4114108774306633,
+      "learning_rate": 1.985192735344467e-05,
+      "loss": 5.9011,
+      "step": 208
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.125731707197647,
+      "learning_rate": 1.98496979418139e-05,
+      "loss": 5.5251,
+      "step": 209
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.558175716829582,
+      "learning_rate": 1.9847451999183692e-05,
+      "loss": 5.5834,
+      "step": 210
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1705697609316523,
+      "learning_rate": 1.9845189529323473e-05,
+      "loss": 5.6806,
+      "step": 211
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.375356341487538,
+      "learning_rate": 1.98429105360304e-05,
+      "loss": 5.6382,
+      "step": 212
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.886412279702093,
+      "learning_rate": 1.9840615023129372e-05,
+      "loss": 5.5464,
+      "step": 213
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.003221004452615,
+      "learning_rate": 1.9838302994473e-05,
+      "loss": 5.5527,
+      "step": 214
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2142950313355976,
+      "learning_rate": 1.9835974453941623e-05,
+      "loss": 5.4318,
+      "step": 215
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.124707418757366,
+      "learning_rate": 1.9833629405443283e-05,
+      "loss": 5.4974,
+      "step": 216
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.162365896102337,
+      "learning_rate": 1.983126785291375e-05,
+      "loss": 5.5908,
+      "step": 217
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5671782583753178,
+      "learning_rate": 1.9828889800316467e-05,
+      "loss": 5.5886,
+      "step": 218
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.552839596348838,
+      "learning_rate": 1.982649525164258e-05,
+      "loss": 5.527,
+      "step": 219
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.338000185805563,
+      "learning_rate": 1.9824084210910924e-05,
+      "loss": 5.6058,
+      "step": 220
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.314810218621173,
+      "learning_rate": 1.9821656682168013e-05,
+      "loss": 5.6443,
+      "step": 221
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.124905233622171,
+      "learning_rate": 1.9819212669488026e-05,
+      "loss": 5.8652,
+      "step": 222
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.601461559913127,
+      "learning_rate": 1.9816752176972815e-05,
+      "loss": 5.6487,
+      "step": 223
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.682447349841513,
+      "learning_rate": 1.9814275208751882e-05,
+      "loss": 5.4415,
+      "step": 224
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.092388938276944,
+      "learning_rate": 1.9811781768982392e-05,
+      "loss": 5.4588,
+      "step": 225
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.5748442841522134,
+      "learning_rate": 1.9809271861849147e-05,
+      "loss": 5.4366,
+      "step": 226
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6795744789993554,
+      "learning_rate": 1.9806745491564588e-05,
+      "loss": 5.5378,
+      "step": 227
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.236814926224578,
+      "learning_rate": 1.9804202662368782e-05,
+      "loss": 5.5051,
+      "step": 228
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8825245603921315,
+      "learning_rate": 1.980164337852943e-05,
+      "loss": 5.4982,
+      "step": 229
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.689216752889219,
+      "learning_rate": 1.9799067644341844e-05,
+      "loss": 5.4037,
+      "step": 230
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.498751852842296,
+      "learning_rate": 1.9796475464128943e-05,
+      "loss": 5.5379,
+      "step": 231
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2926041035928995,
+      "learning_rate": 1.9793866842241245e-05,
+      "loss": 5.6656,
+      "step": 232
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.563969254045201,
+      "learning_rate": 1.9791241783056874e-05,
+      "loss": 5.5352,
+      "step": 233
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9349353489105567,
+      "learning_rate": 1.9788600290981525e-05,
+      "loss": 5.6132,
+      "step": 234
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.461095251707113,
+      "learning_rate": 1.978594237044849e-05,
+      "loss": 5.5776,
+      "step": 235
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.128877934256824,
+      "learning_rate": 1.9783268025918622e-05,
+      "loss": 5.3886,
+      "step": 236
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.340131228111977,
+      "learning_rate": 1.9780577261880336e-05,
+      "loss": 5.4318,
+      "step": 237
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5405936438145473,
+      "learning_rate": 1.977787008284962e-05,
+      "loss": 5.612,
+      "step": 238
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9277621763337125,
+      "learning_rate": 1.9775146493369996e-05,
+      "loss": 5.5956,
+      "step": 239
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.189175924053646,
+      "learning_rate": 1.977240649801253e-05,
+      "loss": 5.4193,
+      "step": 240
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7261802456797044,
+      "learning_rate": 1.9769650101375835e-05,
+      "loss": 5.4232,
+      "step": 241
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.461796387404502,
+      "learning_rate": 1.9766877308086038e-05,
+      "loss": 5.5134,
+      "step": 242
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.061775651187604,
+      "learning_rate": 1.9764088122796785e-05,
+      "loss": 5.4679,
+      "step": 243
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.965102040882132,
+      "learning_rate": 1.976128255018924e-05,
+      "loss": 5.5467,
+      "step": 244
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.676829205300839,
+      "learning_rate": 1.9758460594972068e-05,
+      "loss": 5.4451,
+      "step": 245
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.911400602977502,
+      "learning_rate": 1.975562226188143e-05,
+      "loss": 5.4803,
+      "step": 246
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.282418377084345,
+      "learning_rate": 1.9752767555680967e-05,
+      "loss": 5.286,
+      "step": 247
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.308118562484986,
+      "learning_rate": 1.9749896481161807e-05,
+      "loss": 5.4824,
+      "step": 248
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.263589886497265,
+      "learning_rate": 1.9747009043142556e-05,
+      "loss": 5.5256,
+      "step": 249
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.827111939511474,
+      "learning_rate": 1.9744105246469264e-05,
+      "loss": 5.5244,
+      "step": 250
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.403518844265166,
+      "learning_rate": 1.974118509601545e-05,
+      "loss": 5.3989,
+      "step": 251
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.720678665989358,
+      "learning_rate": 1.9738248596682078e-05,
+      "loss": 5.4725,
+      "step": 252
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.631390734909404,
+      "learning_rate": 1.973529575339755e-05,
+      "loss": 5.4297,
+      "step": 253
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.629825341493495,
+      "learning_rate": 1.9732326571117703e-05,
+      "loss": 5.3291,
+      "step": 254
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.527878508159066,
+      "learning_rate": 1.9729341054825783e-05,
+      "loss": 5.4482,
+      "step": 255
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.041074384352786,
+      "learning_rate": 1.9726339209532462e-05,
+      "loss": 5.4306,
+      "step": 256
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7876647301759494,
+      "learning_rate": 1.9723321040275816e-05,
+      "loss": 5.3598,
+      "step": 257
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9349814631193314,
+      "learning_rate": 1.972028655212131e-05,
+      "loss": 5.2161,
+      "step": 258
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.196007883835419,
+      "learning_rate": 1.9717235750161808e-05,
+      "loss": 5.1868,
+      "step": 259
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.128921591316096,
+      "learning_rate": 1.9714168639517543e-05,
+      "loss": 5.274,
+      "step": 260
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.42696847333355,
+      "learning_rate": 1.971108522533613e-05,
+      "loss": 5.4778,
+      "step": 261
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0050007534041336,
+      "learning_rate": 1.9707985512792544e-05,
+      "loss": 5.5596,
+      "step": 262
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.669616234270172,
+      "learning_rate": 1.9704869507089105e-05,
+      "loss": 5.3656,
+      "step": 263
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7645184234700584,
+      "learning_rate": 1.970173721345549e-05,
+      "loss": 5.3022,
+      "step": 264
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0530948203414745,
+      "learning_rate": 1.9698588637148705e-05,
+      "loss": 5.374,
+      "step": 265
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8200253090291936,
+      "learning_rate": 1.9695423783453086e-05,
+      "loss": 5.2075,
+      "step": 266
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.315490825813043,
+      "learning_rate": 1.9692242657680286e-05,
+      "loss": 5.2312,
+      "step": 267
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.5436673528211955,
+      "learning_rate": 1.9689045265169272e-05,
+      "loss": 5.3477,
+      "step": 268
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6433231953876364,
+      "learning_rate": 1.9685831611286312e-05,
+      "loss": 5.3378,
+      "step": 269
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.116580849675184,
+      "learning_rate": 1.9682601701424958e-05,
+      "loss": 5.3044,
+      "step": 270
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3227233442460795,
+      "learning_rate": 1.9679355541006056e-05,
+      "loss": 5.3657,
+      "step": 271
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7533500222827096,
+      "learning_rate": 1.9676093135477713e-05,
+      "loss": 5.255,
+      "step": 272
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.564992658943727,
+      "learning_rate": 1.9672814490315312e-05,
+      "loss": 5.3761,
+      "step": 273
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8094013405705556,
+      "learning_rate": 1.9669519611021485e-05,
+      "loss": 5.2762,
+      "step": 274
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.43100443242049,
+      "learning_rate": 1.9666208503126115e-05,
+      "loss": 5.1598,
+      "step": 275
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.2699612429758895,
+      "learning_rate": 1.9662881172186313e-05,
+      "loss": 5.3879,
+      "step": 276
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.922986884818375,
+      "learning_rate": 1.9659537623786428e-05,
+      "loss": 5.4821,
+      "step": 277
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.4516156643641995,
+      "learning_rate": 1.9656177863538025e-05,
+      "loss": 5.3939,
+      "step": 278
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.939693355525761,
+      "learning_rate": 1.965280189707987e-05,
+      "loss": 5.2649,
+      "step": 279
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3627846721747887,
+      "learning_rate": 1.9649409730077934e-05,
+      "loss": 5.3007,
+      "step": 280
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.506360077258328,
+      "learning_rate": 1.9646001368225382e-05,
+      "loss": 5.4097,
+      "step": 281
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.214193129635226,
+      "learning_rate": 1.9642576817242553e-05,
+      "loss": 5.316,
+      "step": 282
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.2863221250018135,
+      "learning_rate": 1.9639136082876954e-05,
+      "loss": 5.3198,
+      "step": 283
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.826690727773217,
+      "learning_rate": 1.9635679170903258e-05,
+      "loss": 5.3665,
+      "step": 284
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.521409795613998,
+      "learning_rate": 1.9632206087123296e-05,
+      "loss": 5.1958,
+      "step": 285
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.195903307493117,
+      "learning_rate": 1.962871683736603e-05,
+      "loss": 5.1238,
+      "step": 286
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.701677642348548,
+      "learning_rate": 1.962521142748755e-05,
+      "loss": 5.1754,
+      "step": 287
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.650371964203301,
+      "learning_rate": 1.9621689863371083e-05,
+      "loss": 5.2856,
+      "step": 288
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.643759454410139,
+      "learning_rate": 1.9618152150926953e-05,
+      "loss": 5.4458,
+      "step": 289
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.0810750026652345,
+      "learning_rate": 1.9614598296092603e-05,
+      "loss": 5.3881,
+      "step": 290
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.540921451453481,
+      "learning_rate": 1.9611028304832547e-05,
+      "loss": 5.3133,
+      "step": 291
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.449089021608378,
+      "learning_rate": 1.9607442183138403e-05,
+      "loss": 5.2165,
+      "step": 292
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.303653173381438,
+      "learning_rate": 1.960383993702884e-05,
+      "loss": 5.2999,
+      "step": 293
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.777311506166449,
+      "learning_rate": 1.9600221572549607e-05,
+      "loss": 5.1426,
+      "step": 294
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.756333904685148,
+      "learning_rate": 1.9596587095773496e-05,
+      "loss": 5.3915,
+      "step": 295
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.378184736178621,
+      "learning_rate": 1.959293651280034e-05,
+      "loss": 5.138,
+      "step": 296
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.883143470991173,
+      "learning_rate": 1.958926982975701e-05,
+      "loss": 5.1153,
+      "step": 297
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.929401125683602,
+      "learning_rate": 1.958558705279739e-05,
+      "loss": 5.2267,
+      "step": 298
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.781275244606163,
+      "learning_rate": 1.9581888188102375e-05,
+      "loss": 5.2871,
+      "step": 299
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.234221317741716,
+      "learning_rate": 1.957817324187987e-05,
+      "loss": 5.1617,
+      "step": 300
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.159465777574264,
+      "learning_rate": 1.9574442220364768e-05,
+      "loss": 5.1299,
+      "step": 301
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.256329018916455,
+      "learning_rate": 1.9570695129818928e-05,
+      "loss": 5.2352,
+      "step": 302
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.133018933678203,
+      "learning_rate": 1.956693197653119e-05,
+      "loss": 5.2031,
+      "step": 303
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.87755014809663,
+      "learning_rate": 1.9563152766817356e-05,
+      "loss": 5.2025,
+      "step": 304
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9463633578914727,
+      "learning_rate": 1.9559357507020163e-05,
+      "loss": 5.0225,
+      "step": 305
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.274298024612477,
+      "learning_rate": 1.9555546203509297e-05,
+      "loss": 5.1675,
+      "step": 306
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.571969693205734,
+      "learning_rate": 1.9551718862681363e-05,
+      "loss": 5.1387,
+      "step": 307
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.441878767657563,
+      "learning_rate": 1.9547875490959884e-05,
+      "loss": 5.3539,
+      "step": 308
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.627572689101822,
+      "learning_rate": 1.9544016094795294e-05,
+      "loss": 5.1321,
+      "step": 309
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.781514761663345,
+      "learning_rate": 1.9540140680664915e-05,
+      "loss": 5.1977,
+      "step": 310
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.140560769424146,
+      "learning_rate": 1.953624925507295e-05,
+      "loss": 5.0422,
+      "step": 311
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.907850959965522,
+      "learning_rate": 1.953234182455048e-05,
+      "loss": 5.3369,
+      "step": 312
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.691324420893443,
+      "learning_rate": 1.9528418395655443e-05,
+      "loss": 5.2007,
+      "step": 313
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.0983085273471955,
+      "learning_rate": 1.952447897497263e-05,
+      "loss": 5.1451,
+      "step": 314
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.354850165008846,
+      "learning_rate": 1.952052356911368e-05,
+      "loss": 5.1415,
+      "step": 315
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.8975948202258897,
+      "learning_rate": 1.9516552184717036e-05,
+      "loss": 5.0287,
+      "step": 316
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.873422769676822,
+      "learning_rate": 1.951256482844799e-05,
+      "loss": 4.9794,
+      "step": 317
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.683604880793774,
+      "learning_rate": 1.9508561506998613e-05,
+      "loss": 5.105,
+      "step": 318
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.194250897520733,
+      "learning_rate": 1.950454222708778e-05,
+      "loss": 5.2785,
+      "step": 319
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.560112427184304,
+      "learning_rate": 1.950050699546116e-05,
+      "loss": 4.9859,
+      "step": 320
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.652738444026526,
+      "learning_rate": 1.949645581889118e-05,
+      "loss": 5.0616,
+      "step": 321
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.8360003224169783,
+      "learning_rate": 1.9492388704177036e-05,
+      "loss": 5.0376,
+      "step": 322
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.744616157206043,
+      "learning_rate": 1.9488305658144666e-05,
+      "loss": 5.0505,
+      "step": 323
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.197730198094315,
+      "learning_rate": 1.9484206687646753e-05,
+      "loss": 5.2067,
+      "step": 324
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.344506328513772,
+      "learning_rate": 1.9480091799562706e-05,
+      "loss": 5.2547,
+      "step": 325
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.033536417957443,
+      "learning_rate": 1.9475961000798645e-05,
+      "loss": 5.1837,
+      "step": 326
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.220408063420163,
+      "learning_rate": 1.947181429828739e-05,
+      "loss": 5.0591,
+      "step": 327
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.20562324255212,
+      "learning_rate": 1.9467651698988464e-05,
+      "loss": 5.2423,
+      "step": 328
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.741171703804873,
+      "learning_rate": 1.9463473209888063e-05,
+      "loss": 5.0776,
+      "step": 329
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.352390084872122,
+      "learning_rate": 1.9459278837999048e-05,
+      "loss": 5.1908,
+      "step": 330
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.336766536879142,
+      "learning_rate": 1.9455068590360943e-05,
+      "loss": 5.0963,
+      "step": 331
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.323080508118235,
+      "learning_rate": 1.9450842474039914e-05,
+      "loss": 5.0653,
+      "step": 332
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.748250010976563,
+      "learning_rate": 1.944660049612876e-05,
+      "loss": 5.1332,
+      "step": 333
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.7527834896436865,
+      "learning_rate": 1.9442342663746903e-05,
+      "loss": 4.9813,
+      "step": 334
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.5313121566876555,
+      "learning_rate": 1.9438068984040366e-05,
+      "loss": 4.8612,
+      "step": 335
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.769650373642717,
+      "learning_rate": 1.943377946418178e-05,
+      "loss": 5.1038,
+      "step": 336
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.254447255879978,
+      "learning_rate": 1.942947411137035e-05,
+      "loss": 4.9938,
+      "step": 337
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.232316770724755,
+      "learning_rate": 1.942515293283187e-05,
+      "loss": 4.9693,
+      "step": 338
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.84144956221707,
+      "learning_rate": 1.9420815935818673e-05,
+      "loss": 5.0987,
+      "step": 339
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.639198534380874,
+      "learning_rate": 1.9416463127609655e-05,
+      "loss": 5.0796,
+      "step": 340
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.293798925443441,
+      "learning_rate": 1.941209451551025e-05,
+      "loss": 4.9292,
+      "step": 341
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.903633350612866,
+      "learning_rate": 1.9407710106852405e-05,
+      "loss": 5.129,
+      "step": 342
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.126542780485656,
+      "learning_rate": 1.940330990899459e-05,
+      "loss": 4.9603,
+      "step": 343
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.288069212982325,
+      "learning_rate": 1.9398893929321763e-05,
+      "loss": 4.9396,
+      "step": 344
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9523613385551686,
+      "learning_rate": 1.9394462175245382e-05,
+      "loss": 5.1768,
+      "step": 345
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.996899954396938,
+      "learning_rate": 1.939001465420337e-05,
+      "loss": 4.9787,
+      "step": 346
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.569032903501148,
+      "learning_rate": 1.9385551373660113e-05,
+      "loss": 4.9764,
+      "step": 347
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.176747620302349,
+      "learning_rate": 1.9381072341106453e-05,
+      "loss": 5.1336,
+      "step": 348
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.154232238794425,
+      "learning_rate": 1.937657756405966e-05,
+      "loss": 5.058,
+      "step": 349
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8984462605590084,
+      "learning_rate": 1.937206705006344e-05,
+      "loss": 4.9333,
+      "step": 350
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.690682787780671,
+      "learning_rate": 1.9367540806687894e-05,
+      "loss": 4.974,
+      "step": 351
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.031124140257661,
+      "learning_rate": 1.9362998841529542e-05,
+      "loss": 5.1886,
+      "step": 352
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.992297500197864,
+      "learning_rate": 1.935844116221127e-05,
+      "loss": 5.0888,
+      "step": 353
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.818973294829996,
+      "learning_rate": 1.9353867776382357e-05,
+      "loss": 5.1499,
+      "step": 354
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.619804549551641,
+      "learning_rate": 1.9349278691718426e-05,
+      "loss": 4.9794,
+      "step": 355
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.592935767722761,
+      "learning_rate": 1.934467391592146e-05,
+      "loss": 4.8134,
+      "step": 356
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.796223984937865,
+      "learning_rate": 1.9340053456719768e-05,
+      "loss": 5.0538,
+      "step": 357
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2741434670658767,
+      "learning_rate": 1.9335417321867988e-05,
+      "loss": 4.8695,
+      "step": 358
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4290531934102613,
+      "learning_rate": 1.9330765519147058e-05,
+      "loss": 4.8899,
+      "step": 359
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8359637299600364,
+      "learning_rate": 1.9326098056364224e-05,
+      "loss": 5.0173,
+      "step": 360
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.249562964456709,
+      "learning_rate": 1.9321414941353006e-05,
+      "loss": 4.9638,
+      "step": 361
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.552155369652054,
+      "learning_rate": 1.931671618197319e-05,
+      "loss": 4.8706,
+      "step": 362
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.94673706390727,
+      "learning_rate": 1.931200178611083e-05,
+      "loss": 4.9603,
+      "step": 363
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.587398864725715,
+      "learning_rate": 1.9307271761678214e-05,
+      "loss": 5.1801,
+      "step": 364
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.786932404767004,
+      "learning_rate": 1.9302526116613863e-05,
+      "loss": 4.8651,
+      "step": 365
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.4252520799983905,
+      "learning_rate": 1.9297764858882516e-05,
+      "loss": 5.1099,
+      "step": 366
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.1820317143483927,
+      "learning_rate": 1.9292987996475113e-05,
+      "loss": 4.8881,
+      "step": 367
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.807420527910623,
+      "learning_rate": 1.928819553740878e-05,
+      "loss": 4.9814,
+      "step": 368
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.2681398731258975,
+      "learning_rate": 1.9283387489726827e-05,
+      "loss": 4.8342,
+      "step": 369
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.2039721487863915,
+      "learning_rate": 1.9278563861498726e-05,
+      "loss": 4.9355,
+      "step": 370
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.125766177541317,
+      "learning_rate": 1.9273724660820086e-05,
+      "loss": 5.0716,
+      "step": 371
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.872774946960404,
+      "learning_rate": 1.9268869895812673e-05,
+      "loss": 4.8714,
+      "step": 372
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.4800006941199735,
+      "learning_rate": 1.9263999574624357e-05,
+      "loss": 4.7984,
+      "step": 373
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.493398067066022,
+      "learning_rate": 1.925911370542912e-05,
+      "loss": 5.1133,
+      "step": 374
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.493654763578316,
+      "learning_rate": 1.9254212296427043e-05,
+      "loss": 4.7568,
+      "step": 375
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.658644058543485,
+      "learning_rate": 1.9249295355844286e-05,
+      "loss": 4.8862,
+      "step": 376
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.74943938464457,
+      "learning_rate": 1.9244362891933077e-05,
+      "loss": 4.9593,
+      "step": 377
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.3633638724612815,
+      "learning_rate": 1.9239414912971697e-05,
+      "loss": 4.8014,
+      "step": 378
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8515410115588504,
+      "learning_rate": 1.923445142726446e-05,
+      "loss": 5.0227,
+      "step": 379
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.555780430950132,
+      "learning_rate": 1.922947244314172e-05,
+      "loss": 4.8137,
+      "step": 380
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.809535670084992,
+      "learning_rate": 1.922447796895982e-05,
+      "loss": 4.9212,
+      "step": 381
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.688606257617834,
+      "learning_rate": 1.9219468013101123e-05,
+      "loss": 5.0151,
+      "step": 382
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.349121714522054,
+      "learning_rate": 1.9214442583973965e-05,
+      "loss": 4.8788,
+      "step": 383
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.863385327006022,
+      "learning_rate": 1.920940169001265e-05,
+      "loss": 4.8467,
+      "step": 384
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.852291860004504,
+      "learning_rate": 1.9204345339677442e-05,
+      "loss": 5.0244,
+      "step": 385
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.064040378679961,
+      "learning_rate": 1.919927354145454e-05,
+      "loss": 4.8919,
+      "step": 386
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.802953333476715,
+      "learning_rate": 1.919418630385607e-05,
+      "loss": 4.9746,
+      "step": 387
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.651834945637443,
+      "learning_rate": 1.9189083635420077e-05,
+      "loss": 4.7957,
+      "step": 388
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.565966174891365,
+      "learning_rate": 1.9183965544710495e-05,
+      "loss": 4.9191,
+      "step": 389
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.027565446582875,
+      "learning_rate": 1.9178832040317153e-05,
+      "loss": 4.9894,
+      "step": 390
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.785579850054768,
+      "learning_rate": 1.9173683130855737e-05,
+      "loss": 4.8868,
+      "step": 391
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.764008160209594,
+      "learning_rate": 1.9168518824967797e-05,
+      "loss": 4.9743,
+      "step": 392
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.849385545583676,
+      "learning_rate": 1.916333913132072e-05,
+      "loss": 4.9855,
+      "step": 393
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.172608067530522,
+      "learning_rate": 1.915814405860771e-05,
+      "loss": 4.8692,
+      "step": 394
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.252142640840125,
+      "learning_rate": 1.91529336155478e-05,
+      "loss": 4.7972,
+      "step": 395
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7879672350261075,
+      "learning_rate": 1.9147707810885798e-05,
+      "loss": 4.8605,
+      "step": 396
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.846876109888715,
+      "learning_rate": 1.9142466653392317e-05,
+      "loss": 4.7777,
+      "step": 397
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.271047397565751,
+      "learning_rate": 1.913721015186372e-05,
+      "loss": 4.7847,
+      "step": 398
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.792822143066731,
+      "learning_rate": 1.913193831512213e-05,
+      "loss": 4.8928,
+      "step": 399
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.818952980145011,
+      "learning_rate": 1.9126651152015404e-05,
+      "loss": 4.7874,
+      "step": 400
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.5004007269500335,
+      "learning_rate": 1.912134867141712e-05,
+      "loss": 4.9484,
+      "step": 401
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8619583600345053,
+      "learning_rate": 1.911603088222657e-05,
+      "loss": 4.8863,
+      "step": 402
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.068544846748943,
+      "learning_rate": 1.9110697793368733e-05,
+      "loss": 4.9838,
+      "step": 403
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.678329405936819,
+      "learning_rate": 1.9105349413794272e-05,
+      "loss": 4.8077,
+      "step": 404
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.1845179577894065,
+      "learning_rate": 1.9099985752479505e-05,
+      "loss": 4.9151,
+      "step": 405
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.25200584743188,
+      "learning_rate": 1.9094606818426403e-05,
+      "loss": 4.9883,
+      "step": 406
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.469433610330476,
+      "learning_rate": 1.908921262066257e-05,
+      "loss": 4.9338,
+      "step": 407
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9569422357282615,
+      "learning_rate": 1.9083803168241225e-05,
+      "loss": 4.7066,
+      "step": 408
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.162760399856056,
+      "learning_rate": 1.9078378470241183e-05,
+      "loss": 4.8367,
+      "step": 409
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5660794253037156,
+      "learning_rate": 1.9072938535766864e-05,
+      "loss": 4.7543,
+      "step": 410
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.797315689316765,
+      "learning_rate": 1.9067483373948245e-05,
+      "loss": 4.9331,
+      "step": 411
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.111914062880509,
+      "learning_rate": 1.906201299394086e-05,
+      "loss": 4.7696,
+      "step": 412
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.539315392333889,
+      "learning_rate": 1.9056527404925788e-05,
+      "loss": 5.0399,
+      "step": 413
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.059585069189535,
+      "learning_rate": 1.9051026616109637e-05,
+      "loss": 4.757,
+      "step": 414
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.3260168539805735,
+      "learning_rate": 1.904551063672452e-05,
+      "loss": 4.702,
+      "step": 415
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.549834876237694,
+      "learning_rate": 1.9039979476028044e-05,
+      "loss": 4.6862,
+      "step": 416
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6787257685993997,
+      "learning_rate": 1.90344331433033e-05,
+      "loss": 4.7645,
+      "step": 417
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.4154234577817055,
+      "learning_rate": 1.9028871647858836e-05,
+      "loss": 4.7376,
+      "step": 418
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.9661339259733364,
+      "learning_rate": 1.9023294999028654e-05,
+      "loss": 4.7945,
+      "step": 419
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.932704209466543,
+      "learning_rate": 1.9017703206172187e-05,
+      "loss": 4.7688,
+      "step": 420
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.861972016862522,
+      "learning_rate": 1.9012096278674283e-05,
+      "loss": 4.8772,
+      "step": 421
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.548438857043438,
+      "learning_rate": 1.900647422594519e-05,
+      "loss": 4.8837,
+      "step": 422
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.6790725280248475,
+      "learning_rate": 1.900083705742054e-05,
+      "loss": 4.8103,
+      "step": 423
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.539654010468228,
+      "learning_rate": 1.8995184782561343e-05,
+      "loss": 4.893,
+      "step": 424
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.68762323673155,
+      "learning_rate": 1.8989517410853956e-05,
+      "loss": 4.8705,
+      "step": 425
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.36298700012937,
+      "learning_rate": 1.8983834951810068e-05,
+      "loss": 4.8289,
+      "step": 426
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.5203690816614435,
+      "learning_rate": 1.89781374149667e-05,
+      "loss": 4.7811,
+      "step": 427
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.99630344415343,
+      "learning_rate": 1.897242480988617e-05,
+      "loss": 4.6708,
+      "step": 428
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.297946514995462,
+      "learning_rate": 1.8966697146156092e-05,
+      "loss": 4.7412,
+      "step": 429
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.807089182475005,
+      "learning_rate": 1.896095443338935e-05,
+      "loss": 4.8245,
+      "step": 430
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.442015701587801,
+      "learning_rate": 1.895519668122408e-05,
+      "loss": 4.741,
+      "step": 431
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.627904445348331,
+      "learning_rate": 1.894942389932367e-05,
+      "loss": 4.9728,
+      "step": 432
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.3819001573345915,
+      "learning_rate": 1.8943636097376728e-05,
+      "loss": 4.4982,
+      "step": 433
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.9790732584475466,
+      "learning_rate": 1.8937833285097067e-05,
+      "loss": 4.9344,
+      "step": 434
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.418435773767789,
+      "learning_rate": 1.8932015472223692e-05,
+      "loss": 4.7223,
+      "step": 435
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.379821395443412,
+      "learning_rate": 1.8926182668520794e-05,
+      "loss": 4.6574,
+      "step": 436
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.002954466901974,
+      "learning_rate": 1.892033488377771e-05,
+      "loss": 4.7822,
+      "step": 437
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.36170107372927,
+      "learning_rate": 1.891447212780893e-05,
+      "loss": 4.6219,
+      "step": 438
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.555620399748319,
+      "learning_rate": 1.8908594410454068e-05,
+      "loss": 4.8075,
+      "step": 439
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.765240769452751,
+      "learning_rate": 1.8902701741577844e-05,
+      "loss": 4.8486,
+      "step": 440
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.479000049549065,
+      "learning_rate": 1.8896794131070073e-05,
+      "loss": 4.8357,
+      "step": 441
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.846180547802962,
+      "learning_rate": 1.8890871588845653e-05,
+      "loss": 4.8586,
+      "step": 442
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.4394887924737105,
+      "learning_rate": 1.8884934124844534e-05,
+      "loss": 4.8759,
+      "step": 443
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.771942523822463,
+      "learning_rate": 1.8878981749031718e-05,
+      "loss": 4.6933,
+      "step": 444
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.222815576090352,
+      "learning_rate": 1.8873014471397225e-05,
+      "loss": 4.6762,
+      "step": 445
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8429763054526256,
+      "learning_rate": 1.886703230195609e-05,
+      "loss": 4.963,
+      "step": 446
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.934157371040181,
+      "learning_rate": 1.8861035250748343e-05,
+      "loss": 4.8002,
+      "step": 447
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.717891500942569,
+      "learning_rate": 1.8855023327838984e-05,
+      "loss": 4.7741,
+      "step": 448
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.000372249774135,
+      "learning_rate": 1.8848996543317982e-05,
+      "loss": 4.6798,
+      "step": 449
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7911306427773557,
+      "learning_rate": 1.8842954907300236e-05,
+      "loss": 4.7683,
+      "step": 450
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.059160382982438,
+      "learning_rate": 1.8836898429925586e-05,
+      "loss": 4.7103,
+      "step": 451
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.154690931845284,
+      "learning_rate": 1.883082712135877e-05,
+      "loss": 4.6826,
+      "step": 452
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.718997675526316,
+      "learning_rate": 1.8824740991789417e-05,
+      "loss": 4.6789,
+      "step": 453
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.9877134941447205,
+      "learning_rate": 1.8818640051432036e-05,
+      "loss": 4.7378,
+      "step": 454
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.986902956873382,
+      "learning_rate": 1.881252431052599e-05,
+      "loss": 4.8052,
+      "step": 455
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.581422466169945,
+      "learning_rate": 1.8806393779335483e-05,
+      "loss": 4.7852,
+      "step": 456
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.657535097217237,
+      "learning_rate": 1.8800248468149545e-05,
+      "loss": 4.8474,
+      "step": 457
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.965672082104629,
+      "learning_rate": 1.8794088387282e-05,
+      "loss": 4.7586,
+      "step": 458
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.277918670720924,
+      "learning_rate": 1.8787913547071485e-05,
+      "loss": 4.629,
+      "step": 459
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.159016365365489,
+      "learning_rate": 1.8781723957881374e-05,
+      "loss": 4.7081,
+      "step": 460
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.816953315704852,
+      "learning_rate": 1.8775519630099822e-05,
+      "loss": 4.5669,
+      "step": 461
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.130734125757001,
+      "learning_rate": 1.876930057413971e-05,
+      "loss": 4.598,
+      "step": 462
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.014308605597524,
+      "learning_rate": 1.8763066800438638e-05,
+      "loss": 4.8269,
+      "step": 463
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.896624107406017,
+      "learning_rate": 1.875681831945891e-05,
+      "loss": 4.5851,
+      "step": 464
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.525332134586662,
+      "learning_rate": 1.87505551416875e-05,
+      "loss": 4.662,
+      "step": 465
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.06267120156479,
+      "learning_rate": 1.874427727763607e-05,
+      "loss": 4.694,
+      "step": 466
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9585474393739455,
+      "learning_rate": 1.873798473784092e-05,
+      "loss": 4.7157,
+      "step": 467
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.831951830757437,
+      "learning_rate": 1.8731677532862975e-05,
+      "loss": 4.8873,
+      "step": 468
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.762319748001275,
+      "learning_rate": 1.872535567328778e-05,
+      "loss": 4.5554,
+      "step": 469
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9742662708177727,
+      "learning_rate": 1.871901916972547e-05,
+      "loss": 4.8751,
+      "step": 470
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8286378361260334,
+      "learning_rate": 1.8712668032810767e-05,
+      "loss": 4.67,
+      "step": 471
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.917732126677209,
+      "learning_rate": 1.870630227320294e-05,
+      "loss": 4.8138,
+      "step": 472
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9483842076440188,
+      "learning_rate": 1.8699921901585814e-05,
+      "loss": 4.6845,
+      "step": 473
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8546046730369565,
+      "learning_rate": 1.8693526928667724e-05,
+      "loss": 4.5114,
+      "step": 474
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.077329695550013,
+      "learning_rate": 1.8687117365181514e-05,
+      "loss": 4.4996,
+      "step": 475
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.131904872521195,
+      "learning_rate": 1.868069322188452e-05,
+      "loss": 4.6157,
+      "step": 476
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.167070219544355,
+      "learning_rate": 1.8674254509558544e-05,
+      "loss": 4.635,
+      "step": 477
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.080283219503045,
+      "learning_rate": 1.8667801239009845e-05,
+      "loss": 4.7456,
+      "step": 478
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.617084882050906,
+      "learning_rate": 1.866133342106911e-05,
+      "loss": 4.6322,
+      "step": 479
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.877045676966755,
+      "learning_rate": 1.865485106659145e-05,
+      "loss": 4.6529,
+      "step": 480
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.8546137617071095,
+      "learning_rate": 1.864835418645635e-05,
+      "loss": 4.5812,
+      "step": 481
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.024800269628755,
+      "learning_rate": 1.86418427915677e-05,
+      "loss": 4.7275,
+      "step": 482
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.104346808360058,
+      "learning_rate": 1.863531689285374e-05,
+      "loss": 4.4275,
+      "step": 483
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.443182810817108,
+      "learning_rate": 1.8628776501267052e-05,
+      "loss": 4.5006,
+      "step": 484
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7630454146072916,
+      "learning_rate": 1.862222162778454e-05,
+      "loss": 4.6852,
+      "step": 485
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.377530807213343,
+      "learning_rate": 1.861565228340742e-05,
+      "loss": 4.4133,
+      "step": 486
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.074254813435036,
+      "learning_rate": 1.8609068479161182e-05,
+      "loss": 4.7537,
+      "step": 487
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.8222231116074195,
+      "learning_rate": 1.8602470226095602e-05,
+      "loss": 4.5551,
+      "step": 488
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.37608449638734,
+      "learning_rate": 1.8595857535284692e-05,
+      "loss": 4.53,
+      "step": 489
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.128267316594894,
+      "learning_rate": 1.85892304178267e-05,
+      "loss": 4.6295,
+      "step": 490
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.860631729644693,
+      "learning_rate": 1.8582588884844086e-05,
+      "loss": 4.4502,
+      "step": 491
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.9510406392102295,
+      "learning_rate": 1.8575932947483503e-05,
+      "loss": 4.7843,
+      "step": 492
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.948182711081336,
+      "learning_rate": 1.8569262616915784e-05,
+      "loss": 4.7088,
+      "step": 493
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.802694876664713,
+      "learning_rate": 1.8562577904335913e-05,
+      "loss": 4.4809,
+      "step": 494
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.871236430723091,
+      "learning_rate": 1.8555878820963014e-05,
+      "loss": 4.5609,
+      "step": 495
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.297328298890337,
+      "learning_rate": 1.8549165378040328e-05,
+      "loss": 4.5167,
+      "step": 496
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.604119686459511,
+      "learning_rate": 1.8542437586835202e-05,
+      "loss": 4.7448,
+      "step": 497
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.557903280049274,
+      "learning_rate": 1.8535695458639056e-05,
+      "loss": 4.6444,
+      "step": 498
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.8352904144837736,
+      "learning_rate": 1.8528939004767377e-05,
+      "loss": 4.5181,
+      "step": 499
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.4315903519140605,
+      "learning_rate": 1.8522168236559693e-05,
+      "loss": 4.5232,
+      "step": 500
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.5003343526651225,
+      "learning_rate": 1.851538316537956e-05,
+      "loss": 4.5658,
+      "step": 501
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.655561193479631,
+      "learning_rate": 1.8508583802614534e-05,
+      "loss": 4.6049,
+      "step": 502
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.336819493633404,
+      "learning_rate": 1.8501770159676157e-05,
+      "loss": 4.385,
+      "step": 503
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.213922156906962,
+      "learning_rate": 1.849494224799994e-05,
+      "loss": 4.5663,
+      "step": 504
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.8801853326394733,
+      "learning_rate": 1.8488100079045345e-05,
+      "loss": 4.4578,
+      "step": 505
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.638341688055788,
+      "learning_rate": 1.848124366429576e-05,
+      "loss": 4.4374,
+      "step": 506
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.1259886569912405,
+      "learning_rate": 1.8474373015258472e-05,
+      "loss": 4.5498,
+      "step": 507
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.335260758484502,
+      "learning_rate": 1.846748814346468e-05,
+      "loss": 4.4976,
+      "step": 508
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5227862557003005,
+      "learning_rate": 1.846058906046943e-05,
+      "loss": 4.4723,
+      "step": 509
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.391719508660204,
+      "learning_rate": 1.8453675777851627e-05,
+      "loss": 4.6382,
+      "step": 510
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.119916316504547,
+      "learning_rate": 1.844674830721402e-05,
+      "loss": 4.5931,
+      "step": 511
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5972094158028476,
+      "learning_rate": 1.843980666018315e-05,
+      "loss": 4.4852,
+      "step": 512
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.437761861563065,
+      "learning_rate": 1.8432850848409367e-05,
+      "loss": 4.5205,
+      "step": 513
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4619487319523112,
+      "learning_rate": 1.8425880883566784e-05,
+      "loss": 4.5162,
+      "step": 514
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.246560330909532,
+      "learning_rate": 1.8418896777353272e-05,
+      "loss": 4.6419,
+      "step": 515
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.504108212476313,
+      "learning_rate": 1.8411898541490433e-05,
+      "loss": 4.5368,
+      "step": 516
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.7748515692704,
+      "learning_rate": 1.840488618772359e-05,
+      "loss": 4.325,
+      "step": 517
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.400533553399993,
+      "learning_rate": 1.8397859727821747e-05,
+      "loss": 4.7751,
+      "step": 518
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.796804590072757,
+      "learning_rate": 1.83908191735776e-05,
+      "loss": 4.6476,
+      "step": 519
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.318892580222769,
+      "learning_rate": 1.8383764536807486e-05,
+      "loss": 4.7816,
+      "step": 520
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.755398814310214,
+      "learning_rate": 1.8376695829351378e-05,
+      "loss": 4.4902,
+      "step": 521
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.66174991561927,
+      "learning_rate": 1.8369613063072875e-05,
+      "loss": 4.5982,
+      "step": 522
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.16713458203689,
+      "learning_rate": 1.8362516249859164e-05,
+      "loss": 4.4873,
+      "step": 523
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.714662973498544,
+      "learning_rate": 1.8355405401621e-05,
+      "loss": 4.5149,
+      "step": 524
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.496778637749604,
+      "learning_rate": 1.8348280530292712e-05,
+      "loss": 4.5553,
+      "step": 525
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.908462628514215,
+      "learning_rate": 1.834114164783215e-05,
+      "loss": 4.5851,
+      "step": 526
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.0522977320295945,
+      "learning_rate": 1.8333988766220676e-05,
+      "loss": 4.526,
+      "step": 527
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.000357193034575,
+      "learning_rate": 1.832682189746316e-05,
+      "loss": 4.4757,
+      "step": 528
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.2730152597229,
+      "learning_rate": 1.831964105358794e-05,
+      "loss": 4.5307,
+      "step": 529
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.0852057209225,
+      "learning_rate": 1.831244624664681e-05,
+      "loss": 4.5722,
+      "step": 530
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.9744696789409755,
+      "learning_rate": 1.8305237488714995e-05,
+      "loss": 4.4842,
+      "step": 531
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.004849793326939,
+      "learning_rate": 1.8298014791891138e-05,
+      "loss": 4.3618,
+      "step": 532
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.7948194156817605,
+      "learning_rate": 1.829077816829728e-05,
+      "loss": 4.7279,
+      "step": 533
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.8258386178589605,
+      "learning_rate": 1.8283527630078827e-05,
+      "loss": 4.4468,
+      "step": 534
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.047435784425765,
+      "learning_rate": 1.827626318940454e-05,
+      "loss": 4.242,
+      "step": 535
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.245582218556195,
+      "learning_rate": 1.8268984858466524e-05,
+      "loss": 4.4191,
+      "step": 536
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.212873965419334,
+      "learning_rate": 1.8261692649480174e-05,
+      "loss": 4.3846,
+      "step": 537
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.973443022794257,
+      "learning_rate": 1.8254386574684205e-05,
+      "loss": 4.4775,
+      "step": 538
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.655778663416384,
+      "learning_rate": 1.824706664634058e-05,
+      "loss": 4.6469,
+      "step": 539
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7745155639127477,
+      "learning_rate": 1.8239732876734525e-05,
+      "loss": 4.4211,
+      "step": 540
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8928105110404876,
+      "learning_rate": 1.823238527817449e-05,
+      "loss": 4.5191,
+      "step": 541
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.012209741102411,
+      "learning_rate": 1.822502386299214e-05,
+      "loss": 4.7298,
+      "step": 542
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.09491355504511,
+      "learning_rate": 1.8217648643542326e-05,
+      "loss": 4.4712,
+      "step": 543
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.9623455755711756,
+      "learning_rate": 1.8210259632203063e-05,
+      "loss": 4.5201,
+      "step": 544
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.278338142567551,
+      "learning_rate": 1.8202856841375517e-05,
+      "loss": 4.5629,
+      "step": 545
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.470833412471676,
+      "learning_rate": 1.819544028348399e-05,
+      "loss": 4.5509,
+      "step": 546
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.557667082733272,
+      "learning_rate": 1.818800997097587e-05,
+      "loss": 4.4987,
+      "step": 547
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.307435410054455,
+      "learning_rate": 1.8180565916321646e-05,
+      "loss": 4.5655,
+      "step": 548
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.8047001558686695,
+      "learning_rate": 1.817310813201486e-05,
+      "loss": 4.647,
+      "step": 549
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.232758498965093,
+      "learning_rate": 1.816563663057211e-05,
+      "loss": 4.5642,
+      "step": 550
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.9694403182709084,
+      "learning_rate": 1.8158151424533002e-05,
+      "loss": 4.3808,
+      "step": 551
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.511990232061259,
+      "learning_rate": 1.8150652526460146e-05,
+      "loss": 4.6089,
+      "step": 552
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.589363784630558,
+      "learning_rate": 1.8143139948939138e-05,
+      "loss": 4.424,
+      "step": 553
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.391501846213585,
+      "learning_rate": 1.8135613704578525e-05,
+      "loss": 4.6266,
+      "step": 554
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.081782646799652,
+      "learning_rate": 1.81280738060098e-05,
+      "loss": 4.5131,
+      "step": 555
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.701357773481821,
+      "learning_rate": 1.8120520265887364e-05,
+      "loss": 4.7093,
+      "step": 556
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.560945380217301,
+      "learning_rate": 1.8112953096888517e-05,
+      "loss": 4.5045,
+      "step": 557
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.7367764771999035,
+      "learning_rate": 1.810537231171343e-05,
+      "loss": 4.4186,
+      "step": 558
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.383419123230545,
+      "learning_rate": 1.809777792308513e-05,
+      "loss": 4.4929,
+      "step": 559
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.130267653307888,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 4.4062,
+      "step": 560
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.531928301060983,
+      "learning_rate": 1.808254838647513e-05,
+      "loss": 4.3487,
+      "step": 561
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.1958588839876905,
+      "learning_rate": 1.8074913264053547e-05,
+      "loss": 4.5398,
+      "step": 562
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.145503981737899,
+      "learning_rate": 1.8067264589298945e-05,
+      "loss": 4.6086,
+      "step": 563
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.874804141420842,
+      "learning_rate": 1.8059602375048294e-05,
+      "loss": 4.3948,
+      "step": 564
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.710773936481747,
+      "learning_rate": 1.8051926634161282e-05,
+      "loss": 4.4046,
+      "step": 565
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.7682629786468107,
+      "learning_rate": 1.8044237379520305e-05,
+      "loss": 4.5396,
+      "step": 566
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.9988341208652693,
+      "learning_rate": 1.8036534624030428e-05,
+      "loss": 4.5059,
+      "step": 567
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.393107693200345,
+      "learning_rate": 1.802881838061939e-05,
+      "loss": 4.3796,
+      "step": 568
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.574902532810956,
+      "learning_rate": 1.802108866223755e-05,
+      "loss": 4.3632,
+      "step": 569
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.8128396404739546,
+      "learning_rate": 1.8013345481857903e-05,
+      "loss": 4.3907,
+      "step": 570
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.33414503434105,
+      "learning_rate": 1.8005588852476018e-05,
+      "loss": 4.4354,
+      "step": 571
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.301332379657507,
+      "learning_rate": 1.7997818787110043e-05,
+      "loss": 4.4156,
+      "step": 572
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.718201275219629,
+      "learning_rate": 1.7990035298800682e-05,
+      "loss": 4.6335,
+      "step": 573
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.82328878051179,
+      "learning_rate": 1.798223840061116e-05,
+      "loss": 4.3818,
+      "step": 574
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.009302341530525,
+      "learning_rate": 1.797442810562721e-05,
+      "loss": 4.4935,
+      "step": 575
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.470726269769983,
+      "learning_rate": 1.796660442695705e-05,
+      "loss": 4.4081,
+      "step": 576
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.144773016409501,
+      "learning_rate": 1.795876737773136e-05,
+      "loss": 4.5555,
+      "step": 577
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.540527520249997,
+      "learning_rate": 1.795091697110326e-05,
+      "loss": 4.4273,
+      "step": 578
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.539566055605331,
+      "learning_rate": 1.7943053220248284e-05,
+      "loss": 4.583,
+      "step": 579
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.932252647457211,
+      "learning_rate": 1.793517613836437e-05,
+      "loss": 4.4988,
+      "step": 580
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.217312041012894,
+      "learning_rate": 1.7927285738671825e-05,
+      "loss": 4.4821,
+      "step": 581
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.185334769654126,
+      "learning_rate": 1.7919382034413306e-05,
+      "loss": 4.558,
+      "step": 582
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.583654330854395,
+      "learning_rate": 1.7911465038853805e-05,
+      "loss": 4.4156,
+      "step": 583
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.437316693787005,
+      "learning_rate": 1.7903534765280616e-05,
+      "loss": 4.3659,
+      "step": 584
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.442733348394627,
+      "learning_rate": 1.7895591227003316e-05,
+      "loss": 4.2313,
+      "step": 585
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.898831347856921,
+      "learning_rate": 1.7887634437353754e-05,
+      "loss": 4.3839,
+      "step": 586
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.187477019121502,
+      "learning_rate": 1.7879664409686007e-05,
+      "loss": 4.5723,
+      "step": 587
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.317936443725569,
+      "learning_rate": 1.7871681157376382e-05,
+      "loss": 4.2695,
+      "step": 588
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.427780735865234,
+      "learning_rate": 1.7863684693823375e-05,
+      "loss": 4.5148,
+      "step": 589
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.2148392567243835,
+      "learning_rate": 1.7855675032447648e-05,
+      "loss": 4.3736,
+      "step": 590
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.085066691271566,
+      "learning_rate": 1.7847652186692025e-05,
+      "loss": 4.4033,
+      "step": 591
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.106345509099876,
+      "learning_rate": 1.7839616170021452e-05,
+      "loss": 4.3073,
+      "step": 592
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.795043396096179,
+      "learning_rate": 1.7831566995922983e-05,
+      "loss": 4.4571,
+      "step": 593
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.040391570866654,
+      "learning_rate": 1.782350467790575e-05,
+      "loss": 4.5967,
+      "step": 594
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.172887527977414,
+      "learning_rate": 1.7815429229500946e-05,
+      "loss": 4.4391,
+      "step": 595
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.8558696182852272,
+      "learning_rate": 1.78073406642618e-05,
+      "loss": 4.5532,
+      "step": 596
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.324716483505234,
+      "learning_rate": 1.779923899576357e-05,
+      "loss": 4.3905,
+      "step": 597
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.876541883027683,
+      "learning_rate": 1.7791124237603477e-05,
+      "loss": 4.32,
+      "step": 598
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.7679255531997913,
+      "learning_rate": 1.7782996403400737e-05,
+      "loss": 4.3041,
+      "step": 599
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.222611374687213,
+      "learning_rate": 1.7774855506796497e-05,
+      "loss": 4.4244,
+      "step": 600
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.304244924899713,
+      "learning_rate": 1.776670156145383e-05,
+      "loss": 4.5282,
+      "step": 601
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.981595214348723,
+      "learning_rate": 1.775853458105772e-05,
+      "loss": 4.4572,
+      "step": 602
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.183377523359722,
+      "learning_rate": 1.7750354579315004e-05,
+      "loss": 4.6287,
+      "step": 603
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5751970777478586,
+      "learning_rate": 1.77421615699544e-05,
+      "loss": 4.2279,
+      "step": 604
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.8162581902884876,
+      "learning_rate": 1.7733955566726438e-05,
+      "loss": 4.2113,
+      "step": 605
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.545507736912509,
+      "learning_rate": 1.772573658340347e-05,
+      "loss": 4.3691,
+      "step": 606
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.224629578143067,
+      "learning_rate": 1.7717504633779618e-05,
+      "loss": 4.3486,
+      "step": 607
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.610588665876807,
+      "learning_rate": 1.7709259731670774e-05,
+      "loss": 4.38,
+      "step": 608
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.70075196848019,
+      "learning_rate": 1.770100189091457e-05,
+      "loss": 4.1483,
+      "step": 609
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2360932038131613,
+      "learning_rate": 1.7692731125370355e-05,
+      "loss": 4.3603,
+      "step": 610
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.8433889218472244,
+      "learning_rate": 1.7684447448919156e-05,
+      "loss": 4.4584,
+      "step": 611
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.286041363522565,
+      "learning_rate": 1.7676150875463688e-05,
+      "loss": 4.3522,
+      "step": 612
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.324312507834647,
+      "learning_rate": 1.7667841418928292e-05,
+      "loss": 4.34,
+      "step": 613
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.826648128672422,
+      "learning_rate": 1.765951909325895e-05,
+      "loss": 4.3366,
+      "step": 614
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.326699626627627,
+      "learning_rate": 1.7651183912423228e-05,
+      "loss": 4.22,
+      "step": 615
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.502208687825821,
+      "learning_rate": 1.764283589041028e-05,
+      "loss": 4.434,
+      "step": 616
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.9726600059699697,
+      "learning_rate": 1.7634475041230796e-05,
+      "loss": 4.4077,
+      "step": 617
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.794767934344185,
+      "learning_rate": 1.7626101378917004e-05,
+      "loss": 4.3801,
+      "step": 618
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.828415872023352,
+      "learning_rate": 1.761771491752264e-05,
+      "loss": 4.5012,
+      "step": 619
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.864618059858188,
+      "learning_rate": 1.7609315671122912e-05,
+      "loss": 4.1654,
+      "step": 620
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5444444341424144,
+      "learning_rate": 1.760090365381449e-05,
+      "loss": 4.2813,
+      "step": 621
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0754856971618674,
+      "learning_rate": 1.759247887971548e-05,
+      "loss": 4.4173,
+      "step": 622
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.051093782169358,
+      "learning_rate": 1.7584041362965397e-05,
+      "loss": 4.2801,
+      "step": 623
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.703146776128272,
+      "learning_rate": 1.7575591117725132e-05,
+      "loss": 4.2089,
+      "step": 624
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.194884286403555,
+      "learning_rate": 1.7567128158176955e-05,
+      "loss": 4.4067,
+      "step": 625
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.7197087523579073,
+      "learning_rate": 1.7558652498524464e-05,
+      "loss": 4.5624,
+      "step": 626
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.9997767290885187,
+      "learning_rate": 1.7550164152992573e-05,
+      "loss": 3.9206,
+      "step": 627
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.723995563236504,
+      "learning_rate": 1.7541663135827493e-05,
+      "loss": 4.2779,
+      "step": 628
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8566179329855492,
+      "learning_rate": 1.75331494612967e-05,
+      "loss": 4.4607,
+      "step": 629
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6837802961843953,
+      "learning_rate": 1.7524623143688905e-05,
+      "loss": 4.3756,
+      "step": 630
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3643578361540993,
+      "learning_rate": 1.7516084197314044e-05,
+      "loss": 4.3226,
+      "step": 631
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.13921138996773,
+      "learning_rate": 1.7507532636503256e-05,
+      "loss": 3.9717,
+      "step": 632
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.7857046012154534,
+      "learning_rate": 1.749896847560884e-05,
+      "loss": 4.2997,
+      "step": 633
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4312849218009567,
+      "learning_rate": 1.7490391729004242e-05,
+      "loss": 4.3715,
+      "step": 634
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.031659106872237,
+      "learning_rate": 1.748180241108404e-05,
+      "loss": 4.4076,
+      "step": 635
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.29646662187367,
+      "learning_rate": 1.7473200536263905e-05,
+      "loss": 4.3464,
+      "step": 636
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.453734263665285,
+      "learning_rate": 1.746458611898058e-05,
+      "loss": 4.497,
+      "step": 637
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.267571577595243,
+      "learning_rate": 1.7455959173691863e-05,
+      "loss": 4.326,
+      "step": 638
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.037988035302162,
+      "learning_rate": 1.744731971487658e-05,
+      "loss": 4.2036,
+      "step": 639
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.6499546934773,
+      "learning_rate": 1.7438667757034547e-05,
+      "loss": 4.1384,
+      "step": 640
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.193992253945382,
+      "learning_rate": 1.743000331468657e-05,
+      "loss": 4.1671,
+      "step": 641
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.9238693224223846,
+      "learning_rate": 1.7421326402374406e-05,
+      "loss": 4.2942,
+      "step": 642
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.7221881431043866,
+      "learning_rate": 1.7412637034660735e-05,
+      "loss": 4.2449,
+      "step": 643
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.33969588119697,
+      "learning_rate": 1.740393522612915e-05,
+      "loss": 4.397,
+      "step": 644
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.0812195963488773,
+      "learning_rate": 1.739522099138411e-05,
+      "loss": 4.3136,
+      "step": 645
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.248841778221737,
+      "learning_rate": 1.7386494345050944e-05,
+      "loss": 4.2977,
+      "step": 646
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.4512523414792726,
+      "learning_rate": 1.73777553017758e-05,
+      "loss": 4.181,
+      "step": 647
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.056102712659825,
+      "learning_rate": 1.7369003876225644e-05,
+      "loss": 4.262,
+      "step": 648
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.7592475887210575,
+      "learning_rate": 1.7360240083088213e-05,
+      "loss": 4.2013,
+      "step": 649
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.661785505723803,
+      "learning_rate": 1.7351463937072008e-05,
+      "loss": 4.4768,
+      "step": 650
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6526178793870967,
+      "learning_rate": 1.734267545290625e-05,
+      "loss": 4.1434,
+      "step": 651
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.362778318040067,
+      "learning_rate": 1.7333874645340886e-05,
+      "loss": 4.5816,
+      "step": 652
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6785134226263883,
+      "learning_rate": 1.7325061529146528e-05,
+      "loss": 4.1977,
+      "step": 653
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6203914812532583,
+      "learning_rate": 1.7316236119114466e-05,
+      "loss": 4.1502,
+      "step": 654
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.909042527107989,
+      "learning_rate": 1.7307398430056595e-05,
+      "loss": 4.3402,
+      "step": 655
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.343420425268675,
+      "learning_rate": 1.7298548476805446e-05,
+      "loss": 4.1416,
+      "step": 656
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8231798578591825,
+      "learning_rate": 1.7289686274214116e-05,
+      "loss": 4.4234,
+      "step": 657
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.104953811214553,
+      "learning_rate": 1.7280811837156268e-05,
+      "loss": 4.3791,
+      "step": 658
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.006233314995586,
+      "learning_rate": 1.7271925180526094e-05,
+      "loss": 4.1491,
+      "step": 659
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.493796073974652,
+      "learning_rate": 1.72630263192383e-05,
+      "loss": 4.0866,
+      "step": 660
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.5861790784869365,
+      "learning_rate": 1.7254115268228073e-05,
+      "loss": 4.1643,
+      "step": 661
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.094790170581724,
+      "learning_rate": 1.724519204245105e-05,
+      "loss": 4.3313,
+      "step": 662
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.8169362681688614,
+      "learning_rate": 1.723625665688331e-05,
+      "loss": 4.2777,
+      "step": 663
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5415226420284696,
+      "learning_rate": 1.7227309126521347e-05,
+      "loss": 4.2133,
+      "step": 664
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.447252681075696,
+      "learning_rate": 1.7218349466382024e-05,
+      "loss": 4.4371,
+      "step": 665
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.857931839153315,
+      "learning_rate": 1.7209377691502565e-05,
+      "loss": 4.3564,
+      "step": 666
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.868644542943814,
+      "learning_rate": 1.720039381694053e-05,
+      "loss": 4.3347,
+      "step": 667
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.9872569129304427,
+      "learning_rate": 1.7191397857773787e-05,
+      "loss": 4.1822,
+      "step": 668
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.377738459178347,
+      "learning_rate": 1.7182389829100484e-05,
+      "loss": 4.2361,
+      "step": 669
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.526630194902284,
+      "learning_rate": 1.7173369746039026e-05,
+      "loss": 4.2334,
+      "step": 670
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.903151798762359,
+      "learning_rate": 1.7164337623728044e-05,
+      "loss": 4.2507,
+      "step": 671
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.416704913811221,
+      "learning_rate": 1.7155293477326385e-05,
+      "loss": 4.3865,
+      "step": 672
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.668815207748128,
+      "learning_rate": 1.714623732201307e-05,
+      "loss": 4.1203,
+      "step": 673
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.835149742451225,
+      "learning_rate": 1.713716917298727e-05,
+      "loss": 4.1392,
+      "step": 674
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.881418868309846,
+      "learning_rate": 1.7128089045468294e-05,
+      "loss": 4.2875,
+      "step": 675
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.034412844093579,
+      "learning_rate": 1.7118996954695553e-05,
+      "loss": 4.0848,
+      "step": 676
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.518013043332243,
+      "learning_rate": 1.7109892915928535e-05,
+      "loss": 4.3151,
+      "step": 677
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.478304353081878,
+      "learning_rate": 1.7100776944446783e-05,
+      "loss": 4.3224,
+      "step": 678
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.048591652502835,
+      "learning_rate": 1.709164905554986e-05,
+      "loss": 4.0666,
+      "step": 679
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.0357109087365535,
+      "learning_rate": 1.7082509264557333e-05,
+      "loss": 4.0443,
+      "step": 680
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.3409258985175585,
+      "learning_rate": 1.7073357586808753e-05,
+      "loss": 4.2371,
+      "step": 681
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.582253486092752,
+      "learning_rate": 1.706419403766361e-05,
+      "loss": 4.077,
+      "step": 682
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.801114373409387,
+      "learning_rate": 1.7055018632501326e-05,
+      "loss": 4.2553,
+      "step": 683
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.8747080745349978,
+      "learning_rate": 1.7045831386721213e-05,
+      "loss": 4.1561,
+      "step": 684
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.696948918314867,
+      "learning_rate": 1.7036632315742464e-05,
+      "loss": 4.2312,
+      "step": 685
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.258212104642618,
+      "learning_rate": 1.7027421435004114e-05,
+      "loss": 4.1866,
+      "step": 686
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.388331452291105,
+      "learning_rate": 1.7018198759965018e-05,
+      "loss": 4.1992,
+      "step": 687
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.647950158243207,
+      "learning_rate": 1.7008964306103823e-05,
+      "loss": 4.3468,
+      "step": 688
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.168886137147408,
+      "learning_rate": 1.6999718088918956e-05,
+      "loss": 4.0308,
+      "step": 689
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.357151028206129,
+      "learning_rate": 1.6990460123928577e-05,
+      "loss": 4.3588,
+      "step": 690
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.530147627269544,
+      "learning_rate": 1.698119042667056e-05,
+      "loss": 4.3144,
+      "step": 691
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.817151116073827,
+      "learning_rate": 1.6971909012702483e-05,
+      "loss": 4.2138,
+      "step": 692
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.092987342888455,
+      "learning_rate": 1.6962615897601573e-05,
+      "loss": 4.2911,
+      "step": 693
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.613134484273773,
+      "learning_rate": 1.6953311096964706e-05,
+      "loss": 4.3282,
+      "step": 694
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.173274414167209,
+      "learning_rate": 1.6943994626408365e-05,
+      "loss": 4.0758,
+      "step": 695
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.190457374578632,
+      "learning_rate": 1.6934666501568618e-05,
+      "loss": 4.3804,
+      "step": 696
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.601597668292631,
+      "learning_rate": 1.69253267381011e-05,
+      "loss": 4.3887,
+      "step": 697
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.210265170570191,
+      "learning_rate": 1.6915975351680968e-05,
+      "loss": 4.1695,
+      "step": 698
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.137670168126125,
+      "learning_rate": 1.69066123580029e-05,
+      "loss": 4.3003,
+      "step": 699
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.747988286996686,
+      "learning_rate": 1.6897237772781046e-05,
+      "loss": 4.4487,
+      "step": 700
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.594341934110022,
+      "learning_rate": 1.6887851611749005e-05,
+      "loss": 4.1977,
+      "step": 701
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.693710767713602,
+      "learning_rate": 1.6878453890659815e-05,
+      "loss": 4.4077,
+      "step": 702
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9460120139260026,
+      "learning_rate": 1.686904462528591e-05,
+      "loss": 4.3575,
+      "step": 703
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.903534992356641,
+      "learning_rate": 1.68596238314191e-05,
+      "loss": 4.2751,
+      "step": 704
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.6197148558188004,
+      "learning_rate": 1.6850191524870548e-05,
+      "loss": 4.3067,
+      "step": 705
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.013695095794239,
+      "learning_rate": 1.6840747721470733e-05,
+      "loss": 4.3629,
+      "step": 706
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.793940986645831,
+      "learning_rate": 1.6831292437069425e-05,
+      "loss": 4.1589,
+      "step": 707
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.068061320502639,
+      "learning_rate": 1.6821825687535675e-05,
+      "loss": 4.3191,
+      "step": 708
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.835071559565869,
+      "learning_rate": 1.6812347488757774e-05,
+      "loss": 4.3855,
+      "step": 709
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.352222600656303,
+      "learning_rate": 1.6802857856643214e-05,
+      "loss": 4.1151,
+      "step": 710
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.223271549995184,
+      "learning_rate": 1.6793356807118695e-05,
+      "loss": 4.2358,
+      "step": 711
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.697294907663005,
+      "learning_rate": 1.6783844356130073e-05,
+      "loss": 4.0528,
+      "step": 712
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.019657412383279,
+      "learning_rate": 1.677432051964233e-05,
+      "loss": 4.3906,
+      "step": 713
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.330982781472389,
+      "learning_rate": 1.6764785313639568e-05,
+      "loss": 4.2145,
+      "step": 714
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.159400687537892,
+      "learning_rate": 1.6755238754124965e-05,
+      "loss": 4.15,
+      "step": 715
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.743019702246863,
+      "learning_rate": 1.6745680857120757e-05,
+      "loss": 4.3182,
+      "step": 716
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.376804285660271,
+      "learning_rate": 1.6736111638668203e-05,
+      "loss": 4.1689,
+      "step": 717
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.358221275967151,
+      "learning_rate": 1.6726531114827572e-05,
+      "loss": 4.1588,
+      "step": 718
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.0442497727339966,
+      "learning_rate": 1.6716939301678098e-05,
+      "loss": 4.4582,
+      "step": 719
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.046366033000383,
+      "learning_rate": 1.6707336215317968e-05,
+      "loss": 4.2036,
+      "step": 720
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.445621373056556,
+      "learning_rate": 1.6697721871864286e-05,
+      "loss": 4.348,
+      "step": 721
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.8971967069923776,
+      "learning_rate": 1.6688096287453048e-05,
+      "loss": 4.1313,
+      "step": 722
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.9511536886764964,
+      "learning_rate": 1.6678459478239116e-05,
+      "loss": 4.1277,
+      "step": 723
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.7444284748312713,
+      "learning_rate": 1.6668811460396202e-05,
+      "loss": 4.3639,
+      "step": 724
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.535042203472182,
+      "learning_rate": 1.665915225011681e-05,
+      "loss": 4.2301,
+      "step": 725
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.344050025912771,
+      "learning_rate": 1.664948186361225e-05,
+      "loss": 4.084,
+      "step": 726
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.583266578664799,
+      "learning_rate": 1.663980031711257e-05,
+      "loss": 4.1723,
+      "step": 727
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.094665681041583,
+      "learning_rate": 1.6630107626866558e-05,
+      "loss": 4.3757,
+      "step": 728
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.904919081388526,
+      "learning_rate": 1.6620403809141707e-05,
+      "loss": 4.1186,
+      "step": 729
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.399804648357315,
+      "learning_rate": 1.6610688880224178e-05,
+      "loss": 4.0947,
+      "step": 730
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.923219829805552,
+      "learning_rate": 1.6600962856418782e-05,
+      "loss": 4.335,
+      "step": 731
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.427663133020312,
+      "learning_rate": 1.6591225754048963e-05,
+      "loss": 4.1266,
+      "step": 732
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.097746227281494,
+      "learning_rate": 1.6581477589456737e-05,
+      "loss": 4.1285,
+      "step": 733
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.745457635840251,
+      "learning_rate": 1.6571718379002705e-05,
+      "loss": 4.3592,
+      "step": 734
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.758424791186076,
+      "learning_rate": 1.6561948139065997e-05,
+      "loss": 3.9558,
+      "step": 735
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.6602131909735185,
+      "learning_rate": 1.6552166886044253e-05,
+      "loss": 4.2786,
+      "step": 736
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.331991964126386,
+      "learning_rate": 1.6542374636353605e-05,
+      "loss": 4.205,
+      "step": 737
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.742271026876533,
+      "learning_rate": 1.653257140642863e-05,
+      "loss": 4.1134,
+      "step": 738
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.128592064049964,
+      "learning_rate": 1.6522757212722346e-05,
+      "loss": 4.2645,
+      "step": 739
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.128026840094599,
+      "learning_rate": 1.6512932071706153e-05,
+      "loss": 4.1295,
+      "step": 740
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.73216243413116,
+      "learning_rate": 1.650309599986985e-05,
+      "loss": 4.3763,
+      "step": 741
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.406532996935937,
+      "learning_rate": 1.6493249013721558e-05,
+      "loss": 4.2893,
+      "step": 742
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.892317898810144,
+      "learning_rate": 1.6483391129787725e-05,
+      "loss": 4.0252,
+      "step": 743
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.655969865077262,
+      "learning_rate": 1.64735223646131e-05,
+      "loss": 4.0342,
+      "step": 744
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.237746678866634,
+      "learning_rate": 1.646364273476067e-05,
+      "loss": 4.2777,
+      "step": 745
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.7276325571184055,
+      "learning_rate": 1.6453752256811676e-05,
+      "loss": 4.1562,
+      "step": 746
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.040221659089892,
+      "learning_rate": 1.644385094736556e-05,
+      "loss": 4.0892,
+      "step": 747
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.219234733858174,
+      "learning_rate": 1.6433938823039942e-05,
+      "loss": 4.2405,
+      "step": 748
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.014880781194069,
+      "learning_rate": 1.642401590047059e-05,
+      "loss": 4.1244,
+      "step": 749
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.603922662946277,
+      "learning_rate": 1.6414082196311402e-05,
+      "loss": 4.1623,
+      "step": 750
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.169752318365569,
+      "learning_rate": 1.6404137727234366e-05,
+      "loss": 4.1791,
+      "step": 751
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.800322679915261,
+      "learning_rate": 1.639418250992954e-05,
+      "loss": 4.0497,
+      "step": 752
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.819362617310608,
+      "learning_rate": 1.6384216561105014e-05,
+      "loss": 4.2569,
+      "step": 753
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8943247801354315,
+      "learning_rate": 1.63742398974869e-05,
+      "loss": 4.2013,
+      "step": 754
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.237703478343301,
+      "learning_rate": 1.6364252535819284e-05,
+      "loss": 4.2323,
+      "step": 755
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.289135372736193,
+      "learning_rate": 1.635425449286421e-05,
+      "loss": 4.2226,
+      "step": 756
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.03279421005735,
+      "learning_rate": 1.6344245785401653e-05,
+      "loss": 4.0047,
+      "step": 757
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.12143068811916,
+      "learning_rate": 1.6334226430229475e-05,
+      "loss": 4.1518,
+      "step": 758
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.045018027459975,
+      "learning_rate": 1.632419644416342e-05,
+      "loss": 3.9894,
+      "step": 759
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.3634370875446935,
+      "learning_rate": 1.6314155844037074e-05,
+      "loss": 4.1409,
+      "step": 760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.424185767990525,
+      "learning_rate": 1.6304104646701818e-05,
+      "loss": 4.0657,
+      "step": 761
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.8362370349072434,
+      "learning_rate": 1.629404286902685e-05,
+      "loss": 4.2315,
+      "step": 762
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.054880336997564,
+      "learning_rate": 1.62839705278991e-05,
+      "loss": 4.1762,
+      "step": 763
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.455945336863922,
+      "learning_rate": 1.627388764022323e-05,
+      "loss": 3.9945,
+      "step": 764
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.089347168700492,
+      "learning_rate": 1.626379422292162e-05,
+      "loss": 4.1632,
+      "step": 765
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.9432572091307634,
+      "learning_rate": 1.6253690292934303e-05,
+      "loss": 4.1629,
+      "step": 766
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.0237380868373296,
+      "learning_rate": 1.624357586721896e-05,
+      "loss": 4.3338,
+      "step": 767
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.6907839068821535,
+      "learning_rate": 1.6233450962750895e-05,
+      "loss": 4.0344,
+      "step": 768
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.241780550240553,
+      "learning_rate": 1.622331559652299e-05,
+      "loss": 4.1167,
+      "step": 769
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.354524584576279,
+      "learning_rate": 1.6213169785545688e-05,
+      "loss": 4.3221,
+      "step": 770
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.787194573086892,
+      "learning_rate": 1.6203013546846967e-05,
+      "loss": 4.0784,
+      "step": 771
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.173642942286069,
+      "learning_rate": 1.61928468974723e-05,
+      "loss": 4.1752,
+      "step": 772
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.9260357041238607,
+      "learning_rate": 1.618266985448463e-05,
+      "loss": 4.2805,
+      "step": 773
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.923440227675206,
+      "learning_rate": 1.6172482434964353e-05,
+      "loss": 4.0797,
+      "step": 774
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.2466075816715,
+      "learning_rate": 1.6162284656009276e-05,
+      "loss": 4.2478,
+      "step": 775
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.969458945548505,
+      "learning_rate": 1.6152076534734585e-05,
+      "loss": 4.2096,
+      "step": 776
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.042108075929687,
+      "learning_rate": 1.6141858088272838e-05,
+      "loss": 3.8777,
+      "step": 777
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.9542589642669697,
+      "learning_rate": 1.6131629333773908e-05,
+      "loss": 4.1506,
+      "step": 778
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.219621112468623,
+      "learning_rate": 1.612139028840498e-05,
+      "loss": 4.0092,
+      "step": 779
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.6474863522574874,
+      "learning_rate": 1.6111140969350504e-05,
+      "loss": 4.1307,
+      "step": 780
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.5346835812296193,
+      "learning_rate": 1.610088139381217e-05,
+      "loss": 4.1175,
+      "step": 781
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.725366186832991,
+      "learning_rate": 1.609061157900889e-05,
+      "loss": 4.0711,
+      "step": 782
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.667673965148235,
+      "learning_rate": 1.6080331542176754e-05,
+      "loss": 3.918,
+      "step": 783
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.085941511402021,
+      "learning_rate": 1.6070041300569014e-05,
+      "loss": 3.9859,
+      "step": 784
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.663966021746003,
+      "learning_rate": 1.6059740871456035e-05,
+      "loss": 3.9296,
+      "step": 785
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.171734077729142,
+      "learning_rate": 1.60494302721253e-05,
+      "loss": 4.25,
+      "step": 786
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.470608583836035,
+      "learning_rate": 1.603910951988135e-05,
+      "loss": 4.1841,
+      "step": 787
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.944785987346097,
+      "learning_rate": 1.602877863204576e-05,
+      "loss": 3.9134,
+      "step": 788
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.679669019856182,
+      "learning_rate": 1.6018437625957135e-05,
+      "loss": 4.1185,
+      "step": 789
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.8456129680158755,
+      "learning_rate": 1.6008086518971037e-05,
+      "loss": 3.9203,
+      "step": 790
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.038185112299229,
+      "learning_rate": 1.599772532846e-05,
+      "loss": 4.1326,
+      "step": 791
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.110740634200455,
+      "learning_rate": 1.598735407181347e-05,
+      "loss": 4.188,
+      "step": 792
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.088496773536847,
+      "learning_rate": 1.5976972766437796e-05,
+      "loss": 3.9539,
+      "step": 793
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.548087741743393,
+      "learning_rate": 1.596658142975618e-05,
+      "loss": 4.0481,
+      "step": 794
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.368884980792822,
+      "learning_rate": 1.5956180079208684e-05,
+      "loss": 4.162,
+      "step": 795
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.320305188255113,
+      "learning_rate": 1.5945768732252144e-05,
+      "loss": 4.143,
+      "step": 796
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.9640549005931587,
+      "learning_rate": 1.5935347406360192e-05,
+      "loss": 3.9122,
+      "step": 797
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.222874057950338,
+      "learning_rate": 1.5924916119023214e-05,
+      "loss": 4.2344,
+      "step": 798
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.5720174004747354,
+      "learning_rate": 1.5914474887748297e-05,
+      "loss": 4.1743,
+      "step": 799
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.231737588287248,
+      "learning_rate": 1.5904023730059227e-05,
+      "loss": 4.0582,
+      "step": 800
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.5855444742625484,
+      "learning_rate": 1.589356266349645e-05,
+      "loss": 4.1918,
+      "step": 801
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.6077052786289086,
+      "learning_rate": 1.5883091705617045e-05,
+      "loss": 3.9826,
+      "step": 802
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.462258780982009,
+      "learning_rate": 1.5872610873994685e-05,
+      "loss": 4.2167,
+      "step": 803
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.915782876006761,
+      "learning_rate": 1.5862120186219614e-05,
+      "loss": 4.0085,
+      "step": 804
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.072359468444531,
+      "learning_rate": 1.5851619659898623e-05,
+      "loss": 3.8627,
+      "step": 805
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.9217692589848068,
+      "learning_rate": 1.5841109312655017e-05,
+      "loss": 3.9516,
+      "step": 806
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.086637237247661,
+      "learning_rate": 1.5830589162128574e-05,
+      "loss": 4.0158,
+      "step": 807
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.674894390048435,
+      "learning_rate": 1.582005922597553e-05,
+      "loss": 3.9575,
+      "step": 808
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.18175058284887,
+      "learning_rate": 1.580951952186856e-05,
+      "loss": 4.0599,
+      "step": 809
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.8499748638554805,
+      "learning_rate": 1.57989700674967e-05,
+      "loss": 4.0656,
+      "step": 810
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.716337060023653,
+      "learning_rate": 1.578841088056538e-05,
+      "loss": 4.157,
+      "step": 811
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.305254563957067,
+      "learning_rate": 1.5777841978796348e-05,
+      "loss": 4.1538,
+      "step": 812
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.559347319241137,
+      "learning_rate": 1.5767263379927663e-05,
+      "loss": 3.9864,
+      "step": 813
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.9159846433046575,
+      "learning_rate": 1.5756675101713657e-05,
+      "loss": 4.1094,
+      "step": 814
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.299425354443105,
+      "learning_rate": 1.5746077161924905e-05,
+      "loss": 4.1367,
+      "step": 815
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.316883826806145,
+      "learning_rate": 1.573546957834821e-05,
+      "loss": 4.0427,
+      "step": 816
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.878824079534932,
+      "learning_rate": 1.572485236878654e-05,
+      "loss": 4.1057,
+      "step": 817
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.495346619133286,
+      "learning_rate": 1.5714225551059027e-05,
+      "loss": 4.0545,
+      "step": 818
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.9758999194804545,
+      "learning_rate": 1.570358914300094e-05,
+      "loss": 4.0423,
+      "step": 819
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.604552161772746,
+      "learning_rate": 1.5692943162463628e-05,
+      "loss": 3.9757,
+      "step": 820
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.851176884663419,
+      "learning_rate": 1.5682287627314513e-05,
+      "loss": 3.9113,
+      "step": 821
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.289288179124551,
+      "learning_rate": 1.5671622555437055e-05,
+      "loss": 4.2302,
+      "step": 822
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7811955845085796,
+      "learning_rate": 1.566094796473071e-05,
+      "loss": 4.0358,
+      "step": 823
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.674141666896367,
+      "learning_rate": 1.565026387311092e-05,
+      "loss": 4.1475,
+      "step": 824
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.5027858102986764,
+      "learning_rate": 1.5639570298509067e-05,
+      "loss": 4.053,
+      "step": 825
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6916507627982593,
+      "learning_rate": 1.562886725887245e-05,
+      "loss": 4.2153,
+      "step": 826
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.561565009887269,
+      "learning_rate": 1.5618154772164257e-05,
+      "loss": 4.1343,
+      "step": 827
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.8244028110454558,
+      "learning_rate": 1.5607432856363523e-05,
+      "loss": 3.8806,
+      "step": 828
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.460749430124214,
+      "learning_rate": 1.559670152946512e-05,
+      "loss": 3.9809,
+      "step": 829
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.755014824313866,
+      "learning_rate": 1.5585960809479698e-05,
+      "loss": 4.0223,
+      "step": 830
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.104852010995523,
+      "learning_rate": 1.5575210714433687e-05,
+      "loss": 4.0399,
+      "step": 831
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7760475308616637,
+      "learning_rate": 1.5564451262369247e-05,
+      "loss": 4.0801,
+      "step": 832
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.2793544157021355,
+      "learning_rate": 1.5553682471344237e-05,
+      "loss": 4.0501,
+      "step": 833
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.629557175983418,
+      "learning_rate": 1.5542904359432198e-05,
+      "loss": 4.0355,
+      "step": 834
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7883624814021895,
+      "learning_rate": 1.5532116944722308e-05,
+      "loss": 4.1627,
+      "step": 835
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.7953856863825313,
+      "learning_rate": 1.5521320245319364e-05,
+      "loss": 3.9935,
+      "step": 836
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.788743717332434,
+      "learning_rate": 1.5510514279343736e-05,
+      "loss": 4.0584,
+      "step": 837
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.138431050489037,
+      "learning_rate": 1.5499699064931354e-05,
+      "loss": 3.9832,
+      "step": 838
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2686778021991447,
+      "learning_rate": 1.5488874620233674e-05,
+      "loss": 4.0388,
+      "step": 839
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.340524461273947,
+      "learning_rate": 1.547804096341763e-05,
+      "loss": 4.1873,
+      "step": 840
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.9077568345088065,
+      "learning_rate": 1.5467198112665632e-05,
+      "loss": 4.0613,
+      "step": 841
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.12493499904231,
+      "learning_rate": 1.5456346086175508e-05,
+      "loss": 4.2309,
+      "step": 842
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.341152194190719,
+      "learning_rate": 1.5445484902160494e-05,
+      "loss": 3.8967,
+      "step": 843
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8927081089261724,
+      "learning_rate": 1.543461457884919e-05,
+      "loss": 3.9914,
+      "step": 844
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.07972545980503,
+      "learning_rate": 1.5423735134485537e-05,
+      "loss": 3.8737,
+      "step": 845
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.153125798186397,
+      "learning_rate": 1.541284658732878e-05,
+      "loss": 4.1366,
+      "step": 846
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.9414313278213275,
+      "learning_rate": 1.540194895565346e-05,
+      "loss": 4.0546,
+      "step": 847
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.007997536551065,
+      "learning_rate": 1.5391042257749338e-05,
+      "loss": 4.0203,
+      "step": 848
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7302176266174256,
+      "learning_rate": 1.5380126511921404e-05,
+      "loss": 4.0391,
+      "step": 849
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.785687015338851,
+      "learning_rate": 1.536920173648984e-05,
+      "loss": 4.0755,
+      "step": 850
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.949326413164564,
+      "learning_rate": 1.5358267949789968e-05,
+      "loss": 3.9168,
+      "step": 851
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.339362469992984,
+      "learning_rate": 1.5347325170172246e-05,
+      "loss": 4.044,
+      "step": 852
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.894035836096662,
+      "learning_rate": 1.533637341600221e-05,
+      "loss": 3.9281,
+      "step": 853
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.585643645017433,
+      "learning_rate": 1.532541270566049e-05,
+      "loss": 3.958,
+      "step": 854
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.369189203074978,
+      "learning_rate": 1.5314443057542703e-05,
+      "loss": 4.1421,
+      "step": 855
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6856868676372194,
+      "learning_rate": 1.5303464490059506e-05,
+      "loss": 4.0606,
+      "step": 856
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7601211848512306,
+      "learning_rate": 1.5292477021636498e-05,
+      "loss": 3.9721,
+      "step": 857
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8571531489482274,
+      "learning_rate": 1.528148067071423e-05,
+      "loss": 3.8854,
+      "step": 858
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.285596884849708,
+      "learning_rate": 1.5270475455748165e-05,
+      "loss": 3.9812,
+      "step": 859
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.5621322826993085,
+      "learning_rate": 1.5259461395208628e-05,
+      "loss": 4.0867,
+      "step": 860
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7099938172783467,
+      "learning_rate": 1.5248438507580806e-05,
+      "loss": 4.1262,
+      "step": 861
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.822370952909795,
+      "learning_rate": 1.5237406811364682e-05,
+      "loss": 3.8992,
+      "step": 862
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.615930344249073,
+      "learning_rate": 1.5226366325075042e-05,
+      "loss": 3.9119,
+      "step": 863
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.141109925790369,
+      "learning_rate": 1.5215317067241415e-05,
+      "loss": 3.922,
+      "step": 864
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.91182469534649,
+      "learning_rate": 1.5204259056408046e-05,
+      "loss": 4.1266,
+      "step": 865
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.739014813917722,
+      "learning_rate": 1.5193192311133884e-05,
+      "loss": 4.1846,
+      "step": 866
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.733700050825112,
+      "learning_rate": 1.5182116849992528e-05,
+      "loss": 4.1189,
+      "step": 867
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.577400393069631,
+      "learning_rate": 1.5171032691572207e-05,
+      "loss": 3.8833,
+      "step": 868
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.9216623914448743,
+      "learning_rate": 1.5159939854475743e-05,
+      "loss": 3.8559,
+      "step": 869
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.248938111614794,
+      "learning_rate": 1.5148838357320537e-05,
+      "loss": 3.9905,
+      "step": 870
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7734583568786384,
+      "learning_rate": 1.5137728218738504e-05,
+      "loss": 4.1165,
+      "step": 871
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3465062079897963,
+      "learning_rate": 1.512660945737608e-05,
+      "loss": 4.1162,
+      "step": 872
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.716306651251567,
+      "learning_rate": 1.5115482091894164e-05,
+      "loss": 4.1707,
+      "step": 873
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.776023574626814,
+      "learning_rate": 1.5104346140968096e-05,
+      "loss": 3.9443,
+      "step": 874
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.3972508182391574,
+      "learning_rate": 1.5093201623287631e-05,
+      "loss": 4.0875,
+      "step": 875
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.557313751621321,
+      "learning_rate": 1.5082048557556892e-05,
+      "loss": 3.9307,
+      "step": 876
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.009659574014854,
+      "learning_rate": 1.507088696249436e-05,
+      "loss": 3.9834,
+      "step": 877
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.4421878630309575,
+      "learning_rate": 1.505971685683282e-05,
+      "loss": 3.9976,
+      "step": 878
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.090750290145485,
+      "learning_rate": 1.5048538259319347e-05,
+      "loss": 3.7983,
+      "step": 879
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.084453101262391,
+      "learning_rate": 1.5037351188715265e-05,
+      "loss": 3.8341,
+      "step": 880
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.7389574740374556,
+      "learning_rate": 1.5026155663796123e-05,
+      "loss": 3.6981,
+      "step": 881
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2705691585414263,
+      "learning_rate": 1.5014951703351655e-05,
+      "loss": 3.9908,
+      "step": 882
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.6909533401956316,
+      "learning_rate": 1.500373932618575e-05,
+      "loss": 3.8345,
+      "step": 883
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8291194048755024,
+      "learning_rate": 1.4992518551116436e-05,
+      "loss": 4.0007,
+      "step": 884
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.9216771533735,
+      "learning_rate": 1.4981289396975818e-05,
+      "loss": 4.0789,
+      "step": 885
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.5513257467131147,
+      "learning_rate": 1.4970051882610073e-05,
+      "loss": 4.0285,
+      "step": 886
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.938085563815545,
+      "learning_rate": 1.4958806026879411e-05,
+      "loss": 3.9651,
+      "step": 887
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.200770892095531,
+      "learning_rate": 1.4947551848658036e-05,
+      "loss": 4.0959,
+      "step": 888
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.9964458964377174,
+      "learning_rate": 1.4936289366834123e-05,
+      "loss": 3.8196,
+      "step": 889
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.3213563792790763,
+      "learning_rate": 1.4925018600309784e-05,
+      "loss": 4.109,
+      "step": 890
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.041337943923999,
+      "learning_rate": 1.4913739568001034e-05,
+      "loss": 4.0301,
+      "step": 891
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.4764085757009324,
+      "learning_rate": 1.4902452288837761e-05,
+      "loss": 3.8225,
+      "step": 892
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.592918007172055,
+      "learning_rate": 1.4891156781763692e-05,
+      "loss": 3.8364,
+      "step": 893
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.592437141265646,
+      "learning_rate": 1.4879853065736366e-05,
+      "loss": 4.0147,
+      "step": 894
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7811593209134333,
+      "learning_rate": 1.4868541159727097e-05,
+      "loss": 3.8975,
+      "step": 895
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.17581172791539,
+      "learning_rate": 1.485722108272095e-05,
+      "loss": 4.0324,
+      "step": 896
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.631807623212032,
+      "learning_rate": 1.4845892853716692e-05,
+      "loss": 3.9237,
+      "step": 897
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7247974116244467,
+      "learning_rate": 1.4834556491726781e-05,
+      "loss": 3.9782,
+      "step": 898
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.953397859539653,
+      "learning_rate": 1.482321201577733e-05,
+      "loss": 4.0207,
+      "step": 899
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.286067218324245,
+      "learning_rate": 1.4811859444908053e-05,
+      "loss": 3.9184,
+      "step": 900
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.9689525465502546,
+      "learning_rate": 1.4800498798172263e-05,
+      "loss": 3.9761,
+      "step": 901
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.9432696950576682,
+      "learning_rate": 1.478913009463682e-05,
+      "loss": 3.8415,
+      "step": 902
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.784967591738362,
+      "learning_rate": 1.4777753353382121e-05,
+      "loss": 3.9511,
+      "step": 903
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.649524601044959,
+      "learning_rate": 1.4766368593502028e-05,
+      "loss": 3.9241,
+      "step": 904
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.125027971868248,
+      "learning_rate": 1.4754975834103877e-05,
+      "loss": 3.8264,
+      "step": 905
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.167521116754602,
+      "learning_rate": 1.474357509430843e-05,
+      "loss": 3.9653,
+      "step": 906
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.694796156291042,
+      "learning_rate": 1.473216639324984e-05,
+      "loss": 3.8908,
+      "step": 907
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.880655396365558,
+      "learning_rate": 1.472074975007562e-05,
+      "loss": 3.8734,
+      "step": 908
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7017895815934265,
+      "learning_rate": 1.4709325183946613e-05,
+      "loss": 3.9442,
+      "step": 909
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.055857844353811,
+      "learning_rate": 1.4697892714036959e-05,
+      "loss": 3.9686,
+      "step": 910
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.938238712811652,
+      "learning_rate": 1.4686452359534067e-05,
+      "loss": 3.9718,
+      "step": 911
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.633536668094118,
+      "learning_rate": 1.467500413963857e-05,
+      "loss": 4.0176,
+      "step": 912
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.039475965540521,
+      "learning_rate": 1.4663548073564316e-05,
+      "loss": 3.8204,
+      "step": 913
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.586728291833797,
+      "learning_rate": 1.4652084180538304e-05,
+      "loss": 3.9394,
+      "step": 914
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.6290231319278643,
+      "learning_rate": 1.4640612479800686e-05,
+      "loss": 3.9651,
+      "step": 915
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.523810100980663,
+      "learning_rate": 1.4629132990604706e-05,
+      "loss": 3.9046,
+      "step": 916
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3352008780739277,
+      "learning_rate": 1.4617645732216686e-05,
+      "loss": 3.9908,
+      "step": 917
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.216385007164741,
+      "learning_rate": 1.4606150723915984e-05,
+      "loss": 4.1919,
+      "step": 918
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3379333114880017,
+      "learning_rate": 1.4594647984994966e-05,
+      "loss": 3.8451,
+      "step": 919
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.326212397572884,
+      "learning_rate": 1.4583137534758968e-05,
+      "loss": 4.0293,
+      "step": 920
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.104354270670977,
+      "learning_rate": 1.4571619392526279e-05,
+      "loss": 4.0689,
+      "step": 921
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.855954672044245,
+      "learning_rate": 1.456009357762809e-05,
+      "loss": 3.8103,
+      "step": 922
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4625438644586413,
+      "learning_rate": 1.4548560109408465e-05,
+      "loss": 3.9354,
+      "step": 923
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.244228225779234,
+      "learning_rate": 1.4537019007224324e-05,
+      "loss": 4.1405,
+      "step": 924
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.736402044407656,
+      "learning_rate": 1.4525470290445392e-05,
+      "loss": 3.9756,
+      "step": 925
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.804301068297979,
+      "learning_rate": 1.4513913978454169e-05,
+      "loss": 4.1198,
+      "step": 926
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.157084681702177,
+      "learning_rate": 1.4502350090645919e-05,
+      "loss": 3.8946,
+      "step": 927
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4439326017172562,
+      "learning_rate": 1.4490778646428601e-05,
+      "loss": 3.9797,
+      "step": 928
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.855937635878492,
+      "learning_rate": 1.4479199665222869e-05,
+      "loss": 3.8229,
+      "step": 929
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.52379856141619,
+      "learning_rate": 1.4467613166462024e-05,
+      "loss": 4.1031,
+      "step": 930
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.343080438472811,
+      "learning_rate": 1.445601916959198e-05,
+      "loss": 3.8626,
+      "step": 931
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.3607075037651586,
+      "learning_rate": 1.4444417694071242e-05,
+      "loss": 4.0184,
+      "step": 932
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.7952187164245066,
+      "learning_rate": 1.4432808759370853e-05,
+      "loss": 3.7856,
+      "step": 933
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.346280524599407,
+      "learning_rate": 1.4421192384974396e-05,
+      "loss": 3.8906,
+      "step": 934
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.4646318285115116,
+      "learning_rate": 1.4409568590377918e-05,
+      "loss": 4.0531,
+      "step": 935
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.182963241175465,
+      "learning_rate": 1.439793739508994e-05,
+      "loss": 3.8542,
+      "step": 936
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.7121887516870933,
+      "learning_rate": 1.4386298818631388e-05,
+      "loss": 3.9009,
+      "step": 937
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.281667995792813,
+      "learning_rate": 1.437465288053558e-05,
+      "loss": 3.9757,
+      "step": 938
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.356495016960309,
+      "learning_rate": 1.4362999600348198e-05,
+      "loss": 3.8267,
+      "step": 939
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.711219010948904,
+      "learning_rate": 1.4351338997627233e-05,
+      "loss": 3.9867,
+      "step": 940
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.556481836187361,
+      "learning_rate": 1.433967109194298e-05,
+      "loss": 3.9451,
+      "step": 941
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.265019987798413,
+      "learning_rate": 1.4327995902877972e-05,
+      "loss": 3.8217,
+      "step": 942
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.703379886530379,
+      "learning_rate": 1.4316313450026986e-05,
+      "loss": 3.8829,
+      "step": 943
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.5889354627304004,
+      "learning_rate": 1.4304623752996974e-05,
+      "loss": 4.1259,
+      "step": 944
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.097019004750559,
+      "learning_rate": 1.429292683140706e-05,
+      "loss": 3.7848,
+      "step": 945
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.7604395396181123,
+      "learning_rate": 1.428122270488848e-05,
+      "loss": 4.0124,
+      "step": 946
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1999883377450113,
+      "learning_rate": 1.4269511393084572e-05,
+      "loss": 4.0831,
+      "step": 947
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.382731690898,
+      "learning_rate": 1.4257792915650728e-05,
+      "loss": 3.8673,
+      "step": 948
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.513739284257866,
+      "learning_rate": 1.4246067292254367e-05,
+      "loss": 3.8318,
+      "step": 949
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.723969538475621,
+      "learning_rate": 1.4234334542574906e-05,
+      "loss": 3.9158,
+      "step": 950
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.389218087656833,
+      "learning_rate": 1.4222594686303707e-05,
+      "loss": 3.9276,
+      "step": 951
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.042149116988585,
+      "learning_rate": 1.4210847743144087e-05,
+      "loss": 3.8478,
+      "step": 952
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.823734180174385,
+      "learning_rate": 1.4199093732811227e-05,
+      "loss": 3.8644,
+      "step": 953
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4666325491254772,
+      "learning_rate": 1.4187332675032189e-05,
+      "loss": 3.9167,
+      "step": 954
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.565609657289887,
+      "learning_rate": 1.4175564589545853e-05,
+      "loss": 3.7594,
+      "step": 955
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.279210813025605,
+      "learning_rate": 1.4163789496102902e-05,
+      "loss": 3.9157,
+      "step": 956
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.6109420515584296,
+      "learning_rate": 1.4152007414465771e-05,
+      "loss": 3.775,
+      "step": 957
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.5183027997433562,
+      "learning_rate": 1.4140218364408634e-05,
+      "loss": 3.8613,
+      "step": 958
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.7680112519128635,
+      "learning_rate": 1.4128422365717346e-05,
+      "loss": 3.8931,
+      "step": 959
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.065893326610565,
+      "learning_rate": 1.411661943818944e-05,
+      "loss": 3.926,
+      "step": 960
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.062397562509862,
+      "learning_rate": 1.4104809601634069e-05,
+      "loss": 3.8562,
+      "step": 961
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.143634258192227,
+      "learning_rate": 1.409299287587198e-05,
+      "loss": 3.7753,
+      "step": 962
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.103196921879452,
+      "learning_rate": 1.4081169280735488e-05,
+      "loss": 4.0006,
+      "step": 963
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7702584463005606,
+      "learning_rate": 1.4069338836068434e-05,
+      "loss": 3.8562,
+      "step": 964
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.664516815419518,
+      "learning_rate": 1.4057501561726157e-05,
+      "loss": 4.1171,
+      "step": 965
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.196260439411363,
+      "learning_rate": 1.404565747757545e-05,
+      "loss": 3.872,
+      "step": 966
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.837246468651308,
+      "learning_rate": 1.403380660349455e-05,
+      "loss": 4.0388,
+      "step": 967
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.665536340443557,
+      "learning_rate": 1.4021948959373075e-05,
+      "loss": 3.9043,
+      "step": 968
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.561871423373831,
+      "learning_rate": 1.4010084565112018e-05,
+      "loss": 4.0049,
+      "step": 969
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.31227711238914,
+      "learning_rate": 1.3998213440623691e-05,
+      "loss": 3.9867,
+      "step": 970
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3638177448238977,
+      "learning_rate": 1.3986335605831707e-05,
+      "loss": 4.2255,
+      "step": 971
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7103949192821384,
+      "learning_rate": 1.3974451080670934e-05,
+      "loss": 3.9219,
+      "step": 972
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.981342474110125,
+      "learning_rate": 1.3962559885087482e-05,
+      "loss": 3.8157,
+      "step": 973
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.679894552303274,
+      "learning_rate": 1.3950662039038643e-05,
+      "loss": 3.9717,
+      "step": 974
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.367030262675531,
+      "learning_rate": 1.3938757562492873e-05,
+      "loss": 3.8756,
+      "step": 975
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7171472220445456,
+      "learning_rate": 1.3926846475429767e-05,
+      "loss": 3.8459,
+      "step": 976
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.508351826758392,
+      "learning_rate": 1.3914928797839996e-05,
+      "loss": 3.6319,
+      "step": 977
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.474974790718888,
+      "learning_rate": 1.3903004549725313e-05,
+      "loss": 3.8134,
+      "step": 978
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.555949869772792,
+      "learning_rate": 1.3891073751098481e-05,
+      "loss": 3.917,
+      "step": 979
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.225391171204873,
+      "learning_rate": 1.3879136421983265e-05,
+      "loss": 3.9551,
+      "step": 980
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.666887533774541,
+      "learning_rate": 1.3867192582414393e-05,
+      "loss": 3.8632,
+      "step": 981
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.802718705221923,
+      "learning_rate": 1.3855242252437511e-05,
+      "loss": 3.9409,
+      "step": 982
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5413928934272914,
+      "learning_rate": 1.3843285452109166e-05,
+      "loss": 3.8817,
+      "step": 983
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.159489305465269,
+      "learning_rate": 1.3831322201496757e-05,
+      "loss": 3.857,
+      "step": 984
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.823310889696872,
+      "learning_rate": 1.3819352520678519e-05,
+      "loss": 4.1045,
+      "step": 985
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3447101805626867,
+      "learning_rate": 1.3807376429743467e-05,
+      "loss": 3.9642,
+      "step": 986
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.869554340354115,
+      "learning_rate": 1.3795393948791382e-05,
+      "loss": 3.8103,
+      "step": 987
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.624295111996602,
+      "learning_rate": 1.3783405097932772e-05,
+      "loss": 3.7571,
+      "step": 988
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.8202389870193754,
+      "learning_rate": 1.3771409897288823e-05,
+      "loss": 3.7456,
+      "step": 989
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4182725294539593,
+      "learning_rate": 1.3759408366991391e-05,
+      "loss": 3.9806,
+      "step": 990
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.899990351720217,
+      "learning_rate": 1.3747400527182952e-05,
+      "loss": 4.1462,
+      "step": 991
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2462833506393514,
+      "learning_rate": 1.373538639801657e-05,
+      "loss": 4.1141,
+      "step": 992
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.778151533518385,
+      "learning_rate": 1.3723365999655859e-05,
+      "loss": 3.8378,
+      "step": 993
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4612512836444447,
+      "learning_rate": 1.3711339352274969e-05,
+      "loss": 3.961,
+      "step": 994
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4892336558699677,
+      "learning_rate": 1.3699306476058523e-05,
+      "loss": 3.8107,
+      "step": 995
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.667163737052047,
+      "learning_rate": 1.3687267391201604e-05,
+      "loss": 3.8194,
+      "step": 996
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7753395478216945,
+      "learning_rate": 1.3675222117909716e-05,
+      "loss": 3.8419,
+      "step": 997
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.210105830897785,
+      "learning_rate": 1.366317067639875e-05,
+      "loss": 3.8023,
+      "step": 998
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.723917366613343,
+      "learning_rate": 1.3651113086894951e-05,
+      "loss": 4.0149,
+      "step": 999
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.244202977967538,
+      "learning_rate": 1.3639049369634878e-05,
+      "loss": 4.0083,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.02953633918017,
+      "learning_rate": 1.3626979544865369e-05,
+      "loss": 3.6242,
+      "step": 1001
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.8798268517813534,
+      "learning_rate": 1.3614903632843523e-05,
+      "loss": 3.8001,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.9228004237258767,
+      "learning_rate": 1.3602821653836654e-05,
+      "loss": 3.8272,
+      "step": 1003
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.904727851654214,
+      "learning_rate": 1.3590733628122253e-05,
+      "loss": 3.9567,
+      "step": 1004
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7882268070089697,
+      "learning_rate": 1.357863957598796e-05,
+      "loss": 3.8891,
+      "step": 1005
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.706967784285879,
+      "learning_rate": 1.3566539517731536e-05,
+      "loss": 3.8622,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.748018650903142,
+      "learning_rate": 1.3554433473660818e-05,
+      "loss": 3.7923,
+      "step": 1007
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4758174617619013,
+      "learning_rate": 1.354232146409368e-05,
+      "loss": 3.8247,
+      "step": 1008
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.646296934730249,
+      "learning_rate": 1.353020350935803e-05,
+      "loss": 3.828,
+      "step": 1009
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.470041626731096,
+      "learning_rate": 1.3518079629791725e-05,
+      "loss": 3.7678,
+      "step": 1010
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.079568258122262,
+      "learning_rate": 1.3505949845742599e-05,
+      "loss": 3.7364,
+      "step": 1011
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.8456370828497772,
+      "learning_rate": 1.3493814177568365e-05,
+      "loss": 3.9458,
+      "step": 1012
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.6309196422021865,
+      "learning_rate": 1.3481672645636627e-05,
+      "loss": 3.8194,
+      "step": 1013
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.163434543886125,
+      "learning_rate": 1.3469525270324835e-05,
+      "loss": 3.728,
+      "step": 1014
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.21450380310228,
+      "learning_rate": 1.345737207202023e-05,
+      "loss": 3.9219,
+      "step": 1015
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.974894908073055,
+      "learning_rate": 1.3445213071119841e-05,
+      "loss": 3.795,
+      "step": 1016
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.7309967078096324,
+      "learning_rate": 1.3433048288030424e-05,
+      "loss": 3.9291,
+      "step": 1017
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.882915305390519,
+      "learning_rate": 1.342087774316845e-05,
+      "loss": 3.9465,
+      "step": 1018
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.298426712564282,
+      "learning_rate": 1.3408701456960052e-05,
+      "loss": 3.9188,
+      "step": 1019
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.075499444985152,
+      "learning_rate": 1.3396519449841006e-05,
+      "loss": 3.8499,
+      "step": 1020
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.058588213246766,
+      "learning_rate": 1.338433174225668e-05,
+      "loss": 3.9857,
+      "step": 1021
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.441001425059062,
+      "learning_rate": 1.3372138354662018e-05,
+      "loss": 3.6495,
+      "step": 1022
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.737403468138393,
+      "learning_rate": 1.3359939307521494e-05,
+      "loss": 3.5714,
+      "step": 1023
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.489644852015869,
+      "learning_rate": 1.3347734621309076e-05,
+      "loss": 3.7954,
+      "step": 1024
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.698554077016783,
+      "learning_rate": 1.3335524316508208e-05,
+      "loss": 3.7721,
+      "step": 1025
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.537685736192624,
+      "learning_rate": 1.3323308413611748e-05,
+      "loss": 3.6806,
+      "step": 1026
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.6844980885590415,
+      "learning_rate": 1.3311086933121961e-05,
+      "loss": 3.95,
+      "step": 1027
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.7048407750955152,
+      "learning_rate": 1.3298859895550473e-05,
+      "loss": 3.9469,
+      "step": 1028
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.66973118936307,
+      "learning_rate": 1.3286627321418229e-05,
+      "loss": 3.8629,
+      "step": 1029
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.8653046493517436,
+      "learning_rate": 1.3274389231255466e-05,
+      "loss": 3.8781,
+      "step": 1030
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.60833520162824,
+      "learning_rate": 1.3262145645601693e-05,
+      "loss": 3.786,
+      "step": 1031
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.5817623342333627,
+      "learning_rate": 1.3249896585005628e-05,
+      "loss": 3.877,
+      "step": 1032
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.387212721078502,
+      "learning_rate": 1.3237642070025183e-05,
+      "loss": 3.8994,
+      "step": 1033
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.159428683537451,
+      "learning_rate": 1.322538212122742e-05,
+      "loss": 3.7944,
+      "step": 1034
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.6203299759058987,
+      "learning_rate": 1.3213116759188525e-05,
+      "loss": 3.985,
+      "step": 1035
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.1657884558644,
+      "learning_rate": 1.320084600449377e-05,
+      "loss": 3.7315,
+      "step": 1036
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.281052074286255,
+      "learning_rate": 1.3188569877737474e-05,
+      "loss": 3.7459,
+      "step": 1037
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.7677672464400835,
+      "learning_rate": 1.3176288399522975e-05,
+      "loss": 3.5716,
+      "step": 1038
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.428918038754764,
+      "learning_rate": 1.3164001590462592e-05,
+      "loss": 3.8393,
+      "step": 1039
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.298395603507693,
+      "learning_rate": 1.3151709471177589e-05,
+      "loss": 3.8122,
+      "step": 1040
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.280164110278819,
+      "learning_rate": 1.3139412062298141e-05,
+      "loss": 3.5995,
+      "step": 1041
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.500876125985926,
+      "learning_rate": 1.312710938446331e-05,
+      "loss": 3.8619,
+      "step": 1042
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.6210925789968,
+      "learning_rate": 1.3114801458320988e-05,
+      "loss": 3.6235,
+      "step": 1043
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.7574064252584147,
+      "learning_rate": 1.3102488304527883e-05,
+      "loss": 3.8639,
+      "step": 1044
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.1328056061368095,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 3.987,
+      "step": 1045
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.01823522119479,
+      "learning_rate": 1.3077846396659986e-05,
+      "loss": 3.7406,
+      "step": 1046
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.1690337753538556,
+      "learning_rate": 1.3065517683942339e-05,
+      "loss": 3.8761,
+      "step": 1047
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.118559827596486,
+      "learning_rate": 1.3053183826288124e-05,
+      "loss": 3.8011,
+      "step": 1048
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.382326156860178,
+      "learning_rate": 1.3040844844397573e-05,
+      "loss": 3.7709,
+      "step": 1049
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.4798186185746225,
+      "learning_rate": 1.3028500758979507e-05,
+      "loss": 3.9834,
+      "step": 1050
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.40810523021627,
+      "learning_rate": 1.3016151590751332e-05,
+      "loss": 3.7962,
+      "step": 1051
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.8039902850152925,
+      "learning_rate": 1.3003797360438961e-05,
+      "loss": 3.9683,
+      "step": 1052
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.894557739361905,
+      "learning_rate": 1.2991438088776818e-05,
+      "loss": 3.6801,
+      "step": 1053
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.239443292560934,
+      "learning_rate": 1.2979073796507786e-05,
+      "loss": 3.8762,
+      "step": 1054
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5021977746058317,
+      "learning_rate": 1.296670450438317e-05,
+      "loss": 3.774,
+      "step": 1055
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.1109819520202056,
+      "learning_rate": 1.2954330233162669e-05,
+      "loss": 3.7251,
+      "step": 1056
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.8759541874915167,
+      "learning_rate": 1.2941951003614337e-05,
+      "loss": 3.7179,
+      "step": 1057
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.6162051772187995,
+      "learning_rate": 1.2929566836514556e-05,
+      "loss": 3.7607,
+      "step": 1058
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.801727408834567,
+      "learning_rate": 1.291717775264798e-05,
+      "loss": 3.8958,
+      "step": 1059
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.407215697298154,
+      "learning_rate": 1.2904783772807534e-05,
+      "loss": 3.8602,
+      "step": 1060
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.560458114043935,
+      "learning_rate": 1.2892384917794347e-05,
+      "loss": 3.9323,
+      "step": 1061
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.8926615511691383,
+      "learning_rate": 1.2879981208417735e-05,
+      "loss": 3.6922,
+      "step": 1062
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.022272943906542,
+      "learning_rate": 1.2867572665495156e-05,
+      "loss": 3.8297,
+      "step": 1063
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.174392840890098,
+      "learning_rate": 1.285515930985219e-05,
+      "loss": 3.7822,
+      "step": 1064
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.48534689846332,
+      "learning_rate": 1.2842741162322487e-05,
+      "loss": 3.9492,
+      "step": 1065
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7857352405447555,
+      "learning_rate": 1.2830318243747736e-05,
+      "loss": 3.8072,
+      "step": 1066
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.607574037963277,
+      "learning_rate": 1.2817890574977648e-05,
+      "loss": 3.9189,
+      "step": 1067
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.4592415194671173,
+      "learning_rate": 1.2805458176869885e-05,
+      "loss": 3.8008,
+      "step": 1068
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7386853887528777,
+      "learning_rate": 1.2793021070290065e-05,
+      "loss": 3.8409,
+      "step": 1069
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.53759324100156,
+      "learning_rate": 1.2780579276111702e-05,
+      "loss": 3.6507,
+      "step": 1070
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.945952450142829,
+      "learning_rate": 1.2768132815216174e-05,
+      "loss": 3.8486,
+      "step": 1071
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.8818735387168606,
+      "learning_rate": 1.2755681708492696e-05,
+      "loss": 3.751,
+      "step": 1072
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.159697782524046,
+      "learning_rate": 1.2743225976838277e-05,
+      "loss": 3.7202,
+      "step": 1073
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.688463314558805,
+      "learning_rate": 1.2730765641157689e-05,
+      "loss": 3.8645,
+      "step": 1074
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.8295978281953693,
+      "learning_rate": 1.2718300722363431e-05,
+      "loss": 3.7975,
+      "step": 1075
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.924531322625543,
+      "learning_rate": 1.2705831241375695e-05,
+      "loss": 3.5147,
+      "step": 1076
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.318083146218096,
+      "learning_rate": 1.2693357219122331e-05,
+      "loss": 3.9557,
+      "step": 1077
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.677130448505102,
+      "learning_rate": 1.2680878676538804e-05,
+      "loss": 3.7197,
+      "step": 1078
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.8463986020512575,
+      "learning_rate": 1.2668395634568175e-05,
+      "loss": 3.777,
+      "step": 1079
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.058312752920507,
+      "learning_rate": 1.2655908114161053e-05,
+      "loss": 4.0007,
+      "step": 1080
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.638632419006762,
+      "learning_rate": 1.2643416136275557e-05,
+      "loss": 3.7884,
+      "step": 1081
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5887543927846846,
+      "learning_rate": 1.2630919721877299e-05,
+      "loss": 3.8393,
+      "step": 1082
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.149681723639608,
+      "learning_rate": 1.261841889193932e-05,
+      "loss": 3.7256,
+      "step": 1083
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.897842387997689,
+      "learning_rate": 1.2605913667442096e-05,
+      "loss": 3.817,
+      "step": 1084
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.088190503564847,
+      "learning_rate": 1.2593404069373452e-05,
+      "loss": 3.7328,
+      "step": 1085
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.346641744766435,
+      "learning_rate": 1.2580890118728572e-05,
+      "loss": 3.6659,
+      "step": 1086
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.9172265316872497,
+      "learning_rate": 1.2568371836509936e-05,
+      "loss": 3.6883,
+      "step": 1087
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.0186942695070895,
+      "learning_rate": 1.2555849243727298e-05,
+      "loss": 3.756,
+      "step": 1088
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.2788739628746395,
+      "learning_rate": 1.2543322361397648e-05,
+      "loss": 3.8754,
+      "step": 1089
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.187805152986359,
+      "learning_rate": 1.2530791210545163e-05,
+      "loss": 3.7108,
+      "step": 1090
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.32176195886882,
+      "learning_rate": 1.2518255812201203e-05,
+      "loss": 3.648,
+      "step": 1091
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.967619501198415,
+      "learning_rate": 1.2505716187404242e-05,
+      "loss": 3.9049,
+      "step": 1092
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.8773142998279955,
+      "learning_rate": 1.2493172357199856e-05,
+      "loss": 3.6919,
+      "step": 1093
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6144010815109398,
+      "learning_rate": 1.2480624342640673e-05,
+      "loss": 3.7658,
+      "step": 1094
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.2419265154497285,
+      "learning_rate": 1.2468072164786342e-05,
+      "loss": 3.8289,
+      "step": 1095
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.481230443413657,
+      "learning_rate": 1.2455515844703512e-05,
+      "loss": 3.8465,
+      "step": 1096
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.165489059771342,
+      "learning_rate": 1.2442955403465768e-05,
+      "loss": 3.6346,
+      "step": 1097
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.8231188657999677,
+      "learning_rate": 1.2430390862153625e-05,
+      "loss": 3.8528,
+      "step": 1098
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.896380824909673,
+      "learning_rate": 1.2417822241854466e-05,
+      "loss": 3.785,
+      "step": 1099
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.190572844245639,
+      "learning_rate": 1.2405249563662539e-05,
+      "loss": 3.9942,
+      "step": 1100
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.049608251353112,
+      "learning_rate": 1.2392672848678877e-05,
+      "loss": 3.8083,
+      "step": 1101
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1717023508880833,
+      "learning_rate": 1.238009211801131e-05,
+      "loss": 3.8426,
+      "step": 1102
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.570540640509857,
+      "learning_rate": 1.2367507392774398e-05,
+      "loss": 3.5437,
+      "step": 1103
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5967168685654696,
+      "learning_rate": 1.2354918694089406e-05,
+      "loss": 3.7432,
+      "step": 1104
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6697195244996283,
+      "learning_rate": 1.2342326043084268e-05,
+      "loss": 3.7574,
+      "step": 1105
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.536043300169957,
+      "learning_rate": 1.2329729460893552e-05,
+      "loss": 3.6837,
+      "step": 1106
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.736616862422548,
+      "learning_rate": 1.2317128968658424e-05,
+      "loss": 3.9007,
+      "step": 1107
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6992203544786433,
+      "learning_rate": 1.2304524587526609e-05,
+      "loss": 3.9669,
+      "step": 1108
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.499394891696174,
+      "learning_rate": 1.2291916338652365e-05,
+      "loss": 3.7316,
+      "step": 1109
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.8376821119143663,
+      "learning_rate": 1.2279304243196438e-05,
+      "loss": 3.952,
+      "step": 1110
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.12009679545591,
+      "learning_rate": 1.2266688322326024e-05,
+      "loss": 3.8621,
+      "step": 1111
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.411885666351011,
+      "learning_rate": 1.225406859721475e-05,
+      "loss": 3.8947,
+      "step": 1112
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.975578622769097,
+      "learning_rate": 1.2241445089042623e-05,
+      "loss": 3.636,
+      "step": 1113
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.414469682007408,
+      "learning_rate": 1.2228817818995998e-05,
+      "loss": 3.806,
+      "step": 1114
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.831971776801647,
+      "learning_rate": 1.2216186808267544e-05,
+      "loss": 3.9733,
+      "step": 1115
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.810781796334439,
+      "learning_rate": 1.2203552078056209e-05,
+      "loss": 3.772,
+      "step": 1116
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.884988279276643,
+      "learning_rate": 1.2190913649567185e-05,
+      "loss": 3.932,
+      "step": 1117
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.5364222816956867,
+      "learning_rate": 1.2178271544011864e-05,
+      "loss": 3.8833,
+      "step": 1118
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.086827186062164,
+      "learning_rate": 1.2165625782607817e-05,
+      "loss": 3.8387,
+      "step": 1119
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.2399315656735865,
+      "learning_rate": 1.215297638657875e-05,
+      "loss": 3.7389,
+      "step": 1120
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.317299756870864,
+      "learning_rate": 1.2140323377154467e-05,
+      "loss": 3.9934,
+      "step": 1121
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.755435455736925,
+      "learning_rate": 1.2127666775570837e-05,
+      "loss": 3.6894,
+      "step": 1122
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.702006572771475,
+      "learning_rate": 1.211500660306975e-05,
+      "loss": 3.7638,
+      "step": 1123
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.6732674412435555,
+      "learning_rate": 1.210234288089911e-05,
+      "loss": 3.5788,
+      "step": 1124
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.733951889792287,
+      "learning_rate": 1.2089675630312755e-05,
+      "loss": 3.7181,
+      "step": 1125
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.5632865766418425,
+      "learning_rate": 1.2077004872570454e-05,
+      "loss": 3.7373,
+      "step": 1126
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.682464186358648,
+      "learning_rate": 1.206433062893787e-05,
+      "loss": 3.7892,
+      "step": 1127
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8827651672073524,
+      "learning_rate": 1.2051652920686505e-05,
+      "loss": 3.6032,
+      "step": 1128
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.079285241985473,
+      "learning_rate": 1.2038971769093685e-05,
+      "loss": 3.8538,
+      "step": 1129
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.306943261397105,
+      "learning_rate": 1.2026287195442503e-05,
+      "loss": 3.7325,
+      "step": 1130
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.058850469182537,
+      "learning_rate": 1.201359922102181e-05,
+      "loss": 3.7228,
+      "step": 1131
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8833854615216725,
+      "learning_rate": 1.200090786712615e-05,
+      "loss": 3.9011,
+      "step": 1132
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.401892254030929,
+      "learning_rate": 1.1988213155055754e-05,
+      "loss": 3.7903,
+      "step": 1133
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.239794919660805,
+      "learning_rate": 1.1975515106116472e-05,
+      "loss": 3.7108,
+      "step": 1134
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.857128357765898,
+      "learning_rate": 1.1962813741619777e-05,
+      "loss": 3.8152,
+      "step": 1135
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.7166704461674835,
+      "learning_rate": 1.1950109082882681e-05,
+      "loss": 3.692,
+      "step": 1136
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.529483667781026,
+      "learning_rate": 1.193740115122774e-05,
+      "loss": 4.073,
+      "step": 1137
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.79176223989985,
+      "learning_rate": 1.1924689967983006e-05,
+      "loss": 3.8253,
+      "step": 1138
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.603835067055338,
+      "learning_rate": 1.191197555448197e-05,
+      "loss": 3.8131,
+      "step": 1139
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6152376513863014,
+      "learning_rate": 1.189925793206357e-05,
+      "loss": 3.7461,
+      "step": 1140
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.410975002301851,
+      "learning_rate": 1.1886537122072106e-05,
+      "loss": 3.901,
+      "step": 1141
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3186385039853183,
+      "learning_rate": 1.187381314585725e-05,
+      "loss": 3.9929,
+      "step": 1142
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.7670461549530887,
+      "learning_rate": 1.1861086024773963e-05,
+      "loss": 3.7552,
+      "step": 1143
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.564484757844791,
+      "learning_rate": 1.1848355780182502e-05,
+      "loss": 3.8756,
+      "step": 1144
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.719641849142067,
+      "learning_rate": 1.1835622433448361e-05,
+      "loss": 3.6392,
+      "step": 1145
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.699068447698048,
+      "learning_rate": 1.1822886005942244e-05,
+      "loss": 3.6641,
+      "step": 1146
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.7519943682708763,
+      "learning_rate": 1.1810146519040023e-05,
+      "loss": 3.8413,
+      "step": 1147
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.5090371131676132,
+      "learning_rate": 1.1797403994122698e-05,
+      "loss": 3.7514,
+      "step": 1148
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.483919091866205,
+      "learning_rate": 1.178465845257638e-05,
+      "loss": 3.8969,
+      "step": 1149
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.9152410753151496,
+      "learning_rate": 1.177190991579223e-05,
+      "loss": 3.7882,
+      "step": 1150
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.7833928552124285,
+      "learning_rate": 1.1759158405166446e-05,
+      "loss": 3.8664,
+      "step": 1151
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.217725364864304,
+      "learning_rate": 1.1746403942100215e-05,
+      "loss": 3.807,
+      "step": 1152
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4669536263663465,
+      "learning_rate": 1.1733646547999678e-05,
+      "loss": 3.7389,
+      "step": 1153
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4928261841706183,
+      "learning_rate": 1.1720886244275893e-05,
+      "loss": 3.7921,
+      "step": 1154
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.989073154700081,
+      "learning_rate": 1.1708123052344803e-05,
+      "loss": 3.6648,
+      "step": 1155
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.693225989563176,
+      "learning_rate": 1.1695356993627203e-05,
+      "loss": 3.9464,
+      "step": 1156
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.5615429339832385,
+      "learning_rate": 1.1682588089548692e-05,
+      "loss": 3.6917,
+      "step": 1157
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.680963349085811,
+      "learning_rate": 1.1669816361539647e-05,
+      "loss": 3.7782,
+      "step": 1158
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.598393270654093,
+      "learning_rate": 1.1657041831035186e-05,
+      "loss": 3.7299,
+      "step": 1159
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.466531188320706,
+      "learning_rate": 1.164426451947513e-05,
+      "loss": 3.6484,
+      "step": 1160
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.035699717288,
+      "learning_rate": 1.1631484448303964e-05,
+      "loss": 3.62,
+      "step": 1161
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.121097309892934,
+      "learning_rate": 1.1618701638970815e-05,
+      "loss": 3.6402,
+      "step": 1162
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.323877433118681,
+      "learning_rate": 1.1605916112929388e-05,
+      "loss": 3.6339,
+      "step": 1163
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.3841070809544185,
+      "learning_rate": 1.1593127891637968e-05,
+      "loss": 3.8707,
+      "step": 1164
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.361551441812719,
+      "learning_rate": 1.1580336996559343e-05,
+      "loss": 3.702,
+      "step": 1165
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.073065467204595,
+      "learning_rate": 1.156754344916081e-05,
+      "loss": 3.8247,
+      "step": 1166
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.520324161669763,
+      "learning_rate": 1.1554747270914098e-05,
+      "loss": 3.5686,
+      "step": 1167
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.654913388365398,
+      "learning_rate": 1.1541948483295358e-05,
+      "loss": 3.8035,
+      "step": 1168
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2461626059038178,
+      "learning_rate": 1.1529147107785129e-05,
+      "loss": 3.7129,
+      "step": 1169
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.180515399607351,
+      "learning_rate": 1.151634316586828e-05,
+      "loss": 3.8563,
+      "step": 1170
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.5441403758051315,
+      "learning_rate": 1.1503536679034e-05,
+      "loss": 3.6948,
+      "step": 1171
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.078399152054442,
+      "learning_rate": 1.1490727668775735e-05,
+      "loss": 3.7647,
+      "step": 1172
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.17839698744333,
+      "learning_rate": 1.147791615659118e-05,
+      "loss": 3.6863,
+      "step": 1173
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.411722230592826,
+      "learning_rate": 1.1465102163982218e-05,
+      "loss": 3.7029,
+      "step": 1174
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.572334871578432,
+      "learning_rate": 1.1452285712454905e-05,
+      "loss": 3.7514,
+      "step": 1175
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.010392827821349,
+      "learning_rate": 1.1439466823519414e-05,
+      "loss": 3.7983,
+      "step": 1176
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.15222057397017,
+      "learning_rate": 1.1426645518690015e-05,
+      "loss": 3.6849,
+      "step": 1177
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.519535347933383,
+      "learning_rate": 1.1413821819485035e-05,
+      "loss": 3.6528,
+      "step": 1178
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.14465571981964,
+      "learning_rate": 1.140099574742681e-05,
+      "loss": 3.8408,
+      "step": 1179
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.8855608293576704,
+      "learning_rate": 1.138816732404167e-05,
+      "loss": 3.7254,
+      "step": 1180
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.66606300571839,
+      "learning_rate": 1.1375336570859877e-05,
+      "loss": 3.6236,
+      "step": 1181
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.159327154865983,
+      "learning_rate": 1.136250350941562e-05,
+      "loss": 3.6226,
+      "step": 1182
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.8058559454023757,
+      "learning_rate": 1.1349668161246945e-05,
+      "loss": 3.7024,
+      "step": 1183
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.583400400011528,
+      "learning_rate": 1.1336830547895752e-05,
+      "loss": 3.5792,
+      "step": 1184
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7655923073287294,
+      "learning_rate": 1.1323990690907734e-05,
+      "loss": 3.7676,
+      "step": 1185
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.466605227178359,
+      "learning_rate": 1.1311148611832346e-05,
+      "loss": 3.7292,
+      "step": 1186
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7647284604385653,
+      "learning_rate": 1.129830433222278e-05,
+      "loss": 3.9269,
+      "step": 1187
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.733028020978156,
+      "learning_rate": 1.128545787363592e-05,
+      "loss": 3.6541,
+      "step": 1188
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.784316104630818,
+      "learning_rate": 1.1272609257632305e-05,
+      "loss": 3.7619,
+      "step": 1189
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.6535503281391337,
+      "learning_rate": 1.1259758505776092e-05,
+      "loss": 4.0232,
+      "step": 1190
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.9417157274952443,
+      "learning_rate": 1.1246905639635029e-05,
+      "loss": 3.8453,
+      "step": 1191
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8028096902179827,
+      "learning_rate": 1.1234050680780407e-05,
+      "loss": 3.7516,
+      "step": 1192
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.259264025052849,
+      "learning_rate": 1.1221193650787032e-05,
+      "loss": 3.823,
+      "step": 1193
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.238159290220217,
+      "learning_rate": 1.1208334571233186e-05,
+      "loss": 3.9508,
+      "step": 1194
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.763516705043185,
+      "learning_rate": 1.119547346370059e-05,
+      "loss": 3.6778,
+      "step": 1195
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.801301507686709,
+      "learning_rate": 1.118261034977437e-05,
+      "loss": 3.5536,
+      "step": 1196
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.0995292505809795,
+      "learning_rate": 1.116974525104302e-05,
+      "loss": 3.5847,
+      "step": 1197
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.052881858623986,
+      "learning_rate": 1.1156878189098357e-05,
+      "loss": 3.7869,
+      "step": 1198
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.759889351369552,
+      "learning_rate": 1.114400918553551e-05,
+      "loss": 3.7752,
+      "step": 1199
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.641606335216125,
+      "learning_rate": 1.1131138261952845e-05,
+      "loss": 3.7229,
+      "step": 1200
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2624956103832945,
+      "learning_rate": 1.1118265439951968e-05,
+      "loss": 3.5944,
+      "step": 1201
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.1450940216557415,
+      "learning_rate": 1.110539074113766e-05,
+      "loss": 3.7048,
+      "step": 1202
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3447326327359326,
+      "learning_rate": 1.1092514187117865e-05,
+      "loss": 3.6985,
+      "step": 1203
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.5461055814418447,
+      "learning_rate": 1.1079635799503625e-05,
+      "loss": 3.7647,
+      "step": 1204
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3556065313219903,
+      "learning_rate": 1.1066755599909065e-05,
+      "loss": 3.734,
+      "step": 1205
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8317618472216592,
+      "learning_rate": 1.1053873609951362e-05,
+      "loss": 3.5126,
+      "step": 1206
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8356744970212726,
+      "learning_rate": 1.1040989851250678e-05,
+      "loss": 3.8545,
+      "step": 1207
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8078493058296803,
+      "learning_rate": 1.1028104345430161e-05,
+      "loss": 3.6943,
+      "step": 1208
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.806203159509349,
+      "learning_rate": 1.1015217114115884e-05,
+      "loss": 3.5924,
+      "step": 1209
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4329163591887006,
+      "learning_rate": 1.1002328178936813e-05,
+      "loss": 3.5889,
+      "step": 1210
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2311989382515263,
+      "learning_rate": 1.0989437561524776e-05,
+      "loss": 3.8365,
+      "step": 1211
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.7876161541916766,
+      "learning_rate": 1.097654528351443e-05,
+      "loss": 3.7401,
+      "step": 1212
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.450794072226459,
+      "learning_rate": 1.0963651366543214e-05,
+      "loss": 3.6452,
+      "step": 1213
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4025224897119974,
+      "learning_rate": 1.095075583225131e-05,
+      "loss": 3.7985,
+      "step": 1214
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.266609502557084,
+      "learning_rate": 1.0937858702281631e-05,
+      "loss": 3.653,
+      "step": 1215
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.352530022697998,
+      "learning_rate": 1.0924959998279754e-05,
+      "loss": 3.5712,
+      "step": 1216
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.7920904023153685,
+      "learning_rate": 1.0912059741893908e-05,
+      "loss": 3.5353,
+      "step": 1217
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.27009447990584,
+      "learning_rate": 1.089915795477492e-05,
+      "loss": 3.6817,
+      "step": 1218
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.093104784383785,
+      "learning_rate": 1.0886254658576186e-05,
+      "loss": 3.5539,
+      "step": 1219
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.121766404390022,
+      "learning_rate": 1.087334987495364e-05,
+      "loss": 3.6973,
+      "step": 1220
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.6097287264286777,
+      "learning_rate": 1.0860443625565712e-05,
+      "loss": 3.7054,
+      "step": 1221
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.099101696343389,
+      "learning_rate": 1.0847535932073288e-05,
+      "loss": 3.821,
+      "step": 1222
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.7239613892649963,
+      "learning_rate": 1.0834626816139678e-05,
+      "loss": 3.8649,
+      "step": 1223
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.927013781158877,
+      "learning_rate": 1.0821716299430577e-05,
+      "loss": 3.7919,
+      "step": 1224
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.926068050669173,
+      "learning_rate": 1.0808804403614044e-05,
+      "loss": 3.7943,
+      "step": 1225
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.9343822616688335,
+      "learning_rate": 1.0795891150360435e-05,
+      "loss": 3.6685,
+      "step": 1226
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.816742376232799,
+      "learning_rate": 1.0782976561342398e-05,
+      "loss": 3.5803,
+      "step": 1227
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3950844811577676,
+      "learning_rate": 1.0770060658234815e-05,
+      "loss": 3.6176,
+      "step": 1228
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.9333804501497704,
+      "learning_rate": 1.0757143462714777e-05,
+      "loss": 3.5919,
+      "step": 1229
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.418557233366642,
+      "learning_rate": 1.0744224996461541e-05,
+      "loss": 3.6716,
+      "step": 1230
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.7898331498735236,
+      "learning_rate": 1.0731305281156499e-05,
+      "loss": 3.6412,
+      "step": 1231
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.7969582329269076,
+      "learning_rate": 1.0718384338483141e-05,
+      "loss": 3.7543,
+      "step": 1232
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.148636874878519,
+      "learning_rate": 1.0705462190127011e-05,
+      "loss": 3.7497,
+      "step": 1233
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3035604956864364,
+      "learning_rate": 1.0692538857775685e-05,
+      "loss": 3.8701,
+      "step": 1234
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.008854288704277,
+      "learning_rate": 1.0679614363118718e-05,
+      "loss": 3.4629,
+      "step": 1235
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.8217933247978877,
+      "learning_rate": 1.066668872784762e-05,
+      "loss": 3.599,
+      "step": 1236
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.505699888668807,
+      "learning_rate": 1.0653761973655819e-05,
+      "loss": 3.61,
+      "step": 1237
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.630452722607848,
+      "learning_rate": 1.0640834122238606e-05,
+      "loss": 3.6772,
+      "step": 1238
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4696551759013063,
+      "learning_rate": 1.0627905195293135e-05,
+      "loss": 3.8177,
+      "step": 1239
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.600962380822384,
+      "learning_rate": 1.061497521451835e-05,
+      "loss": 3.7467,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5615429684192623,
+      "learning_rate": 1.0602044201614965e-05,
+      "loss": 3.7444,
+      "step": 1241
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6893702737397125,
+      "learning_rate": 1.0589112178285432e-05,
+      "loss": 3.7873,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.088874847636349,
+      "learning_rate": 1.0576179166233895e-05,
+      "loss": 3.7347,
+      "step": 1243
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6194124320453978,
+      "learning_rate": 1.056324518716616e-05,
+      "loss": 3.6616,
+      "step": 1244
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.709738290667363,
+      "learning_rate": 1.055031026278965e-05,
+      "loss": 3.5815,
+      "step": 1245
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.3126884197713675,
+      "learning_rate": 1.0537374414813384e-05,
+      "loss": 3.7129,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.126610206839559,
+      "learning_rate": 1.0524437664947918e-05,
+      "loss": 3.8666,
+      "step": 1247
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.228485912654699,
+      "learning_rate": 1.051150003490534e-05,
+      "loss": 3.8475,
+      "step": 1248
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.414002348692545,
+      "learning_rate": 1.0498561546399194e-05,
+      "loss": 3.7105,
+      "step": 1249
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.544218783563698,
+      "learning_rate": 1.0485622221144485e-05,
+      "loss": 3.6683,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6293603164028023,
+      "learning_rate": 1.0472682080857606e-05,
+      "loss": 3.7261,
+      "step": 1251
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.119603916787359,
+      "learning_rate": 1.0459741147256325e-05,
+      "loss": 3.645,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.309856421704084,
+      "learning_rate": 1.044679944205975e-05,
+      "loss": 3.7916,
+      "step": 1253
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3126208630112717,
+      "learning_rate": 1.043385698698826e-05,
+      "loss": 3.5428,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.435020152190706,
+      "learning_rate": 1.0420913803763522e-05,
+      "loss": 3.6746,
+      "step": 1255
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.442250663980389,
+      "learning_rate": 1.04079699141084e-05,
+      "loss": 3.8303,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6876733935381405,
+      "learning_rate": 1.0395025339746965e-05,
+      "loss": 3.7032,
+      "step": 1257
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.680657228628987,
+      "learning_rate": 1.0382080102404417e-05,
+      "loss": 3.6895,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4319284847026,
+      "learning_rate": 1.0369134223807082e-05,
+      "loss": 3.5902,
+      "step": 1259
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.146045404231594,
+      "learning_rate": 1.0356187725682359e-05,
+      "loss": 3.7994,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.279354700190127,
+      "learning_rate": 1.0343240629758683e-05,
+      "loss": 3.6323,
+      "step": 1261
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6423447713494674,
+      "learning_rate": 1.0330292957765502e-05,
+      "loss": 3.7522,
+      "step": 1262
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.9929601866704547,
+      "learning_rate": 1.0317344731433217e-05,
+      "loss": 3.6607,
+      "step": 1263
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.884660944168367,
+      "learning_rate": 1.0304395972493172e-05,
+      "loss": 3.9169,
+      "step": 1264
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4149693145698077,
+      "learning_rate": 1.0291446702677598e-05,
+      "loss": 3.6962,
+      "step": 1265
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.45271671752356,
+      "learning_rate": 1.0278496943719585e-05,
+      "loss": 3.7441,
+      "step": 1266
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.6133398318052428,
+      "learning_rate": 1.0265546717353041e-05,
+      "loss": 3.683,
+      "step": 1267
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.371173249364838,
+      "learning_rate": 1.0252596045312666e-05,
+      "loss": 3.6712,
+      "step": 1268
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.83144634459696,
+      "learning_rate": 1.02396449493339e-05,
+      "loss": 3.6621,
+      "step": 1269
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5042655204060793,
+      "learning_rate": 1.02266934511529e-05,
+      "loss": 3.6081,
+      "step": 1270
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.901785995386197,
+      "learning_rate": 1.0213741572506497e-05,
+      "loss": 3.5902,
+      "step": 1271
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.597912776915605,
+      "learning_rate": 1.0200789335132157e-05,
+      "loss": 3.6509,
+      "step": 1272
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.576944142026537,
+      "learning_rate": 1.0187836760767954e-05,
+      "loss": 3.6438,
+      "step": 1273
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.171711826080111,
+      "learning_rate": 1.0174883871152517e-05,
+      "loss": 3.6199,
+      "step": 1274
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.997960682784523,
+      "learning_rate": 1.0161930688025018e-05,
+      "loss": 3.597,
+      "step": 1275
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.118551729202294,
+      "learning_rate": 1.014897723312511e-05,
+      "loss": 3.7327,
+      "step": 1276
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.115281655837042,
+      "learning_rate": 1.013602352819291e-05,
+      "loss": 3.6384,
+      "step": 1277
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.8326653216963527,
+      "learning_rate": 1.0123069594968952e-05,
+      "loss": 3.6624,
+      "step": 1278
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.420306455859883,
+      "learning_rate": 1.0110115455194157e-05,
+      "loss": 3.6849,
+      "step": 1279
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.8078583089194438,
+      "learning_rate": 1.0097161130609774e-05,
+      "loss": 3.6237,
+      "step": 1280
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4296770703932693,
+      "learning_rate": 1.0084206642957393e-05,
+      "loss": 3.772,
+      "step": 1281
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.127642158329446,
+      "learning_rate": 1.0071252013978852e-05,
+      "loss": 3.571,
+      "step": 1282
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.7661312311309287,
+      "learning_rate": 1.0058297265416234e-05,
+      "loss": 3.6234,
+      "step": 1283
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.543027913063287,
+      "learning_rate": 1.0045342419011832e-05,
+      "loss": 3.5259,
+      "step": 1284
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.157795877165396,
+      "learning_rate": 1.003238749650809e-05,
+      "loss": 3.5826,
+      "step": 1285
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.9799792666825966,
+      "learning_rate": 1.0019432519647585e-05,
+      "loss": 3.6644,
+      "step": 1286
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.922820795521154,
+      "learning_rate": 1.0006477510172984e-05,
+      "loss": 3.7645,
+      "step": 1287
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2795127075695625,
+      "learning_rate": 9.993522489827016e-06,
+      "loss": 3.7082,
+      "step": 1288
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.160932501828583,
+      "learning_rate": 9.980567480352417e-06,
+      "loss": 3.7132,
+      "step": 1289
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.9551672926611743,
+      "learning_rate": 9.967612503491915e-06,
+      "loss": 3.6536,
+      "step": 1290
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.567688655247914,
+      "learning_rate": 9.954657580988171e-06,
+      "loss": 3.6158,
+      "step": 1291
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.504476394633079,
+      "learning_rate": 9.941702734583771e-06,
+      "loss": 3.6681,
+      "step": 1292
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.8738546664048155,
+      "learning_rate": 9.928747986021153e-06,
+      "loss": 3.9168,
+      "step": 1293
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.681866936469431,
+      "learning_rate": 9.91579335704261e-06,
+      "loss": 3.5911,
+      "step": 1294
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.8543107213799708,
+      "learning_rate": 9.90283886939023e-06,
+      "loss": 3.649,
+      "step": 1295
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.213819491800523,
+      "learning_rate": 9.88988454480585e-06,
+      "loss": 3.6681,
+      "step": 1296
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.672911779908602,
+      "learning_rate": 9.876930405031047e-06,
+      "loss": 3.4539,
+      "step": 1297
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.242243787102731,
+      "learning_rate": 9.86397647180709e-06,
+      "loss": 3.6487,
+      "step": 1298
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2004720405539255,
+      "learning_rate": 9.851022766874892e-06,
+      "loss": 3.8127,
+      "step": 1299
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.09162308157697,
+      "learning_rate": 9.838069311974986e-06,
+      "loss": 3.6898,
+      "step": 1300
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.548750561025053,
+      "learning_rate": 9.825116128847488e-06,
+      "loss": 3.6147,
+      "step": 1301
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.8328336344463603,
+      "learning_rate": 9.812163239232051e-06,
+      "loss": 3.5539,
+      "step": 1302
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2443184038824855,
+      "learning_rate": 9.799210664867844e-06,
+      "loss": 3.6883,
+      "step": 1303
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.878823223982976,
+      "learning_rate": 9.786258427493505e-06,
+      "loss": 3.823,
+      "step": 1304
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4116361274295186,
+      "learning_rate": 9.773306548847102e-06,
+      "loss": 3.578,
+      "step": 1305
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.24034110874559,
+      "learning_rate": 9.760355050666102e-06,
+      "loss": 3.727,
+      "step": 1306
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.634306245889952,
+      "learning_rate": 9.747403954687334e-06,
+      "loss": 3.7736,
+      "step": 1307
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.749776046989676,
+      "learning_rate": 9.734453282646962e-06,
+      "loss": 3.6961,
+      "step": 1308
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.932417095522838,
+      "learning_rate": 9.721503056280418e-06,
+      "loss": 3.6418,
+      "step": 1309
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.542879382334435,
+      "learning_rate": 9.708553297322407e-06,
+      "loss": 3.5859,
+      "step": 1310
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5258395671961127,
+      "learning_rate": 9.69560402750683e-06,
+      "loss": 3.7562,
+      "step": 1311
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.135406118551599,
+      "learning_rate": 9.682655268566783e-06,
+      "loss": 3.5696,
+      "step": 1312
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.8528419671684597,
+      "learning_rate": 9.669707042234502e-06,
+      "loss": 3.6807,
+      "step": 1313
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.201555676577082,
+      "learning_rate": 9.656759370241318e-06,
+      "loss": 3.5825,
+      "step": 1314
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.64562077172273,
+      "learning_rate": 9.643812274317644e-06,
+      "loss": 3.4712,
+      "step": 1315
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.8396792440839103,
+      "learning_rate": 9.630865776192918e-06,
+      "loss": 3.7105,
+      "step": 1316
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.4577250263526444,
+      "learning_rate": 9.617919897595586e-06,
+      "loss": 3.5812,
+      "step": 1317
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.39166849313604,
+      "learning_rate": 9.604974660253039e-06,
+      "loss": 3.5843,
+      "step": 1318
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1918890159235995,
+      "learning_rate": 9.592030085891602e-06,
+      "loss": 3.6049,
+      "step": 1319
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.664646495100284,
+      "learning_rate": 9.579086196236483e-06,
+      "loss": 3.6161,
+      "step": 1320
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.318782755233096,
+      "learning_rate": 9.56614301301174e-06,
+      "loss": 3.8907,
+      "step": 1321
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.1588859391321322,
+      "learning_rate": 9.553200557940254e-06,
+      "loss": 3.7873,
+      "step": 1322
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.967215300253095,
+      "learning_rate": 9.540258852743676e-06,
+      "loss": 3.5864,
+      "step": 1323
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.941485118974778,
+      "learning_rate": 9.527317919142398e-06,
+      "loss": 3.9642,
+      "step": 1324
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.592458924548226,
+      "learning_rate": 9.514377778855521e-06,
+      "loss": 3.3537,
+      "step": 1325
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.7732706931466202,
+      "learning_rate": 9.501438453600808e-06,
+      "loss": 3.6944,
+      "step": 1326
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.7412543307988915,
+      "learning_rate": 9.488499965094664e-06,
+      "loss": 3.6244,
+      "step": 1327
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.9239890999288694,
+      "learning_rate": 9.475562335052086e-06,
+      "loss": 3.5095,
+      "step": 1328
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5112467552110247,
+      "learning_rate": 9.462625585186621e-06,
+      "loss": 3.7438,
+      "step": 1329
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6543160157821206,
+      "learning_rate": 9.449689737210352e-06,
+      "loss": 3.6419,
+      "step": 1330
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.988862136866014,
+      "learning_rate": 9.436754812833843e-06,
+      "loss": 3.6131,
+      "step": 1331
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3299291715841157,
+      "learning_rate": 9.423820833766108e-06,
+      "loss": 3.7091,
+      "step": 1332
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.9403033777529255,
+      "learning_rate": 9.410887821714571e-06,
+      "loss": 3.6045,
+      "step": 1333
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.8157108095673604,
+      "learning_rate": 9.39795579838504e-06,
+      "loss": 3.5397,
+      "step": 1334
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2338840832210667,
+      "learning_rate": 9.385024785481653e-06,
+      "loss": 3.6673,
+      "step": 1335
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6445781586174437,
+      "learning_rate": 9.372094804706867e-06,
+      "loss": 3.5268,
+      "step": 1336
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.9066716946435434,
+      "learning_rate": 9.359165877761396e-06,
+      "loss": 3.6854,
+      "step": 1337
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2966925584112547,
+      "learning_rate": 9.346238026344186e-06,
+      "loss": 3.5206,
+      "step": 1338
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.299893845106976,
+      "learning_rate": 9.333311272152385e-06,
+      "loss": 3.6807,
+      "step": 1339
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.899885401635834,
+      "learning_rate": 9.320385636881283e-06,
+      "loss": 3.5119,
+      "step": 1340
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.026905163814134,
+      "learning_rate": 9.307461142224318e-06,
+      "loss": 3.442,
+      "step": 1341
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.880216888977279,
+      "learning_rate": 9.29453780987299e-06,
+      "loss": 3.5532,
+      "step": 1342
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.387280773854988,
+      "learning_rate": 9.281615661516866e-06,
+      "loss": 3.7189,
+      "step": 1343
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.6555244517963943,
+      "learning_rate": 9.268694718843503e-06,
+      "loss": 3.6117,
+      "step": 1344
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.277666315779683,
+      "learning_rate": 9.255775003538462e-06,
+      "loss": 3.4797,
+      "step": 1345
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.6156687964961898,
+      "learning_rate": 9.242856537285227e-06,
+      "loss": 3.6104,
+      "step": 1346
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5831485134788332,
+      "learning_rate": 9.229939341765188e-06,
+      "loss": 3.7844,
+      "step": 1347
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.6448623997034795,
+      "learning_rate": 9.217023438657606e-06,
+      "loss": 3.483,
+      "step": 1348
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.37271213698544,
+      "learning_rate": 9.204108849639565e-06,
+      "loss": 3.7646,
+      "step": 1349
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.945295172360582,
+      "learning_rate": 9.19119559638596e-06,
+      "loss": 3.5386,
+      "step": 1350
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.2435047963682893,
+      "learning_rate": 9.178283700569424e-06,
+      "loss": 3.5083,
+      "step": 1351
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.6441695000819863,
+      "learning_rate": 9.165373183860329e-06,
+      "loss": 3.6747,
+      "step": 1352
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.758895378834221,
+      "learning_rate": 9.152464067926717e-06,
+      "loss": 3.5133,
+      "step": 1353
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.8045119099664175,
+      "learning_rate": 9.139556374434288e-06,
+      "loss": 3.6009,
+      "step": 1354
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.862260842586843,
+      "learning_rate": 9.126650125046361e-06,
+      "loss": 3.4762,
+      "step": 1355
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.5145809371669103,
+      "learning_rate": 9.113745341423816e-06,
+      "loss": 3.6713,
+      "step": 1356
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.828220691263425,
+      "learning_rate": 9.100842045225084e-06,
+      "loss": 3.7191,
+      "step": 1357
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.774746236211917,
+      "learning_rate": 9.087940258106093e-06,
+      "loss": 3.4438,
+      "step": 1358
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.7795138083033657,
+      "learning_rate": 9.075040001720247e-06,
+      "loss": 3.4842,
+      "step": 1359
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.94093397333636,
+      "learning_rate": 9.062141297718372e-06,
+      "loss": 3.6428,
+      "step": 1360
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.471037661246279,
+      "learning_rate": 9.049244167748694e-06,
+      "loss": 3.5629,
+      "step": 1361
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.392239101547643,
+      "learning_rate": 9.036348633456791e-06,
+      "loss": 3.5099,
+      "step": 1362
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.437481937306805,
+      "learning_rate": 9.023454716485572e-06,
+      "loss": 3.6523,
+      "step": 1363
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.716482112384828,
+      "learning_rate": 9.010562438475225e-06,
+      "loss": 3.6393,
+      "step": 1364
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.20173417003461,
+      "learning_rate": 8.99767182106319e-06,
+      "loss": 3.4431,
+      "step": 1365
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.601247359299966,
+      "learning_rate": 8.984782885884119e-06,
+      "loss": 3.5896,
+      "step": 1366
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4135121264981185,
+      "learning_rate": 8.971895654569842e-06,
+      "loss": 3.7265,
+      "step": 1367
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.9001461297693223,
+      "learning_rate": 8.959010148749324e-06,
+      "loss": 3.5086,
+      "step": 1368
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.03454895841195,
+      "learning_rate": 8.94612639004864e-06,
+      "loss": 3.677,
+      "step": 1369
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.650697251071297,
+      "learning_rate": 8.933244400090937e-06,
+      "loss": 3.6443,
+      "step": 1370
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.46434005555104,
+      "learning_rate": 8.92036420049638e-06,
+      "loss": 3.5778,
+      "step": 1371
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.93551298155282,
+      "learning_rate": 8.907485812882137e-06,
+      "loss": 3.444,
+      "step": 1372
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.137595514523684,
+      "learning_rate": 8.89460925886234e-06,
+      "loss": 3.5835,
+      "step": 1373
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1221207733061735,
+      "learning_rate": 8.881734560048037e-06,
+      "loss": 3.4763,
+      "step": 1374
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6063693650856203,
+      "learning_rate": 8.868861738047158e-06,
+      "loss": 3.512,
+      "step": 1375
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8006663498045854,
+      "learning_rate": 8.855990814464497e-06,
+      "loss": 3.6161,
+      "step": 1376
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.689346473741649,
+      "learning_rate": 8.843121810901643e-06,
+      "loss": 3.4761,
+      "step": 1377
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.5443205545427614,
+      "learning_rate": 8.830254748956983e-06,
+      "loss": 3.5129,
+      "step": 1378
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.461742310621672,
+      "learning_rate": 8.817389650225631e-06,
+      "loss": 3.5642,
+      "step": 1379
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.0981378966007,
+      "learning_rate": 8.804526536299413e-06,
+      "loss": 3.5141,
+      "step": 1380
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8550298590996155,
+      "learning_rate": 8.79166542876682e-06,
+      "loss": 3.5965,
+      "step": 1381
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.045038429383355,
+      "learning_rate": 8.778806349212968e-06,
+      "loss": 3.6084,
+      "step": 1382
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.955366018201757,
+      "learning_rate": 8.765949319219595e-06,
+      "loss": 3.5752,
+      "step": 1383
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.457754380490334,
+      "learning_rate": 8.753094360364973e-06,
+      "loss": 3.7728,
+      "step": 1384
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.786851175476542,
+      "learning_rate": 8.740241494223911e-06,
+      "loss": 3.5675,
+      "step": 1385
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.660440730641834,
+      "learning_rate": 8.727390742367698e-06,
+      "loss": 3.6431,
+      "step": 1386
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.084737719287929,
+      "learning_rate": 8.71454212636408e-06,
+      "loss": 3.5924,
+      "step": 1387
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2622905116568877,
+      "learning_rate": 8.701695667777221e-06,
+      "loss": 3.7183,
+      "step": 1388
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.55476228024875,
+      "learning_rate": 8.688851388167658e-06,
+      "loss": 3.4861,
+      "step": 1389
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6473137503472546,
+      "learning_rate": 8.676009309092273e-06,
+      "loss": 3.6158,
+      "step": 1390
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7945829817862524,
+      "learning_rate": 8.663169452104248e-06,
+      "loss": 3.6302,
+      "step": 1391
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.640805329322732,
+      "learning_rate": 8.650331838753057e-06,
+      "loss": 3.7547,
+      "step": 1392
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8124210910003455,
+      "learning_rate": 8.637496490584385e-06,
+      "loss": 3.5456,
+      "step": 1393
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.9704591066995767,
+      "learning_rate": 8.624663429140128e-06,
+      "loss": 3.4268,
+      "step": 1394
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.049723146662677,
+      "learning_rate": 8.611832675958335e-06,
+      "loss": 3.5364,
+      "step": 1395
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.533547042548108,
+      "learning_rate": 8.599004252573191e-06,
+      "loss": 3.6708,
+      "step": 1396
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.58267549077645,
+      "learning_rate": 8.586178180514968e-06,
+      "loss": 3.5749,
+      "step": 1397
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.9116869644342795,
+      "learning_rate": 8.573354481309986e-06,
+      "loss": 3.5463,
+      "step": 1398
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.055182017276201,
+      "learning_rate": 8.560533176480588e-06,
+      "loss": 3.8375,
+      "step": 1399
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.627516942271804,
+      "learning_rate": 8.5477142875451e-06,
+      "loss": 3.5717,
+      "step": 1400
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.812255836731826,
+      "learning_rate": 8.534897836017784e-06,
+      "loss": 3.5476,
+      "step": 1401
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.078113719579199,
+      "learning_rate": 8.522083843408823e-06,
+      "loss": 3.6681,
+      "step": 1402
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5598802470842936,
+      "learning_rate": 8.50927233122427e-06,
+      "loss": 3.3416,
+      "step": 1403
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.648309157965994,
+      "learning_rate": 8.496463320966004e-06,
+      "loss": 3.5791,
+      "step": 1404
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7101771856350076,
+      "learning_rate": 8.48365683413172e-06,
+      "loss": 3.674,
+      "step": 1405
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.9313354547428485,
+      "learning_rate": 8.470852892214875e-06,
+      "loss": 3.6009,
+      "step": 1406
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.395437754771472,
+      "learning_rate": 8.458051516704644e-06,
+      "loss": 3.6776,
+      "step": 1407
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.269641226339698,
+      "learning_rate": 8.445252729085907e-06,
+      "loss": 3.5882,
+      "step": 1408
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.238326516283703,
+      "learning_rate": 8.432456550839196e-06,
+      "loss": 3.8776,
+      "step": 1409
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.850496269302347,
+      "learning_rate": 8.419663003440657e-06,
+      "loss": 3.5365,
+      "step": 1410
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.651812598740248,
+      "learning_rate": 8.406872108362034e-06,
+      "loss": 3.6866,
+      "step": 1411
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.310329473688571,
+      "learning_rate": 8.394083887070614e-06,
+      "loss": 3.5812,
+      "step": 1412
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1603001545864196,
+      "learning_rate": 8.38129836102919e-06,
+      "loss": 3.6525,
+      "step": 1413
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.259331885962401,
+      "learning_rate": 8.36851555169604e-06,
+      "loss": 3.3805,
+      "step": 1414
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.6620436971814896,
+      "learning_rate": 8.355735480524874e-06,
+      "loss": 3.5445,
+      "step": 1415
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.124009902890339,
+      "learning_rate": 8.342958168964816e-06,
+      "loss": 3.6279,
+      "step": 1416
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.515479094471101,
+      "learning_rate": 8.330183638460356e-06,
+      "loss": 3.3126,
+      "step": 1417
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.510007177496425,
+      "learning_rate": 8.317411910451313e-06,
+      "loss": 3.5948,
+      "step": 1418
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9870743550694567,
+      "learning_rate": 8.304643006372797e-06,
+      "loss": 3.5617,
+      "step": 1419
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.31871331046679,
+      "learning_rate": 8.291876947655197e-06,
+      "loss": 3.6182,
+      "step": 1420
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.657244073140862,
+      "learning_rate": 8.27911375572411e-06,
+      "loss": 3.4358,
+      "step": 1421
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.6885031118944838,
+      "learning_rate": 8.266353452000326e-06,
+      "loss": 3.4941,
+      "step": 1422
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4968539016285725,
+      "learning_rate": 8.253596057899788e-06,
+      "loss": 3.4569,
+      "step": 1423
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.241081464957698,
+      "learning_rate": 8.240841594833554e-06,
+      "loss": 3.5976,
+      "step": 1424
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.992209502684881,
+      "learning_rate": 8.228090084207773e-06,
+      "loss": 3.5895,
+      "step": 1425
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.6248505694886615,
+      "learning_rate": 8.215341547423624e-06,
+      "loss": 3.7637,
+      "step": 1426
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.548230202941458,
+      "learning_rate": 8.202596005877307e-06,
+      "loss": 3.6086,
+      "step": 1427
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8281555139288903,
+      "learning_rate": 8.189853480959982e-06,
+      "loss": 3.6207,
+      "step": 1428
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5927761035877155,
+      "learning_rate": 8.177113994057756e-06,
+      "loss": 3.5961,
+      "step": 1429
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.296788483429141,
+      "learning_rate": 8.16437756655164e-06,
+      "loss": 3.6734,
+      "step": 1430
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7427103365741834,
+      "learning_rate": 8.1516442198175e-06,
+      "loss": 3.6577,
+      "step": 1431
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7415928654401127,
+      "learning_rate": 8.138913975226044e-06,
+      "loss": 3.4335,
+      "step": 1432
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7793528215077097,
+      "learning_rate": 8.126186854142752e-06,
+      "loss": 3.4863,
+      "step": 1433
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2918021656616068,
+      "learning_rate": 8.113462877927893e-06,
+      "loss": 3.4836,
+      "step": 1434
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.658566957717168,
+      "learning_rate": 8.100742067936432e-06,
+      "loss": 3.3866,
+      "step": 1435
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.6849277344716795,
+      "learning_rate": 8.088024445518033e-06,
+      "loss": 3.6564,
+      "step": 1436
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.935693476386245,
+      "learning_rate": 8.075310032017e-06,
+      "loss": 3.6148,
+      "step": 1437
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.899449265414525,
+      "learning_rate": 8.062598848772261e-06,
+      "loss": 3.5728,
+      "step": 1438
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.121515481621947,
+      "learning_rate": 8.049890917117322e-06,
+      "loss": 3.6736,
+      "step": 1439
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.806565706148004,
+      "learning_rate": 8.037186258380226e-06,
+      "loss": 3.6357,
+      "step": 1440
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.195115835009729,
+      "learning_rate": 8.02448489388353e-06,
+      "loss": 3.5017,
+      "step": 1441
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.586544394139627,
+      "learning_rate": 8.01178684494425e-06,
+      "loss": 3.685,
+      "step": 1442
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.966560123803234,
+      "learning_rate": 7.999092132873851e-06,
+      "loss": 3.5741,
+      "step": 1443
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.172759572645984,
+      "learning_rate": 7.986400778978192e-06,
+      "loss": 3.6334,
+      "step": 1444
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.244342787839209,
+      "learning_rate": 7.9737128045575e-06,
+      "loss": 3.8758,
+      "step": 1445
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.992017579746834,
+      "learning_rate": 7.96102823090632e-06,
+      "loss": 3.6978,
+      "step": 1446
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.134386091152547,
+      "learning_rate": 7.948347079313494e-06,
+      "loss": 3.6387,
+      "step": 1447
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.282185418813217,
+      "learning_rate": 7.935669371062132e-06,
+      "loss": 3.7487,
+      "step": 1448
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.552626650054134,
+      "learning_rate": 7.922995127429547e-06,
+      "loss": 3.7081,
+      "step": 1449
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.447248945720088,
+      "learning_rate": 7.91032436968725e-06,
+      "loss": 3.3661,
+      "step": 1450
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.113127895541361,
+      "learning_rate": 7.897657119100896e-06,
+      "loss": 3.349,
+      "step": 1451
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3860791856708996,
+      "learning_rate": 7.88499339693025e-06,
+      "loss": 3.7081,
+      "step": 1452
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.2905573209490475,
+      "learning_rate": 7.872333224429166e-06,
+      "loss": 3.6622,
+      "step": 1453
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.364258205258447,
+      "learning_rate": 7.859676622845535e-06,
+      "loss": 3.6177,
+      "step": 1454
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.9009187557616913,
+      "learning_rate": 7.847023613421251e-06,
+      "loss": 3.6914,
+      "step": 1455
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5785002847394134,
+      "learning_rate": 7.834374217392188e-06,
+      "loss": 3.544,
+      "step": 1456
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5018286056581798,
+      "learning_rate": 7.82172845598814e-06,
+      "loss": 3.5752,
+      "step": 1457
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.181096789262609,
+      "learning_rate": 7.80908635043282e-06,
+      "loss": 3.495,
+      "step": 1458
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.9920855934196773,
+      "learning_rate": 7.796447921943793e-06,
+      "loss": 3.529,
+      "step": 1459
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.293480957964132,
+      "learning_rate": 7.78381319173246e-06,
+      "loss": 3.4376,
+      "step": 1460
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.9040102950110485,
+      "learning_rate": 7.771182181004005e-06,
+      "loss": 3.5162,
+      "step": 1461
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.635380000175011,
+      "learning_rate": 7.758554910957378e-06,
+      "loss": 3.8755,
+      "step": 1462
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.32258353400662,
+      "learning_rate": 7.745931402785252e-06,
+      "loss": 3.5296,
+      "step": 1463
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.6660451711288147,
+      "learning_rate": 7.733311677673979e-06,
+      "loss": 3.8214,
+      "step": 1464
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.6681150611869375,
+      "learning_rate": 7.720695756803569e-06,
+      "loss": 3.52,
+      "step": 1465
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.419776479875391,
+      "learning_rate": 7.708083661347637e-06,
+      "loss": 3.3692,
+      "step": 1466
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8962339569352924,
+      "learning_rate": 7.695475412473393e-06,
+      "loss": 3.4717,
+      "step": 1467
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8587481606461775,
+      "learning_rate": 7.682871031341579e-06,
+      "loss": 3.7205,
+      "step": 1468
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0586472108646032,
+      "learning_rate": 7.670270539106452e-06,
+      "loss": 3.4851,
+      "step": 1469
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.455954380269373,
+      "learning_rate": 7.657673956915735e-06,
+      "loss": 3.6466,
+      "step": 1470
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.047129327789186,
+      "learning_rate": 7.645081305910596e-06,
+      "loss": 3.267,
+      "step": 1471
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8169951162740348,
+      "learning_rate": 7.632492607225604e-06,
+      "loss": 3.6871,
+      "step": 1472
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8305487121163644,
+      "learning_rate": 7.619907881988692e-06,
+      "loss": 3.5829,
+      "step": 1473
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2735811073421788,
+      "learning_rate": 7.607327151321127e-06,
+      "loss": 3.6891,
+      "step": 1474
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.9136084760331418,
+      "learning_rate": 7.594750436337467e-06,
+      "loss": 3.4515,
+      "step": 1475
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.828639513589231,
+      "learning_rate": 7.582177758145532e-06,
+      "loss": 3.5019,
+      "step": 1476
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.9437593119820415,
+      "learning_rate": 7.569609137846376e-06,
+      "loss": 3.7731,
+      "step": 1477
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.164958834406506,
+      "learning_rate": 7.557044596534234e-06,
+      "loss": 3.7246,
+      "step": 1478
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.7055840109179368,
+      "learning_rate": 7.544484155296492e-06,
+      "loss": 3.6791,
+      "step": 1479
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.7527652506423284,
+      "learning_rate": 7.531927835213657e-06,
+      "loss": 3.6093,
+      "step": 1480
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.014478944178257,
+      "learning_rate": 7.519375657359331e-06,
+      "loss": 3.5312,
+      "step": 1481
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.246264132846788,
+      "learning_rate": 7.506827642800146e-06,
+      "loss": 3.7168,
+      "step": 1482
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.241663502305347,
+      "learning_rate": 7.49428381259576e-06,
+      "loss": 3.4936,
+      "step": 1483
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.985732421048951,
+      "learning_rate": 7.4817441877988005e-06,
+      "loss": 3.4383,
+      "step": 1484
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.520163919263216,
+      "learning_rate": 7.469208789454838e-06,
+      "loss": 3.5112,
+      "step": 1485
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.100364501529123,
+      "learning_rate": 7.456677638602355e-06,
+      "loss": 3.3978,
+      "step": 1486
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.906400338438948,
+      "learning_rate": 7.444150756272704e-06,
+      "loss": 3.5787,
+      "step": 1487
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.142247742585712,
+      "learning_rate": 7.431628163490067e-06,
+      "loss": 3.5451,
+      "step": 1488
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8484953594054607,
+      "learning_rate": 7.419109881271434e-06,
+      "loss": 3.5031,
+      "step": 1489
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5376977157395064,
+      "learning_rate": 7.40659593062655e-06,
+      "loss": 3.5643,
+      "step": 1490
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.684148878346785,
+      "learning_rate": 7.394086332557907e-06,
+      "loss": 3.7448,
+      "step": 1491
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.7070400002530035,
+      "learning_rate": 7.38158110806068e-06,
+      "loss": 3.5748,
+      "step": 1492
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0700666156104046,
+      "learning_rate": 7.3690802781227056e-06,
+      "loss": 3.6341,
+      "step": 1493
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8456034836325275,
+      "learning_rate": 7.356583863724442e-06,
+      "loss": 3.5115,
+      "step": 1494
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.789836055161562,
+      "learning_rate": 7.344091885838949e-06,
+      "loss": 3.5078,
+      "step": 1495
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.782648371746039,
+      "learning_rate": 7.331604365431826e-06,
+      "loss": 3.4992,
+      "step": 1496
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.399890932099315,
+      "learning_rate": 7.319121323461198e-06,
+      "loss": 3.5788,
+      "step": 1497
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.519272517962314,
+      "learning_rate": 7.3066427808776754e-06,
+      "loss": 3.5133,
+      "step": 1498
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.607027156933682,
+      "learning_rate": 7.294168758624307e-06,
+      "loss": 3.5752,
+      "step": 1499
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.6015766335135573,
+      "learning_rate": 7.2816992776365714e-06,
+      "loss": 3.8539,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.7429477576654637,
+      "learning_rate": 7.269234358842314e-06,
+      "loss": 3.6593,
+      "step": 1501
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.417196523965054,
+      "learning_rate": 7.256774023161728e-06,
+      "loss": 3.8102,
+      "step": 1502
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.902532703962836,
+      "learning_rate": 7.244318291507308e-06,
+      "loss": 3.4636,
+      "step": 1503
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1780987980769613,
+      "learning_rate": 7.231867184783826e-06,
+      "loss": 3.5413,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.195603868320716,
+      "learning_rate": 7.219420723888301e-06,
+      "loss": 3.6185,
+      "step": 1505
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5099189748150685,
+      "learning_rate": 7.2069789297099355e-06,
+      "loss": 3.6203,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.7368146967887084,
+      "learning_rate": 7.194541823130119e-06,
+      "loss": 3.4034,
+      "step": 1507
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2477208654607175,
+      "learning_rate": 7.182109425022357e-06,
+      "loss": 3.5462,
+      "step": 1508
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.954509777924059,
+      "learning_rate": 7.169681756252265e-06,
+      "loss": 3.3927,
+      "step": 1509
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.44799095991217,
+      "learning_rate": 7.157258837677514e-06,
+      "loss": 3.3569,
+      "step": 1510
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.7064912121782094,
+      "learning_rate": 7.144840690147812e-06,
+      "loss": 3.5461,
+      "step": 1511
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8854217452893565,
+      "learning_rate": 7.132427334504846e-06,
+      "loss": 3.8537,
+      "step": 1512
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9978972776658903,
+      "learning_rate": 7.120018791582266e-06,
+      "loss": 3.6769,
+      "step": 1513
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.105775013442488,
+      "learning_rate": 7.107615082205654e-06,
+      "loss": 3.6836,
+      "step": 1514
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.594521170717367,
+      "learning_rate": 7.095216227192467e-06,
+      "loss": 3.5434,
+      "step": 1515
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.083671392728415,
+      "learning_rate": 7.082822247352024e-06,
+      "loss": 3.7165,
+      "step": 1516
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.282797887972504,
+      "learning_rate": 7.070433163485451e-06,
+      "loss": 3.6366,
+      "step": 1517
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.372384984933006,
+      "learning_rate": 7.0580489963856646e-06,
+      "loss": 3.5777,
+      "step": 1518
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.160592298870379,
+      "learning_rate": 7.045669766837333e-06,
+      "loss": 3.3398,
+      "step": 1519
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.020626122432115,
+      "learning_rate": 7.033295495616834e-06,
+      "loss": 3.3516,
+      "step": 1520
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.17132157189251,
+      "learning_rate": 7.020926203492218e-06,
+      "loss": 3.4481,
+      "step": 1521
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.648358936658591,
+      "learning_rate": 7.008561911223186e-06,
+      "loss": 3.6298,
+      "step": 1522
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.03872622210649,
+      "learning_rate": 6.9962026395610416e-06,
+      "loss": 3.6469,
+      "step": 1523
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.796123861492065,
+      "learning_rate": 6.983848409248672e-06,
+      "loss": 3.625,
+      "step": 1524
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.01401433098644,
+      "learning_rate": 6.971499241020495e-06,
+      "loss": 3.4491,
+      "step": 1525
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.600759214686406,
+      "learning_rate": 6.959155155602433e-06,
+      "loss": 3.6289,
+      "step": 1526
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.2385099838211495,
+      "learning_rate": 6.946816173711878e-06,
+      "loss": 3.5884,
+      "step": 1527
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.017534211523324,
+      "learning_rate": 6.934482316057663e-06,
+      "loss": 3.6964,
+      "step": 1528
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.284559410654498,
+      "learning_rate": 6.922153603340016e-06,
+      "loss": 3.509,
+      "step": 1529
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.624699831426992,
+      "learning_rate": 6.909830056250527e-06,
+      "loss": 3.315,
+      "step": 1530
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.9957231893286465,
+      "learning_rate": 6.897511695472124e-06,
+      "loss": 3.3653,
+      "step": 1531
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.7559802871043244,
+      "learning_rate": 6.885198541679016e-06,
+      "loss": 3.5622,
+      "step": 1532
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.515750722822639,
+      "learning_rate": 6.872890615536694e-06,
+      "loss": 3.6934,
+      "step": 1533
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.361687296630023,
+      "learning_rate": 6.860587937701862e-06,
+      "loss": 3.7604,
+      "step": 1534
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.665985775304246,
+      "learning_rate": 6.848290528822417e-06,
+      "loss": 3.6467,
+      "step": 1535
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5742810689464894,
+      "learning_rate": 6.835998409537412e-06,
+      "loss": 3.6129,
+      "step": 1536
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1398746962171953,
+      "learning_rate": 6.823711600477025e-06,
+      "loss": 3.5571,
+      "step": 1537
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8744598615958283,
+      "learning_rate": 6.811430122262529e-06,
+      "loss": 3.4432,
+      "step": 1538
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.680958440614546,
+      "learning_rate": 6.799153995506234e-06,
+      "loss": 3.4893,
+      "step": 1539
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.251229349989716,
+      "learning_rate": 6.786883240811479e-06,
+      "loss": 3.421,
+      "step": 1540
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.363805693156403,
+      "learning_rate": 6.774617878772582e-06,
+      "loss": 3.7277,
+      "step": 1541
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8060970867616395,
+      "learning_rate": 6.76235792997482e-06,
+      "loss": 3.5927,
+      "step": 1542
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.121071467859133,
+      "learning_rate": 6.750103414994374e-06,
+      "loss": 3.6922,
+      "step": 1543
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5635624234510095,
+      "learning_rate": 6.737854354398308e-06,
+      "loss": 3.3097,
+      "step": 1544
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3704037106511224,
+      "learning_rate": 6.725610768744535e-06,
+      "loss": 3.5523,
+      "step": 1545
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.6183604752793395,
+      "learning_rate": 6.713372678581773e-06,
+      "loss": 3.6057,
+      "step": 1546
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8491054462248853,
+      "learning_rate": 6.7011401044495304e-06,
+      "loss": 3.4293,
+      "step": 1547
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4848739882554973,
+      "learning_rate": 6.68891306687804e-06,
+      "loss": 3.6692,
+      "step": 1548
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.7199367426632524,
+      "learning_rate": 6.676691586388255e-06,
+      "loss": 3.449,
+      "step": 1549
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.580681658335725,
+      "learning_rate": 6.664475683491797e-06,
+      "loss": 3.5791,
+      "step": 1550
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5016383532515483,
+      "learning_rate": 6.652265378690923e-06,
+      "loss": 3.6628,
+      "step": 1551
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2121234727237225,
+      "learning_rate": 6.6400606924785095e-06,
+      "loss": 3.3646,
+      "step": 1552
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9771811074622305,
+      "learning_rate": 6.627861645337984e-06,
+      "loss": 3.4899,
+      "step": 1553
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.424344005443578,
+      "learning_rate": 6.615668257743322e-06,
+      "loss": 3.6783,
+      "step": 1554
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9008718425597486,
+      "learning_rate": 6.603480550158995e-06,
+      "loss": 3.4927,
+      "step": 1555
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5524895975215767,
+      "learning_rate": 6.591298543039949e-06,
+      "loss": 3.6305,
+      "step": 1556
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.199136364840206,
+      "learning_rate": 6.579122256831551e-06,
+      "loss": 3.469,
+      "step": 1557
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.136039927950006,
+      "learning_rate": 6.566951711969581e-06,
+      "loss": 3.8279,
+      "step": 1558
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5719592951558004,
+      "learning_rate": 6.554786928880165e-06,
+      "loss": 3.6708,
+      "step": 1559
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.265159798599461,
+      "learning_rate": 6.542627927979772e-06,
+      "loss": 3.3514,
+      "step": 1560
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.705950892355711,
+      "learning_rate": 6.530474729675167e-06,
+      "loss": 3.6671,
+      "step": 1561
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.27802567413278,
+      "learning_rate": 6.518327354363374e-06,
+      "loss": 3.4778,
+      "step": 1562
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4155897157303494,
+      "learning_rate": 6.506185822431639e-06,
+      "loss": 3.6436,
+      "step": 1563
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4645941258945476,
+      "learning_rate": 6.494050154257408e-06,
+      "loss": 3.5203,
+      "step": 1564
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5880428732526735,
+      "learning_rate": 6.481920370208274e-06,
+      "loss": 3.292,
+      "step": 1565
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.453501383628343,
+      "learning_rate": 6.469796490641974e-06,
+      "loss": 3.8199,
+      "step": 1566
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.7078788389918738,
+      "learning_rate": 6.4576785359063225e-06,
+      "loss": 3.629,
+      "step": 1567
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.268044566714057,
+      "learning_rate": 6.445566526339187e-06,
+      "loss": 3.4953,
+      "step": 1568
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1001497410904943,
+      "learning_rate": 6.4334604822684645e-06,
+      "loss": 3.6877,
+      "step": 1569
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.245953209024202,
+      "learning_rate": 6.421360424012039e-06,
+      "loss": 3.4786,
+      "step": 1570
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2821730226566808,
+      "learning_rate": 6.409266371877751e-06,
+      "loss": 3.4555,
+      "step": 1571
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.182794902814907,
+      "learning_rate": 6.397178346163348e-06,
+      "loss": 3.3958,
+      "step": 1572
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.766092498437075,
+      "learning_rate": 6.38509636715648e-06,
+      "loss": 3.4831,
+      "step": 1573
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3471215501567446,
+      "learning_rate": 6.373020455134633e-06,
+      "loss": 3.3831,
+      "step": 1574
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.9241836485526003,
+      "learning_rate": 6.360950630365126e-06,
+      "loss": 3.6346,
+      "step": 1575
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3124279763521547,
+      "learning_rate": 6.3488869131050505e-06,
+      "loss": 3.3606,
+      "step": 1576
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8207348660033094,
+      "learning_rate": 6.33682932360125e-06,
+      "loss": 3.4873,
+      "step": 1577
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.313411276985893,
+      "learning_rate": 6.324777882090287e-06,
+      "loss": 3.547,
+      "step": 1578
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8492618965358716,
+      "learning_rate": 6.3127326087983974e-06,
+      "loss": 3.4882,
+      "step": 1579
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.7440406276540643,
+      "learning_rate": 6.300693523941481e-06,
+      "loss": 3.5546,
+      "step": 1580
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.256194385654137,
+      "learning_rate": 6.2886606477250345e-06,
+      "loss": 3.542,
+      "step": 1581
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.100847067766242,
+      "learning_rate": 6.276634000344144e-06,
+      "loss": 3.4619,
+      "step": 1582
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.863838035866122,
+      "learning_rate": 6.264613601983435e-06,
+      "loss": 3.5449,
+      "step": 1583
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4931068818315585,
+      "learning_rate": 6.2525994728170495e-06,
+      "loss": 3.5324,
+      "step": 1584
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3210210896064334,
+      "learning_rate": 6.2405916330086106e-06,
+      "loss": 3.4511,
+      "step": 1585
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.435646053815424,
+      "learning_rate": 6.2285901027111806e-06,
+      "loss": 3.2911,
+      "step": 1586
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.954806179069221,
+      "learning_rate": 6.216594902067233e-06,
+      "loss": 3.5649,
+      "step": 1587
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.220561982446002,
+      "learning_rate": 6.204606051208617e-06,
+      "loss": 3.4769,
+      "step": 1588
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5975537055749225,
+      "learning_rate": 6.192623570256535e-06,
+      "loss": 3.5477,
+      "step": 1589
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.8078229439619187,
+      "learning_rate": 6.180647479321484e-06,
+      "loss": 3.534,
+      "step": 1590
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5466201216437394,
+      "learning_rate": 6.168677798503246e-06,
+      "loss": 3.5194,
+      "step": 1591
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6779273264121106,
+      "learning_rate": 6.156714547890838e-06,
+      "loss": 3.4331,
+      "step": 1592
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7255338440199077,
+      "learning_rate": 6.14475774756249e-06,
+      "loss": 3.5986,
+      "step": 1593
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.626610821101966,
+      "learning_rate": 6.13280741758561e-06,
+      "loss": 3.497,
+      "step": 1594
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.147161579623455,
+      "learning_rate": 6.120863578016736e-06,
+      "loss": 3.7874,
+      "step": 1595
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6261110954731866,
+      "learning_rate": 6.108926248901521e-06,
+      "loss": 3.4446,
+      "step": 1596
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.4428710403247615,
+      "learning_rate": 6.0969954502746916e-06,
+      "loss": 3.4673,
+      "step": 1597
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5732663274462007,
+      "learning_rate": 6.0850712021600044e-06,
+      "loss": 3.4875,
+      "step": 1598
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.507148824224458,
+      "learning_rate": 6.073153524570236e-06,
+      "loss": 3.5483,
+      "step": 1599
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5971890810464346,
+      "learning_rate": 6.061242437507131e-06,
+      "loss": 3.3854,
+      "step": 1600
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.158234085592182,
+      "learning_rate": 6.049337960961362e-06,
+      "loss": 3.3238,
+      "step": 1601
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.8319143470427215,
+      "learning_rate": 6.037440114912521e-06,
+      "loss": 3.5227,
+      "step": 1602
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3015817963676706,
+      "learning_rate": 6.0255489193290675e-06,
+      "loss": 3.6485,
+      "step": 1603
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6210568854004803,
+      "learning_rate": 6.013664394168297e-06,
+      "loss": 3.5156,
+      "step": 1604
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5389233431815312,
+      "learning_rate": 6.00178655937631e-06,
+      "loss": 3.5009,
+      "step": 1605
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6576189067512854,
+      "learning_rate": 5.989915434887985e-06,
+      "loss": 3.3947,
+      "step": 1606
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.339883862987916,
+      "learning_rate": 5.9780510406269245e-06,
+      "loss": 3.4721,
+      "step": 1607
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.52954677971464,
+      "learning_rate": 5.966193396505452e-06,
+      "loss": 3.3105,
+      "step": 1608
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.797901067597484,
+      "learning_rate": 5.954342522424553e-06,
+      "loss": 3.5331,
+      "step": 1609
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.025259163522358,
+      "learning_rate": 5.942498438273849e-06,
+      "loss": 3.4357,
+      "step": 1610
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.883933535536824,
+      "learning_rate": 5.930661163931572e-06,
+      "loss": 3.5552,
+      "step": 1611
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.9176809770017296,
+      "learning_rate": 5.918830719264514e-06,
+      "loss": 3.6888,
+      "step": 1612
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.158931640923236,
+      "learning_rate": 5.9070071241280235e-06,
+      "loss": 3.745,
+      "step": 1613
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.087354007066473,
+      "learning_rate": 5.895190398365935e-06,
+      "loss": 3.6546,
+      "step": 1614
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.238168845466041,
+      "learning_rate": 5.8833805618105635e-06,
+      "loss": 3.6311,
+      "step": 1615
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.921391574264145,
+      "learning_rate": 5.871577634282655e-06,
+      "loss": 3.3619,
+      "step": 1616
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.222025825827338,
+      "learning_rate": 5.8597816355913685e-06,
+      "loss": 3.5381,
+      "step": 1617
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.526603701604458,
+      "learning_rate": 5.84799258553423e-06,
+      "loss": 3.2635,
+      "step": 1618
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.328920440404449,
+      "learning_rate": 5.836210503897099e-06,
+      "loss": 3.3941,
+      "step": 1619
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.694703294976491,
+      "learning_rate": 5.82443541045415e-06,
+      "loss": 3.6353,
+      "step": 1620
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.046321767564445,
+      "learning_rate": 5.812667324967813e-06,
+      "loss": 3.56,
+      "step": 1621
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.7512923075952695,
+      "learning_rate": 5.800906267188773e-06,
+      "loss": 3.5772,
+      "step": 1622
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.919553279950237,
+      "learning_rate": 5.789152256855917e-06,
+      "loss": 3.4645,
+      "step": 1623
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9665901288915806,
+      "learning_rate": 5.777405313696294e-06,
+      "loss": 3.4751,
+      "step": 1624
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.9438397686222832,
+      "learning_rate": 5.765665457425102e-06,
+      "loss": 3.5324,
+      "step": 1625
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.720691178895891,
+      "learning_rate": 5.753932707745635e-06,
+      "loss": 3.3672,
+      "step": 1626
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4053190270389604,
+      "learning_rate": 5.742207084349274e-06,
+      "loss": 3.5297,
+      "step": 1627
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.8384090142645753,
+      "learning_rate": 5.73048860691543e-06,
+      "loss": 3.6019,
+      "step": 1628
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.155590506087914,
+      "learning_rate": 5.718777295111524e-06,
+      "loss": 3.541,
+      "step": 1629
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.935875106073164,
+      "learning_rate": 5.707073168592943e-06,
+      "loss": 3.4477,
+      "step": 1630
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.0601087611821365,
+      "learning_rate": 5.695376247003025e-06,
+      "loss": 3.5035,
+      "step": 1631
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.005956841096344,
+      "learning_rate": 5.683686549973018e-06,
+      "loss": 3.6031,
+      "step": 1632
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.018384680043002,
+      "learning_rate": 5.672004097122033e-06,
+      "loss": 3.4305,
+      "step": 1633
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.089213051550755,
+      "learning_rate": 5.6603289080570274e-06,
+      "loss": 3.4664,
+      "step": 1634
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.0324336465554556,
+      "learning_rate": 5.648661002372769e-06,
+      "loss": 3.5037,
+      "step": 1635
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.8702545614500976,
+      "learning_rate": 5.637000399651804e-06,
+      "loss": 3.3685,
+      "step": 1636
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.044818800764605,
+      "learning_rate": 5.625347119464422e-06,
+      "loss": 3.4022,
+      "step": 1637
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4606581789411743,
+      "learning_rate": 5.613701181368618e-06,
+      "loss": 3.7145,
+      "step": 1638
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.083316905669718,
+      "learning_rate": 5.602062604910064e-06,
+      "loss": 3.6496,
+      "step": 1639
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.959151798357818,
+      "learning_rate": 5.590431409622081e-06,
+      "loss": 3.8787,
+      "step": 1640
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.588318519706414,
+      "learning_rate": 5.5788076150256075e-06,
+      "loss": 3.6376,
+      "step": 1641
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.939429818845424,
+      "learning_rate": 5.567191240629151e-06,
+      "loss": 3.3604,
+      "step": 1642
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.024057438227261,
+      "learning_rate": 5.555582305928766e-06,
+      "loss": 3.3678,
+      "step": 1643
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.702973125759723,
+      "learning_rate": 5.5439808304080225e-06,
+      "loss": 3.6029,
+      "step": 1644
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.553114471083393,
+      "learning_rate": 5.5323868335379775e-06,
+      "loss": 3.4881,
+      "step": 1645
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.531936177964785,
+      "learning_rate": 5.520800334777132e-06,
+      "loss": 3.7625,
+      "step": 1646
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7173152360158865,
+      "learning_rate": 5.509221353571404e-06,
+      "loss": 3.2961,
+      "step": 1647
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9439784637905064,
+      "learning_rate": 5.497649909354084e-06,
+      "loss": 3.3936,
+      "step": 1648
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.78139801390855,
+      "learning_rate": 5.486086021545829e-06,
+      "loss": 3.5477,
+      "step": 1649
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8954021579269464,
+      "learning_rate": 5.4745297095546125e-06,
+      "loss": 3.3632,
+      "step": 1650
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9087623151938984,
+      "learning_rate": 5.4629809927756794e-06,
+      "loss": 3.4839,
+      "step": 1651
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.635931087210787,
+      "learning_rate": 5.451439890591539e-06,
+      "loss": 3.4566,
+      "step": 1652
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.58703249038166,
+      "learning_rate": 5.439906422371914e-06,
+      "loss": 3.6018,
+      "step": 1653
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3918410665471743,
+      "learning_rate": 5.42838060747372e-06,
+      "loss": 3.3492,
+      "step": 1654
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.307113537697726,
+      "learning_rate": 5.416862465241033e-06,
+      "loss": 3.4002,
+      "step": 1655
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4614695792966033,
+      "learning_rate": 5.405352015005039e-06,
+      "loss": 3.5605,
+      "step": 1656
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5235364563297984,
+      "learning_rate": 5.3938492760840176e-06,
+      "loss": 3.4043,
+      "step": 1657
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6270246557742647,
+      "learning_rate": 5.382354267783316e-06,
+      "loss": 3.6153,
+      "step": 1658
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7458961603016636,
+      "learning_rate": 5.370867009395294e-06,
+      "loss": 3.3846,
+      "step": 1659
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9253003162704854,
+      "learning_rate": 5.359387520199317e-06,
+      "loss": 3.5766,
+      "step": 1660
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.66451016542722,
+      "learning_rate": 5.3479158194617e-06,
+      "loss": 3.3781,
+      "step": 1661
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.913591319050972,
+      "learning_rate": 5.336451926435688e-06,
+      "loss": 3.5858,
+      "step": 1662
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6297933301525984,
+      "learning_rate": 5.32499586036143e-06,
+      "loss": 3.3996,
+      "step": 1663
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.409251034623019,
+      "learning_rate": 5.313547640465937e-06,
+      "loss": 3.6274,
+      "step": 1664
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.5973973468416776,
+      "learning_rate": 5.302107285963045e-06,
+      "loss": 3.3889,
+      "step": 1665
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1962049167523157,
+      "learning_rate": 5.2906748160533895e-06,
+      "loss": 3.4495,
+      "step": 1666
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.0059543728084375,
+      "learning_rate": 5.279250249924384e-06,
+      "loss": 3.5386,
+      "step": 1667
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.7194107001174936,
+      "learning_rate": 5.26783360675016e-06,
+      "loss": 3.5265,
+      "step": 1668
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.7921834437433484,
+      "learning_rate": 5.2564249056915704e-06,
+      "loss": 3.5549,
+      "step": 1669
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.403776989679724,
+      "learning_rate": 5.245024165896126e-06,
+      "loss": 3.3479,
+      "step": 1670
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.8704830411141518,
+      "learning_rate": 5.2336314064979766e-06,
+      "loss": 3.3499,
+      "step": 1671
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.7989300337853664,
+      "learning_rate": 5.222246646617886e-06,
+      "loss": 3.5178,
+      "step": 1672
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.734045664578303,
+      "learning_rate": 5.210869905363178e-06,
+      "loss": 3.464,
+      "step": 1673
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.832535478682915,
+      "learning_rate": 5.199501201827741e-06,
+      "loss": 3.7964,
+      "step": 1674
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.7729187124032797,
+      "learning_rate": 5.18814055509195e-06,
+      "loss": 3.4932,
+      "step": 1675
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.7973719375176183,
+      "learning_rate": 5.1767879842226745e-06,
+      "loss": 3.4163,
+      "step": 1676
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.2259816427355,
+      "learning_rate": 5.165443508273218e-06,
+      "loss": 3.3001,
+      "step": 1677
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.251724589557688,
+      "learning_rate": 5.154107146283311e-06,
+      "loss": 3.1482,
+      "step": 1678
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6809587313873013,
+      "learning_rate": 5.1427789172790565e-06,
+      "loss": 3.6134,
+      "step": 1679
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.951286411251701,
+      "learning_rate": 5.131458840272905e-06,
+      "loss": 3.4604,
+      "step": 1680
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.002690650992529,
+      "learning_rate": 5.120146934263638e-06,
+      "loss": 3.4269,
+      "step": 1681
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.587007733749125,
+      "learning_rate": 5.10884321823631e-06,
+      "loss": 3.4809,
+      "step": 1682
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.033586550699142,
+      "learning_rate": 5.097547711162243e-06,
+      "loss": 3.495,
+      "step": 1683
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.358904503933148,
+      "learning_rate": 5.086260431998967e-06,
+      "loss": 3.4295,
+      "step": 1684
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.092011940657899,
+      "learning_rate": 5.074981399690219e-06,
+      "loss": 3.4061,
+      "step": 1685
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.509859863620395,
+      "learning_rate": 5.0637106331658815e-06,
+      "loss": 3.3268,
+      "step": 1686
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4988932754363944,
+      "learning_rate": 5.0524481513419675e-06,
+      "loss": 3.5498,
+      "step": 1687
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5371819602716257,
+      "learning_rate": 5.041193973120595e-06,
+      "loss": 3.6162,
+      "step": 1688
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7999904525150257,
+      "learning_rate": 5.02994811738993e-06,
+      "loss": 3.4791,
+      "step": 1689
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.34916869882989,
+      "learning_rate": 5.018710603024187e-06,
+      "loss": 3.4989,
+      "step": 1690
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.3696963886037965,
+      "learning_rate": 5.007481448883567e-06,
+      "loss": 3.458,
+      "step": 1691
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.2633588151069315,
+      "learning_rate": 4.99626067381425e-06,
+      "loss": 3.4841,
+      "step": 1692
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6953948473727296,
+      "learning_rate": 4.985048296648346e-06,
+      "loss": 3.3859,
+      "step": 1693
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7639597092063384,
+      "learning_rate": 4.973844336203879e-06,
+      "loss": 3.5064,
+      "step": 1694
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6696256156495695,
+      "learning_rate": 4.9626488112847384e-06,
+      "loss": 3.4375,
+      "step": 1695
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.165077667236303,
+      "learning_rate": 4.951461740680655e-06,
+      "loss": 3.4031,
+      "step": 1696
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6209308184846605,
+      "learning_rate": 4.9402831431671834e-06,
+      "loss": 3.4249,
+      "step": 1697
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4591739980588274,
+      "learning_rate": 4.929113037505642e-06,
+      "loss": 3.3922,
+      "step": 1698
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.2079804852931355,
+      "learning_rate": 4.91795144244311e-06,
+      "loss": 3.4299,
+      "step": 1699
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.873718125760766,
+      "learning_rate": 4.9067983767123736e-06,
+      "loss": 3.5628,
+      "step": 1700
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1974630418683976,
+      "learning_rate": 4.895653859031906e-06,
+      "loss": 3.6201,
+      "step": 1701
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2799431494589464,
+      "learning_rate": 4.884517908105837e-06,
+      "loss": 3.43,
+      "step": 1702
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6073965976368547,
+      "learning_rate": 4.873390542623922e-06,
+      "loss": 3.4691,
+      "step": 1703
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6439670250442,
+      "learning_rate": 4.8622717812615e-06,
+      "loss": 3.2538,
+      "step": 1704
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.884889324274793,
+      "learning_rate": 4.851161642679466e-06,
+      "loss": 3.3603,
+      "step": 1705
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.371748248373679,
+      "learning_rate": 4.840060145524254e-06,
+      "loss": 3.4982,
+      "step": 1706
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.07931994670271,
+      "learning_rate": 4.828967308427795e-06,
+      "loss": 3.2928,
+      "step": 1707
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.704691505269074,
+      "learning_rate": 4.817883150007474e-06,
+      "loss": 3.3268,
+      "step": 1708
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.609481060520609,
+      "learning_rate": 4.806807688866119e-06,
+      "loss": 3.5597,
+      "step": 1709
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.287120337678626,
+      "learning_rate": 4.795740943591955e-06,
+      "loss": 3.7014,
+      "step": 1710
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2417233501828395,
+      "learning_rate": 4.784682932758588e-06,
+      "loss": 3.4061,
+      "step": 1711
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.099954943224899,
+      "learning_rate": 4.77363367492496e-06,
+      "loss": 3.3982,
+      "step": 1712
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4674350398982248,
+      "learning_rate": 4.7625931886353215e-06,
+      "loss": 3.5787,
+      "step": 1713
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.8299806212591014,
+      "learning_rate": 4.7515614924192026e-06,
+      "loss": 3.2252,
+      "step": 1714
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3829512664027224,
+      "learning_rate": 4.740538604791371e-06,
+      "loss": 3.585,
+      "step": 1715
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5722636561234493,
+      "learning_rate": 4.729524544251837e-06,
+      "loss": 3.4258,
+      "step": 1716
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.050179515902923,
+      "learning_rate": 4.718519329285771e-06,
+      "loss": 3.3031,
+      "step": 1717
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.128925748003555,
+      "learning_rate": 4.707522978363508e-06,
+      "loss": 3.5009,
+      "step": 1718
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.002009858329216,
+      "learning_rate": 4.696535509940499e-06,
+      "loss": 3.4921,
+      "step": 1719
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6431409914843242,
+      "learning_rate": 4.685556942457296e-06,
+      "loss": 3.5246,
+      "step": 1720
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.8368553209068117,
+      "learning_rate": 4.674587294339513e-06,
+      "loss": 3.6919,
+      "step": 1721
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.9066283355125186,
+      "learning_rate": 4.663626583997789e-06,
+      "loss": 3.388,
+      "step": 1722
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.337474491676217,
+      "learning_rate": 4.652674829827762e-06,
+      "loss": 3.6723,
+      "step": 1723
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.8716302465293615,
+      "learning_rate": 4.641732050210032e-06,
+      "loss": 3.5052,
+      "step": 1724
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5067245303171237,
+      "learning_rate": 4.630798263510162e-06,
+      "loss": 3.6115,
+      "step": 1725
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7223703218552573,
+      "learning_rate": 4.619873488078597e-06,
+      "loss": 3.3937,
+      "step": 1726
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.510075235746018,
+      "learning_rate": 4.608957742250667e-06,
+      "loss": 3.5372,
+      "step": 1727
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.9874952880001366,
+      "learning_rate": 4.598051044346542e-06,
+      "loss": 3.3717,
+      "step": 1728
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.885816752060376,
+      "learning_rate": 4.587153412671217e-06,
+      "loss": 3.3987,
+      "step": 1729
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7827071572281445,
+      "learning_rate": 4.576264865514467e-06,
+      "loss": 3.1763,
+      "step": 1730
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.890063791368322,
+      "learning_rate": 4.565385421150817e-06,
+      "loss": 3.5614,
+      "step": 1731
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.783650330588731,
+      "learning_rate": 4.554515097839511e-06,
+      "loss": 3.4039,
+      "step": 1732
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.973268905334733,
+      "learning_rate": 4.543653913824496e-06,
+      "loss": 3.3783,
+      "step": 1733
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.422375253114076,
+      "learning_rate": 4.53280188733437e-06,
+      "loss": 3.4258,
+      "step": 1734
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6091058977682224,
+      "learning_rate": 4.521959036582372e-06,
+      "loss": 3.5038,
+      "step": 1735
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.449899993245845,
+      "learning_rate": 4.511125379766332e-06,
+      "loss": 3.3051,
+      "step": 1736
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3716372041379747,
+      "learning_rate": 4.500300935068647e-06,
+      "loss": 3.3973,
+      "step": 1737
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5533894107507735,
+      "learning_rate": 4.489485720656266e-06,
+      "loss": 3.4855,
+      "step": 1738
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.559773371573033,
+      "learning_rate": 4.478679754680639e-06,
+      "loss": 3.4554,
+      "step": 1739
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.138358956281304,
+      "learning_rate": 4.467883055277696e-06,
+      "loss": 3.5556,
+      "step": 1740
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.049507644251867,
+      "learning_rate": 4.457095640567804e-06,
+      "loss": 3.5686,
+      "step": 1741
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.11767434244862,
+      "learning_rate": 4.4463175286557654e-06,
+      "loss": 3.6089,
+      "step": 1742
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3567108518742557,
+      "learning_rate": 4.435548737630756e-06,
+      "loss": 3.4967,
+      "step": 1743
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6967941537174314,
+      "learning_rate": 4.4247892855663164e-06,
+      "loss": 3.3797,
+      "step": 1744
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.179141600245913,
+      "learning_rate": 4.414039190520308e-06,
+      "loss": 3.5925,
+      "step": 1745
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.453056516550807,
+      "learning_rate": 4.403298470534885e-06,
+      "loss": 3.6349,
+      "step": 1746
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5712888486599774,
+      "learning_rate": 4.39256714363648e-06,
+      "loss": 3.6258,
+      "step": 1747
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.31493960973444,
+      "learning_rate": 4.3818452278357445e-06,
+      "loss": 3.4372,
+      "step": 1748
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.171432917668506,
+      "learning_rate": 4.371132741127553e-06,
+      "loss": 3.3394,
+      "step": 1749
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.656724448043528,
+      "learning_rate": 4.360429701490935e-06,
+      "loss": 3.3471,
+      "step": 1750
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.186499874648426,
+      "learning_rate": 4.349736126889084e-06,
+      "loss": 3.3913,
+      "step": 1751
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.8208946244534188,
+      "learning_rate": 4.339052035269291e-06,
+      "loss": 3.6172,
+      "step": 1752
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.116100858707486,
+      "learning_rate": 4.328377444562948e-06,
+      "loss": 3.4454,
+      "step": 1753
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.161503924768385,
+      "learning_rate": 4.31771237268549e-06,
+      "loss": 3.6466,
+      "step": 1754
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.494899046455282,
+      "learning_rate": 4.307056837536373e-06,
+      "loss": 3.4494,
+      "step": 1755
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.767898841687319,
+      "learning_rate": 4.296410856999062e-06,
+      "loss": 3.3613,
+      "step": 1756
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.9982391161187314,
+      "learning_rate": 4.2857744489409725e-06,
+      "loss": 3.6314,
+      "step": 1757
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5058775605789996,
+      "learning_rate": 4.2751476312134655e-06,
+      "loss": 3.4211,
+      "step": 1758
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.697058493781567,
+      "learning_rate": 4.264530421651792e-06,
+      "loss": 3.5555,
+      "step": 1759
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.011432271491016,
+      "learning_rate": 4.2539228380750955e-06,
+      "loss": 3.241,
+      "step": 1760
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.9185907235318886,
+      "learning_rate": 4.243324898286349e-06,
+      "loss": 3.4969,
+      "step": 1761
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5614235766322233,
+      "learning_rate": 4.2327366200723404e-06,
+      "loss": 3.3976,
+      "step": 1762
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.8839560924349317,
+      "learning_rate": 4.222158021203657e-06,
+      "loss": 3.4789,
+      "step": 1763
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.671243277730213,
+      "learning_rate": 4.211589119434622e-06,
+      "loss": 3.5638,
+      "step": 1764
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3993535836751265,
+      "learning_rate": 4.201029932503303e-06,
+      "loss": 3.5107,
+      "step": 1765
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.013661425202177,
+      "learning_rate": 4.190480478131443e-06,
+      "loss": 3.2718,
+      "step": 1766
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3719456559452223,
+      "learning_rate": 4.179940774024469e-06,
+      "loss": 3.5625,
+      "step": 1767
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2931148487027095,
+      "learning_rate": 4.169410837871427e-06,
+      "loss": 3.4839,
+      "step": 1768
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2833663912883853,
+      "learning_rate": 4.158890687344986e-06,
+      "loss": 3.3587,
+      "step": 1769
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.8882103322283528,
+      "learning_rate": 4.14838034010138e-06,
+      "loss": 3.3663,
+      "step": 1770
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.8955310931768823,
+      "learning_rate": 4.137879813780388e-06,
+      "loss": 3.4513,
+      "step": 1771
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.636956183849931,
+      "learning_rate": 4.127389126005319e-06,
+      "loss": 3.5105,
+      "step": 1772
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6484577281266577,
+      "learning_rate": 4.116908294382956e-06,
+      "loss": 3.6188,
+      "step": 1773
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.0266571534966005,
+      "learning_rate": 4.10643733650355e-06,
+      "loss": 3.304,
+      "step": 1774
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.88428003537644,
+      "learning_rate": 4.095976269940777e-06,
+      "loss": 3.3975,
+      "step": 1775
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6192466745520893,
+      "learning_rate": 4.085525112251706e-06,
+      "loss": 3.5155,
+      "step": 1776
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.704237885509808,
+      "learning_rate": 4.0750838809767875e-06,
+      "loss": 3.5143,
+      "step": 1777
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.678360702162961,
+      "learning_rate": 4.0646525936398086e-06,
+      "loss": 3.542,
+      "step": 1778
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.389314698208116,
+      "learning_rate": 4.054231267747862e-06,
+      "loss": 3.5115,
+      "step": 1779
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.701937280915647,
+      "learning_rate": 4.043819920791322e-06,
+      "loss": 3.4142,
+      "step": 1780
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.9218694258931435,
+      "learning_rate": 4.033418570243819e-06,
+      "loss": 3.4976,
+      "step": 1781
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.721482825872019,
+      "learning_rate": 4.0230272335622065e-06,
+      "loss": 3.4643,
+      "step": 1782
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6470247224867243,
+      "learning_rate": 4.012645928186533e-06,
+      "loss": 3.5088,
+      "step": 1783
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.7696088401191385,
+      "learning_rate": 4.002274671540006e-06,
+      "loss": 3.2971,
+      "step": 1784
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.7052611261579202,
+      "learning_rate": 3.991913481028965e-06,
+      "loss": 3.5161,
+      "step": 1785
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.7585081410257684,
+      "learning_rate": 3.981562374042867e-06,
+      "loss": 3.1938,
+      "step": 1786
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.9585861280160684,
+      "learning_rate": 3.971221367954239e-06,
+      "loss": 3.4439,
+      "step": 1787
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.283636045591956,
+      "learning_rate": 3.960890480118653e-06,
+      "loss": 3.6489,
+      "step": 1788
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.654578635777536,
+      "learning_rate": 3.950569727874704e-06,
+      "loss": 3.4375,
+      "step": 1789
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.169022532820428,
+      "learning_rate": 3.940259128543967e-06,
+      "loss": 3.499,
+      "step": 1790
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3344543139240477,
+      "learning_rate": 3.9299586994309905e-06,
+      "loss": 3.4579,
+      "step": 1791
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.746352581841832,
+      "learning_rate": 3.919668457823248e-06,
+      "loss": 3.4583,
+      "step": 1792
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6192979524953657,
+      "learning_rate": 3.909388420991113e-06,
+      "loss": 3.4426,
+      "step": 1793
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.736628337328482,
+      "learning_rate": 3.899118606187832e-06,
+      "loss": 3.5413,
+      "step": 1794
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4436637095613287,
+      "learning_rate": 3.888859030649498e-06,
+      "loss": 3.4048,
+      "step": 1795
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2072769723867838,
+      "learning_rate": 3.878609711595022e-06,
+      "loss": 3.3558,
+      "step": 1796
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.7640368299387825,
+      "learning_rate": 3.8683706662260945e-06,
+      "loss": 3.4847,
+      "step": 1797
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.436055272932363,
+      "learning_rate": 3.858141911727168e-06,
+      "loss": 3.2878,
+      "step": 1798
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.529156547637179,
+      "learning_rate": 3.847923465265418e-06,
+      "loss": 3.7081,
+      "step": 1799
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.468301049973078,
+      "learning_rate": 3.837715343990727e-06,
+      "loss": 3.3457,
+      "step": 1800
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5741394728587204,
+      "learning_rate": 3.8275175650356485e-06,
+      "loss": 3.523,
+      "step": 1801
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6469623554292134,
+      "learning_rate": 3.817330145515374e-06,
+      "loss": 3.3227,
+      "step": 1802
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.332529000908109,
+      "learning_rate": 3.807153102527704e-06,
+      "loss": 3.3699,
+      "step": 1803
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.208393616053263,
+      "learning_rate": 3.7969864531530344e-06,
+      "loss": 3.4386,
+      "step": 1804
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.512302703646855,
+      "learning_rate": 3.7868302144543146e-06,
+      "loss": 3.6729,
+      "step": 1805
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.7091168641858503,
+      "learning_rate": 3.7766844034770155e-06,
+      "loss": 3.7557,
+      "step": 1806
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.132021104400233,
+      "learning_rate": 3.766549037249112e-06,
+      "loss": 3.3382,
+      "step": 1807
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.002942906418598,
+      "learning_rate": 3.7564241327810436e-06,
+      "loss": 3.4619,
+      "step": 1808
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.9271576290560906,
+      "learning_rate": 3.7463097070656995e-06,
+      "loss": 3.5258,
+      "step": 1809
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.98998238474593,
+      "learning_rate": 3.736205777078381e-06,
+      "loss": 3.4328,
+      "step": 1810
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.765986861107831,
+      "learning_rate": 3.72611235977677e-06,
+      "loss": 3.371,
+      "step": 1811
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.13175319308379,
+      "learning_rate": 3.7160294721009026e-06,
+      "loss": 3.2895,
+      "step": 1812
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.5676925128829895,
+      "learning_rate": 3.705957130973149e-06,
+      "loss": 3.3525,
+      "step": 1813
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.065579675390456,
+      "learning_rate": 3.69589535329818e-06,
+      "loss": 3.4717,
+      "step": 1814
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.6964342085032156,
+      "learning_rate": 3.685844155962931e-06,
+      "loss": 3.374,
+      "step": 1815
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.5454709243268656,
+      "learning_rate": 3.675803555836582e-06,
+      "loss": 3.2699,
+      "step": 1816
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.6910173645241513,
+      "learning_rate": 3.6657735697705267e-06,
+      "loss": 3.2236,
+      "step": 1817
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.588311501677217,
+      "learning_rate": 3.6557542145983495e-06,
+      "loss": 3.5066,
+      "step": 1818
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.651925132300316,
+      "learning_rate": 3.6457455071357918e-06,
+      "loss": 3.3634,
+      "step": 1819
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.250683821727388,
+      "learning_rate": 3.63574746418072e-06,
+      "loss": 3.267,
+      "step": 1820
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.429181473137611,
+      "learning_rate": 3.625760102513103e-06,
+      "loss": 3.5255,
+      "step": 1821
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1222267069199283,
+      "learning_rate": 3.6157834388949907e-06,
+      "loss": 3.5681,
+      "step": 1822
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.43599969407311,
+      "learning_rate": 3.6058174900704646e-06,
+      "loss": 3.5532,
+      "step": 1823
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.9924024367280544,
+      "learning_rate": 3.595862272765638e-06,
+      "loss": 3.3557,
+      "step": 1824
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.335774679391719,
+      "learning_rate": 3.585917803688603e-06,
+      "loss": 3.1922,
+      "step": 1825
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.317027772024,
+      "learning_rate": 3.5759840995294136e-06,
+      "loss": 3.3783,
+      "step": 1826
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.805915177072342,
+      "learning_rate": 3.5660611769600604e-06,
+      "loss": 3.5855,
+      "step": 1827
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.9192673343873063,
+      "learning_rate": 3.556149052634443e-06,
+      "loss": 3.4866,
+      "step": 1828
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.484350084902121,
+      "learning_rate": 3.546247743188328e-06,
+      "loss": 3.5675,
+      "step": 1829
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.103281813841342,
+      "learning_rate": 3.536357265239333e-06,
+      "loss": 3.5552,
+      "step": 1830
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.450382295757395,
+      "learning_rate": 3.5264776353869046e-06,
+      "loss": 3.4087,
+      "step": 1831
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.437727669001523,
+      "learning_rate": 3.5166088702122738e-06,
+      "loss": 3.3265,
+      "step": 1832
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.346381212758512,
+      "learning_rate": 3.5067509862784455e-06,
+      "loss": 3.4703,
+      "step": 1833
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.587343604875265,
+      "learning_rate": 3.4969040001301513e-06,
+      "loss": 3.6085,
+      "step": 1834
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.854682116713474,
+      "learning_rate": 3.487067928293848e-06,
+      "loss": 3.4987,
+      "step": 1835
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.9170447966343542,
+      "learning_rate": 3.4772427872776606e-06,
+      "loss": 3.3871,
+      "step": 1836
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.652408689543764,
+      "learning_rate": 3.4674285935713715e-06,
+      "loss": 3.5386,
+      "step": 1837
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.6076480873164796,
+      "learning_rate": 3.4576253636463996e-06,
+      "loss": 3.4289,
+      "step": 1838
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.633468327300971,
+      "learning_rate": 3.4478331139557475e-06,
+      "loss": 3.4698,
+      "step": 1839
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.584011854682265,
+      "learning_rate": 3.4380518609340076e-06,
+      "loss": 3.3809,
+      "step": 1840
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4086643945235453,
+      "learning_rate": 3.428281620997296e-06,
+      "loss": 3.4472,
+      "step": 1841
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7236703417587798,
+      "learning_rate": 3.418522410543266e-06,
+      "loss": 3.3683,
+      "step": 1842
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.516193849667952,
+      "learning_rate": 3.4087742459510396e-06,
+      "loss": 3.3648,
+      "step": 1843
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.0536722168971995,
+      "learning_rate": 3.3990371435812185e-06,
+      "loss": 3.3732,
+      "step": 1844
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7285338094616565,
+      "learning_rate": 3.3893111197758276e-06,
+      "loss": 3.3925,
+      "step": 1845
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7303715346811033,
+      "learning_rate": 3.3795961908582965e-06,
+      "loss": 3.3005,
+      "step": 1846
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.9474329011980913,
+      "learning_rate": 3.3698923731334453e-06,
+      "loss": 3.4754,
+      "step": 1847
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.5606404505123774,
+      "learning_rate": 3.360199682887433e-06,
+      "loss": 3.675,
+      "step": 1848
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7856691217724734,
+      "learning_rate": 3.3505181363877536e-06,
+      "loss": 3.342,
+      "step": 1849
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.300592236425659,
+      "learning_rate": 3.3408477498831917e-06,
+      "loss": 3.2336,
+      "step": 1850
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.511252927635383,
+      "learning_rate": 3.3311885396038002e-06,
+      "loss": 3.47,
+      "step": 1851
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.496345516082245,
+      "learning_rate": 3.321540521760883e-06,
+      "loss": 3.4671,
+      "step": 1852
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.20909351432755,
+      "learning_rate": 3.3119037125469553e-06,
+      "loss": 3.6159,
+      "step": 1853
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.5867603857436006,
+      "learning_rate": 3.3022781281357184e-06,
+      "loss": 3.7095,
+      "step": 1854
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4734426795727376,
+      "learning_rate": 3.2926637846820366e-06,
+      "loss": 3.5205,
+      "step": 1855
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.055599777957748,
+      "learning_rate": 3.2830606983219038e-06,
+      "loss": 3.4546,
+      "step": 1856
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.196656750316166,
+      "learning_rate": 3.2734688851724273e-06,
+      "loss": 3.3104,
+      "step": 1857
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.761411862813457,
+      "learning_rate": 3.2638883613317974e-06,
+      "loss": 3.4606,
+      "step": 1858
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2224622867474353,
+      "learning_rate": 3.2543191428792466e-06,
+      "loss": 3.6587,
+      "step": 1859
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.252120608521762,
+      "learning_rate": 3.2447612458750365e-06,
+      "loss": 3.3622,
+      "step": 1860
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4210000097935085,
+      "learning_rate": 3.2352146863604317e-06,
+      "loss": 3.3387,
+      "step": 1861
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.9081317042211747,
+      "learning_rate": 3.2256794803576707e-06,
+      "loss": 3.5707,
+      "step": 1862
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.2730047394313795,
+      "learning_rate": 3.2161556438699303e-06,
+      "loss": 3.5105,
+      "step": 1863
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.566834764286657,
+      "learning_rate": 3.2066431928813068e-06,
+      "loss": 3.5039,
+      "step": 1864
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.4615537663089855,
+      "learning_rate": 3.197142143356787e-06,
+      "loss": 3.2387,
+      "step": 1865
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.7727034614513606,
+      "learning_rate": 3.1876525112422283e-06,
+      "loss": 3.4536,
+      "step": 1866
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.850992472141217,
+      "learning_rate": 3.178174312464326e-06,
+      "loss": 3.4271,
+      "step": 1867
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.4404782089592967,
+      "learning_rate": 3.1687075629305787e-06,
+      "loss": 3.5452,
+      "step": 1868
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.4309534748380246,
+      "learning_rate": 3.1592522785292714e-06,
+      "loss": 3.5336,
+      "step": 1869
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.9067620491946853,
+      "learning_rate": 3.1498084751294523e-06,
+      "loss": 3.598,
+      "step": 1870
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.448718948350417,
+      "learning_rate": 3.1403761685809007e-06,
+      "loss": 3.3707,
+      "step": 1871
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3574537673903415,
+      "learning_rate": 3.130955374714094e-06,
+      "loss": 3.4151,
+      "step": 1872
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.542471370777354,
+      "learning_rate": 3.1215461093401913e-06,
+      "loss": 3.2864,
+      "step": 1873
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.96498352917791,
+      "learning_rate": 3.1121483882509996e-06,
+      "loss": 3.4146,
+      "step": 1874
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9306184568657394,
+      "learning_rate": 3.1027622272189572e-06,
+      "loss": 3.6675,
+      "step": 1875
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.563712587378033,
+      "learning_rate": 3.0933876419971008e-06,
+      "loss": 3.5339,
+      "step": 1876
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.656991955040363,
+      "learning_rate": 3.0840246483190338e-06,
+      "loss": 3.4484,
+      "step": 1877
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8382126444965428,
+      "learning_rate": 3.074673261898903e-06,
+      "loss": 3.4853,
+      "step": 1878
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.775782565231201,
+      "learning_rate": 3.065333498431381e-06,
+      "loss": 3.193,
+      "step": 1879
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.55715379265718,
+      "learning_rate": 3.0560053735916372e-06,
+      "loss": 3.3536,
+      "step": 1880
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.852296770838666,
+      "learning_rate": 3.0466889030352976e-06,
+      "loss": 3.5737,
+      "step": 1881
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.9813297344776477,
+      "learning_rate": 3.037384102398431e-06,
+      "loss": 3.6859,
+      "step": 1882
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.7225704836822513,
+      "learning_rate": 3.0280909872975194e-06,
+      "loss": 3.4207,
+      "step": 1883
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.424724501437131,
+      "learning_rate": 3.0188095733294388e-06,
+      "loss": 3.4869,
+      "step": 1884
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5295304976839144,
+      "learning_rate": 3.009539876071427e-06,
+      "loss": 3.3601,
+      "step": 1885
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5881747955842727,
+      "learning_rate": 3.0002819110810475e-06,
+      "loss": 3.6793,
+      "step": 1886
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.9583141250494607,
+      "learning_rate": 2.9910356938961782e-06,
+      "loss": 3.0941,
+      "step": 1887
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.912358664094153,
+      "learning_rate": 2.981801240034985e-06,
+      "loss": 3.3709,
+      "step": 1888
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.263720825562735,
+      "learning_rate": 2.9725785649958895e-06,
+      "loss": 3.3977,
+      "step": 1889
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.509980044518029,
+      "learning_rate": 2.9633676842575386e-06,
+      "loss": 3.3369,
+      "step": 1890
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1011270995474804,
+      "learning_rate": 2.9541686132787907e-06,
+      "loss": 3.4079,
+      "step": 1891
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0679947729345174,
+      "learning_rate": 2.944981367498677e-06,
+      "loss": 3.3624,
+      "step": 1892
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.288148926844505,
+      "learning_rate": 2.93580596233639e-06,
+      "loss": 3.4671,
+      "step": 1893
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6214058326443292,
+      "learning_rate": 2.9266424131912495e-06,
+      "loss": 3.3457,
+      "step": 1894
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.53592212281632,
+      "learning_rate": 2.9174907354426696e-06,
+      "loss": 3.3712,
+      "step": 1895
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5987696162272527,
+      "learning_rate": 2.9083509444501433e-06,
+      "loss": 3.3665,
+      "step": 1896
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.085596638238194,
+      "learning_rate": 2.899223055553221e-06,
+      "loss": 3.6694,
+      "step": 1897
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.4340867270724083,
+      "learning_rate": 2.890107084071465e-06,
+      "loss": 3.316,
+      "step": 1898
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.285488215778594,
+      "learning_rate": 2.881003045304448e-06,
+      "loss": 3.4162,
+      "step": 1899
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6478133045438454,
+      "learning_rate": 2.8719109545317102e-06,
+      "loss": 3.4196,
+      "step": 1900
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.4821976881960612,
+      "learning_rate": 2.8628308270127335e-06,
+      "loss": 3.2749,
+      "step": 1901
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8625654362844153,
+      "learning_rate": 2.853762677986932e-06,
+      "loss": 3.1237,
+      "step": 1902
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.9299865920722934,
+      "learning_rate": 2.844706522673616e-06,
+      "loss": 3.5478,
+      "step": 1903
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.410719964347605,
+      "learning_rate": 2.835662376271957e-06,
+      "loss": 3.7158,
+      "step": 1904
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.873022927865709,
+      "learning_rate": 2.8266302539609747e-06,
+      "loss": 3.4461,
+      "step": 1905
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.67026539624037,
+      "learning_rate": 2.8176101708995174e-06,
+      "loss": 3.4867,
+      "step": 1906
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.9182061324797237,
+      "learning_rate": 2.808602142226212e-06,
+      "loss": 3.2738,
+      "step": 1907
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.124918425721532,
+      "learning_rate": 2.7996061830594714e-06,
+      "loss": 3.4517,
+      "step": 1908
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.105538309675829,
+      "learning_rate": 2.7906223084974405e-06,
+      "loss": 3.2271,
+      "step": 1909
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3859102048877254,
+      "learning_rate": 2.78165053361798e-06,
+      "loss": 3.3504,
+      "step": 1910
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.7425966269377486,
+      "learning_rate": 2.772690873478656e-06,
+      "loss": 3.3253,
+      "step": 1911
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.310424851528983,
+      "learning_rate": 2.7637433431166903e-06,
+      "loss": 3.2082,
+      "step": 1912
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.221875373344164,
+      "learning_rate": 2.754807957548955e-06,
+      "loss": 3.6851,
+      "step": 1913
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5123775223492832,
+      "learning_rate": 2.745884731771931e-06,
+      "loss": 3.5482,
+      "step": 1914
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.044351047721393,
+      "learning_rate": 2.736973680761702e-06,
+      "loss": 3.4826,
+      "step": 1915
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.1297812159014615,
+      "learning_rate": 2.728074819473908e-06,
+      "loss": 3.4117,
+      "step": 1916
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8714590716432395,
+      "learning_rate": 2.7191881628437335e-06,
+      "loss": 3.479,
+      "step": 1917
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5456711355596324,
+      "learning_rate": 2.7103137257858867e-06,
+      "loss": 3.4652,
+      "step": 1918
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8344813934172226,
+      "learning_rate": 2.7014515231945557e-06,
+      "loss": 3.4678,
+      "step": 1919
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4176638962488086,
+      "learning_rate": 2.692601569943407e-06,
+      "loss": 3.5222,
+      "step": 1920
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.1479220590942765,
+      "learning_rate": 2.683763880885538e-06,
+      "loss": 3.5278,
+      "step": 1921
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7676113482167803,
+      "learning_rate": 2.674938470853472e-06,
+      "loss": 3.5333,
+      "step": 1922
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.506946240907651,
+      "learning_rate": 2.6661253546591158e-06,
+      "loss": 3.6749,
+      "step": 1923
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.6055160050017987,
+      "learning_rate": 2.6573245470937527e-06,
+      "loss": 3.518,
+      "step": 1924
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7106478573047106,
+      "learning_rate": 2.648536062927999e-06,
+      "loss": 3.54,
+      "step": 1925
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5451801376739276,
+      "learning_rate": 2.639759916911788e-06,
+      "loss": 3.5237,
+      "step": 1926
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8798623404933728,
+      "learning_rate": 2.6309961237743587e-06,
+      "loss": 3.491,
+      "step": 1927
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5768159862331284,
+      "learning_rate": 2.6222446982242e-06,
+      "loss": 3.2704,
+      "step": 1928
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.269484671851689,
+      "learning_rate": 2.61350565494906e-06,
+      "loss": 3.4937,
+      "step": 1929
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.332641827711261,
+      "learning_rate": 2.604779008615895e-06,
+      "loss": 3.5345,
+      "step": 1930
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.829641692493816,
+      "learning_rate": 2.5960647738708553e-06,
+      "loss": 3.3914,
+      "step": 1931
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.231876182233534,
+      "learning_rate": 2.5873629653392653e-06,
+      "loss": 3.5457,
+      "step": 1932
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.92942692841208,
+      "learning_rate": 2.578673597625597e-06,
+      "loss": 3.4388,
+      "step": 1933
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.207353567728463,
+      "learning_rate": 2.569996685313434e-06,
+      "loss": 3.4373,
+      "step": 1934
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.81926065122999,
+      "learning_rate": 2.5613322429654573e-06,
+      "loss": 3.4003,
+      "step": 1935
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.736117305431284,
+      "learning_rate": 2.5526802851234268e-06,
+      "loss": 3.259,
+      "step": 1936
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8281798178776683,
+      "learning_rate": 2.5440408263081385e-06,
+      "loss": 3.5264,
+      "step": 1937
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.418895548861695,
+      "learning_rate": 2.535413881019423e-06,
+      "loss": 3.3591,
+      "step": 1938
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4143628309124843,
+      "learning_rate": 2.526799463736099e-06,
+      "loss": 3.3624,
+      "step": 1939
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.041022539593813,
+      "learning_rate": 2.5181975889159615e-06,
+      "loss": 3.5832,
+      "step": 1940
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6739434254765775,
+      "learning_rate": 2.509608270995758e-06,
+      "loss": 3.3057,
+      "step": 1941
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.493683119459877,
+      "learning_rate": 2.501031524391163e-06,
+      "loss": 3.1586,
+      "step": 1942
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.8597452477414396,
+      "learning_rate": 2.492467363496747e-06,
+      "loss": 3.3849,
+      "step": 1943
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.817698743749816,
+      "learning_rate": 2.483915802685959e-06,
+      "loss": 3.46,
+      "step": 1944
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.0309323747808055,
+      "learning_rate": 2.475376856311097e-06,
+      "loss": 3.5345,
+      "step": 1945
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4469163019967795,
+      "learning_rate": 2.4668505387033025e-06,
+      "loss": 3.3929,
+      "step": 1946
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.761660769645593,
+      "learning_rate": 2.458336864172508e-06,
+      "loss": 3.4265,
+      "step": 1947
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.534763224235543,
+      "learning_rate": 2.44983584700743e-06,
+      "loss": 3.696,
+      "step": 1948
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3341157071801257,
+      "learning_rate": 2.4413475014755396e-06,
+      "loss": 3.385,
+      "step": 1949
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.9798835388607525,
+      "learning_rate": 2.432871841823047e-06,
+      "loss": 3.2967,
+      "step": 1950
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.9600306056181904,
+      "learning_rate": 2.42440888227487e-06,
+      "loss": 3.2849,
+      "step": 1951
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.104498375827391,
+      "learning_rate": 2.415958637034609e-06,
+      "loss": 3.4203,
+      "step": 1952
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.7502083975036826,
+      "learning_rate": 2.407521120284523e-06,
+      "loss": 3.3641,
+      "step": 1953
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5881591927638516,
+      "learning_rate": 2.3990963461855075e-06,
+      "loss": 3.2882,
+      "step": 1954
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.428645383919292,
+      "learning_rate": 2.390684328877089e-06,
+      "loss": 3.4263,
+      "step": 1955
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.721307523264979,
+      "learning_rate": 2.3822850824773623e-06,
+      "loss": 3.4291,
+      "step": 1956
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6339505062874577,
+      "learning_rate": 2.3738986210829997e-06,
+      "loss": 3.5408,
+      "step": 1957
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.7015671534693193,
+      "learning_rate": 2.3655249587692073e-06,
+      "loss": 3.367,
+      "step": 1958
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4071277961892066,
+      "learning_rate": 2.3571641095897223e-06,
+      "loss": 3.1734,
+      "step": 1959
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.704927639099572,
+      "learning_rate": 2.3488160875767717e-06,
+      "loss": 3.4119,
+      "step": 1960
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.1341363150404877,
+      "learning_rate": 2.340480906741053e-06,
+      "loss": 3.2445,
+      "step": 1961
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.58868146610442,
+      "learning_rate": 2.332158581071712e-06,
+      "loss": 3.5284,
+      "step": 1962
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.31203736033518,
+      "learning_rate": 2.323849124536315e-06,
+      "loss": 3.4336,
+      "step": 1963
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.813832593935863,
+      "learning_rate": 2.3155525510808453e-06,
+      "loss": 3.509,
+      "step": 1964
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.0666930929331615,
+      "learning_rate": 2.307268874629649e-06,
+      "loss": 3.3478,
+      "step": 1965
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.479016340534677,
+      "learning_rate": 2.2989981090854306e-06,
+      "loss": 3.5218,
+      "step": 1966
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.087492566542144,
+      "learning_rate": 2.2907402683292268e-06,
+      "loss": 3.4544,
+      "step": 1967
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.227066132714872,
+      "learning_rate": 2.2824953662203832e-06,
+      "loss": 3.4573,
+      "step": 1968
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.859067379027617,
+      "learning_rate": 2.2742634165965317e-06,
+      "loss": 3.4539,
+      "step": 1969
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.413802073964399,
+      "learning_rate": 2.266044433273562e-06,
+      "loss": 3.3942,
+      "step": 1970
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.086380726334998,
+      "learning_rate": 2.2578384300456014e-06,
+      "loss": 3.5544,
+      "step": 1971
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.591331242244212,
+      "learning_rate": 2.249645420684998e-06,
+      "loss": 3.1801,
+      "step": 1972
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.990643884284416,
+      "learning_rate": 2.2414654189422845e-06,
+      "loss": 3.3328,
+      "step": 1973
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.7177062831411893,
+      "learning_rate": 2.233298438546172e-06,
+      "loss": 3.5575,
+      "step": 1974
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.057689293818614,
+      "learning_rate": 2.2251444932035094e-06,
+      "loss": 3.5605,
+      "step": 1975
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5873009841450383,
+      "learning_rate": 2.2170035965992674e-06,
+      "loss": 3.5157,
+      "step": 1976
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.539103717358677,
+      "learning_rate": 2.2088757623965263e-06,
+      "loss": 3.3443,
+      "step": 1977
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.0314317779393285,
+      "learning_rate": 2.2007610042364337e-06,
+      "loss": 3.4664,
+      "step": 1978
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.319751629440479,
+      "learning_rate": 2.1926593357382e-06,
+      "loss": 3.3711,
+      "step": 1979
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5932358520391667,
+      "learning_rate": 2.184570770499056e-06,
+      "loss": 3.4833,
+      "step": 1980
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.367779707746548,
+      "learning_rate": 2.176495322094254e-06,
+      "loss": 3.3852,
+      "step": 1981
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3471031613965927,
+      "learning_rate": 2.1684330040770183e-06,
+      "loss": 3.5292,
+      "step": 1982
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3645699007477976,
+      "learning_rate": 2.1603838299785486e-06,
+      "loss": 3.3293,
+      "step": 1983
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.175500406999358,
+      "learning_rate": 2.1523478133079776e-06,
+      "loss": 3.4913,
+      "step": 1984
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.455711229312703,
+      "learning_rate": 2.1443249675523536e-06,
+      "loss": 3.4397,
+      "step": 1985
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3649792008508643,
+      "learning_rate": 2.1363153061766297e-06,
+      "loss": 3.2848,
+      "step": 1986
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.389255501639541,
+      "learning_rate": 2.128318842623618e-06,
+      "loss": 3.3366,
+      "step": 1987
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9054400348160923,
+      "learning_rate": 2.1203355903139934e-06,
+      "loss": 3.3767,
+      "step": 1988
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.202178172426877,
+      "learning_rate": 2.112365562646248e-06,
+      "loss": 3.3604,
+      "step": 1989
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6981929021416575,
+      "learning_rate": 2.1044087729966856e-06,
+      "loss": 3.5228,
+      "step": 1990
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9855355748316836,
+      "learning_rate": 2.0964652347193894e-06,
+      "loss": 3.4482,
+      "step": 1991
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.522924579103291,
+      "learning_rate": 2.088534961146197e-06,
+      "loss": 3.4379,
+      "step": 1992
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2621294434046266,
+      "learning_rate": 2.0806179655866964e-06,
+      "loss": 3.4508,
+      "step": 1993
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5103053461158957,
+      "learning_rate": 2.072714261328177e-06,
+      "loss": 3.2115,
+      "step": 1994
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6445225225407483,
+      "learning_rate": 2.064823861635633e-06,
+      "loss": 3.4337,
+      "step": 1995
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.760005549912783,
+      "learning_rate": 2.0569467797517173e-06,
+      "loss": 3.3243,
+      "step": 1996
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2248284331492862,
+      "learning_rate": 2.0490830288967443e-06,
+      "loss": 3.1467,
+      "step": 1997
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.170608568490723,
+      "learning_rate": 2.041232622268642e-06,
+      "loss": 3.1809,
+      "step": 1998
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6936094239380752,
+      "learning_rate": 2.033395573042952e-06,
+      "loss": 3.3979,
+      "step": 1999
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0024131234462748,
+      "learning_rate": 2.025571894372794e-06,
+      "loss": 3.473,
+      "step": 2000
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5589140858814954,
+      "learning_rate": 2.017761599388842e-06,
+      "loss": 3.3617,
+      "step": 2001
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.517662937978228,
+      "learning_rate": 2.0099647011993217e-06,
+      "loss": 3.3516,
+      "step": 2002
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.243366018931923,
+      "learning_rate": 2.00218121288996e-06,
+      "loss": 3.3382,
+      "step": 2003
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5996319864021356,
+      "learning_rate": 1.994411147523987e-06,
+      "loss": 3.4044,
+      "step": 2004
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.931460370263261,
+      "learning_rate": 1.9866545181421016e-06,
+      "loss": 3.394,
+      "step": 2005
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.521015795117093,
+      "learning_rate": 1.97891133776245e-06,
+      "loss": 3.5298,
+      "step": 2006
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.625159681693936,
+      "learning_rate": 1.971181619380611e-06,
+      "loss": 3.192,
+      "step": 2007
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.8264105777364743,
+      "learning_rate": 1.963465375969572e-06,
+      "loss": 3.3629,
+      "step": 2008
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.676670729415169,
+      "learning_rate": 1.955762620479699e-06,
+      "loss": 3.4384,
+      "step": 2009
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.394143019524009,
+      "learning_rate": 1.9480733658387175e-06,
+      "loss": 3.3489,
+      "step": 2010
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2006120148546935,
+      "learning_rate": 1.940397624951709e-06,
+      "loss": 3.4182,
+      "step": 2011
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6169611140886224,
+      "learning_rate": 1.9327354107010566e-06,
+      "loss": 3.4577,
+      "step": 2012
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.518943092013396,
+      "learning_rate": 1.9250867359464575e-06,
+      "loss": 3.3019,
+      "step": 2013
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.257401642789152,
+      "learning_rate": 1.9174516135248745e-06,
+      "loss": 3.5141,
+      "step": 2014
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.6641653222885147,
+      "learning_rate": 1.9098300562505266e-06,
+      "loss": 3.3903,
+      "step": 2015
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.7826868132289713,
+      "learning_rate": 1.902222076914869e-06,
+      "loss": 3.4909,
+      "step": 2016
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5520282434078796,
+      "learning_rate": 1.894627688286571e-06,
+      "loss": 3.312,
+      "step": 2017
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.6540670643755684,
+      "learning_rate": 1.8870469031114868e-06,
+      "loss": 3.3058,
+      "step": 2018
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.735593509692127,
+      "learning_rate": 1.8794797341126403e-06,
+      "loss": 3.423,
+      "step": 2019
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.328850503165228,
+      "learning_rate": 1.8719261939902023e-06,
+      "loss": 3.3835,
+      "step": 2020
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5312614490842464,
+      "learning_rate": 1.8643862954214754e-06,
+      "loss": 3.4535,
+      "step": 2021
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.530286273299819,
+      "learning_rate": 1.8568600510608659e-06,
+      "loss": 3.1957,
+      "step": 2022
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3364923534174658,
+      "learning_rate": 1.8493474735398575e-06,
+      "loss": 3.4549,
+      "step": 2023
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.495691119616844,
+      "learning_rate": 1.8418485754670013e-06,
+      "loss": 3.4042,
+      "step": 2024
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8025310219035995,
+      "learning_rate": 1.8343633694278895e-06,
+      "loss": 3.7479,
+      "step": 2025
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8728585066365886,
+      "learning_rate": 1.8268918679851388e-06,
+      "loss": 3.4093,
+      "step": 2026
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.7564618280991113,
+      "learning_rate": 1.8194340836783565e-06,
+      "loss": 3.3054,
+      "step": 2027
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.681863260494089,
+      "learning_rate": 1.8119900290241331e-06,
+      "loss": 3.5343,
+      "step": 2028
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5455921958019068,
+      "learning_rate": 1.8045597165160134e-06,
+      "loss": 3.3716,
+      "step": 2029
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.356563608206486,
+      "learning_rate": 1.7971431586244814e-06,
+      "loss": 3.4031,
+      "step": 2030
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2452422437785047,
+      "learning_rate": 1.7897403677969405e-06,
+      "loss": 3.3848,
+      "step": 2031
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.914647472140249,
+      "learning_rate": 1.7823513564576788e-06,
+      "loss": 3.3433,
+      "step": 2032
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.856800399645909,
+      "learning_rate": 1.774976137007861e-06,
+      "loss": 3.3345,
+      "step": 2033
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.45709066528513,
+      "learning_rate": 1.7676147218255092e-06,
+      "loss": 3.4962,
+      "step": 2034
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.766310680972499,
+      "learning_rate": 1.7602671232654755e-06,
+      "loss": 3.4752,
+      "step": 2035
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1432477155905585,
+      "learning_rate": 1.7529333536594217e-06,
+      "loss": 3.4079,
+      "step": 2036
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.178186217696704,
+      "learning_rate": 1.7456134253157976e-06,
+      "loss": 3.4548,
+      "step": 2037
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.075392019420518,
+      "learning_rate": 1.7383073505198255e-06,
+      "loss": 3.387,
+      "step": 2038
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.822195351239234,
+      "learning_rate": 1.7310151415334798e-06,
+      "loss": 3.4997,
+      "step": 2039
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.4265966877599485,
+      "learning_rate": 1.723736810595461e-06,
+      "loss": 3.484,
+      "step": 2040
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9330877865873854,
+      "learning_rate": 1.7164723699211782e-06,
+      "loss": 3.4072,
+      "step": 2041
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.25944471322417,
+      "learning_rate": 1.709221831702723e-06,
+      "loss": 3.2468,
+      "step": 2042
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.6246985984462836,
+      "learning_rate": 1.7019852081088616e-06,
+      "loss": 3.429,
+      "step": 2043
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.455585489352441,
+      "learning_rate": 1.6947625112850074e-06,
+      "loss": 3.5424,
+      "step": 2044
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.443555610888772,
+      "learning_rate": 1.687553753353195e-06,
+      "loss": 3.5144,
+      "step": 2045
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.818876261129042,
+      "learning_rate": 1.680358946412064e-06,
+      "loss": 3.454,
+      "step": 2046
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.343897423268437,
+      "learning_rate": 1.6731781025368422e-06,
+      "loss": 3.42,
+      "step": 2047
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.4004073563010193,
+      "learning_rate": 1.6660112337793256e-06,
+      "loss": 3.5292,
+      "step": 2048
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.6316960690058284,
+      "learning_rate": 1.6588583521678536e-06,
+      "loss": 3.571,
+      "step": 2049
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5002122945295335,
+      "learning_rate": 1.6517194697072903e-06,
+      "loss": 3.5268,
+      "step": 2050
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2795958724183882,
+      "learning_rate": 1.644594598378999e-06,
+      "loss": 3.3957,
+      "step": 2051
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.6550159003093974,
+      "learning_rate": 1.6374837501408403e-06,
+      "loss": 3.7017,
+      "step": 2052
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.488224452276052,
+      "learning_rate": 1.6303869369271264e-06,
+      "loss": 3.3991,
+      "step": 2053
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.357249271462046,
+      "learning_rate": 1.6233041706486253e-06,
+      "loss": 3.5832,
+      "step": 2054
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.266031195079106,
+      "learning_rate": 1.6162354631925203e-06,
+      "loss": 3.3827,
+      "step": 2055
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.6682563446568577,
+      "learning_rate": 1.609180826422404e-06,
+      "loss": 3.412,
+      "step": 2056
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.033994422891126,
+      "learning_rate": 1.602140272178253e-06,
+      "loss": 3.5786,
+      "step": 2057
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.9632696082851995,
+      "learning_rate": 1.5951138122764132e-06,
+      "loss": 3.2678,
+      "step": 2058
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.8403714532035282,
+      "learning_rate": 1.58810145850957e-06,
+      "loss": 3.4687,
+      "step": 2059
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.193247452585058,
+      "learning_rate": 1.5811032226467304e-06,
+      "loss": 3.6549,
+      "step": 2060
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2398053957952713,
+      "learning_rate": 1.5741191164332192e-06,
+      "loss": 3.2968,
+      "step": 2061
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.6586197558812015,
+      "learning_rate": 1.5671491515906355e-06,
+      "loss": 3.4885,
+      "step": 2062
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.61379309237948,
+      "learning_rate": 1.5601933398168523e-06,
+      "loss": 3.5474,
+      "step": 2063
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.558606413314661,
+      "learning_rate": 1.5532516927859853e-06,
+      "loss": 3.3917,
+      "step": 2064
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.051200426686327,
+      "learning_rate": 1.5463242221483742e-06,
+      "loss": 3.1919,
+      "step": 2065
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.502973571132373,
+      "learning_rate": 1.5394109395305757e-06,
+      "loss": 3.2573,
+      "step": 2066
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.683744743231853,
+      "learning_rate": 1.5325118565353237e-06,
+      "loss": 3.5408,
+      "step": 2067
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5358865096945524,
+      "learning_rate": 1.5256269847415283e-06,
+      "loss": 3.3673,
+      "step": 2068
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.8122495212068617,
+      "learning_rate": 1.5187563357042423e-06,
+      "loss": 3.436,
+      "step": 2069
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.406497985667074,
+      "learning_rate": 1.511899920954656e-06,
+      "loss": 3.6248,
+      "step": 2070
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.690777983132295,
+      "learning_rate": 1.5050577520000608e-06,
+      "loss": 3.3869,
+      "step": 2071
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.171874530283066,
+      "learning_rate": 1.498229840323847e-06,
+      "loss": 3.4189,
+      "step": 2072
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.492803640863754,
+      "learning_rate": 1.4914161973854714e-06,
+      "loss": 3.2746,
+      "step": 2073
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4706085858486273,
+      "learning_rate": 1.4846168346204425e-06,
+      "loss": 3.388,
+      "step": 2074
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.0469632561414475,
+      "learning_rate": 1.4778317634403082e-06,
+      "loss": 3.2526,
+      "step": 2075
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3027608809487523,
+      "learning_rate": 1.4710609952326239e-06,
+      "loss": 3.4284,
+      "step": 2076
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4405453383957854,
+      "learning_rate": 1.464304541360946e-06,
+      "loss": 3.4081,
+      "step": 2077
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.6129503858999756,
+      "learning_rate": 1.457562413164799e-06,
+      "loss": 3.6202,
+      "step": 2078
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.508733737407462,
+      "learning_rate": 1.4508346219596725e-06,
+      "loss": 3.3593,
+      "step": 2079
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5745164993431753,
+      "learning_rate": 1.4441211790369892e-06,
+      "loss": 3.4305,
+      "step": 2080
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.000893397603999,
+      "learning_rate": 1.4374220956640895e-06,
+      "loss": 3.3896,
+      "step": 2081
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.482970629880628,
+      "learning_rate": 1.4307373830842174e-06,
+      "loss": 3.2945,
+      "step": 2082
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.9659815944283183,
+      "learning_rate": 1.424067052516499e-06,
+      "loss": 3.3155,
+      "step": 2083
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3102034971632275,
+      "learning_rate": 1.4174111151559188e-06,
+      "loss": 3.3685,
+      "step": 2084
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1837029622658832,
+      "learning_rate": 1.4107695821733026e-06,
+      "loss": 3.2757,
+      "step": 2085
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.267635574821952,
+      "learning_rate": 1.4041424647153112e-06,
+      "loss": 3.3863,
+      "step": 2086
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4560576602866955,
+      "learning_rate": 1.3975297739043992e-06,
+      "loss": 3.435,
+      "step": 2087
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.989060712448598,
+      "learning_rate": 1.3909315208388185e-06,
+      "loss": 3.419,
+      "step": 2088
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.30029757789045,
+      "learning_rate": 1.3843477165925846e-06,
+      "loss": 3.5192,
+      "step": 2089
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4550946449825046,
+      "learning_rate": 1.3777783722154603e-06,
+      "loss": 3.5344,
+      "step": 2090
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.44034970355981,
+      "learning_rate": 1.3712234987329486e-06,
+      "loss": 3.6119,
+      "step": 2091
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3075509781316645,
+      "learning_rate": 1.3646831071462606e-06,
+      "loss": 3.4391,
+      "step": 2092
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4657392511161556,
+      "learning_rate": 1.3581572084323014e-06,
+      "loss": 3.4166,
+      "step": 2093
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1882512927254556,
+      "learning_rate": 1.3516458135436539e-06,
+      "loss": 3.3661,
+      "step": 2094
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.174656242832297,
+      "learning_rate": 1.3451489334085555e-06,
+      "loss": 3.2017,
+      "step": 2095
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.757745275768527,
+      "learning_rate": 1.3386665789308885e-06,
+      "loss": 3.4389,
+      "step": 2096
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4372378669298747,
+      "learning_rate": 1.3321987609901553e-06,
+      "loss": 3.4804,
+      "step": 2097
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.961745026644603,
+      "learning_rate": 1.325745490441458e-06,
+      "loss": 3.3249,
+      "step": 2098
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.9095905978056797,
+      "learning_rate": 1.3193067781154835e-06,
+      "loss": 3.3281,
+      "step": 2099
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.546582952524952,
+      "learning_rate": 1.3128826348184886e-06,
+      "loss": 3.1577,
+      "step": 2100
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.161408052005003,
+      "learning_rate": 1.3064730713322793e-06,
+      "loss": 3.3952,
+      "step": 2101
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4427792034132265,
+      "learning_rate": 1.3000780984141881e-06,
+      "loss": 3.4078,
+      "step": 2102
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.8874612168561593,
+      "learning_rate": 1.2936977267970597e-06,
+      "loss": 3.4509,
+      "step": 2103
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.018729625106304,
+      "learning_rate": 1.2873319671892337e-06,
+      "loss": 3.2307,
+      "step": 2104
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1584635393788365,
+      "learning_rate": 1.2809808302745298e-06,
+      "loss": 3.4466,
+      "step": 2105
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.5165182729856266,
+      "learning_rate": 1.2746443267122233e-06,
+      "loss": 3.4201,
+      "step": 2106
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.6148373495878054,
+      "learning_rate": 1.2683224671370286e-06,
+      "loss": 3.3471,
+      "step": 2107
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.520082081357915,
+      "learning_rate": 1.262015262159082e-06,
+      "loss": 3.2257,
+      "step": 2108
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.5479686710776432,
+      "learning_rate": 1.255722722363929e-06,
+      "loss": 3.2691,
+      "step": 2109
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.956232806174273,
+      "learning_rate": 1.249444858312502e-06,
+      "loss": 3.4043,
+      "step": 2110
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.855373187723047,
+      "learning_rate": 1.2431816805410968e-06,
+      "loss": 3.4917,
+      "step": 2111
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.1965735034591716,
+      "learning_rate": 1.2369331995613664e-06,
+      "loss": 3.3842,
+      "step": 2112
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.028949144756169,
+      "learning_rate": 1.2306994258602922e-06,
+      "loss": 3.3098,
+      "step": 2113
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.362284334661307,
+      "learning_rate": 1.2244803699001785e-06,
+      "loss": 3.2978,
+      "step": 2114
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.779755730336405,
+      "learning_rate": 1.218276042118629e-06,
+      "loss": 3.3034,
+      "step": 2115
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5439051586156336,
+      "learning_rate": 1.2120864529285203e-06,
+      "loss": 3.3475,
+      "step": 2116
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3052105946171926,
+      "learning_rate": 1.2059116127179993e-06,
+      "loss": 3.3798,
+      "step": 2117
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.218353846479608,
+      "learning_rate": 1.199751531850457e-06,
+      "loss": 3.2223,
+      "step": 2118
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.300669411980882,
+      "learning_rate": 1.1936062206645183e-06,
+      "loss": 3.4321,
+      "step": 2119
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5403763231563574,
+      "learning_rate": 1.1874756894740137e-06,
+      "loss": 3.2812,
+      "step": 2120
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2628393837539784,
+      "learning_rate": 1.1813599485679684e-06,
+      "loss": 3.5328,
+      "step": 2121
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.507796005006337,
+      "learning_rate": 1.1752590082105863e-06,
+      "loss": 3.2593,
+      "step": 2122
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.77010141147154,
+      "learning_rate": 1.1691728786412315e-06,
+      "loss": 3.4443,
+      "step": 2123
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2616859923362456,
+      "learning_rate": 1.1631015700744153e-06,
+      "loss": 3.5765,
+      "step": 2124
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.345595498088119,
+      "learning_rate": 1.1570450926997657e-06,
+      "loss": 3.4364,
+      "step": 2125
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.502868345524852,
+      "learning_rate": 1.1510034566820205e-06,
+      "loss": 3.2767,
+      "step": 2126
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.7828869424110594,
+      "learning_rate": 1.144976672161019e-06,
+      "loss": 3.379,
+      "step": 2127
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4442099961092265,
+      "learning_rate": 1.1389647492516598e-06,
+      "loss": 3.4932,
+      "step": 2128
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5130471167919657,
+      "learning_rate": 1.132967698043913e-06,
+      "loss": 3.1723,
+      "step": 2129
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.7375709878896384,
+      "learning_rate": 1.1269855286027798e-06,
+      "loss": 3.3686,
+      "step": 2130
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6802875325445674,
+      "learning_rate": 1.1210182509682854e-06,
+      "loss": 3.2742,
+      "step": 2131
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3259916423660383,
+      "learning_rate": 1.1150658751554667e-06,
+      "loss": 3.4796,
+      "step": 2132
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.234854692123814,
+      "learning_rate": 1.1091284111543499e-06,
+      "loss": 3.3168,
+      "step": 2133
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.632315148384429,
+      "learning_rate": 1.1032058689299297e-06,
+      "loss": 3.4312,
+      "step": 2134
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1620238621266283,
+      "learning_rate": 1.0972982584221592e-06,
+      "loss": 3.5919,
+      "step": 2135
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3617634363233004,
+      "learning_rate": 1.0914055895459353e-06,
+      "loss": 3.383,
+      "step": 2136
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.922553259067895,
+      "learning_rate": 1.08552787219107e-06,
+      "loss": 3.2105,
+      "step": 2137
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.37654662519368,
+      "learning_rate": 1.0796651162222916e-06,
+      "loss": 3.4856,
+      "step": 2138
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7966863859829796,
+      "learning_rate": 1.07381733147921e-06,
+      "loss": 3.501,
+      "step": 2139
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.840564360081184,
+      "learning_rate": 1.067984527776309e-06,
+      "loss": 3.4146,
+      "step": 2140
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4357237633421547,
+      "learning_rate": 1.062166714902938e-06,
+      "loss": 3.5836,
+      "step": 2141
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2728261173976994,
+      "learning_rate": 1.0563639026232742e-06,
+      "loss": 3.214,
+      "step": 2142
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.915665784391846,
+      "learning_rate": 1.0505761006763315e-06,
+      "loss": 3.342,
+      "step": 2143
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2202663116738095,
+      "learning_rate": 1.044803318775922e-06,
+      "loss": 3.3531,
+      "step": 2144
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.65403499358109,
+      "learning_rate": 1.0390455666106547e-06,
+      "loss": 3.4521,
+      "step": 2145
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.662658535408528,
+      "learning_rate": 1.0333028538439093e-06,
+      "loss": 3.3767,
+      "step": 2146
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3917540438518996,
+      "learning_rate": 1.027575190113832e-06,
+      "loss": 3.5931,
+      "step": 2147
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.397753021489659,
+      "learning_rate": 1.021862585033304e-06,
+      "loss": 3.3456,
+      "step": 2148
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.2907386279763573,
+      "learning_rate": 1.0161650481899344e-06,
+      "loss": 3.4717,
+      "step": 2149
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.316699462123251,
+      "learning_rate": 1.010482589146048e-06,
+      "loss": 3.4836,
+      "step": 2150
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.5992791259273185,
+      "learning_rate": 1.0048152174386584e-06,
+      "loss": 3.5198,
+      "step": 2151
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.897952562761025,
+      "learning_rate": 9.991629425794624e-07,
+      "loss": 3.5266,
+      "step": 2152
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.645336662391037,
+      "learning_rate": 9.935257740548143e-07,
+      "loss": 3.6137,
+      "step": 2153
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.079152642547377,
+      "learning_rate": 9.879037213257214e-07,
+      "loss": 3.3159,
+      "step": 2154
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.9483279351701177,
+      "learning_rate": 9.822967938278172e-07,
+      "loss": 3.2368,
+      "step": 2155
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4047908302860526,
+      "learning_rate": 9.767050009713476e-07,
+      "loss": 3.1575,
+      "step": 2156
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1730170838584604,
+      "learning_rate": 9.711283521411674e-07,
+      "loss": 3.5803,
+      "step": 2157
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.5454281828947,
+      "learning_rate": 9.655668566967026e-07,
+      "loss": 3.4175,
+      "step": 2158
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.6667236181520173,
+      "learning_rate": 9.600205239719584e-07,
+      "loss": 3.4001,
+      "step": 2159
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3294844387626874,
+      "learning_rate": 9.544893632754816e-07,
+      "loss": 3.3551,
+      "step": 2160
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.654141927545411,
+      "learning_rate": 9.489733838903648e-07,
+      "loss": 3.6266,
+      "step": 2161
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.8045320643118186,
+      "learning_rate": 9.434725950742119e-07,
+      "loss": 3.4512,
+      "step": 2162
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.6024403263194764,
+      "learning_rate": 9.379870060591434e-07,
+      "loss": 3.3326,
+      "step": 2163
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.386820474420187,
+      "learning_rate": 9.325166260517593e-07,
+      "loss": 3.428,
+      "step": 2164
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.651991887916642,
+      "learning_rate": 9.270614642331377e-07,
+      "loss": 3.6325,
+      "step": 2165
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3168626955308005,
+      "learning_rate": 9.216215297588182e-07,
+      "loss": 3.3719,
+      "step": 2166
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.808956179858258,
+      "learning_rate": 9.161968317587788e-07,
+      "loss": 3.1597,
+      "step": 2167
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5534169108074485,
+      "learning_rate": 9.107873793374322e-07,
+      "loss": 3.4365,
+      "step": 2168
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.713641504637213,
+      "learning_rate": 9.053931815735995e-07,
+      "loss": 3.3859,
+      "step": 2169
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.291244265002922,
+      "learning_rate": 9.000142475204965e-07,
+      "loss": 3.2342,
+      "step": 2170
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.6304683284911143,
+      "learning_rate": 8.946505862057286e-07,
+      "loss": 3.3497,
+      "step": 2171
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.036041707321972,
+      "learning_rate": 8.893022066312674e-07,
+      "loss": 3.297,
+      "step": 2172
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5995359541359333,
+      "learning_rate": 8.839691177734322e-07,
+      "loss": 3.3432,
+      "step": 2173
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5355815132434922,
+      "learning_rate": 8.786513285828835e-07,
+      "loss": 3.4101,
+      "step": 2174
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2445046978749734,
+      "learning_rate": 8.733488479845997e-07,
+      "loss": 3.3971,
+      "step": 2175
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3838590014717567,
+      "learning_rate": 8.680616848778711e-07,
+      "loss": 3.3631,
+      "step": 2176
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.808898429237898,
+      "learning_rate": 8.627898481362817e-07,
+      "loss": 3.5121,
+      "step": 2177
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.148682068047069,
+      "learning_rate": 8.575333466076863e-07,
+      "loss": 3.5631,
+      "step": 2178
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.349487356821742,
+      "learning_rate": 8.522921891142034e-07,
+      "loss": 3.3379,
+      "step": 2179
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.335460985598822,
+      "learning_rate": 8.470663844522053e-07,
+      "loss": 3.2859,
+      "step": 2180
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5146350091888077,
+      "learning_rate": 8.418559413922933e-07,
+      "loss": 3.4165,
+      "step": 2181
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0467244644452345,
+      "learning_rate": 8.366608686792854e-07,
+      "loss": 3.491,
+      "step": 2182
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.37795744884803,
+      "learning_rate": 8.31481175032206e-07,
+      "loss": 3.3244,
+      "step": 2183
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.8806468735539363,
+      "learning_rate": 8.263168691442624e-07,
+      "loss": 3.2478,
+      "step": 2184
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.399507866382731,
+      "learning_rate": 8.211679596828481e-07,
+      "loss": 3.493,
+      "step": 2185
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.6180760047276843,
+      "learning_rate": 8.160344552895061e-07,
+      "loss": 3.4451,
+      "step": 2186
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.975696190654819,
+      "learning_rate": 8.109163645799267e-07,
+      "loss": 3.5058,
+      "step": 2187
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.290590938027743,
+      "learning_rate": 8.058136961439333e-07,
+      "loss": 3.4879,
+      "step": 2188
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9777779041564334,
+      "learning_rate": 8.007264585454632e-07,
+      "loss": 3.3098,
+      "step": 2189
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.7754034334706503,
+      "learning_rate": 7.956546603225601e-07,
+      "loss": 3.3984,
+      "step": 2190
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.098136120772462,
+      "learning_rate": 7.905983099873504e-07,
+      "loss": 3.3151,
+      "step": 2191
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.840743601541193,
+      "learning_rate": 7.855574160260371e-07,
+      "loss": 3.492,
+      "step": 2192
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.6265418650700787,
+      "learning_rate": 7.805319868988759e-07,
+      "loss": 3.4926,
+      "step": 2193
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.85235771264861,
+      "learning_rate": 7.755220310401812e-07,
+      "loss": 3.4783,
+      "step": 2194
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5818307264323272,
+      "learning_rate": 7.705275568582848e-07,
+      "loss": 3.2884,
+      "step": 2195
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.270453873784988,
+      "learning_rate": 7.655485727355416e-07,
+      "loss": 3.2947,
+      "step": 2196
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.6452666439703183,
+      "learning_rate": 7.60585087028305e-07,
+      "loss": 3.1528,
+      "step": 2197
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9837151835317623,
+      "learning_rate": 7.556371080669222e-07,
+      "loss": 3.3995,
+      "step": 2198
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.505327878248056,
+      "learning_rate": 7.507046441557142e-07,
+      "loss": 3.3998,
+      "step": 2199
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1591895177001366,
+      "learning_rate": 7.457877035729588e-07,
+      "loss": 3.317,
+      "step": 2200
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5187197845554508,
+      "learning_rate": 7.408862945708839e-07,
+      "loss": 3.0716,
+      "step": 2201
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.4083125288162908,
+      "learning_rate": 7.360004253756459e-07,
+      "loss": 3.3461,
+      "step": 2202
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0850483605956924,
+      "learning_rate": 7.311301041873276e-07,
+      "loss": 3.432,
+      "step": 2203
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.8955991470826112,
+      "learning_rate": 7.262753391799127e-07,
+      "loss": 3.3223,
+      "step": 2204
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.171206654714713,
+      "learning_rate": 7.21436138501278e-07,
+      "loss": 3.5402,
+      "step": 2205
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3432430917547724,
+      "learning_rate": 7.166125102731735e-07,
+      "loss": 3.3176,
+      "step": 2206
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.694322018316045,
+      "learning_rate": 7.118044625912213e-07,
+      "loss": 3.5505,
+      "step": 2207
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.359500371733772,
+      "learning_rate": 7.070120035248906e-07,
+      "loss": 3.4174,
+      "step": 2208
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.010062042226531,
+      "learning_rate": 7.022351411174866e-07,
+      "loss": 3.0835,
+      "step": 2209
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5776919424054485,
+      "learning_rate": 6.974738833861383e-07,
+      "loss": 3.5002,
+      "step": 2210
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0916541703552607,
+      "learning_rate": 6.927282383217893e-07,
+      "loss": 3.3724,
+      "step": 2211
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.483857548891637,
+      "learning_rate": 6.879982138891717e-07,
+      "loss": 3.2362,
+      "step": 2212
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.783788104808641,
+      "learning_rate": 6.83283818026812e-07,
+      "loss": 3.4115,
+      "step": 2213
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.924748913054718,
+      "learning_rate": 6.785850586469989e-07,
+      "loss": 3.3712,
+      "step": 2214
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2175573933479846,
+      "learning_rate": 6.739019436357774e-07,
+      "loss": 3.4548,
+      "step": 2215
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.843047627078977,
+      "learning_rate": 6.692344808529427e-07,
+      "loss": 3.4471,
+      "step": 2216
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.746682602581423,
+      "learning_rate": 6.645826781320141e-07,
+      "loss": 3.2257,
+      "step": 2217
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.632430383751459,
+      "learning_rate": 6.599465432802332e-07,
+      "loss": 3.5574,
+      "step": 2218
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.591676174959998,
+      "learning_rate": 6.553260840785414e-07,
+      "loss": 3.2546,
+      "step": 2219
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.739713355457886,
+      "learning_rate": 6.507213082815745e-07,
+      "loss": 3.3196,
+      "step": 2220
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.5008219313024265,
+      "learning_rate": 6.461322236176438e-07,
+      "loss": 3.465,
+      "step": 2221
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.310600462482619,
+      "learning_rate": 6.415588377887305e-07,
+      "loss": 3.433,
+      "step": 2222
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.4841994321670917,
+      "learning_rate": 6.370011584704617e-07,
+      "loss": 3.3722,
+      "step": 2223
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6842707477566856,
+      "learning_rate": 6.324591933121072e-07,
+      "loss": 3.3919,
+      "step": 2224
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.8566346491530994,
+      "learning_rate": 6.279329499365649e-07,
+      "loss": 3.6028,
+      "step": 2225
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.3556102440025666,
+      "learning_rate": 6.234224359403407e-07,
+      "loss": 3.2679,
+      "step": 2226
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.248424488867721,
+      "learning_rate": 6.1892765889355e-07,
+      "loss": 3.3181,
+      "step": 2227
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.433953805154286,
+      "learning_rate": 6.144486263398886e-07,
+      "loss": 3.3922,
+      "step": 2228
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.606794192325152,
+      "learning_rate": 6.099853457966342e-07,
+      "loss": 3.4089,
+      "step": 2229
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2863483621832756,
+      "learning_rate": 6.055378247546217e-07,
+      "loss": 3.3874,
+      "step": 2230
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.5462763362220646,
+      "learning_rate": 6.01106070678239e-07,
+      "loss": 3.2217,
+      "step": 2231
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.18770247047478,
+      "learning_rate": 5.966900910054141e-07,
+      "loss": 3.414,
+      "step": 2232
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.3584454420282794,
+      "learning_rate": 5.922898931475973e-07,
+      "loss": 3.548,
+      "step": 2233
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2032427721789545,
+      "learning_rate": 5.879054844897536e-07,
+      "loss": 3.4794,
+      "step": 2234
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.172170956432926,
+      "learning_rate": 5.835368723903456e-07,
+      "loss": 3.478,
+      "step": 2235
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.8577441327169724,
+      "learning_rate": 5.791840641813295e-07,
+      "loss": 3.4523,
+      "step": 2236
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.5517894641461263,
+      "learning_rate": 5.748470671681328e-07,
+      "loss": 3.367,
+      "step": 2237
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3612955774508038,
+      "learning_rate": 5.705258886296494e-07,
+      "loss": 3.2858,
+      "step": 2238
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.620965017753635,
+      "learning_rate": 5.662205358182226e-07,
+      "loss": 3.5161,
+      "step": 2239
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.306799867032343,
+      "learning_rate": 5.619310159596358e-07,
+      "loss": 3.3049,
+      "step": 2240
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.892165227199948,
+      "learning_rate": 5.576573362531001e-07,
+      "loss": 3.2364,
+      "step": 2241
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.418545935942755,
+      "learning_rate": 5.533995038712403e-07,
+      "loss": 3.5193,
+      "step": 2242
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.605324441619377,
+      "learning_rate": 5.491575259600879e-07,
+      "loss": 3.2923,
+      "step": 2243
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2932202213739408,
+      "learning_rate": 5.449314096390601e-07,
+      "loss": 3.3651,
+      "step": 2244
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7819035993265215,
+      "learning_rate": 5.407211620009545e-07,
+      "loss": 3.1878,
+      "step": 2245
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.342693554550334,
+      "learning_rate": 5.365267901119398e-07,
+      "loss": 3.497,
+      "step": 2246
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.0434406319244784,
+      "learning_rate": 5.323483010115382e-07,
+      "loss": 3.4321,
+      "step": 2247
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.9546557643627587,
+      "learning_rate": 5.281857017126124e-07,
+      "loss": 3.4684,
+      "step": 2248
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.202307405158531,
+      "learning_rate": 5.240389992013606e-07,
+      "loss": 3.4182,
+      "step": 2249
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6682108167874126,
+      "learning_rate": 5.199082004372958e-07,
+      "loss": 3.5151,
+      "step": 2250
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.517041176431839,
+      "learning_rate": 5.157933123532466e-07,
+      "loss": 3.5793,
+      "step": 2251
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.060991767391451,
+      "learning_rate": 5.116943418553355e-07,
+      "loss": 3.1333,
+      "step": 2252
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.4617925682458828,
+      "learning_rate": 5.076112958229673e-07,
+      "loss": 3.4533,
+      "step": 2253
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.8325002574788374,
+      "learning_rate": 5.035441811088204e-07,
+      "loss": 3.3571,
+      "step": 2254
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.87591566298437,
+      "learning_rate": 4.994930045388414e-07,
+      "loss": 3.2436,
+      "step": 2255
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2740730847041695,
+      "learning_rate": 4.954577729122212e-07,
+      "loss": 3.363,
+      "step": 2256
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.9086146571113733,
+      "learning_rate": 4.914384930013927e-07,
+      "loss": 3.3876,
+      "step": 2257
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.9553725583407164,
+      "learning_rate": 4.874351715520154e-07,
+      "loss": 3.4666,
+      "step": 2258
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.409566328810979,
+      "learning_rate": 4.834478152829658e-07,
+      "loss": 3.3281,
+      "step": 2259
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6652751960125713,
+      "learning_rate": 4.794764308863242e-07,
+      "loss": 3.3374,
+      "step": 2260
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7376935408690137,
+      "learning_rate": 4.755210250273701e-07,
+      "loss": 3.3913,
+      "step": 2261
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.5403464753418685,
+      "learning_rate": 4.715816043445609e-07,
+      "loss": 3.3002,
+      "step": 2262
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.488337281386899,
+      "learning_rate": 4.676581754495235e-07,
+      "loss": 3.4845,
+      "step": 2263
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2385604249717814,
+      "learning_rate": 4.6375074492705173e-07,
+      "loss": 3.4416,
+      "step": 2264
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.8998958534577404,
+      "learning_rate": 4.5985931933508757e-07,
+      "loss": 3.2112,
+      "step": 2265
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4381788905534654,
+      "learning_rate": 4.559839052047066e-07,
+      "loss": 3.24,
+      "step": 2266
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9753181370969592,
+      "learning_rate": 4.521245090401172e-07,
+      "loss": 3.372,
+      "step": 2267
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.220562802271603,
+      "learning_rate": 4.482811373186402e-07,
+      "loss": 3.4169,
+      "step": 2268
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4575228301184597,
+      "learning_rate": 4.4445379649070587e-07,
+      "loss": 3.3371,
+      "step": 2269
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4887598974576735,
+      "learning_rate": 4.406424929798403e-07,
+      "loss": 3.2589,
+      "step": 2270
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.285211766969659,
+      "learning_rate": 4.368472331826479e-07,
+      "loss": 3.2891,
+      "step": 2271
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2239734105128406,
+      "learning_rate": 4.3306802346881116e-07,
+      "loss": 3.3581,
+      "step": 2272
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.484669516554814,
+      "learning_rate": 4.2930487018107425e-07,
+      "loss": 3.4378,
+      "step": 2273
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.487229626033901,
+      "learning_rate": 4.2555777963523506e-07,
+      "loss": 3.4464,
+      "step": 2274
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.1419345073346125,
+      "learning_rate": 4.218267581201296e-07,
+      "loss": 3.4621,
+      "step": 2275
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.219933497970078,
+      "learning_rate": 4.1811181189762684e-07,
+      "loss": 3.2655,
+      "step": 2276
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.594623370400675,
+      "learning_rate": 4.1441294720261373e-07,
+      "loss": 3.3844,
+      "step": 2277
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.7548340359274093,
+      "learning_rate": 4.107301702429922e-07,
+      "loss": 3.3651,
+      "step": 2278
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3666014345227353,
+      "learning_rate": 4.070634871996615e-07,
+      "loss": 3.4483,
+      "step": 2279
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2939439702731037,
+      "learning_rate": 4.034129042265067e-07,
+      "loss": 3.1616,
+      "step": 2280
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.7606232331352465,
+      "learning_rate": 3.9977842745039464e-07,
+      "loss": 3.2778,
+      "step": 2281
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4635150898245977,
+      "learning_rate": 3.961600629711615e-07,
+      "loss": 3.2342,
+      "step": 2282
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5295595342864443,
+      "learning_rate": 3.925578168616007e-07,
+      "loss": 3.319,
+      "step": 2283
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.792507271237906,
+      "learning_rate": 3.889716951674549e-07,
+      "loss": 3.3114,
+      "step": 2284
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.228923211595772,
+      "learning_rate": 3.8540170390740097e-07,
+      "loss": 3.5547,
+      "step": 2285
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.442052688731442,
+      "learning_rate": 3.8184784907304704e-07,
+      "loss": 3.457,
+      "step": 2286
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3026272696031045,
+      "learning_rate": 3.783101366289199e-07,
+      "loss": 3.3291,
+      "step": 2287
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.436857110291015,
+      "learning_rate": 3.747885725124523e-07,
+      "loss": 3.2545,
+      "step": 2288
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.569007423588782,
+      "learning_rate": 3.712831626339752e-07,
+      "loss": 3.5329,
+      "step": 2289
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.413741939457016,
+      "learning_rate": 3.67793912876705e-07,
+      "loss": 3.4219,
+      "step": 2290
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9608789183114315,
+      "learning_rate": 3.643208290967415e-07,
+      "loss": 3.4362,
+      "step": 2291
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.037967550302035,
+      "learning_rate": 3.608639171230488e-07,
+      "loss": 3.4947,
+      "step": 2292
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.963217206035839,
+      "learning_rate": 3.5742318275745147e-07,
+      "loss": 3.4222,
+      "step": 2293
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.404616702862916,
+      "learning_rate": 3.5399863177462024e-07,
+      "loss": 3.5048,
+      "step": 2294
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.5130147839058874,
+      "learning_rate": 3.5059026992206645e-07,
+      "loss": 3.5803,
+      "step": 2295
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.145610428854735,
+      "learning_rate": 3.4719810292013214e-07,
+      "loss": 3.3987,
+      "step": 2296
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1675486907915675,
+      "learning_rate": 3.438221364619776e-07,
+      "loss": 3.3274,
+      "step": 2297
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.5141499328252555,
+      "learning_rate": 3.404623762135728e-07,
+      "loss": 3.2946,
+      "step": 2298
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.5931881815304108,
+      "learning_rate": 3.371188278136883e-07,
+      "loss": 3.4986,
+      "step": 2299
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.237549412583975,
+      "learning_rate": 3.3379149687388866e-07,
+      "loss": 3.4054,
+      "step": 2300
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.4046964323023268,
+      "learning_rate": 3.3048038897851576e-07,
+      "loss": 3.3034,
+      "step": 2301
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.288139297934954,
+      "learning_rate": 3.271855096846899e-07,
+      "loss": 3.3689,
+      "step": 2302
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.512789471000361,
+      "learning_rate": 3.2390686452228983e-07,
+      "loss": 3.4543,
+      "step": 2303
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.843382156078518,
+      "learning_rate": 3.2064445899394723e-07,
+      "loss": 3.1137,
+      "step": 2304
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.054753404859892,
+      "learning_rate": 3.1739829857504235e-07,
+      "loss": 3.2675,
+      "step": 2305
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.567866951457264,
+      "learning_rate": 3.1416838871368925e-07,
+      "loss": 3.3668,
+      "step": 2306
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.588050596796032,
+      "learning_rate": 3.1095473483072733e-07,
+      "loss": 3.5633,
+      "step": 2307
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3237287020633057,
+      "learning_rate": 3.0775734231971443e-07,
+      "loss": 3.498,
+      "step": 2308
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.5006180356875145,
+      "learning_rate": 3.045762165469168e-07,
+      "loss": 3.351,
+      "step": 2309
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.783894516690254,
+      "learning_rate": 3.0141136285129825e-07,
+      "loss": 3.3718,
+      "step": 2310
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.8365546836380733,
+      "learning_rate": 2.982627865445109e-07,
+      "loss": 3.3811,
+      "step": 2311
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.6036011049940466,
+      "learning_rate": 2.951304929108956e-07,
+      "loss": 3.3192,
+      "step": 2312
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.117468304701286,
+      "learning_rate": 2.9201448720745706e-07,
+      "loss": 3.3399,
+      "step": 2313
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.360859819170253,
+      "learning_rate": 2.8891477466386987e-07,
+      "loss": 3.6341,
+      "step": 2314
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.7695749961891507,
+      "learning_rate": 2.8583136048245697e-07,
+      "loss": 3.2734,
+      "step": 2315
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6133036111449046,
+      "learning_rate": 2.827642498381955e-07,
+      "loss": 3.2883,
+      "step": 2316
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4475207107442603,
+      "learning_rate": 2.7971344787869114e-07,
+      "loss": 3.231,
+      "step": 2317
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.569809544449258,
+      "learning_rate": 2.76678959724187e-07,
+      "loss": 3.1757,
+      "step": 2318
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.310205615755986,
+      "learning_rate": 2.7366079046753925e-07,
+      "loss": 3.4045,
+      "step": 2319
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.043808757165406,
+      "learning_rate": 2.706589451742181e-07,
+      "loss": 3.341,
+      "step": 2320
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4378814841858576,
+      "learning_rate": 2.6767342888229907e-07,
+      "loss": 3.4172,
+      "step": 2321
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3665108545388387,
+      "learning_rate": 2.647042466024485e-07,
+      "loss": 3.4747,
+      "step": 2322
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0931178224328666,
+      "learning_rate": 2.617514033179236e-07,
+      "loss": 3.3271,
+      "step": 2323
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4596327927003623,
+      "learning_rate": 2.588149039845533e-07,
+      "loss": 3.3217,
+      "step": 2324
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.611293101153788,
+      "learning_rate": 2.5589475353073987e-07,
+      "loss": 3.5053,
+      "step": 2325
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4077089320299736,
+      "learning_rate": 2.5299095685744734e-07,
+      "loss": 3.5835,
+      "step": 2326
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.362519772264958,
+      "learning_rate": 2.5010351883819283e-07,
+      "loss": 3.441,
+      "step": 2327
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6639173507916767,
+      "learning_rate": 2.472324443190355e-07,
+      "loss": 3.5446,
+      "step": 2328
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.5669895297852423,
+      "learning_rate": 2.4437773811857304e-07,
+      "loss": 3.3679,
+      "step": 2329
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9325205560382717,
+      "learning_rate": 2.4153940502793185e-07,
+      "loss": 3.3625,
+      "step": 2330
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4262239886835975,
+      "learning_rate": 2.387174498107614e-07,
+      "loss": 3.1708,
+      "step": 2331
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.382901267977727,
+      "learning_rate": 2.359118772032176e-07,
+      "loss": 3.6955,
+      "step": 2332
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.8769220455422917,
+      "learning_rate": 2.3312269191396619e-07,
+      "loss": 3.4398,
+      "step": 2333
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2645249976733792,
+      "learning_rate": 2.30349898624167e-07,
+      "loss": 3.5903,
+      "step": 2334
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2437369619688132,
+      "learning_rate": 2.2759350198746978e-07,
+      "loss": 3.2821,
+      "step": 2335
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.585064031562485,
+      "learning_rate": 2.2485350663000727e-07,
+      "loss": 3.3699,
+      "step": 2336
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.529688315884326,
+      "learning_rate": 2.2212991715038324e-07,
+      "loss": 3.3273,
+      "step": 2337
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.5455952725918034,
+      "learning_rate": 2.1942273811966563e-07,
+      "loss": 3.3869,
+      "step": 2338
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7542843609631973,
+      "learning_rate": 2.1673197408138115e-07,
+      "loss": 3.5519,
+      "step": 2339
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7024764625310995,
+      "learning_rate": 2.1405762955151178e-07,
+      "loss": 3.3642,
+      "step": 2340
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.2576909980611903,
+      "learning_rate": 2.1139970901847607e-07,
+      "loss": 3.1324,
+      "step": 2341
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.291794267112212,
+      "learning_rate": 2.0875821694313014e-07,
+      "loss": 3.492,
+      "step": 2342
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.851766738786476,
+      "learning_rate": 2.0613315775875665e-07,
+      "loss": 3.4191,
+      "step": 2343
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3503396542169863,
+      "learning_rate": 2.0352453587105914e-07,
+      "loss": 3.4286,
+      "step": 2344
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.0832706870783757,
+      "learning_rate": 2.009323556581566e-07,
+      "loss": 3.4376,
+      "step": 2345
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.2176834985322738,
+      "learning_rate": 1.9835662147057012e-07,
+      "loss": 3.7042,
+      "step": 2346
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.508572081745489,
+      "learning_rate": 1.9579733763121943e-07,
+      "loss": 3.2487,
+      "step": 2347
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3815565196920945,
+      "learning_rate": 1.932545084354154e-07,
+      "loss": 3.3544,
+      "step": 2348
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.224405710797126,
+      "learning_rate": 1.9072813815085523e-07,
+      "loss": 3.3338,
+      "step": 2349
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.587434527550172,
+      "learning_rate": 1.8821823101760949e-07,
+      "loss": 3.2779,
+      "step": 2350
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.0105448519410976,
+      "learning_rate": 1.857247912481197e-07,
+      "loss": 3.5875,
+      "step": 2351
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3545775574783443,
+      "learning_rate": 1.8324782302718835e-07,
+      "loss": 3.6504,
+      "step": 2352
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.772545381786392,
+      "learning_rate": 1.8078733051197561e-07,
+      "loss": 3.2604,
+      "step": 2353
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.514296507565377,
+      "learning_rate": 1.7834331783198933e-07,
+      "loss": 3.4913,
+      "step": 2354
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.2130027457938604,
+      "learning_rate": 1.7591578908907724e-07,
+      "loss": 3.2063,
+      "step": 2355
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.473408546211247,
+      "learning_rate": 1.735047483574215e-07,
+      "loss": 3.4455,
+      "step": 2356
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.611810751851434,
+      "learning_rate": 1.7111019968353625e-07,
+      "loss": 3.4056,
+      "step": 2357
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.448733256723853,
+      "learning_rate": 1.687321470862524e-07,
+      "loss": 3.4959,
+      "step": 2358
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1270683895468423,
+      "learning_rate": 1.6637059455671623e-07,
+      "loss": 3.4274,
+      "step": 2359
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7016371313782153,
+      "learning_rate": 1.6402554605838173e-07,
+      "loss": 3.5068,
+      "step": 2360
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7952189658057955,
+      "learning_rate": 1.6169700552700284e-07,
+      "loss": 3.3875,
+      "step": 2361
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.6167241598160005,
+      "learning_rate": 1.5938497687062905e-07,
+      "loss": 3.4302,
+      "step": 2362
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.899382383808761,
+      "learning_rate": 1.5708946396959856e-07,
+      "loss": 3.112,
+      "step": 2363
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.61072616759305,
+      "learning_rate": 1.5481047067652744e-07,
+      "loss": 3.0877,
+      "step": 2364
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6528911414970993,
+      "learning_rate": 1.5254800081630828e-07,
+      "loss": 3.1509,
+      "step": 2365
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.7677655630715092,
+      "learning_rate": 1.5030205818610255e-07,
+      "loss": 3.286,
+      "step": 2366
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.0290147877844635,
+      "learning_rate": 1.4807264655533282e-07,
+      "loss": 3.4737,
+      "step": 2367
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.3276357529009775,
+      "learning_rate": 1.4585976966567826e-07,
+      "loss": 3.292,
+      "step": 2368
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.5897484848827896,
+      "learning_rate": 1.4366343123106697e-07,
+      "loss": 3.1764,
+      "step": 2369
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.931259962842155,
+      "learning_rate": 1.4148363493766803e-07,
+      "loss": 3.2836,
+      "step": 2370
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8715543592371913,
+      "learning_rate": 1.3932038444389063e-07,
+      "loss": 3.456,
+      "step": 2371
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6658305690245285,
+      "learning_rate": 1.3717368338037163e-07,
+      "loss": 3.5232,
+      "step": 2372
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.5552091059925846,
+      "learning_rate": 1.3504353534997682e-07,
+      "loss": 3.4265,
+      "step": 2373
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.060853993648709,
+      "learning_rate": 1.3292994392778535e-07,
+      "loss": 3.6187,
+      "step": 2374
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4127774044560337,
+      "learning_rate": 1.30832912661093e-07,
+      "loss": 3.4672,
+      "step": 2375
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.3298217946063042,
+      "learning_rate": 1.287524450694011e-07,
+      "loss": 3.3182,
+      "step": 2376
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.765897310420544,
+      "learning_rate": 1.2668854464441104e-07,
+      "loss": 3.5001,
+      "step": 2377
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6799767075818983,
+      "learning_rate": 1.246412148500198e-07,
+      "loss": 3.3275,
+      "step": 2378
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.271650758753963,
+      "learning_rate": 1.2261045912231318e-07,
+      "loss": 3.4898,
+      "step": 2379
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4900191809890475,
+      "learning_rate": 1.2059628086956044e-07,
+      "loss": 3.3488,
+      "step": 2380
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.432667422028782,
+      "learning_rate": 1.1859868347220749e-07,
+      "loss": 3.3059,
+      "step": 2381
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1095534136728507,
+      "learning_rate": 1.1661767028287363e-07,
+      "loss": 3.3028,
+      "step": 2382
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2825282377319884,
+      "learning_rate": 1.1465324462634375e-07,
+      "loss": 3.2163,
+      "step": 2383
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.348291541570649,
+      "learning_rate": 1.1270540979956501e-07,
+      "loss": 3.4357,
+      "step": 2384
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.8133420709248758,
+      "learning_rate": 1.1077416907163573e-07,
+      "loss": 3.3492,
+      "step": 2385
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.416116463150174,
+      "learning_rate": 1.0885952568380764e-07,
+      "loss": 3.2941,
+      "step": 2386
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2336397899674707,
+      "learning_rate": 1.0696148284947694e-07,
+      "loss": 3.2592,
+      "step": 2387
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.290423744841851,
+      "learning_rate": 1.0508004375417546e-07,
+      "loss": 3.4197,
+      "step": 2388
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4718862466754437,
+      "learning_rate": 1.032152115555718e-07,
+      "loss": 3.1517,
+      "step": 2389
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4536745523599475,
+      "learning_rate": 1.0136698938346012e-07,
+      "loss": 3.3263,
+      "step": 2390
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4892773617678046,
+      "learning_rate": 9.953538033975918e-08,
+      "loss": 3.548,
+      "step": 2391
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.274808744119864,
+      "learning_rate": 9.772038749850665e-08,
+      "loss": 3.4609,
+      "step": 2392
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.1066954963040736,
+      "learning_rate": 9.59220139058492e-08,
+      "loss": 3.5136,
+      "step": 2393
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.819428406700386,
+      "learning_rate": 9.414026258004583e-08,
+      "loss": 3.3541,
+      "step": 2394
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5996884033585483,
+      "learning_rate": 9.237513651145224e-08,
+      "loss": 3.5706,
+      "step": 2395
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.280278075970925,
+      "learning_rate": 9.062663866252541e-08,
+      "loss": 3.422,
+      "step": 2396
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6535216978039577,
+      "learning_rate": 8.889477196781571e-08,
+      "loss": 3.2123,
+      "step": 2397
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6186688736381183,
+      "learning_rate": 8.717953933395695e-08,
+      "loss": 3.5313,
+      "step": 2398
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4552877616197284,
+      "learning_rate": 8.548094363966974e-08,
+      "loss": 3.2186,
+      "step": 2399
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.614807339334542,
+      "learning_rate": 8.379898773574924e-08,
+      "loss": 3.4121,
+      "step": 2400
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.560606455212878,
+      "learning_rate": 8.213367444506515e-08,
+      "loss": 3.4588,
+      "step": 2401
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.894001650217336,
+      "learning_rate": 8.04850065625551e-08,
+      "loss": 3.3264,
+      "step": 2402
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.7294219687904273,
+      "learning_rate": 7.885298685522235e-08,
+      "loss": 3.268,
+      "step": 2403
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3727386747414427,
+      "learning_rate": 7.723761806212371e-08,
+      "loss": 3.5093,
+      "step": 2404
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6980302225240456,
+      "learning_rate": 7.563890289437825e-08,
+      "loss": 3.2537,
+      "step": 2405
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3141221393707134,
+      "learning_rate": 7.405684403514635e-08,
+      "loss": 3.3421,
+      "step": 2406
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.943541524063672,
+      "learning_rate": 7.24914441396396e-08,
+      "loss": 3.3136,
+      "step": 2407
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.621160257306377,
+      "learning_rate": 7.094270583510976e-08,
+      "loss": 3.4558,
+      "step": 2408
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.611032570898534,
+      "learning_rate": 6.941063172084094e-08,
+      "loss": 3.5023,
+      "step": 2409
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.069405550611829,
+      "learning_rate": 6.78952243681541e-08,
+      "loss": 3.5456,
+      "step": 2410
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.2993895937672195,
+      "learning_rate": 6.639648632039697e-08,
+      "loss": 3.5449,
+      "step": 2411
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.673634537765846,
+      "learning_rate": 6.491442009293858e-08,
+      "loss": 3.4421,
+      "step": 2412
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.7532147515910252,
+      "learning_rate": 6.344902817316811e-08,
+      "loss": 3.3273,
+      "step": 2413
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.7110501713964834,
+      "learning_rate": 6.200031302049048e-08,
+      "loss": 3.2886,
+      "step": 2414
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.438293669662373,
+      "learning_rate": 6.056827706632185e-08,
+      "loss": 3.5594,
+      "step": 2415
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.508072088402561,
+      "learning_rate": 5.915292271408524e-08,
+      "loss": 3.3602,
+      "step": 2416
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.5684197432611096,
+      "learning_rate": 5.7754252339204955e-08,
+      "loss": 3.414,
+      "step": 2417
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.319063366225016,
+      "learning_rate": 5.637226828910436e-08,
+      "loss": 3.3698,
+      "step": 2418
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.8336668797631637,
+      "learning_rate": 5.5006972883204776e-08,
+      "loss": 3.3801,
+      "step": 2419
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.6058057493735585,
+      "learning_rate": 5.365836841291439e-08,
+      "loss": 3.4924,
+      "step": 2420
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.4600163464627554,
+      "learning_rate": 5.232645714163265e-08,
+      "loss": 3.4741,
+      "step": 2421
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.480906233508525,
+      "learning_rate": 5.1011241304738115e-08,
+      "loss": 3.3203,
+      "step": 2422
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3298471961537905,
+      "learning_rate": 4.9712723109590636e-08,
+      "loss": 3.4202,
+      "step": 2423
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.337131802937764,
+      "learning_rate": 4.843090473552914e-08,
+      "loss": 3.4015,
+      "step": 2424
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2352102980270843,
+      "learning_rate": 4.716578833386054e-08,
+      "loss": 3.2217,
+      "step": 2425
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.436029228983509,
+      "learning_rate": 4.5917376027861945e-08,
+      "loss": 3.2638,
+      "step": 2426
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.8479008982912006,
+      "learning_rate": 4.468566991277512e-08,
+      "loss": 3.2157,
+      "step": 2427
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.500510734627635,
+      "learning_rate": 4.347067205580424e-08,
+      "loss": 3.4505,
+      "step": 2428
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.057314345010156,
+      "learning_rate": 4.2272384496112597e-08,
+      "loss": 3.4177,
+      "step": 2429
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.5742843904617185,
+      "learning_rate": 4.109080924481479e-08,
+      "loss": 3.2047,
+      "step": 2430
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.980513447383297,
+      "learning_rate": 3.9925948284980086e-08,
+      "loss": 3.3224,
+      "step": 2431
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.523660677303886,
+      "learning_rate": 3.877780357162353e-08,
+      "loss": 3.1595,
+      "step": 2432
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.7222719612091675,
+      "learning_rate": 3.764637703170593e-08,
+      "loss": 3.4171,
+      "step": 2433
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.294507134265397,
+      "learning_rate": 3.653167056413054e-08,
+      "loss": 3.4806,
+      "step": 2434
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.4418301658508055,
+      "learning_rate": 3.543368603973529e-08,
+      "loss": 3.2865,
+      "step": 2435
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3616588149744073,
+      "learning_rate": 3.435242530129723e-08,
+      "loss": 3.4426,
+      "step": 2436
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.9010966248921273,
+      "learning_rate": 3.3287890163523626e-08,
+      "loss": 3.5075,
+      "step": 2437
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.116521733307449,
+      "learning_rate": 3.224008241304977e-08,
+      "loss": 3.2828,
+      "step": 2438
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.285127613209865,
+      "learning_rate": 3.120900380844116e-08,
+      "loss": 3.3661,
+      "step": 2439
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.5071884384835306,
+      "learning_rate": 3.019465608018024e-08,
+      "loss": 3.5639,
+      "step": 2440
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1870037421829043,
+      "learning_rate": 2.9197040930674102e-08,
+      "loss": 3.1974,
+      "step": 2441
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.509012040156903,
+      "learning_rate": 2.8216160034244544e-08,
+      "loss": 3.3545,
+      "step": 2442
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.70637322606787,
+      "learning_rate": 2.7252015037131373e-08,
+      "loss": 3.3309,
+      "step": 2443
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.8258205503720206,
+      "learning_rate": 2.6304607557481322e-08,
+      "loss": 3.3279,
+      "step": 2444
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3015147026301306,
+      "learning_rate": 2.537393918535358e-08,
+      "loss": 3.3331,
+      "step": 2445
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.558951915848977,
+      "learning_rate": 2.4460011482713153e-08,
+      "loss": 3.2497,
+      "step": 2446
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.989173866915556,
+      "learning_rate": 2.3562825983427517e-08,
+      "loss": 3.5516,
+      "step": 2447
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.639158558560462,
+      "learning_rate": 2.2682384193266625e-08,
+      "loss": 3.4531,
+      "step": 2448
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.331617858251365,
+      "learning_rate": 2.1818687589896248e-08,
+      "loss": 3.4148,
+      "step": 2449
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.5011814560248915,
+      "learning_rate": 2.0971737622883515e-08,
+      "loss": 3.2685,
+      "step": 2450
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.4780309027220215,
+      "learning_rate": 2.01415357136836e-08,
+      "loss": 3.2917,
+      "step": 2451
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.40132084644993,
+      "learning_rate": 1.93280832556475e-08,
+      "loss": 3.3335,
+      "step": 2452
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.359641935311238,
+      "learning_rate": 1.8531381614013133e-08,
+      "loss": 3.4068,
+      "step": 2453
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1694399725333087,
+      "learning_rate": 1.7751432125903133e-08,
+      "loss": 3.3822,
+      "step": 2454
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.587399184824128,
+      "learning_rate": 1.698823610032929e-08,
+      "loss": 3.4473,
+      "step": 2455
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.302820237000699,
+      "learning_rate": 1.6241794818180333e-08,
+      "loss": 3.2511,
+      "step": 2456
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.546048636973119,
+      "learning_rate": 1.5512109532229703e-08,
+      "loss": 3.3763,
+      "step": 2457
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.6815940730038124,
+      "learning_rate": 1.4799181467125557e-08,
+      "loss": 3.5656,
+      "step": 2458
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.23628539901506,
+      "learning_rate": 1.4103011819395218e-08,
+      "loss": 3.652,
+      "step": 2459
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.047804798596946,
+      "learning_rate": 1.3423601757436289e-08,
+      "loss": 3.2655,
+      "step": 2460
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.528437058663026,
+      "learning_rate": 1.276095242151998e-08,
+      "loss": 3.1583,
+      "step": 2461
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.4972696524475113,
+      "learning_rate": 1.2115064923787778e-08,
+      "loss": 3.344,
+      "step": 2462
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9816851292985413,
+      "learning_rate": 1.1485940348249235e-08,
+      "loss": 3.2459,
+      "step": 2463
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5442232507559304,
+      "learning_rate": 1.087357975078085e-08,
+      "loss": 3.352,
+      "step": 2464
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4841578124819246,
+      "learning_rate": 1.0277984159122734e-08,
+      "loss": 3.423,
+      "step": 2465
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5554413670728247,
+      "learning_rate": 9.699154572877511e-09,
+      "loss": 3.3301,
+      "step": 2466
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.6871360033669887,
+      "learning_rate": 9.137091963510314e-09,
+      "loss": 3.2715,
+      "step": 2467
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.165684460893977,
+      "learning_rate": 8.591797274344338e-09,
+      "loss": 3.4842,
+      "step": 2468
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9791606895583,
+      "learning_rate": 8.063271420563068e-09,
+      "loss": 3.4481,
+      "step": 2469
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9882821622236264,
+      "learning_rate": 7.551515289203615e-09,
+      "loss": 3.5619,
+      "step": 2470
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.7516938084672558,
+      "learning_rate": 7.056529739158935e-09,
+      "loss": 3.488,
+      "step": 2471
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.80255477519852,
+      "learning_rate": 6.5783156011778315e-09,
+      "loss": 3.3546,
+      "step": 2472
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.145429658546217,
+      "learning_rate": 6.116873677858293e-09,
+      "loss": 3.3526,
+      "step": 2473
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.6512344972109707,
+      "learning_rate": 5.6722047436497115e-09,
+      "loss": 3.3866,
+      "step": 2474
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.9255532464199225,
+      "learning_rate": 5.2443095448506674e-09,
+      "loss": 3.4207,
+      "step": 2475
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.500865341292617,
+      "learning_rate": 4.833188799610033e-09,
+      "loss": 3.5687,
+      "step": 2476
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4307967357678644,
+      "learning_rate": 4.438843197922538e-09,
+      "loss": 3.3044,
+      "step": 2477
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3415100972592366,
+      "learning_rate": 4.061273401627653e-09,
+      "loss": 3.3692,
+      "step": 2478
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3171125128650845,
+      "learning_rate": 3.7004800444095933e-09,
+      "loss": 3.3511,
+      "step": 2479
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.563985957009853,
+      "learning_rate": 3.3564637317984318e-09,
+      "loss": 3.2359,
+      "step": 2480
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2396809393878967,
+      "learning_rate": 3.0292250411645406e-09,
+      "loss": 3.5031,
+      "step": 2481
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.487495476298278,
+      "learning_rate": 2.7187645217219283e-09,
+      "loss": 3.2537,
+      "step": 2482
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.462183993392574,
+      "learning_rate": 2.4250826945226847e-09,
+      "loss": 3.4758,
+      "step": 2483
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.45474989584317,
+      "learning_rate": 2.148180052462534e-09,
+      "loss": 3.4833,
+      "step": 2484
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5726004000973304,
+      "learning_rate": 1.888057060274173e-09,
+      "loss": 3.227,
+      "step": 2485
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2144747119002104,
+      "learning_rate": 1.6447141545272717e-09,
+      "loss": 3.168,
+      "step": 2486
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.608690469923541,
+      "learning_rate": 1.4181517436306913e-09,
+      "loss": 3.2924,
+      "step": 2487
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.523006771698765,
+      "learning_rate": 1.2083702078302673e-09,
+      "loss": 3.4771,
+      "step": 2488
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.393916464258931,
+      "learning_rate": 1.0153698992088069e-09,
+      "loss": 3.4871,
+      "step": 2489
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.5257507681162927,
+      "learning_rate": 8.391511416816489e-10,
+      "loss": 3.3328,
+      "step": 2490
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.5827728588499856,
+      "learning_rate": 6.797142310022154e-10,
+      "loss": 3.0827,
+      "step": 2491
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.412401860436348,
+      "learning_rate": 5.370594347575697e-10,
+      "loss": 3.4849,
+      "step": 2492
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.3902248151074974,
+      "learning_rate": 4.1118699236841753e-10,
+      "loss": 3.478,
+      "step": 2493
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.1063910939480706,
+      "learning_rate": 3.0209711509132657e-10,
+      "loss": 3.4109,
+      "step": 2494
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.5735345416169753,
+      "learning_rate": 2.0978998601206558e-10,
+      "loss": 3.3083,
+      "step": 2495
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.4576775876063928,
+      "learning_rate": 1.342657600544861e-10,
+      "loss": 3.51,
+      "step": 2496
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.835368532804796,
+      "learning_rate": 7.552456397053042e-11,
+      "loss": 3.3847,
+      "step": 2497
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.9813514920634563,
+      "learning_rate": 3.3566496349113355e-11,
+      "loss": 3.5756,
+      "step": 2498
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.2882708884674345,
+      "learning_rate": 8.39162760835066e-12,
+      "loss": 3.4625,
+      "step": 2499
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.528204116206639,
+      "learning_rate": 0.0,
+      "loss": 3.2533,
+      "step": 2500
+    },
+    {
+      "epoch": 1.0,
+      "step": 2500,
+      "total_flos": 5.835554469802148e+17,
+      "train_loss": 4.1818853558540345,
+      "train_runtime": 17904.7495,
+      "train_samples_per_second": 4.468,
+      "train_steps_per_second": 0.14
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2500,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 240,
+  "total_flos": 5.835554469802148e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}